FABRICATION AND ELECTRICAL CHARACTERISATION OF

GRAPHITIC SCHOTTKY CONTACTS TO

SILICON, CARBIDE AND GALLIUM OXIDE

A thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy

Hung Viet Pham

Master of Electrical and Computer Engineering by RMIT University

School of Engineering College of Science, Engineering and Health RMIT University

August 2019 AUTHOR’S DECLARATION

I certify that except where due acknowledgement has been made, the work is that of the author alone; the work has not been submitted previously, in whole or in part, to qualify for any other academic award; the content of the thesis is the result of work, which has been carried out since the official commencement date of the approved research program; any editorial work, paid or unpaid, carried out by a third party is acknowledged.

Hung Viet Pham

02/08/2019

ii

ACKNOWLEDGMENTS

This PhD thesis cannot be accomplished without the support, advice and encouragement from many sincere and generous people.

I want to show special appreciation to my supervisor, Associate Professor Anthony

Holland, for his constant academic guidance throughout my PhD candidature at RMIT

University. I will always be impressed and grateful for your dedication toward student despite of distance and time.

I would also like to send the deepest gratitude to my co-senior supervisor, Doctor

James Partridge. Your professional and disciplined hard-working attitude as well as your broad Physics insight has become the model for me to learn and follow.

My appreciation extends to many staffs and friends at RMIT Vietnam who support in laboratory and administration, especially Head of School of Science and Technology

Associate Professor Alex Stocevski and Professor Eric Dimla who provided strong academic environment to support PhD students.

Within a short period of time in RMIT Melbourne, I had the opportunities to work intensively in RMMF (RMIT Microscopy and Microanalysis Facility) to fulfill my PhD.

This visit cannot happen without sponsorship from Doctor James Partridge and Professor

Dougal McCulloch, and relentless effort of Associate Professor Anthony Holland. I also want to thank Hiep Tran, Tuan Bui and Phuong Le for their enthusiastic support during my time in Melbourne.

Finally, I send special thanks to all my friends and colleagues who walk along with me through the long journey.

iii

PUBLICATIONS

Journal publication:

[1] H. V. Pham et al., "Temperature dependent electrical characteristics of rectifying

graphitic contacts to p-type silicon," Science and Technology, vol.

34, no. 1, p. 015003, 2018.

[2] H. Pham, H. N. Tran, A. S. Holland, and J. G. Partridge, "Temperature-Dependent

Electrical Characteristics and Extraction of Richardson Constant from Graphitic-

C/n-Type 6H-SiC Schottky ," Journal of Electronic Materials, pp. 1-6, 2019.

This article has been chosen by the Editors of the Journal of Electronic Materials

to appear on Springer's website as a "Free Access" article.

IEEE Conference Proceeding

[3] Hung V. Pham, N. Le Huy, A. Stojcevski, and A. S. Holland, "Impact of impurities

in 4H, 6H and 3C-SiC substrate on reverse recovery time of pn junction," in

Information Science and Technology (ICIST), 2017 Seventh International

Conference on, 2017, pp. 299-303: IEEE.

[4] Hung V. Pham, A. S. Holland, H. L. Nguyen, J. G. Partridge, and H. N. Tran,

"Modified electrical characteristics of filtered cathodic vacuum arc amorphous

film on n-Si (100) by heat treatment," in Micro and Nanoelectronics (RSM),

2017 IEEE Regional Symposium on, 2017, pp. 38-41: IEEE.

[5] Hung V. Pham, S. N. Demidenko, and G. M. Merola, "Eliminating Re-Burn-In in

semiconductor manufacturing through statistical analysis of production test data,"

iv

in Instrumentation and Measurement Technology Conference (I2MTC), 2017 IEEE

International, 2017, pp. 1-6: IEEE.

[6] H. V. Pham, S. Luong, A. S. Holland, and H. L. Nguyen, "Impact of temperature

on electrical performance of Ni film on n-type 4H-SÌC contacts in terms of

micropipes density," in Recent Advances in Signal Processing,

Telecommunications & Computing (SigTelCom), 2018 2nd International

Conference on, 2018, pp. 132-135: IEEE.

v

TABLE OF CONTENTS

AUTHOR’S DECLARATION ...... ii

ACKNOWLEDGMENTS ...... iii

PUBLICATIONS ...... iv

TABLE OF CONTENTS ...... vi

LIST OF FIGURES ...... x

LIST OF TABLES ...... xv

LIST OF SYMBOLS ...... xvi

LIST OF SYNONYMS...... xviii

ABSTRACT ...... xix

CHAPTER 1 INTRODUCTION ...... 1

1.1 Motivation and rationale ...... 1

1.2 Silicon ...... 2

1.3 Silicon Carbide ...... 4

1.4 Gallium Oxide ...... 8

1.5 Challenges for high power devices ...... 10

1.6 Schottky diodes ...... 11

1.6.1 Brief overview of -semiconductor junction formation and electrical properties ...... 14

1.6.2 Application ...... 16

1.7 Carbon materials ...... 16

vi

1.8 Experiment activities ...... 19

1.9 Thesis organization ...... 20

1.10 References ...... 21

CHAPTER 2 BACKGROUND AND RECENT ADVANCES IN CARBON-BASED SCHOTTKY DEVICES ...... 31

2.1 Introduction ...... 31

2.2 Schottky ...... 31

2.2.1 Interface states ...... 34

2.2.2 Image force lowering ...... 36

2.2.3 Current-voltage characteristics ...... 38

2.2.4 Current transportation ...... 39

2.2.5 Schottky barrier inhomogeneity ...... 45

2.3 Carbon-semiconductor Schottky diodes ...... 48

2.4 Summary ...... 57

2.5 References ...... 58

CHAPTER 3 FABRICATION & CHARACTERISATION METHODS ...... 65

3.1 fabrication ...... 65

3.1.1 Substrate dicing ...... 65

3.1.2 Wafer cleaning ...... 66

3.1.3 Spin coating ...... 67

3.1.4 Mask alignment and UV exposure ...... 68

3.1.5 Energetic carbon deposition using a filtered cathodic vacuum arc (FCVA) 70

3.1.6 Ion beam sputtering of metallic device contact layers ...... 73

3.2 Electrical characterization ...... 74

3.2.1 Current-voltage characterization ...... 74

3.2.2 Temperature dependent electrical measurements ...... 75 vii

3.2.3 Capacitance-voltage characterization ...... 75

3.3 Material characterization ...... 76

3.3.1 Introduction ...... 76

3.3.2 Scanning Electron Microscopy (SEM) ...... 78

3.3.3 X-ray Photoelectron Spectroscopy (XPS) ...... 80

3.3.4 Raman spectroscopy ...... 82

3.3.5 Film thickness measurement ...... 84

3.3.6 Hall measurement ...... 85

3.4 Summary ...... 86

3.5 References ...... 87

CHAPTER 4 GRAPHITE ON SILICON SCHOTTKY DIODE ...... 93

4.1 I-V characteristics ...... 93

4.2 C-V characteristics ...... 98

4.3 Richardson constant ...... 99

4.4 Summary ...... 103

4.5 References ...... 103

CHAPTER 5 GRAPHITE ON SILICON CARBIDE SCHOTTKY DIODE ...... 106

5.1 I-V characteristics ...... 106

5.2 C-V characteristics ...... 113

5.3 Richardson constant ...... 115

5.4 Summary ...... 117

5.5 References ...... 118

CHAPTER 6 GRAPHITE ON GALLIUM OXIDE SCHOTTKY DIODE ...... 121

6.1 I-V characteristics ...... 121

6.2 Richardson constant ...... 126

viii

6.3 Summary ...... 128

6.4 References ...... 129

CHAPTER 7 CONCLUSIONS AND FUTURE WORK ...... 131

7.1 Conclusions ...... 131

7.2 Future Work ...... 132

7.3 References ...... 133

APPENDIX A WAFER SPECS ...... 134

A.1 Wafer Specs ...... 134

ix

LIST OF FIGURES

Figure 1-1: Structure of Silicon (reproduced from [24])...... 3

Figure 1-2: Illustration of SiC polytypism with Si-C bilayer stacking along the c [0001] axis for the three main SiC polytypes 3C, 4H and 6H [43]. h and k stand for hexagonal and cubic style of stacking, respectively...... 6

Figure 1-3: SiC focused journal publications per year since 1970 (Source: ScopusTM search engine with titles “SiC” or “Silicon Carbide”, April 02, 2019) ...... 7

Figure 1-4: A graphical comparison of the key material properties for β-Ga2O3 – the , electron mobility, breakdown field, melting point, and thermal conductivity, as compared with Si, SiC and GaN. The numbered values refer to Ga2O3 while here these areas are all normalized to silicon (blue pentagon) [reproduced from 53]...... 9

Figure 1-5: Structure of a Schottky diode with Schottky (rectifying) and Ohmic (non- rectifying) back metal contact [77-79]...... 13

Figure 1-6: Thermal equilibrium energy for metal contact with n-type and p-type semiconductor (modified from [23])...... 15

Figure 1-7: Typical current-voltage characteristic of n-type Schottky diode. The inset is a simplified schematic of a n-type 6H-SiC Schottky diode...... 16

Figure 1-8: Carbon journal publications per year since 1970 (Source: ScopusTM search engine with titles “Carbon”, “Graphite”, “Graphene”, “Carbon Nanotube” or “Diamond”, April 02, 2019) ...... 17

Figure 1-9: Hybridization of the carbon atom [118, 144, 145]...... 18

Figure 1-10: Some allotropes of carbon: a) diamond; b) graphite; c) graphene; d) amorphous carbon; e) carbon nanotube [146-150]...... 19

Figure 2-1: (a) Energy band diagram of an isolated metal adjacent to an isolated n- type semiconductor. (b) Energy band diagram of the same metal and semiconductor in contact (reproduced from [3])...... 33

Figure 2-2: Barrier height versus electronegativity of deposited on Si, GaSe, and SiO2 [41]...... 36

x

Figure 2-3: Energy-band diagram incorporating the Schottky effect for a n-type metal semiconductor contact under different biasing conditions (reproduced from [41]). The intrinsic barrier height is ΦBn0. The barrier height at thermal equilibrium is ΦBn. The barrier is lowered under forward and reverse biases by ΔΦF and ΔΦR respectively...... 38

Figure 2-4: Energy band diagram showing three current transport mechanisms in n- type Schottky contact: (TE), thermionic field emission (TFE) and field emission (FE) [3]...... 40

Figure 2-5: Space charge effect. a) Band diagram for electron injection and b) J-V characteristic due to the space charge effect [3]...... 43

Figure 2-6: Energy band diagram showing transport mechanisms of (a) direct tunneling, (b) Fowler-Nordheim tunneling, (c) thermionic emission and (d) Poole- Frenkle emission [3]...... 45

Figure 2-7: Schottky barrier heights (SBHs) and ideality factors derived from the fitting as a function of temperature for CoSi2 contacts on (a) Si (100) and (b) Si (111) substrates. The modified SBH data and their linear fits are also shown [64]...... 46

Figure 2-8: Ideality factors and SBHs as a function of the ...... 47

Figure 2-9: Effective barrier heights and ideality factors as a function of temperature for a graphene–MoS2 contact [26]...... 47

Figure 2-10: Graphene on Silicon Schottky diode by Stelzer and Kreupl [65, 66]...... 48

Figure 2-11: Current voltage characteristic of graphene on a silicon Schottky diode after various series of current pulses by Stelzer and Kreupl [65, 66]. After up to 500 million pulses the electrical behavior is almost unaffected. Only a small shift of the SBH of 0.01 eV is observable, leading to a lower reverse current. After 545 million pulses, the reverse leakage current increases...... 49

Figure 2-12: SEM image of the top view of C-Si Schottky diodes with Ti/Cu/Au top metallization after the diodes failed. (a) shows a device after 548 million pulses (100ns) with no obvious visible damage in the top metallization. The device in (b) demonstrates the damage of a 300 ns pulse after 1.5 million pulses where the metal started to melt. (c) illustrates a sample where the molten copper was spit on the periphery and a circular crater was formed after a series of 500 ns pulses. (d) shows J- V characteristics of graphene on Si diodes after they reached the threshold for a failed diode [65, 66]...... 50

Figure 2-13: (a–d) Schematic of the process flow for metal-semiconductor-metal photodetectors with CNT film electrodes along the dashed line AB shown in (e): (a) SiN isolation layer deposited on a GaAs substrate, (b) CNT film prepared by vacuum filtration deposited on the substrate after opening the active windows in the SiN layer, (c) CNT film patterned into interdigitated electrode fingers by photolithography and

xi inductively coupled plasma etching, and (d) Cr/Pd metal contacts patterned on the nanotube film contact pads using photolithography, e-beam evaporation, and subsequent lift-off. (e) Optical microscope image of the finished metal- semiconductor-metal photodetector, showing the various device dimensions. (f) Atomic force microscope (AFM) image showing the area between two CNT film electrode fingers of the metal-semiconductor-metal device of part (e) [77]...... 52

Figure 2-14: (a) Dark current (log scale) vs applied voltage measured at six different temperatures between 230 and 340 K for a CNT film–GaAs metal-semiconductor- metal device with W = S = 15 μm and FL = FW = 300 μm. The inset shows the dark I-V characteristics of this device at room temperature 294K in linear scale. (b) The Richardson plot of log I/T2 vs 1/T at V= 3 V voltage bias in the temperature range of 280–340 K for three CNT film–GaAs metal-semiconductor-metal devices of identical active area (FL = FW = 300 μm) but different finger widths W and spacings S, as labeled in the figure. The inset shows log current vs 1/T for the metal-semiconductor- metal device with W=S= 20 um in a wider temperature window (from 150 to 340K) at V= 3 V bias [77]...... 53

Figure 2-15: J-V characteristic of HOPG contact at room temperature on (a) n-type Si/graphite (red squares), (b) n-type GaAs/graphite (blue circles), and (c) n type 4H- SiC/graphite junctions (black triangles). Insets: J-V plots on semi-logarithmic axes [80]...... 56

Figure 2-16: J-V characteristic of the graphite/ZnO(O-face) Schottky diode over the temperature range of 300–420 K [96]...... 57

Figure 3-1: Dicing machine...... 66

Figure 3-2: Laurell WS-650MZ spin coater...... 67

Figure 3-3: Cr on glass masks...... 68

Figure 3-4: Karl Suss MJB-3 mask aligner...... 69

Figure 3-5: Circular test device structure...... 69

Figure 3-6: Schematic of a filtered cathodic vacuum arc deposition system, reproduced from [58]...... 71

Figure 3-7: FCVA deposition system installed in RMIT University, Melbourne...... 72

Figure 3-8: Gatan 682 precision coating system...... 73

Figure 3-9: Through film I-V measurement circuit...... 74

Figure 3-10: Linkam temperature variating system...... 75

Figure 3-11: Boonton 7200 C-V measurement system...... 76

xii

Figure 3-12: Common microanalysis techniques with different resolution. Figure reproduced from [59]...... 77

Figure 3-13: SEM system schematic showing the electron gun emitting electrons which are then accelerated as a beam towards the sample/specimen. Resulting electrons from the sample are collected to form an image on a computer screen. Figure reproduced from [59]...... 78

Figure 3-14: The Philips XL30 SEM tool at the RMIT University RMMF (RMIT Microscopy and Microanalysis Facility)...... 79

Figure 3-15: Demonstration of a XPS system’s operation [64]...... 80

Figure 3-16: Typical XPS data from C films energetically deposited with ion energies of (a) -500 eV and (b) -1.0 keV substrate biases applied...... 82

Figure 3-17: Photograph of a Horiba Scientific LabRAM HR Evolution Raman system, such as used at RMIT University...... 83

Figure 3-18: Raman spectra taken from carbon films deposited at (a) 25 ºC/0.52 keV and (b) 100 ºC/1.0 keV. The peak from the Si substrate and the D and G peaks are labeled...... 84

Figure 3-19: KLA Tencor P-16+ surface profiler...... 85

Figure 3-20: Hall measurement HMS-3000 system...... 86

Figure 4-1: Room temperature I-V characteristics of an energetically deposited C/p- Si junction. The inset shows the structure of the device and the measurement circuit. .... 94

Figure 4-2: (a) Forward bias I-V characteristics of a C/p-Si Schottky diode (C deposited at 100 ºC and 1.0 keV) with linear fits showing transport dominated by thermionic emission in the temperature range 98 - 498 K. (b) Richardson plot used to extract the effective Richardson constant...... 96

Figure 4-3: Theoretical fits of direct tunneling model for the I-V-T characteristics of a C/p-Si diode in the region of the reverse-bias voltages. Reverse currents at temperature 98 K are below the noise threshold 2 pA...... 96

Figure 4-4: Temperature dependent barrier height and ideality factor of the energetically deposited (100 ºC/1.0 keV) C/p-Si Schottky diode...... 97

Figure 4-5: Plots of inverse capacitance square versus reverse bias voltage measured from Ag/C/p-Si contacts with C deposited at the energy/temperature indicated...... 99

Figure 4-6: Barrier height versus 1/2kT plot for the energetically deposited 100 ºC/1.0 keV C/p-Si Schottky diode...... 100

xiii

Figure 4-7: Modified Richardson plot for the 100 ºC/1.0 keV C/p-Si Schottky diode. 101

Figure 5-1: I-V characteristics of energetically deposited C/6H-SiC diodes with direct bias (grey) and remote bias (black) applied during deposition and (inset) schematic of the device measurement circuit...... 108

Figure 5-2: a) Forward bias portions of the I-V-T characteristics of the C/6H-SiC Schottky diode with linear fits showing transport by TE and b) temperature- dependent differential device resistance versus voltage characteristics. The energy difference between the donor level and conduction band in the n-type 6H-SiC substrate is determined from the inset Arrhenius plot...... 109

Figure 5-3: Temperature dependent a) I-V characteristics, b) rectification ratios and c) ideality factors/apparent barrier heights of a C/6H-SiC diode...... 111

Figure 5-4: The temperature dependent current voltage characteristics of C/6H-SiC diodes with reverse-bias voltages of a) 3V > |V| > 0.6V and b) 0.6V > |V| > 0V. The linearities in a) and b) are consistent with transport dominated by Schottky emission and direct tunneling, respectively...... 113

Figure 5-5: The capacitance voltage characteristic of C/6H-SiC junction under reverse bias at room temperature measured at 1 MHz...... 114

Figure 5-6: a) Richardson plot constructed from the I-V-T data of a C/6H-SiC Schottky diode and b) linear correlation between ln(A*eff) and ΦBeff obtained from several devices and ΦBeff vs. n relationship used to determine the barrier height characteristic of a laterally homogeneous energetically deposited C/6H-SiC Schottky diode...... 116

Figure 6-1: I-V characteristic of C/Ga2O3 Schottky diode with elevated temperature from 143K to 443K...... 122

Figure 6-2: Temperature-dependent differential device resistance versus voltage characteristics of C/Ga2O3 Schottky diode...... 123

Figure 6-3: a) Rectification ratios and b) ideality factors/apparent barrier heights of a C/Ga2O3 diode...... 124

Figure 6-4: (a) Forward bias I-V characteristics of a C/Ga2O3 Schottky diode with linear fits showing transport dominated by thermionic emission in the temperature range 283 - 443 K. (b) Richardson plot used to extract the effective Richardson constant. Lower temperature data are unreliable for linear fit due to high series resistances...... 125

Figure 6-5: Barrier height versus 1/2kT plot for the C/Ga2O3 Schottky diode...... 126

Figure 6-6: Modified Richardson plot for the C/Ga2O3 Schottky diode...... 127

xiv

LIST OF TABLES

Table 1-1: Electrical properties of Si and three polytypes of SiC [4, 23, 36-39] ...... 5

Table 1-2: Maximum electronic operating temperature for high-temperature applications of Si and SiC [4, 23, 36-39, 47, 48]...... 7

Table A-1: Properties of Si wafers...... 134

Table A-2: Properties of SiC wafers...... 134

Table A-3: Properties of β-Ga2O3 wafers...... 134

xv

LIST OF SYMBOLS

Symbol Description

A Diode contact area

A* Richardson constant

A*eff Effective Richardson constant

A*o Homogeneous Richardson constant

C Capacitance

Cm Measured capacitance

Energy difference between the active EA donor level and conduction band

EC Energy at conduction band

EF

EV Energy at valence band

I Current

Is Saturation current k Boltzmann’s constant n Ideality factor

Concentration of uncompensated donor ND – NA (Donor – Acceptor concentration)

Ns Surface density of electron q Electron charge

xvi

Electron charge accumulated with a unit Qss area of the interface

Rdiff Differential device resistance

RR Rectification ratio

Rs Series resistance

T Temperature

V Voltage

Vbb Band bending parameter

Vbi Built in potential

Vc Cut off voltage

β Parameter (q/KT)

1/3 β1 Parameter (Vbb/η)

ϵs constant of semiconductor

η Overall patch strength

χs of semiconductor

ω Angular frequency

фB Barrier height

фBeff Effective barrier height

фBo Homogeneous barrier height

фM of carbon

xvii

LIST OF SYNONYMS

Symbol Description

AFM Atomic Force Microscopy

BH Barrier Height

CNT Carbon NanoTubes

C-V Capacitance-Voltage

DI water De-Ionsed water

IPA Isopropanol

I-V Current-Voltage

I-V-T Current-Voltage-Temperature

MIGS Metal Induced Gap States

RMMF RMIT Microscopy and Microanalysis Facility

SBH Schottky Barrier Height

Si Silicon

SiC Silicon Carbide

xviii

ABSTRACT

There is an increasing demand for power electronic devices that exhibit high current rectification and high environmental/thermal stability. This demand has driven research into novel materials that provide stable performance in extreme environments. Abundant and low-cost carbon thin film materials are potential candidates because they exhibit a wide range of electrical characteristics that depend systematically on their microstructure/bonding. More specifically, their electrical characteristics are sensitive to the ratio of diamond-like to graphitic bonding within the film. The electrical resistivity of the carbon film can be varied by many orders of magnitude within a readily accessible range of microstructures. Energetic deposition methods are one way to control the bonding ratios in carbon thin films and energetic deposition has been exploited for much of the work in this thesis.

The electronic devices that were fabricated and tested during this project were

Schottky diodes. These devices, normally consisting of metal-semiconductor junctions, were instead formed between energetically deposited graphitic carbon (with high graphite- like bonding fraction) and the semiconductor materials silicon, silicon carbide and gallium oxide. These materials were selected due to their proven or anticipated application in power devices. Once fabricated, the graphitic-C/semiconductor junctions were characterized micro-structurally and electrically. Subsequent analysis of the measurements revealed reasons for the excellent device performance and/or guidance for further improvements.

xix

CHAPTER 1

INTRODUCTION

1.1 Motivation and rationale

All semiconductor devices require electrical contacts (except some photonic devices). These contacts may fulfill simple input-output duties or more elaborate functions.

They are often formed between dissimilar materials and may provide either linear or non- linear current-voltage characteristics depending on the device application. Contacts are formed on semiconductor surfaces which contain electronic states that differ significantly from those in the semiconductor bulk. This presents challenges in their design, fabrication and optimization. Operational efficiency and reliability depends on a thorough understanding of the structure, chemistry and electrical transport in electrical contacts.

Hence, extensive theoretical and experimental research activity has been devoted to electrical contact design and fabrication.

This thesis describes research focused on developing a new method for forming carbon-based electrical contacts to . Carbon offers advantages including thermal and chemical stability as well as being an abundant low-cost material. Its ability to withstand high operating temperatures, voltages and current-densities has been noted by the power electronics community [1-3]. In the project work described in this thesis, carbon- based contacts were formed on semiconducting Silicon (Si), Silicon Carbide (6H-SiC), and

1

Gallium Oxide (β-Ga2O3). These materials were selected as they are important for recent and future advances in the power electronics industry.

Silicon is the dominant semiconductor material in power electronics as in micro- electronics but in recent years, Silicon Carbide has emerged as a power electronic material due to its larger band gap, lower dielectric constant, higher thermal conductivity, higher critical electric field and higher carrier saturation velocity compared to Si [4]. With improved growth methods and defect control, 6H-SiC and 4H-SiC have become a common choice for high power applications and have shown potential to develop further into a

mainstream semiconductors [5]. More recently, Ga2O3 has attracted attention from researchers working in the power electronics field due to its exceptional breakdown characteristics and comparative ease of production [6-9]. The remaining sections in this chapter summarize the materials and devices central to the thesis and then provide an overview of the project objectives and thesis structure.

1.2 Silicon

Si has long been the foundation material for microelectronics. Si and Ge were both exploited as detector materials in radar technology beginning around 1930 [10]. Later, with the emergence of the transistor [11-13] and improvements in Si production techniques [14-

16], Si overtook Ge and became the dominant semiconducting material that it is today. The advantages of Si over Ge (wider bandgap and more abundant resource) in power devices were revealed back in 1953 [17, 18]. As previously stated, devices cannot be made from Si alone. As a consequence, alongside the development of Si as a semiconducting material, there has been ongoing development in rectifying and ohmic contacts to this material. With

2 a stable thermal oxide and a well characterized surface, Si can be interfaced with metals, insulators and other semiconductors to form functional electronic junctions.

Si mostly exists in various forms of silicon dioxide or silicates. Silicates are one of the most abundant materials in the Earth’s crust hence making Si a low cost material for electronic applications [19-21]. Si is a typical semiconductor with the Fermi level within the energy band gap and possesses four valence electrons. Si with P introduces one extra electron per dopant, and these may then be excited into the conduction band either thermally or photoexcitedly, creating an n-type semiconductor. Similarly, doping Si with

B or Al creates a hole that traps electrons, creating a p-type semiconductor [22, 23].

The physical properties of Si also contribute to its popularity and usage as a semiconductor material. Si possesses a moderate energy band gap of 1.12eV. This makes

Si a stable element when compared to Ge and reduces leakage current. The crystalline structure of Si, as shown in Figure 1-1, consists of a face centric cubic lattice structure with 34% packing density. This lattice structure allows easy substitution of impurity’ atoms onto silicon lattice sites, and the doping concentration in silicon can reach 1021 cm-2.

Figure 1-1: Structure of Silicon (reproduced from [24]).

3

In terms of Schottky diode application, Si was used widely as a base semiconductor with various contact metals such as Al, Ni, or Ti [25-30]. Due to the fact that Si forms compounds with most elements in the Periodic Table, metal silicides that form an interface with Si that is stable and free from a native oxide have also been exploited [24].

One of the main focuses of this project was forming Si Schottky diodes with carbon-based contacts. Schottky diodes are briefly reviewed in section 1.6. More details on Schottky diode formation and characterization are provided in Chapter 2 (section 2.2).

