Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Photoluminescence from MEH-PPV and DP-PPV in bulk and encapsulated in porous alumina Qi, Dong Feng 2001

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2001-0500.pdf [ 4.08MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0085715.json
JSON-LD: 1.0085715+ld.json
RDF/XML (Pretty): 1.0085715.xml
RDF/JSON: 1.0085715+rdf.json
Turtle: 1.0085715+rdf-turtle.txt
N-Triples: 1.0085715+rdf-ntriples.txt
Original Record: 1.0085715 +original-record.json
Full Text
1.0085715.txt
Citation
1.0085715.ris

Full Text

P H O T O L U M I N E S C E N C E F R O M M E H - P P V A N D D P - P P V IN B U L K A N D E N C A P S U L A T E D IN P O R O U S A L U M I N A by Dong Feng Q i B.Sc. Zhongshan University, Guangzhou, China, 1990 M.Sc. Jinan University, Guangzhou, China, 1993 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E REQUIREMENTS FOR T H E DEGREE OF M A S T E R OF APPLIED  SCIENCE  in T H E F A C U L T Y O F G R A D U A T E STUDIES D E P A R T M E N T O F PHYSICS A N D A S T R O N O M Y  We accept this thesis as conforming to the required standard  T H E U N I V E R S I T Y O F BRITISH C O L U M B I A September, 2001 © D o n g Feng Qi, 2001  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying o f this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  Pty'C*  & Afr»**>ntj  The University of British Columbia Vancouver, Canada  Abstract T w o P P V derivatives, M E H - P P V and D P - P P V , are encapsulated in the nanopores of anodized alumina with different pore sizes (50/200 nm) and depth (10/60 urn). Time-integrated and  time resolved optical emission properties of these structures  have been studied and  compared with those of thin films of the polymers. Significant blue shifts of the P L spectra (0.034 eV-0.183 eV) are found in the porous materials as compared to the bulk. The relative intensity of the one-phonon assisted transition increases in comparison with the intensity of the no-phonon transition in the nanostructured samples. The time constants of the optical emission from the bulk M E H - P P V film are very similar in their energy dependence to results obtained previously on bulk P P V films when considered with respect to the excess energy above the peak of the emission spectrum.  The  decay times are, however, two to three times shorter in the parent. The energy dependence of the time constants measured in D P - P P V films and D P - P P V encapsulated in nanopores (pore diameter 200 nm, pore depth 60 um) are also very similar to each other, and to the results obtained on the M E H - P P V .  These results suggest that the intrachain diffusion of electron-hole  excitations in all of these polymers, either in bulk or encapsulated form, is very similar.  Ill  Table of Contents ABSTRACT...,  ii  T A B L E OF C O N T E N T S  iii  LIST O F FIGURES  v  LIST O F T A B L E S ,  viii  A C K N O W L E D G E M E N T S  CHAPTER 1  ix  INTRODUCTION  I  CHAPTER 2 THEORY  5  2 1 SUM F R E Q U E N C Y G E N E R A T I O N A N DU P C O N V E R S I O N  5  2.2  OPTICAL PROPERTIES  OF  CONJUGATED  POLYMERS  11  C H A P T E R 3 E X P E R I M E N T A N DS A M P L E P R E P A R A T I O N  14  3.1  14  OPTICAL S Y S T E M OVERVIEW  3.2 M A J O R C H A L L E N G E S A N D S O L U T I O N S . . .  19  3.2.1  GENERAL ALIGNMENT  19  3.2.2  G A T E A N D P L SPOT W A L K - O F F O N UPCONVERSION C R Y S T A L  19  3.2.3  S H A P E OF E X C I T A T I O N SPOT  20  3.2.4  NONLINEAR CRYSTAL FOR UPCONVERSION  20  3.2.5  GATE A N D P L BEAM GEOMETRY  21  3.2.6  P M T DARK CURRENT  23  3.2.7  L N / C C D AS A DETECTOR FORE X T R E M E L Y W E A K SIGNALS  23  3.3  UPCONVERSION D A T A ACQUISITION  25  3.3.1  T H E Z E R O D E L A Y POINT: L A S E R U P C O N V E R S I O N  25  3.3.2  A L I G N M E N T OF C O L L E C T I O N OPTICS  27  3.3.3  M A X I M I Z I N G T H E INTENSITY OF P L  28  3.3.4  T H ES U M FREQUENCY SIGNAL  28  3.3.5  DATA COLLECTION  29  3.3.6  WHITE LIGHT UPCONVERSION  30  3.4  SAMPLE PREPARATION  32  3.4.1  B U L K S A M P L E S : F I L M O N MICROSCOPE-SLIDES  32  3.4.2  HOST MATERIAL PREPARATION  34  3.4.3  E N C A P S U L A T I N G M E H - P P V INTO A N O D I Z E D A L U M I N A O R A N O D I S C  39  3.4.4  E N C A P S U L A T I N G D P - P P V INTO A N O D I Z E D A L U M I N A O R A N O D I S C  39  IV  CHAPTER 4 EXPERIMENTAL RESULTS  40  4.1  41  TIME-INTEGRATED  SPECTRA  4 . 1 . 1 . E F F E C T S OF N A N O P O R O U S HOSTS  41  4.1.2. TEMPERATURE DEPENDENCE  47  4.1.3 E X C I T A T I O N INTENSITY D E P E N D E N C E  48  4.1.4 C O M P A R I S O N OF P L INTENSITY F R O M DIFFERENT SAMPLES  49  4.1.5 P H O T O D E G R A D A T I O N  50  4.2 T I M E - R E S O L V E D S P E C T R A 4.2.1 P H A S E - M A T C H I N G C A L I B R A T I O N  54 54  4.2.2 B U L K M E H - P P V  57  4.2.3 B U L K D P - P P V  64  4.2.4 A N O D I S C D P - P P V  68  4.2.5 B U L K V S . A N O D I S C D P - P P V : L I F E T I M E D E P E N D E N C E O N E N E R G Y  72  4.2.6 B U L K M E H - P P V vs. B U L K P P V : L I F E T I M E D E P E N D E N C E O N E N E R G Y  73  4.2.7 L I F E T I M E D E P E N D E N C E O N E N E R G Y FOR P P V A N D DERIVATIVES  74  CHAPTER 5 S U M M A R Y A N DCONCLUSIONS  75  5.1 S U M M A R Y  75  5.1.1 E X P E R I M E N T A L A P P A R A T U S  75  5.1.2 E F F E C T S OF E N C A P S U L A T I O N O N THE T I M E - I N T E G R A T E D L U M I N E S C E N C E S P E C T R U M OF M E H - P P V A N D D P - P P V A T 7 7 K  76  5.1.3 E F F E C T S OF E N C A P S U L A T I O N O N T H E T I M E - R E S O L V E D L U M I N E S C E N C E S P E C T R U M OF M E H - P P V A N D D P - P P V A T 7 7 K  77  5.2 C O N C L U S I O N S  78  REFERENCE  80  APPENDIX A  82  APPENDIX B  84  APPENDIX C  86  V  List of Figures 1.1: Chemical structure of P P V , M E H - P P V and D P - P P V  2  2.1: Schematic optical setup of sum frequency experiment using a nonlinear crystal  5  2.2: Optics of an uniaxial crystal  7  2.3: Negative uniaxial crystal  8  2.4: Geometry of the gate and P L beams for determination of phase matching  10  2.5: Photoluminescence processes of organic molecules  11  2.6: Schematic diagram showing the segments of polymer chain  13  3.1: Schematic set-up for femto-second time-resolved photoluminescence spectroscopy  18  3.2: Polar plot of the phase matching angle as a function of the P L wavelength  21  3.3: The angular difference between the sum frequency propagation and the gate propagation as a function of P L wavelength under phase matched conditions  22  3.4: Schematic Setup for White Light Upconversion  31  3.5: Synthesis of M E H - P P V  33  3.6: Synthesis of D P - P P V  33  3.7: Set-up for Anodization of Aluminum  34  3.8: Schematic graph of sample before and after anodizing  35  3.9: S E M pictures of anodized alumina on aluminum foil  36  3.10: Pores of Anodisc  37  3.11: Chemical reaction route of silanization of Anodisc  38  VI  4.1: Photoluminescence Spectra of bulk M E H - P P V and bulk D P - P P V at 77 K  41  4.2: Photoluminescence Spectra of Anodisc M E H - P P V and Anodisc D P - P P V at 77 K  42  4.3: Two-Gaussian fits for bulk M E H - P P V and bulk D P - P P V P L at 77 K and three-Gaussian fits for bulk M E H - P P V and bulk D P - P P V P L at 77 K  44  4.4: Two-Gaussian fits for Anodisc M E H - P P V and Anodisc D P - P P V P L at 77 K and three-Gaussian fits for Anodisc M E H - P P V and Anodisc D P - P P V P L at 77 K 4.5: Normalized 77 K P L spectra of M E H - P P V and D P - P P V  45 46  4.6: Effects of Temperature on the time-integrated photoluminescence of bulk D P - P P V , Anodized D P - P P V and bulk M E H - P P V 4.7: Intensity o f bulk D P - P P V P L at 77 K versus excitation intensity  47 48  4.8: 77 K Photoluminescence spectra of M E H - P P V and D P - P P V in bulk and nanostructured forms  49  4.9: Evolution of Spectra of Anodisc M E H - P P V and Anodisc D P - P P V with excitation time  52  4.10: Photo-Degradation of Anodisc M E H - P P V and Anodisc D P - P P V  53  4.11: White Light Upconversion spectra at different crystal orientation angles  55  4.12: Wavelength at which the mixing efficiency is maximum for a given angular position of the nonlinear crystal  56  4.13: Short time regime of time-resolved bulk M E H - P P V photoluminescence at 77 K  58  4.14: Bi-exponential fits o f time-resolved bulk M E H - P P V photoluminescence at 77 K  59  4.15: Time constants extracted from the time-resolved decay profiles using the Bi-exponential fits  60  4.16: Population lifetime versus estimated electronic energy  61  4.17: Population lifetime versus E C L energy for bulk P P V  63  4.19: Extracted time constants versus probe energy for bulk D P - P P V P L at 77 K  65  4.18: Time-resolved profiles for bulk D P - P P V photoluminescence at 77 K  66  4.20: Population Lifetime versus estimated E C L energy  67  4.22: Extracted time constants versus probe energy for Anodisc D P - P P V P L at 77 K  69  vii  4.21: Decay profiles for Anodisc D P - P P V photoluminescence at 77 K  70  4.23: Population Lifetime versus estimated E C L energy for Anodisc D P - P P V  71  4.24: Comparison of population lifetimes of bulk and Anodisc D P - P P V  72  4.25: Comparison of population lifetimes of bulk P P V and bulk M E H - P P V  73  4.26: Comparison of population lifetimes of P P V and its derivatives  74  /  viii  List of Tables Table 1.1 Full names of the polymers studied  2  Table 3.1  Specifications of optical components in Figure 3.1  17  Table 4.1  Sample Properties and Sample Labels  40  Table 4.2 Comparison of Spectra of Bulk and Nano structured M E H - P P V and D P - P P V  43  ix  Acknowledgements I would like to thank my supervisor D r . Jeff F . Young for his continuous guidance and encouragement in the past two years. I am so fortunate to have worked with him, through whom I could expose myself to the academic and industrial atmosphere of Canada and North America. I am also in debt to D r . Katja Rademacher, M r . Andras Pattantyus and M s . Keri Kwong in D r . Michael O . W o l f s group in Chemistry Department of U B C for preparing various conjugated polymer samples used in this thesis. I would also like to thank all my colleagues in Photonics Nanostructures Lab, especially M r . Alex Busch, D r . Xiaonong Shen and M s . Lilian Fan, for their constructive discussion and resourceful help. In addition, I would like to sincerely thank my dear wife Qiaomin, for her full support through out my graduate study in U B C , my mother Shaowen, who flow from China to Vancouver, helping me take care of my new born son, so that I could concentrate on the research work.  dedicated to Qiaomin, Hao, Ming-Hau and Wing-Hau  1  Chapter 1 Introduction  Chapter 1 Introduction Conjugated polymers have been the subject of intensive research in the last twenty years, mainly due to their semiconductor-like optical and electronic properties. These organic materials offer promise for use in opto-electronic devices such as organic light emitting diodes (OLEDs)  [1],  light  emitting  electrochemical  cells  [2],  photodiodes  [3],  sensors  [18],  waveguides [5] and lasers [6], etc. As compared to the traditional inorganic semiconductors, such as G a A s , conjugated polymer based devices are easier to fabricate. Currently, full color O L E D based products are commercially available, such as cell phone displays [7]. However, the lifetime of such conjugated-polymer-based products, e.g. display,  is  much shorter as compared to the L C D / C R T  manufacturing are inconsistent.  the  OLED  based display and yields  for  For example, L C D s tend to operate for 10,000 to 15,000  hours, but O L E D s last only 5,000 hours [8].  Research is needed to find new materials and  structures to improve the device performance, to increase the device lifetime and make the device more robust. T o do that, conjugated polymers have been mixed with nanoparticles such as C  6 0  [9,32], carbon nanotubes  [10],  CdS [11]  and T i 0  2  [12],  etc.  to form polymer -  nanoparticle-composites. Conjugated polymers are also encapsulated in nanoporous materials, such as porous silica [13,14,23,31] or porous GaP [21], to modify their optical properties In this thesis, the optical properties of P P V derivatives, M E H - P P V and D P - P P V are compared in bulk form and when encapsulated in the nanopores of anodized alumina.  The  samples were synthesized by members of D r . M . O . W o l f s research group in the Chemistry  Chapter 1 Introduction  Department of U B C . PPV.  Figure 1.1 shows the chemical structures of P P V , M E H - P P V and D P -  The full names for them are given in Table 1.1. The polymers are prepared in two  forms, bulk form, by spin-coating from the polymer solutions onto microscopic glass and nanostructured form, by encapsulating the polymer into porous alumina.  Table 1.1 Full names of the polymers studied  Abbreviation  Full Name  PPV  po/y(p-phenylenevinylence)  MEH-PPV DP-PPV  poly( 1 -methoxy-4-(2-ethylhexyloxy-2,  5-phenylenevinylene)  /?o/y(2,3-diphenyl-/?-phenylenyvinylene)  Figure 1.1: Chemical structures of (a) PPV, (b) MEH-PPV, and (c) DP-PPV.  Chapter 1 Introduction  3  This thesis reports a study of the time-integrated and time-resolved optical emission properties of these samples, predominantly obtained at 77 K . Owing to the relatively weak emission from these, as compared to bulk P P V samples, modifications had to be made to the time-resolved optical detection system originally developed by McCutcheon [18] for studying bulk P P V . Chapter 2 describes some of the basic optical properties of the materials used in the nonlinear optical sum-frequency system and the polymers themselves. Chapter 3 describes how the time-resolved photoluminescence system was modified in order to increase its sensitivity by more than a factor of three times. Chapter 4 presents the time-integrated and time-resolved results obtained in bulk and nanostructured M E H - P P V and D P - P P V , and compares them with those previously obtained from bulk P P V . The bulk spectra of the three types of polymer are all similarly shaped, but are centered  at different wavelengths  in the visible part of the spectrum. The energy  dependence of the decay time constants obtained from the time-resolved experiments are remarkably similar for all samples when the energy scales are shifted so the peaks of the emission spectra overlap.  These time constants decrease almost exponentially with excess  energy above the peak in the density of states associated with the inhomogeneously broadened distribution of coherent segment lengdis in the polymer chains. While these dynamics are similar in bulk and nanostructured forms of the polymers, the time-integrated spectra are changed significantly when the polymers are encapsulated in the nanoporous material. Chapter 4 summarizes the blue shifts,  broadening, and shape changes  integrated spectra from the nanostructured samples.  that occur in these time-  Chapter 1 Introduction  4  In Chapter 5, the conclusions drawn from the experimental results are summarized and some future work is suggested. Together with the previous work on bulk P P V , the present results  motivate  further, more  detailed  studies of  PPV-based polymers  encapsulated  in  nanoporous alumina. The aim of these studies should be to understand better the interaction of the polymers with the pore walls, and how they influence their optical properties. Appendix A and B present the Maple code to numerically calculate the phase matching conditions for the sum frequency  generation.  anodized alumina.  Appendix C presents the chemical process  to  silanizing  the  Chapter 2 Theory  5  Chapter 2 Theory  2.1 Sum Frequency Generation and Upconversion The sum frequency generation technique based on non-linear optical crystals was first used by Mahr and Hirsch [16] in 1975. J. Shah has given a detailed review in 1988 on this  Delay i  Ultrashort Laser Pulse (100 fs) (gate)  P L of sample excited by Ultra-short Pump Pulse  Figure 2.1: Schematic optical setup of sum frequency experiment using a nonlinear crystal.  technique [17], which is shown schematically in Figure 2.1. The photoluminescence (PL) excited by an ultra-short pump pulse* is mixed with an ultra-short gate pulse in a properly oriented nonlinear crystal, such as beta-barium borate,  (3-BaB 04 (BBO), to generate sum 2  frequency  radiation. There are two conditions to be satisfied for efficient sum frequency generation.  * The ultra-short pump pulse originates from the ultra-short laser pulse by frequency doubling.  Chapter 2 Theory  6  The mixing process only takes place when both the P L and gate pulses are present and interact in the crystal at exactly the same time. This can be achieved by varying and monitoring the light-path lengths o f the ultra-short laser pulse (gate) and the ultra-short pump pulse (excitation/PL) beams, as described in Chapter 3. This is the condition that allows the technique to be used to achieve ultrafast time resolution. However, even i f the P L and gate beams temporally and spatially overlap in the crystal, a sum frequency signal will only be generated i f the two are "phase matched". The phase matching condition is given by equations (2.1) and (2.2). COg+0)pl=COs  (2.1)  K + Ki = K  (2.2)  where co , co i and co are the frequencies of the gate, P L and sum frequency beams respectively, g  and k ,k g  pl  p  s  and k  s  are the corresponding wave vectors of the gate, P L and sum frequency  beams in the nonlinear crystal. T o simplify the analysis, we consider a collinear phase matching, which requires,  k +k =k g  pl  (2.3)  s  In terms of the refractive indices «(co), (2.3) can be written as [15]  ^ [n(o)J-n(co )] 8  g  + o) [n(coJ-n(a) )] pl  pl  =0  (2.4)  In transparent isotropic materials with normal dispersion, the index of refraction n{co) increases with co, so the above relation can never be satisfied. B y using a negative uniaxial crystal, such as B B O , and choosing the co beam to be polarized in the extraordinary direction, it s  is possible to arrange for n(co ) - n(co ) and n(co ) - n(co ) to have opposite signs. Two types s  g  s  pt  7  Chapter 2 Theory  of phase matching are commonly used. In type I, both n(a> ) and n(co ) are ordinary. In type g  II, either  pl  n(co ) or n{co ) is ordinary [15]. g  pJ  T o further understand the mechanism o f sum frequency generation using negative uniaxial crystals, such as B B O , one should know the optics of the uniaxial crystals, which was  Principal plane  Beam propagation wavevector  9. Figure 2.2: Optics of uniaxial crystals. described in Reference [33]. A s shown in Figure 2.2, in uniaxial crystals, there exists a specific direction called the optic axis (the Z axis). The plane containing the Z axis and the wave vector of the light wave is termed the principal plane. The light beam whose polarization is normal to the principal plane is called an ordinary beam or an o-beam. The beam polarized in the principal plan is called an extraordinary beam or e-beam. The refractive index o f the o-beam, which is denoted by n , does not depend on the propagation direction, while the refractive index of the e0  beam depends on the propagation direction. The difference between the refractive indices of the ordinary and extraordinary beams is known as the birefringence An. The value o f An is zero along the optic axis Z and reaches a maximum, An= n - n , in the direction normal to the Z axis. 0  e  Chapter 2 Theory  8  In an uniaxial crystal, the index of refraction of the e-beam, n .b m(Q) depends on the angle e  ea  between the propagation beam and the Z axis, denoted as 0, as shown in Figure 2.2-2.3, and is determined by the ellipse defined by n and n , the principal values of the refractive indices [33] . a  e  In this work, non-collinear type I phase matching was used, with both the gate and P L beams polarized as o-beams. Upconversion was accomplished in a negative (n >n ) uniaxial a  e  B B O crystal cut at an angle (9 ut=38 with respect to the optic axis (the Z axis) as shown in 0  C  Figure 2.3.  Figure 2.3: Negative uniaxial crystal: Principal values of refractive indices. Index ellipse. Cut angle with respect to the optic axis.  9  Chapter 2 Theory  As shown in Figure 2.4, the gate and P L beams are not collinear. T o achieve phase matching for such a geometry, the nonlinear crystal must be oriented to the right phase matching angle, which is given by [17] [18]  sm(0J  - »:r-«r  -  (2 5)  (  where 0  is the phase matching angle (the internal angle between sum frequency wave vector  m  and the optical Z axis of the nonlinear crystal), n  H e  and n" refer to the principal values of 0  refractive index at the sum frequency co , respectively, and n (0 ) s  n  s  s  is determined by  m  =k cla>,  (2.6)  s  where co is determined by (2.1), k is given by equation (2.2), and k s  s  g  and k  pl  are magnitudes  of the gate and the P L wave vectors inside the nonlinear crystal, determined by (for Type I phase matching)  k =n° co lc  (2.7)  k =n (D lc  (2.8)  g  s  g  0  pl  where  pl  pl  n° and n ° are the ordinary indices of the gate and P L inside the nonlinear crystal. The pl  indices of refraction, n ° and n °, g  p  are wavelength dependent but are simply determined by  using the empirical dispersion relations given in Reference [33] for B B O . Referring to Figure 2.4, the external angle between the gate and P L beams ( 9 + 6 ) can be easily obtained by u  2i  using the method described in Chapter 3. The internal angle between the gate and P L beams ( B + 9 ) is determined by refraction at the interface according to Snell's law. Based on the u  2t  equations (2.1), (2.2) and (2.5)-(2.8), McCutcheon [18] developed a useful Maple code to find  \Q  Chapter 2 Theory numerically the incident angle of the P L beam, 9  l b  required to phase match a specific  frequency component of the P L . This code was used throughout this thesis work and it was modified to extract the angular difference between the sum frequency propagation and the gate propagation under phase matching conditions. The Maple codes are attached in Appendix A and B . A more detailed description of the code can be found in the Appendix of Reference [18].  Incident angle of PL corresponding to the phasematching condition given by the code (Appendix A)  Entrance surface of Nonlinear Crystal  . I J I  Exit surface of Nonlinear Crystal  Figure 2.4: Geometry of the gate and PL beams for determination of phase matching.  Chapter 2 Theory  \\  2.2 Optical Properties of Conjugated Polymers Conjugated polymers exhibit many semiconductor-like properties. However, the optical  Energy  Sj  Anti-bonding  State (TI* orbital)  Photon Emission  Photo Excitation  So Bonding (71 orbital)  hv  State  Atomic coordinates  Figure 2.5: Photoluminescence processes between the S and Si electronic states of organic molecule, 0  properties are more conveniently described by a molecular picture, which was outlined in Reference [26] by Greenham and Friend. First consider a single isolated benzene ring ( C H ) , which ultimately forms part of the 6  6  backbone of the polymer chain. The electronic structure can be conveniently described in terms of o bonds formed by overlap of sp hybrid orbitals and n bonding formed by overlap 2  of p orbitals on adjacent carbon atoms. The cr orbitals are relatively deep bonding orbitals. The z  semiconductor-like properties of conjugated polymer arise from the derealization of the relatively shallow re orbital electrons. The estates in the benzene molecule act very much like  Chapter 2 Theory  •  12  the valence (bonding TC orbital) and conduction (anti-bonding TC orbital) states in inorganic semiconductors, such as GaAs. The promotion of electrons from occupied TC to unoccupied TC states typically results in a change in the quasi equilibrium geometry of the molecule. The transitions associated with photoluminescence from an organic molecule can be represented as in Figure 2.5. The energies of the ground (estate) and excited (TC state) states are shown as a function of the configuration coordinates  of the molecule system,  which represents  the  positions of the atomic nuclei. The parabolic shape of the energy bands indicates an energy increase due to displacements from the equilibrium configuration for a specific electronic state. Each electronic state is associated with a set of vibrational sublevels that are spaced much closer than the electronic energy levels. When the molecule is excited (absorption of high energy photons), electrons are promoted from the bottom of the S  0  state to high lying Sj  sublevels, from which they non-radiatively decay very quickly (< 10" s) down to the bottom 13  of Sj. Radiative emission from S,° can result from a transition to any one of the vibrational sublevels in S . The transition from the bottom of S, to the bottom of S is a purely electronic 0  0  transition, and is denoted as Sj-^S /0—>0 or SQ. . The transition from the bottom of S to the 0  0  ;  other vibrational sublevels in S are denoted as 5 ->5' /0->i, S ^S /0->2, 0  i  0  1  0  etc.  Polymer chains are formed by bonding identical monomer molecules in a linear chain. When defect free, the 7r electrons can be delocalized over several monomers. The number of monomer units over which the orbitals are delocalized is determined by defects on the polymer chain, such as twists, bends or impurities, as shown schematically in Figure 2.6. The effective conjugation length ( E C L ) will determine the energy spacing of the S and Sj levels, and thus 0  the emission energies from the segments [30]. The transition energy is smaller for longer  13  Chapter 2 Theory effective  conjugation lengths  and is larger for short effective  conjugation lengths.  The  distribution of conjugation lengths, typically between 10 and 30 monomers [30], produces an inhomogeneously broadened density of states (DOS) distribution.  Defects  Coherent Segments  Figure 2.6: Schematic diagram showing the segments of a polymer chain jointed by defects such as twists, bends or impurities.  14  Chapter 3 Experiment and Sample Preparation Reference [18] gives a detailed description of the basic upconversion system used in this work. This chapter will give a brief review and discuss the improvements made to optimize the pre-existing system to achieve a better sensitivity and to reduce the noise floor.  3.1 Optical System Overview Figure 3.1  shows a schematic  representation of the setup for the  time-resolving  experiment. A Spectra-Physics Mode-locked Titanium Sapphire (Ti:Sapph) laser is pumped by an A r  +  laser with a constant pumping power of 7 W . The Titanium Sapphire laser is tunable  over a range of 735-820 nm with a maximum output power of 800 m W at 800 nm. When the laser system is well mode-locked, it emits "100 femtosecond (full width at half maximum intensity, sech (f) shaped) ultra-short laser pulses with a repetition rate of 82 M H z . 2  The TkSapph laser beam is split by BS1 into an excitation beam and a gate beam. The gate beam first hits a retro-reflector mounted on a translation stage, then is directed by mirrors M 3 , M 4 , M 5 and M 8 , and finally focused by L 3 onto the surface of the non-linear crystal, C 2 , for upconversion. The excitation beam first hits M 6 and another retro-reflector mounted on a mini-shaker whose frequency can be adjusted from 5 H z to 30 H z . This mini-shaker is useful in finding the zero delay point described later in section 3.3. Directed by M 7 , the excitation beam is focused by L I onto a non-linear crystal C I where it is frequency doubled to generate  Chapter 3 Experiment and Sample Preparation  \5  U V pulses with the same repetition rate as the gate beam. The U V beam is directed by M 9 and M 1 0 , and focused by L 2 onto the sample in the cryostat. The photoluminescence excited by the U V pulse is collimated by a 9 0 ° parabolic mirror (PM), directed by M 1 1 / M 1 2 and focused by L 4 onto the same spot as the gate beam on crystal C 2 . The sum frequency signal generated in C2 is collimated by L 5 , directed by M 1 3 / M 1 4 and focused by L 6 onto the entrance slit of the monochromator. Filter F l is used to filter out the second harmonic background from the strong gate pulses (even though the crystal C I is oriented to phase match co and (o , there is significant, non-phasematched second harmonic g  pl  generation at 2co due to the strength of the gate beam). Photon counting is done by either a g  photo-multiplier tube ( P M T ) connected to a computer-monitored discriminator or a liquid nitrogen cooled C C D detector. Table 3.1 gives the major specifications  of all die optical  components used in the time-resolved photoluminescence experiments illustrated in Figure 3.1. The whole experimental system is controlled by Labview programs via computer interfaces,  including rotation of the non-linear crystal C 2 , pulse counting from the P M T  discriminators and positioning the monochromator for a given detection wavelength. A C C D detector cooled by liquid nitrogen is also used for detecting very weak sum frequency signals. The C C D offers lower noise, but it is not as convenient as the P M T for measuring full spectra at a fixed delay. The P M T is made by Thorn E M I Electron Tubes L t d . (Type: 9784A). The C C D is from Princeton Instruments Inc. (Model: Specl0:100BR). The upconversion crystal C 2 is mounted on a three-dimensional translation stage on top of a motor-driven rotation stage. The surface of the crystal is parallel to the x direction of the translation stage. Z and y directions are used to align the crystal so that the focused spots of the  Chapter 3 Experiment and Sample Preparation gate and P L do not walk off when the crystal is rotated.  \6 A small piece of Teflon tape  (~2 x 4 mm) is placed on the surface of the crystal to image P L and gate pulse spots. A P a n a s o n i c ® C C D camera is used to view the surface of the crystal on a monitor. The sample is mounted using Teflon tape at the end of a one-meter-long probe inserted down the center of the cryostat. The cryostat was designed to cool the samples to 77 K with liquid nitrogen or to 4 K with liquid Helium. Polymer samples studied in this thesis were cooled to 77 K to avoid photo-degradation.  Chapter 3 Experiment and Sample Preparation  Table 3.1 Specifications of optical components in Figure 3.1  Symbols in Figure 3.1  Specifications  M1-M8 M12  Newport Round Pyrex Mirror 25.4 mm Diameter, 1/5 wave, 0.5-18 pm  Mil  50 x 50 mm Square Flat Mirror, 1/5 wave, 0.5-18 jum  M9-M10  Newport Round Pyrex Mirror 25.4 mm Diameter, 1/5 wave, 250-600 nm  M13-M14  Newport Round Pyrex Mirror 50.8 mm Diameter, 1/5 wave, 250-600 nm  CI  5x5x0.5 mm Beta-Barium Borate (BBO) Crystal, cut at 9 = 2 9 . 2 °  C2  5x5x0.5 mm Beta-Barium Borate (BBO) Crystal, cut* at 9 = 3 8 °  RF1-RF2  Newport Gold-coated retro-reflector  f  Ll  Focal length 5 cm  L2-L3  Focal length 10 cm  L4  Focal length 15 cm  L5  Focal length 7.5 cm, 250-430 nm, A R coated  L6  Focal length 7.5 cm, 250-430 nm, A R coated GSI-Lumonics Inc. Gold Coated Aluminum, 9 0 ° off Axis Parabolic  PM  Mirror 2.875 cm Focal length  BS1  50/50 beam splitter  FI  Newport UG11 filter or Schott D U G 1 1 X filter  Refer to Figure 2.3 for definition of cut angle. Refer to Figure 2.3 for definition of cut angle.  Chapter 3 Experiment and Sample Preparation  Figure 3.1: Schematic set-up for femto-second time resolved photoluminescence spectroscopy.  18  Chapter 3 Experiment and Sample Preparation  \g  3.2 Major Challenges and Solutions The previous setup used in Reference [18] had a sufficient signal to noise ratio to timeresolve the transient P L emitted by bulk P P V . The P L strength of the signal from samples studied in this work was considerably less than that of bulk P P V . Several changes were made to increase the system sensitivity by a factor of approximately three times. Details of these improvements are described in this section.  3.2.1 General Alignment Proper alignment of all components  is critical to achieving high sensitivity  in this  complex optical setup. The first phase of the procedure involves ensuring that all beam paths lie in a common plane parallel to the optical table, and that all focusing optics are aligned precisely on the optical axis of the beam passing through them.  3.2.2 Gate and PL Spot walk-off on Upconversion Crystal Usually, during an upconversion experiment, the nonlinear crystal C 2 needs to be rotated to maximize the signal at different wavelengths. The focused gate and P L spots on the crystal must not "walk off" each other as the crystal is rotated. Gate and P L beams must therefore be focused right on the rotation axis of the rotation stage. T o achieve this, first rotate crystal C 2 so that the gate beam is perpendicular to the surface of the crystal. Second, translate crystal C 2 in the x direction, so that the spot is focused on the Teflon and imaged by the C C D camera. Mark the position of spot on the screen. Now rotate crystal C 2 and adjust the x or y positions of the crystal until the gate spot does not walk when crystal C 2 is rotated. Finally,  20  Chapter 3 Experiment and Sample Preparation  adjust the position and focus of the P L beam so that it overlaps with the gate spot on the rotation stage axis.  3.2.3 Shape of Excitation Spot The U V excitation beam is generated by focusing the IR laser beam onto the non-linear crystal C I . In order to make sure the shape of the excitation spot on the sample is a circle for optimum overlap with the gate beam, the Beta-Barium Borate (BBO) crystal is cut at an angle of 2 9 . 2 °  to the optic axis. Such a cut angle allows normal incidence second harmonic  generation with "800 nm laser pulses, which minimizes the distortion in the second harmonic generation beam.  3.2.4 Nonlinear crystal for upconversion A negative uniaxial crystal ( B B O , n  e  < n ) was used for upconversion. For optimum a  results, the crystal needs to be cut properly so that it can upconvert a wide frequency range of the P L spectrum. In our upconversion experiment, the B B O crystal is cut at 3 8 ° . With such a cut angle, this B B O crystal can be used to upconvert P L wavelengths ranging from the visible (500 nm) to the IR (1200 nm), when used in combination with gate wavelengths from 745 nm800 nm. The polar plot in Figure 3.2 gives the phase matching angle as a function of the P L wavelength for a 798 nm gate pulse (thick solid curve) as obtained using the code in Appendix A . The angle between the gate and P L beams is taken to be 2 0 ° . Less than 3 0 ° of rotation of the crystal can cover the P L wavelength range of 500 nm-1200 nm.  Chapter 3 Experiment and Sample Preparation  21  Normal of Upconversion Crystal Gate Beam Incident Angle ofPLGnOfor Phase Matching  Figure 3.2: Polar plot of the phase matching angle as a function of the PL wavelength (thick solid curve). Wavelength of the gate was 798 nm. The external angle between the gate and the PL beams was 20°.  3.2.5 Gate and P L beam geometry T o measure the angle between the gate and the P L , the most accurate method is to determine the angles at which the crystal surface reflects the gate and the P L beams back on themselves,  using  the  Labview  program  rotate.vi to  rotate  the  crystal  and  using  MM2000TellPosition.vi to indicate the position of the crystal. This method is more straightforward and quicker than that described in Reference  [18]  which was based on second  harmonic generation from the gate and the reflected laser beam from the sample surface. The P L spectra of the organic semiconductors studied in this work are 150 nm -200 nm wide (taking M E H - P P V as an example, the range is 550 nm - 700 nm). For an angle of 1 8 ° between the gate and the P L beams, the angular difference between the sum frequency  Chapter 3 Experiment and Sample Preparation  22  propagation and the gate propagation under phase matching condition is plotted as a function of the P L wavelengths in Figure 3.3. The direction of sum frequency propagation only varies by ~1°  for P L wavelengths  between  550 nm to 700 nm. Thus,  once the collection optics  (M13/M14/L6) are well aligned for upconversion of one P L wavelength, even though the crystal C 2 is rotated for upconverting different parts (wavelengths) of the photoluminescence spectra, the adjustment of the collection optics is minor. W e just need to rotate the crystal C2 to the right angle  (calculated by the Maple code  listed in Appendix A ) , position the  monochromator to the right wavelength, and set the delay line translation stage to ~3 picosecond delay. A very minor adjustment of M 1 3 / M 1 4 then results in an optimized upconverted signal for the next P L wavelength. The same principal applies to other polymer samples which have a similar P L bandwidth.  Figure 3.3: The angular difference between the sum frequency propagation and the gate propagation as a function of PL wavelength under phase matched conditions. The wavelength of the gate was 798 nm, and the external angle between the gate and the PL was 18°.  Chapter 3 Experiment and Sample Preparation  23  3.2.6 PMT Dark Current The dark count rate of the detector directly impacts the achievable signal-to-noise-ratio (SNR). T h e P M T dark count rate is sensitive to the ambient temperature and internal heating effects. Practice has shown that if the room temperature is kept stable below 18° and the P M T is turned on immediately before the time-resolved experiment, the dark count rate starts at only 2-5 counts per second, and increases slowly to around 10 counts per second within three hours. A  typical run involves acquiring signal counts in 20 second integrations for each of ~70  different delays, where the background dark counts over the same 20 seconds are monitored every five delay points. Between each run at fixed wavelength, the P M T is turned off and a fan is used to cool the P M T before the next run.  3.2.7 LN/CCD as a Detector for Extremely Weak Signals  Although much effort has been implemented to reduce the background count rate of the P M T , it is still not low enough to obtain a good signal to noise ratio from some of the weaker sources studied here, which yield peak signals of <3 counts per second. T o overcome this problem, a liquid nitrogen cooled C C D detector from Princeton Instruments Inc. (Model: Specl0:100BR) was used in place of the P M T as a detector when the upconverted signal was very weak. T h e C C D detector is M g F l coated for broad-band operation in the 200 nm - 1100 nm range. Liquid nitrogen was used to cool the detector, so that it effectively has zero dark counts. The detector consists of a 1340 x 100 array of pixels and is specifically designed for spectroscopy measurements. The size of each pixel is 20 x 20 jum.  Chapter 3 Experiment and Sample Preparation  24  T o use the detector in upconversion experiments, we want to integrate every photon that reaches the detector. 35,000 (350 x 100) pixels are therefore combined to form a "super pixel" by setting an appropriate region of interest (ROD in the C C D control software [27]. T o correctly set the R O I , the monochromator is positioned to the peak wavelength of the S H G of gate, and the monochromator output is imaged on the full C C D chip. M 1 3 / M 1 4 are adjusted so that the image on the C C D chip is sharp and strong (typically over 1000 counts per second). A s the upconverted signal and the S H G originate from the same spot on the nonlinear crystal and pass through the same optics, they should be imaged at the same position on the C C D chip. Considering the wavelength difference between S H G and upconverted signal and the chromatic aberration of the optics, a ROI which is larger than the S H G image is set to make sure the region can detect every photon of the signal. In our experiment, a 350 x  100  rectangular R O I was centered on the S H G image. A complication arises due to the relatively large IR response of the C C D detector as compared to the P M T . T o isolate the detector from any unwanted light in the lab, three layers of black electrical tape were used to seal the gap between the C C D window and the wall of the monochromator. A l l the edges of the monochromator were also sealed with black electrical tape. The height of the entrance slit of the monochromator is narrowed down to ~2 mm by black electrical tape. Several pieces of black cardboard were used to shield the P L , scattered gate beam and the light from the computer monitor from entering the entrance slit of the monochromator. Most importantly, a D U G 1 1 X filter from Schott Glass Technology Inc. was used in place of the U G 1 1 filter to filter out the strong IR background of the gate. The  Chapter 3 Experiment and Sample Preparation D U G 1 1 X is a U G 1 1 filter with a special coating, which only has one transmission window in the 250 nm - 400 nm range. The C C D detector is equipped with a shutter in front of the chip to control photon integration time. For the above mentioned 350 x 100 R O I , the pure dark count rate with the shutter totally closed is 385+10/30 sec. With the D U G 1 1 X filter inserted in place of the U G 1 1 filter and with the pure dark count rate (385+10/30 sec) subtracted by the control software [27], the gate beam background is below 100/30 sec while the net upconverted signal is in the range of 50-250/30 sec, depending on the wavelength being upconverted and the delay. Data acquisition with the C C D detector is performed in Mode I described in Section 3.3.5. A typical run consisting of 30-40 delays and a 30 second integration time per delay takes about one hour. Occasionally, the integration is corrupted by high intensity spikes due to cosmic rays.  3.3 Upconversion Data Acquisition 3.3.1 The zero delay point: Laser upconversion After all the optical components preceding crystal C 2 are properly aligned, the next step is to determine the position of the delay translation stage ( D L T S ) that corresponds to zero time delay. The laser beam is split at BS1 into gate and excitation beams. The excitation beam hits the sample before being focused onto the upconversion crystal, C 2 . It is not practical to measure and set the two path lengths to be closer than ~3 cm. Therefore, laser upconversion is used to determine the exact zero delay point.  25  Chapter 3 Experiment and Sample Preparation  26  • Referring back to Figure 3.1, the frequency doubling crystal C I and focusing lens L I are removed from the excitation beam path. The sample in the cryostat is lifted slightly so that the laser is focused on the Teflon below the sample. The scattered laser beam is then focused on C 2 in the same manner as the P L . After the scattered spot and gate spot are overlapped on crystal C 2 and the crystal is rotated to the right angle for laser upconversion, a lens with a focal length of "200 mm focuses the non-phase-matched second harmonic generated from the gate onto a Si-detector. As the laser upconversion signal is generated at the same point in the crystal, and is the same wavelength as the S H G of the gate pulses,  the scattered laser  upconversion signal will also be focused on the Si-detector when it is available. T h e Si-detector is A C coupled to an oscilloscope, and the mini-shaker M S is turned on. The delay line translation stage D L T S is slowly moved, and as the translation stage passes through the zero delay point, a number of spikes appear on the oscilloscope screen. The spikes appear in pairs, one of each pair represents the forward and backward pass of the shaker through the zero delay point. T o find exactly the zero delay point, the delay translation stage is moved so all the spikes are evenly spaced on the oscilloscope screen. If the mini-shaker is then turned off and the Si-detector is D C coupled into the oscilloscope, the upconverted signal appears as a D C level. A combination of fine tweaking of the delay line and M 1 2 , and fine rotations (in 0 . 1 ° increments) of crystal C 2 can result in a scattered laser upconverted signal close to 1 volt. This scattered laser upconverted signal is visible to the eye on florescent paper and is a valuable tool for aligning the collection optics, the monochromator, and for positioning the P M T . The zero delay point determined from the scattered laser upconversion is not exactly the same as that for the actual P L upconversion, after the frequency doubling crystal C I and  Chapter 3 Experiment and Sample Preparation  27  L l are re-inserted. Experience has shown that they are within 5 mm apart on the delay line translation stage. The real zero delay point of the actual P L upconversion is determined after the P L upconversion signal is observed. Details of this are discussed in section 3.3.4.  3.3.2 Alignment of Collection Optics The collection optics consists of M 1 3 , M14 and L 6 . M 1 3 and M 1 4 are 2 inch diameter U V mirrors. L 6 is a U V lens mounted on a 3-D translation stage. Positioning L 6 is particularly important. The laser upconversion signal (LUS) is used to align the collection optics, including M 1 3 / M 1 4 and the P M T . This is preferred over using the second harmonic generation (SHG) of the gate alone, as the L U S is a dot while the S H G of the gate is ring shaped. Firstly, the entrance and middle slits of the monochromator are opened to 2000 pm,  and the lid of the  monochromator and the focus lens L 6 are removed. The L U S is collimated by L 5 , reflected by M 1 3 / M 1 4 to the entrance slit of the monochromator. By tweaking M 1 3 / M 1 4 , the L U S is aligned to be parallel to the optical stage and perpendicular to the entrance  slit of the  monochromator. After positioning the monochromator to the peak wavelength of the L U S , adjust the U V mirror and the U V lens inside the monochromator for maximum signal on P M T . Finally, replace the lid, and adjust L 6 for maximum signal on the P M T . Normally, 3-5 §  million counts per second can be achieved from the L U S .  The second grating of the monochromator was bypassed by placing a U V mirror and lens after the middle slit, for the reasons described in Reference [18]. The signal is directed to the detector through a hole in the side of the lid. §  Chapter 3 Experiment and Sample Preparation  28  3.3.3 Maximizing the Intensity of PL  The  nonlinear crystal C I and lens L l are re-inserted  for the  P L upconversion  experiment. The intensity of the P L should be maximized before being upconverted. The P L excited by U V pulses is collimated by the parabolic mirror (PM), reflected by M 1 1 / M 1 2 and focused by L 4 onto crystal C 2 . C2 is translated so that the P L focused on the Teflon tape is imaged by the C C D camera. A combination of tweaking L 2 and L 4 and moving the sample up and down results in a tightly focused, bright and round spot on the screen.  3.3.4 The Sum Frequency Signal  T o obtain good overlap of the P L and the gate beams on crystal C 2 , neutral density filters should be inserted into the gate path to reduce the intensity of the gate spot being imaged. When the attenuated gate and maximized P L spots are overlapped on the surface of the Teflon tape using the C C D camera, their respective diameters on the monitor should be ~1 cm and ~0.5 cm. C 2 is then translated back and the neutral density filters are removed so that the P L and gate beams are focused on the surface of the nonlinear crystal. The geometry of the gate, the P L and the crystal C 2 was described in section 3.2.5. C 2 is rotated to the phase matching angle for the wavelength 6f interest, as calculated by the Maple code (Appendix A ) . The delay line translation stage ( D L T S ) is set to approximately +5  ps  delay  from  the zero delay point determined  by laser  upconversion.  Next,  the  monochromator is positioned to the peak wavelength of the S H G from the gate alone and the P M T signal is maximized by adjusting M 1 3 / M 1 4 . Then the monochromator is set to the wavelength  corresponding  to  the  desired  upconverted  signal.  The  wavelength  of  the  Chapter 3 Experiment and Sample Preparation  29  upconverted signal for M E H - P P V P L (550 nm - 700 nm) with an 800 nm gate is in the range of 325-370 nm. A s the upconverted signal is much weaker than both the S H G and the P L background, a Newport UG11 filter that has a transmission window between 250 nm and 400 nm is inserted between M13 and L 5 to filter both of them. Most likely, a small signal should be observed.  If not,  adjust the delay line.  Once a small signal  is available, a  combination of adjusting M 1 3 / M 1 4 , rotating C 2 , and adjusting M 1 2 for better gate/PL overlap results in an optimized upconverted signal. The zero delay point of the upconverted signal of the P L is approximately determined by moving the D L T S  towards the "negative delay"  direction until the upconverted signal suddenly becomes equal to the background. The precise zero delay point position is not determined, and is not crucial for interpreting the data obtained to this point.  3.3.5 Data Collection  Two modes of data collection were used. Mode I consists of fixing the monochromator and varying the delay. Mode II involves scanning the monochromator at a given delay. For Mode I, with a high signal-to-noise ratio, the P M T counts were monitored in 20 second bins for each delay. The dark count rate of the P M T is controlled to be as low as possible using the method described in section 3.2.5. The P M T dark count background was recorded every five data points by blocking the excitation beam. The temperature of the sample is controlled within 0.1K. The best temperature stability was achieved when the liquid nitrogen level was above the bottom of the sample holder but well below the sample itself. A typical run consisting of 60-70 data points takes about one hour.  30  Chapter 3 Experiment and Sample Preparation For Mode II, the upconverted spectrum is scanned at given delays with a step-size of 0.5  nm. A t the shorter delays  (<30  ps), the more intensive spectra were scanned at 1  second/point. A t longer delays (50-200 ps), the weaker spectra was scanned at 2 second/point. Each spectrum consists of 80 data points.  3.3.6 White light upconversion The  spectra taken at fixed delay (Mode II) require calibration if they are to be  quantitatively analyzed. The reason that more than a single frequency can be upconverted for a given crystal orientation has partly to do with the large cone angle ( ~ 1 0 ° ) with which the P L signal is focused onto the upconversion crystal. The P L in fact is not incident at a single angle, but over a range of angles, and so different frequency components of the P L spectrum are simultaneously phase-matched. There is also a finite extent of ~ ± 1 . 5 ° to the phase-matching range for a given frequency at a fixed angle of incidence. In principal, the system could be calibrated by replacing the P L source with a diffuse, broadband white light source. This proved to be difficult to achieve without removing the entire cryostat. Instead, to perform white light upconversion, as shown in Figure 3.4, the white light of a 100W tungsten quartz halogen bulb (equivalent to a 3200 K blackbody, Model Oriel 77501) was coupled into an Ocean Optics Inc. fiber (model: P600-2-VIS/NIR) with a facet diameter of 100 pm. The light output from the fiber was collimated by a lens L  w  (focal length  25 mm) and then focused by L4 onto the nonlinear crystal, in the same manner as the P L was focused. The main difference between this and the P L source was the f-number of L4 imaged on the crystal. In the P L setup, it was ~f4, while in the white light setup it was ~f8. A l l other  Chapter 3 Experiment and Sample Preparation conditions such as the wavelength of the gate beam, angle between the gate beam and the P L beam (now the white light beam), collection optics and insertion of the UG11 filter were unchanged from the P L upconversion setup. The spectra of upconverted photons are measured in Mode II. A s the spectrum of polymer P L is in the visible range, a Newport GG.455 filter was inserted in the tungsten lamp light path to cut off wavelengths shorter than 455 nm, in order to eliminate strong white light background in the sum frequency beam. The spectral response curves were taken at different phase-matching angles, as presented in Chapter 4.  Monochromator  Figure 3.4: Schematic Setup for White Light Upconversion.  3\  Chapter 3 Experiment and Sample Preparation  32  3.4 Sample preparation In this thesis, two types of P P V derivatives, M E H - P P V and D P - P P V are studied. They were prepared in three forms. 1.  Bulk film on microscope slide glass;  2.  Encapsulated in anodized alumina on aluminum foil;  3.  Encapsulated in commercially anodized disks (Anodisc 13) from Whatman Inc.;[28]  4.  Encapsulated in silanized Anodisc material.  A l l the samples described below were prepared by members of D r . M . O . W o l f s group in the Department of Chemistry, University of British Columbia, Canada.  3.4.1 Bulk Samples: Film on microscope-slides 1. M E H - P P V  M E H - P P V films were synthesized using the method described in Reference [19]. 0.6%  by weight solution of M E H - P P V  in T H F (Tetrahydrofuran) was  spin-coated  A  onto  microscope slides. The chemical reaction route used to prepare the M E H - P P V is illustrated in Figure 3.5. 2. D P - P P V D P - P P V films were synthesized following a chlorine precursor process described in the literature [22]. First, a soluble chlorine precursor polymer was synthesized and dissolved in toluene. 0.03% by weight prepolymer in toluene was spin-coated onto microscope slides. After drying in vacuum, the sample was heated to 2 8 0 ° C for 2 hours. The chemical reaction route used to prepare the polymer is illustrated in Figure 3.6.  Chapter 3 Experiment and Sample Preparation  Figure3.6: Synthesis of DP-PPV.  33  Chapter 3 Experiment and Sample Preparation  34  3.4.2 Host Material Preparation  layer (anode)  Figure3.7: Set-up for Anodization of Aluminum. 1. Anodized Alumina on Aluminum Foil Anodization is a process in which an aluminum layer is oxidized electrochemically. Under certain conditions, self-organized nanopore structures are formed in the aluminumoxide (alumina) layer [20]. The aluminum foil, which is 0.13 mm thick with a purity of 99.99%, is first degreased with acetone, then it is anodized with 0.3 M oxalic acid ( C H 0 ) as electrolyte, 2  2  4  with an  applied voltage of 40-60 V . The anodization is complete once the current starts dropping from its steady value. This process normally takes about one hour. The sample is rinsed with distilled water. The pores are then widened by dipping the oxidized aluminum layer into 5 % phosphoric acid ( H P 0 ) for 45 minutes. A thin aluminum layer sometimes remains at the 3  4  bottom of the alumina matrix.  Figure 3.7 gives a schematic setup for the anodizing process.  35  Chapter 3 Experiment and Sample Preparation Figure 3.8 gives a schematic diagram of the aluminum sample before and after the anodization process. Figure 3.9 gives S E M pictures of anodized alumina on aluminum foil with both top and cross sectional views. The diameter of the pores varies between 50-90 nm across the sample. The pores are self-organized with a roughly hexagonal order but they are not uniform across the whole sample. The depth of the pores of the anodized alumina is approximately 10  pm.  A l foil  Alumina  Figure 3.8: Schematic graph of sample before (a) and after (b) anodizing.  Air Hole  Chapter 3 Experiment and Sample Preparation  Figure 3.9: SEM pictures of anodized alumina on aluminum foil. (a) Top view and (b) cross sectional view. (Images obtained by Keri Kwong)  36  Chapter 3 Experiment and Sample Preparation 2. Anodisc 13 membranes from Whatman Inc. Anodisc 13 membranes, manufactured by Whatman Inc., consist of a thin wafer of  Figure3.10: Pores of Anodisc [28].  porous alumina, which can be used as a host to light emitting materials. These porous membranes have an asymmetric structure [35]. The majority of the membrane is comprised of pores with 200 nm pore diameters with a "60 pm depth. The other side of the membrane contains ~20 nm pore diameters with a 2 pm depth. Figure 3.10 shows the pores in the anodisc membrane [28]. More information about this material can be found in References [28] [35]. 2.  Silanization of Anodisc  Silanization is a chemical process used to modify the internal walls of porous materials, such as alumina or silica. This process allows the interim of the pores to be modified with alkyl groups, as shown in Figure 3.11,  before the polymer chains are encapsulated. The  purpose of silanization of porous materials is to help encapsulate more polymers into the pores by modifying the interaction of the polymer chain with the silanized pore walls. In principal, the silanization process also offers better protection to the encapsulated polymer chains because  37  38  Chapter 3 Experiment and Sample Preparation the oxygen in the alumina, which is believed to be involved in the degradation of the polymer, is screened from the polymer chain by silanization [34] [23]. Some of the Anodisc membranes were silanized with M e S i C l (trimethylchlorosilane). Full details are described in Appendix C . 3  Figure 3.11  shows the chemical reaction route used for the silanization. Trithylammine  (N(CH ) ) is used as a catalyst which reacts first with the silane and then reacts with the 3  3  alumina. In this thesis, the Anodisc used to host D P - P P V was silanized, while the Anodisc used to host M E H - P P V was not silanized. R R-Si—R CI  +  N(CH ) 3  3  »•  R  R-Si-R +N(CH ) 3  R 1  D _ q i _  01  K  ,  OH  R  r\  I  \  +  I  OH  OH  —Al—O-AI—O—Al—  + |\|(OH)3 3  Porous alumina wall  ..N(CH )  R  I  r^  »~  3  3  gj  1  3  |_j  1  O O OH !, _ !, _ !, —Al—O-AI—O— Al—  Note: R= Methyl (TMS)  Figure3.11: Chemical reaction route for silanization of Anodisc.  Chapter 3 Experiment and Sample Preparation  39  3.4.3 Encapsulating MEH-PPV into anodized Alumina or Anodisc T o encapsulate M E H - P P V into the host materials described in section 3.4.2, the host material is immersed in a 0.025wt% M E H - P P V T H F solution for two days in the dark, under nitrogen atmosphere to avoid photodegradation. The sample is then rinsed with distilled water, dried and stored in a sealed bottle filled with nitrogen.  3.4.4 Encapsulating DP-PPV into anodized Alumina or Anodisc A s D P - P P V is not soluble in T H F , a soluble D P - P P V prepolymer was synthesized as described in section 3.4.1. The host material is immersed in 0.03%  by weight D P - P P V  prepolymer toluene solution for two days in the dark, nitrogen atmosphere to avoid photo degradation. The sample was then rinsed with distilled water, dried in vacuum and heated to 2 8 0 ° C for 2 hours to convert the prepolymer to D P - P P V .  40  Chapter 4 Experimental Results Table 4.1 summarizes the properties of the M E H - P P V and D P - P P V samples and their labels used in this work. Methods of preparing these samples were described in Section 3.4. In the rest of this thesis, bulk M E H - P P V or bulk D P - P P V refers to the polymer on microscopic slides, Anodisc M E H - P P V or Anodisc D P - P P V refers to the polymer encapsulated in the Anodisc 13 membranes from Whatman Inc., and Anodized M E H - P P V or Anodized D P - P P V refers  to the polymer encapsulated  in the anodized alumina made by members of D r .  M . O . W o l f ' s group in U B C .  Table 4.1 Sample Properties and Sample Labels ^ - ^ ^ ^  Sample 1 .ibeh  Hulk  Anudi/cd  \nodisc  Bulk  -\iiodi/vd  A iiud isc  MLH-PPV  MEH-PPV  MEH-PPY  DP-PPY  DP-PPY  DP-PP\  Pol j met Type  MEH-PPV  MEH-PPV  MEH-PPV  DP-PPV  DP-PPV  DP-PPV  HoslTspc  Glass Slide  Anodized Alumina on  Anodisc 13 (Whatman Inc.)  Anodized Alumina on Aluminum Foil  Anodisc 13 (Whatman Inc.)  50-90 nm diameter pores, 10 nm thick  Side 1: -200 nm diameter pores, —60 ^m thick Side 2: -20 nm diameter pores, —2 |xm thick  No  Yes  I'lopcMlc-i  Glass Slide  Aluminum Foil HUM  Flat  50-90 nm diameter pores, 10 (im thick  Side l:~200nm diameter pores, ~60 Mm thick Side 2:-20 nm diameter pores,  Flat  ~2 (mi thick Sil.-nii/ed''  -  No  No  -  Note: Anodized samples and anodisc samples are both referred to as "nanostructured samples".  Chapter 4 Experimental Results  41  4.1 Time-integrated Spectra 4.1.1. Effects of Nanoporous Hosts The time-integrated photoluminescence spectra for bulk M E H - P P V and bulk D P - P P V at 77 K are plotted in Figure 4.1. Their shape is quite similar to the published [16] spectra from other conjugated polymers such as P P V . Referring to Section 2.2, in each of the spectra, the large peaks at the highest energies (2.05 e V for M E H - P P V and 2.45 e V for D P - P P V ) result from a purely electronic transition from the bottom of the first excited state S  to the  2  bottom of the ground state So (the S _ or the 0-phonon transition). The less intensive peaks at 0  0  lower energies (1.90 e V for M E H - P P V and 2.30 e V for D P - P P V ) are attributed to a onephonon assisted transition from the bottom of 5  to the first vibrational state in S , and are  7  0  0-phonon 2.45 eV  0-phonon 2.05 eV  § o  Z  0.0 1.8  2.0  2.2  2.4  1.8  Energy (eV)  2.0  2.2  2.4  2.6  2.8  3.0  Energy (eV)  Figure 4.1: Photoluminescence Spectra of (a) bulk MEH-PPV and (b) bulk DP-PPV at 77 K. Excitation wavelength was 398 nm. referred to as the S . or the 1-phonon transitions. The less obvious little "bump" near 2.11 e V 0 }  for D P - P P V is attributed to a two-phonon assisted transition and is referred to as the S _ or the 0  2  2-phonon transition. For bulk M E H - P P V , the S _ or 2-phonon transition was not observed. 0  2  The broad peaks reflect the inhomogeneously broadened density of states (DOS) due to the distribution of effective conjugation lengths ( E C L ) in the polymer chains.  Chapter 4 Experimental Results  Energy (eV)  42  Energy (eV)  Figure 4.2: Photoluminescence Spectra of (a)Anodisc MEH-PPV and (b) Anodisc DP-PPV at 77 K. Excitation wavelength was 398 nm.  The time-integrated photoluminescence spectra of Anodisc M E H - P P V and Anodisc DPPPV at 77 K are plotted in Figure 4.2. The encapsulation clearly results in a significant effect of the on the shape and energies of the spectra. In order to quantitatively analyze these spectra, multiple Gaussian functions are used to fit the time-integrated spectra of bulk and Anodisc samples for M E H - P P V and DP-PPV. In all cases considered, two-Gaussian functions do not yield as good fits as those obtained with threeGaussians. With 3-Gaussians, two of the three (labeled 1 and 2) always have similar line widths and are separated by the expected phonon energy, while the third has a different line width and energy. For direct comparison, the normalized P L spectra of M E H - P P V and DP-PPV at 77 K for both bulk and nanostructured samples are plotted in Figure 4.5. One can observe that there is a significant blue-shift for the P L of the nanostructured samples, ranging from 0.034 - 0.183 eV. A similar blue-shift (approximately 0.15 eV) was also observed by Zojer et al. on P P V chains isolated in a self-assembled lyotropic liquid crystal host [24,25]. In the nanostructured  Chapter 4 Experimental Results  43  samples, the amount of blue-shift of M E H - P P V is greater than that of D P - P P V . For the same guest polymer, the Anodisc samples result in a greater blue shift than the Anodized samples. The relative (with respect to the 0-0 transition peak) intensity of the 0-1 transition peak increases in the nanostructured samples compared to the bulk samples. Similar results were also observed by Zojor on isolated P P V polymer chain [25]. In addition, it is observed that the spectra of the nanostructured samples become wider, indicating a greater degree of inhomogeneous broadening of the E C L energy density of states. Similar results were also observed by Zojor on isolated P P V polymer chain [25]. Table 4.2 summarizes the observed blue shifts and the broadening of the 0-0 transition peaks, the relative amplitude of the 0-1 peak with respect to the 0-0 peak, and the blue shift of the third Gaussian for both M E H - P P V and D P - P P V in different host materials.  Table 4.2 Comparison of Spectra of Bulk and Nanostructured M E H - P P V and D P - P P V  Sample Label Anodized MEH-PPV Anodisc MEH-PPV Anodized DP-PPV Anodisc DP-PPV Bulk MEH-PPV Hulk DP-PPV  Blucsliift of 0-0 peak  Broadening of 0-0 peak  \mnliuulc 1-0 Amplitude 0-0  Blucsliift of third Gaussian  Broadening of third Gaussian  +100%  81%  0.163 e V  +115%  +35%  58%  0.101 e V  +10%  +0.058 e V +0.183 e V +0.034 e V +0.068 e V  25% 35%  Chapter 4 Experimental Results  1.7  1.8  44  1.9  2.0  2.1  2.2 2.3  2.0  2.2  2.4  2.6  2.8  1.9  2.0 2.1  2.3  2.0  2.2  2.4  2.6  2.8  Energy (eV)  Energy (eV)  Figure 4.3: Two-Gaussian tits for (a) bulk MEH-PPV and (b) bulk DP-PPV P L at 77 K and three-Gaussian fits for (c) bulk MEH-PPV and (d) bulk DP-PPV P L at 77 K. In all cases, the dashed lines show the individual Gaussians, and the solid lines show the sum of the Gaussians.  Figure 4.4: Two-Gaussian fits for (a) Anodisc MEH-PPV and (b) Anodisc DP-PPV PL at 77 K and three-Gaussian fits for (c) Anodisc MEH-PPV and (d) Anodisc DP-PPV P L at 77 K. In all cases, the dashed lines show the individual Gaussians, and the solid lines show the sum of the Gaussians.  Chapter 4 Experimental Results  1.8  2.0  T  1  T  2.2  2.4  2.6  2.8  3.0  Energy(eV)  Figure 4.5: Normalized 77 K PL spectra of MEH-PPV and DP-PPV for bulk (solid lines), Anodized (dashed lines) and Anodisc (dash-dotted lines) samples. Excitation wavelength was 398 nm.  Chapter 4 Experimental Results  47  4.1.2. Temperature Dependence The effects of temperature on the time-integrated photoluminescence of bulk M E H P P V , bulk D P - P P V and anodized D P - P P V are plotted in Figure 4.6. The data shows that both bulk and nanostructured samples exhibit similar temperature dependences. T h e S _ and S _j 0  0  0  peaks red-shift with decreasing temperature, which is attributed to the freezing out of torsional modes of the phenylene rings [29], resulting in a longer effective conjugation length and a lower emission energy.  2.5 -  I  2.0 tc  1.5  to c  1.0  I  I  77K - • • 110K 180K 210K  l  S /  '  /  I  J  ••'  \\v  4  —i  77K  - • • 120K  150K  \  210K  \  —— ——— —  0.5 0.0  / (a)  s - '  :  -  A \  / !\\  (C)  — "  i  i  i  i  i  1.8  2.0  2.2  2.4  2.6  1 2.8  1.8  2.0  2.2  2.4  2.6  Energy ( e V )  "T  3  c CD  0.0 h 2.0  2.2  2.4  2.6  2.8  Energy(eV) Figure 4.6 : Effects of Temperature on the time-integrated photoluminescence of (a) bulk DP-PPV, (b) anodized DP-PPV and (c) bulk MEH-PPV. Excitation wavelength was 398 nm.  2.8  Chapter 4 Experimental Results  48  4.1.3 Excitation Intensity Dependence With an excitation wavelength of 397 nm, the 77 K P L spectra of bulk D P - P P V have been measured at different excitation powers, as shown in Figure 4.7 (a). The intensity of the P L (at 510 nm) as a function of the excitation intensity is plotted in Figure 4.7 (b). The intensity of excitation I  ex  was monitored by inserting neutral density slides into the excitation  beam path. The power of the excitation without any neutral density filter in place was 0.1 m W . Figure 4.7  (a) shows that the shape of the P L spectrum does not depend on the  excitation intensity. In Figure 4.7 (b), I  pl  is found to proportional to l)£ , where the power 1.2  has an error of ± 0.2. The response of the system is therefore linear within the experimental uncertainty.  Figure 4.7: Intensity of bulk DP-PPV P L at 77 K versus excitation intensity. Wavelength of excitation was 397 nm.  Chapter 4 Experimental Results  49  4.1.4 Comparison of PL intensity from different samples With the same excitation power of ~0.1  m W and same excitation wavelength  of  400 nm, the intensity of the P L for different polymers in different forms varies significantly. In Figure 4.8, the P L spectra of bulk and nanostructured M E H - P P V and D P - P P V at 77 K are plotted. A m o n g the six spectra, the P L intensity of bulk M E H - P P V is the strongest, almost ten times stronger than bulk D P - P P V . The P L intensities of the two nanostructured M E H - P P V  400  450  500  550  600  650  700  750  Wavelength (nm) Figure 4.8: 77 K Photoluminescence spectra of MEH-PPV and DP-PPV in bulk and nanostructured forms. Solid line: bulk MEH-PPV, Short dashed line: Anodisc MEH-PPV, Dotted line: Anodized MEH-PPV, Long dashed line: Anodized DP-PPV, Solid line single dot: bulk DP-PPV, Solid line double dots: Anodisc DPPPV. Excitation wavelength was 400 nm.  Chapter 4 Experimental Results  50  samples reduced by factors of 100 and 1000 from the bulk for the anodized and anodisc hosts, respectively. O n the other hand, the intensities of the bulk and both nanostructured D P - P P V samples are similar, and considerably stronger that nanostructured M E H - P P V . It is tempting to interpret these results as implying that encapsulation has a strong quenching effect on M E H P P V , but not on D P - P P V . Unfortunately, there is currently too little knowledge about the relative or absolute polymer incorporation efficiencies in the nanoporous materials to draw any firm conclusions from this data. Its main significance here is to explain why it was impossible to obtain time-resolved data on nanostructured M E H - P P V .  4.1.5 Photodegradation While both bulk and nanostructured samples exhibit photodegradation, the effect is stronger in the nanostructured samples. Figure 4.9 shows a series of 77 K P L spectra obtained at different delays under constant excitation conditions for Anodisc M E H - P P V and Anodisc D P - P P V . It took about one minute to obtain each spectrum. It is noticed that the shape of the spectra does not change appreciably with photo-degradation. The intensity of the P L at the S .  0 0  peaks of Anodisc M E H - P P V and Anodisc D P - P P V are plotted versus exposure time in Figure 4.10. The intensity of the Anodisc M E H - P P V P L dropped by a factor of 2.3 in 60 minutes after which it continues to decay, but at a slower rate. O n the other hand, the intensity of Anodisc D P - P P V P L dropped by a factor of 2 in 20 minutes, after which it remains quite stable. W h e n the time-resolved experiment was carried out on Anodisc D P - P P V ,  for each  detection wavelength, a fresh spot on the sample was exposed to the excitation beam for 20 minutes. Data in Mode-I was then acquired after the P L intensity stabilized.  Chapter 4 Experimental Results In summary, based on the experimental results in Section 4.1.4 and 4.1.5, bulk M E H P P V is the most intense emitter. Time-resolved experiments were performed on bulk M E H P P V using the P M T as the sum frequency signal detector, bulk and Anodisc D P - P P V are the less intense emitters, so time-resolved experiments on them were performed using the liquid nitrogen cooled C C D . nanostructured M E H - P P V s are the least intense emitters and their upconverted signals could not be detected even using the C C D .  51  Chapter 4 Experimental Results  52  450  Wave'  1.2e+6 1.0e+6 3  8.0e+5  (/> -»-» c  O O  650  Wave' Figure 4.9: Evolution of Spectra of (a) Anodisc MEH-PPV and (b) Anodisc DP-PPV with excitation time. Wavelength of excitation was 400 nm.  Chapter 4 Experimental Results  140 Anodisc MEH-PPV  120  5 100 to c CD  80 60  •• •  40 0  20  40  60  80  100 120  1200 Anodisc DP-PPV  1000 Z3  800  -t—>  600 400 200 0 0  20 40 60 80 100 120 Excitation Time (min)  Figure 4.10: Photodegradation of Anodisc MEH-PPV and Anodisc DP-PPV as obtained by the intensity of their respective S .o peaks. 0  Chapter 4 Experimental Results  54  4.2 Time-resolved Spectra 4.2.1 Phase-matching calibration As  described in section 3.3.6, a tungsten lamp, which is equivalent to a 3200K-  blackbody radiation source, was used to evaluate the effectiveness of the Maple code used to estimate the phase-matching angle. The angle between the gate beam and the white light was set to 1 6 . 5 ° ± 0 . 1 ° , the same angle used in time-resolving experiments performed on bulk M E H P P V film. The measured upconverted spectra at different crystal orientation angles are plotted in Figure 4.11.  I = I + A exp 0  '  Considering the asymmetric nature of the spectra, a function of form  ( 1 - 1  0  )  2  A  + aX was used to fit them. Figure 4.12 plots the wavelength (k in  l  0  ^  J  the equation) at which the upconversion mixing efficiency is the maximum for a given angular position of the nonlinear crystal, as obtained from the fits, compared to the predictions from the Maple code in Appendix A . The above results show that the Maple code can give a good prediction for the phase matching angle with a small offset. The actual wavelengths of peak conversion efficiency are all ~ 0 . 2 ° greater than that predicted by the Maple Code for a given gate/PL geometry. In Figure 4.12, the error bars originate from the accuracy of the gate/PL angle (± 0 . 1 ° ) and the curve fitting to the white-light upconversion spectra have been taken into consideration.  Chapter 4 Experimental Results  Wavelength (nm) Figure 4.11 : White light upconversion spectra at different crystal orientation angles (incident angle of the gate, 0 i in Figure 2.4). Solid line is the fit obtained as described section 4.2.1. Wavelength of the gate: 797.5 nm. 2  Chapter 4 Experimental Results  56  380 H 700  370  «  360  H  650  350 h-  600  340 h  330 h 550  320 h6  7  8  Crystal Position (Degrees)  Figure 4.12: Solid dot: Wavelength at which the mixing efficiency is maximum for a given angular position of the nonlinear crystal, obtained from Figure 4.7. Solid line: results predicted by the Maple code in Appendix A (angle between the gate and the PL beams is 16.5°). Dotted lines: results predicted by the Maple code when the angles between the gate and PL beams are 16.4° and 16.6°.  Chapter 4 Experimental Results  57  4.2.2 Bulk MEH-PPV The  photoluminescence of bulk M E H - P P V  at 7 7 K was  time-resolved using the  upconversion technique described in Chapter 3. W e first consider the decay profiles that are measured using data collection Mode I. The decay at short time delays (0-20 ps) is plotted in Figure 4.13. The full decay curves (0-500 ps) are plotted in Figure 4.14. F r o m these plots one can see two distinct regimes ih the data: a fast decay which occurs on timescales less than 20 ps, followed by a slower tail which extends to hundreds of picoseconds. approach  of  McCutcheon's on bulk  P P V decay  profiles  [18],  the  Following the  data that  decrease  monotonically with delay were fit to a bi-exponential function of the form  I =I +w e-  t/Ti  0  where I  0  l  +w e~ ,  (4.1)  t/T2  2  is a constant background, z  }  and r  2  represent the short and long time constants,  respectively, and w and w are the respective weighting factors. The fits are superimposed on ;  2  the data points in Figure 4.14, and match the experimental data well. In Figure 4.15(b), the extracted fast and slow time constants are plotted versus the corresponding detection energies. These time constant data points can be categorized into three groups; each group is represented by a circle. Based on the model described in Reference [18], the time constants in the 0-0 group are associated with the pure electronic transition S . , while 0 0  the time constants in the 0-1 group are associated with one-phonon assisted transition  , and  the time constants in the 0-2 group are associated with two-phonon assisted transition S . . 2 0  Chapter 4 Experimental Results  58  800  594nm  580nm 400  *•»*•••  o < u </>  ^  O CO  r 200  600  . . .  •• ••  • •. .  400  o  200 \  o 0 10  15  20  800  10  15  20  800  611nm 600  •  •••  400 200  0  * 10  800  15  20  1200  624nm  640nm  1000  600 \  800 o 400  600  -I  400 O 200  200 0 5  10 Delay (ps)  15  20  5  10  15  Delay (ps)  Figure 4.13: Short time regime of time-resolved bulk MEH-PPV photoluminescence at 77 K.  20  Chapter 4 Experimental Results  „ I §. | o  °  1200 1000 800 600 400 200 0  800 1.94 ev  1.99 ev  600 400 200 0  0  100 200 300 400 500  0  Delay (ps)  800  •  100 200 300 400 500 Delay (ps)  800 2.05 ev  600 400 200 0 0  100 200 300 400 500 Delay (ps)  0  100 200 300 400 500 Delay (ps)  Figure 4.14: Bi-exponential fits of time-resolved bulk MEH-PPV photoluminescence at 77 K.  Chapter 4 Experimental Results  3 ro  1.2  c  0.8  cu  0.4 1.71(eV).  £  0.0  60  r  0-phonon 2.05(eV)  0)  2-phonon  1-phonon 1.90(eV)  N  1.6  1.8  2,0  2.2  Energy(eV)  1000  r  0-0 Group  to CL  100  -I—'  c to c o O ro -»—<  CD  E  10 0-2 Group  •  1 -  0.1  L  1.6  1.8  2.0  2.2  2.4  Energy(eV) Figure 4.15: Time constants extracted from the time-resolved decay profiles using the Bi-exponential fits.  Chapter 4 Experimental Results  13  CO  c  CD T3 CD N  E  o  1.7  z  1.8 1.9 2.0 2.1  2.2 2.3 2.4  Enerjgy (eV) co CD  E .CD  c o '  CL O CL  1000 (b)  100 .1  10 h 1 0.1  It  1.7  I  J  L  I  1.8 1.9 2.0 2.1 2.2 2.3 2.4 Energy(eV)  Figure 4.16 : Population lifetime versus estimated electronic energy. Time-integrated spectrum (a) includes the 3-Gaussians used to obtain an excellent fit to the data, (see Section 4.1).  Chapter 4 Experimental Results W e consider the time constants to be associated with a single D O S of the E C L of the coherent segments of the polymer chains, but that D O S is manifest via 0-0, 0-1 and 0-2 transitions. Thus to obtain a consistent energy dependence for the time constants, the points in 0-1 and 0-2 groups are translated one phonon energy 0.153 e V and two phonon energies 0.337 e V to higher energy, respectively. The one phonon energy was determined by measuring the energy separation between first and second Gaussians in the 3-Gaussian fit discussed in Section 4.1.1. As the 2-phonon transition peak is not in the range of our measurement, the two phonon energy of 0.337 e V was derived from the photoluminescence spectra in Reference [37]. In Figure 4.16, the translated time constants are plotted versus the estimated energy of the associated coherent segments. For energies above the 0-0 peak in the time-integrated spectrum (actually above the intermediate energy Gaussian shown in Figure 4.15(a)), the population lifetime decreases with increasing energy. This behavior is very similar to that obtained on P P V , the parent conjugation polymer of M E H - P P V , by McCutcheon [18], as shown in Figure 4.17. Reference [18].  The explanation of our results reinforces the hypothesis raised in  62  Chapter 4 Experimental Results  63  Energy (eV) Figure 4.17: Population lifetime versus E C L energy for bulk PPV, the parent polymer of MEH-PPV, at 77 K, from Reference [18]. (a) Bulk PPV P L at 77 K with three Gaussian fit in dashed line, (b) The dashed line is the estimated E C L density of states (DOS) from the 0-0 transition peak of the bulk PPV P L at 77 K.  Chapter 4 Experimental Results  64  4.2.3 Bulk DP-PPV Time-resolved experiments on bulk D P - P P V at 77 K were performed with probing energies at 2.23 e V , 2.27 e V , 2.30 e V , 2.36 e V , 2.43 e V and 2.50 e V , as shown in Figure 4.18. The intensity of the P L from bulk D P - P P V is less than one tenth that of bulk M E H - P P V . The peak intensity of the upconverted signal is ~3 counts per second, which was almost the same as the dark counts from the P M T . A L N / C C D detector was therefore used in place of the P M T as the photo detector (see Section 3.3.7). A s in Section 4.2.2, a bi-exponential function was used to fit the decay profiles, and the results are plotted in Figure 4.18, along with the data. Within the latitude allowed by the signal to noise ratio, the bi-exponential fits reproduce the data points reasonably well. However, for the 495 nm (2.50 eV) profile, a tri-exponential function was used to give a better fit, resulting in a 20% reduction of the chi-squared value. A s the error of the longest time constant is very large (1500+13500 ps), the longest life constant is omitted. In Figure 4.19 the extracted time constants are plotted versus probe energy. Again, the time constants can be categorized into three groups. After translating them by the onephonon energy (0.15 eV) for the 0-1 group and the two-phonon energy (0.34  eV) for the 0-2  group, respectively, we obtain the time constants as a function of coherent segment energy, as shown in Figure 4.20.  The one-phonon energy (0.15 eV) was obtained by measuring the  separation of the first and second Gaussians of the three-Gaussian fits to the time-integrated P L spectrum of bulk D P - P P V , as discussed in Section 4.1.1. The 2-phonon peak is actually a small "bump". The position of the "bump" is at 2.11 eV, the two-phonon energy is therefore  Chapter 4 Experimental Results  65  estimated to be 0.34 e V , very close to the results quoted in Reference [36]. T h e symmetric inhomogeneously broadened E C L density of states obtained from the three-Gaussian fit in Section 4.1.1  is also plotted. Again, the population lifetime decreases with the increasing  energy, or the decreasing E C L of the coherent segments in the polymer.  1.8  2.0  2.2  2.4  2.6  2.8  3.0  Energy (eV) Figure 4.19 : Extracted time constants versus probe energy for bulk DP-PPV PL at 77 K.  Chapter 4 Experimental Results  -100  100  200  66  300  300  300  -100  100  200  300  I  I  I  140  •  2.50 e V  •s.  100 Delay (ps)  200  300  -100  "  —  i  i  +  100  200  300  Delay (ps)  Figure 4.18: Time-resolved profiles for bulk DP-PPV photoluminescence at 77 K.  Chapter 4 Experimental Results  67  A  ^ m  ^ 1.0 g . 3  o.8 & 0  Q.  0.6 m O  0.2 0.1 t 0.0 1.8 2.0 2.2 2.4 2.6 2.8 3.0 E n e r g y (eV) Figure 4.20: Population Lifetime versus estimated energy of the corresponding coherent segments for bulk DP-PPV at 77 K.  o CD c  Chapter 4 Experimental Results  68  4.2.4 Anodisc DP-PPV  Time-resolved experiments were performed on Anodisc D P - P P V at 77 K at probe energies of 2.30 e V , 2.36 e V , 2.43 e V , 2.54 eV and 2.58 e V . A t 77 K the intensity of the Anodisc D P - P P V P L is similar to that of the bulk D P - P P V .  The L N / C C D  detector  was  therefore used to detect the upconverted signal. The bi-exponential function was again used to fit the decay profiles, and the results are shown with the data in Figure 4.21. Similar to bulk D P - P P V and bulk M E H - P P V ,  the bi-exponential fits reproduce the data points reasonably  well. However, here for the 480 nm (2.58 eV) profile, a tri-exponential function was used to give a better fit resulting in a 20% reduction of the chi-squared value. But as the longest lifetime has a very large error (940+6000 ps), it is omitted. For the 540 nm (2.30 eV) profile, a single exponential was used. In Figure 4.22, the extracted time constants are plotted versus probe energy. Again, the time constants are categorized into three groups and translated by the one-phonon energy (0.15 eV) for the 0-1 group and the two-phonon energy (0.34 eV) for the 0-2 group, respectively. Again, the one-phonon energy (0.15 eV) was obtained by measuring the separation of the first and second Gaussians of the three-Gaussian fits to the time-integrated P L spectrum of Anodisc D P - P P V , as per discussed in Section 4.1.1. A s expected the onephonon energy is the same as that obtained from bulk D P - P P V . The 2-phonon transition peak in Anodisc D P - P P V time-integrated P L spectrum is not obvious. A s the encapsulation of polymer chains into the nanopores should not change the structures of the material, it is assumed that both the bulk and Anodisc D P - P P V have the same energy separation between the first and second vibrational levels in the S state. Thus the two-phonon energy was taken to be 0  the same as that obtained from bulk D P - P P V .  Chapter 4 Experimental Results The  results  after  shifting  69 are  shown  in  Figure  4.23.  Again,  the  symmetric  inhomogeneously broadened E C L density of states obtained from the three-Gaussian fit in Section 4.1.1 is also plotted, and the population lifetime decreases with the increasing energy, or the decreasing E C L of the coherent segments in the polymer. It is noticed that the dependence on the E C L energy below the D O S peak becomes more flat as compared to that above the D O S peak.  1000 CO Q _  CO  -«—•  0-0 group  100  CD  -i— i  CO  c o  O  CD  E  10  0-1 group  1 -  o.i L 1.8  0-2 group  2.0  2.2  2.4  2.6  2.8 3.0  Energy (eV) Figure 4.22: Extracted time constants versus probe energy for Anodisc DP-PPV P L at 77 K.  Chapter 4 Experimental Results  70  200 o <u co  o  o CO  O CO  c o  c o  CD  (0  O  70 FF o  50  O CO  40  c  30  O  200  •  60  O  100 h  o  100  Q> CO  150 h  300  100  200  300  100  200  300  2.43 e V cu co o CO  13 o  20 -  O  10 -100  200  300  -100  0  Delay (ps)  o a> co o co CO  c o  O  0  100  300  Delay (ps) Figure 4.21 : Decay profiles for Anodisc DP-PPV photoluminescence at 77 K.  Chapter 4 Experimental Results  -*—*  w  £Z CD  CD  "ro E 2.8  3.0  Energy (ejV) 1000 F CO CL  E  100  I  I  I  1  4 1.0 7j  i  10 i i i  -4—•  CD  0.1  1  ~ \ \  •f-  t"  4  1.8  2.0  2.2  0)  0.6  m O o O co  2.6  0.2  i—»-  CD Q.  0.0 b  'l  2.4  0.8  0.4 \  m  2.8  3.0  Energy (eV) Figure 4.23 : Population lifetime versus estimated energy of the corresponding coherent segment for Anodisc DP-PPV.  Chapter 4 Experimental Results  72  4.2.5 Bulk vs. Anodisc DP-PPV: Lifetime Dependence on Energy The population lifetime dependence on energy for both bulk D P - P P V and Anodisc D P P P V shown in Figure 4.20 and Figure 4.23 look similar. Figure 4.24 shows both sets of data on the same graph for easy comparison. The time constants for Anodisc D P - P P V have been shifted -0.068 e V , the blue-shift energy extracted from the time-integrated spectra of bulk and Anodisc D P - P P V as discussed in section 4.1.1, so that the peaks of D O S s associated with the 0-0 peaks are aligned. The dependence of population lifetime on the estimated E C L energy for the coherent segments, near and above the respective E C L energy D O S peaks, are  remarkably  similar for the bulk and Anodisc D P - P P V .  1000  o —i  100  i I  CO Q_  o  O •  <  10  3  Bulk DP-PPV Anodisc DP-PPV shifted -0.068ev from raw data  Q3_  1.0  \  1  Q.  m CO  If 1  N CD  0) r-t-  *L  0.5 02  k  CD CL  m O D O (J)  v \  H 0.0 0.1  l_  2.2  2.4  2.6  2.8  3.0  Energy ( e V ) Figure 4.24:Comparison of population lifetimes of bulk and Anodisc DP-PPV. Time constants of Anodisc DP-PPV have been shifted -0.068 eV, the blue-shift energy of the 0-0 transition extracted from the time-integrated spectra of bulk and Anodisc DP-PPV, to align the DOS peaks, shown in dashed line for bulk DP-PPV and dot-dashed line for Anodisc DP-PPV.  Chapter 4 Experimental Results  73  4.2.6 Bulk MEH-PPV vs. Bulk PPV: Lifetime Dependence on Energy The population lifetime dependence on energy for both bulk M E H - P P V and bulk P P V shown in Figure 4.16 and Figure 4.17 also look similar. T o compare them, Figure 4.25 shows their population lifetime constants on the same graph for easy comparison. The time constants of bulk M E H - P P V have been shifted +0.289 e V , the energy difference of the respective 0-0 transition peaks, so that the estimated E C L D O S peaks are aligned. The population lifetimes of bulk M E H - P P V are significant longer that those of bulk P P V , but their dependence on the energy with respect to the peak in the respective D O S is very similar.  1000  100  Mi  y-  10 •I  • O  I  *  Bulk P P V Bulk M E H - P P V shifted +0.289ev from raw data  1.0  £  H 0.5  ui.  ii •l \  \ \  /l  0.0  0.1 2.2  2.3  2.4  2.5  2.6  2.7  2.8 2.9  Energy (eV)  Figure 4.25:Comparison of population lifetimes of bulk PPV and bulk MEH-PPV. Time constants of bulk MEH-PPV have been shifted +0.289 eV, the energy difference of the 0-0 transition extracted from the timeintegrated spectra of bulk PPV and bulk MEH-PPV, to align the DOS peaks, shown in dashed line for bulk PPV and dot-dot-dashed line for bulk MEH-PPV.  Chapter 4 Experimental Results  74  4.2.7 Lifetime dependence on Energy for PPV and derivatives In Figure 4.26, the population lifetime constants of bulk P P V , bulk M E H - P P V , bulk D P - P P V and Anodisc D P - P P V are shown on the same graph with the respective 0-0 transition peaks aligned for easy comparison. For clarity, all the error bars are omitted. The energy dependences of the lifetimes in P P V and these two P P V derivatives in bulk and different encapsulated forms are quite similar.  1000 O BULK DP-PPV Anodisc DP-PPV Bulk PPV • Bulk MEH-PPV A DOS Bulk DP-PPV DOS Anodisc DP-PPV DOS Bulk PPV DOS MEH-PPV  A  100 b- •  1.5  3 1.0  Q.  £  £D_ N' CD Q.  m a) .—»3 SD  10  CD Q.  .CD  0.5  m o r~ D O cn •QT  O  c  0.0 0.1 2.2  2.4  2.6  2.8  3.0  3.2  Energy(eV) Figure 4.26:Comparison of population lifetimes of PPV and its derivatives MEH-PPV and DP-PPV, in bulk or in Anodisc forms. Time constants of bulk PPV, bulk MEH-PPV and Anodisc DP-PPV have been shifted +0.128 eV, +0.416 eV and -0.068 eV, respectively, to align their respective 0-0 transition peaks.  75  Chapter 5 Summary and Conclusions 5.1 Summary 5.1.1 Experimental Apparatus A pre-existing photoluminescence upconversion system [18] used to time-resolve the optical emission from bulk P P V polymers was enhanced to increase the achievable signal to noise ratio by a factor of more than three. Using a liquid nitrogen cooled C C D detector, a new nonlinear upconversion crystal, and new alignment procedures, upconverted signals less than "0.3  counts  per second  have  been  detected.  The tuning curve for the  new  nonlinear  upconversion crystal ( B B O cut at 3 8 ° ) was calibrated by recording the upconverted spectrum obtained using a white-light source at various crystal angles. These enhancements  made it  possible to time-resolve the emission from M E H - P P V and D P - P P V light emitting polymers, even when they are encapsulated in pores formed in two different types of electrochemically etched alumina.  Chapter 5 Summary and Conclusions  76  5.1.2 Effects of Encapsulation on the Time-Integrated Luminescence Spectrum of MEH-PPV and DP-PPV at 77 K Members of D r . M . O . Wolf's group in the Chemistry Department of U B C successfully encapsulated two types of light emitting polymers ( M E H - P P V and D P - P P V ) in nanoporous. alumina. Porous alumina structures made at U B C had pore sizes of 50-90 nm, while porous alumina obtained commercially predominantly consisted of pores with diameters on the order of 200 nm. The time-integrated photoluminescence spectra of both types of polymers, in bulk form and when they are encapsulated in the two different alumina hosts, were obtained under identical excitation conditions at 77 K . For both polymers, there is a significant blue shift of the 0-0 transition peaks in both nanostructured samples as compared to the bulk samples. Referring to Table 4.2, for the same guest polymer, the Anodisc 13 alumina membrane from Whatman Inc. results in a larger blue shift than the "home-made" anodized alumina. The relative intensity of the 0-1 transition peak (with respective to the 0-0 transition peak) is increased for both nanostructured samples as compared to bulk samples. The relative intensity of the 0-1 transition increases from 25% (bulk) to 81% (Anodisc) for M E H - P P V and from 35% (bulk) to 58% (Anodisc) for D P - P P V . In addition, the spectra of the nanostructured samples are broader than their bulk counterparts. The F W H M s of the 0-0 peaks of Anodisc M E H - P P V and Anodisc D P - P P V P L increased by 100% and 35%, as compared to those of bulk M E H - P P V and bulk D P - P P V , respectively.  A relatively strong Gaussian shaped feature  distinct from the usual 0-n electronic/vibrational transitions is found to help in quantitatively  Chapter 5 Summary and Conclusions  77  fitting the data from the nanostructured samples.  There is some evidence of a weaker  contribution from this additional feature in the bulk M E H - P P V and D P - P P V spectra. By comparing the intensities of time-integrated P L for different polymers in different forms, it is tempting to imply that encapsulation has a strong quenching effect on M E H - P P V , but not on D P - P P V . Unfortunately, there is currently too little knowledge about the relative or absolute  polymer  incorporation efficiencies  in  the  nanoporous  materials  to  draw  any  conclusions from this data.  5.1.3 Effects of Encapsulation on the Time-Resolved Luminescence Spectrum of MEH-PPV and DP-PPV at 77 K Time-resolved upconversion experiments were performed on bulk M E H - P P V ,  bulk  D P - P P V and Anodisc D P - P P V . It was found that the population lifetime increases as the E C L of the coherent segments increases above the average segment length, for both bulk and nanostructured polymer samples. When the energy dependence of the population lifetime data for both bulk D P - P P V and Anodisc D P - P P V are shifted in energy by the difference in the respective 0-0 transitions, they exhibit very similar behavior. Applying the same shift to the population lifetime dependence on segment energy for bulk P P V and bulk M E H - P P V , it is found that the shape of the curves are almost identical, but the lifetimes of M E H - P P V are two to three times longer than those of the pure P P V . The similarity of the energy dependence of the lifetimes in pure P P V and these two P P V derivatives in bulk and different encapsulated forms is quite striking.  Chapter 5 Summary and Conclusions  78  5.2 Conclusions The time-integrated photoluminescence polymers  are substantially  nanoporous alumina.  spectra from both M E H - P P V  altered from their bulk form when they  and D P - P P V  are encapsulated  in  The changes, including blue shifts, broadening, and enhancements  in  phonon assisted transitions, are all greater for both P P V derivatives when encapsulated in commercial Anodisc material with randomly distributed pores,  "200  nm in diameter,  as  compared to alumina films with 50-90 nm diameter pores arranged in a regular hexagonal array.  A s the pores in both alumina hosts are considerably larger than the diameter of the  polymer molecules, the changes are most likely due to interactions between the polymers and the inside walls of the alumina pores. These results are very encouraging, and suggest that there is significant potential for modifying the optical emission properties of PPV-based polymers by controlling their interaction with alumina surfaces. The significance of the distinct new feature in the time-integrated data from the nanostructured samples is not understood, but it is a relatively large effect that should be pursued. It is unlikely that this new feature plays a significant role in the emission at short times (less than a few hundred picoseconds), as the energy dependence of the time-resolved data is remarkably similar for all bulk and nanostructured samples when considered with respect to the peak in the respective D O S . A l l four samples studied with the upconversion system exhibit lifetimes that decrease monotonically, and nearly exponentially with excess energy above the peak in the D O S .  This is qualitatively consistent with the interpretation of Kersting et al.  [38,39] who attribute the decay dynamics in terms of an intrachain diffusion mechanism of the  Chapter 5 Summary and Conclusions  79  photoexcited electron-hole pairs. These results suggest that the interactions with the alumina pore walls does not affect these intrachain dynamics on sub-nanosecond timescales. These results motivate future studies that should include sample characterization using complementary interface sensitive techniques.  The substantial blue shifts observed here in the  nanoporous materials, up to 0.183 e V , may prove useful in light emitting device applications if methods are developed to improve the emission efficiency and to reduce photodegradation. Further optical experiments, including analysis of polarization dependences and the influence of magnetic fields at temperatures from 4 K to room temperature are also warranted.  80  Reference [I]  D . Braun, et al. Appl. Phys. Lett., 58(1991) 1982  [2]  Q. Pei, etal. Science, 269(1995) 1086  [3]  G . Y u , etal. Science, 270 (1995) 1789  [4]  G . Y u , et al. Synthetic Metals, 111-112 (2000) 133  [5]  F . Hide, etal. Science, 273 (1996) 1833  [6]  T. Gabler, etal. Opt. Comm., 137(1997) 31  [7]  G . Y u , et al. Appl. Phys. Lett., 64(1994) 3422  [8]  Industrial analysis web site, e.g. http://news.cnet.com/news/0-1006-201-4745832-0.html  [9]  C . Lee, et al. Molecular Crystal and Liquid Crystal, 256(1994) 85  [10]  H . Ago, et al. Phys. Rev. B , 61(2000) 2286  [II]  M . Gao, et al. J. of Phys. Chem. B , 102(1998) 4096  [12]  M . Baraton, et al. Nanotechnology, 9(1998) 356  [13]  T . Nguyen, et al. Appl. Phys. Lett., 76(2000)2454  [14]  B . Schwartz, et al. Synthetic Metals, 116(2001) 35  [15]  Y . R. Shen, "The Principles of Non-linear Optics", N e w York: Wiley 1984  [16]  H . Mahr et al. Hirch, Opt. C o m m . , 13 (1975) 96  [17]  J. Shah, I E E E J. of Quantum Electronics, 24 (1998) 276  [18]  Murray McCutcheon, M . S c . thesis, University of British Columbia, 1999  [19]  U S Patent # 5189136, N . N . Barashkov etc. Synthetic Metals 75 (1995) 153  [20]  A . P. L i , et al, J. of Appl. Phys., 84 (1998) 6023  [21]  L . Rendu, etal. Opt. Materials, 17(2001) 175  Reference  81  [22]  Bing R. Hsieh, et. al. Macromolecules 31 (1998), 631  [23]  Junjun W u , et al. , J. Phys. Chem. B 103(1999), 2374  [24]  D . Gin etal, Synthetic Metals 101(1999) 52  [25]  E . Zojer etal, Synthetic Metals 102(1999)1270  [26]  N . C . Greenham, R H . Friend, Solid State Physics, 29(1995) 32  [27]  User's Manual, WinSpec, Princeton Instruments Spectroscopic Software, Version 2.4.M, September 20, 1999 Chapter 11  [28]  http ://www. whatman. com  [29]  K . Pichler et al. J. of Phys.: Condensed Matter, 5(1993) 7155  [30]  G . R. Hayes et al. Synthetic Metals 84(1997) 889  [31]  T. Nguyen et al. Science, 288(2000) 652  [32]  C . Y . Yanga, et al. Synthetic Metals 83(1996) 85  [33]  V . G . Dmitriev et al. "Handbook of Nonlinear Optical Crystals", Berlin: Springer, 1997  [34]  Personal communication with D r . Katja Rademacher and Keri Kwong in M . O . W o l f s group  [35]  A . W . Ott et al. , Chem. Materials, 9(1997), 707  [36]  B . R. Hsieh, et al., Advanced Materials, 7( 1995), 3 6  [37]  A . K . Sheridan, et al,  [38]  R. Kersting, et al, J. of Chem. Phys., 106(1997), 2850  [39]  R. Kersting, et al, Phys. Rev. Lett., 70(1993), 3820  Synthetic Metals, 111-112(2000), 531  82  Appendix A Maple code for calculating phase matching angle for B B O crystal cut at 3 8 ° .  Incident angle of PL corresponding to the right phasematching condition given by the code (Appendix A)  Entrance surface of Nonlinear Crystal  Exit surface of Nonlinear Crystal  restart; n02:=proc(x); # o r d i n a r y index squared f o r BBO 2.7359+0.01878/(x 2-0.01822)-0.01354*x 2; end: ne2:=proc(x); # e x t r a o r d i n a r y index squared f o r BBO 2.3753+0.01224/(x 2-0.01667)-0.01516*x 2: end: A  A  A  A  lambdaG:=0.798: lambdaPL:=0.610: anglediff:=20.5: phasematch:=proc(thetar)  Appendix local  83  kPL, kG, kS, nS, thetaM,lambdas,theta2i,thetali,alpha,beta, thetalt,theta2t,gamma,thetarn;  lambdas:=(1/lambdaG + 1/lambdaPL) (-1); # sum frequency beam thetali:=thetar*Pi/180; theta2i:=thetali - anglediff*Pi/180; theta2t:=arcsin(-sin(theta2i)/sqrt(n02(lambdaG))); thetalt:= arcsin(-sin(thetali)/sqrt(n02(lambdaPL))); alpha:=theta2t - t h e t a l t ; kPL:=sqrt(n02(lambdaPL))/lambdaPL: # PL beam wavevector kG:=sqrt(n02(lambdaG))/lambdaG: # Gate beam wavevector kS:=sqrt(kPL 2 + kG 2 - 2*kPL*kG*cos(Pi-alpha)); # Sum beam nS:=kS*lambdas; # Index o f sum r a y beta:=arccos((—kPL 2 + kG 2 + kS 2)/(2*kG*kS)); thetam:= -38*Pi/180 + beta - t h e t a 2 t ; A  A  A  A  A  A  (sin(thetarn)) 2 A  (l/nS 2-l/n02(lambdas))/(l/ne2(lambdas)-l/n02(lambdas)) ; A  end: angle:=fsolve(phasematch(thetar)=0,thetar); #Solve n u m e r i c a l l y f o r t h e t a r rotatefromgate:=anglediff-angle; lambdas:=(1/lambdaG + 1/lambdaPL) (—1)*1000; A  angle := 8.833027055  rotatefromgate := 11.66697295  lambdaS := 345.7244319  Appendix  84  Appendix B Maple code for calculating angular difference between the sum frequency propagation and the gate propagation, as a function of the P L wavelength. Wavelength of the gate is 798 nm. Angle between the gate and the P L is 1 8 ° .  restart; n02:=proc(x); # o r d i n a r y index squared f o r BBO 2.7359+0.01878/(x 2-0.01822)-0.01354*x 2; end: A  A  ne2:=proc(x); # e x t r a o r d i n a r y index squared 2.3753+0.01224/(x 2-0.01667)-0.01516*x 2: end: A  f o r BBO  A  lambdaG:=0.798 : anglediff:=20: phasematch:=proc(thetar) local kPL, kG, kS, nS, thetaM,lambdas,theta2i,thetali, alpha,beta,thetalt,theta2t,gamma,thetarn; lambdas:=(1/lambdaG + 1/lambdaPL) (-1); # sum frequency beam thetali:=thetar*Pi/180; theta2i:=thetali - anglediff*Pi/180; theta2t:=arcsin(-sin(theta2i)/sqrt(n02(lambdaG))); thetalt:= arcsin(-sin(thetali)/sqrt(n02(lambdaPL))); alpha:=theta2t - t h e t a l t ; kPL:=sqrt(n02(lambdaPL))/lambdaPL: # PL beam wavevector kG:=sqrt(n02(lambdaG))/lambdaG: # Gate beam wavevector kS:=sqrt(kPL 2 + kG 2 - 2*kPL*kG*cos(Pi-alpha)); # Sum beam nS:=kS*lambdaS; # Index o f sum r a y b e t a : = a r c c o s ( ( - k P L 2 + kG 2 + kS 2)/(2*kG*kS)); thetam:= -cutangle*Pi/180 + beta - t h e t a 2 t ; (sin(thetam)) 2 (l/nS 2-l/n02(lambdaS))/(l/ne2(lambdaS)-l/n02(lambdas)); end: A  A  A  A  A  A  A  A  Angle:=[seq(x[i],i=l..31)]: f o r i from 1 by 1 to 31 do lambdaPL:=0.550+(i-1)/200; PhaseMatchingl:=fsolve(phasematch(thetar)=0,thetar); lambdaS:=(1/lambdaG + 1/lambdaPL) (-1); thetalil:=PhaseMatchingl*Pi/180: theta2il:=thetalil - anglediff*Pi/180: A  Appendix  85  theta2tl:=arcsin(-sin(theta2il)/sqrt(n02(lambdaG))); thetaltl:=arcsin(-sin(thetalil)/sqrt(n02(lambdaPL))): alphal:=theta2tl - t h e t a l t l : kPLl:=sqrt(n02(lambdaPL))/lambdaPL: kGl:=sqrt(n02(lambdaG))/lambdaG: kSl:=sqrt(kPLl 2 + kGl 2 - 2*kPLl*kGl*cos(Pi-alphal)): nSl:=simplify(kSl*lambdaS); betal:=simplify(arccos((-kPLl 2 + kGl 2 + kSl 2)/(2*kGl*kSl))); thetaSl:=simplify(betal-theta2tl); thetaOutl:=simplify(arcsin(-sin(thetaSl)*nSl)); Angle[i]:=simplify(anglediff-PhaseMatchingl-thetaOutl*180/Pi); od: print(Angle); A  A  A  A  A  10.68959556,  10.64635300,  10.60345869,  10.56090737,  10.51869484,  10.47681629,  10.43526706,  10.39404323,  10.35314017,  10.31255381,  10.27228097,  10.23231600,  10.19265726,  10.15329943,  10.11423950,  10.07547399,  10.03699892,  9.998810817,  9.960906510,  9.923282806,  9.885936326,  9.848863880,  9.812062344,  9.775527888,  9.739259016,  9.703251993,  9.667502964,  9.632010977,  9.596771544,  9.561783216,  9 . 527041740  Appendix  86  Appendix C Procedures for Silanization of Anodisc 13 purchased from Whatman Inc. 1. Place Anodisc under vacuum and heat to 1 0 0 ° C for 24 hours to remove residual moisture. 2. Under N  2  environment, add 0.1 ml of dry trithylammine to catalyze the reaction  between alumina surface groups and silane. Then 1 ml trimethylchlorosilane was added to the Anodisc. The Anodisc was allowed to dark under N  2  for 2 hours and the excess was removed  by vacuum. 3. Wash the Anodisc two times with dry hexane. 4. D r y the Anodisc and heat to 1 0 0 ° C for 1 hour under N . 2  5. Wash the Anodisc with ethanol and distilled water and dry under vacuum.  

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 13 0
China 3 4
Ireland 2 0
City Views Downloads
Ashburn 10 0
Dublin 2 0
Shenzhen 2 4
Beijing 1 0
Sunnyvale 1 0
Redmond 1 0
Unknown 1 12

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0085715/manifest

Comment

Related Items