Acknowledgment: I would like to thank to my professor Dr.Somnath Chattopadhayay for being my advisor and guide.

I am grateful to him for his continuous support and invaluable inputs he has been providing me through the development of the project. This work would not have been possible without his support and encouragement. I also would like to thank him for showing me some examples that related to the topic of my project. Besides, I would like to thank the Department Chair Prof.Ali Amini of Electrical and Computer Engineering for providing me with a good environment and facilities to complete this project. It gave me an opportunity to participate and learn about the software Matlab. In addition, I would like to thank him that he provides me valuable information as the guidance of my project.

CONTENTS:

1. Introduction

2. Materials Technology for GaN

3. Physics of P-N Junction Diode

23

4. Analysis of electrical characteristics of GaN diode

32

5. Results and Discussion

39

6. References

51

7. Matlab Code

57

INTRODUCTION: Material properties of GaN-based and related semiconductors make them very attractive for high temperature applications. The material quality has been steadily improving over the past few years, and this improvement opened up new opportunities for a rapid progress in GaN related devices. GaN pn-junction diodes have been grown on GaN and sapphire substrates by metal organic vapor phase epitaxy and their electrical characteristics have been studied. For the diode on the GaN substrate, the reverse leakage current is lower and the breakdown voltage VB is higher than those on the sapphire substrate owing to the lower dislocation density. The breakdown voltage is further improved with decreasing Mg concentration in pGaN layers. Gallium nitride (GaN) is an attractive material for high-frequency and high-power applications because of its wide-band gap about three times higher than silicon (Si) and high breakdown electric field about ten times that of Si and high electron saturation velocity has been demonstrated. GaN based power devices with much higher breakdown voltage and lower on-resistance than conventional Si-based devices are also expected to be developed [1,2]. Because on-resistance is theoretically inversely proportional to the third power of breakdown electric field, GaN-based power devices have the potential of having the onresistance that is three orders lower than that of Si-based ones. However, GaN-based electronic devices, such as FETs, Schottky barrier diodes (SBDs) and p-n junction diodes, are generally fabricated by hetero-epitaxial growth on different material substrates such as sapphire and silicon carbide (SiC). Because the lattice mismatch and the difference in thermal expansion coefficients will cause dislocations in GaN epitaxial layers at a high density between 109 to 1010 cm-2, it is difficult to realize good crystalline quality of GaN epitaxial layer. Due to this problem, it is highly possible that previously reported GaN-based power devices could not utilize the excellent material characteristics of GaN enough [3-6]. It is presumed that the way to explore the intrinsic performance of GaN can be opened by using the GaN homo-epitaxial layers grown on low-dislocation-density GaN substrates [7]. In this study, we have demonstrated that the epitaxial layers grown on GaN substrates were much higher in purity and lower in dislocation density.
3

Despite very extensive applications in just about every sphere of life, electronic devices from Si, GaAs, and their alloys are intolerant of elevated temperatures and caustic environments. Devices from wide bandgap semi- conductors such as III-V nitrides are, however, potentially free from consideration, the aerospace these shortcomings. Taking this advantage into the

and power companies have continuously provided

impetus for the development of advanced high-power technologies capable of operating at high temperatures and hostile environments [7]. Among the candidates are SiC, other wide bandgap with indirect material property and GaN based technologies. It is expected that, when matured, a suitable high-temperature nitride based technology would allow bulky aircraft hydraulics and mechanical control systems to be replaced with heat tolerant on site control electronics in what is dubbed as the all electric aircraft. On-site electronics,

actuators, and sensors would thus reduce complexity and increase reliability. Hydraulic systems, a fire hazard in aircraft, and heat radiators in satellites would then be greatly reduced in size and number leading to considerable reductions in weights. Moreover, direct bandgap nature of nitrides renders them suitable for green and shorter wavelengths emitters and detectors making investment in this class of materials well worthwhile. As members of the 111-V nitrides semiconductor family, aluminum nitride, gallium nitride, and their alloys are all wide bandgap semiconductors. They crystallize in both wurtzite and zincblende polytypes structure. Wurtzite GaN, AlN, and InN have direct room temperature bandgaps of 3.4, 6.2, and 1.9 eV, respectively. In cubic form, GaN and InN have direct bandgaps, while AIN is indirect. In view of the available wide range of

bandgaps, when GaN is alloyed with A1N and InN, the 111-V nitrides thus formed may span a continuous range of direct bandgap energies throughout much of the visible

spectrum well into the ultraviolet (UV) wavelengths. This is one of the reasons fueling the recent interest in GaN, AlN, InN, and their tertiary alloys for short wavelength

optoelectronic device applications. These optoelectronic devices, especially emitters such as the light emitting diodes (LEDs) and lasers, which are active in the green, blue, and UV wavelengths, are essential for developing full color displays and, in the case of coherent sources, high density optical storage technologies, and very likely devices for signal and illumination applications. Digital storage may especially benefit from short wavelength coherent sources based on 111-V nitrides, because the diffraction limited optical storage density will then increase to a first extent quadratically as the probe laser wavelength is reduced. Notably, although yellow lasers based on InGaAlP heterostructures have been
4

successfully demonstrated, this material system is limited to about 550 nm. Although the ZnSe based laser technology can cover shorter wavelengths, they are capable of operation only in the green and fringes of blue wavelengths because higher fractions of Mg in the solid leads to high resistivity films and rapid oxidation. [4] Evolution of GaN technology: Recently, Group III nitride-based semiconductors have emerged as the leading material for the production of blue LEDs, blue laser diodes, and high-power, hightemperature electronics. The achievement of high brightness blue InGaN LEDs has basically caused a revolution in LED technology and opened up enormous new, previously inaccessible markets. The use of InGaN/GaN double heterostructures in LEDs in 1994 by Nakamura et al. (1994) and the achievement of p-doping in GaN by Akasaki (Amano et al. 1989) are widely credited with re-igniting the III-V nitride system. The historical evolution of GaN materials and device technology in Japan in the early 1990s is regarded as one of the key developments in solid-state devices today. These remarkable advancements in GaNbased LED materials have depended on several keybreakthroughs in materials synthesis and fabrication. The lack of p-n junctions in Group III nitrides and their poor crystal quality slowed research for many decades. The recent realization of blue lasers has taken over 20 years from the time the first optical pumped stimulated emission was observed in GaN crystals (Dingle et al. 1994) and the first LEDs (Pankove et al. 1971) were fabricated. [4] Contribution of GaN for High power device: Wide bandgap GaN has long been sought for its applications to blue and UV emitters and high temperature high power electronic devices. GaN based electronic devices such as high power and heat tolerant heterojunction bipolar transistors (HBT's) can be important components of integrated systems designed for high frequency and high speed applications, for example, in satellites and all electric aircraft. GaN based field-effect transistors (FETs) are projected to be highly useful for power amplification and switching in high temperature and high power environment. [5] There is a strong development effort on widegap power devices, with lesser efforts in GaN for achieving the higher stand-off voltages, which should have benefits that Si-based or electromechanical power electronics cannot attain. The higher stand-off voltages should eliminate the need for series stacking of devices and the associated packaging difficulties. In

addition, these wide bandgap devices should have higher switching frequency in pulse-widthmodulated rectifiers and inverters. [6] Wide band gap semiconductors (as evidenced by this issue) are extremely attractive for the gamut of power electronics applications from power conditioning to microwave transmitters for communications and RADAR. Of the various materials and device technologies, the AlGaN/GaN HEMT seems the most promising. There is a strong interest in developing wide band gap power devices for use in the electric power utility industry. With the onset of deregulation in the industry, there will be increasing numbers of transactions on the power grid in the US, with different companies buying and selling power. The main applications are in the primary distribution system ~1002000 kVA! and in subsidiary transmission systems ~150 MVA!. A major problem in the current grid is momentary voltage sags, which affect motor drives, computers, and digital controls. Therefore, a system for eliminating power sags and switching transients would dramatically improve power quality. [710] GaN based FET structures offer the potential of high speed, high power and high temperature operation beyond that of Si and GaAs. Theoretical considerations concerning FETs predict an output power density in the range of 20W/mm, with a power density above 6 W/mm being obtained at three laboratories (Cree, Nitres, HRL) and 9.8 W/mm being the present record. Such high power densities may result in high channel temperatures and the thermal stability of the structure becomes important as well as the thermal management. Indicative for the thermal stability may be the materials Debye-temperature, which is approx. 750C in the case of GaN. [12] A 140-watt highly-efficient gallium nitride (GaN) RF unmatched power transistor (UPT) with superior performance is now competing the GaAs and silicon power technologies. And then 48-volt high power-density GaN semiconductor process -- featuring a unique combination of high RF power density and efficiency, low capacitance and high thermal conductivity. Such features enable the development of compact and efficient high power amplifiers (HPAs) for a broad range of applications, including public mobile radio (PMR), 3G/4G wireless infrastructure, ISM (industrial scientific & medical), military and civilian radar and CATV transmission networks. [13]

MATERIALS PROPERTIES: Energy Band Structure/Lattice Constants: The bandgap in the (Al,Ga,In) N-based materials system ranges from 1.9 eV (InN) to 3.4 eV (GaN) to 6.2 eV (AlN). The band structure is currently thought to be a direct bandgap across the entire alloy range. This direct bandgap is especially fortuitous as it allows for high quantum efficiency light emitters to be fabricated in this system. [14] GaN Physical Properties: Group III nitrides possess several remarkable physical properties that make them particularly attractive for reliable solid state device applications. The wide bandgap materials possess low dielectric constants with high thermal conductivity pathways. Group III nitrides exhibit fairly high bond strengths and very high melting temperatures. The large bond strengths could possibly inhibit dislocation motion and improve reliability in comparison to other II-VI and III-V materials. In addition, the nitrides are resistant to chemical etching and should allow GaN-based devices to be operated in harsh environments. These properties may lead to devices with superior reliability [15]. GaN p-n junction diodes: The p-n junction diodes on GaN and on sapphire substrates have been fabricated shown in Figure.1. The thickness and donor concentration of the n-GaN layer were 7 m and about 31016 cm-3, respectively. The p-GaN layer had a thickness of 0.5 m, and its Mg impurity concentration had been optimized to 71017 cm-3 based on the result of an investigation on the dependence of breakdown voltage on Mg concentration. Mesa structures were fabricated by reactive-ion etching (RIE), and Ni/Au and Ti/Al ohmic contact electrodes were deposited by means of standard photolithography and vacuum evaporation techniques. The size of the Ni/Au electrode was 400 m400 m. [16]

Fig.1 Structures of SBDs (a) GaN substrate (b) Sapphire substrate

Figure 2 shows (a) reverse and (b) forward I-V characteristics of the p-n junction diodes fabricated on GaN and sapphire substrates. While the p-n junction diode on sapphire substrate had a large leakage current that increased with reverse voltage and exhibited a breakdown voltage of 588V, the diode on GaN substrate had an extremely low leakage current and exhibited an ideal breakdown voltage of 925 V. This breakdown voltage is the highest among the values achieved by the GaN based p-n junction diodes ever reported. The breakdown electric field of this p-n junction diode on GaN substrate was estimated to be 3.27 MV/cm, which is considerably close to the theoretical value 3.3 MV/cm. [17]

Fig.2 (a) Reverse and (b) Forward I-V characteristics of GaN PN junction diode

Fig.3 CL images of P-N junction diodes (a) on GaN substrate (b) on sapphire substrate.

