Structural and Optical Enhancement of Organic Light Emitting Diodes

MISSING IMAGE

Material Information

Title:
Structural and Optical Enhancement of Organic Light Emitting Diodes
Physical Description:
1 online resource (136 p.)
Language:
english
Creator:
Wrzesniewski, Edward Joseph, III
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Materials Science and Engineering
Committee Chair:
Xue, Jiangeng
Committee Members:
Singh, Rajiv K
So, Franky
Douglas, Elliot P
Jiang, Peng

Subjects

Subjects / Keywords:
microlens -- oled -- outcoupling
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre:
Materials Science and Engineering thesis, Ph.D.
Electronic Thesis or Dissertation
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )

Notes

Abstract:
Cost effective,highly efficient, and color accurate lighting sources are quickly reaching high demand.  One possible method of achieving these disparate goals is through organic light emitting devices (OLED).  In this work, methods to improve the luminous power and external quantum efficiency of OLEDs were investigated.  Specifically, alternative device structures and methods that would alter the path of light through a device in order to extract more light to air were designed.  First, a transparent electrode consisting of an oxide / metal / oxide structure was considered for use in top-emitting OLED (TOLED) architectures to replace the commonly used transparent electrode, indium tin oxide (ITO), which has limited processability in TOLEDs.  The electrical and optical characteristics of electrodes consisting of molybdenum oxide (MoO3) / gold or silver / MoO3 were shown to have low sheet resistance (<10 Ohm/sq) Next, molded, UV cured, transparent, hemispherical microlens arrays were applied to TOLEDs using a soft lithography process in order to reduce internal reflection at the device air interface.  Previously, bottom-emitting devices with applied microlens arrays show a 50-70% enhancement in efficiency.  By eliminating an internal interface that causes waveguided mode formation, a significantly higher enhancement of 2.6fold can be seen when applying arrays to the TOLED architecture. The current enhancement has been shown to only be limited by the index of refraction of the lens material. Finally an alternative device structure of blue green OLEDs showing nearly 100% internal quantum efficiency was examined.  Through the use of multiple pathways for charge transportation, exciton confinement and abundant sites for radiative recombination,the device efficiency can be maximized. The potential of this structure when combined with down conversion microlens arrays for high efficiency white OLEDs was also examined.  While leading to a CRI value of over 80, the limited quantum yield of the emitters used limits the viability of this technique.
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.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Xue, Jiangeng.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-12-31
Statement of Responsibility:
by Edward Joseph Wrzesniewski III.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044661:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 STRUCTURAL AND OPTICAL ENHANCEMENT OF ORGANIC LIGHT EMITTING DIODES By EDWARD JOSEPH WRZESNIEWSKI III A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Edward Joseph Wrzesniewski III

PAGE 3

3 To my supporters

PAGE 4

4 ACKNOWLEDGMENTS My research and my development as a scientist are not simply the culmination of my efforts but also directly influenced by many people in my life. Dr. Jiangeng Xue, my PhD advisor, whose high standards and critical attention to detail, taught me that to be a researcher, one must continue to question assumptions and always dig deeper to form a more compl ete understanding of a problem. Without his support or that of my group members I could not have accomplished this task. I must thank my ear ly mentor and collaborator Sang Hyun Eom, whose guidance allowed me to transition from undergraduate novice to ski lled graduate student and whose example of graduate success all while starting a family is something any researcher can commend. Additionally, Jason M yers and Bill Hammond for their support in the lab and for the reminder that even the most brilliant re searcher will still end up cov ered in vacuum oil now and then; Weiran Cao for his insightful discussions and collaborations; and John Mudrick my brother in arms. Ying Zheng, Yixing Yang, Renjia Zhou, Matt Rippe, Nate Shewmon, Sangjun Lee, and Justin Denn ison have all certainly impacted my journey and I wish them much success in their own research efforts. I also must acknowledge my other advisory committee members Drs. Franky So, Elliot Douglas, Rajiv Singh, and Peng Jiang for their time and efforts. Fi nancially, I must thank the US Department of Energy, National Science Foundation, and the UF Alumni Fellowship for allowing me to conduct my research. While the lab is where research is conducted, it is shaped by the world at large and those in it. For th is reason, I must thank Dr. Barron Humphries for his advice and support and helping me to adapt to life in Gainesville; Chris, Sarah, and Jonah Cook for being great friends and providing much needed stress relief; Drs. Danny and Jennifer

PAGE 5

5 Coenen for remindi ng me that even academics need to let their hair down once and while; Joe Simmons for expanding my horizons; and Chris Glass and Doug Foster for our weekly therapy sessions. My friends, Glenn Bean, Hunter Henderson, Jon Webb, Julie Hill, Amanda Samerson, Meagan Harris, Felipe Martinez, Alex Blandina, Katie thanks for keeping me sane. Without my parents, Ed and Laura Wrzesniewski, whose years of tireless support and enthusiasm h ave allowed me to weather all storms, I could not have reached this point and I am forever indebted to them. Charlie Wrzesniewski has shown me the direct path is not always the best path and that sometimes getting lost is exactly where one need s to be. C arolyn Smith has been an inspiration. Her optimism and support brought me to this point and I am a better person because of her My CMU friends, Jason Glasser, Dan Frank, Adam Aaron, and Aaron Johnson, through their example, challenge me t o reach my own goals and remind me of my modest beginnings. To all those who have made an impact on my life, either by removing hurdles in front of me or perhaps laying some of their own, I could not be the person I am today without your involvement.

PAGE 6

6 TABLE OF CONTEN TS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 13 ABSTRACT ................................ ................................ ................................ ................... 17 CHAPTER 1 INTRODUCTION TO ORGANIC ELECTRONICS ................................ .................. 19 1.1 The Need for High Efficiency/Cost Effective Lighting Sources .......................... 1 9 1.2. Organic Electronics as a Viable Lighting Replacement ................................ .... 22 1.2.1 Advantages of Organic Technologies ................................ ...................... 22 1.2.2 Potential applications for Organic Light Sources ................................ ..... 22 1.2.2.1 Display technology ................................ ................................ ......... 23 1.2.2.2 Commercial and residential lighting ................................ ............... 24 1.3 Current State of Organic LEDs ................................ ................................ ...... 24 2 FUNDAMENTAL PHYSICS OF ORGANIC SEMICONDUCTORS ......................... 27 2.1 Introduction ................................ ................................ ................................ ....... 27 2.2 Electronic Structure and Bonding ................................ ................................ ..... 27 2.3 Charge Transport Models ................................ ................................ ................. 28 2.3.1 Band Structures ................................ ................................ ....................... 29 2.3.2 Amorphous Material Charge Transport ................................ ................... 30 2.3.3 Electronic versus Optical Gaps ................................ ............................... 33 2.4 Excitons ................................ ................................ ................................ ............ 34 2.4.1 Formation and Types ................................ ................................ ............... 34 2.4.2 Singlets and Triplets ................................ ................................ ................ 35 2.4.3 Heavy Metal Complexes ................................ ................................ .......... 36 2.5 Energetic Systems and Interactions ................................ ................................ .. 37 2.5.1 Intramolecular Processes ................................ ................................ ........ 37 2.5.1.1 Energy absorbtion ................................ ................................ .......... 38 2.5.1.2 Fluorescent decay ................................ ................................ .......... 38 2.5 .1.3 Intersystem crossing ................................ ................................ ...... 39 2.5.1.4 Phosphorescent decay ................................ ................................ ... 39 2.5.1.5 Vibrational shifts ................................ ................................ ............. 39 2.5.2 Intermolecular Processes ................................ ................................ ........ 41 rster energy transfer ................................ ................................ ... 41 2.5.2.2 Dexter energy transfer ................................ ................................ ... 42

PAGE 7

7 3 LIGHT EMITTING ELECTRONIC DEVICES BASED ON ORGANIC SEMICONDUCTORS ................................ ................................ ............................. 43 3.1 Fabrication Methods for Organic Electronics ................................ .................... 43 3.2 OLED Device Architecture ................................ ................................ ................ 45 3.2.1 Operation Principles ................................ ................................ ................ 45 3.2.2 The Multilayer Stack ................................ ................................ ................ 47 3.2.2.1 Electrodes ................................ ................................ ...................... 47 3.2.2.2 Charge injection layers ................................ ................................ ... 48 3.2.2.3 Charge transport layers ................................ ................................ .. 48 3.2.2.4 Emissive l ayer ................................ ................................ ................ 49 3.3 Measurement of OLEDs ................................ ................................ ................... 50 3.3.1 Colorimetry ................................ ................................ .............................. 50 3.3.1.1 Responsivity of the human eye ................................ ...................... 52 3.3.1.2 Angular emission ................................ ................................ ............ 53 3.3.1.3 The ideal Lambertian emitter ................................ ......................... 54 3.3.1.4 CIE coordinates and color rendering index ................................ .... 55 3.3.2 Device Characterization ................................ ................................ .......... 57 3.3. 2.1 Current voltage measurement ................................ ....................... 57 3.3.2.2 Luminance measurement and calibration ................................ ...... 58 3.3.2.3 Emission spectra ................................ ................................ ............ 58 3.3.2.4 Luminance efficiency ................................ ................................ ...... 59 3.3.2.5 Luminous power efficiency ................................ ............................. 59 3.3.2.6 Ext ernal Quantum Efficiency ................................ .......................... 61 4 OLED OPTICAL PATHS AND MEANS OF ENHANCEMENT ................................ 63 4.1 Light Path in Basic Bottom Emitting OLED Archit ecture ................................ ... 63 4.2 Fabrication Methods to Alter Optical Modes ................................ ..................... 65 4.2.1 Cathode Surface Plasmons ................................ ................................ ..... 65 4.2.2 High Index Glass ................................ ................................ ..................... 67 4.2.2 Low Index Grids ................................ ................................ ...................... 67 4.2.3 Quarter Wave Stacks (Bragg Reflector) ................................ .................. 68 4.2.4 Microlens Arrays ................................ ................................ ...................... 69 5 TRANSPARENT OXIDE/METAL/OXIDE TRILAYER ELECTRODE FOR USE IN TOP EMITTING ORGANIC LIGHT EMITTING DIOD ES ................................ ........ 71 5.1 Introduction ................................ ................................ ................................ ....... 71 5.2 Experimental Methods ................................ ................................ ...................... 73 5.3 Optoel ectronic properties of multilayer structures ................................ ............. 76 5.4 OLED devices with trilayer transparent electrodes ................................ ........... 79 5.5 Summary ................................ ................................ ................................ ......... 82 6 ENHANCING LIGHT EXTRACTION IN TOP EMITTING ORGANIC LIGHT EMITTING DEVICES USING MOLDED TRANSPARENT POLYMER MICROLENS ARRAYS ................................ ................................ ........................... 83

PAGE 8

8 6.1 In troduction ................................ ................................ ................................ ....... 83 6.2 Experimental Methods ................................ ................................ ...................... 83 6.3 Comparison of Bottom and Top Emitting OLEDs ................................ .............. 89 6.4 Simulation of Microlens Enhancement on Top OLEDs ................................ ..... 92 6.5 Bottom Emitting Devices with Microlens Arrays ................................ ................ 93 6.5 Large Area TOLEDs with MLA ................................ ................................ .......... 96 6.6 Wavelength and Angular Dependence of MLA ................................ ................. 98 6.7 White OLEDs with Lens Enhancemen t ................................ ........................... 101 6.8 Summary ................................ ................................ ................................ ........ 102 7 ALTERNATIVE OLED ARCHITECTURES FOR IMPROVED EFFICIENCY ........ 104 7.1 Introduction ................................ ................................ ................................ ..... 104 7.2 Experimental Methods ................................ ................................ .................... 104 7.3 Blue/ Green Bottom OLEDs with Near Maximum External Qua ntum Efficiency ................................ ................................ ................................ ........... 105 7.4 Down Conversion Red Microlens Arrays ................................ ........................ 113 7.5 Summary ................................ ................................ ................................ ........ 118 8 CONCLUSIONS ................................ ................................ ................................ ... 119 8.1 Oxide / Metal/ Oxide Transparent Electrodes ................................ ................ 119 8.2 Light Extraction Enhancement in TOLE Ds ................................ ...................... 119 8.3 Alternative OLED Architectures ................................ ................................ ...... 120 8.4 Future Work ................................ ................................ ................................ .... 121 8.4.1 Quantum Dot LEDs ................................ ................................ ............... 121 APPENDIX A LIST OF PUBLICATIONS ................................ ................................ ..................... 123 REFERENCES ................................ ................................ ................................ ............ 124 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 136

PAGE 9

9 LIST OF TABLES Table page 3 1 Comparison of the various terms and units used for photometry and radiometry. ................................ ................................ ................................ .......... 51 4 1 Recorded enhancement of various outcoupling methods over traditional OLEDs. ................................ ................................ ................................ ............... 70

PAGE 10

10 LIST OF FIGURES Figure page 1 1 US annual power consumption for various lighting applications and by lamp type in 2010 ................................ ................................ ................................ ........ 19 1 2 T imeline of lighting technology efficiencies with predicted ef ficiencies of future solid state lighting technologies as dashed lines ................................ ..... 21 1 3 Examples of OLED display technology from Sony: the Xel 1, the first OLED TV and an example of a flexible OLED display ................................ .................. 23 2 1 Examples of organic molec ules with increasing complexity ............................... 27 2 2 Benzene molecule with perpendicular p z orbita ls and their delocalized nature 28 2 3 Evolution of discrete energy levels to energy bands with increased molecular intera ction. ................................ ................................ ................................ .......... 29 2 4 Splitting of energy levels as two molecules begin to interact. ............................. 31 2 5 Energy gap for single molecules, crystals, a nd amorphous solids. ..................... 32 2 6 Distinction between Wannier and Frenkel excitons. ................................ ........... 34 2 7 Singlet and triplet excitons with radiative relaxation. ................................ .......... 36 2 8 Two types of excitons in heavy metal complexes. Type 1 is ligand only interaction while type 2 shows metal ligand charge transfer. .............................. 36 2 9 A Jablonski energy diagram illustrating the various intramolecular processes that can occur. ................................ ................................ ................................ .... 38 2 10 Configurational diagram showing the absorbtion of photonic e nergy and subsequent shifted emission energy ................................ ................................ .. 40 2 11 Intermolecular energy transfer processes ................................ .......................... 41 3 1 Schematic of a thermal evaporation chamber with the significant components labeled. ................................ ................................ ................................ ............... 44 3 2 Diagram of spin coating deposition. ................................ ................................ .... 45 3 3 Basic OLED operation. ................................ ................................ ....................... 46 3 4 The entire electromagnetic spectrum with t he visible spectrum highlighted ..... 50 3 5 The normalized scotopic and photopic curves ................................ ................... 53

PAGE 11

11 3 6 Comparison of the planar angle to solid angle ................................ ................... 54 3 7 Two dimensional representation of a uniform Lambertian emission source ..... 54 3 8 The tristimulus curves used to determine CIE coordinates ................................ 55 3 9 Two dimensional CIE coordinate plane with black line indicating the Planckian locus ................................ ................................ ................................ .. 57 3 10 A two dimensional Lambertian source with light emitting area A and surface area dS to define the solid angle ................................ ................................ ........ 60 4 1 Optical mo des in bottom emitting device. ................................ ........................... 64 4 2 Comparison of planar and surface plasmon enhanced device structures ......... 66 4 3 Light path through an OLED with incorporated low index grid. ........................... 68 4 4 Demonstration of light p ath through a Bragg reflector. ................................ ....... 68 5 1 Schematic device structures for the standard bottom emitting architecture based on the commercial ITO electrode and the top emitting architecture based on the MoO 3 /metal/MoO 3 trilayer transparent electrode. ......................... 74 5 2 Transmittance of the ITO and MoO 3 /Au/MoO 3 electrode with various Au layer thickness. ................................ ................................ ................................ ........... 76 5 3 Sheet resistance and transmittance at 600nm as a function of Au layer thickness. ................................ ................................ ................................ ........... 77 5 4 Wavelength dependant transmittance of mul tilayer structures with Au, Au/Ag, and Ag intermediate layers. ................................ ................................ .... 79 5 5 Device Properties of Bottom and Top Emitting Devices ................................ .... 80 5 6 High efficiency p i n green phosphorescent devices. ................................ ......... 81 6 1 Schematic o f microlens fabrication process ................................ ...................... 86 6 2 Silicon wafer w ith polystyrene bead mono layer ................................ ................. 87 6 3 Scanning Electron microscope images o f microlens formation process ............. 88 6 4 Bottom and top emitting OLEDs. ................................ ................................ ...... 91 6 5 Normalized EQE and enhancement factor of BOL EDs with added microlens arrays ................................ ................................ ................................ ................ 94 6 6 Microlens enhancement of large area OLEDs ................................ .................... 95

PAGE 12

12 6 7 Effect of a microlens array on light extraction in top emitting green OLED ........ 96 6 8 Wavelength dependence of the light extraction enhancement induced by the microlens array. ................................ ................................ ................................ .. 98 6 9 Performance of a top emitting white OLED with a microlens array. .................. 101 7 1 Blue g reen d evice features. ................................ ................................ ............. 106 7 2 Ext ernal quantum efficiencies of devices with varied hole and electron transport materials ................................ ................................ ............................ 107 7 3 Comparison of host materials CBP and mCP for use in green OLEDs based on external quantum and p ower efficiency. ................................ ...................... 108 7 4 Comparison of green (G), blue (B) and blue green devices with various doping concentrations ................................ ................................ ...................... 109 7 5 Blue green device structures showing both separation of blue dopant from ETL and removal of the mCP exciton blocking layer ................................ ........ 110 7 7 Current injection of equally doped red (PQIr), green (Ir(pp y) 3 ), and blue (FIr6) OLEDs. ................................ ................................ ................................ ............. 113 7 8 Comparison of the emission spectra of the B G device, and with either rubrene or PQIr doped microlens array ................................ ............................ 115 7 9 Comparison of white light emission spectra with rubrene and DCJTB doping .. 116 7 10 Large area blue green OLED luminance and luminance with attached DCJTB lens array ................................ ................................ ................................ .......... 118

PAGE 13

13 LIST OF ABBREVIATION S Ag silver Al aluminum AMOLED active matrix organic light emitting device Au gold BCP bathocuproine BPhen b atho phenanthroline CB conduction band CBP 4,4' bis(carbazol 9 yl)biphenyl CCT correlated color temperature CIE C ommission Cs Cesium CsCO 3 Cesium Carbonate CRI color rendering index CRT cathode ray tube CVD chemical vapor depo sition EIL e lectron injection layer EL Electroluminescent EML emissive layer ETL electron transporting layer FIr6 iridium(III) bis(4 ,6 difluorophenylpyridinato)tetrakis(1 pyrazolyl)borate FOLED fluorescent organic light emitting device F 4 TCNQ tetraf luo ro tetracyanoquinodimethane HIL h ole injection layer HOMO highest occupied molecular orbital

