Optical and magneto-optical studies of unconventional semiconductors for optoelectronic applications

We measured resonant Raman scattering (RRS) and doping-induced absorption (DIA) in pristine films of π-conjugated donor-acceptor (D-A) copolymers, as well as the photoinduced absorption (PIA) spectra of their blend with fullerene PCBM molecules. We found that dominant charge carriers in these copolymers are polaron excitations. In addition, we also found that the D-A copolymer chains contain strongly coupled vibrational modes having relatively strong Raman scattering intensity. The lower energy-induced polaron absorption band overlaps with the renormalized vibrational modes; they appear as antiresonance lines in both PIA and DIA spectra. We show that the RRS, DIA, and PIA spectra of PTB7 (fluorinated poly-thienothiophene-benzodithiophene) are well explained by the amplitude mode model. We also studied magnetic field effect in (D-A) type TADF compounds based organic light emitting diodes (OLED). Magneto-electroluminescence (MEL) and magnetophotoluminescence (MPL) in thin films of these compounds are enhanced thermally, and the response is interpreted as due to the Δg mechanism. TADF-based OLED doped with fluorescent emitters with various concentrations were also investigated. We found that both MEL and MPL responses are thermally activated with substantially lower activation energy compared to the pristine D-A TADF host blend. However, both MPL and MEL steeply decrease with the emitter's concentration indicating the existence of a loss mechanism in iv the OLED device associated with energy transfer directly into the nonemissive triplet level of the emitter. Using optical spectroscopies such as electroabsorption (EA) and PIA, we studied the primary (excitons) and long-lived (free carriers) photoexcitations in thin films of 2D lead perovskite, namely (C6H5C2H4NH3)2PbI4, which form natural "multiple quantum wells" having strong spin-orbit coupling that may lead to "Rashba-splitting" in the electron bands. From the EA spectrum, we found that the exciton binding energy is 190 meV for the 1s exciton, whereas the continuum shows Franz-Keldysh oscillation that unambiguously reveals the band-edge energy. We found a strong PIA band at 0.15 eV that is due to long-lived free carrier absorption, caused by the Rashba-splitting in this material. We obtained a Rashba-splitting energy of (40 ± 5) meV and Rashba parameter of (1.6 ± 0.1) eV·Å in this compound.

OPTICAL AND MAGNETO-OPTICAL STUDIES OF UNCONVENTIONAL SEMICONDUCTORS FOR OPTOELECTRONIC APPLICATIONS by Sangita Baniya A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Physics Department of Physics and Astronomy The University of Utah December 2018 Copyright © Sangita Baniya 2018 All Rights Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL STATEMENT OF DISSERTATION APPROVAL The following faculty members served as the supervisory committee chair and members for the dissertation of Sangita Baniya Dates at right indicate the members’ approval of the dissertation. Zeev Valentine Vardeny , Chair 05/01/2018 Date Christoph Boehme , Member 05/01/2018 Date Michael D. Morse , Member 05/01/2018 Date Yan Sarah Li , Member 05/01/2018 Date Yong Si Wu , Member Date The dissertation has also been approved by Chair of the Department of and by David B. Kieda, Dean of The Graduate School. Peter Trapa Physics and Astronomy ABSTRACT We measured resonant Raman scattering (RRS) and doping-induced absorption (DIA) in pristine films of π-conjugated donor-acceptor (D-A) copolymers, as well as the photoinduced absorption (PIA) spectra of their blend with fullerene PCBM molecules. We found that dominant charge carriers in these copolymers are polaron excitations. In addition, we also found that the D-A copolymer chains contain strongly coupled vibrational modes having relatively strong Raman scattering intensity. The lower energy-induced polaron absorption band overlaps with the renormalized vibrational modes; they appear as antiresonance lines in both PIA and DIA spectra. We show that the RRS, DIA, and PIA spectra of PTB7 (fluorinated poly-thienothiophene-benzodithiophene) are well explained by the amplitude mode model. We also studied magnetic field effect in (D-A) type TADF compounds based organic light emitting diodes (OLED). Magneto-electroluminescence (MEL) and magnetophotoluminescence (MPL) in thin films of these compounds are enhanced thermally, and the response is interpreted as due to the Δg mechanism. TADF-based OLED doped with fluorescent emitters with various concentrations were also investigated. We found that both MEL and MPL responses are thermally activated with substantially lower activation energy compared to the pristine D–A TADF host blend. However, both MPL and MEL steeply decrease with the emitter’s concentration indicating the existence of a loss mechanism in the OLED device associated with energy transfer directly into the nonemissive triplet level of the emitter. Using optical spectroscopies such as electroabsorption (EA) and PIA, we studied the primary (excitons) and long-lived (free carriers) photoexcitations in thin films of 2D lead perovskite, namely (C6H5C2H4NH3)2PbI4, which form natural “multiple quantum wells” having strong spin-orbit coupling that may lead to “Rashba-splitting” in the electron bands. From the EA spectrum, we found that the exciton binding energy is 190 meV for the 1s exciton, whereas the continuum shows Franz-Keldysh oscillation that unambiguously reveals the band-edge energy. We found a strong PIA band at 0.15 eV that is due to long-lived free carrier absorption, caused by the Rashba-splitting in this material. We obtained a Rashba-splitting energy of (40 ± 5) meV and Rashba parameter of (1.6 ± 0.1) eV·Å in this compound. iv To my Parents, my Husband, and my Sons TABLE OF CONTENTS ABSTRACT ....................................................................................................................... iii LIST OF TABLES ............................................................................................................. ix LIST OF ACRONYMS ...................................................................................................... x ACKNOWLEDGEMENTS .............................................................................................. xii Chapters 1. INTRODUCTION .......................................................................................................... 1 1.1 π-Conjugated Polymers ............................................................................................ 1 1.1.1 Excitation Models for π-Conjugated Polymers ...............................................4 1.1.2 Major Excitations in the Class of π-Conjugated Polymers ............................ 6 1.1.2.1 Excitons........................................................................................... 6 1.1.2.2 Polaron .......................................................................................... 10 1.2 TADF Organic Light Emitting Diodes .................................................................. 10 1.2.1 Fluorescence ................................................................................................ 12 1.2.2 Phosphorescence .......................................................................................... 13 1.2.3 Thermally Activated Delayed Fluorescence ................................................ 14 1.2.3.1 Reverse Intersystem Crossing (RISC) .......................................... 15 1.2.3.2 The Singlet-Triplet Energy Splitting (ΔEST) ............................... 16 1.2.3.3 D-A Molecular System for TADF emission ................................. 17 1.2.3.3.1 Intermolecular D-A System ........................................... 17 1.2.3.3.2 Intramolecular D-A Systems.......................................... 19 1.2.3.4 Energy Transfer Processes ............................................................ 21 1.2.3.5 Magnetic Field Effect ................................................................... 22 1.3 Hybrid Organic-Inorganic Perovskites ................................................................... 23 1.3.1 Two-Dimensional (2D) Perovskites Crystal Structure ................................ 26 1.3.2 Photoexcitation in Hybrid Perovskites......................................................... 27 1.3.2.1 Free Carrier Absorption (FCA) ..................................................... 28 1.3.2.2 Exciton in 2D Perovskites ............................................................. 30 1.3.3 Rashba Effect in Hybrid Perovskites ........................................................... 32 1.4 Outline of Dissertation............................................................................................ 34 2. EXPERIMENTAL TECHNIQUES ...............................................................................37 2.1 Materials ................................................................................................................. 37 2.1.1 π-Conjugated Polymers ............................................................................... 37 2.1.2 TADF Materials ........................................................................................... 38 2.1.3 Hybrid Organic-Inorganic Perovskite ......................................................... 38 2.1.4 Thin Film Fabrication .................................................................................. 40 2.1.4.1 One-Step Solution Process ........................................................... 40 2.1.4.2 Co-deposition Process ................................................................. 41 2.2 Linear Absorption and Photoluminescence Measurement ..................................... 42 2.3 Vibration Spectroscopy .......................................................................................... 44 2.3.1 Fourier Transform Infrared (FTIR) Absorption .......................................... 44 2.3.2 Resonant Raman Scattering (RRS).............................................................. 46 2.4 Continuous Wave Photoinduced Absorption ......................................................... 48 2.4.1 Kinetic Analysis .......................................................................................... 53 2.4.1.1 Monomolecular Recombination ................................................... 53 2.4.1.2 Bimolecular Recombination ......................................................... 54 2.5 Doping-induced Absorption (DIA) ........................................................................ 55 2.6 Electroabsorption (EA) ........................................................................................... 55 2.7 Device Fabrication .................................................................................................. 58 2.8 OLED Characterization .......................................................................................... 59 3. AMPLITUDE MODES SPECTROSCOPY OF CHARGE EXCITATIONS IN PTB7 πCONJUGATED DONOR-ACCEPTOR COPOLYMER FOR PHOTOVOLTAIC APPLICATIONS.......................................................................................................... 61 3.1 Introduction ............................................................................................................ 62 3.2 Experimental........................................................................................................... 66 3.3 Results and Discussion .......................................................................................... 67 3.3.1 Analysis Using the Amplitude Mode Model ............................................... 74 3.4 Conclusions ........................................................................................................... 80 4. MAGNETIC FIELD EFFECT IN ORGANIC LIGHT EMITTING DIODES BASED ON ELECTRON DONOR-ACCEPTOR EXCIPLEX CHROMOPHORES DOPED WITH FLUORESCENT EMITTERS ..................................................................................... 81 4.1 Introduction ............................................................................................................ 82 4.2 Experimental Section.............................................................................................. 85 4.3 Results and Discussion ........................................................................................... 86 4.3.1 Magneto-Photoluminescence Studies.......................................................... 86 4.3.2 Magneto-Electroluminescence Studies ....................................................... 92 4.5 Conclusions ............................................................................................................ 98 5.OPTICAL STUDIES OF TWO-DIMENSIONAL LAYERED HALIDE HYBRID PEROVSKITE SEMICONDUCTORS ...................................................................... 100 5.1 Introduction .......................................................................................................... 101 5.2 Experimental ........................................................................................................ 102 vii 5.2.1 Sample Preparation..................................................................................... 102 5.2.2 Electroabsorption (EA) Measurements ...................................................... 102 5.2.3 CW Photomodulation (PM) Measurements ............................................... 103 5.3 Results and Discussion ......................................................................................... 103 5.3.1 Electroabsorption Spectra........................................................................... 107 5.3.2 Free Carrier Absorption(FCA) ................................................................... 116 5.3.2.1 Fitting the FCA and Transient PA1 Band ................................... 122 5.3.2.2 Density Functional Theory (DFT) Calculations ..................... 123 5.3.2.3 Inter-Rashba Optical Transitions in 2D Perovskites .................. 127 5.4 Conclusions .......................................................................................................... 129 6. CONCLUSIONS......................................................................................................... 131 6.1 Conclusions .......................................................................................................... 131 6.2 Future Works ........................................................................................................ 133 REFERENCES ............................................................................................................... 135 viii LIST OF TABLES 2.1: Detector and filter combinations used for measuring PA spectrum in a broad spectral range. ..................................................................................................................................50 3.1: The best fitting parameters for the bare phonon propagator, D0(ω) that describes the six most strongly coupled vibrational modes in PTB7. ω are the bare frequencies, λn/λ are their relative e-p coupling, and δn are their natural width used to fit the RRS spectrum. ..79 4.1: The room temperature MEL max and E act extracted from MEL( T ); room temperature MPL max of exciplex host band, MPL max of emitter PL band, E act extracted from MPL( T ) of exciplex host band and E act extracted from MPL( T ) of the emitter PL band, at various DBP emitter concentrations. .................................................................................92 LIST OF ACRONYMS HOMO Highest Occupied Molecular Orbital LUMO Lowest Unoccupied Molecular Orbital TADF Thermally Activated Delayed Fluorescence HFI Hyperfine Interaction SOC Spin Orbit Coupling RISC Reverse Intersystem Crossing PIA Photoinduced Absorption DIA Doping-Induced Absorption FTIR Fourier Transform Infrared RRS Resonant Raman Scattering D-A Donor-Acceptor EA Electro Absorption FCA Free Carrier Absorption ISC Intersystem Crossing PL Photo Luminescence EL Electro Luminescence 2D Two-Dimensional MFE Magnetic Field Effect AMM Amplitude Mode Model OLED Organic Light Emitting Diodes OPV Organic Photo Voltaic OSV Organic Spin Valve PF Prompt Fluorescence DF Delayed Fluorescence IQE Internal Quantum Efficiency CW Continuous Wave KBr Potassium Bromide O.D. Optical Density CB Conduction Band GS Ground State SE Singlet Exciton TE Triplet Exciton PCE Power Conversion Efficiency AC Alternating Current xi ACKNOWLEDGEMENTS I would like to express my deepest gratitude and heartfelt thanks to my honorable advisor, Prof. Zeev Valy Vardeny, for his exemplary guidance, constant encouragement, care and support through all the years of my Ph.D program. The research opportunity provided by Prof. Vardeny is invaluable to me. I was so fortunate to experience and learn from his knowledge, patience, and critical thinking that led to the completion of this study. I also express my sincere gratitude to my supervisory committee members, Professors Christoph Boehme, Wong Shi Wu, Michael Morse, and Sarah Li, for their support and valuable suggestions. I need to also thank Matt Delong and Zhiheng Liu for their priceless technical instructions and assistance. I would like to thank Dr. Tek Basel for introducing me to the Continuous Wave optics and fabrication of OLED devices used for most of the measurements in this work. My sincere gratitude to Dr. Eitan Ehrenfreund, Dr. Dali Sun, Dr. Chuang Zhang, Dr. ChuanXiang Sheng, Dr. Ajay Nahata, Dr. Luisa Whittaker, Dr. Ryan Maclaughlin, Dr. Yaxin Zhai, Shai R. Vardeny, Ashish Chanana, and Eric Amerling for their valuable suggestions, support, and collaboration. My thanks is extended to my current group members, Dr. Evan Lafalce, Dr. Haoliang Liu, Dr. Jane Wang, Dr. Uyen Huynh, Matt Groesbeck, Xiaojie Liu, Qingji Zeng, Dipak Khanal, Xin Pan and Jaspal Singh, for their collaborating ideas, motivation, experience and interaction. I would like to thank my husband, Dr. Bijaya Thapa, for his continuous support, love, care, and encouragement during my studies and thanks to my lovely son Reyansh Thapa who always relieved my tiredness with his sweet smile and hug whenever I got back home from my work. I am indebted to my parents and family members for their love, patience, and inspiration with their best wishes through all these years of my Ph.D. xiii CHAPTER 1 INTRODUCTION 1.1. π-Conjugated Polymers In 1977, Alan J. Heeger, Alan MacDiarmid, and Hideki Shirakawa received the 2000 Nobel Prize in Chemistry “for the discovery and development of conductive polymers”. Before their discovery, all known polymers were associated with electrical insulating behavior. To date, inorganic semiconductors such as silicon are most popular for electronics because of their high carrier mobility and device stability. However, the production of pure silicon is expensive since it occurs exclusively in oxidized states and its chemical reduction requires a large amount of energy [1]. Meanwhile, fabrication of electronic devices out of inorganic semiconductors requires multiple etching, deposition, and lithographic steps. Conducting polymers have attracted interest for a number of novel optoelectronics application such as organic light emitting diodes (OLED) [2,3], organic photovoltaics (OPV) [4,5], organic field effect transistors (OFET), organic spin valves (OSV) [6,7] and biological sensors because of their unique properties such as mechanical flexibility, solubility, easily tunable bandgap, low cost, light weight and relatively inexpensive fabrication process. 2 The building blocks of organic semiconductors are carbon-containing atoms that have 4 valence electrons: one s-electron and three p-electrons. Organic semiconductors are divided into two types, namely, small molecules and polymers. Small molecules are materials that have very well-defined molecular structure and weight, usually less than 1000g/mol, which can be deposited by thermal evaporation. On the other hand, polymers consist of varying-length chains of repeating molecular units with molecular weight larger than 1000g/mol, which are soluble and can be deposited easily. π-conjugated semiconductors are unsaturated carbon compounds with alternating single and double bonds between the carbon atoms, as shown in Figure 1.1. Usually, in organic semiconductors, if all the valence electrons are singly bonded (sp3 hybridization), they form strong σ-bonding leading to large bandgap energies (approximately 5eV) rendering those insulators. However, in the π-conjugated semiconductors, there are double bonds due to sp2 hybridization causing three electrons to establish strong planar σbonds with neighboring atoms and one electron to be bound in a π-bond perpendicular to the polymer backbone. The π-electrons are delocalized over many carbon atoms along the chain, giving the relatively high conducting properties, and are responsible for the electronic and optical properties of the polymer [8]. Since the π-bond between two carbon atoms results from the linear superposition of the wave functions of the two π -electrons in the pz orbital of each carbon atom that leads to the formation of two states, a low energy state where the probability of finding an electron between nuclei is high, called bonding π– orbitals, and a high energy state where the probability of finding an electron between nuclei is low, called antibonding π*–orbitals. The bonding π–orbitals form the highest occupied molecular orbitals (HOMO) and antibonding π*–orbitals form the lowest unoccupied 3 Figure 1.1. Electronic orbitals and bonds in carbon atoms (a) sp2 hybridization in excited states. (b) A conjugated backbone with overlapped Pz orbitals. Adapted from www.orgworld.de. (c) Chemical structure of trans-polyacetylene; a π-conjugated polymer showing alternation of carbon-carbon single and double bonds. molecular orbitals (LUMO), which are roughly equivalent to the inorganic semiconductor’s valence and conduction band-edges, respectively. The π-electrons are delocalized over many carbon atoms over the chain and hence, the quantum mechanical wave function is confined to a single chain. π -conjugated organic semiconductors are often treated as one-dimensional systems with half-filled electronic bands as there is one πelectron per carbon atom. By taking an account of either electron-phonon interaction or 4 electron-electron interactions among the π-electrons, the formation of the bandgap can be explained. Changing the extent of delocalization, the gap between bonding and antibonding orbitals can be altered, which makes them promising for optoelectronic applications. The energy gap between HOMO and LUMO lies in the range 1.4-3.0 eV in most of the π -conjugated organic semiconductors. 1.1.1 Excitation Models for π-Conjugated Polymers Several models have been developed to explain the excitations in π-conjugated polymers. Assuming the electron–electron interaction is negligible, Su, Schrieffer, and Heeger proposed a model, named the SSH model, for trans-polyacetylene (t-(CH)x), based on a tight-binding approximation calculation by taking an account of electron-phonon interaction and a restoring energy [9]. In this model, we consider a π-conjugated chain described by Hamiltonian, HSSH, in the semiclassical Huckel approximation. As the Huckel model only considers the nearest neighbor interactions based on the tight binding approximation [8], it can explain the bandgap of the polymer but cannot explain the excitation energy levels. The Hamiltonian HSSH contains the lattice kinetic energy, which is treated classically, and the electron-phonon interaction, which is treated quantum mechanically, as shown in equation (1.1): = ∑ − + ∑ , , + −∑ , , , + − 1.1 where t0 is the hopping integral between the nearest neighbors for an undistorted chain, α is the electron lattice coupling constant, , / , are the creation / annihilation operators of an electron on site n with spin s, k is the spring constant due to π-electrons, 5 and un is the deviation of nth site from the equilibrium position in an undistorted chain with equal distance between sites. According to the SSH model, dimerization caused by strong electron-phonon interaction lowers the system energy and creates an energy gap at Fermi level Eg=2Δ where Δ= 4αu and u is the dimerization amplitude in equilibrium. It can be seen that the gap size is directly proportional to the phonon coupling constant α, i.e., the stronger is the coupling, the larger is the gap. Since, dimerization lowers the system energy, occupied electronic states in equilibrium are lowered, resulting in a more stable configuration where the lattice constant is doubled. Therefore, the system no longer acts as a one-dimensional metal, but instead behaves as a semiconductor with a direct energy gap. On the other hand, the Hubbard model can also explain the energy levels of charged and neutral excitations by including Coulomb repulsion of two electrons on the same site. Hubbard Hamiltonian is: $$ = % & '(,↑ '(,↓ ( 1.2 In this model, the electron-electron interaction and 3D intrachain coupling are included, but the electron-phonon interaction is ignored, even though this interaction is quite strong in the polymer system. The SSH model, on the other hand, ignores Coulomb repulsion. So, the model that includes both interactions, i.e., combination of SSH and the Hubbard model, is more realistic to explain the energy levels of excitations in the class of π-conjugated polymers. Such a model is the Pariser-Parr-Pople (PPP) model [10]. This model can explain many aspects of the excited states in polymer systems, and was also used to explain the energy levels in fullerene molecules and carbon nanotubes. 6 1.1.2 Major Excitations in the Class of π-Conjugated Polymers Two kinds of electronic excited states (excitations) are dominant in πconjugated polymers, namely charged (polarons) and neutral (excitons). When a photon with energy higher than the bandgap is absorbed in the π-conjugated polymers, neutral, spinless (S=0) excitations called singlet excitons (SE) are generated. The SE may either radiatively recombine; or convert into long-lived neutral excitations, i.e., triplet excitons (TE) via intersystem crossing; or dissociate into positive and negative charge excitations called polarons, some of which may form long-lived polaron pairs (neutral excitations). Upon electrical excitation (i.e., current injection), charged excitations are injected. These may recombine to form neutral excitations that later recombine to form singlet and triplet excitons. Here we discuss these two types of electronic excitations. 1.1.2.1 Excitons When an electron is promoted from lower energy level (HOMO) to higher energy level (LUMO) upon photon absorption, electron hole pairs called excitons, which are bound through their mutual Coulombic interaction, are formed. The excitons have lower energy than the bandgap of a material, and the difference is defined as the exciton binding energy, whose value is in between 0.3-0.5 eV in most π-conjugated polymers. There are three types of excitons depending on their spatial states, namely Frenkel excitons, chargetransfer excitons, and Wannier-Mott excitons [11,12]. Out of these three, Frenkel excitons are common in organic semiconductors since they have small dielectric constants and the Coulomb interaction between the electron and the hole is strong. These kind of excitons have large binding energies of ~1eV and localize in a small radius 7 comparable to the size of unit cell. On the other hand, Wannier-Mott excitons have much larger radii of 40-100 Å and can delocalize over many atoms or molecules. The dielectric constant in inorganic semiconductors is generally large, and the electric field screening tends to reduce the Coulomb interaction between electrons and holes, which results in the Wannier exciton. Considering that the effective mass of electrons in semiconductors is usually small, the binding energy of a Wannier exciton is significantly reduced and typically on the order of ~50 meV. The excitons in a 3D semiconductor can be described as hydrogen-like states [13]. ,-. '$ , / = ,0 − 12∗ 1 ℏ/ + '4 26 1.3 in which the principle quantum number '8, exciton Rydberg energy 12∗ , translational mass M, wave vector K, and reduced exciton mass 9 are: '4 = 1,2,3 … 1.4 6 = ?- + ?B , / = C- + CB (1.6) 12∗ = 13.6=> 9= 9 ? @ ?- ?B ?- + ?B Therefore, the oscillator strength f and Bohr radius E8 are given by F '4 ∝ '4 H , E4 ∝ E4 '4 1.5 1.7 1.8 The spin states of the excited excitons can be described by quantum mechanics. Depending on the spin statistics, an electron and hole in an exciton may form a singlet or a triplet state with total spin 0 or 1, respectively; both excitations still remain neutral. The wave-function describing these two particle systems (excitons) is asymmetric in spin and electronic coordinates and can be obtained from the Slater determinant: 8 J= 1 J( M N( O L √2 JQ M NQ O J( M′ N( O′ L JQ M′ NQ O′ 1.9 where J( M and N( O are the spatial and the spin part of the wavefunction, respectively. We can get spin S=0 (antisymmetric singlet states) and S=1, (symmetric triplet state) wavefunctions. The wave-functions that have a different total spin quantum number, S, constructed from the above equation are: J S J S J S J S = 1 TJ 1 J 2 + J 1 J 2 U T↑ 1 ↓ 2 −↑ 2 ↓ 1 U 2 1.10 = 1 TJ 1 J 2 + J 1 J 2 U T↑ 1 ↑ 2 U 2 1.12 = 1 TJ 1 J 2 − J 1 J 2 U T↑ 1 ↓ 2 +↑ 2 ↓ 1 U 2 = TJ 1 J 2 + J 1 J 2 U T↓ 1 ↓ 2 U 1.11 1.13 where ↑ and ↓ symbols represent the spin projection of χ as up and down, respectively. Singlet and triplet energy levels are degenerate in the noninteracting case. However, in the presence of spin-spin interaction such as an exchange interaction, they are nondegenerate with the triplet level taking the lower energy. The energy bands in excitons are shown in the left panel of Figure 1.2. In the energy band diagram, 1Ag is the ground state and 1Bu, mAg, and kAg are the excited singlet states, where g stands for gerade (even parity) and u stands for ungerade (odd parity). There is radiative recombination or photoluminescence (PL) from singlet state to the ground state that occurs immediately after photoexcitation, which is a fast process on the order of 100 picoseconds (ps). However, still there is a possible chance to convert a singlet state into a long-lived triplet exciton within ~10 ns or less via intersystem-crossing that results by different mechanism such as hyperfine interaction, spin flip of one of the electrons 9 Figure 1.2. Various photo-excitations in conjugated polymers: exciton bands on the left and polaron excitation with P1 and P2 transitions on the right. The triplet state has a lower energy than the singlet state. We only can see long lived photo-excitations with the lifetime of the order of 1ms such as absorption due to triplet exciton (T1 and T2) and polaron: P1 and P2 transitions with CW photoinduced absorption. However, we cannot detect PA1 and PA2 transitions in the singlet manifold because of their short lifetimes so ultrafast pump probe is used. involved in the exciton due to spin orbit coupling, or the existence of radical impurities on the polymer chains. Decay from the triplet exciton to the ground state is spin forbidden but can radiatively decay through phosphorescence (PH) in the presence of a spin flipping mechanism such as spin-orbit interactions; this transition typically has longer lifetime, on the order of milliseconds [14]. Phosphorescence is not a general case in organic materials, so we have to use the photoinduced absorption technique to monitor the density of the triplet excitons in the material. Hence, photoinduced absorption is the transition from Bu to higher lying Ag states, which are shown in the Figure 1.2 by the vertical arrows. Since 10 singlet excitons are short lived, they can be observed by ultrafast pump-probe techniques, but triplet excitons are long-lived species and thus can be probed with CW pump-probe techniques. 1.1.2.2 Polaron Organic materials have less rigid structure than inorganic materials such as silicon. This is because of Van der Waals forces between the neighboring molecules in organic materials in solid state, which are much weaker than the covalent and ionic bonds in inorganic materials. Consequently, a moving charge carrier that propagates in an organic material can cause local distortion and form a charged quasi-particle with nonzero spin called polaron. A polaron can be charged positively (P+) or negatively (P-) and has spin ½. It has two symmetrical, localized states within the gap and has two allowed belowgap optical transitions P1 and P2, as shown at the right in Figure 1.2. Polaron move from chain to chain through hopping between the localized states, and they are the major charge/current carriers in organic device applications such as organic photovoltaic devices. They can be created by various methods, such as chemically doping, photo-doping, and electrically carrier injection through metallic electrodes. Polaron transitions can be observed by the CW pump-probe method. We can identify them by Doping-Induced Absorption (DIA) and Magneto-Photoinduced Absorption (MPA). 1.2 TADF Organic Light Emitting Diodes Organic light emitting diodes (OLEDs) have attracted interest for over the last couple of decades both in academia and industries. Interest in OLED was initiated by Tang 11 and Vanslyke in 1987 [15] where they used Alq3 as the organic layer. Later, polymer LEDs were introduced [3] and consistent efforts aimed to optimize the device working condition and maximize electroluminescence quantum efficiency. These materials are ideal candidates for display and solid state lightning technologies due to their unique qualities. Electronic companies such as Apple, Samsung, LG, and Panasonic are utilizing OLEDs in their electronic devices such as tablet, smart watches, digital cameras, and television. OLEDs are electroluminescent devices consisting of an emitting layer sandwiched between two electrodes (an anode and a cathode). A typical device structure and the working processes are shown in Figure 1.3. Materials for anode, cathode, and carrier transport layers are selected by matching the HOMO/LUMO of the semiconductor and the work-function of the metal electrode. Holes injected from the anode and electrons from the cathode are transported through the organic material under the forward bias voltage, V, with the positive terminal connected to the anode and negative terminal to cathode. These charges in the organic material move towards the center of the device until they recombine, in which case they may form the following three different species: (i) polaron pairs formed by oppositely charged carriers in adjacent chains bounded by Coulombic interaction, (ii) bipolarons formed by pairs of the same charge in the segment of the chain, and (iii) excitons formed by the two closely separated oppositely charged carriers on the same chain due to Coulombic interactions. Singlet excitons within the Coulomb capture radius rc, given by equation (1.14) will recombine and the excited states of excitons decay through the emission of photons that exit through the semitransparent anode or cathode. MW = X 4YZ[ Z /\ 1.14 12 Figure 1.3. Organic light emitting diode. (a) Device structure. (b) Working scheme: 1. Charge carrier injection. 2. Charge carrier transport. 3. e-h pair, and then exciton formation. 4. Exciton decay (light emission). The EL internal quantum efficiency, ηint, of the OLED is the product of the three factors [16]: ]( = ] ] ]H 1.15 where η1 is the fraction of excitons formed to the number of carriers injected, η2 is the fraction of radiative singlet excitons to the total number of excitons, and η3 is the quantum efficiency of fluorescence. EL external quantum efficiency (ELQE) is the product of ηint and the light output coupling factor, ηph~20%. If all and only the singlet excitons are emissive, ηint approaches 25% and ELQE approaches 5%. Phosphorescent OLEDs [17] and delayed fluorescence [18] are intensively studied to harvest both singlets and triplets so that ηint reaches 100% and the ELQE limit increases to 20%. 1.2.1 Fluorescence The charge recombination statistics generates 1:3 singlet S1 to triplet T1 excitons ratio [19] in OLEDs, i.e., 25% of all the excitons created in the organic semiconductor 13 reside in the singlet state of the material. Radiative emission from singlet excitons with zero spin occurs when singlet excited states decay to ground states S0. Nonradiative relaxation of excitons is also possible, which leads to the generation of heat. In the fluorescent materials, (∆EST) is as large as ~0.7 eV, so the contribution in the emission is only 25% [20] . Hence, the internal quantum efficiency (IQE) is 25% in these materials. Therefore, harvesting the triplet excitons in OLED devices is crucial for better performance. All three different emission processes are shown in Figure 1.4. 1.2.2 Phosphorescence Seventy-five percent of triplet excitons do not undergo radiative recombination due to spin selection rules in the presence of very weak spin-orbit coupling (SOC) in the organic semiconductors. One way of achieving triplet exciton emission, i.e, phosphorescence, is by incorporating a heavy molecule, generally platinum and iridium [21], which have large SOC, along the backbone of the luminescent polymer in the device. These heavy molecules breaks the transition rules between S1 and T1 excitons. All the S1 excitons can be transferred to T1 through intersystem crossing (ISC) where we now have 100% of all excited excitons that can emit from T1 to ground S0 leading to 100% IQE. Emission from the triplet is necessary not only to enhance the luminance but also to produce the necessary broad white light emission (WOLED devices). Several methods to generate such a broad spectrum and harvest light from the triplet exciton exist. Some of these techniques include using a blue emitters OLED [22,23], materials having dimer or excimer emission [24], blended single layer or multilayer devices [17], or using high-efficiency diodes with delayed phosphorescence [18]. Phosphorescent OLEDs have superior efficiencies to 14 Figure 1.4. Three different emission processes: Fluorescence, Phosphorescence, and TADF. fluorescent devices but they contain heavy-metal complexes that make them expensive and suffer from efficiency degradation at high applied bias voltage. Meanwhile, devices based on blue-emitting materials have poor operational lifetimes [20,25]. 1.2.3 Thermally Activated Delayed Fluorescence We can achieve 100% IQE from the fluorescence state through the process called thermally activated delayed fluorescence (TADF) [20,26–28]. In TADF, triplet excitons are converted into the singlet excitons by a process called reverse intersystem crossing (RISC), which is efficient when the energy splitting between singlet to triplet (∆EST) < 100 meV and it is thermally activated. This process results in two different decay mechanisms, namely prompt (PF) and delayed fluorescence (DF). In the next section, we will discuss the different processes in the TADF such as RISC and small ∆EST. 15 1.2.3.1 Reverse Intersystem Crossing (RISC) The key process in the TADF mechanism is facilitated by RISC process from triplet (T1) to singlet (S1) i.e (T1→S1) when the energy difference between S1 and T1 is small and T1 is stable enough. TADF process is temperature sensitive and facilitates this endothermic transition at high temperature. The dependence of the kRISC on temperature can be expressed in a Boltzmann distribution relation: C^_ ` ∝= ∆bcd ef d 1.16 Since reverse intersystem crossing (RISC) is a nonradiative transition from the lower vibrational level of T1 to the higher one of S1 in an excited molecule, the delayed fluorescence efficiency (ΦDF) and its transient lifetime ( τDF) are highly dependent on the RISC rate constant ( kRISC). The RISC rate can be determined as follow: /^_ ` = Cgh Cih jgh C_ ` jih 1.17 Here kPF, kDF, and kISC are the rate constant of prompt, delayed component, and ISC. The rate of reverse intersystem crossing (kRISC) in TADF materials is on the order of 103-106 s1 and rate of intersystem crossing (KISC) is on the order of 106–1011 s−1 [20,26]. Rate constants (kPF and kDF) can be obtained experimentally. Cih = jih kih Cgh = jgh kgh 1.18 1.19 kih and kgh are fluorescent lifetime for prompt and delayed fluorescence, which can be obtained by fitting the decay curve of the time-resolved PL spectrum. 16 1.2.3.2 The Singlet-Triplet Energy Splitting (ΔEST) According to equation (1.16), fast RISC process require small energy splitting between S1 and T1 for efficient TADF emission. To obtain small ΔEST, special approaches for construction of TADF molecules need to be considered. The molecular energy of the lowest singlet ( ES) and triplet ( ET) excited state shown in equation (1.20) and (1.21 ) can be decided by the orbital energy ( E), electron repulsion energy ( K), and exchange energy ( J) of the two unpaired electrons at the excited states. , =,+/+l 1.20 ,m = , + / − l ∆, m 1.21 = , − ,m = 2l 1.22 Due to same electron arrangement of the singlet and triplet in one molecule, E, K, and J at the two excited states are the same with each other. However, due to the same spin states of the unpaired electrons in T1, the ET is reduced (equation (1.21)) in comparison with increased ES in S1 (equation (1.20)). ∆, m is the difference between ES and ET , equal to twice of J as shown in equation (1.22) . Two unpaired electrons S1 and T1 are distributed on the frontier orbitals of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO), respectively, with the same J value regardless of different spin states. Exchange energy of these two electrons at HOMO and LUMO levels can be calculated by equation (1.23). l = n op 1 o 2 q = r op 2 o MM 1 sM sM 1.23 17 Here op and op represent wave functions of electron at HOMO and LUMO level, respectively, e is electron charge. It is clear that small energy splitting can be resulted via a small overlap integral of< op |o >, i.e, via spatial wave function separation of HOMO and LUMO. To obtain these spatially separated HOMO and LUMO orbitals, a donoracceptor system with twist/spiro/bulky connection is introduced with special engineering that can reduce overlap between them and enhance charge-transfer state [26,29]. 1.2.3.3 D-A Molecular System for TADF Emission Two types of TADF molecules, namely intramolecular and intermolecular, are suitable to realize small ΔEST through separated HOMO and LUMO for TADF emission without need of heavy metals. According to equation (1.23), the HOMO of the donor and the LUMO of the acceptor contribute to the exchange energy of the D-A molecular system at the excited state of S1 and T1. Hence, the small overlap between the frontier orbitals for small ΔEST is shown in the bottom panel of Figure 1.5, which enhances the RISC process in the TADF materials, whereas large overlap between the frontier orbitals that gives large ΔEST is shown in the top panel of Figure 1.5. 1.2.3.3.1 Intermolecular D-A System Electron donor (D) and acceptor (A) molecules form an intermolecular D-A structure through charge transfer, known as an exciplex. The exciplex-based OLED can be fabricated with either a single-layer composed of a mixture of D and A molecules, or with a bilayer of two separate layers of donor and acceptor molecules. The emission spectrum from the exciplex OLED is usually broad and red-shifted relative to the emission of the 18 Figure 1.5. Molecular designs for large ΔEST (top) and small ΔEST (bottom). 19 individual acceptor or the donor and the emission efficiency is generally low [30,31]. This is because the electron transition from the LUMO of the acceptor to the HOMO of a donor in a large electron hole separation distance should provide small exchange energy, thereby forming small ΔEST for efficient TADF emission via RISC process from Triplet T1 to Singlet S1. Exciplex-based OLED can be tuned for the stability, color, and emission strength via the choice of appropriate donor and acceptor molecules. An example for donor and acceptor molecules that provide small ΔEST is shown in Figure 1.6. 4,4',4"-tris(N,Nphenyl-m tolylamino)triphenylamine (m-MTDATA) is used as the donor molecule and tris-[3-(3-pyridyl)mesityl]borane (3TPYMB) as the acceptor, resulting in ΔEST reported in Wang et al. [32] of approximately 42meV. This satisfies the requirement of TADF-based OLED, which should be less than 100meV. Electron-donating and accepting molecules should have a higher triplet energy level than a triplet exciplex energy level to prevent triplet energy back-transfer to the donor or acceptor [26]. Moreover, high PL efficiency, shallow HOMO levels in the donor molecules, and deep LUMO levels in the acceptor molecule are significant for exciplex-based OLEDs shown in Figure 1.7. 1.2.3.3.2 Intramolecular D-A Systems Donor-acceptor type TADF materials were realized via intramolecular chargetransfer state enabling small ΔEST, which can be enhanced by connecting donor and acceptor moieties by steric hindrance such as twist, bulky, or Spiro junction. Steric hindrance effectively separated the spatial overlaps between HOMO and LUMO [33,34]. Redox potential and π-conjugation length of the donor and acceptor moieties should also be taken into account for small ΔEST. Intramolecular TADF systems containing spatially 20 Figure 1.6. Donor and acceptor molecules which provide small ΔEST. Figure 1.7. Shallow HOMO levels in the donor molecules and deep LUMO levels in the acceptor molecule are significant for exciplex-based OLEDs [26]. 21 separated donor and acceptor have intramolecular charge transfer resulting in low PL efficiency due to limited frontier orbital overlaps. Hence, the smart choice of bridge for the connection between the donor and acceptor component is crucial. For efficient TADF emission, small ΔEST, stable T1, highly luminescent S1, and a large steric hindrance between donor and acceptor is required for strong CT emission along with limited structure distortions for nonradiative decay of the excited states [26]. 1.2.3.4 Energy Transfer Processes Although the OLEDs offer many advantages over the LED, there are still many challenges that should be overcome before they are used in the market. Similar to LED, OLEDs also experience decrease in efficiency with increasing current called “efficiency roll-off”. In addition, OLEDs experience challenges in operation lifetime and performances. Hence, in the TADF material, fluorescent emitter is doped in order to increase the stability of the devices, to achieve better performance, and also to cover the wide range of colors. The doped material is also called the ‘guest’ and is involved in the energy transfer from the host by two processes: Forster resonant energy transfer (FRET) and Dexter energy transfer process (DET) [35,36]. The energy transfer from host TADF molecule to the fluorescent emitter is known as Forster transfer. This is called a ‘long range’ process because it generally occurs at a distance of up to 100 Å. The probability of Forster energy transfer decays with R-6, where R is the distance between photoexcited exciplexes and the emitter molecules. In contrast, DET process is a short-range process and described as a loss mechanism in the OLED. These two processes are studied in Chapter 4 using magnetic field effect in the TADF OLED devices. 22 1.2.3.5 Magnetic Field Effect The magnetic field effect (MFE) technique has growing interest for studying spindependent generation and recombination processes of spin-pair species in organic semiconductors. Magnetic fields can change physical quantities in the optoelectronic devices in the form of magneto-electroluminescence (MEL) [37] in OLED, magnetophotoluminescence (MPL) and magneto-photoinduced absorption (MPA) in films, and magneto-conductivity/resistivity (MC/MR) and magneto-photoconductivity (MPC) in organic photovoltaic cells [38]. Several mechanisms for explaining MFE have been proposed such as (i) Hyperfine Interaction (HFI), (ii) the difference, Δg , between the g -factor of the spin 1/2 electron and hole that leads to different precession frequencies of their respective spins about the applied magnetic field, B [39,40], and (iii) a number of mechanisms involving triplet excitons: exciton charge interaction, [41] triplet-triplet annihilation, [42], and (iv) spin-orbit coupling in small molecules containing heavy atoms [43]. The underlying mechanism for the magnetic field effect (MFE) in an organic semiconductor is magnetic field modulation, which requires the existence of ‘spin-pair’ species (such as polaron-pairs, PP), and interaction between them, in which there is a spinmixing activity among the spin-pair sublevels. Another requirement is symmetry breaking, i.e., the dissociation rates, d, or/and the radiative recombination rates, r, are not equal for PP singlet (PPS) and PP triplet ( PPT) dS ≠dT, rS ≠rT as shown in Figure 1.8 [40]. When a magnetic field is applied, change in the spin-mixing between PPS and PPT occurs due to hyperfine interaction after lifting the degeneracy of the triplet levels (T+, T0, T-), leading to change of the steady state physical properties such as PC in solar cells, EL in OLED, and 23 Figure 1.8. Schematic presentation of charge dissociation and recombination processes and spin-mixing mechanism between the singlet and triplet e–h states. If the dissociation rates, d, of the singlet and triplet e–h pairs are not equal to each other, dS ≠dT, when B is applied, then they lead to change in photocurrent (MPC). Furthermore, if radiative recombination rates rS ≠rT, the spin-mixing at B yields MEL(B) in OLED and MPL(B) in films. PL in films. In donor and acceptor (D-A) blends, the electron (e) and hole (h), have different environments, and thus they possess different g-values. The difference of g-factor between e and h leads to unequal Larmor frequency of the e and h spins in a magnetic field. The difference in spin frequency (g1-g2) µBB/ℏ changes the relative orientation of the e- and h-spins, leading to interconversion of singlet/triplet polaron pairs at magnetic fields known as Δg mechanism. 1.3 Hybrid Organic-Inorganic Perovskites The organic-inorganic hybrid perovskites have attracted interest because of their promising application in optoelectronics such as high-efficiency photovoltaic cells, light emitting diodes (LED) [44], photodetectors, and field effect transistor (FET) [45]. The 24 perovskite provides important characteristics such as solution-processability, high color purity, and tunable bandgap, which works in both meso-structured and thin-film devices. The perovskite itself can absorb light and generate charges. The first reports of stable solid state solar cells based on CH3NH3PbI3 perovskite in mid-2012 and its PCEs have already exceeded 15%, which is leapfrogging every other solution-processed solar cell technology [46]. Altering or mixing the halide contents in the solution changes the bandgap of the perovskite, enabling the bandgap from 1.6eV with X=I to 2.3eV with X-Br and to 3.1eV with X=Cl. MAPbBr3 perovskites show the intense PL, emission thereby making this material favorable for LED applications [47]. The chemical structure of the perovskite is ABX3 shown in Figure 1.9 where A is the organic material such as methylammonium (CH3NH3: MA), B represents lead (Pb), and X is the halides (I, Br, Cl). The name perovskite comes after the name of Russian mineralogist L. A. Perovski who discovered a mineral called CaTiO3 in 1839 at the Ural Mountains in Russia. MAPbX3 has three major crystal structures: cubic, tetragonal, and orthorhombic. However, different halogen has different transition temperature. MAPbI3 has cubic phase above 330K; at this temperature, the MA+ and halogen ions both are disordered, leading to a lower symmetry of local structure in MAPbI3. With decreasing temperature, the cubic phase transformed to tetragonal phase. In the tetragonal phase, the halogen ions are ordered but not the MA+ ions. Tetragonal phase is transformed to orthorhombic phase below phase transition temperature 160K at which organic MA+ ions are fully ordered. 25 Figure 1.9. Crystal structure of the Hybrid Halide Perovskite (CH3NH3PbX3). A is organic ion, B is lead, and X = I, Br, Cl. [48]. Both MAPbBr3 and MAPbCl3 crystal structure possess cubic structure at room temperature and transform to tetragonal phase at 236 K and 177 K and from tetragonal to orthorhombic phase at 150 K and 172K, respectively. Depending on the ratio of organic and inorganic in the compound, different dimension perovskite structures can be fabricated. The three-dimensional (3D) perovskite contains a divalent inorganic cation and organic cation of equal charge. Lower dimensional structures such as two-dimensional (2D), one dimensional (1D), or zero-dimensional (0D) structures can be formed if the methylammonium organic cation is large enough [30]. In the layered 2D structure known as Ruddlesden-Popper phase, the exciton binding energies are large, which affects their optoelectronic properties significantly through the anisotropic properties of the crystal and strong quantum confinement. 26 1.3.1 Two-Dimensional (2D) Perovskites Crystal Structure 2D materials have a general formula of (RNH3)2(CH3NH3)n-1AnX3n+1, where R is an alkyl or aromatic moiety(C4H9), A is a metal cation such as Pb, and X is a halide. The variable n indicates the number of metal cation layers between the two layers of the organic chains [49] forming different layers. Schematic crystal structure of 2D butylammonium lead iodide perovskite is shown in Figure 1.10. In the extreme case where n = ∞, the structure becomes a three-dimensionally bonded perovskite crystal with a structure similar to BaTiO3. When MAI in 3D perovskite is replaced by larger organic molecules such as butylammonium iodide ((C4H9NH3I) or C6H5C2H4-NH3I), the inorganic octahedra layers can be separated by the organic moieties and form a lower dimensional Ruddlesden-Popper type 2D structure compound and reported as orthorhombic and non-centrosymmetric at room temperature [50–53]. In the opposite extreme where n = 1, the structure becomes an ideal quantum well with only one atomic layer of AX42– separated by organic chains, in which the adjacent layers are held together by weak van der Waals forces. 2D compounds are reported as orthorhombic and non-centrosymmetric at room temperature [37-40]. The thickness of the quantum well can be changed, thereby changing the optical properties of the 2D perovskite. The space group for n=1 are Cc2m with a = 8.9470(4) Å, b = 39.347(2) Å, c = 8.8589(6) Å for n=2, C2cb with a = 8.9275(6) Å, b = 51.959(4) Å, c = 8.8777(6) Å for n=3, and Cc2m with a = 8.9274(4) Å, b = 64.383(4) Å, c = 8.8816(4) Å for n=4. 27 Figure 1.10. Schematic crystal structure of 2D butylammonium lead iodide (C4H9NH3)2PbI4 . 1.3.2 Photoexcitation in Hybrid Perovskites Hybrid perovskites are also a kind of semiconductor where usually two kinds of photoexcitation species exist following photon absorption. One is photocarriers, which are typically common in inorganic semiconductors such as Si and GaAs. The second type of photoexcitation is the excitons, whose PA bands are in the mid-IR range and originates from the interband and intersub-band transitions [55]. The study of these excitations in hybrid perovskites is crucial to understand the charge transport in optoelectronic applications that make use of photoexcited states. Charge transport can be facilitated if the photoexcitation is mobile in the materials. 28 1.3.2.1 Free Carrier Absorption (FCA) The photocarriers do not show structured photoinduced absorption (PA) bands in the mid-IR spectral range. Instead, their optical signature is free carrier absorption (FCA), which is described by the Drude model. FCA may be written as, w x~ z T1 + {k U 1.24 where N is the photocarrier density, ω is the photon frequency, and τs is the momentum scattering time. Thus, the FCA contribution is limited to the spectral range of ω in which ωτs < 1. For high mobility semiconductors such as GaAs and Si, strong FCA occurs in the THz range since τs is of the order 1 ps. From the hole mobility, μ, obtained in crystalline perovskites (μ~100 cm2/vs) [56], we deduce τs ≈10 fs in these compounds, and consequently the FCA spectrum would be in the spectral range of a few tens of meV in these materials. Free-carrier absorption is an optical absorption process that occurs when a material absorbs a photon. Instead of generating electron-hole pairs; the photon energy is absorbed by free-carriers in either the conduction or valence band, moving the carrier to a higher energy state within that band [57], which is shown in the top panel of Figure 1.11. This type of transition is intraband transition. As we can see, the free electron is typically an indirect, phonon-assisted transition, but the free hole absorption can occur in a direct transition from the heavy hole band to the light hole band. FCA is stronger for holes than for electrons. Intraband transitions are different from interband absorption, which is usually explained by two absorption mechanisms, namely 1) intrinsic absorption where the photon 29 Figure 1.11. Intraband transition in the conduction band showing free carrier absorption (FCA) spectrum at low energy range. 30 energy must be higher than the bandgap, and 2) impurity level-to-band absorption, used in extrinsic infrared (IR) photodetectors [58]. Free carrier absorption is usually a performance degrading mechanism in many photodetectors. This is especially true for IR detectors because the absorption coefficient increases as the square of the wavelength of light absorbed. However, this type of absorption is proportional to the carrier concentration; it has been used to measure the spatial distribution of the carrier concentration in power semiconductor devices. The free carrier absorption coefficient in semiconductors is given, in MKS units, by the equation [1.3.2] [58]. = | } ~• € • •‚ • ƒ„ W } …∗ † (1.25) where λ is wavelength, ρ is density of free carriers (either electrons or holes), n is refractive index, m* is effective mass, and μ is mobility. 1.3.2.2 Exciton in 2D Perovskites The second kind of excitation is the exciton that can be observed in linear absorption. Excitons may have PA bands in the mid-IR range that originate from excitonic intersub-band or/and interband transitions similar to π-conjugated polymers [59]. Exciton PA bands may also originate from transitions into the continuum band at energies that correspond to high density of states [59]. From equation (1.5) described in section 1.1, we can calculate the exciton binding energy for the 3D semiconductor. In the 3D (isotropic) semiconductors based on Pbhalides, the excitons have been considered to be Wannier-like [60], whereas in section 1.1, 31 we came to know that the polymers’ exciton is Frenkel-like. Exciton binding energy for hybrid 3D perovskites (MAPbI3) is still under debate, although its bandgap is 1.5eV. So far, the binding energy reported in the literature for MAPbI3 varies from 10-55meV [60– 65]. In MAPbBr3 perovskites with a bandgap at ~2.3eV, however, the binding energy has been reported to be on the order of ~100meV [60,64]. Considering thermal energy kBT = 25 meV at room temperature, the photoexcitations should be mainly free carriers in MAPbI3 and excitons for MAPbBr3 at room temperature. We can also see that the exciton binding energy usually increases with the bandgap of a semiconductor. In a 2D semiconductor, the binding energy can be significantly enhanced due to spatial confinement as ,-. '4 , / = ,0 − 12∗ F '4 1 ∝ ˆ'4 − ‰ 2 1 1 '4 − 2 H , + ℏ /. + /2 + ,‡ , 26 1 E4 ∝ E4 ˆ'4 − ‰ , 2 1.26 1.27 where ,‡ is the quantization energy. When the dimension decreases from 3D to 2D, 12∗ increased to 412∗ , even without considering the screening effect. The screening effect caused by the difference in the @ (from equation (1.5)) of the quantum wells and barriers would enhance the exciton binding energy even more. The spatial confinement quadruples the exciton binding energy in the two-dimensional limit. The obtained enhancement factor is > 6 [66–69]. The large 1s exciton binding energy is ~200meV [68,69] for 2D hybrid perovskites due to effective dielectric confinement. In Chapter 5, we study the PA band from FCA and the binding energy of PEPI perovskite showing large oscillator strength. Hence, from the different PA bands for photocarriers and excitons, we can extract the characteristic properties of the photoexcitations [59,70]. 32 1.3.3 Rashba Effect in Hybrid Perovskites Larger Rashba-splittings are known to exist at several metal surfaces and in ultrathin metal films [71]. Not only in metals, large Rashba effect can be observed in layered polar semiconductor such as BiTeI having Rashba parameter(αR) of 3.8 eVÅ [72] and the surface of topological insulators such as Bi2Se3, which has αR 1.3 eVÅ [73]. But in the organic semiconductors, the Rashba effect is very small. Recently, hybrid perovskites having heavy atoms, namely Pb and halides, have attracted interest for application in spintronics as a result of the large spin-orbit coupling (SOC). This strong SOC causes the splitting of the spin degenerate bands in the presence of structural ‘inversion asymmetry’, which was observed by Rashba in the interface or surface of the materials [74] referred to as Rashba effect and in bulk of materials76 called Dresselhaus effect (Figure 1.12). Hence, study of the Rashba effect in hybrid perovskites is crucial to understand the spin manipulation in spintronic devices. The extrema points in the conduction and valence bands are of utmost importance in determining the optical, spin, and transport properties of semiconductors [12]. The electron dispersion relation, E(k), near these extrema points is usually described by the effective mass approximation, where the electrons and holes are treated as ‘free carriers’ having an effective mass, m*, that leads to spin-degenerate parabolic dispersion (Figure 1.12 (a)). In the case of Rashba-Dresselhaus effects, the electron (or/and hole) dispersion ‹Œ = ħ ‹Œ ± relation may be described by, ± C …∗ • • ‹Œ ^ ŽC Ž, where ^ is the Rashba-splitting parameter. This formulation yields new extrema at a momentum offset (k0) and energy splitting (ER) that are related to each other via the relation ^ = 2ER/k0 (Figure 1.12 (b)). 33 Figure 1.12. Rashba-splitting in semiconductors. (a) Schematic electron dispersion relation of a regular conduction band (CB) that shows a doubly spin-degenerate parabolic band having a single minimum at k=0. (b) Same as in (a) but subject to Rashba-splitting, in which two parabolic branches having opposite spin sense are formed. The Rashba-splitting energy (ER) and momentum offset (k0) are denoted. Importantly, the two Rashba-split branches have opposite spin sense; therefore, even though it is a small correction to the band structure, the Rashba effect can drive a wide variety of novel physical phenomena and influence the optical response and magnetic properties. Various theoretical and experimental methods are applied to study the Rashba effect in the hybrid perovskites and other materials. Experimental methods such as timeof-flight and spin-resolved angle-resolved photoemission spectroscopy (ARPES) are used to observe Rashba effect in Bi2Se3 and doped Bi2Se3 [73]. By using the similar surfacesensitive ARPES technique , D. Niesner et al. reported a spin splitting of the highest-energy valence band (VB) in single crystal MAPbBr3 [75]. The Rashba parameters αR were 34 measured as 7 ± 1 eV∙Å and 11 ± 4 eV∙Å in the orthorhombic and the cubic phases, respectively, which are amongst the highest values reported. According to D. Niesner et al. the Rashba-splitting arises when orbitals with SOC are subject to symmetry breaking electric fields in MAPbBr3 perovskites. Hybrid perovskite MAPbI3 has an indirect to direct bandgap transition that arises due to distortion of the lead iodide framework, leading to an electric field across the Pb atom and splitting the spin-degenerate conduction bands occurs indicating the presence of Rashba effect. This effect is observed via optical measurements such as steady state photoluminescence, time-resolved photoluminescence (TRPL), and microwave conductance (TRMC) and absorbance by tuning pressure and temperature [76,77]. Rashba effect is enhanced in the 2D materials due to intrinsic quantum confinement with many interfaces and large SOC. We study the Rashba effect in 2D perovskites using optical spectroscopies in Chapter 5. Moreover, our collaborator from NIST helped us in the theoretical part by calculating band structure and showing symmetry breaking in the form of Pb atom displacement from the octahedra center. We determined the Rashbasplitting energy, ER = (40 ± 5) meV and Rashba-splitting parameter αR= (1.6 ± 0.1) eV·Å in PEPI 2D hybrid perovskite. 1.4 Outline of Dissertation There are six chapters in this dissertation. Chapter 1 is the introduction of three different semiconductors and optical their properties. We introduce π-conjugated polymers, thermally-activated delayed fluorescence (TADF) materials, and also hybrid organic-inorganic perovskites, with more detailed discussions provided in later chapters. 35 Chapter 2 introduces the main experimental techniques as well as sample fabrications and characterizations. Chapter 3 describes the study of vibration modes from π-conjugated copolymers, namely poly-thienothiophene benzodithiophene 7 (PTB7). We measure photo-induced absorption (PIA), doping-induced absorption (DIA), and Raman and Infrared spectra in order to explain the vibration modes, in this copolymers. The PIA spectra below 0.5eV overlapped with the phonon modes, hence the antiresonance appears in the polaron band. We applied an amplitude mode model to explain these vibration modes in the π-conjugated copolymers. Chapter 4 presents the magnetic field effect study from the TADF organic light emitting diodes (OLED). We chose donor-acceptor type (intermolecular) TADF materials to fabricate OLED. Both electroluminescence (EL) and photoluminescence are enhanced by magnetic field effect, which is faciliated by reverse intersystem crossing (RISC). We doped the TADF materials with fluorescent molecules and we increase the stability of the OLED. From the temperature-dependent measurement of the MEL, we obtained that the energy splitting of singlet to triplet is ~15meV. In Chapter 5, we discussed the optical properties of 2D hybrid perovskite thin film, called (C6H5C2H4NH3)2PbI4 or PEPI, in which a giant Rashba effect has been observed. We applied all optical methods such as electro-absorption (EA) spectroscopy and cw photoinduced absorption (PA) spectroscopy to study the Hybrid 2D perovskite. We obtained binding energy of 1s exciton of 190meV. In addition, PA FCA at 0.15eV is a direct transition from Rashba sub-band from which we obtained Rashba-splitting energy of ~37 meV and Rashba parameter of ~2.7 eV∙Å, which are among the highest Rashbasplitting size parameters reported so far. 36 Finally, Chapter 6 summarizes the overall achievements, limitation of the dissertation, and suggestions for future work. CHAPTER 2 EXPERIMENTAL TECHNIQUES We used different experimental technique to explore the optical and electronic properties of organic semiconductors. This chapter introduces the experimental techniques applied in characterizing films of polymers, TADF materials, and hybrid organic-inorganic perovskites that can be used in organic photovoltaics and organic light emitting diodes. Thin films are prepared using glass, sapphire, cesium-iodide, or KBr substrates by either drop casting or spin coating in inert nitrogen atmosphere inside a glove box with oxygen level less than 1ppm. 2.1 Materials 2.1.1 π-Conjugated Polymers In this dissertation, we use π-conjugated polymers that have applications on organic photovoltaics to study the vibration modes and different optical transitions. Copolymers materials studied in this dissertation include poly-thienothiophene benzodithiophene 7 (PTB7), poly[(5,6-difluoro-2,1,3-benzothiadiazol-4,7-diyl)-alt-(3,3000-di(2octyldodecyl) -2,20;50,200;500,2000-quaterthiophen-5,5000-diyl)] (PffBT4T-2OD), and poly- benzodithio-phene fluorinated benzotriazole (PBnDT-FTAZ), homo polymers are poly (2methoxy- 5-(2-ethylhexyl- oxy)-1,4-phenylene-vinylene) (MEH-PPV) and region-regular 38 and region-random–Poly-(3-hexylthiophene) (RR, RRa, P3HT), and fullerene C60 molecule, [6,6]-phenyl C61 butyric acid methyl ester (PC61BM), and [6,6]-phenyl C71 butyric acid methyl ester (PC71BM) are tabulated in Figure 2.1. 2.1.2 TADF Materials In order to study magnetic field effects in the organic light emitting diodes, we use TADF materials whose ΔEST is small. We use donor (MeoTPD) and acceptor (3TPYMB) molecules and also a fluorescent emitter as the dopant whose molecular structure is shown in Figure 2.2. These molecules can be fabricated in two ways: 1) Solution Process 2) Coevaporation Process under high vacuum. 2.1.3 Hybrid Organic-Inorganic Perovskite We studied the free carrier absorption (FCA) and electroabsorption spectroscopy from a hybrid organic inorganic 2D (C6H5C2H4NH3)2PbI4 (PEPI) perovskite. The 2D perovskites precursors were synthesized by mixing PbI2, CH3NH3I, and C6H5C2H2NH3I in a dimethylformamide (DMF) solution with a typical concentration of 0.5 M. The amounts of precursors were kept at corresponding stoichiometric ratios to form (C6H5C2H4NH3)2(CH3NH3)n-1 (PbI4)n where n =1, 2, 3. Figure 2.3 shows the molecular structure of the 2D PEPI. 39 Figure 2.1. Molecular structures. PC61BM, PC71BM, MEHPPV (top row). PffBT4T-2OD (middle row). PBnDT-FTAZ and PTB7 (bottom row) P3HT, Figure 2.2. Molecular structures of TADF materials: MeO-TPD, 3TPYMB, and DBP, respectively. 40 Figure 2.3. Molecular structure of PEPI with alternating organic and inorganic layers, forming multiple quantum wells onto the substrate [79]. 2.1.4 Thin Film Fabrication 2.1.4.1 One-Step Solution Process Organic photovoltaic and organic light emitting devices from π-conjugated materials and TADF materials can be fabricated with a one-step solution process. This process is easy, low cost, and is widely used even in hybrid organic-inorganic perovskite thin film fabrication. One-step solution thin film fabrication is shown in Figure 2.4 (a). πconjugated polymers and TADF materials molecules are dissolved in different solvent such as (ortho di-chloro benzene (ODCB), chlorobenzene (CB), tri-chloro benzene), stirred overnight, and finally spin cast or drop cast on the substrates either sapphire or KBr, 41 Figure 2.4. Thin film fabrication of pi conjugated polymers, TADF materials, and hybrid perovskite. (a) Spin cast process (b) Co-deposition process. depending on the spectral range. Similarly for the hybrid organic-inorganic perovskites, the precursor solution is directly deposited on a pretreated substrate and followed by thermal annealing to form the perovskite phase. Because the precursors can fill the space within the pores, the solution process is attractive in mesoporous structures. 2.1.4.2 Co-deposition Process TADF OLEDs can be fabricated with co-deposition process as shown schematically in Figure 2.4 (b). Sources of two organic materials are thermally evaporated simultaneously in an ultra-high vacuum chamber. The deposition rate as well as the thickness were monitored with two sensors (sensor 1 and sensor2). The same process is applied to fabricate hybrid perovskite. Since it is a hybrid that contains both organic and inorganic material, one source contains organic materials and another with inorganic material that are thermally evaporated and deposited on the substrate, forming hybrid perovskite thin film. 42 2.2 Linear Absorption and Photoluminescence Measurement Absorption spectra measurements are the basic firsthand technique for measuring the optical gap of the material, as well as electronic excited states of interest. When the semiconducting material absorbs light, it promotes an electron from the ground state S0 to the excited state S1, which occurs depending on the oscillator strength of the particular transition, appropriate parity, and spin angular momentum. The material absorption spectrum in the UV-visible and near infrared spectral range was measured using a CARY 17 spectrophotometer. The absorption of a medium is quantified by measuring the optical density (O.D.), which is also called absorbance. In order to eliminate substrate effects and the system response, the absorption of a blank substrate is measured first and then this absorption is automatically subtracted from the absorption spectrum of the sample. Scattering and reflection from the films are assumed to be small and therefore are not accounted for. The absorbance ‘A’ was then calculated using the relation A=log (T0/T), where T0 and T are the wavelength-dependent transmitted signal intensity for the substrate and the sample, respectively. The absorbance is related to the film’s thickness ‘d’ and the absorption coefficient α according to the Beer-Lambert law A (λ) =O.D. =αd. α is defined by the density of optical absorbers N times the optical cross-section of those absorbers σ at wavelength λ, as α(λ)=Ν(λ) σ(λ). Therefore, the absorption that is measured in the unit of O.D. is given by the relation, \ = \ =•• ‘’ 2.1 The radiative decay from the excited state species (S1) to the ground state S0 emits a photon called photoluminescence (PL) and can be measured using a photo-detector. As When the polymer or TADF film is excited by a CW laser beam with above bandgap 43 photon energy, it generates steady state singlet excitons (SE: S0 S1) giving prompt fluorescence or (from exciplex in TADF films) giving delayed fluorescence. Besides radiative recombination, there could be also nonradiative decay in the form of heat energy. For measuring the fluorescence spectrum, a sample is deposited onto a sapphire substrate and kept inside a cryostat, since usually, organic films are sensitive to air. Appropriate pump energy that matches the material optical gap is used to excite the sample. Following Kasha’s rule, the emission spectrum begins at the lowest vibrational level. Therefore, the PL emission is red shifted as compared to absorption, and depending upon the vibrational overlap, several phonon replicas can be observed. The emission intensity depends on the overlap integral of the wave functions of the vibrational levels involved in transitions, which is expressed in terms of the Huang-Rhys parameter, S, determined by the excitonphonon coupling strength. Since PL originates from singlet exciton (singlet exciplex in TADF) radiative recombination, magneto-photoluminscence (MPL) cannot directly originate from SE (Spin (S)=0) (which is B-independent), but rather is caused indirectly by nonradiative decay channels of singlet excitons collisions with triplet excitons (TE) or polaron pairs, of which density varies with B. In TADF materials, triplet exciplex undergoes reverse intersystem crossing (RISC) process to singlet exciplex, which is a magnetic field-dependent process, and has another channel similar to polymers called polaron pairs, which is also magnetic field dependent. Due to these two channels, we observed large MPL in TADF materials. 44 2.3 Vibration Spectroscopy Vibrational spectroscopy is the collective term for Infrared (IR) and Raman spectroscopy. Both of these spectroscopic techniques are nondestructive, noninvasive tools that provide information about the molecular composition, structure, and interactions within a sample. They provide complementary information about molecular structure since their working principles are different. For IR absorption, sample is irradiated with polychromatic light and a photon of light is absorbed, leading to a change in molecular dipole moment during vibrations that are infrared active modes. For Raman scattering, sample is irradiated with monochromatic light and the photons are scattered changing the polarizability during the vibrations that are Raman active modes. Experimental setups for IR absorption and Resonant Raman scattering (RRS) are described in section 2.3.1 and 2.3.2, respectively. 2.3.1 Fourier Transform Infrared (FTIR) Absorption Fourier transform infrared spectroscopy (FTIR) is used to measure the IR absorption spectrum in the mid- and far–infrared spectral range. This technique is based on a Michelson interferometer, shown in Figure 2.5. Incident beam I0 from the source is split by a beam splitter into two beams; one beam (I1) is reflected from the stationary mirror, and the other beam (I2) is reflected from the movable mirror, causing the change in optical travel length denoted by x. The two reflected beams pass through the sample, generating an interferogram spectrum as a function of mirror displacement. Fourier transform of this interferogram spectrum gives the absorption as a function of wavelength in units of wavenumber (or cm-1). FTIR spectroscopy uses the Fast Fourier transform 45 Figure 2.5. Schematics of Michelson interferometer. One mirror is fixed and the other is movable. I0 is incident beam through the beam splitter. 46 (FFT) to convolute the spectrum with a limitation on resolution, Δνmax. In our measurements, we used FTIR with resolution of 4 cm-1 to measure the charged excited states in the polymer after it is doped with some oxidizing agent. 2.3.2 Resonant Raman Scattering (RRS) When an electromagnetic wave penetrates a medium, it generates a polarization, P(t), and this polarization acts as the source for the secondary wave, which interferes with the original wave giving rise to a transmitted wave. Any perturbation on a homogeneous medium can cause scattering and this scattering light contains frequencies different from the incident light [80]. If the scattered frequency is same as that of original incident light, then the phenomena is called Rayleigh scattering. However, there is a small part of scattered light whose frequency is shifted from the incident frequency by rotational or vibrational quanta. This process is called Raman scattering and the shift in frequency is called Raman shift, which is of fundamental importance in material characterization. Raman scattering was first observed in 1828 by Indian physicist Sir C. V. Raman who performed a series of measurements focusing sunlight on a liquid sample. He used a monochromatic filter that let only a specific wavelength of light to reach the sample. The measured scattered light showed a broader spectrum with additional wavelengths. A second filter (emission filter) behind the sample allowed blocking of the incident wavelength. The observed residual scattered light is clearly distinguished from the incident light. Hence, the Raman Effect is based on light interacting with the chemical bonds of a sample. The Raman spectrum is unique for each chemical composition, which provides a chemical “fingerprint” of the investigated compounds and can provide qualitative and 47 quantitative information of the material. Depending on the measurement techniques and properties of the experimental setup, we can observe Raman shifts spanning from tens of wavenumbers to thousands. However, it is very difficult to measure Raman below 100 wavenumbers because of strong excitation and the central Rayleigh peak [80]. Let us use a model suggested by Placzek in 1934 that provides a simple and intuitive explanation for Raman scattering. Assume that the molecular polarizability p depends on vibration coordinate Q giving • = @ “ = + ”• ”‡ “ , . Expanding this “ term gives “, again assuming vibration motion is harmonic “– cos {… . The molecular dipole induced by an external monochromatic electric field at frequency { leads to • = @ ,– cos { +@ š “– ,– cos { ± {… , which describes a time- dependent dipole moment and eventually a medium polarization. Here the first term explains the oscillation at the same frequency of the monochromatic excitation and contributes to Rayleigh scattering whereas the second terms oscillates at frequency ({ − {… ) and { +{… and accounts for the Raman effect, namely Stokes scattering and antiStokes scattering, respectively. However, this classical picture cannot explain a few important observations that are as follows [80]: i) When the excitation frequency is tuned, Raman peaks shoot up in amplitude dominating the spectrum and this process is called resonant Raman scattering. ii) Stokes Raman scattering is stronger that anti Stokes. This anti-Stokes is temperature dependent. iii) We know that not all vibrations are Raman active, and they can be Infrared (IR) active. Out of raman active vibrations, only few are resonant and when this happens, series 48 of overtones and combination bands can be observed with large intensity. Therefore, we need to apply quantum mechanics that describe the Raman scattering as a two-photon process. When a photon of energy ħ{ is destroyed and ħ{ is created together with material phonons ħ{… , then this process is called a Stokes event. Hence, the Stokes Raman energy is ħ{ = ħ{ − ħ{… . However, in anti-Stokes events, photon energy ħ{ and phonon energy ħ{… are destroyed and ħ{› is created. Hence, the antiStokes Raman energy is ħ{› = ħ{ + ħ{… [80]. In our measurement, we use a WiTec micro-Raman system to measure Resonant Raman scattering from π-conjugated polymers. The experimental geometry is in reflected mode with CCD camera as detector. Excitation wavelength is usually 486nm or 532 nm. It contains two gratings; one is 600nm and the other 1800nm. 2.4 Continuous Wave Photoinduced Absorption Continuous wave (CW) Photo-induced Absorption (PA) is the experimental technique used to detect long-lived photoexcitation species (with lifetimes of the order of 1ms at low temperatures) such as triplet and polaron excitations. The photogenerated species density is substantially higher at low temperature, because of increase in lifetime. CW diode laser is used in most cases as a pump beam to promote electrons from the ground to an excited state. Tungsten or Xenon lamps or Globar are used as probe light to excite the sample from excited state to higher excited states. Tungsten lamp covers the wavelength range from 250nm to 4.2µm and Globar covers the range from 4um to 15um. The experimental setup is presented in Figure 2.6. Both pump and probe beams are overlapped on the sample placed in vacuum inside of a cryostat equipped with a closed- 49 Figure 2.6. Experimental setup for CW PA and MPA measurement. Cw diode laser is used as the pump beam and tungsten lamp or Globar is used as probe lamp. cycle refrigeration system for reaching cryogenic temperatures. The probe beam that passes through the sample is collected with spherical mirrors and focused on a monochromator slit. The appropriate long pass filters are placed at the slit of the monochromator to block the pump beam. The light is then collected at the exit of the monochromator with a suitable photodetector: The monochromator grating, detector, and filters are adjusted according to the spectral range of interest, which spans from 450 nm to 12µm. Combinations of detectors and filters are shown in Table 2.1. Since Germanium (Ge), Indium Antimonide 50 Table 2.1. Detector and filter combinations used for measuring PA spectrum in a broad spectral range. Spectral Range(nm) 550-1000 1000-1500 1300-2500 2500-5000 4500-9000 Monochromator Grating 500/750nm 1um/1.6um 1um/1.6um 4um 8um 8500-12000 8um 250025000(FTIR range) FTIR spectrometer Detector Silicon (Si) Germanium (Ge) Indium Antimonide (InSb) Indium Antimonide (InSb) Mercury Cadmium Telluride (MCT) Mercury Cadmium Telluride (MCT) DTGS with KBr window and KBr beam splitter Filter 550 long pass 800 long pass 1300 long pass 2500 long pass 4500 long pass 7500 long pass FTIR spectrometer (InSb), and Mercury Cadmium Telluride (MCT) detectors are semiconductor detectors which have low bandgap, they are cooled by liquid nitrogen to reduce the thermal noise. The signal from the detector is fed to a preamplifier in order to linearize the voltage and sent to a lock-in amplifier that is connected to a computer. Measurements are done as a function of temperature using a closed cycle cryostat that operates from 40K to room temperature. In order for the signal to be detected by the lock-in amplifier, the pump light has to be modulated with a frequency corresponding to the lifetime of photoexcitation, usually around 305Hz by mechanical chopper. The absorption of the photoexcited species is essentially the difference in transmission (ΔT) when the sample is illuminated with both the pump and the probe (TL) and when sample is only illuminated with the probe (TD). The relation between the induced absorption coefficient (Δα) and difference in transmission (ΔT) is shown in equation (2.6). 51 The number of excited species is directly proportional to the change in transmission as shown in equations (2.2)-(2.6). Δ\ = \p − \g \p = \g = 2.2 2.3 •• From equations (2.2) and (2.3), ⟹∆αd=- ln (1+ As ΔT<<TD, 1+ ∆T ) TD ΔT == TD ž• ∆ s = −m ∆m 2.4 2.5 2.6 Usually, a PA spectrum contains both PA when Δα >0 and photoinduced bleaching PB, when Δα<0. In the case of PA, new transient photoexcitations are generated, whereas PB is due to depletion of the ground state optical transition by the pump. Fourier transform infrared (FTIR) Photoinduced Absorption (PA) is measured with probe beam of Globar from the FTIR spectrometer and DTGS detector. Experimental setup is shown in Figure 2.7. For modulating the pump beam, an external shutter is used. Since the signal in the infrared region is small and noisy, we need to average the signal more than 6000 scans with resolution of 4 cm-1. An applied magnetic field induces changes in PA, which is dubbed magneto-PA (MPA); it is measured using the same setup as the PA. Electromagnet pole pieces are placed in plane to the sample across the cryostat. The PA spectrum with and without the magnetic 52 Figure 2.7. Experimental setup for FTIR PA and DIA measurement. field is measured and divided by the PA intensity without magnetic field to obtain the MPA (percentage), i.e., 6¡x 8 = i› 4 i› i› x100 2.7 To obtain the desired magnetic field response of the PA spectrum in films, the monochromator was fixed at the desired wavelength where the triplet exciton PA or the polaron PA band were assigned, and the MPA is recorded while sweeping the magnetic field. The PA is calculated from the negative fractional change in transmission, which is subject to the relation: − ∆\ =∆ s=z O , s \ 2.8 where NSS is the species steady state density, σ(E) is the photoexcitation optical cross- 53 section, and E is the probe beam photon energy. Therefore, in a magnetic field, B, PA(B) is determined by the density NSS(B), which, in turn, is controlled by the photoexcitation species decay rate coefficient, k(B) [NSS=G/k], where G is the generation rate. The X species (X may be a polaron pair, triplet, or triplet-pair) has an excited state transition X0 X1, which is activated by the probe beam. For B ≠0, the X0 level splits according to the relevant spin multiplicity, L (L=3, 4, and 9, respectively, for the S=1 TE; PP composed of two S=½ polarons; and a pair of TEs). Consequently, through specific spin-mixing processes, the spin content of each sublevel Nss, its decay rate k, and thus PA, all become B-dependent. 2.4.1 Kinetic Analysis We can describe the photoexcitation dynamics with a simple rate equation as following: sz =£ s −1 z 2.9 Equations (2.8) and (2.9) are important equations that govern the PA. For steady state PA, N is in equilibrium and therefore G=R= aI0, which is proportional to the pump intensity I0. In transient absorption, however, N=N(t) and G is zero after excitation ends. 2.4.1.1 Monomolecular Recombination Monomolecular recombination (MR) is the simplest channel where only one species is involved in photoexcitation recombination, and therefore, R=N/τ. In steady-state, 54 sz z |ss = £ − =0 s k 2.10 z = £k = E¤ k having solution 2.11 This clearly shows that in cw-PA measurements, the monomolecular recombination process gives a signal proportional to the pump laser intensity. However, in transient spectroscopy, we expect the monomolecular recombination gives a signal that is proportional to exponential decay as follows: sz s giving solution of z t =− z = k 2.12 =z=¦ 2.13 2.4.1.2 Bimolecular Recombination When two photoexcited states are involved in one recombination, then it is Bimolecular recombination and in this case, R=bN2. In steady-state, sz |ss = £ − §z = 0 s 2.14 z = ¨£©§ = ªE¤ ⁄§ 2.15 Bimolecular recombination process gives signal proportional to the square root of the pump laser intensity in the cw –PA measurement. In the transient spectroscopy, it is not related to laser intensity. z = sz s = −§z z 1 + §z 2.17 2.16 55 2.5 Doping-induced Absorption (DIA) Polaron bands in CW PA can be identified by doping-induced absorption (DIA). In this experiment, the films were doped by iodine for several minutes to induce charges on the polymer chains, then absorption was measured. The difference in absorption before and after doping forms a band at the absorption peak of the induced charges. ¬¤x = -¬ ®¯- − -¬ ®¯- (2.18) The absorption in the lower energy (<0.5 eV) is measured with a FTIR spectrometer and normalized to match with the absorption data at high energy, which is measured using the CARY spectrophotometer. Iodine serves as the p-type dopant where iodine ions may isolate polymer and iodine counter-ions may localize the induced polaron excitation. As a result, a DIA spectrum consists of two polaron bands at low and high photon energy, P1 and P2. The same spectrum is obtained when transmission with and without doping is measured as formed by photogenerated polarons in the PA setup. 2.6 Electroabsorption (EA) Electroabsorption (EA) is an important tool to characterize the optical properties of hybrid organic-inorganic perovskite and π-conjugated copolymers. A cw electric field acts as a perturbation on the electronic structure, which induces a shift of the energy levels. This is known as either Stark shift or Frank-Keldysh effect for discrete or continuum states, respectively, and breaks the symmetry mixing state characters. In this technique, an electric field is applied to the sample and the change in transmission is measured with the probe lamp. Depending on desired wavelength range, either xenon lamp for UV region or tungsten lamp for visible region was used as a probe light, which are cw lights. EA 56 spectrum gives information for both even and odd parity states. The electric field F introduces a symmetry breaking that relaxes the optical transition restrictions so additional energy states, which are not unraveled in the linear absorption measurement because of optical selection rules, can be observed in the EA measurement. The external electric field perturbs the wave function and associated energies that are given by, JQ = E °JQ + & ∆,Q = & ⟨J² Žμ ‹Œ(Q . wŒ ŽJQ ⟩ ,Q − ,² |⟨J² |μ ‹Œ(Q . wŒ |JQ ⟩| ,Q − ,² J² µ J² = •Q w 2 (2.19) (2.20) where ψ· is the zero-field wavefunction, → is the transition dipole moment, and pj is µ ij the polarizability of the states resulting from the nonvanishing dipole matrix elements. The signal that is electric field-induced change in absorption (-ΔT/T) is proportional to the imaginary part of the third-order nonlinear susceptibility tensor χ(3), which is given in equation (2.21). The EA spectrum is usually proportional to the derivative of the absorption spectrum with respect to the photon energy (dα/dE) [81] : 4Y{ ∆\ ≅ −∆ s = ¤?»χH −{; {, 0,0 ¾ w s \ 'º 2.21 where d is film thickness and n is refractive index ; zeros in the frequency specify the cw field. The experiment needs a special device with an interdigitated gold electrodes array as shown in Figure 2.8(a). 12 ×12 mm2 quartz is used as substrate and 0.5 nm titanium and 150 nm of gold are RF sputtered on top of it. Interdigitated gold electrodes array are patterned using photo-lithography and etched with a gap of 40µm. A very thin film is 57 Figure 2.8. EA measurement. (a) EA substrate; (b) Experimental setup 58 prepared by spin coating, and the sample is mounted inside a cold finger cryostat equipped with electrical connections. The experimental setup for the EA measurement is shown in Figure 2.8 (b), which is in fact a modified version of the cw-PA setup. The electric field is used instead of the laser to access the excited states. The probe light is dispersed via the monochromator and guided through the sample but not directly dispersed to the sample in order to reduce the effect of heating the sample. From the sample, the beam is collected by curved mirror and focused on the UV enhanced silicon detector. An electric field of order of ~105V/cm is easily achieved between two electrodes because of their small spacing gap using an AC field of 200-300V. The AC signal is modulated at 1017 Hz and fed through the step up transformer whose output is connected to the sample electrode. We use a square wave for the AC signal. This AC modulation frequency is used as the reference to the phase sensitivity lock in amplifier and detected at the second harmonic frequency since the conjugated polymer has mirror symmetry, and therefore, the EA ~ V2. Thus, they respond to a sinusoidal field of frequency f with the EA signals at 2f, where the fundamental frequency is very much suppressed. 2.7 Device Fabrication We use 2mm Indium Tin Oxide (ITO) patterned on 12.5mmX12.5mm glass substrates, purchased from Kinetec Company, as the anode and the thermally evaporated calcium and aluminum as the cathode when fabricating OLEDs with active layer of TADF material. Substrates are ultrasonically cleaned with acetone, 2% micro-soap solution, deionized water, and methanol for 10 minutes each. A thin layer of about 50nm thickness of PEDOT: PSS is spin-coated on the substrate. A thin layer about 50nm thickness of PEDOT: 59 PSS is spin-coated on the substrate, which serves as the hole transport layer because its HOMO level lies between the ITO work function and the HOMO of most of polymers. The substrates are then transferred to the glove-box (<1 ppm oxygen) for the active layer deposition. The polymer and TADF solution are spin-coated on top of the substrate at different spin speeds to vary the thickness. Small molecules such as Alq3 cannot be spincoated as they are hard to dissolve in the solvent. Therefore, they are thermally evaporated inside the vacuum deposition system at high vacuum of 10-6 Torr. TADF material solutions are prepared at different concentrations and have been experimentally optimized. After the active layer deposition, 25nm calcium and 80nm aluminum is thermally evaporated where Al serves as the capping layer to protect Ca from oxidization. Occasionally, the device is also encapsulated with a microscope cover glass using optically adhesive UV curable glue if the measurement needs to be performed in air. Typical UV exposure time is ~45 seconds. 2.8 OLED Characterization In this dissertation, we measured the OLED IV curve and MEL at different temperatures varying from 320K to 10K. The OLED device is mounted in a closed cycle He-cooled cryostat with proper electrical connection to measure MEL at low temperature. In order to measure at high temperature above 300K, we have a home-built oven that also has proper electrical connection. For optical detection, we have a silicon photo-detector with the preamplifier, or the Ocean Optics spectrometer, to measure the MEL and electroluminescence spectrum. Magnetic field is applied using an electromagnet up to ~2000 Gauss, and the current for MC and the EL for MEL are scanned by sweeping the magnetic field in plane to the device. Experimental setup is shown in Figure 2.9. MC is 60 Figure 2.9. Experimental setup for the Magneto-electroluminescence (MEL) of TADF OLED. defined as, 6 = and MEL is defined as, 6,¿ = ¤ 8 −¤ 0 ¤ 0 ,¿ 8 − ,¿ 0 ,¿ 0 2.22 2.23 CHAPTER 3 AMPLITUDE MODES SPECTROSCOPY OF CHARGE EXCITATIONS IN PTB7 π-CONJUGATED DONOR-ACCEPTOR COPOLYMER FOR PHOTOVOLTAIC APPLICATIONS This chapter is a reprint of a paper published in Physical Review Applied in the year 2017 authored by Sangita Baniya, Shai R. Vardeny, Evan Lafalce, Nasser Peygambarian and Zeev Valy Vardeny1. We measured the spectra of resonant Raman scattering and doping-induced absorption of pristine films of the π-conjugated donor-acceptor (D-A) copolymer, namely Thieno[3,4 b]thiophene-alt-benzodithiophene (PTB7), as well as photoinduced absorption spectrum in blend of PTB7 with fullerene PCBM molecules used for organic photovoltaic (OPV) applications. We found that the D-A copolymer contains six strongly coupled vibrational modes having relatively strong Raman scattering intensity, which are renormalized upon adding charge polarons onto the copolymer chains either by doping or photogeneration. Since the lower energy charge polaron absorption band overlaps with the 1 Sangita Baniya, Shai R. Vardeny, Evan Lafalce, Nasser Peygambarian, and Z. Valy Vardeny Phys. Rev. Applied 7, 064031 (2017). Published by American Physical Society. Copyright © 2017 American Physical Society. 62 renormalized vibrational modes, they appear as antiresonance lines superposed onto the induced polaron absorption band in the photoinduced absorption spectrum, but less so in the doping-induced absorption spectrum. We show that the Raman scattering doping and photoinduced absorption spectra of PTB7 are well explained by the amplitude mode model, where a single vibrational propagator describes the renormalized modes and their related intensities in detail. From the relative strengths of the induced infrared activity of the polaron-related vibrations and electronic transitions, we obtained the polaron effective kinetic mass in PTB7 using the amplitude mode model to be ~3.8m*, where m* is the electron effective mass. The enhanced polaronic mass in PTB7 may limit the charge mobility, which, in turn, reduces the OPV solar cell efficiency based on PTB7/fullerene blend. 3.1 Introduction Organic photovoltaics (OPVs) based on blend films of organic semiconductors such as π-conjugated polymers and fullerene molecules have attracted significant interest as a potential lightweight, low-cost renewable energy source [82–86]. The donor polymer materials were homopolymers such as polythiophene that absorb in the visible range of the solar spectrum [87]. However, homopolymers have recently been replaced by block copolymers whose repeat units consist of alternating donor (D) and acceptor (A) moieties [88–97]. This chain architecture reduces the optical gap drastically, and thus the D-A copolymers absorb in the near infrared, where the largest fraction of the photons emitted by the Sun lies. The power conversion efficiencies (PCE) of organic solar cells with D-A copolymers as donor materials and fullerene molecules as acceptors have 63 exceeded 10% [97] approaching the 12% threshold for commercial viability [83,98] . Despite the increasing interest in device applications, there is only limited understanding of the molecular doping of D-A conjugated copolymers [99], or the nature and properties of the charge excitations (polarons) in these materials [100–108]. This requires deeper understanding of the vibrational modes associated with the polaron excitation in the D-A copolymer chains since the strongly coupled vibrations may enhance the polaron mass thereby inhibiting its mobility. The polaron excitation in π-conjugated polymers distorts the chain dimerization pattern and this renormalizes the strongly coupled vibrations [109]. Consequently, Raman active modes having even parity character (i.e, Ag symmetry modes) become infrared active with very large oscillation strengths [110]. These modes, dubbed infrared active vibrations (IRAVs) appear upon doping or charge photogeneration and are thus signature of charges added onto the polymer chain [111,112]. One way to describe the IRAVs in πconjugated polymers is via the amplitude mode model (AMM) that was successfully advanced by Horovitz and collaborators for polyacetylene, (CH)x [110,111]. The AMM has had enormous success in describing the vibrations of the ground and excited states of trans- and cis-(CH)x. In addition, it has also been used to analyze the strong dispersion of the Raman modes with the laser excitation frequency in t-(CH)x as measured by resonant Raman scattering (RRS) spectroscopy [110,111,113]. The IRAVs in this and other polymers appear in the absorption spectrum as intense absorption lines if the absorption bands of the charge polaron do not overlap with the vibrational modes. However, when the charge polaron lower absorption band overlap with the vibrational modes then quantum interference between the two features may occur in the form of multiple Fano effect, where 64 the IRAVs manifest themselves in the form of antiresonances (ARs) superimposed on the polaron lower absorption band [109,114]. Since the lower polaron absorption band in low bandgap D-A copolymers should be in resonance with the IRAVs, it is expected that ARs would dominate the absorption spectrum when charges are added to the copolymer chains. Yet the anticipated ARs have not been studied in D-A copolymers, and the suitability of the AMM to describe the vibrational modes in these compounds has not been validated. Therefore, important information on the charge polaron in these copolymers could not be obtained. To achieve p-type molecular doping in π-conjugated polymers, highly electronegative molecules are mixed in low concentration with the organic chains [99] . Here we focus on p-type doping, namely the introduction of excess positive polarons into the copolymer chains. The most commonly accepted picture is the ground-state transfer of one electron from the highest occupied molecular orbital (HOMO) of the polymer to the lowest unoccupied molecular orbital (LUMO) of the electronegative dopant molecule, thereby enhancing the polymer film conductivity. However, recent observations show evidence that the effect of doping on conductivity enhancement in D-A copolymers is lower than that for homopolymers [99] . It is thus interesting whether the inability of ptype doping is a common property for the π-conjugated copolymers that are used in OPV applications. For our investigations we have chosen the D-A copolymer thieno[3,4b]thiophenealt- benzodithiophene (PTB7); see structure in Figure 3.1(a) [115]. PTB7 is a low bandgap copolymer with optical gap in the near-IR spectral range, namely at ~ 1.7 eV (Figure 3.1(b)) [115] , [91]. When blended with PCBM, the PTB7 blend has achieved 65 Figure 3.1. (a) The backbone structure of the D-A copolymer that contains two different moieties. (b) The photoluminescence and absorption spectra of PTB7 film at ambient conditions. photovoltaic PCE of up to 9% in an optimized device architecture [91]. Alas a recent study demonstrated that PTB7 degrades under accelerated conditions, showing a rapid loss of absorption compared to homopolymers such as P3HT and PPV [116,117]. To characterize the strongly coupled vibrational modes of the copolymer that influence the polaronic mass we measured the RRS, doping-induced absorption (DIA) and photoinduced absorption (PIA) spectra in films of pristine PTB7 and PTB7/PCBM blend, respectively. We show that the main charge excitation of the PTB7 chains is the polaron having two characteristic absorption bands below the gap [106]. Since the low-energy polaron absorption band indeed overlaps with the strongly coupled vibrational modes, these modes mostly appear in the DIA and PIA spectra as ARs superimposed on the polaron absorption band. Nevertheless, we show that the AMM can still provide an excellent basis for understanding the various vibrational spectra, since it describes all three spectra in detail. Using the AMM parameters and the relative strengths of the IRAVs compared to the polaron absorption bands, we calculate the polaron effective mass in PTB7 to be ~3.8m*, where m* is the 66 effective electron mass in this copolymer chain [106]. The polaronic mass enhancement may inhibit the charge mobility in this compound, which, in turn may reduce the obtained OPV solar cell efficiency. 3.2 Experiment The PTB7 copolymer powder was purchased from Solarmer and used without further purification. Thin films of PTB7 were spin cast or drop cast from PTB7 solution in 1,2-Dichlorobenzene (ODCB) at concentration of 10 mg/ml onto sapphire or KBr substrates, depending on the spectral range of interest. For the PTB7/PCBM blend we used a solution of PTB7 and PC60BM (1:1.5 in weight) with the same concentration in ODCB, which was stirred in Nitrogen atmosphere overnight. The sample films were placed in a cryostat equipped with transparent windows in the visible to mid-IR, where the temperature could be regulated between 40-300K. The photoluminescence (PL) and absorption spectra of the pristine films were measured using a steady state photomodulation pump-probe setup [118], where the pump was a diode laser at 486 nm and intensity of ~ 100 mW/cm2. For PIA, the probe beam was derived from an incandescent Tungsten lamp (visible spectral range) or glow-bar (mid-IR range) and passed through a ¼ met monochromator optimized at various wavelengths throughout the visible, near-IR and mid-IR spectral ranges. A variety of semiconductor detectors such as Si, Ge and InSb were used to monitor the transmission, T through the sample film, and the changes, ∆T induced by the pump beam. The PIA spectrum was subsequently calculated as -∆T/T. To cover the spectral range of 500-4000 cm-1 we used an FTIR spectrometer. For obtaining the PIA spectrum, we used a shutter to modulate the 67 laser illumination on the film, and we signal-averaged the IR absorption spectrum for 6000 scans [114]. For the DIA spectrum, we doped the film p-type by exposing it to iodine vapor for various time durations ranging from a few seconds to one hour. The Raman scattering spectrum was measured using a micro-Raman spectrometer equipped with a cw laser at 486 nm where the inelastic scattered light was detected with a filter and a photomultiplier tube. 3.3 Results and Discussion Figure 3.1 shows the absorption and PL spectra of pristine PTB7 film. Figure 3.1(a) shows the copolymer backbone repeat unit; the repeat unit is composed of donor and acceptor moieties [34]. In Figure 3.1(b) it is seen that the absorption onset is at 1.7 eV, whereas the absorption peaks at ~ 1.8 eV with the maximum slope at 1.75 eV; this is the mean exciton energy in PTB7. One of the reasons for the low optical gap here is the difference in electron affinity between the donor and acceptor moieties in the copolymer chain [108]. The absorption spectrum shows a second pronounced band ~150 meV higher than the first band, which we consider to be a ‘vibration replica’. We note that a third band occurs in the absorption spectrum at ~ 3.1 eV, which may be due to a second transition of the copolymer chain [119,120]. The PL spectrum shown in Figure 3.1(b) is composed of two pronounced bands that are in the form of ‘mirror image’ of the absorption spectrum close to the band-edge, that are ~150 meV apart. The PL main peak is at 1.55 eV having an apparent ‘Stokes shift’ of ~ 250 meV from the main absorption peak. However, this large energy difference is mainly caused by exciton diffusion to sites on the chain that have the lowest energy, rather 68 than a natural, intrinsic Stokes shift. The relatively low optical gap of PTB7 makes it attractive for organic photovoltaic applications [115,120]. The expected difference in electron affinity of the donor and acceptor moieties raises the question whether the inversion symmetry is broken in this copolymer chains [108]. This can be checked by Raman scattering and IR-absorption. If the chain inversion symmetry still holds, then the Raman active vibrations are not IR-active, and vice versa. Figure 3.2 shows the RRS spectrum compared to the IR-absorption spectrum in pristine PTB7 film on KBr substrate. The RRS spectrum contains six strongly coupled modes that are labeled 1-6 (Figure 3.2(a)). We note that the two modes above 1600 cm-1 have been identified as due to impurities and defects on the chains [118] , and thus we do not analyze them in detail. It is interesting to note that the C=C stretching mode in PTB7 shows a group of Raman-active modes stretching from 1400 to 1600 cm-1. This is probably caused by a distribution of the C=C frequency due to the different environments on the copolymer chain close to the donor, acceptor and the ‘in between D-A’ place [118] . On the contrary, the IR absorption spectrum shown in Figure 3.2(b) contains many more IRactive modes than in the RRS spectrum; and there is no resemblance of the two spectra. We therefore conclude that the copolymer chain approximately retains the inversion symmetry operation. Figure 3.3 shows the evolution of the IR-absorption spectrum in PTB7 upon iodine doping for various durations. We note that doping of PTB7 has occurred relatively easy here, similar to doping of many homopolar polymers; and in contrast to other copolymers discussed in the literature [99]. We can divide the DIA spectrum in PTB7 into two regions; namely the vibrational spectral range up to ~1500 cm-1 and the electronic range at higher 69 Figure 3.2. The Raman scattering (a) and IR-absorption spectra (b) of PTB7 film measured at ambient conditions. The most strongly coupled modes 1 to 6 are assigned. The (*) symbol denotes impurity-related mode [117]. 70 Figure 3.3. The evolution of the doping-induced absorption of PTB7 film upon exposure to iodine vapor at various time durations as given. The P1 polaron band and related IRAVs are assigned. 71 photon energy. The absorption strength in both ranges intensifies similarly with the iodine exposure time, showing that they are correlated. The absorption increase also shows that both ranges lure their oscillator strength from the ground state absorption due to charges that are added onto the chains by the iodine molecule dopants. We identify the strong vibration as IRAVs showing that the added charges onto the copolymer chains give rise to new IR-active vibrations in agreement with the extra charges [111]. It is rather surprising that the IRAVs oscillator strengths are similar (but weaker overall) to that of the absorption band associated with the charge excitation. The immense vibrational oscillator strength indicates that the added charges are ‘easy to move’, pointing to a relatively small effective kinetic mass, despite the relaxation energy associated with the added charges on the chain, namely the ‘polaronic effect’. A more complete picture of the DIA spectrum related with the added charges on the chains is shown in Figure 3.4(a), where the DIA is displaced on a broader spectrum. It is seen that in fact the DIA spectrum of the charge excitations contain two bands, namely P1 peaked at ~0.4 eV and P2 with a peak at ~1.15 eV [121,122]. We therefore identify the charges on the copolymer chains as polarons since these excitations have two allowed absorption bands, as seen in many DIA and PIA spectra of homopolymers [114], and calculated for the copolymers [106]. From P1 band we can extract the polaron relaxation energy in PTB7 copolymer upon doping to be ~ 0.3 eV (see Figure 3.4(c) inset). This lower energy estimate takes into account the large dip superimposed on the P1 band at ~ 1600 cm-1 (or 0.2 eV). It is noteworthy that the appropriate sum of the polaron absorption energies according to Figure 3.4(c) inset, namely 2E(P1)+E(P2)= 1.75 eV, which is the optical gap energy. This shows that the polaron energies are symmetrically located in the 72 Figure 3.4. The doping-induced absorption (DIA) spectra of PTB7 film (a) and (b), and photoinduced absorption (PIA) spectra of PTB7/PCBM blend film (c) and (d). The polaron bands P1 and P2 are assigned (see inset in panel (c)); the triplet PIA band, T is also denoted in (c). The IRAVs are shown in more detail in panels (b) and (d), where the antiresonances (ARs) are denoted with arrows pointing to dips superimposed on the polaron P1 band. copolymer gap [114], as depicted in Figure 3.4(c) inset. In addition, we also note that the relative strengths of the IRAVs to that of the two polaron bands is approximately 1:7. This indicates a reasonably small polaron kinetic effective mass, Mp that can be estimated using the AMM (see below). A more detailed DIA spectrum of the vibrational modes shows that in fact some IRAVs do not reveal themselves as sharp absorption bands, but are in the form of dips (or ARs) superimposed onto the P1 DIA band of the polaron (see Figure 3.4(b)) [109] . It is clearly seen that there are at least four large ARs in the DIA spectrum; and Figure 3.3 shows that they grow upon iodine exposure. These ARs replace the traditional IRAVs that are formed upon doping in many π-conjugated homopolymers. Interestingly we note that there are no pronounced ARs in the lower frequency range below 1200 cm-1. This indicates that the ARs are due to interference between the IRAVs and the polaron P1 73 band [109,123]. Since the polaron band in the DIA spectrum does not extend to frequencies smaller than 1200 cm-1, there are no ARs that are formed below this frequency. In this case, we may consider the peak seen at ~1115 cm-1 as IRAV rather than AR, and take it into account in calculating the relative oscillation strengths of the vibration to that of the electronic absorption bands. A detailed ARs spectrum is barely possible via DIA spectroscopy since doping increases inhomogeneity and disorder in the copolymer chains, which broadens the AR dips. A better ARs spectrum can be obtained via photogeneration. However, the photon absorption in the pristine copolymer does not usually lead to charge excitation, since the binding energy of the excitons is quite large (of order 0.4 eV [108]). An alternative way for polarons photogeneration is therefore in blend of the copolymer donor with fullerene acceptor molecules, such as PCBM. In this case, the photogenerated excitons in the copolymer may dissociate at the copolymer/fullerene interface, where an electron is donated to the fullerene acceptor and consequently a hole polaron is left on the copolymer chains [87]. Under steady state conditions, hole polarons are photogenerated onto the copolymer chains in the same manner as doping with iodine, except that the disorder in the blend is less acute than that formed upon doping. Figure 3.4(c) shows the PIA spectrum of PTB7/PCBM blend. The spectrum contains the two polaron bands P1 and P2, the related ARs, and another band at ~ 1.2 eV that was identified before as due to triplet excitons [121,122]. We checked that P1, P2 and ARs in the PIA spectrum have the same dependence on the laser excitation intensity and temperature, and therefore they originate from the same photoexcitation species, which we identify as photogenerated hole polarons on the PTB7 chains [121] . A more detailed PIA 74 spectrum in the frequency range of the vibrational modes is shown in Figure 3.4(d). At least five ARs can now be clearly identified, at frequencies slightly lower than those in the DIA spectrum of Figure 3.4(b). 3.3.1 Analysis Using the Amplitude Mode Model We now introduce the AMM for analyzing the strongly coupled vibrations that are revealed as peaks in the copolymer RRS spectrum and ARs in the DIA and PIA spectra, respectively. In previous applications of the AMM it has been explicitly assumed that the adiabatic approximation holds true [111]. This was a correct assumption since the Raman frequencies are much smaller than the electronic absorption bands. This approximation however does not hold in our case; since the vibrational frequencies are in resonance with the P1 polaron band. We thus need to use the modified AMM that includes non-adiabatic effects [109], [123]. An important ingredient of the AMM is that all IRAVs are interconnected by being coupled to the same phonon propagator. The bare phonon propagator in the copolymer chain is given by [110]: ¬ { =& À { À { − { − Á{ , 3.1 where { , δn, and λn are the bare frequencies, their natural width and electron-phonon (e- p) coupling constants; and ∑λn=λ that is the total e-p coupling constant. The index, n varies from 1 to N, the number of coupled modes in the polymer chain; for example N=6 for PTB7 copolymer. The bare frequencies in the coupled e-p chain are renormalized, since they interact with the electronic gap, so that their modified propagator is given by a ‘Dysontype’ equation [113]: 75 ¬ { = ¬ { , 1 + 1 − 2À∗ ¬ { 3.2 where the renormalization parameter, 2λ*=(1-2λ, šš (2∆0)), , šš is the second derivative of the total electron-phonon system vs. the dimerization amplitude, ∆, and 2∆0 is the dimerization gap at equilibrium [111] . The poles of equation (3.2) are the renormalized Raman frequencies, {^ of the copolymer, given by the relation [113]: , ¬ { = − 1 − 2À∗ 3.3 The renormalized AR frequencies, ωn,AR appear as zeroes in the conductivity response, σ(ω) [109], [123] : O { ~ 1+¬ { 1− š , 1 + ¬ { T1 + c1 − αU which are given by the relation: where š ¬ { =− 1− š , is a non-adiabatic renormalization parameter. It is noteworthy that 3.4 š 3.5 need not be the same for the ARs in PIA and DIA spectra, since the polaron excitation in PIA is ‘free’, whereas the polaron excitation in DIA is bound to the dopant molecule [109] . In contrast, the IRAV frequencies appear as poles in Eq. (3.4), given by the relation D0 (ω) = -[1+c1−α]1 (where c1 and α are constants), which are practically suppressed by their proximity to the ARs. When plotting the function D0 (ω) (Figure 3.5(b)), we note that it has poles at the bare frequencies, { and approaches the value (-1) at low ω. This is because at ω=0 the numerator and denominator in D0 (ω) cancel each other, and ∑-λn/λ= -1. Equations (3.3) and (3.4) are in fact polynomial of order N in ω2, and thus it is much easier to solve for the renormalized vibrational mode frequencies by drawing a horizontal line at different values 76 Figure 3.5. The measured spectrum (lines) and calculated vibrational mode frequencies (symbols) in PTB7 film for (a) photoinduced absorption (PIA); (c) doping-induced absorption (DIA); and (d) resonant Raman scattering. The ‘bare’ phonon propagator, D0(ω) is shown in (b), where the horizontal lines are obtained using 2λ*=0.73 (RRS, down triangles), š (PIA)=0.44 (ARs, circles), and š (DIA)=0.44 (ARs, up triangles). The modes 1 to 6 are assigned. D0 (ω) parameters are given in Table 3.1. The (*) symbol in panels ((b) and (c) denotes an IRAV at 1115 cm-1, that is not an AR dip (see text for discussion). 77 that correspond to the parameters λ* or š (see Figure 3.5(b)). To fit the various renormalized frequencies using the AMM we have to find 2N-1 parameters associated with D0 (ω), and three additional parameters, namely λ* and two different š . This has been a formidable task; but when considering also the RRS intensities, it has made it much easier. The best fitting function D0(ω) that describes the six most strongly coupled vibrations in PTB7 is shown in Figure 3.5(b), together with three horizontal lines that represent the cases of RRS, PIA and DIA Figure 3.5(a),(c),(d) (Eqs. (3.3) and (3.5)). The fitting parameters for the best D0 (ω) are given in Table 3.1. The fit with the experimental frequencies is excellent. We therefore conclude that the AMM is a good basic model for describing the strongly coupled vibrations in PTB7. We found that, in general 2λ*> and š (DIA)> š (PIA); this agrees with other polymers, where š š , is in fact α(pinning) upon doping or photogeneration [109]. It is interesting to note that the horizontal line that describes the ARs in the DIA spectrum in fact has an intersection at ~ 1180 cm-1; alas, there is no AR dip that is obtained at this frequency. The reason for the apparent ‘failure’ to form AR dip at that frequency is that the polaron DIA band does not extend to such low frequency; and hence interference with the amplitude mode #1 does not occur. Instead of an AR, the amplitude mode #1 appears in the DIA as a renormalized IRAV at 1115 cm-1 (denoted by a (*) in Figure 3.5 panels (b) and (c)) that can be generated from the phonon propagator at a different š . One of the benefits of using the AMM to describe the most coupled vibrations of the copolymer chain is that it can predict the relative scattering intensities in the RRS spectrum [110]; this is a unique virtue, since, in general, this is rather impossible to predict. The reason behind this extraordinary ability is that all renormalized modes are related to each other through the same phonon propagator, which is not the case 78 Table 3.1 The best fitting parameters for the bare phonon propagator, D0(ω) that describes the six most strongly coupled vibrational modes in PTB7. ω are the bare frequencies, λn/λ are their relative e-p coupling, and δn are their natural width used to fit the RRS spectrum (Figure 3.6(b)). Mode index 1 2 3 4 5 6 ω (bare frequency) cm-1 1240 1316 1445 1520 1575 1760 À © À 0.005 0.002 0.008 0.03 0.002 0.953 δn (RRS) cm-1 9 10 25 60 13 20 in RRS spectra of most compounds. It turns out that the Raman scattering intensity, In of each renormalized mode is inversely proportional to the slope of D0(ω) at the renormalized frequencies [110] ; namely In~|dD0(ω)/dω| at ω={^ . Since the function D0 (ω) is known for PTB7 and we also found the proper parameter 2λ* (=0.73) that describes the RRS frequencies, it is straightforward to calculate the RRS spectrum based on the AMM. This is nicely shown in Figure 3.6, where the calculated RRS spectrum Figure 3.6(b) is compared to the experimental spectrum Figure 3.6(a). The fit is superb; and this validates the conclusion that the AMM is a reliable theoretical model for the vibrational frequencies in PTB7, and probably for many other D-A copolymers. Another benefit of using the AMM for the vibrations in the class of the πconjugated polymers is the ability to estimate the polaronic effective kinetic mass, Mp from the intensity ratio of the IRAVs to that of the electronic absorption bands. This is given by the following relation [112,113]: 6¯ Y ¤¯ Å = ˆ ‰ , ∗ ? 5.6 ¤Ä Å 3.6 79 Figure 3.6. The RRS spectrum of PTB7 as measured (a) and calculated (b) using the AMM parameters given in Table 3.1. where Ip/Iv is the ratio of the polaronic absorption band, Ip to that of the IRAV Iv, m* is the electron effective mass (~0.1 me in π-conjugated polymers) and the frequencies Ω1 and Ω0are given by the relations [111]: Å−2 0 = & ' À' À {0' −2 , Å21 = & ' À' À {0' 2 , 3.7 We obtain from the DIA spectrum shown in Figure 3.4(b) Ip/Iv≈6.7; and from the AMM parameters given in Table 3.1 we calculate (Ω1/Ω0)2≈1. Using Eq. (3.6) we therefore estimate MP/m* ≈ 3.8 for the polaron effective kinetic mass in PTB7. This shows that the polaronic mass is enhanced over the electron effective mass, m* by ~3.8, which is intermediately large indicating that the ‘polaronic effect’, namely the polaron relaxation 80 energy in PTB7 is still large, in agreement with the P1 band in the PIA spectrum (Figure 3.4(c)). It is interesting to compare the obtained polaronic mass in PTB7 (MP/m*≈3.8) to the soliton mass MS in t-(CH)x (MS/m*≈6), which was estimated using the same method [110]; the smaller polaronic mass here indicates a larger carrier mobility in PTB7. In addition the polaronic mass enhancement that we obtained for doped PTB7 may explain the low hole mobility in its blend with PCBM, which was measured to be 2x10-4 cm2/Vs [124] that limits its application in OPV solar cells. We thus conclude that better growth of the PTB7 film may lead to a more planar structure of the copolymer chains, which, in turn may also increase the OPV solar cell efficiency based on this copolymer. 3.4 Conclusions We studied the most strongly coupled vibrations in the pristine D-A copolymer PTB7 using resonant Raman scattering and doping-induced absorption spectra, and in PTB7/PCBM blend using the photoinduced absorption technique. The six Raman active modes are renormalized when charge polarons are added to the copolymer chains upon doping and photogeneration. They revealed themselves as antiresonant dips superposed on the lower polaron absorption band. We have shown that the amplitude mode model well describes the vibrational modes in all three spectra with a single phonon propagator in which all six modes are coupled together. From the AMM parameters in PTB7, we calculated the polaronic mass enhancement for charge excitation in this compound. The estimation of the polaronic kinetic mass from the application of the AMM to the polaron vibration and electronic spectra may be an effective tool for studying the charge mobility in π-conjugated D-A copolymers that are used in various forms for OPV applications. CHAPTER 4 MAGNETIC FIELD EFFECT IN ORGANIC LIGHT-EMITTING DIODES BASED ON ELECTRON DONOR-ACCEPTOR EXCIPLEX CHROMOPHORES DOPED WITH FLUORESCENT EMITTERS This chapter is a reprint of a paper published in Advanced Functional Materials in the year 2016 authored by Sangita Baniya, Zhiyong Pang, Dali Sun, Yaxin Zhai, , Ohyun Kwon , Hyeonho Choi, Byoungki Choi, Sangyoon Lee and Zeev Valy Vardeny2. A new type of organic light-emitting diode (OLED) has emerged that shows enhanced operational stability and large internal quantum efficiency approaching 100%, which is based on thermally activated delayed fluorescence (TADF) compounds doped with fluorescent emitters. Magneto-electroluminescence (MEL) in such TADF-based OLEDs and magneto-photoluminescence (MPL) in thin films based on donor–acceptor (D–A) exciplexes doped with fluorescent emitters with various concentrations are investi gated. It has been found that both MEL and MPL responses are thermally activated with 2 Sangita Baniya , Zhiyong Pang , Dali Sun , Yaxin Zhai , Ohyun Kwon , Hyeonho Choi , Byoungki Choi , Sangyoon Lee , and Z. Valy Vardeny. Adv. Funct. Mater. 26, 6930, (2016) Published by wileyonlinelibrary.com. Copyright © 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim. Reproduced with permission. 82 substantially lower activation energy compared to that in the pristine undoped D–A exciplex host blend. In addition, both MPL and MEL steeply decrease with the emitter’s concentration. This indicates the existence of a loss mechanism, whereby the triplet chargetransfer state in the exciplex host blend may directly decay to the lowest, nonemissive triplet state of the fluorescent emitter molecules. 4.1 Introduction In the last decade organic light-emitting diodes (OLEDs) have attracted intensive research interest for the displays and solid state lighting applications [15]. The electrically injected charge carriers in OLEDs recombine into singlet and triplet excitons following the respective ratio of 1:3. This limits the internal electroluminescence (EL) quantum efficiency of OLEDs based on singlet exciton emission to 25%, because the emission from triplet excitons is usually spin forbidden [19,125,126]. It has been shown that electrophosphorescence from triplet excitons may be realized in OLEDs via strong spin-orbit coupling, by incorporating heavy transition-metal atoms such as iridium (Ir) and platinum (Pt) in the organic emitter materials [127–130]. However, most of the heavy transitionmetal atoms are expensive and unsustainable that leads to higher cost of commercial products based on phosphorescent OLEDs. Several alternative methods of harvesting singlet excitons from triplet excitons have been recently introduced such as triplet–triplet annihilation and other up-conversion techniques [131–133]. Among these methods, thermally activated delayed fluorescence (TADF) has attracted great interest, because it has the potential to achieve 100% internal quantum efficiency with metal-free organic emitter materials, via reverse intersystem crossing (RISC) [20,34,134–138]. In TADF- 83 based OLEDs, the lowest triplet state (T1) and singlet state (S1) are close in energy, and this leads to thermally induced RISC thereby generating delayed fluorescence from the singlet excitons [28,139–143]. The RISC process occurs in compounds with relatively small spin exchange energy, and thus small singlet–triplet energy splitting, Δ EST [20]. In general, two types of TADF-based compounds have been discussed; these are “intrinsic” and “extrinsic” TADF materials. The former occurs intrinsically in a single molecule, whereas the latter occurs in blend of designed donor (D) and acceptor (A) molecules forming charge-transfer (CT) excitons dubbed exciplexes (EX) [137,139,143–145]. Recently, another promising route for triplet harvesting has been demonstrated by adding fluorescent emitter molecules in TADF-based host compounds. This leads to the next generation of TADF-based OLEDs, namely TADF-assisted fluorescent OLEDs, or super-fluorescent-OLEDs (SF-OLED) [27,35,36,146–150]. In SFOLED, triplet CT excitons undergo RISC to the singlet CT state, which subsequently transfers to the lowest singlet state (S1) of the fluorescent emitter molecules via Förster resonance energy transfer (FRET) process that results in more efficient fluorescent emission [27,35,36,146–150]. SF-OLEDs have shown enhanced device performance and internal quantum efficiency of nearly 100% [27,35,36,146–150]. In 2014, Adachi and co-workers reported SF-OLEDs with external quantum efficiency of 14%–18% for blue, green, yellow, and red emission bands [27]. Subsequently blue, green, yellow, red, and white SF-OLEDs with improved external quantum efficiency and device properties have been developed [35,36,146–150]. In parallel with the studies of SF-OLEDs performance, recently it has been reported that in TADF-based OLEDs there are two spin-mixing processes which may be susceptible to external magnetic field, leading to a remarkable magnetic field effect 84 (MFE) [32,151,152]. These routes are an upper-energy spin mixing process related to polaron-pair species which proceeds via the hyperfine interaction, and a lower-energy spinmixing process that involves exciplexes, via the so called “Δg mechanism” [151]. When an external magnetic field is applied to the TADF-OLED, the combination of the two spinmixing processes dramatically enhances the EL emission, which is particularly attractive in the field of organic optoelectronics [153]. We note that the MFE is not only an alternative method to enhance the EL [151] but also serves as an effective tool for probing spin-related phenomena in organic semiconductor films and devices [40]. In view of energy transfer processes induced by the fluorescent emitter molecules such as FRET and Dexter energy transfer (DET), [35] the exciton dynamics in SF-OLED are more complicated than in pristine undoped TADF-OLED devices. In this work we studied the MFE such as magneto-electroluminescence (MEL) in SF-OLED devices, and magneto-photoluminesce (MPL) in thin films of TADF D–A exciplex-based host blend doped with fluorescent emitter molecules. The host blend is composed of N, N, N ′, N ′-tetrakis(4-methoxyphenyl) benzidine (MeO-TPD) as the donor and tris-[3-(3-pyridyl)mesityl] borane (3TPYMB) as the acceptor. The fluorescent emitter is tetraphenyldibenzoperiflanthene (DBP). We found a decrease in the MFE activation energy (Eact) compared to that in the undoped D–A exciplex blend, which may explain the reported enhanced quantum efficiency in SF-OLED devices. However, as the concentration of emitter molecules increases, we found that both MPL and MEL responses substantially decrease. We explain the MFE decrease with the emitter concentration by inferring the existence of a DET process from the triplet state of the D–A exciplex blend to the lowest triplet state of the emitter molecule (dark exciton), which drains the magneto- 85 RISC (M-RISC) process in the exciplex. 4.2 Experimental Section The blend active layer was prepared using a solution-based method [151]. The donor molecule MeO-TPD (99.99%, Lumtec Corporation), acceptor molecule 3TPYMB (99.99%, Lumtec Corporation), and emitter molecule DBP (99.99%, Lumtec Corporation) were dissolved together in an orthodichloro-benzene solvent at a concentration of 7 mg mL −1. The ratio of D to A molecules in the exciplex blend was kept at 1:4, [151] and the concentration of DBP emitter molecules was varied between 0.5% and 6%. Subsequently the solution was stirred overnight prior to film preparation. The SF-OLED structure was in the form of Indium tin oxide (ITO)/poly (3, 4ethylenedioxythiophene): polystyrene sulfonate (PEDOT: PSS, 40 nm)/D–A exciplex blend with DBP/Ca (20 nm)/Al (100 nm). The SF-OLED devices were fabricated in a glovebox in nitrogen atmosphere. The ITO anode substrate was cleaned by ultrasonic treatment with acetone, 2% micro-90 soap, deionized water, methanol, and propanol, each for 10 min. The PEDOT: PSS layer was spin-coated onto the ITO substrate, followed by an annealing treatment at 140 °C for 20 min. Subsequently the active layer was spin-coated at various speeds. The Ca and Al layers were deposited using a thermal evaporation belljar positioned inside the glovebox. The active device area was 2 mm × 2 mm. For PL and MPL measurements, the thin films were excited using a 40 mW UV laser at 349 nm and the PL emission spectrum was recorded by an Ocean Optics USB4000 spectrometer. The MPL (B) response was measured by applying an external magnetic field, B parallel to the film surface. MPL( B) is defined by the relation: MPL( B) =[PL( B) −PL(0)]/PL(0), where PL( B) [PL(0)] is the PL intensity at field B[or B =0]. MPL at low 86 temperature, T, was measured using a closed cycle He refrigerator cryostat, where T was varied between 30 and 300 K. For the EL spectra and MEL measurements, a positive bias voltage was applied to the SF-OLED ITO electrode. The EL emission was detected using an Ocean Optics USB4000 spectrometer (for the EL spectrum) and a silicon detector (for the MEL response). MEL is defined as MEL( B) =[EL( B) – EL(0)]/EL(0), where EL( B) [EL(0)] is the EL intensity at field B [or B =0]. For EL and MEL at high temperature (T > 300 K), a home-made oven was used. 4.3 Results and Discussion 4.3.1 Magneto-Photoluminescence Studies The molecular structures of MeO-TPD, 3TPYMB, and DBP molecules are shown in Figure 4.1(a). Figure 4.1(b) shows the normalized room temperature (RT) absorption and photoluminescence (PL) spectra of MeO-TPD/3TPYMB blend host and DBP emitter. The PL emission band of MeO-TPD/3TPYMB blend and DBP each covers a broad spectral range showing peaks at ≈418 and ≈694 nm, respectively. We note that the DBP absorption band overlaps with the PL emission band from the exciplexes in the blend host, and this indicates that an effective FRET may occur from 1EX of the host to the S1 state of the molecule emitter. Figure 4.1(c) shows the normalized room temperature PL spectra of DBP-doped MeO-TPD/3TPYMB blend as a function of the emitter concentration, C. The PL spectra contain two emission bands; the shorter wavelength band originates from the D–A exciplex, whereas the longer wavelength band originates from the fluorescent emission of the emitter molecules. As the concentration C increases, the exciplex emission 87 Figure 4.1. a) The chemical structure of MeO-TPD, 3TPYMB, and DBP molecules. b) The optical absorption and photoluminescence spectra of MeO-TPD/3TPYMB blend and DBP molecules. c) The normalized PL spectra of DBP-doped MeO-TPD/3TPYMB (1:4) at various DBP concentrations. d) Schematics of the photoexcitation dynamics and M-RISC process that leads to MPL in D–A blend host doped with fluorescent emitter molecules. band decreases, indicating that more photogenerated 1EX transfer to the emitter molecules. Figure 4.1(d) shows an energy diagram that describes the processes associated with the DBP-doped D–A blend upon photoexcitation. For the pristine D–A exciplex blend, the initial photo excitation decay occurs very fast from the excited donor states to the singlet exciplex 1EX, followed by fluorescent emission. It is tempting to assume that the majority of the photogenerated 1EX produces fluorescence emission; under this condition, however there would not be any magnetic field effect since the singlet emission itself is not senstive 88 to the magnetic field (total S =0). However, the intersystem crossing from 1EX to the triplet exciplex (3EX) is quite feasible due to the small ΔEST between 1EX and 3EX states. This leads to substantial population of 3EX which is generated in parallel with the 1EX emission process. Consequently, magnetic field induced RISC process (M-RISC) from 3EX into 1EX may occur (Figure 4.1(d)) that leads to the obtained MPL response in the D–A exciplex host blend [151]. When the emitter molecules (DBP) are added to the D–A blend host, then the FRET channel from 1EX into the S1 state of DBP becomes available, resulting in an efficient fluorescent emission from the DBP molecules due to their high PL quantum efficiency. In this case the DET process from the 3EX state of the host that may eventually become populated, into the lowest triplet states ( 3S1) of the fluorescent molecule is considered to be a loss mechanism; since 3S1 is a nonemissive, dark state. This may decrease the device performance at higher emitter concentration because the DET process is a short-range energy transfer that becomes stronger with the concentration C [35,154]. Actually, the reported emitter concentration, C in SF OLEDs is usually very small, in order to avoid this loss mechanism caused by the DET process [27,35,146–148] . Figure 4.2(a) shows the RT MPL (B) response of the PL emission bands of D–A exciplex host and the DBP emitter in a film of 1% DBP doped D–A blend. Here MPL( B) is defined by the relation: MPL( B) =[PL( B) −PL(0)]/PL(0), where PL( B) [PL(0)] is the PL intensity at field B[or B =0]. The maximum MPL (MPL max) for the exciplex host and emitter bands are ≈2% and ≈5%, respectively. We note that MPLmax for the exciplex host band is smaller here compared with that of films based on pristine undoped D–A blend (≈3.5%), whereas MPL max of the fluorescent emitter is larger. The decrease (increase) of MPL max from the host (emitter) reveals that there is an effective FRET process from 1EX 89 Figure 4.2. The MPL (B) response of 1% DBP-doped MeO-TPD/3TPYMB (1:4) thin films. a) Room temperature MPL( B) response of the pristine exciplex host (black), pristine emitter, 1% DBP doped exciplex host (gray triangles), and emitter (gray circles) emission bands. b) The MPL (B) response of the emitter PL band measured at various temperatures. c,d) The temperature dependence of MPL max of the exciplex host and emitter PL band, respectively. The insets show the Arrhenius plot of MPLmax (T) of the two emission bands. The linear fits at high temperatures give the activation energy, Eact≈19 ±1 meV for the exciplex host PL band and ≈17 ±1 meV for the emitter PL band, respectively. 90 of the D–A host, of which population is mainly influenced via M-RISC from 3EX, to the S1 of the emitter molecules (Figure 4.1(d)). The obtained MPL response for the emitter emission band also reveals that the FRET process from 1EX to S1 is more efficient than the direct rapid energy transfer from the photoexcited singlet state, Sn of the host molecules into S1 of the emitter, since this latter process would not show any magnetic field effect. Figure 4.2 (b) shows the temperature dependence of the MPL (B) response for the emitter PL band; whereas the MPL (B) response of the D–A host behaves similarly (not shown here). The MPL response of the host (and emitter) PL band monotonically increases with the temperature, especially when the temperature is above 200 K (Figure 4.2(c), (d)), showing a thermally activated behavior related with the M-RISC [151]. The activation energy, Eact extracted from the Arrhenius plot of MPL( T) shown in the inset of Figure 4.2 (c),(d) is ≈19 ±1 meV (host) and ≈17 ±1 meV (emitter), respectively (see Table 4.1). These values are equal within the measurement uncertainty indicating that they originate from the same M-RISC process that occurs in the exciplexes. We note that Eact here is smaller than that of pristine MeO-TPD/3TPYMB blend (≈30 meV) [151]. We speculate that the proximity of the emitter molecules to the photogenerated exciplexes causes a reduction of the activation energy; this is clearly exposed in the doped TADF blend since the emission from the fluorescent emitters preferentially originates from emitters that are closer to the photogenerated exciplexes. Smaller Eact indicates a more efficient RISC process, which may explain the reported literature [27,35,36,146–150]. high quantum efficiency of SF-OLED in the 91 Table 4.1. The room temperature MEL max and E act extracted from MEL( T ); room temperature MPL max of exciplex host band, MPL max of emitter PL band, E act extracted from MPL( T ) of exciplex host band and E act extracted from MPL( T ) of the emitter PL band, at various DBP emitter concentrations. Figure 4.3(a), (b) shows the room temperature MPL (B) response and MPL max of the emitter PL band measured at various emitters’ concentrations (C) ranging from 0.5% to 6%. The obtained MPL max values and calculated Eact as a function of care summarized in Table 4.1. It is clear that the MPL response decreases with C revealing a weakened MRISC process, which exposes the existence of a loss process in the triplet harvesting when the D–A blend is doped with fluorescent emitter molecules at high C. The decrease in MPL and thus M-RISC with Cis also reflected in the MPL max values of the D–A host. As shown in Table 4.1, MPL max of the emitter molecules substantially decreases for C > 2%. This indicates that a DET process from 3EX of the host to 3S1 of the emitter molecules becomes more effective at higher C (see Figure 4.1 d). [35,154] Importantly, the average distance, R between the photoexcited exciplexes and the emitter molecules decreases with C; R ≤1.6 nm for C ≥2%. Consequently the short-range DET rate from 3EX to 3S1 becomes increasingly more effective with C, thereby weakens the M-RISC process that, in turn leads to the reduced MPL response. An alternative explanation that involves a direct rapid energy 92 Figure 4.3. a) The room temperature MPL( B) response of the emitter PL band in DBPdoped MeO-TPD/3TPYMB (1:4) thin films measured at various emitter concentrations. b) The MPL max values extracted from (a) versus emitter concentration. transfer from the photoexcited singlet (Sn) of the host molecules to 1S1 of the emitter molecules at high C can be excluded for the following reasons: (i) the overall PL intensity, which is composed of emission contributions from the exciplex host and emitter molecules, decreases monotonically with C. (ii) MPL response of the emitter PL band at high C still shows a thermally activated behavior (Table 4.1 ) indicating that the emitter emission mainly originates from a slow RISC process rather than a rapid energy transfer from Sn of the host molecules. 4.3.2. Magneto-Electroluminescence Studies Figure 4.4(a) depicts the I– V and EL-V responses of the 1% DBP doped MeOTPD/3TPYMB (1:4) SF-OLED device; the device structure is presented in the inset. We note that the operating voltage of the SF-OLED device is lower than that of the pristine undoped MeO-TPD/3TPYMB exciplex-OLED devices, [151] indicating that the 93 Figure 4.4. a) I– V and EL-V response of SF-OLED device based on 1% DBP-doped MeO-TPD/3TPYMB (1:4). The inset shows the device structure. b) The EL spectra of MeO-TPD/3TPYMB with and without DBP dopants. c) Schematic of various electrically injected excitation mechanism in DBP-doped exciplex blend. The notations are the same as in Figure 4.1 (d). 94 fluorescent emitters enhance the device performance. Figure 4.4 (b) shows the normalized EL spectra of the DBP-doped SF-OLED device compared to that of the undoped exciplexOLED device. The EL spectrum of the exciplex-OLED is relatively broad and featureless with maximum at ≈550 nm showing typical characteristics of a delayed fluorescence, and this band is effectively suppressed in the EL spectrum of the SF-OLED device upon doping. In addition the EL band of the SF-OLED device is much narrower showing a peak at ≈602 nm followed by a vibrational replica at 660 nm indicating that an efficient FRET process from the D–A exciplexes host to the DBP emitter molecules may occur. Figure 4.4(c) shows a schematic of the electrical-excitation dynamics processes which influence the EL emission and MEL properties in the SF-OLED devices. MEL is defined here as MEL(B) =[EL(B) −EL(0)]/EL(0), where EL( B) [EL(0)] is the EL intensity at field B[or B =0]. The injected electrons and holes in the D–A host blend firstly form exciton species ( 1S n and 3S n) that subsequently decay into respective exciplex states 1EX and 3EX on the D–A host molecules, which follow a ratio of 1:3. In the case of undoped pristine D–A exciplex host, thermal (and magnetic-related) RISC to 1EX state occurs that enhances the EL efficiency of the device [151]. When the DBP emitter molecules are introduced into the D–A host blend, the electrically formed 1EX on the host D–A molecules may reach the 1S1 state of the DBP molecules via an effective FRET process that results in efficient fluorescent EL emission from the emitter molecules. On the other hand the electrically generated 3EX may decay via a DET process to the dark 3S1 of the emitter molecules thus diminishing the M-RISC process, similar as the loss process that we identified above in the PL and MPL responses. This may explain the reduced MEL response at high C. 95 Figure 4.5(a), (b) shows the MEL(B) response of 1% DBP-doped MeOTPD/3TPYMB (1:4) SF-OLED device as a function of the applied bias voltage. It is clear that MEL increases with the bias voltage, reaches a plateau and then decreases slightly at higher voltage. The maximum response is observed around 4.5 V, which may originate from saturation of the FRET process at higher current density due to the limited emitter molecules density. We note that the maximum MEL (MEL max) in a device is larger than the MPL max obtained in the same DBP-doped MeO-TPD/3TPYMB films. This indicates that the upper spin mixing channel is also active in MEL, but not in MPL [40,151]. Figure 4.5(c) shows the MEL( B) response of the 1% DBP-doped SFOLED as a function of temperature; the resulting temperature dependent MEL max is plotted in Figure 4.5(d). We note that MEL max increases with the temperature, similar as MPLmax( T) response shown in Figure 4.2 (d). The inset of Figure 4.5(d) shows an Arrhenius type plot from which Eact of ≈14.6 ±0.4 meV is extracted. This value is consistent with the value obtained from MPL (T), and furthermore confirms that Eact here is much smaller than that of the undoped TADF-OLED (where Eact≈25 meV was extracted from MEL(T) response) [151]. Figure 4.6(a) shows the MEL( B) response of DBP-doped MeOTPD/3TPYMB (1:4) SF-OLED devices as a function of the DBP concentration, C. The extracted MEL max values and corresponding Eact values are summarized in Table 4.1. It is clear that MEL max steeply decreases with C, as seen in Figure 4.6(b). The short-range DET process is the most probable cause that explains this MEL( C) characteristic behavior, [155,156] similar as in MPL( C) discussed above. The direct carrier formation and recombination on the DBP emitter molecules that increases with C, may be another possible loss process that reduces the MEL response here [35]. 96 Figure 4.5 a) MEL( B) response of 1% DBP-doped MeO-TPD/3TPYMB (1:4) SF-OLED at various bias voltages, V, measured at room temperature. b) MEL maxextracted from (a) versus voltage. c) MEL( B) response at various temperatures. d) The temperature dependence of MEL max extracted from (c). The inset shows the Arrhenius plot of MEL max( T). The linear fit at high temperatures gives the activation energy, Eact≈14.6 ±0.4 meV. Figure 4.6. a) The room temperature MEL( B) response of DBP-doped MeOTPD/3TPYMB (1:4) SF-OLED at various DBP emitter concentrations. b) Room temperature MEL maxat various DBP emitter concentrations extracted from (a). 97 The MPL(B) and MEL(B) responses in 1% DBP-doped MeO-TPD/3TPYMB (1:4) were also measured at temperatures T >300 K using a home-made oven, as shown in Figure 4.7(a), (b). We note that compared to the low temperature measurements, both MPL max and MEL max values are reduced here; this is due to the limited available magnetic field strength (≈50 mT) using our home-made optical-oven setup. MPL max increases with T in the range of 300 K < T< 320 K showing the same thermally activated behavior as for T< 300 K that is caused by the M-RISC process. As the temperature increases further, however MPL max remains unchanged revealing that the doped D–A film are more stable than the undoped counterparts. The MEL( T) measurements in the range of 300 K < T< 320 K also confirm the good device stability; [27], [36] this behavior in MEL( T) of SF-OLEDs contrasts MEL( T) in pristine TADF-OLED in which the MEL steeply decreases with T for T >300 K. [151] This result is consistent with the reported good operational stability of SF-OLED devices. [27], [36]However for T >320 K MEL steeply decreases, in contrast to MPL. The different behavior of MPL (T) and MEL (T) responses indicate that in addition to the thermally activated behavior from TADF exciplex molecules which increases MELmax, there exist other processes that influence the MEL response in the SF-OLED which are triggered at high temperatures. For example, the electrical transport may be activated at higher temperature. Also at high temperature nonradiative recombination channels that compete with the TADF emission may be also activated that significantly reduce the MEL response in SF-OLED, but do not affect the MPL response in thin films. Another possibility for the MEL reduction at high T may be from excess 3EX loss channels during the operating device at higher temperature, which do not exist in thin films. 98 Figure 4.7. a) MPL max of 1% DBP-doped MeO-TPD/3TPYMB (1:4) thin film versus T measured at T >300 K. b) Same as in panel (a) but for MEL max measured in a SF-OLED based on the same DBP doped exciplex blend compound as in panel (a). The maximum magnetic field strength is 50 mT, limited by the oven setup. 4.4 Conclusions We studied MPL and MEL responses in thin films and SFOLED devices based on D–A exciplex blend host of MeOTPD/3TPYMB (1:4) doped with DBP fluorescent emitter molecules at various concentrations, C. We found that both MPL and MEL are activated at high temperature exhibiting a smaller activation energy (Eact) than that in pristine undoped TADF-OLED. The rapid FRET process from the exciplex 1EX of the exciplex host blend to the singlet state ( 1S1) of the emitter molecules leads to enhanced fluorescence but smaller MEL and MPL responses compared to those in films and devices based on the pristine undoped exciplex blend counterpart. As C increases, we identified a new loss mechanism, namely a DET type process from 3EX of the exciplex host to the triplet state (3S1) of the DBP molecules that competes with the original FRET process from the 1EX to the emitter 1S1 and results in a decrease of both MPL and MEL responses in the doped D– 99 A blend. We believe that our finding may be an alternative approach to achieve higher luminous efficiency in OLEDs and may further improve their overall performance. CHAPTER 5 OPTICAL STUDIES OF TWO-DIMENSIONAL LAYERED HALIDE HYBRID PEROVSKITE SEMICONDUCTORS Two-dimensional (2D) layered hybrid organic-inorganic halide perovskite semiconductors form natural “multiple quantum wells” that have strong spin-orbit coupling due to the heavy elements in their building blocks. This may lead to “Rashba-splitting” close to the extrema in the electron bands. We have used a plethora of nonlinear optical spectroscopies and theoretical calculations to study the primary (excitons) and long-lived (free carriers) photoexcitations in thin films of 2D perovskite, namely, phenethylammonium lead iodide (C6H5C2H4NH3)2PbI4. The density functional theory calculation shows the occurrence of Rashba-splitting in the plane perpendicular to the 2D barrier. From the electroabsorption spectrum and photoinduced absorption spectra from excitons and free carriers, we obtain a giant Rashba-splitting in this compound, with energy splitting of (40 ± 5) meV and Rashba parameter of (1.6 ± 0.1) eV·Å, which are among the highest Rashba-splitting size parameters reported so far. This finding shows that 2D hybrid perovskites have great promise for potential applications in spintronics. 101 5.1 Introduction Large Rashba-splitting with αR of few eV·Å has been observed in only a handful of materials such as ultrathin metallic films [71], surfaces of topological insulators such as Bi2Se3 [73], and surfaces of the polar semiconductor BiTeI [72]. We note that Rashbasplitting is very small in traditional III-V semiconductors. For example, the Rashbasplitting parameter αR at the interface of InAlAs/InGaAs was measured to be ≈0.07 eV·Å with ER ≈1 meV [157]. In contrast, here we discovered a giant size Rashba-splitting in a 2D semiconducting layered hybrid organic-inorganic perovskite. The 3D hybrid organic-inorganic perovskites (hereafter hybrid perovskites) such as methyl-ammonium lead halogen (MAPbX3; X is halogen) have recently attracted immense attention because of several promising optoelectronic device applications such as photovoltaic solar cells, light emitting diodes, and lasers [158,159]. Importantly, these compounds possess strong SOC due to the heavy elements (Pb, X) that determine the electron bands near their extrema points, which may lead to large Rashba-splitting if the structure lacks inversion symmetry [160]; this is in fact realized in some 2D and 3D hybrid perovskites [161]. Interest in the hybrid perovskites for spintronics applications has only recently begun, with promising spin and magnetic field phenomena [40,162]; therefore, solid evidence of Rashba-splitting existence in these compounds would further boost this interest. Recently, the 2D hybrid perovskites have come into focus [163,164]. These compounds self-assemble into alternating organic and inorganic layers that form natural ‘multiple-quantum wells’, with outstanding optoelectronic characteristics and stability at ambient conditions. Having strong SOC, high-charge mobility, and intrinsic quantum well 102 structures with many interfaces and facile solution processability, the 2D hybrid perovskites may be promising candidates for room-temperature spintronic applications [165]. Our work shows that these materials may also exhibit giant Rashbasplitting, which could further enhance their appeal for this field. 5.2 Experimental 5.2.1 Sample Preparation The 2D hybrid perovskite films were fabricated in a nitrogen-filled glove box with oxygen and moisture levels < 1 ppm. We mixed R-NH3I (where R is C6H5C2H4) and PbI2 in a 2:1 mole ratio in N,N-dimethylformamide to form solutions with a concentration of 0.5 mol/ml. The solutions were spin-coated on an Oxygen plasma pretreated sapphire or KBr substrate at 420 rad/sec and 90 s, to form 100 nm thick films; the obtained films were subsequently annealed at 100°C for 30 m. The purchased chemicals were used without further purification. 5.2.2 Electroabsorption (EA) Measurements We used a 2D perovskite film fabricated on a substrate with patterned metallic electrodes. The EA substrate consisted of two interdigitated sets of a few hundred of 10 μm wide gold electrodes, which were patterned on a sapphire substrate. The device was placed in a cryostat for low temperature measurements. By applying a potential, V, to the electrodes, a typical electric field, F ≈ 105 Volt/cm, was generated with V = 300 Volts and f = 1 kHz parallel to the film. That is, the AC voltage bias was applied in the longitudinal geometry with the perturbing electric field parallel to the [PbI6]4− film. For probing the EA spectrum, we used an incandescent light source from a Xe lamp, which was dispersed 103 through a monochromator, focused on the sample, and detected by a UV-enhanced silicon photodiode. We measured the changes, ∆T, in the transmission spectrum T using a lockin amplifier, set to twice the frequency (2f) of the applied field, and verified that no EA signal was observed at f or 3f. We note that detection in the second harmonic can be useful for materials that may exhibit piezoelectricity, since these effects are linear and hence are not present in the quadratic EA signal. ∆T and T spectra were measured separately and the EA spectrum was obtained from the ratio ∆T/T. 5.2.2 CW Photomodulation (PM) Measurements The excitation pump was provided by a diode laser with ћω=2.8 eV, and the probe beam was provided by an incandescent tungsten/halogen lamp (for the visible-near infrared) or a globar light source (for mid-infrared). The sample films grown on KBr substrates were put in a cryostat under vacuum. The sample temperature was varied from 50 K to 300 K. The pump and probe beams were overlapped on the sample films, and the transmitted probe beam was filtered through a monochromator and detected by a Si, InGaAs, or HgCdTe detector for different probe spectral ranges. The transmission, T , and change of the transmitted probe beam (ΔT), which was caused by the modulated pump beam, was detected by a lock-in amplifier. The PA spectrum was then calculated from ΔT/T. 5.3 Results and Discussion We have employed a variety of steady state linear and nonlinear optical spectroscopies as well as theoretical calculations for studying the optical characteristics of 104 the primary (excitons) and long-lived (free-carriers) photoexcitations in 2D hybrid perovskite films [166]. Figure 5.1 (a) shows the molecular structure of the 2D hybrid perovskite that we studied here, namely (C6H5C2H4NH3)2PbI4 (PEPI). Figure 5.1(b) shows schematically the 2D hybrid perovskite where the organic (C6H5C2H4NH3+) and inorganic [PbI6]4- octahedron layers form ‘multiple quantum wells’ with thickness of ≈1 nm (barrier) and ≈0.6 nm (well), respectively [67]. The spectroscopies and model calculations employed in this work provide compelling evidence that the continuum band-edge above the exciton level (CB bottom and/or VB top) possesses surprising optical characteristic properties, which result from a large Rashba-splitting energy, ER≈40 meV. Our work provides a general all-optical method for studying the Rashba-splitting effect in semiconductors. The room-temperature photoluminescence (PL) and absorption spectra of the 2D hybrid perovskite PEPI film are dominated by an exciton band at ≈2.4 eV with large oscillator strength, consistent with a relatively large exciton binding energy, EB of ≈0.2 eV (Figure 5.1(c)) [167], followed by a slow increase in the absorption with an onset at ≈2.6 eV. From the temperature-dependent absorption measurement, we observed two peaks at the exciton band at low temperatures (T<110 K). We determined that the main peak originates from the 1s exciton absorption and side peak arises from the phonon side-band due to electron-phonon interaction. The phonon side-band of about the same values has been reported previously in both single crystals and thin films of 2D perovskites [168] . Moreover, the absorption spectrum shows two step-like absorption edges in the spectral range of 2.45 eV to 2.65 eV (see Figure 5.2a). The step-like feature located at ~2.53 eV is assigned as the 2s exciton absorption [68]. The second step-like feature located from 2.55 105 Figure 5.1. PEPI having alternating organic (C6H5C2H4NH3+) and inorganic ([PbI6]4- layers that form ‘multiple quantum wells’. (a) The molecular structure and (b) schematical structure. (c) The absorption (Abs) and photoluminescence (PL) spectra of PEPI film at room temperature (RT). eV to 2.60 eV may be interpreted as the interband (IB) transition, from which one can get the information of bandgap EIB (Eg). We can extract the binding energy (Figure 5.2b) from the difference between exciton resonant energy and bandgap energy EIB. However, this bandgap energy is not sharp enough for us to extract the exact values even at 10K. We thus used a modulation spectroscopy, namely the electroabsorption (EA), to separate these delicate absorption features (namely EIB, E1s and E2S ) from the broad spectral background. 106 Figure 5.2. Absorption and energy levels of PEPI film. (a) Absorption spectra at 10K, 110K, 180K, and 290 K. The 1s and 2s exciton (E1s and E2s, respectively) and an interband transition (IB) are assigned. (b) Estimated energy levels of the excitons (E1s and E2s) and interband transition (E(IB)) are assigned with respect to the ground-state (GS). (Figure 5.2(a) Measured By Yaxin Zhai) 107 5.3.1 Electroabsorption Spectra Modulation spectroscopy techniques such as EA spectroscopy provide distinctive and much sharper derivative-like spectral features, and thus can often separate the exciton from the continuum band. In general, EA spectroscopy enhances the “fine structure” in the absorption spectrum, because it is extremely sensitive to Coulomb correlation effects in the excitons, as well as surface/interface electric fields [169]. The effect of an electric field on excitons differs greatly from that on delocalized states at the band-edge, as illustrated in Figure 5.3 (a). The absorption spectra of an exciton and band-edge states are shown with and without an applied electric field. We anticipate two distinct electric-field-induced effects, namely on the exciton and band-edge states; These are described by the Stark effect and Franz−Keldysh (FK) oscillation, respectively. The Stark effect can be thought of as the shifting of a localized state (i.e., excitons) in the presence of an electric field. The change in the exciton energy, ΔE, with the field, F, is described as, 1 ∆, w ∝ −∆9w − ∆•w 2 5.1 where Δμ is the change in dipole moment and Δp is the change in polarizability [170]. Depending on the sign of the polarization change, the applied electric field interacts with that state either positively or negatively, which shifts the energy of that state to higher or lower energies, respectively. If all excitons in the perturbed sample are polarized in the same manner, then we can expect the entire exciton absorption band to shift uniformly. This uniform shift corresponds to an EA spectrum in the form of the first-derivative of the absorption spectrum with respect to the photon energy, as shown in Figure 5.3 (b). This is typical of tightly bound excitons with large binding energy, such as the 1s exciton 108 Figure 5.3. Schematic illustration of (a) expected absorption response of a direct gap semiconductor having relatively large exciton binding energy upon the application of an external electric field. The exciton undergoes a red shift due to the Stark effect, whereas the interband absorption onset acquires a pronounced oscillation caused by the FK effect. (b) EA spectrum showing a derivative-like feature caused by the Stark effect, and damped oscillation due to the FK effect. 109 in 2D semiconductors [68]. By contrast, if the perturbed states do not show a uniform polarization change, then the subpopulations of those states will interact with the applied electric field in discrete ways; this results in inhomogeneous broadening of the absorption feature. In the electroabsorption spectrum, this leads to line shape that resembles the second-order derivative of the absorption spectrum with respect to the photon energy, which is indicative of weakly bound states such as defects or trap states. From equation (5.1), it is clear that, at sufficiently high applied fields, the amplitude of the absorption signal will follow a F2 dependence. As such, the physical origin of the Stark effect in a quantum well material, such as (C4H9NH3)2PbI4, is in the separation of the electrons and holes to either side of the quantum well; this is known as the quantum confined Stark effect (QCSE) [171]. The charge separation results in a decrease in the overlap of the wave functions of the electron−hole pair, which produces a redshift in the absorption energy of an exciton. Conversely, the effect of an electric field on delocalized states near the continuum band-edge in a direct gap semiconductor is described by the FK effect. Near the fundamental bandgap, when a uniform electric field (F) is applied along a linear axis, the change can be represented by variations in the dielectric constant: ∆\ ≈ E∆@[ \ 5.2 where a is the Seraphin coefficient and εr is the dielectric constant. The change in dielectric constant is related to the applied electric field through the electro-optical energy (hθ): where ℎÈ H = - • ℏ• h • † ∆Z[ , − ,0 , w ∝ ℎÈ 5.3 and μ is the reduced mass at the band-edge [172,173]. The result of this relation is oscillations in the absorption spectrum (see Figure 5.3 (b)) that cross the unperturbed absorption spectrum and broadens as F2/3, whereas the EA amplitude increases 110 as F1/3. Consequently, the signature of the FK effect in the EA spectrum close to the energy gap is oscillatory spectral response that broadens with F2/3 and amplitude that follows F1/3. Although FK oscillations have been investigated in InGaAs/GaAs quantum wells [172], understanding the observed line shapes in the absorption that are related to this effect is challenging. Various factors such as energy-dependent broadening, Coulomb interaction, surface effects, photovoltaic effects, nonflat modulation, nonuniform field effect, among others, are known to affect the line shape of the FK oscillations. To date, there are no theoretical models that are able to accurately describe the FK oscillations observed in 2D quantum wells. Several studies have utilized an oversimplified theory that, although it provides a good fit, it is somewhat misleading due to the inability to capture the complete physical contributions of Coulomb interactions and surface effects. The room-temperature EA spectra of 2D PEPI film as a function of field strength are shown in Figure 5.4. Clear EA features occur at the exciton peak with zero-crossing at Eex = 2.38 eV, and near the band-edge at Eg = 2.53 eV, consistent with the EA spectra of similar 2D organic-inorganic hybrid perovskite [68]. The inset shows the high energy EA spectrum at E > Eg; the FK oscillations are not clearly visible at room temperature due to wavefunction localization caused by defects and imperfections. Nevertheless, the EA line shape does suggest there are oscillations in the spectrum that are being quenched. At low temperature, these interfering effects are subdued and consequently, the FK oscillations are more clearly visible. From the energies of the band-edge and exciton extracted from the EA analysis, we determine the exciton binding energy is Eg − Eex = Eb = 150 meV. From the obtained Eb, we can now calculate the reduced exciton mass for the room temperature 2D PEPI crystal phase using the 2D hydrogen model: 111 Figure 5.4. Electroabsorption spectra for various field strength measured at 2f. Inset is the high energy spectra range where FK oscillation is not resolved. ,$ = 9= • 8ℎ @[ @ 5.4 where μ is the reduced mass, e is the electron charge, h is Planck’s constant, and ε0 is the permittivity of free space. εr is the dielectric constant calculated as per the following equation: @[ = @É ¿É + @$ ¿$ ¿É + ¿$ 5.5 where L and ε are the width and dielectric constant. For the inorganic [PbI6]4− well, εw = 6.1 and Lw = 0.64nm, and for the organic barrier, C6H5C2H4NH3+ εb = 2.5 and Lw = 1nm, 112 respectively [165,174]. As such, the εr for (C6H5C2H4NH3)2PbI4 is calculated to be 3.91, which yields an exciton reduced mass of μ =0.169m0 for the room-temperature 2D PEPI crystal phase. To investigate the electronic structure near the band-edge of the low-temperature 2D PEPI crystal phase, EA measurements were carried out at low-temperature (50K) EA spectra at various ac field strengths as shown in Figure 5.5 (a). Based on the EA dependence on the field strength (or V, the applied voltage), we indeed identify two distinctive EA spectral ranges. The EA scales with V2 (Figure.5.6(a)-(c)) for ħω<2.55 eV [175]; however, it saturates at large field for ħω>2.55 eV (see Figure 5.5 (c) (inset)). The EA spectrum in the low-energy spectral region (<2.55 eV) shows a ‘first derivative-like’ feature consistent with an exciton Stark shift, having a zero-crossing at 2.38 eV that we assign as the 1s exciton energy, E1s (Figure 5.2(a)). We note that in this spectral region, there is a second ‘derivative-like’ feature with ‘trending’ zero-crossing at ≈2.53 eV, in agreement with the lowest step-like feature in the absorption spectrum (Figure 5.2(a)). We identify this EA feature as originating from the 2s exciton in PEPI [176], as shown in Figure 5.5(e) [161,177]. In contrast, the large oscillatory-like EA feature at ħω>2.55 eV having multiple zero-crossings is due to FK oscillation above the direct band-edge [178], [179]. These features are commonly observed in polycrystalline thin film samples, and they represent unassigned energy transitions into the continuum energy states, high above the band-edge. Importantly, the oscillation energy period, δE, shows ‘field-broadening’ that scales with V2/3 (see Figures 5.5(b), 5.5 (c), and 5.5(d)). This broadening and the EA saturation at large V are typical characteristic properties of the FK oscillation in the EA spectrum, unique 113 Figure 5.5. Electroabsorption (EA) spectra of PEPI film. : (a) EA spectra of PEPI measured at 50K at various applied electric fields (~ the applied voltage, V). Various EA spectral features are assigned, where FK stands for the Franz-Keldysh oscillation. (b) EA spectra close to the zero-crossing energy measured at various field strengths; broadening of the FK oscillation is clearly seen. “d” symbolizes the high-energy FK oscillation that blue-shifts with increasing field. (c) Field broadening of the EA features related to the FK oscillation; ‘a’, ‘c’, and ‘b’ and ‘d’ are assigned as zero-crossing energies and peak positions, respectively. The inset shows the peak values of EA vs. V2 of bands ‘b’ and ‘d’, which saturate at large V. (d) The energy difference δEac and δEbc plotted vs V2/3. (e) Energy levels of the excitons (E1s, E2s) and interband transition (E(IB)) are assigned with respect to the ground state (GS). 114 Figure 5.6 The dependence of the EA signal on V2 at various energies below the IB edge at 2.55 eV, where V is the applied voltage. 115 to bandgap modulation, therefore confirming that this feature is from the band-edge (Interband (IB) transition in Figure 5.5(a)), and not from another exciton state where the peak at energy below the first zero-crossing determines the energy gap value. We thus locate the band-edge of the PEPI film at 2.57 eV (Figure 5.5(a)). From E(IB), E1s, and E2s, we can now obtain the 1s and 2s Wannier exciton binding energies: Eb(1s) = (190 ± 8) meV and Eb(2s)= (45 ± 8) meV, respectively. The uncertainty originates mainly from the optical resolution of our spectrometer (output resolution is 2 nm, due to the entrance slit width of 100um) and the film inhomogeneity; all uncertainties are reported as one standard deviation. At low temperature, the organic barrier width changed and d spacing reduced by 1.3Å, hence Lw = 0.87nm. εr for (C6H5C2H4NH3)2PbI4 is calculated to be 4.02; this yields a exciton reduced mass μ =0.227m0 for the low-temperature 2D PEPI crystal phase. This value is comparable to the reduced mass extracted using DFT calculation. The polarizability (Δp) of the 1s exciton was calculated from the EA spectra via: ∆, = 1 ∆•w 2 5.6 where ΔE is the Stark shift, and F is the applied field. Fields of 90 kVcm−1 shift the exciton energy by about 10 meV (since without the field, the exciton peak is at 2.4eV, but with the field, the 1st zero crossing for 1s exciton is at 2.39eV), which corresponds to a polarizability of Δp ≈ 3.95 × 10−31 Cm2V−1 in SI units. The applied electric field perturbation is parallel to the [PbI6]4− sheets; therefore, this large polarizability indicates that the exciton in this 2D perovskite is rather extended in the ab plane. This is verified by DFT calculation in section 5.3.3. As the interlayer spacing in (C6H5C2H4NH3)2PbI4 decreases, the exciton binding energy increases due to increased quantum confinement effects on the excitons. As the 116 quantum well width narrows, there is a smaller amount of dispersion in the bands, resulting in a widening of the bandgap [69]. Due to quantum confinement of the 1s exciton, application of an electric field redshifts the exciton absorption, which is described by the Stark effect [170]. The interlayer spacing of (C6H5C2H4NH3)2PbI4 is significantly smaller than the exciton radius; therefore, we are able to observe significant Stark shift to occur (Figure 5.5) [69]. On the other hand, the bandgap modulation is dominated by the FK effect, evident by the F2/3 broadening with the field. To the best of our knowledge, this is the first evidence of the FK effect in (C6H5C2H4NH3)2PbI4 quantum wells. ΔT/T. 5.3.2 Free Carrier Absorption (FCA) Furthermore, we expect that the presence of Rashba-splitting would affect the process that leads to free carrier absorption (FCA). To investigate this assumption, we studied the properties of long-lived photoexcitations in PEPI using the technique of steady state PM . The long-lived photoexcitations should be free-carriers since the excitons have a sufficiently long time to ionize into free electrons and holes that can contribute to photocarriers, especially at grain boundaries. This has been verified in PV cells based on 2D perovskites that have shown power conversion efficiency larger than 10 % [163]. In this case the PM spectrum would be due to photogenerated FCA. To verify that we can indeed measure photoinduced FCA by our PM technique, we measured, as a ‘control experiment’, the steady state PM spectrum in crystalline Si at 300K and 45 K, as seen in Figure 5.7 (a),(b). We could readily fit this PA spectrum by a Drudetype FCA response in which the PA spectrum varies as ω-2 [180]; this validates our approach. Similarly, PEPI film on top of Si substrate, also shows the Drude type FCA, 117 Figure 5.7. Steady State Free carrier absorption from Si substrate at (a) 300K and (b)50K, respectively. Red line is the fitting with Drude model (1/ω2). which is due to exciton dissociation in the Si substrate, resulting in free carrier absorption shown in Figure 5.8 (a),(b),(c). In contrast, the steady state PA spectrum (PAFCA) in PEPI film grown on KBr substrate shows a sharp dip at low photon energy, forming a peak at ħω≈0.15 eV 50K. A slight asymmetry is also observed for the PAFCA band in Figure 5.8(d), which suggests that the electron quasi momentum k near the bottom of the lower Rashbasplit band at k0 may also extend in k-space. This may be due to localization or the quasiFermi level of electrons in this band. We consider this surprising FCA(ω) response a ‘smoking-gun’ verification of the Rashba-splitting that exists in the PEPI continuum band. Since optical transitions within the same branch can only be Drude-like, that is allowed because of the mixture of s and p states in the CB [166]; we ascribe this PA band to FCA with an onset at the vertical transition from the bottom of one branch of the CB into the other branch, as shown 118 Figure 5.8 Steady state photomodulation spectroscopy of PEPI film on Si substrate and KBr substrate excited at 2.8 eV. (a), (c) PM spectrum of PEPI film on Si and (b), (d) on KBr substrate measured at modulation frequency of 310 Hz, and temperature of 300K and 50K, respectively. The solid lines through the data points are fits using the Drude model (PA~ω-2) for the 2D Iodide on Si wafer, whereas we use equation (5.10) for PAFCA spectrum of 2D on KBr substrate. (e) Schematic electron energy bands with Rashbasplitting that explains the FCA in PEPI. The Rashba energy (ER) and momentum offset (k0 = ∆q/2) are assigned. 119 schematically in Figure 5.8 (e). This transition is allowed since it involves states of nonzero k value for the upper and lower spin-split branches of the CB and therefore mixture of s and p states, which is explained in section 5.3.2.3. As described in section 5.3.2.1, equation (5.10) is used to fit the asymmetric PAFCA spectrum where m* is the electron effective mass, and ∆0 is the transition energy considering a Gaussian distribution F(∆-∆0) of ∆ around ∆0 having width δ∆. From the fit, we get ∆0 = 0.15 eV, and the distribution width δ∆= 0.03 eV. The “localization length (l)” or “wavefunction extent” of the electron that we obtained from the fit is ~ 14 nm. This shows that the electron is barely localized, consistent with the free-electron model of electron in conduction band. We have performed temperature-dependent PAFCA whose intensity decreases with increasing temperature and vanishes at 300K as shown in Figure 5.9. From the FCA peak at 0.15 eV, we obtain Rashba-splitting energy, using ∆E=4ER, which is ER≈(38 ± 3) meV. Another source of uncertainty in this measurement is the unknown value of the quasi-Fermi level at the CB bottom, which may well be of order 2-3 meV at low temperature. We further study the excited-state properties of PEPI close to the continuum band minima. The transient PA (PA1) band at mid-IR from the excitons into the continuum band peaks (Figure 5.10(a)) at (350 ± 2) meV (the uncertainty comes from the 150-fs pulse duration), which cannot be ascribed to the vertical transition from the 1s exciton into the lowest continuum band because from our EA studies, a transition into the lowest continuum band should appear at 190 meV. We therefore assign PA1 to an optical transition from the 1s exciton to a second, upper electron continuum branch, which is split from the lower band by Rashba SOC, as shown schematically in Figure 5.10(b). 120 Figure 5.9. Temperature-dependent PM spectra of PEPI film on KBr substrate. Figure 5.10. PA1 transition in PEPI film. (a) Asymmetric PA1 band is fitted by “k-space extension model” in equation (5.10), yielding the delocalization length of exciton, l~16 nm. (b) Schematic energy diagram with Rashba-splitting that explains the PA1 transition. The Rashba energy (ER) may be obtained from the 1/4 of energy difference (∆E) between PA1 transition and the 1s exciton binding energy (Eb). 121 We can now obtain the Rashba-splitting energy, ER, from the PA1 band at 350 meV, because PA1 should be pushed to higher energies by an energy, ΔE =4ER, namely E(PA1)= Eb+4ER (Figure 5.10(b)). Using this relation, Eb (=190meV from the EA spectrum) and E(PA1)=350meV (from the transient PA spectrum), we determine ER=(40±5)meV in 2D PEPI, which is in good agreement with the value obtained from the Free carrier absorption measurement. We note that the PA1 spectrum is asymmetric; this may be due to the transition from the exciton discrete level to the continuum band, where the exciton wave function is spread in k-space by a “k-localization length,” Δk, as determined by its localization length l in real space and Δk ≈ 1/ l. The solid line through the PA1 data points is a fit using the optical transition model for a 2D semiconductor described in equation (5.10) from which we obtain the exciton localization length l ≈ 10 nm. This relatively large l value indicates that the exciton is quite delocalized in the quantum well in a direction perpendicular to the barriers but localized in k-space. From the obtained ER value, we can readily estimate the offset, k0 in the momentum space using a parabolic dispersion relation with an electron effective mass obtained using density functional theory m* = 0.25 m0 (where m0 is the bare electron mass); we thus obtained k0= (0.051 ± 0.004) Å-1. Consequently, we estimate the Rashba-splitting parameter αR=(1.6 ± 0.1) eV·Å. These values are comparable to the recently measured Rashba parameters in MAPbBr3 using the surface-sensitive angle-resolved photo-electron spectroscopy [181]. 122 5.3.2.1 Fitting the FCA and Transient PA1 Band In the momentum (k) space of a quantum well, the energy-dependent absorption coefficient α(E) of electron from the exciton state into the continuum can be expressed equation (5.9) as, , ∝ 1 Ê s C|6 C | ËF ,-. C , − F ,W$ C Ì ,-. C − ,W$ C − , 5.7 where Eex and Ecb are the energy of the exciton and continuum band (CB or VB), respectively, f(E) is the Fermi-Dirac distribution function, and M is the matrix element for the transition. The exciton wave function is localized in real space over a distance, l, and thus the exciton is not a δ-function in k-space, but instead is spread over a range, ∆k in k, where ∆k≈1/l. Therefore, we have to integrate over k< ∆k , ∝ 1 Ê s X|6 X | | , X |  ,-. X − ,W$ X − , 5.8 where C(q) is a ‘mixture function’ that is spread over ∆k, and determines the k-space extension of the exciton wavefunction. We replace the Fermi-Dirac distribution function with a step function at Eb, if the exciton temperature is low compared to exciton binding energy, Eb. If we assume that the exciton wave function in real space has the form of ∝exp(r/l), where l is the ‘localization length’, then C(q) is given by the Fourier transform of the real space extension, X ∝ 1+Í X 5.9 Consequently, the k-space exciton extension, ∆k =l-1, and the absorption, α(E), for a 2D system can be then expressed by equation (5.10): 123 , ∝ ,−∆ , 1 Î ,−∆ + ħ 2Í ?∗ Ï 5.10 where m* is the electron effective mass, and ∆0 is the transition energy. We calculated the exciton PA using m* = 0.25 m0, where m0 is the free electron mass [equation (5.11)], and ∆0 = 0.35 eV. The asymmetric PA1 band is fitted using equation (5.10) considering a Gaussian distribution F(∆ -∆ 0) of ∆ around ∆ 0 (see Figure 5.10(a)) having width δ∆ . From the fit, we get the exciton localization length, l=10 nm, ∆ 0 = 0.35 eV, and the distribution width δ∆ = 0.05 eV. 5.3.2.2 Density Functional Theory (DFT) Calculations To help understand the origin of the Rashba spin-orbit splitting of the CB, our collaborator Paul Haney group from the Center for Nanoscale Science and Technology, National Institute of Standards and Technology (NIST) carried out first-principles DFT calculations using local density approximation (LDA) in the form of ultrasoft pseudopotentials, as implemented using Quantum ESPRESSO. An energy cutoff of 80 Ry and a 6 × 6 × 1 grid was employed for the plane wave basis expansion and for the Brillouin zone sampling during structural relaxation, respectively. All atoms in the unit cell were allowed to move until the force on each is less than 0.5 eV/nm. The lattice parameters are calculated to be a = 0.619 nm, b = 0.623 nm, c = 3.025 nm, and 99.67 ° for the angle between lattice vectors a and b, in good agreement with the experimental measurements (a = 0.613 nm, b = 0.619, c = 3.251 nm, and 93.80° (equation (5.7)). Similar to the threedimensional halide perovskites, the near-gap energy states in (C6H5C2H2NH3)2PbI4 are dominated by the orbitals in the two-dimensional inorganic framework and therefore, we 124 are concerned with the symmetry properties of the inorganic framework. As shown in Figure 5.10(a), inversion symmetry is present in the z-direction, normal to the twodimensional inorganic framework. The displacement of the Pb atom off the octahedral center leads to inversion symmetry breaking in the x-y plane. The symmetry breaking direction is roughly along the a + b direction. The electronic structure exhibits a direct bandgap at the R point in the BZ [2π/a, 2π/b, 0], where a and b are the lattice constants. The near-gap conduction band states are composed of the p orbitals of Pb while the valence bands states are derived from the Pbs orbital and Ip orbitals. The spin-orbit coupling splits degenerate conduction band states orbital quantum number (L = 1) into lower orbital angular momentum J = 1/2 and upper J = 3/2 bands, leading to a J =1/2 conduction band and S =1/2 valence band. In order to observe symmetry breaking effect on the band structure, bands along two paths are plotted as shown in Figure 5.10 (c). One is aligned with symmetry breaking direction from X=(0.0, 0.0, 0.0) to R=(0.5, 0.5, 0.0) and the other is along the normal direction from R=(0.5, 0.5, 0.0) to Y=(0.0, 1.0, 0.0). The band along X-R is nearly degenerate, whereas bands along R-Y exhibit Rashba-like splitting. We now consider the photogenerated free carrier absorption, i.e., the transition between Rashba-split conduction bands. Using DFT, we calculated the momentum matrix element between Rashba-split bands. We use a plane wave basis and the Bloch wave function is described as |Ψ Ñ ⟩ = ∑Ó Ò Ñ + Ó =( Ñ Ó ⋅Õ , where > is the crystal volume, G is the reciprocal lattice vector, k is the crystal momentum, and coefficient of plane wave = ( ⟨Ψ… |•̂• |Ψ ⟩ = ℏ ∑Ó ∗ … Ñ Ó ⋅Õ Ñ+Ó Ñ + Ó is the . The momentum matrix element was calculated as Ñ + Ó Ñ + Ó • , where refers to the momentum 125 direction •, ×, or Ù. In Figure 5.10(d), we show the dipole transition magnitude as a function of k along the direction with largest splitting that is also the direction normal to the inversion symmetry breaking. As seen, the optical transition between the two Rashbasplit CBs is allowed, in agreement with the tight-binding model described in section 5.3.2.3. Hence, we can say that Figure 5.11(a) shows the geometry of the relaxed structure. We find that the inversion symmetry is broken due to the Pb atom displacement from the octahedral center. The displacement is in the 2D plane, roughly in the direction of a + b, where a and b are the inplane lattice vectors. This leads to the Rashba band splitting for states with crystal momentum oriented perpendicularly to the symmetry breaking direction, as shown in Figure 5.11 (c). Figure 5.11 (b) is the CB energy dispersion near the R point in the Brillouin zone, where k1(2) is directed along the a +(−) b direction. The dashed red and blue lines in Figure 5.11 (d) show the optical transition matrix elements within the lowest CB, which, as discussed above, vanish at the minimum of the energy dispersion. The solid green line denotes the interband (IB) matrix element, which does not vanish at this point, indicating that the optical transition is allowed, which is seen in mid-IR photoinduced absorption (PA1) (Figure 5.10). An effective tight-binding model described in section 5.3.4 shows that the source of this optical transition is sp hybridization present in the CB eigenstates at the band minimum. From our model calculation, we obtain band splitting energy, ER = 160 meV, which is larger than the experimentally determined splitting. We attribute this discrepancy to approximations used in LDA. For more quantitatively accurate description of the electronic structure, a calculation at the level of quasi-particle GW approximation (Green's function G and the screened Coulomb 126 Figure 5.11. DFT calculations on the 2D perovskite. (a) One layer of the Pb-I octahedra that describes the relaxed structure of the PEPI used in the DFT calculations. The Pb atom (gray sphere) is displaced from the octahedra center along the a + b direction, which breaks off the inversion symmetry, resulting in Rashba-splitting caused by SOC. The unit cell vectors a and b lie in the x-y plane with an angle of 99.7° between them. (b) Schematic of the CB energy dispersion near the R point in the Brillouin zone, where k1(2) is directed along the a +(−) b direction. (c) Electronic band structure near the R point, which shows the Rashba-splitting along a direction perpendicular to the symmetry breaking direction; c1 and c2 represent the lower and upper Rashba bands, respectively. (d) DFT-calculated momentum matrix elements versus k near the band minimum (at k0 = 0.07 Å−1) away from the R point along the (1, −1) direction. Red and blue lines correspond to x and y component of the momentum matrix element between lowest CB c1 and itself, showing the vanishing transition between the exciton and lowest Rashba-split CB at k = k0. The green curve is the z component of the momentum matrix element between the Rashba-split bands c1 and c2, which is nonzero for all k. The y axis is dimensionless, with the computed momentum p presented in terms of its value in Rydberg units: p0 = 1.99 × 10−24 kg/(m·s).(Calculated by Paul Haney Group NIST) 127 interaction W) is likely required. 5.3.2.3 Inter-Rashba Optical Transitions in 2D Perovskites To illustrate the essential features of the electronic structure and optical response of the system, we present an effective tight-binding model for the conduction and valence bands of the 2D perovskite. As in the 3D perovskite CH3NH3PbI3 in equation (5.12), the conduction band is derived from Pb •-orbitals. Due to the high spin-orbit of the heavy Pb atom, the l = 1/2 split off band is well separated in energy from the l = 3/2 bands, and forms the basis for the conduction band states. In terms of orbitals •.,2,Ú and spin ↑, ↓, the |l± ⁄ ⟩ states are: Žl ⁄ Ü= Žl √H ⁄ |•. , ↓⟩ + Á|•. , ↓⟩ + |•Ú , ↑⟩ Ü= √H |•. , ↓⟩ − Á|•. , ↓⟩ − |•Ú , ↓⟩ The valence band is comprised of Pb and I orbitals with Ý = 1 symmetry: ŽÞ ŽÞ ⁄ ⁄ (5.11) Ü = |Ý, ↑⟩ Ü = |Ý, ↓⟩ (5.12) It suffices to consider a square 2D lattice. The ss, pp, and sp hopping terms are denoted , ¯¯ , and ß, respectively. Breaking inversion symmetry introduces additional spin- dependent hopping terms. For structural inversion symmetry breaking along the •- direction, spin-orbit coupling leads to a C2 -dependent effective magnetic field in the Ùdirection acting on the l = ± 1⁄2 states. We parameterize this Rashba spin-orbit coupling with a constant à. The symmetry breaking also leads to additional terms coupling Ý and • 128 states, which we parameterize with à′. The tight-binding Hamiltonian then takes the following form (where the basis ordering is: ŽÞ − C 0 0 š áß −ÁC + à ã =â 0 ⁄ Ü, ŽÞ ⁄ 0 ß −ÁC + à š ¯¯ C + àC2 + Z 0 − C ß ÁC + à š 0 where C± = C. ± ÁC2 , C = C. + C2 . Ü, Žl ⁄ Ü, Žl ⁄ Ü ): ß ÁC + à š 0 æ å 0 ¯¯ C − àC2 + Z ä 5.13 C is dimensionless, scaled by the lattice constant a, and we assume CE ≪ 1. The constant Z determines the bandgap. To make analytical progress, we consider a perturbation expansion in ß/Z . (Note that in equation (5.7) we rescaled the inversion asymmetry Ý• hopping parameter à š to ßà š so that we can do an expansion of the conduction-valence coupling in terms of the single parameter ß.) Assuming C. = 0, the (unnormalized) conduction band wave functions to lowest order in ß are: where = + ¯¯ . JW = Žl ⁄ Ü+ˆ è ( é êš ‰ ŽÞ • é ê é ƒ„ ⁄ Ü, JW = Žl ⁄ Ü+ˆ è ( é êš ‰ ŽÞ ê é ƒ„ ⁄ Ü • é 5.14 The lowest order contribution to the conduction and valence band energies enters as ß : ,W = ¯¯ C2 ,Ä = − ± àŽC2 Ž + Z C2 + è • êë • • é ±ê é + • ƒ„ è • êë • • é ±ê é • ƒ„ (5.15) (5.16) 129 Note that the valence band degeneracy is lifted through the hybridization with the conduction band. The minimum of the conduction C is found using equation (5.15), and given here to lowest order in ß: C = ê ìì + è • ê ìì ê• • íí ìì êë • • ìì ƒ„ íí ê • ìì ê • • ƒ • • ìì „ (5.17) We may now compute the dipole matrix element between conduction band states at C . We find that only the Ù-component is nonzero: îJW |Ù|JW ⟩ = 2 Ú ¯¯ ßs ¯ ˆ íí (ê ê• ( ìì êë • • • ƒ ìì ê ìì „ + íí ê• ( ìì êë • ƒ ‰ H ìì ê• • ìì „ (ê (5.18) where s Ú¯ = îÝ|Ù|•Ú ⟩. Although the expression above is cumbersome, the significant result is simply that it is nonzero. This can also be understood by inspecting the form of the wave functions given in equation (5.14). The Ý• hybridization present in JW and JW enable an optical transition between the two. We note that the incident light is mostly polarized in the •× plane; however, the light is incoherent and diffuse, so that a Ù-component of the polarization is also generically present. 5.4 Conclusions In conclusion, the room-temperature EA spectra of (C4H9NH3)2PbI4 crystal phase gives a bandgap of 2.52 eV with an exciton binding energy of 150 meV and a reduced mass of 0.169m0. Driven by the rearrangement of the C4H9NH3+ +, there is a decrease in lattice spacing and increase in exciton binding energy due to quantum confinement effects. The low temperature (C4H9NH3)2PbI4 crystal phase possesses a bandgap of 2.57 eV with an exciton binding energy of 190 meV and reduced mass of 0.227m0. Finally, the EA signals 130 for the low-temperature (C4H9NH3)2PbI4 crystal phase show evidence of the quantum Stark and FK effects on exciton and bandgap absorption, respectively. Also, strong spin orbit coupling in the 2D perovskite PEPI causes Rashba-splitting in the continuum band, where the spin-degenerate parabolic band splits into two branches with opposite spin-aligned electronic states. This causes both the optical transitions of excitons into the continuum band, and free carrier absorption within the continuum band to acquire an ‘add-on’ energy term of 4ER. From the peak of FCA measured by steady state PM, we have determined the Rashba-splitting energy, ER=(40±5) meV in PEPI 2D hybrid perovskite, which is among the highest values reported so far. Our work provides a comprehensive, all optical method for studying the Rashba-splitting effect in semiconductors. CHAPTER 6 CONCLUSIONS 6.1 Conclusions In Chapter 3, we studied the vibrational and electronic properties of the πconjugated copolymers that are used in photovoltaic application. We found that the main charge carriers in these copolymers are polaron excitations. For these kinds of materials, the power conversion efficiency can reach 9%, but there are always degradation processes involved in these organic materials. Therefore, it is crucial to study the polaronic mass. For that purpose, we studied infrared vibration modes using Resonant Raman scattering, doping-induced absorption, and photoinduced absorption tools to study 6 different vibrational modes in the donor-acceptor copolymer PTB7. These six vibration modes are easily explained by amplitude mode model. The vibrational peaks for the Raman and IR absorption are not at the same energy but are shifted. We observed two polaron bands, P1 and P2, after doping with iodine at various duration from a few minutes to hours. This work is the first to report the amplitude mode model to explain the vibration properties of the πconjugated copolymers. Also, the polaronic mass is about 3.8me where me is electron effective mass. This value is smaller than the soliton effective mass, which is 6me, which indicates a larger carrier mobility in PTB7. In addition, the polaronic mass that we obtained 132 for doped PTB7 may explain the low hole mobility in its blend and the enhancement that we obtained for doped PTB7 may explain the low hole mobility in its blend with PCBM, which was measured to be 2x10-4 cm2/Vs and is a limiting factor in its application in OPV solar cells. We thus conclude that better growth of the PTB7 film may lead to a more planar structure of the copolymer chains, which, in turn, may also increase the OPV solar cell efficiency based on this copolymer. In Chapter 4, we discussed the TADF organic light emitting diodes. We studied the photoluminescence (PL) and electroluminescence (EL) from both films and OLED devices based on TADF materials. The magnetic field effect (MFE) shows that both PL and EL increase with MFE. PL and EL also increase with increasing temperature, which indicated the RISC process is thermally activated as well as susceptible to MFE. Doping with the fluorescent emitter caused the increase in stability of the OLED device. Also from the concentration dependence of the fluorescent emitter, we identified the energy transfer process from host TADF to the fluorescent emitter. FRET process is the energy transfer from the singlet exciplex state to the singlet state of the fluorescent emitter, which is a longrange process. DET is the short-range process, which is the energy transfer from singlet exciplex to the triplet of the fluorescent material. Hence, the process is a loss mechanism and we observed small value of MEL at high concentration of dopant. In Chapter 5, we applied different optical spectroscopies to study the lead halide hybrid 2D perovskite. Specifically, we applied temperature-dependent linear absorption spectroscopy, electroabsorption spectroscopy, and photomodulation spectroscopy. In the 2D layered perovskite, we observed large Rashba-splitting effect due to the inversion asymmetry present due to large spin orbit coupling. This SOC is induced by lead (Pb) atom 133 whose atomic number is large (82). Electroabsorption (EA) spectroscopy is used to observe the 1s exciton binding energy, which is 190meV. We observed two features from the EA spectrum, namely Stark shift of the exciton and Franz-Keldysh oscillation (FK oscillation). This FK oscillation in the EA spectrum implied that the material is crystalline with a periodic lattice. We observed the energy of the 1s and 2s excitons to be E1s=2.38eV and E2s=2.53eV and additionally, we obtained a precise measure of the bandgap energy of Eg=2.57eV, which is difficult to obtain from the linear absorption measurement. In addition, we performed CW-photomodulation spectroscopy to observe the free carrier absorption (FCA) spectrum for 2D halide perovskite. We saw a band at 0.15eV and analyzed it as a transition from the lower conduction band to the split-off upper conduction band. From DFT calculations, this transition is allowed and the relation with Rashba energy is EFCA/4. The Rashba-splitting energy thus obtained is ER~0.15/4=37±3meV, where the uncertainty comes from the spectral resolution of the spectrometer. Another source of uncertainty in this measurement is the unknown value of the quasi-Fermi level at the CB bottom, which may well be of order of 2 to 3meV at low temperature. We therefore conclude that the FCA results agree with ER = (40 ± 5) meV determined from the transient picosecond spectroscopy. 6.2 Future Works Detailed study of vibrational modes of donor-acceptor copolymers is also crucial to understand the symmetry breaking in the polymer chains. We have chosen three copolymers, namely PffBT4T, PTB7, and PBnDT-FTAZ, as well as homopolymers, namely P3HT and MEHPPV, for the comparison of the vibration modes and optical 134 transitions due to iodine doping. We expect to observe different polaron bands for copolymers and homopolymers, which would give us an idea about symmetry breaking in the copolymers due to the two nonequivalent moieties in the chain. Recently, hybrid perovskites have attracted interest because of their usefulness for application in OPV, LED, lasing, and spintronics. Here it is very important to study free carrier absorption properties at far infrared (FIR) range. For this, we need a polyethylene beamsplitter and DTGS detector, since polyethylene has transmission from 700cm-1 to 50cm-1. Due to water absorption effect in the spectrum in the far infrared region, we need to build a vacuum chamber to minimize this effect covering both spectrometer and detector. A proper cryostat (polyethylene window) to work at low temperature should be built, which is under construction. REFERENCES [1] C. J. Brabec, N. S. Sariciftci, and J. C. Hummelen, Adv. Funct. Mater. 11, 15 (2001). [2] D. Braun and A. J. Heeger, Appl. Phys. Lett. 58, 1982 (1991). [3] J. Burroughes, D. Bradley, A. Brown, R. Marks, K. Mackay, R. Friend, P. Burns, and A. Holmes, Lett. to Nat. 347, 539 (1990). [4] J. Peet, J. Y. Kim, N. E. Coates, W. L. Ma, D. Moses, A. J. Heeger, and G. C. Bazan, Nat. Mater. 6, 497 (2007). [5] N. S. Sariciftci, L. Smilowitz, A. J. Heeger, and F. Wudl, Science (80-. ). 258, 1474 (1992). [6] Z. H. Xiong, D. Wu, Z. V. Vardeny, and J. Shi, Nature 427, 821 (2004). [7] V. Dediu, M. Murgia, F. C. Matacotta, C. Taliani, and S. Barbanera, Solid State Commun. 122, 181 (2002). [8] A. J. Heeger, S. Kivelson, J. R. Schrieffer, and W.-P. Su, Rev. Mod. Phys. 60, 781 (1988). [9] W. P. Su, J. R. Schrieffer, and A. J. Heeger, Phys. Rev. B 22, 2099 (1980). [10] S. Mazumdar, Phys. Rev. Lett. 51, 292 (1983). [11] J. I. Pankove, Optical Processes in Semiconductors (Dover, New York, 1971). [12] C. KITTEL, Introduction to Solid State Physics, 8th ed. (Wiley, New Jersey, 2004). [13] C. F. Klingshirn, Semiconductor Optics (Springer, Berlin, 2006). [14] D. Campbell and A. Bishop, Nucl. Phys. B 200, 297 (1982). [15] C. W. Tang and S. A. Vanslyke, Appl. Phys. Lett. 51, 913 (19 [16] R. H. Friend, R. W. Gymer, A. B. Holmes, J. H. Burroughes, R. N. Marks, C. 136 Taliani, D. D. C. Bradley, D. A. Dos Santos, J. L. Bredas, M. Logdlund, and W. R. Salaneck, Nature 397, 121 (1999). [17] Y. Sun, N. C. Giebink, H. Kanno, B. Ma, M. E. Thompson, and S. R. Forrest, Nature 440, 908 (2006). [18] T. Drori, E. Gershman, C. X. Sheng, Y. Eichen, Z. V. Vardeny, and E. Ehrenfreund, Phys. Rev. B 76, 033203 (2007). [19] M. A. Baldo, D. F. O’Brien, M. E. Thompson, and S. R. Forrest, Phys. Rev. B 60, 14422 (1999). [20] H. Uoyama, K. Goushi, K. Shizu, H. Nomura, and C. Adachi, Nature 492, 234 (2012). [21] R. J. Holmes, B. W. D’Andrade, S. R. Forrest, X. Ren, J. Li, and M. E. Thompson, Appl. Phys. Lett. 83, 3818 (2003). [22] B. C. Krummacher, V.-E. Choong, M. K. Mathai, S. A. Choulis, F. So, F. Jermann, T. Fiedler, and M. Zachau, Appl. Phys. Lett. 88, 113506 (2006). [23] Y. B. Yuan, S. Li, Z. Wang, H. T. Xu, and X. Zhou, Opt. Express 17, 1577 (2009). [24] E. L. Williams, K. Haavisto, J. Li, and G. E. Jabbour, Adv. Mater. 19, 197 (2007). [25] C. Murawski, K. Leo, and M. C. Gather, Adv. Mater. 25, 6801 (2013). [26] Y. Tao, K. Yuan, T. Chen, P. Xu, H. Li, R. Chen, C. Zheng, L. Zhang, and W. Huang, Adv. Mater. 26, 7931 (2014). [27] H. Nakanotani, T. Higuchi, T. Furukawa, K. Masui, K. Morimoto, M. Numata, H. Tanaka, Y. Sagara, T. Yasuda, and C. Adachi, Nat. Commun. 5, 4016 (2014). [28] S. Hirata, Y. Sakai, K. Masui, H. Tanaka, S. Y. Lee, H. Nomura, N. Nakamura, M. Yasumatsu, H. Nakanotani, Q. Zhang, K. Shizu, H. Miyazaki, and C. Adachi, Nat. Mater. 14, 330 (2015). [29] M. Klessiger, Angew. Chem. Int. Ed. 34, 549 (1995). [30] G. Li, C. H. Kim, Z. Zhou, J. Shinar, K. Okumoto, and Y. Shirota, Appl. Phys. Lett. 88, 253505 (2006). [31] D. Kolosov, V. Adamovich, P. Djurovich, M. E. Thompson, and C. Adachi, J. Am. Chem. Soc. 124, 9945 (2002). [32] Y. Wang, K. Sahin-Tiras, N. J. Harmon, M. Wohlgenannt, and M. E. Flatte, Phys. 137 Rev. X 6, 011011 (2016). [33] T. P. I. Saragi, T. Spehr, A. Siebert, T. Fuhrmann-Lieker, and J. Salbeck, Chem. Rev. 107, 1011 (2007). [34] F. B. Dias, K. N. Bourdakos, V. Jankus, K. C. Moss, K. T. Kamtekar, V. Bhalla, J. Santos, M. R. Bryce, and A. P. Monkman, Adv. Mater. 25, 3707 (2013). [35] D. Zhang, L. Duan, C. Li, Y. Li, H. Li, D. Zhang, and Y. Qiu, Adv. Mater. 26, 5050 (2014). [36] T. Furukawa, H. Nakanotani, M. Inoue, and C. Adachi, Sci. Rep. 5, 8429 (2015). [37] J. Wang, A. Chepelianskii, F. Gao, and N. C. Greenham, Nat. Commun. 3, 1191 (2012). [38] A. H. Devir-Wolfman, B. Khachatryan, B. R. Gautam, L. Tzabary, A. Keren, N. Tessler, Z. V. Vardeny, and E. Ehrenfreund, Nat. Commun. 5, 4529 (2014). [39] F. J. Wang, H. Bassler, and Z. V. Vardeny, Phys. Rev. Lett. 101, 236805 (2008). [40] C. Zhang, D. Sun, C.-X. Sheng, Y. X. Zhai, K. Mielczarek, A. Zakhidov, and Z. V. Vardeny, Nat. Phys. 11, 427 (2015). [41] P. Desai, P. Shakya, T. Kreouzis, W. P. Gillin, N. A. Morley, and M. R. J. Gibbs, Phys. Rev. B - Condens. Matter Mater. Phys. 75, 094423 (2007). [42] S. Zhang, T. Kreouzis, and W. Gillin, Synth. Met. 173, 46 (2013). [43] J. Rybicki, T. D. Nguyen, Y. Sheng, and M. Wohlgenannt, Synth. Met. 160, 280 (2010). [44] Z.-K. Tan, R. S. Moghaddam, M. L. Lai, P. Docampo, R. Higler, F. Deschler, M. Price, A. Sadhanala, L. M. Pazos, D. Credgington, F. Hanusch, T. Bein, H. J. Snaith, and R. H. Friend, Nat. Nanotechnol. 9, 687 (2014). [45] Y. Mei, C. Zhang, Z. V. Vardeny, and O. D. Jurchescu, MRS Commun. 5, 297 (2015). [46] T. C. Sum and N. Mathews, Energy Environ. Sci. 7, 2518 (2014). [47] D. M. Jang, K. Park, D. H. Kim, J. Park, F. Shojaei, H. S. Kang, J. P. Ahn, J. W. Lee, and J. K. Song, Nano Lett. 15, 5191 (2015). [48] L. Etgar, Energy Environ. Sci. 11, 234 (2018). 138 [49] T. Ishihara, J. Takahashi, and T. Goto, Solid State Commun. 69, 933 (1989). [50] J.-C. Blancon, H. Tsai, W. Nie, C. C. Stoumpos, L. Pedesseau, C. Katan, M. Kepenekian, C. M. M. Soe, K. Appavoo, M. Y. Sfeir, S. Tretiak, P. M. Ajayan, M. G. Kanatzidis, J. Even, J. J. Crochet, and A. D. Mohite, Science (80-. ). 355, 1288 (2017). [51] H. Hu, T. Salim, B. Chen, and Y. M. Lam, Sci. Rep. 6, 1 (2016). [52] F. De Angelis, Acc. Chem. Res. 47, 3349 (2014). [53] C. C. Stoumpos, D. H. Cao, D. J. Clark, J. Young, J. M. Rondinelli, J. I. Jang, J. T. Hupp, and M. G. Kanatzidis, Chem. Mater. 28, 2852 (2016). [54] A. M. Ganose, C. N. Savory, and D. O. Scanlon, Chem. Commun. 53, 20 (2017). [55] C. Sheng, Y. Zhai, U. Huynh, C. Zhang, and Z. V. Vardeny, Synth. Met. 216, 31 (2016). [56] C. C. Stoumpos, C. D. Malliakas, and M. G. Kanatzidis, Inorg. Chem. 52, 9019 (2013). [57] D. a Clugston and P. A. Basore, Prog. Photovoltaics Res. Appl. 5, 229 (1997). [58] D. K. Schroder, R. N. Thomas, and J. C. Swartz, IEEE Trans. Electron. Devices 25, 254 (1978). [59] C. X. Sheng, M. Tong, S. Singh, and Z. V. Vardeny, Phys. Rev. B - Condens. Matter Mater. Phys. 75, 085206 (2007). [60] I. B. Koutselas, L. Ducasse, and G. C. Papavassiliou, J. Phys. Condens. Matter 8, 1217 (1996). [61] S. Sun, T. Salim, N. Mathews, M. Duchamp, C. Boothroyd, G. Xing, T. C. Sum, and Y. M. Lam, Energy Environ. Sci. 7, 399 (2014). [62] M. Hirasawa, R. Ishihara, T. Goto, K. Uchida, and N. Miura, Phys. B 201, 427 (1994). [63] K. Tanaka, T. Takahashi, T. Ban, T. Kondo, K. Uchida, and N. Miura, Solid State Commun. 127, 619 (2003). [64] J. Even, L. Pedesseau, and C. Katan, J. Phys. Chem. C 118, 11566 (2014). [65] Y. Yamada, T. Nakamura, M. Endo, A. Wakamiya, and Y. Kanemitsu, IEEE J. Photovoltaics 5, 401 (2015). 139 [66] M. Shimizu, J. I. Fujisawa, and J. Ishi-Hayase, Phys. Rev. B - Condens. Matter Mater. Phys. 71, 205306 (2005). [67] X. Hong, T. Ishihara, and A. V. Nurmikko, Phys. Rev. B 45, 6961 (1992). [68] K. Tanaka, F. Sano, T. Takahashi, and K. E. Takashi Kondo, Ryoichi Ito, Solid State Commun. 122, 249 (2002). [69] K. Tanaka and T. Kondo, Sci. Technol. Adv. Mater. 4, 599 (2003). [70] R. A. Kaindl, M. A. Carnahan, D. Hӓgele, R. Lövenich, and D. S. Chemla, Nature 423, 734 (2003). [71] E. Frantzeskakis, S. Pons, H. Mirhosseini, J. Henk, C. R. Ast, and M. Grioni, Phys. Rev. Lett. 101, 1 (2008). [72] K. Ishizaka, M. S. Bahramy, H. Murakawa, M. Sakano, T. Shimojima, T. Sonobe, K. Koizumi, S. Shin, H. Miyahara, A. Kimura, K. Miyamoto, T. Okuda, H. Namatame, M. Taniguchi, R. Arita, N. Nagaosa, K. Kobayashi, Y. Murakami, R. Kumai, Y. Kaneko, Y. Onose, and Y. Tokura, Nat. Mater. 10, 521 (2011). [73] P. D. C. King, R. C. Hatch, M. Bianchi, R. Ovsyannikov, C. Lupulescu, G. Landolt, B. Slomski, J. H. Dil, D. Guan, J. L. Mi, E. D. L. Rienks, J. Fink, A. Lindblad, S. Svensson, S. Bao, G. Balakrishnan, B. B. Iversen, J. Osterwalder, W. Eberhardt, F. Baumberger, and P. Hofmann, Phys. Rev. Lett. 107, 1 (2011). [74] E. I. Rashba, Phys. Solid State 2, 1224 (1960). [75] D. Niesner, M. Wilhelm, I. Levchuk, A. Osvet, S. Shrestha, M. Batentschuk, C. Brabec, and T. Fauster, Phys. Rev. Lett. 117, 1 (2016). [76] T. Wang, B. Daiber, J. M. Frost, S. A. Mann, E. C. Garnett, A. Walsh, and B. Ehrler, Energy Environ. Sci. 10, 509 (2017). [77] E. M. Hutter, M. C. Gélvez-Rueda, A. Osherov, V. Bulović, F. C. Grozema, S. D. Stranks, and T. J. Savenije, Nat. Mater. 16, 115 (2017). [78] J. S. Lee, G. A. H. Schober, M. S. Bahramy, H. Murakawa, Y. Onose, R. Arita, N. Nagaosa, and Y. Tokura, Phys. Rev. Lett. 107, 1 (2011). [79] Y. Zhai, S. Baniya, C. Zhang, J. Li, P. Haney, C.-X. Sheng, E. Ehrenfreund, and Z. V. Vardeny, Sci. Adv. 3, e1700704 (2017). [80] G. Lanzani, The Photophysics behind Photovolatic and Photonics (Wiley-VCH Verlag GmbH and Co KGaA, 2012). 140 [81] M. Liess, S. Jeglinski, Z. Vardeny, M. Ozaki, K. Yoshino, Y. Ding, and T. Barton, Phys. Rev. B 56, 15712 (1997). [82] G. Yu, J. Gao, J. C. Hummelen, F. Wudl, and A. J. Heeger, Science (80-. ). 270, 1789 (1995). [83] C. J. Brabec, Sol. Energy Mater. Sol. Cells 83, 273 (2004). [84] J. Peet, J. Y. Kim, N. E. Coates, W. L. Ma, D. Moses, A. J. Heeger, and G. C. Bazan, Nat. Mater. 6, 497 (2007). [85] L. M. Peter, Phil. Trans. R. Soc. A 369, 1840 (2011). [86] F. He and L. Yu, J. Phys. Chem. Lett. 2, 3102 (2011). [87] B. W. Ma, C. Yang, X. Gong, K. Lee, and A. J. Heeger, Adv. Funct. Mater. 15, 1617 (2005). [88] B. N. Blouin, A. Michaud, and M. Leclerc, Adv. Mater. 19, 2295 (2007). [89] H. Chen, J. Hou, S. Zhang, Y. Liang, G. Yang, and Y. Yang, Nat. Photonics 3, 649 (2009). [90] Y. Zhang, S. K. Hau, H. Yip, Y. Sun, O. Acton, and A. K. Jen, Chem. Mater. 22, 2696 (2010). [91] W. Chen, T. Xu, F. He, W. Wang, C. Wang, J. Strzalka, Y. Liu, J. Wen, D. J. Miller, J. Chen, O. K. Hong, O. L. Yu, and S. B. Darling, Nano Lett. 11, 3707 (2011). [92] H. Zhou, Y. Zhang, J. Seifter, S. D. Collins, C. Luo, G. C. Bazan, T. Nguyen, and A. J. Heeger, Adv. Mater. 25, 1646 (2013). [93] J. You, L. Dou, K. Yoshimura, T. Kato, K. Ohya, T. Moriarty, K. Emery, C. Chen, J. Gao, G. Li, and Y. Yang, Nat. Commun. 4, 1446 (2013). [94] L. Dou, J. You, Z. Hong, Z. Xu, G. Li, R. A. Street, and Y. Yang, Adv. Mater. 25, 6642 (2013). [95] R. S. Kularatne, H. D. Magurudeniya, P. Sista, M. C. Biewer, and M. C. Stefan, J. Polym. Sci. Part A Polym. Chem. 51, 743 (2013). [96] S. A. Hawks, F. Deledalle, J. Yao, D. G. Rebois, G. Li, J. Nelson, Y. Yang, T. Kirchartz, and J. R. Durrant, Adv. Energy Mater. 3, 1201 (2013). [97] Y. Liu, J. Zhao, Z. Li, C. Mu, W. Ma, H. Hu, K. Jiang, and H. Lin, Nat. Commun. 141 5, 5293 (2014). [98] Z. He, C. Zhong, S. Su, M. Xu, H. Wu, and Y. Cao, Nat. Photonics 6, 591 (2012). [99] D. Di Nuzzo, C. Fontanesi, R. Jones, S. Allard, I. Dumsch, U. Scherf, E. Von Hauff, S. Schumacher, and E. Da Como, Nat. Commun. 6, 6460 (2015). [100] M. D. M. and J.-L. B. Chad Risko, Chem. Sci. 2, 1200 (2011). [101] N. Blouin, A. Michaud, D. Gendron, S. Wakim, E. Blair, R. Neagu-plesu, M. Bellete, G. Durocher, Y. Tao, M. Leclerc, and D. Chimie, J. Am. Chem. Soc 130, 732 (2008). [102] B. P. Karsten, J. C. Bijleveld, and L. Viani, J. Mater. Chem. 19, 5343 (2009). [103] T. M. Pappenfus, J. A. Schmidt, R. E. Koehn, and J. D. Alia, Macromolecules 44, 2354 (2011). [104] N. M. O. Boyle, C. M. Campbell, and R. Hutchison, J. Phys. Chem. C 115, 16200 (2011). [105] N. Banerji, E. Gagnon, P. Morgantini, S. Valouch, A. R. Mohebbi, J. Seo, M. Leclerc, and A. J. Heeger, J. Phys. Chem. C 116, 11456 (2012). [106] C. Wiebeler, R. Tautz, J. Feldmann, E. Von Hau, E. Da Como, S. Schumacher, and W. Strasse, J. Phys. Chem. B 117, 4454 (2013). [107] R. Tautz, E. Da Como, T. Limmer, J. Feldmann, H. Egelhaaf, E. Von Hauff, V. Lemaur, D. Beljonne, S. Yilmaz, I. Dumsch, S. Allard, and U. Scherf, Nat. Commun. 3, 970 (2012). [108] K. Aryanpour, T. Dutta, U. N. V Huynh, Z. V. Vardeny, and S. Mazumdar, Phys. Rev. Lett. 115, 267401 (2015). [109] R. Österbacka, X. M. Jiang, C. P. An, B. Horovitz, and Z. V Vardeny, Phys. Rev. Lett. 88, 226401 (2002). [110] E. Ehrenfreund, Z. Vardeny, and O. Brafman, Phys. Rev. B 36, 1535 (1987). [111] B. Horovitz, Solid State Commun. 41, 729 (1982). [112] Z. Vardeny and Joseph Orenstein and G. L. Baker, Phys. Rev. Lett. 50, 2032 (1983). [113] Z. Vardeny, E. Ehrenfreund, O. Brafman, and B. Horovitz, Phys. Rev. Lett. 51, 2326 (1983). 142 [114] R. Österbacka, C. P. An, X. M. Jiang, and Z. V. Vardeny, Science (80-. ). 287, 839 (2000). [115] Y. Liang, Z. Xu, J. Xia, S. Tsai, Y. Wu, G. Li, C. Ray, and L. Yu, Adv. Energy Mater. 22, E135 (2010). [116] Y. W. Soon, H. Cho, J. Low, H. Bronstein, I. Mcculloch, and J. R. Durrant, Chem. Commun. 49, 1291 (2013). [117] J. Razzell-hollis, J. Wade, W. C. Tsoi, Y. Soon, J. Durrant, and J. Kim, J. Mater. Chem. A Mater. Energy Sustain. 2, 20189 (2014). [118] T. Drori, J. Holt, and Z. V Vardeny, Phys. Rev. B 82, 075207 (2010). [119] G. Grancini, M. Maiuri, D. Fazzi, A. Petrozza, H.-J. Egelhaaf, D. Brida, G. and Cerullo, and G. Lanzani, Nat. Mater. 12, 29 (2013). [120] D. Fazzi, G. Grancini, M. Maiuri, D. Brida, G. Cerullo, and G. Lanzani, Phys. Chem. C Hem. P Hys 14, 6367 (2012). [121] T. Basel, U. Huynh, T. Zheng, T. Xu, L. Yu, and Z. V. Vardeny, Adv. Funct. Mater. 25, 1895 (2015). [122] B. R. Gautam, T. D. Nguyen, E. Ehrenfreund, and Z. V. Vardeny, Phys. Rev. B 85, 205207 (2012). [123] B. Horovitz, R. Österbacka, and Z. V Vardeny, Synth. Met. 141, 179 (2004). [124] B. Ebenhoch, S. A. J. Thomson, G. Kristijonas, G. Juška, and I. D. W. Samuel, Org. Electron. 22, 62 (2015). [125] C. J. Zheng, J. Wang, J. Ye, M. F. Lo, X. K. Liu, M. K. Fung, X. H. Zhang, and C. S. Lee, Adv. Mater. 25, 2205 (2013). [126] S. R. Forrest, M. A. Baldo, D. F. O’Brien, Y. You, A. A. Shoustikov, S. Sibley, and M. E. Thompson, Nature 395, 151 (1998). [127] M. Baldo, S. Lamansky, P. E. Burrows, M. E. Thompson, and S. R. Forrest, Appl. Phys. Lett. 75, 4 (1999). [128] C. Adachi, M. A. Baldo, M. E. Thompson, and S. R. Forrest, J. Appl. Phys. 90, 5048 (2001). [129] T. Tsuzuki, Y. Nakayama, J. Nakamura, T. Iwata, and S. Tokito, Appl. Phys. Lett. 88, 243511 (2006). 143 [130] J. Kalinowski, V. Fattori, M. Cocchi, and J. A. G. Williams, Coord. Chem. Rev 255, 2401 (2011). [131] D. Y. Kondakov, T. D. Pawlik, T. K. Hatwar, and J. P. Spindler, J. Appl. Phys. 106, 124510 (2009). [132] S. M. King, M. Cass, M. Pintani, C. Coward, F. B. Dias, A. P. Monkman, and M. Roberts, J. Appl. Phys. 109, 074502 (2011). [133] Y. Zhang, Y. Lei, Q. Zhang, and Z. Xiong, Org. Electron. 15, 577 (2014). [134] V. Jankus, P. Data, D. Graves, C. McGuinness, J. Santos, M. R. Bryce, F. B. Dias, and A. P. Monkman, Adv. Funct. Mater. 24, 6178 (2014). [135] S. A. Crooker, F. Liu, M. R. Kelley, N. J. D. Martinez, W. Nie, A. Mohite, I. H. Nayyar, S. Tretiak, D. L. Smith, and P. P. Ruden, Appl. Phys. Lett. 105, 153304 (2014). [136] H. Kaji, H. Suzuki, T. Fukushima, K. Shizu, K. Suzuki, S. Kubo, T. Komino, H. Oiwa, F. Suzuki, A. Wakamiya, Y. Murata, and C. Adachi, Nat. Commun. 6, 8476 (2015). [137] K. Goushi, K. Yoshida, K. Sato, and C. Adachi, Nat. Photonics 6, 253 (2012). [138] Q. Zhang, B. Li, S. Huang, H. Nomura, H. Tanaka, and C. Adachi, Nat. Photonics 8, 326 (2014). [139] Y. Seino, S. Inomata, H. Sasabe, Y. J. Pu, and J. Kido, Adv. Mater. 28, 2638 (2016). [140] M. R. A. E. Nikolaenko , M. Cass , F. Bourcet , D. Mohamad, Adv.Mater 27, 7236 (2015). [141] Y. H. Kim, C. Wolf, H. Cho, S. H. Jeong, and T. W. Lee, Adv. Mater. 28, 734 (2016). [142] G. Xie, X. Li, D. Chen, Z. Wang, X. Cai, D. Chen, Y. Li, K. Liu, Y. Cao, and S. J. Su, Adv. Mater. 28, 181 (2016). [143] X. K. Liu, Z. Chen, J. Qing, W. J. Zhang, B. Wu, H. L. Tam, F. Zhu, X. H. Zhang, and C. S. Lee, Adv. Mater. 27, 7079 (2015). [144] X. K. Liu, Z. Chen, C. J. Zheng, C. L. Liu, C. S. Lee, F. Li, X. M. Ou, and X. H. Zhang, Adv. Mater. 27, 2378 (2015). [145] P. B. Deotare, W. Chang, E. Hontz, D. N. Congreve, L. Shi, P. D. Reusswig, B. 144 Modtland, M. E. Bahlke, C. K. Lee, A. P. Willard, V. Bulović, T. Van Voorhis, and M. A. Baldo, Nat. Mater. 14, 1130 (2015). [146] W. Song, I. Lee, and J. Y. Lee, Adv. Mater. 27, 4358 (2015). [147] T. Higuchi, H. Nakanotani, and C. Adachi, Adv. Mater. 27, 2019 (2015). [148] W. Song, I. H. Lee, S. H. Hwang, and J. Y. Lee, Org. Electron. 23, 138 (2015). [149] X. K. Liu, Z. Chen, C. J. Zheng, M. Chen, W. Liu, X. H. Zhang, and C. S. Lee, Adv. Mater. 27, 2025 (2015). [150] D. R. Lee, B. S. Kim, C. W. Lee, Y. Im, K. S. Yook, S. H. Hwang, and J. Y. Lee, Appl. Mater. Interfaces 7, 9625 (2015). [151] T. Basel, D. Sun, S. Baniya, R. McLaughlin, H. Choi, O. Kwon, and Z. V. Vardeny, Adv. Electron. Mater. 2, 1500248 (2016). [152] Y. Ling, Y. Lei, Q. Zhang, L. Chen, Q. Song, and Z. Xiong, Appl. Phys. Lett. 107, 213301 (2015). [153] T. D. Nguyen, E. Ehrenfreund, and Z. V. Vardeny, Science (80-. ). 337, 204 (2012). [154] Y. Zhang, M. Slootsky, and S. R. Forrest, Appl. Phys. Lett. 99, 223303 (2011). [155] H. Katayama, S. Ito, and M. Yamamoto, J. Phys. Chem. 96, 10115 (1992). [156] K. Hisada, A. Tsuchida, S. Ito, and M. Yamamoto, J. Phys. Chem. B 102, 2640 (1998). [157] J. Nitta, T. Akazaki, H. Takayanagi, and T. Enoki, Phys. Rev. Lett. 78, 1335 (1997). [158] S. D. Stranks and H. J. Snaith, Nat. Nanotechnol. 10, 391 (2015). [159] N.-G. Park, M. Grätzel, T. Miyasaka, K. Zhu, and K. Emery, Nat. Energy 1, 16152 (2016). [160] Z. G. Yu, Sci. Rep. 6, 28576 (2016). [161] M. Kepenekian, R. Robles, C. Katan, D. Sapori, L. Pedesseau, and J. Even, ACS Nano 9, 11557 (2015). [162] Y. C. Hsiao, T. Wu, M. Li, and B. Hu, Adv. Mater. 27, 2899 (2015). 145 [163] H. Tsai, W. Nie, J.-C. Blancon, C. C. Stoumpos, R. Asadpour, B. Harutyunyan, A. J. Neukirch, R. Verduzco, J. J. Crochet, S. Tretiak, L. Pedesseau, J. Even, M. A. Alam, G. Gupta, J. Lou, P. M. Ajayan, M. J. Bedzyk, M. G. Kanatzidis, and A. D. Mohite, Nature 536, 312 (2016). [164] D. H. Cao, C. C. Stoumpos, O. K. Farha, J. T. Hupp, and M. G. Kanatzidis, J. Am. Chem. Soc. 137, 7843 (2015). [165] D. Giovanni, W. K. Chong, H. A. Dewi, K. Thirumal, I. Neogi, R. Ramesh, S. Mhaisalkar, N. Mathews, and T. C. Sum, Sci. Adv. 2, e1600477 (2016). [166] J. -i. Fujisawa and T. Ishihara, Phys. Rev. B 70, 205330 (2004). [167] L. S. and G. Weiser, Phys. Rev. Lett. 46, 1156 (1981). [168] D. B. Straus, S. Hurtado Parra, N. Iotov, J. Gebhardt, A. M. Rappe, J. E. Subotnik, J. M. Kikkawa, and C. R. Kagan, J. Am. Chem. Soc. 138, 13798 (2016). [169] D. F. Blossey, Phys. Rev. B 3, 1382 (1971). [170] G. U. Bublitz and S. G. Boxer, Annu. Rev. Phys. Chem. 48, 213 (1997). [171] V. Perebeinos and P. Avouris, Nano Lett. 7, 609 (2007). [172] H. Shen and M. Dutta, J. Appl. Phys. 78, 2151 (1995). [173] M Cardona, Modulation Spectroscopy of Semiconductors (Springer Berlin Heidelberg: Berlin, Heidelberg, 1970). [174] T. Ishihara, J. Takahashi, and T. Goto, Phys. Rev. B 42, 11099 (1990). [175] O. Yaffe, A. Chernikov, Z. M. Norman, Y. Zhong, A. Velauthapillai, A. Van Der Zande, J. S. Owen, and T. F. Heinz, Phys. Rev. B - Condens. Matter Mater. Phys. 92, 1 (2015). [176] J. Bordas and E. A. Davis, Solid State Commun. 12, 717 (1973). [177] Y. Yacoby, Phys. Rev 140, A263 (1965). [178] M. E. Ziffer, J. C. Mohammed, and D. S. Ginger, ACS Photonics 3, 1060 (2016). [179] D. E. Aspnes, Phys. Rev 972 (1967). [180] M. A. Fox, Optical Properties of Solids (2001). [181] G. Bastard, Wave Mechanics Applied to Semiconductor Heterostructures (1988).