1.3 Silicon Carbide

Improving the efficiency of power electronic devices is vital to reduce losses incurred during switching. Whilst Si is by far the most widely used semiconductor material for power devices, its material limits are being reached. Hence, engineers and researchers have searched for alternatives to Si for power devices [31]. Wide bandgap (WBG) semiconductors such as Silicon Carbide (SiC) and Gallium Nitride (GaN) offer significant benefits over Si including more stable operation at higher temperatures, high dielectric strength and high saturation drift velocity [4]. SiC has arguably become the material of choice for high performance power semiconductor devices. The wider bandgap, higher thermal conductivity and larger critical electric field allow SiC devices to operate at higher temperature with higher current density and higher blocking voltage [32-35]. The outstanding electrical characteristics of SiC, which are listed in Table 1-1, provide numerous and novel combinations of attributes for a semiconductor material for both optoelectronic and for high-power, frequency, temperature, speed, and radiation hard

4 microelectronic devices. The band gap of SiC (up to 3.26eV for 4H-SiC) is significantly wider than Si (1.12eV). The higher energy gap gives devices the ability to operate at higher temperatures, and for some applications, allows devices to at larger voltages. Wide bandgap semiconductors such as SiC are associated with a high breakdown voltage.

Table 1-1: Electrical properties of Si and three polytypes of SiC [4, 23, 36-39]

PROPERTIES Si 3C-SiC 4H-SiC 6H-SiC Bandgap 1.12 2.40 3.26 3.02 (eV) Electric Field Breakdown 0.25 2.12 2.2 2.5 Strength (MVcm-1) Thermal Conductivity 1.5 3.2 3.7 4.9 At 300K (Wcm-1K-1) Intrinsic Carrier Concentration 10 -1 -9 -6 At 300K (cm-3) 1.0 × 10 1.5 × 10 5 × 10 1.6 × 10 Saturation Velocity 7 7 7 7 (cms-1) 1.0 × 10 2.0 × 10 2.0 × 10 2.0 × 10 Electron Mobility 1400 800 1000 400 (cm2V-1s-1) Hole Mobility 471 40 115 101 (cm2V-1s-1) Dielectric Constant 11.7 9.72 9.66 9.66

Comprising equal parts (covalently bonded) silicon and carbon with a highly ordered bonding configuration, single crystal SiC is the third hardest substance on earth.

More than 170 different polytypes exist and depending on the crystal structure, SiC has an energy gap of 2.0 to 3.3 eV [40, 41]. Among the polytypes, 3C, 4H and 6H are used most widely for device production since large wafers can be made [36]. The polytypism of these three SiC polytypes are shown in Figure 1-2 [36-38, 42].

5

Figure 1-2: Illustration of SiC polytypism with Si-C bilayer stacking along the c [0001] axis for the three main SiC polytypes 3C, 4H and 6H [43]. h and k stand for hexagonal and cubic style of stacking, respectively.

Many companies now manufacture SiC substrates (e.g. Cree) with very low

‘micropipe’ defect densities, a problem that hindered early development [44, 45]. The thermal conductivity of SiC is triple that of Si which assists in heat dissipation and enables the device package size to be reduced [38, 46]. A comparison between Si and the three main SiC polytypes (3C-SiC, 4H-SiC and 6H-SiC) in terms of maximum operating temperature is shown in Table 1-2. SiC has demonstrated a strong capability to operate under much harsher thermal environments (up to 750oC) than Si (under 150oC). However, in a device, the contacts/encapsulation may impose lower thermal limits so that it is important to develop high performance contact materials.

6

Table 1-2: Maximum electronic operating temperature for high-temperature

applications of Si and SiC [4, 23, 36-39, 47, 48].

PROPERTIES BANDGAP MAXIMUM OPERATING PROCESS (EV) TEMPERATURE (OC) MATURITY SI 1.12 150 Very high 3C-SIC 2.40 600 Low 4H-SIC 3.26 750 Medium 6H-SIC 3.02 700 Medium

There are many methods to manufacture SiC wafers, and each type of SiC requires distinct manufacturing processes. The simplest SiC manufacturing method is to synthesize the mix of silica (SiO2) with carbon in an electric resistance furnace at high temperatures from 1400oC to 2500oC [49-52]. With aforementioned properties, SiC has become the focus of much research as shown in Figure 1-3.

Figure 1-3: SiC focused journal publications per year since 1970 (Source: ScopusTM search engine with titles “SiC” or “Silicon Carbide”, April 02, 2019)

7

1.4 Gallium Oxide

Power electronics is currently dominated by Si-based devices due to the low-cost, versatility and established processing methods of Si. However, as a power device material

Si is not particularly efficient. Alternatives such as SiC and GaN exist but their commercial uptake has been limited by their relatively high cost. Wide band-gap semiconducting Ga2O3 has recently emerged as a material with the potential to replace Si, as well as other semiconductors such as SiC and GaN.

The bandgap of Ga2O3 (4.9 eV) significantly exceeds that of both SiC (3.3 eV) and

GaN (3.4 eV), meaning it can withstand higher applied electric fields. This is a significant advantage because thinner layers can be employed within devices, and these can provide lower resistance and hence higher efficiency. Single crystal Ga2O3 is known to form five polymorphs, commonly labelled with α, β, γ, δ, ε. The β-phase is the most stable while the other polymorphs are meta-stable. With regard to commercial applications, bulk β-Ga2O3 substrates can be produced using low-cost melt methods so cost barriers are unlikely to prevent uptake.

In Figure 1-4, five key properties of β-Ga2O3, Si, SiC and GaN are compared. The numerical values listed are for β-Ga2O3 while each concentric pentagon represents a doubling of the relevant value in the outward direction. As shown, the bandgap and breakdown field are highest in β-Ga2O3. However, its carrier mobility and thermal conductivity require improvement and/or strategic device design.

8

Figure 1-4: A graphical comparison of the key material properties for β-Ga2O3 – the band gap, electron mobility, breakdown field, melting point, and thermal conductivity, as compared with Si, SiC and GaN. The numbered values refer to Ga2O3 while here these areas are all normalized to silicon (blue pentagon) [53].

For high quality devices to be realised using β-Ga2O3, suitable electrical contacts must be designed and implemented. Ohmic contacts with a low are a must, as are Schottky contacts with a controllable barrier height, low-leakage and high endurance. Regarding the latter, Farzana et al. [54] fabricated Pd, Ni, Pt and Au contacts to β-Ga2O3. They demonstrated high current rectification and determined the respective barrier heights from electrical measurements. In devices in which transport occurred by thermionic emission, a peak barrier height of 1.5 – 1.6 eV (depending on method employed) was reported for Pt. Recently, oxidised and non-oxidised noble metal Schottky contacts to β-Ga2O3 were compared using capacitance-voltage and temperature dependent current-voltage measurements [55]. In general, the reactively deposited (partially oxidised)

9 contacts exhibited higher barriers (up to 2.5 eV) and ideality factors that more closely approached unity (indicating higher lateral homogeneity). These devices exhibited record rectification ratios and in more recent work by the same group, IrOx Schottky contacts to

β-Ga2O3 were demonstrated to provide >10 orders of rectification at ± 3.0 V whilst operating above 300 ºC [56].

Despite these impressive performance figures, there remains a need for alternative

Schottky contact materials for β-Ga2O3. The cost and scarcity of Pd, Pt, Rh and Ir would likely limit commercial applications of the aforementioned devices. Hence, abundant, low- cost contact materials that enable high-performance β-Ga2O3 Schottky diodes with high power/temperature endurance are required.

1.5 Challenges for high power devices

Reviewing the literature shows that thermal stability of metal-semiconductor contacts has increasingly become a demand for emerging applications in extreme environments (such as nuclear facilities, space travel, or high temperature related industries). Si as a device layer is unstable under operating temperature > 250oC [48]. SiC possesses exceptional chemical and physical properties such as high thermal conductivity, a wide band gap, high breakdown field, high saturation velocity, and chemical stability

[57]. Therefore, metal-SiC Schottky contacts should be more suitable than metal-Silicon

Schottky contacts for electrical devices operating in harsh environments such as high voltage , UV radiation detectors, signal mixers, and high temperature gas sensors

[57, 58]. β-Ga2O3 is a relative newcomer and has an even wider band gap, stronger critical

10 field strength and significant cost advantages. However, thermal management for β-Ga2O3 appears to be a great concern due to its poor thermal conductivity [9, 59].

Carbon is a material with great potential for high-power contact applications.

Schottky diodes using carbon as a contact material to the wide band gap and high breakdown field semiconductors such as SiC and β-Ga2O3 can promise thermally stable and high-performance devices. Therefore, in this project, carbon on Si, 6H-SiC and β-

Ga2O3 Schottky diodes were fabricated and studied.

1.6 Schottky diodes

It is likely that the first Schottky diode ever made was a point-contact diode made by Braun in the 1870s [60]. He formed metal semiconductor junctions using lead sulphide and iron sulphide as semiconductors and metallic ‘whiskers’ for the point contact. When microwave radar was developed during World War Two, the metal-semiconductor point contact diode became important once again (having seen little use since the vacuum tube was invented two decades prior). These devices were mostly used as frequency converters and small signal microwave detector diodes and when compared with vacuum tube devices, offered lower noise operation and hence, greater suitability as square-law detectors of weak microwave pulses [61-63]. In 1948, Bardeen and Brattain reported the carrier injection phenomenon in point contact germanium diodes and they realised the point contact Ge transistor shortly afterwards [64]. In the 1950s, fabrication of Si p-n junctions became a mature technology and metal-semiconductor contacts were then largely designed as ohmic contacts [3]. Rectifiers based on pn junctions were more reliable and easier to understand than metal-semiconductor point contact rectifiers. The mixture of injection phenomena

11 associated with the metal-semiconductor barrier, unintentional p-n junctions and damage- induced generation-recombination all contributed to undesirable variability in point contact diodes. Several factors contributed to a revival of interest in Schottky diodes. The development of the planar Schottky diode process (intended primarily for field effect transistors) led to a better understanding of interfaces and associated cleaning techniques

[65-69]. As a result, large area metal-semiconductor rectifiers quickly became electrically reliable with stable and reproducible characteristics. The discovery and implementation of low temperature silicide formation enabled subsurface intermetallic-alloy-to-silicon contacts which were influenced much less strongly by surface contamination [70-72].

In the 1960s, planar Schottky diodes led to several important applications. Baird patented a Schottky barrier clamp integrated with a silicon bipolar transistor in 1964 [73].

Mead reported the metal-semiconductor field effect transistor (MESFET) in 1966 [74]. The development of miniature planar Schottky varactors and varistors for microwave applications also took off in the 1960s [75, 76]. The constant demand for higher frequency and faster switching semiconductor devices has led to commercial applications for

Schottky diodes including saturation-preventing clamps in high speed bipolar integrated circuits for computers and other switching networks, a variety of high power Schottky diodes used for circuit protection, discrete high frequency diodes and transistors for signal detection and amplification circuits in microwave communication systems [3].

A complete Schottky diode suitable for embedding in a circuit includes the rectifying (Schottky) metal/semiconductor junction and a non-rectifying (Ohmic) contact

(see Figure 1-5). In the figure, a SiO2 passivation layer is shown in the device. This feature is common in many Schottky diodes to assist in isolation and can also help to increase the

12 efficiency of Schottky solar cells [77-79]. The guard ring helps to alleviate edge enhanced electric field effects and improves the reverse breakdown threshold [80, 81]). Typical metals used to establish Schottky contact with semiconductors are Aluminum (Al) [82-84],

Tungsten (W) [85-87], Nickel (Ni) [88-90] and metal silicides [91-94]. More recently, carbon and its polytypes such as graphite [95-97], graphene [98-100] and carbon nanotube

[101-103] have been employed as the Schottky contact materials to a variety of semiconducting materials. Chemical and thermal stability are primary drivers in these investigations but the wide range of electrical properties available when forming carbon thin films is also a distinct benefit when comparing this material with metallic contact layers. The design, fabrication and characterization of energetically deposited carbon

Schottky contacts has been the main focus of the work reported in this thesis.

Figure 1-5: Structure of a Schottky diode with Schottky (rectifying) and Ohmic (non- rectifying) back metal contact [77-79].

13

1.6.1 Brief overview of metal-semiconductor junction formation and electrical properties

Metal-semiconductor contacts in semiconductor devices can be rectifying Schottky barrier contacts or Ohmic contacts [104]. The linear behaviour of Ohmic contacts is indispensable in semiconductor devices [105]. Ohmic behaviour depends on the relationship between the work function of the metal (휙푚) and the work function of the semiconductor (휙푠) [22]. In the case where the metal work function is greater than the semiconductor work function (휙푚 > 휙푠), the energy bands of an n-type semiconductor bend upwards and hence cause a barrier which hinders electrons crossing the junction.

Given the same difference in work functions and a p-type semiconductor holes can move freely across the junction [3]. Hence, the n-type semiconductor junction is rectifying

(Figure 1-6a) while Ohmic behaviour is exhibited by the same metal on the p-type semiconductor (Figure 1-6b). Similarly, for the case where the metal work function is less than semiconductor work function (휙푚 > 휙푠), Ohmic behavior is exhibited by the metal on the n-type semiconductor (Figure 1-6c) and rectifying behaviour is exhibited by the same metal on the p-type semiconductor (Figure 1-6d). Therefore, to make an , it is necessary to find metal-semiconductor combination with ϕm < ϕs when the semiconductor is n-type or ϕm > ϕs when it is p-type. However, there are few combinations with these properties [106, 107].

The majority of Ohmic contacts are made by creating a thin highly doped semiconductor layer adjacent to the metal [106]. Among numerous methods, this layer can be created by heat treatment [108]. Under high temperature, the metal and the

14 semiconductor form an alloy which contains a high concentration of electrically active elements that cause low electrical resistance [109].

Figure 1-6: Thermal equilibrium energy band diagram for metal contact with n-type and p-type semiconductor (modified from [23]).

According to Sze [22, 23], the typical current voltage characteristic of a Schottky diode follows the form shown in Figure 1-7. Typical Schottky diodes have rectifying characteristics that allow current to flow in forward bias (for voltages greater than a characteristic threshold voltage) and prevent current flow in reverse bias.

15

Figure 1-7: Typical current-voltage characteristic of n-type Schottky diode. The inset is a simplified schematic of a n-type 6H-SiC Schottky diode.

1.6.2 Applications

Schottky diodes are used in a wide range of applications from basic bipolar integrated circuits to optical, nuclear, and microwave applications [3, 110, 111]. Schottky diodes also exhibit advantages over p-n diodes because they are majority carrier devices

[112]. These advantages include: fast switching speed, low forward voltage drop, and compact device size [39]. Disadvantages of Schottky diodes include their high dependency on surface contamination and the performance instability at high operating temperatures

(>300oC) [106].

1.7 Carbon materials

Carbon materials such as graphite or diamond have existing industrial applications

16 due to their outstanding durability. Carbon thin films have been utilized in highly efficient batteries [113-116], electrodes [117-120], sensors [121-124] and aerospace/assembly applications [125, 126]. Many investigators believe that there still remains untapped potential in these materials for further electronic applications. The discoveries of carbon nanotubes and graphene have reinforced this belief and the enormous research interest is reflected in the number of publications focusing on carbon-based materials (three time greater than Silicon publications) (Figure 1-8) [127, 128].

Figure 1-8: Carbon journal publications per year since 1970 (Source: ScopusTM search engine with titles “Carbon”, “Graphite”, “Graphene”, “Carbon Nanotube” or “Diamond”,

April 02, 2019)

In regard to Schottky contact materials, beyond popular choices such as Ni, Al, Ti,

Au or Pt [129-134], carbon appears to be an excellent candidate. Carbon and its polytypes

17 such as graphite, graphene, or amorphous carbon are establishing themselves as novel materials in semiconductor research [2, 135, 136]. The literature shows that interconnections coated by graphene can provide suitable electrical and endurance properties to improve or replace conventional Copper interconnects [137, 138]. Graphene, with low contact resistance and high-power compliance, was recently reported to be an advantageous material as a Schottky contact to Si [99, 139, 140].

Furthermore, carbon is one of the most commonly occurring elements and can be found in many different forms. The atoms of carbon bond together in several ways, creating different polytypes. Among these polytypes, the best known are graphite, diamond and fullerenes [141]. Experimental studies show that graphite exhibits good electrical conduction while diamond, carbon nanotubes and graphene have high thermal conductivity

[142, 143] and are considered as promising alternatives for use in future high-performance devices.

Carbon has three different bonding states: sp, sp2 and sp3 which are shown schematically in Figure 1-9 [118, 144, 145]. Each hybridization state of the carbon atom exhibits distinct electrical and chemical characteristic.

Figure 1-9: Hybridization of the carbon atom [118, 144, 145].

18

In carbon thin film applications, graphitic carbon, amorphous carbon, graphene, carbon nanotubes and diamond have been exploited. Structures of these allotropes are shown in Figure 1-10 [1, 118].

Figure 1-10: Some allotropes of carbon: a) diamond; b) graphite; c) graphene; d)

amorphous carbon; e) carbon nanotube [146-150].

1.8 Experiment activities

In this work, Schottky diodes were fabricated on Si, 6H-SiC and β-Ga2O3 substrates using energetic deposition to make (largely graphitic) carbon contacts. Deposition conditions were controlled to manipulate the bonding within the carbon layers. The fabrication processes included wafer cleaning, die dicing, spin coating, mask alignment, optical lithography, energetic deposition and metal coating. Devices were characterized electrically by current-voltage (I-V) and capacitance-voltage (C-V) measurements under varying temperatures. Material characterizations were also conducted, including scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy and physical characteristic measurements (Hall and film thickness

19 measurement). Details of these fabrication and characterization processes are provided in chapter 3.

1.9 Thesis organization

This thesis is organized as follows:

• This first chapter provides a background together with motivations and objectives

of the thesis.

• Chapter 2 provides a review of relevant literature including previous reports on

devices using carbon device layers.

• Chapter 3 describes the fabrication and characterization techniques utilized in this

project to fabricate and analyze Schottky devices.

• Chapter 4 describes the temperature dependent electrical characteristics of

graphitic-C on Si Schottky diodes. Analysis based on these measurements is

discussed and compared with other reported characteristics from similar devices.

• Chapter 5 examines the temperature dependent electrical characteristics of

graphitic-C on 6H-SiC Schottky diodes.

• Chapter 6 analyzes the temperature dependent electrical characteristics of

graphitic-C on β-Ga2O3 Schottky diodes.

• Chapter 7 summarizes the thesis findings and discusses possible future work.

20

1.10 References

[1] F. Kreupl, "Carbon-based materials as key-enabler for “more than moore”," MRS Online Proceedings Library Archive, vol. 1303, 2011. [2] M. Stelzer and F. Kreupl, "Graphenic carbon-silicon contacts for reliability improvement of metal-silicon junctions," in 2016 IEEE International Electron Devices Meeting (IEDM), 2016, pp. 21.7.1-21.7.4. [3] V. Rideout, "A review of the theory, technology and applications of metal- semiconductor rectifiers," Thin Solid Films, vol. 48, no. 3, pp. 261-291, 1978. [4] J. Biela, M. Schweizer, S. Waffler, B. Wrzecionko, and J. W. Kolar, "SiC vs. Si - Evaluation of Potentials for Performance Improvement of Power Electronics Converter Systems by SiC Power Semiconductors," Materials Science Forum, vol. 645-648, pp. 1101-1106, 2010. [5] S. Dimitrijev, "Silicon carbide as a material for mainstream electronics," Microelectronic Engineering, vol. 83, no. 1, pp. 123-125, 2006/01/01/ 2006. [6] C. Hou, R. M. Gazoni, R. J. Reeves, and M. W. Allen, "Oxidized Metal Schottky Contacts on (010) $\beta $-Ga 2 O 3," IEEE Electron Device Letters, vol. 40, no. 2, pp. 337-340, 2019. [7] T.-H. Hung et al., "Energy band line-up of atomic layer deposited Al2O3 on β- Ga2O3," Applied Physics Letters, vol. 104, no. 16, p. 162106, 2014. [8] M. Higashiwaki et al., "Recent progress in Ga2O3 power devices," Semiconductor Science and Technology, vol. 31, no. 3, p. 034001, 2016. [9] S. Pearton et al., "A review of Ga2O3 materials, processing, and devices," Applied Physics Reviews, vol. 5, no. 1, p. 011301, 2018. [10] R. R. Haberecht, Semiconductor silicon. Electronics Division, Electrochemical Society, 1969. [11] G. Teal, "GK Teal and JB Little Bull. Am. Phys. Soc. 25, No. 2, 16 (1950)," Bull. Am. Phys. Soc., vol. 25, p. 16, 1950. [12] J. Bardeen and W. H. Brattain, "Physical principles involved in transistor action," Physical Review, vol. 75, no. 8, p. 1208, 1949. [13] W. Shockley, J. Bardeen, and W. Brattain, "The electronic theory of the transistor," in Science, 1948, vol. 108, no. 2816, pp. 678-679: AMER ASSOC ADVANCEMENT SCIENCE 1200 NEW YORK AVE, NW, WASHINGTON, DC 20005. [14] H. Theuerer, "Method of processing semiconductive materials," US Patent, vol. 3, p. 123, 1962. [15] P. H. Keck and M. J. Golay, "Crystallization of silicon from a floating liquid zone," Physical Review, vol. 89, no. 6, p. 1297, 1953. [16] W. G. Pfann, "Principles of zone-melting," JOM, vol. 4, no. 7, pp. 747-753, 1952. [17] E. Spenke, C. Pulling, and Z. Refining, "History and Future Needs in Silicon Technology," in Semiconductor Silicon: The Electrochemical Society, 1969, p. 12.

21

[18] H. John, "Silicon power device material problems," Proceedings of the IEEE, vol. 55, no. 8, pp. 1249-1271, 1967. [19] C. Jacoboni, C. Canali, G. Ottaviani, and A. A. Quaranta, "A review of some charge transport properties of silicon," Solid-State Electronics, vol. 20, no. 2, pp. 77-89, 1977. [20] R. Hull, Properties of crystalline silicon (no. 20). IET, 1999. [21] H. S. Nalwa, Silicon-based material and devices, two-volume set: materials and processing, Properties and Devices. Academic Press, 2001. [22] S. M. Sze, Semiconductor devices: physics and technology. John Wiley & Sons, 2008. [23] S. M. Sze and K. K. Ng, Physics of semiconductor devices, 3rd ed. John Wiley & Sons, 2006. [24] B. Sharma, Metal-semiconductor Schottky barrier junctions and their applications. Springer Science & Business Media, 2013. [25] C. Ting and M. Wittmer, "The use of titanium-based contact barrier layers in silicon technology," Thin Solid Films, vol. 96, no. 4, pp. 327-345, 1982. [26] A. Cowley, "Titanium-silicon Schottky barrier diodes," Solid-State Electronics, vol. 13, no. 4, pp. 403-414, 1970. [27] M. Sinha, E. F. Chor, and Y.-C. Yeo, "Tuning the Schottky barrier height of nickel silicide on p-silicon by aluminum segregation," Applied Physics Letters, vol. 92, no. 22, p. 222114, 2008. [28] Y.-J. Chang and J. Erskine, "Diffusion layers and the Schottky-barrier height in nickel silicide—silicon interfaces," Physical Review B, vol. 28, no. 10, p. 5766, 1983. [29] E. Philofsky and E. Hall, "A review of the limitations of aluminum thin films on semiconductor devices," IEEE Transactions on Parts, Hybrids, and Packaging, vol. 11, no. 4, pp. 281-290, 1975. [30] S. Riegel, F. Mutter, T. Lauermann, B. Terheiden, and G. Hahn, "Review on screen printed metallization on p-type silicon," Energy Procedia, vol. 21, pp. 14- 23, 2012. [31] J. W. Palmour, "Silicon carbide power device development for industrial markets," in 2014 IEEE International Electron Devices Meeting, 2014, pp. 1.1. 1- 1.1. 8: IEEE. [32] J. Millán, P. Godignon, X. Perpiñà, A. Pérez-Tomás, and J. Rebollo, "A survey of wide bandgap power semiconductor devices," IEEE transactions on Power Electronics, vol. 29, no. 5, pp. 2155-2163, 2013. [33] T. Nakamura, M. Sasagawa, Y. Nakano, T. Otsuka, and M. Miura, "Large current SiC power devices for automobile applications," in The 2010 International Power Electronics Conference-ECCE ASIA-, 2010, pp. 1023-1026: IEEE. [34] Y.-B. Guo, K. P. Bhat, A. Aravamudhan, D. C. Hopkins, and D. R. Hazelmyer, "High current and thermal transient design of a SiC SSPC for aircraft application," in 2011 Twenty-Sixth Annual IEEE Applied Power Electronics Conference and Exposition (APEC), 2011, pp. 1290-1297: IEEE.

22

[35] T. P. Chow, "SiC and GaN high-voltage power switching devices," in Materials Science Forum, 2000, vol. 338, pp. 1155-1160: Trans Tech Publications Ltd., Zurich-Uetikon, Switzerland. [36] M. Bhatnagar and B. J. Baliga, "Comparison of 6H-SiC, 3C-SiC, and Si for power devices," IEEE Transactions on Electron Devices, vol. 40, no. 3, pp. 645-655, 1993. [37] D. M. Brown, M. Ghezzo, J. Kretchmer, E. Downey, J. Pimbley, and J. Palmour, "SiC MOS interface characteristics," IEEE Transactions on Electron Devices, vol. 41, no. 4, pp. 618-620, 1994. [38] J. Casady and R. W. Johnson, "Status of silicon carbide (SiC) as a wide-bandgap semiconductor for high-temperature applications: A review," Solid-State Electronics, vol. 39, no. 10, pp. 1409-1422, 1996. [39] R. Singh, J. A. Cooper, M. R. Melloch, T. Chow, and J. W. Palmour, "SiC power Schottky and PiN diodes," IEEE Transactions on Electron Devices, vol. 49, no. 4, pp. 665-672, 2002. [40] J. D. van Wyk and F. C. Lee, "On a future for power electronics," IEEE Journal of Emerging and Selected Topics in Power Electronics, vol. 1, no. 2, pp. 59-72, 2013. [41] E. Bahat-Treidel, GaN-Based HEMTs for High Voltage Operation: Design, Technology and Characterization. Cuvillier Verlag, 2012. [42] M. Bhatnagar and B. J. Baliga, "Comparison of 6H-SiC, 3C-SiC, and Si for power devices," Electron Devices, IEEE Transactions on, vol. 40, no. 3, pp. 645-655, 1993. [43] G. Ferro and D. Chaussende, "A new model for in situ nitrogen incorporation into 4H-SiC during epitaxy," Scientific reports, vol. 7, p. 43069, 2017. [44] C. Carter Jr et al., "Progress in SiC: from material growth to commercial device development," Materials Science and Engineering: B, vol. 61, pp. 1-8, 1999. [45] M. Brady et al., "Silicon carbide crystal and substrate technology: A survey of recent advances," in Materials Science Forum, 2004, vol. 457, pp. 3-8: Trans Tech Publ. [46] J. Li, L. Porter, and S. Yip, "Atomistic modeling of finite-temperature properties of crystalline β-SiC: II. Thermal conductivity and effects of point defects," Journal of Nuclear Materials, vol. 255, no. 2-3, pp. 139-152, 1998. [47] G. Rutsch and M. Werner, "Perspectives in rugged microsystems for harsh environments," MST News, vol. 2, pp. 4-6, 1998. [48] M. R. Werner and W. R. Fahrner, "Review on materials, microsensors, systems and devices for high-temperature and harsh-environment applications," IEEE Transactions on Industrial Electronics, vol. 48, no. 2, pp. 249-257, 2001. [49] V. D. Krstic, "Production of fine, high‐purity beta silicon carbide powders," Journal of the American Ceramic Society, vol. 75, no. 1, pp. 170-174, 1992. [50] H. Gotou, "Method for fabricating a silicon carbide substrate," ed: Google Patents, 1991.