PAGE 14

14 HTL hole transporting layer IC internal conversion Ir(ppy) 3 fac tris (phenylpyridine) iridium ISC intersystem crossing ITO i ndium tin oxide LCD liquid crystal dis play LUMO lowest uno ccupied molecular orbital MBE molecular beam epitaxy mCP dicarbazolyl 3,5 benzene MeO TPD N,N' diphenyl N,N' bis(3 methylphenyl ) [1,1' biphenyl] 4,4' diamine MLCT metal ligand charge transfer MoO 3 Molybdenum Trioxide NPB N, N' bis (naphthalen 1 yl) N,N' bis(phenyl) benzidine OLED organic light emitting device OMBD Organic Molecular Beam Deposition PDP plasma display panel PHOLED phosphorescent organic light emitting device PL photoluminescent PLD pulsed laser deposition PLED polymer based OLED P LQY photoluminescen ce quantum yield PQIr iridium(III) bis (2 phenylquinolyl N,C 2 ) acetylacetonate S EML single emissive layer SMOLED small molecule based OLED SSL solid state lighting

PAGE 15

15 TAPC 1,1 bis (di 4 tolylaminophenyl)cyclohexane TCO transp arent conducting oxide UGH2 p bis(triphenylsilyly)benzene VB valence band VDW van der Waals VTE vacuum thermal evaporation WOLED white organic light emitting device 3TPYMB tris[3 (3 pyridyl)mesityl]borane conversion factor E f f ermi energy level f geometr ic factor G( ) photopic response G ( ) scotopic response I det photocurrent I D OLED device current J cu rrent density L luminance solid angle S 0 ground state S 1 excited singlet state S( ) spectrum T 1 excited triplet state mobility V voltage V 2 O 5 vanadium oxide

PAGE 16

16 current efficiency (or luminance efficiency) luminous power efficiency internal quantum efficiency external quantum efficiency ou tcoupling efficiency

PAGE 17

17 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy STRUCTURAL AND OPTICAL ENHANCEMENT OF ORGANI C LIGHT EMITTING DIODES By Edward Joseph Wrzesniewski III August 2012 Chair: Jiangeng Xue Major: Materials Science and Engineering Cost effective, highly efficient, and color accurate lighting sources are quickly reaching high demand. One possible met hod of achieving these disparate goals is through organic light emitting devices (OLED). In this work, methods to improve the luminous power and external quantum efficiency of OLEDs were investigated. Specifically, alternative device structures and meth ods that would alter the path of light through a device in order to extract more light to air were designed First, a transparent electrode consisting of an oxide / metal / oxide structure was considered for use in top emitting OLED (TOLED) archi tectures t o replace the commonly used transparent electrode, indium tin oxide (ITO), which has limited processability in TOLEDs The electrical and optical characteristics of electrodes consisting of molybdenum oxide (MoO 3 ) / gold or silver / MoO 3 were shown to h ave low sheet resistance (<10 ) using thin 10 nm thick metal layers Additionally, the transparency of these electrodes has been shown to reach 90% for some wavelengths depending on the metal used When used in a TOLED, the device behavior was sh own to be comparable to bottom emitting counterparts with only 20% reduction in efficiency due to charge injection issues.

PAGE 18

18 Next, molded UV cured transparent, hemispherical microlens arrays were applied to TOLEDs using a soft lithography process in order to reduce internal reflection at the device air interface. Previously, bottom emitting devices with applied microlens arrays show a 50 70% enhancement in efficiency. By eliminating an internal interface that causes waveguided mode formation a significan tly higher enhancement of 2.6 fold can be seen when applying arrays to the TOLED architecture. The current enhancement has been shown to only be limited by the index of refraction of the lens material. Finally an alternative device structure of blue green OLEDs showing nearly 100% internal quantum efficiency was examined. Through the use of multiple pathways for charge transportation, exciton confinement and abundant sites for radiative recombination, the device efficiency can be maximized. The potential of this structure when combined with down conversion microlens arrays for high efficiency white OLEDs was also examined. While leading to a CRI value of over 80, the limited quantum yield of the emitters used limits the viability of this technique.

PAGE 19

19 CHAP TER 1 INTRODUCTION TO ORGA NIC ELECTRONICS 1.1 The N eed for H igh E fficiency / C ost E ffective L ighting S ources In a world of declining resources and growing demand for said resources; new, more efficient methods of utilization are necessary. This statement i s no more significant than in the realm of energy generation and consumption, where a growing world population dependent on electricity and electronic consumer devices has put an ever growing strain on electrical infrastructure. While power storage techno logies such as fuel cells and lithium ion batteries continue to advance in capacity [1,2] a balanced approach of also reducing the power consumption of existing products is prudent. Portable electronics, large screen displays, and a growing desire fo r ultrathin compact technologies all require high efficiency components to meet power, form and lifetime demands. One of the most significant contributors to the growing demand for power is something often taken for granted, namely lighting. Figure 1 1 i llustrates this point rather succinctly showing nearly 700 terawatt hours per year simply on lighting [3] Figure 1 1 US annual power c onsumption for various lighting applications and by lamp type in 2010 [3]

PAGE 20

20 Humans in particular rely on light more heavily than most organi sms, but for most of human history light sources have been limited to the sun or combustion of various sources of organic matter. With the advent of incandescent light bulbs, light could be generated on a whim and could last hours into the night. Lighti ng has progressed over the years, to increase brightness, and duration, and more recently to reduce power consumption. The result has been a virtually endless daytime in the industrialized world with cities such as Tokyo and New York ever illuminated, wit h dynamic glowing billboards presenting an ever changing series of advertisements Lighting contributes nearly 22% of all power consumption in the United States with an estimated 1,000 terawatt hours likely to be consumed in by the year 2025 [4] With a light source even 50% efficient at converting electrical power to optical power, this amount could be reduced by 620 billion kilo watt hours per year or the equivalent output of nearly 70 nuclear power plants [5] Clearly, in order to reduce this burden to the power grid, high ly efficient, low cost methods of lighting are necessary. One area of research that is likely to meet the needs of both the electronics and lighting sectors is solid state lighting (SSL). The promise of SSL stems from the materials used, semiconductors, which can transfer electronic energy directly to visible light. Consisting of direct band gap semiconductors (typically group III nitride compounds), supplied electrons directly emit photons when de leading to less indirect losses found in conventional lighting due to heating of a filament or exciting an intermediate gas. The advantage of SSL compared to these methods is twofold both in terms of efficiency (comparing the poor incandescent power conv ersion efficiency of 5%

PAGE 21

21 to the potential 100% internal quantum efficiency of solid state light emitting devices), as compared to up to 90 for SSL sources) [6] Thus great interest in these light emitting devices has been shown for both high efficiency display and lighting applications. Developments in SSL lighting are compared to other light sources in figure 1 2. Figure 1 2 Timeline of lighting technology efficiencies with predicted efficiencies of future solid state lighting technologies as dashed lines [3] While solid state lighting as a whole is a promising avenue of research, it is not without its limitations. Traditional inorganic semiconductors require costly and time consum ing methods such as chemical vapor deposition, RF sputtering, and pulsed laser deposition to produce commercial scale products [7 9] Additionally, the need for expensive and difficult to maintain clean rooms only increases the difficulty of producing quality products. Furthermore, many of the materials required for inorganic semiconductors are either rare or highly toxic. Cadmium, lead, and tellurium are all commonly used, high performing materials in the industry with distinct drawbacks [10

PAGE 22

22 12] Even precursors such as arsene and silane carry as many health and safety risks as the materials in the devices themselves. 1.2. Organic Electronics as a V iable L ighting R eplaceme nt With these considerations in mind, one promising alternative to these materials and methods is the realm of organic semiconductors which present solutions to some of these issues. 1.2.1 Advantages of O rganic T echnologies Based on highly abundant ca rbon, these small molecules or polymers possess controlled structures dependent entirely on organic chemistry synthesis. As a result, by adjusting the moiety of these molecules, various colors, band gaps, and orbital energy levels can be achieved, leading to various functionalities in a light emitting device. Furthermore, these molecules can be deposited using simpler methods such as vacuum thermal evaporation, or solution processed methods such as inkjet printing [13] ultra sonic spray [14] or roll to roll [15] Additionally these films are far thinner (approximately 100 nm) than their inorganic counterparts leading to reduced material consumption, ult ra thin display elements, and even flexible lighting that can conform to surfaces throughout a structure. The result of these features is the ability to produce scalable, thin technologies applicable to both displays as well as lighting applications. 1. 2.2 Potential applications for O rganic Light S ources Organic LEDs have a wide variety of applications, some similar to their inorganic counterparts while others are entirely novel. The two primary applications come in the form of flat panel displays, as w ell as white light sources in order to replace conventional fluorescent tubes and incandescent bulbs.

PAGE 23

23 1.2.2.1 Display t echnology The most developed application for OLEDs is for use in active matrix displays. Display technology has progressed from the us e of cathode ray tubes to current competitors plasma display panels (PDP) and liquid crystal displays (LCD). While currently LCD devices dominate both the mobile and large screen markets, this technology has some inherent limitations. The viewing angle o f LCD screens has been notoriously limited and its slow refresh rate limits the capability of these screens to generate realistic motion or splitting 3D images. OLEDs by contrast have nearly 180 viewing angle and micro to nanosecond radiative life times. Furthermore, by actively generating the light in appropriate intensities rather than filtering intense white light as the liquid crystals do, power consumption is reduced. Companies like Samsung and LG have already begun to produce small scale displays for mobile devices and are increasingly scaling these technologies up for next generation televisions. The thin organic nature of OLEDs also enables new concepts such as flexible displays, able to better withstand the wear and tear mobile dev ices receiv e as can be seen. Figure 1 3 Examples of OLED display technology from Sony: the Xel 1, the first OLED TV and an example of a flexible OLED display [16]

PAGE 24

24 1.2.2.2 Commercial and r esidential l ighting One area of increased interest for OLED development is the realm of commercial and residential lighting. The low power consumption and high color fidelity of OLED d evices as well as the cus tomizabi lity of shape and color lead present the next great challenge in terms of commercialization. While the current state of OLED efficiency may be enough to create display products, work must continue in order to rival fluorescent and other high effic iency lighting technologies. The slim form factor and low cost of production make this technology an attractive predecesso r to current lighting sources. 1.3 Current State of Organic LEDs Historically, the advent of organic light emitting devices came in 1987 when Tang and Van Slyke at Kodak discovered light emission when a current was applied to organic molecules that were being tested for use in organic photovoltaics [17] From this followed the convention of using separate materials for electron and hole transport. This lead to early external quantum efficiencies of nearly 1% (photons out/electrons in) by shifting the recombination zone away from the plasmon quenching cathode as well balancing charges injected for more efficient recombination By 1989, further improvement to the small molecule device was made through the addition of efficient fluorescent dyes which when doped into the e missive layer allowed energy transfer from the host material into dye and allowing for a greater rate of radiative recombination without the loss of the electronic properties [18] Finally, the cathode itself has made the transition from Mg which of course is highly reactive when exposed to air, to Mg/Ag a more stable albeit expensive alternative to finally the introduction of a LiF interl ayer

PAGE 25

25 that allows adequate electron injection even from metals such as Al while simultaneously reducing the operating voltage of these devices [19] To this point, the primary method of light emission relied on fluorescent emission within the active layer. Unfortunately, singlet transmissions that lead to fluorescent emission only account for 25% of the total formed excitons in organic m olecules. The other 75% of excitons were forbidden to radiatively relax due to spin conservation. To further develop OLED technology and overcome this 25% internal efficiency limitation, Stephen Forrest of Princeton University and Mark E. Thompson of the University of Southern California developed emissive molecules that capitalized on the spin orbit coupling between heavy metals and organic ligands [20] The result was that phosphorescent emission from triplet states in these molecules could now be accessed by doping the emissive layer with these new dyes. From this development, internal quantum efficiencies of nearly 100% can be achiev ed [21] While highly efficient green devices could be achieved, wide band gap blue OLEDs still lacked high efficiency. Few potential deep blue emitting dopants had been developed [22] and those that had been still required better exciton confinement, wide band gap host material, charge injection, and charge confinement. In order to combat these issues, high triplet energy hole transport [23] and electro n transport materials [24] were used to better confine excitons, a nd a double emitting regi on were used to force recombination on dopant mo lecules [25] Furthermore, charge injection was improved through the use of p i n device structures [25] which add doped p and n type transporting layers in order to reduce the driving voltage of the device.

PAGE 26

26 As the efficiency of the blue emitter has increased, so too has the possibility of producing a white light OLED. Various techniques have been applied to produce white light broad emission such as blended single emitting layer devices with multiple color dopants [26] single dopant multiple emitting layers [27] or even tandem OLE Ds with optimized device structures for each of the colors emitted [28] By applying these methods, some laboratory devices have already begun to match fluorescent lighting sources in terms of efficiency [29] Unfortunately, most methods to produce these highly efficient devices require expensive or difficult to fabricate optical enhancement mechanisms to overcome the light extraction limitations inherent in OLED design. The focus of the current OLED research push is to find better more effective means of extracting the light generated in these devices [26,29 38] reduce efficiency roll off at high intensities, and improve device stability under operating hours. Much work is also conducted on transferring laboratory scale processes to large scale manufacturi ng facilities with high throughput methods of device creation. Where OLEDs will be in the next 10 years remains to be seen, but with recent advances, it is likely that this technology may be found in any home.

PAGE 27

27 CHAPTER 2 FUNDAMENTAL PHYSICS OF ORGANIC S EMICONDUCTORS 2.1 Introduction In spite of current usages of the word, organic compounds are those consisting primarily of carbon molecules, typically conjugated or aromatic hydrocarbons. Due to the up to four bonds each carbon atom can have, the arrangem ent and iterations of these molecules is nearly endless. The size and complexity can range from the most basic, methane (CH 4 ) to long chain molecules such as polyethylene, to even further complex biological proteins like spider silk [39] Figure 2 1 illustrates this transition in complexity. Figure 2 1 Examples of organic molecules with increasing complexity A ) M ethane B ) P olyethyl ene C ) M yoglobin protein [40 42] 2.2 Electronic Structure and Bonding Solid materials are typically the product of four major bond types: covalent, metallic, ionic, and van der Waals (VDW). While typical materials consist of one bond type, organic solids form from covalently bonded molecules that are linked through VDW forces. This uniqu e interaction would suggest a variety of properties not seen in most other materials [39] While the mechanical and biocompatible prope rties of these materials are certainly impressive areas of study, this chapter will focus specifically on the electronic and photo physical properties these materials may possess.

PAGE 28

28 While most organic compounds are considered electrical insulators, one sub set of note are organic semiconductors. These compounds are characterized by their strong conjugated covalent bonds which form sp 2 hybridized orbitals ( bond s ) that exist in plane with the organic molecule Perpendicular to the molecule, alternatively are the p z orbitals The overlapping of neighboring p z bonds. It is hopping between these conjugated structures electrical conductivity is allowed. Figure 2 2 shows a typical organic semiconducting molecul e with its orbitals displayed. Fig ure 2 2 Benzene molecule with perpendicular p z orbitals and their delocalized nature [43] 2.3 Charge Transport Models While describing the motion of charges using molecular diagrams is perhaps the most accurate method, more succinct energy level diagrams are of value when describing the semiconducting properties and charge transport of these materials. The following sections will illustr ate this method.

PAGE 29

29 2.3.1 Band Structures The methods for describing semiconductors are based upon traditional inorganic semiconducting materials and therefore it of benefit to explain this structure first. Inorganic semiconductors rely on a single bonding m ethod, most commonly covalent although some ionically bonded oxides are also semiconductors. As single molecules and even small numbers of bonded atoms, electrons are only capable of occupying discrete energy levels as described by the Pauli exclusion pri nciple. As more atoms become bonded together to form a solid material the close proximity of electrons causes the energy levels of these electrons to shift in order to perpetuate single occupant levels. Once enough of a crystalline solid material is form ed, the difference between discrete energy levels becomes so minute that in effect it becomes a band of acceptable energy levels. Figure 2 3 illustrates this phenomenon [44] Figure 2 3 Evolution of discrete energy levels to energy bands with increased molecular interaction For metallically bonded materials, this band is continuous allowing charges to become completely delocalized an d able to move easily throughout the material. For this reason, metals are electrically conducting materials. In semiconductors, the number of acceptable states does not form a completely continuous band. Typically a

PAGE 30

30 gap exists between the two bands, on e filled with electrons (the valence band) and the other unoccupied acceptable states for electrons to occupy (conduction band) This band gap, and the amount of energy necessary to excite and electron over the gap, form the basis for semiconducting techn ologies, both in controlling the flow of current and in the interaction of this gap with both the absorption and emission of light. The common terminology used for charge carriers in semiconducting devices are electrons and holes. Electrons, of course, ar e the negatively charged components of atoms and molecules. When these electrons are excited from the valence band into the conduction band, a hole remains behind. This hole is the virtual positive charge carrier (as the absence of a negative ch arge in a balanced system would create a net positive charge). Each of these carriers has its own movement through their respective bands allowing for the flow of electrical current. 2.3.2 Amorphous Material Charge Transport While inorganic semiconductors are known for their valence and conduction bands, the lack of a crystalline structure in most organic materials means that these terms do not exactly apply to organic semiconductors. Instead, organic materials are described by their highest occupied molecular orbi tal (HOMO) and lowest unoccupied molecular orbital (LUMO) which loosely correlate to valance and conduction bands respectively. The interactions between the electrons within each organic molecule define the gap between these two states. Figure 2 4 shows the splitting of energy levels as molecules interact [39]

PAGE 31

31 Figure 2 4 Splitting of energy levels as two molecules begin to interact With enough molecules interacting and a large enough population of electrons, these energy again begin to form band like structures similar to those in their inorganic counterparts. The size of the energy gap in this case is also stron gly dependent on the degree of conjugation, allowing larger molecules to have narrower band gaps. Just as inorganic semiconductors rely on electrons and holes for charge carriers, so too do their organic counterparts. When comparing a single molecule in the gas phase to crystalline solids with excited charge carriers the band gap becomes reduced due to polarization energy in the two charges. If the solid is no longer an ordered crystal but rather amorphous, as is the case in organic solids, the polariza tion changes with the local arrangement of molecules leading to a Gaussian distribution of possible states for charges to transport through figure 2 5 illustrates this transition [45]

PAGE 32

32 Figure 2 5 Energy gap for single molecules, crystals, and amorphous solids Due to the disorder in organic solids and the weak VDW forces between molecules, the transport properties of organics are seve ral orders of magnitude lower than their inorganic counterparts which have delocalization of electronic states in a crystalline solid that are created by the covalent bonding. Organic molecules are known to have two types of transport mechanisms based on the amount of order present in films. Highly amorphous films with weakly interactions between molecules cannot provide a complete band structure for transport, and therefore must rely on intermolecular hopping of charges. The typical carrier mobility ( ) of these films is in the range of 10 3 ~10 10 cm 2 /V s. This mobility is primarily dependent on electric field and temperature [46] This relationship can be expressed mathematically as:

PAGE 33

33 with F representing the electric field, T is temperature, the activation energy necessary for intermolecular hopping to occur, k is the Boltzmann constant and is a numerical constant. Due to the shape of some organic small molecules, a high degree of crystallinity can be achieved in solid films allowing for a somewhat delocalized band structure. These films tend to have high charge carrier mobility on the order of 1~10 2 cm 2 /V s as is the case in molecules like pentacene [47] Due to the delocalized nature, the mobility of these films can be represented mathematically only as a function of temperature [48] : 2.3.3 Electronic v ersus Optical Gaps One important distinction to make in organic semiconducting films is the difference between electrical or transport band gaps ( E tr ) and optical band gaps ( E opt ) While delocalization in bands tends to reduce this difference in inorganic semiconductors to only a few meV, and indicates weak exciton binding, the close proximity of charges on the same molecule lead to high exciton binding energy and notable distinct ion between E tr and E opt [39] Expressed mathematically, the relationship between these two terms is: E opt = (E gap E p ) E ex = E tr E ex In this equation, E gap is the actual energy difference between the HOMO and LUMO levels of the molecule, E p is the energy that is lost to polarization, and E ex is exciton binding energy.

