1 PROCESSING AND CHARACTERIZATION OF OPTOELECTRONIC DEVICES BASED ON INORGANIC NANOCRYSTALS AND ORGANIC SEMICONDUCTORS B Y GALILEO SARASQUETA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN P ARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 1
2 201 1 Galileo Sarasqueta
3 To my wife and family
4 ACKNOWLEDGMENTS First I thank my Lord and Savior, Jesus Christ, for guiding me in my life and bringing me to this point. Without Him, I would not be the person I am today and my life would be very different. I thank Him for everything. I want to thank Dr. Franky So for giving me the opportunity to prove myself and work in his group. I a m thankful for his advice and guidance in research, and in my career in general. I also thank him for supporting me and inculcating the drive to improve consistently through hard work, perseverance, diversity, and creativity. I thank Dr. Amelia Dempere fo r taking me in and supporting me financially for the first two years of my program. I will always be in debt to her for giving me the opportunity to work at MAIC, for giving me advice in life, for always looking out for me, and for treating me as a friend I thank Dr. Norton, Dr. Xue, and Dr. Ziegler for being interested in my research and being part of my committee. Kaushik Roy Choudhury for teaching me about quantum dots and giving me advice in research. I also thank Dr. Subbiah Jegadesan, Dr. Jiyeon Song, Jaewon Lee, Neetu Chopra Alok Gupta, Cephas Small, Dongwoo Song, Michael Hartel, Mikail Shaikh, Song Chen, Pieter De Somer, Verena Giese, Daniel S. Duncan, Frederick St effy, Jessie Manders, and Nikhil Bhandari. I thank them for being my family during research in the lab and for sharing their lives and good times with me. I feel very fortunate to work with and get to know them I thank Fernando Lugo, Hector Martinez, Ser ge Maslov and Joe for being such great friends and being there for me when I needed them the most. I thank them for
5 their good spirit and words of confort and encouragement. My life at UF has been great because of the many good times I shared with them. I thank my father and mother for loving me and for supporting me in all ways possible. I thank them for the words of encouragement, for having faith in me, and for educating me to be the person I am today. Finally, I thank my wife and the love of my lif e, for always being there for me. I thank her for giving me good advice, for challenging me to be a better person, for always laughing and having a good, relaxing time with me, and for always lifting my spirits when I need it.
6 TABLE OF CONTENTS P age ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 15 ABSTRACT ................................ ................................ ................................ ................... 16 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 18 1.1 Organic Electronics ................................ ................................ ........................... 18 1.1.1 Organic Semiconductors ................................ ................................ ......... 18 1.1.2 Advantages and Disadvantages of Organic Se miconductors .................. 22 1.1.3 Purification of Organic Small Molecules ................................ .................. 25 1.2 Inorganic Nanocrystals ................................ ................................ ..................... 25 1.2.1 Background ................................ ................................ ............................. 26 1.2.2 Synthesis of Semiconductor Nanocrystals ................................ .............. 27 1.2.3 Growth Mechanism of Na nocrystals ................................ ........................ 30 1.2.4 Optical and Electronic Properties of Semiconductor Nanocrystals .......... 32 1.2.5 Advantages and Disadvantages of Col loidal Nanocrystals ...................... 33 1.3.1 Charge Carrier Transport ................................ ................................ ........ 35 1.3.2 Excitons ................................ ................................ ................................ ... 37 1.3.3 Energy and Charge Transfer Mechanisms ................................ .............. 38 1.4 Organic Photovoltaic Solar Cells ................................ ................................ ...... 39 1.4.1 Background ................................ ................................ ............................. 39 1.4.2 Nanostructured Organic Solar Cells ................................ ........................ 41 1.4.3 Solar Cell Characterization ................................ ................................ ...... 41 184.108.40.206 Current voltage measurement system ................................ ........... 42 220.127.116.11 Spectral response measurements ................................ .................. 43 1.5 Photodetectors and Infrared t o visible Up conversion Devices ........................ 45 1.5.1 Background ................................ ................................ ............................. 45 1.5.2 Colloidal Nanocrystal Infrared Photodetectors ................................ ........ 46 1.5.3 Photodetector Characterization ................................ ............................... 46 1.5.4 Up conversion Devices ................................ ................................ ............ 48 1.6 Dissert ation Organization ................................ ................................ .................. 49
7 2 METHODS TO SYNTHESIZE NANOCRYSTALLINE STRUCTURES DIRECTLY FROM CO EVAPORATED ORGANIC FILMS FOR SOLAR CELLS AND HIGH K DIELECTRIC ................................ ................................ .................... 61 2.1 Introduction ................................ ................................ ................................ ....... 61 2.2 Synthesis of Metallic Nanocrystals in Organic Thin Films ................................ 61 2.2.1 Fabrica tion of In situ Metallic Nanocrystals in an Organic Matrix ............ 63 2.2.2 Results and Discussion ................................ ................................ ........... 64 2.3 Synthesis of Dense Organic Nan ostructured Films ................................ .......... 66 2.3.1 Procedure to Fabricate a Nanostructured Film ................................ ........ 68 2.3.2 Results and Discussion ................................ ................................ ........... 69 2.4 Summary ................................ ................................ ................................ .......... 71 3 EFFECT OF NANOCRYSTALLINE, HIGH SURFACE AREA DONOR ACCEPTOR HETEROJUNCTION INTERFACE IN ORGANIC SOLAR CELLS ..... 79 3.1 Why Use Interpenetrating Heterojunction Interfaces? ................................ ....... 79 3.2 Fabrication of Bulk Heterojunction Organic Solar Cells ................................ .... 80 3.3 Results and Discussions ................................ ................................ ................... 81 3.4 Summary ................................ ................................ ................................ .......... 84 4 SYNTHESIS SETUP FOR INORGANIC SEMICONDUCTOR COLLOIDAL NANOCRYSTALS AND THEIR CHARACTERIZATION ................................ ......... 89 4.1 Introduction ................................ ................................ ................................ ....... 89 4.2 Nanocrystal Synthesis Setup ................................ ................................ ............ 89 4.3 Characterization of the Nanocrystals ................................ ................................ 91 4.3.1 Transmission Electron Microscopy (TEM) ................................ ............... 91 4.3.2 Spectrophotometry ................................ ................................ .................. 92 4.4 Synthesis of PbSe Nanocrystals ................................ ................................ ....... 94 4.5 Summary ................................ ................................ ................................ .......... 95 5 THE EFFECT OF LIGAND EXCHANGING SOLUTION TREATMENTS ON COLLOIDAL INFRARED PBSE NANOCRYSTAL FILMS AND PHOTODETECTORS ................................ ................................ ........................... 101 5.1 The Problem with Nanocrystal Capping Ligan ds ................................ ............ 101 5.2 Processing and Characterization of Nanocrystal Films and Photodetectors .. 102 5.2.1 Synthesis of PbSe Nanocrystals ................................ ............................ 103 5.2.2 Exchange on Nanocrystal Capping Ligands in Solution ........................ 103 5.2.3 Solid state Capping Ligand Exchange Treatment and Device Fa brication ................................ ................................ ............................. 104 5.2.4 PbSe Nanocrystal Film and Device Characterization ............................ 105 5.3 Results and Discussion ................................ ................................ .................. 106 5.3.1 Effect of Ligand Exchange Processes in Film Properties ...................... 106 5.3.2 Keeping Track of the Ligand Exchange in the Nanocrystal Films .......... 108 5.3.3 Effect of Dithiol Ligand Exchange on Device Characteristics ................ 109
8 5.4 Summary ................................ ................................ ................................ ........ 112 6 REDUCING DARK CURRENT OF INFRARED NANOCRYSTAL PHOTODETECTORS WITH INORGANIC/ORGANIC BLOCKING LAYERS ....... 122 6.1 Dark Current in Infrared Nanocrystal Photodetectors ................................ ..... 122 6.2 Engineering and Processing of PbSe NC Photodetectors with Blocking L ayers ................................ ................................ ................................ ............ 124 6.2.1 Device Fabrication and Processing ................................ ....................... 125 6.2.2 Preparation of the Blocking Layer Solutions and Materials ................... 126 6.2.3 Device Characterization ................................ ................................ ........ 127 6.3 Results and Discussion ................................ ................................ ................... 127 6.3.1 Effect of Blocking Layers on Photodetectors J V Characteristics .......... 127 6.3. 2 Detectivity Calculations for Photodetectors ................................ ........... 128 6.3.3 Effect of Blocking Layers on Stability of PbSe Nanocrystal Photodetectors ................................ ................................ ....................... 133 6.4 Summary ................................ ................................ ................................ ........ 134 7 APPLICATION OF SOLUTION PROCESSED INFRARED SENSING NANOCRYSTALS IN AL L S OLUTION PROCESSED UP CONVERSION DEVICES ................................ ................................ ................................ .............. 140 7.1 Introduction ................................ ................................ ................................ ..... 140 7.2 Fabrication of Solution Processed Up conversion Devices ............................ 141 7.3 Characterization of Reflect ive Up conversion Devices ................................ ... 143 7.4 Results and Discussion ................................ ................................ ................... 143 7.4.1 Device Mechanism and Control of Hole Injection ................................ .. 143 7.4.2 Efficiency of All Solution Processed Up Conversion Device ................. 146 7.4.3 Dark to Visible Switching Factor ................................ ............................ 147 7.4 Summary ................................ ................................ ................................ ........ 148 8 CONCLUSION AND FUTURE WORK ................................ ................................ .. 157 8.1 Conclusion ................................ ................................ ................................ ...... 157 8.2 Future Work ................................ ................................ ................................ .... 159 LIST OF REFERENCES ................................ ................................ ............................. 163 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 175
9 LIST OF TABLES Table P age 1 1 PV characteristics for various types of photovoltaic cells. ................................ .. 60 3 1 Solar c ell parameters for CuPc/PCBM solar cells with interlayers. ..................... 88 3 2 Solar cell parameters for N CuPc/PCBM solar cells with different N CuPc layer thicknesses. ................................ ................................ ............................... 88 6 1 Materials used in this work with their corresponding HOMO and LUMO energies. ................................ ................................ ................................ ........... 139
10 LIST OF FIGURES Figure P age 1 1 Classification of molecules based on complexity 154 ................................ ........... 50 1 2 Conjugated small molecules with alternating single and double bonds. ............. 50 1 3 Formation of Sigma and Pi bonds with sp 2 hybridization. ................................ ... 51 1 4 Schematic showing the interactions of energy levels in a sp 2 hybridization. ...... 51 1 5 Molecular structures of common small molecules used in organic electronics. .. 52 1 6 Structure of some common polymers used in the fabrication of organic electronics. ................................ ................................ ................................ ......... 52 1 7 Organic printing modules from different parties. ................................ ............... 53 1 8 Different photovoltaic solar cells. ................................ ................................ ..... 53 1 9 CIE coordinates for a cyclomated iridium complex with slight modifications in molecular structure. Extracted from reference s 23 25. ................................ ...... 54 1 10 Diff erent purification systems used. ................................ ................................ 54 1 11 constituent concentration and time. Extracted from reference 47. ...................... 55 1 12 Graphic representation of the free energy curve for a particle. ........................... 55 1 13 Ideal density of states for an energy band of a semico nductor with different dimensional bodies. Notice that the states are discrete only when the body is completely confined. Extracted from reference 50. ................................ ............ 56 1 14 A typical absorption spectra fo r PbSe nanocrystals of different sizes. ................ 56 1 15 CdSe nanocrystals with different sizes emitting light across the entire visible spectrum. ................................ ................................ ................................ ............ 57 1 16 Schematic depicting the energy diagram for hopping transport and a charge carrier hopping through localized states under an applied bias. ......................... 57 1 17 Schematic showing the differ ent types of excitons in a solid. ........................... 58 1 18 Schematic of an exciton dissociation process. ................................ ................. 58 1 19 Schematic representations of different types of heterojunctions ........................ 59
11 1 20 Typical J V characteristics for a photovoltaic cell under dark and illumination conditions. ................................ ................................ ................................ .......... 59 1 21 Solar simulator system for photovoltaic cell and photodetector measurements. ................................ ................................ ................................ ........................... 60 2 1 Image of one thermal evaporator used in this work. ................................ ........... 72 2 2 TEM images of Alq 3 blended with silver. ................................ ........................... 73 2 3 TEM image of Alq 3 blended with silver after annealing.. ................................ ..... 73 2 4 TEM electron diffraction patterns from films of Alq 3 blended with silver. ............ 74 2 5 TEM electron diffraction pattern from a neat amorphous Alq 3 film. ..................... 75 2 6 Plot showing the effect of annealing time on the dissipation factor and dielectric constant of composite films with different silver content. ..................... 75 2 7 CuP c films made by different methods. ................................ ............................ 76 2 8 CuPc films annealed at 210 C after being deposited. ................................ ...... 76 2 9 Effect of CuPc:TPD co e vaporation ratio on the resulting film uniformity. ....... 77 2 10 N CuPc films grown on different substrates. a) ITO. b) Glass. c) MoO 3. ......... 78 2 11 SnPc films treated with different methods. ................................ ....................... 78 3 1 Evaporators and glovebox used in the fabrication of the nanostructured CuPc films. ................................ ................................ ................................ ................ 84 3 2 Absorption spectrum of a CuPc:TPD co evaporated film compared with a plain CuPc film and a plain TPD film. ................................ ................................ 85 3 3 Absorption spectrum of a CuPc:TPD after annea ling compared with a plain CuPc film and a slightly doped CuPc with TPD. ................................ ................. 85 3 4 J V characteristics for CuPc/PCBM bilayer solar cells with and without hole extraction interlayers. ................................ ................................ ......................... 86 3 5 Cross sectional SEM images of nanostructured CuPc films. ............................ 86 3 6 J V characteristics for N CuPc/PCBM solar cells with different N Cu Pc layer thicknesses. ................................ ................................ ................................ ........ 87 3 7 External quantum efficiency for a planar CuPc solar cell compared with a N CuPc solar cell. ................................ ................................ ................................ ... 87
12 4 1 Colloidal nanocrystal synthesis setup. ................................ ................................ 96 4 2 Components for the NC synthesis setup. ................................ ......................... 96 4 3 Three neck round flask connected t o the condenser, thermocouple, and heating pad. ................................ ................................ ................................ ........ 97 4 4 Illustration of a TEM system with the main components. ................................ .... 98 4 5 The electro n optical system for a TEM. ................................ ........................... 98 4 6 Copper TEM grids. ................................ ................................ ............................. 99 4 7 Typical absorption spectrum of PbSe NCs. The higher the reaction temperature, the NCs will absorb at longer wavelengths. ................................ ... 99 4 8 TEM images from different NC synthesis with different precursor concentration ratios. ................................ ................................ ......................... 99 4 9 TEM image of PbSe NCs grown at the same temperature but left growing for different times. ................................ ................................ ................................ 100 5 1 Schematic of process to treat NC films with BDT. ................................ ............ 114 5 2 Schematic of PbSe NC photodetector device structure. ................................ ... 114 5 3 Schematics of capping ligand molecules with a respective TEM image of t he PbSe NCs capped with them. ................................ ................................ ........ 115 5 4 AFM images from films of PbSe NCs capped with oleic acid and treated with dithiol solutions. Insets are the corresponding SEM images of the same films which share the same scale as the AFM images. ................................ ........... 115 5 5 AFM images from films of PbSe NCs capped with octylamine and treated with dithiol solutions. Insets are the corresponding SEM images of the same films, which share the same scale as the AFM images. ................................ 116 5 6 Absorption spectrum from PbSe NCs capped with octylamine in solution, with EDT, and BDT. ................................ ................................ ................................ 116 5 7 FTIR spectrum for NC films with oleic acid and after ligand exchange with octylamine. ................................ ................................ ................................ ....... 117 5 8 FTIR spectra for NC films with octylamine and after BD T treatment. ............... 117 5 9 AES spectroscopy data from a BDT treated film. ................................ ........... 118 5 10 J V characteristics for a device with long cappin g ligands and after dithiol treatment. ................................ ................................ ................................ ......... 118
13 5 11 Shcematic of unpassivated dangling bonds that act as mid gap traps in NCs. 119 5 12 J V characteristics for the single carrier devices plotted in log log scale. ....... 119 5 13 Temperature dependence of the dark current for the photodetector devices treated with dithiol molecu les and not treated. ................................ ................ 120 5 14 Dark and photocurrents for the photodetectors treated with EDT and BDT. ..... 120 5 15 Calculated resp onsivity for the photodetectors treated with EDT and BDT. ..... 121 5 16 Spectral responsivity for a BDT treated photodetector at different voltages. .... 121 6 1 A schematic of the energy band diagram necessary to block charge injection under reversed biad for a photodetector and the corresponding structure of the device. ................................ ................................ ................................ ...... 135 6 2 Illustration of some materials used in this work. ................................ ............. 135 6 3 Dark J V characteristics of PbSe NC photodetectors. ................................ .... 136 6 4 Da rk currents, photocurrents, and calculated detectivity values for photodetectors with different blocking layers. Values for 0.5V. ...................... 136 6 5 Detectivity as a function of wavelength for a photod etector with blocking layers compared with another one without blocking layers. Calculated using responsivity values at 0.5V. ................................ ................................ ............. 137 6 6 Responsivity and measured detectivity as a function of w avelength for a photodetector without any blocking layers and with blocking layers. Values measured at 1.0V and with 830 nm source light. ................................ ............. 137 6 7 Photocurrent density and responsivity of a p hotodetctor with blocking layers (structure shown in legend) for different light intensities. ................................ .. 138 6 8 Responsivity as a function of time for devices with and without ZnO NCs as the cathode int erlayer. ................................ ................................ ...................... 138 7 1 Schematic representation of the basic structure for the solution processed up conversion device. ................................ ................................ ....................... 149 7 2 Schema tic of measurement setup for up conversion device. ........................ 149 7 3 Schematic representation of the mechanism by which the up conversion process takes place. ................................ ................................ ......................... 150 7 4 Device structures with their corresponding energy band diagrams for a typical MEH PPV light emitting device and a simple up conversion device. .... 150
14 7 5 LIV chara cteristics for the simple up conversion device under dark and infrared irradiation. ................................ ................................ ............................ 151 7 6 Device structure and corresponding energy band diagram for an up conversion device with a ZnO NC hol e blocking layer. ................................ ..... 152 7 7 Dark current for a typical MEH PPV light emitting device and up conversion devices with and without hole blocking layer. ................................ ................... 153 7 8 Dark and photocurrents under 830 nm monochromatic IR light and IR broadband light above 1000 nm for the up conversion device with ZnO NC hole blocking layer. ................................ ................................ ........................... 153 7 9 Luminance under dark and 830 nm IR light for the up conversion device with ZnO NC hole blocking layer. ................................ ................................ ............. 154 7 10 Luminous efficiency plot for the up conversion device. ................................ .... 154 7 11 Photon to photon efficiency for the up conversion device. ............................... 155 7 12 Plot for the dark to visible switching factor for the up conversion device. ........ 155 7 13 Image of a demonstration up conversion device under 830 nm IR irradiation and at three volts of applied bias. ................................ ................................ ..... 156
15 LIST OF ABBREVIATIONS AFM Atom ic force microscopy Alq 3 Tris(8 hydroxy quinolinato)aluminium BDT Benzenedithiol CuPc Copper phthalocyanine EDT Ethanedithiol FTIR Fourier transform infrared spectroscopy HOMO Highest occupied molecular orbital LUMO Lowest unoccupied molecular orbital MEH PPV Poly[2 methoxy 5 (2' ethyl hexyloxy) 1,4 phenylene vinylene] NC Nanocrystal N CuPc Nanostructured copper phthalocyanine OA Oleic acid PCBM [6 ,6] phenyl C 61 bytyric acid methyl ester PHOLED Phosphorescent organic light emitting diode PLED Polymer light emitting diode Poly TPD (P oly(N,N bis(4 butylphenyl) N,N bis(phenyl)benzidine ) QD Quantum dot SEM Scanning electron microscopy TEM Transmission electron microscopy TFB P oly[(9,9 dioctylfluorenyl 2,7 diyl) co (4,4 (N (4 sec butyl))diphenylamine)]
16 Abs tract 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 PROCESSING AND CHARACTERIZATION OF OPTOELECTRONIC DEVICES BASED ON INORGA NIC NANOCRYSTALS AND ORGANIC SEMICONDUCTORS By Galileo Sarasqueta May 201 1 Chair: Franky So Major: Material s Science and Engineering Organic electronic materials and inorganic colloidal nanocrystals are emerging as promising candidate s for the fabrication of versatile and high performance energy sources and photodetection has become very important and therefore organic electronics and colloidal nanocrystals are interesting for their great advantages in development of those fields. Additionally, organic electronic materials and colloidal nanocrystals have the potential to be used to fabricate large area low cost optoelectronic devices. The focus of this work has been on processing and fabrication of inorganic colloidal infrared sensing nanocrystal based devices and the improvement and characterization of their properties. Different organic and inorganic materials also complement the devices while keeping the ease of processability. Specifically, PbSe colloidal nanocrystals were used to fabricate infrared photodetectors. Chemical treatments were used to improve the nanocrystal film quality, to passivate traps in the nanocrystals, and to improve the device charact eristics. The effects of different
17 nanocrystal capping groups including oleic acid, octylamine, ethanedithiol, and benzenedithiol on the nanocrystal film quality and the photodiode characteristics were studied. It was found that the physical quality of the films was very sensitive to the capping groups used for the nanocrystals as well as to the specific exchange reactions. The spectral responsivity and quantum efficiency for the photodetectors were also measured and characterized. A PbSe solution processed nanocrystal based infrared photodetector incorporating carrier blocking layers was also demonstrated and significant reduction of dark current was achieved. Detectivity values close to 10 12 Jones are achieved by using a series of organic and inorganic mat erials such as poly[(9,9 dioctylfluorenyl 2,7 diyl) co (4,4 (N (4 sec butyl))diphenylamine)] (TFB ) and poly(N,N bis(4 butylphenyl) N,N bis(phenyl)benzidine ( poly TPD ) as the electron blockers ; while the small molecule C 60 and ZnO nanocrystals were use d as the hole blockers. An improvement in lifetime is also observed in the devices with the ZnO NCs hole blocking layer. Infrared sensing nanocrystals were also used to demonstrate the fabrication of the first all solution processed up conversion device. The processing of different materials was carefully controlled so that multiple inorganic/organic layers were solution processed on top of each other to fabricate the device. PbSe nanocrystals were used as infrared sensing mateial, MEH PPV was used as th e light emi tting layer, and ZnO nanocrystals were used as a blocking layer. Based on the solvents polarity and differences in solubitliy of the materials used, the processing of the devices was successful. In addition, ZnO was critical to decrease dark c urrents and enhance the dark to visible switching effect of the up conversion device.
