University of Southern California Dissertations and Theses

First steps of solar energy conversion; primary charge generation processes in condensed phase coupled chromophores

(USC Thesis Other)

First steps of solar energy conversion; primary charge generation processes in condensed phase coupled chromophores

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content First steps of solar energy conversion; primary charge generation processes in condensed phase coupled chromophores by Laura Estergreen A Dissertation Presented to FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (CHEMISTRY) December 2020 Copyright 2020 Laura Estergreen ii Acknowledgements While working on my PhD in chemistry I often sought help or advice from friends, family and coworkers to some end or another. I think that, overall, my experience in graduate school has been positive and I think that is almost entirely due to the support I received. Truth be told, I have a collection of memories which could amount to a book of thanks towards people during this time, but I will try to be brief and compress my thanks to a mere few pages. First and foremost, I have to thank my family. They have been a boulder of consistency and support for my entire life, and they showed me the same support during my graduate studies. My immediate family, which is currently my mom, Joanne, and brother, Phil, have helped me in more ways than I think they even realize. After my dad died when I was in high school, my mom and brother became a major source of support. My mom reinforced the principles my dad taught my brother and I and she tried to be strong for us, though she may have been more impacted than both Phil and me. She is the strongest person I know and is who I wish to become. My brother is someone I can call and talk to anytime I need to. He gives amazing advice and is probably the only person I know who can make me laugh to the point of not being able to breath. I love my family so much and I am truly grateful for them. Family extends to more than just the people you live with. My best friend Becka Shoemaker, who I have known since I was 10 years old, has been there for me through the hardest times in my life. She knew how terrified I was coming to LA to pursue a PhD, but she encouraged me the whole time. She even moved down to LA so we could live together, like proper siblings. She is someone I expect to have in my life for the rest of it as she, like my blood relatives, is constant and will always be family. I am truly thankful to have Becka as a part of my life. iii Family such as my grandparents, aunt, uncle and cousins have always been supportive. My second Cousin Paul, his wife Colleen and his daughter Kristen moved to the LA area during my time in grad school, so I have had family I could spend the holidays with. They have done so much for me throughout my time in grad school and I am so grateful for them. I have made a considerable number of friends during my time in LA. One of my good friends, Molly Dyer, whom I met outside of grad school has been an amazing form of support. Molly moved to LA for a job in advertisement, she was a friend of a friend of my friend, Sean O’Connell, looking for a place to stay and I happened to provide said place. Though we are no longer roommates, we both hang out together and she has become one of my greatest and truest friends. I have come to her for so many things, whenever life gets hard and things become too much, she has always made herself available to listen. I don’t know where I would be without her. Sean O’Connell is someone I would like to thank, not just for getting me in touch with Molly, but for being an awesome person and friend. He is one of the most enjoyable people to hang out with that I know. He has taught me a lot of different card games, he does improv and is really good at. Whenever he would stop by my office to chat, regardless of how the day was going, the conversation would always end in laughter. For giving enjoyment to my day to day, thank you. In graduate school I made many more friends. During my first week of grad school I got to meet the group friends who would be there for me for many years. Alexandra Aloia, Paymaneh Malihi, Erica Howard, Joel Patrow, Jimmy Joy and many more. Jimmy joined Steve’s lab with me and we both worked on similar projects. It was great having him around. I used to refer to Jimmy as a ninja, because he would just stealthily complete experiments and end up with a bunch of really good results. Joel joined Jahan Dawlaty’s lab, but when he would get tired of working, or just needed a break, he would come to my office and talk to either just me or whoever else was in the iv office. It was something I really looked forward to, and I really enjoyed our conversations. Alexandra, Paymaneh, Erica and I used to go on regular hiking trips, which were extremely fun and took us away from the stress that comes with grad school. We used to have lunch together where we could vent our mutual stresses and frustrations and, because we all were going through a similar experience, we could all sympathize with one another. Erica became an amazing friend and source of support. We used to do Krav Maga together and we also lived together. I would sometimes just come to the Melot office (Brent Melot was Erica’s advisor) to hang out with her. It is through her that I made even more friends; JoAnna Milam-Guerrero, Abbey Neer, Kelsey Bass, Nick Bashian, Shilliang Zhou (Nemo), Joseph Stiles, Sabrina Mir, Nicole Spence, Mari Rustebakke and many others. Abbey and JoAnna taught me everything I know about magnetism, which is far less than they know, and also became extremely close friends. JoAnna helped me out when I had a rough time where I ended moving in with her and her family. Her husband Victor became a really great source for advice for finishing up grad school and moving on to the next step. Her son Max is probably the cutest little boy anyone will ever meet. He is sassy and extremely smart for his age, but he has the joy and innocence of a toddler which is a breath of fresh air amidst all the stresses of grad school. I also got to live with Possum and Sophie, JoAnna and Victor’s dogs, they didn’t even have to do anything, and they were a source of stress relief and support. JoAnna, as she was going through the same experiences as me and dealt with some of the same emotional and mental stresses that I do, became a great friend and means of moral and emotional support. Her and her family will always be my extended family and I am so grateful to them. I would like to thank my lab mates in the Bradforth group; Mike Kellogg, Ryan McMullen, Gaurav Kumar, Dhrittiman Bhattacharyya, Jimmy Joy, Saptaparna Das and Anirban Roy. They v have all been great colleagues and I have learned so much from all of them and I can absolutely say that my success was a direct product of what they taught me. During the last couple of years of my PhD we aquired two post docs, Matt Bain and Tillmann Buttersack, and two graduate students, Shivalee Dey and Jose Godinez Castellanos, and they have been great to work with and learn from. Matt helped me with what amounts to the final chapter in my thesis where we worked together to design a new anisotropy set up with dual detection. He was the one to write the labview program and he provided many cups of coffee. He became a great friend and colleague. I would like to thank Mark Thompson and his group, specifically, Nadia Korovina, Jessica Golden and Austin Mencke, as they were my collaborators on a number of projects. Without Mark, I would not have gotten to do the research I was able to do and learn about all the cool excited state behaviors of chromophores. I am truly grateful to Mark because he never makes you feel judged for not knowing something and he is just really good at explaining things. I would like to thank my screening and qualifying exam committee, Jahan Dawlaty, Susumu Takahashi, Jayakanth Ravichandran, Brent Melot and, of course, my advisor Steve Bradforth. I would also like to thank my defense committee, Steve, Mark Thompson and Jayakanth Ravichandran. Thank you all for passing me and ultimately making the achievement of a PhD possible. I would also like to thank Alex Benderskii, Jahan Dawlaty and their group members as they helped me a lot in my research and understanding the complex processes in the systems I got to study. Last but most certainly not least, I would like to thank my advisor, Steve Bradforth. Steve is probably one of the best people at time management I know. During my time in grad school he started as Department Chair and then became Dean of Natural Sciences and Mathematics. Even with all of his administrative duties he still tried to find time to meet with everyone in his group vi individually. He taught me so much in terms of spectroscopy as well as the complexities involved in excited state processes within molecules in solution. He helped me understand pump-probe anisotropy as well as the practical considerations in setting up the experimental apparatus. Aside from being a great research scientist, he is also a good person. He was sympathetic when I was dealing with some of my own personal problems and was understanding when I was feeling down about something, be it related to my research or life. Though he had many time constraints, if I ever needed to talk to him, he would find time in the day to meet with me. I am truly, truly appreciative of Steve as a researcher and advisor, without him, I don’t think my graduate experience would have been as great as it was and I don’t think I would have learned to the depth and degree that I did. So, Steve, thank you. There are many more people I would like to thank, but as stated previously the list and memories are too long to put into a couple of pages. So, thank you everyone! vii Table of Contents Acknowledgements ..................................................................................................................................... ii List of Figures .............................................................................................................................................. x List of Tables ........................................................................................................................................... xvii Abstract ................................................................................................................................................... xviii Chapter 1: Introduction ............................................................................................................................. 1 Solar cell efficiency ................................................................................................................................. 3 Excited State Processes in Organic Chromophores ............................................................................... 11 Experimental Overview ........................................................................................................................ 16 3.1 Time Correlated Single Photon Counting ..................................................................................... 16 3.2 Transient Absorption .................................................................................................................... 18 4 Thesis Outline ..................................................................................................................................... 21 Chapter 1 Bibliography ........................................................................................................................... 24 Chapter 2: Morphological Effects on Singlet Fission within systems of DPT isomers ....................... 27 1. Introduction ...................................................................................................................................... 27 2. Experimental..................................................................................................................................... 32 2.1 Sample Preparation ....................................................................................................................... 32 2.2 Steady-state Characterization ........................................................................................................ 32 2.3 Time Resolved Photoluminescence .............................................................................................. 33 2.4 Femtosecond Transient Absorption .............................................................................................. 33 3. Crystalline DPT................................................................................................................................ 34 3.1 Time resolved photoluminescence ................................................................................................ 34 3.2 Crystal Morphology ...................................................................................................................... 36 3.3 Conceptual description of crystal morphology towards singlet fission ........................................ 37 4. Amorphous Vapor Deposited Thin Films ....................................................................................... 40 4.1 Time resolved photoluminescence ................................................................................................ 40 3.2 Steady-state photophysics ............................................................................................................. 42 3.3 Transient Absorption .................................................................................................................... 44 5. Spincast versus Vapor Deposited 5,12-DPT ................................................................................... 47 Conclusion ............................................................................................................................................. 48 Chapter 2 Bibliography .......................................................................................................................... 50 Chapter 3: Controlling Symmetry Breaking Charge Transfer in BODIPY Pairs.............................. 53 1. Introduction ...................................................................................................................................... 53 2. A Brief Review of SBCT in 9,9’-bianthryl ...................................................................................... 55 viii 3. SBCT in Dipyrrin Dimers ................................................................................................................ 57 3.1 Interchromophore separation ...................................................................................................... 59 3.2 Structural Rigidity and Chromophore Configuration ................................................................. 67 3.3 SBCT Facilitated by Intermediate States .................................................................................... 71 4. Summary & Outlook ........................................................................................................................ 75 5. Experimental..................................................................................................................................... 75 5.1 Femtosecond transient absorption .............................................................................................. 75 5.2 Global analysis ........................................................................................................................... 76 Chapter 3 Bibliography .......................................................................................................................... 78 Chapter 4: Symmetry Breaking Charge transfer in DIPYR dimers.................................................... 83 1. Introduction ...................................................................................................................................... 83 2. Experimental Methods ...................................................................................................................... 87 2.1 Sample Preparation ..................................................................................................................... 87 2.2 Steady State Photophysics .......................................................................................................... 87 2.3 Femtosecond Transient Absorption ............................................................................................ 88 2.4 Nanosecond Transient Absorption .............................................................................................. 88 3. Results and Discussion ..................................................................................................................... 89 3.1 Steady-state Photophysics ............................................................................................................. 89 3.2 Femtosecond Transient Absorption .............................................................................................. 93 3.3 DFT calculations of bis-DIPYR and bis-α-DIPYR .................................................................... 106 3.4 Nanosecond Transient Absorption .............................................................................................. 107 4. Conclusions .................................................................................................................................... 111 Chapter 4 Bibliography ........................................................................................................................ 115 Chapter 5: Pump-probe Anisotropy of Symmetry Breaking Charge Transfer (SBCT) Systems ... 118 Introduction .......................................................................................................................................... 118 Pump-Probe Anisotropy Principles ...................................................................................................... 121 Anisotropy in SBCT Compounds ........................................................................................................ 125 Experimental Apparatus ....................................................................................................................... 127 Anisotropy of Relevant Monomers ...................................................................................................... 129 Anisotropy of SBCT Dimers ................................................................................................................ 132 6.1 Zinc-dipyrrinato dimer (zDIP2) .................................................................................................. 132 6.2 M 8B ....................................................................................................................................................................................................................... 136 6.3 Bis-DIPYR .................................................................................................................................. 140 Improvements to Anisotropy Setup ..................................................................................................... 143 Summary and Future Work .................................................................................................................. 147 ix Chapter 5 Bibiography ......................................................................................................................... 150 x List of Figures Figure 1.1: a) Reflectivity and absorption depth of crystalline silicon as a function of wavelength. b) AM 1.5 G solar irradiance spectrum as a function of wavelength with relative absorptivity of a small molecule (a-DIPYR) and crystalline silicon. Figure 1.2: a) Energy diagram of a p-n junction semiconductor photovoltaic. Excitation can occur in both the p-type and n-type layers where an electron (solid blue circle) is promoted to the conduction band and hole (open circle) to the valence band, forming mobile charges. Electrons flow from the p layer to the n layer and holes from the n layer to the p layer. b) Energy diagram of an organic photovoltaic. Excitations can occur by exciting the electron to the LUMO and holes to the HOMO in either the donor (D) or acceptor (A) layers. The excitons move to the interface, from there the electrons and hole separate where the electrons move to the A layer and holes to the D layer. For simplicity, only excitations in the p-type layer and D layer are depicted, though realistically excitations can occur in either layer. Figure 1.3: a) current (J) vs voltage (V) curve where the product of J and V give the power. The max current, Jmax and max voltage Vmax are compared to the short circuit current JSC and open circuit voltage, VOC. The dashed red lines denote the JSC VOC product or maximal ideal power of a device and are related to the grey box which represents the measured maximum power. The dashed grey curve is the dark current JD (or saturation current) which increases with increasing voltage. b) Illustration of a photovoltaic, taken from Ref. 8, within a circuit where J denotes the current, or movement of charges through the wire and V is the potential and RL is an external load which is collecting power from the device. JD is included to indicate with a voltage present, the dark current will move charges in the opposite direction of the photon induced currednt c) Illustration of the open circuit voltage, VOC, where the wires no longer form a closed loop d) Illustration of the short circuit current, JSC, where there is no potential, V=0. Figure 1.4: Jabloski diagram of allowed photochemical processes. Figure 1.5: a) Diagram of TCSPC apparatus. An excitation source is used to excite a sample as well as act as a reference photon source. The sample fluorescence is emitted into all direction in space; a portion is collected by a lens system which images the fluorescence onto the entrance slit of a double “negative” monochromater. Each grating disperses the fluorescence signal by wavelength. The grating pair are geometrically arranged to introduce zero-time delay across the dispersed colors. Photons at the wavelength selected by a slit at the monochromater exit is then directed to a fast photon multiplier tube (PMT) where the fluorescence is detected and converted into digital ‘shots’ or ‘counts’ which give the histogram. b) Timing scheme of TCSPC apparatus where the delay time, τ, is the period of time between each laser pulse at the reference. When a photon event is detected from the signal, this acts a ‘start’ where after some time the reference pulse is detected, giving a stop. By looking 2 4 7 12 18 xi at the photon events as a function of τ-tn the temporal location for each florescence event can be monitored, giving the final histogram. Figure 1.6: a) Transient absorption spectroscopy set up. A pump pulse acts to excite the sample, where it is blocked by a mechanical chopper for every other shot. A probe pulse arrives at a time delay, τ, and the pump and probe intersect at the sample. The pump is blocked with an iris after the sample; the probe after transmission through the sample is directed to a grating where the white light probe is dispersed into various wavelength components. The dispersed probe is subsequently directed onto a 256- channel diode array to give a spectrally resolved trace. b) The pump-induced transmitted white light spectrum of the sample and the white light signal as it passes through the sample in its ground state. The difference gives the change in absorption (ΔAbs. (λ)) with respect to wavelength (λ) giving a trace comprised of a positive excited state absorption (ESA) and negative ground state bleach (GSB) and stimulated emission (SE). Figure 2.1: Schematic representation of singlet fission. Figure 2.2: Diphenyltetracene systems of interest Figure 2.3: Time resolved photoluminescence measurements of 5,11-DPT (a) and 5,12-DPT (b) as crystals and dissolved in chloroform (CHCl3). From the fit curve, 5,11-DPT gives an initial amplitude of 1 and a final amplitude for the delayed fluorescence of 3.8 ×10 -4 where these amplitudes can be used to determine the ΔGSF for crystalline 5,11-DPT using Equation 2. Figure 2.4: Crystal structure of 5,11-DPT (left) and 5,12-DPT (right) indicating molecular packing. The boxed inset is the crystal structure of o-BETB, a molecular tetracene dimer which has been shown to form the 1 (T1T1) on the isolated molecule. Figure 2.5: Illustrations of the orbital overlap of tetracene molecules in the 'Stacked', 'Slip-stacked' and 'Herringbone' geometries. The illustrated orbitals are based on DFT calculations of tetracene. Figure 2.6: Time resolved photoluminescence of 5,12-DPT and 5,11-DPT. To the right is an energy scheme of geminate recombination back to the excited singlet, where the relative differences in energy between the 5,11 and 5,12-DPT states are indicated by the initial and final photoluminescence amplitudes Figure 2.7: Steady-state absorption and emission of 5,12-DPT (a) and 5,11-DPT (b) in both film (solid black line) and dissolved in chloroform (CHCl3, dotted dark cyan line). Figure 2.8: Time slices of the transient absorption spectral traces for 5,12-DPT (a) and 5,11-DPT (b). Normalized wavelength traces of the S1 absorption (c) and T1 absorption (d) for 5,12-DPT and 5,11-DPT. 20 29 31 35 36 38 40 42 44 xii Figure 2.9: Wavelength slices of the S1 and T1 absorption for 5,12-DPT (a) and 5,11- DPT (b). Δ Absorption traces have been changed to indicate a population density in terms of number of excitations per μm 3 . The relative populations of the S1 and T1 indicate singlet fission efficiencies in both films. Figure 2.10: a) Transient absorption spectra of spincast 5,12-DPT. b) Excitation density of the singlet and triplets as a function of time for both the spin cast (solid line) and vapor deposited (dashed line) 5,12-DPT. c)transient spectra of both spincast and vapor deposited 5,12-DPT at a delay time of 200 fs, where only singlets are present. D)Transient spectra of vapor deposited and spincast 5,12-DPT at a time delay of 750 ps, where the population is primarily triplet (< 1 % are singlet) Figure 3.1: (a) Intramolecular charge transfer (ICT), where the “donor” chromophore is initially excited (D*) followed by electron transfer to an acceptor (A) leaving a positive donor (D + ). (b) Symmetry breaking charge transfer (SBCT) occurs on a set of identical chromophores (Ch) such as anthracene. Either chromophore of the pair can be photoexcited (Ch*), producing a localized excited state (LE) that can undergo SBCT to form a radical cation (Ch + ) and radical anion (Ch - ) Figure 3.2. Materials investigated in this report. Figure 3.3: (a) Jablonskii diagram for SBCT showing the impact of solvent polarity. LE states produced by light can undergo SBCT (kSBCT) or return to the ground state via radiative (kfl) and nonradiative (knr) decay channels. As solvent polarity is increased the SBCT state is stabilized, making its formation from the LE state exoergic. From the SBCT state, the system can decay radiatively (krrec) and nonradiatively (knrrec). (b-e) Absorption and emission spectra of (b) 1,3,5,7- tetramethyl-8-phenyl-BODIPY, (c) m8B, (d) zDIP2, and (e) m8Ph. Absorption spectra of all four systems are similar. Emission spectra of m8B are solvatochromic, indicating an emissive SBCT state while zDIP2 and m8Ph emission is dominated by their LE states. Figure 3.4. Solvent dependent steady state absorption and emission of e4m8B (a) e4m8Ph (b) m4B (c) and m4Ph (d). Figure 3.5. (a) TA spectra of m8Ph in toluene where SBCT is not observed, giving only the LE state transient spectrum. (b) TA spectra of m8Ph in acetonitrile, where m8Ph shows growth of a new ESA feature indicating SBCT. (c) SBCT rate, k SBCT, versus interchromophore separation in acetonitrile. The solid line shows the expected exponential decrease in this rate with increasing interchromophore separation. (d) Decay rate of the SBCT state, krec, (krec = krrec + knrrec,, Fig. 2a) as a function of interchromophore separation. This shows an unexpected result as closely spaced dimers undergo charge recombination more slowly than further spaced dimers. Solid line is a guide to the eye. 45 47 54 57 60 61 64 xiii Figure 3.6: Growth and decay of SBCT state absorption for directly linked dyads m4M and m8M (a) and phenylene-linked dyads m4Ph and m8Ph (b) in acetonitrile (right). The addition of steric groups that hinder rotation slows SBCT and charge recombination for both sets of dyads. Figure 3.7: DFT calculations of the SOMOs of the BODIPY radical anion and radical cation using B3LYP/6-31G(d,p) basis. Molecular orbitals are shown in a parallel (top) and orthogonal (bottom) configuration. The cation and anion orbitals resemble the HOMO and LUMO, respectively, of the neutral BODIPY monomer. Figure 3.8: (a) Illustration of the LE to PCT to SBCT in directly-linked BODIPY dimers. (b) TA spectra of m8B in cyclohexane with overlaid steady-state emission spectra of the 1,3,5,7-tetramethyl-8-phenyl-BODIPY monomer and m8B dimer, highlighting development of a SE band that originates from the PCT state. (c) TA spectra of the BODIPY monomer in THF with overlaid absorption and steady-state emission spectra. (d) TA spectra of e4m8B in toluene with overlaid steady-state emission from the monomer 1,3,5,7-tetramethyl-2,6-ethyl-8-phenyl-BODIPY and e4m8B dimer, highlighting formation of the SE band duet to PCT formation. The monomer emission whas shifted by 20 nm to account for the changes in the ground to excited transition energy when going from the 1,3,5,7-tetramethy-2,6-ethyl-8-phenyl- BODIPY monomer to the e4m8B dimer. Figure 4.1: a) solar irradiance spectrum with SBCT BODIPY dimers. The dotted line to longer wavelengths indicates where we would like our dyes to absorb light. b) Diagram for SBCT and how this could be implemented at a donor/acceptor interface of a device. Figure 4.2: Borondifluoro dipyridylmethene monomer and dimer systems of interest Figure 4.3: Steady-state absorption and emission of DIPYR and bis-DIPYR (a) and α-DIPYR and bis-α-DIPYR (b)in cyclohexane. Solvent dependent absorption and emission of bis-DIPYR (c) and bis-α-DIPYR (d). Figure 4.4: Transient absorption spectra of bis-DIPYR (left) in cyclohexane (a), tetrahydrofuran (c) and acetonitrile (e) with emission spectra of DIPYR monomer (dashed line) and bis-DIPYR (dotted line). Transient absorption of bis-α-DIPYR (right) in cyclohexane (b), tetrahydrofuran (d) and acetonitrile (f) with emission spectra of the α-DIPYR monomer (dashed line) and bis-α-DIPYR (dotted line). Transient traces were normalized to the GSB at the earliest time in each plot. Pump wavelengths for transient absorption experiments were centered at 500 nm. Figure 4.5: SADS of bis-DIPYR in cyclohexane (a) tetrahydrofuran (c) and acetonitrile (e). Spectral slices plotted as a function of time for bis-DIPYR in cyclohexane (b) tetrahydrofuran (d) and acetonitrile (f) indicating ESA (dark blue), GSB (dark cyan) and SE (dark red), with spectral slices taken from global fitting overlaid. 68 70 72 84 87 90 95 97 xiv Figure 4.6: Transient absorption spectra with cation (blue, dashed line) and anion (red, dash-dot line) absorption determined by chronoamperometric spectroelectrochemistry of bis-DIPYR (a) and bis-α-DIPYR (b) in acetonitrile. (c) Transient absorption spectra of bis-DIPYR in acetonitrile with cation (blue, dashed line) and anion (red, dash-dot line) absorption determined from pulsed-radiolysis. (d) TA of bis-α-DIPYR in acetonitrile with the anion absorption overlaid, where the anion was obtained using a different spectroelectrochemical cell from (b). This was the cell used for infrared spectroelectrochemistry (IR-SEC) as reported by Kubiak and coworkers (ref 22). Figure 4.7: Transient absorption spectra of bis-DIPYR in THF (a) and MeCN (c) where only the GSB and SE bands are included (450 – 650 nm). Simulated spectra where excited evolution only include the monomer and dimer emission (b) and simulated spectra with the monomer and dimer emission as well as the cation and anion absorption (d). Figure 4.8: SADS of bis-α-DIPYR in cyclohexane (a) tetrahydrofuran (c) and acetonitrile (e). Spectral slices plotted as a function of time for bis-α-DIPYR in cyclohexane (b) Tetrahydrofuran (d) and acetonitrile (f) indicating ESA (dark cyan/dark blue), GSB (orange), and SE (red), with spectral slices taken from global fitting overlaid. Figure 4.9: DFT calculations of bis-DIPYR (left) and bis-a-DIPYR (right) of their HOMO (bottom) and LUMO (top) orbitals using B3LYP/6-31G** basis (from ref 20). Figure 4.10: Nanosecond transient absotption spectra of bis-α-DIPYR in cyclohexane (a) and tetrahydrofuran (d). SADS of bis-α-DIPYR in cyclohexane (b) and THF (e). Spectral slices of bis-α-DIPYR with global analysis fits overlaid in cyclohexane (c) and THF (f). Spectral slices were taken from wavelengths which intersected ESA (dark blue/cyan), GSB (purple) and SE (red). Figure 4.11: Femtosecond and nanosecond TA time traces of bis-a-DIPYR in cyclohexane (a) and THF (b) with the scaled anion absorption overlaid. Figure 4.12: Proposed Jablonskii diagram for excited state evolution of bis-DIPYR and bis-α-DIPYR. The rate constants denoted by k1and k-1 denote the formation and back transfer of the partial charge transfer, or excimer state with charge transfer character, to and from the initial LE state. From formation of this second intermediate state, k2 denotes relaxation of this intermediate charge transfer state where k2-0 gives the recombination from this intermediate to the ground state. Following this, k3 gives formation of the charge separated state followed by decay to the ground denoted by k3- 0. Figure 5.1: a) Illustration of photo selection with the introduction of a pump pulse. Initially the sample is isotropically distributed, depicted as a sphere. Following this a pump pulse will be introduced to the initially ground state sample and select the chromophores whose transition moments have a projection on the polarization axis giving the cos 2 distribution shown on the right. From there a probe is introduced, 99 102 104 106 110 111 112 122 xv polarized parallel or perpendicular to the photo-selected sample. b) Transition moments on the BODIPY chromophore where the angle θ within the laboratory frame in a) will indicate the angle theta between the pump-induced transition dipole with the probe induced transition moment. c) Diagram of electronic states showing the transition moments attributed to the pump (µpump), ground state bleach (µGSB), excited state absorption (µESA) and stimulated emission (µSE). Figure 5.2: Configuration diagram showing the locally excited (LE) state and the expected absorption processes for the (ground state) radical cation and anion formed by SBCT. The transition corresponding to local HOMO-LUMO promotion is the same for the neutral fragment as for the cationic and anionic fragments of the dimer, and therefore is expected to have the same transition dipole vector direction. The orientation of the transition moments are shown for the LE state (green arrow), the anion absorption (red arrow) and the cation absorption (blue arrow). Figure 5.3: Experimental set up for polarization dependent transient absorption spectroscopy. The pump is sent through a Glan-Taylor polarizer to purify the polarization which is horizontal. Following this the pump is sent through an achromatic λ/2 waveplate so that the pump can be rotated between horizontal and vertical polarizations. The 800 nm pump used for white generation of the probe is sent through a λ/2 waveplate to set the white light polarization as horizontal. Rotating the pump between vertical and horizontal give the perpendicular and parallel configurations, respectively, for the anisotropy experiments. From this the white light is sent to a grating which disperses the light in the horizontal plane onto a 256-pixel silicon diode array. Further description of the transient absorption set up can be found in Chapter 1. Figure 5.4: Parallel (a) and perpendicular (b) transient absorption spectra of gamma- DIPYR (molecular structure shown on the right) in tetrahydrofuran (THF). c) Parallel (red line) intensities of the GSB (square) and SE (circle) plotted as a function of time along with the perpendicular (blue line) intensities of the GSB (upward triangle) and SE (downward triangle) as a function of time. d) Anisotropy of the GSB (blue squares) and SE (green circle) as a function of time. It can be seen that the anisotropies of the GSB and SE follow the predicted behavior where at early times the anisotropy is 0.4, indicating they are parallel transitions to the pump-induced absorption, and 0 at long times due to reorientation of the photo-selected sub-ensemble back to an isotropic distribution. Figure 5.5: Magic angle transient absorption spectra of DIPYR monomer (a) and 1,3,5,7-tetramethyl-8-phenyl-BODIPY (b) in tetrahydrofuran. Anisotropy of DIPYR monomer (c) and 1,3,5,7-tetramethyl-8-phenyl-BODIPY (d) versus time with spectral slices taken from the ESA (black square) GSB (red circle) and SE (dark cyan triangle). As observed for the gamma-DIPYR, the anisotropies indicate that all transitions at early delay are overall parallel to the pump-induced transition moment. At long times, the anisotropy decays to zero, following the expected behavior of monomers in solution. 126 128 129 131 xvi Figure 5.6: Magic angle transient absorption spectra of zDIP2 in THF with the anion absorption from pulsed radiolysis measurements overlaid (red dashed line). b) Anisotropies of the ESA attributed to the formation of the SBCT state. c) Anisotropies of the ESA (black square), GSB (dark red circle) and SE (dark cyan triangle) which are characteristic of the LE trace. d) zDIP2 molecule. Figure 5.7: a) Magic angle transient absorption spectra of m8B in THF with the anion and cation absorption indicated by the red and blue lines, respectively, in accordance to Hattori et. al 28 . b) Anisotropies of the anion (red) and cation (blue) bands, where the {} indicates timescales attributed to PCT formation and [ ] indicate timescales attributed to SBCT formation. c) Anisotropies of the ESA (black square), GSB (dark red circle) and SE (dark cyan triangle) which are characteristic of the LE trace with an inset zooming in on the first picosecond of the anisotropy. d) m8B molecule. Figure 5.8: Magic angle transient absorption spectra of bis-DIPYR in THF with the anion (red dashed line) and cation (blue dashed line) absorptions form pulsed radiolysis measurements overlaid. It can be seen in the case of bis-DIPYR, there is no indication of new absorption features of the in the transient spectrum attributed to the formation of the cation and anion, implying that SBCT does not occur. b) Anisotropies of the wavelengths attributed to anion and cation absorption, if they were formed or two spectral slices through the dimer emission component of the SE. c) Anisotropies of the ESA (black circle), GSB (dark red circle) and SE (dark cyan triangle) which are characteristic of the LE trace. d) bis-DIPYR molecule. Figure 5.9: New detection set up for pump-probe anisotropy experiments. The pump is polarized along the horizontal polarization while the probe is rotated to 45º with respect to the pump. The probe is collimated after the sample and sent through a Wallostan polarizer which separates the horizontal and vertical or parallel and perpendicular components, respectively, of the white light. The horizontal component of the probe is direct through a prism which disperses the white light onto a 256-pixel silicon diode array where the parallel signal can be detected. The vertical component of the probe is first sent through a 90 º periscope to flip the polarization of the probe to horizontal where it is directed to a prism and dispersed onto a 256-pixel silicon diode array for detection of the perpendicular transient signal. The prims used are an identical pair as are the silicon diode arrays. Figure 5.10 Preliminary anisotropy of γ-DIPYR in acetonitrile (left) and γ-DIPYR molecule (right). The anisotropy plotted is for the SE band at 540 nm. It can be seen that the anisotropy follows the predicted behavior as well as the experimental behavior observed from the sequential collection of the parallel and perpendicular signals. That is the initial anisotropy is 0.4 and the final is 0. 133 137 141 144 147 xvii List of Tables Table 3.1:Photophysical Properties of Dipyrrin Dimers. Table 3.2:Kinetic rates of Dipyrrin Dimers Determined by Femtosecond TA Table 4.1:Rate Constants Determined by Femtosecond and Nanosecond Transient Absorption Experiments 62 66 112 xviii Abstract Currently the solar industry is dominated by solar cells made up of crystalline silicon wafers. Though silicon has shown power conversion efficiencies as high as 26 %, it fails to maximally utilize the already diffuse photons provided by the sun. Organic photovoltaics (OPV) offer an alternative as they contain organic chromophores which can readily absorb photons and be deposited as thin films onto flexible substrates, where they may be used in an array of novel geometries. OPVs, however, tend to fall short in terms of their overall power conversion efficiencies. Maximizing the efficiency of an OPV comes from minimizing loss processes associated with excitation to charge generation. One of the challenges in doing this is producing light absorbing molecules whose excited state lifetime is long enough to diffuse to a donor- acceptor interface while minimizing energy loss from charge separation at the interface. In this thesis ultrafast spectroscopy will be used to study excited state phenomena within systems of coupled chromophores. The excited state phenomena studied in this thesis are singlet fission and symmetry breaking charge transfer (SBCT) which are two processes that can both elongate the excited state lifetime while potentially minimizing energy loss in the eventual generation of charge carriers for power conversion. These processes are studied by looking at the efficiencies of singlet fission and SBCT as a function of either morphology or molecular structure of the coupled chromophores. 1 Chapter 1 Introduction The primary source for world energy consumption is from fossil fuels in the form of coal, petroleum and natural gas 1,2 . Fossil fuels, however, have proven to be problematic as they are limited in resource though the energy demand is ever-increasing and will soon exceed the supply of fossil fuels 3 . Not only that, environmental effects from burning fossil fuels have proven detrimental, impacting the health and safety of people and wildlife 1,2 . However, over the past several decades research and development in renewable energy has made strides towards producing energy solutions which will be competitive alternatives. One still developing field is in solar energy, where, despite the magnitude of incident power provided by the sun, it only accounted for about 2% of total energy production in 2018 4 . The solar energy incident on the Earth’s surface amounts to about 3 × 10 24 joules per year, which is enough energy to meet the worlds demands ten thousand times over 5 . However, though the total energy is abundant, the relative intensity at sea level is about 1 kW/m 2 . In other words, in order to produce a gigawatt of energy, assuming a 20% power conversion efficiency, an area of 5 km 2 would be required 2 . In order for solar energy to be a competitive alternative to fossil fuels, absorption efficiencies need to be optimized so that even at low intensities, the device can utilize all possible photons. Part of this pertains to placement and engineering of devices, but this brings merit to research towards optimizing the initial step of photon capture towards charge generation. Currently the solar cell industry is dominated by single junction photovoltaics made of crystalline silicon wafers which run in parallel to create modules with large surface areas. Silicon is an indirect band gap material where the band gap (EG) is about 1.1 eV 6 (1100 nm) which lies in the near infrared (NIR) part of the solar spectrum. That is, the lowest photon energy for absorption 2 is 1.1 eV, implying that absorption will occur across a large range of the solar spectrum, though photon energies greater than EG will result in wasted energy due to relaxation processes, as the generated charges relax to the band edge. The calculated maximum power conversion efficiency of silicon is about 30 % 7,8 where limitations come from the ability for silicon to absorb low energy incident photons and utilize all of the photon energy towards charge generations. In practice, the highest device efficiency of a single junction Si solar cell is about 26 %, thus far 9 . Figure 1.1: a) Reflectivity and absorption depth of crystalline silicon as a function of wavelength. b) AM 1.5 G solar irradiance spectrum as a function of wavelength with relative absorptivity of a small molecule (a-DIPYR 10 ) and crystalline silicon 11 . Though crystallinity is important for high charge mobility, it greatly impacts the ability for devices to absorb photons, where reflection or scattering of photons is more prevalent 12 . This usually results in manufacturing of relatively thick crystals and the use of antireflection coatings to maximize photon absorption 13,14 . It can be seen in Fig. 1a, that the reflectivity of crystalline silicon exponentially increases from 500 – 300 nm and the absorption depth of crystalline Si is about 1 cm around its band gap, but decreases to sub-microns as the photon energy increases 11 (shorter wavelengths). Development of crystalline silicon requires high temperature processing and annealing environments which can have their own environmental drawbacks 4,15 . Second 3 generation photovoltaics, or thin film semiconductors, have been developed from amorphous silicon, II-VI semiconductors, such as CdTe/CdS, and chalcogenides such as CuInGaS (CIGS) 16 . CIGS are examples of second-generation PVs with relatively high efficiencies of up to almost 23 % on glass substrates, where more flexible substrates are currently under investigation 16 . However, deposition onto flexible substrates has proven to impact device efficiency due to the properties of the flexible substrates necessary for high temperature deposition 16 . Organic photovoltaics (OPV) are an alternative to the currently used bulk semiconductor photovoltaics. Low temperature processing and high molecular absorptivity of organics allow for thin film deposition on a variety of surfaces, offering a major advantage over bulk semiconductor devices 17 . One could imagine depositing OPVs onto clothing, backpacks, buildings and many other surfaces in a multitude of geometries. Another advantage of OPVs over bulk semiconductor photovoltaics is that the low temperature processing decreases the overall energy input during manufacturing and maintenance, reducing potential environmental and energetic drawbacks during production. Bearing this in mind, it does pose the question as to why silicon solar cells dominate the solar industry over organic devices. This, in a nutshell, is due to device efficiencies as they currently stand in organic versus bulk semiconductor photovoltaics, where power conversion efficiencies of OPVs fall short. This thesis proposes molecular manipulation which may help in increasing the device efficiencies of OPVs by way of maximizing the efficiency of their light absorbing component. To begin we will overview how device efficiency is determined as it relates to both bulk semiconductor and organic solar cells. 1. Solar Cell Efficiency Ultimately, solar cell efficiency comes down to how much power is output from a device relative to the power input. Shockley and Queisser formulated a thermodynamic description for 4 photovoltaic devices, where a detailed balance limit was calculated to be about 32 % for single- junction devices 8 . This description accounts for both external environmental conditions as well as inherent device properties. The external conditions include temperature, wavelength dependent solar irradiance through the earth’s atmosphere, and geometric constraints in terms of the angle of incidence of the sunlight on a device. The device properties of consideration are temperature, absorption energies (that is the wavelength range from the incident sunlight that is absorbed by the device versus transmitted), exciton generation, charge diffusion and any recombination or relaxation processes which take away from the device efficiency. These device properties and their formulated descriptions as efficiency terms have been modified for OPVs 17 –20 , as they operate differently from typical semiconductor PVs. For the purpose of introducing device limitations towards power conversion efficiency, the properties of a device will be discussed as it relates to photon absorption and charge generation. Figure 1.2: a) Energy diagram of a p-n junction semiconductor photovoltaic. Excitation can occur in both the p-type and n-type layers where an electron (solid blue circle) is promoted to the conduction band and hole (open circle) to the valence band, forming mobile charges. Electrons flow from the p layer to the n layer and holes from the n layer to the p layer. b) Energy diagram of an organic photovoltaic. Excitations can occur by exciting the electron to the LUMO and holes to the HOMO in either the donor (D) or acceptor (A) layers. The excitons move to the interface, from there the electrons and hole separate where the 5 electrons move to the A layer and holes to the D layer. For simplicity, only excitations in the p-type layer and D layer are depicted, though realistically excitations can occur in either layer. A classic example of a photovoltaic device is a semiconducting p-n junction where p and n describe the dopant type in a given layer. The p-type dopant implies that a particle with less valence electrons is introduced at small concentrations, that is far less than a stoichiometric equivalent, to one of the layers essentially introducing holes to the layer and thus increasing the energy of the valence band edge 21 . Conversely an n-type dopant is a dopant that contains more valence electrons, increasing the number of electrons in the layer and decreasing the conduction band edge in energy 21 . The junction of the p and n doped layers creates an energetic gradient which will act to guide the electrons and holes to their respective electrodes. An energy scheme of a p-n junction device is illustrated in Figure 1.1a and described herein. At the interface, the valence and conduction bands of the n- and p-doped layers meet. The bands will ‘bend’ to create a nearly continuous drop in energy as the charges move through the crystal phase of the bulk material 12,21 . Upon excitation an excited electron (solid blue circle, Fig. 1.1a) and hole (open circle, Fig 1.1a) pair is formed, called an exciton, where the electron is excited into the conduction band and the hole is excited into the valence band. If the absorbed photon is greater in energy than the energy difference between the conduction and valence band edges, i.e. the band gap, the electron and hole will relax to the band edges. The excitons formed in bulk semiconductors have a small coulombic attraction, or binding energy, where room temperature (~25 meV) conditions provide enough energy for the electron and hole to dissociate 12 . This results in independent charge carriers which can mobilize through the semiconductor layers via drift and diffusion processes. In accordance to energetics, the electrons will flow towards the n-type doped layer and the holes will flow to the p- typed doped layer from where these mobile charges will be collected for power conversion. 