PAGE 34

34 2.4 Excitons Based on the previous discussion, the question mu st be answered as to what an exciton actually is. The term exciton refers to a bound electron hole pair of charge carriers held together by coulombic interactions. As a result, it can be considered as charge balanced quasi particle capable of transport t hrough a film. The exciton is critically important to OLEDs and therefore a description of its various properties will be presented in the following sections. 2.4.1 Formation and Types Excitons may be divided into two major types: Wannier exciton that i s loosely bound and typically found in inorganic semiconductors and the Frenkel exciton that is highly bound and most commonly found in organic semiconductors. Figure 2 6 displays the difference between these two forms in solid films [49] Figure 2 6 Distinction between Wannier and Frenkel excitons

PAGE 35

35 The large radius of the Wannier exciton (on the order of 100 ) allows for only weak binding of around 10 meV, whereas by confining the charges to a single molecule or two (with a radius of only 10 ), these Frenkel excitons become tightly bound with energies on the order of 1 eV. One less discussed form of exciton, is the charge transfer exciton (CT) which resides in the intermediary energy range between the two extremes considered in the other excitons These charges can be found a few intermolecular distances apart. 2.4.2 Singlets and Triplets Another distinction to be made between ex citons occurs as the result of the quantum spin states that electrons possess. As an exciton is the combination of two [49] the total wave function of the two electron system must be anti symmetric with the interchange of particles. Based on the spin statistics of electrons in an excited state, the anti symmetric and symmetric wave functions can be expressed as: s = 1 2 s = 1 2 symmetric states ( net spin =1 ) s = 1 2 1 2 a = 1 2 1 2 anti symmetric state ( net spin =0 ) Here, a and s refer to anti sy mmetric and symmetric wave functions respectively and n (where n=1,2 ) refers to the spin function with and referring to the spin state of the electron. Due to this difference in spatial symmetry, two different exciton states, the high energy ant i symmetric singlet state and the lower energy symmetric triplet states, exist. The radiative relaxation of these two states then is markedly different. The high

PAGE 36

36 energy singlet state due to the symmetry of its spins, will quickly relax and emit a photon The radiative lifetime of this process is approximately 1 ns and is referred to as fluorescence. Due to the forbidden nature of charges with the same spin state occupying the same position, triplets have a long relaxation time (on order of 1 ms) and is referred to as phosphorescence. Figure 2 7 illustrates these two processes. Figure 2 7 Singlet and triplet excitons with radiative relaxation 2.4.3 Heavy Metal Complexes Figure 2 8 Two types of excitons in heavy metal complexes. Type 1 is ligand o nly interaction while type 2 shows metal ligand charge transfer.

PAGE 37

37 As the radiative relaxation of triplets in organic molecules is a forbidden process, only of the excitons formed will ever be able to emit light. In order to overcome this highly inefficie nt process organometallic compounds with a central heavy metal atom of iridium, platinum, osmium, and ruthenium surrounded by organic ligands can be used. These compounds can access the triplet state as well as the singlet state due to strong spin orbit coupling which scales to the fourth degree with atomic number (Z 4 ) [50 53] Based on the strength of this coupling, organometallic compounds can either have excitons based entirely on the organic compon ent with long relaxation times that may prevent full radiative recombination of charges or with very strong coupling, the metal ligand charge transfer excitons that interact with the metal core can occur, leading to nearly 100% of all excitons radiatively recombining. Figure 2 8 contrasts these two types of complexes [21] 2.5 Energetic Systems and Interactions Now that the exciton is be tter understood, the energy interactions that can excite an electron from the HOMO level, move the electron to other energetic states and finally return to ground state must be better understood. The follow ing sections detail these photo physical processe s. 2.5.1 Intramolecular Processes At the most basic level, these energetic processes can take place on a single molecule. Using a Jablonsky energy diagram, as is displayed in figure 2 9, one can begin to understand the various processes that can occur. F irst some basic symbols must be defined. The S symbol refers to singlet states with S 0 referring to the ground state and S 1 n referring to excited singlet states while T 1 represents the triplet state

PAGE 38

38 Figure 2 9 A Ja b lonski energy diagram illustrating the various intramolecular processes that can occur. 2.5.1.1 Energy absorbtion Absorbtion is the process by which a photon of light with energy higher than the band gap (hc/ > E gap ) of the material leads to the excitation of a charge into a higher energy state. Since a range wavelengths of light with greater energy than the band gap can excite an electron, the single lines for energy states rather than a broad range is mere ly used for simplicity. Symbolically this is represented as S 0 S 1,2 2.5.1.2 Fluorescent decay Once an exciton is formed, series of vibrational relaxation and internal conversion processes occur in order to reach the more stable lowest energy availabl e singlet state. Once these non radiative transitions have occurred, exciton can further relax to a

PAGE 39

39 ground state through emission of a photon of its own. As previously described, this singlet relaxation is referred to as fluorescence and is represented b y the transition, S 1 S 0 This rapid relaxation only occurs in 25% excitons due to the limited number of singlets available. 2.5.1.3 Intersystem crossing In an effort to reduce the energy of the excited electron in a singlet state to a more stable lowe r energy configuration, many singlet excitons transfer to the triplet. This process is called intersystem crossing (ISC) and is the product of a vibrational coupling between the singlet and triplet states. This non radiative process is symbolized by S 1 T 1 2.5.1.4 Phosphorescent decay Once ISC has occurred, the electron goes through the same internal conversion and vibrational relaxation processes to reach the minimum triplet state. Transfer back to the singlet state is not possible due to the large ga p between the singlet state and that of the triplet and a lack of sufficient thermal energy under ambient conditions. Alternatively, through use of spin orbit coupling, the typically forbidden process of relaxing a triplet exciton can occur. This is phos phorescent radiative relaxation and due to the long lifetime of these excitons, may emit light long after the initial excitation. Without this coupling, most triplet excitons would non radiatively recombine over a long lifetime. 2.5.1.5 Vibrational shif ts As previously expressed, during excitation and relaxation processes, the electrons typically u ndergo vibrational relaxation. This relaxation as opposed to its radiative counterpart is a low energy transition that only release s a phonon or heat instea d of

PAGE 40

40 light. Figure 2 10 is a configurational diagram that demonstrates the path of an excited electron and its relaxation [49] Figure 2 10 Configurational diagram showing the absorbtion of photonic energy and subsequent shifted emi ssion energy [54] As an electron is excited from the minimum energy zero order vibrational mode of the ground state to the excited state the corresponding vibrational mode with the same radius is usually higher in order. From this state, the electron rapidly sheds energy through heat to occupy the zero order mode which is of a larger radius than the starting radius. Due to this change i n radius, when the electron emits a photon to reach the ground state, it transitions to a high order vibrational mode of the ground state before again shedding heat to reach the lowest order mode. This process is referred to as the Fran c k Condon shift or Stokes shift and explains reduction in energy between the light absorbed and the light emitted from a film.

PAGE 41

41 2.5.2 Intermolecular Processes Figure 2 11 Intermolecular energy transfer processes A B ) Dexter transfer Now that energy processes within a molecule are better understood, the non radiative energy transfer between molecules can be discussed. This process is typically divided into two types depending on the range ov er which the transitions occur. The longer range transition at around 100 rster energy transfer [49] Transfers at short distances around 10 are called Dexter transfers. Figure 2 11 demonstrates these mechanisms as they are explained in the following sections. rster energy transfer rster transfer is the intermolecular interaction that occurs as the result of resonant dipole dipole interactions. This rapid energy transfer (< 1ns) happens between two singlet excitons as the overlap between the wavelengths of the photons emitted by the so called donor molecule and wavelengths able to be absorbed by the so

PAGE 42

42 called acceptor molecule [55,56] A rate constant at which this occurs can be expressed mathematically as: Where J is the ove rlap between the emission and absorption spectra of the two molecules, K is the orientation factor, n is the refractive index of the medium through which the energy is transferred, is the radiative lifetime of the donor, r is the d istance (cm) between donor and acceptor and, of course, k D A is the rate constant derived. 2.5.2.2 Dexter energy transfer Dexter transfer due to the proximity of the molecules is the actual transfer of electrons between molecules. This process require s the wave functions to overlap between the donor and acceptor molecules. This process most commonly occurs in triplet to triplet transitions [57] The rate constant for this process ( k ET ) is given by following equation: Here previously defined constants remain the same and L and P are numerical

PAGE 43

43 CH APTER 3 LIGHT EMITTING ELECT RONIC DEVICES BASED ON ORGANIC SEMICONDUCTORS 3.1 Fabrication Methods for Organic Electronics While organic semiconducting materials may have many desirable properties that can be used to harness electricity and light, these mat erials must be arranged in a meanin gful manner in order to achieve the desired results. More so than just a necessity, one of the primary advantages of organic semiconductors is in their versatility of processing and rela tive ease with which electronic de vices can be fabricated using these molecules. When discussing processing, the first distinction to be made is between the use of small molecules (SMOLED) and polymers (PLED) Small molecules are typically designated by low molecular weight (less than 100 0g/mol) whereas polymers are long chain molecules with a repeated unit and high molecular weight (greater than 1000g/mol). Based on this separation, the methods for fabrication break down into high vacuum processes such as, organic molecular beam depositi on (OMBD) and vacuum thermal evapor ation (VTE) or solution based processes such spin coating, spray deposition, inkjet printing, vapor jet printing, and roll to roll processing. While small molecules can be functionalized to be soluble, these materials are typically deposited using high vacuum methods that produce highly controlled film layers with pristine inter faces. Due to the low vacuum pressure (approximately 10 6 torr) compared to OMBD (10 8 10 12 torr) VTE is the most common method for small molecu le deposition [58] Figure 3 1 illustrates this method.

PAGE 44

44 Figure 3 1 Schematic of a t hermal e vaporation c hamber with the significant c omponents labeled. Essentially, a metal boat filled with the desired material is resistively heated under vacuum. The sublimed material is able to coat a substrate target above the material, due the high mean free path (on the order of tens of centimeters ) for molecules in the vacuum chamber. Through the use of shadow masks, various features can also be patterned using this method. A quartz crystal monitor ( QCM ) provides real time rate and thickness measurements accurate to .1 /s This method of deposi tion is also notable for its ability to deposit metals such as gold, silver, and aluminum as well as some oxides such as molybdenum trioxide (MoO 3 ), vanadium oxide (V 2 O 5 ), and tungsten oxide (WO 3 ). As a result, a complete device structure including electr ical contacts and encapsulants can be fabricated. Polymers, conversely, are typically deposited using solution based methods, the most common of which on the research scale being sp in coating (figure 3 2) [59] Polymer powders are dissolved into organic solvent and subsequently dropped onto a

PAGE 45

45 by the r ate of spin or solute concentration and typically a thermal annealing period is required in order to drive off excess solvent as well as control morphology. Solution processing is notable for the ability to dissolve multiple materials in the same solutio n which creates unique morphologies that can be controlled to optimize transport mechanisms throughout the film. Electrical contacts must still be deposited using an alternative method often high vacuum method. Figure 3 2 Diagram of spin coating depo sition 3.2 OLED Device Architecture As presented in the previous chapter, in order for a photon to be emitted, the electrons and holes in an exciton must recombine. The issue of course, is how do the charges reach appropriate molecule and efficiently rec ombine. 3.2.1 Operation Principles Figure 3 3 shows the operational structure of a simplified OLED device. Before contact, the Fermi levels of the anode and cathode of the device are not aligned (a), but once contact is made, an equilibrium Fermi level exists, but cannot transport charges

PAGE 46

46 due to a built in voltage barrier, V bi created by the alignment that prevents the movement (b). Once a potential, V, is applied to the electrodes, charges may move as soon as V= V bi (c) and inject into the organic lay er as greater values(d). Figure 3 3 Basic OLED operation A single layer of organic material is often time not sufficient in order meet all of the various functions of an electronic device. In the case of a light emitting diode, electrical contacts, charge injection and transport, and emission regions are all necessary to produce a fully functional and efficient device. Therefore, multilayer stacks wherein each layer serves a specific function are employed. With applied voltage, charges move from t he injection layers through the transport layers, and into the emitting layer where the opposing charges can recombine. Light is then able to through one of the electrodes which must be transparent. While each of these tasks seems simple, the

PAGE 47

47 specific qu alit ies each layer must possess in order efficiently achieve their purpose, are many. 3.2.2 The Multilayer Stack Each section of the device stack will be described in detail for its function as well as the properties for ideal material candidates. 3.2.2.1 Electrodes While seemingly trivial in many discussions of modern device structures, the electrodes are quite critical components of OLED design. Each electrode must be able to not only supply the respective charge, but also have optimal work functions and optical properties for the device stack. The anode or hole injecting electrode must have sufficiently high work function (greater than 4. 8 eV) in order to align with the HOMO level of the hole injecting or transporting material. Additionally, the anode is often used as the transparent electrode through which light generated in the device may escape to air for viewing. Few materials exist that are both transparent and electrically conductive. The most common of which is tin doped indium oxide (ITO) whic h is 90% transparent in the visible spectrum. This material must be treated with UV/Ozone in order to increase the work function from 4.5 eV to 4.8 eV. Further investigation into other transparent conductive oxides has begun due to the limited availabili ty of indium. The cathode or the electron injecting electrode requires just the opposite qualities as the anode. A low work function (~4.1 4.2 eV) metal with high reflectivity is necessary in order to match the LUMO level of the electron injecting or tra nsporting layers while also redirecting light back to the viewing direction. Al and Ag are typically used for this purpose, although a lower work function metal such as lithium or cesium is typically used as an interlayer in order to reduce the injection barrier.

PAGE 48

48 3.2.2.2 Charge injection layers The charge injection layer s while considered optional to device operation, do present a myriad of advantages to device efficiency. Most notably charge injection layers create a high density of charge carriers th at both reduce the turn on voltage and increase power efficiency. As the device stability tends to decrease with increased power, the ability to produce more light at lower voltages allows the device to be less taxed as well as consume less power at norma l operating parameters. The second advantage of charge injection layers is in the ability to more actively control the charge balance through doping concentration. In order to maximize the efficiency of devices, an equal number of holes and ele ctrons are necessary for effective radiative recombination. To produce a charge injecti on layer, host materials with HOMO or LUMO levels closest to the work function of their respective electrodes are used. Furthermore, the mobility of the charge that is to be car ried in the layer must be maximized. Finally, the layer is doped with either a material whose energy levels allow for a multiplication of charges in the layer as is the case in materials such as MeO TPD doped with F4 TCNQ which creates an abundance of cha rge carriers as electrons become trapped on the F4 conductivity creates even higher mobility in the host material, as is the case in BPhen doped with Cs [25] 3.2.2.3 Charge t ransport l ayers Charge transport layers are critical to device operation in a n efficient OLED. Few organic materials have HOMO and LUMO levels together that are close to the work functions of the adjoining electrodes or balanced mobility to bring these charges together and to confine them. Furthermore, such a material would also have a narrow

PAGE 49

49 band gap and would not be capable of emitting visible light. Therefore in o rder to shuttle charges from the electrodes to the emitting region, transport layers are necessary. Possible charge transport layers require t hree major criteria ; t hese layers must be highly mobile to the specific desired charge, prevent movement of excit ons back into the transporting material and act as an i ntermediary energy level for transition from the electrodes to the emitting layer While the first criteria is obvious the second is less so. By using materials with wide band gaps and more importantl y high triplet energies, excitons in the emitting layer will not be able to migrate out and prevent radiative recombination [23,24] 3.2.2.4 Emissive layer The emissive layer is the obvious crux of the OLED device. In this layer the electrons an d holes bind into excitons and finally recombine to emit a photon. The emissive layer of OLED was originally a single material with the recombination of the charges occurring on the same material they were injected into. This method was ultimately shown to be largely inefficient. Thus the properties of the emissive layer have often been separated in order to better accommodate its needs. The first component is the host, which is known for its more balanced electron a nd hole mobility as well as appropria te alignment of HOMO and LUMO levels to allow for charge injection. While excitons form on these molecules, there is little in the layer itself to confine charges to the layer. As a result the host material is usually of wider band gap than desired emiss ion energy. To achieve emission, a second material is doped into the host with lower band gap forcing the excitons to be confined before recombination. As shown in the previous chapter, only small fraction of the excitons formed exist in a singlet state able to produce fluorescent emission. In order to access the triplet states, heavy metal

PAGE 50

50 complexes are used as dopants. The result is an emissive layer that can convert 100% of the charges injected into the desired photons. 3.3 Measurement of OLEDs Fabri cation of an OLED is of course only half the task necessary to understand its behavior. The second component is to characterize i t s various optoelectronic properties. This section will illustrate the various techniques and figures of merit that are used to describe OLED behavior and performance. 3.3.1 Colorimetry Figure 3 4 The entire electromagnetic spectrum with the visible spect rum highlighted [60]

PAGE 51

51 Light, or the visible spectrum, is defined as the region of the electromagnetic radiation spectrum that can be per ceived by the human eye. The wavelengths ( ) of the entire electromagnetic spectrum are shown in figure 3 4 with the visible spectrum expanded. As can be seen, light has a wavelength between 380 nm (violet) and 780 nm (deep red) bordered by the invisible ultraviolet (UV) and infrared (IR) regions respectively. As the purpose of an OLED is to emit light, quantifying and categorizing the light produced is key to evaluating the performance. The first distinction to be made in the measurement of light is the difference between photometry and radiometry. While radiometry is the study of the entire electromagnetic spectrum, photometry is restricted to the visible range [61] Both techniques are used to quantify the energy emitted from a device, but given human perception is a critical component of evaluating the usefulness of OLEDs, the photometric quantities are typically the ones reported. T able 3 1shows the various terms associated with both photometry and radiometry. The units of measure and other terms will be further defined in the following sections. Table 3 1. Comparison of the various terms and units used for photometry and radiometry. Photometry Radiometry Illuminance L x lm/m Irradiance W/m 2 Luminous Flux lm Radiant Flux W J/s Luminous Intensity C d lm/sr Radiant Intensity W/sr Luminance cd/m 2 lm/sr m 2 Radiance W/sr m 2 Luminance Efficiency cd/A Radiance Efficiency W/sr A Luminous Power Efficiency lm/W Power Efficiency W/W

PAGE 52

52 3.3.1. 1 Responsivity of the human eye While light is the product of an OLED, the consumer of the product must be considered in order to tailor the device to its needs. The particular consumer of interest is the human eye which must be able to respond to daytime environments with illuminance of 100,000 lux (lx) while also responding to night time environments with illuminance as low as .0003 lx. In order to accomplish such a feat, the size of the pupil and the s ensitivity of the retina must change with the environment [62] The result of have two distinct modes, the photopic for normal or daylight environments (luminance greater than 3 cd/m 2 ) and the scoptic for low light environments (luminance less than cd/m 2 ) The photopic response, G( ) can be expressed as a function with peak lumi nous efficacy of 0 = 683 lm/W at = 555 nm compared to the scotopic, G ( ) which has a value of 0 = 1700 lm/W at 507 nm. The t erms g ( ) and g ( ) refer to the normalized responses for the photopic and scotopic responses, respecti vely, with peak values of unity at the peak wavelength. Figure 3 5 illustrates these normalized curves. For the purposes of this dissertation the photopic response will be used for luminance calculations.

PAGE 53

53 Figure 3 5 The normalized scotopic (green and left ) and photopic (black, dotted, and dashed to the right ) curves [63] 3.3.1. 2 Angular emission Light emitting devices do not direct photons in a single direction. Therefore an explanation of two dimensional and three dimensional angles is necessary. A plane angle ( ) is the ratio of the length of the arc of a two dimensional circle to its radius described by the SI unit radian (rad), as opposed to a solid angle ( ) which is defined as the ratio of the spherical surface area to the square of the radius of a three di mensional sphere described by the SI unit steradian (sr). A visual representation of this is seen in figure 3 6.

PAGE 54

54 Figure 3 6 Comparison of the planar angle to solid angle A ) a planar angle in two dimensions and B ) a solid angle in 3 dimensions 3.3.1. 3 The ideal Lambertian emitter The ideal emission source for light emitting devices is known as the Lambertian source and is defined as being isotropic with equal luminance in any solid angle. A representation of the luminous intensity can be seen in figur e 3 7. Figure 3 7 Two dimensional representation of a uniform Lambertian emission source [64]

PAGE 55

55 The intensity follows a cosign law when observed from the normal to the emitting surface. While it may seem counter intuitive, the reas on that the luminance does not change whereas as the luminous intensity does as a function of viewing angle, the reason is that the area of the emitter viewed changes as well. 3.3.1. 4 CIE coordinates and color rendering index One final piece to understan ding the light emission from a device is, of course, the color. Light sources may be single color or wide band like white light. Two figures of merit are used to describe color from light emitters. The Commission Internationale de colo r space created in the early 20 th century, allows the hue emitted from a source to be defined using (x,y,z) coordinates. These coordinates are dependent on the tristimulus curves ( ) seen in figure 3 8 as well as by the wavelength dependent intensity of the emission spectrum. Figure 3 8 The tristimulus curves used to determine CIE coordinates [65] The tristimulus valu es (X,Y,Z) are derived integrating the curves by the spectral intensity as seen in the following equation.

PAGE 56

56 From these values the coordinates are calculated as follows: As the summation of x+y+z equals unity, the color space is most often described in two dimensional coordinates (x,y). Using this notation, the color space in figure 3 9 can be rendered. The second term important term related to color is the color rendering index (CRI). The CRI is a quantita tive method of describing how well a light source can represent the colors of an object compared to a reference lighting source. The CRI is then listed as a relative value from 0 to 100 with a higher value corresponding to closer rendering with respect to the reference. Black body radiators which exist along the Planckian locus as is seen in figure 3 9 can show a CRI value of 100. Therefore, a known black body radiator, the incandescent light bulb is typically the reference source used for comparison. T ypically the term CRI actually refers to the average of 14 different color rendering values which compare the rendering of 14 specific colors (CRI 1 14 ). Commonly, only the CRI 1 8 are used for the general CRI.

PAGE 57

57 Figure 3 9 Two dimensional CIE coordinate plane with black line indicating the Planckian locus [66] 3.3.2 Device Characterization Now that the figures of merit have been explained, the question remains how these values as well as the electrical properties of the device are acquired. Through the use of standard measuremen t techniques, consistent comparison can be made between lighting sources from various research and commercial sources. The following sections describe the methods by which devices are characterized based on published methods [67] Equipment between testing centers may differ but the application of these techniques allows for system independent study of devices. 3.3.2.1 Current v oltage m ea surement The current and voltage measurements for OLEDs are the corner stone from which to discuss device behavior. Measurement is conducted in ambient conditions

PAGE 58

58 through the use of a semiconductor parameter analyzer (Agilent 4155C) which is electrically contacted to the device The voltage applied ( V ) to the device is incrementally increased and the current generated ( I d ) is recorded though it is typically reported as the current density which is the current per device area ( J ) By combining this infor mation with the corresponding luminance measured, the operating parameters of the device can be ascertained. 3.3.2.2 Luminance m easurement and c alibration Luminance, as mentioned earlier, is the luminous flux light per unit area per unit solid angle or i n units, candela per meter squared (cd/m 2 ) or lumens per steradian per meter squared (lm/sr m 2 ) Obviously this term is based on photometric units and therefore must be calibrated using the photopic response. This calibration is conducted using a lumina nce meter ( Konica Minolta LS 100 with No. 110 close up lens) and calibrated silicon photodetector (Newport 818 UV). By attaching the photodetector to the parameter analyzer, the photocurrent generated ( I det ) by the device ca n be simultaneously recorded w hile current voltage measurements are conducted. The luminance meter is then used to measure the actual luminance ( L ) at a given photocurrent in order to derive a conversion factor ( ) for the dedicated measurement set up assuming ideal Lambertian emissio n 3.3.2.3 Emission s pectra In order to assess the color emitted from the device as well as adjust the photocurrent measurements by the responsivity of the detector, the emission spectra must be captured. Using a spectrometer (Oc ean Optics Jaz ) calibrated with a tungsten

PAGE 59

59 lamp, the wavelength dependent emission intensity of the device can be measured. Furthermore, using optical software, the CIE coordinates and CRI value can be derived. 3.3.2.4 Lumin ance e fficiency Luminance or C urrent Efficiency ( L ) is the most basic assessment of the capabilities of a fabricated device. It is defined as the ratio of the luminance to the current injected into the device. The following equation illustrates this relationship: This efficiency is reported in cd/A and taken directly from the calibrated measurements. From this efficiency rudimentary understanding of the light out versus the charges put in develops. 3.3.2.5 Luminous p ower e fficiency The luminous power effici ency ( P ) is calculated assuming a two dimensional area of emission that is Lambertian in emission (figure 3 10) The quantity is derived from the following equation:

PAGE 60

60 Figure 3 10 A two dimensional Lambertian source with light emitting area A and surface area dS to define the solid angle In this system, A is the total luminous flux in lumens (lm) from the light source with a certain area ( A ) and P is the electrical power consumed by the device in watts (W). P can also be thought of as radiometric unit radiant efficacy which is optical power out per electrical power in (W/W) multiplied by the photopic luminous efficacy of the light. This relationship illustrates two methods of improving the P of a device: first if more optical power is closely aligned to the photopic response. The following derivation shows how P is calculated using this relationship.

PAGE 61

61 Here G( ) and g( ) are the previously reported wavelength dependent photopic response and its normalized counterpart. S( ) and s ( ) are the spectrum dependent optical power and normalized counterpart. T he device current ( I d ) injected at a given voltage ( V ) and the photocurrent ( I d et ) are from the direct measurements conducted R( ) is the responsivity of the photodector and f is the geometric factor of the measurement set up. For commercial lighting applications, this value is important in determining the energy costs an OLED wil l require in terms of its application. 3.3.2.6 External Quantum Efficiency While the luminance efficiency is a rudimentary way of correlating electricity to light, the most precise quantity to do so is known as the external quantum efficiency ( EQE ) whic h directly shows the ratio of photons emitted to air to the amount of charges injected into the device Since this is based purely on physical quantities, the photopic response is not considered when calculating the value. Alternatively, the photodetecto r used must be carefully chose based on the wavelengths of light to be measured. The following equations show the derivation of this quantity. In this derivation,all previously mentioned terms remain the same with q representing th e fundamental charge of an electron (1.602 10 19 coulombs) h constant (6.626 10 34 J s) and c is the speed of light (299,792,458 m/s). This figure is

PAGE 62

62 ne form of energy to another.

PAGE 63

63 CHAPTER 4 OLED OPTICAL PATHS A ND MEANS OF ENHANCEM ENT 4.1 Light Path in Basic Bottom Emitting OLED Architecture While each of the components of an OLED described in the previous chapter are critical to the electrical performa nce, one aspect not considered is the desirable characteristics necessary to maximize light transmission to air. The term most commonly associated with the amount of light that reaches air is the previously mentioned external quantum efficiency. While a value for EQE can be derived from physical measurements, the term actually encompasses the various physical processes. The two primary terms are the internal quantum efficiency ( I QE ) and the outcoupling efficiency ( outcoupling ) The internal quantum efficiency is more directly the number of photons created per the number of charges injected. This term is dependent on the efficiency of three processes within the device. First i s the charge recombination efficiency ( R ) which requires balanced charge injection from the electrodes in order to form excitons and prevent non radiative relaxation [68 71] As excitons are formed in a one to one ratio, by adjusting the charges injected from either side to be equal, one can potentially 100% recombination efficiency. The second process of interest is the fraction of exploited excitons. As explained previously, organic molecules exi st in two states, the singlet and the triplet. Singlets account for 25% of all excitons formed and triplets 75%. Therefore in order for 100% of all excitons to be accessed, a material that takes advantage of metal ligand charge transfer must be used [27,72] The final process is the photoluminescent quantum yield, which is a ratio of photons generated to photons used to excite a material. While this quantity can be extremely high for materials suspended in solution [73] solid films tend to have reduced values [ 74] In

PAGE 64

64 order to combat this ineffi ci ency, films consisting of a host and emissive dopant layer with optimized doping ratio can increase the yield. While maximizing these terms can produce a nearly 100% internal quantum efficiency, the limiting factor for the EQE is the outcoupling efficiency, whereby the charges generated escape the device stack. Figure 4 1 crudely illustrates the paths light may take when passing through the various layers of a complete bottom emitting (light emits through the substr ate) OLED. Figure 4 1 Optical modes in bottom emitting device. Light is isotropically generated in the emitting layer of the device leading to emission at various angles and directions. The different paths or modes of light travel can be described b y the regions that are significant to the device stack The fir st mode is defined by the organic/ITO layers (n=1.7 2.0) where light can be completely internally reflected as it reaches the lower index g lass substrate (n=1.5 ) [75 77] This mode accounts for 50% to 60% of all the light generated in the device. The second mode is

PAGE 65

65 the define d by the glass substrate itself where light that has escaped the device stack is now internally reflected throughout the substrate due to glass/air interface ( n air = 1.0) and accounting for another 20 30% of the light The final mode is defined by the lig ht that is able to escape the substrate and finally be emitted into the air [21,78] Due to the various interfaces with changes in refractive index, it is easy to see how a device with perfect internal efficiency can lead to such low external ef ficiencies. 4.2 Fabrication Methods to Alter Optical Modes The previous section detailed the reasons which in spite of a highly efficient device, the amount of useable light is extremely limited. In order to access more of the light already generated b y the device structure, various methods of altering the optical modes have been examined to enhance outcoupling The following sections will address some of these methods. 4.2.1 Cathode Surface Plasmons The first method is designed to access some of the light from the first optical mode presented. One reason that much of the light generated never lea ves the organic layers is due to a phenomenon known as surface p lasmon quenching. Essentially, light interacts with the sea of electrons that exists at the metal cathode/organic interface causing the electrons to oscillate. This oscillation extends outwa rd horizontally in the device meaning that the light is dissipated instead of reaching a forward direction. In order to reduce this quench ing, a few methods have been attempted to alter the way that light interacts with the metal electrode. The first is to move the emitting layer as far from the cathode as possible [79] This is in turn reduces the likelihood that quenching will occur in the first place. While a simple solution, the shifting of the emitting layer has the potential to upset the charge transport capabilities of the OLED. The second

PAGE 66

66 method requires modifying the structure of the electrode to incorporate nanoscale feature s that can act as antennae to retransmit the light in the forward direction [80 83] Figure 4 2 Comparison of planar and surface p lasmon enhanced device structures A) Planar. B ) Modified for enhanc ed extraction By redirecting the electromagnetic fields at the cathode/organic interface to incorporate a vertical component, light previous trapped horizontally can now be emitted in the forward direction. This approach has been shown to increase the li ght extracted by 2.35 times that of the standard device architecture [80] The major limitation to this method is that i t typically requi res some kind of template in order to produce the nanoscale features. Such templates require lithographic methods common to inorganic semiconductors and therefore adding significant cost and difficult processing to the OLED fabrication. Furthermore, thes e techniques do not necessarily allow for commercial scale application.

PAGE 67

67 4.2.2 High Index Glass The second method utilized takes the focus of improvement to the other end of device. Due to the reduction in refractive index at the anode/glass interface, muc h of the light generated is reflected back into the organic layers. The easiest way to prevent this is to have the glass match the refractive index of the rest of the device. By removing one of the reflective interfaces, more light reaches the glass subs trate and therefore more light can exit the OLED [84,85] This method has been used in record setting device structures that can match fluorescent tube efficiencies in the lab (90 lm/W) [29] and shows a 2.7 fold increase from the bare device. A lthough the use of high index glass shows a high enhancement, it may present a cost prohibitive challenge to commercial applications 4.2 .2 Low Index Gr id s One further method to improve this interface is change the optical path of light as it passes through the high index organic layer In order to do that, low index two dimensional grids with micron scale spacing allow light that is norma lly transmitted horizontally through the organic layer to instead pass into the sides of the grid, which in turn allows more light to enter the substrate [38,86] Through the application of extremely low (n=1.15) index grids, enhancements of up to 2.9 times the efficiency of conventional devices has been seen. This method still relies on traditional cleanroom processing techniques, either through lithography or oblique angle deposition in order to produce.

PAGE 68

68 Figure 4 3. Light path through an OLED with incorporated low index grid. 4.2.3 Qua rter Wave Stacks (Bragg Reflector) Figure 4 4. Demonstration of light path through a Bragg reflector. The backwards reflected light accounts for multiple reflections at the various interfaces in the stack

PAGE 69

69 One m ethod to improve the outcoupling capabilities of OLEDs relies on selective control of the electromagnetic wave of light in order to create constructive interference. Through the use of alternating layers of high and low index materials at the interface be tween the transparent electrode and substrate, light waves can be selectively reflected back at each interface This wave, due to both the forward directional and back reflected light, will experience constructive interference and by tailoring the device thickness optimize the interference to the wavelength By adjusting the thickness of these alternating high and low index layers to a quarter of the desired wavelength, more light of that wavelength is able to escape the device stack and reach the substra te mode (figure 4 4) This microcavity effect is therefore a positive enhancement for single color light sources showing 1.6 1.7 times enhancement of light while broad band sources such as white light can be limited by the application of the stack [87 89] Therefore, to produce white light, typically a high energy single wavelength source is produced by this method with additional down conversion layers to create white light. 4.2.4 Microlens Arrays The most ideal method of enhancing the outcoupling efficiency of light would be to apply a large high index of refraction lens to the substrate which would allow most light trapped in either the substrate or device stack to reach through the 90 contact angle of the lens. While this technique is ideal, a lens of sufficient size to cover any practical sized device would greatly increase the size of the product. One alternative to produce similar light extraction would be through the use of microscopic lens arrays that could be produced across the surface of the device both emulating the enhancement of the large macro lens and maintaining the form factor of the device itself. While most current attempts to produce microlens arrays require expensive processing techniques that also

PAGE 70

70 limit the scale to the arrays can be printed [37,90] some recent methods utilizing either soft lithographic [30] or inkjet printing methods have begun to be explored. By using monomer materials that have a refractive index that matches that of t he device and can be UV cured into a high contact angle solid state, this method presents the possibility for high efficiency devices that can be scaled to commercial levels. Table 4 1 summarizes these methods. Table 4 1. Recorded enhancement of various outcoupling methods over traditional OLEDs. Technique Enhancement Multiplier Cathode Surface Plasmons 2.3x [80] High Index Glass 2.7x [29] Low Index Grids 2.9x [86] Bragg Reflector 1.7x [87] Microlens Arrays 1.7x [90,91]

PAGE 71

71 CHAPTER 5 TRANSPARENT OXIDE/ME TAL/OXIDE TRILAYER E LE CTRODE FOR USE IN TO P EMITTING ORGANIC LIG HT EMITTING DIODES 5.1 Introduction Organic light emitting diodes (OLEDs) are already commercially available in small area displays and quickly approaching the point of commercial viability for lighting application s. White light devices with efficiencies reaching 100 lumens per watt (lm/W) have been achieved [29] which exceed the maximum efficiencie s of the incandescent (17 lm/W) and fluorescent bulbs (90 lm/W) [92] Soon flat panels with strong color rendering capabilities could repl ace conventional lighting fixtures both at home and in commercial environments. While these devices could one day revolutionize technology, one important component that still requires investigation is the transparent electrode, which must be conductive bu t also transparent so as to facilitate the transmission of light to and from the active organic layers. The most common transparent electrode used for these devices is indium tin oxide (ITO), a degenerately doped metal oxide consisting of 90% In 2 O 3 and 10 % SnO 2 typically deposited in a sputtering system at elevated substrate temperatures to simultaneously optimize the optical transparency and electrical conductivity [93,94] This electrode can be patterned onto glas s or other transparent substrates, and has high transparency of approximately 90% across the visible spectrum and low sheet resistance of 10 for a 100 150 nm thick film; however, the high temperatures induced by the plasma as well the necessary post deposition annealing process will cause the thermal expansion of polymer based s ubstrates and significantly reduce the conductivity of the layer. Typically, OLEDs are designed to emit through ITO and the transparent substrate before reaching air (there is typically a metal electrode on the other side of the organic

PAGE 72

72 multilayer stacks to reflect the light toward the ITO electrode). This leads to three distinct interfaces which light must pass through: one between the organic layers and ITO, one between the ITO and glass, and one between glass and air. The differences in indices of re fraction between these components mean that at each interface light has the potential to be internally reflected, reducing the overall efficiency of the device. If one could remove the interface between glass and substrate by switching the position of the transparent ITO electrode and the reflective metal electrode, light previously trapped internally in wave guided modes could now partially escape the device [95,96] This top emitting structure, wherein the device emits directly from a transp arent electrode to air, has the potential to improve the outcoupling of light [97,98] and drive organic devices towards higher efficiencies. Furthermore these devices can also be fabricated on opaque substrates like silicon, allowing for integration into current active matrix display technologies. Unfortunately, with respect to top emitting organic devices, the ITO electrode does not easily integrate into the fabrication process due to its method of deposition. The sputtering process is highly destr uctive to the soft organic layers underneath [99] Therefore, another transparent electrode is necessary for use in top emitting or s emitransparent organic devices. Recently, the use of multilayer patterned electrodes consisting of thin metal layers sandwiched by oxi de layers has been investigated [100 104] While these electrodes have typically been deposited using e beam deposition and therefore require additional processing steps and cham bers apart from traditional thermal evaporation, Yook et.al. have shown that such a structure can be thermally evaporated as well, which integrates into the fabrication of small molecule organic devices and has less potential to damage

PAGE 73

73 the underlying layer s [102] Typically the metal used has been Ag, while the oxide layers have run the gamut from ITO to WO 3 to ZnO. This chapter seek s to expand the understanding of these multilayer electrode structures, and their integration into top emitting diode design for high efficiency OLEDs. Previously the use of Ag as the intermediate layer showed a wavelength dependent transmittance favoring the blue region (400 500 nm) of the visible spectrum, and the application of this structure as a cathode i n place of the typical Al or Ag [100] By contrast, replacing this layer with Au and sandwiching it between layers of molybdenum oxide (MoO 3 ), a shift in peak transmittance towards the green and red regions (550 600 nm) of the spectrum is shown while maintaining a low sheet resista nce. Furthermore, when ultrathin layers of Au and Ag are stacked in this structure, the peak transmittance can be both broadened and tuned based on the proportions of each layer. Finally, the structure was incorporated as an anode into OLED devices showin g the potential for high efficiency devices with an inverted top emitting structure. 5.2 Experimental Methods All structures were deposited on glass substrates using vacuum thermal evaporation at a pressure < 3.0 10 6 Torr. The substrates were cleaned by submerging the samples in beakers of detergent and water, de ionized water, acetone, and isopropanol successively and each beaker was ultra sonicated for a period of 15 minutes. The transparent oxide / metal / oxide structures were deposited sequentially using 5 nm of MoO 3 followed by various thicknesses of Au or Ag and finally 40 nm of MoO 3 OLEDs were then fabricated using an inverted top emitting structure and compared to a standard bottom emitting architect ure deposited on top of a precoated ITO layer. Figure 5 1 illustrates these two different structures. Both fluorescent and

PAGE 74

74 diphenyl bis(3 methylphenyl) bipheny l] diamine (MeO TPD) doped with tetrafluoro tetracyanoquinodimethane (F 4 TCNQ) as the hole injecting layer (HI L) [25] bis (1 naphthyl) diphenyl1 1,1 biphenyl1 diamine (NPB) as a hole transporting layer (HTL) [105] and tris(8 hydroxyquin olinato) aluminum as the electron transport (ETL) and emitting layer (EML) [17] The respective transparent electrodes (ITO for bottom e mitting devices and the oxide/metal/oxide trilayer for top emitting devices) were used as the anodes, whereas a Cs 2 CO 3 interlayer [24] a nd Al were used as the cathode. Phosphorescent devices consisted of a p i n structure with MeO TPD doped with F 4 TCNQ as the HIL, 1,1 bis (di 4 tolylaminophenyl)cyclohexane (TAPC) [23] dicarbazolyl 3,5 benzene (mCP) doped with 10 wt% fac tris(2 phenylpyridine) iridium (Ir(ppy) 3 ) green p hosphorescent dopant as the EML [106] 4,7 diphenyl 1,10 phenanthroline (BPhen) as ETL and BPhen doped with CsCO 3 electron injection layer (EIL) [24] The same anode and cathode structures as the fluorescent device were also employed. Fig ure 5 1. Schematic device structures for the standard bottom emitting architecture based on the commercial ITO ele ctrode and the top emitting architecture based on the MoO 3 /metal/MoO 3 trilayer transparent electrode

PAGE 75

75 The optical transmittance of the films were obtained by passing the white light from a Oriel Apex Monochromatic Illuminator through an Oriel Cornerstone 2 60 monochromator and using a Newport 818 UV photodetector to measure the monochromatic light intensity. The sheet resistance of the transparent electrodes was measured using the four point probe technique. Luminance (L) current density (J) voltage (V) me asurements were conducted in ambient conditions using an Agilent 4155C semiconductor parameter analyzer and the aforementioned Newport 818 UV photodetector. The luminance of the OLEDs was calibrated using a Konica Minolta LS 100 luminance meter assuming a Lambertian emission pattern. Electroluminescence (EL) spectra were taken using an Ocean Optics Jaz spectrometer. The power and external quantum efficie P EQE respectively) were der ived based on published methods as explained in the previous chapter [67]

PAGE 76

76 5.3 Optoelectronic properties of multilayer structures Figure 5 2. Transmittance of the ITO and MoO 3 /Au/MoO 3 electrode with various Au layer thickness. Figure 5 2 shows the optical transmittance of the trilayer MoO 3 / Au / MoO 3 structures with varying Au layer thicknesses as compared to that of the commercial ITO electrode. With a 5 or 10 nm thick Au layer, the trilayer structure can achieve a maximum transmittance of 85 90%, only slightly lower than that of the ITO. The transmittance of this trilayer structure does show stronger dependencies on the wavelength than the ITO, and the transmittance of trilayers with 5 10 nm thick Au intermediate layers is reduced to approximately 75% at 500 nm. The trilayer

PAGE 77

77 transmittance also shows a strong dependence on the Au layer thickne ss. While the structure with 5 nm thick Au layer shows broad transmittance, as the Au layer thickness increases, a characteristic peak transmittance begins to form around 580 nm, and the overall transmittance of the structure steadily decreases, particular ly for wavelength <500 nm. Figure 5 3. Sheet resistance and transmittance at 600nm as a function of Au layer thickness. Figure 5 3 shows the relationship between the transmittance (at 600 nm) and the sheet resistance for trilayers with varying Au layer thickness. While the 5 nm thick Au

PAGE 78

78 layer, the sheet resistance of the trilayer is r the Au layer thickness, although this is accompanied by a simultaneous reduction in the optical transmittance. Therefore in order to apply this electrode to light emitting devices, a balance between the transmittan ce and sheet resistance must be achieved. One important observation relates to the 10 nm sample, which possesses both sufficiently highest transmittance at 600 nm. This phe nomenon suggests a correlation between the more coherent Au layer and an optical enhancement perhaps related to surface plasmon resonance. This effect is further supported by the change in peak transmittance seen as the metal material is changed, suggesti ng a material dependent resonance. Figure 5 4 shows that by changing the metal intermediate layer used to Ag or depositing sequentially stacked Au/Ag layers, the maximum transmittance wavelength shifts from the red region of the electromagnetic spectrum to the blue (Ag) or green (Au/Ag) region. The 5nm/5nm Au/Ag stacked interlayer appears to be best suited for white light emitting devices as the corresponding transmittance is over 70% for the entire wavelength range of 450 to 650 nm, whereas either Ag or Au only results in a relatively poor transmittance at one end of the visible spectrum (long wavelength end for Ag and short wavelength end for Au). The transmittance of this stacked interlayer appears to average the transmittance of the two components. T his feature is further illustrated by adjusting the proportions the metals in the layer suggesting that this structure can be completely tunable to a variety of niche applica tions and emissive molecules.

PAGE 79

79 Figure 5 4. Wavelength dependant transmittance of multilayer structures with Au, Au/Ag, and Ag intermediate layers. 5.4 OLED devices with trilayer transparent electrodes As this MoO 3 /metal/MoO 3 trilayer structure shows promising electrical and optical properties, its application to OLED devices must al so be examined. First, fluorescent devices were fabricated as previously described using the multilayer electrode with 10 nm Au as the intermediate layer. Figure 5 5(a) shows the L J V characteristics for the bottom emitting Alq 3 device (with a commercia l ITO anode) and its top emitting counterpart with a MoO 3 /Au/MoO 3 anode. The top emitting device shows somewhat lower current than the bottom emitting device, especially at voltages just above the turn on voltages of the devices (~2.2 V). The inferior ch arge injection in the top emitting

PAGE 80

80 device may not be entirely associated with the transparent trilayer electrode. More likely it could be due to the electron injection from the Al cathode. In a typical bottom emitting structure, an interlayer of LiF or Cs 2 CO 3 is first deposited onto the organic layer followed by an Al cathode, and reactions and diffusion may occur upon the deposition of hot Al atoms, which subsequently reduces the electron injection barrier [19,107] When depositing a top emit ting structure with Al followed by an interlayer (such as Cs 2 CO 3 used here), the same series of events may not occur. Nonetheless, the emission spectra of the bottom and top emitting devices as shown in Fig 5 5(b) are nearly identical. This indicates tha t if the injection barrier can be overcome, the transmittance of this structure is unlikely to affect the emission of the device. Figure 5 5. Device Properties of B ottom and Top Emitting Devices A ) L J V characteristics of fluorescent bottom and top emi tting devices. The commercial ITO anode was used in the bottom emitting device whereas the MoO3/Au/MoO3 trilayer electrode was used as the anode in the top emitting device. B ) Electroluminescent (EL) spectra of the two devices.

PAGE 81

81 Figure 5 6. High effic iency p i n green phosphorescent devices A ) Energy level diagram of device structure B ) External quantum and luminous power efficiencies of these two devices. T he trilayer transparent electrode was also applied to a p i n phosphorescent green emitting de vice with both hole and electron injecting layers. The schematic energy level diagram of such devices is shown in Fig. 5 6(a). As shown in Fig. 5 6(b), the external quantum efficiency of the top emitting device was mostly the same as that of the bottom em itting device in the current density range of 0.1 to 100 mA/cm 2 The power efficiency of the top emitting device; however, shows a 20 30% reduction

PAGE 82

82 compared to the bottom emitting device in this current density range, which is attributed to the somewhat h igher drive voltage of the top emitting device similar to the case of the fluorescent devices (Fig. 5 5a). Based on this green p i n structure, the potential application of this electrode to high efficiency light emitting devices can be seen, so long as th e high turn on voltage can be overcome. 5.5 Summary A new architecture for a transparent multilayer electrode for use in thermally evaported top emitting OLEDs was demonstrated While its transparency is wavelength dependent, maximum transmittance up t o 85 90% is achieved along with sheet resistances lower than typical commercial ITO electrodes. Furthermore, the wavelength dependence of the transmittance of such electrodes can be tuned based on the metal material used for the intermediate layer. When Au containing multilayer structures were used as the anode in a top emitting OLED, the emission spectra of the device matched that of the bottom emitting devices with ITO anodes. These top emitting devices also exhibit similar external quantum efficiencie s as the bottom emitting devices; however the somewhat inferior charge injection, likely caused at the Al cathode, leads to higher operating voltages and lower power efficiencies in the top emitting devices. Further examination of these multilayer structur es is clearly necessary as they present a promising avenue for the replacement of ITO in OLEDs

PAGE 83

83 CHAPTER 6 ENHANCING LIGHT EXTR ACTION IN TOP EMITTING ORGANIC LIG HT EMITTING DEVICES USING MOLDED TRANSPARENT POLYMER MICROLENS ARRAYS 6.1 Introduction T he low light extraction efficiency in OLEDs poses a critical limitation to the overall efficiency of these devices. In a typical OLED, only approximately 20% of the emission can escape the device i n the forward viewing direction [108] While v arious methods have been reported to enhance light extraction in OLEDs, the more effective methods often involve using expensive materials or complex processing routines [26,29,30,32 38,109 111] In this chapter, the use of mold ed close packed, hemispherical, transparent polymer microlens array s will be examined in order to e nhance the OLED light extraction efficiency o f top emitting OLEDs. The microlens array does not appreciably alter the thin form factor of OLEDs, and can be readily scaled for low cost, large area device manufacturing. As a result, the optimization of this technique presents a very attractive option for creating solid state white light devices. This chapter is an extended discussion of results previously present [31] 6.2 Experimental Methods Optical simulations were performed using b oth ray tracing and wave optics models to calculate the light extraction efficiency in top emitting OLEDs with a close packed hemis pherical microlens array. Th e T OLED device used for simulation consists of an Al reflecting electrode on a substrate, a 140 nm thick organic layer ( n org = 1.7), a 40 nm thick transparent electrode ( n = 2.0), and a microlens array whose index of refraction is varied from n lens = 1.0 ( meaning no microlenses) to 2.3. Except for Al, all other materials are considered fully transparent in simulations.

PAGE 84

84 For ray tracing, the Monte Carlo simulation started by randomly selecting a position and direction of light emission (a ray), followed by t racing the trajectory of that ray without considering interference. The ray was either reflected or transmitted at a material interface based on probabilities determined by the reflectance and transmittance across that material interface. The light extrac tion efficiency is calculated as the ratio of the number of rays that escape the device through the microlens array to the total number of rays used for simulation (which is on the order of one million). Using a commercial software (Lumerical Solutions I nc., Canada), the finite difference time domain (FDTD) algorithm was also employed to simulate the light (electromagnetic wave) propagation in the OLED multi layer structures with or without a microlens array attached. Emission from the organic molecules was modeled as a sheet of dipoles with random orientation positioned at a fixed distance away from the reflecting metal electrode. The light extraction efficiency is calculated as the ratio of the optical power radiated through the microlens array to the t otal power generated by the dipoles in the active layer. The diameter of the microlenses was set as 2 m (instead of 100 m in actual experiments) to reduce the simulation time and computer memory space. Significant wavelength dependence in the calculated LE was observed due to the increased interference effect with smaller microlenses. The error in LE as shown in Fig ure 6 2 ( c ) was taken from the standard deviation of LE in the wavelength range from 500 nm to 600 nm. Small molecule based, top emitting O LEDs with different color emissions were fabricated using vacuum thermal evaporation in a custom made high vacuum evaporator (base pressure ~ 1 10 7 Torr). The OLEDs generally consist of a 100 nm

PAGE 85

85 thick Al cathode on a pre cleaned glass substrate, a 1 nm t hick interfacial Cs 2 CO 3 layer [24] an organic multilayer stack, and a MoO x (5 nm)/Au (10 nm)/MoO x (40 nm) trilayer transparent anode [58] The organic multilayer stack consists of (in deposition sequence) a 40 nm thick t ris[3 (3 pyridyl)mesityl]borane (3TPYMB) electron tran sport layer [24] a 20 nm thick emitting layer (EML), a 10 nm thick 1,1 bis (di 4 tolylaminophenyl)cyclohexane (TAPC) hole transport lay er, and a 30 nm thick hole injection layer of N, N' diphenyl N,N' bis(3 methylphenyl) [1,1' biphenyl] 4,4' diamine) (MeO TPD) doped with 2 mol% of tetrafluoro tetracyanoquinodimethane (F 4 TCNQ). The green OLED employed an EML of dicarbazolyl 3,5 benze ne (mCP) layer doped with 8 wt% fac tris (phenylpyridine) iridium [Ir(ppy) 3 ] [106] For the blue green OLED, the EML is mCP layer dope difluorophenylpyridinato)tetrakis(1 pyrazolyl)borate ( FIr6 ) [22,23] and 1 wt% Ir(ppy) 3 Combining FIr6 and Ir(ppy) 3 together with the red emitting iridium(III) bis (2 phenylquinoly N,C )dipivaloylmethane (PQI r) in a dual EML structure as r eported previously led to white light emission [26] The typical OLED active device area was 4 mm 2 as defin ed by a cross bar electrode geometry. Larger size devices with an area of 1 cm 2 were also fabricated for direct demonstration of the light extraction enhan cement with the microlens array.

PAGE 86

86 Figure 6 1. Schematic of microlens fabrication process. A) PS bea ds on a silicon substrate. B) Template for a PDMS mold. C) Cured mold with beads. D) Removal of beads. E) Filling with NOA. F) Application of filled mold to device. G)The mold is removed to form the lens array [30] T he previously published soft lithography process to fabricate microlens arrays on top emitting OLEDs [30,91] was adapted here for use A schematic of the process is see n in figure 6 1. First, polystyrene microspheres (PS) in aqueous solution (2.1% for 100 micron diamater spheres) were applied to precleaned, hydrophilic, UV/ozone treated silicon substrates with native oxide layer. Using convective and capillary assempl y, a tightly closed packed monolayer film is formed as is seen in figure 6 2

PAGE 87

87 Figure 6 2 Silicon wafer with polystyrene bead mono layer A ) M acroscopically B ) U nder an optical microscope [91] With domain sizes on the order of 1 mm and a hexagonally closed packed array of beads, this film can be used to create microlens templates. To form this template, ss can be seen in figure 6 3 poly dimethylsiloxane (PDMS) precursors were mixed 10:1 ratio of base polymer to curing agent and poured onto the monolayer film. After 3 hours of vacuum exposure to remove any air bubbles, the precursor was cured at 60 C for 2

PAGE 88

88 hours under vacuum (6 3 a) Us ing common packing tape, the beads can be removed from the curec PDMS leaving concave hemispherical surfaces (6 3 b,c). Figure 6 3 Scanning Electron microscope images of microlens formation process. A) Cur ed PDMS with PS beads. B) Removal of beads from mold. C) Concave PDMS mold. D) Cured microlens array [91] Norland Optical Adhesive (NOA) 68 (n=1.54) monomer liquid is used to fill the t emplate which can be placed on onto the glass substrate of an OLED. Using UV light ( =365 nm), the NOA is then cured and the template pealed back to form a microlens array(6 2d). The luminous, power, and external quantum efficiencies ( L P and EQE respectively) were determined following the recommended methods [67] Current density voltage ( J V ) characteristics of OLEDs were measured under ambient

PAGE 89

89 conditions using an Agilent 4155C semiconductor parameter analyzer Unless otherwise noted, a calibrated Newport 818 UV silicon diode was placed in close proximity to measure the OLED light output, and the luminance was calibrated using a Minolta luminance meter (LS 100). The EL spectra were taken using an Ocean Optics J az spectrometer coupled to an optical fiber. For direct demonstration of the light extraction enhancement with a microlens array on TOLED s, large pixel OLEDs with a 1 cm 2 active device area were fabricated and covered half of the device area with a microl ens array (while leaving the other half uncovered). With the device operated at a constant current (in the range of 1 to 10 mA), a n optical fiber (diameter = 400 m) was positioned perpendicular to the device surface and within 1 mm of the light emitting s urface. The near field EL spectra at multiple spots within the MLA covered and bare regions of the OLED were recorded by an Ocean Optics Jaz spectrometer connected to the receiving end of the fiber. 6.3 Comparison of Bottom and Top Emitting OLEDs Conventio OLED) structure as schematically illustrated in Figure 6 4 ( a ) which consists of a planar substrate such as soda lime glass (with an index of refraction of n glass 1.5), a layer of transparent cond ucting indium tin oxide (ITO) anode ( n ITO 1.8 2.0), multilayers of organic thin films ( n org 1.7 1.8) with a total thickness on the order of 100 nm, and a reflecting metal cathode such as Al. For light emission in directions that have large angles from the substrate normal, the total internal reflection at the ITO/substrate or substrate/air interfaces prevents the light from exiting the device through the substrate surface, leading to waveguiding modes confined in the substrate, ITO and the organic

PAGE 90

90 layer s (modes ii and iii in Figure 6 4 ( a ) ). The light extraction efficiency of the OLED, LE defined as the ratio of the number of photons coupled out of the device in the forward viewing directions to the total number of photons generated in the emissive reg ion, is simply the percentage of light emitted into the external modes (modes i in Fig ure 1 ( a ) ), which has been estimated to be approximately 20% assuming isotropic emission inside the OLED [108,112] There are also surface plasmon modes gen erated at the metal/organic interface, which are resulted from the coupling between the free electrons at the metal surface and the electromagnetic radiation and further reduce the percentage of emission in the external modes [113,114] Method s to alleviate this surface plasmon quenching have been investigated by several groups, primarily by increasing the distance between the emissive layer and cathode [115] or by roughening the metal organic interface [80,116]

PAGE 91

91 Figure 6 4 Bottom and top emitting OLEDs. A ) Schematic device structure of a bottom emitting OLED with three major emission modes B ) Schematic dev ice structure of a top emitting OLED with two major emission C ) Simulation of l ight extraction efficiency, LE of a top emitting OLED with a close packed, hemispherical, transparent microlens By modifying the glass/air interface, such as attaching polyme r microlens arrays [30,37,38,110] or macrolenses [26,35] or introducing light scattering centers [109] a portion of the substrate waveguiding modes can be extracted in addition to the external modes. Microlens arrays are preferred over macrolenses as they are a more practical solution for large area OLEDs and do not signficantly a lter the thin form factor of the devices (which is mostly dictated by the thickness of the substrate, typically around 1

PAGE 92

92 mm). The ITO/organic waveguiding modes, however, remain untapped, as they are separated from the glass/air interface by the substrate t hickness. When a high index of refraction ( n 1.8) substrate is used instead of the more common, lower index soda lime glass ( n 1.5), the ITO/organic waveguiding modes can be effectively coupled into the substrate due to the matching of the refractive indices. In this case, an index matching hemisphere on the substrate surface can be used to significantly enhance LE by up to 140% [29] However, these high index substrates are too costly for use in large area panels, and other much lower cost solutions are still needed. TOLED ) architecture has been considered to be more advantageous than the bo ttom emitting devices for integration with active matrix displays [95] and as will be shown here, it is also more advantageous for possi ble light extraction enhancement. As shown in Figure 6 2 ( b ) with a reflecting metal electrode on the substrate and light emission through the top transparent electrode, the substrate waveguiding modes are eliminated, and the waveguiding modes in the organ ic and transparent electrode layers (mode ii in Figure 1b) are now accessible by modifications, such as attachment of a microlens array, on the light emitting surface (the transparent electrode/air interface). 6.4 Simulation of Microlens Enhancement on To p OLEDs Optical modeling was used to study the impact of attaching a transparent microlens array (MLA) onto TOLED s (see inset of Figure 6 4 ( c ) ) on light extraction. Both Monte Carlo based ray tracing and finite difference time domain (FDTD) wave optics si mulations have been carried out, although neither method is expected to be fully accurate due to various assumptions. The total thickness of the organic multilayer

PAGE 93

93 stack in an OLED (~100 nm) is significantly less than the optical wavelength, which limits the applicability of the ray optics. On the other hand, the microlens size was set to be 2 m in the FDTD simulations (instead of 100 m in actual experiments) due to the simulation constraints. This leads to significant wavelength dependence in the calc ulated light extraction efficiency due to increased interference effect with small microlenses. Nevertheless, as shown in Figure 6 4 ( c ) while there are discrepancies between the two methods especially in the high index region, both methods predict that LE increases with the index of refraction for the microlens material, n lens for n lens n org (which is 1.7 in this calculation). This is anticipated as a higher index of refraction for the microlens material allows a larger portion of the waveguiding mo des in the organic/transparent electrode layers to enter the microlenses and be subsequently extracted. Overall, maximum LE in the range of 70 80% is predicted when high index transparent materials are used for the MLA. 6.5 Bottom Emitting Devices with Microlens Arrays While the inevitable goal of this process is to produce top emitting devices with microlens arrays it is first prudent to examine the effect of the arrays on the standard bottom emitting device [30,91] Figure 6 5 shows the effect o f 100 micron diameter lens arrays on small 2 mm by 2 mm green devices showing both the external quantum efficiency of bare and lens added devices normalized to the maximum EQE measured as well as the enhancement factor seen between the two.

PAGE 94

94 Figure 6 5 Normalized EQE and e nhancement factor of BOLEDs with added microlens arrays [91] Here it can be seen that with the additional light extracted from the lenses, an enhancement factor of f=1.35 + .04 Compared to previous devices with lens array s (f= 1.5 1.7) [38,90] this enhancement is decidedly low. One possible explanation for this reduction in enhancement is due to internal reflections within the OLED architecture. Devices with a large cathode measuring 1 cm by 1 cm were also fab ricated and measured using both and integrating sphere and confirmed using a spectrometer The enhancement due to the lens area in these devices easily reaches f=1.70 + .07, more closely matching other reported values. Figure 6 6 illustrates the enhanceme nt in the emission spectrum as well as the proposed explanation for this increased enhancement.

PAGE 95

95 Figure 6 6 Microlens enhancement of large area OLEDs A) EL spectra with and without microlens. B) Light path in large cathode device. C) Light path in smal l cathode device [91] By expanding the reflective cathode, light previously back reflected into the device as seen in fig. 6 6(c) are n ow able to reflected forward again and possibly extracted to air as in fig. 6 6(b) In summary, 70% enhancement of extracted light is possible through the application of molded microlens arrays to bottom emitting devices which shows comparable enhanceme nt to other methods of microlens deposition. While this enhancement is significant, the simulations for top emitting devices with this structure show the potential for even greater extraction. Therefore the next obvious step is te st the actual capability of top emitting devices with microlens arrays.

PAGE 96

96 6.5 Large Area TOLEDs with MLA Figure 6 7 Effect of a microlens array on light extraction in top emitting green OLED. A ) Schematic illustration of the device configuration (top) and scanning electron mic rograph of the microlens array (bottom, scale bar = 200 m). B ) Comparison of the electroluminescence (EL) spectra of the green OLED in the bare region and in the MLA cov ered region (six spots each). C ) Line scan of the integrated EL intensity across the two regions as illustrated in A ). It has been previously demonstrated that the fabrication of large area (several square inches) arrays of close packed, hemispherical microlenses from Norland optical adhesives (NOA) using a soft l ithography process, and a chieve a 50 70% enhancement

PAGE 97

97 in LE when such MLAs are applied to BE OLEDs [30] Here a 1 cm 2 area, green emitting TOLED based on the structure investigated previously [58] with the phosphorescent fac tris (phenylpyridine) iridium [Ir(ppy) 3 ] [106] as the emitter was fabricated Half of the device emitting surface was covered with a MLA ( n lens = 1.5 6 ) as schematically illustrated in Figure 6 7 ( a ) A scanning electron micrograph (SEM) of the MLA is also shown in Figure 6 7 ( a ) The diameter of the microlenses is 100 m, and close packed arrays of hemispherical shaped microlens can be readily obtained over several square inches [30] Figure 6 7 ( b ) shows the electroluminescence (EL) spectra taken at multiple spots from both the bare and MLA covered regions of the OLED operated at a constant current (see Methods). The nonunif ormity in the OLED leads to a 4 7% variation in the peak or integrated EL intensity among the six spots from either region. On average, the enhancement factor for the EL intensity with the use of the MLA, defined as the ratio of the intensity for the MLA c overed surface region to that for the bare region, is f = 2.3 0.2 based on the peak EL intensity and f = 2.4 0.2 based on the integrated EL intensity. The normalized EL spectra show negligible shift in the peak emission wavelength, although the should er peaks at 540 nm becomes slightly more prominent after the attachment of the MLA. Figure 6 7 ( c ) shows the integrated EL intensity as the optical fiber was scanned across the device surface as illustrated by the arrow in Figure 6 7 ( a ) The two plat eaus in the plot in Figure 6 7 ( c ) correspond to the bare and MLA covered regions, and an enhancement factor of f 2.3 is obtained (EL data in the transition zone are not reliable due to the unavoidable accumulation of microlens material jus t outside the e dge of the MLA).

PAGE 98

98 6.6 Wavelength and Angular Dependence of MLA Figure 6 8 Wavelength dependence of the light extraction enhancement induced by the microlens array. A ) Comparison of the electroluminescence (EL) spectra of a blue green OLED in the bare r egion and in the MLA covered region (three spots each). B ) The MLA induced light extraction enhancement factor, f as a function of the wavelength, for the green and blue green devices, and the transmittance, T of the ITO and MoO x /Au/MoO x trilayers (th e intermediate Au layer thickness is 10 nm for trilayer 1 and 20 nm for trilayer 2). The same MLA structures have also been applied to OLEDs emitting in different difluorophenylpyridinato)t etrakis(1 pyrazolyl)borate ( FIr6 ) [22,23,25] into the green emitting TOLED a broadened emission sp ectrum from blue to green, which enables the possibility to probe possible wavelength dependence of the light extraction efficiency. Shown in Figure 6 8 ( a ) are the comparisons of the EL spectra in the bare and MLA covered regions of a blue green emitting OLED. The integrated EL intensity for the blue green device shows a 2.2 0.2 fold increase with the MLA, similar to that for the green emitting device discussed above. Somewhat different from the green OLED discussed above, the spectral change after the attachment of the MLA is more

PAGE 99

99 appreciable for the blue green device. While the peak at = 458 nm is approximately 20% more intense than the peak at = 491 nm without the MLA, the intensities of the two peaks are approximately equal after the MLA is atta ched. This is further illustrated in Figure 6 8 ( b ) in which the enhancement in the EL intensity at each wavelength has been plotted as a function of the wavelength for both the green and blue green devices. For both devices, the enhancement factor reache s a maximum at 530 540 nm, f max = 2.6 0.2 for the green device and f max = 2.4 0.2 for the blue green device, and decreases towards both shorter and longer wavelengths. As the emission spectrum of the green device mostly occupies the peak enhancemen t region (hence relatively small variation in f ), there is negligle change in the EL spectrum; however, the broader EL spectrum for the blue green device and the significantly lower f in the blue region lead to an appreciable red shift in the EL spectrum u pon the attachment of the MLA. Such wavelength dependence in enhancement is mostly attributed to the transmittance of the transparent electrode used for these TOLED s. ITO commercially pre deposited on glass substrates has been commonly used as the trans parent conductor for BE OLEDs. While methods have been developed to deposit ITO during the rf sputtering process on organic layers without significantly damaging the organics [117,118] MoO x /Au/MoO x trilayer structure [58] as the transparent electrode for the TOLED s was used here based on previous experience which is fully compatible with organic material deposition in high va cuum. As shown in Figure 3 ( b ) the transmittance of such trilayer structure (trilayer 1, with a 10 nm thick intermediate Au layer) is above 80% for 550 nm 700 nm, approaching that of the commercial ITO film. The trilayer transmittance decreases with a 20 nm thick Au layer (trilayer 2), especially in the long

PAGE 100

100 wavelength region, thus leading to maximum transparency at 550 nm to 600 nm. While trilayer 1 is used in the TOLED s here, the waveguided OLED emission may travel through the trilayer electrode m ore than once before it could escape the device. Consequently, the relatively low transmittance of the trilayer at < 500 nm and > 600 nm significantly reduces the EL intensity in these wavelength regions, leading to less efficient extraction enhancemen t. The maximum enhancement factors ( f max = 2.4 or 2.6) obtained at intermediate wavelengths ( ~ 550 nm) are therefore considered to be more indicative of the actual potential of the MLA approach discussed here. If a more transparent electrode such as ITO is used instead of the trilayer, provided that no damage to the organic layers occurs during electrode deposition, a less wavelength dependent and overall a higher enhancement in light extraction across the entire visible spectrum is expected

PAGE 101

101 6.7 White OL EDs with Lens Enhancement Figure 6 9 Performance of a top emitting white OLED with a microlens array. A ) External quantum efficiency, EQE as a function of the current density, J for the triple doped, top emitting white OLED with (w/) or without (w/o ) a microlens array attached to the light emitting surface. B ) Power efficiency, P as a function of the luminance, L for the devices.

PAGE 102

102 The MLA structure was further applied to white OLEDs employing a dual emissive layer structure containing three phospho rescent dopants [26] As shown in Figure 6 9 ( a ) and ( b ) a maximum external quantum efficiency (EQE) of EQE = (20 2)% and a maximum p ower efficiency of P = (46 3) lm/W were achieved for the TOLED which are slightly higher than those of the BE OLED (maximum EQE = 18% and P = 40 lm/W ) [26] Both devices show significant efficiency roll off at higher current densities ( J ) or luminances ( L ), although it is more severe for the TOLED The efficiency roll off behavior is common for phosphorescent OLEDs due to processes such as triplet triplet annihilation [119] However the likely unbalanced charge injection in the non optimized TOLED may have contrib uted further to the efficiency roll off. With an enhancement factor of f = 2.2 0.2, a maximum EQE = 43% can be achieved for the white OLED after the attachment of the MLA. As P decreases with the increase of luminance, the MLA induced enhancement in P at a given luminance becomes even higher. For example, at L = 100 cd/m 2 P was increased from 28 lm/W to 68 lm/W, corresponding to a 2.4 fold increase; at L = 1,000 cd/m 2 P was increased from 17 lm/W to 45 lm/W, corresponding to a 2.6 fold increase. 6.8 Summary In summary a method to increase the light extraction efficiency in OLEDs by up to 2.6 times has been demonstrated As presented in chapter 4, this method is comparable to other light extraction techniques in terms of degree of enhancement and provides a simple means of producing a high efficiency OLED. The integration of close packed, hemispherical, transparent microlens arrays on the light emitting surface of top emitting OLEDs serves to effectively extract the waveguiding modes in the trans parent

PAGE 103

103 electrode and organic layers, without appreciably altering the thin form factor of the devices. Optical simulations predict even higher light extraction enhancement could be achieved when transparent materials with higher index of refraction are use d to fabricate the microlenses. The soft lithography approach used to fabricate the microlens array is simple and low cost, and can be readily scaled up for larger size arrays, making the overall approach compatible w ith low cost fabrication of large area OLED panels, especially for solid state lighting applications.

PAGE 104

104 CHAPTER 7 ALTERNATIVE OLED ARC HITECTURES FOR IMPRO VED EFFICIENCY 7.1 Introduction Methods to improve the efficiency of OLED devices generally follow a natural progression of finding a new mate rial, integrating it into standard device structure, and finally optimizing the device stack for maximum efficiency. While this method has proven to be a useful tool in pushing the progress of OLEDs forward, potential new structures or new ways of generat ing light are often ignored. In this section, alternative OLED structures with unique properties will be examined. First, high efficiency blue/green OLEDs will be examined to determine the reasons behind the high external quantum efficiency and low roll off of these devices. Second, fluorescent doped microlens arrays will examined as a method of both improving device efficiency and reducing OLED fabrication requirements. 7.2 Experimental Methods OLED active layers were deposited by thermal evaporation un der a pressure of 3.0 x 10 6 Torr or less. The substrates were cleaned by submerging the samples in beakers of detergent and water, de ionized water, acetone, and isopropanol successively and each beaker was ultra sonicated for a period of 15 minutes. De vices were deposited on a precoated ITO electrode for standard bottom emission architecture. Device structures consisted of 40 nm thick 1,1 bis (di 4 dicarbazolyl 3,5 benzene (mCP) doped with fac tris(2 phenylpyridine) iridium (Ir(ppy) 3 ) green and difluorophenylpyridinato)tetrakis(1 pyrazolyl)borate ( FIr6 ) blue phosphorescent dopants as the emitting layer and 40 nm thick t ris[3 (3

PAGE 105

105 pyridyl)mesityl]borane (3TPYMB) for the electron transport layer. A 1nm CsCO 3 interlayer followed by 80nm Al were used as the cathode. T he previously described soft lithography process to fabricate microlens arrays on OLEDs [30] was used with additional steps Norland optical adhesive (NOA) was used for the transparent microlenses (with an index of refraction of 1.5 6 ) with the addition of organic fluoresecent dye which had been dissolved in solvent and mixed with the NOA The solvent was allowed to evaporate in a vacuum environment and leaving behind the NOA precursor and dye material. A p olydimethylsiloxane (PDMS) concave template produced using a close packed monolayer of polystyrene microspheres (diameter = 100 m), was placed on top of the liquid NOA on the TOLED surface. After polymerization of the NOA with UV illumination (365 nm, 200 mW/cm 2 ), the PDMS template was pealed away and all that remains was a microlens array Current density voltage ( J V ) characteristics of OLEDs were measured under ambient conditions using an Agilent 4155C semiconductor parameter analyzer The Newport 818 UV silicon diode was placed in close proximity to measure the OLED light output, an d the luminance was calibrated using a Minolta luminance meter (LS 100). The EL spectra were taken using an Ocean Optics Jaz spectrometer coupled to an optical fiber. 7.3 Blue/ Green Bottom OLEDs with Near Maximum External Quantum Efficiency The classic an alysis of OLEDs shows a maximum approximately 20% (although closer to 22%) amount of light generated that can escape the device. Current simulations and experimental results seem to indicate this to be a gross underestimation. Theoretical limits of nearl y 30% appear to possible using an optimized

PAGE 106

106 bottom OLED design [78] Questions still remain as to how the typical 20% barrier can be ove rcome. For example, through the use of high triplet energy transport and emitting layers, an optimized green phosphorescent device can reach the traditional threshold. Of note though, is the effect that adding blue phosphorescent dopant contributes to th is system. Figure 7 1 Blue g reen d evice features A ) T he device energy diagram B ) C omparison of EQE with optimized green and blue green dopants Figure 7 1 shows the difference between an optimized green device architecture (8 wt% doping) and a device incorporating both blue and green dopants (25 wt% blue and 0.5% green) The high quantum efficiency as well as gradual roll off suggest a highly balanced charge injection and recombination and an upper threshold greater than 22% for these devices. The question must then be asked, from where does this enhancement arise? Further investigation of this structure provides some clues into the necessary requirements for greater conventional OLED efficiency.

PAGE 107

107 Figure 7 2. External quantum efficiencies of devic es with varied hole and electron transport materials First an investigation of the green device structure can help illustrate some of the properties leading to conventional maximum efficiency. Figure 7 2 shows first the improvement in efficiency as the tr ansport materials are changed from traditional materials N,N' bis(naphthalen 1 yl) N,N' bis(phenyl) benzidine ( NPB ) and 2,9 dimethyl 4,7 diphenyl 1,10 phenanthroline ( BCP ) as hole and electron transport materials respectively to TAPC and 3TPYMB in mCP devi ces with 8 wt% green doping As has been presented previously with relation to blue doped devices [23,24] replacing even one of these transport materials with its high triplet counterpart improved charge injection and exciton confinement and a llows more efficient recombination. The combination of both high triplet materials allows the green device to reach 22% external

PAGE 108

108 quantum efficiency. Therefore it is clear that exciton confinement is necessary in order to maximize the internal quantum eff iciency of the device. Figure 7 3 Comparison of host materials CBP and mCP for use in green OLEDs based on external quantum and power efficiency. While the assumption so far has been to use mCP as the host material, most common high host material chan ges from the traditional 4,4' bis(carbazol 9 yl)biphenyl (CBP) host to the mCP host with 10 wt% doping in the emitting layer (figure 7 3) Both devices are fabricated using an added mCP exciton blocking layer as presented in literature [22] While the current injection of the CBP devices is higher, the efficiency of the mCP based device is far higher than the CBP counterpart. As has b een suggested in literature [120] this likely due to the reduction in aggregation of dopants within the host material due to the impr oved solvation of mCP molecules. This in turn leads to higher photoluminescent quantum yield as excitons are more efficiently transferred from

PAGE 109

109 host to dopant, as well as reducing triplet triplet annihilation and triplet polaron annihilation. Furthermore, the improved exciton confinement due to the mCP blocking layer, display an even higher maximum EQE than the previous presented devices. From this test two more critical features are seen in the use of an exciton blocking layer and host with reduced aggre gation. Figure 7 4 Comparison of green (G), blue (B) and blue green devices with various doping concentrations Next a comparison of the green device, the blue green and a blue device with changing doping concentrations is used to illustrate some intere sting features (figure 7 4) First, a blue only device doped 15 wt% shows relatively low efficiency compared to

PAGE 110

110 the other devices present. This low efficiency is confirmed by previous reports of blue OLEDs [121] What is more significant is the increased efficiency low doping levels of green only devices with .7 wt% these devices show high EQE in spite of the lack of doping molecules o n which to recombine. The more rapid decline in efficiency suggests that this efficiency is limited by the charge density in the EML where with greater amount of charges and less dopants, annihilation is likely to occur. Furthermore, while the blue dopant is fixed, the green dopant has little effect on the EQE but both devices exceed the efficiencies of the single dopant devices. This combination of dopants th erefore must somehow improve transfer of excitons or charges in devices. Figure 7 5. Blue gre en device structures showing both separation of blue dopant from ETL and removal of the mCP exciton blocking layer Previous examinations of blue dopants [22,68] show that electron transfer directly onto the doping molecules from the electron tran sport layer is not only possible but

PAGE 111

111 necessary for exciton formation in the wide band gap host materials. Figure 7 5 demonstrates that this property is also true of the blue green doped device. By placing even 2 nm of only green doped mCP between the blu e doped region and the ETL the efficiency drops to by nearly half (14% EQE) Clearly for electron transfer and exciton formation, electrons must inject directly into the FIr6. The LUMO level of the dopant at 3.1 eV facilitates more efficient charge trans fer into the EML than into mCP (LUMO=2.4 eV) from 3TPYMB (LUMO=3.3 eV). Thus more of electrons injected are able to efficiently transfer into the EML instead of being trapped at the interface. Furthermore on the opposing side of the EML, the removal of t he mCP exciton blocking shows a clear drop in efficiency to only 22% EQE. This is likely due to even further prevention of green triplets reaching the hole transport layer and perhaps acting as a necessary intermediary for exciton formation on blue dopant molecules with holes transporting to mCP first before reaching the blue dopant. Finally, an examination of the emission spectrum of the blue gr een device shows on e final feature that may explain the high efficiency of these devices. Figure 7 6 shows the dependency of the emission spectrum on increasing current injection. As can be seen, the blue FIr6 emission peak at 460 nm increases with increased amount of charges. This suggests that as previously shown, the green dopant is less capable of handling t he increased number of charges injected. Due to the high density of blue dopant molecules and the propensity for charges to directly transfer to this dopant before transferring as excitons to the green molecules, these extra charges that would once have non radiatively recombine d in the host molecule now occurs on a phosphorescent dopant extacting light in the process.

PAGE 112

112 Figure 7 6. Current dependent EL spectra of blue green device (B:15 wt% and G:.35 wt%). From this analysis of the blue green OLED, it i s clear that while charge transfer, and exciton confinement are necessary to improve efficiency of devices, a single pathway limits the means of charge and exciton transfer whereas through a multiple path system whereby the charge transport is optimized th rough different materials and excitons have both a large density of emitting sites as well as low energy sites to efficiently to transfer to and recombine on, the device can truly maximize its internal quantum efficiency. This phenomenon has been seen in part using mixed host systems in order to better balance charge transport in the emitting layer [122] These devices show the same lo w roll off in efficiency as the blue green mixed systems, indicating that

PAGE 113

113 perhaps ambipolar transport in the EML is one of the final necessary components to maximize OLED efficiency. 7.4 Down Conversion Red Microlens Arrays Figure 7 7. Current injection of equally doped red (PQIr), green (Ir(ppy) 3 ), and blue (FIr6 ) OLEDs. In order to produce desirable white light devices that out compete fluorescent tubes, OLED light sources must broad emission spectra, or at the very least, a high color rendering index that can display blue, as well as green, and even so far as red must be created. While the commonly accepted method of producing high color rendering devices requires the use of three dopants added to the emissive layer of a device [26] this method requires precise control of the doping layers, specifically the red doping layer in order to avoid quenching to red emission, reduction in ef ficiency due to charge trapping, and reduced charge injection, leading to low power efficiencies

PAGE 114

114 As red dopants provide significant challenges to device efficiency, perhaps another method is necessary to accomplish the production of white light. P revio us work has shown [87] that down conversion devices with an outer layer that absorbs light emitted from the device and emits light at a longer wavelength may provide a provide a promising avenue Alone, this method may prevent some of the difficulties encountered by the use of red dopant in a device, but unless photoluminescent efficiency of the down conversion layer is 100%, the external quantum efficiency will be reduced. Therefore, further steps are necessary to produce both white light and a high efficiency de vice. One possible method is to combine the properties of a down conversion outer film with the molded qualities of a microlen s array. Presented here is an investigation of the viability of this technique. The first ste p in this process is to select the appropriate method of down conversion. Due to the need for blue and green light as well as red light, as suspension of a dye within a transparent host was selected to prevent complete absorption of the emitted light from within the device. For the host material, the natural choice is to use the NOA 68 UV cureable lens material previously used to fabricate devices in order to p rovide index matching and forming of lenses With the host material set the next variable to investigate is the red emitting material desireable. Three materials were examined for this purpose: iridium(III) bis (2 phenylquinolyl N,C 2 ) acetylacetonate ( PQIr ) the traditional red phosphorescent dopant used in the EML [26] ; 5,6,11,12 tetraphenylnaphthacene (rubrene), a common orange/ red fluorescent dye [123] ; and 4 ( dicyanomethylene) 2 t butyl 6(1,1,7,7 tetramethyljulolidyl 9 enyl) 4H pyran (DCJTB) [124,125] a deep red dopant also used in fluorescent applications. In order to

PAGE 115

115 suspend the doping material in the host, .7 .8 grams of the NOA were dissolved in 1 ml toluene while the dyes were dissolved in chloroform with an optimized amount of 10mg/1ml in order to prevent aggregation. The two mixtures were then combined and placed in a vacuum oven to evaporate the solvent from the system. The resultant mixture was then cured in a lens mold using the methods presented in the previous chapter. The films were the n applied to the optimized blue green device structure. Figure 7 8 shows a comparison of the emission spectra of the B G device, and with either rubrene or PQIr doped microlens array captured using a spectrometer Figure 7 8 Comparison of the emission spectra of the B G device, and with either rubrene or PQIr doped microlens array From this graph it is clear that PQIr makes a poor choice for an emitting dopant. This problem is likely explained due to the absorption peak of PQIr occurring below 400 nm

PAGE 116

116 ( 279 nm for films and 344 dissolved in tetrahydrafuran). Conversely, the rubrene doped lenses do show a promising broad spectrum emission with peaks in th e blue, green and red regions (CIE coordinates 39,47), but with a color rendering index of only 51. T he low color rendering is likely due to the emission peak of rubrene at 560 nm which limits the red contribution as the emission quickly decays before 600 nm [126] While this demonstrates that such a device is possible, the limited red emission makes for a poor white light device due to its low color rendering. For this reason, DCJTB was considered as an alternative due to its emission peak at 600 nm [125] Figure 7 9 illustrates the difference in emission spectra between the DCJTB and rubrene lens films. Figure 7 9 Comparison of white light emission spectra with rubrene and DCJTB doping with inset of an illuminated device with half the area covered in with DCJTB microlens.

PAGE 117

117 Of note are both the broader emission of the DCJTB doped lens array and additionally the im proved balance of the peaks as well (CIE 38, 38). This leads to a CRI of 8 2 a distinct improvement in color rendering and even exceeds the value for the previous triple doped emitting layer device (79) [26] Now that an ideal emission pattern has been achieved, device performance becomes a concern. Figure 7 10 shows the luminance of the blue green device with half the device cover ed by the microlens array measured using a luminance meter. From this measurement it is clear that in spite of the addition of the MLA luminance of the device was severely hindered. Upon examination of the photoluminescent quantum yield of a solution of chloroform dissolved DCJTB, the yield of this solution shows only 35% compared to 52% for rubrene using an excitation wavelength of 510 nm (optimized to peak DCJTB absorbtion) While solution QY efficiency is not the same as the dopant in a solid film, t he solid films tend to have even lower efficiencies if the host material is incapable of transferring excitons As result, while the CRI of the device is well balanced, much of the emitted light is sacrificed in order to produce the red emission.

PAGE 118

118 Figu re 7 10 Large area blue green OLED luminance and luminance with attached DCJTB lens array 7.5 Summary Here a new device structure for a blue green OLED was investigated after showing near maximum external quantum efficiency (28%). By examing the variou s components of this device structure, it is clear that a multiple path approach whereby more balanced charges are injected, excitons are confined and site s for recombination are widely ava ilable leads to maximum efficiency in OLED devices. Furthermore, b y doping red dopants into microlens arrays, white light devices with high color rendering index are possible. Currently, these doped lens arrays do not show enhancement due to the poor PL quantum yield of the dopants used.

PAGE 119

119 CHAPTER 8 CONCLUSIONS 8.1 Oxi de / Metal/ Oxide Transparent Electrodes The prospect of finding an easily fabricated transparent electrode that can replace the industry standard is certainly a daunting task. The ideal replacement must possess high transmittance, high c onductivity, eas y patternability, and be cost effective. When compared these criteria, the multilayer structure, w hile its transparency has a maximum transmittance up to 85 90% this transparency is highly wavelength dependent and the maximum wavelength must be tuned base d on the metal material, thus reducing its usefulness. The sheet resistance is lower than typical commercial ITO electrodes and a simple shadow mask can be used to define the electrode area When multilayer structures were used as the anode in a T OLED, th e emission spectr um was comparable to that of the ITO counterpart The external quantum efficiencies were shown to be the same as the bottom emitting devices; however the severe charge imbalance leads to higher operating voltages and lower power efficien cies in the TOLEDs This unbalanced nature therefore necessitates the use of hole injecting layers in order to produce comparable charge balance to conventional device structures. While the stack does show some results comparable to those of a standard de vice, overall it is still lacking, and therefore the uniqueness of a n oxide /metal/ oxide electrode is perhaps best left as an academic curiosity. 8.2 Light Extraction Enhancement in TOLEDs The use of outcoupling enhancements to improve OLED efficiency is a promising method of overcoming some of the inherent barriers in OLED design. Perhaps through this route, commercial OLED products will begin to exceed current lighting technology.

PAGE 120

120 In Chapter 6 a method was demonstrated t hat increase s the light extract ion efficiency in OLEDs by up to 2.6 times making the process comparable to other outcoupling enhancement methods Th rough the use of close packed, hemispherical, transparent microlens arrays applied directly to the light emitting surface of top emitting O LEDs more light is effectively extract ed from the waveguiding modes that reside in the device architecture all the while without significantly increasing the form the device The simple nature of the processing for the lens array means that its applicati on would not be a cost or process limiting factor for use in commercial OLEDs. This enhancement appears to be useful for both single emission and broad emission so long as the transparent electrode used is not very wavelength selective. As current high i ndex materials used for the lens array are limited, the enhancements shown are lower compared to the maximum enhancements from the optical simulation predict ions. It seems possible that through further development of cureable small molecule The use of a c heap, easy to handle approach to apply these lenses makes for promising new step in OLED design 8.3 Alternative OLED Architectures In Chapter 7, means of improving device efficiency in OLEDs through the use of unusual device architectural changes were e xamined. A blue green doped OLED has been shown to maximize external quantum efficiency to nearly a 100% internal quantum efficiency limit. By examining the interactions between the various components of the high efficiency device, a better understanding of needs of OLED design has been achieved. The improved charge transport, exciton confinement and availability of recombination sites allows for a unique system in which the interaction between multiple emitters lead to a device more efficient than the s um of its parts. Furthermore, the potential for greater efficiencies for use in white OLEDs was examined through a

PAGE 121

121 combined down conversion / microlens array. While no specific dopant was ideal, it is clear that the potential for a high color rendering h igh extraction can exist. Through such a method, a device capable of competing with fluorescent tubes might be possible. 8.4 Future Work 8.4.1 Quantum Dot LEDs While the work presented here was designed in conjunction with OLEDs, techniques are not limite d simply to organic devices. While OLED technology is reaching a commercially accessible point, the technology still has many limitations, some inherent prevent future growth in the field. The susceptibility to environmental conditions requires that all devices have some form of encapsulation in order to preserve device operation for more than a few hours. This encapsulation therefore limits how the device itself can be interacted with and what types modifications (like a top emitting device with directl y attached microlens array) can be used to improve device efficiency. Additionally, the most efficient OLEDs are small molecule based and rely on vacuum thermal evaporation or other high vacuum techniques in order to produce the desired properties. This further limits how cost effective this technology can be and therefore its ability to supplant conventional incandescent lighting. One area of research that is beginning to garner attention is the use of solution processed inorganic nanoparticles to gene rate light instead [11,127 129] The particles, due to their relation to traditional semiconductors allow for much high processing temperatures and much weaker susceptibility to environmental and chemical changes that may occur in the device. By controlling the size of these molecules, the emission wavelength is highly tuneable in a manner similar to organic materials. The techniques presented here: the use of a top emitting architecture, an d the application of

PAGE 122

122 microlens arrays are just as applicable to this technology as to the organic devices. While the energetics of these quantum dot LEDs is still in the process of being understood, this technology could present a long term alternative to organic technologies.

PAGE 123

123 APPENDIX A LIST OF PUBLICATIONS [1] E. Wrzesniewski, S. H. Eom, W. T. Hammond, emitting organic light emitting Journal of Photonics for Energy 1 011023 (2011) [doi:10.1117/1.3592886]. [2] E. Wrzesniewski, S. H. Eom, W. Cao, W. T. Hammond S. Lee, E. P. Douglas, Emitting Organic Light Emitting Small n/a n/a (2012) [doi:10.1002/smll.201102662]. [3] S. H. Eom, Y. Zheng, E. Wrzesniews ki, J. Lee, N. Chopra, F. So, and J. Xue, emitting devices with dual triple doped Applied Physics Letters 94 153303 153303 3 (2009) [doi:doi:10.1063/1.3120276]. [4] S. H. Eom, Y. Zheng, E. Wrzesniews ki, J. Lee, N. Chopra, F. So, and J. Xue, blue phosphorescent organic light Organic Electronics 10 686 691 (2009) [doi:10.1016/j.orgel.2009.03.002]. [5] S. H. packed hemispherical microlens arrays for light extraction enhancement in organic light Organic Electronics 12 472 476 (2011) [doi:10.1016/j.orgel.2010.12.021]. [6] S. H. Eom, E. Wrzesniewski, light emitting devices via hemispherical microlens arrays fabricated by soft Journal of Photonics for Energy 1 011002 (2011) [doi:10.1117/1.3528267]. [7] W. Cao, J. D. Myers, Y. Zheng, W. T. Hammond, E. Wrzesniewski, and J. Xue, Applied Physics Letters 99 023306 023306 3 (2011) [doi:doi:10.1063/1.3609870]. [8] J. Xue, S. H. Eom, Y. Zheng, E. Wrzesniewski N. Chopra, J. Lee, and F. So, Proceedings of SPIE 7415 741511 741511 10 (2009) [doi:doi:10.1117/12.829598]. [9] Organic Electronics [doi:10.1016/j.orgel.2012.05.047].

PAGE 124

124 REFERENCES [1] ricity Generating Energies 3 462 591 (2010) [doi:10.3390/en3030462]. [2] An effective way of Renewable Energy 34 2414 2421 (2009) [doi:10.1016/j.renene.2009.02.018]. [3] Department of Energy (2012). [4] Prepared by Navigant Con sulting for the US Department of Energy. Office of Energy Efficiency and Renewable Energy Building Technologies Program (2002). [5] J. M. Phillips, P. E. Burrows, R. F. Daves, J. A. Simmons, G. G. Malliaras, F. So, J. A. Misewich, A. V. Nurmikko, and D. Basic Energy Sciences report (US Department of Energy, 2006), http://www. s c. doe. gov/bes/reports/files/SSL_rpt. pdf (2006). [6] in White Organic Light Advanced Materials 22 572 582 (2010) [doi:10.1002/adma.200902148]. [7] C. S. Chang, S. J. Chang, Y. K. Su, Y. C. Lin, Y. P. Hsu, S. C. Shei, S. C. Chen, GaN light emitting diodes with ITO p contact Semiconductor Science and Technology 18 L21 L23 (2003) [doi:10.1088/0268 1242/18/4/102]. [8] Luminosity Blue and Blue Green Gallium Nitri de Light Science 267 51 55 (1995) [doi:10.1126/science.267.5194.51]. [9] H. S. Kim, F. Lugo, S. J. Pearton, D. P. Norton, Y. L. Wang, and F. Ren, Appl ied Physics Letters 92 112108 (2008). [10] emitting diodes made Nature 370 354 357 (1994). [11] K. R. Choudhury, D. W. Song, a processed hybrid polymer nanocrystal near infrared light Organic Electronics 11 23 28 (2010) [doi:10.1016/j.orgel.2009.09.017].

PAGE 125

125 [12] Emitting Diodes Semiconductors and Semimetals Volume 48 pp. 149 C3, Elsevier (1997). [13] Printing Advanced Materials 22 n/a n/a (2010) [doi:10.1002/adma.201090011]. [14] K. X. Steirer, M. O. Reese, B. L. Rupert, N. Kopidakis, D. C. Olson, R. T. Collins, Solar Energy Materials and Solar Cells 9 3 447 453 (2009) [doi:10.1016/j.solmat.2008.10.026]. [15] to roll processed polymer solar cells free from indium tin Organic Electronics 10 761 768 (2009) [doi:10.1016/j.orgel.2009.03.009 ]. [16] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [17] Appl. Phys. Lett. 51 913 (1987) [doi:10.1063/1.98799]. [18] C. W Journal of Applied Physics 65 3610 3616 (1989) [doi:doi:10.1063/1.343409]. [19] organic e Appl. Phys. Lett. 70 152 (1997) [doi:10.1063/1.118344]. [20] on from organic Nature 395 151 154 (1998) [doi:10.1038/25954]. [21] internal phosphorescence efficiency in an organic light Journal of Applied Physics 90 5048 5051 (2001) [doi:doi:10.1063/1.1409582]. [22] blue organic electrophosphorescence by guest Applied Physics Lette rs 83 3818 3820 (2003) [doi:doi:10.1063/1.1624639].

PAGE 126

126 [23] Y. Zheng, S. blue phosphorescent organic light emitting device with improved electron and exciton Appl. Phys. Lett. 92 223301 (2008) [doi:10.1063/1.2937403]. [24] S. H. Eom, Y. Zheng, E. Wrzesniewski, J. Lee, N. Chopra, F. So, and J. Xue, blue phosphorescent organic light Organ ic Electronics 10 686 691 (2009) [doi:10.1016/j.orgel.2009.03.002]. [25] S. very high efficiency deep blue phosphorescent organic light Appl. Phys. Lett. 93 133 309 (2008) [doi:10.1063/1.2996274]. [26] S. H. Eom, Y. Zheng, E. Wrzesniewski, J. Lee, N. Chopra, F. So, and J. Xue, emitting devices with dual triple doped Applied Physics Letters 94 153303 153303 3 (2009) [doi:doi:10.1063/1.3120276]. [27] Y. Sun, N. C. Giebink, H. Kanno, B. Ma, M. E. Thompson, and S. R. Forrest, Nature 440 908 912 (2006) [doi:10.1038/ nature04645]. [28] performance tandem white organic light emitting diode combining highly effective white Journal of Applied Physics 105 0 76101 076101 3 (2009) [doi:doi:10.1063/1.3106051]. [29] S. Reineke, F. Lindner, G. Schwartz, N. Seidler, K. Walzer, B. Lussem, and K. Nature 459 234 238 (2009) [doi:10.1038/nat ure08003]. [30] S. packed hemispherical microlens arrays for light extraction enhancement in organic light Organic Electronics 12 472 476 (2011) [doi:10.1016/j.orgel.2010.12.021]. [31] E. Wrzesniewski, S. H. Eom, W. Cao, W. T. Hammond, S. Lee, E. P. Douglas, Emitting Organic Light Emitting Small n/a n/a (2012) [doi:10.1002/smll.201102 662]. [32] J. Lim, S. S. Oh, D. Y. Kim, S. H. Cho, I. T. Kim, S. H. Han, H. Takezoe, E. H. coupling factor of microcavity organic light Opt. Express 14 6564 6571 (2 006) [doi:10.1364/OE.14.006564].

PAGE 127

127 [33] Y. J. Lee, S. H. Kim, J. Huh, G. H. Kim, Y. H. Lee, S. H. Cho, Y. C. Kim, and Y. extraction efficiency nanopatterned organic light Applied Physics Letters 82 3779 3781 (2003) [doi:doi :10.1063/1.1577823]. [34] G. Gu, D. Z. Garbuzov, P. E. Burrows, S. Venkatesh, S. R. Forrest, and M. E. external quantum efficiency organic light Opt. Lett. 22 396 398 (1997) [doi:10.1364/OL.22.000396]. [35] C. F. Mad igan, M. efficiency of organic light Applied Physics Letters 76 1650 1652 (2000) [doi:doi:10.1063/1.126124]. [36] T. Tsutsui, M. Yahiro, H. Yoko Coupling Out Efficiency in Organic Light Emitting Devices Using a Thin Silica Advanced Materials 13 1149 1152 (2001) [doi:10.1002/1521 4095(200108)13:15<1149::AID ADMA1149>3.0.CO;2 2]. [37] S. coupling in organic light emitting Journal of Applied Physics 91 3324 3327 (2002) [doi:doi:10.1063/1.1435422]. [38] c oupling of organic light emitting devices using embedded low Nature Photonics 2 483 487 (2008) [doi:10.1038/nphoton.2008.132]. [39] M. Pope and C. E. Swenberg, Electronic processes in organic crystals and polymers Oxford University Press (1999). [40] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [41] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [42] Wikipedi Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [43] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [44] S. Kasap, Principles of Electron ic Materials and Devices 3rd ed., McGraw Hill Science/Engineering/Math (2005). [45] W. Brtting, Ed., Physics of Organic Semiconductors 1st ed., Wiley VCH (2005).

PAGE 128

128 [46] J. D. Wright, Molecular crystals Cambridge University Press, Cambridge; New York (1995). [47] on i norganic semiconductor contact barrier diodes. II. Dependence on organic film and metal Journal of Applied Physics 56 543 551 (1984) [doi:doi:10.1063/1.333944]. [48] igh Purity Molecular Crystals and Liquid Crystals Science and Technology. Section A. Molecular Crystals and Liquid Crystals 355 149 173 (2001) [doi:10.1080/10587250108023659]. [49] Y. Divayana, Electroluminescence in Organic Light Emitting Diodes: Basics, Processes, and Optimizations VDM Verlag (2009). [50] S. Lamansky, P. Djurovich, D. Murphy, F. Abdel Razzaq, H. E. Lee, C. Adachi, Cyclometalated J. Am. Chem. Soc. 123 4304 4312 (2001) [doi:10.1021/ja003693s]. [51] S. Lamansky, P. Djurovich, D. Murphy, F. Abdel Razzaq, R. Kwong, I. Tsyba, M. Inorg. Chem. 40 1704 1711 (2001) [doi:10.1021/ic0008969]. [52] resolution optical sp ectroscopy of bis[2 (2 thienyl)pyridinato Inorg. Chem. 32 3081 3087 (1993) [doi:10.1021/ic00066a019]. [53] W. Y. Wong and C. derived from multi component chromopho Coordination Chemistry Reviews 253 1709 1758 (2009) [doi:10.1016/j.ccr.2009.01.013]. [54] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [55] C. Wu, P. I. Djurovich, and Triplet Exciton Diffusion in Hole Advanced Functional Materials 19 3157 3164 (2009) [doi:10.1002/adfm.200900357]. [56] range resonantly enhanced triplet Nature Materials 5 463 466 (2006) [doi:10.1038/nmat1630].

PAGE 129

129 [57] effects on organic elect Phys. Rev. B 70 205303 (2004) [doi:10.1103/PhysRevB.70.205303]. [58] E. Wrzesniewski, S. emitting organic light emitting Journal of Photonics for Energy 1 011023 (2011) [doi:10.1117/1.3592886]. [59] Processed Full Color Polymer Organic Light Emitting Diode Displays Fabricated Advanced Functional Materials 17 191 200 (2007) [doi:10.1002/adfm.20060065 1]. [60] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [61] J. R. Meyer Appl. Opt. 7 2081 (1968) [doi:10.1 364/AO.7.002081]. [62] N. Ohta and A. R. Robertson, J. Wiley, Chichester, West Sussex, England; Hoboken, NJ, USA (2005). [63] Wikipedia, the free encyclopedi a Wikimedia Foundation, Inc. (2012). [64] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [65] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [66] Wikipedia, the free encyclopedia Wikimedia Foundation, Inc. (2012). [67] of Organic L ight Adv. Mater. 15 1043 1048 (2003) [doi:10.1002/adma.200302151]. [68] N. Chopra, J. Lee, Y. Zheng, S. Balance on High Efficiency Blue Phosphorescent Organic Light Emitting AC S Appl. Mater. Interfaces 1 1169 1172 (2009) [doi:10.1021/am900228b].

PAGE 130

130 [69] E. J. W. List, C. H. Kim, W conjugated wide bandgap semiconductors: a Materials Science and Engineering: B 85 218 223 (2001) [doi:10.1016/S0921 5107(01)00 588 8]. [70] B. Krummacher, M. K. Mathai, V. E. Choong, S. A. Choulis, F. So, and A. efficiency of organic Organic Electronics 7 313 318 (2006) [doi:10 .1016/j.orgel.2006.03.011]. [71] exciton quenching in organic phosphorescent light emitting diodes with Ir Phys. Rev. B 75 125328 (2007) [doi:10.1103/PhysRevB.75.125328]. [72] Z. Shuai, D. B Formation Rates in Conjugated Polymer Light Phys. Rev. Lett. 84 131 134 (2000) [doi:10.1103/PhysRevLett.84.131]. [73] S. C. Lo, R. E. Harding, C. P. Shipley, S. G. S tevenson, P. L. Burn, and I. D. W. Triplet Energy Dendrons: Enhancing the Luminescence of Deep J. Am. Chem. Soc. 131 16681 16688 (2009) [doi:10.1021/ja903157e]. [74] S. C. Lo, E. B. Namdas, C. P Shipley, J. P. J. Markham, T. D. Anthopolous, P. Organic Electronics 7 85 98 (2006) [doi:10.1016/j.orgel.2005.11.003]. [75] R. W. Gruhlke Phys. Rev. Lett. 56 2838 2841 (1986) [doi:10.1103/PhysRevLett.56.2838]. [76] alized Surface Plasmon Advanced Materials 16 1685 1706 (2004) [doi:10.1002/adma.200400271]. [77] Phys. Rev. B 25 2297 2300 (1982) [doi:10.1103/Ph ysRevB.25.2297]. [78] S. Y. Kim and J. Organic Electronics 11 1010 1015 (2010) [doi:10.1016/j.orgel.2010.03.023].

PAGE 131

131 [79] M. zation of external coupling and light emission in organic light Journal of Applied Physics 91 595 604 (2002) [doi:doi:10.1063/1.1425448]. [80] P. A. Hobson, S. Wedge, J. A. E. Wasey, I. Sage, and W. L. Barnes, Plasmon Mediated Emission from Organic Light Adv. Mater. 14 1393 1396 (2002) [doi:10.1002/1521 4095(20021002)14:19<1393::AID ADMA1393>3.0.CO;2 B]. [81] B. Wang, L. Ke, and S. patterned organic light emitting diode Journal of Crystal Growth 288 119 122 (2006) [doi:10.1016/j.jcrysgro.2005.12.039]. [82] Y. Jin, J. Feng, X. L. Zhang, Y. G. Bi, Y. Bai, L. Chen, T. Lan, Y. F. Liu, Q. D. Stability Tradeoff in Top Emitting Organic Light Advanced Materials 24 1187 1191 (2012) [doi:10.1002/adma.201103397]. [83] pola riton mediated emission of light from top emitting organic light emitting diode type Organic Electronics 8 136 147 (2007) [doi:10.1016/j.orgel.2006.07.003]. [84] mat Proceedings of SPIE 6999 69992T 69992T 12 (2008) [doi:doi:10.1117/12.780249]. [85] film waveguiding mode light extraction in organic electroluminescent device using high refr active index Journal of Applied Physics 97 054505 054505 6 (2005) [doi:doi:10.1063/1.1858875]. [86] LEDs using an ultra low Optics Letters 35 1052 (2010) [doi:10.1364/OL.35.001052]. [87] J. Lee, N. Chopra, D. Bera, S. Maslov, S. H. Eom, Y. Zheng, P. Holloway, J. Conversion White Organic Light Emitting Diodes Using Advanced Energy Materials 1 174 178 ( 2011) [doi:10.1002/aenm.201000014]. [88] light Semiconductor Science and Technology 21 1094 1097 (2006) [doi:10.108 8/0268 1242/21/8/020].

PAGE 132

132 [89] H. K. Kim, S. H. Cho, J. R. Oh, Y. H. Lee, J. H. Lee, J. G. Lee, S. K. Kim, Y. I. Park, J. Organic Electronics 11 137 145 (2010) [do i:10.1016/j.orgel.2009.10.011]. [90] Journal of Applied Physics 100 073106 073106 6 (2006) [doi:doi:10.1063/1.2356904 ]. [91] S. light emitting devices via hemispherical microlens arrays fabricated by soft Journal of Photonics for Energy 1 011002 (2011) [doi:10.1117/1.3528267]. [ 92] Science 308 1274 1278 (2005) [doi:10.1126/science.1108712]. [93] tin oxide films on polymer substrates for application in plastic Thin Solid Films 397 49 55 (2001) [doi:10.1016/S0040 6090(01)01489 4]. [94] J. Appl. Phys. 95 186 9 (2004) [doi:10.1063/1.1640454]. [95] M. H. Lu, M. S. Weaver, T. X. Zhou, M. Rothman, R. C. Kwong, M. Hack, and J. efficiency top emitting organic light Appl. Phys. Lett. 81 3921 (2002) [doi:10.1063/1.1523150]. [96] top emitting organic light Appl. Phys. Lett. 84 2986 2988 (2004). [97] C. C. Liu, S. H. Liu, K. C. Tien, M. H. Hsu, H. W. Chang, C. K. Chang, C. J. Yang, and C. emitting organic light emitting devices integrated with diffusers for simultaneous enhancement of efficiencies and Appl. Phys. Lett. 94 103302 (2009) [doi:10.1063/1.3097354]. [98] X. W. Chen, W. C and optimal design of top emitting organic light J. Appl. Phys. 101 113107 (2007) [doi:10.1063/1.2739220]. [99] ition of cathodes J. Appl. Phys. 86 4607 (1999) [doi:10.1063/1.371410].

PAGE 133

133 [100] structure of gallium indium oxide/silver/gallium indium Materia ls Letters 61 3897 3900 (2007) [doi:10.1016/j.matlet.2006.12.053]. [101] Optics Communications 282 574 578 (2009) [doi:10. 1016/j.optcom.2008.10.075]. [102] emitting diodes using a multilayer oxide as a low resistance transparent Appl. Phys. Lett. 93 013301 (2008) [doi:10.1063/1.2955528]. [103] K. S. Lee, S. Y. Ryu, S. J. Jo, C. S. Kim, S. H. Choi, J. H. Noh, H. K. Baik, H. S. emitting diodes using Appl. Phys. Lett. 91 093515 (2007) [doi: 10.1063/1.2776347]. [104] C. H. Lee, S. Y. Song, S. H. Choi, S. Y. Park, S. Y. Ryu, J. H. Noh, B. H. Hwang, emitting diodes consisting of Appl. Phys. Lett. 92 023 306 (2008) [doi:10.1063/1.2835044]. [105] Appl. Phys. Lett. 70 1665 (1997) [doi:10.1063/1.118664]. [106] M. A. Baldo, S. Lamansky, P. E. Burrows, M. E. Thompson and S. R. Forrest, efficiency green organic light emitting devices based on Appl. Phys. Lett. 75 4 (1999) [doi:10.1063/1.124258]. [107] el, emitting diodes through J. Appl. Phys. 89 420 (2001) [doi:10.1063/1.1331651]. [108] the Emission from a Conjugated Polymer Light Emitting Diode: Implications for Advanced Materials 6 491 494 (1994). [109] emitting device with an ordered monolayer of silica microsphere Applied Physics Letters 76 1243 1245 (2000) [doi:doi:10.1063/1.125997]. [110] M. K. Wei and I. Opt. Express 12 5777 5782 (2004) [doi:10.1364/OPEX.12.005777].

PAGE 134

134 [111] emitting device luminaire for Applied Physics Letters 88 192908 192908 3 (2006) [doi:doi:10.1063/1.2202722]. [ 112] Emitting Devices for Solid Advanced Materials 16 1585 1595 (2004) [doi:10.1002/adma.200400684]. [113] th Nature 424 824 830 (2003) [doi:10.1038/nature01937]. [114] ASTM International (2010). [115] C. L. Lin, T. Y. Cho, C. H. Chang, and C. organic light emitting devices by locating emitters around the second antinode of Applied Physics Letters 88 081114 081114 3 (2006) [doi:doi: 10.1063/1.2178485]. [116] plasmon polariton mediated emission from phosphorescent dendrimer light Applied Physics Letters 88 161105 (2006) [doi:10.1063/1.21 93795]. [117] V. Bulovic, G. Gu, P. E. Burrows, S. R. Forrest, and M. E. Thompson, Published online: 07 March 1996; | doi:10.1038/380029a0 380 29 29 (1996) [doi:10.1038/380029a0]. [118] G. Parthasarathy, P. E. B metal Applied Physics Letters 72 2138 2140 (1998) [doi:doi:10.1063/1.121301]. [119] f organic electrophosphorescence. II. Transient analysis of triplet Phys. Rev. B 62 10967 10977 (2000) [doi:10.1103/PhysRevB.62.10967]. [120] V. Adamovich, J. Brooks, A. Tamayo, A. M. Alexander, P. I. Djurovich, B. W. single dopant white electrophosphorescent light emitting diodesElectronic supplementary information (ESI) available: emission spectra as a function of doping concentration for 3 in CBP, as well as the absorption and emission spectra of Irppz, CBP and mCP. See New Journal of Chemistry 26 1171 1178 (2002) [doi:10.1039/b204301g].

PAGE 135

135 [121] S. Kwon, K. rgy level Applied Physics Letters 97 023309 (2010) [doi:10.1063/1.3462931]. [122] N. Chopra, J. S. Swensen, E. Polikarpov, L. Cosimbescu, F. So, and A. B. iency and low roll off blue phosphorescent organic light Applied Physics Letters 97 033304 (2010) [doi:10.1063/1.3464969]. [123] H. Kajii, T. Tsukagawa, T. Taneda, K. Yoshino, M. Ozaki, A. Fujii, M. Hikit a, S. Diode \ Using 8 Hydroxyquinoline Aluminum Doped with Rubrene \ as an Electro Japanese Journal of Applied Physic s 41 2746 2748 (2002) [doi:10.1143/JJAP.41.2746]. [124] Thin Solid Films 363 327 331 (2000) [doi:10.1016/S0040 609 0(99)01010 X]. [125] Macromolecular Symposia 125 49 58 (1998) [doi:10.1002/masy.19981250103]. [126] H. Murata, C. D. Merritt, and Z. H. Ka doped molecular organic light emitting diodes: direct carrier recombination at Selected Topics in Quantum Electronics, IEEE Journal of 4 119 124 (1998) [doi:10.1109/2944.669481]. [127] J. M. Ca quantum dot light emitting diodes with metal Nature Photonics 2 247 250 (2008) [doi:10.1038/nphoton.2008.34]. [128] B. O. Dabbousi, J. Rodriguez Viej o, F. V. Mikulec, J. R. Heine, H. Mattoussi, R. Synthesis and Characterization of a Size Series of Highly Luminescent J. Phys. Chem. B 101 9463 9475 (1997) [doi :10.1021/jp971091y]. [129] dot light emitting diodes based on solution Nature Photonics 5 543 548 (2011) [doi:10.1038/nphoton.2011.171].

PAGE 136

136 B IOGRAPHICAL SKETCH Edward Wrzesniewski was born in Delaware County, Pennsylvania. From an early age, the prospect of becoming a mad scientist appealed to him, almost as much as becoming a Teenage Mutant Ninja Turtle. While the desire to be horribly mutat ed by unknown chemicals abated, the need to invent, d iscover, and design only grew. In order to satisfy this need, he enrolled at Carnegie Mellon University in 2004 for his B.S degree in Materials Science and Engineering. Upon graduation, he continued his education by joining the Xue group at the University of Florida. There he completed his M.S. and continued on to complete his PhD. Upon completion of his PhD, Ed plans to move to California and to work in renewable energy or high efficiency lighting.