Figure 3 shows the CL images of the p-n junction diodes fabricated on GaN and sapphire substrates. The dark spots observed on the surfaces of the p-n junction diodes are dislocation defects. The p-n junction diode on GaN substrate had much less dark spots than the one on sapphire. This result indicates that the formation of leakage current paths due to dislocations is much less in the p-n junction diode on the GaN substrate than on sapphire substrate, as was the case of the SBDs described above. From these results it was found that the diodes fabricated on GaN substrates have low leakage current and high breakdown voltage. As for forward I-V characteristics, the ideality factor n in the diode on GaN substrate and that in the diode on sapphire substrate are 1.4 and 2.2, respectively. The value 1.4 shows that a high-quality p-n junction was realized on GaN substrate. The diode on GaN substrate achieved a specific on-resistance of 6.3 m cm2 at 100A/cm2 resulting into a figure of merit VB2/RON of 136 MW/cm2, which is also the highest value among those obtained by the GaN-based p-n junction diodes ever reported [18]. Figure 4 shows the breakdown voltage and the specific on-resistance values achieved by the SBDs and the p-n junction diodes fabricated on GaN substrates in this study and the breakdown voltage and the specific on resistance values previously reported [19], [20], [21], [22] as well as the theoretical Si-limit and GaN-limit[23]. This figure shows that the characteristics obtained in this study are much better than the Si-limit and the values obtained by the previously reported GaN-based SBDs and p-n junction diodes. These results indicate that GaN has great potential for power device applications and that the excellent characteristics of GaN can be utilized enough by using low-dislocation-density GaN substrates.
9

Fig.4 breakdown voltage and specific on-resistance of GaN based devices

MATERIALS TECHNOLOGY FOR GaN: Substrates: Several problems in the epitaxial growth of nitrides originate from the non availability of single crystalline GaN substrates or other high quality single crystalline substrates with the same lattice parameters as GaN. For this reason, so far, most of the epitaxial growth of nitrides has been performed on sapphire or SiC substrates. In both cases, problems due to the lattice mismatch between the nitride epi-layer and the substrate (16% for sapphire and 3.5 % for SiC) have to be overcome. One of the major breakthroughs in the growth of device quality group-III nitride material was the implementation of nucleation layers. Using sapphire substrates, thin AlN or GaN nucleation layers deposited at temperatures between 500 and 750 C showed to remarkably improve the quality of the GaN film grown at temperatures above 1000 C. [24,25] In the case of SiC substrates, the growth is usually initiated with the deposition of a thin AlN nucleation layer at high temperatures. [26] By this means, GaN material of comparable quality on both types of substrates could be achieved. Since so far, most GaN growth has been performed on c-plane sapphire substrates, in the following section just the growth on c-plane sapphire will be discussed. Additional substrate materials are currently being examined to determine if the properties of the GaN thin films can be enhanced by improved structural matching. From Figure 5 it is clear that in addition to sapphire several other substrates offer potentially much better latticed and thermal matching. To this end, 6H-SiC, ZnO, and 3C-SiC, MgO are alternative substrate
10

Fig.5 Bandgap and wavelength of III-V Nitrides versus Lattice Constants.

materials. ZnO has a wurtzite structure with lattice constants of (a=3.32A, c=5.213A) and thus offers an better structural match to the equilibrium wurtzite nitride. 3C-SiC and MgO are both cubic zinc-blende structures having better structural and thermal match to the nitrides than sapphire. 3C-SiC and MgO have cubic lattice constants of a=4.36A and a=4.22A, respectively. Although the nitrides are most commonly observed as the wurtzite (2H) polytype, they can also crystallizes in a metastable zinc-blende structure (a=4.52A) when using non-equilibrium based growth techniques. The identification of a suitable substrate material that is lattice matched and thermally compatible with GaN wurtzite structure (a=3.19A, c=5.185A) will alleviate many of the difficulties associated with the deposition of device quality material. [27] MATERIALS GROWTH: Metal organic Chemical Vapor Deposition (MOCVD) In the past few years, metal-organic chemical vapor deposition (MOCVD) has evolved into a leading technique for production of III-V nitride semiconductor optoelectronic devices and electronic devices. For commercial GaN device applications MOCVD has emerged as the leading candidate because of the achievement of super-bright blue LEDs [28] and the large scale-manufacturing potential of the MOCVD technique. The majority of all GaN
11

based p-n junction light emitting diodes (LEDs) typically employ impurity related transition for blue and green emission. [29,30,31,32] Recently, direct bandgap emission in the bluegreen spectral region has been obtained using high In content in single quantum well (SQW) LEDs and lasers using the two-flow MOCVD technique[33]. Full-color LED displays can now be made entirely with the MOCVD technique when combining the blue and green GaN LEDs with the very high brightness yellow and orange emitting LEDs which were demonstrated in the AlGaInP materials system in the early 1990s. Understanding the growth of AlInGaN/GaN based materials by metal-organic chemical vapor deposition (MOCVD) is therefore of extreme importance in improving the properties of these high temperature electronic devices. [34,35] MOCVD Reaction Chemistry: The basic MOCVD reaction describing the GaN deposition process can be described by the following reaction: Ga(CH3)3(v) + NH3(v) -->GaN(s) + 3CH4(v) where (v)=vapor and (s)=solid This balanced expression ignores that the specific reaction path and reactive species are largely unknown. The details of the reaction are not well known and the intermediate reactions are thought to be complex. A more likely reaction pathway leading to growth of the GaAs epitaxial layers involves the homogeneous decomposition of TMGa as reported in an earlier study on GaAs epitaxy. [36] Ga(CH3)3(v) ---> Ga(CH3)2(v) + CH3(v) Ga(CH3)2(v) ---> GaCH3(v) + CH3(v) Ga(CH3)(v) ---> Ga(v) + CH3(v) The Group-V hydride source, is thought to decompose heterogeneously on the GaN surface or reactor walls to yield atomic nitrogen or a nitrogen containing radical at high growth temperatures. Abstraction of the first hydrogen bond is thought to be the rate limiting step in the decomposition of ammonia. NH3(s/v) ---> NH(3-x)(s/v) + x H(s/v)

12

Therefore one possible growth mechanism of GaN that might occur at the solid vapor interface could be expressed as follows: GaCH3(s/v) + NH(s/v) ---> GaN(s) + 1/2H2 However, the level of understanding of the growth process is inadequate at best. The most difficult topic, and certainly the least developed, is the area of the kinetics of the process and growth mechanisms occurring at the solid/vapor interface during MOCVD growth. Pyrolysis and diffusion of the group-III source through the boundary layer is the main pathway controlling the growth rate. However, parasitic side reactions such as solid adduct formation between TMAl and ammonia will decrease the growth rate by limiting the flux of group-III sources to the growing interface. Optimization of MOCVD growth is typically done by empirical studies of external parameters such as growth temperature, V/III ratio, substrate tilt and mass flow rates. These studies have identified three regions of growth: mass transport limited, desorption and surface kinetically limited regimes. Conventional GaN MOCVD is usually performed in the mass transport limited regime that occurs over a wide temperature range (600C1100C). In this temperature region growth is limited by mass transport of the column III reactant to the growing interface. Because the diffusion process is slightly temperature dependent, there exists a slight increase in the growth rate in this temperature range. [37] MOCVD System and Reactor Design Issues: Both atmospheric-pressure and low-pressure MOCVD reactors are employed by various research and industrial groups in the growth of GaN. Atmospheric pressure reactors are favored because a high partial pressures of ammonia or nitrogen containing precursor is achievable. MOCVD Reactor design for GaN growth must overcome problems presented by high growth temperatures, pre-reactions, flow and film non-uniformity. Typically in GaN growth very high temperatures are required because of the high bond-strength of the N-H bond in ammonia precursors. Compounding this fact is the thermodynamic tendency of ammonia to prereact with the group-III metal-organic compounds to form non-volatile adducts. These factors contribute to the difficulties currently facing researchers in the design and scale-up of III-V nitride deposition systems. Further research and development is needed in the scale-up and understanding of the mechanism of gallium nitride growth by MOCVD.

13

MOCVD Systems for Production: Currently, several types of MOCVD reactor geometries are being developed for the mass production of GaN based materials and devices. Both atmospheric pressure and lowpressure systems are being produced by the major MOCVD equipment manufacturer (Aixtron GmbH, Emcore Corp., Nippon Sanso, and Thomas Swan, Ltd. (Now Aixtron Subsidiary). The two types of geometries used by Aixtron and Swan are illustrated in figure 6. This figure shows a closed space RDR for atmospheric pressure growth, and a two-flow horizontal flow pancake reactor. Both reactor designs are producing high quality GaN materials and it is not the intent of the author to judge one superior to the other. The benefits to each approach will be specific to the ultimate device and materials being grown. The closed space RDR has the benefit of atmospheric pressure operation because the low free height eliminates free convection. The two-flow horizontal planetary rotation TM reactor from Aixtron can also be operated at near atmospheric pressure and can accommodate large wafer volumes (>7 wafers). The selection of any reactor has to be carefully considered against factors such as: material quality, high throughput, reproducibility, maintenance, and source usage.

Fig.6 Two types of MOCVD reactors currently produced in Europe, planetary rotating discs (Aixtron) and closed space RDR by Swan (now Aixtron).