18 CHAPTER 1 INTRODUCTION Th is chapter will serve to introduce and explain the basic concepts of organic semiconductor materials and inorganic quantum dot physics. Additio nal background information will be presented concerning specific optoelectronic devices such as photovoltaic solar cells and photodetectors. These concepts are meant to help the reader understand the principles behind the devices discussed in the followin g chapters of this dissertation. 1.1 Organic Electronics 1.1.1 Organic Semiconductors Historically, electronic devices and industry related products have been dominated by inorganic materials. In fact, there are plenty of inorganic materials that are cu rrently being used in the market worldwide for different purposes 1 Specifically, silicon is one of the dominant materials for applications such as solar cells, transistors, and photodetectors. These inorganic materials are very attractive because they ar e efficient and also because of the extensive knowledge and research that has been gathered about them in the past century. However, times change quickly and today we live in a world where the market wants new products with capabilities beyond the ones th at conventional inorganic materials can offer. As a result, the future of electronic devices is being shaped considerably by the development of new materials known as organic semiconductors. So what is an organic semiconductor? In conventional chemistry books and encyclopedias, an organic compound is defined simply as a substance whose molecules contain one or more (often many more) carbon atoms, with the exception of
19 certain molecules such as carbonates, cyanides, carbides, and a few others 2 3 This defi nition is just too general and does not give merit to the diversity of organic materials. Since organic molecules can vary from simple hydrocarbons such as methane, with just a few carbon atoms, all the way to more complex polymeric chains with multiple c arbon atoms, and considering the diverse properties of different organic materials, we must give a clearer definition and classification to these materials in order to have a better understanding of the subject at hand. Traditionally, solid materials are c lassified based on the nature of the bonding interactions of their constituents whether they are atoms or molecules as covalent, ionic, metallic, or molecular. While it is true that the constituents of organic solids are molecules held together by covalen t bonds, the vast majority of organic solids are molecular in nature because their intermolecular bonding interactions are characterized by weak van der Waals forces 4 Because van der Waals forces are much weaker than the typical bonds present in inorgani c solids (i.e., metallic, ionic, and covalent), the properties of organic solids are very different from the properties of inorganic solids 4 5 Some of the most contrasting properties between organic and inorganic solids are their mechanical, optical, and electrical properties, each of which will be discussed in more detail later. Now that organic solids have been distinguished from other solids based on their characteristic van der Waals intermolecular interactions, they can be further classified based o n the complexity of their constituents and their weight into three subdivisions : small organic molecules, polymers, and biological organic molecules Figure 1 1 shows schematics of these materials. As a rule, the small organic molecules are the simplest
20 organic solids because each constituent molecule consists of only a few to a few hundred atoms. Typically, small molecules are also light in weight (MW < 1000amu) 6 As the organic solids become more complex, the size and weight of their molecules increas e. Polymers are an example of long chain molecules composed by smaller building units called monomers. Very complex polymers can be created with different lengths and weighst (thousands to millions amu) depending on the monomers unit variation, orientati on, and building order. Biological organic molecules are the most complex of organic molecules, usually with complicated structures and large sizes. The DNA structure is the most common example of a biological organic molecule. Even though there have be en some studies regarding the electronic and optical properties of biological molecules 7 8 the field of organic semiconductors is mainly concerned with the subdivisions of small organic molecules and polymers. Furthermore, not all the materials that be long to these two subdivisions are considered organic semiconductors. The materials known as organic semiconductors are characterized by being conjugated in their constituent molecules. The term conjugated refers to the alternating sequence of single a nd double carbon bonds in the molecule as in, for example, the case of benzene and anthracene in Figure 1 2. This sequence is important because when a molecule is conjugated in nature, it leads to the delocalization of electrons in the molecule 4 When tw o carbon atoms are brought together, the 2s orbital mixes with the p x and p y orbitals to form sp 2 hybrid orbitals so that the electrons can reside in the most stable state possible. As a result, the p z orbital left from each carbon atom overlaps with the neighboring one, creating what is called a bond bond s are of great interest because they are characterized by the delocalization of the electron density above and
21 below the plane of the carbon atoms as shown in Figure 1 3 4 This delocalization leads to degeneracy and to the establishment of the highest energy occupied molecular orbital (HOMO) and lowest energy unoccupied molecular orbital (LUMO or anti bonding orbital), which are analogous to the valence band (VB) and the conduction band (CB) for inor ganic materials, respectively 9 Figure 1 4 shows how the orbitals mix to form these molecular energy states. As a consequence of having HOMO and LUMO bands, organic small molecules and polymers that are conjugated have semiconducting properties and are r eferred as organic semiconductors. As mentioned before, polymers and small molecules differ from each other mainly in how complex and heavy their constituents are, with the latter being the simpler and lighter. Because of this difference, small molecules can easily be turned into vapor phase and that is why they are processed into thin films by sublimation techniques such as vacuum thermal evaporation. On the other hand, when polymers are heated up, they decompose and break down before reaching the sublim ation point. Instead, polymers are typically dissolved in organic solvents and used in solution processes such as spin coating and dip coating to fabricate films. Another predominant difference between small molecules and polymers is their crystalline p roperties. The crystallinity of small molecule materials depends on the ordered stacking of their molecules and, therefore, is directly related to how symmetric the constituent molecules are. Generally, asymmetric small molecules do not stack in an order ly fashion and usually form amorphous films while symmetric small molecules can stack on top of each other and form crystalline films 9 12 In fact, depending on the evaporation conditions and different substrates, some small molecule materials can be
22 grow n into large single crystals. Figure 1 5 shows some common amorphous and crystalline small molecules. Pentacene is a symmetric molecule that can form very crystalline films and is used in transistors 10 Alq 3 [tris (8 hydroxyquinoline) aluminum] is an amo rphous molecule that is used in organic light emitting diodes 12 CuPc (Copper Phthalocyanine) is another crystalline molecule that is regularly used for organic solar cells 13 On the other hand, polymer molecules are so much longer and bulkier that the en tire molecules structure cannot stack orderly. Instead, polymers typically coil up and down repeatedly, creating small domains that are considered crystalline. Unfortunately, it is hard to create these crystalline domains in polymer films because when th ey are processed from solutions, the long chains get entangled and ultimately create amorphous films. Nevertheless, the morphology and crystallinity of polymer films can be controlled by manipulating different parameters. Some of these parameters are rel ated to the material itself (molecular weight, chain side groups, regioregularity); while others are related to the processing conditions such as careful selection and mixture of solvents, post fabrication thermal treatments, and surface/interface alterati ons 14 17 Figure 1 6 shows some common polymer materials used in fabricating plastic light emitting diodes and solar cells. 1.1.2 A dvantages and Disadvantages of Organic Semiconductors Organic semiconductor materials have considerable advantages ov er inorganic semiconductor counterparts. Perhaps the most compelling advantages of organic semiconductors are the potential for low cost production and the fact that they are compatible with large area manufacturing processes. As mentioned before, small molecules are usually deposited using simple vacuum thermal evaporation, which typically is done at much lower temperatures compared to inorganic semiconductor
23 processes. Thermal evaporation of small molecules can also be adapted to large area substrates 1 8 Furthermore, since polymers are specifically processed from solution, they are ideal for large area manufacturing processes such as printing, which would increase the production rate and minimize the cost of the products 1,19 Figure 1 7 shows some man ufacturing printing modules for organic semiconductors that are currently being employed and developed by different parties. Another outstanding advantage of using organic semiconductors is that they are soft materials (due to their weak intermolecular bo nding forces) and, therefore, are compatible with flexible devices by using flexible plastic substrates 19 20 In addition, other cheap materials such as glass or metal foil can be used as substrates for organic semiconductors without neglecting the qualit y of the devices. Organic semiconductors possess this advantage because it is not as important for them to match the lattice of the substrates as it is for inorganic semiconductors, in which the atoms at the substrate interface have to bond in a specific way and defects have much more drastic effects on the devices. Due to their light molecules and high absorption coefficients which can go as high as 10 5 cm 1 21 organic semiconductors materials are also great candidates for very thin (<500 nm) and lightwei ght photovoltaic devices. Inorganic materials, on the contrary, are generally heavier, brittle, and when used to fabricate solar cells often much thicker (~1 m). Figure 1 8 shows the comparison of a classic inorganic silicon solar cell with a plastic sol ar sell. As a consequence, organic solar cells are already being targeted for very practical applications embedded in other electronic devices or even windows for houses or cars 22
24 Another important advantage of organic semiconductors is that their consti tuent molecules can be tuned and engineered to specific application requirements. Typically, the optical properties of small molecule materials can be tuned by changing the chemical structures of the constituent molecules. Figure 1 9 shows an example of this optical tuning. The iridium complex is commonly used in OLEDs and can be tuned so that its optical absorption and emission can change to different wavelengths of light across the visible spectrum 23 25 However, in spite of all their advantages, org anic semiconductors materials also have disadvantages. Common drawbacks of organic semiconductors are their low intrinsic carrier density and their low carrier mobility that result from their localized molecular nature and weak intermolecular interactions As a consequence, organic semiconductors typically have high resistivity and high voltages are required when they chemical impurities that are hard to separate from the material desired. It is difficult to control the purification of small molecules because many times it is done through thermal gradient deposition. Stability of organic semiconductors in ambient cond itions and in operation is another problem since many organic materials tend to degrade if they are exposed to moisture and oxygen, while some others simply decompose during operation 26 28 Even though this problem can be aided by encapsulation techniques it is still more convenient and organic semiconductor based devices.
25 1.1.3 Purification of Organic Small Molecules As mentioned earlier, small molecule materials a re difficult to purify and it is not uncommon to purchase commercially available materials that are only 99.9% pure. There are different methods to purify organic materials, including chromatography, zone refining, and thermal gradient sublimation 29 31 Some of the small molecules used in this thesis were purified through the latter. The concept of purifying organic molecules with thermal gradient sublimation consists of separating the chemical impurities from the actual material of interests by sublimat ing them and letting the impurities deposit at different temperatures in a thermal gradient. Volatile impurities usually deposit at colder places and the less volatile impurities deposit at the hotter areas or do not sublimate and are left behind in the cr ucible. At the end, the good material is mechanically collected to be used. Typically, to aid the process and avoid more contamination, the sublimation takes place in a vacuum on the order of 10 6 Torr. A different approach involves running the sublimati on process at atmospheric pressure, but under a constant flow of inert gas. A typical setup for thermal gradient sublimation involves an outer quartz tube where the vacuum is created and the material is placed, an inner tube onto which the material is dep osited, a tube furnace in which the outer tube is placed to create the thermal gradient, and a vacuum pump. The process usually takes several days and sometimes it must be repeated two or three times in order to obtain good quality material. Figure 1 10 shows the two systems that were built and used for this work. 1.2 Inorganic Nanocrystals When investigations on nanocrystals started more than a quarter of a century ago, nobody thought that research on these types of materials would develop and become
26 one of the leading areas of modern technology. Today, we witness nanoscience and nanotechnology being developed and implemented for different applications and areas such as in semiconductor electronic devices, imaging applications, and biomedical application s, just to name a few 32 36 1.2.1 Background A nanocrystal is defined as a cluster that contains from a few hundred to tens of thousands of atoms, ranging anywhere from a couple of nanometers to tens of nanometers in size. Nanocrystals are very attra ctive and interesting because they have unique properties that set them apart from typical bulk materials. These unique properties result because nanocrystals are larger than molecules, but much smaller than bulk materials. As a result, nanocrystals show properties that are somewhere in between the properties of discrete molecules and bulk materials 32 Furthermore, when the size or shape of nanocrystals are changed, their properties can also be altered accordingly, enabling scientists to tune them for spe cific purposes. In addition, it is easy to grow flawless, perfect nanocrystals since their length of scale is so small that there is barely any time to introduce defects 33 34 Because of these reasons, much investigation is dedicated to the study of nanoc rystals and how their special properties can be applied to practical applications. Some of the most interesting nanocrystal properties are their optical, electronic, magnetic, catalytic, and mechanical properties. For example, as the nanocrystal size is decreased, the band gap of the nanocrystals are increased and also, as a consequence, their optical absorption is affected 37 This is why materials such as CdSe nanocrystals can be tuned to emit light ranging through all visible spectrum wavelengths. It has also been shown that magnetic nanocrystals show size dependent magnetization transition temperatures 38 Some nanocrystals have also shown
27 remarkable yield strengths compared to their bulk counterparts, which makes them better candidates for high stres s applications and devices 39 In short, the study and applications of nanocrystal materials have huge potential for the technology of the future. An important factor with nanocrystals what determines their stability and their charge transport is their sur face passivation. By themselves, nanocrystals are not thermodynamically stable because they have very large surface area to volume ratios. As a consequence of this instability, the nanocrystals tend to agglomerate to reduce their high surface energy and sometimes they lose their special quantum properties. To avoid agglomeration, organic or inorganic molecules are used as capping ligands and are attached on the surface of the nanocrystals to stabilize them 32,40 41 Furthermore, inter nanocrystal interact ions, solubility in solutions, and some of the nanocrystal properties will also be affected by these capping ligands. Therefore, controlling the surface passivation of the nanocrystals with capping ligands is of extreme importance in order to achieve any d esired outcome with the use of nanocrystals. Almost any type of material s can be used to grow nanocrystals, ranging from pure metals and metallic alloys to ceramics and semiconductor materials. In this thesis, only semiconductor nanocrystals were used du e to their potential use in semiconductor electronic and optoelectronic devices. 1.2.2 Synthesis of Semiconductor Nanocrystals Depending on the reaction media, the preparation of nanocrystals can be classified into gas phase, liquid phase, and solid phase syntheses. Gas phase synthesis of nanocrystals can be done by many different routes in which the source is usually transformed using flame, plasma, or laser reactors. This reaction is followed by the establishment of supersaturation in the gas phase, con densation of the constituent
28 monomers, and nucleation into nanocrystals 42 Gas phase synthesis of nanocrystals can be beneficial because high purity material can be obtained, but typically it creates particles with wide size distribution, without any liga nds to stabilize the particles in solution, and they tend to be more complicated and expensive. Some gas phase syntheses do not require vacuum as in the case of the electrospray technique, where a source droplet is shot into air through an electric field 4 2 Unfortunately, this method, although inexpensive, also creates very polydisperse particles. While a lot of work has been done in the gas phase techniques and continues to be done most of the major advancements in nanocrystal science have been achieved with the liquid phase syntheses. Liquid phase synthesis of nanocrystals can be performed within a structured media (acting as a template) or in a template free media. Typical structured media are nanoporous membranes, reverse micelles, and microemulsion s 43 45 These methods are very effective and can produce very monodisperse nanocrystals, but some of them cannot produce large yields of nanocrystals and the variation of the nanocrystal size depends on the template. On the other hand, template free meth ods can produce decent yields and the sizes can be tuned easier through the use of surfactants. Liquid phase synthesis can be carried in aqueous solutions or in organic solutions to grow nanocrystals of almost any material 32,37 The organic solution temp late free liquid phase method will be the center of focus here since the nanocrystals used in this work were synthesized using this method. The principle of the synthesis of nanocrystals in organic solutions typically involves the decomposition of molecu lar precursors, i.e., molecules that deliver the
29 monomers of the desired nanocrystal at relatively high temperatures (100 300C ). The precursor solutions are usually mixed together by injecting rapidly one onto another in the presence of a hot solvent whe re the constituent monomers are freed and develop an oversaturation. Under these conditions, the nucleation and growth of nanocrystals is thermodynamically favored. The advantage of using organic solvents comes into play because the process can be carried out at higher temperatures compared to aqueous solutions, which opens the doors to different growth conditions 32 Typically, the constituent monomers are in liquid or in solid form, but when they are mixed with the precursor molecules at high temperature s, they bind with the liquid precursors. This small reaction usually takes place with a change in coloration of the solution. For example, when CdO is mixed with a phosphonic acid as the precursor, the solution turns from dark red to translucent after th e Cd + ions bind to the precursor. In the case of PbO, the solution turns from an orange to a clear solution. The reaction also takes place in the presence of another very important type of solvent molecule that controls the speed of the reaction by dynam ically attaching and detaching from the surface of the growing crystals. These molecules are called surfactants and ultimately they become the surface capping ligands of the nanocrystals. Surfactants are composed of two domains: a non polar domain, which determines the solubility of the nanocrystal and controls the diffusion properties during the reaction; and a polar domain, which controls the binding efficiency to the surface of the nanocrystal and therefore the stability of the nanocrystal 37,40 Using different surfactants can also lead to the growth of multiple nanocrystal shapes based on the idea that different polar groups will have affinities to bind with different crystal facets and therefore favor growth in specific crystal
30 directions 46 In gene ral, the synthesis of nanocrystals is controlled by many different reaction parameters such as temperature, time of reaction, type and concentration of surfactants, pressure, heating uniformly, and stirring rates. However, in order to fully control the sy nthesis of nanocrystals and to know how to manipulate the particle size distribution of the synthesis process, it is important to understand the growth mechanisms of these particles. 1.2.3 Growth Mechanism of Nanocrystals The general growth mechanism of n anocrystals is divided into two parts: the rapid nucleation of the nanocrystals, followed by a slow and controlled growth process of nanocrystals. Nucleation is the event that happens when the concentration of the constituent monomers is high enough that the system reaches a supersaturation state. Based on the classic studies by LaMer and Dinegar 47 when a rapid addition of the constituent monomers increases the concentration above a critical nucleation limit, the monomers assemble into small nuclei to r elieve the supersaturation of the system. When the concentration falls below this critical nucleation limit, the nucleation process stops, but since the system is still supersaturated, the growth process takes over. Figure 1 11 shows an illustration of t he processes. It is important to note that not all the nuclei that are formed become grown nanocrystals. Actually, during the nucleation process, only the nuclei that are above a critical size are allowed to grow; otherwise, they are dissolved back into the solution because they are not stable enough. The critical size of the stable nuclei is determined by the difference in the chemical potential of the constituents in their crystalline and solution phases along with the surface energy 32 This can be be tter understood by reviewing the definition of free energy for a particle, which is given by
31 (1 1) where d is the density of the nanocrystal, r is the radius of the nucleus, c and s are the chemical potential for the crystalline phase and the solution phase, and is the surface tension. It is important to note that this equation assumes a spherical particle with constant surface tension and neglects the differences in crystal facets. When the chemical potential of the monomers in solution phase is larger than the chemical potential of the monomers in the crystalline phase, the first term becomes negative and the free energy reaches a maxima that determines the critical particle size radius r c Hence, nuclei smaller than the cr itical size are dominated by the surface energy term and are forced to dissolve back into the solution. Figure 1 12 shows the nucleation free energy curve. As mentioned earlier, once the nucleation process stops, the growth process takes over. Growth of the nanocrystals is controlled by two different mechanisms: transport of the monomers to the surface of the nanocrystals by diffusion and the reaction at the surface in order for the constituents to be incorporated. When the reaction is still supersaturat ed and the concentration of constituent monomers is high, the growth is limited only by the reaction at the surface and it is independent of particle size, but when the available constituents start to get consumed and their concentration drops, the growth mechanism starts becoming more dependent on diffusion of the constituents 47 The growth of the particles also depends on the critical particle size, which shifts to larger values during the reaction. If all the particles are above the critical size value, then the small particles will grow faster and the system will grow in a size focusing regime However, when the particle sizes are below the critical size value, then the
32 system enters what is called the Oswald ripening regime which is characterized by a broadening of the size distribution and a decrease in particle concentration 48 49 Regardless of the different regimes in the growth process, in order to ensure monodispersity, the nucleation process must be separated from the growth process as much as p ossible. If the growth process starts while the nucleation is still taking place, then a polydisperse distribution will result regardless of any growth regime. In order to do so, nucleation must be kept as fast and short as possible. 1.2.4 Optical and E lectronic Properties of Semiconductor Nanocrystals As mentioned earlier, when the size of bulk materials is reduced to particles with nano scale dimensions (<50 nm), the nanocrystals materials show drastic change in properties such as in their optical, ele ctronic, magnetic, and chemical properties. These changes arise due to what is known as quantum confinement effects and also due to surface effects 32,40,50 51 Quantum confinement refers to the condition under which the electrons and holes are confined in a body with zero dimensions and, therefore, are not able to move freely. In bulk materials, the average separation between an electron and a hole, i. e., the exciton Bohr radius, is much smaller than the dimensions of the materials so they are allowed to move freely and because of that, the density of states is continuous. However, in a nanocrystal, i. e., a zero dimensional body, the exciton Bohr radius is comparable or larger than the dimensions of the body. In such a case, the charges are forced to move in restricted paths and this is why the electron density of states is discrete in quantized energy levels for nanocrystals 50 53 Figure 1 13 shows the density of states for different dimensional bodies. As a result of this quantization, the energy of the discrete density of states also changes based on quantum mechanic conditions and the band gap of the nanocrystal material becomes different compared to
33 that of the bulk. Accordingly, when the band gap of the material is changed due to quantum confi nement, the absorption spectra of the nanocrystals also get quantized. Figure 1 14 shows the typical absorption of inorganic QDs with the ir characteristic ex c itonic peaks. Since this extraordinary phenomenon in nanocrystal materials occurs due to quantum confinement effects, it means that it is also dependent on nanocrystal size and shape. Typically for semiconductor materials, as the nanocrystals get smaller, the band gap of the material gets larger and the first exitonic absorption peak shifts to small er wavelengths in the absorption spectra, respectively. Since not all materials have the same exciton Bohr radii, nor the same electron and hole effective masses, the quantization of the absorption spectra for different nanocrystal materials is different and dependent on specific materials. For example, the exciton Bohr radius for CdS and CdSe are 5 6 nm while the Bohr radius of PbSe is ~ 46nm, which is why PbSe can be tuned over a wider range in the optical spectrum based on nanocrystal size. Another d iference is that in the case of PbSe, the shift in conduction/valence energy levels as a function of particle size can be approximated to be the same because both the electrons and holes are confined strongly. This scenario is not the same for other nanoc rystal materials from the II VI and III V semiconductor groups where the quantum confine ment effect is not the same or comparable for the holes and electrons 51 1.2.5 Advantages and Disadvantages of Colloidal Nanocrystals Solution processed colloidal nano crystals have several advantages over using their respective bulk materials, but probably the main advantage of nanocrystal materials is the ability to tune or change the properties of the material by simply controlling the particle size. In this regard, nanocrystals can be tuned to specific
34 property requirements. For example, the size of CdSe nanocrystals can be tuned to emit any specific color in the visible spectrum for display applications 54 Figure 1 15 shows CdSe nanocrystals with different sizes e mitting different colors. PbSe and PbS can be tuned to absorb specific infrared wavelengths, which are used in telecommunication technology 36 Electronically, the nanocrystal mateials can be tuned to align their conduction/valence bands with other materi als to form Schottky or Ohmic contacts. Another advantage that nanocrystal materials have is their potential for cost effective and versatile applications. Nanocrystals can be synthesized and processed through simpler methods and at much lower cosst comp ared to bulk semiconductors. Just like organic semiconductors, nanocrystal materials can be used in large area and flexible production methods such as printing, which can increase yields and reduce cost of production. Also, nanocrystals are very attractiv e because they have much brighter emissions than some organic dyes, very narrow emission spectra with broad excitation spectra 55 Although nanocrystal materials are really attractive because of the above reasons, they also have some disadvantages. The b iggest disadvantage of nanocrystals is the inevitable presence of capping ligands and the lack of efficient surface passivation. As briefly mentioned before, nanocrystals need to be surrounded by capping ligands in order to be stable and prevent aggregati on. These ligands bond to the surface of the nanocrystals, but it is impossible for them to bond to all atoms at the surface. As a result, multiple dangling bonds are present at the surface of the nanocrystals that act as traps for charges and alter the p roperties of the material. In addition, since most capping ligands are insulating in nature, the charge transport
35 between nanocrystal materials is limited. Although there are ways to get around this problem for certain applications, e. g., using core/she ll structures 56 57 there must be more effort to improve surface passivation in nanocrystals. Another disadvantage that derives from the surface instabilities of nanocrystals is their lack of stability in ambient conditions. Nanocrystals tend to degrade easily if they are exposed to air or certain light sources and they quickly show degradation in their properties 58 Toxicity is another issue of certain nanocrystal materials for biomedical applications. Cadmium and Selenium ions are known to be toxic so when these materials are used for biological applications, they raise a level of concern 32 Other source materials to synthesize the nanocrystals such as lead oxide or dimethyl cadmiun are highly toxic, so many scientists choose to stay away from nanocry stal materials. 1.3 Electronic Processes in Organic and Nanocrystals Semiconductors This section will address the fundamental processes that govern the electrical characteristics of organic and nanocrystal semiconductors materials. The discussion will ma inly cover charge carrier transport mechanisms in molecules and nanocrystals and the fundamentals of excitons, charge, and energy transfer. 1.3.1 Charge Carrier Transport The topic of charge transport in organic and nanocrystal materials is still a develop ing subject and a lot of experimental research is being focused on it. It is known, though, that the charge transport in these materials falls in between two different cases: the band transport model and the hopping transport model. The transport for or ganic materials depends on the intermolecular interaction and the stacking of molecules. Band transport is typically seen in crystalline organic molecules that have stronger intermolecular interactions and therefore typically have a higher degree of charg e
36 delocalization 59 Materials that show band transport usually also have higher mobility values (> 1cm 2 /v s) while organic materials that show hopping transport have much lower mobility values. The band model consists of the formation of narrow transport ing bandwidths due to the delocalization of carriers, similar to the case of inorganic materials. However, the degree of delocalization is not comparable to the case of inorganic materials, which is why only mobility values up to 10 cm 2 /V s can be achieve d in crystalline organic materials. On the other hand, hopping transport is commonly seen in amorphous materials, where the intermolecular interactions are very weak and the charges are very localized. Hopping transport consists of carriers residing in a site for an extended period of time until there is enough energy for it to jump to the next available site. In this model, there is no defined transporting band; the charges are thought to be residing in potential wells 59 The transport requires the cha rges to hop to the next potential well by overcoming energy barriers with each hop. Figure 1 16 illustrates this model. Hopping transport is a thermally activated process since sufficient energy is required for the hopping steps to overcome the barriers. In nanocrystal materials, the inter particle distance also plays a similar role in determining the transport mechanism. Typically, nanocrystals that have been synthesized with long capping ligands show insulating characteristics with poor charge transpo rt. However, in contrast with organic molecules, the interactions between nanocrystals can be controlled by exchanging or modifying the capping ligands that surround the nanocrystal materials. By manipulating the surface chemistry of the nanocrystals, th e charge transport can be modified 60
37 1.3.2 Excitons Excitons are electron hole pairs that are bound by coulombic interactions. As opposed to typical inorganic semiconductor materials in which an excitation event leads to the formation of a free electron and hole, in organic materials and excitons result because of the degree of charge localization 32,50 Excitons are important in opto electronic devices because they can transport energy without a net charge. There are three types of excitons that have b een identified in crystalline materials: Wannier Mott, Frenkel, and charge transfer excitons. Figure 1 17 shows the three kinds of excitons. Wannier Mott excitons are highly delocalized and have very small coulombic interactions. As a result, they are us ually larger than single unit cells and are found in inorganic materials, not in organic materials 61 Frenkel excitons are highly localized and they are comparable to a few unit cell s in size; so in organic molecules, they can be found in a single molecul e. It would be analogous to say that because the organic materials have low dielectric constants, the electron hole interactions are much stronger and therefore the excitons are smaller. The exciton binding energy of a Frenkel exciton is ~ 1.0 eV. In sl ightly more delocalized organic materials, charge transfer (CT) excitons can be found. Here, CT excitons form between two or several molecules, increasing the size of the exciton compared to Frenkel excitons. Just like the transport of single charged car riers, the motion of excitons takes place through band transport or hopping transport as discussed previously. In the case of nanocrystals, the exciton bound energies are between 200 50 meV, which is less than the exciton binding energy for organic molecu les, but much more than the binding energies of Wannier Mott excitons that are found in bulk inorganic semiconductors. Nanocrystal materials are also characterized by having large dielectric constants, which translate to more
38 delocalization of the exciton s. Therefore, excitons in nanocrystal materials are more delocalized than in organic molecules; nevertheless, exciton confinement in nanocrystals is obviously strong, with increasing localization as the nanocrystals get smaller 52 53,90 When excitons are transported through different species of molecules, their energy is transferred through different mechanisms that have different effects important for different applications, as it will be discussed in the next section. 1.3.3 Energy and Charge Transfer Me chanisms When an exciton gets transferred from one species (donor molecule) to another (acceptor molecule), its energy can be transferred in one of three routes: radiative transfer or re absorption, F o rster energy transfer, or Dexter energy transfer 32 Ra diative transfer happens when the exciton at the donor molecule recombines, creating a photon that is then re absorbed by the acceptor molecule. In this case, the transfer happens over a distance of more than 10 nm and there must be an overlap between the emission of one species and the absorption spectrum of the other species. In a F o rster energy transfer, there should also be an overlap between the emission and absorption of both species, but the energy transfer is not radiative. Instead, dipole dipole interactions between the donor and acceptor molecules induce a resonant transition of the donor to the ground state and the acceptor to the excited state. This energy transfer happens in adjacent molecules and well within 10 nm. A Dexter energy transfer is the exchange of an electron in the excited state from the donor to the excited state of the acceptor species. This energy transfer requires an overlap between the energy wave functions of the two species involved in order to conserve the exciton and i s the dominant energy transfer in singlet singlet and triplet triplet energy transfers.
