Chapter 3

Generation-Recombination Processes:
Carrier Lifetime

The competition of generation processes with recombination processes directly affects the
performance of photodetectors, setting up a steady-state concentration of carriers in
semiconductors subjected to the thermal and optical excitation and determining the
kinetics of photogenerated signals. The decay of optically generated carriers due to
recombination reduces the quantum efficiency of the device and photoelectric gain.
Worse, the statistical nature of the generation-recombination processes results in
fluctuation of carrier concentration, causing noise that limits the performance of
photodetectors. The effects of fluctuating recombination rates can be avoided in many
cases by arranging for the recombination process to take place in a region of the device
where it has little effect due to low photoelectric gain − for example, at the contacts in
sweep-out photoconductors or in the neutral regions of the diodes. The generation
process, with its associated fluctuation, however, cannot be avoided. The narrow-gap
semiconductors necessary for infrared photodetectors, especially those operated in the
LWIR range at near-room temperature, exhibit very high inherent thermal generation
rates. To achieve high performance, thermal generation must be suppressed to the lowest
possible level. For practical purposes, the ideal situation is when the thermal generation is
reduced to the level below that of the optical generation. One important example is the
background-limited photodetector (BLIP), in which optical generation due to the
background radiation exceeds the thermal component. The requirements for thermal
generation rate can be highly reduced in heterodyne systems, when optical excitation by
the local oscillator can dominate the generation even for high thermal generation.
Various mechanisms, both fundamental and technology-related, might be responsible
for the generation and recombination rates. The resultant net generation rate G – R due to
various processes is the summation of the bimolecular factor nopo–np multiplied by terms
gk, which are dependent on the generation-recombination processes1:

G−R = ∑ g (n p
k
k o o − np) , (3.1)

where p = po + ∆p and n = no + ∆n are the nonequilibrium carrier densities.

For small deviations from equilibrium, the formulation approximates the summation
of familiar small-signal net generation rates:

83

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 84 Chapter 3

∆n
G−R = ∑− τk k
, (3.2)

where τk are the carrier recombination lifetimes for various mechanisms.

Under the condition of a small excess carrier concentration, the lifetimes associated
with these processes can be treated as independent. Therefore, the resultant recombination
lifetime can be calculated as

R −G
∑τ
1 1
= = . (3.3)
τ k k ∆n

The lifetime of the photoexcited electrons and holes are among the most important
parameters of a semiconductor detector material, since they govern the magnitude and
frequency response of the detector signal.2,3 Electron-hole pairs can recombine in two
ways. First, an electron can drop directly from the conduction band into an unoccupied
state in the valence band; this is known as a direct band-to-band process. Second, an
electron initially makes a transition to an energy level lying deep in the bandgap and
subsequently captures a hole from the valence band; this is known as an indirect process.
In the process of electron-hole pair recombination, an energy equal to the difference
between electron and hole energies is released. This energy can be emitted as a photon, in
which case the recombination is said to be radiative. Alternatively, the energy can be
dissipated to the lattice in the form of phonons. A third possibility is that the energy can
be imparted as a kinetic energy to a third mobile carrier: this process is called the Auger
process. Note that both the phonon and Auger processes are nonradiative.
This chapter considers only the most important generation-recombination mechanisms
in narrow-gap semiconductors. As discussed by many authors,4–8 in narrow-gap
semiconductors there are three essential generation-recombination mechanisms:
Shockley-Read, radiative and Auger (see Fig. 3.1).

Fig. 3.1. Semiconductor generation mechanisms: (a) Shockley-Read-Hall recombination,

(b) radiative recombination, and (c) Auger recombination.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 85

3.1 Shockley-Read mechanism

The Shockley-Read mechanism is not an intrinsic process, as it occurs via levels in the
forbidden energy gap. It may be reduced by lowering concentrations of native defects and
foreign impurities. This can be achieved by low-temperature growth and by progress in
purification of materials. Though a considerable research effort is still necessary, the
Shockley-Read process does not represent a fundamental limit to photodetector
performance; it can be reduced with progress toward purer and higher-quality material,
which is viable in the case of narrow-gap semiconductors.
In the Shockley-Read mechanism, generation and recombination occurs via energy
levels introduced by impurities or lattice defects into the forbidden energy gap [Fig.
3.1(a)]. As recombination centers, they trap electrons and holes; as generation centers,
they successively emit them. The rates of generation and recombination depend on the
individual nature of the center and its predominant occupation state of charge carriers, as
well as on the local densities of those carriers in the bands of the semiconductor. In
general, the transition rates via energy levels for electrons and holes are quite different,
thus causing different lifetimes.
The statistical theory for the generation-recombination processes via intermediate
centers was developed first by Shockley and Read,9 and Hall.10 This type of
recombination is often called Shockley-Read-Hall (SRH) recombination.
The single-level recombination can be described by four processes: electron capture,
electron emission, hole capture and hole emission. The analysis of the various
possibilities, taking into account different capture cross sections for electron and hole σn
and σp , respectively, yields the trap-determined net generation rate

σ n σ p v th (n o po − np )N t
(G − R )SR = , (3.4)
σ n (n + n1 ) + σ p (p + p1 )

where n1 = Ncexp[−Et/kT] and p1 = Nvexp[–(Eg–Et)/kT] are the carrier concentrations for

the Fermi level coincident with trap level; Nc and Nv are the density of states for
conduction and valence band, Et is the trap below the conduction band, Nt is the density
of the traps, and vth is the carrier thermal velocity equal to (8kT/πm*)1/2.
The recombination rate approaches a maximum as the energy level of the
recombination center approaches midgap. Thus, the most effective recombination centers
are those located near the middle of the bandgap. In this case, the net generation rate due
to SRH states is given by

−1
 −E t  
( G − R )SR= ( n i2 − n o po )  τpo  n + n i exp
Et  
 + τno  p + n i exp  .
kT  
(3.5)
  kT  

Equation (3.4) can be simplified under conditions of low carrier injection (i.e., ∆n <<
no or ∆p << po) into material in which the ionization energy of the centers is
comparatively deep relative to kT. In n–type material, for example, when trapping can be
neglected (∆n = ∆p) in sufficiently low temperature (no > n1 > po),

(R − G )SR ≈ σ p N t v th ∆n . (3.6)

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 86 Chapter 3

From the definition of the minority carrier lifetime as the ratio of the excess carrier
density ∆n to the net recombination rate, the trap-limited lifetime in doped n-type material

(
τ po ≈ σ p v th N t )−1 (3.7)

is independent of the concentration of majority carriers and is determined by the lifetime

of minority carriers (holes), which is intuitively clear because in this case all
recombination centers are occupied by electrons. Similarly, for p–type material,

τ no ≈ (σ n v th N t )−1 . (3.8)

Taking into account Eqs. (3.4), (3.7) and (3.8) we can obtain

τpo (n o + n1 ) + τno (po + p1 )

τSR = . (3.9)
n o + po

Let us analyze the dependence of the lifetime on temperature for an n–type

semiconductor in which the trap level is located above midgap (Et > Ei) and assuming that
σn and σp are constant. Three distinct ranges of temperature are to be noticed (Fig. 3.2):

• region I (at low temperature), in which Et < EF < Ec. Then no > po, n1, p1 and the
steady-state carrier lifetime is determined by Eq. (3.6), valid even throughout
freezeout;
• region II, in which Ei < EF < Et. Then no > po, n1 and n1 > po; the carrier lifetime

Nc  E 
τSR = exp − t 
σp v th N t n o  kT 

increases exponentially with temperature. The transition between region I and

region II occurs at the point for which no = n1 (when EF = Et);
• region III (the intrinsic range), in which no = po = ni, n1 > ni, p1 < ni, and the
lifetime

 Eg Et 
τ SR ∝ exp − 
 2kT kT 
 

decreases exponentially with temperature.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 87

Fig. 3.2. Changes in the position of the Fermi level (a) and lifetime of minority carriers (b)
as a function of temperature.

Because of the above considerations, the lifetime of the Shockley-Read mechanism is

a function of four parameters, namely Nt, Et, τno, and τpo. If there is more than a single
recombination level, each level will have its own parameters. The expression valid for a
finite number of both recombination centers and excess carriers can be found in
Blakemore;1 it is considerably more complex than the above expressions. For the case of
a multiple number of recombination levels, the Shockley-Read lifetime becomes

−1
τSR = ∑τ −1
i ,

where τi is the lifetime of the i-th recombination level.

The electron-hole recombination via intermediate centers is particularly pronounced in
indirect bandgap materials because the competing direct band-to-band recombination
process is so low.

3.2 Radiative mechanism

The radiative process takes place [Fig. 3.1(b)] when a free electron and a hole recombine,
emitting the excess energy in the form of a photon. This type of recombination
mechanism (the so-called intrinsic mechanism) is determined by the electronic band
structure of the semiconductor. The radiative recombination is of central importance to
light-emitting-diode operation and may take place in any semiconductor having a direct

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 88 Chapter 3

bandgap.11 The theory of radiative recombination has been developed by van Roosbroeck
and Shockley,12 reviewed by Hall,13 and reexamined by Humpreys.14,15
The recombination rate for bands with spherical symmetry was shown by van
Roosbroeck and Shockley to depend on the absorption coefficient through the equation

8π ε ( E ) α ( E ) E dE
∞ 2

3 3 ∫
GR = , (3.10)
h c 0 exp ( E / kT ) − 1

where α(E) is the absorption coefficient, ε(E) is the relative dielectric constant, h is the
Planck’s constant, and c is the speed of light. Usually dispersion in the dielectric constant
ε(E) is ignored, and the high-frequency value ε∞ is used. For a small departure from
equilibrium, the radiative lifetime is given by

n i2
τR = , (3.11)
G R (n o + p o )

where ni is the intrinsic carrier concentration. The steady-state radiative lifetimes of

excess electrons (τe) and holes (τh) are equal if charge trapping and storage effects at
defects are neglected. For an intrinsic semiconductor

ni
τiR = , (3.12)
2G R

the lifetime achieves the longest possible value.

The value of GR can be obtained either by numerical integration of Eq. (3.10),
employing measured values of α(E), or by analytical techniques using theoretical
expressions for the absorption coefficient in simple semiconductors. The expression for
the dependence of α on E as analyzed by Bardeen et al.16 is

3/ 2
22/3 mq 2  m*e m*h 
1/ 2
 m m   E-E g 
α ( E ) = 1/ 2 2   1+ * + *   2 
(3.13)
3ε∞   m ( m*e +m*h )   m e m h   mc 
 

where m is the free electron mass, and Eg is the material bandgap.

Based on this expression and assuming that Eg > kT and dispersion in the dielectric
constant is ignored, one obtains

3/ 2
   
( )
3/ 2
G R = 5.8 × 10 −13
n i2 ε1∞/ 2 
m  1 + m  300  E 2g + 3kTE g + 3.75k 2 T 2 ,
 * *   *  T
 
 me + mh   m e 
(3.14)

where Eg is in eV. For largebandgap materials, Eg > kT, and therefore only the E g2 term is
needed, as given by Hall.13

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 89

The above calculations suffer from the objection that they are only strictly valid for
parabolic bands and nondegenerate statistics. However, a quantum electrodynamic
calculation for a Kane-type band structure employing degenerate statistics, and
considering only transitions between the conduction band and heavy hole band, yields a
value in good agreement with the simple analysis above.17 Hall’s equation is used
routinely to calculate the radiative lifetime even in HgCdTe.7,18–23 However, the other two
terms in Eq. (3.14), 3kT and 3.75(kT)2, should not be disregarded. For example, for
Hg0.8Cd0.2Te at 77K, GR is increased by more than 20%.20
The minority carrier lifetimes for radiative recombination according to Eq. (3.11) are
given by

2τiR n i2
τhR = = (3.15)
( n o / n i ) + ( n i / n o ) G R ( n o +n i2 / n o )
and
2τiR n i2
τeR = = . (3.16)
( po / n i ) + ( n i / po ) G R ( po +n i2 / po )

From the above equations we can see that the lifetimes follow immediately from a
knowledge of ni and τiR (or GR). The temperature dependence of the radiative lifetime for
intrinsic material will vary as exp(Eg/2kT).
It should be noted that Humpreys recently 14,15 reexamined the van Roosbroeck and
Shockley theory. He indicated that most photons emitted in photodetectors as a result of
radiative decay are immediately readsorbed, so that the observed radiative lifetime is only
a measure of how well photons can escape from the body of a detector. Due to
reabsorption, the radiative lifetime is highly extended, in dependence of the
semiconductor geometry. This implies that further improvements in the state of the art for
infrared detectors are possible, particularly for materials where radiative recombination is
more important, i.e., for short-wavelength low-temperature devices. Achieving this
performance in practice would require exceptionally high material quality, since radiative
efficiencies of better than 95% would be required. The fundamental limit to the detector
performance is reached when the radiation generation is caused principally by the photons
from the scene, while the secondary recombination-generation processes are practically
noiseless. The frequently met calculations of performance based on an incorrectly derived
recombination lifetime by the Van Roosbroeck and Shockley theory leads to a radiatively
limited detectivity that is unduly pessimistic. Until now, the Humpreys theory has not
been included in many papers concerning recombination mechanisms in narrow-gap
semiconductors, so in this book some of the results were obtained assuming the van
Roosbroeck and Shockley theory.

3.3 Auger mechanism

The Auger effect is the second intrinsic recombination mechanism. It is an important
mechanism in determining the performance of light-emitting devices and infrared
detectors made from narrow-gap semiconductors.
Auger recombination can be considered to be the inverse process of impact ionization
where a three-particle Coulomb interaction occurs. Figure 3.1(c) shows one of the Auger

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 90 Chapter 3

transitions for a simple two-band model of a semiconductor. In this mechanism, an

electron in the conduction band recombines with a hole in the valence band. The energy
lost by the electron is transferred to a second conduction-band electron, which is pushed
to an excited state. The second electron does not stay in the excited state, but returns to its
initial state by phonon emission via the electron-lattice interaction.
The Auger process may also take part in a Shockley-Read process, i.e., in a transition
from a band to a recombination center level in the energy gap. However, the band-to-band
Auger effect represents the most efficient intrinsic recombination process in
semiconductors with a small energy gap. In narrow-gap semiconductors a Shockley-Read
process often governs the lifetime of the excess carrier, but this behavior seems to become
less important as the quality of the material improves.

3.3.1 Band-to-band Auger mechanisms

The band-to-band Auger effects are classified in several processes according to related
bands. Beattie24 has determined ten types of photonless Auger recombination mechanisms
that are possible in a semiconductor with a single conduction band and heavy- and light-
hole valence bands. In Fig. 3.3 we show the three most important mechanisms in the case
of this type of band structure. They have the smallest threshold (ET ≈ Eg) and the largest
combined density of states. The CHCC recombination mechanism (also labeled Auger 1)
involves two electrons and a heavy hole and is dominant in n-type material. The CHLH
process (labeled Auger 7) is dominant in p–type material if the spin split-off band can be
ignored. For materials such as InSb and HgCdTe where the spin split-off energy ∆ is
much larger than the bandgap energy Eg, the probability of the Auger transition through
the conduction band/heavy-hole band/spin split-off band mechanism (called CHSH or
Auger S hereafter) may be negligibly small in comparison with that of the CHLH Auger
transition. It was pointed out by Takeshima25 that the spin split-off band plays a far more
important role than the light-hole band for the direct-bandgap materials, especially when
the bandgap energy Eg approaches the spin-orbit splitting ∆ (as in the case of InAs and
InAsSb).

Fig. 3.3. The three band-to-band Auger recombination processes. Arrows indicate
electron transitions; •, occupied state;  , unoccupied state.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 91

The basic theory of the Auger recombination in semiconductors with InSb-like band
structure was developed by Beattie and Landsberg in 1959.26 They considered the CHCC
(Auger 1) mechanism in order to explain the experimental results of Tauc.27 In n–type
materials in which m∗e / m∗h << 1, the process involving a hole as the third carrier can be
ignored and only electron-electron processes are important. The electron-electron
recombination rate in a nondegenerate semiconductor is proportional to n2p and is given
by1

R A =G A1 n 2 p .

The generation (or impact ionization) rate on the other hand is given by

G A =G A1n i2 n .

Hence, the net recombination rate is given by

R A − G A =G A1n ( np − n i2 ) , (3.17)

where1

8(2π)5 / 2 q 4 mo  (me / mo ) F1F2  1  kT 

∗ 2 3/ 2
  1 + 2µ  E g 
G A1 = 3 2 
× exp −    . (3.18)
 2 
h (4πεoεs ) (1 + µ) (1 + 2µ) n i  E g    1 + µ  kT 
1/ 2
 

In the last equation, µ is the ratio of the conduction to the heavy-hole valence-band
effective mass, εs is the relative static dielectric constant, and F1 and F2 are the overlap
integrals of the periodic part of the electron wave functions. The overlap integrals F1 and
F2 cause the biggest uncertainty in the Auger 1 lifetime. The value of F1F2 was found
by Petersen to depend on k, and was determined by him to within a factor of 3.29 In
practice, it is taken as a constant equal to anywhere between 0.1 and 0.3, leading to
changes in the lifetime by an order of magnitude.
The carrier lifetime determined by the Auger process τA = ∆n (R A − G A ) is

n i4
τA = . (3.19)
G eeA ( n o +p o +∆n ) np o

If, in addition, the hole-hole recombination process take place, then

n i4
τA = . (3.20)
( n o +po +∆n ) ( G eeA npo +G Ahh pn o )

To determine the Auger lifetime, it is first necessary to calculate the net transition rate.
The calculation of GA for different types of Auger recombinations involving both the
heavy- and light-hole valence bands has been treated extensively in the

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 92 Chapter 3

literature,1,3,5,26,28–33 and the early results are summarized by Blakemore.1 Different

explicit expressions for the transition rate of the band-to-band recombination can be
found in the literature. The pioneering work of Beattie and Landsberg26 on the Auger 1
process used parabolic band and nondegenerate statistics. The Auger 7 process was first
considered by Beattie and Smith.28 The Auger 7 and Auger S processes have been
discussed by Takeshima,25 using nonparabolic bands and nondegenerate statistics.
Sugimura34,35 has considered the Auger S process, using degenerate statistics,
nonparabolic bands and screening. In the calculations of the transition rate of Auger
processes, different estimates of overlap integrals have been suggested.28–37 These
integrals were evaluated for the threshold momentum values of particles participating in
the Auger recombination. In general, these integrals are dependent on the wave functions,
which are not sufficiently understood, and methods of approximation must be used. Burt
et al.38 have found that the effective-mass sum-role method overestimates the modules
squared of the overlap integrals by at least an order of magnitude. The general solution
for the Auger lifetime for a fully degenerate semiconductor has not been completely
solved. Some other remaining theoretical problems are, for example, screening in
semiconductors having extremely high static dielectric constants, and assuming of a
quasi-equilibrium at low temperatures.4 Also, more accurate values of the band
parameters for narrow-gap semiconductors considered in this book are necessary in order
to obtain a quantitative understanding of the recombination mechanisms.
For the narrow-gap semiconductors considered here, µ = m*e / m*h is << 1, and then in
the extrinsic region, the Auger 1 lifetime is given by1

2n i2 τiA1
τA1 = , (3.21)
( n o +po +∆n ) ( n o +∆n ) +β ( po +∆n )

where ∆n is the excess carrier density that we neglected, since we usually operate in the
small signal condition; the β term equal to

µ1/2 (1 + 2µ )   1 − µ  Eg 
β= exp −   
2+µ   1 + µ  kT 

refers to hole-hole collisions and is also usually neglected. Assuming the above
simplifications, we can obtain the following expression for the Auger 1 recombination
lifetime in nondegenerated material:

2τiA1
τA1 = , (3.22)
1+ ( n o / n i )
2

where τiA1 is the Auger lifetime for intrinsic material given by

 1+2µ  E g 
3.8 × 10−18 εs2 (1+µ ) (1+2µ ) exp 
1/ 2
 
τiA1 =  1+µ  kT  in s. (3.23)
(m / m ) F1F2 ( kT / E )
* 2 3/ 2
e g

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 93

Casselman and Petersen,32,33 following the work of Beattie and Smith,28 argued that
the Auger 7 process is the dominating Auger mechanism for p–type semiconductors with
an InSb bandlike structure. This process involves the recombination between a
conduction electron and a heavy hole, while at the same time an electron is excited from
the light-hole band to an empty state in the heavy-hole band, absorbing the excess energy
and momentum (Fig. 3.3 − CHLH process). Casselman and Petersen’s computations
indicate that the Auger 7 lifetime in the p–type extrinsic region should be on the same
order of magnitude as that in the n–type material with the same carrier concentration.
Similar to the Auger 1 process, the lifetime associated with the Auger 7 process is

2n i2 τiA 7
τA 7 = , (3.24)
(n o + po + ∆n )(po + ∆n )

where τ iA 7 is the Auger 7 lifetime for intrinsic material

16q 4 α (ch) α (co) m∗co3 / 2 η37 / 2 I7 ( η7 )

7 7

(τ )
i −1
= , (3.25)
1+ ( m∗ / m∗ )3 / 2 
A7
π1/ 2  7 ε 2  m∗l ( k 7 )  a 72 ( 2+µ o )
2 3/ 2

 lo h


where η7 = ET/kBT, ET is the Auger 7 threshold energy; m∗co and m∗lo are the conduction-
and light-hole band effective masses at k = 0; μo is the ratio of m∗co / m∗h ; m∗h is the heavy-
hole mass; m∗l (k 7 ) is the light-hole effective mass at threshold;
(
a 7 = m∗l (k 7 ) / m∗co + 1 + µ o / 2 ) ; ε is the dielectric constant; and  is Planck’s constant.
−1

The factors α (ch7 ) and α (lh7 )

are overlap integral coefficients proportional to the oscillator
strength of a transition between the initial and final states in the conduction band c,
heavy-hole valence band h, and light hole valence band l. The factor I7(η7) is given by28

∞
 dm∗l  1/ 2
I7 ( η7 ) = ∫ e −η7 y a 2 ( y ) m∗l 5 / 2 ( y )  m∗y ( y ) +y  y ( y − 1) dy ,
2 (3.26)
1  dy 

where y = E/ET and E is energy.

The above calculations of carrier lifetime for the Auger 7 process are rather
complicated. Casselman and Petersen,32 making some simplifying analytical
approximations, have found that

 5 
m∗e ( E T ) 1 − 
τ  4ηT  .
i
γ= A7
≈2 (3.27)
τi
 3 
A1
m∗eo 1 − 
 2 ηT 

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 94 Chapter 3

For the Auger 7 process, the threshold energy ET ≈ Eg. From Kane’s nonparabolic
approximation m∗e ( E T ) / m∗eo ≈ 3, and, for example, for Hg1–xCdxTe over the range 0.16 ≤
x ≤ 0.40 and 50 K ≤ T ≤ 300 K, 3 ≤ γ ≤ 6.32
The combined lifetime of the two Auger processes (Auger 1 and Auger 7) is given by

2n i2 γτiA1
τA = . (3.28)
(n o + po + ∆n )[(n o + ∆n )γ + (po + ∆n )]
In the limit of a small signal condition

2τiA 7
τA 7 = (3.29)
1 + (po / n i )2

and

2 γ τiA 7
τA = . (3.30)
1 + γ + γ (n o / n i )2 + (po / n i )2

The Auger lifetime decreases very slowly with decreasing temperatures in the
extrinsic range, since it is proportional to T3/2exp[µEg/(1+µ)kT].1 At very low
temperatures, τA starts increasing due to the exponential term. However, since µ << 1 in
narrow-gap semiconductors, this effect should be noticed only at very low temperatures
below 10 K. Gerhardts et al.39 measured an exponential increase in τA in n–type HgCdTe,
but at temperatures much higher than 10 K. The exponential increase in the Auger
lifetime is steeper than in the radiative process, since τA is inversely proportional to the
square of the hole concentration rather than to its first power.
As was previously mentioned, in the case of semiconductors for which the bandgap
energy Eg approaches the spin-orbit splitting ∆, the Auger S process—also called the
CHSH process (see Fig. 3.3)—plays an important role in extrinsic p–type material. The
Auger S recombination lifetime is expressed in a way analogous to that of the Auger 7
recombination

2τiAS
τAS = , (3.31)
1 + (po / n o )2

where

ni
τiAS = . (3.32)
2G AS

According to the Gelmont theory:36

(a) for ∆ > Eg

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 95

27 4 2 q  ms ( ∆ − E g ) exp  − ( ∆ − E g ) / kT 
4 3 ∗5 / 2

G AS = π n i p
ε 2 m∗h3 m∗e3 / 2 kT∆ 2 ( E g +∆ )
5 , (3.33)

where m∗s is the split-off band effective mass;

(b) for ∆ < Eg

G AS =
(
18π q 4m∗e 3 E g + ∆ 2 ) n i2 γ 2p[I1 (γ ) − I 2 (γ ) / 2] . (3.34)
ε 2
m∗h m∗s 2E5g

In this equation,

m∗s E g2 Eg − ∆
γ =2
m h ( E g +∆ )( 3E g − 2∆ ) kT
∗

and the functions I1(γ) and I2(γ) are rather complicated integrals given in Ref. 36.
It can be shown that the net effective Auger generation rate is equal to

n i2 − np  n o po po 
( G=
− R )A  i + i + i . (3.35)
2n i  τA1 τA7 τAS 
2

From the above considerations we see that the band-to-band Auger process is a
complex function of the carrier effective mass, the density of state distribution, the
bandgap energy and a steep function of the majority carrier concentration [see Eqs. (3.22)
and (3.31)]. The band-to-band Auger processes are characterized by a strong temperature
dependence arising from a characteristic “threshold energy” for these processes
conditioned by the momentum and energy conservation of the three particles.

3.3.2 Other Auger recombination mechanisms

The conservation laws give rise to a threshold energy for each of the Auger processes. In
the absence of momentum conservation, there is no threshold energy. Thus, the strong
temperature dependence will not appear. This may be achieved either by phonon
assistance, which makes the process a second order, or if one of the four-particle states is
not a plane wave. The second case arises in Auger recombination via traps.
The phonon-assisted process and the trap-Auger processes do not exhibit such a
strong temperature dependence. Also, the recombination rate of phonon-assisted
processes does not decrease as rapidly with decreasing bandgap as it does with the band-
to-band process. The last two characteristics are used to identify the type of Auger
process.
In wide-gap semiconductors, especially in indirect-gap materials like silicon, the
band-to-band Auger recombination mechanisms do not usually appear to be important.
The upper bound of the minority carrier lifetime in highly doped semiconductors is
defined by the phonon-assisted and trap-assisted Auger recombination mechanisms. It

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 96 Chapter 3

should be stated that minority carrier recombination in heavily doped semiconductors is a

poorly understood subject.
The phonon-assisted Auger process in direct-gap semiconductors was first discussed
by Eagles.40 Lochman41,42 calculated the phonon-assisted Auger rate in GaAs, using
parabolic bands and nondegenerate statistics, and Huldt43 calculated it in Ge.
Auger recombination via traps was first discussed by Bess.44 Sinha and DiDomenico45
included the effect of screening on the Auger recombination of trapped electrons and
bound excitons. Bonch-Bruevich and Glasko46 and Landsberg et al.47–49 have also
discussed Auger recombination involving traps. All possible recombination mechanisms
are considered by Landsberg in his comprehensive monograph.3
The two relevant trap Auger effects are shown in Fig. 3.4. In process (a), a trapped
electron makes a transition to the valence band by giving up its energy to an electron in
the conduction band. The rate of this process can be written as Ctapnnt, where nt
represents the concentration of trapped electrons and Cta is the trap coefficient. This
process becomes important only in a very heavily doped n-type semiconductor with a
high concentration of traps below the Fermi level. Process (b) shows recombination
involving two holes and one electron. Here, the electron makes a transition to an
unoccupied trap level in the bandgap and gives its energy to a hole in the valence band.
Like the above, the rate of process can be expressed as Ctbpnpt, and this process may
acquire importance in heavily doped p-type semiconductors with a large density of traps
located above the Fermi level.

Fig. 3.4. Schematic illustrations of two possible trap-assisted Auger processes.

Landsberg and Robbins49,50 have estimated the trap Auger coefficients of processes (a)
and (b) by assuming a hydrogenlike wave function and a rigid-parabolic model. Their
analysis gives
2
 m  E 5t / 2
C ta =2.23 × 10−26 
2
∗ 
F cm6/sec (3.36)
 εs m e  (E − Et )
3/ 2 4
g E g

and
2
−26  m  1 2
C tb =2.23 × 10  ∗ 
F . (3.37)
 εs m e  E t
3

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 97

In these equations, m∗e is the density of state effective mass, Et is the trap energy measured
from the conduction band edge and |F|2 accounts for wave-function overlap and—
although not precisely known—is estimated to be of the order of 0.1.
The effective carrier lifetime is given by

1 1 1 1
= + + . (3.38)
τef τR τA τSR

3.4 Carrier lifetime in InSb and InAs

By the late 1950s and early 1960s, several papers had been published on the
photoconductive lifetime of InSb and InAs. Since then, the quality of single-crystal
growth and fabrication technology has improved immensely. More recently, Pines and
Stafsudd51 have presented the photoconductive lifetime of a high-quality state-of-the-art
n–type InSb ingot. The measured photoconductive lifetime is the parallel combination of
the SR photoconductive lifetime τ,pc = (μeτe+μhτh)/(μe + μh) (when the lifetimes of the
electron and hole are not equal), the Auger lifetime and the radiative lifetime, and is given
by τpc = (1/τA1 + 1/τR + 1/ τ,pc )–1.
Figure 3.5 shows the photoconductive lifetime as a function of inverse temperature for
various impurity electron concentrations and for density of the recombination centers
8×1013 cm–3. The SR photoconductive lifetime is predominant in the extrinsic
temperatures up to 250 K, but above that temperature, the Auger 1 lifetime becomes the
limiting lifetime. At a higher temperature (T > 200 K), the parallel combination of the SR
photoconductive lifetime and the Auger lifetime is presented by the broken line.
Measured photoconductive lifetime data from previous work by Hollis et al.52 are also
presented in Fig. 3.5. At temperatures below 130 K, the electron and hole lifetimes obey
the expression for the SRH recombination process

4.4 × 108
τh = . (3.39)
Nd

This lifetime dependence on doping has been observed in a number of different samples
fabricated at different laboratories. The electron and hole lifetimes are equal, because in
n–type material there is a single set of recombination centers. To explain the
recombination process in the extrinsic temperature range, Pines and Stafsudd51 have
presented a new two-step recombination model in which a dipole moment interaction
potential and a Coulomb interaction potential are involved.
Zitter et al.53 have measured photoelectromagnetic and photoconductive lifetimes
from 77 to 300 K in p–type InSb with a concentration ranging from less than 1015 cm–3 to
1018 cm–3. Figure 3.6 illustrates the dependence of PC carrier lifetime versus reciprocal
temperature for different p–type InSb samples. In p–type InSb at 77 K, hole and electron
lifetimes are far different, τe being less than 10–9 s, and τh being equal to about 109/po,
where po is the equilibrium hole concentration. This behavior is attributed to the presence
of donorlike recombination centers situated at about 0.05 eV above the valence band and
possibly a second deeper level.52,54 At 200 K, τe ≈ τh ≈ τPEM ≈ τpc, which indicates that

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 98 Chapter 3

Fig. 3.5. N–type InSb photoconductive and Auger lifetime for various impurity
concentrations (after Ref. 51).

Fig. 3.6. P–type InSb carrier lifetime vs reciprocal temperature for various impurity
concentrations (after Ref. 53).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 99

carrier trapping is negligible. In intrinsic InSb above 250 K, recombination is limited by

the interband Auger process. It is expected on theoretical grounds (see Section 3.3.1) that
the Auger 7 process should contribute to recombination in p–type InSb. However, no
evidence for an Auger 7 mechanism has been found.55
InSb detectors have usually been restricted to operating temperatures ranging from 77
K to 90 K. This limitation is mainly due to the small generation lifetimes that result from
impurities and defects in the crystal lattice of available bulk material. It has been found
that these impurities and defects can be greatly reduced by growing epitaxial layers of
InSb on InSb wafers by liquid phase epitaxy.56
Figure 3.7 shows the dependence of the particular carrier lifetimes (determined by
various recombination channels) in InAs on the relative doping of n– and p–type material
at 200 and 300 K. As the temperature increases, the Auger recombination comes to the
fore. In contrast to InSb, for p–type InAs the influence of the CHSH process is apparent.
The contribution of the CHSH process increases with decreasing temperature. On
comparison of the theoretical curves with the experimental data estimated from
photoconductivity measurements of n–type materials, it may be seen that satisfactory
consistency has been achieved. However, oddly enough, on comparing experimental
measurements with a theoretical model, Dalal et al.,57 Wider and Collins58 considered the
influence of the CHCC to be decisive. It is not obvious. In Fig. 3.7 we see that for the
higher carrier concentrations, the influence of the CHCC process and the radiative
recombination is similar, and at 77 K the contribution of the radiative recombination is
even greater.59

Fig. 3.7. Dependence of carrier lifetime on normalized doping concentrations for InAs at
200 and 300 K. The experimental values are taken from photoconductivity measurements
according to Ref. 57 (×) and Ref. 58 (o) (after Ref. 59).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 100 Chapter 3

Rogalski and Orman59 calculated the carrier lifetime in InAs1–xSbx for radiative and
Auger recombination for the temperature range 77–300 K and the composition range 0 ≤
x ≤ 1. The transition rate of the Auger process was calculated according to Gelmont
theory. In the low temperature range, the carrier lifetime is determined by radiative
recombination. At higher temperatures, the CHCC process is dominant in n–type
InAs1–xSbx, but in p–type material, for the composition range 0 ≤ x ≤ 0.3, competition
takes place between CHLH and CHSH processes; for the composition range x > 0.3, the
CHLH process is dominant. Various recombination channels in InAs0.35Sb0.65, which is
potentially capable of operating at the longest cutoff wavelengths, with Eg = 0.1 eV at
room temperature are also discussed in Ref. 59.

3.5 Carrier lifetime in InxGa1–xAs

Radiative and nonradiative lifetimes in InxGa1–xAs (InGaAs) ternary alloys have been
determined both theoretically and experimentally. Most of data are given for
In0.53Ga0.47As. However, considerable discrepancy between different papers exist.60
Recently, Ahrenkiel et al.61 have fitted to the experimentally measured minority carrier
lifetime in n–type and p–type In0.53Ga0.47As the following approximation:

−1
=τ  2.11× 104 + 1.43 × 10−10 N + 8.1× 10−29 N 2  (s), (3.40)

where N is the doping density in cm–3. The corresponding lifetimes ranged from about 20
μs to 10 ps over doping ranges from 1014 to 2×1019 cm–3. It appears however, that the
experimental values given by Gallant and Zemel,62 as well as by Trommer and
Hoffmann,63 for the undoped material are larger than described by Eq. (3.40). To receive
good agreement between experimental data and theoretical calculations for the undoped
material, Rogalski and Ciupa65 have assumed that the radiative coefficient GR is one order
of magnitude lower than described by Eq. (3.14). As was previously suggested (see
Section 3.2), the reason for that can be controlled by photon reabsorption and an increase
of the apparent lifetime over the theoretical radiative value derived by the Van
Roosbroeck and Shockley theory. An estimation of influence of photon recycling given
by Grein et al.66 indicates that the radiative lifetime enhancement can be greater than an
order of magnitude.
Figure 3.8 shows a composite of the lifetime data against carrier concentrations at
room temperature for n–type and p–type In0.53Ga0.47As. In both types of materials in the
doping range below 1016 cm–3, the dominant effect of radiative process is revealed. In the
range of higher carrier concentration, the Auger 1 process is decisive in n–type material,
instead of the Auger S process in p–type material. In Fig. 3.8, experimental values from
the literature are also shown. For n–type In0.53Ga0.47As, the agreement of experimental
data with theoretical calculations in the range of doping concentrations below 1018 cm–3 is
good. In the range of very high carrier concentrations—above 1018 cm–3—the discrepancy
between theory and experimental values increases mainly due to fact that the carrier
lifetimes were calculated assuming the Boltzmann statistic. In the case of degenerate
statistics, the doping dependence of the carrier lifetimes is weaker. We have the same
situation in the case of p–type In0.53Ga0.47As. To our knowledge, there is a lack of
experimental data in the low doping range of p–type material.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 101

(a)

(b)

Fig. 3.8. Dependence of the carrier lifetime on doping concentration for n–type (a) and p–
type (b) In0.53Ga0.47As ternary alloy at room temperature. References of experimental data
are marked inside the figures. The solid curve is theoretically calculated, the dashed
curve is an approximation to experimental data given by Ahrenkiel et al.61 (after Ref. 65).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 102 Chapter 3

In general, the effective carrier lifetimes, in n–type as well as p–type InxGa1–xAs

ternary alloys with composition x > 0.53, are comparable in the range of low doping
concentrations. In the wide wavelength range, the influence of radiative process is
decisive.

3.6 Carrier lifetime in Hg1–xCdxTe

In HgCdTe alloy systems, all three recombination mechanisms take place. During the last
20 years, various papers have been published4–8,17–23,29,32,33,39,67–81 that indicate the Auger
effect as the important intrinsic recombination mechanism in HgCdTe alloys. It is well
known that the Auger 1 process is an important recombination mechanism in n–type
Hg1–xCdxTe, particularly for x around 0.2 and at higher temperatures.4,7,17,23 In Fig. 3.9(a),
experimental results from Kinch et al.17 are compared with theoretical data of the intrinsic
Auger 1 carrier lifetime τiA1 and intrinsic radiative carrier lifetime τiR . There is excellent
agreement between the experimental and numerical data. Even in the extrinsic range at
temperatures below 140 K, the measured lifetime seems to be governed by the Auger 1
effect, where the following relation should hold: τA1 ≈ 2τiA1 (n i / n o ) 2 [see Eq. (3.22)].
Bartoli et al.67 have measured the concentration dependence of carrier lifetime in
Hg0.20Cd0.80Te, and for n > 1.5×1015 cm–3 they found that τ ∼ n–2, which supports Auger
recombination as the limiting mechanism. Also, the measurement of the transit decay of
large excess carrier densities has given a statistical criterion for the dominant Auger
process.66 The above results have been obtained for purer ingots. In more compensated
material (no < 4×1014 cm–3, µ77 < 1.5×105 cm2/Vs), SRH-limited behavior has been
observed. Also, Pratt et al.19 explained a variation of minority carrier lifetimes with
temperature in n–type Hg0.30Cd0.70Te by a combination of SRH recombination through
deep traps with band-to-band Auger and radiative recombinations. The traps have an
activation energy of 0.3– to 0.4Eg above the top of the valence band,69,70 and their density
can be as high as 1016 cm–3.69 For example, Fig. 3.9(b) shows the lifetime as a function of
temperature for an In-doped HgCdTe layer grown by MOCVD with x ≈ 0.227 and Nd ≈
6.6×1014 cm–3. The lifetime was reported to follow an Auger 1 dependence for
temperatures above 200 K and to deviate from Auger dependence for temperature below
200 K. The deviation was attributed to an SRH recombination center with energy 0.8Eg
above the valence band. It should be noticed that dopants have different influences on
carrier lifetime in HgCdTe. Doping with the donors Si, Cl, and In was shown21 to have
little effect on lifetime, but Al doping did reduce lifetime, as did the addition of a high
concentration of inactive Fe.82
More recently, Lopes et al.7 reviews the status of minority carrier lifetime in n–type
and p–type HgCdTe materials grown by bulk and epitaxial techniques. Figure 3.10 shows
the measured lifetime as a function of carrier concentration for n–type LWIR HgCdTe
and MWIR HgCdTe, respectively, at temperatures near 80 K. In the case of LWIR
material, an upper limit in the lifetime is observed. At this upper limit, the lifetime
increases with decreasing carrier concentration. At carrier concentrations less than 1015
cm–3, the lifetimes exhibit a broad range of values for all material types. At carrier
concentration near 5×1014 cm–3, the lifetime is observed to range from approximately 0.4
μs to 8 μs. The range in lifetimes for fixed carrier concentration may be due to different
compositions, measurement temperatures, surface preparations, epitaxial-substrate surface

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 103

(a)

(b)

Fig. 3.9. Theoretical and experimental lifetime data vs temperature for (a) n–type
Hg0.795Cd0.205Te (n77K = 1.7×1014 cm–3, µ77K = 1.42×105 cm2/Vs), solid lines represent theoretical
values for radiative and Auger 1 recombination respectively (after Ref. 17); (b) for an
MOCVD sample with x = 0.227 and Nd = 6.6×1014 cm–3 (after Ref. 76).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 104 Chapter 3

(a)

(b)

Fig. 3.10. (a) Lifetime (near 80 K) as a function of carrier concentration for n–type LWIR
HgCdTe. The full curve is the calculated Auger and radiative lifetime using F1F2 = 0.15 and
x = 0.225. (b) Lifetime (near 80 K) as a function of carrier concentration for n–type MWIR
HgCdTe. The full curve is the calculated Auger and radiative lifetime using F1F2 = 0.15 and
x = 0.3 (after Ref. 7).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 105

recombination velocities, material inhomogeneities or material defects. In addition, high

background fluxes can affect the steady-state carrier concentration and therefore affect the
lifetime. The biggest uncertainly in the theoretical Auger 1 lifetime calculation arises
from the overlap integrals F1 and F2 [see Eq. (3.22)]. To obtain a good fit to the
experimental data a value for the overlap integral F1F2 between 0.1 and 0.3 has been
determined.18,19,29,68,83 The full curve in Fig. 3.10(a) is the calculated Auger (where F1F2 
= 0.15 has been assumed) and radiative lifetimes as a function of carrier concentration.
In the case of MWIR HgCdTe, there are limited lifetime data at low temperatures. The
full curve in Fig. 3.10(b) is the calculated Auger and radiative lifetime as a function of
carrier concentration using F1F2 = 0.15.
Figure 3.11 shows the lifetime as a function of dislocation density at 77 K for n–type
HgCdTe alloys with x = 0.22 and 0.3. Shin et al.84 used the photoconductive decay
technique in MBE HgCdTe samples grown on lattice-matched ZnCdTe and lattice-
mismatched GaAs substrates, and Yamamoto et al.85 used the MIS pulse recovery
technique to relate lifetime and dislocation density in MBE HgCdTe (x = 0.22 and n ≈
1.5×1015 cm–3) and HgCdTe grown by a modified zone melt (x = 0.3, n ≈ 1.0×1015 cm–3),
respectively. In both studies, dislocation recombination was observed for dislocation
densities greater than 5×1015 cm–2. For dislocation densities greater than 5×1015 cm–2,
Yamamoto found τ ∝ ρd−1 , while Shin found τ ∝ ρd−3 / 4 , where ρd is the dislocation
density.
The experimental lifetime data available for n–type HgCdTe in the degenerate regime
are those of Gerhardts et al.39 These are shown in Fig. 3.12, along with the theoretical
predictions of Beattie,72 in which the overlap function term was obtained from the four-
band k⋅p calculations and the possible effect of 300 K background radiation on measured
recombination time was included. The parameter f indicates a fraction of the total
available flux incident on the sample. It is shown that there can be considerable reduction
in recombination time at very low temperatures (3–50 K). In the temperature range 50–
200 K, the background radiation has a negligible effect. The lifetime increases as
expected, when the temperature is lowered from room temperature. At lower
temperatures, the material becomes extrinsic; and the lifetime rolls over and decreases
with a further decrease in temperature. Finally, in the region of extreme degeneracy (at
very low temperatures), the lifetime again increases. We can see that the shape of the
experimental data is consistent with the theory.
For p–type HgCdTe material, conclusions regarding the dominant recombination
mechanisms are still contradictory. Casselman and Petersen32,33 surveyed theoretically
those Auger processes that could occur in nondegenerate p–type HgCdTe, and found that
the Auger 7 might be expected to be important. The results of calculation show that the
intrinsic lifetimes for the two mechanisms (Auger 1 and Auger 7) are on the same order of
magnitude over a wide range of composition and temperature (0.5 < τ iA 7 / τ iA1 < 6) for
0.16 ≤ x ≤ 0.30 and 50 ≤ T ≤ 300 K.
Schacham and Finkman20 have presented experimental evidence for the first time that
the Auger 7 process is a dominating recombination mechanism in p–type Au-doped
Hg0.785Cd0.215Te material grown by the solid state recrystallization method. To obtain a
good agreement with lifetime measurements, combined Auger 7 and radiative
mechanisms, including freezeout and background radiation effects, have been involved.
Figure 3.13 shows the results of lifetime measurements on a Hg0.71Cd0.29Te sample with a

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 106 Chapter 3

Fig. 3.11. Lifetime (near 80 K) as a function of dislocation density for n–type LWIR and
MWIR HgCdTe. Experimental data are taken from Refs. 84 (□) and 85 (ο) (after Ref. 7).

Fig. 3.12. The Auger recombination times in n-type Hg0.8Cd0.2Te for various fractions, f, of
300 K background radiation. The experimental points of Gerhards et al.39 are plotted for
comparison: () Nd = 1.7×1015 cm–3, () Nd = 1.3×1015 cm–3, ( ) Nd = 4×1014 cm–3 (after Ref. 72).


Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 107

Fig. 3.13. Measured lifetime as a function of temperature in p–type Hg0.71Cd0.29Te (after

Ref. 20).

hole concentration of 7.5×1015 cm–3 at 77 K. We can see that the intrinsic region is
dominated entirely by the recombination rate, and the lifetime decreases slightly due to
increases in background flux with the decrease in Eg. Both mechanisms contribute in the
intermediate region. For the ratio τ iA 7 / τ iA1 , the same functional dependence as
Casselman has been used,33 but with a ten-times larger value. Highly compensated
materials are dominated by the SRH recombination.
Adomaitis et al.74 have compared their experimental results obtained from study of
transient photoconductivity in Hg1–xCdxTe (x ≈ 0.20) with theoretical calculations by
Casselman.33 They received quite good agreement between the experiment and theory at a
temperature range of 77–300 K for samples with an equilibrium hole concentration of
1016 – 5×1017 cm–3.
In general, however, numerous experiments concerning the predominant contribution
of the Auger 7 process in recombination mechanisms in p–type HgCdTe material appear
contradictory. Reported excess carrier lifetimes in Hg0.80Cd0.20Te having acceptor
concentrations of ≈ 1016 cm–3 span three orders of magnitude, from 1 ns to 1 µs at 77 K.73
These results may be influenced by the measurement technique or by the quality of the
material. Figure 3.14 shows the measured lifetimes as a function of temperature for bulk,
LPE and MBE p–type LWIR HgCdTe materials. These samples were doped by group II
vacancies and have approximately the same composition and comparable doping.
However, the carrier lifetimes exhibit very different temperature dependences. Chen et
al.86 found that the lifetime increased for temperatures below 30 K and attributed either
the Auger or radiative mechanism as the dominant recombination process. Lacklison and
Capper22 determined the dominant low-temperature recombination mechanism to be SRH.
DeSouza et al.87 found the lifetime to be determined by Auger and SRH recombination
process.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 108 Chapter 3

Fig. 3.14. Lifetime as a function of temperature for p–type LWIR HgCdTe grown by
Bridgman (Ref. 22), LPE (Ref. 86) and MBE (Ref. 87). The reported composition and 77 K
carrier concentrations were 0.21 and 2.3×1015 cm–3 for the Bridgman sample, 0.22 and
1.2×1015 cm–3 for the LPE sample and 0.234 and 4.4×1015 cm–3 for the MBE sample (after
Ref. 7).

Figures 3.15(a) and 3.15(b) show the measured lifetimes as a function of carrier
concentration at temperatures near 80 K for p–type LWIR and MWIR HgCdTe material,
respectively. As in the case of n–type MWIR HgCdTe, there are limited lifetime data for
p–type MWIR HgCdTe near 80 K. Figure 3.15(b) shows the measured lifetime as a
function of carrier concentration for p–type HgCdTe with x = 0.3. Again, a broad range in
lifetime is observed. For carrier concentration near 1×1016 cm–3, the lifetimes range from
approximately 20 ns to 1 µs. The broad range in the reported lifetimes has been attributed
to SRH centers and to charge trapping of excess carriers. The full lines in figures 3.15(a)
and 3.15(b) are the calculated radiative lifetimes as a function of carrier concentration
using equations (3.11) and (3.14) and x = 0.225 and 0.3, respectively.
Generally, very low lifetimes are observed for these undoped bulk HgCdTe materials,
which are made p–type through generation of Hg vacancies by stoichiometric annealing.
It was found, however, that intentionally impurity-doped HgCdTe can be obtained with
relatively high minority carrier lifetime. It is believed that in undoped HgCdTe there is
probably a higher concentration of trapping deep levels that control the excess carrier
lifetimes in SRH recombination processes. The lifetime for the As-doped material is
longer than that of the comparably vacancy-doped material.7,86,88 A similar trend has been
reported by Jones et al.89 for p–type MWIR vacancy- and As-doped HgCdTe. It was
found that the deep-level defects in undoped p–type HgCdTe act donorlike, with electron
capture being larger than the hole capture.89,90
Experimental data obtained in undoped p–type HgCdTe at lower temperatures
indicate that the dominant recombination mechanism is a SRH process involving the
mercury vacancy. Polla et al.18,91,92 have characterized deep levels in Hg1–xCdxTe (0.2 < x
< 0.4) by a number of complementary techniques such as deep-level transient

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 109

(a)

(b)

Fig. 3.15. (a) Lifetime (near 80 K) as a function of carrier concentration for p–type LWIR
HgCdTe. The open symbols are for the group II vacancy doped material, and the full
symbols are for the As-doped material. The full line is calculated radiative lifetime using
x = 0.225. (b) Lifetime (near 80 K) as a function of carrier concentration for MWIR HgCdTe.
The open symbols are for the group II vacancy doped material, and the full symbols are
for the As-doped material. The full line is calculated radiative lifetime using x = 0.3 (after
Ref. 7).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 110 Chapter 3

spectroscopy, admittance spectroscopy, thermally stimulated current, population

modulation spectroscopy and, in addition, have studied minority carrier lifetimes using
diode recovery and population modulation. The results show that two deep levels are
commonly observed, one slightly below midgap and one at approximately (3/4)Eg. The
near-midgap center has a density proportional to the number of Hg vacancies and
determines the excess carrier lifetime in p–type HgCdTe. The fact that deep levels with
the same activation energy are also observed in n–type material with low vacancy
concentrations suggests a mercury-substituted impurity.92 The acceptor ionization energy
Ea in Bridgman-grown HgCdTe increases with x and decreases with N1/a 3 .88,93
Nemirovsky et al.73,75,94–97 indicate that the transient and steady-state lifetimes differ
by more than one order of magnitude at 77 K in p–type Hg1–xCdxTe (x ≈ 0.20). This
difference is attributed to minority-carrier trapping and explains—in part—the large range
of lifetimes reported for this material. In interpreting lifetime data, it is important to make
a distinction between the three commonly measured lifetimes: the excess minority carrier
lifetime, the excess majority carrier lifetime and the transient lifetime. Each of these time
constants refers to a different aspect of electron-hole recombination and therefore
provides different information. The theory of the relevant lifetimes is presented in Refs.
75 and 97.
It is also well known that dislocations originating from material growth and device
processing affect minority carrier lifetime,84,85,98–100 which is directly related to device
performance.85,101–104 The observed dependence of lifetime versus temperature and each-
pit density levels suggests that threading dislocations behave as SRH recombination
centers.100
Remaining nonfundamental recombination mechanisms in HgCdTe such as single-
phonon recombination, plasmon recombination and influence of magnetic field on carrier
lifetime have been reviewed by Nimtz et al.4,105

3.7 Carrier lifetime in lead salts

Although direct one-phonon recombination as well as plasmon recombination have been
proposed theoretically106,107 and identified experimentally108,109 for very small energy gaps
of IV–VI compounds, in samples with gaps of 0.1 eV or more, radiative recombination or
Auger recombination is dominant.4,105,110,111 Nevertheless, in most samples, extrinsic
recombination via local centers (SRH recombination) cannot be ignored. In this section a
theoretical description of radiative and Auger recombination in small-gap IV–VI
semiconductors is given. Some representative experimental data are presented and
discussed next.
Ziep et al.106 have calculated the lifetime determined by radiative recombination in
terms of the Kane-type two-band model case and Boltzmann statistics. It appears,
however, that in the case of mirror-symmetric band structure of lead salts ( m∗e ≈ m∗h ), a
good approximation for recombination rate is the following equation112:

10−15 n r E g2 n i2
in cm3/s,
(kT )3 / 2 K1 / 2 (2 + 1/K )3 / 2 (m∗ / m )
GR = 5/ 2
(3.41)

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 111

where K = m∗l / m∗t is the effective mass anisotropy coefficient. The mass m* can be
determined if we know the longitudinal m∗l and transverse m∗t components of the
effective mass, since m* = [1/3(2/ m∗t +1/ m∗l )]−1. In Eq. (3.41), the values of kT and Eg
should be expressed in electron volts.
For a long time, the Auger process was considered to be a low-efficiency channel for
nonradiative recombination in semiconductors of the IV–VI type. The valence band and
the conduction band with mirror-reflection symmetry occur at point L of the Brillouin
zone (number of valleys w = 4). In such a case, the energy and momentum conservation
laws are difficult to fulfill for impact recombination, especially for carriers near the band
edges when only single-valley interaction is taken into account. Since 1976,113 many
theoretical and experimental works have been published110,114–122 in which the intervalley
interaction of carriers is considered, and it has been found that even at lower temperatures
the lifetime of carriers is determined by impact recombination. According to the
intervalley carrier interaction model (Fig. 3.16), an electron and hole from valley (a),
characterized by ”heavy” mass m∗l , and a third carrier from valley (b) with a “light” mass
m∗t (in PbSnTe mass anisotropy coefficient K > 10) participate in impact recombination
in the given direction. As a result of this interaction the “heavy” electron and hole carriers
recombine, and the liberated energy and momentum are transferred to the “light” carrier.

Fig. 3.16. Intervalley Auger recombination.

Two cases should be considered:

• all scattered carriers are at a definite point of the Brillouin zone,

• the initial carriers are in different valleys of the band.

In the case of Boltzmann statistics,107

(τ )
j −1
A ( )
= C Aj n o2 + 2n o p o , (3.42)

where C Aj is the Auger coefficient for j-th recombination mechanism. Then,

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 112 Chapter 3

( ) + (τ )
τ −A1 = τ aA
−1 b −1
A . (3.43)

Different approximations for CA, especially for the process (b), can be found in the
literature.113,118,119,123–125 For the intervalley process proposed by Emtage within the
parabolic band model,113

w −1 q2 3  −1 
CA = (2π )5 / 2 (kT )1 / 2 − 7/2
E exp − EgK  . (3.44)
w 2 (4π εoε∞ )2 m∗l 1 / 3m∗t 2 / 3  2kT 
g
 

Nonparabolic bands of the Kane type are expected to reduce the Auger transition
rate.117,119 However, the Emtage expression is a good approximation of more exactly
calculated Auger coefficients.118
Ziep et al.123 carried out a comprehensive examination of nonradiative and radiative
recombination mechanisms in the mixed crystals Pb0.78Sn0.22Te and Pb0.,91Sn0.09Se for a
temperature interval 20 < T < 400 K and in a doping range 0 < Nd < 1019 cm–3. In the
calculations, the degeneracy of the carrier gas, anisotropy and to some extent the
nonparabolicity of the band structure were taken into account. Their calculated data for
Pb0.78Sn0.22Te are shown in Fig. 3.17. The radiative lifetime as well as the Auger lifetime
have a maximum in the transition from the extrinsic to the intrinsic region. At low
temperatures in the range of low dopant concentrations, the lifetime is determined by
radiative recombination. As the temperature increases, the Auger recombination comes to
the fore and determines the carrier lifetimes at room temperature. At comparable energy
gaps, the radiative lifetimes differ only slightly in PbSnTe and PbSnSe compounds.
However, due to the smaller anisotropy of the isoenergetic surface in PbSnSe as
compared with PbSnTe, the Auger lifetime is higher in PbSnSe than in PbSnTe. At carrier
concentrations above 1019 cm–3, the plasmon recombination dominates over radiative and
Auger recombinations.
Dmitriev124 has calculated the Auger carrier lifetime in PbSnTe, taking into account
the carrier degeneracy and the exact expressions for the overlap integrals.125 The
calculated lifetimes are greater than those obtained previously.
Experimental investigations of lead salts confirm that the carrier lifetime is determined
by band-to-band recombination as well as by any SRH recombination. The results of the
first extensive investigation of Pb1–xSnxTe (0.17 ≤ x ≤ 0.20) revealed an extremely large
variation of the lifetime (10–12 < τ < 10–8 s at 77 K) measured from photoconductivity and
PEM effects111,116,126 (see Fig. 3.18). The highest observed lifetime fits quite well to the
straight line calculated according to Emtage’s theory of Auger recombination.123 All Cd-
and In-doped samples are characterized by considerably lower lifetimes in spite of their
lower free-carrier concentration. Donor levels with ionization energies from 12 to 25 meV
are presumed to work as recombination centers.
In the case of PbTe epitaxial layers on BaF2 substrates, a good agreement with the
Emtage theory has been obtained by Lischka and Huber.115 However, these authors
observed a second lifetime branch with a much longer lifetime than that for Auger
recombination (Fig. 3.19). This branch was attributed to deep levels acting as minority
carrier traps at intermediate temperatures.110 Zogg et al.120 observed up to four lifetime
branches determined by the transient photoconductivity method in n– and p–type PbSe
epitaxial layers. The shortest lifetimes were in agreement with calculated values for direct

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 113

Fig. 3.17. Calculated lifetime vs temperature for Auger, radiative, and plasmon
recombination in Pb0.78Sn0.22Te (after Ref. 123).

Fig. 3.18. PEM lifetime (T = 100 K) vs doping level for Pb1–xSnxTe (0.17 ≤ x ≤ 0.20) samples of
different origin: Bridgman annealed (•), THM (+), vapor-phase (○), uncompensated (),
and Cd-compensated () (after Ref. 111).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 114 Chapter 3

Fig. 3.19. Carrier lifetime vs temperature for p–PbTe: experimental values of lifetime τ1
•
( ) (p = 9.5×1016 cm–3), solid curves are calculated for Auger recombination assuming
m*l / m*t = 10 (the upper curve) and m*l / m*t = 14 (the lower curve); experimental data of a
second time constant τ2 occurring in the decay of the photocurrent pulse (○) (after Ref.
115).

(Auger and radiative) recombination below ≈ 250 K for samples with carrier
concentrations ≤ 2×1017 cm–3. From the longer observed lifetime branches, three impurity
levels (separated between 20 and 50 meV from the nearer band edge), acting principally
as minority carrier traps at intermediate temperature, were calculated. Also, Shahar et
al.121 confirmed that the carrier lifetime in undoped Pb0.8Sn0.2Te layers grown by liquid
phase epitaxy is determined by band-to-band radiative and Auger recombination
mechanisms in the temperature range 10–110 K, while in In-doped PbTe layers,
recombination takes place via nonradiative centers.
Two other groups, Schlicht et al.127 and Weiser et al.,128 have not obtained agreement
between experimental data on PbSnTe epitaxial layers on (111) BaF2 substrates and the
calculated Auger lifetime. Measurements of the photoconductive decay time yield values
that at 77 K are longer than predicted by the Emtage theory. Also, the dependence of
lifetime on carrier density in the extrinsic range of conductivity was not observed. Weiser
et al.128 suggest that the lifetime enhancement by about two orders of magnitude at T ≈ 14
K is the result of photon recycling, and they explain the lifetime enhancement by about
one order of magnitude at 77 K as overestimation of the recombination rate within the
Emtage theory or by the fact that strains in the sample reduce the importance of the
multivalley Auger process. Genzow et al.129 theoretically investigated the influence of
uniaxial strain on the Auger and radiative recombinations in PbSnTe and obtained quite

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 115

good agreement of their theoretical results with the experimental data given by Weiser et
al.128
We conclude that the mechanism of Auger recombination in small-gap IV-VI
semiconductors is expected to be important but is not fully clarified. Experimental
evidence for this mechanism is not yet unambiguously presented. Some disagreement
between the experimental results reported by the various authors may be related to
differences in sample preparation as well as differences in the experimental methods used
to measure the lifetime. Also, these discrepancies are assumed to be due to the lack of
accurate band parameters for lead salts and to the theoretical description of screening in
the Auger process.4 In spite of these differences, some conclusions may be drawn:

• With decreasing energy gaps and increasing temperature, impact recombination

becomes more important.
• In samples having a large number of lattice defects (impurities, native defects,
misfit dislocations), the recombination mechanism is attributed to SRH centers.
• In defect-free undoped material, carrier lifetime is determined by direct band-to-
band recombination.

The recombination processes in polycrystalline IV–VI films are modified by grain

boundary potential barriers. Different theories of photoconductivity in lead chalcogenide
polycrystalline layers have been proposed. Their brief review is given by Espevik et al.130
and Johnson.131 More recently, Neustroev and Osipov132 developed a theoretical model
based on the knowledge that polycrystalline materials are a two-phase system. Low-ohmic
n–type conductance crystallites are surrounded by oxygen-saturated p–type inversion
layers with a large concentration of acceptor-type states. During irradiation, electrons and
holes generating in crystallites are separated by surface barriers. Due to spatial division of
photocarriers, their lifetime rises abruptly and, consequently, so does their
photosensitivity.
Similarly, as in single-crystal lead salt samples, the carrier lifetime in polycrystalline
materials is determined by three principal recombination processes: SRH, radiative and
Auger. Recently Vaitkus et al.133 have investigated picosecond photoconductivity in
highly excited electrolytically deposited PbS and vacuum-evaporated PbTe
polycrystalline films. They described the dependence of the lifetime on nonequilibrium
concentrations, taking into account linear, trap-assisted Auger, interband radiative, and
Auger recombinations134

+ (γ At N t + γ R )∆N + γ A ∆N 2 ,
1 1
= (3.45)
τ τR

where γAt is the trap Auger coefficient, Nt is the trap density, γR is the radiative
recombination coefficient, and γA is the interband Auger recombination coefficient. Good
agreement of the theoretical curves based on Eq. (3.45) with experimental points in all
investigated films (freshly made and annealed) has been obtained, assuming γAtNt + γR =
2.2×10–11 cm3/s and γA = 5.3×10–29 cm6/s. Interband Auger recombination in the grain
bulk is the main mechanism of carriers. The structure and the film preparation process
influence only the slower recombination. The same results have been obtained for PbTe
films.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 116 Chapter 3

3.8 Carrier lifetimes in superlattices

3.8.1 HgTe/CdTe superlattices

The HgTe/CdTe superlattices (SLs) system belonging to type III superlattices was the first
from a new class of quantum-size structures for IR photoelectronics, which was proposed
as a promising new alternate structure for the construction of long-wave IR detectors to
replace those of HgCdTe alloys.135,136 Jiang et al.137 indicated that for HgCdTe quantum
wells, Auger recombination can be significantly suppressed. They predicted almost an
order of magnitude increase in Auger lifetime in the quantum well case over the bulk in
the midwave infrared. To date, however, attempts to realize HgTe/CdTe SLs with
properties suitable for IR detectors with parameters comparable with HgCdTe-alloy
photodetectors, have been unsuccessful, in spite of considerable amount of fundamental
research in this field. It seems to be determined by interface instabilities of superlattices due
to weak Hg chemical bonding in the material.
In the beginning, the lifetimes of about of 1 to 10 ns were measured either by a
photoconductive decay technique and analysis of photo-Hall data138 for p–type
HgTe/CdTe SL or calculated from dark currents in metal-insulator-semiconductor (MIS)
structures.139 These lifetimes are rather short and are the reason for poor quantum
efficiencies of HgTe/CdTe SL structures. After improving the superlattice technology
with high-quality interfaces between the layers (to minimize interface recombination), the
excess carrier lifetimes of several hundreds of nanoseconds have been reported,140–142
which approaches numbers typically observed in good-quality bulk HgCdTe of
comparable Eg and doping levels.
Figure 3.20 displays experimental lifetime measurements reported by Reisinger et
al.141 for n–type and p–type superlattices with carrier concentration in the low 1015 cm–3
and the low 1016 cm–3 range, respectively. Lifetime near 400 ns has been recorded at
temperatures below about 100 K for n–type SL (carrier concentration of about 3×1015
cm–3 at 77 K) with a λc ≈ 8 µm. Figure 3.20 also shows calculated lifetimes for bulk
HgCdTe of equivalent Eg and similar doping levels for a number of recombination
processes. The observed lifetimes are about one order of magnitude shorter than those
calculated based on the Auger 1 process alone, and are consistent with Shockley-Read-
Hall (SRH) recombination with a midgap trap density not in excess of 5×1014 cm–3. In the
case of intentionally doped p–type material using As, the superlattice had a measured 80
K lifetime of about 40 ns, which is consistent with calculations of Auger 7 lifetime in the
equivalent alloy for the 2×1016 cm–3 concentration and λc of ≈ 8.5 μm.
Results presented by Shin et al.142 reveal that the minority carrier lifetime in both n–
and p–type HgTe/CdTe SLs appears to be limited by a SRH-related recombination
mechanism that is most likely due to interface defects. The n–type SL samples show a
lifetime of 90–100 ns at 77–160 K, which then decreases with increasing temperature
down to 50 ns near 300 K. For p–type SLs, the measured lifetime is approximately 20 ns
at 77 K and increases continuously to 85 ns at 300 K.
More recently published results136 have indicated that the lifetime exhibited by an n–
type (8×1014 cm–3) SL with cutoff wavelength λc = 4.8 µm exceeded the bulk Auger limit,
and the peak of τR = 20 µs at 140 K (17 µs at 77 K) was higher than any observed in
HgCdTe with the same energy gap and doping level.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 117

(a)

(b)

Fig. 3.20. Carrier lifetime versus reciprocal temperature for unintentionally doped
HgTe/CdTe superlattices with λc ≈ 8 μm at 77 K: (a) n–type structure with Nd – Na ≈ 3×1015
cm–3; (b) p–type structure with Na – Nd ≈ 2×1016 cm–3. The lifetime due to various
recombination processes for the equivalent HgCdTe alloys are shown (after Ref. 141).

3.8.2 InAs/Ga1–xInxSb strained layer superlattices

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 118 Chapter 3

Far-IR response in InAs/Ga1–xInxSb strained layer superlattice (SLS) arises due to a

broken-gap type II band alignment and internal strain that lowers the conduction band
minimum of InAs and raises the heavy-hole band in Ga1–xInxSb by the deformation
potential effect.
The calculations143,144 of band-to-band Auger and radiative recombination lifetimes
for InAs/GaInSb SLSs showed that in these objects, the p–type Auger recombination rates
are suppressed by several orders, compared to those of bulk HgCdTe with similar
bandgap, due to the flatness of the light and heavy hole bands whose large splitting
exceeds the energy gap, thus limiting phase space for recombination transitions. The
suppression is large enough that these materials are excellent candidates in the LWIR
range (> 11 µm). By contrast, n–type materials are less advantageous because the Auger
suppression achieved by increasing the Ga1–xInxSb layer widths, thereby flattening the
lowest conduction band, is far less effective than the band structure adjustments possible
in the valence band.144–146
Comparison of theoretically calculated and experimentally observed lifetimes at 77 K
for 10 µm 39Å InAs/25Å Ga0.75In0.25Sb SLS and 10 µm HgCdTe is presented in Fig. 3.21.
The agreement between theory and experiment for carrier densities above 2×1017 cm–3 is
good. The discrepancy between both types of results for lower carrier densities is due to
Shockley-Read recombination processes having a τ ≈ 6×10–9 s, which has been not taken
into account in the calculations. For higher carrier densities, the SL carrier lifetime is two
orders of magnitude longer than in HgCdTe, however, in the low doping region (below
1015 cm–3, necessary in fabrication of high-performance p–on–n HgCdTe photodiodes),
experimentally measured carrier lifetime in HgCdTe is more than two orders of
magnitude longer than in SL.
InAs/GaInSb, SLSs are also employed as the active regions of midwave IR lasers
operating in the 2.5–6-μm spectral region. Recently Meyer et al.147 have experimentally
determined Auger coefficients for InAs/GaInSb quantum well structures with energy gaps
corresponding to 3.1–4.8 μm and have compared their values with that for typical III–V
and II–VI type I superlattices. The Auger coefficient is defined by the expression γ3 ≡
1/τAn2. Figure 3.22 summarizes the Auger coefficients at ≈ 300 K for different material
systems: a wide variety of type I materials (all open points), including bulk and quantum
well III–V semiconductors as well as HgCdTe (upside-down triangles). We can see that
the room-temperature Auger coefficients for seven different InAs/GaInSb SLs are found
to be nearly an order of magnitude lower than typical type I results for the same
wavelength, indicating a significant supression of Auger losses in the antimonide type II
SLs. The data imply that at this temperature, the Auger rate is relatively insensitive to
details of the band structure.

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 119

Fig. 3.21. Comparison of measured and calculated carrier lifetimes of a 39Å InAs/25Å
Ga0.75In0.25Sb superlattice (10 µm energy gap) as a function of carrier density (after Ref.
145).

Fig. 3.22. Auger coefficient vs gap wavelength for type II InAs/GaInSb quantum wells
(filled points) and a variety of type I materials (open points). The filled circles are from the
photoconductive response technique and filled boxes from the lasing thresholds,
whereas the filled triangles are from pump-probe experiments. The solid and dashed
curves are guides to the eye joining data from the present type II and bulk type I
experiments, respectively (after Ref. 147).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 120 Chapter 3

References
1. J. S. Blakemore, Semiconductor Statistics, Pergamon Press, Oxford (1962).
2. R. A. Smith, Semiconductors, Cambridge University Press, London (1978).
3. P. Landsberg, Recombination in Semiconductors, Cambridge University Press,
Cambridge (1991).
4. G. Nimtz, “Recombination in narrow-gap semiconductors,” Phys. Rep. 63, 265–300
(1980).
5. P. E. Petersen, “Auger recombination in mercury cadmium telluride,” in
Semiconductors and Semimetals, Vol. 18, pp. 121–155, edited by R. K. Willardson
and A. C. Beer, Academic Press, New York (1981).
6. N. S. Baryshev, B. L. Gelmont, and M. I. Ibragimova, “Recombination mechanisms
in CdxHg1–xTe,” Fiz. Tekh. Poluprovodn. 24, 209–224 (1990).
7. V. C. Lopes, A. J. Syllaios, and M. C. Chen, “Minority carrier lifetime in mercury
cadmium telluride,” Semicond. Sci. Technol. 8, 824–841 (1993).
8. A. V. Dmitriev and M. Mocker, “Recombination and ionization in narrow gap
semiconductors,” Physics Reports 257, 85–131 (1995).
9. W. Shockley and W. T. Read, “Statistics of the recombination of holes and
electrons,” Phys. Rev. 87, 835–842 (1952).
10. R. N. Hall, “Electron-hole recombination in germanium,” Phys. Rev. 87, 387 (1952).
11. T. S. Moss, G. J. Burrel, and B. Ellis, Semiconductor Optoelectronics, Butterworths,
London (1973).
12. W. Van Roosbroeck and W. Shockley, “Photon-radiative recombination of electrons
and holes in germanium,” Phys. Rev. 94, 1558–1560 (1954).
13. R. N. Hall, “Recombination processes in semiconductors,” Proc. IEE. B106 (Suppl.
17), 923–931 (1959).
14. R. G. Humpreys, “Radiative lifetime in semiconductors for infrared detection,”
Infrared Phys. 23, 171–175 (1983).
15. R. G. Humpreys, “Radiative lifetime in semiconductors for infrared detection,”
Infrared Phys. 26, 337–342 (1986).
16. J. Bardeen, F. J. Blatt, and L. H. Hall, “Indirect transitions from the valence to the
conduction bands,” in Photoconductivity Conference, Atlantic City 1954, pp. 146–
154, edited by R. Breckenridge, B. Russell, and E. Hahn, Wiley, New York (1956).
17. M. A. Kinch, M. J. Brau, and A. Simmons, “Recombination mechanisms in 8–14 µm
HgCdTe,” J. Appl. Phys. 44, 1649–1663 (1973).
18. D. L. Polla, S. P. Tobin, M. B. Reine, and A. K. Sood, “Experimental determination
of minority-carrier lifetime and recombination mechanisms in p–type Hg1–xCdxTe,”
J. Appl. Phys. 52, 5182–5194 (1981).
19. R. G. Pratt, J. Hewett, P. Capper, C. L. Jones, and M. J. Quelch, “Minority carrier
lifetime in n–type Bridgman grown CdxHg1–xTe,” J. Appl. Phys. 54, 5152–5157
(1983).
20. S. E. Schacham and E. Finkman, “Recombination mechanisms in p–type HgCdTe:
Freezeout and background flux effects,” J. Appl. Phys. 57, 2001–2009 (1985).
21. R. G. Pratt, J. Hewett, P. Capper, C. L. Jones, and N. Judd, “Minority-carrier
lifetime in doped and undoped n–type CdxHg1–xTe,” J. Appl. Phys. 60, 2377–2385
(1986).
22. D. E. Lacklinson and P. Capper, “Minority carrier lifetime in doped and undoped p–
type CdxHg1–xTe,” Semicond. Sci. Technol. 2, 33–43 (1987).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 121

23. P. S. Wijewarnasuriya, M. D. Lange, S. Sivananthan, and J. P. Faurie, “Carrier

recombination in indium-doped HgCdTe(211)B epitaxial layers grown by molecular
beam epitaxy, J. Appl. Phys. 75, 1005–1009 (1994).
24. A. R. Beattie, “Quantum efficiency in InSb,” J. Phys. Chem. Solids 23, 1049–1056
(1962).
25. M. Takeshima, “Auger recombination in InAs, GaSb, InP, and GaAs,” J. Appl.
Phys. 43, 4114–4119 (1972).
26. A. Beattie and P. T. Landsberg, “Auger effect in semiconductors,” Proc. Roy. Soc.
A249, 16–29 (1959).
27. J. Tauc, “Electron impact ionization in semiconductors,” J. Phys. Chem. Solids 8,
219–223 (1959).
28. A. R. Beattie and G. Smith, “Recombination in semiconductors by a light hole
Auger transition,” Phys. Stat. Sol. 19, 577–586 (1967).
29. P. E. Petersen, “Auger recombination in Hg1–xCdxTe,” J. Appl. Phys. 41, 3465–3467
(1970).
30. B. L. Gelmont, “Three-band Kane model of Auger recombination,” Zh. Eksper.
Tieor. Fiz. 75, 536–541 (1978).
31. B. L. Gelmont, “Auger recombination in narrow gap p–type semiconductors,” Fiz.
Tekh. Poluprovodn. 15, 1316–1319 (1981).
32. T. N. Casselman and P. E. Petersen, “A comparison of the dominant Auger
transitions in p–type (Hg,Cd)Te,” Solid State Commun. 33, 615–619 (1980).
33. T. N. Casselman, “Calculation of the Auger lifetime in p–type Hg1–xCdxTe,” J. Appl.
Phys. 52, 848–854 (1981).
34. A. Sugimura, “Band-to-band Auger effect in GaSb and InAs lasers,” J. Appl. Phys.
51, 4405–4411 (1980).
35. A. Sugimura, “Band-to-band Auger effects in long wavelength multinary III–V alloy
semiconductor lasers,” IEEE J. Quant. Electr. QE-18, 352–363 (1982).
36. B. L. Gelmont, Z. N. Sokolova, and I. N. Yassievich, “Auger recombination in
direct gap p–type semiconductors,” Fiz. Tekh. Poluprovodn. 16, 592–600 (1982).
37. A. Antoncik and P. T. Landsberg, “Overlap integrals for Bloch electrons,” Proc.
Phys. Soc. 82, 337–342 (1963).
38. M. G. Burt, S. Brand, C. Smith, and R. A. Abram, “Overlap integrals for Auger
recombination in direct-bandgap semiconductors: calculations for conduction and
heavy-hole bands in GaAs and InP,” J. Phys. C: Solid State Phys. 17, 6385–6401
(1984).
39. R. R. Gerhardts, R. Dornhaus, and G. Nimtz, “The Auger-effect in Hg1–xCdxTe,”
Solid-State Electr. 21, 1467–1470 (1978).
40. D. M. Eagles, “The phonon-assisted Auger effect in semiconductors,” Proc. Phys.
Soc. 78, 204–216 (1961).
41. W. Lochman, “Phonon-assisted Auger recombination in semiconductors,” Phys.
Stat. Sol. (a) 40, 285–292 (1977).
42. W. Lochman, “Scattering mechanisms in phonon-assisted Auger recombination,”
Phys. Stat. Sol. (a) 42, 181–185 (1977).
43. L. Huldt, “Phonon-assisted Auger recombination in germanium,” Phys. Stat. Sol. (a)
33, 607–614 (1976).
44. L. Bess, “Possible mechanism for radiationless recombination in semiconductors,”
Phys. Rev. 105, 1469–1475 (1957).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 122 Chapter 3

45. K. P. Sinha and M. DiDomenico, Jr., “Effects of plasma screening and Auger
recombination on the luminescent efficiency in GaP,” Phys. Rev. B1, 2623–2631
(1970).
46. V. L. Bonch-Bruevich and V. B. Glasko, “On the theory of current carriers ‘cascade’
recombination in homopolar semiconductors,” Fiz. Tverd. Tela 2, 513–523 (1960).
47. P. T. Landsberg, C. Rhys-Roberts, and P. Lal, “Auger recombination and impact
ionization involving traps in semiconductors,” Proc. Phys. Soc. 84, 915–931 (1964).
48. P. T. Landsberg, “Non-radiative transitions in semiconductors,” Phys. Stat. Sol. 41,
457–489 (1970).
49. D. J. Robbins and P. T. Landsberg, "Impact ionization and auger recombination
involving traps in semiconductors," J. Phys. C: Solid State Phys. 13, 2425-2439
(1980).
50. P. T. Landsberg and D. J. Robbins, “The first 70 semiconductor Auger processes,”
Solid-State Electr. 21, 1289–1294 (1978).
51. M. Y. Pines and O. M. Stafsudd, “Recombination processes in intrinsic
semiconductors using impact ionization capture cross sections in indium antimonide
and mercury cadmium telluride,” Infrared Phys. 20, 73–91 (1980).
52. J. E. Hollis, S. C. Choo, and E. L. Heasell, “Recombination centers in InSb,” J.
Appl. Phys. 38, 1626–1636 (1967).
53. R. N. Zitter, A. J. Strauss, and A. E. Attard, “Recombination processes in p–type
indium antimonide,” Phys. Rev. 115, 266–273 (1959).
54. P. W. Kruse, “Indium antimonide photoconductive and photoelectromagnetic
detectors,” in Semiconductors and Semimetals, Vol. 5, pp. 15–83, edited by R. K.
Willardson and A. C. Beer, Academic Press, New York (1970).
55. P. Berdahl, “Galvanomagnetic luminescence and the quantum efficiency of radiative
recombination of InSb,” J. Appl. Phys. 63, 5846–5858 (1988).
56. S. R. Jost, V. F. Meikleham, and T. H. Myers, “InSb: A key for IR detector
applications,” Mater. Res. Soc. Sym. Proc. 90, 429–435, (1987).
57. V. L. Dalal, W. A. Hicinbothem, and H. Kressel, “Carrier lifetimes in epitaxial
InAs,” Appl. Phys. Lett. 24, 184–185 (1974).
58. H. H. Wider and D. A. Collins, “Minority carrier lifetime in InAs epilayers,” Appl.
Phys. Lett. 25, 742–743 (1974).
59. A. Rogalski and Z. Orman, “Band-to-band recombination in InAs1–xSbx,” Infrared
Phys. 25, 551–560 (1985).
60. N. Chand, “Carrier diffusion lengths and lifetimes in InGaAs,” in pp. 127–129,
edited by P. Bhattacharya, INSPEC, IEE, London (1993).
61. R. K. Ahrenkiel, R. Ellingson, S. Johnston, and M. Wanlass, “Recombination
lifetime of In0.53Ga0.47As as a function of doping density,” Appl. Phys. Lett. 72,
3470–3472 (1998).
62. M. Gallant and A. Zemel, “Long minority hole diffusion length and evidence for
bulk radiative recombination limited lifetime in InP/InGaAs/InP double
heterostructure,” Appl. Phys. Lett. 52, 1686–1688 (1988).
63. R. Trommer and L. Hoffmann, “Large-hole diffusion length and lifetime in
InGaAs/InP double-heterostructure photodiodes,” Electronics Letters 22, 360–362
(1986).
64. C. H. Henry, R. A. Logan, F. R. Merritt, and C. G. Bethea, “Radiative and
nonradiative lifetimes in n–type and p–type 1.6 μm InGaAs,” Electronics Letters 20,
358–359 (1984).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 123

65. A. Rogalski and R. Ciupa, “Performance limitation of short wavelength infrared

InGaAs and HgCdTe photodiodes,” J. Electron. Mater. 28, 630–636 (1999).
66. C. H. Grein, H. Ehrenreich, and E. Runge, “Radiative lifetime in semiconductors:
influence of photon recycling,” Proc. SPIE 2999, 11–18 (1998).
67. F. Bartoli, R. Allen, L. Esterowitz, and M. Kruer, “Auger-limited carrier lifetimes in
HgCdTe at high excess carrier concentrations,” J. Appl. Phys. 45, 2150–2154
(1974).
68. M. Baker, F. A. Capocci, D. E. Charlton, and J. T. M. Wotherspoon,
“Recombination in cadmium mercury telluride detectors,” Solid-State Electr. 21,
1475–1480 (1978).
69. D. L. Polla and C. E. Jones, “Deep level studies of HgCdTe. I: Narrow-band-gap
space-charge spectroscopy,” J. Appl. Phys. 52, 5118–5131 (1981).
70. D. L. Polla, M. B. Reine, and C. E. Jones, “Deep level studies of HgCdTe. II:
Correlation with photodiode performance,” J. Appl. Phys. 52, 5132–5138 (1981).
71. J. Calas and J. Allegre, “Recombination processes at temperatures lower than 20 K
in Hg0.8Cd0.2Te,” Phys. Stat. Sol. (b) 112, 179–186 (1982).
72. A. R. Beattie, “Auger recombination in Cd0.8Hg0.8Te and the effect of background
radiation on its measurement,” Semicon. Sci. Technol. 2, 281–287 (1987).
73. R. Fastow and Y. Nemirovsky, “The excess carrier lifetime in p-type HgCdTe
measured by photoconductive decay,” J. Appl. Phys. 66, 1705–1710 (1989).
74. E. Adomaitis, K. Grigoras, and A. Krotkus, “Minority carrier recombination in p–
type CdxHg1–xTe,” Semicon. Sci. Technol. 5, 836–841 (1990).
75. R. Fastow, D. Goren, and Y. Nemirovsky, “Shockley-Read recombination and
trapping in p–type HgCdTe,” J. Appl. Phys. 68, 3405–3412 (1990).
76. R. Zucca, D. D. Edwall, J. S. Chen, S. L. Johnston, and C. R. Younger, “Minority
carrier lifetimes of metalorganic chemical vapor deposition long-wavelength
infrared HgCdTe on GaAs,” J. Electron. Materials B9, 1823–1828 (1991).
77. Y. L. Tyan, T. R. Schimert, and L. T. Claiborne, “Analysis of excess carrier lifetime
in p–type HgCdTe using a three-level Schockley-Read model,” J. Vac. Sci. Technol.
10, 1560–1568 (1992).
78. S. Barton, D. Dutton, P. Capper, C. L. Jones, and N. Metcalfe, “Minority carrier
lifetime in doped and undoped epitaxially grown n–type CdxHg1–xTe,” J. Electronic
Materilas 24, 1759–1764 (1995).
79. P. S. Wijewarnasuriya, M.D. Lange, S. Sivananthan, and J. P. Faurie, “Minority
carrier lifetime in indium-doped HgCdTe(211)B epitaxial layers grown by molecular
beam epitaxy,” J. Electronic Materials 24, 545–549 (1995).
80. S. Barton, P. Capper, C. L. Jones, N. Metcalfe, and D. Dutton, “Determination of
Shockley-Read trap parameters in n– and p–type epitaxial CdxHg1–xTe,” Semicond.
Sci. Technol. 11, 1163–1167 (1996).
81. D. D. Edwall, R. E DeWames, W. V. McLevige, J. G. Pasko, and J. M. Arias,
“Measurement of minority carrier lifetime in n–type MBE HgCdTe and its
dependence on annealing,” J. Electronic Materials 27, 698–702 (1998).
82. P. Capper, “The behaviour of selected impurities in CdxHg1–xTe,” J. Cryst. Growth
57, 280–299 (1982).
83. D. Long and J. L. Schmit, “Mercury-cadmium telluride and closely related alloys,”
in Semiconductors and Semimetals, Vol. 5, pp. 175–255, edited by R. K. Willardson
and A. C. Beer, Academic Press, New York (1970).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 124 Chapter 3

84. S. H. Shin, J. M. Arias, M. Zandian, J. G. Pasko, and R. E. DeWames, “Effect of the

dislocation density on minority-carrier lifetime in molecular beam epitaxial
HgCdTe,” Appl. Phys. Lett. 59, 2718–2720 (1991).
85. T. Yamamoto, Y. Miyamoto, and K. Tanikawa, “Minority carrier lifetime in the
region close to the interface between the anodic oxide CdHgTe,” J. Crystal Growth
72, 270–274 (1985).
86. J. S. Chen, J. Bajaj, W. E. Tennant, D. S. Lo, M. Brown, and G. Bostrup, “On the
Auger recombination process in p–type LPE HgCdTe,” Mater. Res. Soc. 90, 287–
293 (1987).
87. M. E. de Souza, M. Boukerche, and J. P. Faurie, “Minority-carrier lifetime in p–type
(111)B HgCdTe grown by molecular-beam epitaxy,” J. Appl. Phys.68, 5195–5199
(1990).
88. P. Capper, ”A review of impurity behavior in bulk and epitaxial Hg1–xCdxTe,” J.
Vac. Sci. Technol. B9, 1667–1681 (1991).
89. C. E. Jones, K. James, J. Merz, R. Braunstein, M. Burd, M. Eetemadi, S. Hutton,
and J. Drumheller, “Status of point defects in HgCdTe,” J. Vac. Sci. Technol. A3,
131–137 (1985).
90. T. Tung, M. H. Kalisher, A. P. Stevens, and P. E. Herning, “Liquid-phase epitaxy of
HgCdTe from Hg solution: A route to infrared detector structures,” Mater. Res. Soc.
Sym. Proc. 90, 321–354 (1987).
91. D. L. Polla, W. Scott, and C. E. Jones, “Nonradiative recombination at deep
impurity levels in p–type Hg1–xCdxTe,” in Lecture Notes in Physics, Vol. 152, pp.
435–438, edited by E. Gornik, H. Heinrich and L. Palmetshofer, Springer-Verlag,
Berlin (1982).
92. D. L. Polla, R. L. Aggarwall, D. A. Nelson, J. F. Shanley, and M. B. Reine, “Hg
vacancy related lifetime in HgCdTe by optical modulation spectroscopy,” Appl.
Phys. Lett. 43, 941–943 (1983).
93. I. Kenworthy, P. Capper, C. L. Jones, J. J. Gosney, and W. G. Coates,
“Determination of acceptor ionisation energies in CdxHg1–xTe,” Semicon. Sci.
Technol. 5, 854–860 (1990).
94. E. Finkman and Y. Nemirovsky, “Electrical properties of shallow levels in p–type
HgCdTe,” J. Appl. Phys. 59, 1205–1211 (1986).
95. R. Fastow and Y. Nemirovsky, “Transient and steady-state excess carrier lifetimes in
p–type HgCdTe,” Appl. Phys. Lett. 55, 1882–1884 (1989).
96. R. Fastow and Y. Nemirovsky, “The excess carrier lifetime in vacancy- and
impurity-doped HgCdTe,” J. Vac. Sci. Technol. A8, 1245–1250 (1990).
97. Y. Nemirovsky, R. Fastow, M. Meyassed, and A. Unikovsky, “Trapping effects in
HgCdTe,” J. Vac. Sci. Technol. B9, 1829–1839 (1991).
98. S. G. Dasan-Zade, E. A. Salkov, and G. A. Shepelski, “Photoelectric phenomena in
n–type CdxHg1–xTe under uniaxial deformation,” Fiz. Tekh. Poluprov. 17, 1913–
1917 (1983).
99. A. V. Lyubchenko, K. A. Myslivets, and Y. M. Olikh, “Recombination of carriers
via acceptor levels of eigen defects in n– CdxHg1–xTe crystals after ultrasound
treatment,” Fiz. Tekh. Poluprov. 24, 171–174 (1990).
100. S. H. Shin, J. M. Arias, D. D. Edwall, M. Zandian, J. G. Pasko, and R. E. DeWames,
“Dislocation reduction in HgCdTe on GaAs and Si,” J. Vac. Sci. Technol. B10,
1492–1498 (1992).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 125

101. A. J. Syllaios and L. Colombo, “The influence of microstructure on the impedance

characteristics of HgCdTe MIS devices,” Tech. Dig. IEDM, 137 (1982).
102. M. J. Małachowski, J. Piotrowski, and A. Rogalski, “Influence of dislocations on the
performance of 3 to 5 µm HgCdTe graded gap photoresistors,” Phys. Stat. Sol. (a)
113, 467–476 (1989).
103. D. Chandra, J. H. Tregilgas, and M. W. Goodwin, “Dislocation density variations in
HgCdTe films grown by dipping liquid phase epitaxy: effects on metal-insulator-
semiconductor properties,” J. Vac. Sci. Technol. B9, 1852–1857 (1991).
104. S. M. Johnson, D. R. Rhiger, J. P. Rosbeck, J. M. Peterson, S. M. Taylor, and M. E.
Boyd, “Effect of dislocations on the electrical and optical properties of long-
wavelength infrared HgCdTe photovoltaic detectors,” J. Vac. Sci. Technol. B10,
1499–1506 (1992).
105. R. Dornhaus, G. Nimtz, and B. Schlicht, Narrow-Gap Semiconductors, Springer-
Verlag, Berlin, 1983.
106. O. Ziep, D. Genzow, M. Mocker and K. H. Herrmann, “Nonradiative and radiative
recombination in lead chalcogenides,” Phys. Stat. Sol. (b) 99, 129–138 (1980).
107. O. Ziep, “Estimation of second-order Auger recombination in lead chalcogenides,”
Phys. Stat. Sol. (b) 115, 161–170 (1983).
108. D. A. Cammack, A. V. Nurmikko, G. W. Pratt, and J. R. Lowney, “Enhanced
interband recombination in Pb1–xSnxTe,” J. Appl. Phys. 46, 3965–3969 (1975).
109. D. M. Gureev, O. I. Davarashvili, I. I. Zasavitskii, B. N. Matsonashvili, and A. R.
Shotov, “Radiative recombination in epitaxial Pb1–xSnxSe (0 ≤ x ≤ 0.4) layers,” Fiz.
Tekh. Poluprov. 13, 1752–1755 (1979).
110. K. Lischka and W. Huber, “Carrier recombination and deep levels in PbTe,” Solid-
State Electr. 21, 1509–1512 (1978).
111. K. H. Harrmann, “Recombination in small-gap Pb1–xSnxTe,” Solid-State Electr. 21,
1487–1491 (1978).
112. I. Melngailis and T. C. Harman, “Single-crystal lead-tin chalcogenides,” in
Semiconductors and Semimetals, Vol. 5, pp. 111–174, edited by R. K. Willardson
and A. C. Berr, Academic Press, New York (1970).
113. P. R. Emtage, “Auger recombination and junction resistance in lead-tin telluride,” J.
Appl. Phys. 47, 2565–2568 (1976).
114. K. Lischka, W. Huber, and H. Heinrich, “Experimental determination of the
minority carrier lifetime in PbTe p–n junctions,” Solid State Commun. 20, 929–
931(1976).
115. K. Lischka and W. Huber, “Auger recombination in PbTe,” J. Appl. Phys. 48, 2632–
2633 (1977).
116. T. X. Hoai and K. H. Herrmann, “Recombination in Pb0.82Sn0.18Te at high levels of
optical excitation,” Phys. Stat. Sol. (b) 83, 465–470 (1977).
117. O. Ziep, M. Mocker, D. Genzow, and K.H. Herrmann, “Auger recombination in
PbSnTe-like semiconductors,” Phys. Stat. Sol. (b) 90, 197–205 (1978).
118. O. Ziep and M. Mocker, “A new approach to Auger recombination,” Phys. Stat. Sol.
(b) 98, 133–142 (1980).
119. R. Rosman and A. Katzir, “Lifetime calculations for Auger recombination in lead-
tin-telluride,” IEEE J. Quant. Electr. QE18, 814–817 (1982).
120. H. Zogg, W. Vogt, and W. Baumgartner, “Carrier recombination in single crystal
PbSe,” Solid-State Electr. 25, 1147–1155 (1982).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 126 Chapter 3

121. A. Shahar, M. Oron, and A. Zussman, “Minority-carrier diffusion-length

measurement and lifetime in Pb0.8Sn0.2Te and indium-doped PbTe liquid phase
layers,” J. Appl. Phys. 54, 2477–2482 (1983).
122. M. Mocker and O. Ziep, “Intrinsic recombination in dependence on doping
concentration and excitation level,” Phys. Stat. Sol. (b) 115, 415–425 (1983).
123. O. Ziep, M. Mocker and D. Genzow, “Theorie der Band-Band-Rekombination in
Bleichalkogeniden,” Wissensch. Zeitschr. Humboldt-Univ. Berlin, Math.-Nataruiss.
Reihe XXX, 81–97 (1981).
124. A.V. Dmitriev, “Calculation of the Auger lifetime of degenerate carriers in the
many-valley narrow gap semiconductors,” Solid State Commun. 74, 577–581
(1990).
125. S. D. Beneslavskii and A. V. Dmitriev, “Auger transitions in narrow-gap
semiconductors with Lax band structures,” Solid State Commun. 39, 811–814
(1981).
126. P. Berndt, D. Genzow, and K. H. Herrmann, “Recombination analysis in 10 µm
Pb1–xSnxTe,” Phys. Stat. Sol. (a) 38, 497–503 (1976).
127. B. Schlicht, R. Dornhaus, G. Nimtz, L.D. Haas, and T. Jakobus, “Lifetime
measurements in PbTe and PbSnTe,” Solid-State Electr. 21, 1481–1485 (1981).
128. K. Weiser, E. Ribak, A. Klein, and M. Ainhorn, “Recombination of photocarriers in
lead-tin telluride,” Infrared Phys. 21, 149-154 (1981).
129. D. Genzow, M. Mocker, and E. Normantas, “The influence of uniaxial strain on the
band to band recombination in Pb1–xSnxTe,” Phys. Stat. Sol. (b) 135, 261–270
(1986).
130. S. Espevik, C. Wu, and R.H. Bube, “Mechanism of photoconductivity in chemically
deposited lead sulfide layers,” J. Appl. Phys. 42, 3513–3529 (1971).
131. T. H. Johnson, “Lead salt detectors and arrays. PbS and PbSe,” Proc. SPIE 443, 60–
94 (1984).
132. L. N. Neustroev and V. V. Osipov, “Physical properties of photosensitive
polycrystalline lead chalcogenide films,” Mikroelektronika 17, 399–416 (1988).
133. J. Vaitkus, M. Petrauskas, R. Tomasiunas, and R. Masteika, “Photoconductivity of
highly excited AIVBVI thin films,” Appl. Phys. A54, 553–555 (1992).
134. J. Vaitkus, R. Tomasiunas, K. Tumkevicius, M. Petrauskas, and R. Masteika,
“Picosecond photoconductivity of polycrystalline PbTe films,” Fiz. Tekh. Poluprov.
24, 1919–1922 (1990).
135. T. H. Myers, J. R. Meyer, and C. A. Hoffman, “The III superlattices for long-
wavelength infrared detectors: The HgTe/CdTe system,” in Semiconductor
Quantum Wells and Superlattices for Long-Wavelength Infrared Detectors, edited
by M. O. Manasreh, pp. 207–259, Artech House, Boston (1993).
136. J. R. Meyer, C. A. Hoffman, and F. J. Bartoli, “Quantum wells and superlattices,” in
Narrow-Gap II-VI Compounds for Optoelectronic and Elektromagnetic
Applications, pp. 363–400, edited by P. Capper, Chapman & Hall, London (1997).
137. Y. Jiang, M. C. Teich, and W. I. Wang, “Carrier lifetimes and threshold currents in
HgCdTe double heterostructure and multiquantum-wave lasers,” J. Appl. Phys. 69,
6869–6875 (1991).
138. C. A. Hoffman, J. R. Meyer, E. R. Youngdale, J. R. Lindle, F. J. Bartoli, K. A.
Harris, J. W. Cook, and J. F. Schetzina, “Photo-Hall investigation of p–type
HgTe/CdTe superlattices,” Phys. Rev. B37, 6933–6940 (1988).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]
 Generation-Recombination Processes: Carrier Lifetime 127

139. M. W. Goodwin, M. A. Kinch, and R. J. Koestner, “Metal–insulator–semiconductor

properties of HgTe/CdTe superlattices,” J. Vac. Sci. Technol. A6, 2685–2692
(1988).
140. K. A. Harris, R. W. Yanka, L. M. Mohnkern, A. R. Reisinger, T. H. Myers, Z.
Yang, Z. Yu, S. Hwang, and J. F. Schetzina, “Properties of (211)B HgTe–CdTe
superlattices grown by photon assisted molecular-beam epitaxy,” J. Vac. Sci.
Technol. B10, 1574–1581 (1992).
141. A. R. Reisinger, K. A. Harris, T. H. Myers, R. W. Yanka, L. M. Mohnkern, and C.
A. Hoffman, “Carrier lifetime in HgTe/CdTe superlattices grown by photoassisted
molecular beam epitaxy,” Appl. Phys. Lett. 61, 699–701 (1992).
142. S. H. Shin, J. M. Arias, M. Zandian, J. G. Pasko, J. Bajaj, and R. E. DeWames,
“Minority carrier lifetime and diffusion length in HgTe/CdTe superlattices by
molecular beam epitaxy,” Appl. Phys. Lett. 61, 1196–1198 (1992).
143. C. H. Grein, P. M. Young, and H. Ehrenreich, “Minority carrier lifetimes in ideal
InGaSb/InAs superlattice,” Appl. Phys. Lett. 61, 2905–2907 (1992).
144. C. H. Grein, P. M. Young, M. E. Flatté, and H. Ehrenreich, “Long wavelength
InAs/InGaSb infrared detectors: Optimization of carrier lifetimes,” J. Appl. Phys.
78, 7143–7152 (1995).
145. E. R. Youngdale, J. R. Meyer, C. A. Hoffman, F. J. Bartoli, C. H. Grein, P. M.
Young, H. Ehrenreich, R. H. Miles, and D. H. Chow, “Auger lifetime enhancement
in InAs–Ga1–xInxSb superlattices,” Appl. Phys. Lett. 64, 3160–3162 (1994).
146. E. R. Youngdale, J. R. Meyer, C. A. Hoffman, F. J. Bartoli, R. H. Miles, and D. H.
Chow, “Recombination lifetime in InAs–Ga1–xInxSb superlattices,” J. Vac. Sci.
Technol. B12, 1129–1132 (1994).
147. J.R. Meyer, C.L. Felix, W.W. Bewley, I. Vurgaftman, E.H. Aifer, L.J. Olafsen, J.R.
Lindle, C.A. Hoffman, M.J. Yang, B.R. Bennett, B.V. Shanabrook, H. Lee, C.H.
Lin, S.S. Pei, and R.H. Miles, “Auger coefficients in type-II InAs/Ga1–xInxSb
quantum wells,” Appl. Phys. Lett. 73, 2857–2859 (1998).

Downloaded From: [Link] on 25 May 2022

Terms of Use: [Link]