GaN p-type and n-type doping: One of the key breakthroughs in the development of GaN technology was the achievement of p-doping using Mg and LEEBI treatment. Amano et al. [38], observed that under low electron beam irradiation (LEEBI) Mgdoped GaN exhibited much lower resistivity and the PL properties drastically improved. This achievement subsequently leads to the development of p/n GaN diodes with good turn on characteristics. Nakamura et al [39] built
14

upon this fundamental breakthrough to achieve even higher p-doping and uniform activation of Mg by using high temperature thermal annealing under a nitrogen ambient. The passivation requires post growth treatment for MOCVD material to activate the dopants. During the growth, interstitial hydrogen is incorporated and a H-Mg acceptor complex forms which passivates the acceptor. This H-Mg bond can be broken by a high temperature annealing step under an inert environment. This work demonstrated that hydrogen compensation of Mg in the MOCVD growth of GaN was the principal problem that plagued previous researchers. High room temperature p-doping is further complicated by the high activation energy of magnesium as the most commonly used dopant (270 meV) and the passivation of acceptors with hydrogen during CVD growth. The binding energies of dopants are dependent of the dielectric constant and effective mass of the material. The nitride system has low dielectric constant (GaN, e(0)=9.5) and large effective masses (GaN, me=0.2 mo, mhh=0.75mo) resulting in large binding energies. This is especially pronounced in ptype doping when comparing GaN to GaAs the acceptor levels are very deep because of the large hole mass. This has led to difficulty in high p-type doping. This is the result of two effects:high n-type background concentration compensating the p-dopant and the incomplete activation of the dopants at room temperature. The low p-type doping (typical values are 1017 /cm3) leads to high contact resistances and problems with current spreading. Further work on increasing the p-doping level and developing new p-dopants will result in substantial payoff in producing LEDs and lasers with lower operating voltages and higher power efficiencies. Doping n-type is rather straightforward in GaN with silicon being the typical ndopant. The as-grown material is typical unintentionally n-type, which is widely believed to be due to intrinsic nitrogen vacancies. The Si donor lies just below the conduction band (Ea = 15-25 meV). Therefore well-controlled n-type doping can be easily accomplished using silicon as the donor. The typical MOCVD precursor for n-type doping are silane (SiH4) and disilane (Si2H6) which are typical diluted with hydrogen in the 200 ppm range. Doping levels between 1(10)17 to 2(10)19 cm-3 are easily achieved in the doping of GaN with silane. [40] Growth of AlGaN and AlGaN/GaN Heterostructures: High quality AlGaN films have been demonstrated by atmospheric pressure MOCVD as well as epitaxy performed under low pressure conditions. At growth temperatures below 1100 C, the mole fraction of aluminum in the AlGaN epitaxial layer was found to be almost directly proportional to the mole fraction of TMAl in the gas phase. At temperatures above
15

1100 C, the incorporation efficiency of gallium atoms decreased. This behavior was explained by a decreased sticking probability of Ga molecules at this high temperatures. [41] High quality AlGaN/GaN heterostructures are characterized by a very high mobility of the two dimensional electron gas at the interface. [42] Values as high as 1500 cm2/Vs at room temperature have been achieved in the authors laboratory. [43] The optical properties of AlGaN/GaN quantum wells [44] were found to be determined by both, quantum and strain related effects. [45,46] MOCVD AlN films showed a full width of half maximum of the (002) x-ray rocking curve peak as low as 97 arc seconds. [47] AlN/GaN Superlattices of high structural and optical quality have also been fabricated by switched atomic layer MOCVD. [48] Growth of InGaN and InGaN/GaN Heterostructures: Growth of high quality InGaN is necessary to obtain good electrical and optical characteristics from LEDs. However, the growth of high quality InGaN has proven to be more difficult than GaN. InGaN growth needs to be performed at much lower temperatures than that of GaN, due to the low dissociation temperature of InN. [49,50] Furthermore, the decomposition of ammonia becomes less efficient with decreasing temperature due to the high kinetic barrier for breaking the nitrogen - hydrogen bonds. The growth of InGaN has to be performed at temperatures below 850 C because of the high volatility of indium at common GaN growth temperatures of above 1000 C. But even on InGaN layers grown at temperatures below 800 C, In droplet formation was observed. [51] GaN epitaxial layers were grown by low-pressure metal-organic vapor phase epitaxy (MOVPE) using trimethylgallium (TMGa), ammonia (NH3), silane (SiH4) and biscyclopentadienylmagnesium (Cp2Mg). The growth temperature, pressure and V/III ratio were in the range of 990C to 1090C, 100 to 500 Torr and 300 to 2000, respectively. The GaN substrates with an n-type carrier concentration of 4 to 61018 cm-3 and a low dislocation density in the order of 106 cm-2 were used. To clarify the effect of the substrates on the characteristics of the epitaxial layers, the GaN layers were grown simultaneously both on GaN substrate and on GaN/sapphire template (GaN layer thickness of about 3 m). The GaN epitaxial layers were characterized by capacitance-voltage (C-V) measurement, secondary ion mass spectrometry (SIMS), atomic force microscopy (AFM) and cathodoluminescence (CL) observation. [52]

16

NITRIDE MATERIALS GROWTH ISSUES: MOCVD Growth: MOCVD is a non-equilibrium growth technique that relies on vapor transport of the precursors and subsequent reactions of Group III alkyls and Group V hydrides in a heated zone. The MOCVD technique originated from the early research of Manasevit (1968) who demonstrated that triethylgallium (TEGa) and arsine deposited single crystal GaAs pyrolytically in an open tube, cold-wall reactor. The basic MOCVD reaction describing the GaN deposition process is: Ga(CH3)3(V) + NH3(V) GaN(S) + 3CH4(V) However, the details of the reaction are not well known, and the intermediate reactions are thought to be complex. Further work is needed to understand the fundamentals of this crystal growth process. [53] Various researchers employ both atmospheric-pressure and low-pressure MOCVD reactors in the growth of GaN. The majority of research groups in Japan utilize atmospheric pressure reactors because of the high partial pressures of ammonia. Nakamura and his colleagues achieved the breakthrough in bright blue LEDs using a modified MOCVD system (Nakamura et al. 1994a). Nichia Chemistries Inc. has employed a novel two-flow approach that yields excellent film quality. MOCVD reactor designs for GaN growth must overcome problems presented by high growth temperatures, pre-reactions, and flow and film non-uniformity. Typically in GaN growth, very high temperatures are required because of the high bond-strength of the N-H bond in ammonia precursors. Compounding this fact is the thermodynamic tendency of ammonia to prereact with Group III metalorganic compounds to form non-volatile adducts. These factors contribute to the difficulties currently facing researchers in the design and scale-up of III-V nitride deposition systems. Much research activity is needed in the scale-up and understanding of the mechanism of gallium nitride growth by MOCVD. [54] As shown in Figure 7, sources are supplied in this DenBaars 19 reactor from a horizontal inlet and from a vertical sub-flow rather than driving the reactants to the growing film surface. [55]

17

Fig.7 Two-Flow atmospheric pressure MOCVD approach at Nichia.

Molecular Beam Epitaxy (MBE): Several researchers in Japan have begun developing MBE for growth of III-V nitrides. Several approaches have been investigated for supplying an atomic source of nitrogen. RF plasma and electron cyclotron resonance (ECR) microwave plasma sources are the two most successful techniques discovered so far (Lei et al. 1991; Strite et al. 1991; Paisley et al. 1989). In these RF plasma MBE systems, the plasma source is used to crack molecular nitrogen. These plasma sources use cylindrical cavity geometry to efficiently pump microwave energy into the nitrogen discharge area. The plasma stream is a complex mixture of atomic, molecular, and ionic N radicals. When using ECR sources, a tradeoff between growth rate and ion damage occurs (Lei et al. 1991). Under normal ECR use, the flux of the low energy reactive N species is so low that only low growth rates of 500 /hr can be achieved. At higher microwave powers, higher growth rates can be achieved, but ion damage leading to deep levels and semi-insulating electrical properties occurs. A major advantage of MBE for nitride growth is the low growth temperature that can be achieved because of the atomic nitrogen source. This is in contrast to MOCVD, which must employ high growth temperatures (>1000C) to crack the ammonia molecules. The lower growth temperatures should result in lower thermal stress upon cooling, less diffusion, and reduced alloy segregation. This lower growth temperature is especially important in the AlGaN alloys that have a large mismatch in their thermal expansion coefficients. [56]

18

GaN Bulk Crystal Growth: In Poland, world-leading bulk GaN crystal growth is being achieved at Unipress. The Unipress research group is located at the High Pressure Research Center (HPRC) in Warsaw, where they employ GaN bulk crystal growth from a melt. As shown in figure 8 pressure as high as 10Kbar are required to growth GaN from a melt. One of the most impressive achievements the panel saw in Europe was the defect free bulk GaN wafers grown from melt under extremely high pressure. Dr. Isabel Gregory gave an overview of the highpressure solution growth method used for depositing the bulk single crystals.

Fig.8 Nitrogen pressure versus Growth temperature for bulk GaN substrate growth.

The size of the substrates has steadily increased up to 1cm in 1999 from 5mm in 1997. In comparison to GaN on sapphire technology which exhibits 1e+9cm -2 defect densities, or ELO films which are in the 1E+6cm-2 to 5E+5cm-2 range the bulk GaN crystals possess 10 to 1 defect per square centimeter. These densities were estimated from etch pit counts and correlated to TEM measurements of higher defect densities found on standard GaN sapphire technology. The team was allowed to observe the bulk crystals under both optical microscopes and field emission SEM. Both n-type bulk crystals and semi-insulating bulk crystal have been grown. The semi-insulating substrates were highly resistive 1E+5ohmcm and could be used for thin film deposition of high power AlGaN/GaN microwave amplifiers. [57] Very high quality homo-epitaxial growth of GaN by MOCVD was obtained on top of these substrates and the films exhibited narrow double-Crystal x-ray diffraction
19

(DCXRD) line-widths as low as 21 arcsec. Figure 9 shows an image of a 1cm diameter crystal grown by the Unipress group.

Fig.9 1cm Diameter Bulk GaN Substrate.

The group at University of Ulm in collaboration with Unipress has obtained thin films which exhibit the narrowest reported PL linewidths at low temperature (0.1meV) which is indicative of the uniform high quality film. [58] MBE has also been performed on the bulk crystals. Both Ga-face and N-face polarity bulk substrates have been produced. The bulk crystals also display extremely smooth cleaved facets with rms roughness of 5 angstroms which would make excellent laser facets. Lateral Epitaxial Overgrowth: Lateral Epitaxial Overgrowth (LEO) is an attractive method to produce GaN films with a low density of extended defects, which is beneficial both to studies of the fundamental properties of the GaInAlN materials system and to GaN-based device technology. The basic concept is to reduce defect propagation in masked regions of the substrate where the laterally overgrowing GaN is defect free. Recent studies in Europe, Japan and USA have confirmed that the density of threading dislocations (TDs) is reduced by 3-4 orders of magnitude in the LEO material grown on 6H-SiC and Al2O3 [59] substrates, and the mechanisms of threading dislocations evolution during LEO have been investigated. Studies of the optical properties of LEO GaN and InGaN quantum wells [60] have revealed that TDs act as non-radiative recombination centers. However, the minority carrier diffusion length (<200 nm) is smaller than the average distance between TDs such that the emission mechanisms of the carriers that do recombine radiatively appear to be unaffected by moderate TD densities (~106-109 cm-2). On the other hand, reducing the TD density has been shown to reduce the reverse leakage current by ~3 orders of magnitude in GaN p-n junctions, InGaN single and multiple quantum
20

well light emitting diodes, and GaN/AlGaN heterojunction field-effect transistors fabricated on LEO GaN. More recently, ultraviolet p-i-n photodetectors fabricated on LEO AlGaN have exhibited a similar reduction of the reverse leakage current by up to 6 orders of magnitude. [61] Effect of threading dislocations: GaN films grown on sapphire typically exhibit a high dislocation density; these threading dislocations (TDs) are likely to be electrically active and create states within the band-gap. Such states may be responsible for the high reverse-bias leakage currents observed. Indeed, a correlation between dark current and dislocation density has been reported in various materials systems including Si, SiGe, HgCdTe, and has been proposed for GaN diodes. In order to investigate this hypothesis, we used lateral epitaxial overgrowth (LEO) to fabricate GaN p-n diodes both with and without threading dislocations. [62] LEO of GaN has attracted considerable attention as a method for reducing the high TD density which has historically limited device performance in the nitride system. In this technique, standard dislocated GaN layers are partially masked and then re-grown upon. GaN grown vertically through the mask openings (windows) retains a high TD density; however material which then grows laterally across the mask has a drastically reduced TD density. For example, work in our laboratory has shown a reduction in TD density from ~4 x 108 cm-2 in the window region to below 106 cm-2 in the LEO region.[63] This reduction is dislocation density is expected to both improve device performance and facilitate investigation of the fundamental properties of GaN films. Indeed the use of LEO GaN has already yielded improved lifetime for laser diodes, and there are indications that it helps eliminate pathways for dopant diffusion. [64] Measurements of the reverse bias current density in a p-n junction should be a sensitive tool for monitoring the benefits of dislocation reduction. The devises were fabricated described in the process flow shown in figure 10. GaN films 2 m thick were grown on c-plane sapphire substrates by MOCVD and subsequently patterned with an SiO2 mask for LEO regrowth. The pattern consisted of stripes with mask openings 5 m wide separated by 45 m of SiO2. In the LEO regrowth step GaN was grown vertically through the mask opening and laterally across the mask, creating bars whose height and overgrowth width were both approximately 8m. The SiO2 mask was then removed in
21

Fig.10 Process flow for diode growth and fabrication.

the region between the LEO bars. This step exposed GaN over the entire (corrugated) surface of the wafer, so that subsequent growth was not selective epitaxy but uniform deposition. We note that despite this precaution the highly non-planar surface morphology may have affected the thickness and doping uniformity of the GaN deposited in the later growth steps. Part (a) : LEO growth. A SiO2 mask is deposited on a GaN film and window stripes are opened to the GaN beneath. During regrowth dislocations propagate vertically GaN above the window region remains dislocated while laterally grown(LEO) material has a reduced dislocation density. Part (b) : SiO2 mask is removed between the LEO bars and the p-n junction layers are grown. Part (c) : Mesa is etched and metal contacts are deposited to form diode structures. Adjacent diodes lie on LEO GaN and on the window region. [65]

22

The sample was then returned to the MOCVD machine a final time and p-n junction layers were grown. The p-n junction consisted of 1 m unintentionally doped n-type GaN followed by 0.5 m Mg-doped p-type GaN. On standard planar films these growth conditions typically yield electron and hole concentrations of approximately 5 x 1016 cm-3 and 7 x 1017 cm-3, respectively. Diode devices were fabricated on this sample using the standard activation and p-contact processes, followed by a mesa etch using CI2 RIE and the deposition of a Ti/Al n-contact on the etched surface. The devices had dimensions of 2 m x 20m, and were placed so that they lay either on the LEO GaN or on the window region. In this way, p-n junction diodes were fabricated in close proximity on both the standard (highly dislocated) GaN in the window regions and on the low-dislocation density LEO GaN in the overgrown regions. The devices are sufficiently small that most of the diodes on LEO GaN contain no threading dislocations whatsoever; the devices on the window region are expected to contain approximately 200 threading dislocations on average. A photograph of the finished devices is shown in figure 11. [65]

Fig.11 Photograph of the finished devices. The top row of diodes are on LEO GaN; those in the bottom row are on the window region (dislocated GaN). The inset (lower right) shows the position of the overgrown GaN and the window region.

PHYSICS OF P-N JUNCTION DIODE: A pn junction is formed by joining P-type and N-type semiconductors together in very close contact. The term junction refers to the boundary interface where the two regions
23

of the semiconductor meet. If they were constructed of two separate pieces this would introduce a grain boundary, so pn junctions are created in a single crystal of semiconductor by doping, for example by ion implantation, diffusion of dopants, or by epitaxy (growing a layer of crystal doped with one type of dopant on top of a layer of crystal doped with another type of dopant). The following figure 12(a) indicates the symbolic representation of a P-N junction diode with two terminals: anode which is the positive terminal and cathode which is the negative terminal.

Fig.12 (a) Symbolic representation of a P-N junction diode

The figure 12(b) represents the physical representation of a P-N junction diode with n-substrate etched by p+ material and a metal contact which is the ohmic contact on the top of the p+ material acting as anode exists. And the cathode is taken from the bottom (n-sub).

Fig.12(b) Physical Structure of a typical P+N junction diode

24

PROPERTIES OF P-N JUNCTION: The pn junction possesses some interesting properties which have useful applications in modern electronics. A p-doped semiconductor is relatively conductive. The same is true of an n-doped semiconductor, but the junction between them can become depleted of charge carriers, and hence nonconductive, depending on the relative voltages of the two semiconductor regions. By manipulating this non-conductive layer, pn junctions are commonly used as diodes: circuit elements that allow a flow of electricity in one direction but not in the other (opposite) direction. This property is explained in terms of forward bias and reverse bias, where the term bias refers to an application of electric voltage to the p n junction. [65] UNDER EQUILIBRIUM (ZERO BIAS): In a pn junction, without an external applied voltage, an equilibrium condition is reached in which a potential difference is formed across the junction. This potential difference is called built-in potential (Vbi).

Fig.13 A pn junction in thermal equilibrium with zero bias voltage applied.

After joining p-type and n-type semiconductors, electrons near the pn interface tend to diffuse into the p region as illustrated in the figure 13. As electrons diffuse, they leave positively charged ions (donors) in the n region. Similarly, holes near the pn interface begin to diffuse into the n-type region leaving fixed ions (acceptors) with negative charge. The

25

regions nearby the pn interfaces lose their neutrality and become charged, forming the space charge region or depletion layer. The space charge region is a zone with a net charge provided by the fixed ions that have been left uncovered by majority carrier diffusion. When equilibrium is reached, the charge density is approximated by the displayed step function. In fact, the region is completely depleted of majority carriers (leaving a charge density equal to the net doping level), and the edge between the space charge region and the neutral region is quite sharp. The space charge region has the same magnitude of charge on both sides of the pn interfaces, thus it extends farther on the less doped side. The following figure 14 represents the energy band diagram of the p-n junction diode in thermal equilibrium with zero applied bias. [66]

Fig.14 Energy band diagram of a p-n junction in thermal equilibrium

To reach thermal equilibrium, electrons/holes close to the metallurgical junction diffuse across the junction into the p-type/n-type region where hardly any electrons/holes are present. This process leaves the ionized donors (acceptors) behind, creating a region around the junction, which is depleted of mobile carriers. We call this region the depletion region, extending from x = -xp to x = xn. The charge due to the ionized donors and acceptors causes an electric field, which in turn causes a drift of carriers in the opposite direction. The

26

diffusion of carriers continues until the drift current balances the diffusion current, thereby reaching thermal equilibrium as indicated by a constant Fermi energy. While in thermal equilibrium no external voltage is applied between the n-type and ptype material, there is an internal potential, which is caused by the workfunction difference between the n-type and p-type semiconductors. This potential is called as the builtin potential. [66] FORWARD BIAS: In forward bias, the p-type is connected with the positive terminal and the n-type is connected with the negative terminal which appears as in the following figure 15.

Fig.15 P-N junction diode in forward bias

With a battery connected this way, the holes in the P-type region and the electrons in the N-type region are pushed towards the junction. This reduces the width of the depletion zone. The positive charge applied to the P-type material repels the holes, while the negative charge applied to the N-type material repels the electrons. As the electrons and holes are pushed towards the junction, the distance between them decreases. This lowers the barrier in potential. With increasing forward-bias voltage, the depletion zone eventually becomes thin enough that the zone's electric field can't counteract charge carrier motion across the pn junction, consequently reducing electrical resistance. The electrons which cross the pn junction into the P-type material (or holes which cross into the N-type material) will diffuse in the near-neutral region. Therefore, the amount of minority diffusion in the near-neutral zones determines the amount of current that may flow through the diode. [66]

27

Only majority carriers (electrons in N-type material or holes in P-type) can flow through a semiconductor for a macroscopic length. With this in mind, consider the flow of electrons across the junction. The forward bias causes a force on the electrons pushing them from the N side toward the P side. With forward bias, the depletion region is narrow enough that electrons can cross the junction and inject into the P-type material. However, they do not continue to flow through the P-type material indefinitely, because it is energetically favorable for them to recombine with holes. The average length an electron travels through the P-type material before recombining is called the diffusion length, and it is typically on the order of microns. Although the electrons penetrate only a short distance into the P-type material, the electric current continues uninterrupted, because holes (the majority carriers) begin to flow in the opposite direction. The total current (the sum of the electron and hole currents) is constant in space, because any variation would cause charge buildup over time (this is Kirchhoff's current law). The flow of holes from the P-type region into the N-type region is exactly analogous to the flow of electrons from N to P (electrons and holes swap roles and the signs of all currents and voltages are reversed). [66] Therefore, the macroscopic picture of the current flow through the diode involves electrons flowing through the N-type region toward the junction, holes flowing through the Ptype region in the opposite direction toward the junction, and the two species of carriers constantly recombining in the vicinity of the junction. The electrons and holes travel in opposite directions, but they also have opposite charges, so the overall current is in the same direction on both sides of the diode, as required. The following figure 16 represents the energy band diagram of a p-n junction under forward bias.

Fig.16 Energy band diagram of a p-n junction under forward bias.

28

REVERSE BIAS: Reverse biased usually refers to how a diode is used in a circuit. If a diode is reverse biased, the voltage at the cathode is higher than that at the anode. Therefore, no current will flow until the diode breaks down. Connecting the P-type region to the negative terminal of the battery and the N-type region to the positive terminal, corresponds to reverse bias. The connections are illustrated in the following figure 17:

Fig.17 P-N junction diode in Reverse bias.

Because the p-type material is now connected to the negative terminal of the power supply, the 'holes' in the P-type material are pulled away from the junction, causing the width of the depletion zone to increase. Similarly, because the N-type region is connected to the positive terminal, the electrons will also be pulled away from the junction. Therefore the depletion region widens, and does so increasingly with increasing reverse-bias voltage. This increases the voltage barrier causing a high resistance to the flow of charge carriers thus allowing minimal electric current to cross the pn junction. The increase in resistance of the p-n junction results in the junction to behave as an insulator. This is important for radiation detection because if current were able to flow, the charged particles would just dissipate into the material. The reverse bias ensures that charged particles are able to make it to the detector system. The strength of the depletion zone electric field increases as the reverse-bias voltage increases. Once the electric field intensity increases beyond a critical level, the pn junction depletion zone breaks-down and current begins to flow, usually by either

the Zener or avalanche breakdown processes. Both of these breakdown processes are nondestructive and are reversible, so long as the amount of current flowing does not reach levels that cause the semiconductor material to overheat and cause thermal damage. [66]
29

This effect is used to one's advantage in zener diode regulator circuits. Zener diodes have a certain - low - breakdown voltage. A standard value for breakdown voltage is for instance 5.6V. This means that the voltage at the cathode can never be more than 5.6V higher than the voltage at the anode, because the diode will break down - and therefore conduct - if the voltage gets any higher. This effectively regulates the voltage over the diode. The following figure 18 represents the energy band diagram of a p-n junction under reverse bias.

Fig.18 Energy band diagram of a p-n junction under reverse bias.

DIFFUSION CURRENT: Diffusion current is a current in a semiconductor caused by the diffusion of charge carriers (holes and/or electrons) in the neutral region of the diode. Diffusion current can be in the same or opposite direction of a drift current and the carrier transport is occurred due the impurity gradient. At equilibrium in a p-n junction, the forward diffusion current in the depletion region is balanced with a reverse drift current, so that the net current is zero. The diffusion current density is given by

Jp(x xn) = - qDp Jn(x -xp) = q Dn

Where q electronic charge. (1.6 x 10-19 C)

30

Dp diffusion coefficient of holes ( Dn diffusion coefficient of electrons (

) )

variation of minority carrier holes concentration with respect to the space charge region distance(x). variation of minority carrier electrons concentration with respect to the space charge region distance(x). DRIFT CURRENT: Drift current is the electric current, or movement of charge carriers, which is due to the applied electric field in the depletion region, often stated as the force over a given distance. It is distinguished from thermally-induced diffusion current, which results from the random Brownian motion of charge carriers independent of electrical stimulus. If an electric field is applied to an electron existing in a free space, it will accelerate the electron in a straight line from the negative terminal to the positive terminal of the applied voltage. But same thing does not happen in the case of electrons available in good conductors. Good conductors have plenty of free electrons moving randomly in between the fixed positive ion cores. This random movement of electrons in a straight line is known as drift current. The diffusion current density is given by

J(-xp<x<xn) = qnnEx + qppEx

Where q electronic charge. (1.6 x 10-19 C)

n,p electron and hole concentration respectively (cm-3) n,p Mobility of electrons and holes respectively (cm2/V-s)
No bias: If no voltage is connected to p-type side, then drift and diffusion cancel: Diffusion = Drift Forward bias: If a positive voltage is connected to p-type side, then the internal field is reduced. Therefore more majority carriers will diffuse into the adjacent region. Drift and diffusion no longer cancel: Diffusion > Drift

31

Reverse bias: If a negative bias is connected to p-type side, then the internal electric field is enhanced. Drift will be stronger than diffusion. No majority carriers will contribute to current: Diode model: The diode equation relates the diode current I of a p-n junction diode to the diode voltage VD. This relationship is the diode I-V characteristic: Drift > Diffusion

Where IS is the saturation current or scale current of the diode (the magnitude of the current that flows for negative VD in excess of a few VT, typically 1012 A). The scale current is proportional to the diode area. Continuing with the symbols: VT is the thermal voltage (kT / q, about 26 mV at normal temperatures), and n is known as the diode ideality factor in the range of 2 to 5 for GaN. ANALYSIS OF ELECTRICAL CHARACTERISTICS OF GaN DIODE: Solving Poissons equation in the case of Non-uniformly doped GaN p-n diode: The general Poissons equation can be given as

(x) = -/ = d2 V/dx2
Where is the charge density,

------- (1)

is the permittivity.
At the same time it is known that,

(x) = -/ = d2 V/dx2 = -

[ ND NA ]

------- (2)

Where ND is the concentration of the donor atoms. NA is the concentration of the acceptor atoms(substrate concentration). ND =

exp[-(
32

)2]

----- (3)

Where Q is the Implant dose. is the straggle parameter. Rp is the range parameter. q is the electronic charge (1.6 x 10-19 C) Substituting equation(3) in equation(2) and integrating the resulting equation in order to get electric fieldE

E=

=-

exp [-(

)2]dx +

A.dx

----- (4)
Since it is non-uniform doping(

>> NA),the error function should be brought

up here. The genereal error function is given as,

Using this error function equation(4) can be integrated as,

E=-

[ erf(

) - erf(

) ] + c1 +

.NA.x + c2 ------(5)

For convenience the equation(5) can be split into two parts and we can find constants

c1 and c2.
Part(1):

E=-

[ erf(

) - erf(

) ] + c1

-----(6)

In case of non-uniform doping, at place of x = xn , E = 0.

c1 =
Part(2):

[ erf(

) - erf(

)]

------ (7)

33

E = .NA.x + c2
In case of non-uniform doping, at place of x = xp, E = 0.

------(8)

c1 = Now substituting equation (7) and (9) in (5),

------(9)

E=

[ erf(

) - erf(

) ] + .NA[x xp ]

------ (10)

Thus the above equation is the electric field equation with the function of the distance of the depletion region in terms of x, xn, xp. Now in order to find Xn, Under thermal equilibrium,

Xne =
Under forward bias,

Xnf =
Under reverse bias,

Xnr =
Where the Built-in Voltage(Vbi) at T=300k is given by,

Vbi = (0.0259) ln[

Now the integration of the equation(10) provides potential() which is,

34

=- (

)[ erf(

) - erf(

)]+

.[

- [

---- (11)

Thus this is the equation to analyze the potential variation with respect to the depletion region distance. In order to analyze the I-V characteristics of the non-uniformly doped GaN p-n diode, it is supposed to consider the following equation, The hole diffusion current is given by,

Jp =

[exp( ) - 1]
)

---- (12)

Where Dp diffusion constant of holes (

Lp diffusion length (cm) which is given by, Lp =

p Lifetime of holes (sec)

pno Minority electron concentration in the P side at thermal equilibrium given by Donor concentration is given by

Similarly The electron diffusion current is given by,

Jn =

[exp( ) - 1]

---- (13)

35

Where Dn diffusion constant of electrons (

Ln diffusion length (cm) which is given by, Ln =

n Lifetime of electrons (sec)

npo Minority hole concentration in the N side at thermal equilibrium given by

It is interesting to note that the hole current is due to injection of holes from the p-side to the n-side, but the magnitude is determined by the properties in the n-side only (Dp ,Lp, Pno,). The total current is given by the sum of equation(12) and (13).

J = Jn + Jp =[ = J = Js [exp( ) - 1]
Where Js Saturation current density. Then in order to get the current equation the cross-sectional area(A) of the semiconductor is multiplied with the above current density equation. Multiplying the equation(14) by the cross-sectional area(A), the resulting equation is,

[exp( ) - 1]] + [ [exp( ) - 1 ]

[exp( ) - 1]]

----- (14)

I=JA I = Js.A.[exp( ) - 1] ------ (15)

36

Thus this is the final current equation through which we can analyze the variation of current with respect to the applied diode voltage variation which will be very useful as it deals with non-uniform doping. And here a factor called as ideality factor(n) is included in the reaction as the electrical characteristics of a junction when the diode is in forward bias operation. The ideality factor(n) can be 1 or 2 or 3. The minority-carrier densities and the current densities for the forward-bias and reverse-bias conditions are shown and illustrated in the following Fig. 19.

(a)

(b)

Fig.19 (a)Carrier distributions and (b)current densities (both linear plots) for forward bias condition.

In the above figure in case of forward bias, the depletion region width is narrower and the minority carrier in the n-side which are holes gradually increases towards the center of the depletion region and the minority carrier in the p-side which are electrons gradually increases towards the center of the depletion region from the edges of the neutral regions. In the other figure the hole current density in the P-side decreases gradually towards the N-side and diminishes finally and in the P-side the minority electron current density is very low. Similarly the electron current density in the N-side decreases gradually towards the P-side and diminishes finally and in the N-side the minority hole current density is very low.

37

Fig.20 Carrier distributions and current densities (both linear plots) for reverse bias condition.

In the above figure 20 in case of reverse bias, the depletion region width is broader and the minority carrier in the n-side which are holes gradually decreases towards the center of the depletion region and the minority carrier in the p-side which are electrons gradually decreases towards the center of the depletion region from the edges of the neutral regions. In the other figure the hole current density in the P-side decreases gradually towards the N-side and diminishes finally and in the P-side the minority electron current density is very low. Similarly the electron current density in the N-side decreases gradually towards the P-side and diminishes finally and in the N-side the minority hole current density is very low. In reverse bias, the important diode figures-of-merit are leakage current and breakdown voltage. The breakdown behavior of GaN p-n diodes has been examined. The reported breakdown voltages are approximately equal to 150V consistent with a breakdown
38

field of approximately 2MV/cm. The leakage current in GaN due to minority carrier diffusion should be extremely small because of the very wide band-gpa. Much higher leakage currents are normally obtained, indicating the presence of trap states with in the gap. The resulting leakage paths may adversely impact the performance of both photodiodes and bipolar transistors;the trap states may also act as recombination centers in forward bias operation. For improved device operation, it is clearly important to identify and eliminate the defects which are responsible for the various trap states. A typical I-V characteristic is shown in figure 21 on a linear scale. The diodes exhibit a sharp turn-on at a voltage near band-gap. The reverse-bias breakdown behavior is quite soft, and occurs at voltages up to ~100V depending on the doping level in the n-type layer. The breakdown voltage is obviously of great importance for developing high-power bipolar transistors.

Fig.21 I-V Characteristic of a typical GaN p-n junction.

RESULTS AND DISCUSSIONS: Using MATLAB the electric field(E) and potential()(equation(10) and (11)) can be computed with respect to the distance of depletion region from the center of the semiconductor(x) using the different straggle parameters() and range parameters(Rp) for different Q values for different energy values in KeV. These values are calculated and notified from a software called SRIM (The Stopping and Range of Ions in Matter).

39

SRIM is a group of programs which calculate the straggling parameters and range of ions (up to 2 GeV/amu) for specific ion energy into matter using a quantum mechanical treatment of ion-atom collisions (assuming a moving atom as an "ion", and all target atoms as "atoms"). This calculation is made very efficient by the use of statistical algorithms which allow the ion to make jumps between calculated collisions and then averaging the collision results over the intervening gap. During the collisions, the ion and atom have a screened Coulomb collision, including exchange and correlation interactions between the overlapping electron shells. The ion has long range interactions creating electron excitations and plasmons within the target. These are described by including a description of the target's collective electronic structure and interatomic bond structure when the calculation is setup (tables of nominal values are supplied). The charge state of the ion within the target is described using the concept of effective charge, which includes a velocity dependent charge state and long range screening due to the collective electron sea of the target. Ion Implantation Energy Loss Mechanisms:

Fig.22 Mechanism of Ion Implantation Energy

Energy Loss and Ion Properties: Light ions/at higher energy more electronic stopping. Heavier ions/at lower energy more nuclear stopping.

40

Fig.23 Mechanism of Energy Loss and properties of ion

Nuclear Stopping: If the ion and the target atoms were bare nuclei, the scattering potential would be given by simple Coulomb potential. Because of electron which surround the target nucleus, the full effect of the positive core potential is screened from the incoming ion and is given by modified Thomas-Fermi model of the atom.

Where a is some screening distance.

Where m1, Z1 refer to the ion and m2,Z2= substrate atom mass and atomic number The energy loss of the incident ion per unit depends on the ion energy. The nuclear energy loss is small at very high energies, because fast particle have Less interaction time with the scattering nucleus. Thus the nuclear energy loss tends to dominant towards the end of the range when ion has lost much of its energy and where nuclear collisions produced most of the damage.

41

Fig.24 Nuclear stopping

Electron Stopping: Drag force caused by charged ion in "sea" of electrons (nonlocal electronic stopping). These collisions are inelastic and result in small energy losses in which the electron are excited or ejected from their shells and then dissipate their energy through thermal vibration of the target. Since electron stopping depends directly on the ion velocity, we can write

Where C and k are parameters depend on ion, the substrate.

Fig.25 Implantation disorder caused by (a) light ions and (b) heavy ions

42

Total stopping power: The total stopping power is composed of the sum of the nuclear and electronic stopping. The nuclear stopping dominates at low energies. Electronic stopping dominates at higher energies which is explained in the following figure 26.

Fig.26 Nuclear and electronic components of S(E) for several common Si dopants as a function of energy.

Electric field versus Depletion region distance: With a constant ion dose (Q) values for different energy values such as 50,75,100,150 KeV, different range parameter (Rp) values and different straggle parameter () values are determined from SRIM program and substituted into the electric field equation. The Figure 20 shows the plot of electric field versus depletion width obtained from equations (10) and (11) using MATLAB software as follows. The plot exhibits the maximum electric field at the junction and becomes minimum at the edge of depletion width at 0.49m, 0.22 m and 0.13 m for the ion dose of Q = 1x1011, 5x1011 and 1.5x1012cm-2 respectively. The plot also shows that the high ion dose gives rise of the impurity concentration reflected as high electric field as well as the depletion width is terminated early compared to the low ion dose.

43

Fig.27 Electric Field versus depletion region distance

The Figure 27 shows the plot of the electric field(E) versus depletion region distance calculated by using the following equation:

E=

[ erf(

) - erf(

) ] + .NA[

] ----(15)

Where = o s = 8.41 x 10-13 F/cm, the permittivity of the material.

o 8.854 x 10-12 F/cm, the permittivity of free space. s 9.498 F/cm, dielectric constant of GaN.
q electronic charge (1.6 x 10-19 C)
n distance of depletion region on n-side from the center(cm) p distance of depletion region on p-side from the center(cm)

NA Substrate concentration.(1 x 1015 cm-3) Q Implant dose. (1 x 1011 cm-2) straggle parameter.

44

Rp range parameter. For 50KeV, Rp1 = 783 x 10-8 cm, 1 = 146 x 10-8 cm. For 75KeV, Rp2 = 1104 x 10-8 cm, 2 = 238 x 10-8 cm. For 100KeV, Rp3 = 1428 x 10-8 cm, 3= 297 x 10-8 cm. For 150KeV, Rp4 = 2091 x 10-8 cm, 4 = 412 x 10-8 cm. Here in the resulting plot the electric field(E) is maximum near the center of the depletion region and becomes smaller at the edge of the space charge region. The low energy to high energy shows different junction depth, which reflects the space charge width. This concept is illustrated with various Rp, values and a constant ion dose(Q) value for various energy values. As the energy values increase, the electric field will be decreasing in a way that it is maximum at the center and exponentially falls down. Thus the electric field plot with respect to the depletion region distance(x) is illustrated as shown in the figure 27. Potential versus depletion region distance:

Fig.28 Potential () versus Space charge region distance

45

The Figure 28 demonstrates the plot of the potential() versus Space charge region distance (x) calculated by using the following equation:

=- (

)[ erf(

) - erf(

)]+

.[

- [
Where = o s = 8.41 x 10-13 F/cm, the permittivity of the material.

----(16)

o 8.854 x 10-12 F/cm, the permittivity of free space. s 9.498 F/cm, dielectric constant of GaN.
q electronic charge (1.6 x 10-19 C)
p distance of depletion region on p-side from the center(cm)

NA Substrate concentration.(1 x 1015 cm-3) straggle parameter. (For 50KeV, = 146 x 10-8 cm) Rp range parameter. (For 50KeV, Rp = 783 x 10-8 cm) Q Implant dose. For 50KeV, Q1 = 1 x 1011 cm-2 Q2 = 5 x 1011 cm-2 Q3 = 1.5 x 1011 cm-2 Similarly the figure 28 explains that the potential() is maximum at the physical junction depth as well as at the center of the depletion region. The potential slowly decreases exponentially with respect to the distance up to the edge of depletion and becomes almost zero at the beginning of the neutral zone. This plot has been generated using constant Rp, values for energy (50KeV) and various ion dose (Q) values. As the value of ion dose (Q) increases, the impurity concentration increases, which in turn increases the electric field and the potential. Thus the potential plot with respect to the depletion region distance(x) clearly
46

shows that the potential becomes almost zero at the edge of space charge width and neutral region. The space charge bounded area is extremely important because electric field and potential for carrier transport by drift mechanism. I-V characteristics of GaN p-n junction diode under forward bias: The Figure 29 shows I-V characteristics of GaN diode under forward bias condition. The built-in voltage found from the plot is approximately in the range of 3.10 3.25V, which exactly agrees well with physical GaN p-n junction diode and the ideality factor lies between 1.5 1.7. The forward resistance has been estimated from the slope and found to be 0.666. The built-in voltage becomes high because the intrinsic carrier concentration is extremely low. Hence, this characteristics shows that the GaN p-n junction can be used as power device for high voltage rectification.

Fig.29 I-V characteristics of GaN diode under forward bias.

I-V characteristics of GaN p-n junction diode under reverse bias: The Figure 30 shows the plot if I-V characteristics of GaN p-n junction diode under reverse bias. The diffusion current (reverse saturation current) is extremely low in the range of 10-12-10-14A and hence the leakage current is extremely low due to the wide-band semiconductor properties. observed. Due to low leakage current, the premature break was not

The exact breakdown was obtained at approximately 30V with substantial

47

increment of current. Compared to other SiC and GaN devices, the breakdown voltage showed a low value, because of low impurity dose, low junction depth and space charge curvature in the bulk and the ratio of active doping concentration and substrate concentration was not properly matched.

Fig.30 I-V characteristics of GaN diode under reverse bias.

The Figure 29 and figure 30 shows the plot of the current (I) versus voltage(V) (I-V characteristics) of a GaN p-n junction diode calculated by using the following equation:

I=JxA
Where A Area of cross section of the material (1 x 10-8cm2)

-----(17)

J current density (A/cm2) which is calculated using, J = Js [exp( ) - 1]

Where q electronic charge (1.6 x 10-19 C) K Boltzmanns constant (1.38 x 10-23J/K) T Temperature (300K)

------(18)

48

Js Saturation current density which is given by Js =

Where equilibrium given by

------(19)

Minority hole concentration in the N side at thermal

.
Minority electron concentration in the P side at thermal

equilibrium given by

. ni intrinsic carrier concentration for GaN (4.6 x 10-11cm-3 )

NA Substrate concentration.(1 x 1015 cm-3) ND Donor concentration which is given by,

exp[-(

)2]

Dn diffusion constant of electrons ( Dp diffusion constant of holes ( )

n Lifetime of electrons (5x10-5sec) p Lifetime of holes (1x10-5sec)

Ln diffusion length (cm) which is given by, Ln = Lp diffusion length (cm) which is given by, Lp =

49

From the current-voltage relation for non-uniformly doped GaN P-N junction diode, an empirical parameter has been incorporated to some possible values in order to manipulate both the forward and reverse bias plots to get the curve to be turned up at built-in voltage (Vbi) and the breakdown (Vbd) to be occurred in case of reverse bias. In forward bias condition, the current will remain zero unless the forward bias voltage reaches the built-in voltage (Vbi) which is 3.1V and then immediately the diode current starts rising exponentially as shown in the figure 29. The MATLAB code for each and every analysis is given at the end of the report. And thus the analytical modeling of Gallium Nitride (GaN) P-N junction diode is simulated successfully and the results are discussed legibly.

50

REFERENCE: [1] Z. Bandic, P. M. Bridger, E. C. Piqette, T. C. McGill, R. P. Vaudo, V. M. Phanse, and J. M. Redwing, High voltage (450 V) GaN Schottky rectifiers , Appl. Phys. Lett. 74 (1999) 1266 [2] P. Zhang, J. W. Johnson, B. Luo, F. Ren, S. J. Pearton, S. S. Park, Y. J. Park, J. -I.Chyi, Vertical and lateral GaN rectifiers on freestanding GaN substrates , Appl. Phys. Lett. 79 (2001) 1555 [3] P. Kozodoy, J. P. Ibbetson, H. Marchand, P. T. Fini, S. Keller, J. S.Speck, S. P. DenBaars, and U. K. Mishra, Electrical characterization of GaN p-n junctions with and without threading dislocations , Appl. Phys. Lett. 73 (1998) 975 [4] J. W. P. Hsu, M. J. Manfra, D. V. Lang, S. Richter, S. N. G. Chu, A.M. Sergent, R. N. Kleiman , L. N. Pfeiffer, R. J. Molnar,Inhomogeneous spatial distribution of reverse bias leakage in GaN Schottky diodes , Appl. Phys. Lett. 78 (2001) 1685. [5] Emerging Gallium Nitride Based Devices s. N. Mohammad, arnel a. Salvador, and hadis morkoc, fellow, ieee. [6] S.C. Binari, L.B. Rowland, W. Kruppa, G. Kelner, K. Doverspike and D.K. Gaskill, Electron. Lett. 30, 1248 (1994).

[7) H. Akagi, IEEE Trans. Power Electron. 13, 345 ~1998!

[8] G. T. Heydt and B. J. Skromme, Mater. Res. Soc. Symp. Proc. 483, 3 ~1998.

[9] E. R. Brown, Solid-State Electron. 43, 1918 ~1998.

[10] Power Semiconductor Devices, edited by S. J. Baliga ~PWS, Boston, 1996. [11] AlGaN/GaN HEMTs: An overview of device operation and applications U.K. Mishra,
P. Parikh, Y.F. Wu

[12] Y.-F. Wu et al., 1999 IEDM, Technical Digist, 925-927

51

[13] RF Micro Devices(R) Expands High Power GaN Product Portfolio, Highly-Efficient 140-Watt Transistor Release. [14] Dingle, D., K.L. Shaklee, R.F. Leheny, and R.B. Zetterstrom. 1994. Appl. Phys. Lett. 64:1377. [15] Kozodoy, P., J. P. Ibbetson, H. Marchand, P. T. Fini, S. Keller, S. P. DenBaars, J. S. Speck, U. K. Mishra.1998. Appl.Phys. Lett. 73:975. [16] Freitas, J. A., Jr., O. H. Nam, R. F. Davis, G. V. Saparin, and S. K. Obyden. 1998. Appl. Phys. Lett. 72:2990 [17] J. C. Zolper, A review of junction field effect transistors for high temperature and highpower electronics , Solid-State Electron. 42 (1998) 2153 [18] Y. Yoshizumi, S. Hashimoto, T. Tanabe and M. Kiyama, Highbreakdown-voltage pnjunction diodes on GaN substrates , Proc.13th ICMOVPE, We-P.69, (2006) 445 [19] I. Akasaki, H. Amano, Y. Koide, H. Hiramatsu, and N. Sawaki, J. Cryst. Growth 89, 209 (1989) [20)] S. Nakamura, Jpn. J. Appl. Phys. 30, L1705 (1991) [21] X.A. Cao, H. Lu, S.F. LeBoeuf, C.Cowen and S.D. Arthur,Growth and characterization of GaN PiN rectifiers on free-standing GaN , Appl. Phys. Lett. 87 (2005) 053503 [22] S. Nakamura, T. Mukai , M. Senoh, Appl.Phys Lett. 64 ,1687 (1994) [23] J.I. Pankove, J. Electrochem. Soc., 119, 1118 (1972) [24] N. Nakamura, T. Mukai, M. Senoh; Appl. Phys. Lett. 64 , 1687(1994) [25] I. Akasaki, H. Amano, H. Murakami, M. Sassa, H. Kato, K. Manabe; J. Cryst. Growth 128,379, (1993) [26] T. W. Weeks, Jr., M.D. Bremser, K.S. Ailey, E. Carlson et al., Appl. Phys. Lett. 67, 401 (1995) [27] M.A. Kahn, Q. Chen, R.A. Skogman, J.N. Kuznia; Appl. Phys. Lett. 66,2046,(1995)

52

[28] S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matshushita, H. Kiyoku, and Y.Sugimoto, Jpn. J. Appl. Phys. Vol 35, L74-L76 (1996) [29] I. Akasaki, H. Amano, Y. Koide, K. Hiramatsu, and N. Sawaki, J. Cryst. Growth, 98, 209 (1989) [30] S. Nakamura, M. Senoh, T. Mukai,Jpn. J. Appl. Phys., Vol 32, pp.L8-11 (1993) [31] K. Hirosawa, K. Hiramatsu, N. Sawaki, and I. Akasaki, Jpn. J. Appl. Phys. 32, L1030 (1993) [32] M.A. Kahn, J.M. Van Hove, J.N. Kuznia, and D.T. Olson, Appl. Phys. Lett. 58, 2408 (1991) [33] Y.-F. Wu, B.P. Keller, S. Keller, D. Kapolnek, P. Kozodoy, S.P. DenBaars and U.K.Mishra, Appl.Phys. Lett. 69, 1438 (1996) [34] M.A. Kahn, R. A. Skogman, and J.M. van Hove, S. Krishnan kutty, and R.M. Kolbas, Appl. Phys.Lett. 56, 1257 (1990) [35] S. Krishnankutty and R.M. Kolbas, M. A. Kahn, J. N. Kuznia, J.M. van Hove and D.T.Olson, J.Electron. Mat. 21, 437 (1992) [36] I. Arslan and N. D. Browning, Role of Oxygen at Screw Dislocations in GaN , Phys. Rev. Lett. 91, (2003) 165501 [37] A. Saxler, P. Kung, C.J. Sun,E. Bigan, and M. Razeghi, Appl. Phys. Lett. 64 , 339 (1994) [38] M. A. Kahn, J. N. Kuznia, and D. T. Olson, T. George and W. T. Pike, Appl. Phys. Lett. 63, 3470 (1993) [39] T. Matsuoka, H. Tanaka, T. Sasaki and A. Katsui, Inst. Phys. Conf. Ser. No 106, GasAs and related compounds , chapter 3, p. 141(1988) [40] A. Koukitu, N. Tkahashi, T. Taki, and H. Seki, Jpn. J. Appl. Phys. 35, L673 (1996) [41] M. Shimizu, K. Hiramatsu and N. Sawaki, J. Cryst. Growth 145,209 (1994) [42] E. Litwin-Staszewska, T. Suski, I. Grzegory, S. Porowski, J. Robert, D. Wasik, A. Witowski,D.Cote, and B. Clerjund, Electrical Properties of GaN Bulk Single Crsytals Doped
53

with Mg Paper We-P082,Proceedings of Intl. Conf. on Nitride Semiconductors, Montpelliar, France (1999), pg 124.) [43] K. Kornitzer, T. Ebner, M. Grehl, K. Thonke, R. Sauer, C. Kirchner, V. Schwelger, M. Kamp, M.Leszczynski, I. Grzergory, S. Porowski, High Resolution PL and Refelectance Spectra of Homoepitaxial GaN Layers Paper Tu-17, Proceedings of Intl. Conf. on Nitride Semiconductors, Montpelliar, France (1999), pg 49 [44] T.S. Zheleva, O.-H. Nam, M.D. Bremser, and R.F. Davis, Appl. Phys. Lett. 71, 2472 (1997). [45)] A. Sakai, H. Sunakawa, and A. Usui, Appl. Phys. Lett. 71, 2259 (1997). [46] H. Marchand, J.P. Ibbetson, P.T. Fini, P. Kozodoy, S. Keller, J.S. Speck, S.P. DenBaars, and U.K.Mishra, MRS Internet J. Nitride Semicond. Res. 3, 3 (1998). [47] H. Marchand, X.-H. Wu, J.P. Ibbetson, P.T. Fini, P. Kozodoy, S. Keller, J.S. Speck, S.P. DenBaars,and U.K. Mishra, Appl. Phys. Lett. 73, 747 (1998). [48] S. Chichibu, H. Marchand, S. Keller, P. Fini, J.P. Ibbetson, M. Minsky, S. Fleischer, J.S. Speck, J.Bowers, E. Hu, U.K. Mishra, S.P. DenBaars, T. Deguchi, T. Sota, and S. Nakamura, 2nd Int. Conf. on Blue Laser and Light Emitting Diodes, Chiba, Japan, Sept. 29 Oct. 2, 1998; S. Chichibu, H.Marchand, M.Minsky, S. Keller, P. Fini, J. Ibbetson, S. Fleisher, J.S. Speck, J.E. Bowers, E. Hu, U.K. Mishra, S.P. DenBaars, T. Deguchi, T. Sota, and S. Nakamura, Appl. Phys. Lett. (in press). [49] See also J.A. Freitas, Jr., O.-H. Nam, R.F. Davis, G.V. Saparin, and S.K. Obyden, Appl. Phys. Lett .72, 2990 (1998); X. Li, S.G. Bishop, and J.J Coleman, Appl. Phys. Lett. 73, 1179 (1998). [50] S.J. Rosner, G. Girolami. H. Marchand, P.T. Fini, J.P. Ibbetson, L. Zhao, S. Keller, U.K. Mishra .S.P. DenBaars, J.S. Speck, Appl. Phys. Lett. (in press). [51] P. Kozodoy, J.P. Ibbetson, H. Marchand, P.T. Fini, S. Keller, S.P. DenBaars, J.S. Speck, U.K.Mishra, Appl. Phys. Lett. 73 975 (1998). [52] (T. Mukai, K. Takekawa, S. Nakamura, Jpn. J. Appl. Phys. 37 L839 (1998 .

54

[53] C. Sasaoka, H. Sumakawa, A. Kimura, M. Nido, A. Usui, and A. Sakai, J. Cryst. Growth 189, 61 .(1998). [54] R. Vetury, H. Marchand, J.P. Ibbetson, P.T. Fini, S. Keller, J.S. Speck, S.P. Denbaars, and U.K .Mishra, 25th Int. Symp. on Compound Semicond., Nara, Japan, Oct 12-16, 1998. [55] G. Parish, S. Keller, P. Kozodoy, J.P. Ibbetson, H. Marchand, P.T. Fini, S.B. Fleischer .S.P. DenBaars, U.K. Mishra, E.J. Tarsa, 1998 Conference on Optoelectronic and Microelectronic Materials And Devices, Perth, Australia, 14-16 Dec. 1998. [56] S. Nakamura, M. Senoh, S. Nagahama, N. Isawa, T. Yamada, T. Matsushita, H. Kiyoku, Y.Sugimoto, T. Kozaki, H. Umemoto, M. Sano, M. Chocho, Appl. Phys. Lett. 72, 211 (1998). [57] J. Neugebauer and C. G. Van de Walle, Role of hydrogen in doping of GaN. Appl.Phys. Lett. 68, 1829-31 (1996). [58] B. S. Simpkins, E. T. Yu, P. Waltereit and J. S. Speck, Correlated scanning Kelvin probe and conductive atomic force microscopy studies of dislocations in gallium nitride , J. Appl. Phys. 94 (2003) 1448. [59] K. Motoki, T. Okahisa, N. Matsumoto, M. Matsushima, H.Kimura, H. Kasai, K. Takemoto, K. Uematsu, T. Hirano, M.Nakayama, S. Nakahata, M.Ueno, D. Hara, Y. Kumagai, A.Koukitu and H. Seki, Preparation of Large Freestanding GaN Substrates by Hydride Vapor Phase Epitaxy Using GaAs as a Starting Substrate ,Jpn. J. Appl. Phys., 40 (2001) L140 [60] S. Hashimoto, Y.Yoshizumi, T.Tanabe and M. Kiyama, High-purity GaN epitaxial layers for power devices on low-dislocation-density GaN substrates , Proc. 13th ICMOVPE, Th-B1.4, (2006) 530 [61] M. Razeghi and M. Henini, Optoelectronic Devices:III-Nitrides (Elsevier, Oxford, 2005), Chapter 12 [62] A. F. Wright, Substitutional and interstitial carbon in wurtzite GaN ,J. Appl. Phys. 92 (2002) 2575 [63] Lei, T., M. Fanciulli, R. J. Molnar, T. D. Moustakas, R. J. Graham, and J. D. Scanlon. 1991. Appl. Physics Lett. 59:94.
55

[64] C.P. Kou, . R.M. Flecther, T.D. Ostenowski, M.C. Lardizabal, M.G. Craford, and V.M. Robbins, Appl. Phys. Lett., 57, 2937,(1990) [65] S. Krishnankutty and R.M. Kolbas, M. A. Kahn, J. N. Kuznia, J.M. van Hove and D.T. Olson, J.Electron. Mat. 21, 609 (1992) [66] Physics of semiconductor by B. Van Zeghbroeck, 2007.

56

CODING PART: CODE 1: For calculating and analyzing the electric field variation with respect to the depletion region distance from the center of the diode with various Q values:
Clc; Clear; Na=1e15; Xp=8e-6; X=0:.0000000001:9.817e-6; eps=8.41e-13; q=1.6021e-19; ni=4.6e-11;

-----------------------------------------------------------------------------------% with 50Kev

Q1=1e11; Rp1=783e-8; sigma1=146e-8; Nd1=Q1/(sigma1*sqrt(2*pi)); Vbi1=0.0259*log((Na*Nd1)/(ni*ni)); Xn1=sqrt((2*eps*Vbi1)/(q*Nd1)); z21=(Xn1-Rp1)/(sigma1*sqrt(2)); z31=(X-Rp1)/(sigma1*sqrt(2)); z41=((q*Na)*(X-Xp))/(eps);
57

z11=((q*Q1)/(2*eps)); E1=z11*(erf(z21)-erf(z31))+z41;

-----------------------------------------------------------------------------------% with 75Kev

Q2=1e11; Rp2=1104e-8; sigma2=238e-8; Nd2=Q2/(sigma2*sqrt(2*pi)); Vbi2=0.0259*log((Na*Nd2)/(ni*ni)); Xn2=sqrt((2*eps*Vbi2)/(q*Nd2)); z22=(Xn2-Rp2)/(sigma2*sqrt(2)); z32=(X-Rp2)/(sigma2*sqrt(2)); z42=((q*Na)*(X-Xp))/(eps); z12=((q*Q2)/(2*eps)); E2=z12*(erf(z22)-erf(z32))+z42;

-----------------------------------------------------------------------------------% with 100Kev

Q3=1e11; Rp3=1428e-8; sigma3=297e-8; Nd3=Q3/(sigma3*sqrt(2*pi)); Vbi3=0.0259*log((Na*Nd3)/(ni*ni));

58

Xn3=sqrt((2*eps*Vbi3)/(q*Nd3)); z23=(Xn3-Rp3)/(sigma3*sqrt(2)); z33=(X-Rp3)/(sigma3*sqrt(2)); z43=((q*Na)*(X-Xp))/(eps); z13=((q*Q3)/(2*eps)); E3=z13*(erf(z23)-erf(z33))+z43;

-----------------------------------------------------------------------------------% with 150Kev

Q4=1e11; Rp4=2091e-8; sigma4=412e-8; Nd4=Q4/(sigma4*sqrt(2*pi)); Vbi4=0.0259*log((Na*Nd4)/(ni*ni)); Xn4=sqrt((2*eps*Vbi4)/(q*Nd4)); z24=(Xn4-Rp4)/(sigma4*sqrt(2)); z34=(X-Rp4)/(sigma4*sqrt(2)); z44=((q*Na)*(X-Xp))/(eps); z14=((q*Q4)/(2*eps)); E4=z14*(erf(z24)-erf(z34))+z44; p=plot(X,E1,X,E2,X,E3,X,E4);

59

CODE 2: For calculating and analyzing the potential variation with respect to the depletion region distance from the center of the diode with various Q values:
Na=1e15; Xp=8e-6; X=0:.000000000001:11e-6; X1=6e-6; eps=8.41e-13; q=1.6021e-19; ni=4.6e-11; Rp=783e-8; sigma=146e-8; % with 50Kev and Q1=1e11 Q1=1e11; Nd1=(Q1/(sigma1*sqrt(2*pi))); Vbi1=(0.0259)*(log((Na*Nd1)/(ni*ni))); Xn1=sqrt((2*eps*Vbi1)/(q*Nd1)); z21=(X-Rp)/(sigma*sqrt(2)); z31=(Xp-Rp)/(sigma*sqrt(2)); z41=((q*Na)*(X-Xp)*(X1-Xp))/(eps); z11=((q*Q1)/(2*eps))*(X1-Rp1); z511=(q*Q1*sigma)/(eps*(sqrt(2*pi))); z512=(exp((-(X-Rp)*(X1-Rp))/(2*sigma*sigma))); z513=(exp((-(Xp-Rp)*(Xp-Rp))/(2*sigma*sigma)));

60

z51=z511*(z512-z513); O1=(z11*(erf(z21)-erf(z31))+z41+z51); % with 50Kev and Q2=5e11 Q2=5e11; Nd2=(Q2/(sigma*sqrt(2*pi))); Vbi2=(0.0259)*(log((Na*Nd2)/(ni*ni))); Xn2=sqrt((2*eps*Vbi2)/(q*Nd2));

z41=((q*Na)*(X-Xp)*(X1-Xp))/(eps); z12=((q*Q2)/(2*eps))*(X1-Rp); z521=((q*Q2*sigma)/(eps*(sqrt(2*pi)))); z522=exp((-(X-Rp)*(X1-Rp))/(2*sigma*sigma)); z513=exp((-(Xp-Rp)*(Xp-Rp))/(2*sigma*sigma)); z52=z521*(z522-z513); O2=z12*(erf(z21)-erf(z31))+z41+z52; % with 50Kev and Q3=1.5e12 Q3=1e12; Nd3=(Q3/(sigma*sqrt(2*pi))); Vbi3=(0.0259)*(log((Na*Nd3)/(ni*ni))); Xn3=sqrt((2*eps*Vbi3)/(q*Nd3)); z21=((X-Rp)/(sigma*sqrt(2))); z31=((Xp-Rp)/(sigma*sqrt(2))); z41=((q*Na)*(X-Xp)*(X1-Xp))/(eps); z13=((q*Q3)/(2*eps))*(X1-Rp); z531=(q*Q3*sigma)/(eps*(sqrt(2*pi))); z532=exp((-(X-Rp)*(X1-Rp))/(2*sigma*sigma));
61

z513=exp((-(Xp-Rp)*(Xp-Rp))/(2*sigma*sigma)); z53=z531*(z532-z513); O3=z13*(erf(z21)-erf(z31))+z41+z51; p=plot(X,O1,X,O2,X,O3);

CODE 3: Current-Voltage characteristics of GaN P-N junction diode in Forward Bias condition.
Dp=10; tp=1e-5; Lp=sqrt(Dp*tp); ni=4.6e-11; q=1.6e-19; Dn=25; tn=5e-5; Ln=sqrt(Dn*tn); Na=1e15; np0=(ni*ni)/Na; A=1e-8; V=0:.0001:10; Q=1e11; sigma1=131e-8; Nd=Q/(sigma1*sqrt(2*pi));

62

pn0=(ni*ni)/Nd; Jp=((q*Dp*pn0)/Lp)*(exp(V/(2*13*0.0259))-1); Jn=((q*Dn*np0)/Ln)*(exp(V/(2*13*0.0259))-1); J=Jn+Jp; I=J/A; p=plot(V,I);

CODE 4: Current-Voltage characteristics of GaN P-N junction diode in Reverse Bias condition.
q=1.6e-19; Na=1e15; Nd=5e17; ni=4.6e-11; t0=3e-9; Q=1e11; eps=11.9*8.854e-14; Vbi=.0259*log((Nd*Na)/(ni*ni)); V=-30:0.0001:0; Vbd=30.1106; Dn=25; Dp=10; Ln=sqrt(Dn*t0); Lp=sqrt(Dp*t0); J1=q*ni*(((Dn)/(Ln*Na))+((Dp)/(Nd*Lp)));
63

W=sqrt(((2*eps*(Vbi-V)/(q*Na)))); Jgen=(q*ni*W)/(2*t0); J2=exp((Vbi-V)/(2*11*0.0259)); Js=J1*J2; Jr=-Js+Jgen; I=Jr; p=plot(V,I);

64

65