23

[51] C. Eddy and D. Gaskill, "Silicon carbide as a platform for power electronics," Science, vol. 324, no. 5933, pp. 1398-1400, 2009. [52] H. Nagasawa, T. Kawahara, and K. Yagi, "Method of manufacturing silicon carbide, silicon carbide, composite material, and semiconductor element," ed: Google Patents, 2004. [53] R. Stevenson, "Power electronics' cool new flavor [news]," IEEE Spectrum, vol. 53, no. 4, pp. 11-12, 2016. [54] E. Farzana, Z. Zhang, P. K. Paul, A. R. Arehart, and S. A. Ringel, "Influence of metal choice on (010) β-Ga2O3 Schottky diode properties," Applied Physics Letters, vol. 110, no. 20, p. 202102, 2017. [55] C. Hou, R. Gazoni, R. Reeves, and M. Allen, "Direct comparison of plain and oxidized metal Schottky contacts on β-Ga2O3," Applied Physics Letters, vol. 114, no. 3, p. 033502, 2019. [56] C. Hou et al., "High-temperature (350° C) oxidized iridium Schottky contacts on β-Ga2O3," Applied Physics Letters, vol. 114, no. 23, p. 233503, 2019. [57] G.-S. Chung and J.-H. Ahn, "Characterization of polycrystalline 3C–SiC thin film diodes for extreme environment applications," Microelectronic Engineering, vol. 85, no. 8, pp. 1772-1775, 2008. [58] I. Josan et al., "Extreme environment temperature sensor based on silicon carbide Schottky diode," in 2009 International Semiconductor Conference, 2009, vol. 2, pp. 525-528: IEEE. [59] B. Chatterjee, A. Jayawardena, E. Heller, D. W. Snyder, S. Dhar, and S. Choi, "Thermal characterization of gallium oxide Schottky barrier diodes," Review of Scientific Instruments, vol. 89, no. 11, p. 114903, 2018. [60] F. Braun, "Ueber die Stromleitung durch Schwefelmetalle," Annalen der Physik, vol. 229, no. 12, pp. 556-563, 1875. [61] J. Bardeen, "Research leading to point-contact transistor," Science, vol. 126, no. 3264, pp. 105-112, 1957. [62] H. C. Torrey and C. A. Whitmer, "Crystal rectifiers," 1948. [63] G. Bussey, Vintage Crystal Sets, 1922-1927. IPC Electrical-Electronic Press Limited, 1976. [64] W. Brattain and J. Bardeen, "Nature of the forward current in germanium point contacts," Physical review, vol. 74, no. 2, p. 231, 1948. [65] C. Crowell and S. Sze, "Hot-electron transport and electron tunneling in thin film structures," Physics of thin films, vol. 4, pp. 325-371, 1967. [66] C. Crowell and S. Sze, "Quantum‐Mechanical Reflection of Electrons at Metal‐ Semiconductor Barriers: Electron Transport in Semiconductor‐Metal‐ Semiconductor Structures," Journal of Applied Physics, vol. 37, no. 7, pp. 2683- 2689, 1966. [67] D. Kahng, "Conduction properties of the Au-n-type—Si Schottky barrier," Solid- State Electronics, vol. 6, no. 3, pp. 281-295, 1963. [68] R. Archer and M. Atalla, "Metals contacts on cleaved silicon surfaces," Annals of the New York Academy of Sciences, vol. 101, no. 3, pp. 697-708, 1963.

24

[69] J. A. Hoerni, "Method of manufacturing semiconductor devices," ed: Google Patents, 1962. [70] J. Andrews, "The role of the metal-semiconductor interface in silicon integrated circuit technology," Journal of Vacuum Science and Technology, vol. 11, no. 6, pp. 972-984, 1974. [71] M. Lepselter and S. Sze, "Silicon Schottky barrier diode with near‐ideal I‐V characteristics," Bell System Technical Journal, vol. 47, no. 2, pp. 195-208, 1968. [72] D. Kahng and M. Lepselter, "BSTJ briefs: Planar epitaxial silicon schottky barrier diodes," The Bell System Technical Journal, vol. 44, no. 7, pp. 1525-1528, 1965. [73] J. R. Biard, "Unitary semiconductor high speed switching device utilizing a barrier diode," ed: Google Patents, 1969. [74] C. Mead, "Schottky barrier gate field effect transistor," Proceedings of the IEEE, vol. 54, no. 2, pp. 307-308, 1966. [75] H. A. Watson and M. R. Barber, Microwave semiconductor devices and their circuit applications. McGraw-Hill, 1969. [76] J. Irvin and N. Vanderwal, "Schottky barrier devices," Microwave Semiconductor Devices and Their Circuit Applications, pp. 340-369, 1969. [77] S. W. Glunz and F. Feldmann, "SiO 2 surface passivation layers–a key technology for silicon solar cells," Solar Energy Materials and Solar Cells, vol. 185, pp. 260-269, 2018. [78] J. Schmidt, A. Merkle, R. Brendel, B. Hoex, M. v. de Sanden, and W. Kessels, "Surface passivation of high‐efficiency silicon solar cells by atomic‐layer‐ deposited Al2O3," Progress in photovoltaics: research and applications, vol. 16, no. 6, pp. 461-466, 2008. [79] J. Schmidt, M. Kerr, and A. Cuevas, "Surface passivation of silicon solar cells using plasma-enhanced chemical-vapour-deposited SiN films and thin thermal SiO2/plasma SiN stacks," Semiconductor Science and Technology, vol. 16, no. 3, p. 164, 2001. [80] A. Yu and C. Mead, "Characteristics of aluminum-silicon Schottky barrier diode," Solid-State Electronics, vol. 13, no. 2, pp. 97-104, 1970. [81] K. Ueno, T. Urushidani, K. Hashimoto, and Y. Seki, "The guard-ring termination for the high-voltage SiC Schottky barrier diodes," IEEE Electron device letters, vol. 16, no. 7, pp. 331-332, 1995. [82] F. Yakuphanoglu, M. Kandaz, and B. Senkal, "Current–voltage and capacitance– voltage characteristics of Al/p-type silicon/organic semiconductor based on phthalocyanine contact," Thin Solid Films, vol. 516, no. 23, pp. 8793- 8796, 2008. [83] M. E. Aydin, F. Yakuphanoglu, J.-H. Eom, and D.-H. Hwang, "Electrical characterization of Al/MEH-PPV/p-Si Schottky diode by current–voltage and capacitance–voltage methods," Physica B: Condensed Matter, vol. 387, no. 1-2, pp. 239-244, 2007.

25

[84] T. Reith and J. Schick, "The electrical effect on Schottky barrier diodes of Si crystallization from Al–Si metal films," Applied Physics Letters, vol. 25, no. 9, pp. 524-526, 1974. [85] M. Khaidar, A. Essafti, A. Bennouna, E. Ameziane, and M. Brunel, "rf‐sputtered tungsten‐amorphous silicon Schottky barrier diodes," Journal of applied physics, vol. 65, no. 8, pp. 3248-3252, 1989. [86] S. Babcock and K. Tu, "Titanium‐tungsten contacts to silicon. II. Stability against aluminum penetration," Journal of applied physics, vol. 59, no. 5, pp. 1599-1605, 1986. [87] S. Babcock and K. Tu, "Titanium‐tungsten contacts to Si: The effects of alloying on Schottky contact and on silicide formation," Journal of Applied Physics, vol. 53, no. 10, pp. 6898-6905, 1982. [88] D. Coe and E. Rhoderick, "Silicide formation in Ni-Si Schottky barrier diodes," Journal of Physics D: Applied Physics, vol. 9, no. 6, p. 965, 1976. [89] S. Woodruff et al., "Nickel and nickel silicide Schottky barrier contacts to n-type silicon ," Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena, vol. 26, no. 4, pp. 1592-1596, 2008. [90] Y. Jia, R. C. Shi, and S. Y. Chou, "Spin-valve effects in nickel/silicon/nickel junctions," IEEE Transactions on Magnetics, vol. 32, no. 5, pp. 4707-4709, 1996. [91] P. S. Ho, E. S. Yang, H. L. Evans, and X. Wu, "Electronic states at silicide-silicon interfaces," Physical review letters, vol. 56, no. 2, p. 177, 1986. [92] I. Ohdomari, T. Kuan, and K. Tu, "Microstructure and Schottky barrier height of iridium silicides formed on silicon," Journal of Applied Physics, vol. 50, no. 11, pp. 7020-7029, 1979. [93] W. Mönch, "Role of virtual gap states and defects in metal-semiconductor contacts," in Electronic Structure of Metal-Semiconductor Contacts: Springer, 1990, pp. 224-227. [94] H. Norde, J. de Sousa Pires, F. d’Heurle, F. Pesavento, S. Petersson, and P. Tove, "The Schottky‐barrier height of the contacts between some rare‐earth metals (and silicides) and p‐type silicon," Applied Physics Letters, vol. 38, no. 11, pp. 865- 866, 1981. [95] N. Ooi, A. Rairkar, and J. B. Adams, "Density functional study of graphite bulk and surface properties," Carbon, vol. 44, no. 2, pp. 231-242, 2006. [96] R. Yatskiv and J. Grym, "Temperature-dependent properties of semimetal graphite-ZnO Schottky diodes," Applied Physics Letters, vol. 101, no. 16, p. 162106, 2012. [97] E. Taft and L. Apker, "Photoelectric emission from polycrystalline graphite," Physical Review, vol. 99, no. 6, p. 1831, 1955. [98] S. J. Sque, R. Jones, and P. R. Briddon, "The transfer doping of graphite and graphene," physica status solidi (a), vol. 204, no. 9, pp. 3078-3084, 2007. [99] D. Sinha and J. U. Lee, "Ideal graphene/silicon Schottky junction diodes," Nano letters, vol. 14, no. 8, pp. 4660-4664, 2014.

26

[100] X. Li et al., "Graphene‐on‐silicon Schottky junction solar cells," Advanced materials, vol. 22, no. 25, pp. 2743-2748, 2010. [101] S. Heinze, J. Tersoff, R. Martel, V. Derycke, J. Appenzeller, and P. Avouris, "Carbon nanotubes as Schottky barrier transistors," Physical Review Letters, vol. 89, no. 10, p. 106801, 2002. [102] M. Yang, K. Teo, W. Milne, and D. Hasko, "Carbon nanotube Schottky diode and directionally dependent field-effect transistor using asymmetrical contacts," Applied Physics Letters, vol. 87, no. 25, p. 253116, 2005. [103] H. M. Manohara, E. W. Wong, E. Schlecht, B. D. Hunt, and P. H. Siegel, "Carbon nanotube Schottky diodes using Ti− Schottky and Pt− Ohmic contacts for high frequency applications," Nano letters, vol. 5, no. 7, pp. 1469-1474, 2005. [104] R. T. Tung, "Recent advances in Schottky barrier concepts," Materials Science and Engineering: R: Reports, vol. 35, no. 1-3, pp. 1-138, 2001. [105] M. M. E. Labotary. (2016). Ohmic contact science. Available: http://www.mtl.kyoto-u.ac.jp/english/laboratory/micro/micro.htm [106] E. H. Rhoderick, "Metal-semiconductor contacts," IEE Proceedings I-Solid-State and Electron Devices, vol. 129, no. 1, p. 1, 1982. [107] M. Kracica, E. Mayes, H. Tran, A. Holland, D. McCulloch, and J. Partridge, "Rectifying electrical contacts to n-type 6H–SiC formed from energetically deposited carbon," Carbon, vol. 102, pp. 141-144, 2016. [108] J.-K. Ho et al., "Low-resistance ohmic contacts to p-type GaN achieved by the oxidation of Ni/Au films," Journal of Applied Physics, vol. 86, no. 8, pp. 4491- 4497, 1999. [109] P. Barnes and A. Cho, "Nonalloyed Ohmic contacts to n‐GaAs by molecular beam epitaxy," Applied Physics Letters, vol. 33, no. 7, pp. 651-653, 1978. [110] L. Lorenz, G. Deboy, and I. Zverev, "Matched pair of CoolMOS transistor with SiC-Schottky diode-advantages in application," IEEE Transactions on Industry Applications, vol. 40, no. 5, pp. 1265-1272, 2004. [111] A. Itoh and H. Matsunami, "Analysis of Schottky barrier heights of metal/SiC contacts and its possible application to high‐voltage rectifying devices," physica status solidi (a), vol. 162, no. 1, pp. 389-408, 1997. [112] J. H. Werner, "Schottky barrier and pn-junctionI/V plots—Small signal evaluation," Applied physics A, vol. 47, no. 3, pp. 291-300, 1988. [113] I. Lahiri et al., "High capacity and excellent stability of lithium ion battery anode using interface-controlled binder-free multiwall carbon nanotubes grown on copper," ACS nano, vol. 4, no. 6, pp. 3440-3446, 2010. [114] D. Pech et al., "Ultrahigh-power micrometre-sized supercapacitors based on onion-like carbon," Nature nanotechnology, vol. 5, no. 9, p. 651, 2010. [115] P. Simon and Y. Gogotsi, "Materials for electrochemical capacitors," in Nanoscience And Technology: A Collection of Reviews from Nature Journals: World Scientific, 2010, pp. 320-329. [116] M. Armand and J.-M. Tarascon, "Building better batteries," nature, vol. 451, no. 7179, p. 652, 2008.

27

[117] P. V. Huong, "Structural studies of diamond films and ultrahard materials by Raman and micro-Raman spectroscopies," Diamond and Related Materials, vol. 1, no. 1, pp. 33-41, 1991. [118] A. Grill, "Diamond-like carbon: state of the art," Diamond and related materials, vol. 8, no. 2-5, pp. 428-434, 1999. [119] G. Che, B. B. Lakshmi, E. R. Fisher, and C. R. Martin, "Carbon nanotubule membranes for electrochemical energy storage and production," Nature, vol. 393, no. 6683, p. 346, 1998. [120] S. Iijima, "Helical microtubules of graphitic carbon," nature, vol. 354, no. 6348, p. 56, 1991. [121] Q. Zhao, Z. Gan, and Q. Zhuang, "Electrochemical sensors based on carbon nanotubes," Electroanalysis: An International Journal Devoted to Fundamental and Practical Aspects of Electroanalysis, vol. 14, no. 23, pp. 1609-1613, 2002. [122] A. Modi, N. Koratkar, E. Lass, B. Wei, and P. M. Ajayan, "Miniaturized gas ionization sensors using carbon nanotubes," Nature, vol. 424, no. 6945, p. 171, 2003. [123] K. G. Ong, K. Zeng, and C. A. Grimes, "A wireless, passive carbon nanotube- based gas sensor," IEEE Sensors Journal, vol. 2, no. 2, pp. 82-88, 2002. [124] A. A. Kosterev et al., "Thermoelectrically cooled quantum-cascade-laser-based sensor for the continuous monitoring of ambient atmospheric carbon monoxide," Applied optics, vol. 41, no. 6, pp. 1169-1173, 2002. [125] M. Mecklenburg et al., "Aerographite: ultra lightweight, flexible nanowall, carbon microtube material with outstanding mechanical performance," Advanced Materials, vol. 24, no. 26, pp. 3486-3490, 2012. [126] A. Krauss et al., "Ultrananocrystalline diamond thin films for MEMS and moving mechanical assembly devices," Diamond and Related Materials, vol. 10, no. 11, pp. 1952-1961, 2001. [127] J. D. Buckley and D. D. Edie, Carbon-carbon materials and composites. William Andrew, 1993. [128] T. W. Tombler et al., "Reversible electromechanical characteristics of carbon nanotubes underlocal-probe manipulation," Nature, vol. 405, no. 6788, p. 769, 2000. [129] M. E. Aydın, N. Yıldırım, and A. Türüt, "Temperature-dependent behavior of Ni/4 H-n SiC Schottky contacts," Journal of applied physics, vol. 102, no. 4, p. 043701, 2007. [130] N. Papanicolaou, A. Christou, and M. Gipe, "Pt and PtSi x Schottky contacts on n‐type β‐SiC," Journal of Applied Physics, vol. 65, no. 9, pp. 3526-3530, 1989. [131] O. Ageev et al., "Au-TiB x-n-6H-SiC Schottky barrier diodes: Specific features of charge transport in rectifying and nonrectifying contacts," Semiconductors, vol. 43, no. 7, pp. 865-871, 2009. [132] J. Waldrop, "Schottky barrier height of metal contacts to p‐type alpha 6H‐SiC," Journal of applied physics, vol. 75, no. 9, pp. 4548-4550, 1994.

28

[133] A. Itoh, T. Kimoto, and H. Matsunami, "High performance of high-voltage 4H- SiC Schottky barrier diodes," IEEE Electron Device Letters, vol. 16, no. 6, pp. 280-282, 1995. [134] V. Saxena, J. N. Su, and A. J. Steckl, "High-voltage Ni-and Pt-SiC Schottky diodes utilizing metal field plate termination," IEEE transactions on electron devices, vol. 46, no. 3, pp. 456-464, 1999. [135] H. Pham, H. N. Tran, A. S. Holland, and J. G. Partridge, "Temperature-Dependent Electrical Characteristics and Extraction of Richardson Constant from Graphitic- C/n-Type 6H-SiC Schottky Diodes," Journal of Electronic Materials, pp. 1-6, 2019. [136] M. Stelzer, M. Jung, and F. Kreupl, "Graphenic Carbon: A Novel Material to Improve the Reliability of Metal-Silicon Contacts," IEEE Journal of the Electron Devices Society, vol. 5, no. 5, pp. 416-425, 2017. [137] J. Jiang, J. Kang, and K. Banerjee, "Characterization of self-heating and current- carrying capacity of intercalation doped graphene-nanoribbon interconnects," in 2017 IEEE International Reliability Physics Symposium (IRPS), 2017, pp. 6B- 1.1-6B-1.6: IEEE. [138] L. Li, Z. Zhu, T. Wang, J. A. Currivan-Incorvia, A. Yoon, and H.-S. P. Wong, "BEOL compatible graphene/Cu with improved electromigration lifetime for future interconnects," in 2016 IEEE International Electron Devices Meeting (IEDM), 2016, pp. 9.5. 1-9.5. 4: IEEE. [139] Y. Song et al., "Role of interfacial oxide in high-efficiency graphene–silicon Schottky barrier solar cells," Nano letters, vol. 15, no. 3, pp. 2104-2110, 2015. [140] C.-C. Chen, M. Aykol, C.-C. Chang, A. Levi, and S. B. Cronin, "Graphene-silicon Schottky diodes," Nano letters, vol. 11, no. 5, pp. 1863-1867, 2011. [141] M. L. Terranova, V. Sessa, and M. Rossi, "The world of carbon nanotubes: an overview of CVD growth methodologies," Chemical Vapor Deposition, vol. 12, no. 6, pp. 315-325, 2006. [142] D. Chung, "Review graphite," Journal of materials science, vol. 37, no. 8, pp. 1475-1489, 2002. [143] R. Sengupta, M. Bhattacharya, S. Bandyopadhyay, and A. K. Bhowmick, "A review on the mechanical and electrical properties of graphite and modified graphite reinforced polymer composites," Progress in polymer science, vol. 36, no. 5, pp. 638-670, 2011. [144] D. R. Lide Jr, "A survey of carbon-carbon bond lengths," Tetrahedron, vol. 17, no. 3-4, pp. 125-134, 1962. [145] R. Haerle, E. Riedo, A. Pasquarello, and A. Baldereschi, "sp 2/s p 3 hybridization ratio in amorphous carbon from C 1 s core-level shifts: X-ray photoelectron spectroscopy and first-principles calculation," Physical Review B, vol. 65, no. 4, p. 045101, 2001. [146] A. Hirsch, "The era of carbon allotropes," Nature materials, vol. 9, no. 11, p. 868, 2010.

29

[147] Y. Liu, G. Wang, Q. Huang, L. Guo, and X. Chen, "Structural and electronic properties of T graphene: a two-dimensional carbon allotrope with tetrarings," Physical review letters, vol. 108, no. 22, p. 225505, 2012. [148] A. N. Enyashin and A. L. Ivanovskii, "Graphene allotropes," physica status solidi (b), vol. 248, no. 8, pp. 1879-1883, 2011. [149] T. Yamabe, K. Fukui, and K. Tanaka, The science and technology of carbon nanotubes. Elsevier, 1999. [150] R. Hoffmann, T. Hughbanks, M. Kertesz, and P. H. Bird, "Hypothetical metallic allotrope of carbon," Journal of the American Chemical Society, vol. 105, no. 14, pp. 4831-4832, 1983.

30

CHAPTER 2

BACKGROUND AND RECENT ADVANCES IN CARBON-BASED SCHOTTKY DEVICES

2.1 Introduction

Interfaces between semiconductors, metals, and/or insulators are ubiquitous in semiconductor technology. The focus of this project is on metal-semiconductor junctions and in particular, current-rectifying Schottky junctions. Carbon has emerged as a material with higher chemical/thermal endurance when compared to more conventional contact materials. In this project, energetically deposited carbon has been used to produce Schottky contacts to semiconductors. Silicon, silicon carbide and gallium oxide were selected as the semiconducting materials as both are extensively used in power electronics, a field in which the aforementioned thermal stability of carbon could provide significant benefits. In the remainder of this chapter, the investigated device materials are described and following this, the band structure, interface- and transport- properties of Schottky diodes are presented and finally, a chapter summary is provided.

2.2 Schottky diode

A Schottky contact is a semiconductor device formed from two layers: a metal (or semi-metal) and a semiconductor. A potential barrier formed from a space charge on the

31 semiconductor surface was first reported by Schottky and Mott in 1938 [1, 2]. In the

Schottky-Mott model for a Schottky junction, the barrier height is given by [1, 2]:

Φ퐵 = Φ푚 − χ Eq. 2-1

Where ΦB is the barrier height of the Schottky contact, Φm is the metal work function and

χ is the electron affinity of the semiconductor.

The potential barrier formed between the metal and semiconductor determines the flow of current and is shown in an energy band diagram in Figure 2-1. In Figure 2-1a, a metal and n-type semiconductor are shown separated. The metal work function Φm is defined as the energy difference between the Fermi level EF and the vacuum level, the electron affinity χ is the energy difference between the conduction band EC and the vacuum level and qVn is the energy difference between conduction band EC and Fermi level EF.

When the two layers are brought into contact, the electrons flow from the semiconductor conduction band to the metal, resulting in a region in the semiconductor near the interface being depleted of mobile charge. As a result, the bands are bent upwards at the interface and a potential barrier is formed, as shown in Figure 2-1b. The Fermi levels of the two materials are levelled in the absence of external bias. An electron must obtain sufficient energy to overcome the potential barrier and to flow from one side of the device to the other. The is described by:

2휖푠 푘푇 푊퐷 = √ (푞푉푏푖 − 푉 − ) Eq. 2-2 푞푁퐷 푞 where WD is the depletion width, εs is the permittivity of semiconductor, q is the electron charge, ND is the donor concentration, Vbi is the built-in potential, V is the applied voltage, k is Boltzmann’s constant, and T is the temperature in Kelvin.

32

Figure 2-1: (a) Energy band diagram of an isolated metal adjacent to an isolated n-type semiconductor. (b) Energy band diagram of the same metal and semiconductor in contact

(reproduced from [3]).

In theory, Eq. 2-1 shows a strong dependency between barrier height and metal work function dΦB/dΦM = 1. However, this relationship is rarely seen in experimental work

[4, 5]. The Schottky barrier height measured in actual experiments implies some relation to the preparation of the semiconductor surface. In other words, the Schottky barrier height depends not only on the metal work function but also the quality of the metal- semiconductor interface. In much of the reported literature, the actual dependence of

33

Schottky barrier height on the metal work function is much weaker (SΦ = dΦB/dΦM << 1) than indicated by the Schottky-Mott theory [4, 6-8]. This phenomenon is known as “Fermi- level pinning” [9, 10]. Furthermore, barrier inhomogeneity is also a problem which complicates the dependency. Barrier inhomogeneity is described as a fluctuation of barrier height of a Schottky contact as temperature changes or as the variation (called ideality factor, > 1) of the barrier height calculated from an I-V characteristic of a Schottky contact

[11-13]. The issue is observed in various kinds of Schottky contacts from traditional Si to novel materials such as SiC and Ga2O3 [8, 12, 14-29].

2.2.1 Interface states

As stated, in ideal cases, the barrier height of a Schottky diode follows the Schottky-

Mott relation Eq. 2-1. In real (‘non-ideal’) metal-semiconductor interfaces, a significant number of surface states are present on the surface of the semiconductor. These states significantly affect the barrier height and cause Fermi level pinning [9, 30]. The phenomenon referred to as "Fermi level pinning" results from the creation of some energy states in the band gap that set (pin) the Fermi level, with a dominating effect over the predicted effect described by Eq. 2-1. This pinning makes the Schottky barrier height almost completely insensitive to the metal work function [31-33]. The Fermi level pinning effect is prominent in many semiconductors such as Ge, Si and GaAs [10, 31, 34, 35]. To alleviate this problem, additional processing steps can be introduced. For example, adding an intermediate insulating layer between the semiconductor and metal is known to facilitate unpinning of the bands [36-38].

34

Interface states can even be created in the absence of native defects by tunneling of metal electron wave functions into the semiconductor gap [39, 40]. These are known as metal induced gap states (MIGS) and have been extensively modeled by Mönch [18].

