Citation
Design, Fabrication and Characterization of Compound Semiconductors for Electronic and Photonic Devices

Material Information

Title:
Design, Fabrication and Characterization of Compound Semiconductors for Electronic and Photonic Devices
Creator:
JANG, SOOHWAN ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Doping ( jstor )
Electric current ( jstor )
Electric potential ( jstor )
Electrons ( jstor )
Indium ( jstor )
Ions ( jstor )
Oxides ( jstor )
Photometers ( jstor )
Quantum efficiency ( jstor )
Semiconductors ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Soohwan Jang. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2008
Resource Identifier:
659814641 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

1 DESIGN, FABRICATION AND CHAR ACTERIZATION OF COMPOUND SEMICONDUCTORS FOR ELECTR ONIC AND PHOTONIC DEVICES By SOOHWAN JANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Soohwan Jang

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGMENTS The accomplishment in Gainesville could not be done without my advisor, Fan Ren. First, I would like to thank him for guiding me with his tremendous experience and knowledge. His academic and personal advices have enabled me to finish this work and will be a pillar in my future life. I am also indebted to my other advisory committee members: drs. Stephen Pearton, Brent Gila, and Timothy Anderson. I am honored to have been advised by such an eminent committee. I appreciate Dr. Pearton for weekly m eetings while Dr. Ren was in hospital. It was my good luck to have a chance to deal with Dr. Gila’s novel oxide and to work with him. Of particular importance to this work are my colleagues and co-workers in the Chemical Engineering Department: Dr. Byoung Sam Kang , Hungta Wang, Travis Anderson, and Wei-Lun Min. I will not forget the memories from Trav is’s party. Thanks are also extended to Dennis Vince, James Hinnant, drs. Suku Kim, Jihyun Kim, Rishabh Mahandru , Jeff LaRoche, and Irokawa Yoshihiro. Many thanks go to Rohit Khanna, Mark Hald, Jerry, Lars Voss, John Wright, Wantae Lim, Luc Stafford, and Andy Gerger in the Materi als Science and Engineering Department for growing films and ICP etching, Dr . Lin and Sangwon Ko in electr ical and computer engineering department for their theoretical advice, and Ivan Kravchenko and Bill Lewis in UF Nanofab for their assistance and advice about their equipment. I enjoyed the time as an intern with drs. Hock Ng and Brijesh Vyas at Bell Labs in New Jersey. Mt deepest gratitude goes to Dr. Ng for giving me the chance to work in a new environment. Also, I appreciate drs. Paul Shen in Army research labs and Nuri Emanetoglu at the University of Maine for their great help for RF and optical measurements.

PAGE 5

5 Really, I am indebted to my family. The physical distance between them and me has not taken away their unsparing support to me. They have always respected my decisions and this has given me great confidence. A great love and appreciati on goes to my wife Mira.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .........9 ABSTRACT....................................................................................................................... ............12 1 INTRODUCTION..................................................................................................................14 1.1. Motivation and Literature Review...................................................................................14 1.2. Background................................................................................................................ ......17 1.2.1. Ion Implantation....................................................................................................17 1.2.2. Diffusion Process...................................................................................................19 1.2.3. Transfer Length Method........................................................................................22 1.2.4. Secondary Ion Mass spectrometry.........................................................................23 1.2.5. Auger Electron Spectroscopy................................................................................24 1.2.6. Atomic Force Microscopy.....................................................................................26 1.3. Dissertation Outline...................................................................................................... ...27 2 ENHANCEMENT-MODE GALLIUM NITRIDE METAL OXIDE SEMICONDUCTOR FIELD EFFECT TRANS ISTER WITH SILICON-DIFFUSION INTO SOURCE AND DRAIN REGIONS............................................................................43 2.1. Introduction.............................................................................................................. ........43 2.2. Silicon Diffusion into Gallium Nitride............................................................................44 2.3. Silicon Diffusion and Contact Properties........................................................................45 2.4. Magnesium Oxide/Gallium Nitride Meta l Oxide Semiconductor Field Effect transistor Using Silicon-Di ffused Source/Drain Regions...................................................46 3 ZINC OXIDE LIGHT-EMITTING DIODE..........................................................................56 3.1. Introduction.............................................................................................................. ........56 3.2. Simulation of Vertical and Latera l Zinc Oxide Light-Emitting Diodes..........................58 3.2.1. ISE TCAD Simulation...........................................................................................58 3.2.2. Comparison of Vertical and Late ral Zinc Oxide Light-emitting Diode................59 3.3. Formation of p-n Homojunctions in n-Zinc Oxide Bulk Single Crystals by Diffusion from Zinc Phosphide (Zn3P2) Source.................................................................61 3.3.1. Diffusion from Zinc Phosphide (Zn3P2) Source into Bulk Zinc Oxide Substrate...................................................................................................................... .61 3.3.2. Phospher Diffused Zinc Oxide p-n Diode.............................................................61 4 INDIUM GALLIUM ARSENIDE-BASED METAL SEMICONDUCTOR METAL OPTOELECTRONIC MIXER...............................................................................................77

PAGE 7

7 4.1. Introduction.............................................................................................................. ........77 4.2 Characterization of Indium Gallium Ar senide Self-Mixing Detectors for Chirp, Amplitude-Modulated LADAR (CAML)...........................................................................79 4.2.1 Indium Gallium Arsenide-based Metal Semiconductor Metal Photodetector Fabrication...................................................................................................................79 4.2.2 DC Response..........................................................................................................80 4.2.3. Mixing Responsivity.............................................................................................80 4.3. Metal Oxide Metal Photodetector with Tran sparent Indium Tin Ox ide Interdigitated Schottky Contact............................................................................................................... ..81 4.3.1. Deposition of Transparent Indium Tin Oxide.......................................................81 4.3.2. Characterization of Transparent Indi um Tin Oxide Films and Indium Tin Oxide Schottky Contact Metal Semi conductor Metal Photodetectors........................82 4.4. Design of Transparent Indium Tin Oxidebased Interdigitated Fingers for Metal Semiconductor Metal Photodetector...................................................................................85 4.4.1. Fabrication of Indium Gallium Arse nide-based Metal Semiconductor Metal Photodetector with Transparent Indium Tin Oxide Electrodes...................................85 4.4.2. Design of Transparent Interdigitated Indium Ti n Oxide Contact Metal Semiconductor Metal Photodetector............................................................................86 5 SUMMARY AND FUTUREWORK...................................................................................108 5.1. Gallium Nitride Enhancement Mode Me tal Oxide Semiconductor Field Effect Trainsiter..................................................................................................................... ......108 5.2. Zinc Oxide Light-Emitting Diode.................................................................................108 5.3. Indium Gallium Arsenide Metal Semiconductor Metal Photodetector.........................110 LIST OF REFERENCES.............................................................................................................112 BIOGRAPHICAL SKETCH.......................................................................................................123

PAGE 8

8 LIST OF TABLES Table page 1-1 Material properties of ZnO and GaN.................................................................................284-1 Sheet resistance, resistivity, transmittance at 1.5 m, and index of refraction of different stack samples.......................................................................................................89

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Intrinsic carrier concentr ation according to temperature of Si, GaAs, and GaN...............29 1-2 Bandgap energy and lattice constant of GaN and ZnO......................................................30 1-3 Schematic diagram of an ion implanter.............................................................................31 1-4 Coordinate of implanted ions.............................................................................................32 1-5 Projected range of Si in GaN as a function of implantation energy..................................33 1-6 Projected straggle and transv erse straggle of Si in GaN as a function of implantation energy......................................................................................................................... ........34 1-7 Distribution of Si in GaN with implantation energy of 350 KeV......................................35 1-8 Nitrogen, gallium, and total vacancies in GaN after 350 KeV Si implantation.................36 1-9 Lateral current flow geometry...........................................................................................37 1-10 Transfer Length Method. A) Transfer length method test structure. B) Plot of total resistance as a function of contact spacing........................................................................38 1-11 Schematic diagram of sec ondary ion mass spectrometry..................................................39 1-12 Emission of Auger electron...............................................................................................40 1-13 Schematic diagram of A uger electron sp ectroscopy..........................................................41 1-14 Schematic diagram of at omic force microscope................................................................42 2-1 Schematic of structure for Si diffusion in GaN.................................................................48 2-2 Optical micrographs of surfaces after Si diffusion and subsequent removal of the SiO2 or SiNx encapsulant and remaining Si layer..............................................................49 2-3 SIMS profiles of Si diffused in GaN at 800, 900 or 1000C for 120 mins........................50 2-4 The AFM surface scans of control GaN a nd surfaces after diffusion of Si at 800 or 900C and subsequent removal of the SiO2 and remaining Si............................................51 2-5 Specific contact resistivity of Ti/Al/Pt/A u contacts on the Si-diffused regions. A) Diffusion temperature dependen ce. B) Si thickness dependence for a fixed diffusion temperature of 900C.........................................................................................................52

PAGE 10

10 2-6 The Si diffused enhancement mode metal oxide semiconductor field effect transistor. A) Schematic cross-section of the MgO/ p-GaN MOSFET. B) Microscope image of the device layout.............................................................................................................. ..53 2-7 The IDS-VDS characteristics from the MgO/GaN-on-Si MOSFET. The gate voltage was changed from 4V in steps of -1 V...............................................................................54 2-8 Transfer characteristics of MgO/GaN-onSi MOSFET at a drain-source voltage of 7V............................................................................................................................. ..........55 3-1 Schematic of vertical geometry ZnO LED used in the simulations...................................64 3-2 Schematic of lateral geometry ZnO LED used in the simulations.....................................65 3-3 Conduction and valence band, and radiativ e recombination of vertical device................66 3-4 Total current density distribution at 5V in the vertical structure.......................................67 3-5 Total current density distribution at 5V in the lateral structure.........................................68 3-6 Active layer thickness dependence at 5V and peak wavelength for both LED geometries. A) Output power . B) Optical intensity...........................................................69 3-7 Simulated electroluminescence spectra as a function of doping level in the p-ZnO layer. A) Vertical geometry structur e. B) Lateral geometry structure...............................70 3-8 p-layer doping dependence at 5V and peak wavelength. A) Output power. B) Optical intensity...................................................................................................................... ........71 3-9 n-layer doping dependence intensity at 5V and peak wavelength. A) Current. B) Output Intensity............................................................................................................... ..72 3-10.Schematic of ZnO p-n junction structure produced by P-diffusion.......................................73 3-11 The I-V characteristics at 25C from ZnO p-n junction. A) Linear Plot. B) Log plot......74 3-12 The I-V characteristics measured from neighboring p-contacts on the P-diffused ZnO layer.......................................................................................................................... ..........75 3-13 The SIMS profiles of P, Cd and As in the ZnO after diffusion from a Cd3P2, arsenic and red phosphorous sources.............................................................................................76 4-1 The DC responsivity measurement setup..........................................................................90 4-2 Dark-current of MSM photodetector.................................................................................91 4-3 The DC responsivity of MSM photodetecto r at different inci dent optical power. Green : 10 nW, black : 100 nW, brown : 1 W, blue : 10 W, red : 100 W...................92

PAGE 11

11 4-4 Mixing responsivity measurement setup...........................................................................93 4-5 The VLO dependence of the mixing responsivity at 100 W incident power. Red : 20 dBm, blue : 22 dBm, brown : 24 dBm, black 26 : dBm, green : 27 dBm. IF is 10 kHz............................................................................................................................ .........94 4-6 Optical power dependence of the mixing responsivity. Blue : 1 W, red : 10 nW...........95 4-7 The IF dependence of th e mixing responsivity at 100 W (dc) incident optical power. Red : 10 kHz, blue : 100 kHz, brown : 1 MHz, black : 10 MHz. VLO is 27 dBm.............96 4-8 Normalized reflectance, transmittance, and absorbance. A) 2000 thick e-beam deposited ITO. B) 2000 thick sputtered ITO.................................................................97 4-9 The AES spectra. A) e-beam deposite d ITO film. B) Sputtered ITO film........................98 4-10 The SEM images. A) e-beam deposited ITO film. B) Sputtered ITO film......................99 4-11 Normalized reflectance, transmittance, and absorbance. A) 100 e-beam ITO/1900 sputtered ITO film. B) 50 Ti/ 50 Au/1900 sputtered ITO film.........................100 4-12 Lifted-off 2000 sputter ITO. A) Optical microscopy image. B) The SEM image........101 4-13 Dark current and optical response. A) The Ti/Au MSM device. B) Sputtered ITO MSM device. C) Compos ite ITO MSM device...............................................................102 4-14 Schematic illustration of an in terdigitated MSM photodetector......................................103 4-15 External quantum efficiency of In0.55Ga0.45As MSM devices with metal-based interdigitated fingers as the function of finger gap spacing and finger width.................104 4-16 External quantum efficiency of In0.55Ga0.45As MSM devices with transparent ITObased interdigitated fingers. A) Finger gap spacing and finger width dependence. B) Contributions to the external quantum ef ficiency of an ITO-based MSM device...........105 4-17 Simulated and experimental external qua ntum efficiency of conventional metal and transparent ITO fingers for 1um finger width MSM phtodetectors.................................106 4-18 Response time dependence on finger ga p spacing and width for 100x100 um active area........................................................................................................................... ........107

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DESIGN, FABRICATION AND CHAR ACTERIZATION OF COMPOUND SEMICONDUCTOR FOR ELECTR ONIC AND PHTONIC DEVICES By Soohwan Jang May 2007 Chair: Fan Ren Major: Chemical Engineering Gallium nitride (GaN)-based n-channel enha ncement mode metal oxide semiconductor field effect transistor (MOSFET) using silicon (Si) diffusion was fabricated and characterized. In addition, performance of lateral and vertical structure zinc oxide (ZnO) light-emitting diodes (LEDs) was compared by using simulations. Photon responsivity of indium gallium arsenide (InGaAs) metal semiconductor metal photo-detect or (MSMPD) was improved dramatically by employing transparent indium tin oxide (ITO) Sc hottky electrodes, and optimum design of the interdigitated ITO contact was studied. Silicon (Si) diffusion into GaN was studied as a function of encapsulant type (SiO2 or SiNX) and diffusion temperature. Using a SiO2 encapsulant, the Si diffusion exhibited an activation energy of 0.57 eV with a prefactor of 2.07x10-4 cm2sec-1 in the temperature range 8001000C. An enhancement mode MgO/GaN-on-Si metal-oxide semiconductor field effect transistor (MOSFET) was fabri cated utilizing Si diffused regi ons under the source and drain contact. The devices showed improved transconduc tance and drain current relative to previous devices with Si-implanted source and drain regions.

PAGE 13

13 The effect of active, na nd p-layer doping level and thic kness of the ZnO-based LEDs on the optical output intensity and current-voltage characteristics of bot h vertical and lateral geometry ZnO LEDs were examined with simulati ons. The current density distribution was more uniform in the vertical structures but there wa s little difference in opt ical output power as a function of doping level or layer thickness between the two geometries. Also, p-type ZnO layers were realized by closed ampoule ph osphor diffusion into n type ZnO. The optical and electrical pr operties and surface morphologies of e-beam deposited and sputtered ITO films were studied. A novel compos ite e-beam deposited/sputtered ITO film was developed which avoided the issue of ion damage during the sputtering process. InGaAs-based MSM devices were fabricated with ITO fingers , and devices showed low dark current and improved photo-response as compared to the MSM devices using Ti/Au fingers. Furthermore, the optimum design of transparent ITO finger MSM photodetectors was discussed and compared with that of conventional metal finger MSM photodetectors.

PAGE 14

14 CHAPTER 1 INTRODUCTION 1.1. Motivation and Literature Review Gallium Nitride (GaN) and related materials ha ve been viewed as highly promising for applications in high power and frequency el ectronics capable of operation at elevated temperatures and in harsh radiative environment. The interest comes from mainly two intrinsic properties of this group of semiconductors. The first is their wide bandgap nature. The wide bandgap materials such as GaN and SiC, are promising for high temperature applications because their intrinsic carrier c oncentration stays at very low le vel at higher temperatures in contrast to that of Ge, Si or GaAs (Figure 1-1). It means that GaN power devices can operate at higher temperature with less expansive package associated with heat dissipation. The second attractive property of III-V nitrid es is that they have high brea kdown fields. The critical electric field of the breakdown scales roughl y with the square of the energy band gap, and is estimated to be >4 MV/cm for GaN1, as compared to 0.2 and 0.4 MV/cm for Si and GaAs, respectively. Also, GaN has excellent electron transport properties, including good mobility, and high saturated drift velocity2, thus making this material suitable fo r general electronics, and promising for microwave rectifiers and am plifiers in particular. Many epitaxial thin film growth methods fo r GaN-based materials have been developed, including molecular beam epitaxy (MBE)3, 4, hydride vapor phase epitaxy (HVPE)5-7, metal organic chemical vapor deposition (MOCVD)8-13, and derivatives of these methods. In the past few years, MOCVD has evolved as a leadi ng technique for production of III-V nitrides optoelectronic and microelectronic devices. Initially, the growth of GaN was performed directly on sapphire and SiC substrates, with a high density of threading def ects initiated at the substrate interface. The wafer usually had rough surfaces mainly caused by the 3D-growth mode. In 1986,

PAGE 15

15 Amano et al.9 succeeded in remarkably improving the GaN surface morphology as well as the electrical and optical properties by deposition of a thin low-temper ature AlN buffer layer prior to the high temperature growth of Ga N. The essential role of this buffer was served as a template for nucleation and promotion of lateral growth of the GaN film due to the decrease in interfacial free energy between the film and the substrate. Efficient n-type doping of GaN through incorpor ation of Si during the growth has proven relatively easy to achieve. High doping can also be achieved by implantation of Si or Group VI donors. Burm et al.14 used a shallow Si implant at hi gh dose to produce a doping level of 420 cm-3, resulting in an extremely low Ohmic contact resistance of 4-8 cm2 using Ti/Au contacts. It was difficult to obtain p-type conductivity and furt her research found that hydrogen played a crucial role in passivating the Mg accep tors. The neutral complex, Mg-H, prevented the formation of holes in GaN.5 For the first time, Amano et al.16 achieved p-type conductivity by activating Mg-doped GaN using low-energy elec tron irradiation. Nakamura demonstrated subsequently the activation of Mg by thermal annealing at 700~800 17 Amano et al.16 demonstrated the first p-n junction LED in 1989. Fo r the electronic devices, there has been also rapid progress in the reali zation of a broad range of GaN-based devices, including heterostructure field effect transistors (HFETs ), Schottky and p-i-n re ctifiers, heterojunction bipolar transistors (HBTs), bipol ar junction transistors (BJTs) and metal-oxide semiconductor field effect transistors (MOSFETs).18-26 Lately, ZnO is attracting attention for its application to UV light-emitters, varistors, transparent high power electronics , surface acoustic wave devices, piezoelectric transducers, gassensing and as a window material for display and solar cells.27-35It has advantages relative to GaN because of its availability in bulk singl e-crystal form, larger exciton binding energy (60

PAGE 16

16 meV) than that of GaN ( 25 meV), and ease of wet-etching.36-55 Recent improvements in the quality and control of c onductivity in bulk ZnO and epitaxial laye r have increased interest in the use of this material for transparent elec tronics and short wavelength light emitters.36-48 In addition, ZnO is lattice-matched to InGaN at an In composition of 22%, rais ing the possibility of integration of the two materials to provide enhanced functionality.56 ZnO is a direct bandgap semiconductor with bandgap of 3.37 eV, which can be tuned via divalent substitution on the cation site as shown in Figure 1-2. Substitution of Zn by Cd leads to the reduction in the bandgap to 3.0 eV. Substituti ng Mg on the Zn site in epitaxial films can increase the bandgap to approximately 4.0 eV while still maintaining the wu rtzite structure. The electron mobility in ZnO single cr ystals is on the order of 200 cm2/V s at room temperature.57 While the electron mobility is slightly lower th an for GaN, ZnO has a hi gher saturation velocity56 (Table 1-1). ZnO normally forms in the hexagona l crystal structure wi th a = 3.25 and c = 5.12 . The Zn atoms are tetrahedra lly coordinated to four O atom s, where the Zn d-electrons hybridize with the O p-electrons. Layers occupied by zinc atoms alternate with layers occupied by oxygen atoms. Electron doping in nominally undoped ZnO has been due to Zn interstitials or oxygen vacancies.45,58-62 The intrinsic defects behaving as n type dopants are positioned approximately 0.01.05 eV below the conduction band . The optical properties of ZnO, studied using photoluminescence, photoconductivity, and absorption, reflect the intrinsic direct bandgap, a strongly-bound exc iton state, and gap states due to point defects.59,63-65 A strong room temperature near bandedge UV photoluminescence peak at 3.2 eV is attributed to an exciton state, as th e exciton binding energy is on the order of 60 meV.66 In addition, visible emission is also observed due to defect states. A blue–green emission, centered at around 500 nm in wavelength, has been explained within the context of transitions

PAGE 17

17 involving self-activated centers formed by a doubly ionized zinc vacancy and an ionized interstitial Zn+59, oxygen vacancies67-70, donor–acceptor pair recombination involving an impurity acceptor71, and/or interstitial O.72-74 A broad orange–red photoluminescence emission was observed and assigned to defect states as well.75-78 Recently, a number of groups have reported hybrid heterojunction LEDs using n-type ZnO deposited on top of p-type layers of GaN, AlGaN or conducting oxides.79-83 Homojunction ZnO LEDs have been reported by Tsukazaki27, 28 with a structure of pZnO/i-ZnO/n-ZnO LED on a ScAlMgO4 substrate. Most of the emission consiste d of bands at 420 and 500nm, with a small shoulder at 395 nm assigned to radiative r ecombination in the p-ZnO through donor-acceptor pair transitions. 1.2. Background 1.2.1. Ion Implantation Ion implantation is the materi al engineering process by whic h energetic charged ions can be implanted into a substrate for the purpose of changing electrical, metallurgical, or chemical properties of the substrate.84 The typical ion energies range from 10 to 400 KeV, and typical ion doses are considered between 1011 to 1016 ions/cm2. The main advantages of ion implantation are low temperature of processing and precise c ontrol of total dose, depth profile, and area uniformity. However, subsequent damage of the substrate, expense, and complexity of the equipment are disadvantages. Figure 1-3 shows a schematic of an ion im plant system. The ion source has a heated filament to break up the source gas into char ged ions. A 40 kV extrac tion voltage causes the charged ions to move out of the ion source cham ber into an acceleration tube, where ions are accelerated to the implantation energy as they move from high voltage to ground. The accelerated ions then enter the mass analyzer. The magnetic field of mass analyzer is selected so

PAGE 18

18 that only ions with desired ma ss-to-charge ratio can travel thr ough it. The filtered ion beam is scanned over the wafer using electostatic deflec tion plates and ions are implanted into the semiconductor. The ion distribution in the semi conductor is given by Equation 1-1.85 ] ) 2 ( exp[ 2 ) (2 p p pR R x R x n (1-1) is the total number of ions per un it area, Rp the projected range, and Rp the projected straggle (Figure 1-4). Figure 1-5 shows the proj ected range of silicon in GaN as a function of implantation energy simulated by SRIM 2006. The projected range increases approximately linearly with the energy. Projecte d straggle and transverse straggl e in GaN for silicon ion are shown in Figure 1-6. The straggles are al so increases with implantation energy. Energetic ions enter a substrat e, lose their energy in semic onductor, and finally come to rest. The distribution of Si implanted with 350 ke V energy into GaN is illustrated in Figure 1-7. The loss of energy in the target material can be explained by two mechanisms. The first is by the interaction of the incident ions with electrons through Coulombic intera ction. The electrons can be excited to higher energy leve ls, or ejected from the atom. The second is by transferring the energy of incident ions to target nuclei, which causes deflection of incident ion and displaces many target nuclei from their original lattice sites. These dislodged atoms may have enough energy to cause cascades of secondary displacement of nearby atoms. As the density of displaced atoms approaches the atomic density of the semiconductor, the material becomes amorphous. Figure 1-8 shows the distribution of vacancies cr eated by incident Si ions, recoiled Ga, and recoiled N in GaN substrate. For example, if 1.014 Si ions /cm2 are implanted with 350 KeV energy, 2.721 Ga site vacancies / cm3, 4.121 N site vacancies / cm3, hence, 6.921 total vacancies / cm3 are created, whereas only 4.4 18 Si ions / cm3 are distributed at 3000 depth

PAGE 19

19 (typically, approximately 10% of im planted Si ions are activated after annealing process). Both repair of ion damage and activation of impl anted ions are done by heating the wafer at appropriate combination of time and temperature. Conventional furnace, rapid thermal annealing (RTA), or high intensity la ser radiation processes can be used for the annealing. 1.2.2. Diffusion Process In diffusion of impurities into the semiconductor substrate, dopant atoms are placed on or near the surface of the wafer by deposition from the gas phase of the dopant or by using doped oxide sources. The doping concentration decreases fr om the surface, and the profile of the dopant distribution is determined mainly by the te mperature and diffusion time. The temperature typically ranges between 600 and 1200 . At the elevated temperature, the lattice atoms start to vibrate around the equilibrium lattice sites. It is possible th at a host atom acquires enough ener gy to leave the lattice site and become an interstitial atom crea ting a vacancy. If an interstitial atom moves from one site to another without occupying a lattic e site, the mechanism is interstitial diffusion. Activation energy (Ea) is related to the movement of dopant atoms from interstitial s ite to another. When a neighboring impurity atom migrates to the v acancy site, the mechanism is called vacancy diffusion. Ea is related to both the energies of mo tion and the energies of formation of vacancies. Therefore, the activation energy for vacancy diffusi on is larger than that for interstitial diffusion usually. The simple one-dimensional diffusion process can be given by Fick’s diffusion equation (Equation 1-2).86 2 2) , ( ) , ( x t x C D t t x C (1-2)

PAGE 20

20 C is the impurity concentration and D the diffusi on coefficient. The diffusion coefficient depends on temperature and impurity concentration. Under low concentration, D becomes independent of impurity concentration and can be expressed as Equation 1-3. ) exp(0kT E D Da (1-3) D0 is the diffusion coefficient in cm2/s, and Ea the activation energy in eV. For Fick’s diffusion equation, there are two sets of initial and boundary conditions fo r which exact solutions can be derived. The first case is the constant surface concentration diffusion for all time greater than zero, and initial and boundary conditions are C(x, o) = 0, (1-4) C(0, t) = Cs, (1-5) C( , t) = 0, (1-6) where Cs is the surface concentr ation, which is independent of time. The solution for these conditions is given by Equation 1-7.87 ) 2 ( ) , ( Dt x erfc C t x Cs (1-7) erfc is the complementary error function and Dt is the diffusion length. The total number of dopant atoms per unit area of the semiconductor can be obtained by the integration of the profile. 02 ) , ( ) ( Dt C dx t x C t Qs (1-8) The second case is the constant total dopant diffusion. Initial and boundary conditions are C(x, o) = 0, (1-9) 0) , ( S dx t x C, (1-10) C( , t)=0, (1-11)

PAGE 21

21 where S is the total amount of dopant per unit area. The solution of the diffusion equation satisfy ing the above conditions is Equation 1-12.84 ) 4 exp( ) , (2Dt x Dt S t x C (1-12) This expression is Gaussian distribution, and the surface concentration decreases with time. Dt S Cs (1-13) Solutions for both constant surface concentr ation diffusion and constant total dopant diffusion are based on the assumption that the diffusivity is independent of the dopant concentration. However, when the impurity con centration, including both the substrate and the dopant, is greater than the intrinsic carrier c oncentration, the semiconductor becomes extrinsic and the extrinsic diffusivity de pends on the dopant con centration. In the extrinsic diffusion region, the diffusion profiles are more complicated. In this case, diffusion coefficient is not independent of C any more, and can be written as Equation 1-13. r s sC C D D ) ( (1-13) Cs is the surface concentration, Ds is the diffu sion coefficient at the surface, and r is the parameter related to the concentration depe ndence. When the diffusion is made into a background of an opposite impurity type, the junction depth is given by84 t D xs j6 . 1 for r = 1, (1-14) t D xs j1 . 1 for r = 2, (1-15) t D xs j87 . 0 for r=3. (1-16)

PAGE 22

22 1.2.3. Transfer Length Method When the current flows between metals through semiconductor, it sees the contact resistance and sheet resistance as shown in Figure 1-9. Transfer length method (TLM) originally proposed by Schockely is used to determine the c ontact resistance, sheet re sistance, and transfer length.88 The potential distribution under the contact is determined by specific contact resistivity ( c) and sheet resistance (Rs) according to Equation 1-17.89 ) sinh( ] ) ( cosh[ ) (T T s cL L Z L x L R I x V (1-17) L is the contact length, Z the contact width, and I the current flowing into contact. It is obvious that the voltage is highest at the contact edge (x=0). The 1/e distance of the voltage curve is defined as the transfer length, Lt. Thus, most of current flows from the semiconductor to metal within the transfer length. Figure 1-10 (A) shows the tran sfer method test structure. The total resistance, R, is measured for various contact spacings d, and plo tted as a function of d as illustrated in Figure 110 (B). The total resistan ce between any two contacts is given by Equation 1-18. c sR d Z R 2 (1-18) From the slope ( d R ), Rs is determined as z slope in / . The intercept at d = 0 gives Rc = intercept / 2 in , and leads to the transfer resistivity, Rt = RcZ in mm. LT is obtained as – [intercept (d) /2] in m at R=0. From the contact resistance and sheet re sistance, specific contact resistivity is give n by Equation 1-19. 2 2) (T s s c cL R R Z R in cm2 (1-19)

PAGE 23

23 Typical specific contact resistivities are less than 10-6 cm2 for good contacts, and transfer length is on the order of 1m or less for such contacts. 1.2.4. Secondary Ion Mass spectrometry Secondary ion mass spectroscopy (SIMS), al so known as ion micro probe and ion microscope, was developed independently by Cast aing and Slodzian at University of Paris and by Herzog at the GCA corp. in the United States in the early 1960’s.90,91 It is very sensitive technique with detection limits for some elements in the range of 1014 to 1015 cm-3. Lateral resolution is typically 100 m but can be as small as 0.5 m w ith the depth resolution of 5 to 10 nm.88 As shown in Figure 1-11, the removal of ma terial from the sample surface by sputtering and the analysis of the material by a mass analy zer are the basis of SIMS. A primary ion beam impinges on the sample and atoms from the surf ace are sputtered or ejected from the sample. 99% of the ejected atoms are neutral and cannot be detected by conventional SIMS. Only a small amount of ions are ejected as positive or negative , and the mass to charge ratio of the ions is analyzed as a mass spectrum, detected as a count, or displayed on a CRT. Cs+, O2+, Oand Ar+ are used for the primary ion source, and its ion nature has an important effect on the secondary ions polarity. O2 + primary ions are generally used to produce positive secondary ions whereas Cs+ primary ions are used to create negative ions. Sputtering takes place when atoms near th e surface receive sufficient energy from the incident ion to be ejected from the sample. The escape depth of the sputte red atoms is generally a few monolayers for primary energies of 10 to 20 KeV. Ion bombardment leads to not only sputtering, but also ion implanta tion and lattice damage. Some of the primary ions strike sample atoms and displace them from th eir lattice sites, and this caus es a homogenization of all atoms within the depth affected by the collision cascade. Dopant atoms originally present at a given

PAGE 24

24 depth in the sample will be distributed throughout this mixing depth as sputtering proceeds, and dopant profile will give a deeper distribution than true distri bution. Thus, it is important for the primary ion penetration depth to be kept to a minimum. Also, high vacuum should be maintained unless gaseous species like H, C, N, and O from the vacuum chamber can be detected. Quantitative depth profiling is the major st rength of SIMS. The conversion of signal intensity to density of intere st dopant can be calculated knowing primary ion beam current, sputter yield, ionization efficiency , atomic fraction of the ions to be analyzed, and instrumental factor. Using ion implanted standards with composition and matrices identical or similar to the known, some of above factors can be obtained. The time-to-depth conversion is usually made by measuring the sputter crater depth after the analysis is complete. 1.2.5. Auger Electron Spectroscopy Auger electron spectrosc opy (AES) is analytical technique for the characterization of chemical and compositional properties of material s, based on the Auger effect discovered by Auger in 1925.92 The detection limit of AES is about 0.1% but it varies greatly from element to element. It is also influenced by the beam current and analysis time. 5% accuracy for the standard semiconductor sample which has 10% of interest element was reported.93 Figure 1-12 shows the process of Auger electron emission. A primary electron from an electron gun ejects an electron from the core K shell and leaves a vacancy. The K shell vacancy is filled by an outer shell electron ( in this case, electron in L1 ) or valence band electron. The energy equivalent to EL1-EK is transferred to the third elect ron, which is Auger electron and ejected from the L2,3 level. The energy can also result in the emission of an X-ray photon. Auger emission dominates over X-ray emission for the low atomic number (Z) elements. The atom remains in a doubly ionized state and entire proce ss is labeled as “KLL”. In the KLL transition, the L shell ends up with two vacancies, which ca n be filled by valence band electrons (LVV).

PAGE 25

25 Since Auger process is three-electron process, hydrogen and helium cannot be detected by AES. For 3 < Z < 14, KLL transitions take a place, LMM for 14 < Z < 40, and MNN transitions for 40 < Z < 82. Auger electron spectro scopy instrumentation consists of electron gun, electron beam control, electron ener gy analyzer, and data analysis electronics.88 The typical incident electron beam energy is 1 to 5 keV. Higher beam energy electrons produce Auger el ectrons deeper in the sample with little chance of es caping. The focused electron be am diameter is dependent upon electron source, beam energy, electron optics, and beam current. For non-scanning AES, the diameter of electron beam is on the order of 100 m, and it is smaller for a scanning system. With the field emission source, a 10 nm beam at 1nA beam current can be achieved. The cylindrical mirror analyzer (CMA) is most co mmon analyzer shown in Figure 1-13. A coaxial configuration with the analyzer wrapped around the electron gun reduces shadowing and allows room for positioning the ion sputter gun. Auger elec trons enter the inlet aperture between the two concentric cylinders and are focused by a negative potential that creates a cylindrical electric field between the coaxial electrodes. Although the sampling depth of AES is typical ly 0.5 to 5nm, depth profile of interest element can be obtained by sputtering with an inert ion beam. When th e surface is sputtered during AES depth profiling, sputter induced ar tifacts may appear, which include crater wall effect, re-deposition of sputtere d material, surface roughening, pr eferential sputtering, atomic mixing, charging effect, and specimen damage. AES measurement is made in a high vacuum environment of 10-12-10-10 torr to avoid the formation of contamination films on the surface like carbonaceous material. Since Auger electrons can only escape from very sh allow surface layer, the presence of surface contamination interferes with the Auger Signal.

PAGE 26

26 1.2.6. Atomic Force Microscopy Atomic force microscopy (AFM) operates by me asuring the force betw een a probe and the sample for imaging, measuring and mani pulating of the sample at the nanoscale.88,94 AFM is suitable for conducting as well as insulating samp les. Figure 1-14 shows the schematic diagram of AFM. A tip is brought into continuous or in termittent contact with the sample and scanned across the sample surface. Th e sample is moved in the z direction while maintaining a constant force by piezoelectric tube scanner, and in the x and y directions for scanning of the sample. The motion of the cantilever is sens ed by a segmented position senstiv e photodetector. The cantilever is typically made from silicon, silicon oxide , or silicon nitride and is around 100~200 m long and 0.5 to 5 m think. Silicon nitride can tilever and tip are formed by depositing Si3N4 on the silicon surface containing a pyramid etch pit. The sampe height variation is measured by holding the signal constant, equivalent to constant cantilever deflection, that is, by varying the sample height through a feedback arrangement. In contact mode, the sample topography is measured by scanning the tip, which contacts the sample surface. Usually, samples are cove red with a thin laye r of water or other contaminants. When the probe touches the surface, it is pulled toward the sample by a capillary force. This force coupled with electrostatic force, can create a substantial frictional force and lead to possible sample damage. In the non-cont act mode, the instrument senses van der Waals force between the tip and sample surface. The re solution is lower than the contact mode. In tapping mode, the cantilever is ex cited close to its resonant frequency by an external signal applied to a piezoelectric ceramic to which the cantilever is attached. Resonance frequencies depend on the mechanical property of cantilevers and range from 15kHz to 500 kHz.

PAGE 27

27 1.3. Dissertation Outline This dissertation consists of three parts; n-channel enhancement mode GaN MOSFET, simulation of ZnO LED, and InGaAs MSM photodet ector. Chapter 2 discusses Si diffusion into GaN and current voltage charac teristics of Si diffused nor mally-off GaN MOSFET. Chapter 3 describes the simulated results of vertical and lateral geometry ZnO LED , and device fabrication and performance of P diffused ZnO diodes. St udy on device structure of InGaAs-based MSM optoelectric mixer, properties of transparent ITO films as interdig itated electrodes, and optimum design of ITO fingers is present in chapter 4. Chapter 5 presents a brief summary and suggestions for future work.

PAGE 28

28 Table 1-1. Material prop erties of ZnO and GaN. ZnO GaN Bandgap (eV) 3.37 eV, direct 3.39 eV, direct Crystal structure () wurtzite a=3.250 , c=5.205 (c/a=1.602) wurtzite a=3.189 , c=5.206 (c/a=1.632) Excition binding energy (meV) 60 meV 25 meV Effective mass m*=0.24m o (e) m*=0.59m o (h) m*=0.20m o (e) m*=0.80m0 (h) Mobility (cm2/Vs) 200 (e) 5~50 (h) 1000 (e) 30 (h) Saturation Velocity (cm/s) 3.27 2.57 Thermal conductivity (W/cmK) 0.3 1.3 Thermal expansion coefficient (10 -6 /K) a 6.5 c 3.7 a 5.59 c 3.17

PAGE 29

29 Figure 1-1. Intrinsic carrier concentration according to temp erature of Si, GaAs, and GaN

PAGE 30

30 Figure 1-2. Bandgap energy and latt ice constant of GaN and ZnO MgO ZnO CdO MgO ZnO CdO

PAGE 31

31 Figure 1-3. Schematic diagram of an ion implanter

PAGE 32

32 Figure 1-4. Coordinate of implanted ions Rp Rp Ion Beam Ion Concentration y x

PAGE 33

33 101001000 0.01 0.1 1 Projected Range ( m )Energy (KeV) Figure 1-5. Projected range of Si in GaN as a function of implantation energy

PAGE 34

34 101001000 0.01 0.1 Projected Straggle Transverse Straggle Projected and ransvers e Straggle ( m )Energy (KeV) Figure 1-6. Projected straggle and transverse straggle of Si in Ga N as a function of implantation energy.

PAGE 35

35 010002000300040005000 0.0 1.0x10-42.0x10-43.0x10-44.0x10-45.0x10-4 Target Depth ( ) Number / Ion Figure 1-7. Distribution of Si in GaN with implantation energy of 350 KeV.

PAGE 36

36 010002000300040005000 0.0 0.2 0.4 0.6 0.8 1.0 N Vacancies Total Vacancies Ga Vacancies Number / IonTarget Depth ( ) Figure 1-8. Nitrogen, gallium, and total vacan cies in GaN after 350 KeV Si implantation

PAGE 37

37 Figure 1-9. Lateral current flow geometry RcRc Rs Metal Contact Region causing RcBulk Resistance

PAGE 38

38 A B Figure 1-10. Transfer Length Method. A) Transfer le ngth method test structure. B) Plot of total resistance as a function of contact spacing. L Z d1 d2 d3 d4d5 R Intercept = 2Rc Slope = Rs/Z Lt d

PAGE 39

39 Figure 1-11. Schematic diagram of secondary ion mass spectrometry. Mass Filter Primary Ions Secondary Ions Mass Spectrometer Ion detector Sample Ion Beam

PAGE 40

40 Figure 1-12. Emission of Auger electron. Auger Electron Primary Electron EVacECEL2 , 3EVEL1 EK

PAGE 41

41 Figure 1-13. Schematic diagram of Auger electr on spectroscopy. Sample VaTo Electron Multiplier Sputter Ions Magnetic Shield Incident Electrons

PAGE 42

42 Figure 1-14. Schematic diagram of atomic force microscope. Laser Light Cantilever Position Sensitive Photodetector Sample Piezoelectric Tube Scanner

PAGE 43

43 CHAPTER 2 ENHANCEMENT-MODE GALLIUM NITRIDE METAL OXIDE SEMICONDUCTOR FIELD EFFECT TRANSISTER WITH SILICON-DIFFUSION INTO SOURCE AND DRAIN REGIONS 2.1. Introduction There is strong interest in the developm ent of GaN-based normally-off (enhancement mode) transistors for reducing power consumption and gate leakage current as well as improving voltage swing relative to the more conventional depletion-mode metal gate transistors. The use of an oxide gate enhances the device perfor mance at high temperature because of the higher effective barrier height and better thermal stability as compared to Schottky gate transistors. Metal-oxide semiconductor (MOS) heterostructur e field effect transistors (HFETs) have exhibited approximately four orde rs of magnitude lower gate leak age current at 300 C than that of conventional HFETs.26,95-118 Many insulators, such as Ga2O3 (Gd2O3), AlN, SiO2, Si3N4, MgO and Sc2O3 113 have been used in GaN-based MOSFET. MgO and Sc2O3 produced interface state densities in the range 2x1011/eVcm2 (at Ec-Et = 0.2eV from the Terman method) to 6x1011/eVcm2 (at Ec-Et = 0.42eV from AC conductance measurements).119-123 Inversion characteristics with gate-contro lled MgO/GaN diodes was observed121,123 and initial results on MgO/GaN e-mode MOSFETs showed the inversion of the channel was achieved for gate voltages above 6V.124 The transconductance was limited to 5.4 S.mm-1 at a drain-source voltage of 5V, comparable to the initial values reported for GaAs-based MOSFETs.125 One of the weaknesses in the processing of those devices wa s the need for high temperature annealing (> 1300C ) to activate the Si implanted under the source-drain regions, wh ich roughened the GaN surface and made it difficult to gr ow high quality gate oxides. Another option for local n-type doping of the source/drain regions is Si diffusion. Lin et al.126 showed greatly improved cont act resistance for Ni/Ti/Al contacts to Si-diffused GaN.

PAGE 44

44 The diffusion temperatures were 1000C, significantly below the implant activation temperature. Silicon (Si) diffusion into GaN and the use of this process to fabricate enhancement mode MOSFETs using MgO as a gate dielectric on und oped GaN-on-Si are studie d. The devices show good dc characteristics and demonstrate the hi gh quality of gate oxi des grown by molecular beam epitaxy (MBE). 2.2. Silicon Diffusion in to Gallium Nitride Several different types of samples were us ed in these experiments. For the initial development of the Si diffusion process, we used ~2 m thick nominally undoped GaN (n ~5 x 1016 cm-3) grown on c-plane sapphire substrate. The GaN was grown by metal organic chemical vapor deposition (MOCVD) at 1040 C. The sample s were cleaned in buffered oxide etchant (BOE) prior to depositing 100500 of Si by e-beam evaporation. The samples were encapsulated with 1000 of SiO2 or SiNX deposited by plasma enhanced chemical vapor deposition and then annealed at 700-1000 C under a flowing N2 ambient. A schematic of the process is shown in Figure 2-1. After BOE removal of the SiO2 and HF/HNO3/CH3COOH wet etch of the Si, secondary ion mass spectrometry (S IMS) was performed to obtain the Si diffusion depth. On these samples, we also measured the contact resistance of Ti/Al/Pt/Au contacts deposited by e-beam evaporation on the diffused regions and patterned by liftoff and dry etch delineation of a transmission line pattern. The second type of sample was 2 m of high resistivity GaN grown on a buffer layer on Si substrate. The template resistivity was >106 .cm. On this sample we performed the optimized Si diffusion condition determined from the prev ious sample. The e-beam deposited n-ohmic (Ti/Al/Pt/Au) was annealed at 850. The MgO gate dielectric was grown by MBE at 100C on

PAGE 45

45 the undoped GaN layer as the gate dielectric. Th e MgO precursors were elemental Mg and radiofrequency plasma-activated oxygen. The GaN sa mples were cleaned initially with a 3 min chemical etch in HCl/H2O(1:1), H2O rinse, UV-ozone exposure fo r 25 min, rinse in buffered oxide etch solution(6:1, NH4F/HF) and rinsed in H2O. The samples were then loaded into the MBE system and heated at 650C to ensure oxide removal. A st andard effusion cell operating at 380C was used for evaporation of the Mg, while the O2 source was operated at 300W forward power (13.56MHz) and 2.5x10-5 Torr. The deposited MgO layers were ~80nm thick. Following this step, the MgO was removed in the source and drain contact areas, using H3PO4 etchant and an AZ 5214-E resist mask. The e-beam deposited ga te(Pt/Au) metals were patterned by lift-off. An Agilent 4156C parameter analyzer was em ployed to measure current-voltage (I-V) characteristics of devices with gate length 6 m and gate width 100 m. 2.3. Silicon Diffusion and Contact Properties Silicon dioxide (SiO2) provided far superior results as an encapsulating layer than SiNX, as shown in the optical micrographs of the su rfaces after diffusion at 900C for 120 mins and removal of both the cap layer and the remaining Si (Figure 2-2).The Si was removed in a mixture of HF/HNO3/CH3COOH. SiO2 was used as the encapsulant in all subsequent work. Figure 2-3 shows the SIMS profile s of Si in GaN as a function of diffusion temperature for a fixed initial Si thickness of 200. Defining the diffusion distance x as the distance at which the Si concentration reaches the background sensitivity of ~5x1016 cm-3 and equating this to (Dt)0.5, where D is the diffusion coefficient at temperat ure T and t is the diffusion time, then we can represent the data in Figu re 2-3 by the relation D=DO exp(-EA/kT), where the prefactor DO is 2.07x10-4 cm2/V.s and the activation energy EA is 0.57eV. This is faster than observed in Si-

PAGE 46

46 implanted GaN, where the creation of vacancies by the ion-stopping process means there is a greater chance of Si occupying a substitutional lattice site.127 Figure 2-4 shows that the GaN surface after di ffusion and removal of the encapsulant and remaining Si did not show any significant roughening up to diffusi on temperatures of 900C, as measured by Atomic Force Microscopy (AFM).This is ideal for the subse quent growth of the MgO gate dielectric. Figure 2-5 shows some of the Ohmic contact data for Si-diffused GaN. The Si diffusion process lowers the specific contac t resistivity by about a factor of 2 over the control sample with the same contacts that was as grown undoped GaN. Th e final value of specif ic contact resistivity was not strongly dependent on temperature in the range 750-900C (Fi gure 2-5 (A)). The Si thickness dependence of specific co ntact resistivity for a fixed di ffusion temperature of 900 C is shown at Figure 2-5 (B). Beyond a thickness of 100 there was also not a strong dependence of the contact properties, so we used 200 in subsequent experiments. 2.4. Magnesium Oxide/Gallium Ni tride Metal Oxide Semiconducto r Field Effect transistor Using Silicon-Diffused Source/Drain Regions The MOSFET was fabricated on the GaN/Si template. A schematic of the final structure is shown in Figure 2-6(A) and a microscope image is shown at Figure 2-6 (B). The Si diffusion process was carried out at 1000C for 120 mins with 200 of Si encapsulated with SiO2. For the MgO thickness of 80nm, the corresponding oxide breakdown field was 1.85 MV/cm at a current density of 15 mA/cm2. The gate contact exhibited good rect ification, demons trating the MgO produces excellent insulator char acteristics and a much larger breakdown than if the Pt/Au was placed directly on the GaN without any MgO . 300K capacitance-voltage characteristics measured at 1 MHz and with a sweep rate of 30 mV/s performed on companion diodes showed clear modulation from accumulation to depletion .Using the relation, Cox = oox A/Tox, (where o

PAGE 47

47 is the permittivity in a vacuum, Cox is the oxide capacitance, A is the cross-sectional area of oxide, and Tox is the oxide thickness), the di electric constant of the oxide, ox, was calculated to be 9.7, a value which is in agreement w ith the tabulated value for MgO (9.8). The drain-source I-V characteristics from the MOSFET are shown in Figure 2-7. The channel is formed at a gate voltage of > +1V a nd the device can be modulated to + 10V of gate voltage. There is no indication of negative resi stance effects and thus self-heating does not appear to be a problem under the biasing condi tions employed here. The transconductance of the MOSFET (i.e. the change in drain current for a change in gate voltage) is maximized for high electron mobilities, thinner MgO dielectric layers and larger values of ga te length/gate width. MEDICI simulations indicate that the drain current should be at least an order of magnitude higher for the structure used here and the lower currents obtained show that further improvement is possible for increasing carrier mobility and the n-type-doping level obtained in the source/drain regions, which affects the series resistance. The transfer characteristics of the MOSFETs are shown in Figure 2-8. The maximum drain current was ~35 A/mm at 7V gate voltage a nd the transconductance was 7.0 S.mm-1. This is an improvement over previous GaN MOSFET reports.124 Clearly much more work has to be done to increase the channel mobility.

PAGE 48

48 Figure 2-1. Schematic of struct ure for Si diffusion in GaN. 700~1000 , N2Ambient Si 1000SiO2or SiNx GaN Buffer Layer Sapphire 700~1000 , N2Ambient 700~1000 , N2Ambient Si 1000SiO2or SiNx GaN Buffer Layer Sapphire GaN Buffer Layer Sapphire

PAGE 49

49 Figure 2-2. Optical micrographs of surfaces afte r Si diffusion and subsequent removal of the SiO2 or SiNx encapsulant and remaining Si layer. After Diffusion After Removing SiNx After Removing Si 100 SiNx After Diffusion After Removing SiO2After Removing Si Si/SiO2 SiO2 Si SiO2

PAGE 50

50 Figure 2-3. SIMS profiles of Si diffused in GaN at 800, 900 or 1000C for 120 mins.

PAGE 51

51 Figure 2-4. The AFM surface scans of control Ga N and surfaces after diffusion of Si at 800 or 900C and subsequent removal of the SiO2 and remaining Si.

PAGE 52

52 Figure 2-5. Specific contact resistivity of Ti/A l/Pt/Au contacts on the Si-diffused regions. A) Diffusion temperature dependen ce. B) Si thickness depe ndence for a fixed diffusion temperature of 900C. A B

PAGE 53

53 Figure 2-6. The Si diffused enha ncement mode metal oxide semic onductor field effect transistor. A) Schematic cross-section of the MgO/p -GaN MOSFET. B) Microscope image of the device layout. Silicon substrate UndopedGaN n+ n+ MgO Pt/Au Ti/Al /Pt/Au Ti/Al /Pt/Au Silicon substrate UndopedGaN n+ n+ MgO Pt/Au Ti/Al /Pt/Au Ti/Al /Pt/Au Silicon substrate UndopedGaN n+ n+ MgO Pt/Au Ti/Al /Pt/Au Ti/Al /Pt/Au

PAGE 54

54 Figure 2-7. The IDS-VDS characteristics from the MgO/GaN -on-Si MOSFET. The gate voltage was changed from 4V in steps of -1 V. 01234567 0 5 10 15 20 25 30 35 Drain Voltage (V)Drain Current (uA/mm)MgO/GaN MSOFET Vg= 4V Step -1V

PAGE 55

55 Figure 2-8. Transfer characteristics of MgO/G aN-on-Si MOSFET at a drain-source voltage of 7V.

PAGE 56

56 CHAPTER 3 ZINC OXIDE LIGHT-EMITTING DIODE 3.1. Introduction There has been great development of solid state sources for lighting and general illumination. One approach involves blue/UV lig ht-emitting diodes (LEDs) coupled to phosphors for visible color and white light generation. Th e first LEDs and laser diodes (LDs) operating in the blue/green spectral range were fabricated from c ubic II-VI (MgZnCdSSe) compounds128 but the reliability of these devi ces was poor due to strong degradation processes associated with the high mobility of native point defects in the material. Gr oup III-Nitride (AlInGaN) LEDs have become the standard materials for light emitters in the blue/green/UV spectral range due to their excellent reliability.129,130 The spectral range of the III-nitride LEDs now extends to the near-UV (340-400 nm) and the deep-UV (250-300 nm) regions.131-133 One issue with nitride LEDs is the presence of a high dislocation density due to the la ck of native GaN or AlN substrat es. The use of optimized buffer layers and GaN “templates” (thick epilayers normally grown by hydride vapor phase epitaxy) reduces the typical threading dislocation density of ~109 cm-2 in layers grown on sapphire or SiC substrates to the 107 cm-2 range. This is still too high, however , to eliminate the effect of defects on the internal quantum effici ency (IQE) of nitride LEDs.133,134 The threading dislocations may also affect the overall quality of epitaxial mate rials, and reduce the life time of devices operating at high current density, i.e., of high-power LEDs and LDs. An alternative approach uses ZnO, which ha s a bandgap of 3.37 eV at room temperatures and is suitable for light emission in the near-UV (~370 nm) spectral range.27,28,49,53,79,80,82,83,134-139 High quality bulk substrates of Zn O are commercially available and the use of Mg or Cd addition produces a heterostructure system. For example, MgZnO has a wider bandgap and a relatively

PAGE 57

57 low lattice constant mismatch with ZnO, e.g., ~0.4% for Mg0.2Zn0.8O. Thus, MgZnO/ZnO/MgZnO heterostructur es with effective carrier confinement in the ZnO active region can be fabricated. More over, by introducing a CdZnO active layer the emission spectral range may be extended into the visible regime. A number of groups have recently demonstrated emission from all-ZnO or hybrid ZnO/GaN LED s. ZnO has a much higher exciton binding energy than GaN and thus is expected to display more robus t band-edge emission.27,28,53,8083,138,139,143 There are two basic structures of interest -the first is a vertical geometry LED on a ZnO substrate which would exploit the low dislo cation density of the substrate to improve IQE relative to GaN and the second is a lateral geometry LED which could be deposited on a cheap substrate such as glass and w ould represent a low-cost appro ach. Simulation and design aspects of these two approaches are discussed. If efficient lasers or LEDs can be made of ZnO, this UV light could be used to excite phosphors to produce white light.40,53,140-142 Efforts to make thin films of p-ZnO have involved several deposition methods (e.g., molecular-beam ep itaxy, chemical-vapor deposition, sputtering, spray deposition, diffusion, oxidation of Zn3N2), and use of different dopants (e.g., nitrogen, phosphorous, arsenic, antimony, lithium,).In additi on, co-doping with Ga-N, Al-N, and In-N has been reported, varying degrees of success.27,40,53,140-142 ZnO-based p-n junctions have been reported, as has the emission of near-band-edge light emission.27,79,80,118,143 ZnO has asymmetric doping limitations that strongly favor n-type conductivity. Low formation energies for intrinsic donor defects, such as oxygen vacancies (VO) and zinc interstitials (Zni), produce donor states that compensate acceptors.145 Growth conditions for p-type films are also favorable for the formation of native donors when attempting anionic doping.146 While there have been significant activities focused on nitrogen doping, few reported effort s have addressed phosphorus doping.49,

PAGE 58

58 140, 147,147-159 p-n junction-like behavior ha s been reported between a n-type ZnO substrate and a surface layer that was heavily doped with phosphorus.49 Kim et al.147 showed that postactivation annealing produced p-type in P-doped ZnO films grown by sputter deposition. Clearly more effort is needed to develop reliable p-doping me thods for formation of p-n junctions in ZnO. The electrical properties of ZnO homojuncti ons formed in bulk single-crystal substrates by P-diffusion to form the p-type layer are investigated. Junctions with good rect ification ratios are formed by P diffusion at relatively low temperatures (550C). 3.2. Simulation of Vertical and Latera l Zinc Oxide Light-Emitting Diodes 3.2.1. ISE TCAD Simulation Schematics of the two structures used for simulation are shown in Figures 3-1 and 3-2. We have not defined a substrate for the lateral geometry device, but as mentioned above it could be a cheap material such as glass. The dc ch aracteristics were simulated using the ISE TCAD code, which is based on solving the Poisson and continuity equations of a 2D-structure. The Poisson equation is given by Equation 3-1. ) ( A DN N n p q (3-1) is the electrical permittivity, the electrostatic potential, q th e elementary electric charge, p the hole density, n the electron density, ND+ the number of ionized donors and NA+ the number of ionized acceptors. The electron and hole con tinuity equations are given by Equation 3-2 and Equation 3-3. t p q qR Jp (3-2) t n q qR Jn (3-3)

PAGE 59

59 R is the net electron–hole recombination rate and Jn,p are the electron and hole current densities. Physical models included drift-diffusion transport with Fermi-Dirac statistics, surface recombination, Shockley-Read-Hall recomb ination, Auger recombination and bandgap narrowing at high doping levels. The optical gene ration along a ray (z-axis) was calculated from Equation 3-4. z z optdz z z z y x J t Z G0] ) , ( exp[ ) , ( ) , , ( ) , (0 (3-4) ( ,z ) is absorption coefficient, J ( x,y,z0) beam spatial variation of intensity, and z0 position along the ray. The photon rate equa tion is given by Equation (3-5). ) ( ) ) ( ( w T c S L w G t Ssp r (3-5) is the spontaneous emission factor, G(w) the modal gain, L is the optical loss and Tsp is the spontaneous emission. The best curre ntly available values of propert ies such as density of states, recombination coefficients, surface velocity, low field mobility and effective mass were employed.160 3.2.2. Comparison of Vertical and Lat eral Zinc Oxide Light-emitting Diode Both the structures showed a fl at-band condition at 5V forward bias and the majority of the carrier recombination was found to occur in the active layer near the p-layer due to the higher mobility for electrons in ZnO (Figure 3-3). Figur e 3-4 shows the total current density at 5V forward bias in the vertical st ructure. As expected, the curren t distribution (and the radiative recombination rate and electric field distribution) is quite unif orm. By contrast the lateral structure (Figure 3-5) has a highe r current density near the edge of the mesa. The recombination rates in the latter structure we re found to be dominated by the area encompassing the p-region, active region and n-region above the mesa.

PAGE 60

60 The calculated current-voltage (I-V) characteristics form both structures showed little dependence of current on undoped active layer thic kness in the range of 0-200 nm. As a result, Figure 3-6 shows there was little difference in ou tput power and optical intensity for the two structures as a function of active layer thickness. The LED output power increases as the active layer thickness increases since there is a higher probability for electron and hole recombination. The effect of active layer doping was also exam ined and no significant difference in the I-V characteristics over the range 5x1015-5x1018 cm-3 was found. This is an attractive resu lt since the control over epi growth residual dopi ng in this layer is not stringent and one can focus instead on having high crystal quality. The doping level on the p-side of the junc tion had a strong influence on the emission spectrum intensity for both geometries, as shown in Figure 3-7. The recombination rate Rr and hence output power increases rapidly as a functi on of hole concentration through the relation Rr = Bnp where n and p are the carrier concentrations in the n and p respectively. The current at a given voltage shows a slight decrease with increas ing p-layer doping because of the drop in hole mobility due to scattering, but this is outweighe d by the increased recombination which leads to increased output intensit y at the peak wavelength as shown in Figure 3-8. The doping level in the n-ZnO layer also had a significant effect on the LED performance. Figure 3-9 shows the current through the n-layer at 5V forward bi as (A) and output intensity (B) as a function of n-layer doping. The current increases with n layer doping. For the lateral structure, the voltage drop acro ss the Ohmic contact gives more effect on current (current crowding). The decrease at very high doping levels is due to the mobility drop from scattering. Once again, the recombination rate increases wi th electron concentratio n, leading to higher power and output intensity.

PAGE 61

61 3.3. Formation of p-n Homojunctions in n-Zinc Oxide Bulk Single Crystals by Diffusion from Zinc Phosphide (Zn3P2) Source 3.3.1. Diffusion from Zinc Phosphide (Zn3P2) Source into Bulk Zinc Oxide Substrate The samples were (0001) undoped grade I quality bulk, single-crystal ZnO crystals from Cermet. They were epiready w ith one-side-Znface-polished by the manufacturer. The room temperature electron concentration and mobility established by van der Pauw measurements were 1017 cm and 190 cm2/V s, respectively. The P diffusion was performed by sealed ampoule at a temperature of 550C for 30 mins under N2, similar to that described in detail previously.156 The incorporation depth of the P wa s determined by Secondary Ion Mass Spectrometry (SIMS) and quantified using an ion-implanted standard. Th e diodes were fabricated using full-area back contacts of e-beam evaporated Ti (200 )/Au(2000) and front-side Ni(200)/Au (800) contacts also deposited by e-beam evaporati on and lithographically pa tterned by lift-off. A shallow mesa for current flow delineation was etched using HCl/H2O. The contacts were annealed at 200C under flowing N2 ambient. A schematic of the diodes is shown in Figure 3-10. 3.3.2. Phospher Diffused Zinc Oxide p-n Diode A typical current densit y-voltage characteristic from the diffused structures at 300 K is shown in Figure 3-11 in both linear (A) and log (B) form. These devices show excellent diodelike behavior. The forward turn -on voltage defined at 100 Acm-2 was ~4 V with an on-state resistance of ~21 -3 cm2. The on/off rectification ratio was ~70 at +3/-5V. The forward current density JF at low forward voltage V and temperature T is determined by a current contribution due to recombination in the space charge region and by a diffusion contribution due to recombination close to the space charge region. ) exp( 1 ) ( ) 2 exp( kT eV e kT e kT eV W en Jp p r i F (3-6)

PAGE 62

62 e is the electronic charge, ni the intrinsic carrier density, W th e depletion depth, k Boltzmann’s constant, p the hole mobility and p the hole lifetime. The theore tical forward I-V’s expected from our junctions with different mechanisms in the current transport (i.e. Shockley-Read-Hall recombination, Auger recombination, and inco mplete P dopant ionization) were modeled by MEDICI but these details were found only to affect the low-current regime and to have little influence on the ultimate value of turn-on voltage. The forward current density-voltage characteristics at low curr ent densities (< 0.1 Acm-2) showed thermally activated behavior with activation energy ~ 1.4 eV and n 2, which is consistent with the presence of several current transport mechanisms in the junction such as defect-assisted tunneling and conventional carrier recombination in the space-charge region via midgap deep le vels. At higher current densities, the rectifiers showed excess current with slope of exponential dependen ce. This is consistent with recombination via multilevel centers and has been reported previously for SiC rectifiers and heteroepitaxial GaN p-i-n rectifiers.161,162 The bulk component of reverse leakage in a p-n junction rectifier (IR) can be expressed as Equation 3-7. m D i D p i RN V n B N n A I ) (2 (3-7) A and B are constants and the effective lifetime in the space-charge region. The first term represents the diffusion current component and the second term is the generation current component. In our diodes, IR was roughly proportional to V0.5 up to ~ 5 V. For biases up to this value, the reverse current was directly proporti onal to the contact di ameter, indicating the dominance of surface perimeter leakage.

PAGE 63

63 The polarity of the I-V characteristics is consistent with the ZnO being p-type, with measurement in a lateral direction producing Ohmic behavior (Figure 3-12) while the vertical structure showed rectifying charact eristics, confirming the presence of a p-n junction. The n contacts were Ohmic even in the as-deposited state, facilitated by the high n-type doping of the substrate. The observation of light emi ssion from the junctions was attempted, but any emission at room temperature could not be detected. SIMS profiles of the P are shown in Figure 3-13. The presence of As and Cd which are often found as impurities in Zn3P2 was profiled, but these were not detected above the background sensitivity. The P is in corporated to a depth of ~200nm. If we use the simple relation of diffusion distance being (4Dt) 0.5, this translates to a P diffu sion coefficient at 550C in ZnO of 5.6x10-13 cm2/s. Of course the diffusivity might be fa ster in polycrystalline ZnO due to the presence of grain boundaries.

PAGE 64

64 Figure 3-1. Schematic of vertical geom etry ZnO LED used in the simulations.

PAGE 65

65 Figure 3-2. Schematic of lateral geomet ry ZnO LED used in the simulations.

PAGE 66

66 0.00.20.40.60.8 -4 -3 -2 -1 0 1 2 3 4 Y ( um ) Energy (eV)0.0 2.0x10224.0x10226.0x10228.0x10221.0x10231.2x1023 0V 3V 5V Radiative Recombination (cm-3s-1) Figure 3-3. Conduction and valenc e band, and radiativ e recombination of vertical device.

PAGE 67

67 Figure 3-4. Total current de nsity distribution at 5V in the vertical structure.

PAGE 68

68 Figure 3-5. Total current de nsity distribution at 5V in the lateral structure.

PAGE 69

69 Figure 3-6. Active layer thickness dependence at 5V and peak wavelength for both LED geometries. A) Output pow er. B) Optical intensity. 050100150200 0.0 5.0x10-81.0x10-71.5x10-72.0x10-72.5x10-73.0x10-7 Vertical Lateral Active Layer Thickness (nm)Power (W)050100150200 6.0x10-69.0x10-61.2x10-51.5x10-51.8x10-52.1x10-5 Vertical LateralActive Layer Thickness (nm)Intensity (a.u.) A B

PAGE 70

70 Figure 3-7. Simulated electrolu minescence spectra as a functi on of doping level in the p-ZnO layer. A) Vertical geometry structure. B) Lateral geometry structure. 320330340350360370380390400410 0.00 5.00x10-51.00x10-41.50x10-42.00x10-4 1e16 2e17 8e17 5e18Wavelength (nm)Intensity (a.u.) 320330340350360370380390400410 0.00 5.00x10-51.00x10-41.50x10-42.00x10-4 1e16 2e17 8e17 5e18Wavelength (nm)Intensity (a.u.) A B

PAGE 71

71 Figure 3-8. p-layer doping dependen ce at 5V and peak wavelength. A) Output power. B) Optical intensity. 101610171018 0.00 5.00x10-71.00x10-61.50x10-62.00x10-62.50x10-6 Vertical Lateralp layer DopingPower (W) 1E161E171E181E19 0.00 5.00x10-51.00x10-41.50x10-42.00x10-4 Vertical Lateralp Layer Doping (cm-3)Intensity (a.u.) A B

PAGE 72

72 Figure 3-9. n-layer doping depende nce intensity at 5V and peak wavelength. A) Current. B) Output Intensity. A B

PAGE 73

73 Figure 3-10.Schematic of ZnO p-n junc tion structure produ ced by P-diffusion

PAGE 74

74 Figure 3-11. The I-V characteristics at 25C from ZnO p-n junction. A) Lin ear Plot. B) Log plot. -6-4-20246 0 1 2 3 4 Current(mA)Voltage (V)ZnO pn diode-6-4-20246 10-410-310-210-1100101 ZnO pn diode Current(mA)Voltage (V)A B

PAGE 75

75 -2-1012 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 p type to p type Current(mA)Voltage (V) Figure 3-12. The I-V characteris tics measured from neighbori ng p-contacts on the P-diffused ZnO layer.

PAGE 76

76 Figure 3-13. The SIMS profiles of P, Cd a nd As in the ZnO after diffusion from a Cd3P2, arsenic and red phosphorous sources.

PAGE 77

77 CHAPTER 4 INDIUM GALLIUM ARSENIDE-BASED METAL SEMICONDUCTOR METAL OPTOELECTRONIC MIXER 4.1. Introduction Metal semiconductor metal photo detectors (MSM-PDs) have traditionally been employed as receiver elements in opto-electronic inte grated circuits (OEI C) for long wavelength telecommunication applications. This work is fo cused on their non-traditional use as detector elements in laser detection and ranging (LADAR) syst ems. Important aspect in this application is the range information attached to each pixel of data, which cannot be determined from ordinary imaging. Hence, it can be used in collision detection systems in automobiles, surveillance sensors for Robots, holographic im aging, and terrain mapping. Possi ble choices for opto-electric mixers include conventional photodiode like p-n junction diodes, pi n diodes, avalanche photodetectors (APD), and elec tron bombarded active pixel sensors (EBAPS). One major disadvantage suffered by all of them is asymme tric mixing. The othe r disadvantage is high power needed to operate APD and EBAPS. EBA PS has high gain and low noise, but they are expensive to manufacture. In cont rast, MSM detectors are self-mixi ng, they require low power to operate, they have low parasitic capacitance, they are compact and cheap to manufacture. 1.55micro meter wavelength is one of the key require ments in this system, which would make the operation of these devices eye-safe, and In0.55GaAs-based material systems were used on our system for 1.55 m applications. Schottky barrier height on InGaAs is quite low (~0.1-0.2 eV) leading to high dark current and, hence, low si gnal-to-noise ratio. To reduce dark current, the Schottky barrier height must be enhanced. Th is can be accomplished by growing (e.g. via molecular beam epitaxy) highband-gap, lattice-matched In0.52Al0.48As schottky enhancement layers (SELs) on top of the InGaAs layer. Photod etectors using SELs have been shown to yield low dark current, high res ponsivity, and high bandwidth.163-171

PAGE 78

78 To improve the performance of MSM PDs or OE Mixers, the first step is to maximize the signal to noise ratio by minimizing dark current a nd parasitics of the MSM device. Dark current and parasitic capacitance were further decreased by using “smart designs” for the MSM device by removing the semiconductor layer underneath th e final metal and finger tips. Optoelectronic mixing characteristics of InGaAs-based meta l semiconductor metal photodetector (MSMPD) employing Schottky enhancement layer and smart deigns was analyzed. Blocking of incoming light by the conventi onal Schottky metal was the most biggest hindrance in photon absorption of active semic onductor layer for MSMPDs with interdigitated fingers. Transparent conducting ox ide can be an alternative of conventional finger metal. There has been strong interest in tr ansparent conducting oxides (TCO) for optical device applications. For example, multiple junction solar cells base d on polycrystalline semiconductors represent a low-cost approach to achievi ng high efficiency photovoltaic technology, with estimates on conversion efficiencies as high as 50%. Among th e various TCO’s, indium tin oxide (ITO) is commonly used electrode for the transparent cont acts due to its excellent properties in optical transmittance, electrical conductivity, substrate adhesion, hardness, and chemical inertness.172-179 A variety of ITO deposition t echniques have been reported, including e-beam evaporation180, dc or rf sputtering181, reactive ion plating182, chemical vapor deposition183 and pulsed laser deposition.184 Sputtering and e-beam evaporation t echniques are the mostly commonly used deposition methods. ITO has also been used fo r GaAs-based metal semiconductor metal photodetector (MSMPD) to improve the responsivity by minimizing the blocking of incoming light by the inter-digitated finger contacts.112 However, the surface damage introduced by ion bombardment during the ITO spu tter deposition reduced the barrier height and subsequently increased the low dark current of the MSMPDs.36,185

PAGE 79

79 Optical and electrical properties and su rface morphologies for ITO films deposited by sputtering, e-beam evaporation and sputtered-e-beam deposited composite deposition were compared. Also, a high yield process for lifting off sputter-deposited ITO for InGaAs-based MSMPD applications with standard toluene soak ing of the resist befo re exposure during the lithography process was demonstrated. Then, MS MPDs with transparent ITO fingers were fabricated and compared with conventional in terdigitated Schottky metal contact MSMPDs. It is true that interdigitated ITO elect rodes on metal semiconductor metal photodetectors (MSM) can improve the photon responsivity by mi nimizing the shadow effect of conventional metal finger electrodes.112 However, studies of the optimal geometric design of transparent ITO finger MSM photodetectors have not been reported to date. The optimized designs employed for the more conventional metalbased interdigitated fingers186,187 are not suitable for the ITO-based transparent contacts due to the absence of shadowing effects in the latter case. Study of the effects of using ITO as the inte rdigitated electrodes for InGaAs-based MSM photodectors in terms of both external quantum efficiency and response time is presented. InGaAs-based MSM photodectors with conventiona l metal and transparen t ITO interdigitated electrodes were fabricated with various confi gurations to verify the simulation results. 4.2 Characterization of Indium Gallium Ar senide Self-Mixing De tectors for Chirp, Amplitude-Modulated LADAR (CAML) 4.2.1 Indium Gallium Arsenide-based Metal Semiconductor Metal Photodetector Fabrication InAlAs buffer layer, 2m thick In0.55Ga0.45As absorption layer, and 100 In0.52Al0.48As Schottky enhancement layer were grown on a semi-i nsulating InP:Fe substrate with a molecularbeam epitaxy (MBE) system. The as-grown sample had 62.4% absorption at 1550 nm. 2000 Si3N4 was employed for the smart design by plas ma enhanced chemical vapor deposition

PAGE 80

80 (PECVD). Windows for interdigitate d electrode contact were etched by reactive ion etcher (RIE) with CF4/O2 finished by BOE wet-etching, and 200 Ti/ 2000 Au Schottky fingers were deposited by e-beam evaporator. 1980 Si3N4 anti-reflection layer was coated, and Ti/Au final metal pads were defined after Si3N4 window etching for final pad c onnection. All levels of the patterning processes were made by standard optical lithography. 4.2.2 DC Response 90 m 90 m active area, 1 m wide fingers, and 2 m spacing device was chosen for the following DC and mixing reponse measuremen t. Fingure 4-1. shows the diagram of DC responsivity measurement setup. LDC is laser diode controller (ILX Lightwave LDC-3752), LD is laser diode(Multiple x MTX510EW fiber pigtailed 1550 nm lase r), and FA is fiber attenuator. Also, Programmable voltage source (Keithley 230), Laser diode controller (ILX Lightwave LDC-3752), Variable optical attenuator (JDS Op tics 6500L), and Optical power meter (Newport multifunction optical meter 2835-C) were used for measurement. Dark-current of MSMPD is shown in Figure 4-2. Very small knee voltage whic h is defined as the bias at which the current equals 5% of the current at the second region wher e dI/dV is low is observed. The dark-current at 2V was 2.3nA. DC responsivity of device at 10 nW, 100 nW, 1 W, 10 W, and 100 W incident optical power is illustrated in Figure 43. DC responsivity was approximately 0.5 A/W. At power levels of interest (10 nW and below) , the knee voltage was less than 80 mV. Even at 100 W, the knee voltage was 0.26 V. 4.2.3. Mixing Responsivity Mixing responsivity setup for MSM optoelectric mixer is s hown in Figure 4-4. SA is spectrum analyzer (Advantest R3271), SGRF is signal generator, SGLO is signal generator (Wavetek 2500A), TZA is trans-impedance amplifie r (Stanford Research Systems 570), and Opt.

PAGE 81

81 cap. is optional capacitor for th e low-pass filtering to remove the RF components from MSM output. Mixing responsivity was measured for severa l different incident optical power levels, 10 nW, 1 W and 100 W. Figure 4-5 shows the mixing responsivity at 100 W dc optical power (corresponding to ~ 35.4 W rms AC power) for VLO of 20, 22, 24, 26 and 27 dBm, and a constant IF of 10 kHz. The data has been normalized with the la ser’s frequency response curve. Mixing response improved considerably with increasing VLO. The maximum mixing responsivity was 0.24 A/W. Mixing responsivity at 1 W and 10 nW dc optical power (corresponding to ~ 354 nW and 3.54 nW rms AC power, respectively) for a VLO of 26 dBm, and an IF of 10 kHz is plotted in Figure 4-6. The mixing responsivity is quite flat, with a peak va lue of approximately 0.22 A/W. As can be seen from figures 4-5 and 4-6, a relati vely flat mixing responsivity can be obtained at a VLO of 26 dBm, independent of the incident op tical power. Figure 4-7 sh ows the IF dependence of the mixing responsivity at 100 W dc optical power (corresponding to ~ 35.4 W rms AC power). The IF frequencies rang e from 10 kHz to 10 MHz. There is a slight decrease in mixing responsivity with increasing IF frequency. Howeve r, the responsivity levels off at an IF of 1 MHz, with a peak responsivity of 0.21 A/W. 4.3. Metal Oxide Metal Photod etector with Transparent Indi um Tin Oxide Interdigitated Schottky Contact 4.3.1. Deposition of Transparent Indium Tin Oxide For the e-beam evaporation, the ITO films were deposited onto silic on substrates with a CHA Mark II system. A chamber pressure of 10-7 Torr was maintained during deposition and a deposition rate of ~1/sec was used in all cases . For the sputter deposition, the ITO films were deposited by Ar plasma 3 % O2 chemically-assisted rf sputtering at pressures of 4 mTorr and an rf power of 125 W at 13.56 MHz. The composition of the target was 90 % In2O3 and 10 % SnO2

PAGE 82

82 and the deposition rates were also ~1 /sec. For the composite film, a 100 e-beam deposited ITO film was first deposited on th e substrate and a thicker overlay er ITO film was subsequently deposited in the sputtering system. Typically, a total thickness of 2000 for the ITO films was used throughout this work. The composition of the ITO films was analyzed with a Physical Electronics 660 scanning Auger mi croprobe. The resistivity of th e ITO films was obtained from standard four-point measurements at 300K. Tran smittance and reflectance of the films deposited by the different methods were measured using a Perkin-Elmer spectrometer 19. In addition to the thin film deposition studies, patterned ITO inter-digitated fingers were also fabricated with the sta ndard toluene soaking process du ring the lithography process using S1808 photoresist. Scanning Electron Microscopy (SEM) images were taken before and after the ITO lift-off. 4.3.2. Characterization of Transparent I ndium Tin Oxide Films and Indium Tin Oxide Schottky Contact Metal Semiconductor Metal Photodetectors Figure 4-8 shows the normalized reflectance, transmittance, and absorbance of 2000 of both the e-beam deposited films as well as 2000 sputtered ITO films. The sputtered ITO film showed 83% transmission, whic h was significant higher than the value of 13% of the transmittance for the e-beam deposited ITO. A resistivity of 6.6210-4 cm-1 was obtained for the sputtered ITO films, which was almost twoorders of magnitude less than the value of 1.6410-2 cm-1 for the ITO deposited with e-beam eva poration. Table 1 su mmarizes the optical and electrical properties for the e-beam and s putter-deposited ITO. Thes e significant differences of optical and electrical prope rties between e-beam and sputtered ITO were not caused by differences in the oxide composition. AES surf ace scans and depth profiles were used to evaluate the composition of the ITO films. Th ere was no measurable difference obtained, as

PAGE 83

83 illustrated in Figure 4-9, which shows the Auger surface scans of e-beam (A) and sputtered (B) ITO films. However, the morphology of ITO films displa yed a strong dependence on the deposition method. Figure 4-10 shows SEM images of the ebeam and sputtered ITO films. The grain size of sputtered ITO film was much smaller that in the e-beam deposited ITO. The grain size distribution was also more unifo rm for sputtered ITO films a nd the film was more densely packed. This could be the reason why the sputte red films showed lower resistivity and better transmission, since grain boundaries are known to degrade carrier mobility and also to increase absorption of light in TCOs.36 Although the sputtere d ITO films have previously shown superior transmittance and conductivity, the ion bombardment damage introduced by the sputtering has been shown to limit the use of spu ttered ITO for MSMPD applications.185,188 We proposed to use e-beam/sputtering deposit ed composite ITO films to reduce the ion bombardment damage. Although the e-beam deposite d ITO had inferior op tical and electrical properties, there was no ion bombardment da mage during the deposit ion. A 100 e-beam deposited ITO was deposited on the top of the se miconductor and the thicke r sputtered ITO film was subsequently deposited on the top of e-beam deposited ITO film. Since the e-beam deposited film was quite thin, around 5% of th e entire ITO film thic kness, the optical and electrical properties only degraded a small amount , as illustrated in Table 1. A resistivity of 7.60-4 4 cm-1 was obtained, which was 15% larger than that of the sputtered ITO films. The transmission at 1.55 m was reduced from 83% to 77% for sputtered and composite ITO film, respectively, as shown in Figur e 4-11 (A). This 100 e-beam deposited ITO film was thick enough to shield the semiconductor surface from ion bombardment damage during the sputterdeposition.

PAGE 84

84 Thin Ni/Au (50/50) or Ti/A u (50/50) bilayers have be en employed as transparent contacts for GaN-based light-emitting diodes.129 The resistivity of a composite 50 Ti/50 Au/1900 sputtered ITO stack was 3.08-4 cm-1, which was the best among all the ITO films in our study. However, the light reflectance incr eased significantly and transmittance reduced to only around 35%, as shown in Figur e 4-11 (B). This woul d degrade the MSMPD performance considerably. Typically, etch-back processes are used to make the patterns on sputtered films. However, the etch-back process introduces ion bombar dment damage on the semiconductor during the sputter deposition stage. To avoid this probl em, the standard positive photoresist lift-off technique had to be used to define the pattern of the composite ITO films. With the assistance of the conventional toluene soaking effect during lithography, 1m line and space inter-digitated composite ITO fingers were successfully demonstr ated, as illustrated in Figure 4-12. Excellent edge definition and yield >99% were obtained. 60 finger InGaAs-based MSM devices with finger dimension of 90 m 2 m and 2 m gap between fingers using Ti/Au, composite ITO, or sputtered ITO as the finger were fabricated. The sputtering power was kept low to mini mize surface damage during the ITO deposition. Figure 4-13 shows the dark current of the MSM de vices and optical responses. The dark current of these three devices were in a similar range, around 1.5 nA. The device with sputtered ITO did not show an appreciably higher dark current level, however it displayed an early breakdown voltage around 0.2V. This could result from the low bias voltage used for the ITO sputtering. The photo-response of both composite and sputte red ITO MSM showed more than double the photo response for the MSM device with Ti/Au fingers, since the dimension of the gaps and

PAGE 85

85 fingers of the MSM was the same. The shadow effect of the Ti/Au fi ngers reduces the photo response in those devices. 4.4. Design of Transparent Indium Tin Ox ide-based Interdigitated Fingers for Metal Semiconductor Metal Photodetector 4.4.1. Fabrication of Indium Gallium Ar senide-based Metal Semiconductor Metal Photodetector with Transparent Indium Tin Oxide Electrodes The InGaAs MSM photodetector structure cons isted of 150 of InAlAs as a Schottky enhancement layer, a 230 In(G a,Al)As graded layer, a 1.0 m InGaAs absorption layer, and a 0.3m InAlAs buffer layer grown on a semi-insulat ing InP:Fe substrate with a molecular beam epitaxy system(MBE).171 The device fabrication started w ith a blanket deposition of SiNx using a low power plasma enhanced chemical vapo r deposition (PECVD) system. BOE-based wet chemical etching was used for opening the interdigitated finger contact-window. A 200 Ti/1800Au deposition with an electron-beam evapor ator was used for the contacts to the conventional metal-based interdigit ated electrode devices. For th e devices with transparent ITO fingers, 2000 ITO was deposited with a plasma-assis ted rf sputtering system using Ar/3% O2 discharges at a pressure of 4 mTorr and 125 W of rf power (13. 56 MHz). The sputtering target for the ITO was a composite target consisting of 90 % In2O3 and 10 % SnO2. The testing contact pads and the tips of the interdigitated electrode s were sitting on the SiNx layer to reduce dark current and parasitic capacitance.167 The dc photo-currents of both Ti/Au and ITO-based MSM photodetectors were measured with an HP4156C parameter analyzer.

PAGE 86

86 4.4.2. Design of Transparent Interdig itated Indium Tin Oxide Contact Metal Semiconductor Metal Photodetector The configuration of the simulated MSM device us ed in this work is illustrated in Figure 414. In this Figure, D is the finger width of the metal electrodes and t is the gap spacing between fingers. The active area of the MSM device is th en L (t + D) total number of the metal electrodes. The shadowing of th e active area by the interdigitated metal electrodes is the key drawback of conventional MSM detectors. Due to the shadowing effect, the external quantum efficiency, ex, is reduced by a factor of t/(T + D), wh ich can be expressed as Equation 4-1. )] exp( 1 [ ) 1 (d D t t rex (4-1) r is the reflectivity at the measurement wavelength, is the absorption coefficient at this wavelength, and d is the thickne ss of semiconductor active layer187,189 . Figure 4-15 shows the external quantum efficiency of an 1 m thick InGaAs-based MSM device as a function of finger gap spacing as well as finger widths for reflectiv ity and absorption coefficient of 0.288 and 1.8 m-1, respectively.190,191 The external quantum efficiency in creases with finger gap spacing since a larger gap provides more abso rption area. The narrow finge r width reduces the shadowing effect of the metal fingers and therefore exhibits higher external quantum efficiency. However, the resistance of the finger increases significan tly when the finger becomes too narrow and the high-speed performance of the MSM device will degrade. By replacing the metal fingers with tran sparent ITO fingers the external quantum efficiency can be significantly improved, as sh own in Figure 4-16 (A). The improved external quantum efficiency can be expressed as Equation 4-2. )] exp( 1 [ ) 1 ( 825 . 0 )] exp( 1 [ ) 1 (d D t D r d D t t rex (4-2)

PAGE 87

87 The external quantum efficiency of the MSM devices with ITO fingers consists of the components from the active layer between the fingers and from the active layer under the transparent contacts, as shown in Figure 4-16 (B). The coefficien t, 0.825, in the second term of the equation is the measured transmittance of 2000 thick ITO fingers for the 1.55 m wavelength light used in this work. As show n in the Figure 4-15 and Figure 4-16 (A), the external quantum efficiency of the MSM devices with ITO fingers is much higher than that of the MSM devices with metal-based interdigitated fingers. Unlike the MSM devices with metalbased interdigitated fingers, the ITO-based MSM device does not require th e use of very narrow fingers to obtain a higher external quantum effi ciency. The external quantum efficiency of the MSM devices with ITO fingers is almost i ndependent of the finger gap spacing and the dependence of the external quantum efficiency on the finger width is also minimal, as illustrated in Figure 4-16 (A). This provi des a great advantage in that it is possible to use wider interdigitated ITO fingers to reduce the finger parasitic resistance, maintain the high speed operation of the devices and at the same time not sacrifice the external quantum efficiency. Figure 4-17 shows the simulated and measured ex ternal quantum efficiency for MSMs with 1m wide ITO and metal fingers for various finger gap spacing. There is excellent agreement between simulated and measured results. Transparent ITO MSM devices exhibit much higher external quantum efficiency than that of conventional metal finger MSM devices. The speed of the MSM devices is limited by the transit time of the optically generated carriers and the RC-time constant of the interdig itated fingers. Besides the external quantum efficiency, high speed operation is also one of critical factors for MSM photodetectors. To achieve this goal, a trade-off between minimizi ng transit time and RC time delay is required.

PAGE 88

88 The total response time, , can be determined by combining carrier transit time, tr and RC time, RC 187,189,192,193 2 / 1 2 2) (RC tr (4-3) V ttr2 , RCRC2 . 2 , ) 1 1 ( 2 1 14 4 4h eV V V , is the carrier drift corrective coefficient, and Ve and Vh are saturation velocity of electrons and ho les, respectively. For RC time constant, the capacitance of the interdigit ated contact system can be obtained from Equation 4-4.67,187,194,195 ) 37 . 2 08 . 1 5 . 6 )( 1 ( 226 . 02 0 sNL C (4-4) ) ( 2 D t L N , and D t D , 0 is the permittivity in vacuum, and s is the relative dielectric constant of semiconductor, and L is finger length. Figure 4-18 illustrates the response time depe ndences of finger gap spacing and finger width for an ITO-based device wi th 2m finger width, the same dimension for the finger spacing and 100 m 100 um active area. For the regi on of smaller finger gaps, the RC time delay dominates the response time. For the devices with la rger finger gaps, the e ffect of carrier transit time dominates the response time. In other words, to the left of the minimum, the response is dominated by RC time constant, whil e the right side is controlled by carrier transit time. For the ITO MSM devices, high speed and high external quantum efficien cy devices can be achieved with very relaxed 2 m finger width design. For the metal-ba sed MSM device, submicron finger width design is needed to obt ain high external quantum efficiency. However the use of submicron finger width will significantly degrade the high speed performance.

PAGE 89

89 Table 4-1. Sheet resistance, resistivity, transmittance at 1.5 m, and index of refraction of different stack samples Sample Stack Sheet Resistance ( / ) Resistivity ( cm) Transmittance at 1.55 m (%) Index of Refraction 2000 e-beam ITO 821.0 1.64-2 13 2.248 2000 sputter ITO 33.1 6.62-4 83 2.008 50 Ti / 50 Au / 1900 sputter ITO 15.4 3.08-4 28 100 e-Beam ITO / 1900 sputter ITO 38.0 7.60-4 77 2.109

PAGE 90

90 A LDC LD FA PoptLens system 50Vdc Figure 4-1. The DC responsivity measurement setup.

PAGE 91

91 Figure 4-2. Dark-curre nt of MSM photodetector Bias (V)

PAGE 92

92 Figure 4-3. The DC responsivity of MSM photodetector at diffe rent incident optical power. Green : 10 nW, black : 100 nW, brown : 1 W, blue : 10 W, red : 100 W. Bias (V)

PAGE 93

93 LDC LD FA PoptLens system 50 SGRFSA SGLO Opt. cap. TZA Figure 4-4. Mixing responsiv ity measurement setup.

PAGE 94

94 Figure 4-5. The VLO dependence of the mixing responsivity at 100 W incident power. Red : 20 dBm, blue : 22 dBm, brown : 24 dBm, black 26 : dBm, green : 27 dBm. IF is 10 kHz.

PAGE 95

95 Figure 4-6. Optical power dependence of the mixing responsivity. Blue : 1 W, red : 10 nW.

PAGE 96

96 Figure 4-7. The IF dependence of the mixing responsivity at 100 W (dc) incident optical power. Red : 10 kHz, blue : 100 kHz, br own : 1 MHz, black : 10 MHz. VLO is 27 dBm.

PAGE 97

97 Figure 4-8. Normalized reflectance, transmittanc e, and absorbance. A) 2000 thick e-beam deposited ITO. B) 2000 thick sputtered ITO. 8001200160020002400 0 10 20 30 40 50 60 70 80 90 100 Reflectance Transmittance Absobance Percentage (%)Wavelength (nm)8001200160020002400 0 10 20 30 40 50 60 70 80 90 100 Reflectance Transmittance Absobance Percentage (%)Wavelength (nm)A B

PAGE 98

98 Figure 4-9. The AES spectra. A) e-beam deposited ITO film. B) Sputtered ITO film. 02004006008001000 -3000 -2000 -1000 0 1000 2000 3000 In Sn O Intensity (a.u.)Kinetic Energy (eV) C02004006008001000 -2000 -1000 0 1000 2000 3000 Intensity (a.u.)Kinetic Energy (eV)In Sn O CA B

PAGE 99

99 A B A B Figure 4-10. The SEM images. A) e-beam de posited ITO film. B) Sputtered ITO film.

PAGE 100

100 Figure 4-11. Normalized reflectance, transmittan ce, and absorbance. A) 100 e-beam ITO/1900 sputtered ITO film. B) 50 Ti/ 50 Au/1900 sputtered ITO film. 8001200160020002400 0 10 20 30 40 50 60 70 80 90 100 Reflectance Transmittance Absobance Percentage (%)Wavelength (nm)8001200160020002400 0 10 20 30 40 50 60 70 80 90 100 Reflectance Transmittance Absobance Percentage (%)Wavelength (nm)A B

PAGE 101

101 Figure 4-12. Lifted-off 2000 sputter ITO. A) Op tical microscopy image. B) The SEM image.

PAGE 102

102 Figure 4-13. Dark current and optical response. A) The Ti/Au MSM device. B) Sputtered ITO MSM device. C) Compos ite ITO MSM device. A B 1 0 -111 0 -1010-910-810-7 Optical Response Darkcurrent 10-1110-1010-910-810-7 Optical Response Darkcurrent Current (A) 0.00.51.01.52.0 1 0 -111 0 -1010-910-810-7 Optical Response Darkcurrent Voltage (V) C

PAGE 103

103 Figure 4-14. Schematic illustration of an interdigitated MSM photodetector. t L D

PAGE 104

104 012345 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Finger Width 0.1 um 1 um 2 um 3 um External Quantum EfficiencyFinger Gap (um) Figure 4-15. External quantum efficiency of In0.55Ga0.45As MSM devices with metal-based interdigitated fingers as the function of finger gap spacing and finger width.

PAGE 105

105 Figure 4-16. External quantum efficiency of In0.55Ga0.45As MSM devices with transparent ITObased interdigitated fingers. A) Finger gap spacing and finger width dependence. B) Contributions to the extern al quantum efficiency of an ITO-based MSM device. 012345 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 Finger Width 0.1 um 0.5 um 1 um 2 um 3 umExternal Quantum EfficiencyFinger Gap (um) 012345 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Efficiency of Gap Area Efficiency of Finger Area Total Efficiency 2um Wide FingersFinger Gap (um)External Quantum Efficiency A B

PAGE 106

106 012345 0.0 0.1 0.2 0.3 0.4 0.5 0.6 1um Finger Width Calculated Efficiency of ITO Measured Efficiency of ITO Calculated Efficiency of metal Measured Efficiency of metalFinger Gap (um)External Quantum Efficiency Figure 4-17. Simulated and experimental external quantum efficiency of conventional metal and transparent ITO fingers for 1um finger width MSM phtodetectors.

PAGE 107

107 012345 0.0 2.0x10-114.0x10-116.0x10-118.0x10-111.0x10-10 Transit Time RC Time Total Response TimeActive Area 100 x 100 um 2 um Finger WidthResponse Time (s)Finger Gap (um) Figure 4-18. Response time dependence on finge r gap spacing and width for 100x100 um active area.

PAGE 108

108 CHAPTER 5 SUMMARY AND FUTUREWORK 5.1. Gallium Nitride Enhancement Mode Metal Oxide Semiconductor Field Effect Trainsiter The optimum condition for Si diffusion into GaN to reduce the contact resistance of source-drain regions and to obtain the smooth surface around gate regions of GaN MOSFET was studied. The effect of encapsulant type, Si s ource layer thickness, and temperature for Si diffusion on the conductivity and morphology of GaN was examined. After Si diffusion into GaN at 1000 C for 120 mins with 200 Si source layer and 1000 SiO2 encapsulant, the specific contact resistivity was 6.9-5 cm2 and the root mean square roughness was 1.5 nm. The diffusion coefficient of 2.07x10-4 cm2sec-1 and the activation energy of 0.57 eV was calculated from the SIMS results. An n-cha nnel enhancement mode MgO/GaN-on-Si MOSFET using Si diffusion to convert th e source-drain region to smooth n+ was demonstrated. The maximum drain current was around 35 A/mm at 7V gate voltage and the transconductance was 7.0 S.mm-1. The devices showed good dc characteristi cs and demonstrate the high quality of gate oxides grown by Molecular Beam Epitaxy (MBE). Future work will focus on the effect of temperature on breakdown and I-V characteristics, since the MOS-GaN transistors should have ex cellent high temperatur e performance. Also, molecular beam epitaxy (MBE) grown AlN encapsulant over a thin silicon layer for the diffusion is proposed as an alternative to avoid the passivation of magnesium acceptors by hydrogen under the gate regions. 5.2. Zinc Oxide Light-Emitting Diode ISE TCAD simulations were performed to ex amine the effect of active, nand p-layer doping and thickness on the optical ou tput intensity and current-vol tage characteristics of both vertical and lateral geometry ZnO LEDs. The latter geometry is attractive for ultra-low structures

PAGE 109

109 with the ZnO deposited on glass substrates. The cu rrent density distribution is more uniform in the vertical structures but there is little difference in optical ou tput power as a function of doping or layer thickness betwee n the two geometries. The results of our simulations of ZnO LED structures may be summarized as follows: i. There is little difference in the output intensity from the vertical and lateral geometry LEDs under the same bias condi tions, although this does not take into account that the latter will be grown on a cheap, latticemismatched substrate such as glass in which case the materials qualit y will be poorer and w ill have an effect on emission intensity. This may also limit the ability to achieve acceptable p-type doping levels, although the p-layer will be on the top and benefit from having a significant amount of ZnO underneath as is the case with nitride-based LEDs. ii. The simulations show the importance of having highly doped layers on both sides of the junction, through their effect on recombination rate. iii. The emission is mainly from the region close to the p-layer because of the higher electron mobility and in the case of th e mesa (lateral) structure all of the recombination occurs in the region adjacen t to the mesa, emphasizing the need for ring contact structures an d transparent conducting oxides for spreading of the current in the p-layer. p-n junction diodes have been re alized in bulk n-type crysta ls using P diffusion. p-type ZnO layers have been formed in lightly n-type (1017 cm-3) bulk, single-crystal ZnO substrates by diffusion of P from a Cd3P2 , arsenic and red phosphorous dopan t source in a closed-ampoule system. The P incorporation depth was found to be ~200 nm after diffusion at 550C for 30 mins, as determined by Secondary Ion Mass Spec trometry profiling. The resulting structures

PAGE 110

110 show rectification, with on-o ff current ratios of ~70 at +3/5V. The forward current ideality factor was 2, consistent with multiple current transport mechanisms present in the junction, such as defect-assisted tunneling and conventional carrier recombination in the space-charge region via midgap deep levels. The forward turn-on voltage, VF was ~ 4 V at 300 K with a specific on-state resistance ( RON) of ~21 m cm2 .The activation energy of the forward current at low forward biases was ~1.4 eV. This is also co nsistent with carrier recombination in the space charge region via a mid-gap deep level. For the future work, forming gas annealing af ter P diffusion can be employed to passivate the defects in ZnO and enhance the radiative reco mbination. Group I elements such as Li and Na can be diffused or implanted into bulk ZnO to form the p-type dopants substituting zinc sites since the ionization energy of those elements is shallower than that of nitrogen in ZnO. 5.3. Indium Gallium Arsenide Metal Semiconductor Metal Photodetector InGaAs-based Metal Semic onductor Metal (MSM) Photodetect ors were fabricated and tested for use of opto-electronic mixers. Higher energy bandgap, lattice-matched InAlAs Schottky enhancement layers (SEL’s) grown on top of the InGaAs layer were employed to improve the barrier height. Dark current and parasitic capacitance were further minimized by removing the semiconductor underneath the final metal using Si3N4 dieletric layer. Photodetector using Schottky enhancement layer was shown to yi eld low dark current, high dc responsivity of 0.5 A/W and flat mixing responsivit y with an average of 0.2A/W. The optical and electrical properties and surface morphol ogy of sputtered, e-beam deposited and composite e-beam/sputtered IT O films for Metal-Semiconductor-Metal (MSM) photo-detector applications were compared. The resistivity of sputtered ITO was almost twoorders of magnitude less than that of e-beam deposited ITO, while the resistivity of the composite films was only 15% higher than that of sputtered films. The transmittance at 1.55 m

PAGE 111

111 of sputtered and composite ITO films was 83% and 77% respectively, which was much higher than the value of 13% for the e-beam deposite d ITO. The composition of ITO films for both deposition techniques was very similar, as de termined from Auger el ectron spectroscopy surface scans and depth profiles. The s puttered films showed much smoother surfaces and smaller grain size. InGaAs-based MSM devices with ITO fing ers have been fabricated and showed higher photo response than that of MSM devices with conventional Ti/Au fingers. Simulations were performed to determine the optimum geometric design of In0.55Ga0.45As metal-semiconductormetal (MSM) photodetectors. The purpose was to maximize external quantum efficiency with transparent indium tin oxide (ITO) based in terdigitated fingers. Both ITO finger and conventional metal finger MSM photodetectors were fa bricated to verify the simulated external quantum efficiency and good agreement between experimental and simulated results was obtained. In all cases, the IT O provided higher effici encies, particularly at short finger gap spacing. Using the transparent ITO fingers, high external quantum efficiency and high peed performance can obtained with very relaxed design. Annealing of ITO electrode should be studied to improve the conductivity and transmittance for the future work. Instead of tran sparent ITO interdigitat ed fingers, a backside illumination MSM photodetector can be fabricated using flip chip bonding. In this case, only a minimum response time will be the focus on fo r the optimum design of MSM photodetectors.

PAGE 112

112 LIST OF REFERENCES 1 Z. Z. Bandic, P. M. Bridger, E. C. Piquette, T. C. McGill, R. P. Vaudo, V. M. Phanse, and J. M. Redwing, Appl. Phys. Lett. 74, 1266 (1998). 2 R. Gaska, J. W. Yang, A. Osinsky, Q. Chen, M. A. Khan, A. O. Orlov, G. L. Snider, and M.S. Shur, Appl. Phys. Lett. 72, 707 (1998). 3 T.D. Moustakas, T. Lei, and R. J. Molnar, Physica B 185, 39 (1993). 4 J. M. V. Hove, R. Hickman, J. J. Klaassen, P. P. Chow, and P. P. Ruden, Appl. Phys. Lett.70, 282 (1997). 5 W. Gotz, L.T. Romano, J. Walker, N.M. Johnson, and R.J. Molnar, Appl. Phys. Lett. 72, 1214 (1998). 6 C. Wetzel, D. Volm, B. K. Meyer, K. Pressel, S. Nilss on, E. N. Mokhov, and P. G. Baranov, Appl. Phys. Lett. 65, 1033 (1994). 7 Y. A. Vodakov, E. N. Mokhov, A. D. Roenkov, M. E. Boiko, and P. G. Baranov, J. Cryst. Growth, 10, 189 (1998). 8 M. Morita, S. Isogai, N. Shimizu, K. Ts ubouchi, and N. Mikohiba, Jpn. J. Appl. Phys. 19, 173 (1981). 10 S. Keller, P. B. Keller, Y. F. Wu, B. Heying, D. Kapolnek, J. S. Speck, U. K. Mishra, and S. P. DenBaars, Appl. Phys. Lett. 68, 1525 (1996). 11 S. Nakamura, T. Mukai, and M. Senoh, Appl. Phys. Lett. 64, 1687 (1994). 12 Y. Kato, S. Kitamura, K. Hiramatsu, and N. Sawaki, J. Cryst. Growth 144, 133 (1994). 13 H. Marchand, J. P. Ibbetson, P. T. Fini, P. Kozodoy, S. Keller, S. DenBaars, J. S. Speck, and U. K. Mishra, Mater. Res. So c. Int. J. Nitride Semi Res. 3, 3 (1998). 14 J. Burm, K. Chu, W. Davis, W. J. Schaff, L. F. Eastman, and T. J. Eustis, Appl. Phys. Lett. 70, 464 (1997). 15 W. Gotz, N. M. Johnson, J. Walker, D. P. Bour, H. Amano, and I. Akasaki, Appl. Phys. Lett. 67, 2606 (1995). 16 H. Amano, M. Kito, K. Hiramatsu, I. Akaski, Jpn. J. Appl. Phys. 28, 2112 (1989). 17 S. Nakamura, T. Muskai, M. Senoh, and N. Iwasa, Jpn. J. Appl. Phys. 31, 139 (1992). 18 M. A. Khan, A. R. Bhattarai, J. N. Kuznia, and D. T. Olson, Appl. Phys. Lett. 72, 211, (1998).

PAGE 113

113 19 M. A. Khan, J. M. Kuznia, D. T. Olson, W. Schatt, J. Burm, and M. S. Surf, Appl. Phys. Lett. 63, 1214, (1993). 20 S. T. Sheppard, K. Doverspike, W. L. Pribble, S. T. Allen, J. W. Palmour, L. T. Kehias, and T. J. Jenkins, IEEE. Electron. Dev. Lett. 20, 161, (1999). 21 L. Daumiller, C. Kirchner, M. Kamp, K. J. Eb eling, L. Pond, C. E. Weitzel, and E. Kohn, Porceedings of the 56th Development Research Confern ece, Charlottesville, VA, (1998). 22 Z. Z. Bandic, P. M. Bridger, E. C. Pi queette, and T. C. McGill, Appl. Phys. Lett. 74, 1266, (1999). 23 T. G. Zhu, D. J. H. Lambert, B.S. Shelt on, M.M. Wong, U. Chowdrury, H. K. Kwon, and R.D. Dupuis, IEEE. Electron. Lett. 36, 1972, (2000). 24 L.S. McCarthy, P. Kozodoy, M. J. W. Rodwe ll, and S. P. DenBaars, IEEE. Electron. Dev. Lett. 20, 227, (1999). 25 J. B. Limb, H. Xing, B. Morau, L. Mccart hy, S. D. Denbaars, and U. K. Mishro, Appl. Phys. Lett. 76, 2457, (2000). 26 J. W. Johnson, B. P. Gilla, B. Luo, K. P. Lee, C. R. Abernathy, S. J. Pearton, J. I. Chyi, T. E.Nee, C. M. Lee, C. C. Cho, and F. Ren, J. Electrochemi. Soc. 148, 303, (2001). 27 A. Tsukazaki, A. Ohtomo, T. Onuma, M. Ohtani, T. Makino, M. Sumiya, K. Ohtani, S. F. Chichibu, S. Fuke, Y. Segawa, H. Ohno, H. Koinum a, and M. Kawasaki, Nat. Mater. 4, 42 (2005). 28 A. Tsukazaki, M. Kubota, A. Ohtomo, T. Onum a, K. Ohtani, H. Ohno, S. F. Chichibu, and M. Kawasaki, Jpn. J. Appl. Phys. 21, 643 (2005). 29 E. Olsson and G. L. Dunlop, Appl. Phys. Lett, 66, 3666, (1999). 30 J. Koike, K. Shimoe, and H. Ieki, Jpn. J. Appl. Phys. 37, 2337, (1993). 31 M. P. Roe, B. Wacoqnem, and C. N. Pannell, IEEE. Phot. Tech. Lett. 2, 1026, (1996). 32 Y. Nakanishi, A. Miyake, H. Kaminami, T. Aoki, Y. Hatataka, and G. Shimaoka, Appl. Surf. Sci. 142, 233, (1999). 33 J. Xu, Q. Pan, Y. Shun, and Z. Tian, Sens. Actu. B. 66, 277, (2000). 35 E. M. C. Fortuato, P. M. C. Ba rguinha, A. C. M. B. G. Pimetel, A. M. F. Gocalves, A. J. S. Marques, R. F. P. Martins, and L. M. N. Pereira, Appl. Phys. Lett., 85, 2541, (2004). 36 H. L. Hartnagel, A. L. Dawar, A. K. Ja in, and C. Jagadish, Semiconducting Transparent Thin Films (IOP Publishing, Bristol, 1995). 37 D. C. Look, Mater Sci. Eng. B 80, 383 (2001).

PAGE 114

114 38 P. Zu, Z. K. Tang, G. K. L Wong, M. Kawasaki, A. Ohtomo , and K. Koinuma, Solid State Commun. 103, 459 (1997). 39 D. M. Bagnall, Y. R. Chen, Z. Zhu, T. Yao, S. Koyama, and M. Y. Shen, Appl. Phys. Lett. 70, 2230 (1997). 40 M. Wraback, H. Shen, S. Liang, C. R. Gorla, and Y. Lu, Appl. Phys. Lett. 74, 507 (1999). 41 J. M. Lee, K. K. Kim, S. J. Park , and W. K. Choi, Appl. Phys. Lett. 78, 2842 (2001). 42 J. E. Nause, III–V’s Rev 12, 28 (1999). 43 Y. Chen, D. Bagnell, and T. Yao, Mater Sci. Eng. B 75, 190 (2000). 44 D. C. Look, D. C. Reynolds, J. W. Hemsky, R. L. Jones, and J. R. Sizelove, Appl. Phys. Lett. 75, 811 (1999). 45 D. C. Look, J. W. Hemsky, and J. R. Sizelove, Phys. Rev. Lett. 82, 2552 (1999). 46 F. D. Auret, S.A. Goodman, M. Hayes, M. J. Legodi, H. A. van Laarhoven, and D. C. Look,Appl. Phys. Lett. 80, 956 (2002). 47 S. O. Kucheyev, J. E. Bradley, J. S. Williams, C. S. Jagadish, and M. V. Swain, Appl. Phys. Lett. 80, 956 (2002). 48 D. C. Reynolds, D. C. Look, a nd B. Jogai, Solid State Commun. 99, 873 (1996). 49 H. Amano, N. Sawaki, I. Akasak i, Y. Toyada, Appl. Phys. Lett. 48, 353 (1986). 49 T. Aoki D.C. Look, Y. Hatanaka, Appl. Phys. Lett. 76, 3257 (2000). 50 C. C. Chang, and Y. E. Chen, IEEE Tran s. Ultrasonics. Ferroel ectr. Freq. Control. 44, 624 (1997). 51 P. M. Verghese, and D. R. Clarke, J. Appl. Phys. 87, 4430 (2000). 52 C. R. Gorla, N. W. Emanetogl u, S. Liang, W. E. Mayo, Y. Lu, and M. Wraback, J. Appl. Phys. 85, 2595 (1999). 53 H. Ohta, K. Kawamura, M. Orita, M. Hirano, N. Sarukura, and H. Hosono, Appl. Phys. Lett. 77, 475 (2000). 54 M. Joseph, H. Tabata, and T. Kawai, Jap. J. Appl. Phys. 38, 1205 (1999). 55 S. Krishnamoorthy, A. A. Iliadis, A. Inumpudi, S. Choopun, R. D. Vispute, and T. Venkatesan , Solid State Electron. 46, 1631 (2002). 56 Y. Li, G. S. Tompa, S. Liang, C. Gorla, C. Lu, and J. Doyle, J. Vac. Sci. Techol. A 15, 1663 (1997).

PAGE 115

115 57 L. K. Singh and H. Mohan, Indian J. Pure. Appl. Phys. 13, 486 (1975). 58 A. R. Hutson, Phys. Rev. 108, 222 (1957). 59 B. J. Jin, S. H. Bae, S. Y. Lee, and S. Im, Mater Sci. Eng. B 71, 301 (2000). 60 D. M. Hofmann, A. Hofstaetter, F. Leiter, H. Zhou, F. Henecker, and B. K. Meyer BK, Phys. Rev. Lett. 88, 45504 (2002). 61 C. G. V. Walle, Phys. Stat. Solidi. B 229, 221 (2002). 62 S. F. J. Cox, F. A. Davis, P. J. C. King, J. M. Gil, H. V. Alberto, a nd R. C. Vilao, J. Phys. Cond. Mater. 13, 9001 (2001). 63 C. G. V. Walle, Phys. Rev. Lett. 85, 1012 (2000). 64 Gutowski, N. Presser, an d I. Broser, Phys. Rev. B 38, 9746 (1988). 65 S. Bethke, H. Pan, and B. W. Wessels, Appl. Phys. Lett. 52, 138 (1988). 66 B. J. Jin, S. Im, S. Y. Lee, Thin Solid Films. 366, 107 (2000). 67 D. M. Bagnall, Y. F. Chen, M. Y. Shen, Z. Zhu, T. Goto, and T. Yao, J. Cryst. Growth. 184,605 (1998). 68 E. G. Bylander, J. Appl. Phys. 49, 1188 (1978). 69 N. Riehl, and O. Z. Ortman, Electrochem. 60, 149 (1952). 70 F. A. KrCoger, and H. J. Vink, J. Chem. Phys. 22, 250 (1954). 71 I. Y. Prosanov, and A. A. Politov, Inorg. Mater. 31, 663 (1995). 72 N. Y. Garces, L. Wang, L. Bai, N. C. Giles, L.E. Halliburton, and G. Cantwell, Appl. Phys. Lett. 81, 622 (2002). 73 H. J. Egelhaaf, and D. J. Oelkrug, J. Cryst. Growth, 161, 190 (1996). 74 D. Hahn and R. Nink, Phys. Cond. Mater, 3, 311 (1965). 75 B.J. Pierce and R.L. Hengehold, J. Appl. Phys. 47, 644, (1976). 76 S. A. Studenikin, N. Golego, and M. Cocivera, J. Appl.Phys. 84, 2287, (1998). 77 Y. G. Wang, S. P. Lau ,X. H. Zhang, H. W. Lee, S. F. Yu, B. K. Tay, and H. H. Hng, Chem. Phys. Lett. 375, 113, (2003). 78 Q. P. Wang, D. H. Zhang, Z. Y. Xue, X. J. Zhang, Opti. Mater. 26, 23, (2004).

PAGE 116

116 79 Y. I. Alivov, E. V. Kalinina, A. E. Cherenkov, D. C. Look, B. M. Ataev, A. K. Omaev, M. V. Chukichev, and D. M. Bagnall, Appl. Phys. Lett. 83, 4719 (2003). 80 A. Osinsky, J. W. Dong, M. Z. Kauser, B. Hertog, A. M. Dabiran, P. P. Chow, S. J. Pearton, O. Lopatiuk, and L. Chernyak, Appl. Phys. Lett. 85, 4272 (2004). 81 H. Hosono, H. Ohta, K. Hayashi, M. Orita, and M. Hirano, J. Cryst. Growth, 496, 237 (2001). 82 D. K. Hwang, S. H. Kang, J. H. Lim, E. J. Ya ng, J. Y. Oh, J. H. Yang, and S. J. Park, Appl. Phys. Lett. 86, 222101 (2005). 83 H. S. Yang, D. P. Norton, S. J. P earton, and F. Ren, Appl. Phys. Lett. 87, 212106(2005). 84 S. M. Sze, Physics of Semiconductor Devices, (Wiley, New York, 1981) 85 K. A. Pickar, Ion Implantation in SilionPhysics (Appl Soli Stat Sci, New York, 1975). 86 H. S. Carslaw and J. C Jaeger, Cond uc. Heat. Solids, (Oxford, London, 1959). 87 A. S. Grove, Physics and Technology of Semiconductor Devices, (Wiley, New York, 1967). 88 D. K. Schroder, Semiconductor Material a nd Device Characterizati on, (Wiley, New York, 1998). 89 J. M. Pimbley, IEEE Tr ans. Electron Dev. Lett. 33, 986, (1986). 90 R. Castaing, B. Jouffrey, a nd G. Slodzian, Compt. Rend. 251, 1010 (1962). 91 R. K. Herzog and H Liebl, J Appl Phys 34, 2893, (1963). 92 P. Auger, J. Phys. Radium, 6, 205, (1925). 93 E. Minnni, Appl Surf Sci 15, 270, (1983). 94 D. Sarid, Scanning Force Microscopy with A pplications to Electric, Magnetic, and Atomic Forces, (Oxford, New York, 1991). 95 G. Simin, A. Tarakji, X. Hu, A. Koudymov, J. Yang, M. A. Khan, M.S. Shur, and R. Gaska, Phys. Stat. Sol. A, 188, 219(2001). 96 A. Tarakji, X. Hu, A. Koudymov, G. Simin, J. Yang, M. A. Khan, M.S. Shur, and R. Gaska, Solid State Electron. 46, 1211 (2002). 97 A. Koudymov, X. Hu, K. Simin, G. Simin, M. Ali, J. Yang, and M. A. Khan, IEEE Elec. Dev. Lett. 23, 449 (2002). 98 G. Simin, A. Koudymov, H. Fatima, J. P. Zha ng, J. Yang, M. A. Khan, X. Hu, A. Tarakji, R. Gaska, and M. S. Shur, IEEE Elec. Dev. Lett. 23, 458 (2002).

PAGE 117

117 99 M. Asif Khan, X. Hu, A. Tarakji, G. Simin, J. Yang, R. Gaska, and M. S. Shur, Appl. Phys.Lett. 77, 1339(2000). 100 X. Hu, A. Koudymov, G. Simin, J. Yang, M. A. Khan, A. Tarakji, M. S. Shur, and R. Gaska, Appl. Phys. Lett. 79, 2832 (2001). 101 F. Ren, M. Hong, S. N. G. C hu, M. A. Marcus, M. J. Schurman, A. Baca, S. J. Pearton, and C. R. Abernathy, Appl. Phys. Lett. 73, 3839 (1998). 102 J. W. Johnson, B. Luo, F. Ren, B. P. Gila , V. Krishnamoorthy, C. R. Abernathy, S. J. Pearton, J. I. Chyi, T. E. Nee, C. M. Lee, and C. C. Chuo, Appl. Phys. Lett. 77, 3230 (2000). 103 N. Q. Zhang, S. Keller, G. Parish, S. Heik man, S. P. DenBaars, and U. K. Mishra, IEEE Electron. Device Lett. 21, 421(2000). 104 H. C. Casey, G. G. Fountain, R. G. Alley, B. P. Keller, and S. P. DenBaars, Appl. Phys. Lett. 68, 1850 (1996). 105 R. Therrien, G. Lucovsky, and R. F. Davis, Phys. Stat. Solidi. A 176, 793(1999). 106 T. Hashizume, E. Alekseev, D. Pavlidis, K. S. Boutros, and J. Re dwing, J. Appl. Phys. 88, 1983 (2000). 107 M. A. Khan, X. Hu, G. Simin, G. Lunev, J. Yang, R. Gaska, and M. S. Shur, IEEE Electron. Dev. Lett. 21, 63(2000). 108 T. S. Lay, M. Hong, T. Kwo, T. P.Lee, W. Hung, and D. T. Huang, Solid State Electron. 45,1679 (2001). 109 C. Steven, C. Binari, K. Dove rspike, G. Kelner, H. B. Dietri ch, and A. E. Wickenden, Soli. Stat. Elec. 41, 177 (1997). 110 S. Arulkumaran, T. Egawa, H. Ishikawa, T. Jimbo, and M. Umeno, Appl. Phys. Lett. 73, 809 (1998). 111 P.A. Ivanov, M. E. Levinshtei n, G. Simin, X. Hu, J. Yang, M. A. Khan, S. C. Rumyantsev, M. S. Shur, and R. Gaska, Elect. Lett. 37, 1479 (2001). 112 C. H. Chen, S. J. Chang, Y. K. Su, G. C. Chi, J. Y. Chi, C. A. Chang, J. K. Sheu, J. F. Chen, IEEE Photon. Technol. Lett. 13, 848 (2001). 113 D. Mistele, T. Rotter, R. Fe rretti, F. Fedler, H. Klausiag, O. K. Semchinova, J. Stemmer, 20. J. Aderhold, and J. Graul, Mat. Res. Soc. Symp. Proc. 639, 1142(2001). 114 B. P. Gila, J. W. Johnson, R. Mehandru, B. Luo, A.H. Onstine, K. K. Allums, V. Krishnamoorthy, S. Bates, C. R. Abernathy, F. Ren and S. J. Pearton, Phys. Stat. Solidi. A 188, 239 (2001).

PAGE 118

118 115 M. Hong, K. A. Anselm, J. Kwo, H. M. Ng, J. N. Baillargeon, A. R. Kortan, J. P. Mannaerts, A. Y. Cho, C. M. Lee, J. I. Chyi , and T. S. Lay , J. Vac. Sci. Technol. B. 18, 1453 (2000). 116 Y. Irokawa and Y. Nakano, Solid State Elect. 46, 1467 (2002). 117 Y. Nakano, T. Kachi, and T. Jimbo, J. Vac. Sci. Techol. B 21, 2220 (2003). 118 D. J. Fu, Y. H. Kwon, T. W. Kang, C. J. Park , K. H. Baek, H. Y. C ho, D. H. Shin, C. H. Lee, and K. S. Chung, Appl. Phys. Lett. 80, 446 (2002). 119 B. Luo, J. W. Johnson, J. Kim, R. M. Mehandru, F. Ren, B. P. Gila, A. H. Onstine, C. R. Abernathy, S. J. Pearton, A. G. Baca, R. D. Briggs, R. J. Shul, C. Monier, and J. Han, Appl. Phys. Lett. 80, 1661(2002). 120 R. Mehandru, B. P. Gila, J. Kim, J. W. Johnson, K. P. Lee, B. Luo, A. H. Onstine, C. R. Abernathy, S. J. Pearton, and F. Ren, Electrochem. Solid State Lett. 5, 51 (2002). 121 J. Kim, R. Mehandru, B. Luo, F. Ren, B. P. Gila, A. H. Ons tine, C. R. Abernathy, S. J. Pearton, and Y. Irokawa, Appl. Phys. Lett. 81, 373 (2002). 122 J. Kim, R. Mehandru, B. Luo, F. Ren, B. P. Gila, A. H. Ons tine, C. R. Abernathy, S. J. Pearton and Y. Irokawa, Appl. Phys. Lett., 80, 4555 (2002). 123 J. Kim, B. P. Gila, R. Mehandru, J. W. Johns on, J. H. Shin, K. P. Lee, B. Luo, A. H. Onstine, C. R. Abernathy, S. J. Pearton and F. Ren, J. Electrochem. Soc. 149, 482 (2002). 124 Y. Irokawa, Y. Nakano, M. Ishiko, T. Kachi, J. Kim, F. Ren, B. Gila, A. Onstine, C. R. Abernathy, S. J. Pearton, C. Pan, G. Chen, and J. I. Chyi, Appl. Phys. Lett. 84, 2919 (2004). 125 F. Ren, M. Hong, W. S. Hobson, J. M. Kuo, J. R. Lothian, J. P. Mannaer ts, J. Kuo, S. N. G. Chu , Y. K. Chen, and A. Y. Cho, Tech. Di g. Int. Electron. Devi ces Meeting, 943 (1996). 126 C. F. Lin, H. C. Cheng, G. C. Chi, C. J. Bu, and M. S. Feng, Appl. Phys. Lett. 76, 1878 (2000). 127 S. J. Pearton, J. C. Zolper, R. J. Shul , and F. Ren, J. Appl. Phys. 86,1 (1999). 128 A. Nurmikko, and R. L. Gunshor, Semicond. Sci. Technol. 12, 1337 (1997). 129 S. Nakamura, S. J. Pearton, and G. Faso l, The Blue Laser Diode-the Complete Story (Springer, Berlin, 2000). 130 T. Nishida, N. Kobayashi, and T. Ban, Appl. Phys. Lett. 82, 1 (2003). 131 M. Iwaya, S. Takanami, A. Miyazaki, Y. Watanabe, S. Kamiyama, H. Amano, and I. Akasaki, Jpn. J. Appl. Phys. 42, 400 (2003).

PAGE 119

119 132 J. P. Zhang, A. Chitnis, V. Adivarahan, S. Wu, V. Mandavilli, R. Pachipulusu, M. Shatalov,G. Simin, J. W. Yang, a nd M. Asif Khan, Appl. Phys. Lett. 81, 4910 (2002). 133 H. Amano, S. Takanami, M. Iwaya, S. Ka miyama, and I. Akasaki, Phys. Stat. Soli. A 195, 491 (2003). 134 T. Makino, Y. Segawa, M. Kawasaki, and H. Koinuma, Semicond. Sci. Technol. 20, 78 (2005). 134 K. Ramanathan, M. A. Contreras, C. L. Perk ins, S. Asher, F. S. Hasoon, J. Keane, D. Young, M. Romero, W. Metzger, R. Nouti, J. Wards, and A. Duda, Prog. Photovolt; Res. Appl. 11, 225, (2003). 135 A. Ohmoto, and A. Tsukazaki, Semicond. Sci. Technol. 20, 1 (2005). 136 D.C. Look, Semicond. Sci. Technol. 20, 55 (2005). 137 X. L. Guo, J. H. Choi, H. Tabata, and T. Kawai, Jpn. J. Appl. Phys. 40, L177 (2001). 138 S.Y. Karpov, K. A. Bulashevich, I.A. Zhma kin, M.O. Nestoklon, V.F. Mymrin, and Y. N. Makarov, Phys. Stat. Solidi. B 241, 2668 (2004). 139 J. W. Dong, A. Osinsky, B. Hertog, A. M. Dabiran, P. P. Chow, Y.W. Heo, D. P. Norton, and S. J. Pearton, J. Electron. Mat. 34, 416 (2005). 140 D. C. Look, B. Claflin, Y. I. Alivov, and S. J. Park, Phys. Status Solidi. A 201, 2203 (2004). 141 Z. K. Tang, G. K. L. Wong, P. Yu, M. Kawasak i, A. Ohtomo, H. Koinuma, and Y. Segawa,Appl. Phys. Lett. 72, 3270 (1998). 142 K. K. Kim, J. H. Song, W. K. Choi, S. J. Park, and J. H. Song, J. Appl. Phys. 87, 3573 (2000). 143 Y. I. Alivov, J. E. Van Nostrand, D. C. Look, M. V. Chukichev, and B. M. Ataev, Appl. Phys. Lett. 83, 2943 (2003). 144 K. Ip, Y. W. Heo, D. P. Norton, S. J. Peart on, J. R. LaRoche, and F. Ren , Appl. Phys. Lett. 85, 1169 (2004). 145 C. H. Park, S. B. Zhang, and S. H. Wei, Phys. Rev. B 66, 073202 (2002). 146 D. J. Dai, Phys. Rev. B 59, 15181 (1999). 147 K. K. Kim, H. S. Kim, D. K. Hwang, J. H. Lim, and S. J. Park, Appl. Phys. Lett. 83, 63 (2003). 148 Y. W. Heo, Y. W. Kwon, Y. J. Li, J. Pearton, and D. P. Norton, Appl. Phys. Lett. 84, 3474 (2004). 149 D. C. Look, D. C. Reynolds, C. W. Litton, R. L. Jones, D. B. Eason, and G. Cantwell, Appl.Phys. Lett. 81, 1830 (2002). 150 C. C. Lin, S. Y. Chen, S. Y. Cheng, and H. Y. Lee, Appl. Phys. Lett. 84, 5040 (2004).

PAGE 120

120 151 S. Limpijumnong, S. B. Zhang, S. H. Wei, and C. H. Park, Phys. Rev. Lett. 92, 155504 (2004). 152 Y. R. Ryu, T. S. Lee, J. H. Leem, and H. W. White, Appl. Phys. Lett. 83, 4032 (2003). 153 T. Ohshima, T. Ikegami, K. Ebihara, J. Asmus sen, and R. K. Thareja, Thin Solid Films 435, 49 (2003). 154 X. Li, Y. Yan, T. Gessert, C. DeHart, C. L. Perkins, D. Young, and T. J. Coutts, Electrochem. Solid State Lett. 6, 56 (2003). 155 X. Li, Y. Yan, T. A. Gessert, C. L. Perkin s, D. L. Young, C. DeHa rt, M. Young, and T. J. Coutts, J. Vac. Sci. Technol. A 21, 1342 (2003). 156 K. H. Bang, D. K. Hwang, M. C. Park, Y. D. Ko, I. Yun, and J. M. Myoung, Appl. Surf. Sci. 210, 177 (2003). 157 T. Yamamoto and H. Yoshida, J. Cryst. Growth, 215, 552 (2000). 158 F. Oba, S. R. Nishitani, S. Iostani, H. Adachi, and I. Tanaka, J. Appl. Phys. 90, 824 (2001). 159 Y. J. Li, Y. W. Heo, Y. Kwon, K. Ip, S. J. Pearton, and D. P. Norton, Appl. Phys. Lett. 87, 072101 (2005). 160 S. Y. Han, H. Yang, D. P. Norton, S. J. Pearton, F. Ren, A. Osinsky, J.W. Dong, B. Hertog,and P. P. Chow, J. Vac. Sci. Technol. B 23, 2504 (2005). 161 A. P. Zhang, G. T. Dang, F. Ren, H. Cho, K. P. Lee, S. J. Pearton, J. I. Chyi, T. Y. Nee, and C. C. Chuo, IEEE Trans. Electron. Dev. 48, 407 (2001). 162 V. Khemka, R. Patel, T. P. Chow an d R. J. Gutmann, Solid State Electron, 43, 1945 (1999). 163 B. L. Stann, W. C. Ruff, a nd Z. G. Sztankay, Optic. Eng. 35, 3270 (1996). 164 B. L. Stann, Proc. of Adv. Sensors Consortium, 25, (2000). 165 P. H. Shen, Proc. SPIE, 4028, 426 (2000). 166 D. Soole, B. Julian, and H. Sc humacher, IEEE J. Quant. Electron. 27, 7373 (1991). 167 W. A. Wohlmuth, P. Fay, C. Can eau, and I. Adesida, Electon. Lett. 32, 249 (1996). 168 E. Drge, E. H. Bttcher, St. Kollakowski, A. Strittmater, D. Bibmberg, O. Reimann, and R. Steingrber, Elect. Lett. 34, 2241 (1998). 169 K. Aliberti, M. Wraback, M. Stead, P. Newman, and H. Shen, Appl. Phys. Lett. 80, 2846 (2002).

PAGE 121

121 170 W. A. Wohlmuth, P. Fay, C. Caneau and I. Adesida, Electron. Lett, 32, 249 (1996). 171 H. Shen, K. Aliberti, B. Stann, P. Newman , R. Mehandru, and F. Ren, Appl. Phys. Lett. 82,3814 (2003). 172 J. L. Hamberg and C. G. Granqvist, J. Appl. Phys. 60, R123 (1986). 173 J. L. Vossen, Phys. Thin Films 9, 1 (1977). 174 G. Haacke, Ann. Rev. Mater. Sci. 7, 73 (1977). 175 C. G. Granqvist, Appl. Opt. 20, 2606 (1981). 176 C.M. Lampert, Solar Energy Mater. 6, 1 (1981). 177 Z. M. Jarzebski, Phys. Status. Solid. A 71, 13 (1982). 178 J. C. Manifacier, Thin Solid Films 90, 297 (1982). 179 K. L. Chopra, S. Major, and D. K. Pandya, Thin Soli. Films 102, 1 (1983). 180 M. Quaas, C. Eggs, and H. Wulff, Thin Solid Films 332, 277 (1998). 181 K. Ichihara, N. Inoue, and M. Okubo, Thin Solid Films 245, 152 (1994). 182 J. E. Morris, M. I. Ridge, C. A. Bi shop, and R. P. Howson, J. Appl. Phys. 54, 3497 (1983). 183 V. M. Donnelly, M. Geva, J. Long, and R. F. Karlicek, Appl. Phys. Lett. 44, 951 (1984). 184 J. P. Zheng, and H. S. Kwok. Thin Solid Films, 232, 99 (1993). 185 W. Gao, A. S. Khan, R. Paul , R. Berger, R. G. Hunsperger, G. Zydzik, H. M. O’Bryan, D. Sivco, A. Y. Cho, Appl. Phys. Lett. 65, 1930 (1994). 186 K. Aliberti, B. Stann, S. Svensson, R. Me handru, F. Ren and H. Shen, Micr. Opti. Tech. Lett. 39, 08 (2003). 187 S.V. Averine, Y.C. Chen, and Y. L. Lan, Soli. Stat. Elec. 46, 441 (2001). 188 R. H.Yuang, J. I. Chyi, Y. J. Chan, W. Lin, and Y. K. Tu, IEEE Photon. Technol. Lett. 7, 1333 (1995). 189 J. E. Bowers and C. A. Burrus, J. Ligh. Tech. 5, 1339 (1987). 190 S. Adachi, Physical Properties of III-Y Semiconductor Compounds, 2nd ed (Wiley, New York, 1992). 191 E. Zielinski, H. Schweizer, K.Streubel, H.Eisele, and G.Weima nn, J. Appl. Phys. 59, 2196 (1986).

PAGE 122

122 192 J. Burm, K. I. Litvin, W. J. Kshaff, and L. F. Eastman, IEEE Tr ans. Electron. Dev. 6, 722 (1994). 193 H. Beneking, IEEE Trans. Elec. Dev. 29, 1420 (1982). 194 G. W. Farnell, I. A. Cermac, P. Silvester, and S. K. Wong, IEEE Trans. Sonics. Ultrason, 17, 188 (1970). 195 H. Hodora, Fibre Integ. Opt. 4, 233 (1983).

PAGE 123

123 BIOGRAPHICAL SKETCH Soohwan Jang was born on July 27, 1977 in Se oul, Korea. He matriculated at Soul National University in Seoul, Ko rea, where he received his Bach elor’s Degree in February, 2003. He joined the Army in 1999 and wa s retired as a sergeant in 2001. In 2003, he entered the Ph.D. program at the University of Florida and joined Professor Ren’s group. He is co-author of approximately 30 journal and conference papers dealing with compound semiconductor device technology. He did his internship with Dr. Hock Min Ng at Bell Laboratories during the spring of 2007. GaN MOSFET, ZnO LED, ZnO nanorod solution growth, and InGaAs MSM photodetector were researched in his PH.D. pursuing. He gradua ted from the University of Florida with a doctoral degree in Chemical Engineering in May 2007.