The Pennsylvania State University

The Graduate School

College of Engineering

ELECTRICAL STUDIES OF SCHOTTKY BARRIER DIODES

(SBDs) ON GALLIUM NITRIDE (GaN)

A Thesis in

Electrical Engineering

by

Asim Mohammed A. Noor Elahi

© 2015 Asim Mohammed A. Noor Elahi

Submitted in Partial Fulfillment of the Requirements for the Degree of

Master of Science

May 2015

The thesis of Asim Mohammed A. Noor Elahi was reviewed and approved* by the following:

Zhiwen Liu Professor of Electrical Engineering Thesis Co-Adviser

Osama O. Awadelkarim Professor of Engineering Science and Mechanics Thesis Co-Adviser

Jian Xu Associate Professor of Engineering Science and Mechanics and Adjunct Professor of Electrical Engineering Thesis Co-Adviser

Kultegin Aydin Professor of Electrical Engineering Head of the Department of Electrical Engineering

*Signatures are on file in the Graduate School

ii

ABSTRACT

In this work, the thesis describes experiments made on both GaN Schottky barrier diodes

(SBDs) and commercially available SiC Schottky barrier diodes (SBDs). The electrical characterizations on both devices were investigated. Current – Voltage technique was used for finding the barrier height and the ideality factor. Capacitance – Voltage characterization technique is also used to obtain the value of the carrier concentration of both GaN and SiC SBDs and also. Thermally Stimulated Capacitance (TSCAP) graph was used on GaN SBDs device to detect the traps and their concentrations. Charge based – Deep Level Transients Spectroscopy

(Q-DLTS) mechanism was applied to both GaN and SiC SBDs for the investigation of the deep charge trapping levels in both devices. The measurements employed included Schottky output characteristics at room temperature and at different temperature values.

It is concluded from the experiments that the barrier height for both devices is increasing with the increase of the temperature whereas the ideality factor is decreasing with the increase of the temperature. The values of the barrier height and the ideality factor of GaN Schottky diode are 0.35 eV and 1.2 at 120K and 0.93 eV and 0.47 at 430K, respectively. The value of the barrier height and the ideality factor of SiC Schottky diode are 0.36 eV and 1.5 at 120K and 1.14 eV and

0.4 at 430K, respectively. Three different regions were selected to calculate the carrier concentration of the SiC and GaN SBDs from the C-V characteristics at room temperature. The carrier concentration of the SiC remains constant through the three regions while the carrier concentration of GaN device increases as the reverse bias increases. Two traps have been found by applying the TSCAP technique to GaN Schottky barrier diodes. The first trap was located at

200 K with a concentration of 2.28x1018 cm-3 and the second trap was located at 300 K with a iii

concentration of 3.56x1017 cm-3. For Q-DLTS measurements, unfortunately no traps have been detected for both the GaN and SiC SBDs and therefore no DLTS signals can be shown from the this experiment.

iv

TABLE OF CONTENTS

List of Figures………...……………...……………………………………………………...…...vii

List of Tables……..……………...………………………………………………………………..x

Acknowledgements………………………...…………………………………………………...... xi

Chapter 1: INTRODUCTION………………………………………………………………...…..1

Chapter 2: Gallium Nitride: Composition, Properties and Applications……..………...………....3

2.1 Why GaN?...... ……………………………………………….3

2.2 Structural and Physical Properties of GaN…..………………………………………..4

2.3 Applications of GaN……………………………………………………………....…..8

2.3.1 Light Emitting Diodes……………………………………………………....8

2.3.2 Color Displays………………………………………………………………8

2.3.3 Solar Cell Devices…………………………………………………………..9

2.3.4 Photodetectors………………………………………………………………9

Chapter 3: SHOTTCKY AND OHMIC CONTACTS…...……………………………………...10

3.1 Introduction……………………………………………………………………...…..10

3.2 Schottky Contacts……………………………………………………………………10

3.3 Ohmic Contacts…………………………………………………………………...…18

3.4 Current – Voltage (I-V) Relationship in Schottky Contact…….……………..……..21

3.5 Capacitance – Voltage Characteristics (C-V) of a Schottky Contact………………..23

Chapter 4: EXPERIMENTAL PROCEDURE..……...………………………………………….25

4.1 The Structure of LED Wafer………,……………………...………………………...25

4.2 The Schottky Diode Fabrication………………………..……...…………………….28

4.3 Current-Capacitance-Voltage Characterization Methods and Equipment……..…….32 v

4.4 Charge Deep-Level Transient Spectroscopy Measurement Methods and Equipment...... 34

Chapter 5: RESULTS AND DISCUSSIONS…………………....………………………………35

5.1 Current – Voltage Characteristics……………………………………………………35

5.2 Capacitance – Voltage Characteristics…………………………………....………….48

5.3 Charge – Deep Level Transient Spectroscopy (QDLTS) Characteristics………..….59

Chapter 6: CONCLUSION AND FUTURE WORK……………………………………………63

6.1 Conclusion…………………………………………………………………………...63

6.2 Future Work………………………………………………………………………….64

REFERENCES…………………………………………………………………………………..66

vi

LIST OF FIGURES

Figure 2.1: GaN Zinc blende structure…...………………………………………………...…...... 5

Figure 2.2: GaN Wurtzite structure ……………………...……………………………………….5

Figure 3.1: Band Diagram of Metal – Semiconductor junction (a) before contact the metal to semiconductor contact is made, and (b) after the contact is established………………………...11

Figure 3.2: Ideal I-V characteristic for Schottky diode………………….………………………12

Figure 3.3: Band diagram of metal semiconductor contact under forward bias conditions.…….15

Figure 3.4: Band diagram of metal semiconductor contact under reverse bias conditions.……..16

Figure 3.5: I-V curve for Ohmic contact……………………...…………………………………18

Figure 3.6: Band diagram of metal-semiconductor Ohmic contact before contact ……...... ….19

Figure 3.7: Band diagram of metal-semiconductor Ohmic contact after contact and at thermal equilibrium…………………………………………………………………………..……….…..19

Figure 3.8: Band diagram of a tunneling Ohmic contact……………………………...…………20

Figure 3.9: C-V and 1/C2-V of schottky diode…..………………………………………………23

Figure 4.1: Structure of the LED device in the commercial wafer…,,,,,,………………………..26

Figure 4.2: SEM Picture of the LED device in the commercial wafer…………………………..27

Figure 4.3: The fabrication processes of the Schottky diode……………..………………..28 & 29

Figure 5.1: Current – Voltage measurements for SiC Schottky diodes at three different temperature values……………………………………………………………………………….37

Figure 5.2: Current – Voltage measurements for GaN Schottky diodes at three different temperature values……………………………………………………………………………….38

vii

Figure 5.3: Forward Current values for GaN and SiC Schottky diodes at different temperature values…………………………………………………………………………………………….39

Figure 5.4: Leakage Current values for GaN and SiC Schottky diodes at different temperature values after saturation…………………………………………………………………………....40

Figure 5.5: Semi-log of (I) vs. (V) for GaN and SiC Schottky diodes at room temperature...... 44

Figure 5.6: Ideality factor vs. Temperature for GaN and SiC Schottky diodes……………..…...45

Figure 5.7: Ln (Is/T2) vs. 1/T for GaN and SiC Schottky diodes………….…………………….46

Figure 5.8: Barrier height vs. Temperature for GaN and SiC Schottky diodes..………………...47

Figure 5.9: Capacitance – Voltage characteristics (C-V curve) for SiC Schottky barrier diodes at room temperature………………………………………………………………………………...50

Figure 5.10: Capacitance – Voltage characteristics (C-V curve) for GaN Schottky barrier diodes at room temperature……………………………………………………………………………...51

Figure 5.11: 1/C2 vs. V for three different regions of SiC Schottky barrier diodes at room temperature………………………………………………………………………………………52

Figure 5.12: 1/C2 vs. V for three different regions of GaN Schottky barrier diodes at room temperature………………………………………………………………………………………53

Figure 5.13: TSCAP for SiC Schottky diode at reverse bias of 1 V……..……..….…....……….56

Figure 5.14: TSCAP for GaN Schottky diode at reverse bias of 1 V………………...….………57

Figure 5.15: Derivative of TSCAP (w.r.t. Temp.) versus temperature measurement for GaN

Schottky diode………….………………………………………………………………………..58

Figure 5.16 : Charge Transients vs. Time for GaN Schottky diode at different temperature values……………………………………………………………………………………...... 61

viii

Figure 5.17 : Charge Transients vs. Time for SiC Schottky diode at different temperature values……………………………………………………………………………………...... 62

ix

LIST OF TABLES

Table 2.1: Basic parameters of GaN in zinc blende and Wurzite structures.………….…..…...…6

Table 2.2: Material Properties of Si, SiC and GaN…………………………………..…...………7

Table 3.1: Workfunctions for some Metals………….…….…………….………….…………...13

Table 3.2: Electron Affinities for some semiconductors……….………………………….…….13

Table 3.3: Richardson constants value for various semiconductors……………………………..22

Table 4.1: Different metals that are used for forming ohmic contact on n-GaN………………...31

Table 5.1: Barrier height and Ideality factor values for GaN and SiC Schottky diodes at different temperatures………………………………………………………………………………...……43

Table 5.2: Carrier Concentration values for both GaN and SiC Schottky barrier diodes…..……49

x

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude first to my advisors, Professor Osama

Awadelkarim and Professor Jian Xu, for advising and guiding me during the course of my research at Penn State. With the absence of their support and leadership, this thesis would not be possible.

I would like to send my appreciation to Professor Zhiwen Liu for his help and effort for selecting me to be a part of The Pennsylvania State University graduate students and also for being the chair of my committee and his suggestion to improve this work.

I am also lucky to be a member of one of the inspiring, supportive and hardworking group of graduate students and postdoctoral fellows. I am so grateful to work with them as my colleagues and friends. Those people include: Dr. Jie Liu, Dr. Zhenyu Jiang, Dr. Guanjun You,

Dr. Li Wang and Mahmoud R. M. Atalla.

It is an honor for me to thank Taibah University and the Ministry of Education in the

Kingdom of Saudi Arabia for the selection and provision of financial support to me. Without the financial help from them, all the accomplishments and the hard work that I have done in my studies would not be possible.

I would like to express my gratitude and appreciation to the Department of Electrical

Engineering, the Department of Engineering Science & Mechanics and the Millennium Science

Complex at The Pennsylvania State University for the provision of courses and laboratory

xi

support. I would like to thank especially Dr. Taka Okamoto, Mr. Derek, Mr. Jeffery Long, and

Mr. Steven Perini for giving me the technical help with experiments and data acquisition.

Finally, I would like to dedicate this thesis to my parents, my wife Amany and my daughter Lauren for their love, support and encouragement.

xii

CHAPTER 1

INTRODUCTION

Group III-V compound semiconductors have been under intense investigation for a long time, but recently they have been used in a number of optical and electronic applications. This group of compound semiconductors has become very important to the fabrication of semiconductors devices. Gallium Nitride (GaN), as a member of group III-V semiconductors, is one of the materials that offered a great potential for improving the performance and efficiency of light emitting diodes (LEDs). Different structure designs of alternating current LEDs (AC –

LEDs) have been proposed to result in low cost fabrication of these devices with high performance qualities. The different fabrication methods of AC – LEDs such as anti – parallel

AC-LEDs, Wheatstone bridge (WB) circuit, and the separation growth of LED epitaxial layers and Schottky barrier diodes (SBDs) on different region of the same substrate have been developed to overcome the AC driving problems and achieving high breakdown voltage SBDs with high efficiency LEDs on the same wafer [1 – 7]. However, in order to achieve the high efficiency LEDs, good Schottky contacts and ohmic contacts are required. There are several studies that have already been published of achieving good Schottky contacts and stable, low resistance ohmic contacts. Thus, investigation of choosing the metal materials of SBDs and ohmic contacts is important to achieve high performance LEDs.

In this thesis, the fabrication of Nickel/Aluminum/Titanium/Gold (Ni/Al/Ti/Au) SBDs on unintentionally doped – Gallium Nitride (u-GaN) is examined. The performance of the fabricated

SBD is characterized using Current – Voltage (I-V) and Capacitance – Voltage (C-V) measurements performed at different temperatures. Similar experiments were performed on commercial Silicon Carbide (SiC) SBDs and the results of these experiments are compared to the 1 results obtained on the u-GaN SBDs. The comparison included important SBD parameters, such as barrier height, ideality factor and doping concentration.

This thesis is organized in the following way:

Chapter 2 introduces GaN in terms of its composition and properties. The Chapter, also, presents some important applications of GaN. The importance of Gallium Nitride in optoelectronics and its suitability for optical signal generation and detection are emphasized.

Chapter 3 describes the theory of the metal – semiconductor contacts e.g. Schottky contacts and ohmic contacts in terms of structure design, band diagrams and the operation of both contacts under forward and reverse biases. Moreover, the Chapter discusses the different characterization techniques that can be applied to characterizing metal – semiconductor contacts are described. These techniques include a current – voltage (I-V) measurements, capacitance - voltage (C-V) measurements, as well as charge based deep level transient spectroscopy (Q-

DLTS).

Chapter 4 discusses the experimental procedures that are used from etching down the

GaN material system to fabricating the Schottcky barrier diodes on top of the u-GaN layer. In

Chapter 4, a discussion on why we choose certain types of metals for the contacts is presented.

The Chapter presents the experimental set-up and procedures used in this study.

Chapter 5 presents the results of the experiments that were done and the discussion of the results obtained in the GaN SBDs. The performance of the GaN SBDs is compared to that of commercial silicon carbide SBDs.

Chapter 6 gives the conclusion and the summary of the work described in this thesis.

2

CHAPTER 2

GALLIUM NITRIDE: COMPOSITION, PROPERATIES, AND APPLICATIONS

2.1 Why GaN?

Gallium Nitride (GaN) is one of the III-V group and it is a direct bandgap semiconductor that is been used in light applications particularly in light emitting diodes (LEDs) since 1980s[8].

GaN can be considered as the second most important semiconductor material after silicon because it can be operated at high temperatures and it has wide band gap energy, which is equal to 3.4 eV. Also, it is the key material for the next generation of high frequency and high power transistors. Deposition of GaN began in 1969 on a sapphire substrate by using hydride vapor phase epitaxy (HVPE) [9], and since then different deposition techniques have been developed such as metal organic chemical vapor deposition (MOCVD). The deposited GaN was used in lighting applications until 1993 when the first high brightness blue LEDs were introduced [9].

GaN offers many advantages that make it the most important semiconductor after silicon. These advantages include:

• High melting temperature.

• High decomposition pressure.

• High breakdown voltage (it can reach over 100V).

• High operating temperature conditions (it can operate above 150°C).

However GaN does not exist as a single bulk crystal like silicon and, hence, it becomes difficult to achieve the full potential of GaN and its use is limited.

3

The breakdown field of GaN materials is 3.5x106 V/cm compared to 3x105 and 4x105

V/cm for Silicon (Si) and Gallium Arsenide (GaAs)[10], respectively. All the advantages mentioned above that GaN offers are the reasons of why researchers are motivated to work on

GaN and also why GaN is considered to have a great potential for use in optoelectronics.

2.2 Structural and Physical Properties of GaN

GaN crystallizes in two different crystal structures:

1. Zinc blende crystal structure.

2. Wurtzite crystal structure.

Both crystallographic structures are similar to each other and there are small differences between the zinc blende GaN and wurtzite GaN structures. The zinc blende structure has a cubic unit cell and one lattice constant a, whereas the Wurtzite structure has a hexagonal unit cell and two lattice constants a and c. In the zinc blende structure, each atom type (Ga or N) is surrounded by four atoms of the other type forming a regular tetrahedron. Also, in the Wurtzite structure, each atom type is surrounded by four atoms of the other type. However, in the

Wurtzite structure atoms are located at the edges of the tetrahedron and the structure is hexagonal. Figures 2.1 and 2.2 below show the zinc blend GaN and Wurtzite GaN structures, respectively.

4

Figure 2.1 GaN Zinc blende Structure (Figure taken from [11])

Figure 2.2 GaN Wurtzite Structure (Figure taken from [11])

5

GaN in the zinc blende structure is metastable, whereas in its Wurzite structure GaN is stable. The basic parameters for both zinc blende and Wurtzite GaN structures are summarized in

Table 2.1 below and it can be seen from the Table that there is a temperature dependence on the parameters of both structures.

Property / Material Cubic GaN Hexagonal GaN

Structure Zinc Blende Wurtzite

3.28 eV at 0K 3.47 eV at 0K Energy Gap Eg 3.2 eV at 300K 3.39 eV at 300K

Electron Affinity 4.1 eV at 300K 4.1 eV at 300K

a = 0.3189 nm Lattice parameters 0.450 nm at 300K c = 0.5185 nm

Density 6.10 g.cm-3 6.095 g.cm-3

Nature of Energy gap Direct Direct

Refractive Index 2.9 at 3 eV 2.67 at 3.38 eV

Table 2.1 Basic parameters of GaN in zinc blende and Wurzite structures

In this thesis, GaN is compared to SiC, which is another wide band gap semiconductor that shows a great promise for future power applications. Table 2.2 below shows a comparison between the properties of GaN and the properties of silicon and silicon carbide.

6

Material Property Si SiC GaN

Band Gap (eV) 1.1 3.2 3.4

Breakdown Field (V/cm) 0.3x106 3x106 3.5x106

Electron Mobility (cm2/v-sec) 1450 900 2000

Thermal Conductivity (Watt/cm2 K) 1.5 5 1.3

Electron Saturation Velocity (cm/sec) 10x106 22x106 25x106

Table 2.2 Material Properties of Si, SiC and GaN

From the table above, we can see clearly that the electron saturation velocities of GaN and SiC are higher than the electron saturation velocity of Si, which allow the devices that made on GaN and SiC to operate at higher frequencies. Similarly, the high breakdown fields for both

GaN and SiC compared to Si enable GaN and SiC devices to operate at high power because they can sustain higher voltages and lower leakage currents.

The bonding of GaN is a mix between the ionic and covalent bondings. Based on the distance between the gallium and nitrogen atoms in both zinc blende and wurtzite structures, the

Ga – N bond is much stronger than the Ga – Ga bonds or N – N Bonds. The energy gap of a semiconductor is determined by the bonding strength and the large ban gap of GaN, ~ 3.4 eV, is an evidence of the strong bonding in GaN. Because of the wide band gap, GaN is used for violet, blue and green light emitting diodes. Moreover, GaN is an excellent material for high temperature device applications.

7

2.3 Applications of GaN

GaN is used in several applications including LEDs, color displays and traffic lights [12] as well as in military applications [8]. The large band gap feature of GaN makes it particularly suitable for applications in power electronics and optoelectronics industries. Below are some prominent applications involving GaN.

2.3.1 Light Emitting Diodes (LEDs)

The most important application that utilizes GaN is light emission. GaN is used for making violet, green, and blue light emitting diodes. Different companies around the world are involved in fabricating GaN based LEDs devices and efforts are focused in developing the new technology of white GaN based LEDs. This technology will replace current lighting technologies and is projected to reduce the electricity consumption worldwide by 50% in 2030 [13].

2.3.2 Color Displays

Another important use of GaN is in color display technology. The short wavelength emission and absorption capabilities of GaN give it a wide application potential in color conversion [14]. Several developments had been made in the recent years to use GaN for full color displays. This was not the case before since it was not possible to cover the necessary color spectrum using the same material.

8

2.3.3 Solar Cell Devices

One of the important applications of GaN technology is in GaN based solar cell devices.

