1
CHAPTER 1
INTRODUCTION
This chapter explains about the background information of nanotechnology, definition of nanowires, fabrication of NWs, problem statement, objectives, scopes, and the significance of the study that had been conducted.
1.1
General In Information
Nanot Nanotech echnol nology ogy generat generates es a lot of attent attention ion these these days days and therefo therefore re it expectation is not only in the academic community but also among investors, the governmen governments ts and industries. industries. It is predicted that nanotechno nanotechnology logy would be able to pro provi vid de
bette etterr
last lastin ing, g,
clea clean ner, er,
safe saferr
and and
smart marter er
prod roducts ucts
for for
homes omes,,
comm commun unica icati tion ons, s, medi medici cine ne,, tran transp spor ortat tatio ion, n, and and agri agricu cult ltur ure. e. This This atom atomic ic scal scalee structure has produced novel materials and devices with great potential applications in a wide number of fields as stated by Serrano E. et al. (2009).
2
Actually nanotechnology is a field which concentrated on nanometer scale of natural and artificial structures. It is stated that the range of nano scale is in the range of one nanometer (10 −9 m) to a few hundreds of nanometer. 1 nm is approximately the distance from one end to the other of a line of five neighbouring atoms in an ordinary solid (Bruus H., 2004) .
NW, which is one of the basic products, plays a key role in this area. NW is a two-di two-dimen mensio sional nal nano nano scale scale struct structure ure that that has future future applic applicatio ation n in integr integrated ated electronics, electronics, and optoelectro optoelectronic nic devices, devices, such as light emitting diodes (LEDs), solar cells and sensors. There are two approaches to produce quantum wire structures, which is either top-down or bottom-up. Top-down method involves lithography, metallization metallization,, and etching. Bottom-up Bottom-up methods methods begin with arrangement arrangement of atoms or molecules to build up nanostructures and in some cases through smart use of selforganization.
As time pass by, a lot of advancement has been made in instrumentation for observation and manipulation of individual atoms and molecules such as atomic force microscope (AFM), field emission-scanning electron microscope (FE-SEM) and transmission electron microscope (TEM).
1.1.1
Definition of a Nanowire
Samuelson L. (2003) had reported that, a NW can be defined as an extremely thin wire with a diameter in the order of a few nanometers (nm) to ~100nm. A free standing NW is also called nano whisker. Earlier, NWs have a promising future as one-dimensional (1D) building blocks for a wide range of applications such as in electro electronics nics and photon photonics ics devices devices includ including ing photo photo detect detectors ors,, chemic chemical al and gas sensors, waveguides, LEDs, micro cavity lasers, solar cells and non linear optical converters converters also in in life-sciences life-sciences as describ described ed by Yan R. et al. (2009). NWs have
2
Actually nanotechnology is a field which concentrated on nanometer scale of natural and artificial structures. It is stated that the range of nano scale is in the range of one nanometer (10 −9 m) to a few hundreds of nanometer. 1 nm is approximately the distance from one end to the other of a line of five neighbouring atoms in an ordinary solid (Bruus H., 2004) .
NW, which is one of the basic products, plays a key role in this area. NW is a two-di two-dimen mensio sional nal nano nano scale scale struct structure ure that that has future future applic applicatio ation n in integr integrated ated electronics, electronics, and optoelectro optoelectronic nic devices, devices, such as light emitting diodes (LEDs), solar cells and sensors. There are two approaches to produce quantum wire structures, which is either top-down or bottom-up. Top-down method involves lithography, metallization metallization,, and etching. Bottom-up Bottom-up methods methods begin with arrangement arrangement of atoms or molecules to build up nanostructures and in some cases through smart use of selforganization.
As time pass by, a lot of advancement has been made in instrumentation for observation and manipulation of individual atoms and molecules such as atomic force microscope (AFM), field emission-scanning electron microscope (FE-SEM) and transmission electron microscope (TEM).
1.1.1
Definition of a Nanowire
Samuelson L. (2003) had reported that, a NW can be defined as an extremely thin wire with a diameter in the order of a few nanometers (nm) to ~100nm. A free standing NW is also called nano whisker. Earlier, NWs have a promising future as one-dimensional (1D) building blocks for a wide range of applications such as in electro electronics nics and photon photonics ics devices devices includ including ing photo photo detect detectors ors,, chemic chemical al and gas sensors, waveguides, LEDs, micro cavity lasers, solar cells and non linear optical converters converters also in in life-sciences life-sciences as describ described ed by Yan R. et al. (2009). NWs have
3
many many intere interesti sting ng proper propertie tiess that that are not seen in bulk bulk or 3-D materials materials becaus becausee electrons in NWs are quantum confined laterally and thus occupy discrete energy levels.
There are three types of nanostructures. They are NWs, quantum dots (QDs) and quantum wells (QW). A quantum dot is a semiconductor whose semiconductor whose excitations are confined in all three spatial dimensions. dimensions. As a result, they have properties that are betwee between n those those of bulk bulk semico semicondu nducto ctors rs and those those of discret discretee molecules, molecules, whil whilee quantum well is a potential well that confines particles, which were originally free to move in three dimensions (3D), to two dimensions (2D), forcing them to occupy a planar region as described by Kang J.H. et al. (2009).
In Science and Technology Dictionary, Quantum Wire (QWR) is defined as a strip of conducting material about 10 nanometers or less in width and thickness that displa displays ys quantu quantum-m m-mech echani anical cal effects effects and univer universal sal conduc conductan tance ce fluctu fluctuati ations ons.. A quantum wire can also be described as an electrically conducting wire, in which quantum effects affects transport transport properties by limiting limiting the motion motion of electrons or holes in two spatial spatial directions directions or in two-dimensiona two-dimensionall (2D) but at the same time still allowing free propagation in the third-dimensions (3D).
1.2
Statements of Problem
Differe Different nt method methodss have have been been used used to synthe synthesiz sizee group group III-V III-V materi materials als nanostructures. That includes metal-organic chemical vapor deposition (MOCVD). NWs NWs have have been been grow grown n with with the the aid of catal catalys ystt befo before re and and the the kine kineti tics cs and and mechanism of III-V NWs growth using Au catalyst has also been investigated and is relatively well understood. Gold (Au) colloid usage as the catalyst is one of the vital aspects in NWs growth. The main function of these gold colloids is as a controller on where the wires growth is significant.
4
Many research before had been focusing on the effect of the growth time, growth temperature, and also the pressure variation on the growth of NWs. But not much of them are really discussing in depth the tapering factor and phenomenon due to the source mole fraction that occurring in the InGaAs NWs and the angle of grown InGaAs NWs on GaAs (100) substrate. The source mole fraction and the substrate orientation were observed to be an important factor for controlling the morphologies of InGaAs structures.
The aim of this work is to investigate the tapering phenomenon due to Indium mole fraction in InGaAs NWs using liquid Au droplets and to calculate the angle of grown InGaAs NWs on GaAs (100) substrate. FE-SEM reveals that the shapes of the nanostructures can be precisely controlled in the MOCVD process to synthesize wires.
Therefore, the growth of InGaAs will be conducted in MOCVD system to establish the mechanism that will help out wire growth and to optimize each growth conditions. It is being proposed that the crystal morphology is determined by the relative rates of InGaAs deposition via the vapor-liquid-solid (VLS).
1.3
Objectives
The objectives of this research are as follows: i.
To grow the InGaAs NWs on GaAs (111)B substrate with different Indium mole fraction.
i.
To determine the tapering factor in InGaAs NWs due to different Indium mole fraction grown on GaAs (111)B substrate.
ii.
To grow InGaAs NWs on GaAs (100) substrate.
5
iii.
To determine the growth direction of InGaAs NWs with respect to GaAs (111)B and (100) substrate.
1.4
Scope of study
In line with the objectives, the experiments will be done under the following scopes; i.
Growing of the InGaAs NWs on GaAs (111) and GaAs (100) substrates using the MOCVD technique.
ii.
Observation on the distribution, size of the wires, tapering factor and uniformity of the InGaAs NWs using FE-SEM.
1.5
Significant of Study
This study may help other researchers to find out the tapering phenomenon in InGaAs NWs using liquid gold (Au) droplets and the growth direction of InGaAs NWs with respect to its substrates grown by MOCVD technique. This study will help to determine the tapering factor in NWs thus lead to the development of nanotechnology in the near future especially in semiconductor fields.
6
CHAPTER II
LITERATURE REVIEW
This chapter explains the theories and the early information related to the study which helps in the pursuing of this project. The information includes NWs, vapor-liquid-solid (VLS) mechanism, gold (Au) colloid as catalyst, MOCVD system, tapering phenomenon in NWs and the relationship between the angle of NWs and substrate orientation, also the analysis and characterization of NWs.
2.1
Nanowires (NWs)
Semiconductor NWs represent a unique system for exploring phenomena at the nano scale and are also expected to play a critical role in future. It is predicted to
7
have lower threshold currents, lower temperature sensitivity, and narrower spectral line widths rather than quantum well devices. The properties make NWs an extremely attractive in future applications involving large scale integration of laser diodes, sensors and other photonic and electronic devices.
As defined earlier, NW is defined as a strip of conducting material about 10 nanometers (nm) or less in width and thickness that displays quantum-mechanical effects. Works by Nemanja Isailovic (2004), have shown that the basic difference between quantum wire and the classical wire is that quantum information cannot be copied instead, it must be transported which means destroying the information at the source and recreating it at the destination. Inside quantum wire there are electrons confined to a narrow one-dimensional (1D) channel with motion perpendicular to the channel quantum mechanically frozen out.
As time past by, modern semiconductor technologies could be applied to fabricate the wires. There are several types of NWs which were metallic, semiconducting, insulating and molecular NWs. Metal NWs are made from platinum (Pt), gold (Au), or nickel (Ni). Whereas semi-conducting NWs are made by the composition of phosphate (InP), arsenide (GaAs), silicon (Si), or nitride (GaN). Insulating NWs are made up from silicon dioxide (SiO 2) or titanium dioxide (TiO), while molecular NWs involves process of repeating organic or inorganic molecular units in particular order. As for this project we used InGaAs and GaAs NWs, it can be stated that, semi-conducting NWs were used for the project.
Many systematic studies have been published on the growth of NWs using different methods of growth. As mentioned before, the fabrication methods are generally classified as top-down and bottom-up approaches. The later approach refers to the arrangement of smaller components into more complex assemblies, which include chemical vapor deposition (CVD) and also molecular beam epitaxy (MBE). There are a number of forms of CVD used. Each process differs in the means by which chemical reactions are initiated and the process conditions. For this project, MOCVD is used rather then MBE due to its versatility to the sources also because of its flux uniformity is better than MBE.
8
2.2
Vapor-Liquid-Solid (VLS) Mechanism
The vapor-liquid-solid method (VLS) is a mechanism for the growth of onedimensional (1D) structures. It can be stated as one of the most popular method commonly used to fabricate semiconductor NWs. This mechanism was introduced by Wagner & Ellis in 1964. According to this method, a small nanometer size Au particle on the semiconductor surface forms a eutectic liquid alloy with the host material at growth temperature. Based on Wagner R. S. and Ellis W. C. (1964) the growth of a crystal through direct adsorption of a gas phase onto a solid surface is generally very slow. Thus, VLS mechanism improved this by introducing a catalytic liquid alloy phase which can rapidly adsorb a vapor to supersaturation level, and from which crystal growth can subsequently occur from nucleated seeds at the liquid-solid interface.
