University of Southern California Dissertations and Theses

Singlet fission in disordered acene films: from photophysics to organic photovoltaics

(USC Thesis Other)

Singlet fission in disordered acene films: from photophysics to organic photovoltaics

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content SINGLET FISSION IN DISORDERED ACENE FILMS: FROM PHOTOPHYSICS TO ORGANIC PHOTOVOLTAICS by Robert Eric McAnally A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (CHEMISTRY) August 2016 Copyright 2016 Robert Eric McAnally ii Table of Contents List of Tables v List of Figures vi Chapter 1. Introduction and Overview of Organic Photovoltaics and their Governing Photophysical Processes 1 1.1 The Case for Solar 1 1.2 Current Solar Options 4 1.2.1 Inorganic Photovoltaics 4 1.2.2 Organic Photovoltaics 6 1.3 Mechanisms of Photoelectrical Conversion 8 1.3.1 Power Conversion Mechanism of the p-n junction 8 1.3.2 Power Conversion Mechanism of Organic Photovoltaics 10 1.3.3 Generalized Description 11 1.3.4 Light Absorption 12 1.3.5 Exciton Diffusion 13 1.3.6 Charge Transfer and Separation 16 1.3.7 Charge Migration 18 1.3.8 Charge Collection 19 1.4 JV Response and Efficiency 20 1.5 The Shockley Quessier Limit 23 1.6 Alternate Routes to Efficiency 25 1.7 References 28 Chapter 2. Molecular Aspects of Singlet Fission: Photophysics and Devices 34 2.1 Overview of Singlet Fission 34 2.1.1 Introduction 34 2.1.2 Thermodynamic Requirements for Singlet Fission 35 2.1.3 Singlet Fission Chromophores 38 2.1.4 Mechanism of Singlet Fission 40 2.1.5 Magnetic Field Effects 43 2.2 Singlet Fission in OPVs 48 2.2.1 Designing a Singlet Fission OPV 48 2.2.2 Singlet Fission OPVs 49 2.2.3 Tetracene Singlet Fission OPVs 50 2.2.4 Pentacene Singlet Fission OPVs 53 2.3 References 57 iii Chapter 3. Photophysical Investigation of Singlet Fission in a Disordered Acene Film 63 3.1. Singlet Fission in Thin Films of 5-12 Diphenyltetracene 63 3.2 Morphology of DPT and Tetracene Vapor Deposited Thin Films 65 3.3 Solution and Condensed Phase Stead-State Absorption and Emission 66 3.4 Time-resolved Emission and Ultrafast Transient Absorption 67 3.5 Kinetic Model of Singlet Fission in DPT Thin Films 70 3.6 DPT Crystal Structure 78 3.7 References 80 Chapter 4. Photophysics of Platinum Porphyrin-Doped DPT Films 84 4.1 Host-Guest Sensitization of Triplet Excitons 84 4.2 Pt(TPBP):DPT Doped Films 85 4.3 Ultrafast Transient Absorption Spectroscopy of Pt(TPBP) Doped DPT Films 87 4.4 OPVs with Pt(TPBP):DPT Doped Donor Layers 91 4.5 References 95 Chapter 5. Construction of a Custom Physical Property Measurement System Solar Cell Testing Station 97 5.1 Testing Station Construction Motivation 97 5.2 The Physical Property Measurement System 98 5.3 PPMS Solar Cell Testing Station Design and Construction 100 5.3.1 OPV Substrate Description and Parameters 101 5.3.2 Substrate Holder and Probe Construction 102 5.3.3 Light Delivery 105 5.3.4 Fully Assembled System Generalized Description of Operation 107 5.4 References 108 Chapter 6. Measurement of Organic Photovoltaic Charge Transfer State Energies and Singlet Fission Response Using a Custom Physical Property Measurement System Solar Cell Testing Station 109 6.1 Variable Temperature Device Testing 109 6.2 Extrapolating Charge Transfer State Energy 115 6.2.1 OPV Device Architecture and Testing Procedure 115 6.2.2 CuPc/C 60 116 6.2.3 Rubrene/C 60 118 6.2.4 Tetracene/C 60 and Pentacene/C 60 121 6.2.5 Summary 123 6.3 Probing Singlet Fission Dynamics in Acene OPVs Using the Physical Property Measurement System Solar Cell Testing Station 125 iv 6.3.1 Singlet Fission in Tetracene/C 60 OPVs 126 6.3.2 Singlet Fission in Rubrene/C 60 OPVs 128 6.3.3 Singlet Fission in DPT/C 60 OPVs 129 6.3.4 Singlet Fission in Pentacene/C 60 OPVs 131 6.3.5 Summary 132 6.4 References 134 v List of Tables Table 3.1: Best fit parameters for the kinetic model constructed to fit ultrafast DPT transient absorption data. 77 Table 4.1: Summary of J-V metrics collected for organic photovoltaics employing both variable thickness 5% Pt(TPBP):DPT donor layers and a neat DPT donor layer. 92 Table 6.1: Tabulated values of experimentally measured charge transfer state energies compared to literature reports and the magnitude of the cell voltage under 1 sun illumination conditions 123 vi List of Figures Figure 1.1: Graphical representation of the total estimated planetary energy stores, both finite and renewable. 3 Figure 1.2: Compilation of photographs highlighting various qualities of organic photovoltaic devices, including roll-to-roll printing, substrate flexability, integration into consumer products and variable chromaticity. 6 Figure 1.3: Chart of all solar cell technologies certified by the National Renewables Laboratory (NREL) over the past 40 years. 8 Figure 1.4: Cartoon representing the sequential steps of photo-to-electrical conversion operant in an organic photovoltaic. 11 Figure 1.5: Cartoon depiction of the two main device architectures constructed for organic photovoltaics: lamellar and bulk heterojunctions. 14 Figure 1.6: Schematic representation of the Förster and Dexter energy transfer mechanisms. 15 Figure 1.7: Current vs. voltage curve plotting the electrical response of an idealized organic photovoltaic diode with graphical overlays accentuating the short- circuit current (J sc , green dot), open-circuit voltage (V oc , green dot), and fill factor (FF, yellow square). 22 Figure 1.8: Graphic representation of the thermalization process that dissipates energy as heat in a semiconductor after absorption of a photon in excess of the bandgap, E g . 24 Figure 1.9: Schematic (left) of the singlet fission process minimizing thermalization losses and the additional efficiency gain to a photovoltaic employing advantageous singlet fission (right). 27 Figure 2.1: Schematic illustrating the singlet fission process. 36 Figure 2.2: Cartoon representation of the four orbital model describing the direct or mediated mechanisms by which singlet fission can proceed. 41 Figure 2.3: Relative measure of tetracene fluorescence intensity in the presence of a variable magnitude magnetic field. 46 Figure 2.4: Compilation of a singlet-fission-sensitized tetracne/CuPc/C 60 organic photovolaic’s device structure (top), EQE (bottom left), and IV curves (bottom right). 50 vii Figure 2.5: TPTPA/rubrene/PDI-CN 2 device EQE and IV data with diagramed device architecture. 51 Figure 2.6: Device architecture (left) and EQE measurement (right) for a P3HT/pentacene/C 60 solar cell. 54 Figure 3.1: XRD traces and TEM images collected on DPT and tetracene vapor deposited thin films. 65 Figure 3.2: Chloroform solution and vapor deposited thin film steady-state absorption and emission spectra collected for DPT and tetracene samples. 66 Figure 3.3: TCSPC decay curves for vapor deposited DPT thin films at short and long delays 68 Figure 3.4: Ultrafast transient absorption curves collected on DPT films and the residual triplet spectrum of the DPT T 1 state. 69 Figure 3.5: DPT differential extinction spectra and fits to TA data at various pump-probe delays. 71 Figure 3.6: Relative amplitudes of DPT S 1 and T 1 populations extracted from TA data revealing ~70% efficient singlet fission at low pump fluences. The extracted S 1 data follows the TCSPC decay confirming the validity of the fitting procedure. 72 Figure 3.7: Diagram of the kinetic model used to model singlet fission kinetic in DPT thin films. 74 Figure 3.8: DPT Singlet and triplet populations at various pump-probe delays s a function of fluence overlayed with the kinetic model fits to the data. 76 Figure 3.9: Comparison of vapor grown and xylene grown DPT crystals. 79 Figure 3.10 Structure of DPT vapor-grown crystals as viewed down the crystallographic a-axis highlighting cofacial molecular arrangement (left) and the eclipsed versus staggered dimer molecular pairs (right) 80 Figure 4.1: Molecular structures of 5-12 diphenyltetracene, DPT, and platinum tetraphenylbenzoporphyrin, Pt(TPBP) 85 Figure 4.2: Absorption spectrum of neat DPT, 5% doped Pt(TPBP):DPT, and 20% doped Pt(TPBP):TFS thin films. 86 Figure 4.3: Ultrafast transient absorption spectra of a 20% Pt(TPBP):TFS at various pump-probe delays following porphyrin excitation. 87 viii Figure 4.4: Ultrafast transient absorption spectra of a 5% Pt(TPBP):DPT thin film at various pump-probe delays following porphyrin excitation, and the residual DPT triplet absorption spectrum. 88 Figure 4.5: Ultrafast transient absorption spectra of a 5% Pt(TPBP):DPT thin film at various pump-probe delays following DPT excitation. 90 Figure 4.6: J-V and EQE curves for OPVs employing Pt(TPBP):DPT doped donor layers. 92 Figure 4.7: Schematic of the various photophysical processes and their respective time constants operating in Pt(TPBP):DPT doped films following either porphyrin or acene excitation. 94 Figure 5.1: Compilation picture of the Physical Property Measurement System (PPMS) and the instrument’s bore and electrical bridge used for loading and testing samples. 98 Figure 5.2: Photographs of the patterned ITO substrates used in the construction of OPVs. 101 Figure 5.3: Photographs of the custom built substrate holder used to house OPV substrates in the PPMS solar cell testing station assembly. 103 Figure 5.4: Photographs of the fiber optic cable light delivery system providing photons to the OPV device at the bottom of the PPMS bore. 104 Figure 5.5: Pictures of the top of the custom PPMS solar cell testing station assembly highlighting the flange vacuum seal interface and fiber optic cable feed through. 106 Figure 5.6: Photograph of the fully assembled PPMS solar cell testing station including the white-light lamp housing and PPMS console control. 107 Figure 6.1: Semi-log J-V curves and molecular models of CuPc versus Pt(TPBP) devices and tetracene versus rubrene devices. 111 Figure 6.2: Semi-log J-V curve and EQE response highlighting sub-band gap photoresponse in tetracene/C 60 and rubrene/C 60 devices. 113 Figure 6.3: Room temperature I-V curves for CuPc/C 60 devices under 1-sun illumination and inside the PPMS solar cell testing station being excited by white light provided through the fiber optic cable assembly. 116 ix Figure 6.4: Variable temperature I-V curves for CuPc/C 60 devices inside the PPMS solar cell testing station under white light excitation provided through the fiber optic cable assembly. Fits to the extracted open-circuit voltages versus temperature extrapolate the interfacial charge transfer state energy. 117 Figure 6.5: Room temperature I-V curves for rubrene/C 60 devices under 1-sun illumination and inside the PPMS solar cell testing station being excited by white light provided through the fiber optic cable assembly. 118 Figure 6.6: Variable temperature I-V curves for rubrene/C 60 devices inside the PPMS solar cell testing station under white light excitation provided through the fiber optic cable assembly. Fits to the extracted open-circuit voltages versus temperature extrapolate the interfacial charge transfer state energy. 120 Figure 6.7: Room temperature I-V curves for a tetracene/C 60 OPV device under 1-sun illumination and inside the PPMS solar cell testing station being excited by white light provided through the fiber optic cable assembly. Fits to the extracted open-circuit voltages as a function of temperature from variable temperature I-V curves extrapolate the interfacial charge transfer state energy. 121 Figure 6.8: Room temperature I-V curves for pentacene/C 60 devices under 1-sun illumination and inside the PPMS solar cell testing station being excited by white light provided through the fiber optic cable assembly. 122 Figure 6.9: Variable temperature I-V curves for pentacene/C 60 devices inside the PPMS solar cell testing station under white light excitation provided through the fiber optic cable assembly. Fits to the extracted open-circuit voltages versus temperature extrapolate the interfacial charge transfer state energy. 122 Figure 6.10 Graphs a tetracene/C 60 OPV current response as a function of variable magnetic field strength inside the PPMS solar cell testing station. The device is illuminated by white light provided to the substrate through the fiber optic cable assembly after being passed through a λ = 500 nm longpass filter. 126 Figure 6.11 Graphs a rubrene/C 60 OPV current response as a function of variable magnetic field strength inside the PPMS solar cell testing station. The device is illuminated by white light provided to the substrate through the fiber optic cable assembly after being passed through a λ = 500 nm longpass filter. 128 Figure 6.12 Graphs a DPT/C 60 OPV current response as a function of variable magnetic field strength inside the PPMS solar cell testing station. The device is x illuminated by white light provided to the substrate through the fiber optic cable assembly after being passed through a λ = 500 nm longpass filter. 129 Figure 6.13 Graphs a pentacene/C 60 OPV current response as a function of variable magnetic field strength inside the PPMS solar cell testing station. The device is illuminated by white light provided to the substrate through the fiber optic cable assembly after being passed through a λ = 600 nm longpass filter. 131 1 Chapter 1. Introduction and Overview of Organic Photovoltaics and their Governing Photophysical Processes 1.1 The Case for Solar: In March 2011 a 15-foot tsunami caused three reactors at the Fukushima Daiichi nuclear plant in Japan to melt and leak radioactive material. The accident caused the evacuations of over a hundred thousand citizens in the immediate area and aroused fears of a potentially permanent radioactive-contaminated area of Japan and the surrounding ocean hundreds of square miles in area. 1 The event was so catastrophic that it was rated a ‘7 (major accident)’ on the International Nuclear and Radiological Event Scale, with the 1986 Chernobyl nuclear disaster in Ukraine as the only other such event in history to be similarly rated. Amidst the disaster public anxiety of nuclear power grew. The apparently cataclysmic consequences of any sort of accident, or terrorist attack, on any nuclear reactor was far too egregious in many opinions to justify further nuclear plant construction and lead to international protests of programs aimed at such projects, including in Germany, which saw the parliament immediately decommission the eight oldest German nuclear reactors within days of the Fukushima disaster and expedite the closing of the remainder. 2 This policy decision has left Germany with higher retail energy costs than those of its comparable European counterparts. 3 Fukushima, along with Japan’s 50+ other nuclear reactors, were responsible for approximately 30% of Japan’s total electrical power production in 2011. 4 As of 2016 only three Japanese nuclear reactors remain active and Japan has had to accommodate for this loss of energy production by increasing the burning of fossil fuels, a problem exacerbated by the fact that Japan lacks an abundance of natural gas and crude oil and must import these fuel sources. All of these factors in concert have seen the cost of electricity in Japan increase significantly. 2 For centuries man’s approach to fill the demand for energy has been the same as that of the Japanese response of the 21 st century: utilize the high energy density located in the molecular bonds of ‘fossil fuels’ (crude oil, coal, natural gas, etc.) through combustion, turning exothermic chemical reactions into heat for inefficient useable electrical power. The myriad problems inherent to such an approach are glaring. They include, to be brief, the ravishing of local ecosystems where such fossil fuels are mined or extracted, the cost of transportation, logistical concerns and international affairs dilemmas arising from disparities between the locus of fuel caches and the destination of their global demand, and the fact that such fossil fuels are by definition limited in their supply. Finally, the combustion of these materials lead to the production of so-called ‘greenhouse gases,’ such as carbon dioxide, which has seen a massive increase in global atmospheric concentrations to over 400 ppm in the last two centuries due to human industrialization. 5 The increase concentration of greenhouse gases in the atmosphere has been identified as the source for the gradual warming of the Earth’s average temperature and linked to drastic shifts in climate and weather, the destabilization of ecosystems, acidification of the ocean, ice cap melting, and sea levels rising. These changes not only lead to the destruction and destabilization of Earth’s ecosystems but also threaten the safety and potential posterity of human civilizations themselves, so much so that the United States Pentagon even identifies climate change as an immediate threat to national security. 6 The continual combustion of fossil fuels and the resulting slow poisoning of the planet is an untenable source for energy. With developing nations such as China, India, and parts of Africa becoming larger energy consumers as their economies and populations continue to grow, and the world population as a whole on a steady incline, it is apparent that sources of energy in the future need to be both renewable, and as environmentally benign, or “green,” as possible. The technology must be scalable and must 3 be implementable in the locality of the demand. If recent events in Japan have swayed public opinion enough to deny nuclear power as a viable option then this sustainable, green, scalable energy technology must come from some other source. Thankfully there exists an infinite (from a human perspective) source of energy a mere 93 million miles away from Earth, the Sun. In one hour the sun delivers more energy to the Earth’s surface than humanity consumes in the course of a year from all sources combined. Direct photon-to-electrical conversion of the sun’s energy is clean with no chemical pollutant by- products and infinitely renewable. It is estimated that global energy demand will double by 2100, reaching a global consumption of approximately 40 TW. 7 Of all renewable energy sources the only source great enough to supply power on this scale of demand is solar (Figure 1.1). 8 Figure 1.1. Graphical representation of the estimated total planetary energy stores, both finite and renewable. Numbers quoted are in Terawatt-years. Of all potential supplies, solar energy is the only available source large enough to meet humanity’s ever-growing energy needs. 8 4 While other renewable sources can assuredly play prominent roles in a sustainable, global energy landscape, electricity from solar is the only sustainable pool large enough in volume to meet the energy requirements of the 21 st century and beyond. For this reason research into new solar technologies as well as methodologies to improve current solar options is of great consequence. 1.2 Current Solar Options 1.2.1 Inorganic Photovoltaics Modern solar solutions convert sunlight to electricity in two different ways, by either directly turning sunlight into electrical power through the photovoltaic effect or through a solar thermal approach wherein the thermal energy of sunlight is used to produce steam which is then turned into electricity through a turbine. 9,10 In the photovoltaic approach the photovoltaic effect (the production of a voltage difference between electrodes in a material as a result of exposure to electromagnetic radiation) is exploited in a device to produce charge carriers at a given electrochemical potential from photons of a sufficient energy. A device that accomplishes this using electromagnetic radiation from the sun is known as a “solar cell.” The number of charge carriers produced, their electrochemical potential, and the photon energy necessary to produce them are all material and device-dependent parameters. Inorganic photovoltaics is a class of device that uses inorganic materials such as silicon (both crystalline (Si-c) and multicrystalline(Si-m)), III-V compounds of InP, GaAs, and chalcogenide CuInGaSe 2 11, 12 (CIGS), 13 and CdTe 14 as the main light absorbing and charge- producing material of the solar cell. Characteristic of these materials are strong covalently- bound lattice sites in a periodic crystal giving rise to electron wavefunction delocalization in valence (electron-filled) and conduction (electron-unfilled) bands. Charge carriers created through the photo-generation process are highly mobile due to the band electronic structure and 5 relatively high dielectric constants of these materials, allowing for efficient screening and charge separation at room temperature, kT. Inorganic devices are capable of high power efficiencies (Si-c 26%, Si-m 22%, GaAs 29%, InP 22%, CdTe 21%, CIGS 21%) 15 but at the tradeoff of financial cost to the consumer. Silicon is by far the most abundant inorganic semiconductor used in commercial photovoltaics and Si-wafer based photovoltaics accounted for over 90% of total production in 2015 alone. 16 While ubiquitous it is by no means a preferred choice for a solar-absorbing material. Silicon’s indirect band gap causes it to have a low absorption-coefficient in the wavelength ranges most important for solar energy conversion (λ = 300-1400 nm). As a result, silicon solar cells must be manufactured hundreds of microns thick in order to ensure enough light absorption to generate appreciable power. In order for diffusive charge carriers to not be trapped or quenched at unwanted impurity sights, the silicon and its crystal lattice must be as regular and devoid of imperfections as possible, demanding high-temperature crystal growth methods resulting in increased manufactured cost. 17, 18 The terrestrial abundance of naturally occurring silicon in conjunction with a developed silicon microprocessors industry have helped to establish silicon solar cells as a market-available technology, however the cost of the cell manufacture makes global implementation to meet the energy demands of the 21 st century economically unviable. 7,19, 20 In order for solar to be cost-competitive with fossil fuels the cost of materials and cost of manufacture must be dramatically decreased. 6 1.2.2 Organic Photovoltaics A burgeoning new technology that has emerged as a potential replacement of, or at least supplement to, current commercial inorganic photovoltaic options is organic photovoltaics (OPVs). 21 These devices use carbon-based organic dye and pigment molecules with high extinction coefficients for visible photons as the main light absorbing medium of the solar cell. These species can be molecular, polymeric, or oligomeric in nature. Compared to silicon photovoltaics, OPVs can absorb a large amount of incident photons with very thin active layers of absorbing material. The organic layers can be processed by low-temperature vapor or solution-based techniques, allowing for low-cost roll-to-roll processing 22 on flexible substrates 23, 24 and consumer products 25 with varied form factors, light weights, and potentially lower cost (Figure 1.2). Furthermore, the full diversity of synthetic organic chemistry can be used to create chromophores specifically tailored to address solar cell optimization and function, or pure Figure 1.2. Photographs of representative OPV devices highlighting roll-to-roll processing (upper left), 22 flexible form factors (upper right), 24 integration into consumer products (lower right), 25 and the unique tunability of design aesthetic and color (lower left). 7 aesthetics. Being able to sample from the vast molecular toolbox of organic semiconductors allows for a unique design space in OPVs that couples both judicious materials choice for cell optimization and the appearance of a consumer product. While OPVs offer a variety of advantages when compared to their inorganic counterparts, overall power conversion efficiencies must be improved. Recently, efficiencies over 10% have been reported for laboratory-scale OPVs 26, 27 and efficiencies continue to inch closer to values projected necessary for OPVs to be considered commercially feasible. However, while improving efficiencies are encouraging, the long-term stability of these devices tends to be poor due to the sensitivity of the organic chromophores to oxidation and degradation from exposure to a combination of moisture, oxygen, and high-energy photons. 28, 29 OPVs’ performance further decline when laboratory devices are scaled-up to large modules. 23,30 Finally, complicated, multi- step synthetic procedures necessary to create the organic chromophores utilized in the highest performing OPVs inflates the overall cost of the technology. Improving device longevity, module efficiency, and decreasing synthetic demand are all mandatory steps to improve OPVs’ prospects. The demands placed on OPV improvement seem daunting but the flourish of organic photovoltaic research in the past twenty years has yielded drastic increases in device performance over a relatively short period of time (Figure 1.3). 27 OPV prospects are further aided from evidence of other organic-based electronics technologies that have already found market viability in the form of organic light-emitting diodes (OLEDs), a technology ubiquitous in modern cell phone, computer, and television displays. 31 OLED’s successful market penetration could act as a harbinger for OPV acceptance if the aforementioned device improvements are realized. 8 1.3 Mechanisms of Photoelectrical Conversion If efficiencies are to be improved it is pertinent to understand the underlying physics governing the charge generation process in an OPV. The photocurrent generation process will be briefly outlined here for a pn-junction and compared to the operating mechanism employed in OPVs with differences highlighted. 1.3.1 Power Conversion Mechanism of the pn-junction One of the most common solar cell device architectures is a pn-junction, characterized by an inorganic material such as silicon doped with impurities to cause increased conductivity and an excess of charge carriers on opposite sides of the junction. When a group V impurity such as phosphorous is introduced into a tetrahedral site of the group IV silicon lattice, excess negative charge is accumulated due to the extra electron of the phosphorous atom. A concentrated region Figure 1.3. Compilation chart of all solar cell technologies’ power conversion efficiencies certified at the National Renewable Laboratory since 1975. While OPV device efficiencies (represented by solid red circles) languish behind their inorganic counterparts, active research in the field has only been conducted over the past two decades and OPVs remain a nascent technology. However, in that short time OPVs have been able to achieve double-digit efficiencies for laboratory scale devices. 27 9 of negative charge in a semiconductor is known as an ‘n-type’ material. Conversely, when a Group III impurity such as boron is introduced into the silicon lattice excess positive charge (also known as a ‘hole’) is accumulated due to the absence of valence electrons from the boron atom. A concentrated region of positive charge (holes) in a semiconductor is known as a ‘p-type’ material. When separate the n-type and p-type materials have the same conduction and valence band energies but differing Fermi energies due to the disparate doping. Combining an n- and p-type material together aligns the materials’ Fermi energies at equilibrium and creates a junction between the two materials. In the center of the junction excess holes from the p-type region spatially diffuse into the n-type region and excess electrons from the n-type region spatially diffuse into the p-type region. These diffusive charge carriers then recombine with their oppositely charged counterparts. However, in the lattice sites from which each carrier migrated is left an oppositely charged ion as demanded by charge neutrality. These covalently bound charged lattice sites are immobile and result in the creation of an electric field within the region of static charge build-up in the middle of the junction and acts to oppose further charge diffusion and sweeps out any free carriers. Known as the ‘depletion region’ this segment of the device can be hundreds of microns thick in an inorganic photovoltaic. Electrically, the pn-junction behaves as a diode in the dark, allowing uni-directional current flow due to the electrical field within the device. When illuminated, an inorganic photovoltaic produces useable power from absorbed photons of energy greater than the band gap (E g ) of the material. First, the absorbed photon promotes an electron from the valence band to conduction band, creating an electron-hole pair. This oppositely charged pair is Coulombically bound but easily separated at ambient temperatures from screening provided by the high dielectric constant of inorganic crystalline materials. The charges are then separated by the 10 electric field of the junction and collected at their respective contacts, providing useable power in the external circuit. 1.3.2 Power Conversion Mechanism of Organic Photovoltaics The electronic structure of organic semiconductors is highly localized, with excited state wavefunctions sometimes confined to single molecules. 32 Given the low dielectric constants of condensed organic media, electron-hole pairs created through photon absorption in organic semiconductors are Coulombically bound, with very large exciton binding energies on the order of 0.5 -1.5 eV. Unlike their inorganic counterparts, the thermal bath at ambient temperature is insufficient to liberate organic excited states from this exciton binding energy to generate free charges. Therefore, in order to generate free charges from the excited states of molecular species, a heterojunction using favorable energy alignment is employed. In this scheme a photoexcited molecule is paired with a second molecular species with a suitable reduction potential. Since a molecule in its electronic excited state is a potent reductant, photoinduced electron transfer can occur between the excited molecule and the ground state electron-accepting species. In this mode charges are separated at a heterojunction between an electron donating species (termed a Donor, D) and an electron accepting species (termed an Acceptor, A). 33 This D/A heterojunction has remained as the defining feature for OPV devices since its advent in 1986 by Tang. 34 The photon-to-charge conversion process in prototypical organic D/A systems can be visualized as a series of distinct photo or electrochemical steps: I. Light Absorption II. Exciton Diffusion III. Charge Transfer and Separation 11 IV. Charge Migration V. Charge Collection A generalized description for the power generation process in an OPV will be given here, with a more detailed analysis of each step to follow. 1.3.3 Generalized Description Figure 1.4. Cartoon representation of the power conversion process in an OPV. First, photon capture creates a molecular exciton with the active layers of the device. The exciton is a tightly bound electron-hole pair that is Coulombically stabilized at room temperature. Then, the exciton migrates through an energy transfer mechanism throughout the active layer until it arrives at the D/A interface. Next, an excited state electron transfer occurs from the exciton to the ground state species, creating an interfacial charge transfer state of energy E CT at the Donor/Acceptor heterojunction. Further charge dissociation separates the electron and hole from one another at the interface, resulting in a fully oxidize Donor (D + ) and fully reduced Acceptor (A - ). Sequential electron exchanges transfer the polaron species through their respective layers where they are collected at electrodes and harnessed in an external circuit as electrical work. 35 Photon absorption within the active layers of an OPV leads to a locally excited state on either the Donor (D*) or Acceptor (A*). Exciton diffusion channels the excited state through an energy transfer mechanism to the D/A interface where photoinduced charge transfer occurs creating a Coulombically-bound interfacial charge transfer (CT) state between geminate positively and negatively charged polarons residing on the Donor and Acceptor [D + /A - ], respectively. Once formed, further charge separation separates the [D + /A - ] pair into individual polaron species resulting in a fully oxidized donor (hole, D + ) and a fully reduced acceptor (electron, A - ). Sequential charge transfer at localized sites transports charge through percolation 12 pathways in the Donor and Acceptor layers until carriers are collected at electrodes and harnessed in an external circuit (Figure 1.4). 1.3.4 Light Absorption The absorptive transitions of π-conjugated small molecules and polymeric systems are typically very intense due to strong ground and excited state wavefunction overlap and can be quite broad from geometric relaxations in the excited state, affording increase photon capture across the visible spectrum. Upon absorption, this geometric relaxation quickly thermalizes excess energy as vibrational heat, until the relaxed molecular equilibrium geometry of the lowest vibronic state (E 00 ) is reached. In solar cells, this thermalization of excess energy that is not harnessed as work in an electrical circuit is a large loss mechanism. The energy difference of this fully relaxed E 00 excited state to the ground state is taken as the energy of the exciton (either D* or A* following the notion previously introduced). In practice this excited state can be measured as the energy of intersection of thin film absorption and emission spectra. For most of the organic semiconductors commonplace in OPVs the overall spin multiplicity of the closed-shell electronic ground state is one and termed a singlet (S 0 ). Since photon absorption occurs with no change in spin, the prepared electronic excited state is also of singlet character, denoted S 1, and typically lives for a characteristic lifetime (τ) of nanoseconds. Triplet manifolds (spin multiplicity three) exist at lower energies than those of the singlets with the lowest-energy triplet (T 1 ) residing a few tenths of an eV lower in energy than S 1 for prototypical organic molecules (there are classes of molecules with very large S 1 -T 1 gaps on the order of well over 1.0 eV, as will be seen in subsequent chapters of this document. However, for most organic chromophores employed in OPVs the S 1 -T 1 gap is much smaller). In molecules such as porphyrins and phthalocyanines with heavy atoms and dominate spin-orbit couplings, 13 rapid intersystem crossing can populate T 1 from the initially prepared S 1 . 36, 37 For pure hydrocarbon molecular systems with vanishingly small spin-orbit coupling the rate of intersystem crossing tends to be orders of magnitude slower than other excited state processes and triplets minimally form. Weak coupling with the singlet ground state further slows the deactivation pathways of triplet excited states, allowing lifetimes (τ) on micro- to millisecond timescales once formed. 1.3.5 Exciton Diffusion In order for an exciton to be split into a separate electron and hole the created excited state must diffuse to the D/A interface from its initial absorption location in an OPV. The distance an exciton can migrate in its lifetime is known as the exciton diffusion length, L D . Appreciable photon capture must be realized within L D of the D/A heterojunction to ensure significant current production and high power conversion efficiencies; photons absorbed outside L D of the D/A interface will not contribute to power production and will decay within the characteristic lifetime (τ) of the excited state. Given the high absorption coefficients of π- conjugated organic semiconductors for visible photon capture substantial light absorption can be realized with several 10-100s nm of active material, thus decreasing the distance an exciton must diffuse before being harnessed. 14 Figure 1.5. OPVs devices are constructed with a D/A heterojunction that is used to separate excitons into free charges. The Donor and Acceptor materials are typically deposited as either discrete, neat layers and the heterojunction formed at the interface (left), or blended together in a bulk heterojunction with nanoscale morphology (right). 38 Device architectures can also be engineered to ensure excitons are created within L D of the D/A heterojunction (Figure 1.5). In a lamellar, or planar, type of device architecture, discrete, distinct layers of neat Donor and Acceptor are deposited. The D/A heterojunction, therefore, is formed only at the interface between these two neat layers of Donor and Acceptor. Such a device architecture demands individual Donor and Acceptor layer thicknesses to be on the order of L D for their respective excited state species. While cartoons of lamellar device structures typically imply a clean D/A interface with no mixing of molecular species, in reality molecular dynamics simulation 39 and neutron diffraction 40 experiments have shown that Donor and Acceptors can mix over nanometer length scales, increasing the volume of D/A interaction and presenting more charge transfer opportunities for excitons. Alternate device architectures include the bulk heterojunction. In this device construction Donor and Acceptor layers are intimately mixed to form an interpenetrating, phase-separated network with nanoscale morphology. This intimately mixed D:A phase is distributed throughout the bulk of the active layer and drastically reduces the distance excitons must travel before they dissociate. As a result, bulk heterojunctions can be made microns thick to maximize light 15 capture without fear of losing absorbed photons to exciton diffusion lengths outside of L D to the D/A heterojunction. Figure 1.6. Cartoon representing the main energy transfer mechanisms operant in exciton migration in OPVs. (Above) Förster resonance energy transfer couples dipoles between excited (S 1 ) and ground state molecules (S 0 ). The energy transfer between the donor ( 1 D*) and acceptor (A) can occur over distances beyond the combined van der Waals radii of the molecular species and is commonly the mechanism by which singlet excited states diffuse. (Bottom) Triplet excited state energy (T 1 ) migrates through a Dexter mechanism which demands close molecular proximity and wavefunction overlap to facilitate a concerted double electron transfer between the triplet donor ( 3 D*) and acceptor (A). The distance an exciton can travel (L D ) is highly dependent on the nature of exciton formed. Singlet excitons commonly diffuse by a transition dipole coupling between excited and ground state molecules over distances larger than the combined van der Waals radii of the molecules (Figure 1.6, top). Known as Förster resonance energy transfer (FRET), the rate of this mode of energy transfer can be expressed as: (1) 16 where ‘n’ is the index of refraction, τ o is the radiative lifetime, ‘r’ is the intermolecular separation, ‘j’ is the spectral overlap integral, and ‘κ’ the orientation factor. Going as 1/τ o , FRET rates for energy transfer for states with long radiative lifetimes such as triplets (τ o = μs-ms) are small. Consequently, triplet excited states diffuse predominately by a Dexter mechanism requiring proximal molecules and wavefunction overlap between energy donor and acceptor. In this mode concerted electron exchange between donor and acceptor is responsible for triplet energy migration (Figure 1.6, bottom). The energy of triplet states, the diffusivity of these excited states, and their overall role in the mechanism of power conversion in an OPV is an important field of consideration 41 and touched upon in Chapter 4 and Chapter 6 of this document. 1.3.6 Charge Transfer and Separation Excitons at the D/A interface can dissociate by undergoing an electron transfer from the molecular excited state (either D* or A*) to the neutral species to form an interfacial charge transfer state (denoted [D + /A - ]) of energy E CT . The energy of this interfacial [D + /A - ] charge transfer state must be less than the Donor or Acceptor excited state from which it was formed through the charge transfer process. For a D* excited state, the electron transfer occurs from the excited Donor reducing the LUMO of the Acceptor (D* + A  [D + /A - ]). For an A* excited state, the electron transfer occurs from the excited Acceptor oxidizing the HOMO of the Donor (D + A*  [D + /A - ]). After electron transfer a charge transfer state species is created between a partially oxidized Donor and partially reduced Acceptor, [D + /A - ]. While an energy level offset was successfully used to separate the 17 hole-electron pair of the exciton, the geminate polaron species of the interfacial CT state are still electrostatically attracted and must still be further separated into charge separated species. 42 Charge transfer at the D/A interface will only be exothermic if: i. the energy of the Donor excited state is sufficiently reducing compared to the reduction potential of the electron Acceptor and ii. the energy of the interfacial charge transfer state ([D + /A - ]) is less than the Donor excited state (D*). Case i can be easily verified by adding the excited state energy (E 00 ) to the ionization potential of ground state Donor to give the ionization potential of D*. If this value is greater than the electron affinity of the Acceptor then forward charge transfer from D* to A will be exothermic. The oxidation and reduction potential for organic semiconductors can be arrived at using various techniques including ultraviolet photoelectron spectroscopy (oxidation) 43 and inverse photoelectron spectroscopy (reduction) 44 or solution phase cyclic voltammetry. 