MIGS have a tendency to pin the barrier height close to the charge neutrality level ΦCNL of the semiconductor. In a semiconductor, the most common and most prominent sources of charge are electrons, holes, ionized acceptors and ionized donors. When all the charge in a given volume sum to zero, charge neutrality level is achieved [3, 41]. In other words, in a uniformly doped semiconductor the negative charge associated with an electron or ionized acceptor would be canceled by the positive charge associated with a hole or ionized donor.

As a result, the homogeneous barrier height is modeled as:

Φ퐵 = S(푋푚 − 푋푠) + Φ퐶푁퐿 Eq. 2-3

Where S = dΦB/dXm is the slope parameter, while the electronegativity of the metal Xm usually varies linearly with Φm [18]. ΦCNL is the energy level at which the MIGS change from exhibiting a donor to an acceptor like nature and its value is often close to the center of the gap (Eg/2) [42]. The slope parameter S between barrier height and electronegativity can be observed in Figure 2-2.

35

Figure 2-2: Barrier height versus electronegativity of metals deposited on Si, GaSe, and

SiO2 [41].

2.2.2 Image force lowering

Image-force lowering, also known as Schottky-barrier lowering, is an image-force induced by electrons near the metal surface hence causing the lowering of barrier energy

(at a metal-semiconductor interface). Sze explained that when an electron is positioned at a distance x from the metal, a positive charge is induced on the metal surface; hence the attractive force of the electron toward the metal can be expressed as [41]:

−q2 Eq. 2-4 F(x) = 2 16휋휀표푥

Where εo is the permittivity of free space, x is the distance from metal surface and q is the electron charge. According to Sze, the magnitude of the image force lowering ΔΦ and the

36 location of the lowering xm can be calculated by evaluating the maximum field ξm of the metal-semiconductor interface from the potential energy and external electric field ξ [41]:

푞휉 ΔΦ = √ 푚 Eq. 2-5 4휋휀푠

푞 x푚 = √ Eq. 2-6 16휋휀표|휉푚|

In non-ideal Schottky diodes, the maximum electric field at the surface can be approximated by [41]:

2푞푁퐷|Ψ푠| ξ푚 = √ Eq. 2-7 휀푠

Where ψs is the surface potential at the metal-semiconductor interface. Therefore, the image force lowering can be calculated from:

1/4 푞3푁 |Ψ | 퐷 푠 Eq. 2-8 ΔΦ = ( 2 3 ) 8휋 휀푠

The Schottky barrier lowering effect for a metal on n-type semiconductor under different biasing conditions is illustrated in Figure 2-3. For forward bias (V > 0), the field and the image force are smaller hence the barrier height is slightly larger than the barrier height at zero bias. For reverse bias (V < 0), the field and the image force are larger hence the barrier height is smaller than the barrier height at zero bias.

37

Figure 2-3: Energy-band diagram incorporating the Schottky effect for a n-type metal semiconductor contact under different biasing conditions (reproduced from [41]). The intrinsic barrier height is ΦBn0. The barrier height at thermal equilibrium is ΦBn. The barrier is lowered under forward and reverse biases by ΔΦF and ΔΦR respectively.

2.2.3 Current-voltage characteristics

The current-voltage characteristics of a Schottky diode are described by [41]:

푞Φ 푞푉 I = AA∗푇2 exp (− 퐵) 푒푥푝 [( ) − 1] Eq. 2-9 푘푇 푛푘푇 where I is the current, V is the voltage, A is the contact area, A* the Richardson constant, T is the temperature, k is Boltzmann’s constant, and n is the ideality factor. In non-ideal

Schottky diodes, the series resistance Rs can be accounted for using:

푞(푉 − 퐼푅 ) I = I 푒푥푝 [( 푠 ) − 1] Eq. 2-10 푠 푛푘푇

With Is being the saturation current:

38

푞Φ I = AA∗푇2 exp (− 퐵) Eq. 2-11 푠 푘푇

As a result, the barrier height and ideality factor of the Schottky diode can be estimated as:

푘푇 AA∗푇2 Φ퐵 = ln ( ) Eq. 2-12 푞 I푠

푞 푑푉 푛 = Eq. 2-13 푘푇 푑(푙푛퐼)

In reverse bias, if the barrier height is substantially smaller than the semiconductor bandgap, the reverse current is proportional to the reverse bias due mainly to image force lowering [3, 41]:

푞(Φ − √푞ξ /4휋휀 I = AA∗푇2 exp (− 퐵 푚 푠) Eq. 2-14 푅 푘푇

For practical Schottky diodes, the reverse current is typically dominated by leakage current due to the sharp edge around the periphery of the metal contact. A high field region develops at the metal-semiconductor-oxide interface giving rise to excess leakage currents and lowered breakdown voltages [43].

2.2.4 Current transportation

There are two basic mechanisms for charge transportation across a Schottky barrier: thermionic emission, in which electrons/holes are emitted over the barrier and tunneling

(field emission) in which electrons/holes quantum mechanically tunnel through the barrier.

The relative contributions of thermionic and field components depend on both temperature and doping level. A parameter Eoo is defined as:

39

푞ℏ 푁 Eq. 2-15 퐸표표 = √ ∗ 2 푚 휀푠

* Where N is density of states, m is the electron effective mass. When Eoo << kT, thermionic emission dominates and tunneling is negligible. When kT << Eoo, field emission (or tunneling) dominates. When kT = Eoo, thermionic-field emission is the main mechanism which is a combination of thermionic and field emission.

Figure 2-4: Energy band diagram showing three current transport mechanisms in n-type

Schottky contact: thermionic emission (TE), thermionic field emission (TFE) and field emission (FE) [3].

In a non-ideal Schottky contact, the existence of an insulating layer at the interface significantly alters the device characteristics and a modified theory is required to describe these characteristics.

40

2.2.4.1 Thermionic (Schottky) emission

The thermionic (Schottky) emission process, which was first described by

Richardson and by more advanced theory later on [44-47], concerns the transport of electrons with energies sufficient to overcome the metal- barrier or the insulator- semiconductor barrier as shown in Figure 2-6c. There were some assumptions that Bethe

[48] made in regard to the thermionic emission theory: the barrier height ΦB is much larger than kT, thermal equilibrium is established and the net current flow does not affect this equilibrium. According to Sze, the thermionic emission current is described by the following equation [41]:

푞Φ 푞푉 I = 퐴퐴∗푇2푒푥푝 (− 퐵) 푒푥푝 ( ) Eq. 2-16 푘푇 푘푇

4휋푞푚∗푘2 A∗ = Eq. 2-17 ℎ3

Considering image force lowering, thermionic Schottky emission is described by the following equation [49, 50]:

푞Φ − 훽 √퐸 I = 퐴퐴∗푇2푒푥푝 (− 퐵 푠 ) ~exp⁡(√푉) Eq. 2-18 푘푇

3 1/2 Where βs is the Schottky coefficient = (q /4πεoεr) and E is electric field.

The Richardson constant A* in Eq. 2-17 is derived theoretically, but is in practice often multiplied by a correction factor depending on the semiconductor material [51]. The work function, ΦB, must be modified in the presence of a surface electric field to account for the Schottky effect which describes how the electric field decreases the effective work function by the Schottky coefficient βs [52].

41

2.2.4.2 Direct tunneling

If the metal-semiconductor contact is sufficiently small, electrons in the semiconductor-side may transport across the barrier and move to the other side, even if the electron energy is much less than the barrier height. This phenomenon is called tunneling transportation. For heavily doped semiconductors, the barrier width becomes thinner hence electrons have higher probability to tunnel through the barrier. In metal-semiconductor

Ohmic contacts, tunneling is generally the dominant mechanism [53]. For direct tunneling, electrons tunnel through the barrier as in Figure 2-6a. The direct tunneling current is described by [54]:

3 4푑√2푚푒Φ퐵 퐼~푉2푒푥푝 − Eq. 2-19 3ℏ푞푉 ( ) where d is the barrier/insulator width and ℏ is Planck’s constant divided by 2π.

2.2.4.3 Space charge effect

The space charge effect in a semiconductor is determined by both ionized impurity concentrations (donor concentration ND and acceptor concentration NA) and the carrier concentrations (n and p). In the presence of space charge, the current is called the space charge limited current if it is dominated by the drift current due to the injected carriers. In

Figure 2-5a, the energy band diagram for the case of electron injection is presented. With the boundary condition V(x)=V at x=L, the formula of the space charge limited current can be obtained [3]:

9휀 휇푉2 퐽 = 푠 ~푉2 Eq. 2-20 8퐿3

42 where μ is electron mobility. As indicated in Eq. 2-20, the space charge limited current is proportional to the square of the applied voltage. This relationship is represented in Figure

2-5b.

Figure 2-5: Space charge effect. a) Band diagram for electron injection and b) J-V characteristic due to the space charge effect [3].

2.2.4.4 Fowler-Nordheim tunneling

In Fowler-Nordheim tunneling, the carrier only travels through a part of the barrier.

In this case, both the average potential barrier height and the tunneling distance are reduced from those of direct tunneling. The barrier is assumed to have a triangular shape as shown in Figure 2-6b and the current density is described by the following equation [55-60]:

4 √2푚(푒Φ )3/2 퐽 = 푐표푛푠푡. {퐸2푒푥푝 [− 퐵 ]} Eq. 2-21 3 ℏ푒퐸

Fowler and Nordheim originally ignored the effects of image charge on the shape and width of the surface barrier, thereby underestimating the current density to a great

43 degree, but Nordheim later included the Schottky barrier into the field emission model [58,

59]. Forbes investigated the properties of Fowler–Nordheim type equations, generalizing it for barrier shape and correcting for effects of temperature, electron band structure and occupation states [60].

2.2.4.5 Frenkel-Poole emission

In Frenkel-Poole (or Poole-Frenkel) emission, electrons are trapped in localized states. By thermal excitation, electrons receive enough energy to escape their localized states and move to the conduction band. The emission process is similar to that in the

Schottky emission but now the barrier height is the depth of the trap potential well instead of the difference between the metal and semiconductor’s Fermi level. The band diagram describing this transport mechanism is shown in Figure 2-6d. The current density of

Frenkel-Poole emission follows [41, 61, 62]:

−푞(Φ − √푞퐸/(휋휀)) 퐽⁡~푒푥푝 ( 퐵 ) Eq. 2-22 푘푇

44

Figure 2-6: Energy band diagram showing transport mechanisms of (a) direct tunneling,

(b) Fowler-Nordheim tunneling, (c) thermionic emission and (d) Poole-Frenkle emission

[3].

2.2.5 Schottky barrier inhomogeneity

Non-ideal Schottky contacts are likely to contain inhomogeneities that result in dissimilarity in barrier height. Inhomogeneities have been described by numerous reports with a broad range of contact materials [14, 20, 24, 29]. Henisch [63] suggested barrier disparities occur even in the most rigorously fabricated Schottky contacts. According to

Tung, Werner and Guttler [12, 14, 15], there are many contributors to barrier inhomogeneities including structural defects and contamination at the semiconductor interface, intrinsic defect clusters and extrinsic surface defects.

As an effect of barrier inhomogeneity, the discrepancy between barrier height derived from I-V and C-V data is observed [11, 17, 28, 29]. This can be explained by the assumption that the current in the I-V measurement is dominated by the current which flows through the regions of low Schottky barrier height. Since the low Schottky barrier 45 height patch is pinched-off, the effective Schottky barrier height of the patch is the potential at the saddle point, which has a potential with respect to the metal Fermi level. As a result, the I-V barrier height through the low patch is significantly lower than C-V barrier height

[14, 15].

Another observed effect is the dependency of barrier height and ideality factor on temperature. Shiyang et al [64] reported on inhomogeneity in Schottky contacts consisting of CoSi2 on Si. The degradation of barrier height and ideality factor at lower temperature can be observed in Figure 2-7. Similar temperature dependency of barrier height and ideality factor was also observed in Ti on 4H-SiC Schottky contacts (Figure 2-8) [19]. For

Carbon based Schottky diodes, this effect has also appeared in the literature. For example, graphene on MoS2 Schottky contacts also exhibit temperature-dependent barrier heights and ideality factors as in Figure 2-9 [26].

Figure 2-7: Schottky barrier heights (SBHs) and ideality factors derived from the fitting as a function of temperature for CoSi2 contacts on (a) Si (100) and (b) Si (111) substrates.

The modified SBH data and their linear fits are also shown [64].

46

Figure 2-8: Ideality factors and SBHs as a function of the

absolute temperature for Ti/4H–SiC contacts [19].

Figure 2-9: Effective barrier heights and ideality factors as a function of temperature for

a graphene–MoS2 contact [26].

47

2.3 Carbon-semiconductor Schottky diodes

Stelzer and Kreupl demonstrated that Carbon on Silicon (C-Si) contacts have almost the same electrical properties compared to the well-established TiSi on Si contacts but show a much improved reliability and temperature stability [65, 66]. The structure of the graphene on Si Schottky diode from Stelzer and Kreupl’s work is shown in Figure

2-10. This Schottky diode was fabricated with graphene deposited by chemical vapor deposition (CVD), described in [67, 68], but otherwise identical to a commercial BAT15 diode structure from Infineon. The diode was then subjected to a series of 3.5 MA/cm2 current pulses with 100ns width.

Figure 2-10: Graphene on Silicon Schottky diode by Stelzer and Kreupl [65, 66].

The graphene on Si diode endured up to 500 million pulses before the first significant increase in reverse current was observed, as shown in Figure 2-11. The conventional TiSi-based diode was damaged and hence ‘short-circuited’ after 2 – 4 pulses

48

[65, 66]. The degradation in the graphene on Si diode was observed after 545 million pulses and was attributed to dopant diffusion from the substrate or to the diffusion of the top metallization.

Figure 2-11: Current voltage characteristic of graphene on a silicon Schottky diode after various series of current pulses by Stelzer and Kreupl [65, 66]. After up to 500 million pulses the electrical behavior is almost unaffected. Only a small shift of the SBH of 0.01 eV is observable, leading to a lower reverse current. After 545 million pulses, the reverse leakage current increases.

Stelzer and Kreupl stressed the device further (with up to 500ns pulse width current) until failure and under certain conditions, the top Ti/Cu/Au layer melted. The scanning electron microscope (SEM) image in Figure 2-12 illustrates how the top metallization layer was physically damaged after the series of 3.5 MA/cm2 current pulses were applied.

Using 100 ns current pulses, the graphene on silicon diode endured 548 million pulses

49 before failure with little damage visible (Figure 2-12a). In Figure 2-12b using 300ns current pulses and Figure 2-12c using 500ns current pulses, the top Ti/Cu/Au layers melted and this is clearly visible. The melting point of Cu is approximately 1360K, and the temperature at the graphene/Si interface is much higher [65, 66]. In the COMSOL simulation by Stelzer and Kreupl, the temperature of the graphene/Si interface was estimated to peak at 1700 K. Figure 2-12d reveals all devices had failed, showing short- circuit electrical behavior. The important point revealed in the microscopy and electrical measurements is that the carbon layer endured greater temperatures than the surrounding metallic layers.

Figure 2-12: SEM image of the top view of C-Si Schottky diodes with Ti/Cu/Au top metallization after the diodes failed. (a) shows a device after 548 million pulses (100ns)

50 with no obvious visible damage in the top metallization. The device in (b) demonstrates the damage of a 300 ns pulse after 1.5 million pulses where the metal started to melt. (c) illustrates a sample where the molten copper was spit on the periphery and a circular crater was formed after a series of 500 ns pulses. (d) shows J-V characteristics of graphene on Si diodes after they reached the threshold for a failed diode [65, 66].

Carbon nanotubes (CNT) are another form of carbon that have been used intensively as a promising material to reduce short-channel effects, minimize leakage currents and control process parameter variation in high performance electronics [69-72].

CNTs have been applied in the fabrication of field-effect transistors (FET) to create

Schottky barriers at their gates [73-76]. Behnam [77] used single-walled CNTs on GaAs as a back to back device (Figure 2-13) and characterized the device from 230K to 340K

(Figure 2-14). As indicated in Figure 2-13, in order to study the effect of device geometry on the dark current, metal-semiconductor-metal devices with different active area widths

FW, finger lengths FL, finger widths W, and finger spacing’s S were fabricated.

51

Figure 2-13: (a–d) Schematic of the process flow for metal-semiconductor-metal photodetectors with CNT film electrodes along the dashed line AB shown in (e): (a) SiN isolation layer deposited on a GaAs substrate, (b) CNT film prepared by vacuum filtration deposited on the substrate after opening the active windows in the SiN layer, (c) CNT film patterned into interdigitated electrode fingers by photolithography and inductively coupled plasma etching, and (d) Cr/Pd metal contacts patterned on the nanotube film contact pads using photolithography, e-beam evaporation, and subsequent lift-off. (e) Optical microscope image of the finished metal-semiconductor-metal photodetector, showing the various device dimensions. (f) Atomic force microscope (AFM) image showing the area between two CNT film electrode fingers of the metal-semiconductor-metal device of part

(e) [77].

52

In Figure 2-14a, the Schottky characteristic of a CNT on GaAs contact is shown.

The data exhibits the characteristic I-V curves of the back-to-back Schottky diodes making up the metal-semiconductor-metal photodetector. Figure 2-14b shows the Richardson plot for the temperature range 280K to 340K with linear dashed lines showing the best fit to the experimental data. From the I-V data with thermionic mechanism, the SBHs of the CNT on GaAs contacts were estimated to be 0.53, 0.54, and 0.54 eV for devices with W = S =

15, 20, and 30 μm, respectively. The work function of the CNT films was calculated to be

4.6eV, in line with previously reported values [78, 79]. In the inset of Figure 2-14b, the current starts to saturate at temperatures lower than 260 K, which suggests that tunneling, which depends weakly on temperature, begins to dominate the transport across the CNT film–GaAs junction at lower temperatures [77].

Figure 2-14: (a) Dark current (log scale) vs applied voltage measured at six different temperatures between 230 and 340 K for a CNT film–GaAs metal-semiconductor-metal device with W = S = 15 μm and FL = FW = 300 μm. The inset shows the dark I-V characteristics of this device at room temperature 294K in linear scale. (b) The Richardson plot of log I/T2 vs 1/T at V= 3 V voltage bias in the temperature range of 280–340 K for

53 three CNT film–GaAs metal-semiconductor-metal devices of identical active area (FL =

FW = 300 μm) but different finger widths W and spacings S, as labeled in the figure. The inset shows log current vs 1/T for the metal-semiconductor-metal device with W=S= 20 um in a wider temperature window (from 150 to 340K) at V= 3 V bias [77].

Graphitic carbon like graphene, CNTs or ‘graphenic’ carbon is chemically inert and stable at extreme temperatures, making it a suitable material for electronic device contacts

[25, 66, 80-84]. High quality graphitic device contacts have been demonstrated on a variety of technologically important semiconducting materials including Si, GaAs and GaN [80], graphene [81, 82] and graphitic carbon [66, 80]. C formed on Si diodes have exhibited excellent chemical- and thermal- stability with ideality factors approaching unity [66, 80].

Graphitic device contacts have been formed in numerous ways, including by application of colloidal graphite [80], high temperature chemical vapour deposition [66,

85], pulsed laser deposition [86, 87] and by pyrolizing photoresist [85]. Whilst effective in demonstrating the excellent properties and potential applications of these devices, these methods require highly elevated substrate temperatures and/or slow exfoliation/application processes and to some extent limit applicability. It has been previously shown that energetic physical vapour deposition methods, in which C is deposited from a highly ionized flux, enable accurate control over the sp2 (graphitic) fraction in C films. This can be achieved using moderate (< 100 °C) growth temperatures that are compatible with photolithography [88-90]. The application of a direct or remote bias (DC or pulsed) enables the energy of the arriving flux to be controlled and consequently, synthesis of C films with prescribed sp2/sp3 (graphite-like/diamond-like) bonding fractions, microstructure and

54 electrical properties [84, 89, 90]. Using this approach, highly rectifying graphitic Schottky contacts to p-type Si were fabricated and tested [91].

Tongay et al [80] performed a comparison of highly oriented pyrolytic graphite

(HOPG) contacts on Si, GaAs, and 4H-SiC substrates. A HOPG contact was fabricated by either “soft-placing” a sheet of HOPG or painting a graphite powder blended in residue- free 2-butoxyethyl acetate and octyl acetate onto the substrate. Graphite formed a good rectification barrier at room temperature for all three substrates as shown in Figure 2-15.

In these devices, electron transport over the barrier height is dominated by thermionic emission as the semilogarithmic J-V curves usually display a sufficiently linear portion in the forward-bias region. Tongay [80] reported ideality factors ranging from 1.25 to 2.0 and barrier heights of 0.60eV, 0.78eV and 1.60eV for Si, GaAs and 4H-SiC respectively; with the high value being attributed to bias dependent barrier heights, generation-recombination, thermally assisted tunneling, and image force lowering. These claims are also similar to

Tung’s work on Schottky barrier theory [4, 92]. From the extracted barrier height value,

Tongay arrived with the work function for graphite falling in the range of 4.6 to 4.8eV, which compares well with some other reported graphite work functions [93-95].

Yatskiv [96] demonstrated graphite on Zinc oxide (ZnO) Schottky diodes with ZnO grown by the hydrothermal method. The current-voltage characteristic of a graphite on

ZnO diode is shown in Figure 2-16. From J-V data obtained at different temperatures from

300 K to 420 K, Yatskiv calculated the barrier heights and ideality factors to be around 0.9 eV and 1.1 respectively. In this case, the barrier heights and ideality factors remained stable in the temperature range 300 K to 420 K. As current transport is dominated by thermionic emission, the extracted work function for graphite following the Schottky-Mott relation

55

(described in Eq. 2-1) was determined to be 4.95 eV, which is within the range of values reported in the literature [94, 97].

Figure 2-15: J-V characteristic of HOPG contact at room temperature on (a) n-type

Si/graphite (red squares), (b) n-type GaAs/graphite (blue circles), and (c) n type 4H-

SiC/graphite junctions (black triangles). Insets: J-V plots on semi-logarithmic axes [80].

56

Figure 2-16: J-V characteristic of the graphite/ZnO(O-face) Schottky diode over the

temperature range of 300–420 K [96].

2.4 Summary

In Chapter 2, the structure and underlying current transport mechanisms of

Schottky diodes have been presented. Recently, graphitic forms of carbon have been employed as Schottky contacts to semiconductors and the chapter also contained a review of work in this field. Due to its extreme stability, graphite has the potential to become a widely used contact material for high power devices and/or devices operating under harsh conditions. The remainder of this thesis describes the experimental methods and the results obtained after fabrication and testing of graphitic contacts to Si, 6H-SiC and 훽-Ga2O3.

These results not only support the assertion made by numerous authors that carbon materials are a key enabler for next generation electronics but also address some obstacles

57 as well as suggest promising development of these novel materials in the high-power sector.

2.5 References

[1] W. Schottky, "Semi-conductor theory in barrier layers," Naturwissenschaften, vol. 26, no. 843, p. 280, 1938. [2] N. Mott, "Note on the contact between a metal and an insulator or semi- conductor," in Mathematical Proceedings of the Cambridge Philosophical Society, 1938, vol. 34, no. 4, pp. 568-572: Cambridge University Press. [3] S. M. Sze, Semiconductor devices: physics and technology. John Wiley & Sons, 2008. [4] R. T. Tung, "Recent advances in Schottky barrier concepts," Materials Science and Engineering: R: Reports, vol. 35, no. 1-3, pp. 1-138, 2001. [5] L. Brillson, "The structure and properties of metal-semiconductor interfaces," Surface Science Reports, vol. 2, no. 2, pp. 123-326, 1982. [6] A. Cowley and S. Sze, "Surface states and barrier height of metal‐semiconductor systems," Journal of Applied Physics, vol. 36, no. 10, pp. 3212-3220, 1965. [7] J. Freeouf and J. Woodall, "Schottky barriers: An effective work function model," Applied Physics Letters, vol. 39, no. 9, pp. 727-729, 1981. [8] A. Itoh and H. Matsunami, "Analysis of Schottky barrier heights of metal/SiC contacts and its possible application to high‐voltage rectifying devices," physica status solidi (a), vol. 162, no. 1, pp. 389-408, 1997. [9] F. Léonard and J. Tersoff, "Role of Fermi-level pinning in nanotube Schottky diodes," Physical Review Letters, vol. 84, no. 20, p. 4693, 2000. [10] T. Nishimura, K. Kita, and A. Toriumi, "Evidence for strong Fermi-level pinning due to metal-induced gap states at metal/germanium interface," Applied Physics Letters, vol. 91, no. 12, p. 123123, 2007. [11] A. B. Ulusan and A. Tataroglu, "Analysis of barrier inhomogeneities in AuGe/n- Ge Schottky diode," Indian Journal of Physics, vol. 92, no. 11, pp. 1397-1402, 2018. [12] J. H. Werner and H. H. Güttler, "Barrier inhomogeneities at Schottky contacts," Journal of applied physics, vol. 69, no. 3, pp. 1522-1533, 1991. [13] R. Tung, "Schottky barrier height—do we really understand what we measure?," Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena, vol. 11, no. 4, pp. 1546- 1552, 1993. [14] J. P. Sullivan, R. T. Tung, M. R. Pinto, and W. R. Graham, "Electron transport of inhomogeneous Schottky barriers: A numerical study," Journal of Applied Physics, vol. 70, no. 12, pp. 7403-7424, 1991/12/15 1991. 58

[15] R. T. Tung, A. F. J. Levi, J. P. Sullivan, and F. Schrey, "Schottky-barrier inhomogeneity at epitaxial NiS2 interfaces on Si(100)," Physical Review Letters, vol. 66, no. 1, pp. 72-75, 01/07/ 1991. [16] M. Bhatnagar, B. J. Baliga, H. R. Kirk, and G. A. Rozgonyi, "Effect of surface inhomogeneities on the electrical characteristics of SiC Schottky contacts," IEEE Transactions on Electron Devices, vol. 43, no. 1, pp. 150-156, 1996. [17] S. Chand and J. Kumar, "Evidence for the double distribution of barrier heights in Pd2Si/n-Si Schottky diodes from I-V-T measurements," Semiconductor Science and Technology, vol. 11, no. 8, p. 1203, 1996. [18] W. Mönch, "Barrier heights of real Schottky contacts explained by metal-induced gap states and lateral inhomogeneities," Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena, vol. 17, no. 4, pp. 1867-1876, 1999. [19] F. Roccaforte, F. L. Via, V. Raineri, R. Pierobon, and E. Zanoni, "Richardson’s constant in inhomogeneous silicon carbide Schottky contacts," Journal of Applied Physics, vol. 93, no. 11, pp. 9137-9144, 2003. [20] A. F. Hamida, Z. Ouennoughi, A. Sellai, R. Weiss, and H. Ryssel, "Barrier inhomogeneities of tungsten Schottky diodes on 4H-SiC," Semiconductor Science and Technology, vol. 23, no. 4, p. 045005, 2008. [21] H. Kim, A. Sohn, and D.-W. Kim, "Silver Schottky contacts to Zn-polar and O- polar bulk ZnO grown by pressurized melt-growth method," Semiconductor Science and Technology, vol. 27, no. 3, p. 035010, 2012. [22] S. Tongay, T. Schumann, and A. Hebard, "Graphite based Schottky diodes formed on Si, GaAs, and 4H-SiC substrates," Applied Physics Letters, vol. 95, no. 22, p. 222103, 2009. [23] A. Hattab, J. Perrossier, F. Meyer, M. Barthula, H. Osten, and J. Griesche, "Schottky barrier inhomogeneities at contacts to carbon-containing silicon/germanium alloys," Materials Science and Engineering: B, vol. 89, no. 1- 3, pp. 284-287, 2002. [24] Y.-J. Lin and J.-H. Lin, "Annealing effect on Schottky barrier inhomogeneity of graphene/n-type Si Schottky diodes," Applied Surface Science, vol. 311, pp. 224- 229, 2014. [25] D. Sinha and J. U. Lee, "Ideal graphene/silicon Schottky junction diodes," Nano letters, vol. 14, no. 8, pp. 4660-4664, 2014. [26] Y.-F. Lin et al., "Barrier inhomogeneities at vertically stacked graphene-based heterostructures," Nanoscale, vol. 6, no. 2, pp. 795-799, 2014. [27] S. Shivaraman, L. H. Herman, F. Rana, J. Park, and M. G. Spencer, "Schottky barrier inhomogeneities at the interface of few layer epitaxial graphene and silicon carbide," Applied Physics Letters, vol. 100, no. 18, p. 183112, 2012. [28] D. Defives et al., "Barrier inhomogeneities and electrical characteristics of Ti/4H- SiC Schottky rectifiers," IEEE Transactions on Electron Devices, vol. 46, no. 3, pp. 449-455, 1999.

59

[29] L. Boussouar, Z. Ouennoughi, N. Rouag, A. Sellai, R. Weiss, and H. Ryssel, "Investigation of barrier inhomogeneities in Mo/4H–SiC Schottky diodes," Microelectronic Engineering, vol. 88, no. 6, pp. 969-975, 2011. [30] F. Himpsel, G. Hollinger, and R. Pollak, "Determination of the Fermi-level pinning position at Si (111) surfaces," Physical Review B, vol. 28, no. 12, p. 7014, 1983. [31] C. Hobbs et al., "Fermi level pinning at the polySi/metal oxide interface," in 2003 Symposium on VLSI Technology. Digest of Technical Papers (IEEE Cat. No. 03CH37407), 2003, pp. 9-10: IEEE. [32] C. C. Hobbs et al., "Fermi-level pinning at the polysilicon/metal-oxide interface- Part II," IEEE Transactions on Electron Devices, vol. 51, no. 6, pp. 978-984, 2004. [33] A. Dimoulas, P. Tsipas, A. Sotiropoulos, and E. Evangelou, "Fermi-level pinning and charge neutrality level in germanium," Applied physics letters, vol. 89, no. 25, p. 252110, 2006. [34] J. Woodall, G. Pettit, T. Jackson, C. Lanza, K. Kavanagh, and J. Mayer, "Fermi- level pinning by misfit dislocations at GaAs interfaces," Physical review letters, vol. 51, no. 19, p. 1783, 1983. [35] M. Pashley, K. Haberern, R. Feenstra, and P. Kirchner, "Different Fermi-level pinning behavior on n-and p-type GaAs (001)," Physical Review B, vol. 48, no. 7, p. 4612, 1993. [36] R. Lieten, S. Degroote, M. Kuijk, and G. Borghs, "Ohmic contact formation on n- type Ge," Applied Physics Letters, vol. 92, no. 2, p. 022106, 2008. [37] N. A. Ives, G. W. Stupian, and M. S. Leung, "Unpinning of the Fermi level on GaAs by flowing water," Applied physics letters, vol. 50, no. 5, pp. 256-258, 1987. [38] S. Tanaka, J. A. Bruce, and M. S. Wrighton, "Deliberate modification of the behavior of n-type cadmium telluride/electrolyte interfaces by surface etching. Removal of Fermi level pinning," The Journal of Physical Chemistry, vol. 85, no. 25, pp. 3778-3787, 1981. [39] V. Heine, "Theory of surface states," Physical Review, vol. 138, no. 6A, p. A1689, 1965. [40] J. Tersoff, "Schottky barrier heights and the continuum of gap states," Physical Review Letters, vol. 52, no. 6, p. 465, 1984. [41] S. M. Sze and K. K. Ng, Physics of semiconductor devices, 3rd ed. John Wiley & Sons, 2006. [42] J. Robertson and B. Falabretti, "Band offsets of high K gate oxides on III-V semiconductors," Journal of applied physics, vol. 100, no. 1, p. 014111, 2006. [43] J. Saltich and L. Clark, "Use of a double diffused guard ring to obtain near ideal I–V characteristics in Schottky barrier diodes," Solid-State Electronics, vol. 13, no. 6, pp. 857-863, 1970. [44] O. W. Richardson, On the Negative Radiation from Hot Platinum. University Press, 1901.

60

[45] O. Richardson, "Electron emission from metals as a function of temperature," Physical Review, vol. 23, no. 2, p. 153, 1924. [46] S. Dushman, "Thermionic emission," Reviews of Modern Physics, vol. 2, no. 4, p. 381, 1930. [47] R. Fowler, "RH Fowler, Proc. R. Soc. London, Ser. A 159, 229 (1937)," Proc. R. Soc. London, Ser. A, vol. 159, p. 229, 1937. [48] H. A. Bethe, "Theory of the boundary layer of crystal rectifiers," in Semiconductor Devices: Pioneering Papers: World Scientific, 1991, pp. 387-399. [49] W. Schottky, "Über kalte und warme Elektronenentladungen," Zeitschrift Für Physik A Hadrons and Nuclei, vol. 14, no. 1, pp. 63-106, 1923. [50] W. Schottky, "Vereinfachte und erweiterte Theorie der Randschicht- gleichrichter," Zeitschrift für Physik A Hadrons and Nuclei, vol. 118, no. 9, pp. 539-592, 1942. [51] P. Zhang, Á. Valfells, L. Ang, J. Luginsland, and Y. Lau, "100 years of the physics of diodes," Applied Physics Reviews, vol. 4, no. 1, p. 011304, 2017. [52] W. Schottky, "Über den Einfluss von Strukturwirkungen, besonders der Thomsonschen Bildkraft, auf die Elektronenemission der Metalle," Physik. Zeitschr, vol. 15, pp. 872-878, 1914. [53] R. Clerc, A. Spinelli, G. Ghibaudo, and G. Pananakakis, "Theory of direct tunneling current in metal–oxide–semiconductor structures," Journal of applied physics, vol. 91, no. 3, pp. 1400-1409, 2002. [54] J. M. Beebe, B. Kim, J. W. Gadzuk, C. D. Frisbie, and J. G. Kushmerick, "Transition from direct tunneling to field emission in metal-molecule-metal junctions," Physical review letters, vol. 97, no. 2, p. 026801, 2006. [55] L. Nordheim, "Die theorie der elektronenemission der metalle," Physikalische Zeitschrift, vol. 30, pp. 117-196, 1929. [56] I. D. Parker, "Carrier tunneling and device characteristics in polymer light‐ emitting diodes," Journal of Applied Physics, vol. 75, no. 3, pp. 1656-1666, 1994. [57] R. Blum, M. Sprave, J. Sablotny, and M. Eich, "High-electric-field poling of nonlinear optical polymers," JOSA B, vol. 15, no. 1, pp. 318-328, 1998. [58] R. Ellis, "Fowler-Nordheim emission from non-planar surfaces," IEEE Electron Device Letters, vol. 3, no. 11, pp. 330-332, 1982. [59] L. Nordheim, "LW Nordheim, Proc. R. Soc. London Ser. A 121, 626 (1928)," Proc. R. Soc. London Ser. A, vol. 121, p. 626, 1928. [60] R. G. Forbes, "Physics of generalized Fowler-Nordheim-type equations," Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena, vol. 26, no. 2, pp. 788-793, 2008. [61] J. Frenkel, "On pre-breakdown phenomena in insulators and electronic semi- conductors," Physical Review, vol. 54, no. 8, p. 647, 1938. [62] P. Rottländer, M. Hehn, and A. Schuhl, "Determining the interfacial barrier height and its relation to tunnel magnetoresistance," Physical Review B, vol. 65, no. 5, p. 054422, 2002.

61

[63] H. Henisch, S. Rahimi, Y. Moreau, and L. Szepessy, "Schottky barrier ideality, real and imagined," Solid State Electronics, vol. 27, pp. 1033-1034, 1984. [64] S. Zhu et al., "Barrier height inhomogeneities of epitaxial CoSi2 Schottky contacts on n-Si (100) and (111)," Solid-State Electronics, vol. 44, no. 4, pp. 663- 671, 2000. [65] M. Stelzer and F. Kreupl, "Graphenic carbon-silicon contacts for reliability improvement of metal-silicon junctions," in 2016 IEEE International Electron Devices Meeting (IEDM), 2016, pp. 21.7.1-21.7.4. [66] M. Stelzer, M. Jung, and F. Kreupl, "Graphenic Carbon: A Novel Material to Improve the Reliability of Metal-Silicon Contacts," IEEE Journal of the Electron Devices Society, vol. 5, no. 5, pp. 416-425, 2017. [67] S. Huebner, N. Miyakawa, A. Pahlke, and F. Kreupl, "Performance improvement of graphenic carbon X-ray transmission windows," MRS Advances, vol. 1, no. 20, pp. 1441-1446, 2016. [68] F. Kreupl, "Carbon-based materials as key-enabler for “more than moore”," MRS Online Proceedings Library Archive, vol. 1303, 2011. [69] C. Chen, Y. Lu, E. S. Kong, Y. Zhang, and S. T. Lee, "Nanowelded carbon‐ nanotube‐based solar microcells," Small, vol. 4, no. 9, pp. 1313-1318, 2008. [70] Y. Jia et al., "Achieving high efficiency silicon-carbon nanotube solar cells by acid doping," Nano letters, vol. 11, no. 5, pp. 1901-1905, 2011. [71] A. Raychowdhury and K. Roy, "Carbon nanotube electronics: design of high- performance and low-power digital circuits," IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 54, no. 11, pp. 2391-2401, 2007. [72] J. Guo, S. Datta, and M. Lundstrom, "A numerical study of scaling issues for Schottky-barrier carbon nanotube transistors," IEEE transactions on electron devices, vol. 51, no. 2, pp. 172-177, 2004. [73] P. Avouris, M. Freitag, and V. Perebeinos, "Carbon-nanotube photonics and optoelectronics," Nature photonics, vol. 2, no. 6, p. 341, 2008. [74] S. Heinze, J. Tersoff, R. Martel, V. Derycke, J. Appenzeller, and P. Avouris, "Carbon nanotubes as Schottky barrier transistors," Physical Review Letters, vol. 89, no. 10, p. 106801, 2002. [75] M. Yang, K. Teo, W. Milne, and D. Hasko, "Carbon nanotube Schottky diode and directionally dependent field-effect transistor using asymmetrical contacts," Applied Physics Letters, vol. 87, no. 25, p. 253116, 2005. [76] H. M. Manohara, E. W. Wong, E. Schlecht, B. D. Hunt, and P. H. Siegel, "Carbon nanotube Schottky diodes using Ti− Schottky and Pt− Ohmic contacts for high frequency applications," Nano letters, vol. 5, no. 7, pp. 1469-1474, 2005. [77] A. Behnam et al., "Metal-semiconductor-metal photodetectors based on single- walled carbon nanotube film–GaAs Schottky contacts," Journal of Applied Physics, vol. 103, no. 11, p. 114315, 2008. [78] V. Barone, J. E. Peralta, J. Uddin, and G. E. Scuseria, "Screened exchange hybrid density-functional study of the work function of pristine and doped single-walled

62

carbon nanotubes," The Journal of chemical physics, vol. 124, no. 2, p. 024709, 2006. [79] B. Shan and K. Cho, "First principles study of work functions of single wall carbon nanotubes," Physical review letters, vol. 94, no. 23, p. 236602, 2005. [80] S. Tongay, T. Schumann, and A. F. Hebard, "Graphite based Schottky diodes formed on Si, GaAs, and 4H-SiC substrates," Applied Physics Letters, vol. 95, no. 22, p. 222103, 2009/11/30 2009. [81] C.-C. Chen, M. Aykol, C.-C. Chang, A. Levi, and S. B. Cronin, "Graphene-silicon Schottky diodes," Nano letters, vol. 11, no. 5, pp. 1863-1867, 2011. [82] Y. Song et al., "Role of interfacial oxide in high-efficiency graphene–silicon Schottky barrier solar cells," Nano letters, vol. 15, no. 3, pp. 2104-2110, 2015. [83] M. Kracica, E. L. H. Mayes, H. N. Tran, A. S. Holland, D. G. McCulloch, and J. G. Partridge, "Rectifying electrical contacts to n-type 6H–SiC formed from energetically deposited carbon," Carbon, vol. 102, pp. 141-144, 2016/06/01/ 2016. [84] T. J. Raeber, Z. C. Zhao, B. J. Murdoch, D. R. McKenzie, D. G. McCulloch, and J. G. Partridge, "Resistive switching and transport characteristics of an all-carbon ," Carbon, vol. 136, pp. 280-285, 2018. [85] C. Yim, E. Rezvani, S. Kumar, N. McEvoy, and G. S. Duesberg, "Investigation of carbon–silicon Schottky barrier diodes," Physica Status Solidi B: Basic Research, vol. 249, no. 12, pp. 2553-2557, 2012. [86] R. K. Gupta, K. Ghosh, and P. K. Kahol, "Current–voltage characteristics of p- Si/carbon junctions fabricated by pulsed laser deposition," Microelectronic Engineering, vol. 87, no. 2, pp. 221-224, 2010. [87] R. A. Ismail, W. K. Hamoudi, and K. K. Saleh, "Effect of rapid thermal annealing on the characteristics of amorphous carbon/n-type crystalline silicon heterojunction solar cells," Materials Science in Semiconductor Processing, vol. 21, pp. 194-199, 2014. [88] D. W. M. Lau et al., "Abrupt Stress Induced Transformation in Amorphous Carbon Films with a Highly Conductive Transition Phase," Physical Review Letters, vol. 100, no. 17, p. 176101, 04/28/ 2008. [89] D. W. M. Lau et al., "The structural phases of non-crystalline carbon prepared by physical vapour deposition," Carbon, vol. 47, no. 14, pp. 3263-3270, 2009. [90] D. W. M. Lau et al., "Microstructural investigation supporting an abrupt stress induced transformation in amorphous carbon films," Journal of Applied Physics, vol. 105, no. 8, p. 084302, 2009/04/15 2009. [91] M. S. N. Alnassar et al., "Graphitic Schottky contacts to Si formed by energetic deposition," MRS Online Proceedings Library Archive, vol. 1786, pp. 51-56, 2015. [92] R. T. Tung, "The physics and chemistry of the Schottky barrier height," Applied Physics Reviews, vol. 1, no. 1, p. 011304, 2014. [93] E. Taft and L. Apker, "Photoelectric emission from polycrystalline graphite," Physical Review, vol. 99, no. 6, p. 1831, 1955.

63

[94] S. J. Sque, R. Jones, and P. R. Briddon, "The transfer doping of graphite and graphene," physica status solidi (a), vol. 204, no. 9, pp. 3078-3084, 2007. [95] S. Suzuki, "S. Suzuki, C. Bower, T. Kiyokura, KG Nath, Y. Watanabe, and O. Zhou, J. Electron Spectrosc. Relat. Phenom. 114, 225 (2001)," J. Electron Spectrosc. Relat. Phenom., vol. 114, p. 225, 2001. [96] R. Yatskiv and J. Grym, "Temperature-dependent properties of semimetal graphite-ZnO Schottky diodes," Applied Physics Letters, vol. 101, no. 16, p. 162106, 2012. [97] N. Ooi, A. Rairkar, and J. B. Adams, "Density functional study of graphite bulk and surface properties," Carbon, vol. 44, no. 2, pp. 231-242, 2006.

64

CHAPTER 3

FABRICATION & CHARACTERISATION METHODS

This chapter describes the experimental work carried out during this project including device fabrication, materials analysis, microscopy and electrical characterization. The devices were patterned using optical lithography and the layers were then deposited by physical vapor deposition. Energetic deposition employed for the graphitic carbon layers, is described in detail later in this chapter. Device measurements including current-voltage (I-V), capacitance-voltage (C-V) and Hall measurements were used to examine the performance of the devices and to enable extraction of material parameters from the devices. Scanning electron microscopy (SEM), Raman spectroscopy and X-ray photoelectron spectroscopy (XPS) were all essential characterization methods required to support the device measurements and an overview of these is also provided.

3.1 Schottky diode fabrication

3.1.1 Substrate dicing

Prior to device fabrication and materials analysis, the wafers (wafers’ specs are listed in Appendix A) were cleaved or diced to produce substrates approximately 10 × 10 mm2 in area. Materials were coated with photoresist before being diced/cleaved. The wafers were mounted in a Disco Dac 321 automatic dicing machine (as shown in Figure

65

3-1). Once diced or cleaved, the substrates were immersed in acetone to remove the protective photoresist coating and were then prepared for cleaning.

Figure 3-1: Dicing machine.

3.1.2 Wafer cleaning

The standard RCA cleaning procedure was applied to remove organic contaminants, oxide layers and ionic contamination from the substrates [1, 2]. They were then rinsed in acetone for 5 minutes to remove organic impurities from both the polished and unpolished surfaces. Isopropanol (IPA) was used to rinse the substrates to dissolve any excess acetone on their surfaces. Finally, they were cleaned in de-ionised (DI) water and dried using a dry nitrogen gun.

66

3.1.3 Spin coating

Spin coating is widely used in semiconductor fabrication to produce uniform thin films on substrates from liquid materials applied during the spinning cycle [3-6]. For photolithography, a layer of photoresist is formed with a thickness that is dependent on its weight/viscosity and the rotational frequency [7, 8]. A Laurell WS-650MZ spin coater, as shown in Figure 3-2 was used for this procedure in this project.

Figure 3-2: Laurell WS-650MZ spin coater.

The substrate was held on a spinning vacuum chuck inside the coater. The photoresist was applied from a micro-pipette onto the spinning substrate through an aperture in the lid of the coater. AZ1512 positive photoresist was selected for optical lithography. When exposed to light with wavelengths ranging from 310nm to 440nm, the

67 resist becomes soluble in an accompanying developer [9, 10]. The AZ1512 coated samples were baked for 1 minute at 95oC on a hot-plate prior to UV exposure.

3.1.4 Mask alignment and UV exposure

The desired device patterns were translated from a Cr-on-glass mask (shown in

Figure 3-3) to the substrates using a Karl Suss MJB-3 mask aligner, shown in Figure 3-4.

The UV exposure system uses a 350 W high pressure mercury lamp. Stage movement relative to the mask is controlled by x, y and theta aligning verniers. Exposure time was typically 30 seconds. After UV exposure was completed, AZ400 developer (diluted in DI water 1:4) was used to develop the patterns in the photoresist.

Figure 3-3: Cr on glass masks.

68

Figure 3-4: Karl Suss MJB-3 mask aligner.

The patterns used for the Schottky anodes were arrays of 40 µm, 80 µm and 160

µm diameter apertures, shown in Figure 3-5.

Figure 3-5: Circular test device structure.

69

3.1.5 Energetic carbon deposition using a filtered cathodic vacuum arc (FCVA)

In a filtered cathodic vacuum arc (FCVA), a high current arc discharge is initiated on a cathode source to form a high density plasma. The depositing flux consists mainly of ions so the energy of the depositing species can be tuned using either a substrate bias or a biased grid placed in front of the substrate within the plasma stream [11]. Energetic deposition is particularly beneficial when depositing carbon. It has been found that even at moderate growth temperatures, the structure and properties of carbon films can be varied over a wide range (from ~ 90% graphitic bonding to ~ 80% diamond-like bonding) by controlling the energy of the depositing flux [12, 13]. Energetic methods which have been used to produce highly activated carbon plasmas include filtered cathodic arc deposition

[14-18], laser arc deposition [19-22], pulsed laser ablation [23-30], laser vaporization [31-

37], plasma enhanced chemical vapour deposition (PECVD) [38-40], arc discharge [21,

41, 42], and mixed mode cathodic arc sputtering [43-45]. A FCVA (shown in Figure 3-6) is capable of producing thin films of a wide range of materials (both reactively and non- reactively) including metals, oxides, and carbon [46-55]. During FCVA carbon deposition processes, carbon ions, electrons, and other particles in the plasma travel toward the substrate. A curved solenoidal magnetic filter allows only ions and electrons to reach the target substrate where the ions condense to form the film. For the carbon depositions performed in this project, the deposition current through the 68 mm-diameter 99.99% pure carbon cathode within the RMIT FCVA system was 60 A DC and the voltage was 27±3 V.

The chamber/chassis served as the anode which was grounded. Pulsed substrate bias voltages of -500 V and -1000 V were used to increase the ion arrival energy and the pulsing alleviated charging and discharging across the insulating photoresist [56]. Energies of the

70 depositing fluxes were assumed to be in the range from 0.12 to 1.0 keV due to the fact that about 95% C ions produced by the system are singly charged (C+) ions [57]. A photograph of the FCVA system installed in RMIT University, Melbourne is shown in Figure 3-7. The magnetic filter and cathode chamber are on the right of this image with current, voltage and heating controllers on the left.

Figure 3-6: Schematic of a filtered cathodic vacuum arc deposition system, reproduced from [58].

71

Figure 3-7: FCVA deposition system installed in RMIT University, Melbourne.

All carbon films in this project were deposited on substrates using FCVA deposition. The bonding fractions and electrical properties of the carbon films were strongly determined by the FCVA deposition conditions, as reported previously. Notably, the fraction of graphite-like bonding (sp2 hybridisation) could be altered from ~20% to more than ~90% by increasing the magnitude of the substrate bias from floating (~25 V) to 1.0 kV. This enabled insulating (low sp2 fraction) and highly conducting (high sp2 fraction) films to be synthesized on substrates near room temperature and therefore provided compatibility with substrates that were photo-lithographically patterned.

72

3.1.6 Ion beam sputtering of metallic device contact layers

A Gatan 682 precision sputter coating system as in Figure 3-8 was utilized to coat

Pt, Au or Ag onto the surface of the devices. The system is equipped with three ion guns operating at 10kV and a base pressure of 10−5 to 10-6 Torr. The metals are deposited with a typical deposition rate of 10 nm/min. The coating thickness can be controlled via the system controller. For the devices described in this thesis, Pt layers of thickness 20nm were deposited on top of the C device layer to improve optical contrast for device probing.

Figure 3-8: Gatan 682 precision coating system.

73

3.2 Electrical characterization

3.2.1 Current-voltage characterization

Current-voltage (I-V) characterization was performed using a probe-station and a

Keysight B2902A twin channel source-measurement unit (SMU). The SMU was interfaced with a personal computer to enable data extraction and manipulation. The schematic in

Figure 3-9 shows the typical two-probe measurement circuit. One probe was placed on the top side Schottky contact pad, and the other probe was placed on high conductivity silver paste contacting reverse-side Ohmic contacts. Dual Al contacts were initially deposited on the backside of the semiconductor substrates (with resistivity ~ 1 Ω.cm) and subsequently annealed at 350 °C in forming gas for 30 seconds. To determine that the contacts were

Ohmic, their room-temperature I-V characteristics were measured in two-probe configuration prior to the Schottky diode measurements.

Figure 3-9: Through film I-V measurement circuit.

74

3.2.2 Temperature dependent electrical measurements

Variable temperature device measurements were performed with the aid of a

Linkam PE95 temperature controller (shown in Figure 3-10) coupled with a device test chamber. The latter was equipped with a sample stage with liquid nitrogen cooling and electrical heating. This sample stage also featured miniature electrical probe arms. The temperature range of the system was ~80 K to 480 K with stability at set-point temperatures of ± 0.5 K. The miniature probes were connected to the aforementioned Keysight SMU to record I-V data at the different operating temperatures.

Figure 3-10: Linkam temperature variating system.

3.2.3 Capacitance-voltage characterization

A Boonton 7200 C-V meter (shown in Figure 3-11) was available for this project.

It is a microprocessor-based capacitance meter enabling fully automatic 1.0 MHz

75 capacitance measurement. Its operating capacitance range is ~5 fF to 2000 pF. This meter was connected to the probes on the probe-station for capacitance measurements on

Schottky diodes. The capacitance of the test leads and probes was de-embedded prior to device measurements. From the C-V measurements, doping densities and surface/interface charge densities were calculated.

Figure 3-11: Boonton 7200 C-V measurement system.

3.3 Material characterization

3.3.1 Introduction

In this section, different material characterization methods such as Scanning

Electron Microscopy (SEM), X-ray Photoelectron Spectroscopy (XPS) and Raman spectroscopy, will be introduced. These methods were utilized to analyze composition, surface and structure of devices in the research project. These analyses assisted in assessing the impact of different parameters on the performance of the devices. In microanalysis, the analytical spot size ranges from approximately 100 μm down to 100 nm and the depth of

76 the analysis can be as small as a few nanometers. Sample preparation is required for many of the analysis techniques. Figure 3-12 shows the most common materials analysis techniques on a plot indicating spot size (sample collection area) and sensitivity. In this project, SEM, Raman and XPS were employed.

Figure 3-12: Common microanalysis techniques with different resolution. Figure reproduced from [59].

77

3.3.2 Scanning Electron Microscopy (SEM)

The SEM uses a beam of high energy electrons generated by an electron gun, passed through magnetic lenses, focused at the specimen surface and systematically scanned

(rastered) across the surface of a specimen as shown in Figure 3-13. The SEM image is formed from a serial data stream i.e. it is an electronic image. Increased magnification is produced by decreasing the size of the area scanned together with increasing the energy level of the electron beam.

Figure 3-13: SEM system schematic showing the electron gun emitting electrons which are then accelerated as a beam towards the sample/specimen. Resulting electrons from the sample are collected to form an image on a computer screen. Figure reproduced from [59].

78

There are several imaging modes of the SEM such as secondary electron (SE) imaging, backscattered electron (BSE) imaging and energy dispersive X-ray spectroscopy

(EDX or EDS).

