Electrical and optical characteristics of InP Interband
Nanowire Infrared Photodetectors
Thesis submitted to the
Halmstad University, Sweden
For the degree of Masters of Science
In
Electrical Engineering
by
Mohammed Nurul Amin
&
Md. Obaidul Alam
School of Information Science, Computer and Electrical-
Engineering, Halmstad University
Electrical and optical characteristics of InP Interband
Nanowire Infrared Photodetectors
Master’s thesis in Electrical Engineering
School of Information Science, Computer and Electrical Engineering
Halmstad University
PO Box 823, S-301 18 Halmstad, Sweden
Preface
First of all we would like to show our deepest gratitude to our supervisor and mentor Dr. Lars Landin and we are also grateful to him for his precious time and cordial help in a number of ways. We are extremely glad that he has made it possible for us to work in this new photodetector technology project based on nanowires. We also owe our deepest gratitude to Professor Håkan Pettersson who proposed the project and invited us to carry out the project work in MPE-Lab at Halmstad University, Sweden. It was a real pleasure to work with those who made this thesis possible and to learn vast knowledge from such experienced persons. Finally, we would like to thank our parents and family members for their understanding and support to us throughout our lives. Without help of them, we would face many difficulties while doing this project.
Mohammed Nurul Amin, MD.Obaidul Alam
Halmstad University, October 2010
Abstract
We have investigated two devices for detection of radiation, typically in the infrared range, Photons are absorbed in an active region of semiconductor devices such that the absorption induces inter band electronic transitions and generate photo-excited charge carriers. A photocurrent is generated between the conducting contacts through the active region of the devices. We worked on infrared photodectors based on nanowires. This type of photodectors can be used for optical communication and to detect atmospheric pollution by absorption of the polluting molecules in the infrared region (0.7µm-1µm).
In this project we have used Fourier transform infrared spectroscopy to study and compared the photoresponse of two different types of interband nanowire infrared photodetectors 8samples 6080 and 6074). Fourier Transform infrared (FTIR) spectroscopy is a measurement technique that allows one to record infrared spectra in all wavelengths at then same time. The basic task was to compare and analyse the electrical and optical characteristics of these two detectors at different temperatures (78K-300K) corresponding to the wavelength.
Abbreviations:
NWs: Nanowires
CVD: Chemical Vapour Deposition
MOVPE: Metal Organic Vapor Phase epitaxy
VLS: Vapor Liquid Solid
TMI: Tri-Methylindium
TMG: Tri- Methylgallium
TBP: Triar-buthylphosphine
TBA: Teriar-buthylarsine
Epi-layer: Epitaxial layer
ITO: Indium Tin oxide
FTIR: Fourier Transform Infrared
W: Wavenumber
NIR: Near-Infrared
SWIR: Short Wave Infrared
MIWR: Medium Wave Infrared
LWIR: Long Wave Infrared
VLWIR: Very Long Wave Infrared
C
ONTENTS
:
1.
I
NTRODUCTION………..9
1.1 PROJECT OBJECTIVES………11 1.2 PROJECT GOALS……….11 1.3 PROJECT SCOPE……….112
.
B
ACKGROUND………...13
2.1THERMAL DETECTORS………...13 2.2PHOTON DETECTORS... ...14
2.2.1EXTERNAL PHOTOELECTRIC EFFECT………..15
2.2.2INTERNAL PHOTOELECTRIC EFFECT………...16
2.2.3THE P-N PHOTODIODE………...17
2.2.3.1PHOTOVOLTAIC EFFECT………..18
2.2.4THE P-I-N PHOTODIODE………...18
2.3 SOLID STATE PROPERTIES OF IR MATERIALS………...20
2.4ELECTRONIC BAND STRUCTURES OF THE SOLID MATERIALS………...20
3.
T
HE PERFORMANCE CHARACTERISTICS OF PHOTODETECTOR…………...24
3.1 RESPONSIVITY………..24
3.1.1 QUANTUM EFFICIENCY………25
3.1.2 INTERNAL GAIN………...25
3.1.3 RELATION AMONG THE IMPORTANT IR DETECTOR PARAMETERS………...25
3.2 DARK CURRENT……….26
3.3 NOISE……….27
3.4 NOISE EQUIVALENT POWER………...28
3.5 DETECTIVITY………...28
4.
N
ANOWIRE………..30
4.1 DEVELOPMENT OF NANOWIRES………...30
4.2APPLICATION OF NANOWIRES………...31
4.3GROWTH TECHNIQUE OF NANOWIRES…………...31
4.3.2METAL ORGANIC VAPOR PHASE EPITAXY(MOVPE)………..31
5. SAMPLES DESCRIPTION………...33
5.1 FOR THE SAMPLE 6080…………...33
5.2FOR THE SAMPLE 6074………....37
5.
E
XPERIMENTAL SETUP………...38
5.1FOURIER TRANSFORM INFRARED SPECTROSCOPY (FTIR)………..38
6.
R
ESULTS………...45
6.1CURRENT-VOLTAGE (I-V) CHARACTERISTICS……….45
6.1.1FOR THE SAMPLE 6080………...45
6.1.2FOR THE SAMPLE 6074………...52
6.2OPTICAL CHARACTERISTICS….……….56
6.2.1FOR THE SAMPLE 6080………...56
6.2.2FOR THE SAMPLE 6074………...60
7.
C
ONCLUSION………...64
R
EFERENCES………...66
Chapter 1
Introduction
An infrared detector is a photodetector that works based on the absorption of electromagnetic infrared radiation. Infrared radiation has longer wavelength (higher energy) than visible light. Electromagnetic waves are classified according to their wavelength or energy in the electromagnetic radiation spectrum as shown in the figure below:
Figure1: Electromagnetic spectrum
The energy of the infrared radiation is expressed by the following equation
)
(
24
.
1
)
(
m
hc
hv
eV
E
Eq-1
The range between 750 nm and 1mm wavelength is known as infrared radiation. It is usually sub-divided into 5 regions:
Near-infrared (0.7 µm – 1 µm) , Short-wave infrared (1 µm – 3 µm), Mid-wave infrared (3 µm – 5 µm), Long-wave infrared (8 µm -12 µm) and Very-long wave infrared (14 µm – 1000 µm).
Infrared radiation is emitted from all objects as a function of temperature in certain wavelength. It is invisible to the human eye and formed due to the vibration and rotation of molecules. The higher the temperature of an object, (higher temperature = more powerful motion) the more infrared energy is emitted. At absolute zero temperature, the infrared radiation energy is very low
INTRODUCTION Applications based on infrared radiation detection include surveillance, night vision, remote temperature sensing, spectroscopy astronomy, optical communication and medical diagnosis. Infrared sensors are can be made by III-V semiconductor elements as extrinsic or intrinsic infrared photodetectors, these types of detectors have been improved and widely used for many years. Recently a new kind of detector based on nanowires has been proposed [1,11]. In fiber optic-communication, NIR is most commonly used because of low dispersion and attenuation. The detector we investigated in this thesis project was fabricated to work in NIR.
INTRODUCTION
1.1 Project objectives:
There are several main objectives for this project:
To investigate and understand the electrical and optical characteristics of the both samples.
To look for advantages of the nanowire photodetectors compared to the conventional photodetectors.
To investigate the dark current level in darkness at different temperature and biases.
To compare the electrical and optical characteristics between the p-n and p-i-n photodiodes (photodetectors).
1.2 Project goals:
The main goal of the project was to analyze and compare the electrical and optical characteristics at a different temperature for two different infrared photodetectors based on nanowires. To analyze their electro-optical response in infrared radiation, we have used FTIR. We observed and investigated which sample showed the best spectral photoresponse for different bias voltages.1.3 Project scope:
Both photodetectors can be used in optical communication and the detection of the atmospheric pollution. In optical communication, nanowire photodetectors can be used in the near infrared region and they are supposed to give better photo response, which means better responsitivity and lower dark current than conventional photodetectors.INTRODUCTION
Chapter 2
Background
Infrared detection works because most everything in our known world gives off heat, by radiation in the infrared band. The first IR detector actually dates back to 1800 when prisms were used to detect this band. By 1900, objects could be detected as far as a quarter mile in distance. Nowadays, infrared radiation is widely used in many applications such as night vision, thermal cameras, remote temperature sensing, and medical diagnosis. It can be used to detect irregularities in machinery, ice on aircraft wings and faults in circuit boards. As it has a low energy, long wavelength, it is invisible to the human eye [1, 11]. A typical system for infrared radiation detection is configured as shown below.
Infrared sensors can be classified into two categories; one is thermal detectors that have no wavelength dependence and the other one is photodetectors also called quantum detectors that are wavelength dependent.
2.1 Thermal detectors
Thermal detectors operate are based on energy conversion. The radiation energy is converted to heat, which changes the temperature of the detector. This temperature change will result in a resistance change that ca be easily measured. Examples of such detectors are bolometers and microbolometers. When two miscellaneous metals are aligned together and then applied to different bias voltages, heat can be produced due to the thermoelectric effect. Thermocouples and thermopiles are devices based on this effect. The sensitivity of thermal detectors are most useful and appreciable for building sensors that are independent of the wavelength of the incident radiation [3,4]. However, most thermal detectors show low sensitivity and relatively slow response because of the time required for changing their temperature. The simplest representation of the thermal detector is shown in the figure below:
T
Figure 2: Thermal detector action
Infrared source Transmission system
Optical system Detector Signal processing
BACKGROUND
In a thermal detector, the photon flux is transformed into heat by absorption (see above fig). Now, at most, the absorbed signal power is Ps = Pinc, leading to a temperature increase ∆T
given by: T g T dt d H Pinc Eq-2
Where H is the thermal mass and g is the thermal conductance of the device and ∆T is the temperature rise due to incident radiative flux. So ∆T is expressed by:
2 20 2 1/2 ) (Gth Cth T Eq-3
Here ε is the detector emissivity; ω is the radiant frequency, s-1; Φ0 is the incident radiative
flux; Cth is the thermal capacity of the detector and Gth is the thermal coupling to the detector
surroundings
2.2 Photon detectors
Photo detectors work due to the photoelectric effect. When a semiconductor material absorbs photons, this causes excitation of charge carrier (electrons and holes) from low energy levels to higher energy levels and thus an electric current can flow under the influence of an electric field. The photoelectric process in the material can be external or internal. In the external process, all electrons escape from the material as free electrons due to the photon excitation on to the metal surface. On the other hand, all free carriers are produced inside the material by the internal process which leads to a photoconductivity signal. In the advanced detector technology, we can find more detectors using nanowire which provides high sensitivity, fast response and low cost of fabrication [5,7]. However, the dark current is raised by increased temperature. In order to reduce the dark current of the photodetector, it needs to be cooled to cryogenic temperatures. The simplest representation of the photon detector is shown in figure below:
Figure 3
:
Photo-electron emissionBACKGROUND
2.2.1 External photoelectric effect:
When photons, with higher energy than the work function, strike a metal or semiconductor surface electrons can be emitted from the surface. The emitted electrons are called “photoelectrons” and the corresponding current is called “photoelectric current”. This effect was first observed by Heinrich Hertz in 1887. Photodetectors that works due to the external photoelectric effect can operate in the near infrared as well as in the visible and ultraviolet regions. Devices based on the external photoelectric effect are phototubes and photomultipliers. These devices have a photo-emissive cathode that emits electrons as a result of the incoming photons. The electrons travel to another electrode called the anode that is maintained at higher potential. The result is an electric current proportional to the incident photon flux. The photomultiplier is basically a phototube with more metal or semiconductor surfaces called dynodes. When incident photons strike the photocathode material, electrons are produced due to the photoelectric effect. These photons are directed by the focusing electrode towards the dynods, where the numbers of electrons are increased by the process of secondary emission [6].
Figure 4: Phototube
BACKGROUND Sensors based on the photoelectric emission can measure small photon fluxes due to a high amplification factor of the order of 107. They are also very fast due to the high electric fields used for their operation. This type of sensor is commonly bulky and too expensive to be used in non-military applications.