As formulated by Wagner & Ellis (1964), the VLS mechanism is typically described in three stages. The stages include the preparation of a liquid alloy droplet, usually (Au) upon the substrate from which a wire is to be grown. Next is an introduction of a vapor (sources) to the substance which is then adsorbs on to the liquid surface, and thus diffuses into the droplet. Finally is the formation of supersaturation and nucleation at the liquid or solid interface leading to axial crystal growth to produce NWs. A schematic of the elementary processes that occur during the deposition is presented in Figure 2.1.
Recently, the VLS growth technique has been extended to nanometer-sized Au particles on a variety of semiconductor materials including III-V compound.
9
With a reduction in size, unique electrical, mechanical, chemical and optical properties would result in large surface area and quantum confinement effects.
Figure 2.1: Schematic diagram of the VLS growth mechanisms for Self-Catalyzed InP NWs in MOCVD (after Robyn Woo L. et al. (2009).
2.3
Catalyst
There are several requirements for catalyst particles. As wrote by Wagner R. S. et al . (1975), catalyst particle must form a liquid solution with the crystalline material to be grown at the NWs growth temperature. The solid solubility of the catalyzing agent needs to be low in the solid and liquid phases of the substrate material as well. The equilibrium vapor pressure of the catalyst over the liquid alloy also have to be small, so that the droplet does not vaporize, shrink in volume, and decrease the radius of the growing wire until, ultimately, growth is terminated. Above and beyond, the catalyst must be inert (non-reacting) to the reaction products during CVD NWs growth.
10
Works by Yicheng L. and Jian Z. (2004) have shown that during VLS growth process, the rate of wire growth is dependent on its diameter. The larger the wire’s diameters, the faster the NWs grow axially. This is due to the fact that the supersaturation of the catalyst is the main driving force for NWs growth and decreases with decreasing wire diameter. Usually, for the growth of NWs, gold (Au) colloid is used as catalyst. Spirkosa D. et
al . (2009) in their report have stated that Au is used because it works well with wide range of semiconductor materials and this surely helps in better performance of devices. It is in line with what Zhang X. et al . (2006) had described that the concentration of gold (Au) colloid particles certainly gives significant consequence to the height of NWs.
2.4
Metal-Organic Chemical Vapor Deposition (MOCVD)
Chemical vapor deposition (CVD) is a chemical process and can be classed as a system where chemical precursors react either in gas phase or on the substrate to produce a high-purity, high performance solid materials or film which has a composition different from any of the precursors. It is a technique to produce II-VI and III-V semiconductor structures.
In recent years, a number of different CVDs are widely used. These processes differ in the means by which chemical reaction are initiated and process conditions. In the semiconductor industry two major techniques are well established. There are MOCVD and MBE. The difference of these two epitaxy techniques is that MOCVD are associated with the epitaxy of compound semiconductors using gas sources like
11
arsenides (AsH3), phosphides (PH 3), and nitrides (NH3) and MBE are associated with high purity solid source materials such as arsenic, gallium, indium or aluminium.
The MOCVD is a process that utilizes the chemical reaction between the Metal-Organics (MO) and or between the MOs and hydrides. MOCVD is distinguished by the chemical nature of the precursor gases. As the name implies, metal organic compounds like trimethlygallium (TMGa), trimethlyindium (TMIn), and trimethylaluminium (TMAl) are employed. They react with group V hydrides. As what Ramdani J. and Vaccari G.
wrote in Semiconductor Manufacturing
Handbook (2004), for III-V compound semiconductor epitaxy, an additional parameter which is the V/III ratio simply the mole fraction ratio of group V elements over group III elements must be taken into account. The V/III ratio needs to be high enough to compensate for the low decomposition rates of hydrides and to drive carbon and oxygen from group V lattice sites. For example GaAs grown at a typical temperature of 700°C, a V/III of about 100 is sufficient while for GaN using ammonia for group V, a V/III of about 1000 at 1050°C is needed for high-quality material.
It is desired that the chemical reaction does not occur at room temperature but occurs only at the elevated temperatures. So the reactant is supplied to the substrates that are heated to the desired temperature. Then when the reactants reach on the surface of the substrate, they will undergo thermal decomposition and react on it to form the film. The metal organic (MO) sources are delivered to the reactor chamber by the carrier gas (high purity hydrogen) that passes through the MO bubblers. The hydrides are usually gaseous and directly delivered from the cylinder. When different kinds of compounds are desired to be grown on the same substrate, the injections of reactants are switched by the injection manifold. The processes parameters include the temperature, pressure, and the amount of reactants supplied to the substrate. Since it is desired to grow films with uniform thickness and composition on a large area of substrate, the reactants are fed into the growth chamber through the shower head.
12
The MOCVD reactor is usually equipped with the components listed below in order to fulfill the function above: i.
Growth chamber with a shower head, a heater, a susceptor on which the substrate is sit.
ii.
Vacuum system that control the pressure of the growth chamber.
iii.
Injection manifold that switches the reactants into the growth chamber.
iv.
MO bubblers in the temperature bath.
v.
Hydrides in the cylinders.
vi.
Gas distribution system equipped with Mass Flow Controllers (MFCs), Electronic Pressure Controllers (EPCs), valves, a nd SUS tubing.
vii.
Exhaust lines through which the by-product of the chemical reaction passes out.
The gas manifold is made of semiconductor-grade stainless steel tubing with pneumatic valves, mass flow controllers, and pressure regulators and transducers. The growth chamber is made of either quartz or stainless steel and consists of a susceptor and elements of heating (resistive, induction heating, and lamp heating) and finally the exhaust system. Figure 2.2 (a) and (b) below shows the pressure controller and toxic gas detector used at Ibnu Sina Institute MOCVD system.