6 In an organic system, instead of bands, there are discrete electronic states 22 where, analogous to a p-n junction, there is a donor (D) and an acceptor (A) layer made up of different molecules 19 . These molecules are defined as being either D or A, based on their electronic properties. The donor (D) molecules exhibit low-lying ionization potentials, where the highest occupied molecular orbital (HOMO) is higher in energy, or destabilized, compared to the acceptor molecules 17 . The acceptor (A) molecules have high electron affinities, giving a lowest unoccupied molecular orbital (LUMO) that is lower in energy compared to the donor molecules 17 . In Figure 1.1b an energy scheme of the OPV analogue to a p-n junction device is illustrated. The interface of the D/A layers gives an energetic offset in the HOMO and LUMO such that it is downhill in energy for the electrons to move from D to A and holes to move from A to D. Upon excitation, like the p-n junction solar cell, an excited electron and hole are formed in either the donor or acceptor layers. The excited electron and hole may lower their energy as the D (or A) molecules undergo relaxation, due to either stabilization processes occurring after excitation or effects caused at the interface of the D and A layers, such as energy bending. The product of the initial excitation is an exciton which moves through either the D or A layer. Finally, at the interface the electrons will move to the A layer (or the holes to the D layer). Unlike the p-n junction cells described above, in molecular systems the formed excitons (called Frenkel excitons) have a high binding energy (~100s meV). Instead of independent electrons and holes moving through the different layers, the coulombically bound excitons will act as the mobile energy carrier and the electron and holes will only separate at the interface where it is energetically favorable to do so. 7 Figure 1.3: a) current (J) vs voltage (V) curve where the product of J and V give the power. The max current, J max and max voltage V max are compared to the short circuit current J SC and open circuit voltage, V OC. The dashed red lines denote the J SC V OC product or maximal ideal power of a device and are related to the grey box which represents the measured maximum power. The dashed grey curve is the dark current J D (or saturation current) which increases with increasing voltage. b) Illustration of a photovoltaic, taken from Ref. 8, within a circuit where J denotes the current, or movement of charges through the wire and V is the potential and R L is an external load which is collecting power from the device. J D is included to indicate with a voltage present, the dark current will move charges in the opposite direction of the photon induced currednt c) Illustration of the open circuit voltage, V OC, where the wires no longer form a closed loop d) Illustration of the short circuit current, J SC, where there is no potential, V=0. The efficiency of a device, η, is determined by the maximal power output from a device, Pmax, relative to the incident power provided by the sun, Pinc. 𝜂 = 𝑃 𝑚𝑎𝑥 𝑃 𝑖𝑛𝑐 8 Let us say that the photovoltaic device is held within a circuit (Figure 1.3a) which has a voltage, V, which indicates the differences in fermi levels of electrons and holes in a p-n junction or HOMO and LUMO energies in an OPV. The wires connect the PV to an external load, RL, where the charges move through the wires giving a current, J, to the external RL. The dark current or saturation current, JD, is the current moving in the opposite direction when the cell is not under illumination and increases with increasing voltage (Figure 1.3 a & b). This results in a description of the power of the device as a product of the current and voltage (i.e. P=JV). We can now put the maximum power output of the device in terms of a power term that accounts for certain physical limitations as it relates to photon absorption and charge generation compared to the measured or true power, Pmeas, of the device. That is 𝑃 𝑚𝑎𝑥 =𝑃 𝑚𝑒𝑎𝑠 𝐽 𝑆𝐶 𝑉 𝑂𝐶 𝐽 𝑆𝐶 𝑉 𝑂𝐶 = 𝑃 𝑚𝑒𝑎𝑠 𝐽 𝑆𝐶 𝑉 𝑂𝐶 ×𝐽 𝑆𝐶 𝑉 𝑂𝐶 =𝐹𝐹 ×𝐽 𝑆𝐶 𝑉 𝑂𝐶 where JSC is the short circuit current and VOC is the open circuit voltage. The product of the JSC and VOC can be thought of as a device specific ideal or maximal power. The fill factor, FF, relates the measured power, Pmeas, to this ideal device power. The fill factor is illustrated in Figure 1.3a as the ratio of the area of the grey box to the box formed by the red dotted lines. This essentially indicates the ratio of a measured maximal power output to an ideal power determined by JSC and VOC. By combining equation (1) and (2) we get the following definition for the power conversion efficiency. 𝜂 =𝐹𝐹 × 𝐽 𝑆𝐶 𝑉 𝑂𝐶 𝑃 𝑖𝑛𝑐 The VOC can be described as the case in which the photovoltaic device is connected to some wires where the wires do not form a complete loop, giving an open circuit (Figure 1.3c). The 9 voltage then is the positive and negative charge build up at the interface of the donor and acceptor layers resulting from photon absorption. In the case of the open circuit, because there is a voltage due to charge build up, JD will be present, where an increased amount of potential build up will direct charges in the opposite direction of the photon induced bias. The JSC is where the system is now a closed loop where there is no external load. That is the current is induced by the spontaneous flow of charges strictly due to photon absorption. By isolating the VOC and JSC components, we can focus on the physical properties of the materials that make up the device to optimize the V OC and JSC and therefore maximize the ideal device efficiency. Now let us understand JSC and VOC in terms of a more physical picture for organic photovoltaics. JSC can be described by the following equation 17 𝐽 𝑆𝐶 =∫ 𝑒 𝜂 𝐸𝑄𝐸 (𝜆 )𝑁 𝑃 ℎ (𝜆 )𝑑𝜆 𝜆 ′ 𝐴𝑀 1.5 where e is the electronic charge, ηEQE(λ) describes how efficiently an incident photon results in mobile charges and NPh(λ) is the photon flux density at wavelength, λ, based on the incident AM 1.5G solar spectrum, integrated over the full solar spectrum 17 . For organic photovoltaics, the ηEQE is the efficiency term which accounts for photon absorption, exciton diffusion, exciton dissociation to free charge carriers, and charge transport and collection 17,20 . That is, of the photons incident on the device, the nEQE accounts for the fraction of photons absorbed by the system as well as the fraction of absorbed photons which result in exciton generation. Since the excitons are the mobile carriers, the nEQE term also encompasses the efficiency or fraction of mobile excitons which make it to the D/A layer and ultimately produce separated charges. Thus, nEQE is really a product of multiple efficiency terms where loss processes are accounted for in terms of JD and exciton recombination/trapping or any processes that inhibit charge generation. 10 The VOC can be described by the following 8,17 𝑉 𝑂𝐶 = 1 𝑒 𝑛 𝑛 ′ Δ𝐸 𝐷𝐴 (1−ln( 𝐽 𝐷 𝐽 𝑆𝐶 )) where ΔEDA is the difference in energy between the HOMO energy of the D molecule and LUMO energy of the A molecule. The factors n and n’ are essentially activity coefficients which account for deviation from ideality due to device properties and junction or interfacial changes, respectively. It can be seen that the VOC is optimized when the ΔEDA is large and JD is small. The VOC does not account for photon absorption explicitly, except for its dependence on JSC. It essentially describes the energy requirement at the interface for the exciton to separate into free charges. When ΔEDA is large, this implies that the energy separation between the resulting electrons and holes is large. However, it should be noted that an inherent energy loss will arise from loss processes associated with moving the electrons and holes to lower energy states 18 , which will be a thermodynamic limit in the device efficiency as well, though this loss is necessary for exciton dissociation. In order to maximize device efficiency, it is clear that the VOC and J SC terms be maximized as they describe the efficiency of generation of separated charges from photon absorption. They also account for loss processes due to reverse current or energy losses going between the D and A layers in the device. The motivation for this thesis is to examine photophysical phenomena which are dictated by the excited state landscape of certain light absorbing chromophores to optimize the first step of photon capture toward either exciton or charge generation. In the following section, the excited state landscape of organic chromophores will be discussed, specifically how certain excited state phenomenon induced by the coupling of chromophores can enable optimization of the JSC and VOC. 11 2. Excited State Processes in Organic Chromophores In the above description of device efficiency, organic photovoltaics were conceptually compared to p-n junction bulk semiconductor photovoltaics. They were described as circuits containing a voltage source which biased the direction of current. This was then broken into specific limiting cases where the JSC and VOC were determined as device properties to be maximized, optimizing the power conversion efficiency (η). JSC was described in terms of nEQE(λ) which is the wavelength dependent efficiency accounting for a multitude of efficiency terms but essentially describes the probability of photon absorption to charge generation. VOC was described in terms of the energy difference between the donor (D) HOMO and acceptor (A) LUMO energy levels where negative effects were due to back transfer, that is an electron moving from A to D (or a hole moving from D to A) and any recombination processes that result in loss of mobile excitons. In short, to optimize photon conversion towards charge generation, exciton diffusion and charge separation must outcompete other decay processes. 12 Figure 1.4: Jabloski diagram of allowed photochemical processes When a photon is absorbed, this results in a perturbation of the electrons within a molecule, causing the molecule to access higher energy electronic states. Ideally, if the chromophore, or light absorbing molecule, is useful as an absorbing medium for applications in OPV systems the chromophore should have a long-lived excited state. This usually leads to an experimental observable such as fluorescence or phosphorescence where radiative recombination occurs from the first singlet excited state (S1) or the triplet state (T1), respectively. Typical excited state lifetimes for systems which decay via fluorescence are several nanoseconds (ns) to tens of ns while systems which decay via phosphorescence have excited state lifetimes of several microseconds (μs) to several milliseconds (ms) 22 . As the initial electronic excitation typically occurs on femtosecond (fs) timescales, fluorescence and phosphorescence indicate excited state lifetimes that are 6 – 12 orders of magnitude longer than this, thus, on electronic time scales, these excitons are very long lived. 13 There are many dyes which can absorb in the visible region of the solar spectrum (300 – 700 nm) that have long lived excited states. However, as shown in Figure 1.4, once the molecule is in its excited state, there are a multitude of processes an excited chromophore can undergo that induce energy loss from the initial excitation. Upon absorption of a photon, the system will populate a singlet excited state (S1 or Sn) depending on the photon frequency 23 . If excited above S1, the system can internally convert (IC) to a lower lying state within the singlet manifold. IC proceeds both by electronic state change and vibrational relaxation of excess energy to its surroundings 23 . Another process which can occur, though is typically slow is intersystem crossing (ISC) where the initial singlet excited state can cross surfaces to an excited triplet state (T 1 – Tn). ISC has structural and electronic requirements which can enable an electron to change its spin. This usually requires coupling of angular momentum provided by an atomic or molecular orbital within the chromophore, i.e. spin-orbit intersystem crossing (SO-ISC), to change the spin configuration 23 . Faster ISC processes can be achieved if population of one of the triplet excited states (Tn) is symmetry allowed which has been observed in photoexcited quinones 24 , though this is uncommon. Finally, excited state deactivation, as discussed above, can occur radiatively via fluorescence and phosphorescence, or it can also occur noradiatively via IC to the ground state (S0). First, let us consider these excited state processes as they relate to JSC. Simply put, J SC is optimized when the energy required for exciton generation is minimized, the fraction of excitons formed from photon absorption is maximized and the fraction of excitons towards the formation of separated charges is large. If we consider the above Jablonski diagram (Figure 1.4), this essentially means minimizing the S0→S1 transition energy in a chromophore, while reducing the rate of IC to the ground. This also means that the exciton needs to be able to diffuse to the interface 14 before recombining, thus requiring relatively large diffusion lengths. Longer diffusion lengths imply larger exciton mobility, yielding a higher JSC. It should be noted that singlet diffusion lengths are usually confined to Förster distances, so several nanometers 25 . Triplets, however, have diffusion lengths that are on the order of microns 26,27 , which pose an advantage to triplet formation. Triplets usually are formed via ISC which is a relatively slow process where the probability of IC prior to ISC is high. However, if triplet formation could occur on faster timescales, outcompeting IC, this will aid in maximizing the JSC efficiency term and overall efficiency. Now let us discuss optimization of the VOC in terms of excited state processes of a chromophore. The VOC is optimized when, according to the previous section, the energy between the donor HOMO and acceptor LUMO is large and charge separation at the interface occurs with minimal energy loss. That is to say, the LUMO state of D must be higher in energy than that of A, though not too high so energy loss is minimized. However, as we decrease the energy difference between the LUMO of D and the LUMO of A, the rate of charge transfer at the interface will decrease, and back transfer or recombination may more readily occur, which will take away from the JSC and VOC terms. Molecular manipulation must be carried out in order to both decrease the energy loss when moving an electron from D to A (or hole from A to D) while outcompeting back transfer of the charges or recombination. The probability of these events decreases when the rate of charge separation at an interface is increased. In short, optimization of the efficiency parameters can be achieved when exciton diffusion and charge transfer outcompete any decay processes within in the D and A layers (depending on which chromophore was excited). Also, optimization of photon absorption is required to maximize the nEQE(λ) term for JSC. Synthetic manipulations can be carried out where the organic chromophores maximally utilize the photons absorbed resulting in excitons that have long 15 diffusion lengths, charge transfer at the D/A interface outcompetes relaxation or recombination processes, and energy losses inherent to charges moving across the D/A interface are minimized. One means of manipulation is to use excited state phenomena which arise from direct coupling of two or more chromophores. This allows manipulation of the excited state architecture where loss processes can be minimized. In this thesis the two processes that will be investigated will be Singlet Fission (SF, Chapter 2) and Symmetry Breaking Charge transfer (SBCT, Chapter 3 – 5). Singlet Fission is used as a means of coupling neighboring chromophores to get two excitons from one photon, effectively doubling the photon conversion efficiency, accounted for in nEQE within JSC. This is done by way of coupling a ground state chromophore (S0) to an excited singlet chromophore (S1) where through interactions of the excited and neighboring ground state chromophores the photoexcited system can transform into two correlated triplets ( 1 (T1T1)) whose total spin configuration is still a singlet. Thus, this is a spin-allowed IC process, unlike ISC, and can occur on picosecond (ps) timescales. Following the formation of the correlated triplet, these triplets can lose coherence and form two separated triplets (T1 + T1). As stated previously, triplet excitons are advantageous relative to singlet excitons as their diffusion lengths are longer, leading to higher exciton mobility and thus larger JSC. They also have longer excited state lifetimes, which will contribute to the VOC as recombination events are minimized. Symmetry breaking charge transfer is a biological phenomenon which enables efficient production of separated charges within photosynthetic organisms powering the biosynthetic processes necessary for their survival. SBCT allows for charge transfer to occur within a molecule that is driven primarily by its environment. It requires a closely associated pair of identical chromophores, or a dimer, where photon absorption forms an initial exciton from which charge 16 transfer subsequently occurs. That is, the formed excited electron and hole pair spatially separate across the dimer to form a radical cation and anion within the overall molecular assembly. Because excitons formed in molecular systems tend to have large Coulombic binding energies, electronic energy is necessarily lost when subsequent separation of the electron and hole is achieved. SBCT spatially separates the electron and hole on the excited dimer molecule, reducing this binding energy and, therefore, reducing the energy loss necessary for charge separation at an interface. This would allow for a decrease in the S0→S1 transition energy, which would have a positive effect on JSC as exciton generation could be done at lower energies, while maintaining long excited state lifetimes. The VOC will also benefit from this since charge separation at the interface can be performed with minimal energy loss. Typical energy loss processes due to charge separation are 1 – 2 eV, while SBCT could reduce this to 100s of meV and still maintain the lifetime of the separated charges, or conversely, reduce back transfer and recombination processes. The process of photon absorption towards exciton and charge generation occur on ultrafast (fs – ps) timescales. In order to understand dynamics and efficiencies of the above processes, probing on ultrafast timescales is required. The following section will outline the ultrafast spectroscopic processes used to study SBCT and SF so that general trends and mechanisms can be extracted and will be described in the following chapters. 3. Experimental Overview The primary ultrafast spectroscopy techniques used in this work were time correlated single photon counting (TCSPC), alternatively known as time-resolved photoluminescence (TRPL), and transient absorption (TA). The concept and methodology will be described herein. 3.1 Time Correlated Single Photon Counting (TCSPC) 17 TCSPC is a very powerful method which monitors the kinetics of radiative and nonradiative decay processes as it occurs over a narrow energy region. This method monitors the emission from a specific excited electronic state as a function of time. The specificity of TCSPC (transitions from a single excited state to the ground) is very advantageous as other ultrafast techniques usually monitor signals that are attributed to absorption and stimulated emission between multiple electronic excited states and their respective processes, making analysis of the data nontrivial. TCSPC is a statistical technique that relates excited state population to the number of detected of photons, where this relative population is monitored over the course of several to hundreds of ns. The concept of photon detection 28 –30 arises from the fact that emission processes are diffuse and occur over a range of angles from the sample, thus the probability of detection of even one photon is small. By using a periodic source, in our case a regenerative amplifier with a 250 kHz repetition rate, a photon detection event is correlated to some time interval. Our set up 31 is specifically called a ‘reverse start-stop’, as it is opposite to classical TCSPC architectures for lower repetition rate lasers or nanosecond flash sources, but the concept and overall mechanism is essentially the same. One can consider a countdown timer, where there is a ‘start’ and a ‘stop’ and the time between the start and stop is recorded. In the ‘reverse start-stop’ configuration, the start is from a signal or emitted photon from the excited sample while the ‘stop’ is due to photon detection from the excitation source. When a signal is detected there is a time-to-amplitude converter which produces a linear ramp in voltage as time passes such that when the ‘stop’ pulse is detected the voltage ramp stops. This gives a position in time with respect to the reference. As multiple excitation events occur and emitted photon signals are accumulated, a histogram is 18 constructed with respect to time, giving a statistical picture of decay events, or population, as a function of time (Figure 1.5). A full description of the apparatus in our lab is described in Ref 31. Figure 1.5: a) Diagram of TCSPC apparatus. An excitation source is used to excite a sample as well as act as a reference photon source. The sample fluorescence is emitted into all direction in space; a portion is collected by a lens system which images the fluorescence onto the entrance slit of a double “negative” monochromater. Each grating disperses the fluorescence signal by wavelength. The grating pair are geometrically arranged to introduce zero-time delay across the dispersed colors. Photons at the wavelength selected by a slit at the monochromater exit is then directed to a fast photon multiplier tube (PMT) where the fluorescence is detected and converted into digital ‘shots’ or ‘counts’ which give the histogram. b) Timing scheme of TCSPC apparatus where the delay time, τ, is the period of time between each laser pulse at the reference. When a photon event is detected from the signal, this acts a ‘start’ where after some time the reference pulse is detected, giving a stop. By looking at the photon events as a function of τ-t n the temporal location for each florescence event can be monitored, giving the final histogram. 3.2 Transient Absorption Spectroscopy (TA) Transient absorption spectroscopy is another very powerful technique in monitoring excited state behavior of different systems. Essentially, one could think of a typical steady-state 19 absorption process where an incoherent source of white light is passed through a sample and the absorbed light is detected as a function of wavelength. Transient absorption spectroscopy monitors absorption processes after a sample has been excited and is illustrated in Figure 1.6a. Initial excitation occurs by using a pump pulse which is on resonance with a transition from the ground state to an excited state in the sample, usually populating the S1 state of a chromophore. Following this excitation, a white light continuum pulse probes the now excited sample. The time delay between the pump and probe pulse is iteratively increased, allowing optical transition of the excited state to be monitored as a function of time. By comparing the white light spectrum as it passes through the sample in its ground versus excited state, we can monitor excited state transitions as the excited state evolves. Changes in transmitted intensities when comparing the white light spectrum of the ground state versus excited state sample are typically very small making the spectral differences hard to distinguish. In order to emphasize these changes one can instead subtract the ground state spectrum from the excited state spectrum to better examine spectral difference between the ground and excited samples (Figure 1.6b). The difference spectra are obtained by introducing a mechanical chopper which is set to half the repetition rate of the laser and blocks the pump pulse every other shot. This allows us to detect both the ‘pump on’ and ‘pump off’ or excited and ground state spectra, respectively, at each time point. This gives live acquisition of the difference spectrum per probe time delay 32,33 . 20 Figure 1.6: a) Transient absorption spectroscopy set up. A pump pulse acts to excite the sample, where it is blocked by a mechanical chopper for every other shot. A probe pulse arrives at a time delay, τ, and the pump and probe intersect at the sample. The pump is blocked with an iris after the sample; the probe after transmission through the sample is directed to a grating where the white light probe is dispersed into various wavelength components. The dispersed probe is subsequently directed onto a 256-channel diode array to give a spectrally resolved trace. b) The pump-induced transmitted white light spectrum of the sample and the white light signal as it passes through the sample in its ground state. The difference gives the change in absorption (ΔAbs. (λ)) with respect to wavelength (λ) giving a trace comprised of a positive excited state absorption (ESA) and negative ground state bleach (GSB) and stimulated emission (SE). The acquired difference spectrum is calculated to give the difference in the ground and excited state absorption (ΔAbs. (λ)) as a function of wavelength, where the spectra contain both positive and negative features. The positive features are excited state absorption (ESA, blue filled curve in Figure 1.6b) processes where the probe continuum has energies on resonance with absorption processes to higher lying excited states. The ground state bleach (GSB, green-filled curve in Figure 1.6b) is a negative feature that is indicative of successful population of the excited state as reflected in a lower population (and thus absorption) from the ground state, resulting in a negative signal. Another negative feature is the stimulated emission (SE, yellow-filled curve in 21 Figure 1.6b) which is where colors within the probe continuum are resonant with and can induce stimulated emission processes from the excited state. The superposition of the ESA, GSB and SE features give an excited state trace. As the probe is delayed in time and the excited state evolves, this trace will change indicating formation of new excited state species. The temporal and spectral information is used to postulate excited state population processes taking place in the molecular sample and infer mechanism. By making adjustments to the most basic transient absorption experiment outlined above, the experiment can be configured to be sensitive to different light activated effects. One that will be utilized and outlined in detail later is a polarization dependent study, where the polarization of the probe response is measured with respect to different polarizations of the pump. This allows us consider the polarization of different excited state transitions, lending insight into the orientation of the relevant transition moments with respect to the molecular frame. This technique is called pump-probe anisotropy where polarization dependent photo-selection combined with known arrangements of different transition moments can help to resolve or further establish character of different excited state species. 4. Thesis Outline: In this thesis the main goal is to understand excited state phenomena resulting from coupling molecular chromophores and how that relates to efficiency terms towards solar power conversion. Specifically, mechanistic understandings of these processes will be sought as well as the establishment of general trends as they relate to structure of the molecular systems. Chapter 2 is an overview of singlet fission within diphenyl-tetracene (DPT) systems, in different morphologies. DPT isomers will be compared based on crystal packing arrangements as 22 well as amorphous thin films as it relates to singlet fission efficiencies. Following this, singlet fission in DPT films will be discussed as it relates to different methodologies of deposition, specifically chemical vapor deposition relative to spincasting. This will help in comparing conceptually the configurational requirements of a ground and excited chromophore towards singlet fission. This will also answer the question of low temperature processing versus higher temperature deposition when comparing spincast and vapor deposited DPT films on their overall rates of singlet fission. Chapter 3 is an overview of symmetry breaking charge transfer (SBCT) in dipyrrin dimers as a function of their molecular properties. Specifically, the rates and lifetimes of the SBCT states will be compared as a function of distance, rigidity and molecular conformation. Then a section pertaining to the observation of an intermediate state will be discussed as it relates to what has been observed in the literature as this compares to the experimental evidence herein. Chapter 4 overviews partial charge transfer (PCT) and SBCT within borondifuoro dipyridylmethene (DIPYR) dimers which are directly linked through their meso positions. These systems are compared to the dipyrrin dimers discussed in chapter 3 in terms of their rates of PCT and SBCT in different solvent environments. This section discusses proposed methods for the observation of charge transfer or charge separation as it relates to spectroelectorchemistry and pulsed radiolysis data combined with nanosecond transient absorption experiments. Chapter 5 discusses polarization dependent behaviors of transient species. Specifically, pump-probe anisotropy is used to understand the orientation of different excited transition moments of SBCT compounds. Specifically, we hope to understand if in some of or all of the SBCT systems there is a preferred direction for hole or electron transfer. In principle as the donor and acceptor chromophores within an SBCT system are identical, there should be no initial driving 23 force for hole or electron transfer, implying both should be equally probable. However, with the observation of an intermediate state, there may be a preferential direction of charge transfer. Part of this chapter relates to design and implementation of a new detection scheme where the different polarization configurations (parallel and perpendicular) can be measured simultaneously. By using ultrafast spectroscopic techniques where light manipulation can be controlled, the complexities within the excited state landscape of coupled chromophores can be disentangled. The understanding of these excited state phenomena will aid in producing systems which can maximize photon absorption to charge generation, either by increasing the photon conversion efficiency or by using site-specific generation of charge carriers at an interface. Ultimately, we can attribute excited state behavior to impacts of efficiency parameters outlined in the previous sections of this chapter. 24 Chapter 1 Bibliography (1) Moomaw, W.; Yamba, F.; Kamimoto, M.; Maurice, L.; Nyboer, J.; Urama, K.; Weir, T.; Bruckner, T.; Jäger-Waldau, A.; Krey, V.; Sims, R.; Steckel, J.; Sterner, M.; Stratton, R.; Verbruggen, A.; Wiser, R.; Pan, J.; Ypersele, J.-P. van. Renewable Energy Sources and Climate Change Mitigation. 2009, 161–208. https://doi.org/10.1017/cbo9781139151153.005. (2) Freris, L.; Infield, D. Renewable Energy in Power Systems; John Wiley & Sons, Ltd: John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom, 2008. (3) Wagner, L.; Ross, I.; Foster, J.; Hankamer, B. Trading Off Global Fuel Supply, CO2 Emissions and Sustainable Development. Plos One 2016, 11 (3), e0149406. https://doi.org/10.1371/journal.pone.0149406. (4) Dupont, E.; Koppelaar, R.; Jeanmart, H. Global Available Solar Energy under Physical and Energy Return on Investment Constraints. Appl Energ 2020, 257, 113968. https://doi.org/10.1016/j.apenergy.2019.113968. (5) Grätzel, M. Mesoscopic Solar Cells for Electricity and Hydrogen Production from Sunlight. Chem Lett 2005, 34 (1), 8–13. https://doi.org/10.1246/cl.2005.8. (6) Tiedje, T.; Yablonovitch, E.; Cody, G. D.; Brooks, B. G. Limiting Efficiency of Silicon Solar Cells. Ieee T Electron Dev 1984, 31 (5), 711–716. https://doi.org/10.1109/t-ed.1984.21594. (7) Tiedje, T.; Yablonovitch, E.; Cody, G. D.; Brooks, B. G. Limiting Efficiency of Silicon Solar Cells. Ieee T Electron Dev 1984, 31 (5), 711–716. https://doi.org/10.1109/t-ed.1984.21594. (8) Shockley, W.; Queisser, H. J. Detailed Balance Limit of Efficiency of P‐n Junction Solar Cells. J Appl Phys 1961, 32 (3), 510–519. https://doi.org/10.1063/1.1736034. (9) Hollemann, C.; Haase, F.; Rienäcker, M.; Barnscheidt, V.; Krügener, J.; Folchert, N.; Brendel, R.; Richter, S.; Großer, S.; Sauter, E.; Hübner, J.; Oestreich, M.; Peibst, R. Separating the Two Polarities of the POLO Contacts of an 26.1%-Efficient IBC Solar Cell. Sci Rep-uk 2020, 10 (1), 658. https://doi.org/10.1038/s41598-019-57310-0. (10) Golden, J. H.; Facendola, J. W.; R, D. S. M.; Baez, C. Q.; Djurovich, P. I.; Thompson, M. E. Boron Dipyridylmethene (DIPYR) Dyes: Shedding Light on Pyridine-Based Chromophores. J Org Chem 2017, 82 (14), 7215–7222. https://doi.org/10.1021/acs.joc.7b00786. (11) Green, M. A. Self-Consistent Optical Parameters of Intrinsic Silicon at 300K Including Temperature Coefficients. Sol Energ Mat Sol C 2008, 92 (11), 1305–1310. https://doi.org/10.1016/j.solmat.2008.06.009. 25 (12) Yu, P. Y.; Cardona, M. Fundamentals of Semiconductors, Physics and Materials Properties. 2010. https://doi.org/10.1007/978-3-642-00710-1. (13) Wang, K. X.; Yu, Z.; Liu, V.; Cui, Y.; Fan, S. Absorption Enhancement in Ultrathin Crystalline Silicon Solar Cells with Antireflection and Light-Trapping Nanocone Gratings. Nano Lett 2012, 12 (3), 1616–1619. https://doi.org/10.1021/nl204550q. (14) Ranjan, S.; Balaji, S.; Panella, R. A.; Ydstie, B. E. Silicon Solar Cell Production. Comput Chem Eng 2011, 35 (8), 1439–1453. https://doi.org/10.1016/j.compchemeng.2011.04.017. (15) Tsuo, Y. S.; Gee, J. M.; Menna, P.; Strebkov, D. S.; Pinov, A.; Zadde, V. Environmentally Benign Silicon Solar Cell Manufacturing; 1998. (16) Ramanujam, J.; Bishop, D. M.; Todorov, T. K.; Gunawan, O.; Rath, J.; Nekovei, R.; Artegiani, E.; Romeo, A. Flexible CIGS, CdTe and a-Si:H Based Thin Film Solar Cells: A Review. Prog Mater Sci 2019, 110, 100619. https://doi.org/10.1016/j.pmatsci.2019.100619. (17) Kippelen, B.; Brédas, J.-L. Organic Photovoltaics. Energ Environ Sci 2009, 2 (3), 251–261. https://doi.org/10.1039/b812502n. (18) Elumalai, N. K.; Uddin, A. Open Circuit Voltage of Organic Solar Cells: An in-Depth Review. Energ Environ Sci 2016, 9 (2), 391–410. https://doi.org/10.1039/c5ee02871j. (19) Organic Photovoltaics, Concepts and Realization. 2003. https://doi.org/10.1007/978-3-662- 05187-0. (20) Lu, N.; Li, L.; Sun, P.; Liu, M. Short-Circuit Current Model of Organic Solar Cells. Chem Phys Lett 2014, 614, 27–30. https://doi.org/10.1016/j.cplett.2014.08.070. (21) Sharon, M. An Introduction to the Physics and Electrochemistry of Semiconductors: Fundamentals and Applications; 2016. https://doi.org/10.1002/9781119274360.ch3. (22) Applied Photochemistry. 2013. https://doi.org/10.1007/978-90-481-3830-2. (23) Kelley, A. M. Condensed-Phase Molecular Spectroscopy and Photophysics, 1st ed.; Wiley: Hoboken, N.J., 2012. https://doi.org/10.1002/9781118493052.ch9. (24) Hubig, S. M.; Bockman, T. M.; Kochi, J. K. Identification of Photoexcited Singlet Quinones and Their Ultrafast Electron-Transfer vs Intersystem-Crossing Rates. J Am Chem Soc 1997, 119 (12), 2926–2935. https://doi.org/10.1021/ja963907z. (25) Fravventura, M. C.; Hwang, J.; Suijkerbuijk, J. W. A.; Erk, P.; Siebbeles, L. D. A.; Savenije, T. J. Determination of Singlet Exciton Diffusion Length in Thin Evaporated C 60 Films for Photovoltaics. J Phys Chem Lett 2012, 3 (17), 2367–2373. https://doi.org/10.1021/jz300820n. 26 (26) Ern, V. Anisotropy of Triplet Exciton Diffusion in Anthracene. Phys Rev Lett 1969, 22 (8), 343–345. https://doi.org/10.1103/physrevlett.22.343. (27) Irkhin, P.; Biaggio, I. Direct Imaging of Anisotropic Exciton Diffusion and Triplet Diffusion Length in Rubrene Single Crystals. Phys Rev Lett 2011, 107 (1), 017402. https://doi.org/10.1103/physrevlett.107.017402. (28) Phillips, D.; Drake, R. C.; O’Connor, D. V.; Christensen, R. L. Time Correlated Single- Photon Counting (Tcspc) Using Laser Excitation. Instrum Sci Technol 2008, 14 (3–4), 267–292. https://doi.org/10.1080/10739148508543581. (29) Becker, W. Advanced Time-Correlated Single Photon Counting Techniques. 2005, 11–25. https://doi.org/10.1007/3-540-28882-1_2. (30) O’Connor, D. V.; Phillips, D. Time-Correlated Single Photon Counting. 1984, 36–54. https://doi.org/10.1016/b978-0-12-524140-3.50006-x. (31) Kloepfer, J. A.; Bradforth, S. E.; Nadeau, J. L. Photophysical Properties of Biologically Compatible CdSe Quantum Dot Structures. J Phys Chem B 2005, 109 (20), 9996–10003. https://doi.org/10.1021/jp044581g. (32) Chen, X.; Larsen, D. S.; Bradforth, S. E.; Stokkum, I. H. M. van. Broadband Spectral Probing Revealing Ultrafast Photochemical Branching after Ultraviolet Excitation of the Aqueous Phenolate Anion. J Phys Chem 2011, 115 (16), 3807–3819. https://doi.org/10.1021/jp107935f. (33) Roberts, S. T.; McAnally, R. E.; Mastron, J. N.; Webber, D. H.; Whited, M. T.; Brutchey, R. L.; Thompson, M. E.; Bradforth, S. E. Efficient Singlet Fission Discovered in a Disordered Acene Film. J Am Chem Soc 2012, 134 (14), 6388–6400. https://doi.org/10.1021/ja300504t. 27 Chapter2: Morphological Effects on Singlet Fission within DPT Isomers 1. Introduction One way to increase the overall efficiency of organic photovoltaics (OPVs) is to increase the overall photon conversion efficiency. That is, increasing the excitations generated per photon absorbed, effectively overcoming the Shockley-Queisser limit, which gives a 32% power conversion efficiency limit for a single junction photovoltaic 1 (PV). The process of multiple electrons out for one photon absorbed has been suggested as a means of exceeding the Shockley- Queisser limit by producing multiple excitations from one photon whose energy exceeds the band gap of a device. This has specifically pertained to both bulk semiconductors as well as inorganic quantum dots in the form of multiple exciton generation (MEG) 2,3 . However, in the context of semiconductor materials MEG would require photons whose energy is multiples of the bandgap energy with very little gain. Typically, due to how easily accessible optical phonons are, these high energy photons produce electrons which simply internally convert or readily thermalize to a lower energy band or to the conduction band edge 4 . To put it simply, the energy input from a single high- energy photon should theoretically output several electrons, however, in practice this primarily amounts to only one electron out, where much of the incident photon energy is lost, resulting in a net negative effect for power conversion efficiency. For organic photovoltaics, singlet fission has been proposed as a method for producing a multiple exciton state 5 –7 . That is a photon is absorbed on resonance with the lowest lying excitation where, due to electronic coupling between chromophores, this initial excitation will form two excitations which can diffuse. It has been proposed that the detailed balance limit of a photovoltaic device would be increased from 32 % to about 45 % using singlet fission in the presence of a single-junction PV 5 . Singlet Fission is the process by which a singlet excited chromophore couples 28 to a ground state singlet chromophore and by an internal conversion mechanism produces two excited state triplets (Figure 2.1). In order for singlet fission to occur within a system of coupled chromophores conservation of energy must be maintained that is E(S1) ≥ 2E(T1), where E(S1) is the energy of the first excited singlet state and E(T1) is the energy of the triplet state. There must also be sufficient electronic coupling between the coupled S0S1 states with the correlated triplet pair, 1 (T1T1), for the system to readily internally convert and outcompete other excited state deactivation processes. Finally, loss of coherence between the correlated triplets must be accessible such that the two triplets may diffuse independently. Triplets are advantageous in that they typically have longer diffusion lengths (several microns) 8,9 than singlet states (several nanometers) 10 and recombine on much slower timescales. If singlet fission can occur with unity quantum efficiencies, this would result in long lived separated excitons which will ultimately result in twice the electrons out per photon absorbed or doubling the photon conversion efficiency. It should be noted that herein, singlet fission is the mechanism as described in Figure 2.1, where the final product is two uncorrelated triplets. This is not to be confused with certain representations in the literature where singlet fission is considered to only account for the formation of the ME state. 29 Figure 2.1: Schematic representation of singlet fission. Ideally, one would want to look at systems whose energies are such that there is no inherent energy loss when going from the singlet excited state of the chromophore to the correlated triplet of the two chromophores. That is where singlet fission is isoergic. For instance, a popular singlet fission material to study is pentacene 11 –15 . Pentacene dimers or films readily undergo singlet fission because twice the energy of the triplet in pentacene is much less than that of the excited singlet. So, in accordance with the linear free energy relation, singlet fission is extremely fast, but this is with a substantial energy loss which will ultimately be taken from the overall power conversion efficiency of the device. However, in the case of tetracene, twice the energy of the triplet state is slightly uphill 5,16 –18 , though tetracene has been shown to readily undergo singlet fission. In crystalline tetracene singlet fission has been shown to occur on the order of 100s of picosecond depending on the morphology 19,20 . It has been proposed that though E(S1) ≲ E(T1), singlet fission is observed due to a combination of entropic effects 16,17 as well as excitonic splitting of the S1 states due to coupling of neighboring chromophores in the crystalline arrangement 5 . The packing arrangement and electronic coupling or wavefunction overlap is what allows a constructive 30 interaction between tetracene chromophores within the crystal. This idea will be expanded on later in this chapter. Another common idea, due to the above morphological effects, is that crystallinity is better for singlet fission 21,22 . That is ordered and optimal packing will result in the most efficient singlet fission processes over large crystal domains. However, our group has demonstrated that in amorphous films of vapor deposited 5,12-diphenyltetracene (5,12-DPT) efficient singlet fission was observed 23 . Singlet fission occurred in these disordered systems with a time constant of 800 femtoseconds (fs), which is several orders of magnitude faster than singlet fission in crystalline and polycrystalline tetracene 20 . Upon excitation of the disordered film, two processes occur. Singlet fission and singlet exciton diffusion. That is some of the DPT molecules were prearranged within the distribution of randomly oriented systems, to undergo rapid singlet fission. Some of the population, however, proceeded to diffuse to find a chromophore which would constructively couple towards singlet fission, or decayed back to the ground. Though some fraction of the singlet fission activity is diffusion limited in the amorphous DPT systems, the singlet fission rate is almost an order of magnitude faster than what has been reported for crystalline systems and, even more interesting, is that diffusion or formation of the uncoupled triplets is highly favorable in the amorphous systems relative to crystalline systems. For amorphous DPT there was no distinct spectral component observed for a multiple exciton (correlated triplet) state, as triplet dissociation is assumed to be rapid after forming the ME state. This is similar to what is observed in tetracene films and crystals at room temperature. In tetracene, the ME state is only distinguishable at lower temperatures, where triplet dissociation is slow compared to the ME formation. That is, the first step in singlet fission is temperature independent, that is S0S1 → 1 (T1T1), while dissociation of the ME state to the uncorrelated triplets 31 is temperature dependent 19,24 . The ME state has been observed in molecular dimers in dilute solution, where dissociation of a triplet pair cannot take place 25,26 . To identify the ME state, that is to verify singlet fission is taking place, these dimers were deposited as a neat film, where dissociation of the ME state into a pair of separated triplets was observed 26 ; once again, triplet dissociation was faster than the ME state formation, therefore resulting in an undetectable ME intermediate population. Figure 2.2: Diphenyltetracene systems of interest Herein we will investigate the effects of morphology on two isomers of DPT (Figure 2.2). The systems to be studied are tetracene derivatives with phenyl substituents in the 5, 11 and 5, 12 positions, giving two isomers of diphenyltetracene (DPT). We will compare the relative effects of each system in their crystalline phase, to understand how molecular arrangement impacts singlet fission. This will be understood based on previous concepts proposed my Michl and Smith 5 , where we can think of orbital and wavefunction overlap and crystal arrangements to understand constructive versus deconstructive interactions coupling the initial ground (S0) and excited (S1) chromophores to the correlated triplet ( 1 (T1T1)). Following we will discuss the differences, if any, between the DPT isomers as it relates to singlet fission in their amorphous, vapor deposited films. 32 Finally, we will discuss the effects of spin casting versus vapor deposition on singlet fission in 5,12-DPT. 2. Experimental 2.1 Sample Preparation: Synthesis of 5,11-DPT and 5,12-DPT followed procedures previously published 23,27,28 . Samples were prepared by vapor deposition of 5,11-DPT and 5,12-DPT on 1/16” thick quartz substrate to ~100 nm thickness, where the OD at 500 nm ~0.3 OD. The films were epoxy sealed between two quartz substrates to prevent exposure to oxygen which would quench triplet formation. DPT crystals were grown as described previously 29 , in brief, the crystals were grown through vacuum train sublimation in a three-zone tube furnace. The chamber is evacuated to 10 -7 Torr and the sample is placed in zone 1 set to the highest temperature of 250º C, where the sample sublimes to form molecular vapor. It is transported to the next zone, zone 2 set to 210º C, where the sample condenses. Slow vapor diffusion and condensation allows for direct crystal grown from the molecular vapor. This resulted in relatively large crystals (about 0.5 – 1 mm 3 ). 2.2 Steady-state characterization Steady-state absorption experiments were obtained with a Cary 50 UV-vis spectrometer in transmission geometry for thin films and solution. For dispersed emission of the DPT thin films samples were excited at 435 nm and 440 nm. The slits for both the excitation and emission were set to a bandpass of 3 nm for both amorphous films of 5,12-DPT and 5,11-DPT, as well as the systems dissolved in solution. All step sizes were 0.5 nm with integration times of 0.5 seconds. 33 2.3 Time resolved Photoluminescence Fluorescence lifetime measurements were carried out on the vapor deposited films of 5,11-DPT and 5,12-DPT films and crystals where the measurements on crystalline DPT were done under nitrogen or in a purged 1 mm cuvette. The samples were excited with 500 nm generated from an optical parametric amplifier (Coherent OPA 9450) which is pumped by a 250 kHz regenerative Ti: sapphire amplifier (Coherent RegA 9050). Time resolved emission data were collected using a time correlated singlet photon counting (TCSPC) set up (Becker and Hickl SPC 630) which was triggered with respect to the RegA 9050 amplifier pulse train. Samples were collected at 530 nm for selective detection of S1 emission. Emitted photons were detected at right angle through a 0.125 mm double monochromator (Digikrom CM 112) and detected using a Hamamatsu R3809U-50 photon multiplier tube with a ~20 ps instrument response time. 2.4 Femtosecond transient absorption Transient absorption experiments were conducted for amorphous thin film samples of DPT. Pump and probe pulses were generated from the output of a 1 kHz Ti:sapphire regenerative amplifier (Coherent Legend Elite, 3.5 mJ, 35 fs). About 10 % of the amplified 800 nm was used to pump a type II OPA (Spectra physics OPA-800C) generating excitation pulses centered at 500 nm with a ~ 10 nm bandwidth. The white light continuum probe (320 – 950 nm) was generated by focusing 800 nm onto a rotating 2mm thick CaF2 disk. The white light was sent to a pair of off axis aluminum coated parabolic mirrors where the white light was collimated then focused onto the sample while a CaF2 lense focuses the pump onto the sample. After the sample the white light is collimated and refocused through a slit into a spectrograph (Oriel MS1271) disperses the white- light continuum onto a 256 pixel silicon array (Hamamatsu). The pump energies were set to 34 ~250 nJ with a spot size of 350 µm. To prevent photodamage, the sample was slowly translated perpendicular to the probe. Transient absorption measurements were attempted from DPT crystals in a reflection geometry; however, the large scatter overwhelmed the transient signals. On the other hand, the crystals were adequate for study by time resolved photoluminescence (TRPL). 