|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 INVESTIGATING THE PHOTODYNAMICS OF CONJUGATED POLYELECTROLYTE TITANIA FILMS By DEREK AUSTIN LAMONTAGNE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR T HE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009
2 2009 Derek Austin LaMontagne
3 To everyone who ever stopped to smell the pine trees
4 ACKNOWLEDGMENTS Many people have supported me and believed in me throughout my life I have had s o many teachers, coaches, and friends that I cannot name them all, but I would like to specifically mention the people who have helped me here at the University of Florida. My committee has been helpful in teaching me much about physical chemistry and lasers through the class es I took with them. They have also given me additional guidance when I have come to them with questions. So I would like to thank Dr. Alex Angerhofer for teaching me how to use EPR and for being very approachable with any question I had. I would like to thank Dr. Nic o l Omenetto for teaching me laser fundamentals and helping me calibrate my detector. I would like to thank Dr. So Hirata for teaching me spectroscopy and guiding me to a better fundamental understanding of physical chem istry. And I would like to thank Dr. Valeria Kleiman for being my advisor. And of course my life would not be fulfilling if it werent for my friends. So I would like to thank all of my group members: Sevnur, Shiori, Jonathan, Jaired, and Allison. I als o should thank some previous group members who helped me a bunch: Daniel and Aysun taught me how to use the laser and setup, and post -docs Gustavo and C.P. were great mentors and even better friends. I also would like to thank all of my fencing teammates f or trying to give me a challenge, and my best friend David for being cool Chemistry and sports are nice but family is more important. S o I would like to thank my entire family for their support. My dad has always had confidence in me and taught me to b e inquisitive in life. My mom has given me everything and always loved me with all her heart. My Nana has always been able to help me with anything, and my brother Dustin is simply the best.
5 TABLE OF CONTENTS page ACKNOW LEDGMENTS ...................................................................................................... 4 LIST OF TABLES ................................................................................................................ 7 LIST OF FIGURES .............................................................................................................. 8 LIST OF ABBREVIAT IONS .............................................................................................. 10 ABSTRACT ........................................................................................................................ 11 CHAPTER 1 BACKGROUND .......................................................................................................... 13 Energy Problems and Solutions ................................................................................. 13 Solar Cell Efficiency ............................................................................................. 15 The Theory of Sensitization ................................................................................. 16 Conjugated Polyelectrolytes ....................................................................................... 17 Polyelectrolytes as Sensitizers in Solar Cells ..................................................... 18 Titanium Dioxide (TiO2) ........................................................................................ 20 Excitons ....................................................................................................................... 23 Energy Transfer ................................................................................................... 23 Excimers ............................................................................................................... 24 Aims ............................................................................................................................ 24 2 FILM FABRICATION .................................................................................................. 26 Film Preparation Methodology: Layer -By -Layer Technique ...................................... 26 Film Thickness Calculations ....................................................................................... 29 3 LASER SET -UP AND SYSTEM MODIFICATIONS .................................................. 32 St eady -State Experiments .......................................................................................... 32 Decay of Films over Time .................................................................................... 33 Transient Absorption .................................................................................................. 35 Introduction ........................................................................................................... 35 Femtosecond Laser Source................................................................................. 37 Pump -Probe Set up ............................................................................................. 39 Detection System ................................................................................................. 41 Modifications ........................................................................................................ 42 Atomic Force Microscopy ........................................................................................... 43
6 4 TRANSIENT ABSORPTION RESULTS AND ANALYSIS ........................................ 47 Transient Absorption Results for 30 Bilayer Film ...................................................... 47 Bleach ................................................................................................................... 47 Photo induced Absorption (PIA) .......................................................................... 50 Stimulated Emission (SE) .................................................................................... 51 Assign ing Bands .................................................................................................. 53 Exciton Exciton Annihilation ....................................................................................... 54 Exciton Density Calculation ................................................................................. 55 Kinetics ................................................................................................................. 57 Data Fitting to Find Rate Constants .................................................................... 59 Exciton Diffusion Length Calculations ................................................................. 63 Effect of Changing Number of Bilayers ...................................................................... 66 5 CONCLUSIONS AND FUTURE WORK .................................................................... 71 Experimental D iscussion ............................................................................................ 71 Future Experiments .................................................................................................... 72 Conclusions ................................................................................................................ 75 Exciton Diffus ion Calculations ............................................................................. 75 PPE-SO3 Bilayer Film Dynamics ......................................................................... 76 Solar Cells ............................................................................................................ 76 LIST OF REFERENCES ................................................................................................... 77 BIOGRAPHICAL SKETCH ................................................................................................ 83
7 LIST OF TABLES Table page 2 -1 Cal culation of film layer thickness ......................................................................... 31 4 -1 Calculation of annihilation rate constant for 30 bilayer film. ................................. 63 4 -2 Calculation of diffusion length for 30 bilayer film. .................................................. 65 4 -3 Lifetimes of states using single exponential decay fittings. .................................. 70
8 LIST OF FIGURES Figure page 1 -1 Solar spectrum with PPE-SO3 overlap zoom shown as the inset. ....................... 14 1 -2 Dye Sensitized Solar Cell (DSC). .......................................................................... 17 1 -3 PPE-SO3 and PDDA+ polymer repeat units. ........................................................ 19 1 -4 Diagram of relative band gap energies of polymer/titania/FTO glass. ................. 20 1 -5 Steady state absorption a nd fluorescence spectra of TiO2 films on FTO glass. ....................................................................................................................... 21 1 -6 AFM image of TiO2 on FTO glass, showing mesoporous structure ..................... 22 2 -1 Layer by -Layer process. ........................................................................................ 27 3 -1 Absorption and fluorescence emission spectra of PDDA/PPE -SO3 films on TiO2-coated FTO -glass for 5, 10, 15, 20, 25, and 30 bilayer films. ...................... 33 3 -2 Absorption spectra of FTO -TiO2 PDDA/PPE-SO3 films before and after one week. ....................................................................................................................... 34 3 -3 Fluorescence spectra of FTO -TiO2 PDDA/PPE-SO3 films initially and one week after fabrication. ............................................................................................ 34 3 -4 Possible Transient Absorption detectable signals ................................................ 36 3 -5 Transient Absorption set up. .................................................................................. 39 3 -6 AFM image for A) 5, B) 10, C) 20, and D) 30 Bilayers. ......................................... 44 3 -7 AFM step -height measurement on 10 bilayer film .............................................. 45 4 -1 Complete TA data for 30 BL film excited with 50 nJ pump energy. ..................... 48 4 -2 30 BL film excited with 50 nJ pump energy, plotted selecting time cross sections to show the dynamics. ............................................................................. 49 4 -3 30 bilayer film excited with 50 nJ pump energy, plotted selecting wavelength cros s -sections to show the decays. ....................................................................... 50 4 -4 30 BL film excited with 50 nJ pump energy, plotted selecting time cross sections to show the dynamics. ............................................................................. 52 4 -5 Spectral band assignment from TA data using a Perrin-Jablonski diagram. ....... 53
9 4 -6 30 bilayer film bleach signal energy dependence ................................................. 56 4 -7 30 bilayers film excited with 50 nJ energy from 420 nm pump beam. ................. 60 4 -8 30 bilayer film excited with 30 nJ energy from 420 nm pump beam. ................... 60 4 -9 30 bilayer film excited with 15 nJ energy from 420 nm pump beam. ................... 61 4 -10 K values over time, for excitation energies of 50 nJ, 30 nJ, and 15 nJ. ............... 62 4 -11 30 bilayer film excited with 12 nJ at 420 nm ......................................................... 67 4 -12 20 bilayer film excited with 12 nJ at 420 nm. ........................................................ 68 4 -13 10 bilayer film excited with 12 nJ at 420 nm. ........................................................ 69 4 -14 5 bilayer film excited with 12 nJ at 420 nm. .......................................................... 69
10 LIST OF ABBREVIATIONS BL Bilayer CP Conjugated polymer CPE Conjugated polyelectrolyte EEA Exciton exciton annihilation ET Energy transfer FTO Fluorine -doped tin oxide FWHM Full width at half maximum LbL Layer by -Layer OPA Optical parametric amplifier PIA Photo in duced absorption PRU Polymer repeat unit SE Stimulated emission TA Transient absorption UV Ultra violet WLG White light generation/generator fs Femtosecond nm Nanometer ps Picosecond
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INVESTIGATING THE PHOTODYNAMICS OF CONJUGATED POLYELECTROLYTE TITANIA FILMS By Derek Austin LaMontagne Dec ember 2009 Chair: Valeria D. Kleiman Major: Chemistry This work reported in this thesis has used ultrafast transient absorption pumpprobe spectroscopy to generate a better underst anding of the crucial aspect of energy dynamics within the polymer bilayers and titania-sensitizer interface in multi bilayer poly mer -sensitized solar cells. Photodynamic characterizat ions are evaluated based on pump excitation dependence and variations i n the number of sensitizing bilayers used to increase the light harvesting efficiency and structural stabil ity of these coatings The sensitizing chromophoric conjugated polyelectrolyte (CPE) chosen for testing wa s PPESO3 because of its ability to form bilayers (with cationic molecules) which can increase total net absorption because of its recent studies in solution for comparison, and because of its possible potential for photovoltaics and biosensor use. An oppositely charged non-chromophoric polymer ic molecule, PDDA+, is needed to counterbalance the PPESO3 and allow build up of the film s bilayers using electrostatic forces. Besides increasing the ability of the chromophore to harvest light the Layer -b y L ayer synthesis of these films is shown to be an effective and simple approach to streamline production while improving device performance. Increasing the number of
12 layers is an easy way to increase photon collection efficiency in most cases, and the exact trends and limits to this technique have been evaluated so as to find an optimal layer count before recombination and other quenching processes begin to negatively affect the energy transfer efficiency. Analysis of ultrafast solid -state transient absorption experimental data on these film s ha s been carried out to describe e xciton migration pathways and lifetim es through the polymer layers as well as other photodynamic properties like the electron injection into the titania layer
13 CHAPTER 1 BACKGROUND Energy Problems and Solutions The need fo r inexpensive, clean, and renewable energy is considered one of the greatest problems the world will face during the next century. Harnessing solar energy is regarded as one of the best solutions, because of its global abundance and lack of pollutant bypr oducts. The goal, then, of engineers and scientists during the next generation will be to optimize solar cells to achieve the most efficient and economically m arketable solar cells possible. In 2008, the United States of Ameri ca used more than one hundred billion, b illion Joules of energy, more than 20% of the global consumption, and less than 7% of that energy came from renewable resources.1 2 The use of fossil fuels such as petroleum and coal in the U.S.A. alone released roughly six billion metric tons of carbon dioxide into the earths atmosphere.1 This behavior is unsustainable on two counts: 1) because there is a limited supply of non -renewable energy sources like fossil fuels, and 2) the burning of such emits harmful gases into the atmosphere which contribute to poorer quality air to b reathe as well as enhancing the greenhouse effect that warms the planet. To avoid the catastrophic effects associated with changes in the climate from these anthropogenic problems, scientists must create viable alternatives to compete with the ease of burning fossil fuels. The amount of energy radiated from our sun from the nuclear fusion reaction of H2 to He is staggering, some 4x1020 Joules every second (W).3 The intensity of that electromagnetic radiation when it hi ts E arth is 1,353 Wm2; however, absorption and scattering from our atmosphere attenuates that to lower that 1,000 Wm2 depending on
14 the angle of incidence. Figure 11 shows the solar irradiance spectrum of light which passes through the atmosphere.3,4 D espite that attenuation, across the globe, and on a yearly basis, 1.2x1017 W of potential energy reach the surface,2 and when one compares this number to the total human yearly demand for energy of 1.3x1013 W (roughly 10,000 times less), the true potential for solar cells to solve the energy crisis is seen.6 Even with the expected increase in world demand due to an increasing population and greater demand from developing countries, covering just 0.2% of the planets surface with 10% efficient solar cells s hould provide enough energy for the planets needs And this relatively stable energy source is not going anywhere, at least not for more than 10 billion years.3 500 1000 1500 2000 2500 3000 3500 4000 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 400 500 600 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Absorption and Flux (Irradiance: Wm-2nm-1)Wavelength (nm)Solar Spectrum PDDA+/PPE-SO3 Spectral Irradiance (Wm-2nm-1)Wavelength (nm)Direct Spectral Solar Irradiance (AM 1.5)Solar Spectrum Figure 11. Solar s pectrum with PPE SO3 overlap zoom shown as t he inset
15 Solar Cell Efficiency Since one of the main goals of scientists working on solar cells is to improve a device is given by the ratio of the maximum pow er output to the power input:3,8,32,53 in F OC SC inP F V I P P max (1 -1) where ISC = short -circuit current, VOC = open -circuit voltage, and FF is the fill factor, which is a measure of the square ness denoting the area, under the I -V curve. To obtain a large VOC, and thereby a better efficiency, a large energy gap in the chosen absorption material is necessary. Once the choices of absorber and corresponding redox electrolyte are made, the ISC is given by the photo-induced current of electrons throu gh the system. The photocurrent is influenced by many factors, a few of which will be explored in this proposal in detail, the most notable being the effect of nano crystalline film morphology here exemplified by TiO2 nanoparticles .8 As of 2008, the most efficient solar cell on record wa s a concentrated GaInP/GaInAs/Ge two-terminal cell which can harness air mass 1.5 solar light at an efficiency of 40.7 2.4%.5 Using single cell s ilicon, the original solar cell semiconductor material, an efficiency of 2 7.3 1% has been achieved over larger areas.5 Although these cells are very promising, it is important to note that they suffer from one or more of these common problems: they are expensive, they are hard to manufacture on a large scale, they are inflexible, and they cannot coexist with light harvesting plants or animals. Cheap, flexible, polymer -sensitized devices can fix these problems.
16 The Theory of Sensitization Dye -sensitized solar cells (DSCs) are also known as Grtzel cells, after Michael Grtzel who was the first (and still a prominent) developer of them.6 11 Unlike traditional first gene ration silicon solar cells, DSCs work by separating the processes of light harvesting and electron transfer by using a dye layer attached to a semiconductor oxid e film.10 The dye film layer absorbs a photon of light, which then almost immediately (<1012 s) moves via electron injection into the conduction band of the semiconductor (here TiO2). An interpenetrating electrolyte couple (often I3 -/3I-) then returns the oxidized dye to its ground state by donating an electron in a redox process (1010 s), two layers of transparent conducting glass on either side establish the electrodes, and an external load completes the circuit. A diagram o f this cell is found in F igure 12 If the dye is adsorbed onto a flat surface, it cannot efficiently absorb more than a small percentage of the incoming photons because its cross-sectional volume is essentially two -dimensional severely limiting the possible absorbance.8 A (1 -2) This equation for the typical absorbance (A) of a dye covered film shows that to cm2) or find a sensi tizer with a better molar cross -sect 2 mol1), the latter being simply the molar extinction coefficient multiplied by 1000. To change the former, one can increase the surface roughness by coating the film onto a nanoporous semiconductor, like fractal or nanocrystalline TiO2. G rtz el has shown that a yellow ruthenium dye (RuL3, L = 2,2 -bipyridyl 4,4 dicarboxylate) deposited on sol gel prepared fractal TiO2 harvests light with an efficiency near unity at the sensitizers absorption maximum of 470 nm.9 Grtzels group8 and others17,34 have recently
17 demonstrated even greater efficiencies of nanocrystalline TiO2 from colloidal suspensions as an undoped, wide band gap charge acceptor in DSCs and polymer sensitized cells. Figure 12 Dye -Sensitized Solar C ell (DSC). Using this idea of sensitization, new molecules are currently being designed to act as the sensitizing layer. One such class of these molecules is conjugated polyelectrolytes, and by building up layers of these to increase absorption, solar cel l efficiency may be able to be increased. Conjugated Polyelectrolytes Conjugated poly mers (CPs ) are a subclass of polymers which demonstrate c onjugation which refers to how the electron cloud along the covalently bonded e flow Electrolyte Conducting Glass Semiconductor Oxide Dy e Conducting Glass E (V) h Load I (current) S 0 /S + Recombination Electron Injection S Semiconductor F ermi Level Conduction Band Mediato r Regeneration C A T H O D E Redox Max Voltage F Doped SnO 2 (FTO)
18 backbone interact s through delocalization allowing for rapid excit on transfer As a matter of fact, these polyunsaturated compounds conductive properties have gained considerable interest since they were first developed by Shirakawa, MacDiarmid, and Heeger in 1977. CPs have been used to replace other semiconductors in such devices as light emitting diodes, batteries, field effect transistors, and sensors, specifically in the biological sciences. To elicit a wider range of applications, i f one adds ionic side chains to CPs to increase their solubility in polar solvents, then these polymers are called conjugated polyelectrolytes (CPEs).16 ,23 Polyelectrolytes as Sensitizers in Solar Cells CPEs as the sensitizing (film) layer for use in solar cells and other optoelectronic devices provide manufacturers with the benefits of macroscopic physical flexibility, high tunability via wavelength control from choice of CPE band gap, and in most cases a reduction in cost of materials.15 Recent studies conducted all over the world,13,24 26, 30 47 include detailed work creating novel conjugated polymers and polyelectrolytes,14 19 as well as the creation and study of Buckminsterfullerene derivatives and CPE blends for use in optoelectronic devices .24,41,45 Extended conjugation in CPEs reduces the minim um energy required to excite an electron from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUM O). Moreover th rough the addition of specially chosen aryl groups to the backbone of the CPE, the usual range of the band gap for this excitation is in the visible part of the electromagnetic spectrum, making CPEs ideal for solar cells or inversely tuna ble for light emitting diodes .76 In particular, the conjugated polyelectrolyte PPE -SO3 -, which contains a poly(pphenylene ethynylene) backbone with anionic 3 -sulfonatopropyloxy side groups,21 has
19 been the focus of numerous studies at the University of Florida .15,21 PPE-SO3 (Figure 1 -3, and often written without the formal charge) demonstrates an ab ility to be rapidly quenched by cationic species in solution and to form aggregates in certain solutions .15 17 PPE-SO3 has also been shown to exhibit man y useful properties, including strong light absorption and fluorescence, high solubility in biologically favorable solutions and high mobility of photo excit at ions However, ultrafast studies of the solid -state films made from this polyelectroly t e, which will ultimately be necessary to understand the complete energy transfer picture in this molecule should it be used in applications such as solar cell s have yet to be completed. Figure 13 PPE-SO3 (left) and PDDA+ (right) polymer repeat units For solar light harvesting, i t is necessary to evaluate the spectral overlap of the absorption of PPE-SO3 with the solar spectrum which can be seen in the inset of Figure 11 For optimal, single absorber solar cells, absorption over the entire spectrum cre ates the most efficient cells Admittedly, PPE-SO3 only absorbs at a relatively narrow region on the higher energy side of the spectrum. Nevertheless, the potential of this CPE to play a role in optoelectronic devices may still be sought through use in tandem cells or by maximizing the absorption at the wavelengths it does absorb.6
20 To do the latter, a design which maximizes the absorption cross-section has been implemented here by use of a mesoporous TiO2 interface and multiple bilayer build-up Titanium Dioxide (TiO2) Titanium dioxide, or titania, or TiO2, comes in multiple forms, but the research reported here start ed with it as a nan oparticle. Titania is useful in optoelectronic dev ices such as solar cells for three reasons: 1) it increases the absorption cross section of sensitizers attached to it compared to having just a flat, two-dime nsional sensitizing surface, 2) it facilitates charge transfer through a circuit becaus e of its conductive properties, and 3) it is abundant and a relatively cheap semiconductor. For an effective solar cell, the HOMO of the CPE must reside within the band gap of the titania, while the LUMO must reside within the conduction band.69,71 Ch ecking literature (redox) values for PPE SO3 -, TiO2, and the substrate, Figure 14 is produced.16 Figure 14. Diagram of relative band gap energies of polymer /titania/FTO glass. WF is the work function of FTO glass, CBE and VBE are the conduction and valence band edges of titania, and the difference between the HOMO and LUMO in PPE -SO3 is 2.6 eV = 420 nm, the excitation wavelength. The gray arrows indicate downhill energy migration via electron transfer. The difference between the valence and conduction band edge energies in titania is 3.20 eV, which corr esponds to near -UV absorption, so it will be transparent to 420 nm FTO CBE Vacuum Level (not to scale) PPE SO3 3.74 TiO 2 4.40 VBE 3.20 5.80 6.9 4 HOMO LUMO WF hv Energy (eV) relative to vacuum level
21 excitation. Furthermore, a downhill, favorable electron migration is possible due to the position of the band energies. Th e fluorinedoped tin oxide (FTO) glass substrate used in the work reported here was coated with 20 nm diameter nanoparticles (Solaronix ) which were immobilized in their nanocrystalline form through heating. This created a film transparent to the collection wavelengths. FTO -coated glass wa s chosen for its conductive pr operties and potential to accept electrons should th is system become an optoelectronic device. The titania needs to be heated (sintered) for 45 minutes to generate the anatase nanocrystalline structure on the glass Figure 15 shows the steady state abso rption and fluorescence of films produced this way, without any CPE layers. 300 400 500 600 700 800 0.0 0.5 1.0 1.5 2.0 2.5 Absorption Spectra of FTO Glass Slides with only TiO2Absorption (OD)Wavelength (nm) 500 600 700 800 0 5000 10000 15000 20000 Fluorescence (counts)Wavelength (nm)Fluorescence of FTO Glass Slide with only TiO2Peak ~532 nm Peak ~695 nm Figure 15 Steady state absorption (left) and fluorescence s pectra (right) of TiO2 f ilms on FTO glass The absorption shows that variations in 14 different titania films show little variation, and all absorb in the UV. The fluorescence with 420 nm excitation for four different spots on the same film show peaks at 532 and 695 nm. In Figure 15, t he absorption spectra graph plot s 14 different trials, one from each of the slides which were made. They all show UV absorption lower than 400 nm To assure reproducibility, t he front -face fluorescence spectra were collected upon
22 excitation of four different spots on the same film It can be seen on the right graph of Figure 15 that fluorescence uniformity throughout the film is not a problem. The fluorescence for TiO2, which show s two broad peaks at 532 and 695 nm, has low counts and detection in the transient absorption spectr a is not expected.67 Figure 16 AFM image of TiO2 on FTO glass, showing mesoporous structure. The raised areas, shown in brighter pink are the tops of the TiO2 nanoparticles, while the darker regions approach the glass substrate but still contain a l ayer of TiO2. Figure 16 shows an AFM image of a coating of TiO2 nanoparticles on FTO glass, made to the same thickness as that for the studies in this paper. The clusters of titania can clearly be seen, and this porous scaffold is what increases the surf ace area onto which the CPEs can adhere.
23 Excitons The idea of excitons as excited electron -hole bound pairs comes from molecular crystal studies.73 74 In those studies, the concept started as a description of the migration of the photo induced excitation from one molecule to another. In low energy excited crystal s tates the electrically neutral bound pair ( exciton ) is not a stationary state, but instead can be thought of as an excitation wave traveling throughout the crystal from the initial point of excitation.79 J. I. Frenkel was one of the first p eople to describe this phenomenon and the Frenkel exciton is used today to describe tightly bound electron hole pairs.74 In chemistry terms, however, the idea of an exciton can be thought of as a deloc alization of the energy in a conjugated polymer. In more detail, however, the exciton can be thought of as new wavefunctions generated from the coupling of chromophores. For PPE-SO3 -, the average excitation delocalization length is 5 polymer repeat units .75 Energy Transfer There are in general two me chanism s by which excitons move through films. The first is by energy transfer in the Coulombic coupling limit and the second is by energy transfer from electron exchange.54 Excitons in a triplet excited s tate usually undergo exchange, which is an intramolecular switching of the donor units excited electron with the acceptor units ground state electron. Because of the relativel y slow times for this to occur compared to energy transfer and because triplets are not expected to be seen because of lack of heavy metal atoms to promote intersystem crossing, thi s mechanism will not be discussed further in this thesis Coulombic e nergy transfer (ET), however, is a much faster process, and is favored by the shorter lived singlet excited states.51,54 In this non radiative ET, Coulombic
24 interactions between the donor unit and acceptor unit give rise to long range dipole dipole interaction, and result in transfer of the exciton s energy to the acceptor unit. The rate of ET depends on several factors, including the spectral overlap of the donors emi ssion and acceptors absorption and the distance between t he units (by a n inverse power of 6 for weak coupling) Excimers Because PPE -SO3 exhibits more efficient packi ng compared to less rigid polymers like polyfluorenes, PPE SO3 is more likely to form excimers An excimer is a dimer only in the excited state. The fluorescence emission band of an excimer is fe atureless and red-shifted, and for PPE -SO3 it can be seen in chapter 3 as a broad band at 525 nm. The unaggregated emission band for PPE -SO3 was found to be around 455 nm ( it can be seen as a shoulder in the fluorescence of the aggregates ). PPE-SO3 -s limited ability to rotate or fold on itself leads to -stacking, where phenyl rings of neighboring molecules line up face to face, which leads to increased aggregation and formation of excimers.51,75 Aims Characterization of the energy transfer processes in PDDA/PPE -SO3 films wa s the overall goal of this re search. If one can understand the speed at which these processes take place, one can design better materials for optimal solar cell s Specifically, the experiments being pr esented here show the excited state lifetimes and photo dynamic processes of the c hosen CPE. In addition, structural, AFM on the film surface and calculation of actual film thicknesses are presented. I t wa s anticipated that increa sing the number of bilayers would increase the absorption in a linear fashion. After a point, however, fil ms that are too thick may
25 negatively affect the electron injection process because the excitons will be unable to reach the semiconductor (TiO2) and therefore will not be able to cleave and contribute to the circuit of the cell. Finding this limit is impo rtant for designing better solar cells. After studying PPE -SO3 in depth, certain trends and equations were evaluated which help give a better picture of the energy transfer process in pol yelectrolyte films in general.
26 CHAPTER 2 FILM FABRICATION Steady s tate absorption, emission, and t ime -resolved femtosecond t ransient a bsorption (TA) wer e the focus of the lab work in characterizing these films. However, the production of the test films in a relatively easy and effic ient way was also d emonstrated. Film P reparation Methodology: Layer By -Layer Technique In order for these films to be produced on a large scale, as is the goal should they be used in the photovoltaics industry, the production method must be as free from unnecessary stress as possible. For thi s reason, spin-coating and evaporation techniques, which require significant mechanical energy use, were relinquished for the less strenuous method of Layer by L ayer deposition (Figure 2 -1 ).49 The Layer -by L ayer (LbL) technique can be done with no additional equipment other than the polymers, vials, solvent, pipettes, tweezers, and a dedicated scientist. No heating or temperature fluctuations are necessary. Neither are any mechanical devices, although this process can surely be automated for reproduc ibility. It should be noted however that a robotic dipper is available, but due to its rapid use of solvents and openness to the atmosphere, thereb y causing increased aggregation and titania degradation due to dissolved water, this resource was not employ ed. To start the Layer -by -L ayer process, a preliminary coating of TiO2 is doctor bladed and sintered at 45 0C for 30 minutes (to change TiO2 from the rutile to the more easily adsorbing anatase surface)27 onto a 1.75 cm x 1.26 cm area on a clean FTO co ated soda lime glass slide.51 The FTO -glass is cleaned via sonication in solutions of soap, deionized water, acetone, and isopropanol, followed by 20 minutes in a plasma
27 cleaner. After taking a steady state absorption spectrum on a film run in parallel (for later ba ckground correction), the slide s we re then ready to be coated with the first polyelectrolyte layers. Figure 21. Layer -by Layer process. The adsorption experiments involve using alternating layers of oppositely char ged polymers to create bilayers which can theoretically be built to any thickness. As mentioned before, the polyelectr olyte chromophore chosen was PPE -SO3 because of its degree of s tudy in solution for comparison,21 as well as its promise to be used for fabrication of sensors for ions, peptides, nucleic acids, and proteins.15 To create a nanostructured multilayer which binds electrostaticly, a non-chromophoric aliphatic TiO 2 TiO 2 Wash in Methanol Wash + + + + + TiO 2 TiO 2 + + + + + + FTO FTO FTO FTO
28 polycation, poly(diallyldimethylammonium chloride), or PDDA+,52 was chosen so as not to inter fere with the PPE SO3 (Figure 1 3 ). The cationic layer of PDDA+ was added first to the slide to adsorb onto the anatase form of TiO2. To achieve this, one allows the bottom half of the slide containing the TiO2 surface to soak in a 1mM solution (in methanol) of the counter cation PDDA+. After the prescribed wait period of 25 minutes, the slide is then run through three separate methanol washes (3, 1, and 1 minutes) and finally placed in to a 1mM solution (in methanol) of the PPE -SO3 anion polyelectrolyte for 25 minutes After this layer is added, the slide is run back through clean methanol vials and a new bilayer is ready to begin. The total time for one bilayer to be coated is one hour. Variations in the soak times were considered but becaus e this study focuse d on the photodynamics, previous working methods were decided to be better than new, untested soak times. It is expected, though, that increasing the thickness of a single bilayer merely creates a greater number of aggregated states as the polymers cluster together on the film. Aggregation may then quench the excitons of PPE -SO3 in a nonlinear fashion with increasing layer thickness which is undesirable. The relat ively short time of 25 minutes was chosen in order to limit aggregation while still providing complete film layer coverage. Using this strategy, an individual bi layer thickness should be about 2 to 3 nm, and films containing 100 or more bilayers can be fabricated. But for the s e experiment s the focus was to create films of 5, 10, 15, 20, 25, and 30 bilayer thicknesses ( in order to compare exciton diffusion before it reach es it s limit expected to be under 30 bilayers)13 and evaluate their linear optical properties.
29 Film Thickness Calculations Although the thickness of the ti tania layer of a few microns determines the main scaffold on which the films are built, it is possible and useful to calculate the relatively small additions to that thickness generated by the polymer bilayers. The Layer by Layer process impl emented here wa s expected to generate thickness as low as a few nanometers upwards of 50 nm.57,58 Thickness es are affected by length of time in the dipping solution, concentration of the solution, and the amount of interlayer diffusion which takes place during the coating process. It wa s further expected that if the interlayer diffusion is low, then a linear relation would be seen as bilayers were built up.59,60 The absorbance of a 10 M PPESO3 in methanol solution at its maximum absorption wavelength of 420 nm is 0.57 although this value has been found to vary between 0.40 and 0.60 for different stock solutions .75 Using A = 0.57 and the fact that it was taken in a 1 cm cuvette, one can calculate the molar absorption coefficient ( ) from: lc A (2 -1) Where A is the absorbance, l is the length of the cuvette, and c is the concentration. Equation 21 is a form of the Beer Lambert Law.54 From this equation the molar absorption coefficient was calculated to be about 57 ,000 M1cm1. It should be noted that this value is based on one polymer repeat unit (PRU) because that was used in the solution studies, although absorption occurs in a chromophore longer than one PRU .75 It is standard convention to use one PRU values in most calculati ons, despite the fact that a chromophore covers about 5 PRU for PPESO3 -.51 And although t his value is for the unaggregated CPE in methanol solution, it was used in the calculation of thickness for the solid state films which show aggregation because th e films were
30 produc ed using methanol solution, because studies to determine in a solution with aggregation have not yet been done, and because changes is absorption due to aggregati on have been shown to be hard to calculate due to lack of data reproducib ility .29 The background corrected (TiO2 and glass subtracted) absorption maxima of the films at the chosen bilayer numbers are shown with the calculated thickness in Table 2 -1. The absorption spectra can be found in the next chapter in Figure 31. An e xample of the calculation using the 30 bilayer film, is as follows: nm m x cm x M cm M c A l 0 41 10 10 4 10 10 4 ) 11 2 )( 57000 ( 492 08 6 1 1 (2 -2) The molecular weight w as taken as the formula weight for a repeat unit of PPE SO3 -, 474. 51 g for C22H18O8S2 (sodium was the original counter -ion, but during LbL it reenters solution when PDDA+ is adsorbed) and the generally accepted value of film densities was also used, which is 1 g/cm3.72 This density value used is a n assumption, b ut Malone and Albert have shown that a wide range of polymers all yield this value within 25% error.72 Because the polymers used here are CPEs with charges, this assumption may contain an even larger error, but any more exact value has yet to be found, so 1 g/cm3 was used in the calculations. Thus the polymer repeat unit conc entration value was determined to be: M g mol dm cm cm g 11 2 ) 51 474 ( 1 1 10 1 13 3 3 3 (2 -3) The absorption values found in Table 21 really only describe how much PPE SO3 the incident beam (during steady -state absorption studies) encounters and makes no distinction about w hether or not there are large distances between the CPEs. I t is
31 possible that the beam passes through multiple TiO2 particles which are each coated with a few nanometers of thickness so that the sum of the thicknesses equals the final thickness; however, it is un likely because it assumes that there are empty spaces underneath nanoparticles in the TiO2 scaffold. Once the TiO2 was heated, it became the more interconnected anatase form, which inhibits penetration into the TiO2, but creates a large rough sur face area to promote adsorption. It is therefore assumed that the entire thickness of the film is connected (not found as separate, disconnected vertical layers) on the outermost side of the scaffold. It is further assumed that the flexible PDDA+, because it is a less bulky polymer than PPE -SO3 -, will not separate the PPE -SO3 layers by more than one or two nanometer thickness; thus in Table 2-1, the thicknesses shown may only be that of a single layer (of PPE -SO3 -), not an entire bilayer. The true bilayer thickness is unknown because PDDA+ did not absorb, but it can be estimated to be double that of a single layer In Table 21, the increase in thickness is relatively lin ear, with an increase of about 7 nm per 5 layer s or about 1.3 nm per layer This in dicates low interlayer diffusion during the coating process. Table 2 1. Calculation of film layer thickness Number of B i layers Max Absorption (OD) Thickness (nm) Thickness per layer (nm) 30 0.492 41.0 1.37 25 0.376 31.3 1.25 20 0.308 25.6 1.28 15 0.23 6 19.6 1.31 10 0.165 13.7 1.37 5 0.081 6.7 1.34
32 CHAPTER 3 LASER SET -UP AND SYSTEM MODIFI CATIONS T he experiments in this work only look ed at the FTO/titania/CPE interface, which could potentially act as the photoanode in a solar cell as in Figure 12 The dynamics of the photoinduced sensitizer excitation of an electron from the HOMO to LUMO of the conjugated polyelectrolyte and subsequent electron injection into the Fermi level of the TiO2 acceptor w ere evaluated. Although not part of the investigation in this proposal, a fully functional solar cell would need to have an electrolyte redox couple such as an I3 -/ 3Ior a hole transporting layer to effectively produce a circuit.11 It is possible that future tests on the complete cells using solar si mulators can be done, possibly through a collaboration. Steady -State Experiments Steady state absorption spectra were taken on a UV -Vis Varian -Cary 100 spectromete r. The wavelength range w as detected from 300 nm to 800 nm by 1 nm intervals to cover the vi sible light spectrum. Fluorescence emission spectra w ere taken using a Jobin Yvon Spex -Fluorolog-3 in the range from 430 nm to 800 nm.50 All experiments were done at room temperature, and an average of four spots on the films was used as the final spectr al value.51 Figure 31 shows steady state data for PDDA+/PPE -SO3 films with the glass/TiO2 background subtracted out The absorption peak at 420 nm will be the wavelength of pump beam excitation in the TA setup. The emission spectra were excited with 420 nm. It should be noted that the 525 nm fluorescence peak is due to emission of the aggregate form, which is more prevalent at a higher number of bilayers.51 The unaggregated peak, at 4 55 nm, is seen as a shoulder. Absorption rises
33 linearly, while fluo rescence rises monotonically but not linearly with increasing bilayers. This can be seen in the inset graph of Figure 31 which includes two points per number of bilayers (absorption: 420 nm; fluorescence: 515 nm) because there were two films created for each number of bilayers 400 500 600 700 800 0.0 0.1 0.2 0.3 0.4 0.5 10 15 20 25 5 10 15 20 25 30 Normalized Abs/FluorNumber of BL 420nm 515nm0 100000 200000 300000 400000 500000 600000 700000 Fluorescence (counts)30 BLAbsorption (OD units)Wavelength (nm)Abs. Peak: 420 nm Fluor. Peak: 520 nm 5 BL Figure 31. Absorption and fluor escence emission spectra of PDDA/PPE -SO3 f ilms on TiO2-coated FTO -glass for 5, 10, 15, 20, 25, and 30 bilayer films Each line is an average of four scans at different p oints in the film. Excitation for the fluorescence was 420 nm. TiO2/FTO ba ckgrounds were subtracted for the absorption. Inset shows absorption/fluorescence rises including points from an additional set of films Decay of Films over Time A transient abs orption study of these films would not be possible if they naturally degraded over time. Therefore, the steady state absorption and fluorescence spectra
34 were monitored every week. The first weeks results are found in Figures 3-2 and 33. It should be n oted that during the synthesis process, two films were made for each 400 500 600 700 800 0.0 0.3 0.6 0.9 1.2 1.5 Absorption Spectra on first dayAbsorption (unitless)Wavelength (nm) 30BL Film#1 30BL Film#2 20BL Film#1 20BL Film#2 10BL Film#1 10BL Film#2 5BL Film#1 5BL Film#2 420 nm 400 500 600 700 800 0.0 0.3 0.6 0.9 1.2 1.5 Absorption Spectra after 1 weekAbsorption (unitless)Wavelength (nm) 30BL Film#2 20BL Film#2 10BL Film#1 5BL Film#2 420 nm Figure 32 Absorption spectra of FTO -TiO2 PDDA/PPESO3 films before (left) and after (right) one week. No background subtraction was done here only to show effect of glass/titania Each curve is the average of four scans. The right graph just shows the best films out of the two that were made at each bilayer number. Curves shown are for 5, 10, 20, and 30 bilayers 400 500 600 700 800 0 100000 200000 300000 400000 500000 600000 700000 800000 Fluorescence Spectra on first dayFluorescence (counts)Wavelength (nm) 30BL Film#1 30BL Film#2 20BL Film#1 20BL Film#2 10BL Film#1 10BL Film#2 5BL Film#1 5BL Film#2 400 500 600 700 800 0 100000 200000 300000 400000 500000 600000 700000 800000 Fluorescence Spectra after 1 weekFluorescence (counts)Wavelength (nm) 30BL Film#2 20BL Film#2 10BL Film#1 5BL Film#2 Figure 33 Fluorescence spectra of FTO -TiO2 PDDA/PPE-SO3 films initially (left) and one week after (right) fabrication. No background subtraction was needed because TiO2 fluoresces relatively weakly ex citation = 420 nm. The r ight graph shows just the best films, while the left graph shows some unaggregated signal around 455 nm. Each curve is the average o f front -face collected data at four different locations on the film. Curves shown are for 5, 10, 20, and 30 bilayers.
35 bilayer number. It was then decided based on the spectral shape and amplitude which of the two was better, and that one was used in experimentations; hence, why the graphs on the right for after one week only show one film p er bilayer number the graphs only contain data for the better film s Again, the fluorescence (excited with 420 nm) is dominated by the 525 nm signal from the aggregated PPE -SO3 -. The unaltered absorption spectra (Figure 3-2, without the FTO -TiO2 background subtraction), is shown for reference purposes to demonstrate how much the glass and titania themselves absorb. One week after fabrication, the spectra barely change. In Figure 33, each films fluorescence spectra decayed by roughly one third of the original just af ter the first week. It is believed that this is due to degradation of the titania support structure from water in the atmosphere.65 66 This effect was limited as much as possible by storing the films in a vacuum desiccator at all times when not in use T he TA experiments were also r un as soon as possible after film fabrication. Transient Absorption Introduction Transient absorption spectroscopy experiment s also known as pump-probe spectroscopy rely on two ultrashort pulses, a pump and a probe. The pump is more intense, and perturbs the sample at some time, called time zero, while a whitelight probe beam monitors the excited state dynamics through changes in its absorption spectrum. The relative time between the two pulses is varied u sing a n opt ical delay stage to see the time resolved characteristics in the spectrum.50 After the initial excitation, parts of the whitelight probe spectrum will either show an increase or decrease in absorption. As shown in Figure 34, increase in absorption is due to photo induced absorption into a higher excited state, while a decrease in
36 absorption may be due to either ground state bleaching (fewer molecules in S0 to absorb) or because of stimulated emission at the wavelengths of fluorescence. Unlike fluorescenc e which is a spontaneous process, stimulated emission depends on the fluence of the excitation beam. Figure 34. Possible Transient Absorption detectable signals. The transient absorption technique has been used before on all sorts of materials including work on C60/GaAs bilayers24,40,41 and nanosecond studies on polyelectolytes,51 but an ultrafast TA experiment that explores the dynamics of the chosen polyelectrolyte film on a titania scaffold has n ot yet been done. Determin ing how far and how fast the photo-induced excitons can travel will also demonstrate the quality of the films, because it is expected that if aggregation is more prevalent, then the exciton lifetime, and thereby anode efficiency, will be reduced somewhat To create a higher irradiance to increase the number of excitons created the pump beam was focused at the sample position using an off -axis parabolic mirror (focal length = 15. 2 4 cm) onto the surface of the films to be tested. The exciton dynamics can be studied through analysis of time -resolved data from excitonexciton annihilation, while monitoring the spectral change in absorption as a function of time. S1 S0 t Excitation Stimulated Emission S n h 1 h 2 h 1 Ground State Bleach h 3 Photoinduced Absorption A > 0
37 For the TA experiment, wavelength intensity data are collected after a grating monochromator by a C CD detector in the form of relative change in transmission, or (s ) 3 -1.50 The intensity of the probe beam (I) is detected with and without the pump beam on (turned on/off with a chopper) T he reference beam (I0), an identical beam to the probe which does not go through the volume excited by the pump, is used to limit shot -to -shot power fluctuation noise. The signal is given by E quation 3 -2.50 0 0() ()loglog() () () () () I AT I I T I (3 -1 ) ,, 00 01tpumptnopump pumpnopump tpump tnopump nopump tnopumpII TT I II T I TT I I (3 -2 ) 1 log T T A (3 -3 ) Finally, the change in absorption (in m OD) is calculated by Equation 3-3. Femtosecond Laser Source Ultrafast ( ~ 55 fs) laser light pulses for use in the transient absorption experiments were created using a commercially available laser system made by Spectra Physics A Millennia, a continuous wave neodymium:yttrium vanadate (Nd:YVO4) solid state laser, provides 532 nm output with powers of several Watts. The Millennia pumps a mode locke d titanium -doped s apphire osc illator called a Tsunami which produces pulses centered at 790 nm that are 35 nm wide (FWHM) and ~35 fs in time, have a 82
38 MHz repetition rate, and which are used to seed a regenerative amplifier called a Spitfire. 77 The S pitfire is a titanium doped sapphire regeneration a mplifier pumped by an Evolution X, a Q -s witched neodymium:yttrium lithium fluoride ( Nd:YLF ) diodepumped laser.50 The Evolution X provides 6 W of power at 527 nm and a repetition rate of 1 kHz. Inside the Spitfire, the pulses undergo stretching, amplification, and then compression, before they are released with an energy of 0.840 mJ per pulse centered at 795 nm with a pulse width of 55 fs and repetition rate of 1 kHz. This energy is split in half using a beam splitter for use in two independent OPA systems. The 0.420 mJ energy after the first beam splitter passes another beam splitter which takes 4% to be used for TA whitelight generation. T he 96% of the 0.420 mJ has a Gaussian beam spectral output centered at 795 nm and m ust be changed to a wavelength at which the sample absorbs so an optical parametric ampl ifier (OPA) wa s used for wavelength control This OPA provides near transform limited and high energy output pulses.50 Unlike a laser, an OPA s gain comes from nonlinear processes such second harmonic generation (SHG), fourth harmonic generation (FHG), sum frequency mixing (SFM), and difference frequency mixing (DFM).77 In these processes, a whit elight continuum seeds a barium borate (BBO) cry stal as the nonlinear medium. Using these processes and a proper selection of the signal or idler beam, phase matching, polarization, and number of harmonic crystals, any desired wavelength within a range of just under 300 nm to just above 10 can be generated.50 The excitation wavelength chosen here (420 nm) is based on the most optimal absorption of the molecules as seen in th eir absorption spectra (Figure 31 ), and is created by using two BBO crystals to generate the FHG of the idler beam An adjust able waveplate was
39 fixed inside the OPA before the SHG crystal to allow for control of the output energy for later use in energy dependent studies. Figure 35. Transient Absorption set -up. The sample and WLG (CaF2 window) are on rotating stages. An 800 nm wavelength, 1 kHz repetition rate, 0.42 mJ beam splits to generate WL and seed the OPA. Prior to reaching the WLG, the beam is attenuated using a waveplate and polarizer. A 420 nm pump beam is made through second and fourth harmonic generation of the OPA idler beam. Pump, probe, and reference are focused on the near side of the film using an off axis parabolic mirror, then photons are collected using a spectrograph and CCD (charge -coupled device) camera. PumpProbe Set up The 4% of the 0.420 mJ energy (that was separated before the OPA ) passes through a waveplate (to lower the energy to about 4 J) and polarizer (set to the magic angle, 54.7 ) and is then focused (using a 100 mm focal length lens) onto the far side of a 1 inch diameter, 1.5 mm thick rotating CaF2 window to generate a supercontinuum (white light spectrum) Rotating the window removes shot -to -shot fluctuations, while f ocusing on the far side of the window avoids unwanted dispersion effects After OPA Delay stage Parabolic Mirror WLG Rotating Film Sample Probe beam Reference beam Pump beam Incoming Beam Chopper Beam Stopper CCD Grating monochromator
40 collimation by an off axis parabolic mirror, t his whitelight beam is then split 45:5 5 into a probe and reference beam. T he reference passes through two extra mirrors so because of losses from less than 100% reflectivity in the mirrors, it needs about 55% of the wh itelight from the beam splitter to have the same amount of whitelight counts as the probe when it reaches the sample. The probe and reference run parallel a long a vib ration control led table until reaching another off axis parabolic mirror (f = 15.24 cm) t o focus them at sample position. They are not overlapped, but they are as close as possible to each other to allow for the best results. The purpose of the probe beam is to monitor the perturbation in the sample created by the pump beam, while the reference beam passes through an unperturbed part of the sample to act as a reference in the calculations. After the sample, lenses are used to collimate and then refocus the beams on the entrance slit of a spectrograph. T he 420 nm pump beam out of the OPA is m easured to travel the same distance to the sample as that of the probe and reference beams so that the beams can be temporally overlapped. The mirrors it encounters are chosen to provide the best reflection at 420 nm, while the mirrors for the whitelight must be able to reflect a broader spectrum of light. The pump first passes through a telescope made up of two quartz lenses (concave: f = -50 nm; convex: f = 150 nm) in order to decrease the beam size and to collimate it. The probe beam then passes through a n optical delay stage which contains two perpendicular mounted mirrors on a computer -controlled horizontal translation stage.77 This enables one to change the time delay between the pump and probe up to 800 ps during experimentation. After the delay stage, the pump beam passes a mechanical chopper wheel with repetition rate of 8.4 Hz which allows for data
41 collection with the pump on (not blocked) and off (blocked). Then the pump beam hits the same parabolic mirror (f = 15.24 cm) as the probe and reference, and is focused so that it completely overlaps the probe beam. The diameter of the pump was measured using a razor -blade edge technique to be 276 microns.77 The film sample is placed film side facing towards the incoming beam a t the focus of the p arabolic mirror Any group velocity dispersion from the glass is corrected for during data analysis using a homemade Labview program. After alignment, the pump and probe beams are now overlapped on the part of the film to be tested with the reference b eam right below (Figure 3 -5 ). It is essential that the probe beam is smaller than the pump beam (d = 276 m) and that their overlap is optimized to make sure every molecule in the probed volume is excited. Partial overlap will give a low signal to noise ratio from a much weaker change in the collected signals Detection System After the sample, the pump beam is blocked with a beam stopper, and the transmission through the sample of both the pr obe and reference beams are collimated by a lens, then focused by another lens onto a 150 m entrance slit of a n Andor iStar Shamrock spectrograph. The beams arrive at two different heights. After a n internal grating with line density of 300 lines/mm disperses the wavelengths, t hey are then independently detecte d o n different rows of pixels by a Charge -Coupled Device (CCD) camera and read by Andor iStar and LabView programs. The CCD is a silicon based semiconductor chip bearing 256 rows and 1024 columns of photo -sensors (pixels) The whitelight beams cover about 13 rows each (which are integrated for data collection), and are separated by at least 50 rows between them. The grating angle in the monochromator of the Shamrock can be changed to collect the desired wavelengths
42 of light and is chosen here to show the spectral coverage from 350.49 to 631.21 nm, because the range of counts generated from the CaF2 window only show measurable counts between ~350 nm and ~600 nm, with the highest counts around 450 nm Averaging was implemented to improve the signal to noise ratio. For each experiment, unless otherwise noted, an average of 80 points at each time-step was taken Then, at least 10 experiments were run for each study, and the best 10 experiments were taken and averaged. About 50 time steps were used to show t he entire range from 50 ps to 800 ps (after ~200 ps there was little signal) with time steps much closer to each other at times just following excitation at time zero If significant degradation of the magnitude of the data was seen from experiment to experiment, the film was moved horizontally perpendicular to the beam in order to hit a fresh spot. Modifications The film sample is secured on a homemade three-dimensional adjustable translation stage for proper alignment. A circularly rotating stag e atta ched to this translation stage is positioned so that film sample s can be taped to it Rotation of the films benefits the data collection in three ways: 1) it prevents the film from burning because the slide will be moving and the f ocus will hit different parts 2) it provide s an average for the entire film, which is useful because often the film surfaces may not be the exact same thick ness throughout, and 3) it help s assure that each shot is exciting a new part of the film which starts in the ground state. T he range of wavelengths collected by the CCD after the monochromators grating dispersion are from ~350 nm to ~630 nm, which cover s the bleach signal at 420 nm and possible stimulated emission peak found at 525 nm. Furthermore, emission from the TiO2 q uencher may also be seen from about 500 nm up to the longer
43 wavelength edge of the detectable spectrum ;42 therefore, these wavelengths are also monitored for any stimulated emission which may result. Lastly, numerical corrections to the TA data, including that f or chirp, ( which is a consequence of the temporal distribution of the different wavelengths of the whitelight, which causes the signal to appear earlier at shor ter wavelengths than longer ones due to the beams passing through materials other than vacuum ,) are implement ed using a homemade LabView program .50,2 8 An instrument response function (IRF) was found in previous solution studies,77 but could not be done for films because there is no solvent to test for coherent artifacts. Based on that prev ious work, it is assumed that the IRF is somewhere between 100 and 250 fs. Finally, a calculation of Equation 33 to convert change in transmission to change in absorption was done to all the TA data. Fluences are defined as energy per unit area. For the pump diameter of 276 microns and energies measured using an Ophir power meter (at 420 nm) of 50 nJ, 30 nJ, 15 nJ, and 7 nJ at the sample position, one calculates fluences of 84 J cm2, 50 J cm2, 25 J cm2, and 12 J cm2, respectively. These flu ences may be lower than expected because of reflection losses in some of the optical components, and because on certain days full power from the Spitfire could not be obtained. Because of the 1 kHz repetition rate, the irradiances, in units of power/area, would simply be 84 m W cm2, 50 mW cm2, 25 mW cm2, and 12 mW cm2, respectively. Atomic Force Microscopy Measurements done using Atomic Force Microscopy (AFM) were employed in order to elucidate the film surface structure. It is important in thes e studies that a uniformly rough surface is obtained, otherwise the TA results would not be reproducible. During the TA data collection, the rotating sample stage was added in order to average
44 the film absorption, so if the film only contains minor impuri ties or changes in roughness those would not greatly affect data results. Figure 36. AFM image for A) 5, B) 10, C) 20, and D) 30 Bilayers. Darker areas are sha llow regions, while the bright pink areas are the tops of TiO2 nanoparticle clusters. It is not possible to clearly distinguish the polymer layers on the films. Figure 36 shows AFM images of four films, one of each of 5, 10, 20, and 30 bilayers of PDDA/PPE-SO3. They were taken after a week of the initial fabrication. The 5 bilayer film appear s stretched from side to side, which is not necessarily bad for the
45 TA studies, but is likely due t o the doctor blading process which drags titania material across the slide In general, the size of the clusters and film roughness stay the same, or poss ibly a very slight trend can be seen going from fewer to greater number of bilayers, in which more bilayers equals larger clusters. In the bottom of picture D, the top half of a large cluster can be seen as a bright pink spot thus showing that the larg est clusters are found with the highest number (30) of bilayers It is b elieved that overall, the titanium dioxide nanoparticles make up the majority of t he film thickness and structure, but t hrough aggregation and a simple accumulation of material, the additional bilayers appear as a small, not quite conclusively visible increase in the thickness throughout the film structure.65 This is because the CPE bilayers in dividually account for only tens of nanometers on the overall micron thickness made from the 20 nm TiO2 nanoparticles, Figure 37. AFM stepheight measurement on 10 bilayer film, taken at three different spots on the film, indicated by the three lines. All lines give ~1.60 m as the height between the glass and the surface.
46 A scratch method using the side of a razor blade to remove the film down to the glass was employed, and then the AFM was run in stepheight calculation mode. It required a couple trials to find a good spot, which was assisted by blowing off of residue from the scratch. After some simple analysis using the provided AFM software, Figure 37 was obtained. In Figure 37, the dark area represents the scratched away part, showing the glass surface, while the orange part is the remaining film. From a calculation of the di fference in the z value (height), a film thickness of a little more than 1.60 microns was found. This matches well with the literature, where titania coatings were made using a few microns as the thickness.65, 66, 68
47 CHAPTER 4 TRANSIENT ABSORPTION R ESULTS AND ANALYSIS This chapter outlines the results of the transient absorption experiments performed on the PDDA+/PPE -SO3 bilayer titania films. Transient Absorption Results for 30 Bilayer Film In the first experiment, the 30 bilayer film was excited with 50 nJ of energy measured at 420 nm at the sample position. This was the maximum energy that could be attained, so all other scans which required a change in energy were done by reducing the energy using a waveplate placed inside the OPA. A complete threedi mensional representation using Matlab software on the data is found in Figure 4-1, followed by two two dimensional cross -sectional plots in Figures 4 -2 and 4 -3, made by select ing either time s or wavelengths The TA data were collected from a range of -20 ps to 200 ps, and a wavelength r ange based on the probes white light spectrum which gave usable counts between 350 and 600 nm. Th e first thing to note about these data plots is that they show a narrow peak at ~420 nm which is simply pump scatter In Figure 41, this was removed by setting absorption values arbitrarily to zero, while in Figure 42, the full magnitude of this scatter is seen. Despite setting up numerous beam blockers, this could not be fully eliminated. Fortunately, it does not hi nder the analysis of wavelengths outside of the narrow range of about 414 to 422 nm Bleach Ignoring those scatter points, one sees a decrease in absorption from about 380 to 460 nm due to the bleach signal. This negative signal, centered at the 420 nm pe ak of the ground state absorption, corresponds to the population of chromophores being
48 unable to continue to absorb at those wavelengths because many are in the first excited singlet state, and if there is a state with energy difference E = h above the singlet to which the population can further absorb, the signal is not strong enough to detect it Figure 41. Complete TA data for 30 BL film excited with 50 nJ pump energy. The positive absorption (in red) can be seen to rise sharply at 0 ps, and then dec rease in intensity out towards 200 ps The negative absorption signal (in blue) is also strongest right after 0 ps, and then becomes less negative as it approaches 200 ps. The pump scatter was removed around 420 nm wavelengths, and i s seen as an arbitrary green band. The signals at wavelengths lower than 400 nm is noisy and is not assumed real. Upon closer inspection of Figure 4-2 the bleach signal can be seen not to match the steady state absorption because of the narrowing on the shorter wavelength side (380 to 410 nm) and possible narrowing on longer wavelength side around 450 to 460 nm Both may be d ue to a PIA band of the CPE being superimposed with the bleach
49 signal, thus giving a summation of two opposite magnitude absorptio n signals in the spectrum. The shorter wavelength narrowing may also include contribution due to absorption from the TiO2 or glass. Thus, the bleach signal contribution, because it is not a similar shape to the linear absorption band in Figure 31 (included normalized to 57 mOD in Figure 4-2) may also contain anot her process below 410 nm and above 450 nm, causing the total TA signal to appear less negative To separate these signals, further computational methods, outside the scope of this paper, would be required. 400 500 600 -80 -60 -40 -20 0 20 40 60 30 Bilayer Film at 50 nJ 0ps 0.9ps 0.95ps 1.2ps 2ps 4ps 20ps 100ps 200ps Abs Norm to 57mOD Absorption (mOD)Wavelength (nm) Figure 42. 30 BL film excited with 50 nJ pump energy plotted selecting time crosssections to show the dynamics Pump scatter is seen around 420 nm. Ignoring this, a broad negative absorption bleach band is seen between 380 and 460 nm, while a PIA positive absorption signal, with possible stimulated emission contribution near 525 nm, is seen between 460 and 600 nm. The steady state absorption normalized to 57 mOD is shown for comparison.
50 0 5 10 15 20 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 Bilayer Film at 50 nJ Absorption (mOD)Time (ps) 400nm 425nm 450nm 475nm 525nm 575nm Figure 43. 30 bilayer film excited with 50 nJ pump energy, plotted selecting wavelength cross-sections to show the decays. At 400, 425, and 450 nm, the bleach signal decays from negative absorption, while at the 475, 525, and 575 nm, the PIA signal decays with a weaker and positive signal. Photo -induced Absorption (PIA) At 475 nm, after initial excitation (0.9 ps on the plot) the positive absorption signal decay s in a few picoseconds in Figure 4 2 To determine what causes this PIA it is necessary to list the possibilities. It could be due to: 1) absorption from the titania layer after electron injection occurs67 (before electron injection occurs the TiO2 absorbs in the near -UV re gion ), 2) absorption from the PPE -SO3 + after electron injection occur s, or 3) absorption of light by the singlet state of PPESO3 into a higher energy excited state. The first thing to note is that the titania layer is ruled out from causing the strong 475 nm signal because literature value show TiO2 absorption to be broad and centered around
51 620 nm.65 67 Second, molecules similar to the cationic PPE -SO3 + have been studied, and they also absorb above 600 nm,18 out of the range of this test so that can not be the source of PIA Therefore it can be concluded that the obs erved photo-induced absorption must be due to PPESO3 excited state absorption. The broadness of this absorption is convoluted with the bleach on the shorter wavelengths around 460 nm and perhaps some stimulated emission centered at 525 nm which essential ly subtracts from the absorption signal seen. Stimulated Emission (SE) Although the signal in Figure 42 shows positive absorption (PIA) at wavelengths above 460 nm after pump excitation, the signal is less positive around 525 nm. Because this corresponds to the peak in the fluorescence signal of PPE -SO3, stimulated emission from PPE SO3 is the most likely cause. The broad shape of the dip also points to a subtraction in detected absorption from the broad aggregated emission from PPE-SO3. An additional c ontribution may be due to the titania stimulated emission ( after electron acceptance ) To determine which contributes to the signal, the shapes need to be compared to that of the TA data. This is done i f Figure 4 4, where Figure 4 -2 is adapted to show the 30 bilayers TA data with the fluorescence of TiO2 and PPE SO3 overlapped and normalized to 32 mOD Subsequently the first interpretation of the signal at 525 nm is as a combination of positive absorption PIA of the singlet state PPE -SO3, plus a negati ve signal due to stimulated emission from the excited state of PPE-SO3. An alternate interpretation of the signal above 460 nm is that of two PIAs with no stimulated emission necessary. The first PIA (475 nm) would be due to the excited state absorption and the second (575 nm) would be due to the negatively charged titania. As mentioned earlier, TiO2
52 absorbs in a broad, featureless shape between 400 to 800 nm .67 Thus, it could be that that absorption is partially hidden in the spectra, and only detecte d alone (not as a combination of signals) at wavelengths greater than 525 nm. A different interpretation still is that of PIA of PPE SO3 with TiO2 SE, but based on the lack of shape similarity near 520 nm in Figure 4 -4 this is least likely. Future tests are necessary to visualize the PIA above 600 nm to decide for certain what the cause is. Probing higher wavelengths would additionally be beneficial in elucidating the dynamics of PPE -SO3 +. 400 500 600 -80 -60 -40 -20 0 20 40 60 30 Bilayer Film at 50 nJ Absorption (mOD)Wavelength (nm) 0ps 0.9ps 2ps 20ps 200ps TiO2 PPE-SO3 Figure 44 30 BL film excited with 50 nJ pump energy, plotted selecting time cross sections to show the dynamics. In this plot, the curves for titania and PPE S03 fluorescence are added to try to account for stimulated emission. Both are normalized to 32 mOD. They are close in shape, but SE from PPE-SO3 is believed to be the main cause.
53 Assigning Bands The initial absorption of 425 nm light by PPE -SO3 from the ground state to the fi rst electronic excited state c an be represented by a Perrin-Ja blonski diagram, as shown in Figure 4 5 Here PIA excite s the first excited singlet state of PPE -SO3 t o a higher energy excited state through the absorption of 475 nm light. Within 1012 s to 1010 s after both of these absorptions, vibrational relaxation to the lowest vibrational level of the elec tronic excited state occurs,50 and in the case of the first singlet, stimulated emission from it (or one of the aggregated states to which the energy moved) to one of the ground states vibrational levels can be detected at 525 nm. From the TA data, it ca n not be concluded definitely if other states, whether a triplet of PPE -SO3 or simply titania, are present, but SE from the titania is not completely ruled out because higher wavelength dynamics are not yet known. Triplets however are much less probable b ecause their characteristic longer lifetimes were not seen. Figure 45 Spectral band assignment from TA data using a Perrin-Jablonski diagram The bleach occurs at 420 nm, the photo induced absorption at 475 nm, and the stimu lated emission at 525 nm. It is not clear from data collected wh ich other states and processes are present. PIA SE S1 S0 Bleach Sn 475 nm 425 nm 525 nm Excited Triplet or TiO2 State present? Possible ISC or e injection Aggregated State ET
54 ExcitonExciton Annihilation Exciton motion and diffusion determination can give important information about the nanoscale structure of polymer fil ms. However, no universal method has been determined to calculate fully the diffusion process, although many suggestions, like by adding quenching centers, testing with a metal present and depolarization experiments have been made.61 65 One method to determine exciton diffusion is through studies of the excitation energy dependence on exciton exciton annihilation rates. Exciton exciton annihilation (EEA) occurs when two excitons diffuse to within close proximity of each other, known as an annihilation radius.80 Once there, the two excitons can interact via dipoledipole processes causing the energy of one exciton to be transferred to the other.81 The exciton in the higher excited state quickly relaxes non -radiatively back to the lowest excited state. The net result of this bimolecular process is the loss of one exciton to heat.70,80 81 Because there are fewer excitons after EEA because some returned to the ground state, it is expected that the greater the extent of EEA, the shorter the lifetime of ti me resolved signals (such as bleach) will be. Bleach signals are seen in TA experiments because of a reduction of the ground state population, so as EEA occurs, the ground state population will return more quickly, and a faster decay of the bleach signal should be seen By solving exciton rate equations and using TA data, Gulbinas et al. have been able to use time-resolved signals measured for different fluences to calculate exciton diffusion values.62 These equations and PPE-SO3 film TA data ar e present ed in the next few section s to calculate annihilation rate constants and determine preliminary values for exciton diffusion in a similar manner
55 It is important to know about the diffusion of excitons generated by photons in thin films because it is this process which contributes to photocurrent, should the films used in photovoltaic devices by allowing excitons generated by photoexcitation in the bulk of the polymer to diffuse to the polymer acceptor interface where electronhole separation occurs. The higher the average distance traveled during their "random walk,"29 the higher the number of neutral excitons which will reach the bulk interface with titania and be able to be cleaved in to unbound electrons and holes. Exciton Density Calculation In order f or EEA to occur, exciton densities of a sufficiently high value must be achieved. In the literature, densities of at least 1017 cm3 are usually needed.80 81 Therefore, in order to see if EEA is taking place in these PPE SO3 films, the exciton density needs to be calculated. First, using the typical density of polymer films72 of 1 g cm3 3 20 23 310 54 2 1 10 022 6 5 1 51 474 1 1 1 cm x mole x PRU e chromophor g molPRU cm g (4 -1 ) The exciton density can be calculated based on the number of photons absorbed. One can use = c/ and E = h to determine that one photon of 420 nm wavelength has an energy of about 4.73x1019 J. In 50 nJ, the highest excitation energy used, there are then 1.06x1011 photons. Furthermore, using A = log(I0/I) where A i s absorbance and I0 and I are intensity entering and leaving the excitation volume, one can determine that 67.8% of photons are absorbed based on A = 0.492 (background subtracted) at 420 nm. 67.8% of 1.06x1011 photons equals 7.19x1010 photons. This i s how many photons were absorbed per cm3 for 50 nJ energy excitation. From the AFM step -height measurement,
56 the thickness of the films, including titania, is about 1.60 microns; however, of that volume, only about a 41 nm thickness contains PPE -SO3 interw oven amid the titania scaffold, so that can be assumed to be the absorption layer thickness (see Table 2-1). From razor blade experiments in the lab which determined the focused pump beam diameter to be 276 microns, one can approximate the active cylindrical volume as: 3 9 3 3 2 2 210 45 2 10 45 2 41 ) 138 ( cm x m x nm m t r V (4 -2 ) The final step to calculate the exciton density is to take the number of photons absorbed by PPE SO3 (each of which became an exciton) and divide by the active volume, which for 50 nJ excitation gives 2.93 x1 019 cm3. For 30 nJ and 15 nJ excitation, t he exciton density values are 1 76 x1019 cm3 and 8.78 x1018 cm3, respectively. All of these numbers are above the minimum 1017 cm3 density v alue expected for EEA to occur; thus, it is concluded that EEA occurs under the experimental conditions presented here. -2 0 2 4 6 8 10 12 14 16 18 20 0.00 0.05 0.10 0.15 30 Bilayers Film Bleach Transmission (unitless)Time (ps) 50nJ 30nJ 15nJ 7nJ -2 0 2 4 6 8 10 12 14 16 18 20 0.0 0.2 0.4 0.6 0.8 1.0 30 Bilayers Film Bleach, Normalized Transmission (unitless)Time (ps) 50nJ 30nJ 15nJ 7nJ Figure 46. 30 bilayer film bleach signal ( ex citation detection = 427 nm) from four different excitation energies, 7, 15, 30, and 50 nJ. Each curve is the average of 10 experiments, 80 points averaged per time-step, per experiment. These data are raw and uncorrected to show the EEA in the first few picoseconds. The graph on the right is normalized data, showing 50 nJ decays fastest.
57 To further show that EEA is happening, Figure 4 6 is presented to show energy dependence on the EEA of the bleach signal for the 30 bilayer film. The data are not corrected for chirp or converted to change in absorption yet because the purpose here is just to show the energy dependence. The presence of a fast decay component appears upon increased excitation, as seen in the sharper decay for the 50 nJ ener gy excitation in the first few picoseconds after excitation. The lower energy excitations show a slower decay indicating weaker EEA ; this is also shown in the graph on the right of Figure 46, which simply normalized the data on the left Gulbinas et al. showed that EEA in films occur at fluences orders of magnitude lower than for solutions.62 Their polymer films, methyl -substituted polyparaphenylene (m -LPPP), begi n to show EEA at fluences of 30 J cm2. The fluences, calculated from the excitation energies and spot size radius of 138 m, used in the PPE -SO3 films are: 84 J cm2 (50 nJ), 50 J cm2 (30 nJ), 25 J cm2 (15 nJ), and 12 J cm2 (7 nJ). So besides the visual appearance of EEA at the first few picoseconds in Figure 46, the fluenc es calculat ed show that at least for the two highest excitations, there should be EEA. Unfortunately, higher fluences could not be obtained during these tests, but hopefully future tests can be done using higher values to verify the results found here. Kinetics The 30 BL films ( and 20 BL films not shown) were excited at different energie s to determine the extent of exciton diffusion through annihilation studies To understand how this is done, it is first necessary to examine the kinetic equations describing the pa thways of e xciton relaxation. Equation 4-3 shows one of the possible deactivations while Equation 4 -4 models the annihilation process. There are numerous pathways for the exciton to decay, including PIA, stimulated and spontaneous emission, internal
58 con version, intersystem crossing, and quenching by the titania layer. All of these deactivation pathways are accumulated in a linear rate constant, k, which separate them from the annihilation rate constant .56,80 ] 3 [ *] 3 [1 1PPESO PPESO (4 -3 ) heat PPESO PPESO PPESO PPESO PPESO PPESO ] 3 [ *] 3 [ ] 3 [ *] 3 [ *] 3 [ *] 3 [1 1 1 1 1 1 (4 -4 ) Exciton exciton annihilation can be expressed as Equation 45 :62,80 2*] [ *] [ *] [ n n k dt n d (4 -5 ) where [ n* ] is the exciton density. The annihilation term has an exciton density squared dependence because EEA obvious ly requires two excitons to interact. An initial pumping term is not included in Equation 4 5, but it is known that pumping to create excitons occurs at the initial time. T he goal is t o isolate and solve for so first both sides of Equation 4 -5 are divided by [ n* ] Then the exciton relaxation rate, K, is defined as in Equation 4 6: dt n d n K *] [ *] [ 1 (4 -6 ) which leads to an annihilation rate constant as in Equation 4 -7 : *] [ ) ( n k K (4 -7 ) Although [n*], the exciton density, has time dependence, the rate constants k and are assumed to be time independent. When necessary, the exci ton density is thus taken as the value at the initial t ime and for K calculations, is incorporated into the relative derivative which is obtained from data.
59 At this point, K (from smooth ing the bleach signal decay and differentiating) and [ n* ] can be obtained from the data and measurements but k is still unk nown, unless one determines additional parameters by comparing two excitation energies thus isolating the k term It was seen in Figure 4-6 that at different excitation energies the bleach decay rates are different at short timescales, so this means diff erent K values can be obtained. Putting together two equations at different energie s, noted with subscripts 1 and 2, one can get Equations 4 -8 and 4-9 :62 *] [ *] [2 1 2 1n n K K (4 -8 ) *] [ *] [ *] [ *] [2 1 1 2 2 1n n n K n K k (4 -9 ) Data Fitting to Find R ate Constants Now the TA data are analyzed for different fluences to give values for K, which will then be used along with the exciton density to give an estimate of the annihilation rate coefficient. Appropriately, Figures 4-7 4 8, and 49 show the (10 experiment averaged) data obtained during TA data collection (at bleach wavelength) for the 30 bilayer film excited with 50 30, and 15 nJ pump excitation energy respectively. The 50 and 30 nJ energies data was plotted for 425 nm detection, while the 15 nJ data was plotted using 427 nm because of a clearer signal seen at that wavelength. A three point adjacent averaging smoothing was implemented in order better evaluate the derivative of the data. This was done using the Origin software program. Res iduals for the smoothings are also plotted as insets in these figures to show that the deviation from the actual data in uniformly random, so the smoothing can be trusted.
60 1 2 3 4 5 -50 -40 -30 -20 -10 0 1 2 3 4 5 -4 -2 0 2 4 Residuals (mOD)Time (ps) Absorption (mOD)Time (ps) 50nJ bleach at 425nm 3 point smoothing50 nJ Bleach Fitting Figure 47. 30 bilayers film excited with 50 nJ energy f rom 420 nm pump beam. Detected at 425 nm. 3 point adjacent averaging smoothing is shown as the red line. Residuals between data and smoothing are shown in the inset. 1 2 3 4 5 -50 -40 -30 -20 -10 0 30 nJ Bleach Fitting1 2 3 4 5 -4 -2 0 2 4 Residuals (mOD)Time (ps) Absorption (mOD)Time (ps) 30nJ bleach at 425nm 3 point smoothing Figure 48. 30 bilayer film excited with 30 nJ energy fro m 420 nm pump beam. Detected at 425 nm. 3 point adjacent averaging smoothing is shown as the red line. Residuals between data and smoothing are shown in the inset.
61 1 2 3 4 5 -50 -40 -30 -20 -10 0 1 2 3 4 5 -4 -2 0 2 4 Residuals (mOD)Time (ps) 15 nJ Bleach Fitting Absorption (mOD)Time (ps) 15nJ bleach at 427nm 3 point smoothing Figure 49. 30 bilayer film excited with 15 nJ energy from 420 nm pump beam. Detected at 427 nm. 3 point adjacent averaging smoothing is shown as the red line. Residuals between data and smoothing are shown in the inset. Once the simple smoothings were obtained for each excitation energy, the derivatives were taken in order to give values for K, the exciton relaxation rate. The s e are plotted in Figure 410. The derivatives are highest right after the initial excitation as expected. It is clear that the relative derivatives, which are taken as the K values a ccording to Equation 46, vary over time, but the relatively large errors present a significant drawback in evaluating the numerator in Equation 48. Because the assumption is being made that the annihilation rate is time independent during the first few picoseconds (based on EEA film work by Gulbinas et al.),62 the best that can be done is to choose K values at a point in time at which the derivatives show the least amount of error. The point chosen is 0.6 ps after excitation. At this value, the derivat ives in Figure 4 10 do not show the large error seen at earlier time, show a clear difference in values as opposed to later times, and align such that the higher excitation
62 energy data derivatives yield a higher derivative and K value. The last reason mak es sense because the excitation with higher energy (50 nJ) should create more excitons, each of which should see more excitons than the system with fewer excitons created (15 nJ). The K values 0.6 ps after excitation are: K50 = 17 .821x1012 s1, K3 0 = 13.045x1012 s1, and K15 = 10.096x1012 s1. 1 2 3 4 5 0 10 20 30 40 50 60 70 Derivative Value (x10-12s-1)Time (ps) 50nJ smoothing derivative 30nJ smoothing derivative 15nJ smoothing derivativeValues of K over time 0.6 ps after excitation Figure 410. K values over time, for excitation energies of 50 nJ, 30 nJ, and 15 nJ Equation 48 can now be evaluated. The difference in K values gives the numerator, and the differenc e in exciton densities gives the denominator. Table 41 shows these values calculated for each of the three pairs of excitations. The absolute value of the annihilation rate constant is estimated from Equation 4-8 but should be noted to have high error. The values are similar to each other and the average of them is 3.73x107 cm3 s1. The linear relaxation rate f rom Equation 49 is not calculated because it is not needed in diffusion coefficient calculations in the next section, and because it is expected to be unreliable due to the noise in the TA data.
63 Table 4 1. Calculation of annihilation rate constant for 30 bilayer film. Excitation Pair K1K2 n1* n2* 50 & 30 nJ 4.78x1012s1 1.17x1019cm3 4.09 x107cm3 s1 50 & 15 nJ 7 73x1012s1 2.05x1019cm3 3.77x107cm3 s1 30 & 15 nJ 2 95x1012s1 8.82x1018cm3 3.34x107cm3 s1 Exciton Diffusion Length Calculations The 50 nJ and 15 nJ excitation energies were chosen because they were the farthest stable energie s apart which could be obtained, considering the irradiated area, both energies would result in fluence values that should still produce annihilation, and their difference should show the greatest change in data due to fluence dependence. This proved fruitful because the average of the three pairs of ex citation ended up being remarkably close to the obtained simply between the 50 nJ and 15 nJ pair. Thus it is expected in future studies that using excitation pairs with the greatest energy difference between them will yield the best values for annihilat ion. It should be remembered that t he labels 50nJ and 15nJ refer to the energy measure at 420 nm at the sample position during exper imentation using an Ophir Nova II power meter. Because the annihilation coefficient is now known (although it may cont ain significant error) the diffusion coefficient, D, can be calculated from the following equation:62,7 8 ,80 81 DR 4 (4 -10 ) Here, R is the exciton annihilation radius, the distance at which annihilation is faster than diffusio n.80 The equation assumes that the excitons have to diffuse some distance before annihilation can occur. In the literature, Valencia et al. have assumed R to be the exciton delocalization length,81 and Lewis et al. have found it using R = a(Re/Rg)3/2,
64 wh ere a is the interchain hopping distance, Re is the F rster radius for ET onto the excited segment and Rg is the F rster radius for ET onto the ground state segment.80 In general, most organic materials have F rster radii less than 5 nm,82 but because t hey are not exactly known for the films studied here, a simple approach of assuming r = 6.0 nm,23,62,75 which is the length of one chromophore unit, or 5 PRUs, should give an decent upper estimate. Equation 4 11 relates diffusion length (LD) to the known diffusion coefficient (D), based on the unquenched fluorescence lifetime ( FL = 550 ps ), taken from literature values of the PPE -SO3 in water solution.13,26,30 62,75 This solution lifetime was used because PPE -SO3 is aggregated in water, and it appears that the films made, despite being made from methanol solutions, also exhibi t strong aggregation. D d LD (4 -11 ) This model, based on random walk of the excitons, can be used for any dimensional diffusion because d is the dimensionality term. If one-dimensional diffusion is assumed, d = 1, if threedimensional diffusion is assumed, d = 6. Here, if thr eedimensional diffusion is assumed it allow s for both interchain and intrachain diffusion to take place with equal probability For the most part, interchain and intrachain diffusion take place at very dif ferent rates, and it is hard to model them together. Therefore, both dimensionalities are presented below From E quation 4 -11 values of LD are summarized in Table 42 along with the values of variables and constants calculated up to this point. Value s of LD on the order of magnitude of 1 to 100 nm are expected from repor ts of other conjugated polymers,13,26,30 ,51 so LD = 52.2 nm is a reasonable value, but LD =
65 128 nm is extremely high, and is considered inaccurate. 52.2 nm may be a little on the high s ide compared to most singlet excitons, but that is good when dealing with solar cells where longer exciton diffusion is preferred. The distance of 52.2 nm in terms of number of bilayers from calculations from Table 2-1 (1.3 nm per layer) can be approximat ed to about 20 bilayers. However, this number is extremely unreliable, and no true conclusion should be drawn from it until further studies have verified these results. However, if the data and calculations are correct, then at least qualitatively one ca n say that in 20 bilayer PDDA/PPE -SO3 films the excitons generated have a chance to diffuse through the CPE bilayers and reach the titania interface where they can cleave and undergo electron injection. Creating films with more than this number of bilayer s would not improve collection efficiency or device performance. Table 4 2 Calculation of diffusion length for 30 bilayer film K50 K30 K15 n50* n30* n15* 17.82e12 s1 13.05e12 s1 10.10e12 s1 2.93e19 cm3 1.76e19cm3 8.78e18cm3 R D FL LD (d=1) LD (d=6) 3.73e 7cm3s1 6.0 nm 4.95e 2cm2s1 550 ps 52.2 nm 128 nm Once again, the errors which manifested themselves in these calculations included but were not limited to: low fluences creating low EEA and large amounts of n oise, use of constants from solutions even though these are films, an estimation of the chromophore length to be 6.0 nm, and the use of that number as the annihilation radius. In the future, higher and more stable fluences and more accurate kinetic modeli ng should improve the calculations.
66 Effect of Changing Number of Bilayers Transient absorption experiments were run with 30, 20, 10, and 5 bilayer films to help confirm the dynamics seen on the pure 30 bilayer film. All data shown are the average of 10 sc ans, with 50 points taken at each time step. The best signal to noise r atios were obtained in the higher number of bilayer films with decreasing resolution as the number of bilayers was reduced. This is because fewer photons were collected by the fewer number of bilayers, so there was a lower signal compared to the same amount of noise. Although it does not affect the data it should be noted that time zero was shifted between previous and the following experimentations so that the peak excitation tak es place at 0.2 instead of 0.9 ps on the time axis. In Figure 411 the 30 bilayer film was excited with 12 nJ energy (the maximum attainable at the time) at 420 nm. All of the experiments in this section were run using that energy for consistency. In th e left graph of Figure 41 1 three wavelengths corresponding to bleach (400, 425, and 450 nm) and three wavelengths corresponding to PIA (475, 525, and 575 nm) were plotted. Out of the bleach signals, 450 nm gave the largest magnitude upon initial excitat ion, while 400 nm, where the white light counts were lowest, shows considerable noise. The PIA signal curves are decidedly smaller amplitude than the 50 nJ test in Figure 4-3. Again, this is attributed to an undesirable lower irradiance from the pump. On the right graph in Figure 411 the same TA data is plotted versus wavelength and specified times to show the dynamics. The bleach is still apparent, but the previously strong PIA at 475 nm is absent, while the rest of the PIA above 460 nm wavelengths app ears but weakly. It is believed that the 475 nm PIA of singlet state PPE-SO3 to a higher excited state is missing because the excitation energy is not
67 enough to generate as many excitons as the 50 nJ excitation did, so fewer excitons are around to absorb the 475 nm light. 0 5 10 15 20 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 Absorption (mOD)Time (ps) 400nm 425nm 450nm 475nm 525nm 575nm30 Bilayer Film, 12 nJ 400 500 600 -40 -30 -20 -10 0 10 20 30 Bilayer Film, 12 nJ Absorption (mOD)Wavelength (nm) -2ps 0ps 0.2ps 0.5 1ps 3ps 20ps 200ps Figure 411 30 bilayer film excited with 12 nJ The left graph plots three bleach signals (400, 425, and 450 nm) and three PIA signals (475, 525, and 575 nm) versus time The bleac h signal to noise is better than the PIA signals, and the fast component of bleach decay takes place within 5 ps. The right graph plots the wavelength spectrum at chosen times to show the dynamics. Single exponential decays were fitted to the bleach sig nal for the first 5 ps for these data as a better visual reference than smoothing techniques For 400 nm, the lifetime ( ) was 1.492 ps with an r2 = 0.511, which indicates a poor fitting. r2 values approaching unity indicate perfect fit, so this r2 value for bleach decay signifies that the lifetime is likely incorrect. This is probably due to the poor signal to noise at t he bleach wavelength of 400 nm. To improve this, lifetimes were calculated at 425 and 450 nm bleach signals, and gave 1.388 ps (r2 = 0.825) and 0.530 ps (r2 = 0.922). Thus, the better signal to noise at those wavelengths produced clearer exponential decay data. Similarly, exponential decays of the PIA signals were fitted for the 475, 525, and 575 nm data. Only 575 nm gave a nonnegligible value, and it was 6.553 ps (r2 = 0.515), which is a poor fitting, and can not be trusted.
68 0 5 10 15 20 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 Bilayer Film, 12 nJ Absorption (mOD)Time (ps) 400nm 425nm 450nm 475nm 525nm 575nm 400 500 600 -40 -30 -20 -10 0 10 20 20 Bilayer Film, 12 nJ Absorption (mOD)Wavelength (nm) -2ps 0ps 0.2ps 0.5 1ps 3ps 20ps 200ps Figure 41 2 20 bilayer film excited with 12 nJ at 420 nm. The left graph plots three bleach signals (400, 425, and 450 nm) and three PIA signals (475, 525, and 575 nm) versus time. Noise is considerable, yet bleach decay appears to take place within 5 ps. The right graph plots the wavelength spectrum at chosen times to show the dynamics. In Figure 41 2 the TA data from the 20 bilayer film is plotted versus time and wavelength once again. Noise begins to become a bigger issu e compared to the 30 bilayer film. Nevertheless, bleach lifetime decay components were obtained for the bleach signal at 425 nm and 450 nm as 0.735 ps (r2 = 0.827) and 0.851 ps (r2 = 0.721). Notably, the degree of fitness is worse, likely from the lower signal to noise ratio. Also, the only SE fit which could be obtained was = 0.213 with (r2 = 0.433) at 575 nm. The bleach signal is still clear on the right graph in Figure 4-1 2 but this is not the case for the 10 bilayer film in Figure 4-13 Noise below 425 nm wavelengths appears as large as the true signal itself, reaching as low as 10 mOD, while PIA is all but impossible to evaluate. Looking at the tim e plot on the left of Figure 4-1 3 the highlighted 450 nm bleach decay is seen up to 5 ps. Fitting of these data with a single exponential decay over the first 5 ps yields = 0.187 ps (r2 = 0.515), which is very short compared to previous values.
69 0 5 10 15 20 -15 -10 -5 0 5 10 10 Bilayer Film, 12 nJ Absorption (mOD)Time (ps) 400nm 425nm 450nm 475nm 525nm 575nm 400 500 600 -15 -10 -5 0 5 10 15 10 Bilayer Film, 12 nJ Absorption (mOD)Wavelength (nm) -2ps 0ps 0.2ps 0.5 1ps 3ps 20ps 200ps Figure 41 3 10 bilayer film excited with 12 nJ at 420 nm. The left graph plots three bleach signals (400, 425, and 450 nm) and three PIA signals (475, 525, and 575 nm) versus time. Noise is very strong, yet bleach decay can be seen within 5 ps in the highlighted 450 nm data. The right graph plots the wavelength spectrum at chosen times to show the dynamics. 0 5 10 15 20 -15 -10 -5 0 5 10 5 Bilayer Film, 12 nJ Absorption (mOD)Time (ps) 400nm 425nm 450nm 475nm 525nm 575nm 400 500 600 -15 -10 -5 0 5 10 15 5 Bilayer Film, 12 nJ Absorption (mOD)Wavelength (nm) -2ps 0ps 0.2ps 0.5 1ps 3ps 20ps 200ps Figure 41 4 5 bilayer film excited with 12 nJ at 420 nm. The left graph plots three bleach signals (400, 425, and 450 nm) and three PIA signals (475, 525, and 575 nm) versus time. Signal to noise is considerable, yet bleach decay appears to take place within 5 ps. The right graph plots the wavelength spectrum at chosen times to show the dynamics. One would think that the trend of decreasing signal to noise with decreasing bilayer count would make the 5 bilayer film uninteresting if not unreadable. However, in Figure 41 4 an interesting band was detected. Unaggregated stimulated emission
70 within the first picosecond wa s observed as a negative signal around 450 nm. As previously noted, this corresponds to unaggregat ed PPE -SO3 stimulated emission. This is very interesting because this signal was not seen in any of the other bilayer film s The 450 nm signal must mean that the po lymer exists with little or no aggregation at 5 bilayers which is logical because aggregation has been shown not to form until many bilayers are built -up and the polymers have had opportunity to interact and -stack with neighbor chains for instance .51 Thus, when finding the bleach lifetime of the 5 bilayer film at 450 nm, an unexpectedly good fitting was obtained, with = 0.207 ps. Table 4 3 summarizes the lifetimes obtained at the selected bleach and PIA signals of 425, 450, 475, and 575 nm for the 4 types of bilayer films. 525 nm was also fit, but produced no real numbers. In general, the higher the BL number, the better the datas signal to noise, and hence the better the fit and the higher the r2 valu e. As noted before, the 450 nm signal for the unaggregated 5 bilayer film shows a stimulated emission signal. Table 4 3 Lifetimes of states using single exponential decay fittings. BL 425nm (r2) 4 50nm (r2) 4 7 5 nm (r2) 575 nm (r2) 30 1.388ps (0.825) 0.530ps (0.922) 6.553ps (0.515) 20 0. 735ps (0. 827) 0. 851ps (0. 721) 0.213ps (0.433) 10 0.187ps ( 0.587) 0.524ps (0.080) 5 0. 207ps ( 0.729) 6.99ps (0.191)
71 CHAPTER 5 CONCLUSIONS AND FUTURE WORK Experimental Discussion To keep the films as fresh as possible, they were stored in a vacuum desicc ator when not in use; however, while on the TA set up, they were exposed to the atmosphere. Because it took several days to reach the optimal TA setting, water in the air may have degraded the titania structure, even at the low pressures in the desiccator. Also, the films were dried for 12 hours between coatings of 10 bilayers but whenever possible, all layers made in the future should b e c oated in a single trial. Humidity must be kept at a minimum. The rotating stage proved fruitful, and tape held the films to the back of it surprisingly well. The only trouble was in finding the correct overlap position of pump and probe on the thin fi lm. This was accomplished by alignment through a 100 pinhole (inside the circular rotating cell, but not rotating) while detecting with a photo diode/oscilloscope placed behind the pinhole and adjusting the probe white light to go through the center. Then, without moving the stationary pinhole, the pump beam was aligned through it. Finally, the entire stage was translated along the beam axis a measur ed number of millimeters to place the taped film at the previous pinhole position. Additionally, a major problem was discovered part way through experimentation with the optical delay stage. It appeared that somehow it was changing alignment as it moved, which meant at long timescales a reproducible decay due to loss of overlap was constantly seen. To fix this problem, two TA set -up mirrors before the optic al delay stage were iteratively adjusted while the beam after the stage, measure d at long distances (taken to a far wall with an inserted mirror), was monitored. To check that it
72 was fixed, a solution of perylene was used as the sample and near constant b leach signal was seen due to the long lifetime the molecule. B ecause the optical delay stage problem only became significant at longer (>100 ps) timescales, short er timescales, measured in this paper were ultimately unaffected and are still considered re liable. The generation of white light for the probe also became a problem, as part way through experimentation the CaF2 rotating stage power supply was lost. Apparently, the original power supply was much too powerful for the little rotating motor, so a n ew power supply, which uses less then 5 V, was installed. It was actually connected in parallel with the homemade sample rotating stage motor, which makes experimentation more streamlined The CaF2 window was also replaced, changing the whitelight counts The grating inside the Andor iStar Shamrock also gave trouble. In the first few experiments, it detected the 420 nm pump scatter signal at values closer to 414 nm. Not only that, but the detected wavelength actually changed upon moving the diffraction grating range This is a serious problem which was discovered to be caused by an internal, computer -controlled sine-bar which controls the grating. If this were a mechanical sine bar, it could have been fixed by hand, but because it is internally control led it instead requires the whole Shamrock be sent to the company for fixing. In lieu of that down time, the grating was fixed to the range of ~360 nm to ~630 nm, and calibrated once with a Mercury lamp. Future Experiments Some conditions which could have an affect on the results are temperature and wait time between creating the films and testing them. To control these variables, future tests should be done at constant room temperature for optimum laser operating stability conditions (to avoid Florida su mmer weather dependence fluctuations ). S table pressure
73 and humidity would also help and all tests should be done within one week after the films have been prepared. As it was, the Tsunami and Spitfire had to constantly be monitored and adjusted, which c ould have been avoided with favorab le room stability. And even th ough auto -correlation experiments were run to check the Spitfire pulse width, constant adjustment meant variable instrument response functions in these experiments. Further areas of testing the PPE-SO3/TiO2/FTO glass films include experiments at lower temperatures which would limit the exciton diffusion to its "downhill migration" onl y, in which excitons travel to lower energy states until they reach the one of greatest population.13 ,23 At the lower temperatures, they would be "trapped" and kept from further "hopping" to other states.23 So in the tests performed in the work of this paper at room temperature, a second, thermally a ctivated hopping process likely contributed a second component to the exciton diffusion. Currently, numerous organic conducting polymers are being developed for use in optoelectronic devices .57 60, 68 A new set of poly(arylene ethynylene) conjugated polymers with carboxylic side groups has been developed at the Univer sity of Florida, and preliminary studies have been done. The author of this manuscript was also working on characterizing the ultrafast photo dynamics of those carboxylates at the time of this publication. Additional work to follow this paper would include test s over that entire family of polyelectrolytes (each with slight changes in monomer length or substituents ) to evaluate the validity of the diffusion equations and to determine best choice of sensitizer in polymer -sensitized solar cells.
74 As mentioned before, there are multiple ways to calculate exciton diffusion length, and one of those way s relies of calculating fluorescence lifetimes in the s teady state in the presence and absence of a quencher. Titania acts as quencher, but films without it, or some kind of scaffold to create a mesoporous adsorption surface for increased polymer deposition for TA detection, could not act as a ba ckground signal for the unquenched calculations. Therefore, a similar nanoparticle or surface must be developed or found which can act as titania for structural purposes, but not be involved in any sort of quenching. The search for such a material will be analyzed based on the literature, and more importantly, on TA experiments which try them. From the data collected in thi s paper, the bleach signal was clear, as is a photoinduced absorption into an excited state. But t o truly determine which stimulated emission process appears (from PPE -SO3 or TiO2) additio nal tests would need to be done which include probing higher wavel engths or co nducting photocurrent studies. Instead of changing the grating diffusion angle to not select lower wavelengths studied in this paper, it is instead proposed that the grating be changed altogether. The current set -up grating is 300 lines/mm, s o to increase the wavelength range of signal detection, a grating with fewer lines/mm should be used. Additionally, a technique known as Singular Value Decomposition (SVD) has been developed t o separate multiple signals, like those that appeared in the TA data.77 T he author of this paper has been researching the method, and plans to use it to separate the components of bleach, PIA, and SE. Ideally, though, averaging of more than 10 data sets and using higher energies would improve the signal to noise rati o, and may be needed to get useful SVD results.
75 Conclusions Exciton Diffusion Calculations In conjunction with using computer software to calculate derivatives of the smoothings of the films at different pump excitation energies exciton exciton annihilati on formulas were evaluated and a diffusion coeffi cient (D) of 4.95 x102 cm2 s1 and a one dimensional diffusion length (LD) of 52.2 nm were found. 52.2 nm qualitatively corresponds to about 2 0 bilayers, so that is the proposed limit of efficient build -up of bilayers to improve light collection for use in solar cells. It needs to be noted that it is only a rough estimate because of all of the approximations which were made in the calculations. To find th e number s above, the thickness of the CPE films were not taken as close to that of the pure titania thickness (1.6 microns) but instead were calculated to be 4 1 nm for the PPE -SO3 layers in the 30 bilayer film Additiona lly, the annihilation radius, R, could in the future be based on calculated Frster radii for energy transfer between two polymer segments, but here was based on the assumption found in the literature that the annihilation radius equals one chromophore unit, which is 5 PRUs for PPE -SO3 and corresponds to 6.0 nm .81 The diffusion length found is higher than values for many singlet excitons,51 which can be attributed to assuming a cylindrical (and therefore smaller) absorption volume and because the excitation energies (50 30, and 15 nJ) could have been higher to give clearer excitonexciton annihilation. Furthermore, based on the model,62 other deactivation mechanisms were not explicitly acco unted for, which may be necessary to further refine the calculation s. These include the true effect of TiO2 quenching and quenching by other traps and structural centers within the film.
76 PPE-SO3 Bilayer Film Dynamics To summarize, films of 5, 10, 20, and 30 bilayers were fabricated and tested using TA. The trend was that a greater number of bilayers showed greater signal, with the 30 bilayer film showi ng the strongest signal s The bleach lifetime of that film measured at 450 nm was found to be 530 fs, giving the best fit of any of the lifetimes calculated. Most lifetimes were less than 1 ps, but due to low energy of excitation, signal to noise was con siderable and exponential decay fitting was mostly unreliable at the lower number of bilayers. However, one interesting finding was that the 5 bilayer film showed stimulated emission from the unaggregated form of PPE -SO3, which did not occur in any other film. It is therefore determined that somewhere between 5 and 10 bilayers aggregation becomes dominant in these films. This can be reaf firmed within the literature.15 16 Solar Cells I n the future, design of optimal polymers from theory (calculating the H OMO and LUMO levels)31,37,46 and novel synthetic techniques to make them will provide an increasing need to understand experimentally the photophysical properties of conjugated polyelectrolytes in films. Transient Absorption is a useful technique for unde rstanding these properties, and has been proven to be promising once again. It is hoped that together in the future, theory and experiment can determine optimal compositions for solar cells and other devices, which can then be used to solve the energy cri si s through intelligent product designs that can be implemented with confidence.
77 LIST OF REFERENCES 1. U.S. Department of Energy website U.S. Department of Ener gy. 2009. 10/7/2009. http://www.energy.gov 2. The Future of Power and Energy in the World Lewis, Nate. Presentation. 2004. 10/24/2009. http://online.kitp.ucsb.edu/online/colloq/lewis1/ 3. Sze, S. M. and Ng, Kwok K. Physics of Semiconductor Devices ; Wiley: Hoboken, NJ, 2007. 4. Reference Solar Spectral Irradiance: Air Mass 1.5 NREL. 2008. 10/23/2009 http://rredc.nrel.gov/solar/spectra/am1.5/ 5. Green, Martin A.; Emer y, Keith; Hishikawa, Yosh ihiro; Warta, Wilhelm. Progress in Photovoltaics: Resear ch and Applications; 2008 16, 435-440. 6. Grtzel, Michael. Chemistry Letters 2005 34(1), 8-13. 7. Grtzel, Michael. Photoelectrochemical Solar Conversion by Dye Sensitization. Future Generation Photovoltaic Te chnologies: First NREL Conference ; American Institute of Physics: Woodbury, NY, 1997. 8. Grtzel, Michael. Journal of Photochemistry and Photobiology A: Chemistry 2004 164 3-14. 9. ORegan, Brian and Grtzel, Michael. Nature 1991 353 737-740. 10. Grtzel, Michael. Nature 2001 414 338-344. 11. Wang, Mingkui; Chen, Peter; Humphry-Baker, Robin; Za keeruddin, Shaik; Grtzel, Michael. ChemPhysChem 2009 10, 290-299. 12. Kim, Sanghoon; Lee, Jae Kwan; Kang, Sang Ook; Ko, Jaejung; Yum, J. H.; Fantacci, Simona; De Angelis, Filippo; Di Ce nso, D.; Nazeeruddin, Md. K.; Grtzel, Michael. JACS 2006 128 16701-16707. 13. Mikhnenko, O. V.; Cordella, F.; Sieval, A. B.; Hummelen, J. C.; Blom, P.W.M.; and Loi, M. Journal of Physical Chemistry B 2008 112 11601-11604. 14. Baur, Jeffery; Rubner, Michael F.; Reynolds, John R.; Kim, Seungho. Langm uir 1999 15, 6460-6469. 15. Tan, Chunyan; Pinto, Maur icio R.; Schanze, Kirk S. Chemical Communications 2002 5 446-447.
78 16. Mwaura, Jeremiah K.; Pinto, Mauricio R.; Witker, David; Ananthakrishnan, Nisha; Schanze, Kirk S.; Reynolds, John R. Langmuir 2005 21 10119-10126. 17. Mwaura, Jeremiah K.; Zhao, Xiaoyong; Jiang, Hui; Schanze, Kirk S.; Reynolds, John R. Chemical Materials 2006, 18 61096111. 18. Ramey, Michael B.; Hiller, JeriAnn; Rubner, Michael F.; Tan, Chunyan; Schanze, Kirk S.; Reynolds, John R. Macromolecules 2005 38, 234243. 19. Taranekar, Prasad; Qiao, Qiquan; Jiang, Hui; Ghiviriga, Ion; Schanze, Kirk S.; Reynolds, John R. JACS 2007 129 8958 8959. 20. Furukawa, Naoki; Mair, Chad E.; Takeda, Jun; Kleiman, Valeria D. Applied Physics Letters 2004 85 ( 20), 4645 -4647. 21. Tan, Ch unyan; Atas, Evrim; Mller, Jrgen G.; Pinto, Mauricio R.; Kleiman, Valeria D.; Schanze, Kirk S. JACS 2004 126 13685-13694. 22. Hardison, Lindsay M.; Zhao, Xiaoyong; Jiang, Hui; Schanze, Kirk S.; Kleiman, Valeria D. Journal of Physical Chemistry C 2008 112, 1614016147. 23. Mller, Jrgen G.; Tan, Chunyan; Atas, Evrim; Kleiman, Valeria D.; Schanze, Kirk S JACS 2006 128, 4007 -4016. 24. Mller, J. G.; Lupton, J. M.; Feldmann, J Physical Review B 2005 72, 195208. 25. Mller, J. G.; Lemmer, U.; Raschke, G.; Anni, M.; Scherf, U.; Lupton, J. M.; Feldmann, J Physical Review Letters 2003 91 (26 ), 267403. 26. Holzhey, A.; Uhrich, C.; Brier, E.; Reinhold, E.; Buerle P.; Leo, K.; Hoffman, Michael Journal of Applied Physics 2008 104, 064510. 27. Fahmi, Adil; Minot, Christian. Surface Science 1994 304(3), 343359. 28. Lorenc, M.; Ziolek, M.; Naskrecki, R.; Karolczak, J.; Kubicki, J.; Maciejewski, A Applied Physics B 2002, 74 1927. 29. Nelson, Jenny. Physical Review B 1999 59 (23 ), 1537415380. 30. Scully, Shawn R. and McGehee, Michael D. Journal of Applied Physics 2006 100, 034907. 31. Hagberg, Daniel P.; Marindo, Tannia; Karlsson, Karl Martin; Nonomura, Kazuteru; Qin, Peng; Boschloo, Gerrit; Brinck, Tore; Hagfeldt, Anders; and Sun, Licheng. Journal of Organic Chemistry 2007 72 95509556.
79 32. Henry, C. H. Journal of Applied Physics 1980 51(8), 4494 -4499. 33. Westenhoff, Sebastian; Howard, Ian A.; Hodgkiss, Justin M.; Kirov, Kiril R.; Bronstein, Hugo A.; Williams, Charlotte K.; Greenham, Neil C.; and Friend, Richard H JACS 2008 130 13653-13658. 34. Ai, Xin; Guo, Jianchang; Anderson, Neil A.; and Lian, Tuanquan. Journal of Physical Chemistry B 2004, 108 12795 -12803. 35. Gilot, Jan; Barbu, Ionut; Wienk, Martijn M.; Janssen, Ren A. J. Applied Physics Letters 2007 91, 113520. 36. Veldman, Dirk; Bastiaansen, J. A. M.; Langeveld-Voss, Bea; Sweelssen, Jrgen; Koetse, Marc M.; Meskers, Stefan C. J.; Janssen, Ren A. J Thin Solid Films 2006, 511512, 581586. 37. Meskers, Stefan C. J .; Hbner, Jens; Oestreich, Michael; Bassler, Heinz. The Journal of Physical Chemistry B 2001, 105 9139. 38. Gilot, Jan; Wienk, Martijn M.; Janssen, Ren A. J. Applied Physics Letters 2007 90, 143512. 39. Lanzani, G.; Cabanillas -Gonzalez, J.; Virgil i, T.; Yeates, S.; Ariu, M .; Nelson, J.; Bradley, D. D. C. Physical Review B 2005 71 014211. 40. Gadermaier, Christoph; Lanzani, Gugliemo. Journal of Physics: Condensed Matter 2002, 14 97859802. 41. Sessi, P.; Brambilla, A; Finazzi, M.; Du, L.; Cab anillas -Gonzalez, J.; Egelhaaf, H. J.; Ciccacci, F.; Lanzani, G Journal of Chemical Physics 2008, 466 6567. 42. Ghosh, Hirendra Nath. BARC Newsletter : Founders Day Special Issue 2001 94 100. 43. Yang, Shujiang; Oli shevski, Pavel; Kertesz, Miklos Synthetic Metals 2004 141 171-177. 44. Wakizaka, Daiki; Fushimi, Toshiki; Ohkita, Hideo; and Ito, Shinzaburo. Polymer 2004, 45 (25 ), 8561 -8565. 45. Drori, T.; Sheng, C. X.; Ndobe, A.; Sin gh, S.; Holt, J.; Vardeny, Z. V Physical Review Letters 2008 1 01, 037401. 46. Beljonne, D.; Pourtols, G.; Silva, C.; Hennebicq, E; Herz, L. M.; Friend, R. H.; Scholes, G. D.; Setayes h, S.; Mllen, K.; Brdas, J. L PNAS 2002 99, 1098210987.
80 47. Heeger, A. J.; Kivelson, S.; Schrieffer, J. R.; Su, W. -P. Review of Modern Physics 1988, 60 (3 ) 781-851. 48. Wunnicke, O.; Mller -Buschbaum, P.; Wolkenhauer, M.; Lorenz Haas, C.; C ubitt, R.; Leiner, V.; Stamm, M. Langmuir 2003 19 85118520. 49. Ariga, K.; Hill, J.; Ji, Q Physical Chemistry Chemical Physics 2007, 9 ( 1 9 ), 2319 2340. 50. Atas, Evrim. "Ultrafast Time Resolved Dynamics in Conjugated Dendrimers." Thesis. University of Florida, 2006. 51. Vella, Jarrett "Singlet and Triplet Excitons in Organic and Organometallic Thin Films: Diffusion and Triplet Sensization ." Thesis University of Florida 2009. 52. Bricaud, Quentin, John R. Reynolds post -doctoral associate. "Department of Energy Meeting." Personal communication. 9 April 2009. 53. Bube, Richard H Photovoltaic Materials ; Imperial College Press: Stanford, C A,1998. 54. Valeur, B Molecular Fluorescence: Principles and Applications Wiley: New York, NY, 2002. 55. Forrest, S. R.; Giebink, N. C.; Sun Y. Organic Electronics 2006, 7 375-386. 56. Yamamoto, M.; Tsuchida, A.; Liebe, W.; Burkhart, R.; Tsubakiyama, K. Macromolecules 1996 29, 1589 -1594. 57. Arys, X.; Laschewsky, A.; Jonas, A. M. Macromolecules 2 001, 34 33183330. 58. Liu, G. M.; Zhao, J. P.; Sun, Q. Y.; Zhang, G. Z. J. Phys. Chem. B 2008 112, 3333 3338. 59. Jourdainne, L.; Lecuyer, S.; Arntz, Y.; Picart, C.; Schaaf, P.; Senger, B.; Voegel, J. C.; Lavalle, P.; Charitat, T. Langmuir 2008 24 7842 7847. 60. Caruso, F.; Niikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997 13 34223426. 61. Sundstr m, V.; Benko, G.; Myllyperkio, P.; Pan, J.; Yartsev, P. JACS 2003 125, 11181119. 62. Gulbinas, V.; Sundstr m, V.; Mineviciute, I.; Hertel, D.; Wellander, R.; Yartsev, A. Journal of Chemical Physics 2007 127 144907.
81 63. Sundstr m, V.; Westenhoff, S.; Beenken, W.; Friend, R.; Greenham, N.; Yarsev, A. Physical Review Letters 2006 97, 166804. 64. Sundstr m, V.; Jespersen, K.; Beenken, W.; Zaushitsyn, Y.; Yartsev, A.; Andersson, M.; Pullerits, T. Journal of Chemical Physics 2004 121(24), 1261312617. 65. Sundstr m, V.; Hilgenforff, M. Journal of Physical Chemistry B 1998 102, 1050510514. 66. Gratzel, M.; Hagfeldt, A. Chemical Reviews 1995 95 (1 ), 4968. 67. Gratzel, M.; Rothenberger, G.; Moser, J.; Serpone, N.; Sharma, D. JACS 1985 107(26), 8054 -8059. 68. Schanze, K. S.; Reynolds, J. R.; Jiang, H.; Zhao, X.; Shelton, A.; Lee, S. H. ACS Applied Materials and Interfaces 2009 1(2) 381-387. 69. Snook, J. H.; Samuelson, L.; Kumar, J.; Kim, Y -G.; Witten, J. Organics Electronics 2005, 6 55 -64. 70. Schwartz, B.; Smith, A.; Martini, I. Physical Review B 2004 69 035204. 71. Anderson, A.; Johansson, N.; Broms, P.; Yu, N.; Lupo, D.; Salaneck, W. R. Advanced Materials 1998 10 (11) 859863. 72. Malone, W. M.; Albert, R. Journal of Applied Polymer Science 1973 17 (8), 24572461. 73. Kenkre, V.; Reineker, P.; Exciton Dynamics in Molecular Crystals and Aggregates ; Springer: New York, NY, 1982. 74. Davydov, A. S. Theory of Molecular Excitons ; McGraw -Hill: New York, NY, 1962. 75. Tan, Chunyan Conjugated Polyelectrolytes: Amplified Quenching, Self -Assembly and Sensor Applications ." Thesis. University of Florida 2004 76. Schanze, K S.; Reynolds, J. R.; Kleiman, V. D.; Zhao, X.; Pinto, M.; Hardison, L.; Mwaura, J.; Muller, J.; Jiang, H.; Witker, D. Macromolecules 2006 39 63556366. 77. Altan, Aysun Ultrafast Studies of a Photochromic Oxazine in Solution ." Thesis. University of Florida 2009. 78. Suna, A. Physical Review B 1970, 1 (4), 1716. 79. Craig, D. P. and Walmsley, S. H. Excitons in Molecular Crystals ; W. A. Benjamin, Inc.: New York, NY, 1968.
82 80. Lewis, A. J.; Ruseckas, A.; Gaudin, O. P. M.; Webster, G. R.; Burn, P. L .; Samuel, I. D. W. Organic Electronics 2006, 7 452456. 81. Valencia, V. S.; Kepler, R. G.; Jacobs, S. J.; Beeson, P. M. Exciton Annihilation Studies in Poly(p -Phenylene Vinylene) ; Donnelly Corporation: Tucson, AZ, 1995. 82. Lakowicz, J. R.; Principl es of Fluorescence Spectroscopy ; Kluwer Academic / Plenum Press: New York, NY, 1999.
83 BIOGRAPHICAL SKETCH Derek Austin LaMontagne was born in State College, Pennsylvania in the autumn of 1985. He was raised in New Hampshire and Florida, where he learned to appreciate politics, art, sports, and science. In high school, he played varsity baseball and graduated valedictorian in 2003 earning an International Baccalaureate diploma. He got his Bachelor of Science degree in c hemistry and m athematics from the C ollege of William and Mary in Virginia in 2006 after three years of study during which time he did research for Dr. David Thompson and Dr. Junping Shi He then immediately enrolled in the chemistry graduate program at the University of Florida. There he joined Dr. Valeria Kleimans physical chemistry research group in 2007. He enjoys and still plays many sports, especially saber fencing, through which he has competed with the W&M and UF team s He has won many tournaments, and enjoys teaching the sport t o undergraduates at UF He hopes to continue on at UF to get his doctorate. He also hopes to some day go to Mars save the environment, and become President of the United States of America.