Sunteți pe pagina 1din 14

1

Mercury Cadmium Telluride (HgCdTe) based Photovoltaic Detectors


Submitted by: Muhammad Shoaib and Malik M, Haris Amir,
Submitted to: Dr Adnan Masood
Department of Electrical Engineering, College of Electrical & Mechanical Engineering,
National University of Sciences & Technology, Rawalpindi

1.

Introduction
There are several semiconductor devices that can be used to detect the presence of

photons. These devices, known as photodetectors, are used to convert optical signals to electrical
signals. When excess electrons and holes are generated in a semiconductor due to illumination
by photons, the conductivity of the semiconductor material is increased. This principle is the
basis of operation of photodetectors[1].
The main types of photodetectors are thermal and photonic detectors. The thermal effects
of the incident radiation can be followed through many temperature dependent phenomena.
Bolometers and micro bolometers are based on changes in resistance. Thermocouples and
thermopiles use the thermoelectric effect. Golay cells follow thermal expansion. In IR,
spectrometers the pyroelectric detectors are the most widespread.
The response time and sensitivity of photonic detectors can be much higher, but usually
these have to be cooled to cut thermal noise. The materials in these are semiconductors with
narrow band gaps. Incident IR photons can cause electronic excitations. In photoconductive
detectors, the resistivity of the detector element is monitored. Photovoltaic detectors contain a
p-n junction on which photoelectric current appears upon illumination.
HgCdTe[2] is a pseudo-binary alloy semiconductor that crystallizes in a zinc-blende
structure. Because of its band gap tunalbility with Hg1xCdxTe has evolved to become the most
important/versatile material for detector applications over the entire IR range. As the Cd
composition increases, the energy gap for Hg1xCdxTe gradually increases from a negative value

for HgTe to a positive value for CdTe. The band gap energy tunability results in IR detector
applications that span the short wavelength IR (SWIR: 13 m), middle wavelength (MWIR: 3
5 m), long wavelength (LWIR: 814 m) and very long wavelength (VLWIR:1430 m)
ranges. At present HgCdTe is the most widely used variable gap semiconductor for IR
photodetectors. This paper will focus on photonic detectors based on Mercury Cadmium
Telluride (HgCdTe, also known as MERCATEL in various texts)[3].
This paper summarizes the fundamental properties of HgCdTe semiconductor, and relates
these material properties, which have successful applications as an IR detector material. The
intent is to concentrate on mainly on photovoltaic detectors.
2.

Historical Perspective[6]
The 1959 publication by Lawson and coworkers triggered the development of variable

bandgap Hg1xCdxTe (HgCdTe) alloys providing an unprecedented degree of freedom in infrared


detector design. Early recognition of the importance of this technology led to intensive
development of HgCdTe detectors in a number of countries including England, France,
Germany, Poland, the former Soviet Union, and the United States. However, little has been
written about the early development years; as work going on in the United States was classified
until the late 1960s. Figure 1 gives approximate dates of significant development efforts for
HgCdTe IR detectors.

Figure 1. Development of HgCdTe detectors.

3.

Fundamental HgCdTe Properties


HgCdTe ternary alloy is a nearly ideal IR detector material system. Its position is

conditioned by three key features

tailorable energy band gap over the 130 m range,

large optical coefficients that enable high quantum efficiency and

favorable inherent recombination mechanisms that lead to high operating


temperature.

Table 1 (attached at Annexure A) summarizes various material properties of Hg1xCdxTe.


Table 2 (attached as Annexure B) compares important parameters of HgCdTe with other narrow
gap semiconductors used in IR detector fabrication.
3.1

Energy Band Diagram


The electrical and optical properties of Hg1xCdxTe are determined by the energy gap

structure in the vicinity of the -point of the Brillouin zone. The energy gap of this compound at
4.2 K ranges from 0.300 eV for semi-metallic HgTe, goes through zero at about x = 0.15 and
opens up to 1.648 eV for CdTe. The band structure of Hg1xCdxTe is shown in Figure 2.

Figure 2. The bandgap structure of Hg1xCdxTe near the -point for three different values of
the Eg.