3. Crystalline DPT 3.1 Time Resolved Photoluminescence One of the earliest methods for detecting singlet fission was by time-resolved photoluminescence (TRPL) 30 where a key indicator of singlet fission is the observation of a delayed fluorescence in the singlet emission. As singlet fission occurs, a fast, initial decay where nonradiative formation of the correlated triplet pair occurs. Either the initially formed correlated triplet pair, or those later formed by geminate recombination, can fuse back to the original singlet excited state and this results in delayed fluorescence. An approximate equilibrium is established between the two states (Eqution 1) and this will determine the relative magnitude of the delayed fluorescence component compared to the fast decaying component. [𝑆 1 ]⇌[𝑇 1 + 𝑇 1 ] (1) Δ𝐺 𝑆𝐹 =−𝑅𝑇𝑙𝑛 ( [𝑇 1 +𝑇 1 ] [𝑆 1 ] )==−𝑅𝑇𝑙𝑛 ( 𝐴 𝑑𝑒𝑙𝑎𝑦𝑒𝑑 𝐴 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 ) (2) It has been shown in crystalline or polycrystalline tetracene that singlet fission occurs. This is attributed to the manner in which the tetracene moieties are stacked. Part of this excited state interaction is indicated by the ground state absorption where a Davydov split state is observed. This is marked by the large red shift and lack of structure in the steady-state absorption, indicative 35 of J-aggregate like behavior. The slipped stack geometry of the tetracenes is such that the electronic coupling between an excited and ground state geometry is constructive towards singlet fission 5 . Here, crystalline 5,11-DPT and 5,12-DPT are compared as to show the effect of differences in stacking geometries with respect to singlet fission. Preferential geometry will be shown to enable singlet fission whereas a different arrangement inhibits singlet fission. It has been proposed by Figure 2.3:Time resolved photoluminescence measurements of 5,11-DPT (a) and 5,12-DPT (b) as crystals and dissolved in chloroform (CHCl 3). From the fit curve, 5,11-DPT gives an initial amplitude of 1 and a final amplitude for the delayed fluorescence of 3.8 ×10 -4 where these amplitudes can be used to determine the ΔG SF for crystalline 5,11-DPT using Equation 2. Decay traces for the photoluminescence measurements for the different DPT systems provide an interesting contrast. If we compare the TRPL trace of 5,12-DPT dissolved in chloroform and in its crystal form, the traces are nearly identical. That is the photoluminescence in both cases follow nearly single exponential decay. It has been proposed by Casanova that crystalline 5,12-DPT could undergo singlet fission with proper changes to the morphology of the stacked tetracenes or intramolecular motion 31 . The similarity in the TRPL decay traces between crystalline 5,12-DPT and 5,12-DPT dissolved in chloroform (CHCl3), however, imply that 5,12- DPT molecules in their crystal phase are overall excited to their S1 state and decay back to the ground without population of other excited states. However, in crystalline 5,11-DPT, we observe 36 a multiexponential decay in the PL. It is the observation of much faster primary decay of the singlet excited state lifetime along with delayed fluorescence that suggests that crystalline 5,11-DPT, unlike 5,12-DPT, undergoes singlet fission. We can use the final (Adelayed, Equation 2) and initial (Ainitial, Equation 2) amplitudes in the TRPL traces to get the change in energy towards singlet fission (ΔGSF) as these are indicative of the relative population or concentration of the triplet and singlet states (Equation 1 and 2). For the case of crystalline 5,11-DPT, the ΔGSF = -RTln(3.8×10 - 4 /1) ~200 meV, assuming the temperature is room temperature (T= 298 K). 3.2 Crystal Morphology Figure 2.4: Crystal structure of 5,11-DPT (left) and 5,12-DPT (right) indicating molecular packing. The boxed inset is the crystal structure of o-BETB, a molecular tetracene dimer which has been shown to rapidly form the 1 (T 1T 1) on the isolated molecule by singlet fission. Examining the crystal structures of each points to a possible explanation for this difference. Looking at the crystal structure of one of the sites, it is clear that the molecular packing in 5,12- DPT versus the 5,11-DPT is completely different (Figure 2.4). In the case of the 5,12-DPT system, the tetracene molecules are stacked almost right on top of one another, which, in accordance to Michl and Smith 5 , results in a deconstructive interaction of the chromophores towards singlet 37 fission. However, in 5,11-DPT the alignment of the tetracene chromophores is similar to that in alkynyl-phenyl bridged tetracenes (Figure 2.4) where the two alkynyltetracenes are linked at the ortho positions on the phenyl bridge (o-BETB). These o-BETB systems have been shown to undergo rapid singlet fission with a relatively long-lived correlated triplet pairs ( 1 (T1T1)) 25,26 . Similar to o-BETB molecular dimers, the tetracenes in crystalline 5,11-DPT are locked in a herringbone geometry that should also be constructive towards singlet fission. As it is a bulk crystal, subsequent steps to triplet dissociation and geminate recombination can take place as manifested by the delayed fluorescence in the TRPL measurements. 3.3 Conceptual description of crystal morphology towards singlet fission Constructive and deconstructive interactions towards singlet fission can be thought of in the following way, in accordance to Smith and Michl 5 . The probability of transitioning between the coupled ground and excited singlet chromophores (S0S1) to the correlated triplet pair 1 (T1T1) is described by the following equation <𝑆 1 𝑆 0 |ℋ 𝑒𝑙 | 1 (𝑇 1 𝑇 1 )> ∝ <𝑙 𝐴 𝑙 𝐵 | 𝑒 2 𝑟 12 ⁄ |ℎ 𝐵 𝑙 𝐴 > −<ℎ 𝐴 ℎ 𝐵 | 𝑒 2 𝑟 12 ⁄ |ℎ 𝐵 𝑙 𝐴 > Where S0 and S1 are the ground and excited singlet states of chromophores A and B, ℋ 𝑒𝑙 , is the electronic contribution from the Hamiltonian operator, e is the electronic charge and r12 is the distance between electrons 1 and 2 for molecules A and B in their respective electronic states. It can be seen here that the electronic Hamiltonian is approximated to only contain the electronic repulsion/attraction terms, as that is the coupling or off-diagonal components in the final matrix, which mathematically indicates the projection of the initial hBlA state to the final hBhA/lBlA after operation of ℋ 𝑒𝑙 . The HOMO, h, and LUMO, l, are localized on chromophore A and B where the probability of transitioning to the final correlated triplet from the initial S1S0 state can be described 38 as the difference between the probability of populating both LUMOs, lA and lB, and relaxing the system so both chromophores are in the ground, in other words populating both HOMOs, hA and hB. When the overlap of the chromophores is such that this difference is greater than zero, there will be significant probability of singlet fission. Figure 2.5: Illustrations of the orbital overlap of tetracene molecules in the 'Stacked', 'Slip-stacked' and 'Herringbone' geometries. The illustrated orbitals are based on DFT calculations of tetracene. For a pictorial description one can refer to Fig. 2.5 where different stacking geometries of these chromophores lead to different overlap of these HOMO and LUMO orbitals. In the case of the entirely overlapped stacked system (Fig. 2.5, ‘Stacked’) there is equal overlap of both the HOMO orbitals of A and B and the LUMO orbitals of A and B. In accordance to the above equation this results in a net zero probability of 1 (T1T1) state formation. That is there is as much attraction as there is repulsion of the orbitals, giving equal probability of transitioning to the excited state in both chromophores <𝑙 𝐴 𝑙 𝐵 | 𝑒 2 𝑟 12 ⁄ |ℎ 𝐵 𝑙 𝐴 > and ground state of both chromophores < 39 ℎ 𝐴 ℎ 𝐵 | 𝑒 2 𝑟 12 ⁄ |ℎ 𝐵 𝑙 𝐴 >, producing a net-zero interaction towards singlet fission. This first arrangement is like that of crystalline 5,12-DPT which corresponds to the lack of singlet fission observed. The second arrangement is the slip-stack geometry like what is observed in crystalline tetracene. This results in a greater overlap of the LUMO densities giving a net positive probability of singlet fission. The third and final orientation is like that of 5,11-DPT. In Figure 2.5, the relative overlap appears larger between the LUMOs of the tetracene chromophores than the HOMOs, though only slightly more, enough so that there is a positive coupling towards singlet fission. It has been suggested by Smith and Michl that the overlap of these orbitals requires a ‘slip’ along the transition dipole which couples the S0 to the S1 state. In the case of tetracene this lies along the short axis of the tetracene moieties. For the slip-stacked geometry observed in crystalline tetracene, there is an obvious offset in the direction of the transition moments relative to the completely overlapped orientation observed in 5,12-DPT (labeled ‘Stacked’ above). In the case of the herringbone structure, the transition moments are nearly orthogonal, which would initially imply that there is no overlap for constructive coupling towards singlet fission. However, the offset of the overlap, such that the first ring of one of the tetracenes is stacked over the second ring in the neighboring chromophore as well as the tetracene moieties having a slight angle between each other, i.e. they are not stacked where the tetracene planes are parallel, making it so the transition moments are not entirely orthogonal. This gives a constructive interaction such that the LUMO orbitals have more overlap than the HOMO orbitals resulting in a positive interaction towards singlet fission. By considering the structure-property relationship within the organic crystals, we can conceptually achieve why singlet fission is observed in one of the DPT isomers over the other. 40 This idea is a reinforcement of a simplified understanding of the constructive interactions of coupled chromophores towards singlet fission. This further enhances that orientation of the chromophores is an extremely important parameter when considering systems which can undergo singlet fission. 4. Amorphous Vapor Deposited Thin Films The amorphous 5,11-DPT and 5,12-DPT were done independently of the systems previously published by Roberts et. al 23 . 4.1 Time resolved photoluminescence Figure 2.6: Time resolved photoluminescence of 5,12-DPT and 5,11-DPT. To the right is an energy scheme of geminate recombination back to the excited singlet, where the relative differences in energy between the 5,11 and 5,12-DPT states are indicated by the intial and final photoluminescence amplitudes As discussed for the crystalline DPT systems, TRPL can be a good indicator of whether singlet fission is occurring by the observation of delayed fluorescence. Delayed fluorescence following a prompt decay of the TRPL experiments indicates geminate recombination of the 41 dissociated triplet pair back to the singlet excited state. In both 5,11-DPT and 5,12-DPT the photoluminescence gives three time constants (Figure 2.6), one for a prompt initial decay due to a fast loss of population in the S1 state, usually indicative a rapid population of a new state, a slightly slower time constant (hundreds of picoseconds) and finally a long delayed fluorescence, several nanoseconds. Both 5,11 and 5,12-DPT show a prompt decay in the S1 emission that is within the instrument limited response, which is responsible for ~20 - 30 % amplitude loss where this loss will be considered in more detail in the transient absorption section. The second temporal component decays with a time constant of hundreds of picoseconds. Within 5 ns 99% of the total PL intensity decayed for both 5,11-DPT and 5,12-DPT. However, the delayed fluorescence lifetimes for these seemingly similar systems is quite different. For 5,12-DPT the time constant associated with the triplet recombination is about 5.7 ns while in 5,11-DPT this time constant increases to almost 13 ns. The relative amplitudes are also different (Figure 2.6). As stated previously, the relative amplitudes of the initial and final trace are indicative of population or concentration as it relates to the formation of the singlet and triplet states allowing us to ultimately calculate the change in free energy between the initial singlet excitation and the separated triplets (ΔGSF). Qualitatively speaking, it can be seen that singlet fission, or rather the uncorrelated triplet state, is lower in energy relative to the singlet state in 5,12-DPT than 5,11-DPT (Figure 2.6, energy scheme). Assuming that T=298 K, the ΔGSF are 240 meV and 200 meV for 5,12-DPT and 5,11-DPT, respectively. Based on the TRPL measurements alone, singlet fission, or rather the formation of diffuse triplets is more downhill in the amorphous 5,12-DPT than both the crystalline and 5,11-DPT systems. However, the amorphous 5,11-DPT appears to have about the same change in free energy regardless of whether it is crystalline or amorphous. 42 4.2 Steady-state Photophysics Figure 2.7: Steady-state absorption and emission of 5,12-DPT (a) and 5,11-DPT (b) in both film and dissolved in chloroform (CHCl 3). The above steady-state absorption and emission (Figure 2.7) are of the disordered 5,12 and 5,11-DPT films prepared by vapor deposition as well as the isomers dissolved in solution. The absorption of dissolved 5,12-DPT and 5,11-DPT are similar to their absorption spectra as amorphous films, where they both show a vibrational progression as is observed in solution. The only difference is that both DPT films give a red-shifted absorption relative to solution, where the red-shift is about 380 cm -1 and 400 cm -1 for 5,12-DPT and 5,11-DPT respectively. The steady- state absorption suggests that the films are disordered as there appears to be no ground state coupling of the DPT molecules. Tetracene films have shown evidence of crystal packing, where the stacking of the tetracene molecules results in a J-aggregate formation observed in the steady- state absorption, causing the absorption spectrum to be strongly red-shifted and lose vibrational character, like that of crystalline tetracene. This suggests that the films of DPT are amorphous and ground state coupling is minimal. The most obvious difference between the isomers really comes from the emission spectra. In 5,12-DPT the 0-0 vibrational band in both the absorption and emission completely overlay, just 43 like 5,12-DPT dissolved in solution (Figure 2.7a). Though the emission spectrum of 5,12-DPT is not a complete mirror images of the absorption, the observation of three bands in the emission spectrum indicates that the first excited states are for the most part localized on a single DPT chromophore. However, the observed broadening, especially to the red end of the emission spectrum, indicates that there may be some degree of coupling between neighboring DPT molecules in their excited states. The broadening of the emission line shape may be due to some aggregation of the sample or potential inhomogeneous broadening caused by the film environment. The emission spectrum of 5,11-DPT displays a very large stokes shift and a large degree of broadening (Figure 2.7b). 5,11-DPT has a tendency to crystallize, so there may be some crystal packing, which causes aggregate like behavior in the emission, giving a broadened red-shifted emission. Considering the 5,11-DPT molecules stack in a herringbone geometry (Figure 2.4), it would be unsurprising that the ground state absorption would not indicate crystal packing since, as discussed in the above section, the transition moments of the S0 →S1 state are nearly orthogonal. 4.3 Transient Absorption Though time resolved photoluminescence allows us to observe the presence of singlet fission, extracting the mechanisms of singlet fission requires faster timescales. Since our TRPL set up has an instrument limited response of ~20 ps, if we want to look at processes which occur on sub picosecond timescales, femtosecond transient absorption is required. We know from previous work done by our group that in the 5,12-DPT amorphous films, singlet fission occurs with a time constant of 800 fs, where ME state formation is limiting, but the triplet diffusion process is much faster than the instrument limitation, as the only spectral observables were the initial S1 absorption trace and the final triplet trace. Here we will compare and see if this holds true for the 5,11-DPT films, or if potential crystal packing enables an observable ME state, that is decreasing the rate of 44 triplet diffusion. Or potentially, the placement of the phenyl substituents on the tetracene changes the electronic structure of the initial singlet changing coupling towards singlet fission within the amorphous film configuration. Figure 2.8: Time slices of the transient aborption spectral traces for 5,12-DPT (a) and 5,11-DPT (b). Normalized wavelength traces of the S 1 absorption (c) and T 1 absorption (d) for 5,12-DPT and 5,11-DPT. Both 5,11-DPT and 5,12-DPT give similar transient behavior (Figure 2.8 a&b). In both cases the excited state absorption (ESA) attributed to the S1-Sn absorption (380 – 450 nm and 570 – 680 nm) decays within the first 10 ps. Within 5 ps there is a loss in GSB centered around 500 nm where a new ESA feature is observed. The final trace shown at about 750 ps is a structured absorption which matches the ESA of the triplet state found by triplet sensitization. The line shape gives an absorption from about 450 – 530 nm, where some of the structure or vibrational character 45 arises from the superimposed GSB, like what was observed in the steady-state absorption experiments. The final trace for bot 5,11 and 5,12-DPT match the sensitized trace for the triplet absorption 23,32 , implying that within 10s – 100s of picoseconds separated triplets within both DPT films are observed. That is, for both 5,11-DPT and 5,12-DPT, there was no observation of the correlated triplet pair, the only distinguishable spectral traces match that for the S1 trace and independent T1. Figure 2.9: Wavelength slices of the S 1 and T 1 absorption for 5,12-DPT (a) and 5,11-DPT (b). Δ Absorption traces have been changed to indicate a population density in terms of number of excitations per μm 3 . The relative populations of the S 1 and T 1 indicate singlet fission efficiencies in both films. However, there are some differences between the two systems. Looking at the normalized triplet and singlet traces of the 5,11-DPT and 5,12-DPT systems, it is clear that 5,11-DPT has a slower singlet decay and triplet rise. Examining the relative populations of the triplet states relative to the populations of the singlet states (Figure 2.9), it can be seen that 5,12-DPT is more efficient than 5,11-DPT at singlet fission. That is 5,12-DPT gives more triplet excitations at later times than 5,11-DPT per singlet excitation. In principle, if both systems are entirely amorphous, their efficiencies should be similar as one could approximate the interactions of the singlet excitons to be primarily dependent on the electronic structure of the tetracene moieties and diffusion of said 46 excitons. However, there is an issue of differences in the emission and the TRPL data between samples of the DPT isomers. The TRPL traces indicate that the 5,11-DPT had a smaller ΔGSF than 5,12-DPT, which is consistent with the observation of less triplets in 5,11-DPT than 5,12-DPT per singlet. However, if we look at the steady-state spectra as it relates to the two films, it can be seen that though both show nearly independent ground state configurations, from the steady-state absorption, the excited state in 5,11-DPT appears to be more delocalized than that of 5,12-DPT, as the emission spectrum is more aggregate- or excimer-like. It should also be noted that the amorphous 5,11-DPT has a tendency to crystalize while the disordered 5,12-DPT was very stable. This can be sort of hinted at in the crystal structures as the stacked 5,12-DPT leads to a more frustrated crystal structure relative to 5,11-DPT. That is all to say that it could be the 5,11-DPT systems are somewhat amorphous with sites of crystallinity, this could lead to trapping of the initial singlet exciton within the 5,11-DPT film. Or, conversely, in the case of 5,11-DPT, singlet fission occurs preferentially within more crystalline arrangements. Regardless, the arguably more amorphous 5,12-DPT is more efficient or energetically favorable towards singlet fission compared to both the 5,11-DPT film and crystal. That is to say, that, in accordance to what has been previously published, there is no disadvantage in amorphous DPT towards singlet fission, in fact it singlet fission is about 100 times faster in amorphous DPT than crystalline tetracene. Though the population of initial singlets immediately undergoing singlet fission is only a fraction of the initially excited population where the other singlet fission processes are diffusion limited, once a constructive chromophore pair is found towards singlet fission, this process occurs very rapidly. 47 5. Spin Cast 5,12-DPT versus Vapor Deposited 5,12-DPT Figure 2.10: a) Transient absorption spectra of spincast 5,12-DPT. b) Excitation density of the singlet and triplets as a function of time for both the spin cast (solid line) and vapor deposited (dashed line) 5,12-DPT. c)transient spectra of both spincast and vapor deposited 5,12-DPT at a delay time of 200 fs, where only singlets are present. D)Transient spectra of vapor deposited and spincast 5,12-DPT at a time delay of 750 ps, where the population is primarily triplet (< 1 % are singlet) One advantage that was shown in the above section where efficient singlet fission is observed in an amorphous film of DPT, is the implication that disorder does not inhibit singlet fission. In fact, here and in previous work by our group indicates that singlet fission can readily occur, even in films where the molecules are not initially arranged for optimal interaction towards singlet fission, as in crystalline systems. One thing that has been proposed as an advantage for organic photovoltaic devices versus inorganic devices is the idea that they can operate efficiently 48 with low temperature deposition. In this section we test this idea by comparing the vapor deposited thin film of 5,12-DPT to a spincast thin film. Vapor deposited thin films, as the name implies, requires deposition of systems at their sublimation temperatures, which, on an industrial scale, could be energetically and environmentally taxing. Spincasting on the other hand simply takes a room temperature, dissolved solution where the solution is dispersed onto a substrate (in this case quartz) while the substrate is spinning. This will in turn create an inequal distribution of the DPT molecules on the substrate where various stacking or packing arrangements can occur. In the Figure 2.10, the vapor deposited and spincast 5,12-DPT are compared. The initial trace at 200 fs for both the spincast and vapor deposited sample give an identical trace indicating successful population of the S1 state (Figure 2.10 c). The traces at 750 ps, indicative of triplet population, are also identical between the vapor deposited and spincast samples (Figure 2.10 d). Thus, singlet fission is occurring, regardless of sample preparation. By comparing the relative number of excitations for the triplet and singlet traces, it can be seen that though the relative excitations are different, both samples give about the same number of triplets per singlet (Figure 2.10 b). This implies that regardless of sample preparation, for 5,12-DPT, singlet fission occurs with nearly the same efficiency. 6. Conclusions: The most interesting difference was actually in the observation of the singlet fission in one of the crystals of the isomers over the other. In crystalline 5,12-DPT the TRPL decay is nearly identical to that of 5,12-DPT dissolved in solution, indicating no singlet fission is occurring. 5,11- DPT, however, gave a multiexponential decay where a prompt decay followed by delayed fluorescence is observed, indicating that singlet fission is an allowed process. Following this we the molecular packing was related to this behavior. The directly stacked arrangement of crystalline 49 5,12-DPT gave a large degree of coupling between DPT molecules, where the interaction was equally attractive as repulsive resulting in a deconstructive interaction towards singlet fission. 5,11-DPT crystals, however, packed in a herringbone arrangement, where the coupling was not too strong or too week, giving more attractive character. This resulted in a constructive process towards singlet fission. In the amorphous systems it was also indicated that there is an advantage to a large distribution of chromophore arrangements as singlet fission readily and efficiently occurred in 5,12-DPT. In the ‘amorphous’ films in both 5,11-DPT and 5,12-DPT showed singlet fission occurring, but 5,11-DPT was slower and resulted in significantly fewer triplets per singlet excitation relative to 5,12-DPT. This can be attributed to a couple of effects. One being that, in accordance to the TRPL measurements, 5,11-DPT has a smaller ΔGSF compared to 5,12-DPT, which would result in a decreased population and rate associated with triplet formation. However, as implied by the steady-state emission spectra, it could be that the 5,11-DPT films had some crystal packing where, since it was not entirely crystalline or entirely amorphous, acted to trap some of the singlet states as the excitons diffused. Different sample preparations were also compared between vapor deposited and spin cast films of 5,12-DPT. It was shown that the relative efficiencies of singlet fission in the spin cast and vapor deposited systems were very similar. Singlet fission was slightly slower in the spin cast films relative to the vapor deposited, but overall the number triplet excitations per singlet excitation were about the same. This is a promising indicator for low temperature sample preparation as it pertains to singlet fission in these films. 50 Chapter 2 Bibliography (1) Shockley, W.; Queisser, H. J. Detailed Balance Limit of Efficiency of P‐n Junction Solar Cells. J Appl Phys 1961, 32 (3), 510–519. https://doi.org/10.1063/1.1736034. (2) Beard, M. C.; Midgett, A. G.; Hanna, M. C.; Luther, J. M.; Hughes, B. K.; Nozik, A. J. Comparing Multiple Exciton Generation in Quantum Dots To Impact Ionization in Bulk Semiconductors: Implications for Enhancement of Solar Energy Conversion. Nano Lett 2010, 10 (8), 3019–3027. https://doi.org/10.1021/nl101490z. (3) Nozik, A. J. Multiple Exciton Generation in Semiconductor Quantum Dots. Chem Phys Lett 2008, 457 (1–3), 3–11. https://doi.org/10.1016/j.cplett.2008.03.094. (4) Prezhdo, O. V. Multiple Excitons and the Electron–Phonon Bottleneck in Semiconductor Quantum Dots: An Ab Initio Perspective. Chem Phys Lett 2008, 460 (1–3), 1–9. https://doi.org/10.1016/j.cplett.2008.03.099. (5) Smith, M. B.; Michl, J. Singlet Fission. Chem Rev 2010, 110 (11), 6891–6936. https://doi.org/10.1021/cr1002613. (6) Smith, M. B.; Michl, J. Recent Advances in Singlet Fission. Annu Rev Phys Chem 2013, 64 (1), 361–386. https://doi.org/10.1146/annurev-physchem-040412-110130. (7) Johnson, J. C.; Nozik, A. J.; Michl, J. The Role of Chromophore Coupling in Singlet Fission. Accounts Chem Res 2013, 46 (6), 1290–1299. https://doi.org/10.1021/ar300193r. (8) Ern, V. Anisotropy of Triplet Exciton Diffusion in Anthracene. Phys Rev Lett 1969, 22 (8), 343–345. https://doi.org/10.1103/physrevlett.22.343. (9) Irkhin, P.; Biaggio, I. Direct Imaging of Anisotropic Exciton Diffusion and Triplet Diffusion Length in Rubrene Single Crystals. Phys Rev Lett 2011, 107 (1), 017402. https://doi.org/10.1103/physrevlett.107.017402. (10) Fravventura, M. C.; Hwang, J.; Suijkerbuijk, J. W. A.; Erk, P.; Siebbeles, L. D. A.; Savenije, T. J. Determination of Singlet Exciton Diffusion Length in Thin Evaporated C 60 Films for Photovoltaics. J Phys Chem Lett 2012, 3 (17), 2367–2373. https://doi.org/10.1021/jz300820n. (11) Herz, J.; Buckup, T.; Paulus, F.; Engelhart, J.; Bunz, U. H. F.; Motzkus, M. Acceleration of Singlet Fission in an Aza-Derivative of TIPS-Pentacene. J Phys Chem Lett 2014, 5 (14), 2425– 2430. https://doi.org/10.1021/jz501102r. (12) Zimmerman, P. M.; Zhang, Z.; Musgrave, C. B. Singlet Fission in Pentacene through Multi- Exciton Quantum States. Nat Chem 2010, 2 (8), 648–652. https://doi.org/10.1038/nchem.694. 51 (13) Johnson, J. C.; Nozik, A. J.; Michl, J. The Role of Chromophore Coupling in Singlet Fission. Accounts Chem Res 2013, 46 (6), 1290–1299. https://doi.org/10.1021/ar300193r. (14) Gilligan, A. T.; Miller, E. G.; Sammakia, T.; Damrauer, N. H. Using Structurally Well- Defined Norbornyl-Bridged Acene Dimers to Map a Mechanistic Landscape for Correlated Triplet Formation in Singlet Fission. J Am Chem Soc 2019, 141 (14), 5961–5971. https://doi.org/10.1021/jacs.9b00904. (15) Zimmerman, P. M.; Bell, F.; Casanova, D.; Head-Gordon, M. Mechanism for Singlet Fission in Pentacene and Tetracene: From Single Exciton to Two Triplets. J Am Chem Soc 2011, 133 (49), 19944–19952. https://doi.org/10.1021/ja208431r. (16) Kolomeisky, A. B.; Feng, X.; Krylov, A. I. A Simple Kinetic Model for Singlet Fission: A Role of Electronic and Entropic Contributions to Macroscopic Rates. J Phys Chem C 2014, 118 (10), 5188–5195. https://doi.org/10.1021/jp4128176. (17) Chan, W.-L.; Ligges, M.; Zhu, X.-Y. The Energy Barrier in Singlet Fission Can Be Overcome through Coherent Coupling and Entropic Gain. Nat Chem 2012, 4 (10), 840–845. https://doi.org/10.1038/nchem.1436. (18) Grumstrup, E. M.; Johnson, J. C.; Damrauer, N. H. Enhanced Triplet Formation in Polycrystalline Tetracene Films by Femtosecond Optical-Pulse Shaping. Phys Rev Lett 2010, 105 (25), 257403. https://doi.org/10.1103/physrevlett.105.257403. (19) Burdett, J. J.; Gosztola, D.; Bardeen, C. J. The Dependence of Singlet Exciton Relaxation on Excitation Density and Temperature in Polycrystalline Tetracene Thin Films: Kinetic Evidence for a Dark Intermediate State and Implications for Singlet Fission. J Chem Phys 2011, 135 (21), 214508. https://doi.org/10.1063/1.3664630. (20) Piland, G. B.; Bardeen, C. J. How Morphology Affects Singlet Fission in Crystalline Tetracene. J Phys Chem Lett 2015, 6 (10), 1841–1846. https://doi.org/10.1021/acs.jpclett.5b00569. (21) Piland, G. B.; Bardeen, C. J. How Morphology Affects Singlet Fission in Crystalline Tetracene. J Phys Chem Lett 2015, 6 (10), 1841–1846. https://doi.org/10.1021/acs.jpclett.5b00569. (22) Thorsmølle, V. K.; Averitt, R. D.; Demsar, J.; Smith, D. L.; Tretiak, S.; Martin, R. L.; Chi, X.; Crone, B. K.; Ramirez, A. P.; Taylor, A. J. Morphology Effectively Controls Singlet-Triplet Exciton Relaxation and Charge Transport in Organic Semiconductors. Phys Rev Lett 2009, 102 (1), 017401. https://doi.org/10.1103/physrevlett.102.017401. (23) Roberts, S. T.; McAnally, R. E.; Mastron, J. N.; Webber, D. H.; Whited, M. T.; Brutchey, R. L.; Thompson, M. E.; Bradforth, S. E. Efficient Singlet Fission Discovered in a Disordered Acene Film. J Am Chem Soc 2012, 134 (14), 6388–6400. https://doi.org/10.1021/ja300504t. 52 (24) Wilson, M. W. B.; Rao, A.; Johnson, K.; Gélinas, S.; Pietro, R. di; Clark, J.; Friend, R. H. Temperature-Independent Singlet Exciton Fission in Tetracene. J Am Chem Soc 2013, 135 (44), 16680–16688. https://doi.org/10.1021/ja408854u. (25) Korovina, N. V.; Joy, J.; Feng, X.; Feltenberger, C.; Krylov, A. I.; Bradforth, S. E.; Thompson, M. E. Linker-Dependent Singlet Fission in Tetracene Dimers. J Am Chem Soc 2018, 140 (32), 10179–10190. https://doi.org/10.1021/jacs.8b04401. (26) Korovina, N. V.; Das, S.; Nett, Z.; Feng, X.; Joy, J.; Haiges, R.; Krylov, A. I.; Bradforth, S. E.; Thompson, M. E. Singlet Fission in a Covalently Linked Cofacial Alkynyltetracene Dimer. J Am Chem Soc 2016, 138 (2), 617–627. https://doi.org/10.1021/jacs.5b10550. (27) Anthony, J. E.; Eaton, D. L.; Parkin, S. R. A Road Map to Stable, Soluble, Easily Crystallized Pentacene Derivatives. Org Lett 2002, 4 (1), 15–18. https://doi.org/10.1021/ol0167356. (28) Barlier, V. S.; Schlenker, C. W.; Chin, S. W.; Thompson, M. E. Acetylide-Bridged Tetracene Dimers. Chem Commun 2011, 47 (13), 3754–3756. https://doi.org/10.1039/c0cc05164k. (29) McAnally, R. E.; Bender, J. A.; Estergreen, L.; Haiges, R.; Bradforth, S. E.; Dawlaty, J. M.; Roberts, S. T.; Rury, A. S. Defects Cause Subgap Luminescence from a Crystalline Tetracene Derivative. J Phys Chem Lett 2017, 8 (24), 5993–6001. https://doi.org/10.1021/acs.jpclett.7b02718. (30) Singh, S.; Jones, W. J.; Siebrand, W.; Stoicheff, B. P.; Schneider, W. G. Laser Generation of Excitons and Fluorescence in Anthracene Crystals. J Chem Phys 1965, 42 (1), 330–342. https://doi.org/10.1063/1.1695695. (31) Casanova, D. Electronic Structure Study of Singlet Fission in Tetracene Derivatives. J Chem Theory Comput 2013, 10 (1), 324–334. https://doi.org/10.1021/ct4007635. (32) Burgdorff, C.; Kircher, T.; Löhmannsröben, H.-G. Photophysical Properties of Tetracene Derivatives in Solution. Spectrochimica Acta Part Mol Spectrosc 1988, 44 (11), 1137–1141. https://doi.org/10.1016/0584-8539(88)80084-9. 53 Chapter 3: Controlling Symmetry Breaking Charge Transfer in BODIPY pairs 1. Introduction In nature, several organisms employ solar energy towards biosynthesis. Photosystem II (PS II) in plants 1–3 and the photosystems of Rhodobacter sphaeroides 2,4–7 and Rhodopseudomonas viridis 7,8 are some of the most well-studied photosynthetic systems for converting solar energy into chemical fuels 1–3 . These systems contain light harvesting (LHC) and reaction centers (RC) that work in tandem to synthesize adenosine triphosphate (ATP) 1 from optical excitations created by light by using them to drive unidirectional charge transfer 1,2 . Within PS II’s RC resides a chlorophyll dimer surrounded by a pair of accessory chlorophylls and pheophytins. The multiplex is held in a C2 symmetry by the protein frame, though only one side is active in electron transfer 1 . This central dimer is typically referred to as a ‘special pair’ as the two cofacial chlorophylls use energy delivered to the RC from the LHC to drive charge separation. Generation of ATP by RCs is incredibly efficient. In plants, this process occurs with 90% thermodynamic efficiency 1 while bacteria achieve 100% quantum efficiency for charge transfer per photon absorbed 9,10 . Their impressive ability to separate charge suggests photosynthetic RCs can serve as a functional model for artificial systems that harvest light to produce solar fuels. A RC’s special pair uses its environmental asymmetry to enable light-driven charge transfer, breaking the pair’s symmetry. As this charge transfer produces an electron and hole sufficiently uncoupled as to behave independently, it is referred to as symmetry breaking charge separation (SBCS). Synthetic approaches have been used to explore SBCS to further our understanding of this process and advance our ability to create systems that readily separate charge 9,11–14 . Several asymmetric donor-acceptor (D-A) systems have been made where electron and hole transfer are 54 energetically dictated to be unidirectional (D-A + h → + D-A - ) 11,15–20 . Such intramolecular charge Figure 3.1: (a) Intramolecular charge transfer (ICT), where the “donor” chromophore is initially excited (D*) followed by electron transfer to an acceptor (A) leaving a positive donor (D + ). (b) Symmetry breaking charge transfer (SBCT) between a set of identical chromophores (Ch) such as anthracene. In weakly coupled dimers, light absorption generally excites one of the dimer pairs (Ch*), producing a localized excited state (LE) that undergoes SBCT to form a radical cation (Ch + ) and radical anion (Ch - ). transfer (ICT) systems generally employ differences in redox properties of the electron donor and acceptor to conduct electrons and holes along distinct, unidirectional paths (Figure 3.1 a). However, ICT has also been achieved in entirely symmetric molecular dimers (Figure 3.1 b). Weakly coupled dimers typically have absorption lineshapes similar to that of a single chromophore (Ch) of the pair molecule, suggesting the excited state formed by light absorption is localized to single member of the pair (localized excited state, LE) rather than being spread over the full dimer. On their own, these systems possess no energetic driving force for charge separation. Rather, fluctuations of the local solvent environment surrounding the dimer create an asymmetry of its units that can drive separation and stabilize the resulting electron and hole. In 55 this sense, these symmetric systems are not unlike the special pair, wherein the protein scaffold holding the pair creates an asymmetric energy gradient that separates charge. We refer to this type of ICT in symmetric dimers as symmetry breaking charge transfer (SBCT). While related, SBCT and SBCS are distinguished experimentally based on the emission properties of the charge transfer state produced. 20,24 SBCT yields an emissive charge transfer state while SBCS produces a charge-separated state that decays nonradiatively. This distinction partially reflects the degree of coupling between the spatially separated electron and hole. SBCS generally produces electrons and holes that are fully decoupled and behave independently while SBCT produces electron-hole pairs that remain weakly coupled and can be thought of as a Wannier-Mott or charge transfer exciton. Herein, SBCS will be treated throughout this review as a type of SBCT, where the resulting charges are in the “decoupled” limit. Here, we explore how the formation rate and longevity of SBCT states created between dipyrrin dimers is dictated by their structure. We briefly overview work on a UV-light absorbing model system, 9,9’-bianthryl (BA), where nuclear motion and solvent fluctuations have been characterized to create a mechanistic picture for SBCT. We then use this work to frame studies by our group exploring SBCT in visible-light absorbing dipyrrin dimers. The impact of both interchromophore separation and chromophore linker rigidity on SBCT will be discussed as well as the population of stable intermediates along the SBCT pathway. The outcome of our work is a set of general guidelines for rapid formation of long-lived SBCT states needed for solar fuel and photocatalytic applications. 2. A Brief Review of SBCT in 9,9’-Bianthryl Among compounds that undergo SBCT, 9,9'-binathryl (BA) is perhaps the most well- studied. Emission spectra of BA show a strong solvent dependence indicative of a CT state 18,21–23 56 that can be viewed via femtosecond transient absorption (TA) 18,22 and time-resolved photoluminescence spectroscopy 23 . In nonpolar solvents emission spectra of BA are broadened and do not resemble their corresponding absorption lineshapes. This broadening has been attributed to torsion about the dihedral angle of the anthracene planes, 18,21 a motion observed with high-resolution TA spectroscopy 24 . Flash-photolysis time-resolved microwave conductivity experiments of BA indicate it possesses a significant static dipole moment in its excited state when dissolved in nonpolar solvents 25 even though anthracene, which comprises BA, lacks a dipole moment in its excited S1 state. BA’s dipole in its LE state has been attributed to it containing charge resonance contributions (Ch + -Ch - and Ch - -Ch + ) whose degeneracy is lifted by intramolecular torsion and solvent fluctuations, producing a net dipole moment in BA 25 . Fluorescence upconversion experiments coupled to molecular dynamics simulations have also explored the nature of these stabilizing solvent fluctuations, showing that solvent molecules modulate SBCT by rotating to create local electric fields. 26 Once formed, the dipole moment associated with the SBCT state stabilizes these solvent arrangements, making reformation of the LE state unfavorable. In addition, near-infrared pump-probe anisotropy experiments 27 coupled to time-resolved microwave conductivity experiments 25 suggest formation of a unique excited state prior to SBCT that is distinct from the LE state. In nonpolar solvents, BA does not evolve beyond population of this intermediate state while in polar media SBCT continues to completion. 30,31 . This intermediate has been denoted as a partial charge transfer (PCT) state and is comprised of a charge transfer state that is strongly coupled to the LE state. PCT formation is proposed to be driven by a transient, unequal interaction of BA’s anthracene moieties with their solvent environment. 27,28 Thus, the proposed mechanism for SBCT in BA is formation of an LE state that evolves to a PCT state from 57 which complete charge transfer occurs (LE → PCT → SBCT) 27,29 . Taken together, these findings highlight that a solvent can play an active role in SBCT by stabilizing key charge-separated intermediates via the characteristic structural fluctuations that it undergoes over time. 3. SBCT in Dipyrrin Dimers Dipyrrins and BODIPY dyes are known for their tunability, high molar absorptivities, long- lived excited states 30 , and quasi-reversible redox potentials 31 , making them great candidates for light harvesting applications that require SBCT. BODIPY dimers have also shown utility as triplet sensitizers for processes such as singlet oxygen generation as these states can be formed by charge recombination from ICT states 32,33 . Prior work from our group has shown dimers structurally similar to BA, prepared with either DIPYR (borondifluoro dipyridylmethene) 34 or BODIPY (borondifluoro dipyrromethene), 35,36 undergo either PCT or full SBCT depending on their structure and local environment. In these dimers the meso-positions of the chromophores can be bridged by a single bond as in bis-DIPYR and mnB or by a phenylene spacer as in mnPh (Figure 3.2). Importantly, structural changes, even to a minor degree, greatly impact the ability of a BODIPY Figure 3.2: Materials investigated in this report. 58 pair to interconvert between LE and SBCT states. This provides a chemical handle by which rates for SBCT, charge recombination, and formation of key states such as triplets, can be manipulated to meet the demands of a given application. For example, torsion of the chromophores about their connecting bridge can be sterically constrained to extend the lifetime of SBCT states (vide infra). Likewise, through-bond coupling, which can tune rates of SBCT via orbital overlap, can be adjusted through linking bridge design 35 or even outright eliminated via use of simple ions, such as Zn 2+ , as linkers 30,36,37 (zDIP, Figure 3.2). Prior work by our group investigated SBCT within m4B and m4Ph 35 . We found SBCT was significantly faster in m4B than m4Ph and ascribed this to differences in interchromophore spatial separation. However, oddly enough, the lifetime of the SBCT state in m4Ph was much shorter than in m4B which is counterintuitive. This was attributed to rotation of the phenylene bridge in m4Ph, which acts as a nonradiative decay path. Below, we test this hypothesis by exploring systems where steric constraints limit bridge rotation. Our group has also looked at a series of zDIPs where formation of an SBCT state was identified by spectroelectrochemistry and femtosecond-to-nanosecond TA 30 . Reported zDIPs were discussed in terms of their solvent-dependent excited state properties as a function of substitution on the DIPY ligands 38 . In another study zDIP2 was used to probe how the free energy difference between the SBCT and LE states changed with solvent dielectric to identify a solvent environment that gave no driving force for their interconversion (i.e. G = 0 for the LE ⇄ SBCT reaction), 36 which was found in a mixture of 20% tetrahydrofuran (THF) to 80% cyclohexane (CHX) with an ET(30) ~ 34. Herein, we expand upon this work by comparing the ground-state structure and 59 photoexcited dynamics of six different BODIPY dimers, three directly linked through their meso positions (Scheme 1: m4B, m8B, e4m8B) and three incorporating a phenylene bridge between the dimers (Scheme 1: m4Ph, m8Ph, e4m8Ph). Due to steric hindrance, the BODIPY units of mnB are held in near orthogonal configurations while the phenyl bridge in mnPh both increases dimer spatial separation and allows the BODIPY dyes to occupy a common plane. We also compare the dynamics of mnB and mnPh to analogous zinc-bridged dipyrrinato dimers (zDIP2/zDIP3) that afford a similar spatial separation between the BODIPY chromophores but orient them in a rigid, orthogonal geometry. 3.1 Interchromophore Separation Chromophore spatial separation within a molecular dimer is expected to play a major role in SBCT 17,39 . Electron transfer is promoted by orbital overlap, which has an exponential dependence on donor and acceptor separation 40–42 . Going from mnB to zDIP to mnPh, the distance separating the chromophores of each dimer increases from 1.5 Å (C – C meso bridge) to 3.4 Å (N – N distance) to 5.8 Å (C – Phenyl – C meso bridge). To begin, we compare m8B, zDIP2 and m8Ph to assess the impact of interchromophore spatial separation on SBCT in systems with a similar degree of steric encumbrance. 60 Figure 3.3: (a) Jablonskii diagram for SBCT showing the impact of solvent polarity. LE states produced by light can undergo SBCT (k SBCT) or return to the ground state via radiative (k fl) and nonradiative (k nr) decay channels. As solvent polarity is increased the SBCT state is stabilized, making its formation from the LE state exoergic. From the SBCT state, the system can decay radiatively (k rrec) and nonradiatively (k nrrec). (b-e) Absorption and emission spectra of (b) 1,3,5,7-tetramethyl-8-phenyl-BODIPY, (c) m 8B, (d) zDIP2, and (e) m 8Ph. Absorption spectra of all four systems are similar. Emission spectra of m 8B are solvatochromic, indicating an emissive SBCT state while zDIP2 and m 8Ph emission is dominated by their LE states. Steady-state absorption spectra of m8B, zDIP2 and m8Ph provide insight into interdimer electronic coupling (Figures 3.3 b-e). Spectra of m8Ph (Figure 3.3 e) are very similar to the absorption of the 1,3,5,7-tetramethyl-8-phenyl BODIPY monomer 35 (Figure 3.3 b), indicating minimal electronic coupling between m8Ph’s phenylene-bridged BODIPY chromophores. This behavior can be contrasted with that of m8B (Figure 3.3 c), which contains identical chromophores but lacks the central phenyl ring of m8Ph. Absorption spectra of m8B are bathochromically shifted with respect to that of m8Ph, suggesting enhanced interchromophore coupling within m8B stemming from the proximity of its BODIPY units. This spectral redshift is observed for all mnB systems (Figure 3.4) 61 Figure 3.4. Solvent dependent steady state absorption and emission of e 4m 8B (a) e 4m 8Ph (b) m 4B (c) and m 4Ph (d). Interestingly, absorption spectra of zDIP2 exhibit a slight blue shift (~50 meV) relative to the BODIPY monomer, m8Ph, and m8B (Figure 3.3d). This could stem from either complexation of its dipyrrin units to a zinc center in place of boron, or the mesityl substitution in the meso position. The former of these two scenarios seems most likely as the mesitylene group is rigorously orthogonal to the DIPY and thus is expected to negligibly contribute to DIPY’s valence orbitals. In contrast, binding of a Zn 2+ ion to a dipyrrin unit is expected to have less of an impact on its orbital structure than binding of a F2B + unit, which would favor an absorption lineshape change between zDIPY and BODIPY chromophores. While spectra of all three dimers display slight spectral shifts relative to their parent monomer, the absorption profile of each shows a vibronic progression resembling that of the monomer, suggesting the dimers possess an absorptive LE state 62 Table 1:Photophysical Properties of Dipyrrin Dimers. Solvent Absorption λmax1(λmax2) (nm) Emission λmax(nm) Φ Pl -ΔG 0 CT (eV) c m 4 B Toluene DCM MeCN 533 (544) 530 (543) 527 ( 541) 577 619 -- 0.62 a 0.087 a -- 0.36 a m 4 Ph Toluene DCM MeCN 515 513 508 515 513 508 0.095 a 0.069 a -- 0.18 a m 8 B Cyclohexane THF DCM MeCN 514 515 512 515 566 602 (646) 613 (647) 654 0.70 0.14 0.14 0.014 0.52 d m 8 Ph Toluene DCM MeCN 503 501 495 520 513 (520) 520 0.52 0.31 0.027 0.34 d e 4 m 8 B Toluene DCM MeCN 543 541 538 592 634 684 -- 0.42 d e 4 m 8 Ph Toluene DCM MeCN 525 526 520 546 544 538 -- 0.24 d zDIP2 Cyclohexane THF MeCN 493 493 490 506 507 508 0.66 b 0.09 b -- 0.30 b zDIP3 Cyclohexane THF DCM 489 489 488 507 511 509 (653) 0.16 b 0.15 b 0.02 b 0.29 b a) Quantum yield and redox potentials from Ref. 50. b) Quantum yield and redox potentials From Ref. 45. c) Using the equation ∆𝐺 𝐸𝑇 0 =𝐸 1/2 𝑜𝑥 −𝐸 1/2 𝑟𝑒𝑑 −𝐸 0,0 ∗ −𝑒 2 4𝜋 𝜀 0 𝜀 𝑠 𝑅 ⁄ where 𝐸 1/2 𝑜𝑥 and 𝐸 1/2 𝑟𝑒𝑑 are the oxidation and reduction potentials of the monomer BODIPY chromophores, e is the electronic charge, 𝜀 0 𝑎𝑛𝑑 𝜀 𝑠 are the permittivity of the vacuum and solvent dielectric, R is the center-to-center distance between Dipyrrins. The electrochemical potentials used here were recorded in DCM. d) Redox potentials taken from Ref 46. 63 akin in character to the excited monomer. Steady-state emission spectra were recorded to assess charge transfer in these systems (Figure 3.3 b, right). Population transfer from the LE state to an emissive, low-energy state will yield a large emission Stokes shift relative to the parent monomer. If this emissive state carries significant SBCT character, placing the dimer in a polar solvent will further redshift the emission (Figure 3.3 a, krrec) and broaden its linewidth relative to that of the monomer 17,18 . This behavior is observed for m8B (Figure 3.3 c), whose emission spectra show a progressive bathochromic shift with increasing solvent polarity, signaling formation of an emissive SBCT state. If SBCT instead results in charges that are decoupled from one another, recombination will likely proceed nonradiatively 17 (Figure 2a, knrrec), yielding a decrease in dimer photoluminescence quantum yield (ΦPL) with increasing stability of the SBCT state. zDIP2 and m8Ph display emission spectra that are near mirror images of the LE state absorption with small Stokes-shifts, suggesting emission originates from the LE state itself (Figure 2a, kfl/knr). This LE emission ΦPL decreases with increasing solvent polarity, indicating competition between LE emission and SBCT. By increasing the spatial separation of the BODIPY chromophores, coupling between the radical pair produced by SBCT weakens. This can allow population of other states, such as spin- triplet levels, in addition to nonradiative recombination (Figure 3.3 a, knrrec). Weak coupling among separated charges in the zinc- and phenyl-bridged systems could allow radical-pair or spin- orbit intersystem crossing to a triplet state while more strongly coupled systems may favor charge recombination to the ground state. Distinction of such scenarios requires tracking the excited state dynamics of each dimer, which lead us to employ femtosecond TA. Figures 3.5 a and 3.5 b plot TA spectra of m8Ph in toluene and acetonitrile (MeCN), 64 respectively, following excitation at 500 nm. At short time delays, a strong negative signal is seen at 490 nm that corresponds to the dimer’s ground state bleach (GSB) and signifies formation of excited molecules. A negative shoulder at ~525 nm denotes stimulated emission (SE) that indicates these molecules occupy the LE state. In toluene, TA spectra remain invariant as a function of time. In MeCN, however, the SE feature disappears while the GSB is unchanged. This disappearance is Figure 3.5: (a) TA spectra of m 8Ph in toluene where SBCT is not observed, giving only the LE state transient spectrum. (b) TA spectra of m 8Ph in acetonitrile, where m 8Ph shows growth of a new ESA feature indicating SBCT. (c) SBCT rate, k SBCT, versus interchromophore separation in acetonitrile. The solid line shows the expected exponential decrease in this rate with increasing interchromophore separation. (d) Decay rate of the SBCT state, k rec, (k rec = k rrec + k nrrec,, Fig. 2a) as a function of interchromophore separation. This shows an unexpected result as closely spaced dimers undergo charge recombination more slowly than further spaced dimers. Solid line is a guide to the eye. attributed to a rise in excited state absorption (ESA) superimposed with the SE, signaling a 65 transition from the LE state to a new excited state we have previously identified as the SBCT state 15,35 . Figures 3.5 c and 3.5 d summarize rates for formation and decay of the SBCT state, kSBCT and krec (krec = krrec + knrrec), extracted from TA data recorded in MeCN for several dyads with differing interchromophore spatial separation. In polar MeCN the SBCT state is highly stabilized and TA spectra are best fit by a model that postulates population transfer from the LE state to SBCT state followed by decay of the SBCT state (LE → SBCT → Ground State). As expected, kSBCT decreases with increasing interchromophore spatial separation (Figure 3.5 c). In MeCN, kSBCT for m4B, zDIP2 and m4Ph appears to exhibit an exponential dependence on interchromophore separation. However, m8B, zDIP3 and m8Ph do not follow this trend as m8B and zDIP3 have similar rates of SBCT, 0.8 ps and 1 ps, respectively, though their chromophores are separated by distances of 1.5 Å and 3.4 Å. Moving to less polar solvents, we see kSBCT slows for each dyad (Table 2), but the overall trend of kSBCT decreasing as interchromophore separation increases is maintained. This is expected as the rate of SBCT is dependent on both the ability of solvents to reorganize to stabilize the SBCT state as well as interchromophore orbital overlap, which is dependent on dimer spatial separation. Overall, we find kSBCT is fastest for mnB systems as its chromophores have the smallest spatial separation. We would also expect these dimers to show the fastest krec among our dimer series as the resulting electron and hole are held in the closest proximity to one another, aiding their recombination. Likewise, the mnB SBCT state lies closer to the ground state in energy relative to the SBCT state of mnPh dimers. This too should lead to a faster krec for mnB dimers according to the energy gap law. However, in all solvents increasing interchromophore separation showed behavior opposite to this expectation. As interchromophore separation is decreased, the lifetime of 66 the SBCT state increases. In polar MeCN, this effect is nonnegligible, as the m8B SBCT state exhibits a 1.8 ns lifetime that is 2× longer than zDIP2 (0.9 ns) and 9× longer than that of m8Ph (0.2 ns). Clearly, interchromophore spatial separation alone is not a sufficient indicator for the decay Table 2: Kinetic rates for transitions between excited states of Dipyrrin dimers determined by femtosecond TA. a Ref. 50, b Ref. 45, c Ref 24. Solvent 1/k PCT (1/k PCR ) (ps) 1/k SBCT (1/k BET ) (ps) 1/(k rrec +k nrrec ) (ps) 1/(k fl +k nr ) (ps) M4m Toluene DCM Acetonitrile 0.8 <0.2 -- 6.5 1 <0.2 6700 650 -- M8m Cyclohexane Toluene THF Acetonitrile 1.6 (4) 0.3 0.2 0.07 -- 7 1.4 0.8 10000 12000 9600 1900 -- M8mE Toluene DCM Acetonitrile 1.5 0.3 <0.15 -- 1.3 0.8 9400 9400 1700 -- M4Ph Toluene DCM Acetonitrile -- -- 18 4.5 -- -- 34 845 a 1000 a M8Ph Toluene DCM Acetonitrile -- -- 130 (140) 50 -- -- 200 4800 5100 M8PhmE Toluene DCM Acetonitrile -- -- 140 (240) 75 -- -- 280 9400 9400 zDIP2 Cyclohexane b Toluene b THF c Acetonitrile b -- -- 9 (16) 2 (27) 1.1 -- 4000 2200 900 4800 3900 4300 zDIP3 Cyclohexane b Toluene b Acetonitrile b -- -- 2.3 (11.2) 1.0 -- 2500 1400 1400 2100 rate of the SBCT state, highlighting additional factors play a role in its relaxation. The zinc-bridged 67 dipyrrin systems are a classic case of an SBCS system, where the resulting charges are decoupled as the zinc acts as a noninteracting spacer that prevents any through-bond linkage of the dipyrrins. zDIP2 and other zinc-bridged systems, show population of a triplet state following SBCT 30,43 , which is encompassed in krec, whereas m8Ph and m8B decay directly to the ground state. Decay to a triplet state close in energy to the SBCT state may rationalize why zDIP2 exhibits a faster than expected value for krec but does not explain the fast decay rate of m8Ph relative to the directly linked mnB dimers. Looking at other structural differences between m8Ph and m8B, we can identify the degree of torsion allowed about their linkage as a key difference as electron and hole recombination requires spatial overlap between the wavefunction of each charge carrier. The enhanced rigidity of the linking groups for both m8B and zDIP2 relative to m8Ph could reduce the potential for achieving such overlap, thereby slowing their rate of charge recombination relative to m8Ph, consistent with our observations. We test this hypothesis below by modifying the structure of each of these three dimers with steric groups that impact their ability to twist about their meso linkage. 3.2 Structural Rigidity and Chromophore Configuration We have identified a decrease in the longevity of the SBCT state with increasing chromophore separation that runs counter to expectation. Here, we discuss the role of molecular conformation and active nuclear modes on interchromophore wavefunction overlap, which modulates the rate of decay from the SBCT state (krrec & knrrec). Steric rigidity of meso-bridged BODIPY dimers was increased by substituting methyl groups that hinder dihedral rotation about the C – C bond tethering directly-linked dimers (m8B, e4m8B) and rotation about the bridge of phenylene-linked systems (m8Ph, e4m8Ph). By increasing steric hinderance of these otherwise active rotational modes, nonradiative decay driven by torsion, which increases the orbital overlap 68 of chromophores within a dimer, is expected to decrease, increasing the lifetime of the SBCT state. The structural rigidity of the directly-linked dimers is apparent in their steady-state spectra. m4B’s absorption spectra have a primary absorption peak (λmax1, Table 1) with a weaker red shoulder (λmax2, Table 1) in likeness to an H-aggregate. This reflects excitonic splitting from coupling of the BODIPY transition moments, signaling that its chromophores are rigidly Figure 3.6: (a) Growth and decay of SBCT state absorption for directly-linked dyads m 4M and m 8M and (b) phenyl-linked dyads m 4Ph and m 8Ph in acetonitrile. The addition of steric groups that hinder rotation slows SBCT and charge recombination for both sets of dyads. orthogonal (Figure 3.4 c). This contrasts with m8B and e4m8B, where only a single transition appears in absorption spectra (Figure 3.3 c and 3.4 a). With increasing steric hindrance about the meso bridge, directly-linked dimers show a decrease in their SBCT rate (Figure 3.6). For BA, it has been proposed that the dihedral angle of its anthracene planes acts as an active coordinate for 69 SBCT. BA’s LE state possesses minima along this coordinate at 60º and 120º and motion between these minima facilitates crossing to the SBCT state 18,19,26 . The reduced rate of SBCT in the more rigid BODIPY dimers implies surface crossing to the SBCT state is disfavored due to the increased intramolecular reorganization energy, making kSBCT nearly an order of magnitude slower in m8B and e4m8B (0.8 ps, MeCN) relative to m4B (0.1 ps, MeCN). Though kSBCT is slower in the rigid, directly-linked dyads, the lifetime of the SBCT state is substantially longer (Figure 3.6 a). In MeCN, krec is almost three times slower in m8B (1.9 ns) and e4m8B (1.7 ns) than m4B (0.65 ns). This can stem from hindrance of the torsional mode reducing nonradiative decay or hindering access to regions of the SBCT potential where crossing to the ground state potential energy surface is likely. Also, fixation of the BODIPY planes in an orthogonal geometry would decrease wavefunction overlap of the separated cation and anion, decreasing recombination. Similarly, for phenylene-bridged dimers, when the bridge is fixed to a rigid, orthogonal geometry that prevents coupling to its pendant BODIPY moieties, kSBCT is reduced (Figure 3.6 b). In MeCN, the time constants for SBCT in m4Ph and m8Ph differ by an order of magnitude, 4.5 ps and 50 ps, respectively. As with the directly-linked dimers, the lifetime of the SBCT state also increases with decreasing rotational freedom, from 34 ps to 200 ps moving from m4Ph to m8Ph. As rotation of a pendant phenylene group can facilitate nonradiative deactivation of the excited state in monomeric phenyl-BODIPY chromophores 35,44 it is perhaps unsurprising that increased rigidity results in a substantially longer SBCT state lifetime (Figure 3.6). However, even with fixation of the phenylene bridge, the phenylene-bridged dimers still have the shortest lifetime, by orders of magnitude, relative to the directly-linked and zinc-bridged dimers. Previously, we employed TD-DFT calculations to examine the HOMO and LUMO of m4B, 70 m4Ph and zDIP2 to better understand the influence of torsion on SBCT and charge recombination 35,36 . These calculations found the BODIPY HOMO has a node at the meso carbon while the LUMO has significant -orbital charge density at the meso position. The BODIPY ground state radical cation and radical anion SOMOs are similar to the HOMO and LUMO of the BODIPY dyad, respectively, and are depicted in Figure 5. In a rigidly held perpendicular dimer Figure 3.7: DFT calculations of the SOMOs of the BODIPY radical anion and radical cation using a B3LYP/6-31G(d,p) basis. Molecular orbitals are shown in a parallel (top) and orthogonal (bottom) configuration. The cation and anion orbitals resemble the HOMO and LUMO, respectively, of the neutral BODIPY monomer. linked at the meso position, such as m8B, this meso charge density has no net orbital overlap with the neighboring BODIPY unit. However, if a dimer undergoes excited state activated torsion, the 71 meso carbons can increase their orbital overlap, facilitating SBCT and, inadvertently, charge recombination (Figure 3.7). Thus, for systems where the BODIPY dyads are orthogonal, the rate of SBCT will be slower, but the longevity of this state will increase due to the small degree of orbital overlap of the anion and cation. In the case of the phenylene-bridged systems, a similar result was found when examining torsion of the phenylene bridge. Charge delocalization across the bridge facilitated by dimer planarization will give rapid SBCT and charge recombination. 3.3 SBCT Facilitated by Intermediate States In BA, under certain conditions SBCT is preceded by formation of an intermediate referred to as a partial charge transfer (PCT) state 27,29 . This state has been noted to give rise to BA’s excited-state static dipole in nonpolar solvents 25 and has been speculated to underlie lineshape differences between BA’s absorption and emission spectra in nonpolar solvents. As we highlight below, all of the directly-linked meso-bridged dimers reported here show population of an intermediate state in the course of SBCT, similar to BA (Figure 3.8). The characterization of this state has proven difficult as excited state signals can arise from several phenomena, including excimer formation. However, excimer states typically act to drive excited state decay to the ground state 45,46 whereas the intermediate we observe does not act to drive excited population decay, but rather enables formation of the SBCT state. States of mixed LE and charge transfer character are typically observed in molecular aggregates or strongly-coupled chromophores 47,48 and can serve as doorway states for full charge separation, 27 suggesting the intermediate state we observe corresponds to a PCT state. Evidence for an intermediate state involved in SBCT comes from rates for this process (Figure 3.5 b). The log of these rates shows a clear deviation from linearity with increasing dimer spatial separation, which goes against expectation if SBCT formation is a one-step process driven 72 by interdimer orbital overlap. TA spectra of the directly-linked dimers, m4B, m8B and e4m8B, provide more direct evidence of a PCT state in weakly polar to polar solvents as a clear spectral intermediate can be identified that evolves to the SBCT state. In solvents such as cyclohexane and toluene, where SBCT is inhibited, this intermediate is sustained for m8B and e4m8B, manifesting itself via a decay of stimulated emission (SE) attributed to the LE state followed by an increase in SE at lower energies (Figure 3.8 b). This new SE band matches the steady-state emission spectra of m8B and e4m8B (Figure 3.8 b and d) in cyclohexane and toluene, respectively. One could Figure 3.8: (a) Illustration of the LE to PCT to SBCT in directly-linked BODIPY dimers. (b) TA spectra of m 8B in cyclohexane with overlaid steady-state emission spectra of the 1,3,5,7-tetramethyl-8-phenyl- BODIPY monomer and m 8B dimer, highlighting development of a SE band that originates from the PCT state. (c) TA spectra of the BODIPY monomer in THF with overlaid absorption and steady-state emission spectra. (d) TA spectra of e 4m 8B in toluene with overlaid steady-state emission from the monomer 1,3,5,7- tetramethyl-2,6-ethyl-8-phenyl-BODIPY and e 4m 8B dimer, highlighting formation of the SE band duet to PCT formation. The monomer emission whas shifted by 20 nm to account for the changes in the ground to 73 excited transition energy when going from the 1,3,5,7-tetramethy-2,6-ethyl-8-phenyl-BODIPY monomer to the e 4m 8B dimer. presume this redshifted emission is due to a relaxation process caused by solvation or intramolecular nuclear relaxation on the LE excited state potential. However, the isosbestic nature of the loss and gain of the emission bands indicated by a time invariant point at 557 nm (Figure 3.8 b) signals population of a new state as opposed to adiabatic relaxation on a single potential surface. As observed for BA, this intermediate state appears to be a precursor to complete SBCT and therefore we refer to it as a PCT state. For m8B in cyclohexane and e4m8B in toluene, the PCT state lives for 10 and 9.4 ns, respectively, whereas for m4B in cyclohexane, steady-state emission implies the PCT state is not long-lived. This may be due in part to the structural rigidity of m8B and e4m8B whereas torsional relaxation of the less sterically constrained m4B may favor recombination to the LE state. However, placing m4B in solvents with moderately higher polarity than cyclohexane (toluene and DCM) causes it to clearly form a PCT state preceding SBCT (Table 2). Interestingly, increasing the solvent polarity further, to MeCN, induces m4B to show rapid formation of a SBCT with no apparent PCT intermediate. This could reflect formation and decay of the PCT state occurring faster than the instrument limited response or direct population of the SBCT from the LE state. Evidence for this latter scenario comes from the observation that MeCN is the only solvent wherein kSBCT is significantly faster for m4B than the other directly-linked systems. In all other solvents, directly-linked dimers exhibit similar SBCT rates (Table 2), implying that for m4B in MeCN, the SBCT state is formed directly from the LE state without need for a PCT intermediate. This could be a result of dynamic torsion of the excited m4B while m8B and e4m8B are more rigorously orthogonal and show population of the PCT state prior to SBCT, even in polar MeCN. That is, active torsion with the increased stability of the SBCT state in MeCN 74 for m4B may enable direct population from the LE to the SBCT state. This behavior for directly-linked systems can be contrasted with that of m8Ph (Figure 3.5 a, Table 2) and e4m8Ph in toluene or zDIP2/3 in cyclohexane 30,36 where SBCT is not observed (Table 2). For these cases, a model wherein the LE state is taken to directly return to the ground state fits the data well. When these dimers are placed in more polar solvents, they show two-state kinetics, where the LE state decays via SBCT without an observed intermediate formed during this exchange. This implies building significant population in a PCT state may require an uninterrupted through-bond linkage as the bridge may introduce an active coordinate that facilitates PCT. For BA, it was posed that PCT state formation was enabled via the activated torsion of the anthracene dyes about the meso bridge. Based on TA data of the dimers we have investigated, we conclude that steric encumbrance of this torsion does not inhibit PCT formation in these systems, but rather may prevent recombination to the LE state in nonpolar or weakly polar environments. This would also explain the nonlinearity of the distance dependence with respect to ln(kSBCT) for these systems as SBCT requires electronic coupling between the SBCT and PCT states, not the LE and SBCT states. Interestingly, SBCT seems to occur as a sequential step-down in energy for the mnB systems (Figure 6a). Figure 6a plots the LE, PCT and SBCT state energies relative to the ground state based on steady-state emission data for m8B and e4m8B. PCT state energies were respectively taken in cyclohexane and toluene for m8B and e4m8B, where SBCT state emission is not observed. SBCT state energies are from emission experiments recorded in DCM. The PCT state’s ability to serve as an intermediate between the LE and SBCT states may aid in facilitating efficient charge separation in these systems. This is in likeness to a z-scheme in photosynthesis where electron transfer occurs between neighboring acceptors with small decreases in energy, leading to an 75 overall efficient system favoring electron transfer. For mnB dimers, this step-down in energy leads to rapid SBCT, regardless of increased intramolecular reorganization or changes in electronic energy of the LE state. When PCT is seen, similar kSBCT are observed despite the reduced interchromophore coupling induced by dimer rigidity. For instance, kSBCT is nearly equivalent in m8B and e4m8B in solvents of similar polarity even though the LE state is about 0.13 eV (~30 nm) lower in energy in e4m8B than m8B. 4. Summary & Outlook The excited state and electrochemical properties of dipyrrins make them advantageous for synthetic approaches for solar energy production and generation of solar fuels via artificial photosynthesis. By learning from a biologically established phenomenon, SBCS, we can use solvent or environmental manipulations to tune the efficiency of charge generation and separation within synthetic dimers. We have found SBCT efficiency to be dependent on several intramolecular properties, including interdimer separation as well as chromophore coupling induced by chemical bonding and molecular structure. The longevity of the SBCT state was found to be less dependent on interdimer spatial separation and more dependent on wavefunction overlap. A rigidly orthogonal dimer, even at close proximity, can give rise to long-lived SBCT states. We have also shown formation of an intermediate PCT state may play a key role in rapid SBCT. These design principles can be used in the design of new molecular systems that rapidly form long-lived, spatially-separated charges when photoexcited. 5. Experimental 5.1 Femtosecond Transient Absorption The general transient absorption set up for m4B, e4m8B, m4Ph, m8Ph and e4m8Ph is described in (1). In brief, femtosecond experiments were performed using a Ti:sapphire regenerative amplifier 76 (Coherent Legend Elite, 1kHz, 35 fs, 3.5 W). The output from the amplifier was split such that about 10% was used to pump a Type II OPA (Spectra Physics OPA-800C) resulting in pump pulses resonant with the S0→S1 transition in the dimer systems, ranging from 500 – 550 nm (For m8B the pump source was from a home build, double pass noncolinear OPA, NOPA). Probe pulses were generated by pumping a rotating CaF2 plate giving a broad-band white light continuum (320 – 950 nm). A pair of off-axis parabolic mirror were used to collimate and focus the probe onto the sample. The overlapping pump-pulse was focused using a CaF2 lens. Data were collected with the probe at magic angle with respect to the pump (with the exception of m4B in acetonitrile which was collected at perpendicular geometry to minimize scatter). A spectrograph (MS1271) was used to disperse the probe continuum onto a 256 pixel silicon diode array (Hamamatsu). Differential detection of pump-induced signal was achieved using a chopper set to have the repetition rate of the laser, 500 Hz. Samples were prepared in a 1 mm (for m4Ph, samples were purged with nitrogen in a custom 1 cm cuvette) cuvette where maximum absorbance ranged between 0.1 – 0.3 OD. The cross-correlation of the pump and probe pulses, based off time zero coherent artifacts, was ~ 150 – 180 fs (~300 fs for 1 cm cuvettes and ~ 70 fs for m8B with NOPA pump source). 5.2 Global Analysis Transient spectra for BODIPY dimers (mnB and mnPh) compared to that of the zDIP systems showed that the rate of charge transfer was dependent on distance and intramolecular conformation, both of which effect electronic coupling between the locally excited (LE) and symmetry breaking charge transfer (SBCT) states. It was also discussed in the main text the discovery of an intermediate state in all of the directly linked (mnB) dimers which introduced an 77 efficient step-down to the final SBCT state. This intermediate state was denoted the partial charge transfer (PCT) state in accordance to literature on 9,9’-bianthryl. In order to determine the rates of formation and decay of these states, global analysis of the transient data was performed. Specifically, a linear decomposition of the transient signal in terms of the spectral and temporal componenets: S(λ,t)=∑ C n (t)σ n (λ) n Where S(λ,t) is the transient signal as a function of time (t) and wavelength (λ), is decomposed into time independent species associated difference spectra (SADS), σ n (λ) , and time dependent concentration, C n (t) , for each state n. Each SADS corresponds to a characteristic spectrum of a given state which consists of positive features, attributed to excited state absorption, and negative features due to stimulated emission and ground state bleach. C n (t) is calculated assuming population occurs as sequential first order rate processes from the initially excited state. That is: C 1 (t) k 1 → C 2 (t) k 2 → C 3 (t) k 3 → … k n−1 → C n (t) The kinetic model used to describe each system was dependent on solvent polarity or type of dimer. The phenyl-bridged dimers m8Ph and e4m8Ph fit to a one state model in toluene and a two state model in polar solvents while m4Ph fit to a two and three state model in toluene and DCM, respectively and a two-state model in acetonitrile. The directly linked dimers showed an intermediate population which introduced a third state in polar solvents and in non-polar solvents m8B and e4m8B fit to a two-state model where the PCT was the final state. 78 Chapter 3 Bibliography (1) Brinkert, K. Energy Conversion in Natural and Artificial Photosynthesis. 2018. https://doi.org/10.1007/978-3-319-77980-5. (2) Grondelle, R. van; Dekker, J. P.; Gillbro, T.; Sundstrom, V. Energy Transfer and Trapping in Photosynthesis. Biochimica Et Biophysica Acta Bba - Bioenergetics 1994, 1187 (1), 1–65. https://doi.org/10.1016/0005-2728(94)90166-x. (3) Suga, M.; Akita, F.; Hirata, K.; Ueno, G.; Murakami, H.; Nakajima, Y.; Shimizu, T.; Yamashita, K.; Yamamoto, M.; Ago, H.; Shen, J.-R. Native Structure of Photosystem II at 1.95 Å Resolution Viewed by Femtosecond X-Ray Pulses. Nature 2015, 517 (7532), 99–103. https://doi.org/10.1038/nature13991. (4) El-Kabbani, O.; Chang, C. H.; Tiede, D.; Norris, J.; Schiffer, M. Comparison of Reaction Centers from Rhodobacter Sphaeroides and Rhodopseudomonas Viridis: Overall Architecture and Protein-Pigment Interactions. Biochemistry-us 1991, 30 (22), 5361–5369. https://doi.org/10.1021/bi00236a006. (5) Deisenhofer, J.; Epp, O.; Miki, K.; Huber, R.; Michel, H. X-Ray Structure Analysis of a Membrane Protein Complex Electron Density Map at 3 Å Resolution and a Model of the Chromophores of the Photosynthetic Reaction Center from Rhodopseudomonas Viridis. J Mol Biol 1984, 180 (2), 385–398. https://doi.org/10.1016/s0022-2836(84)80011-x. (6) Chang, C.-H.; Tiede, D.; Tang, J.; Smith, U.; Norris, J.; Schiffer, M. Structure of Rhodopseudomonas Sphaeroides R‐26 Reaction Center. Febs Lett 1986, 205 (1), 82–86. https://doi.org/10.1016/0014-5793(86)80870-5. (7) Reaction Centers of Phtosynthetic Bacteria, 1st ed.; Mchel-Beyerle, M.-E., Ed.; Springer Series in Biophysics; Springer-Verlag Berlin Heidelberg, 1990; Vol. 6. https://doi.org/10.1007/978-3-642-61297-8_43. (8) Breton, J.; Martin, J.-L.; Migus, A.; Antonetti, A.; Orszag, A. Femtosecond Spectroscopy of Excitation Energy Transfer and Initial Charge Separation in the Reaction Center of the Photosynthetic Bacterium Rhodopseudomonas Viridis. Proc National Acad Sci 1986, 83 (14), 5121–5125. https://doi.org/10.1073/pnas.83.14.5121. (9) Wasielewski, M. R. Photoinduced Electron Transfer in Supramolecular Systems for Artificial Photosynthesis. Chem Rev 1992, 92 (3), 435–461. https://doi.org/10.1021/cr00011a005. (10) Deisenhofer, J.; Norris, J. R. Photosynthetic Reaction Center ; Academic Press: San Diego, 1993; Vol. 2. https://doi.org/https://doi.org/10.1016/C2009-0-02601-4. (11) Veldkamp, B. S.; Han, W.-S.; Dyar, S. M.; Eaton, S. W.; Ratner, M. A.; Wasielewski, M. R. Photoinitiated Multi-Step Charge Separation and Ultrafast Charge Transfer Induced Dissociation 79 in a Pyridyl -Linked Photosensitizer–Cobaloxime Assembly. Energ Environ Sci 2013, 6 (6), 1917–1928. https://doi.org/10.1039/c3ee40378e. (12) Kim, T.; Kim, W.; Mori, H.; Osuka, A.; Kim, D. Solvent and Structural Fluctuations Induced Symmetry-Breaking Charge Transfer in a Porphyrin Triad. J Phys Chem C 2018, 122 (34), 19409–19415. https://doi.org/10.1021/acs.jpcc.8b05363. (13) Yushchenko, O.; Villamaina, D.; Sakai, N.; Matile, S.; Vauthey, E. Comparison of Charge- Transfer Dynamics of Naphthalenediimide Triads in Solution and π-Stack Architectures on Solid Surfaces. J Phys Chem C 2015, 119 (27), 14999–15008. https://doi.org/10.1021/acs.jpcc.5b04060. (14) Giaimo, J. M.; Gusev, A. V.; Wasielewski, M. R. Excited-State Symmetry Breaking in Cofacial and Linear Dimers of a Green Perylenediimide Chlorophyll Analogue Leading to Ultrafast Charge Separation. J Am Chem Soc 2002, 124 (29), 8530–8531. https://doi.org/10.1021/ja026422l. (15) Hattori, S.; Ohkubo, K.; Urano, Y.; Sunahara, H.; Nagano, T.; Wada, Y.; Tkachenko, N. V.; Lemmetyinen, H.; Fukuzumi, S. Charge Separation in a Nonfluorescent Donor−Acceptor Dyad Derived from Boron Dipyrromethene Dye, Leading to Photocurrent Generation. J Phys Chem B 2005, 109 (32), 15368–15375. https://doi.org/10.1021/jp050952x. (16) Hu, R.; Lager, E.; Aguilar-Aguilar, A.; Liu, J.; Lam, J. W. Y.; Sung, H. H. Y.; Williams, I. D.; Zhong, Y.; Wong, K. S.; Peñ a-Cabrera, E.; Tang, B. Z. Twisted Intramolecular Charge Transfer and Aggregation-Induced Emission of BODIPY Derivatives. J Phys Chem C 2009, 113 (36), 15845–15853. https://doi.org/10.1021/jp902962h. (17) Kumpulainen, T.; Lang, B.; Rosspeintner, A.; Vauthey, E. Ultrafast Elementary Photochemical Processes of Organic Molecules in Liquid Solution. Chem Rev 2016, 117 (16), 10826–10939. https://doi.org/10.1021/acs.chemrev.6b00491. (18) Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W. Structural Changes Accompanying Intramolecular Electron Transfer: Focus on Twisted Intramolecular Charge-Transfer States and Structures. Chem Rev 2003, 103 (10), 3899–4032. https://doi.org/10.1021/cr940745l. (19) Dereka, B.; Svechkarev, D.; Rosspeintner, A.; Tromayer, M.; Liska, R.; Mohs, A. M.; Vauthey, E. Direct Observation of a Photochemical Alkyne–Allene Reaction and of a Twisted and Rehybridized Intramolecular Charge-Transfer State in a Donor–Acceptor Dyad. J Am Chem Soc 2017, 139 (46), 16885–16893. https://doi.org/10.1021/jacs.7b09591. (20) Fakis, M.; Beckwith, J. S.; Seintis, K.; Martinou, E.; Nançoz, C.; Karakostas, N.; Petsalakis, I.; Pistolis, G.; Vauthey, E. Energy Transfer and Charge Separation Dynamics in Photoexcited Pyrene–Bodipy Molecular Dyads. Phys Chem Chem Phys 2017, 20 (2), 837–849. https://doi.org/10.1039/c7cp06914f. 80 (21) Müller, S.; Heinze, J. Fluorescence Analysis of Two New TICT Systems: 10-Cyano-9,9′- Bianthryl (CBA) and 10, 10′-Dicyano-9,9′-Bianthryl (DCBA). Chem Phys 1991, 157 (1–2), 231– 242. https://doi.org/10.1016/0301-0104(91)87147-n. (22) Jurczok, M.; Plaza, P.; Martin, M. M.; Meyer, Y. H.; Rettig, W. Excited State Relaxation Paths in 9,9′-Bianthryl and 9-Carbazolyl-Anthracene: A Sub-Ps Transient Absorption Study. Chem Phys 2000, 253 (2–3), 339–349. https://doi.org/10.1016/s0301-0104(99)00386-9. (23) Khara, D. C.; Paul, A.; Santhosh, K.; Samanta, A. Excited State Dynamics of 9,9′-Bianthryl in Room Temperature Ionic Liquids as Revealed by Picosecond Time-Resolved Fluorescence Study. J Chem Sci 2009, 121 (3), 309–315. https://doi.org/10.1007/s12039-009-0035-6. (24) Hashimoto, S.; Yabushita, A.; Kobayashi, T.; Okamura, K.; Iwakura, I. Direct Observation of the Change in Transient Molecular Structure of 9,9′-Bianthryl Using a 10 fs Pulse UV Laser. Chem Phys 2018, 512 (Bull. Chem. Soc. J. 49 4 1976), 128–134. https://doi.org/10.1016/j.chemphys.2017.12.016. (25) Piet, J. J.; Schuddeboom, W.; Wegewijs, B. R.; Grozema, F. C.; Warman, J. M. Symmetry Breaking in the Relaxed S 1 Excited State of Bianthryl Derivatives in Weakly Polar Solvents. J Am Chem Soc 2001, 123 (22), 5337–5347. https://doi.org/10.1021/ja004341o. (26) Lee, C.; Choi, C. H.; Joo, T. A Solvent–Solute Cooperative Mechanism for Symmetry- Breaking Charge Transfer. Phys Chem Chem Phys 2019, 22 (3), 1115–1121. https://doi.org/10.1039/c9cp05090f. (27) Takaya, T.; Hamaguchi, H.; Iwata, K. Femtosecond Time-Resolved Absorption Anisotropy Spectroscopy on 9,9′-Bianthryl: Detection of Partial Intramolecular Charge Transfer in Polar and Nonpolar Solvents. J Chem Phys 2009, 130 (1), 014501. https://doi.org/10.1063/1.3043368. (28) Fujiwara, T.; Egashira, K.; Ohshima, Y.; Kajimoto, O. Effects of a Solvent Molecule on the Torsional Potential of 9,9′-Bianthryl. Phys Chem Chem Phys 2000, 2 (7), 1365–1373. https://doi.org/10.1039/a910327i. (29) Asami, N.; Takaya, T.; Yabumoto, S.; Shigeto, S.; Hamaguchi, H.; Iwata, K. Two Different Charge Transfer States of Photoexcited 9,9′-Bianthryl in Polar and Nonpolar Solvents Characterized by Nanosecond Time-Resolved Near-IR Spectroscopy in the 4500−10 500 Cm −1 Region. J Phys Chem 2010, 114 (22), 6351–6355. https://doi.org/10.1021/jp912173h. (30) Trinh, C.; Kirlikovali, K.; Das, S.; Ener, M. E.; Gray, H. B.; Djurovich, P.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Charge Transfer of Visible Light Absorbing Systems: Zinc Dipyrrins. J Phys Chem C 2014, 118 (38), 21834–21845. https://doi.org/10.1021/jp506855t. (31) Nepomnyashchii, A. B.; Bard, A. J. Electrochemistry and Electrogenerated Chemiluminescence of BODIPY Dyes. Accounts Chem Res 2012, 45 (11), 1844–1853. https://doi.org/10.1021/ar200278b. 81 (32) Kandrashkin, Y. E.; Wang, Z.; Sukhanov, A. A.; Hou, Y.; Zhang, X.; Liu, Y.; Voronkova, V. K.; Zhao, J. Balance between Triplet States in Photoexcited Orthogonal BODIPY Dimers. J Phys Chem Lett 2019, 10 (15), 4157–4163. https://doi.org/10.1021/acs.jpclett.9b01741. (33) Filatov, M. A. Heavy-Atom-Free BODIPY Photosensitizers with Intersystem Crossing Mediated by Intramolecular Photoinduced Electron Transfer. Org Biomol Chem 2019, 18 (1), 10–27. https://doi.org/10.1039/c9ob02170a. (34) Golden, J. H.; Estergreen, L.; Porter, T.; Tadle, A. C.; R, D. S. M.; Facendola, J. W.; Kubiak, C. P.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Charge Transfer in Boron Dipyridylmethene (DIPYR) Dimers. Acs Appl Energy Mater 2018, 1 (3), 1083–1095. https://doi.org/10.1021/acsaem.7b00214. (35) Whited, M. T.; Patel, N. M.; Roberts, S. T.; Allen, K.; Djurovich, P. I.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Intramolecular Charge Transfer in the Excited State of Meso -Linked BODIPY Dyads. Chem Commun 2011, 48 (2), 284–286. https://doi.org/10.1039/c1cc12260f. (36) Kellogg, M.; Akil, A.; Ravinson, D. S. M.; Estergreen, L.; Bradforth, S. E.; Thompson, M. E. Symmetry Breaking Charge Transfer as a Means to Study Electron Transfer with No Driving Force. Faraday Discuss 2019, 216 (0), 379–394. https://doi.org/10.1039/c8fd00201k. (37) Bartynski, A. N.; Gruber, M.; Das, S.; Rangan, S.; Mollinger, S.; Trinh, C.; Bradforth, S. E.; Vandewal, K.; Salleo, A.; Bartynski, R. A.; Bruetting, W.; Thompson, M. E. Symmetry- Breaking Charge Transfer in a Zinc Chlorodipyrrin Acceptor for High Open Circuit Voltage Organic Photovoltaics. J Am Chem Soc 2015, 137 (16), 5397–5405. https://doi.org/10.1021/jacs.5b00146. (38) Trinh, C.; Kirlikovali, K.; Das, S.; Ener, M. E.; Gray, H. B.; Djurovich, P.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Charge Transfer of Visible Light Absorbing Systems: Zinc Dipyrrins. J Phys Chem C 2014, 118 (38), 21834–21845. https://doi.org/10.1021/jp506855t. (39) Marcus, R. A. ON THE THEORY OF ELECTROCHEMICAL AND CHEMICAL ELECTRON TRANSFER PROCESSES. Can J Chem 1959, 37 (1), 155–163. https://doi.org/10.1139/v59-022. (40) Noginov, M. A.; Dewar, G.; McCall, M. W.; Zheludev, N. I.; Andrews, D. Tutorials in Complex Photonic Media. 2009, 439–478. https://doi.org/10.1117/3.832717.ch14. (41) Closs, G. L.; Calcaterra, L. T.; Green, N. J.; Penfield, K. W.; Miller, J. R. Distance, Stereoelectronic Effects, and the Marcus Inverted Region in Intramolecular Electron Transfer in Organic Radical Anions. J Phys Chem 1986, 90 (16), 3673–3683. https://doi.org/10.1021/j100407a039. (42) CLOSS, G. L.; MILLER, J. R. Intramolecular Long-Distance Electron Transfer in Organic Molecules. Science 1988, 240 (4851), 440–447. https://doi.org/10.1126/science.240.4851.440. 82 (43) Das, S.; Thornbury, W. G.; Bartynski, A. N.; Thompson, M. E.; Bradforth, S. E. Manipulating Triplet Yield through Control of Symmetry-Breaking Charge Transfer. J Phys Chem Lett 2018, 9 (12), 3264–3270. https://doi.org/10.1021/acs.jpclett.8b01237. (44) Kee, H. L.; Kirmaier, C.; Yu, L.; Thamyongkit, P.; Youngblood, W. J.; Calder, M. E.; Ramos, L.; Noll, B. C.; Bocian, D. F.; Scheidt, W. R.; Birge, R. R.; Lindsey, J. S.; Holten, D. Structural Control of the Photodynamics of Boron−Dipyrrin Complexes. J Phys Chem B 2005, 109 (43), 20433–20443. https://doi.org/10.1021/jp0525078. (45) Cook, R. E.; Phelan, B. T.; Kamire, R. J.; Majewski, M. B.; Young, R. M.; Wasielewski, M. R. Excimer Formation and Symmetry-Breaking Charge Transfer in Cofacial Perylene Dimers. J Phys Chem 2017, 121 (8), 1607–1615. https://doi.org/10.1021/acs.jpca.6b12644. (46) Aster, A.; Licari, G.; Zinna, F.; Brun, E.; Kumpulainen, T.; Tajkhorshid, E.; Lacour, J.; Vauthey, E. Tuning Symmetry Breaking Charge Separation in Perylene Bichromophores by Conformational Control. Chem Sci 2019, 10 (45), 10629–10639. https://doi.org/10.1039/c9sc03913a. (47) Hestand, N. J.; Spano, F. C. Expanded Theory of H- and J-Molecular Aggregates: The Effects of Vibronic Coupling and Intermolecular Charge Transfer. Chem Rev 2018, 118 (15), 7069–7163. https://doi.org/10.1021/acs.chemrev.7b00581. (48) Engels, B.; Engel, V. The Dimer-Approach to Characterize Opto-Electronic Properties of and Exciton Trapping and Diffusion in Organic Semiconductor Aggregates and Crystals. Phys Chem Chem Phys 2017, 19 (20), 12604–12619. https://doi.org/10.1039/c7cp01599b. 83 CHAPTER 4: Symmetry Breaking Charge Transfer in DIPYR dimers 1. Introduction: Typically, the red to near infrared (NIR) region of the solar spectrum is underutilized in organic photovoltaics. This is because NIR absorbing dyes readily undergo internal conversion (IC) processes which truncates the lifetime of the excited state 1 . This results in dyes whose excited state lifetime is too short to be utilized for charge generation and ultimately power conversion. Since much of the solar irradiance lies in the NIR (800 – 2000 nm, Figure 1a) this absorption limitation greatly impacts the efficiencies of organic dyes for photovoltaic (PV) applications, as these photons will simply go unabsorbed. There is another advantage to reducing the energy requirement for charge generation as this will inherently increase the short circuit current (JSC) 2,3 . As was discussed in the introductory chapter, to maximize device efficiency the open circuit voltage (VOC) and short circuit current (JSC) of a device need to be maximized, as their product gives the maximal power for a given device. By decreasing the absorption energy requirement for exciton formation, diffusion and eventual collection of charges, the JSC will be maximized. However, as shown by Shockley and Queisser in their thermodynamic description of PV efficiency 4 , the JSC and VOC are coupled, that is changing one parameter will affect the other. The VOC is typically optimized by increasing the energy difference between the valence and conduction band in a solid-state medium 4 , or, for organic materials, this would be defined as the energy gap between donor and acceptor layers 2 . However, a better description of the VOC accounts for the energy of charge transfer at the interface, accounting for energy loss when the electron or hole is transferred to the acceptor or donor layers, respectively 5 . One way to decrease the energy requirement for exciton formation, thus increasing the JSC while maximizing the VOC, is to use symmetric systems where loss due to charge transfer 84 is minimized. Charge transfer processes in dimers have been observed in biological systems at high efficiencies. Figure 4.1: a) solar irradiance spectrum with SBCT BODIPY dimers. The dotted line to longer wavelengths indicates where we would like our dyes to absorb light. b) Diagram for SBCT and how this could be implemented at a donor/acceptor interface of a device. Nature has found a way to utilized red to near infrared (NIR) photons to form separated charges, which catalyze a series of biosynthetic reactions through the process of photosynthesis. In autotrophic bacteria this has been shown to occur via optical absorptions ranging from 800 – 1000 nm (1.55 – 1.24 eV) 6 –9 . Photosynthetic organisms, both plants and bacteria, contain a multi- chromophoric system in their reaction centers which absorbs optical excitations transferred from the light harvesting system. The multichromophoric system contains a dimer of cofacial bacteriochlorophyll, called the special pair, attached to this dimer is a pair of accessory bacteriochlorophylls which has a pair of bacteriopheophytin on either side 10,11 . This complex with the special pair at the center is held in a C2 symmetry by the protein scaffold. Optical excitations transferred from the light harvesting system ultimately end at the special pair 12 –14 . Following excitation, an electron is transferred from the special pair to the neighboring bacteriopheophytin that is on the active side of the complex. From this initial charge transfer a series of charge transfer 85 processes ensue which ultimately result in synthesis of adenosine triphosphate (ATP) which is used as the energy source for survival of the bacteria. The special pair in Rhodobacter sphaeroides optically absorbs at 865 nm (1.43 eV) 8 while the special pair in Rhodopseudomonas viridis absorbs down to about 1000 nm (1.24 eV) 7 . Both of these systems give quantum efficiencies of 100 %, that is, per photon absorbed an electron is transferred The special pair in the reaction center uses an excited state phenomenon for highly efficient charge separation known as symmetry breaking charge separation (SBCS). SBCS is a process in which a pair of closely associated chromophores are excited. Following this initial excitation, electronic coupling of the chromophores combined with the driving force provided by the environment enables charge to transfer from, or in more general cases within, the dimer system to a neighboring chromophore such that the transferred and remaining charge act independently. The case in which a dimer system undergoes charge transfer where the spatially separated charges are still coupled, either by confinement due to the size of the dimer or wavefunction overlap, gives the charge transfer analogue, SBCT. The primary distinction between SBCT and SBCS is in that SBCT state is fluorescent, giving a solvent dependent emission spectrum, while SBCS is nonemissive 15 . Like in SBCS, SBCT requires a closely associated pair of chromophores where upon excitation of the dimer, the excitation localizes, either immediately or after very rapid electronic dephasing of a superposition of the two excitonically coupled states, on one of the chromophores producing a localized excited (LE) state. This is then followed by hole or electron transfer to the neighboring chromophore. That is, the localized electron-hole pair is spatially delocalized across the dimer system, reducing the coulombic attraction of the electron and hole and thus reducing the 86 probability of recombination, increasing of the excited state lifetime. This yields an increase in lifetime of charges useful towards power conversion. By reducing the probability of recombination while reducing the energy requirement for exciton generation and diffusion, the JSC can be increased without having a negative impact on the VOC of a device. In organic photovoltaics, using symmetry breaking charge transfer therefore could be beneficial. Typically, charge transfer requires a polar environment to stabilize a break in symmetry in the dimer charge distribution. This would also result in selection of this process that may occur preferentially at a donor-acceptor interface of a device (Figure 4.1b). Previously our group has looked at dipyrrin derivatives, where the dipyrrin chromophores are either bridged through a noninteracting zinc (II) center through the nitrogen terminals 16 –18 or covalently through the meso positions 19 . In the previous chapter, structure-property relationships with respect to tuning of the bridging groups and changing of substitution were studied with respect to the efficiency of SBCT. In light of this it was found that short distances, or direct linkage, resulted in the fastest rate of SBCT while rigid orthogonality of the dimers leads to the longest lifetimes of the SBCT states. One other benefit to direct linkage was found in the existence of an intermediate state which allowed rapid formation of the SBCT state regardless of the energy gap between the LE and SBCT state. This intermediate we have termed a partial charge transfer (PCT) state, in accordance to literature on 9,9’-biantrhryl. Our group has worked on dimer systems made of dipyridylmethene (DIPYR) moieties to aid in decreasing the excitation energy requirement for these systems while doing symmetry breaking charge transfer 20 . Two dimer systems were studied, bis-DIPYR and bis- α-DIPYR (Fig 4.2) using steady-state absorption/emission, spectroelectrochemistry and transient absorption experiments. Previous work on these systems reported that both dimers undergo rapid SBCT, 87 regardless of solvent and in bis- α-DIPYR this ultimately results in formation of a triplet 20 . However, the recent work by our group on BODIPY dimers has resulted in further understanding of potential excited state processes that lead to excited state signals observed, and this in turn may provide further insight into the process of charge transfer within these systems. Figure 4.2: Borondifluoro dipyridylmethene monomer and dimer systems of interest 2. Experimental Methods 2.1 Sample preparation DIPYR monomers and dimers (Figure 4.2) were synthesized as described previously 20,21 . Samples were prepared for transient absorption experiments such that optical densities in a 1mm path remained within 0.2 – 0.4 OD range at the wavelength of maximum absorption. For nanosecond TA experiments, samples were prepared on a Schlenk line where the samples were purged with N2 and dissolved in dry solvents such that the maximum absorbance did not exceed 0.5 OD. All samples for transient absorption experiments were made in a capped 1 mm quartz cuvette. 2.2 Steady-state photophysics Steady-state emission and absorption experiments have been described elsewhere 20 . UV- visible spectra were recorded on a Hewlett‑Packard 4853 diode array spectrometer. Photoluminescence spectra were measured using a QuantaMaster Photon Technology 88 International phosphorescence/fluorescence spectrofluorimeter. Quantum yield measurements were carried out using a Hamamatsu C9920 system equipped with a xenon lamp, calibrated integrating sphere and model C10027 photonic multi-channel analyzer (PMA). 2.3 Femtosecond Transient Absorption Pump and probe pulses were obtained from the output of a Ti:Sapphire regenerative amplifier (Coherent Legend Elite, 1 kHz, 3.2 mJ, 35 fs). The pump pulses were generated by pumping a type II OPA (Spectra Physics OPA-800C) with 10% of the 800 nm amplifier output and mixing the OPA signal with the 800 nm residual on a type II BBO to generate 500 nm and 485 nm. 500 nm was used to pump bis-DIPYR, bis-α-DIPYR and α-DIPYR monomer while 485 nm was used to pump the DIPYR monomer. A white light supercontinuum (320-950 nm) was generated by focusing a small amount of the 800 nm on a rotating CaF2 disk. The supercontinuum probe was collimated and focused by a pair of off-axis parabolic mirrors onto the sample, while the pump was focused before the sample with a 25 cm CaF2 lens. The polarization of the probe was rotated to magic angle (54.6º) with respect to the pump using a λ/2 waveplate in order to avoid dynamics associated with molecular reorientation. The supercontinuum probe was dispersed using a spectrograph (Oriel MS127i) onto a 256-pixel silicon diode array (Hamamatsu) for multiplexed detection of the probe. During data collection the samples were moved on a motorized stage perpendicular to the pump to decrease photodamage caused by the pump. 2.4 Nanosecond Transient Absorption Samples were sealed in a capped 1 mm quartz cuvette which was translated backwards and forwards on a stage. Nanosecond pump pulse generation was performed using a 532 nm output from a Nd:YAG laser (Alphalas, 532 nm, 1 kHz, 700 ps) which is externally triggered and 89 synchronized with the femtosecond amplifier. The pump pulse is delayed with respect to the femtosecond supercontinuum probe using a delay generator DG 645 (Stanford Research Systems). 3. Results and Discussion 3.1 Steady-state absorption and emission: The steady-state absorption of the dimers relative to their respective monomers can help give insight into ground state orientation and allowed excited state transitions. First, let us consider the monomer absorption of the DIPYR and α-DIPYR dyes which has been discussed extensively elsewhere 21 . As shown in Figure 4.3 a & b, the lowest energy absorption bands in both monomers is the S0 ν=0 →S1 ν=0 transition where the next lowest energy absorption band transition is the S0 ν=0 →S1 ν=1 transition. In DIPYR, the S0 ν=0 → S1 ν=1 transition is superimposed by an S0→S2 transition. This is based off the relative geometry of the S0 → S2 transition moment being along the short axis of the DIPYR which is parallel to the molecular dipole moment giving a medium dielectric dependence to the S0 →S2 absorption energy 21 . In the α-DPYR dye, there is no such solvent dependence so this second band only contains the S0 ν=0 → S1 ν=1 transition. Looking at the bis-DIPYR dimer, the absorption line shape has similar characteristics to the monomer. Both the primary and secondary absorption peaks red shift by a similar magnitude of about 60 meV (~483 cm -1 ) from the monomer spectrum (Figure 4.3a). However, the relative intensities are different between the two bands and there is very little solvent dependence on the energy and intensity of the secondary peak (Figure 4.3c). The effect on the second peak could be due to the two S0→S2 transition moments being along the intra-dimer bridge which should result in excitonic coupling of the two transition moments, inducing a red-shift, as it is J-aggregate like in character. In the case of the dimer this peak shows little-to-no solvent dependence, unlike that of the monomer. This has been explained previously to be due to the dimer having a net zero 90 molecular dipole as the two dipole moments of each DIPYR chromophore cancels out 20 . The S0→S1 transition is red-shifted by nearly the same amount, which could be due to destabilization of the ground state caused by the proximity of the two DIPYR species. This same effect was observed in all of the directly linked BODIPY dimers in the previous chapter, regardless of substitution or excitonic coupling. Steady state absorption in bis-α-DIPYR compared to the α-DIPYR monomer red shifts by about 100 meV (~806 cm -1 ) going from monomer to dimer, implying that there may be significant coupling in the ground state configuration leading to a destabilized ground state (Figure 4.3b). Interestingly, bis-α-DIPYR gives an absorption spectrum that is broadened and the band that is Figure 4.3: Steady-state absorption and emission of DIPYR and bis-DIPYR (a) and α-DIPYR and bis-α- DIPYR (b) in cyclohexane. Solvent dependent absorption and emission of bis-DIPYR (c) and bis-α- DIPYR (d). 91 attributed to the S0 ν=0 → S1 ν=1 absorption in the parent monomer, has also increased in amplitude. The broadened and red-shifted absorption may be a result of allowed ground state torsion in bis-α-DIPYR. Dimerization of the α-DIPYRs may give rise to a larger Frank-Condon overlap of the S0 ν=0 →S1 ν=1 transition giving a larger observed absorption amplitude. Alternatively, increased absorption to the blue of the primary absorption band could be due to H- aggregate-like excitonic splitting from coupling of the transition moments of the α-DIPYR chromophores. The steady-state emission spectrum of bis-DIPYR in cyclohexane is strikingly different from its parent monomer. In the monomer, the emission is a Stokes-shifted (Δν = 129 cm -1 ) mirror image. The dimer’s emission spectrum is largely red shifted (Δ ν= 1210 cm -1 ), broadened and not mirrored to the absorption spectrum (Fig. 4.3 a & c). In the previous chapter this was observed in 1,3,5,7-tetramethyl-bis-BODIPY (m8B) in cyclohexane and was attributed to formation of a partial charge transfer (PCT) state. However, unlike m8B, there are no methyl groups to produce a rigid frame and therefore torsion about the meso bridge is fully allowed. We should then expect a greater likeness with the emission spectrum of 1,7-dimethyl-bis-BODIPY (m4B) in cyclohexane 19 , which looks to be emission from the localized excited (LE) state. However, here this does not seem to be the case. Torsion can lead to broadening, however, the degree to which the emission spectrum is red shifted compared to the absorption; the lack of any vibrational character we argue implies that the fluorescent state is not the S1 or LE state. Similar to bis-DIPYR, bis-α-DIPYR gives a significantly red-shifted emission (Δν = 880 cm -1 ) relative to the Stokes shift observed in α-DIPYR monomer (Δν = 40 cm -1 ). The emission spectrum of bis-α-DIPYR also is not a mirror- image of the absorption and has a similar line shape to the bis-DIPYR emission. Bis-α-DIPYR also has accessible torsion, which could be 92 an explanation to the broadening of the emission spectrum, or it could be that the emissive state is also from a state other than the S1 or LE. As suggested for bis-DIPYR, this too could be emission from a PCT state, in analogy to the directly linked BODIPY dyes in the previous chapter. Fluorescence is a great indicator of SBCT if monitored with respect to solvent polarity. Since, in principle, a CT state is fluorescent, then the fluorescence will occur at longer wavelengths (lower transition energies) with increasing solvent polarity. If this solvatochromic behavior is not observed, fluorescence is only observed from the LE state and the photoluminescent quantum efficiency decreases with increasing polarity, then the system is most likely undergoing SBCS. In the case of bis-DIPYR, increasing solvent polarity gives negligible changes to the shape of the fluorescence spectra while the quantum efficiency decreases substantially, from ΦPL= 0.26 in Methyl-cyclohexane to ΦPL=0.03 in acetonitrile 20 . However, as stated previously, the emission spectrum does not appear to trace back to the monomer-like LE state. Bis-α-DIPYR also shows negligible emission spectra shape change with respect to solvent polarity and the quantum yield only decreases slightly from ΦPL= 0.57 in Methyl-cyclohexane to ΦPL=0.31 in acetonitrile. Though the overall drop in quantum efficiency in bis-α-DIPYR is about the same as bis-DIPYR, the quantum efficiencies are still much larger in polar acetonitrile than one would expect from a system which undergoes SBCS. Since the fluorescence in bis-a-DIPYR is not solvatochromic and the quantum efficiency does not indicate complete quenching of the fluorescent state, this poses many questions we hope to address: which state is the emissive state, why is there no solvatochromism and why is there not a more significant drop in ΦPL? In order to directly look at the excited state, transient absorption (TA) spectroscopy is used. 93 3.2 Femtosecond Transient Absorption In the previous chapter and in other work by our group, zinc dipyrrinato dimers show a strong solvent dependency for SBCS where the photoluminescence quantum yield decreases significantly with increasing polarity of the environment 16,18 . In a nonpolar environment the dimers show trivial transient behavior, S1 population followed by decay back to the ground state. In more polar solvents, a two-state model is observed within the first nanosecond of excited state evolution. The observed transient behavior corresponds to population of the LE state, where the exciton is localized on one of the chromophores, followed by population of the SBCS state. The rate at which the SBCS state is populated can be modulated by the solvent polarity. However, in the directly linked BODIPY dimers, population of an intermediate state, regardless of the solvent, was observed. We discussed in the previous chapter that this intermediate partial charge transfer state requires direct, uninterrupted, through-bond interaction. It is believed to be an excited state manifestation which is due to charge transfer between the two chromophores that is still coupled to the LE state. In the BODIPY dimers, the transient behavior in nonpolar solvents rendered an isosbestic point when going from LE to PCT, implying that the PCT state is in fact a distinct electronic state, rather than simple nuclear relaxation/solvation on the initial LE which should be manifested by continuous spectral shifting. Here we will consider the possibility of partial charge transfer as the DIPYR dimers analyzed herein are also directly meso bridged, implying that, like 9,9’-bianthryl and the BODIPY dimers, an intermediate state is likely to occur prior to complete charge transfer or separation. Previous measurements of the DIPYR and α-DIPYR monomers 20 , which comprise the bis- DIPYR and bis-α-DIPYR dimers, respectively, were done to examine potential nontrivial behaviors of the monomers which may attribute to the transient behavior of the dimers. It was 94 shown previously that α-DIPYR follows a trivial excited state behavior, that is population of the S1 state followed by decay to the ground, independent of solvent. DIPYR on the other hand follows a nontrivial behavior where the excited state behavior is attributed to symmetry allowed intersystem crossing, which is observed within a nanosecond of excitation. This may contribute to the time dependent excited state behavior of the bis-DIPYR and will be considered when analyzing the transient data of bis-DIPYR. 95 Figure 4.4: Transient absorption spectra of bis-DIPYR (left) in cyclohexane (a), tetrahydrofuran (c) and acetonitrile (e) with emission spectra of DIPYR monomer (dashed line) and bis-DIPYR (dotted line). Transient absorption of bis-α-DIPYR (right) in cyclohexane (b), tetrahydrofuran (d) and acetonitrile (f) with emission spectra of the α-DIPYR monomer (dashed line) and bis-α-DIPYR (dotted line). Transient traces were normalized to the GSB at the earliest time in each plot. Pump wavelengths for transient absorption experiments were centered at 500 nm. To begin, the excited state dynamics of bis-DIPYR will be looked at in nonpolar cyclohexane (Figure 4.4 a) which is an environment where the SBCT state should be destabilized 96 with respect to the LE state. This should result in no observed formation of the SBCT state and, in principle, should give a trivial excited state behavior where only population and decay of the LE state is allowed. However, like 9,9’-bianthryl and the directly linked BODIPY dimers, evidence of nontrivial behavior is hinted at in the steady-state fluorescence spectra. In bis-DIPYR at early times there is an excited state absorption (ESA) from 350 – 450 nm, indicative of the S1→Sn absorption. The negative feature from 450 – 500 nm corresponds to the ground state bleach (GSB), and from 510 – 650 nm the negative feature corresponds to stimulated emission (SE). The SE band just red of the GSB (500 – 520 nm) is spectrally similar to emission of the monomer DIPYR, though the band shape is lacking in vibrational structure seen in either the fluorescence spectrum or stimulated emission pattern for the monomer 20 (Fig 4.4a). Within the first 5 ps, the SE band of the LE state decays over several ps while a SE band further to the red (520 – 600 nm) remains and matches reasonably well the steady state emission observed for the dimer. Following the decay of the LE emission, the transient spectra decays where a decrease in amplitude of all components is observed. 97 Figure 4.5: SADS of bis-DIPYR in cyclohexane (a) tetrahydrofuran (c) and acetonitrile (e). Spectral slices plotted as a function of time for bis-DIPYR in cyclohexane (b) tetrahydrofuran (d) and acetonitrile (f) indicating ESA (dark blue), GSB (dark cyan) and SE (dark red), with spectral slices taken from global fitting overlaid. From global analysis fitting, the species associated difference spectrum (SADS), which mark the transient trace for a given state and time dependent concentrations, mechanisms can be fit for the different excited state processes. For bis-DIPYR in cyclohexane (Figure 4.5 a & b), 98 the data fits decently to a two-state equilibrium model between the first (σ 1) and second (σ 2) SADS where the forward and backward time constants are 2 ps and 140 ps, respectively. The forward rate corresponds to the formation of the DIPYR PCT intermediate state, where, like m8B, it does not evolve beyond this state in cyclohexane. Interestingly, the backward rate is significantly slower in bis-DIPYR (140 ps) than m8B (4 ps). This implies that return back to the LE state is less favorable in bis-DIPYR than m8B even though there is nothing inhibiting torsion in the case of bis-DIPYR. Bis-DIPYR in more polar tetrahydrofuran (THF) also follows a two-state equilibrium model (Figure 4.5c). As was observed in bis-DIPYR in cyclohexane, initial decay of the SE band between 500 – 520 nm is observed indicating loss of monomer-like LE emission (Figure 4.4 c and 4.5 c) where the dimer emission remains. An interesting difference in spectral evolution for the THF data compared to the cyclohexane transient spectra, is that along with decay of the SE attributed the LE emission, there is a blue shift in the ESA or loss of ESA at around 450 nm, which resulted in the increase in amplitude of the GSB band for the S0 ν=0 → S1 ν=1 transition. This could either correspond to a loss in ESA from the S1 state, further indicating population of a new electronic state that is more pronounced in polar THF relative to nonpolar cyclohexane. The backward and forward time constants from SADS (σ 1) to this second SADS (σ 2) are 1.2 ps and 43 ps, respectively. However, unlike what was observed for the directly linked BODIPY dimers, the system does not seem to evolve to a third state, but rather remains in this equilibrium state. Since the reshaping of the ESA from 350-450 nm was not observed in nonpolar cyclohexane, perhaps the second state in THF is the full SBCT, but where the characteristics of the PCT state and SBCT trace are similar. To expand on this further, bis-DIPYR in acetonitrile (MeCN) will be 99 discussed to see if in very polar MeCN there is any further evolution of the excited state to a third state or if bis-DIPYR remains as an equilibrium two-state model. Figure 4.6: Transient absorption spectra with cation (blue, dashed line) and anion (red, dash-dot line) absorption determined by chronoamperometric spectroelectrochemistry of bis-DIPYR (a) and bis-α-DIPYR (b) in acetonitrile. (c) Transient absorption spectra of bis-DIPYR in acetonitrile with cation (blue, dashed line) and anion (red, dash-dot line) absorption determined from pulsed-radiolysis. (d) TA of bis-α-DIPYR in acetonitrile with the anion absorption overlaid, where the anion was obtained using a different spectroelectrochemical cell from (b). This was the cell used for infrared spectroelectrochemistry (IR-SEC) as reported by Kubiak and coworkers (ref 22). The transient absorption of bis-DIPYR in MeCN can be fit by a two-state target model (Figure 4.5 e & f) where back transfer to the initial state is not necessary. The excited state absorption region from 350 – 450 nm reshaping seen in THF happens more rapidly in MeCN and there is no significant evolution of the TA from 5 ps onwards. Due to the pump scatter, it is difficult 100 to assess if the reshaping of the SE band from 510 – 650 nm is due to the loss of the monomer emission and gain of the dimer emission, which we attribute to PCT, or if there is a rise in absorption superimposed with this SE band. In MeCN, the second state (σ2 in Figure 4.5e) is fit to occur with a time constant of 0.8 ps. The faster rate of formation of this second state in MeCN relative to THF and cyclohexane implies that this second state has some charge transfer character. However, it is unclear as to whether or not in all solvents bis-DIPYR forms the PCT state without evolving to full SBCT within a nanosecond, or if in the case of the THF and MeCN data, the second state is SBCT and it is observed by a loss of ESA around 420 – 450 nm. To assess whether or not SBCT is occurring in the more polar solvent environments chronoamperometric spectroelectrochemical experiments were done. The anion and cation spectra were obtained by comparing the neutral ground state spectrum to the biased spectra. In Figure 4.6 a, the cation and anion spectra obtained for bis-DIPYR using chronoamperometric spectroelectochemistry are overlaid with the transient spectra of bis-DIPYR in acetonitrile in order to compare the characteristic absorption features for the cation and anion relative to the transient spectra. The anion spectrum gives a positive feature in the range of 500 – 650 nm, which may imply that the loss in amplitude of the SE band observed in THF and acetonitrile may be attributed to absorption of the formed anion which would indicate full SBCT. We can compare this to a more quantitative approach to producing the cation and anion spectra, where pulsed radiolysis measurements were made to create the charged species in bis-DIPYR and the anion and cation spectra were obtained (Figure 4.6 c). Qualitatively, the chronoamperometric experiments seemed to reproduce the anion absorption, however, the cation absorptions are different, suggesting that the cation spectrum from the spectroelectrochemical experiments resulted primarily in decomposed matter with a relatively small concentration of cations. From both experiments, it can 101 be seen that the cation and anion absorption have primary character in the red, and they have similar absorption energies. In the case of bis-DIPYR, there is little-to-no change in shape of the SE band, though in principle the SE band should change shape, not just decrease in amplitude. This may imply that the in the case of bis-DIPYR, the excited state does not evolve beyond the PCT state within a nanosecond. To try and answer this further we can simulate the transient spectra. This is done by using the inverted steady-state absorption, the steady state emission of the monomer and the emission of the dimer to simulate the GSB and SE bands. The emission spectra were scaled to the steady-state absorption where it is assumed that the transition strength of the monomer emission and absorption are about equal and the steady-state emission of the dimer, or emission from the intermediate partial charge transfer state, is about 0.3 × the transition strength of the monomer emission and GSB. Following this, incorporation of the cation and anion absorption was done to see if the spectra could be reproduced with an absorption component superimposed with the emission. From pulsed radiolysis measurements the absorptivity of the cation and anion absorption were determined and give us a handle on how this will scale the transitions with respect to the ground state absorption. The cation spectrum has a transition strength that is about 0.1 × that of the ground state absorption and the anion is 0.06× that of the ground state absorption. 102 Figure 4.7: Transient absorption spectra of bis-DIPYR in THF (a) and MeCN (c) where only the GSB and SE bands are included (450 – 650 nm). Simulated spectra where excited evolution only include the monomer and dimer emission (b) and simulated spectra with the monomer and dimer emission as well as the cation and anion absorption (d). It can be seen from the simulated spectra (Figure 4.7 b and d) that if complete charge transfer occurs, which should render the radical cation and anion absorption, the band from 510 – 650 should structurally change from absorption of the cation and anion. That is, as shown in Figure 4.7 d, the shape of the SE band changes where two new absorption bands are formed in accordance to the cation and anion spectra. If full SBCT were occurring, this absorption should be observed. Also, the simulated behavior where only the emission bands and ground state bleach are present (Figure 4.7 b), looks more closely to that of the experimental transient absorption (Figure 4.7 a and c). This further strengthens the idea that the second state is not full charge transfer, but rather 103 an intermediate with charge transfer character, which we have thus far been referring to as the partial charge transfer state and is consistent with the steady-state emission. Since SBCT does not seem to occur, solvatochromic behavior is only minimally observed with a decrease in photoluminescent quantum yield. This is consistent with the timescales for the formation of the PCT state going from cyclohexane to THF to MeCN, where the time constants decrease (or rates increase) from 2.6 ps to 1.3 ps to 0.8 ps, respectively. Comparing to the directly linked BODIPY dimers, the behavior of bis-DIPYR is strikingly different. That is, regardless of solvent polarity, the DIPYR dimer does not seem to evolve beyond the PCT sate within a nanosecond. Bis-α-DIPYR in cyclohexane (Figure 4.4 b), like bis-DIPYR, shows an initial change in SE over 5ps from the more monomer-like emission to that in better in agreement with the steady- state emission of the dimer, from 560 – 640 nm. Interestingly, unlike the other directly linked dimers in nonpolar solvents, bis-α-DIPYR also shows a change in shape of the ESA where there appears to be an increase in amplitude from 470 – 500 nm with a decrease in ESA at the band centered at ~515 nm. This same behavior is seen in THF as well, where within the first 5 ps, there is a rise and reshaping of the ESA between 470 – 500 nm and an increase in SE between 550 – 600 nm to better match the emission observed in the steady-state. Fitting with global analysis, bis- α-DIPYR in cyclohexane (Figure 4.8 a & b) can be described by an equilibrium being set up between two SADS where the forward and backwards processes occur with time constants of 3.7 and 6.7 ps, respectively. In THF formation of the second SADS (Figure 4.8 c, σ 2) requires no backward rate process for optimal fitting and the second state is formed with a time constant of 2.2 ps. Comparing solely the cyclohexane and THF TA spectra, and noting the similarities in SADS traces, one may conclude that bis-α-DIPYR does not proceed to SBCT in THF, as was observed in bis-DIPYR. 104 Figure 4.8: SADS of bis-α-DIPYR in cyclohexane (a) tetrahydrofuran (c) and acetonitrile (e). Spectral slices plotted as a function of time for bis-α-DIPYR in cyclohexane (b) Tetrahydrofuran (d) and acetonitrile (f) indicating ESA (dark cyan/dark blue), GSB (orange), and SE (red), with spectral slices taken from global fitting overlaid. Bis-α-DIPYR in MeCN, however, best fits to a three-state model (Figure 4.10e), where the first SADS is very similar to that observed in both cyclohexane and THF. Unlike the observation in THF and cyclohexane, the ESA from 470 – 500 nm experiences little to no change in shape and amplitude. Following this initial LE state, there is a decrease in SE neighboring the GSB, consistent with the second SADS observed in the other two solvents and with a time constant 105 of 1 ps, slightly faster than observed in the first two solvents. However, unlike cyclohexane and THF, a third SADS is required to model the data where a SE band experiences a slight loss in amplitude from 550 – 570 nm. This third state may be an evolution to a new electronic state, or nuclear relaxation processes attributed to the formation of the newly formed excited state which occur with a time constant of about 90 ps. As we did for the bis-DIPYR, we can compare the transient absorption experiments to the chronoamperometric spectroelectrochemistry experiments. The anion spectrum has a slight positive feature between 550 – 570 nm (Figure 4.6 b), similar to the anion observed in bis-DIPYR. Due to a limited supply of bis-α-DIPYR and the difficulty of synthesis, pulsed radiolysis experiments were not performed. The anion spectrum was reproduced using a different cell which is based on what was used for infrared spectroelectrochemistry (IR-SEC) as published by Kubiak and coworkers 22 (Figure 4.6 d). In Figure 4.6 b the anion spectrum was obtained using a technique where an electrode was placed in a cuvette with the bis-α-DIPYR in an electrolyte solution and a constant bias was applied and the was changed in the direction for either reduction or oxidation of the system, where spectra were measured near the electrode and the absorption of the charged species could be obtained. However, this method can lead to build up of decomposed product around the electrode, where signals associated with the cation and anion can be very small. The anion spectrum shown in Figure 4.6 d was measured in a cell where the potential was scanned and thin-layer bulk electrolysis was monitored by reflectance UV/visible spectroscopy off a polished platinum working electrode where the technique and cell are described in depth in Ref. 20 and 22. This technique resulted in less decomposed matter in the anion spectrum and gave a larger absorption from 550 – 650 nm. This second method for determining the anion spectrum, overall, reproduced the absorption observed with chronoamperometric measurements, but with 106 higher signals of the anion. This absorption may infer that the change in shape of the SE band could indicate formation of the anion spectrum, but seeing as there once again is no change to the lineshape of the SE centered around 580 nm, which would correspond to the rise in anion absorption, the implication is that the excited state does not evolve beyond an intermediate partial charge transfer state. 3.3 DFT calculations of bis-DIPYR and bis-α-DIPYR Figure 4.9: DFT calculations of bis-DIPYR (left) and bis-α-DIPYR (right) of their HOMO (bottom) and LUMO (top) orbitals using B3LYP/6-31G** basis (from ref 20). 107 To further understand these differences between the DIPYR and BODIPY dimers in terms of their excited state behavior, DFT calculations were performed for bis-DIPYR and bis-α-DIPYR to calculate the relevant molecular orbitals (Figure 4.9). Unlike the BODIPY systems, the HOMO has substantial charge density on the meso carbon while the LUMO has no charge density on the meso position. In BODIPY, this was in part why the PCT state was a necessary intermediate, as charge delocalization along the bridge would be beneficial towards charge transfer and, more importantly, an orthogonal configuration would inhibit charge recombination, as it would reduce wavefunction overlap of the anion and cation towards recombination. Here, the LUMO does not have charge density along the bridge, so the twisting of the DIPYR moieties is necessary for sufficient wavefunction overlap to allow for charge transfer or partial charge transfer. This is the largest difference between the BODIPY and DIPYR systems and may also be why SBCT appears to be so inefficient, or not occurring, in the DIPYR systems relative to the BODIPY systems. That is to say that charge localization on the bridge in the LUMO, may act to enable PCT and eventual SBCT, thus reducing the importance of torsion about the meso position. This difference in electronic structure and wavefunction overlap of the DIPYR and α-DIPYR chromophores seems to have a large effect on the efficacy of SBCT. That is, the BODIPY dimers undergo SBCT very efficiently with an intermediate PCT state which aids in SBCT formation while the DIPYR dimers seem to get stuck at this intermediate PCT state, at least within the first nanosecond of the excited state. 3.4 Nanosecond Transient Absorption Since SBCT appears to be very slow in bis-α-DIPYR, nanosecond transient absorption was performed to see if in THF and cyclohexane, SBCT is occurring on longer timescales. Ideally both bis-DIPYR and bis-α-DIPYR would have been analyzed, but due to limitations in the nanosecond 108 pump laser, a pump wavelength on resonance with bis-DIPYR was not achievable. This is unfortunate as full SBCT was not clearly observed within a nanosecond in any solvent, regardless of polarity. Though the third trace in MeCN may indicate formation of a new state from the slight decay in SE, it is not clear if that is the case or if this third state is due to some slow diffusive or nuclear relaxation process of the second state formed in bis-α-DIPYR, which is not observed in the other solvent environments, simply due to their respective dielectric responses. Bis-α-DIPYR in both cyclohexane and THF give a four-state sequential model with the additional information provided by nanosecond TA experiments (Figure 4.10 a, b, d, e). The shape of the third SADS observed in cyclohexane and THF is similar to that of the third SADS observed in MeCN where slight absorption or loss of stimulated emission about 570 nm is observed, though again there is no apparent absorption matching that of the anion trace observed in the spectroelectrochemistry experiments (Figure 4.6 b and d). In THF, this third trace occurs in 100s of ps while in cyclohexane this is observed in nanoseconds. These time scales may correspond to slow nuclear relaxation processes where the different timescales are in accordance to the differences in solvation or solvent reorganization energies between cyclohexane and THF. This would explain the small differences in the SE bands as a function of time. The final state observed in the nanosecond TA experiments is formed over several to tens of nanoseconds and has lifetimes of several to tens of microseconds. This final state gives a new ESA from 550 – 650 nm. Based on these time scales, this new state could be formation of a triplet state, where this ESA is the T1→Tn absorption. An interesting observation in bis-α-DIPYR is a new emission feature observed at 77K which is not present at room temperature, which may indicate emission from a triplet state 20 . Unlike the DIPYR monomer where triplet formation is a symmetry allowed process and observed readily, the α-DIPYR monomer required sensitization 109 combined with gated detection for phosphorescence to be observed, and even then, the triplet yield was quite low 21 . That is the α-DIPYR triplet states are energetically unfavorable, so it was unexpected to see phosphorescence readily from the dimer. This implies that the final state formed is most likely a triplet where the most likely mechanism of ISC would be via a charge transfer state. However, the newly formed ESA band (550 – 650 nm) also overlaps with the absorption observed for the radical anion from the above spectroelectrochemistry experiments (Figure 4.11). Potentially, this fourth state is the charge separated state, however, the timescales of formation and decay are very different than any of the systems we have studied thus far, implying that in even nonpolar cyclohexane SBCS can occur. Perylene dimers give charge separated state formation occurring over several nanoseconds where the perylene chromophores have large interchromophore separation. This increased separation both decreases coupling towards SBCS and coupling of the resultant charge separated products, giving and SBCS lifetime of about 1 µs 23 . Here the lifetimes are several to 10s of µs and formation of this fourth state occurs on the order of tens of nanoseconds, which could also be explained by triplet state formation and decay. Formation of this fourth transient component occurs in cyclohexane and THF with a time constant of 13 and 21 ns, respectively. In cyclohexane this state decays with a time constant of about 3 μs (Figure 4.10 a & c) and in THF it decays with a time constant of about 20 μs (Figure 4.10 d & f). 110 Figure 4.10: Nanosecond transient absorption spectra of bis-α-DIPYR in cyclohexane (a) and tetrahydrofuran (d). SADS of bis-α-DIPYR in cyclohexane (b) and THF (e). Spectral slices of bis-α-DIPYR with global analysis fits overlaid in cyclohexane (c) and THF (f). Spectral slices were taken from wavelengths which intersected ESA (dark blue/cyan), GSB (purple) and SE (red). Pump wavelengths for these experiments were centered at 532 nm. An alternative explanation for the four-state model and observed lifetimes where the final product is the SBCS state is by way of an excimer or exciplex state. This has been shown in strongly coupled dimers where excimer formation occurs as an intermediate to charge separation 24,25 . These follow a four-state model from formation of the excimer, followed by nuclear relaxation of the excimer which leads to formation of the charge separated state. A charge transfer resonant enhanced excimer state has been observed in perylene aggregates 26 , which may act as a secondary hypothesis for the solvent dependent excited state character in bis-α-DIPYR. This would explain line shape of the ground state absorption spectrum for bis-α-DIPYR and solvent dependent transient and emissive behavior. This excimer with charge transfer character could also significantly slow down the time scales for SBCS, even in very polar solvents, as it too would be 111 stabilized. It may also require a larger degree of activation in order to go from an exciplex state to a charge separated state both due to required intramolecular coupling of the α-DIPYR chromophores and coupling of the solvent environment to drive SBCS. Figure 4.11: Femtosecond and nanosecond TA time traces of bis-a-DIPYR in cyclohexane (a) and THF (b) with the scaled anion absorption overlaid. 4. Conclusions DIPYR dimers were used to increase conjugation in order to decrease the photon energy requirement of the S0 – S1 transition and in principle reduce the overall energy requirement to generate mobile charges. However, unlike what was expected, these DIPYR systems did not behave like the bis-BODPIY systems. This in part is due to the differences in electronic structure between the DIPYR and BODIPY systems, where coupling of the two chromophores which comprise the dimers interact differently or couple differently relative to that of the BODIPY dimers. BODIPY dimers showed rapid formation of an SBCT state which occurred via an intermediate PCT state, while the DIPYR dimers did not undergo complete SBCT within a nanosecond. 112 Figure 4.12: Proposed Jablonskii diagram for excited state evolution of bis-DIPYR and bis-α-DIPYR. The rate constants denoted by k 1and k -1 denote the formation and back transfer of the partial charge transfer, or excimer state with charge transfer character, to and from the initial LE state. From formation of this second intermediate state, k 2 denotes relaxation of this intermediate charge transfer state where k 2-0 gives the recombination from this intermediate to the ground state. Following this, k 3 gives formation of the charge separated state followed by decay to the ground denoted by k 3-0. Table 4.3:Rate Constants Determined by Femtosecond and Nanosecond Transient Absorption Experiments Solvent 1/k 1 (k -1 ) (ps) 1/k 2 (ps) 1/k 2-0 (ps) 1/k 3 (1/k 3-0 ) (ns) bis-DIPYR Cyclohexane THF Acetonitrile PMMA 2.0 (140) 1.2 (43) 0.8 200 -- 3000 3000 1000 4000 -- bis- α-DIPYR Cyclohexane THF Acetonitrile Polystyrene 3.7 (6.7) 2.2 1 500 4600 270 90 4000 4000 -- 4000 13(2700) 21 (27000) Bis-DIPYR fit to a two-state equilibrium model in cyclohexane and THF where the rate of formation of the second trace increases with increasing solvent polarity. In acetonitrile this second state is formed faster than in cyclohexane and THF, and, unlike in cyclohexane and THF, the acetonitrile data did not require a back transfer to the LE state to best model the data. The second 113 state in acetonitrile is similar in character to the second traces observed in cyclohexane and THF. Due to the solvent dependent formation of this second trace as well as the solvent dependent emission spectra, we can say that the second trace may have charge transfer character. By simulating the transient spectra by incorporating the steady-state emission lines of the DIPYR monomer and DIPYR dimer as well as the cation and anion absorption determined from pulsed radiolysis measurements, it can be concluded that the second trace does no encompass the radical cation and anion. That is to say bis-DIPYR, unlike the bis-BODIPY dimers, populates a partial charge transfer intermediate and does not evolve beyond this within a nanosecond. Benzannulation of the DIPYR structure to obtain bis-α-DIPYR was employed primarily to reduce the S0 →S1 excitation energy. This change however leads to a nontrivial change in the excited state behavior; even in cyclohexane the excited state dynamics requires a four-state sequential model at minimum to fit the spectral evolution out into the microsecond regime. Previously, we reported that complete SBCT was occurring in cyclohexane, unlike any other bis- chromophoric system, allowing for relatively high triplet yields. However, upon revisiting these results with insights on intermediate PCT states learned from directly linked BODIPY dimers, the loss of the LE SE band and gain of the dimer-like fluorescence band in the SE is instead now better assigned as PCT state formation. Like bis-DIPYR, it appears that within the first nanosecond only formation of the PCT state is observed. However, with additional nanosecond experiments and the formation of the anion absorption band over 10s of nanoseconds, we can say that SBCS occurs very inefficiently, where the lifetime of the SBCS is several to tens of microseconds, depending on the solvent. Here we claim that a potential mechanism for SBCS in bis-α-DIPYR occurs by way of population of an intermediate excimer state which greatly reduces the efficacy of charge 114 separation. Based on the slight solvent dependent formation of the second state and solvent dependent emission, it can be said that this second state has charge transfer character and can be thought of as an excimer-like species with charge transfer character. From there, in both cyclohexane and THF, it is believed that the final state is the SBCS state which occurs on slower timescales in THF than cyclohexane. This might be due to the increased stability of a partial charge transfer intermediate in THF relative to cyclohexane where solvent and intramolecular reorganization may increase the activation energy required to go from the PCT or charge transfer enhanced excimer state to the fully charge separated product. 115 Chapter 4 Bibliography (1) Lee, H.; Berezin, M. Y.; Henary, M.; Strekowski, L.; Achilefu, S. Fluorescence Lifetime Properties of Near-Infrared Cyanine Dyes in Relation to Their Structures. J Photochem Photobiology Chem 2008, 200 (2–3), 438–444. https://doi.org/10.1016/j.jphotochem.2008.09.008. (2) Kippelen, B.; Brédas, J.-L. Organic Photovoltaics. Energ Environ Sci 2009, 2 (3), 251–261. https://doi.org/10.1039/b812502n. (3) Lu, N.; Li, L.; Sun, P.; Liu, M. Short-Circuit Current Model of Organic Solar Cells. Chem Phys Lett 2014, 614, 27–30. https://doi.org/10.1016/j.cplett.2014.08.070. (4) Shockley, W.; Queisser, H. J. Detailed Balance Limit of Efficiency of P‐n Junction Solar Cells. J Appl Phys 1961, 32 (3), 510–519. https://doi.org/10.1063/1.1736034. (5) Elumalai, N. K.; Uddin, A. Open Circuit Voltage of Organic Solar Cells: An in-Depth Review. Energ Environ Sci 2016, 9 (2), 391–410. https://doi.org/10.1039/c5ee02871j. (6) Antennas and Reaction Centers of Photosynthetic Bacteria Structure, Interactions, and Dynamics; Michel-Beyerle, M. E., Goldanskee, V. I., Shafer, F. P., Toennies, J. P., Eds.; Schafer, F. P., Series Ed.; Springer Series in Chemical Physics; 1985; Vol. 42. https://doi.org/10.1007/978-3-642-82688-7. (7) Reaction Centers of Phtosynthetic Bacteria, 1st ed.; Mchel-Beyerle, M.-E., Ed.; Springer Series in Biophysics; Springer-Verlag Berlin Heidelberg, 1990; Vol. 6. https://doi.org/10.1007/978-3-642-61297-8_43. (8) Jonas, D. M.; Lang, M. J.; Nagasawa, Y.; Joo, T.; Fleming, G. R. Pump−Probe Polarization Anisotropy Study of Femtosecond Energy Transfer within the Photosynthetic Reaction Center of Rhodobacter Sphaeroides R26. J Phys Chem 1996, 100 (30), 12660–12673. https://doi.org/10.1021/jp960708t. (9) Wasielewski, M. R. Photoinduced Electron Transfer in Supramolecular Systems for Artificial Photosynthesis. Chem Rev 1992, 92 (3), 435–461. https://doi.org/10.1021/cr00011a005. (10) Chang, C.-H.; Tiede, D.; Tang, J.; Smith, U.; Norris, J.; Schiffer, M. Structure of Rhodopseudomonas Sphaeroides R‐26 Reaction Center. Febs Lett 1986, 205 (1), 82–86. https://doi.org/10.1016/0014-5793(86)80870-5. (11) El-Kabbani, O.; Chang, C. H.; Tiede, D.; Norris, J.; Schiffer, M. Comparison of Reaction Centers from Rhodobacter Sphaeroides and Rhodopseudomonas Viridis: Overall Architecture and Protein-Pigment Interactions. Biochemistry-us 1991, 30 (22), 5361–5369. https://doi.org/10.1021/bi00236a006. 116 (12) Bradforth, S. E.; Jimenez, R.; Mourik, F. van; Grondelle, R. van; Fleming, G. R. Excitation Transfer in the Core Light-Harvesting Complex (LH-1) of Rhodobacter Sphaeroides: An Ultrafast Fluorescence Depolarization and Annihilation Study. J Phys Chem 1995, 99 (43), 16179–16191. https://doi.org/10.1021/j100043a071. (13) Grondelle, R. van; Dekker, J. P.; Gillbro, T.; Sundstrom, V. Energy Transfer and Trapping in Photosynthesis. Biochimica Et Biophysica Acta Bba - Bioenergetics 1994, 1187 (1), 1–65. https://doi.org/10.1016/0005-2728(94)90166-x. (14) Brinkert, K. Energy Conversion in Natural and Artificial Photosynthesis. 2018. https://doi.org/10.1007/978-3-319-77980-5. (15) Kumpulainen, T.; Lang, B.; Rosspeintner, A.; Vauthey, E. Ultrafast Elementary Photochemical Processes of Organic Molecules in Liquid Solution. Chem Rev 2016, 117 (16), 10826–10939. https://doi.org/10.1021/acs.chemrev.6b00491. (16) Trinh, C.; Kirlikovali, K.; Das, S.; Ener, M. E.; Gray, H. B.; Djurovich, P.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Charge Transfer of Visible Light Absorbing Systems: Zinc Dipyrrins. J Phys Chem C 2014, 118 (38), 21834–21845. https://doi.org/10.1021/jp506855t. (17) Bartynski, A. N.; Gruber, M.; Das, S.; Rangan, S.; Mollinger, S.; Trinh, C.; Bradforth, S. E.; Vandewal, K.; Salleo, A.; Bartynski, R. A.; Bruetting, W.; Thompson, M. E. Symmetry- Breaking Charge Transfer in a Zinc Chlorodipyrrin Acceptor for High Open Circuit Voltage Organic Photovoltaics. J Am Chem Soc 2015, 137 (16), 5397–5405. https://doi.org/10.1021/jacs.5b00146. (18) Kellogg, M.; Akil, A.; Ravinson, D. S. M.; Estergreen, L.; Bradforth, S. E.; Thompson, M. E. Symmetry Breaking Charge Transfer as a Means to Study Electron Transfer with No Driving Force. Faraday Discuss 2019, 216 (0), 379–394. https://doi.org/10.1039/c8fd00201k. (19) Whited, M. T.; Patel, N. M.; Roberts, S. T.; Allen, K.; Djurovich, P. I.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Intramolecular Charge Transfer in the Excited State of Meso -Linked BODIPY Dyads. Chem Commun 2011, 48 (2), 284–286. https://doi.org/10.1039/c1cc12260f. (20) Golden, J. H.; Estergreen, L.; Porter, T.; Tadle, A. C.; R, D. S. M.; Facendola, J. W.; Kubiak, C. P.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Charge Transfer in Boron Dipyridylmethene (DIPYR) Dimers. Acs Appl Energy Mater 2018, 1 (3), 1083–1095. https://doi.org/10.1021/acsaem.7b00214. (21) Golden, J. H.; Facendola, J. W.; R, D. S. M.; Baez, C. Q.; Djurovich, P. I.; Thompson, M. E. Boron Dipyridylmethene (DIPYR) Dyes: Shedding Light on Pyridine-Based Chromophores. J Org Chem 2017, 82 (14), 7215–7222. https://doi.org/10.1021/acs.joc.7b00786. (22) Machan, C. W.; Sampson, M. D.; Chabolla, S. A.; Dang, T.; Kubiak, C. P. Developing a Mechanistic Understanding of Molecular Electrocatalysts for CO2 Reduction Using Infrared 117 Spectroelectrochemistry. Organometallics 2014, 33 (18), 4550–4559. https://doi.org/10.1021/om500044a. (23) Aster, A.; Licari, G.; Zinna, F.; Brun, E.; Kumpulainen, T.; Tajkhorshid, E.; Lacour, J.; Vauthey, E. Tuning Symmetry Breaking Charge Separation in Perylene Bichromophores by Conformational Control. Chem Sci 2019, 10 (45), 10629–10639. https://doi.org/10.1039/c9sc03913a. (24) Cook, R. E.; Phelan, B. T.; Kamire, R. J.; Majewski, M. B.; Young, R. M.; Wasielewski, M. R. Excimer Formation and Symmetry-Breaking Charge Transfer in Cofacial Perylene Dimers. J Phys Chem 2017, 121 (8), 1607–1615. https://doi.org/10.1021/acs.jpca.6b12644. (25) Yu, Y.; Chien, S.-C.; Sun, J.; Hettiaratchy, E. C.; Myers, R. C.; Lin, L.-C.; Wu, Y. Excimer- Mediated Intermolecular Charge Transfer in Self-Assembled Donor–Acceptor Dyes on Metal Oxides. J Am Chem Soc 2019, 141 (22), 8727–8731. https://doi.org/10.1021/jacs.9b03729. (26) Kim, W.; Nowak-Król, A.; Hong, Y.; Schlosser, F.; Würthner, F.; Kim, D. Solvent- Modulated Charge-Transfer Resonance Enhancement in the Excimer State of a Bay-Substituted Perylene Bisimide Cyclophane. J Phys Chem Lett 2019, 10 (8), 1919–1927. https://doi.org/10.1021/acs.jpclett.9b00357. 118 Chapter 5: Pump-probe anisotropy of Symmetry Breaking Charge Transfer (SBCT) systems 1. Introduction In the previous two chapters, symmetry breaking charge transfer (SBCT) was an explored excited state phenomenon in systems with different parent chromophores. SBCT is an excited state process which occurs in systems containing identical coupled chromophores where excitation of the dimer system results in an initial localization of the excitation followed by separation of the excited electron and hole pair across the dimer, producing a charge transfer state 1 –7 . In the case of Chapters 3 and 4 the systems studied were dimers containing dipyrrinato (DIP), borondifluoro dipyrromethene (BODIPY) and borondifluoro dipyridylmethene (DIPYR) chromophores. SBCT can be used in device applications for solar energy conversion as it would be a means to reduce the exciton binding energy of a system at a donor-acceptor (D/A) interface, thus increasing the rate of charge separation while minimizing energy loss at the interface. In both Chapters 3 and 4, an intermediate state or partial charge transfer state (PCT) was observed as either a precursor towards efficient SBCT in the directly linked BODIPY dimers, or as a potential intermediate for charge separation in the directly linked DIPYR dimer systems, where SBCS is very slow. In the BODIPY directly linked dimers, SBCT was observed on femto-to-picosecond timescales and occurred via a PCT intermediate. In the case of the BODIPY dimers, this intermediate enabled full SBCT as it energetically behaved like a z-scheme observed within the photosystem of photosynthetic organisms 8 , where there is a step down in energy going from the LE to PCT to full SBCT. In the DIPYR dimers, however, this intermediate behaved more akin to an excimer state 9 –11 which may have some enhanced charge transfer character. That is, it acts to either inhibit or slow the process of charge transfer or separation to the nanosecond timescale where evidence of a charge separated species was only observed in bis-α-DIPYR within tens of 119 nanoseconds. Interestingly, the anion absorption spectrum, which indicates charge separation, was observed even in nonpolar cyclohexane. The chromophores that comprise the BODIPY and DIPYR dimers vary primarily with the ring structures of the nitrogen containing aromatics. BODIPY contains a dipyrromethene where the dipyrromethene comprises two 5-member nitrogen containing rings that are bridged by a borondifluoro group at the nitrogen terminals 12 . The DIPYR contains a dipyridylmethene group where the dipyridylemethene group comprises two 6-membered nitrogen containing rings that are bridged by a borondifluoro group at the nitrogen terminals 13 . The HOMO and LUMO structures for both BODIPY and DIPYR were calculated where the character of the HOMO and LUMO were different for each chromophore. In the case of the BODIPY dye (Chapter 3, Figure 3.5), the HOMO gives a node at the meso carbon while the LUMO shows significant electron density at the meso carbon 2,6 . The structure of the LUMO, which is similar to the anion structure in Chapter 3, was used to understand why torsion about the meso bridge would aid in rapid formation of the SBCT state while also reducing the lifetime of the SBCT state. In the DIPYR structures (Chapter 4, Figure 4.9), however, there is a large degree of electron density on the meso bridge in the HOMO of the DIPYR dimers while the LUMO structure shows almost no electron density on or near the bridge 4,13 . These differences in molecular orbitals may indicate why in the case of the BODIPY dimers the intermediate state enables rapid formation of the SBCT state, while in the DIPYR dimers, the intermediate state is rapidly formed but in all cases full SBCT or SBCS is not observed within a nanosecond of the excited state lifetime. 9,9’-bianthryl (BA) gives electron density at the meso carbon in both its HOMO and LUMO structures 14 , where torsion is known to occur within the LE state of BA and allows curve crossing to the ultimate SBCT state 15 –18 . This electron density along the bridge has also been shown to produce an intermediate partial charge 120 transfer state 19 , which in BA enables rapid formation of the SBCT state. It is this delocalization of electron density along the bridge which connects the chromophores. So, in the case of the BODIPY, where the LUMO gives electron density on the bridge, this acts as a precursor to motion of that electron density across the bridge while in the DIPYR the LUMO gives no electron density which may act to inhibit movement of charge density along the bridge to the other chromophore. The presence of this intermediate state poses an interesting question; though these are symmetric systems where, in principle, there should be no preferential direction for electron or hole transfer, does the presence of this state indicate that one charged particle will move from the initial LE as opposed to the other? It may be that the efficiency of SBCT (SBCS) is dependent on whether it is the hole or electron that is moving from the initial locally excited state. It may also have to do with the relative orientation of the transition moments which allow interconversion or population transfer between electronic states. In the case of the BODIPY dimers it is apparent that the PCT state is a new, distinct electronic state, as there is an isosbestic point in the transient absorption data. However, no such isosbestic point is observed in the DIPYR dimers, just the same characteristic growth and decay of the dimer and monomer emission bands, respectively, similarly observed in BODIPY dimers. The lack of isosbestic point may mean that in the case of the DIPYR dimers, the PCT state is more like a ‘relaxed LE’ state that has charge transfer character, or it may be a type of excimer state that has some charge transfer enhancement. One technique that has been implemented to study charge transfer systems is pump-probe anisotropy 19 –23 . Pump-probe anisotropy essentially utilizes the principle that transition dipole moments have an orientation within a molecular frame where this transition dipole couples electronic states induced by the incident light. The orientation or differences in orientation of the transition moments between the BODIPY and DIPYR systems may allow us to understand 121 differences in symmetry and behavior of the excited state processes. Herein, this technique will be implemented with the goal of indicating whether there is a preferential hole or electron transfer from the observed PCT state. This may also give insight into the very different efficiencies of SBCT between the DIPYR and BODIPY dimers in terms of the relative orientations of the PCT transition moments which may aid in further characterizing the PCT state or determine differences in intermediate states observed in the directly linked BODIPY and DIPYR dimers. 2. Pump-Probe Anisotropy Principles The following description of anisotropy is based on concepts taken from Kelley and Lakowicz (Ref. 24 and 25). Pump-probe anisotropy uses the principle of photo-selection induced by using a polarized, coherent light source. Coupling of electronic states uses transition dipoles to allow non-zero population of the new or final electronic state, where these transition moments have a defined orientation within a molecule. The transition dipoles for the molecules that are excited are oriented preferential by the linear polarization of the incident electric field, even though the molecules are oriented randomly and isotropically in the starting solution. Thus, a pump pulse photo-selects a oriented sub-ensemble that can be probed 24 . When the pump pulse is introduced to electronically excite the sample, usually this is resonant with the S0→S1 transition, it will photo select the molecules whose transition moments have some projection along the polarization of the pump-pulse (Figure 5.1 a). This photo selection gives an angular distribution, f(θ), of these transition moments described by the following equation 25 𝑓 (𝜃 )=𝑐𝑜𝑠 2 (𝜃 ) where θ is the angle between the transition moment vectors and the polarization axis. Following this photo-selection process, a probe pulse is introduced either parallel or perpendicular to the pump where the probe will interact with different electronic state transitions, be it emission or 122 absorption, providing the relative orientation of those transition moments with respect to the orientation of the pump-induced transition moment. Figure 5.1: a) Illustration of photo selection with the introduction of a pump pulse. Initially the sample is isotropically distributed, depicted as a sphere. Following this a pump pulse will be introduced to the initially ground state sample and select the chromophores whose transition moments have a projection on the polarization axis giving the cos 2 distribution shown on the right. From there a probe is introduced, polarized parallel or perpendicular to the photo-selected sample. b) Transition moments on the BODIPY chromophore where the angle θ within the laboratory frame in a) will indicate the angle theta between the pump-induced transition dipole with the probe induced transition moment. c) Diagram of electronic states showing the transition moments attributed to the pump (µ pump), ground state bleach (µ GSB), excited state absorption (µ ESA) and stimulated emission (µ SE). 123 From the distribution within the sample or laboratory frame, we can compare the orientation of the probe-induced transitions with respect to the pump-induced transition moment within the frame of the molecule. Thus, it is important that the pump-induced transition is of a known orientation within the molecular frame, providing a reference for the molecular orientation of the probe-induced transitions. By taking the difference between the parallel (𝐼 ∥ (𝜆 ,𝑡 ) ) and perpendicular (𝐼 ⊥ (𝜆 ,𝑡 ) ) amplitudes and dividing by the isotropic or total intensity we can calculate the anisotropy (𝑟 (𝜆 ,𝑡 ) ) as a function of its spectral, λ, and temporal, t, components 24,25 . 𝑟 (𝜆 ,𝑡 )= 𝐼 ∥ (𝜆 ,𝑡 )−𝐼 ⊥ (𝜆 ,𝑡 ) 𝐼 ∥ (𝜆 ,𝑡 )+2𝐼 ⊥ (𝜆 ,𝑡 ) By looking at the above equation we can discern that if parallel transitions (e.g., S1 S0 absoprtion and S0 S1 stimulated emission) will give larger intensity in the parallel pump-probe signal relative to the perpendicularly probed intensity (that is 𝐼 ∥ >𝐼 ⊥ ) and will result in a positive anisotropy. Conversely, perpendicular transitions will give a negative anisotropy. We can then define the parallel and perpendicular intensities in terms of the angular distribution of the transition moments relative to the polarization axis 25 , denoted by θ. 