CHARGES and EXCITONS in ORGANIC and PERVOSKITE LIGHT EMITTING MATERIALS and DEVICES by JUNWEI XU a Dissertation Submitted To

CHARGES AND EXCITONS IN ORGANIC AND PERVOSKITE LIGHT EMITTING

MATERIALS AND DEVICES

JUNWEI XU

A Dissertation Submitted to the Graduate Faculty of

WAKE FOREST UNIVERSITY GRADUATE SCHOOL OF ARTS AND SCIENCES

in Partial Fulfillment of the Requirements

for the Degree of

DOCTOR OF PHILOSOPHY

Physics

May, 2018

Winston-Salem, North Carolina

Approved By:

David L. Carroll, Ph.D., Advisor

Scott M.Geyer, Ph.D., Chair

Martin Gulthold, Ph.D.

K. Burak Üçer, Ph.D.

Richard T. Williams, Ph.D.

Dedication and Acknowledgements

Most of all, I would like to thank my advisor, Dr. David Carroll, a talented and passionate scientist who took me into Wake Forest University back in 2013. For that, I am very thankful for the fresh new opportunity he offered that I never thought I would have. During my thesis work, I thank him for appreciating my research strengths and patiently encouraging me to improve in my weaker areas. His strong support of my own ideas and research directions, and confidence in my abilities were benefits not all thesis students enjoy (but should). I am also very much indebted to Dr. Richard Williams, who is very patient and thoughtful to guide me to understand these exciting findings of perovskite materials. My time working on optical measurement in ultrafast laser lab was extremely positive and fun. I also would like to thank the members of my dissertation committee, Dr. Burak Üçer, Dr. Scott Geyer, and Dr. Martin Gulthold, not only for their time and patience, but for their intellectual contributions to my development as an independent researcher.

To my lab buddies: Yonghua Chen, Corey Hewitt, Jiwen Liu, Gregory Smith,

Yuan Li, Wenxiao Huang, Chaochao Dun, Hyunsu Lee, Yue Cui, Drew Onken, Yang

Guo, Gabriel Marcus, Linqi Shao, Stephen McCarthy, Tianxiao Shen and so many others

I can’t list here. Thanks for all the fun, great company, discussion and helps every day in lab and office. Without their strong opinions and sharp insights, my research project might not have ever come to fruition. I cannot adequately express how thankful I am.

Finally, but not least, I want to thank my parents. I thank my father for encouraging me to be an independent thinker, and having confidence and courage to go

II after new things that may sound crazy. I thank my mother for teaching me to face the complex world and try to leave the world just a little better than when came into it, and how my dream should a worthy part of that pursuit.

And my most heartfelt thanks to my wife, Peiyun.

Dedicated to the most intellectually exciting and memorable 5 years in my life.

2013-2018.

III

Table of Contents

List of Illustrations and Tables………………….………………………………………VII

Table of Abbreviations………………………………………………………………..XXII

Abstract…………………………………………………………………………...….XXIV

Chapter 1 Basics of Radiative Emission in Semiconductors ...... 1

Spontaneous Emission ...... 2

Charge Carrier Transport ...... 4

Exciton Generation and Recombination ...... 9

The Intersystem Crossing (ISC) Model in Singlet-Triplet Dynamics ...... 12

Exciton Oscillator strength ...... 13

Lead Halide Perovskite Emitter ...... 15

Chapter 2 Charge Balance in AC- Field Driven Organic Light Sources ...... 18

Introduction of Alternating Current Driven Organic Electroluminescence (AC-OEL)

Device ...... 18

Materials and Characterization ...... 21

Carrier Gate Structure in AC-OEL Device ...... 22

Gate Carrier Valve Mechanism ...... 28

No Gate vs Semiconductor Gate vs Insulator Gate ...... 33

Chapter Summary ...... 38

Chapter 3 Tailoring Spin Mixtures in Color-Tunable AC-OEL Devices ...... 39

Spin Remixing in Electroluminescence ...... 39

Materials and Methods ...... 42

Fluorescence-Phosphorescence (F-P) Hybrid Emission Layer ...... 45

Maxwell Magnetic Field Coupling ...... 47

Frequency Controlled Color Tunability ...... 51

Ion Enhanced Color Tunability ...... 60

Energy Transfer Mechanism of Magnetic Dependent ISC ...... 63

Chapter Summary ...... 68

Chapter 4 Solution Possessing Organic Small Molecular Emitter ...... 70

Solution Based AC-OEL Devices ...... 70

Solubility of Alq3 in Different Solvents ...... 72

Morpholgy of Spun-cast Alq3 Thin Film and Wash-out Effect ...... 75

Resonant Effect and Carrier Gate Mechanism in Alq3 Device ...... 77

Methods and Characterizations ...... 84

Chapter 5 Flexible, Stable, White Electroluminescent Devices Under High Frequency . 86

Challenges for Achieving Flexibility ...... 87

Construction of Metal/Nanowires Composite Electrode ...... 90

Bending Test on Nano-Network Composite Cathode ...... 97

Iterative Bending Tests on Devices with Al/MWCNTs/Al Electrode ...... 103

Flexible Carrier Gate in White AC-EL ...... 104

N-type Polymeric Gate Structure and Application ...... 106

Chapter Summary ...... 120

Chapter 6 Imbedded Nanocrystals of CsPbBr3 in Cs4PbBr6 Matrix in the Application of

LED ...... 121

Discovery of CsPbBr3-Cs4PbBr6 nanocompsite ...... 121

Methods and Measurement ...... 123

Solution growth and properties of Cs-Pb-Br perovskite thin films...... 126

CsPbBr3 and Cs4PbBr6 DMSO-washed powder samples...... 130

Morphology of Thin Films Comprised of CsPbBr3/Cs4PbBr6 Nano-inclusion Grains.

...... 132

Light Emitting Diodes (LEDs) using CsPbBr3 nanocrystals in Cs4PbBr6 host...... 134

Optical response and kinetics of pure CsPbBr3 and the CsPbBr3/Cs4PbBr6 nano-

inclusion composite under high-intensity photoexcitation...... 139

PL quantum efficiency and decay time dependence on size of the Cs4PbBr6 host micro-

grain...... 145

Enhanced Oscillator Strength of CsPbBr3 Imbedded Nanocrystals Relative to Extended

CsPbBr3...... 147

Chapter 7 Cs4PbBr6 Single Crystal with CsPbBr3 Nano-Inclusions ...... 154

Growth of Cs4PbBr6 Single Crystal imbedded with CsPbBr3 Nanoinclusions ...... 155

Compositional analysis of the nano-composite ...... 159

Exction Kinetics and Recombination Analysis ...... 161

CsPbBr3 Phase Evolution in Host Matrix ...... 168

Reference………………………………………………………………………...……..174

Curriculum Vitae……………………………………………………………………….206

List of Illustrations and Tables

Figure 1-1 | Different lighting technologies. (a) Incandescent bulb; (Image courtesy of thorinside via Flickr) (b) Gas discharge tube; (Image courtesy of Tatalux Lighting) (c)

Light emitting diode; (Image courtesy of Adafruit) (d) Organic light emitting device.

(Image courtesy of LG Display) ...... 1

Figure 1-2 | Scheme of electron hole recombination resulting in spontaneous light emission in a semiconductor...... 2

Figure 1-3 | P-N junction in zero bias (a) and forward bias (b)...... 3

Figure 1-4 | Carrier transfer in trap controlled transport mechanism...... 6

Figure 1-5 | Density of states with deep carrier traps...... 7

Figure 1-6 | Carrier hopping and tunneling mechanism in organic semicondcutors ...... 8

Figure 1-7 | Singlet excited state (S=0) and triplet excited states (Sz=+1,0,-1)...... 10

Figure 1-8 | Energy diagram illustrating fluorescence, phosphorescence, internal conversion (IC) and intersystem crossing (ISC)...... 11

Figure 1-9 | Electron and hole spin precessions in an external magnetic field B for singlet and triplet e–h pairs...... 12

Figure 1-10 | (a) Lattice structure of orthogonal Pervoskite structure.28 (b) Red and green

Pervoskite LEDs.29 ...... 15

Figure 1-11 | The color tunable fluorescence of CsPbX3 nanocrystals based on halid ion exchange.30 ...... 16

Figure 1-12 | Quantum size effects in low dimensional perovskite materials.35...... 17

VII

Figure 2-1 | The high-brightness AC-OEL device configuration consists of ZnO gate and organic emitter complex. The photo of AC-OEL device has a luminance over 10,000 cd/m2. SEM image of ZnO NPs shows the morphology of gate layer...... 22

Figure 2-2 | The thicknesses of ZnO gate layers spun coat by different concentrations of

ZnO NPs solution (1wt% and 5wt%)...... 23

Figure 2-3 | Organic molecule structures employed in this work ...... 24

Figure 2-4 | The energy level diagram of the AC-OEL devices ...... 25

Figure 2-5 | Performance of phosphorescent green AC-OEL devices with various thicknesses of ZnO gate. (a) The luminance as function of driving frequency (from 50Hz to 70kHz). (b) The RMS current density as function of driving frequency (from 50Hz to

70kHz)...... 26

Figure 2-6 | Current density versus voltage characteristic under DC driving in the devices of ITO (100nm)/ZnO (~100nm)/Poly-TPD:F4TCNQ (~70nm)/MoO3 (15nm)/Au

(100nm) and ITO (100nm)/ZnO (~100nm)/MoO3 (15nm)/Au (100nm)...... 28

Figure 2-7 | Energy band diagrams of AC-OEL devices in forward and reverse bias. .... 29

Figure 2-8 | Time response of cell current in ITO (100nm)/ZnO (~100nm)/Poly-

TPD:F4TCNQ (~70nm)/Au (100nm) and ITO (100nm)/ZnO (~100nm)/Au (100nm). ... 30

Figure 2-9 | I-V performance and theoretical models of the AC-OEL devices with various hole generation layers (HGL). (a) Experimental RMS current density as function of RMS voltage utilizing Poly-TPD:F4TCNQ, Poly TPD, F4TCNQ as HGL or absence, respectively. (b), (c), (d), and (e) are electronic band structures analysis with various

HGLs. (f) Theoretical RMS current density as function of RMS voltage by fitting in tunneling model and P-N junction model...... 31

VIII

Figure 2-10 | (a) RMS current density as function of RMS voltage. (b) Luminance as function of RMS voltage. (c) Luminance as function of frequency...... 34

Figure 2-11 | The carrier transportation mechanisms of AC-OEL devices with ZnO gate and P(VDF-TrFE-CFE) insulator, and AC-OLED...... 35

Figure 2-12 | (a) Power efficiency plot as function of luminance. Inset shows the phase angle characteristics. (b) Luminance versus RMS current density characteristics of the

AC-OEL devices with ZnO gate and P(VDF-TrFE-CFE) insulator, and AC-OLED while an increasing DC voltage offset are added to AC driving voltage. 7.1V and 4.2V of DC voltages are added to AC voltages when AC-OLED and ZnO device reach the maximum luminance respectively...... 37

Figure 3-1 | Compositional schematic of devices with fluorescence-phosphorescence (F-

P) emission interface driven by AC power source (50Hz ~ 70,000Hz). Singlet spins (blue circle pairs) and triplet spins (red circle pairs) are formed at the PVK/PFN-Br interface

(or F-P interface)...... 45

Figure 3-2 | AFM images of PVK:Ir(MDQ)2(acac) before annealing (a) and after 100 degrees annealing (b); Surface morphology of PFN-Br before annealing (c) and after 100 degress annealing (d)...... 46

Figure 3-3 | Chemical structures of PFN-Br and Ir(MDQ)2(acac). Photo luminescence of

PFN-Br and Ir(MDQ)2(acac) under 365nm UV lamp...... 46

Figure 3-4 | Lifetime of PFN-Br’s fluorescence and Ir(MDQ)2(acac)’s phosphorescence.

...... 47

Figure 3-5 | Schematic diagram of dynamic field analysis at p-n magnetic interface...... 48

Figure 3-6 | Simulated results of magnetic and electric fields at p-n junction at 60,000Hz.

The maximum self-generated magnetic field is 0.85mT. In the positive and negative half of AC driving cycle, the directions of vertex magnetic fields are clock wise (z > 0) and counter clock wise (z < 0) respectively...... 49

Figure 3-7 | Electroluminescence spectrum with PVK: 3wt% Firpic/PFN:DOF: 3wt%

Ir(MDQ)2(acac)...... 50

Figure 3-8 | Images of the blue emission and red emission at the frequency of 50Hz and

60,000 Hz respectively. Electrical and magnetic field properties of PVK/PFN-Br interface...... 51

Figure 3-9 | (a) At 50Hz, electroluminescence (EL) spectrum with RMS voltage varying from 15.4V to 20.4V corresponding the luminance of 15 cd/m2 to 210 cd/m2. Hot carrier injection is only current contribution. (b) At 60,000Hz, EL spectrum with RMS voltage varying from 17.9V to 22.0V corresponding the luminance of 40 cd/m2 to 240 cd/m2. .. 52

Figure 3-10 | (a) Time-resolved current at 50Hz. Hot carrier injection is only current contribution. (b) Time-resolved current at 60,000Hz with various voltages (17.9V to

22.0V). Secondary carriers (Jdc) and hot injection carriers (Jsine) have mutual contributions in EL spectrum and current density characteristics...... 53

Figure 3-11 | Jrms-Vrms and L-Vrms characteristics of AC-OEL devices at 50Hz and

60,000Hz, respectively...... 54

Figure 3-12 | Luminance plots as the function of frequency. Current-frequency characteristics of AC-OEL devices and its effective current analysis...... 55

Figure 3-13 | Current and voltage transient properties of AC-OEL device at frequency of

50Hz (a) and 60,000Hz (b)...... 55

Figure 3-14 | Electroluminescence (PL) spectral shift of AC-OEL devices with driving frequency of 50Hz, 100Hz, 500Hz, 1,000Hz, 10,000Hz, 20,000Hz, 30,000Hz, 40,000Hz,

50,000Hz, and 60,000Hz...... 57

Figure 3-15 | (a) Images of the color tuning emission in the forward and reverse frequency sweeps. (b) CIE 1931 x-y chromaticity diagram showing the color shift from

(0.23, 0.34) to (0.53, 0.40) with frequency variation between 50Hz and 60,000Hz...... 58

Figure 3-16 | EL spectra shift of AC-OEL devices with 30nm (a) and 80nm (b) PFN-Br spun-cast from Me(OH) solution...... 59

Figure 3-17 | Intensity ratio between 474nm peak and 590nm peak as a function of frequency in the PFN-Br’s concentrations of 20nm, 50nm, and 70nm...... 60

Figure 3-18 | Photoluminescence (PL) spectroscopy for organic emitting materials and energy transfer analysis. (a) Excitation and emission spectra for PVK film and PFN-Br film on glass. (b) PL spectra for glass/PVK: 3% Ir(MDQ)2(acac)/PFN-Br with two excitation: 347nm and 380nm. (c) Current-voltage characteristics of PFN-Br (ionized) and PFN-DOF (non-ionized). (d) EL spectral shifts of PFN-Br (ionized) and PFN-DOF

(non-ionized) at 50Hz (magnetic free) and 60,000Hz (magnetic engaged)...... 61

Figure 3-19 | PL spectroscopy of PVK: Ir(MDQ)2(acac)/PFN-Br and resonance energy transfer (FRET) efficiency...... 62

Figure 3-20 | Scheme for e-h pairs excited energy transfer mechanisms between PVK

(host), Ir(MDQ)2(acac) (phosphorescent dopant), and PFN-Br (fluorescent material) with magnetic-suppressive ISC. (e-h)1 and (e-h)3 represent singlet and triplet intermolecular e- h pairs. S and T are singlet and triplet excitons...... 63

Figure 3-21 | SEM image of Fe3O4 magnetic nanoparticles (NPs) in PFN-Br polymer matrix...... 64

Figure 3-22 | XRD pattern of Fe3O4 magnetic nanoparticles (NPs)...... 64

Figure 3-23 | (a) Time-integrated PL spectra measured at 437nm for pure PFN-Br thin film, PFN-Br doping 0.125wt% Fe3O4 NPs film, and PFN-Br doping 0.375wt% Fe3O4

NPs film after constant excitation at 280nm (10Hz, 500fs, 130 μJ/cm2). (b) Time-resolved

PL decay transients at 410nm~500nm integration...... 65

Figure 3-24 | (a) Time-integrated PL spectrum (>1us) for glass/ PVK:3wt%

Ir(MDQ)2(acac)/PFN-Br, glass/PVK:3wt% Ir(MDQ)2(acac)/PFN-Br:0.125wt%Fe3O4, and glass /PVK:3wt% Ir(MDQ)2(acac) /PFN-Br:0.375wt%Fe3O4. The streak images for glass/ PVK:3wt% Ir(MDQ)2(acac)/PFN-Br (b), glass/PVK:3wt% Ir(MDQ)2(acac)/PFN-

Br:0.125wt%Fe3O4 (c), and glass /PVK:3wt% Ir(MDQ)2(acac) /PFN-Br:0.375wt%Fe3O4

(d) in 6ns...... 67

Figure 3-25 | EL spectrum with no magnetic field, internal AC magnetic field, and external magnetic field due to Fe3O4 NPs...... 68

Figure 4-1 | (a) Chemical structure of Alq3. (b) The solubility of Alq3 (10 mg mL−1) in dimethylformamide (DMF), chloroform, and acetone...... 72

Figure 4-2 | (a) Current density characteristics in a function of RMS voltage at 50 Hz.

Inset: device structure of the AC-OEL devices. (b) The relation between RMS voltage and brightness at 50 Hz. Inset: the power efficiency...... 73

XII thermally (d) evaporated Alq3, (e) DMF spun-cast Alq3 film, and (f) chloroform spun- cast Alq3 film. The scale bar is 1 µm. (g–i) The corresponding 3D AFM images...... 75

Figure 4-4 | AFM image (5μm × 5 μm) of PEDOT:PSS...... 76

Figure 4-5 | XPS patterns of Poly-TPD/Alq3 from DMF(a), Poly-TPD/Alq3 from

Chloroform (b), and Poly-TPD/Alq3 deposited by evaporation (c)...... 77

Figure 4-6 | Frequency response of the AC driven OEL devices with and without a gate layer. (a) Current density. (b) Normalized luminous intensity. Note that the applied voltages were maintained at 22 and 8 V with and without the ZnO gate, respectively. ... 78

Figure 4-7 | The resonant behavior of ZnO-gated AC driven OEL devices over the frequency range 50Hz to 75,000Hz...... 79

Figure 4-8 | Hot carriers (50 Hz) versus polarization carriers (30 000 or 50 000 Hz) in electroluminescence (EL) characteristics. (a) AC driven OEL device with PEDOT:PSS injection layer. (b) 70 nm ZnO-gated AC-OEL device. Insets: current density in the function of AC voltage. Time-resolved voltage and current of ZnO-gated device at (c) 50

Hz and (d) 50 000 Hz...... 80

Figure 4-9 | Energy diagram of AC driven OEL devices (a) with hole injection layer

(PEDOT:PSS) and (b) with n-type gate layer (ZnO)...... 82

Figure 4-10 | EL spectrum of ZnO-gated AC-OEL device with Alq3 emitter doping 1wt%

C545T (a) and corresponding CIE diagram (b)...... 84

Figure 5-1 | AC-driven organic EL device on flexible 1 inch × 1 inch plastic substrate. (a)

PET/ITO(100nm)/PEDOT:PSS doped ZnO NPs (80nm)/PVK:Ir(ppy)3 (150nm)/TPBi

(35nm)/Triplet-layer cathode (~200nm). Electric fields due to AC driver in the capacitor-

XIII based device are changing in high frequency rather than pointing a unique direction. (b)

The definition of the bending angle used in this report...... 90

Figure 5-2| Multi-step fabrication of triple-layer composite electrode. (a) Thermal deposition of 50nm Al layer. (b) The formation of nano-matrix of MWCNTs with spraying. (c) Top 100 nm Al layer coating. SEM images of multi-step fabrication of triple-layer composite electrode. (d) Al, (e) Al/MWCNTs, (f) Al/MWCNTs/Al, corresponding to (a), (b) and (c)...... 92

Figure 5-3 | The electrical resistivity of Al electrode and Al/MWCNTs/Al before and after bending...... 96

Figure 5-4 | Performance of the green phosphorescent flexible organic AC-EL devices with and without MWCNTs driven at a frequency of 50kHz during bending (flat, 60°,

90°, 120°, recover). Luminance with Al (a) or Al/MWCNT/Al (b) as function of voltage for flexible organic AC-EL devices. Red arrows represent the luminance peak shifts between the maximum luminance of the Al-based device and that of the Al/MWCNT/Al- based device. (c) and (d) show RMS current density vs RMS voltage characteristics of flexible organic AC-OEL devices before and after bending...... 98

Figure 5-5 | Metallurgical microscopy images of top electrodes in flexible organic AC-EL devices before (insets) and after a full cycle of 120° bending. (a) Al (150nm) electrode shows 20μm width creases due to bending. (b) Al (50nm)/MWCNT/Al (100nm) composite electrode shows outstanding mechanical property in bending...... 99

Figure 5-6 | Power efficiencies with Al (a) or Al/MWCNT/Al (b) as function of luminance for flexible AC-OEL devices...... 101

XIV

Figure 5-7 | Electroluminescence spectrum of flexible organic AC-El device with

Al/MWCNTs/Al cathode at different bending angles...... 102

Figure 5-8 | Bending test of flexible organic AC-EL devices with and without MWCNT nano-network modification. (a) Luminance as a function of the number of cycles (0, 20,

25, 50, 100, 200, 500, and 1000). The insert shows the photograph of the flexible device with MWCNTs after 1000 bends. Luminance - current density - voltage characteristic of flexible organic AC-OEL devices before (b) and after (c) 1300 cycles of bend...... 104

Figure 5-9 | (a) Chemical structure of PFN-Br. (b) Morphology of spin-coated PFN-Br thin film...... 107

Figure 5-10 | Energy band diagrams of gate functionality in forward and reverse bias. 107

Figure 5-11 | I-V demonstration of PFN-Br polymeric gate in DC mode...... 108

Figure 5-12 | Current density characteristic of ITO/Poly-TPD:F4TCNQ/MoO3/ Au in forward and reversed bias...... 109

PFN-Br gate layer. (c) and (d) are the relation between RMS voltage versus brightness and RMS voltage versus RMS current density at 40kHz respectively...... 110

Figure 5-14 | Brightness relation with RMS AC voltage in the white AC-OEL devices with 1wt%, 2wt%, and 5wt% phosphorescent dopants [Ir(ppy)3:Ir(MDQ)2(acac) = 2:1].

...... 112

Figure 5-15 | (a) Degradation of AC-OEL device with different concentrations of phosphorescent: 5%, 2%, and 1wt% in weight. (b) Energy transfer mechanism of dual doping white emission system...... 112

Figure 5-16 | (a) and (b) Frequency response of brightness and RMS current density in three doping ratios of Ir(ppy)3 and Ir(MDQ)2(acac): 3:1, 2:1, and 1:1 at a constant voltage of 11.5V (±1V). (c) and (d) The brightness and RMS current density characteristics in the function of AC electric field intensity ...... 113

Figure 5-17 | The EL spectrum of white emission AC-OEL device with different doping ratios of Ir(ppy)3 and Ir(MDQ)2(acac): (a) 3:1, (b) 2:1, and (c) 1:1...... 114

Figure 5-18 | The photos of white emission AC-OEL device with different doping ratios of Ir(ppy)3 and Ir(MDQ)2(acac): (a) 3:1, (b) 2:1, and (c) 1:1...... 116

Figure 5-19 | (a) Device structure of PET-based AC-OEL device. (b) The photos of devices are bent in convex position and concave position. The scale bar is 10mm...... 116

Figure 5-20 | The luminescent and electrical stability in the function of bending angle from -90deg to +90deg. (a) PFN-Br gate; (b) ZnO gate...... 117

Figure 5-21 | (a) Device performance over frequency in 5 different bending positions: flat, 90deg, -90deg, 45deg, and -45deg. (b) Brightness and power efficiency variation with the change of bending angles. (c) The EL spectrum measurement with the variation of bending angles...... 118

Figure 5-22 | SEM images of PET-based flexible ITO substrates coated ZnO gate (a) and

PFNBr gate (b) after one full cycle bending...... 119

Figure 5-23| Iterative bending tests on the AC-OEL devices with different gate materials.

...... 120

XVI

Figure 6-1 | The fabrication process of Cs-Pb-Br perovskite thin films...... 127

Figure 6-2 | Cs-Pb-Br perovskite thin films on glass slides in room light (a) and uniform

365nm ultra-violet light (b) at room temperature. (c) AFM measurement of film thinkness. (d) XRD patterns of Cs-Pb-Br perovskite thin films from precursor solutions.

...... 128

Figure 6-3 | (a) Quantum yield (QY) of Cs-Pb-Br perovskite thin films from precursor solutions with different PbBr2-CsBr molar ratios (1:1, 1:1.2, 1:1.5, and 1:2). (b)

Photoluminescence (PL) shift of different Cs-Pb-Br perovskite thin films...... 129

Figure 6-4 | The XPS analysis of thin films...... 130

Figure 6-5 | Photos of pure CsPbBr3 (Cs1) and Cs4PbBr6 crystals with imbedded

CsPbBr3 nano-inclusions (Cs4) in room light (a) and UV light (b)...... 131

Figure 6-6 | Morphology of thin films comprised of CsPbBr3/Cs4PbBr6 nano-inclusion grains. AFM (a) and TEM (b) images of 1:1 thin film. Scale bar is 200nm. (c) Select- area electron diffraction pattern of CsPbBr3. AFM (d) and TEM (e) images of 1:2 thin film. Scale bar is 100nm. (f) Select-area electron diffraction pattern of Cs4PbBr6 crystal with CsPbBr3 nanocrystals...... 132

Figure 6-7 | SEM images of pervoskite thin films in a variety of PbB2-CsBr molar ratios:

(a) 1:1, (b) 1:1.2, (c) 1:1.5, (d) 1:2, (e) 1:3...... 133

Figure 6-8 | The device structure of the Cs-Pb-Br pervoskite LEDs...... 134

Figure 6-9 | (a) Luminance and current density of devices with emission layer fabricated from 1:1, 1:1.2, 1:1.5, and 1:2 PbBr2-CsBr precursors. Current efficiency (b) and external quantum efficiency (c) of the LEDs. (d) Luminance measurement via on-off method in a slow voltage sweep on nano-inclusion and bulk CsPbBr3 LEDs...... 135

XVII

Figure 6-10| TEM of CsPbBr3 inclusions imbedded Cs4PbBr6. Scale bar is 100nm...... 136

Figure 6-11 | 0.15 cm2 small area (a) and 2 cm2 large area pervoskite LEDs (b). (c)

Device spectrum in a ramping voltage...... 137

Figure 6-12 | (a) Schematic of operational LED device with Cs4PbBr6 crystal imbedded

CsPbBr3 nano-inclusions. (b) Radiative energy transition with CsPbBr3 imbedded

Cs4PbBr6 in a conjectured Type I heterostructure...... 138

Figure 6-13 | (a) EL spectra of LEDs with varying molar ratio of PbBr2 and CsBr. (b)

Summary of spectral blue shifts in PL and EL...... 139

Figure 6-14 | Optical images and SEM images of pure CsPbBr3 (a)(c) and purified

Cs4PbBr6 imbedded with CsPbBr3 nano-inclusions (b)(d). Scale bars are 2μm...... 140

Figure 6-15 | Subnanosecond luminescence under 300 fs, 60 μJ/cm2, 420 nm excitation.

Streak camera images of CsPbBr3 micro-crystals (a) and CsPbBr3 nano-crystals in

Cs4PbBr6 perovskite (b) at room temperature. (c) PL lifetime decay curves of Cs-Pb-Br thin films with different molar ratios, CsPbBr3 micro-crystals and CsPbBr3 nano- inclusions in Cs4PbBr6 crystals. (d) Fitting analysis on exciton dynamics of CsPbBr3 micro-crystals and CsPbBr3 nano-inclusions in Cs4PbBr6 crystals...... 142

Figure 6-16 | Dependence of nano-inclusion photoluminescence on size of the Cs4PbBr6 host micro-grains. TEM images of Cs4PbBr6 crystals in different sizes: 40nm (a), 100nm

(b), 2000nm (c), and 5000nm (d). (e) The image of Cs4PbBr6 crystal suspensions in

DMSO under 365nm UV lamp at room temperature. (f) Quantum yield in a function of

Cs4PbBr6 crystal sizes, and the PL lifetime versus Cs4PbBr6 crystal sizes. Streak camera images of Cs4PbBr6 crystal suspensions with a size of 40nm (g), 100nm (h), 2000nm (i), and 5000nm (j) at room temperature...... 146

XVIII

Figure 6-17 | CsPbBr3 inclusions in Cs4PbBr6 perovskite crystals. (a) Absorption coefficient of Cs-Pb-Br perovskite thin films with different precursor ratios. (b) TEM images of Cs4PbBr6 chip crystal with CsPbBr3 nano-crystals. Scale bar is 100nm. .... 148

Figure 7-1 | Steps for CsPbBr3/ Cs4PbBr6 crystal growth...... 155

Figure 7-2 | Morphology of the CsPbBr3/ Cs4PbBr6 crystals...... 156

Figure 7-3 | Fluorescent images of CsPbBr3/ Cs4PbBr6 crystals...... 157

Figure 7-4 | Molar ratio influence on quantum yield of CsPbBr3/ Cs4PbBr6 crystals .... 157

Figure 7-5 | Confocal microscopy images of CsPbBr3/ Cs4PbBr6 crystal. Scale bar is 5

μm...... 158

Figure 7-6 | (a) Powder X ray diffraction patterns of the ground Cs4PbBr6 host crystals imbedded with ~5% CsPbBr3 nanocrystals. (b) Crystal structures of Cs4PbB6 and

CsPbBr3. (c) HRTEM images of nanocomposite...... 159

Figure 7-7 | EDS images of Cs-Pb-Br nanocomposite...... 160

Figure 7-8 | (a) PL spectrum of emission of CsPbBr3 nanocrystals. (b) Absorbance and photoluminescence of Cs4PbBr6 crystal with CsPbBr3 nanoinclusions. (c) Excitation spectrum of the crystal. The inset is PL spectrum of emission of CsPbBr3 nanocrystals

...... 161

Figure 7-9 | Streak images of CsPbBr3 nanocrystals in Cs4PbBr6 crystal under 420nm laser excitation in a 300fs pulse in a different of fluences: 26.8 μJ/cm2 and 0.768 μJ/cm2.

...... 162

Figure 7-10 | Fitting analysis on exciton dynamics of CsPbBr3 nanocrystals in Cs4PbBr6 crystal in a function of laser fluence...... 163

Figure 7-11 | Power dependent analysis...... 164

XIX

Figure 7-12 | CsPbBr3/Cs4PbBr6 crystals sealed in PDMS...... 166

Figure 7-13 | X-ray luminesence and decay of the X-ray luminescence...... 166

Figure 7-14 | Snapshots of CsPbBr3/Cs4PbBr6 nanocomposite sample under UV lamp within 1 hour...... 168

Figure 7-15 | Time dependent PL spectrum of CsPbBr3/Cs4PbBr6 nanocomposite...... 169

Figure 7-16 | Scheme of CsPbBr3 nanocrystal growth in Cs4PbBr6 matrix...... 169

Figure 7-17 | Optical images of phase transition within 10 min. Scale bar is 50μm...... 170

Figure 7-18 | XRD pattern of Cs4PbBr6 and CsPbBr3 microcrystals mixture...... 170

Figure 7-19 | Phase transition between CsPbBr3/Cs4PbBr6 nanocomposite and CsPbBr3 microcrystals...... 171

Figure 7-20 | (a) Photos of pure dye molecules and CsPbBr3/Cs4PbBr6 nanocomposite sealed in epoxy in room light and 365nm UV lamp. (b) The fluorescent spectra of the dye molecules and CsPbBr3/Cs4PbBr6 nanocomposite...... 172

Figure 7-21 | (a) Photos of a series of the samples mixing the different ratios of dye molecules and CsPbBr3/Cs4PbBr6 nanocomposite in room light and 365nm UV. The ratio of the dye molecules and the CsPbBr3/Cs4PbBr6 nanocomposite increases from sample#1 to #4. (b) The fluorescent spectra of mixture samples...... 173

Table 1 | Summary of the performance of AC-OEL with different gates at high frequency and low frequency...... 36

Table 2 | Summary of the performance of flexible organic AC-EL devices with different cathode structures...... 94

Table 3 | Summary of electrical and luminescent properties of white emission AC-OEL devices...... 114

Table 4 | Thickness of thin film with different molar ratio precursors...... 148

Table 5 The PL decay effective lifetime (t 1/2) under 420nm 300fs laser pulse excitation.

First order (k1) and Third order Auger recombination constant (k3)...... 164

XXI

Table of Abbreviations

förster resonance energy AC alternating current FRET transfer full width of half- AFM atomic force microscope FWHM maximum Al aluminum GOS giant oscillation strength tris(8- Alq3 hydroxyquinolinato) HGL hole generation later aluminium high-resolution Bphen bathophenanthroline HRTEM transmission electron microscopy highest energy occupied CE current efficiency HOMO molecular orbital CRI color rendering index ISC intersystem crossing

Ir(MDQ)2(aca bis(2-methyldibenzo[f,h] CNT carbon nanotubes quinoxaline)(acetylacetona c) te) iridium(iii). CB conduction band LEDs light emitting diodes

D/A donor/acceptor ITO indium tin oxide

DC direct current LSM laser scanning microscopy

DMF dimethylformamide LPW limens per watt lowest energy unoccupied DMSO dimethyl sulfoxide LUMO molecular orbital n,n,n',n'-tetrakis(4- energy-dispersive x-ray EDS MeO-TPD methoxy- spectroscopy phenyl)benzidine 4,4′,4′′-tris[phenyl(m- e-h electron-holes m-MTDATA tolyl)amino]triphenylamin e multi-wall carbon EML emission layer MWCNTs nanotubes ETL electron transport layer NCs nanocrystals

EL electroluminescence NPs nanoparticles

external quantum organic light emitting EQE OLEDs efficiency diodes

XXII

2,3,5,6-tetrafluoro- 7,7,8,8- organic F4TCNQ OEL tetracyanoquinodimethan electroluminescence e

fluorescence - poly(3-hexylthiophene- F-P P3HT phosphorescence 2,5-diyl)

poly[n,n'-bis(4- transmission electron Poly-TPD butylphenyl)-n,n'- TEM microscopy bis(phenyl)-benzidine]

PVDF polyvinylidene fluoride TTA triplet-triplet annihilation poly(3,4- 2,2',2"-(1,3,5- PEDOT:PSS ethylenedioxythiophene) TPBi benzinetriyl)-tris(1-phenyl- polystyrene sulfonate 1-h-benzimidazole)

poly[(9,9-bis(3'-((n,n- dimethyl)-n- PFN-Br ethylammonium)- UV ultra-violet propyl)-2,7-fluorene)-alt- 2,7-(9,9-dioctylfluorene)]

poly[(9,9-bis(3'-(n,n- dimethylamino)propyl)- x-ray photoelectron PFN-DOF XPS 2,7-fluorene)-alt-2,7-(9,9- spectroscopy dioctylfluorene)] PVK poly(9-vinylcarbazole) XRD x-ray diffraction polyethylene PET ZnO zinc oxide terephthalate photoluminescence PLQY quantum yield PDMS polydimethylsiloxane

QDs quantum dots

RMS the root mean square

RT room temperature scanning electron SEM microscope selected area (electron) SAED diffraction

XXIII

Abstract

Charge balance in organic light emitting structures is essential to simultaneously achieving high brightness and high efficiency. In DC-driven OLEDs, this is relatively straight forward. However, in the newly emerging, capacitive, field-activated AC-driven organic devices, charge balance can be a challenge. We introduced the concept of gating the compensation charge in AC-driven organic devices and demonstrated that this can result in exceptional increases in device performance. To do this we replaced the insulator layer in a typical field-activated organic light emitting device with a nanostructured, wide band gap semiconductor layer. This layer acts as a gate between the emitter layer and the voltage contact. Time resolved device characterization shows that, at high-frequencies (over 40,000Hz), the semiconductor layer allows for charge accumulation in the forward bias, light generating part of the AC cycle and charge compensation in the negative, quiescent part of the AC cycle.

We also showed that solution-grown films of CsPbBr3 pervoskite nanocrystals imbedded in wide band-gap Cs4PbBr6 can be incorporated as the recombination layer in light-emitting diode structures. More importantly, optical response was studied to shed light on why the very poor light emitter CsPbBr3 becomes a high-efficiency fast emitter when imbedded as nanocrystals in the wider gap host Cs4PbBr6. Kinetics at high carrier density of pure (extended) CsPbBr3 and the nano-inclusion composite were measured and analyzed, indicating second order kinetics in extended and mainly first order kinetics in the confined CsPbBr3, respectively. In these terms, the nano-confinement might be viewed as enforcing mainly geminate recombination of electrons and holes in a material

(CsPbBr3) that did not support stable excitons at room temperature. The resulting first-

XXIV order recombination competes with trapping much more effectively than does bimolecular recombination at the moderate carrier densities typical of LEDs and usual photo-excitation. Analysis of absorption strength of this all-perovskite, all-inorganic imbedded nanocrystal composite relative to pure CsPbBr3 indicated enhanced oscillator strength consistent with earlier published attribution of the subnanosecond exciton radiative lifetime in nano-precipitates of CsPbBr3 in melt-grown CsBr host crystals and

CsPbBr3 evaporated films. Single crystals of the CsPbBr3 - Cs4PbBr6 nanocomposite were also grown in the solution based method which are potentially used as new generation high energy particle detectors.

XXV

Chapter 1 Basics of Radiative Emission in Semiconductors

Conventional lightings rely on either incandescence (Figure 1-1a) or discharge in gases (Figure 1-1b). Both phenomena are strongly linked with significant quantity of energy losses which are essentially inherent because of the extremely high temperatures or large Stokes shift. Semiconductors provide an alternative way of light generation, so- called spontaneous light emission (Figure 1-1c and 1-1d).

Figure 1-1 | Different lighting technologies. (a) Incandescent bulb; (Image courtesy of thorinside via Flickr) (b) Gas discharge tube; (Image courtesy of Tatalux Lighting) (c) Light emitting diode; (Image courtesy of Adafruit) (d) Organic light emitting device. (Image courtesy of LG Display)

As shown in Figure 1-2, spontaneous light emission in semiconductor is the result of radiative recombination of free electrons “negative charges” and holes

“imaginary positive charges”. Both of the free charges are created by current injection overcoming energy barriers. Subsequent radiative recombination of the injected carriers may attain quantum yields close to unity. This phenomenon, called injection

1 luminescence, is the basis of operation of all kinds of thin film light emitting devices.

Therefore, the efficient carrier injection and exciton recombination is essential to highly efficient, bright, long-lived solid-state lighting technology.

Figure 1-2 | Scheme of electron hole recombination resulting in spontaneous light emission in a semiconductor.

Spontaneous Emission

The basic element of a thin film light emission device is a semiconductor electroluminescent structure that comprises, at least, a region of radiative recombination regions of different conductivity type (p and n) that supply the recombining carriers. In the simplest design, the structure relies on a junction between a p-type semiconductor and an n-type semiconductor of the same kind (p-n homojection) with on or both conductivity regions employed as the radiative-recombination region or regions.

Figure 1-3a depicts a band diagram of a p-n homojunction. Under zero bias, the majority electrons from the n-region diffuse into the p-region and majority holes diffuse in the opposite direction. This process creates depleted regions on both sides of the interface. The space charge of the depleted regions creates an internal electric field that counteracts the diffusion. In equilibrium, when the potential barrier is smaller than the

2 bandgap energy, the diffusion current is counterbalanced by the reverse current of minority carriers that drift in this internal electric field.

Figure 1-3 | P-N junction in zero bias (a) and forward bias (b).

When the homojunction is biased in the forward direction as shown in Figure 1-

3b, the reverse currents of the minority carriers change negligibly. Meanwhile, the barrier for majority carriers decreases by qV. Consequently, the majority carrier diffusion current increases by a factor of exp(qV/ kB T) . Enhancement of the diffusion due to the electric field, called injection, results in an excess density of minority carriers on both sides of the homojection. For luminescence, the injected electrons and holes recombine radiatively. Thus, the net forward I-V characteristic can be described by the Equation 1-1 with the consideration of electron-hole nonradiative recombination at the interfaces of the junction.

VIVIRs Rs IIIIIn  p  nr (In0  I p0 )[exp( )  1]  nr 0 [exp( )  1] VVTT

Eqn. (1-1)

3 where In0 and Ip0 are reverse currents for minority electrons and holes, respectively.

VTB k T/ q . Inr is the reverse non-radiative recombination current and  is the ideality factor for the recombination current. (12 )

Charge Carrier Transport

Electrons in the conduction band and holes in the valence band are able to move upon thermal activation, a gradient or an applied electric field. In this section, the concepts of electronic transport in crystalline materials will be described.

Electrons in the conduction band or holes in the valence band can essentially be treated as free carriers or free particles. Even in the absence of an electric field the carriers follow a thermally activated random motion. In thermal equilibrium the average thermal energy of a particle (electron or hole) can be obtained from the theorem for equipartition of the average thermal energy of an electron or hole. The thermal energy of the particle is equal to its kinetic energy, so that the velocity of the particle can be calculated. The mass of the electron is equal to the effective mass of the electron.

Furthermore, the velocity of the electron corresponds to the thermal velocity of the electron, so that the thermal velocity can be determined by

2 mveff t Ek  2

Eqn. (1-2)

At room temperature the average thermal velocity of an electron can be determined by

3kT v  t m eff 4

Eqn. (1-3)

Thermal motion of free carriers can be seen as random collision (scattering) of the free carriers with the crystal lattice. A random motion of an electron or hole leads to zero net displacement of the free carrier over a sufficient long distance / period of time. The average distance between two collisions within the crystal lattice is called mean free path.

Associated to the mean free path we can introduce a mean free time τ. A typical mean free path is in the range of 100nm and the mean free time is in the range of 1ps.

Generally, there are three types of charge carrier transport which are band transport, trap controlled transport and hopping model.

Band-like transport: When a small electric field is applied to the semiconductor material, each free carrier will experience an electro static force which accelerate the carriers along or opposite direction of the field based on the charge of carrier (electron or hole). An additional velocity component caused by an applied electric field (E) will be superimposed upon the thermal motion of the carrier. The additional component is called drift velocity. The drift of the electrons can be described by a steady state motion since the gained momentum is lost due to collisions of the electrons and the lattice. Based on momentum conservation the drift velocity can be calculated. Therefore, the electron and hole mobility can be expressed as

vp vn   and p  n E E

Eqn. (1-4)

5 where 푣푛 and 푣푝 are drifting velocity of electrons and holes. The mobility is directly related to the mean free time between two collisions, which is determined by various scattering mechanisms. The most important scattering mechanisms are lattice scattering and impurity scattering. Lattice scattering is caused by thermal vibrations of the lattice atoms at any temperature above 0 K. Normally, the carrier moves in a highly delocalized plane wave in a broad carrier band with a relatively large mean free path. For organics we expect in the ideal case that the HOMO corresponds to the conduction band and the

LUMO to the valence band.

Trap controlled transport: In contrast to charge carriers immobilized in deep traps, charge carriers occupying shallow states can be excited and therefore contribute to charge-carrier transport.1,2

Figure 1-4 | Carrier transfer in trap controlled transport mechanism.

In disordered semiconductors such transport is usually characterized by an average mobility, μ, and the diffusivity, D. Both μ and D are equally relevant for the proper description of conductivity and recombination.3 The following Einstein equation is widely applied in the physics of conventional and organic semiconductors:

D kT  B  q

Eqn. (1-5)

here kB denotes the Boltzmann constant, and, T represents the temperature q defines the electron charge. In a system characterized with the help of a density-of-states function, this relation may be generalized for an arbitrary temperature to the form of

D n  c qn(/) c

Eqn. (1-6)

where  is the chemical potential and nc represents the total density of mobile charge carriers.4

Figure 1-5 | Density of states with deep carrier traps.

Hoping transport: In inorganic semiconductors, strong covalent bonds hold atoms together with well-ordered configurations. The energy band in this case extends continuously in the bulk and therefore the delocalized carriers can freely move along the band with a relative high mobility.

Figure 1-6 | Carrier hopping and tunneling mechanism in organic semicondcutors

However, for most organic semiconductors, weak intermolecular forces are predominant among molecules or polymers. In this case, discrete energy band structure is dominant in the bulk. The freely propagating wave of charge as usually seen in inorganic semiconductor no longer exists. As a result, carrier transport in organic semiconductors becomes a hopping process that involves thermionic emission and tunneling of carriers between localized sites, Figure 1-6. Many of the hopping models for organic semiconductors are based on the Miller-Abrahams formalism that was originally used to describe the hopping process in deep traps in inorganic semiconductors.5,6 The photon assisted hopping rate between two localized sites is:

vvexp(  2 r  (    ) / k T)  ij 0 ij j i B ji

Eqn. (1-7)

vvexp( 2 r )  ij 0 ij ji

Eqn. (1-8)

12 where v0 is the attempt-to-jump frequency of the order of the phonon frequency 10 ~

13 -1 10 s , γ is the inverse localization radius, rij is the distance between sites i and j, εi and

εj are the site energies. The Eqn. (1-7) describes the tunneling probability from a localized site i to another localized site j separated by a distance rij. The Eqn. (1-8) depicts the thermally assisted hopping rate. For an energetically upward hop, εj > εi, the hopping rate follows an exponential function. For an energetically downward hop, εj < εi, the hopping rate is equal to unity, that is independent of the energy difference of the hopping sites and the temperature.

Exciton Generation and Recombination

Migrating hole and electron charge carriers that encounter each other within the

Coulombic capture radius (rc) combine to form an electron-hole pair sometimes called an exciton.7 This occurs when the Coulombic binding energy is to equal or greater than the exciton thermal dissociation energy kT.8,9 The Coulombic capture radius is defined as follow,

e2 rc  40kT

Eqn. (1-9)

where e is the electron charge, ε is the dielectric constant, ε0 is the permittivity of free space, k is the Boltzmann constant, and T is the temperature. Based on the Eqn. (1-9) the thermal energy increases with the Coulombic capture radius decreases. As the dielectric constants of organic semiconductors are typically low (ε ~ 3), the capture radius is typically found to be around 20 nm.10,11 Studies on the rates of recombination have been treated within the context of recombination of statistically independent opposite charges.12,13 The bimolecular recombination rate can be specified in terms of the product

9 of the hole and electron densities and the bimolecular rate constant defined as γ = e(µ+ +

µ−)/εε0 ; where e represents the electron charge, µ+ and µ− represent the mobilities of the holes and electrons, ε is the dielectric constant, ε0 is the permittivity of free space. From the point of view of molecular states, recombination of holes and electrons can generate either singlet or triplet excitons that may form according to spin statistics and lead to a one-to-three ratio of these states.14 Thus, four possible excited states may exist as shown in Figure 1-7.

Figure 1-7 | Singlet excited state (S=0) and triplet excited states (Sz=+1,0,-1).

The bound hole-electron pair (exciton) is stabilized (relative to unbound and charged states) by a Coulombic attraction known as the exciton-binding energy. At temperatures where the exciton-binding energy is greater than the thermal energy kT, excitons will not dissociate appreciably after formation. Light can be generated from the excitons states depending on the type of materials employed is fluorescent or phosphorescent. The possible pathways for relaxation can be depicted in the Jablonski diagram15 as follows:

Figure 1-8 | Energy diagram illustrating fluorescence, phosphorescence, internal conversion (IC) and intersystem crossing (ISC).

Excluding considerations of allowed transitions or of spin-orbit coupling, the available decay pathways for a singlet exciton include radiative fluorescence and/or nonradiative internal conversion (IC). In addition, intersystem crossing (ISC) from the excited singlet state to the triplet state may occur followed by either radiative phosphorescence and/or non-radiative ISC to the ground state. For the triplet excitons, decay to the ground state can occur via radiative (phosphorescent) and/or non-radiative

ISC. It was believed that phosphorescence is rarely observed at room temperature for π- conjugated organic emitters due to their typically low rate of intersystem crossing associated with weak spin-orbit coupling. The prediction is established on the presumption that the triplet-to-ground state transition is not spin conserving and therefore a forbidden process. However, it was later proven that it is possible to capture both singlet and triplet excitons by using metal organic semiconductors (such as Ir with strong spin-orbit coupling) doped in host materials. It opened the door for extremely efficient

OLEDs having 100% internal quantum efficiency of the exciton recombination.16,17

The Intersystem Crossing (ISC) Model in Singlet-Triplet Dynamics

The ISC model is based on the magnetic field change of the spin dependent intersystem crossing in intermolecular excited states (polaron pairs). Therefore, the intersystem crossing model also is called polaron pair model. The intersystem crossing is the transition between the singlet and triplet levels. In the intramolecular excited state under photo excitation, most of the excitons are in singlet configuration. But in the intermolecular excited state under electrical excitation, the singlet - triplet ratio has the limit of 25%, if the spin orbital coupling effect is not considered. However, the singlet excitons and triplet excitons can be mutually converted when the spin flip mechanism is present due to hyperfine coupling or spin-orbital coupling.18

Figure 1-9 | Electron and hole spin precessions in an external magnetic field B for singlet and triplet e–h pairs.

The Figure 1-9 illustrates that electron and hole spin polarizations coherently process around both internal nuclear magnetic field and external magnetic field. The singlets and triplets excited states are created at e-h capture. In essence, the e–h spin– exchange interaction is accountable for the coherent spin orientations and precessions within an e–h pair. There are basically two roads to influence singlet - triplet ratios.19

First, changing electron and hole coherent spin orientations can directly affect the e–h

12 spin-pairing configurations, and the resultant singlet and triplet ratios at formation.

Second, perturbing the spin correlation between electrons and holes can lead to a mutual conversion between singlet and triplet excited states, and redistribute the singlet and triplet populations. Therefore, an external magnetic field is able to impact de-excitation or revolution processes of singlet and triplet pairs by varying the singlet and triplet ratios involved in excited processes and charge transport.

When an external applied magnetic field is comparable to the internal magnetic interaction in strength, an external magnetic field can further increase the triplet splitting

(namely, external Zeeman effect), and consequently modifies the ISC. In principle, magnetic-field-dependent singlet and triplet ratios can consist of two contributions20, from spin-dependent formation of excited states and from the field-sensitive ISC. There are two necessary conditions, namely magnetic requirements, for the ISC to be magnetic- field dependent. First, the external Zeeman splitting must be larger than the internal

Zeeman splitting. Second, the external Zeeman splitting should be comparable to the singlet–triplet energy difference caused by the e–h separation-distance-dependent spin- exchange interaction.

This magnetic field phenomena has drawn strong interest toward the development of organic spintronics in which electronic, optical, and magnetic properties can be integrated for the development of next-generation semiconductor devices for energy conversion, sensing, and optical-communication applications.

Exciton Oscillator strength

Oscillator strength is a dimensionless quantity that expresses the probability of absorption or emission of electromagnetic radiation in transitions between energy levels

13 of an atom or molecule.21 It can also be defined as the ratio between the quantum mechanical transition rate and the classical absorption/emission rate of a single electron oscillator with the same frequency as the transition.22

In direct bandgap semiconductors, a hydrogen-like series of the transitions to s- states of Wannier-Mott excitons accounts for the fundamental absorption.23 In addition to the free exciton lines, there are surprisingly strong additional absorption lines in the same spectral region.24 They are due to the excitons weakly bound to impurities or defects.

Anomalously high intensity of the impurity-exciton lines indicate their giant oscillator strength of about 1 per impurity center while the oscillator strength of free excitons is only of about 10-4 per unit cell. Shallow impurity-exciton states are working as antennas borrowing their giant oscillator strength from vast areas of the crystal around them. They were predicted by Emmanuel Rashba first for molecular excitons25 and afterwards for excitons in semiconductors.26 Giant oscillator strengths of impurity excitons normally ties with the ultra-short radiative lifetimes less than 1 ns. The relation between oscillator strength and exciton lifetime is given

3 3mc0 i  22 2e nii f

Eqn (1-10)

where m0 is the electron mass, c is the speed of light, n is the refraction index, and i is the frequency of light. Typical values of exciton lifetime are down to nanoseconds range, and the fast luminescence decay favors the radiative recombination of excitons over the

14 non-radiative process.27 With a high quantum yield, the emission process is also named as resonance fluorescence.

Lead Halide Perovskite Emitter

In order to promote color purity of display and lighting devices, research on new emitting materials that can emit light with narrow full width at half- maximum (FWHM) have been a huge trend. Inorganic quantum dot or quantum well emitters with narrow spectra (FWHM ∼ 30 nm) have been significantly studied following organic emitters

(FWHM > 40 nm); however, size-sensitive color purity, difficult size uniformity control, and expensive material costs of inorganic quantum dot (QD) emitters retard the progress for wide use in industry. Therefore, new emitters with size-insensitively high color purity

(FWHM <20 nm) and low material cost should be developed. Among many candidates, metal halide perovskites have gained great attentions and show the possibility for future high-color purity emitters.

Figure 1-10 | (a) Lattice structure of orthogonal Pervoskite structure.28 (b) Red and green Pervoskite LEDs.29

Halide perovskites have a general formula of ABX3, where A and B are monovalent and divalent cations, respectively, and X is a monovalent halide (Cl, Br, I) anion. The three-dimensional (3D) crystal structure of lead halide perovskites is shown in

Figure 1-10, where the B cation, commonly Pb, but also Sn, is coordinated to six halide ions in an octahedral configuration. The octahedra are corner-sharing, with the A cation located in between those octahedra. Lead halide perovskites are further classified into either organic–inorganic (hybrid) or all-inorganic, depending on whether the A cation is

3+ an organic molecule, most commonly methylammonium (MA, CH3NH ), or an inorganic cation (commonly Cs+), respectively.

Figure 1-11 | The color tunable fluorescence of CsPbX3 nanocrystals based on halid ion exchange.30

The optical (fluorescence) and electronic properties of perovskites are tunable

(Figure 1-11) by varying the composition of constituted halide ions and to a smaller

16 degree of the cations.30–32 In addition, the size and dimensionality of perovskites can also be used to tune their optical properties, similar to conventional metal chalcogenide semiconductors.33,34

Figure 1-12 | Quantum size effects in low dimensional perovskite materials.35

In principle, the dimensionality (D) of the inorganic framework can vary from three to zero. Small-sized organic or inorganic cations can fit into the PbX6 octahedra of the 3D framework. If the cation is too large, the 3D perovskite structure would be inappropriate and the dimensionality of the inorganic framework would change to 2D,

1D, or 0D. In the 0-D hybrid perovskite structure, the PbX6 octahedra are isolated, while in the 1D the PbX6 octahedra are connected in a chain, and in the 2D the PbX6 octahedra are connected in layered sheets at the corners.36 The stoichiometry of the perovskite and radiative emission change with the dimension of the inorganic framework as shown in

Figure 1-12.

Chapter 2 Charge Balance in AC- Field Driven Organic Light Sources

Charge balance in organic light emitting structures is essential to simultaneously achieving high brightness and high efficiency. In DC-driven OLEDs, this is relatively straight forward. However, in the newly emerging, capacitive, field-activated AC-driven organic devices, charge balance can be a challenge. In this chapter we introduce the concept of gating the compensation charge in AC-driven organic devices and demonstrate that this can result in exceptional increases in device performance. To do this we replace the insulator layer in a typical field-activated organic light emitting device with a nanostructured, wide band gap semiconductor layer. This layer acts as a gate between the emitter layer and the voltage contact. Time resolved device characterization shows that, at high-frequencies (over 40 kHz), the semiconductor layer allows for charge accumulation in the forward bias, light generating part of the AC cycle and charge compensation in the negative, quiescent part of the AC cycle. Such gated AC organic devices can achieve a non-output coupled luminance of 25,900 cd/m2 with power efficiencies that exceed both the insulator-based AC devices and OLEDs using the same emitters. This work clearly demonstrates that by realizing balanced management of charge, AC-driven organic light emitting devices may well be able to rival today’s

OLEDs in performance.

Introduction of Alternating Current Driven Organic Electroluminescence (AC-

OEL) Device

Carrier drift under uniform external electric field is the dominates mechanism of free carrier injection and transportation in OLEDs.37–39 Electrons and holes are injected as polarons from highly conductive, metallic, cathodes and anodes. They are then

18 transported through the HOMO and LUMO levels of organic semiconductors to a light emission layer respectively, resulting in the formation of excitons and radiative recombination transitions.16,40 Thus, it is crucial to reduce barrier heights for the carrier injection at the metal/organic and organic/organic interfaces.12,41–44 For the same reason, heat dissipation at injection barriers interfaces is not trivial for energy loss in OLEDs.45

Hot carriers exceed the energy gap of the emitting material, so the excess energy is wasted thermally, giving rise to a lower power efficiency. The reduction of the carrier injection barriers by decreasing the difference between adjacent energy levels is now a common method to optimize the efficiency of OLEDs.41,46–49

However, recently many researchers have been drawn to understanding AC-OEL which avoid losses at interfaces by the use of an AC field to create polarization currents in the active organic layer for light emission.50–59 In these devices, power loss is limited to dielectric losses and efficiency and luminance of the AC-OEL has been attributed to the cross-section for the creation of free carrier by the field. Thus they are said to be

“field-activated.”

For most defect free, light emitting, semiconducting polymers, the cross-section for the field generation of a polaron or exciton is small and with the massive reduction of injected carriers from electrodes this means the emitting layer will need to be doped.

This can be done by blending dopants into the emitter or using a hole or electron generation layer as a carrier source.50,53,54,60. Normal hole generation layers include

N,N,N',N'-tetrakis(4-Methoxy-phenyl)benzidine (MeO-TPD): tetrafluoro- tetracyanoquinodimethane (F4TCNQ)50, Poly(3-hexylthiophene) (P3HT):F4TCNQ53,and

4,4',4''-Tris(N-3-methylphenyl-N-phenylamino)triphenylamine (m-

61 50 62 MTDATA):F4TCNQ . Bathophenanthroline (BPhen):Cs , Cs2CO3:BPhen , and

63 2,2′,2"-(1,3,5-Benzinetriyl)-tris(1-phenyl-1-H-benzimidazole) (TPBi):Cs2CO3 are good choices for electron generation electron-donor/electron-acceptor (D/A) system. But, in order to prevent hot carrier direct injection into the device, single or double dielectric layers (in asymmetric or symmetric structures) should be utilized generally64.

50 Unfortunately, the high dielectric constant of such materials (HfO2 , P(VDF-TrFE-

65 59 62 CFE) , SiO2 , and LiF ), means that a nontrivial applied voltage drop on the dielectric layer, rather than functional organic layers, results and high driving voltages are necessary for the insulating and capacitive AC-OEL. Moreover, dielectric losses can be large. Most importantly, however, the insulating nature of this layer means that there is little charge compensation allowed in the device and a space charge can accumulate during operation at high brightness.

In this chapter, we demonstrate that n-type ZnO, a typical wide band gap (~3.2 eV) semiconductor, can be used as an ideal substitution of high-k dielectric layer in AC-

OEL devices, serving the function of a “passive carrier management gate”. By utilizing a

ZnO nanoparticles (NPs) layer between anode and HGL not only hot holes are prevented from direct injection in the negative half of AC cycle, but electron extraction in HGL

[Poly(4-butylphenyl-diphenyl-amine) (Poly-TPD):F4TCNQ] is facilitated. As we show, this essentially balances the charge in the device. This function of the ZnO gate is demonstrated in the Poly-TPD:F4TCNQ heterojunction system for this work, but may be applied generally.

Materials and Characterization

All devices were fabricated on a glass substrate with a pre-coated 100 nm ITO film with a sheet resistance of approximately 10 Ω. The substrates were cleaned in an ultrasonic bath with acetone followed by methanol and isopropanol for 1 hour each. The

ITO substrates subsequently were dried in a vacuum oven for 2 hours and treated with

UV-ozone for 15 min. ZnO NPs aqueous solution (50 wt%, average diameter ~35 nm,

Aldrich) was purchased and diluted to 1 wt% and 5 wt%. The ZnO gate layer was spun coat with 1 wt% and 5 wt% ZnO NPs aqueous solution (50 wt%, average diameter ~35 nm, Aldrich). Then, a doped hole generation layer of doped Poly(4-butylphenyl- diphenyl-amine) (Poly-TPD)/ 2,3,5,6-Tetrafluoro-7,7,8,8- tetracyanoquinodimethane

(F4TCNQ) is fabricated at 2500 rpm in a weight ratio of 10:1. The emission layer consisted of a blend of poly(N-vinylcarbazole) (PVK) as a co-host and 10 wt% fac-tris(2- phenylpyridinato)iridium(III) (Ir(ppy)3) as the dopant. The emitting layers were obtained by spin coating the 18 mg/mL PVK:Ir(ppy)3 blend in chlorobenzene at 3000 rpm. As an electron injection layer, TPBi is dissolved in a mixture of formic acid and water (5:1) and spin cast at 3500 rpm before moving to a heated vacuum oven for 30 min.

Then, a 1 nm layer of LiF and a 100 nm layer of Al was deposited as electrode in high vacuum. The active area of the AC-OEL pixels was 15.0 mm2.

The AC driven field-induced polymer EL devices in this report are measured in ambient air at atmospheric pressure and room temperature (25 °C) without sealing. A 200

MHz function/arbitrary waveform generator (Agilent 33220A) connected to an amplifier

(Trek PZD700A M/S) provides a sinusoidal signal with suitable voltage and frequency. A power analyzer (Zimmer LMG95) is utilized to read the root mean square (RMS) value

21 of voltage and current on the AC-OEL devices. At the same time, voltage waveforms and current waveforms are recorded from a Tektronix TPS 2024B oscilloscope to validate the voltage and current values. A photometer (ILT 1400-A) is used to measure the out- coupling luminance. The entire system is connected and controlled by a computer. In order to maintain accurate and reliable measurements of luminance and efficiency, each turn-on measurement of the pixels was integrated over 2000 ms and averaged 5 times instead of fast sweeping for good-looking curves. SEM and AFM images. A scanning electron microscope (SEM, JEOL 6330) was utilized to measure the morphology of ZnO

NPs gate. The thicknesses of ZnO gates were calibrated by an atom force microscope

(AFM, Asylum Research) and a surface profiler (Alpha-Step 500)

Carrier Gate Structure in AC-OEL Device

The specific configuration of the AC-OEL used in this work is shown in Figure

2-1, with the three-layer emitter complex, the gate structure, and the electrodes marked separately. To fabricate the structure we begin with spin-coating a layer of ZnO NPs 22

(SEM image is shown Fig. 1(a)) onto the 2.5 cm × 2.5 cm glass substrate which has been pre-coated with 100 nm of ITO. This will function as a gate for carrier management.

AFM images typical of 50 nm and 100 nm gate layers spun cast using 1 wt% and 5 wt%

ZnO NP solutions in solvent, are shown in Figure 2-2.

Figure 2-2 | The thicknesses of ZnO gate layers spun coat by different concentrations of ZnO NPs solution (1wt% and 5wt%).

Poly-TPD doped with 10 wt% F4TCNQ was used as HGL and spun cast to a thickness of 70 nm on top of the ZnO layer (using a concentration of 10 mg/mL in chlorobenzene). This layer will function as the main source of holes in our devices.

A guest-host system was used in the light-emission layer (EML). This 150 nm layer contained the highly efficient metallo-organic phosphores, Ir(ppy)3, and a well- known p-type host polymer, PVK. Thus, both singlet and triplet excitons could be harvested during device operation.

Finally, a 30 nm electron transport layer (ETL) of TPBi was spun cast in a glove box and this was then transferred into a high vacuum (~2 × 10−7 Torr) evaporator where a

LiF/Al electrode was deposited. To be clear, we show the molecular structures of the functional organic materials employed in the work in Figure 2-3.

Figure 2-3 | Organic molecule structures employed in this work

We note here that while the overall construction seems similar to those seen in insulating AC organic EL devices50,58 where the HGL and ETL are effectively charge sources activated through a Zener breakdown process with applied field66, in this system the insulators have been replaced by a single gate structure and the HGL and ETL layers must be chosen specifically to allow barrier formation at the interfaces.

To fully understand the mechanisms of charge generation in D/A system (Poly-

TPD:F4TCNQ), we must first consider the two sources of holes as shown in Figure 2-4:

Figure 2-4 | The energy level diagram of the AC-OEL devices

Source I (Polarization current)

The first charge generation process is associated with the polarization current which normally exists in materials under an applied time-varying electric field.

Molecules are being polarized under the influence of the electric field and displaced positive and negative charges are accumulating on sites throughout the volume randomly, as a bound charge density. These electron-hole pairs can become unbound in sufficiently large field assuming there is a local defect to assure momentum conservation.

Alternatively, the bound pair can migrate in the field to the Poly-TPD/F4TCNQ interfaces.67 At the interface, these electron-hole pairs may form charge transfer states that dissociate to mobile carriers.68 Since the electric field varies with time, then the charge displacement is time dependent. The displacement yields a polarization current which is proportional to the time derivative of the field (퐼 = 푑퐸⃑ ⁄푑푡).69

Source II (Electron-donor/electron-acceptor heterojunctions)

Poly-TPD and F4TCNQ are known as a strong electron donor and accepter respectively. When 10% F4TNQ is doped in Poly-TPD, electrons can easily tunnel from

HOMO states of electron donor (Poly-TPD) to the LUMO states of electron acceptor

(F4TCNQ) through a narrow depletion zone.70 This suggests that an electron-hole pair generated in doped organic p-n heterojunctions, dissociates under intense electric field.

The electric-field-assisted bipolar charges spout from the internal charge separation.71,72

The combination of these two sources of charge generation is responsible for the electron-hole pair population in the HGL. The unstable electron-hole pairs are easily dissociated into free holes in the HOMO state of Poly-TPD and electrons in the LUMO state of F4TCNQ at the interface of Poly-TPD/F4TCNQ. Under external electric field, free holes drift into EML for light emission.

The gate structure, however, changes the balance of these charge sources according to frequency and applied voltage. The frequency-dependent characteristics of

AC-OEL devices with ZnO gates of different thicknesses (50 nm, 70 nm, 100 nm, and

120 nm) are shown in Fig. 2(a,b). There is a noticeable difference in the AC-OEL devices’ current and emissive response with frequency as the gate thickness is increased.

For instance, Figure 2-5 shows a strong thickness-dependence in carrier injection

(current) in the low frequency range (below 500 Hz). Near DC frequencies (< 50 Hz), the

AC-OEL devices with 50 nm, 70 nm, 100 nm, and 120 nm gates all show relatively low levels of light generation (470 cd/m2, 620 cd/m2, 650 cd/m2, and 760 cd/m2) at current densities of 10.4 mA/cm2, 13.0 mA/cm2, 16.7 mA/cm2, and 18.5 mA/cm2, respectively.

However, these luminances increase dramatically at 40 kHz for instance (1,130 cd/m2,

1,230 cd/m2, 1,420 cd/m2, and 1,550 cd/m2), nearly twice that of low frequency driving, while the current densities of the devices remain at the nearly same level (18.7 mA/cm2,

18.5 mA/cm2, 14.5 mA/cm2, and 14.7 mA/cm2, respectively). Thus, it is reasonable to suggest that the current densities at very low and DC frequencies mainly depend on the contribution of external hot carrier injection, rather than carrier generation within the devices. This is made clearer in Figure 2-6, α , β , θ , and δ are slopes of lines connecting the data points between at 50 Hz and the data points locating at luminance peaks at high frequency. The larger the slope the higher the injection barrier for free carriers near DC.

We attribute this graphical result: α > β > θ > δ , to the fact that the voltage drop over

ZnO layer dramatically increases. Note that the dielectric constant (~3) of the organic multi-layer is incompatible with that of the inorganic semiconductor gate layer (~8). So, the tunneling probability of the positive charges through gate is greatly suppressed. This gives a lower current density (for near DC frequencies) with the thicker ZnO gates. Thus, we reason that carrier injection is the dominant mechanism of free charge in these ZnO

27 gated AC-OEL devices at low frequencies and the behavior is similar to that of an AC driven OLED63.

Conversely, the high-frequency electric field (where the field is held constant

RMS magnitude) induces a large population of the free carriers due to the contributions of charge injection and polarization. The luminance and current peaks in the frequency sweep are the direct evidence to show this point.

Gate Carrier Valve Mechanism

Figure 2-6 | Current density versus voltage characteristic under DC driving in the devices of ITO (100nm)/ZnO (~100nm)/Poly-TPD:F4TCNQ (~70nm)/MoO3 (15nm)/Au (100nm) and ITO (100nm)/ZnO (~100nm)/MoO3 (15nm)/Au (100nm).

To further investigate the carrier manipulation of semiconductor gate, the hole- only devices with a ZnO gate were studied using the structure: ITO/ZnO NPs(~100 nm)/Poly-TPD: F4TCNQ(~70 nm)/MoO3(15 nm)/ Au(100 nm) and ITO/Poly-TPD:

F4TCNQ(~70 nm)/MoO3(15 nm)/Au(100 nm). Figure 2-6 show the JRMS as a function of

VRMS for devices with and without ZnO gate layer under the forward and reverse DC bias. In the absence of ZnO gate, the device exhibits nearly symmetric curve (584.6 mA/cm2 at 3.57 V versus -293.87 mA/cm2 at − 3.77 V). This is because the symmetric energy barriers at the interface (~0.3 eV) of ITO and HGL and that of HGL and Au (~0.1 eV) for carrier injection. However, the device with the ZnO semiconductor shows a

“gating” effect in the I-V characteristics (8.2 mA/cm2 at 3.67 V versus − 101.5 mA/cm2 at − 3.78 V). We illustrate this using a band alignment diagram of the hole-only devices in Figure 2-6.

Figure 2-7 | Energy band diagrams of AC-OEL devices in forward and reverse bias.

In the forward cycles, the direction of external electric field points from ITO towards Au. The high density of positive charges in the HOMO state of Poly-TPD are strongly drifted to MoO3 then to Au. On the other hand, the excess electrons trapped in

HGL are transferred by the strong electric field and encounter the energy barrier at

HGL/ZnO interface, then tunnel to the CB of ZnO. The electron extraction in HGL dramatically facilitates the electron hopping rate from Poly-TPD to F4TCNQ, resulting in 29 a larger population of positive charge. In other words, the ZnO gate “opens the door” of electron extraction and hole regeneration in HGL. As the applied AC voltage switches to reversed fields in the power cycle, the electric field tilts the Fermi level. Holes in the

LUMO of Poly-TPD are driven towards the opposite direction and are accumulated at the interface of ZnO and HGL because of high tunneling barrier (~2.2 eV). The ZnO gate

“shuts the door” for electron extraction before the next forward cycle arrives. A high electron injection barrier (~2.5 eV) is observed as well. The accumulated carriers trapped in HGL will be moved or neutralized in the next cycle of charge regeneration. Therefore, the ZnO layer plays a role of electron extraction layer (“gate on”) in the forward bias and works as dielectric layer (“gate off”) in the reverse cycle.

Figure 2-8 | Time response of cell current in ITO (100nm)/ZnO (~100nm)/Poly- TPD:F4TCNQ (~70nm)/Au (100nm) and ITO (100nm)/ZnO (~100nm)/Au (100nm).

Figure 2-8 shows the time-resolved current characteristics under AC driving.

Waveforms of current density indicate not only increased amplitude with rising RMS driving voltage but are also positive (without ZnO) or negative (with ZnO). More importantly, the negative current density in the device with ZnO gate suggests that the free electrons are transferred from Au side to ITO side. In the perspective of AC driven devices, the ZnO gate facilitates or “pumps” the excess electrons trapped in HGL, resulting in a high population of positive charge.

To further validate the electron extractor and hole blocker function of the ZnO gate, under intense high-frequency electric field in the gated AC-OEL devices, the Poly-

TPD:F4TCNQ (doped HGL) was substituted by pure Poly-TPD or F4TCNQ in AC-OEL devices or absence. The experimental JRMS-VRMS curves of these devices are shown in

Figure 2-9. In the AC-OLED device, without the electron blocking layer, electrons are

31 injected from cathode, then transferred to the anode without recombination, which causes a poor EL performance and a low breakdown voltage (5.5 V). This, we suggest, is explained by the fact that the device missing HGL layer, establishes an electron passage through the LUMOs of functional organic layers to CB of ZnO (shown in Figure 2-9).

Meanwhile, the same tunneling behavior was observed when a layer of F4TCNQ was sandwiched by ZnO and PVK:Ir(ppy)3. The JRMS of the device increases rapidly as function of VRMS because electrons injection from Ir(ppy)3 can be drifted to ZnO NPs through F4TCNQ molecules which is an excellent electron acceptor (shown in Figure 2-

9). An identical curve was found for pure Poly-TPD since free electrons dissociated from the electron-hole pairs due to AC electric field have to stay in the LUMO state of Poly-

TPD (shown in Figure 2-9), rather than be transferred to the HOMO states of the electron acceptor (F4TCNQ) in D/A system (shown in Figure 2-9). The high density of electrons in Poly-TPD enormously increases the possibility of electron breakdown in the device. A turning point in J–V curve can be observed and confirms the fact that the device is breaking down at a voltage of 6.7 V. In contrast, in the Poly-TPD:F4TCNQ device, current density increases without breakdown while applied AC voltage increases since free charges are well-managed at the ZnO/HGL interface. The electron-hole pairs are dissociated into free electrons and holes staying in the HOMO state of F4TCNQ and the

LUMO state of Poly-TPD, respectively, illustrated in Figure 2-9. In the forward bias of

AC cycles, the injected electrons in PVK will not break through the barrier to reach ZnO, but instead, hop to the LUMO state of Ir(ppy)3 for exciton recombination.

We use a previously established tunneling model73,74 coupled with a P-N junction

75 model to further interpret the experimental data. The modeling results of the JRMS -

VRMS characteristics of the devices are in excellent agreement with experimental measurements in Figure 2-9. Together with the experimental demonstration, this suggests that the charge generation mechanism due to field polarization indeed has great impact on majority of free hole carrier population, and that the ZnO gate has a non-trivial influence on carrier extraction and balancing of the dual carrier injection and transport under an AC electric field.

No Gate vs Semiconductor Gate vs Insulator Gate

So far, we have demonstrated the electronic functionality of the ZnO semiconductor “gate” in AC-OEL devices. Specifically, the ZnO gate is responsible for carrier management in the Poly-TPD:F4TCNQ by facilitating (gate on) and blocking

(gate off) the electron transport passage in different halves of AC power cycle.

Furthermore, it is worth comparing different gates, 100 nm ZnO, 100 nm P(VDF-TrFE-

CFE) in a same AC-OEL devices configuration, and AC-OLED directly.

(a) (b) (c)

Figure 2-10 | (a) RMS current density as function of RMS voltage. (b) Luminance as function of RMS voltage. (c) Luminance as function of frequency.

First, the “turn-on” voltage: in Figure 2-10, the turn-on voltage at 40 kHz of the device with ZnO gate is 10.5 V (RMS value) which is much lower than that (29.1 V) of the device with P(VDF-TRFE-CFE) insulator. However, turn-on voltages of the devices with the ZnO gate (10.5 V) and without (9.0 V) are nearly identical. Second, the maximum brightness: Figure 2-10 shows that the peak brightness at 40 kHz of the AC-

OEL device with the ZnO gate is as high as 25,900 cd/m2 at 28.4 V while the P(VDF-

TRFE-CFE) device has a luminance of 770 cd/m2 at 41.9 V (maximum brightness 6,030 cd/m2 at 108.9 V is not indicated in the plot). As a reminder, we state that these devices are built on glass without the aid of output coupling and output was measured using a fiber optic light probe. Thus light captured by the glass substrate is not accounted for.

The devices without gate or insulator exhibit a maximum luminance of 13,940 cd/m2 at

23.8 V. However, the current characteristics in Figure 2-10 show the opposite trend; that the devices with ZnO gate, P(VDF-TRFE-CFE) insulator, and AC-OLED have current densities of 231.0 mA/cm2, 258.7 mA/cm2, and 215.2 mA/cm2, respectively. This suggests the carrier transport within the devices have become unbalanced shifting the recombination zone away from the emitter layer. Third, the low frequency response: In

Figure 2-10, we observed that the luminance is greatly suppressed to 10 cd/m2 at 50 Hz by the P(VDF-TRFE-CFE) insulator, whereas, AC-driven OLED shows a much higher luminance of 2,240 cd/m2 due to the dramatically enhanced carrier injection. As would be anticipated, a luminance of 680 cd/m2 is obtained in the device with ZnO gate, which

34 demonstrates the gate function of ZnO as distinguished from the insulator or direct injection devices.

(a) (b) (c) Figure 2-11 | The carrier transportation mechanisms of AC-OEL devices with ZnO gate and P(VDF-TrFE-CFE) insulator, and AC-OLED.

This suggests that the significant optical and electric differences of the various charge injection responses are due to the charge transportation properties and band gap structures of semiconductor and insulator. The dielectric constant of P(VDF-TrFE-CFE), as high as 40 at 40 kHz, means that the carrier injection is limited and the applied electric field is not enough to effectively generate carriers within the emitter. There is enormous free charge accumulated at the interface of the P(VDF-TrFE-CFE) and HGL in Figure 2-

11. In the case of ZnO, the gate facilitates electron extraction from HGL (Poly-

TPD:F4TCNQ) which greatly enhances the generation rate of positive charge. Electron extraction over ZnO and hole regeneration in Poly-TPD:F4TCNQ tremendously promotes carrier injection in positive half of AC power cycle at low voltage (see in

Figure 2-11). In the reversed bias, the hole carriers are trapped at the ZnO/HGL interface

35 resulting in a lower current density in Figure 2-11 compared with the devices that have no gate or utilize an insulator. The theoretical analysis on carrier injection and transportation is confirmed by power efficiency in Figure 2-11. The 40 kHz driven lighting device with ZnO gate exhibits the highest efficiency of 72.9 lm/W, compared

25.4 lm/W with P(VDF-TrFE-CFE) and 50.3 lm/W in AC-drive OLED. Table 1 summaries the performances of AC-OEL devices with different gates at high frequency

(40 kHz) or low frequency (50 Hz).With comparison to the power efficiency as low as to

2 lm/W at 50 Hz, we conclude that polarization current contribution facilitates more efficient carriers injection than the direct injection of hot carriers, which exhibits dramatically high power efficiency at 40 kHz.

Table 1 | Summary of the performance of AC-OEL with different gates at high frequency and low frequency.

High frequency (at 40kHz) Low frequency (at 50Hz) Power Max. Turn-on Max. Turn-on Max. eff. at Gate Power voltage brightness voltage brightness Max. material eff. [V] [cd/m2] [V] [cd/m2] brightness [lm/W] [lm/W] 25,900 @ 15,620 @ ZnO 10.5 72.9 7.9 1.2 28.4 V 26.8 V P(VDF- 6,030 @ TrFE- 29.1 25.4 63.3 < 100 < 0.1 108.9V CFE) AC- 13,940 @ 20,130 @ 9.0 50.3 6.7 1.7 OLED 23.8 V 19.7 V

The crucial condition for the ZnO gate to be “open” or “closed” is band alignment of the AC-OEL devices and the direction of net electric field applied to the device. Thus, a reasonable external DC electric field can promote the electron extraction from HGL to

ZnO, resulting in a higher density of hole carriers. Conversely, the direct current injection

36 can be significantly suppressed due to the high k insulator. So it is not hard to understand the inset in Figure 2-12, which shows the luminance - VRMS curve where a DC offset has been added to the device during AC driving.

v v

(a) (b) Figure 2-12 | (a) Power efficiency plot as function of luminance. Inset shows the phase angle characteristics. (b) Luminance versus RMS current density characteristics of the AC-OEL devices with ZnO gate and P(VDF-TrFE-CFE) insulator, and AC-OLED while an increasing DC voltage offset are added to AC driving voltage. 7.1V and 4.2V of DC voltages are added to AC voltages when AC-OLED and ZnO device reach the maximum luminance respectively. However, the analysis of the luminance - VRMS characteristics comparison between ZnO gated and AC-OLED in Figure 2-12 is more complicated. This is because the luminance - VRMS curves are divided into parts (zone “1” and “2” for ZnO gate; zone “I”, “II”, and “III” for AC-OLED). For the device with ZnO gate, an external DC electric field accelerates the rate of electron extraction over the ZnO and establishes hole accumulation, leading to the increasing brightness (zone “1”). In the device without gate, the DC electric field promotes electron extraction as well as hole extraction, which continuously increases current density. However, the luminance of the AC-OLED drops rapidly after a small bump (between zone “I” and “II”) because the positive external DC

37 offset trades off against the slight negative offset in AC waveforms in the first stage, resulting in a reduced total voltage. When the external DC electric field is nontrivial compared to AC field, both of the devices (with and without ZnO gate) work as traditional OLEDs and show the peak luminances of 12,300 cd/m2 at 309.3 mA/cm2 and

6,990 cd/m2 at 575.8 mA/cm2 (zone “2” and “III”), respectively.

Chapter Summary

In this chapter, we propose that, due to the gating effect at the interface between n-type and p-type semiconductors, ZnO is an ideal gate material to replace traditional dielectric layers in the AC-driven EL devices. Based on general band alignment in AC-

OEL devices, the energy band diagrams at the interface of gate and HGL in the forward and reverse cycles of AC voltage are given and demonstrated in hole-only device experiments under positive and negative bias. Four distinct types of devices with different HGLs (F4TCNQ, Poly-TPD, Poly-TPD: F4TCNQ, or absence) were fabricated to confirm the validity of ZnO gate function and hole generation mechanism due to polarization in the experiment and modeling simulation. A comparison between AC-

OLED and AC-OEL devices with ZnO and P(VDF-TrFE-CFE) reveals distinct trends in the performance of luminance, frequency-dependent, and DC-offset characteristics. From the viewpoint of the broader electroluminescence community, inorganic semiconductor gates provide a novel strategy to utilize free charge injection as well as polarization current in organic thin film EL devices for the development of next generation, high brightness, high power efficiency, AC-DC hybrid electroluminescence devices.

Chapter 3 Tailoring Spin Mixtures in Color-Tunable AC-OEL Devices

In this chapter, we show that the spin dynamics of excitons can be dramatically altered by Maxwell magnetic field coupling together with an ion-enhanced, low internal splitting energy organic semiconducting emitter. By employing an unique, AC-OEL device architecture that optimizes this magnetic field coupling, almost complete control over the singlet to triplet ratio – from fluorescent to phosphorescent emission in a single device is realized. We attribute this spin population control to magnetically sensitive polaron-spin pair ISC that can be directly manipulated through external driving conditions. As an illustration of the utility of this approach to spin-tailoring, we demonstrate a simple hybrid (double layer) fluorescence-phosphorescence (F-P) device using a polyfluorene-based emitter with a strong external Zeeman effect and ion-induced long carrier diffusion. Remarkable control over de-excitation pathways through control of the device driving frequency is achieved resulting in complete blue-red color tunability of the emission. The picosecond photoluminescence (PL) spectroscopy directly confirms that this color control derives from the magnetic manipulation of the singlet to triplet ratios. These results may point the way to far more exotic organic devices with magnetic field-coupled organic systems poised to usher in an era of dynamic spintronics at room temperature.76

Spin Remixing in Electroluminescence

In today’s current-driven organic electroluminescence devices, or OELs of which organic lighting diodes are a subset, free charges are injected through metal contacts into an organic semiconductor where they can recombine into an electron-hole (e-h) pair and decay to emit a photon37,38 (or they can be transported through the layer without

39 interacting). Quantum statistics dictates that the fraction of spin pairs formed in the spin- triplet excited state is generally fixed at 75%. The spin-singlet states make up the remaining 25% of the excited state population.77,78 This means, when re-combining injected, spin-1/2, polaron-spin pairs for electroluminescence, three out of four possible spin-combinations will be triplets (with nonradiative or IR decay routes) and one of the four will be a singlet (fluorescent emission). This has prompted the use of triplet scavenging dyes that resonantly transfer the triplet energy to a metal complex (e.g. Pt, Os,

Ir, Au, Pd, or Ru) with strong L-S coupling16,17,40, allowing for phosphorescent emission and higher device efficiency.

The resonant transfer of energy is quite useful for lighting and display applications. However, we must realize that lifetimes involved in triplet transfer or decay, limits on current flux and dye concentrations due to co-localization of triplet energy, and subsequent quenching, among other considerations, means that resonant energy transfer doesn't address issues that spin-triplets cause for high performance light emitting applications. These potential but unrealized, applications include the electrically stimulated organic laser,79 the photonic spin valve, organically based optical computing paradigms and telecommunications. Quite simply, without more complete spin- population control, it is unlikely we will see any of these technologies realized using the organic platform.

Until recently, there didn't seem to be any way to beat the statistics of this problem. But, in 2007 20 a hint was provided experimentally in negative magnetoresistance of organic semiconductor device. The typical 75%:25% ratio of triplets to singlets is not maintained when external, static magnetic fields are present.

This has been attributed to a spin remixing in the polaron-spin pairs that is typically found in organic systems.18,20,80,81 Simply, the external magnetic field can perturb the coherent relationship between electron and hole spin precessions.82–84 Therefore, an enhanced singlet-related emission can be observed in a florescent semiconducting polymer that has an low internal splitting energy or strong Zeeman effect. However, phosphorescent organic semiconductors do not exhibit the spin remixing property in external magnetic field since the heavy metal atom can significantly raise the internal splitting energy of the organic compound.18 Moreover, in magnetically-coupled OEL devices, the redistribution of singlet and triplet electron-hole pairs can give rise to a significant change in electrical current in the semiconductor, through dissociation85–89 and charge reaction90–93. The real key to utility, however, is to extend this control over the spin populations.

By utilizing the magnetic field spin remixing phenomena, in our present work, a novel device architecture is optimized to self-couple the time dependent, Maxwell field into the ion-assisted emitter. With the right choice of emitters in such structure, a high color contrast in spin-state ratios for a single compact device structure can be achieved.

Color tunable pixels have an unique potential for use in ultra-high-resolution information displays39,51 since they allow for an optimal fill factor and impart further momentum to realizing HD micro-displays. One of most common color-tuning concepts is to apply voltage-dependent color shift by a spatial shift of recombination zone, or exciton redistribution within a multilayerd device.94,95 However, this leads to inevitable brightness change and quenching in the high fields, making the voltage-dependent color tunable devices more difficult and expensive to commercialize. Another method seen in

41 the literature is to stack two or three color independent OEL segments into a tandem device.51,96 However, stacked tandem EL structures requires complicated, multiple thin film, deposition and designs of electrical power for multiple electrodes.

In the following section, a color-change hybrid AC driven, OEL device is driven with a high frequency, AC electric field. This yields a Maxwell AC magnetic field which is coupled to the organic emitter - a p-n emitting interface. The internal AC magnetic field allows for the manipulation of the ratio between singlet and triplet electron-hole pairs in an n-type fluorescent material by ISC, yielding the generation of diffusive secondary carriers. Then, these ion-enhanced secondary carriers produce triplet-spin excitons in a proximal phosphorescent organic matrix. In the magnetically-coupled F-P hybrid OEL device, we successfully shifted CIE coordinates of the radiative output from

(0.23, 0.34) to (0.53, 0.40) by manipulating driving frequency from 50Hz to 60,000Hz with no significant brightness change.

Materials and Methods

Materials and device fabrication

AC-OEL devices were built on a 2.54cm × 2.54cm glass substrate pre-coated with

140nm thick layer of ITO having a sheet resistance ~ 10 Ω/□. These ITO glass substrate are cleaned in an ultrasonic bath with acetone followed by methanol and isopropanol for

1 hour each, and then dry-cleaned for 30 min by exposure to an UV-ozone ambient. To efficiently control the carriers transport under AC driving, PEDOT: PSS doped with

18wt% ZnO NPs (~35nm) was spun onto the substrate to form a gate and hole generation layer. As to dual emission layers unit, a layer of PVK with 3wt% Ir(MDQ)2(acac) was

42 spin-coated using 10 mg/mL in chlorobenzene at 2000 rpm, followed by baking at 100 ℃ for 30min. The second emission layer was obtained by spin coating the 5 mg/mL (20nm),

8mg/mL (50nm), or 10mg/mL (70nm) of PFN-Br blend in methanol at 3000 rpm and dried at 95℃ for 20min. The Iron oxide magnetic nanopowder (Fe3O4 NPs ~ 30nm) was purchased from Sigma-Aldrich and was pre-functionalized by N-succinimidyl ester for easy dispersion in suitable solvent. A 24 mg/mL electron-transport material (TPBi) was dissolved in formic acid: deionized water (FA:H2O = 3:1) mixture and spun cast onto the

EML at a spin speed of 4000 rpm followed by drying at 120 ºC for 30 min. The 150nm top Al electrode was deposited at a rate of 2 Å/s by thermal evaporation through a shadow mask with 0.16cm2 opening. Before carried out for test, the devices were sealed by quartz caps with UV curing adhesive at nitrogen atmosphere in glove box.

EL and PL measurement

The AC driven field-induced polymer EL devices in this report are measured in ambient air at atmospheric pressure and room temperature (25℃) with encapsulation. A

200 MHz function/arbitrary waveform generator (Agilent 33220A) connected to an amplifier (Trek PZD700A M/S) provides a sinusoidal signal with suitable voltage and frequency. A power analyzer (Zimmer LMG95) is utilized to read the RMS value of voltage and current on the AC-OEL devices. At the same time, voltage waveforms and current waveforms are recorded from a Tektronix TPS 2024B oscilloscope to validate the voltage and current values. A photometer (ILT 1400-A) is used to measure the out- coupling luminance. The entire system is connected and controlled by a computer. In order to maintain accurate and reliable measurements of luminance, each turn-on measurement of the pixels was integrated over 2000 ms and averaged 5 times instead of

43 fast sweeping for good-looking curves. The EL spectra for color tunable AC-OEL devices were collected by an ILT 950 spectroradiometer (International Light

Technologies) while the driving frequency was varied easily by function waveform generator. The PL spectra for PVK, PFN-Br, and PVK:Ir(MDQ)2(acac)/PFN-Br thin films were measured by a fluorescence spectrometer (PerkinElmer LS50B).

Streak camera test

For measurement of the transient PL characteristics, short-pulse excitation with a pulse width of 200 fs and a wavelength of 280 nm was used in combination with a

Hamamatsu C2830 streak camera system mainly consisting of C2830 streak camera and

M2547 fast sweep unit (resolution < 10ps). Fluorescence and phosphorescence lifetime measurements were carried out by a photodiode with UV laser illumination in dark-room ambient at 25℃.

SEM and AFM images

A scanning electron microscope (JEOL 6330) was utilized to analyze the morphology of Fe3O4 NPs dispersed in PFN-Br methanol solution. The surface morphology analyses of PVK and PFN-Br were made by an atom force microscope

(Asylum Research MFP-3D-BIO).

XRD pattern

The crystalline phase analysis was performed by an X-ray powder diffractometer

(Bruker D2 PHASER) in ambient. The result was averaged over 10 times.

Fluorescence-Phosphorescence (F-P) Hybrid Emission Layer

Figure 3-1 | Compositional schematic of devices with fluorescence-phosphorescence (F- P) emission interface driven by AC power source (50Hz ~ 70,000Hz). Singlet spins (blue circle pairs) and triplet spins (red circle pairs) are formed at the PVK/PFN-Br interface (or F-P interface).

Our device architecture (Figure 3-1) borrows from the “gated” AC-driven organic emitters introduced recently in literature.97 In this case, however, the gated structure is coupled to a bilayer organic emitter consisting of PVK : Ir(MDQ)2(acac) and PFN-Br.

Surface morphology analysis of PVK: Ir(MDQ)2(acac) and PFN-Br emissive layers are made via AFM and the images can be found in Figure 3-2.

(a) (b)

(c) (d) Figure 3-2 | AFM images of PVK:Ir(MDQ)2(acac) before annealing (a) and after 100 degrees annealing (b); Surface morphology of PFN-Br before annealing (c) and after 100 degress annealing (d).

Figure 3-3 | Chemical structures of PFN-Br and Ir(MDQ)2(acac). Photo luminescence of PFN-Br and Ir(MDQ)2(acac) under 365nm UV lamp.

As in the earlier work98 the carrier gate layer was composed of PEDOT:PSS loaded with ZnO NPs, and the electron transporting layer was TPBi. The two conductive electrodes were ITO and Al. The chemical structures of emitters, PFN-Br and

Ir(MDQ)2(acac), are listed in Figure 3-3. With 365nm UV excitation, PFN-Br and

Ir(MDQ)2(acac) show strong fluorescent and phosphorescent luminescence peaking at

474nm and 585nm respectively (as shown in Figure 3-3). The lifetimes of short-live blue fluorescence and long-live red phosphorescence are given in Figure 3-4 for 0.31ns

46 and 1.87us respectively. With ZnO NPs (~35nm in diameter) gating the interface of

ITO/PEDOT:PSS so that injected charges can be efficiently manipulated in the forward and reversed bias of AC cycles.50,52,98 The AC-OEL devices are driven by sinusoidal voltage with a wide frequency range from 50Hz, which causes the gate to allow for bipolar injection and the device acts like diode, to 60,000Hz in which most light is created by field-generated polarons/excitons with little diffusive transport in the active volume of the emitter.

Figure 3-4 | Lifetime of PFN-Br’s fluorescence and Ir(MDQ)2(acac)’s phosphorescence.

Maxwell Magnetic Field Coupling

We note that PFN-Br is a high performance ionized electron transporting polymer99,100 with electron mobility of 1.41 ×10-7 cm2V-1s-1 and PVK is a typical p-type semiconductor with a hole mobility in the order of 1.0 ×10-6 cm2V-1s-1. Thus, electrons and holes are transferred and interact at the PVK/PFN-Br interface under external electric field. At different driving frequencies, there is more or less time for accumulation at this interface. However, as already noted at high frequencies, the gate of the system allows

47 only for field-generated carrier injection into the emitting volume, while at lower frequencies the gate allows for direct injection from the contacts.

Figure 3-5 | Schematic diagram of dynamic field analysis at p-n magnetic interface.

Never-the-less, both conditions result in drifting charge heading toward the interface. The electrons and holes are transferred to the heterointerface in the positive cycle of AC electric field and drifted along the opposite direction in reversed bias.101–103

Therefore, time-dependent electric field generates an interfacial magnetic field at

PVK:Ir(MDQ)2(acac)/PFN-Br heterointerface based on Maxwell’s equations. In an ideal case, the pixel dimension is 4mm × 4mm, significantly larger than its thickness (~300 nm) – so it is reasonable to ignore fringing effects (infinite area parallel plate capacitor assumption). Via engaging a high frequency driving (60,000 Hz), and strong AC

48 electrical field (1.6×108 V/m), the temporal and spatial characteristics of the internal magnetic field are shown in Figure 3-6. The upper and lower half plane represent the opposite “clock directions” of magnetic field in the positive and negative halves of an AC cycle. The amplitude of the magnetic field is estimated to be approximately 0.85 mT.

Figure 3-6 | Simulated results of magnetic and electric fields at p-n junction at 60,000Hz. The maximum self-generated magnetic field is 0.85mT. In the positive and negative half of AC driving cycle, the directions of vertex magnetic fields are clock wise (z > 0) and counter clock wise (z < 0) respectively.

Meanwhile, the heterointerface is also playing the role of an electron-hole pair recombination zone for hot carrier injection as shown in energy level diagram in Figure

3-5. These electron-hole pairs move in the applied electric field, but also experience the induced magnetic field. The device shows a blue emission due to the PFN-Br fluorescence at near DC driving (50Hz) since the dissociation rate of electron-hole pairs is trivial in the absence of induced magnetic field ( < 0.00005mT). When this internal

AC magnetic field is of the same order as the nuclear hyperfine field (~1mT), ISC suppression should occur80 and this would naturally lead to singlet-spin electron-hole pair accumulation. A large number of secondary carriers will be produced in PFN-Br through the magnetically-mediated dissociation of the electron-hole pairs. The secondary charges

49 are diffused to nearby Ir(MDQ)2(acac) sites, which yields decay of triplet-state excitons.

We note that there is no significant position shift of recombination zone in the device as proven in Figure 3-7.

Figure 3-7 | Electroluminescence spectrum with PVK: 3wt% Firpic/PFN:DOF: 3wt% Ir(MDQ)2(acac).

Frequency Controlled Color Tunability

Figure 3-8 | Images of the blue emission and red emission at the frequency of 50Hz and 60,000 Hz respectively. Electrical and magnetic field properties of PVK/PFN-Br interface. So as illustrated in Figure 3-5, in the low frequency driving regime (50Hz ~

1,000Hz), hot carrier injection is the main mechanism for fluorescent excitons in PFN-Br.

In the high frequency regime (30,000Hz ~ 70,000Hz), the high intensity AC magnetic field at the F-P interface greatly populates singlet-excited e-h pairs via ISC suppression, which leads to secondary carriers. The secondary carriers exist in form of bonded electrons in PFN-Br polymer matrix, more specifically with Br atoms which are strong electron acceptors. The charged Br ions significantly improves the carrier diffusion length104,105, resulting in movable negative charges across interfacial energy barrier.

Consequently, the secondary carriers are transferred to Ir(MDQ)2(acac) for red phosphorescent emission. For the same reason, the charged movable Br ions greatly

51 facilitate magnetic-field current even in very subtle magnetic intensity with non-ionized polymer which normally needs over hundreds of mT.106

(a) (b) Figure 3-9 | (a) At 50Hz, electroluminescence (EL) spectrum with RMS voltage varying from 15.4V to 20.4V corresponding the luminance of 15 cd/m2 to 210 cd/m2. Hot carrier injection is only current contribution. (b) At 60,000Hz, EL spectrum with RMS voltage varying from 17.9V to 22.0V corresponding the luminance of 40 cd/m2 to 240 cd/m2.

Figure 3-9a and 3-9b show the evolution of the device’s EL spectrum with increased electric field at low and high frequencies. At the frequency of 50Hz, there are trivial spectral shifts in Figure 3-9a when applied voltage varies from 15.4V to 20.4V.

Meanwhile, no phosphorescence is observed. This is attributed to the dominant hot carrier injections as shown in Figure 3-10a. Electrons and holes are injected in the positive halves of voltage cycles while no current injection occurs in the reverse bias.

Clearly, hot carrier injection is strongly tied to the 474nm-wavelength fluorescence.

(a) (b) Figure 3-10 | (a) Time-resolved current at 50Hz. Hot carrier injection is only current contribution. (b) Time-resolved current at 60,000Hz with various voltages (17.9V to 22.0V). Secondary carriers (Jdc) and hot injection carriers (Jsine) have mutual contributions in EL spectrum and current density characteristics.

As for 60,000Hz (secondary carrier related), as seen in Figure 3-9b the EL intensity of the fluorescent emission at 474nm is nearly identical to that seen at 50Hz. But the 595nm phosphorescent peak boosts with the raising of the external electric field.

Examining current transients in Figure 3-10b, the total current density is the sum of

RMS value of sinusoidal waveform and DC offset (Jtot = Jdc + Jsine). Under the lower voltage of 17.9V, the DC current density and RMS sine current density are 60.2mA/cm2 and 36.6mA/cm2 respectively. Compared to higher voltages (18.9V, 19.8V, 20.6V,

21.5V, and 22.0V), the DC current densities are dramatically increased to 82.3mA/cm2,

114.9mA/cm2, 138.5mA/cm2, 164.1mA/cm2, and 184.6mA/cm2 while a relative constant

RMS sinusoidal current is observed at: 37.8mA/cm2, 39.4mA/cm2, 42.1mA/cm2,

43.7mA/cm2, and 45.2mA/cm2, respectively. It seems straightforward to suggest that the tremendous enhancement of secondary charge current produces the growth of 595nm- wavelength phosphorescence.

Figure 3-11 | Jrms-Vrms and L-Vrms characteristics of AC-OEL devices at 50Hz and 60,000Hz, respectively.

Plus, Jrms-L-Vrms characteristics at low frequency (50Hz) and high frequency

(60,000Hz) are shown in Figure 3-11 in which the maximum brightness, 360cd/m2 in blue and 600cd/m2 in red, are of the order necessary for devices for personal display use for instance.

Figure 3-12 | Luminance plots as the function of frequency. Current-frequency characteristics of AC-OEL devices and its effective current analysis.

The luminance-frequency characteristic of the color tunable AC-OEL device is shown in Figure 3-12. The total emission of the device shifts between fluorescent and phosphorescent contributions, but the luminance is relatively stable over the frequency.

Corresponding to the F-P shift, the frequency characteristics are also shown with current density in Figure 3-12. Analyzing the low-frequency regime (below 1,000Hz) first, we note that low frequencies lead to dominant blue fluorescence, since, as we suspected, the device under low frequency driving acts more like a carrier injection type diode in forward and reverse bias (Figure 3-13).

(a) (b) Figure 3-13 | Current and voltage transient properties of AC-OEL device at frequency of 50Hz (a) and 60,000Hz (b).

At higher frequencies (over 10,000Hz), the current density consists of a sinusoidal contribution and a DC offset (see in Figure 3-13), essentially reflecting both displacement of direct current injection and secondary charge current respectively. In

Figure 3-12, the DC offset component of the current through the device starts at a very

55 low level (13.8 mA/cm2) at 10,000Hz and then increases to 226.1 mA/cm2 at 45,000Hz.

In contrast, the RMS value of the sinusoidal component of the current waveform drops from 286.1 mA/cm2 at 10,000Hz all way down to 106.8 mA/cm2 at 45,000Hz, which suggests the significantly reduced contribution of hot carrier injection in total current at high frequency. This opposite trends illustrate that electric field above 20,000Hz applied on the capacitive device is sufficient to generate a magnetic field strong enough to yield secondary charge diffusion. The stronger AC magnetic field (due to higher frequency) suppresses ISC between singlet-state and triplet-state electron-hole pairs in the PFN-Br, resulting in population enhancement of singlet electron-hole pairs at the F-P interface.

The elevated singlet-triplet ratio promotes the generation of secondary charge carriers.

The hopping transport of secondary electrons and holes in the organic semiconductor is acutely tied to the generation of radical triplet excitons in Ir(MDQ)2(acac) leading to the red phosphorescence (as indicated in Figure 3-12). It is worth noting that the coupling of the driver to the capacitive device has been taken into consideration because dominant capacitance of device (1~3nF) compared with parallel external capacitance (530pF) in the driver. The driving frequencies over 50,000Hz cause the insufficient carrier injection resulting the decrease of electron-hole pairs as well as the sum current reduction in

Figure 3-12.

Figure 3-14 | Electroluminescence (PL) spectral shift of AC-OEL devices with driving frequency of 50Hz, 100Hz, 500Hz, 1,000Hz, 10,000Hz, 20,000Hz, 30,000Hz, 40,000Hz, 50,000Hz, and 60,000Hz.

In Figure 3-14, the EL spectrum of the AC-OEL device shows a dramatic color change when the driving field frequency varies from 50Hz, 100Hz, 500Hz, 1,000Hz,

10,000Hz, 20,000Hz, 30,000Hz, 40,000Hz, 50,000Hz, up to 60,000Hz. The 474nm emission band is dominant at low frequency of 50Hz. As driving frequency increases, the peaks at 430nm, 474nm, and 531nm (due to PFN-Br’s fluorescent emission) are weakened, in contrast, the 595nm peak (corresponding to Ir(MDQ)2(acac)’s phosphorescent emission) grows rapidly and becomes the dominant emission band. We note that all spectra are measured under 100 cd/m2 and integrated for 500ms with 5 times average. The operating pixel images in Figure 3-15a provide us a clearer picture of the color change with frequency in the forward and reverse sweeps. As indicated by the marked circles, the CIE coordinates in Figure 3-15b starts from (0.23, 0.34) at 50Hz, across the white zone then towards (0.53, 0.40) in red zone at 60,000Hz. The color change effect is also impacted by the thickness of PFN-Br as shown in Figure 3-16 and

3-17.

(a)

(b) Figure 3-15 | (a) Images of the color tuning emission in the forward and reverse frequency sweeps. (b) CIE 1931 x-y chromaticity diagram showing the color shift from (0.23, 0.34) to (0.53, 0.40) with frequency variation between 50Hz and 60,000Hz.

(a)

(b) Figure 3-16 | EL spectra shift of AC-OEL devices with 30nm (a) and 80nm (b) PFN-Br spun-cast from Me(OH) solution.

Figure 3-17 | Intensity ratio between 474nm peak and 590nm peak as a function of frequency in the PFN-Br’s concentrations of 20nm, 50nm, and 70nm.

Ion Enhanced Color Tunability

To further study the energy transition at the F-P interface, photoluminescence

(PL) spectroscopy is carried out and the results are shown in Figure 3-18a and 3-18b.

The absorption of PFN-Br (peaks at 398nm) shows a large overlap with PL spectra of the host PVK (shows a wide peak at 404nm) in Figure 3-18a. This implies an efficient

Fӧrster energy transfer route between PVK and PFN-Br (Fӧrster resonance energy transfer efficiency ~ 21.1% calculated from Figure 3-19). In the PL spectra for PVK:

3wt% Ir(MDQ)2(acac)/PFN-Br seen in Figure 3-18b, an extra 585nm peak due to

Ir(MDQ)2(acac) was detected when the sample was excited by 347nm (PVK’s strongest absorption), in contrast, 380nm excitation (PFN-Br’s strongest absorption) yields the absence of this 585nm peak. This suggests that the direct energy transfer between PFN-

Br and Ir(MDQ)2(acac) is not allowed. The Fӧrster energy transfer from PVK to PFN-Br and the forbidden transfer between PFN-Br and Ir(MDQ)2(acac) explains the blue florescence due to PFN-Br in hot carrier injection.

Figure 3-19 | PL spectroscopy of PVK: Ir(MDQ)2(acac)/PFN-Br and resonance energy transfer (FRET) efficiency.

Because of the full ionization of poly[(9,9-bis(3'-(N,N -dimethylamino)propyl)-

2,7-fluorene)-alt-2,7-(9,9-dioctylfluorene)] (PFN-DOF) by Br atoms, PFN-Br possesses a great number of moveable negative charges among main polymers. The Figure 3-18c shows the electron mobility enhancement by the diffusive Br negative ions comparing with PFN-DOF. Because of the long distance diffusion, the movable Br ions are able to populate the free secondary electrons leading to the dissociation of more excitons, thereby they facilitate the carrier diffusion to the phosphorescent emission layer. To contrast PFN-Br devices, we studied the spectral shift in PFN-DOF devices between magnetic free device (50Hz) and AC magnetic field coupled device (60,000Hz) in the same architecture. In Figure 3-18d, the fluorescence of PFN-DOF (at 423nm and

476nm) is promoted implying the promoted singlet-state excitons by suppressing ISC of

PFN-DOF. However, the color change is almost unnoticeable in PFN-DOF, which suggests the secondary carriers are unable to reach the phosphorescent sites without the aid of Br ions.

Energy Transfer Mechanism of Magnetic Dependent ISC

Figure 3-20 | Scheme for e-h pairs excited energy transfer mechanisms between PVK (host), Ir(MDQ)2(acac) (phosphorescent dopant), and PFN-Br (fluorescent material) with magnetic-suppressive ISC. (e-h)1 and (e-h)3 represent singlet and triplet intermolecular e- h pairs. S and T are singlet and triplet excitons.

Figure 3-20 shows the overall energy transfer at the PVK: Ir(MDQ)2(acac)/PFN-

Br heterojunction. There are three processes need to address:

(i) ISC of electron-hole pairs in PFN-Br is magnetic field sensitive. In the presence of magnetic field, energetically inaccessible T- and T+ states yield a redistribution of singlet and triplet excited states (S0:T0=1/2:1/2) in PFN-Br, as has been predicated theoretically106–111;

(ii) Accumulated singlet-spin electron-hole pairs are dissociated into diffusive secondary carriers with the assistance of Br ions;

(iii) The energy transfer of electron-hole pairs between PVK and PFN-Br is efficient but trivial because of insufficient singlet state excitons in PVK.

Figure 3-21 | SEM image of Fe3O4 magnetic nanoparticles (NPs) in PFN-Br polymer matrix.

Figure 3-22 | XRD pattern of Fe3O4 magnetic nanoparticles (NPs).

An instructive and elegant confirmation of magnetic field effects on F-P tunable emission, can be gained through the use of Fe3O4 NPs acting as a magnetic field alternative. The nano magnetite-induced magnetic field is extremely local, and so it will not introduce extra, unpredicted magnetic field effects as a uniform external magnetic field may do (for instance, possible magnetic field effects on the PVK. It has been reported that magnetic field gradients due to Fe3O4 nanocrystals are more effective in suppressing ISC processes than static external fields are for DC-devices.76,107 The ferromagnetic Fe3O4 magnetic NPs used here have average diameter of 30nm (TEM shown in Figure 3-21), which has been demonstrated by XRD pattern (Figure 3-22) and

64 widely accepted as a ferromagnetic material since superparamagnetic behavior only

112–114 exists in Fe3O4 magnetic NPs with diameter less than 10nm.

(a) (b) Figure 3-23 | (a) Time-integrated PL spectra measured at 437nm for pure PFN-Br thin film, PFN-Br doping 0.125wt% Fe3O4 NPs film, and PFN-Br doping 0.375wt% Fe3O4 NPs film after constant excitation at 280nm (10Hz, 500fs, 130 μJ/cm2). (b) Time-resolved PL decay transients at 410nm~500nm integration.

The magnetic enhanced singlet-state excitons in PFN-Br were directly observed in time-integrated PL spectra in Figure 3-23a by 1.6 fold while 0.125wt% Fe3O4 NPs was doped in PFN-Br. A slight PL intensity is increased in 0.375wt% Fe3O4 NPs doping

PFN-Br, indicating the 50% theoretical up limit of singlet excitons. Worth to note a constant excitation was used in experiment: λex. = 280 nm. In Figure 3-23b, we show the picosecond-transient PL intensity of PVK: Ir(MDQ)2(acac)/ PFN-Br (τ1=319.7ps, where

τ0 is the time taken for the PL to fall to 1/e of initial intensity) and PFN-Br (τ2=171.7ps) for 410nm~500nm emission (corresponding to PFN-Br’s fluorescence). We assign the PL radiative delay to short exciton diffusion in PFN-Br, which has trivial impact on

Ir(MDQ)2(acac)’s phosphorescence. Furthermore, the relative reduced PL intensity of

Ir(MDQ)2(acac)’s phosphorescent emission (580nm - 600nm) provides the evidence that no diffusive exciton involves color tunbility even though singlet-state exictons are saturated with doping magnetic NPs in PFN-Br. Therefore, the possibility of diffusive exciton across interface is excluded, reversely the Br ion-enhanced secondary carriers is demonstrated. More than that, we analyzed the streak image for the sample of PVK:3 wt% Ir(MDQ)2(acac)/PFN-Br in Figure 3-24, no emission is observed except for an intense emission band (350nm~425nm) which corresponds to PVK’s fluorescence.

However, PFN-Br’s fluorescent emission (410nm~500nm) becomes observable when

0.125wt% Fe3O4 or 0.375wt% Fe3O4 are blended into PFN-Br. This demonstrates that massive numbers of radical singlet-spin pairs are formed in PFN-Br as the influence of the magnetic field on F-P emitting unit. Note that the Ir(MDQ)2(acac)’s phosphorescent emission cannot be captured in the first several nanoseconds in both cases because of long-live triplet excitons, but appears at 600nm in time integrated spectra.

(a) (b)

(c) (d) Figure 3-24 | (a) Time-integrated PL spectrum (>1us) for glass/ PVK:3wt% Ir(MDQ)2(acac)/PFN-Br, glass/PVK:3wt% Ir(MDQ)2(acac)/PFN-Br:0.125wt%Fe3O4, and glass /PVK:3wt% Ir(MDQ)2(acac) /PFN-Br:0.375wt%Fe3O4. The streak images for glass/ PVK:3wt% Ir(MDQ)2(acac)/PFN-Br (b), glass/PVK:3wt% Ir(MDQ)2(acac)/PFN- Br:0.125wt%Fe3O4 (c), and glass /PVK:3wt% Ir(MDQ)2(acac) /PFN-Br:0.375wt%Fe3O4 (d) in 6ns.

To confirm the importance of the magnetic effect in F-P emitting unit, we measured the EL spectrum of analogous devices in Figure 3-25 with zero magnetic field coupling (DC driving), internal AC magnetic field (60,000Hz driving) and external NP magnetic field (DC driving). A large F-P intensity ratio (If/Ip) of 5.5 was observed when the magnetic field was eliminated. In contrast, the F-P intensity ratios are 0.65 and 0.93 in AC magnetic field and Fe3O4 NC magnetic field, respectively. This implies the low F-

P intensity ratio is a sign that secondary charges, originating from singlet electron-hole pair accumulation, are able to facilitate the generation of triplet excitons in the

Ir(MDQ)2(acac) molecules.

Figure 3-25 | EL spectrum with no magnetic field, internal AC magnetic field, and external magnetic field due to Fe3O4 NPs.

Chapter Summary

In summary, the application of internally generated, AC magnetic fields in a unique AC-OEL architecture is demonstrated as an efficient method to manipulate the ratio of singlet and triplet spin pairs. The magnetically sensitive ISC can modify the probability of singlet-spin electron-hole pairs in organics (PFN-Br), which significantly populates diffusive secondary charges to achieve phosphorescence emission

(Ir(MDQ)2(acac)). By varying AC magnetic fields coupled at an F-P heterointerface in the AC-OEL device, the emission can be shifted between blue fluorescence (singlet-spin related) and red phosphorescence (triplet-spin related) while no significant brightness

68 changes. Conventional contributions for color shifts associated with spatial shift of recombination zones or exciton redistribution and quenching under high voltage have been excluded by our experiments. While the unusual properties of magnetically-coupled

AC-OEL devices may lead to other breakthrough devices, our work opens new doors for a more detailed understanding of radical pairs manipulation in the quantum electrodynamics and spintronics of organic materials.

Chapter 4 Solution Possessing Organic Small Molecular Emitter

With the significant advantages of low-cost manufacturing and large device area, field induced AC-OEL based on solution processed polymeric emitters have been intensively stuided over the past few years.52,57,59,97,115–117 However, the choice of polymeric emitter for full color devices is restricted when compared to organic small molecules emitters. The thermal evaporation of organic small molecules has critical drawbacks as a fabrication technique such as high vacuum manufacturing cost and being time consuming. In this chapter, state-of-the-art solution processed small-molecule films are fabricated and compared to their vacuum-deposited counterpart. The proper solvent for tris(8-hydroxyquinolinato)aluminium (Alq3) has been studied, which has an adequate solubility but also avoids the interfacial mixing between top emission layer and bottom

HGL. Comprehensive non-gated and ZnO-gated AC-OEL devices based on Alq3 emitter were fabricated and the underlying mechanism of hole carrier manipulation at ZnO/HGL interfaces was investigated in time-resolved current and voltage characteristics.

Solution Based AC-OEL Devices

The key-advantage of field induced inorganic electroluminescent devices is an easy solution-fabrication process, a dramatic manufacturing cost saver which meant that they were rapidly commercialized after this was first reported by Toshio Inoguchi in

1974.118 The solution processing of these devices is effectively used to fabricate large- area lighting panels or displays with good robustness and long-term reliability. Recently, instead of inorganic phosphors, organic semiconducting polymers have been given these traditional advantages through a variety of solution-process thin film fabrication methods, such as spin-coating38, blade coating119, spraying120, ink-jet printing121 etc. In contrast

70 with polymers, organic small molecule materials provide more advantages, particularly high stability and a wide selection of emitter materials which span the entire visible spectrum.122,123 The major drawback of organic small molecule thin film devices is the considerably strict manufacturing requirement on high vacuum evaporation.50 Therefore, it is necessary to develop an AC-OEL device based on soluble organic small molecule emitters.

Due to the insulators sandwiching a p-i-n structure, the operating voltage in AC-

OEL has to be very high (over 100V in the most cases)53,124 this is because of the inefficient carrier transportation in AC cycles which causes a great number of hot carriers that need to be neutralized at insulator/HGL interfaces. This in turn leads to irradiative recombination of electron-hole pairs. In this respect, the use of a n-type, wide bandgap semiconducting gate has been demonstrated as a good strategy to manipulate effective carriers and to lower driving voltage97,98, which solves the device’s overheating issue.

Here, Alq3 was chosen as the emitting host material, as it is the most common

37,125 used emitter in vacuum-deposition. The soluble properties of Alq3 in different organic solvents were studied and the morphological properties of the films using solution process were compared to the evaporated film. The non-gated and ZnO-gated

AC-OEL devices were also fabricated with solution processed Alq3 and the electrical and optical responses over driving frequency were further investigated. The analysis of the performance of the AC-OEL devices near DC and at the resonant frequency revealed that carriers are generated under an AC electric field, and at the same time, electrons and holes are efficiently manipulated at the ZnO gate/HGL interface.

Solubility of Alq3 in Different Solvents

Figure 4-1 | (a) Chemical structure of Alq3. (b) The solubility of Alq3 (10 mg mL−1) in dimethylformamide (DMF), chloroform, and acetone.

The choice of solvent is an initial but critical issue for solution processing technique, because a considerable amount of organic small molecules need to be able to be dissolved into it. Figure 4-1a shows the molecular structure of Alq3, the common solvents reported in literature for which are chloroform 126,127 and acetone128, which have a high polarity index of 4.1 and 5.1 respectively. However, both of these options have some inevitable technical drawbacks. Chloroform has quite good solubility, not just for

Alq3 but for a variety of organic compounds including polymers and small molecules, which could cause interface damage during spin-coating. This would lead to inferior film quality and a shorted current during device’s operation. With regards to acetone, the solubility of Alq3 in it is quite low (as 10mg/mL shown in Figure 4-1b) to the point that it is unable to satisfy the thickness of spun-cast film where sufficient radiative electron- hole recombination occurs129.

(a) (b) Figure 4-2 | (a) Current density characteristics in a function of RMS voltage at 50 Hz. Inset: device structure of the AC-OEL devices. (b) The relation between RMS voltage and brightness at 50 Hz. Inset: the power efficiency.

So far, dimethylformamide (DMF) has not been fully reported as a solvent for

Alq3 in solution-processed organic lighting devices. Due to its high vapor point (~153℃), the DMF spun-cast thin film shows a few hundred nanometer thickness with smooth morphology. In addition, there are a number of hole transport organics that show a resistance to dissolving in DMF, which prevents the dual films mixing. Taking these into consideration, AC-OEL devices were built with solution-processed Alq3 emitters from chloroform, acetone, and DMF. These devices were driven under a sinusoidal voltage with frequency between 50Hz and 60,000Hz and their structure is shown in the inset of

Figure 4-2a. In order to provide sufficient hole carriers, poly[N,N’-bis(4-butylphenyl)-

N,N’-bis(phenyl)-benzidine] (poly-TPD) was pre-coated as a HGL on top of gate layer prior to coating with Alq3. Missing a HGL layer normally gives an imbalanced electron- hole injection into emitter, resulting in the missing of an efficient recombination zone.53

The current characteristics of AC-OEL devices with Alq3 emitters, either from chloroform, acetone, or DMF are compared in Figure 4-2a. The DMF device shows the

73 best carrier injection at 50Hz, while a reduced current density was found in the acetone device. The current density of the chloroform device is consistently between the DMF and acetone device. It is difficult to draw any convincing conclusion based on current characteristics alone. Therefore, the corresponding plot of luminance versus applied voltage (L-V) is presented in Figure 4-2b. All devices show a similar turn-on voltage at around 4V. However, the brightness boosts to 2,080 cd/m2 at 8.5 V in the DMF device, whereas, a maximum luminance of only 71 cd/m2 is obtained in the chloroform device at

6.7V. Combined with the tremendous current in Figure 4-2a, we attribute that to the shortening of the device due to washed-off poly-TPD by chloroform. The exposure of poly(3,4-ethylenedioxythiophene):polystyrene sulfonate (PEDOT:PSS) causes direct contact with Alq3 so that electrons can easily break through the entire device through the

LUMO levels. On the other hand, the acetone-based device lights up to 370 cd/m2 with non-trivial current injection (238.6 mA/cm2). Considering the thin emission layer from dilute Alq3-acetone solution, it seems reasonable to suggest that electron and hole carriers are simply being transported from one electrode to another in the lack of presence of an efficient recombination interface. More importantly, it is worth comparing the Alq3 solution-processed device performance with the one with evaporated Alq3 emitter. The

Alq3 evaporated device shows a lower turn-on voltage and a faster increase in L-V plot, exhibiting a maximum power efficiency of 0.29lm/W which is higher than that in Alq3 solution-processed devices (0.22 lm/W in Acetone, 0.03 lm/W in Chloroform, 0.23 lm/W in DMF). This attributed to the trivial defects on evaporated Alq3 thin film. Only the

DMF based device can compete against the evaporated device since its maximum efficiency shows at the high brightness.

Morpholgy of Spun-cast Alq3 Thin Film and Wash-out Effect

Figure 4-3 | Surface morphology of spun-cast Alq3 films from a variety of organic solvents. The optical microscopy (×1000 magnification) of thermally (a) evaporated Alq3 film and (b) spun-cast Alq3 films from DMF and (c) chloroform. AFM topography of thermally (d) evaporated Alq3, (e) DMF spun-cast Alq3 film, and (f) chloroform spun- cast Alq3 film. The scale bar is 1 µm. (g–i) The corresponding 3D AFM images.

To prove the distinctive influence of solvent choice on solution-processed Alq3 film, the morphology of the solution-processed films were looked at on the top of poly-

TPD, and are compared to its vacuum-deposited counterpart. First, the optical microscopic images of Alq3 from thermal evaporation (Figure 4-3a), DMF spin coating

(Figure 4-3b), and chloroform spin coating (Figure 4-3c) were compared. Under × 1000

75 magnification, both the evaporated and DMF spin coated Alq3 films exhibited a good film quality over a large area. The roughness values (Ra) of the Alq3 layers from evaporated and spin-casted obtained DMF are 0.14nm and 0.19nm, respectively, given by the AFM surface morphology shown in Figure 4-3d and 4-3e. As for chloroform spin coated Alq3, the film surface is full of eye detectable defects which led to a high Ra of

~3.98nm (shown in Figure 4-3f). The chloroform spun-cast Alq3 film indicates a polymeric morphology which noticeably contrasts with small molecules, but is quite similar to PEDOT:PSS (shown in Figure 4-4). The corresponding 3D AFM images are shown in Figure 4-3g-i for better observation. Also, the corresponding XPS results (in

Figure 4-5) shows the atomic concentration increases in the peak of C 1s and the decreases in the peaks of N 1s and O 1s which indicates the wash-off effect on Poly-TPD.

Figure 4-4 | AFM image (5μm × 5 μm) of PEDOT:PSS.

(a)

(b)

(c) Figure 4-5 | XPS patterns of Poly-TPD/Alq3 from DMF(a), Poly-TPD/Alq3 from Chloroform (b), and Poly-TPD/Alq3 deposited by evaporation (c).

Resonant Effect and Carrier Gate Mechanism in Alq3 Device

In our previous work, it has been demonstrated that ZnO is an ideal gate material to replace traditional dielectric layers in AC-OEL devices for high brightness, and high power efficient lighting.97 With such a gated structure, carrier manipulation at the interface can lead to high brightness and efficiency in an AC-OEL. We demonstrated that the role of the gate is to promote carrier transport in the forward bias and block carriers in the reverse bias, thereby using part of the power cycle as a charge accumulation phase creating light, and part of the power cycle as a charge compensation phase to retain charge balance in the device. The theoretical mechanism of carrier manipulation by the gate was tested with hole-only devices in the absence and present of ZnO gate under AC

77 power cycles and comparisons of the performance of AC-OEL devices with ZnO gate,

P(VDF-TrFE-CFE), and with no gate are made as a function of frequency response and electric field.

A 70nm-ZnO-gated AC-OEL devices having solution-processed Alq3 emitters

(from DMF) were also fabricated and studied in this report with time-resolved current and voltage characteristics which has not been done yet. Figure 4-6a shows the frequency response to current density with the ZnO gate layer and the PEDOT:PSS hole injection layer (HIL). Compared to the non-gated device with HIL, the ZnO-gated device presents a dramatic current cut-off at the low frequency range (< 10,000Hz). At a constant voltage, the current density is as low as 10.0 mA/cm2 at 1,000Hz with the ZnO gate, while conversely, the PEDOT:PSS device has169.7 mA/cm2 at 1,000Hz. The current density in ZnO-gated device is not maintained at low levels over a frequency range, but promotes rapidly to the peak (169.3 mA/cm2) at the resonant frequency of

50,000Hz.

Figure 4-7 | The resonant behavior of ZnO-gated AC driven OEL devices over the frequency range 50Hz to 75,000Hz.

This resonant behavior is strongly related to the enhancement of polarization current due to the time dependent electric field. Because of strong carrier injection, the current density of a non-gated device consistently stays at a high level (over 100mA/cm2 on average). However, the resonant behavior remains in the absence of ZnO layers, it shows up at a lower frequency of 30,000Hz. This remaining resonant effect again can be explained by the polarized nature of the hole generation material. A similar resonant behavior is also discovered in a plot of EL characteristics shown in Figure 4-6b. Besides

79 a slight resonant frequency shift, the cut-off behavior at low frequencies still exists in the

ZnO-gated device.

The functional ZnO gate was initially designed to serve as carrier balance in resonant frequency driven OEL devices. Therefore, it is worth investigating the resonant behavior and DC behavior in both a non-gated device and a ZnO-gated device. Figure 4-

8a shows the L-V graph of a non-gated device at 50Hz (near DC) and 30,000Hz

(resonant frequency as found from Figure 4-6a). Both L-V curves are similar in shape, but the 30,000Hz curve increases a little bit later than the 50Hz one. This is mainly because at high frequencies the injection carriers are more easily drained in the reversed bias before the electron-hole recombination occurs, resulting in a reduction of the effective current.

However, the carrier balance has been totally altered by the ZnO gate layer. As shown in Figure 4-8b, the device (50Hz driven) is turned on at a relatively high voltage of 25V caused by a high hole injection barrier. This energy barrier (∆E) can reach 2.7eV in ZnO-gated devices (Figure 4-9b) which is significant in comparison to 0.3eV in non- gated device (Figure 4-9a). Also, the 50Hz time-resolved voltage and current characteristics are shown in Figure 4-8c. Unlike normal lighting diodes (non-gated devices shown in Figure 4-8e), the current in the reversed bias is not trivial but a noticeable negative value (about -0.13mA), which results from the hole-electron neutralization in the p-n junction at ZnO/poly-TPD interfaces. This only happens when a great number of carriers are pushed toward the junction under an electrical field until a thin enough depletion zone forms for a reduced electrical resistance.

Figure 4-9 | Energy diagram of AC driven OEL devices (a) with hole injection layer (PEDOT:PSS) and (b) with n-type gate layer (ZnO).

At the resonant frequency of 50,000Hz, luminance increases earlier and more rapidly than at 50Hz in the L-V curve in Figure 4-8b. This is mainly attributed to the promotion of polarization carriers and electron extraction of the ZnO gate under high frequency AC cycles. More specifically, electrons and holes are generated in polarized

HGL (poly-TPD), and they occupy the LUMO and HOMO levels respectively. When the bias is applied in the forward direction, the electrons are extracted via n-type ZnO towards the ITO electrode and the holes are directly injected into the Alq3 emitter. Then, when the electric field flips over, the holes are drifted and accumulate at the ZnO/poly-

TPD interface because of the forbidden transfer shown in Figure 4-9. Each electrical cycle was approximately 20 μs in period which inhibits the over-saturation of carriers.

Hence, together with energy level alignment, it was predicted that a dominant current would be present in ZnO-gated AC-OEL at the resonant frequency, which has been demonstrated by the current characteristic inserted in Figure 4-8b. In contrast to the trivial current density below 20V at 50Hz, a linear current increase from 16.2 mA/cm2 at

2.4V to 146.2 mA/cm2 at 21.6V can be seen at 50,000Hz. Meanwhile, another piece of evidence was observed in the time-resolved current and voltage properties at 50,000Hz shown in Figure 4-8d. It reveals two axisymmetric current waveforms (yellow and purple) in the top and bottom of the voltage cycles. Contrasting the resonant-frequency- resolved current (at 30,00Hz) of the non-gated device in Figure 4-8f, both halves of the current of gated device remain positive in AC power, which suggests that the current flow through the device is consistently in the forward direction, resulting from hole blocking effect at the ZnO/poly-TPD interfaces. Thus, it is clear to conclude that functional ZnO gates facilitate hole injection in forward bias as well as hole blocking in reversed bias.

Cavity effects may have an influence on the emission of thin-film lighting devices. A preliminary comparison of the AC-OEL’s spectra (in Figure 4-10) shows equal emissive profiles, indicating a negligible impact of cavity effects on the device’s optical performance.

(a) (b)

Figure 4-10 | EL spectrum of ZnO-gated AC-OEL device with Alq3 emitter doping 1wt% C545T (a) and corresponding CIE diagram (b).

Methods and Characterizations

Device measurement

The AC-OEL devices were measured in a nitrogen filled glove box at room temperature (25 °C) without sealing. A 200 MHz function/arbitrary waveform generator

(Agilent 33220A) connected to a Trek PZD700A M/S amplifier provided a sinusoidal signal with suitable voltage and frequency. The actual applied voltage and current were measured from a power analyzer (Zimmer LMG95). An ILT 1400-A photometer

(International Light Technologies) is used to measure the out-coupling luminance. EL spectra were collected using an ILT 950 spectroradiometer (International Light

Technologies). The entire system was connected and controlled by a computer. In order to obtain accurate and reliable measurements of luminance and current, each turn-on measurement of the pixels was integrated over 2000 ms and averaged five times.

Surface Morphology

An optical microscope (Olympus BX60M) and an AFM (JEOL JSPM-5200) were used to measure the morphology of spun-cast and evaporated Alq3 films.

XPS Measurement

An XPS (AXIS ULtrabld) was performed on the Alq3 thin films from spun-cast and evaporation. To maintain the accuracy, each XPS pattern was averaged from 15 times measurements.

Chapter 5 Flexible, Stable, White Electroluminescent Devices Under

High Frequency

In this chapter, the application of an aluminum (Al)/multiwall carbon nanotube

(MWCNT)/Al, multi-layered electrode to flexible, high-efficiency,(AC-OEL is discussed. The electrode was fabricated by sandwiching a spray-cast nano-network film of MWCNTs between two evaporated layers of Al. The resulting composite film facilitates a uniform charge distribution across a robust crack-free electrode under various bending angles. We demonstrate that these composite electrodes stabilize the power efficiency of flexible devices for bending angles up to 120°, with AC-OEL device power efficiencies of approximately 22 lm/W at luminances of ~ 4000 cd/m2 (using no output coupling). Microscopic examination of the Al/MWCNTs/Al electrode after bending of up to 1300 cycles, suggests the nanotubes significantly enhance the mechanical properties of the thin Al layers while providing a moderate modification to the work function of the metal. While the realization of robust, high brightness and high efficiency AC-OEL devices is potentially important in their future lighting applications, we anticipate this to also have significant impact in standard OLEDs lighting applications.

Meanwhile, a novel flexible polymeric gate layer for AC-OEL devices is presented. The charge-valve mechanism of carrier manipulation in forward and reversed bias of the applied AC cycles is fully detailed. With a low-doping strategy of two phosphors in the fluorescent host, a flexible PET-based AC-OEL device with PFN-Br gate exhibits incredible stability on EL performance with a superior color rendering index

(CRI), over 81 at 2800K color temperature and a power efficiency of 2.8 lm/W at 1,000 cd/m2 with high bending ability (-90° to +90°).

Challenges for Achieving Flexibility

AC-OEL devices have received tremendous interest lately by the organic lighting community.50,52,56–59 The devices which use an AC field to create polarization currents in an electroluminescent organic layer and thus generate light emission 53,55,130,131 have shown promise as high-quality panel lighting due to their property of simultaneously increasing brightness and power efficiency. 54,65,124 Furthermore, they may present potential advantages in driver circuit simplicity when interfacing to standard AC power sources. However, as with all ultra-thin or organic lighting approaches, a widely sought property of the AC-OEL is flexibility.

The concept of low-cost, flexible, high-efficiency light sources based on organic solid-state lighting devices is already well establish.132–137 Indeed, both standard OLEDs as well as AC-OELs have demonstrated brightness and efficiency numbers of commercial interest. However, the slow transition from state-of-art laboratory research to a modern manufacturing environment should somehow indicate the magnitude of the challenges faced by the technologies to meet with robust utilization in the consumer world.138 Two key problems that developers are facing are addressed in this work:

The first is how to achieve extremely high-power efficiencies in lighting devices when internal quantum efficiencies are already reaching levels approaching 100% (both the spin-symmetric and anti-symmetric molecular excitations are fully used for photon emission16,17,40,139,140). In any organic EL device, there are many energy loss mechanisms from carrier transport to multiple non-radiative processes that convert singlet and triplet

87 excitations into the ground state.16,17,40,139 Among these mechanisms, an important problem in OLEDs is direct injection of hot carriers, which form bound electron-hole pairs or excitons after tunneling through energy barriers at multiple interfaces.141–144

These hot carriers exceed the energy gap of the emitting material, so the excess energy is wasted thermally, giving rise to a lower power efficiency. The reduction of the carrier injection barrier by decreasing the difference between adjacent energy levels (HOMO or

LUMO) is now a common method to optimize the efficiency of OLEDs.46,47,145–147

However, as in the case with most technology demonstrators, what works best in the lab is difficult to implement in commercial settings, so placing Li or Ca at the metal interface is undesirable.

In the case of capacitively coupled AC-OEL devices, as described here, the large induced polarization currents effectively control the rate and energy of carrier “injection”.

While power efficiencies as high as 29.3 lm/W at 20500 cd/m2 have been reported in green phosphorescent AC-OEL devices, it has been anticipated that higher values are possible using asymmetric contact designs that allow for better charge compensation in the devices.52–54 But it must be recognized that facile charge compensation in such asymmetric devices also requires engineering of the work function at the metal contact.

And, again, we wish to avoid the use of typical laboratory solutions to this problem such as Li or Ca.

The second problem we would like to address is what the ideal flexible, reflective electrode for organic devices might be. A great deal of research has been devoted to the development of highly flexible electrodes with high transparency including metal nanowire grids148–153, graphene sheets154–158, and carbon nanotube (CNT) networks.159–162

However, development of flexible cathode materials (the back reflector) with excellent optical, electrical, and mechanical properties is rarely reported. Generally, flexible optoelectronics, such as OLEDs and organic solar cells (OPVs), utilize surprisingly thin metal layers as the flexible back cathode. Metals that have been most widely studied for this use include Al132,133, calcium (Ca)/Al163, nickel-chromium (Ni-Cr)44 and sliver

(Ag).164 Interestingly, the balance between flexibility and conductivity is difficult to achieve. Metal electrodes can have residual stress due to work hardening and become rigid or brittle, leading unavoidably to cracking after repeated bending or stretching.

Voids, gaps and uneven thin film morphologies that result can then lead to non-uniform current distributions in the electrodes: essential for performance in both AC and DC EL devices. In the laboratory setting one would choose the metal with most desirable mechanical properties then modify its work function using the methods described above.

However, again, in a commercial sense this may not be possible.

To address the two challenges above, we examine the application of novel composite electrode structures, composed of metal/nanowires (nanotubes)/metal. We demonstrate that as a reliable alternative structure for the cathode, this triple-layer composite electrode gives an excellent balance between mechanical stability, electrical conductivity, and optical functionality even after thousands of bending cycles. We further demonstrate these Al (50nm)/ MWCNTs/Al (100nm) cathodes on a flexible AC-OEL device structure which achieves and maintains a power efficiency of approximately 22.2 lm/W at 4050 cd/m2 after full bending cycles of 120° for 10 minutes. However, we emphasize the generality of this approach and their potential use in OLEDs as well.

Construction of Metal/Nanowires Composite Electrode

(a)

(b) Figure 5-1 | AC-driven organic EL device on flexible 1 inch × 1 inch plastic substrate. (a) PET/ITO(100nm)/PEDOT:PSS doped ZnO NPs (80nm)/PVK:Ir(ppy)3 (150nm)/TPBi (35nm)/Triplet-layer cathode (~200nm). Electric fields due to AC driver in the capacitor- based device are changing in high frequency rather than pointing a unique direction. (b) The definition of the bending angle used in this report.

Figure 5-1a shows a diagram of the capacitive flexible AC-driven OEL device with a triple-layer composite cathode. The fabrication of the device starts with a clean polyethylene terephthalate (PET) substrate pre-coated with 100 nm of ITO. Commercial

PEDOT:PSS doped with 36 wt.% ZnO NPs (35 nm average diameter) is spun cast on

ITO as not only a hole generation layer, but also as an inorganic semiconductor gate, which is required for a field-induced EL device. A more complete discussion and analysis

90 of the function of insulating or semiconducting gate layers in AC-OEL devices can be

52–54 found in our previous work. In this work PVK : Ir(ppy)3) is used as the electroluminescent layer with Ir(pp)3 as the triplet scavenging phosphorescent dye. TPBi, is an electron-rich (electronegative) dopant and transport layer used to help in charge generation at the interface to the emitter. The PEDOT serves an analogous function for hole injection. The multi-layer composite electrode (metal/MWCNTs/metal) is deposited on the functional organic layers in a three step process before the devices are tested. This procedure is shown in Figure 5-2 and is carried out as follows: a) A 50nm layer of Al is pre-deposited through a shadow mask over the TPBi layer with a rate of 2 Å/s under high vacuum (~2 × 10-7 Torr). b) A spray gun is used to deposit the MWCNT nano-matrix at different volumes of 0.5 mL, 1.0 mL and 1.5 mL on the substrates. The MWCNTs are dispersed in acetone at a concentration of 0.2 mg/mL through ultrasonication. c) After post annealing in air for 10min at 60° to evaporate excess solvent, the PET substrate is placed back into high vacuum for thermal evaporation of the final 100nm of

Al.

SEM images in Figure 5-2 show the three deposition steps in the fabrication and surface morphology of the Al/MWCNTs/Al composite cathode. The Al/MWCNTs/Al sandwich structure is asymmetric in that the bottom Al layer is flat and the top Al layer firmly coats the protruding MWCNTs network, which leaves a rough surface.

The property that we seek in this electrode is outstanding durability under multiple bending cycles. Generally, we would expect that the interlaced bundles of

MWCNTs would exhibit high structural integrity and mechanical robustness as has already been reported in numerous applications.165–167 We would also anticipate that mobile dislocations within the thin metal films will be pinned by the nanometer inclusions, thereby limiting failure due to work hardening of the film.

Electrically, it is well-known that MWCNT mats can exhibit a rather high electrical conductivity at room temperature, but generally not as high as a metal like Al.

This limits the application of the Al/MWCNTs/Al structure in DC-driven OLEDs since carrier injection efficiency is greatly affected by electrode resistance. MWCNTs also 92 have a relatively high work function, between 4.40 and 4.95eV, as compared to LiF/Al

(2.6 eV), Mg (3.72eV), Mg/Ag (4.12 eV) and Al (4.28eV)).168 169,170 This suggests pure nanotubes would be unsuitable to replace the metal as a top cathode in flexible lighting devices since there would exist a large electron injection barrier at the metal- semiconductor interface and the overall conductivity across the electrode would be inadequate. We add here that while the asymmetric AC-OEL devices with their unique capacitive nature, utilize a field-induced polarization current which contributes significantly to the light output, they also have some direct injection. Thus, even though lower carrier injection efficiency has less influence on the AC-OEL device’s performance, it does have some. Furthermore, MWCNT mats’ porous and low dimensional carbon structure produces a very low reflectivity surface for visible light

(wavelength ~ 400 nm -700 nm). Therefore, clearly we want the mechanical properties of MWCNTs but not the electronic properties for this application.

However, a composite of nanotubes and a metal such as Al should result in a blending of the work functions of the two materials, while providing a surface with the reflectivity of the metal (for relatively small amounts of nanotubes) and finally the mechanical properties we seek, as described above. Unlike symmetric field-induced EL devices with two insulators or asymmetric El devices with a single insulator, the devices used in our studies are an asymmetric structure with a ZnO (semiconductor) gate. A unique feature of these asymmetric AC-OEL devices is that the contribution of injected charges has a nontrivial influence on the formation of excitons in the light emission layer, and thus the work function is of importance.

In order to avoid high temperature processing of a composite structure, our composite is formed by layering. This means the Al is exposed to air between the application of the MWNTs and the second layer, leading to a very thin alumina film.

However we expect that the mechanical properties to be dominated by the metal and the

MWNTs. Figure 5-2 show the formation of the composite layers using field emission

SEM. We note here that the Al covered mat is relatively rough, but covered with Al, leaving behind a reflective film.

The AC-OEL device’s flexibility was determined by measuring the performance parameters of brightness and lumens per watt (LPW) as a function of the number of bends at a specified bend angle as defined in Figure 5-1b above. Resistivity of the triple layer contact as well as microscopic analysis of the devices was carried out and correlated with the performance at specific bend conditions.

Table 2 | Summary of the performance of flexible organic AC-EL devices with different cathode structures.

Turn-on Max. Max. Power Optimum Cathode Structure RMS Brightness Efficiency Frequency Voltage 2 (V) (cd/m ) (lm/W) (Hz) Al(150nm) 5.76 4070 21.8 50k Al(50nm)/0.5mLMWCNTs/Al(100nm) 6.63 4050 22.2 60k Al(50nm)/1.0mLMWCNTs/Al(100nm) 10.71 3376 16.1 55k Al(50nm)/1.5mLMWCNTs/Al(100nm) 11.00 1982 7.3 50k

We first examine the effects of the triple-layer cathode on un-bent geometries and these data are tabulated in Table 2. 0.5mL, 1.0mL and 1.5mL volumes of 0.2 wt. %

MWCNTs dispersions are used to create triple layer cathodes with different nanotube contents for comparison. Notice that the optimum operational frequency is where the AC

94 driver is in resonance with the capacitive device for the maximum reactive power. This occurs at similar frequencies in all cases suggesting that the capacitance is roughly the same from device to device as we would expect. However, we note that the brightness decreases slightly with nanotubes in the metal layer while the turn-on voltage increases.

This can be understood in terms of the relatively poor conductivity of MWCNTs (0.5~1 kΩ/□)171 compared to Al (~0.3 Ω/□). Quite simply, as more of the Al volume is replaced by the poorer conducting nanotubes, the overall conductivity of the system begins to drop proportionally and there is a potential drop now across the electrode. Consequently the voltage drop seen across the emitting layer has gone down for a given voltage applied to the device. Electrical resistivity characteristics of Al electrode and Al/MWCNTs/Al are given in Figure 5-3.

Figure 5-3 | The electrical resistivity of Al electrode and Al/MWCNTs/Al before and after bending.

We expect three other factors have non-negligible effects in raising the turn-on voltage while the light emission is reduced. The first of these is obviously due to the effects the increasing content of the nanotubes have on the reflectivity of the back cathode. The MWNTs are within 50 nm (less than the skin depth for Al) of the interface.

So, while scattering generally helps output coupling in such devices, in this case the addition of a black absorber will surely reduce the output.

The second factor is the effect of increasing the work function of the cathode. In the AC-OEL charge neutrality must be maintained throughout the power cycle. In the

96 asymmetric configuration used here, this charge balance is established using makeup charge from the cathode – thereby requiring a low barrier to charge transport. If the injection barrier is raised, the compensating charge is hindered.

Finally, when the device is bent, there occurs a change in device capacitance.

Since these devices operate at a resonance between the driver and the lamp capacitor, the frequency at which they should be driven for optimal light output changes as the capacitance changes. Bending can change the capacitance slightly and therefore move the device off of resonance. We expect this to yield a decrease in overall brightness since the device is being driven off the resonance. However, in this last case, we expect the capacitance to recover once the device has relaxed back (recovered) from the bending.

Bending Test on Nano-Network Composite Cathode

To minimize the negative effects of resistance in the MWCNT component of the composite, while examining their potential to enhance the robustness of device flexibility, AC-OEL devices with 0.5mL of 0.2 wt. % MWCNTs are chosen as our basis for comparison during bending. Clearly, from Table 2 AC-OEL devices with and without

MWCNTs at this loading level in the cathode exhibited brightness’s of the same order

2 2 (Lmax_flat_Al=3330 cd/m and Lmax_flat_Al/MWCNTs/Al=2700 cd/m . In Figure 5-4, we present the brightness data for Al only cathodes (Figure 5-4a) and compare to brightness for MWCNT cathodes (Figure 5-4b) for all voltages tested. When tested initially in an unbent state, the flat nanotube and flat Al only devices (open squares) are very nearly the same as would be expected, except for the slight offset associated with resistance in the cathode. Notice too, in Figure 5-4c and 5-4d, the current voltage characteristic associated

97 with the flat devices without and with nanotubes respectively, are linear and differ only by the resistance of the cathode.

(a) (b)

(c) (d) Figure 5-4 | Performance of the green phosphorescent flexible organic AC-EL devices with and without MWCNTs driven at a frequency of 50kHz during bending (flat, 60°, 90°, 120°, recover). Luminance with Al (a) or Al/MWCNT/Al (b) as function of voltage for flexible organic AC-EL devices. Red arrows represent the luminance peak shifts between the maximum luminance of the Al-based device and that of the Al/MWCNT/Al- based device. (c) and (d) show RMS current density vs RMS voltage characteristics of flexible organic AC-OEL devices before and after bending.

A slight enhancement of the luminance is observed in both cases as the devices are bent from 0° to 120°. This is measuring the brightness at the center of curvature of the bend. The capacitance change of flexible AC-OEL devices with shape change of substrates should lead to a decrease in luminance as stated in our expectations above.

Therefore, we attribute this slight increase to increased conductivity of the compressed electrodes induced by external mechanical deformation of the film and enhanced field density in this area. It should also be recognized that the output coupling of the light from the surface of the curved substrate during bending could be quite different from that of the flat state device, thereby leading to an observed increase in light output.

(a) (b) Figure 5-5 | Metallurgical microscopy images of top electrodes in flexible organic AC-EL devices before (insets) and after a full cycle of 120° bending. (a) Al (150nm) electrode shows 20μm width creases due to bending. (b) Al (50nm)/MWCNT/Al (100nm) composite electrode shows outstanding mechanical property in bending.

When the bent Al-based device is released, its brightness-voltage and current- voltage curves recover to their original shape, but the maximum luminance shifts from

4054 cd/m2 at 12.7 V at 90° to 3328 cd/m2 at 16.3 V in the recovered state. The lower optimal luminance and higher voltage required to achieve optimal luminance is consistent

99 with the view that the cathode has formed cracks and deformations that have reduced its overall conductivity – thereby reducing the total voltage dropped across the emitter. This morphology change is confirmed in optical microscopy under 200× magnification shown in Figure 5-5. The formation of these cracks and creases in Figure 5-5a is an irreversible process, and thus poor conductivity results when the device is flattened after severe bends.

In contrast, consistent EL performance of MWCNT-based devices is observed with bending and recovery. A small luminance peak shift from 4002 cd/m2 at 13.1 V at

90° to 4060 cd/m2 at 13.9 V in the recovered state, as shown in Figure 5-4b, is seen, but peak shape as well as optimal luminance is, intriguingly, maintained. We attribute this to the mechanical properties of the networked structure of the Al/MWCNT/Al electrode.

The composite cathode keeps the surface morphology crack-free even under large-angle bends since dislocations within the Al are being pinned and additional some stress is taken by the nanotubes – not the malleable metal. Again, optical micrographs at 200× magnification confirms the morphology as seen in Figure 5-5b. However, the EL performances in MWCNT-based devices do show a slight shift in the voltage where the maximum luminance occurs under different bending angles compared with Al-based devices. We attribute the difference to the reflections on back electrodes: normal reflection to Al and random scattering to Al/MWCNTs/Al (Insets in Figure 5-4a and 4b).

There is one final issue. AC-OEL devices can be thought of as a light emitting capacitor. High frequency electric fields induce a polarization current within the device to generate light. Therefore the electrodes must be just as highly conductive and homogeneous as in the case of OLEDs. Creases and voids in the Al film due poor

100 mechanical properties can cause local potential variations in the device which show up as brightness inhomogeneity. The network of MWCNTs clearly imparts some mechanical stability on the Al film during bending leading to uniformity in the composite cathodes spreading resistance. To see this point more clearly, RMS current density vs the RMS voltage (J-V) of Al-based devices and composite electrode-based devices are compared in Figure 5-4c and 4d. The difference between J-V curves of Al-based case before and after a full cycle of 120° bend is an unmistakable increase in the resistivity (slope of the line). However, the J-V characteristics of Al/MWCNTs/Al device before and after bending are nearly identical suggesting not very much permanent morphological change has taken place.

(a) (b) Figure 5-6 | Power efficiencies with Al (a) or Al/MWCNT/Al (b) as function of luminance for flexible AC-OEL devices.

In measuring the power efficiency of the devices, we use the standard convention of measuring the current voltage and the phase angle between them.50 For both cathode types (with and without MWCNTs) the overall power efficiency of flat devices are quite similar except for small fluctuations from device to device. However, in Figure 5-6, we

101 show what happens to the power efficiency of composite cathode devices as compared to pure Al cathodes under bending. For example, a pure Al cathode device flat that has not been bent achieves approximately 22 lm/W at 3250 cd/m2 but yields only 18 lm/W at

3250 cd/m2 after bending once to 120° and fully relaxing back. Notice that this is not the case for the composite cathode devices that tracks almost exactly in its recovery from

120° bending.

Figure 5-7 | Electroluminescence spectrum of flexible organic AC-El device with Al/MWCNTs/Al cathode at different bending angles.

Importantly, the EL spectra of devices with MWCNT cathodes taken at different bending angles and displayed in Figure 5-7, do show subtle differences. Most notably the recovered spectra from a previously bent device seems to suggest some damage is done to the device, perhaps delamination at the interfaces. However, it is important to realize that this is for one set of loadings only and optimization of MWCNT content has yet to be completely optimized.

102

Iterative Bending Tests on Devices with Al/MWCNTs/Al Electrode

Iterative bending tests on the AC-OEL devices were performed holding the bending angle at a constant 90° (radius of curvature ~ 9 mm), for each bend. Shown in

Figure 5-8a, the luminance remains very stable in devices with MWCNT composite electrodes, requiring over 500 bending cycles to drop to half of the initial luminance

(1250 cd/m2). In comparison, devices with pure Al electrodes exhibit a luminance drop of the same magnitude (half of the initial brightness (1270 cd/m2)) with only 50 bending cycles. The insert in Figure 5-8a shows an illuminated AC-OEL device after 1000 bending cycles.

(a) (b)

(c)

103

Figure 5-8 | Bending test of flexible organic AC-EL devices with and without MWCNT nano-network modification. (a) Luminance as a function of the number of cycles (0, 20, 25, 50, 100, 200, 500, and 1000). The insert shows the photograph of the flexible device with MWCNTs after 1000 bends. Luminance - current density - voltage characteristic of flexible organic AC-OEL devices before (b) and after (c) 1300 cycles of bend.

A voltage sweep at 50 kHz driving frequency was made before and after 1300 bending cycles and L-J-V characteristics are shown in Figure 5-8b and 8c respectively.

Compared with identical performances before bends (Figure 5-8b), after continuous

1300 bending cycles the device with a MWCNT composite cathode shows a luminance and current density of over 1290 cd/m2 at 133 mA/cm2 while the Al-based cathode device only produced 600cd/m2 at 93 mA/cm2 (Figure 5-8c). The superior mechanical reliability of the AC-OEL device with an Al/MWCNT/Al composite cathode, again, can be attributed to the strong mechanical strength of the MWCNT network together with the nanoinclusions ability to pin mobile defects in the metal films.

Flexible Carrier Gate in White AC-EL

Solar-Spectral, flexible, indoor illumination panels, wearable display technologies and roll-up cell phones are all a part of the imaginative “form-factor advantage” that has brought so much attention to the field of organic EL.51,98 Traditionally, it has been recognized that wide-panel organic light emitting diodes (OLEDs) face specific challenges associated with controlling large currents across large areas. The comparatively newer, capacitive and AC-OEL offer some advantages in this respect over the OLED. However, flexibility has generally required the use of a more conventional approach to AC-OELs, utilizing symmetric50 or asymmetric52,116 organic insulators sandwiching the emitting complex. PVDF, a highly non-reactive thermoplastic

104 fluoropolymer, has been popular in flexible AC-OEL device studies117, but, its high dielectric constant leads to high driving voltages that can make the device lossy.54

It has recently been shown that AC-OEL devices gated by n-type wide band gap semiconducting layers in place of the insulator layers, exhibit greater carrier manipulation and luminescent performance advantages in AC power cycles.97,172 The “model” for this improvement is rather straightforward. By utilizing a semiconductor gate between anode and HGL, hot holes are prevented from direct injection in the negative half of the AC cycle, and electron extraction in HGL is facilitated. Of course, inorganic crystalline gate materials such as ZnO, NiO, and TiO2 have been the focus of this approach and this makes the full realization of the flexible potential of AC-OEL devices difficult.

In the following section, we present an AC-OEL structure that utilizes a novel polymeric gate layer for efficient free carrier manipulation in high frequency power cycles. The n-type PFN-Br has been used to demonstrate passive carrier management as a gate in a highly flexible platform. This gate system has also been used in this work, to construct a high CRI white light emission AC-OEL device comprised of a solution- processed single emission layer with phosphorescent emitters lightly-doped into a fluorescent blue host, simultaneously emitting fluorescence and phosphorescence.

To harvest triplet excitons for power-efficient white light, organic sources, metallo-organic phosphors (e.g. Pt, Os, Ir, Au, Pd, and Ru complexes) are used to achieve near 100% internal quantum efficiency.16,17 Blue173,174, green175, and red176 phosphorescent dopants are typically employed in a single host matrix for wide spectral output.177 Heavy doping of these phosphors (over 5wt%) is a common method to improve the brightness and light quality of OEL devices. But, triplet-triplet annihilation (TTA)

105 dramatically quenches radiative excitons, or yields delayed fluorescence, and substantially diminishes the luminescent efficiency and accelerates degradation.178,179 The low endothermic triplet transfer rate also makes it difficult to achieve high CRIs in a multi-blend emission layer devices.180

However, the optical design seems to take advantage of the integration of superior flexibility a very low-doping recipe, and AC-driving to yield a white AC-OEL device that can bend ±90° and with a high CRI (over 81 at 2800K). We note that in these studies little emphasis has been given to impedance matching of the driving circuit.

Consequently the maximum power efficiency measured was only 2.8 lm/W at 1,000 cd/m2, relatively low.

N-type Polymeric Gate Structure and Application

PFN-Br is a polyfluorene-based conjugated polyelectrolyte n-type semiconductor

(chemical structure is given in Figure 5-9a). Its HOMO and LUMO levels are located at

-5.73eV and -2.11eV indicating a wide band gap of 3.62 eV.100 The surface morphology of PFN-Br spin-coated onto ITO is shown in Figure 5-9b. PFN-Br has a rather high electron mobility of 1.41 X 10-7cm2V-1s-1, measured using an “electron-only” device for

PFN-Br.

106

Figure 5-9 | (a) Chemical structure of PFN-Br. (b) Morphology of spin-coated PFN-Br thin film.

The wide band gap and high electron mobility makes PFN-Br an excellent candidate for the carrier management gate layer in the gated AC-OEL as Figure 5-10. As an external electric field is applied, pointing from the ITO side to the metal anode, the

“forward” direction of the device-diode, a mass of positive carriers drift toward the low work function metal electrode through the MoO3 buffer layer. Simultaneously, the excess electrons trapped in the HGL are transferred by the strong electric field toward the energy barrier at the HGL/PFN-Br interface and tunnel into PFN-Br’s LUMO level.

Electron extraction in the HGL dramatically facilitates the electron hopping rate from

Poly-TPD to F4TCNQ, resulting in a larger population of holes. In other words, the gate layer “opens the door” for electron extraction.

Gate on Gate off

Figure 5-10 | Energy band diagrams of gate functionality in forward and reverse bias.

Conversely, when the applied AC voltage changes polarity, and the electrical field is reversed across the entire device, the Fermi level tilts the opposite way around. Holes

107 in the HOMO of Poly-TPD are driven in the opposite direction and are accumulated at the interface of PFN-Br and HGL because of the high tunneling barrier (~ 0.43 eV). The

PFN-Br gate thus, “shuts the door” for electron extraction before the next forward cycle arrives. Therefore, the PFN-Br gate plays a role of electron extraction layer (“gate on”) in the forward bias and works as dielectric layer (“gate off”) in the reverse cycle.

Figure 5-11 | I-V demonstration of PFN-Br polymeric gate in DC mode.

Experimentally, we investigated current density for forward and reversed voltage of the gate structure: ITO/PFN-Br(40nm)/Poly-TPD: F4TCNQ(~70 nm)/MoO3(10 nm)/

Au(60 nm) in Figure 5-11. The peak current is 161.9 mA/cm2 in the forward bias of 3V, and crosses through the origin. As the bias is reversed to 3V, the current becomes independent from the applied voltage, maintaining a low level contant (less than

0.67mA/cm2). This is indicative of the “gate off” condition in reverse bias. In comparison, the I-V characteristics of a “hole-only” device without the PFN-Br gate shows the lack of this current cut-off ability in negative bias (Figure 5-12).

108

Figure 5-12 | Current density characteristic of ITO/Poly-TPD:F4TCNQ/MoO3/ Au in forward and reversed bias.

The PFN-Br gate structure is utilized in an AC-OEL device as illustrated in

Figure 5-13a. The device fabrication details are listed in the supplementary experimental section. The electron-hole pairs are generated in the Poly-TPD:F4TCNQ layer via electrical polarization and donor-acceptor heterojunctions. Holes and electrons are transported toward the emission layers in the “forward” part of the AC cycle. There, they recombine using Ir(ppy)3 molecules for scavenging of the triplet-state excitons. As discussed previously, PFN-Br guarantees exceptional carrier manipulation in AC power cycles, resulting in carrier balance throughout the cycle. The normalized EL intensity versus frequency characteristics of AC-OEL devices with a variety of thicknesses of

PFN-Br gate, made from polymer solutions of different concentrations, are shown in

Figure 5-13b. Comparing a non-gated device and a device with a gate (one made from a

0.5mg/mL solution of PFN-Br in solvent, for example), the device without PFN-Br clearly seems to have a greater injection current at low frequency. This is expected due to the loss of the hole barrier that the gate layer provided. Unexpectedly though, for very low frequencies (that is, near DC) this direct injection of current gets stronger with

109 increasing gate thickness (via solution concentration: 1mg/mL, 2mg/mL, and 5mg/mL).

Of course, in the experiment the voltage across the entire device is increasing for the increasing gate thicknesses so as to make light emission comparable since the field across the emitter must be the same. However, total resistance in the gate layer is increasing and this means the voltage dropped across this layer is increasing. This will naturally result in an increase in the direct current injection in this layer.

Figure 5-13 | Performance of PFN-Br gated green AC-OEL device. (a) The energy level diagram of the PFN-Br gated AC-OEL device with Ir(ppy)3 green phosphorescent dopant. (b) Frequency response of normalized EL intensity with different thickness of PFN-Br gate layer. (c) and (d) are the relation between RMS voltage versus brightness and RMS voltage versus RMS current density at 40kHz respectively.

110

Meanwhile, JRMS vs. VRMS measurements were conducted at 40kHz and the results are shown in Figure 5-13c and 13d respectively. The gate thickness-dependent characteristics also can be observed in “turn-on” voltage that increases from 7.8 V

(0.5mg/mL) to 11.5 V (5mg/mL). Meanwhile, the PFN-Br gated device exhibits a maximum brightness of 5,840 cd/m2 at 205.3 mA/cm2, comparing to 4,860 cd/m2 at 221.7 mA/cm2 (gate-free), which implies a superior hole carrier facilitation by the gate layer at the resonant frequency.

The doping strategy of phosphorescent molecules is a crucial factor in efficient triplet harvesting for such a device. Normally there is a tradeoff between insufficient

Förster transfer and exciton-exciton annihilation. Figure 5-15a shows the normlized lifetimes of 1wt% (insufficient Förster transfer), 3wt%, and 5wt% (TTA) doped AC-OEL device. Not surprisingly, both insufficient Förster transfer and TTA process impact EL luminance (as shown in Figure 5-14), more importantly, they accelerate the degradation of the OEL device. Thus, we choose to utilize 2wt% of phosphoresecent dopants in the

PVK host matrix, including the green Ir(ppy)3 and an additional red dye, Ir(MDQ)2(acac).

That suggests the excited e-h pairs are initially formed in PVK and then transferred to

Ir(ppy)3 and Ir(MDQ)2(acac) molecules, as band diagram illustrated in Figure 5-15b.

Then, the recombination of excitons takes place via PVK, Ir(ppy)3, or Ir(MDQ)2(acac) in a specified ratio and the white emission is generated consequently.

111

Figure 5-14 | Brightness relation with RMS AC voltage in the white AC-OEL devices with 1wt%, 2wt%, and 5wt% phosphorescent dopants [Ir(ppy)3:Ir(MDQ)2(acac) = 2:1].

Figure 5-15 | (a) Degradation of AC-OEL device with different concentrations of phosphorescent: 5%, 2%, and 1wt% in weight. (b) Energy transfer mechanism of dual doping white emission system.

Figure 5-16a presents the EL luminance as a function of frequency for the white

AC-OEL devices. The total phosphorescent doping weight percentage is pinned down to

2wt%, meanwhile, the weight ratios of Ir(ppy)3: Ir(MDQ)2(acac) are precisely tuned from

3:1, 2:1, and 1:1. At constant voltage (~11.5V), the luminance intensity of all devices rises as the driving frequency increases up to 40kHz, where the peak brightness occurs.

112

These peak brightness’ are: 1,200 cd/m2, 930 cd/m2, and 690 cd/m2 for 3:1, 2:1, and 1:1 respectively. Also, identical JRMS - frequency resonant behaviors are found in Figure 5-

16b. It is worth noting that the effective current consists of two contributions: a DC component (direct injection related) and AC component (polarization current related).

The relations between luminance and electrical field intensity are also shown in Figure

5-16c. At resonant frequency, the devices with 3:1 ratio show the highest luminance of

4,010 cd/m2 under 5.5×107 V/m. The devices with 2:1 and 1:1 doping exhibit the maximum brightness, 3,650 cd/m2 under 5.4×107 V/m and 2,870 cd/m2 under 5.3×107

V/m, respectively. The corresponding current properties are given in Figure 5-16d.

Figure 5-16 | (a) and (b) Frequency response of brightness and RMS current density in three doping ratios of Ir(ppy)3 and Ir(MDQ)2(acac): 3:1, 2:1, and 1:1 at a constant

113 voltage of 11.5V (±1V). (c) and (d) The brightness and RMS current density characteristics in the function of AC electric field intensity

(a) (b)

(c) Figure 5-17 | The EL spectrum of white emission AC-OEL device with different doping ratios of Ir(ppy)3 and Ir(MDQ)2(acac): (a) 3:1, (b) 2:1, and (c) 1:1.

Table 3 | Summary of electrical and luminescent properties of white emission AC-OEL devices.

Doping Frequency Max. L Powder Power Color Color CIE ratio (Hz) (cd/m2) factor efficiency rendering temperatur (lm/W) index e (K) 3:1 40,000 4,010 0.0628 5.0 80.86 3,077 0.45,0.47 2:1 40,000 3,365 0.0642 4.0 85.12 2,582 0.48,0.44 1:1 40,000 2,870 0.0660 3.4 75.98 1,986 0.52,0.41

114

Figure 5-17 shows the EL spectrum of AC-OEL devices with 3:1, 2:1, and 1:1 doping ratios and there is no significant spectral shift in luminance of 100, 500, 1000, and

2000 cd m−2. Even though the relative peak intensity may change slightly with different doping ratios, each spectrum consists of three main contributions: weak PVK’s blue fluorescence (410 nm), 510 nm green phosphorescence due to Ir(ppy)3, and 600 nm red phosphorescence from Ir(MDQ)2(acac). In the co-doping configuration of Ir(ppy)3:

Ir(MDQ)2(acac) = 3:1, the Ir(MDQ)2(acac) peak is slightly stronger than Ir(ppy)3 peak which eventually gives us a cold white with CRI of 80.86. With heavier doping of

Ir(MDQ)2(acac) (2:1 ratio), Ir(ppy)3’s emission is weakened consequently, which demonstrates a 2582 K white emission with a CRI as high as 85.12. However, the device’s output becomes a reddish warm white with a CCT of 1986 K and a CRIof 75.98 with too many Ir(MDQ)2(acac) molecules (1:1 ratio). The electrical luminescent data of white AC-OEL devices aresummarized in Table 1. Therefore, the enhancement of

Ir(ppy)3 doping contribution has a great impact on overall performance. This is attributed to the high quantum efficiency of Ir(ppy)3 molecules compared with Ir(MDQ)2(acac) molecules as shown in Table 2. More importantly, it was demonstrated that the For- ster resonance energy transfer efficiency (FRET) from PVK toIr(ppy)3 boost from 21.7% to

99.1% while the doping concentration increases from 0.1 to 2 wt%, which enables a very efficient radiative excitons generation.

115

(a) (b) (c) Figure 5-18 | The photos of white emission AC-OEL device with different doping ratios of Ir(ppy)3 and Ir(MDQ)2(acac): (a) 3:1, (b) 2:1, and (c) 1:1.

(a) (b) Figure 5-19 | (a) Device structure of PET-based AC-OEL device. (b) The photos of devices are bent in convex position and concave position. The scale bar is 10mm.

116

(a) (b) Figure 5-20 | The luminescent and electrical stability in the function of bending angle from -90deg to +90deg. (a) PFN-Br gate; (b) ZnO gate.

(a) (b)

(c)

117

The improved flexibility of PFN-Br polymeric layer eno mously advances the inorganic gate layer and allows us to fabricate mechanically flexible gated AC-OEL devices with relatively high performance. The PFN-Br gate was laid down on the cleaned ITO/PET substrate by spin-coating. A 70 nm layer of poly-TPD:F4TCNQ was then added on top of this gate layer to act as an HGL. We employed the doping level of 2:1 of Ir(ppy)3 and

Ir(MDQ)2(acac) in consideration of high luminance and CRI. The device was completed after thermal deposition of Bphen and Ca/Al contacts. The entire flexible structure is shown in Figure 5-19a. Since the AC-OEL device was fabricated on one side of the PET substrate, there are generally two basic bending positions: convex and concave (shown in

Figure 5-19b). The bending angle is defined as negative (from −90° to 0°) in convex bending and as positive (from 0° to +90°) in concave bending. Figure 5 D shows the luminance and current variation as a function of bending angle which ranges from −90° to +90° in initial no-bending luminance of 200 cd m−2 (red dots and line) and 500 cd m−2

(black dots and line). As shown in Figure 5D, the polymer gated flexible device shows stable luminescent output (200 ± 20 and 410 ± 90 cd m−2) compared with ZnO-gated device in Figure S3 (Supporting Information). The unstable current under convex bending (72–64.3 mA cm−2) results in some luminescent fluctuation (500–340 cd m−2).

This is attributed to the ITO’s tensile deformation during bending which is confirmed by the SEM images shown in Figure 5-22. In addition, iterative bending tests on the AC-

OEL devices were performed with sputtered ZnO gate and PFN-Br gate. As shown in

118

Figure 5-23, the luminance remains very stable in devices with PFN-Br with 200 bending. In comparison, devices with sputtered ZnO gate exhibit a significant luminance drop (over the half of the initial brightness) within only five bending cycles. Furthermore, the frequency responses of the flexible AC-OEL device were tested at the bending angles of 0°, ±90°, and ±45°. The results are shown in Figure 5-21a where a trivial resonant frequency shift and no current variations were found. The L–V curves and power efficiency were also characterized and shown in Figure 5-21b. The small deviation from unflexed device parameters demonstrates the high bending stability of PFN-Br-gated AC-

OEL devices. Finally, the white EL spectra is shown to remain intact and stable in convex and concave bending (Figure 5-21c).

(a) (b) Figure 5-22 | SEM images of PET-based flexible ITO substrates coated ZnO gate (a) and PFNBr gate (b) after one full cycle bending.

119

Figure 5-23| Iterative bending tests on the AC-OEL devices with different gate materials.

In addition, iterative bending tests on the AC-OEL devices were performed on sputtered

ZnO gate and PFN-Br gate devices shown in Figure 3b. The luminance drops less 30% from the initial in PFN-Br device with 200 bending cycles. In comparison, devices with sputtered ZnO exhibit a significant luminance degradation (over the half of the initial brightness) within only 5 bending cycles.

Chapter Summary

In this chapter, we showed that Al/MWCNTs/Al nano matrix is potential working as extremely durable flexible electrode in thin film lighting device. We also demonstrated the white emission of gated AC-OEL devices on both rigid glass and flexible PET substrates using an n-type conjugate polymer, PFN-Br, as the gate layer. The polymeric gate between the ITO electrode and HGL has been shown to provide superior carrier manipulation in hole blocking and electron extraction for positive and negative bias of the AC power cycle of the device. We also studied a low studied a low concentration doping strategy of phosphorescent dyes in the gated AC-OEL device to balance the tradeoff between insufficient Förster transfer and exciton annihilation.

120

Chapter 6 Imbedded Nanocrystals of CsPbBr3 in Cs4PbBr6 Matrix in the Application of LED

Solution-grown films of CsPbBr3 nanocrystals imbedded in Cs4PbBr6 are incorporated as the recombination layer in light-emitting diode structures. Analysis of absorption strength of this all-perovskite, all-inorganic imbedded nanocrystal composite relative to pure CsPbBr3 indicates enhanced oscillator strength consistent with earlier published attribution of the subnanosecond exciton radiative lifetime in nano-precipitates of CsPbBr3 in melt-grown CsBr host crystals and CsPbBr3 evaporated films. Kinetics at high carrier density of pure (extended) CsPbBr3 and the nano-inclusion composite are measured and analyzed, indicating second order kinetics in extended and mainly first order kinetics in the confined CsPbBr3, respectively. Carriers injected or excited in the wider gap Cs4PbBr6 host can be concentrated by collection into the heterostructure wells corresponding to the smaller gap CsPbBr3 inclusions, raising the recombination rate and further enabling high quantum yield in competition with traps.

Discovery of CsPbBr3-Cs4PbBr6 nanocompsite

Recent work has shown that nanocrystals of CsPbBr3 imbedded in a Cs4PbBr6 host matrix achieve photoluminescence quantum yield (PLQY) up to 90% 181 in contrast

182–185 to very low light yield from bulk single crystal or polycrystalline CsPbBr3 .

Furthermore, the unusually fast (~200 ps) exciton luminescence decay time measured in

186,187 CsPbBr3 nanocrystal precipitates in a melt-grown CsBr host is seen to be preserved even at the high PLQY achieved in solution-processed samples of our work and Ref.181

This implies that enhanced exciton oscillator strength rather than quenching/trapping is

121 mainly responsible for the short lifetime of luminescence, in line with the previous

186,187 184,188 attribution by Nikl et al and Kondo et al in CsPbBr3 nanocrystals in melt- grown CsBr or evaporated CsPbBr3 films.

Optical properties of bulk CsPbBr3 and Cs4PbBr6 single crystals and melted or annealed film perovskites are known from a series of studies from 1996 onward.182,184,188–

191 In addition to determining band gaps and exciton luminescence energies of the two compounds, the occurrence of green luminescence from an impurity phase of CsPbBr3 in

Cs4PbBr6 crystals was noted. Nikl et al remarked that such an impurity phase is likely unavoidable in melt-grown Cs4PbBr6 because of incongruent melting in its phase

190 diagram. Unusually fast (90-300ps at 10K) luminescence lifetime in CsPbBr3 was observed and the possibility that exciton superradiance could be involved was remarked.184,189 The light yield of ~ 2.35-eV green luminescence from bulk crystalline

182 CsPbBr3 was characterized as weak even at low temperature. The exciton binding

192 193 energy in the semiconductor CsPbB3 has been reported as 19 meV to 62 meV , so at room temperature (including the present study) excitons should quickly be ionized to free carriers and/or transferred to more deeply trapped excitons at defect sites The exciton luminescence of Cs4PbBr6 was observed at 3.3 eV and its binding energy is deeper than

194 195 in CsPbB3, reported as 353 meV or 222 meV . Nikl et al prepared nanocrystals of

CsPbBr3 precipitated in melt-grown CsBr, but the solubility of Pb in solid CsBr limits the

186 concentration to about 0.2% . Kondo et al produced high concentrations of CsPbBr3 nanocrystals in films produced by moderate annealing of CsPbBr3 evaporated on cold substrates, and demonstrated stimulated emission in the film of imbedded CsPbBr3 nanocrystallites.196

122

In this chapter, we adds to the observations and resulting conclusions on the imbedded nanocrystal origin of the fast and efficient green luminescence reported by

Quan et al 181as well as preceding empirical observations.194,197 Going further, we describe construction and characterization of LEDs utilizing the CsPbBr3 nano-inclusions in Cs4PbBr6 as the active medium. Analysis of absorption strength of this all-perovskite, all-inorganic imbedded nanocrystal composite relative to pure CsPbBr3 demonstrates enhanced oscillator strength, to which the unusually fast subnanosecond exciton radiative lifetime in nano-precipitates of CsPbBr3 in melt-grown CsBr and partially annealed

CsPbBr3 evaporated films can also be attributed. High density excitation of pure extended CsPbBr3 is examined and analyzed on the basis of bimolecular radiative recombination in competition with trapping and Auger recombination.198 Analysis of radiative kinetics in the nano-inclusion composite indicate first-order recombination despite the known thermal instability of excitons at room temperature in CsPbBr3.

Methods and Measurement

Photoluminescence Quantum Yield (PLQY)

An integrating sphere was employed to determine the QY of perovskite samples.

The sample was placed in a thin cuvette and was excited with 400 nm blue LED light.

The scattered excitation, and the emission, were collected together using a fiber optic coupled spectrometer (FLAME-S-XR1-ES). The whole spectrum was collected sequentially for a reference (cuvette with DMSO only) and for the sample.

The ratio of the number of emission photons over the number of absorbed photons was used to determine the QY as shown in the equation below,

123

푆 − 푅 푄푌 = 푒푚 푒푚 푅푒푥푐 − 푆푒푥푐

Eqn. (6-1)

where Sem and Sexc are the integrated signal in the emission and excitation region, respectively, for the sample, and Rem and Rexc are the integrated signal in the emission and excitation region, respectively, for the reference. As a check for the accuracy of the system, the QY of Rhodamine 6G in ethanol was found to be 94.4%, close to the literature value of 95%. For the power dependent PLQY measurement, the intensity of the excitation UV beam is controlled by the current passing through UV diode, measuring the power with a Newport power meter (Model 843-R).

X-ray photoelectron spectroscopy (XPS)

XPS (AXIS ULtrabld) measurements were performed on the Cs-Pb-Br perovkite thin films. To maintain the accuracy, each XPS pattern was averaged from 15 measurements.

Absorption coefficient

The high transparency of the Cs-Pb-Br pervoskite thin film enabled us to record its UV-Vis absorbance in transmission mode with a Cary 50 UV-Vis spectrophotometer, while an identical glass slide was used as a baseline reference for the absorption measurements. With equation S2, we carried out the calculation of absorption coefficient of Cs-Pb-Br perovkite thin films built from different PbBr-CsBr molar ratio precursors.

-1 -1 Absorption Coefficient = d ln(I0/I) (cm )

Eqn. (6-2)

124 where d is the thickness of pervoskite thin film, I/Io is transmittance.

Time-resolved PL measurement

For measurement of the transient PL characteristics, short-pulse frequency- doubled Ti-Sapphire excitation with a pulse width of 300 fs and a wavelength of 420 nm was used in combination with a Hamamatsu C2830 streak camera system. The measurement was carried out in ambient at room temperature. The excitation fluence

(10Hz) was kept to ~ 60μJ/cm2 in collimated beam/

LED fabrication

Devices were built on a 2.54cm X 2.54cm glass substrate pre-coated with 140nm thickness layer of ITO having a sheet resistance 10 Ohm/square. The ITO glass substrates are cleaned in an ultrasonic bath with acetone followed by methanol and isopropanol for

1 hour each, and then dry-cleaned for 30 min by exposure to an UV-ozone ambient.

PEDOT:PSS solutions (filtered through a 0.45 µm filter) were spin-coated onto the ITO- coated glass substrates at 4000 rpm for 30s and baked at 100 C for 30 min in a vacuum oven. The hole transporting layer was prepared by spin-coating Poly-TPD chlorobenzene solution (concentration: 5 mg/mL) at 2000 rpm for 30s. After baking, the substrates were transferred into glove box for further depositions. Perovskite layers were deposited by spin-coating at 4000 rpm for 60s with 70°C annealing for 5min. TPBi (30 nm) and Al electrode (100 nm) were deposited using a thermal evaporation system through a shadow mask under a vacuum of 2 × 10-7 Torr.

LED characterization

125

The LED devices in this report are measured in on-off cycles inside of glove box at room temperature (25℃) without encapsulation. Current versus voltage characteristics were measured using a Keithley 2400 unit. Simultaneously, luminance (in cd/m2) was directly measured using a photometer (ILT 1400-A) with an optical fiber facing the light- emitting pixel. The EL spectra for the devices were collected by an ILT 950 spectroradiometer (International Light Technologies).

Solution growth and properties of Cs-Pb-Br perovskite thin films.

We fabricated the Cs-Pb-Br films from solutions made by dissolving PbBr2 and

CsBr in DMSO at molar ratios of 1:1, 1:1.2, 1:1.5, and 1:2. Films were spun on glass substrates at room temperature in a nitrogen-filled glovebox as shown Figure 6-1a. The formation of Cs-Pb-Br perovskites was initiated by dripping acetone for fast crystallization during spin-coating.

(a)

126

(b) Figure 6-1 | The fabrication process of Cs-Pb-Br perovskite thin films.

The resultant thin films are transparent with a light yellow tint in room light as shown in Figure 6-2a. The film thickness is approximately 100 nm as measured by

AFM at a scratch edge as Figure 6-2c. Under uniform UV excitation (λ = 365nm), the equi-molar (1:1 precursor) sample exhibits extremely low luminescence, whereas the other three samples with excess CsBr in the precursor solution glow brightly as shown in

Figure 6-2b.

(a) (b)

127

(c)

(d) Figure 6-2 | Cs-Pb-Br perovskite thin films on glass slides in room light (a) and uniform 365nm ultra-violet light (b) at room temperature. (c) AFM measurement of film thinkness. (d) XRD patterns of Cs-Pb-Br perovskite thin films from precursor solutions.

X-ray diffraction (XRD) tests were carried out on these thin films to determine crystalline composition, and the results are shown in Figure 6-2d. The XRD pattern of the (1:1) CsBr-PbBr2 thin film matches well with the orthorhombic crystal structure of

CsPbBr3. When CsBr is in excess (1:1.2 up to 1:2 shown), the main peak (112) of

CsPbBr3 becomes weak and disappears effectively into the background. New peaks at

128

12.4°, 19.9°, 25.3°, 28.5°, and 30.1° become observable, which identifies the Cs4PbBr6 phase of pervoskite as the majority constituent in the 1:2 film. If XRD peak heights are taken as a rough indicator of fractional composition, the data for the 1:2 sample suggests an upper limit for a minority phase CsPbBr3 of about 10% in the Cs4PbBr6 majority phase. The quantum yield of photoluminescence (PLQY) of the films under 400 nm excitation increased from 0.1% (1:1 sample) to 29.4% (1:2 sample) as shown in Figure

6-3a. The PL of the 1:2 composite thin film shows a blue shift toward 2.41 eV from 2.36 eV as the fraction of the minority phase CsPbBr3 in Cs4PbBr6 increases as shown in

Figure 6-3b. XPS measurements were carried out on the thin films as shown in Figure

6-4.

(a) (b) Figure 6-3 | (a) Quantum yield (QY) of Cs-Pb-Br perovskite thin films from precursor solutions with different PbBr2-CsBr molar ratios (1:1, 1:1.2, 1:1.5, and 1:2). (b) Photoluminescence (PL) shift of different Cs-Pb-Br perovskite thin films.

129

(a) (b)

CsPbBr3 and Cs4PbBr6 DMSO-washed powder samples.

Additionally, the CsPbBr3 and Cs4PbBr6 crystals were synthesized in precipitated powder form and then purified by DMSO washing as described in Ref. 194 to remove exposed CsPbBr3 while leaving less soluble Cs4PbBr6. Upon doing this, we found that the green 2.4 eV PLQY of the nominal Cs4PbBr6 powder prepared from 1:2 precursor increased from 27% (unwashed) up to 70% (DMSO-washed). The fact that removing exposed grains of CsPbBr3 makes the acknowledged CsPbBr3 PLQY increase dramatically implies that the active CsPbBr3 is contained and protected within the less soluble Cs4PbBr6 host grains.

130

(a) (b)

Figure 6-5 | Photos of pure CsPbBr3 (Cs1) and Cs4PbBr6 crystals with imbedded CsPbBr3 nano-inclusions (Cs4) in room light (a) and UV light (b).

The CsPbBr3 and Cs4PbBr6 powders in room light display orange and light green color respectively. With uniform UV excitation (~365nm), the nominal Cs4PbBr6 powder exhibits bright green luminescence, contrasting no glow from CsPbBr3 powder in identical excitation condition (shown in Figure 6-5). The DMSO-washed Cs4PbBr6 powder has a PLQY of 70% with 400nm excitation while nearly zero was found from

CsPbBr3.

131

Morphology of Thin Films Comprised of CsPbBr3/Cs4PbBr6 Nano-inclusion Grains.

To prevent high current leakage through the emitter in a LED device, a compact thin film of crystallites is required. An AFM image of cubic CsPbBr3 thin film (1:1) is shown in Figure 6-6a. A TEM image (Figure 6-6b) and SAED (Figure 6-6c) of a specimen made by scraping a 1:1 thin film shows the similar cubic CsPbBr3 structure.

The 1:2 ratio thin film sample shows an average crystallite size of 70nm in Figure 6-6d.

The TEM images of the crystals in Figure 6-6e show approximately 10 nm nano- inclusions around the edges of the host grains. The SAED (Figure 6-6f) identifies the coexistence of Cs4PbBr6 and CsPbBr3 conclusively. The growth direction of 100 is

132 favorable for CsPbBr3 in Cs4PbBr6 crystal composed thin film. Further microscopic images of a group of thin films are include in Figure 6-7.

(a) (b)

(e)

Figure 6-7 | SEM images of pervoskite thin films in a variety of PbB2-CsBr molar ratios: (a) 1:1, (b) 1:1.2, (c) 1:1.5, (d) 1:2, (e) 1:3.

133

Light Emitting Diodes (LEDs) using CsPbBr3 nanocrystals in Cs4PbBr6 host.

Figure 6-8 | The device structure of the Cs-Pb-Br pervoskite LEDs.

Cs-Pb-Br pervoskite based LEDs were fabricated from the different molar ratio precursor solutions with the device structure shown in Figure 6-8. Under static electric field, hot carriers (holes and electrons) are injected from ITO (anode) and Al (cathode) over HTL and EIL, and generate radiative excitons in the Cs-Pb-Br perovskite emission layer. The LED fabricated from CsPbBr3 solution (1:1) showed poor luminous characteristics (maximum brightness = 20 cd/m2 at 106 mA/cm2) in Figure 6-7a, mainly due to large CsPbBr3 crystals with weak exciton binding facilitating exciton dissociation.

However, the CsPbBr3 nanocrystals in Cs4PbBr6 host show a significant improvement in luminous performance (850 cd/cm2 at 136mA/cm2). In Figure 6-7b and 6-7c, the LED fabricated from the equimolar precursor shows a poor luminous characteristic (maximum current efficiency = 3.33 × 10-3 cd/A, maximum EQE = 3.5 × 10-7). In contrast, the CE

-5 and EQE is enhanced to 0.30 cd/A and 2.4 × 10 respectively with the CsPbBr3 nano- inclusion in Cs4PbBr6 composite emitter.

134

The nonradiative recombination rates of biexcitons and charged excitons in Cs-

Pb-halide perovskite nanocrystals have been shown to seriously limit to efficiency of

199 luminescence in CsPbBr3 nanocrystals. The loss from Auger recombination is detrimental to prospective LED efficiency using traditional quantum dots.200 In the well- studied system of CdSe/CdS core/shell nanocrystals, it has been found that energy transfer loss through Auger recombination can be effectively inhibited by providing a

201–203 thick shell. The passivation distance between imbedded CsPbBr3 inclusions in our samples is ~ 7.3nm estimated from TEM images (Figure 6-10).

135

Figure 6-10| TEM of CsPbBr3 inclusions imbedded Cs4PbBr6. Scale bar is 100nm.

This can assist in suppression of unwanted energy transfer among neighboring nanocrystals and reduces the charging effect in nano-inclusions under device operation.

The low external quantum efficiency of our present devices might be attributed to insufficient isolation between nanocrystals. Further study of the influence of spatial distribution of CsPbBr3 inclusions in Cs4PbBr6 on the luminescence efficiency will be useful. Another reason for low radiative efficiency vs injection current can be Shockley-

Real-Hall (bimolecular) e-h recombination on mid-gap states, which Stranks et al204 have found to be important in hybrid Pb-halide perovskites. The luminance-voltage characteristics of the LEDs are measured in alternating on-off cycles at increasing voltage within 60 seconds (Figure 6-9d), which is known to give a more conservative efficiency than a simple voltage sweep. This gives the brightness of 1150 cd/m2 for our champion device. This enables us to scale up the pixel to a large-area of 2cm2 in Figure 6-11.

136

(a) (b)

(c) Figure 6-11 | 0.15 cm2 small area (a) and 2 cm2 large area pervoskite LEDs (b). (c) Device spectrum in a ramping voltage.

Our powder samples were prepared from DMSO precursor solution in air, and as a result they dried from the solvent somewhat slowly. Microcrystallites of CsPbBr3 were found to grow in the powder over a least 2 minutes after precipitation of the Cs4PbBr6 nanocomposite material. Subsequent exposure of the dried powder to DMSO vapor removes the CsPbBr3 microcrystals. The resulting purified powder retains stable optical properties including high PLQY for months. The films were spun down in a dry box leading to rapid drying. They did not have a color change that would correspond to

137

CsPbBr3 microcrystal growth, and if not subjected to voltage in an LED, they retained stable optical properties for 3 months. The degradation under applied voltage in air in an

LED configuration is rather rapid (less than 1 hours). The turn-on voltages of LEDs with

CsPbBr3 nanocrystals imbedded the Cs4PbBr6 host are between 3V and 4V, which is consistent with carrier injection into the Cs4PbBr6 host (Figure 6-12, band gap ≈ 3.7eV

181,191 ). The Cs4PbBr6 would be collected at the nanocrystal inclusions assisted by the band-edge offsets as supposed in Figure 6-12 illustrating a Type I heterostructure. Type I heterostructure could boost carrier recombination well, helping overcome the

198 “bimolecular bottleneck” . Proof of Type I band offsets in this CsPbBr3/Cs4PbBr6 heterostructure has been givn by K. Biswas based on the hybrid functional band structures with heterostructure slab supercell.205

(a) (b)

Figure 6-12 | (a) Schematic of operational LED device with Cs4PbBr6 crystal imbedded CsPbBr3 nano-inclusions. (b) Radiative energy transition with CsPbBr3 imbedded Cs4PbBr6 in a conjectured Type I heterostructure.

We also carried out the EL spectrum measurement in Figure 6-13a using a spectrometer under a constant current density of 50 mA/cm2.The LEDs show narrow full width at half maximum (FWHM) of about 18nm. The EL spectra have blue shift from

522nm, 521nm, 518nm, to 514nm in the precursor ratio sequence from 1:1 to 1:2 138 resulting a CIE coordinate shift. We compare the luminescent peak position shifts 0f the

EL and PL spectra in Figure 6-13b. The same magnitude of blue shifts occurs in PL (~

9nm) and EL (~ 8nm).

(a) (b)

Figure 6-13 | (a) EL spectra of LEDs with varying molar ratio of PbBr2 and CsBr. (b) Summary of spectral blue shifts in PL and EL.

Optical response and kinetics of pure CsPbBr3 and the CsPbBr3/Cs4PbBr6 nano- inclusion composite under high-intensity photoexcitation.

As shown in the preceding discussion and in published accounts, the PLQY of the

CsPbBr3 nano-inclusion composite under steady blue or UV illumination (as well as EL under current injection) is much higher than that of pure extended CsPbBr3. In thin-film form under modest intensity 0.1 W/cm2, 400nm LED excitation, we found a ratio of 290 between PLQY of the 1:2 film and that of the 1:1 film, and an even larger ratio between

PLQY of DMSO-washed powders of CsPbBr3 nano-inclusions in Cs4PbBr6 and pure

CsPbBr3 (powder morphology is shown in Figure 6-14).

139

(a) (b)

Figure 6-14 | Optical images and SEM images of pure CsPbBr3 (a)(c) and purified Cs4PbBr6 imbedded with CsPbBr3 nano-inclusions (b)(d). Scale bars are 2μm.

However, under high-intensity femtosecond laser excitation at 60 μJ/cm2 in 300 fs or ~ 0.12 GW/cm2 at 420 nm, the streak camera images in Figure 6-15a and 6-15b show that the pure CsPbBr3 powder emits roughly equal peak intensity and slightly larger time- integrated luminescence yield than the nano-composite powder which had been approximately 600 times brighter in steady state. The phenomena at work to produce this should include the competition of carrier trapping, bimolecular radiative recombination, and Auger decay such as discussed recently by Xing et al for organic/inorganic hybrid lead-iodide perovskite electroluminescence198. Transcribing their discussion to the inorganic perovskite light emitters of present interest, we first note that the low exciton

192 193 binding energy reported as 19meV to 62meV in pure CsPbBr3 means that at room temperature the recombination kinetics are expected to be mainly those of free and

140 trapped carriers, not excitons. Radiative recombination should thus obey bimolecular kinetics and so the Xing et al rate equation 198 applies. Qualitatively in this picture, the very low light yield of pure CsPbBr3 at micro-grain and larger size for low carrier density typical of steady-state optical excitation or LED operation with a homogeneous active recombination layer is a result of the intensity – dependent bimolecular recombination rate being smaller than the trapping rate at defects. Using the bimolecular rate constant

-10 3 -1 198 of k2 = 7 × 10 cm s tabulated and reported by Xing et al as typical for hybrid lead iodide perovskites, and LED injected carrier densities of n = 1015 cm-3 also typical of hybrid perovskite LEDs with homogeneous active layers, the initial rate of descent of the

1/e -1 bimolecular luminescence can be characterized by an effective initial decay rate (τPL )

5 -1 = k2n = 7 × 10 s . This is significantly slower than the carrier trapping rate achieved without taking unusual steps to exclude defects, and thus corresponds (in the present case) to very low light yield in pure extended CsPbBr3.

141

(a) (b)

Figure 6-15 | Subnanosecond luminescence under 300 fs, 60 μJ/cm2, 420 nm excitation. Streak camera images of CsPbBr3 micro-crystals (a) and CsPbBr3 nano-crystals in Cs4PbBr6 perovskite (b) at room temperature. (c) PL lifetime decay curves of Cs-Pb-Br thin films with different molar ratios, CsPbBr3 micro-crystals and CsPbBr3 nano- inclusions in Cs4PbBr6 crystals. (d) Fitting analysis on exciton dynamics of CsPbBr3 micro-crystals and CsPbBr3 nano-inclusions in Cs4PbBr6 crystals.

In our experimental conditions, the 60 μJ/cm2 of 420 nm laser light in a 300 fs pulse, taking into account the absorption coefficient of 3.3 × 104 cm-1 measured and reflection loss measured in the course of PLQY determination, produced approximately

18 3 1.4 × 10 e-h pairs per cm in CsPbBr3. Putting this initial carrier density into the

-10 3 -1 bimolecular rate equation with the rate constant k2 = 7 × 10 cm s , we obtain a model

1/e -1 prediction of the effective initial time constant τPL = (k2n) = 1 ns for the conditions of our experiment in Figure 6-15. The observed effective initial lifetime seen in Figure 6-

142

15a, 15c, and 15d is 0.97 ns. This agreement with the prediction based on excitation density and estimated rate constant suggests that the simple rate equation model properly describes pure CsPbBr3 powder of micro-grain size such as was used in our streak camera experiment. The bimolecular radiative decay time becomes extremely fast (~ 1 ns) for dense pumping and thus competes very well against the trapping rate, resulting in a dramatic increase of PLQY. The full 2nd order decay curve of course soon bends toward slower rate of decay as time progresses and the carrier density falls. By fitting the streak

nd nd data for pure CsPbBr3 powder to 2 order kinetics we obtain the fitted 2 order curve in

Figure 6-15d (green curve) and the time constant for decay to half value, t1/2 = 0.69 ns.

We want to underscore that in fitting Figure 6-15d (green curve) for pure CsPbBr3 powder, the initial excitation density was fixed at the measured value 1.4 ×1018 e-h pairs per cm3, so that there was only one variable fitting parameter, the 2nd order rate constant

-10 3 -1 -10 3 -1 k2. The fitting specifies k2 = 4.6 × 10 cm s , to be compared to 7 × 10 cm s , reported by Xing et al198 for extended (3D) hybrid lead-iodide perovskites.

Note that the imbedded nanocrystal composite powder labeled Cs4PbBr6 in the streak camera data did not fare as relatively well in gaining luminescence intensity versus pumping density as did the pure CsPbBr3 powder. This is partly a statement that the composite started out quite bright at low intensity pumping by steady illumination or

LED injection, and then remained high in PLQY as the excitation density was increased, possibly decreasing slightly due to dipole-dipole biexciton or Auger recombination.

There was simply not headroom or opportunity for such dramatic improvement of PLQY upon increasing the excitation density in the nano-inclusion composite that started out with high PLQY for low excitation density. The point is that confinement effects favoring

143 geminate electron-hole recombination in the composite already overcame the slow recombination that bimolecular kinetics would have predicted at low injection or excitation density. The nanocrystal boundaries in the Cs4PbBr6-encapsulated inclusions of CsPbBr3 enforce geminate recombination despite the weak exciton binding energy and consequently produce first-order decay kinetics even at low excitation density in a material whose excitons are thermally ionized in the bulk at room temperature. The data for the nano-composite powder in Figure 6-15c and 15d (red points) could not be well fitted with bimolecular kinetics alone, but did allow a fitting of the main part (longer component) with first-order kinetics (exponential decay time τ = 345 ps). The faster initial decay could be fit with addition of second-order kinetics having rate constant k2 =

-9 3 -1 1/e -1 st nd 1 × 10 cm s , corresponding to τPL = (k2n0) = 710 ps. The sum of the 1 and 2 order rates gives the steepening which can be seen at earliest time in Figure 15d (red).

The imbedded nanocrystals exhibit first order decay of the geminate e-h pairs enforced at room temperature by nano-crystal confinement. The initial second-order decay at the highest density is attributed to biexciton and Auger quenching of multiple geminate pairs per nanocrystal.

In summary, the reason that the CsPbBr3 nano-inclusion composite offers better performance at low excitation density appears to be partly because confinement encourages first-order recombination kinetics in a material that does not otherwise support stable excitons at room temperature. For LED applications, the collection of injected carriers from the Cs4PbBr6 host into Type I heterostructure islands of lower-gap

CsPbBr3 , if they occurs, could offer further advantage by concentrating carrier density into recombination wells. The flip side is that there is not as much room or reason for

144 improvement in this case as excitation density increases. Auger quenching defeat the advantage of quantum confinement in the imbedded nanocrystals at high excitation density. This may set an ultimate limit of radiative yield unless one takes measure to adjacent nanoparticles as shown in core/thick-shell systems.201–203 Improving isolation or engineering enhanced oscillator strength to help radiative decay compete with Auger decay might be other possibilities for high yield.

PL quantum efficiency and decay time dependence on size of the Cs4PbBr6 host micro-grain.

Cs4PbBr6 crystals were synthesized in a variety of sizes as imaged by TEM in

Figure 6-16. The size of the Cs4PbBr6 crystals is defined here as the longest distance between hexagonal corners which are 40nm, 100nm, 2000nm, and 5000nm for the examples shown. Under the excitation of 365nm UV, PL of Cs4PbBr6 crystals in

DMSO solution varies with crystal size, which can be observed in Figure 6-16e. We quantified the PL quantum yields of these Cs4PbBr6 crystal suspensions and the results are plotted in Figure 6-16f. The light yields are 0.01%, 34%, 55%, and 60% with ascending Cs4PbBr6 crystal size. Furthermore, the streak camera measurements of PL decay are shown in Figure 6-16, and the luminescence decay time (1/e) increases with the increase of Cs4PbBr6 host micro-grain size in the order 256ps, 301ps, 474ps, and

621ps. The increasing emission lifetime with increase of the host micro-grain size suggests that isolation of the nano-inclusions from the micro-grain boundary layer and associated defects may be important. The parallel increase of PLQY supports this hypothesis.

145

Figure 6-16 | Dependence of nano-inclusion photoluminescence on size of the Cs4PbBr6 host micro-grains. TEM images of Cs4PbBr6 crystals in different sizes: 40nm (a), 100nm (b), 2000nm (c), and 5000nm (d). (e) The image of Cs4PbBr6 crystal suspensions in DMSO under 365nm UV lamp at room temperature. (f) Quantum yield in a function of Cs4PbBr6 crystal sizes, and the PL lifetime versus Cs4PbBr6 crystal sizes. Streak camera images of Cs4PbBr6 crystal suspensions with a size of 40nm (g), 100nm (h), 2000nm (i), and 5000nm (j) at room temperature.

There are already some excellent works that demonstrate the suppression of blinking behavior and nonradiative Auger recombination in traditional core/shell nanocrystals by increasing shell thickness.201–203 The strong shell-thickness dependence is attributed to suppression of recombination rates of biexcitons. The Biexciton emission

146 efficiency is enhanced through electron localization around the core due to the strong

Coulomb potential of the two confined holes.202,206 According to established theories of radiative decay of quantum dots, radiative biexciton lifetimes should also be reduced compared to radiative exciton lifetimes by only a factor of 2 to 4.207 Our data indicate that PLQY improves with larger composite micro-grain size (possibly thicker Cs4PbBr6).

Meanwhile, the largest composite micro-grains show a 2.4 times increase of radiative decay lifetime compared to the smallest crystals. However, pending more thorough study of spatial distribution of CsPbBr3 nanocrystals in the Cs4PbBr6 matrix, we are unable to conclude yet that the large composite micro-grains directly diminish nonradiative Auger recombination.

Enhanced Oscillator Strength of CsPbBr3 Imbedded Nanocrystals Relative to

Extended CsPbBr3.

Figure 6-17a shows the absorption spectra for a series of (1:1), (1:1.2), (1:1.5), and (1:2) films measured with a Cary 50 UV-Vis spectrophotometer. The data are

-1 -1 expressed as absorption coefficient α = d ln(I0/I) (cm ) where d is film thickness measured by AFM at a scratch edge, I0 is power transmitted through a reference substrate, and I is power transmitted through the film on substrate. The deposited films sometimes had a cloudy appearance, which was manifested in the spectra as attenuation at wavelengths longer than 560 nm where both CsPbBr3 and CsPbBr6 are transparent.

Correction for Rayleigh scattering (⁓ λ-4) and Tyndall scattering (approximately independent of wavelength) was made using the FluorTool software extrapolating from data in the long-wavelength transparency range. Note that the 520 nm exciton absorption peak does not have a wavelength shift commensurate with the PL peaks. This suggests

147 that the cause of the main PL shift may be other than quantum confinement, possibly self- absorption by the sloping Urbach edge.

(a) (b)

Table 4 | Thickness of thin film with different molar ratio precursors.

PbBr-CsBr ratio 1:1 1:1.2 1:1.5 1:2 Thickness 130 ± 20nm 90 ± 15nm 120 ± 15nm 80 ± 10nm

The TEM image in Figure 6B supports that there are dispersed nanocrystals within the matrix of the larger host grain. Analysis of the images of multiple host micro- grains leads to an average 11.6 nm diameter of the nano-crystal inclusions. Regarding the micrograins shown as disks in the TEM image of Figure 6-17b, and assuming that the dark inclusions are also disks extending through the full thickness of the micro-disk, we obtain from multiple grain analyses an upper estimate of the fractional volume of the inclusion phase as 12%. But if the nano-inclusions regarded as CsPbBr3 extend all the way to the surfaces of the micro-disk, they should not have withstood the DMSO

148 washing. In a model of nano-inclusions that are completely encapsulated by the less soluble Cs4PbBr6 host, an assumption that the inclusions have half the thickness of this micro-grain in the TEM photo leads to an estimate of 6% fraction of the volume of the host micro-grain which is CsPbBr3 nano-inclusions. Since the unit cell of Cs4PbBr6 is considerably larger than that of CsPbBr3, this volume fraction corresponds to a molar fraction of 14%. This is close to the fraction deduced by taking the ratio of XRD resolvable peak height at the position of the [112] peak in the (1:2) film relative to the

(1:1) film. It is in line with ~15% molar fraction in the CsPbBr3/Cs4PbBr6 nanocomposite reported by Quan et al181. The absorption spectrum between 340 nm and 530 nm for all 4 films is characteristically that of CsPbBr3, notably the exciton peak at 520 nm with interband absorption extending to shorter wavelength until the transparency gap of Cs4PbBr6 ends at about 320 nm as seen in the (1:2) and (1:1.5) films. If the strength of the CsPbBr3 absorption spectrum were simply proportional to the number of unit cells of

CsPbBr3 in a unit volume of film, then the nearly 100 % CsPbBr3 film (1:1) should be several times off scale compared to the (1:2) film that is plotted. Instead, they are of comparable strength in the measured data.

It is useful to define an absorption cross section σ per unit cell (of CsPbBr3 in this case) where

푑퐼 휎(푐푚2)푁(푐푚−3) = − = 훼 푑푧

Eqn. (6-3) with N unit cells per unit volume. We will label properties of the (1:1) and (1:2) films as

1 and 2 respectively. From comparison of the absorption strengths of the characteristic

CsPbBr3 spectral features,

149

훼 휎 푁 1 ≈ 1 = 1 1 훼2 휎2 푁2

Eqn. (6-4)

푁1 With the estimate already made of in the (1:2) sample, we conclude that σ2 ≈ 16 σ1 in 푁2 that case. There is apparently something about the arrangement of the minority of

CsPbBr3 unit cells in the composite film 2 that conveys greater absorption strength per unit cell compared to bulk CsPbBr3 (film 1). Absorption cross section scales with oscillator strength. There are at least four well-established phenomena that can modify the oscillator strength per unit cell when the material is a nanocrystal and/or is imbedded in a dielectric. We review them briefly below:

(1) The local effective field produced by a dielectric medium surrounding a transition dipole which is atomic in size with oscillator strength f0 in isolation modifies the absorption coefficient α (and cross section per unit cell σ) according to the Smakula equation which can be written

(푛2+2)2 9푚푐 훼 9푚푐 푓 = 푊 = 푊휎 0 푛 2푒2 푁 2푒2

Eqn. (6-5) where n is the surrounding refractive index and W is the transition band width. A similar effective field dependence of the radiative lifetime of an atomic dipole transition was

208 3+ presented by Meltzer et al and compared to experimental shortening of the Eu : Y2O3 lifetime as the index n of the surrounding medium was increased. If we regard the

(푛2+2)2 product 푓 as an effective oscillator strength of the immersed atomic dipole, then 0 푛 oscillator strength is enhanced by increasing the index of the host medium according to this mechanism. However, the transition of interest in CsPbBr3 is an exciton in a

150 semiconductor rather than an atom in a host. The exciton’s internal coordinates as well as its external (translational) coordinates are affected by spatial confinement and/or surrounding dielectric constant. Takagahara has treated this situation in detail.209,210

Takagahara divided the phenomena responsible for oscillator strength modification into two categories: quantum confinement and dielectric confinement. His treatment of dielectric confinement focused on modification of the exciton wavefunction by the dielectric interface at the nanocrystal boundary, and interestingly the resulting

휖 dependence of oscillator strength on the ratio 1 of dielectric functions in the 휖2 semiconductor nanocrystal (ε1) and surrounding dielectric medium (ε2) is opposite in direction to the effective field phenomenon discussed in the paragraph above and measured for imbedded compact atomic dipoles e.g. by Meltzer et al208. Takagahara showed that lowering the surrounding dielectric constant with respect to a given semiconductor generally increases the oscillator strength of excitons in an imbedded nanocrystal.

(2) Quantum confinement increases electron-hole overlap in the exciton and therefore increases oscillator strength, without invoking dielectric constant.

(3) The main part of Takagahara’s theory concerns dielectric confinement as he defined it, which includes effects both on the internal exciton coordinates governing electron-hole overlap and on the translational coordinates governing spatial extent of the exciton wavefunction (coherence over multiple unit cells). For partly historical reasons going back to roots of “giant oscillator strength” in defect-bound excitons in semiconductors, we will separate those two aspects in the brief discussion following. In

Takagahara’s treatment, the image charge modification of electric field lines intersecting

151 the dielectric boundary modifies electron-hole correlation and thus the e-h overlap. The calculated results establish that this effect becomes stronger if the interface transition is toward lower index of the surrounding medium.

(4) Translational extent of the exciton wavefunction and associated coherence volume – The phenomenon of so-called giant oscillator strength was discovered decades ago for defect bound excitons in semiconductors. The corresponding oscillator strength enhancement is proportional to the volume coherently linked by center-of-mass translation of the exciton. Rashba and Gurgenishvili211, first advanced (and named) the giant oscillator strength theory to account for anomalously strong absorption by a small number of donor-bound excitons. Henry and Nassau extended the GOS or exciton coherence volume theory to account for 500-picosecond-scale spontaneous radiative emission rates of bound excitons in CdS.212 Itoh et al213 and Nakamura et al214 extended its applicability to describe experiments in semiconductor nanocrystals. Nikl et al186 first suggested its role in explaining the very short exciton radiative lifetime in CsPbBr3 nanocrystals precipitated in melt-grown CsBr host. Takagahara’s general theory of dielectric confinement in semiconductor quantum dots includes this contribution to enhanced oscillator strength associated with exciton translational motion and increasing with crystallite volume. His treatment showed in addition the interaction between the

휀1 dielectric ratio and the magnitude of GOS. Smaller surrounding ε2 increases the GOS 휀2 effect for a given semiconductor ε1. Effective mass ratio and exciton Bohr radius relative to nanocrystal radius are also factors.

A goal of follow-on work with this material system will be to control and characterize crystallite size and shape well enough to test scaling of oscillator strength

152 with crystallite volume. The comparison will also require more complete spectral dielectric function data and effective mass data for the host Cs4PbBr6 in addition to

CsPbBr3. As we suggested earlier in this paper, engineering of GOS for radiative dipole transitions might be one route to help radiative processes compete with Auger.

153

Chapter 7 Cs4PbBr6 Single Crystal with CsPbBr3 Nano-Inclusions

Melt-grown single-crystal CsPbBr3 is under development, e.g. at Argonne and

Lawrence Berkeley Labs, as a semiconductor spectroscopic radiation detector, measuring induced current in a biased crystal proportional to the energy of a fully absorbed gamma ray.215 The alternative approach for spectroscopic gamma ray detectors is scintillation.

Each approach has its advantages. But melt-grown CsPbBr3 and indeed any spatially extended bulk crystal or pure powder of CsPbBr3 exhibits very low luminescence efficiency, so scintillation from the pure bulk semiconductor had never seemed a possibility. Recently in work directed mainly toward photo- and electro-luminescence,

181,216 we showed along with others that CsPbBr3 nano-inclusions in a wider-gap Cs4PbBr6 host achieve photoluminescence quantum yield up to 70% also reported up to 90% for higher CsPbBr3 loading. Using sub-picosecond laser pulse excitation, we explained the different kinetics of extended and nano-inclusion CsPbBr3 and how that accounts for the extraordinary quantum yield as well as the subnanosecond lifetime of bright green luminescence in this solution-grown nanocomposite. Photoluminescence of the nanocomposite exhibits 345 ps exponential decay, suggesting potential for fast trigger time-of- flight applications in gamma-ray detection and localization. As described here, we have achieved the first macroscopic solution-grown crystals of the Cs4PbBr6 host with

CsPbBr3 nano-inclusions. This nano-inclusion composite in a crystalline host is entirely composed of high-Z, earth-abundant elements and it can be grown from solution at room temperature. We explore suitability of this nanocomposite as a fast, dense, possibly low- cost scintillation detector of ionizing radiation with especially promising potential as a fast trigger for time-of-flight spatial localization.

154

Growth of Cs4PbBr6 Single Crystal imbedded with CsPbBr3 Nanoinclusions

Figure 7-1 | Steps for CsPbBr3/ Cs4PbBr6 crystal growth.

The Cs-Pb-Br precursor solution is prepared by dissolving CsBr and PbBr2 powders into DMSO solvent in Figure 7-1. The CsBr - PbBr2 molar ratio plays a main role of determining crystal phase (CsPbBr3 or/and Cs4PbBr6) in thin films and micro scale crystals based on previous reports. Excessive CsBr in the precursor is helpful to produce large size and high quality single crystals. By exposing precursor solution

(optimized 0.2M) in anti-solvent vapor (acetone) at room temperature, Cs-Pb-Br crystal nuclei are quickly formed. In 24 hours, clear and hexagonal crystals were found and carefully collected from the bottom of crystal growth beaker (Figure7-2a). The resultant crystals have an average size of 3 mm with the largest so far about 6 mm in diameter

(Figure 7-2b). The crystals have a light yellow tint in room light. The body of the crystal has a smooth surface suggesting facets (Figure 7-2c) and a fairly clear bulk transparency. We have demonstrated that the crystals can be grown in a relatively large quantity from 3ml of precursor in one batch without high pressure or high temperature, which are particularly important factors for a low cost method for potential applications.

155

(a) (b)

(c)

Figure 7-2 | Morphology of the CsPbBr3/ Cs4PbBr6 crystals.

Figure 7-3a shows the fluorescent image of the Cs-Pb-Br single crystals with

365nm excitation, which emit intensive green luminescence. Figure 7-3b shows more details of one Cs-Pb-Br single crystal under an optical microscope. The entire crystal body emits macroscopic uniform green fluorescence in the wavelength of 520nm which is corresponding to the 2.4eV bandgap of cubic CsPbBr3. PL quantum yield of the crystal is measured by integration sphere with a UV-blue LED about 26.7% (Figure 7-4).

156

(a) (b)

Figure 7-3 | Fluorescent images of CsPbBr3/ Cs4PbBr6 crystals.

Figure 7-4 | Molar ratio influence on quantum yield of CsPbBr3/ Cs4PbBr6 crystals

In Figure 7-5, the LSM was employed to map luminescent locations of the crystal. In order to get a wide view of the entire crystal body, a small size crystal (~ 10 um) was chosen to image. The laser excitation wavelength was set to 420 nm and detection wavelength of window filter was tuned to 520 nm which is aligned with the peak of luminescence of the crystal. The focal plane was raised from the bottom of crystal (Z = 0.86 μm) to the top of the crystal (Z = 6.02 μm) in a minimum increment of

0.86 μm (limited by the instrument) for a compete Z-scan. Each fluorescent image was

157 overlapped with the corresponding transmittance image to confirm where the luminescent areas were inside the crystal. All transmittance images show an identical hexagonal shape which is consistent with large crystal shape. Near the crystal bottom (Figure 7-

5a), most of the area of the slice was covered by the green luminescence. The luminescent area decreases with the increase of the Z distance while didn’t trespass lager hexagonal boundaries. In a higher Z scanning (Figure 7-5e), the active area starts to spread into connecting dots which is unable to be observed in Figure 7-5h when the camera is saturated by high intensity luminescence. These observations suggest that the luminescence of the crystal is due to the dot-wise inclusions instead the matrix crystal.

Figure 7-5 | Confocal microscopy images of CsPbBr3/ Cs4PbBr6 crystal. Scale bar is 5 μm.

158

Compositional analysis of the nano-composite

Figure 7-6 | (a) Powder X ray diffraction patterns of the ground Cs4PbBr6 host crystals imbedded with ~5% CsPbBr3 nanocrystals. (b) Crystal structures of Cs4PbB6 and CsPbBr3. (c) HRTEM images of nanocomposite.

The XRD measurement was first carried out for the crystal to identify the lattice structure of the composite. Please note that the crystal was grounded into powder for XRD measurement due to the strong x-ray absorption of the single crystal. As seen in Figure

7-6a, the XRD peaks are in a good agreement with calculated Cs4PbBr6 trigonal structure. However, we did find the unaligned peaks as the red arrows mark, which fits the orthorhombic structure of CsPbBr3 phase. The HRTEM images collected from the composite sample (Figure 7-6c) show that there are CsPbBr3 nano-inclusion imbedded into Cs4PbBr6 matrix just as the thin films we showed in last chapter. The size pf the

CsPbBr3 nanocrystals is about 9~10 nm in average. We used the Reference Intensity

159

Ratio (RIR) method to quantitatively analyze the weight ratio of CsPbBr3 (noted as phase a) and Cs4PbBr6 (noted as phase b). The diffraction data of both CsPbBr3 and Cs4PbBr6 were scaled to the standard Corundum as a factor which is Intensity Analyte over

Intensity Corundum (I/Ic) at strongest peaks. In Crystallography Open Database, the I/Ic values for CsPbBr3 and Cs4PbBr6 are 6.08 and 5.11 respectively. In our experiment, the intensity ratio (Ia/Ib) of the strongest lines between CsPbBr3 (30.44°) and Cs4PbBr6

(21.59°) is about 13.48 from the XRD result in Figure 2A. By applying the equation,

IX()II a ca a IIIX() bc b b

Eqn. (7-1)

the weight ratio of the two compounds (Xa/Xb) can be calculated, which indicates that the weight fraction of CsPbBr3 nano-inclusions in the composite is about 5.9%.

Figure 7-7 | EDS images of Cs-Pb-Br nanocomposite.

We also show the EDS mapping of the thinned sample in Figure 7-7. The spacemen is uniformly rich in three elements: Pb, Br, and Cs which we certainly expect.

160

Exction Kinetics and Recombination Analysis

Figure 7-8a shows the PL spectrum of the CsPbBr3/Cs4PbBr6 crystal at 520nm with a

FWHM of 23.1 nm which is quite consistent with emission CsPbBr3 quantum dots in the literatures. The absorbance of the crystal (Figure 7-8b) shows a clear band edge at

2.35eV and second band edge at 3.87eV which are attributed to CsPbBr3 and Cs4PbBr6 respectively. The relative instability of absorption signal around 290nm is because dominant Cs4PbBr6 phase strongly absorbs in this range reaching the upper limit of the equipment. In the same figure, under 290 nm excitation, not only the emission of

161

Cs4PbBr6 host was seen between 350nm and 400nm but also a narrow photoluminescence band was detected peaking at 520 nm as CsPbBr3 quantum dots. One possibility is the

290 nm UV light excited the CsPbBr3 inclusions directly leading to a radiative recombination. However, the contribution of energy transfer from Cs4PbBr6 host to

CsPbBr3 cannot be ignored considering the inner band gap of CsPbBr3 which could potentially cause an efficient energy transfer from host to inclusion. In Figure 7-8c, we investigated the excitation spectrum of the crystal while the emission is fixed at 520nm.

It suggests that the 520nm emission is relevant to both Cs4PbBr6’s and CsPbBr3’s intrinsic absorptions. The additional peak appeared at 295nm (in Figure 7-8c) aligning with the absorption peak of Cs4PbBr6 (in Figure 7-8b) could be a sign of conversion down to 2.35 eV.

(a) (b)

Figure 7-9 | Streak images of CsPbBr3 nanocrystals in Cs4PbBr6 crystal under 420nm laser excitation in a 300fs pulse in a different of fluences: 26.8 μJ/cm2 and 0.768 μJ/cm2.

162

Figure 7-10 | Fitting analysis on exciton dynamics of CsPbBr3 nanocrystals in Cs4PbBr6 crystal in a function of laser fluence.

By generating excitons in the CsPbBr3 nanoinclusions only (via photoexcitation using photon energies lower than Cs4PbBr6 host band gap, ~ 2.95 eV), we investigate carrier dynamics confined in CsPbBr3 nanocrystals. The steak images in Figure 7-9 shows that

CsPbBr3 emission was enhanced with a stronger laser pulse. Meanwhile, the decay

2 lifetime of CsPbBr3 fluorescence significantly decreased with 0.768 8μJ/cm intensity as shown in Figure 7-10. The decay curves under different excitation densities were fitted by 1st order kinetics with initial 3rd order kinetics (Auger recombination). 2nd order kinetics was unable to provide a satisfactory fit. The fittings in the nanocomposite differ with that in extended CsPbBr3 crystal where the PL decay can be only fitted by 2nd order bimolecular kinetics due to dissociation of excitons at room temperature. The large population of electrons and holes under high excitation promotes the Auger recombination in CsPbBr3 nano-inclusions. More power dependence can be found in

Figure 7-11.

163

(a) (b) Figure 7-11 | Power dependent analysis..

Table 5 The PL decay effective lifetime (t 1/2) under 420nm 300fs laser pulse excitation. First order (k1) and Third order Auger recombination constant (k3).

Intensity -1 6 -1 t 1/2 (ps) t 1/2 fitting Relative 2 k1 (s ) k3 (cm s ) n0 2 (uJ/cm ) =1.5/(k3*n0 ) (ps) Lumeffic 26.8 3.0E+08 3.5E-27 1.38E+18 225 207 0.57 4.26 2.3E+08 7E-26 2.2E+17 443 403 0.71 1.85 1.6E+08 2.5E-25 9.5E+16 667 561 0.82 0.768 1.5E+08 7.5E-25 3.94E+16 1293 952 0.85

The decay curves under different excitations were fitted by 1st order kinetics mainly with additional 3rd order kinetics, as black curves in Figure 7-10. Please note that 2nd order kinetics is not satisfactory, which rules out the dominant contribution of dipole-dipole biexcitons. The first order exponential recombination constant was found to be in the order of 108 s-1 in the slower later decay with a slight drop with lower excitation intensities

(Table 1). This has been attributed to the quantum confinement effects in such nanoscale system. The physical confinement favors localization of carriers, hereby enforcing geminate e-h recombination despite the weak exciton binding energy of CsPbBr3 at room temperature and consequently producing 1st order kinetics. The constant luminescence efficiency at low intensities is geminate in its nature and an increased efficiency is

164 anticipated at higher intensity due to nongeminate recombination. However, we failed to observe an improved efficiency with intensity from our results. This may be because when the bleaching of absorption sets in, the luminescence efficiency decreases because the incident photons simply pass through the host without further exciting CsPbBr3 nanocrystals. Considering power-dependent decay lifetime (Figure 7-11), the additional contribution of Auger recombination should be considered. The effect of Auger recombination is to increase the nonradiative recombination at high intensities and, therefore, to decrease the initial carrier density which decreases the bimolecular radiative decay lifetime until it meets the Auger decay lifetime limit (~ 200ps). At laser intensity of 26.8 uJ/cm2 (in high intensity range), 1.38 X 1018 e-h pairs per cm3 are generated in

CsPbBr3. Putting this initial carrier density into the 3rd order recombination rate equation

-27 6 -1 with rate constant k3 = 3.5 X 10 cm s , we obtain a model prediction of effective initial time constant t1/2 = 1.5/(k3*n2) = 225 ps for the conditions of our experiment. The observed effective initial lifetime is 207 ps showing an agreement with the prediction.

With decrease of the generated carriers, the effective initial constants are also close to experimental effective initial lifetimes, which are summarized in Table 5. This agreement with the prediction based on excitation density and estimated rate constant suggests that the Auger recombination model properly describes the multi-exciton effects in CsPbBr3 nanocrystals.

165

Figure 7-12 | CsPbBr3/Cs4PbBr6 crystals sealed in PDMS.

(a) (b) Figure 7-13 | X-ray luminesence spectrum and decay of the x-ray luminescence.

To understand the energy transfer, we studied the x-ray luminescence of the

CsPbBr3/Cs4PbBr6 nanocomposite in collaborative experiments with Federico Moretti at

Lawrence Berkeley National Laboratory. Since the 3mm crystal is difficult to align the

166 beam and collect enough signal, the sample we prepared an assemby of several separated crystals sealed by PDMS transparent epoxy as shown in Figure 7-12. The spectrum of luminescence under x-ray excitation easured at LBNL is exhibited in Figure 7-13a.

Regardless of the background noise, there are two main peaks at the wavelength of

340nm and 520nm which are due to de-excitation of electrons in Cs4PbBr6 and CsPbBr3 respectively. The 520nm peak is 10 times higher than the 340nm peak while volume percentage of the CsPbBr3 nanocrystals is only 5%. This indicates the possibility of energy flow from Cs4PbBr6 to CsPbBr3. However, we are unable to provide the energy transfer efficiency at the Type I heterjunction at the moment due to the missing of quantum efficiency of Cs4PbBr6. Figure 7-13b shows the decay of the CsPbBr3 emission measured at LBNL using a picosecond pulsed x-ray source the time resolved spectrometer (Hamamatsu). The data can be fitted to a tri-exponential function. The first component of the decay has a lifetime of 107 ps, which is attributed to ultrafast recombination of excitons in the CsPbBr3 confinement. The decay constants increases significantly in the second (764ps) and third (15.5ns) components. The delayed process could very likely be related with the exciton transfer over the heterjunction which normally takes extra time. The type of exciton transfer (dipole-dipole transfer, radiative transfer, or exciton migration) still remains unclear and more experiments.

167

CsPbBr3 Phase Evolution in Host Matrix

Figure 7-14 | Snapshots of CsPbBr3/Cs4PbBr6 nanocomposite sample under UV lamp within 1 hour.

To form the CsPbBr3/Cs4PbBr6 nanocomposite, we were dripping precursor solution into chlorobenzene (serving as anti-solvent). The fact interests us is that there is almost no luminescence with UV excitation in the first 2 minutes of the synthesis, whereas there is strong luminescence found after 1 hour. So we recorded the revolution of emissive intensity in Figure 7-14 which shows the brightness increasing over time.

The corresponding PL spectrum is shown in Figure 7-15. The intensity of the emission dramatically increases within 1 hour of precipitating the microcrystals of Cs4PbBr6. We attribute this to the time for growth CsPbBr3 NCs (emissive centers) inside of the mocrocrystals. Besides that, a slight red shift was noticed while the intensity of the peak was increasing, which could be linked to the CsPbBr3 NCs growth in size at room temperature. The size of CsPbBr3 NCs determines the quantum confinement of excitons trapped. Oscillator strength enhancement (GOS) initially grows in proportion to the

168 volume of a nanocrystal (the coherence volume)Therefore, the red-shift and improved brightness is the result of the growth of the CsPbBr3 NCs in terms of size and density, as depicted in Figure 7-16.

Figure 7-15 | Time dependent PL spectrum of CsPbBr3/Cs4PbBr6 nanocomposite.

Figure 7-16 | Scheme of CsPbBr3 nanocrystal growth in Cs4PbBr6 matrix.

On the other hand, microcrystallites of CsPbBr3 were also found to grow in the powder over at least 2 minutes after precipitation of the CsPbBr3/Cs4PbBr6 nanocomposite material. The transient optical images of this phase transition is shown in Figure 7-17.

From the XRD result in Figure 7-18 on the final precipitation, the peaks due to both

CsPbBr3 and Cs4PbBr6 were detected and fitted with the calculated patterns which confirmed the mixture of the two phases.

169

Figure 7-17 | Optical images of phase transition within 10 min. Scale bar is 50μm.

Figure 7-18 | XRD pattern of Cs4PbBr6 and CsPbBr3 microcrystals mixture.

170

This phase transition may be attributed to the continuous growth of CsPbBr3 crystals from nanoscale to microscale (Figure 7-19). The promotion of CsPbBr3 microcrystals negatively impacts the quantum efficiency of the sample since the electron- hole recombination rate is reduced in the extended CsPbBr3 crystals, ad discussed earlier.

Optical properties of the extended CsPbBr3 has been studied in Chapter 6.

Figure 7-19 | Phase transition between CsPbBr3/Cs4PbBr6 nanocomposite and CsPbBr3 microcrystals.

In future work, this nanocomposite could be extended to the application of a dense, fast, cheap scintillator, possibly cast as powder suspended in plastic. Aside from general transparency noted as desirable above, there is a particular challenge to get scintillation light out of a composite in which the emitted light is near the band edge of the recombination islands. Some of our preliminary work have been done by mixing dye molecules and CsPbBr3/Cs4PbBr6 nanocomposite as shown in Figure 7-20 and 7-21. All of these issues need investigation to assess potential of this material system as a spectroscopic scintillator, and this just makes a start.

171

(a)

(b)

172

(a)

(b) Figure 7-21 | (a) Photos of a series of the samples mixing the different ratios of dye molecules and CsPbBr3/Cs4PbBr6 nanocomposite in room light and 365nm UV. The ratio of the dye molecules and the CsPbBr3/Cs4PbBr6 nanocomposite increases from sample#1 to #4. (b) The fluorescent spectra of the mixture samples.

173

Reference

[1] Pope, M.., Swenberg, C., Electronic Processes in Organic Crystals and Polymers,

2ed ed., Oxford University Press (1999).

[2] Karl, N., Festkörperprobleme: Advanced Solid State Physics, H. J. Queisser, Ed.,

Pergamon/Vieweg, Braunschweig (1974).

[3] Neumann, F., Genenko, Y. A., Schmechel, R.., Seggern, H. von., “Role of

diffusion on SCLC transport in double injection devices,” Synth. Met. 150(3),

291–296 (2005).

[4] Ashcrof, N. W.., Mermin, N. D., Solid State Physics, Holt, Rinehart and Winston,

New York (1976).

[5] Miller, A.., Abrahams, E., “Impurity Conduction at Low Concentrations,” Phys.

Rev. 120(3), 745–755, American Physical Society (1960).

[6] Gutmann, F.., Lyons, L. E., Organic Semiconductors, Wiley, New York (1967).

[7] Brown, A. R., Pichler, K., Greenham, N. C., Bradley, D. D. C., Friend, R. H..,

Holmes, A. B., “Optical spectroscopy of triplet excitons and charged excitations in

poly(p-phenylenevinylene) light-emitting diodes,” Chem. Phys. Lett. 210(1–3),

61–66 (1993).

[8] Scott, J. C., Brock, P. J., Salem, J. R., Ramos, S., Malliaras, G. G., Carter, S. A..,

Bozano, L., “Charge transport processes in organic light-emitting devices,” Synth.

Met. 111, 289–293 (2000).

[9] Bässler, H., Tak, Y. H., Khramtchenkov, D. V., Nikitenko, V. R., “Models of 174

organic light-emitting diodes,” Synth. Met. 91(1), 173–179 (1997).

[10] Hill, I. G., Kahn, A., Soos, Z. G.., Pascal, Jr, R. A., “Charge-separation energy in

films of π-conjugated organic molecules,” Chem. Phys. Lett. 327(3–4), 181–188

(2000).

[11] Kohler, A.., Bassler, H., “Triplet states in organic semiconductors,” Mater. Sci.

Eng. R Reports 66(4–6), 71 (2009).

[12] Malliaras, G. G.., Scott, J. C., “The roles of injection and mobility in organic light

emitting diodes,” J. Appl. Phys. 83(10), 5399 (1998).

[13] Albrecht, U.., Bässler, H., “Efficiency of charge recombination in organic light

emitting diodes,” Chem. Phys. 199(2), 207–214 (1995).

[14] Atkins, P., Paula, J.., Keeler, J., Atkins’ Physical Chemistry, 11th ed., Oxford

University Press (2017).

[15] Turro, N. J., Modern Molecular Photochemistry, University Science Books,

Sausalito (1991).

[16] Adachi, C., Baldo, M. a., Thompson, M. E.., Forrest, S. R., “Nearly 100% internal

phosphorescence efficiency in an organic light-emitting device,” J. Appl. Phys.

90(10), 5048 (2001).

[17] Kawamura, Y., Goushi, K., Brooks, J., Brown, J. J., Sasabe, H.., Adachi, C.,

“100% phosphorescence quantum efficiency of Ir(III) complexes in organic

semiconductor films,” Appl. Phys. Lett. 86(7), 71104 (2005).

[18] Wu, Y., Xu, Z., Hu, B.., Howe, J., “Tuning magnetoresistance and magnetic-field-

175

dependent electroluminescence through mixing a strong-spin-orbital-coupling

molecule and a weak-spin-orbital-coupling polymer,” Phys. Rev. B 75(3), 35214

(2007).

[19] Hu, B., Yan, L.., Shao, M., “Magnetic-Field Effects in Organic Semiconducting

Materials and Devices,” Adv. Mater. 21(14–15), 1500–1516 (2009).

[20] Hu, B.., Wu, Y., “Tuning magnetoresistance between positive and negative values

in organic semiconductors.,” Nat. Mater. 6(12), 985–991 (2007).

[21] Demtröder, W., Laser Spectroscopy : Basic Concepts and Instrumentation,

Wolfgang Demtröder (1981).

[22] Hilborn, R. C., “Einstein Coefficients, Cross Sections, f Values, Dipole Moments,

and All That,” Am. J. Phys. 50(982) (1982).

[23] Elliott, R. J., “Intensity of Optical Absorption by Excitons,” Phys. Rev. 108(6),

1384–1389, American Physical Society (1957).

[24] Broude, V. L., Eremenko, V. V.., Rashba, É. I., “The Absorption of Light by CdS

Crystals,” Sov. Phys. Dokl. 2, 239 (1957).

[25] Rashba, E. I., “Theory of The Impurity Absorption of Light in Molecular

Crystals,” Opt. Specktrosk 2, 568–577 (1957).

[26] Rashba, E. I.., Gurgenishvili, G. E., “To The Theory of The Edge Absorption in

Semiconductors,” Sov. Phys. - Solid State 4, 759–760 (1962).

[27] Rashba, E. I., “Giant Oscillator Strengths Associated with Exciton Complexes,”

Sov. Phys. Semicond. 8, 807–816 (1975).

176

[28] Demic, S., Ozcivan, A. N., Can, M., Ozbek, C.., Karakaya, M., “Recent Progresses

in Perovskite Solar Cells,” N. B. T.-N. S. C. Das, Ed., InTech, Rijeka, Ch. 13

(2017).

[29] Tan, Z.-K., Moghaddam, R. S., Lai, M. L., Docampo, P., Higler, R., Deschler, F.,

Price, M., Sadhanala, A., Pazos, L. M., et al., “Bright light-emitting diodes based

on organometal halide perovskite,” Nat. Nanotechnol. 9(August), 1–6, Nature

Publishing Group (2014).

[30] Protesescu, L., Yakunin, S., Bodnarchuk, M. I., Krieg, F., Caputo, R., Hendon, C.

H., Yang, R. X., Walsh, A.., Kovalenko, M. V., “Nanocrystals of Cesium Lead

Halide Perovskites (CsPbX3, X = Cl, Br, and I): Novel Optoelectronic

Materials Showing Bright Emission with Wide Color Gamut,” Nano Lett. 15(6),

3692–3696 (2015).

[31] Zhang, F., Zhong, H., Chen, C., Wu, X. G., Hu, X., Huang, H., Han, J., Zou, B..,

Dong, Y., “Brightly luminescent and color-tunable colloidal

CH3NH3PbX3 (X = Br, I, Cl) quantum dots:

Potential alternatives for display technology,” ACS Nano 9(4), 4533–4542 (2015).

[32] Rainoì, G., Nedelcu, G., Protesescu, L., Bodnarchuk, M. I., Kovalenko, M. V.,

Mahrt, R. F.., Stöferle, T., “Single Cesium Lead Halide Perovskite Nanocrystals at

Low Temperature: Fast Single-Photon Emission, Reduced Blinking, and Exciton

Fine Structure,” ACS Nano 10(2), 2485–2490 (2016).

[33] Sichert, J. A., Tong, Y., Mutz, N., Vollmer, M., Fischer, S., Milowska, K. Z.,

García Cortadella, R., Nickel, B., Cardenas-Daw, C., et al., “Quantum Size Effect

177

in Organometal Halide Perovskite Nanoplatelets,” Nano Lett. 15(10), 6521–6527

(2015).

[34] Huang, H., Susha, A. S., Kershaw, S. V., Hung, T. F.., Rogach, A. L., “Control of

Emission Color of High Quantum Yield CH 3 NH 3 PbBr 3 Perovskite Quantum

Dots by Precipitation Temperature,” Adv. Sci. 2(9), 1500194 (2015).

[35] Levchuk, I., Osvet, A., Tang, X., Brandl, M., Perea, J. D., Hoegl, F., Matt, G. J.,

Hock, R., Batentschuk, M., et al., “Brightly Luminescent and Color-Tunable

Formamidinium Lead Halide Perovskite FAPbX3 (X = Cl, Br, I) Colloidal

Nanocrystals,” Nano Lett. 17(5), 2765–2770, American Chemical Society (2017).

[36] Liang, K., Mitzi, D. B.., Prikas, M. T., “Synthesis and Characterization of

Organic−Inorganic Perovskite Thin Films Prepared Using a Versatile Two-Step

Dipping Technique,” Chem. Mater. 10(1), 403–411, American Chemical Society

(1998).

[37] Tang, C. W.., VanSlyke, S. A., “Organic electroluminescent diodes,” Appl. Phys.

Lett. 51(12), 931 (1987).

[38] Friend, R. H., Gymer, R. W., Holmes, A. B., Burroughes, J. H., Marks, R. N.,

Taliani, C., Bradley, D. D. C., Lo, M., Salaneck, W. R., et al.,

“Electroluminescence in conjugated polymers,” Nature 397, 121–128 (1999).

[39] Shen, Z., Burrows, P. E., Bulović, V., Forrest, S. R.., Thompson, M. E., “Three-

color, tunable, organic light-emitting devices,” Science (80-. ). 276(5321), 2009–

2011 (1997).

178

[40] Baldo, M. A.., Forrest, S. R., “Highly efficient phosphorescent emission from

organic electroluminescent devices,” Nature 395(September), 151–154 (1998).

[41] Oehzelt, M., Koch, N.., Heimel, G., “Organic semiconductor density of states

controls the energy level alignment at electrode interfaces.,” Nat. Commun.

5(May), 4174, Nature Publishing Group (2014).

[42] Helander, M. G., Wang, Z. B., Qiu, J.., Lu, Z. H., “Band alignment at

metal/organic and metal/oxide/organic interfaces,” Appl. Phys. Lett. 93(May

2009), 2008–2010 (2008).

[43] Aziz, H., Popovic, Z., Hu, N.-X., Hor, A.-M.., Xu, G., “Degradation Mechanism of

Small Molecule – Based Organic Light-Emitting Devices,” Science (80-. ).

283(19), 1900–1903 (2008).

[44] Xu, J., Zhang, L., Zhong, J.., Lin, H., “High-luminance flexible top emission

organic light emitting diode with nickel-chromium alloy thin film on aluminum

anode,” Opt. Rev. 19(6), 358–360, Optical Society of Japan (2012).

[45] Meerheim, R., Furno, M., Hofmann, S., Lüssem, B.., Leo, K., “Quantification of

energy loss mechanisms in organic light-emitting diodes,” Appl. Phys. Lett. 97,

25–28 (2010).

[46] Scott, J. C., “Metal–organic interface and charge injection in organic electronic

devices,” J. Vac. Sci. Technol. A 21(3), 521 (2003).

[47] Huang, J., Pfeiffer, M., Werner, A., Blochwitz, J., Leo, K.., Liu, S., “Low-voltage

organic electroluminescent devices using pin structures,” Appl. Phys. Lett. 80(1),

179

139 (2002).

[48] Koch, N., “Organic electronic devices and their functional interfaces,”

ChemPhysChem 8, 1438–1455 (2007).

[49] Baldo, M.., Forrest, S., “Interface-limited injection in amorphous organic

semiconductors,” Phys. Rev. B 64, 85201 (2001).

[50] Perumal, A., Fröbel, M., Gorantla, S., Gemming, T., Lüssem, B., Eckert, J.., Leo,

K., “Novel Approach for Alternating Current (AC)-Driven Organic Light-Emitting

Devices,” Adv. Funct. Mater. 22(1), 210–217 (2012).

[51] Fröbel, M., Schwab, T., Kliem, M., Hofmann, S., Leo, K.., Gather, M. C., “Get it

white: color-tunable AC/DC OLEDs,” Light Sci. Appl. 4(September 2014), e247

(2015).

[52] Chen, Y., Xia, Y., Smith, G. M.., Carroll, D. L., “Frequency-Dependent,

Alternating Current-Driven, Field-Induced Polymer Electroluminescent Devices

with High Power Efficiency,” Adv. Mater. 26, 8133–8140 (2014).

[53] Chen, Y., Xia, Y., Smith, G. M., Sun, H., Yang, D., Ma, D., Li, Y., Huang, W..,

Carroll, D. L., “Solution-Processable Hole-Generation Layer and Electron-

Transporting Layer: Towards High-Performance, Alternating-Current-Driven,

Field-Induced Polymer Electroluminescent Devices,” Adv. Funct. Mater. 24(18),

2677–1688 (2014).

[54] Chen, Y., Xia, Y., Sun, H., Smith, G. M., Yang, D., Ma, D.., Carroll, D. L.,

“Solution-Processed Highly Efficient Alternating Current-Driven Field-Induced

180

Polymer Electroluminescent Devices Employing High- k Relaxor Ferroelectric

Polymer Dielectric,” Adv. Funct. Mater. 24(11), 1501–1508 (2014).

[55] Cho, S. H., Sung, J., Hwang, I., Kim, R. H., Choi, Y. S., Jo, S. S., Lee, T. W..,

Park, C., “High performance AC electroluminescence from colloidal quantum dot

hybrids.,” Adv. Mater. 24(33), 4540–4546 (2012).

[56] Yamao, T., Shimizu, Y., Terasaki, K.., Hotta, S., “Organic light-emitting field-

effect transistors operated by alternating-current gate voltages,” Adv. Mater. 20,

4109–4112 (2008).

[57] Wood, V., Panzer, M. J., Bozyigit, D., Shirasaki, Y., Rousseau, I., Geyer, S.,

Bawendi, M. G.., Bulović, V., “Electroluminescence from nanoscale materials via

field-driven ionization,” Nano Lett. 11, 2927–2932 (2011).

[58] Zhang, L., Nakanotani, H., Yoshida, K.., Adachi, C., “Analysis of alternating

current driven electroluminescence in organic light emitting diodes: A comparative

study,” Org. Electron. 15(8), 1815–1821, Elsevier B.V. (2014).

[59] Sung, J., Choi, Y. S., Kang, S. J., Cho, S. H., Lee, T.-W.., Park, C., “AC field-

induced polymer electroluminescence with single wall carbon nanotubes.,” Nano

Lett. 11(3), 966–972 (2011).

[60] Perumal, A., Lüssem, B.., Leo, K., “High brightness alternating current

electroluminescence with organic light emitting material,” Appl. Phys. Lett.

100(10), 103307 (2012).

[61] Xu, J., Zhang, L., Lin, H.., Jiang, Q., “Nanoscale Control of Doped Organic Hole-

181

Injection Layer in Thermally Annealed Top Emission Organic Light Emitting

Diode,” Optoelectron. Adv. Mater. 7(11), 817–821 (2013).

[62] Liu, S.-Y., Chang, J.-H., -Wen Wu, I.., Wu, C.-I., “Alternating Current Driven

Organic Light Emitting Diodes Using Lithium Fluoride Insulating Layers,” Sci.

Rep. 4, 7559 (2014).

[63] Zhao, Y., Chen, R., Gao, Y., Leck, K. S., Yang, X., Liu, S., Abiyasa, A. P.,

Divayana, Y., Mutlugun, E., et al., “AC-driven, color- and brightness-tunable

organic light-emitting diodes constructed from an electron only device,” Org.

Electron. 14(12), 3195–3200, Elsevier B.V. (2013).

[64] Chen, Y., Smith, G. M., Loughman, E., Li, Y., Nie, W.., Carroll, D. L., “Effect of

multi-walled carbon nanotubes on electron injection and charge generation in AC

field-induced polymer electroluminescence,” Org. Electron. 14(1), 8–18, Elsevier

B.V. (2013).

[65] Chen, Y., Xia, Y., Smith, G. M., Gu, Y., Yang, C.., Carroll, D. L., “Emission

characteristics in solution-processed asymmetric white alternating current field-

induced polymer electroluminescent devices,” Appl. Phys. Lett. 102(1), 13307

(2013).

[66] Fröbel, M., Hofmann, S., Leo, K.., Gather, M. C., “Optimizing the internal electric

field distribution of alternating current driven organic light-emitting devices for a

reduced operating voltage,” Appl. Phys. Lett. 104(7), 71105 (2014).

[67] Wood, C.., Jena, D., Polarization Effects in Semiconductors (2008).

182

[68] Nie, W., Gupta, G., Crone, B. K., Liu, F., Smith, D. L., Ruden, P. P., Kuo, C.,

Tsai, H., Wang, H., et al., “Interface Design Principles for High-Performance

Organic Semiconductor Devices,” Adv. Sci. 2(6), 1500024 (2015).

[69] Katsenelenbaum, B. Z., High-frequency Electrodynamics (2006).

[70] Gao, X. D., Zhou, J., Xie, Z. T., Ding, B. F., Qian, Y. C., Ding, X. M.., Hou, X.

Y., “Mechanism of charge generation in p -type doped layer in the connection unit

of tandem-type organic light-emitting devices,” Appl. Phys. Lett. 93, 2–5 (2008).

[71] Kröger, M., Hamwi, S., Meyer, J., Dobbertin, T., Riedl, T., Kowalsky, W..,

Johannes, H. H., “Temperature-independent field-induced charge separation at

doped organic/organic interfaces: Experimental modeling of electrical properties,”

Phys. Rev. B 75, 235321 (2007).

[72] Terai, M.., Tsutsui, T., “Electric-field-assisted bipolar charge generation from

internal charge separation zone composed of doped organic bilayer,” Appl. Phys.

Lett. 90(2007), 83502 (2007).

[73] Parker, I. D., “Carrier tunneling and device characteristics in polymer light‐

emitting diodes,” J. Appl. Phys. 75(3), 1656 (1994).

[74] Braun, D.., Heeger, A. J., “Visible light emission from semiconducting polymer

diodes,” Appl. Phys. Lett. 58(18), 1982 (1991).

[75] Blom, P. W. M.., Jong, M. J. M. De., “Electrical characterization of polymer light-

emitting diodes,” IEEE J. Sel. Top. Quantum Electron. 4(1), 105–112 (1998).

[76] Bobbert, B. P. A., “Manipulating spin in organic spintronics,” Science (80-. ).

183

(2014).

[77] Pope, M., Kallmann, H. P.., Magnante, P., “Electroluminescence in Organic

Crystals,” J. Chem. Phys. 38(8) (1963).

[78] Baldo, M. A., Thompson, M. E.., Forrest, S. R., “Excitonic singlet-triplet ratio in a

semiconducting organic thin film,” Phys. Rev. B 60(20), 422–428 (1999).

[79] Forget, S.., Chénais, S., Organic Solid-State Lasers, Springer Berlin Heidelberg,

Berlin, Heidelberg (2013).

[80] Nguyen, T. D., Gautam, B. R., Ehrenfreund, E.., Vardeny, Z. V.,

“Magnetoconductance response in unipolar and bipolar organic diodes at

ultrasmall fields,” Phys. Rev. Lett. 105(16), 1–4 (2010).

[81] Bobbert, P., Nguyen, T., van Oost, F., Koopmans, B.., Wohlgenannt, M.,

“Bipolaron Mechanism for Organic Magnetoresistance,” Phys. Rev. Lett. 99(21),

216801 (2007).

[82] Mahato, R. N., Lülf, H., Siekman, M. H., Kersten, S. P., Bobbert, P. a., de Jong,

M. P., De Cola, L.., van der Wiel, W. G., “Ultrahigh magnetoresistance at room

temperature in molecular wires.,” Science 341(6143), 257–260 (2013).

[83] Prigodin, V. N., Bergeson, J. D., Lincoln, D. M.., Epstein, a. J., “Anomalous room

temperature magnetoresistance in organic semiconductors,” Synth. Met. 156(9–

10), 757–761 (2006).

[84] Janssen, P., Cox, M., Wouters, S. H. W., Kemerink, M., Wienk, M. M..,

Koopmans, B., “Tuning organic magnetoresistance in polymer-fullerene blends by

184

controlling spin reaction pathways.,” Nat. Commun. 4, 2286, Nature Publishing

Group (2013).

[85] Kalinowski, J., Szmytkowski, J.., Stampor, W., “Magnetic hyperfine modulation

of charge photogeneration in solid films of Alq3,” Chem. Phys. Lett. 378(3–4),

380–387 (2003).

[86] Wilkinson, J., Davis, a. H., Bussmann, K.., Long, J. P., “Evidence for charge-

carrier mediated magnetic-field modulation of electroluminescence in organic

light-emitting diodes,” Appl. Phys. Lett. 86(11), 111109 (2005).

[87] Müller, J., Lupton, J., Feldmann, J., Lemmer, U., Scharber, M., Sariciftci, N.,

Brabec, C.., Scherf, U., “Ultrafast dynamics of charge carrier photogeneration and

geminate recombination in conjugated polymer:fullerene solar cells,” Phys. Rev. B

72(19), 195208 (2005).

[88] Szmytkowski, J., Stampor, W., Kalinowski, J.., Kafafi, Z. H., “Electric field-

assisted dissociation of singlet excitons in tris-(8-hydroxyquinolinato) aluminum

(III),” Appl. Phys. Lett. 80(8), 1465 (2002).

[89] Stampor, W., “Electromodulation of Fuorescence in hole-transporting materials

( TPD , TAPC ) for organic light-emitting diodes,” 351–362 (2000).

[90] Kalinowski, J., Stampor, W., Szmytkowski, J., Virgili, D., Cocchi, M., Fattori, V..,

Sabatini, C., “Coexistence of dissociation and annihilation of excitons on charge

carriers in organic phosphorescent emitters,” Phys. Rev. B 74(8), 85316 (2006).

[91] Wittmer, M.., Zschokke‐Gränacher, I., “Exciton–charge carrier interactions in the

185

electroluminescence of crystalline anthracene,” J. Chem. Phys. 63(10) (1975).

[92] Tolstov, I. V., Belov, a. V., Kaplunov, M. G., Yakuschenko, I. K., Spitsina, N. G.,

Triebel, M. M.., Frankevich, E. L., “On the role of magnetic field spin effect in

photoconductivity of composite films of MEH-PPV and nanosized particles of

PbS,” J. Lumin. 112(1–4), 368–371 (2005).

[93] Levinson, J., “Determination of the Triplet Exciton-Trapped Electron Interaction

Rate Constant in Anthracene Crystals,” J. Chem. Phys. 52(5), 2794 (1970).

[94] Berggren, M., Inganäs, O., Gustafsson, G., Rasmusson, J., Andersson, M. R.,

Hjertberg, T.., Wennerström, O., “Light-emitting diodes with variable colours

from polymer blends,” Nature 372(6505), 444–446 (1994).

[95] Al Attar, H. a., Monkman, A. P., Tavasli, M., Bettington, S.., Bryce, M. R., “White

polymeric light-emitting diode based on a fluorene polymerIr complex blend

system,” Appl. Phys. Lett. 86(12), 1–3 (2005).

[96] Parthasarathy, G., Gu, G.., Forrest, S. R., “Full-color transparent metal-free

stacked organic light emitting device with simplified pixel biasing,” Adv. Mater.

11(11), 907–910 (1999).

[97] Xu, J., Carroll, D. L., Smith, G. M., Dun, C.., Cui, Y., “Achieving High

Performance in AC-Field Driven Organic Light Sources,” Sci. Rep. 6, 24116,

Nature Publishing Group (2016).

[98] Xu, J., Smith, G. M., Dun, C., Cui, Y., Liu, J., Huang, H., Huang, W.., Carroll, D.

L., “Layered , Nanonetwork Composite Cathodes for Flexible , High-Efficiency ,

186

Organic Light Emitting Devices,” Adv. Funct. Mater. 25(28), 4397–4404 (2015).

[99] Tian, Y., Xu, X., Wang, J., Yao, C.., Li, L., “Solution-processed white organic

light-emitting diodes with enhanced efficiency by using quaternary ammonium salt

doped conjugated polyelectrolyte,” ACS Appl. Mater. Interfaces 6(11), 8631–8638

(2014).

[100] Huang, F., Wu, H., Wang, D., Yang, W.., Cao, Y., “Novel Electroluminescent

Conjugated Polyelectrolytes Based on Polyfluorene,” Chem. Mater. 16(8), 708–

716 (2004).

[101] Khramtchenkov, D. V., Arkhipov, V. I.., Bässler, H., “Charge carrier

recombination in organic bilayer electroluminescent diodes. I. Theory,” J. Appl.

Phys. 81(10) (1997).

[102] Jonda, C.., Mayer, A. B. R., “Investigation of the Electronic Properties of Organic

Light-Emitting Devices by Impedance Spectroscopy,” Chem. Mater. 74(5), 2429–

2435 (1999).

[103] Pingree, L. S. C., Scott, B. J., Russell, M. T., Marks, T. J.., Hersam, M. C.,

“Negative capacitance in organic light-emitting diodes,” Appl. Phys. Lett. 86(7),

1–3 (2005).

[104] Gong, X., Yang, Z., Walters, G., Comin, R., Ning, Z., Beauregard, E., Adinolfi,

V., Voznyy, O.., Sargent, E. H., “Highly efficient quantum dot near-infrared light-

emitting diodes,” Nat Phot. advance on, Nature Publishing Group (2016).

[105] Luo, H., Yu, C., Liu, Z., Zhang, G., Geng, H., Yi, Y., Broch, K., Hu, Y.,

187

Sadhanala, A., et al., “Remarkable enhancement of charge carrier mobility of

conjugated polymer field-effect transistors upon incorporating an ionic additive,”

Sci. Adv.(May), 1–11 (2016).

[106] Wang, J., Chepelianskii, A., Gao, F.., Greenham, N. C., “Control of exciton spin

statistics through spin polarization in organic optoelectronic devices,” Nat.

Commun. 3, 1191, Nature Publishing Group (2012).

[107] Cohen, A. E., “Nanomagnetic control of intersystem crossing.,” J. Phys. Chem. A

113(41), 11084–11092 (2009).

[108] Manolopoulos, D. E.., Hore, P. J., “An improved semiclassical theory of radical

pair recombination reactions,” J. Chem. Phys. 139(12), 1–23 (2013).

[109] Thomas, J., “Excited States and Reactions in Liquids,” Ann. Rev. Phys. Chem.

21(17) (1970).

[110] Ding, B. F., Yao, Y., Sun, X. Y., Sun, Z. Y., Gao, X. D., Xie, Z. T., Wang, Z. J.,

Ding, X. M., Wu, Y. Z., et al., “Magnetic field-modulated exciton generation in

organic semiconductors: an intermolecular quantum correlation effect,” 16 (2009).

[111] Song, J. Y., Stingelin, N., Drew, a. J., Kreouzis, T.., Gillin, W. P., “Effect of

excited states and applied magnetic fields on the measured hole mobility in an

organic semiconductor,” Phys. Rev. B - Condens. Matter Mater. Phys. 82(8), 1–5

(2010).

[112] Caruntu, D., Caruntu, G.., O’Connor, C. J., “Magnetic properties of variable-sized

Fe3O4 nanoparticles synthesized from non-aqueous homogeneous solutions of

188

polyols,” J. Phys. D. Appl. Phys. 40(19), 5801–5809 (2007).

[113] Goya, G. F., Lima, E., Arelaro, A. D., Torres, T., Rechenberg, H. R., Rossi, L.,

Marquina, C.., Ibarra, M. R., “Magnetic hyperthermia with Fe3O4 nanoparticles:

The influence of particle size on energy absorption,” IEEE Trans. Magn. 44(11

PART 2), 4444–4447 (2008).

[114] Iida, H., Takayanagi, K., Nakanishi, T.., Osaka, T., “Synthesis of Fe3O4

nanoparticles with various sizes and magnetic properties by controlled hydrolysis,”

J. Colloid Interface Sci. 314(1), 274–280 (2007).

[115] Nie, W., Chen, Y., Smith, G., Xia, Y., Hewitt, C.., Carroll, D., “Nano graphite

platelets enhanced blue emission in alternating current field induced polymer

based electroluminescence devices using Poly (9,9-dioctylfluorenyl-2,7-diyl) as

the emitter,” Org. Electron. 15(1), 99–104, Elsevier B.V. (2014).

[116] Cho, S. H., Sung, J., Hwang, I., Kim, R. H., Choi, Y. S., Jo, S. S., Lee, T. W..,

Park, C., “High performance AC electroluminescence from colloidal quantum dot

hybrids.,” Adv. Mater. 24(33), 4540–4546 (2012).

[117] Lee, J. H., Cho, S. H., Kim, R. H., Jeong, B., Hwang, S. K., Hwang, I., Kim, K. L.,

Kim, E. H., Lee, T.-W., et al., “A Field-induced Hole Generation Layer for High

Performance Alternating Current Polymer Electroluminescence and Its

Application to Extremely Flexible Devices,” J. Mater. Chem. C 4, 4434–4441,

Royal Society of Chemistry (2016).

[118] Suzuki, C., Inogochi, T.., Mito, S., “Thin Film EL Displays,” Inf. Display, SID Int.

Symp. 13, 14–19 (1977).

189

[119] Tseng, S. R., Meng, H. F., Lee, K. C.., Horng, S. F., “Multilayer polymer light-

emitting diodes by blade coating method,” Appl. Phys. Lett. 93(15) (2008).

[120] Ju, B. J., Yamagata, Y.., Higuchi, T., “Thin-Film Fabrication Method for Organic

Light-Emitting Diodes Using Electrospray Deposition,” Adv. Mater. 21, 4343–

4347 (2009).

[121] Hebner, T. R., Wu, C. C., Marcy, D., Lu, M. H.., Sturm, J. C., “Ink-jet printing of

doped polymers for organic light emitting devices,” Appl. Phys. Lett. 72(5), 519–

521 (1998).

[122] Lee, T. W., Noh, T., Shin, H. W., Kwon, O., Park, J. J., Choi, B. K., Kim, M. S.,

Shin, D. W.., Kim, Y. R., “Characteristics of solution-processed small-molecule

organic films and light-emitting diodes compared with their vacuum-deposited

counterparts,” Adv. Funct. Mater. 19(10), 1625–1630 (2009).

[123] Duan, L., Hou, L., Lee, T.-W., Qiao, J., Zhang, D., Dong, G., Wang, L.., Qiu, Y.,

“Solution processable small molecules for organic light-emitting diodes,” J. Mater.

Chem. 20(31), 6392–6407 (2010).

[124] Chen, Y., Xia, Y., Smith, G. M., Hewitt, C. a., Fu, Q., Liu, Y., Sun, H., Ma, D.,

Gu, Y., et al., “High-color-quality white emission in AC-driven field-induced

polymer electroluminescent devices,” Org. Electron. 15(1), 182–188 (2014).

[125] Xu, J., Zhang, L., Zhong, J.., Lin, H., “High-Luminance Flexible Top Emission

Organic Light Emitting Diode with Nickel – Chromium Alloy Thin Film on

Aluminum Anode,” Opt. Rev. 19(6), 358–360 (2012).

190

[126] He, L., Liu, J., Wu, Z., Wang, D., Liang, S., Zhang, X., Jiao, B., Wang, D.., Hou,

X., “Solution-processed small molecule thin films and their light-emitting

devices,” Thin Solid Films 518(14), 3886–3890, Elsevier B.V. (2010).

[127] Noguchi, T., Ohnishi, T., Uchida, M., Ohmori, Y.., Yoshino, K., “Color-variable

light-emitting diode utilizing conducting polymer containing fluorescent dye,” Jpn.

J. Appl. Phys. 32(7), L921–L924 (1993).

[128] Piao, J., Katori, S., Ikenoue, T.., Fujita, S., “Ultrasonic spray-assisted solution-

based vapor-deposition of aluminum tris(8-hydroxyquinoline) thin films,” Jpn. J.

Appl. Phys. 50(2) (2011).

[129] Höfle, S., Lutz, T., Egel, A., Nickel, F., Kettlitz, S. W., Gomard, G., Lemmer, U..,

Colsmann, A., “Influence of the Emission Layer Thickness on the Optoelectronic

Properties of Solution Processed Organic Light-Emitting Diodes,” ACS Photonics

1(10), 968–973 (2014).

[130] Fröbel, M., Perumal, A., Schwab, T., Gather, M. C., Lüssem, B.., Leo, K.,

“Enhancing the efficiency of alternating current driven organic light-emitting

devices by optimizing the operation frequency,” Org. Electron. 14(3), 809–813

(2013).

[131] Perumal, A., Lüssem, B.., Leo, K., “Ultra-bright alternating current organic

electroluminescence,” Org. Electron. 13(9), 1589–1593, Elsevier B.V. (2012).

[132] Wang, Z. B., Helander, M. G., Qiu, J., Puzzo, D. P., Greiner, M. T., Hudson, Z.

M., Wang, S., Liu, Z. W.., Lu, Z. H., “Unlocking the full potential of organic light-

emitting diodes on flexible plastic,” Nat. Photonics 5(October), 753–757 (2011).

191

[133] Han, T.-H., Lee, Y., Choi, M.-R., Woo, S.-H., Bae, S.-H., Hong, B. H., Ahn, J.-

H.., Lee, T.-W., “Extremely efficient flexible organic light-emitting diodes with

modified graphene anode,” Nat. Photonics 6(2), 105–110, Nature Publishing

Group (2012).

[134] Lee, B. R., Jung, E. D., Park, J. S., Nam, Y. S., Min, S. H., Kim, B.-S., Lee, K.-M.,

Jeong, J.-R., Friend, R. H., et al., “Highly efficient inverted polymer light-emitting

diodes using surface modifications of ZnO layer.,” Nat. Commun. 5, 4840, Nature

Publishing Group (2014).

[135] Chang, H., Wang, G., Yang, A., Tao, X., Liu, X., Shen, Y.., Zheng, Z., “A

Transparent, Flexible, Low-Temperature, and Solution-Processible Graphene

Composite Electrode,” Adv. Funct. Mater. 20(17), 2893–2902 (2010).

[136] Kwak, K., Cho, K.., Kim, S., “Stable bending performance of flexible organic

light-emitting diodes using IZO anodes.,” Sci. Rep. 3, 2787 (2013).

[137] Ho, M. D., Kim, D., Kim, N., Cho, S. M.., Chae, H., “Polymer and small molecule

mixture for organic hole transport layers in quantum dot light-emitting diodes.,”

ACS Appl. Mater. Interfaces 5(23), 12369–12374 (2013).

[138] Forrest, S. R., “The path to ubiquitous and low-cost organic electronic appliances

on plastic,” Nature 428(April), 911–918 (2004).

[139] Baldo, M., Thompson, M.., Forrest, S., “High-efficiency fluorescent organic light-

emitting devices using a phosphorescent sensitizer,” Nature 403(6771), 750–753

(2000).

192

[140] Uoyama, H., Goushi, K., Shizu, K., Nomura, H.., Adachi, C., “Highly efficient

organic light-emitting diodes from delayed fluorescence.,” Nature 492(7428), 234–

238, Nature Publishing Group (2012).

[141] Kim, C.-H.., Shinar, J., “Bright small molecular white organic light-emitting

devices with two emission zones,” Appl. Phys. Lett. 80(12) (2002).

[142] Choong, V.-E., Shi, S., Curless, J., Shieh, C.-L., Lee, H.-C., So, F., Shen, J..,

Yang, J., “Organic light-emitting diodes with a bipolar transport layer,” Appl.

Phys. Lett. 75(2) (1999).

[143] He, G., Pfeiffer, M., Leo, K., Hofmann, M., Birnstock, J., Pudzich, R.., Salbeck, J.,

“High-efficiency and low-voltage p‐i‐n electrophosphorescent organic light-

emitting diodes with double-emission layers,” Appl. Phys. Lett. 85(17) (2004).

[144] Kido, J., Kimura, M.., Nagai, K., “Multilayer white light-emitting organic

electroluminescent Device,” Science (80-. ). 267(5202), 1332–1334 (1995).

[145] Ding, X. M., Hung, L. M., Cheng, L. F., Deng, Z. B., Hou, X. Y., Lee, C. S.., Lee,

S. T., “Modification of the hole injection barrier in organic light-emitting devices

studied by ultraviolet photoelectron spectroscopy,” Appl. Phys. Lett. 76(19)

(2000).

[146] Su, S. J., Chiba, T., Takeda, T.., Kido, J., “Pyridine-containing triphenylbenzene

derivatives with high electron mobility for highly efficient phosphorescent

OLEDs,” Adv. Mater. 20, 2125–2130 (2008).

[147] Wong, W.-Y.., Ho, C.-L., “Functional metallophosphors for effective charge

193

carrier injection/transport: new robust OLED materials with emerging

applications,” J. Mater. Chem. 19(26), 4457 (2009).

[148] De, S., Higgins, T. M., Lyons, P. E., Doherty, E. M., Nirmalraj, P. N., Blau, W. J.,

Boland, J. J.., Coleman, J. N., “Silver Nanowire Networks as Flexible ,” ACSNano

3(7), 1767–1774 (2009).

[149] Nanowire, S., Hu, L., Kim, H. S., Lee, J., Peumans, P.., Cui, Y., “Scalable Coating

and Properties of transparent Ag nanowire,” ACS Nano 4(5), 2955–2963 (2010).

[150] Krantz, J., Stubhan, T., Richter, M., Spallek, S., Litzov, I., Matt, G. J., Spiecker,

E.., Brabec, C. J., “Spray-Coated Silver Nanowires as Top Electrode Layer in

Semitransparent P3HT:PCBM-Based Organic Solar Cell Devices,” Adv. Funct.

Mater. 23(13), 1711–1717 (2013).

[151] Lee, J. Y., Connor, S. T., Cui, Y.., Peumans, P., “Solution-processed metal

nanowire mesh transparent electrodes,” Nano Lett. 8, 689–692 (2008).

[152] Ju, S., Li, J., Liu, J., Chen, P. C., Ha, Y. G., Ishikawa, F., Chang, H., Zhou, C.,

Facchetti, A., et al., “Transparent active matrix organic light-emitting diode

displays driven by nanowire transistor circuitry,” Nano Lett. 8, 997–1004 (2008).

[153] Zhou, L., Xiang, H., Shen, S., Li, Y., Chen, J., Xie, H., Goldthorpe, I. A., Chen, L.,

Lee, S., et al., “High-Performance Flexible Organic Light-Emitting Diodes Using

Embedded Silver Network Transparent,” 12796–12805 (2014).

[154] Eda, G., Lin, Y.-Y., Miller, S., Chen, C.-W., Su, W.-F.., Chhowalla, M.,

“Transparent and conducting electrodes for organic electronics from reduced

194

graphene oxide,” Appl. Phys. Lett. 92(23) (2008).

[155] Geim, A. K.., Novoselov, K. S., “The rise of graphene,” Nat Mater 6(3), 183–191

(2007).

[156] Li, X., Zhang, G., Bai, X., Sun, X., Wang, X., Wang, E.., Dai, H., “Highly

conducting graphene sheets and Langmuir-Blodgett films,” Nat Nano 3(9), 538–

542, Nature Publishing Group (2008).

[157] Meyer, J. C., Geim, a K., Katsnelson, M. I., Novoselov, K. S., Booth, T. J.., Roth,

S., “The structure of suspended graphene sheets.,” Nature 446(March), 60–63

(2007).

[158] Wang, X., Zhi, L.., Müllen, K., “Transparent, Conductive Graphene Electrodes for

Dye-Sensitized Solar Cells,” Nano Lett. 8(1), 323–327 (2008).

[159] Cao, Q., Zhu, Z.-T., Lemaitre, M. G., Xia, M.-G., Shim, M.., Rogers, J. A.,

“Transparent flexible organic thin-film transistors that use printed single-walled

carbon nanotube electrodes,” Appl. Phys. Lett. 88(11) (2006).

[160] Hu, L., Hecht, D. S.., Grüner, G., “Percolation in Transparent and Conducting

Carbon Nanotube Networks,” Nano Lett. 4(12), 2513–2517 (2004).

[161] Wu, Z., Chen, Z., Du, X., Logan, J. M., Sippel, J., Nikolou, M., Kamaras, K.,

Reynolds, J. R., Tanner, D. B., et al., “Transparent, conductive carbon nanotube

films.,” Science 305(5688), 1273–1276 (2004).

[162] Dan, B., Irvin, G. C.., Pasquali, M., “Continuous and Scalable Fabrication of

Transparent Conducting Carbon Nanotube Films,” ACS Nano 3(4), 835–843

195

(2009).

[163] Lee, J., Shin, H., Noh, Y., Na, S.., Kim, H., “Cells Brush painting of transparent

PEDOT / Ag nanowire / PEDOT multilayer electrodes for flexible organic solar

cells,” Sol. Energy Mater. Sol. 114, 15–23 (2013).

[164] Mazzeo, M., Mariano, F., Genco, A., Carallo, S.., Gigli, G., “High efficiency ITO-

free flexible white organic light-emitting diodes based on multi-cavity

technology,” Org. Electron. 14(11), 2840–2846, Elsevier B.V. (2013).

[165] Toth, G., Mäklin, J., Halonen, N., Palosaari, J., Juuti, J., Jantunen, H., Kordas, K.,

Sawyer, W. G., Vajtai, R., et al., “Carbon-nanotube-based electrical brush

contacts,” Adv. Mater. 21, 2054–2058 (2009).

[166] Dresselhaus, M. S., “Applied physics: nanotube antennas.,” Nature

432(December), 959–960 (2004).

[167] Chen, J., Minett, A., Liu, Y., Lynam, C., Sherrell, P., Wang, C.., Wallace, G.,

“Direct Growth of Flexible Carbon Nanotube Electrodes,” Adv. Mater. 20, 566–

570 (2008).

[168] Ma, L., “Organic light-emitting diode with enhanced efficiency,” Google Patents

(2011).

[169] Ago, H., Kugler, T., Cacialli, F., Salaneck, W. R., Shaffer, M. S. P., Windle, A.

H.., Friend, R. H., “Work Functions and Surface Functional Groups of Multiwall

Carbon Nanotubes,” J. Phys. Chem. B 103(38), 8116–8121 (1999).

[170] Shiraishi, M.., Ata, M., “Work function of carbon nanotubes,” MRS Proc. 633(July

196

2000), 1913–1917 (2000).

[171] Jung, D., Lee, K. H., Kim, D., Burk, D., Overzet, L. J.., Lee, G. S., “Highly

Conductive Flexible Multi-Walled Carbon Nanotube Sheet Films for Transparent

Touch Screen,” Jpn. J. Appl. Phys. 52(3S), 03BC03 (2013).

[172] Xu, J., Carroll, D. L., Li, P., Nolan, L. V., Shao, L.., Dun, C., “Solution Processing

Small-Molecule Organic Emitter in Field-Induced, Carrier Gated Lighting

Devices,” Adv. Opt. Mater. 5(6), 1600917 (2017).

[173] Baldo, M. a., Lamansky, S., Burrows, P. E., Thompson, M. E.., Forrest, S. R.,

“Very high-efficiency green organic light-emitting devices based on

electrophosphorescence,” Appl. Phys. Lett. 75(1), 4 (1999).

[174] Lee, J., Chen, H.-F., Batagoda, T., Coburn, C., Djurovich, P. I., Thompson, M. E..,

Forrest, S. R., “Deep blue phosphorescent organic light-emitting diodes with very

high brightness and efficiency,” Nat. Mater.(October), 1–8 (2015).

[175] Adachi, C., Kwong, R. C., Djurovich, P., Adamovich, V., Baldo, M. A.,

Thompson, M. E.., Forrest, S. R., “Endothermic energy transfer: A mechanism for

generating very efficient high-energy phosphorescent emission in organic

materials,” Appl. Phys. Lett. 79(13), 2082–2084 (2001).

[176] Duan, J.-P., Sun, P.-P.., Cheng, C.-H., “New Iridium Complexes as Highly

Efficienct Orange-Red Emmitters in Organic Light-Emmitting Diodes,” Adv.

Mater. 15(89), 224–228 (2003).

[177] Reineke, S., Lindner, F., Schwartz, G., Seidler, N., Walzer, K., Lüssem, B.., Leo,

197

K., “White organic light-emitting diodes with fluorescent tube efficiency.,” Nature

459(7244), 234–238 (2009).

[178] Reineke, S., Schwartz, G., Walzer, K.., Leo, K., “Reduced efficiency roll-off in

phosphorescent organic light emitting diodes by suppression of triplet-triplet

annihilation,” Appl. Phys. Lett. 91(12), 1–4 (2007).

[179] King, S. M., Cass, M., Pintani, M., Coward, C., Dias, F. B., Monkman, a. P..,

Roberts, M., “The contribution of triplet-triplet annihilation to the lifetime and

efficiency of fluorescent polymer organic light emitting diodes,” J. Appl. Phys.

109(7), 14–19 (2011).

[180] Schwartz, G., Reineke, S., Rosenow, T. C., Walzer, K.., Leo, K., “Triplet

harvesting in hybrid white organic light-emitting diodes,” Adv. Funct. Mater.

19(9), 1319–1333 (2009).

[181] Quan, L. N., Quintero-Bermudez, R., Voznyy, O., Walters, G., Jain, A., Fan, J. Z.,

Zheng, X., Yang, Z.., Sargent, E. H., “Highly Emissive Green Perovskite

Nanocrystals in a Solid State Crystalline Matrix,” Adv. Mater. 29(21), 1605945

(2017).

[182] Nitsch, K., Hamplová, V., Nikl, M., Polák, K.., Rodová, M., “Lead bromide and

ternary alkali lead bromide single crystals — growth and emission properties,”

Chem. Phys. Lett. 258(3–4), 518–522 (1996).

[183] Akkerman, Q. A., Radicchi, E., Park, S., Nunzi, F., Mosconi, E., De Angelis, F.,

Brescia, R., Rastogi, P., Prato, M., et al., “Nearly Monodisperse Insulator

Cs4PbX6 (X = Cl, Br, I) Nanocrystals, Their Mixed Halide Compositions, and

198

Their Transformation into CsPbX3 Nanocrystals,” Nano Lett. 17(3), 1924–1930

(2017).

[184] Kondo, S. I., Kakuchi, M., Masaki, A.., Saito, T., “Strongly enhanced free-exciton

luminescence in microcrystalline CsPbBr 3 films,” J. Phys. Soc. Japan 72(7),

1789–1791 (2003).

[185] Cha, J.-H., Han, J. H., Yin, W., Park, C., Park, Y., Ahn, T. K., Cho, J. H.., Jung,

D.-Y., “Photoresponse of CsPbBr3 and Cs4PbBr6 Perovskite Single Crystals,” J.

Phys. Chem. Lett. 8, 565–570 (2017).

[186] Nikl, M., Nitsch, K., Fabeni, P., Pazzi, G. P., Gurioli, M., Santucci, S., Phani, R.,

Scacco, a., Somma, F., “Luminescence of CsPbBr 3 -like quantum dots in CsBr

single crystals,” Phys. E 4, 323–331 (1999).

[187] Nikl, M., Nitsch, K., Mihokova, E., Polak, K., Fabeni, P., Pazzi, G. P., Gurioli, M.,

Phani, R., Santucci, S., et al., “Optical properties of Pb 2+ -based aggregated

phases in CsBr Thin film and single crystal matrices,” Radiat. Eff. Defects Solids

150(1–4), 341–345 (1999).

[188] Kondo, S., Nakagawa, H., Saito, T.., Asada, H., “Extremely photo-luminescent

microcrystalline CsPbX3 (X=Cl, Br) films obtained by amorphous-to-crystalline

transformation,” Curr. Appl. Phys. 4(5), 439–444 (2004).

[189] Nikl, M., Nitsch, K., Mihokova, E., Polak, K., Fabeni, P., Pazzi, G. P., Gurioli, M.,

Phani, R., Santucci, S., et al., “Optical properties of Pb2+-based aggregated phases

in CsBr Thin film and single crystal matrices,” Radiat. Eff. Defects Solids 150(1–

4), 341–345, Taylor & Francis (1999).

199

[190] Nikl, M., Mihokova, E., Nitsch, K., Somma, F., Giampaolo, C., Pazzi, G. P.,

Fabeni, P.., Zazubovich, S., “Photoluminescence of Cs 4 PbBr 6 crystals and thin

films,” Chem. Phys. Lett. 306(June), 280–284 (1999).

[191] Kondo, S., Amaya, K.., Saito, T., “Localized optical absorption in Cs4PbBr6,” J.

Phys. Condens. Matter 14, 2093–2099 (2002).

[192] Yettapu, G. R., Talukdar, D., Sarkar, S., Swarnkar, A., Nag, A., Ghosh, P..,

Mandal, P., “Terahertz Conductivity within Colloidal CsPbBr3 Perovskite

Nanocrystals: Remarkably High Carrier Mobilities and Large Diffusion Lengths,”

Nano Lett. 16(8), 4838–4848 (2016).

[193] Li, J., Yuan, X., Jing, P., Li, J., Wei, M., Hua, J., Zhao, J.., Tian, L., “Temperature-

dependent photoluminescence of inorganic perovskite nanocrystal films,” RSC

Adv. 6(82), 78311–78316 (2016).

[194] Saidaminov, M. I., Almutlaq, J., Sarmah, S. P., Dursun, I., Zhumekenov, A. A.,

Begum, R., Pan, J., Cho, N., Mohammed, O. F., et al., “Pure Cs 4 PbBr 6 : Highly

Luminescent Zero-Dimensional Perovskite Solids,” ACS Energy Lett. 1(4), 840–

845 (2016).

[195] Chen, D., Wan, Z., Chen, X., Yuan, Y.., Zhong, J., “Large-scale room-temperature

synthesis and optical properties of perovskite-related Cs 4 PbBr 6 fluorophores,” J.

Mater. Chem. C 4(45), 10646–10653, Royal Society of Chemistry (2016).

[196] Kondo, S., Takahashi, K., Nakanish, T., Saito, T., Asada, H.., Nakagawa, H.,

“High intensity photoluminescence of microcrystalline CsPbBr3 films: Evidence

for enhanced stimulated emission at room temperature,” Curr. Appl. Phys. 7(1), 1–

200

5 (2007).

[197] Zhang, Y., Saidaminov, M. I., Dursun, I., Yang, H., Murali, B., Alarousu, E.,

Yengel, E., Alshankiti, B. A., Bakr, O. M., et al., “Zero-Dimensional Cs4PbBr6

Perovskite Nanocrystals,” J. Phys. Chem. Lett., acs.jpclett.7b00105 (2017).

[198] Xing, G., Wu, B., Wu, X., Li, M., Du, B., Wei, Q., Guo, J., Yeow, E. K. L., Sum,

T. C., et al., “Transcending the slow bimolecular recombination in lead-halide

perovskites for electroluminescence,” Nat. Commun. 8, 14558 (2017).

[199] Makarov, N. S., Guo, S., Isaienko, O., Liu, W., Robel, I.., Klimov, V. I., “Spectral

and Dynamical Properties of Single Excitons, Biexcitons, and Trions in Cesium-

Lead-Halide Perovskite Quantum Dots,” Nano Lett. 16(4), 2349–2362 (2016).

[200] Pietryga, J. M., Park, Y. S., Lim, J., Fidler, A. F., Bae, W. K., Brovelli, S..,

Klimov, V. I., “Spectroscopic and device aspects of nanocrystal quantum dots,”

Chem. Rev. 116(18), 10513–10622 (2016).

[201] Dennis, A. M., Mangum, B. D., Piryatinski, A., Park, Y. S., Hannah, D. C.,

Casson, L., Williams, D. J., Schaller, R. D., Htoon, H., et al., “SI-Suppressed

Blinking and Auger Recombination in Near-Infrared Type-II InP / CdS

Nanocrystal Quantum Dots,” Nano Lett. 12(11), 5545–5551 (2012).

[202] Hiroshige, N., Ihara, T., Saruyama, M., Teranishi, T.., Kanemitsu, Y., “Coulomb-

Enhanced Radiative Recombination of Biexcitons in Single Giant-Shell CdSe/CdS

Core/Shell Nanocrystals,” J. Phys. Chem. Lett. 8(9), 1961–1966 (2017).

[203] Lim, J., Jeong, B. G., Park, M., Kim, J. K., Pietryga, J. M., Park, Y. S., Klimov, V.

201

I., Lee, C., Lee, D. C., et al., “Influence of shell thickness on the performance of

light-emitting devices based on CdSe/Zn 1- X Cd x S core/shell heterostructured

quantum dots,” Adv. Mater. 26(47), 8034–8040 (2014).

[204] Stranks, S. D., Burlakov, V. M., Leijtens, T., Ball, J. M., Goriely, A.., Snaith, H. J.,

“Recombination Kinetics in Organic-Inorganic Perovskites: Excitons, Free

Charge, and Subgap States,” Phys. Rev. Appl. 2(3), 1–8 (2014).

[205] Kang, B.., Biswas, K., “Exploring Polaronic, Excitonic Structures and

Luminescence in Cs4PbBr6/CsPbBr3,” J. Phys. Chem. Lett. 9(4), 830–836,

American Chemical Society (2018).

[206] Piryatinski, A., Ivanov, S. A., Tretiak, S.., Klimov, V. I., “Effect of Quantum and

Dielectric Confinement on the Exciton − Exciton Interaction Energy in Type II

Core / Shell Semiconductor Nanocrystals,” Nano Lett. 7(1), 108–115 (2007).

[207] Micic, O. I., Cheong, H. M., Fu, H., Zunger, A., Sprague, J. R., Mascarenhas, A..,

Nozik, a J., “Size-Dependent Spectroscopy of InP Quantum Dots,” J. Phys. Chem.

B 5647(97), 4904–4912 (1997).

[208] Meltzer, R., Feofilov, S., Tissue, B.., Yuan, H., “Dependence of fluorescence

lifetimes of Y2O3:Eu3+ nanoparticles on the surrounding medium,” Phys. Rev. B

60(20), R14012–R14015 (1999).

[209] Takagahara, T., “Effects of dielectric confinement and electron-hole exchange

interaction on excitonic states in semiconductor quantum dots,” Phys. Rev. B

47(8), 4569–4584 (1993).

202

[210] Takagahara, T., “Nonlocal theory of the size and temperature dependence of the

radiative decay rate of excitons in semiconductor quantum dots,” Phys. Rev. B

47(24), 16639–16642 (1993).

[211] Rashba, E. I.., Gurgenishvili, G. E., “Edge Absorption Theory in Semiconductors,”

Sov. Phys. - Solide State 4, 759–760 (1962).

[212] Henry, C. H.., Nassau, K., “Lifetimes of Bound Excitons in CdS,” Phys. Rev. B

1(4), 1628–1634 (1970).

[213] Itoh, T., Ikehara, T.., Iwabuchi, Y., “Quantum Confinement of Excitons and Their

Relaxation Processes,” J. Lumin. 45, 29–33 (1990).

[214] Nakamura, A., Yamada, H.., Tokizaki, T., “Size-dependent radiative decay of

excitons in CuCl semiconducting quantum spheres embedded in glasses,” Phys.

Rev. B 40(12), 8585–8588 (1989).

[215] Stoumpos, C. C., Malliakas, C. D., Peters, J. A., Liu, Z., Sebastian, M., Im, J.,

Chasapis, T. C., Wibowo, A. C., Chung, D. Y., et al., “Crystal Growth of the

Perovskite Semiconductor CsPbBr 3 : A New Material for High-Energy Radiation

Detection,” Cryst. Growth Des. 13, 2722−2727 (2013).

[216] Chen, X., Zhang, F., Ge, Y., Shi, L., Huang, S., Tang, J.., Lv, Z., “Centimeter-

Sized Cs 4 PbBr 6 Crystals with Embedded CsPbBr 3 Nanocrystals Showing

Superior Photoluminescence : Nonstoichiometry Induced Transformation and

Light-Emitting Applications,” Adv. Funct. Mater., 1–7 (2018).

[217] Xu, J., Huang, W., Li, P., Oken, D. R., Dun, C., Guo, Y., Ucer, K. B., Lu, C.,

203

Geyer, S., et al., “Imbedded Nanocrystals of CsPbBr3 in Cs4PbBr6 – Enhanced

Oscillator Strength, Kinetics, and Application in Light Emitting Diodes,” Adv.

Mater. 29(443), 1703703 (2017).

[218] Guo, Y., Dun, C., Xu, J., Mu, J., Li, P., Gu, L., Hou, C., Hewitt, C. A., Zhang, Q.,

et al., “Ultrathin, Washable, and Large-Area Graphene Papers for Personal

Thermal Management,” Small, 1702645 (2017).

[219] Dun, C., Hewitt, C. A., Guo, Y., Jiang, Q., Xu, J., Marcus, G., Schall, D. C..,

Carroll, D. L., “Self-Assembled Heterostructures: Selective Growth of Metallic

Nanoparticles on V2-VI3 Nanopates,” Adv. Mater. 29(38), 1702968 (2017).

[220] Dun, C., Hewitt, C. A., Li, Q., Xu, J., Schall, D. C., Lee, H., Jiang, Q.., Carroll, D.

L., “2D Chalcogenide Nanoplate Assemblies for Thermoelectric Applications,”

Adv. Mater. 29(21), 1700070 (2017).

[221] Xu, J., Carroll, D. L., Shao, L., Li, P., Dun, C.., Nolan, L., “Polymer Gating White

Flexible Field-Induced Lighting Device,” Adv. Mater. Technol. 2(8), 1700017

(2017).

[222] Dun, C., Hewitt, C. a., Huang, H., Xu, J., Montgomery, D. S., Nie, W., Jiang, Q..,

Carroll, D. L., “Layered Bi2Se3 Nanoplate/Polyvinylidene Fluoride Composite

Based n-type Thermoelectric Fabrics,” ACS Appl. Mater. Interfaces 7(13), 7054–

7059 (2015).

[223] Huang, H., Dun, C., Huang, W., Cui, Y., Xu, J., Jiang, Q., Xu, C., Zhang, H., Wen,

S., et al., “Solution-Processed Yellow-White Light-Emitting Diodes Based on

Mixed-Solvent Dispersed Luminescent ZnO Nanocrystals,” Appl. Phys. Lett.

204

106(26), 263506 (2015).

[224] Dun, C., Hewitt, C. a., Huang, H., Montgomery, D. S., Xu, J.., Carroll, D. L.,

“Flexible thermoelectric fabrics based on self-assembled tellurium nanorods with a

large power factor,” Phys. Chem. Chem. Phys. 17(14), 8591–8595, Royal Society

of Chemistry (2015).

[225] Dun, C., Huang, W., Huang, H., Xu, J., Zhou, N., Zheng, Y., Tsai, H., Nie, W.,

Onken, D. R., et al., “Hydrazine-Free Surface Modi fi cation of CZTSe

Nanocrystals with All- Inorganic Ligand,” J. Phys. Chem. C 118(51), 30302–

30308 (2014).

205

Junwei Xu Gender: Male Date of Birth: June 1st, 1989 Address: 2551A Owen Drive, 27106 Winston-Salem, NC Email: [email protected] Mobile Phone: +1 (314) 435-6202

EDUCATION  Wake Forest University – Center for Nanotechnology and Molecular Materials, Department of Physics, Winston-Salem, USA PhD in Physics Advised by Prof. David L. Carroll Graduation Date: Mar. 2018  University of Electronic Science and Technology of China, Chengdu, China Bachelor of Information Display and Opto-Electronic Technology Graduation Date: Jul. 2012 RESEARCH INTERESTS optoelectronics; LED; nanotechnology and nanomaterials; thin film depostion;

RESEARCH EXPERIENCE  Center for Nanotechnology and Molecular Materials and Department of Physics at Wake Forest University Oct. 2013-present ♦ Developed Cs-Pb-Br perovskite core-shell thin film and single crystal. ♦ Studied excitonic oscillator strength and carrier transport of perovskite nanocomposite. ♦ Applied the perovskite core-shell structure in the light emitting device and scintillation detector. ♦ Achieved the color-tunable organic electroluminescent devices via the ion- enhanced Maxwell magnetic coupling effect. ♦ Developed ultra flexible, stretchable perovskite light emitting smart skin. ♦ Studied carrier transport mechanism under high frequency in AC driven organic electroluminescence devices.

 Key Laboratory of Display Science and Technology of Sichuan Province Oct. 2010-Aug.2013 ♦ Designed flexible anode stack structure for top emission OLED on Poly (ether sulfones) substrate. ♦ Optimized etching process on Aluminum and Nickel-Chromium alloy electrode on flexible substrate. ♦ Researched on quick-response liquid power display.

206

 State Key Laboratory of OLED Science and Technology & OLED Joint Laboratory of UESTC and BOE Mar. 2011- Aug.2012 ♦ Studied hole-injection efficiency.and metal-organic interfacial morphology of OLEDs. ♦ Fabricated the flexible OLED arrays on flexible substrate. ♦ Optimized OLED encapsulation technology.

 State Key Laboratory of Electronic Thin Films and Integrated Devices of China Apr. 2012-Jul.2012 ♦ Fabricated organic solar cells and organic field-effect transistors. ♦ Built organic photo-absorption layer by solution processing and determined light absorption of the photoactive layers.

SKILLS AND TECHNIQUES  Thin film and Integrated device fabrication, test and optimization based on solution and vacuum process (spin-coating, thermal evaporation, RF sputtering deposition, ALD, spray-coating, print-coating, blade-coating, CVD, et al);  Photolithography process operation: U-V lithography machine; Spinning and drying processor;  Characterization techniques (SEM, TEM, HRTEM, AFM, XRD, XPS, UV-vis, PL, Abs, LSM et al);  Synthetic techniques (nanocrystals, nanoplates, single crystal);  Ultra-fast laser set-up and optimization, transient process measurement and analysis (streak camera, transient absorption, Z-scan, et al);  Experience on Labview and Matlab;  Cleanroom work experience (BOE OLED pilot line and Wake Forest Nanotech Center).

COLLABORATIONS  Imbedded nanocrystals of pervoskite nanocomposite Collaborate with Berkley Lawrence National Laboratory (USA), Los Alamos National Laboratory (USA), Oak Ridge National Laboratory (USA), and Fisk University (USA)  Field-induced pervoskite top emission lighting source Collaborate with Trinity College Dublin (Ireland), Donghua University (China), and State Key Laboratory for Modification of Chemical Fibers and Polymer Materials (China)  Tailoring spin mixtures in color-tunable organic EL devices

207

Collaborate with University of Stuttgart (Germany), Key Laboratory of Luminescence and Optical Information (China), Beijing Jiaotong University (China), University of Dayton (USA), Donghua University (China), and State Key Laboratory for Modification of Chemical Fibers and Polymer Materials (China)

TEACHING EXPERIENCE  Instructed undergraduate level Physics (PHY113) course as a lab lecturer (3 semesters, 2013-2014).  Worked as a tutor for PHY113 course (3 semesters, 2013-2015).

HONORS AND AWARDS  SPIE Student Member (since 2017)  APS Student Member (since 2014)  Scholarship from Graduate School of Art and Science at Wake Forest University (since 2013)  UESTC Excellent Graduate 2012  National Inspirational Scholarship (2009 & 2011)  People’s Scholarship-First Level (2010)  Outstanding Undergraduate Graduation Thesis (2012)  Excellent Student of School of Optoelectronic Information (2010)  IEEE Student Member of the Chinese Institute of Electronics (2012-2013) RELATED COURSES Thin Film Technology; Solid-state and Semiconductor Physics; Vacuum Technology; Quantum Mechanics; Information Display Technology; Crystal Display Technology; Display Devices Drive Technology; Fundamentals of Irradiancy Theory; Fundamental of Analog Circuit; Digital Logic Desgn and Application, et al.

INVITED TALKS AND PRESENTATIONS [1] “Imbedded nanocrystals of pervoskite nanocomposite – enhanced oscillator strength, kinetics and LED application”, MRS Spring 2018, April 2 -6, Phenoix, USA

[2] “Polymeric Semiconducting Gate in Flexible, White Emission, AC-Driven Field Induced Electroluminescence Devices”, SPIE Optics&Photonics 2017, Sep 6 - 11, San Diego, USA.

[3] “Spintronics and Magnetic Application in High Frequency Field Induced Organic Lighting Devices”, Sturrgart Nanodays Workshop (Invited talk in organizer section) 2016, Sep 14 - 16, Stuttgart, Germany.

[4] “Recent Progess in Perovskite Light Emitting Diode”, Sturrgart Nanodays Workshop (Invited talk) 2016, Sep 14 - 16, Stuttgart, Germany.

208

SELECTED PUBLICATIONS [1] J. Xu, D. L. Carroll, Y. Cui, G. M. Smith, P. Li, C. Dun, W. Huang, L. Shao, Y. Guo, and H. Wang, “Tailoring Spin Mixtures by Ion-Enhanced Maxwell Magnetic Coupling in Color-Tunable Organic Electroluminescent Devices,” Light Sci. Appl., In revision, 2018. [2] C. Dun, C. A. Hewitt, Q. Jiang, Y. Guo, J. Xu, Y. Li, Q. Li, H. Wang, and D. L. Carroll, “Bi2Te3 Plates with Single Nanopore: The Formation of Surface Defects and Self- Repair Growth,” Chem. Mater., Just accepted, 2018. [3] J. Xu, W. Huang, P. Li, D. R. Oken, C. Dun, Y. Guo, K. B. Ucer, C. Lu, S. Geyer, R. T. Williams, and D. L. Carroll, “Imbedded Nanocrystals of CsPbBr3 in Cs4PbBr6 – Enhanced Oscillator Strength, Kinetics, and Application in Light Emitting Diodes,” Adv. Mater., vol. 29, no. 443, p. 1703703, 2017. [4] Y. Guo, C. Dun, J. Xu, J. Mu, P. Li, L. Gu, C. Hou, C. A. Hewitt, Q. Zhang, Y. Li, D. L. Carroll, and H. Wang, “Ultrathin, Washable, and Large-Area Graphene Papers for Personal Thermal Management,” Small, p. 1702645, 2017. (Cover highlight) [5] J. Xu, D. L. Carroll, L. Shao, P. Li, C. Dun, and L. Nolan, “Polymer Gating White Flexible Field-Induced Lighting Device,” Adv. Mater. Technol., vol. 2, no. 8, p. 1700017, 2017. [6] C. Dun, C. A. Hewitt, Y. Guo, Q. Jiang, J. Xu, G. Marcus, D. C. Schall, and D. L. Carroll, “Self-Assembled Heterostructures: Selective Growth of Metallic Nanoparticles on V2-VI3 Nanopates,” Adv. Mater., vol. 29, no. 38, p. 1702968, 2017. [7] J. Xu, D. L. Carroll, P. Li, L. V. Nolan, L. Shao, and C. Dun, “Solution Processing Small-Molecule Organic Emitter in Field-Induced, Carrier Gated Lighting Devices,” Adv. Opt. Mater., vol. 5, no. 6, p. 1600917, 2017. [8] C. Dun, C. A. Hewitt, Q. Li, J. Xu, D. C. Schall, H. Lee, Q. Jiang, and D. L. Carroll, “2D Chalcogenide Nanoplate Assemblies for Thermoelectric Applications,” Adv. Mater., vol. 29, no. 21, p. 1700070, 2017. (Cover highlight) [9] J. Xu, D. L. Carroll, G. M. Smith, C. Dun, and Y. Cui, “Achieving High Performance in AC-Field Driven Organic Light Sources,” Sci. Rep., vol. 6, p. 24116, 2016. [10] J. Xu, G. M. Smith, C. Dun, Y. Cui, J. Liu, H. Huang, W. Huang, and D. L. Carroll, “Layered , Nanonetwork Composite Cathodes for Flexible , High-Efficiency , Organic Light Emitting Devices,” Adv. Funct. Mater., vol. 25, no. 28, pp. 4397–4404, 2015. (Cover highlight) [11] C. Dun, C. a. Hewitt, H. Huang, J. Xu, D. S. Montgomery, W. Nie, Q. Jiang, and D. L. Carroll, “Layered Bi2Se3 Nanoplate/Polyvinylidene Fluoride Composite Based n- type Thermoelectric Fabrics,” ACS Appl. Mater. Interfaces, vol. 7, no. 13, pp. 7054–7059, 2015.

209

[12] H. Huang, C. Dun, W. Huang, Y. Cui, J. Xu, Q. Jiang, C. Xu, H. Zhang, S. Wen, and D. L. Carroll, “Solution-Processed Yellow-White Light-Emitting Diodes Based on Mixed-Solvent Dispersed Luminescent ZnO Nanocrystals,” Appl. Phys. Lett., vol. 106, no. 26, p. 263506, 2015.

210