In analyzing the photoelectric effect quantitatively using Einstein's method, the following equivalent equations are used:
0 2
2
1
mmv
hf
hf
Eq-4 Using physicists' symbols:hf EK Eq-5
Where h is Planck's constant, f0 is threshold frequency for the photoelectric effect to occur, φ
is the work function, or minimum energy requires removing electron from atomic binding, and Ek is maximum kinetic energy observed.
2.2.2 Internal photoelectric effect:
The internal photoelectric effect is based on the fact that electron can be excited from the valance band to the conduction band in the semiconductor material due to the absorption of photons. In the equilibrium condition, all charge carriers (electrons) occupies the valance band in the semiconductor. When the material is illuminated by light and absorbs the photon, an electron is kicked up from the valence band to the conduction band, leaving behind a free hole in the valence band. In the presence of an external electric field, the photocurrent flows through the circuit. Such kinds of devices are, for example p-n photodiodes, p-i-n photodiodes, heterostructure photodiodes, Schottky-Barrier photodiodes. On the basis of internal photoelectric effect, the measured current can be amplified due to internal gain mechanism [7].
Nowadays, many photodetectors operate based on the internal photoelectric effect because it has high responsivity and sensitivity. However, these kinds of photodetectors produce the dark current due to thermal excitation. To reduce the dark current it can be necessary to cool the detector by using thermo-electric cooling or cryogenic liquids.
BACKGROUND
2.2.3 The p-n Photodiode:
The p-n photodiodes are semiconductor light sensors with reversed bias. When the p-n junction in the semiconductor is illuminated by light with high photon energy in reverse bias condition, it excites an electron, and thereby a free electron and a free hole is created. In the presence of a high electric field, charge carriers will drift quickly in opposite directions in the depletion region. That means holes move to the left in the p-doped region and electrons to the right in the n-doped region until they are collected at the electrode, and hence generate a flow of current in an external circuit proportional to the incident power (see fig below). Photodiodes are often used for accurate measurement of light intensity in science and industry. They generally have a better, more linear response than photoconductors [4,5,13]. Basically, p-n photodiode is the same as p-n diode. It is formed by combining a p-doped and n-doped region in a direct band gap type III-V photosensitive material. Hence, the Fermi levels in the two regions are different. The figure below shows the basic construction and energy band diagram;
Figure 7: p-n photodiode under illumination with reverse bias
BACKGROUND
2.2.3.1 Photovoltaic effect:
When p-n photodiode is under illumination with zero bias, the flow of external circuit current called “photocurrent” is restricted and a voltage builds up. In the depletion region, due to the low electric field, the diode becomes forward biased and the dark current begins to flow across the junction in the direction opposite to the photocurrent. This action is responsible for the photovoltaic effect, which is the basis for solar cells – in fact; a solar cell is just a large area photodiode [2]. The photovoltaic effect is nowadays a hot research area where scientists try to improve solar energy systems,
Figure 9: Energy band of photovoltaic effect realized by a p-n junction
2.2.4 The p-i-n Photodiode:
We know that the most important part of the photodiode is the active region or depletion region, where the light is absorbed and the photon energy excite the electron and, hence, generate the external circuit current called photocurrent. For this reason, if we want to have a high photocurrent, we should make a photodiode with a wider depletion or active region. As a result we can trap more photons and get a higher photocurrent. In order to have a wider depletion region, we add an intrinsic layer (i layer) between the p and the n doped regions which in turn provide more efficient conversion of photons to excited charge carriers [13]. Furthermore, the p-i-n photodiodes have more advantages compared to p-n photodiodes, for the following reasons:
Its increased depletion region reduces the junction capacitance so the RC delay is also reduced.
Due to increase in the depletion region, the electric field and the drift velocity of the charge carriers also increases.
BACKGROUND Due to reducing the RC time delay, it will have a very high sensitivity and
fast response time.
Figure 10: Cross-section view of p-i-n photodiode under reverse biased
Figure 11: Energy band diagram of p-i-n photodiode under reverse biased
BACKGROUND
2.3 Solid state properties of the IR detector material:
Nowadays, semiconductor materials play an important role in optoelectronics due to their specific interaction with light. These interaction involves phenomena such as photoconductivity, light emission capability and various electro-optic and nonlinear effects. In large scale applications, photodetectors, light-emitting diodes, lasers, photovoltaic solar cells are included. One application of most interest is fiber-optical communication system where we need to convert electrical signals into optical ones in the transmitter, and transform light signal back to electrical signal in the receiver [21].
2.3.1 Electronic band structure of the solid materials:
Electronic band structure of solid state materials is a very important feature, and solid state electrical and optical properties are explained in terms of band structure. The figure below shows the band structure of solid-state materials:
Figure 12: Band structure of the solid
The valence band and conduction band can be seen as ground and excited states of an atom. In order for charge to flow, there must be available states in the conduction band and/or empty states in the valence band. Since the band gap of an insulator is very large, no electrons can flow because there is not enough energy for an electron to exceed the band transition. In a conductor, the two bands overlap so electrons flow very easily. However, the semiconductor lies between the insulator and the metal with a relatively small bandgap. Therefore, electrons can much more easily be elevated into the conduction band than in the case of the insulator. This will leave holes in the valence band, which are effectively positive charge carriers. When these holes are present in the material, charge can flow, and a current can be detected.
An important energy level in the band structure is the Fermi level. The position of the Fermi level is the most important factor to determine the electrical properties. Semiconductor materials have two kinds of band gap, one is direct band gap and the other is indirect band gap. The figure below shows the direct and indirect bandgap material energy-momentum diagram:
BACKGROUND
Figure 13: Energy-momentum diagram of the two semiconductor materials
In equilibrium, the charge carriers occupy their lowest energy states; electrons at the bottom of the conduction band, and holes at the top of the valence band. In silicon (Si), the electrons cannot shift from lowest energy state in the conduction band to the highest energy state in the valance band without change in momentum. Therefore, if a recombination is to result in the emission of a photon, which has little momentum, a quantum of lattice vibration (a phonon) must also be created to carry away the excess momentum. This is known as an “indirect process” and such semiconductors are known as “indirect bandgap semiconductors”. On the other hand, in GaAs, the electrons easily shift from the conduction band to the valance band without change in momentum. This is called “direct process” and such kinds of semiconductors are “direct bandgap semiconductors”. This material is used for optoelectronics and photonics devices because each photon gives one electron and no heat dissipation in the material due to carrier recombination [14,19].
The bandgap in a semiconductor photodiode determines the way the photons of different wavelengths are detected. A simple quantum mechanical model involving a direct bandgap states that the absorption coefficient has a simple dependence on photon energy
1 2 / 1
)
(
.
)
(
C
E
gcm
)
(
)
(
K
E
K
E
c
v
BACKGROUND
6
...
...
...
...
2
)
2
/
(
2 2 2 2
Eq
m
K
m
K
E
h e g
Here, C is the light velocity within the semiconductor, me is the effective mass of electron, mh
is the effective mass of hole, h is the reduced plank constant, K is the Boltzmann constant and Eg is the band gap energy in units of electron volts (eV). ). The detection can only begin when the photon energy exceeds the bandgap, e.g which in case of a bulk semiconductor is a fixed number. In figure 14, the bandgap values for some simple and binary compound semiconductors are shown:
Figure 14: Diagram of band gaps and lattice constants for semiconductors in the III-V-group
Alloys of binary compound semiconductors (that are ternary or quaternary semiconductors) provide some flexibility in terms of available bandgap. At the same time, a serious limitation comes from the necessity of a lattice-matched substrate for these compounds, since the lattice parameter generally changes with alloy composition. In addition, for most semiconductors the bandgap corresponds to at least visible or near IR light.
BACKGROUND
Chapter 3
The performance characteristics of photodetector
We can describe every system by measuring its efficiency. Efficiency means the ratio between the output signal and the input signal of the system. The incident power to the photodetector is the input signal, and the output signal is the amount of photocurrent in the external circuit. In order to have an indication of the performance of IR photodetectors, there are a number of performance criteria which are used. The responsivity, dark current and detectivity (D*) are commonly used for a single pixel device. On the other hand, the noise equivalent temperature difference (NETD) is more commonly used for detector arrays [16, 17].
3.1 Responsivity or photo sensitivity:
The responsivity of the detector is to determine the detector response, when it is irradiated with a certain power of radiation. Therefore, responsivity is the output voltage or output current per watt (measure in V/W or A/W) of incident energy without noise consideration.
R=
PA S
(V/W) Eq-7 Where S is the signal output, P is the incident power and A is detector active area.
For a photoconductive mode detector, output signal as a photocurrent is usually extracted from the detector. The photosensitivity is,
R=
PA IPh
(A/W) Eq-8
Where, IPh is the photocurrent, when light is entered in to the detector at a given wavelength. For a photovoltaic mode detector, output signals are extracted as a voltage, so the responsivity is, R= PA VO (V/W) Eq-9
Where, VO is the changing voltage in a detector, when it is to an under illumination. The detector responsivity also can be measured in terms of two important parameters, quantum efficiency and photoconductive or internal gain, which are explained in the following section.
THE PERFORMANCE CHARACTERISTICS OF PHOTODETECTOR
3.1.1Quantum efficiency:
For the photodetector, quantum efficiency is defined as the ratio of the number of photoelectrons transferred from the detector to the number of incident photons on to the detector, i.e. the quantum efficiency is defined by:
= Number of photoelectrons transferred /Number of incident photons
The value of for a detector is also determined by the conduction and light absorption properties of the material and is related to the absorption quantum efficiency. Under illumination, light is absorbed by the material and the intensity I of the light decreases as
I
I
0e
L, where I0 is the initial intensity of the light, is the absorption coefficient and L is the thickness of the material. So the absorption quantum efficiency is:L L a
e
I
e
I
I
1
0 0 0 Eq-10This is useful when comparing measurement results with theoretically calculated values of the quantum efficiency and the absorption coefficient, respectively.
A good photodetector should have large values for , approximately 1. For this reason, all excited electrons are liberated with high energy due to the photon absorption in the material. If is low, that means that the incident photons are not absorbed sufficiently in the detector material..
3.1.2 Internal gain:
Generally, gain is defined by the ratio between the number of free excited electrons and the number of absorbed photons in the detector, when the detector is under illumination. However, internal or photoconductive gain is given by the ratio of the free carrier lifetime τ to the transit time between theT electrodes. Therefore, in the interband transition like p-n photodiode, the gain is almost one. This value shows how well the generated electron-hole pairs are used to generate a current response in a photodetector. An infrared photodetector with internal gain higher than one is the avalanche photodiode other example are photomultiplier tubes and the phototransistors. Value shows how well the generated electron-hole pairs are used to generate the curre
3.1.3 Relation among the infrared detector parameters:
The photocurrent is the part of the current that is caused by the light. When light enters into the detector at a certain wavelength, the photocurrent (IPh ) is expressed by the following equation
THE PERFORMANCE CHARACTERISTICS OF PHOTODETECTOR M hc qPA M c h PA q IPh Eq-11 Where, is the quantum efficiency, P is the incident energy, A is the effective area of detector, is the light wavelength, M is the internal gain,c is the light velocity and h is the planck constant.
However, photocurrent measurement is not the only way to classify the detector. To get better classification, you can consider the responsivity of the detector. The responsivity R is given by: R= M PA IPh = hc M q M hc N I P Ph * (A/W) Eq-12
Instead of incident energy, we use power density P [W/cm2].
The magnitude of the responsivity depends on the gain and the quantum efficiency, that means detector performance, depends on both high absorption and a long lifetime of the excited charge carriers.
3.2 Dark current:
Dark current is the relatively small electric current that flows through a reverse biased phpotodetector due to a high electric field even if no photo-excitation occurs. It is also called the “reverse bias leakage current”. Higher reverse bias voltages result in higher dark currents The dark current is raised with increasing temperature. In zero bias, dark current is not present in the detector.
In photodiodes, due to high electric field at different temperatures, the dark current can be produced by the following different contribution factors:
Thermally generated diffusion current from both sides of the junction; Thermal generation through band gap states in the junction depletion region; Generation from the diode surfaces;
Tunnelling in the depletion region via band gap states;
Direct tunnelling between the valence and conduction bands in the depletion region. In general, dark current denoted by
i
0 , it can be modelled by a Poisson distribution with noise due to dark current,
THE PERFORMANCE CHARACTERISTICS OF PHOTODETECTOR 0 i
=2 e
i
0 vWhere, is the bandwidth of the photodetector. The dark current is varies several to v
hundred of nanometers in the basis of photodetector design.
Dark current in interband nanowire IR photodetector: in equilibrium (at room temperature) with no bias, the electrons occupy lower energy states in the conduction band and the holes upper energy state in the valance band. According to Fermi-Dirac distribution, if temperature is increased, the electrons can occupy their higher energy states. This is called “thermal emission”. The dark current is consequently increased with increasing temperature due to thermal emission. The dark current in the interband nanowires photodector can be minimized by cooling the detector.
3.3 Noise:
In order to achieve good performance of the photodetectors, you have to make sure that it has present high bandwidth, high responsivity, low noise and high saturation power. Most objects which are capable of allowing the flow of electrical current will exhibit noise. This occurs as some electrons will have a random motion, causing fluctuating voltage and currents. Noise is the most important factor for the photodetector. Usually various kind of noises are exhibited when either flowing charge carriers or temperature rises in the photodetector. Now we are going to describe the most important noises below:
Thermal noise: this thermal noise exists in any resistive material and other material that has shown black body curves and possesses thermal energy regardless of any applied voltage due to the random motion or thermal agitation of the charge carriers. This was discovered by Johnson in 1927 and explained by Nyquist. That is why it is also called Johnson noise or Nyquist noise.
Thermal noise can be measured when a voltage source
v
th is connected in series with anideal resistor. As well as
v
th has a Gaussian distribution with a mean value of zero. In thiscase, v
k
BTR
v
th 4
2
Eq-13Where
vth is the standard deviation of the voltage kB is the Boltzmann’s constant in joulesper Kelvin and T is the resistor’s absolute temperature in Kelvins. It can also measure when a current source
i
th in parallel with R:R T kB ith 4 2 Eq-14 v
Shot noise: Photons are emitted randomly from the light source and are not constant, but exhibit detectable statistical fluctuations. During biasing condition,
THE PERFORMANCE CHARACTERISTICS OF PHOTODETECTOR
photodetector exhibits the photon shot noise, when the charge carriers (mainly electrons) are fluctuated by its nature of random motion and in the presence of an electric field.
This is especially predominant in semiconductor diodes that rely on p-n junctions. Shot noise is given by
I
2noise=2
q
2
v
Eq-15 Where q is the charge of an electron, is the photon flux, is the band width. v
Generation-recombination noise: in semiconductor based photon detectors, the numbers of excited carriers (electrons) is fluctuating due to the generation and recombination process, and these results in a generation-recombination noise. In extrinsic detectors, this happens via the doping level while in intrinsic detectors,it is due to the carrier fluctuations in the interband transition.
Flicker noise (1/f-noise): this is exhibited in the detector due to the low frequency in which the noise amplitude declines linearly with frequency (1/f region). It is often characterized by the corner frequency and is caused mainly by traps associated with contamination and crystal defects. In order to reduce it, the high quality monocrystalline materials will be used.
3.4 Noise equivalent power (NEP):
this is defined as the signal power or incident light power required to obtain signal to noise ratio to be unity. It can be expressed by the following equation:. NEP=
R
T
k
i
i
q
q
hc
v
P
B s N S opt4
2
0 1
Eq-16Where, v is the bandwidth, i and S i are signal and dark current respectively. The smallest 0
value of the NEP, the better is the detector.
3.5 Detectivity:
this is one of the most important performance parameter for the photodetector. It is used to characterize normalized signal to ratio performance of a detector and it can be defined as;
D
= n I v A R =A
v
i
ndarki
nBhc
M
q
, 2 , 2/
Eq-17THE PERFORMANCE CHARACTERISTICS OF PHOTODETECTOR
Where, R is the responsivity, is the bandwidth, M is the gain, is the quantum v
efficiency
i
n,dark and thei
n,Bis dark current noise and the detector noise respectively.The detectivity is also defined as equal to the reciprocal of noise equivalent power. i.e.
D
=NEP v A
, where A is area of the photosensitive region of the detector.
Chapter-4
Nanowire
A nanowire is a very thin wire in which constrained diameter is a few tens of nanometers (10−9 m) or less and the length in unconstrained. This shows length to width ratio of 100 or more, which makes it a one dimensional material that has different and interesting properties compared to the bulk and 2-D materials. It is basically either a single quantum wire or array of quantum wire where quantum mechanical effects are exhibited. The final nanowire is likely to be a chain of single atoms. Nanowire technology is being investigated for faster and smaller electronic devices. There are many types of nanowires, semiconducting (InP, Si), metallic (Pt, Au), and insulating (SiO2) etc. The material of nanowires is varied with respect
to their application, for example, InP nanowires are used in photoconductors and LED’s and, SnO2 nanowires are used in Lithium Ion batteries [20].
4.1 Development of Nanowire:
Nanowires can be developed through many processes such as lithography, solution chemistry, and extruded or drawn. Vapor-Liquid-Solid (VLS) and template growth processes are well known. In the VLS and the template growth processes, nanowires are fabricated up atom-by-atom.
Vapour-Liquid-Solid (VLS) processes: Most of the nanowires fabrications are done by Vapour-Liquid-Solid (VLS) processes. In this process a liquid catalyst is used to figure out components from a vapour phase and place them in to a solid phase. Silicon nanowires are produced through VLS processes. A substrate is covered with gold (Au) nano-particles after that heated up until the gold melts and forms tiny droplets. A silicon vapour, silane or SiH4, is passed above the substrate. At the vapour-gold interface silicon atoms are exposed from the silane and absorbed into the nanoparticles, then silicon precipitates out of the droplet onto the substrate. The silicon atoms build up rapidly and rise up the catalyst from the substrate up to a vertical wire. The diameter of the growth is directly related to the diameter of the catalyst droplet.
2
NANOWIRE Template growth process: another often used method of nanowire production is template growth process. This means that nanowires are developed on top or within another nano-scale structure. Generally, two template growth methods are used such as, injection of a molten phase under pressure and electrochemical deposition. In template growth, anodic alumina is used as a substrate material since it is comparatively inert and its pore distribution can be controlled, thus the size of nanowires can be controlled. Vapour-Liquid-Solid (VLS), template growth process are not exclusive but it is rear. This is to combine the two processes in which the catalytic nanoparticles can be established into a porous template. These combine fabrication techniques are remarkable for reducing the requirement to manipulate individual nanowires.
4.2 Application of Nanowire:
There is many successful uses of nanowires such as memory devices and transistors. The NASA Ames Research Center and the University of Southern California have produced a memory device that they expect will be capable of storing 40 gigabits per square centimetre. Nanowire can be used to produce nanoscale sensors for different purposes like the detection of cancer in the biomedical field. A rising field of nanowire techniques is for future uses in various sensors, electrical, thermo electrical, chemical, sanitation, home and business construction, more efficient solar cells and almost invisible computer chips.
4.3 Growth technique of Nanowire:
4.3.1Metal Organic Vapor Phase Epitaxy (MOVPE):
Metal organic vapour phase epitaxy (MOVPE) is mainly a chemical vapour deposition method. Metalorganic precursor vapours give the basic atoms of the layer in metal organic vapour phase epitaxy (MOVPE). In this process, trimethylindium (TMI), trimethylgallium (TMG), tertiarybutylphosphine (TBP), and tertiarybutylarsine (TBA), are generally used for the growth of GaInAsP which is a III-V semiconductor. A schematic diagram of the MOVPE system is shown bellow in Fig 16.
NANOWIRE
Figure 16: Schematic diagram of the MOVPE system
There are some gas lines and a flat quartz glass reactor in this system. Hydrogen gas is added through bubblers which consist of metalorganic precursors, hydrogen or nitrogen carrier gas bringing the precursors vapours into the reactor. A mass flow controller is used to control the gas flows. The bubblers are placed in a temperature controlled pool to get a known vapour pressure for every metalorganic precursor. The substrate is placed in the reactor on graphite which is heated up by a halogen lamp. A thermocouple is used to control the temperature. The thermocouple temperature is a few degrees higher than the substrate surface. Two dimensional III-V group semiconductor layers develop at about 650°C temperature. Nitride needs higher temperature. Mass transport from the vapour phase to the substrate surface, precursor molecule decomposition and atom adsorption, atom diffusion on the substrate surface, and atom desorption from the substrate surface are some main processes in MOVPE growth. MOVPE for Vapour liquid-solid growth of InP with metal catalyst particles is done at lower than 500°C temperature. The decomposition temperature is very important for nanowire fabrication. At low temperatures the metalorganic molecules decompose moderately, TMI and TBP molecules are decomposed at respective temperatures 325°C and 475°C. Catalyst-free InP nanowires typically grow at 350°C and TMI decomposition efficiency is greatly dependent on temperature. At 350°C decomposition does not take place for almost all TBP molecules. For expected chemical catalytic action of metal nanoparticle the temperature of decomposition of metalorganic molecules should be significantly lower.
Thermocouple TMI TMG TBP TBA Halogen Heater Quartz Reactor Substrate 2 H 2 N Bubblers in Temperature Stabilized Bathes Exhaust
NANOWIRE
5. Description of the samples:
During our project, we have investigated two different types of photodetectors based on nanowires. One was the basically a p-n photodiode (sample 6074). The other one was a p-i-n diod , a p-n diod with an intrinsic layer in between the n and p-layer (sample 6080).
As we can see the main difference is the presence and absence of the intrinsic layer which is placed in the middle of the sample structure. However, they are made by the same compound III-IV semiconductor materials, InP. The sample structures are shown in figures below.
5.1 Sample 6080:
This was a first sample that we measured and investigated in the laboratory. It was made by InP which is III-IV group semiconductor material. The substrate and the nanowires in this sample are made of the same material to obtain a greater lattice match. These types of photodetectors are supposed to have good photoresponse due to the carrier confinement. 19 10 17 10 * 5 18 10 * 5The upper part of the substrate layer, which has grown by homoepitaxy process is called epi-layer. The epi-layer is to more lightly doped (5*1017) than the substrate layer (5*1018) to maintaining the low collector resistance to and to get a higher breakdown voltage across the collector-substrate junction. Meanwhile, the Lower collector resistance allows a higher operating speed with the same current. The nanowires are grown on the epi-layer with the density of 0.28NW m2 . Finally, nanowires are grown in 3µm length with a 40 nm diameter. All the nanowires are isolated by SiO2which is a very good insulator. This prevents
NANOWIRE
2
2
When the insolation process is completed, all spaces between the nanowires are filled by indium tin oxide (ITO) in order to make a metal contact. The ITO has good metallic properties to provide an ohmic contact to increase the carrier collection efficiency as well as showing good transparency, which is necessary in the many optical devices. The circuit connection of the complete cross-section sample at a reverse bias is shown below. The energy band diagram is also shown in fig.
NANOWIRE
2
The energy band diagram of the sample under reverse bias condition is shown below:
R
qv
FE
NANOWIRE
5.2 Sample 6074:
This was fabricated in the same way as sample 6080, but there was no intrinsic layer. It is simply a p-n photodiode which had schematic diagram and a plate diagram shown below:19 10 17 10 * 5 18 10 * 5
Chapter-5
Experimental setup
In our experiments, we investigated how much photocurrent was generated as a function of wavelength at different temperatures, when the infrared light is passed through the sample. In this experiment, we have used Fourier Transform Infrared Spectroscopy.
5.1 Fourier Transform Infrared spectroscopy (FTIR): i
nfrared spectroscopy is a simple and reliable measurement technique that use the interference between two light beams. When the signal is detected, an interferogram is formed. This data signal is collected and converted from an interferogram to a spectrum [18]. For the infrared spectral analysis, FTIR (Fourier Transform Infrared) spectroscopy is a more advantageous modern technique than the conventional or dispersive technique because: It is a constructive technique
It does not require external calibration, so it gives precise measurement. It is faster and, as a result, the collecting data can scan every second
The signal to noise ratio is good enough because it has high fast time resolution and greater optical throughput
Its construction is very simple with only one moving part used, so the possibility of the mechanical breakdown is very limited.
A Fourier Transform Infrared (FTIR) spectrometer basically works is based on the Michelson interferometer. It obtains infrared spectra from the sample by using the interference of light with an interferometer to form an interferogram, which is a unique signal that has all of the infrared frequencies encoded simultaneously at a time. During this measurement, signal is measured very quickly due to the order of one second or so, since the overall time of the sample measurement has been reduced to a few seconds rather than the several minutes [18]. The figure below shows the basic mechanism of the FTIR spectrometer:
EXPERIMENTAL SETUP
Figure 24: Basic mechanism of FTIR
In our project, we used the spectrometer VERTEX 80V FTIR, manufactured by BRUKAR OPTICS.
The basic parts of the FTIR are shown in fig 24.
The Source: The instrument is equipped with two different IR sources. One is working in near infrared and the other in mid infrared. In the source compartment, we can choose an aperture so that the radiation effect can be controlled.
.
The Interferometer: This is basically a Michelson interferometer, which consists of a beam splitter and two mirrors. One mirror is fixed and the other one is moved a short distance. A collimated beam from the infrared radiation source is split into two beams by the beam splitter. This beam splitter is suited for measurements with wave numbers between 10000 and 380 cm-1. In our project, the used sample is considered as a (photo) detector. The modulated output radiation from the Michelson interferometer is converted into an electrical current that has been shown by the computer as a spectral distribution of the generated photocurrent by using the mathematical Fourier Transformation technique.
EXPERIMENTAL SETUP We can see above in the figure, that if the beam reflected by the moving mirror has different path lengths as a result of the motion of the mirror. The output signal of the interferometer will be a function of this difference . The Fourier expression can be x
calculated following two ways:
I. If the light source is monochromatic and the wavenumber is
1
, the intensity of the modulated signal becomes:
I
x =
x
v
E
E
E
E
E
2 12 2 22
1 2cos
2
This can be simplified to,
I
x =2 x v
2 cos 1 Eq-18 II. If we used a broad brand source instead of monochromatic source, the outputsignals will be:
I x
kx
Gkdk 0 ) ( 1 cos( )This can be simplified to
I G k e dk I x ( ) ikx 2 1 2 1 0 ) ( Eq-19 Where, I is the intensity of the partial beam and 0 G(k)is the converted intensity corresponding to theI x
The Sample: the sample is placed in the sample chamber. The resulting beam from the interferometer is partly transmitted when the radiation is either reflected or absorbed by the sample,.
The Detector: the transmitted radiation (the interferogram) is detected by the detector. The detector is not used in our project as we use the sample itself as a detector.
The Computer: the detected signal is displayed on computer either as a interferogram or as a spectrum.
EXPERIMENTAL SETUP
The measuring process:
An infrared radiation light source produces a highly coherent beam of light by using collimating mirror. This beam is divided into two beams when it comes to the beam splitter. One of these beams is reflected by a fixed mirror and the other one is reflected by a mirror which moving back and forth within a short distance.
After the reflection of the two beams from their respective mirrors, they recombine and interfere with each other at the beam splitter to create a modulated beam, called interferogram, which is then directed to the sample. The different wavelength of the interferogram signal is either refracted, reflected of absorbed by sample and the transmitted interferogram signal finally arrives at the detector and is measured by detector. In our measurements the sample itself acts as the detector and the optical interferogram is transferred to an electrical interferogram, the measured photocurrent. The detected interferogram signal has to be shown in the form of a spectral response of the infrared radiation by using a well known mathematical interpretation, the Fourier Transformation technique. As a result, the computer shows the infrared spectrum. That is the photocurrent for each frequency component or wavenumber of the infrared radiation.
In order to show the spectral response of the photocurrent as a function of the radiation energy we used the formula:
E hc hcW
Eq-20 Where, 1 1 Wavenumber cm W As we can see the photon energy is directly proportional to the wavenumber and the reciprocal of the wavelength.
5.2 Cryostat:
a cryostat is like a vessel in a vacuum to maintain cold cryogenic temperature (less than 90K) by using cryogenic fluid such as liquid helium. It is a useful application particularly for low temperature absorption, fluorescence, photoluminescence spectroscopy and Fourier Transform Infrared Spectroscopy. In our experiment, we used the OptistatDN cryostat manufactured by Oxford Instrument [22] This cryostate is well suited for the FTIR-spectrometer. A schematic diagram of the cryostat is shown on the following page.EXPERIMENTAL SETUP
EXPERIMENTAL SETUP As we can see in the above, a typical OptistatDN cryostat consists of:
Sample rod entry port: This is the radial and hollow port surrounded by a steel shell where the sample rod can either entered or removed with the sample.
Sample holder: The sample can be fitted in a precise position in the sample chamber by the sample rod. This can be done by using the two sample rod options: one controlling the height of the sample and one controlling the angular position of the sample.
Sample space: This is the place in the cryostat where the sample is placed to be cooled by a separate exchange gas, liquid helium in our case, so that the circulating cryogen does not make contact with the sample.
Vacuum chamber: To maintain the good thermal isolation between cooled part of the cryostate and the surrounding room temperature, the vacuum chamber is needed to be pumped to a high vacuum level (lower than 104mbar).
Electrical access: For the electrical measurement, the wires are connected from top to the bottom of the sample holder where sample and pin connectors are placed respectively. This makes it possible to measure with maximum flexibility for different configuration.
Optical access: This cryostat has an exquisite optical access (f/1) for light collection measurements over different regions of the optical spectrum. It consists of four sets of radial and one set of axial windows. For the small signal FTIR measurement, the large clear access is used in a large illumination area.
Due to the operation mode, the sample is allocated in to the central space of the cryostat which is surrounded by exchange gas to provide extremely uniform cooling. A liquid nitrogen reservoir on the upper part of the central sample tube is placed so that liquid nitrogen goes through by capillary tube to the heat exchanger. The flow of liquid is influenced by gravity and it can be controlled by the gas exhaust valve that is placed on the top of the sample tube. An outer vacuum chamber surrounds the nitrogen reservoir and the sample space to maintain good thermal isolation and prevent the room temperature from heating up the sample space. In order to cool the sample by conduction through the exchange gas, the vacuum chamber needs to be pumped to a high vacuum level.
This cryostat model is designed to operate 77K to 500K. However, in our experiment, we operated it from77 K up to room temperature. The design parameters of optistat DN are given below:
EXPERIMENTAL SETUP
Parameter
Units
Specification
Sample space mm 20 in diameter Temperature range Kelvin 77-500 Temperature stability Kelvin 0.1 Cool down from ambient min ~20 Liquid nitrogen capacity L 1.2 Hold time at 77K Hours ≥15 Sample change time Min 5 Cryostat weight Kg 5
Optical port - 5(4 radial, 1 axial) Radial Optical Access - f/1
The whole experimental setup that we were used to do our measurement is shown in below:
Chapter -6
Results
In our project, we have measured two different characteristics for the both samples. One is the spectral or photoresponse at different biases and temperatures by using Fourier Transform Infrared Spectroscopy. The other is electrical, or I-V, characteristics which were carried out at different temperatures in darkness as well as under illumination.
6.1 Current-Voltage (I-V) characteristics:
We will first present and discuss the I-V characteristics of the both samples. The measurements were carried out in darkness and under illumination at temperatures from 77 K to room temperature..
6.1.1 For the sample 6080:
This sample is structured like a plate which has many sub-samples on it (shown in figure 21). Every sub-sample consists of a large number of wires but behaves like an individual photodiode. We checked randomly the three sub-samples that are marked in the sample plate structure. Although the sub-samples have been fabricated at the same time, they showed result with some differences.The I-V curves can be divided in:
Forward bias state, in which the current increases exponentially. This state is also known as forward bias mode.
Reverse bias state, in which the reverse saturation current is appeared. It can be related to the dark current as:
Eq-21
Under the illumination, I-V curve is shifted by the amount of photocurrent which is called solar cell curves due to the photovoltaic action. Thus, this relation by the equation is
Eq-22
When the applied reverse bias increases we reach a point where the reversed current start to increase rapidly, this point is referred to as the breakdown voltage. For sample 1 we can see that, in darkness, the breakdown voltage is almost -3V at 78K. However, for the 300K, the breakdown voltage is approximately -0.5V.
It is quite clear that the breakdown voltage is decreased due to a rise in the temperature. Before the breakdown voltage, we get the straight line which shows the photodetector stability.
For the sample 2, the breakdown voltage is 1V at 300K. By comparison, it is clear that the breakdown voltage is different for the both samples due to the sample defect.
RESULTS
(a)
(b) Fig 1: I-V characteristics of the sample 1 for different temperature
a) In darkness
RESULTS
a)
b)
Fig 2: I-V characteristics of the sample 1 for different temperature a) In darkness
RESULTS
a)
b
) Figure 3: I-V curves for sample 3 for the different temperaturea) In darkness
RESULTS
Figure 6: ln(I)-V curves for sample1 in darkness and under illumination
RESULTS
Fig 8: ln(I) vs V curves in darkness and illumination for sample 3
Clearly, the most stable sample is the third one in both darkness and under illumination. It has a high breakdown voltage (-2V) at room temperature compared to the other two. The ln(I) vs V curves for the 3 samples in darkness and under illumination are shown in fig 8.
Real diodes do not behave like an ideal diodes (Eq-21) . So the equation for the device has to be modified by using a parameter to show whether the diode behaves closely or apart from the ideal case. This parameter is known as the ideal factor or slope parameter ( ). This factor is affected by many physical processes, such as carrier generation -recombination, tunneling, presence of interface states etc. Usually, it is a function of temperature and the applied bias. For a real diode, the current flow under forward bias in dark, is represented by the equation
Eq-23 To find the ideality factor we can rewrite this as:
Eq-24
RESULTS
So, we can select randomly one of the above curves, and the ideality factor can be calculated by using the eq-24.
For a non ideal case, the current and voltage are affected by the shunt and series resistance; hence the relation is given by
Sh s s R IR V KT IR V q I I 0exp ( ) ! Eq-25
As we can see in ln(I)-V figure regarding Eq-25, the curves are more linear in low bias level than in the high voltage level. This is because the current is increased with respect to the increase voltage and hence, this leads to a voltage drop across the series resistance.
At low voltages, the device’s performances are dominated due to the shunt resistance (Rshunt)
effects, which cause a large peak and the curves are more linear. In a practical sense, it is usually not possible to correct the shunt resistance effects.
At high voltages, the series resistance is dominating device performance and this also causes a large peak but the curves are more non-linear. This effect can be eased by using the photovoltaic action, which is shown above the all ln(I)-V graphs.Usually, we can see that the all graphs are more linear under illumination.
RESULTS
6.1.2 Sample 6074:
(a)
(b) Fig 1: I-V characteristics for sample 1
a) In darkness
RESULTS
(a)
(b) Fig 2: I-V characteristics for sample 2
a) In darkness b) Under illumination
RESULTS
As we can see above in the figures for both samples in darkness and under illumination, Sample 2 has more stable properties than the sample 1. Then breakdown voltage is almost -0.5V of sample 2, while sample 1 has -0.25V in darkness as well as under illumination. However, the curve is shifted downwards by the photocurrent under illumination.
Apparently the pin structure behaves more like a diode than the p-n structure; sample 6080 has better responsivity and higher breakdown voltage than sample 6074. Sample 6080 has intrinsic layer formed between the p and the n layer, while 6074 has no intrinsic layer. We believe the higher breakdown voltage for 6080 is caused by the intrinsic layer. Sample 6074, which has no intrinsic layer creates a larger amount of leakage current, low response and a low breakdown voltage. However, sample 6080 shows more electronic noise due to an increase in the depletion region.
Figure 3 shows the ln(I)-V. In darkness, the sample 2 has followed the characteristic in the forward direction with starting from zero level, whereas sample 1 does not follow the characteristics in the same way due to the sample defect. However, both samples are following the characteristics in the same way under illumination.
RESULTS
(a)
(b) Fig 3: ln(I) vs V for sample 1 and 2
a) In darkness
b) Under illumination