Figure 2.2 (a): Pressure controller
13
Figure 2.2 (b): Toxic Gas detector
Nowadays, two reactor designs are available which are the vertical reactors in which the gas flow is perpendicular to the substrate surface and horizontal reactors in which the gas flow is parallel to the substrate. In addition, for batch or multiwafer processes, two types of susceptor designs exist which are the barrel type and the pancake type. These reactors are suitable for small diameter substrates (less than 200 mm) standard complementary metal oxide semiconductors (CMOS) and for discrete device applications where the requirements in thickness and resistivity uniformity are not tight. Advanced CMOS epitaxy systems for 200 and 300 mm diameter wafers are of the horizontal type equipped with cassette load locks, a transfer chamber, and a cool-down chamber. Figure 2.3 and 2.4 below show the MOCVD system and its schematic diagram located at Ibnu Sina Institute, Universiti Teknologi Malaysia used for the fabrication of III-V compound semiconductor.
14
Figure 2.3: MOCVD system at Ibnu Sina Institute for Fundamental Science Studies
15
Figure 2.4: Schematic Diagram of MOCVD Gas lines
16
2.5
Tapering in the Growth of NWs
Figure 2.5 below had been reported by Cho H. S. et al. (2010). In this study, Ge nanocones were grown by an Au-catalyzed process in a chemical vapor deposition (CVD) chamber using germane (GeH 4) as the source gas with temperature variation. The corresponding variation of taper in the segments was observed. When the growth was conducted at an initial temperature of 350 oC then increased to 400 oC, faceted pillars consisting of a base segment of largely uniform diameter and a conical tapered tip were formed. When the growth was conducted at 395 oC and 375 oC, then at 350oC, the NWs featured a base with a wider taper and tips with a smaller taper.
Figure 2.5: SEM images and respective schematic diagrams of Ge NWs grown at:
(a) 380oC, (b) 350oC, then increases to 400 oC, and (c) 395, 375, then 350 oC (After Cho H. S. et al., 2010).
17
Thus, from all of the images, Cho H. S. et al. (2010) in their report had concluded that, the tapering phenomenon occur depending on the growth temperature. Catalyzed unidirectional growth at lower temperatures approaching the Au–Ge eutectic point, and increasingly isotropic and facet-bounded epitaxial grow that higher temperatures up to 400 oC.
The effect of temperature during the growth is not only seen through the morphology of the quantum wires produced, but it could also influence the sources ratios (net carrier concentration). Kasemset D. et al. (1984) in their report had reported that the growth temperature does affect the carrier concentration for different V/III sources ratios In general, the net carrier concentration decreases at .
higher growth temperature. This could be due to the differences in the relative incorporation of donors and acceptors. At higher growth temperatures, the concentrations of donors and acceptors move closer together resulting in lower net carrier concentration.
In other research conducted by Plante M.C. et al. (2007), the Au-assisted growth of GaAs NWs by gas source molecular beam epitaxy has been investigated. This paper generally investigates about the tapering, sidewall faceting and crystal structure of GaAs NWs. Figure 2.6 shows 45° titled view of GaAs NWs grown on GaAs (111) B substrate with growth time variations.
18
Figure 2.6: 45°tilted view of GaAs NWs grown on a GaAs (111) B substrate for (a)
3 min (b)10 min and(c) 30 min at a temperature of 525 ºC and V/III ratio of 1.5. The insets are TEM images illustrating the typical morphology of the NWs from each specimen (After Plante M.C. et al., 2007).
It is important to note that the NWs diameter is essentially constant up to a certain height from the NWs base, and than reduces rapidly near the tip thus proposed that the radial growth occurs at some expense of axial growth via layer-bylayer and step-flow growth modes on the NWs sidewalls.
2.6
Substrate Orientation
Mostly NWs growth is carried out on (111) substrates that commonly results orthogonal oriented NWs to the substrate surface (Rosnita Muhammad et al., 2009). In addition, NWs with cubic crystal structure are often epitaxially grown on (111) substrates to achieve vertically aligned, out-of-plane NWs growth along the surface
19
normal. Vertically oriented NWs on (100) substrates is also preferred to grow. But unfortunately, NWs growth on (100) substrates were typically yields tilted growing along the equivalent (111) crystallographic directions (Fortuna S. A. et al., 2008). Based on Wagner and Ellis (1964), NWs generally grow in the crystal direction that minimizes the total free energy which, in most cases, is dominated by the surface free energy of the interface between the semiconductor and the metal catalyst. Additionally, in the origin of thermodynamics, a system or any material was in it most stable state only when it is in a state with the lowest energy. This were supported by research conducted by (Ghosh S. C. et al., 2009) where the NWs growth on polar III–V substrates such as GaAs (100) results in NWs grown along the (111)B directions where As-faced planes are preferable for nucleation.
2.7
Characterization of NWs
The information on nanostructures has open up ideas for possibility of measuring the parameters associated with relevant theories and mechanism. Selected techniques have to be used to characterize the NWs since the surface morphology of the NWs cannot be seen by unaided human eyes.
For the characterization of NWs, SEM is used to characterize the surface morphology of the grown NWs. Interaction of electrons with atoms that make up the sample producing signals that contain information about the sample’s surface topography, composition and other properties.
The SEM has many advantages over traditional microscopes. Some of its main advantages are that, the SEM has a large depth of field, which allows more of a specimen to be in focus at one time. The SEM also has much higher resolution, so
20
closely spaced specimens can be magnified at much higher levels. Figure 2.7 shows an image of NWs that being characterized using SEM.
Figure 2.7: Image of InP NWs (After Woo R. L. et al., 2009).
Additionally, the NWs samples also can be viewed with the aid of FE-SEM (Rosnita Muhammad et al., 2008). A field-emission cathode in the electron gun of a scanning electron microscope provides narrower probing beams at low as well as high electron energy, resulting in both improved spatial resolution and minimized sample charging and damage. FE-SEM produces clearer, less electrostatically distorted images with spatial resolution down to 1 1/2 nm, which is 3 to 6 times better than conventional SEM. High quality, low voltage images are obtained with negligible electrical charging of samples. Figure 2.8 shows the image of NWs that being characterized using FE-SEM.
21
Figure 2.8: Image of GaAs NWs (After Rosnita Muhammad et al., 2008)
CHAPTER III
METHODOLOGY
This chapter discusses the research method that was used while conducting the study. It includes the sample preparation of GaAs (111) and (100) substrates until the analysis and characterization of InGaAs (InGaAs) quantum wires using FE-SEM.
22
The experiments were done in RF heated reactor cell under low pressure MOCVD condition (100 mbar). Trimethylgallium (TMGa), Trimethylindium (TMIn), Arsine (AsH 3) and GaAs (111) and (100) substrates were used in this experiment.
3.1
Sample Preparation
The experiments started with semi insulating GaAs (111) and (100) wafer immersed in 0.1% poly-L-lysine (PLL) solution for 3 minutes. After washing with de-ionized (DI) water and subsequently blow dried with N 2, the wafer was cut into desired sizes depending on application. The 30 nm diameter gold colloids were dispersed onto the wafer surface by using micro littre pipette and immediately washed after 20 seconds, followed by drying with N 2. The PLL layer on the surface is positively charged and attracts the negatively charged of the gold colloids.
The growth recipes were first keyed-in into the computer. It consisted of all information needed for the experiment including the pressure (Torr), temperature (oC), time and also the sources ratio for the experiment to the controller computer that was connected and controlling the MOCVD machine.
3.2
Substrate Loading
Before the chamber was opened and the substrate was loaded into the chamber, the pressure in the growth chamber should always be set to the atmospheric pressure. It was opened by lowering the SUS plate at the bottom of the SUS growth
23
chamber using the up-down system operated by N 2 gas. It was important to flush the growth chamber first when it were opened. Since N 2 is an inert gas, N 2 purge was the best for this propose. It was also possible to used H 2 gas rather than N 2 gas but, it should be made sure that the H 2 gas was properly ventilated.
Next, the substrate was loaded onto the graphite susceptor by using tweezers. Afterwards, the growth chamber were closed by raising the lower SUS plate by using the up-down system and the carrier gas was started to flow into the chamber. Usually, H2 gas was used as the carrier gas and it was supposed that the carrier gas flows through all the lines that will be used to grow.
After the chamber was closed, the growth chamber was evacuated. The carrier gas flow was started and the pressure was set to the desired value. This process was done by opening the valves in between the growth chamber and the vacuum pump. The vacuum pump was supposed to be run all the times unless in maintenance process or in an emergency case. The pressure in the growth chamber was automatically controlled by the pressure control system that were composed of the Baratron (Absolute Pressure Transducer), Pressure Controller, and the butterfly valve.
After the substrate was loaded, the growth chamber was then evacuated to the desired pressure. Then the valve in between the injection manifold and the growth chamber was opened in order to flow H 2 gas into the growth chamber. Next, wait until the pressure settles down to the desired value.
3.3
MOCVD Growth Process
24
The heater of MOCVD was turned on and the temperature was raised to the pre-bake or generally known as soft bake temperature. The temperature was automatically controlled by the Temperature Controller that could be manually or computer set. Before it was heated up to the actual process temperature, the temperature was usually set around 250 to 300 oC. The rationale of this temperature is to prepare the susceptor and the growth chamber in the MOCVD. Pre-bake or soft bake temperature usually holds for about 10 minutes.
To provide overpressure, the flow of hydrides was started. If the substrate contained the atomic element that easily evaporates, then reactants were needed to be supplied to compensate its evaporation. As in this project, InGaAs NW was the desired outcome, therefore AsH3 or As reactant was supplied to provide overpressure. The overpressure is then kept until the growth run were finished. The substrate was then annealed or heated in-situ at 600 oC for 10 minutes under AsH3 ambient to desorb surface contaminants. This was also called as the hard baking process. At the same time, Au particles were alloyed with GaAs, primarily by an up-take of Ga particle in the Au droplet. It was stabilized for 10 minutes after the process.
The carrier gas was then started to flow through the MO bubbler. The stabilization period was needed for the MOs in the bubbler to be saturated by the carrier gas, usually about 10 minutes. With the overpressure, the temperature was raised for the process. The reactants which were trimethylgallium (TMGa) and trimethylindium (TMIn) were then flowed into the process by switching the injection manifold valves to the run position from the vent position that was done through the run program prepared on the computer.
25
3.4
Sample Unloading
Before the growth chamber was opened, the carrier gas was switched from H 2 to N2. In order to open the growth chamber to unload the grown sample, it was required to follow the same procedure as the sample loading process. Subsequently, when the susceptor was already cooled down to room temperature, the growth chamber was then ready to open to remove the sample out from the growth chamber.
When the sample was taken out, the growth chamber was evacuated to maintain it in vacuum condition. The air in the growth chamber could oxidize the reactants in the growth chamber wall. Later then, the growth chamber is filled back with H2 or N2 gas.
3.5
Sample Analysis and Characterization Technique
In this project, the samples were analyzed by using Field-Emission Scanning Electron Microscope (FE-SEM) as it has much higher resolution, therefore closely spaced specimens can be magnified at much higher level. Besides, it has an ability to produce strikingly clear images. In order to analyze the sample, it should be prepared first by coating the sample with Auto Fine Coater (JOEL JFC 1600). The purpose of Auto Fine Coater coating was to hold the grown nanowires in place by increasing the conductivity between the nanowires and the surface of the substrate. After the coating process was finished, the sample was then attached to the FE-SEM specimen holder to be observed.
26
CHAPTER IV
RESULTS AND DISCUSSION
This chapter showed the results of the experiment that had been conducted. The results were in the form of the FE-SEM images of the NWs which showed the distribution and area density (number of NWs per unit area) of NWs in each sample. From the FE-SEM images the length and tapering factor (diameter) of NWs were determined including the direction or orientation of the NWs.
27
4.1
FE-SEM Images of InGaAs NWs
The images of InGaAs NWs grown using gold (Au) droplets as a function of Indium mole fraction on GaAs (111) substrate as characterized by Field-Emission Scanning Electron Microscope (FE-SEM) were shown in Figure 4.1 below. Morphology of NWs looks very different even grown at narrow range of Indium mole fractions. At 0.41 the InGaAs NWs grown were up straight with relatively cylindrical shape. While at 0.47, it grown with bigger size at the bottom leaded to tapered wire. At 0.65, InGaAs NWs grown with low height and highly tapered form whereby the top of the NWs was highly small compared to the based of the NWs. Figure 4.2 shows the image of InGaAs NWs grown using gold (Au) on GaAs (100) substrate. The NWs growth on (100) substrates were shown to yields tilted growing along the equivalent crystallographic directions.
a
b
28
c
Figure 4.1: FE-SEM images of InGaAs NWs grown with different value of Indium
mole fraction: (a) 0.41, (b) 0.47, and (c) 0.65. Images were viewed at an angle of 50 o from the surface normal.
Figure 4.2: Images of InGaAs NWs grown on GaAs (100) substrate
29
From these images, there are several parameters that can be discussed as in section below. Those parameters are the area density, growth direction, length and tapering factor of the grown NWs.
4.2
Effect of Indium Mole Fraction Value towards Area Density of InGaAs NWs
The morphology of the InGaAs NWs varies depending on the process condition. From Figure 4.1 the area density initially increases from 2.24 x 10 12 m-2 at indium mole fraction 0.41 to 2.53 x 10 12 m-2 at indium mole fraction 0.47. However, when the indium mole fraction is increased to 0.65 the area density decreased to 1.19 x 1012 m-2. The change in the trend of area density was due to the change in the growth of the NWs. Besides, it can be observed that In xGa1-x As NWs grown were different in diameter, surface density and length.
Figure 4.3: The change of area density with Indium mole fraction.
30
4.3
Effect of Indium Mole Fraction Value to the Length and Tapering of InGaAs NWs
The NWs’ length and diameter can be simply controlled by adjusting the growth parameter. In this study, the NWs were varied by the change in Indium mole fraction. To study the gradation of NWs tapering, the tapering factor ( TF ) is used (Wibowo Edy, 2011). The smaller the tapering factor value, the lower the NWs tapering. The poorer value that TF can take place is zero (0.00). Meaning that, the diameters along the grown NWs become more uniform at much lower value of TF. Therefore, the tapering factor TF can be calculated using equation 4.1 (c) below (Wibowo Edy, 2011). The average length, average diameter, and tapering factor of the NWs grown with different Indium mole fraction ( x) are shown in Table 4.1.
=
4.1 (a)
TF i=
4.1 (b)
=
: Tapering factor (
4.1 (c)
0.00)
TF i
: Tapering factor for individual NW
n
: The number of NW sample : Average diameter of individual NW
31
dT ,i dM ,i dBi
: Diameter of tip, middle and bottom of individual NW, respectively
(After Wibowo Edy, 2011)
Table 4.1: The length, diameter, and tapering factor of the InGaAs NWs at different
Indium mole fraction ( x)
Indium Mole
Average Length
Average
Tapering Factor,
Fraction ( x)
(nm)
Diameter (nm)
(TF )
0.41
1000
1070
11.00
0.47
1180
966
26.59
0.65
455
1140
30.78
Figure 4.4 shows that, the dependence of length and tapering of the NWs towards the Indium mole fraction value. Based on the figure, the length of InGaAs NWs were increased when the Indium mole fraction increasing from 0.41 up to 0.47, while started to decreased when the Indium mole fraction were further increased from 0.47 to 0.65. At the same time the tapering of the NWs increases continuously from 0.41 up to 0.65.
32
Figure 4.4: Plot of InGaAs NWs length versus Indium mole fraction
The change in the trend of the length and tapering was due to the change in the growth mechanism from direct impinging to combination of direct impinging and sideway diffusion of source atoms. At lower Indium mole fraction most atoms were favorable for the direct impinging mechanism (Figure 4.5(a)). As the Indium mole fraction increased to 0.65 the mechanism has changed to a combination of direct impinging and sideways diffusion (Figure 4.5(b)).
1 2 3
(a)
1 2 3
(b) Substrate
(c)
33
Figure 4.5: Different growth mechanism models for sources atom incorporated into
the growth of In xGa1-xAs NWs. The metal droplet (seed particle) is either in solid or partially molten state, (a) direct impinging mechanism, (b) combination of direct impinging and diffusion of source atoms from substrate surface (c) domination of diffusion of source atoms mechanism from substrate surface.
Figure 4.6: Plot of InGaAs NWs tapering factor versus Indium mole fraction
Based on the above results, the tapering of the NWs at high value of indium mole fraction is due to the effect of radial growth. A significant amount of reaction species arrived at the NWs sidewall or GaAs (111) substrate and diffused towards the gold nanoparticle. During the migration process, the incorporation of the reaction species into the GaAs (111) surface and NWs sidewall takes place. This incorporation was known as the 2-D growth, results in the tapering of NWs (Tan H. H. et al . 2006). Besides, the decrease of apparent NWs length from 0.47 to 0.65 is due to the increase of 2-D growth at GaAs (111) substrate.
34
4.4
Effect of Different Substrate Orientation to the Growth Direction of InGaAs NWs
The substrate often chose such that the NWs growth direction were normal to the growth plane. From this experiment, the growth direction for InGaAs NWs grown on GaAs (111) were in (111) direction which is perpendicular (90°) to the substrate, while for InGaAs NWs grown on GaAs (100) they grew at 52° to the substrate surface.
The above result were obtained by the condition that NWs generally grow in the crystal direction that minimizes the total free energy which in most cases, it were dominated by the surface free energy of the interface between the semiconductor and the metal catalyst. For zinc-blende crystals like GaAs, it has been widely observed that the semiconductor catalyst interface often forms a single surface at the lowestenergy (111) plane and thus NWs tend to grow in the ( 111) direction for most growth conditions (Wagner and Ellis, 1964).
For compound zinc-blende semiconductors and their alloys the (111)B is the lowest energy plane (Braun W et al ., 2001) and therefore NWs have been generally observed to grow in the (1 11)B direction (Hiruma K et al ., 1995). For the case of InGaAs NWs grown on GaAs (100) substrate, the growth direction were in (111) direction but with an angle of 38° to the normal or simply 52° from the substrate surface. This was obtained by the condition that GaAs (100) substrate has much higher surface energy compared to that of GaAs (111) substrate.
35
CHAPTER V
CONCLUSION AND SUGGESTION
36
This chapter presented particular conclusions that had been obtained from all the results of the conducted experiment. Several suggestions were also presented in the purpose for future research to improve the previous conducted research.
5.1
Conclusion
From the results and discussions that had been presented, there are several conclusions that can be attained: i) The growth of InGaAs using vapor-liquid-solid (VLS) technique by MOCVD can be done by setting the growth temperature at 400°C and the growth time at 30 minutes using GaAs (111)B substrate. ii) For the increasing of Indium mole fractions value from 0.41 to 0.47, both length and the tapering (diameter) of NWs were increased while further increased of Indium mole fractions from 0.47 to 0.65 leaded to the decreased in NWs length but, on the contrary the tapering were continuously increased. iii) The growth of InGaAs using vapor-liquid-solid (VLS) mechanism by MOCVD can be done by setting the growth temperature at 400°C and the growth time at 30 minutes using GaAs (100) substrate. iv) The InGaAs NWs were grown at an angle of 38° to the normal or simply 52° from the GaAs (100) substrate surface.
37
5.2
Suggestions
Based on the experiment, there are some suggestions for improving the research for future purposes: i) Further investigation on the effect of Indium mole fraction towards NWs with more samples and wider range of Indium mole fraction to other material besides InGaAs NWs. ii) Investigate more on the characterization of NWs, such as EDAX or other characterization techniques rather
than
FE-SEM only
for
advance
understanding of NWs composition. iii) Investigate the relationship of InGaAs NWs growth direction towards various substrate orientations besides GaAs (100) and GaAs (111) substrate.
REFERENCES
Bauer J., Gottshalch V., Paetzelt H., Wagner G., Fuhrmann B., Leipner H. S.,
MOVPE Growth and Real Structure of Vertical-Aligned GaAsNWs. Journal of Crystal Growth 298, (2007) 625-630.
38
Braun W, Kaganer V M, Trampert A, Sch¨onherr H-P, Gong Q, N¨otzel R, D¨aweritz L and Ploog K H, Diffusion and incorporation: shape evolution
during overgrowth on structured substrates, J. Cryst. Growth (2001) 227–228. Dapeng Wei, Qing Chen, The temperature dependence of 1D germanium
nanostructures grown in a small-diameter quartz tube cavity by vapor deposition, Journalof Crystal Growth 312. (2010) 2315–2319. Dr Stormer H., and Prof. Laggner P., Arriving At a Crossroads in Science , Joint EC/NSF Workshop on Nanotechnologies, Toulouse. (2000) 9-11. Elena Serrano, Guillermo Rus, Javier Garcı´a-Martı´nez, Nanotechnology for
sustainable energy, Renewable and Sustainable Energy Reviews 13. ( 2009) 2373–2384.
Ghosh S C, Kruse P., and LaPierre R R., The effect of GaAs(100) surface
preparation on the growth of NWs, Nanotechnology 20, (2009).
Hans Cho S., Theodore.Kamins, In situ control of Au-catalyzed chemical vapor
deposited (CVD) Genanocone morphology by growth temperature variation , Journal of Crystal Growth 312, (2010) 2494–2497. Hiruma K, Yazawa M, Katsuyama T, Ogawa K, Haraguchi K, Koguchi M and Kakibayashi H., Growth and optical properties of nanometer-scale GaAs
and InAs whiskers , J. Appl. Phy,.Vol. 77, (1995) 447–62. Jabeen F., Rubini S., Martelli F. , Growth of III-V Semiconductor NWs by
Molecular Beam Epitaxy , Microelectronics Journal 40. (2009) 442-445. Kang J.H., Gao Q, Joyce H.J., Tan H.H., Jagadish C., Kim Y.,. Choi D.Y, Guo Y., Zu H., Zou J., Fickenscher M.A., Smith L.M., Jackson H.E. and Yarrison Rice J.M., Novel Growth and Properties of GaAsNWs on Si Substrates, Nanotechnology, IOP Publishing. (2009). Kasemset D., Hess K. L., Mohammed K. and Merz J. L. , The Effects Of V/Iii
Ratio And Growth Temperature On The Electrical And Optical Properties Of
39
Inp Grown By Low-Pressure Metalorganic Chemical Vapor Deposition, Journal of Electronic Materials, Vol. 13, No. 4. (1984). Lars Samuelson, Self-forming nanoscale devices , Materials today, (2003). Lu, Yicheng; Zhong, Jian (2004). Todd Steiner. Edition. Semiconductor
Nanostructures for Optoelectronic Applications. Norwood, MA: Artech House, Inc. pg. 191–192. ISBN 978-1-580-53751-3. Nemanja Isailovic, Mark Whitney, Yatish Patel, and JohnKubiatowicz, Datapath
and control for quantum wires , ACM Transactions on Architecture and Code Optimization (TACO) Vol 1, Issue1, (March 2004) 34 – 61. Paek J.H., Nishiwaki T., Yamaguchi M. and Sawaki N., Catalyst Free MBE-VLS
Growth of GaAs Nanowires on Si (111) Substrate, Phys. Status Solidi C 6, No. 6, 1436–1440. (2009) 89-3693. (2009) Paiman S., Gao Q., Tan H.H, Jagadish C., Pemasiri K., Montazeri M., Jackson H.E., Smith L.M., Yarrison Rice J.M., Zhang X. and Zhou J., The Effect of
V/III Ratio and Catalyst Particle Size on Crystal Structure and Optical Properties InPNWs, Nanotechnology, IOP Publishing. (2009). Peter Krogstrup, Jun Yamasaki, Claus B. Sorensen, Erik Johnson, Jakob B. Wagner, Robert Pennington, Martin Aagesen, Nobuo Tanaka and Jesper Nygard, Junctions in Axial III-V HeterostructureNWs Obtained via
Interchange of Group III Elements, Nano Letters, Vol.9, No.11, pg 36. Robyn Woo L., Li Gao Niti Goel, Mantu Hudait K., Kang Wang L., Suneel Kodambaka, and Robert Hicks F., Kinetic Control of Self-Catalyzed Indium
PhosphideNWs, Nanocones, and Nanopillars, Nano Letters. (2009).
Rosnita Muhammad, Zulkafli Othaman, SamsudiSakrani, YussofWahab , Vapor-
liquid solid mechanism using gold colloids for the growth of GaAsNWs, Journal of Fundamental Sciences. (2008).
Rosnita Muhammad, Zulkafli Othaman, Yussof Wahab, Samsudi Sakrani, W. Faizal W. Ahmad & Mohd Nazri , GaAs NWs Formed by Au-assisted Metal-
40
organic Chemical Vapor Deposition: Effect of Growth Temperature, Modern Applied Science, Vol. 3, No. 7. (2009). Rossi F. and Molinari E., Coulomb-Induced Suppression of Band-Edge
Singularities in the Optical Spectra of Realistic Quantum-Wire Structures Phys. Rev. Lett. 76. (1996) 3642. Ruoxue Yan, Daniel Gargas F. & Peidong Yang, Nature Photonics3, (2009) 569 – 576. Seth Fortuna A., Jianguo Wen,Ik Su Chun, and Xiu ling Li, Planar GaAs NWs on
GaAs (100)Substrates: Self-Aligned, NearlyTwin-Defect Free, and Transfer Printable, Nano Letter, Vol. 8, No. 12, (2008) 4421-4427.
Seth Fortuna A.and Xiu ling Li, Metal-catalyzed semiconductor NWs: A review
on the control of growth directions , Semiconductor Sciences and Technology, 25 (2010) 024005.
Spirkosa D., Abstreite G. and Font cuberta Morral A., GaAs NWs and Related
Prismatic Heterostructure, Nanotechnology, IOP Publishing (2009). Wagner R S and Ellis W C, Vapor–liquid–solid mechanism of single crystal
growth, Appl. Phys. Lett.4. (1964) 89–90. Wei Lu and Charles M Lieber, Semiconductor NWs, Journal of Physic D: Applied. Physics, Vol. 39. (2006). Wibowo Edy (2011). Gold Seed-Particles Assisted Growth of InGaAsNWs. UniversitiTeknologi Malaysia: Master Thesis.
Wibowo Edy, Zulkafli Othaman, Samsudi Sakrani, Amira Saryati Ameruddin, Didik Aryanto, Rosnita Muhammad, Imam Sumpono and Habib Hamidi nezhad, Morphorlogical Evolution of Ternary InGaAs NWs grown with Au-
particle assisted using vertical chamber MOCVD, Journal of Fundamental Science, Vol. 6, No. 2 (2010) 131-135.
41
Xinpin Zhang, Baoquan Sun and Richard H. Friend, Nano Letters, Vol.6, No.4, (2006) 651-655.
APPENDIX A
MOCVD Software System
42
Auto Fine Coater (JEOL JFC 1600)
Coating Process
43
44
APPENDIX B
InGaAsNWs Images by FE-SEM
•
Temperature : 400°C
•
Time : 30 minutes
•
Type :InGaAs NWs
•
Substrate :GaAs {111}
•
Indium mole fraction : 0.41
45
•
Temperature : 400°C
•
Time : 30 minutes
•
Type :InGaAs NWs
•
Substrate :GaAs {111}
•
Indium mole fraction : 0.47
46
•
Temperature : 400°C
•
Time : 30 minutes
•
Type :InGaAs NWs
•
Substrate :GaAs {111}
•
Indium mole fraction : 0.65
47
•
Temperature : 400°C
•
Time : 30 minutes
•
Type :InGaAs NWs
•
Substrate = GaAs (100)
•
Indium mole fraction = 0.41
APPENDIX C
48
Graphs of Area Density, Length and Tapering Factor of InGaAs NWs versus Indium Mole Fraction