45 Evaluating Case ii and whether the energy of interfacial CT state is less than the optically prepared excited state (D* or A*) is not as straight forward. The energy of CT states and their role in the power conversion process in OPVs is of great research interest. However, because of poor ground state coupling with molecular species absorption into, or photoluminescence from, CT states is weak and not easily observed. This problem is further exacerbated considering CT states are only formed within nanometer domains of intimately mixed D/A regions and their density within OPV active layers very low. Bulk heterojunctions increase CT state density but not appreciably enough to be observed with common absorption or emission measurements. Alternative methods such as electroluminescence, 46 EQE modeling, photo-thermal deflection spectroscopy, and temperature-dependent device measurements have been used to probe the energy of charge transfer states and their effect on OPV performance. 18 While measuring the [D + /A - ] CT state directly is nontrivial, being able to determine its role in the power conversion process of an OPV is paramount. Once formed the geminate polaron pair [D + /A - ] can either separate into charge separated (not Coulombically bound) polarons, or recombine to a charge-neutral excited or ground state species lower in energy than the CT state. The former pathway produces free charges that can be collected at electrodes and harnessed in electrical work. The latter constitutes a loss mechanism and severely affects the electrical response of the device, scavenging efficiencies. OPV materials commonly absorb high energy photons (2.0 -3.0 eV), but cell photovoltages rarely exceed 1.0 V. Non-radiative recombination pathways of interfacial CT excitons depress achievable photovoltages and must be suppressed in order to maximize efficiencies. Further, the cell voltage cannot exceed the magnitude of the lowest energy state participant in the photo-electrical conversion process. This means that the energy of the charge transfer state (E CT ) sets a thermodynamic upper limit on the achievable photovoltage of a cell. The realized cell voltage will by default then be energy of the charge transfer state minus recombination losses. Therefore it is mandatory to both secure a high energy interfacial charge transfer state energy and minimal recombination losses to realize the full open-circuit potential of an OPV. 1.3.7 Charge Migration Once separated, holes and electrons must migrate through a continuous percolation network of Donor and Acceptor (respectively) to be collected at electrodes. During this process the energy of the electron and hole are taken to be the equivalent to the Donor ionization potential and Acceptor electron affinity. Owing to the low dielectric media and localized electronic structure of organic semiconductors, charge transport occurs through sequential local site hopping, the rate of which is governed by nonadiabatic Marcus theory of electron transfer. 47, 19 48 Morphology irregularities and inhomogeneities within amorphous films of organic semiconductors can give rise to local charge trapping sites, decreasing the rate of charge migration and deleteriously impacting device performance with increased resistances. For bulk heterojunction devices, intimate mixing of Donor and Acceptor phases can lead to isolated islands of one, pure material and as a result form incomplete percolation pathways that do not allow for charge collection. Minimizing such loss pathways is crucial for high performing OPVs, unfortunately device fabrication processes afford little to no control over material morphology. 1.3.8 Charge Collection Carrier collection in OPVs is typically accomplished with a semi-transparent electrode and vapor-deposited metal electrode. The optical properties of these electrodes are integral to the functioning of the device: semi-transparent electrodes allow for light penetration into the cell and a reflective metallic electrode creates a second pass for photons to be absorbed by the solar cell active layers. Commonly, indium-doped tin oxide (ITO) or fluorine doped tin oxide (FTO) 49 deposited on rigid glass substrates is employed as the semi-transparent electrode. Due to the cost of indium, 50 and the mechanical limitations of metal oxides, 51 the desire to move away from ITO and rigid substrates has fueled research into other electrode materials such as carbon nanotubes, 52, 53 nanowires, 54 and grapheme. 55 Regardless of the materials employed, the electrode/molecule interface responsible for charge collection at terminals presents a complicated electronic structure due to interfacial polarization effects. 55 Charge-injection interlayers between electrodes and active layers have been employed to aid in charge extraction efficiency as well as prevent undesired exciton quenching. 56, 57 Most commonly, thin buffer layers of bathocuproine 58 20 or bathophenanthroline 59 are employed at the metallic cathode-organic interface to prevent metal deposition from damaging the active layer of the solar cell. 1.4 J-V Response and Efficiency The electrical response of OPVs is characterized by measuring the current response of the device as a function of applied voltage. Typically, the voltage is swept from reverse bias to forward bias and the current density of the cell recorded both in the dark and under illumination. In reverse bias holes are injected into the Acceptor at the metallic cathode and electrons injected into the Donor at the semi-transparent anode. When in forward bias holes are injected into the Donor layer at the anode and electrons are injected into the Acceptor layer at cathode. In the dark, the cell behaves as a diode, producing dark current, or saturation current (J s ) with minimal voltage dependence in reverse bias and an exponential current density increase upon rectification at some applied potential in forward bias. For minimal resistive losses the current density (J) and saturation current density (J s ) can be expressed with the generalized Shockley equation as: (2) (3) where 'q' is the elementary charge, 'V' is the applied bias, 'n' the ideality factor of the diode, 'k' Boltzmann constant, 'T' the temperature, and J 00 is a material-system dependent parameter. ∆E is the energy barrier at the interface which must be overcome to generate charge carriers and represents the energy offset at the D/A interface. For molecular systems this energy offset is the difference between the Donor HOMO (hole carrier) and Acceptor LUMO (electron carrier). For diodes with idealities factors ≈ 2, J s is characteristic of charge carrier recombination. Practically, 21 Equations 2 and 3 demonstrate that the magnitude of J s determines the amount of current (J) present in the dark, and an inverse exponential scaling of J s with an increasing energy offset at the D/A interface. Under illumination the OPV generates photocurrent by the operating mechanism detailed in Section 1.3. The magnitude of the current density and its voltage dependence can be simplified as (4) with J ph being the photogenerated current. At no applied bias (V = 0), the magnitude of the current density is the difference between J s and J ph . Since J ph >> J s , at no applied bias the short circuit current (J sc ) is assumed to be J sc = -J ph under illumination. With no external load the cell will output no net current while illuminated and the voltage drop between the contacts of the cell is equivalent to the open-circuit voltage (V oc ). At open- circuit conditions, J = 0 and Equation 1 simplifies to: (5) Here it can be seen that the magnitude of the open-circuit voltage (V oc ) in the light is dependent upon a ratio of the generated photocurrent (J sc ) and the dark saturation current (J s ), or more simply, the ratio of generation charges extracted to charges lost to recombination. Therefore, in order to maximize the open-circuit voltage, it is paramount to minimize the amount of charge carrier recombination loss. Insertion of Equation 2 into (4) yields: (6) assuming exp(qV oc /nkT) >> 1. Analysis of Equation 6 reveals that ∆E/q is the limit of V oc as T 0. The thermodynamic limits affecting the open-circuit voltage and its relation to the energy 22 of the interfacial charge transfer state ([D + /A - ]) introduced in Section 1.3 will be further discussed in Chapter 6. Kinetically, open-circuit conditions are reached when the rate of charge injection into the cell supplied by an external bias equals the photocurrent generation rate under illumination such that no net current flows (J = 0). The product of current (amps) and voltage (volts) is power (watts). In an OPV the maximum theoretical power output of the cell (P theo ) occurs at J sc x V oc . However, at J sc (V = 0) and at V oc (J = 0) no power is generated. For voltages greater than zero but less than V oc the cell outputs power J x V with the maximum obtainable power occurring at P max . The fill factor, FF, represents the fractional amount of actual power output compared to the theoretical power capacity (P max /P theo ) of the device and can be shown graphically (Figure 1.7xx). These three metrics together constitute the power conversion efficiency (η) of the solar cell: (7) Figure 1.7. Idealized current-voltage response of an OPV. In the dark the device behaves as a diode, with minimal voltage-dependent current response until a rectifying voltage applied in forward bias allows free carriers to flow. In the light the cell outputs photocurrent J sc at zero applied bias. At some voltage in forward bias (V oc ) the applied potential counters the electrical output of the cell and no current flows. The product of J sc and V oc represents the theoretical maximum power of the cell (grey square) which is ultimately unobtainable. At some voltage, V < V oc 23 the maximum achievable power output of the cell will be realized and the fraction of this achieved power to the theoretical power is known as the fill factor (yellow square). Graphically, the fill factor can be represented as the proportion of the area of the yellow square to the area of the grey square. Together, all three device metrics constitute the overall power conversion efficiency of the OPV, η = FF x J sc x V oc . In order to maximize the overall power conversion efficiency it is pertinent to maximize each the FF, V oc , and J sc . The molecular aspects governing FF, V oc , and J sc are quite complex and often interrelated. Bathochromic shifting absorption of OPV active layers to capture longer wavelengths of solar flux and increase spectral coverage increases short-circuit current density. However, 0.5-1.0 eV energy offsets necessary for exothermic electron transfer to overcome large exciton binding energy prohibit low energy excited states from providing large open circuit voltages. Ordered molecular morphologies can increase charge conductivity, increasing FF, however the deposition techniques common to organic device processing afford little control over morphology and typically result in either amorphous or polycrystalline phases, the latter exhibiting crystallite grain boundaries responsible for exciton and charge quenching sites. A complementary measurement technique important for evaluating the performance of a solar cell is an External Quantum Efficiency (EQE) measurement. This technique quantifies the ratio of incident photons to electrical charges collected by the solar cell for a given wavelength of light. While the J sc of the device quantifies the amount of current produced by the cell, the EQE indicates which incident photons are responsible for that current. 1.5 The Shockley Queisser Limit With OPV efficiencies steadily increasing the question of the ultimate efficiency limits of such devices becomes an important consideration. For inorganic semiconductors this question was answered by Shockley and Queisser half a century ago when they were able to predict a 24 thermodynamic upper limit on power conversion efficiencies of single junction devices using a detailed-balance approach with minimal assumptions. 60 Among these assumptions were: 1. A semiconductor of bandgap E g absorbs photons of energy hν > E g with unit efficiency 2. Absorbed photons greater than E g thermalize to the band edge, dissipating excess energy as heat. 3. All photons absorbed contribute exactly one electron-hole pair to the external circuit 4. All photons below the bandgap energy (hν < E g ) are not absorbed 5. All recombination losses are radiative With these few assumptions Shockley-Quessier were able to derive the thermodynamic efficiency limits of single junction semiconductor devices with E g as the only variable. For a material with a band gap such as silicon (E g = 1.1 eV) the Shockley-Quessier limit (SQ) predicts a maximum power conversion efficiency of ~32%. While an appreciable percentage of available photons can be captured assuming unity absorption by a semiconductor with a E g = 1.1 eV band gap, the overall power conversion efficiency of such a device is fundamentally limited by Figure 1.8. Graphic representation of the thermalization process that occurs when a photon in excess energy of a semiconductor’s bandgap (hν > E g ) is absorbed and dissipated as heat as free carriers cool to the band edge. 61 25 excessive heat dissipation of all photons with energy in excess of E g (Figure 1.8). Future high efficiency devices must therefore circumvent the large energy dissipation problem fundamentally crippling single junction devices in the SQ treatment. The excitonic nature of organic semiconductor excited states demands modifications to the Shockley-Quessier treatment for an accurate description of efficiency limits for OPVs. 62 Among these is the introduction of a Donor-Acceptor heterojunction to facilitate exothermic charge transfer and a CT state of energy E CT as an intermediate in the power-production process. 63 In much the same Shockley-Quessier treatment, all absorbed photons are assumed to thermalize to E CT , dissipating excess energy as heat to the surrounding media, and each absorbed photon contributes one electron-hole pair to the external circuit. Given the further restrictions a heterojunction demanding exothermic electron transfer imposes, single junction OPV efficiency limits have been predicted to be 20 - 25%. While certainly lower than their idealized inorganic counterparts, these calculated efficiencies demonstrate how far the OPV technology has yet to improve and the ceiling it can potentially achieve. 1.6 Alternate Routes to Efficiency Efficiency calculations for both inorganic and organic photovoltaics traditionally hinge upon assumptions such as thermalization losses and a one-for-one photon-to-electron conversion ratio. With proper materials choices and device engineering researchers have approached avenues to push device efficiencies beyond those of the SQ limit. The construction of tandem devices is one common approach to help minimize thermalization losses. 64 Such devices use subcells each with bandgaps engineered to selectively absorb particular regions of the solar spectrum. Having individual subcells with bandgaps selected to properly absorb particular 26 wavelengths instead of demanding panchromatic absorption from a single material with an intrinsic bandgap diminishes thermalization losses. When the subcells are electrically connected in series the overall tandem cell voltage is additive, but the total current output is limited by the lowest current production of any one subcell due to the necessity of transparent interlayers between cells to act as recombination centers for charge balancing (known as current matching). 65 Creating tandem solar cells with properly balanced bandgaps and current matching presents large engineering challenges that could be relaxed by using carrier multiplication schemes that accomplish a similar feat. The ability to harness more than one electron from a photon through a carrier multiplication scheme is another way to potentially circumvent the SQ limit. 66 In this approach high energy photons (hν) interacting with a semiconductor create multiple low energy excited states (E 1 , E 2 , …, E n ) instead of one high energy excited state. Since energy must be conserved, the summation of energy over all created states must be less than or equal to the energy of high energy excitation photon (E 1 + E 2 + … + E n < hν). Normally, when a semiconductor absorbs a photon of energy in excess of its band gap, the additional energy is rapidly dissipated as heat and constitutes a fundamental loss mechanism for solar cells, as mentioned previously. When the same photon’s excess energy is instead used to create multiple, low energy excited states, thermalization losses are minimized as the full energy photon has been preserved in the form of multiple excited states and not lost as heat. In this approach the solar cell outputs more than one electron-hole pair per absorbed photon. Inorganic quantum dots have been shown to create multiple excitations when irradiated with photons in excess of its band gap 61, 67 and photovoltaic devices exhibiting external photocurrent quantum efficiencies in excess of 100% have even been constructed. 68 In certain organic chromophores a photophysical process known as ‘singlet 27 fission’ (SF) creates two triplet excited states from one initial singlet excited state. 69, 70 Effectively utilizing SF in an OPV would mitigate the large thermalization losses experienced from absorption of high energy photons dissipating excess heat by preserving the full energy of the photon as two low energy triplet excited states. When paired with a complementary absorbing chromophore to ensure panchromatic photon capture, SF has been calculated to achieve theoretical power conversion efficiencies over 40% for devices (Figure 1.9). 71 If organic photovoltaics are to ever be a Figure 1.9. (Left) Schematic representing the operating principle of advantageous singlet fission in OPVs. Under the SQ treatment, all photon energies absorbed in excess of the S 1 state are thermalized as excess heat. Introduction of a singlet fission material converts high energy photons to two low energy triplets, preserving the full energy of the photon. When complemented with a second, complementary absorption material, singlet fission increases the efficiency of single junction devices. (Right) Plot of the total solar irradiance as a function of wavelength. Full photon capture by a semiconductor of band gap E g = 1.1 eV operates with an efficiency of ~ 32% (red area). Introduction of a SF chromophore than converts 2.2 eV singlets to 1.1 eV triplets increases the overall power conversion efficiency of the cell to ~42% (red + blue area) when paired with a complementary second absorber. market viable technology, their overall power conversion efficiencies must be drastically improved. Harnessing a phenomenon such as singlet fission to potentially achieve efficiencies beyond the thermodynamic limit attainable for a single junction offers an avenue to realize an efficient device without demanding the engineering of tandem devices. For this reason the prospects of exploiting singlet fission in OPVs has garnered significant research attention. 28 The remainder of this document is devoted to an investigation of the molecular aspects governing singlet fission in organic chromophores, particularly with an emphasis on understanding the phenomenon so that it may be ultimately exploited in next generation, high efficiency OPVs. To this end, Chapter 2 introduces the molecular properties governing SF, and the pertinent photophysical behaviors to be understood for its use in solar cells. Subsequently, Chapters 3 and 4 document investigations of SF in prototypical condensed-phase, disordered acene films, and a novel energy transfer scheme between a SF chromophore and a complementary low-energy absorber, respectively. Chapter 6 examines the operating mechanism of SF in representative OPVs using a custom-built device testing apparatus whose design and construction is described in Chapter 5. However, effective use of SF in OPVs can only be realized once the pertinent aspects governing the power conversion process and device design practices paramount to OPVs as introduced in this Chapter are first understood. 1.7 References (1) https://www.fas.org/sgp/crs/nuke/R41694.pdf. (2) http://www.world-nuclear.org/information-library/country-profiles/countries-g- n/germany.aspx. (3) http://www.eia.gov/todayinenergy/detail.cfm?id=26372. (4) http://www.world-nuclear.org/information-library/country-profiles/countries-g-n/japan- nuclear-power.aspx. (5) http://climate.nasa.gov/vital-signs/carbon-dioxide/. (6) http://archive.defense.gov/pubs/150724-congressional-report-on-national-implications- of-climate-change.pdf. (7) Lewis, N. S.; Nocera, D. G.; Powering the Planet: Chemical Challenges in Solar Energy Utilization. P. Natl. Acad. Sci. USA, 2006, 103, 15729-15735. 29 (8) http://www.asrc.albany.edu/people/faculty/perez/Kit/pdf/a-fundamental-look-at%20the- planetary-energy-reserves.pdf. (9) Nelson, J. The Physics of Solar Cells, World Scientific, Singapore 2003. (10) Würfel, P. Physics of Solar Cells, Wiley-VCH, Weinheim, Germany 2005. (11) Slaoui, A.; Collins, R. T. Advanced Inorganic Materials for Photovoltaics, MRS Bull., 2007, 32, 211-218. (12) Ramanathan, K.; Contreras, M. A.; Perkins, C. L.; Asher, S.; Hasoon, F. S.; Keane, J.; Young, D.; Romero, M.; Metzger, W.; Noufi, R.; Ward, J.; Duda, A. Properties of 19.2% Efficiency ZnO/CdS/CuInGaSe 2 Thin-Film Solar Cells, Prog. Photovoltaics, 2003, 11, 225-230. (13) Cuo, A.; Ford, G. M.; Hillhouse, H. W.; Agrawal, R. Sulfide Nanocrystal Inks for Dense Cu(In 1-x Ga x )(S 1-y Se y ) 2 Absorber Films and Their Photovoltaic Performance. Nano Lett., 2009, 9, 3060-3065. (14) Aramoto, T.; Kumazawa, S.; Higuchi, H.; Arita, T.; Shibutani, S.; Nishio, T.; Nakajima, J.; Tsuji, M.; Hanafusa, A.; Hibino, T.; Omura, K.; Ohyama, H.; Murozono, M. 16.0% Efficient Thin-Film CdS/CdTe Solar Cells, Jpn. J. Appl. Phys., 1997, 36, 6304-6305. (15) Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W.; Dunlop, E. D. Solar Cell Efficiency Tables (Version 47). Prog. Photovolt. Res. Appl., 2016, 24, 3-11. (16) https://www.ise.fraunhofer.de/de/downloads/pdf-files/aktuelles/photovoltaics-report-in- englischer-sprache.pdf. (17) Komp, R. J.; Perlin, J. Practical Photovoltaics. Electricity from Solar Cells 3 rd Ed. Aatec Publications, 2001. (18) Goetzberger, A.; Hebling, C.; Schock, H.-W. Photovoltaic Materials, History, Status and Outlook, Mater. Sci. Eng. Rep., 2003, 40, 1-46. (19) Hoffert, M. I.; Calderira, K.; Benford, G.; Criswell, D, R.; Gree, C.; Herzog, H.; Jain, A. K.; Kheshgi, H. S.; Lackner, K. S.; Lewis, J. S.; Lightfoot, H. D.; Manheimer, W.; Mankins J. C.; Mauel, M. E.; Perkins, L. J.; Schlesinger, M. E.; Volk, T.; Wigley, T. M. L. Science, 2002, 298, 981-987. (20) Lewis, N. S.; Toward Cost-Effective Solar Energy Use Science, 2007, 315, 798-799. (21) Hoppe, H.; Sariciftci, N. S.; Organic Solar Cells: An Overview. J. Mater. Res., 2004, 19, 1924-1945. (22) http://www.gm.dk/en/solar. 30 (23) Angmo, D.; Gevorgyan, S. A.; Larsen-Olsen, T. T.; Sondergaard, R. R.; Hosel, M.; Jorgensen, M.; Gupta, R.; Kulkarni, G. U.; Krebs, F. C.; Scalability and Stability of Very Thin, Roll-to-Roll Processed, Large Area, Indium-Tin-Oxide Free Polymer Solar Cell Modules. Org. Elec. 2013, 14, 984-994. (24) Kippelen, B.; Bredas, J-L.; Organic Photovoltaics, Energy Environ. Sci, 2009, 2, 251- 261. (25) http://www.greentechmedia.com/articles/read/konarka-gets-20m-more-for-organic-solar- cells. (26) http://www.heliatek.com. (27) http://www.nrel.gov/ncpv/images/efficiency_chart.jpg. (28) Jorgensen, M.; Norrman, K.; Krebs, F. C.; Stability/Degredation of Polymer Solar Cells. Solar Energy Mat and Solar Cells. 2008, 92, 686-714. (29) Gevorgyan, S. A.; Madsen, M. V.; Dam, H. F.; Jorgensen, M.; Fell, C. J.; Anderson, K. F.; Mescheloff, A.; Katz, E. A.; Elschner, A.; Roesch, R.; Hoppe, H.; Hermenau, M.; Riede, M.; Krebs, F. C.; Interlaboratory Outdoor Stability Studies of Flexible Roll-to- Roll Coated Organic Photovoltaic Modules: Stability over 10,000 h. Solar Energy Mat. and Solar Cells. 2013, 116, 187-196. (30) Tipnis, R.; Bernkopf, J.; Jia, S.; Krieg, J.; Li, S.; Storch, M.; Laird, D.; Large-area Organic Photovoltaic Module-Fabrication and Performance. Solar Energy Mat and Solar Cells. 2009, 93, 442-446. (31) Forrest, S. R. The Path to Ubiquitous and Low-Cost Organic Electronic Appliances on Plastic. Nature, 2004, 428, 911-918. (32) Bardeen, C. J.; The Structure and Dynamics of Molecular Excitons. Annu. Rev. Phys. Chem. 2014, 65, 127-148. (33) Rand, B.P.; Burk, D. P.; Forrest, S. R. Offset Energies at Organic Semiconductor Heterojunctions and Their Influence on the Open-Circuit Voltage of Thin-Film Solar Cells. Phys Rev. B 2007, 75, 115327:1-11. (34) Tang, C. W.; Two-layer Organic Photovoltaic Cell. Appl Phys Lett. 1986, 48, 183-185. (35) Opitz, A.; Wagner, J.; Brutting, W.; Salzmann, I.; Koch, N.; Manara, J.; Pflaum, J.; Hinderhofer, A.; Schreiber, F.; Charge Separation at Molecular Donor-Acceptor Interfaces: Correlation Between Morphology and Solar Cell Performance. IEEE J. Sel. Top. Quant. Elec, 2010, 16, 1707-1717. 31 (36) Frackowiak, D.; Planner, A.; Waszkowiak, A; Boguta, A.; Ion, R.-M.; Wiktorowicz, K. Yield of Intersystem (Singlet-Triplet) Crossing in Phthalocyanines Evaluated on the Basis of a Time in Resolved Photothermal Method. J. Photoch. Photobiol. A, 2001, 141, 101-108. (37) Forster, L. S; Intersystem Crossing in Transition Metal Complexes. Coord. Chem. Rev. 2006, 250, 2023-2033. (38) Schlenker, C. W.; Thompson, M. E.; The Molecular Nature of Photovoltage Losses in Organic Solar Cells. Chem Comm. 2011, 47, 3702-3716. (39) Fu, Y.-T.; Risko, C.; Bredas, J.-L.; Intermixing at the Pentacene-Fullerene Bilayer Interface: A Molecular Dynamics Study. Adv. Mater. 2013, 25, 878-882. (40) Sai, N.; Gearba, R.; Dolocan, A.; Tritsch, J. R.; Chan, W.-L.; Chelikowsky, J. R.; Leung, K.; Zhu, X. Understanding the Interface Dipole of Copper Phthalocyanine (CuPc)/C 60 : Theory and Experiment. J. Phys. Chem. Lett., 2012, 3, 2173. (41) Schlenker, C. W.; Chen, K-S.; Yip, H-L.; Li, C-Z.; Bradshaw, L. R.; Ochsenbein, S. T.; Ding, F.; Li, X. S.; Gamelin, D. R.; Jen, A. K-Y.; Ginger, D. S.; Polymer Triplet Energy Levels Need Not Limit Photocurrent Collection in Organic Solar Cells. JACS, 2012, 134, 19661-19668. (42) Bassler, H.; Kohler, A.; “Hot or Cold”: How do Charge Transfer States at the Donor- Acceptor Interface of an Organic Solar Cell Dissociate? Phys. Chem. Chem. Phys., 2015, 17, 28451-28462. (43) Tadayyon, S. M.; Gradndin, H. M.; Griffiths, K.; Coatsworth, L. L.; Norton, P. R.; Aziz, H.; Popovic, Z. D.; Reliable and Reproducible Determination of Work Function and Ionization Potentials of Layers and Surfaces Relevant to Organic Light Emitting Diodes. Org. Elect. 2004, 5, 199-205. (44) Cahen, D.; Kahn, A. Electron Energetics at Surfaces and Interfaces: Concepts and Experiments. Adv. Mater., 2003, 15, 271-277. (45) Djurovich, P. I.; Mayo, E. I.; Forrest, S. R.; Thompson, M. E.; Measurement of the Lowest Unoccupied Molecular Orbital Energies of Molecular Organic Semiconductors. Org. Elect. 2009, 10, 515-520. (46) Tvingstedt, K.; Vandewal, K.; Gadisa, A.; Zhang, F.; Manca, J.; Inganas, O.; Electroluminescence from Charge Transfer States in Polymer Solar Cells. JACS, 2009, 131, 11819-11824. (47) Marcus, R. A. On the Theory of Oxidation-Reduction Reactions Involving Electron Transfer, J. Chem. Phys., 1956, 24, 966-978. 32 (48) Marcus, R. A.; Sutin, N. Electron Transfers in Chemistry and Biology. Biochim. Biophys. Acta, 1985, 811, 265-322. (49) Andersson, A.; Johansson, N.; Broms, P.; Yu, N.; Luop, D.; Salaneck, W. R. Fluorine Tin Oxide as an Alternative to Indium Tin Oxide in Polymer LEDs, Adv. Mater., 1998, 10, 859-863. (50) http://www.compoundsemiconductor.net/article/-Indium-price-soars-as-demand-for- displays-continues-to-grow.html. (51) Scott, J. C.; Kaufman,J.H.; Brock, P. J.; DiPietro, R.; Salem, J.; Goitia, J. A. Degradation and Failure of MEH-PPV Light Emitting Diodes. J. Appl. Phys., 1996, 79, 2745-2751. (52) Wu, Z.; Chen, Z.; Du, X.; Logan, J.M.; Sippel, J. Nikolou, M.; Kamaras, K. Reynolds, J. R.; Tanner, D. B.; Hebard, A. F.; Rinzler, A. G. Transparent, Conductive Carbon Nanotube Films, Science, 2004, 305, 1273-1276. (53) Rowell M. W.; Topinka, M. A.; McGehee, M. D.; Prall, H.-J.; Dennler, G.; Sarciftci, N. S.; Hu, L.; Gruner, G. Organic Solar Cells with Carbon Nanotube Network Electrodes. Appl. Phys. Lett., 2006, 88, 233506. (54) Lee, J.-Y.; Connor, S. T.; Cui, Y.; Peumans, P. Solution-Processed Metal Nanowire Mesh Transparent Electrodes. Nano Lett., 2008, 8, 689-692. (55) De Arco, L. G.; Zhang, Y.; Schlenker, C. W.; Ryu, K.; Thompson, M. E.; Zhou, C.; Continuous, Highly Felxible, and Transparent Graphene Films by Chemical Vapor Deposition for Organic Photovoltaics. ACS Nano, 2010, 5, 2865-2873. (56) Grob, S.; Gruber, M.; Bartynski, A. N.; Hormann, U.; Linderl, T.; Thompson, M. E.; Brutting, W.; Amorphous vs Crystalline Exciton Blocking Layers at the Anode Interface in Planar and Planar-Mixed Heterojunction Organic Solar Cells. Appl. Phys. Lett. 2014, 104, 213304:1-5. (57) Tress, W.; Leo, K.; Riede, M.; Influence of Hole-Transport Layers and Donor Materials on Open-Circuit Voltage and Shape of I-V Curves of Organic Solar Cells. Adv. Funct. Mater., 2011, 21, 2140-2149. (58) Gommans, H.; Verreet, B.; Rand, B. P.; Muller, R.; Poortmans, J.; Heremans, P.; Genoe, J.; On the Role of Bathocuproine in Organic Photovoltaic Cells. Adv. Funct. Mater., 2008, 18, 3686-3691. (59) Naka, S.; Okada, H.; Onnagawa, H.; Tsutsui, T., High Electron Mobility in Bathophenanthroline, Appl. Phys. Lett., 2000, 76, 197-198. 33 (60) Shockley, W.; Queisser, H. J. Detailed Balance Limit of Efficiency of p-n Junction Solar Cells. J. Appl. Phys. 1961, 32, 510-519. (61) Beard, M. C.; Johnson, J. C.; Luther, J. M.; Nozik, A. J.; Multiple Exciton Generation in Quantum Dots Versus Singlet Fission in Molecular Chromophores for Solar Photon Conversion. Phil. Trans. R. Soc. A. 373:20140412. (62) Gregg, B. A.; Hanna, M. C.; Comparing Organic to Inorganic Photovoltaic Cells: Theory, Experiment, and Simulation. J. Appl. Phys. 2003, 93, 3605-3614. (63) Koster, L. J. A.; Shaheen, S. E.; Hummelen, J. C. Pathways to a New Efficiency Regime for Organic Solar Cells. Adv. Energy Mater., 2012, 2, 1246-1253. (64) Zuojia, K. L.; Feng, K.; X. Xiaopeng, X. Wang, L.; Peng, Q.; Development of Large Band-Gap Conjugated Copolymers for Efficient Regular Single and Tandem Organic Solar Cells. JACS. 2013, 135, 13549-13557. (65) Zhou, Y.; Fuentes-Hernandez, C.; Shim, J. W.; Khan, T. M.; Kippelen, B.; High Performance Polymeric Charge Recombination Layer for Organic Tandem Solar Cells. Energy Environ. Sci, 2012, 5, 9827-9832. (66) Nelson, C. A.; Monahan, N. R.; Zhu, X.-Y.; Exceeding the Schockly-Quessier Limit in Solar Energy Conversion. Energy Environ. Sci. 2013, 6, 3508-3519. (67) McGuire, J. A.; Joo, J.; Pietryga, J. M.; Schaller, R. D.; Klimov, V. I.; New Aspects of Carrier Multiplication in Semiconductor Nanocrystals. Acc. Chem Res. 2008, 41, 1810- 1819. (68) Semonin, O. E.; Luther, J. M.; Choi, S.; Chen, H-Y.; Goa, J.; Nozik, A. J.; Beard, M. C.; Peak External Photocurrent Quantum Efficiency Exceeding 100% via MEG in a Quantum Dot Solar Cell. Science, 2011, 334, 1530-1533. (69) Smith, M. B.; Michl, J. Singlet Fission Chem. Rev., 2010, 110, 6891-6939. (70) Smith, M. B.; Michl, J.; Recent Advances in Singlet Fission. Annu. Rev. Phys. Chem,. 2013, 64, 61-86. (71) Hanna, M. C.; Nozik, A. J.; Solar Conversion Efficiency of Photovoltaic and Photoelectrolysis Cells with Carrier Multiplication Absorbers. J. Appl. Phys., 2006, 100, 074510. 34 Chapter 2. Molecular Aspects of Singlet Fission: Photophysics and Devices 2.1 Overview of Singlet Fission 2.1.1 Introduction Singlet fission (SF) is a molecular photophysical process involving organic chromophores wherein one molecule in its singlet excited state (S 1 ) shares the energy of the excited state wavefunction with another molecule in its singlet ground state (S 0 ) then “fissions,” leaving both molecules in their distinct, separate triplet excited states (T 1 + T 1 ). 1,2 SF is a spin- conserving process (i.e. no electron has to flip its spin) and as a result can be very fast, occurring on sub-picosecond timescales, and can readily out-compete other S 1 deactivation pathways such as fluorescence (τ ~ ns). Since SF demands at least two molecular species for triplet production, spectroscopic studies are commonly carried out on neat thin films, single crystals, or molecular aggregates of qualifying chromophores. Over time increasing amounts of covalently-bound dimers of SF molecules have been synthesized in an attempt to probe the dynamics of SF with the two molecules constrained to a particular mutual geometry. 3-8 These dimers tend to be highly soluble in common organic solvents and offer a unique opportunity to investigate singlet fission dynamics in fluid media of variable polarity. While dimers in solution offer a conceptually simplified platform, their low yielding synthesis and instability compound the difficulty of their study. The phenomenon was first observed in 1965 in an attempt to understand the excited state photophysics of anthracene crystals. 9 Shortly thereafter singlet fission was observed in another member of the acene family, tetracene. 10-12 In the past half century since its initial discovery, research into singlet fission was minimal at best with little progress being made until recently. 35 The advent of organic electronics coupled with singlet fission’s potential utility in modern photovoltaic devices has seen a renaissance in fission research. In the past decade significant advances have been made towards understanding the mechanism of singlet fission, identifying and synthesizing chromophores capable of SF, and exploring the utilization of SF in optoelectronic devices. 1,2 This chapter will introduce the pertinent aspects of molecular singlet fission. Subsequent chapters will discuss photophysical investigations of SF in disordered thin films of a substituted acene derivative (Chapter 3) and the interplay between singlet fission and a complementary low-energy sensitizer (Chapter 4), and then concludes with the construction of hardware to probe SF in operating organic photovoltaics using magnetic fields (Chapters 5, 6). 2.1.2 Thermodynamic Requirements for Singlet Fission During singlet fission a molecule in its singlet excited state shares it energy with a neighboring molecule in its singlet ground state and then fissions, leaving both molecules in their respective triplet excited states (Figure 2.1). Typically, the two molecules are the same type of chromophore and the SF process can generally be written as: (1) with the ‘a’ and ‘b’ subscripts denote and designate the two molecules involved in the process (Note that a process involving fission between two different types of chromophores is known to exist, deemed ‘heterofission 3,4,13,14 but will not be the subject of further investigation in this document) . 36 Figure 2.1. Cartoon schematic illustrating the singlet fission process: Absorption of a photon (E = hν) promotes a chromophore from its singlet ground state (S 0A ) to the singlet excited state (S 1A , green arrow). The energy of the singlet excited state wavefunction is shared with a neighboring molecule in its singlet electronic ground state (S 0B ) then fissions (red arrows), leaving both molecules in their distinct, separate triplet excited states (T 1A + T 1B ). Since energy must be conserved, in order for SF to be thermodynamically favorable an immediate energy requirement is apparent: the energy of the molecule’s lowest triplet excited state (T 1 ) must be approximately half the energy of the lowest energy singlet excited state (S 1 ). (2) Obviously, the energy of higher lying singlet states (S n , where n >1) can easily satisfy Equation 2, but rapid vibrational relaxation converting higher lying state energy to lower energy states plus dissipated heat (S n →S 1 + ∆) demands only S 1 and T 1 meet this stringent energy requirement. In order for triplet yields from SF to be high not only does the fission process need to be efficient, but the back reaction of Equation 1, known as triplet-triplet annihilation, must be suppressed. 15-17 Triplet-triplet annihilation is a well-known process in OLEDs and deleteriously impacts performance at high current densities in those devices. 18 Annihilation also negatively impacts singlet fission as triplet excitons created from the fission process can fuse and diminish the advantageous return of two triplets per one singlet. Since energy must, once again, be 37 conserved during a triplet-triplet annihilation event, many different high energy excited states can be accessed including higher lying singlet states, triplet states, and even quintet states. For SF molecules where E(S 1 ) > 2E(T 1 ), triplet-triplet annihilation is too endothermic to repopulate S 1a + S 0b . Further, T 1a + T 1b deactivating back to ground state (S 0a + S 0b ) will be slow as it is considerably exothermic and occurs in the Marcus inverted region. However, if E(T 2 ) ≈ 2E(T 1 ) then triplet-triplet annihilation could feasibly repopulate T 2 . If two triplets produced from singlet fission annihilate such that (3) then one excitation has been lost and T 2 will rapidly internally convert to T 1 . Therefore, in addition to SF being exothermic via E(S 1 ) > 2E(T 1 ), it is preferential also for E(T 2 ) > 2E(T 1 ) to minimize recombination pathways. In order for triplet yields from SF to be high the fission process must kinetically out- compete other deactivation pathways of the S 1 state such as fluorescence or internal conversion. Intersystem crossing from the singlet manifold to the triplet manifold is another unfavorable deactivation pathway. While intersystem crossing produces triplets from singlets, it only does so in a 1:1 ratio (one singlet produces one triplet) instead of the 1:2 ratio (one singlet produces two triplets) possible from fission and thus should be avoided. Since singlet fission is a spin conserving process, triplet production occurs orders of magnitude faster than the nanosecond fluorescent lifetimes or microsecond rates of intersystem crossing typical in organic hydrocarbons. For this reason it is desirable for a targeted SF molecule to have a high fluorescence quantum yield, and as a result a low non-radiative recombination rate, as an isolated chromophore in solution. If the SF compound exhibits minimal non-radiative recombination 38 pathways as an isolated, solvated molecule, then there are no native recombination pathways to kinetically compete with SF, and once in condensed media the fission process will have the full radiative lifetime of the molecule to occur. Indeed, a large magnitude of quenched fluorescence from solution to condensed media can be one initial check to suggest positive triplet production from SF. 2.1.3 Singlet Fission Chromophores Since the resurgence in singlet fission research numerous organic chromophores have been shown to undergo fission. The most common of these is the acene family including tetracene, pentacene and recently, hexacene. 19,20 Since the absorptive transitions of tetracene and pentacene fall in the visible spectrum, these acenes and their derivatives have found common use as OPV active layer Donors. Ultrafast transient absorption experiments have shown triplet formation in tetracene films and crystals occurs between 40-100 ps with near unity quantum yield (100% of singlet excited states undergo fission, or 200% yield of triplet excited states from initially prepared singlets). 21 Interestingly, SF in tetracene is slightly endothermic since its triplet energy (E T1 = 1.23 eV) is greater than half the singlet energy (E S1 = 2.32 eV). Photophysical studies on tetracene thin films and OPVs have shown pronounced temperature dependence on triplet yield and production rate as a result of SF being thermally activated. 22 Unlike tetracene, SF in pentacene systems is highly exoergic (E S1 = 1.83 eV, E T1 = 0.86 eV) and as a result occurs an order of magnitude faster than tetracene, ~80 fs, also with unity quantum yield. The process is so fast, even, that the fluorescence quantum yield of solid pentacene is effectively zero. Hexacene, owing to its unique energy levels, has a fission time constant of about 5 ps. Substituted acens such as rubrene, 23 5,12-diphenyltetracene, 24,25 and TIPS- pentacene 26 have also been shown to undergo fission in moderate to high yields. By chemically 39 substituting the acene backbone it is possible to induce new packing motifs between neighboring molecules with minimal perturbation of state energies. In this way it is possible to study the effect of inter-chromophore orientation and coupling on SF without altering thermodynamics. 27- 30 1,3-diphenylisobenzofuran is a biradicaloid species whose SF yield has been shown to drastically depend upon polymorphism. 30,31 Thin films of 1,3-diphenylisobenzofuran grown on sapphire exhibit triplet formation on the order of 2-25 ps, with a yield of approximately 200 ± 30%. On other substrates, the triplet yield is an order of magnitude smaller even though X-ray powder patterns show similar, but not identical, crystal structures as films grown on sapphire. Carotenoids are long oligomeric molecules containing linearly conjugated double bonds. Their chemical structure gives rise to a unique electronic configuration wherein the S 1 of 2A g - character is symmetry forbidden from undergoing one-photon absorption from the ground state. The highly allowed 1Bu + absorption is the first optically allowed transition and occurs at a much higher energy than S 1 . As a result, S 1 population occurs through 1Bu + excitation followed by internal conversion. In aggregates of 3R,3’R-zeaxanthin (a carotenoid with 11 conjugated double bonds, all of anti configuration), picosecond resonance-Raman spectroscopy has shown triplet yields of ~200% with T 1 signal growth over multiexponential time constants (5-7 ps, 600- 700 ps, 3+ ns). 32-33 The quest for more SF chromophores has yielded an even larger molecular toolbox than is meant to be summarized here. Quinoidal bithiophenes, 34 diphenylhexatriene, 29 perylene dicarboximides (PDIs), 26,28 and polymers 35 have all been shown to exhibit SF to some degree. The reader is encouraged to reference the literature for details pertaining to a particular system of interest. 40 2.1.4 Mechanism of Singlet Fission The mechanistic details pertinent to the singlet fission process have been the subject of numerous physical and computational investigations. Even with such active research a definitive picture of the fission process has yet to be realized. The full details of the SF process are beyond the scope of this document but will be touched upon in brief. The conversion of one singlet to two triplets via singlet fission can be expressed by expanding Equation 1 to the form: (4) where 1 (TT) represents a correlated triplet pair existing of two triplets on different molecules but of pure singlet spin. 36 Simply, 1 (TT) represents a multi-exciton state (both molecules in their triplet excited states) where both molecules’ spin functions are still coupled, while “T 1a + T 1b ” represents a state where both triplets are now independent and their spins have decohered. Commonly, 1 (TT) is also referred to as a “multiexciton state” or a “doubly excited state.” Given its multiexcitonic nature 1 (TT) tends to be one-photon spectroscopically “dark” from molecular closed-shell singlet ground states and not easily observable. The existence of 1 (TT) and its role in singlet fission has been probed by other methods, however, such as two-photon photoemission 37,38 and 2-D electronic spectroscopy. 39 41 Figure 2.2. Schematic representation of the two possible mechanisms by which singlet fission proceeds. In the direct mechanism the rate of fission is dependent on the size of the two electron coupling <T 1 T 1 |H|S 1 S 0 > matrix element and is modulated by the percent of orbital overlap between the two molecules. The mediated case can proceed through charge transfer states as either real or virtual intermediates that exchange population through a superexchange mechanism. Using a simple four frontier orbital model with four electrons it is possible to derive two potential mechanisms for the conversion of S 1a + S 0b to 1 (TT): a direct mechanism facilitated by direct two-electron coupling between S 1 S 0 and 1 (TT) or a mediated CT mechanism demanding the involvement of charge transfer intermediates as either real or virtual states (Figure 2.2). To date conclusive evidence favoring either mechanism has yet to be obtained, or even concluded if a single mechanism is operant across all fission materials. In the direct mechanism case, the rate of fission will be fast if the matrix element <S 0 S 1 |H el | 1 T 1 T 1 > is large. Consideration of electrostatic interactions shows that spatial overlap of SF chromophore orbitals is critical to maximize <S 0 S 1 |H el | 1 T 1 T 1 >. In acenes this matrix element appears maximum with a slip-stacked arrangement and zero with perfectly eclipsed backbones. 1,2 The nature of <S 0 S 1 |H el | 1 T 1 T 1 > was evaluated for pentacene 40 and tetracene 41 and lead to opposite conclusions for the involvement of charge transfer states in the direct SF process. For pentacene, the authors believed weak mixing of the charge transfer states with the locally excited states excluded their necessity from the process. In tetracene, separating out the 42 relative contributions of the direct and mediated interactions to <S 0 S 1 |H el | 1 T 1 T 1 > lead to the conclusion that charge transfer states are important governors to SF. In the mediated case, charge transfer states coupling to S 1 and 1 (TT) can transfer population through a superexchange pathway without ever populating the CT directly. 42 Little evidence has yet to be brought forth for the mediated case. The best examples arises from 1,3- diphenylisobenzofuran dimers in solution. It was observed that triplet formation was negligent in non-polar mediated, however in polar DMSO triplet yields of about ~10% could be measured. 43,44 The triplet formation appeared to proceed after a low energy charge transfer state was populated after the initial excitation and was irreversible at low temperature. It has become apparent that the spatial arrangement of neighboring molecules must permit orbital overlap between the two species to promote SF. Further, inter-chromophore coupling must be strong enough to permit SF to occur on rapid timescales to out-compete other singlet deactivation pathways, but must not be too strong as to allow the formation of excimeric states that would be too low in energy to participate in SF and act as an alternate pathway to siphon singlet excitation density. Given the exogernicity of pentacene SF, excimer creation has been postulated 40,45 and simultaneously denied 46 as an intermediate step, further illuminating the elusive nature of the SF mechanism. In an attempt to ascertain requisite coupling strengths a density functional theory (DFT) study was carried out on tetracene and 1,3- diphenylisobenzofuran dimers connected in a host of geometries. The investigation was able to qualitatively approximate the SF rates witnessed for these systems from calculated interaction energies and coupling strengths. 47 43 2.1.5 Magnetic Field Effects In 1969 Johnson-Merrifield provided a density matrix treatment of triplet-triplet annihilation and SF to explain observed magnetic field effects on tetracene crystals’ prompt and delayed fluorescence. 11 Expansion of the treatment invoked the existence of a nine possible spin states produced by the pairing of two triplet excitons ( 1 TT, 3 TT, 5 TT). 36 Each pair state has equal probability of population at room temperature, therefore the rate of production of each pair state by triplet collisions is where ‘n’ is the triplet exciton concentration. 48,49 Further, when two triplets collide there are two different resulting possibilities: 1.) scattering, with a transition rate, 2.) annihilation, with transition rate, where is the amplitude of the singlet component of the lth state. Thus, the probability of annihilation from the lth pair is: (5) Taking the product of the total annihilation probability and the collision rate constant gives the total annihilation rate constant: (6) It is useful to analyze the total annihilation rate constant (γ) in two extremes: the first ( ) when all nine pair states contain equal singlet character and the second ( ) when only one state is the singlet. Making the observation that the must sum to unity, it is possible to derive: 44 (7) From Equation 7 it can be seen that and that the annihilation rate constant for reforming the singlet state is greater the more uniform the singlet character is over the nine triplet pair states. 48 Given the treatment above, it is pertinent to know the amount of singlet character in a given triplet pair state since increased singlet character increases the rate of S 1 reforming. Similarly, the probability that one of the nine triplet sublevels ( 1 TT, 3 TT, 5 TT) is occupied is proportional to the amount of 1 (TT) character it contains. The spin Hamiltonian ( ) contains the operator of Zeeman interaction with an outside magnetic field, and since the initial wave function of 1 (TT) is a coherent superposition of the nine triplet pair sublevels, the 1 TT amplitude in each of the nine sublevels is affected by the strength and orientation of an external magnetic field. Simply, an external magnetic field can modulate the amount of singlet character contained in each of the triplet pair sublevels and affects the rate of SF. 50 In the absence of a magnetic field dipole-dipole interaction between two electrons of the triplet lead to the zero field Hamiltonian with three eigenstates: 23,51 45 (8) where and denote the single electron spin functions. For two molecules with two triplets (four electrons), the nine possible triplet pair states are just the product states of the zero field eigenstates: (9) with the overall singlet pair state: (10) Equation 10 is the quantum mechanical explanation for how a singlet exciton can be mapped onto a triplet manifold without having to flip spin and lead to a rapid process like SF. 52 As the magnetic field is applied, the zero field states mix and additional states gain singlet character, resulting in an increased SF rate in accordance with Equation 7. Larger magnetic fields further mix the states until only one pure singlet and one pure quintet state remain, decreasing the total amount of states with singlet character from three in the zero-field case to two in the high field case. In accordance with Equation 7, the SF rate decreases in the high field limit. 46 Figure 2.3. Measured relative fluorescence intensity from a tetracene crystal as a function of applied external magnetic field. The data is normalized to the magnitude of fluorescence intensity gathered at zero field. At large applied fields the fluorescence intensity increases as less of the nine triplet pair states ( 1 TT) retain singlet character. Magnetic field-dependent fluorescence intensity is a common way to confirm active singlet fission. 11 The overall effect of an external magnetic field on the singlet fission process can be seen in (Figure 2.3) for a fluorescence curve of representative tetracene crystals. 11 The magnitude of the tetracene crystal fluorescence in the absence of a magnetic field is normalized to one. As the field is slowly applied the number of triple pair states with singlet character increases, resulting in an increase in SF rate, siphoning off fluorescence intensity from the S 1 . Further field increases start to decrease the amount of pair states with singlet character, retarding the rate of SF and allowing for more singlets to fluoresce. Measuring magnetic field effects on fluorescence intensity is one common way to probe for SF in condensed phases of organic chromophores. 23,24 In fact, magnetic field effects can be used to probe for fission in other circumstances. If there is an alternate deactivation pathway from S 1 that competes with SF, then the relative magnitude of triplet production from fission to the amount of S 1 deactivation through an alternate path is governed by a ratio of rates between the two processes. In an OPV properly constructed to advantageously exploit SF, triplets produced from fission charge separate to produce polarons that are harnessed in an external circuit. Application of an external magnetic field to such a device while illuminated will decrease the amount of triplets produced from fission and thus 47 diminish the amount of triplet excited states available to be harnessed as charge. This truth has a profound impact on the possibility to directly probe the role SF is plays in a functioning OPV: by comparing the amount of photocurrent the solar cell produces with no applied external magnetic field to the amount of photocurrent produced when the field is turned on, it is possible to determine whether the low energy triplets produced from SF positively or negatively impact the photocurrent of the cell. When compared to its zero-field value, if the magnitude of the photocurrent drops when the magnetic field is applied, then the multiexciton generation from SF leads to multiple collected polaron pairs from an single absorbed photon in that device. Put another way, application of a magnetic field slows the rate of triplet production in the device and if as a result, the magnitude of generated photocurrent decreases, then the triplets produced from SF must have been active participants in the photocurrent generation mechanism of the device. Oppositely, compared to its zero-field value, if the magnitude of the photocurrent increases when the magnetic field is applied, then the multiexciton generation from SF leads to multiple low energy triplet excited states that are not being successfully converted into collected charges by the OPV. Since application of a magnetic field slows the rate of triplet production in the device and enhances the relative singlet population, an increase in photocurrent under these conditions means that singlets are positive contributors to photocurrent and any triplets produced from SF are not producing photocurrent in the cell. By plotting the cell photocurrent as a function of applied magnetic field and noting whether the magnitude of the current deviates positively or negatively from its initial zero-field value, it is possible to quickly determine the behavior of SF in the OPV. Being able to test OPVs in the presence of a magnetic field offers great insight to the operant photo-electrical conversion mechanism of SF in these devices. Design and construction 48 of a custom testing station apparatus capable of carrying out these measurements is detailed in Chapter 5 of this document with results summarized in Chapter 6. 2.2 Singlet Fission in OPVs 2.2.1 Designing a Singlet Fission OPV The mechanism of SF offers a challenging and fascinating puzzle for modern theorists and spectroscopist as well as synthetic chemists engaged in the challenge of making new SF chromophores. While certainly interesting, SF could offer more than just an academic curiosity. The ability to turn one high energy photon into two low energy triplet excited states in an advantageous carrier multiplication scheme with minimal thermalization losses has recently been noted as a potential way to improve OPV performance and even surpass the SQ limit. 53 Calculations have revealed that effective employment of SF materials in solar cells could increase efficiencies close to 45%, well above the thermodynamic limit of 32% for single junction devices with no carrier multiplication. 54 While certainly impressive, these quoted high efficiencies hinge on the same detailed balance assumptions of the SQ limit such as complete photon absorption, the only losses being radiative, etc, with the added complication of assuming 200% triplet quantum yield from SF AND unity charge collection from the 200% triplets. Further, for these devices a second molecular chromophore with a complementary absorption must be introduced to the device to absorb the photons that fall within the spectral range of the large energy gap between the fission material’s S 1 and T 1 states. The complementary absorber must capture photons lower in energy than the SF material S 1 and as a result will most likely be a good FRET acceptor for energy transfer. If FRET transfer from the fission material’s S 1 to the red complementary absorber occurs then SF will be detrimentally circumvented. Further, charge transfer from the SF chromphore S 1 , reducing the molecular acceptor must be kinetically out- 49 competed by SF. Therefore, the OPV must be constructed in such a way to allow for efficient fission, without permitting energy transfer or charge transfer deactivation. 26 If accomplished, in this mode of device construction a low energy triplet state on either the SF material or the red absorber is the final molecular excited state to reduce the electron Acceptor at the D/A interface. Since this final triplet excited state is a low energy exciton, the achievable photovoltage is fundamentally limited in an OPV. If SF is meant to increase efficiencies, halving the possible photovoltage to boost current is not necessarily a clear advantage as overall cell efficiency is the product of current and voltage (η = FF x J sc x V oc ). It is clear that in order for SF to be a competent player in OPVs a detailed balance must be struck between the packing and coupling governing the efficient SF process, the D/A nanoscale morphologies promoting charge separation with minimal recombination losses, and the careful pairing of complimentary bandgaps and energy level offsets to ensure panchromatic photon absorption and exothermic charge transfer. While all these simultaneous criteria appear daunting, meeting them all concurrently is the only way to realize efficiencies that surpass the SQ limit. 2.2.2 Singlet Fission OPVs While the design criteria outlined above may seem daunting, SF has found traction in academic OPV research. This has mostly been accomplished using acenes as the SF material. Tetracene and pentacene absorb visible photons and produce 200% triplets in polycrystalline thin films. Chemical substitutions such as phenyl or triisopropylsilylethynyl groups can also be made to the acene backbone to create molecules such as rubrene 23,55 or TIPS-pentacene 26,56 with different packing motifs and thin film morphologies, affecting both SF production and OPV performance simultaneously. Literature results encompassing both tetracene and pentacene devices will be outlined here as a brief survey of successful implementation of SF in OPVs. 50 2.2.3 Tetracene Singlet Fission OPVs Figure 2.4. (Upper left) Cartoon illustrating the role of singlet fission in the OPV. Photon absorption by tetracene creates singlets that undergo fission, producing two triplets. The triplets diffuse to the CuPc interface where they energy transfer. Subsequent triplet migration with the CuPc layer presents CuPc triplets at the C60 Acceptor where they are subsequently separated into free charges. (Upper right) Architecture of the OPV, complete with oxidation and reduction potentials and singlet and triplet energies for the active layer materials of interest. (Bottom) EQE (left) and IV curves (right) showing tetracene photoresponse in devices both including and excluding the CuPc layer. Magnetic field modulated photocurrent measurements track with the tetracene absorption profile, confirming the role of SF in the device. 57 When paired with copper phthalocyanine (CuPc) in an OPV it was shown that triplets produced via SF in a tetracene neat layer (E T = 1.2 eV) could be sensitized onto CuPc (E T = 1.1 eV) and then separated into free charges at the CuPc/C 60 interface (Figure 2.4). 57 Complementary CuPc absorption acted well to cover the visible spectrum and capture photons too low in energy for the acene to absorb. In the device tetracene fissioned with 71 ± 18% efficiency (140% triplet production) at room temperature and 23 ± 7% at 175K owing to endothermic SF. Magnetic field 51 measurements show the cell photocurrent is modulated with applied field and only sensitive to tetracene excitation, confirming SF in the cell. Figure 2.5. (Top left) Diagram of the photo and electrical conversion processes operant in the TPTPA/rubrene/PDI- CN2 OPV. Singlet excitons on rubrene fission to produce two triplets which are then charge separated at the PDI- CN2 interface. TPTPA absorption is also sensitized onto rubrene through singlet energy transfer and allows for high energy photons not captured by the acene to be converted into triplet excited states. (Upper right) Architecture of the device, complete with oxidation, reduction and work function potentials and layer thicknesses. (Bottom left) EQEs showing a doubling of TPTPA photoresponse when the neat rubrene layer is included (red) versus when the rubrene is omitted (blue). (Bottom right) Magnetic field modulated photocurrent measurements prove the function of SF in the device. Both TPTPA (blue solid) and rubrene (green solid) photoresponses are magnetic field dependent when rubrene is included in the device. When rubrene is removed the device exhibits no magnetic field dependent electrical response (blue and green dashed). 58 In a separate experiment the high energy absorber tris[4-(5-phenylthiophen-2- yl)phenyl]amine (TPTPA, E S1 = 2.8eV) was used to sensitize singlets onto rubrene (E S1 = 2.3eV, E T1 = 1.2 eV) (Figure 2.5). 58 In this scheme singlet excited states were created on the rubrene fission material through singlet energy transfer sensitization instead of direct photoexcitation, allowing for fission to occur from photons not directly absorbed by the SF material. After fission the produced triplets were then separated into charge at a rubrene/PDI-CN2 interface 52 (PDI-CN 2 = dicyanoperylene diimide, Acceptor). With the rubrene layer the photocurrent of the cell increased 69% compared to a similarly constructed device that omitted the acene. External quantum efficiency measurements show the origin of the photocurrent enhancement to be a doubling of TPTPA response from fission. This was confirmed through magnetic field measurements that saw photocurrent decrease as a function of applied field when either the rubrene (λ = 500 nm) or TPTPA (λ = 365 nm) was preferentially excited. Once rubrene was removed the OPV did not have any magnetic field dependent electrical response. These two case studies highlight how proper OPV construction can lead to advantageous SF providing more than one collected charge per absorbed photon. Both devices use a neat layer of the acene material to maximize triplet production through fission, and a neat layer of the complementary absorber. This design spatially separates the two materials in the device and minimizes unwanted deactivation pathways from the acene S 1 before fission. For tetracene/CuPc this is critical as the two materials make a good FRET pair and only by spatially excluding the materials can SF be guaranteed over energy transfer populating the CuPc S 1 . For TPTPA/rubrene, energy transfer from the rubrene S 1 to the TPTPA S 1 is energetically prohibited and not a concern. However, energy transfer from TPTPA S 1 to rubrene S 1 is highly favorable and acts as a way to sensitize singlet excited states onto a fission material through a complementary high energy absorber. In both devices once triplets are formed they must be converted into free polarons by charge separating at the D/A interface. In the tetracene/CuPc device this is accomplished by Dexter-type energy transfer from the tetracene T 1 to the CuPc T 1 . CuPc triplets then diffuse to the CuPc/C 60 interface. In this device C 60 acts as an electron Acceptor and CuPc triplets are then separated into free charges and percolate back through their respective layers to be collected at electrodes. Here it is critical that the oxidation potential of 53 tetracene and CuPc allow for efficient charge migration and not present an energetic barrier for hole transfer between the materials. For TPTPA/rubrene, triplets residing on rubrene after fission diffuse to the rubrene/PDI-CN 2 interface where the triplet excited states are separated into free charges. The EQE of the device shows that being able to complement the rubrene absorption profile with the high energy absorbing TPTPA and capture photons that would not have otherwise been absorbed to fuel the SF process leads to a large photocurrent response in this spectral region. Finally, the oxidation potential of rubrene and TPTPA are suitable and ensure no hole trapping occurs between the materials. The above two devices simultaneously marry the complementary absorption profiles of multiple chromophores, ensure high yielding triplet production from SF, engineer exothermic electron transfer from the low energy triplet states at a suitable Acceptor interface, and permit charge collection with minimal thermodynamic barriers in a successful attempt to harness triplets from singlet fission in an OPV. 2.2.4 Pentacene Singlet Fission OPVs Ultrafast SF in pentacene neat layers has also been exploited in optoelectronic devices. Photodetectors incorporating thin (~5Å) alternating Donor-Acceptor layers of pentacene and C 60 exhibit exciton multiplication by 145 ± 7%. 59 Given the intimate mixing of pentacene/C 60 in this photodetector device architecture, the high fission yield of (145 ± 7%) is important to note as SF actively competes with other pentacene S 1 deactivation pathways such as charge transfer to C 60 . Even with this kinetically competing pathway, pentacene fission occurs rapid enough to produce a high yield of triplets. If the rate of charge transfer to C 60 is taken as , then the rate of triplet formation was calculated to be (0.8 ± 0.2%) . 54 Neat pentacene has also been employed as a fission material in a P3HT/pentacene/C 60 OPV (P3HT = poly(3-hexylthiophene-2,5-diyl) (Figure 2.6). 60 In this device the pentacene layer is only 150Å thick and absorbs ~49% of 670nm photons at this thickness. In order to enhance the SF process external mirrors were positioned around the solar cell to afford multiple passes of light through the cell and increase pentacene absorption. Further, P3HT was employed as a photoabsorber and singlet sensitizer onto pentacene. The polymer’s large triplet energy (E T1 = 1.5eV) also acts to confine triplet excitons in the pentacene layer to ensure triplet charge transfer at C 60 . It was found from optical modeling that the IQE (IQE = quantum efficiency of photons absorbed to charge carriers collected) for photoexcitation of pentacene and P3HT is 160 ± 10% and 150 ± 10% respectively. Remarkably, the peak EQE of the device at λ = 670 nm is 109 ± 1%. Since EQE is a measure of the number of charge carriers collected versus number of Figure 2.6. (Left) Device architecture including oxidation, reduction and work function potentials as well as layer thickness (in nm) for each component of the OPV. In the device pentacene singlets fission into two triplets which are then charge separated at the C 60 interface. P3HT acts to both block pentacene triplets from diffusing away from the Acceptor, and to sensitize singlet excitations on to the acene. (Right) EQE measurements for P3HT/pentacene/C 60 devices. Pentacene photoresponse at λ = 670 nm exceeds 100 % (red trace) when external mirrors are assembled to direct incident light through the cell multiple times to ensure maximum light absorption. Producing more charges in the external circuit than there were even photons of that wavelength incident on the device speaks to the potential SF could have in OPVs. 60 55 incident photons available, a 109 ± 1% efficiency at λ = 670 means the solar cell is producing more charges from 670nm photons than there are even 670nm photons incident on the device. Such a result is only possible from carrier multiplication schemes like singlet fission. It is quite remarkable that EQEs exceeding 100% can be realized with such thin layers of a highly efficient SF material such as pentacene. However, such an advance was only possible by configuring mirrors around the OPV to take advantage of multiple light passes through the cell. Another way to increase light absorption would be to increase the thickness of the pentacene layer beyond 150 Å. In fact for pentacene/C 60 heterojunctions SF fission efficiency increases with pentacene thickness; thicker pentacene layers allow for the fission process to occur unmolested by other competing process such as charge transfer from the pentacene singlet, a charge production avenue responsible for one-to-one photon-to-charge conversion. Therefore, it seems logical to increase the thickness of pentacene neat layers to both increase light absorption and SF quantum yield. However, as mentioned in Chapter 1 the materials in active layers of OPV must strike a delicate balance between maximizing multiple intrinsic properties, including light absorption, exciton diffusion, and charge migration. Optical modeling has suggested pentacene triplets to be diffusive on 400Å length scales, 61 however thick pentacene devices have been shown to suffer from parasitic triplet-charge annihilation as revealed from magnetic field photocurrent measurements. 62 The use of colloidal quantum dots in modern solar cells is another technology alongside OPVs that has received increased research interest in recent years. In fact, quantum dots have been shown to be competent Acceptors when paired with semiconducting polymer donors in organic/inorganic hybrid cells. The ability to tune quantum dot band gap through quantum 56 confinement offers a unique opportunity to tailor redox levels of the inorganic nanocrystal to best compliment the energy levels of an organic Donor molecule. To this end PbS semiconductor nanocrystals (E g = 0.7 eV) have been synthesized to act as an electron Acceptor for pentacene triplets produced from SF. 63 In such a device a neat layer of pentacene absorbs light, undergoes SF, and reduces PbS quantum dots from the excited state. The low lying PbS conduction band ensures electron transfer from pentacene triplets to be exothermic. The device further benefits from complimentary bandgap profiles with photons between the singlet of pentacene (E S1 = 1.83 eV) and greater than the PbS bandgap being able to be absorbed by the quantum dot. Characteristic pentacene vibronic progression appearing in the EQE confirms pentacene photoresponse but does not verify the pentacene contribution to be from triplets produced from fission. SF is instead inferred indirectly by noting that a control device using PbS quantum dots with a 1.3 eV bandgap does not exhibit pentacene response. In this architecture the higher lying conduction band of the 1.3 eV bandgap quantum dot is not low enough in energy to be reduced by pentacene triplets, but would be reduced by pentacene singlets. Since Förster energy transfer from pentacene singlets to the PbS nanocrystals is forbidden, a lack of pentacene photoresponse with 1.3 eV bandgap quantum dots implies pentacene undergoes rapid SF, resulting in too low of energy triplets to be useful. When the bandgap of PbS is lowered such that the pentacene triplet is sufficiently reducing to the PbS conduction band, pentacene photoresponse is observed because of triplet charge trasnfer. In a similar experiment a series of PDI derivatives, C 60 , and quantum dots of PbS and PbSe with different redox potentials where used as Acceptors when paired with pentacene and diphenylpentacene SF donor materials. 62 The aim was to scan a host of acceptor materials with varying LUMO/conduction band energies to see which materials would dissociate pentacene 57 triplets. The results were then correlated with the energy level difference between the donor HOMO and acceptor LUMO for each device as a pseudo metric for the energy of the charge transfer state formed upon electron transfer at the D/A interface ( 3 Pentacene + A  [Pentacene + /A - ]). It was found that materials combinations with Donor HOMO-Acceptor LUMO offsets of > 1.0 eV could not effectively harness pentacene triplets, correlating well with the pentacene triplet energy (E T = 0.86 eV). 2.3 References (1) Smith, M. B.; Michl, J. Singlet Fission Chem. Rev., 2010, 110, 6891-6939. (2) Smith, M. B.; Michl, J.; Recent Advances in Singlet Fission. Annu. Rev. Phys. Chem,. 2013, 64, 61-86. (3) Sanders, S. N.; Kumarasamy, E.; Pun, A. B.; Steigerwald, M. L.; Sfeir, M. Y.; Campos, L. M. Intramolecular Singlet Fission in Oligoacene Heterodimers. Angew. Chem. Int. Ed., 2016, 55, 3373-3377. (4) Sanders, S. N.; Kumarasamy, E.; Pun, A. B.; Appavoo, K.; Steigerwald, M. L.; Campos, L. M.; Sfeir, M. Y.; Exciton Correlations in Intramolecular Singlet Fission. JACS, 2016, 138, 7289-7297. (5) Sakuma, T.; Sakai, H.; Araki, Y.; Mori, T.; Wada, T.; Tkachenko, N. V.; Hasobe, T. Long-Lived Triplet Excited States of Bent-Shaped Pentacene Dimers by Intramolecular Singlet Fission. J. Phys. Chem. A, 2016, 120, 1867-1875. (6) Lukman, S.; Musser, A. J.; Chen, K.; Athanasopoulos, S.; Yong, C. K.; Zeng, Z.; Chi, C.; Hodgkiss, J. M.; Wu, J.; Friend, R. H.; Greenham, N. C. Tuneable Singlet Exciton Fission and Triplet-Triplet Annihilation in an Orthogonal Pentacene Dimer. Adv. Funct. Mater. 2015, 25, 5452-5461. (7) Korovina, N. V.; Das, S.; Nett, Z.; Feng, X.; Joy, J.; Haiges, R.; Krylov, A. I.; Bradforth, S. E.; Thompson, M. E. Singlet Fission in a Covalently Linked Cofacial Alkynyltetracene Dimer. JACS, 2016, 138, 617-627. (8) Sanders, S. N.; Kumarasamy, E.; Pun, A. B.; Trinh, M. T.; Choi, B.; Xia, J.; Taffet, E. J.; Low, J. Z.; Miller, J. R.; Roy, X.; Zhu, X.-Y.; Steigerwald, M. L.; Sfeir, M. Y.; Campos, 58 L. M., Quantitative Intramolecular Singlet Fission in Bipentacenes. JACS, 2015, 137, 8965-8972. (9) Singh, S.; Jones, W. J.; Siebrand, W.; Stoicheff, B. P.; Schneider, W. G. Laser Generation of Excitons and Fluorescence in Anthracene Crystals. J. Chem. Phys., 1965, 42, 330-342. (10) Geacintov, N.; Pope, M.; Vogel, F. E., Effect of Magnetic Field on the Fluorescence of Tetracene Crystals: Exciton Fission Phys. Rev. Lett., 1969, 22, 593-596. (11) Merrified, R. E.; Avakian, P.; Groff, R. P. Chem. Phys. Lett., 1969, 3, 386-388. (12) Swenberg C.E.; Stacy W. T. Bimolecular Radiationless Transitions in Crystalline Tetracene. Chem. Phys. Lett., 1968, 2, 327–328. (13) Chabr, M.; Williams, D. F.; Singlet Exciton Trapping and Heterofission in Tetracene Doped Anthracene Crystals. Chem. Phys. Lett., 1977, 49, 599-603. (14) Burg, K.; Granacher, Z.; Singlet Exciton Fission in Pure and Doped Anthracene, J. Chem. Phys., 1979, 70, 3807-3811. (15) Bayliss, S. L.; Chepelianskii, A. D.; Sepe, A.; Walker, B. J.; Ehrler, B.; Bruzek, M. J.; Anthony, J. E.; Greenham, N. C. Geminate and Nongeminate Recombination of Triplet Excitons Formed by Singlet Fission. Phys. Rev. Lett., 2014, 112, 238701. (16) Fleming, G. R.; Millar, D.; Morris, G. C.; Morris, J. M.; Robinson, G. W. Exciton Fission and Annihilation in Crystalline Tetracene. Aust. J. Chem., 1977, 30, 2353-2359. (17) Pope, M. Geacintov, N.; Vogel, F. Singlet Exciton Fission and Triplet-Triplet Exciton Yield in Fission and Fusion in Tetracene. Mol. Cryst. Liquid Cryst., 1969, 6, 83-104. (18) Staroske, W.; Pfeiffer, M.; Leo, K.; Hoffmann, M. Single-Step Triplet-Trilet Annihilation: An Intrinsic Limit for the High Brightness Efficiency of Phosphorescent Organic Light Emitting Diodes. Phys. Rev. Lett., 2007, 98, 197402. (19) Lee, J.; Bruzek, M. J.; Thompson, N. J.; Sfeir, M. Y.; Anthony, J. E.; Baldo, M. A. Singlet Exciton Fission in a Hexacene Derivative. Adv. Mater., 2013, 25, 1445-1448. (20) Busby, E.; Berkelbach, T. C.; Kumar, B.; Chernikov, A.; Zhong, Y.; Hlaing, H.; Zhu, X.- Y.; Heinz, T. F.; Hybertsen, M. S.; Sfeir, M. Y.; Reichman, D. R.; Nuckolls, C.; Yaffe, O. Multiphonon Relaxation Slows Singlet Fission in Crystalline Hexacene. JACS, 2014, 136, 10654-10660. 59 (21) Burdett J.J.; Muller, A.M.; Gosztola D.; Bardeen C. J. 2010. Excited State Dynamics in Solid and Monomeric Tetracene: The Roles of Superradiance and Exciton Fission. J. Chem. Phys., 2010, 133, 144506. (22) Burdett. J. J.; Goxztola, D.; Bardeen, C. J. The Dependence of Singlet Exciton Relaxation on Excitation Density and Temperature in Polycrystalline Tetracene Thin Films: Kinetic Evidence for a Dark Intermediate State and Implications for Singlet Fission. J. Chem. Phys. 2011, 135, 214508. (23) Piland, G. B.; Burdett, J. J.; Kurunthu, D.; Bardeen, C. J., Magnetic Field Effects on the Singlet Fission and Fluorescence Decay Dynamics in Amorphous Rubrene. J. Phys. Chem. C. 2013, 117, 1224-1236. (24) Thompson, N. J.; Hontz, E.; Chang, W.; Van Voorhis, T. ; Baldo, M. A.; Magnetic Field Dependence of Singlet Fission in Solutions of Diphenyl Tetracene. Phil. Trans. R. Soc. A., 2015, 373, 20140323. (25) Roberts, S. T.; McAnally, R. E.; Mastron, J. N.; Webber, D. H.; Whited, M. T.; Brutchey, R. L.; Thompson, M. E.; Bradforth, S. E. Efficient Singlet Fission Discovered in a Disordered Acene Film. JACS, 2012, 134, 6388-6400. (26) Ramanan, C.; Smeigh, A. L.; Anthony, J. E.; Marks, T. J.; Wasielewski, M. R. Competition Between Singlet Fission and Charge Separation in Solution-Processed Blend Films of 6,13-Bis(triisopropylsilylethynyl)-pentacene with Sterically-Encumbered Perylene-3,4:9,10-bis(dicarboximide)s. JACS, 2012, 134, 386-397. (27) Yost, S. R.; Lee, J.; Wilson, M. W. B.; Wu, T.; McMahon, D. P.; Parkhurst, R. R.; Thompson, N. J.; Congreve, D. N.; Rao, A.; Johnson, K.; Sfeir, M. Y.; Bawendi, M. G.; Swager, T. M.; Friend, R. H.; Baldo, M. A.; Van Voorhis, T. A Transferable Model for Singlet-Fission Kinetics. Nat. Chem., 2014, 6, 492-497. (28) Renaud, N.; Sherratt, P. A.; Ratner, M A. Mapping the Relation Between Stacking Geometries and Singlet Fission Yield in a Class of Organic Crystals. J. Phys. Chem. Lett., 2013, 4, 1065-1069. (29) Dillion, R. J.; Piland, G. B.; Bardeen, C. J. Different Rates of Singlet Fission in Monoclininc Versus Orthorhombic Crystal Forms of Diphenylhexariene. JACS, 2013, 135, 17278-17281. (30) Ryerson, J. L.; Schrauben, J. N.; Ferguson, A. J.; Sahoo, S. C.; Naumov, P.; Havlas, Zdenek, Michl, J.; Nozik, A. J.; Johnson, J. C. Two Thin Film Polymorphs of the Singlet Fission Compound 1,3-Diphenylisobenzofuran. J. Phys. Chem. C., 2014, 118, 12121- 12132. 60 (31) Johnson J. C.; Nozik A. J.; Michl J. High Triplet Yield from Singlet Fission in a Thin Film of 1,3-Diphenylisobenzofuran. J. Am. Chem. Soc., 2010, 132, 16302–16303. (32) Wang, C.; Tauber, M. J.; High-Yield Singlet Fission in a Zeaxanthin Aggregate Observed by Picoseconds Resonance Raman Spectroscopy. JACS, 2010, 132, 13988–13991. (33) Musser, A. J.; Maiuri, M.; Brida, D.; Cerullo, G.; Friend, R. H. The Nature of Singlet Exciton Fission in Carotenoid Aggregates. JACS, 2015, 137, 5130-5139. (34) Varnavaski, O. Abeyasinghe, N.; Arago, J.; Serrano-Perez, J. J.; Orti, E.; Navarrete, J. T. L.; Takimiya, K.; Casanova, D.; Casado, J.; Goodson III, T. High Yield Ultrafast Intramolecular Singlet Exciton Fission in a Quinodial Bithiophene. J. Phys. Chem. Lett. 2015, 6, 1375-1384. (35) Busby, E.; Xia, J.; Wu, Q.; Low, J. Z.; Song, R.; Miller, J. R.; Zhu, X.-Y.; Campos, L. M.; Sfeir, M. Y., A Design Strategy for Intramolecular Singlet Fission Mediated by Charge-Transfer States in Donor-Acceptor Organic Materials. Nat. Mater,. 2015, 14, 426-433. (36) Suna A. Kinematics of Exciton-Exciton Annihilation in Molecular Crystals. Phys. Rev. B, 1970, 1, 1716-1739. (37) Chan, W.-L.; Ligges, M.; Jailaubekov, A.; Kaake, L.; Miaja-Avila, L.; Zhu, X.-Y. Observing the Multiexciton State in Singlet Fission and Ensuing Ultrafast Multielectron Transfer. Science, 2011, 334, 1541-1545. (38) Chan, W.-L.; Ligges, M.; Zhu, X.-Y.; The Energy Barrier in Singlet Fission can be Overcome Through Coherent Coupling and Entropic Gain. Nat. Chem, 2012, 4, 840-845. (39) Bakulin, A. A.; Morgan, S. E.; Kehoe, T. B.; Wilson, M. W. B.; Chin, A. W.; Zigmantas, D.; Egorova, D. Rao, A. Real-Time Observation of Multiexcitonic States in Ultrafast Singlet Fission Using Coherent 2D Electronic Spectroscopy. Nat. Chem., 2016, 8, 16-23. (40) Zimmerman, P. M.; Bell, F.; Casanova, D.; Head-Gordon, M. Mechanism for Singlet Fission in Pentacene and Tetracene: From Single Exciton to Two Triplets. JACS, 2011, 133, 19944-19952. (41) Havenith, R. W. A.; de Gier, H. K.; Broer, R. Explorative Computational Study of the Singlet Fission Process. Mol. Phys., 2012, 110, 2445-2454. (42) Berkelbach, T. C.; Hybertsen, M. S.; Riechman, D. R. Microscopic Theory of Singlet Exciton Fission. II. Application to Pentacene Dimers and the Role of Superexchange. J. Chem. Phys., 2013, 138, 114103. 61 (43) Johnson, J. C.; Nozik, A. J.; Michl, J. The Role of Chromophore Coupling in Singlet Fission. Acc. Chem. Res., 2013, 46, 1290-1299. (44) Johnson, J. C.; Akdag, A.; Zamadar, M.; Chen, X.; Schwerin, A. F.; Paci, I.; Smith, M. B.; Havlas, Z.; Miller, J. R.; Ratner, M. A.; Nozik, A. J.; Michl, J. Toward Designed Singlet Fission: Solution Photophysics of Two Indirectly Coupled Covalent Dimers of 1,3-Diphenylisobenzofuran. J. Phys. Chem. B, 2013, 117, 4680-4695. (45) Marciniak, H.; Pugliesi, I.; Nickel, B.; Lochbrunner, S. Ultrafast Singlet and Triplet Dynamics in Microcrystalline Pentacene Films. Phys. Rev. B, 2009, 79, 235318. (46) Kuhlman, T. S.; Kongsted, J.; Mikkelsen, K. V.; Moller, K. B.; Solling, T. I. Interpretation of the Ultrafast Photoinduced Processes in Pentacene Thin Films. JACS, 2010, 132, 3431-3439. (47) Greyson, E. C; Stepp, B. R.; Chen, X.; Schwerin, A. F.; Paci, I.; Smith, M. B.; Akdag, A.; Johnson, J. C.; Nozik, A. J.; Michl, J.; Ratner, M. A. Singlet Exciton Fission for Solar Cell Applications: Energy Aspects of Interchromophore Coupling. J. Phys. Chem. B., 2010, 114, 14223-142232. (48) Merrifield, R. E.; Theory of Magnetic Field Effects on the Mutual Annihilation of Triplet Excitons. J. Chem. Phys., 1968, 48, 4318-4319. (49) Johnson, R. C.; Merrifield, R. E Effects of Magnetic Fields on the Mutual Annihilation of Triplet Excitons in Anthracene Crystals. Phys. Rev. B., 1970, 1, 896-902. (50) Merrifield, R. E. Magnetic Effects on Triplet Exciton Interactions. Pure Appl. Chem., 1971, 27, 481-498. (51) Burdett, J. J.; Bardeen, C. J.; Quantum Beats in Crystalline Tetracene Delayed Fluorescence Due to Triplet Pair Coherences Produced by Direct Singlet Fission. JACS, 2012, 134, 8597-8607. (52) Burdett, J. J.; Piland, G. B.; Bardeen, C. J. Magnetic Field Effects and the Role of Spin States in Singlet Fission. Chem. Phys. Lett. 2013, 585, 1-10. (53) Shockley, W.; Queisser, H. J. Detailed Balance Limit of Efficiency of p-n Junction Solar Cells. J. Appl. Phys. 1961, 32, 510-519. (54) Hanna, M. C.; Nozik, A. J.; Solar Conversion Efficiency of Photovoltaic and Photoelectrolysis Cells with Carrier Multiplication Absorbers. J. Appl. Phys., 2006, 100, 074510:1-8. (55) Pandey, A. K. Highly Efficient Spin-Conversion Effect Leading to Energy Up-Converted Electroluminescence in Singlet Fisison Photovoltaics. Sci. Rep., 2015, 5, 1-6. 62 (56) Yang, L.; Tabachnyk, M.; Bayliss, S. L.; Bohm, M. L.; Broch, K.; Greenham, N. C.; Friend, R. H.; Ehrler, Br.; Solution-Processable Singlet Fission Photovoltaic Devices. Nano Lett. 2015, 15, 354-358. (57) Jadhav, P. J.; Mohanty, A.; Sussman, J.; Lee, J.; Baldo, M. A. Singlet Exciton Fission in Nanostructured Organic Solar Cells. Nano Lett., 2011, 11, 1495-1498. (58) Reusswig, P. D.; Congreve, D. N.; Thompson, N. J.; Baldo, M. A. Enhanced External Quantum Efficiency in an Organic Photovoltaic Cell via Singlet Fission Exciton Sensitizer. Appl. Phys. Lett., 2012, 101, 113304. (59) Lee, J.; Jadhav, P.; Baldo, M. A. High Efficiency Organic Multilayer Photodetectors Based on Singlet Exciton Fission. Appl. Phys. Lett., 2009, 95, 033301. (60) Congreve, D. N.; Lee, J.; Thompson, N. J.; Hontz, E.; Yost, S. R.; Reusswig, P. D.; Bahlke, M. E.; Reineke, S.; Van Voorhis, T.; Baldo, M. A. Baldo External Quantum Efficiency Above 100% in a Singlet-Exciton-Fission-Based Organic Photovoltaic Cell. Science, 2013, 340, 334-337. (61) Tabachnyk, M.; Ehrler, B.; Bayliss, S.; Friend, R. H.; Greenham, N. C. Triplet Diffusion in Singlet Exciton Fission Sensitized Pentacene Solar Cells. Appl. Phys. Lett., 2013, 103, 153302. (62) Thompson, N. J.; Hontz, E.; Congreve, D. N. Bahlke, M. e.; Reineke, S.; Voorhis, T. V.; Baldo, M. A.; Nanostructured Singlet Fission Photovoltaics Subject to Triplet-Charge Annihilation. Adv. Mater., 2014, 26, 1366-1371. (63) Ehrler, B.; Wilson, M. W. B.; Rao, A.; Friend, R. H.; Greenham, N. C. Singlet Exciton Fission-Sensitized Infrared Quantum Dot Solar Cells. Nano Lett. 2012, 12, 1053-1057. 63 Chapter 3. Photophysical Investigation of Singlet Fission in a Disordered Acene Film* *Published in Journal of the American Chemical Society, 2012, 134, 6388-6400. 3.1 Singlet Fission in thin films of 5-12 Diphenyltetracene The mutual chromophore geometry and state couplings necessary to promote efficient singlet fission is one of the most contested, and highly investigated, aspects of fission research. Because of this, being able to probe fission in environments with well defined molecular packing is extremely powerful as it allows a one-to-one structure to property relationship between the molecular arrangement and photophysical process. By tethering two SF chromophores together with a covalently-bound molecular scaffold it is possible to design a dimer system that constrains the molecules to a particular mutual geometry. Research results garnered from interrogating such dimers allows for conclusions to be made toward the effect that molecular geometry has on the SF process. Aside from dimers, single crystals of molecular species also present a unique opportunity to explore fission in a well-ordered environment. X-ray crystallography can elucidate the structure of a molecular crystal and subsequent photophyiscal studies reveals how singlet fission behaves in a well-ordered infinite lattice of these chromophores. Numerous such studies have been performed on single crystals of the archetypical SF chromophores tetracene and pentacene. It has been demonstrated that infinite lattices of herringbone-patterened tetracene and pentacene single crystals fission with unity yield on ultrafast timescales. The question of how imperative the herringbone-packing motif is for efficient and fast fission in acene systems is provocative. It is possible to alter molecular packing by growing crystals under different crystallization conditions. Chemical substitutions along the acene backbone can also be added to force different crystallization polymorphs. For tetracene, tetraphenyl substitution along the backbone affords rubrene. For pentacene, triiisopropylsilylethynyl substitutions create TIPS- 64 pentacene, a well-studied molecule ubiquitous in the organic electronics literature. Even with these substitutions to the parent compound, both rubrene and TIPS-pentacene have been shown to undergo singlet fission. Investigating the dynamics of SF in well-ordered molecular dimers or single crystal condensed media in order to make predictive and testable claims on the role chromophore geometry has on the photophysical process is a compelling and logical motivation. However, one aspect that is somewhat disregarded is how prominent SF can be in highly disordered media and how pliable are the restrictions to efficient SF. In fact, if SF is to ever be a major player in OPV devices, it is reasonable to demand fission occur quickly and efficiently in disordered systems since deposition conditions used in the fabrication of such devices affords minimal control over thin film morphology and often results in highly amorphous layers. This chapter explores the photophysics of another substituted acene, 5,12-diphenyltetracene (DPT). Specifically, the morphology of vapor-deposited DPT films is discovered to be amorphous, yet fission is measured to proceed despite the disordered media. The results are compared to similar experiments carried out on polycrystalline thin films and single crystals of tetracene to highlight how morphology impacts SF compared to the herringbone motif for the parent tetracene analogue. 65 3.2 Morphology of DPT and Tetracene Vapor Deposited Thin Films Figure 3.1. (Left) X-ray diffraction patterns for vapor deposited tetracene and DPT films. Tetracene films exhibit diffraction peaks corresponding to the (001) and (002) planes while DPT shows no diffraction peaks. (Right) Electron diffraction images measured for tetracene and DPT. Tetracene thin films diffract the electron beam while DPT films do not, suggesting an amorphous morphology for DPT vapor deposited films. The morphology of DPT vapor deposited films were characterized by X-ray and transmission electron microscopy measurements and compared to similar data collected for vapor deposited tetracene thin films (Figure 3.1). XRD data shows two strong diffraction peaks at 2θ values of 7.1º and 14.5º for tetracene films that correspond to tetracene’s 001 and 002 diffraction planes respectively. 1 Conversely, no discernible X-ray diffraction peaks are observed in DPT thin films, suggesting the film is structurally amorphous. These findings are further confirmed by TEM-SAED measurements. Images recorded for tetracene show strong diffraction spots while DPT films exhibit no discernible electron beam diffraction. These results taken together confirm the disordered morphology of amorphous DPT thin films and the crystalline nature of tetracene films. 66 3.3 Solution and Condensed Phase Steady-State Absorption and Emission Figure 3.2. Absorption and emission spectra of tetracene (left) and DPT (right) chloroform solutions and vapor deposited thin films. Thin film absorption and emission lineshapes for tetracene samples are red-shifted from solution measurements due to pronounced Davydov coupling between neighboring chromophores. DPT film absorption and emission lineshape does not deviate drastically from solution measurements, suggesting minimal DPT inter-chromophore coupling. Absorption measurements are conducted on thick films (100 nm) to maximize signal intensity while emission measurements are conducted on thin films (10 nm) to minimize reabsorption effects. . The optical properties of DPT thin films were characterized in absorption and emission experiments and compared to tetracene (Figure 3.2). When dissolved in CHCl 3 , tetracene solutions show the presence of a pronounced vibrational progression characteristic of isolated polyacenes. 2,3 When vacuum deposited, tetracene crystallite formation 4,5 in the thin film permits intermolecular electronic coupling between neighboring tetracene molecules, resulting in a pronounced red-shift in the film absorption spectrum due to Davydov coupling. 6,7 This coupling is also present in the tetracene thin film emission spectrum which also shifts to lower energy emission wavelengths due to molecular aggregation. The extent of exciton delocalization has been estimated on the basis of line shape changes between thin film and solution emission spectra. For tetracene, the change in fluorescence lineshapes between solution and thin film suggest singlet excitons are delocalized over ~10 molecules in aggregates. 8,9 67 CHCl 3 solutions of DPT show the same vibronic progression as seen for tetracene solutions but with peak values slightly shifted to lower energy by 810 cm -1 (19 nm) and 917 cm -1 (21 nm) for absorption and emission, respectively. These small spectroscopic shifts suggest the phenyl substitutions minimally perturb the electronic structure of the DPT acene backbone. However, unlike tetracene, DPT thin films have almost identical absorption spectral lineshapes between solution and thin film and exhibit no evidence of Davydov splitting. Similarly, DPT thin film emission spectra are near-identical to solution. The lack of change in absorption and emission spectra from solution to thin films suggests that DPT ground state and excited states are relatively uncoupled in condensed media, with wavefunctions likely localized on a single molecule. Whether it is the amorphous nature of DPT thin films or the increased spacing between DPT π-backbones afforded by the pendant phenyl groups causing of the lack of coupling is unclear. 3.4 Time-resolved Emission and Ultrafast Transient Absorption Time-resolved fluorescence emission experiments were used to probe for the presence of delayed fluorescence in DPT thin films. During SF in acene crystals 10-12 the ultrafast fission process depletes the singlet population, leading to accelerated decay of crystal fluorescence over short timescales. However, if the triplet energy is approximately half the singlet energy then the two produced triplets can recombine over long time scales and reform the singlet, leading to fluorescence that can last orders of magnitude longer than the fluorescence lifetime of the isolated molecule. 68 Figure 3.3. (Left) TCSPC measurements of DPT solutions (blue) and vapor deposited thin films (red). The instrument’s response function is plotted in black. Decay in DPT solutions occurs fits to a single exponential decay with a characteristic lifetime of 11 ns (green dash) while thin film decay is nonexponential. (Right) The same decay plotted on a log-log scale highlighting the long emission tail in DPT thin films fitted to a t -1.8 power law (green dash). For DPT solutions the acene fluoresces with a characteristic lifetime of 11 ns (Figure 3.3). For DPT thin films emission is largely quenched and highly non-exponential, with 25% of the initial decay occurring within the instrument response time (20ps) and 99% of the total decay occurring within 3 ns. Even with the fast initial decay, a long-lived fluorescence tail persists for over 150ns in the emission trace, over an order of magnitude longer that the radiative lifetime of DPT in solution. Persistent emission over such time scales is attributed to delayed fluorescence. The emission tail can be fit with a power law decay with an exponent of -1.8 and is highly suggestive of triplet-triplet annihilation, which has been shown to cause emission decay as t -2 in rubrene and tetracene crystalline films at long times (> 300 ns). 13 69 Figure 3.4. (Upper left) Transient absorption spectra of DPT CHCl 3 solutions following excitation at 500 nm. The induced absorption at 416 nm corresponds to a S 1  Sn transition and verifies the existence of the DPT S 1 state. Over the 1 ns experimental window excited DPT molecules in solution remain in the S 1 state. (Upper right) Transient absorption spectra of DPT thin films. After excitation the induced absorption at 416 nm indicative of S 1 population rapidly decreases over ~100 ps. Concomitantly, a new induced absorption feature not witnessed in solution appears at ~510 nm. (Bottom) Comparison of the transient spectrum for vapor deposited DPT films at ∆t = 750 ps (black dashed) with the transient line shape expected for DPT triplet induced absorption (red). To further verify the existence of SF in DPT thin films, transient absorption (TA) experiments were used to characterize DPT excited state dynamics over femtosecond time scales (Figure 3.4). As a control, TA experiments were first conducted on DPT CHCl 3 solutions. DPT excitation at 500 nm populates the singlet excited state, reflected in ground state photobleaches at 460 and 498 nm corresponding to the vibronic progression present in DPT’s absorption spectrum. A third photobleach feature at 540 nm is due to stimulated emission of the S 1 . At 416 nm a pronounced induced absorption feature appears immediately following photoexcitation and is consequently assigned to a S 1  S n transition. Over the course of 1 ns, the transient signal exhibits no spectral evolution, except for a slight decrease in magnitude consistent with the 11 ns S 1 lifetime measured in TCSPC experiments. 70 TA measurement carried out on vapor deposited DPT thin films show drastically different behavior than isolated DPT in solution. Following excitation, the transient signal resembles DPT in CHCl 3 . Then, over ~100 ps the strong initial transition S 1  S n transition at 421 nm decays, suggesting population transfer out of the S 1 . Concurrent with the change, a new induced absorption band appears between 470 and 520 nm. From literature it is known that tetracene triplets absorb in the range of 400-500nm and suggests the same origin for the similar signal witnessed in DPT thin films. 14-16 To verify this hypothesis, the transient signal at a time delay of 750 ps was compared to an experiment that used a triplet sensitizer, platinum tetraphenylbenzoporphyrin (PtTPBP), doped into DPT to selectively populate the DPT T 1 . 17 The transient signal at 490nm of neat DPT and Pt(TPBP) doped DPT thin films match well, suggesting that triplet excitons are indeed produced in DPT following photoexcitation. The subnanosecond rate of triplet formation is much faster than the rate expected from triplet produced via intersystem crossing (k ISC = 89 ns). 15 Fast triplet production and characteristic delayed fluorescence confirm amorphous DPT films undergo SF. 3.5 Kinetic Model of Singlet Fission in DPT Thin Films In order to extract the kinetics of singlet fission in amorphous DPT thin films, TA and TCSPC measurements must be fit with a competent kinetic model. This task is complicated by singlet-singlet annihilation (exacerbated a high excitation pump fluencies) and triplet-triplet annihilation affecting both the singlet exciton populations and the spectral overlap between singlet and triplet induced absorptions. 71 Figure 3.5. (Left) Differential extinction spectra highlighting the changes in absorption profile for DPT vapor deposited thin films resulting from excitation of the DPT lowest energy singlet for triplet state. (Right) Comparison of transient spectra for DPT vapor deposited thin films (black dash) with a fit produced from a linear combination of the singlet and triplet spectra at various time delays. To compensate, the time dependence of the singlet and triplet populations were extracted by fitting DPT transient spectra as a linear combination of spectra representative of S 1 and T 1 differential lineshapes (Figure 3.5). The representative S 1 and T 1 differential lineshapes were arrived at experimentally from TA measurements independent of amorphous DPT films. Since the vapor deposited DPT TA spectra closely resembles that of DPT in CHCl 3 at short delays, the time-integrated transient spectrum of solution DPT was chose to represent the S 1 DPT basis spectrum. Comparison of the magnitude of the ground-state bleach of this spectrum with published extinction spectra of DPT’s ground state 15 allows for scaling of the amplitude of the basis spectrum so that it represents the change in the molar absorptivity of DPT following excitation to S 1 . Further, published triplet extinction spectrum measured in nanosecond TA experiments of DPT in solution 15 has a line shape almost identical to that measured in experiments on PtTPBP-doped DPT films (Chapter 4). 17 Thus, the triplet spectrum of PtTPBP- doped DPT scaled in amplitude to the literature extinction spectra was chosen as the basis spectrum of DPT’s T 1 state for fitting purposes. 72 At a fluence of 7.5 μJ/cm 2 a linear combination of the S 1 and T 1 basis spectra just outlined reproduces the TA spectral lineshape excellently over a host of pump-probe delays (Figure 3.5, right). The fits also allow for extraction of the time dependence of S 1 and T 1 populations as can be seen in Figure 3.6. Figure 3.6. (Left) Singlet (blue squares) and triplet (red circles) population densities in DPT vapor deposited thin films extracted from transient spectra. The green line denotes a biexponential function and is included as a guide to the eye. (Middle) Comparison of the triplet population extracted from transient spectra (red) to the TCSPC emission measured for DPT thin films (blue). (Right) Ratio of the triplet excitons observed at a delay of 0.9 ns to the initial singlet population prepared in the pump excitation pulse as a function of pump fluence. Extrapolation to an annihilation-free value (zero excitation fluence) gives an estimation of 1.22 triplets per singlet exciton (61% SF yield). Error bars are based on one standard deviation in the values of the singlet and triplet populations extracted from the transient absorption data. Interestingly, at 1 ns the triplet population is approximately 1.2x the initial S 1 population prepared in the pump pulse (Figure 3.6, left). The creation of more triplet states than initial singlet states prepared further confirms the presence of singlet fission in amorphous DPT films. Aside from a slight discrepancy at short times due to convolution of the measured signal with the TCSPC instrument response function (20 ps), singlet population determined from fitting the TA data matches well to the TCSPC data set. The independent confirmation of S 1 populations across two disparate measurement methods confirms the measured triplet yield is not an artifact of the chosen fitting method. With the fitting procedure in place the SF yield in amorphous DPT films can now be estimated by comparing the number of triplet excitons present at long time delays to the initial 73 singlet population. By extracting this ratio over a range of pump excitation fluences contributions to the SF yield estimate from processes such as singlet-singlet annihilation can be removed. At a 900 ps delay, the ratio of triplet population density to initial singlet density is quite high across all investigated pump fluences, ranging from 0.82 at high pump fluences to 1.25 at low pump fluences (Figure 3.6, right). Ratios of 0.82 and 1.25 for triplet to singlet populations mean 41% and 64.5% SF yields respectively. Given the amorphous nature of DPT thin films these large singlet fission yields are quite impressive and suggests that highly ordered phases present in single crystal and poly-crystalline samples is not necessary for efficient SF. Other than the triplet yield, another interesting aspect of (Figure 3.6, left) is the fact that triplet production appears to occur with two distinct time constants, with ~50% of the triplet population evolving within 3 ps and the remaining over ~100 ps. As a guide for the eye, a biexponential fit with time constants 1.3 and 105 ps is artificially superimposed over the data. In polycrystalline acene films SF typically occurs with a single primary rate. 14,18,19 Since the rate of SF is expected to be highly dependent on chromophore orientation and coupling, in a highly disordered media such as amorphous DPT thin films it seems highly probable that certain DPT neighboring dimers will be orientated appropriately to encourage fast SF, while other neighboring chromophores will adopt orientations highly disfavoring an efficient SF process. In order to explain triplet production over two time scales a kinetic model demanding exciton diffusion was constructed (Figure 3.7). At the core of the model is the assumption that due to structural disorder, only a subset of DPT molecular pairs will adopt geometries favorable for undergoing SF. After excitation, the singlet excitons created proximate to these favorable molecular pairs can fission rapidly, leading to an initial exponential increase in triplet population. Meanwhile, singlets created at excitation sites with DPT molecular pairs not in a favorable 74 geometry to undergo rapid SF diffuse through a Förster energy transfer mechanism among other DPT molecules. Once this diffusive singlet finds a DPT molecular pair of the appropriate geometry for fission, a second growth of diffusion-limited triplet production will occur. Figure 3.7. Diagram of the energy transfer and relaxation processes used to model the kinetics of singlet fission in amorphous DPT thin films. Photon absorption (blue arrows) populates DPT singlet sites. Energy transfer between DPT molecules (k D (t)) and relaxation to the ground state (k fl ) or to the triplet manifold via SF (k sf ) redistributes singlet population density between singlet sites (green arrows). Annihilation events act to depopulate diffusive singlet sites (k SS ) and the triplet manifold (k TS , k TT ) (orange arrows). As long as the DPT film is on average homogeneous, the rate at which diffusive singlets encounter fission sites is governed by the Smoluchowski theory of diffusion-limited reactions 20,21 (1) In the above equation D stands for the singlet exciton diffusion constant, c sf is the concentration of SF sites in the film, and R denotes an encounter distance related to the maximum distance a singlet exciton can be from a fission site and still undergo SF. Given this, a set of coupled rate equations for the interchange of singlet and triplet populations within amorphous DPT thin films can be written: 75 (2) These equations parse the initial photoexcited singlet populations in two categories: those that occupy fission sites (S sf ) and those that do not (S D ). S sf can decay via singlet fission (k sf ), or through other deactivation pathways such as fluorescence, internal conversion, and intersystem crossing. These latter processes are accounted for by the sum of their rates, k fl . Intersystem crossing in isolated DPT molecules proceeds quite slowly (1/k ISC = 89 ns), 15 orders of magnitude slower than triplets produced from fission in DPT thin films and is therefore neglected as a triplet production channel. S f can be regenerated by the diffusion of S D to fission sites as dictated by the time dependent rate k D (t). S D can also decay via the pathways encompassed in k fl . The starting population of S sf is given by δS 0 where S 0 is the total singlet population prepared in the excitation pulse and ‘δ’ is a parameter that varies between 0 and 1 to appropriately fit the data. The initial population of S D is therefore (1 - δ)S 0 . Exciton annihilation events affecting singlet and triplet populations are also accounted for in the model. Decay of the singlet population to ground state through singlet-singlet annihilation proceeds with rate constant k SS . Triplet-triplet annihilation, on the other hand, can act to either repopulate singlet states (with rate constant k TS ) or not (with rate constant k TT ) and annihilate to the ground state. For triplet-triplet annihilation events that regenerate singlets, 10 it is assumed that the annihilation event occurs between uncorrelated triplet pairs and therefore only S D is restored, not S sf . 76 Equation 2 was integrated numerically to yield the singlet and triplet populations as a function of time. The populations were used in conjunction with extinction spectra to reconstruct the TA data and a least-squares minimization routine was used to optimize the values of the rate coefficients as well as the ratio between S sf and S D . k D (t) contains three unknowns, c sf , R, and D, but only two of these are independent. For the purposes of fitting k D (t) was parameterized as k D (t) = a + bt -1/2 , with ‘a’ and ‘b’ as free parameters. This fit was performed simultaneously for excitation fluences ranging from 3.7 to 96 μJ/cm 2 with only the initial singlet population density allowed to vary in order to account for exciton annihilation. The best fit parameters can be summarized in Table 3.1. Figure 3.8. (Left) Singlet populations extracted from transient absorption experiments over a host of excitation pump fluences compared to the singlet population calculated based on the kinetic model (blue line). (Right) Triplet populations extracted from transient absorption experiments over a host of excitation pump fluences compared to the triplet population calculated based on the kinetic model (blue line). The fit was performed simultaneously to all data sets. The singlet and triplet populations predicted by the model shows good qualitative agreement with the data arrived at by the fitting procedure outlined previously, tracking both the singlet population decay and triplet population growth over multiple time scales (Figure 3.8). The model, however, does not properly reproduce the amplitude of the triplet populations measured in TA experiments at high excitation fluence. This discrepancy most likely arises from the model using time-independent rate coefficients to describe exciton annihilation processes and 77 overlooks diffusive contributions to these events. 22,23 Resultantly, the fits overestimate the amount of triplet-triplet annihilation present at long delays. While the fit to the data varies most notably at high excitation fluences, at lower excitation densities where exciton annihilation events are suppressed the fit improves. In fact, the model reproduces 99% of the singlet exciton decay observed in the time-resolved fluorescence measurements, which are collected with excitation densities 1-2 orders of magnitude smaller than those employed in TA experiments. Table 3.1. Best-fit parameters for the kinetic model described in Section 3.5. parameter best-fit value 1/k sf 0.8 ps a 1.9 x 10 -3 ps -1 b 1.2 x 10 -2 ps -1/2 k SS 3.9 x 10 -9 cm 3 /s k TT 9.0 x 10 -11 cm 3 /s k TS 4.6 x 10 -11 cm 3 /s δ 0.33 R 4.3 Å D 1.5 x 10 -5 cm 2 /s Comparison of the best fit values from Table 3.1 with those reported for thin films of tetracene and pentacene highlights interesting differences involving SF in DPT thin films. SF occurs in 0.8 ps in DPT, more than an order of magnitude faster than the ~40-90 ps time scales cited for tetracene polycrystalline thin films, 18,24,25 but still considerably slower that the 80 fs fission rate reported for pentacene films. 19,26 The drastic increase of SF rate going from tetracene and pentacene can be understood in part from the thermodynamics of the problem. In tetracene, 2E(T 1 ) – E(S 1 ) ≈ 0.19 – 0.24 eV and SF is endothermic and requires thermal activation. 11,27,28 For pentacene, 2E(T 1 ) – E(S 1 ) = -0.11 eV and fission proceeds with an exothermic driving force. For DPT in a benzene solution, 2E(T 1 ) – E(S 1 ) = -87 meV. 15 The monomeric nature of DPT thin film absorption and emission spectra suggests small inter-chromophore coupling and electronic states that are minimally perturb in the condensed media. Therefore, if the DPT state energies 78 remain mostly unchanged from solution to thin film, DPT SF might not require significant thermal activation like in tetracene. Information on the energetics of SF can also be gleaned from analysis of the rate of the reverse reaction (k TS ), triplet-triplet annihilation. Exothermic SF demands endothermic triplet- triplet annihilation. In time-resolved emission measurement of tetracene films a model treating exciton annihilation using time-independent rate coefficients qualitatively reproduced data with an annihilation constant of k TS = 5.0 x 10 -10 cm 2 /s. 14 This value is roughly 10x larger than the best-fit value obtained for DPT thin films, suggesting a larger thermodynamic barrier for triplet- triplet annihilation in DPT than tetracene. Interestingly, the model predicts that one-third (δ = 0.33) of the initially prepared singlet excitons are created at dimer sites capable of undergoing SF and may be surprising given the high degree of conformational disorder in the DPT film. It is possible that a high percentage of DPT dimer pairs pack with their π-systems in close proximity but do not impose any long-range order on the film. Using the parameters contained within the expression for k D (t), the singlet diffusion constant (D) was extracted to be 1.5 x 10 -5 cm 2 /s and the SF encounter radius, R = 4.3 Å. The extracted value for R matches well with literature results from ab initio calculations on pentacene dimers. 29 For DPT, however, an encounter radius of 4.3 Å is much larger than the average spacing between molecules of 10.9 Å, as calculated on the basis of the film’s concentration as measured by optical absorption. This suggests the DPT dimer pair geometry responsible for promoting fission are sites where two acenes closely associate. 3.6 DPT Crystal Structure To garner insight into the possible close-approach, dimer pair geometry responsible for SF in DPT thin films, DPT crystals were grown and crystal structure solved. The DPT crystals 79 were grown using a low-pressure, vapor-phase vacuum sublimation approach and is a method similarly employed in the growth of vapor deposited thin films. Similar crystals were also grown by Kitamura et al. from xylene solvent. 30 In the vapor-grown crystals, fourteen of the eighteen carbons composing the acene backbone lie within 0.08 Å of a least-squares plane (Figure 3.9). The remaining four carbons are bent upwards from this plane by up to 0.37 Å. This arrangement is also witnessed in the xylene-grown crystals from literature. The main difference between the vapor and xylene grown crystals arises from the DPT phenyl rings. For sublimed crystals, the phenyl rings are twisted in the same direction, constituting a 70º dihedral angle. In contrast, xylene crystallized samples have phenyl rings canted in opposite directions, with dihedral angles of 75º and 94º. Figure 3.9. Comparison between DPT crystal structures from samples grown through vapor-phase deposition or in xylene solution. The molecular packing of DPT in vapor-grown crystals deviates drastically from the herringbone arrangement of tetracene crystals that places the π-systems of neighboring chromophores 51º apart. 31,32 DPT stacks cofacially along the crystal’s a-axis and alternate between being eclipsed and staggered (Figure 3.10). For eclipsed pairs, the pendant phenyl groups prevent close approach of the acene backbones and they reside approximately 4.00 Å away from one another. For staggered pairs the spacing between the acene backbones is even 80 closer, 3.68 Å. The xylene-grown DPT crystals from literature also adopt alternating eclipsed and staggered conformations of DPT molecules, with a 4.00 Å spacing for the eclipsed pair and a 3.82 Å spacing for the staggered pair. Figure 3.10. Structure of vapor-grown DPT crystals as determined from X-ray diffraction experiments. (Left) View down the crystal lattice a-axis highlighting the cofacial arrangement of DPT molecules. (Right) The spacing between DPT acene cores alternates between eclipsed and staggered at a distance of 4.00 Å and 3.68 Å respectively. 3.7 References (1) Milita, S.; Servidori, M.; Cicoira, F.; Santato, C.; Pifferi, A. Synchrotron X-ray Investigation of Tetracene Thin Films Grown at Different Deposition Fluxes. Nuc. Instrum. Methods Phys. Res., Sect. B 2006, 246, 101-105. (2) Klevens, H. B.; Platt, J. R. Substituted Borazole Spectra: Gain and Loss of Aromatic Character J. Chem. Phys. 1949, 17, 470-481. (3) Tanaka, J. The Electronic Spectra of Pyrene, Chrysene, Azulene, Coronene and Tetracene Crystals. Bull. Chem. Soc. Jpn. 1965, 38, 86-103. (4) Cicoira, F.; Santato, C.; Dinelli, F.; Murgia, M.; Loi, M. A.; Biscarnini, F.; Zomboni, R.; Heremans, P.; Muccini, M. Morphology and Field-Effect-Transistor Mobility in Tetracene Thin Films, Adv. Funct. Mater., 2005, 15, 375-380. 81 (5) Shao, Y.; Sista, S.; Chu, C.-W.; Sievers, D.; Yang, Y. Enhancement of Tetracene Photovoltaic Devices with Heat Treatment, Appl. Phys. Lett., 2007, 90, 103501. (6) Schlosser, D. W.; Philpott, M. R. Singlet Excitons in Crystalline Naphthalene, Anthracene, Tetracene and Pentacene. Chem. Phys., 1980, 49, 181-199. (7) Yamagata, H.; Norton, J.; Hontz, E.; Olivier, Y.; Beljonne, D.; Bredas, J. L.; Silbey, R. J.; Spano, F. C.; Bardeen, C. J. The Nature of Singlet Excitons in Oligoacene Molecular Crystals. J. Chem. Phys., 2011, 134, 204703. (8) Spano, F. C. Excitons in Conjugated Oligomer Aggregates, Films, and Crystals. Annu. Rev. Phys. Chem. 2006, 57, 217-243. (9) Lime, S.-H.; Bjorklund, T. G.; Spano, F. C.; Bardeen, C. J. Exciton Delocalization and Superradiance in Tetracene Thin Films and Nanoaggregates. Phys. Rev. Lett. 2004, 92, 107402. (10) Merrifield, R. E. Magnetic Effects on Triplet Exciton Interactions. Pure Appl. Chem., 1971, 27, 481-498. (11) Groff, R. P.; Avakian, P.; Merrifield, R. E. Coexistence of Exciton Fission and Fusion in Tetracene Crystals. Phys. Rev. B, 1970, 1, 815-817. (12) Johnson, R. C.; Merrifield, R. E. Effects of Magnetic Fields on the Mutual Annihilation of Triplet Excitons in Anthracene Crystals. Phys. Rev. B 1970, 1, 896-902. (13) Ryasnyanskiy, A.; Biaggio, I. Photophysics of Pentacene Thin Films: The Role of Exciton Fission and Heating Effects. Phys. Rev. B, 2011, 84, 193203. (14) Burdett, J. J.; Muller, A. M.; Gosztola, D.; Bardeen, C. J. Excited State Dynamics in Solid and Monomeric Tetracene; The Roles of Superradiance and Exciton Fission. J. Chem. Phys., 2010, 133, 144506. (15) Burgdorff, C.; Kircher, T.; Lohmannsroben, H.-G. Photophysical Properties of Tetracene Derivatives in Solution. Spectrochim. Acta, Part A, 1988, 44, 1137-1141. (16) Burgdorff, C.; Lohmannsroben, H.-G. Photophysical Properties of Tetracene Derivatives in Solution III. Thermally Activated Nonradiative Processes and Triplet State Properties. J. Lumin., 1994, 59, 201-208. 82 (17) Roberts, S. T.; Schlenker, C. W.; Barlier, V.; McAnally, R. E.; Zhang, Y.; Mastron, J. N; Thompson, M. C.; Bradforth, S. E. Observation of Triplet Exciton Formation in a Platinum-Sensitized Organic Photovoltaic Device. J. Phys. Chem. Lett. 2011, 2, 48-54. (18) Grumstrup, E. M.; Johnson, J. C.; Damrauer, N. H. Enhanced Triplet Formation in Polycrystalline Tetracene Films by Femtosecond Optical-Pulse Shaping. Phys. Rev. Lett., 2010, 105, 257403. (19) Wilson, M. W. B.; Rao, A.; Clark, J.; Kumar, R. S. S.; Brida, D.; Cerullo, G.; Friend, R. H. Ultrafast Dynamics of Exciton Fission in Polycrystalline Pentacene. JACS, 2011, 133, 11830-11833. (20) Powell, R. C.; Soos, Z. G. Singlet Exciton Energy Transfer in Organic Solids. J. Lumin. 1975, 11, 1-45. (21) Rice, S. A. Diffusion-Limited Reactions, Elsevier: Amsterdam, 1985; Vol. 25. (22) Engel, E.; Leo, K.; Hoffmann, M. Ultrafast Relaxation and Exciton-Exciton Annihilation in PTCDA Thin Films. Chem. Phys., 2006, 325, 170-177. (23) Marciniak, H.; Pugliesi, I.; Nickel, B.; Lochbrunner, S. Ultrafast Singlet and Triplet Dynamics in Microcrystalline Pentacene Films. Phys. Rev. B, 2009, 79, 235318. (24) Thorsmolle, V. K.; Averitt, R. D.; Demsar, J.; Smith, D. L.; Tretiak, S.; Martin, R. L.; Chi, X.; Crone, B. K.; Ramirez, A. P.; Taylor, A. J. Morphology Effectively Controls Singlet-Triplet Exciton Relaxation and Charge Transport in Organic Semiconductors. Phys. Rev. Lett., 2009, 102, 017401. (25) Burdett, J. J.; Gosztola, D.; Bardeen, C. J. The Dependence of Singlet Exciton Relaxation on Excitation Density and Temperature in Polycrystalline Tetracene Thin Films: Kinetic Evidence for a Dark Intermediate State and Implications for Singlet Fission. J. Chem. Phys. 2011, 135, 214508. (26) Rao, A.; Wilson, M. W. B.; Hodgkiss, J. M.; Albert-Seifried, S.; Bassler, H.; Friend, R. H. Exciton Fission and Charge Generation via Triplet Excitons in Pentacene/C 60 Bilayers. JACS, 2010, 132, 12698-12703. (27) Tomkiewicz, Y.; Groff, R. P.; Avakian, P. Spectroscopic Approach to Energetics of Exciton Fission and Fusion in Tetracene Crystals. J. Chem. Phys. 1971, 54, 4504-4507. 83 (28) Jundt, C.; Klein, G.; Sipp, B.; Moigne, J. L.; Joucla, M.; Villaeys, A. A. Exciton Dynamics in Pentacene Thin Films Studied by Pump-Probe Spectroscopy. Chem. Phys. Lett., 1995, 241, 84-88. (29) Zimmerman, P. M.; Zhang, Z.; Musgrave, C. B. Singlet Fission in Pentacene Through Multi-Exciton Quantum States. Nat. Chem., 2010, 2, 648-652. (30) Kitamura, C.; Matsumoto, C.; Kawatsuki, N.; Yoneda, A.; Kobayashi, T.; Naito, H. Crystal Structure of 5,12-Diphenyltetracene. Anal. Sci. 2006, 22, 5-6. (31) Robertson, J. M; Sinclair, V. C.; Trotter, J. The Crystal and Molecular Structure of Tetracene. Acta Crystallogr., 1961, 14, 697-704. (32) Campbell, R. B.; Robertson, J. M.; Trotter, J. The Crystal Structure of Hexacene, and a Revision of the Crystallographic Data for Tetracene. Acta Crystallogr., 1962, 15, 289- 290. 84 Chapter 4. Photophysics of Platinum Porphyrin-Doped DPT Films* *Published in Journal of Physical Chemistry Letters, 2011, 2, 48-54. 4.1 Host-Guest Sensitization of Triplet Excitons In the absence of strong spin-orbit coupling triplet excitons of chromophores most commonly used in organic photovoltaics persist for μs-ms due to forbidden relaxation to the closed-shell singlet electronic ground state of these molecules. 1 These lifetimes are orders of magnitude longer than the fluorescent (τ = ns) lifetimes since radiative recombination between the singlet excited and singlet ground states is allowed. Giving excited states prolonged lives and a greater opportunity to be charge separated at the D/A interface through the purposeful population of triplets could offer a potential advantage for OPVs. However, owing to the short- range Dexter type mechanism 2 of triplet energy migration it is unclear if prolonged lifetimes result in increased exciton diffusion lengths. Triple diffusion lengths on the order of a few nanometers in condensed organic films, 3 to several microns in molecular single crystals 4 have been reported. Further, measured triplet diffusion lengths for a given material can vary drastically depending on the methodology of their measure. 5-8 Therefore, being able to probe and understand the excited state behaviors of triplet excitons in condensed organic phases is important for their utilization in OPVs. 9 This truth is ever more apparent for SF, a photophysical process that results in two low energy triplet excited states and demands the inclusion of other complementary absorbing chromophores in order to boost the efficiency of an OPV. In order for SF to be a useful contributor to organic solar cell efficiencies the interplay between fission and other photophysical processes such as energy and charge transfer to molecular dopants or Acceptors must be understood. If another S 1 deactivation channel kinetically competes with singlet fission then an OPV will not be able to advantageously harness 85 fission to increase the power conversion efficiency of the cell. In this Chapter the interaction and excited state dynamics between a host-guest system of the singlet fission material 5-12 diphenyltetracene (DPT), and the complementary, red-absorbing platinum tetraphenylbenzoporphyrin (Pt(TPBP)) are probed using femtosecond transient absorption spectroscopy (TA). Doping the porphyrin sensitizer at a low percentage (5%) into the DPT host creates a host-guest system in close molecular contact. Excitation of either chromophore creates excited state population density that can be tracked through energy transfer processes to reveal the ultimate fate of excitons in the a Pt(TPBP):DPT doped film. Uncovering the photophysical processes operant upon film excitation divulges the photo-electrical conversion mechanism operant in an organic photovoltaic employing Pt(TPBP):DPT doped layers as Donor materials. 4.2 Pt(TPBP):DPT Doped Films Figure 4.1. Molecular structures of the platinum porphyrin (PtTPBP) sensitizer and acene host (DPT) used triplet sensitization host-guest experiments. To this end a model host-guest architecture consisting of Pt(TPBP) doped into a host of DPT (Figure 4.1) was investigated using ultrafast TA and steady-state photocurrent response measurements. The materials were carefully chosen to given complementary photon absorption with minimal overlapping features. It is well known that transition metal complexes such as porphyrins and phthalocyanines give high triplet yields from strong spin-orbit coupling enhancing the intersystem crossing rate. 2,10,11 In our model system complementary absorption of 86 DPT host and Pt(TPBP) dopant allows for exclusive photoexcitation of the porphyrin. From solution experiments it is known that Pt(TPBP) excitation results in phosphorescence at 769 nm (1.6 eV) 12 with minimal fluorescence due to rapid inter-system crossing converting the prepared porphyrin singlet excited state into a triplet. 13,14 Energy transfer from the porphyrin triplet to the DPT host triplet (1.2 eV) 15 is energetically favorable and allows for exclusive population of the acene triplet via excitation of a guest sensitizer. Such a scheme has been well studied for intramolecular energy transfer between dyads of energy donor and accepting chromophores covalently bound in the same molecule. 16-18 Here the same mechanism is probed in condensed media for an intermolecular host-guest system. Figure 4.2. Thin film absorption spectra for 5% Pt(TPBP):DPT doped films (black solid), 20% Pt(TPBP):TFS doped films (green dashed), and neat DPT (red dashed). Figure 4.2 shows an absorption spectrum typical of Pt(TPBP) doped DPT films. DPT exhibits strong absorption between 400-500 nm. The porphyrin Q and Soret band absorption appears prominently at 618 nm and 435 nm respectively. No new spectral features are seen for Pt(TPBP):DPT doped films when compared to neat DPT or Pt(TPBP) doped into a non- interacting TFS host matrix (TFS = tetra(9,9’-dimethylfluoren-2-yl)silane), indicating that the Pt(TPBP):DPT doping does not produce any new electronic states not present in the separate species. 87 4.3 Ultrafast Transient Absorption Spectroscopy of Pt(TPBP) Doped Films Figure 4.3. TA spectra of 20% Pt(TPBP) in TFS at various time delays. Upper left, spectra collected to a time delay of 10 ps. Upper right, close-up of the induced absorption features occurring between 470 – 570 nm at fast delays. Bottom, long delays out to 600 ps showing minimal spectral evolution. First, Pt(TPBP) excited state dynamics were probed in a TFS matrix. Because TFS’s triplet state (2.9 eV) is higher in energy than Pt(TPBP) (1.6 eV) the porphyrin excited state absorption and kinetics of triplet formation could be characterized uninfluenced by the host. Pumping Pt(TPBP) at 618 nm results in an immediate photobleach of the Q (618 nm) and Soret (435 nm) bands due to the depopulation of the porphyrin ground state. Between these ground state bleaches an induced absorption feature is observed to evolution with time. Spectral evolution over the first 10 ps reveals the presence of an isosbestic point at 515 nm. Beyond 10 ps little spectral evolution is witnessed and the spectral lineshape at long delays (750 ps) matches literature reported spectra for the Pt(TPBP) triplet spectrum using ns TA. 19 Therefore, it is concluded the isosbestic point witnessed at early delays is due to inter-system crossing from the Pt(TPBP) singlet state to the triplet. Fitting the decay traces with a two-state kinetic model gives an inter-system crossing rate of 1/k ISC = 400 fs. 88 Figure 4.4. Left: TA spectra for 5% Pt(TPBP):DPT films. Spectral evolution occurs with new induced absorption features appearing at 450 and 490 nm at long delays not seen in films of isolated Pt(TPBP). Right: Comparison between DPT triplet exciton lineshape taken from literature (red) and the residual of TA spectra at long delays. With the excited state kinetics of isolated Pt(TPBP) analyzed, Pt(TPBP) doped DPT films can now be explored. For Pt(TPBP) doped DPT films excitation of the porphyrin gives a TA signal similar to that seen for Pt(TPBP) in the non-interacting TFS matrix. However, over the course of 200 ps the Q and Soret photobleaches of Pt(TPBP) recover, indicating repopulation of the porphyrin ground state. Concomitantly, the induced absorption band between 440 and 600 nm evolves from the spectrum associated with the Pt(TPBP) triplet to a different lineshape with peaks at 450 and 490 nm. A previous solution-phase nanosecond TA study of DPT identified the acene’s triplet absorption spectrum to occur between 450 and 500 nm, with a peak at 490 nm and was able to quantify its cross-section. 15 Similar lineshapes seen in Pt(TPBP):DPT films suggest Pt(TPBP) triplet energy transfer to DPT. Triplet sensitization was confirmed by averaging the TA signal of Pt(TPBP):DPT films between 600 ps and 750 ps. A lack of spectral evolution after 600 ps suggests that the energy transfer process is largely complete. Subtraction of the Pt(TPBP) triplet signal acquired from the TFS experiments at these delays removes the contribution of the Pt(TPBP) T 1 , and leaves a 89 residual spectrum. When overlaid with the DPT T 1 line shape expected from literature extinction spectra, the agreement is excellent, matching well the relative scaling between the peaks at 490 and 450 nm. 15 Using a three state kinetic model: the TA spectra could be fit to obtain values for the rate of intersystem in Pt(TPBP) (k ISC ) and the rate of triplet energy transfer from Pt(TPBP) to DPT (k TT ). The best fit values for 1/k ISC is 400 fs while the best fit for 1/k TT is 35 ps. Along with tracking the triplet population transfer, the total yield of DPT triplet excitons could be calculated. The ratio between the literature DPT T 1 extinction spectra and the residual TA signal at long delays provides an estimate of the concentration of triplet excitons formed via Pt(TPBP) photoexcitation. Similarly, literature extinction spectra for Pt(TPBP) compared the initial photobleach of the Q band provides an estimate of the concentration of excited Pt(TPBP) prepared in the pump pulse. 13 Comparison of these two values reveals triplet energy transfer from Pt(TPBP) to DPT is highly efficient, with 85 ± 6% of excited Pt(TPBP) molecules forming DPT triplet excitons. 90 In Chapter 3 TA experiments on neat DPT thin films revealed efficient SF (~ 140% triplet production at low pump fluences) occurring with two time constants, ~ 1 ps for fission occurring promptly at photoexcited dimer sites, and ~ 100 ps for excitons diffusing to those dimer sites. It was just shown that excitation of Pt(TPBP) in Pt(TPBP):DPT doped films leads to fast porphyrin intersystem crossing followed by triplet sensitization to the acene. The question remains as to what the fate of DPT excitations in the presence of the platinum porphyrin sensitizer. Given that the DPT singlet excited state (E S1 = 2.4 eV) is greater in energy than the Pt(TPBP) singlet state (E S1 = 1.9 eV), photoexcitation of DPT in Pt(TPBP):DPT films could lead to FRET to the dopant, sensitizing the porphyrin S 1 and circumvent the SF pathway. Figure 4.5 shows TA spectra collected for Pt(TPBP):DPT films at various time delays following excitation of DPT. After the DPT excitation, porphyrin Q and Soret photobleach occurs on sub-picosecond time scales, confirming singlet energy transfer from the acene to the porphyrin. Fits to the decay reveal a time constant of ~ 4.6 ps for singlet energy transfer. Given that the fastest, prompt Figure 4.5. TA spectra for 5% Pt(TPBP):DPT films following DPT excitation. Prompt Pt(TPBP) Q and Soret photobleach indicates singlet energy transfer occurs from DPT with a rate constant of 4.6 ps. At longer delays the porphyin photobleach recovers due to intersystem crossing following triplet energy transfer repopulating the Pt(TPBP) ground state. The energy transfer scheme results in approximately 78.5 DPT triplets being produced per 100 DPT S 1 excitations prepared in the pump pulse. 91 component of triplet production via SF in neat DPT films occurs at about 1 ps, a 4.6 ps time constant for singlet energy transfer actively competes with the SF process, and effectively circumvents the delay (~100 ps) component of SF. After 10 ps the porphyrin ground state bleach recedes, indicating repopulation of the prophyrin ground state via intersystem crossing followed by triplet sensitization back to the acene host. In total, excitation of DPT in the presence of 5% Pt(TPBP) dopant results in a 74% triplet production efficiency (100 DPT S 1 excitations lead to 74 triplets produced). This yield compares disfavorably to a 140% triplet production efficiency when DPT is excited in the absence of a dopant. The ability for a complementary, low-energy absorber to siphon excitation density away from an advantageous SF channel through energy transfer highlights the importance of proper device design and engineering when constructing SF OPVs. 4.4 OPVs with Pt(TPBP):DPT Doped Donor Layers With the excited state dynamics of Pt(TPBP) doped DPT films characterized, the relative contribution of triplet excitons from Pt(TPBP) absorption to the photocurrent of OPV devices were explored (Figure 4.6). Devices of the architecture glass/ITO/donor/C 60 (17.5 nm)/BCP(10 nm)/Al (100 nm) were fabricated through thermal evaporation of materials in a high-vacuum chamber. Here, ‘donor’ is either neat DPT or DPT doped with 5% Pt(TPBP) by volume and BCP = 2,9-dimethyl-4,7-diphenyl-1,10phenanthroline. 92 Figure 4.6. Left: EQE spectra of OPVs with neat DPT or 5% Pt(TPBP):DPT as donor layers. Pt(TPBP) photoresponse at 625nm increases with increasing doped donor thickness. Devices with Pt(TPBP) increase DPT photoresponse in the wavelength regions 400-500 nm. Right: J-V curves of the same devices under 600nm long- pass excitation. As the doped donor layer thickness increases, Jsc correspondingly increases with minimal drop in voltage. EQE measurements of the doped devices show photocurrent response at 625 nm from the Pt(TPBP) Q band that increases with increasing donor layer thickness, demonstrating that Pt(TPBP) photoexcitation leads to charge generation. For devices with a doped donor layer, DPT photoresponse (λ = 400-500nm) increases drastically from a device with a neat DPT donor layer and also scales with doped donor layer thickness, suggesting that charge generation by photoexcited DPT molecules is aided by the presence of Pt(TPBP). Table 4.1. J-V Metrics Measured for neat DPT or Pt(TPBP):DPT doped donor layers OPVs of the architecture donor/C 60 (17.5 nm)/BCP (10 nm)/Al (100 nm). Devices are measured under 600nm longpass filtered Xe Arc illumination. donor thickness (nm) J sc (mA/cm 2 ) V oc (V) fill factor DPT 25 0.030 0.83 0.40 5% Pt(TPBP)/DPT 10 0.043 0.73 0.53 5% Pt(TPBP)/DPT 15 0.047 0.76 0.53 5% Pt(TPBP)/DPT 20 0.053 0.77 0.53 5% Pt(TPBP)/DPT 25 0.066 0.77 0.50 93 J-V characteristics of doped versus undoped devices were measured under Xe arc illumination filtered through a 600 nm long-pass filter, allowing for preferential excitation of the Pt(TPBP) guest material (Table 4.1). For devices with a neat DPT donor layer, excitation at λ > 600 nm leads to non-zero photoresponse due to C 60 absorption. Upon addition of 5% Pt(TPBP) to the DPT, the short-circuit current (J sc ) under long-pass illumination more than doubles as a result of Pt(TPBP) absorption and sensitized DPT triplet diffusion to the D/A interface. As the doped donor layer thickness is increased from 15 to 25 nm, J sc increases accordingly and the open-circuit voltage remains unchanged. The large photovoltage of DPT devices eclipse those measured for similar devices based on planar conjugated acenes such as tetracene 20 and pentacene. 21 In part this voltage increase originates from DPT’s deep HOMO energy (-5.4 eV) and inaccessibility of the backbone π- system limiting back electron transfer after charge dissociation due to the steric bulk of the phenyl groups. 12 V oc values for doped and undoped DPT devices are comparable (0.77 and 0.73 V respectively), suggesting a mutual charge transfer state is created from the dissociation of singlet and triplet DPT excitons. 94 These results demonstrate the viability of intermolecular host-guest systems to enhance exciton collection efficiencies in OPV devices. By judicious choice of complementary absorbing materials with properly aligned state energies the interplay between singlet and triplet excited states can be advantageously used to generate photocurrent in OPVs. For neat DPT it was shown that photoexcitation leads to triplet production with two time constants via SF, a prompt component ~1 ps and a diffusive component ~100 ps (Figure 4.7). Excitation of DPT in the presence of Pt(TPBP) sensitizes singlets to the porphyin via energy transfer with in ~ 4.6 ps and kinetically competes with DPT SF. Pt(TPBP) singlets intersystem cross (1/kisc = 400 fs) and energy transfer back to the acene host, populating the DPT T 1 with a rate constant of 1/kTT = 35 ps. In an OPV the resulting excitons can be effectively charge separated at a C 60 interface, and harness as electrical work, producing a cell voltage of 0.7 V. Figure 4.7. Schematic showing the various photophysical processes operant in Pt(TPBP):DPT doped thin films. Photoexcitation of DPT (dashed red line) creates singlet excited states that decay via singlet energy transfer to Pt(TPBP) in doped films and prompt and diffusive singlet fission in undoped, neat films (green arrow). The rate of singlet energy transfer from the acene to the porphyrin (4.6 ps) kinetically competes with prompt SF (1 ps) in neat DPT films and circumvents the diffusive component (100 ps). Singlet excitations on the porphyrin, whether sensitized through energy transfer or photoexcited directly (dashed blue line) quickly intersystem cross (400 fs) producing triplets that are then sensitized through energy transfer (35 ps) to the DPT host. Organic photovoltaics using Pt(TPBP):DPT active layers separate molecular excited states, resulting in an open-circuit voltage of 0.7 V. 95 4.5 References (1) Kroeze, J. E.; Savenije, T. J.; Warman, J. M. Efficient Charge Separation in a Smooth- TiO 2 /Palladium-Porphyrin Bilayer Via Long-Distance Triplet-State Diffusion. Adv. Mater. 2002, 14, 1760-1763. (2) Turro, N. J. Modern Molecular Photochemistry: University Science Books: Sausalito, CA, 1991. (3) Wunsche, J.; Reineke, S.; Lussem, B.; Leo, K. Measurement of Triplet Exciton Diffusion in Organic Light-Emitting Diodes. Phys. Rev. B, 2010, 81, 245201. (4) Irkhin, P.; Biaggio, I. Direct Imaging of Anisotropic Exciton Diffusion and Triplet Diffusion Length in Rubrene Single Crystals. Phys. Rev. Lett., 2011, 107, 017402. (5) Matsusue, N; Ikame, S.; Suzuki, Y.; Naito, H. Charge-Carrier Transport and Triplet Exciton Diffusion in a Blue Electrophosphorescent Emitting Layer, J. Appl. Phys., 2005, 97, 123512. (6) Giebink, N. C.; Sun, Y.; Forrest, S. R. Transient Analysis of Triplet Exciton Dynamics in Amorphous Organic Semiconductor Thin Films. Org. Electron., 2006, 7, 375-386. (7) Zhou, Y. C.; Ma, L. L.; Zhou, J.; Ding, X. M.; Hou, X. Y. Effects of a Sensing Layer on Triplet Exciton Diffusion in Organic Films. Phys. Rev. B, 2007, 75, 132202. (8) D’Andrade, B. W.; Thompson, M. E.; Forrest, S. R. Controlling Exciton Diffusion in Multilayer White Phosphorescent Organic Light Emitting Devices, Adv.Mater., 2002, 14, 147-151. (9) Schlenker, C. W.; Chen, K-S.; Yip, H-L.; Li, C-Z.; Bradshaw, L. R.; Ochsenbein, S. T.; Ding, F.; Li, X. S.; Gamelin, D. R.; Jen, A. K-Y.; Ginger, D. S.; Polymer Triplet Energy Levels Need Not Limit Photocurrent Collection in Organic Solar Cells. JACS, 2012, 134, 19661-19668. (10) Frackowia, D.; Planner, A.; Waszkowiak, A.; Boguta, A.; Ion, R.-M.; Wiktorowicz, K. Yield of Intersystem (Singlet-Triplet) Crossing in Phthalocyanines Evaluated on the Basis of a Time Resolved Photothermal Method. J. Photoch. Photobiol. A, 2001, 141, 101-108. (11) Forster, L. S. Intersystem Crossing in Transition Metal Complexes. Coord. Chem. Rev., 2006, 250, 2023-2033. (12) Perez, M. D.; Borek, C.; Forrest, S. R.; Thompson, M. E. Molecular and Morphological Influences on the Open Circuit Voltages of Organic Photovoltaic Devices. JACS, 2009, 131, 9281-9286. 96 (13) Borek, C.; Hanson, K.; Djurovich, P. I; Thompson, M. E.; Aznavour, K.; Bau, R.; Sun, Y.; Forrest, S. R.; Brooks, J.; Michalski, L.; Brown, J. Highly Efficienct, Near-Infrared Electrophosphorescence from a pt-Metalloporphyrin Complex. Angew. Chem. Int. Ed. 2007, 46, 1109-1112. (14) Perez, M. D.; Borek, C.; Djuorovich, P. I.; Mayo E. I.; Lunt, R. R.; Forrest, S. R.; Thompson, M. E. Organic Photovoltaics Using Tetraphenylbenzoporphyrin Complexes as Donor Layers. Adv. Mater., 2009, 21, 1517-1520. (15) Burgdorff, C.; Kircher, T.; Lohmannsroben, H.-G, Photophysical Properties of Tetracene Derivatives in Solution. Spectrochim. Acta, Part A, 1988, 44A, 1137-1141. (16) Whited, M. T.; Djurovich, P. I.; Roberts, S. T.; Durrell, A. C.; Schlenker, C. W.; Braforth, S. E.; Thompson, M. E. Singlet and Triplet Excitation Management in a Bichromophoric Near-Infrared-Phosphorescent BODIPY-Benzoporphyrin Platinum Complex, JACS, 2010, 133, 88-96. (17) Cotlet, M.; Vosch, T.; Habuchi, S.; Weil, T.; Mullen, K.; Hofkens, J.; De Schryver, F. Probing Intramolecular Forster Resonance Energy Transfer in a Naphthaleneimide- Peryleneimide-Terrylenediimide-Based Dendrimer by Ensemble and Single-Molecule Fluorescence Spectroscopy, JACS, 2005, 127, 9760-9768. (18) Adronov, A.; Frechet, J. M. J. Chem. Commun., 2000, 1701-1710. (19) Singh-Rachford, T. N.; Castellano, F. N. Supar-Nanosecond Dynamics of a Red-to-Blue Photon Upconversion System. Inorg. Chem., 2009, 48, 2541-2548. (20) Chu, C.-W.; Shao, Y.; Shrotriya, V.; Yang, Y. Efficient Photovoltaic Energy Conversion in Tetracene-C 60 Based Heterojunctions. Appl. Phys. Lett., 2005, 86, 243506. (21) Postcavage, W. J., Jr.; Yoo, S.; Kippelen, B. Origin of the Open-Circuit Voltage in Multilayer Heterojunction Organic Solar Cells. Appl. Phys. Lett., 2008, 93, 193308. 97 Chapter 5. Construction of a Custom Physical Property Measurement System Solar Cell Testing Station 5.1 Testing Station Construction Motivation The possibility of OPVs to offer a low-cost, scalable solution for solar energy conversion will only be realized if power conversion efficiencies are improved. Since the power conversion efficiency of the cell is the mathematical product of the fill factor, short-circuit current, and open-circuit voltage (η = FF x J sc x V oc ), the three device metrics must be maximized together to achieve high η. Singlet fission offers a unique way to accomplish this, however a careful balance between increased current and depressed photovoltage must be found for SF to truly be an advantageous phenomenon. Understanding the operating mechanism and molecular electronic states participant in the power conversion process in OPVs is key to making high efficiency devices. Specifically, being able to probe the CT state formed at the D/A interface and measure its energy to discern its impact on the cell open-circuit voltage is paramount. Since [D + /A - ] CT state formation is a mandatory step along the power conversion process in OPVs, the CT state energy sets an upper limit on the achievable electrochemical potential of the separate hole and electron (V oc ). For OPVs that employ SF in a carrier multiplication scheme, measuring the CT energy is even more crucial as these devices must separate low energy triplets. To this end we set out to construct a testing station apparatus that allows for simultaneous probing of CT state energies in functioning OPVs, and the operating role SF plays in those devices. Magnetic field-dependent photocurrent measurements have been shown to provide insight into the role SF plays in functioning OPVs (Chapter 2). Further, variable temperature device testing has been shown as an effective means to extract CT state energies. Therefore, our 98 testing apparatus seeks to combine the variable temperature capabilities of a cryostat with applied magnetic fields necessary to explore SF. 5.2 The Physical Property Measurement System Figure 5.1. Photographs of the Physical Property Measurement System (left) and the 2.6 cm diameter bore opening at the top of the instrument (upper right). (Bottom right) At the bottom of the bore sits a 12-pin electrical lead used for interfacing samples with testing equipment. A Quantum Design Physical Properties Measurement System (PPMS) offers researchers a “unique concept in laboratory equipment” with, “… an open architecture, variable temperature- field system, and designed to perform a variety of automated measurements.” The system boasts sample environmental controls including magnetic fields up to ±16 tesla and temperature range of 1.9 - 400K. A modular design affords the system dynamic versatility and allows researchers to adapt the equipment to accommodate innumerable custom experiments. 1 Given the unique possibilities of the PPMS, we set out to build a custom solar cell testing station inside such a system (Figure 5.1, left). Since the PPMS allows for dynamic temperature control and magnetic field application, such a custom-built system could be used for normal OPV testing, to measure CT state energies through variable temperature testing, and to probe for 99 SF in properly constructed devices. The main challenge was to figure out how to adapt the equipment normally used in table-top solar cell testing stations to be accommodated for use with the PPMS. A PPMS is constructed with a long cylindrical metal bore running vertically down the center of a superconducting magnet (Figure 5.1). The bore is fairly narrow, only 2.6 cm in diameter (Figure 5.1, upper right). At the bottom of the bore a 12-pin electrical lead is built into the cryostat insert and is used to make electrical contact to samples (Figure 5.1, lower right). In order for samples to be able to reach the electrical pins at the bottom of the bore and interface with the system, rod-like probes several feet in length are constructed. These probes house the sample of interest in a properly designed substrate holder and terminate at their base with a universal puck that interfaces with the 12-pin electrical lead. The top of these probes terminate in a vacuum tight seal and allows for a low pressure, inert argon atmosphere to be established inside the bore through several pump-purge cycles. The purging of ambient gases and the establishment of an inert argon atmosphere is mandatory for low temperature device testing. Therefore, the maintenance of a vacuum-tight seal is between the probe and PPMS is essential. It was our aim to construct one of these rod-like probes to house our OPV devices and interface with the PPMS for testing purposes. The task was, more succinctly, to design a solar cell testing station probe assembly that would fit into the bottom of a 3 foot long, 2.6 cm diameter cylindrical bore where the 12-pin electrical lead is housed. The probe must be able to maintain a vacuum seal that cannot be compromised. Therefore, visible photons would have to be delivered to the device inside the bore through a vacuum-tight feed through mechanism. All materials used in the construction of the testing station probe assembly must be non-ferric owing to the magnetic field strengths of the PPMS. 100 5.3 PPMS Solar Cell Testing Station Design and Construction The construction of a PPMS probe assembly capable of housing an OPV, establishing and maintaining a low pressure, inert atmosphere inside the PPMS bore while having photons supplied to the device demands particular design criteria to meet each of these goals. First, a substrate holder must be designed to accommodate the geometry of the substrate used in the fabrication of OPV devices. The substrate holder must be able to make electrical connection to the devices on the substrate and be tied into the 12-pin electrical lead at the base of PPMS bore. In order for the electrical contacts to reach the bottom of the bore the substrate holder must be mounted on a rod several feet in length. At the top of this rod a vacuum tight seal must be made with the PPMS. Simultaneously, a light delivery system must be devised to transport photons to the OPV substrate at the bottom of the bore without jeopardizing the established vacuum. The following section outlines the design decisions and considerations made for the construction of the PPMS solar cell testing probe assembly to simultaneously achieve each one of the outlined criteria previously mentioned. Photographs of the constructed probe are provided to illustrate the operational details of the functioning probe. The final section outlines how the fully constructed assembly is used to test OPVs and gather experimental device data. 101 5.3.1 OPV Substrate Description and Parameters Figure 5.2. (Left) Photograph of vapor-deposited organic devices on patterned ITO substrates. The purple material is the deposited organic material and only covers one quadrant of the glass substrate. Patterned ITO appears as pale green vertical stripes. The dark black horizontal stripe is the vapor deposited aluminum cathode (note: the cathode is highly reflective but appears opaque due to the contrast of the photograph. (Right) The same photograph with overlays highlighting the areas of 2 mm ITO striping (red), 4 mm ITO striping (green) and the deposited aluminum cathode (blue). The intersection of the each 2 mm ITO stripe with the deposited metallic cathode defines a device’s active area, accented in yellow. This fabrication method creates four functioning devices, all of the same architecture, per substrate. OPVs fabricated in our research group are normally constructed using patterned ITO stripes on glass substrates as a semi-transparent electrode (Figure 5.2, left). There are five total ITO stripes, 4 stripes each approximately ~2 mm wide, and one thicker stripe ~4 mm wide. During OPV fabrication organic material is vapor deposited onto the patterned ITO glass substrate through a shadow mask such that the device is constructed on one quadrant of the substrate, across all four of the ~2 mm ITO stripes. The mask is aligned so that the thicker ~ 4mm ITO stripe does not have any organic material deposited on it. The device is then completed by vapor depositing a top metal cathode through a second shadow mask to create a ~2 mm thick metallic stripe perpendicular to the ITO striping. This process creates four distinct devices (all of the same architecture) on one glass substrate, each with an approximately 4 mm 2 active area defined by the spatial overlap of the metal cathode with the perpendicular ITO striping (Figure 5.2, right). The thicker ~4 mm ITO stripe is left devoid of deposited organic material, but makes an ohmic contact with the metal cathode. During testing this mode of device 102 fabrication allows for testing of any of the four devices by simply making electrical contact to the thick (~4 mm) ITO stripe and the thin ITO stripe corresponding to the device of interest. Charges of one polarity are collected through the ITO of the device. Concurrently, charges of the opposite polarity are collected through the top metal cathode and then harvested at the ohmic contact of the thick ITO stripe. 5.3.2 Substrate Holder and Probe Construction A solar cell testing station created inside the bore of a PPMS must be able to accommodate our glass substrate geometry. Since our square glass substrates are 2.2 cm on a side, the substrate must be housed in a vertical fashion inside the PPMS bore. With this in mind we set out to construct a custom testing probe that consisted of several parts: a substrate holder to house the device mounted in a vertical fashion, a bracket mount that housed the substrate holder and could be outfitted with electrical leads to tie into the 12-pin lead at the base of the PPMS bore, and a supporting rod several feet in length to reach the PPMS bore bottom. 103 Figure 5.3. Photographs of the substrate holder and copper mounting bracket apparatus used to house OPV devices in the PPMS testing station. (Left) Picture of the Macor substrate holder (orange arrow) outfitted with an OPV glass substrate described in Section 5.3.1 (blue arrow) mounted in the copper bracket. Photons guided down the fiber optic cable deflect 90º off of an aluminum mirror (purple arrow) to fall incident on the device. Electrical contact to each of the four devices per substrate is made individually through brass tabs soldered to blue wires that connect to the universal puck (red arrow) located at the base of the assembly. (Right) Opposite view of the assembly highlighting the variable height screw holes in the copper mounting bracket and the brass tabs (magenta arrows) used to make electrical contact with the ITO striping. First, a mounting bracket was constructed to house the substrate holder for the OPV device. Through close collaboration with the USC Machine Shop a sample mount bracket out of oxygen-free copper with minimal tolerances against the PPMS bore inner diameter was fabricated. The copper bracket is highly modular, with screw holes in the bottom and top of the bracket that allow for variable rod and base attachments. Set screws up the side of the bracket allow the researcher to mount any substrate at the appropriate location for a given experiment. Next, a substrate holder was fabricated from Macor, a machinable glass ceramic that can be formed using generic metalworking tools. The OPV glass substrate outlined in Section 5.3.1 fits vertically into a narrow slot of the substrate holder and is machined to be slightly larger than the thickness of the glass. Inside this slot, five brass tabs are spatially arranged to make contact 104 to the five ITO stripes on the glass substrate in a one-to-one manner and are the source of electrical contact to the device. Soldered wires connect these brass tabs to a universal sample puck mounted on the bottom of the copper mounting bracket. When the puck is plugged into the 12-pin lead at the base of the PPMS bore, the researcher can use external instruments interfaced with the PPMS control software to probe the electrical response of the device. Figure 5.3 photographs the fully constructed assembly used to house OPV devices in the PPMS testing apparatus. Figure 5.4. Photographs of the light delivery system used to direct photons to the OPV substrate. (Left) Picture showing the fiber optic cable hanging parallel to the central rod (white) and being directed through the aluminum baffles (purple). (Right) Photograph highlighting the path length of the cable itself. The top of the copper mounting bracket is outfitted with an aluminum cap affixed to a ¼ inch hollow stainless steel rod. At regular intervals aluminum baffles are screwed to the stainless steel rod and act to dissipate heat from the top of the rod in contact with ambient atmosphere away from the sample at the base of the bore that could be at cryogenic temperatures. The rod 105 terminates at the top of the bore with a centering o-ring ISO KF 16 flange mounted on spring to ensure a quality vacuum seal with the PPMS. 5.3.3 Light Delivery With the testing station assembly in hand, a system to deliver light to the glass substrate in the bottom of the PPMS bore was devised. The delivery method could not compromise the vacuum seal necessary to guarantee inert testing conditions, could not be made of ferric materials due to the applied magnetic fields, and must be able to withstand cryogenic temperatures. Given these criteria we decided to deliver photons to the device using a fiber optic cable custom designed to withstand cryogenic temperatures. The 400 μm diameter silica core/silica clad cryogenic cable was purchased from Fiberguide and is capable of a 300-2400 nm wavelength transmission range between temperatures of -269 ºC to +400 ºC (4 – 673 K). 2 The cable runs the length of the support rod, through guide holes cut in the aluminum baffles, terminates at the top with an SMA connector and has an exposed, polished cable termination at the bottom. Photons exiting the fiber optic cable are deflected by a 45º polished aluminum mirror in order to fall incident upon the device. 106 Figure 5.5. Photographs of the top of the PPMS testing station probe. (Left) Picture highlighting the ISO KF 16 flange (red) used to create a vacuum seal at the top of the PPMS bore. From the flange the stainless steel rod extends down into the bore and aluminum baffles (purple) act to dissipate heat to the sidewalls of the bore. At the top, the vacuum feed through capped off by a fiber optic coupler (yellow) gives the researcher an access point to pipe light down the bore. (Middle) Picture showing the fiber optic cable vacuum feed through (white) portion of the assembly and where the cryogenic fiber optic cable (magenta) is guided down the length of the stainless steel rod. (Right) Photograph of the fully constructed PPMS solar cells testing station probe. At the top of the bore the cryogenic fiber optic cable interfaces with a commercially available vacuum-compatible fiber-optic feed through. The feed through houses a 400 μm diameter fused silica cable with 200-800 nm operating wavelength and terminates in a male SMA 905 connector at both ends. Vacuum is maintained with a nested o-ring pressed against the underside of the assembly housing and can hold 30 mTorr worth of pressure. The exposed end of the fiber optic feed-through acts as the access point where the researcher can supply light to the solar cell at the base of the PPMS bore. This design choice allows for any excitation source to be used as long as it can be coupled into a fiber optic cable. For our purposes we employed an Oriel Instruments Xe Arc Lamp as a white-light illumination source coupled into an 800 μm fused silica cable through a focusing assembly. A filter housing in the light’s path length allows 107 the excitation source supplied to the device to be broadband or monochromatic per the user’s choice. 5.3.4 Fully Assembled System Generalized Description of Operation Figure 5.6. Photograph of the fully assembled PPMS solar cell testing apparatus. The light delivery and device testing process proceeds as follows: Xe arc lamp (A) generates white light. The lamp output spectrum can either be broadband or attenuated using filters. Generated photons are focused into a fiber optic cable using a commercially available assembly (B) that is married to the fiber optic feed through at the top of the PPMS (C). Light passes through the feed-through and into the cryogenic cable inside the PPMS bore (D) where it is delivered to the solar cell after reflecting off a 45º polished aluminum mirror. Current-voltage testing using a Keithley 2635B Source Meter is carried out through the PPMS automated control software (E). Fully assembled, the custom built PPMS testing station operates as follows: photons produced from the Xe arc lamp are focused into an 800 μm fused silica fiber optic cable. This cable is coupled to the exposed end of the fiber optic feedthrough at the top of the PPMS bore. Photons exiting the 800 μm cable, enter the feedthrough where they are then coupled into the 400 μm Fiberguide cryogenic cable inside the PPMS bore. Light is then fed down the cryogenic fiber where it exits at the 45º degree mirror and is directed toward the device substrate at the bottom of the bore. Here the device is subject to 30 mTorr of inert helium atmosphere and the magnetic 108 field strength (0 ± 16 T) or temperature (1.9 – 400 K) chosen by the user. Electrical contact is made to the device through the brass tabs wired to the universal puck plugged into the 12-pin electrical lead at the base of the PPMS. A Keithley 2635B Source Meter interfaced with the PPMS control software records the electrical response of the device under measuring conditions as specified by the user. 5.4 References (1) http://www.qdusa.com/products/ppms.html (2) http://www.fiberguide.com 109 Chapter 6. Measurement of Organic Photovoltaic Charge Transfer State Energies and Singlet Fission Response Using a Custom Physical Property Measurement System Solar Cell Testing Station 6.1 Variable Temperature Device Testing A molecular interpretation of properties influential to the photovoltage of OPVs has slowly formed over the past two decades of research. Early on it was demonstrated that the magnitude of the open-circuit voltage was only minimally affected by the nature and work function of the electrode contacts. 1,2 It was determined that the D/A heterojunction was more important, specifically, that the cell voltage was dictated by the energy difference between the HOMO of the Donor and LUMO of the Acceptor, the two orbitals responsible for conducting free polarons after charge separation. 3-5 This energy difference, commonly referred to as ∆E DA , has since been established with a clear correlation to the magnitude of V oc . 6 It seems a logical conclusion that an effective avenue to increase cell photovoltages would be to maximize ∆E DA . While effective at increasing voltage, this approach is not without a tradeoff. As ∆E DA increases so too must the optical gaps of the active layer materials in order to maintain an exothermic driving force for photo-induced charge separation. Larger optical gaps forfeit absorption of low energy photons, deleteriously impacting J sc . Therefore, it is clear that a balance between optimized photocurrent and photovoltage must be found for greatest cell efficiency. Recently, experiments have started to elucidate the impact kinetics has on the magnitude of obtainable photovoltages. 7-11 Since (1) 110 it is apparent V oc will be maximized when the reverse saturation current (J s ) is minimized. In forward bias for well-behaved diodes charge recombination occurs bimolecularly (n ≈ 2) and the magnitude of reverse saturation current is dictated by the rate of charge recombination. This recombination is proposed to only occur at the D/A interface, not in the bulk, and proceeds through forming an intermolecular charge transfer state [D + /A - ] that recombines at some rate k rec to form low-energy species ‘D’ and ‘A’ such that E(D), E(A) < E([D + /A - ]). The process is given by: (2) It is straight forward to then conclude that the magnitude of J s is dictated by the kinetics of charge recombination versus charge separation at the D/A interface as governed by Marcus theory for out sphere electron transfer. In the Marcus formulism 12,13 (3) where h is Planck’s constant, H ij is the electronic coupling matrix element between the CT state and the final neutral state species, λ is the reorganization energy, and ∆Gº is the total free energy change of the reaction. From Equation 3 it is apparent that in order for k rec to be depressed, H ij (the electronic coupling of the CT state to ground state) must be minimized. However, coupling of the molecular excited state to CT state manifold must be high to encourage forward charge transfer to generate photocurrent in an OPV. Modern computational methods have been applied to probe these seemingly contradictory possibilities. In a pentacene/C 60 pairing it was found that geometric orientation between the two chromophores affects both ground state to CT coupling 111 and excited state to CT coupling independently. 14 While mutual orientation of molecular species seems like a promising avenue to control relative state coupling strengths, processing conditions used in the deposition of organic materials afford little control over the nanoscale morphology, and therefore orientation. 15-17 Another proposed method of modifying relative state coupling strengths is through minimizing the π-π interaction of the Donor and Acceptor materials. By adding steric bulk to a molecular backbone it is possible to spatially separate the Donor and Acceptor materials, and hence their wavefunctions, to decrease their coupling (Figure 6.1). 11,18,19 In a representative example, the ∆E DA value for a CuPc/C 60 OPV is approximately 1.7 eV (CuPc = copper Figure 6.1. (Top left) Molecular structures of CuPc and PtTPBP highlighting the pendant phenyl groups supplying added steric bulk to the porphyrin core. (Top right) Semi-log IV curves collected under 1 sun illumination (solid) and in the dark (open). OPVs constructed with PtTPBP Donor layers produce less dark current and higher open- circuit voltages than similarly constructed devices employing CuPc as the molecular Donor even though the porphyrin has a smaller ∆E DA gap than the phthalocyanine when paired with C 60 . (Bottom left) OPVs constructed with rubrene produce less dark current and higher open-circuit voltages than similarly constructed devices employing tetracene as the molecular Donor even though the two acenes have similar oxidation potentials. The steric bulk of the four pendant phenyl rings of rubrene act to physically separate the acene backbone from the Acceptor and minimizes recombination current (Bottom right). 18 112 phthalocyanine). For Pt(TPBP)/C 60 the ∆E DA difference is 1.4 eV. From the ∆E DA values it is expected for the porphyrin device to have a lower photovoltage than the phthalocyanine device when paired with the same Acceptor. In reality, the Pt(TPBP)/C 60 cell voltage approaches 0.69 V 18,20 while the CuPc/C 60 cell voltage languishes at 0.48 V. In another example, solar cells with tetracene versus rubrene donor layers exhibit drastically different photovoltages when paired with the same Acceptor (C 60 ), even though the oxidation potential of both acenes is approximately equal (-5.48 eV for rubrene, -5.33 eV for tetracene). The V oc for the rubrene/C 60 cell (V oc = 0.92 V) is markedly larger than the tetracene/C 60 cell (V oc = 0.55 V). In order to understand these pronounced voltage differences in the absence of a satisfying thermodynamic rationalization based on electrochemical potentials, experimental fits to the electrical behavior of the devices were made. As introduced in Chapter 1, an expression for V oc can be obtained: (4) where ∆E is the energy barrier at the interface which must be overcome to generate charge carriers, n is diode ideality factor, k Boltzmann constant, T the absolute temperature, and J 00 a phenomenological term that must be assessed experimentally. J sc is the short circuit current density of the cell. This voltage expression shows well the relationship of V oc to ∆E and an inverse linear relationship expected between V oc and temperature. Fits of experimental J-V curves to Equation 4 reveal J 00 values an order of magnitude larger for tetracene/C 60 versus rubrene/C 60 and three orders of magnitude larger for CuPc/C 60 versus Pt(TPBP)/C 60 . Larger J 00 coupling values correlate to increased recombination current and lower photovoltages. 20 113 Recently, the uniqueness of the tetracene/C 60 versus rubrene/C 60 system was again analyzed, this time using high sensitivity EQE measurements that are able to measure current response from sub-band gap photons. 21 Using a generalized Marcus expression the EQE spectra could be fit to reveal an interfacial CT energy of 1.23 eV for tetracene and 1.48 eV. This time, device modeling gave almost identical values for J 00 for the two donor molecules. Referring back to Equation 4 above, the authors attributed the large difference in V oc values for the two solar cells to the energy of the interfacial CT state between Donor and Acceptor, [D + /A - ]. The energy of the CT state has a large impact on solar cell efficiency as it sets the upper limit of achievable photovoltage 22 since the energies of separated, free polarons cannot exceed the energy of the bound state from which they arise. Measuring the energy of this state, therefore, illuminates the magnitude of radiative and non-radiative losses scavenging voltage from its thermodynamic potential. To this end, the energy of the interfacial CT state has been probed for a variety of organic solar cells, including those with polymeric and small molecular active layers, in both bulk heterojunction and lamellar device architectures. 23-27 Almost Figure 6.2. (Left) Semi-log tetracene/C 60 and rubrene/C 60 IV curves collected under 1 sun illumination (solid) and in the dark (dashed). (Right) EQE spectra for tetracene/C 60 and rubrene/C 60 OPVs with fits to a generalized Marcus expression used to extract the energy of the interfacial charge transfer state for each device. 21 114 ubiquitously, the magnitude of the obtained photovoltage is equal to the CT energy minus some loss, on the order of 0.3 - 0.8 eV, leading to the commonly accepted guideline of: qV oc = E CT – 0.5 ± 0.3 eV (5) The origin of the non-radiative contribution to this loss mechanism has be evaluated using spectroscopically accessible quantities at the Donor-Acceptor interface, including the energy of the CT state, the reorganization energy in forming the CT state by photon absorption, and a parameter ‘f’ that is proportional to the number of CT states in the device as well as the electronic coupling between CT and ground state. Using this approach it was found that the open-circuit voltage measured at room temperature was between 0.5 and 0.6 V less than the CT energy in polymer-fullerene blended solar cells. 23 Measuring the CT state energy directly is non-trivial due to low absorption coefficients and low state densities. Electroluminescence of the CT state can be detected given sensitive enough equipment with detection capabilities in the spectral regions where these states emit (0.5- 1.0 eV). 23-25,27 Commonly, the energy of the CT state is extrapolated, either through a fit to the device external quantum efficiency measured for sub-band gap photons, 26,27 or by measuring the device electrical response as a function of temperature. 23-25 As Equation 4 implies, the open- circuit voltage of the solar cell is expected to increase as a function of decreasing temperature in some linear regime such that qV oc (T  0) = E CT . By plotting the obtained V oc versus temperature the energy of the CT state can be inferred by extrapolating the linear fit back to 0 K. The variable temperature capability of the Physical Property Measurement System solar testing station described in Chapter 5 of this document allows evaluation of E CT in organic photovoltaic 115 devices through this method. Further, the PPMS magnetic field capability allows for simultaneous detecting of singlet fission operant in the device as a function of temperature. 6.2 Extrapolating Charge Transfer State Energy 6.2.1 OPV Device Architecture and Testing Procedure The charge transfer state energy (E CT ) of several molecular lamellar Donor/Acceptor OPVs were extrapolated using variable temperature device testing in the custom-built PPMS solar cell testing station probe assembly. All devices were of the architecture ITO/Donor/Acceptor/Buffer/Al where BCP = bathocuproine and Acceptor = C 60 . CuPc, tetracene, rubrene, DPT, and pentacene were investigated as Donors. Devices were grown on solvent-cleaned patterned ITO substrates (described in Section 5.3.1). All materials were deposited in a high-vacuum chamber at base pressure of (3 x 10 -6 Torr) with individual rates, CuPc (2.0 Å/s), tetracene (20.0 Å/s), rubrene (2.0 Å/s), DPT (2.0 Å/s), pentacene (2.0 Å/s), C 60 (2.0 Å/s), BCP (1.0 Å/s), Al (2.0 Å/s) as monitored by in situ calibrated quartz crystal microbalance assemblies. Device testing procedures were carried out as follows: First, current-voltage measurements were made using a Keithley 2420 source meter in air at 25 ºC in the dark and under 1000 W/m 2 (1 sun) white light illumination from an AM1.5G filtered 300 W xenon arc lamp (Asahi Spectra HAL-320W) to establish the electrical response of the device under standard testing conditions. Then, the device was transferred to the substrate holder of the PPMS testing station probe apparatus, the assembly fitted into the instrument, and the PPMS bore was pumped down to 30 mTorr, with residual pressure supplied from an inert argon atmosphere. Electrical response was measured in the dark and under illumination from photons supplied to the device by an Oreil Instruments Xe Arc Lamp through the fiber optic coupling system as 116 previously described (Section 5.3.4). Measurements were carried out with long (~ 30 mins) dwell time between temperature adjustments to ensure thermal equilibrium of the device. Once PPMS testing was concluded the devices were retested in air under standard testing conditions to probe for changes in the electrical response of the device caused by the testing protocol. 6.2.2 CuPc/C 60 Figure 6.3. (Left panel) I-V curves of CuPc/C 60 devices under 1 sun (1000 W/m 2 ) testing conditions in open air at room temperature (red traces) and inside the PPMS custom testing apparatus using white light supplied to the substrate through the fiber optic cable coupling scheme (black traces). Curves are for two different devices on the same substrate. (Right panel) The same curves zoomed in to accentuate the magnitude of J sc for devices inside the PPMS. Figure 6.3 shows the I-V curves of CuPc (40 nm)/C 60 (40 nm) devices under standard 1sun illumination conditions (red trace) and under white light illumination conditions inside the PPMS (black trace). Both measurements were carried out at room temperature for two different devices on the same substrate. Since the PPMS testing conditions deviate drastically from standardized testing conditions and illumination within the bore is not uniform across all devices, no mention of device efficiencies are made and current response is reported in raw intensity 117 instead of normalized by device area. Under 1 sun conditions the CuPc/C 60 devices produce approximately 100 μA of short-circuit current while inside the PPMS the same devices are only able to produce ~6 μA of short-circuit current. The open-circuit voltage also drops from 0.45 V to 0.33 V. These drastic differences highlight the illumination disparity between 1000 W/m 2 1 sun conditions and the photon intensity able to reach the substrate inside the bore of the PPMS due to losses between fiber optic coupling and poor spatial alignment of the emitted light with the substrate. The open-circuit voltage subsequently drops since V oc ln(J sc /J s ). While illumination conditions inside the PPMS do not approach 1 sun intensity, they can still be used to make temperature dependent device measurements to extrapolate E CT . Figure 6.4. (Left panel) Variable temperature IV curves for CuPc/C 60 OPVs collected in the PPMS solar cell testing apparatus using white light illumination from provided through the fiber optic coupling scheme. The top and bottom graphs represent data from two different devices on the same substrate. (Middle panel) IV curves zoomed in to accentuate the value of the open-circuit voltage (I = 0, black squares) for the devices as a function of changing temperature. (Right panel) Plot of the extracted open-circuit voltages as a function of temperature. The red line represents a linear regression fit to the data of the displayed fitting function. Extrapolation of the fit to T = 0 K gives a representative measure of the charge transfer state energy formed at the D/A interface. Figure 6.4 shows an overlay of I-V curves collected for CuPc/C 60 devices as a function of temperature. It can be seen that as temperature decreases the cell current trends downward, 118 although not monotonically. It can also be seen that as temperature decreases the cell photovoltage trends upward, as predicted by Equation 4. While the supplied white light intensity inside the PPMS solar cell testing apparatus does not adhere to 1 sun illumination conditions, it can be seen that plotting the obtained V oc as a function of temperature still reveals a linear trend that when fitted allows for E CT to be extrapolated from the linear regime as T  0 K. For CuPc/C 60 , values of 1.0 - 1.1 eV were extracted for E CT from the two devices measured. These estimated charge-transfer state energies match well to literature values of 0.92 V and 1.08 V obtained for bulk heterojunction and lamellar device architectures, respectively, of CuPc/C 60 . 24,25 6.2.3 Rubrene/C 60 Figure 6.5. (Left panel) I-V curves of rubrene/C 60 devices under 1 sun (1000 W/m 2 ) testing conditions in open air at room temperature (red traces) and inside the PPMS custom testing apparatus using white light supplied to the substrate through the fiber coupling scheme (black traces). Curves are for two different devices on the same substrate. (Right panel) The same curves zoomed in to accentuate the magnitude of J sc for devices inside the PPMS. The same generalized testing protocol was carried out for the remaining Donor/C 60 devices. Figure 5xx reports the IV curves collected for rubrene (30 nm)/C 60 (25 nm) devices 119 tested under both simulated 1 sun illumination intensity, open air conditions (red trace) and inside the PPMS solar cell testing station apparatus (black trace) for two different devices on the same substrate. As previously witnessed, under 1 sun conditions the rubrene/C 60 devices produce approximately 100 μA of short-circuit current while inside the PPMS the same devices are only able to produce ~5 μA of short-circuit current. The open-circuit voltage also drops from 0.92 V to 0.84 V. 120 Figure 6.6. (Left panel) Variable temperature IV curves for rubrene/C 60 OPVs collected in the PPMS solar cell testing apparatus using white light illumination from provided through the fiber optic coupling scheme. The top and bottom graphs represent data from two different devices on the same substrate. (Middle panel) IV curves zoomed in to accentuate the value of the open-circuit voltage (I = 0, black squares) for the devices as a function of changing temperature. (Right panel) Plot of the extracted open-circuit voltages as a function of temperature. The red line represents a linear regression fit to the data of the displayed fitting function. Extrapolation of the fit to T = 0 K gives a representative measure of the charge transfer state energy formed at the D/A interface. Figure 6.6 shows an overlay of I-V curves collected for rubrene/C 60 devices as a function of temperature. Once again, as temperature decreases the cell current trends downward and the cell photovoltage trends upward, as predicted by Equation 4. Plotting the obtained V oc as a function of temperature reveals a linear trend that when fitted allows for E CT to be extrapolated from the linear regime as T  0 K. The roll-off in voltage witnessed for low temperatures could be from resistances becoming non-negligible and the electrical response deviating from ideal behavior. For rubrene/C 60 , values of 1.40 – 1.50 eV were extracted for E CT from the two devices measured. These estimated charge-transfer state energies match well to literature values of 1.48 eV as measured through sub-band gap EQE device measurements obtained for a lamellar architecture of rubrene/C 60 . 21 121 6.2.4 Tetracene/C 60 and Pentacene/C 60 Figure 6.7. (Left) I-V curves of tetracene/C 60 devices under 1 sun (1000 W/m 2 ) testing conditions in open air at room temperature (red traces) and inside the PPMS custom testing apparatus using white light supplied to the substrate through the fiber coupling scheme (black traces). (Middle) Variable temperature IV curves for tetracene/C 60 OPVs collected in the PPMS solar cell testing apparatus using white light illumination from provided through the fiber optic coupling scheme. (Right) Plot of the extracted open-circuit voltages as a function of temperature. The red line represents a linear regression fit to the data of the displayed fitting function. Extrapolation of the fit to T = 0 K gives a representative measure of the charge transfer state energy formed at the D/A interface. The final two D/A systems investigated were the parent acene Donors tetracene and pentacene, individually paired with C 60 as the molecular Acceptor (Figure 6.7). When tested under 1 sun illumination conditions the tetracene (60 nm)/C 60 (20 nm) cell produces ~100 μA of photocurrent and only ~30 μA at J sc when illuminated in the PPMS solar cell testing station apparatus. As the temperature is lowered in the PPMS testing apparatus, the tetracene/C 60 device produces less short-circuit current and the operating open-circuit voltage systematically increases. Plotting the V oc as a function of temperature and extrapolating the linear fit to the data back to T = 0 K reveals an interfacial charge transfer state energy of 1.12 eV. This value of E CT compares well to literature experiments using sub-bandgap EQE measurements to extract an interfacial charge transfer state energy of 1.23 eV for lamellar tetracene/C 60 OPVs. 21 122 Figure 6.8. (Left) I-V curves of pentacene/C 60 devices under 1 sun (1000 W/m 2 ) testing conditions in open air at room temperature (red traces) and inside the PPMS custom testing apparatus using white light supplied to the substrate through the fiber coupling scheme (black traces). Curves are for two different devices on the same substrate. (Right) The same curves zoomed in to accentuate the magnitude of J sc for devices inside the PPMS. Figure 6.9. (Left) Variable temperature IV curves for pentacene/C 60 OPVs collected in the PPMS solar cell testing apparatus using white light illumination from provided through the fiber optic coupling scheme. The top and bottom graphs represent data from two different devices on the same substrate. (Middle) IV curves zoomed in to accentuate the value of the open-circuit voltage (I = 0, black squares) for the devices as a function of changing temperature. (Right) Plot of the extracted open-circuit voltages as a function of temperature. The red line represents a linear regression fit to the data of the displayed fitting function. Extrapolation of the fit to T = 0 K gives a representative measure of the charge transfer state energy formed at the D/A interface. Finally, pentacene (20 nm)/C 60 (22.5 nm) lamellar devices tested under 1 sun illumination conditions (Figure 6.8) produces ~150 μA of photocurrent and only ~3.5 μA at J sc when illuminated in the PPMS solar cell testing station apparatus. As with every other device, as the temperature is lowered in the PPMS testing apparatus, the pentacene/C 60 device produces less short-circuit current and the operating open-circuit voltage systematically increases (Figure 6.9). Plotting the V oc as a function of temperature and extrapolating the linear fit to the data back to T 123 = 0 K reveals an interfacial charge transfer state energy of 0.76 eV and compares well to 0.89 V obtained from literature results for pentacene/C 60 bilayer devices. 24 6.2.5 Summary It was shown that the PPMS solar cell testing station could be effectively used to measure interfacial charge transfer state energies at lamellar D/A heterojunctions. Table 6.1 summarizes the obtained experimental charge transfer state energy measurements with literature values obtained on comparable devices with a lamellar architecture. The literature values were extracted using either variable temperature device measurements or fits to sub-band gap photoresponse in device EQEs. In all cases the experimental results obtained from the PPMS testing station match excellently with literature results. 21,24 Table 6.1. Device open-circuit voltage metrics obtained for lamellar OPVs with active layers of D/A molecular pairings. Donor and Acceptor layer thickness are given in (nm). Device testing was carried out inside the PPMS testing apparatus with white light supplied to the substrate through the fiber optic coupling scheme, and on a 1 sun (1000 W/m 2 ) open-air solar cell testing station. qVoc denotes the open-circuit voltage of each cell when tested under 1 sun illumination conditions while. E CT denotes the energy of the charge-transfer state as extracted from a linear extrapolation of the temperature-dependent V oc towards T = 0 K from temperature-dependent I-V measurements conducted inside the PPMS. Results are compared to measured to interfacial charge transfer state energies (E CT lit) taken from literature for OPVs of lamellar architecture with the same D/A pairings. 21,24 Table 6.1 also summarizes the experimentally determined E CT , and compares it to the 1 sun intensity open-circuit voltage value for each D/A combination. In every case, the charge transfer state energy is systematically ~ 0.5 – 0.6 eV larger than the devices’ achieved photovoltages, a trend that has been witnessed in both molecular and polymeric systems with Donor/Acceptor qV oc (eV, at 1 sun) E CT (eV) exp E CT (eV) lit CuPc (40 nm)/C 60 (40 nm) 0.44 1.0 – 1.1 0.89, 1.08 Rubrene(30 nm)/C 60 (25 nm) 0.93 1.40 – 1.50 1.48 Tetracene (60 nm)/C 60 (20 nm) 0.68 1.12 1.23 Pentacene (20 nm)/C 60 (22.5 nm) 0.28 0.76 0.89 124 fullerene and non-fullerene Acceptors. While the organic systems investigated here are quite similar to one another (planar, molecular Donors with delocalized π-systems paired with C 60 ), the consistency of the magnitude of energy loss between E CT and V oc over a host of disparate materials implies a common origin and that the relative loss with respect to the photovoltaic gap will be larger for smaller gaps. The origin of these losses arises from charge recombination pathways that do not lead to free carriers. Of this combination, radiative loss pathways are unavoidable, as demanded by detailed balance. The non-radiative recombination, however, is theoretically avoidable and would increase device efficiencies if ameliorated. Quantification of the radiative versus nonradiative components of the recombination loss has shown the nonradiative component to be up to 80% of the loss magnitude, accounting for 0.4 – 0.6 V of photovoltage gains that could be realized if properly addressed. In a final note it should be mentioned DPT/C 60 was another material set investigated using the PPMS testing station for the purposes of extracting E CT . Unlike the other D/A combinations DPT/C 60 shows no clear linear regime with respect to its photovoltage as a function of temperature. This could be due to the resistivity of amorphous DPT films to charge migration at depressed temperatures, structural reorganizations within the film, or inconsistent contact to the device from the PPMS testing station due to thermal contractions around the substrate. Considering the V oc of DPT/C 60 OPVs is ~0.93 V under 1 sun conditions, the charge transfer state energy is estimated to be ~1.43 eV as long as the observation that qV oc = E CT – 0.5 ± 0.3 eV holds. 125 6.3 Probing Singlet Fission Dynamics in Acene OPVs Using the Physical Property Measurement System Solar Cell Testing Station The PPMS testing station was shown to successfully extract OPV charge transfer state energies through variable temperature device testing. Application of an external magnetic field further allows the PPMS testing station assembly the opportunity to inspect the mechanism of singlet fission operant in OPVs in conjunction with variable temperature testing. Since applied magnetic fields depress the rate of SF, more singlet excited states are produced than would otherwise exist in the absence of the field, and the S 1 excited state therefore has a greater opportunity to decay by some other mechanism. For OPVs with SF chromophores, the relative change in the magnitude of cell photocurrent under applied magnetic fields signals the role of fission in the device. Devices with a negative change in photocurrent are effectively harnessing triplets produced from SF while devices with a positive change in photocurrent are not converting triplets into charge (i.e., since the magnetic field produces more singlets, a positive change in photocurrent means the cell photocurrent benefits from the presence of a greater amount of a singlets. When the opposite trend is witnessed the opposite conclusion holds true). Of the D/A pairings explored in Section 6.2 tetracene, rubrene, DPT, and pentacene are all acene materials shown to undergo SF in neat thin films. The devices employing these acenes as molecular Donors in an OPV were also analyzed under magnetic fields of the PPMS testing station assembly to garner insight into the role fission plays in photocurrent production. In this type of experiment it is preferential to use monochromatic excitation to excite only the fission chromophore in the device. If the only photocurrent generation mechanism operating in the device is a SF-modulated channel, the magnitude of the relative current modulation under 126 the magnetic field will be maximized. This can be guaranteed by only exciting the SF chromophore. However, if the amount of current produced by the cell through a SF-modulated channel is small compared to other current producing avenues in the device, then the percent change in current output due to the magnetic field modulation will be small compared to the static, unchanged magnitude of current. Such a circumstance is equivalent to measuring a minor change in signal on top of a large background. Given this, the acene devices are excited with filtered white light to create as monochromatic an excitation source as possible, and fabricated with a thin Acceptor layer since photocurrent generation through C 60 absorption is a non- negligible in OPVs. 6.3.1 Singlet Fission in Tetracene/C 60 OPVs Figure 6.10. Graphs of cell photocurrent as a function of applied magnetic field for tetracene/C 60 devices at 300K (left) 275K (middle) and 250 K (right) illuminated with white light passed through a λ = 500 nm longpass filter and supplied to the device through the fiber optic cable coupling scheme. The dashed red line acts as a guide to the eye to compare how the magnitude of the photocurrent changes with applied magnetic field from its initial zero field value. Figure 6.10 shows the magnetic field modulated photocurrent data collected for a tetracene (60 nm)/C 60 (20 nm) device. For tetracene, the white light excitation source is filtered through a λ = 500 onset longpass filter (designated OG515). The C 60 acceptor layer is ~20 nm thick (compared to ~ 40 nm thickness common for lamellar OPVs). It can be seen that at low fields the measured cell photocurrent is highly variable, fluctuating several percentage points with minimal field application most likely due to unfavorable resistances in the testing apparatus 127 having non-negligible impact on the cell photocurrent at these low light intensity illumination conditions. The operating mechanism of SF in molecular OPVs is determined by evaluating the nature of the current change under applied magnetic field from its zero-field value (photocurrent at 0 Oe). Variable current output at near zero field application makes it difficult to ascertain what the initial photocurrent value should be for comparison purposes. For this reason all experimental data shown here is reported in amperes and is not normalized against an initial value zero-field value. Instead, a low-field current value is chosen for comparisons as graphically accented with a dotted red line. Evaluations of general trends in the current response are made by considering the sign of the signal change compared to this low-field current value without normalization. At 300K and no applied field the tetracene/C 60 device produces ~ -4.7 x 10 -6 A of photocurrent. As the applied magnetic field strength increases, the cell photocurrent decreases and eventually plateaus at ~ -4.2x10 -6 A at 1000 Oe (0.1 T), where it remains approximately constant out to 10,000 Oe (1.0 T). Decreasing photocurrent under applied magnetic field indicates triplets produced during SF in tetracene are contributing to the photocurrent of the device. The same relative behavior is witnessed for the same device tested at 275 K and 250 K. The magnitude of the zero-field photocurrent value decreases as temperature decreases, from ~ - 4.7 x 10 -6 A at 300K, to ~ -3.9 x10 -6 A at 275K, and ~ -3.4x10 -6 A at 250 K. This current decrease could be due to the necessity for thermal activation in tetracene to undergo SF, or from increased resistances to exciton diffusion and charge migration experienced in organic thin films at depressed temperatures. Regardless of the origin, at 275 K and 250 K tetracene/C 60 photocurrent drops from its initial zero-field value under increasing magnetic field. Once again, this conclusion is based off initial zero-field photocurrent values chosen at each temperature. 128 6.3.2 Singlet Fission in Rubrene/C 60 OPVs Figure 6.11. Graphs of cell photocurrent as a function of applied magnetic field for rubrene/C 60 devices at 300K (upper left), 275K (upper right), 250 K (lower left) and 225K (lower right) illuminated with white light passed through a λ = 500 nm longpass filter and supplied to the device through the fiber optic cable coupling scheme. The dashed red line acts as a guide to the eye to compare how the magnitude of the photocurrent changes with applied magnetic field from its initial zero field value. In another example, rubrene (30 nm)/C 60 (17.5 nm) devices tested in the PPMS exhibit magnetic field modulated photocurrent responses similar to those reported in literature. At 300K the cell zero-field photocurrent is approximately ~ -9.3 x 10 -7 A. With minimal applied fields (< 0.1 T) the photocurrent initially decreases by ~2 % to ~ -9.1 x 10 -7 A. Further applied fields increase the photocurrent until it reaches its initial zero-field magnitude at ~0.2 T. Beyond this point greater applied fields see the cell photocurrent increase beyond its initial zero-field value, indicating triplet dissociation at the rubrene/C 60 interface is thermodynamically unfavorable. This result is intuitive considering the energy of rubrene triplets E T = 1.23 eV is lower than the measured charge transfer state energy of a rubrene/C 60 cell, E CT = 1.4 - 1.50 eV as reported in Section 6.2.3. Rubrene triplets are not energetic enough to form the [D + /A - ] charge transfer state mandatory in the photo-electrical conversion process. At 275 K, 250 K, and 225 K this trend in 129 photocurrent response is also witnessed, differing in only the magnitude of the initial zero-field value (current at 0 Oe). 6.3.3 Singlet Fission in DPT/C 60 OPVs Figure 6.12. Graphs of cell photocurrent as a function of applied magnetic field for DPT/C 60 devices at 300K (upper left), 270K (upper right), 240 K (lower left) and 225K (lower right) illuminated with white light passed through a λ = 500 nm longpass filter and supplied to the device through the fiber optic cable coupling scheme. DPT (22.5 nm)/C 60 (22.5 nm) devices tested in the PPMS show no discernible trend between applied magnetic field and cell photocurrent (Figure 6.12). At room temperature (300K) the cell produces ~4.9 x 10 -7 A at 0 Oe applied field, however as the field strength is increased the current deviates between ± 5% out to 4000 Oe where the photocurrent drops -10% to ~ -4.38 x 10 -7 A where it remains out to 1000 Oe. This response does not mirror the magnetic field current response seen for tetracene/C 60 or rubrene/C 60 devices outlined earlier. At 270 K the device photocurrent deviates from its initial zero-field value (~ -7.0 x 10 -7 A) by 10% at 2000 Oe. While the photocurrent drops by similar relative magnitudes at 300 K and 270 K, the change 130 occurs at half the applied field strength when the device held at 270 K. At the lower temperatures the cell photocurrent positively deviates from its initial zero field value, in opposition to the behavior witnessed at 300 K and 270 K. The magnitude of the cell photocurrent is comparable to tetracene/C 60 and rubrene/C 60 , and the measured IV response under 1 sun behaves as expected for a DPT/C 60 device. This suggests that the cell’s electrical response at room temperature behaves as expected and therefore the aberrant electrical response with applied magnetic field is most likely due to the testing station setup and not the device itself. This further explains why the DPT/C 60 temperature-dependent open circuit voltage was highly non-linear. Since the supplied white light is filtered to be as monochromatic as possible the illumination intensity is low. Further, there is still an appreciable amount of C 60 excitation occurring through the longpass filtration. These two facts in concert could conspire to create a condition where the change in magnitude of the photocurrent due to the magnetic field is a small relative to a considerable, unchanging C 60 photoresponse background. Considering DPT/C 60 cell has a V oc = 0.9 V under 1 sun illumination, the energy of the interfacial charge transfer state is expected to be ~1.4 eV, (qV oc = E CT – 0.5 ± 0.3 eV) which is greater than the DPT triplet energy of 1.23 eV. Because of this DPT triplet dissociation at the C 60 interface is expected to be energetically disfavored and the magnitude of the cell’s photocurrent diminished as a function of applied field strength similar to rubrene/C 60 . 131 6.3.4 Singlet Fission in Pentacene/C 60 OPVs Figure 6.13. Graphs of cell photocurrent as a function of applied magnetic field for pentacene/C 60 devices at 300K (upper left), 285K (upper middle), 255 K (upper right), 240 K (lower left), and 225K (lower right) illuminated with white light passed through a λ = 500 nm longpass filter and supplied to the device through the fiber optic cable coupling scheme. Figure 6.13 shows the magnetic field modulated photocurrent data collected for a pentacene (20 nm)/C 60 (22.5 nm) device. For a pentacene Donor, the white light excitation source is filtered through a λ = 600 onset longpass filter and the C 60 acceptor layer is ~22.5 nm thick. It can be seen that at no applied field the cell outputs between 5.5 x 10 -7 A and 7.0 x 10 -7 A of photocurrent across all temperatures investigated. As the field is applied the magnitude of the photocurrent increases for all temperatures. Since an applied magnetic field decreases the rate of singlet fission and increases the relative population of singlet excited states, a corresponding increase in photocurrent implies that triplets are not dissociated to produce free charges in the device. It was previously noted that literature results showed under appropriate testing conditions with proper device construction pentacene/C 60 OPVs are capable of exhibiting external quantum efficiencies over 100 % from pentacene photoresponse (λ = 670 nm). 28 132 Obviously, if an EQE exceeds over 100 %, a carrier multiplication scheme such as SF must be operant in the device. Also, considering our variable temperature device measurements were able to extract the energy of the charge transfer state formed at the pentacene/C 60 interface (E CT = 0.76 eV) and found it to be approximately 0.01 eV lower in energy than the pentacene triplet (E T1 = 0.86 eV), charge transfer from the pentacene T 1 reducing the C 60 molecular Acceptor is expected to be a thermodynamically favorable process. The opposite trend witnessed in the PPMS data most likely results from the relatively thin layer of pentacene (20 nm) being deposited directly onto ITO leading to charge recombination and/or exciton quenching that prohibits triplet excitons from properly dissociating into free charges and being collected as current. Further, the low light intensity of the fiber optic cable illumination transmitted through a λ = 600 nm longpass filter results in a J sc photocurrent three orders of magnitude lower (5.5 x 10 -7 A) than achieved at 1 sun illumination (1.5 x 10 -4 A). While C 60 photoresponse is minimal at these long wavelengths it is still non-zero and therefore the miniscule photocurrent produced is an ad-mix of both pentacene and C 60 response. Application of a magnetic field will modulate only the pentacene photocurrent production and not the fullerene, resulting in a small measured signal change on top of a static background. Taken in concert with probable exciton and charge quenching, it is reasonable to anticipate aberrant device response under these testing conditions. 6.3.5 Summary It was shown that the PPMS solar cell testing station could be effectively used to probe the function of SF in the photocurrent production of acene-Donor molecular OPVs by applying an external magnetic field across the OPV and measuring the percent change in photocurrent away from its initial zero-field value. An increase in photocurrent under applied fields signals triplets produced from SF do not positively add to the photocurrent of the device. Conversely, a 133 decrease in photocurrent under applied fields signals triplets produced from SF add positively to the photocurrent of the device. In a tetracene/C 60 solar cell, fission of singlets produces tetracene triplets with an energy of 1.23 eV. The interfacial charge transfer state energy of this molecular pair was measured to be 1.14 eV (Section 6.2.4) and excited state charge transfer of the tetracene triplet reducing C 60 is expected to be exothermic and energetically favorable. Indeed, magnetic field device measurements shown a decrease in photocurrent as a function of applied field, suggesting tetracene triplets do contribute to OPV photocurrent. Oppositely, a rubrene/C 60 cell was exhibited increased photocurrent under applied magnetic field (i.e. rubrene triplets do not produce photocurrent) even though the rubrene triplet is almost isoenergetic to the tetracene triplet (1.23 eV). The difference between the two device behaviors can be rationalized from the energy of the interfacial charge transfer state energy. While the tetracene/C 60 E CT is lower in energy than the tetracene triplet, the rubrene/C 60 E CT was measured to be 1.40 – 1.50 eV (Section 6.2.3) and thus fullerene reduction from the rubrene triplet is expected to be an energetically disfavored process. Both variable temperature device measurements extracting the pentacene/C 60 charge transfer state energy (0.76 eV, Section 6.2.4) and literature reports of external quantum efficiencies over 100% demonstrate pentacene triplet (E T1 = 0.86 eV) charge transfer to C 60 is exothermic and triplets contribute positively to cell photocurrent. 28 This would be corroborated with magnetic field measurements that saw a decrease in cell photocurrent; however our results follow the opposite trend. This discrepancy is attributed to injurious exciton quenching in the device and a poor signal-to-noise threshold inherent to the low light intensity of the PPMS testing environment. Finally, DPT/C 60 device photocurrent changes exhibited no consistent under applied magnetic field. Further, variable temperature device measurements were not able to elucidate the energy of the DPT/C 60 charge transfer state energy as no linear 134 regime of open circuit voltage with applied temperature was witnessed. The failing of both techniques to resolve the experimental result of interest could be due to the resistivity of highly disordered DPT films being non-negligible contributors to the electrical response of the device (especially at depressed temperatures). Under 1-sun conditions the V oc of DPT/C 60 OPVs is approximately ~0.93 eV. The general trend witnessed for both polymeric and small molecule OPVs is that the energy of the [D + /A - ] charge transfer state is qV oc = E CT – 0.5 ± 0.3 eV. Assumed to be true, this would mean the DPT/C 60 interfacial charge transfer state enegy is ~1.43 eV, a full 200 meV larger than the DPT triplet excited state. Because of this DPT triplets produced from SF are not expected to be energetically competent to reduce C 60 and should not add to cell photocurrent. 6.4 References (1) Tang, C. W.; Two-layer Organic Photovoltaic Cell. Appl Phys Lett. 1986, 48, 183-185. (2) Potscavage, W. J.; Sharma, A.; Kippelen, B. Critical Interfaces in Organic Solar Cells and Their Influence on the Open-Circuit Voltage. Acc. Chem Res. 2009, 42, 1758-1767. (3) Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.; Thompson, M. E. Enhanced Open-Circuit Voltage in Subphthalocyanine/C 60 Organic Photovoltaic Cells, JACS, 2006, 128, 8108-8109. (4) Brabec, C. J.; Cravino, A.; Meissner, D.; Sariciftci, N. S.; Fromherz, T.; Rispens, M. T.; Sanchez, L.; Hummelen, J. C. Origin of the Open Circuit Voltage of Plastic Solar Cells, Adv. Funct. Mater. 2001, 11, 374-380. (5) Rand, B. P.; Burk, D. P.; Forrest, S. R. Offset Energies at Organic Semiconductor Heterojunctions and Their Influence on the Open-Circuit Voltage of Thin-Film Solar Cells. Phys. Rev. B, 2007, 75, 115327. (6) Scharber, M. C.; Wuhlbacher, D.; Koppe, M.; Denk, P.; Waldauf, C.; Heeger, A. J.; Brabec, C. L.; Design Rules for Donors in Bulk-Heterojunction Solar Cells-Towards 10% Energy-Conversion Efficiency. Adv. Mater. 2006, 18, 789-794. 135 (7) Schlenker, C. W.; Thompson, M. E. The Molecular Nature of Photovoltage Losses in Organic Solar Cells. Chem. Commun., 2011, 47, 3702-3716. (8) Koster, L. J. A.; Kemerink, M.; Weik, M. M.; Maturova, K.; Janssen, R. A. J. Quantifying Bimolecular Recombination Losses in Organic Bulk Heterojunction Solar Cells. Adv. Mater. 2011, 23, 1670-1674. (9) Proctor, C. M.; Kuik, M.; Nguyen, T.-Q. Charge Carrier Recombination in Organic Solar Cells, Prog. Poly. Sci., 2013, 38, 1941-1960. (10) Maurano, A.; Hamilton, R.; Shuttle, C. G.; Ballantyne, A. M.; Nelson, J.; O’Regan, B.; Zhang, W.; McCulloch, I.; Azimi, H.; Morana, M.; Brabec, C. J.; Durrant, J. R. Recombination Dynamics as a Key Determinant of Open Circuit Voltage in Organic Bulk Heterojunction Solar Cells: A Comparison of Four Different Donor Polymers. Adv. Mater., 2010, 22, 4987-4992. (11) Erwin, P.; Thompson, M. E.; Elucidating the Interplay Between Dark Current Coupling and Open-Circuit Voltage in Organic Photovoltaics, Appl. Phys. Lett., 2011, 98, 223305. (12) Marcus, R. A. On the Theory of Oxidation-Reduction Reactions Involving Electron Transfer, J. Chem. Phys., 1956, 24, 966-978. (13) Marcus, R. A.; Sutin, N. Electron Transfers in Chemistry and Biology. Biochim. Biophys. Acta, 1985, 811, 265-322. (14) Yi, Y. P.; Coropceanu, V.; Bredas, J. L.Exciton-Dissociation and Charge-Recombination Processes in Pentacene/C 60 Solar Cells: Theoretical Insight into the Impact of Interfacial Geometry. JACS, 2009, 131, 15777-15783. (15) Trinh, C.; Whited, M. T.; Steiner, A.; Tassone, C. T.; Toney, M. F.; Thompson, M. E. Chemical Annealing of Zinc Tetraphenylporphyrin Films: Effects on Film Morphology and Organic Photovoltaic Performance. Chem. Mater., 2012, 24, 2583-2591. (16) Hormann, U.; Lorch, C.; Hinderhofer, A.; Gerlach, A.; Gruber, M.; Kraus, J.; Sykora, B.; Grob, S.; Linderl, T.; Wilke, A.; Opitz, A.; Hansson, R.; Anselmo, A. S.; Ozawa, Y.; Nakayama, Y.; Ishii, H.; Koch, N.; Moons, E.; Schreiber, F.; Brutting, W. V oc from a Morphology Point of View: the Influence of Molecular Orientation on the Open Circuit Voltage of Organic Planar Heterojunction Solar Cells. J. Phys. Chem. C, 2014, 118, 26462-26470. (17) Zimmerman, J. D.; Xiao, X.; Renshaw, C. K.; Wang, S.; Diev, V.; Thompson, M. E.; Forrest, S. R. Independent Control of Bulk and Interfacial Morphologies of Small Molecular Weight Organic Heterojunction Solar Cells, Nano Lett., 2012, 12, 4366-4371. 136 (18) Perez, M. D.; Borek, C.; Forrest, S. R.; Thompson, M. E. Molecular and Morphological Influences on the Open Circuit Voltages of Organic Photovoltaic Devices. JACS, 2009, 131, 9281- 9286. (19) Schlenker, C. W.; Barlier, V. S.; Chin, S. W.; Whited, M. T.; McAnally, R. E.; Forrest, S. R.; Thompson, M. E. Cascade Organic Solar Cells, Chem. Mater. 2011, 23, 4132-4140. (20) Perez, D. M.; Borek, C.; Djurovich, P. I.; Mayo, E. I.; Lunt, R. R.; Forrest, S. R.; Thompson, M. E. Organic Photovoltaics Using Tetraphenylbenzoporphyrin Complexes as Donor Layers. Adv. Mater., 2009, 21, 1517-1520. (21) Graham, K. R.; Erwin, P.; Nordlund, D.; Vandewal, K.; Li, R.; Ndjawa, G. O. N.; Hoke, E. T.; Salleo, A.; Thompson, M. E.; McGehee, M. D.; Amassian, A. Re-Evaluating the Role of Sterics and Electronic Coupling in Determining the Open-Circuit Voltage of Organic Solar Cells. Adv. Mater., 2013, 25, 6076-6082. (22) Vandewal, K.; Tvingstedt, K.; Manca, J. V.; Inganas, O. Charge-Transfer States and Upper Limit of the Open-Circuit Voltage in Polymer:Fullerene Organic Solar Cells. IEEE J. Sel. Top. Quantum Electron. 2010, 16, 1676-1684. (23) Vandewal, K.; Tvingstedt, K.; Gadisa, A.; Inganas, O.; Manca, J. V. Relating the Open- Circuit Voltage to Interface Molecular Properties of Donor:Acceptor Bulk Heterojunction Solar Cells. Phys. Rev. B, 2010, 81, 125204. (24) Hormann, U.; Kraus, J.; Gruber, M.; Schuhmair, C.; Linderl, T.; Grob, S.; Kapfinger, S.; Klein, K.; Stutzman, M.; Krenner, H. J.; Brutting, W. Quantification of Energy Losses in Organic Solar Cells from Temperature-Dependent Device Characteristics. Phys. Rev. B, 2013, 88, 235307. (25) Zou, Y.; Holmes, R. J. Correlation Between the Open-Circuit Voltage and Charge Transfer State Energy in Organic Photovoltaic Cells. ACS Appl. Mater. Interfaces, 2015, 7, 18306-18311. (26) Vandewal, K.; Tvingstedt, K.; Gadisa, A.; Inganas, O.; Manca, J. V. On the Origin of the Open-Circuit Voltage of Polymer-Fullerene Solar Cells. Nat. Mater., 2009, 8, 904-909. (27) Vandewal, K.; Albrecht, S.; Hoke, E. T.; Graham, K. R.; Widmer, J.; Douglas, J. D.; Schubert, M.; Mateker, W. R.; Bloking, J. T.; Burkhard, G. F.; Sellinger, A.; Frechet, J. M. J.; Amassian, A.; Riede, M. K.; McGehee, M. D.; Neher, D.; Salleo, A. Efficient Charge Generation by Relaxed Charge-Transfer States at Organic Interfaces. Nat. Mater., 2014, 13, 63-68. (28) Congreve, D. N.; Lee, J.; Thompson, N. J.; Hontz, E.; Yost, S. R.; Reusswig, P. D.; Bahlke, M. E.; Reineke, S.; Van Voorhis, T.; Baldo, M. A. Baldo External Quantum 137 Efficiency Above 100% in a Singlet-Exciton-Fission-Based Organic Photovoltaic Cell. Science, 2013, 340, 334-337.