𝐼 ∥ =∫ 𝑓 (𝜃 )𝑐𝑜𝑠 2 (𝜃 )𝑑𝜃 𝜋 /2 0 =𝑘 <𝑐𝑜𝑠 2 (𝜃 )> 𝐼 ⊥ = 1 2 ∫ 𝑓 (𝜃 )𝑠𝑖𝑛 2 (𝜃 )𝑑𝜃 𝜋 /2 0 = 𝑘 2 <𝑠𝑖𝑛 2 (𝜃 )> This gives the expectation value of a transition with respect to a given orientation. By putting the parallel and perpendicular intensities in the form of their angular distributions, the anisotropy can also be described in terms of its angular distribution 25 . 𝑟 (𝜃 )= 2 5 ( 3<𝑐𝑜𝑠 2 (𝜃 )> 2 − 1 2 ) 124 This θ comes from the distribution of probe induced transition moments as they relate to the pump selected transition moments. This inherently gives an angle of the probe-induced transition relative to the pump-induced transition within the molecular frame (Figure 5.1 b). From this we can conclude the two extreme cases where the probe transition is parallel or perpendicular. When the probe transition in parallel, θ = 0, r(0) = 0.4 and when the probe transition is perpendicular, θ = 90º, r(90º) = -0.2. As shown in Figure 5.1 c these transition moments describe coupling between two electronic states, leading to population of a final state. It can be seen that in pump-probe anisotropy, which compares two transient absorption signals, one with the pump and probe parallel and another with them perpendicular, the transition moments will be observed in the ground state bleach (GSB) stimulated emission (SE) and excited state absorption (ESA). The GSB and SE, so long as the SE is from the initially populated state, couple the same two states as the pump, and at initial times, or t = 0, these features should give anisotropies of 0.4, as these transitions will be parallel. However, the ESA need not be parallel or perpendicular as the molecular frame transition dipole orientations depend on the symmetries of the initial and final electronic states. As the excited state of the chromophore evolves, where interconversion between different electronic states can occur, the orientation of the SE and ESA bands will change and help give insight into orientations of these new states as they are populated. One thing to also keep in mind is that in the case of the systems to be analyzed, the molecules are free to rotate in solution. The molecules will then reorient themselves over time back to an isotropic distribution, where the anisotropies of all states will become 0 (𝑟 𝑡 =∞ (𝜃 )=0). The two extreme cases of parallel and perpendicular anisotropies described above correspond to initial times, where the system has not begun to reorient. The time in which the system reorients 125 will depend on the solvent (usually viscosity or other intermolecular/solvation properties) and the size of the chromophore in question 24 . To both understand these timescales as well as establish polarization purities for the experiment, monomers which simply populate the pump-induced excited state and decay back to the ground can be used as a control or calibrator. 3. Anisotropy in SBCT Compounds As discussed in the previous section, pump-probe anisotropy can give insight into the orientations of transition moments within a molecule relative to a known pump-induced transition moment. Here we use this principle to discern whether or not the electron or hole is moving by tracking the orientation of the absorptions which correspond to the anion and cation. In the dimer systems investigated, the pump-induced transition moment forms a localized excited state, where the transition moment is along the long-axis of the DIPYR/dipyrrin molecules (Figure 5.1 c and Figure 5.2). In the dimers of interest, the dimers have orthogonal moments, which provide a tool to determine the direction of charge transfer. 126 Figure 5.2: Configuration diagram showing the locally excited (LE) state and the expected absorption processes for the (ground state) radical cation and anion formed by SBCT. The transition corresponding to local HOMO-LUMO promotion is the same for the neutral fragment as for the cationic and anionic fragments of the dimer, and therefore is expected to have the same transition dipole vector direction. The orientation of the transition moments are shown for the LE state (green arrow), the anion absorption (red arrow) and the cation absorption (blue arrow). First, it should be understood that the transitions resulting in the product state absorption (we will still call this ESA) of the charge transfer state are assumed to be the sum the absorption transitions belonging to the ground state radical cation and anion. One assumption that is made is that within the spectral range of interest, the absorption processes for the resultant cation and anion, involve the same HOMO to LUMO transitions (Figure 5.2) and the transition energy is simply shifted due to the different overall charge of the fragment. This implies that if the electron or hole moves, the absorption of the destination fragment will give a perpendicular transition relative to the pump-induced LE state, giving a negative anisotropy. For instance, if a hole is transferred, that cation will be localized on the chromophore perpendicular to that which had the initial localized excitation, thus resulting in a perpendicular anisotropy (Figure 5.2, top). The remaining anion will 127 then give a parallel anisotropy as the anion will remain on the chromophore which had the local excitation. Conversely, if an electron is transferred, the anion will give a perpendicular anisotropy and the cation will give a parallel (Figure 5.2, bottom). Herein we will first discuss the experimental apparatus for measuring transient absorption anisotropy. Then we will discuss the anisotropy of the relevant monomers which were used to calibrate the set up as well as show the reorientation times within the monomer systems. We will proceed with the assumption that the SBCT dimers are expected to reorient on the same timescale or slower than the monomers which comprise them. Following the monomers, the anisotropy of the SBCT dimers will be compared between zDIP2, which is a well-behaved systems that follows a two-state model towards SBCT, 1,3,5,7-tetramethyl-bis-BODIPY (m8B) and bis-DIPYR, which have shown the population of an intermediate ‘partial charge transfer’ state. From the analysis it will be discussed whether or not there is a preferential direction towards charge transfer and whether or not the intermediate PCT state results in said preference. Following this, a new pump- probe anisotropy set up will be proposed to potentially account for differences in parallel and perpendicular amplitudes caused by uncorrelated changes within the experiment. 4. Experimental Apparatus Anisotropy experiments were carried out using the same transient absorption set up described in Chapter 1. In earlier chapters, transient absorption experiments were performed such that pump and probe polarizations were oriented at magic angle (54.7º) to extract population dynamics of the excited state processes without reorientation of the system potentially compromising the kinetics. Here we are intentionally avoiding magic angle so that the experiment is sensitive to the orientation of different transitions. 128 Figure 5.3: Experimental set up for polarization dependent transient absorption spectroscopy. The pump is sent through a Glan-Taylor polarizer to purify the polarization which is horizontal. Following this the pump is sent through an achromatic λ/2 waveplate so that the pump can be rotated between horizontal and vertical polarizations. The 800 nm pump used for white generation of the probe is sent through a λ/2 waveplate to set the white light polarization as horizontal. Rotating the pump between vertical and horizontal give the perpendicular and parallel configurations, respectively, for the anisotropy experiments. From this the white light is sent to a grating which disperses the light in the horizontal plane onto a 256-pixel silicon diode array. Further description of the transient absorption set up can be found in Chapter 1. The parallel and perpendicular measurements were taken sequentially in two separate TA experiments. The 800 nm used to pump a CaF2 crystal for white light generation was sent through a λ/2 waveplate such that the 800 nm was oriented to a horizontal polarization. The probe was left horizontally polarized as the grating within the spectrograph more efficiently disperses light along the horizontal polarization. The pump polarization was first purified by a Glan-Taylor polarizer, where the pump is horizontally polarized. This is then sent through an achromatic zero-order λ/2 waveplate where the pump is rotated between the horizontal and vertical polarization. This gives the parallel configuration where the pump and probe are both horizontal and the perpendicular is given when the pump is vertical and probe horizontal. The anisotropy was calculated by post-processing for each molecular system characterized. Initially, pump-probe anisotropy experiments were done using a γ-DIPYR monomer which gives trivial excited state behavior, that is population of the S1 state followed by direct decay back to the ground state and was used to verify polarization purity. This should result in anisotropy within the 129 GSB and SE parts of the transient spectrum to be initially 0.4 and then decay to zero at long times due to molecular tumbling in the liquid. The ESA features, as stated previously, need not give parallel transition moments, but the anisotropy at long times for all features across the spectrum should approach zero. Once the polarizations were verified for the monomer and reproduced, the SBCT/SBCS dimer systems were studied using the same experimental set up determined by the monomer experiments. 5. Anisotropy of Relevant Monomers By examining the transient spectra at different time points, we can see that the parallel and perpendicular transient spectra give different amplitudes where the amplitudes appear to decay over different periods of time. Figure 5.4: Parallel (a) and perpendicular (b) transient absorption spectra of gamma-DIPYR (molecular structure shown on the right) in tetrahydrofuran (THF). c) Parallel (red line) intensities of the GSB (square) and SE (circle) plotted as a function of time along with the perpendicular (blue line) intensities of the GSB (upward triangle) and SE (downward triangle) as a function of time. d) Anisotropy of the GSB (blue 130 squares) and SE (green circle) as a function of time. It can be seen that the anisotropies of the GSB and SE follow the predicted behavior where at early times the anisotropy is 0.4, indicating they are parallel transitions to the pump-induced absorption, and 0 at long times due to reorientation of the photo-selected sub-ensemble back to an isotropic distribution. The spectral slices are qualitatively the same, where the ESA is observed from 345 – 450 nm, GSB from 450 – 500 nm and SE from 500 – 575 nm. It can also be seen that the features of the transient traces in both the parallel and perpendicular spectra are invariant, indicating that there is no population of a new excited state (Figure 5.4 a and b). If we plot wavelength slices which intersect the SE and GSB for both the parallel and perpendicular spectra as a function of time we see that the initial amplitudes are very different, but at longer times the spectra converge to the same amplitude (Figure 5.4 c). This is due to molecular reorientation: the evolution from the initial photo-selected distribution to an isotropic distribution. From these intensities we can calculate the anisotropy as a function of time. In Figure 5.4 d the anisotropy of the GSB and SE band are plotted where both start with an initial anisotropy of 0.4 (within the signal to noise) and decay to 0 at long times. This indicates that the polarization purity of our laser beams are good, and the expected anisotropic behaviors are observed. 131 Figure 5.5: Magic angle transient absorption spectra of DIPYR monomer (a) and 1,3,5,7-tetramethyl-8- phenyl-BODIPY (b) in tetrahydrofuran. Anisotropy of DIPYR monomer (c) and 1,3,5,7-tetramethyl-8- phenyl-BODIPY (d) versus time with spectral slices taken from the ESA (black square) GSB (red circle) and SE (dark cyan triangle). As observed for the gamma-DIPYR, the anisotropies indicate that all transitions at early delay are overall parallel to the pump-induced transition moment. At long times, the anisotropy decays to zero, following the expected behavior of monomers in solution. Next, we will discuss the anisotropy of the DIPYR and BODIPY monomers dissolved in THF to see if there is any unexpected behavior of the monomers that comprise the SBCT dimers to be analyzed. The BODIPY monomer is the 1,3,5,7-tetramethyl-8-phenyl-BODIPY (Figure 5.5 a and c) discussed in Chapter 3 while the DIPYR monomer is the unsubstituted borondifluoro- dipyridylmethene monomer which comprises bis-DIPYR (Figure 5.5 b and d). In the anisotropy, the DIPYR monomer shows a and initial decay to 0.3 within the first 500 fs following this, the 132 anisotropy decays to zero by about 100 ps. Similarly, the BODIPY monomer shows an initial rapid decay in the anisotropy to 0.3 within the first 100 fs followed by similar decay times as the other monomers. These initial fast decays in the anisotropy could be due to some vibrational relaxation process, though it is unclear what the source to this initial decay is. Dephasing can occur due to vibronic coupling within a chromophore giving initial, rapid decay in the anisotropy 26,27 , so in the case of the monomers, this initial decay may be due to vibrational coupling or relaxation, though, again, this has not been established. The monomers indicate that initial dephasing to 0.3 in the anisotropy, could be due to vibronic coupling of the excited state or nuclear relaxation of the local exciton within the dimers. However, decays which involve large changes in anisotropy may be attributed to either electronic coupling between the chromophores in the dimer or due to population transfer as PCT and SBCT proceed. After about 10 ps, all systems will undergo free tumbling, with timescales depending on molecular size and shape as well as the solvent viscosity, that will lead to decays in the anisotropy of the dimers. Because PCT and SBCT are formed on timescales much faster than 10 ps, changes in anisotropy can be effectively associated with the formation of these charge transfer states. For full SBCT the ESA features are the approximate absorptions attributed to the ground state cation and anion. Based on the above explanation, this should give a handle on discerning which charged particle is moving, the electron or the hole. 6. Anisotropy of SBCT Dimers 6.1 Zinc-dipyrrinato dimer (zDIP2) In Chapter 3 the different dipyrrin dimers were overviewed in terms of structural effects on the rate of SBCT and lifetime of the SBCT state. It was shown that the zinc dipyrrinato dimer was described by a two-state model where the initial state is the localized excited (LE) state 133 followed by formation of the SBCT state where the SBCT state is in the decoupled limit, which is technically charge separation (SBCS). This is based on the fact that the emissive state is the LE state where the photoluminescent quantum efficiency decreases with increasing solvent polarity 3,6 . The behavior of the emission combined with observed triplet formation 3 indicate that the radical cation and anion formed are uncoupled and therefore the state formed is charge-separated. Within the analysis of the anisotropy of zDIP2, we will assume that the dipyrrinato dimers behave as independent chromophores. That is, they will be treated effectively as orthogonal monomers which have little to no electronic interaction with one another. Figure 5.6: Magic angle transient absorption spectra of zDIP2 in THF with the anion absorption from pulsed radiolysis measurements overlaid (red dashed line). b) Anisotropies of the ESA attributed to the formation of the SBCT state. c) Anisotropies of the ESA (black square), GSB (dark red circle) and SE (dark cyan triangle) which are characteristic of the LE trace. d) zDIP2 molecule. 134 The zinc dipyrrinato (zDIP) systems do not show formation of an intermediate or partial charge transfer state prior to SBCS. That is because the zinc bridge interrupts any through-bond linkage between the dipyrrinato chromophores where charge separation proceeds by way of a through-space interaction between the two dipyrrinatos. Here we can establish whether in the case of the zDIP2, where there is no PCT state, there is a larger probability of the electron or hole moving from the initial localized excited state. Before discussing the anisotropy of the charge transfer state, it is important to establish what distinguishes the decoupled SBCT or SBCS state from the initially formed LE state in the transient spectra. Previously our group studied different zinc dipyrrinato dimers as a function of solvent as well as changes in substitution on the dipyrrin chromophore 3 . Using spectroelectrochemistry the spectral signature of a given charged species can be identified. This was done by taking a neutral ground state absorption spectrum followed by a spectrum where a bias was applied then, by taking the difference between the neutral and biased spectra, the anion or cation absorption can be resolved as well as the signature of the bleached neutral. In the case of the zDIP systems, a negative bias was applied and therefore the anion spectrum was determined 3 . The spectral trace looks similar in shape to a transient absorption trace as there is a negative ground state bleach and positive features from new formed species. From this we can attribute the anion absorption in the transient spectra as the absorption observed from 520 – 550 nm. However, the absorption further to the blue (350 – 390 nm) may be another anion transition or it could be the cation absorption (Figure 5.6 a). We have further reinforced that the anion absorption is the absorption from 520 – 550 nm as it is consistent with the anion spectrum measured from pulsed radiolysis measurements (Figure 5.6 a, red dashed line). 135 Since the anion spectrum has been established, we can look at the anisotropy of the effectively characterized anion absorption and we can compare to the absorption of the other SBCT absorption feature to the blue, centered at ~375 nm. This gives the anisotropy observed in Figure 5.6 b and c which is plotted as a function of time. What we see is on the timescale of SBCT, there is a rapid decay which asymptotes to an anisotropy of about 0.2. Previously, Vauthey and group also observed an anisotropy of about 0.2 for a perylene dimer system which undergoes SBCS 20 . They attributed their result to an equivalent probability of the electron and hole moving across the system. That is, the anisotropy of the absorption of a single charge species will be 0.2, because there is an equal probability of the charged species both staying on the chromophore of the initially localized excited state and moving to the neighboring chromophore resulting in an anisotropy of 0.2. That is the total anisotropy would be the sum of the parallel anisotropy and perpendicular anisotropy, or 0.4 + -0.2, giving an anisotropy of 0.2. However, in the case of the perylene dimers, the transient absorption of the cation and anion were superimposed with each other as well as the ESA of the LE state 20 . This can lead to complications when trying to attribute changes in anisotropy as the recorded anisotropy will have contributions from several excited state processes. However, in the case of zDIP2, the band centered at about 530 nm does not appear to be superimposed with any other spectral features of different absorption/emission processes. That is the SE attributed to the monomer emission has little to no amplitude by about 5 ps, indicating that the ESA at 375 nm and 530 nm are nearly pure ESA of the CT state. Furthermore, in both bands the trace gives an anisotropy of about 0.2 at longer times. Though the band at 375 nm has yet to be assigned, we can infer that this transition is oriented along the same direction as the band at 530 nm. That is to say, that in order for the anisotropy to be 0.2 for both bands, the transition moment must be along the long axis of the DIP chromophore, which 136 is the same as the S0→S1 transition. If this state were to a higher lying anion absorption, say the S0→S2, this transition should give an entirely perpendicular, or negative, anisotropy. Thus, as depicted in Figure 5.2 the bluer CT band centered at 375 nm may be the HOMO to LUMO transition of the cation, which would be along the same direction of the pump-induced transition moment. Of course, this is not conclusive as experiments determining the cation absorption within this spectral range have yet to be done. Thus far, however, it can be hypothesized that in the case of zDIP2, electron and hole transfer occur with an equal probability. 6.2 M8B As discussed above for the directly linked dipyrrin dimers, there was an observed intermediate state which precedes SBCT. We have speculated that this intermediate state may resemble the partial charge transfer state observed in BA, where PCT is due to charge delocalization across the meso bridge connecting the anthracene chromophores in the dimer 19 . If this is the case with the BODIPY dimers, it may imply that charge transfer occurs in a preferred direction. That is, the electron or hole will move from the initial excited state which is localized on a single chromophore. Here, we will analyze the anisotropy of the m8B system on different timescales to look at the direction of the transition moments of both the PCT and SBCT state. 137 Figure 5.7: a) Magic angle transient absorption spectra of m 8B in THF with the anion and cation absorption indicated by the red and blue lines, respectively, in accordance to Hattori et. al 28 . b) Anisotropies of the anion (red) and cation (blue) bands, where the {} indicates timescales attributed to PCT formation and [ ] indicate timescales attributed to SBCT formation. c) Anisotropies of the ESA (black square), GSB (dark red circle) and SE (dark cyan triangle) which are characteristic of the LE trace with an inset zooming in on the first picosecond of the anisotropy. d) m 8B molecule. In the case of the zDIP2 system, the zinc ion acts as a scaffold to hold both ligands at right angles, leading to the dipyrrinato chromophores behaving independently of each other. This implies that the zDIP2 system could be treated as a system containing two noninteracting monomers. When the system is weakly coupled where the chromophores give degenerate perpendicular transitions there can be an initial dephasing of the anisotropy, where the rate of dephasing will depend on the degree of coupling, and the anisotropy will asymptote to 0.1 prior to complete randomization. In the case of m8B such dephasing needs to be taken into consideration, 138 as the chromophores are directly linked and therefore are more strongly coupled to one another. Historically, this type of behavior in the anisotropy has been shown where dephasing is attributed to electronic coupling of perpendicular transitions 29 –32 either within a single chromophore or coupled dimers. The anisotropy of coupled chromophores or systems containing identical orthogonal transition moments may lead to anisotropies of less than 0.4 26,31 , where the energy splitting due to electronic coupling is relatively small, or when the pump pulse couples to vibrational modes in the S1 state, giving rapid dephasing beyond the instrument response. Returning to the m8B dimer, the anisotropies of the ESA, GSB and SE transitions shown in Figure 5.7 c are the transitions originating from the S1 state localized on the BODIPY chromophore. None of the anisotropies start at 0.4 and the GSB and SE both asymptote to 0.1. The ESA does not asymptote to 0.1, but rather to an anisotropy of about 0.05 and there is some evidence of beating, or the end of an oscillation, in the anisotropy of the ESA within the first 200 fs. In coupled chromophores, excited state coupling can result in excitonic splitting producing two new eigenstates which consist of the symmetric and antisymmetric combinations of the excitonically coupled pair. The beating-like behavior, specifically in the ESA where the anisotropy dips to near 0 and rises to 0.05 within the first 200 fs, may be attributed to this excitonic splitting where, by about 200 fs, the system collapses to populate one of the new eigenstates formed. As stated previously, the ESA may encompass transitions which are both parallel and perpendicular, so the final anisotropy is a composite over several transitions, explaining why it too does not asymptote to 0.1. From these anisotropies, we can infer that the chromophores, though orthogonal, have some degree of electronic coupling, unlike the dipyrrinato chromophores in zDIP2. There is an added complexity when interpreting anisotropy in coupled chromophores for the above reasons. However, it is important to keep in mind the magic angle data which measures 139 the excited state dynamics as it relates to population of excited states. For m8B in THF, the excited state is described by a three-state model where the initial state is the LE state, which is the pump- induced transition, followed by population of a PCT intermediate followed by complete SBCT. Population of the PCT state occurs with a time constant of 200 fs which is within the time of the anisotropy dephasing (Figure 5.7 b). The decay to 0.1 anisotropy (Figure 5.7 b and c) implies that there is an equal projection of the probe-induced transition moment onto both orthogonal transitions, that is the transition is 45 º with respect to the pump-induced transition, and inherently it is also 45 º with the equivalent orthogonal transition. In order for that to be true the transition moment must be normal to the meso bridge and at 45 º with the long axis of each BODIPY dye. A PCT in likeness that observed in BA, where the transition moment is along the meso bridge, can’t necessarily be ruled out as this would give a transition that is perpendicular to both the molecule where initial exciton was localized and the perpendicular chromophore, which could give equivalent projections onto their orthogonal transitions. However, further consideration of the anisotropies should be done where simulations of the transient anisotropy combined with fluorescence anisotropy measurements, since the PCT state is emissive, could be done to gain further insight into the orientation of this transition. On the time scale of SBCT formation, it can be seen that the bands attributed to the anion and cation absorption behave differently (Figure 5.7 b). The anion band is the most interesting as it appears that for several picoseconds after PCT the anion band is negative, while the cation band remains at an anisotropy of 0.1. The negative anisotropy for the anion absorption implies that the electron has moved to the perpendicular chromophore. Between 1.7 – 2.2 ps, the points were deleted in the anisotropy because the anisotropy diverges in association with the transient absorption amplitudes crossing zero for both the parallel and perpendicular signals, giving an 140 anisotropy, 𝑟 (𝜆 ,𝑡 )= 0 0 . From the PCT and SBCT anisotropies, we can claim that the PCT is due to some transition that has an equivalent projection onto the identical perpendicular transitions within the dimers which is result of electronic coupling between the directly linked chromophores. This PCT state results primary electron transfer, suggesting that the PCT, like in BA, is due to electronic delocalization across the meso bridge. 6.3 Bis-DIPYR Bis-DIPYR is another directly linked system which forms an intermediate state. This was identified by the red-shifted, broadened steady-state emission spectrum in nonpolar cyclohexane, like what was observed for the m8B dimer. However, unlike m8B, the emission experienced little- to-no solvatochromic behavior though the photoluminescence quantum yield decreased with increasing solvent polarity. The solvent dependent behavior of the steady-state emission implies that the bis-DIPYR is undergoing SBCS, as it is in more likeness to that of zDIP2. However, unlike zDIP2, the emissive state is not the localized excited state and there was no evidence of charge separation within a nanosecond of the excited state lifetime. In the previous chapter we related this to the differences in electronic structure between the BODIPY and DIPYR dimers. The character of the molecular orbitals may imply that the intermediate state which has the same characteristic transient trace as the intermediate in m8B, may suppress complete SBCT/SBCS. That is, the intermediate seems to almost act as a trap state in bis-DIPYR while it acts to enhance SBCT in m8B. 141 Figure 5.8: Magic angle transient absorption spectra of bis-DIPYR in THF with the anion (red dashed line) and cation (blue dashed line) absorptions form pulsed radiolysis measurements overlaid. It can be seen in the case of bis-DIPYR, there is no indication of new absorption features of the in the transient spectrum attributed to the formation of the cation and anion, implying that SBCT does not occur. b) Anisotropies of the wavelengths attributed to anion and cation absorption, if they were formed or two spectral slices through the dimer emission component of the SE. c) Anisotropies of the ESA (black circle), GSB (dark red circle) and SE (dark cyan triangle) which are characteristic of the LE trace. d) bis-DIPYR molecule. Here we will look at the anisotropy of bis-DIPYR in THF (Figure 5.8 b and c). Similar to what was observed in the m8B dimer, there is a rapid decay in the anisotropy within the first 300 fs observed in the GSB and SE. The ESA anisotropy has some oscillatory behavior within the first 600 fs, which, like m8B, is indicative of an excitonically split state attributed to close coupling between the DIPYR chromophores. The anisotropy of the GSB asymptotes to 0.1, which is an expected anisotropic behavior of coupled chromophores. However, in the case of the ESA and SE, 142 the anisotropy of 0.2 implies a superposition of perpendicular and parallel transitions. In the case of the bis-DIPYR systems, there are no methyl groups to inhibit the excited state torsion processes which could allow an increase of excited state coupling between the DIPYR chromophores, explaining persistence of the oscillations in the ESA anisotropy. Similar to m8B, the rapid decay or dephasing in the anisotropy of the dimer system can be due to electronic coupling between the two DIPYR chromophores, where the PCT state is a product of said electronic coupling. An anisotropy of 0.2 has been shown to be due to superpositions of parallel and perpendicular transitions. In m8B, the SE, GSB and ESA bands attributed to the anion and cation transitions all gave anisotropy decays to 0.1, however, bis-DIPYR only gives an anisotropy of 0.1 in the GSB. The SE bands are a superposition of both the monomer and dimer emission. The ESA may also be attributed to absorption from the intermediate or PCT state or a superposition of the LE and PCT absorption. In Figure 5.8 b the anisotropies where the cation and anion absorption would be, if they were present in the transient absorption, are from the SE attributed to PCT. This shows that initially, the anisotropy decays to 0.2 but over about 10 ps, this anisotropy decays to 0.1. The behavior observed here would initially suggest that the initial anisotropy shows the excitation either localized on both DIPYR chromophores or potentially delocalized across the dimer. Since torsion is fully allowed, delocalization of the excitation is possible. However, as stated for the m8B system, characterizing this intermediate may require simulations combined with fluorescence anisotropy to disentangle the anisotropies of the local excitation with this PCT state. In analyzing the anisotropy data, it can be seen that the bis-DIPYR data compared to the m8B dimer is quite noisy. This ‘noise’ can be due to couple of effects. One is uncorrelated sources of noise or drifts in power and compression of the pump/probe source between the parallel and perpendicular spectra as they were taken sequentially. Alternatively, this ‘noise’ or rapid changes 143 in anisotropy has been shown by Matro and Cina 27 to be attributed to anisotropies of coupled chromophores whose transitions couple to vibrational modes or these are oscillations due to electronic coupling of the DIPYR chromophores. In order to distinguish noise due to sequential acquisition of the parallel and perpendicular transient signals from variations of anisotropic signals due to physical changes within the molecule being probed, a new anisotropy set up has been proposed. 7. Improvements to Anisotropy Setup One of the major advantages of anisotropy as a technique is that it is ratiometric, where experimental noise and errors can be canceled out. However, this is only true if the noise or drifts within the experiment are the same in both the parallel and perpendicular intensities. This is impossible to eliminate entirely when the parallel and perpendicular transient spectra are obtained sequentially. If, say there are drifts in power or compression between experiments, this can compromise the physical interpretation of the anisotropy. One way overcome this is to collect the parallel and perpendicular signals simultaneously. 144 Figure 5.9: New detection set up for pump-probe anisotropy experiments. The pump is polarized along the horizontal polarization while the probe is rotated to 45º with respect to the pump. The probe is collimated after the sample and sent through a Wallostan polarizer which separates the horizontal and vertical or parallel and perpendicular components, respectively, of the white light. The horizontal component of the probe is direct through a prism which disperses the white light onto a 256-pixel silicon diode array where the parallel signal can be detected. The vertical component of the probe is first sent through a 90 º periscope to flip the polarization of the probe to horizontal where it is directed to a prism and dispersed onto a 256- pixel silicon diode array for detection of the perpendicular transient signal. The prims used are an identical pair as are the silicon diode arrays. Here we implement a new detection scheme that will track correlated noise and drifts between the parallel and perpendicular signals which should increase the overall signal to noise of the anisotropy. In the sequential collection of the parallel and perpendicular signals, the probe pulse was set to a horizontal polarization and held there while the pump was rotated using an achromatic waveplate between vertical and horizontal polarizations. In the dual detection set up, the pump is held at a horizontal polarization as that is the polarization output from the noncolinear optical parametric amplifier (NOPA), and the probe will be held at 45 º with respect to the pump. 145 The pump is directed through a Glan-Taylor polarizer to purify the horizontal polarization prior to the sample. The probe is set to 45 º by sending the 800 nm through a λ/2 waveplate prior to generating white light in a CaF2 disc. Thus, at the sample the pump and probe are at 45 º with respect to one another giving an equivalent projection of intensity in both the vertical (perpendicular) and horizontal (parallel) polarizations. After the sample, the probe is collimated and directed to a Wollaston polarizer. The polarizer is rotated such that the horizontal and vertical components, or parallel and perpendicular signals, respectively, are separated at equal intensities. Spatially the parallel and perpendicular components are separated along a horizontal plane. The parallel, or horizontal, component of the white light is directed to a prism which disperses the white light onto a 256-pixel silicon diode array. The perpendicular, or vertical, component is directed through a 90 degree periscope first in order to change the polarization to horizontal in the lab frame. This is then sent through a prism, where the prism used to disperse the perpendicular signal is a matched pair with the prism dispersing the parallel signal. The perpendicular signal is dispersed onto a 256-pixel silicon diode array, identical to diode array for parallel signal detection. In setting up this new apparatus, there are a couple of things to keep in mind. The distance between the detectors and the prisms should be the same for both probe polarizations and the angle of the prisms with respect to the incident beams and diode arrays should be equivalent as this will give an equivalent dispersion onto both diode arrays. It is also important that the white light is collimated prior to the Wollaston polarizer such that the beam size of the parallel and perpendicular components are identical as this may affect intensity of the detected parallel and perpendicular probe. The last thing, though potentially the most complicated, is the degree and direction of spectral chirp induced by the Wollaston polarizer. If the beams are horizontally dispersed through 146 the polarizer, this introduces chirping of the white light along the horizontal axis. Thus, for the output that is polarized horizontal and sent through the prism, this chirping will have negligible effect on the dispersed signal. However, for the vertically polarized output, because the chirp and polarization are along orthogonal axes, after the periscope the polarization will be horizontal, but the chirp will flip to vertical. This will primarily manifest itself in the bluer wavelengths, where there is a visible curve up in the blue region of the dispersed white light, while the redder wavelengths are more-or-less in plane. The chirp discussed previously will have an effect on the ability for the differently polarized signals to be identically dispersed on their respective diode arrays. It was found that wavelengths red of 450 nm could be spectrally matched on the identical photodiode arrays, but wavelengths to the blue of 450 nm could not be spectrally matched. There are a couple of ways to mitigate this, and one is to use identical cylindrical lenses where the dispersed signal can be compressed to a common plane, thus reducing the vertical chirp in the perpendicular signal. However, for signals red of this, so long as the transient signals match, the anisotropy can be obtained without regards for the bluer wavelengths in the transient spectra. 147 Figure 5.10 Preliminary anisotropy of γ-DIPYR in acetonitrile (left) and γ-DIPYR molecule (right). The anisotropy plotted is for the SE band at 540 nm. It can be seen that the anisotropy follows the predicted behavior as well as the experimental behavior observed from the sequential collection of the parallel and perpendicular signals. That is the initial anisotropy is 0.4 and the final is 0. We have shown for the γ-DIPYR system in acetonitrile, where there was no chirp correction, the expected anisotropy was obtained. Like what was observed for the sequential acquisition of the parallel and perpendicular transient spectra, the anisotropy for the SE band started at 0.4 and decayed to 0 (Figure 5.10). These results are preliminary and with the added suggested changes to help match the dispersed light onto the identical arrays, polarization dependent studies could be used to obtain anisotropy data that has high signal to noise and can account for power, compression and experimental noise between the parallel and perpendicular transient signals. 8. Summary and Future Work Pump-probe anisotropy is a powerful technique which has been used to understand the orientation of different transition moments for different excited state processes. Here we used pump-probe anisotropy with the goal in mind of both determining whether or not there was a preferential direction for charge transfer, especially in the BODIPY dimer where a PCT 148 intermediate was observed. Initially, the zinc dipyrrinato dimer was analyzed and it was found that, most likely, electron and hole transfer were equally probable, as would be expected. This is with the assumption that the anisotropic decay is due to formation of the charge separated state, where the dipyrrinato chromophores behave independently. In the case of the BODIPY dimer, however, the anisotropy decayed to about 0.1 over timescales of PCT formation. Following this the anion band seemed to be negative on timescales prior to reorientation where SBCT should occur. This may imply that the BODIPY dimers preferentially undergo electron transfer as appose to hole transfer. We also hoped to understand the differences in orientation or character of the PCT state which is observed in the BODIPY and DIPYR dimers. As stated previously, the anisotropy of the BODIPY dimer seemed to immediately decay to 0.1. For anisotropy in dimers this implies that there is projection onto a common plane of a given transition moment, or rather there is equal projection of the transition moment with the pump induced transition as well as that of the identical perpendicular transition, giving a 45 º angle. This could be due to, like what is observed in BA, electron density along the bridge. The DIPYR dimers gave an anisotropy of about 0.1 in the GSB, but an anisotropy of 0.2 in the SE and ESA bands. Unlike the BODIPY systems, DIPYR does not indicate that the PCT state is along a transition which has an equal interaction with equivalent perpendicular transitions, but rather behaves as a state which is due to a superposition of both a parallel and perpendicular transition. This may mean that in the case of m8B, the PCT state is along the bridge, in likeness to BA, while in bis-DIPYR the anisotropy may indicate more of a delocalization of the excitation where it is observed on both the parallel and perpendicular chromophores. 149 However, the above anisotropy data were obtained from sequential acquisition of the parallel and perpendicular TA data. Thus, any drifts in power and compression or uncorrelated noise between the parallel and perpendicular experiments will directly affect the anisotropy data and may compromise the physical interpretation of the data. In order to mitigate this a new experimental set up for simultaneous collection of the parallel and perpendicular TA signals has been proposed. As a proof of concept, it was shown that the dual detection works, as the expected anisotropy behavior of the SE in γ-DIPYR monomer, that is an initial anisotropy of 0.4 and decay to 0 at long times, was achieved. This will help in either reproducing the anisotropy results from above or may help in further understanding of the systems as all noise will be correlated resulting in an overall higher signal-to-noise of the anisotropy. This can be used in the future to study the orientations of the transition moments in excited state charge transfer, where there will be even higher certainty in the anisotropic behavior of these systems. This will provide a new technique which can allow sensitivity to the character and behavior of excited state charge transfer in engineering efficient systems. 150 Chapter 5 Bibliography (1) Kumpulainen, T.; Lang, B.; Rosspeintner, A.; Vauthey, E. Ultrafast Elementary Photochemical Processes of Organic Molecules in Liquid Solution. Chem Rev 2016, 117 (16), 10826–10939. https://doi.org/10.1021/acs.chemrev.6b00491. (2) Whited, M. T.; Patel, N. M.; Roberts, S. T.; Allen, K.; Djurovich, P. I.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Intramolecular Charge Transfer in the Excited State of Meso -Linked BODIPY Dyads. Chem Commun 2011, 48 (2), 284–286. https://doi.org/10.1039/c1cc12260f. (3) Trinh, C.; Kirlikovali, K.; Das, S.; Ener, M. E.; Gray, H. B.; Djurovich, P.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Charge Transfer of Visible Light Absorbing Systems: Zinc Dipyrrins. J Phys Chem C 2014, 118 (38), 21834–21845. https://doi.org/10.1021/jp506855t. (4) Golden, J. H.; Estergreen, L.; Porter, T.; Tadle, A. C.; R, D. S. M.; Facendola, J. W.; Kubiak, C. P.; Bradforth, S. E.; Thompson, M. E. Symmetry-Breaking Charge Transfer in Boron Dipyridylmethene (DIPYR) Dimers. Acs Appl Energy Mater 2018, 1 (3), 1083–1095. https://doi.org/10.1021/acsaem.7b00214. (5) Bartynski, A. N.; Gruber, M.; Das, S.; Rangan, S.; Mollinger, S.; Trinh, C.; Bradforth, S. E.; Vandewal, K.; Salleo, A.; Bartynski, R. A.; Bruetting, W.; Thompson, M. E. Symmetry- Breaking Charge Transfer in a Zinc Chlorodipyrrin Acceptor for High Open Circuit Voltage Organic Photovoltaics. J Am Chem Soc 2015, 137 (16), 5397–5405. https://doi.org/10.1021/jacs.5b00146. (6) Kellogg, M.; Akil, A.; Ravinson, D. S. M.; Estergreen, L.; Bradforth, S. E.; Thompson, M. E. Symmetry Breaking Charge Transfer as a Means to Study Electron Transfer with No Driving Force. Faraday Discuss 2019, 216 (0), 379–394. https://doi.org/10.1039/c8fd00201k. (7) Vauthey, E. Photoinduced Symmetry‐Breaking Charge Separation. Chemphyschem 2012, 13 (8), 2001–2011. https://doi.org/10.1002/cphc.201200106. (8) Brinkert, K. Energy Conversion in Natural and Artificial Photosynthesis. 2018. https://doi.org/10.1007/978-3-319-77980-5. (9) Cook, R. E.; Phelan, B. T.; Kamire, R. J.; Majewski, M. B.; Young, R. M.; Wasielewski, M. R. Excimer Formation and Symmetry-Breaking Charge Transfer in Cofacial Perylene Dimers. J Phys Chem 2017, 121 (8), 1607–1615. https://doi.org/10.1021/acs.jpca.6b12644. (10) Yu, Y.; Chien, S.-C.; Sun, J.; Hettiaratchy, E. C.; Myers, R. C.; Lin, L.-C.; Wu, Y. Excimer- Mediated Intermolecular Charge Transfer in Self-Assembled Donor–Acceptor Dyes on Metal Oxides. J Am Chem Soc 2019, 141 (22), 8727–8731. https://doi.org/10.1021/jacs.9b03729. 151 (11) Aster, A.; Licari, G.; Zinna, F.; Brun, E.; Kumpulainen, T.; Tajkhorshid, E.; Lacour, J.; Vauthey, E. Tuning Symmetry Breaking Charge Separation in Perylene Bichromophores by Conformational Control. Chem Sci 2019, 10 (45), 10629–10639. https://doi.org/10.1039/c9sc03913a. (12) Loudet, A.; Burgess, K. BODIPY Dyes and Their Derivatives: Syntheses and Spectroscopic Properties. Chem Rev 2007, 107 (11), 4891–4932. https://doi.org/10.1021/cr078381n. (13) Golden, J. H.; Facendola, J. W.; R, D. S. M.; Baez, C. Q.; Djurovich, P. I.; Thompson, M. E. Boron Dipyridylmethene (DIPYR) Dyes: Shedding Light on Pyridine-Based Chromophores. J Org Chem 2017, 82 (14), 7215–7222. https://doi.org/10.1021/acs.joc.7b00786. (14) Dupuy, N.; Bouaouli, S.; Mauri, F.; Sorella, S.; Casula, M. Vertical and Adiabatic Excitations in Anthracene from Quantum Monte Carlo: Constrained Energy Minimization for Structural and Electronic Excited-State Properties in the JAGP Ansatz. J Chem Phys 2015, 142 (21), 214109. https://doi.org/10.1063/1.4922048. (15) Lee, C.; Choi, C. H.; Joo, T. A Solvent–Solute Cooperative Mechanism for Symmetry- Breaking Charge Transfer. Phys Chem Chem Phys 2019, 22 (3), 1115–1121. https://doi.org/10.1039/c9cp05090f. (16) Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W. Structural Changes Accompanying Intramolecular Electron Transfer: Focus on Twisted Intramolecular Charge-Transfer States and Structures. Chem Rev 2003, 103 (10), 3899–4032. https://doi.org/10.1021/cr940745l. (17) Dereka, B.; Svechkarev, D.; Rosspeintner, A.; Tromayer, M.; Liska, R.; Mohs, A. M.; Vauthey, E. Direct Observation of a Photochemical Alkyne–Allene Reaction and of a Twisted and Rehybridized Intramolecular Charge-Transfer State in a Donor–Acceptor Dyad. J Am Chem Soc 2017, 139 (46), 16885–16893. https://doi.org/10.1021/jacs.7b09591. (18) Piet, J. J.; Schuddeboom, W.; Wegewijs, B. R.; Grozema, F. C.; Warman, J. M. Symmetry Breaking in the Relaxed S 1 Excited State of Bianthryl Derivatives in Weakly Polar Solvents. J Am Chem Soc 2001, 123 (22), 5337–5347. https://doi.org/10.1021/ja004341o. (19) Takaya, T.; Hamaguchi, H.; Iwata, K. Femtosecond Time-Resolved Absorption Anisotropy Spectroscopy on 9,9′-Bianthryl: Detection of Partial Intramolecular Charge Transfer in Polar and Nonpolar Solvents. J Chem Phys 2009, 130 (1), 014501. https://doi.org/10.1063/1.3043368. (20) Markovic, V.; Villamaina, D.; Barabanov, I.; Lawson Daku, L. M.; Vauthey, E. Photoinduced Symmetry‐Breaking Charge Separation: The Direction of the Charge Transfer. Angew Chem-ger Edit 2011, 123 (33), 7738–7740. https://doi.org/10.1002/ange.201102601. (21) Jonas, D. M.; Lang, M. J.; Nagasawa, Y.; Joo, T.; Fleming, G. R. Pump−Probe Polarization Anisotropy Study of Femtosecond Energy Transfer within the Photosynthetic Reaction Center of Rhodobacter Sphaeroides R26. J Phys Chem 1996, 100 (30), 12660–12673. https://doi.org/10.1021/jp960708t. 152 (22) Yeh, A. T.; Shank, C. V.; McCusker, J. K. Ultrafast Electron Localization Dynamics Following Photo-Induced Charge Transfer. Science 2000, 289 (5481), 935–938. https://doi.org/10.1126/science.289.5481.935. (23) Zhang, W.; Lan, Z.; Sun, Z.; Gaffney, K. J. Resolving Photo-Induced Twisted Intramolecular Charge Transfer with Vibrational Anisotropy and TDDFT. J Phys Chem B 2012, 116 (37), 11527–11536. https://doi.org/10.1021/jp306455m. (24) Kelley, A. M. Condensed-Phase Molecular Spectroscopy and Photophysics, 1st ed.; Wiley: Hoboken, N.J., 2012. https://doi.org/10.1002/9781118493052.ch9. (25) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd Edition.; Lacowicz, J. R., Ed.; 2006. https://doi.org/10.1007/978-0-387-46312-4. (26) Smith, E. R.; Jonas, D. M. Alignment, Vibronic Level Splitting, and Coherent Coupling Effects on the Pump−Probe Polarization Anisotropy. J Phys Chem 2011, 115 (16), 4101–4113. https://doi.org/10.1021/jp201928s. (27) Matro, A.; Cina, J. A. Theoretical Study of Time-Resolved Fluorescence Anisotropy from Coupled Chromophore Pairs. J Phys Chem 1995, 99 (9), 2568–2582. https://doi.org/10.1021/j100009a015. (28) Hattori, S.; Ohkubo, K.; Urano, Y.; Sunahara, H.; Nagano, T.; Wada, Y.; Tkachenko, N. V.; Lemmetyinen, H.; Fukuzumi, S. Charge Separation in a Nonfluorescent Donor−Acceptor Dyad Derived from Boron Dipyrromethene Dye, Leading to Photocurrent Generation. J Phys Chem B 2005, 109 (32), 15368–15375. https://doi.org/10.1021/jp050952x. (29) Wynne, K.; Hochstrasser, R. M. Coherence Effects in the Anisotropy of Optical Experiments. Chem Phys 1993, 171 (1–2), 179–188. https://doi.org/10.1016/0301- 0104(93)85142-u. (30) Wynne, K.; Hochstrasser, R. M. Anisotropy as an Ultrafast Probe of Electronic Coherence in Degenerate Systems Exhibiting Raman Scattering, Fluorescence, Transient Absorption and Chemical Reactions. J Raman Spectrosc 1995, 26 (7), 561–569. https://doi.org/10.1002/jrs.1250260711. (31) Qian, W.; Jonas, D. M. Role of Cyclic Sets of Transition Dipoles in the Pump–Probe Polarization Anisotropy: Application to Square Symmetric Molecules and Perpendicular Chromophore Pairs. J Chem Phys 2003, 119 (3), 1611–1622. https://doi.org/10.1063/1.1581854. (32) Galli, C.; Wynne, K.; LeCours, S. M.; Therien, M. J.; Hochstrasser, R. M. Direct Measurement of Electronic Dephasing Using Anisotropy. Chem Phys Lett 1993, 206 (5–6), 493– 499. https://doi.org/10.1016/0009-2614(93)80174-n.

Abstract (if available)

Abstract Currently the solar industry is dominated by solar cells made up of crystalline silicon wafers. Though silicon has shown power conversion efficiencies as high as 26%, it fails to maximally utilize the already diffuse photons provided by the sun. Organic photovoltaics (OPV) offer an alternative as they contain organic chromophores which can readily absorb photons and be deposited as thin films onto flexible substrates, where they may be used in an array of novel geometries. OPVs, however, tend to fall short in terms of their overall power conversion efficiencies. Maximizing the efficiency of an OPV comes from minimizing loss processes associated with excitation to charge generation. One of the challenges in doing this is producing light absorbing molecules whose excited state lifetime is long enough to diffuse to a donor-acceptor interface while minimizing energy loss from charge separation at the interface. ❧ In this thesis ultrafast spectroscopy will be used to study excited state phenomena within systems of coupled chromophores. The excited state phenomena studied in this thesis are singlet fission and symmetry breaking charge transfer (SBCT) which are two processes that can both elongate the excited state lifetime while potentially minimizing energy loss in the eventual generation of charge carriers for power conversion. These processes are studied by looking at the efficiencies of singlet fission and SBCT as a function of either morphology or molecular structure of the coupled chromophores.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Singlet fission in disordered acene films: from photophysics to organic photovoltaics

PDF

Singlet fission in covalently-linked tetracenes

PDF

Organic solar cells: molecular electronic processes and device development

PDF

Excited state dynamics of organic photovoltaics probed using ultrafast spectroscopy

PDF

Improving the field of organic photovoltaics through the development of new active layer materials with unique photophysical properties

PDF

Exciton dynamics in photovoltaic materials

PDF

Organic photovoltaics: from materials development to device application

PDF

Molecular design for organic photovoltaics: tuning state energies and charge transfer processes in heteroaromatic chromophores

PDF

Two-dimensional metal dithiolene metal-organic frameworks as conductive materials for solar energy conversion

PDF

Capturing the sun: leveraging excited state dynamics

PDF

Multilayer grown ultrathin nanostructured GaAs solar cells towards high-efficiency, cost-competitive III-V photovoltaics

PDF

Plasmonic enhancement of catalysis and solar energy conversion

PDF

Electrochemical pathways for sustainable energy storage and energy conversion

Asset Metadata

Creator Estergreen, Laura Kathleen (author)

Core Title First steps of solar energy conversion; primary charge generation processes in condensed phase coupled chromophores

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Chemistry

Publication Date 12/01/2020

Defense Date 09/21/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,singlet fission,solar energy,spectroscopy,symmetry breaking charge transfer

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bradforth, Stephen (committee chair), Ravichandran, Jayakanth (committee member), Thompson, Mark (committee member)

Creator Email estergre@usc.edu,lottestergr@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-401626

Unique identifier UC11666686

Identifier etd-Estergreen-9170.pdf (filename),usctheses-c89-401626 (legacy record id)

Legacy Identifier etd-Estergreen-9170.pdf

Dmrecord 401626

Document Type Dissertation

Rights Estergreen, Laura Kathleen

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA