Magnetic field effect in organic films and devices

In this work, we focused on the magnetic field effect in organic films and devices, including organic light emitting diodes (OLEDs) and organic photovoltaic (OPV) cells. We measured magnetic field effect (MFE) such as magnetoconductance (MC) and magneto-electroluminescence (MEL) in OLEDs based on several π-conjugated polymers and small molecules for fields B<100 mT. We found that both MC(B) and MEL(B) responses in bipolar devices and MC(B) response in unipolar devices are composed of two B-regions: (i) an ‘ultra-small' region at |B| < 1-2 mT, and (ii) a monotonic response region at |B| >∼2mT. Magnetic field effect (MFE) measured on three isotopes of Poly (dioctyloxy) phenylenevinylene (DOO-PPV) showed that both regular and ultra-small effects are isotope dependent. This indicates that MFE response in OLED is mainly due to the hyperfine interaction (HFI). We also performed spectroscopy of the MFE including magneto-photoinduced absorption (MPA) and magneto-photoluminescence (MPL) at steady state conditions in several systems. This includes pristine Poly[2-methoxy-5-(2-ethylhexyl-oxy)-1,4-phenylene-vinylene] (MEH-PPV) films, MEH-PPV films subjected to prolonged illumination, and MEH-PPV/[6,6]-Phenyl C61 butyric acid methyl ester (PCBM) blend, as well as annealed and pristine C60 thin films. For comparison, we also measured MC and MEL in organic diodes based on the same materials. By directly comparing the MPA and MPL responses in films to MC and MEL in organic diodes based on the same active layers, we are able to relate the MFE in organic diodes to the spin densities of the excitations formed in the device, regardless of whether they are formed by photon absorption or carrier injection from the electrodes. We also studied magneto-photocurrent (MPC) and power conversion efficiency (PCE) of a 'standard' Poly (3-hexylthiophene)/PCBM device at various Galvinoxyl radical wt%. We found that the MPC reduction with Galvinoxyl wt% follows the same trend as that of the PCE enhancement. In addition, we also measured the MPC response of a series of OPV cells. We attribute the observed broad MPC to short-lived charge transfer complex species, where spin mixing is caused by the difference, Δg of the donor/acceptor g factors; whereas narrow MPC is due to HFI within long-lived polaron-pairs.

MAGNETIC FIELD EFFECT IN ORGANIC FILMS AND DEVICES by Bhoj Raj Gautam A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Physics Department of Physics and Astronomy The University of Utah August 2013Copyright © Bhoj Raj Gautam 2013 All Rights Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL STATEMENT OF DISSERTATION APPROVAL The dissertation of Bhoj Raj Gautam has been approved by the following supervisory committee members : Zeev Valy Vardeny , Chair 04/24/2013 Date Christoph Boehme , Member 04/24/2013 Date Oleg Starykh , Member 04/24/2013 Date Kyle Dawson , Member 04/24/2013 Date Ashutosh Tiwari , Member 04/24/2013 Date and by David Kieda , Chair of the Department of Physics and Astronomy and by Donna M. White, Interim Dean of The Graduate School. ABSTRACT In this work, we focused on the magnetic field effect in organic films and devices, including organic light emitting diodes (OLEDs) and organic photovoltaic (OPV) cells. We measured magnetic field effect (MFE) such as magnetoconductance (MC) and magneto-electroluminescence (MEL) in OLEDs based on several π-conjugated polymers and small molecules for fields B<100 mT. We found that both MC(B) and MEL(B) responses in bipolar devices and MC(B) response in unipolar devices are composed of two B-regions: (i) an ‘ultra-small' region at |B| < 1-2 mT, and (ii) a monotonic response region at |B| >∼2mT. Magnetic field effect (MFE) measured on three isotopes of Poly (dioctyloxy) phenylenevinylene (DOO-PPV) showed that both regular and ultra-small effects are isotope dependent. This indicates that MFE response in OLED is mainly due to the hyperfine interaction (HFI). We also performed spectroscopy of the MFE including magneto-photoinduced absorption (MPA) and magneto-photoluminescence (MPL) at steady state conditions in several systems. This includes pristine Poly[2-methoxy-5-(2-ethylhexyl-oxy)-1,4-phenylene-vinylene] (MEH-PPV) films, MEH-PPV films subjected to prolonged illumination, and MEH-PPV/[6,6]-Phenyl C61 butyric acid methyl ester (PCBM) blend, as well as annealed and pristine C60 thin films. For comparison, we also measured MC and MEL in organic diodes based on the same materials. By directly comparing the MPA and MPL responses in films to MC and MEL in organic diodes based on the same active layers, we are able to relate the MFE in organic diodes to the spin densities of the excitations formed in the device, regardless of whether they are formed by photon absorption or carrier injection from the electrodes. We also studied magneto-photocurrent (MPC) and power conversion efficiency (PCE) of a 'standard' Poly (3-hexylthiophene)/PCBM device at various Galvinoxyl radical wt%. We found that the MPC reduction with Galvinoxyl wt% follows the same trend as that of the PCE enhancement. In addition, we also measured the MPC response of a series of OPV cells. We attribute the observed broad MPC to short-lived charge transfer complex species, where spin mixing is caused by the difference, Δg of the donor/acceptor g factors; whereas narrow MPC is due to HFI within long-lived polaron-pairs. iv To my parents, and family TABLE OF CONTENTS ABSTRACT ............................................................................................................. iii LIST OF FIGURES ................................................................................................ ix ACKNOWLEDGEMENTS ................................................................................. xvi CHAPTER 1. INTRODUCTION ............................................................................................... 1 1.1 π-Conjugated Organic Semiconductors ....................................................... 1 1.2 Excitation Models for π-Conjugated Polymers ............................................ 5 1.3 Major Excitations in π-Conjugated Polymers .............................................. 6 1.3.1 Excitons ................................................................................................ 7 1.3.2 Polarons, Polaron Pairs and Bipolarons ............................................... 9 1.4 Charge Transport in Organic Semiconductors ........................................... 12 1.5 Organic Light Emitting Diodes .................................................................. 13 1.6 Magnetic Field Effects ............................................................................... 15 1.7 Organic Photovoltaics ................................................................................ 17 2. EXPERIMENTAL TECHNIQUES ................................................................. 20 2.1 Materials ..................................................................................................... 20 2.2 Organic Light Emitting Diodes Fabrication ............................................... 21 2.3 Organic Light Emitting Diodes Characterization ...................................... 26 2.3.1 Current-Voltage and Electroluminescence-Voltage Characteristics .. 27 2.3.2 Magnetoconductance and Magneto-electroluminescence .................. 28 2.4 Organic Photovoltaic Device Fabrication .................................................. 32 2.5 Organic Photovoltaic Device Characterization .......................................... 32 2.5.1 Current-Voltage Characteristics ......................................................... 32 2.5.2 Magneto-photocurrent Measurement ................................................. 35 2.6 Material Characterization ........................................................................... 36 2.6.1 Linear Absorption Measurement ........................................................ 36 2.6.2 Photoinduced Absorption Measurement ............................................ 38 2.6.3 Magneto-photoinduced Absorption Measurement ............................ 41 2.6.4 Photoluminescence and Magneto-photoluminescence Measurement 43 2.6.5 X-ray Diffraction Measurement ......................................................... 44 3. MAGNETIC FIELD EFFECT IN ORGANIC DIODES ............................... 48 3.1 Introduction ................................................................................................ 48 3.2 Experimental .............................................................................................. 49 3.3 Experimental Results and Discussion ........................................................ 50 3.3.1 Magnetoconductance Response in Organic Diodes at Ultra-small Fields .......................................................................................................... 50 3.3.2 Magneto-electroluminescence Response in Organic Diodes at Ultra-small Fields ........................................................................................ 62 3.3.3 Illumination Effect on Magnetoconductance Response of MEH-PPV Devices ..................................................................................... 67 3.4 Conclusion ................................................................................................. 72 4. MAGNETIC FIELD EFFECT IN ORGANIC FILMS ................................. 73 4.1 Magnetic Field Effect on Excited State Spectroscopies of π-Conjugated Polymer Films .................................................................................................. 73 4.1.1 Introduction ........................................................................................ 73 4.1.2 Experimental ...................................................................................... 77 4.1.3 Experimental Results .......................................................................... 79 4.1.3.1 Pristine MEH-PPV Films ........................................................... 79 4.1.3.2 Irradiated Pristine MEHPPV Films and Devices ....................... 81 4.1.3.3 Films and Devices of MEH-PPV/PCBM Blends ....................... 84 4.1.4 Discussion .......................................................................................... 86 4.1.5 Conclusion .......................................................................................... 92 4.2 Magnetic Field Effect Spectroscopy of C60 Based Films and Devices ...... 93 4.2.1 Introduction ........................................................................................ 93 4.2.2 Experimental ...................................................................................... 95 4.2.3 Experimental Results .......................................................................... 97 4.2.4 Discussion ........................................................................................ 105 4.2.4.1 Magneto-photoinduced Absorption; Narrow Component ........ 107 4.2.4.2 Magneto-photoinduced Absorption; Δg Mechanism ............... 109 4.2.4.3 Magnetoconductance and Magneto-photoluminescence .......... 111 4.2.5 Conclusion ............................................................................................ 111 5. ORGANIC PHOTOVOLTAIC DEVICES ................................................... 113 5.1 Efficiency Enhancement in Organic Bulk Heterojunction Photovoltaic Devices ........................................................................................................... 113 vii 5.1.1 Introduction ...................................................................................... 113 5.1.2 Experimental .................................................................................... 115 5.1.3 Experimental Results and Discussion .............................................. 117 5.1.3.1 Spin Enhanced Organic Bulk Heterojunction Photovoltaic Devices ................................................................................................. 117 5.1.3.2 Low Band Gap Organic Bulk Heterojunction Photovoltaic Devices ................................................................................................. 127 5.1.4 Conclusion ........................................................................................ 131 5.2 Magneto-photocurrent of Charge Transfer Complex in Organic Blends for Photovoltaic Applications ......................................................................... 132 5.2.1 Introduction ...................................................................................... 132 5.2.2 Experimental .................................................................................... 133 5.2.3 Experimental Results and Discussion .............................................. 134 5.2.4 Conclusion ........................................................................................ 143 6. CONCLUSION ................................................................................................ 144 APPENDIX: LIST OF PUBLICATIONS ......................................................... 148 REFERENCES..................................................................................................... 149 viii LIST OF FIGURES Figure 1.1 Chemical structures of π-conjugated organic semiconductors: Polyfluorine and P3HT are π-conjugated polymers whereas Alq3 and Pentacene are small molecular semiconductor ........................................................................................ 3 1.2 Electronic orbitals and bonds in sp2 hybridized carbon atoms, adapted from www.orgworld.de (a). A conjugated backbone with overlapped Pz orbitals (b). Chemical structure of trans-polyacetylene showing the alternation of carbon- carbon single and double bonds (c) ........................................................................ 3 1.3 Various photoexcitations in conjugated polymers: polaron excitation (charge manifold; uncorrelated) with P1 and P2 transitions on the left and exciton (neutral manifold; correlated) bands on the right ................................................................. 9 1.4 Energy level diagrams and possible optical transitions in polaron-pairs .............. 11 1.5 Energy levels and associated optical transitions of (a) positive, and (b) negative bipolarons .............................................................................................................. 11 1.6 Typical device structure of an organic light emitting diode .................................. 14 1.7 Schematic of different processes (recombination, dissociation and intersystem crossing) in an OLED device ................................................................................ 14 1.8 Summary of magnetic field effects on the singlet-triplet conversion of polaron pairs and physical effect yield with B ................................................................... 17 1.9 Typical device structure of BHJ OPV solar cell ................................................... 19 1.10 Energy level diagram for typical OPV showing different electronic processes, figure adapted from Ref. [37] ............................................................................... 19 2.1 Chemical structures of (a) H-DOO-PPV, (b) D-DOO-PPV, (c) C-13-DOO-PPV, (d) Rubrene, (e) PFO, (f) MEHPPV, (g) RR P3HT, (h) RRa P3HT, (i) PTB7, (j) C60 , (k) PC61BM, and (l) PC71BM .................................................................. 22 2.2 ITO pattern on the substrate after etching (a). Top view of completed OLED device (b). The typical OLED device structure (c) ................................... 25 2.3 Typical I-V and EL-V characteristics of OLED device based on MEHPPV as an active layer at 10 K .......................................................................................... 28 2.4 The experimental set up for measuring the organic magnetic field effect in films and devices ............................................................................................................ 30 2.5 Typical MC and MEL responses of an OLED based on MEHPPV as the active organic interlayer, measured at 10 K .................................................................... 31 2. 6 Standard ‘AM1.5 spectrum', under which the integrated illumination intensity is 100 mW/cm2 ......................................................................................................... 33 2.7 The experimental set-up for measuring the I-V response of an OPV cell ............ 34 2.8 Typical I-V characteristics of PTB7/PC70BM OPV cell with 3 wt% of 1,8- diiodooctane, under ‘AM 1.5' illumination spectrum .......................................... 34 2.9 MPC(B) response of PTB7/PC70BM-based OPV cell with 3 wt% of 1,8- diiodooctane, measured at room temperature ....................................................... 37 2.10 Typical optical density spectrum of MEHPPV film ............................................. 38 2.11 The experimental setup for measuring the PA spectrum ...................................... 40 2.12 The photoinduced absorption spectrum of an irradiated MEH-PPV film. The PA bands P1 and T+P2 are denoted ........................................................................... 42 2.13 Magneto-photoinduced absorption response of irradiated MEHPPV film measured at50 K ................................................................................................... 44 2.14 Photoluminscence spectrum of Pristine MEHPPV film at 50 K .......................... 45 2.15 Magneto-photoluminscence response of pristine MEHPPV film measured at 50 K ..................................................................................................................... 45 2.16 X-ray diffraction patterns of a C60 film, where the background scattering was removed for clarity. The numbers represents various (hkl) Bragg bands ............. 47 3.1 Magnetoconductance (MC) response vs. field, B in bipolar organic diodes based on: three isotopes of DOO-PPV Panel (a) shows MC(B) for B<50 mT; whereas panel (b) shows the normalized MC(B) measured with high field resolution, for B<3 mT (some MC responses are shifted vertically for clarity); MCmax is the saturation MC value at large B. ΔB is the HWHM for the normal MC(B) response, x as defined in (a); whereas MCmin and Bm are for the USMFE response, as defined in (b). Panel (c) summarizes Bm vs. ΔB for the MC(B) responses in (a) and (b); the straight line is guide to the eye ............................................................................. 53 3.2 Magnetoconductance (MC) response vs. field, B in bipolar organic diodes based on MEH-PPV, PFO (MCx3), and rubrene RBRN; (MCx8). Panel (a) shows MC(B) for B<50 mT; whereas panel (b) shows the normalized MC(B) measured with high field resolution, for B<3 mT (some MC responses are shifted vertically for clarity). Panels (c) summarizes Bm vs. ΔB for the MC(B) responses in (a) and (b) ......................................................................................... 55 3.3 Normalized MC(B) response of a bipolar diode based on D-DOO-PPV for B<0.5 mT at (a) various bias voltages at T=10 K, and (b) various temperatures at V= 3.4 Volt; MCmax is defined in Figure 3.1. The insets in (a) and (b), respectively summarize MCmin/MCmax at various voltages at 10 K, and various temperatures at 3.4 Volt ........................................................................................ 56 3.4 Normalized MC(B) response for (a) B<30 mT, and (b) B<2 mT of hole- and electron-only unipolar diodes based on MEH-PPV, measured at room temperature and V=3 Volt and 20 Volt, respectively. For clarity, the MC(B) responses are multiplied by a factor in (b) ........................................................... 57 3.5 Normalized MC(B) response for (a) |B| < 60mT, and (b) |B| < 3mT of hole-only unipolar diodes based on H- and C13-DOO-PPV, measured at room temperature ........................................................................................................... 59 3.6 Calculated spin energy levels and magnetoconductance. (a)Example of calculated spin energy levels vs. B for a spin pair with isotropic HFI; a1=3, a2=3 mT, and J=0. Note the multiple level-crossing at B=0. (b) Calculated MC(B) response for a SP with axially symmetric HFI averaged over all magnetic field directions. The isotropic HFI is the same as in (a). The anisotropic HFI component is azz=0.15ai for the respective SP constituent ........................................................................... 63 3.7 Magneto-electroluminescence (MEL) response vs. field, B in bipolar organic diodes based on MEHPPV polymer as an active layer. Panel (a) shows MEL(B) for B<60 mT; panel (b) shows the MEL(B) measured with high field resolution, for B<3 mT .......................................................................................................... 65 3.8 Magneto-electroluminescence (MEL) response vs. field, B in bipolar organic diodes based on rubrene small molecule as active layer. Panel (a) shows MEL(B) for B<60 mT; panel (b) shows the MEL(B) measured with high field resolution, for B<3 mT .......................................................................................................... 66 3.9 Magneto-conductance MC(B) response in bipolar, hole-only and electron-only unipolar organic diodes based on MEHPPV. The latter responses are multiplied by a factor of 100 and 10, respectively ................................................................. 69 xi 3.10 Magneto-conductance MC(B) response in pristine and UV irradiated (20 minutes) MEH-PPV OLED, measured at 10 K ................................................................... 70 3.11 MC(B) response of (a) hole-only and (b) electron-only unipolar devices at 10 K, illuminated with 532 nm laser for 30 min at different power ............................... 71 4.1 Schematic illustration of the magnetic field dependent pump-probe PA processes. (a) The pump beam with above gap photon energy hυL excites the polymer MEH-PPV to the singlet exciton (SE) level (S0S1). The SE relaxes via intersystem crossing to a triplet exciton (TE) or ionizes into separate charges forming polaron pair, PP (S1X0). The steady state density of the X species is controlled by the spin dependent decay coefficient, κ. The incandescent probe beam monitors the photoinduced absorption, PA (X0X1, PAX), which is proportional to the X0 steady state density. In a magnetic field B>0, X0 splits according to its spin multiplicity, and the decay rate of each spin sub-level becomes field dependent, resulting in a B-dependent density and PAX (thus forming MPAX) ..................................................................................................... 76 4.2 Excited state spectra (PA and PL) and magnetic field effects in pristine MEH- PPV films. (a) The triplet PA band, PAT at B=0 and 100 mT (black and red lines, respectively), respectively, generated using a laser excitation at hυL=2.54 eV @ IL=200 mW/cm2, and their difference spectrum ΔPAT=[PAT(100mT)-PAT(0)] (blue line). The region near the peak is magnified (within a circle). Right inset: PL spectrum at B=0 (black line) and 100 mT (red line), respectively. The lines in the circles show the data on a higher resolution scale. (b) MPAT(B) response measured at 1.37 eV probe, for various laser excitation intensities (normalized). (c) MPL(B) response measured at 2.05 eV probe for various laser excitation intensities (normalized). (d) Model calculations of MPAT(B) response using the TE mechanism (blue line, corresponds to the 10 mW data in (b)) and TTA mechanism (green line, corresponds to the 400 mW data in (b)) mechanisms; see text. (e) Model calculation of MPL(B) response using the model of singlet exciton quenching by TE (SE-TE collision, see text) ........................................... 78 4.3 Excited state spectra and magnetic field effects in UV irradiated MEH-PPV film and in organic light emitting diode. (a) PA spectrum at IL=100 mW/cm2 for B=0 (black line) and B=100 mT (red line), respectively, and their difference spectrum, ΔPA=[PA(100mT)-PA(0)] (blue line) in MEH-PPV film. (b) MPA(B) response measured at 1.4 eV probe for various laser excitation intensities (normalized). (c) MEL(B) and MC(B) responses in MEH-PPV diode. (d) Model calculations of MPAPP(B) response in MEH films using the PP mechanism (see text). (e) MPA(B) response at 1.1 eV probe up to B=1.5 mT (filled squares) and B=60 mT (blue line, inset) .................................................................................................... 82 4.4 Excited state spectra and magnetic field effects in MEH-PPV/PCBM film and diode. (a) PA spectrum of MEH-PPV film at IL=mW/cm2 for B=0 (black line) and B=15 mT (red line), respectively, and their difference spectrum, ΔPA=PA(15mT)-PA(0) (blue line). (b) MPA(B) response measured at 1.37 eV xii probe for various laser excitation intensities (normalized). Inset: high resolution data, showing USMPA peaks at |B|~0.1 mT. This data was measured upon shielding from the earth magnetic field and any stray field. (c) MC(B) response in a diode at various bias voltages, V. (d) and (e) Model calculations of MPAPP(B) and MC(B) response, respectively, using the ‘Δg + HFI' mechanism (see text, Section 4.1.4) ........................................................................................................ 85 4.5 The X-ray diffraction pattern of annealed and pristine C60 films in the range (a) 2θ= 6-250, (b) 2θ= 8-130; the miller indices are denoted on the Bragg scattering bands. The inset in (b) shows the chemical structure of C60. TEM images of annealed (c) and pristine (d) C60 films; the grey grains are C60 microcrystallites. The scale bar is 50 nm. Also shown are the grain size distributions extracted from the TEM images for the annealed (e) and pristine (f) C60 films ........................................................................................................... 98 4.6 Photomodulation spectra of annealed (a) and pristine (b) C60 films at T=50 K and IL=0.2W/cm2 for B=0 (black lines) and B=180 mT (red lines). The blue negative lines are the difference spectra ΔPA=PA(B=180 mT)-PA(0) .............. 100 4.7 MPA(B) response of an annealed C60 film at various pump excitation intensities, measured at photon energy E=1.8 eV and T=50 K. (b) The spectra ΔPA(B1,B2,E) for B1=0, B2=20 mT (black line, lower curves), and B1=20 mT, B2=180 mT (blue line, upper curves) for IL=1.5 W/cm2. The smooth green and red lines through the data are to guide the eye, and show the TE- and polaron-related MPA bands, respectively .................................................................................... 101 4.8 MC(B) response of an annealed C60 diode for various bias voltages measured at T=10 K. (a) high resolution for |B|<0.2 T; (b) low resolution for |B|<1 T. (c) MC(B) response of devices based on 13C-rich C60 (black line) compared with that of devices based on regular C60 (red line) for |B|<40 mT ........................... 104 4.9 PL spectrum (a) and MPL(B) response (b) of annealed C60 film at T=50 K ...... 106 4.10 Model fitting for MPA(B) of C60. (a) Low field, |B|<40 mT. The blue line is calculated based on the TE mechanism (see text); the black points are measured MPA(B) taken from Fig. 4.7 (a). (b) Intermediate field, |B|<0.2 T. Blue line: calculated using the ‘Δg mechanism' (see text); black line: measured MPA(B) respectively ......................................................................................................... 109 5.1 J-V characteristics of P3HT:PCBM OPV devices at different percentages of PCBM under AM 1.5 illumination ..................................................................... 118 5.2 J-V characteristics of OPV solar cells of pristine P3HT/PCBM blend (η = 3.4%, Black line), the blend doped with 3 wt% galvinoxyl radicals (η = 4.0%, Red line) and the blend doped with 3 wt% precursor (η = 2.8%, Blue line) under AM1.5 xiii ‘sun illumination' condition. The inset shows the galvinoxyl molecular structure .. ............................................................................................................................. 119 5.3 The change in OPV device properties with galvinoxyl-additive concentration; Jsc (triangles) and η (squares) are shown versus galvinoxyl wt% in the P3HT/PCBM blend. η of OPV devices doped with galvinoxyl precursor that does not possess spin 1/2 radical is also shown for comparison (circles). ..................................... 120 5.4 MPC response of OPV devices doped with galvinoxyl up to field, B of 190 mT. The inset summarizes the MPC value at 190 mT versus galvinoxyl wt%.......... 121 5.5 The UV/Vis absorption spectrum of pure galvinoxyl (dash-dot line), pristine (dashed line) and doped (solid line) P3HT/PCBM blend (a). The EQE spectrum of OPV solar cells based on pristine (dashed line) and galvinoxyl-doped (solid line) P3HT/PCBM blend (b). The XRD pattern of pristine (green dash) and doped (red solid) P3HT/PCBM films (c). PL spectrum of pristine (dashed line) and doped (solid line) P3HT/PCBM. The phonon replicas are assigned. Norm., normalized (d) ......................................................................... 123 5.6 The spin exchange mechanism where the photogenerated PP at the D-A domain interface changes its spin configuration from singlet to triplet augmented by the galvinoxyl spin 1/2 radical (a). The calculated HOMO, LUMO, and SOMO levels of P3HT, PCBM, and galvinoxyl that show a clear resonance between the radical and acceptor LUMO levels (b). .............................................................. 126 5.7 Linear absorption spectrum of P3HT and PTB7 polymer .................................. 128 5.8 Linear absorption spectrum of PTB7 and its blend with PC71BM...................... 128 5.9 J-V characteristics of PTB7/ PC71BM device under AM 1.5 illumination ......... 129 5.10 J-V characteristics of PTB7/PC71BM OPV devices at different percentage of dio under AM 1.5 illumination, inset shows the chemical structure of dio .............. 130 5.11 The PCE and MPC(B) response of RRa P3HT/PC61BM (1:2) based OPV cell . 135 5.12 The PCE and MPC(B) response of MEH-PPV/PC61BM (1:4) based OPV cell . 136 5.13 The PCE and MPC(B) response of RR P3HT/PC61BM (1.2:1) based OPV cell.137 5.14 The PCE and MPC(B) response of PTB7/PC71BM (1:1.5) based OPV cell ..... 138 5.15 The PCE and MPC(B) response of PTB7/PC71BM (1:1.5) based OPV cell with 3 wt% of 1,8-diiodooctane .................................................................................. 139 xiv 5.16 PA spectra of a ~100 nm thick film of P3HT/PCBM (1.2:1 by weight) blend at B=0 (black line) and B=150 mT (red line). The difference ΔPA (enlarged) is plotted as a blue line (a). The MPA(B) response monitored at E1=0.35 eV (b) . 140 xv ACKNOWLEDGEMENTS I would like to express my utmost gratitude to my advisor, Prof. Zeev Valy Vardeny, for his sincere and unfailing support throughout my thesis and throughout my years at the University of Utah. His knowledge, patience, and encouragement were instrumental to the completion of this study. I must express my sincere thanks to Prof. Eitan Ehrenfreund for his theoretical guidance and Prof. Tho Nyugen for his research guidance. I would also like to thank my supervisory committee members, Professors Christoph Boehme, Oleg Starykh, Kyle Dawson, and Ashutosh Tiwari for their encouraging discussions and suggestions. I would like to thank Dr. Randy Polson, Dr. Mathew Delong, and Mr. Leonard Wojcik for their help and suggestions. My sincere thanks is extended to my current group members, Tek Prasad Basel, Dr. Dali Sun, Ella Olejnik, Uyen Huynh, Yaxin Zhai, Ryan McLaughlin, and Peter Peroncik for their help, support, and suggestions. I would not forget my former group members Dr. Bill Pandit, Dr. Golda Hukic Markosian, Dr. Ye Zang, Dr. Sanjeev Singh, and Dr. Maria Navas for their help. I am indebted to my family for their love, care, support, and inspiration. I must express my gratitude to Dipa Sharma, my wife, for her continuous support and encouragement. Also, thanks goes to my brother in law Deepesh Poudel and friend Sajjan Koirala for their valuable suggestions. CHAPTER 1 INTRODUCTION For a long time, organic materials have been associated with electrical insulators. Research on organic semiconductors was boosted after the discovery of the highly conducting oxidized iodine-doped polyacetylene [1]. Although to date, inorganic semiconductors are still the most popular materials in the electronic industry, the unique properties of organic semiconductors such as electroluminescent properties, flexibility, solubility, light weight, low cost, and easily modified band gap make these semiconductors very attractive for a number of novel optoelectronic applications such as: organic light emitting diodes (OLEDs) [2, 3], organic field effect transistors (OFETs) [4], organic photovoltaic cells (OPVs) [5, 6], organic spin valves [7, 8], thin film magnetometers [9], biological sensors, etc. In this chapter, we will give a brief review of π-conjugated organic semiconductors and their use in OLEDs and OPV cells. The focus of this work will be the magnetic field effect in organic semiconductor films and devices. 1.1 π-Conjugated Organic Semiconductors Π-conjugated organic semiconductors are divided into two groups based on their molecular weight, namely polymers and small molecules. Chain-like macromolecules 2 with high molecular weight (>1000 g/mol) are polymers that are soluble and can be deposited easily, whereas materials with molecular weight less than 1000 g/mol are small molecules and are usually deposited by thermal evaporation. Both of these groups have a cocommon π-conjugated chemical structure, as shown in Figure 1.1. Π-conjugated semiconductors are unsaturated carbon compounds with alternating single and double bonds between the carbon atoms, as shown in Figure 1.2. The sp2pz hybridization causes three electrons to establish strong planar σ-bonds with neighboring atoms and one electron to be bound in π-bond perpendicular to the polymer backbone. The π-electrons are delocalized over many carbon atoms along the chain, giving the relatively high conducting properties [10]. These delocalized electrons occupy the bonding π-orbitals while antibonding π*-orbitals remain empty. The bonding π-orbitals form the highest occupied molecular orbitals (HOMO) and antibonding π*-orbitals form the lowest unoccupied molecular orbitals (LUMO), which are roughly equivalent to the inorganic semiconductor's valence and conduction band edges, respectively. The energy gap between HOMO and LUMO lies in the range 1.4-3.0 eV in most of the organic semiconductors, which makes them promising for applications in optoelectronic operation in the visible spectral range By changing the extent of delocalization, the gap between occupied and empty states can be altered, which makes them interesting in both academic and industrial research. The π- electrons are delocalized over many carbon atoms over the chain and hence, the quantum mechanical wave function is confined to a single chain. Π-conjugated organic semiconductors are often treated as one-dimensional systems with half-filled electronic bands as there is one π-electron per carbon atom. By taking an account of either electron-phonon interaction or electron-electron interactions 3 Figure 1.1. Chemical structures of π-conjugated organic semiconductors: Polyfluorine and P3HT are π-conjugated polymers whereas Alq3 and Pentacene are small molecular semiconductors. (a) Figure 1.2. Electronic orbitals and bonds in sp2 hybridized carbon atoms, adapted from www.orgworld.de (a). A conjugated backbone with overlapped Pz orbitals. (c) Chemical structure of trans-polyacetylene showing the alternation of carbon-carbon single and double bonds (b). 4 (b) (c) Figure 1.2. Continued 5 among the π-electrons, the formation of the band gap can be explained. 1.2 Excitation Models for π-Conjugated Polymers Excitations in π-conjugated polymers are described by using several models. Su, Schrieffer, and Heeger proposed a model, named SSH model, for trans-polyacetylene (t- (CH)x), based on tight binding approximation calculation by taking an account of electron phonon interaction and neglecting the electron-electron interaction [11]. In this model, they applied a semiclassical Huckel Hamiltonian. The Hamiltonian contains the lattice kinetic energy, which is treated classically, and the electron-phonon interaction, which is treated quantum mechanically, as written in Equation 1.1: ( ) ( ( ))( ) 2 ( ) 2 , 0 1 1, , , 1, 2 2 1 n s n n n s n s n s n s n SSH n n n n t u u C C C C dt M du u u k H                 (1.1) where t0 is the hopping integral between the nearest neighbors for an undistorted chain, α is the electron lattice coupling constant, and  n s C , and n s C , are the creation and annihilation operators of an electron on site n with spin s. k is the spring constant due to π-electrons and un is the deviation of nth site from the equilibrium position in an undistorted chain with equal distance between sites. According to the SSH model, dimerization caused by strong electron-phonon interaction lowers the system energy and creates an energy gap Eg=4αu where u is the dimerization amplitude in equilibrium. Thus, the occupied electronic states in equilibrium are lowered, resulting in a more stable configuration. Therefore, the system no longer acts 6 as a one-dimensional metal, but instead behaves as a semiconductor with a direct energy gap. On the other hand, the Hubbard model that includes electron-electron interaction and 3D intrachain coupling can also explain the energy levels of charged and neutral excitations. Although this model includes the coulomb repulsion of two electrons on the same site, it ignores the electron-phonon interaction, which is quite strong in the polymer system. The model which includes both interactions, i.e., combination of SSH and the Hubbard model, is more realistic to explain the energy levels of excitations in the class of π-conjugated polymers. The Pariser-Parr-Pople (PPP) is such a model [12]. 1.3 Major Excitations in π-Conjugated Polymers Two kinds of electronic excited states (excitations), namely charged (polarons) and neutral (excitons), are dominant in π-conjugated polymers. Upon photoexcitation (with above-gap photon energy), neutral, spinless excitations called singlet excitons (SE) are generated. The SE may either radiatively recombine; or convert into long-lived neutral excitations, i.e., triplet excitons (TE) via intersystem crossing; or separate into positive and negative charge excitations (polarons), some of which may form long-lived polaron pairs (neutral excitations). On the other hand, upon electrical excitation, charged excitations are injected; these may recombine to form neutral excitations or other types of charged excitations [13]. In the following, we summarize the main properties of the charged and neutral photoexcitations. 7 1.3.1 Excitons Excitons are electron-hole pairs that are bound through their mutual coulombic interaction. Upon photon absorption, an electron is promoted from lower energy level to higher energy level and an exciton is generated. This excitation causes structural relaxation of the surrounding geometry, which leads to an exciton binding energy Eb. Typical Eb is between 0.3-0.5 eV in most π-conjugated polymers. Depending upon the mutual spin configuration, an electron and hole in an exciton may form singlet or triplet state with total spin 0 or 1, respectively; both species are neutral. The wave function describing two particle systems (exciton) is asymmetric in spin and electronic coordinates and can be obtained from Slater determinant: ( ) ( ) ( ') ( ') ( ) ( ) ( ') ( ') 2 1             j j j j i i i i r r r r   (1.2) where ψi (r) and σi (r) represent the electronic and spin part of wave function. The wave functions that have a different total quantum number, S, constructed from the above equation are: [ (1) (2) (1) (2)][ (1) (2) (2) (1)] 2 1 1 2 2 1 0              S (1.3) [ (1) (2) (1) (2)][ (1) (2) (2) (1)] 2 1 1 2 2 1 1              S (1.4) 8 [ (1) (2) (1) (2)][ (1) (2)] 2 1 1 2 2 1 1           S (1.5) [ (1) (2) (1) (2)][ (1) (2)] 2 1 1 2 2 1 1           S (1.6) where ↑and ↓ represent the spin up and spin down projection of χ. Singlet and triplet energy levels are degenerate in the noninteracting case. However, in the presence of spin-spin interaction such as an exchange interaction, they are nondegenerate with triplet taking the lower energy. The energy bands in excitons are shown in the right panel of Figure 1.3. Although a singlet exciton is formed immediately after photoexcitation, it may convert into a long-lived triplet exciton within ~10 ns or less via intersystem-crossing that results by a spin flip of one of the electrons involved in the exciton due to spin orbit coupling, hyperfine interaction, or the existence of radical impurities on the chains. The excited singlet state may recombine radiatively by emitting light in the form of fluorescence (PL). This process is usually fast with a lifetime of ~100 picoseconds. As the optical transition from the triplet lower state to the ground state is forbidden, the radiative emission from the excited triplet state, namely phosphorescence (PH), is usually weaker in organic materials. The transition may be possible if one of the two paired electrons spins flips due to spin orbit interaction. However, the optical transition of the triplet exciton is relatively small, resulting in long lifetime, of the order of milliseconds [14]. 9 Figure 1.3. Various photoexcitations in -conjugated polymers: polaron excitation (charge manifold; uncorrelated) having P1 and P2 transitions on the left, and exciton (neutral manifold; correlated) bands on the right. 1.3.2 Polarons, Polaron Pairs and Bipolarons The interaction between neighboring molecules in an organic material in solid state is due to Van der Waals forces, which are much weaker than the covalent and ionic bonds in inorganic materials. As a consequence of this, organic materials are less rigid than inorganic materials. Therefore, the charge carrier that propagates in organic material is able to distort the host material and thus form a quasi-particle called a polaron. The polaron is charged negative (P-) or positive (P+), and has spin ½. It has two symmetrical, localized states within the gap and has two allowed below-gap optical transitions P1 and P2, as shown in Figure 1.3. Doping-induced absorption, charge injection through metallic electrodes, photo-doping, i.e., exciting the sample with photon Nonradiative Decay 10 absorption, are some methods for creating polarons in organic materials. Polaron transport from one chain to another is usually described by the hopping process between the localized states. A Polaron pair (PP) is a bound pair of two oppositely charged polarons (P+ and P-), formed on two adjacent chains. The PP binding energy is mainly Coulombic. The PPs are the intermediate step between free polarons and excitons. These are the prerequisite for the formation of singlet and triplet excitons in OLEDs, and hence, their related physics is very important for device applications. In optical excitation, PPs are generated by the relaxation of higher energy singlet excitons. The species keeps the original spin 0 configuration and is hence dubbed a geminate pair. Upon electrical excitation, the electrons and holes that are injected into the active layer via the metal electrodes capture each other by Coulomb interaction and form PPs; these are nongeminate PPs. The nongeminate PPs can have spin 0 or 1 with high probability of having triplet configuration because of the degeneracy of the spin sublevels (in fact 3 to 1). The energy levels and possible transitions for PPs are shown in Figure 1.4. When two polarons with the same charge come together with opposite spins on the same site, the resulting species with energy lower than two separate polarons is called a bipolaron. A bipolaron can either be doubly positive (BP+ +) or doubly negative (BP- -). A bipolaron has two in-gap electronic states (like the polaron), but has only one allowed optical transition, as shown in Figure 1.5. 11 P+ P+ P- P- Figure 1.4. Energy level diagrams and possible optical transitions in polaron-pairs. Figure 1.5. Energy levels and associated optical transitions of (a) positive, and (b) negative bipolarons 12 1.4 Charge Transport in Organic Semiconductors Most of the organic semiconductor films are grown either by vacuum deposited small molecules or from solution processed polymers. Irregular packing of molecules due to vacuum deposition or spin coating causes energetic disorder in the HOMO and LUMO levels. The distribution of chain length, kinks, and twists present in polymer chains also causes structural, and hence, energetic disorder. Therefore, the HOMO and LUMO levels are distributed, and the band conduction transport concept (i.e., relaxation time approximation) does not apply to organic semiconductors. As a result of the energetic disorder in organic semiconductors, charges are localized on molecular sites. Charge transport occurs via phonon-assisted tunneling or hopping [15] between the localized states in disordered materials and depends strongly on the parameters like temperature, electric field, traps present in the material, etc. For hopping, charge transport is very poor. The probability to hop from one energetic site to another, described by Miller-Abrahams formalism [16], is given by: p= υ0exp(2αRij){ j i j B j i E E E Ei K T E E     1 exp( ) (1.7) where α is the inversion localization radius of the electronic wave function, Rij is the distance between the localized sites i and j, and KB is the Boltzmann constant. The disorder in the position and energy of hopping sites leads to much smaller mobility than via delocalized band states, as in inorganic semiconductors. 13 1.5 Organic Light Emitting Diodes The first organic light emitting diode (OLED) based on the small molecule material (Alq3) was demonstrated by Tang and Van Slyke in 1987 [17] and a polymer- based OLED was demonstrated three years later by Burroughs et al [2]. Extensive research activities were then carried out to optimize the device parameters and understand the physical processes that occur during OLED operation. Charge carrier injection, charge carrier transport, polaron pair, followed by exciton formation and exciton decay (light emission) are the four important electronic processes that occur when OLED is in operation. Figure 1.6 shows a typical OLED structure. Here, the large work function material PEDOT:PSS is used as a hole transport layer whereas the low work function metal calcium is used as an electron transport layer. Electroluminescence in the OLEDs results from recombination of polaron pairs (PP) in the spin singlet configuration. The electrons and holes that are injected into the active layer via the metal electrodes can form loosely bound polaron pairs, which are the precursor of excitons. Following the PP generation, they may undergo three possible processes. They (i) may combine to form excited state singlet excitons (SE) and triplet excitons (TE), (ii) can dissociate into free charge carriers again, or (iii) can exchange spins via intersystem crossing (ISC). SEs may decay radiatively, resulting in electroluminescence. The long-lived TEs may decay nonradiatively or show delayed fluorescence via the process of triplet-triplet annihilation. The schematic of the different electronic processes in OLED such as recombination, dissociation, and intersystem crossing are shown in Figure 1.7. 14 Figure 1.6. Typical device structure of an organic light emitting diode. Figure 1.7. Schematic of different processes (recombination, dissociation, and intersystem crossing) in an OLED device. Figure adapted from Ref. [18]. 15 1.6 Magnetic Field Effects Magnetic field effects include the field-induced changes in chemical and bio-chemical reaction yields, magneto-conductance, magneto-electroluminescence, magneto-phosphorescence, magneto-photoconductance, etc. and have been intensively studied over the recent years [19-33]. Various mechanisms that account for the magnetic field effects (MFE) have emerged from these studies. This includes (a) spin-mixing by the hyperfine (HF) interaction within polaron pairs (PP) and bipolarons, (b) the difference, Δg, in the electron and hole g-factors, and (c) a number of mechanisms that involve triplet excitons (TE). The polaron pair mechanism accounts for the spin mixing between the singlet-triplet (S-T) polaron pairs that can be influenced by weak magnetic interactions such as Zeeman and hyperfine. If neither spin couples to any magnetic nuclei (hyperfine coupling constant Ai is zero for both of them), then to have S-T conversion, they must have different g values. The precession frequency of the two individual spins transforms singlet pair state to triplet pair state, and vice versa, driven by the difference in precision frequencies, i.e., (g1-g2)μBΔgB/ħ. It can be seen that this mechanism is applicable only in the presence of external magnetic fields (B), and that the spin mixing frequency increases with increase in external magnetic field. In contrast, in the presence of hyperfine interaction, S-T conversion at zero B may occur between the singlet state and all three triplet sublevels. With increasing B field, the hyperfine interaction (HFI)-driven spin mixing decreases and saturates at fields higher than the hyperfine coupling constant. If we take into account the exchange interaction, J, there is no S-T conversion at zero B, as S and T levels are separated by exchange energy (2J). However, with increasing B field, 16 S and T+1 or T-1 can exchange the spin at a certain field called the level-crossing field (BLC). S-T conversion rate increases suddenly at BLC through the hyperfine field. Figure 1.8 shows the magnetic field effects on the S-T mixing of the polaron pairs and the physical effect yield with B [31]. The bipolaron model proposed by Bobbert et al. [24] based on the experimental observation of magnetic field effect in unipolar devices suggests the influence of magnetic field on the mobility of charge carriers and hence the current. The model is based on the competition between B-dependent bipolaron formation and B-independent hopping to empty sites. Using Monte Carlo simulation, two different line shapes in agreement with the experimental observation were shown to exist. A crucial point of this model is that carrier mobility and current density in a device are directly affected by the probability of magnetic field dependent bipolaron formation. Desai et al. [22] proposed another model to explain the magnetic field effect observed in organic diodes. These authors considered the reduction in carrier mobility by polaron scattering from triplet excitons. They assumed that magnetic field acts on the intersystem crossing of singlet and triplet excitons, thereby decreasing the triplet concentration, consequently decreasing the scattering and, in turn, increasing the mobility. Recently, we observed the magneto-photoinduced absorption (MPA) and magneto magneto-photoluminscence (MPL) in organic polymer films. We explained the MPA observation in terms of triplet-triplet annihilation (TTA) and spin-mixing among the triplet spin sublevels, in addition to the spin mixing in PP and Δg mechanism that are viable in polymer/fullerene blends. Since the PL is affected by the nonradiative decay 17 Figure 1.8. Summary of magnetic field effects on the singlet-triplet conversion of polaron pairs and physical effect yield with B. channel of singlet excitons' collisions with triplet excitons (TE), of which density varies with B, MPL (B) can be explained by the magnetic field dependent TE density. 1.7 Organic Photovoltaics Tang was the first to implement a bilayer heterojunction solar cell device [34] in 1986 and achieved 1% power conversion efficiency (PCE). After this discovery, intensive research on solar cells comprising organic semiconductors has been carried out and 10.6% PCE [35] including tandem structure is the record value to the date. 18 The working principle of bulk heterojunction (the active layer consists of donor and acceptor) organic photovoltaics (OPVs) starts with photoexcitation of donor material. The photons that are absorbed in the active layer excite the polymer and form a coulombically bound electron-hole pair, known as an exciton. Dissociation of the photogenerated excitons is facilitated by the energy level difference between the LUMO of the donor and acceptor, as well as between their HOMO levels. The exciton diffuses to the donor-acceptor (D-A) interface within few picoseconds [36, 37], and forms a charge transfer exciton (CT) upon arrival [38, 39]. Initially, the CT excitons separate into more loosely-bound polaron pairs (PPs), the intermediate species that exist at the donor/acceptor interface. Subsequently, PPs separate into "free" electrons and holes that are available for transport. In the blend, the donor acts as electron donor and hole transporter, whereas the fullerene derivative is an electron acceptor and transporter; thus, the photogenerated electrons and holes can be readily collected at the anode and cathode, respectively. The typical device structure of bulk heterojunction (BHJ) OPVs is shown in Figure 1.9. The charge photogeneration process upon photoexcitation in BHJ solar cell devices is shown in Figure 1.10. 19 Figure 1.9. Typical device structure of bulk-heterojunction OPV solar cell. Figure 1.10. Energy level diagram of donor and acceptor in typical OPV, and the related electronic processes. The figure is adapted from Ref [37].CHAPTER 2 EXPERIMENTAL TECHNIQUES In this chapter, we describe most of the experimental techniques used in this PhD thesis. In particular, we focus on the fabrication of organic light emitting diodes (OLEDs) and bulk heterojunction organic photovoltaic (OPV) devices; and experiments performed using the magnetic field effect (MFE) in OLEDs, organic films, and OPV devices. 2.1 Materials The materials used in this thesis are either polymers or small molecules. Polymers are chain-like macromolecules with high molecular weight (>1000g/mol). They are soluble in organic solvents and can be deposited easily. Small molecules have molecular weight less than 1000g/mol and are usually deposited using thermal evaporator in vacuum. Isotopes of Poly (dioctyloxy) phenylenevinylene (DOO-PPV), Poly (2-methoxy-5-(2-ethylhexyl-oxy)-1,4-phenylene-vinylene)) (MEH-PPV), Rubrene, Polyfluorene (PFO), Regio-Regular -Poly-(3-hexylthiophene) (RR P3HT), Regio-Random-Poly-(3-hexylthiophene) (RRa P3HT), Poly-thienothiophene benzodithiophene 7 (PTB7), C60 molecule, [6,6]-Phenyl C61 butyric acid methyl ester (PC61BM), and [6,6]-Phenyl C71 butyric acid methyl ester (PC71BM) are materials used in this thesis, which are tabulated 21 in Figures 2.1(a)- (l) along with their chemical structures. Isotopes of DOO-PPV were synthesized by chemist Leonard Wojcik in our lab. RR P3HT was supplied by Plextronics; it has excellent properties compared to other commercial suppliers. MEH-PPV, PFO, C60, PC61BM, and PC71BM were purchased from American Dye Source (ADS). RRa P3HT and Rubrene were from Sigma Aldrich and PTB7 was from 1-Material. The synthetic reagents and solvents were procured from Aldrich Chemical as reagent grade and used as received. To prevent oxidation and other possible material contaminations, all handling processes were done in an inert nitrogen (N2) atmosphere inside a glove box with oxygen level less than 0.7 ppm. 2.2 Organic Light Emitting Diodes Fabrication The typical OLED device that we have investigated consists of a thin film of organic layer sandwiched between two nonmagnetic electrodes. A glass substrate partially coated with Indium Tin Oxide (ITO) with resistivity 8-12 Ω/cm was purchased from Delta Technologies. ITO is used as an anode to inject holes into the organic layer, because of its high work function (4.8-5.1 eV). To detect the light coming out of an OLED, a transparent electrode must be also used. As ITO has high transparency (>85%), it is suitable to be used as a transparent electrode. Patterning of ITO was done using either photolithography or a tape as an ‘etch mask'. For regular size (1 mm x1 mm) devices, the portion of the substrate to be used as bottom electrode was covered with ‘nail polish' and the rest was covered with a tape. The substrate was then immersed into a solution of hydrochloric acid (80% by volume) and water (20% by volume) for 10 minutes, for etching the exposed portion of the ITO. After 22 (a) Protonated Poly(dioctyloxy) (b) Deuterated Poly(dioctyloxy) phenylenevinylene (H-DOO-PPV) phenylenevinylene (D-DOO-PPV) (c) C13-rich Poly(dioctyloxy) (d) Poly[2-methoxy-5- phenylenevinylene (C13-DOO-PPV) (2-ethylhexyl-oxy)- 1,4- phenylene-vinylene]] (MEH-PPV) (e) Rubrene (f) Polyfluorene (PFO) Figure 2.1. Chemical structures of (a) H-DOO-PPV, (b) D-DOO-PPV, (c) C-13- DOO-PPV, (d) Rubrene, (e) PFO, (f) MEHPPV, (g) RR P3HT, (h) RRa P3HT, (i) PTB7, (j) C60 , (k) PC61BM, and (l) PC71BM. 23 (g) Regio-Regular -Poly- (h) Regio-Random-Poly- (3-hexylthiophene) RR P3HT (3-hexylthiophene) RRa P3HT (i) Poly-thienothiophene- (j) C60 molecule benzodithiophene 7 (PTB7) (k) [6,6]-Phenyl C61 butyric acid (l) [6,6]-Phenyl C71 butyric methyl ester (PC61BM) acid methyl ester(PC71BM) Figure 2.1. Continued. 24 etching, the ‘nail polish' was cleaned with acetone, and the patterned ITO was ‘diced' into 12.5 x12.5 mm2, as shown in Figure 2.2(a). For fabricating miniature devices, the ITO substrate was cleaned and photoresist was applied by spin casting. The photoresist was then dried by heating the substrate at 120 oC for 2 minutes. After baking, the sustrate was exposed to intense UV and developed by AZ 352 developer for the desired pattern. Finally, the substrate was etched using dilute hydrochloric acid, and the residual photoresist was removed using acetone. One percent micro soap cleaning solution, acetone, methanol, and propanol were consecutively used in ultrasonic hot baths for 15 minutes each to remove occasional organic and inorganic dirt from the substrate. Compressed nitrogen gas was blown to dry the substrate in clean room. Subsequently, oxygen plasma cleaning of the substrate was performed to remove any remaining dirt and organic solvents. Following the above-mentioned cleaning procedures, a thin layer of PEDOT:PSS (70:30) was spin-coated at 5000 rpm for 40 sec. The thickness of this layer was about 50 nm, as indicated by a ‘thickness profilometer'. This layer acts as the hole transporter into the organic layer. The spin coated substrate was then transferred into the glove box. In order to remove water molecules, the substrate was heated at 110 oC for 30 minutes inside the glove box. A solution of luminescent π- conjugated polymer was made by dissolving the appropriate chemical powder with suitable organic solvent. Based on the material used for the organic layer, different solvents were used such as toluene, chloroform, chlorobenzene, and 1, 2-dichlorobenzene. The thin organic layer (80-150 nm) was made by spin casting the solution onto the substrate. For small molecules such as Alq3, C60 etc., 25 Figure 2.2. ITO pattern on the substrate after etching (a). Top view of completed OLED device (b). The typical OLED device structure (c). 26 the powder was thermally evaporated to produce thin films using a slow evaporation rate. To deposit the top electrode, the quoted substrate was put in a thermal evaporator. The evaporator was then pumped down to 2 x 10-6 torr before evaporation. Low work function metal calcium was evaporated at the rate of 2-3 Ao/s on top of the organic layer, which served as an electron transporter into the organic layer. 100 nm of aluminum was then deposited on top of the calcium layer to serve as a ‘capping layer' for protection against oxidation. The film thickness of deposited metals was measured using an Inficon XTM quartz crystal deposition monitor mounted at the same height as the samples in the evaporation chamber. The top view of a typical completed device is shown in Figure 2.2(b). Three kinds of organic diodes were fabricated with the configuration ITO/PEDOT/organic layer/Ca/Al, ITO/PEDOT/ organic layer /Au, and Glass/Al/LiF/organic layer/Ca/Al for bipolar (OLED), for hole unipolar and electron unipolar diodes, respectively. The typical device structure of OLED is shown in Figure 2.2(c). In order to reduce the penetration of oxygen and water to the device, the completed device was encapsulated using microscope cover glass and UV curable glue purchased from Norland, which was exposed to UV light for 30 seconds. 2.3 Organic Light Emitting Diodes Characterization To characterize the performance of the fabricated OLEDs, the following measurements were performed on the completed device. 27 2.3.1 Current-Voltage and Electroluminescence- Voltage Characteristics The completed device was mounted on the sample holder and the electrical connections for the measurements were done. The device was then placed in a closed cycle Helium cryostat. I-V measurement was performed on the device using Keithley 236 apparatus. A silicon photo-detector connected with the oriel preamplifier and Keithley 2400 system was used to measure the electroluminescence from the bipolar devices. Figure 2.3 shows typical current-voltage (I-V) and electroluminescence-voltage (EL-V) characteristics of a MEH-PPV OLED (ITO/PEDOT/MEH-PPV/Ca/Al). The charge transport in the organic layer under electric field is mainly due to hopping, which is limited by shallow and deep traps, recombination, morphology, temperature, etc. When the applied bias voltage is smaller than the ‘built-in voltage', V0, then the current flow in the device is linear with the voltage, which may be due to some leakage current superimposed on the injection current. Upon increasing the bias voltage, injected carriers form a space charge layer near the injecting metal/organic interface due to the low carrier mobility. The current flow is then governed by space charge limited current (SCLC) described by Mott-Gurney law [40] for current density, i.e., 3 2 8 9 L V J   (2.1) where ε is the electric permittivity, μ is the carrier mobility, and L is the organic layer thickness. 28 Figure 2.3. Typical I-V and EL-V characteristics of OLED device based on MEHPPV as an active layer at 10 K. In the bipolar injection regime, needed for electroluminescence emission, the relation of current density is modified as [41], 3 2 2 ( ) 8 9 L V J r e h e h         (2.2) where μr=r (μe+μh) is the recombination mobility, and r<<1 is a constant. 2.3.2. Magnetoconductance and Magneto-electroluminescence Magnetoconductance (MC) and magneto-electroluminescence (MEL) of OLED devices is typically measured by sweeping the magnetic field at a constant bias voltage 0 1 2 3 4 5 6 7 8 0 20 40 60 80 I EL Voltage (V) I (A) 0 1 2 3 EL(Arb.Units) 29 using Keithley 236 apparatus. MC or MEL is defined as the fractional change in the field induced current or electroluminescence, respectively. For performing such measurement, the devices were mounted onto a cryostat placed in between the poles of an electromagnet with the magnetic field perpendicular to the current flow through the device. Magnetic field up to 300 mT was produced using an electromagnet, and measured using the Hall probe Gaussmeter. A temperature controller unit was connected to the cryostat for measuring the MC and MEL temperature dependences. The schematic of the experimental set-up for the measuring organic magnetic field effect (MFE) is shown in Figure 2.4. The change in current at a constant bias voltage, V for different magnetic field, B was measured using Keithley 236 apparatus. Magnetic field-induced fractional change in current or electroluminescence, ΔX/X (dubbed MX) is defined by ( 0) ( ) ( 0)      X B X B X B X X (2.3) which is positive or negative depending on the value of X(B) compared to X(0), where X=I or EL. Figure 2.5 shows typical magnetoconductance, MC, and magneto-electroluminescence, MEL, responses of an OLED device. To characterize the magnetic field dependence of current flow through the device or electroluminescence output, either a Lorentzian, (2.4) 2 0 2 2 ( ) B B MX B MX B    30 Figure 2.4. The experimental set-up for measuring the organic magnetic field effect in films and devices 31 Figure 2.5. Typical MC and MEL responses of an OLED based on MEHPPV as the active organic interlayer, measured at 10 K. where B0 is half width at half maximum and MX∞ is MX at infinite magnetic field, or the non-Lorentzian line shape (2.5) where B0 is half width at quarter maximum were reported for most of the polymers and small molecules used here. The parameter B0 is about 3-10 mT for most of the investigated polymers [19, 22, 25, 33, 42]. It was shown in the literature that this parameter is related with a process that involves a spin flip mechanism caused by the hyperfine interaction. -180 -120 -60 0 60 120 180 0 2 4 6 8 10 12 14 16 MC MEL B (mT) MC (%) 0 4 8 12 16 20 24 28 MEL (%) 2 0 2 ( ) ( ) B B MX B MX B    32 2.4 Organic Photovoltaic Device Fabrication The fabrication procedure of an organic photovoltaic (OPV) cell is roughly the same as fabrication of an OLED device. The only difference is the active material. The active material used in an OPV cell is a suitable blend of organic donor and acceptor. Depending upon the donor/ accepter system, either the spin casted layer was annealed, or a few percentages of additives were added onto a solution of the blend in order to improve the morphology and hence to facilitate the nanoscale phase separation between the polymer donors and fullerene aggregates acceptors. 2.5 Organic Photovoltaic Device Characterization 2.5.1 Current-Voltage (I-V) Characteristics To characterize the power conversion efficiency (PCE) of OPV solar cells, the OPV devices were illuminated under a standard AM 1.5 condition shown in Figure 2.6. This illumination condition was generated in our lab using a xenon lamp having a broad spectral range (300-1000 nm). After passing through the AM 1.5 filter, the light has a spectrum close to a standard AM 1.5 spectrum. Using a NREL-certified Si photovoltaic cell, the xenon lamp output was calibrated to get a light intensity of 100 mW/cm2, appropriate to the sun illumination intensity on the Earth at sea level. The experimental set-up for the current-voltage (I-V) characteristics of OPV devices is shown in Figure 2.7. The typical I-V characteristics of a PTB7/PC70BM with 3 wt% of 1,8-diiodooctane (dio) device measured using the Keithley 236 apparatus is shown in Figure 2.8. Three parameters are used to characterize the PCE of OPV devices. 33 250 500 750 1000 1250 1500 1750 2000 0.0 0.5 1.0 1.5 Spectral irradiance (Wm-2nm-1) Wavelength (nm) Figure 2.6. Standard ‘AM1.5 spectrum', under which the integrated illumination intensity is 100 mW/cm2 34 Xenon lamp Keithley 236 Computer Figure 2.7. The experimental set-up for measuring the I-V response of an OPV cell. -0.25 0.00 0.25 0.50 0.75 -15 -10 -5 0 PTB7/PC70BM 3% dio PCE=6.87% Jsc(mA/cm2 ) Voltage (V) Jsc Voc Figure 2.8. Typical I-V characteristics of a PTB7/PC70BM OPV cell with 3 wt% of 1,8-diiodooctane, under ‘AM 1.5' sun-like illumination. 35 These are: short circuit current density (Jsc), open circuit voltage (Voc), and fill factor (FF), which is defined by relation, (2.6) where Pmax is the largest power output from the device, as shown in Figure 2.6 by the shaded region. The power conversion efficiency (η) of OPV cell is defined as, in sc oc P J V   FF (2.7) where Pin is the optical irradiance of incident light from the sun (100 mW/cm2). 2.5.2. Magneto-photocurrent (MPC) Measurement In order to measure the magnetic field effect on photocurrent (PC) of OPV devices, the fabricated OPV device was transferred into the cryostat placed in between the pole pieces of an electromagnet. The experimental set-up is the same as that for measuring MC in OLEDs, except that the OPV device was illuminated either with a tungsten lamp or with a laser of suitable wavelength (depending upon the absorption spectrum of polymer). By setting the bias voltage to zero (short circuit condition) using the Keithley 236 apparatus, the field-induced fractional change in photocurrent was measured when sweeping the magnetic field. MPC is defined by the relation: sc oc J V P FF max  36 1 (0) ( ) ( )   PC PC B MPC B (2.8) Typical MPC(B) response for an OPV cell based on PTB7/PC70BM with 3 wt% of 1,8-diiodooctane (dio) molecules is shown in Figure 2.9. The MPC response shows the broad nonsaturating response with magnetic field, typical of g spin mixing mechanism (see below). 2.6 Material Characterization 2.6.1 Linear Absorption Measurement The absorption of a medium is quantified by measuring the optical density (OD), which is also called absorbance. In general, the absorption spectrum gives general information about the band gap (material compound) and the electronic excited states of the material of interest. When the -conjugated polymer absorbs light, it promotes an electron from the ground state S0 to the excited state S1 that is dipolar-coupled with the ground state. The transitions from the ground state S0 to the higher singlet states Sn occur depending on the oscillator strength of particular transition, appropriate parity, and spin angular momentum. A Cary-17 spectrophotometer from Olis. Co. was used for the absorption measurement in the spectral range 300-2400 nm, which was carried out at ambient conditions. In order to remove the substrate effect and system response, background transmittance T0 of a glass substrate was measured first as a function of wavelength. The sample was then deposited on the glass substrate and the transmittance through the 37 -2000 -1000 0 1000 2000 -0.02 0.00 0.02 0.04 0.06 0.08 PTB7/PC70BM 3% dio MPC (%) B (Gauss) Figure 2.9. MPC(B) response of a PTB7/PC70BM-based OPV cell with 3 wt% of 1,8- diiodooctane, measured at room temperature. sample, T1, was measured again. The reflection and scattering from the sample was neglected, assuming their negligible values. The absorbance ‘A' was then calculated using the relation A=log (T0/T1). The absorbance is related to the film's thickness‘d' and the absorption coefficient (α) according to the Beer-Lambert law A(λ)=OD=αd. So, the absorption which is measured in the unit of OD is given by the relation, T1=T0 exp (-αd) (2.9) A typical absorbance spectrum of MEHPPV film is shown in Figure 2.10. 38 2.5 3.0 3.5 4.0 0.0 0.5 1.0 Normalized Intensity Energy(eV) Figure 2.10. Typical optical density spectrum of MEHPPV film. 2.6.2 Photoinduced Absorption Measurement Continuous wave (CW) photoinduced absorption (PA) studies the change in absorption caused by long-lived photoexcitation species such as triplet excitons and polarons in the film. The difference in the transmission (ΔT) when the sample is illuminated with both the pump and the probe (TL) and when the sample is illuminated only with the probe (TD), i.e., ΔT=TL-TD gives the photoinduced absorption of the photoexcited species. Assuming the change in transmission is associated with a light-induced change in absorption coefficient (Δα), we have 39 TL=TD e-Δαd (2.10) d D D D e T T T T T        1 (2.11) ln(1 ) D T T d      (2.12) When the difference in the transmission is much smaller than the transmission, i.e., ΔT<<TD, D T T d     (2.13) We can have two types of signals depending upon the sign of Δα. If Δα<0, then it is photoinduced absorption (PA), which is associated with the absorption due to creation of new states; if Δα>0, it is photobleaching (PB), which is caused when the lower of the two energy states involved in the optical transition (usually the ground state) is depleted by another process. The experimental set-up for the PA measurement is shown in Figure 2.11. The sample (thin film) was transferred into the He cryostat, and cooled down to cryogenic temperatures using a close-cycle refrigerator. Two light beams were used for the PA measurement. A cw Ar+ laser was used as a pump to excite the material (i.e., to promote electron from ground state to excited state) and another cw beam from an incandescent 40 Figure 2.11. The experimental set-up for measuring the PA spectrum. halogen tungsten lamp or xenon lamp to cover wavelength range from 550 nm to 4.2 μm was used to probe the PA of long-lived photoexcitations. The transmitted light was spectrally resolved by an Acton 300 monochromator and monitored by Si, Ge, or InSb detectors with corresponding amplifier, long pass filter, and grating set, depending on the wavelength probed. Si 10 D photodiode, Ge, and InSb detectors were used to cover the wavelength 550 nm to 1.05 μm, 800 nm to 1.6μm, and 1 μm to 4.2 μm, respectively. The signal was converted from current to voltage and amplified using a preamplifier. The 41 amplified signal was then fed into a lock-in amplifier SR 830 together with the phase reference of a modulated laser beam which is usually modulated with a frequency that corresponds to the life time of photoexcitations, which was usually set at 300 Hz. The cw photo-modulation (PM) spectrum measured in UV irradiated MEH-PPV film using above gap (2.5 eV) pump excitation is shown in Figure 2.12. The PM spectrum consists of two broad PA bands; one centered at ~0.4 eV, which is assigned to the lower polaron transition (marked ‘P1'); and the other is asymmetric with a peak at ~1.4 eV (marked ‘T+P2'), which is composed of the polaron P2 transition centered at ~1.55 eV, and the remnant of the triplet exciton transition. 2.6.3 Magneto-photoinduced Absorption Measurement The negative fractional change in transmission, also called PA, is given by the relation: PA(E)=(-T/T)=Δαd=NSS σ(E), (2.14) where NSS is the species steady state density, σ(E) is the photoexcitation optical cross- section, and E is the probe beam photon energy. Therefore, in a magnetic field, B, PAX(B) is determined by the density NSS(B); which, in turn, is controlled by thephotoexcitationspecies (polaron pair (PP), triplet exciton (TE), or pair of triplet excitons) decay rate coefficient, κ(B) [NSS=G/ κ] where G is the generation rate, and X stands for species such as PP, TE, and pairs of TEs. The X species has an excited state transition X0X1 (PAX), which is activated by a weak probe beam. For B ≠0, the X0 42 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 -2 0 2 4 6 8 10 12 14 PB T+P2 PA 4 (-T/T) Photon Energy (eV) P1 Figure 2.12. The photoinduced absorption spectrum of an irradiated MEH-PPV film. The PA bands P1 and T+P2 are denoted. level splits according to the relevant spin multiplicity, L (L=3, 4 and 9, respectively for the S=1 TE; PP composed of two S=½ polarons; and a pair of TEs). Consequently, through specific spin-mixing processes, the spin content of each sublevel, its decay rate κ, and thus NSS and consequently, PA all become B-dependent, i.e., MPAX(B)[PAX(B)-PAX(0)]/PAX(0) (2.15) In order to measure the magneto-photoinduced absorption (MPA), an electromagnet producing the magnetic field up to 200 mT was placed across the cryostat. 43 The PA spectrum with and without the magnetic field is measured to obtain the MPA spectrum. To obtain the desired magnetic field response of the PA spectrum in films, the monochromator was fixed at the desired wavelength where either the triplet exciton or the polaron band was assigned, and PA(E) spectrum was measured while sweeping B. Figure 2.13 shows the typical MPA(B) response of an irradiated MEH-PPV film. The MPA response is similar to the MC and MEL response of OLED made from the same active layer (MEH-PPV), which indicates that they share a common origin. 2.6.4. Photoluminscence and Magneto-photoluminscence Measurements When the polymer film is excited by a continuous wave (cw) laser beam with above-gap photon energy, it generates steady state singlet excitons (SE; S0S1). The SE may either recombine radiatively (S1S0), giving photoluminscence (PL) emission, or undergo nonradiative processes. A CW Ar+ laser with minimum energy corresponding to the S0S1 transition was used for measuring the PL spectrum. Since PL originates from singlet exciton radiative recombination, magneto-photoluminscence (MPL) cannot directly originate from SE (S=0) (which is B-independent); but rather is caused indirectly by nonradiative decay channel of singlet excitons collisions with TE or polaron pairs, of which density varies with B. The same experimental set-up used for PA and MPA measurement was used for measuring the MPL response. The PL emitted light was collected using spherical mirrors and was dispersed by using an Acton 300 monochromator for measuring the spectrum. The dispersed light was measured using appropriate solid-state photodetectors such as 44 -80 -40 0 40 80 0 5 10 MPA (%) B (mT) Figure 2.13. Magneto-photoinduced absorption response of irradiated MEHPPV film measured at 50 K. silicon or germanium photodiodes. In order to measure the MPL response, the monochromator was fixed at a PL band, and PL was measured while sweeping the magnetic field. The PL spetrum of MEHPPV film is shown in Figure 2.14. The transitions involving the creation of vibrational quanta in the ground state are assigned as 0-0, 0-1, and 0-2 in the PL spectrum. Figure 2.15 shows the typical MPL response of MEH-PPV film. 2.6.5. X-ray Diffraction Measurement X-ray diffraction (XRD) technique is used to identify the crystalline phases, determine, the lattice constant and microcrystalline grain size. We can estimate the inter-planar spacing ‘d' and hence lattice constant according to Bragg's law of diffraction, 45 1.2 1.6 2.0 2.4 0 1 2 3 0-2 0-1 PL (arb.units) Photon Energy (eV) 0-0 Figure 2.14. Photoluminscence spectrum of Pristine MEHPPV film at 50 K. -150 -100 -50 0 50 100 150 -0.6 -0.4 -0.2 0.0 MPL (%) B (mT) Figure 2.15. Magneto-photoluminscence response of pristine MEHPPV film measured at 50 K. 46 2dsinθ = nλ (2.16) where 2θ is the scattered angle between the incident and scattered X-ray, n is the diffraction order, and λ (=0.154 nm) is the wavelength of the incident beam. The grain size ‘L' of the polymer crystallite may be estimated using the Scherrer's relation L~    cos 0.9 2  (2.17) where Δ2θ is the full width at half maximum (FWHM) of the peak. The peak position and Δ2θ values are analyzed using the X'Pert Plus crystallographic analysis software. The morphology of the semicrystalline polymer films was studied by XRD technique. For the XRD measurements, about 200 nm thick polymer film was made on a glass substrate (2.5mm X 2.5mm area) either with spin-coating or with thermal evaporation. The XRD pattern was then obtained using a Philips powder diffractometer equipped with CuK α source at 45 kV and 40 mA power setting. The grazing incidence method was used to measure the XRD pattern from thin films. Figure 2.16 shows the typical grazing incidence XRD patterns from C60 films using the CuK α X-ray line at λ=0.154 nm as a function of the diffraction angle (2θ). 47 5 10 15 20 25 0 500 1000 1500 Intensity (Counts) 2(Degrees) (111) (220) (311) Figure 2.16. X-ray diffraction patterns of a C60 film, where the background scattering was removed for clarity. The numbers represents various (hkl) Bragg bands. CHAPTER 3 MAGNETIC FIELD EFFECT IN ORGANIC DIODES 3.1 Introduction Magnetoconductance (MC) and magneto-electroluminescence (MEL) in organic light emitting diodes (OLEDs) [19-33] are two aspects of the broader research area of "magnetic field effect" (MFE) in the organics [31], which includes field-induced changes in chemical and bio-chemical reaction yields, magneto-luminescence and magneto-phosphorescence, magneto-photoconductance, etc. Typically, the organic MFE response has been observed at relatively low fields (B <100 mT) and various temperatures, and may be as large as ~20% [32]. It has been generally accepted that the organic MFE originates from the field influence on long-lived radical spins in solutions [31], or polarons in organic solids and devices [27, 28]. For obtaining substantial MFE response, the electron spin relaxation rate should be sufficiently small so that magnetic field- induced spin manipulation may occur [31]. Various models have been proposed for explaining the MFE response in devices where the active layers are π-conjugated organic semiconductors (OSEC). Most of these models are based on the hyperfine interaction (HFI) between the injected spin ½ carriers 49 and the nuclear spins in the OSEC layer [23-28]. The most common model considers the HFI mixing of spin sublevels of bound polaron-pairs (PP), where the level-mixing becomes less effective as B increases [27]. Recently [33], by replacing the protons (H) with deuterons (D) in the π-conjugated polymer interlayer, where the D-polymer has a smaller HFI constant, aHFI, it was unambiguously demonstrated that the HFI indeed plays a crucial role in the MFE of polymer diodes. 3.2 Experimental The devices used in our measurements were 5 mm2 diodes, where the OSEC spacers were deposited on a hole transport layer: poly(3,4-ethylenedioxythiophene) [PEDOT]-poly(styrene sulphonate) [PSS]. For the bipolar devices, we capped the bilayer structure with a transparent anode: indium tin oxide [ITO], and a cathode: calcium (protected by aluminum film). The hole-unipolar device was in the form of ITO/PEDOT-PSS/organic layer/Au; whereas the electron-unipolar device was Al/LiF(~2nm)/organic layer/Ca/Al. Very weak or no EL was detected in these unipolar devices. The organic diodes were transferred to a cryostat with variable temperature that was placed in between the two poles of an electromagnet producing magnetic fields up to ~300 mT with a 0.1 mT resolution. By increasing the distance between the two magnetic poles, we improved the resolution down to 0.01 mT; in all cases, B was determined with a calibrated magnetometer. Device I-V characteristics were measured using a Keithley 236 Source-Measure unit. A silicon photo-detector connected with the oriel preamplifier and Keithley 2400 system was used to measure the electroluminescence from the bipolar devices. The devices were driven at constant bias, V, using a Keithley 236 apparatus; and 50 the current, I, and electroluminescence, EL, were simultaneously measured while sweeping B. Magnetic field-induced fractional change in current or electroluminescence, ΔX/X (dubbed MX), is defined by ( 0) ( ) ( 0)      X B X B X B X X where X=I or EL. 3.3 Experimental Results and Discussion 3.3.1 Magnetoconductance Response in Organic Diodes at Ultra-small Fields While most research activities in the field of MFE in OLED have been focused on an intermediate magnetic field regime (~100 mT) [19-30], less attention has been given to understand the effect of magnetic fields that are comparable to the earth's magnetic field and much smaller than the hyperfine coupling. The existence of ultra-small magnetic field effect (USMFE) opposite to the normal magnetic field effect was first predicted by Brocklehurst in 1976 [43] and USMFE has been observed for variety of reactions in solutions [31, 44, 45]. After the recognition of magnetic sense in animals and concern over the possible health hazard due to electromagnetic fields, it has been investigated in detail. These observations motivated us to study the USMFE in organic devices. 51 Here, we include in our study very small fields (B<1-2 mT) and extend our measurements to a variety of unipolar and bipolar organic devices. We show that the MC(B) response in fact contains a peculiar sign reversal at (B<1-2 mT), similar to that reported earlier in the MEL response of polymer OLED [33]. This ultra-small MFE (or USMFE) component manifests itself as MC sign reversal from positive (negative) to negative (positive) in bipolar (unipolar) devices, forming a dip (peak) at Bm that scales with the half-width at half-maximum, ΔB, of the normal MC(B) response. We found, however, that the USMFE in polymers has different width in electron- and hole-unipolar polymer diodes, indicating different hyperfine interaction constant for the electron-polaron and hole-polaron in these materials. We explain the complete MC(B) response using a model Hamiltonian based on "spin pairs" of loosely bound spin ½ polarons with small exchange, having HFI with several strongly coupled nuclear spins. The spin-pairs are composed of either same charges (unipolar devices) or opposite charges (bipolar devices). In this model, the intermixing between the hyperfine-split spin sublevels increases at very small B due to level-crossing at B=0, thereby causing a MC sign reversal. We have studied MC in organic diodes based on a variety of π-conjugated polymers and small molecule spacers. The polymers include: polyfluorene, two derivatives of poly(phenylene-vinylene) [PPV], namely 2-methoxy-5-(2'-ethylhexyloxy) [MEH-PPV], and three isotope enriched 2-methoxy-5-(2'-dioctyloxy) [DOO-PPV]. The latter include H-DOO-PPV (fully protonated-hydrogen), D-DOO-PPV (deuterated-hydrogen rich), and C13-DOO-PPV (13C-carbon rich). The three isotope rich DOO-PPV polymers have different aHFI since skeletal protons (nuclear spin I=½) are replaced by 52 deuterium (I=1) in D-DOO-PPV (causing smaller aHFI); whereas some of the 12C nuclei (I=0, no HFI) are replaced by 13C nuclei (I=½ having substantial HFI), thus increasing the effect of the HFI. The small molecules that we studied include tetracene, pentacene, rubrene, and several fullerenes (only a subset is shown here). We fabricated organic diodes from all of these materials, and subsequently measured the MC response with high field resolution at various bias voltages and temperatures. By shielding the measuring apparatus from the earth magnetic field (BE0.053 mT in Utah) using mu-metal shield, we verified that the USMFE is not caused by BE. Figure 3.1 shows the MC(B) response of several bipolar diodes for B<50 mT at room temperature and V>VBI, where VBI is the device built-in potential, at which both positive and negative charges are injected into the active layer [7]. For |B|>~2 mT, MC is positive, reaching a saturation level, MCmax, at large B. This is the normal MC(B) response [19-30] that is characterized by HWHM, ΔB ranging from 2.8 mT for D-DOO-PPV, to 6.2 mT for H-DOO-PPV, to 9.1 mT for 13C-DOO-PPV; as summarized in Figure 3.1(c). The isotope-dependent ΔB (where ΔB increases with aHFI) for the three DOO-PPV polymers shows that the HFI plays a crucial role in determining the MC response in polymeric organic diodes, as reported in [33] for EL(B) response. However, a surprising MC(B) response is observed at B<1-2 mT (Figure 3.1(b)): where upon decreasing B, the MC reverses its sign, reaching a minimum, MCmin at B=Bm, followed by an increase toward zero MC at B=0. We have measured a number of devices for each material and found the results to be reproducible. When the USMFE response is summarized by plotting Bm vs. ΔB (Figure 3.1(c)), it is apparent that Bm increases with ΔB (i.e., larger aHFI). 53 -40 -20 0 20 40 0.0 0.4 0.8 D H C13 MC(%) (a) DOO-PPV 2B -3 -2 -1 0 1 2 3 0.0 0.3 0.6 MCmin (b) B (mT) MC/MCmax B (mT) Bm 5 10 0 1 C13 H Bm (mT) B (mT) (c) D Figure 3.1. Magnetoconductance (MC) response vs. field, B in bipolar organic diodes based on: three isotopes of DOO-PPV Panel (a) shows MC(B) for B<50 mT; whereas panel (b) shows the normalized MC(B) measured with high field resolution, for B<3 mT (some MC responses are shifted vertically for clarity); MCmax is the saturation MC value at large B. ΔB is the HWHM for the normal MC(B) response, as defined in (a); whereas MCmin and Bm are for the USMFE response, as defined in (b). Panel (c) summarizes Bm vs. ΔB for the MC(B) responses in (a) and (b); the straight line is guide to the eye. 54 The USMFE response component was obtained in most organic devices based on various polymers and small molecules. The normal and ultra-small MC(B) response of three additional devices are shown in Figure 3.2(a) and 3.2(b), respectively. Figure 3.2(c) summarizes Bm vs. ΔB for the MC(B) responses in (a) and (b). The USMFE component in the MC(B) response depends on both bias voltage and temperature (Figure 3.3 for D-DOOPPV). At 10 K, we found that |MCmin| decreases by a factor of 2 as the bias increases from 3.4 to 4.4 V, whereas Bm does not change much. At V=3.4 V, we found that |MCmin| increases as the temperature increases from 10 to 300 K, whereas Bm is not affected by the temperature. Importantly, the dependence of MCmin with V and T is found to follow the same dependencies as the saturation value, MCmax; so the ratio, MC/MCmax is independent on V and T (Figure 3.3 insets). This indicates that the USMFE is correlated with the normal MC response, and therefore is also determined by the HFI in the polaron-pair species. We thus conclude that any viable model which explains the normal MC(B) response needs to also explain the USMFE response component. The USMFE response is not limited to bipolar devices. In Figure 3.4, we show MC responses of hole-only and electron-only MEH-PPV diodes. The high-field MC in both devices is negative (Figure 3.4(a)) [28], and thus the USMFE appears as ‘negative-to-positive' sign reversal with maximum at Bm~0.8 mT for the electron-only device, and Bm~0.1 mT for the hole-only device (Figure 3.4(b)). Importantly, ΔB is smaller in the hole-only device compared to the electron-only device; this is consistent with smaller aHF for holes than for electrons that was recently measured in a similar polymer [46]. We therefore conclude that Bm increases with ΔB in unipolar devices similar to bipolar 55 -40 -20 0 20 40 0 1 2 3 MC (%) RBRN MEH PFO (a) -2 -1 0 1 2 0.0 0.1 0.2 B (mT) IMCI/MCmax (%) B (mT) (b) 0 5 10 0 1 PFO MEH Bm (mT) B (mT) (c) RBRN Figure 3.2. Magnetoconductance (MC) response vs. field, B in bipolar organic diodes based on MEH-PPV, PFO (MCx3), and rubrene RBRN; (MCx8). Panel (a) shows MC(B) for B<50 mT; whereas panel (b) shows the normalized MC(B) measured with high-field resolution, for B<3 mT (some MC responses are shifted vertically for clarity). Panels (c) summarizes Bm vs. ΔB for the MC(B) responses in (a) and (b). 56 0 4 8 -0.4 -0.2 0.0 0.2 0.4 0 4 8 0 100 200 300 0.0 1.0 2.0 3.0 3.5 4.0 4.5 0.0 1.0 2.0 3.0 (a) 3.4V 3.8V 4.4V 10K MC/MC max (%) (b) B (mT) 3.4V 10K 100K 200K 300K T (K) |MC|min/MCmax (%) |MC|min/MCmax (%) bias (V) Figure 3.3. Normalized MC(B) response of a bipolar diode based on D-DOO-PPV for B<0.5 mT at (a) various bias voltages at T=10 K, and (b) various temperatures at V= 3.4 Volt; MCmax is defined in Figure 3.1. The insets in (a) and (b), respectively summarize MCmin/MCmax at various voltages at 10 K, and various temperatures at 3.4 Volt. 57 . -30 -20 -10 0 10 20 30 -1.2 -0.8 -0.4 0.0 -2 -1 0 1 2 -0.4 -0.2 0.0 0.2 MC (%) Hole only (x40) Electron only (a) B (mT) MEH-PPV (b) MC/MC max x10 x2 Figure 3.4. Normalized MC(B) response for (a) B<30 mT, and (b) B<2 mT of hole-and electron-only unipolar diodes based on MEH-PPV, measured at room temperature and V=3 Volt and 20 Volt, respectively. For clarity, the MC(B) responses are multiplied by a factor in (b). 58 devices. We also observed the isotope dependence for H- DOO-PPV and C13-DOO-PPV h-unipolar devices shown in Figure 3.5. The monotonic, high-field MC component in these unipolar devices is also negative (Figure 3.5(a)) [28], and thus the USMFE response here appears as‘negative-to-positive' sign reversal with a pronounced maximum at Bm. Figure 3.5(b) shows that Bm ~ 0.15mT for the H-DOO-PPV hole-only device, whereas Bm ~ 0.4mT for the C13-DOO-PPV hole-only device. In the traditional view of organic MC, the injected spin ½ carriers form weakly bound polaron spin pairs, SP, in either singlet (SP)S or triplet (SP)T spin configuration. As B increases, the intermixing between the singlet and triplet configurations (S-T intermixing) decreases due to the increased Zeeman contribution, thereby affecting their respective populations; this leads to a monotonous, MCM(B), response [27, 28]. However, if the exchange interaction constant J0, then a new MCLC(B) component emerges at BBLC=J, where a singlet-triplet level-crossing (LC) occurs giving rise to excess spin intermixing between the singlet and triplet SP manifolds. The MCLC(B) component has therefore an opposite sign with respect to MCM(B) response, which results in a strong MC(B) modulation at B=BLC [31]. By explicitly taking into account the HFI between each of the SP constituents and N (1) strongly coupled neighboring nuclei, we explain the newly discovered USMFE response as due to a level-crossing response at B=0. Our model is based on the time evolution of the SP spin sublevels in a magnetic field. For bipolar devices, the SP species is the polaron-pair, whereas for unipolar devices, the SP species is a π-dimer (i.e. biradical, or bipolaron [24, 28]). The SP spin Hamiltonian, H, includes exchange interaction (EX), HFI and Zeeman terms: H = 59 Figure 3.5. Normalized MC(B) response for (a) |B| < 60mT, and (b) |B| < 3mT of hole-only unipolar diodes based on H- and C13-DOO-PPV, measured at room temperature. HZeeman+HHFI+Hex; where        2 1 1 ] ~ [ i Ni HF j i ij j I A S H is the HFI term, A ~ is the hyperfine tensor describing the HFI between polaron (i) with spin Si (=½) and Ni neighboring nuclei, each with spin Ij, having isotropic aHFI constant; HZeeman = g1μBBS1z+ g2μBBS2z is the electronic Zeeman interaction component; gi is the g-factor of each of the polarons in the SP specie (we choose here g1 = g2); μB is the Bohr magneton; Hex=JS1· S2 is the isotropic exchange interaction; and B is along the z-axis. All parameters in the 60 Hamiltonian H are given in units of magnetic field (mT). An example of the spin energy sublevels using the spin Hamiltonian H for N1=N2=1, and I=½ (namely, overall 16 wavefunctions) is shown in Fig. 3.6(a). Note the multiple level-crossings that occur at B=0. Other level-crossings appear at larger B, but those are between mostly triplet sublevels that hardly change the S-T intermixing rate and consequent (SP)S and (SP)T populations. The steady state (SP)S and (SP)T populations are determined by the spin-dependent generation and decay rates. The effective decay rate constant, k, is composed of dissociation rate (that contributes to the device current density [47]) and recombination rate (for bipolar diodes); these two processes eliminate the SP species. The SP spin sublevel populations are also influenced by the S-T intermixing coupling. Any change of the S-T intermixing rate, such as produced by increasing B, may perturb the overall relative steady state spin sublevel populations; and through the SP dissociation mechanism, it may consequently contribute to MC(B) . To obtain sizable MC value, k < aHFI. The USMFE response in this model results from the strong coherent S-T inter-conversion of nearly degenerate levels at B<<aHFI/g μB where aHFI is the isotropic HFI constant. The relevant time evolution of the S-T intermixing that determines the steady state SPS population is obtained in our model via the time dependent density matrix (t). Solving the spin Hamiltonian, H, for the energies En and wavefunctions n, we express the time evolution of the singlet population S(t) as [31, 48]:     M m n mn S mn S S P t M t Tr t P , 1 2 | | cos 4  ( ) [ ( ) ]  , (3.1) 61 where PS mn are the matrix elements of the (SP)S projection operator, mn=(E.1n-Em)/ħ, and M is the number of spin configurations included in the SP species (for I=½ M=2N+2). In the absence of a spin decay mechanism, Equation (3.1) contains many rapidly oscillating terms that do not contribute to the singlet steady state population, and two important terms that do not oscillate in time. These are: <S(t=)>=4m|PS mm|2/M +4mn|PS mn|2/M, where the second summation is restricted to accidental degenerate levels, for which mn(B)=0. The first (diagonal) term contributes to the "normal" monotonous MCM(B) response, whereas the second ("level crossing") term contributes to MCLC(B) response that modulates <S(t=)> primarily at B=0, where the S-T degeneracy is relatively high (see Fig. 3.6(a)). The combination of the monotonous MCM(B) and MCLC(B) components at B~0 explains in principle the USMFE response in organic devices. When the SP spin species decays, S(t) in Equation (3.1) needs to be multiplied by a decay function f(t). Under these conditions, the steady state (SP)S decay yield,     0 k (t) f (t)dt S S  is given by:      M m M n mn S S mn B M P f 1 1 2 ( ) (4 / ) | | ( ) (3.2) where    0 f () k cost f (t)dt . When SPS elimination is controlled by an exponentially decaying function f(t)exp(-kt), we have f()=k2/(k2+2). The triplet yield in this model is given by, T(B) =[1-S(B)] [33]. If the SP singlet and triplet dissociation rates into polarons are equal to each other, then their 62 relative contribution to the device conductivity would not change with B in spite of their field-induced population change, resulting in null MC(B) response. We account for the dissociation rate difference by expressing MC(B) as the weighted average [33]: 1, (0) (0) ( ) ( ) ( )         S TS T S TS T B B MC B   (3.3) where S(B) is given by Equation (3.2) and TS is the triplet-singlet "symmetry breaking" parameter that describes the relative S-T contributions to the device conductance via dissociation into free polarons. Figure 3.6(b) shows the calculated MC(B) response using Equations 3.1-3.3 for an axially symmetric anisotropic HFI with N1=N2=1 (I=½; M=16), where aHFI(electron)=3aHFI (hole)=3 mT, J=0, TS=0.96, and an exponential SP decay /  0.001 HFI k a . The calculated MC(B) response captures both the obtained experimental USMFE response at small B, as well as an approximate B2/(B0 2+B2) shape at larger B, where B0  1.5aHFI  4.5 mT. The excellent agreement between theory and experiment, including both Bm and the USMFE shape and relative amplitude, validates the model used. 3.3.2 Magneto-electroluminescence (MEL) Response in Organic Diodes at Ultra-small Fields Electroluminescence in the OLEDs results from recombination of polaron pairs (PP) in the spin singlet configuration. The electrons and holes that are injected into the 63 -6 -3 0 3 6 0.0 0.5 MC (%) B (mT) a 1 =3mT a 2 =1mT (b) -6 -3 0 3 6 -6 0 6 E/g B (mT) (a) Figure 3.6. Calculated spin energy levels and magnetoconductance. (a) Example of calculated spin energy levels vs. B for a spin pair with isotropic HFI; a1=3, a2=3 mT, and J=0. Note the multiple level-crossing at B=0. (b) Calculated MC(B) response for a SP with axially symmetric HFI averaged over all magnetic field directions. The isotropic HFI is the same as in (a). The anisotropic HFI component is azz=0.15ai for the respective SP constituent. active layer via the metal electrodes can form loosely bound singlet (PPs) and triplet polaron pairs (PPT) depending upon the mutual polarons' spin configuration. Following the PP generation, they may undergo three possible processes. They (i) may combine to form excited state singlet excitons (SE) and triplet excitons (TE), (ii) can dissociate into free charge carriers again, or (iii) can exchange spins via intersystem crossing (ISC). SEs may decay radiatively, resulting in electroluminescence. The long-lived TEs may decay nonradiatively or show delayed fluorescence via the process of triplet-triplet annihilation. The steady state PP density depends on the PPS and PPT "effective rate constant", γ, which is the sum of the formation, dissociation, and recombination rate constants, as 64 well as triplet-singlet (T-S) mixing via intersystem crossing (ISC). If the effective rates, γS for PPS and γT for PPT, are not identical to each other, then any disturbance of the T-S mixing rate, such as by the application of an external magnetic field, B, would perturb the dynamic steady state equilibrium that consequently results in a change of the device electro-luminescence (MEL), as well as the conductance (MC). It has been generally accepted that the organic MEL originates from the field-induced changes in the dynamics of long-lived loosely coupled polaron pairs (PP) in organic solids and devices [27, 32]. In a recent paper [33], it has been experimentally shown that the hyperfine interaction is responsible for the mixing of the spin sublevels of the PP species. This was achieved by replacing protons with deuterons (D) in the π-conjugated polymer chains, where the D-polymer has smaller HFI constant, aHFI. The obtained MEL(B) response was narrower in the D-polymer, in accordance with the reduced HFI constant. In this section, using high magnetic field resolution, we show USMFE response component in MEL(B) response in most organic devices based on various polymers and small molecules. We measured a number of devices for each material and found the results to be reproducible. Figure 3.7 shows the normal and ultra-small MEL response of OLED devices based on MEHPPV polymer as an active layer. Similar response was observed in MC(B) of devices based on the same active layer. We also measured MEL(B) response of the small molecule rubrene, as shown in Figure 3.8. The MEL(B) response in both cases is composed of two regions: (i) a "sign-reversal" region at |B| < 1-2 mT, where MEL(B) reverses its sign reaching a maximum absolute value |MEL|m at B = Bm, and (ii) a monotonic region at |B| >∼2mT, where MEL(B) monotonically increases having an approximate Lorentzian line shape with half 65 -60 -40 -20 0 20 40 60 0 10 20 30 MEL (%) B (mT) (a) -3 -2 -1 0 1 2 3 0 1 2 3 MEL (%) B (mT) (b) Figure 3.7. Magneto-electroluminescence (MEL) response vs. field, B in bipolar organic diodes based on MEHPPV polymer as an active layer. Panel (a) shows MEL(B) for B<60 mT; panel (b) shows the MEL(B) measured with high field resolution, for B<3 mT. 66 -60 -40 -20 0 20 40 60 0 2 4 6 (a) MEL (%) B (mT) -3 -2 -1 0 1 2 3 0.0 0.5 1.0 1.5 2.0 (b) MEL (%) B (mT) Figure 3.8. Magneto-electroluminescence (MEL) response vs. field, B in bipolar organic diodes based on rubrene small molecule as active layer. Panel (a) shows MEL(B) for B<60 mT; panel (b) shows the MEL(B) measured with high field resolution, for B<3 mT. 67 width at half maximum, ΔB. We explained the entire MEL(B) response, including the "normal" monotonic region, as well as the "sign reversal" region using a simple model Hamiltonian based on PP having HFI with several nuclear spins (same as explained detail in Section 3.3.1). In this model, the intermixing between the hyperfine-split spin sublevels increases at very small B due to level-crossing at B = 0, thereby causing a sign reversal. 3.3.3 Illumination Effect on Magnetoconductance Response of MEHPPV Devices MEH-PPV films are somewhat unusual in the class of -conjugated polymers since their photoinduced absorption (PA) spectrum may change according to the environment/mixture used, as previously shown in detail [49]. Films of pristine MEH-PPV that are kept in the dark for a long time show fairly strong PL emission (quantum efficiency of about 25%), and their PA spectrum consists of long-lived triplet excitons; but do not support long-lived photogenerated polarons, probably because of small density of imperfections and impurities in the film. However, if the same films are exposed to prolonged UV illumination, a meta-stable state is formed due to photoinduced native defects in the film, in which long-lived polarons are photogenerated and the photoluminescence (PL) emission is considerably quenched [49]. Here, we make use of this property of MEH-PPV and measured the effect of illumination on MC(B) response in three different types of organic diodes with the configuration of ITO/PEDOT/MEH-PPV/Ca/Al, ITO/PEDOT/MEH-PPV/Au, and ITO/AL/MEH-PPV/Ca/ Al for bipolar, hole unipolar, and electron unipolar diodes, 68 respectively. Figure 3.9 shows the MC(B) response of these devices based on pristine MEH-PPV polymer. The bipolar diode shows positive MC, whereas negative MC was observed for both e-unipolar and h-unipolar diodes [28, 32]. Figure 3.10 shows the MC(B) response of a pristine and UV irradiated MEH-PPV bipolar device. It is clearly seen that upon UV illumination, there is a significant increase in MC of the organic device. Similar effects were observed in X-ray exposed organic diode based on Alq3 [50], and electrically conditioned PPV devices [51, 52]. This enhancement in MC can be explained by defect generation within the organic active layer upon irradiation. Figure 3.11 (a) is the effect of the illumination on the MC(B) response in h-unipolar device. In the dark, the MC response of this device is negative. Upon prolonged illumination, however, we obtained a gradual change in the MC(B) magnitude the h-unipolar MEH-PPV device; the MC first decreases then changes sign from negative to positive. A possible mechanism for this effect is that the light-induced metastable polarons [49] in the illuminated polymer initiate more polaron pairs generation having opposite charge in the device upon current injection, and these PP species are responsible for the obtained positive MC with illumination. Figure 3.11 (b) shows the effect of illumination on the MC response of the e-unipolar device. There is no sign reversal in MC of these unipolar devices upon prolonged illumination. This could be due to the creation of metastable electron defects in the polymer layer upon illumination, so that PP species of different charges are not formed. 69 -60 -40 -20 0 20 40 60 -8 -6 -4 -2 0 2 4 6 e-Unipolar MC (%) Magnetic Field (mT) x100 x10 Bipolar h-Unipolar Figure 3.9. Magneto-conductance MC(B) response in bipolar, hole-only, and electron-only unipolar organic diodes based on MEHPPV. The latter responses are multiplied by a factor of 100 and 10, respectively. 70 -120 -60 0 60 120 0 4 8 12 16 20 24 0 2 4 6 0 4 8 12 16 20 min Pristine MC (%) B (mT) Current (uA) Voltage (V) Figure 3.10. Magnetoconductance MC(B) response in pristine and UV irradiated (20 min) MEHPPV OLED, measured at 10 K. 71 -60 -30 0 30 60 -3 0 3 6 9 12 15 100*MC(%) B (mT) dark W 4W 9.6W (a) -100 -50 0 50 100 -0.8 -0.6 -0.4 -0.2 0.0 Dark 13.7W 552.8 W 1500W MC (%) B (mT) (b) Figure 3.11. MC(B) response of (a) hole-only and (b) electron-only unipolar devices at 10 K, illuminated with 532 nm laser for 30 minutes at different power. 72 3.4 Conclusion In summary, we found a novel USMFE response at B<<aHFI in many bipolar and unipolar organic diodes, which demonstrates that MC(B) and MEL(B) response is much richer than anticipated before. The USMFE component scales with the more regular MC(B) response, and is thus also due to the HFI influence of the SP pairs. Our simple model explicitly includes in the SP Hamiltonian the most strongly interacting nuclear spins, and is capable of reproducing the entire MFE(B) response, including the new USMFE component. Our findings show that, via the USMFE component, relatively small B is capable of substantially altering both electrical and electro-optical response in organic diodes, as well as chemical, and biological reactions discussed elsewhere[31], and thus should be seriously considered. In fact, a chemical USMFE has been proposed to be at the heart of the ‘avian magnetic compass' in migratory birds. In this respect, our work shows that the USMFE appears in MFE response of many more organic compounds that has been thought before. We also found that prolonged illumination of the organic layer dramatically changes the performance of the organic devices. We found enhancement in MC of bipolar device, and sign reversal in h-unipolar device upon illumination. Positive MC observed in irradiated unipolar device supports the polaron-pair mechanism. CHAPTER 4 MAGNETIC FIELD EFFECT IN ORGANIC FILMS 4.1 Magnetic Field Effect on Excited State Spectroscopies of -Conjugated Polymer Films 4.1.1 Introduction The intensive studies of magnetic field effect, such as magnetoconductance (MC) and magneto-electroluminescence (MEL) in organic light emitting diodes [19-33], was boosted in 2004 as the first prototype organic spin valve was demonstrated revealing the existence of relatively long spin coherence length in the organics [7]. Various mechanisms responsible for the MC and MEL in organic diodes have emerged from these studies. Some models emphasized the influence of magnetic field on carrier mobility in the device [24, 27, 53, 54], while other models emphasized the influence of the magnetic field on the carrier density, brought about by spin-dependent microscopic processes among polaron-pairs (PP) or triplet excitons (TE) [22, 33, 55]. A variety of spin-mixing mechanisms have been proposed, including the hyperfine interaction (HFI) between polarons and the skeleton protons in -conjugated polymers [42, 33]; the difference, Δg, in the electron and hole g-factors in polymer/fullerene blends [31]; a number of mechanisms that involve TE [22, 55]; and the spin-orbit coupling in small molecules that 74 in the electron and hole g-factors in polymer/fullerene blends [31]; a number of mechanisms that involve TE [22, 55]; and the spin-orbit coupling in small molecules that contain heavy atoms [56]. Thus, the magnetic field effect in organic diodes has proven to be an especially rich and interesting research field. Here, we report a novel magnetic field effect of spectrally resolved photoinduced absorption (PA) and photoluminescence (PL) [dubbed hereafter MPA and MPL, respectively] in -conjugated polymer films (as opposed to the previously studied organic diodes [57]), and apply it to study a number of spin-dependent processes. This ‘spectroscopic-sensitive' magnetic field effect technique differs from the previously studied ‘transport-related' MC and MEL in devices in two important respects. (i) Since PA and PL measure directly the density of the photoexcitations (such as PP or TE), then MPA and MPL can be directly related to the photoexcitation spin density. Consequently, by directly comparing the MPA and MPL responses in films to those of MC and MEL in organic diodes based on the same organic active layer, we are able to relate the magnetic field effect in organic diodes to the spin densities of the excitations formed in the device. (ii) Being a spectroscopic technique, we can use the MPA as a new tool to discern various long-lived photoexcitations in organic semiconductor films. In addition, we deduce the main spin-dependent species and/or spin-mixing mechanism that determine the MPA (MPL) response in three different forms of a -conjugated polymer, including spin-mixing in PP species, triplet-triplet annihilation, spin-mixing among the triplet spin sublevel, and Δg mechanism of PP in polymer/fullerene blends. We studied MPA and MPL responses in a prototype -conjugated polymer, namely MEH-PPV, which is a derivative of poly(phenylene vinylene). The three different 75 forms that we studied are: pristine film; film exposed to prolonged UV illumination; and electron donor in MEH-PPV/PCBM blend having weight ratio 1:1. The chemical structures of MEHPPV and PCBM are shown in Figure 2.1 [(d), and (k), respectively (Chapter-2)]. A schematic diagram of the philosophy underlying the MPA technique is presented in Figure 4.1. For obtaining PA, the film is excited by a continuous wave (cw) laser beam with above-gap photon energy that generates steady state singlet excitons (SE; S0S1). The SE may either radiatively recombine (S1S0); or convert into long-lived TE via intersystem crossing; or separate into positive and negative charge polarons, some of which may form long-lived PP. These various secondary reactions are symbolized by S1X0, where X stands for species such as PP, TE, and pairs of TEs. The X species has an excited state transition X0X1 (PAX), which is activated by a weak probe beam. PA is defined as the negative fractional change in transmission, T: PA(E)=(-T/T)=NSSβ(E), where NSS is the species steady state density, β (E) is the photoexcitation optical cross-section, and E is the probe beam photon energy. Therefore, in a magnetic field, B, PAX(B) is determined by the density NSS(B); which, in turn is controlled by the X species decay rate coefficient, κ(B) [Nss=G/κ] For B≠0, the X0 level splits according to the relevant spin multiplicity, L (L=3, 4, and 9, respectively, for the S=1 TE; PP composed of two S=½ polarons; and a pair of TEs). Consequently, through specific spin-mixing processes, the spin content of each sublevel, its decay rate κ, and thus NSS and consequently PA all become B-dependent, and consequently MPAX(B)[PAX(B)-PAX(0)]/PAX(0) is formed. In contrast, since it originates from singlet exciton radiative recombination, MPL(B) cannot directly originate from SE (S=0) (which is B-independent); but rather is caused indirectly, for example via SE collision with TE. 76 Figure 4.1. Schematic illustration of the magnetic field dependent pump-probe PA processes. (a) The pump beam with above gap photon energy hυL excites the polymer MEH-PPV to the singlet exciton (SE) level (S0S1). The SE relaxes via intersystem crossing to a triplet exciton (TE) or ionizes into separate charges forming polaron pair, PP (S1X0). The steady state density of the X species is controlled by the spin-dependent decay coefficient, κ. The incandescent probe beam monitors the photoinduced absorption, PA (X0X1, PAX), which is proportional to the X0 steady state density. In a magnetic field B>0, X0 splits according to its spin multiplicity, and the decay rate of each spin sub-level becomes field dependent, resulting in a B-dependent density and PAX (thus forming MPAX). PA(B)PAPAX(B)PAXhLhPROBEhPROBE B>0B=0X1X0S1 S077 The work is arranged as follows. The experimental technique is described in Section 4.1.2. In Section 4.1.3 we describe our experimental results on the three forms of MEH-PPV, including comparative studies of films and devices. In pristine MEH-PPV films, we assign the MPA as due to the TTA mechanism, while the MPL is assigned to TE-polaron scattering. In irradiated MEH-PPV films, we propose that the PP mechanism with hyperfine interaction-mediated spin mixing is responsible for the obtained MPA. The same mechanism combined with a mechanism related to the different g-values of positive and negative polarons (‘g mechanism') play a dominant role in the MEH-PPV/PCBM blend film. In Section 4.1.4 we describe an all-purpose quantum mechanical model which may explain the magnetic field effect obtained in the three MEH-PPV polymer forms. The model is based on the time evolution of the photogenerated species spin-sublevels in a magnetic field in the presence of spin-dependent decay mechanism. This model is viable for both MPA measurements in films as well as MC and MEL in devices made of the same polymers. Using this model, we show that the magnetic field dependent excitation density may account for the measured magnetic effect in the MEH-PPV system, including MPA, MPL, MC, and MEL. 4.1.2 Experimental For the MC and MEL measurements, we fabricated ~5 mm2 diodes, where the organic spacers were deposited on a hole transport layer: poly(3,4-ethylenedioxythiophene) [PEDOT]-poly(styrene sulphonate) [PSS]. We capped the bilayer structure with a transparent anode: indium tin oxide [ITO], and a cathode: calcium (protected by aluminum film). The devices were driven at constant bias, V. For the PL 78 and PA measurements, we used a standard photomodulation set-up described in Section 2.6.2. For excitation, we used a cw Ar+ laser pump beam at ћL=2.54 eV that was modulated at frequency f; and an incandescent tungsten/halogen lamp as the probe. The PA signal, T/T is the fractional change, T in transmission, T, which is negative for PA, and positive for photobleaching (PB). The PA signal was measured using a lock-in amplifier referenced at f, a monochromator, and various combinations of gratings, filters, and solid-state photodetectors spanning the spectral range 0.3<ћ(probe)<2.3 eV. This set-up was also used for measuring the PL spectrum. The device (or film) was placed in a cryostat in between the two poles of an external magnetic field up to 300 mT. For obtaining the desired magnetic field response, the measured quantity, such as PA and PL in films, and EL and current in diodes, was measured while sweeping B. MEH-PPV films are somewhat unusual in the class of -conjugated polymers since their PA spectrum may change according to the environment/mixture used, as previously shown in detail [49]. Films of pristine MEH-PPV that are kept in the dark for a long time show fairly strong PL emission (quantum efficiency of about 25%), and their PA spectrum consists of long-lived triplet excitons, namely PAT (Figure 4.2 (a)); but do not support long-lived photogenerated polarons, probably because of small density of imperfections and impurities in the film. However, if the same films are exposed to prolonged UV illumination, a meta-stable state is formed due to photoinduced native defects in the film, in which the PA spectrum also contains substantial long-lived photogenerated polarons having two characteristic PA bands (PAP) that are formed on the expense of both PL and PAT [49]. The process is reversible when subjected to elevated temperatures in the dark. Furthermore, when the MEH-PPV donor-like polymer is mixed 79 with a fullerene acceptor-like molecule forming bulk heterojunction morphology, then the photogenerated excitons ionize to form positive polarons on the polymer and negative polarons on the fullerene molecule [6]. We took advantage of these MEH-PPV film properties to obtain MPA of various photoexcitation species using the same polymer film; namely, before and after prolonged UV illumination, and in blend with fullerene molecules, namely [6,6]-phenyl-C61-butyric acid methyl ester (PCBM). 4.1.3 Experimental Results 4.1.3.1 Pristine MEH-PPV Films In Figure 4.2 (a), we show the PA spectra of pristine MEH-PPV film at B=0 and 100 mT, respectively. The spectrum consists of a broad PA band centered at ~1.37 eV (marked T) that is assigned to TE transition (PAT) [49]; no other PA bands were obtained down to 0.2 eV, attesting to the good quality of the polymer used here. The B=100 mT spectrum is identical in shape to that of B=0, except that is slightly weaker. The difference, ΔPA spectrum is similar to PAT, demonstrating that it relates to the TE density. As seen in Figure 4.2 (b), the magnetic field response, MPAT(B) ΔPA/PAT, varies strongly with the laser excitation intensity, IL, and thus with NSS (which is proportional to IL). NSS is also inversely proportional to the sublevel TE effective recombination rate constant, κ=κα (α=1,..,L), which are B-dependent. At small IL, MPAT(B) monotonically decreases, but it gradually transforms into a more complex response at large IL where two components are resolved; a low-field MPA component that decreases with B, and a high-field component that increases with B. We thus conclude that MPAT(B) is dominated by two different spin-mixing mechanisms related 80 1.2 1.4 1.6 1.8 0 1 2 3 4 -10 3 T/T B=0 Photon Energy (eV) (a) B=100 mT T PA x10 -150 -100 -50 0 50 100 150 -4 -3 -2 -1 0 x1.4 x2.9 10 mW MPA (%) B (mT) (b) -150 -100 -50 0 50 100 150 -0.6 -0.4 -0.2 0.0 1.5 1.8 2.1 1 2 3 B=0 B=100 mT PL (arb. units) Photon Energy (eV) x2.2 x3.3 x1.4 400 mW 50 mW 200 mW x4.7 TE TTA 10 mW 50 mW 200 mW 400 mW MPL (%) B (mT) (c) -150 -100 -50 0 50 100 150 -1.0 -0.5 0.0 MPA (%) B (mT) (d) -150 -100 -50 0 50 100 150 -0.4 -0.2 0.0 TE MPL (%) B (mT) (e) Figure 4.2. Excited state spectra (PA and PL) and magnetic field effects in pristine MEH-PPV films. (a) The triplet PA band, PAT at B=0 and 100 mT (black and red lines, respectively), respectively, generated using a laser excitation at hυL=2.54 eV @ IL=200 mW/cm2, and their difference spectrum ΔPAT=[PAT(100mT)-PAT(0)] (blue line). The region near the peak is magnified (within a circle). Right inset: PL spectrum at B=0 (black line) and 100 mT (red line), respectively. The lines in the circles show the data on a higher resolution scale. (b) MPAT(B) response measured at 1.37 eV probe, for various laser excitation intensities (normalized). (c) MPL(B) response measured at 2.05 eV probe for various laser excitation intensities (normalized). (d) Model calculations of MPAT(B) response using the TE mechanism (blue line, corresponds to the 10 mW data in (b)) and TTA mechanism (green line, corresponds to the 400 mW data in (b)) mechanisms; see text. (e) Model calculation of MPL(B) response using the model of singlet exciton quenching by TE (SE-TE collision, see text). 81 with TE species; one mechanism that dominates at low IL, which may be a ‘single-TE' process; and the other mechanism that increases at large IL, and therefore most likely involves ‘triplet-triplet annihilation' (TTA) process. The same pristine MEH-PPV film also shows MPL response. Figure 4.2 (a) inset displays the PL spectrum at B=0 and 100 mT, respectively, that consists of several vibronic replicas, with 0-0 transition at 2.05 eV. The difference, ΔPL spectrum follows the PL spectrum, and is thus assigned to the S1S0 transition (Figure 4.1). Unlike MPAT(B), however, Figure 4.2 (c) shows that MPL(B) does not change with IL; it monotonically decreases with B, similar to the low intensity MPAT(B), i.e., the low-field component. Since singlet excitons alone cannot depend on the magnetic field, we therefore assign this MPL(B) response as due to SE nonradiative decay that is activated by ‘collisions' with TE species, of which density NSS(B) also determines the MPAT(B) response at low IL. 4.1.3.2 Irradiated MEH-PPV Films and Devices Entirely different characteristic PA and MPA properties were measured in the same MEH-PPV film after prolonged UV irradiation (~150 minutes using a Xenon lamp at 50 K), which supports photogenerated polaron species [49]. Figure 4.3 (a) shows the PA spectrum of irradiated MEH-PPV film at B=0 and 100 mT, respectively, at similar excitation intensities as used above for the pristine film. The spectrum in this case consists of two broad PA bands; one centered at ~0.4 eV, which is assigned to the lower polaron transition (marked ‘P1'); and the other is asymmetric with a peak at ~1.4 eV (marked ‘T+P2'), which is composed of the polaron P2 transition centered at ~1.55 eV, 82 Figure 4.3. Excited state spectra and magnetic field effects in UV irradiated MEH-PPV film and in organic light emitting diode. (a) PA spectrum at IL=100 mW/cm2 for B=0 (black line) and B=100 mT (red line), respectively, and their difference spectrum, ΔPA=[PA(100mT)-PA(0)] (blue line) in MEH-PPV film. (b) MPA(B) response measured at 1.4 eV probe for various laser excitation intensities (normalized). (c) MEL(B) and MC(B) responses in MEH-PPV diode. (d) Model calculations of MPAPP(B) response in MEH films using the PP mechanism (see text). (e) MPA(B) response at 1.1 eV probe up to B=1.5 mT (filled squares) and B=60 mT (blue line, inset). -40040010200.40.81.21.62.004812P1T+P2B=0 mTB=100 mTx4PAP2-104 T/TPhoton Energy (eV)(a)-80-40040800510x1.4 40 mW 100mW 300 mWMPA (%)(b)x1.1-80-4004080024 MEL MCx3.8MC,MEL (%)(c)-80-400400246MPA (%)(d)-1010.00.20.4 MPA (%)B (mT)(e)83 and the remnant of the TE transition, PAT [49]. The spectrally resolved difference ΔPA (Figure 4.3 (a)) shows that MPA in this MEH-PPV form is correlated only with the two polaron PA bands, P1 and P2, but not with that of PAT. This is one of the MPA technique advantages; its ability to spectrally resolve the dominant species and spin-dependent process. We assign ΔPA spectrum here to magnetic field dependence of the PP's density, namely ΔPAPP. Unlike the negative ΔPAT of the pristine sample (Figure 4.2 (a)), we found ΔPAPP>0 in the irradiated sample, which suggests that a different spin-mixing mechanism is dominant in the present case. The positive, monotonically increasing MPAPP(B) (Figure 4.3 (b)) is naturally explained by the PP mechanism, in which the spin-mixing is governed by the HFI [33] (see below). For comparison, we also show MC(B) and MEL(B) (Figure 4.3 (c)) obtained in MEH-PPV diodes. The MC and MEL responses are identical to each other; and, in addition, are very similar to the MPAPP(B) response shown in Figure 4.3 (b)). This indicates that all three magnetic field effects share a common origin. Since MPA(B) does not involve carrier transport, we conclude that MC(B) and MEL(B) obtained in the devices need not involve transport. All three responses can be explained equally well by the microscopic PP model presented below, that involves magnetic field dependence of the species' spin sublevel character and their density, rather than transport related mechanism through the organic interlayer in the device. A salient feature of the low field (B<1.2 mT) MPAPP(B) response is shown in Figure 4.3 (e). Interestingly, this response (dubbed here ultra-small MPA, or USMPA) was measured at 1.1 eV probe photon energy, where the PA spectrum actually shows photo-bleaching (PB, Figure 4.3 (a)). The 1.1 eV MPA is shown on a larger B-scale in 84 Figure 4.3 (e) inset; it has, in fact the same response as MPA at 1.4 eV. The USMPA response decreases at B<0.6 mT before increasing again to form the monotonic response seen at larger fields. Similar nonmonotonic response was previously observed in both MC(B) and MEL(B) in organic diodes [42, 58], and was explained as due to level-crossing at B=0 that involves spin sublevels formed by the polaron-proton HFI in the polymer chains. Thus, the same explanation is viable also for the USMPA component here. We note that the USMPA is not related to transport in an organic device; in addition, it occurs at field values close to the earth magnetic field (0.05 mT). We thus infer that the USMPA in polymers (and other organic molecules [42]) could, in principle, be used by a variety of living creatures on earth that may take advantage of the earth magnetic field to augment their activity; such as navigation for example, as shown previously [59]. 4.1.3.3 Films and Devices of MEH-PPV/PCBM Blends Yet, a third type of MPA response is viable in films of MEH-PPV/PCBM blend. Upon laser excitation of the polymer (PCBM does not absorb in the visible spectral range), the singlet excitons quickly dissociate into hole-polarons on the MEH-PPV chains and electron-polarons on the PCBM molecules [6]. This weakens the PL intensity of the MEH-PPV chains, and completely eliminates the triplet PAT band from the PA spectrum [60]. Thus, the PA spectrum in this case (Figure 4.4 (a)) consists of PA of positive polarons on the MEH-PPV chains (P1 at ~0.4 eV, and P2 at ~1.37 eV, respectively), as well as PA band of negative polarons on the PCBM (C61- at ~1.2 eV).