A Philips XL30 equipped with a tungsten filament and a Nova NanoSEM equipped with a field emission gun were both used in this project. The XL30 can collect secondary electron (topography) images using a standard chamber mounted Everhart-Thornley detector, or a backscatter detector can be mounted to the bottom of the column to collect backscattered electron images. Backscattered electron images contain atomic number contrast (Z-contrast), and can enable elemental identification.

Figure 3-14: The Philips XL30 SEM tool at the RMIT University RMMF (RMIT

Microscopy and Microanalysis Facility).

79

3.3.3 X-ray Photoelectron Spectroscopy (XPS)

X-ray Photoelectron Spectroscopy (XPS) is a technique used to measure chemical and electronic characteristics of elements within materials. As shown in Figure 3-15, photo-electrons from core atomic energy levels are collected in a high to ultra high vacuum environment (typically from 10-8 mbar to less than 10-9 mbar). A beam of X-rays is focused on the sample surface and this causes photo-electrons to be ejected from the constituent atoms within the material. These ejected photo-electrons are received at a collection lens with different intensities. An electron energy analyzer measures the kinetic energy of electrons and detectors quantify the numbers of electrons that arrive within small energy ranges. The position and shape of characteristic peaks in electron binding energy level plots provide information on the elements and compounds in the sample and acquired XPS spectra can be compared with reference spectra from the literature [60-63]. In this project,

XPS was performed using a Thermo Scientific K-alpha system with a monochromated Al

K-α 1487 eV source.

Figure 3-15: Demonstration of a XPS system’s operation [64].

80

Quantification of the graphitic order/bonding fractions in the carbon contact materials was an important application of XPS in this project. For this, the C 1s spectrum was collected and peak fitting was performed to reveal the contributions from sp2 and sp3 bonded C [65-69]. Figure 3-16a shows that after peak fitting, the sp2:sp3 ratio in a carbon film deposited near room temperature from a C bias of 500 eV is determined to be approximately 2:1 while the ratio for a carbon film deposited at 1.0 keV is approximately

3:1 (Figure 3-16b). Besides controlling the substrate bias voltage within the FCVA system, it is also possible to control the sp2:sp3 ratio of carbon by adjusting other parameters such as substrate temperature, arc current, and source to substrate distance

[70, 71] or by other methodologies such as selective oxidation or by injecting argon or hydrogen during deposition [72, 73].

81

Figure 3-16: Typical XPS data from C films energetically deposited with ion energies of

(a) -500 eV and (b) -1.0 keV substrate biases applied.

3.3.4 Raman spectroscopy

Raman spectroscopy provides a structural fingerprint. The method relies on measuring the change of frequency that occurs in light that is inelastically scattered via photon interactions from a material [74, 75]. The laser light source employed in Raman spectroscopy is usually in the visible range. The laser light interacts with the sample’s surface by transferring energy between the photon and the molecule, resulting in the energy of the re-emitted photons being shifted relative to the incident photon energy [74, 76, 77].

82

The shift in energy of the re-emitted photons provides information about the microstructure of the samples [76-79]. Raman spectroscopy was performed using a Horiba Scientific

LabRAM HR Evolution equipped with a 532 nm laser (shown in Figure 3-17).

Figure 3-17: Photograph of a Horiba Scientific LabRAM HR Evolution Raman system,

such as used at RMIT University.

Figure 3-18 shows representative Raman and X-ray photoelectron spectra (XPS) taken from different carbon films. Figure 3-18 (a) and (b), show the Raman spectra from the C films deposited at 25 ºC/0.52 keV and 100 ºC/1.0 keV. D and G peaks in these spectra are characteristic of graphite with some structural order [80]. The intensity of the D peak divided by the intensity of the G peak provides a measure of the structural order [80]. This number is 0.84 for the 25 ºC/0.52 keV and 1.1 for the 100 ºC/1.0 keV layers showing (as expected) that the higher energy deposition provides greater structural order. The additional peaks in these spectra are attributed to surface-bound carbon-oxygen contaminants that result from and remain after photolithography.

83

Figure 3-18: Raman spectra taken from carbon films deposited at (a) 25 ºC/0.52 keV and

(b) 100 ºC/1.0 keV. The peak from the Si substrate and the D and G peaks are labeled.

3.3.5 Film thickness measurement

A KLA Tencor P-16+ surface profiler (Figure 3-19) was used in this project to measure the thickness of deposited films and device layers. From these measurements, deposition rates were inferred based on the duration of the deposition process.

84

Figure 3-19: KLA Tencor P-16+ surface profiler.

3.3.6 Hall measurement

An Ecopia HMS-3000 Hall Measurement System shown in Figure 3-20 was available for sheet resistance, carrier concentration and carrier mobility measurements. The

HMS-3000 also includes a I-V curve capability which was used to verify the linearity of ohmic substrate contacts. The measurement procedure follows the four-point Van der Pauw method [81-84]. A spring clip board shown in the lower right corner of Figure 3-20 was used to make electrical contacts with the sample substrate during measurement. In this project, Hall measurements were mainly used to measure the doping concentrations of the

Si, 6H-SiC and β-Ga2O3 substrates.

85

Figure 3-20: Hall measurement HMS-3000 system.

3.4 Summary

In this chapter, the principle fabrication and characterization methodologies used during the project were described. The Schottky diodes were fabricated using standard photolithographic processes. The FCVA deposition used to deposit the carbon device layers is not a method used in CMOS fabrication but proved compatible with conventionally patterned photoresist (the temperatures reached during energetic deposition were moderate and exposure to the plasma did not compromise the photoresist).

Importantly, by manipulating the FCVA deposition’s conditions, carbon contacts with differing electrical properties could be produced. I-V-T characterization, C-V

86 characterization, SEM analysis, XPS analysis and Raman spectroscopy were then employed to study the films and devices.

3.5 References

[1] W. Kern, "The evolution of silicon wafer cleaning technology," Journal of the Electrochemical Society, vol. 137, no. 6, pp. 1887-1892, 1990. [2] W. Kern, "Handbook of semiconductor wafer cleaning technology," New Jersey: Noyes Publication, pp. 111-196, 1993. [3] D. Bornside, C. Macosko, and L. Scriven, "Spin coating: One‐dimensional model," Journal of Applied Physics, vol. 66, no. 11, pp. 5185-5193, 1989. [4] D. B. Mitzi, L. L. Kosbar, C. E. Murray, M. Copel, and A. Afzali, "High-mobility ultrathin semiconducting films prepared by spin coating," Nature, vol. 428, no. 6980, p. 299, 2004. [5] L. Scriven, "Physics and applications of dip coating and spin coating," MRS Online Proceedings Library Archive, vol. 121, 1988. [6] D. B. Hall, P. Underhill, and J. M. Torkelson, "Spin coating of thin and ultrathin polymer films," Polymer Engineering & Science, vol. 38, no. 12, pp. 2039-2045, 1998. [7] F. Zhang et al., "Ultrathin film organic transistors: precise control of semiconductor thickness via spin‐coating," Advanced Materials, vol. 25, no. 10, pp. 1401-1407, 2013. [8] D. Schubert, "Spin coating as a method for polymer molecular weight determination," Polymer bulletin, vol. 38, no. 2, pp. 177-184, 1997. [9] J.-S. Sohn et al., "Design and fabrication of diffractive optical elements by use of gray-scale photolithography," Applied optics, vol. 44, no. 4, pp. 506-511, 2005. [10] R. Voelkel et al., "Advanced mask aligner lithography (AMALITH)," in Optical Microlithography XXV, 2012, vol. 8326, p. 83261Y: International Society for Optics and Photonics. [11] A. Moafi, D. W. M. Lau, A. Z. Sadek, J. G. Partridge, D. R. McKenzie, and D. G. McCulloch, "Energetic deposition of carbon in a cathodic vacuum arc with a biased mesh," Journal of applied physics, vol. 109, no. 7, p. 073309, 2011. [12] D. L. Smith, Thin-film deposition: principles and practice. McGraw Hill Professional, 1995. [13] R. Lossy, D. L. Pappas, R. A. Roy, J. P. Doyle, J. J. Cuomo, and J. Bruley, "Properties of amorphous diamond films prepared by a filtered cathodic arc," Journal of applied physics, vol. 77, no. 9, pp. 4750-4756, 1995. [14] M. Keidar, I. Levchenko, T. Arbel, M. Alexander, A. M. Waas, and K. Ostrikov, "Increasing the length of single-wall carbon nanotubes in a magnetically enhanced arc discharge," Applied Physics Letters, vol. 92, no. 4, p. 043129, 2008. 87

[15] S. T. Huxtable et al., "Interfacial heat flow in carbon nanotube suspensions," Nature materials, vol. 2, no. 11, p. 731, 2003. [16] C. T. White and T. N. Todorov, "Carbon nanotubes as long ballistic conductors," Nature, vol. 393, no. 6682, p. 240, 1998. [17] S. Iijima and T. Ichihashi, "Single-shell carbon nanotubes of 1-nm diameter," nature, vol. 363, no. 6430, p. 603, 1993. [18] E. Dujardin, T. W. Ebbesen, A. Krishnan, and M. M. Treacy, "Wetting of single shell carbon nanotubes," Advanced Materials, vol. 10, no. 17, pp. 1472-1475, 1998. [19] A. Grill, "Diamond-like carbon: state of the art," Diamond and related materials, vol. 8, no. 2-5, pp. 428-434, 1999. [20] Q. Wei, R. Narayan, A. Sharma, J. Sankar, and J. Narayan, "Preparation and mechanical properties of composite diamond-like carbon thin films," Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films, vol. 17, no. 6, pp. 3406-3414, 1999. [21] A. Voevodin and M. Donley, "Preparation of amorphous diamond-like carbon by pulsed laser deposition: a critical review," Surface and Coatings Technology, vol. 82, no. 3, pp. 199-213, 1996. [22] W. Kautek, S. Pentzien, A. Conradi, J. Krüger, and K.-W. Brzezinka, "Pulsed- laser deposition and boron-blending of diamond-like carbon (DLC) thin films," Applied surface science, vol. 106, pp. 158-165, 1996. [23] K. Awasthi et al., "Synthesis of nano-carbon (nanotubes, nanofibres, graphene) materials," Bulletin of Materials Science, vol. 34, no. 4, p. 607, 2011. [24] K. Murata et al., "Molecular potential structures of heat-treated single-wall carbon nanohorn assemblies," The Journal of Physical Chemistry B, vol. 105, no. 42, pp. 10210-10216, 2001. [25] N. Mutsukura and K. Miyatani, "Deposition of diamond-like carbon film in CH4 He rf plasma," Diamond and related materials, vol. 4, no. 4, pp. 342-345, 1995. [26] H.-J. Scheibe and B. Schultrich, "DLC film deposition by laser-arc and study of properties," Thin Solid Films, vol. 246, no. 1-2, pp. 92-102, 1994. [27] T. Witke, T. Schuelke, J. Berthold, C. Meyer, and B. Schultrich, "Deposition of hard amorphous carbon coatings by laser and arc methods," Surface and Coatings Technology, vol. 116, pp. 609-613, 1999. [28] P. Banks et al., "Short-pulse laser deposition of diamond-like carbon thin films," Applied Physics A, vol. 69, no. 1, pp. S347-S353, 1999. [29] A. V. Rode, B. Luther-Davies, and E. G. Gamaly, "Ultrafast ablation with high- pulse-rate lasers. Part II: Experiments on laser deposition of amorphous carbon films," Journal of Applied Physics, vol. 85, no. 8, pp. 4222-4230, 1999. [30] M. Moreno-Couranjou, M. Monthioux, J. Gonzalez-Aguilar, and L. Fulcheri, "A non-thermal plasma process for the gas phase synthesis of carbon ," Carbon, vol. 47, no. 10, pp. 2310-2321, 2009.

88

[31] A. T. Chuang, B. O. Boskovic, and J. Robertson, "Freestanding carbon nanowalls by microwave plasma-enhanced chemical vapour deposition," Diamond and Related Materials, vol. 15, no. 4-8, pp. 1103-1106, 2006. [32] A. Puretzky, D. Geohegan, X. Fan, and S. Pennycook, "In situ imaging and spectroscopy of single-wall carbon nanotube synthesis by laser vaporization," Applied Physics Letters, vol. 76, no. 2, pp. 182-184, 2000. [33] A. Puretzky, D. Geohegan, X. Fan, and S. Pennycook, "Dynamics of single-wall carbon nanotube synthesis by laser vaporization," Applied Physics A, vol. 70, no. 2, pp. 153-160, 2000. [34] H. Kataura et al., "Optical properties of single-wall carbon nanotubes," Synthetic metals, vol. 103, no. 1-3, pp. 2555-2558, 1999. [35] V. Merkulov, D. Lowndes, Y. Wei, G. Eres, and E. Voelkl, "Patterned growth of individual and multiple vertically aligned carbon nanofibers," Applied Physics Letters, vol. 76, no. 24, pp. 3555-3557, 2000. [36] A. V. Melechko et al., "Vertically aligned carbon nanofibers and related structures: controlled synthesis and directed assembly," Journal of applied physics, vol. 97, no. 4, p. 3, 2005. [37] G. Cocorullo, F. Della Corte, I. Rendina, C. Minarini, A. Rubino, and E. Terzini, "Amorphous silicon waveguides and light modulators for integrated photonics realized by low-temperature plasma-enhanced chemical-vapor deposition," Optics Letters, vol. 21, no. 24, pp. 2002-2004, 1996. [38] Y. Wu, P. Qiao, T. Chong, and Z. Shen, "Carbon nanowalls grown by microwave plasma enhanced chemical vapor deposition," Advanced materials, vol. 14, no. 1, pp. 64-67, 2002. [39] P. Pai, S. S. Chao, Y. Takagi, and G. Lucovsky, "Infrared spectroscopic study of SiO x films produced by plasma enhanced chemical vapor deposition," Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films, vol. 4, no. 3, pp. 689-694, 1986. [40] M. Chhowalla et al., "Growth process conditions of vertically aligned carbon nanotubes using plasma enhanced chemical vapor deposition," Journal of applied physics, vol. 90, no. 10, pp. 5308-5317, 2001. [41] F. Xiong, Y. Wang, V. Leppert, and R. Chang, "Pulsed laser deposition of amorphous diamond-like carbon films with ArF (193 nm) excimer laser," Journal of materials research, vol. 8, no. 9, pp. 2265-2272, 1993. [42] D. L. Pappas et al., "Pulsed laser deposition of diamond‐like carbon films," Journal of applied physics, vol. 71, no. 11, pp. 5675-5684, 1992. [43] M. Lattemann, B. Abendroth, A. Moafi, D. McCulloch, and D. McKenzie, "Controlled glow to arc transition in sputtering for high rate deposition of carbon films," Diamond and Related Materials, vol. 20, no. 2, pp. 68-74, 2011. [44] R. Ganesan et al., "Synthesis of highly tetrahedral amorphous carbon by mixed- mode HiPIMS sputtering," Journal of Physics D: Applied Physics, vol. 48, no. 44, p. 442001, 2015.

89

[45] M. Lattemann, A. Moafi, M. Bilek, D. McCulloch, and D. McKenzie, "Energetic deposition of carbon clusters with preferred orientation using a new mixed mode cathodic arc–Sputtering process," Carbon, vol. 48, no. 3, pp. 918-921, 2010. [46] P. Wang, X. Wang, Y. Chen, G. Zhang, W. Liu, and J. Zhang, "The effect of applied negative bias voltage on the structure of Ti-doped aC: H films deposited by FCVA," Applied Surface Science, vol. 253, no. 7, pp. 3722-3726, 2007. [47] D. Bootkul, B. Supsermpol, N. Saenphinit, C. Aramwit, and S. Intarasiri, "Nitrogen doping for adhesion improvement of DLC film deposited on Si substrate by Filtered Cathodic Vacuum Arc (FCVA) technique," Applied Surface Science, vol. 310, pp. 284-292, 2014. [48] R. McCann, S. Roy, P. Papakonstantinou, G. Abbas, and J. McLaughlin, "The effect of thickness and arc current on the structural properties of FCVA synthesised ta-C and ta-C: N films," Diamond and related materials, vol. 14, no. 3-7, pp. 983-988, 2005. [49] M. Bilek and W. Milne, "Filtered cathodic vacuum arc (FCVA) deposition of thin film silicon," Thin solid films, vol. 290, pp. 299-304, 1996. [50] I. Aksenov, V. Belous, V. Padalka, and V. Khoroshikh, "Transport of plasma streams in a curvilinear plasma-optics system," Soviet Journal of Plasma Physics, vol. 4, pp. 758-763, 1978. [51] S. Falabella and D. Sanders, "Comparison of two filtered cathodic arc sources," Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films, vol. 10, no. 2, pp. 394-397, 1992. [52] D. M. Sanders and A. Anders, "Review of cathodic arc deposition technology at the start of the new millennium," Surface and Coatings Technology, vol. 133, pp. 78-90, 2000. [53] P. Martin and A. Bendavid, "Review of the filtered vacuum arc process and materials deposition," Thin solid films, vol. 394, no. 1-2, pp. 1-14, 2001. [54] R. Boxman et al., "Recent progress in filtered vacuum arc deposition," Surface and Coatings Technology, vol. 86, pp. 243-253, 1996. [55] M. Chhowalla, Y. Yin, G. Amaratunga, D. McKenzie, and T. Frauenheim, "Highly tetrahedral amorphous carbon films with low stress," Applied physics letters, vol. 69, no. 16, pp. 2344-2346, 1996. [56] M. Kracica, E. Mayes, H. Tran, A. Holland, D. McCulloch, and J. Partridge, "Rectifying electrical contacts to n-type 6H–SiC formed from energetically deposited carbon," Carbon, vol. 102, pp. 141-144, 2016. [57] S. Xu et al., "Properties of carbon ion deposited tetrahedral amorphous carbon films as a function of ion energy," Journal of applied physics, vol. 79, no. 9, pp. 7234-7240, 1996. [58] Y. Wang et al., "Comprehensive study of ZnO films prepared by filtered cathodic vacuum arc at room temperature," Journal of applied physics, vol. 94, no. 3, pp. 1597-1604, 2003. [59] RMMF, "Myscope Training," https://myscope.training/.

90

[60] B. V. Crist, "Advanced peak-fitting of monochromatic XPS spectra," Journal of Surface Analysis, vol. 4, no. 3, pp. 428-434, 1998. [61] B. V. Crist, "Handbooks of Monochromatic XPS Spectra. Volume 2— Commercially pure binary oxides," XPS International, LLC, Mountain View, 2005. [62] B. V. Crist, "Handbook of Monochromatic XPS Spectra, Semiconductors," Handbook of Monochromatic XPS Spectra, Semiconductors, by B. Vincent Crist, pp. 568. ISBN 0-471-49266-3. Wiley-VCH, October 2000., p. 568, 2000. [63] B. V. Crist and D. B. Crisst, Handbook of monochromatic XPS spectra. Wiley New York, 2000. [64] L. H. Yahia and L. K. Mireles, "4 - X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectrometry (ToF SIMS)," in Characterization of Polymeric Biomaterials, M. C. Tanzi and S. Farè, Eds.: Woodhead Publishing, 2017, pp. 83-97. [65] D. Lau et al., "Microstructural investigation supporting an abrupt stress induced transformation in amorphous carbon films," Journal of Applied Physics, vol. 105, no. 8, p. 084302, 2009. [66] D. Lau et al., "Abrupt stress induced transformation in amorphous carbon films with a highly conductive transition phase," Physical review letters, vol. 100, no. 17, p. 176101, 2008. [67] M. S. N. Alnassar et al., "Graphitic Schottky contacts to Si formed by energetic deposition," MRS Online Proceedings Library Archive, vol. 1786, pp. 51-56, 2015. [68] D. W. M. Lau et al., "Abrupt Stress Induced Transformation in Amorphous Carbon Films with a Highly Conductive Transition Phase," Physical Review Letters, vol. 100, no. 17, p. 176101, 04/28/ 2008. [69] J. Diaz, G. Paolicelli, S. Ferrer, and F. Comin, "Separation of the sp 3 and sp 2 components in the C1s photoemission spectra of amorphous carbon films," Physical Review B: Condensed Matter, vol. 54, no. 11, p. 8064, 1996. [70] Y. Umehara, S. Murai, Y. Koide, and M. Murakami, "Effects of sp2/sp3 bonding ratios on field emission properties of diamond-like carbon films grown by microwave plasma chemical vapor deposition," Diamond and related materials, vol. 11, no. 7, pp. 1429-1435, 2002. [71] C. Surdu-Bob, R. Vladoiu, M. Badulescu, and G. Musa, "Control over the sp2/sp3 ratio by tuning plasma parameters of the Thermoionic Vacuum Arc," Diamond and Related Materials, vol. 17, no. 7-10, pp. 1625-1628, 2008. [72] S. Osswald, G. Yushin, V. Mochalin, S. O. Kucheyev, and Y. Gogotsi, "Control of sp2/sp3 carbon ratio and surface chemistry of nanodiamond powders by selective oxidation in air," Journal of the American Chemical Society, vol. 128, no. 35, pp. 11635-11642, 2006. [73] S. Scaglione, R. Giorgi, J. Lascovich, and G. Ottaviani, "Study of the sp2-to-sp3 ratio of dual-ion-beam sputtered hydrogenated amorphous carbon films," Surface and Coatings Technology, vol. 47, no. 1-3, pp. 287-291, 1991.

91

[74] D. A. Long, "Raman spectroscopy," New York, pp. 1-12, 1977. [75] N. Colthup, Introduction to infrared and Raman spectroscopy. Elsevier, 2012. [76] M. S. Dresselhaus, G. Dresselhaus, R. Saito, and A. Jorio, "Raman spectroscopy of carbon nanotubes," Physics reports, vol. 409, no. 2, pp. 47-99, 2005. [77] L. Malard, M. Pimenta, G. Dresselhaus, and M. Dresselhaus, "Raman spectroscopy in graphene," Physics Reports, vol. 473, no. 5-6, pp. 51-87, 2009. [78] A. C. Ferrari, "Raman spectroscopy of graphene and graphite: disorder, electron– phonon coupling, doping and nonadiabatic effects," Solid state communications, vol. 143, no. 1-2, pp. 47-57, 2007. [79] D. S. Knight and W. B. White, "Characterization of diamond films by Raman spectroscopy," Journal of Materials Research, vol. 4, no. 2, pp. 385-393, 1989. [80] F. Tuinstra and J. L. Koenig, "Raman spectrum of graphite," Journal of Chemical Physics, vol. 53, no. 3, pp. 1126-1130, 1970. [81] P. M. Hemenger, "Measurement of high resistivity semiconductors using the van der Pauw method," Review of Scientific Instruments, vol. 44, no. 6, pp. 698-700, 1973. [82] O. Philips’Gloeilampenfabrieken, "A method of measuring specific resistivity and Hall effect of discs of arbitrary shape," Philips Res. Rep, vol. 13, no. 1, pp. 1-9, 1958. [83] R. Chwang, B. Smith, and C. Crowell, "Contact size effects on the van der Pauw method for resistivity and Hall coefficient measurement," Solid-State Electronics, vol. 17, no. 12, pp. 1217-1227, 1974. [84] H. Ogawa, M. Nishikawa, and A. Abe, "Hall measurement studies and an electrical conduction model of tin oxide ultrafine particle films," Journal of Applied Physics, vol. 53, no. 6, pp. 4448-4455, 1982.

92

CHAPTER 4

GRAPHITE ON SILICON SCHOTTKY DIODE

This chapter presents temperature dependent electrical characteristics of energetic

C on p-Si diode. Using energetic deposition, highly rectifying graphitic Schottky contacts to p-type Si are demonstrated. After analysis, this data is used to elucidate the properties of the C-Si interface, including lateral homogeneity, and explain the near-ideal device characteristics.

4.1 I-V characteristics

A typical room temperature current-voltage (I-V) characteristic from a C/p-Si diode is shown in Figure 4-1 and is consistent with previously reported characteristics [1]. The ideality factor (n) of the diode was obtained (assuming thermionic emission over a Schottky barrier and neglecting shunt resistance) from the slope of the linear portion of the forward

I-V characteristic in Figure 4-1 and using I = Is [exp(q(V-IRs)/nkT) –1] where V is the voltage across the diode, q is electron charge, k is Boltzmann's constant and T is temperature. The barrier height, ΦB, of the diode was calculated to be 0.60 eV from Is =

2 AA*effT exp(-qΦB/kT) where A is the active diode area and A*eff is the effective Richardson constant. The saturation current (Is) was determined to be 20 pA by extrapolation of the linear region in forward bias to the y-intercept [2].

93

Figure 4-1: Room temperature I-V characteristics of an energetically deposited C/p-Si junction. The inset shows the structure of the device and the measurement circuit.

With an ideality factor of n = 1.05 and five orders of magnitude current rectification at ±1.0V, these devices are amongst the best reported for graphitic/graphenic junctions to

Si. For comparison, Sinha et al. [3] reported ‘ideal’ graphene-Si Schottky diodes with ideality factors of n = 1.08. Non-energetically deposited (colloidal) graphitic carbon for

C/p-Si devices formed on the same substrate as energetically deposited graphitic carbon exhibit inferior characteristics [1], consistent with the presence of more defects at the interfaces. This result suggests that the energetic C flux affects the interface. Electron microscopy on sectioned samples has shown negligible surface oxide between the C and

Si [4]. Since a native oxide removal process (such as wet etching in HF acid) was not performed prior to the energetic C deposition, the oxide must have been removed by the energetic plasma during deposition.

94

I-V-T (I-V at different temperatures) measurements were performed on the C/p-Si

Schottky devices to further investigate the properties of the interfaces. Figure 4-2(a) shows the low forward bias region of the I-V characteristics of one such device. (The current is limited at higher forward bias by the series resistance). The validity of the Schottky emission model is demonstrated by the linearity of the data across the range of temperatures

2 investigated. According to thermionic emission theory, ln(Is/AT ) = -qΦBeff/kT + ln(A*eff)

[2] and an effective Richardson constant for the p-Si substrate can therefore be extracted from the I-V-T data (as shown in Figure 4-2(b)). This produces a value for A*eff of 27 ± 8

Acm-2K-2, in agreement with the reported value of 32 Acm-2K-2 for holes in p-type Si [5].

The effective barrier height determined from Figure 4-4 is ΦBeff = 0.54 ± 0.006 eV. Under reverse bias (Figure 4-3), the dominant transport mechanism was direct tunnelling, as expected for a Schottky diode with low reverse leakage current.

95

Figure 4-2: (a) Forward bias I-V characteristics of a C/p-Si Schottky diode (C deposited at 100 ºC and 1.0 keV) with linear fits showing transport dominated by thermionic emission in the temperature range 98 - 498 K. (b) Richardson plot used to extract the effective Richardson constant.

Figure 4-3: Theoretical fits of direct tunneling model for the I-V-T characteristics of a C/p-Si diode in the region of the reverse-bias voltages. Reverse currents at temperature

98 K are below the noise threshold 2 pA.

96

The aforementioned extraction method for A*eff and ΦBeff is known to be affected by barrier height inhomogeneity [6]. Figure 4-4 shows apparent barrier heights and ideality factors of the same C/p-Si Schottky device evaluated at temperatures from 98 - 498 K.

Over this temperature range, the apparent barrier height varies from 0.2 to 0.8 eV while the ideality factor varies from 1.2 to 6.0. Similar variations in these parameters have been reported for Schottky devices on many differing semiconducting materials with many differing metallic and semi-metallic contacts (see for example [7-9]). In laterally inhomogeneous junctions, regions of the barrier with lower height transmit more current at lower temperature and as the device temperature increases, more current is transmitted over higher barriers leading to an increase in the effective device area [7].

Figure 4-4: Temperature dependent barrier height and ideality factor of the energetically deposited (100 ºC/1.0 keV) C/p-Si Schottky diode.

97

4.2 C-V characteristics

Figure 4-5 shows plots of 1/C2 versus applied bias obtained from the energetically deposited devices. The difference in slope of the 0.52 keV and 1.0 keV is attributed to the differing microstructures resulting from the differing deposition conditions as described in

Section 3.3.3. These are approximately linear, consistent with an approximately uniform doping concentration (2.6 × 1016 cm-3 which agrees with manufacturing doping concentration 5.0 × 1016 cm-3) in the p-type Si substrates. A linear extrapolation to the intercept with the x-axis gives the barrier height (фC-V) and built in potential (Vbi) according to the relationship фC-V = Vbi + Vo + kT/q where Vo is the bulk potential of Si based on the effective density of states [2]. The C-V measurements yielded higher (flat-band) barrier heights (1.0eV) than those obtained from the I-V measurements (0.6eV). This disparity in barrier height measured by I-V and C-V is often observed and is consistent with the presence of a thin, resistive layer at the interface between the C contact and p-type Si substrate [10]. Molecular dynamics simulations have shown that the first few monolayers of C deposited onto a surface at high energy are disordered and are not sp2-rich [11]. As the deposition continues, the orientation and high sp2 bonding develop in tandem (see supp. movie 3 in [11]). Therefore it is believed that these a-C interfacial layers exist in these devices and leads to the aforementioned disparity. Previous work suggests this layer may also be a contributory factor in the high performance of the devices. Song et al. [12] reported that an increase (from 0.5 to 1.5 nm) in the thickness of an interfacial resistive layer in graphene/n-Si diodes resulted in significantly improved device characteristics with

-6 -12 n changing from 1.8 to 1.1 and Is changing from 10 to 10 A.

98

Figure 4-5: Plots of inverse capacitance square versus reverse bias voltage measured from

Ag/C/p-Si contacts with C deposited at the energy/temperature indicated.

4.3 Richardson constant

Since inhomogeneity can be minimised but never eradicated, numerous analysis methods have been developed to clarify the effects of inhomogeneity using I-V-T characteristics [6, 13]. The approach of Kim et al. [13] is followed to calculate the

Richardson constant by assuming a Gaussian distribution of barrier heights over the

Schottky junction area. This Schottky junction is then characterized by a zero-bias mean barrier height (ΦBmean) and a standard deviation (σo), and the apparent barrier height ΦB

2 then equals ΦBmean − qσo /2kT [13, 14]. Fitting the data plotted in Figure 4-6 results in the values of ΦBmean = 1.0 eV and σo = 0.13 V. This σo is significantly smaller than ΦBmean

(only contribute to 1% difference of apparent barrier height ΦB), consistent with a near- ideal Schottky interface. For comparison, similar analysis performed on high performance

99

Al/Si3N4/p-Si Schottky contacts yielded σo = 0.091 V [15] and on Al/TiO2/p-Si Schottky contacts, σo = 0.137 V [16].

Figure 4-6: Barrier height versus 1/2kT plot for the energetically deposited 100

ºC/1.0 keV C/p-Si Schottky diode.

2 2 2 2 A modified Richardson (ln(Is/T )−q σo /2(kT) versus q/kT) plot, in which σo is obtained from Figure 4-6 (as in [14]), is shown in Figure 4-7. Linear fitting yields ΦBmean

= 1.0 eV, which agrees with the ΦBmean value obtained from Figure 4-6. Extrapolating to the intercept gives an effective Richardson value of 34 ± 6 Acm-2K-2 which again agrees with the theoretical value of 32 Acm-2K-2 for p-Si [5]. The correspondence between the extracted and theoretical Richardson constants shows that the underlying surface of Si substrate is not significantly damaged by the energetic deposition process. Hence, the energetic deposition method provides the aforementioned surface oxide/contamination removal without causing damage to the Si device layer itself.

100

Figure 4-7: Modified Richardson plot for the 100 ºC/1.0 keV C/p-Si Schottky diode.

Luongo et al. [17] recently published an investigation of graphene/p-Si Schottky diodes in which the extracted Richardson constant differed significantly from the theoretical value. The authors included a tunneling attenuation factor in a redefined

Richardson equation to account for an insulating interlayer in their devices. They hence determined the insulating interlayer to be 16 Å in thickness. In this thesis, the experimentally extracted and theoretical Richardson constants are in close agreement and using the approach in ref. [17] to calculate an interlayer thickness returns a negligible value

(<0.1 Å). The devices reported in [17] exhibited higher ideality factors and lower rectification ratios than those reported here and these characteristics are consistent with a non-negligible interlayer thickness.

Self-sputtering is a well reported effect in the energetic deposition of metals [18].

When depositing carbon in a cathodic arc system, it is possible to select a bias voltage (> 101

1 kV) where self-sputtering by the energetic ions is sufficient to entirely prevent film growth. Molecular dynamics simulations of oriented graphitic film growth clearly show this self-sputtering [11]. Sputtering of silicon/oxygen by C ions is less well reported/characterised but cross-sectioned devices have shown that the native oxide can be effectively removed from the semiconductor surface [4]. The I-V-T analysis presented here confirms the absence of a SiO2 insulating interlayer and suggests that its removal is possible without incurring significant damage to the underlying Si.

Other authors have found that hydrogen passivation of surface defects can affect the electrical characteristics of Schottky diodes (see for example [19]). However, annealing at 620 K (for 1-hour) performed in vacuum and Ar on the C/p-Si diodes studied here, resulted only in a very small (<1%) reduction in series resistance (observed at high forward voltage) and otherwise unaltered I-V characteristics. This thermal stability contrasts with the results reported in [19] which indicated that even at moderately elevated temperatures hydrogen was removed, surface defects were reactivated and as a consequence, the characteristics of the devices were altered significantly.

Using the Schottky-Mott equation (Eq. 2-1), the effective work function of the energetically deposited carbon is calculated at 4.59 eV with effective barrier height 0.54 eV and electron affinity of Si of approximately 4.05 eV at room temperature [20-22]. As the generally accepted range for graphite work function is from 4.5 eV to 4.8 eV [23], the calculated value is inside this range.

102

4.4 Summary

Temperature dependent current-voltage (I-V-T) measurements have been performed on energetically deposited graphitic contacts to p-type Si. These contacts exhibited room-temperature current rectification ratios exceeding 105:1 at ±1 V and ideality factors as low as n = 1.05. I-V-T measurements and subsequent analysis showed that the junctions exhibited low lateral inhomogeneity when compared with other high quality

Schottky contacts. Using two differing approaches, Richardson constants for the p-Si substrate were extracted from the I-V-T data and these agreed with the theoretical value.

This showed that deleterious contamination/native-oxide was removed from the Si surface by the energetic plasma without incurring significant damage to the bulk Si. Hence, this method for forming C-based contacts provides the high-quality interfaces necessary for

Schottky diodes with room-temperature ideality factors approaching unity.

4.5 References

[1] M. S. N. Alnassar et al., "Graphitic Schottky contacts to Si formed by energetic deposition," MRS Online Proceedings Library Archive, vol. 1786, pp. 51-56, 2015. [2] S. M. Sze and K. K. Ng, Physics of semiconductor devices, 3rd ed. John Wiley & Sons, 2006. [3] D. Sinha and J. U. Lee, "Ideal graphene/silicon Schottky junction diodes," Nano letters, vol. 14, no. 8, pp. 4660-4664, 2014. [4] M. Kracica, E. L. H. Mayes, H. N. Tran, A. S. Holland, D. G. McCulloch, and J. G. Partridge, "Rectifying electrical contacts to n-type 6H–SiC formed from energetically deposited carbon," Carbon, vol. 102, pp. 141-144, 2016/06/01/ 2016. [5] H. Tecimer, Ö. Vural, A. Kaya, and Ş. Altındal, "Current-transport mechanism in Au/V-doped PVC+ TCNQ/p-Si structures," International Journal of Modern Physics B, vol. 29, no. 13, p. 1550076, 2015.

103

[6] K. Sarpatwari, O. O. Awadelkarim, M. W. Allen, S. M. Durbin, and S. E. Mohney, "Extracting the Richardson constant: IrO x/n-ZnO Schottky diodes," Applied physics letters, vol. 94, no. 24, p. 242110, 2009. [7] J. P. Sullivan, R. T. Tung, M. R. Pinto, and W. R. Graham, "Electron transport of inhomogeneous Schottky barriers: A numerical study," Journal of Applied Physics, vol. 70, no. 12, pp. 7403-7424, 1991/12/15 1991. [8] R. T. Tung, A. F. J. Levi, J. P. Sullivan, and F. Schrey, "Schottky-barrier inhomogeneity at epitaxial NiS2 interfaces on Si(100)," Physical Review Letters, vol. 66, no. 1, pp. 72-75, 01/07/ 1991. [9] F. Roccaforte, F. L. Via, V. Raineri, R. Pierobon, and E. Zanoni, "Richardson’s constant in inhomogeneous silicon carbide Schottky contacts," Journal of Applied Physics, vol. 93, no. 11, pp. 9137-9144, 2003. [10] P. Hacke, T. Detchprohm, K. Hiramatsu, and N. Sawaki, "Schottky barrier on n‐ type GaN grown by hydride vapor phase epitaxy," Applied physics letters, vol. 63, no. 19, pp. 2676-2678, 1993. [11] D. W. M. Lau et al., "Abrupt Stress Induced Transformation in Amorphous Carbon Films with a Highly Conductive Transition Phase," Physical Review Letters, vol. 100, no. 17, p. 176101, 04/28/ 2008. [12] Y. Song et al., "Role of interfacial oxide in high-efficiency graphene–silicon Schottky barrier solar cells," Nano letters, vol. 15, no. 3, pp. 2104-2110, 2015. [13] H. Kim, A. Sohn, and D.-W. Kim, "Silver Schottky contacts to Zn-polar and O- polar bulk ZnO grown by pressurized melt-growth method," Semiconductor Science and Technology, vol. 27, no. 3, p. 035010, 2012. [14] S. Chand and J. Kumar, "Evidence for the double distribution of barrier heights in Pd2Si/n-Si Schottky diodes from I-V-T measurements," Semiconductor Science and Technology, vol. 11, no. 8, p. 1203, 1996. [15] S. Zeyrek, Ş. Altındal, H. Yüzer, and M. M. Bülbül, "Current transport mechanism in Al/Si3N4/p-Si (MIS) Schottky barrier diodes at low temperatures," Applied Surface Science, vol. 252, no. 8, pp. 2999-3010, 2006/02/15/ 2006. [16] O. Pakma, N. Serin, T. Serin, and Ş. Altındal, "The double Gaussian distribution of barrier heights in Al/TiO2/p-Si (metal-insulator-semiconductor) structures at low temperatures," Journal of Applied Physics, vol. 104, no. 1, p. 014501, 2008/07/01 2008. [17] G. Luongo, A. Di Bartolomeo, F. Giubileo, C. A. Chavarin, and C. Wenger, "Electronic properties of graphene/p-silicon Schottky junction," Journal of Physics D: Applied Physics, vol. 51, no. 25, p. 255305, 2018. [18] A. Anders, "Observation of self-sputtering in energetic condensation of metal ions," Applied physics letters, vol. 85, no. 25, pp. 6137-6139, 2004. [19] R. Van Meirhaeghe, W. Laflere, and F. Cardon, "Influence of defect passivation by hydrogen on the Schottky barrier height of GaAs and InP contacts," Journal of applied physics, vol. 76, no. 1, pp. 403-406, 1994. [20] R. Hunger, R. Fritsche, B. Jaeckel, W. Jaegermann, L. J. Webb, and N. S. Lewis, "Chemical and electronic characterization of methyl-terminated Si (111) surfaces

104

by high-resolution synchrotron photoelectron spectroscopy," Physical Review B, vol. 72, no. 4, p. 045317, 2005. [21] W. Zhu, T.-P. Ma, T. Tamagawa, J. Kim, and Y. Di, "Current transport in metal/hafnium oxide/silicon structure," IEEE Electron Device Letters, vol. 23, no. 2, pp. 97-99, 2002. [22] X. Li et al., "Graphene‐on‐silicon Schottky junction solar cells," Advanced materials, vol. 22, no. 25, pp. 2743-2748, 2010. [23] E. Taft and L. Apker, "Photoelectric emission from polycrystalline graphite," Physical Review, vol. 99, no. 6, p. 1831, 1955.

105

CHAPTER 5

GRAPHITE ON SILICON CARBIDE SCHOTTKY DIODE

In this chapter, temperature-dependent current– voltage (I–V–T) and capacitance– voltage (C–V) characteristics have been collected from energetically deposited, lithographically defined C/6H-SiC contacts, while the barrier height, ideality factor and

Richardson constant have been extracted from the measurement data. In addition, we use an approach similar to that adopted by Sarpatwari et al. [1] to investigate the effects of lateral inhomogeneity on these parameters and to arrive at modified values based on laterally homogeneous C/ 6H-SiC devices.

5.1 I-V characteristics

The I-V characteristics measured at room temperature for two C/6H-SiC devices, both 200 µm in diameter are shown in Figure 5-1 with the measurement circuit shown in the inset. These devices differ in how the substrate bias was applied during C deposition.

The data shown in grey originates from a device deposited with the 1.0 kV bias applied directly to the substrate holder. The black curve originates from a device deposited with

1.0 kV bias applied remotely to a conducting C coated mesh in front of the substrate (with the substrate at floating potential) due to the fact that mesh bias enables more effective bias than possible with substrate bias applied to a semiconducting or semi-insulating substrate

[2]. The deposition then largely takes place with energetic neutrals. The latter curve produces a broader distribution of ion energies, with more ions arriving at the substrate

106 with lower energy. This lower average energy has resulted in a rectifying device but not a

Schottky barrier device, since the I-V characteristic is non-linear at low forward bias in the semi-logarithmic I-V plot. This non-linear behavior indicates that a field independent barrier exists in this device which then consequently resembles a metal-insulator- semiconductor diode. Some non-linearity was observed in the reverse bias current and this supports this assertion. In this structure, it is likely that the insulator is an interfacial oxygen rich layer, since the native oxide on the 6H-SiC was not removed prior to C deposition.

The device formed with direct bias does exhibit linearity at low forward bias and its characteristics can be fitted well assuming thermionic emission (TE) over a Schottky barrier [3]. From both device characteristics, various room-temperature device parameters can be extracted. Firstly, the rectification ratios at ± 1.5 V are respectively 2k and 10k for the remote and direct DC biased devices whilst their series resistances are 170 and 80 Ω.

The ideality factor of the direct bias deposited device is 1.8, significantly closer to unity than the ideality factor of the device deposited with remote bias (3.1). This is consistent with the differing dominant transport mechanisms in the two devices. Ideality factors between 1 and 2 are generally taken as indicative of TE over a Schottky barrier [3]. The differences in the device parameters show that the energy of deposition plays a significant role in the device characteristics. Transmission electron microscopy has been used to show that energetic C fluxes can remove native oxide layers from device interfaces as deposition takes place [4, 5] and the Schottky behaviour of the direct biased device is attributed to this effect. Finally, the apparent (room temperature) barrier heights are 0.67 and 0.71 eV with the higher value belonging to the device deposited with direct bias.

107

Figure 5-1: I-V characteristics of energetically deposited C/6H-SiC diodes with direct bias (grey) and remote bias (black) applied during deposition and (inset) schematic of the device measurement circuit.

Following the room temperature measurements, I-V-T characterisation was performed on the (direct biased) C/6H-SiC Schottky barrier device to further investigate its electrical properties. The 0.0 to 0.5V forward bias portions of the temperature dependent

I-V characteristics from this device (measured from 98 – 498 K) are shown in Figure 5-2a.

All characteristics are shown with linear fits, corresponding to TE over a Schottky barrier.

Assuming TE, the forward current-voltage (I-V) characteristics of the rectifying

C/6H-SiC diode are expressed by:

푞(푉 − 퐼푅 ) 퐼 = 퐼 exp⁡ [( 푠 ) − 1] Eq. 5-1 푠 푛푘푇 where I is the current, T is the temperature, q is the electron charge, k is the Boltzmann constant, n is the ideality factor, V is the applied voltage, Rs is the series resistance and Is is the saturation current, given by:

108

푞훷 퐼 = 퐴퐴∗푇2exp (− 퐵) ⁡ Eq. 5-2 푠 푘푇 where A is the contact area, A* is the Richardson constant with theoretical value of 194

-2 -2 Acm K [6] and ΦB is the barrier height.

The temperature dependent differential device resistance (Rdiff = dV/dI) is shown in

Figure 5-2b. The series resistance of the device, Rs, is determined from each curve at saturation and ranges from 48 Ω at 498 K to 1.18 kΩ at 98 K. The dominant contribution to this series resistance is from the bulk 6H-SiC substrate and (as shown in the Figure 5-2 inset) is proportional to –EA/kT where EA is the energy difference between the active donor level and conduction band of the n-type 6H-SiC [7]. Here, EA is equal to 0.032 ± 0.002 eV.

Figure 5-2: a) Forward bias portions of the I-V-T characteristics of the C/6H-SiC

Schottky diode with linear fits showing transport by TE and b) temperature-dependent differential device resistance versus voltage characteristics. The energy difference between the donor level and conduction band in the n-type 6H-SiC substrate is determined from the inset Arrhenius plot.

109

Next, the rectification ratio, barrier height and ideality factors based on the I-V-T characteristics (see Figure 5-3) are re-evaluated. The temperature dependent I-V characteristics and rectification ratio are shown in Figure 5-3a and Figure 5-3b, respectively. The reverse current (number of charge carriers overcoming the barrier) increases appreciably as temperature is increased [8]. The rectification ratio, calculated using fixed bias voltages ± 1.5 V is therefore maximum (~106) at the minimum temperature

(98K) while at room temperature, it exceeds 104. This reduction of rectification ratio is at least in part due to narrowing of the depletion region in the 6H-SiC as the carrier concentration in the 6H-SiC increases with increased temperature. Since the depletion width in a Schottky diode is inversely proportional to the square root of the carrier concentration in the semiconductor [3, 9, 10], increasing the temperature leads to a reduction in depletion layer width. In reverse bias, a narrower depletion region enables substantially increased reverse tunneling current. This reduces the calculated rectification ratio since the forward current is dominated by thermionic emission and hence increases less significantly with increased temperature. This is most clearly seen in the combined forward and reverse bias I-V-T plot shown in Figure 5-3a.

Referring to Eq. 5-1, the barrier height and ideality factor at each temperature are determined from:

푘푇 퐴퐴∗푇2 휙퐵 = ln ( ) ⁡ Eq. 5-3 푞 퐼푠 and

푞 푑푉 푛 = ⁡ Eq. 5-4 푘푇 푑(푙푛퐼)

110 as shown in Figure 5-3c.

Figure 5-3: Temperature dependent a) I-V characteristics, b) rectification ratios and c) ideality factors/apparent barrier heights of a C/6H-SiC diode.

Figure 5-3c shows that the barrier appears to increase with elevated temperature.

This is again indicative of barrier inhomogeneity [11-13]. If the contact exhibits location dependent interface properties, then it can be viewed as an array of patches with differing barrier heights [14]. At lower temperatures, a larger fraction of the total device current will come from carriers crossing patches with lower barrier heights and therefore the apparent barrier height will decrease (as shown in Figure 5-3c). The ideality factor increases as the temperature decreases since a greater proportion of the (smaller) device current results from carriers transported by mechanisms other than TE.

111

Figure 5-4a shows the reverse bias I-V-T characteristic of C/6H-SiC diodes at 3V

> |V| > 0.6V. When Schottky emission is the dominant transport mechanism, the current is described by [15]:

푞 푞푉 퐼⁡ ∝ 푇2푒푥푝 ( √ ) ⁡ Eq. 5-5 2푘푇 휋휖휖표푑

where ϵ is the relative permittivity, ϵo is the permittivity of free space and d is thickness of the potential barrier.

This Schottky emission plot shown in Figure 5-4a exhibits linearity throughout the reverse voltage region and across all measurement temperatures. Current over the lowered barrier is therefore dominant. Similar characteristics in the reverse bias region were reported for Ti/Ni/SiC [16], CdS [17], Au/GaAs [18], Au/PMI/n-Si and Al/ZnO/Au

Schottky diodes [19-21]. As shown in Figure 5-4b, the dominant transport mechanism in the region |V| < 0.6V is described by the direct tunneling equation [22]:

퐼 1 2푑√2푚 휙 푙푛 ( ) ∝ 푙푛 ( ) − 푒 퐵 ⁡ Eq. 5-6 푉2 푉 ℏ where me is the electron effective mass and ℏ is Planck’s constant divided by 2π.

In junctions with significant inhomogeneity, regions (‘patches’) within the total junction area may exhibit barriers thin enough to enable such a tunneling current [23].

112

Figure 5-4: The temperature dependent current voltage characteristics of C/6H-

SiC diodes with reverse-bias voltages of a) 3V > |V| > 0.6V and b) 0.6V > |V| > 0V. The linearities in a) and b) are consistent with transport dominated by Schottky emission and direct tunneling, respectively.

During cycling the temperature, there is small changes in the device characteristics

(less than 0.2 current order of magnitude). This suggests the device perform stably at this temperature range 98 to 498 K.

5.2 C-V characteristics

Capacitance-voltage (C-V) characterization was performed to further investigate the properties of the interface/barrier in the C/6H-SiC devices. The ‘true’ capacitance (C) is evaluated using the formula [24, 25]:

113

1/2 퐶−2 − 2푅2휔2 − √(퐶−2 − 2푅2휔2)2 − 42푅4휔4 푚 푠 푚 푠 푠 Eq. 5-7 퐶 = [ 4 4 ] ⁡ 2푅푠 휔 where Cm is measured capacitance, Rs is the device series resistance and ω is angular frequency.

The room temperature 1/C2 vs V relationship (Figure 5-5) was linear over the reverse bias measurement range but the barrier height obtained from these measurements, significantly exceeded that calculated from the room temperature I-V measurements (1.46 eV versus 0.7 eV). According to Tung’s model [14, 26-29] this can be explained by the presence of barrier inhomogeneity since (during the I-V measurements) current transport occurs mainly over the low Schottky barrier height ‘patches’ within the larger device area

[12, 26]. Thus, the discrepancy in barrier heights determined from I-V and C-V measurements is consistent with the presence of lateral inhomogeneity and provides further justification for investigation of this effect.

Figure 5-5: The capacitance voltage characteristic of C/6H-SiC junction under reverse bias at room temperature measured at 1 MHz.

114

5.3 Richardson constant

With confirmed barrier inhomogeneity, the approach reported by Sarpatwari et al.

[1] was followed to extract the effective values of Richardson constant A*eff and barrier height ΦBeff, using the following relationship:

퐼 푞훷퐵푒푓푓 푙푛 ( 푠 ) = − + ln⁡(A∗ )⁡⁡ Eq. 5-8 퐴푇2 푘푇 푒푓푓

which includes saturation currents (Is) extracted from device I-V measurements and the device area (A). A plot of this relationship (Figure 5-6a) provides values for A*eff of

-2 -2 0.45 Acm K and ΦBeff of 0.52 eV. The A*eff value is significantly less than the reported

-2 -2 A* value for 6H-SiC of 194 Acm K [6] while ΦBeff is also significantly lower than the barrier height calculated from I-V-T data. Similar disparity has been reported in numerous

Schottky diodes based on a variety of materials [29, 30] and once again lateral inhomogeneity of the junction is a known cause. It is reasonable to assume that this is the case when the device ideality factor (n) exceeds 1.3 [1, 23]. The relationship between A*eff and ΦBeff is described by [1]:

1 ∗ ( ∗ ) ln(A 푒푓푓) = ln A 표 + (훽 − 2 2) (훷퐵푒푓푓 − 훷퐵표)⁡⁡ Eq. 5-9 훽훽1 휎

where A*o and ΦBo are the homogeneous Richardson constant and barrier height. β

1/3 equals q/kT, β1 equals (Vbb/η) with the Vbb term being the band bending at the measurement bias and η = ϵs/qND the doping density dependent parameter [14]. Parameter

η determines the overall patch strength and at higher doping levels (low η), conduction through the low barrier regions is enhanced. From Eq. 5-9, the plot in Figure 5-6b is constructed to exhibit a linear relationship between ln(A*eff) and ΦBeff. As the barrier height

115 becomes more homogeneous, A*eff and ΦBeff moves closer to A*o and ΦBo. When the barrier height becomes homogeneous, the ideality factor converges to a value of 1.0. To determine

ΦBo, the ΦBeff vs. n plot is extrapolated to n =1.05 (which is the lowest reported ideality factor for a Schottky contact to 6H-SiC [23, 31, 32]) to yield ΦBo= 1.05 ± 0.015 eV. This

-2 - ΦBo is then used to extract a homogeneous Richardson constant A*o of 187 ± 42 Acm K

2. This range encompasses the theoretical value of 194 Acm-2K-2.

Figure 5-6: a) Richardson plot constructed from the I-V-T data of a C/6H-SiC

Schottky diode and b) linear correlation between ln(A*eff) and ΦBeff obtained from several devices and ΦBeff vs. n relationship used to determine the barrier height characteristic of a laterally homogeneous energetically deposited C/6H-SiC Schottky diode.

116

Using the Schottky-Mott equation (Eq. 2-1), the effective work function of the energetically deposited carbon is calculated to be 4.52 eV from barrier height 0.52 eV and electron affinity of 6H-SiC is approximately 4 eV at room temperature [33, 34]. As the general accepted range for the graphite work function is from 4.5 eV to 4.8 eV [35], the calculated value is near the lower end of reported value. This discrepancy may be mitigated with the improvement of the barrier height homogeneity.

5.4 Summary

This chapter reports on the investigation of the temperature-dependent electrical properties of energetically deposited C/6H-SiC Schottky diodes. Lateral inhomogeneity of the junction/barrier manifests itself in the I–V–T characteristics and causes disparity between barrier heights determined from the I–V and C–V measurements. Using established methods, and from the relationships between barrier height and ideality factor and barrier height versus Richardson constant, the homogeneous barrier height and

Richardson constant were calculated from the measurement data. The calculated

Richardson constant (187 ± 42 A cm-2 K-2) agreed with the theoretical value of 194 Acm-2

K-2 for n-type 6H-SiC. This work demonstrates the potential to form high performance

Schottky contacts to 6H-SiC using energetically deposited carbon, but shows that high lateral homogeneity across the junctions of these devices is vital in achieving higher rectification and ideality factors approaching unity.

117

5.5 References

[1] K. Sarpatwari, O. O. Awadelkarim, M. W. Allen, S. M. Durbin, and S. E. Mohney, "Extracting the Richardson constant: IrO x/n-ZnO Schottky diodes," Applied physics letters, vol. 94, no. 24, p. 242110, 2009. [2] A. Moafi, D. W. M. Lau, A. Z. Sadek, J. G. Partridge, D. R. McKenzie, and D. G. McCulloch, "Energetic deposition of carbon in a cathodic vacuum arc with a biased mesh," Journal of applied physics, vol. 109, no. 7, p. 073309, 2011. [3] S. M. Sze and K. K. Ng, Physics of semiconductor devices, 3rd ed. John Wiley & Sons, 2006. [4] D. W. M. Lau et al., "Microstructural investigation supporting an abrupt stress induced transformation in amorphous carbon films," Journal of Applied Physics, vol. 105, no. 8, p. 084302, 2009/04/15 2009. [5] D. W. M. Lau et al., "Abrupt Stress Induced Transformation in Amorphous Carbon Films with a Highly Conductive Transition Phase," Physical Review Letters, vol. 100, no. 17, p. 176101, 04/28/ 2008. [6] T. Teraji, S. Hara, H. Okushi, and K. Kajimura, "Ideal Ohmic contact to n-type 6H-SiC by reduction of Schottky barrier height," Applied Physics Letters, vol. 71, no. 5, pp. 689-691, 1997/08/04 1997. [7] W. P. Kang, J. L. Davidson, Y. Gurbuz, and D. V. Kerns, "Temperature dependence and effect of series resistance on the electrical characteristics of a polycrystalline diamond metal‐insulator‐ semiconductor diode," Journal of Applied Physics, vol. 78, no. 2, pp. 1101-1107, 1995/07/15 1995. [8] A. Di Bartolomeo, "Graphene Schottky diodes: An experimental review of the rectifying graphene/semiconductor heterojunction," Physics Reports, vol. 606, pp. 1-58, 2016/01/08/ 2016. [9] S. M. Sze, Semiconductor devices: physics and technology. John Wiley & Sons, 2008. [10] C. Crowell and S. Sze, "Current transport in metal-semiconductor barriers," Solid- state electronics, vol. 9, no. 11-12, pp. 1035-1048, 1966. [11] J. P. Sullivan, R. T. Tung, M. R. Pinto, and W. R. Graham, "Electron transport of inhomogeneous Schottky barriers: A numerical study," Journal of Applied Physics, vol. 70, no. 12, pp. 7403-7424, 1991/12/15 1991. [12] J. H. Werner and H. H. Güttler, "Barrier inhomogeneities at Schottky contacts," Journal of Applied Physics, vol. 69, no. 3, pp. 1522-1533, 1991/02/01 1991. [13] J. H. Werner and H. H. Güttler, "Temperature dependence of Schottky barrier heights on silicon," Journal of Applied Physics, vol. 73, no. 3, pp. 1315-1319, 1993/02/01 1993.

118

[14] R. T. Tung, "Electron transport at metal-semiconductor interfaces: General theory," Physical Review B, vol. 45, no. 23, pp. 13509-13523, 06/15/ 1992. [15] H.-D. Lee, "Characterization of shallow silicided junctions for sub-quarter micron ULSI technology. Extraction of silicidation induced Schottky contact area," IEEE Transactions on Electron Devices, vol. 47, no. 4, pp. 762-767, 2000. [16] K. P. Schoen, J. M. Woodall, J. A. Cooper, and M. R. Melloch, "Design considerations and experimental analysis of high-voltage SiC Schottky barrier rectifiers," IEEE Transactions on Electron Devices, vol. 45, no. 7, pp. 1595-1604, 1998. [17] F. A. Padovani, "Thermionic emission in Au GaAs Schottky barriers," Solid- State Electronics, vol. 11, no. 2, pp. 193-200, 1968/02/01/ 1968. [18] R. Zuleeg and R. S. Muller, "Space-charge-limited currents and Schottky- emission currents in thin-film CdS diodes," Solid-State Electronics, vol. 7, no. 8, pp. 575-578, 1964/08/01/ 1964. [19] K. H. Yoo, K. S. Kang, Y. Chen, K. J. Han, and J. Kim, "The effect of TiO2 concentration on conduction mechanism for TiO2-polymer diode," Applied Physics Letters, vol. 93, no. 19, p. 192113, 2008/11/10 2008. [20] K. J. Han, K. S. Kang, Y. Chen, K. H. Yoo, and K. Jaehwan, "Effect of annealing temperature on the conduction mechanism for a sol–gel driven ZnO Schottky diode," Journal of Physics D: Applied Physics, vol. 42, no. 12, p. 125110, 2009. [21] Ö. F. Yüksel, M. Kuş, N. Şimşir, H. Şafak, M. Şahin, and E. Yenel, "A detailed analysis of current-voltage characteristics of Au/perylene-monoimide/n-Si Schottky barrier diodes over a wide temperature range," Journal of Applied Physics, vol. 110, no. 2, p. 024507, 2011/07/15 2011. [22] J. M. Beebe, B. Kim, J. W. Gadzuk, C. D. Frisbie, and J. G. Kushmerick, "Transition from direct tunneling to field emission in metal-molecule-metal junctions," Physical review letters, vol. 97, no. 2, p. 026801, 2006. [23] L. Zheng, R. P. Joshi, and C. Fazi, "Effects of barrier height fluctuations and electron tunneling on the reverse characteristics of 6H–SiC Schottky contacts," Journal of Applied Physics, vol. 85, no. 7, pp. 3701-3707, 1999/04/01 1999. [24] M. M. Solovan, N. M. Gavaleshko, V. V. Brus, A. I. Mostovyi, P. D. Maryanchuk, and E. Tresso, "Fabrication and investigation of photosensitive MoO x /n-CdTe ," Semiconductor Science and Technology, vol. 31, no. 10, p. 105006, 2016. [25] M. N. Solovan, G. O. Andrushchak, A. I. Mostovyi, T. T. Kovaliuk, V. V. Brus, and P. D. Maryanchuk, "Graphite/p-SiC Schottky Diodes Prepared by Transferring Drawn Graphite Films onto SiC," Semiconductors, vol. 52, no. 2, pp. 236-241, 2018/02/01 2018. [26] R. T. Tung, A. F. J. Levi, J. P. Sullivan, and F. Schrey, "Schottky-barrier inhomogeneity at epitaxial NiS2 interfaces on Si(100)," Physical Review Letters, vol. 66, no. 1, pp. 72-75, 01/07/ 1991. [27] R. Tung, "Schottky barrier height—do we really understand what we measure?," Journal of Vacuum Science & Technology B: Microelectronics and Nanometer

119

Structures Processing, Measurement, and Phenomena, vol. 11, no. 4, pp. 1546- 1552, 1993. [28] R. T. Tung, "Recent advances in Schottky barrier concepts," Materials Science and Engineering: R: Reports, vol. 35, no. 1-3, pp. 1-138, 2001. [29] A. F. Hamida, Z. Ouennoughi, A. Sellai, R. Weiss, and H. Ryssel, "Barrier inhomogeneities of tungsten Schottky diodes on 4H-SiC," Semiconductor Science and Technology, vol. 23, no. 4, p. 045005, 2008. [30] F. Roccaforte, F. L. Via, V. Raineri, R. Pierobon, and E. Zanoni, "Richardson’s constant in inhomogeneous silicon carbide Schottky contacts," Journal of Applied Physics, vol. 93, no. 11, pp. 9137-9144, 2003. [31] M. Bhatnagar, B. J. Baliga, H. R. Kirk, and G. A. Rozgonyi, "Effect of surface inhomogeneities on the electrical characteristics of SiC Schottky contacts," IEEE Transactions on Electron Devices, vol. 43, no. 1, pp. 150-156, 1996. [32] C. Raynaud, K. Isoird, M. Lazar, C. M. Johnson, and N. Wright, "Barrier height determination of SiC Schottky diodes by capacitance and current–voltage measurements," Journal of Applied Physics, vol. 91, no. 12, pp. 9841-9847, 2002/06/15 2002. [33] J. R. Waldrop, R. W. Grant, Y. C. Wang, and R. F. Davis, "Metal Schottky barrier contacts to alpha 6H‐SiC," Journal of Applied Physics, vol. 72, no. 10, pp. 4757- 4760, 1992/11/15 1992. [34] D. M. Brown, M. Ghezzo, J. Kretchmer, E. Downey, J. Pimbley, and J. Palmour, "SiC MOS interface characteristics," IEEE Transactions on Electron Devices, vol. 41, no. 4, pp. 618-620, 1994. [35] E. Taft and L. Apker, "Photoelectric emission from polycrystalline graphite," Physical Review, vol. 99, no. 6, p. 1831, 1955.

120

CHAPTER 6

GRAPHITE ON GALLIUM OXIDE SCHOTTKY DIODE

The following chapter describes measurements performed on sputtered graphitic contacts to β-Ga2O3. Sputtering was chosen as the deposition method for these carbon contacted devices largely due to a fault and extended outage that occurred on the FCVA energetic deposition system towards the end of the PhD project. As in Chapters 4 and 5, the fabrication methods used for these devices are not restated here but were described in detail in Chapter 3 section 1. The characterization/measurement methods employed were described in Chapter 3, section 2.

6.1 I-V characteristics

The temperature dependent I-V characteristics of a sputtered graphitic C/n-type β-

Ga2O3 Schottky diode are shown from 143 – 443 K in Figure 6-1. The rise in the forward current with temperature is similar to that observed in other Ga2O3 Schottky based diodes

[1-5]. In the reverse bias region, between 363K and 403K, the reverse current sharply increases which is indicative of an activated impurity/defect that facilitates conduction through the depletion region. The noise-floor of the apparatus is approximately 2pA so the current measurements below this value can be disregarded.

121

Figure 6-1: I-V characteristic of C/Ga2O3 Schottky diode with elevated temperature from

143K to 443K.

The temperature dependent differential device resistance (Rdiff = dV/dI) is shown in

Figure 6-2. The series resistance of the device, Rs, is determined from each curve at saturation and ranges from approximately 1010 Ω at 143 K to 104 Ω at 443 K. The dominant contribution to this series resistance is from the bulk β-Ga2O3 substrate. This is due to its wide band-gap (4.9eV).

122

Figure 6-2: Temperature-dependent differential device resistance versus voltage characteristics of C/Ga2O3 Schottky diode.

The forward I-V-T characteristics in Figure 6-1 were analyzed using thermionic emission theory to determine the barrier height and the ideality factor at each temperature.

In Figure 6-3a, the rectification ratio (at ± 1 V) of the C/Ga2O3 diode is shown. The constant rise in forward current explains the increase of rectification ratio up to 1:105 at a temperature of 363K. At temperatures of 403K and 443K the rectification ratio decreases to 1:103 due to the rise in reverse current. According to the theory of thermionic emission

Eq. 5-1, the extracted apparent barrier heights vary from 0.5 eV to 1.1 eV while the ideality factor varies from 8 to 1.4 over the evaluated temperature range of 143K to 443K, as shown in Figure 6-3b. Similar behavior has been observed in Ni/Ga2O3 Schottky diodes [6], and

Cu/Ga2O3 Schottky diodes [7]. The apparent barrier heights are in the same range as other reported Ga2O3 Schottky diodes [1-4]. However, only the ideality factors from room temperature up to 443 K are comparable with other reported Ga2O3 Schottky diodes. The

123 ideality factors from room temperature to lower temperature cannot be evaluated reliably due to the presence of measurement noise in the sub-threshold region (forward current less than 10-12 A) and the extremely high series resistance (up to 1010 Ω at 143 K).

Figure 6-3: a) Rectification ratios and b) ideality factors/apparent barrier heights of a C/Ga2O3 diode.

Due to the high series resistance at lower temperatures only data from room temperature up to 443K was chosen for fitting in the forward voltage region (0V to 0.7 V) as shown in Figure 6-4a. The Schottky emission model is validated by the linearity of the

2 data across this range of temperatures. According to thermionic emission theory, ln(Is/AT )

124

= -qΦBeff/kT + ln(A*eff) [8] and an effective Richardson constant for the n-type β-Ga2O3 substrate can therefore be extracted from the I-V-T data (as shown in Figure 6-4b). This

-2 -2 produces a value for A*eff of 37 ± 9 Acm K , in agreement with the reported theoretical

-2 -2 value of 41 Acm K for β-Ga2O3 assuming an effective electron mass of 0.34 me [9, 10].

The effective barrier height determined from Figure 6-3b is ΦBeff = 0.98 ± 0.11 eV.

Figure 6-4: (a) Forward bias I-V characteristics of a C/Ga2O3 Schottky diode with linear fits showing transport dominated by thermionic emission in the temperature range 283 -

443 K. (b) Richardson plot used to extract the effective Richardson constant. Lower temperature data are unreliable for linear fit due to high series resistances.

125

6.2 Richardson constant

Since lateral inhomogeneity was observed in the C/Ga2O3 Schottky diodes, the approach of Kim et al. [11] was again followed to calculate the Richardson constant. A

Gaussian distribution of barrier heights was assumed over the Schottky junction area. This

Schottky junction was then characterized by a zero-bias mean barrier height (ΦBmean) and

2 a standard deviation (σo), and the apparent barrier height ΦB then equals ΦBmean − qσo /2kT

[11, 12]. Fitting the data plotted in Figure 6-5 results in the values of ΦBmean = 1.22 eV and

σo = 0.11 V. This σo is significantly smaller than ΦBmean, indicative of a high quality

Schottky interface and the value of σo is similar to those previously reported from high quality Schottky contacts on other semiconductors [13, 14].

Figure 6-5: Barrier height versus 1/2kT plot for the C/Ga2O3 Schottky diode.

126

2 2 2 2 Next, a modified Richardson (ln(Is/T )−q σo /2(kT) versus q/kT) plot, in which σo is obtained from Figure 6-5 (as in [12]), is shown in Figure 6-6. Linear fitting yields

ΦBmean = 1.2 eV, which agrees with the ΦBmean value obtained from Figure 6-5.

Extrapolating to the intercept gives an effective Richardson value of 43 ± 6 Acm-2K-2 which

-2 -2 also agrees with the theoretical value of 41 Acm K for β-Ga2O3 [9, 10].

Figure 6-6: Modified Richardson plot for the C/Ga2O3 Schottky diode.

Using the Schottky-Mott equation (Eq. 2-1), the effective work function of the energetically deposited carbon was calculated to be 4.4 eV to 4.9 eV. In this calculation, the barrier height was estimated to be 0.9 eV and the electron affinity of β-Ga2O3 was assumed to be in the range 3.5 eV to 4 eV at room temperature [15-18]. The range

127 calculated for the work function of graphitic carbon agrees with the generally accepted range (from 4.5 eV to 4.8 eV [19].

Besides lateral inhomogeneity, other authors have discussed the role of oxygen vacancies in determining the characteristics of Schottky contacts to β-Ga2O3 [20-22]. Hou et al. found that the barrier heights of Ag, Au, Pt, Pd, Ru, and Ir Schottky contacts to β-

Ga2O3 did not vary with the differing work-functions of these materials but instead adhered closely to the Vo (2+/0) transition level of Ec - 0.7 eV [3]. The authors proposed that oxygen

2+ vacancies and in particular, Vo (III) (the most stable charged state) were significant in

2+ pinning the Fermi level of metal Schottky contacts on β-Ga2O3. They also noted that Vo

(III) were most likely to occur near the surface of β-Ga2O3 and hence close to the Schottky interface. Since the barrier heights measured from our diodes also lie close to the Vo (2+/0) transition level of Ec - 0.7 eV, we believe that Fermi level pinning is significant in the graphitic C-contacted devices as well. It would be beneficial to test this theory by employing reactive deposition to provide oxygenated C contacts and/or to anneal the devices in the presence of oxygen and monitor the reverse leakage and barrier heights at increasing temperatures.

6.3 Summary

The experimental investigation described in this chapter involved measuring the I-

V-T characteristics of graphitic carbon contacts to β-Ga2O3. Rectification ratios at ± 1V achieved were as high as 1:105 with ideality factors as low as 1.4. Richardson constants for the β-Ga2O3 substrate were extracted from the I-V-T data using two differing approaches and both results agreed with the theoretical value 41 Acm-2K-2. This investigation

128 demonstrates the capability to form Schottky contacts to β-Ga2O3 using graphitic carbon using standard deposition and lithographic methods.

6.4 References

[1] S. Pearton et al., "A review of Ga2O3 materials, processing, and devices," Applied Physics Reviews, vol. 5, no. 1, p. 011301, 2018. [2] C. Hou et al., "High-temperature (350° C) oxidized iridium Schottky contacts on β-Ga2O3," Applied Physics Letters, vol. 114, no. 23, p. 233503, 2019. [3] C. Hou, R. Gazoni, R. Reeves, and M. Allen, "Direct comparison of plain and oxidized metal Schottky contacts on β-Ga2O3," Applied Physics Letters, vol. 114, no. 3, p. 033502, 2019. [4] E. Farzana, Z. Zhang, P. K. Paul, A. R. Arehart, and S. A. Ringel, "Influence of metal choice on (010) β-Ga2O3 Schottky diode properties," Applied Physics Letters, vol. 110, no. 20, p. 202102, 2017. [5] C. Hou, R. M. Gazoni, R. J. Reeves, and M. W. Allen, "Oxidized Metal Schottky Contacts on (010) $\beta $-Ga 2 O 3," IEEE Electron Device Letters, vol. 40, no. 2, pp. 337-340, 2019. [6] E. Ahmadi, Y. Oshima, F. Wu, and J. S. Speck, "Schottky barrier height of Ni to β-(AlxGa1− x) 2O3 with different compositions grown by plasma-assisted molecular beam epitaxy," Semiconductor Science and Technology, vol. 32, no. 3, p. 035004, 2017. [7] D. Splith et al., "Determination of the mean and the homogeneous barrier height of Cu Schottky contacts on heteroepitaxial β‐Ga2O3 thin films grown by pulsed laser deposition," physica status solidi (a), vol. 211, no. 1, pp. 40-47, 2014. [8] S. M. Sze and K. K. Ng, Physics of semiconductor devices, 3rd ed. John Wiley & Sons, 2006. [9] M. Higashiwaki et al., "Temperature-dependent capacitance–voltage and current– voltage characteristics of Pt/Ga2O3 (001) Schottky barrier diodes fabricated on n– –Ga2O3 drift layers grown by halide vapor phase epitaxy," Applied Physics Letters, vol. 108, no. 13, p. 133503, 2016. [10] T. Oishi, Y. Koga, K. Harada, and M. Kasu, "High-mobility β-Ga2O3 () single crystals grown by edge-defined film-fed growth method and their Schottky barrier diodes with Ni contact," Applied Physics Express, vol. 8, no. 3, p. 031101, 2015. [11] H. Kim, A. Sohn, and D.-W. Kim, "Silver Schottky contacts to Zn-polar and O- polar bulk ZnO grown by pressurized melt-growth method," Semiconductor Science and Technology, vol. 27, no. 3, p. 035010, 2012. [12] S. Chand and J. Kumar, "Evidence for the double distribution of barrier heights in Pd2Si/n-Si Schottky diodes from I-V-T measurements," Semiconductor Science and Technology, vol. 11, no. 8, p. 1203, 1996.

129

[13] S. Zeyrek, Ş. Altındal, H. Yüzer, and M. M. Bülbül, "Current transport mechanism in Al/Si3N4/p-Si (MIS) Schottky barrier diodes at low temperatures," Applied Surface Science, vol. 252, no. 8, pp. 2999-3010, 2006/02/15/ 2006. [14] O. Pakma, N. Serin, T. Serin, and Ş. Altındal, "The double Gaussian distribution of barrier heights in Al/TiO2/p-Si (metal-insulator-semiconductor) structures at low temperatures," Journal of Applied Physics, vol. 104, no. 1, p. 014501, 2008/07/01 2008. [15] I. López, E. Nogales, P. Hidalgo, B. Méndez, and J. Piqueras, "Field emission properties of gallium oxide micro‐and nanostructures in the scanning electron microscope," physica status solidi (a), vol. 209, no. 1, pp. 113-117, 2012. [16] T.-H. Hung et al., "Energy band line-up of atomic layer deposited Al2O3 on β- Ga2O3," Applied Physics Letters, vol. 104, no. 16, p. 162106, 2014. [17] K. D. Chabak et al., "Enhancement-mode Ga2O3 wrap-gate fin field-effect transistors on native (100) β-Ga2O3 substrate with high breakdown voltage," Applied Physics Letters, vol. 109, no. 21, p. 213501, 2016. [18] M. Mohamed, K. Irmscher, C. Janowitz, Z. Galazka, R. Manzke, and R. Fornari, "Schottky barrier height of Au on the transparent semiconducting oxide β- Ga2O3," Applied Physics Letters, vol. 101, no. 13, p. 132106, 2012. [19] E. Taft and L. Apker, "Photoelectric emission from polycrystalline graphite," Physical Review, vol. 99, no. 6, p. 1831, 1955. [20] D. Guo et al., "Oxygen vacancy tuned Ohmic-Schottky conversion for enhanced performance in β-Ga2O3 solar-blind ultraviolet photodetectors," Applied Physics Letters, vol. 105, no. 2, p. 023507, 2014. [21] K. Irmscher, Z. Galazka, M. Pietsch, R. Uecker, and R. Fornari, "Electrical properties of β-Ga2O3 single crystals grown by the Czochralski method," Journal of Applied Physics, vol. 110, no. 6, p. 063720, 2011. [22] L. Dong, R. Jia, B. Xin, B. Peng, and Y. Zhang, "Effects of oxygen vacancies on the structural and optical properties of β-Ga 2 O 3," Scientific reports, vol. 7, p. 40160, 2017.

130

CHAPTER 7

CONCLUSIONS AND FUTURE WORK

7.1 Conclusions

The aim of the project described in this thesis was to improve understanding of graphitic Schottky contacts to semiconductors. Graphitic carbon is known for its high temperature endurance and chemical inertness and these properties suggest potential for high power devices and/or devices operating at high temperatures. Three semiconducting materials with power electronics applications were selected for devices and testing; Si, 6H-

SiC and β-Ga2O3. I-V-T measurement is a reliable means to extract key parameters from

Schottky diodes and has been used extensively in this project.

The graphitic contacts demonstrated rectifying behavior with high rectification ratio when interfaced with Si, 6H-SiC and β-Ga2O3. Thermionic emission was observed in the sub-threshold region of forward bias with extracted barrier heights and Richardson constants agreeing with theoretical values. The barrier heights and ideality factors of the devices on all three substrate materials were temperature dependent, consistent with the existence of lateral inhomogeneity at the interfaces. Quantification of this lateral inhomogeneity showed that it was small (relative to other high-performance devices reported in the literature) on the Si and β-Ga2O3 devices and larger on the 6H-SiC. It is therefore likely that the quality of the 6H-SiC substrate surface accounts for this difference.

131

The work function of graphite extracted from all three devices agreed with the generally accepted range in other reported work.

7.2 Future Work

All devices in this project were subjected to investigation under elevated temperature. To further assess the performance/reliability of energetically deposited graphitic contacts, high power/ high current-density endurance experiments similar to those reported by Stelzer would yield valuable information [1]. Subjecting the energetically deposited graphitic contacts at high temperatures for extended durations would help in quantifying the thermal reliability of the graphite contacts and help to determine whether they are viable for commercial applications. It also might be useful to process the sputtering method on Si or SiC to see whether it resulted in better quality contacts compared to the energetically deposited graphitic contacts.

Oxidised noble metal contacts to wide band gap semiconductors including ZnO and

Ga2O3 have exhibited extremely high rectification ratios. This has been attributed to sub- surface oxygen vacancy passivation because oxygen vacancies are known to cause Fermi level pinning in these devices/materials [2, 3]. Therefore, introducing oxygen into graphitic carbon during deposition may enable increased barrier heights to be obtained in Schottky devices formed between the material and Ga2O3. Graphitic contacts offer cost advantages over noble metal contacts and if defect passivation can be incorporated into the deposition process for improved device performance, this would represent a significant finding.

132

7.3 References

[1] M. Stelzer and F. Kreupl, "Graphenic carbon-silicon contacts for reliability improvement of metal-silicon junctions," in 2016 IEEE International Electron Devices Meeting (IEDM), 2016, pp. 21.7.1-21.7.4. [2] C. Hou, R. M. Gazoni, R. J. Reeves, and M. W. Allen, "Oxidized Metal Schottky Contacts on (010) $\beta $-Ga 2 O 3," IEEE Electron Device Letters, vol. 40, no. 2, pp. 337-340, 2019. [3] C. Hou, R. Gazoni, R. Reeves, and M. Allen, "Direct comparison of plain and oxidized metal Schottky contacts on β-Ga2O3," Applied Physics Letters, vol. 114, no. 3, p. 033502, 2019.

133

APPENDIX A

WAFER SPECS

A.1 Wafer Specs

The properties of Si, SiC and Ga2O3 wafers are listed in below tables:

Table A-1: Properties of Si wafers.

Si wafer Properties from Silicon Wafer Inc. Diameter 50.8 mm Thickness 250 ± 15 um Resistivity 0.01 ~ 0.02 Ωcm Dopant Boron Carrier concentration 5 × 1016 cm-3 Type p-type Orientation <100> two sides polished

Table A-2: Properties of SiC wafers.

6H-SiC wafer Properties from Semiconductor Wafer Inc. Diameter 50.8 mm Thickness 330 ± 25 um Resistivity 0.02 ~ 1 Ωcm Dopant Nitrogen Carrier concentration 2 × 1017 cm-3 Polytype 6H Orientation <0001> two sides polished

Table A-3: Properties of β-Ga2O3 wafers.

β-Ga2O3 wafer Properties from Tamura Corp. Diameter 50.8 mm Thickness 0.68 ± 0.02 mm

134

Dopant Sn Carrier concentration 1.1 × 1018 cm-3 Type n-type Orientation (2̅01)

135