3.2

Intrinsic Carrier Concentrations and Effective Masses of Electrons and Holes


The intrinsic carrier concentration ni (cm-3) of HgCdTe is given by

ni = (5.585 3.82x + 0.001753T 0.001364xT) (1014Eg3/4T3/2 exp (Eg/2KBT)

(Eq. 1)

The effective mass of heavy hole mhh is high; the measured values range between
0.3 and .7mo. The value of mhh= 0.55 mo is frequently used in modeling of IR detectors. The
electron me and light hole mlh effective masses in the narrow gap mercury compounds are close
and they can be approximated by me/m 0.071 Eg, where Eg is in electronvolts.
3.3

Electron Affinity
The expression for HgCdTe electron affinity (eV) is can be approximated by the

following expression
(x,T) = 4.32 0.813 [Eg(x,T) 0.083]
3.4

(Eq. 2)

Mobilities
Due to small effective masses the electron mobilities in HgCdTe are remarkably high,

while heavy-hole mobilities are two orders of magnitude lower. The transport properties of holes
are less studied than those of electrons mainly because the contribution of holes to the electrical
conduction is relatively small due to their low mobility. Very often in transport measurements,
the electron contribution predominates even in p-type materials unless the electron density is
sufficiently low. The electron mobility in Hg1xCdxTe (expressed in cm2 V1 s1), in composition
range 0.2 x 0.6 and temperature range T >50 K, can be approximated as:

The empirical formula (valid for 0.18 x 0.25) for the variation of e with x at 300 K is:
e = 104(8.754x 1.044)1 in cm2(V-s)1

(Eq. 3)

The hole mobilities at room temperature range from 40 to 80 cm2(V-s)1, and the
temperature dependence is relatively weak. At 77 K, hole mobility is by one order of magnitude
higher. For modeling IR photodetectors, the hole mobility is usually calculated assuming that the
electron-to-hole mobility ratio b = e/h is constant and equal to 100.

3.5

Optical Properties
Optical absorption coefficient data for several alloy compositions of Hg1xCdxTe versus

wavelength are plotted in Figure 3. We can see that the absorption strength generally decreases
as the gap becomes smaller.

Figure 3. Optical absorption coefficient plotted versus wavelength

3.6

Lattice Constants
The use of CdTe for passivation is very promising since it has high resistivity, is lattice

matched, and is chemically compatible to Hg1xCdxTe. Barriers can be found in both the
conduction and valence band. The comparison of lattice constants of various materials is shown
in Figure 4.

Figure 4. Comparison of lattice constants of various semiconductors vs bandgap energy and


wavelength

3.7

GenerationRecombination Processes
The generation processes which compete against the recombination processes directly affect

the performance of photodetectors, setting up a steady-state concentration of carriers in


semiconductors subjected to thermal and optical excitation and, frequently determining the
kinetics of photo generated signals.
3.7.1 Radiative Processes
In HgCdTe, the internal radiative processes although of fundamental nature, do not limit the
ultimate performance of IR detectors.
3.7.2 Auger Processes
Auger mechanisms dominate generation and recombination processes in Hg1xCdxTe at near
room temperatures. The Auger generation and recombination rates are strongly dependent on
temperature via dependence of carrier concentration and intrinsic time on temperature.
Therefore, cooling is a natural and a very effective way to suppress Auger processes.
3.7.3 Carrier Lifetimes
The total carrier lifetime

total can be calculated from the independent lifetimes attributed to

each recombination mechanism. The sum of all contributions is given by


(Eq. 4)

4.

Operation of HgCdTe p-n Junction Photodiode[4]


The photovoltaic p-n junction photodiode is representative of high performance HgCdTe

IR detector technology. Figure 5 illustrates the basic operation of a simple p-n homojunction
detector under illuminated zero-bias and reverse bias conditions. Incident photons with energy
greater than the material bandgap (Eph>Eg) generate electron-hole pairs in the material on either
side of the p-n junction.
Carriers generated within a diffusion length of the space-charge region diffuse to the
electrical junction and are separated by the strong electric field. The electron and hole minority
carrier diffusion lengths Ln and Lp (cm) are given by

(Eq. 5)

where n and p are the electron and hole minority carrier recombination lifetimes(s), and Dn and
Dp are the electron and hole minority carrier diffusion coefficients (cm2s-1).

Figure 5. Schematic of structure and energy band diagram of a HgCdTe p-n homo-junction
photodetector device at zero-bias (inset) and under reverse bias and illumination.

Figure 6. Schematic representation of current-voltage characteristics of photodiode

The optically excited minority carriers that are injected into the opposite side of the
junction by the Efield become majority carriers. By this process, a photocurrent density Jph is
generated by the incident radiation, increasing the dark saturation current value as illustrated in
Figure 6.
The total current density Jtotal (A/cm2) in the p-n junction photodiode is given by[3]

Jtotal(Vbias) = Jdark(Vbias) -Jph()

(Eq.6)

where Jdark depends on Vbias, and Jph depends on the incident photon flux density (ph/cm2s). In a
simple photovoltaic detector with no gain, the photocurrent is equivalent to[3]
Iph = JphA = nqA
4.1

(Eq. 7)

Detector Structure
Most commonly used photovoltaic HgCdTe photodiodes are fabricated in a homojunction

configuration with a highly doped n+ cap layer in contact with a lightly doped, narrow bandgap
p-type absorber (base) layer, or in a heterojunction configuration with wide bandgap (denoted by
capitalized letter) P+ cap layer in contact with a lightly doped, narrow bandgap n-type absorber
layer, as shown in Figure 7.
In these structures, the absorber layer determines the dark current and photo current of
the device. The base layer doping in both structures must be below1016 cm-3 to avoid tunneling
current contribution, and the doping

Figure 7. Schematic of energy band diagrams of commonly used unbiased (a) n-on-p
homojunction and (b) P-on-n heterojunction photodiodes under backside illumination

values used are typically ~51015 cm-3 and ~51014 cm-3 for p-type and n-type layers,
respectively.

Figure 8. HgCdTe DLHJ structure on Si substrate for FPA integration

Double layer heterojunction (DLHJ) architectures, shown in Figure 8, have been


developed for high performance LWIR detectors due to low thermal leakage currents[5].
4.2

Current Contribution Mechanisms


The diffusion current Idiff (A) is described by the basic diode equation,

where
and npo and pno are the equilibrium electron and hole concentrations (cm-3) on the p- and n-sides
of the junction, respectively. These mechanisms, several of which are illustrated in Figure 9,
include generation-recombination (G-R) in the depletion region, band-to-band (BTB) tunneling,
trap-assisted tunneling (TAT), and surface recombination. The total dark current can be
expressed as the sum of all contributions

Materials with narrow bandgaps are especially susceptible to thermal G-R processes that
dominate the dark current at high temperatures. The dark current contribution to overall current
is illustrated in Figure 10.

10

Figure 9. Schematic of structure and energy band diagram of a HgCdTep-n junction DLPH
photodetector device under reverse bias and illumination.

Figure 10 - Dark current contribution mechanisms in reverse-biased p-n junction

Band-to-band tunneling occurs when electrons at a given energy level tunnel across the
bandgap from the valence band on one side of a junction to the conduction band on the other to
fill an unoccupied state at the same energy level. Band-to-band tunneling exhibits a strong
dependence on Eg and Efield, where the field strength depends on Vbias and the doping
concentration of the junction. Trap assisted tunneling is caused by carriers tunneling within the
depletion region via intermediate trap states created by impurities or defects. The tunneling
occurs in a two-step process involving one thermal transition from the valence band to the trap
state and a second transition from the trap state to the conduction band. Because the required
transition energies are smaller, trap-assisted tunneling occurs at lower fields than band-to-band

11

tunneling. The mechanism is heavily dependent on defect density Nt (cm-3) and the trap level
Etrap (eV) within the material bandgap.
Surface current contributions can create defects or trap states that affect detector
performance. Dangling bonds at the material surface can result in a large density of interface
states that generate current via the SRH mechanism. Moreover, the presence of a net surface
charge may invert or accumulate the bandedges near the surface creating current channels or
affect the position of the junction depletion regions which intersect the surface, facilitating
tunneling or field-induced G-R currents. Wide bandgap insulators typically used to neutralize the
material surface and reduce the density of surface states.
5.

HgCdTe Detectors Figures of Merit

5.1

Responsitivity
The responsivity of an infrared detector is defined as the ratio of the root mean square

(rms) value of the fundamental component of the electrical output signal of the detector to the
rms value of the fundamental component of the input radiation power. The units of responsivity
are volts per watt (V/W) or ampers per watt (amp/W).
5.2

Noise Equivalent Power


The noise equivalent power (NEP) is the incident power on the detector generating a

signal output equal to the rms noise output, or the NEP is the signal level that produces a signalto-noise ratio of 1. The NEP for HgCdTe detectors is about two orders of magnitude better than
any other uncooled device and only a factor of three worse than devices cooled to 80 K[6].
5.3

Detectivity
The detectivity is the reciprocal of NEP. The advantage in material preparation of

HgCdTe and detector technology have led to devices approaching theoretical limits of
responsivity and detectivity over wide ranges of temperature and background[6].
6.

Advantages of HgCdTe Detectors


The specific advantages of HgCdTe are the direct energy gap, ability to obtain both low

and high carrier concentrations, high mobility of electrons and low dielectric constant. The

12

extremely small change of lattice constant with composition makes it possible to grow high
quality layers and heterostructures. HgCdTe can be used for detectors operated at various modes,
and can be optimized for operation at the extremely wide range of the IR spectrum (130 m)
and at temperatures ranging from that of liquid helium to room temperature.
7.

Limitations of HgCdTe Detectors


Thermal G-R processes, namely intrinsic Auger generation, dominate the dark current

density in narrow bandgap materials, fundamentally limiting the detector performance at higher
temperatures. Thus, high performance detectors are typically operated at cryogenic temperatures
(77 K or lower). These cooling systems are extremely bulky, require high input power, and are
prohibitively costly to implement. Increasing the detector operating temperature to a range where
the cooling requirements can be met by TE coolers (193 K-253 K) is advantageous as these
systems are smaller, simpler, and less expensive to integrate
At the processing level, achieving stable, well-controlled, extrinsic p-type doping
required for p-n junction PV structures by ex situ ion implantation or in situ doping via molecular
beam epitaxy (MBE) poses a significant challenge for HgCdTe detector technology. Ex situ
doping techniques (implantation, in-diffusion) cannot easily provide the level of control over
doping profiles necessary for abrupt junction, multi-layer structures and can introduce material
damage, reducing the carrier mobilities and minority carrier lifetimes.
8.

Conclusion
HgCdTe is an important material for high performance IR photodetectors and continues

to outperform competing materials, achieving near theoretical performance limits. Properties


such as the high absorption coefficient, high carrier mobility, and lattice matched bandgap
tunability make HgCdTe an attractive option for IR detector applications. However, extrinsic and
intrinsic G-R processes such as SRH, radiative, and Auger mechanisms are major factors which
affect detector performance in narrow bandgap materials and must be taken into account.

13

Annexure A

Annexure B

14

References
[1]Donald

A. Neaman Semiconductor Physics and Devices 3rd Edition, 2003, McGrawHill,

ISBN 0-07-232107-5
[2]Antoni

Rogalski, HgCdTe Infrared Detector Material: History, Status And Outlook, Reports

on Progress in Physics, Pages from 2269 to 2331, Rep. Prog. Phys. 68 (2005) 22672336,
Published 22 August 2005, Institute Of Physics Publishing.
[3]

[Online] Available at https://en.wikipedia.org/wiki/Infrared_detector.

[4]

Anne M. Itsuno, Bandgap-Engineered HgCdTe Infrared DetectorStructures for Reduced

Cooling Requirements, 2012, The University of Michigan


[5]

P. Capper and J. Garland, editors. Mercury Cadmium Telluride, Growth, Properties and

Applications. Wiley, 2011.


[6] Antoni

Rogalski, Infrared Detectors, 2nd Edition, CRC Press, ISBN 978-1-4200-7671-4