39 Alternatively, excitons may also dissociate into charge carriers. This dissociation usually occurs in the presence of an electric field, but it may also occur in the p resence of an energy offset between the wave functions of the donor and acceptor species, which induces the separation of the exciton. This type of exciton dissociation is especially useful in organic solar cells, which will be covered in the next section 1.4 Organic Photovoltaic Solar Cells 1.4.1 Background Photovoltaic solar cells are devices that transform solar energy into electrical energy. The photovoltaic effect occurs in organic molecules when photons excite molecules creating excitons, which the n dissociate into separate charges at the interface with another molecule. The dissociation of an exciton at the interface between a donor and an acceptor (i.e., a heterojunction) occurs when the energy level offset between the low ionization potential do nor material and the high electron affinity acceptor material is high enough to overcome the binding energy of the exciton 62 63 In such a case, the electron gets transferred to the acceptor material and the hole stays in the donor molecule as it is depic ted in Figure 1 18. Recently, organic photovoltaic devices have gotten special attention because they have great potential for low cost production of solar cells. Organic solar cells can also be versatile because of the attributes of organic based devic es such as flexibility and the option to produce large area devices with mass production printing fabrication methods. Since the demonstration of the first organic solar cell by Tang et. al, organic solar cells have improved tremendously. In less than 30 years, organic solar cells have been reported to have power conversion efficiencies of almost eight percent, so it is clear that organic solar cells can compete with amorphous silicon (a Si) in the industry
40 of solar energy 64 Table 1 1 compares the perfo rmance of the main inorganic based solar cell devices with that of organic solar cells. If the development in organic solar cells keeps the pace it has had in recent years, the efficiency of organic solar cells will certainly surpass that of a Si. Orga nic solar cells can be classified depending on their structure in two types: bilayer heterojunction solar cells and bulk heterojunction solar cells. The first type consists of a two layer stack formed by a donor material and an acceptor material. Each ma terial forms a distinct layer and the interface where they meet is called a heterojunction, which is where the excitons dissociate as explained before. Bilayer cells are limited because only the excitons that are formed within the exciton diffusion length (1 10 nm) from the heterojunction will be able to dissociate. Otherwise, any exciton that is created far away from the heterojunction is lost to recombination mechanisms. On the other hand, bulk heterojunctions are a mixture of donor and acceptor materi als in one single layer 65 In small molecules, this is achieved by co evaporation and with polymers this is achieved simply by mixing the two materials in solution. The advantage of bulk heterojunction cells is that the total heterojunction area is much greater than in bilayer cells. As a result, most excitons can dissociate and contribute to the current of the device, making it more efficient. In order to improve the efficiency of organic solar cells, the morphology of these devices must be optimized so that exciton dissociation is maximized while the collection of carriers is not compromised. In chapter 3 of this dissertation, we will discuss methods to modify the morphology of organic materials and how they can be controlled for specific applications such as solar cells.
41 1.4.2 Nanostructured Organic Solar Cells As mentioned earlier, bilayer devices suffer from having limited dissociation of the excitons that are created within their layers. This happens because the exciton diffusion length is too sma ll in organic materials and the excitons cannot reach the heterojunction before recombining. One way to get around this problem in bilayer structures is to modify the morphology of the heterojunction interface by creating ordered interpenetrating features from each layer that form the heterojunction 66 67 Figure 1 19 shows such an interpenetrating structure in comparison with a bulk heterojunction and a bilayer heterojunction. In such a scenario, the heterojunction surface area is increased and the excit ons created may have a better probability of reaching it in order to dissociate. In addition, by having an ordered structure, the generated free carriers can be transported with ease to the electrodes, which may not be the case in some bulk heterojunction systems where the charge transport and collection is not very efficient. Furthermore, if the crystallinity of the nanosctructure features is controlled, it may be possible to increase the exciton diffusion length and, therefore, increase the exciton diss ociation events. In chapter 4, we will discuss more in detail the effects of using crystalline nanostructured heterojunctions in solar cells. 1.4.3 Solar Cell Characterization Basically, solar cells are photodiodes that can be operated at zero bias. This means that current voltage ( I V ) measurements show rectification with a net current value passing through the device at zero volts under illumination. Figure 1 20 shows a characteristic I V data for a solar cell. For a solar cell, one of the most important parameters is the power conversion efficiency (PCE), which is calculated by the following formula:
42 (1 2) where J sc is the short circuit current density, V oc is the open circuit voltage, P inc is the incident power density in mW/cm 2 and the FF is the fill factor, which is defined as follows: (1 3) In equation 1 3, the J max and V max correspond to the values of current density and voltage that gives the maximum value of power produced by the device u nder illumination. Figure 1 18 also indicates all the different parameters just described, making it obvious that in order to calculate the power conversion efficiency of a cell, all that it is needed is a photo current voltage solar set up. The standard testing illumination condition for a solar cell set up requires that the active area is irradiated by 100mW/cm 2 of light under air mass 1.5 global illumination (AM 1.5G) at room temperature (298K). 18.104.22.168 Current voltage m easurement s ystem All the photo I V measurements done in this dissertation were carried out under the standard illumination requirement described above. In order to calibrate the solar simulator (AM 1.5G), a CIGS cell was used as a reference cell. Once the solar simulator was calibrat ed, the power intensity coming from it was varied by using a set of neutral density filters. The power was measured by a 70260 radiant power meter, combined with a 70268 probe in order to keep track of the different power intensities sh one at the solar ce lls besides the standard 100 mW/cm 2 Figure 1 21 shows an image of the system used for the measurements in this dissertation. For each
43 measurement of the photo I V characteristics, the devices were swept from negative to positive voltage in voltage mode using a Keithley semiconductor parameter analyzer. All the parameters discussed previously were collected from these I V characteristics. 22.214.171.124 Spectral response measurements The incident photon to current efficiency (IPCE) spectral response is an i mportant parameter that describes how efficient the cell is in transforming each photon of light into collected charge carriers as a function of photon wavelength. In order to measure this parameter, a grating monochromatic system was used coupled with th e solar simulator as the source of light. With this set up, it was possible to shine light of different wavelengths from 400 nm to 1800 nm with steps between 5 nm and 10 nm. Broad band filters and a light diffuser were used at the exit slit of the monoch romatic system to block the undesired second harmonics coming from the monochromator and to create a beam of light with uniform intensity, respectively. The photocurrents generated at the solar cells were measured with a lock in amplifier. The signal was also controlled by using a chopper that runs at 400Hz and a current to voltage amplifier to amplify the low signals. In order to measure the power of the light source as a function of wavelength, calibrated silicon and germanium photodetectors were also u sed in the test plane coupled with the lock in amplifier. The silicon photodetector was used for wavelengths in the visible up to 1100 nm and the germanium one was used for wavelengths in the infrared portion of the spectrum up to 1800 nm. The power dens ity was calculated by taking the photocurrent density function of wavelength, which was previously supplied by the vendor.
44 To calculate the spectral responsivity of the te sted solar cell (i.e., the responsivity as a function of wavelength), the photocurrent density generated from the tested solar cell is simply divided by the calculated power density as in the following equation: (1 4) where J cell ( ) is the photocurrent density from the tested solar cell and P inc ( ) is the incident power density measured with the photodetectors as a function of wavelength. The IPCE can then be calculated using the following expression: (1 5) where h is the Planck constant, c is the speed of light, q is the electron charge, and is the wavelength. A good practice when measuring the spectral responsivity and the IPCE of solar cells is to calculate the J SC and cross reference it with the measu red value from the photo I V measurements. In that way, it is easier to point out discrepancies and potential problems in the calibration of the systems. To calculate the J SC of the tested cells, the product of the shone power and the calculated spectral responsivity of the cell must be integrated as in the following expression: (1 7) where E ref is the reference spectral irradiance, which was obtained from the American Society of Testing and Materials (ASTM) standard G173.
45 1.5 Ph otodetectors and Infrared to visible Up conversion Devices 1.5.1 Background Just like solar cells, photodetectors are photodiodes that show rectification in their I V characteristics. However, these type of photodiodes are operated under reversed bias and their efficiency not only depends on the generation of current upon illumination, but also on the magnitude of current that is passed through the device when it is not under illumination (dark condition). Photodetectors have been around for a long time an d many different inorganic materials are used to fabricate photodetectors depending on specific applications. However, conventional inorganic photodetectors cannot be processed on flexible substrates, nor can they be produced with high throughput manufact uring methods that may require large areas. On the other hand, organic materials and nanocrystal based inorganic materials have all the qualities that are required for processing thin, flexible devices, and in large areas for low production cost. Differe nt organic materials have already been used to fabricate photodetectors with good results 68 70 but there has been a lack of organic materials that can be used in the short wavelength infrared (SWIR) region of the spectrum. Recently 71 72 there has been a n attempt to use organic materials with absorption spectra that extend into the near infrared (NIR), but it has been a challenge to create small molecule or polymers that can absorb beyond 1.3 m, which is where the ideal range for telecommunication applic ations is. In contrast, colloidal inorganic nanocrystals based on III V and IV VI semiconductors can overcome this limitation by extending the detector sensitivities to longer wavelengths and, in addition, colloidal nanocrystals can be processed at lower temperatures compared to some organic materials. Nanocrystal materials such as PbSe and PbS are great candidates for use in infrared photodetectors because they
46 can be tuned to absorb light all the way up to 2.0 m. Nevertheless, there are several facto rs that limit the performance of infrared inorganic nanocrystal photodetectors. 1.5.2 Colloidal Nanocrystal Infrared Photodetectors Recently, there have been multiple attempts to use nanocrystals for photodetectors or even solar cells applications that are active on the infrared 35 36, 73 75 Photodetectors have been demonstrated using neat nanocrystal thin films as well as thin films of hybrid bulk heterojunction blend composites with polymers 76 78 The polymer based composite devices often have a limi ted absorption, low carrier mobility, and poor device stability compared with crystalline inorganic semiconductor photodetectors. On the other hand, photoconductive devices based on neat nanocrystal thin films with high efficiencies have been reported. Ho wever, these devices suffer from high dark currents. To obtain photodetectors with high signal to noise (S/N) ratio, the detector responsivity needs to be high while the dark current needs to be low. Therefore, there is a great need to control the levels of dark current in these nanocrystal based photodetectors. In chapter 5 and 6 of this dissertation, we will discuss different methods to control the dark current of infrared nanocrystal photodetectors. 1.5.3 Photodetector Characterization Similar to sola r cells, photodetectors generate current when they are illuminated with light. However, photodetectors differ from solar cells in that they are operated under reversed bias, which means that there is an associated power loss whenever they are used. There fore, it is very important for photodetectors to minimize this loss by having low current passing through the device when it is not under illumination. Another difference is that since a photodetector is meant to operate at specific wavelength
47 ranges, the re is no use in shining it with a solar simulator as in photovoltaic devices. Instead, a monochromatic system is used to measure its efficiency at specific wavelengths. Just like in solar cells, the responsivity of a photodetector is an important measure of how much current is generated upon incident power of light on the device. The spectral responsivity of the photodetectors in this work was measured with the same procedure described for the solar cells. However, for specific wavelengths of light, the responsivity was measured as follows: (1 8) where J ph and J d are the photocurrent generated and the dark current respectively taken from I V characteristics measured with a Keithley semiconductor analyzer. The P inc is the power d ensity measured with the silicon photodetector coupled with a Newport Optical Power Meter 840 E for the specific wavelength of choice. The quantum efficiency could then be calculated as described earlier as in the case for solar cells. It is also import ant to note that since photodetectors are operated under reversed bias, their characteristics may also be dependent on voltage applied. Therefore, when the I V characteristics were measured, the devices were swept from 0V to 5V and subsequently from 0V t o +5V. Likewise, the spectral responsivity measurements were also carried out at different biases. In order to measure the performance of the photodetectors at different irradiated power densities, neutral density filters were used along with the monochr omatic system for low powers. However, for high power densities, lasers were used to illuminate the photodetectors.
48 1.5.4 Up conversion Devices Up conversion devices are devices that can convert low energy emissions of light into high energy emissions. This means that when photons with long wavelength enter the device, they get transformed into shorter wavelength photons 79 82 Because of this reason, up conversion devices have tremendous potential for different optoelectronic applications such as sensor s, homeland security, and especially in night vision technologies in which the conversion of IR to visible light is crucial. In the past, scientists have employed materials such as GaAs, AlGaAs, and other inorganic semiconductors to fabricate up conversion devices 82 85 Unfortunately, conventional processing and fabrication methods using such inorganic materials are very expensive and lack the versatility that organic materials offer. With recent developments in organic light emitting diodes, there has be en some effort in fabricating organic based up conversion devices that can convert IR to visible light 86 89 However, there are a limited number of organic materials that can absorb in the IR and most of them cannot absorb beyond the NIR. In addition, dev ice fabrication through solution processing methods can have great potential in order to reduce the fabrication costs and have large area device capabilities, but this has not been done until this work. As mentioned before, infrared sensing inorganic nano crystal materials are a viable option to sense infrared light in the NIR and also in the SWIR region of the spectrum. Therefore chapter 7 of this dissertation describes successfully using inorganic nanocrystals in the fabrication of the first ever solutio n processed IR to visible up conversion devices.
49 1.6 Dissertation Organization This dissertation covers topics on organic photovoltaic solar cells, inorganic nanocrystal photodetectors, and inorganic/organic hybrid devices including photodetectors and up conversion devices. Chapter 2 of the dissertation explains two novel methods that were developed to synthesize nanostructures in organic thin films and to modify the film morphology. The chapter also reports the physical characterization of the films The effect of modifying the heterojunction interface in bilayer organic solar cells is addressed in chapter 3. Chapter 4 provides a detailed description of the set up and the procedure to synthesize the colloidal infrared sensing inorganic nanocrystals used in this dissertation. Chapter 5 and 6 cover the fabrication of inorganic nanocrystal photodetectors with an emphasis in the control of the nanocrystal capping ligands and the use of blocking layers to suppress dark currents. Chapter 7 describes the fabrication of solution processed up conversion devices based on inorganic nanocrystal materials and organic light emitting polymers. Finally, the conclusions are presented in chapter 8 of this dissertation.
50 Figure 1 1. Classification of molecules based on complexity. 154 Figure 1 2. Conjugated small molecules with alternating single and double bonds. Images extracrted from reference 155.
51 Figure 1 3. Formation of Sigma and Pi bonds with sp 2 hybridization. a) Between two carbon atoms in E thane. b) Formation of conjugated Pi bonds leading to delocalization in benzene. Image extracted from reference 156. Figure 1 4. Schematic showing the interactions of energy levels in a sp 2 hybridization.
52 Figure 1 5. Molecular structures of comm on small molecules used in organic electronics. Figure 1 6. Structure of some common polymers used in the fabrication of organic electronics.
53 Figure 1 to roll organic solar cel l manufacturing module. 22 solar cell printing modules 19 Figure 1 8. Different photovoltaic solar cells. a) A typical inorganic silicon photovoltaic panel. b) A flexible organic based photovoltaic cell. Images extrac ted from 157 and 158.
54 Figure 1 9. CIE coordinates for a cyclomated iridium complex with slight modifications in molecular structure. Extracted from reference s 23 25. Figure 1 10. Different purification systems used. a) Small molecule high vacuum t hermal gradient purification system with a single heating zone. b) Constant inert gas flow thermal gradient purification system with three heating zones.
55 Figure 1 of consti tuent concentration and time. Extracted from reference 47. Figure 1 12. Graphic representation of the free energy curve for a particle. Extracted from referende 163.
56 Figure 1 13. Ideal density of states for an energy band of a semiconductor with dif ferent dimensional bodies. Notice that the states are discrete only when the body is completely confined. Extracted from reference 50. Figure 1 14. T ypical absorption spectra for PbSe nanocrystals of different sizes.
57 Figu re 1 15. CdSe nanocrystals with different sizes emitting light across the entire visible spectrum. Image extracted from reference 159. Figure 1 16. Schematic depicting the energy diagram for hopping transport and a charge carrier hopping through loca lized states under an applied bias.
58 Figure 1 17. Schematic showing the different types of excitons in a solid. a) Frenkel exciton. b) Charge transfer exciton. c) Wannier Mott exciton. Figure 1 18. Schematic of an exciton dissociation process. a) Exciton diffuses towards the heterojunction. b) Exciton reaches the heterojunction and dissociates into separate charges. c) Separate charge carriers get transported away from the heterojunction.
59 Figure 1 19. Schematic representations of differen t types of heterojunctions. a) Bulk heterojunction formed by phase separation from a mixed donor acceptor. b) Ordered interpenetrating heterojunction formed by controlling the interface. c) Planar bilayer heterojunction. Extracted from reference 66 F igure 1 20. Typical J V characteristics for a photovoltaic cell under dark and illumination conditions.
60 Figure 1 21. Solar simulator system for photovoltaic cell and photodetector measurements. Table 1 1. PV characteristics for various types of pho tovoltaic cells
61 CHAPTER 2 METHODS TO SYNTHESIZE NANOCRYSTALLINE STRUCTURES DIRECTLY FROM CO EVAPORATED ORGANIC FILMS FOR SOLAR CELLS AND HIGH K DIELECTRIC 2.1 Introduction Nanocrystal and other type of nanostructures have been used for many purposes an d different devices in organic electronics such as photodetectors, high K dielectrics, light emitting diodes, and photovoltaic devices, among others 34 36, 93 95 In many cases, nanostructures or nanocrystals are used along with organic materials in the fo rm of single layers to complement a function in a given device, as in the case of the photodetectors that will be discussed in chapters 5 7, or as in the case of microlenses used for light outcoupling in organic light emitting diodes 95 In such cases, inc orporation of the nanocrystals can be achieved with relative ease since the processing only involves a plain nanocrystal layer. On the other hand, incorporating or creating embedded nanosctructures in thin films, as in the case of bulk heterojunctions or blended nanocrystal devices, is more complicated to manage. In this chapter, we will discuss some of the problems regarding inorganic nanocrystals in organic blends, the manipulation of nanostructures in organic films and heterojunctions, and then we will describe two methods that were developed to create nanocrystals in thin films: the synthesis of metallic nanostructures in organic matrices and the synthesis of dense organic nanostructured films. 2.2 Synthesis of Metallic Nanocrystals in Organic Thin Fi lms Recently, there has been development on organic/inorganic nanocrystal composite materials for different applications. The demand for versatile and compact systems has driven intense research efforts to explore new materials with enhanced electrical an d mechanical properties. Mechanical flexibility and chemical tunability have
62 made polymers a preferred choice for many applications 96 However, in order to obtain the good mechanical properties from organic materials and the electronic properties from in organic materials, composite films consisting of blended inorganic nanocrystals in polymer matrices can be created. Specifically in the fabrication of high K dielectrics, a typical strategy is to form percolate composites by loading polymer matrices with ceramic fillers that posses high dielectric constants (for example, BaTiO 3 ) or with metal nanocrystals 93,97 As mentioned in chapter 1, inorganic nanocrystals exhibit fascinating size controlled electrical, optical, and magnetic properties due to their qu antum size effect, leading to their use in a variety of nanoengineered device embodiments. For the case of high K dielectric composites, metal nanocrystals that are dispersed in a polymer matrix offer unique properties like the Coulumb blockade effect 98 1 00 which can reduce electron tunneling between conductive fillers and thereby inhibit conduction loss in the dielectric system. However, a uniform dispersion of colloidal nanocrystals in a polymer blend, which is essential for controllable and reliable el ectrical properties, is difficult to achieve due to the propensity of the nanocrystals to agglomerate. While this problem other, a uniform dispersion is still not guarante ed. Moreover, nanocrystal passivation requires precise control of the passivating layer thickness on the nanocrystals and it poses strict synthetic challenges for colloidal nanocrystals 101 We can overcome these drawbacks by resorting to a technique of i n situ formation of metallic nanocrystals during fabrication in an insulating organic matrix, facilitating a more uniform dispersion of the nanocrystals. Here, metal/organic composite films are formed by co evaporating Alq 3 and silver in a vacuum. The me tallic nanocrystals are subsequently formed in situ
63 by heating treatments of the composite films at relatively elevated temperatures, resulting in uniformly dispersed nanocrystals embedded in an organic matrix. In addition, since these composites are co e vaporated, precise control over the loading can be possible. 2.2.1 Fabrication of In s itu Metallic Nanocrystals in an Organic Matrix The composite thin films were fabricated from Alq 3 (organic) and metallic silver (inorganic) through a simple thermal co ev aporation technique. Figure 2 1 shows the thermal evaporator used in this work. The materials were co evaporated from separate sources at a chamber pressure below 5.0 10 6 Torr on thoroughly cleaned glass substrates that had been previously patterned wit h an indium tin oxide (ITO) layer. To ensure uniform distribution of the silver metal in the organic matrix throughout the thickness of the films (200 nm), the co evaporation process was performed at constant and steady (<5% change) deposition rates for bo th materials. Next, a thin Al layer of 100 nm was evaporated through a shadow mask as the top contact electrode. All materials were used as purchased from eRay and Sigma Aldrich. To achieve different doping concentrations, the deposition rate of Ag was varied from 0.2 to 0.64 /s while the rate of Alq 3 was fixed at 4.0 /s yielding composite organic/inorganic films with a Ag/Alq 3 ratio varying from 5 to 20 volume percent. The thicknesses of the films were confirmed with a Tencor profilometer. Annealing was carried on inside a nitrogen glove box on a hot plate. The temperature was set to 100 C and each substrate was annealed for different periods of time from 10 minutes up to 3 days.