This type of solar cells can achieve more than 50% conversion rate in theory if it is doped with

Indium (In)[15]. Also, multilayers are required in order to achieve high efficiency and the lattice matching is not an issue in this type of solar cells. These InGaN solar cells have many advantages including high heat capacity, high efficiency, and resistance to degradation by radiation. The disadvantages of GaN-based solar cells, however, are stresses generated by the multilayer stack necessary for the solar cell and the high fabrication cost.

2.3.4 Photodetecors

Photodetectors are another application that can utilize GaN in their fabrication.

Photodetectors are defined as sensors of photons. The performance of photodetectors is characterized in terms of two factors: first, the noise resulting from device architecture and operating environment; and secondly, the light detection sensitivity [16]. Some of the applications of GaN based photodetectors have already been used for include engine/flame monitoring and detection, plant/ vegetation growth monitoring, UV astronomy, and gas detection [17]. GaN-based photodetectors find several applications in harsh operating environment where many other types of photodetctors fail to operate.

9

CHAPTER 3

SHOTTCKY AND OHMIC CONTACTS

3.1 Introduction

Metal – semiconductor contacts are the most important component in the structure of the semiconductor devices to be connected with the outside world. Two types of metal – semiconductor junctions are often present in each semiconductor device and involved in the fabrication of solid-state devices. The first type is the rectifying contact, which is known as

Schottky contact and the second type is the non-rectifying contact that is known as Ohmic contact. The rectifying contact or the Schottky contact allows current to flow in one biasing direction and blocks current in the opposite biasing direction. This implies that the resistance to current flow in the blocking direction is very large. The Ohmic contact, on the other hand, has low resistance that allows current to flow in both biasing directions equally well. The qualities of both of types of contacts play important role in the performance of the integrated circuit devices

[18]. In this chapter, we will discuss the properties of Shottky and Ohmic contacts and the characterization techniques that are used to test the integrity of these contacts.

3.2 Schottky Contacts

Schottky contacts play an important function in all the GaN devices. An important attribute of the Schottky contact is the barrier height that is shown in the band diagram of figure

3.1. Part (a) of the figure is the metal – semiconductor junction before the contact is made between the metal and the semiconductor, and part (b) of the figure is the thermal equilibrium band diagram of the Schottky contact for the case of a larger workfunction of the metal than that of the semiconductor [18]. 10

Figure 3.1: Band Diagram of Metal – Semiconductor junction (a) before contact the metal to semiconductor contact is made, and (b) after the contact is established (Figure taken from [18])

11

The ideal current – voltage curve of the Schottky diode is shown in figure 3.2 and more details about the current-voltage (I-V) characteristics are given later in this Chapter.

Figure 3.2: Ideal I-V characteristic for Schottky diode (Figure adopted from [20])

The vacuum level in the figure 3.1a is the reference for the band diagram. ∅m is the work function of the metal and ∅s is the work function of the semiconductor (both of the workfunctions are measured in volts). � is an abbreviation for the electron affinity. Table 3.1 and

Table 3.2 show the workfunction values for some metals and the electron affinity for various semiconductors.

12

Metal Elements Workfunction Value (measured in Volts)

Ag, Silver 4.26 Al, Aluminium 4.28

Au, Gold 5.1 Cr, Chromium 4.5 Mo, Molybdenum 4.6 Ni, Nickel 5.15 Pd, Palladium 5.12 Pt, Platium 5.65 Ti, Titanium 4.33 W, Tungsten 4.55

Table 3.1: Workfunctions for some Metals

Semiconductor Elements Electron Affinity Value

Ge, Germanium 4.13

Si, Silicon 4.01

GaAs, Gallium Arsenide 4.07

AlAs, Aluminum Arsenide 3.5

Table 3.2: Electron Affinities for some semiconductors

13

∅B0 is called the barrier height of the semiconductor side of the contact and it is calculated by the difference between the work function of the metal (∅m) and the electron affinity of the semiconductor (�) [18]:

∅B0 = ∅m − � (3.1)

The function of the barrier height is to control the movement of the electrons from the metal side to the semiconductor side in n-type semiconductor material (electrons are majority carriers and holes are minority carriers). In p-type semiconductor material (holes are majority carriers and electrons are minority carriers), the barrier height controls the transport of the holes from the metal side to the semiconductor side. The electrical characteristics of the metal – semiconductor interface are determined by the barrier height and because of that it is considered

[19] the most crucial parameter of the Schottky contacts . Vbi is the built-in potential barrier that the electrons need to surmount in order to move from the semiconductor side to the metal side.

The built-in potential is calculated by the difference between the barrier height and the work function of the semiconductor.

Vbi = ∅B0 - ∅s (3.2)

When a voltage is applied, the metal – semiconductor junction is no longer in equilibrium and difference in the Fermi level (EF) position in the semiconductor and the Fermi level in the metal is determined by the applied bias. In a Schottky contact in an n-type semiconductor under a forward bias (V > 0) condition, which means the positive voltage applied to the metal side and the negative voltage applied to the semiconductor side, the built-in potential is reduced and the electrons can flow easily from the semiconductor to the metal. In forward bias condition in a

Schottky contact on an n-type semiconductor, the Fermi level position in the semiconductor is

14 higher than the Fermi level position in the metal. As the forward bias is increased, the current also increases. On the other hand, a Schottky contact on n-type semiconductor under reverse bias

(V < 0), whereby positive voltage is applied to the semiconductor side and a negative voltage applied to metal side, the Fermi level position in the semiconductor is below that of the Fermi level in the metal side. Because of this the built-in potential increases and the current from the semiconductor to the metal becomes very small. The applied voltage does not affect the barrier height and it remains constant for both bias conditions. The equations for the built in potential under forward and reverse bias conditions are given by:

Vbi – Va = ∅B0 - ∅s (Forward Bias) (3.3)

Vbi + VR = ∅B0 - ∅s (Reverse Bias) (3.4) where Va is the forward-bias voltage and VR is the reverse-bias voltage applied to the Schottky.

Figures 3.3 and 3.4 below show the band diagrams of the metal – semiconductor junction under forward- and reverse-bias conditions, respectively.

Figure 3.3: Band diagram of metal semiconductor contact under forward bias conditions (Figure

adopted from [18]) 15

Figure 3.4: Band diagram of metal semiconductor contact under reverse bias conditions (Figure

adopted from [18])

W is an abbreviation for the depletion layer width or space charge region width. It is defined as the depth from the metal-semiconductor interface over which band bending takes place inside the semiconductor. See figure 3.1b where the depletion region is located between the interface of the metal and the semiconductor, at x=0, and the edge of the depletion layer in the semiconductor. The depletion is analyzed by applying Gauss’s law to find the electric field and the potential across the semiconductor as functions of position (x). Then by applying the

16 condition that the potential across the semiconductor must equals to the built-in potential minus the applied voltage, the depletion region width found to be [18]:

!∈!(!!! !!) �! = (3.5) !!!

Where ∈! is the dielectric constant of the semiconductor, �! is the built-in potential, �! is the applied voltage and �! is the donor density in the n-type semiconductor.

To have Schottky diode operating optimally, the following conditions are to be satisfied[20] :

• The ideality factor should equal to one.

• Large breakdown voltages.

• High switching speed.

• Small turn-on voltage.

• Very short time for reverse recovery (pico- or nano-second).

The above conditions will be discussed in greater details in the following sections.

Schottky diodes are used in different applications including:

• Solar cells.

• Photodetectors,

• Light emitting diodes (LEDs),

• Metal oxide semiconductor field effect transistors (MOSFETs),

• Metal gate field effect transistors (MESFETs),

• Digital switches.

17

3.3 Ohmic Contacts

Ohmic contact is the other type of the metal – semiconductor contacts and it is a non- rectifying contact since it has low resistance, which is independent of the biasing polarity. The

Ohmic contact is important to complete the fabrication process of any semiconductor device and to apply any electric signal to the device. The current – voltage relationship of the Ohmic contact, shown in figure 3.5, is linear irrespective of the sign of the applied voltage signal. The energy band diagram for the Ohmic contact before the metal and semiconductor come in contact and after contact and at thermal equilibrium for n-type material under the condition (�! < �!) are shown in figures 3.6 and 3.7, respectively [18].

I

V

Figure 3.5: I-V curve for Ohmic contact

18

Figure 3.6: Band diagram of metal-semiconductor Ohmic contact before contact (Figure adopted

from [18])

Figure 3.7: Band Diagram of metal-semiconductor Ohmic contact after contact and at thermal

equilibrium (Figure adopted from [18]) 19

Metal – semiconductor Ohmic contacts can be fabricated in two ways [18]. The first way of obtaining an Ohmic contact is shown in figure 3.7. In this type of Ohmic contact, if the positive voltage is applied to the metal side, the electrons can easily move from the semiconductor to the metal since there is no barrier formed, but if the positive voltage is applied to the semiconductor, there is a small barrier height and the electrons can easily flow from the metal to the semiconductor [18]. The second type of Ohmic contacts is the tunneling ohmic contacts. In this type, the depletion region width is very small because the n-type semiconductor material is heavily doped so that electrons can tunnel through the barrier back and forth between the metal and the semiconductor. If the doping concentration increases that means the semiconductor become more heavily doped, the deletion region width decreases and the probability for an electron to tunnel becomes large. Figure 3.8 shows the band diagram of a tunneling Ohmic contact.

Figure 3.8: Band diagram of a tunneling Ohmic contact. (Figure adopted from [18]) 20

3.4 Current – Voltage (I-V) Relationship in Schottky Contact

The current – voltage characteristics of Schottky diode are discussed and analyzed in ref.

21. The current-voltage relationship in a Schottky is [21]:

!! � = � [exp( ! ) - 1] (3.6) ! !"#

where,

• q is the charge of the electron.

• VD is the applied voltage across the junction.

• K is the Boltzmann constant.

• T is the temperature.

• Is is the reverse saturation current

• n is the ideality factor and it is a dimensionless parameter.

This forward bias current –voltage in the Schottky diode results from the thermionic emission theory of the electrons over the potential barrier. The reverse saturation current is expressed by [21]:

∗ ! −�∅! �! = �� � exp[ ��] (3.7)

where

• A is the area of the diode.

• A* is the Richardson constant.

• ∅! is the effective barrier height.

21

Table 3.3 summarize some of the Richardson constants value for various semiconductors:

Semiconductor A* (A/cm2. K2) Reference

n-Si 112 (±6) 22

P-Si 32 (±2) 22

n-GaAs 4-8 23

n-GaAs 0.41 (±0.15) 24

P-GaAs 7 (±1.5) 24

n-InP 10.7 25

n-GaN 26.4 26

Table 3.3: Richardson constants value for various semiconductors

The barrier height of the Schottky diode can be calculated from the reverse saturation current equation and it is independent of the bias. The equation of the barrier height is expressed as:

!" !!∗!! ∅! = ln( ) (3.8) ! !!

The value of the thermal voltage, which is KT/q at room temperature is ~ 26 mV and the

Richardson constant for n-type GaN is 26.4 A/cm2. K2. The value of the reverse saturation current can be found by plotting the semilog of I versus V and extrapolating the semilog of I to

V=0. The ideality factor can be determined from the slope of the linear curve of semilog of I versus V and it can be expressed by:

22

� = ! (3.9) !.! !"#$% !"/!

The barrier height and the ideality factor are dependent on temperature. The barrier height increases with increasing temperature, whereas the ideality factor decreases with increasing temperature. Another way to obtain the barrier height is to plot the natural log of the current density / square of the temperature versus the inverse of the temperature and the slope of that linear curve is equal to the barrier height.

3.5 Capacitance – Voltage Characteristics (C-V) of a Schottky Contact

The ideal capacitance – voltage characteristic curve of a Schottky diode is shown in figure 3.9:

Figure 3.9: C-V and 1/C2-V of Schottky diode (Figure adopted from [27])

23

As shown in figure 3.9, the relationship between 1/C2 versus voltage is a straight line and the carrier concentration can be determined from the slope of the linear line. The carrier concentration is given by [27]:

! �! = ! (3.10) !∈!∈!! !"#$%

where A is the area of the diode, ∈! is the permittivity of the of free space and ∈! is the permittivity of the semiconductor which is 9.7 or 8.9 for GaN in zinc blende or Wurtzite structures, respectively [28]. The flat barrier height of a Schottky diode is given by:

�!" = �!" + �! (3.11)

Vbi is the built in potential and it is equal to the difference between the thermal voltage and the intercept value of the straight line of (1/C2) versus (V) at (1/C2=0) [27].

�!" = �! − �! (3.12)

So the barrier height equals to:

�!" = −�! + �! + �! (3.13)

�! = ��/� is the thermal voltage and its value at room temperature is 26mV. �! can be calculated using the equation [27]:

!" !! �! = ln (3.14) ! !! where Nc is the effective density of states in the conduction band and its value for GaN is 2.8 x

1018 cm-3. The effective density of states is given by [27]:

∗ � = 2(!!! !")!/! (3.15) ! !!

∗ � is the electron effective mass in GaN and it is equal to m0 0.22 where m0 is the electron rest mass, which is 9.11x10-31 kg.

24

CHAPTER 4

EXPERIMENTAL PROCEDURE

In this chapter, the GaN Schottky diode fabrication process is described. In addition to fabrication, the experimental procedures followed in characterizing the diodes are also described in this chapter in order to understand the behavior of the GaN Schottky diode and compare to a commercial SiC Schottky diode. The characterization experiments performed are current-voltage

(I-V) and capacitance-voltage (C-V) measurements done at different temperatures, including room temperature to study the effects of temperature on the output of the device and extract some important diode parameters. Charge deep-level transient spectroscopy (Q-DLTS) is used to investigate the defects in the GaN diode and find their concentrations and locations in the diode, whether at the metal/GaN interface or in the bulk GaN, as well as in the bandgap. Q-DLTS gives information on the defect charge state and its carrier trapping with respect to charge transition as the carrier is captured – electron trap or hole trap.

4.1 The structure of LED wafer

The LED wafer is purchased from a company in China and the Schottky barrier diodes are fabricated on the top of this commercial 2-inch LED wafer. The material system of the commercial 2-inch LED wafer is shown in figure 4.1 and the SEM picture of the wafer is shown

18 -3 in figure 4.2. From top to bottom the wafer layers are: a ~200 nm p-type (Na> 1×10 cm ) GaN

17 -3 capping layer; a ~200 nm p-type (Na≈ 5×10 cm ) GaN layer; a multiple quantum well (MQW) emissive layer consisted by fifteen pairs of 2.5nm In0.1Ga0.9N/12 nm-GaN, a ~2.4 um n-type (Nd

≈ 1×1018cm-3) GaN layer; and a ~2.5 um GaN buffer layer on a patterned sapphire substrate.

25

p-GaN

MQW

n-GaN

u-GaN Sapphire

Figure 4.1: StructureSubstrate of the LED device in the commercial wafer

26

Figure 4.2: SEM Picture of the LED device in the commercial wafer

27

4.2 The Schottky Diode Fabrication

In order to fabricate the Schottky barrier diodes on the LED commercial wafer, several processes are involved such as etching cycles, KOH treatments, and depositions. The process of fabricating the Schottky diode atop the u-GaN layer is shown in figures 4.3.

p-GaN ICP etching

MQW

n- GaN

u-GaN

Sapphire Substrate

28

p-GaN

MQW

Ohmic n -GaN Schottky

u-GaN

Sapphire Substrate

Figure 4.3: The fabrication processes of the Schottky diode

The first step in the diode is etching through selective regions of p-type GaN and MQW to expose the n-type GaN using Chlorine based Inductive Coupled Plasma Reactive Ion Etching

(ICP RIE) system. The etch rate is 2 nm/s and the etch depth is about 1.2 µm. In order to expose the unintentionally doped GaN acting as the Schottky contact surface, the n-type GaN needs to be dry etched down to the u-GaN. However, the ICP RIE process would introduce etching defects and consequently decrease the breakdown voltage of the resulted SBDs. In order to remove the etching defects, various post treatment methods have been used, such as annealing and KOH surface treatment. Nevertheless, these methods are not entirely sufficient for producing a SBD on LED wafer with high enough breakdown voltage. 29

The mixed etching procedure is composed of three mixed etching cycles. In each cycle, the Schottky contact area was dry etched with Chlorine based ICP RIE tool while the unetched area is protected with Ni etch mask. After dry etching, the surface of the Schottky contact area is exposed to boiling 0.5 M KOH solution for 10 minutes for surface treatment. Employing this mixed etching technique, the surface is “refreshed” after each dry etching cycle, thus the accumulation of the etching defects is prevented. In order to further decrease the introduced etching defects, the etch rate is set at 2 nm/s for the first two cycles, and the etch depth is 0.9 µm for each cycle. In the last mixed etching cycle, the etch rate is decreased to 0.2 nm/s and the etch depth is 0.1 µm. With this etching rate setting, the process time would not be too long and the introduced defects can be decreased due to the lower etching rate of the last etching cycle. Next the device is annealed at 750°C for 2 minutes in Nitrogen (N2) atmosphere in a rapid thermal annealing (RTA) furnace.

The metals that are deposited on the top of the u-GaN layer to form the Schottky contact are Ni/Al/Ti/Au (50nm/500nm/100nm/200nm) and the electrode area of the Schottky contact is equal to 2.5 x 10-5 cm-2. The Ohmic contact is formed on the top of the n-type GaN layer by e-

o beam evaporation followed by 500 C annealing for 1 minute in N2 environment and the metals that formed the Ohmic contact are Ti/Al/Ti/Au (10nm/40nm/40nm/100nm). The reason of fabricating the Schottky diodes and the LEDs on the same wafer is to get high switching speed for bi-directional optical wireless communication (OWC) [29]. Several metals are used for

Schottky diodes contacts and among these metals are Pt[30 – 32], Pd[31 – 33], Au[34 – 37], Ni[36 -39], Re

[40], and Ag [37]. Also, several metals have been used for the ohmic contact on n-GaN material and these metals are summarized in table 4.1[41].

30

Metallization (nm) Annealing Conditions Carrier Concentration (cm-3) Ti/Al/Ni/Au (15/220/40/50) 900oC, 30 sec 4 x 1017

Ti/Al (20/100) 900 oC, 30 sec 1017

Ti/Al (35/115) 600 oC, 15sec 5 x 1017

Ti (20) 975 oC, 30 sec 5 x 1017

Ti/TiN (5/200) 800 oC, 60 sec 7 x 1017

Ti/Ag (15/150) No anneal 1.7 x 1019

Ti/Ni (5/25) 1040 oC,30 sec 1 x 1018

Ti/Au (3/300) No anneal 4 x 1020

Ti/Pd/Ni (5/5/25) 990 oC, 20 sec 1 x 1018

TiN (200) 800 oC, 60 sec 7 x 1017

Al (150) 600 oC, 60-480 sec 7 x 1017 in Ar/H2 Al (250) No anneal 5 x 1019

W (50) 600-1000 oC,60sec 1.5 x 1019

Zr/ZrN (20/80) 1000 oC,60sec 2 x 1018

Pd/Al (12.5/100) 650 oC, 30 sec 2.8 x 1017

Ta/Al (35/115) 600 oC, 15sec 7 x 1017

Table 4.1: Different metals that are used for forming Ohmic contact on n-GaN

31

4.3 Current-Capacitance-Voltage Characterization Methods and Equipment

In this section, the equipment used in characterizing the Schottky diodes is described.

The Schottky diodes are mounted and tested in a micromanipulator. The SBD device is placed on a conducting chuck of the micromanipulator, and is held down by a sucking vacuum to the chuck. Two measurement probes are applied to the SBD by the help of a microscope. The microscope is necessary for the proper placement of the probes, as the devices are too small to be viewed with the naked eye. The two probes connect the device to Keithley 238 Source

Measuring Units (SMUs) and Keithly 590 Capacitance – Voltage (CV) measurement unit. The

SMU’s and the C-V measuring unit, when connected to a power supply, are capable of providing the voltage scan and measuring the currents and the capacitance, respectively. The SMUs and the

C-V measuring unit are also capable of interfacing with computer software for easy use.

Two software programs were used for the collection of data. The first, and most important program is metric ICS software. The ICS software features allow the users to perform device characterization, process monitoring, failure analysis, incoming inspection, and process development [42]. The software is capable of connecting to the SMU’s, and could control all the necessary parameters needed for experimentation by instructing which probe would apply voltage to the SBD for a set period of time. The software is also capable of providing sweeps of finite voltages, with a difference between data points specified by the software, while measuring a current at each point.

The second equipment is also used for measuring I-V characteristics at room temperature and also at different temperatures ranging from 40 K to 320 K is done at another probe station, which is a Keithley 2612 current source meter. The use of Keithley 2612 enables measuring currents in the ampere range which is higher than that measured by Keithley 238. The voltage 32 and current sweeps and measurement are controlled by LabTracer Software. For different temperature measurements, a Lake Shore 336 temperature controller source is connected to the probe station and the chamber evacuated and thermally insulated. Helium gas is used to cool down the sample and cycle the temperature between 40 k and 320 K.

The third equipment is used for measuring the I-V and C-V at room temperature and also at different temperatures. The temperature values for this experiment are ranging from liquid nitrogen (around 70 K) to 473 K. The device is placed inside a sample holder, which is a small box that consists of Aluminum and the sample holder is placed inside a heater furnace dd9010.

The furnace is connected to pA meter HP4140, CD meter HP4284 and to a temperature controller HP34401. All of these devices are connected to Hewlett-Packard (HP computer). The pA meter and the CD meter are connected together through a switch NI6008 and both of them are connected to the furnace and a temperature controller. The temperature is lowered inside the furnace by flowing liquid nitrogen through the pipes, which are connected to the furnace. Inside the sample holder, the temperature is detected and recorded by a K-type thermocouple. This thermocouple is held down near to the SBDs device and also is connected to the temperature controller in order to give the temperature of the sample.

Three software programs were utilized in the I-V and C-V measurements and for collecting data. The first program is Notepad that is used for writing the commands and the second program is GADD software that is used to run the notepad commands for doing the measurements. The third software is called Visualize program and it is used for providing the graphs of both I-V and C-V. The Notepad output file may be opened by Visualize program for the I-V and C-V curves or the Notepad data may be transferred to an Excel spread sheet.

33

4.4 Charge Deep-Level Transient Spectroscopy Measurement Methods and Equipment

Another experiment that used the same equipment as above and also it is performed in this work on both GaN and SiC SBDs is the Charge-Deep Level Transient Spectroscopy (Q-

DLTS). The experiment used the same furnace dd9010, the same temperature controller

HP34401 and the new equipment, which is the DLTS source unit GAD d6341. The DLTS source unit is connected to the furnace and also to the computer for controlling the parameters that are set for the measurement. The SBD device is placed in the sample holder and the sample holder is put inside the furnace. The thermocouple is used to measure the temperature of the sample holder and is connected to the temperature controller outside the furnace. All the three-software programs that are used for the I-V and C-V measurements are also used for the DLTS measurements. The DLTS signal is measured over the temperature range 120 K to 430 K.

34

CHAPTER 5

RESULTS AND DISCUSSIONS

In this chapter, the data that obtained from the experimental results are presented and discussed. First, I will show the results that found from the Current – voltage (I-V) characteristics of both GaN and SiC Schottky diodes. The two devices will be compared in terms of the forward and leakage currents. Also, the ideality factors and the barrier heights of both devices are presented and discussed in details. Second, both devices will be discussed individually in terms of Capacitance – Voltage (C-V) characteristics. The results that have been found from this experimental measurement will show the doping concentration of both devices and also the relationship between the traps and their activation energies. Finally, the results of

Charge deep level transient spectroscopy (Q-DLTS) will be presented and the reasons of how this measurement is not successful for both devices will also be described. All the electrical measurements of the three techniques are explained in the previous chapter and those experiments are done at room temperature and also at different temperatures.

5.1 Current – Voltage Characteristics

The current – voltage measurements of both SiC and GaN Schottky diodes at different temperature values with the inclusion of the room temperature are shown in figure 5.1 and figure

5.2, respectively. The temperature values that have been selected in both graphs are 120K, 300K and 430K in order to see if there any differences of the behaviors of both devices I-V curves with the change of the temperature. It is clear from the figures that the behaviors are not changing and also the turn-on voltages are approximately the same for both devices, which are around 0.5 V.

35

Figure 5.3 is the forward current at different temperature values for both Schottky diodes devices at one selected voltage value, which in this case is selected to be at 1V since it is bigger than the turn – on voltage value. Theses graphs show the behavior of both devices after the turn – on voltage at different temperatures to find if the current values increasing smoothly or if there are any unexpected behaviors that will affect the performance of the device. From the figure, the commercial SiC device is increasing linearly which is an obvious case since the forward current is depending on the temperature value. For GaN device, the forward current is also increasing with the increase of the temperature but this behavior is different from the behavior of the SiC device because the values of the forward current of GaN device show a jumping increase especially between temperatures of 160K and 250K and also between the temperature values of

250K and 350K. This means that there are some kinds of traps between these temperature values and this result will be discussed in details later because it is also related to the capacitance vs. temperature graph at one selected voltage value. Figure 5.4 shows the behavior of the leakage currents of the GaN and SiC Schottky diodes after saturation at different temperature values. It is clear from the figure that both devices are saturated at V= -24.5 V and the current values at this voltage value are in the range of nA and pA for GaN and SiC devices, respectively.

36

I-V Characteristics for SiC

2.50E-03

2.00E-03

1.50E-03

120 K 1.00E-03 300 K

Current Current 430 K 5.00E-04

0.00E+00 -3.00E+01 -2.50E+01 -2.00E+01 -1.50E+01 -1.00E+01 -5.00E+00 0.00E+00 5.00E+00 1.00E+01

-5.00E-04 Voltage

Figure 5.1: Current – Voltage measurements for SiC Schottky diodes at three different

temperature values.

37

I-V Characteristic for GaN

2.50E-02

2.00E-02

1.50E-02

120 K 1.00E-02 300 K

Current Current 430 K 5.00E-03

0.00E+00 -30 -25 -20 -15 -10 -5 0 5

-5.00E-03 Voltage

Figure 5.2: Current – Voltage measurements for GaN Schottky diodes at three different

temperature values.

38

Forward current at Voltage = 1 V

4.00E-04

3.50E-04

3.00E-04

2.50E-04

2.00E-04 GaN 1.50E-04 Current Current SiC 1.00E-04

5.00E-05

0.00E+00 0.00E+00 1.00E+02 2.00E+02 3.00E+02 4.00E+02 5.00E+02 Temperature

Figure 5.3: Forward Current values for GaN and SiC Schottky diodes at different temperature

values.

39

Leakage Current

1.00E-08

0.00E+00

-1.00E-08

-2.00E-08

-3.00E-08 GaN

current current SiC -4.00E-08

-5.00E-08

-6.00E-08 0.00E+00 1.00E+02 2.00E+02 3.00E+02 4.00E+02 5.00E+02 Temperature

Figure 5.4: Leakage Current values for GaN and SiC Schottky diodes at different temperature

values after saturation.

40

One of the important parameters that can be extracted from the I-V characteristics of the

Schottky diodes is the ideality factor. The ideality factor is a parameter that can determine the deviation of the device from an ideal diode case (n=1), which means that there are impurities and defects in the device that reduces the current and causes recombination. The value of the ideality factor of both GaN and SiC Schottky diodes at room temperature are calculated from the slope of the semi-log of (I) versus (V) graph that is shown in figure 5.5. Both devices show different values at two different regimes for the same devices. The first regime that is assumed to be the correct region of calculating the ideality factor shows the values of 0.66 and 0.76 while the second regime of the same graphs shows the values of 5.15 and 2.1 for both GaN device and SiC device, respectively. The first region is claimed to be the accurate region of calculating the ideality factor because it shows the domination of the thermionic emission in the conduction process of the Schottky diodes while the second region shows different factors such as the increase of defects number that affect the conduction process of the devices [43 – 44]. More defects means more recombination in the space charge region of the semiconductor and because of that the values of the ideality factor are much greater than unity and they can’t be considered as the accurate values of the ideality factor. Figure 5.6 shows the dependency of the ideality factor on the temperature since the ideality factor is inversely proportional to the temperature. As the temperature increases, the ideality factor should decrease and as the temperature decreases, the ideality factor should increase and it is exactly the case in this work for both GaN and SiC devices. The value of the ideality factor for GaN Schottky diodes device at 120K is 1.2 and at

430K, the ideality factor decreases to 0.47. For SiC Schottky diodes device, the value of the ideality factor at 120K is 1.5 and it drops to 0.4 when the temperature increases to 430K. The reasons of why the ideality factor decreases with the increase of the temperature are the increase 41 of the current flow in the space charge region and also the carrier recombination in the bulk region caused by the defects [45]. The decrease of the ideality factor with the increase of the temperature shows that the thermionic emission is responsible for the transport mechanism of both Schottky diodes [46].

Another important parameter that is determined by the I-V characteristics is the barrier

2 height of the Schottky diodes. Figure 5.7 shows Ln (Is / T ) vs. (1/T) for both GaN and SiC

Schottky diodes and the barrier heights are calculated from the slopes of the lines of both devices. The values of the barrier heights of GaN and SiC, which are calculated from the slopes, are ranging from 0.2 to 0.9 for the GaN device and from 0.4 to 0.8 for the SiC device. Another method of calculating the barrier height that have been done in this work is by using equation 3.8 since every parameter in the equation is already known and the value of the saturation current can be obtained by plotting the semi-log of I versus V and extrapolating the semi-log of I to V=0.

The values of the barrier heights at room temperature that are found by using the equation of the barrier height are 0.82 for GaN Schottky diodes and 0.78 for SiC Schottky diodes. The first method of calculating the barrier heights, which is called the Richardson plot sometimes is the most common way that is used to find the barrier height values for most of the Schottky barrier diodes. From this plot, the Richardson constant can also be found but the extraction of the

Richardson constant value is impossible when the barrier height and ideality factor depend on the temperature [21]. Figure 5.8 is the dependence of the barrier height values on the temperature for both GaN and SiC devices. The barrier heights of both Schottky diodes are increasing with the increase of the temperature. The value of the barrier height of GaN Schottky diodes is 0.35 eV and goes up to 0.93 eV at 430K. For SiC Schottky diodes, the value of the barrier height is 0.36 eV at 120K and it increases to 1.14 eV at 430K. Another researchers were also done similar 42 experiments for SiC schottky diodes to extract the barrier height and the ideality factor values.

The value of the barrier height and the ideality factor that have been found by them are 0.529 and

4.6 at 150K, respectively and also their values at 400K are 1.1 and 1.9, respectively [47]. The barrier height and the ideality factor values of both GaN and SiC schottky diodes at different temperatures are summarized in Table 5.1. There is a small error that needs to be added to the values of both barrier heights and the ideality factor of both devices. This error is resulting from repeating the experiments of I-V to find the barrier height and the ideality factor. The error values equal to ± 0.1 for the barrier height and ± 0.2 for the ideality factor values for both devices of GaN and SiC Schottky diodes and it is shown in the table below.

Temperature GaN SiC

Value /K n BH n BH

120 1.2 ± 0.2 0.35 ± 0.1 1.5 ± 0.2 0.36 ± 0.1

160 0.9 ± 0.2 0.5 ± 0.1 1.08 ± 0.2 0.45 ± 0.1

200 0.71 ± 0.2 0.62 ± 0.1 0.91 ± 0.2 0.56 ± 0.1

250 0.69 ± 0.2 0.7 ± 0.1 0.8 ± 0.2 0.68 ± 0.1

300 0.66 ± 0.2 0.82 ± 0.1 0.76 ± 0.2 0.78 ± 0.1

350 0.58 ± 0.2 0.89 ± 0.1 0.56 ± 0.2 0.99 ± 0.1

430 0.47 ± 0.2 0.93 ± 0.1 0.4 ± 0.2 1.14 ± 0.1

Table 5.1: Barrier height and Ideality factor values for GaN and SiC Schottky diodes at different

temperatures.

43

Semilog of (I) Vs. (V) at Room Temp.

0 0 0.5 1 1.5 2

-5 GaN region 1

-10 GaN region 2

SiC region 1

-15 SiC region 2

Linear (GaN region 1) Current Current -20 Linear (GaN region 2)

Linear (SiC region 1) -25 Linear (SiC region 2)

-30 Voltage

Figure 5.5: Semi-log of (I) vs. (V) for GaN and SiC Schottky diodes at room temperature.

44

Ideality Factor Vs. Temp.

1.6

1.4

1.2

1

0.8 GaN

0.6 SiC

Ideality Factor Factor Ideality 0.4

0.2

0 0 50 100 150 200 250 300 350 400 450 500 Temperature

Figure 5.6: Ideality factor vs. Temperature for GaN and SiC Schottky diodes.

45

2 Ln (Is/T ) vs. (1/T)

0 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009

-10

-20

) SiC 2 -30 /T

s GaN

-40 Linear (SiC) Ln (I

-50 Linear (GaN)

-60

-70 1/T

Figure 5.7: Ln (Is/T2) vs. 1/T for GaN and SiC Schottky diodes.

46

Barrier Height Vs. Temp.

1.2

1

0.8

0.6 GaN

0.4 SiC Barrier Height Height Barrier

0.2

0 0 100 200 300 400 500 Temperature

Figure 5.8: Barrier height vs. Temperature for GaN and SiC Schottky diodes.

47

5.2 Capacitance – Voltage Characteristics

The Capacitance – Voltage Characteristics for SiC and GaN Schottky Diodes at room temperature are shown in figure 5.9 and figure 5.10, respectively. The capacitance of the GaN

Schottky diodes is smaller than the capacitance of the SiC Schottky diodes. From the figures, it can be detected that the capacitance values of the GaN device are in the range of the peco-farad

(pF) while the capacitance values of the SiC device are in the range of nano-farad (nF). The capacitances of both devices were measured at one frequency that is 1MHz and at voltages sweep range from -25V to 0V.

Figure 5.11 and figure 5.12 shows 1/C2 vs. V at room temperature for SiC Schottky diodes and GaN Schottky diodes, respectively. It can be observed from figure 5.11 that the sample is uniform through all the three different regions of the straight line. The three different regions were selected to find if there are any differences of the carrier concentration values. The carrier concentration of the device was calculated from the slope of the linear curve and it is in the range of 1020 cm-3 for all the three regions. On the other hand, the GaN sample is not uniform through all the three regions and this is can be clearly seen in figure 5.12. The curve of 1/C2 vs.

V is not linear as the reverse bias voltages increases towards the final voltage value of -25V. The carrier concentration that was calculated from the slopes of the three regions shows different values as the reverse bias increases. The results of the calculated values of the carrier concentration of both SiC and GaN Schottky barrier diodes for the different selected regions are summarized in Table 5.2.

48

Regions Nd For GaN (cm-3) Nd For SiC (cm-3)

Region 1 1.27x1018 7.70x1020

Region 2 6.30x1017 7.70x1020

Region 3 4.20x1017 7.70x1020

Table 5.2: Carrier Concentration values for both GaN and SiC Schottky barrier diodes.

The carrier concentration of the GaN device shows an increase as the voltage goes from region to region. The reason of the increasing the carrier concentration and the non-uniformity of

1/C2 vs. V is the existence of the defects such as electron traps in the depletion region width of the semiconductor. The presence of those electron traps are due to the etching process that was done for the LED commercial wafer in order to build the Schottky and Ohmic contacts on the top of the u-GaN and n-GaN layers, respectively.

49

C-V Characteristic for SiC

1.60E-10

1.40E-10

1.20E-10

1.00E-10

8.00E-11

6.00E-11 SiC Capacitence Capacitence 4.00E-11

2.00E-11

0.00E+00 -3.00E+01 -2.50E+01 -2.00E+01 -1.50E+01 -1.00E+01 -5.00E+00 0.00E+00 Voltage

Figure 5.9: Capacitance – Voltage characteristics (C-V curve) for SiC Schottky barrier diodes at

room temperature.

50

C-V Charasterictic for GaN

4.00E-12

3.50E-12

3.00E-12

2.50E-12

2.00E-12

GaN 1.50E-12 Capacitence Capacitence 1.00E-12

5.00E-13

0.00E+00 -3.00E+01 -2.50E+01 -2.00E+01 -1.50E+01 -1.00E+01 -5.00E+00 0.00E+00 Voltage

Figure 5.10: Capacitance – Voltage characteristics (C-V curve) for GaN Schottky barrier diodes

at room temperature.

51

1/C^2 vs. V for SiC

2.00E+20

1.80E+20

1.60E+20

1.40E+20

1.20E+20 Region 1 Region 2 1.00E+20 Region 3

1/C^2 8.00E+19 Linear (Region 1)

6.00E+19 Linear (Region 2)

4.00E+19 Linear (Region 3)

2.00E+19

0.00E+00 -5.00E+00 -4.00E+00 -3.00E+00 -2.00E+00 -1.00E+00 0.00E+00 Voltage

Figure 5.11: 1/C2 vs. V for three different regions of SiC Schottky barrier diodes at room

temperature.

52

1/C^2 Vs. V for GaN

3.50E+23

3.00E+23

2.50E+23

Region 1 2.00E+23 "Region 2"

Region 3 1.50E+23 1/C^2 Linear (Region 1)

1.00E+23 Linear ("Region 2") Linear (Region 3)

5.00E+22

0.00E+00 -5.00E+00 -4.00E+00 -3.00E+00 -2.00E+00 -1.00E+00 0.00E+00 Voltage

Figure 5.12: 1/C2 vs. V for three different regions of GaN Schottky barrier diodes at room

temperature.

53

Those traps are discussed and described by using the thermally stimulated capacitance

(TSCAP) technique by drawing the capacitance vs. different temperatures at one selected voltage value. TSCAP is a technique that gives rapid information about the concentrations and the

[48] activation energies of the defects found in the material . The ionized temperature Tm of the

[49] defects is given when there are observed capacitance steps . The midpoint temperature (Tm) is

[50] related to the activation energy (∆� = �! − �!) of the defect by the equation :

! !"!! ∆� = ��! ln [ ] (5.1) !(∆!!!!!!)

Where EC is the energy level of the conduction band, �! is the energy level of the trap state, b is the heating rate and v is the rate for the electron emission. The trap concentration (Nt) can be calculated by using the relation of Buehler and Philips (1976) [51}:

!!! !! �! = 2�! ( ) (5.2) !!

Where Nd is the carrier concentration of the Schottky diodes, Cf and Ci are the final and initial capacitance. The TSCAP for both SiC and GaN Schottky diodes are shown in figures 5.13 and 5.14, respectively for a reverse bias of 1V and a temperature range from 120K to 430K.

From figure 5.13, the capacitance of the SiC schottky diode is increasing linearly with the increase of the temperature and there is no step increase of the capacitance with respect to the increase of the temperature and this is because there are no traps in the SiC schottky diodes. For

GaN Schottky diode device as shown in the TSCAP figure 5.14, the capacitance is also increasing with the increase of the temperature but there are step increases of the capacitance with respect to the increase of the temperature. These step increases can be explained to the effect of the emission of deep levels. This behavior follows the same behavior of the forward

54 current vs. temperature at 1V that is shown in figure 5.3 and the step increase of forward current is the same step increase of the capacitance at the same temperature values. Two traps can be detected from the TSCAP graph of the GaN Schottky diodes between the temperature values of

160K and 250K and also between the values of the 250K and 350K. To get more details about the traps, the derivative of the TSCAP with respect to the temperature vs. temperature should be drawn in order to get the peaks of the traps. The TSCAP derivative with respect to the temperature vs. temperature is shown in figure 5.15. From the figure, it is observed that the two traps are located at temperatures Tm1 ≅ 200K and Tm2 ≅ 300K. The plot also confirms that the concentration of the trap at Tm1 is higher than the concentration of the trap at Tm2. The concentrations of both traps are calculated by equation 5.2. The concentration of the first trap equals to 2.28x1018 cm-3 while the concentration of the second trap is 3.56x1017 cm-3. The activation energy of both traps are difficult to estimate by using equation 5.1 since the rate of the electron emission rate depends on the capture cross section of the defects which is unknown variable in this case. Zahid etal. classified the defects on n-type GaN material based on their activation energies and temperature range into three categories. Those categories are: (i) the defects that are found between the temperature range of 100 K to 200 K have activation energies smaller than 0.35 eV. (ii) The defects that appear between the temperature range of 200 K and

350 K have activation energies between 0.4 – 0.7 eV. (iii) and the defects that are found between

350 K and 550 K have activation energies of 1.0 eV and higher [52]. From this classification the activation energies of both traps of GaN found by the TSCAP graph can be assumed to be in the range of 0.4 to 0.7 eV.

55

TSCAP Graph of SiC

1.30E-10

1.28E-10

1.26E-10

1.24E-10

1.22E-10

1.20E-10

1.18E-10 SiC 1.16E-10 Capacitence Capacitence 1.14E-10

1.12E-10

1.10E-10

1.08E-10 0 50 100 150 200 250 300 350 400 450 500 Temperature

Figure 5.13: TSCAP for SiC Schottky diode at reverse bias of -1 V.

56

TSCAP Graph of GaN

4.00E+00

3.50E+00

3.00E+00

2.50E+00

2.00E+00

GaN 1.50E+00

1.00E+00 Capacitence (pF) Capacitence

5.00E-01

0.00E+00 0 100 200 300 400 500 Temperature

Figure 5.14: TSCAP for GaN Schottky diode at reverse bias of -1 V.

57

dC/dT versus T measurements 3.50E-02

3.00E-02

2.50E-02

2.00E-02

1.50E-02

GaN 1.00E-02

dC/dT (pF/K) (pF/K) dC/dT 5.00E-03

0.00E+00 100 150 200 250 300 350 400 -5.00E-03 Temperature (K)

Figure 5.15: Derivative of TSCAP (w.r.t. Temp.) versus temperature measurement for GaN

Schottky diode

58

5.3 Charge – Deep Level Transient spectroscopy (QDLTS) Characteristics

The charge deep level transient spectroscopy measurements show unsuccessful results for both GaN and SiC Schottky diodes. The reason of applying the QDLTS technique to both devices is to study and investigate the traps in both devices and to know if the traps are located in the in metal/GaN interface or in the bulk of the GaN. The charge transients of 5V applied external reverse bias for both GaN and SiC Schottky diodes at pulse width equals to 100 micro seconds are shown in figure 5.16 and figure 5.17, respectively. It can be seen clearly from the figures that the transients of the junction is still constant and not falling down to its relaxing state after the pulse is off for both devices which are not the correct transients that both devices should act. The charge transients should fall down exponentially after the pulse is off if there are traps that filled out by electrons by the applied reverse bias and if there are no traps, the transients should come back to the initial voltage value which are not shown in both figures. Because of the same behavior of both devices, no DLTS signal can be detected from both devices and no traps can be found in the band gap or in the surface of the GaN and SiC Schottky diodes. Different reasons have been discussed of why those transients didn’t show the exponential decay or sharp decrease to its initial value. One of the reasons that the charge range of the schottky diodes is too small compared to the charge range of the QDLTS source unit. The staff members who are responsible for the device setup have resolved this problem and they have identified different charge ranges that can detect even the small charges of any schottky diodes. Another reason is that there are some problems with the connection of the device to the QDLTS system but the results of the I-V and C-V characteristics that are shown in this chapter are taken with the same connection and the same equipment. One more reason is that the QDLTS technique is not a familiar technique that can study the traps of Schottky diodes but this reason is not correct since 59 there are several publications that use the QDLTS technique to study the deep level traps of the different Schottky diodes devices. The experiment has been done again after all the problems have been solved and with different applied revers bias voltages and also with different pulse width time but still the transients are not changing and keep to show a constant value of the charge after the pulse is off and the results still the same.

60

Charge Transients vs. Time for GaN Schottky diode

4.50E-08

4.00E-08

3.50E-08

3.00E-08 25 2.50E-08 -21

2.00E-08 -59

-97 Charge 1.50E-08 -121 1.00E-08 -149 5.00E-09

0.00E+00 0 0.0001 0.0002 0.0003 0.0004 0.0005 -5.00E-09 Time

Figure 5.16: Charge Transients vs. Time for GaN Schottky diode at different temperature

values

61

Charge Transients vs. Time for SiC Schottky diode

3.00E-09

2.50E-09

2.00E-09 -30

0 1.50E-09 30

1.00E-09 60 Charge 5.00E-10 90 120 0.00E+00 150

-5.00E-10 0.000000 0.000100 0.000200 0.000300 0.000400 0.000500 Time

Figure 5.17: Charge Transients vs. Time for SiC Schottky diode at different temperature

values

62

CHAPTER 6

CONCLUSION AND FUTURE WORK

6.1 Conclusion

This thesis described experiments made on the GaN Schottky barrier diodes and commercial SiC Schottky barrier diodes to investigate and compare the behavior of the both devices. The Electrical measurements such as I-V, C-V and Q-DLTS were applied to both devices to extract important diode parameters. All of those characterization techniques were done at room temperature and different temperatures ranging from 130 K up to 430 K to study the effect of the temperature on the performance of each device. The fabrication of the GaN

Schottky diodes involves three etching cycles and KOH treatment to etch down the LED commercial wafer until the unintentionally GaN layer has been exposed. Then, the Schottky contact and the Ohmic contact were deposited on the top of the GaN layers. The experiments performed included the measurements the forward current and the leakage current at different temperatures for selected one voltage value. The experiments also included the measurements of the capacitance values at different temperatures at a reverse bias of 1V. The ideality factor and the barrier height were extracted from the I-V measurements and plotted as cumulative plots.

Also, the carrier concentrations of both GaN and SiC Schottky diodes were extracted during the

C-V measurements. The conclusion arrived at as follows:

• The ideality factors of both GaN and SiC SBDs devices are decreasing with the increase

of the temperature.

• The barrier heights of both GaN and SiC SBDs are increasing with the increase of the

temperature.

63

• The carrier concentration of the SiC Schottky diodes is the same for the three selected

regions whereas the carrier concentration of the GaN Schottky diode increases as the

reverse bias increases.

• Two electron traps have been found by applying the TSCAP measurements to the GaN

Schottky diodes and located at 200 K and at 300 K with concentrations of 2.28x1018 cm-3

and 3.56x1017 cm-3, respectively. The reason of those traps is the etching process that was

done to the LED commercial wafer.

6.2 Future Work

Since the Charge-Deep Level Transient Spectroscopy technique was not successful to study the deep level traps in both GaN and SiC Schottky diodes, several investigations are needed in order to find more reasons and to get the desired charge transients so the DLTS signals can be extracted by using the Q-DLTS technique. Another idea is to send the devices to some companies that are specialized for applying the DLTS technique such as SULA technologies that are located in Oregon State in USA to measure the same devices that are available right now and to see if they can get the DLTS signal for both GaN and SiC Schottky barrier diodes. Also, the devices can be sent to different groups in different universities around the world that are experts in applying the DLTS technique to find and locate the traps in the GaN devices and what are the capture cross sections of those traps. Once the traps are located and their activation energies are known, the reasons of why those traps are appeared can be easily understood. Those devices can be restudied by using the same characterization techniques that are used in this thesis after the traps are removed by processing the device again. Once the traps can be detected and their locations are known by knowing their activation energies, they can be easily removed by processing the device again and then the performance of the devices can be enhanced after 64 removing the traps. Another idea is to purchase the typical capacitance DLTS system with the capabilities that help the researchers for studying the deep level traps in any semiconductor materials and use it instead of the charge based deep DLTS technique to measure both GaN and

SiC Schottky diodes.

65

REFERENCES

[1] J. Cho, J. Jung, J.H. Chae, H. Kim, H. Kim, J.W. Lee, S. Yoon, C. Sone, T. Jang, Y. Park, E.

Yoon, “Alternating-current Light Emitting Diodes with a Diode Bridge Circuitry”, Jpn J Appl

Phys, 46 (2007) L1194.

[2] H.-H. Yen, W.-Y. Yeh, H.-C. Kuo, “GaN alternating current light-emitting device”, physica status solidi (a), 204 (2007) 2077-2081.

[3] W.-Y. Yeh, H.-H. Yen, Y.-J. Chan, “The development of monolithic alternating current light- emitting diode”, (2011) 793910-793910.

[4] H.-H. Yen, H.-C. Kuo, W.-Y. Yeh, “Characteristics of Single-Chip GaN-Based Alternating

Current Light-Emitting Diode”, Jpn J Appl Phys, 47 (2008) 8808.

[5] Y. Hsi-Hsuan, K. Hao-Chung, W.-Y. Yeh, “Particular Failure Mechanism of GaN-Based

Alternating Current Light-Emitting Diode Induced by GaOx Oxidation”, Photonics Technology

Letters, IEEE, 22 (2010) 1168-1170.

[6] H.-H. Yen, W.-Y. Yeh, “Alternating current light-emitting device and fabrication method thereof”, U.S. Patent No. 7,855,388, (2010).

[7] J.-i. Chyi, G.-y. Lee, W.-s. Lin, “Method for fabricating integrated alternating-current light- emitting diode module”, U.S. Pat. No. 8,426,226, (2013).

[8] Http://www.digikey.com/en/articles/techzone/2011/aug/introduction-to-gan-technology

[9] S. Nakamura, M.R. Krames, “History of Gallium–Nitride-Based Light-Emitting Diodes for

Illumination”, IEEE. (2013) 0018-9219.

[10] Ralph Williams, "Modern GaAs Processing Methods," Aetech House (1990).

[11] Feng Shi (2011). GaN Nanowires Fabricated by Magnetron Sputtering Deposition,

Nanowires - Fundamental Research, Dr. Abbass Hashim (Ed.), ISBN: 978-953-307-327-9, 66

InTech, DOI: 10.5772/16355. Available from: http://www.intechopen.com/books/nanowires- fundamental-research/gan-nanowires-fabricated-by-magnetron-sputtering-deposition.

[12] R. Dylewicz, S. Patek, R. Paszkiewicz, “Applications of GaN-based materials in modern optoelectronics” Proc. of SPIE Vol. 5484, (2004).

[13] U.S. Department of Energy, Solid-State Lighting Research and Development Multi-Year

Program Plan, Office of Energy Efficiency and Renewable Energy, 1-84 (April 2013).

[14] Guha, S.; Bojarczuk, N.A.; Haight, R., "Hybrid organic-inorganic GaN LED based color downconversion for displays," IEEE (1997), 619244.

[15] R. M. Farrell et al, “ InGaN-based solar cells for ultrahigh efficiency multijunction solar cell applications” UC Solar Symposlum, (2011), University of California, Santa Barbra.

[16] Z. Jiang, “ Light Emitting Diodes and Photodetectors Based on III-Nitride and Colloidal

Quantum Dot Materials” Electronic Theses and Dissertation –The Pennsylvania State University,

(2014).

[17] B. Butun, T. Tut, E. Ulker, T. Yelboga, and E. Ozbay, “High-performance visible-blind

GaN-based p-i-n photodetectors” Appl. Phys. Lett. 92, 033507 (2008).

[18] D. Neamen, “Semiconductor Physics and Devices: Basic Principles – 3rd Edition” McGraw

Hill, (2003).

[19] R. T. Tung, “The physics and chemistry of the Schottky barrier height” Appl. Phys. Rev. 1,

011304 (2014).

[20] I. Poole, “Schottky Diode Characteristics & Specifications” Radio-electronics: Resources and analysis for electronics engineers, Available from http://www.radio- electronics.com/info/data/semicond/schottky_diode/characteristics-specifications-parameters.php

67

[21] Dieter K. Schroder, “Semiconductor Material and Device Characterization, Third Edition:

Chapter 3 – Contact Resistance and Schottky Barrier” 2006, John Wiley & Sons, Inc.

[22] J.M. Andrews and M.P. Lepselter, “Reverse Current-Voltage Characteristics of Metal-

Silicide Schottky Diodes,” Solid-State Electron. 13, 1011–1023, July 1970.

[23] A.K. Srivastava, B.M. Arora, and S. Guha, “Measurement of Richardson Constant of GaAs

Schottky Barriers,” Solid-State Electron. 24, 185–191, Feb. 1981.

[24] M. Missous and E.H. Rhoderick, “On the Richardson Constant for Aluminum/Gallium

Arsenide Schottky Diodes,” J. Appl. Phys. 69, 7142–7145, May 1991.

[25] Y.P. Song, R.L. Van Meirhaeghe, W.H. Lafl`ere and F. Cardon, “On the Difference in

Apparent Barrier Height as Obtained from Capacitance-Voltage and Current-Voltage-

Temperature Measurements on Al/p-InP Schottky Barriers,” Solid-State Electron. 29, 633–638,

June 1986.

[26] A. C. Schmitz, A. T. Ping, M. Asif Khan, Q. Chen, J. W. Yang and I Adesida, “Schottky barrier properties of various metals on n-type GaN” Semicond. Sci. Technol. 11 (1996) 1464–

1467.

[27] B. Van. Zeghbrock, “Principle of Semiconductor Devices: Chapter 4 –P – N junction”

(2011), from http://ecee.colorado.edu/~bart/book/pncap.htm

[28] Bougrov V., Levinshtein M.E., Rumyantsev S.L., Zubrilov A., in Properties of Advanced

SemiconductorMaterials GaN, AlN, InN, BN, SiC, SiGe . Eds. Levinshtein M.E., Rumyantsev

S.L., Shur M.S., John Wiley & Sons, Inc., New York, 2001, 1-30.

[29] Zhenyu Jiang, Mahmoud R M Atalla, Guanjun You, Li Wang, Xiaoyun Li, Jie Liu, Asim

M Elahi, Lai Wei, Jian Xu, “Monolithic integration of nitride light emitting diodes and photodetectors for bi-directional optical communication” Optical Letters (2014), Vol. 39, No. 19. 68

[30] J.D.Guo, M.S.Feng, R.J.Guo, F.M.Pan, and C.Y.Chang, Appl. Phys. Lett. 67, 2657(1995).

[31] L.Wang, M.I.Nathan, T H.Lim, M. Akhan, and Q.Chen, Appl. Phys. Lett. 68, 1267 (1996).

[32] S.N.Mohammad, Z.Fan, AE.Botchkarev, W.Kim, O.Aktas, ASalvador, and

H.Morkov, Electron. Lett. 32, 598 (1996).

[33] AT.Ping, AC.Schmitz, M.AKhan, andI.Adesida, Electron. Lett. 32, 68(1996).

[34] Z.Z.Bandic, P.M.Bridger, E.C.Piquette, and T.C.McGiH, Appl. Phys. Lett. 74, 1266 (1999).

[35] P.Hacke, T.Detchprohm, K.Hiramatsu, and N. Sawaki, Appl. Phys. Lett 63, 2676 (1993).

[36] AC.Schmitz, AT.Ping, M.AKhan, J.W.Yang, and I.Adesida [Semicond. Sci. technol. (UK)

Vol.11 (1996)p. 1464-7].

[37] AC.Schmitz, A.T.Ping, M.A.Khan,Q.Chen, J.W.Yang, I.Adesida [J. Electron. Mater. (USA) vol.27 (1998) p.255-60.].

[38] J.D.Guo, F.MPan, M.S.Feng, R.J.Guo,P.F.Chou, C.Y.Chang, J.Appl. Phys. (USA) 80,

1623-7 (1996).

[39] Q.Z.Liu, l.S.Yu, F.Deng and S.S.Lay, Appl. Phys. Lett. 84, 881 (1998).

[40] HS.Venugopalan and S.E. Mohney, Appl. Phys. Lett. 73, 1242 (1998).

[41] James H.Edgar, Samuel Strite, Isamu Akasaki, Hirushi Amano and Christian Wetzel,

"Properties, Processing and Applications of Gallium Nitride and related Semiconductors."

INSPEC, London, United Kingdom. (1999).

[42] Interactive Characterization Software (ICS). Metrics Technology [cited; Available from: http://www.metricstech.com/ics/ics.shtml]

[43] D. Zhu, J. Xu, A.N. Noemaun, J.K. Kim, E.F. Schubert, M.H. Crawford, et al, “ The origin of the high diode-ideality factors in GaInN/GaN multiple quantum well light-

69 emitting diodes ”, Appl Phys Lett (2009), p. 081113.

[44] A. Gokarna , N. R. Pavaskar , S. D. Sathaye , V. Ganesan , and S. V. Bhoraskar,

“Electroluminescence from heterojunctions of nanocrystalline CdS and ZnS with porous silicon”

Journal of Applied Physics 92 , 2118 (2002); doi: 10.1063/1.1483381.

[45] P. Choi, H. Kim, D. Baek, and B. Choi, “A Study on the Electrical Characteristic Analysis of c-Si Solar Cell Diodes” Journal of Semiconductor Technology and Science, vol.12, no.1,

March, 2012.

[46] S. Altindal, S. Karadeniz, N. Tugluogiu, A. Tataroglu: Solid-state Electronics Vol. 47

(2003), p. 1847-1854.

[47] Wang Y., Zhang Y., Zhang Y., Song Q. and Jia R., “Al/Ti/4H SiC Schottky barrier diodes with inhomogeneous barrier heights”, Chin. Phys. B Vol. 20, No. 8 (2011).

[48] P. S. Nayar, “Deep Level Studies of Undoped CdTe” SPIE 0524, Spectroscopic

Characterization Techniques for Semiconductor Technology II, 68 (June 28, 1985); doi:10.1117/12.946321.

[49] Emre Gur, C. Coskun and S. Tuzemen, “High energy electron irradiation effects on electrical properties of Au/n-ZnO Schottky diodes” J. Phys. D: Appl. Phys. 41 (2008)

105301 (5pp).

[50] Dieter K. Schroder, “Semiconductor Material and Device Characterization, 3rd

Edition” ISBN: 978-0-471-74908-0, John Wiley & Sons, February 2006.

[51] U. P. Singh, “Hydrogen Passivation Studies in Pd/n-GaAs devices” Bulletin of

Materials Science April 1998, Volume 21, Issue 2, pp 155-160.

70

[52] Z.H. Mahmood, A.P. Shah, A. Kadir, M.R. Gokhale, A. Bhattacharya, B.M. Arora, Charge deep level transient spectroscopy of electron traps in MOVPE grown n-GaN on sapphire, physica status solidi (b), 245 (2008) 2567-2571.

71