Abstract (if available)

Abstract Singlet fission produces low-energy triplet excited states from higher energy absorbed photons through an ultrafast photophysical process wherein two triplets are produced from one initially prepared singlet. Appropriately harnessed in organic photovoltaics (OPV) solar cells, singlet fission can augment power conversion efficiencies beyond that of the thermodynamic Shockley-Quessier limit for single junction devices by limiting thermalization losses. However, a delicate balance must be found as a low-energy triplet may not be energetically competent to separate into free charges at an OPV D/A interface owing to higher-energy charge transfer states. ❧ Amorphous thin films of vapro-deposited 5,12-Diphenyltetracene (DPT) were found to exhibit no X-ray or electron diffraction, contrasting directly with the strong diffraction witnessed for tetracene. Concurrently, DPT thin film absorption and emission show little spectral change from solution measurements whereas tetracene film absorption and emission red-shifts to lower energy wavelengths due to aggregation and inter-chromophore coupling in the solid state. Ultra-fast transient absorption measurements reveal triplet production in DPT thin films over two timescales, approximately 1 ps and 100 ps, faster than inter-system crossing. Fits to the transient decay allow for extraction of DPT triplet population densities. At low excitation densities DPT triplet production approaches 140%, confirming active singlet fission in the thin film.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Singlet fission in covalently-linked tetracenes

PDF

First steps of solar energy conversion; primary charge generation processes in condensed phase coupled chromophores

PDF

Organic photovoltaics: from materials development to device application

PDF

Improving the field of organic photovoltaics through the development of new active layer materials with unique photophysical properties

PDF

Porphyrin based near infrared‐absorbing materials for organic photovoltaics

PDF

Predictive electronic structure methods for strongly correlated systems: method development and applications to singlet fission

PDF

Organic solar cells: molecular electronic processes and device development

PDF

Properties and applications of dipyrrin-based molecules in organic photovoltaics

PDF

Exciton dynamics in photovoltaic materials

PDF

Excited state dynamics of organic photovoltaics probed using ultrafast spectroscopy

PDF

Squaraines and their applications to organic photovoltaics

PDF

Charge separation in transition metal and quantum dot systems

PDF

Capturing the sun: leveraging excited state dynamics

PDF

Molecular and morphological effects on the operational parameters of organic solar cells

PDF

Nanocrystal surface engineering as a route towards improved photovoltaics

PDF

Understanding organic and inorganic semiconductor surfaces by coherent nonlinear spectroscopy

PDF

Molecular and polymeric donors for bulk heterojunction organic solar cells

PDF

Electronic structure analysis of challenging open-shell systems: from excited states of copper oxide anions to exchange-coupling in binuclear copper complexes

PDF

Molecular design of pyridine and quinoline based compounds for organic optoelectronics

PDF

Expanding the library of surface ligands for semiconductor nanocrystal synthesis and photovoltaic applications

Asset Metadata

Creator McAnally, Robert Eric (author)

Core Title Singlet fission in disordered acene films: from photophysics to organic photovoltaics

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Chemistry

Publication Date 08/03/2016

Defense Date 07/28/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag acene,energy,OAI-PMH Harvest,organic photovoltaic,photophysics,singlet fission,solar cell

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Thompson, Mark E. (committee chair), Bradforth, Stephen (committee member), Haas, Stephen (committee member)

Creator Email mcanally@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-295780

Unique identifier UC11279623

Identifier etd-McAnallyRo-4726.pdf (filename),usctheses-c40-295780 (legacy record id)

Legacy Identifier etd-McAnallyRo-4726.pdf

Dmrecord 295780

Document Type Dissertation

Format application/pdf (imt)

Rights McAnally, Robert Eric

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA