<%BANNER%>

Molybdenum-Trioxide Assisted Charge Injection from Metal Anodes into Organic Light Emitting Diodes

Permanent Link: http://ufdc.ufl.edu/UFE0024689/00001

Material Information

Title: Molybdenum-Trioxide Assisted Charge Injection from Metal Anodes into Organic Light Emitting Diodes
Physical Description: 1 online resource (100 p.)
Language: english
Creator: Shaikh, Mikail
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aluminum, anode, barrier, capacitance, charge, composite, cupc, diodes, dipole, disorder, emitting, encapsulation, evaporation, gold, hole, indium, injecting, injection, interfacial, lifetime, light, material, metal, molybdenum, moo3, npb, oled, organic, oxide, pedot, polymer, silver, tin, transport, trioxide, voltage
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The current-voltage and luminance characteristics were studied of green fluorescent light emitting diodes fabricating using the archetypal N, N?-di (1-naphthyl)-N, N?-diphenyl-1,1?-biphenyl -4,4?-diamine (?-NPD) and Tris-(8-hydroxyquinoline)-aluminum (Alq3) as the hole transporting and electron transporting layers respectively. The performance of Molybdenum trioxide (MoO3) as a hole-injection layer was compared to conventional hole injection materials like copper phthalocyanine (CuPc) and Poly (3,4-ethylene-dioxythiophene):poly (styrenesulfonate) (PEDOT:PSS). The effect of the thickness of MoO3 on the electronic properties of the devices was also examined. OLEDs made with the traditional Indium-Tin-Oxide (ITO) anodes were compared to OLEDs fabricated with transparent metal anodes of thermally evaporated Gold, Silver and Aluminum. Finally, the operating lifetime of OLEDs made with different hole-injection polymers was studied as a function of luminance decay and operating voltage rise. It was observed that MoO3 performance is comparable to PEDOT:PSS and CuPc in terms of efficiency and operating voltages. There is a dramatic improvement in OLED characteristics on the introduction of a nanometer-thick layer of MoO3, but the enhancement dies away with increasing thickness. OLEDs fabricated with the different metal anodes behave almost identically to ITO-based OLEDs, in spite of the large variation in metal work functions, indicating the presence of MoO3 somehow neutralizes that difference. Comparison of operating lifetimes seems to indicate at least an empirical relation between degradation and the hole-injection barrier. Capacitance-Voltage measurements also confirm that the rate of degradation of the organic material is retarded when efficiency is improved via a hole injection material, establishing a link between the two main problems that plague OLED development.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mikail Shaikh.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: So, Franky.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024689:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024689/00001

Material Information

Title: Molybdenum-Trioxide Assisted Charge Injection from Metal Anodes into Organic Light Emitting Diodes
Physical Description: 1 online resource (100 p.)
Language: english
Creator: Shaikh, Mikail
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aluminum, anode, barrier, capacitance, charge, composite, cupc, diodes, dipole, disorder, emitting, encapsulation, evaporation, gold, hole, indium, injecting, injection, interfacial, lifetime, light, material, metal, molybdenum, moo3, npb, oled, organic, oxide, pedot, polymer, silver, tin, transport, trioxide, voltage
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The current-voltage and luminance characteristics were studied of green fluorescent light emitting diodes fabricating using the archetypal N, N?-di (1-naphthyl)-N, N?-diphenyl-1,1?-biphenyl -4,4?-diamine (?-NPD) and Tris-(8-hydroxyquinoline)-aluminum (Alq3) as the hole transporting and electron transporting layers respectively. The performance of Molybdenum trioxide (MoO3) as a hole-injection layer was compared to conventional hole injection materials like copper phthalocyanine (CuPc) and Poly (3,4-ethylene-dioxythiophene):poly (styrenesulfonate) (PEDOT:PSS). The effect of the thickness of MoO3 on the electronic properties of the devices was also examined. OLEDs made with the traditional Indium-Tin-Oxide (ITO) anodes were compared to OLEDs fabricated with transparent metal anodes of thermally evaporated Gold, Silver and Aluminum. Finally, the operating lifetime of OLEDs made with different hole-injection polymers was studied as a function of luminance decay and operating voltage rise. It was observed that MoO3 performance is comparable to PEDOT:PSS and CuPc in terms of efficiency and operating voltages. There is a dramatic improvement in OLED characteristics on the introduction of a nanometer-thick layer of MoO3, but the enhancement dies away with increasing thickness. OLEDs fabricated with the different metal anodes behave almost identically to ITO-based OLEDs, in spite of the large variation in metal work functions, indicating the presence of MoO3 somehow neutralizes that difference. Comparison of operating lifetimes seems to indicate at least an empirical relation between degradation and the hole-injection barrier. Capacitance-Voltage measurements also confirm that the rate of degradation of the organic material is retarded when efficiency is improved via a hole injection material, establishing a link between the two main problems that plague OLED development.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mikail Shaikh.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: So, Franky.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024689:00001


This item has the following downloads:


Full Text

PAGE 1

1 MOLYBDENUM TRIOXIDE ASSISTED CHARGE INJECTION FROM METAL ANODES INTO ORGANIC LIGHT EMITTING DIODES By MIKAIL A. SHAIKH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE R EQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Mikail A. Shaikh

PAGE 3

3 To my mother, Zarin, the rock on which I stand tall

PAGE 4

4 ACKNOWLEDGMENTS I am eternally grateful to those who care. My parents top the list, my father being my counsel and inspiration in the face of chaos and the epitome of hard work, and my mother, who believed in me even when I faltered. The sacrifice they made of allowing their eldest to fly away and pursu e a dream. I pray I am so strong. Id also like to name, among others, Jasmine, a friend beyond measure, and my greatest treasure. Id like to express my sincere gratitude and respect for Dr. Franky So, my graduate advisor, for giving me the right kind of nudge in the right direction at the right time. I also cannot pass lightly over the influence of Jaewon, Kaushik, Neetu and all the other people of the Organic Electronics group. I appreciate them being there.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 8 LIST OF FIGURES .............................................................................................................................. 9 ABSTRACT ........................................................................................................................................ 11 CHAPTER 1 INTRODUCTION ....................................................................................................................... 13 Introduction ................................................................................................................................. 13 Purpose of this Study .................................................................................................................. 13 Historical Review ........................................................................................................................ 14 Theory of Electroluminescence .................................................................................................. 15 Device Architectures ................................................................................................................... 17 2 LITERATURE REVIEW -HOLE INJECTION INTO ORGANIC SEMICONDUCTORS ................................................................................................................ 19 Introduction ................................................................................................................................. 19 Conceptual Aspects ..................................................................................................................... 19 Electronic Structure ............................................................................................................. 19 Energy Alignment at a Metal Organic Interface ............................................................... 20 Band Bending in an Organic Layer .................................................................................... 21 Application of Ultraviolet Photoelectron Spectroscopy in the Study of Interfaces ......... 22 Relevant Metal Organic Interfaces ............................................................................................ 23 NPB/Alq3 Interface .............................................................................................................. 23 Alq3 / LiF / Al Interfaces ..................................................................................................... 24 Review of Hole Injection Ma terials ........................................................................................... 25 Indium Tin Oxide (ITO) Anodes ........................................................................................ 25 Copper Phthalocyanine as a Hole Injection Material ........................................................ 26 Polymeric Anodes as Ohmic Contacts for NPB: ............................................................... 27 Nickel Ox ide as a transparent p -type Conducting Oxide .................................................. 27 Conducting/Insulating Nature of WO3 ............................................................................... 28 Charge Injection through Plasma Polymerized Fluorocarbon Films ............................... 28 The Use of MoO3 as an Anode Buffer Layer ..................................................................... 29 3 LITERATURE RE VIEW -ALTERNATE MATERIALS FOR ANODES ............................. 41 Charge Injection .......................................................................................................................... 41 Introduction .......................................................................................................................... 41 Conventional Approach to Carrier Injection ...................................................................... 41

PAGE 6

6 Observed El ectron Injection from Cathodes ...................................................................... 42 The Interface Injection Model ............................................................................................ 43 Carrier Injection into NPB .............................................................................................. 44 Review of Alternate Anodes ................................................................................................... 45 Introduction .......................................................................................................................... 45 Use of p Type Silicon .......................................................................................................... 45 Semitransparent Silver Anode ............................................................................................ 46 Metal / Fullerene Electrodes ............................................................................................... 46 Oxides Of Underlying Meta l Anodes a s Effective Hole Injection Contacts ................... 47 Charge Generation Interfaces .............................................................................................. 48 Feasibility of a Metal Anode using Molybdenum Trioxide .............................................. 49 4 EXPERIMENTAL METHODS ................................................................................................. 57 Introduction ................................................................................................................................. 57 Device Preparation ...................................................................................................................... 57 Substrate Cleaning ............................................................................................................... 57 Surface Treatment ................................................................................................................ 58 Addition of Polymeric Hole Injection Layer ..................................................................... 58 Thermal Evaporation ........................................................................................................... 59 Encapsulation ....................................................................................................................... 60 Device Materials and Architecture ............................................................................................. 60 Characterization Te chniques ...................................................................................................... 61 LIV Measurements .............................................................................................................. 61 Sample Holder ..................................................................................................................... 61 Analysis of Spectrum .......................................................................................................... 62 Capacitance Voltage Measurements ................................................................................... 62 Experimental Procedure .............................................................................................................. 63 Thickness Variation of the Molybdenum Oxide Hole Injection Layer ............................ 63 Study of Metallic Anodes .................................................................................................... 63 Lifetime Measurement of OLEDs ...................................................................................... 63 5 RESULTS AND DISCUSSION ................................................................................................ 70 Molybdenum Oxide as a HoleInjection Material for Green OLEDs...................................... 70 Background .......................................................................................................................... 70 Results .................................................................................................................................. 70 Discussion ............................................................................................................................ 71 Metal Based Composite Anodes ................................................................................................ 73 Background .......................................................................................................................... 73 Results .................................................................................................................................. 73 Discussion ............................................................................................................................ 75 Lifetime and Degradation of OLEDs Based on the Performance of Hole Injection Materials ................................................................................................................................... 78 Background .......................................................................................................................... 78 Results .................................................................................................................................. 78 Discussion ............................................................................................................................ 79

PAGE 7

7 6 SUMMARY AND FUTURE WORK ....................................................................................... 91 Summary of Results .................................................................................................................... 91 Future Work ................................................................................................................................. 92 REFERENCES ................................................................................................................................... 95 BIOGRAPHICAL SKETCH ........................................................................................................... 100

PAGE 8

8 LIST OF TABLES Table page 4 1 Specification of Organic materials and films ....................................................................... 65 4 2 Specifications of Metals and films ........................................................................................ 65 4 3 Structures of Organic Compounds ........................................................................................ 65

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Schematic of the theory of Electroluminescence ................................................................. 18 2 1 Plot of Coulombic Energy versus distance for nucleus ....................................................... 32 2 2 Vacuum level at infinity and close to the solid .................................................................... 32 2 3 Interface formation between a metal and an organic material ............................................ 33 2 4 Band Bending at a metal / organic interface......................................................................... 34 2 5 UPS diagram of energy levels of metal and overlaid organic material .............................. 34 2 6 Energy level offset at NPB/Alq3 interface ............................................................................ 35 2 7 Energy Offset at the Alq3/LiF/Al interface ........................................................................... 36 2 8 Energy level alignment of ITO, CuPc and PEDOT to NPB ................................................ 37 2 9 Formation of Ohmic Contact of NPB with PEDOT ............................................................ 37 2 10 Transmission through MoO3 and CuPc based films on NPB .............................................. 38 2 11 AFM images showing surface roughness ............................................................................. 38 2 12 UPS data for ITO, ITO/NPB, ITO/CuPc, ITO/MoO3, ITO/MoO3:NPB ............................ 39 2 13 UPS data for ITO/NPB, ITO/CuPc/NPB, ITO/MoO3/NPB, ITO/MoO3:NPB/NPB ......... 39 2 14 Energy diagram showing offsets for ITO/NPB, ITO/CuPC, ITO/MoO3/NPB and ITO/MoO3:NPB/NPB ............................................................................................................ 40 2 15 XPS data showing energy shift of Indium 3d peaks ............................................................ 40 3 1 Electron injection from various cathodes into Alq3 ............................................................. 51 3 2 Electron Injection from various cathodes at 30K and 290K ............................................... 51 3 3 Interface Injection Model, showing the energy diagram modified by interfacial states caused by dipoles (a), (b) and the energy distribution for the first few layers of the o rganic material (c). ............................................................................................................... 52 3 4 Inductance spectroscopy of ITO/NPB showing filling and depopulation at different bias voltage s ........................................................................................................................... 53 3 5 Energy diagram of (a)p Si/TPD and (b) ITO/TPD ............................................................. 54

PAGE 10

10 3 6 Energy diagram of Au/HAT CN/NPB interface, showing charge generation possibilities due to alignment of Au Fermi level, HAT CN LUMO and NPB HOMO .... 54 3 7 UPS studies of ITO/MoO3/NPB interface, showing shift in cut off position and the formation of a gap state within NPB ..................................................................................... 55 3 8 Energy level diagram of ITO/NPB and ITO/MoO3/NPB showing a reduced hole injection barrier and the location of the gap state. ............................................................... 55 3 9 UPS studies of MoO3 deposited on clean Au, showing the shift in the cut off position and the presence of gap states in the MoO3 layer. ............................................................... 56 3 10 Change in work function of Au and Al as a function of MoO3 thickness (a) and the energy level diagram (b) ........................................................................................................ 56 4 1 Schematic representation of device architecture .................................................................. 66 4 2 Thermal Evaporation Set up .................................................................................................. 67 4 3 L -I -V measurement setup ...................................................................................................... 68 4 4 Diagrammatic Representation of L I -V Setup ..................................................................... 68 4 5 Sample Holder with 4 OLEDs (seen with the blue color) ................................................... 69 4 6 Sample holder with one metal anode OLED illuminated .................................................... 69 5 1 Current Density plots as a function of thickness of MoO3 layer ......................................... 82 5 2 Luminance Plots as a function of thickness of the MoO3 layer .......................................... 82 5 3 Current Efficiency Plots as a function of MoO3 layer thickness ......................................... 83 5 4 Current Density plots of OLEDS with different anodes ...................................................... 84 5 5 Luminance plots of OLEDs with different anodes .............................................................. 84 5 6 Current Efficiency plots of OLEDs with different anodes .................................................. 85 5 8 Current Density and Luminance plots for OLEDs made with various polymeric hole injection materials INSET: Luminance response of the OLEDs ....................................... 87 5 10 Lifetime Decay of luminance of OLEDs with different polymeric hole -injection materials .................................................................................................................................. 89 5 11 Lifetime Voltage rise of OLEDs with different polymeric hole injection materials ......... 89 5 12 Capacitance-Voltage Data of OLEDs under electrical stress: ............................................. 90

PAGE 11

11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science MOLYBDENUM TRIOXIDE ASSISTED CHARGE INJECTION FROM METAL ANODES INTO ORGANIC LIGHT EMITTING DIODES By Mikail A. Shaikh August 2009 Chair: Franky So Major: Material Science and Engineering The current voltage and luminance characteristics were studied of green fluorescent light emitting diodes fabricating using the archetypal N, N -di (1 -naphthyl) N, N -diphenyl [1,1 biphenyl] 4,4 -NPD) and Tris (8 hydroxyquinoline) alumin um (Alq3) as the hole transporting and electron transporting layers respectively. The performance of Molybdenum trioxide (MoO3) as a hole injection layer was compared to conventional hole injection materials like copper phthalocyanine (CuPc) and Poly (3,4 -ethylene dioxythiophene):poly (styrenesulfonate) (PEDOT:PSS). The effect of the thickness of MoO3 on the electronic properties of the devices was al so examined. OLEDs made with the traditional Indium Tin Oxide (ITO) anodes were compared to OLEDs fabricated with transparent metal anodes of thermally evaporated Gold, Silver and Aluminum. Finally, the operating lifetime of OLEDs made with different hole injection polymers was studied as a function of luminance decay and operating voltage rise. It was observed that MoO3 performance is comparable to PEDOT:PSS and CuPc in terms of efficiency and operating voltages. There is a dramatic improvement in OLED ch aracteristics on the introduction of a nanometer thick layer of MoO3, but the enhancement dies away with increasing thickness. OLEDs fabricated with the different metal anodes behave almost

PAGE 12

12 identically to ITO -based OLEDs, in spite of the large variation i n metal work functions, indicating the presence of MoO3 somehow neutralizes that difference. Comparison of operating lifetimes seems to indicate at least an empirical relation between degradation and the hole injection barrier. Capacitance -Voltage measurem ents also confirm that the rate of degradation of the organic material is retarded when efficiency is improved via a hole injection material, establishing a link between the two main problems that plague OLED development.

PAGE 13

13 CHAPTER 1 INTRODUCTION Introducti on Electronic devices components made of organic materials are rapidly evolving as energy efficient and cost effective replacements for a number of conventional electronic applications. While the present trend of this so called organic electronics field does not seek to supplant silicon electronics in the high speed and high performance arena, it instead hopes to augment it for large scale, flexible and low cost systems. Progress in the field of Organic Light Emitting Diodes (OLED) is promising because of the comparatively less stringent physical laws governing its fabrication, as well as the capability to tune its properties due to the extremely versatile nature of organic molecular design. The challenge today, as this technology enters the market, is tha t its efficiency and long term degradation hamper its widespread use in large area as well as portable electronics. Purpose of this Study It has been recognized that the efficiency and the lifetime of these devices are directly linked to the physics of it s operation, namely, the process of electroluminescence. Therefore it is necessary to optimize all intermediate steps that contribute to it. Yet, though the concepts are broadly understood, it is the intricacies that limit further development. This study focuses on the injection of electrical charge carriers in small molecule OLEDs. Transporting charges over a hetero -interface such as the one that exists at a metal/semiconductor junction is a phenomenon that poses inherent barriers to seamless electron flo w, and an understanding of the nature of barrier physics would lead to its control via optimization. We hereby consider the injection characteristics of a conducting layer of Molybdenum Oxide (MoO3), used as a medium to aid the injection of positive charge carriers in green fluorescent OLEDs. This material has recently been

PAGE 14

14 gaining a lot of interest ever since it was discovered that it is an efficient hole injection material in spite of being a strong electron acceptor. We examine and attempt to explain the nature of this improvement, and based on these findings, we designed an experiment to support our interpretation of the mechanism of MoO3assisted charge injection. We hope that this study provides insight for improving the performance of OLEDs. Historica l Review Organic material based light emission in itself is not a new phenomenon, with the first organic electroluminescence itself being studied in the 1950s by Bernanose et al. [1] using molecularly dispersed polymer films. In the 1960s, research moved o n to carrier injection based electroluminescence, when the light emission from a highly purified single crystal of condensed, aromatic anthracene was studied by Pope [2, 3] and Helfrich [4, 5] and co-workers. It was established that the basic process of el ectroluminescence is based on the injection of electrons and holes from electrodes, recombination of both these charge carriers, formation of singlet and triplet excitons and the radiative decay of these to emit light. However, anthracene based devices had some inherent disadvantages including high operating voltages, and attention shifted away from organic crystals to organic thin films. However, thin films of polycrystalline materials like anthracene that was vacuum deposited showed morphology that was too rough to achieve satisfactory current injection. Similarly, other methods of fabrication like the Langmuir Blodget method and vacuum sublimed polycrystalline films were examined and found to make unstable films too [6,7]. Then in 1982, Partridge reported on the electroluminescence of poly(vinyl carbazole) thin films doped with fluorescent molecules as emission centers[8]. Multi layer structures began to appear in the 80s, and Hajashi et al. made thin films of polythiophene electropolymerized material betw een anode and perylene films to form light emitting devices, with a dramatic reduction in operating voltages [9]. But it was the landmark paper by Tang et al.

PAGE 15

15 [10] that prototyped all subsequent organic light emitting diodes. They designed a double layer s tructure with separate hole and electron transporting materials. Quantum efficiency and luminescence efficiency of 1% and 1lm/W were achieved, respectively A vast number of papers elucidating the optimization of this structure quickly followed, along with the rapid development of alternate materials and device structures. In 1990, Burroughs et al. developed the first polymer based OLED to open up possibilities of low cost printable devices [11]. Today, green OLEDs can operate at a brightness of >106 cd/m2 and a power efficiency of 53lm/W [12]. The lifetime of these devices reaches 20,000 hours at a brightness of 50 100 cd/m2. The internal quantum efficiency of these OLEDs is now almost 100%. Phosphorescence harnesses all the available energy to make devi ces of unprecedented efficiencies, now nearly equaling the parameters of conventional fluorescent lighting technology [13]. Red, Blue, Green and White color OLEDs are all being fabricated for display and lighting applications. [14,15,16,17] Theory of Elect roluminescence Organic solid films are wide energy -gap semiconductors, which means there are no charges intrinsically present in these materials and charges have to be injected. Therefore OLEDs are treated as being in dynamic charge equilibrium rather than electrostatic equilibrium. Unlike inorganic LEDs, emission in OLEDs occurs through the radiative recombination of charge carriers, and there has been no evidence of radiative decay of individual charges in OLEDs. When a sufficient external voltage is appl ied across the electrodes, electrons from the organic materials move into the anode under the influence of the electric field and therefore, in effect, holes are injected from the anode. Electrons are injected into the organic material at the cathode. Th e ease of injection and subsequent transport of holes and electrons in organic films depends on the ionization potentials and electron affinities of the materials. The work function of

PAGE 16

16 the materials relative to the work function of the electrodes also play s a significant role in charge injection. Charge transport through organic materials is generally accepted to occur via a hopping process between the molecules in the amorphous film. The semiconducting behavior of organic molecules is dependent on the pre sence of double bonds in the material structure. Conjugated systems like those found in aromatic compounds are especially favorable. The overlap of electrons in the -orbitals leads to a delocalization of the -electron cloud over the macromolecule, yet is short range and confined to the molecule, unlike crystalline delocalization, which extends throughout the crystal [18]. This allows surplus electrons to travel easily along the carbon backbone, but forces a jump to the next molecule. The disorder in the p osition and the energy fluctuations of the molecules has a huge effect on the mobility of the charges. Energetic disorder is caused by the combined effect of variation in molecular geometry due to bond rotation and variation in dipole energy and Van -der Wa als interaction between molecules [19]. After injection, electrons and holes travel through their respective materials under the influence of the electric field and recombine at a material interface to form an electrically neutral quasi -particle called a n exciton or electron -hole pair. If the charges are generated far from each other (as in the case of charges injected from electrodes), volume recombination occurs, as opposed to geminate recombination, the process following the initial carrier separation from an unstable locally excited state to form a charge transfer state (as in photoconduction). In the case of OLEDs the process of volume recombination is kinetically bimolecular, as the charges are statistically independent of each other. These excito ns are said to have a singlet or triplet energy state depending on the spin states of the carriers, with the singlet and triplet energy states produced in the ratio of 1:3. Singlet states

PAGE 17

17 decay radiatively, emitting light within the material via fluorescen ce, or are deactivated by thermally dissipating their energy. The triplet states on the other hand will decay non -radiatively. These triplet excitons can undergo triplet triplet annihilation, occurring in approximately 15% of the cases [19], leading to del ayed fluorescence. They can also undergo phosphorescence in the presence of dopants, which also leads to emission of light, or they can be deactivated by thermal dissipation. The photons emitted from the excitons will either be dissipated within the materi al by the process of reabsorption of the energy, or they will escape from the material and be detected. Device Architectures Ever since Tang et al. introduced the concept of fabrication of bi layer OLEDs in 1987 [10], almost all devices have used multiple layers. This eliminates the need of having a single material with good electron and hole mobility, as well as a bandgap corresponding to the desired wavelength. The basic OLED consists of a transparent conducting oxide like Indium Tin Oxide (ITO) as an ano de [10], a hole transporting material with good hole mobility and poor electron mobility, an electron transport layer with good electron mobility and poor hole mobility, and a metal cathode. The metal cathode is usually a composite metal alloy like Mg:Ag o r Li:Al [20]. The recombination of electrons and holes occurs near the interface of the hole and electron transporting layers, with the material in which the actual recombination region lies acting as the emissive layer. In advanced architectures, there mi ght be a dedicated emissive layer, with a difference in energy relative to the neighboring layer, to tune the emission wavelength [21]. A thin layer of organic or inorganic material may be introduced between the anode and the hole transport layer to improv e or retard the injection of holes, depending on the requirement of charge injection or charge balance respectively [22,23,24,25,26]. Moreover, electron or hole blocking layers may be used on either side of the emissive layer to block passage of the carrie rs

PAGE 18

18 through to the opposite electrode, and confine them to the active region [27]. Fluorescent dyes are sometimes added as emissive centers to either transport layer in order to increase efficiency and modify the emission spectra [28]. Phosphorescent materi als are also added to transport layers to harvest the energy of nonradiative processes [29, 30]. Figure 1 1. Schematic of the theory of Electroluminescence

PAGE 19

19 CHAPTER 2 LITERATURE REVIEW -HOLE INJECTION INTO ORGANIC SEMICONDUCTO RS Introduction The underst anding of electronic material interfaces is crucial for the improvement of electroluminescent devices. They can principally be divided into metal/organic and organic/organic interfaces. The latter is important as most recombination of electrons and holes o ccurs at this interface, and the former is crucial to carrier injection and retrieval, and also for the wiring and commercial application of organic electronics. The subject of interfacial structures can be broken into two aspects: the energy level alignme nt at the interface, and the band bending in a thicker region. Alignment is important in electroluminescent devices for carrier injection, and band bending is more important for carrier separation as in the case of photovoltaic devices and solar cells. Con ceptual A spects Electronic Structure In the basic electronic structure of a Hydrogen atom, we can see that in a plot of energy versus nuclear distance, the nucleus sits in a potential well caused by its Columbic forces (Figure 2 1). Various atomic orbitals are formed at a distance above the nucleus, and in the case of Hydrogen the single electron occupies the lowest 1s orbital (Figure 2 1a). The upper horizontal part of the well represents the vacuum level, which is the energy with which the electron can es cape the nuclear binding force. For a polyatomic molecule (Figure 2 1b), the walls of the potential wells of adjacent nuclei merge at higher energies. Deep levels of atomic orbitals are still localized in the individual potential wells, but the upper orbit als interact to form delocalized molecular orbitals. Again the electrons will inhabit the lowest possible orbitals, and there will be free orbitals available within the merged part of the potential well. The highest level of the

PAGE 20

20 occupied orbitals is called the HOMO, and the lowest unoccupied molecular orbital is called the LUMO. The energy separation from the HOMO to the vacuum level (VL) is called the ionization energy (IE) and the energy separation from the LUMO and the VL is called the Electron Affinity (EA). When molecules or polymers come together in a solid (Figure 2 1c), the valence band and part of the conduction band is still localized to the individual molecules, and preserves its electronic structure, while only the upper conduction band levels overlap, leading to limited application of the regular band theory which assumes itinerant electrons [31]. The value of IE and EA is also different from isolated molecules due to a multi-electronic effect. The electronic polarization of the molecules surrounding an ionized molecule stabilizes the ion, leading to a lowering of IE and an increase in EA as compared to gas phase molecules [32]. Since the electrons still fill the orbitals following the Fermi statistics, the concept of Fermi level is still valid, a nd the separation of the Fermi Level from the VL is called the work function The VL for an electron at rest at an infinite distance from a solid is invariant and is called VL ( ), and the energy of an electron at rest just outside a solid is still affe cted by the potential solid, and is denoted as VL(s) [33,34] (Figure 2 2). For a metal, this difference is mostly due to a surface dipole being formed by the tailing of the electron cloud at the surface, and the tailing causes the vacuum side to be negativ e [35,36]. The existence of such a surface dipole in organic solids comparable to metals has not yet been observed. Energy Alignment at a Metal -Organic Interface When a metal and an organic solid are far away, their energy levels are aligned to share a vac uum level VL ( ) (Figure 2 3a). When the solids come in contact without rearrangement of electric charge, the organic layer is now in the potential dipole of the metal, and its energy levels are raised to have a common VL (s) in an extremely narrow interfa cial gap (Figure 2 3 b,c). At

PAGE 21

21 actual contact, a dipole layer is formed at the interface, due to various origins (Figure 2 3d), like charge transfer across the interface, redistribution of the electron cloud, interfacial chemical reaction and other sources of electronic rearrangement [33,37]. This dipole is an additional layer on top of the pre -existing dipole layer at the surface of the metal surface. With such a dipole formation, there will be an abrupt shift in the potential across the dipole layer, lead ing to a VL ( ), determined by the magnitude of the shift of the vacuum levels relative to each other [38]. Band Bending in a n Organic Layer For an interface with a thick organic layer, band bending also must be considered. In general, the work function of metals and organics are not at equilibrium, i.e., when the Fermi levels would be at the same energy. If the number of available charge carriers in the organic layer is sufficient, there will be charge redistribution around the interface in a short period of time after contact [32]. If the work function of the metal is higher than that of the organic, electrons will flow to the metal, creating a negative charge in the metal and a positive charge in the organic material, and vice versa, until the Fermi level s are aligned. As a result, a diffusion layer forms with band bending to align the Fermi levels with a built in potential V bi in the organic material (Figure 2 4) [39,40,41]. This corresponds to the difference in the work functions. The thickness of the d iffusion layer depends on the built in potential V bi the dielectric constant of the material and the spatial distribution of charges. However, the requirement of band bending being the availability of mobile carriers, the minimum reported thickness re quired for band bending is 10nm [42], and for layers with thickness less than this, flat band condition may be approximated [43].

PAGE 22

22 Application o f Ultraviolet Photoelectron Spectroscopy in the Study o f Interfaces In UPS, a sample in vacuum is irradiated with high -energy monochromatic light, and the energy distribution of the emitted photoelectrons is measured [44,45,46]. The typical light sources are rare gas discharges and monochromatic synchrotron radiation. The energy analysis is typically done using an el ectrostatic analyzer or a retarding field analyzer. The electronic structure for a metal is shown in Figure 2 5a along with the photoemission spectrum. The electrons in the occupied states, that corresponds to the Fermi level E F in metals are excited by in cident radiation of energy h and those with energies above the vacuum level E vac m can escape the surface. The kinetic energy of this photoelectron will be: Ek = h Eb (2 1) Where Eb is the binding energy before excitation relative to the VL. The electron with the maximum energy EK max (metal) comes off the Fermi level and the low energy cut -off is defined by the VL EK vac. The work function of a metal can be determined as m = h EK max (metal) (2 2) When a thin organic layer is deposit ed on top of a metal, the electrons that are ejected from the metal are scattered into the organic, and the spectrum is dominated by photoelectrons ejected from the HOMO level of the organic layer(Figure 2 5b). Now the electrons with the highest energy EK m ax (org) correspond to the electrons from the HOMO. The shift in the VL can be determined from the shift in the low energy cut off of the spectra. The ionization energy of the organic can be determined as IE = h EK max (org) (2 3)

PAGE 23

23 The energy difference of the fastest electron of the metal and the organic layer gives the energy of the top of the valence band of the organic layer V F relative to the Fermi level of the metal. This difference corresponds to the injection barrier for holes into the organic layer. The energy of the vacuum level of the organic layer can also be given as vac F = I V F (2 4) This quantity is not necessarily the work function of the organic layer since the Fermi level alignment is generally not achieved w ith such thin films. In this way multiple layers can also be analyzed, while keeping in mind that thicker samples are limited by sample charging. Relevant Metal -Organic Interfaces UPS spectra from 100 thick films of each organic material are first record ed to determine the IE and the valence band line shape IE is defined as the energy difference between E vac measured by linear extrapolation of the low kinetic energy onset of the photoemission spectrum, and the top of the occupied states, measured by lin ear extrapolation of the high kinetic energy side of the HOMO peak. The photoemission onset is measured with the sample biased at 3 V to clear the detector work function. The electron affinity (EA) is defined as IE minus the optical band gap. NPB/Alq3 Int erface Figure 2 6 shows an incremental deposition of NPB on Alq3. A 4 deposition shows a superposition of peaks of both organic layers, while a 20 layer shows the characteristics of a thick layer of NPB [47]. The offset between the HOMO of NPB and Alq3 c an be determined by the direct comparison of the linear extrapolation of the top of the HOMO on the substrate Alq3 and the 20 NPB layer; and also by measuring the energy difference between the prominent peaks at about 10eV below EF and the tops of the HOM O of thick films and comparing their

PAGE 24

24 relative positions at the interface. Both methods lead to the determination of the same offset of 0.5 0.05 eV, with the NPB HOMO being above that of Alq. The same result can be obtained if the order of deposition is r eversed. The photoemission offset shows a 0.25 0.05 eV rise of Evac from Alq3 to NPB. Alq3 / LiF / Al I nterfaces It is well known that Ca and Mg make excellent electron injection contacts [48], but from a commercial standpoint, it makes sense to elimina te IA and IIA reactive metals. In this sense a pure Al cathode would be desirable, but it is seen that such an electrode makes poor electronic contact, though studies of Alq3 deposited on Al [49] should have a very small barrier to electronic injection (figure 2 7a). This has been attributed to a destructive reaction of the hot Al atoms on the Alq3 structure, and as little as 0.4 of deposited metal is enough to disrupt the spectral structure of Alq3 [50]. The presence of a small amount of LiF or MgO at t he Alq3 / Al interface has been shown to dramatically improve the device performance, lower the drive voltage and enhance the EL efficiency [51]. The improvements have largely been ascribed to two mechanisms: the interaction of the fluoride with the metal and organic layer to form low work function contacts (Figure 2 7b), or the generation of a dipole. The liberation of Li is expected to dope the organic material to form a low work function contact, but direct thermal evaporation on a Li metal contact instead of a less reactive fluoride can lead to a more extensive doping, which is seen to cause unwanted quenching of the luminescence [52]. The reaction of the fluoride with the Aluminum [53] also affects the transport properties of the interlayer, and increas es its conductivity dramatically. The second mechanism that can also account for reduction in Al work function in the presence of LiF and does not require dissociation of the fluoride is a dipole

PAGE 25

25 induced work function change. The shift in the vacuum level between the metal electrode and organic layer can be induced by the large dipole moment of the oriented LiF molecule [54 ]and interfacial charge transfer from the adsorbed fluoride [55] and particularly the Li atom to the Al cathode (Figure 2 7c). This re sults in a shift to lower work function of the LiF/Al bilayer and an increase in VBI [52]. Review of Hole -Injection Materials Indium Tin Oxide (ITO) Anodes The most commonly used anode material today is ITO, which is a degenerate ntype semiconductor with a low electrical resistivity of 2 4 x 104 cm. This low value is due to high carrier concentration because the Fermi level lies above the conduction band and the degeneracy is due to the oxygen vacancies and substitutional tin dopants created during fil m deposition. Moreover, ITO is a wide gap semiconductor with high transmission in the visible and near infrared regions [56]. There are several deposition techniques like chemical vapor deposition [57], magnetron sputtering [58], evaporation [59], thermal pyrolysis [60] and pulsed laser deposition [61]. Ever since it was discovered that the charge injection from ITO depended strongly on its work function, which in turn was very sensitive to surface states, extensive research has been conducted on the modula tion of the work function. Numerous methods have been developed like the wet treatment [62], the low -pressure plasma treatment [63], the atmospheric pressure plasma treatment [64], the ultraviolet ozone treatment [65] and the self assembling monolayer coa ting treatment [66]. These methods increase the work function of ITO from about 4.7eV to about 5.1 eV. However, this means that the device performance is strongly dependent of processing conditions and these surface treatment methods suffer from instabili ty and inconsistency. Therefore, alternate methods are being developed to improve injection, and it is seen that inserting a thin layer of a material with a work function intermediate to that of ITO

PAGE 26

26 and the hole transport material sets up a ladder type of injection barrier, and this seems to facilitate overall injection. Copper Phthalocyanine as a Hole Injection Material It has been suggested that CuPc as a buffer layer lowers the drive voltage by reducing the effective barrier between ITO and NPB. NPB gro wth on plain ITO is by island formation, but on CuPc, NPB grows layer by layer, leading to better adhesion of NPB. UPS puts the Ionization potential of 2 3nm of CuPc at about 4.8 5eV. In UPS measurements, ITO features are suppressed after just 2nm of CuPc, which indicates that CuPc wets ITO very well. A vacuum level shift of 0.55eV is seen for ITO/CuPc, 0.35eV for ITO/NPB and just 0.05eV for CuPc/NPB. Based on this, its HOMO should lie between the Fermi level of ITO and NPB, thereby reducing the hole inject ion barrier (Figure 2 8). Conversely, 15nm thick CuPc is shown to retard the injection of holes into NPB, thereby achieving better efficiency through charge balance [67]. Time of Flight data showed that CuPc layers do not affect the transport properties of NPB. Dark Injection measurements confirm that 15nm CuPc reduces hole injection efficiency. Tadayyon et al. proposed a possible explanation for the discrepancy in the methods of increasing efficiency [68]. They suggested that there is increase in injection efficiency of CuPc regardless of the thickness. CuPc hole mobility and trapping concentration is reported to vary over two orders of magnitude, depending largely on deposition conditions like substrate temperature and rate of deposition. Thicker films mi ght mean more trapping centers and lower mobility, thereby retarding the flow of hole current, improving charge balance and hence less holes reach the cathode and are available for the formation of excitons. Moreover, the recombination takes place further away from the cathode. This increases the probability of radiative decay and therefore increases efficiency through charge balance.

PAGE 27

27 Polymeric Anodes as Ohmic Contacts for NPB: Poly (3,4 -ethylenedioxythiophene) (PEDOT), is a p conjugated polymer that is wid ely used as a conducting polymer anode, especially in the application of printed electronics. It has typically high conductivity along with excellent optical transparency. It typically has low solubility, but it is alleviated by the use of a water soluble polyelectrolyte poly(styrene sulfonic acid) PSS as a charge balancing dopant. Conductive PEDOT oligomers consisting of 5 15 repeat units are electrostatically attached to twisted long insulating PSS chains, which form grains with a PSS rich surface. PEDOT :PSS has a work function of about 5.2eV. The conductivity of the PEDOT polymer can be modulated by controlling the solvents, and can very from 0.01 500 S/cm. PEDOT forms a perfectly ohmic contact with m MTDATA, and a quasi -ohmic contact with NPB. As seen in the diagram (Figure 2 9), when different anodes or hole injection materials are used to inject holes into NPB, only PEDOT can form an ohmic contact [69]. This is because the hole injection barrier is less than 0.3eV, computer simulations show that the contact is capable of supplying the organic material with space charge limited current [70]. Other conducting polymer s like polythienothiophene (PTT) doped with poly(perfluoroethylene -perfluoroethersulfonic acid) (PFFSA ) have also been used as efficient p olymeric hole injection materials [71]. The work function of this polymer can be varied from 5.2 eV to 5.5 eV and has been shown to improve efficiency in fluorescent and phosphorescent OLEDs as well as extend their lifetime as compared to PEDOT:PSS and C uPc based OLEDs. Nickel Oxide as a transparent p -type Conducting Oxide Nickel Oxide is a well known ptype transparent semiconducting oxide. NiO has a higher work function of ~5eV as compared to the work function of ITO that is 4.7eV, therefore should be m ore favorable to charge injection into organic materials. NiO is deposited on ITO by reactive magnetron sputtering to a thickness of less than 10nm, as NiO is not as transparent as ITO.

PAGE 28

28 Typically NiO based OLEDs show better current injection due to the low ered injection barriers between NiO and TPD as compared to ITO based OLEDS of similar structure, but as the transmission is lower, the current efficiency of NiO devices also appears to be lower. [72] Conducting/Insulating Nature of WO3 Very high current de nsities have been reported for conducting metal oxides like NiO and CuOx, but the low optical transmittance of these materials creates difficulties with optimizing their efficiency and performance as OLED buffer materials. Typically the transmittance of WO3 on glass is very similar to bare glass, with the transmittance dropping to just 83% for a 50nm thick layer of WO3 [73]. Annealing the films upto 4500C does not change the transmittance of bare ITO and thin films of WO3, with the 50nm film dropping to about 75%. X -Ray Diffraction suggests that the morphology of WO3 film changes from amorphous to crystalline after annealing. The electrical conductivity also increases from 3.5x1011S/cm for amorphous films to 1.6x108S/cm for thermally annealed films. OLEDs that were fabricated with annealed WO3 showed a lowered turn on voltage for a 1nm thick WO3 anode buffer layer as compared to bare ITO and regular WO3 OLEDs. It was concluded that regular WO3 films act as electrical insulators, charges accumulate on the a node side of the buffer layer, and then inject through thin films via tunneling. On the other hand, annealed films act as a semiconductor with a work function 0.27eV higher than ITO, leading to injection by reducing energy barrier to NPB [73]. Charge Injec tion through Plasma Polymerized Fluorocarbon Films UVtreatment of ITO substrates is a good method for improvement of hole injection via reduction of the hole injection barrier, and raises the work function from 4.7eV to almost 5.1 eV. It has been found that the introduction of a 0.3nm layer of plasma polymerized fluorocarbons also achieves a similar increase in work function [74]. Polymerization is carried out in a parallel plate reactor with the precursor CHF3, detailed in ref [75]. Barrier heights determined by UPS is

PAGE 29

29 0.56eV for UV treated ITO/NPB and 0.5eV for plain ITO/CFx/NPB. This shows that the CFx film can replace UV treatment, and is a more reproducible chemical rather than physical treatment like UV treatment. CFx treatment also enhances injection from Au, which is very poor otherwise. Though the work function of Au is about 5.3 eV and therefore higher than ITO, an interface dipole of 1.24eV forms at the Au/NPB interface, which the Cfx films helps to eliminate by phyical separation of the materials At the ITO/CFx interface, UPS studies show a vacuum level shift of 0.61eV, suggesting the formation of a dipole oriented outward from the ITO surface due to the high concentration of negatively charged fluorine. A similar effect is seen in UV Ozone treat ment that coats the ITO surface with negatively charged oxygen ions, which explains the similarity in effect. However, the oxygen -coated surface is hydrophilic, and is rapidly degraded in atmospheric moisture. The inertness of the CFx film makes it air -sta ble and therefore is a viable method for ITO surface modification. The Use of MoO3 as an Anode Buffer Layer Tokito et al. reported the first case of introduction of MoO3 layer between ITO and a hole transporting layer in 1996 [76], where they show reduced operational voltage and enhanced efficiency. They attributed this to the lowered injection barrier for hole injection at the ITO HTL interface. In 2002, Reynolds et al. used MoO3 as an electron extraction barrier layer in PLEDs [77]. The same improvement was seen, which was attributed to the build up of electrons at the MoO3/emmisive polymer interface, resulting in a local field and a consequent enhancement in hole injection and transport. Matsushima et al. reported that a 0.75nm layer of MoO3 on ITO led t o the formation of an ohmic contact [78], and reported that the enhancement of hole injection is due to the electron transfer from ITO and the organic layer to MoO3. When compared to the traditional CuPc hole injection material, OLEDs with MoO3 as a buffe r layer show higher current densities, luminance and efficiencies when compared to corresponding CuPc based

PAGE 30

30 OLEDs. The MoO3 OLEDs are also more stable at higher temperatures, as they prevent the crystallization of NPB, which is a leading cause of small mol ecule OLED degradation. To understand the improvement caused by MoO3 OLEDs, the transmission and morphological characteristics were analyzed [79]. For light outcoupling efficiency, a high transmission film is desired, and insertion of a MoO3 film does not change the transmission much in the visible light region of 400 800 nm, whereas CuPc absorbs light in this range (Figure 2 10). The MoO3 film also improves the morphology and stability of the NPB film, and can effectively suppress the crystallization of NPB. AFM images of annealed NPB on MoO3 and CuPc show that buffer less NPB and CuPcNPB films show a greater degree of roughness, and crystallization after annealing, whereas MoO3 NPB films look smooth and featureless (Figure 2 11). UPS was used to study the change of energy levels when CuPc, MoO3 and MoO3 doped in NPB are deposited on plain ITO (Figure 2 12). The work function of ITO was found to be 4.7 eV and the buffer layers cause the energy level to shift to 4.55 eV, 5.2 eV and 5.1 eV respt. The vacu um level shifts down for CuPc and shifts up for MoO3, and the shift has been attributed to the formation of interface dipoles. For actual devices, an interfacial dipole of 0.15eV is formed at the ITO/NPB and CuPc/NPB interfaces, and a dipole of 0.05 eV is formed at the MoO3/NPB interface and there is no dipole for the MoO3:NPB/NPB interface (Figure 2 -13). Therefore the hole injection barrier for ITO/NPB devices is 0.55eV. The introduction of CuPc increases the barrier to 0.7eV, which is consistent with the reported explanation that CuPc blocks hole injection. However, the use of MoO3 and MoO3:NPB reduces the barrier from 0.55eV to 0.25eV and 0.3eV respectively (Figure 2 14).

PAGE 31

31 An explanation of the interfacial dipoles in MoO3 can be obtained from XPS data of 1nm of MoO3 deposited on ITO. It can be seen that the In 3d and the Sn 3d peaks shift to a higher binding energy after the deposition of MoO3 (Figure 2 15). This shift of binding energy indicates that there has been an electron transfer from ITO to MoO3 a t the interface, which is consistent with the electron accepting reputation of MoO3. The ITO side is consequently positively charged, and the MoO3 side becomes negatively charged, leading to an abrupt shift in the potential across the dipole layer [32].

PAGE 32

32 Figure 2 1. Plot of Coulombic Energy versus distance for nucleus Figure 2 2. Vacuum level at infinity and close to the solid

PAGE 33

33 Figure 2 3. Interface formation between a metal and an organic material

PAGE 34

34 Figure 2 4. Band Bending at a metal / organic interface Figure 2 5. UPS diagram of energy levels of metal and overlaid organic material

PAGE 35

35 Figure 2 6. Energy level offset at NPB/Alq3 interface

PAGE 36

36 Figure 2 7. Energy Offset at the Alq3/LiF/Al interface Alq with 5A0 of LiF ( ) and after 2A0 of Al ( )

PAGE 37

37 Figure 2 8. Energy level alignment of ITO, CuPc and PEDOT to NPB Figure 2 9. Formation of Ohmic Contact of NPB with PEDOT The solid line is the theoretical trap -free SCLC in NPB

PAGE 38

38 Figure 2 10. Transmission through MoO3 and CuPc based films on NPB Figure 2 11. AFM images showing surface roughness (I)NPB (II)CuPc (10 nm)/NPB (50 nm) (III)MoO3 (6 nm)/NPB (50 nm) (IV) MoO3:NPB (40 nm)/NPB (50 nm)

PAGE 39

39 Figure 2 12. UPS data for ITO, ITO/NPB, ITO/CuPc, ITO/MoO3, ITO/MoO3:NPB Figure 2 13. UPS data for ITO/NPB, ITO/CuPc/NPB, ITO/MoO3/NPB, ITO/MoO3:NPB/NPB

PAGE 40

40 Figure 2 14. Energy diagram showing offsets for ITO/NPB, ITO/CuPC, ITO/MoO3/NPB and ITO/MoO3:NPB/NPB Figure 2 15. XPS data showing energy shift of Indium 3d peaks

PAGE 41

41 CHAPTER 3 LITERATURE REVIEW -A LTERNATE MATERIALS F OR ANODES Charge Injection Introduction Under low electrical fields, typically less than 104 V/cm, almost all molecular materials behave like insulators. Purified and vacuum deposited materials have no intrinsic charge carriers, leading to very high resistivities around the order of 1015 cm, unless doped with donor or acceptor molecules. The mobility is also typically very low, on the order of 103 to 107 cm2/Vs. The contradictory high electrical currents seen in organic films (upto 1 A/cm2) are explained on the basis of carrier injection, space charge limited current and trap charge limited current. Conventional Approach to Carrier Injection For a typical metal organic interface, like Alq3/Mg, the vacuum level shift is about 1eV [32], which is too high for conventional injection to explain the high current densities observed. The generally accepted theories suggest that carriers are injected into organics via Richardson Schottky Thermal injection and Fowler Nordheim tunneling injection. The Fowler -Nordheim injection theory [80] ignores image charge effects and states that on application of an external field, a triangular energy barrier is formed as the energy levels of the materials distort to neutralize the built in potential, and the e lectrons from the metal Fermi level can then tunnel through this thin barrier and hop to available energy states in the bulk. Richardson -Schottky thermal injection [81] is made possible by localized levels introduced by structural defects and impurities. U pon deposition, an image force potential is formed due to Coulombic interaction between electrons injected and the holes left behind in the metal after the injection. The thermal emission theory says that an electron can be injected when it has acquired su fficient energy to overcome the potential barrier resulting from the superposition of the image force and the

PAGE 42

42 external potential. At very low temperatures, or for very large barriers, it is believed that electron tunneling through the energy barrier inject s the electrons. At room temperature, the injection is thermally activated and thermionic emission over the barrier injects electrons. The tunneling model neglects the effect of barrier lowering due to electric field, as well as makes assumptions about the existence of a triangular barrier and the availability of continuum unbound states into which the carriers can tunnel. On the other hand the Schottky model fails to account for inelastic scattering in the potential well, which is of crucial importance in organic solids [82]. Observed Electron Injection from Cathodes In the very widely used electron transporting material Alq3, it has been established that injection at the metal/organic interface is thermally activated at room temperature, thereby ruling out tunneling. However for the temperature range T<300K there is a marked difference from Arrhenius behavior, and the injection characteristics undergo a transition at ~100K, where it appears to become temperature independent [83]. Also, if we consider the el ectron hop from metal to organic to occur as a single, energetically significant event, then the injection efficiencies of various cathodes is not as expected [84]. It has often been proved that the injection efficiency decreases with an increase in work f unction of the cathode materials, and for a constant electric field, the injection efficiency should be an exponential function of the energy barrier. By this argument, as the work function of ITO is about 1eV more than Mg, we should expect that the inject ion efficiency from ITO is worse than Mg by a factor of 1018, when in fact it is just 105 (Figure 3 1) [83]. Therefore it is reasonable to assume that injection might not be via a single hop, but rather via intermediate energy states. Studies have shown that the presence of an interfacial dipole induces these intermediate states that participate in the injection process, and hopping out of a Gaussian distribution of intermediate states within the organic material near the

PAGE 43

43 metal interface produces the macros copically observed power law behavior of the current voltage curves. The Interface Injection Model UPS studies show a vacuum level shift of about 1eV at every metal/Alq3 interface [32], which is evidence for dipole formation at the interface due to charge separation. It has been proposed that due to this, injection can be modeled as a two-step process: metal to dipole and dipole to bulk states. There will be variations in the local dipole strength due to randomness of the interface morphology, and this disorder broadens the energy distribution of the transport states in the interfacial layer. As a consequence of the disorder, the initial hop from the metal Fermi level is not the energetically costly event as the interfacial states are in equilibrium with the cathode, and according to this model, we can therefore ignore the barrier between the metal and the interface states. The charges from the metal are readily injected into a broad distribution of interfacial energy states and then they must hop to a narrow distribution of higher energy bulk transport states (Figure 3 3c). In the figure it is also shown that the electric field is screened by the metal contact that lowers the HOMO and LUMO splitting energy at the interface by Ep ~0.2eV relative to the bulk [85]. It has been found that at 30K, the current is injection limited, yet it is seen that at very low temperatures, all cathodes possess nearly the same current voltage characteristics as seen in the Figure 3 2. This breaks the link between the cathode work function and the injection efficiency, because at that temperature, electrons are more settled at the Fermi level, and yet the similar injection characteristics is ignoring the difference in EF and this proves that the initial hop from the Metal Fermi lev el to the organic is not as significant. The similarities at ~30K also suggests that the organic interfacial site distributions is the same for all contacts.

PAGE 44

44 Carrier Injection into -NPB NPB is extensively used as a hole injection and transporting materi al between an anode and Alq3 in green OLEDs as well as a host for luminescent dopant materials. The contact between ITO and NPB can at best be described as a partially ohmic contact. It is well known that holes make up the dominant charge carrier in NPB, a s their mobility of 104 (cm2/V s) is orders of magnitude higher than that of electrons, and that transport of holes in NPB is governed by space charge limited conduction. The injection from the ITO Fermi level to the HOMO also occurs sequentially via int erface states [86,87]. It has been seen that low frequency (equilibrium) capacitance is proportional to the variation of the interfacial state occupancy with the bias voltage [88]. These interface states are filled at low bias voltages, which is seen as i ncreasingly positive capacitance in the low frequency limit of impedance spectroscopy, since the interfacial density of states increases in the direction of variation of the Fermi level at increasing bias voltages. At some bias voltage, an increment of the bias voltage will trigger depopulation of the interfacial states by a net influx of holes into the bulk organic layer. The low frequency capacitance is then seen to become negative (Figure 3 4). The interfacial states are pinned at the organic surface, an d their occupation reaches a maximum at ~0.4eV below the ITO work function and exhibit a Gaussian distribution. These levels can be interpreted as forming an effective hole injection barrier, which is determined by hole -injection kinetics. There are two m ain differences between these two analyses, and the interface injection model described in the previous section [83]. Firstly, as the name suggests, the interface injection theory of Baldo et al. assumes steady state injection from the electrode into the i nterfacial states, which is not energetically significant, that is, the interfacial states in the latter are assumed to be

PAGE 45

45 in equilibrium with the metal contact. This analysis [87] suggests that the interfacial states are filled at low bias energy and are depopulated at a certain bias voltage, implying that both steps are energy dependent. The location of the interfacial states is also disputed. The steady state injection model locates the Interfacial Density of States in the optical bandgap of the organic material, which are influenced by the interface dipoles to undergo broadening of the energy distribution to 3 4 times the bulk Gaussian distribution. The filling/depopulation model locates the interfacial states within the dipole layer formed at the interf ace. Nonetheless, both theories agree that the interfacial dipole and interfacial states are related, and injection via interfacial states has successfully explained the observed current -voltage characteristics. Review of Alternate Anodes Introduction ITO is the preferred anode because of its high transparency, high conductivity and most importantly, high work function. Some other oxides have also been evaluated as potential anode candidates, like aluminum doped zinc oxide, zirconium doped zinc oxide, fluor ine -doped tin oxide and zinc -doped indium oxide [72]. However, given our understanding of charge injection from electrodes, a number of novel composite anodes have been developed. This section provides a review of a few promising approaches. Use of p -Type Silicon Because of the prevalence of existing Si -based optoelectronics, OLEDs have been fabricated using p or n type silicon as electrodes for carrier injection [89]. When compared to an ITO based OLED, Si anode OLEDs showed lower turn -on voltage and higher current densities for the same forward voltage bias. However, for Si anode devices, both the Si anode and the Al electrode are relatively opaque as compared to ITO. The work function of both p Si and ITO is 4.7eV, and the IP for TPD is 5.5eV. The surfac e Fermi level of p -Si is usually pinned at 0.3eV

PAGE 46

46 above the top of the valence band, while the Fermi level in the bulk is 0.2eV above the valence band, resulting in a downward bending of 0.1eV. The effective barrier to injection is therefore 0.5eV. There is reasonable surface band bending upwards in ITO of about 0.1eV at the surface, and 0.03eV at the bulk. Moreover, the Fermi level is at the top of the conduction band (Figure 3 5). Therefore, the electrons see a barrier of 0.9eV between TPD and ITO, which a ccounts for better hole injection in p Si as compared to ITO. Semitransparent Silver Anode Silver is an optimum metal for transparent anodes because it has minimal absorption in the visible light region and has the lowest resistivity of all metals. Althou gh the transmittance is less than that of ITO, it can be improved by combining Ag with a oxide semiconductor with a high refractive index. An OLED structure of TiO2 20nm / Ag 10nm as the anode, CuPc as a hole injection layer and NPB as the hole transporting layer was fabricated [90]. Current density of Ag OLED is higher than ITO OLED, but luminance is comparable. The work function of the Ag anode was found to be 4.7eV and almost equal to ITO. Maximum transmittance of the TiO2/Ag anode is 74% at = 487 nm, and EL efficiency is expected to be lower due to this lower transmittance. A composite electrode consisting of (TiO2 20 nm / Ag 10 nm / TiO2 15nm) has a maximum transmittance of 87% at = 519 nm. However, when using the composite anode, the luminance is f ound to decrease as the TiO2 acts as an insulating layer between the organic material and the metal. Metal / Fullerene Electrodes C60 Fullerene is the only organic material to be used in all types of organic devices. C60 is a closed cage of 60 sp2 hybridi zed carbon atoms with icosahedral symmetry. Various devices were fabricated using the structure

PAGE 47

47 metal/C60(0,3nm)/NPB(45nm)/Alq3(45nm)/LiF(1nm)/Al(100nm). ITO, Au(15nm) and Mg (15nm) were used as the metal anode [91]. It was found that inclusion of a 3nm C60 interlayer reduced driving voltage for an Au/C60 anode by about 7V for 20mA/cm2. The current efficiency of the Au/C60 anode is also around 1 cd/A higher than for ITO anodes, while plain Au anodes are substantially inefficient. The power efficiency of Au/C60 anodes is also ~0.5 lm/W higher, despite a higher driving voltage, which implies that Au/C60 might actually limit the holes at the anode, resulting in better balance of holes and electrons. The same effect is seen for Mg and Mg/C60 anodes, and a reduc tion of 7V is seen again for the composite anode. From single carrier devices, it was seen that the current density at a given voltage was increased by more than two orders of magnitude as a result of the C60 interlayer for both Au and Mg anodes. The devic es with ITO showed the highest density, which is consistent with the hole current limiting model of C60. The work function of the Au/C60 anode was found to be 4.6eV as compared to the work function of Au, which is 5.1eV. It is well known that C60 forms a s trong dipole with metal surfaces. In the case of metal/C60 anode, the dipole pins the Fermi level of the metal to the LUMO of C60. Therefore the surface modification layer effectively modifies the work function of the underlying metal. Since the Fermi leve l is pinned, it was expected that the effective work function should be similar regardless of the metal work function. Oxides Of Underlying Metal Anodes As Effective Hole Injection Contacts Silver has the highest reflectivity of all metals in the visible region, as well as one of the lowest resistivities, therefore it has great potential as the bottom electrode in top emitting OLEDs. However it has a low work function of 4.3eV and therefore is not generally used as an anode material. Ag 2 O on the other hand demonstrates semiconducting properties. It has an ionization potential of 5.3eV, which is about one eV higher than Ag, and has a band gap of about 1.3 eV. UV Ozone treatment of Ag films is used to form the Ag2 O films. Treatment of the a

PAGE 48

48 films for more th an 3 minutes leads to a reduction of reflectivity and a yellowish film that indicates that the oxide film is over 100A. Devices were fabricated using the Ag/Ag2O as the anode, and m MTDATA as an additionally hole injecting layer. The cathode structure uses 1nm LiF and 1nm Al to form the electron injection contact, followed by 20nm Ag because it has highest conductivity and relatively low optical absorption. The electrode is then capped with a refractive index matching or antireflective layer of 40nm of TeO2 [92]. The EL spectra of the devices show higher current saturation as compared to bottom emitting devices and ITO devices due to stronger metal cavity effect. The emission pattern of the bottom -emitting device is close to a Lambertian distribution, while that of the top-emitting device shows enhancement in the forward direction again due to the stronger microcavity effect As a result, the top emitting devices with Ag/Ag2 O anodes show identical operating voltages and efficiencies as compared to ITO bottom emitting diodes. Charge Generation Interfaces T he use of 1,4,5,8,9,11-hexaazatriphe nylene -hexacarbonitrile (HAT CN) as the connecting layer in some of the tandem OLEDs for higher power efficiency was successfully demonstrated recently HAT CN was deposited on Gold and NPB was deposited on top to study the interface characteristics via UPS [93]. It was shown that the lowest molecular orbital of HAT CN is either pinned at, or is very close to the Fermi level of Au, while the HOMO of NPB is only 0.3eV below the Fermi level at the interface of HAT CN/NPB. This small HOMO -LUMO offset allows for electrons and holes to be generated and separated at the interface when external bias is applied, which is known as charge carrier generation. The Au work function is 5.2 eV and the maximum work function for the HAT CN film was about 6.1eV. Even at 80nm, the HAT CN does not cause charging of the film, which is in contrast to other organic materials like Alq3 which causes charging at 15 20nm and NPB which shows charging at 1.5nm. Such good

PAGE 49

49 conduction properties means HAT CN thickness can be used to tune the optical properties of the organic stack. The optical gap is smaller than the measured HOMO position, which makes the material almost like a metallic conductor. The six nitrile groups ensure that the material is a good electron acceptor. The HOMO level is 3.8eV and UV -vis spectra suggests that the LUMO is very close to the Fermi level, making it an n type material. The deposition of even 1.5 nm of NPB causes the HOMO to shif t to 0.3eV. This indicates that the electrons at the NPB HOMO can easily be excited to the HAT -CN LUMO. Feasibility of a Metal Anode using Molybdenum Trioxide Ohmic injection has been observed when a thin layer of MoO3 is inserted between ITO anode and N PB [78]. The energy alignment and the lowering of the injection barrier are dependent on the interfacial electronic structures at the ITO/MoO3/NPB interface and affects charge injection from the anode into the organic. A good way to investigate interfaces and the associated energy changes is via UPS studies of stepwise deposition of the layers. Wang et al. have shown a shift in the energy levels caused by deposition of these layers [79], and energy level changes within the layers is studied by Lee et al., who incrementally deposited ultrathin layers of NPB on 5nm of MoO3, and studied the energy level changes [94]. They found that there is an abrupt energy level shift of the cut -off position of 2.35eV when 5nm of MoO3 is deposited on cleaned ITO, indicating the existence of a huge interfacial dipole (Figure 37a). When NPB thickness was increased from no NPB to 2nm NPB, the cut off position relaxes back to higher energies, with the total secondary cut off shift caused by NPB deposition was about 1.90eV. A ver y significant consequence of this study of the observation of a gap state, formed at 0.15eV when 0.1nm and 0.3nm of NPB was deposited on MoO3 (Figure 3 7c). This gap state disappears at larger thicknesses of NPB, and was not seen when NPB was deposited dir ectly on ITO. It implies that the gap state depends of MoO3 and exists at a short range from the MoO3/NPB

PAGE 50

50 interface. Moreover, the gap state is formed by weak interaction, as XPS did not show any strong chemical reactions. Figure 3 8 shows the energy level diagram and the location of the gap state. Vacuum level shift due to the presence of interfacial dipoles has therefore been confirmed for MoO3 deposited on ITO. Yet until recently, studies had been confined to ITO anodes, and alternate metal anodes had no t been considered. With the recent success of MoO3 as a hole injection layer and a connection layer for tandem OLEDs, the feasibility of using MoO3 to facilitate charge injection from metal anodes is being reviewed. The performance of such a composite anode depends largely on the energy alignment and to understand this, incremental layers of MoO3 were deposited on Au and Al and the energy changes were studied using in -situ UPS [95]. As seen in Figure 3 9a, the secondary cut off shifts to a lower binding ene rgy on the deposition of MoO3 on Au, with a total vacuum level shift of 1.56eV. The valence band spectrum for this system is shown in Figure 3 9c. For a very thin layer of MoO3, a gap state appears at 0.8eV, very close to the Fermi level, which is ascribed to the reduced Mo5+ phase. With increasing MoO3 thickness, a second state appears at 1.5eV, while the first vanishes. These gap states assist in charge injection from Au to MoO3. The existence of a reduced phase of Molybdenum is also seen from XPS studies where the binding energy of the Mo 3d peak is similar to those published for Mo5+ in the initial layers of MoO3, and then shifts to Mo6+. With regards to charge transport, there is an overlap of the two gap states at 2nm, suggesting the transport can occ ur between these two states. Figure 3 10a shows the work function increase of Au and Al when MoO3 is deposited on the metal. Both show a huge increase in work function, with identical final work functions. Figure 3 10b shows the resulting energy diagram an d the associated gap states.

PAGE 51

51 Figure 3 1. Electron injection from various cathodes into Alq3 Figure 3 2. Electron Injection from various cathodes at 30K and 290K

PAGE 52

52 Figure 3 3. Interface Injection Model, showing the energy diagram modified by interfa cial states caused by dipoles (a), (b) and the energy distribution for the first few layers of the organic material (c).

PAGE 53

53 Figure 3 4. Inductance spectroscopy of ITO/NPB showing filling and depopulation at different bias voltages

PAGE 54

54 Figure 3 5. Energy diagram of (a)p -Si/TPD and (b) ITO/TPD Figure 3 6. Energy diagram of Au/HAT CN/NPB interface, showing charge generation possibilities due to alignment of Au Fermi level, HAT CN LUMO and NPB HOMO

PAGE 55

55 Figure 3 7. UPS studies of ITO/MoO3/NPB interface, sho wing shift in cut off position and the formation of a gap state within NPB Figure 3 8. Energy level diagram of ITO/NPB and ITO/MoO3/NPB showing a reduced hole injection barrier and the location of the gap state.

PAGE 56

56 Figure 3 9. UPS studies of MoO3 depos ited on clean Au, showing the shift in the cut off position and the presence of gap states in the MoO3 layer. Figure 3 10. Change in work function of Au and Al as a function of MoO3 thickness (a) and the energy level diagram (b)

PAGE 57

57 CHAPTER 4 EXPERIMENTAL METHODS Introduction This chapter covers actual process of fabrication and characterization of organic light emitting diodes in the Organic Electronics Laboratory at the University of Florida. The initial sections will cover the sample preparation methods along with a description of the tools and the equipment. The device architecture, the materials and chemicals that were used and the procedure for the fabrication are outlined. The next section talks about characterization of the devices. There is a descr iption of the testing setup and the standardized measurement techniques. Device Preparation Substrate Cleaning ITO patterned glass substrates were purchased from Tinwell Technologies, Ltd. The substrates measured 1 x 1 inch square and were sputter coated w ith 80nm of ITO with a reported sheet resistance of ~40 square. For the study of metal anodes, plain glass substrates were used, but all subsequent steps are the same. These films are inherently rough, so the cleaning process begins with mechanical rubbing with lint free cleaning wipes to clean and partly pla narize the surface. The glass substrates are then submerged in electronic grade Acetone purchased from Fischer Scientific and subjected to ultrasonication for 15 minutes in a Branson 3510 sonicator. This process is repeated for 15 minutes in laboratory gra de 2 isopropyl alcohol Ultrasonication uses the principle of cavitation to ensure removal of dust, contaminants and organic films that accumulate on the surface. The glass is then washed with de-ionized water to remove the solvent and then dried using ult ra high purity nitrogen.

PAGE 58

58 Surface Treatment Plain sputtered ITO has a work function of 4.3eV and is unsuitable for use as a hole injection contact for most organic devices, so they must undergo surface treatment to modulate their work function. The samples are therefore loaded into a UV Ozone treatment unit, made by the Jelight Company, Inc. Model 42, just prior to polymer deposition or thermal evaporation. They are irradiated for a period of 15 minutes. During this treatment, the ITO films are bombarded wi th high -energy oxygen ions that increase the oxygen concentration on the surface. Due to this, ITO undergoes physical and chemical changes that in turn cause modification of electronic properties. Addition of Polymeric Hole -Injection Layer In the case of t he devices that utilized a polymeric hole injection layer like PEDOT:PSS, the polymer is spin coated onto the cleaned and treated ITO substrates. The polymer plays the dual role of acting as a hole injection layer as well as planarizing the ITO surface. Th e latter role is crucial in the subsequent thermal evaporation of small molecules and metals on the surface. The planarization removes the inhomogeneities of the ITO surface, allowing uniform deposition. Its thickness also prevents the penetration of the hot cathode metal through to the anode and thereby also reduces the occurrence of electrical shorts. The polymer is spin coated using a Speciality Coating systems coater Model number KW 4A. About 0.5 mL of the polymer is cast onto the stationary substrate t remove impurities and restrict the size of the polymer chains that are deposited. The spin rates are ramped and set to control the thickness of the films. The excess solvent and polymer are remove d by centrifugal force. The samples are then baked in on a hot plate for 15 minutes at 1800C to remove all traces of the solvent and to anneal and harden the film. Other polymer films

PAGE 59

59 will require a different spin coating speed as their viscosities are dif ferent and the baking times will also vary, based on the composition of the preparation. Thermal Evaporation Thermal evaporation of small molecules, oxides and metals is carried out inside a vacuum chamber. The materials are patterned using a set of stainl ess steel shadow masks that are placed in close contact with the substrates. The anode mask corresponds to the ITO patterning on the glass and was used only for the metal anodes. The overlap of the anode and cathode masks creates four usable devices per su bstrate, each having an effective device area of 0.4cm2. The chamber is pumped down to a pressure of 106 Torr using a cryogenic pump, backed by oil sealed roughing pump. The materials are placed in evaporation boats that are radially arranged at the base of the evaporator, and the substrates are suspended above them in a face-down manner in a substrate holder. The materials and the substrates are separated by means of a movable shutter that allows deposition of the material only under specified conditions. The materials are placed in tungsten or boron nitride boats and connected between two posts to create an electrical connection. High electrical currents passing through the boats cause the materials to heat and evaporate. The evaporation rate is controlle d by a quartz oscillator and a feedback circuit, and the thickness correlations were tooled in using a profilometer. Due to the number of material s involved and the non-rotational nature of this particular evaporator, not all substrates are in the direct p lume of the evaporated materials, so there is always minor variation in the thicknesses. Upon completion of the evaporation, the chamber is backfilled with 1atm of nitrogen before the samples can be removed and transferred to the glove box for encapsulation.

PAGE 60

60 Encapsulation Organic devices are very susceptible to degradation due to atmospheric moisture and therefore must be encapsulated to create a physical separation between the organic materials and the atmosphere. Even then, diffusion of water molecules occurs through the encapsulant, but it is severely retarded. The samples are moved from the evaporator to the MBraun LabMaster 130 as expeditiously as possible. The glove box is filled with Nitrogen gas, and it maintains a water concentration of <0.1ppm. It utilizes a load lock system to minimize the introduction of moisture during the transport of objects into and out of the glovebox. Under these conditions a drop of epoxy resin (ELC 2500) obtained from ElectroClear is put on the organic stack via a micropi pette and is covered lightly with a glass cover slip. The resin is pre -dried to remove all traces of moisture from it by contacting it with a desiccating agent and then filtering it into a bottle that is permanently stored in the glovebox. The substrates a re then subjected to 2 minutes of ultraviolet light. This activates the epoxy, which is UV sensitive, and it begins to harden. The substrates are put back into the load lock vacuum or the glove box for about 5 minutes for the curing process of the epoxy to complete, and the devices are ready for characterization. Device Materials and Architecture The device architecture of the OLED consists of organic materials evaporated on the substrate in the case of small molecule materials and spin coated in the case o f polymeric materials. In case of polymeric hole injection materials, PEDOT:PSS (CLEVIOS -Al4083) purchased from Baytron. PEDOT is spun at a rate of 8000rpm, which makes a thin film of about 4050nm. It is then baked for 15 minutes at 1800C. The basic orga nic device structure is oxide or metal anode, hole injection layer, hole transport layer, electron transport layer, and cathode. The Hole injection layer is PEDOT:PSS or Molybdenum ( VI ) Oxide. The hole transport layer is naphthyl phenyl diamine (NPD) and t he electron transport layer is hydroxyquinoline aluminum

PAGE 61

61 (Alq3). The catho de consists of a thin layer of lithium f luoride covered by aluminum. The specifications are shown in Table 4 1. All metal anode materials were deposited at 0.5A/s, MoO3 was deposite d at 0.5A/s, NPB and Alq3 were deposited at 2A/s, LiF at 0.01A/s and the aluminum cathode at 2A/s. The different anode materials were ITO or metal anodes. The metal specifications are shown in Table 4 2. In the case of the metal anodes, the metal was cover ed with a 5nm thick layer of MoO3, which along with the metal formed a composite an ode. The MoO3 doubles as a hole injection layer. The structures of the organic materials are shown in Table 4 3. Characterization Techniques LIV M easurements Current Voltage Luminance measurements form the basis of the analysis of the feasibility of metal anodes. Current Voltage (I -V) data was collected using a two probe technique to determine the response of the device to electrical bias, using a Keithley Series 2400 sourcem eter Probes are connected such that t he diode is run in the forward bias. The voltage is swept in the positive direction, from 0V to 8V with a step of 100mV and a 0.001s source delay. The compliance of the sourcemeter was set at 100mA. The luminance was measured by a silicon photodiode, connected to a Keithley Series 6485 picoammeter to measure the response photocurrent. The photodiode is calibrated using a luminance meter. The sourcemeter and the picoammeter are controlled via GBIP interface and a comput er using an in -house developed LabView program. Sample H older The sample holder used in the characterization is a custom device holder, as shown in the figures and pictures. It consists of a plastic receptacle, in which gold pins are embedded. The pins are connected internally to corresponding pins on the external side of the holder, which are

PAGE 62

62 connected to the sourcemeter via alligator clips. The substrate is placed face down in the holder, so that the electrodes come in contact with the pins. There is a sp ring loaded cover on the holder that ensures that the glass substrate is firmly pressed down onto the pins. This way, the aluminum cathode ends up on the bottom and the semitransparent anode faces up. This configuration is ideal for bottom emitting diodes. The silicon diode is positioned above the center of the substrate, so that it receives only the perpendicularly emitting light waves. The photodiode, the substrate holder and the substrate are housed in a closed black box to optically isolate the system f rom the surrounding ambient light. Analysis of Spectrum The color spectrum of the light emitted from the diodes was measured via a spectrometer manufactured by Ocean Optics, Inc. It consists a dedicated photodiode connected via a 200micron optical fibre. The photodiode is calibrated via a white light simulator to get a bright and dark spectrum. Then the corrected spectrum is collected under similar conditions of optical isolation. The data is analyzed using the SpectraScan software by Ocean Op tics, Inc. T he spectrum obtained is cropped to the visible region, as that is the only region of interest, and is then normalized to the maximum value. Capacitance Voltage Measurements Capacitance-Voltage measurements were made to analyze the degradation of the organi c devices under electrical stressing. The capacitance was measured using a probe station, connected to a Agilent 4282A precision LCR meter. The LCR meter is first calibrated by doing a scan with the two probes not touching anything to calculate an open sig nal correction and then the two probes are kept in direct contact to get a short signal correction The probes are connected to the LCR meter using a 1m cable. The capacitance was measured in both the series and the parallel mode. The d.c signal was swept from 2V to 5V with a step of 0.01V in the

PAGE 63

63 forward direction. It is swept from the negative in order to avoid charging effects before the measurement. The a.c. signal is set at 100mV and 200 Hz. The LCR meter was controlled via an in -house developed LabVi ew program. Experimental Procedure Thickness Variation of the Molybdenum Oxide Hole Injection Layer The OLED devices were made with the architecture as described above, and the thickness of the MoO3 layer was varied from 0nm, that is direct contact of N PB onto the ITO substrate, to 5nm thick layer of MoO3. The devices were tested for current -voltage characteristics. The experiments were repeated to ensure reliability, and representative devices are presented in the next section. Study of Metallic Anodes In the development of metallic anodes, the substrates were cleaned as described and loaded into the evaporator. 5nm of the respective metals were deposited using an anode pattern. The MoO3 and the subsequent organic layers were all deposited using the orga nic mask and the aluminum cathode was deposited using a cathode pattern. This layer of MoO3 forms a composite anode with the metal and doubles as a hole injection layer. In the case of silver, a layer of MoO3 was deposited in the anode pattern below the si lver metal to improve adhesion to the glass and reduce the diffusion of silver into the organic stack. The lower layer of MoO3 also might play a role in increasing the transmittance of the metal anode by improving the index matching of the anode and the gl ass. Lifetime Measurement of OLEDs For the lifetime measurement of the OLEDs, the devices were fabricated as described above and encapsulated in the glovebox. After the L I -V and C -V measurement of the devices

PAGE 64

64 under pristine conditions, they were returned to the glovebox. In the glove box they were connected into an electrical circuit via alligator clips, and attached to the glass window of the glovebox using vacuum tape, such that the light emitted could be measured from outside the glovebox. The electric al circuit consisted of a constant voltage source and a 100ohm resistor in series with the diode. The voltage read across the resistor was converted to current density passing through the circuit knowing the resistance of the resistor, and the applied volt age was read off the constant voltage source. The devices were run at a constant current density of 20mA/cm2. As the device degrades, the voltage required to maintain that current density increases. The current density was monitored via the voltage across the resistor, and the voltage at the source was accordingly increased. The luminance of the devices was measured using a luminance meter. The glass does absorb some of the light and hence the readings from the luminance meter and the calibrated photodiode do not match But as the luminance for the lifetime study was always measured through the glass, the absorbance by the glass is constant, and the comparison of the relative output of light for all the devices at all measurement points still remains valid. A t periodic intervals, the devices were removed from the glovebox for the measurement of capacitance. The encapsulation ensured that the device did not degrade due to the atmospheric moisture in the short periods that the samples were not in the moisture fr ee glove box environment. Due to the frequent application and removal of the alligator clips, the electrodes tend to get scraped off, and therefore a contact blob of silver paste was put down on the electrodes before the start of the experiment, and was fo und to be effective and intact till the end of the study.

PAGE 65

65 Table 4 1. Specification of Organic materials and films Compound Thickness Manufacturer Material Code Specifications MoO 3 5nm Sigma Aldrich CAS 1313 27 5 99.99% metal basis NPB 75nm Luminescent T echnology Corporation. LT E101 Alq 3 50nm Luminescent Technology Corporation. LT 401 LiF 1nm Sigma Aldrich CAS 7789 24 4 <5 microns, 99.98+% Al 100nm Sigma Aldrich CAS 7429 90 5 5 15mm, 99.9% metal basis Table 4 2. Specifications of Metals and fil ms Compound Thickness Manufacturer Material Code Specifications Al 10nm Sigma Aldrich CAS 7429 90 5 5 15mm, 99.9% metal basis Au 10nm Sigma Aldrich CAS 7440 57 5 1 6mm, 99.999% metal basis Ag 10nm Sigma Aldrich CAS 7440 22 4 99.9999% metal basis Table 4 3. Structures of Organic Compounds Name PEDOT:PSS NPB Alq 3 MoO 3 Structure

PAGE 66

66 Figure 4 1. Schematic representation of device architecture

PAGE 67

67 Figure 4 2. Thermal Evaporation Set up

PAGE 68

68 Figure 4 3. L I -V measurement setup Figure 4 4. Dia grammatic Representation of L I -V Setup G

PAGE 69

69 Figure 4 5. Sample Holder with 4 OLEDs (seen with the blue color) Figure 4 6. Sample holder with one metal anode OLED illuminated

PAGE 70

70 CHAPTER 5 RESULTS AND DISCUSSI ON Molybdenum Oxide as a Hole -Injection Material For Green OLED s Background It is seen from the discussion on the theory of electroluminescence that the efficiency of organic light emitting diodes is largely dependent on the charge injection and charge balance of the holes and electrons into the organic material. In the course of research into efficient charge injectors, Molybdenum Oxide has begun to receive some attention in the last few years. It is a wide -gap material that behaves as an electron acceptor that is frequently used on the cathode side, an d as a connecting interlayer for tandem OLEDs. Its use as an efficient hole injection material has only recently begun to be investigated. The purpose of this study was to verify the applicability of MoO3 as an efficient hole injection layer, and try to understand the reason behind the enhancement of hole injection Results OLEDs were originally fabricated with and without a 5nm MoO3 buffer layer, and the electrical and luminescent characteristics were observed. Once the benefits of MoO3 were established, an intermediate thickness of 1nm MoO3 was used, so as to generate a gradation of the thickness of the buffer layer. The Current -Voltage Characteristics are shown in Graph 51. An increase in the current density was seen right from 0V upon the introduction of the 5nm MoO3 layer, which is clearly illustrated from the semi log plot. However, changing the thickness to 1nm increases the current density even more, reaching the compliance of 100mA before 8V. This is seen as the sudden flattening out of the I -V curv e at around 7.5V. The 1nm buffer layer OLED shows higher currents even before the turn-on voltage of around 2.2V. If the flattening is ignored and the curve is extrapolated, then it is seen that the introduction of a 1nm

PAGE 71

71 buffer layer increases the current density by about one decade at 8V, and a 5nm layer shows about half a decade improvement. The luminescent response of the OLEDs is shown in Graph 52. Similar enhancement of the light emission is seen for the MoO3 buffer layers, with the compliance cut of f of the current leading to corresponding flattening of the luminance curve too. The current efficiency plots are shown in Graph 5 3. The OLEDs with 5nm of the buffer layer show an efficiency of 4cd/A, which is nominally greater than the OLEDs without a bu ffer layer. The OLEDs with a buffer layer of 1nm show a current efficiency of 3cd/A. but over a much larger range of current densities. Discussion From this set of experiments, the benefits of a MoO3 buffer layer are obvious. Addition of even a single nano meter of MoO3 increases the current injection. Various studies on the effects of MoO3 have also noted that there is usually a dramatic increase of current injection and transport characteristics of organic devices when MoO3 is used as an interlayer or a ho le injection layer. In terms of just carrier injection, it seems that a thinner layer of the buffer layer is preferred. Studies on MoO3 interlayers has demonstrated that the introduction of 1nm of MoO3 shows the greatest enhancement over no MoO3 buffer lay er, with the relative effect of MoO3 dropping off as the thickness of the buffer layer is increased. This effect is also seen in this study. In another study in which the MoO3 buffer layer was sputtered rather than thermally evaporated, the results indicat e that a 0.75nm thick layer is optimum. Oxides are typically insulating materials, with low carrier mobility. From this study, we can conclude that the enhancement effect of MoO3 is more of an interface effect rather than a bulk effect, and therefore a thi nner layer minimizes any retardation of the carriers in the bulk. In the absence of dopants within the hole transporting material NPB, the luminance of OLEDs is based on the recombination of holes and electrons. Since OLEDs of this structure are believed t o be hole dominated, and therefore

PAGE 72

72 already in possession of an excess of holes, the observed increase in luminance is an interesting phenomenon. In NPB/Alq3 type OLEDs, there have been numerous reports of hole build up near the recombination interface, wit h the luminance being limited by the electrons reaching the interface. With MoO3 enhancing the injection of holes, an internal electric field might be set up within the organic stack to augment the existing external field, and promote the extraction of ele ctrons from the LiF/Al electrode. The increased density of electrons might explain the increased luminance of OLEDs with buffer layers. The current efficiency plots on first glance might seem to be contradictory to the rest of the data, with the efficiency of the 1nm buffer layer having an efficiency worse than the OLEDs with 5nm buffer and even no buffer layer. This is simply an effect of the definition of efficiency, being the ratio of the luminance over current density. In OLEDs with MoO3 buffer layer, t here is an increase in the current density and a corresponding increase in luminance, probably due to the reason outlined above. However, the improvement in luminance does not seem to be commensurate with the hole injection enhancement. For the magnitude o f increase of current density in 1nm MoO3 OLEDs, there has not been an appropriate increase in magnitude of luminance. Therefore, even though the 1nm and 5nm buffer OLEDs show enhanced luminance, the increase in luminance is low relative to the increase in current density and therefore by definition of efficiency, the 1nm buffer OLED shows lower efficiency and the 5nm shows efficiency comparable to no buffer layer OLEDs. It should therefore be noted that in this case, the lower efficiency is in no way atten uating the capability of MoO3 to inject holes, with the current density curves providing a much better measure.

PAGE 73

73 Metal -Based Composite Anodes Background Indium Tin Oxide (ITO) is currently the industry standard anode material. It possesses the advantages o f being a transparent conducting oxide having low sheet resistance and low optical absorbance. This makes it potentially very viable as an anode for bottom emitting devices. However, ITO has a comparatively low work function as compared to the HOMO of most organic materials, even after the extensively used surface modification techniques like plasma and UV -Ozone treatment. This difference presents a relatively large injection barrier for holes. Moreover, ITO also makes a poor contact with organic materials, forming only a partially ohmic contact with NPB. In order to reduce the hole -injection barrier as well as to improve the contact between the anode and the organic semiconductor, various buffer materials are introduced between the anode and the hole transp orting materials. Materials like PEDOT:PSS notably reduce the injection barrier by presenting an effective work function intermediate to the anode and the organic solid, and make an almost ohmic contact with NPB by planarizing the anode. With the success o f MoO3 as a buffer layer than can be thermally evaporated, we investigated the possibility of replacing ITO with metallic anodes, in an effort to displace the commercial dependence on a rapidly diminishing supply of Indium based electrodes. Results In a se ries of experiments, we made green emitting OLEDs using ITO as anode, and then replaced the ITO electrode with thermally evaporated metals, while keeping the remaining architecture of the OLEDs the same. The first series of devices were made without a MoO3 buffer layer. The ITO based OLEDs showed electroluminescence, indicating the injection of both carriers without the presence of a buffer layer. On the other hand, the metal anode OLEDs did not show any luminescent response, and the current voltage curves show negligible currents

PAGE 74

74 even at 8V. The open circuit response of these OLEDs showed that a buffer layer of MoO3 is required for the realization of metal anode OLEDs, and all subsequent OLEDs were fabricated with a 5nm buffer layer. This thickness might no t be optimum, but it provides efficiency values comparable to not having a buffer layer even though carrier injection is enhanced, as shown in the previous data. Three metals having very different work functions were chosen. The current voltage plots are s hown in Graph 5 4. It is observed that OLEDs with all the different anodes show the same shape of the J -V curve. All the anodes show the same turn-on voltage of around 2.25V. Though there is a variation in the characteristics before turn-on, the character istics are remarkably similar after turn -on, with the variation being less than half a decade at 8V. The luminescent responses are shown in Graph 55. The shape of the curves is again similar, with very low variation in the luminescent output for the diff erent anodic materials. It is interesting to note that the materials showing a relatively higher current density are not the same as those showing higher luminance. The current efficiencies were calculated and are shown in Graph 5 6. The ITO OLEDs are com parable to the previously made OLEDs. Though the OLEDs with Gold anodes have lower efficiency than ITO OLEDs, it is within the acceptable range for this species of OLED. The aluminum OLEDs are slightly lower than expected, but the silver anodes are actually better than ITO OLEDs at least on the basis of current efficiency, approaching 5 cd/A at higher densities. The normalized spectra of the light emitted from OLEDs with the different anodes are shown next in Graph 5 7. The OLEDs do not emit at any other w avelength, so the spectrum can be cropped to the visible region. All the emissions are peaked at about the same wavelength, and the spectral curves are similarly shaped. The Aluminum spectrum matches the ITO spectrum exactly, while the Gold and the Silver spectra are very similar. Both the latter are

PAGE 75

75 slightly red -shifted, and show emission over a slightly broader range, rather than having a sharp peak in the intensity. Discussion The reported values of the work functions of Aluminum, Silver and Gold are 4.28, 4.26 and 5.3 respectively. ITO has a work function ranging from 4.5 5eV after surface treatment. The conductivities of Aluminum, Silver and Gold are 37.8, 63.01 and 45.2 (x106 S/m). Aluminum anode OLEDs without a MoO3 buffer layer do not show any elec troluminescent response within the range of testing. However they do show weak luminance at approximately 15V. As a comparison, ITO based OLEDs show luminance at around 2.2 2.5 V. This implies that the injection barrier for an aluminum anode without the presence of MoO3 is very large, but under sufficient external bias, it will inject electrons. But from Graph 5 4, the presence of MoO3 brings down the turn-on voltage to around 2V. From this we can conclude that the MoO3 buffer layer reduces the injection barrier. However Gold anode OLEDs without the buffer layer also do not show electroluminescence. This implies that although Gold and ITO have comparable work functions, gold anodes are poor injectors. In that case MoO3 probably forms a composite anode with the metal and thereby improves injection. It is therefore very significant that the current voltage characteristics of all the different anodes align so well. The fact that they all have the same turn -on voltage implies that the effective injection barrie r is either the same or very similar. The fact that there is no injection barrier at the metal/MoO3 interface, though the work function difference between Aluminum and Moo3 is large, also points towards the formation of a composite anode. The Luminance vol tage curves also mirror these observations. Since luminance depends on recombination of holes and electrons, similar luminance curves imply that holes are being injected from MoO3 into NPB at the same efficiency. So MoO3 itself is maintaining its bulk prop erties and its interactions with NPB, and therefore the effect is

PAGE 76

76 interfacial in nature and is seen only at the metal /MoO3 interface. Another very interesting outcome from analysis of the luminance curves is that all metal anode OLEDs display comparable or higher luminance than ITO OLEDs. This should not have been the case, since the OLED architecture is bottom emitting, and all the metal anodes have lower transmission coefficients than ITO. When combined with the spectral results, the cause of this discr epancy can be explained via the formation of a metal cavity. When the unbiased OLED pixels are visually observed, the ITO OLEDs have a gray color due to the underlying metal cathode, but the metal anodes OLEDs have an almost opaque violet color that underw ent a red shift when the overall thickness of the stack was increased. The existence of a metal cavity would mean that there will be internal reflection of light, and the emitted light will be collimated due to constructive interference within the cavity. This means the will be less destructive interference and divergence after the light leaves the device, and consequently, more light reaches the detector. The metal cavity promotes a non Lambertian distribution of the emission intensified along the normal d irection, which also translates to increased photo -detection. Another effect of the metal cavity can be seen in the emission spectra. The ITO OLED and the Aluminum OLED which shows only a very weak cavity effect shows a peak in the spectrum at 512nm, but t he Silver and Gold anode show a red shift as well as a broadening of the maximum intensity to show a range 525nm to 550nm, rather than a peak. This effect is contrary to the conventional metal cavity effect that shows a narrowing of the emission spectrum i nstead. Silver has the highest reflectivity of all metals, and therefore forms the best mirror for the metal cavity, and this is seen in the luminance plots, where the luminance of Silver anode OLEDs is comparatively higher. The higher reflectivity and conductivity of Silver do contribute to the slight differences in Luminance and current data, and the efficiencies are therefore understandably higher.

PAGE 77

77 It should be pointed out that the efficiency calculations for this set of experiments could be misleading for two reasons. First, the effect of the metal cavity is as described above. Due to this, the luminance that is measured for the Silver and Gold OLEDs cannot be compared to the luminance of the ITO OLED that lacks the metal cavity enhancement. This is an optical effect and has no real bearing on the physics of the metal anodes being studied. Therefore to equate these quantities, we can qualitatively assume that the luminance of the Silver and Gold OLEDs is in fact lower, which would mean that their effici encies have been overestimated. On the other hand, the absorption and transmission of the metal anodes is not accounted for, and as the ITO anode is substantially more transparent, the luminescent emission from all the metal anode OLEDs might be damped by the translucence of the anodes, again causing misrepresentation of the efficiency. Therefore, the efficiencies of the metal anodes should not be directly compared to ITO OLEDs. Another important consideration in this study is the quality of the metal anod es. The ITO anodes are commercially sputtered to a thickness of 80nm, and therefore are sufficiently thick and uniform. The thickness of the metal anodes was deliberately kept low to improve the transmittance and so they were only 10nm thick, and were ther mally evaporated onto glass substrates. The effect of minor thickness variation would be much more pronounced. It is therefore reasonable to assume that the contact between the metal and subsequently deposited materials would be worse than the ITO anodes. The second concern is that since the metal anodes were so thin, the current would be passing through a much smaller cross section. The metal anode OLEDs were therefore subjected to a higher cross sectional current density in order to compare them to ITO OL EDs. However, due to this, the effect of joule heating would be significantly more pronounced, and the metal anode OLEDs were also expected to degrade

PAGE 78

78 faster, and indeed degradation could be seen in current characteristics after a single sweep in some case s. Lifetime and Degradation of OLEDs Based o n the Performance of Hole -Injection Materials Background Apart from efficiency, the operating lifetime of OLEDs is the other obstacle to its commercial application, and there have been concerted effort to underst and the nature of the degradation of OLEDs. Sometimes a compromise must be made between efficiency and lifetime, as in the case of hole transporting materials. OLEDs made using TPD, a material with a lower glass transition temperature, which translates to faster crystallization usually have lower lifetime, yet typically show higher efficiencies. The case for NPB is the exact opposite. In another landmark paper by Kondakov et al., [96] degradation has been attributed to the build up of holes at the NPB/Alq3 interface. By extension of this, there should be a build up of holes at the anode/hole transport material interface too. By varying the hole injection material, we can vary the effective barrier for hole injection and therefore vary the concentration of holes at this interface. This study provides a preliminary investigation of degradation at the anode/semiconductor interface. Results In this study, we fabricated OLEDs using a number of commercial and proprietary hole injection materials. Among these were P EDOT:PSS, polythienothiophene (PTT) doped with poly(perfluoroethylene -perfluoroethersulfonic acid) ( PTT: PFFSA) [71], MoO3 and CuPc. We made a comparison of OLEDs using polymeric hole injection materials. The architecture was a little different, with the NP B and Alq3 layers being 40nm and 50nm respectively. The polymers were spin coated using recipes optimized for them individually, and the resulting annealed layers were around 70nm thick. Graph 5 8 shows the current voltage curves for them, and an OLED

PAGE 79

79 usin g 5nm MoO3 as a buffer layer is included for comparison. The plot also shows the luminescent response, and it is seen that the luminance is roughly proportional to the current density. It can also be seen that the current densities and the luminance of all the OLEDs is very similar, and the turn on voltage for all devices is the same. The current efficiencies are shown in Graph 5 9, and again they are remarkably similar. The MoO3 OLED does show a higher efficiency in spite of lower current density and lumin ance values. For the purpose of lifetime comparison, an OLED was fabricated without a hole injection layer, and was compared to the OLEDs with polymeric buffer layers. The devices were operated continuously at 20mA/cm 2 and the measurement study was conduct ed as outlined in the experimental section. The normalized luminescent decay is shown in Graph 5 10, and the voltage rise data is summarized in Graph 511. It is obvious that the lifetime of devices with similar performance is quite different. The Baytron PEDOT:PSS (CH8000) reaches half life (50% luminance) in less 20 hours, the OLED without a buffer layer degrades 45% and PTT:PFFSA 2 degrades 10% after more than 120 hours. We also compared the degradation of OLEDs without a buffer layer and polythienothi ophene (PTT) doped with poly(perfluoroethylene perfluoroethersulfonic acid) (PTT: PFFSA 2 ), using C -V analysis. These two devices represent the largest and smallest effective hole -injection barrier respectively. The C -V data is presented in Graph 5 12. The important characteristic is the turn -on voltage, and we can see that it shifts to the right over the course of the degradation, although the rate of the shift is significantly slower for PTT:PFFSA 2 OLEDs than for OLEDs without a buffer layer. Discussion T he operating lifetime has so far been linked mainly to the degradation of the organic materials. A very common hole transporting material TPD has a lower glass transition temperature as compared to NPB, which leads to more rapid crystallization. Therefore the higher

PAGE 80

80 efficiency of TPD OLEDs is compromised by its shorter lifetime. In Kondakovs paper, [97] the chemical degradation of NPB is due to the presence of excitons formed after the recombination of holes and electrons at the interface of NPB/Alq3, whic h leads to fragmentation of the NPB molecule progressively away from the interface. This leads to the decay of luminance over time. In our experiment, the OLED structure for all devices is the same apart from the hole injection layer, and yet we see a diff erence in lifetime. Based on this, we can say a priori that there is also some change occurring near the anode/semiconductor interface. The primary difference between these materials is their work function, thereby presenting a different effective injectio n barrier. In 1995, in their work on different hole transporting materials, Adachi et al. suggested that the hole injection barrier causes joule heating near the anode interface, which causes local crystallization of the material, and subsequent degradatio n [98]. The correlation of the hole injection barrier and the rate of degradation does in part confirm this model. But it contradicts the very convincing argument put forth by Kondakov, that states both holes and electrons are required to initiate degradat ion. It was also observed by Scott et al. that oxygen diffuses out of ITO into a polymeric OLED, causing it to fragment, reduce carrier mobility and quench luminescence [99]. UV Ozone treatment is a common method to increase the work function of ITO, by w hich the surface of ITO is enriched by oxygen atoms. It has been observed the work function of treated ITO will reduce if the anode is left in atmosphere, indicating that the adsorbed oxygen is being released. Under electrical stress, it is possible that o xygen is released from the anode and diffuses into the organic material. This would have a combined effect of degrading the organic material as well as lowering the work function of ITO, which would explain the loss in luminance and rise in driving voltage It has also been observed that Indium diffuses out of ITO into PEDOT:PSS,

PAGE 81

81 particularly during the spin coating process[100], to the same destructive effect. The different hole -injection materials have different resistances to these diffusive elements and this may contribute to the difference observed in their lifetimes. The second evidence we present is analysis of degradation via C -V measurements in Figure XII. S hrotriya et al. analyzed the C -V response of OLEDs [101] and have attributed the turn on rise of capacitance as the onset of injection of holes, and the capacitance peak as the point of carrier recombination. The shift of the capacitance turn -on voltage implies that more voltage is required to inject holes. This increase in resistance to carrier injection could be due to the chemical degradation of the materials or the change in the work function of the anodes. Therefore if the anode/organic interface a location for degradation, this could affect the interfacial dipole and therefore hampe r charge injection. Based on the current understanding of interfaces, it could be anything from modification of the characteristics of the interfacial dipoles to the redistribution of the energies of the intermediate states associated with them. Without fu rther investigation, we cannot comment with any certainty about the mechanism of this degradation. All we can conclude is that better hole injection retards this particular cause of degradation.

PAGE 82

82 Figure 5 1. Current Density plots as a function of thickness of MoO3 layer Figure 5 2. Luminance Plots as a function of thickness of the MoO3 layer 0 2 4 6 8 1 10 100 1000 10000 LuminanceVoltage ITO ITO / Moly 1nm ITO / MoO3 5nm 0 2 4 6 8 1E-5 1E-4 1E-3 0.01 0.1 1 10 100 1000 Current DensityVoltage ITO ITO/MoO3 1nm ITO/ MoO3 5nm

PAGE 83

83 Figure 5 3. Current Efficiency Plots as a function of MoO3 layer thickness 0 10 20 30 40 50 0 1 2 3 4 5 Current EfficiencyCurrent Density ITO ITO/MoO3 1nm ITO / MoO3 5nm

PAGE 84

84 Fi gure 5 4. Current Density plots of OLEDS with different anodes Figure 5 5. Luminance plots of OLEDs with different anodes 0 2 4 6 8 1 10 100 1000 10000 Luminance (cd/m2)Voltage (V) ITO / MoO3 Al / MoO3 Au / MoO3 Ag / MoO3 0 2 4 6 8 1E-6 1E-5 1E-4 1E-3 0.01 0.1 1 10 100 1000 Current Density (mA/cm2)Voltage (V) ITO / MoO3 Al / MoO3 Au / MoO3 Ag / MoO3

PAGE 85

85 Figure 5 6. Current Efficiency plots of OLEDs with different anodes 20 40 60 80 100 1 2 3 4 5 Current Efficiency (cd/A)Current Density (mA/cm2) ITO / MoO3 Al / MoO3 Au / MoO3 Ag / MoO3

PAGE 86

86 Figure 5 7. Emission Spectra of OLEDs with different anodes 350 400 450 500 550 600 650 700 750 800 850 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Normalized IntensityWavelength (nm) ITO / MoO3 Al / MoO3 Au / MoO3 Ag / MoO3

PAGE 87

87 Figure 5 8. Current Density and Luminance plots for OLEDs made with various polymeric hole injection materials. INSET: Luminance response of the OLEDs 0 2 4 6 8 101E-3 0.01 0.1 1 10 100 1000 APHIL98 APHIL136 PEDOT8000 PEDOT4083 MoO3 5nmCurrent density [mA/cm2]Voltage [V]100 1000 10000 100000 Luminance [cd/m2]

PAGE 88

8 8 Figure 5 9. Current Efficiency plots of OLEDs with polymeric hole injection materials 0.1 1 10 100 1 2 3 4 5 APHIL98 APHIL136 PEDOT8000 PEDOT4083 MoO3 5nmCurrent efficiency [cd/A]Current density [mA/cm2]

PAGE 89

89 Figure 5 1 0 L i fe t i me Decay of luminance of OLEDs with different polymeric hole injection materials Figure 5 11. Lifetime Voltage rise of OLEDs with different polymeric hole -injection materials 0 20 40 60 80 100 120 140 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 Normalized LuminanceTime (hr) No HIL PEDOT 8000 PEDOT 4083 AP HIL 98 Aphil 136 0 20 40 60 80 100 120 140 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50 Normalized VoltageTime (hr) NO HIL PEDOT 8000 PEDOT 4083 APHIL 98 APHIL 136

PAGE 90

90 -2 -1 0 1 2 6.00E-010 8.00E-010 1.00E-009 1.20E-009 1.40E-009 1.60E-009 1.80E-009 2.00E-009 2.20E-009 No HIL 00 hours 24 hours 48 hours 72 hours 96 hoursCapacitance [F]Voltage [V] Figure 5 12. Capacitance -Voltage Data of OLEDs under electrical stress: -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 1.50E-009 2.00E-009 2.50E-009 3.00E-009 3.50E-009 4.00E-009 4.50E-009 5.00E-009 Voltage [V]Capacitance [F] 00 hours 48 hours 72 hours 96 hours 120 hours APHIL136

PAGE 91

91 CHAPTER 6 SUMMARY AND FUTURE W ORK Summary of Results Green fluorescent OLEDs were fa bricated with NPB and Alq3 as the hole transporting and electron transporting materials respectively to study the nature of injection across a metal/organic semiconductor interface. It was found that performance of MoO3 was comparable to contemporary mater ials like CuPc and PEDOT:PSS, and they all have similar operating voltages and efficiencies. MoO3 combines the advantages of high transparency and ohmic contact nature of PEDOT:PSS and the ability to be thermally evaporated like CuPc. The effect of the thi ckness of MoO3 on the electronic properties of OLEDs was examined, and it was found that the presence of the buffer layer improves the performance of the device significantly. There is a dramatic enhancement upon the introduction of a nanometer -thick layer of MoO3, and the relative improvement drops off with increasing thickness. This implies that the electron accepting MoO3 retards holes in its bulk and its entire efficacy lies in its role as an interface modification agent. ITO has a work function ranging from 4.5 5eV after surface treatment, whereas Aluminum, Silver and Gold have work functions of 4.28, 4.26 and 5.1 respectively. The HOMO level of NPB is 5.4eV. Yet in the presence of MoO3, OLEDs fabricated using all these materials as anodes behave identically. Performance curves and the emission spectra are very similar if you take into account the formation of the metal cavity. These results suggest that the presence of MoO3 neutralizes the difference in the work functions of the anodes, and the only e ffective work function is that of MoO3. A comparative study of the lifetime of OLEDs was conducted using a number of conducting polymers to inject holes. It was observed that the rate of degradation of the OLED is

PAGE 92

92 retarded when the efficiency of the device is improved via better hole -injection characteristics. This study concludes that at least an empirical relation exists between the performance, operating lifetime and the injection barrier in organic devices. Future Work Despite being a strong electron acceptor, Molybdenum trioxide has been used extensively has a hole -injection material in conjunction with ITO. This study shows that in the presence of MoO3, any thin conducting metal may replace the ITO anode. This has great implications in display techno logy, as ITO is the predominant transparent anode material. If we account for the metal cavity effect, we conclude that the efficiency of the ITO -based OLEDS is significantly better; nonetheless this study serves as a proof of concept, that with optimizati on, metal anodes can be engineered to be a viable substitute for ITO. As a sample improvement, if the metal anode were to be sputtered rather than thermally evaporated, the resulting metal film would be of higher quality, as well as more transparent. It mi ght be interesting to repeat this experiment, using these films to see whether the result can be enhanced. Sputtering would also allow for a thicker metal film to be used without compromising the transparency, and would avoid the deleterious effects of jou le heating. The lifetime study of MoO3 based OLEDs is also worthwhile, because of the relation of efficiency, lifetime and injection barrier. Currently, metal anode OLEDs are too unstable to undergo any degree of prolonged electrical stressing and degrade after a few sweeps, but as the methods of depositing metal anodes are improved, and the interaction between the metal and oxide is better understood, this might be attempted. Clearly, the catalyst of this study is the Molybdenum trioxide film. Yet, we have an incomplete understating about the nature of the interfaces between MoO3 and ITO, MoO3 and the metal, and MoO3 and the organic material. In recent findings, it has been shown that MoO3

PAGE 93

93 accepts electrons from Indium from ITO to cause the In 3d peaks in a n XPS spectrum to move to a higher energy [79]. Gain of electrons by the MoO3 means it undergoes reduction. In this context, it would mean that Mo6+ ions are being reduced to Mo4+ ions, which means Molybdenum trioxide might now be Molybdenum dioxide. Indee d, studies have shown than amorphous MoO3 does reduce to a crystalline MoO2 in the presence of hydrogen [102], which could also be brought about by organic hydrocarbons as well as ITO. MoO2 is known to have a higher conductivity than MoO3. The conversion of MoO3 to MoO2 is probably limited to a few atomic layers, which helps explain why thinner films show higher current densities. XPS studies of MoO3 films on Al, Au and Ag should show energy shift of a similar nature. Similarly, deposition of an organic lay er on MoO3 should show an energy shift in the energy of Mo peaks, with Mo moving to a lower energy state. If this reduction is better understood, it may be possible to manipulate it and thereby improve on the performance of current MoOx devices. The shift of the work function of the metal with respect to that of MoO3 is also a crucial piece of evidence for this work. To qualitatively study this, a photovoltaic measurement may be conducted that would help to determine the open circuit voltage of the device. In this technique, the photocurrents of the devices are measured under illumination by a simulated solar lamp. If all the parameters are kept the same and the interface dipole effect is ignored, then the VOC reflects the built in potential between the elec trodes. A study previously conducted on MoO3 OLEDs shows an increase on VOC as compared to an ITO OLED without the buffer layer. An increase in VOC translates to a smaller hole injection barrier as the effective work function of the anode is increased. If we make OLEDs with metal anodes and MoO3 and measure VOC, and see the same built in potential for all OLEDs, that would tell us something about the nature of the shift in work function of the metal.

PAGE 94

94 Similarly, UPS studies also tell us a great deal about the work function of the anode structure, and can be used in conjunction to show that metals in the presence of MoO3 show the same effective work function and therefore the same barrier to hole injection. There have been studies conducted on the existence of gap states in MoO3 when deposited on metals and the presence of gap states in NPB when deposited on MoO3. UPS studies of the complete system conducted incrementally would help explain the nature of carrier injection The lifetime study of OLEDs as a function of hole injection barriers is very significant from an economic point of view. S ingle carrier devices fabricated without Alq3 and a very high work function metal like Au as cathode would eliminate the NPB/Alq3 interface. C harge transport would then be in a single material an d with a single metal/organic interface. The voltage rise data and C -V data should show any change in driving voltage, a change indicating deterioration of either the interface o r the material itself. C hange in work function or built in potential combined with the voltage data would i so late the effect o f this interface and provide deeper insight into the nature of the degradation since this is a simpler system as compared to a bi layer, dual carrier system. An informative study would be to electrically stress the devices and compare the concentrations of Indium and oxygen in the polymer and in the NPB layer via XPS. T he work function changes of the anodes may be measur ed by conducting a Kelvin probe or a temperature dependent study, which would non invasively give the work function change as a function of time of electrical stress. I f diffusion of elements were indeed found to be the cause of degradation, it may be possible to design a better and more resistant hole injection material and processing method that retards a node etching that wou l d significantly enhan ce its lifetime and perform ance

PAGE 95

95 REFERENCES 1 A. Bernanose, M. Comte a nd P. Vouaux, J. Chem Phys. 50, 65 (1953) 2 M. Pope, H. Kallmann, and P. Magnate, J. Chem. Phys. 38, 2042 (1963) 3 M. Sano, M. Pope, and H. Kallmann, J. Chem. Phys. 43, 2920 (1965) 4 W. Helfrich and W.G. Schneider, J. Chem Phys. 14, 229 (1965) 5 W.G. Schneider and W. Helfrich, J. Chem. Phys 44, 2902 (1966) 6 G.G. Roberts, M.M. McGinnity, W.A. Barlow, and P.S. Vincett, Solid State Commun. 32, 683 (1979). 7 P.S. Vincett, W.A. Barlow, R.A. Hann, and G.G. Roberts, Thin Solid Films 94, 171 (1982) 8 R.H. Partridge, Po lymer 24, 748, (1983) 9 S. Hayashi, H. Etoh, and S. Saito, Jpn. J. Appl. Phys. 25, L773 (1986) 10. C.W. Tang, and S.A. VanSlyke, Appl. Phys. Lett. 51, 913, (1987). 11. J.H. Burroughes, Nature, 347, 539 (1990). 12. G He, O Schneider, D Qin, X Zhou, M Pfeiffer, and K Leo, J. Appl. Phys. 95, 5773 (2004) 13. R.F. Service, Science, 310, 1762 (2005) 14. J Kido, H Hayase, K Honyawa, K Nagai, and K Okuyama, Appl.Phys.Lett., 65, 2124 (1994) 15. Y.Yang, Q.Pei, and A.J.Heeger, J.Appl.Phys. 79, 934 (1996) 16. J. Kido, H. Shionoga, and K. Nagai, Appl. Phys. Lett., 67, 2281(1995) 17. C-C Wu, C Wei Chen, and T Y Cho, Appl. Phys. Lett., 83, 611 (2003) 18. K.C. Kao, W. Hwang, Electrical Transport in Solids with Particular Reference to Organic Semiconductors, Pergamon Press, New York, 1981. 19. Z. Kafaf i, Organic Electroluminescence CRC press, 2005 20. L. S. Hung C. W. Tang and M. G. Mason Appl. Phys. Lett. 70, 152 (1997) 21. C.W. Tang, C.H. Chen, R. Goswami, US Patent 4,769,292 (1988).

PAGE 96

96 22. S.A. VanSlyke, C.H. Chen, C.W. Tang, Appl. Phys. Lett. 69 (1996) 2160 23. Y. Shirota, Y. Kuwabara, H. Ina da, Appl. Phys. Lett. 65, 807 (1994). 24. Y. Yang, A.J. Heeger, Appl. Phys. Lett. 64, 1245 (1994). 25. Y. Cao, G. Yu, C. Zhang, R. Menon, A.J. Heeger, Synth. Met. 87, 171 (1997) 26. A. Elschner, F. Bruder, H. W. Heuer, F. Jonas, A. Karbach, S. Kirchmeyer, S. Thurm R. Wehrmann, Synth. Met. 111, 139 (2000) 27. Y Hamada T Sano K Shibata K Kuroki Jpn. J. Appl. Phys. 34 L824 (1995) 28. M Yoshida, A Fujii, Y Ohmori, and K Yoshino, Appl. Phys. Lett. 69, 734 (1996) 29. M.A. Baldo, ME Thompson, SR Forrest Pure Appl. Chem., 71, 11, 2095, (1999) 30. M.A. Baldo, M.E. Thompson, S.R. Forrest, Nature 403, 750 (2000) 31. C. B. Duke, L. B. Schein, Phys. T oday 33, 42 (1980). 32. H Ishii, K Sugiyama, E Ito, K Seki, Adv Mater. 11,8, 605 (1999) 33. D. P. Woodruff, T.A. Delchar, Modern Techniques of Surface Science, Cambridge University Press, Cambridge (1986). 34. K. Wandelt, in Thin Metal Films and Gas Chemisorption (Ed: P. Wissmann), Elsevier, Amsterdam (1987). 35. N. Lang,W. Kohn, Phys. Rev. B 1, 4555 (1970). 36. N. D. Lang,W. Kohn, Phys. Rev. B, 3, 1215 (1971) 37. J. Hoelzel, F. K. Schulte, H. Wagner, Solid State Surface Physics, Springer, Berlin (1979). 38. C. Slater, N. H. Frank, Electromagnetism, McGraw Hill, New York, (1947). 39. S. M. Sze, Physics of Semiconductor Devices, 2nd ed.,Wiley, New York (1981). 40. M. Schott, in Organic Conductors: Fundamentals and Applications (Ed: P. Farges), Marcel Dekker, New York (1994). 41. K. C. K ao, W. Hwang, Electrical Transport in Solids, Pergamon, Oxford (1981) 42. Y. Park, V. Choong, E. Ettedgui, Y. Gao, B. R. Hsieh, T. Wehrmeister, K. Muellen, Appl. Phys. Lett. 69, 1080 (1996). 43. A. Rajagopal, C. I. Wu, A. Kahn, J. Appl. Phys. 83, 2649. (1998)

PAGE 97

97 44. K Seki, Mol. Cryst. Liq. Cryst. 1989, 171, 255. 45. K. Seki, in Optical Techniques to Characterize Polymer Systems (Ed: H. Baessler), Elsevier, Amsterdam (1989) 46. S. H fner, Photoelectron Spectroscopy, 2nd ed., Springer, Berlin (1996). 47. A. Rajagopal C. I. Wu and A. Kahn J. Appl. Phys. 83, 2649 (1998) 48. I. D. Parker, J. Appl. Phys. 75, 1656 (1994). 49. Q. T. Le, L. Yan, Y. Gao, M. G. Mason, D. J. Giesen, and C. W. Tang, J. Appl. Phys. 87, 375 (2000). 50. M. G. Mason, C. W. Tang, L. S. Hung, P. Raychaudhuri, J. Madathil, D. J. Giesen, L. Yan, Q. T. Le, Y. Gao, S. Y. Lee, L. S. Liao, L. F. Cheng, W. R. Salaneck, D. A. d. Santos, J. L. Bredas J. Appl. Phys. 89, 2756 (2001) 51. L. S. Hung, C. W. Tang, and M. G. Mason, Appl. Phys. Lett. 70, 152 (1997) 52. T. M. Brown, R. H. Friend, I. S. Millard, D. J. Lacey, T. B utler, J. H. Burroughes, and F. Cacialli, J. Appl. Phys. 93, 6159 (2003) 53. L. S. Hung, C. W. Tang, M. G. Mason, P. Raychaudhuri, and J. Madathil,Appl. Phys. Lett. 78, 544 (2001) 54. S. E. Shaheen, G. E. Jabbour, M. M. Morrell, Y. Kawabe, B. Kippelen, N. Peygha mbarian, M. F. Nabor, R. Schlaf, E. A. Mash, and N. R. Armstrong,J. Appl. Phys. 84, 2324 (1998) 55. T. M. Brown, R. H. Friend, I. S. Millard, D. J. Lacey, J. H. Burroughes, and F. Cacialli, Appl. Phys. Lett. 79, 174 (2001) 56. H. Kim, C. M. Gilmore, A. Pique, J. S. Horwitz, H. Mattoussi, H. Murata, Z. H. Kafafi, and D. B. Chrisey, J. Appl. Phys. 86, 6451 (1999) 57. T. Maruyama and K. Fukui, Thin Solid Films 203, 297 (1991) 58. W. F. Wu, B. -S. Chiou, and S. T. Hsieh, Semicond. Sci. Technol. 9, 1242 (1994) 59. P. Nath, R. F Bunshah, B. M. Basol, and O. M. Staffsud, Thin Solid Films 72, 463 (1980) 60. V. Vasu and A. Subrahmanyam, Thin Solid Films 193/194, 696 (1990) 61. J. P. Zheng and H. S. Kwok, Appl. Phys. Lett. 63, 1 (1993) 62. F. Li, H. Tang, J. Shinar, O. Resto, and S. Z. Weisz Appl. Phys. Lett. 70,2741 (1997)

PAGE 98

98 63. C. C. Wu, C. I. Wu, J. C. Strum, and A. Kahn, Appl. Phys. Lett. 70, 1348 (1997) 64. I.M. Chan, W. C. Cheng, and F. C. Hong, Appl. Phys. Lett. 80, 13 (2002). 65. S. K. So, W. K. Choi, C. H. Cheng, L. M. Leung, and C. F. Wong, A ppl. Phys. A 68, 447 (1999) 66. S. F. J. Appleyard and M. R. Willis, Opt. Mater. 9, 120 (1998) 67. E. W. Forsythe, M. A. Abkowitz, Yongli Gao, and C. W. Tang, J. Vac. Sci. Technol. A 18, 1869 (2000) 68. S. M. Tadayyon, H. M. Grandin, K. Griffiths, P. R. Norton, H. Aziz and Z. D. Popovic, Org Elect, Vol 5, Iss 4,157 (2004) 69. S. C. Tse, S. W. Tsang, and S. K. So, J. Appl. Phys. 100, 063708 (2006) 70. G. G. Malliaras and J. C. Scott, J. Appl. Phys. 85, 7246 (1999) 71. K.R. Choudhury, J. Lee, N. Chopra, A. Gupta, X. Jiang, F. Amy, F. So, Adv Func. Mat. 19, 3, 491 (2009) 72. I-Min Chan, Tsung -Yi Hsu, and Franklin C. Hong, Appl. Phys. Lett. 81, 1899 (2002) 73. J Li, M Yahiro, K Ishida, H Yamada, K Matsushige Synthetic Metals, 151,2,141 (2005) 74. J. X. Tang, Y. Q. Li, L. R. Zheng, and L S. Hung, J. Appl. Phys. 95, 4397 (2004) 75. L. S. Hung, L. R. Zheng, and M. G. Mason, Appl. Phys. Lett. 78, 673 (2001) 76. S. Tokito, K. Noda, Y. Taga, J. Phys. D: Appl. Phys. 29 2750 (1996) 77. K.J. Reynolds, J.A. Barker, N.C. Greenham, R.H. Friend, G.L. Frey, J Appl. Phys. 92 7556 (2002) 78. T. Matsushima, Y. Kinoshita, H. Murata, Appl. Phys. Lett. 91 253504 (2007) 79. F Wang, X Qiao, T Xiong, D Ma, Organic Electronics, 9, 6, 985 (2008) 80. I. D. Parker, J. Appl. Phys. 75, 1656 (1994) 81. S. M. Sze, Physics of Semicondu ctor Devices, Wiley, New York, (1981) 82. V. I. Arkhipov, E. V. Emelianova, Y. H. Tak, and H. Bassler, J. Appl. Phys. 84, 848 (1998) 83. M.A. Baldo, S. R. Forrest, Phys. Rev B 64, 8, (2001) 84. U. Wolf, V. I. Arkhipov, and H. Baessler, Phys. Rev. B 59, 7507 (1999)

PAGE 99

99 85. I. G. Hill, A. J. Makinen, and Z. H. Kafafi, Appl. Phys. Lett. 77,1825 (2000) 86. G. Garcia Belmonte, J. Bisquert, P.R. Bueno, C.F.O. Graeff, Chemical Physics Letters 455, 242 (2008) 87. G Garcia -Belmonte, J Bisquert, P R. Bueno, C.F.O. Graeff, F.A. Cas tro, Synthetic Metals, 159, 5, 480 (2009) 88. J. Bisquert, G. Garcia Belmonte, A. Pitarch, H.J. Bolink, Chemical Physics Letters 422, 184 (2006) 89. X. Zhou, J. He, L. S. Liao, M. Lu, Z. H. Xiong, X. M. Ding, X. Y. Hou, F. G. Tao, C. E. Zhou, and S. T. Lee, App l. Phys. Lett. 74, 609 (1999) 90. K Tamano, D Cho, T Mori, T Mizutani, M Sugiyama, Thin Solid Films, 438 439, 182 (2003) 91. M.G. Helander, Z. B, Wang, Z. H. Lu,. Organic Light Emitting Materials and Devices XII, Proceedings of the SPIE, 7051, 70510Z (2008) 92. C-W Chen, P Y Hsieh, H H Chiang, C -L Lin, H -M Wu, and C C Wu, Appl. Phys. Lett. 83, 5127 (2003) 93. Y -K Kim, J W Kim, and Y Park, Appl Phys. Lett. 94, 063305 (2009) 94. H Lee, S W Cho, K Han, P E Jeon, C N Whang, K Jeong, K Cho, and Y Yi, Appl. Phys. Lett. 93, 043308 (2008) 95. Y Yi, P E Jeon H Lee K Han S Kim and S W Cho J. Chem. Phys. 130, 094704 (2009) 96. D. Y. Kondakov, J. R. Sandifer, C. W. Tang, and R. H. Young, J. Appl. Phys. 93, 1108 (2003) 97. D. Y. Kondakov, J. Appl. Phys. 104, 084520 (2008) 98. C. Adachi, K. Nagai, and N. Tamoto, Appl. Phys. Lett. 66, 2679 (1995) 99. J. C. Scott, J. H. Kaufman, P. J. Brock, R. DiPietro, J. Salem, and J. A. Goitia, J. Appl. Phys. 79, 2745 (1996) 100. M. P. de Jong, L. J. van IJzendoorn, and M. J. A. de Voigt, Appl. Phys. Lett. 77, 2255 (2000) 101. V. Shrotriya and Y. Yang, J. Appl. Phys. 97, 054504 (2005) 102. J Haber, E Lalik, Catalysis Today, 33, 1 3, 119 (1997)

PAGE 100

100 BIOGRAPHICAL SKETCH Mikail Shaikh was born to Asif and Zarin Shaikh in Pune, India in 1985, where he was succeeded by his younger brother Saif. He has resided in Pune from birth to age 22. He attended St. Vincents high school and Junior College for 12 years and graduated in 2003. He attended the Vishwakarma Institute of Technology (VIT), an engineering college under the University of Pune for his undergraduate education. He undertook an internship at VIT and at the National Che mical Laboratory, India in his final year of engineering. He graduated in August 2007 with a degree in chemical engineering. Mikail immediately began his graduate studies in the University of Florida in Fall 2007, in the Department of Material Science and Engineering, where he specialized in electronic materials. Under the guidance of his advisor Dr Franky So, he gained a deeper understanding of the nascent field of organic electronics, and performed research in small molecule organic light emitting diodes. Upon completion of his masters degree, Mikail looks to pursue his interest in organic electronics in an industrial or academic setting