64 The capacitors were stored in an inert nitrogen environment inside a g love box during the entire processing stage. The dielectric constant and the dissipation factor of the composite thin films were measured in the ambient using a Hioki 3532 50 LCR meter, at frequencies ranging from 100Hz to 1MHz. I V measurements were car ried out on the devices at room temperature s using a Keithley 2400 source meter. A JEOL TEM2010F high resolution TEM was used to characterize the microstructure of the composite films and to analyze how it affected their dielectric properties. The compos ite films were also imaged and analyzed with electron diffraction before and after annealing. The samples made for TEM had to be thinner (~ 20 nm ) to ensure electron transparency and were deposited directly on top of a carbon coated copper TEM grids. 2.2. 2 Results and Discussion As shown in Figure 2 2, silver is uniformly distributed in the Alq 3 matrix. There are some darker spots that represent a higher content of silver, but overall the films are uniform. The images clearly indicate that using co evapor ation is a good method to uniformly distribute the filler in a given matrix. However, for the purpose of high K dielectrics, having just uniformly distributed metal atoms in such an organic matrix increases the conductivity and the dissipation losses of t he films too much. Instead, discrete metal particles are needed in order to maintain low dissipation losses and still benefit from the presence of the metal filler. In order to induce phase separation between the metal and the organic components, the com posite films were annealed. Thermal annealing enables inter diffusion of the components and enhances phase segregation of the silver, leading to the formation of silver nanocrystals. Figure 2 3 shows the dramatic change in the film morphology after annea ling. The TEM images
65 also show that after the nanocrystal formation, there is a uniform distribution of the nanocrystals in the film over a larger area. This result shows that the uniform distribution of metal nanocrystals inside the matrix is better tha n using colloidal nanocrystal/polymer blends. Moreover, t here is no indication that silver would prefe rentially diffuse towards specific end s of the film; in fact, silver would diffuse in all directions due to the amorphous nature of the organic matrix an d the high diffusivity of silver. In this manner, we can assume an even dis tribution of the metal nanocrystals in the entire film volume In order to find out if the nanostructures were in fact crystalline, we took the electron diffraction patterns of t he composite films and compared the data before and after annealing. Figure 2 4 shows the electron diffraction patterns for the composite films for high concentration of silver before and after annealing. For comparison, the electron diffraction of a nea t amorphous Alq 3 film is also shown in Figure 2 5. The diffraction patterns in Figure 2 4 correspond to the diffraction pattern of silver, confirming the presence of the metal in the films. In addition, the sharpening of the rings and the strong diffract ion spots after annealing clearly indicate and confirm the crystallinity of the silver nanostructures. Figure 2 4 also shows a zoomed image of evidence of the crystalli nity of the nanostructures formed by this method. An additional advantage of this method is that the nanocrystal concentration can be increased by simply increasing the doping concentration during the co evaporation. Phase separation can also be contr olled by annealing at different temperatures and for different annealing times. Figure 2 6 shows the effect of annealing the composite films
66 with different contents of silver concentration for different amounts of time. Increasing the concentration of na nocrystals has a direct effect on the dielectric properties of the films by increasing the dielectric constant and the dissipation factor. However, the dissipation losses of the films are decreased after annealing for long periods of time. This decrease in dissipation losses indicates that the phase segregation is further improved with longer annealing times. As a result, the formation of metal nanocrystals can be controlled by annealing for long periods of time, even with higher metal concentrations in the films. 2.3 Synthesis of Dense Organic Nanostructured Films In recent years, organic materials have become very popular due to their unique properties, versatility, and their impact as potential alternative materials for optoelectronic devices. Specif ically, there has been great interest in the potential applications of organic materials in solar cell devices and photodetectors. In the field of photovoltaics, there have been multiple efforts to find ways to control the morphology and crystallinity of organic thin films. It has been shown in recent studies that, by controlling the morphology of the films in organic solar cell devices, the charge pathways from the heterojunctions to the electrodes can be altered such that the collection of charges becom es more efficient 66,102 Likewise, by controlling the morphology of the films, the surface area at the donor acceptor interface can be increased, leading to a higher number of exciton dissociation events 67,103 These changes in morphology usually require altering the film growth into some form of island growth to create rods, tubes, particles, or some type of nanostructure. Other attempts to modify the morphology of films have included enlarging the grain sizes in a specific direction, inducing nucleatio n and formation of nanocrystals, and others 102,104 It has
67 also been shown in recent studies that crystalline organic films have a higher charge mobility due to the order in their structure 105,106 Moreover, highly crystalline films can also be expected to have greater exciton diffusion lengths, which are very small in typical organic materials (~10 nm). In the past, different methods have been used to grow organic thin films and control their morphology and crystallinity. Some of these methods have inc luded MOCVD, vacuum thermal evaporation (VTE), metal seed induced growth of nanorods, substrate heating, post evaporation annealing, and growth from scaffolds, among others 107 109 Some of these methods can be very expensive and time consuming, while othe rs usually create film morphologies with features that are too big, not dense enough, or are non uniform across large areas. Furthermore, some of these methods are also very sensitive and give little space for error in order to efficiently reproduce the r esults. In this section, we have developed a simple alternate method to synthesize nanocrystalline films from organic molecules. The films that undergo this method show a high concentration of nanostructures that form a network all throughout the film. This process consists of co evaporating two materials to form a mixed bulk film. One of these materials is the desired material used to make the final film for any given device, while the second material is just filler that is removed afterwards. This se cond material should have a lower sublimation temperature than the first one in order to be removed efficiently by baking at the right temperature. By doing this, the resulting film is more like a network of nanostructures; hence, changing the morphology and surface area of the film. Figure 2 7 shows AFM images of a CuPc film grown through this method compared with a typical planar CuPc film. Utilizing this method, we have found that we
68 can create very uniform films with a high density of nanocrystalline features. In addition, this method is simple, very reproducible, and independent of which substrate is used to grow the films. 2.3.1 Procedure to Fabricate Nanostructured Film s C opper phthalocyanine (CuPc) was chosen as the testing material since it is a very common material used in organic solar cells. For the filler material, N,N' diphenyl N,N' di(m tolyl)benzidine (TPD) was chosen since it has an evaporation temperature that is 100 C lower than that of CuPc. Specifically, CuPc starts sublimating ar ound 280 C under vacuum, whereas TPD starts evaporating from liquid phase around 170 C. In order to clean the substrates they were sonicated with acetone and isopropanol for 15 minutes each and subsequently exposed to a UV ozone treatment for 20 minutes. The substrates were immediately introduced into an evaporator where the two different organic materials were co evaporated. CuPc was evaporated at a rate of 0.75 A/s while TPD was evaporated at rates of 0A/s, 0.25 A/s 0.5 A/s, and 2.25 A/s for differen t variations for the TPD as filler materials. Both source materials were heated by passing high current through tungsten boats. Sublimation and/or evaporation temperatures were not recorded, however the precise control of depositon rate was carefully mon itered in situ with gold crystal sensors. After the films were made, they were taken out from the evaporator and baked for 12 minutes each in a furnace at 210 C under a vacuum of 3 X 10 6 Torr. To obtain films with different filler concentrations ranging from 0 vol% to 75 vol%, the evaporation rate of TPD was changed accordingly. The total thickness of the films varied depending on the amount of TPD present on the film, but all samples contained the same amount of CuPc (30 nm) for comparison purposes. Th e co evaporated films
69 were also baked at different temperatures and deposited on different substrates to test the outcome of the process under different parameters. Specifically, the samples were baked at 170C and 210C while different substrates such as ITO, glass, MoO 3 and silicon were used to deposit the films. The process was also tested with another material, SnPc, and a clear effect on the films was observed as well. The nanocrystalline CuPc films (N CuPc) were characterized using the scanning el ectron microscope (SEM) and atomic force microscopy (AFM) to analyze the features of the films and changes in morphology. 2.3.2 Results and Discussion CuPc is a small molecule material that is crystalline in nature. In typical deposition methods, the growth of CuPc nanocrystals can be controlled by changing substrate temperature, pressure, and sublimation rate during the deposition of the material. However, it is a real challenge to maximize the number of nucleation sites to obtain a high density of crystals while at the same time be able to control their size, growth, uniformity, and thickness of the overall film. Furthermore, the films that are formed through post annealing methods do not have a high density of nanostructures, they are not unifo rm through larger areas, or they have crystals that are too big to be used in thin films 107 111 By co evaporating TPD with CuPc, we saw that the size of the structures formed changed drastically compared to what would be expected from annealing a pure Cu Pc film. Figure 2 8 shows the comparison between the crystals formed through post annealing methods and our co evaporation method. When pure CuPc films are annealed, the length of the crystals is on average three times as large and twice as thick as the crystals formed when TPD was present in the mixed layer, which ranges
70 from 20 50 nm in size. We also found that, by varying the relative amount of TPD mixed with CuPc, there is a significant difference in the film uniformity and density of crystals. Figur e 2 9 also shows the resulting CuPc films if they were co evaporated with 25 percent, 50 percent, and 75 percent TPD. When there is no TPD present in the mixed layer, the crystals and the film morphology are the least uniform and as the relative amount of TPD is increased, the uniformity of the films in the micro rage is increased as well as the number of crystals present per unit area. In order to broaden our understanding about the role of TPD on the changes in the morphology of the films, we also stud ied the effect of changing the rate at which TPD is evaporating away from the mixed layer when annealed. TPD turns into liquid phase at around 155C and then starts evaporating around 165 170C. We baked some samples at 175C, where it would take longer for TPD to evaporate completely. Other substrates were baked at 210C, at which TPD would evaporate within ten minutes. We found no significant differences between the samples baked at different temperatures. This implies that the effect that TPD has on the formation of the peculiar crystalline films is independent of TPD evaporation rate. Figure 2 10 shows N CuPc films reproduced on different substrates, including glass, indium tin oxide (ITO), and MoO 3 This result suggests that the process by which the nanocrystals are formed is more dependent on the amount of TPD present rather than on the substrate on which the films were deposited. Figure 2 11 shows a resulting film of SnPc that was synthesized through the same co evaporation method as the N CuPc films. The result also shows a marked difference in the film morphology compared to the simple post annealing of the SnPc film. Since SnPc is not as
71 crystalline as CuPc, the resulting SnPc films do not show needle like nanostructures. Nonetheless, this result indicates that the processing method does affect the film morphology and can be generalized for different organic materials. The reason for the marked differences on the films when the TPD:CuPc ratio is changed is not entirely understood, but it is believed that the mechanism is as follows. When the films are baked, there is enough thermal energy for the CuPc molecules to start crystallization. At the same time, TPD turns to its liquid phase, allowing the CuPc molecules to diffuse more efficiently Furthermore, when the TPD molecules start evaporating, they leave ga p s behind where they used to occupy space. Hence, when the relative amount of TPD is larger, it can be expected to observe the formation of a more open structure resembling an interpen etrating network of nanocrystals. This is, indeed, what we have seen in our films. This mechanism also explains the fact that the films are more uniform on the micro range when the amount of TPD is larger. As the relative amount of TPD is larger during the co evaporation, the CuPc molecules are more dispersed from each other on the film and are not allowed to come together to form big crystals. As a consequence, when the films are baked, the CuPc molecules form many more nanocrystals with smaller dimens ions and are more uniformly distributed across larger areas. 2.4 Summary In summary we have demonstrated two alternative and easy approaches to fabricate metal nanostructures embedded in organic matrices and to alter the growth of organic nanostructured films using simple thermal co evaporation. In the case for high K dielectric films from inorganic/organic composites, it was shown that optimized thermal annealing of co evaporated conductive metal in an
72 organic matrix enhances phase segregation of the m etal constituents. This segregation resulted in the formation of discrete metal nanocrystals that were successful on increasing the dielectric constant and reducing the dissipation factor. This method can be controlled by annealing temperature and anneal ing time. We have also developed a processing method that allows us to modify the tendency to crystallize. Different materials can be synthesized using this meth od and the results can be reproducible on different substrate materials, and with good uniformity. Figure 2 1. Image of one thermal evaporator used in this work.
73 Figure 2 2. TEM images of Alq 3 blended with silver. a) Blended with no silver. b) B lended with 5% silver. c) Blended with 20% silver. Figure 2 3. TEM image of Alq 3 blended with silver after annealing. a) Blended with 5% silver. b) Blended with 5% silver (larger area). c) Blended with 15% silver.
74 Figure 2 4. TEM electron diffrac tion patterns from films of Alq 3 blended with silver. a) Blended with 20% silver before annealing. b) Blended with 20% silver after annealing. c) Blended with 50% silver after annealing. d) Image showing the ordered atomic planes in the silver nanocrys tals.
75 Figure 2 5. TEM electron diffraction pattern from a neat amorphous Alq 3 film. Figure 2 6. Plot showing the effect of annealing time on the dissipation factor and dielectric constant of composite films with different silver content.
76 Figure 2 7. CuPc films made by different methods. a) As deposited neat film. b) Co evaporation and re evaporation of film. Figure 2 8. CuPc films annealed at 210 C after being deposited. a) Neat CuPc film. b) CuPc:TPD film.
77 Figure 2 9. Effect of CuPc:TPD co evaporation ratio on the resulting film uniformity. a) 75:25. b) 50:50. c) 25:75.
78 Figure 2 10. N CuPc films grown on different substrates. a) ITO. b) Glass. c) MoO 3 Figure 2 11. SnPc films treated with diffe rent methods. a) As deposited neat film. b) Annealed after being deposited. c) Annealed after being co evaporated with TPD.
79 CHAPTER 3 EFFECT OF NANOCRYSTA LLINE, HIGH SURFACE AREA DONOR ACCEPTOR HETEROJUNCTION INTER FACE IN ORGANIC SOLA R CELLS 3.1 Why U se Interpenetrating Heterojunction Interfaces? As mentioned previously, there has been great interest and multiple efforts control the morphology and crystallinity of organic films in photovoltaic cells. Specifically, there have been efforts to control th e heterojunction interface in organic solar cells so that it resembles ordered interpenetrating structures 66 67,102 103,112 The idea behind this approach is similar to the formation of a bulk heterojunction in that the purpose is to increase the donor ac ceptor surface area and minimize the distance that excitons need to diffuse so that they can dissociate before being lost to recombination processes. However, this approach is different from bulk heterojunctions because ordered structures also have the ad vantage of collecting all charges, where as in bulk heterojunctions there might be isolated phase segregation or cul de sacs that result in poor charge transport to the electrodes. However, the efficiency of the devices with these interpenetrating nanos tructures is still relatively low compared with bulk heterojunction solar cells. These low efficiencies could be explained because of the long lateral distance between nanostructures is still too big for efficient exciton diffusion. Another reason may be poor infiltration of the second material through the nanostructures, leaving behind gaps of unfilled regions that limit exciton dissociation and transport. In this chapter we have used the previously described CuPc nanocrystalline structures to fabricate organic solar cells. For the fabrication of the devices, we used [6 ,6] phenyl C 61 bytyric acid methyl ester (PCBM), which is a solution processed derivative of C 60 in order to ensure better infiltration through the N CuPc films. Here, an increase in PCE of 54 percent was
80 observed in the devices with the N CuPc compared to the devices fabricated with a simple planar bilayer structure. 3.2 Fabrication of Bulk Heterojunction Organic Solar Cells To fabricate devices, ITO substrates were cleaned and treated as described in chapter 2. Prior to the co evaporation of CuPc and TPD, the substrates were coated with either PEDOT or MoO 3 In case of PEDOT, the aqueous solution was spincoated in a hood at a speed of 8000 rpm, which creates a film of approximately 30 nm. After the spin coating of the PEDOT film, the substrate was annealed at 180 C for 20 minutes. For the case of MoO 3 The substrates were loaded into the evaporator chamber, where MoO 3 was evaporated at a rate of 0.2 A/s under 1X10 6 Torr. In case of MoO 3 the films were 10nm in thickness. Following the deposition of PEDOT or MoO 3 the co evaporation of the CuPc:TPD films took place and subsequently the substrates were annealed at 175 C as described in chapter 2 in order to recreate the nanocstructured films. After annealing, the devices were introduced into a glovebox where the PCBM layer was spincoated on top. This PCBM solution was made with chlorobenzene at a concentration of 22.0 mg/ml and was filtered with a 0.45 m PVDF filter. To spincoad the f ilm, we used a speed of 2000 rpm for 60 seconds with a first speed of 700 rpms for 3 seconds. Figure 3 1 shows the evaporators and the glovebox used for these experiments. As the PCBM infiltrates the nanosctructure, the interpenetrating heterojunction is formed. After this, the films would be annealed at low temperatures (40 50 C) to dry the presence of any solvent away. The samples would then be introduced into an evaporator once again to deposit the aluminum cathode. In case of the aluminum, we eveapo rated 100 nm of Aluminum at a rate of 1.0 A/s under 1X10 6
81 Torr. The aluminum was evaporated from a tungsten dimple boat. The thickness of the N CuPc layer was also varied to study the effectiveness of the films for thick devices. The devices were char acterized with cross section SEM to analyze the features of the films and keep track of the interfaces in the device structure. We also kept track of the relative amount of TPD left in the films after being processed and baked by taking their absorption s pectrum. In order to get the photo I V data, the devices were characterized with a Keithley 4200 semiconductor analyzer under AM 1.5G sun illumination To obtain the spectral response, a Newport monochromatic system connected to a Stanford Research Syste m SR810 DSP Lock in amplifier was used. A chopper running at 400 Hz was used. 3.3 Results and Discussions Since the N CuPc films were fabricated through the method described in chapter 2, the films were exposed to large amounts of TPD. It is also very important to have pure films in order to have good performance in solar cells. Therefore, to make sure that the films were pure after the fabrication of the N CuPc films, the amount of TPD that was left in the films after annealing had to be tracked. Fig ure 3 2 shows the absorption of a TPD:CuPc mixed layer composed by mostly TPD, recreating the conditions to fabricate the N CuPc films (25:75% CuPc:TPD). For comparison, Figure 3 2 also shows the absorption spectrum of plain CuPc and TPD films with simila r thicknesses as in the mixed film. It is important to note the relative difference in absorption intensity between the TPD and CuPc regions and also the overlap between TPD and CuPc around 350 nm. Figure 3 3 shows the comparison between the absorption s pectrum of a plain CuPc film against that of the exact same mixed film from Figure 3 2 after it had been annealed for 15 minutes. The absorption spectra show that there is very little difference
82 in absorption between the two films around 350 nm, which is the range where the TPD aborption is This result implies that most of the TPD, if not all of it, evaporates out of the film during annealing. To get a better idea of how much TPD might still be present in the N CuPc films, Figure 3 3 also shows the abso rption data of a 10 percent TPD doped with CuPc for comparison. The absorption of th is mixed film containing 10 percent TPD is much larger around 350 nm than the absorption of the N CuPc film after being annealed. It is also worth noting that the main Q band absorption peaks of the N CuPc film show pronounced shoulders at the edges. These shoulders may indicate that the intermolecular interactions are stronger in these films 113 114 Before fabricating the solar cells with the N CuPc films, planar bilaye r devices were fabricated to determine the effect of MoO 3 as a hole extraction layer for the CuPc/PCBM system based on previous reports 115 117 Figure 3 4 shows the J V characteristics for solar cells with PEDOT, MoO 3 and without any interlayer. From th e results, it was determined that MoO 3 serves well to increase the efficiency of the solar cells and it was chosen for the reference cell because of increased fill factor. Table 3 1 shows the detail parameters of the solar cells. The structure of the fi rst nanostrucured solar cell fabricated consisted of ITO/MoO 3 (10nm)/N CuPc (20nm)/PCBM(60nm)/Al. Unfortunately, this device resulted in very poor performance since the N CuPc film is not continuous and has bad contact at the interface with MoO 3 To ensu re a better contact and fix this problem, a planar CuPc film was introduced in between MoO 3 and the N CuPc. In order to minimize crystallization of this planar CuPc film during the annealing and fabrication of the N CuPc film, a lower temperature of 175 C was used. Figure 3 5 shows the cross section
83 SEM image of an N CuPc after being annealed, with a planar CuPc film underneath. After the planar CuPc film was incorpored into the structure of the device, different thicknesses of N CuPc were tested. Figu re 3 6 shows the J V characteristics of the solar cells with different N CuPc thicknesses and table 3 2 shows the photovoltaic parameters for the respective devices. The results indicate that as the N CuPc film gets thicker, the series resistance increase s and the fill factor decreases. This effect is counterintuitive since the cells were expected to perform better with thicker N CuPc layers. However, it may be due to poor infiltration of the PCBM throughout the nanostructures. It is important to note t hat the nanostructures are randomly oriented, making it even harder for adequate infiltration. Nonetheless, the data shows that the devices with the N CuPc layer performed better than the planar device. There was an associated increase in short circuit c urrent and the fill factor compared to the planar device, which comes from the increased exciton dissociation and charge collection. Since the average distance that the excitons have to diffuse to reach the heterojunction interface is shorter when N CuPc films are used, more excitons can be dissociated. Additionally, the crystalline nature of each nanosctructure may improve the exciton diffusion length for CuPc. As a result, we see an improvement in the short circuit current. To further analyze the sour ce of the increase in the short circuit current in our device; we measured the spectral response and the external quantum efficiency. Figure 3 7 shows the EQE for both the control device and the device with the N CuPc film. It is clear to see that the N CuPc device has higher efficiency. It is important to note that most of the improvement in efficiency originates at the wavelength range where CuPc is absorbing light. In contrast, PCBM does not contribute much to the
84 improvement in efficiency, which ind icates that most of the enhancement in efficiency is due to the crystalline nature of the N CuPc layer. 3.4 Summary In summary, we demonstrated an alternative processing method that allows us to modify the morphology and crystallinity of organic films i n the fabrication of solar cells with a nanocrystalline structured heterojunction interface. The purity of the N CuPc films was roughly estimated tracking the absorption of TPD and it was found that almost all TPD gets removed after annealing of the films The organic photovoltaic cells that were fabricated with the N CuPc films showed an improvement compared to the control devices, but the nanostructures are not ideal for the device since they are randomly oriented throughout the film. Additional improv ement may also be obtained if exciton blocking layers were used in the device structure. Figure 3 1. Evaporators and glovebox used in the fabrication of the nanostructured CuPc films. a) Stand alone evaporator. b) Evaporator linked to a glovebox.
85 F igure 3 2. Absorption spectrum of a CuPc:TPD co evaporated film compared with a plain CuPc film and a plain TPD film. Figure 3 3. Absorption spectrum of a CuPc:TPD after annealing compared with a plain CuPc film and a slightly doped CuPc with TPD.
86 F igure 3 4. J V characteristics for CuPc/PCBM bilayer solar cells with and without hole extraction interlayers. Figure 3 5. Cross sectional SEM images of nanostructured CuPc films. a) Single nanostructured CuPc film. b) Planar CuPc film underneath a nanostructured CuPc film. Scale bar is 100 nm.
87 Figure 3 6. J V characteristics for N CuPc/PCBM solar cells with different N CuPc layer thicknesses. Figure 3 7. External quantum efficiency for a planar CuPc solar cell compared with a N CuPc solar c ell.
88 Table 3 1. Solar cell parameters for CuPc/PCBM solar cells with interlayers. Table 3 2. Solar cell parameters for N CuPc/PCBM solar cells with different N CuPc layer thicknesses.
89 CHAPTER 4 SYNTHESIS SETUP FOR INORGANIC SEMICONDUCTOR COLLOI DAL NANOCRYSTALS AND THEIR CHARACTERIZATION 4.1 Introduction Spherical semiconductor nanocrystals (NCs), also known as quantum dots (QDs), have attracted an increasing amount of scientific interest because of their unique size dependent optical and electro nic properties and their potential in biological applications and electronic devices. One of the first successful methods for controlling the growth of semiconductor NCs was developed by Steigerwald and Brus in the 1980s. The reaction consisted of mixing metal and chalcogen organic compounds in organic coordinating solvents and heating to reflux 118 119 However, Murray and Bawendi were the first to describe the injection method to produce highly crystalline and monodisperse colloidal NCs 37 In their work they prepared CdE (E = Cd, Se, Te) nanocrystals by quickly injecting precursor solutions into a preheated hot reaction medium. This method resulted in rapid nucleation and separation of the nucleation and growth stages, as described in chapter 1 of this dissertation. Later, Alivisatos et al., showed that monomer concentration and surfactants play key roles in controlling the size distribution, shape, and other properties of the NCs 46,49 50,120 Today, different II IV, III V, and IV VI group semiconducto r NCs are being synthesized through the rapid injection method with slight variations in the materials used for the process or in the process itself This chapter describes the setup and the procedure used for the synthesis of PbSe NCs through this rapid injection method. 4.2 Nanocrystal Synthesis Setup Figure 4 1 shows an image of the entire setup used to synthesize the nanocrystals for used in this dissertation. It can be seen that the setup consists of multiple
90 components. Some of these components a re more critical than others, but all of them will be described for the sake of understanding the system. An important requirement for any NC synthesis is that the reactions should take place in an inert environment to avoid contamination of the reaction and oxidation. To provide this clean environment, a constant argon gas flow is supplied to the reaction. To ensure optimal results, it would be ideal that the setup should be inside a glovebox, but since it was not possible, ultra high purity argon gas w as used instead. In addition, a bubbler containing some mercury was adapted in between the gas line from the tank and the reaction vessel to filtrate the gas from any particulate impurity. A condenser was also placed in between the bubbler outlet and con nected directly to the reaction vessel. This condenser was filled with cold water to prevent any vapors form the reaction to escape. Figure 4 2 shows close up images of the bubbler and the condenser. For the reaction vessel, a 100 ml 3 neck round flask was used. During a reaction, the three necks are sealed by connecting them to the condenser, a septum stopper, and a thermocouple holder. However, to ensure a constant gas flow, a needle is placed through the septum stopper. Another important aspect of t he synthesis is to have good control of the temperature of the reaction vessel. To do so, a heating pad is placed under the 3 neck flask for optimal heat transfer. The current applied to the heating pad is controlled by a J KEM temperature controller, wh ich is also connected and receiving feedback from a thermocouple that measures the temperature inside the 3 neck flask. Figure 4 3 shows the 3 neck flask with all the connections to it. To ensure that there is an accurate reading of the temperature, the thermocouple is fitted tightly into the glass thermocouple
91 holder, which is dipped in the reacting solution. Typically, the reaction is stirred to have uniform mixing and heating. 4.3 Characterization of the Nanocrystals The characterization of NCs inclu des the identification of the crystal morphology, the crystalline structure and the measurement of optical properties. This section focuses on the characterization of the NCs with transmission electron microscopy (TEM) and spectrophotometry, which are two important techniques used in this research. 4.3.1 Transmission Electron Microscopy (TEM) Transmission electron microscopy (TEM) provides a direct measure of the morphology of NCs, including the size, the size distribution and the shape of the particles, as well as their crystal lattice spacing. From an image, the sizes of the particles can be determined by comparison to the scale bar; the size distribution can be evaluated from the statistical measurement of the particles. The resolution of a TEM is defined by the Rayleigh criterion: (4 1) where is the wavelength of the radiation, is the refractive index of the viewing medium and angle of collection of the magnifying lens. For TEM, the refractive index and semi angle are fixed; thus, the resolution is mainly dependent on the wavelength of the radiation source. In turn, the electron wavelength is related to the electro n energy, described as: (4 2)
92 w here the energy is in electron volts (eV) and is determined by the accelerating voltage of the microscope. This means that as the accelerating voltage is higher, the resolution of the microscope is h igher 121 Figure 4 4 illustrates the components of a basic TEM system. An electron beam is formed at the gun and it reaches the specimen, where it interacts with the specimen and is transmitted. The transmitted electrons interact with a series of lenses a nd apertures to create diffraction patterns or images on a screen. If the back focus plane of the objective lens acts as the object plane for the intermediate lens, then the diffraction pattern will be obtained on the screen. If the image plane of the obje ctive lens acts as the object plane of the intermediate lens, an image will be projected onto the screen. Figure 4 5 shows the two scenarios for diffraction mode and image mode. Since the TEM images are obtained by the direct electron beam, the sample mu st be made sufficiently thin to allow electron transparency. The samples were easily prepared by diluting an NC solution and simply dropping a drop of the solution on TEM grids and leaving them under vacuum overnight before taking them to the instrument. The TEM grids were purchased from Ted Pella Inc. and were carbon coated, type II copper grids. Figure 4 6 shows an image of a copper grid used to prepare the TEM samples. 4.3.2 Spectrophotometry Spectrophotometry can provide a direct measure of the NC op tical and electronic properties, including the absorption spectra of the NCs, the particle size, and particle size distribution. From the absorption spectra of the nanocrystals, it is also possible to approximate the energy band gap of the nanocrystals by calculating the energy of the
93 absorbed photons at the first excitonic absorption peak of the NCs. The energy of a photon can simply be calculated by: (4 3) where h, c, and are the Planck constant, the speed of light in, and the photon wavelength, respectively. Figure 4 7 shows the characteristic absorption spectrum of NCs grown at different temperatures. When the the absorption peak is broad, it is indicative that the particle size distribution is polydisperse in nature. On t he other hand, if the absorption peak is narrow and there are clear secondary absorption peaks, the particle size distribution is monodisperse. Absorption measurements in spectrophotometry work on the basis of light being shone and interacting with a s ample, before being collected by a photodetector. In a typical setup, light from a broadband source lamp is passed through a monochromator, which diffracts the light into different wavelengths. Then, these discrete wavelengths of light are passed selecti vely through the sample and the intensity of the transmitted light is measured with a photodetectors. In order to have an accurate measurement of the light absorption, the measurement is compared against a reference sample. The NC samples are usually dil uted solutions in a solvent that does not have interference background absorption with the NC material, such as tetrachloroethylene (TCE). As an alternative, the samples may also be prepared in thin film form coated on glass substrates.
94 4.4 Synthesis of PbSe Nanocrystals The synthesis of PbSe NCs was achieved using a modified version of established organometallic synthetic routes 122 For a typical reaction, t ri n octylphosphine (TOP) is mixed with elemental sel enium (Se) powder to form TOP=Se as the selen ium precursor. In our case, we mixed 1.975 g of selenium powder into a 25 ml small bottle of TOP to make a 1M TOP=Se solution. To increase the yield of PbSe NCs, diphenylphosphine (DPP) may also be mixed into the TOP=Se solution as a reducing agent that increases the amount of elemental species that form the monomers 122 The TOP=Se solution was mixed inside a glovebox to avoid oxidation and was sealed with a septum stopper and paraffin wax while it was left stirring vigorously for 24 hours. To prepare th e lead precursor, lead oxide powder was mixed with oleic acid (OA) and octadecene (ODE) as the solvent. The lead precursor solution was mixed in a 3 neck round flask and left degassing under heating and a constant flow of argon gas. After the lead soluti on had been degassing for 30 minutes and had reached the reaction temperature, the TOP=Se was quickly injected into the 3 neck flask. The initially clear solution quickly turned dark brown and then black over time as the PbSe NCs formed. The reaction was terminated by diluting the solution with a cold solvent such as toluene or chloroform. Finally, the NC solution was washed from unreacted precursors and byproducts by centrifuging several times. To synthesise approximately 2 mmol of material, we followe d the following recepe: 0.444 mg of PbO mixed with 1.66 ml of OA and 15 ml of ODE for the lead oxide precursor solution. This solution was heated to 160 C for bigger particles (~5 6nm) or to 120 C for smaller particles (~3nm). For the TOP=Se solution, w e used 6 mmol of the solution (6ml) and mixed it with 56 l of DPP. Once the PbO precursor
95 solution reached the right temperature, the TOP=Se solution was rapidly injected with a large syringe with 18 guage needle and the solution was left to react for 1 minute. The solution would then be poured into 15ml of cold toluene (~30 F) to stop the reaction. As mentioned before, the quality of the NCs could be affected by the relative concentration of elemental species present in the solution as well as the co ncentration of surfactants. In order to optimize the recipe to make high quality spherical NCs, different ratios of precursor concentration as well as surfactant were used. The final size of the NCs was determined by precise control of temperature and re action time. Figure 4 8 shows TEM images of PbSe NCs synthesized with different ratios of precursor material. When the reactivity of the precursors is not similar, a slow nucleation occurs and a polydisperse batch results. On the other hand, if the ratio of the precursors is tuned, they will react more efficiently and nucleation will be faster, resulting in a more monodisperse batch of nanocrystals. Figure 4 9 shows TEM images for NC synthesized at the same temperature, but left growing for different amo unts of time. It is clear that if reaction time is increased, the NCs will grow accordingly. However, if the NCs are left growing for too long, monodispersity might be lost due to the Oswalt ripening effect. 4.5 Summary In summary, the nanocrystal setup s ystem for the synthesis of colloidal nanocrystal through the quick hot injection method was described in detail. In addition, the procedure to synthesize PbSe nanocrystals was also described in detail. Different synthesis conditions resulted in varying N C quality batches during the process to optimize the standard synthesis recipe. The nanocrystals were characterized primarily
96 through transmission electron microscopy and spectrophotometry to measure their morphology and optical properties. Figure 4 1. Colloidal nanocrystal synthesis setup. Figure 4 2. Components for the NC synthesis setup. a) Bubbler. b) Condenser.
97 Figure 4 3. Three neck round flask connected to the condenser, thermocouple, and heating pad.
98 Figure 4 4. Illustration o f a TEM system with the main components. Image extracted from reference 160. Figure 4 5. The electron optical system for a TEM. a) Diffraction mode. b) Image mode. Images extracted from reference 161.
99 Figure 4 6. Copper TEM grids. Images extracte d from reference 162. Figure 4 7. Typical absorption spectrum of PbSe NCs. The higher the reaction temperature, the NCs will absorb at longer wavelengths. Figure 4 8. TEM images from different NC synthesis with different precursor concentration rat ios. a) Polydisperse NCs. b) Monodisperse NCs.
100 Figure 4 9. TEM image s of PbSe NCs grown at the same temperature but left growing for different times. a) NCs grown for 30 seconds. b) NCs grown for 90 seconds.
101 CHAPTER 5 THE EFFECT OF LIGAND EXCHANGIN G SOLUTION TREATMENT S ON COLLOIDAL INFRARED PBSE NANOCR YSTAL FILMS AND PHOT ODETECTORS 5.1 The Problem with Nanocrystal Capping Ligands The ability to detect infrared light with high sensitivity is of paramount importance for diagnostic and therapeutic medi cal, remote sensing and night vision applications. Solution processed light sensing semiconductor optoelectronic devices offer the possibility of low cost manufacturing and compatibility with flexible electronics. While significant progress has been made in solution processed organic based photodetectors the bandgap of these materials limits their use to wavelengths shorter than 1 .3 m 123 124 Colloidal inorganic nanocrystals based on III V and IV VI semiconductors can overcome this limitation by extendi ng the detector sensitivities to longer wavelengths 125 126 These nanocrystals can be synthesized using wet chemical techniques like the one described in chapter 4 and can be processed from solutions at low temperatures which makes them cheap to produce. They also offer the advantage of bandgap tunability over a wide range of wavelength s due to the quantum size effect, allowing for spectral selectivity. However, o ne of the challenges that needs to be overcome in employing colloidal NCs in optoelectronic d evices is the poor charge transport properties of neat nanocrystal films 127 Due to the characteristic hopping mechanism 128 responsible for charge transport between nanocrystals, the electronic nature of the capping ligands plays an important role in det ermining the transport properties. The spatial separation between neighboring nanocrystals determines the width of the energy barrier that carriers need to tunnel through to reach an adjacent nanocrystal. This affects the overall mobility as well as the c onductivity of the nanocrystal films. Semiconductor nanocrystals are typically synthesized with long
102 aliphatic ligands to enable easy processing and prevent aggregation in solution. However, since the spacing between individual nanocrystals is determined by the length of the organic ligands, films made from such nanocrystals are highly insulating and generate poor photocurrent Exchange of the long ligands with shorter and preferably more conducting ones is required to ensure good carrier transport in the nanocrystal films. Recent ly, use of thiol terminated capping groups and other short ligands to improve the responsivity of nanocrystal photodetectors has been reported 78, 129 131 However, processing of high quality NC thin films is problematic and chall enging when replacing NC capping ligands. Two main considerations need to be kept in mind while performing such an exchange of surface ligands : passivation of the nanocrystal surface s during the ligand exchange reaction in order to maintain their chemical stability; and loss of free volume leading to the development of cracks in the nanocrystal films. In addition to processing problems, NC yield, and loss of NC quantum confinement, the exchange of capping ligands with short ones usually also results in an increase in dark current for the devices, which is counteractive for photodetector applications. However, the issue with high dark current in photodetectors will be addressed more in detail in chapter 6. In this chapter we study the effect of different organic surface capping ligands including oleic acid, octylamine, ethanedithiol (EDT) and benzenedithiol (BDT) on film formation. In addition, we have also studied the effects on the electrical properties of the treated films and the device characteristic s of the resulting infrared photodiodes. 5.2 Processing and Characterization of Nanocrystal Films and Photodetectors We investigated the effect of capping moieties on the device characteristics of PbSe nanocrystal infrared photodetectors. The colloidally synt hesized PbSe
103 nanocrystals used in this study are passivated with long chain aliphatic oleic acid capping groups. Since close packing among nanocrystals is required for efficient charge carrier transport, we replaced the insulating oleate ligands with short er ligands such as octylamine, ethanedithiol and benzenedithiol. Exchange with octylamine can be achieved in solution, but the dithiol ligands render the NCs insoluble so they have to be exchanged in film form. 5.2.1 Synthesis of PbSe Nanocrystals PbSe nanocry stals were synthesized following the procedure described in chapter 4 and adding DPP as a catalyst. In the reaction, lead oxide (2 mmol) was dissolved in a mixture of octadecene (15 ml) and oleic acid (6 mmol) under uniform h eating, vigorous stirring and a constant flow of argon Once the temperature reached 140C, 6 mmol of the TOP=Se mixed with 56 l of DPP were rapidly injected into the mixture to initiate nucleation of nanocrystals. The size of the nanocrystals is determined by the reaction time wh ich we range from 40 seconds up to 2 minutes. To terminate the reaction cold toluene was injected into the reaction mixture. The resulting nanocrystals were subsequently washed by precipitating with acetone centrifuging at 8000 rpm, dumping all the supe rnatant, and then re dispersin g in toluene three to four times. 5.2.2 Exchange on Nanocrystal Capping Ligands in Solution During the ligand exchange reaction in solution, the bulky (~2 nm in length) oleate ligands were exchanged with shorter chain (~1 nm) octylamine. This post synthetic solution phase exchange was performed in a nitrogen glovebox. To replace the ligands with octylamine, the nanocrystals were precipitated using acetone and then redispersed in 10 ml of pure octylamine. The ligand exchange s olution was left stirring at room temperature for 48 hours to allow the maximum exchange of ligands. Subsequently, the
104 nanocrystals were precipitated with acetone and finally redispersed in chloroform, yielding a typical concentration of 60 mg/ml. The ex change of oleate passivating groups with octylamine resulted in a clear dispersion with no agglomeration of nanocrystals. It is important to keep special attention to the solvents used as old solvents or strong polar solvents can strip the weak amine grou ps from the NCs. 5.2.3 Solid state Capping Ligand Exchange Treatment and Device Fabrication To treat the NC films with the dithiol ligand exchange treatment, PbSe nanocrystals with either oleic acid or octylamine capping groups were spin coated on UV ozon e treated ITO coated glass substrates inside a nitrogen glove box. To spin coat the 60 mg/ml QD solution, we used a speed of 1500 rpm for 40 seconds with an initial rate of 700 rpms for 3 seconds. Subsequently, the samples were left to dry for 5 minutes and then immersed in a 0.1M solution of EDT or BDT in acetonitrile for 30 seconds. The films were then rinsed with acetonitrile or chloroform to remove any loose agglomerates and residue left from the reaction. Specifically, the films would be spin coated with bare chloroform solvent at speeds of 500 rpm. The samples were subsequently dried inside the glovebox by annealing at 40 50 C for 10 15 minutes As briefly mentioned before, this solid state treatment rendered the nanocrystal layer insoluble in solv ents used for further device processing and multiple nanocrystal layers were able to be spincoated on top of each other. Figure 5 1 shows a schematic of the steps to process an NC film and treat it with a dithiol ligand exchange. The resulting films had a thickness of ca. 200 nm, as measured with a surface profilometer. To fabricate devices, a 100 nm thick Al cathode was thermally deposited on top of the NC film in a vacuum evaporator chamber at a deposition rate of 1 A/s and a pressure of
105 ~10 6 Torr thro ugh a shadow mask with an active area of 4 mm 2 Figure 5 2 shows a schematic of the structure of the devices (ITO/PbSe/Al) once they had been finished. To study the carrier transport properties, single carrier devices were also fabricated. To fabricate ho le only devices, the Al electrode was replaced by an Au cathode (50 nm). The gold was deposited at a rate of 0.5 A/s and also at a pressure of ~10 6 Torr. To fabricate electron only devices, the ITO substrates were cleaned without UV ozone treatment to s uppress hole injection. 5.2.4 PbSe Nanocrystal Film and Device Characterization Achieving defect free, spatially uniform, smooth films is a key requirement to process good quality devices. Due to the difference in the length of the ligands and their che mical nature, it is expected that different ligands lead to different film morphologies. The effect of different surface passivating ligands on the interparticle distance, optical properties, and the film morphology quality was investigated through spectro photometry, scanning electron microscopy (SEM), atomic force spectroscopy (AFM), and transmission electron spectroscopy (TEM). To confirm the exchange of organic ligands on the surface of nanocrystals Fourier transform infrared spectroscopy (FTIR) was us ed. The FTIR samples were prepared by spincoating the NC films on polished aluminum substrates. To investigate the penetration depth from the dithiol solution treatments, Auger spectroscopy (AES) was performed to do depth profiling of the dithiol treated NC films. For this measurement, the NC films were spincoated on silicon substrates. The current density voltage ( J V ) characteristics of the devices were measured with a Keithley 4200 semiconductor parameter analyzer. The devices were irradiated with mon ochromatic light from a Newport monochromator using an Oriel solar simulator
106 as a source For photocurrent measurements, the devices were illuminated with monochromatic light at 830 nm with an intensity of ~200 w/cm 2 The i llumination intensities were mea sured using two calibrated Newport 918D photodiodes one for the visible and the other for the infrared part of the spectrum. To obtain the spectral response of the photodetectors, light from the monochromator was chopped at 400 Hz to modulate the optical signal. The p hotocurrent response as a function of bias voltage was measured using a Stanford Research System SR810 DSP lock in amplifier. In order to obtain further insight into the mechanism of charge transport in the devices, the dark J V characteristi cs were measured as a function of temperature using a cryostat interfaced with a LakeShore 321 Autotuning temperature controller. The o perating temperature was varied from 80 K to 310 K. 5.3 Results and Discussion 5.3.1 Effect of Ligand Exchange Processes o n Film Properties Figure 5 3 shows the structure of the different capping ligands used in this work and the TEM images of the nanocrystals capped with each of them. As is evident from the TEM images, replacing the oleate ligands with octylamine clearly r educes the spacing between adjacent nanocrystals. It is important to note that in the case of the dithiol capping group in Figure 5 3c (BDT), the inter particle spacing is further reduced because the two thiol end groups act as a bridge and cross link the adjacent nanocrystals. This cross linking reaction results in nanocrystal films that are insoluble in common organic solvents. However, i n order to obtain smooth, defect free, good quality films from a solution of dispersed nanocrystals, one also needs t o consider the loss of free volume during the ligand exchanging process. We observed that the ligand exchange of oleic acid with dithiol ligands results in cracks and pinholes formation due
107 to the strain developed in the films as a result of the solid sta te ligand exchange reaction These effects are apparent in figure 5 4, which shows AFM images of oleic acid capped nanocrystal films treated with the dithiol solutions. To avoid this detrimental effect on the NC film morphology, we add an intermediate ste p to exchange the oleate capping groups with shorter ones (octylamine) prior to the dithiol treatments. When this intermediate solution phase exchange reaction was performed prior to the treatments with EDT or BDT, pinholes and film cracking were greatly reduced, as shown in figure 5 5. It is also worth noting the difference in film quality between the EDT treated films and the BDT treated films. Independent of whether the films were deposited from a batch of oleic acid capped nanocrystals or octylamine capped nanocrystals, it is clear from figures 5 4 and 5 5 that the EDT treatment affects film morphology more strongly than the BDT treatment. Two factors should be noted: first, the short dithiol molecules replace the longer capping ligands; second, the dithiol molecules also draw the nanocrystals closer together by cross linking them with their thiol end groups. Since EDT molecules are shorter and less bulky than BDT, the free volume loss in the EDT treated films is greater, resulting in greater residua l stress compared with the BDT treated films. As a result, we see that in identical nanocrystal films (figure 5 4) cracking is more prevalent in the EDT treated films. Figure 5 6 also shows the absorption spectrum of the NCs in solution and in film. It i s interesting to note that there is a shift in the absorption peak of the films to longer wavelengths after they are treated with the dithiol solutions. This shift is expected, since the treatment with EDT or BDT cross links the nanocrystals and reduces i nterparticle distances to increase the effective cluster size. These shifts also suggest
108 that in addition to nanocrystal size, the cut off wavelength of the photodetectors can also be slightly modified through the solvent treatments. 5.3.2 Keeping Track of the Ligand Exchange in the Nanocrystal Films In order to confirm the chemical changes on the surface of the nanocrystals and in the films brought about by the exchange of organic ligands, we investigated films of PbSe nanocrystals capped with different capp ing ligands by FTIR in the reflectance mode. Figure 5 7 compares the FTIR spectrum of a film of PbSe nanocrystals capped with oleic acid with that of another film with nanocrystals capped with octylamine ligands. As expected, there is a decrease in the i ntensity of the C H peaks after the exchange of oleic acid ligands with shorter chain octylamine, due to the lower content of carbon atoms in the latter. There is also a very broad background peak from 2500 3500 cm 1 in the film with oleic acid capped NCs that is gone in the film with octylamine capped NCs. This broad peak is characteristic of the O H stretch peak from oleic acid 35, 122 In addition, the N H peak from the amine group appears in the film containing nanocrystals capped with o ctylamine. Fig ure 5 8 confirms the second stage of ligand exchange. It compares the film of octylamine capped nanocrystals with a nanocrystal film treated with BDT. The data from the latter shows a further reduction of the C H peaks along with the appearance of sharp C =C peaks originating from the benzene ring of the dithiol capp ing ligand. The presence of a C O O stretch peak s at 1520 and 1410 cm 1 in the octylamine capped nanocrystal film shows that there is some residual oleic acid in that film However, these peak s a re completely suppressed in the spectrum from the BDT treated film, confirming the efficient ligand exchange with the dithiol molecule To determine if the dithiol treatment affects only the surface of the NC films or if it penetrates into the film, an NC film was spincoated on a silicon substrate and treated
109 with BDT to perform a depth profile on it. Figure 5 9 shows the depth profile for the BDT treated film. The different elements in the film were identified and, in order to keep track of the presence of the dithiol molecules as a function of sample thickness, sulfur was monitored as the film was being sputtered away. The measurement shows that the content of sulfur deviates very little from being constant all throughout the film, which indicates that the dithiol treatment does penetrate the film and it interacts with the NCs as opposed to just affecting the surface of the NC film. A slight increase in carbon content as a function of film thickness might indicate that some of the longer capping ligand s are trapped within the NC film. 5.3.3 Effect of Dithiol Ligand Exchange on Device Characteristics Replacing the long insulating ligands with interpenetrating and cross linking short ones not only affects the film quality, it also significantly affects the charge transport properties. Figure 5 10 compares the dark steady state J V characteristics of two ITO/PbSe/Al nanocrystal photodetectors with and without the BDT treatment. The plots show the effects of dithiol treatment on the transport characteri stics of the nanocrystal films. First, under forward bias, the current density in the device treated with BDT is two orders of magnitude higher than that in the untreated device. Second, the untreated device shows a symmetrical J V curve. The device trea ted with BDT, on the other hand, exhibits distinct rectification in the J V characteristics. The enhanced device current under forward bias can be attributed to a decrease in inter particle spacing brought about by the short chain BDT ligand, which leads to enhanced carrier mobility and film conductivity 132 The absence of rectification in the untreated devices is probably due to a large density of trap states present at the surfaces of the nanocrystals 127,133 which serve as transport pathways and recombi nation centers. This situation is shown
110 schematically in figure 5 11. The presence of these deep traps can provide pathways for carrier transport through the mid gap states. Because of the presence of these mid gap states the Fermi level is pinned and carriers are injected directly into these gap states under both forward and reverse biases leading to symmetrical J V characteristics without any rectification. Here, the BDT treatment serves two functions: it reduces the spacing between adjacent nanocrys tals and passivates the surface dangling bonds. As a consequence, carrier transport is via electronic states close to the conduction and valence bands of the nanocrystals, resulting in rectification in the J V characteristics. To further understand how t he carrier transport was affected by the BDT treatment in the nanocrystal films, we fabricated both electron and hole only devices using the PbSe nanocrystals. Figure 5 12 shows the J V characteristics for these single carrier devices. Estimated from the space charge limited current, the hole and electron mobilities in the BDT treated devices were aproximated to be 2 10 4 cm 2 /Vs and 6.5 10 4 cm 2 /Vs, respectively. As expected, compared to untreated PbSe films, the mobilities in BDT treated films are signif icantly higher, with a 20 fold increase in hole mobility and an 80 fold increase in electron mobility. The increase in carrier mobility upon treating the nanocrystal films with BDT further demonstrates the effect of the thiol terminated cross linker in dec reasing inter particle spacing and reducing the trap density in the PbSe nanocrystal films. In order to gain further insight into the role of the thiol terminated capping groups in passivating traps in the nanocrystal thin films, we investigated the te mperature dependence of the dark current in the devices. Figure 5 13 shows the dark current densities as a function of temperature in the films with and without BDT treatment,
111 respectively. In the devices treated with BDT, the dark current exhibits a tem perature dependence with an activation energy of ~25 meV, indicating that the transport is a thermally activated process. Similar results were obtained for EDT treated devices as well On the other hand, in the device without the dithiol treatment, the c urrent does not show any temperature dependence. The absence of temperature dependence of the device current suggests that the transport is via tunneling through the gap states. These results indicate, as stated earlier, that the dithiol treatments help reduce the density of gap states. For photodetector applications, it is desirable to achieve the lowest dark current while maintaining the maximum possible photocurrents. As discussed in the previous section, photodiode characteristics were only ob tained when the PbSe films were treated with EDT and BDT. Here, the photoresponse of the resulting NC photodetectors from the two dithiol treatments are compared. Figure 5 14 shows the dark and photocurrents for photodetectors treated with EDT and BDT. The results clearly show that the EDT treatment has much higher dark current in the devices compared to the B DT treat ed device. This difference the NCs closer together compared to BDT. The responsiv ity of the photodetectors was also determined using the equation 1 8 described in chapter 1. Figure 5 15 shows the responsivities for the two devices as a function of bias voltage. The EDT treated device exhibit s a responsivity larger than 0.67 A/W at low volta ges, indicating that there was gain in the photoresponse at the measuring conditions for reverse bias voltages larger than 0.26 V. Gain in some organic photodetectors with high dark currents has been observed and attributed to enhanced charge injection f rom the electrode due to
112 accumulation of photo generated carriers trapped at the electrode interface 134 137 On the other hand, gain was not observed in BDT treated devices at low voltages because of the lower dark currents in comparison with the EDT devi ce. However, at higher voltages, the dark current increases, resulting in gain at voltages beyond 1.7 V. It is important to note that for some applications, gain may not be useful if the dark currents are high; therefore, there is still a need to decrea se the dark currents for the PbSe NC photodetectors. An important performance parameter in photodetectors is the spectral width over which they are active. PbSe nanocrystals, by virtue of quantum size effects, allow tuning of the absorption edge over a broad range of wavelengths in the infrared. Consequently, the cut off wavelength of the photodetectors can also be tuned easily. Also, due to the spectral sensitivity of the nanocrystals all the way into visible wavelengths, the detectors display a broad w indow of activity. Figure 5 16 shows the spectral response of a photodetector device at different reverse bias voltages. The shape of the responsivity closely follows the absorption spectrum (shown in figure 5 5) of the BDT treated nanocrystal films, ind icating that the photoactivity of the photodetector is dependent on the absorption spectrum of the NCs in the infrared. The peak in absorption and responsivity at 1450 nm is attributed to the first excitonic transition in the nanocrystals. These data con firm that the nanocrystals act as IR sensitizers in these devices. 5.4 Summary In summary, the effects of NC surface capping ligand exchange treatments on the NC film properties as well as on the NC photodetector characteristics were demonstrated. More specifically, the fabrication of defect free films was found to be of
113 utmost importance in these solution processed nanocrystal films. This was achieved through an intermediate ligand exchange step in solution prior to the treatment of films in solid stat e. It was also found that the ability of the capping groups to alter the density of trap states in the PbSe nanocrystals is extremely important to obtain rectification in the device characteristics. In addition, moderate dark current densities were achiev ed in IR photodetectors employing short chain BDT capping ligands, maintaining good responsivity in the devices across a broad spectral range; however, lower dark currents need to be achieved for practical applications.
114 Figure 5 1. Schemati c of process to treat NC films with BDT. Figure 5 2. Schematic of PbSe NC photodetector device structure.
115 Figure 5 3. Schematics of capping ligand molecules with a respective TEM image of the PbSe NCs capped with them. a) Oleic acid. b) Octylamine c) BDT. Figure 5 4. AFM images from films of PbSe NCs capped with oleic acid and treated with dithiol solutions. Insets are the corresponding SEM images of the same films, which share the same scale as the AFM images. a) Control film of NCs capped with oleic acid. b) Film after being treated with EDT. c) Film after being treated with BDT.
116 Figure 5 5. AFM images from films of PbSe NCs capped with octylamine and treated with dithiol solutions. Insets are the corresponding SEM images of the sam e films, which share the same scale as the AFM images. a) Control film of NCs capped with octylamine. b) After being treated with EDT. c) After being treated with BDT. Figure 5 6. Absorption spectrum from PbSe NCs capped with octylamine in solution, with EDT, and BDT.
117 Figure 5 7. FTIR spectrum for NC films with oleic acid and after ligand exchange with octylamine. Figure 5 8. FTIR spectra for NC films with octylamine and after BDT treatment.
118 Figure 5 9. AES spectroscopy data from a BDT trea ted film. a) Identification of elements present in the NC film by their characteristic energy. b) Depth profile for the NC film keeping track of the elements present in it. Figure 5 10. J V characteristics for a device with long capping ligands and af ter dithiol treatment.
119 Figure 5 11. Shcematic of unpassivated dangling bonds that act as mid gap traps in NCs. Figure 5 12. J V characteristics for the single carrier devices plotted in log log scale. a) Electron only device. b) Hole only device
120 Figure 5 13. Temperature dependence of the dark current for the photodetector devices treated with dithiol molecules and not treated. a) BDT treated. b) Not treated. Figure 5 14. Dark and photocurrents for the photodetectors treated with EDT and BDT.
121 Figure 5 15. Calculated responsivity for the photodetectors treated with EDT and BDT. Figure 5 16. Spectral responsivity for a BDT treated photodetector at different voltages.
122 CHAPTER 6 REDUCING DARK CURRENT OF INFRARED NANOCRYSTAL PHOTODETEC TORS WITH INORGANIC/ORGANIC BLOCKING LAYERS 6.1 Dark Current in Infrared Nanocrystal Photodetectors Near infrared (NIR) and short wave infrared (SWIR) photodetectors are important for multiple applications such as telecommunications, SWIR imaging, and nigh t vision 71, 77, 123 Currently, epitaxially grown semiconductor materials such as InGaAs are used for commercial SWIR photodetectors. However, the epitaxial deposition techniques employed for device fabrication are expensive, limited to small areas, and th e devices lack physical flexibility. Thus, considerations such as the ease of materials processing, low cost fabrication methods, and large area flexible electronic devices are increasingly becoming important factors for new applications that are not possi ble with conventional inorganic semiconductor materials. Colloidal semiconductor nanocrystal materials, being compatible with low temperature and large area device processing, offer an alternative platform for low cost electronics and optoelectronics 138 139 In this context, colloidal nanocrystals of PbSe have great potential for infrared (IR) photodetection due to their spectral tunability across a wide spectral range in the IR region 36, 73, 138 Depending on the size of the PbSe nanocrystals, their ban dgap energies can be varied from 0.35 eV to 1.3 eV, making them an ideal choice for IR detection. However, due to the processing steps required to improve transport properties and photocurrents in these NC devices i.e., reducing inter particle distance b y controlling the capping ligands as described in chapter 5 the dark currents in PbSe NC photodetectors are high 130,138,140 In addition, the small bandgap energies of infrared sensing NCs impose a fundamental limitation in order to reduce the dark curr ents since electrons can be easily excited to the conduction
123 band under small applied biases. On the other hand, it is desirable to lower the dark there is a great nee d to reduce the dark currents of infrared NC photodetectors. A common approach to decrease the dark currents in devices is to use carrier blocking layers sandwiching the electroactive layer to impede charge injection under reverse bias 72, 141 We used seve ral wide bandgap organic and inorganic materials as carrier blocking layers in IR photodetectors fabricated from PbSe nanocrystals. We chose wide band gap electron transporting materials with a deep highest occupied molecular orbital (HOMO) energy to bloc k holes and wide bandgap hole transporting materials with a shallow lowest unoccupied molecular orbital (LUMO) energy to block electrons 72, 74, 75, 141 In addition to the requirements of proper HOMO and LUMO energy levels, the materials used for blocking layers must also be compatible with the solution process fabrication of the nanocrystal based PbSe photodetectors. Here, we demonstrated substantial reduction in dark current in the IR photodetector when poly[(9,9 dioctylfluorenyl 2,7 diyl) co (4,4 (N (4 sec butyl))diphenylamine)] (TFB) and ( poly(N,N bis(4 butylphenyl) N,N bis(phenyl)benzidine ) (poly TPD) layers were used as the electron blocker layers and a ZnO NC layer was used as a hole blocker. The highest detectivity value of the devices was ~1 10 12 Jones (1 Jones = 1 cm Hz 1/2 /W), which is close to the values in InGaAs photodetectors 142 In addition to reduction in dark current, we found that ZnO nanocrystals can also enhance the operating stability of the resulting devices.
124 6.2 Engineering and Processing of PbSe NC Photodetectors with Blocking Layers In order to suppress dark currents under reverse bias, wide bandgap materials are needed for the blocking layers between the electroactive nanocrystal layer and the electrodes. Figure 6 1 shows the energy band diagram and the general stucture of such a heterostructure photodetector. A typical device has the following structure: ITO/MoO 3 /electron blocker/PbSe NCs/hole blocker/Al. A thin layer of MoO 3 is used to improve the extraction of photogenerat ed holes, following reports on organic based photovoltaic devices 116 117,143 The band diagram highlights the specific energy level requirements that the wide bandgap materials should satisfy to block carriers under reverse bias. For this purpose, it is i mportant to have a large energy offset between the anode work function and the LUMO level/conduction band (CB) of the electron blocking material and a large energy offset between the cathode work function and the HOMO/valence band (VB) energy of the hole b locking material to block carrier injection from the electrodes. This criterion was used to choose a series of materials to be used as blocking layers in the photodetector devices. Figure 6 2 shows the molecular structure of the blocking materials employ ed in this study. Here, we used poly TPD and TFB as the electron blockers while C 60 and ZnO nanocrystals were used as the hole blockers. TFB and poly TPD are polymer materials that were purchased from American Dye Source, C 60 is a small molecule that was also purchased form American Dye Source, the ZnO nanocrystals were synthesized by our collaborator Lei Qian from Dr. Table 6 1 lists the corresponding HOMO/VB and LUMO/CB energy levels of these electron or hole blocking materials. It is important to note that the energy levels fulfill the blocking layer requisites described earlier. Figure 6 2 also depicts the size tunable absorption of the PbSe nanocrystals,
125 also employed in this study as the IR sensi tive material, spanning over a broad range of wavelengths from the visible to 1800 nm in the IR. The inset of Figure 6 2 shows a TEM image of ~4 nm PbSe nanocrystals that were used to fabricate the photodetectors in this work. In addition to the electron ic conditions needed to successfully fabricate the photodetectors with blocking layers, the processing of each individual layer must be taken into consideration. Solution processed heterostructures can be challenging to fabricate since washing of underlyi ng layers is a common issue. In the next section we describe how this issue can be avoided. 6.2.1 Device Fabrication and Processing The ITO patterned glass substrates were cleaned by rubbing them with a cloth and ultrasonicating them in acetone and chlor oform for 15 minutes each. The substrates were exposed to UV ozone treatment for 20 minutes and then immediately introduced into an evaporator chamber to deposit 10 nm of MoO 3 as an extraction interlayer. As stated earlier, the MoO 3 was deposited at a 0. 2 A/s rate. The electron blocking layers were spincoated on top of MoO 3 with a thickness of ~50 nm followed by a thermal annealing at 120 C and 170 C for poly TPD and TFB respectively for 30 minutes each The PbSe NC layer was subsequently spincoated on top of the blocking layer. To avoid washing or dissolving the polymer layers, the PbSe NCs were dispersed in a diluted hexane solution. Since poly TPD and TFB cannot dissolve in hexane, there were no problems with spincoating the subsequent PbSe NC film The substrates were also treated with benzenedithiol (BDT) just as described in the previous chapter and multiple PbSe NC films were stacked on top of each other up to a thickness of ~60 nm. After the
126 active layer was finished, the hole blocking layer was spincoated on top and annealed for 30 minutes at 70 90 C (in the case of ZnO NCs) or deposited through thermal evaporation for the cases of C 60 and BCP. Spincoating the ZnO NCs on top of the PbSe film was not a problem either since the films were mad e insoluble after the crosslinking ligand exchange with the BDT treatment. To spincoat the ZnO, the device was set to a speed of 4000 rpms for 40 seconds with no rampint of initial speeds. The thicknesses of the hole blocking layers were set to ~40 nm wit h the exception of the BCP layer, which was 10 nm thick. Finally, the devices were finished with the evaporation of 100 nm aluminum layer as the cathode. For the control devices, no blocking or extraction layers were inserted so that the final structure was ITO/PbSe/Al. All the spincoating and annealing steps were carried inside an inert nitrogen glovebox environment, except the last annealing step for the ZnO NC layer, which was performed in a hood. 6.2.2 Preparation of the Blocking Layer Solutions an d Materials The two electron blocking layers solutions TFB and poly TPD were prepared using chlorobenzene as the solvent with a concentration of 3 mg/ml and 10 mg/ml, respectively. The solutions were prepared inside a glovebox and were not taken out at any time. For the hole blocking layers, we used ZnO NCs ranging in between 3 5 nm in size, which were synthesized by a sol gel process using precursors of Z inc acetate and t etramethylammonium hydroxide (TMA H) 144 For a typical process, the ZnO nanocrystal s were synthesized by slowly dropwise addition of a stoichiometric amount of TMAH dissolved in ethanol (0.55M) to 0.1M z inc acetate dihydrate dissolved in dimethyl sulfoxide (DMSO) f ollowed by stirring for an hour After being washed the
127 ZnO nanocrystals were dissolved in ethanol and stored outside the glovebox for stability reasons. All solutions were filtered with a 0.45 m filter. C 60 and BCP were also used as hole blocking materials, but they were easily deposited on top of the spincoated films insi de an evaporator chamber. 6.2.3 Device Characterization The J V characteristics of the devices were measured with a Keithley 4200 semiconductor parameter analyzer. The devices were irradiated with monochromatic light from a Newport monochromator using an O riel solar simulator as a source Illumination at 830 nm and 1100 nm were used on the devices with power intensities of ~200 W/cm 2 and ~100 W/cm 2 respectively. The i llumination intensities were measured using two calibrated Newport 918D photodiodes on e for the visible and another for the IR part of the spectrum. The intensity of the incident irradiation was varied by using a set of neutral density filters and a diffuser To obtain the spectral response of the photodetectors, light from the monochromato r was chopped at 400 Hz to modulate the optical signal. The p hotocurrent response as a function of bias voltage was measured using a Stanford Research System SR810 DSP lock in amplifier The spectral response was also used to calculate the spectral detec tivity of the devices across the IR. In order to test the lifetime of the devices, we left the photodetectors with and without ZnO nanocrystal interlayer outside the glovebox being exposed to ambient conditions. 6.3 Results and Discussion 6.3.1 Effect of Blocking Layers on Photodetectors J V Characteristics The blocking layers are very effective in reducing dark currents of the devices and the effect can be clearly observed in figure 6 3. Figure 6 3a shows the dark J V
128 characteristics of a control device, which does not have any charge blocking layers and it has the simple following structure: ITO/PbSe/Al. Figure 6 3b shows the dark J V characteristics for two devices with charge blocking layers. One of the devices has TFB and C 60 as the electron and hol e blocking layers while in the other device has poly TPD and ZnO nanocrystals as the blocking layers. From the comparison between the two plots, it is clear that the use of electron and hole blocking layers significantly reduces the dark current under rev ersed bias. The difference becomes more pronounced at higher reverse biases since there is less leakage in the device with the blocking layers. At 2.0 V applied bias, the dark current is reduced by more than two orders of magnitude compared to that in t he control device. It is important to note that this difference in dark current comes from the fact that in the control device the dark current has a strong bias dependence, while the dark current in the devices with blocking layers is fairly independent of the bias voltage, revealing the effectiveness of blocking layers in suppressing the dark currents and eliminate leakage. It is also worth noting that the device with poly TPD and ZnO nanocrystals shows lower dark currents than the device with TFB and C 60 under reversed bias. This difference arises mainly because the ZnO nanocrystals have a much deeper HOMO compared to C 60 and therefore it is a better hole blocker. 6.3.2 Detectivity Calculations for Photodetectors While responsivity is an important para meter for any photodetector because it determines how efficient the device is in converting incident power to current, it does not fully characterize the performance of a photodetector. To fully determine the performance of a pohotodetector, it is imperat ive to determine the sensitivity of the device by having a notion of the signal to noise ratio. The specific detectivity (D*) is a
129 figure of merit created to characterize the sensitivity of a detector. D* is the inverse of the noise equivalent power (NEP bandwidth and it is expressed as follows 145 : (6 1) where A is the detecting area, f is the bandwidth, and NEP is the noise equivalent power of the photodetector. The NEP is the rad iant power that produces a signal to noise ratio of unity and it is dependent on the noise level of the given photodetector. NEP can be expressed as follows: (6 2) where R is the responsivity, and Ns is the noise signal density co mposed by the shot noise coming from dark currents, the Johnson noise, and Flicker noise. However, due to the small bandgap of PbSe NCs, it can be assumed that the dark current is the main contributor of the noise limiting the sensitivity of photodetector s 72 In this case, the detectivity of the photodetector can be approximated as: (6 3) where q is the electron charge and J d is the dark current. Figure 6 4 shows the dark current, photocurrent, and detectivity values at 0.5 V f or the control device and for devices with the different blocking layers that are listed in table 6 1. The results show the sequence of the different combination of blocking layer materials that were used to fabricate the devices and also demonstrate the effects of
130 devices was approximated with the equation 6 3 and using the measured values of dark, photocurrent, and the intensity of illumination. In the first set of devices in which blocking layers were implemented, we used a combination of TFB and C 60 as electron and hole blocking layers, respectively. TFB is a hole transport material with a shallow LUMO energy and it has been used as an interlayer to facilitate hol e injection in organic light emitting diodes 146 More recently, it has been demonstrated that it is a good electron blocker in polymer solar cells 143 C 60 an electron acceptor used in organic solar cells, has a deep HOMO energy and is expected to block holes. Figure 6 4 shows that the dark current in this photodetector decreased by five fold compared to the control device while the photocurrent increased by over two fold. As stated before, the blocking layers suppress the dark current in the device bec ause of the large energy barriers to hole and electron injection under reverse bias. The photocurrents are also increased because the electron blocking layer prevents the electrons from going to the ITO while the hole blocking layer prevents the holes fro m being lost to the Al under reverse bias 72 These changes in dark current and photocurrent led to an increase in detectivity from 1.510 10 Jones, in the control device, to 8.010 10 Jones for the photodetector with TFB and C 60 as the blocking layers. See king to further reduce the dark current in the device, we then replaced the C 60 layer with a layer of ZnO nanocrystals. As stated earlier, the ZnO nanocrystal material is a wide bandgap n type semiconductor with a deeper lying VB compared to the HOMO energ y of C 60 and is expected to block holes more effectively than C 60 As shown in figure 6 4, the introduction of the ZnO nanocrystals as an electron blocking layer led to an eight fold decrease in dark current and a five fold increase in photocurrent comp ared to the
131 control device. As a result, the detectivity of the devices was improved further to 3.010 11 Jones. In the next device embodiment with blocking layers, the electron blocking layer of TFB was replaced by poly TPD while the ZnO nanocrystals wer e retained as the hole blocking layer. In this device there was a further decrease in dark current, but this was accompanied by a considerable decrease in photocurrent compared to the device with the TFB/ZnO nanocrystal combination of blocking layers. Th is reduction in photocurrent that comes from just switching TFB with poly TPD as electron blocking layers can be attributed to two reasons: first, the HOMO energy level of poly TPD is deeper than the HOMO energy of TFB 143, 147 preventing efficient extract ion of photogenerated holes from the PbSe VB since there is a larger barrier at the PbSe/poly TPD interface compared to the PbSe/TFB interface. The second reason for the lower photocurrents is the difference in hole transporting properties of the two mate rials. The hole mobility of poly TPD is ~ 2.010 3 cm 2 /V s which is an order of magnitude lower than the hole mobility of TFB, ~ 1.0x10 2 cm 2 /V s 148 149 Because of these factors, poly TPD is less efficient in extracting the photogenerated holes, result ing in lower photocurrents. Figure 6 5 compares the detectivity as a function of wavelength of the control photodetector with that of the device with TFB and ZnO nanocrystals as the charge blocking layers because it was the device with the highest value s of detectivity. The device incorporating the blocking layers exhibits detectivity about one order higher than that in the control, over a broad spectral range (600 nm to 1300 nm). The detectivity is greater than 10 11 Jones at wavelengths from 600 nm to 1200 nm, with peak values reaching 10 12 Jones at shorter wavelengths. These detectivity values over a broad
132 spectral range in the SWIR range makes the devices an attractive alternative to conventional photodetectors. Figure 6 6 compares the responsivity as a function of wavelength of these same sets of photodetectors (from figure 6 5). The spectral responses of the two devices are very similar over the entire wavelength range, especially in the IR region. This is due to the fact that the photocurrents i n the two devices, which govern the responsivities, were very similar. In contrast, there is a marked difference in the detectivity of the two devices, shown in the same plot, by more than an order of magnitude at 830 nm. The difference in the two sets o f data serves to illustrate that spectral response measurements can sometimes be misleading in reflecting the performance of photodetectors having significant dark currents. Detectivity, which takes into account contributions from both signal and noise, i s a more practical and useful measure of comparison between devices with different noise levels. To verify the linearity of the photoresponse of the detectors, we measured the photocurrents of the devices as a function of incident light intensity. The li nearity of response is important since it reflects the dynamic range of the device under varying optical signals. The devices were irradiated with light intensities varying over four orders of magnitude. Figure 6 7 shows this photocurrent dependence on op tical power. For the photodetectors, the photogenerated current density does vary linearly with the irradiated light over the entire range of intensities used. This leads to a responsivity of the detector, plotted in figure 6 7 as well, that is practicall y independent of the incident light intensity over this range.
133 6.3.3 Effect of Blocking Layers on Stability of PbSe Nanocrystal Photodetectors It is well known that nanocrystals of PbSe are prone to atmospheric oxidation and degradation, especially when de posited as thin films 58 Hence, improving the stability of the photodetectors fabricated from this material is needed to make them suitable for practical applications. It was found that using ZnO nanocrystals as the hole blocking layer in the photodetect ors significantly increased their storage lifetime under ambient conditions compared to those devices that did not incorporate this blocking layer. Figure 6 8 shows the difference in responsivity as a function of time for the same devices with and without ZnO nanocrystals as the hole blocking layer. Both devices were stored in the ambient and were not encapsulated. The device without the layer of ZnO nanocrystals showed a rapid monotonic decrease in its responsivity over two orders of magnitude within 10 0 hours of storage time. In contrast, the responsivity of the device incorporating ZnO nanocrystals increased to 225 percent of its initial value within the first 24 hours of storage in the ambient. Subsequently, the responsivity reduced at a far slower rate compared to the control device and retained 1.5 times its initial value even after 200 hours. Although the effect of the layer of ZnO nanocrystals is not completely understood, it possibly plays two important roles in enhancing the efficiency in the short term and the stability over several days. Colloidal ZnO nanocrystals have a multitude of surface defects such as oxygen vacancies and/or interstitial Zn atoms 150 When the devices are left in the ambient, molecular oxygen can easily be adsorbed on the ZnO nanocrystal layer, binding with the oxygen vacancies. The dense nanocrystal layer thereby prevents the permeation of atmospheric oxygen and moisture into the inner nanocrystal layer, preventing fast degradation induced by these agents. At the same time, the ZnO nanocrystal layer serves to block UV
134 irradiation, preventing UV induced degradation of the PbSe nanocrystals. Additionally, it has also been demonstrated that ZnO nanocrystals can easily be photodoped by exposure to high energy photon irrad iation 150 151 Photoexcitation of ZnO nanocrystals can enhance the concentration of mobile carriers leading to enhanced conductivity of the ZnO nanocrystal layer. This effect results in better charge collection in the photodetectors, improving the effici ency over the first few hours. However, in order to fully understand the mechanisms in greater detail, more work must be done. 6.4 Summary In summary, several wide bandgap organic and inorganic materials were investigated as carrier blocking layers incor porated into an IR PbSe nanocrystal photodetector architecture. Proper choice of the electron and hole blocking layer materials led to a substantial reduction in dark current in the detectors under operation in reverse bias, leading to high detectivity of the devices over a broad spectral range. Preliminary studies also indicated that a hole blocking layer of ZnO nanocrystals significantly improves the operating lifetime of the devices.
135 Figure 6 1. A schematic of the energy band diagram necessary to block charge injection under reversed biad for a photodetector and the corresponding structure of the device. a) Energy band diagram. b) Schematic of structure. Figure 6 2. Illustration of some materials used in this work. a) Molecular structures of C 60 TFB, and poly TPD. b) Typical absorption spectra of PbSe NCs from different sizes I nset : TEM image of NCs
136 Figure 6 3. Dark J V characteristics of PbSe NC photodetectors. a) W ithout any blocking layers. b) With different blocking layers. Figure 6 4. Dark currents, photocurrents, and calculated detectivity values for photodetectors with different blocking layers. Values for 0.5V.
137 Figure 6 5. Detectivity as a function of wavelength for a photodetector with blocking layers compared with another one without blocking layers. Calculated using responsivity values at 0.5V. Figure 6 6. Responsivity and measured detectivity as a function of wavelength for a photo detector without any blocking layers and with blocking layers. Values m easured at 1.0V and with 830 nm source light.
138 Figure 6 7. Photocurrent density and responsivity of a photodetctor with blocking layers (structure shown in legend) for different light intensities Figure 6 8. Responsivity as a function of time for devices with and without ZnO NC s a s the cathode interlayer.
139 Table 6 1. Materials used in this work with their corresponding HOMO and LUMO energies.
140 CHAPTER 7 APPLICATION OF SOLUT ION PROCESSED INFRARED S ENSING NANOCRYSTALS IN ALL SOLUTION PROCESSED UP CONVERSION DEVICES 7.1 Introduction Light up conversion materials and devices have been studied and explored over the past decades due to their great ability to transform low energy photons on the infrared (IR) range to higher energy emissions ranging close to or in the visible spectrum 7 9 82 Because of this reason, up conversion devices have tremendous potential for different optoelectronic applications such as sensors, homeland security, and especially night vision technologies. In the past, scientists ha ve employed materials such as GaAs, AlGaAs, InP, and other inorganic semiconductors to fabricate up conversion devices 82 85 Unfortunately, due to the equipment required to grow thin films of such materials, the fabrication costs are high and the area of the devices are very limited. Furthermore, inorganic materials are not flexible and therefore are limited in a number of different applications for up conversion devices. Recently, Ban et al., reported a hybrid up converter device that employed an organi c light emitting diode (OLED) as the source of emission of the visible light for the device 86 However, the detecting portion of the device was still grown through MOCVD. Moreover, Chikamatsu et al., and Kim et al., have fabricated all organic up convers ion devices utilizing simple thermal evaporation, taking a step further in trying to use new materials compatible with flexible and large area devices while using fabrication methods that are faster, less complicated, and less expensive 87 89 However, in order to reduce the cost of fabrication even more and to incorporate the ability to process big area devices, solution processed materials must take place in the fabrication of up conversion devices. In this chapter, the fabrication of an all solution pro cessed up conversion device utilizing
141 inorganic nanocrystals and polymers is demonstrated. The inorganic colloidal PbSe nanocrystals were used as the photodetecting material in the IR while the polymer poly[2 methoxy 5 (2' ethyl hexyloxy) 1,4 phenylene vi nylene] (MEH PPV) was used as the light emitting source in the visible. PbSe nanocrystals were chosen because they are well known IR absorbers with great tunability ranging from 0.9 m all the way up to 1.8 m and also because of their potential use in IR photodetector applications 140 Another important limitation for up conversion devices is the operating voltages required to operate them. In most cases, the operating voltages at which the maximum amplification on/off ratio occurred were above 9V, incre asing the overall power required to operate the devices 79 89 We also show in this chapter that, by using the right materials, the operating voltages can be as low as 3V, which makes these up conversion devices much more appealing for practical applicatio ns. To the best of our knowledge, this is the first attempt to make an all solution processed up conversion device. 7.2 Fabrication of Solution Processed Up conversion Devices The basic structure of our solution processed up conversion device is illustrat ed in F igure 7 1. The device architecture is very simple because it only consists of an IR sensing layer and a light emitting layer. However, it might be challenging to stack solution processed single layers back to back. In this case, the advantage of using dithiol treated NC films renders the NC layer insoluble to the common organic solvents used to disperse the NCs. This was shown in the previous chapter, but here the PbSe NC films are turned insoluble and then an MEH PPV polymer solution can be spin coated on top without any problems. To fabricate the devices, the ITO substrates
142 were cleaned and then introduced into a glovebox to spincoat a 40 nm thick PbSe NC film. For this device, we used very diluted solutions of PbSe NC (~20 mg/ml) and spincoate d each film at 1500 rpm to ensure a thin film. After treating the NC film with BDT, the films were rinsed with chloroform and left drying on a hot plate at 50 C for 10 minutes. Next, a 100 nm thick MEH PPV film was spincoated on top of the PbSe NC film a t a speed of 900 rpm for 60 seconds with an initial speed of 500 rpm for 3 seconds. After the film was made, it was annealed at 150 C for 30 minutes and left to cool down to room temperature gradually The MEH PPV solution was prepared using a concentrat ion of 5 mg/ml of chlorobenzene as the solvent and was kept inside the glove box at all times The MEH PPV solution was also filtered with a 0.45 um filter to get rid of all possible agglomerates, however we did not try to use 0.22 um filters because they would break down the polymer chains and result in low efficiency. The device was finished by depositing Li/Al (1nm/100nm) as the cathode contact inside an evaporator chamber. The Li was deposited by evaporating it at a rate of 0.1 A/s and the Al was dep osited at a rate of 1.0 A/s. In order to control hole injection into the device, a ZnO NC layer was added to the device structure in between ITO and the IR sensing PbSe NC layer. In this case, the ZnO NC layers were spincoated at 2000 rpm for 30 seconds instead of 4000 to create a thicker film (~60nm) compared to the previous chapter. Then the ZnO films were annealed at 90 C for 15 mintures outside the glovebox. Spincoating the PbSe NC layer on top of the ZnO layer was facilitated since the ZnO NCs were not soluble in non polar solvents. As a result, the PbSe NC solution (in chloroform) was easily spincoated on top and the rest of the processing was the same as before.
143 7.3 Characterization of Reflective Up conversion Devices Figure 7 2 shows an illustr ation for the measurement setup for the reflective up conversion devices. Since the device only has one transparent electrode (ITO), the IR light source and the visible emission from the device both have to be shined and measured from the ITO side at the same time. For this reason, the up conversion device is called a reflective type. Otherwise, if a transparent cathode is used instead of aluminum, the device can emit from both ends. To characterize the performance of up conversion devices and how effici ent they are in transforming IR light into visible light, each device was tested under dark conditions and then under incident IR irradiation while the light emission from the device was being measured. The l uminance current voltage (LIV) characteristics of the light up conversion devices were measured using a Keithley 2400 source meter for current voltage measurements coupled with a Keithley 6485 picoammeter connected to a calibrated Si photodiode for photocurrent measurements. A Newport LPM830 30C CW di ode laser with a power density of 14.1 mW/cm 2 at 830 nm was used as one IR light source Another source used was an infrared broadband light originating from an Oriel Solar simulator light passing through a 1000 nm filter with a power density of 20mA/cm 2 The devices were not encapsulated and the measurements were carried out at room temperature under an ambient atmosphere. 7.4 Results and Discussion 7.4.1 Device Mechanism and Control of Hole Injection The mechanism by which our up conversion device wor ks is depicted in figure 7 3. Basically, when IR light is irradiated onto the device, the PbSe nanocrystals absorb the photons and generate excitons. Under an applied bias, these excitons dissociate into
144 separate charges: the electrons migrate towards th e ITO anode while the holes go into the light emitting polymer where they recombine with injected electrons from the aluminum cathode to create visible light. It is also important to note that hole injection from the ITO can also contribute to the light e mission. However, it is imperative that hole injection is minimized in order to maximize the up conversion effect due to IR light absorption alone 89 Figure 7 4 shows the comparison between the structures of the up conversion device and a typical MEH PPV light emitting device. The main difference between the two structures is that, for the up conversion device, the PbSe NC layer is added and the PEDOT layer is removed. The PEDOT layer is excluded from the up conversion device structure because it would increase the hole injection, as in the case of a typical MEH PPV device. However, as stated earlier, in the up conversion device the holes are supposed to originate only from the photogenerated excitons. In addition to the exclusion of PEDOT, the ITO sub strates were not treated with UV ozone intentionally so that the hole injection barrier from ITO to the PbSe NCs would be higher. Figure 7 5 shows the LIV characteristics of the up conversion device with such device structure under dark and irradiated con ditions. It is clear that when the device is irradiated with infrared light, the currents and the luminance are higher. This data proves that the photogeneration of charges and the mechanism described earlier is actually taking place in the device. Howe ver, the difference between the dark and irradiated conditions is small and the device does not have a switching transition in which it can go from non light emitting to light emitting. These results suggest that hole injection from the ITO electrode is s till dominant and needs to be suppressed even more in order to see a better switching effect. In order to control the hole injection from ITO
145 and maintain the processability of the device in solution, we used ZnO NC. These nanoparticles are expected to b lock holes because they are n type in nature and have a deep HOMO level. In addition, the ZnO nanoparticles are dispersed in ethanol, making them compatible with our solution fabrication method and also with the PbSe NC, which are dispersed in non polar s olvents. Figure 7 6 shows the structure and the respective energy band diagram of the device using ZnO NCs as a hole blocker. As it can be seen, the ZnO NC layer allows for a 3.3eV energy barrier that could effectively block holes and reduce the dark cur rent in the device. Figure 7 7 shows the reduction in dark current due to the presence of this blocking layer. First, by replacing PEDOT with the PbSe layer and not treating the ITO with UV ozone, the dark currents decrease one order of magnitude. Next, with the addition of the ZnO nanoparticle layer, the dark currents decrease three additional orders of magnitude, proving that the holes are effectively blocked. Once the dark currents in the device are low enough, the majority of the holes transferred in to MEH PPV under irradiated conditions will originate from the photogeneration at the PbSe NC layer. Figure 7 8 shows a clear difference between the dark current and the photocurrents for the up conversion device with ZnO NCs. Also, the characteristic ME H PPV turn on voltage around 2 V can be appreciated clearly only when the device is being irradiated with infrared light, proving that hole injection into the MEH PPV layer is coming from the PbSe nanocrystal layer. Figure 7 9 shows the luminance curve fo r the same device. In this case, when the device is swept under dark conditions, light emission does not happen until six volts are applied to the device; however, under infrared irradiation, light emission can be seen at as low as two volts,
146 indicating t hat there is a four volt range in which the device can be used as a on/off switch up converter. 7.4.2 Efficiency of All Solution Processed Up Conversion Device Figure 7 10 shows the luminance efficiency of the light emitting polymer in the up conversion device. As shown, the efficiency was calculated to be 1.2 cd/A, which is slightly lower, but comparable to typical MEH PPV devices. However, it is important o note that the PbSe NC film will also absorb some of the emission coming from the device so the e fficiency of the emission layer is expected to be higher. This result indicates that the injection of holes from the PbSe nanocrystal layer into the MEH PPV polymer is good and comparable to the injection of holes when using PEDOT. On the other hand, the light power density applied on the device to generate the photocurrents was 14.1mW/cm 2 To calculate the photon to photon efficiency conversion from infrared to visible photons, we can use the following expression: (7 1) where I phot o is the measured photocurrent from the up conversion device, R is the responsivity of the silicon photodetector used to measure the light emission f is the fraction of photons reaching the photodetector, is the wavelength P ir is the incident powe r, IR is the wavelength of the incident infrared light source, h and c is the speed of light. Figure 7 11 shows the calculated photon to photon efficiency for the up converstion device. The reason why the photon to photon efficienc y is low is due to the low efficiency of the MEH PPV along with low responsivity from the photodetecting portion of the device.
147 7.4.3 Dark to Visible Switching Factor The luminance on/off ratio is a parameter that has been reported recently for up convers ion devices and can be calculated by using the following expression: (7 2 ) wh ere L irr and L dark are the luminance under irradiation of infrared light and luminance under dark conditions, respectively. However, the calculation of a parameter can be misleading since the luminance under irradiated light depends on the power density of the light source. Simultaneously, this calculation can only be applied at voltage ranges in which the device is on regardless of whatever it is under dark or irradiation Since no light can be detected below the turn on voltage under dark conditions, thi s expression cannot be used in the range in which the device works as a luminous on/off switch. To solve these two probl ems, a normalized expression using the irradiated power density and the NEP of the measuring photodetector is proposed in order to have an applicable expression in the dark to visible luminance on/off switch ing voltage range. Expression 7 2 can then be modified as follows: (7 3) where I phot o is the photocurrent measured by our photodetector and generaged from the emission of the up conversion device, R is the responsivity of the photodetector, IR is the power density of the irradiated infrared source, is the wavelength, and NEP is the noise equivalent power for the measuring photodetector. This normalized dark to visible switching factor (DVSF) for the up conversion device can be appreciated in f igure 7 12.
148 As it is shown, the switching factor increases the most around 3 5 volts and it proves that the device can be used as a dark to visible switch up converter at low voltages compared to other reports 79 89 It is important to note that below two volts and above six volts, the switching factor cannot apply since the device is always off and it turns on by injection under dark, respectively. Figure 7 13 shows images of a demonstration up conversion device under an applied voltage of three volt s. The device is being irradiated with an 830 nm infrared laser and only the area that the laser hits is emitting orange light. Even though there is an applied voltage to the device, the rest of the device area is not emitting any light at all, demonstra ting the practicality of the device. 7.4 Summary In summary, we have demonstrated the fabrication of an up conversion device that has been completely processed through solutions. With the addition of a ZnO NC hole blocking layer, the device shows low dar k currents and enables it to work as an on/off switch up converter at low voltages. The results show that further improvements can be achieved by improving the photodetection properties of the device and using more efficient light emitting polymers.
149 Figure 7 1. Schematic representation of the basic structure for the solution processed up conversion device. Figure 7 2. Schematic of measurement setup for up conversion device. a) Setup for reflective type. b) Structures for reflective type and tra nsparent type up conversion devices.
150 Figure 7 3. Schematic representation of the mechanism by which the up conversion process takes place. Figure 7 4. Device structures with their corresponding energy band diagrams for a typical MEH PPV light emitti ng device and a simple up conversion device.
151 Figure 7 5. LIV characteristics for the simple up conversion device under dark and infrared irradiation.
152 Figure 7 6. Device structure and corresponding energy band diagram for an up conversion device wit h a ZnO NC hole blocking layer.
153 Figure 7 7. Dark current for a typical MEH PPV light emitting device and up conversion devices with and without hole blocking layer. Figure 7 8. Dark and photocurrents under 830 nm monochrom atic IR light and IR broadband light above 1000 nm for the up conversion device with ZnO NC hole blocking layer.
154 Figure 7 9. Luminance under dark and 830 nm IR light for the up conversion device with ZnO NC hole blocking layer. Figure 7 10. Luminous efficiency plot for the up conversion device.
155 Figure 7 11. Photon to photon efficiency for the up conversion device. Figure 7 12. Plot for the dark to visible switching factor for the up conversion device.
156 Figure 7 13. Image of a demonstration up conversion device under 830 nm IR irradiation and at three volts of applied bias.
157 CHAPTER 8 CONCLUSION AND FUTURE WORK 8.1 Conclusion the most developed countries in the world. As the lifestyle and demands of people expand steadily, technology has matured as well. Specifically the fields of organic electronics and nanotechnology are important contributors to the mission of developing better al ternative energy sources in technology as well as versatile optoelectronic devices. In this work we have mainly focused on the fabrication and processing of infrared sensing inorganic/organic based optoelectronic devices. In addition we have also develope d some methods to synthesize metal nanostructures embedded in organic matrices and to alter the growth of organic nanostructured films using simple thermal co evaporation. In the case for high K dielectric films from inorganic/organic composites, it was s hown that optimized thermal annealing of co evaporated conductive metal in an organic matrix enhances phase segregation of the metal constituents. This segregation resulted in the formation of discrete metal nanocrystals that were successful on increasing the dielectric constant and reducing the dissipation factor of devices. This method can be controlled by annealing temperature and annealing time. We have also developed a processing method that allows us to modify the morphology of organic thin films a crystallize. Different materials can be synthesized using this method and the results can be reproducible on different substrate materials and with good uniformity. This processing met hod can be used to enhance the efficiency of organic solar cells on the
158 basis of increased heterojunction interface. In Chapter 3, this method was used in CuPc based solar cells and it was found that the organic photovoltaic cells that were fabricated wit h the N CuPc films showed an improvement over the control devices. The nanostructure orientation was found to be very important since the penetration of solutions is troubled by randomly oriented structures. The setup system for the synthesis of colloida l nanocrystal through the quick hot injection method was described in detail. In addition, the procedure to synthesize PbSe nanocrystals was also described in detail. Different synthesis conditions resulted in varying NC quality batches during the proces s to optimize the standard synthesis recipe. In general, as the time and temperature of the reaction is higher, the size of the nanocrystals is also higher. On the other hand, careful control of the precursor concentration is crucial to obtain monodisper se nanocrystals. These PbSe nanocrystals were used in the fabrication of defect free films. Exchanging ligands from long to short cross linking chains has a very degrading effect on the morphology of the nanocrystal films. As seen in Chapter 5, an inter mediate ligand exchange step in solution prior to the treatment of films in solid state was found to be a good way to avoid defects in the films. It was also found that the short dithiol molecules EDT and BDT are very good in passivating the dangling bonds at the surface of the nanocrystals and to improve the charge transport in the PbSe nanocrystal films. Wide bandgap organic and inorganic materials were investigated as carrier blocking layers incorporated into an IR PbSe nanocrystal photodetector archite cture. In C hapter 6, proper choice of the electron and hole blocking layer materials led to a substantial reduction in dark current in the detectors under operation in reverse bias,
159 leading to high detectivity of the devices over a broad spectral range. Special solvent selection based on material solubility enabled PbSe in a hexane solution to be processed on top of polymer materials. In addition, preliminary studies also indicated that a hole blocking layer of ZnO nanocrystals significantly improves the operating lifetime of the devices. Finally, we also demonstrated the fabrication of an up conversion device that has been completely processed through solutions. It was found that the selection of solvents based on polarity, and the use of the dithiol tr eatments as crosslinkers are crucial for the processing of the device. Insolubility of the ZnO and PbSe nanocrystals in non polar and polar solvents, respectively, enabled them to be processed both in solution. BDT treatment rendered the PbSe film insolu ble for further solution processing on top. ZnO NC layer was found to be a great hole blocking material and was needed to suppress the dark currents in the up conversion device, which enables it to work as an on/off switch up converter at low voltages. T he results show that further improvements can be achieved by improving the photodetection properties of the device and using more efficient light emitting polymers. 8. 2 Future Work Organic electronics and nanocrystal based devices have evolved quite rapidl y in the past decades. Today, we see that there is much advancement in these fields and many different applications are integrating either organic electronics or inorganic nanocrystals or both. In this dissertation we focused on the processing and characte rization of optoelectronic devices based on inorganic nanocrystals with organic materials; however, there is much more that needs to be addressed for these devices to reach their full potential.
160 One of the main obstacles with these inorganic nanocrystals is their stability and the lifetime of the devices. As we briefly saw in C hapter 6, the devices based on PbSe nanocrystals tend to degrade quite fast. In order for the devices to be operational during thousands of continuous hours, better encapsulation p rocesses must be developed and the degradation mechanisms must be understood better so that it can be prevented. One possible route to increase the stability of these infrared sensing nanocrystals is to passivate them in a core/shell structure. However, more work must be done to achieve sucesfull results. Speciffically, when dealing with PbSe nanocrystals, it is hard to find a semiconductor material that can protect the core from oxygen while having good lattice matching while avoiding diffusion problems in which the shell material creates an alloy with the core Moreover, an ideal shell material for PbSe would also have to fulfill the proper energetic level requirements in order to have proper charge transfer depending on the device applications (charge recombination versus charge extraction). Besides the physical stability of PbSe nanocrystals, the challenge would still remain to prevent the photo bleaching of the materials 56 In addition to the problems of stability, more investigation must be done w ith processing systems so that nanocrystal materials can be manufactured in large area flexible substrates without major defects and in a very reproducible manner QD vision is one company that has had some success in developing manufacturing processes fo r large area displays and lighting applications that are based on nanocrystal materials 152,153 However, more demonstrations must be done for infrared sensing nanocrystal materials like PbSe. Perhaps manufacturing of large devices under ambient condition s is still an option for these not so stable materials.
161 Additional f uture work will also be focusing on the development of up conversion devices, particularly i n all solution processed devices, since the technology has the potential to offer low production and manufacturing cost. Specifically, different device structures can be engineered in order to have more versatility for application integration purposes, for application other than night vision technology, as well as to increase the efficiency of the n anocrystal sensing layes (if NC are being used). However, more efficient materials might have to be developed for both the sensing and for the light emissi on materials in order to achieve very high photon to photon efficiency in these un conversion device s Nanocrystals have the advantage of easy tunability; however, the capping ligands and multiple dangling bonds are still a limiting problem even with special processing treatments. Perhaps, the synthesis of infrared absorbing polymers might be a possibl e route to implement for solution processed up conversion devices as sensing layers. Additionally PHOLEDs are much more efficient than most PLEDS so much work must be focused on improving the efficiency of such emissive materials in order to have solutio n processed high efficien cy up conversion devices that can actually be used in such practical applications as night vision. Other areas of research that needs to be addressed in order to help the development of up conversion devices is on transparent con tacts. Transparent electrodes are needed in order to have transmission of the converted visible light through the up conversion devices. However, it is a challenge to create new materials that have large transmission through the visible spectrum and can perform properly as an electrode. In the past, thin layers of gold or silver have worked patially since their
162 transmission is not very good. Other attempts have included the use of oxide materials along with silver metal, but the outcome have not been en tirely succesfull.
163 L IST OF REFERENCES 1 Gaudiana, R. & Brabec, C. Organic materials: fantastic plastic. Nature Photonics 2 287 289 (2008). 2 Seager, S. L. & Slabaugh, M. R. Chemistry for today: general, organic, and biochemistry 5ed: Brooks Cole (20 04). 3 Encyclopedia Britanica, Organic compound, http://www.britannica.com/EBchecked/topic/431954/organic compound (2010). 4 Silinsh, E. A. Organic molecular crystals: their electron states Springer Verlag Berlin Heidelberg New York, (1980). 5 P ope M. & Swenberg, C. E. Electronic processes in organic crystals and polymers 2ed: Oxford University Press, ( 1999 ) 6 Peumans, P. Organic thin film photodiodes, Ph.D. Thesis, Princeton University, (2004). 7 Giepmans, B. N. G., Adams, S. R., Ellisman, M H., Tsien, R. Y. The fluorescent toolbox for assessing protein location and function Science 312 217 224 (2006). 8 Yang, W., Butle r, J. E., Russel, J. N., Hamers, R. J. Interfacial electrical p roperties of DNA modified d iamond thin f ilms: Intrinsic re sponse and h ybridization Induced field e ffects Langmuir 20 6778 6787 (2004). 9 Brtting, W. Physics of Organic Semiconductors Wiley VCH, (2005). 10 Ruiz, R. et. al. Pentacene thin film growth. Chem. Mater. 16 4497 4508 ( 2004 ) 11 Peis ert, H. et. al. Order on disorder: Copper phthalocyanine thin films on technical substrates. J. Appl. Phys 90 466 469 ( 2001 ) 12 C lle, M. & Br tting, W. Thermal, structural and photophysical properties of the organic semiconductor Alq 3 Phys. Stat. Sol. (a) 201 1095 1115 (2004) 13 Xue, J., Rand, B. P., Uchida, S., Forrest, S. R. A hybrid planar mixed molecular heterojunction pho tovoltaic cell. Adv. Mater. 17 66 71 (2005). 14 Yao, Y., Hou, J., Xu, Z., Li, G., Yang, Y. Effects of solvent mixtures on the nanoscale phase separation in polymer solar cells. Adv. Func. Mater. 18 1783 1789 (2008). 15 Li, G., Shrotriya, V., Yao, Y., Y ang, Y. Investigation of annealing effects and film thickness dependence on polymer solar cells based on poly(3 hexylthiophene). J. Appl. Phys. 89 043704 (2005).
164 16 Kim D. H., Jang, Y., Park, Y. D., Cho, K. Surface induced conformational changes in pol y(3 hexylthiophene) monolayer films. Langmuir 21 3203 3206 (2005). 17 Campoy Quiles, M. et. al. Morphology evolution via self organization and lateral and vertical diffusion in polymer: fullerene solar cell blends. Nature Mater. 7 158 164 (2008). 18 Er itt, M., May, C., Leo, K., Toerker, M., Radehaus, C. OLED manufacturing for large area lightning applications. Thin Solid Films 518 3042 3045 (2010). 19 Krebs, F. C. et. al. A complete process for production of flexible large area polymer solar cells ent irely using screen printing First public demonstration. Solar Energy Materials & Solar Cells 93 422 441 (2009). 20 Helander, M. G. et. al. Oxidizing gold thin films: an effective material for high performance f lexible organic optoelectronics. Adv. Mate r. 21 1 4 (2009). 21 Xue, J. Pure and Mixed Organic Thin Films and Heterojunctions for Optoelectronic Applications, Ph. D. Thesis, Princeton University (2005). 22 Konarka, Capabilities, http://www.konarka.com/media/pdf/Konarka_Capabilities_Broch_8_09.pdf, (2010). 23 Lamansky, S. et. al. Cyclometalated Ir complexes in polymer organic light emitting devices. J. Appl. Phys. 92 1570 1575 (2002). 24 Lamansky, S. et. al. Highly Phosphorescent Bis Cyclometalated Iridium Complexes: Synthesis, Photophysical Characterization, and Use in Organic Light Emitting Diodes. J. Am. Chem. Soc. 123 4304 4312 (2001). 25 Brooks, J. et. al. Synthesis and Characterization of Phos phorescent Cyclometalated Platinum Complexes. Inorg. Chem. 41 3055 3066 (2002). 26 Kondakov, D. Y., Lenhart, W. C., Nichols, W. F. Operational degradation of organic light emitting diodes: Mechanisms and identification of chemical products. J. Appl. Phys 101 024512 (2007). 27 Jorgensen, M., Norrman, K., Krebs, F. C. Stability/degradation of polymer solar cells Solar Energy Materials & Solar Cells 92 686 714 (2008). 28 Aziz, H., Popovic, Z. D., Hu, N. X., Hor, A. M., Xu, G. Degradation Mechanism of S mall Molecule Based Organic Light Emitting Devices. Science 283 1900 1902 (1999). 29 Wagner, H. J., Loutfy, R. O., Tsiao, C. K. Purification and characterization of phthalocyanines. J. Mater. Sci. 17 2781 2791 (1982).
165 30 Ho, C. D., Yeh, H. M., Yeh, T. L. The optimal variation of zone lengths in multipass zone refining processes. Separation and Purification Technology 15 69 78 (1999). 31 Warta, W., Stehle, R., Karl, N. Ultrapure, high mobility organic photoconductors Appl. Phys. A 36 163 170 ( 1985 ) 32 Rogach, A. L. Semiconductor Nanocrystal Quantum Dots: Snthesis, Assembly, Spectroscopy and Applications, Springer Wien New York, (2008). 33 Dragoman, M. & Dragoman, D. Nanoelectronics: Principles and Devices, 2ed Artech House (2009). 34 Alivisatos P. The use of nanocrystals in biological detection. Nature Biotechnology 22 47 52 (2004). 35 Tang, J. et. al. Quantum dot photovoltaics in the extreme quantum confinement regime: The surface chemical origins of exceptional air and light stability. ACS Nano 4 869 878 (2010). 36 Konstantatos, G. et. al. Ultrasensitive solution cast quantum dot photodetectors. Nature 442 180 183 (2006). 37 Murray, C. B., Norris, D. J., Bawendi, M. G. Synthesis and characterization of nearly monodisperse CdE (E= S, Se, Te) semiconductor nanocrystallites. J. Am. Chem. Soc. 115 8706 8715 (1993). 38 Toneguzzo, P. et. al. Static and dynamic magnetic properties of fine CoNi and FeCoNi particles synthesized by the polyol processc. Magn. IEEE Trans. 35 3469 34 71 ( 1999 ) 39 Gerberich, W. W. et. al. Super hard silicon nanospheres. J. Mech. Phys. Solids 51 979 992 ( 2003 ) 40 properties, and perspectives. Chem. Mater. 13 3843 3858 (2001). 41 Kovalen ko, M. V., Scheele, M., Talapin, D. V. Colloidal nanocrystals with molecular metal calchogenide surface ligands. Science 324 1417 1420 (2009). 42 Kruis, F. E., Fissan, H., Peled, A. Synthesis of nanocparticles in the gas phase for electronic, opt ical and magnetic applications --a review. J. Aerosol. Sci. 29 511 535 (1998). 43 Quinlan, F. T. et. al. Reverse miscelle synthesis and characte rization of ZnSe nanoparticles. Langmuir 16 4049 4051 (2000).
166 44 Destree, C., Ghijsen, J., Nagy, J. B. Pr eparation of organic nanoparticles using microemulsions: their potential use in transdermal delivery. Langmuir 23 1965 1973 (2007). 45 Martin, C. R. Membrane based synthesis of nanomaterials. Chem. Mater. 8 1739 1746 (1996). 46 Ying. Y. & Alivisatos, P Colloidal nanocrystal synthesis and the organic inorganic interface. Nature 437 664 670 (2005). 4 7 LaMer, V. K., Dinegar, R. H. Theory, production and mechanism of formation of monodispersed hydrosols. J. Am. Chem. Soc. 72 4847 4854 (1950). 48 Sugimo to, T. Preparation of monodispersed colloidal particles. Adv. Colloid Interf. Sci. 28 65 108 (1987). 49 Peng, X., Wickham, J., Alivisatos, A. P. Kinetics of II VI and III V colloidal semiconductor nanocrystal growth: Focusing of size distributions. J. Am Chem. Soc. 120 5343 5344 (1998). 50 Alivisatos, A. P. Perspectives on the physical chemistry of semiconductor nanocrystals. J. Phys, Chem. 100 13226 13239 (1996). 51 Olkhovets, A., Hsu, R. C., Lipovskii, A., Wise, F. W. Size Dependent Temperature Var iation of the Energy Gap in Lead Salt Quantum Dots. Phys. Rev. Lett. 81 3539 3542 (1998). 52 R ossetti, R., Ell ison, J. L., Gibson, J. M., Brus, L. E. Size effefct in the excited electronic states of small colloidal CdS crystallites. J. Chem. Phys. 80 44 64 4469 (1984). 53 Kayanuma, Y. Quantum size effects of interacting electron and holes in semiconductor microcrystals with spherical shape. Phys. Rev. B 38 9797 9805 (1988). 54 Anikeeva, P. O., Halpert, J. E., Bawendi, M. G., Bulovic, V. Electrolumines cence from a mixed red green blue colloidal quantum dot monolayer. Nano Lett. 7 2196 2200 (2007). 55 Gao, X. et. al. Current Opinion in Biotechnology 16 63 72 (2005). 56 Stouwdam, J. W. et. al. Photostability of Colloidal PbSe and PbSe/PbS Core/Shell Nanocrystals in Solution and in the Solid State J. Phys. Chem. C 111 1086 1092 ( 2007 )
167 57 Hines, M. A. & Guy ot Sionnest, P. Synthesis and Characterization of Strongly Luminescin g ZnS Capped CdSe Nanocrystals. J. Phys. Chem. 100 468 471 ( 1996 ) 58 Dai, Q. et. al. Stability study of PbSe semiconductor nanocrystals over concentration, size, atmosphere, and light Exposure. Langmuir 25 12320 12324 ( 2009 ) 59 J. D. Wright, Molecular Crystals 2nd ed : Cambridge University Press (1995). 60 Romero, H. E., Drndic, M. Coulomb blockade and hopping conduction in PbSe quantum dots. Phys. Rev. Lett. 95 156801 4 (2005). 61 C. Kittel, Introduction to Solid State Physics 7 ed. New York: John Wiley and Sons, Inc. 62 Yu, G., Gao J., Hummelen, J. C., Wudl, F., Heeger, A. J. Polymer Photovoltaic Cells: Enhanced Efficiencies via a Network of Internal Donor Acceptor Heterojunctions. Science 270, 1789 1891 ( 1995 ) 63 Tang, C. W. Two layer organic photovoltaic cel l. Appl. Phys. Lett. 48 183 185 (1986). 64 Solamer, OPV Record, http://www.solarmer.com/ 65 P ark, S. H. et. al. Bulk heterojunctions solar cells with internal quantum efficiency approaching 100%. Nature Photonics 3 297 303 (2009). 66 Y ang, F., Shtein, M., Forrest, S. R. Controlled growth of a molecular bulk heterojunction solar cell. Nature Mater. 4 37 41 (2005). 67 Yang, F., Lunt, R. R., Forrest, S. R. Simultaneous heterojunction organic solar cells with broad spectral sensitivity. A ppl. Phys. Lett. 92 053310 (2008). 68 Halls, J. J. M. et. al. Efficient photodiodes from interpenetrating polymer networks. Nature 376 498 500 (1995). 69 Peumans, P., Bulovic, V., Forrest, S. R. Efficient, high bandwidth organic multilayer photodetect ors. Appl. Phys. Lett. 76 3855 (2000). 70 Yu, G., Cao, Y., Wang, J., lvain, J., Heeger, A. J. High sensitivity polymer photosensors for image sensing applications. Synth. Met. 102 904 907 (1999). 71 Kim, D. Y., Sarasqueta, G., So, F. SnPc:C60 bulk het erojunction organic photovoltaic cell with MoO3 interlayer. Solar Energy Materials & Solar Cells 93 1452 1456 (2009).
168 72 Gong, X. et. al. High detectivity polymer photodetectors with spectral response from 300 nm to 1450 nm. Science 325 1665 1667 (2009 ). 73 Rauch, T. et. al. Near infrared imaging with quantum dot sensitized organic photodiodes. Nature Photonics 3 332 336 (2009). 74 Choi, J. J. et. al. PbSe nanocrystal excitonic solar cells. Nano Lett. 9 3749 3755 (2009). 75 Leschkies, K. S., Beatty T. J., Ka ng, M. S., Norris, D. J., Aydil, E. S. Solar cells based on juntions between colloidal PbSe nanocrystals and thin ZnO films. ACSNano 3 3638 3648 (2009). 76 Choudhury, K. R., Sahoo, Y., Ohulchanskyy, T. Y., Prasad, P. N., Efficient photoconduc tive devices at infrared wavelengths using quantum dot polymer composites. Appl. Phys. Lett. 87 073110 ( 2005 ) 77 McDonald, S. A. et. al. Solution processed PbS quantum dot infrared photodetectos and photovoltaics. Nat. Mater 4 138 142 ( 2005 ) 78 Ko nstantatos, G., Clifford, J., Levina, L., Sargent, E. H. Sensitive solution processed visible wavelength photodetectors. Nat. Photonics 1 531 534 ( 2007 ) 79 Krus e, P. W., Pribble, F. C., Schulze, R. G. Solid state infrared wavelength converter employing high quantum efficiency Ge GaAs heterojunction. J. Appl. Phys. 38 1718 1720 (1967). 80 Mita, Y. Detection of 1.5 excitable phosphors. Appl. Phys. Lett. 39 587 589 (1981). 81 Wang Y. & Ohwaki, J. High efficiency infrared to visible upconversion of Er +3 in BaCl 2 J. Appl. Phys. 74 1272 1278 (1993). 82 Russell, K. J. e t. al. Room temperature electro optic up conversion via internal photoemission. Appl. Phys. Lett. 82 2960 2962 (2003). 83 Sandhu, J. S., Heberle, A. P., Alphenaar, B. W., Cleaver, J. R. A. Near infrared to visible up conversion in a forward bias Schottky diode with a p Appl.Phys. Lett 76 1507 1509 (2000). 84 Liu, H. C., G ao, M., Poole, P. J. 1.5 conversion device. Electron. Lett. 36 1300 1301 (2000). 85 Yang, Y. et. al. Near infrared photon upconversion devices based on GaNAsSb active layer lattice matched to GaAs. Appl. Phys. Lett. 94 093504 (2009).
169 86 Ban, D. et. al. Near infrared to visible light optical upconversion by direct tandem integration of organic light emitting diode and inorganic photodetector. Appl. Phys. Lett. 90 093108 (2007) 87 Ni, J. et. al. Organic light emitting diode with TiOPc layer --A new multifunctional opt oelectronic device. Jpn. J. Appl. Phys. 40 L948 L951 (2001). 88 Chikamatsu, M. et. al. Light up conversion from near infrared to blue using a photoresponsive organic light emitting device. Appl. Phys. Lett. 81 769 771 (2002). 89 Kim, D. Y., Song, D. W., Chopra, N., De Somer, P., So, F. Organic infrared upconversion device. Adv. Mater. 22 2260 2263 (2010). 90 Scholes, G. D. & Rumbles, G. Excitons in nanoscale systems. Nature Mater. 5 683 696 (2006). 91 G reen, M. A., Emery, K., Hishikawa, Y., Warta, W. Solar cell efficiency tables (version 36). Prog. Photovolt: Res. Appl. 18 346 352 (2010). 92 Cheyney, T., OPV Record, 2009, http://www.pv tech.org/news/_a/solarmer_breaks_organic_solar_pv_cell_conversion_efficiency_r ecord_hits_nre/ ( 2010 ) 93 Arbatti, M., Shan, X., Cheng, Z. Ceramic polymer composites with high dielectric constant. Adv. Mater. 19 1369 1372 (2007). 94 Anikeeva, P.O., Madigan, C. F., Halpert, J. E., Bawendi, M. G., Bulovic, V. Electronic and excitonic processes in light emit ting devices based on organic materials and colloidal quantum dots. Phys. Rev. B 78 085434 (2008). 95 Peng, H., Ho, Y. L., Yu, X. J., Wong, M., Kwok, H. S. Coupling efficiency enhancement in organic light emitting devices using microlens array Theory a nd experiment. IEEE J. Disp. Technol. 1 278 282 (2005). 96 Zhang, Q. M. et. al. An all organic composite actuator material with a high dielectric constant. Nature 419 284 287 (2002). 97 Xu, J. & Wong, C. P. Low loss percolative dielectric composite. Ap pl. Phys. Lett. 87 082907 (2005). 98 Daniel, M. C. & Astruc, D. Gold Nanoparticles: assembly, supramolecular chemistry, quantum size re lated properties, and applications toward biology, catalysis, and nanotechnology. Chem. Rev. 104 293 346 (2004). 99 Nicolais, L. & Carotenuto, G. Metal Polymer Nanocomposites. Hoboken, N.J. (2005).
170 100 Beenakker, C. W. J. Theory of coulomb blockade osci llations in the conductance of a quantum dot. Phys. Rev. B 44 1646 1656 (1991). 101 S hen, Y., Lin, Y., Li, M., Nan, C. W. High dielectric performance of polymer composite films induced by a percolating interparticle barrier layer. Adv. Mater. 19 1418 14 22 (2007). 102 Ofuji, M., Ishikawa, K., Takezoe, H., Inaba, K., Omote, K. Crystallite size effect on the hole mobility of uniaxially align copper phthalocyanine thin film field effect transistors. Appl. Phys. Lett. 86 062114 (2005). 103 Hsiao, Y. S., Wh ang, W. T., Suen, S. C., Shiu, J. Y., Chen, C. P. Morphological control of CuPc and its application in solar cells. Nanotechnology 19 416503 (2008). 104 Peumans, P., Uchida, S., Forest, S. R. Efficient bulk heterojunction photovoltaic cells using small m olecular weight organic thin films. Nature 425 158 162 (2003). 105 Yamada, K. et. al. Charge transport of copper phthalocyanine single crystal field effect transistors stable above 100 C. Appl. Phys. Lett 88 122110 (2006). 106 Chen, Y., Shih, I. High mobility organic thin film transistors based on monocrystalline rubrene films grown by low pressure hot wall deposition. Appl. Phys. Lett. 94 083304 (2009). 107 Tong, W. Y., Djurisic, A. B., Ng, A. M. C., Chan, W. K. Synthesis and properties of copper p hthalocyanine nanowires. Thin Solid Films 515 5270 5274 (2007). 108 M benkum, B. N., Barrena, E., Zhang, X., Kelsch, M., Dosch, H. Selective growth of organic 1 D structures on Au nanoparticle arrays. Nano Lett. 6 2852 2855 (2006). 109 Xu, H. B., Chen, H. Z., Xu, W. J., Wang, M. Fabrication of organic copper phthalocyanine nanowire arrays via a simple AAO template based electrophoretic deposition. Chem. Phys. Lett. 412 294 298 (2005). 110 Ambily, S. & Menon, C. S. The effect of growth parameters on the electrical, optical and structural properties of copper phthalocyanin thin films. Thin Solid Films 347 284 288 (1999). 111 Xiao, K., Liu, Y., Yu, G., Zhu, D. Influence of the substrate temperature during deposition on film characteristics of copper phth alocyanine and field effect transistor properties. Appl. Phys. A 77 367 370 (2003). 112 Olson, D. C. et. al. Effect of polymer processing on the performance of Poly(3 hexylthiophene)/ZnO Nanorod photovoltaic devices. J. Phys. Chem. C 111 16640 16654 (2 007).
171 113 Hatta, A., Suzuki, S., Suetaka, W. Polarization modulation electronic absorption study of copper phthalocyanine films on silver by surface plasmon resonance spectroscopy. Appl. Surf. Sci. 40 9 18 (1989). 114 Wojdyla M., Derkows ka, B., Rebarz M., Bratkowski, A., Bala, W. Stationary and modulated absorption spectroscopy of copper phthalocyanine (CuPc) layers grown on transparent substrate. J. Opt. A: Pure Appl. Opt. 7 463 466 (2005). 115 Shrotriya, V., Li, G., Yao, Y., Chu, C. W., Yang, Y. Transition metal oxides as the buffer layer for polymer photovoltaic cells. Appl. Phys. Lett. 88 073508 (2006). 116 Kim, D. Y. et. al. The effect of molybdenum oxide interlayer on organic photovoltaic cells. Appl. Phys, Lett. 95 093304 (2009). 117 Irf an, et. al. Energy level evolution of molybdenum trioxide interlayer between indium tin oxide and organic semiconductor. Appl. Phys. Lett. 96 073304, (2010). 118 Brennan, J. G. et. al. The preparation of large semiconductor clusters via the pyrolysis of a molecular precursor. J. Am. Chem. Soc. 111 4141 4143 (1989). 119 Steigerwald, M. L. & Brus, L. E. Synthesis, stabilization, and electronic structure of quantum semiconductor nanoclusters. Annu. Rev. Mater. Sci. 19 471 495 (1989). 120 Peng, X. et. al. Sha pe control of CdSe nanocrystals. Nature 404 59 61 (2000). 121 Williams, D. B. & Carter, C. B. Transmission Electron Microscopy. Springer (1996). 1 22 Steckel, J. S., Yen, B. K. H., Oertel, D. C., Bawendi, M. G. On the mechanism of lead chalcogenide n anocrystal formation. J. Am. Chem. Soc. 128 13032 13033 (2006). 123 Yao, Y. et. al. Plastic near infrared photodetectors utilizing low band gap polymer. Adv. Mater. 19 3979 3983 (2007). 124 Huang, I. W. et. al. Ultrafast electron transfer and decay dyn amics in a small band gap bulk heterojunction material. Adv. Mater. 19 2307 2312 (2007). 125 Sargent, E. H. Solar cells, photodetectors, and optical sources from infrared colloidal quantum dots. Adv. Mater. 20 3958 3964 (2008). 126 Boberl, M., Kovalenk o, M. V., Gamerith, S., List, E. J. W., Heiss, W. Inkjet printed, nanocrystal photodetectors operating up to 3 m wavelengths. Adv. Mater. 19 3574 3578 (2007). 127 Jarosz, M. V., Porter, V. J., Fisher, B. R., Kastner, M. A., Bawendi, M. G. Photoconductivi ty studies of treated CdSe quantum dot films exhibiting increased exciton ionization efficiency. Phys. Rev. B 70 195327 (2004).
172 128 Yu, D., Wang, C., Wehrenberg, B. L., Guyot Sionnest, P. Variable range hopping conduction in semiconductor nanocrystal sol ids. Phys. Rev. Lett. 92 216802 1 (2004). 129 Barkhouse, D. A. R., Pattantyus Abraham, A. G., Levina, L., Sargent, E. H. Thiols passivate recombination centers, in colloidal quantum dots leading to enhanced photovoltaic device efficiency. ACSNano 2 2356 2362 (2008). 130 Koleilat, G. I. et. a l. Efficient, stable infrared photovoltaics based on solution cast colloidal quantum dots. ACSNano 2 833 740 (2008). 131 Williams, K. J. et. al. Strong electronic coupling in two dimensional assemblies of colloidal PbSe quantum dots. ACSNano 3 1532 1538 (2009). 132 Sargent, E. H. Infrared photovoltaics made by solution processing. Nat. Phot. 3 325 331 (2009). 133 Ginger, D. S., Greenham, N. C. Charge injection and transport in films of CdSe nanocrystals. J. Appl Phys. 87 1361 1368 (2000). 134 Matsukura, Y., Uch iyama, Y., Yamashita, H., Nishino, H., Fujii, T. Responsivity dark current relationship of quantum dot infrared photodetectors (QDIPs). Infrared physics & technology 52 257 259 (2009). 135 Gao, J., Heg mann, F. A. Bulk photoconductive gain in pentacene thin films. Appl. Phys. Lett. 93 223306 (2008). 136 Hiramoto, M., Miki, A., Yoshida, M., Yokoyama, M. Photocurrent multiplication in organic single crystals. Appl. Phys. Lett. 81 ,1500 1502 (2002). 137 H iramoto, M., Imahigashi, T., Yokoyama, M. Photocurrent multiplication in organic pigment films. Appl. Phys. Lett. 64 187 189 (1994). 138 Klem, E. J. D., MacNeil, D. D., Cyr, P. W., Levina, L., Sargent, E. H. Efficient solution processed infrared photovol taic cells: Planarized all inorganic bulk heterojunction devices via inter quantum dot bridging during growth from solution. Appl. Phys. Lett. 90 183113 (2007). 139 Chen, H. Y., Lo, M. K. F., Yang, G., Monbouquette, H. G., Y ang, Y. Nanopartricle assiste d high photoconductive gain in composites of polymer and fullerene. Nat. Nanotechnology 3 543 547 (2008). 140 Sarasqueta, G., Choudhury, K. R., So, F. Effect of solvent treatment on solution processed colloidal PbSe nanocrystal infrared photodetectors. Chem. Mater. 22 3496 3501 (2010).
173 141 Keivan idis, P. E., Khong, S. H., Ho, P. K. H., Greenham, N. C., Friend, R. H. All solution based device engineering of multilayer polymeric photodiodes: Minimizing dark current. Appl. Phys. Lett. 94 173303 (2009). 1 42 Lim, H. C., Tsao S., T aguchi, M., Quivy, A. A., Razeghi, M. InGaAs/InGaP quantum dot infrared photodetector with high detectivity. Proc. SPIE 6127 (2006). 143 Su bbiah, J., Kim, D. Y., Hartel, M., So, F. MoO 3 /poly(9,9 dioctylfluorene co N [4 ( 3 methy lpropyl )] diphenylamine) double interlayer effect on polymer solar cells Appl. Phys. Lett. 96 063303 (2010). 144 Bera, D., Q ian, L., Sabui, S., Santra, S., Holloway, P. H. Photoluminescence of ZnO quantum dots produced by a sol gel process. Opt. Mater. 30 1233 1239 ( 2008 ). 145 Jones, C. Proposal of the detectivity D** for detectors limited by radiation noise. J. Opt. Soc. Am. 50 1058 1059 (1960). 146 Choulis, S. A., Choong, V. E., Mathai, M. K., So, F. The effect of interfacial layer on the performan ce of organic light emitting diodes. Appl. Phys. Lett. 87 113503 (2005). 147 Bae, W. K. et. al Highly efficient green light emitting diodes based on CdSe@ZnS quantum dots with a chemical composition gradient. Adv. Mater. 21 1690 1694 (2009). 148 Zhu, W. W., Xiao S., Shih, I. Field effect mobilities of polyhedral oligomeric silselquioxanes anchored semiconductor polymers. Appl. Surf. Sci 221 358 363 (2004). 149 Fong, H. H., Papadimitratos, A., Hw ang, J., Kahn, A., Malliaras, G. G. Hole injection in a model fluorine triarylamine copolymer Adv. Funct. Mater 19 ,304 310 (2009). 150 Verbakel, F., Meskers, S. C. J., Janssen, R. A. J. Electronic memory effects in diodes from zinc oxide nanoparticle polystarene hybrid material. Appl. Phys. Lett. 89 102103 ( 2006 ) 151 Gilot, J., Wienk, M. M., Janssen, R. A. J. Double and triple junction polymer solar cells processed from solution. Appl. Phys. Lett. 90 ,143512 (2007). 152 Coe Sullivan, S., Steckel, J. S., Woo, W. K., Bawendi, M. G., Bulovic, V. Large area ordered quantum dot monolayers via phase separation during spin casting. Adv. Funct. Mater. 15 1117 1124 (2005). 153 QD Vision, Applications of Quantum Dots, (2010).
174 154 Hipps, K. W. e t. al. Metal d orbital ocupatio n dependent images of the scanning tunneling microscopy of metal pthalocyanines. J. Phys. Chem. 100, 11207 11210 (1996). http://en.wikipedia.org/wiki/Anthracene (2010). 156. Trikha, R. Pi bond, http://www.goiit.com/posts/list/organic chemistry can you please explain why the sp sp2 sp3 905708.htm 157. Top alternative energy sources, Silicon PV, http://www.top alternative energy sources.com/silicon solar cells.html Nov. 1, (2010). 158. Jacoby, M. The Power of Plastic http://pubs.acs.org/cen/coverstory/88/8834cover.html Nov. 1, (2010). 159. Talapin, D. Research Interests, http://chemistry.uchicago.edu/faculty/faculty/person/member/dmitri talapin.html Nov. 1, (2010). 160. TEM components, http://commons.wikimed ia.org/wiki/File:Scheme_TEM_en.png Nov. 1, (2010). 161, Williams, D. B. & Carter, C. B. Transmission Electron Microscopy: A Textbook for Materials Science 2 nd Ed, 2009. 162. TEM grid, http://www.vcbio.science.ru.nl/en/image gallery/preptem/ Nov. 1, (2010). 163. Nucleation, http://en.wikipedia.org/wiki/Nucleation Nov.1 (2010).
175 B IOGRAPHICAL SKETCH Galileo Sarasqueta was born in the Repu b lic of Panama. As he grew up, he always had an affinity towards the sciences and figuring out problems. He graduated high school in the year 2000 with the intention of becoming an engineer and be ing involved at the edge of technology. In 2003, he transferred to the University of Florida from Florida State University to join the Materials Science and Engineering program. As a n MSE student, Galileo specialized on ceramic materials sciences. During the last year of his undergraduate studies, he was exposed to research under the guidance of Dr. Sigmund Wolfgang processing titanium oxide nanowires from sol gel solutions using an electro spinning technique. After being involved in a research environment, he decided to join the Ph. D. program in the same department at the University of Florida and focus in electronic materials. During the first years of the program, Galileo was given the opportunity to work at the Major Analytical Instrumentation Center (MAIC) as an SEM organic electronics. At MAIC galileo learned about vacuum systems, sputtering tools, and different characterization tools. He was also exposed to a wide variety of research th at was being done all through the different engineering and medical departments of the university. During the rest of the graduate program, Galileo had the opportunity to work on different organic electronic devices such as solar cells, photodetectors, un conversion devices, capacitors, and LEDs. He was also involved in the synthesis and application of semiconductor and metallic nanocrystals to devices. He learned to use different processing tools to fabricate devices as well as to characterize their opt ical, electrical, physical, and chemical properties.
176 When he completes his Ph.D. program, Galileo hopes to go into industry and apply his knowledge and systhematic approach to problems in real fabrication processes of electronic devices.