University of Southern California Dissertations and Theses

Multiphoton ionization of tTris(2-phenylpyridine)Iridium

(USC Thesis Other)

Multiphoton ionization of tTris(2-phenylpyridine)Iridium

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i MULTIPHOTON INVESTIGATION OF TRIS(2-PHENYLPYRIDINE)IRIDIUM by Christopher M. Nemirow A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (CHEMISTRY) August 2011 Copyright 2011 Christopher M. Nemirow ii Epigraph I do not believe in a personal God and I have never denied this but have expressed it clearly. If something is in me which can be called religious then it is the unbounded admiration for the structure of the world so far as our science can reveal it. Albert Einstein March 24, 1954 iii Acknowledgements It is an understatement to say that I needed a great deal of guidance throughout my life. I have been incredibly lucky to have always had people to support, encourage, and push me when I needed it the most. The following is a long—but by no means comprehensive—list of mentors, friends, and family that have helped shape my life. First, I would like to thank my advisor Curt Wittig. Not only is Curt a brilliant scientist, he is a wonderful and kind-hearted person. As I am sure anyone who has worked for him will attest it is difficult to imagine a better mentor. Curt never doubted my ability, which helped me learn to believe in myself. I would like to express my appreciation and admiration to the professors of the physical chemistry department, which is friendly and inviting because of these wonderful people. In particular, I am fortunate to have had the opportunity to work with and learn from Professors Anna Krylov and Hanna Reisler. In addition, Professor Steve Bradforth provided me with guidance and encouragement when my self-confidence was severely waning. Steve kept my on track; I am not sure I would be here without his help. The research presented in this dissertation represents the accumulation of much hard work and many late nights; I was not alone in this endeavor. Dr. Zhou Lu and Jordan Fine were instrumental in this regard. After working with both of you, I walk away with a degree, a new library of knowledge, and two new friends. I cannot thank you enough. iv I also learned a great deal working on other projects early in my graduate career. Thank you, Dr. Elena Polykova and Dr. Daniil Stolyarov, for your patience and insight. I also owe much gratitude and many thanks to Anton Zadorozhnyy for making work so much fun. We laughed everyday; I am lucky to have you as a friend. I want to acknowledge the rest of my colleagues, past and present: Dr. Lee-Ann Smith-Freeman, Oscar Rebolledo-Mayoral, Bill Schroeder, Dr. George Kumi, Jessica Quinn, Dr. Sergey Malyk, Jaimie Stomberg, and Stephanie McKean. Each of you made my time in Seaver Science Center unforgettable. Additionally, I would like to acknowledge a number of classmates that made even homework and exams pleasant memories. To Arun Sharma, Lucas Koziol, Laura Edwards, and Diana Warren, thank you for helping me grow and making graduate school fun. Before graduate school, a number of professors and teachers saw potential in me that I did not know existed. No matter what kind of trouble I got into, they never gave up on me, and I would like to thank them for that. At Worcester State College, I would like to recognize Professors Margaret Kerr, Anne Falke, John Goodchild, Alan Cooper, and Arthur Ferguson. At Nashoba Regional High School, despite my refusal to listen to anyone, Loretta Williams, Daniel D’Amore, and Ed Boyce never stopped trying. My friends Matt Taylor and Jeff Dee showed similar unwavering confidence in me. When I was younger, a number of friends and their families showed me kindness and sincere hospitality. Particularly, I would like to thank the families of Wes Schumacher, Adam Frantz, Eric Sliwa, and Heather Meehan (the Hunts). I would also v like to recognize the incredible friendships of Rick Bodine, Jake Eickhorst, Andy Himmer, and Shana Rezac. My life would not be the same without the extraordinary friendships of Matt and Larryn Peterson and Chris Rivera. The world would be a better place if everyone had friends like you. I am looking forward to our future adventures. My family deserves a lot of credit for making me the person I am today. My parents have sacrificed so much without ever asking for anything in return. Your unconditional love and support paved my path. Thank you for believing in me. When I was sixteen, my sister, Tonia, and brother-in-law, Kirk, showed me that happiness existed; my first trip to South Carolina changed my life. Thank you to the rest of my siblings Derek, Taryn, Aaron, and David, and my brother-in-law Jay, for making every other family boring compared to ours. I would like to recognize my nephews Ronan, Ryder, Myles, and Shyloh and my beautiful niece Cadence; just thinking of all of you makes me smile. I am grateful to Joy’s parents and brother for welcoming me into their family. Finally, thank you Joy for your love and support; we make a badass team. I am looking forward to our lives together. I love you. vi Table of Contents Epigraph ii Acknowledgements iii List of Tables viii List of Figures ix Abstract xii Chapter 1: Introduction 1 1.1 Background and Motivation 2 1.1.1 Organic Light Emitting Diodes 2 1.1.2 Tris(2-phenylpyridine)iridium 6 1.2 References 9 Chapter 2: Experimental Techniques 11 2.1 Helium Droplet Isolation Spectroscopy 11 2.1.1 Helium Droplet Production and Characteristics 13 2.1.2 Capturing Foreign Particles 14 2.2 Multistep Photoionization 16 2.2.1 Introduction 17 2.2.2 Obtaining Spectroscopic Information 23 2.2.2.i Stepwise Ionization Spectroscopy 23 2.2.2.ii Multiphoton Ionization Spectroscopy 28 2.2.3 Fragmentation 29 2.2.4 Summary 32 2.3 Time-of-Flight Mass Spectroscopy 32 2.3.1 Overview 33 2.3.2 Resolution 35 2.4 References 40 Chapter 3: Tris(2-phenylpyridine)iridium Helium Droplet Study 42 3.1 Introduction 42 3.2 Experimental Apparatus 45 3.2.1 Vacuum Chamber 46 3.2.2 Laser System 49 3.3 Ir(ppy) 3 Internal Energy 49 3.4 Results 52 3.5 Discussion 58 3.5.1 QMS Helium Droplet Progression 58 3.5.2 Capture Cross Section for Ir(ppy) 3 66 vii 3.6 Future Directions 69 3.7 Conclusion 71 3.8 References 72 Chapter 4: Multiphoton Ionization of Tris(2-phenylpyridine)iridium 74 4.1 Introduction 74 4.2 Experimental Apparatus 75 4.2.1 Laser and Optical Setup 76 4.2.2 Time-of-Flight Mass Spectrometer 77 4.3 Computational Study of Ir(ppy) 3 80 4.3.1 Structure and Molecular Orbitals 80 4.3.2 Excited States 84 4.3.3 Ionization Energies 87 4.4 Results 88 4.5 Discussion 97 4.5.1 Ionization Mechanism 97 4.5.2 Ionization Energy 102 4.5.3 Ir(ppy) 3 + Action Spectrum 104 4.5.4 Ir + Signal 105 4.6 Future Directions 109 4.7 Conclusion 110 4.8 References 112 Bibliography 115 Appendices 121 Appendix A: Ion Yield Derivation 121 Appendix B: Ir(ppy) 3 -Helium Droplet Collision Probability 124 Appendix C: Ir + Action Spectrum 131 Appendices References 135 viii List of Tables Table 3.1 Helium droplet beam parameters. 46 Table 3.2 Deflection angle α and scattering distance d as a function of collision parameters. 63 Table 4.1 Rhodamine 590 and 610 dye mixtures. 77 Table 4.2 TOF spectrometer dimensions and parameters. 78 Table 4.3 Major Ir(ppy) 3 isotopes. 79 Table 4.4 Bond lengths (Å) for the ground state (S 0 ), lowest triplet (T 1 ), and lowest cation states for Ir(ppy) 3 . 81 Table 4.5 Selected excited states (S 0 geometry). 84 Table B.1 Helium droplet expansion conditions and constants. 124 Table B.2 Constants used for Eq. (B.2) – (B.5). 130 Table C.1 Ir + Action Spectrum Assignments 131 ix List of Figures Figure 1.1 Schematic of the structure of a simple OLED. 3 Figure 1.2 Diagram of the low-lying electronic states demonstrating the effects of triplet harvesting. 5 Figure 1.3 Picture of fac-tris(2-phenylpyridine)iridium. 7 Figure 2.1 Average number of helium atoms per droplet for different expansion conditions. The right axis shows the corresponding average droplet diameter (Å). 14 Figure 2.2 Two common multistep ionization schemes: (a) two-color stepwise ionization and (b) one-color multiphoton ionization. 18 Figure 2.3 Two level system: states |i〉 and | j〉 have corresponding ener- gies E i and E j . 20 Figure 2.4 Population loss of the intermediate state S n can compete with photoionization, leading to a depletion of the ion signal. Com- petitive pathways include (a) photon emission; (b) disso- ciation; (c) internal conversion (IC) to the single manifold; and (d) intersystem crossing (ISC) to the triplet manifold. 25 Figure 2.5 Illustration of the effective ionization energy. 27 Figure 2.6 Depiction of two fragmentation schemes for multistep ioni- zation. 30 Figure 2.7 Diagram of the basic time-of-flight (TOF) setup. 33 Figure 2.8 Graphical representation of the TOF mass resolution M achieved for the experiment described in Chapter 4. 37 Figure 2.9 Diagram showing the effects of the initial energy distribution on the TOF mass resolution. 38 Figure 3.1 Schematic representation of the vacuum chamber. Approx- imate dimensions are shown in mm. 45 Figure 3.2 Quadrupole mass spectrum of pure helium droplets with a mean droplet size of 8000 helium atoms. 53 x Figure 3.3 Details of the spectrum presented in Fig. 3.2 (solid red line), along with the analogous spectrum recorded with Ir(ppy) 3 present in the pick-up cell at 500 K (dashed black line). 54 Figure 3.4 TOF spectra of Ir(ppy) 3 recorded with a photon energy of 35,047 cm -1 and a fluence of 2.3 J / cm 2 . 56 Figure 3.5 Action spectra of Ir(ppy) 3 collected with the helium droplet beam on (a) and off (b) with a fluence of 2.3 J / cm 2 and a mean droplet size of N 〈 〉 = 8000. 57 Figure 3.6 Calculated collision probability of k Ir(ppy) 3 molecules as a function of pick-up chamber pressure reading. 61 Figure 3.7 Schematic showing a collision between a helium droplet and Ir(ppy) 3 in the pick-up cell at an incident angle θ, resulting in a deflection angle α that depends on the initial velocities. 62 Figure 3.8 The average impact parameter, b 〈 〉 = 2/3R, is shown for a col- lision between a helium droplet of mean size N 〈 〉 = 8000 atoms and Ir(ppy) 3 . 69 Figure 4.1 Time-of-flight mass spectrum showing the Ir(ppy) 3 + ion peak (solid black line), recorded at 35,673 cm -1 and 415 µJ / cm 2 . The best fit trace results from the addition of four Gaussian curves representing the predominant isotopes of Ir(ppy) 3 . 79 Figure 4.2 Picture of fac-Ir(ppy) 3 . Hydrogen atoms are omitted for clarity. 80 Figure 4.3 Ir(ppy) 3 molecular orbitals: (a) HOMO and (b) LUMO at the S 0 geometry; (c) HOMO and (d) LUMO at the T 1 geometry. 82 Figure 4.4 Ir(ppy) 3 frontier orbitals at the S 0 geometry: (a) six highest occupied MOs and (b) three lowest virtual MOs. 83 Figure 4.5 Calculated oscillator strengths (x 3) of 130 singlet states vs. photon energy (blue lines). Absorption spectrum (black line) is generated by applying a uniform Gaussian width to each state. 85 Figure 4.6 Comparison of the calculated absorption spectrum (black line) taken from Fig. 4.5 and measured spectrum (red line) recorded in dichloromethane. Inset: UV-Vis spectrum of 2-phenyl- pyridine. 86 xi Figure 4.7 Several TDDFT / BNL excited states and ionization energies calculated at the S 0 and T 1 geometries. 87 Figure 4.8 Absorption spectrum of Ir(ppy) 3 in dimethylformamide at 414 K. Inset: increasing temperature has no significant effect on the absorption spectrum in the energy range of interest. 88 Figure 4.9 TOF spectrum of Ir(ppy) 3 collected at 35,774cm -1 with a fluence of 1.5 J / cm 2 . 89 Figure 4.10 Ir + signal intensity vs. photon energy recorded with three dif- ferent fluences. 90 Figure 4.11 Expanded section of the plot shown in the top panel of Fig. 4.10 (black line), along with a stick spectrum of the atomic iridium transitions (red lines). 91 Figure 4.12 TOF spectra of Ir(ppy) 3 collected at hv = 35,357 cm -1 and var- ious fluence. 92 Figure 4.13 Logarithmic plot of the Ir(ppy) 3 + ion signal vs. laser fluence collected using a focused laser, in the range 2 – 15 mJ / cm 2 93 Figure 4.14 Logarithmic plots of the Ir(ppy) 3 + ion signal intensity vs. laser fluence at different photon energies. 94 Figure 4.15 Ir(ppy) 3 + signal intensity vs. hv. 95 Figure 4.16 Selected logarithmic plots of the Ir(ppy) 3 + signal intensity vs. laser fluence at 28,169 cm -1 (355 nm). 96 Figure 4.17 Schematic of the molecular dynamics of Ir(ppy) 3 . 98 Figure 4.18 Schematic of the ionization mechanism. 101 Figure 4.19 Diagram of the ionization and dissociation channels. 107 Figure A.1 Schematic of a three level system. 121 Figure B.1 Relative ion signal of ArHe 9 + , Ar 2 + , and Ar 3 + as a function of pick-up chamber pressure. 125 Figure C.1 Details of the Ir + action spectrum shown in the top panel of Fig. 4.10, recorded with an energy fluence of 130 mJ / cm 2 . 131 xii Abstract Tris(2-phenylpyridine)iridium (Ir(ppy) 3 ) is a popular green phosphor used in or- ganic light emitting diodes (OLEDs). In these applications, strong spin-orbit coupling in- duced by the iridium atom is responsible for this molecule’s high electroluminescent quantum efficiency. A comprehensive understanding of the electronic properties of Ir(ppy) 3 is desirable in this setting. Such knowledge may lead to the smart design of OLED phosphorescent dopants. To achieve this goal, helium droplet isolation and multi- photon ionization (MPI) spectroscopies were applied. The helium droplet study led to im- portant findings regarding the helium droplet capture process: helium droplets containing, on average, 8000 helium atoms are unable to efficiently capture vibrationally excited Ir(ppy) 3 . Gas-phase MPI revealed competing ionization and dissociation channels. Never- theless, the parent Ir(ppy) 3 + ion was isolated and the ionization energy was determined to be < 6.55 eV. This work is complimented by time-dependent density functional theory calculations. 1 Chapter 1 Introduction The primary goal of the work described in this dissertation is an improved under- standing of the electronic properties of tris(2-phenylpyridine)iridium (Ir(ppy) 3 ), a central component in organic light emitting diodes (OLEDs). As is frequently the case with research, the central theme strayed during the course of this work. Nevertheless, the range of results gathered provides interesting insight into the molecular properties as well as fundamental aspects of helium droplet isolation spectroscopy. A variety of spectroscopic tools were employed to probe the excited states of Ir(ppy) 3 . Helium droplet isolation spectroscopy was used with hopes of acquiring high resolution spectral data. While this goal has yet to be realized, strides have been made to this end. The more immediate impact of this work lies within the helium droplet tech- nique itself. Using Ir(ppy) 3 as a projectile provided insight to the helium droplet pick-up process. Effusive Ir(ppy) 3 was examined using multiphoton ionization. This study was complemented by density functional theory (DFT) calculations. This work revealed inter- esting dynamics. After absorption, two competing decay channels are available: ion- ization and dissociation. Isolation of the parent ion signal enabled the ionization scheme to be characterized and molecular information to be obtained. This dissertation proceeds as follows. The remainder of this chapter presents background material on OLEDs and Ir(ppy) 3 . The major experimental techniques 2 employed during this work—namely, helium droplet isolation, multiphoton ionization, and time-of-flight mass spectroscopy—are discussed in detail in Chapter 2. Chapter 3 presents the results of the helium droplet investigation. The gas-phase study is addressed in Chapter 4, which includes a summary of the DFT calculations. 1.1 Background and Motivation The interest in Ir(ppy) 3 stems from its application as a phosphor molecule in OLEDs. 1,2 Therefore, the following discussion begins with an introduction to OLEDs, which includes an examination of the technological significance of organometallic mol- ecules, such as Ir(ppy) 3 . Specific properties of Ir(ppy) 3 as well as a qualitative overview of some previous results are highlighted in the subsequent section. 1.1.1 Organic Light Emitting Diodes OLEDs have been the topic of intense research during the past decade, driven by expectations of this technology replacing existing light displays. 1,2 This section presents an overview of OLED devices and highlights some of their important technological aspects. Particular attention is paid to the phosphorescent emitter molecules that provide large electroluminescent quantum efficiencies. A detailed description of the working principle and device architecture is beyond the scope of this dissertation and can be found elsewhere. 1,2 3 Figure 1.1 Schematic of the structure of a simple OLED. 1 Electrons and holes are injected into the organic material by a voltage applied to the electrodes. Under the driving force of the electric field, the charges mi- grate toward the recombination layer that hosts the emitter molecules. Recombination leads to excitons, which eventually localize to form electronically excited emitter molecules. The excited molecules relax via photon emission (shown as squiggly arrows). A typical OLED device consists of a thin layer (on the order of 100’s of nano- meters 1 ) of electroluminescent organic material sandwiched between two electrodes. 1,2 A simplified structure is shown in Fig. 1.1. The working principle is analogous to a semi- conductor diode. When a voltage is applied to the electrodes, electrons and holes are in- jected into the organic layers. In particular, electrons are injected from the cathode into the lowest unoccupied molecular orbital (LUMO) of the electron-transporting layer, while the anode injects holes by removing electrons from the highest occupied molecular + - Cathode Anode - - - - + + + + * * * - + - + - + Electron transporting layer Hole transporting layer Recombination layer Light output - + * - + Electron Hole Exciton Excited molecule 4 orbital (HOMO) of the hole-transporting layer. 1,2 These charge carriers migrate toward each other, forming excitons in the charge-recombination layer that is doped with emitter molecules. 1,2,3 In a well designed device, excitons lead to population of excited states of the emitter molecules, which subsequently emit photons. 1,2 Each charge carrier has a corresponding spin assignment. In the case of a hole, the spin is defined by the residual electron. Because the charge pairs are uncorrelated, they recombine to form both singlet and triplet excitons. More specifically, the spins of the charge pairs can be combined to give a total of four states: one singlet and three triplet states. 1,2,4 The three triplet states, normally referred to as substates or microstates, differ from one another mainly by their spin orientations (M s = 0, ±1). It is generally accepted that all four states are formed with equal probability. 1,2 As a consequence, three triplet states are created for every singlet state produced. After relaxation according to the re- spective paths, the lowest excited singlet and triplet states are populated. The efficiency of a particular OLED depends largely on the choice of emitter molecule. Consider an organic molecule whose electronic states are represented in Fig. 1.2a. In general, organic molecules exhibit very weak spin-orbit coupling (SOC). 1,2 As a consequence, intersystem crossing (ISC) from S 1 to T 1 is slow, and the S 1 state can ex- hibit radiative emission (fluorescence). On the other hand, weak SOC also corresponds to small emission rates from the T 1 substates (phosphorescence). Therefore, the T 1 sub- states—which represent 75% of charge recombination events—typically experience the unfavorable fate of radiationless deactivation. 1,2,4 Hence, the majority of excitons are wasted as the energy is transferred to heat. 5 Figure 1.2 Diagram of the low-lying electronic states demonstrating the effects of triplet harvesting. Generally, electron-hole recombination leads to a three-to-one mixture of triplet-to-singlet excited states. (a) Organic molecules with weak SOC only emit from the singlet state. 1,2,4 Triplet excitation is lost through radiationless deactivation in the form of heat. (b) An organometallic compound with strong SOC experi- ences fast ISC from S 1 to T 1 (triplet harvesting). Subsequent emission from the triplet substates leads to an improved luminescent quantum efficiency, relative to the organic molecule. 1 The T 1 state can be prompted to do useful work by employing a transition metal complex in which the metal atom induces strong SOC. In such a case (Fig. 1.2b), the ISC rate from S 1 to T 1 is fast, effectively inhibiting fluorescence from the lowest singlet state. 5,6 As a result, both singlet and triplet excitons populate the molecular T 1 substates; this is referred to as triplet harvesting. 1,2 Because strong SOC significantly increases the phosphorescent rate, emission efficiencies can approach 100%. 7,8 A number of organic- transition metal complexes have been used as phosphor dopants in OLEDs to overcome the efficiency limit imposed by the formation of triplet excitons. 7,9,10 S 0 T 1 S 1 slow ISC S 0 T 1 S 1 radiationless deactivation fluorescence fast ISC phosphorescence (a) (b) Organic molecule Organometallic molecule 6 The demonstration of highly efficient OLEDs based on these organometallic com- pounds sparked great interest in the electronic properties of these molecules. A compre- hensive characterization of the electronic states and dynamics enables a better under- standing of the mechanisms that govern the photochemistry. Moreover, such information can potentially lead to smart-design of compounds with specific properties. 1.1.2 Tris(2-phenylpyridine)iridium The most famous organometallic doped emitter is Ir(ppy) 3 . This molecule, which has been central to OLED evolution, 1,2 is shown in Fig. 1.3. The strong SOC induced by the iridium atom is manifest as fast ISC 11 and efficient phosphorescence. 8 These prop- erties, combined with a favorable charge recombination cross section, 12 are responsible for the high electroluminescent quantum yields of OLEDs that employ Ir(ppy) 3 as a phos- phor impurity. 8–10 To a large extent, research on this molecule has focused on the emission mech- anism 12–15 although a handful of empirical studies have examined other aspects of Ir(ppy) 3 . 11,16,17 The broad emission spectrum is relatively invariant over a wide range of temperatures and is centered at 514 nm, 12 making Ir(ppy) 3 a green emitter. On the other hand, a time-resolved pump-probe experiment measuring the intramolecular dynamics determined the ISC lifetime to be around 100 fs, 11 while gas-phase electron energy loss spectroscopy (EELS) measurements established, among other things, an approximate ionization energy (IE) of 7.2 eV. 17 7 Figure 1.3 Picture of fac-tris(2-phenylpyridine)iridium (Ir(ppy) 3 ). The longest axis is approximately 14.75 Å (calculated using the density functional results presented in Sect. 5.3). A large amount of information concerning the electronic properties of Ir(ppy) 3 comes by way of theoretical investigations. 18–22 Density functional theory (DFT) calcu- lations place the IE around 6 eV; 19,23 time-dependent DFT is used to explore the excited electronic states. 18–20,23 In addition, the extent of mixing between the first-order singlet and triplet manifolds is addressed by a number of studies. 18,20,22 Such work is able to re- produce many experimental observations including the radiative lifetime 20 and T 1 sub- state energy separations. 18,20,22 In Sect. 4.3 some important theoretical results are dis- cussed in the context of the DFT calculations that were performed in conjunction with the work presented herein. Research from the application side demonstrates the promise of similar organo- metallic compounds in the phosphorescent emitter role. 5 Adding electron withdrawing and donating groups to the 2-phenylpyridine (ppy) ligand alters the energy gap and 8 therefore the emission wavelength. 5,21,24 Similar effects are obtained when entire ligands are substituted. 7,24 Clearly, Ir(ppy) 3 is interesting from the viewpoint of both physical chemistry and OLED applications. Certainly these pursuits are not mutually exclusive. Theoretical and experimental techniques provide an understanding of the photochemistry and excited state composition of the isolated molecule. Such work serves as a benchmark for other promising electroluminescent molecules. Moreover, with detailed knowledge of the elec- tronic properties it may be possible to design emitter molecules with specific qualities, i.e. emission wavelength and efficiency. 9 1.2 References 1. Yersin, H. Top. Curr. Chem. 2004, 241, 1. 2. Yersin, H.; Finkenzeller, W. J. Highly Efficient OLEDs with Phosphorescent Materials; Yersin, H., Ed.; Wiley-VCH Verlag GmbH & Co.: Wienheim, 2008; pp. 1 – 97. 3. Baldo, M. A.; Forrest, S. R. Phys. Rev. B 2000, 62, 10958. 4. Baldo, M. A.; O’Brien, D. F.; Thompson, M. E.; Forrest, S. R. Phys. Rev. B 1999, 60, 14422. 5. Hedley, G. J.; Ruseckas, A.; Samuel, I. D. W. J. Phys. Chem. A 2009, 113, 2. 6. Tsuboi, T. J. Lumin. 2006, 119–120, 288. 7. Adachi, C.; Baldo, M. A.; Thompson, M. E.; Forrest, S. R. J. Appl. Phys. 2001, 90, 5048. 8. Adachi, C.; Baldo, M. A.; Forrest, S. R. J.; Thompson, M. E. Appl. Phys. Lett. 2000, 77, 904. 9. Stampor, W.; MKżyk, J.; Kalinowski, J.; Cocchi, M.; Virgili, D.; Fattori, V.; DiMarco, P. Macromol. Symp. 2004, 212, 509. 10. Baldo, M. A.; Lamansky, S.; Burrows, P. E.; Thompson, M. E.; Forrest, S. R. Appl. Phys. Lett. 1999, 75, 4. 11. Tang, K.; Liu, K. L.; Chen, I. Chem. Phys. Lett. 2004, 386, 437. 12. Finkenzeller, W. J.; Yersin, H. Chem. Phys. Lett. 2003, 377, 299. 13. Vacha, M.; Koide, Y.; Kotani, M.; Sato, H. J. Lumin. 2004, 107, 51. 14. Tsuboi, T.; Alaroudi, N. Phys. Rev. B 2005, 72, 125109. 15. Hedley, G. J.; Ruseckas, A.; Samuel, I. D. W. Chem. Phys. Lett. 2008, 450, 292. 16. Stampor, W.; MKżyk, J.; Kalinowski, J. Chem. Phys. 2004, 300, 189. 17. Kukhta, A. V.; Kukhta, I. N.; Bagnich, S. A.; Kazakov, S. M.; Andreev, V. A.; Neyra, O. L.; Meza, E. Chem. Phys. Lett. 2007, 434, 11. 10 18. Nozaki, K. J. Chin. Chem. Soc. 2006, 53, 101. 19. Hay, P. J. J. Phys. Chem. A 2002, 106, 1634. 20. Jansson, E.; Minaev, B.; Schrader, S.; Ågren, H. Chem. Phys. 2007, 333, 157. 21. Liu, X.; Feng, J.; Ren, A.; Yang, L.; Yang, B.; Ma, Y. Opt. Mater. 2006, 29, 231. 22. Matsushita, T.; Asada, T.; Koseki, S. J. Phys. Chem. C 2007, 111, 6897. 23. Fine, J.; Nemirow, C.; Krylov, A.; Wittig, C. in preparation. 24. Li, X.; Wu. Z.; Si, Z.; Zhang, H.; Zhou, L.; Liu, X. Inorg. Chem. 2009, 48, 7740. 11 Chapter 2 Experimental Techniques Countless experimental methods and concepts were used to complete the work presented in this dissertation. Exhaustively discussing them is unrealistic and unnecessary as these techniques have all been discussed elsewhere. For example, information on mo- lecular beam methods—both effusive 1,2 and free-jet expansion 3 —as well as electron im- pact ionization, 4 laser spectroscopy, 5 and atomic collision theory 6 is readily available. Therefore, this chapter will concentrate and review only the techniques that are central to understanding the results presented in Chapters 3 and 4. The first section addresses super- fluid helium droplet isolation spectroscopy. The two subsequent sections review multi- step ionization spectroscopy and time-of-flight mass spectroscopy. 2.1 Helium Droplet Isolation Spectroscopy Helium is a remarkable element, exhibiting bizarre behavior at low temperatures. For example, because of weak van der Waals interactions and a large zero-point energy, helium is the only substance that remains liquid at normal pressures down to zero Kel- vin. 7 In addition, 4 He undergoes a unique phase transition at 2.18 K to a superfluid state ( 3 He experiences a similar transition at 3 mK). 8 This state is a fascinating subject in itself, displaying a range of unusual phenomena. These have been well documented 7 and in- clude a vanishingly small viscosity and very large heat conductivity. 8 12 As a result of these unique properties, superfluid helium is an ideal spectroscopic matrix. The ultra-cold environment can efficiently cool molecules to sub-Kelvin temper- atures, while the large quantum mechanical delocalization enables the surrounding heli- um atoms to gently accommodate the impurity, thereby eliminating the hot bands and site-specific broadening that often hinder spectra recorded in solid matrices. In addition, helium is transparent in the spectral range from the far IR to the vacuum UV. 8 Unfor- tunately, impurities move nearly unhindered in the liquid, leading to precipitation on the container walls, aggregation, and even ejection from the liquid. 7 Hence, isolating a for- eign species in superfluid helium is a major obstacle. 8 Helium droplet spectroscopy is an invaluable tool used to overcome these diffi- culties. By expanding cold helium through a small nozzle, a continuous beam of helium droplets ranging in size from several hundred to many thousands of atoms is formed. 8 4 He droplets are superfluid, residing at 0.38 K. 8 Gas-phase impurities are captured upon collision with the droplets, while the extent of doping is easily controlled by varying the experimental conditions. In this way, a large variety of species, ranging from atoms and atomic clusters to large biological molecules, can be isolated and probed in this gentle liquid. 8 Although droplets can be formed with 3 He, these droplets have a temperature of approximately 0.15 K, which is above the superfluid transition temperature for this iso- tope. Therefore, 4 He is used almost exclusively in the helium droplet isolation technique. The following discussion will be restricted to 4 He helium droplets formed with a contin- 13 uous nozzle source. Further information can be found in a multitude of review articles on helium droplets. 7–13 2.1.1 Helium Droplet Production and Characteristics Helium droplets are produced by cryogenically expanding helium through a noz- zle of several microns in diameter at high pressures. 8 There are two major expansion re- gimes, dictated by the source conditions. 7 In the most common operational mode— termed subcritical expansion—gaseous helium is expanded adiabatically, causing cooling and extensive condensation. The droplets cool further via evaporation. For droplets con- taining greater than 10 4 helium atoms, the binding energy per atom approaches the bulk value of 5 cm -1 (7.2 K); the final droplet temperature of 0.38 K is determined by the helium vapor pressure and droplet flight time. 14 Alternatively, in supercritical expansion, liquid helium passes through the nozzle. Although the final droplet temperature is the same in both regimes, in general, the beam properties depend on the type of expansion. Subcritical expansion was used exclusively in this work and will be the primary focus of the following discussion. In the subcritical regime, the droplet velocities lie in the range 200 – 400 m / s, de- pending on the nozzle temperature, 8 and possess a sharp velocity distribution ( / 0.01 0.03 v v ∆ ≈ − ). 8 Likewise, the average droplet size is determined by the source conditions. Droplets containing, on average, 10 3 – 10 7 helium atoms are readily produced by adjusting the nozzle temperature and stagnation pressure, as seen in Fig. 2.1. For a continuous, subcritical expansion, the size distribution is log-normal with a half-width 14 comparable to the most probable droplet size. 15 A spherical droplet containing N helium atoms has a radius of 8 1 3 2.22 R N = Å (2.1) The cross sectional area, which is also referred to as the geometrical cross section, is then ( ) 2 1 3 2.22 tot N σ π = (2.2) Figure 2.1 Average number of helium atoms per droplet for different expansion conditions. The right axis shows the corresponding average droplet diameter (Å). 16 2.1.2 Capturing Foreign Particles Droplets can capture a wide variety of atomic and molecular species upon col- lision. Assuming a sharp cutoff at the droplet edge, the collision cross section is equiv- Supercritical Expansion Subcritical Expansion 15 alent to the geometrical cross section given by Eq. (2.2), i.e. tot coll σ σ ≡ . The capture cross section (also called pick-up cross section) is of similar magnitude, often lying in the range 7 1/ 2 coll capt coll σ σ σ ≤ ≤ (2.3) The deviation between the two is attributed to a small percentage of particles being trans- mitted through the droplet periphery. 17 Collisions between the droplets and foreign spe- cies take place in a pick-up cell, located downstream from the droplet source, where— because of the relatively large pick-up cross sections—low partial pressures (10 -6 – 10 -7 torr) are sufficient for capture. 8 Consequently, a large variety of complexes can be em- bedded in helium droplets; the majority of these impurities reside near the droplet center. 8 When a molecule is captured by a droplet, the internal energy, collision energy, and impurity-helium binding energy are all imparted to the droplet; 8 the latter is relatively inconsequential for large molecules or highly energetic collisions. 17 This energy is dissi- pated by the surrounding helium on a picosecond time scale. 18 Subsequent helium evap- oration (one atom per 5 cm -1 ) brings the system back to 0.38 K. In this way, the droplet acts as a nano-cryostat, efficiently cooling molecules to sub-Kelvin temperatures. How- ever, this process limits the size of droplets that can be used for a particular experiment: the evaporative loss must be less than the pre-collision droplet size. In the limit of a constant capture cross section, i.e. ignoring the evaporative loss caused by molecular capture, the pick-up process is well described with Poisson statistics. The probability of capturing k particles is 16 ( ) ( ) exp ! k capt k capt n L I n L k σ γ σ γ = − (2.4) Here, capt σ is the capture cross section, L is the length of the pick-up cell, n is the number density of the dopant in the pick-up cell, and ~ / rel droplet v v γ accounts for the relative ve- locities in the pick-up cell, where rel v is the average relative velocity, and droplet v is the droplet beam velocity. 17 The number density can be written as B P n k T = (2.5) where P is the partial pressure in the pick-up cell, B k is the Boltzmann constant, and T is the temperature of the gas in the pick-up cell. Eq. (2.4) is plotted for Ir(ppy) 3 in Fig. 3.6. Although the pick-up process is presented above as straightforward, this is hardly the case. Despite being central to the helium droplet technique, the dynamics of the pick- up process remain unresolved. 12 For example, the conditions with which a particle is transmitted through a droplet, as well as the underlying physics involved are unanswered questions. The topic of molecular transmission will be discussed further in Chapter 3. 2.2 Multistep Photoionization The advent of lasers provided spectroscopists with the photon fluxes necessary to ionize atoms and molecules. Because ions are detected with unrivaled ease and 100% probability, 19 this offered a detection sensitivity that was unmatched by other spectro- scopic techniques. However, ionization potentials are frequently larger than the range of easily produced photon energies, so ionization often requires more than one photon. This 17 process, known as multistep ionization, enables a great deal of spectroscopic data to be acquired with a detection sensitivity that was hitherto unobtainable. As a result, multistep ionization spectroscopy has developed into a powerful and versatile tool. 2.2.1 Introduction Molecular multistep ionization refers to a process in which more than one photon is required to ionize a molecule. In general, this can refer to a wide variety of ionization schemes that fall into two broad categories. A distinction is made between multiphoton ionization and stepwise ionization. † Figure 2.2 illustrates these two basic techniques, each of which can be modified to provide a plethora of information. The following discussion provides an overview of these two schemes, while a more detailed description of stepwise ionization is presented, as it is more pertinent to the current work. The simplest scheme, shown in Fig. 2.2a, is stepwise ionization. As the name im- plies, molecular stepwise ionization is a process in which a molecule is ionized through a series of resonant single-photon absorption transitions. Although theoretically any num- ber of steps can be utilized, Fig. 2.2a displays a simple two-step ionization process. In this case, the first photon is absorbed creating an excited electronic state usually referred to as an intermediate state, and the second photon ionizes the molecule. † Unfortunately, the terminology can be ambiguous. While multiphoton is almost exclusively used in the sense it is presented in the text (there are exceptions in early publications on the subject), 20,21 the terms stepwise and multistep are often used interchangeably, although all of the described processes are both multistep and multiphoton, in the literal sense. 18 Figure 2.2 Two common multistep ionization schemes: (a) two-color stepwise ionization and (b) one-color multiphoton ionization. Early on, stepwise ionization was employed as a laser detector 19 and a two-dimen- sional optical mass spectrometer. 22 It offered vast improvements over electron impact ionization. For instance, electron impact ionization is nondiscriminatory: all molecules in the interaction region are subjected to ionization. On the contrary, a stepwise process enables molecules to be selectively ionized provided that the intermediate state has sharp features. 21 In fact, by utilizing unique transitions, isotope selectivity has been demon- strated for 13 C-labeled benzene, 20 among other complexes. 22 Other advantages include (a) (b) M M * M + + e - M M * M + + e - hv 1 hv 2 2hv hv 19 very high ionization efficiencies, enabling strong signals to be acquired from dilute sam- ples, and the ability to ionize without fragmentation, resulting in molecular parent ions. This is in stark contrast to electron impact ionization, where efficiency is low and frag- mentation is more often the rule. This technique, however, is not limited to mass spec- troscopy. Ionization can occur during a single laser pulse or with more than one pulse, enabling the conditions of each transition to be precisely controlled, which allows a large amount of spectroscopic and energetic data to be gleaned (§ 2.2.2). A molecular multiphoton process is defined as an absorption event that requires the simultaneous interaction of two or more photons with a molecule to reach a resonant state. Subsequently, the excited molecule can be ionized with additional absorption. The general scheme shown in Fig. 2.2b is referred to as resonance enhanced multiphoton ion- ization (REMPI). Although this technique appears to be simple, it is complicated by the large photon fluxes required to induce the improbable multiphoton absorption (see Eq. (2.6) and subsequent discussion). Nevertheless, multiphoton ionization has proven to be a highly effective means of obtaining spectroscopic data that may otherwise be elusive. For example, most molecules have interesting electronic transitions that lie in the ultraviolet region of the spectrum. 19,23 Such states may be difficult to probe using single photon absorption because producing high energy photons can be problematic. By applying an intense, focused laser pulse, it is possible to study these high-lying states by invoking a multiphoton process. Another advantage is the different set of selection rules for a two photon as opposed to a one photon transition. For example, transitions between states of the same parity are forbidden when an odd number of photons are absorbed, but allowed 20 when an even number of photons are absorbed. 22,23 A more detailed description of selec- tion rule changes can be found in the literature. 24 These characteristics have proven to be vital in the discovery and study of previously unknown states, 23 such as the 1 E 1g state of benzene. 22 Figure 2.3 Two level system: states |i〉 and | j〉 have corresponding energies E i and E j . Absorption occurs when the interacting photon has an energy hv, such that nhv = ∆ E = E i – E j , where n is the number of photons required to obtain resonance (called the transition order). If |i〉 and | j〉 are stationary states and the laser intensity is sufficient to saturate the transition, there is an equal probability for finding the mol- ecule in either state. The transition rate is proportional to some power of the laser intensity, which strongly affects the multistep ionization process. Consider the two-level system with states |i〉 and | j〉 shown in Fig. 2.3. Assuming that the population of the system initially lies in the ground state |i〉 , when a laser with photon energy hv (where i j E E nh − = ν ) is introduced, the transition probability per unit time, or transition rate, to state | j〉 is 19,25,26 n n W I σ = (2.6) where I is the intensity of the laser in units of photons cm -2 s -1 , n σ is the transition cross section in units of cm 2n s n-1 , and n is the number of photons required for the transition— E nhν ∆ = | i〉 | j〉 i E j E 21 referred to as the order. The absorption cross sections vary drastically for transitions of different order. In general, allowed single photon resonant excitations and single-photon ionizations have absorption cross sections on the order of 1 σ ~ ion σ ~ 10 -17 – 10 -18 cm 2 , while multiphoton processes requiring two and three photons have cross sections on the order of 2 σ ~ 10 -50 cm 4 s and 3 σ ~ 10 -82 cm 6 s 2 , respectively. 19,23 Under normal con- ditions, when the laser intensity is low, the transition rate is proportional to the signal in- tensity. 27 As the intensity is increased, the transition eventually becomes saturated and the signal becomes independent of the laser intensity. 26 Saturation occurs when 1/ sat σ Φ=Φ ≈ (2.7) Here, Φ is the laser fluence, given in units of photons per cm 2 , and is related to the inten- sity by 19 ( ) ( ) t I t dt τ −∞ ′ ′ Φ = ∫ (2.8) When the system is sufficiently saturated, and if | i〉 and | j〉 represent stationary states, the populations of the two states become equivalent. 25,28 Alternatively, if | j〉 represents the ionization continuum, the ionization efficiency approaches unity. 19 Therefore, when a transition is saturated, a further increase in fluence cannot result in additional absorption; the signal intensity is maximized. Clearly then, the laser fluence will have a significant impact on the multistep process. Although there are no exact boundaries and reality is more complex than idealized models, there are three general multistep ionization regimes that depend on the fluence. When the fluence is small and the saturation condition (Eq. (2.7)) is not met, stepwise 22 ionization leads to soft ionization and the formation of a parent ion. As the fluence is in- creased, stepwise ionization becomes highly efficient; usually the newly formed ion con- tinues to absorb photons. A further increase in fluence results in multiphoton transitions. Subsequently, many photons absorbed by the ion. Soft ionization refers to resonant stepwise photoionization in a weak field. In this case, each single photon transition is well below the saturation limit, i.e. sat Φ<<Φ . Because the stepwise process consists of resonant transitions at each step—a condition that is generally not met for the resulting ionic species—the probability of the ion ab- sorbing additional photons is small, and the process is terminated after the ionization step. 21 Consequently, ionization in this regime frequently produces parent molecular ions with no fragmentation. 21 In a weak field, the ionization scheme and signal dependence can be well characterized, although they are sometimes difficult to determine. As a result, this regime is useful for the elucidation of spectroscopic information. The intermediate regime is reached by increasing the laser fluence so that the res- onant transitions are saturated or nearly so. At these intensities, stepwise photoionization becomes highly efficient. After formation, the molecular ions, still bathed in intense radi- ation, often continue to absorb photons; fragmentation is likely to ensue. Under these conditions, a further increase in photon fluence will not necessarily lead to a larger ion yield. Ideally, this regime is used for efficiently ionizing trace amounts of complex mol- ecules in a gas sample. 19 With a further increase in laser intensity, nonlinear absorption begins to occur and multiphoton ionization is realized. The fluence is large enough that any allowed single 23 photon transition is saturated. If excitation to the intermediate state requires a multi- photon absorption, it is typically the rate limiting step because of the low probability for multiphoton transitions. 23 The subsequent ionization is usually resonant or nearly reso- nant and likely saturated, due to the large density of states at high energies and the large photon fluence needed to induce a multiphoton absorption, respectively. 23 As a result, many photons are absorbed after ionization and strong fragmentation results. 2.2.2 Obtaining Spectroscopic Information Information is gathered from multistep ionization in the form of ion signal. There- fore, it is absolutely necessary to understand what factors into the ionization probability. Also important is choosing the proper experimental regime, as this will affect the ion- yield dependence and therefore the available spectroscopic information. Stepwise ioni- zation is extremely versatile because each transition can be treated individually. More than one laser pulse can be employed at various time delays and fluences, allowing for gas-phase dynamic studies and acquiring spectroscopic information. Varying the laser fluence can also help elucidate important information about the absorption process. On the other hand, multiphoton ionization is frequently used for probing states that may otherwise be impossible to investigate. 2.2.2.i Stepwise Ionization Spectroscopy Perhaps the simplest and most general case of multistep ionization is that of the two-color, two-step ionization process shown in Fig. 2.2a. Although the majority of 24 results in this work were acquired with two-step ionization during a single laser pulse (one color), the more general two-color process will be described here because it is easily extended to other schemes (see, for example, Appendix A). In general, the ion signal de- pends on the absorption cross sections of each transition, the intermediate state dynamics, and the laser intensity. For the simple case when the intermediate state is long-lived relative to the ion- ization rate and absorption occurs in a weak field—i.e. sat Φ<<Φ for each tranfsition— the ion yield is 19 0 1 1 1 2 2 ( ) ( ) i ion N N σ ν σ ν = Φ Φ (2.9) where i N is the number of ions formed, 0 N is the number of molecules in the interaction region, ) ( 1 1 ν σ is the absorption cross section for the first transition, ) ( 2 ν σ ion is the ion- ization cross section from the intermediate state, and 1 Φ and 2 Φ are the photon fluences of the excitation laser and ionization laser, respectively. This equation becomes 2 0 1 1 1 ( ) ( ) i ion N N σ ν σ ν ∝ Φ (2.10) when ionization occurs during a single laser pulse (see Appendix A for derivation). The frequency variation of 1 ( ) i σ ν determines the excitation spectrum of the intermediate state. However, the recorded ion signal will not necessarily reflect this because both 1 ( ) i σ ν and ( ) ion i σ ν are frequency dependent. This can make the measured spectrum con- voluted and complicated. The ion yield from the intermediate state depends mainly on Franck-Condon factors. 22 However, it is not uncommon for ( ) ion i σ ν to have a weak spectral depend- 25 ence. 22,24 This is especially true when the ionizing photon takes the molecule to well above the ionization limit. 19 In fact, under these conditions, two-step ionization often reproduces the excitation spectrum of the intermediate state. 19 Still, the frequency de- pendence of the ionization cross section cannot be ignored. One technique to eliminate the ion-yield dependence on ( ) ion i σ ν is the use of a fixed frequency ionization laser, 2 ν h (Fig. 2.2a). By insuring that the ionization transition is saturated and 1 ion I I σ σ >> , 29 the ion yield given by Eq. (2.9) will depend solely on ) ( 1 1 ν σ . Then, the variation of the ion signal with 1 ν h will produce the intermediate-state excitation spectrum. Figure 2.4 Population loss of the intermediate state S n can compete with photoionization, leading to a de- pletion of the ion signal. Competitive pathways include (a) photon emission; (b) dissociation; (c) internal conversion (IC) to the single manifold; and (d) intersystem crossing (ISC) to the triplet manifold. While not always detrimental to the ion signal, these processes complicate the recorded signal. 0 S n S m S n T IP 1 hν 2 hν ( ) a luminescence ( ) b dissociation ( ) c IC ( ) d ISC 26 The ion yield also depends significantly on the intermediate state. 19 Electronically excited states are subject to a variety of relaxation pathways. Often the natural depop- ulation of the intermediate state leads to a decrease in ionization probability, 19 and there- fore knowledge of the excited state dynamics is advantageous. Luckily, if temporally short laser pulses are used, it is possible to study the kinetics of the intermediate state using stepwise ionization. 19,22 For example, the lifetime of the intermediate state can be ascertained by measuring the dependence of the ion signal on the laser pulse delay time. Other methods or schemes, however, may be required to uncover the exact nature of the relaxation pathway. Generally, population loss can stem from luminescence, internal conversion (IC), intersystem crossing (ISC), or photodissociation (Fig. 2.4). Of particular importance to stepwise ionization are radiationless transitions, namely IC and ISC (dissociation can also be significant and will be briefly addressed in Sect. 2.2.3). ISC is only a concern in the rare instance when it occurs on a timescale that is similar to that of the experiment. Also, unlike bound-bound transitions, there are no spin restrictions for photoionization, and therefore IC and ISC can be treated as one. As a result of radiationless relaxation, the molecule is left vibrationally excited. When the ion structure is similar to the neutral, as is often the case for large molecules, 29 Franck-Condon factors—i.e. vibrational overlap—dictate that the ion be formed with sig- nificant vibrational energy. 29 Thus, there will be an effective ionization energy (IE eff ) that is above the actual ionization energy (IE) by an amount close to the energy contained in the vibrations (Fig. 2.5). 23 If the increase in the IE is such that 2 IE eff hν < , the molecule is 27 unlikely to contribute to the ion signal. This can be overcome by increasing the fluence of the ionizing pulse, inducing a multiphoton transition to the ionization continuum. Figure 2.5 Illustration of the effective ionization energy. Absorption of hv 1 creates an excited electronic intermediate state. Radiationless deactivation (IC or ISC) to a lower electronic state leaves the molecule vibrationally hot. As a consequence, the probability of creating an ion in a low vibrational state—near the ionization threshold—is significantly reduced. The ion will likely be formed with considerable vibrational energy, introducing an effective IE that is larger than the IE threshold. / IC ISC IE effective IE 1 hv 2 hv 28 As mentioned previously, laser intensity plays a large role in photoionization. In the limit of a weak radiation field, the total order of the ionization process n is easily de- termined by measuring the change in ion-signal intensity with laser fluence. 19 log( ) log( ) i N n ∂ = ∂ Φ (2.11) This equation is also appropriate for determining the order of individual transitions. For example, a fluence study on the single-laser ionization scheme described by Eq. (2.10) would show that 2 = n , assuming sat Φ<<Φ . If, on the other hand, a two-color scheme was employed (Eq. (2.9)), the signal would have a first order dependence on each tran- sition, i.e. 1 = n . This type of measurement is important because it reveals information about the measured signal. For instance, if both transitions are considerably saturated ( 0 ≈ n ), the ionization efficiency will approach 100%, but spectroscopic information may be impossible to extract. If however, one transition was saturated and the other not, then the spectral signal would depend on the unsaturated transition. Generally, photoionization spectra are recorded in a regime where the ion signal depends significantly on the laser fluence. Consequently, when a spectrum is recorded over any significant frequency range, care must be taken to ensure that the laser intensity does not fluctuate over the region being scanned. 2.2.2.ii Multiphoton Ionization Spectroscopy Frequently, the transition to the intermediate state of interest requires more than one photon, and large laser fluences are required to induce the multiphoton process. 29 Above the intermediate state, the density of states is large, and the transition(s) to the ion- ization continuum is usually near resonance. Perhaps more importantly, the large laser fluence required to initiate the multiphoton absorption is not only sufficient to saturate the subsequent single-photon transitions, but also makes other multiphoton transitions possible. Therefore, as a rule, after the initial multiphoton absorption, the molecule con- tinues to absorb energy even after ionization. 19 This leads to extensive ion fragmentation and complex spectra. In the ideal case, it could be assumed that all other transitions are saturated, and the initial multiphoton absorption is the rate limiting step. Then, a fluence study would reveal the number of photons required to excite the intermediate state, and the ion signal would depend on some power n of the laser intensity and the cross section of the inter- mediate state, ) (ν σ n . Usually however, reliably interpreting the multiphoton spectra is much more challenging. Because high laser intensities induce many transitions, the re- corded spectrum can theoretically come from any combination of transitions. A further complication stems from the fact that there may be more than one multiphoton process that leads to the recorded ion signal. Fortunately, there are a large variety of different experimental schemes that can aid in determining the signal dependence. Further infor- mation on multiphoton ionization spectroscopy can be found in the literature. 24,29 2.2.3 Fragmentation Molecular multistep ionization can lead to diverse photoproducts; as more pho- tons are absorbed, higher appearance potentials are reached. The extent of photofrag- 30 mentation depends mainly, but not solely, on laser fluence. In the soft ionization regime, it is expected that the molecular parent ion will be produced almost exclusively. On the other hand, high photon fluxes can completely obliterate the molecule being studied, leaving only atomic ions. Photofragmentation can significantly alter the information gathered from the ion signal, so determining the conditions under which fragmentation occurs is important. For this reason, it is usually advantageous to record multistep ion signals with a time-of-flight (TOF) mass spectrometer (§ 2.3), enabling a complete mass spectrum to be recorded for every ionization event. Figure 2.6 Depiction of two fragmentation schemes for multistep ionization. (a) Photoabsorption by the parent ion opens up a variety of dissociation channels, which can eventually lead to atomization. (b) Dis- sociation of the intermediate state will lead to neutral fragments that can be ionized and dissociate further. Most often, photofragmentation is a result of the ion absorbing additional pho- tons, creating the various fragmentation pathways displayed in Fig. 2.6a. Because the two-step ionization process in a weak field stops after the ionization step, fragmentation is unusual. When fragmentation does occur, it is typically due to dissociation of the inter- mediate state (Fig 2.6b). This leads to neutral fragments that can then be ionized in the ABC * ABC ABC + * ABC + ** ABC + AB C + + * AB C + + A B C + + + IE ABC * AB C + + A B C + + + IE AB C + * AB C + AB C + + (a) (b) 31 radiation pulse. Here, the resulting ion signal will depend on the intermediate state ab- sorption cross section ) ( 1 ν σ and the absorption of the neutral fragments. The process is further complicated by the fact that the fragmentation can depend on the laser frequency. For example, there may be an energetic barrier for intermediate state dissociation, and with a fix number of absorbed quanta, more energy is deposited into the molecule with higher frequency photons. Increasing the laser fluence is synonymous with increasing the extent of absorp- tion. Therefore, higher fluences result in increased fragmentation. This makes spectro- scopic information, which is best given by the parent ion, 19 difficult to acquire. When the fluence is large enough to induce multiphoton transitions, extensive photofragmentation is the rule. Atomization is not rare, and the fluence dependence of such fragments will often reveal that many transitions are saturated. For instance, the appearance potential of C + from benzene is ~27 eV but the fluence dependence showed n = 3.5 when no less than 9 photons at 3.17 eV were required for the formation of the bare carbon. 19 In the case of organometallic complexes, atomization is the rule, and the resulting frequency depend- ence often reflects the metal transitions. 29 Employing a TOF mass spectrometer for ion detection is beneficial. This method allows for the recording of an entire mass spectrum every time ions are formed. As the laser frequency is scanned, different photofragments may appear, and a two-dimensional mass spectrum is formed. In this way a large amount of data can be collected during an experiment. By doing these measurements at a variety of different laser intensities, another dimension is added, providing the fluence dependence of each fragment ion. This 32 may shed light on the fragmentation process because different fragments have different appearance potentials. Overall, analysis of the mass spectra can aid in understanding the complexities of the multistep process. 22 2.2.4 Summary Molecular multistep ionization is a complex but powerful technique. Myriad ion- ization schemes lead to the production of ions, resulting in unsurpassed detection effi- ciency and making large amounts of data available. By far, the most crucial component in any scheme is the choice of laser fluence. It determines the number of photons absorbed, which transitions are saturated, the extent of photofragmentation, and ultimately what information can be gathered from the resulting ion signal. The desired spectroscopic in- formation must be determined by carefully characterizing the ionization process and choosing the proper experimental conditions. 2.3 Time-of-Flight Mass Spectroscopy Time-of-flight (TOF) mass spectroscopy uses electric fields to separate ions by their mass-to-charge ratio, m / q. Herein it will be assumed that all ions are formed via laser ionization and have a charge of q = 1. A detailed description of this technique can be found in the famous work by Wiley and McLaren. 30 The following discussion highlights some important aspects. 33 Figure 2.7 Diagram of the basic TOF setup. The ionization region has diameter of ∆ s that is centered at a position s, measured from the extractor grid. The ions are accelerated toward the detector by the electric field E 1 , which is controlled by varying the voltages applied to the repeller plate and extractor grid. The ions are accelerated further by electric field E 2 . The field-free region has a length D, and acts to separate ions by mass. A micro-channel plate (MCP) serves as the detector. 2.3.1 Overview A simple TOF mass spectrometer is shown in Fig. 2.7. It consists of an ionization region, a field-free evacuated drift tube, and an ion collector in the form of a micro- channel plate (MCP) detector. Ions are formed in the ionization region and are accel- erated by two electric fields produced by applying a voltage to the repeller plate and ex- tractor grid. The accelerating electric fields can be pulsed on after the ions are formed or can be constant, depending on the required conditions. MCP D d s s ∆ 1 E ↑ 2 E ↑ 0 E = Repeller Ionization Region Extractor 34 Consider the hypothetical case in which ions are created in an infinitely thin plane parallel to the charged plates with no initial kinetic energy. Then, each ion acquires the same energy, independent of mass, given by 1 2 U sE dE = + (2.12) Here, s is the distance the ions travel in the first electric field E 1 , d is the distance the ions travel in the second electric field E 2 , and q is assumed to be 1. Because the acquired translational energy is identical, each ion will have a velocity that depends strictly on its mass m given by 1 2 2U V m   =     (2.13) Thus, as the accelerated ions traverse the field-free region they are separated by mass. The flight times of the different ionic species are given by 1 2 1 2 1 2 0 0 1 2 0 2 ( , ) 1.02 2 2 1 k m T U s k s d D U k     = + +     +     (2.14) where U is given by Eq. (2.12) and 0 k is defined as 0 1 2 1 ( ) / k sE dE sE = + (2.15) Ions arrive at the detector at various times according to their mass. A mass spectrum is formed by recording the ion signal as a function of time and using Eq. (2.14) to convert arrival-time to mass. TOF mass spectrometry has many unique advantages. This method enables an en- tire mass spectrum to be recorded for every ionizing event, with nearly 100% trans- mission. In fact, under typical operating conditions, a TOF mass spectrum is collected in 35 under 50 µs. Thus, data collection is often limited by the laser pulse rate and the acqui- sition electronics. Furthermore, the accuracy of this technique depends on electric circuits as opposed to the precise mechanical construction and alignment that is required for other types of mass spectroscopy. 2.3.2 Resolution Equations (2.12) – (2.15) provide only an approximate description of the TOF technique. According to these equations, ions of the same mass will be detected simul- taneously. However, the ions have initial space and energy distributions, which cause ions with the same mass to have a range of flight times. As a result, ion signals have a time distribution that severely limits the overall mass resolution M. The space distribution is due to the finite geometric cross section of the ionizing laser. If the ionizing region is centered at s and has a diameter s ∆ , as depicted in Fig. 2.7, then ions will be formed at various distances from the extractor grid. The initial positions will range from a maximum value of 1 2 max s s s = + ∆ (2.16) to a minimum value of 1 2 min s s s = − ∆ (2.17) According to Eq. (2.12), the ions formed at s max will gain more translational energy and have a longer flight path than those formed at s min . As a result, ions with the same mass will have different flight times, resulting in a time spread s T ∆ , where 36 ( , ) ( , ) s max min T T U s T U s ∆ = − (2.18) However, because the more energetic ions have a longer flight path, there will be an in- flection point at which the faster ions pass the slower ones. Space focusing is achieved by adjusting the TOF parameters so that this inflection point occurs at the detector, signif- icantly reducing s T ∆ . The focusing condition for which ions of the same mass will reach the detector simultaneously, regardless of energy, is mass independent and given by 0 0 3 2 k d D s k s   − =     (2.19) Usually D, s, and d are fixed. Therefore, 0 k should be adjusted—by altering E 1 and E 2 — so that this condition is met. The space resolution M s is the largest mass m at which there is no overlap between adjacent mass peaks (Fig. 2.8a). That is, M s is the largest mass m that satisfies 1 2 m s m m T T T T m + ∆ = − ≈ (2.20) Molecules in a volume of gas—for instance, in the interaction region—have a velocity distribution given by the Maxwell distribution law. Therefore, molecules possess a range of kinetic energies U 0 that must be taken into account after ionization occurs. After acceleration, the ions will have a translational energy given by 0 t U U U = + (2.21) 37 Figure 2.8 Graphical representation of the TOF mass resolution M achieved for the experiment described in Chapter 4. (a) Two Gaussian curves represent ion signals corresponding to 130 amu (red circles) and 131 amu (blue squares); the overall signal is shown by a solid black curve. Here M is 130 amu because this is the largest mass for which the adjacent mass peaks are completely separated. (b) The highest resolvable mass is 260 amu (red circles). At this mass, the neighboring mass peak, 261 amu (blue squares), overlaps with the 260 peak; however, the overall peak intensities are not affected (solid curve). Therefore, the highest resolvable mass is 2M. Consider two molecules at the same position s with maximum velocities in oppo- site directions, as shown in Fig. 2.9. Once ionized, the ion that was initially moving to- ward the detector will be accelerated, while the ion moving away from the detector will decelerate, change directions, and then accelerate toward the detector. Although both ions will acquire the same final value of U t , they will arrive at the detector with a time dif- ference 0 U T ∆ , where ( ) 0 1 2 0 1 2 2 1.02 U mU T E ∆ = (2.22) The energy resolution M u can then be determined from Eq. (2.20). Specifically, 129 130 131 132 mass / amu 258 259 260 261 262 263 mass / amu (a) (b) 38 0 2 m U U T M T = ∆ (2.23) M u can be increased by introducing a time delay between ionization and the application of E 1 , which is not relevant to the current work, and is described elsewhere. 30 0 U T ∆ could also be reduced by either increasing the energy ratio (U 0 / U) or increasing the flight dis- tance D. However, both of these methods will adversely affect the space resolution. Per- haps the simplest way to improve the energy resolution is to decrease U 0 . This can be achieved by using a highly directional, supersonically expanded molecular beam. 3 Figure 2.9 Diagram showing the effects of the initial energy distribution on the TOF mass resolution. (a) Two identical particles A and B have a translational energy U 0 ; A is moving away from the detector, B is moving toward it. At time t = 0, both particles are ionized at position s. Electric field E 1 causes A to slow down and eventually turnaround, while B is accelerated. (b) At time ∆ T U0 , the particles have moved from their original positions (shown as dotted circles). A arrives at point s with a translational energy U 0 . The ions proceed to the detector separated in time by ∆ T U0 . A B 1 E ↑ r 0 U 0 U A 0 U A B 1 E ↑ r B s 0 t = 0 U t T =∆ Ionization Region s (a) (b) 39 The overall mass resolution M is difficult to calculate, and is best described by an upper and lower bound. The upper limit is the smaller of either M s or M u . The lower bound M s,U is defined as , 1 1 1 s U s U M M M = + (2.24) As Fig. 2.8 shows, masses up to 2M, where M is the actual mass resolution, are resolv- able because the total signal intensities are unaffected by neighboring mass peaks at this mass. 40 2.4 References 1. Anderson, J. B.; Andres, R. P.; Fenn, J. B. Advances in Atomic and Molecular Physics; Bates, D. R. Ed.; Academic Press: New York, NY, 1965; Vol. I; pp. 345 – 359. 2. Ramsey, N. F. Molecular Beams; Oxford University Press: New York, NY, 1985. 3. Miller, D. R. Atomic and Molecular Beams Methods; Scoles, G. Ed.; Oxford University Press: New York, NY, 1988; Vol. I; pp. 14 – 53. 4. Mark, T. D.; Dunn, G. H. Electron Impact Ionization; Springer-Verlag / Wien: New York, NY, 1985. 5. Demtroder, W. Laser Spectroscopy: Basic Concepts and Instrumentation, 2 nd ed.; Springer: New York, NY, 1998. 6. McDaniel, E. W. Atomic Collisions: Electron and Photon Projectiles; John Wiley & Sons, Inc.: New York, NY, 1989. 7. Toennies, J. P.; Vilesov, A. F. Annu. Rev. Phys. Chem. 1998, 49, 1. 8. Toennies, J. P.; Vilesov, A. F. Angew. Chem. Int. Ed. 2004, 43, 2622. 9. Stienkemeier, F.; Vilesov, A. F. J. Chem. Phys. 2001, 115, 10119. 10. Slipchenko, M. N.; Kuma, S.; Momose, T.; Vilesov, A. F. Rev. Sci. Instrum. 2002, 73, 1. 11. Northby, J. A. J. Chem. Phys. 2001, 115, 10065. 12. Barranco, M.; Guardiola, R.; Hernandez, S.; Mayol, R.; Navarro, J.; Pi, M. J. Low Temp. Phys. 2006, 142, 1. 13. Stienkemeier, F.; Lehmann, K. K. J. Phys. B 2006, 39, R127. 14. Close, J. D.; Federmann, F.; Hoffmann, K.; Quaas, N. Chem. Phys. Lett. 1997, 276, 393. 15. Lewerenz, M.; Schilling, B.; Toennies, J. P. Chem. Phys. Lett. 1993, 206, 381. 16. Harms, J.; Toennies, J. P.; Dalfovo, F. Phys. Rev. B 1998, 58, 3341. 17. Lewerenz, M.; Schilling, B.; Toennies, J. P. J. Chem. Phys. 1995, 102, 8191. 41 18. Polyakova, E.; Stolyarov, D.; Zhang, X.; Kresin, V. V.; Wittig, C. Chem. Phys. Lett. 2003, 375, 253. 19. Letokhov, V. S. Laser Photoionization Spectroscopy; Academic Press, Inc.: Orlando, FL, 1987. 20. Boesl, U.; Neusser, H. J.; Schlag, E. W. J. Am. Chem. Soc. 1981, 103, 5058. 21. Schlag, E. W.; Neusser, H. J. Acc. Chem. Res. 1983, 16, 355. 22. Antonov, V. S.; Letokov, V. S. Appl. Phys. 1981, 24, 89. 23. Johnson, P. M. Acc. Chem. Res. 1980, 13, 20. 24. Ashfold, M. N. R.; Howe, J. D. Annu. Rev. Phys. Chem. 1994, 45, 57. 25. Abramczyk, H. Introduction to Laser Spectroscopy; Elsevier: Amsterdam, 2005; pp. 1 – 18. 26. Levenson, M. D. Introduction to Nonlinear Laser Spectroscopy; Academic Press: New York, NY, 1982. 27. Atkins, P. W.; Friedman, R. S. Molecular Quantum Mechanics, 3 rd ed; Oxford University Press: New York, NY, 1997. 28. Hollas, M. High Resolution Spectroscopy; John Wiley & Sons: New York, NY, 1998. 29. Johnson, P. M.; Otis, C. O. Ann. Rev. Phys. Chem. 1981, 32, 139. 30. Wiley, W. C.; McLaren, I. H. Rev. Sci. Instrum. 1955, 26, 1150. 42 Chapter 3 Tris(2-phenylpyridine)iridium Helium Droplet Study 3.1 Introduction Obtaining detailed energetic information from large organic molecules is chal- lenging. Such information is typically acquired via high resolution spectroscopy, which is facilitated by the familiar supersonic molecular beam spectroscopy. However, large mol- ecules have many internal degrees of freedom that may not be efficiently cooled by the rapid expansion. As a result, these spectra are usually congested with broadened features and difficult to interpret. Another problem encountered with this technique lies in obtaining the high vapor pressures necessary for expansion; achieving sufficient cooling requires high source pressures and therefore nonvolatile organic compounds must be heated to high temperatures. This can cause pyrolysis, making adequate signal impossible to achieve. These problems are avoided to a large degree with matrix isolation spectroscopy. The use of crystalline rare-gas or hydrogen matrices enables molecules to be cooled to sub-Kelvin temperatures, 1 while the concentration of the embedded species is easily adjusted by varying the deposition time. This method has a unique set of hurdles, how- ever. Strong interactions between the dopant and host matrix can result in significant changes to the electronic spectrum. Furthermore, embedded molecules can reside in vastly different host sites and / or be frozen in a variety of different conformers, leading to 43 considerable inhomogeneous broadening. Consequently, spectral analysis is difficult and the inferred energetic information ambiguous. Superfluid helium droplet spectroscopy is a nearly perfect marriage of these two techniques. A detailed discussion of this technique is found in Sect. 2.1. Briefly, droplets are produced by supersonically expanding helium at cryogenic temperatures. Adjusting the source temperature and pressure allows the average number of helium atoms per droplet 〈 N 〉 to be varied from 10 3 to 10 7 . The molecules of interest are present in a heated pick-up cell downstream from the droplet source. Collisions in the pick-up cell typically lead to capture. The vapor pressures required for this process are only a fraction of what is necessary for direct molecular expansion; hence, molecular decomposition is less likely. In addition, the helium matrix cools the embedded molecule to 0.38 K, while the unique quantum mechanical delocalization of the helium enables the liquid matrix to adapt itself to the equilibrium geometry of the molecule. These characteristics eliminate most of the hot bands and inhomogeneous broadening that complicate the molecular beam and solid matrix spectra, respectively. Examples of electronic spectra recorded via helium droplet spectroscopy are myr- iad 2 and demonstrate the unique versatility of this method. For instance, helium droplets enabled the vibronically resolved spectra of the 3,4,9,10-perylene-tetracarboxylic dian- hydride (PTCDA) molecule to be recorded for the first time. 2–4 PTCDA films are used as organic semiconductors in light emitting devices and have been studied previously in the gas-phase, 3 in DMSO solution, 5 and on a quartz substrate. 6 Each of these spectra was 44 hampered by broad charge transfer excitonic transitions. In helium droplets, narrow molecular transitions were observed. Helium droplet spectroscopy also shows promise as an effective means for probing large, floppy biological molecules as demonstrated by experiments with the amino acids tryptophan and tyrosine. 7 Other studies demonstrate the ability of droplets to capture and cool larger molecules, including C 60 8 and phthalocyanine (C 32 H 18 N 8 ), 2,9 both of which were examined in relatively large 〈 N 〉 ≥ 20,000 droplets. Tris(2-phenylpyridine)iridium (Ir(ppy) 3 ) is an excellent candidate for helium droplet spectroscopy. Ir(ppy) 3 is arguably the most important molecule used in organic light emitting diode (OLED) displays. Significant strides have been made in under- standing this phosphorescent molecule owing to a plethora of experimental and theoret- ical work. However, high resolution spectra have yet to be recorded. Obtaining the high resolution electronic spectrum of Ir(ppy) 3 by utilizing helium droplet spectroscopy was the original goal of the work presented in this chapter. These goals have yet to be reached; however, during the course of this work a more fundamental aspect of helium droplet spectroscopy was examined. The results pre- sented herein provide insight into the helium droplet pick-up process, which, although touted as straightforward, remains one of the most understudied and least understood facets of helium droplet spectroscopy. 10 Despite the common assumption that most col- lisions lead to capture, it will be shown that droplets with 〈 N 〉 values of 8000 do not efficiently capture hot Ir(ppy) 3 . 45 3.2 Experimental Apparatus The experimental setup used to complete this work includes a helium droplet machine with two mass spectrometers in conjunction with laser and data acquisition sys- tems. This section provides an overview of the vacuum and laser systems; details that are important to the following sections are also presented. Because the experimental arrange- ment has been described in rigorous detail previously, various elements that are omitted here can be found elsewhere. 11,12 Figure 3.1 Schematic representation of the vacuum chamber. Approximate dimensions are shown in mm. Four differentially pumped chambers are labeled I – IV and correspond to the source, scattering, detection, and QMS chambers, respectively. Abbreviated labels are as follows: DP = diffusion pump; TP = turbo- molecular pump; TOF MS = time-of-flight mass spectrometer; QMS = quadrupole mass spectrometer. Nozzle DP TP TP TP Cryostats Pick-up Cell LN 2 TOF MS QMS I II III IV Gate Valves Droplet Beam Laser Beam Cold Shrouds Ionizer 40 360 390 46 3.2.1 Vacuum Chamber A schematic of the helium droplet machine, consisting of four differentially pumped chambers, is presented in Fig. 3.1. The droplet beam is produced in the source chamber (I), where a 6000 L / s diffusion pump (Edwards Diffstack) is used to maintain a pressure of 10 -5 – 10 -6 torr. This chamber is connected to the scattering chamber (II) by a 600 µm skimmer. The pick-up cell, which is used to introduce gas to the droplet beam, lies just beyond the skimmer. Hence, chamber II is also known as the pick-up chamber. Two detection chambers follow. The first (III) houses a time-of-flight (TOF) mass spec- trometer that is used in conjunction with photoionization. Connected to this chamber by a gate valve is (IV) the quadrupole mass spectrometer (QMS) (Balzers QMS 511), equipped with an electron impact ionizer. Table 3.1 Helium droplet beam parameters T nozzle (K) P He (atm) N 〈 〉 a R (Å) b σ tot (Å 2 ) c I (sr -1 s -1 ) 14 v droplet (m / s) 15 16 40 8000 44 6200 7 x 10 15 350 a See Sect. 2.1, and Refs. (4) and (13) b Droplet radius R = 2.22N 1/3 Å 4 c σ tot = πR 2 (Eq. (2.2)) Helium droplets are formed by expanding helium gas (Matheson Tri Gas, 99.9999% pure) through a 5 µm nozzle (National Aper ture). The nozzle is attached to a closed-cycle cryostat (Advanced Research Systems) by a copper braid and the temper- ature is monitored with a silicon diode (Lakeshore). The thermal connection between the nozzle and the cold head is of utmost importance, and is often the largest obstacle to overcome when optimizing the droplet beam. Ideally, nozzle temperatures of 16 K are 47 achieved when the helium gas is pre-cooled by a separate cryostat (CTI Cryogenics). The backing pressure of the helium gas can be varied from 10 – 60 atm, which allows for some control over the mean droplet size (§ 2.1). Data presented in the next section were recorded using a nozzle temperature of 16 K and a helium pressure of 40 atm. Under these expansion conditions, the mean droplet size is 〈 N 〉 = 8000. 13 This and other beam parameters are displayed in Table 3.1. After travelling through the skimmer, the droplets encounter the pick-up cell. This cell is a hollow cylinder with a 1 cm diameter. There are two round holes cut opposite each other so that the droplets can traverse the cell. The Ir(ppy) 3 powder † is placed on the bottom of the cylinder, which is closed by a threaded cap on both ends. The pick-up cell is wrapped in resistive heating wire that is insulated with ceramic beads (Omega). A power supply is used to apply voltage across the wire; the temperature of the pick-up cell is monitored using a thermocouple. Because the pressure in the pick-up cell cannot be measured directly, the temperature is optimized by monitoring the partial pressure of Ir(ppy) 3 with the micro-ion gauge (Granville Phillips) located in the scattering chamber. Previous experiments using this arrangement have shown that a partial pressure of approximately 10 -7 torr is sufficient for particle capture (Appendix B); 11 this pressure is reached when the pick-up cell temperature is held at 500 K. Despite the use of a turbomolecular pump (Leybold Turbovac), the gas-phase Ir(ppy) 3 in the scattering chamber unavoidably effuses into the TOF chamber. The back- † The high purity fac-Ir(ppy) 3 used for these studies was acquired from the research group of Professor Mark Thompson at the University of Southern California. It is also available commercially from Sigma- Aldrich. 48 ground Ir(ppy) 3 in the detection chamber is minimized by the use of two liquid nitrogen cooled copper plates, and a turbomolecular pump (Leybold Turbovac). During the exper- iments, the partial pressure in this chamber is 5 x 10 -9 torr. The TOF mass spectrometer was built in-house, and is arranged according to Fig. 2.7. It was calibrated by photoionizing NO. During the course of this work, voltages of 3 kV and 2 kV were applied to the repeller plate and extraction grid, respectively. Detailed information concerning the operation, dimensions, and optimization of the TOF spectro- meter can be found in Sect. 2.3 and 4.2, as well as Ref. (12). The TOF detector cannot be used to analyze the pure droplet beam because it relies on laser ionization of the impurity. The QMS—located in the last chamber—is equipped with an electron impact ionizer, which is ideal for detecting both pure and doped helium droplets. Therefore, the QMS is employed for beam optimization. To mini- mize background gas signal, the QMS chamber is evacuated with a turbomolecular pump (Leybold Turbovac), and is kept around 10 -9 torr. Obtaining a consistent droplet beam is a challenge with this machine because the nozzle temperature is difficult to stabilize. A related complication is the rather high mini- mum nozzle temperature of 16 K. These problems arise primarily from poor thermal con- tact between the nozzle and the cryostat head, which is most apparent when the nozzle position is adjusted. The problem is exacerbated by radiation emitted from the nearby pick-up cell. According to Fig. 2.1, large droplets are produced by lowering the nozzle temperature or increasing the backing pressure. However, because of the precarious noz- zle setup, when the backing pressure is increased, a larger load is placed on the nozzle 49 causing the temperature to rise. Consequently, helium droplets with 〈 N 〉 values signifi- cantly larger than 8000 are not only difficult to produce, they are impossible to maintain through the course of an experiment. 3.2.2 Laser System Photoionization prior to TOF mass analysis is accomplished with a commercial nanosecond laser system. A 20 Hz Nd:YAG laser (Continuum Powerlite 9020) pumps a dye laser (Continuum ND 6000). The dye output is converted using a frequency doubling crystal inside a UV frequency extender (Continuum UVT-3). A Pellin-Broca prism is set- up outside of the UV extender to counteract the walking of the laser beam as the wave- lengths are scanned. Photons in the range 284 – 285 nm are generated by pumping rhodamine 590 (Exciton) dye with the second harmonic (532 nm) of the Nd:YAG. Doubling the dye out- put resulted in pulse energies of 9 mJ / pulse. The laser was aligned perpendicular to the droplet beam and focused to approximately 0.7 mm with a 60 cm lens, resulting in a pulse fluence of 2.3 J / cm 2 . The position of the laser was optimized with respect to the droplet beam by doping the droplets with NO and maximizing the corresponding TOF signal. 3.3 Ir(ppy) 3 Internal Energy It will be advantageous for the discussion of the experimental results to have an estimate of the internal energy of Ir(ppy) 3 at 500 K. Ir(ppy) 3 is comprised of 61 = N 50 atoms, and thus has 183 3 = N degrees of freedom. Translational and rotational motion each account for 3 degrees of freedom, while the remaining 6 3 − N are attributed to mo- lecular vibrations. According to the equipartition theorem, when at thermal equilibrium, each degree of freedom that appears quadratically in the energy contributes 1/ 2 B k T to the total average internal energy of the system, where B k is the Boltzmann constant and T is the temperature. 16 For the purpose of this discussion, † the internal energy is tr rot vib U U U U = + + (3.1) The translational energy tr U and rotational energy rot U depend quadratically on velocity and angular velocity, respectively. The vibrational energy vib U is the sum of kinetic and potential energy terms, which depend quadratically on momentum and position, respect- tively. 17 Therefore, by the equipartition theorem, the general expression for average inter- nal energy is ( ) ( ) ( ) 1 1 2 2 3 3 3 6 B B B U k T k T N k T = + + − (3.2) which gives 3 ) ( ppy Ir U = 63000 cm -1 at 500 K. However, this is a severe overestimation because it does not take into account the quantization of energy. According to the Boltz- mann distribution, if the spacing between energy levels for a particular degree of freedom i U ∆ is significantly larger than the thermal energy B k T , then that degree of freedom is frozen out and will not contribute to the internal energy. 16 At 500 K the thermal energy B k T is 350 cm -1 . Generally, both tr U ∆ and rot U ∆ are much smaller than this, 16 and † When the pressure of the gas is low, intermolecular interactions are negligible, so the corresponding energy U inter can be neglected. 16 In addition, the energy term due to electronic excitation U el is neglected because it is typically only significant at temperatures above 10 4 K. 16 51 hence, each contribute 3/ 2 B k T to the internal energy. On the other hand, vib U ∆ is often larger than 350 cm -1 . As a consequence, ( ) 3 6 vib B U N k T < − (3.3) Thus, the error in 3 ) ( ppy Ir U is caused by the last term on the right side of Eq. (3.2). Because of the U vib term in Eq. (3.1), calculating the internal energy of a large molecule at moderate temperatures is an arduous task. Fortunately, if it is assumed that the vibrational frequencies of 2-phenylpyridine (ppy) are not significantly altered when coupled to the iridium atom, U vib for Ir(ppy) 3 can be approximated from U vib of the three ppy ligands. This assumption is supported by myriad studies on substituted aromatics that show only small shifts in the normal mode frequencies from those of the respective parent molecules (see Refs. (18), (19), and refer- ences therein). For example, the normal modes of ppy were easily assigned by comparing the measured frequencies with those of substituted benzene and pyridine molecules. 18,19 The vibrational energy of ppy at 500 K can be calculated using the vibrational partition function. For a polyatomic molecule this yields 16 ( ) 3 6 1 1 1 i B N h k T vib i i U h e ν ν − − = = − ∑ (3.4) where h is the Planck constant and ν i is the frequency of the i th vibrational mode. Using the known vibrational frequencies of ppy, 18,19 this gives ppy vib U = 3400 cm -1 . † † Although ppy has 57 normal modes, only 46 are observed by Sett et al. Careful examination of Ref. (18) enables some of the missing frequencies, including three ring modes and five carbon-hydrogen modes, to be approximated. As a result, 54 modes were used in the calculation of vib U for ppy. The remaining three modes are less intuitive, and at most each can contribute B k T to the overall vibrational energy. However, 52 As a first approximation, the vibrational energy of Ir(ppy) 3 is obtained by multi- plying ppy vib U by three. Again, using the assumption that the vibrational frequencies of ppy do not exhibit dramatic shifts when coupled to the iridium atom allows ppy vib U to be used. This approximation is improved by taking into account the modes arising from the iridium-ligand coupling. The large reduced mass of these couplings ensure that the corre- sponding frequencies are low. Assuming each of the three modes contributes the maxi- mum value B k T to the vibrational energy gives 3 ( ) 3 3 Ir ppy ppy vib vib B U U k T = + (3.5) 3 ) ( ppy Ir vib U is 11250 cm -1 . This rough approximation overestimates the contributions to 3 ) ( ppy Ir vib U from the iridium-ligand coupling, but also neglects a total of 12 vibrational modes. Consequently, 11250 cm -1 is probably a low estimate. Nevertheless, by Eqs. (3.1) and (3.5), 3 ) ( ppy Ir U is approximately 12300 cm -1 . 3.4 Results The droplet beam is monitored using the quadrupole mass spectrometer (QMS). The spectrum of pure helium droplets, containing on average 8000 atoms, is shown in Fig. 3.2. The intense signal due to He 2 + is indicative of helium droplet ionization via elec- tron impact. Electron bombardment of droplets also yields larger helium ion clusters: referred to as a helium droplet progression. This is visible as a series of peaks labeled it is unlikely that the actual value differs considerably from 3400 cm -1 . Anyways, for the discussion in Sect. 3.4, a lower limit is more useful. 53 He n + . In addition, the spectrum exhibits H 2 O + , N 2 + , and O 2 + peaks due to background. These are routinely observed in high vacuum systems. Figure 3.2 Quadrupole mass spectrum of pure helium droplets with a mean droplet size of 8000 helium atoms. The labeled He n + peaks, where n ≥ 2, are a signature of helium droplet ionization. Prominent ion peaks corresponding to background gas are also labeled. When the pick-up cell is heated to 500 K, Ir(ppy) 3 is present in the droplet beam path. The effect on the droplet beam is observed by recording another mass spectrum. Specific features from Fig. 3.2 (solid red line) together with the analogous features of the spectrum recorded with a hot pick-up cell (dashed black line) are shown in Fig. 3.3. In the range 15 – 21 amu, the OH + , H 2 O + , and H 3 O + signals increase when the sample is 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 arb. mass / amu He n + N 2 O 2 H 2 O He 2 + 54 heated. This is caused by water that is released from the hot Ir(ppy) 3 sample (and possibly nearby cell walls) being captured by the droplets and carried to the detection chamber. Figure 3.3 Details of the spectrum presented in Fig. 3.2 (solid red line), along with the analogous spectrum recorded with Ir(ppy) 3 present in the pick-up cell at 500 K (dashed black line). The left side of the figure spans the range 15 – 21 amu, and shows ion signals due to H 2 O. These peaks all increase when the pick-up cell is heated. The right side of the figure (47 – 66 amu) has been enlarged 5 times and shows a small part of the helium droplet progression. The H 3 O + signal can only arise from droplets containing more than one water molecule. Thus, under these conditions, collisions between the droplets and H 2 O in the pick-up cell are abundant. After heating and cooling the sample several times under vac- uum, the sample dries, and the water peaks diminish. Because the partial pressure in the pick-up chamber does not decrease during this process, it is assumed that Ir(ppy) 3 was the 16 18 20 48 50 52 54 56 58 60 62 64 66 Pure Helium Droplets Droplets with Ir(ppy) 3 at 500 K arb. mass / amu OH + H 2 O + H 3 O + (x 5) He n + n = 12 – 16 55 major contributor to the partial pressure in the pick-up cell when these data were col- lected. Thus, the H 3 O + peak in Fig. 3.3 is evidence that collisions between the droplets and Ir(ppy) 3 are plentiful. The right side of Fig. 3.3 spans the range 46 – 66 amu and presents a small but representative portion of the droplet progression. Although the progression is still observ- able, its intensity decreases by approximately 40% when gas is present in the pick-up cell. Reasons for such a change include: (1) droplets that have captured a molecule have a lower probability of forming He n + because this process has to compete with ionization of the embedded species; 20,21 (2) collisions with Ir(ppy) 3 scatter droplets away from the ion- ization region; and (3) collisions cause droplets to shrink, which decreases their ion- ization probability. These processes are discussed at length in the next section. Using droplets with 〈 N 〉 = 8000, no signal due to Ir(ppy) 3 ionization is observed up to the maximum QMS mass of 511 amu. On the other hand, evidence of Ir(ppy) 3 is seen for larger droplets. When the helium backing pressure is increased from 40 to 60 atm with the nozzle at 16 K, 〈 N 〉 increases to approximately 12000 (§ 2.1). Ionization of these larger droplets reveals a small peak around 154 amu, which is attributed to ppy + . However, this peak represents well under 1% of the total ion signal. Due to experimental limitations, larger droplets could not be produced. Though the low intensity of the ppy + signal prevented detailed characterization, it is unambiguous that the ppy + signal appears only with the larger droplets. The presence of Ir(ppy) 3 in the vacuum chamber was further confirmed by laser ionization time-of-flight (TOF) mass spectroscopy. Because this technique does not nec- 56 essarily discriminate between Ir(ppy) 3 -doped helium droplets and gas-phase Ir(ppy) 3 that effuses into the vacuum chamber from the pick-up cell, an effort was made to minimize the gas-phase signal. Despite two cold traps near the ionization region, Ir(ppy) 3 back- ground signal persisted. Thus, TOF signals from Ir(ppy) 3 embedded in droplets were ob- tained by subtracting background TOF spectra from total TOF signals. To avoid com- plications from any difference in ionization efficiency between gas-phase and embedded Ir(ppy) 3 , fluence of 2.5 J / cm 2 was used, which ensures efficient ionization. Figure 3.4 TOF spectra of Ir(ppy) 3 recorded with a photon energy of 35,047 cm -1 and a fluence of 2.3 J / cm 2 . Panel (a) was collected with a 500 K pick-up cell containing Ir(ppy) 3 and N 〈 〉 = 8000 droplets. Panel (b) was recorded under similar conditions, but with the helium droplet beam blocked by beam flag. The y- axes for (a) and (b) are identical. The x-axis includes all of the ion peaks attributed to Ir(ppy) 3 : signal corresponding to isotopic Ir + is clearly visible at 191 and 193 amu, while smaller peaks at 203, 205, and 154 amu are due to 191 IrC + , 193 IrC + , and ppy + , respectively. The mass spectrum due to Ir(ppy) 3 doped drop- lets is obtained by subtracting (a) from (b), and is presented in (c). The y-axis in (c) is 40 times smaller than that of the top two panels. The only observable peak in the difference spectrum corresponds to Ir + ; how- ever, the magnitude of this signal is below the noise level of the top two spectra. ppy + Ir + IrC + 140 150 160 170 180 190 200 210 220 -0.5 0.0 0.5 1.0 0 20 40 60 0 20 40 60 mass / amu (c) Difference arb. (b) Droplet Beam Off (a) Droplet Beam On 57 35040 35060 35080 35100 35120 35140 35160 35180 -0.2 0.0 0.2 0.4 0.6 0.9 0.6 0.9 (c) Difference energy / cm -1 (b) Droplet Beam Off signal intensity / arb. (a) Droplet Beam On Figure 3.5 Action spectra of Ir(ppy) 3 collected with the helium droplet beam on (a) and off (b), with a fluence of 2.3 J / cm 2 and a mean droplet size of N 〈 〉 = 8000. Each point is the average of 12,000 laser shots; there are 100 points per spectrum. The difference between the two spectra is shown in (c). Figures 3.4a and 3.4b show spectra recorded with the droplet beam on (〈 N 〉 = 8000) versus blocked by a beam flag, respectively. The y-axis is identical for each of these spectra. The x-axis displays the range 140 – 220 amu, which encompasses peaks that arise from Ir(ppy) 3 ionization. The most prominent peak is split and corresponds to the natural isotopes of iridium at 191 (33%) and 193 amu (67%). Additionally, small peaks at 154 and 205 amu represent ppy + and IrC + , respectively. The difference spectrum, which is enlarged 40 times and shown in Fig. 3.4c, displays a small Ir + peak. Its mag- nitude is only about 1.5% of the total Ir + TOF signal, which is below the noise level of 58 the mass spectra from which it was derived, suggesting that the signal from Ir(ppy) 3 doped helium droplets is absent. In an attempt to observe Ir(ppy) 3 -helium droplet signal, 100 TOF spectra were collected as the photon energy hv was scanned from 35,047 to 35,170 cm -1 . Fig. 3.5a shows Ir + signal intensity versus hv with the droplet beam on, while the background spec- trum in Fig. 3.5b was collected with the droplet beam blocked. The difference spectrum (Fig. 3.5c) shows a flat line. This is not surprising given the lack of signal in Fig. 3.4. 3.5 Discussion Helium droplets capture atoms and molecules with ease, 22 so the absence of signal corresponding to Ir(ppy) 3 doped droplets when 〈 N 〉 = 8000 is interesting. However, this is a unique experimental system because it combines large, energetic molecules with rela- tively small droplets. The results presented in the previous section must be scrutinized in this context. In particular, insight can be gleaned by a thorough examination of the drop- let progression decrease and the limitations of the two detection methods. It will be shown that, because of the large collision energy and the unique ability of helium drop- lets to maintain a temperature of 0.38 K, the capture probability for hot Ir(ppy) 3 is smaller than expected from previous experiments. 3.5.1 QMS Helium Droplet Progression In Fig. 3.3 the helium droplet progression undergoes a marked intensity reduction when Ir(ppy) 3 is introduced. Adding gas to the pick-up cell leads to doped droplets and 59 collisions, each of which can contribute to this effect. In the case of the former, dopant ionization lowers signals from He 2 + (and other helium ions). Alternatively, collisions can lower the intensity of the droplet progression via two processes: (1) collisions scatter droplets out of the beam; (2) collisions with highly internally excited Ir(ppy) 3 cause heli- um evaporation, resulting in a reduced ionization cross section. The following discussion includes an overview of the ionization and charge transfer processes as well as an estima- tion of the collision probability and energy. It is shown that the major contributor to the progression intensity decrease is a lowering of ionization cross sections brought on by collisions with Ir(ppy) 3 . Electron impact ionization of doped helium droplets occurs via three steps. 20,21,23,24 First, a helium atom is ionized within the droplet. The charge then migrates through the droplet, eventually localizing on either the dopant species or another helium atom, forming He n + . Charge transfer to the dopant results in a mass spectrum peak (or peaks) that correspond to the embedded species. For example, dopant peaks are shown for H 2 O on the left side of Fig. 3.3. Localization of the charge on a helium atom leads to the helium progression. A large charge transfer probability (CTP) results in a sharp in- crease in dopant ion peaks and a decrease in the droplet ion peaks relative to the pure droplet spectrum. Therefore, the water peaks observed in Fig. 3.3 could be the cause of the droplet progression decrease. However, the CTP decreases significantly as droplet size increases, and is ≤ 10% for rare gas atoms embedded in droplets with 〈 N 〉 = 3300. 23 Accordingly, for the larger droplets used to record Fig. 3.3 the CTP should be very small 60 and cannot account for the large He n + intensity reduction. † This conclusion is further sup- ported by the observation that as the water peaks decrease, which is a result of the Ir(ppy) 3 sample drying, the reduction in the droplet progression intensity is still observed. The diminished droplet progression intensity, therefore, is caused by collisions between the droplets and Ir(ppy) 3 . Typically, the collision probability is calculated using pressure measurements from the pick-up cell according to Eqs. (2.4) and (2.5). Although the experimental arrangement described in Sect. 3.2 does not allow for direct pressure measurements in the pick-up cell, previous results using argon entrenched helium drop- lets recorded with the same experimental apparatus 11 can be employed to estimate the pressure in the pick-up cell for Ir(ppy) 3 at a given pick-up chamber ion-gauge reading. This enables the collision probability of Ir(ppy) 3 to be calculated for a given set of ob- servable experimental conditions. The argon results, along with a detailed explanation of this process can be found in Appendix B. Figure 3.6 shows the collision probability of helium droplets containing 〈 N 〉 = 8000 traversing a pick-up cell containing Ir(ppy) 3 at 500 K. Plotted on the x-axis is the pick-up chamber ion-gauge reading. Each curve corresponds to the probability of k colli- sions. The highlighted region represents the pressure range used to record the data pre- sented in Sect. 3.4. According to this calculation, when the pressure reading is 1.0 x 10 -7 torr, on average 29% of the droplets will collide with one Ir(ppy) 3 molecule, and another † The CTP for H 2 O is probably larger than that of the rare gas atoms examined (Ne, Ar, and Xe) because H 2 O has a relatively large dipole moment that will attract the positive charge more strongly. In fact, this has been observed for NO, which was found to have a CTP of 15% in N 〈 〉 = 5300 droplets and 2% when N 〈 〉 = 15000. 24 However, the dipole moment for NO is less than 10% of that for H 2 O, 21 and because the available data is limited, the CTP for H 2 O cannot be estimated. 61 6% will undergo two collisions. At a partial pressure of 2.0 x 10 -7 torr, these numbers in- crease to 37% and 16%, respectively. Figure 3.6 Calculated collision probability of k Ir(ppy) 3 molecules as a function of pick-up chamber pres- sure reading. The highlighted region represents the pressure range used to collect the data presented in Sect. 3.4. During any isolated collision, linear momentum is conserved causing trajectories to be altered. For helium droplets, this can cause the droplets to be deflected out of the beam and away from the ionization region, reducing the overall droplet beam intensity. Scattering is largest for inelastic collisions in which the incident angle is 90º with respect to the particle trajectories. This simplified model is presented in Fig. 3.7. Using the root mean squared velocity of Ir(ppy) 3 at 500 K of 138 m / s, and a 8000 atom droplet with a velocity of 350 m / s 15 traveling parallel to the overall droplet beam, the deflection angle 0 10 20 30 40 50 60 70 80 90 100 0.0 0.1 0.2 0.3 0.4 Collision Probability for k Ir(ppy) 3 Pick-up Chamber Ion-Gauge Reading (10 -8 torr) k = 1 k = 2 k = 3 62 will be α = 0.46º. When such a collision occurs in the pick-up cell, which is 75 cm from the ionization region, the droplet will have traveled 0.61 cm normal to its original tra- jectory when it reaches the ionization region. This amount of scattering only slightly hin- ders the droplet intensity in the ionization region. Specifically, under these conditions droplets that form the penumbra will be scattered, while droplets toward the center of the beam will still be detected. Figure 3.7 Schematic showing a collision between a helium droplet and Ir(ppy) 3 in the pick-up cell at an incident angle θ, resulting in a deflection angle α that depends on the initial velocities (Table 3.2). The ioni- zation region is 75 cm from the pick-up cell. The scattering distance at the ionization region is d = 75tan (α). A number of assumptions are inherent in this calculation. Both the lognormal droplet size distribution (§ 2.1) and the Maxwell-Boltzmann Ir(ppy) 3 velocity distribution are ignored; average values are used. Although various droplet sizes and collision veloci- ties will change the result, e.g. see Table 3.2, these effects are negligible when averaged over each distribution. Similarly, there is some nonzero component of the droplet velocity perpendicular to the droplet beam trajectory, which is inconsequential for the same rea- son as stated above. The most significant effect is caused by the assumption that the inci- α pick-up ionization 75 cm 1.4 cm d θ helium Ir(ppy) 3 63 dent angle is 90º. The majority of collisions occur at other angles, all of which result in smaller scattering angles (Table 3.2). Because the result of this calculation is a maximum with respect to the incident angle, the average deflection angle and scattering distance are less than 0.46º and 0.61 cm, respectively. Thus, while some signal is lost due to droplet deflection, most collisions do not lead to significant scattering, and droplet deflection cannot account for the large drop in signal intensity observed in Fig. 3.3. Table 3.2 Deflection angle α and scattering distance d as a function of collision parameters a,b N 〈 〉 (He atoms) m droplet (amu) v droplet (m / s) c 3 ( ) Ir ppy m (amu) 3 ( ) Ir ppy v (m / s) d θ (degrees) α (degrees) d (cm) 5000 20000 350 655 138 90 0.74 0.97 14000 56000 350 655 138 90 0.26 0.35 8000 32000 350 655 138 90 0.46 0.61 8000 32000 350 655 75 90 0.25 0.33 8000 32000 350 655 225 90 0.75 0.99 8000 32000 350 655 138 30 0.23 0.30 8000 32000 350 655 138 140 0.30 0.39 a Refer to Fig. 3.7 b The average experimental conditions are in italics c The droplet beam has a sharp velocity distribution of ∆ v / v < 5% 4 centered around 350 m / s 15 d Root mean square velocity Another consequence of linear momentum conservation is the conversion of ki- netic energy into internal energy during an inelastic collision. 25 The maximum amount of kinetic energy that can be transformed is the relative kinetic energy. 25 For collisions be- tween helium droplets and Ir(ppy) 3 the relative kinetic energy is ( ) 3 2 ( ) 1 2 rel Ir ppy droplet KE v v µ = − (3.6) 64 where µ is the reduced mass, 3 ) ( ppy Ir v is the Ir(ppy) 3 velocity, and droplet v is the droplet ve- locity. The term in parentheses gives the velocity of the incident molecule relative to that of the droplet. This equation is simplified as per Ref. (26). Briefly, the square of the rela- tive velocity is expanded using the cosine theorem and the incident angle θ (Fig. 3.7). The average relative kinetic energy is then obtained by substituting the root mean square velocity for 3 ) ( ppy Ir v and integrating over all angles θ. Finally, because of the relatively large droplet mass, µ is approximately equal to the mass of Ir(ppy) 3 and Eq. (3.6) be- comes 3 ( ) 2 3 2 2 Ir ppy rel droplet m kT KE v 〈 〉 = + (3.7) Eq. (3.7) is the average kinetic energy that is converted to internal energy of the droplet for a completely inelastic collision. The total energy imparted to a droplet in such a col- lision must also include the internal energy of the captured molecule because this energy is efficiently dissipated by the droplet. 4 Specifically, the rovibrational energy rovib U of Ir(ppy) 3 is included because the translational energy term is accounted for by Eq. (3.7). According to Sect. 3.3 rovib vib rot U U U = + (3.8) and rovib U = 11775 cm -1 . Neglecting the relatively small dopant-droplet binding ener- gy, 26 the total available collisional energy is coll rovib rel E U KE 〈 〉= +〈 〉 (3.9) 65 Using Eqs. (3.2) – (3.4), the average energy imparted to a helium droplet during a com- pletely inelastic collision with Ir(ppy) 3 at 500 K is 15600 cm -1 . This energy heats the droplet and induces extensive boiling of helium atoms, cooling the droplet back to its equilibrium temperature of 0.38 K. 4 Because helium atoms are bound to the droplet by approximately 5 cm -1 , 15600 cm -1 corresponds to the evaporation of 3120 helium atoms. The collisions described above decrease 〈 N 〉 from 8000 to 4880. According to Eqs. (2.1) and (2.2), this reduces the corresponding average cross sectional area tot σ from 6200 to 4450 Å 2 . However, this 30% reduction in droplet cross section is a low estimate because the U vib value calculated in Sect. 3.3 and subsequently used in Eq. (3.3) is less than the actual value. Because the electron impact ionization cross section is proportional to the cross sectional area of the droplet, after a collision the ion signal intensity de- creases. In fact, this effect is the working principle of helium droplet depletion spectro- scopy. 4 For example, a similar reduction in droplet size enabled the depletion spectrum of NO 2 in helium droplets to be collected. 27 With the experimental conditions used to record Fig. 3.3, roughly 30 – 40% of the droplets undergo a single collision. This number is increased by 2 – 5% when the colli- sion probability in the pick-up chamber is taken into account. † If 35% of the droplets undergo a single collision and, as a consequence, experience a 30% decrease in size, then † The total collision probability in the pick-up chamber is coll z nL σ γ = . 25,26 Here / B n P k T = is the num- ber density of particles in the pick-up chamber, where P is the corresponding pressure; the helium droplet collision cross section coll σ is equal to the cross sectional area of the droplet; L is the length of the pick- up chamber (17.78 cm); and γ accounts for the velocity dependence of the number of collisions, which is approximated by / rel droplet v v where rel v is the average relative velocity and droplet v is the droplet beam velocity. 26 66 the total ion signal is reduced by 11%. Furthermore, according to Eq. (2.4) and Fig. 3.6, about 10 – 20% of droplets suffer multiple collisions. Because multiple collisions make droplets more susceptible to scattering and further decrease the ionization efficiency, it is assumed that these droplets do not contribute significantly to the ionization signal. There- fore, collisions account for a 20 – 30% decrease in the droplet progression intensity. The foremost source of this signal reduction is a lower ionization probability caused by helium evaporation subsequent to a collision. 3.5.2 Capture Cross Section for Ir(ppy) 3 Despite an abundance of collisions in the pick-up cell, there is no signal corre- sponding to Ir(ppy) 3 embedded droplets in either the quadrupole or TOF spectra when 〈 N 〉 = 8000 droplets are used. For helium droplets, the capture cross section and colli- sion cross section are usually of similar magnitude, suggesting that Ir(ppy) 3 is captured by the droplets but not observed. On the contrary, a close examination of the detection sensitivity and limitations of the quadrupole and TOF methods reveal that this unique ex- perimental system has a low capture probability. The maximum mass range of the QMS is 511 amu, which includes the major Ir(ppy) 3 fragment ions: Ir(ppy) 2 + , Ir(ppy) + , Ir + , and ppy + . However, since the mass of Ir(ppy) 3 + is around 655 amu, there is a possibility that the parent ion is created but not de- tected. Direct electron impact ionization of gas-phase Ir(ppy) 3 produces ion signals at 655, 500, and 155 amu, which are attributed to Ir(ppy) 3 , Ir(ppy) 2 + , and ppy + , respect- tively. 28 However, because molecules in helium droplets ionize via charge transfer from 67 He + , it cannot be assumed that the fragmentation products will be the same as those in the gas-phase. In some instances helium droplets have exhibited a caging effect that reduces dopant fragmentation caused by electron impact. 22 Although not well understood, studies show that this effect is strong for some clusters of small molecules 24,26 and rare gas atoms. 20,21 On the other hand, some large molecules, such as the amino acids tryptophan and tyrosine, display fragmentation patterns that are largely unaffected by the sur- rounding helium. 7 Since Ir(ppy) 3 is larger than tryptophan and tyrosine, these results sug- gest that Ir(ppy) 3 will exhibit fragmentation when ionized within a helium droplet. This conclusion is further supported by the small ppy + peak observed in the mass spectrum when droplets with a 〈 N 〉 value of 12000 are employed. If this is the case, then the ab- sence of signal in the quadrupole mass spectrum can only be explained by a very low Ir(ppy) 3 capture probability. However, because the available data is limited, the possi- bility that the major product of Ir(ppy) 3 ionization is the parent ion and thus out of the mass range, cannot be completely ruled out. There is no discernible signal corresponding to Ir(ppy) 3 doped droplets in the TOF mass spectrum displayed in Fig. 3.4. The sensitivity of this method is limited by the large gas-phase Ir(ppy) 3 signal. When the pick-up cell is hot, the partial pressure of Ir(ppy) 3 in the detection chamber is 5 x 10 -9 torr, which translates to a molecular density of 9.7 x 10 7 cm -3 (Eq. (2.5)). The helium droplet beam intensity is 7 x 10 15 (sr s) -1 , 14 with a corre- sponding droplet density 35 cm from the source of 1.6 x 10 8 cm -3 . For the doped droplet signal to be observable, it should be no less than 10% of the gas-phase signal. Using a 68 high photon fluence that ensures the ionization efficiency is not adversely affected by the surrounding helium, this occurs when 6% of the droplets capture a single Ir(ppy) 3 mole- cule. The capture cross section required for 6% capture in the pressure range highlighted in Fig. 3.6, is determined by replacing coll σ with capt σ in Eq. (B.5) and setting 06 . 0 ) ( 1 = capt I σ . This calculation reveals that a capture cross section of 10% capt coll σ σ ≈ is sufficient for Ir(ppy) 3 -droplet signal to be observed. Therefore, the absence of TOF signal leads to the unambiguous conclusion that the capture cross section is less than 10% of the collision cross section. Previous studies have shown that capture cross sections lie in the range 1/ 2 coll capt coll σ σ σ ≤ ≤ , where coll σ is equivalent to the geometrical cross section given by Eq. (2.2). The results for Ir(ppy) 3 are thus an anomaly and a testament to the uniqueness of the system. The relatively small droplets used in these experiments are inadequate for efficient capture of large, energetic molecules. Consider the collision between Ir(ppy) 3 and a helium droplet depicted in Figs. 3.7 and 3.8. The highly excited Ir(ppy) 3 molecule approaches the droplet with an average rel- ative velocity of 3.76 Å / ps. Assuming a sharp cutoff at the droplet edge, the average im- pact parameter is 2 / 3 b R 〈 〉= , where R is the droplet radius given by Eq. (2.1). Thus, on average, collisions occur on the droplet periphery. After the initial collision, the 14.76 Å long Ir(ppy) 3 (Fig. 1.3), like all closed shell molecules, 4 enters the droplet. The droplet acts as a heat sink and begins to cool the incoming Ir(ppy) 3 on a picosecond timescale. 27 Therefore, in the immediate vicinity of the collision, the droplet becomes excited and helium begins to boil off before Ir(ppy) 3 is fully immersed in the droplet. Because of the 69 local helium evaporation and the large average impact parameter, the effective droplet diameter for such a collision is much smaller than 2R. Evidently, the effective diameter is too small to dissipate all of the collision energy and trap the molecule. In other words, while traversing through the droplet, Ir(ppy) 3 imparts a significant amount of energy and momentum to the surrounding helium, retaining just enough to exit the droplet. Addi- tional evidence to support this picture is the observed decrease in the droplet progression intensity, which indicates that despite the low probability for capture, Ir(ppy) 3 collisions impart a large amount of energy and momentum to the droplets. Figure 3.8 The average impact parameter, b 〈 〉 = 2/3R, is shown for a collision between a helium droplet of mean size N 〈 〉 = 8000 atoms and Ir(ppy) 3 . The diameters of the droplet and Ir(ppy) 3 —about 92 and 15 Å, respectively—and b 〈 〉 are approximately drawn to scale. 3.6 Future Directions These experiments were motivated by a desire to study cold Ir(ppy) 3 in hopes of acquiring high resolution spectroscopic data. Although this goal has yet to be realized, an b 〈 〉 Ir(ppy) 3 Helium Droplet 70 important step in this pursuit has been made. Large droplets, with 〈 N 〉 values of 12000 or more, are necessary to capture hot Ir(ppy) 3 . Creating such droplets requires stable noz- zle temperatures of 12 K or less (§ 2.1) and, hence, a reengineered nozzle assembly. Once formed however, Ir(ppy) 3 doped droplets can be probed by a variety of tech- niques. Photoionization with TOF mass analysis is a viable candidate, provided that the gas-phase signal can be kept to a minimum. Depletion spectroscopy is less desirable in this droplet size range because the depletion percent decreases with increasing droplet size and is difficult to detect. However, if a quadrupole mass spectrometer is to be used, the mass range should include at least the mass of the parent ion. Ideally, the range would extend even further because ions are sometimes formed with helium atoms attached. 20 A more common method of investigating molecules embedded in large helium droplets is laser induced fluorescence (LIF). In the case of Ir(ppy) 3 , this may be the best technique. This work also raises fundamental questions concerning the helium droplet pick- up process. For a given droplet-atom or droplet-molecule collision, an unknown set of parameters exists with which the incident particle will pass through the droplet. A syste- matic study of these parameters would be interesting. For example, an experiment could be set up to determine the capture cross section as a function of mean droplet size, inci- dent species, collision energy, and impact parameter. Such a comprehensive investigation could confirm the results of the research described herein as well as add to the basic understanding of helium droplet capture. 71 3.7 Conclusion New insight to the helium droplet pick-up process was gained by using Ir(ppy) 3 sublimed at 500K as scattering gas. It was demonstrated that, while large droplets with a 〈 N 〉 value of 12000 can capture Ir(ppy) 3 , smaller 〈 N 〉 = 8000 droplets do not efficiently trap the excited molecule. Vibrationally excited Ir(ppy) 3 evaporates the local helium liq- uid leading to a dense helium gas that cannot dissipate the Ir(ppy) 3 kinetic energy; the large molecule passes through the gas with very high efficiency. Therefore, the capture cross section generally depends on the droplet size and collision energy, although these dependencies are not completely understood. Because the pick-up process is often adver- tised as straightforward, these results have implications far beyond the scope of this work. Nevertheless, if large enough droplets are used to ensure efficient capture, the helium droplet isolation spectroscopy remains a promising technique for studying large, energetic OLED molecules. 72 3.8 References 1. Andrews, L.; Moskovitis, M. Chemistry and Physics of Matrix-Isolated Species; North Holland, Amsterdam, 1989; pg. 430. 2. Stienkemeier, F.; Vilesov, A. F. J. Chem. Phys. 2001, 115, 10119. 3. Wewer, M.; Stienkemeier, F. Phys. Rev. B 2003, 67, 125201. 4. Toennies, J. P.; Vilesov, A. F. Angew. Chem. Int. Ed. 2004, 43, 2622. 5. Bulovic, V.; Burrows, P. E.; Forrest, S. R., Cronin, J. A.; Thompson, M. E. Chem. Phys. 1996, 210, 1. 6. Akers, K.; Aroca, R.; Hor, A.; Loutfy, R. O. J. Phys. Chem. 1987, 91, 2954. 7. Lindinger, A.; Toennies, J. P.; Vilesov, A. F. J. Chem. Phys. 1999, 110, 1429. 8. Close, J. D.; Federmann, F.; Hoffmann, K.; Quaas, N. Chem. Phys. Lett. 1997, 276, 393. 9. Hartmann, M.; Lindinger, A.; Toennies, J. P.; Vilesov, A. F. Phys. Chem. Chem. Phys. 2002, 4, 4839. 10. Barranco, M.; Guardiola, R.; Hernandez, S.; Mayol, R.; Navarro, J.; Pi, M. J. Low Temp. Phys. 2006, 142, 1. 11. Zadorozhnyy, A. Energy Transfer Pathways for NO 2 -Rare Gas Complexes in Helium Droplets. Masters Thesis, University of Southern California, Los Angles, CA, 2009. 12. Polyakova, E. Multiphoton Ionization of Molecules Embedded in Superfluid Liquid Helium Droplets. Ph.D. Dissertation, University of Southern California, Los Angeles, CA, 2005. 13. Harms, J.; Toennies, J. P.; Dalfovo, F. Phys. Rev. B 1998, 58, 3341. 14. Slipchenko, M. N.; Kuma, S.; Momose, T.; Vilesov, A. F. Rev. Sci. Instrum. 2002, 73, 1. 15. Buchenau, H.; Knuth, E. L.; Northby, J.; Toennies, J. P.; Winkler, C. J. Chem. Phys. 1990, 92, 6875. 16. Levine, I. N. Physical Chemistry, 5 th ed; McGraw-Hill: New York, NY, 2002. 73 17. Atkins, P. W.; Friedman, R. S. Molecular Quantum Mechanics, 3 rd ed; Oxford University Press: New York, NY, 1997. 18. Sett, P.; Chattopadhyay, S.; Mallick, P. K. Spectro. Acta A 2000, 56, 855. 19. Isaq, M.; Gupta, S. P.; Sharma, S. D.; Yadav, B. S. Orient. J. Chem. 1998, 14, 387. 20. Callicoat, B. E.; Forde, K.; Ruchti, T.; Jung, L.; Janda, K. C.; Halberstadt, N. J. Chem. Phys. 1998, 108, 9371. 21. Ruchti, T.; Forde, K.; Callicoat, B. E.; Ludwigs, H.; Janda, K. C. J. Chem. Phys. 1998, 109, 10679. 22. Toennies, J. P.; Vilesov, A. F. Annu. Rev. Phys. Chem. 1998, 49, 1. 23. Ruchti, T.; Callicoat, B. E.; Janda, K. C. Phys. Chem. Chem. Phys. 2000, 2, 4075. 24. Callicoat, B. E.; Mar, D. D.; Apkarian, V. A.; Janda, K. C. J. Chem. Phys. 1996, 105, 7872. 25. McDaniel, E. W. Atomic Collisions: Electron and Photon Projectiles; John Wiley & Sons, Inc.: New York, NY, 1989. 26. Lewerenz, M.; Schilling, B.; Toennies, J. P. J. Chem. Phys. 1995, 102, 8191. 27. Polyakova, E.; Stolyarov, D.; Zhang, X.; Kresin, V. V.; Wittig, C. Chem. Phys. Lett. 2003, 375, 253. 28. King, K. A.; Spellane, P. J.; Watts, R. J. J. Am. Chem. Soc. 1985, 107, 1431. 74 Chapter 4 Multiphoton Ionization of Tris(2-phenylpyridine)iridium 4.1 Introduction Electroluminescent diodes consisting of organic materials, referred to as organic light emitting diodes (OLEDs), have been the topic of intense research in recent years, owing to their application in light displays. 1,2 These devices consist of organic com- pounds sandwiched between two electrodes (§ 1.1), which inject electrons and holes into the organic layer. 1,2 Ideally, charge recombination leads to an electronically excited dopant molecule that exhibits efficient luminescence. Careful choice of emitter molecules is imperative. Electron-hole recombination produces both singlet and triplet excited states. In general, organic molecules display fluorescence, while phosphorescence is suppressed due to the long lifetime of the triplet state. Consequently, the use of pure organic compounds as emitter molecules severely limits the electroluminescent quantum efficiency. 1,2 On the other hand, because of strong spin-orbit coupling (SOC), organometallic compounds—specifically, transition metal complexes with organic ligands—can exhibit quantum efficiencies of nearly 100%. 1,3 This group of molecules, therefore, has garnered significant attention as phosphor dopants in OLEDs. 4,5 Perhaps the most well known of these complexes is tris(2-phenylpyridine)iridium (Ir(ppy)3). When incorporated in OLEDs, several unique properties of this compound and its close counterparts result in very efficient electroluminescence. For instance, 75 Ir(ppy) 3 possesses a large cross-section for charge recombination, 6 which increases the probability of useful exciton formation. Furthermore, the iridium atom introduces strong SOC, facilitating fast intersystem crossing (ISC) and high phosphorescent efficiency at ambient temperatures. 7,8 Highly efficient OLEDs based on Ir(ppy) 3 are responsible for the scientific com- munity’s great interest in this compound. A thorough understanding of the photochem- istry requires knowledge of the molecular orbital properties, spectra, and excited states. Moreover, the promise of this particular organometallic molecule goes beyond its appli- cation as a green (514 nm) 6 phosphor dopant. A complete understanding of the electronic properties of this benchmark molecule may assist the design of other compounds with different emission wavelengths. 9,10 Gas-phase multiphoton ionization is an excellent technique for gathering infor- mation on isolated molecules (§ 2.2). The results presented herein were obtained by ap- plying this method to effusive Ir(ppy) 3 . Although two-photon excitation leads primarily to molecular dissociation, it is possible to isolate and characterize the parent ion. This experiment, which is complemented by a comprehensive density functional theory (DFT) investigation, 11 led to a determination of the upper limit of the ionization potential and sets the framework for more detailed experiments. 4.2 Experimental Apparatus The experimental setup used for the photoionization of Ir(ppy) 3 is similar to that presented in Sect. 3.2. Minor adjustments were made to both the vacuum chamber and 76 the laser system. In addition, the time-of-flight (TOF) mass spectrometer was calibrated and its settings tuned for maximum resolution. The work described in this chapter required two slight modifications to the helium droplet machine presented in Sect. 3.2.1. First, the pick-up cell was converted to an effu- sive source. This simple alteration was accomplished by blocking the round hole that al- lowed the helium droplet beam to enter the pick-up cell. Effusive Ir(ppy) 3 was obtained by heating the source to 500 K. Second, the TOF mass spectrometer was moved to the pick-up chamber, enabling the experiments to be carried out in chamber II. 4.2.1 Laser and Optical Setup The laser system used for this work is briefly described in Sect. 3.2.2. A number of modifications were made to the setup, including the addition of an attenuator, removal of the focusing lens, and a careful choice of laser dyes. In addition, the laser wavelength was calibrated using a wavemeter. An attenuator consisting of a half-waveplate and a Glan-Thompson polarizer was used to continuously vary the laser fluence. 12 Once the polarizer is set for maximum throughput, the waveplate is inserted between the light source and the polarizer. Turning the waveplate rotates the polarization of the laser beam. Because the polarizer transmits only a specific component of the polarized light, the fluence is easily adjusted. Very low fluence measurements, i.e. less than 1.5 mJ / cm 2 , required removing the focusing lens since focusing limits the lowest fluence that can be obtained. In lieu of fo- cusing, the beam was physically skimmed with a 1 mm diameter aperture. With this 77 method, the ionization area is slightly increased relative to the focused beam, but a much smaller minimum fluence is achievable. The Ir(ppy) 3 + action spectrum was collected using a variety of laser dye mixtures. The Ir(ppy) 3 + signal intensity depends on the second order of the laser fluence, so meas- ures were taken to ensure that the fluence did not affect the action spectrum. To this end, the action spectrum is comprised of seven separate spectra recorded using a variety of rhodamine 590 and rhodamine 610 laser dye mixtures. The ratios of the dyes and the corresponding wavelengths are listed in Table 4.1. Moreover, the laser power was moni- tored and recorded along with the ion signal intensity. In this way, the fluence was kept within 5% of 450 µJ / cm 2 , and the final spectrum was corrected using the simultaneously recorded power data. Table 4.1 Rhodamine 590 and 610 dye mixtures a Rho 590 (mg) Rho 610 (mg) λ (nm) 140 0 279.5 – 283.1 90 10 281.5 – 284.3 80 20 283.2 – 286.8 70 30 286.2 – 288.1 60 40 287.8 – 290.5 0 107 290.2 – 292.8 a The listed amounts were mixed and diluted to 1 L with methanol. The solutions were further diluted in the os- cillator and amplifier to improve the beam profile and free- running, respectively. Exciton dyes were used. 4.2.2 Time-of-Flight Mass Spectrometer The TOF mass spectrometer is shown schematically in Fig. 2.7. The spectrometer was calibrated by a series of gas-phase NO measurements. The dimensions determined 78 from this process are listed in Table 4.2. The following section focuses on the calculated and measured instrument resolution. The working principle, along with general resolution considerations and equations are discussed in detail in Sect. 2.3. Table 4.2 TOF spectrometer dimensions and parameters D (cm) d (cm) s (cm) ∆ s (cm) V ext (kV) V rep (kV) 51.28 0.7 0.62 0.1 2.36 2.00 The resolution is determined from a combination of the initial kinetic-energy and spatial distributions of the molecules in the ionization region. The space distribution is a result of the cross-sectional area of the ionizing laser. Therefore, the increased fluence range—i.e. skimming instead of focusing the laser—comes as a sacrifice to the space resolution. Fortunately, the space resolution can be improved by the focusing condition given by Eq. (2.19). This condition is satisfied by applying voltages of 2.36 and 2.00 kV to the repeller plate and extraction grid, respectively. Because the spectrometer employs constant electric fields to accelerate the ions toward the detector, the energy resolution becomes the limiting factor. † The theoretical resolution range calculated from Eqs. (2.20) and (2.22) – (2.24) is 136 – 142 amu. ‡ Figure 4.1 shows the measured Ir(ppy) 3 + TOF signal (solid black line). The fit is composed of four Gaussian curves that represent the four major Ir(ppy) 3 isotopes listed in † As discussed in Sect. 2.3, improving the energy resolution, which is defined as the mass resolution corre- sponding to the energy distribution of the neutral molecules prior to ionization, requires a pulsed electric field. ‡ By definition, this is the largest mass at which two ion signals spaced by 1 amu will be completely iso- lated. Technically, ion peaks at twice this mass are resolvable. This is explained further by Fig. 2.8 and the supporting text. 79 Table 4.3. The relative intensity of each curve is determined by the isotopic ratios. The fit is optimized when the individual curves have a full width at half maximum (FWHM) of 2 amu. From this fit, the experimental resolution is determined to be 130 amu, which agrees well with the calculated range. Table 4.3 Major Ir(ppy) 3 isotopes formula 13 CC 32 H 24 193 IrN 3 C 33 H 24 193 IrN 3 13 CC 32 H 24 191 IrN 3 C 33 H 24 191 IrN 3 mass (amu) 656 655 654 653 percent 21.1% 41.5% 12.6% 24.6% 645 650 655 660 665 mass / amu Figure 4.1 Time-of-flight mass spectrum showing the Ir(ppy) 3 + ion peak (solid black line), recorded at 35,673 cm -1 and 415 µJ / cm 2 . The best fit trace (♦) results from the addition of four Gaussian curves repre- senting the predominant isotopes of Ir(ppy) 3 : 656 amu (◊); 655 amu (□); 654 amu (∆ ); and 653 amu (○). Each Gaussian curve has a FWHM of 2 amu; the ensuing fit has a FWHM of 3.6 amu. The mass resolution is 260 amu. 80 4.3 Computational Study of Ir(ppy) 3 Ir(ppy) 3 was examined with density functional theory (DFT) with the ωB97X and BNL functionals using the Q-Chem program. 14 The Lan12DZ basis set was used for irid- ium, while the 6-311++G** basis set was employed for the other atoms (ligands). Excited states were calculated using time dependent density functional theory (TDDFT). A full description of the methods and results will be published separately. 11 This section focuses on the results that are pertinent to the current work. In par- ticular, the optimized geometries, frontier molecular orbitals, excited states, and ioni- zation energy are all presented. The results are compared with experimental and other theoretical studies when available. Figure 4.2 Picture of fac-Ir(ppy) 3 with hydrogen atoms omitted for clarity. Atoms labeled for Table 4.4. Adapted from Ref. (11). 4.3.1 Structure and Molecular Orbitals Ir(ppy) 3 can have fac and mer isomers. However, mer-Ir(ppy) 3 is less stable than the fac isomer by approximately 6 kcal / mol. 11,14,15 Furthermore, it is likely that mer- Ir(ppy) 3 is converted to fac-Ir(ppy) 3 upon annealing. 16 Therefore, the following discus- 81 sion treats fac-Ir(ppy) 3 (Fig. 4.2) exclusively. In addition, Ir(ppy) 3 will replace the rather cumbersome fac-Ir(ppy) 3 notation. Table 4.4 Bond lengths (Å) for the ground state (S 0 ), lowest triplet (T 1 ), and lowest cation states for Ir(ppy) 3 a S 0 T 1 Cation Calc. 11 Calc. 17 Calc. 14 Exp. 18 Calc. 11 Calc. 17 Calc. 11 Ir–N2 b 2.168 2.153 2.167 2.132 2.196 2.176 2.149 Ir–N29 2.169 2.154 2.167 2.132 2.175 2.169 2.184 Ir–N42 2.165 2.151 2.167 2.132 2.139 2.116 2.213 Ir–C9 2.025 2.035 2.035 2.024 2.025 2.030 2.028 Ir–C22 2.023 2.035 2.035 2.024 2.036 2.048 1.975 Ir–C49 2.022 2.035 2.035 2.024 1.961 2.000 2.035 a Adapted from Ref. (11) b See Fig. 4.2 for atom labels The molecular structure of Ir(ppy) 3 was optimized at the singlet ground state (S 0 ), lowest triplet state (T 1 ), and lowest cation state using DFT / ωB97X without symmetry constraints. Select geometrical parameters are compared with previous theoretical and experimental values in Table 4.4. The S 0 structure is very close to C 3 symmetry. With ωB97X, the Ir–C bond length coincides well with the measured value, while the Ir–N bond length is similar to the B3LYP result, which is about 0.03 Å longer than the experimental value. Both the T 1 and cation states possess C 1 symmetry. While no reported geometrical parameters for the cation state were found, the T 1 structure agrees qualitatively with previous results; 17 rel- ative to the S 0 geometry, one ligand moves closer to the metal center as the other two are pushed away. 82 These structural changes can be understood by examining the molecular orbitals (MOs) computed with the BNL basis and visualized using the Molden interface. As shown in Fig. 4.3a, the highest occupied molecular orbital (HOMO) is comprised prima- rily of the iridium 5d orbital. In contrast, the dominant contribution to the lowest unoccu- pied molecular orbital (LUMO) are the ligand π*-orbitals (Fig. 4.3b). The T 1 state can be approximately described as a one-electron transition from the HOMO to LUMO, in which case it is primarily of metal-to-ligand charge transfer (MLCT) character. After re- laxation at the T 1 state, the electron distribution localizes on one ligand (Figs. 4.3c and 4.3d), causing the observed geometrical change. Figure 4.3 Ir(ppy) 3 molecular orbitals: (a) HOMO and (b) LUMO at the S 0 geometry; (c) HOMO and (d) LUMO at the T 1 geometry. Adapted from Ref. (11). (c) (a) (b) (d) S 0 T 1 83 A more detailed examination of the frontier orbitals provides a qualitative frame- work for the excited state calculations discussed below. A schematic of the six highest occupied and three lowest unoccupied orbitals is displayed in Fig. 4.4. The MOs are labeled according to their dominant atomic orbital contribution. For example, the HOMO and LUMO described above are labeled d 2 and π 1 *, respectively. However, such labels are somewhat misleading because the MOs consist of a mixture of metal and ligand atomic orbitals. Nevertheless, these results provide a qualitative understanding of the excited states. Specifically, the low-lying excited states are primarily metal-to-ligand charge-transfer (MLCT) states, while the higher states are ligand-centered (LC) states. Figure 4.4 Ir(ppy) 3 frontier orbitals at the S 0 geometry: (a) six highest occupied MOs and (b) three lowest virtual MOs. Energies are not to scale. For orbital labeling, see text. Adapted from Ref. (11). d 2 d 1a d 1b π 1 π 2a π 2b π 1 * π 2a * π 2b * Orbital Energy Orbital Energy (a) (b) 84 4.3.2 Excited States 260 excited states (130 singlet and 130 triplet states) were calculated at the opti- mized S 0 geometry using TDDFT / BNL. An additional 20 excited states were calculated at the optimized T 1 geometry. Table 4.5 lists the vertical excitation energies (calculated at the S 0 geometry), oscillator strengths, and transition character of several selected excited states. The character is assigned based on the dominant orbital transition. Table 4.5 in- cludes the five lowest-lying states and three states with large oscillator strengths within the experimental energy range. As described above, the low-lying states are MLCT in na- ture, while the higher energy states are of LC character. The spin-orbit coupling (SOC) in Ir(ppy) 3 is particularly strong, which has far- reaching effects. For example, SOC efficiently mixes the singlet and triplet state mani- folds, causing states to shift by as much as 0.2 – 0.3 eV. 14 More importantly, this mixing provides oscillator strength to the triplet states, resulting in robust phosphorescence. However, because the TDDFT calculations do not include SOC, the resulting triplet states listed in Table 4.5 all have zero oscillator strength. Table 4.5 Selected excited states (S 0 geometry) state E (eV) f Character T 1 2.56 – MLCT T 2 2.62 – MLCT T 3 2.62 – MLCT S 1 2.69 0.0039 MLCT S 2 2.80 0.0056 MLCT S 55 4.36 0.1031 LC S 58 4.38 0.1259 LC S 68 4.50 0.1150 LC 85 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Calculated Oscillator Strength photon energy / eV Figure 4.5 Calculated oscillator strengths (x 3) of 130 singlet states vs. photon energy (blue lines). Absorp- tion spectrum (black line) is generated by applying a uniform Gaussian width to each state. Adapted from Ref. (11). Despite the exclusion of SOC effects, the TDDFT results provide a reasonable description of the excitations. The 260 states calculated at the S 0 geometry correspond to vertical excitation from the ground state. Therefore, these results can be directly com- pared with the absorption spectrum. In Fig. 4.5, the calculated oscillator strengths of 130 singlet states are plotted as a function of energy. The absorption spectrum is simulated by applying a Gaussian width to each state. The calculated and measured absorption spectra are compared in Fig. 4.6. The in- set shows the absorption spectrum of 2-phenylpyridine (ppy), 19 which bears a striking re- semblance to the Ir(ppy) 3 spectrum. This supports the assignment that the high-lying states result from LC transitions. The calculations reproduce the experimental spectrum 86 well. As mentioned above, the large SOC provides oscillator strength to the triplet states and shifts the excited state energies. Owing to the omission of the SOC in the calculated spectrum, the low energy absorption band that has been assigned to the lowest 3 MLCT state 20 is absent in the computed spectrum. Interestingly, the shifts are rather inconse- quential. Moreover, the absorption spectrum is replicated without the inclusion of Franck-Condon factors. This is due to the high state density and the lack of significant geometrical changes upon excitation. Figure 4.6 Comparison of the calculated absorption spectrum (black line) taken from Fig. 4.5 and meas- ured spectrum (red line) recorded in dichloromethane. Inset: UV-Vis spectrum of 2-phenylpyridine, adapted from Ref. (19). 4.0 0 4.6 0 5.4 0 Photon energy / eV ppy 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 2.5 Calculated Oscillator Strength photon energy / eV Absorbance 87 4.3.3 Ionization Energies The energy of the lowest cation state was calculated at the S 0 , T 1 , and cation ge- ometries using DFT / BNL. These energies are displayed in Fig. 4.7. The energy calcu- lated at the optimized cation geometry (5.86 eV) corresponds to the adiabatic ionization energy (IE). The vertical IEs are 5.87 and 3.58 eV from the S 0 and T 1 states, respectively. The excitations of Ir(ppy) 3 are summarized by the correlation diagram shown in Fig. 4.7. The optimized S 0 and T 1 geometries are shown on the x-axis. All state energies are relative to S 0 ; the dashed line corresponds to the adiabatic IE. Although Ir(ppy) 3 has a very large state density, the diagram is limited to several low-lying states and the cation energies. Figure 4.7 Several TDDFT / BNL excited states calculated at the S 0 and T 1 geometries. High energy states were only calculated at the S 0 geometry. The adiabatic ionization energy (AIE) is represented by a horizon- tal dashed line. Vertical ionization energies are shown as states above the AIE. All energies are relative to the S 0 state. Adapted from Ref. (11). S 0 T 1 Singlet Geometry Triplet Geometry Energy above Singlet Ground State S 0 (eV) 5.87 5.86 (IE) 5.88 2.56 T 2 2.62 T 3 2.62 S 1 2.69 S 55 4.36 T 1 2.30 S 1 2.62 T 2 2.74 T 3 2.78 S 0 0.17 S 68 4.50 88 Figure 4.8 Absorption spectrum of Ir(ppy) 3 in dimethylformamide at 414 K. The highlighted region, 34,150 – 35,775 cm -1 , corresponds to the 1 π-π* band examined in this work. Inset: increasing temperature has no significant effect on the absorption spectrum in the energy range of interest. In order from top trace the temperatures are: 295, 333, 353, 383, 408, and 414 K. 4.4 Results The UV - Vis absorption spectrum of Ir(ppy) 3 in dimethylformamide (DMF) at 414 K is shown in Fig. 4.8. As discussed below, Ir(ppy) 3 was examined via multistep ion- ization in the energy range hv = 34,150 – 35,775 cm -1 . This spectral region corresponds to the π-π* band, and is highlighted in Fig. 4.8. DMF was used because its high boiling point (426 K) allows the temperature dependence of the spectrum to be investigated. The inset shows six offset spectra recorded between 295 and 414 K (from top to bottom) in the frequency range defined by the highlighted region. These spectra are nearly identical; as the temperature is increased, very slight broadening is observed. This is expected in light of the calculated BNL spectrum presented in the last section (Figs. 4.5 and 4.6), 20000 22500 25000 27500 30000 32500 35000 0 1 2 3 Absorbance wavenumber / cm -1 3400 3500 3600 89 which shows that the structure is primarily due to the high state density, and that Franck- Condon factors are of little significance. Figure 4.9 TOF spectrum of Ir(ppy) 3 collected at 35,774cm -1 with a fluence of 1.5 J / cm 2 . The most intense peak corresponds to Ir + isotopes at 191 and 193 amu; IrC + is seen at 205 amu, and lighter peaks are attrib- uted to hydrocarbons. There are no significant peaks at higher mass. Effusive Ir(ppy) 3 was detected via photoionization time-of-flight (TOF) mass spectrometry. Figure 4.9 shows a TOF spectrum recorded with a single laser pulse fo- cused to a diameter of approximately 0.7 mm with a 60 cm lens. Relatively large energy fluence (1.5 J / cm 2 ) was used to ensure favorable ionization conditions. The TOF spec- trum shows two large peaks centered at 192 and 205 amu, corresponding to Ir + and IrC + , respectively. The isotopic structure of these peaks agrees well with the natural abundance of iridium isotopes: 33% 191 Ir and 67% 193 Ir. A cluster of ion peaks below 80 amu is 0 50 100 150 200 250 mass / amu hydrocarbons IrC + Ir + 90 attributed to organic contaminants; these peaks are present even in the absence of Ir(ppy) 3 . The Ir + signal intensity was monitored as a function of photon energy in the range 34,924 – 35,801 cm -1 . The effect of laser fluence on the Ir + action spectrum is illustrated in Fig. 4.10. The bottom plot was recorded with an energy fluence of 1.3 J / cm 2 and con- sists of a mostly structureless continuum superimposed with numerous broad features. Decreasing the fluence by a factor of five (middle panel) significantly reduces the contin- uum absorption. For the top panel, the fluence is decreased to 130 mJ / cm 2 . Here, the spectrum is the least congested, consisting of sharp peaks and a very weak Ir + back- ground. Details of the low fluence spectrum are shown in Appendix C. Figure 4.10 Ir+ signal intensity vs. photon energy recorded with three different fluences. 130 mJ / cm 2 260 mJ / cm 2 1300 mJ / cm 2 35000 35200 35400 35600 0 5 10 15 0 2 4 6 0 4 8 photon energy / cm -1 91 Figure 4.11 shows a portion of the spectrum shown in the top panel of Fig. 4.10 overlaid with a stick spectrum of the known atomic lines of iridium. 21 A complete list of the observed transitions and corresponding atomic assignments is presented in Appendix C. Clearly the sharp features observed in the Ir + action spectra are due to resonance en- hanced multistep ionization (REMPI) of atomic iridium. Therefore, Ir(ppy) 3 must under- go photolysis when excited with UV radiation. These results are discussed at length in Sect. 4.5.4. 35350 35400 35450 35500 35550 35600 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Ir + ion intensity photon energy / cm -1 Figure 4.11 Expanded section of the plot shown in the top panel of Fig. 4.10 (black line), along with a stick spectrum of the atomic iridium transitions (vertical red lines). 21 A further reduction in the laser fluence drastically affects the TOF signal. Figure 4.12 shows five TOF spectra recorded with a broad range of laser fluences. The laser beam was attenuated with a half-lambda waveplate in series with a Glan-Thompson 92 polarizer. The highest fluence spectrum (top plot) is similar to the spectrum presented in Fig. 4.9, consisting mostly of a large Ir + peak. Signal corresponding to Ir(ppy) 2 + and Ir(ppy) 3 + gain intensity as the fluence is decreased. At 88 mJ / cm 2 , the intensities of the Ir + and Ir(ppy) 3 + peaks are nearly identical, and a IrC n + / IrN n + progression emerges. The parent Ir(ppy) 3 + ion is isolated with a fluence of 1.6 mJ / cm 2 or less. Interestingly, only a very weak signal corresponding to the ppy + ion is seen at large fluence (Fig. 3.4) and no signal due to the Ir(ppy) + ion is observed. Figure 4.12 TOF spectra of Ir(ppy) 3 collected at hv = 35,357 cm -1 and various fluence. At low fluence (bottom trace), ionization produces the Ir(ppy) 3 + ion exclusively. As the fluence is increased, fragmentation results in lighter ions being formed including Ir(ppy) 2 + , and eventually Ir + . There are at least two mechanisms responsible for the observed TOF ion signals. As discussed above, the Ir + ion is—at least in part—due to the photodissociation of 200 300 400 500 600 700 800 mass / amu Ir + Ir(ppy) 2 + Ir(ppy) 3 + 1.6 mJ / cm 2 2.0 mJ / cm 2 88 mJ / cm 2 300 mJ / cm 2 1400 mJ / cm 2 93 Ir(ppy) 3 . On the other hand, the parent Ir(ppy) 3 + ion must be the result of multistep ion- ization. Therefore, according to Fig. 4.12, photolysis and photoionization are competing processes. The significance of isolating the parent ion should not be underestimated. Large molecules such as Ir(ppy) 3 have a high density of states (§ 4.3), so terminating the ab- sorption after the ionizing transition is not necessarily realistic. When ions continue to ab- sorb after the ionization step, fragmentation is expected (§ 2.2.3). Therefore, it was not obvious a priori that the parent ion could be obtained. Furthermore, the parent ion is well suited for extracting spectroscopic information because its photon—and therefore fre- quency—dependence is generally less complex than that of smaller ion fragments (§2.2). Figure 4.13 Logarithmic plot of the Ir(ppy) 3 + ion signal vs. laser fluence collected using a focused laser, in the range 2 – 15 mJ / cm 2 . The order is n = 1.5, indicating that one of the two transitions used for ionization is nearly saturated. Figure 4.13 shows the fluence dependence of the Ir(ppy) 3 + signal collected using a focused laser. The signal intensity has a 1.5-order dependence on the energy fluence in log (ion signal) log (µ J / cm 2 ) n = 1.5 94 the range 2 – 15 mJ / cm 2 . Regardless of the ionization scheme, if the transitions are not saturated (Eq. (2.7)), then the order of the fluence dependence will reflect the number of photons absorbed. Therefore, this result is a manifestation of saturation. Figure 4.14 Logarithmic plots of the Ir(ppy) 3 + ion signal intensity vs. laser fluence at different photon energies: (a) 34,207 cm -1 , 76 – 406 µJ / cm 2 ; (b) 34,682 cm -1 ; 425 – 1100 µJ / cm 2 ; (c) 35,420 cm -1 , 215 – 960 µJ / cm 2 . In general, the ion signal depends on the second-order of the fluence (n = 2.0) in the range 34,207 – 35,930 cm -1 under approximately 1000 µJ / cm 2 . Saturation is suppressed by reducing the fluence. However, the focused area of the laser limits the minimum fluence to approximately 1.5 mJ / cm 2 . A smaller fluence is obtained with an unfocused laser; however, this results in a large ionization region that log (ion signal) log (µJ / cm 2 ) log (ion signal) log (µ J / cm 2 ) log (ion signal) log (µJ / cm 2 ) (c) (a) (b) 95 severely hinders the TOF resolution (§2.3). A resolution comparable to that of the fo- cused laser is achieved by skimming the unfocused beam with a 1 mm aperture. The fo- cusing condition given by Eq. (2.19) provides further resolution improvement. Using this setup, a thorough examination of the Ir(ppy) 3 + fluence dependence in the frequency range hv = 34,207 – 35,930 cm -1 reveals a nearly perfect second-order fluence dependence. The three plots presented in Fig. 4.14 represent a small sample of the data collected with a fluence of 75 – 1200 µJ / cm 2 . Decreasing the fluence beyond this range causes the Ir(ppy) 3 + signal to disappear completely, indicating that the parent ion is the result of a two-photon process. 34200 34500 34800 35100 35400 35700 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Ir(ppy) 3 + ion intensity photon energy / cm -1 Figure 4.15 Ir(ppy) 3 + signal intensity vs. hv. This spectrum is comprised of seven spectra. The Ir(ppy) 3 + signal intensity was monitored as the laser frequency was scanned from 34,150 to 35,775 cm -1 . To ensure that the fluence did not deviate more than 5% 96 from 450 µJ / cm 2 , the spectrum shown in Fig. 4.15 is comprised of seven small spectra collected with different laser dye mixtures (Table 4.1). Moreover, the laser power and ion signal were collected simultaneously, which enabled the spectrum to be fluence cor- rected. The limited range of easily accessible photon energies prevented this study from being extended continuously to longer wavelengths. However, the Ir(ppy) 3 + signal was examined using the 355 nm (3.49 eV) output of the Nd:YAG laser. Because the absorp- tion cross section decreases with frequency (Fig. 4.8), a relatively large fluence was re- quired. Ten fluence dependence curves were collected in the range 1 – 4 mJ / cm 2 . These are well represented by the two curves shown in Fig. 4.16. In general, the results were in- consistent; the fluence dependence varied in order from 1.5 to 3. Figure 4.16 Selected logarithmic plots of the Ir(ppy) 3 + signal intensity vs. laser fluence at 28,169 cm -1 (355 nm). A large fluence was required because of the small absorption cross section at this photon energy. log (ion signal) log (µ J / cm 2 ) log (ion signal) log (µ J / cm 2 ) (a) (b) n = 2.1 n = 3.3 1.6 – 3.9 µJ / cm 2 1.4 – 3.4 µJ / cm 2 97 4.5 Discussion 4.5.1 Ionization Mechanism Characterizing the ionization scheme enables the ion signal to be fully understood and provides a framework for the rest of this section. The stepwise ionization process de- pends primarily on the laser fluence and the dynamics of the intermediate state. The fluence determines the transition rate and ultimately the extent of photon absorption, while the latter affects the lifetime of the excited intermediate state. When Ir(ppy) 3 is ionized with photon energies around 4.3 eV, the first photon ex- cites the π–π* (LC) transitions. Therefore, the excited state properties of this absorption band are of primary interest and are addressed in the following discussion. Once the ex- cited state dynamics are examined, the ionization scheme is qualitatively described in terms of transition rate equations. Insight into the photophysical properties of Ir(ppy) 3 is provided by previous experimental studies. Ir(ppy) 3 exhibits intense phosphorescence from the T 1 state with a radiative lifetime of approximately 1.6 µs. 17 Time-resolved transient absorption measure- ments revealed ultrafast dynamics. 22 Photoexcitation with a 400 nm laser pulse accessed the 1 MLCT band. Depending on the wavelength of the probe, transient absorption ap- peared with a time constant of 70 – 100 fs, and remained constant up to the maximum delay of 1.5 ns. The relatively long-lived transient state was assigned to the phosphor- escent 3 MLCT (T 1 ) state, indicating very fast intersystem crossing (ISC). This result is supported by the complete absence of fluorescence, even at low temperatures. 6,23 98 Next, the high-lying LC states were excited with a 266 nm pump pulse. In this case, the appearance of the transient absorption was fit with a biexponential function. Holding one time constant fixed at the ISC value, the other time constant was 300 – 350 fs, which corresponds to internal conversion from 1 LC to 1 MLCT. Importantly, this study demonstrates that photoexcited Ir(ppy) 3 relaxes to the luminescent state on a sub-pico- second timescale. These results are summarized in Fig. 4.17. Figure 4.17 Schematic of the molecular dynamics of Ir(ppy) 3 adapted from Ref. (22). Rapid ISC and the lack of fluorescence are a consequence of strong SOC. A num- ber of theoretical studies, which lend support to the time-resolved results, examined the effects of SOC on Ir(ppy) 3 . 15,17,24 Large SOC induces strong singlet and triplet state 266 nm 400 nm S 0 1 LC 1 MLCT 3 MLCT Phosphorescence τ 1 τ 2 τ 1 = 70 – 100 fs τ 2 = 300 – 350 fs 99 mixing. 15,24 As a result, it has been suggested that photoexcitation produces spin-mixed states directly; the term ISC has little meaning in this molecule. 24 Hence, the sub-pico- second dynamics are not surprising. On the other hand, calculations indicate that the bright states have considerable triplet character. 15,17,24 The luminescence is therefore ade- quately described as phosphorescence. As a testament to the reliability of these compu- tations, the radiative lifetime was qualitatively reproduced when SOC effects are ac- counted for. 17 A general consideration of the transition rate enables the timescales of the ioni- zation and relaxation processes to be compared. The transition probability per unit time, or transition rate W, is given by the absorption cross section σ and laser intensity I / dP dt W I σ = = (4.1) Assuming a square laser pulse with time duration τ p , the fluence given in Eq. (2.8) be- comes p Iτ Φ= . The transition rate is then written in terms of fluence p W σ τ Φ = (4.2) Saturation can be broadly defined as the point in which absorption becomes independent of the laser fluence. A transition approaches saturation when the fluence is such that 1 σ Φ ≥ . 25 Therefore, close to saturation, Eq. (4.2) becomes 1 p W τ ≥ (4.3) As discussed in Sect. 4.4, the Ir(ppy) 3 + TOF signal is the result of two-photon ion- ization. In the fluence range used to record Fig. 4.15, the Ir(ppy) 3 + signal has a second 100 order dependence on the laser fluence. Therefore, neither the excitation nor ionization transition is saturated. In this case, each transition has a rate of 1/ p W τ < . If it is assumed that the photoexcited 1 LC state relaxes to the phosphorescent T 1 state in 500 fs, then the ionization probability from the intermediate state is 500 0 dP Wdt = ∫ ∫ (4.4) where the integration is over the lifetime of the 1 LC state. The probability is roughly ap- proximated using the upper limit of the ionization rate for a non-saturated transition, i.e. 1/ p W τ = . The pulse width τ p is approximately 10 ns. Equation (4.4) becomes 500 ~ 10 fs P ns (4.5) Far from saturation, as is the case for fluence range used to record Figs. 4.14 and 4.15, the ionization probability from 1 LC is much less than 1%. Therefore, ionization occurs from the T 1 state as opposed to the 1 LC state. Figure 4.18 shows the ionization scheme for Ir(ppy) 3 when Φ < 1 mJ / cm 2 . The molecule is initially in the ground electronic state S 0 . Irradiation with a 10 ns UV laser pulse creates a short-lived (< 500 fs) intermediate 1 LC state. While still bathed in radi- ation, relaxation brings the molecule to the metastable (τ = 1.6 µs) T 1 state. Ionization occurs from this state during the same laser pulse. Subsequent to ionization, the absorp- tion process is terminated, as indicated by the lack of fragmentation. Increasing the fluence alters the ionization scheme and complicates its character- ization. For example, the Ir(ppy) 3 + signal in Fig. 4.13 shows signs of saturation: the 101 fluence dependence is less than second order. While the ionization scheme is likely sim- ilar to that shown in Fig. 4.18, it is unclear which state the molecule is ionized from. Although the parent ion is the result of two-photon ionization, the role of the relaxation dynamics is ambiguous. The origins of the smaller fragments that appear at higher fluence are more complex, owing to the competing dissociation channel, and are dis- cussed in Sect. 4.5.4. Figure 4.18 Schematic of the ionization mechanism. UV photon absorption excites the 1 LC band. Ultrafast (< 500 fs) 22 radiationless deactivation populates the lowest triplet state, which is ionized with another photon. See Sect. 4.5.2 for a discussion of the ionization energy. Although these results are arrived at using an oversimplified approach, they are consistent with what would be expected. The lifetime of the intermediate state is much shorter then the laser pulse width. Hence, a very large fluence is required for the ioni- ( 1 LC) 4.0 eV ( 3 MLCT) 2.3 eV Relaxation < 500 fs hv IE < 6.55 eV hv S 0 102 zation probability from the 1 LC state to be significant. With a large fluence, the scheme closely resembles multiphoton ionization: saturation abounds and fragmentation is ex- pected. Alternatively, at low fluence, when the parent ion is isolated and saturation is not realized, the ionization rate cannot compete with the ultrafast relaxation dynamics. 4.5.2 Ionization Energy The ionization scheme depicted in Fig. 4.18 enables the IE of Ir(ppy) 3 to be exam- ined. During photo-irradiation, the first photon excites the intermediate 1 LC state. Subse- quently, ultrafast radiationless deactivation populates the metastable phosphorescent state. For an isolated molecule, energy conservation requires that radiationless electronic relaxation be accompanied by an increase in vibrational energy (E vib ). Therefore, ion- ization occurs from a vibrationally hot metastable state. The geometry of the ion does not differ significantly from the optimized T 1 geom- etry (§ 4.3), as is often the case for large molecules. 26 Therefore, vibrational overlap dic- tates that the ion is formed with a similar amount of vibrational energy as the neutral pre- cursor. Intuitively, this is understood by considering a large vibrationally excited mole- cule; removing one electron is unlikely to significantly affect the vibrational motion. Consequently, when the neutral is vibrationally excited, there is an effective ionization energy (IE eff ) given by eff vib IE IE E ≈ + (4.6) This is shown schematically in Fig. 2.5. 103 The lowest energy photon for which the Ir(ppy) 3 + signal has a second order de- pendence on the fluence is 34,207 cm -1 (4.24 eV), which is shown in Fig. 4.14c. There- fore, two-photon ionization at this photon energy is described well by the mechanism pre- sented above (Fig. 4.18) and provides an upper limit of the ionization energy. Absorption of the first photon populates the metastable T 1 state, leaving the molecule vibrationally excited. Ionization occurs via absorption of another photon. Because two 4.24 eV pho- tons are sufficient for ionization, IE eff has an upper limit of 1 T eff vib IE E hv E ≤ + + (4.7) Here, the energy of the T 1 state ( 1 T E ) is 2.307 eV (Fig. 4.7), and hv is 4.24 eV. Com- bining Eqs. (4.6) and (4.7) gives 1 T IE E hv ≤ + (4.8) because E vib is largely unaffected by ionization. Therefore, IE 6.55 eV ≤ . Although this result is consistent with the calculated IE values of 5.86 eV (§ 5.3) and 5.94 eV, 14 it is less than the previously reported value of 7.2 eV obtained with elec- tron energy loss spectroscopy (EELS). 27 The 7.2 eV result was taken directly from the EELS spectrum, which shows a broad absorption starting close to 6 eV and peaking at 7.2 eV. Although the Ir(ppy) 3 was sublimed in an oven for these measurements, no attempt is made to account for the vibrational energy of the neutral. E vib is not insig- nificant; after all, according to Sect. 3.3, at 500 K E vib amounts to nearly 1.4 eV. Because E vib is nearly constant during the ionization process, the ionization probability—and therefore ion signal—should be small near the true IE and gradually increase as IE eff is 104 approached. Therefore, it is likely that 7.2 eV corresponds to IE eff , while the actual IE is closer to 6 eV, which matches the other values well. At 355 nm (3.49 eV), the Ir(ppy) 3 + signal shows a 1.5 – 3 order power dependence on the laser fluence (Fig. 4.16). One possible explanation of these results is provided by the transition rate discussion made above. Consider a situation in which there is compe- tition between radiationless deactivation and ionization from the intermediate 1 LC state. For an IE of 6 – 6.5 eV, two 3.49 eV photons are sufficient for ionization if radiationless decay is negligible. Because the lifetime of the intermediate state is approximately 100 fs, 22 this ionization mechanism is likely to exhibit saturation—hence the 1.5 order power dependence. On the other hand, when relaxation populates the T 1 state, ionization re- quires three 3.49 eV photons. Depending on the extent of saturation, this leads to a fluence dependence of 3 or less. 4.5.3 Ir(ppy) 3 + Action Spectrum The Ir(ppy) 3 + action spectrum shown in Fig. 4.15 contains four broad peaks over- lapped with a large continuum. In general, the ion signal depends on the absorption cross section of both the excitation and ionization transitions and the laser fluence (Eq. (2.10)). However, because the action spectrum was fluence corrected, the signal depends solely on the absorption cross sections of each transition. In general, ionizing transitions often exhibit only weak frequency dependence. 28,29 This is demonstrated by the striking resemblance between the action and UV-Vis (Figs. 4.8 and 4.15) spectra, which indicates that the Ir(ppy) 3 + action spectrum corresponds to 105 the intermediate excited state. Because the absorption spectrum is reproduced using the calculated electronic states (Fig. 4.6) and temperature has little effect on the absorption spectrum (Fig. 4.8), the structure of the background continuum is due overwhelmingly to the large state density in the spectral region. The broad peaks are separated by approximately 250 cm -1 and are attributed to a low frequency vibrational mode. The observed line spacing suggests that the progression can be assigned to the lowest frequency mode of the ppy molecule. This out-of-plane inter-ring bend has a frequency of 251 cm -1 in free ppy. 30 However, because ppy is a bidentate chelator, the frequency of this mode is likely to be shifted to higher energy when ppy is incorporated in Ir(ppy) 3 . Another possible explanation is a mode arising from the iridium-ligand coupling. As discussed in Sect. 3.3, the frequency of this mode is expected to be low and may account for the vibrational lines observed in the action spectrum. 4.5.4 Ir + Signal In contrast to the Ir(ppy) 3 + action spectrum, the spectra obtained by monitoring the Ir + ion signal intensity exhibits many sharp features (Fig. 4.10). These features corre- spond to atomic iridium transitions, indicating that Ir(ppy) 3 undergoes photolysis when irradiated with nanosecond UV laser pulses. Subsequently, the liberated iridium is probed via REMPI. In light of the two-step ionization mechanism of Ir(ppy) 3 presented above, the production of atomic iridium by photolysis seems surprising. However, previous 106 studies demonstrate that photolysis is a common—and often dominant—pathway for photoexcited organometallic compounds. 31–37 There are two very similar processes by which iridium can be ionized. The ion- ization energy of iridium is 9.1 eV. 21 In the photon energy range used in this work, namely 4.33 – 4.44 eV, iridium is ionized with two or three photons depending on initial atomic state. From the ground state, ionization occurs via one-photon resonant absorption followed by two-photon ionization, or 1+2 REMPI. On the other hand, the first excited state of atomic iridium lies 0.35 eV (2835 cm -1 ) above the ground state 21 and, therefore, can be ionized with either 1+2 REMPI (when hv < 4.38 eV) or one-photon resonant ab- sorption followed by one-photon ionization, 1+1 REMPI (when hv > 4.38 eV). In Table C.1 (Appendix C), the observed transitions from this work are listed along with the corresponding lines and initial energy states taken from Ref. (21). Based on this data, the last column on the right specifies the ionization mechanism. Close in- spection reveals that the majority of transitions originate in excited atomic states, in which case ionization occurs via 1+1 REMPI. It is difficult to draw strong conclusions about the excited state population distri- bution of iridium because the spectral region is small. Moreover, the intensities of the ion signals arising from 1+2 REMPI are necessarily less than for the 1+1 REMPI ion signals because a two-photon ionizing step has a lower probability than a one-photon step. How- ever, a broad distribution of iridium states is detected; the observed transitions have ori- gins ranging from the 4 F 9/2 ground state to the 2 G 9/2 state, which is 23,506 cm -1 (2.91 eV) above the ground state. 107 The results of this work are insufficient for the unambiguous characterization of the photodissociation process. Not only is the dissociation energy unknown, the mech- anism by which Ir(ppy) 3 sheds its ligands is not understood. However, it is likely that photolysis occurs after absorption of two UV photons. In this case, photoexcitation pro- duces an excited state that lies above the ionization threshold. This so-called superexcited state has two decay channels: autoionization or molecular dissociation. On the other hand, the possibility of photolysis occurring after absorption of a single photon cannot be ruled out. At 500 K, Ir(ppy) 3 contains a significant amount of vibrational energy. A single photon imparts an additional 4.3 eV to the molecule, which may be sufficient for dissociation. This process is depicted schematically in Fig. 4.19. Figure 4.19 Diagram of the ionization and dissociation channels. The dissociation pathway is not charac- terized, and is therefore labeled with a question mark. Energy is not to scale. Ir(ppy) 3 Ir(ppy) 3 ** Ir(ppy) 3 * hv hv IE Ir(ppy) 2 + , Ir + nhv Ir(ppy) 3 + + e – ? Ir(ppy) 2 + ppy 3(ppy) + Ir… Ir + + e – 2 – 3 hv Energy 108 In light of the competing dissociation and ionization channels, the vastly different TOF mass spectra presented in Fig. 4.12 can be understood. At low fluence, two photons are absorbed by Ir(ppy) 3 and both channels are accessed. However, the fluence is insuf- ficient for subsequent iridium ionization and the parent ion is isolated. As the fluence is increased, more photons are absorbed and smaller ion fragments appear. In general, the lighter fragments can be the result of either the ionization or dissociation channel. The first fragment that appears is Ir(ppy) 2 + . The intensity of this ion signal is small relative to the parent ion intensity, and is only observed when the parent ion is present. These observations suggest that the Ir(ppy) 2 + fragment is due to Ir(ppy) 3 + absorp- tion followed by fragmentation, i.e. the ionization channel. For the same reason, the weak IrC n + and IrN n + progression seen at 88 mJ / cm 2 is also attributed to Ir(ppy) 3 + fragmen- tation. At high fluence, increased absorption causes the Ir + fragment to become the dom- inant product of the ionization channel; larger ion fragment signals decrease. Evidently, the Ir + ion fragment is the only significant signal corresponding to the dissociation path- way. This scheme is summarized in Fig. 4.19. Interestingly, the ppy + ion is only detected as a very minor product at high fluence (Fig. 3.4) and the Ir(ppy) + fragment is not ob- served. The model illustrated in Fig. 4.19 provides insight into the Ir + action spectra shown in Fig. 4.10. The sharp resonant features correspond to the dissociation channel, while the weak Ir + continuum is attributed to the ionization channel. A comparison of the background Ir + signal intensity (Fig. 4.11) and the Ir(ppy) 3 + signal intensity (Fig. 4.15) 109 supports this conclusion. The large Ir + continuum seen at higher fluence (bottom panel, Fig. 4.10) is due to power broadening. A comparison of the Ir + and Ir(ppy) 3 + signal intensities also allows for a qualita- tive look at the branching ratio of the competing decay channels. The Ir(ppy) 3 + signal in- tensity in Fig. 4.15 provides an approximate measure of the molecular ionization path- way. Although increasing the fluence beyond this range results in a larger ion signal, sat- uration is rapidly realized (Fig. 4.13). This quickly gives way to fragmentation and a de- crease in Ir(ppy) 3 + intensity. The nearly saturated Ir + signal shown in the bottom panel of Fig. 4.10 is a yardstick of the dissociation pathway. Using these indicators, the disso- ciation channel is clearly dominant. This conclusion reiterates how surprising acquiring the parent ion is. 4.6 Future Directions The results presented herein can be better understood by simply extending the current investigation. For example, a comprehensive fluence-dependence study—in par- ticular, monitoring the various ion fragment signals over a broad fluence range at dif- ferent frequencies—will provide insight into the fragmentation process. Such a study would reveal the number of photons required to create each fragment, which will clarify the origin of the ion fragments. Because the parent ion can be isolated with careful choice of ionization con- ditions, a number of high-resolution laser spectroscopy techniques are feasible. Mass an- alyzed threshold ionization spectroscopy (MATI), 38,39 zero electron kinetic energy spec- 110 troscopy (ZEKE), 40 along with a variety of REMPI techniques (§ 2.2) can provide infor- mation on electronic states, vibrational states of the molecular ion, and highly accurate ionization energies. 41 However, obtaining high-resolution data from these techniques often requires a cold sample. Typically, this is done with supersonic expansion, but helium droplets may be a viable option. While the former is difficult owing to pyrolysis, ingenuity and determination should not be underestimated. On the other hand, helium droplet isolation is easily achievable under the right conditions, as discussed in Sect. 3.6. From a molecular dynamics standpoint, the dissociation mechanism is an inter- esting issue. This process can be explored with femtosecond pump-probe spectroscopy. For instance, a cooled gas-phase sample can be excited with a short-duration laser pulse. Any products that are formed are ionized with a second laser pulse, and the ions analyzed with TOF mass spectroscopy. This method was employed to elucidate the ultrafast disso- ciation of Fe(CO) 5 . 37 Experiments based on this general concept will answer key ques- tions surrounding the dissociation process: (1) how many photons are required for disso- ciation; (2) is decomposition sequential or concerted; and (3) what role do vibrations play? Furthermore, a complete characterization of the dissociation mechanism will pro- vide a glimpse into the formation of different fragment ions, which could explain the ab- sence of the Ir(ppy) 2 + fragment in this work. 4.7 Conclusion Multistep ionization of Ir(ppy) 3 reveals competing decay channels. The absorption of two UV photons provides access to the molecular ionization and dissociation path- 111 ways. Photolysis leads predominantly to atomic iridium, which is subsequently probed via REMPI spectroscopy. Alternatively, the ionization channel results in a number of fragments, depending on the energy fluence of the ionizing laser. The branching ratio of the two channels is determined qualitatively by comparing fragment ions that result from the respective processes. The dissociation channel is the principal decay pathway. Isolation of the parent Ir(ppy) 3 + ion is achieved at fluences under 1.5 mJ / cm 2 . The fluence dependence of this signal shows a two-photon dependence, allowing the ion- ization scheme to be deduced. As a consequence, an upper-limit of the ionization energy (6.55 eV) was easily determined. In addition, the frequency variation of the parent ion signal provides molecular excited state information. The importance of isolating the parent ion cannot be overstated. The work pres- ented in this chapter demonstrates that the ionization pathway is only a minor channel. Furthermore, variation of the TOF mass spectra with fluence revealed that the parent ion easily absorbs additional photons, causing significant ion fragmentation. However, not only was parent ion isolated, it was completely characterized, providing a means of ac- quiring high-resolution molecular data. Future experiments may utilize REMPI, MATI, and / or ZEKE, all of which are excellent spectroscopic techniques. 112 4.8 References 1. Yersin, H. Top. Curr. Chem. 2004, 241, 1. 2. Yersin, H.; Finkenzeller, W. J. Highly Efficient OLEDs with Phosphorescent Materials; Yersin, H., Ed.; Wiley-VCH Verlag GmbH & Co.: Wienheim, 2008; pp. 1 – 97. 3. Adachi, C.; Baldo, M. A.; Thompson, M. E.; Forrest, S. R. J. Appl. Phys. 2001, 90, 5048. 4. Baldo, M. A.; Forrest, S. R. Phys. Rev. B 2000, 62, 10958. 5. Baldo, M. A.; O’Brien, D. F.; Thompson, M. E.; Forrest, S. R. Phys. Rev. B 1999, 60, 14422. 6. Finkenzeller, W. J.; Yersin, H. Chem. Phys. Lett. 2003, 377, 299. 7. Adachi, C.; Baldo, M. A.; Forrest, S. R. J.; Thompson, M. E. Appl. Phys. Lett. 2000, 77, 904. 8. King, K. A.; Spellane, P. J.; Watts, R. J. J. Am. Chem. Soc. 1985, 107, 1431. 9. Lamansky, S.; Djurovich, P.; Murphy, D.; Abdel-Razzaq, F.; Lee, H,; Adachi, C.; Burrows, P.; Forrest, S. R.; Thompson, M. E. J. Am. Chem Soc. 2001, 123, 4304. 10. Liu, X.; Feng, J.; Ren, A.; Yang, L.; Yang, B.; Ma, Y. Opt. Mater. 2006, 29, 231. 11. Fine, J.; Nemirow, C.; Krylov, A.; Wittig, C. in preparation. 12. Kolner, B. H. Opt. Lett. 1991, 61, 784. 113 13. Shao, Y.; Molnar, L. F.; Jung, Y.; Kussmann, J.; Ochsenfeld, C.; Brown, S. T.; Gilbert, A. T. B.; Slipchenko, L. V.; Levchenko, S. V.; O’Neill, D. P.; DiStasio, R. A.; Lochan, R. C.; Wang, T.; Beran, G. J. O.; Besley, N. A.; Herbert, J. M.; Lin, C. Y.; Van Voorhis, T.; Chien, S. H.; Sodt, A.; Steele, R. P.; Rassolov, V. A.; Maslen, P. E.; Korambath, P. P.; Adamson, R. D.; Austin, B.; Baker, J.; Byrd, E. F. C.; Dachsel, H.; Doerksen, R. J.; Dreuw, A.; Dunietz, B. D.; Dutoi, A. D.; Furlani, T. R.; Gwaltney, S. R.; Heyden, A.; Hirata, S.; Hsu, C. P.; Kedziora, G.; Khalliulin, R. Z.; Klunzinger, P.; Lee, A. M.; Lee, M. S.; Liang, W.; Lotan, I.; Nair, N.; Peters, B.; Proynov, E. I.; Pieniazek, P. A.; Rhee, Y. M.; Ritchie, J.; Rosta, E.; Sherrill, C. D.; Simmonett, A. C.; Subotnik, J. E.; Woodcock, H. L.; Zhang, W.; Bell, A. T.; Chakraborty, A. K.; Chipman, D. M.; Keil, F. J.; Warshel, A.; Hehre, W. J.; Schaefer, H. F.; Kong, J.; Krylov, A. I.; Gill, P. M. W.; Head- Gordon, M. Phys. Chem. Chem. Phys. 2006, 8, 3172. 14. Hay, P. J. J. Phys. Chem. A 2002, 106, 1634. 15. Matsushita, T.; Asada, T.; Koseki, S. J. Phys. Chem. C 2007, 111, 6897. 16. Tamayo, A. B.; Alleyne, B. D.; Djurovich, P. I. Lamansky, S.; Tsyba, I.; Ho, N. N.; Bau, R.; Thompson, M. E.; J. Am. Chem. Soc. 2003, 125, 7377. 17. Jansson, E.; Minaev, B.; Schrader, S.; Ågren, H. Chem. Phys. 2007, 333, 157. 18. Garces, F. O.; Dedeian, K.; Keder, N. L.; Watts, R. J. Acta. Crystallogr. 1993, C49, 1117. 19. Krumholz, P. J. Am. Chem. Soc. 1951, 73, 3487. 20. Vacha, M.; Koide, Y.; Kotani, M.; Sato, H. J. Lumin. 2004, 107, 51. 21. Van Kleef, T. A. M. Physica 1957, 23, 843. 22. Tang, K.; Liu, K. L.; Chen, I. Chem. Phys. Lett. 2004, 386, 437. 23. Hedley, G. J.; Ruseckas, A.; Samuel, I. D. W. Chem. Phys. Lett. 2008, 450, 292. 24. Nozaki, K. J. Chin. Chem. Soc. 2006, 53, 101. 25. Letokhov, V. S. Laser Photoionization Spectroscopy; Academic Press, Inc.: Orlando, FL, 1987. 26. Johnson, P. M.; Otis, C. O. Ann. Rev. Phys. Chem. 1981, 32, 139. 114 27. Kukhta, A. V.; Kukhta, I. N.; Bagnich, S. A.; Kazakov, S. M.; Andreev, V. A.; Neyra, O. L.; Meza, E. Chem. Phys. Lett. 2007, 434, 11. 28. Antonov, V. S.; Letokov, V. S. Appl. Phys. 1981, 24, 89. 29. Ashfold, M. N. R.; Howe, J. D. Annu. Rev. Phys. Chem. 1994, 45, 57. 30. Sett, P.; Chattopadhyay, S.; Mallick, P. K. Spectro. Acta A 2000, 56, 855. 31. Ketkov, S. Y.; Selzle, H. L.; Schlag, E. W.; Domrachev, G. J. Phys. Chem. A 2003, 107, 4041. 32. Ketkov, S. Y.; Selzle, H. L.; Schlag, E. W.; Titova, S. N. Chem. Phys. 2003, 293, 91. 33. Luetwyler, S. Even, U.; Jortner, J. J. Phys. Chem. 1981, 85, 3026. 34. Luetwyler, S. Even, U.; Jortner, J. Chem. Phys. Lett. 1980, 74, 11. 35. Engelking, P. C. Chem. Phys. Lett. 1980, 74, 207. 36. Braun, J. E.; Neusser, H. J.; Harter, P.; Stockl, M. J. Phys. Chem. A 2000, 104, 2013. 37. Banares, L.; Baumet, T.; Bergt, M.; Kiefer, B.; Gerber, G. J. Chem. Phys. 1998, 108, 5799. 38. Zhu, L.; Johnson, P.M. J. Chem. Phys. 1991, 94, 5769. 39. Krause, H.; Neusser, H.J. J. Chem. Phys. 1992, 97, 5923. 40. Schlag, E. W. ZEKE Spectroscopy; Cambridge University Press: Cambridge, 1998. 41. Ketkov, S. Y.; Selzle, H. L.; Schlag, E. W. Organomettallics. 2006, 25, 1712. 115 Bibliography Abramczyk, H. Introduction to Laser Spectroscopy; Elsevier: Amsterdam, 2005; pp. 1 – 18. Adachi, C.; Baldo, M. A.; Forrest, S. R. J.; Thompson, M. E. Appl. Phys. Lett. 2000, 77, 904. Adachi, C.; Baldo, M. A.; Thompson, M. E.; Forrest, S. R. J. Appl. Phys. 2001, 90, 5048. Akers, K.; Aroca, R.; Hor, A.; Loutfy, R. O. J. Phys. Chem. 1987, 91, 2954. Anderson, J. B.; Andres, R. P.; Fenn, J. B. Advances in Atomic and Molecular Physics; Bates, D. R. Ed.; Academic Press: New York, NY, 1965; Vol. I; pp. 345 – 359. Andrews, L.; Moskovitis, M. Chemistry and Physics of Matrix-Isolated Species; North Holland, Amsterdam, 1989; pg. 430. Antonov, V. S.; Letokov, V. S. Appl. Phys. 1981, 24, 89. Ashfold, M. N. R.; Howe, J. D. Annu. Rev. Phys. Chem. 1994, 45, 57. Atkins, P. W.; Friedman, R. S. Molecular Quantum Mechanics, 3rd ed; Oxford University Press: New York, NY, 1997. Baldo, M. A.; Forrest, S. R. Phys. Rev. B 2000, 62, 10958. Baldo, M. A.; Lamansky, S.; Burrows, P. E.; Thompson, M. E.; Forrest, S. R. Appl. Phys. Lett. 1999, 75, 4. Baldo, M. A.; O’Brien, D. F.; Thompson, M. E.; Forrest, S. R. Phys. Rev. B 1999, 60, 14422. Banares, L.; Baumet, T.; Bergt, M.; Kiefer, B.; Gerber, G. J. Chem. Phys. 1998, 108, 5799. Barranco, M.; Guardiola, R.; Hernandez, S.; Mayol, R.; Navarro, J.; Pi, M. J. Low Temp. Phys. 2006, 142, 1. Boesl, U.; Neusser, H. J.; Schlag, E. W. J. Am. Chem. Soc. 1981, 103, 5058. Braun, J. E.; Neusser, H. J.; Harter, P.; Stockl, M. J. Phys. Chem. A 2000, 104, 2013. 116 Buchenau, H.; Knuth, E. L.; Northby, J.; Toennies, J. P.; Winkler, C. J. Chem. Phys. 1990, 92, 6875. Bulovic, V.; Burrows, P. E.; Forrest, S. R., Cronin, J. A.; Thompson, M. E. Chem. Phys. 1996, 210, 1. Callicoat, B. E.; Forde, K.; Ruchti, T.; Jung, L.; Janda, K. C.; Halberstadt, N. J. Chem. Phys. 1998, 108, 9371. Callicoat, B. E.; Mar, D. D.; Apkarian, V. A.; Janda, K. C. J. Chem. Phys. 1996, 105, 7872. Close, J. D.; Federmann, F.; Hoffmann, K.; Quaas, N. Chem. Phys. Lett. 1997, 276, 393. Demtroder, W. Laser Spectroscopy: Basic Concepts and Instrumentation, 2nd ed.; Springer: New York, NY, 1998. Engelking, P. C. Chem. Phys. Lett. 1980, 74, 207. Fine, J.; Nemirow, C.; Krylov, A.; Wittig, C. in preparation. Finkenzeller, W. J.; Yersin, H. Chem. Phys. Lett. 2003, 377, 299. Garces, F. O.; Dedeian, K.; Keder, N. L.; Watts, R. J. Acta. Crystallogr. 1993, C49, 1117. Harms, J.; Toennies, J. P.; Dalfovo, F. Phys. Rev. B 1998, 58, 3341. Hartmann, M.; Lindinger, A.; Toennies, J. P.; Vilesov, A. F. Phys. Chem. Chem. Phys. 2002, 4, 4839. Hartmann, M.; Miller, R. E.; Toennies, J. P.; Vilesov, A. J. Science 1996, 272, 1631. Hay, P. J. J. Phys. Chem. A 2002, 106, 1634. Hedley, G. J.; Ruseckas, A.; Samuel, I. D. W. Chem. Phys. Lett. 2008, 450, 292. Hedley, G. J.; Ruseckas, A.; Samuel, I. D. W. J. Phys. Chem. A 2009, 113, 2. Hollas, M. High Resolution Spectroscopy; John Wiley & Sons: New York, NY, 1998. Isaq, M.; Gupta, S. P.; Sharma, S. D.; Yadav, B. S. Orient. J. Chem. 1998, 14, 387. Jansson, E.; Minaev, B.; Schrader, S.; Ågren, H. Chem. Phys. 2007, 333, 157. 117 Johnson, P. M. Acc. Chem. Res. 1980, 13, 20. Johnson, P. M.; Otis, C. O. Ann. Rev. Phys. Chem. 1981, 32, 139. Ketkov, S. Y.; Selzle, H. L.; Schlag, E. W. Organomettallics. 2006, 25, 1712. Ketkov, S. Y.; Selzle, H. L.; Schlag, E. W.; Domrachev, G. J. Phys. Chem. A 2003, 107, 4041. Ketkov, S. Y.; Selzle, H. L.; Schlag, E. W.; Titova, S. N. Chem. Phys. 2003, 293, 91. King, K. A.; Spellane, P. J.; Watts, R. J. J. Am. Chem. Soc. 1985, 107, 1431. Kolner, B. H. Opt. Lett. 1991, 61, 784. Krause, H.; Neusser, H.J. J. Chem. Phys. 1992, 97, 5923. Krumholz, P. J. Am. Chem. Soc. 1951, 73, 3487. Kukhta, A. V.; Kukhta, I. N.; Bagnich, S. A.; Kazakov, S. M.; Andreev, V. A.; Neyra, O. L.; Meza, E. Chem. Phys. Lett. 2007, 434, 11. Lamansky, S.; Djurovich, P.; Murphy, D.; Abdel-Razzaq, F.; Lee, H,; Adachi, C.; Burrows, P.; Forrest, S. R.; Thompson, M. E. J. Am. Chem Soc. 2001, 123, 4304. Letokhov, V. S. Laser Photoionization Spectroscopy; Academic Press, Inc.: Orlando, FL, 1987. Levenson, M. D. Introduction to Nonlinear Laser Spectroscopy; Academic Press: New York, NY, 1982. Levine, I. N. Physical Chemistry, 5th ed; McGraw-Hill: New York, NY, 2002. Lewerenz, M.; Schilling, B.; Toennies, J. P. Chem. Phys. Lett. 1993, 206, 381. Lewerenz, M.; Schilling, B.; Toennies, J. P. J. Chem. Phys. 1995, 102, 8191. Li, X.; Wu. Z.; Si, Z.; Zhang, H.; Zhou, L.; Liu, X. Inorg. Chem. 2009, 48, 7740. Lindinger, A.; Toennies, J. P.; Vilesov, A. F. J. Chem. Phys. 1999, 110, 1429. Liu, X.; Feng, J.; Ren, A.; Yang, L.; Yang, B.; Ma, Y. Opt. Mater. 2006, 29, 231. Luetwyler, S. Even, U.; Jortner, J. Chem. Phys. Lett. 1980, 74, 11. 118 Luetwyler, S. Even, U.; Jortner, J. J. Phys. Chem. 1981, 85, 3026. Mark, T. D.; Dunn, G. H. Electron Impact Ionization; Springer-Verlag / Wien: New York, NY, 1985. Matsushita, T.; Asada, T.; Koseki, S. J. Phys. Chem. C 2007, 111, 6897. McDaniel, E. W. Atomic Collisions: Electron and Photon Projectiles; John Wiley & Sons, Inc.: New York, NY, 1989. Miller, D. R. Atomic and Molecular Beams Methods; Scoles, G. Ed.; Oxford University Press: New York, NY, 1988; Vol. I; pp. 14 – 53. Moore, J. H.; Davis, C. C.; Coplan, M. A. Building Scientific Apparatus, 3rd ed.; Perseus Books: Cambridge, MA, 2003. Northby, J. A. J. Chem. Phys. 2001, 115, 10065. Nozaki, K. J. Chin. Chem. Soc. 2006, 53, 101. Polyakova, E. Multiphoton Ionization of Molecules Embedded in Superfliud Liquid Helium Droplets. Ph.D. Dissertation, University of Southern California, Los Angeles, CA, 2005. Polyakova, E.; Stolyarov, D.; Zhang, X.; Kresin, V. V.; Wittig, C. Chem. Phys. Lett. 2003, 375, 253. Ramsey, N. F. Molecular Beams; Oxford University Press: New York, NY, 1985. Ruchti, T.; Callicoat, B. E.; Janda, K. C. Phys. Chem. Chem. Phys. 2000, 2, 4075. Ruchti, T.; Forde, K.; Callicoat, B. E.; Ludwigs, H.; Janda, K. C. J. Chem. Phys. 1998, 109, 10679. Schlag, E. W. ZEKE Spectroscopy; Cambridge University Press: Cambridge, 1998. Schlag, E. W.; Neusser, H. J. Acc. Chem. Res. 1983, 16, 355. Sett, P.; Chattopadhyay, S.; Mallick, P. K. Spectro. Acta A 2000, 56, 855. 119 Shao, Y.; Molnar, L. F.; Jung, Y.; Kussmann, J.; Ochsenfeld, C.; Brown, S. T.; Gilbert, A. T. B.; Slipchenko, L. V.; Levchenko, S. V.; O’Neill, D. P.; DiStasio, R. A.; Lochan, R. C.; Wang, T.; Beran, G. J. O.; Besley, N. A.; Herbert, J. M.; Lin, C. Y.; Van Voorhis, T.; Chien, S. H.; Sodt, A.; Steele, R. P.; Rassolov, V. A.; Maslen, P. E.; Korambath, P. P.; Adamson, R. D.; Austin, B.; Baker, J.; Byrd, E. F. C.; Dachsel, H.; Doerksen, R. J.; Dreuw, A.; Dunietz, B. D.; Dutoi, A. D.; Furlani, T. R.; Gwaltney, S. R.; Heyden, A.; Hirata, S.; Hsu, C. P.; Kedziora, G.; Khalliulin, R. Z.; Klunzinger, P.; Lee, A. M.; Lee, M. S.; Liang, W.; Lotan, I.; Nair, N.; Peters, B.; Proynov, E. I.; Pieniazek, P. A.; Rhee, Y. M.; Ritchie, J.; Rosta, E.; Sherrill, C. D.; Simmonett, A. C.; Subotnik, J. E.; Woodcock, H. L.; Zhang, W.; Bell, A. T.; Chakraborty, A. K.; Chipman, D. M.; Keil, F. J.; Warshel, A.; Hehre, W. J.; Schaefer, H. F.; Kong, J.; Krylov, A. I.; Gill, P. M. W.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2006, 8, 3172. Slipchenko, M. N.; Kuma, S.; Momose, T.; Vilesov, A. F. Rev. Sci. Instrum. 2002, 73, 1. Stampor, W.; MKżyk, J.; Kalinowski, J. Chem. Phys. 2004, 300, 189. Stampor, W.; MKżyk, J.; Kalinowski, J.; Cocchi, M.; Virgili, D.; Fattori, V.; DiMarco, P. Macromol. Symp. 2004, 212, 509. Stienkemeier, F.; Lehmann, K. K. J. Phys. B 2006, 39, R127. Stienkemeier, F.; Vilesov, A. F. J. Chem. Phys. 2001, 115, 10119. Tamayo, A. B.; Alleyne, B. D.; Djurovich, P. I. Lamansky, S.; Tsyba, I.; Ho, N. N.; Bau, R.; Thompson, M. E.; J. Am. Chem. Soc. 2003, 125, 7377. Tang, K.; Liu, K. L.; Chen, I. Chem. Phys. Lett. 2004, 386, 437. Toennies, J. P.; Vilesov, A. F. Angew. Chem. Int. Ed. 2004, 43, 2622. Toennies, J. P.; Vilesov, A. F. Annu. Rev. Phys. Chem. 1998, 49, 1. Tsuboi, T. J. Lumin. 2006, 119–120, 288. Tsuboi, T.; Alaroudi, N. Phys. Rev. B 2005, 72, 125109. Vacha, M.; Koide, Y.; Kotani, M.; Sato, H. J. Lumin. 2004, 107, 51. Van Kleef, T. A. M. Physica 1957, 23, 843. Wewer, M.; Stienkemeier, F. Phys. Rev. B 2003, 67, 125201. Wiley, W. C.; McLaren, I. H. Rev. Sci. Instrum. 1955, 26, 1150. 120 Yersin, H. Top. Curr. Chem. 2004, 241, 1. Yersin, H.; Finkenzeller, W. J. Highly Efficient OLEDs with Phosphorescent Materials; Yersin, H., Ed.; Wiley-VCH Verlag GmbH & Co.: Wienheim, 2008; pp. 1 – 97. Zadorozhnyy, A. Energy Transfer Pathways for NO2-Rare Gas Complexes in Helium Droplets. Master Thesis, University of Southern California, Los Angles, CA, 2009. Zhu, L.; Johnson, P.M. J. Chem. Phys. 1991, 94, 5769. 121 Appendix A Ion Yield Derivation Consider the three-level system shown in Fig. A.1. Assume that the energy differ- ence between the first two states is such that the system lies in the ground state, i.e. 1 0 B E E k T − >> , where E i is the energy of the th i state, B k is the Boltzmann constant, and T is the temperature. Also assume that the final state | 2〉 is a continuum (ionized state) and the intermediate state |1〉 is a steady state, i.e. |1〉 is long-lived with respect to the laser pulse duration p t . Though the described system is idealized, it enables the number of ions produced during a single laser pulse to be examined. Moreover, some experi- mental systems are well represented by this model. Figure A.1 Schematic of a three level system. 1 0 B E E k T − >> |1〉 |2〉 |0〉 122 A laser pulse with intensity ( ) I t having units of photons / cm 2 s is turned on at time t = 0. The transition probability per unit time, or transition rate, from the ground state to the intermediate state is 1 ( ) dP I t dt σ = (A.1) Here, 1 σ is the absorption cross section for the transition to state |1〉 . At low fluence, far from the saturation limit (§ 2.2), stimulated emission can be ignored. In this case, the rate of change of the number of particles in the intermediate state 1 N is 1 0 1 ( ) dN N I t dt σ = (A.2) where 0 N is the number of particles in the ground state. The number of particles occu- pying the intermediate state after time 1 t , where 1 p t t ≤ , is 1 1 0 1 0 ( ) t N N I t dt σ = ∫ (A.3) Assuming a square laser pulse, this simplifies to 1 0 1 1 N N I t σ = (A.4) The rate of ion formation is given by 2 1 2 dN N I dt σ = (A.5) where 2 N is the number of ions, 2 σ is the ionization cross section, and a square laser pulse is assumed. 2 N is obtained by integrating from 1 t to p t , because ionization cannot occur until after 1 N is formed. Eq. (A.5) becomes 123 2 1 2 1 ( ) p N N I t t σ = − (A.6) Plugging Eq. (A.4) into Eq. (A.6) gives ( ) 2 2 2 0 1 2 1 1 p N N I t t t σ σ = − (A.7) The term in parenthesis can be simplified by noting that 1 0 p t t ≤ ≤ , and, therefore, 1 t can be written as ( ) 1 1/ p t x t = , where 1 x≥ (in the case of 1 0 t = , Eq. (A.7) goes to zero). In this case ( ) 2 2 1 1 2 1 1 1 p p t t t t x x x   − = − ≥     (A.8) The right-hand-side of Eq. (A.8) can be written as 2 ( ) p t f x , where the function f (x) varies from 0 to 0.25. For a square laser pulse, the fluence is defined as p It Φ= . There- fore, by Eqs. (A.7) and (A.8), the number of ions formed by a single laser pulse with duration p t is 2 2 0 1 2 ( ) 0 ( ) 0.25 N N f x f x σ σ = Φ ≤ ≤ (A.9) 124 Appendix B Ir(ppy) 3 -Helium Droplet Collision Probability As mentioned in Sect. 2.1, the helium droplet capture probability depends on the pick-up cell pressure, the capture cross section of the droplet, and the length of the pick- up region. The apparatus used for these experiments does not have an ion gauge directly in the pick-up cell, but does have one in the pick-up chamber. When gas is introduced to the pick-up cell there is effusion into the surrounding chamber. Therefore, it is possible to approximate the pressure inside the pick-up cell by measuring capture probability as a function of the pressure in the pick-up chamber. Table B.1 Helium droplet expansion conditions and constants T nozzle (K) P He (atm) N 〈 〉 1 R (Å) a v droplet (m / s) 2 σ coll (Å 2 ) b Ar 15.5 40 9000 46 350 6700 Ir(ppy) 3 16 40 8000 44 350 6200 a Eq. (2.1) b Eq. (2.2) During a previous experiment, 3 using the same experimental setup and similar helium droplet expansion conditions as the current work (Table B.1), the capture proba- bility of argon at room temperature as a function of the pick-up chamber ion-gauge reading was determined. This work relied heavily on the comprehensive study of argon entrenched helium droplets presented in Ref. (4). Briefly, the droplet beam was ionized via electron impact ionization and analyzed using a quadrupole mass spectrometer (QMS). Intensities of the ion peaks at 76, 80 and 120 amu, which correspond to ArHe 9 + , 125 Ar 2 + , and Ar 3 + , respectively, were monitored as a function of the pick-up chamber pres- sure. The experimental results are displayed in Fig. B.1. The intensities were adjusted to enable these data to be fit using the well-known Poisson distribution 5 ( ) ( ) exp ! k capt k capt n L I n L k σ γ σ γ = − (B.1) Here, k is the number of argon atoms captured, capt σ is the capture cross section for argon, L is the length of the pick-up cell, n is the number density in the pick-up cell, and ~ / rel droplet v v γ accounts for the relative velocities in the pick-up cell, where rel v is the average relative velocity and droplet v is the droplet beam velocity. 6 Figure B.1 Relative ion signal of ArHe 9 + (♦), Ar 2 + (●), and Ar 3 + (■) as a function of pick-up chamber pres- sure, recorded using a N 〈 〉 = 9000 helium droplets. These signals are indicative of droplets that have cap- tured 1 atom, 2 or 3 atoms, and 3 or more atoms, respectively. 4 The data was fit (black lines) using Eq. B.2 and Table B.2 (see text). Consequently, for a given pick-up chamber pressure, the capture probability k atoms is determined. 0 10 20 30 40 50 60 0.0 0.1 0.2 0.3 0.4 ArHe 9 + (76 amu) Ar 2 + (80 amu) Ar 3 + (120 amu) Fits Capture Probability for k Ar Pick-up Chamber Ion-Gauge Reading (10 -8 torr) k = 1 k = 2 k = 3 k = 4 126 For argon, the capture cross section is approximately equal to the collision cross section coll σ , 6 which in turn is defined as the cross-sectional area of the droplet, given by Eq. (2.2). For a nozzle temperature of 15.5 K and a helium backing pressure of 40 atm, the average droplet consists of 9000 atoms, and capt σ = 6700 Å 2 . 7 The length of the pick- up cell for the argon experiments was 4 cm. The atomic density n in the pick-up cell can be written as / Ar B x P k T , were P is the pick-up chamber pressure, B k is the Boltzmann constant, and Ar x is the fitting parameter that provides the relationship between the pick- up cell pressure and the pick-up chamber pressure for argon. Equation (B.1) becomes exp ! k Ar capt B Ar k Ar capt B P x L k T P I x L k k T σ γ σ γ         = −     (B.2) which is used to fit, by eye, the experimental data (black lines, Fig. B.1). The parameters used to calculate Eqs. (B.2) – (B.5) are listed in Table B.2. It should be noted that the monitored ion fragments do not unambiguously represent a particular number of argon atoms being captured. In fact, ArHe 9 + is attributed mostly to droplets containing one argon (i.e. ArHe N ) but can have minor contributions from Ar 2 He N droplets; Ar 2 + is representative of Ar 2 He N and Ar 3 He N droplets; and the Ar 3 + signal is primarily due to Ar 4 He N droplets. 4 The data in Fig. B.1 reflect this well. Once the relationship between the pick-up cell and pick-up chamber is estab- lished, for a given pick-up chamber pressure, the effusive rate from the pick-up cell Q, in atoms per second, can be calculated. 8 127 1 4 s Q nvA = (B.3) A s is the area of the effusive source; v is the average velocity of argon in the pick-up cell, given by m kT 3 ; and n is the number density in the pick-up cell, which was determined from the fit. Next, two assumptions are made: (1) the pumping speed for argon and Ir(ppy) 3 is the same, and (2) Ir(ppy) 3 equilibrates to 300 K in the pick-up chamber. Therefore, if the effusive rate of Ir(ppy) 3 is equal to the effusive rate of argon, i.e. 3 ) ( ppy Ir Ar Q Q = , the num- ber density and hence pressure in the pick-up chamber will be equivalent for each gas. Consequently, for Ir(ppy) 3 at a given pick-up chamber pressure, the pick-up cell pressure can be calculated by substituting the appropriate values into Eq. (B.3) from Table B.2. Although the number density in the pick-up chamber for both gases is identical, the ion-gauge pressure reading is not necessarily the same. Because ion gauges rely on electron impact ionization, the pressure reading depends on the ionization efficiency of the gas in the chamber. 9 Specifically, ion gauges are calibrated for air (N 2 ), so that the response of the gauge to another gas is relative to that for N 2 . The actual pressure P actual is related to the ion gauge pressure reading P ion by actual ion P SP = (B.4) where S is the sensitivity factor and equals 1 for both N 2 and Ar. 9 Unfortunately, the elec- tron impact ionization efficiency of Ir(ppy) 3 is unknown and prohibitively difficult to cal- culate. However, a crude approximation can be made with the intuitive realization that the ionization cross section will depend in large part on the ionization energy (IE). In 128 fact, empirical formulas derived for atoms support this assertion. 10 Because the IE of Ir(ppy) 3 is about half of the IE of argon and Ir(ppy) 3 is much larger than argon, the sensitivity factor for Ir(ppy) 3 is assumed to be 3 ( ) 1/ 2 0.5 Ir ppy Ar S S = = (Table B.2). For Ir(ppy) 3 , Eq. (B.1) can be evaluated using the values in Table B.2 and 3 3 3 ( ) ( ) ( ) exp ! k ion Ir ppy coll Ir ppy B ion k Ir ppy coll B P x L S k T P I x L S k k T σ γ σ γ         = −     (B.5) where 3 ) ( ppy Ir x accounts for the pick-up chamber to cell pressure conversion, which was determined using Eq. (B.3). Also, since capt σ for Ir(ppy) 3 is unknown, coll σ , which de- pends solely on the droplet size, is used. Consequently, Eq. (B.5) gives the collision prob- ability for k Ir(ppy) 3 . The set of Eq. (B.5) are displayed in Fig. 3.6 with P ion on the x-axis. There were three assumptions used in order to arrive at Eq. (B.5). The first two assumptions—that the pumping speed is independent of the gaseous species and the tem- perature of Ir(ppy) 3 does not significantly affect the pressure in the pick-up chamber— enabled the pressure of the argon system to be equated to the pressure of the Ir(ppy) 3 system. For turbomolecular pumps, the pumping speed is roughly the same for all gases. 9 Additionally, because of the high collision frequency with the walls of the vacuum cham- ber, 9 Ir(ppy) 3 is expected to lose some kinetic energy. Neither of these approximations is likely to produce a significant error in the final result; regardless, at 500 K the effect of this approximation is not large. The final assumption is for the Ir(ppy) 3 sensitivity factor S. The ionization efficiency is likely to be larger for Ir(ppy) 3 than for argon. However, the 129 actual value of S is unclear. Despite these approximations, the error introduced to the data presented in Fig. 3.6 is small enough that the main results are still valid. 130 Table B.2 Constants used for Eq. (B.2) - (B.5) x Ar / Ir(ppy)3 130 c 679 a γ ~ <v rel > / v droplet 6 b The capture cross section depends on the atomic or molecular species. For argon, σ coll ~ σ capt 6 c Obtained from data fitting d σ capt has not been determined for Ir(ppy) 3 . e This is a conservative estimate. The actual IE is 6 - 7 eV ( Chapter 4) IE (eV) 15.76 7 e S 1 0.5 A s (cm 2 ) π π σ capt (Å 2 ) 6700 b n / a d γ a 1.58 1.08 <v rel > (m / s) 556 376 v rms (m / s) 432 138 L (cm) 4 1 T cell (K) 300 500 Ar Ir(ppy) 3 131 Appendix C Ir + Action Spectrum 34950 35100 35250 35400 35550 35700 0.0 0.5 1.0 1.5 2.0 2.5 Ir + Ion Signal hv / cm -1 Figure C.1 Details of the Ir + action spectrum shown in the top panel of Fig. 4.10, recorded with an energy fluence of 130 mJ / cm 2 . Table C.1 Ir + Action Spectrum Assignment this work Ref. (11) this work transition transition Initial Initial State Ionization hv (cm -1 ) hv (cm -1 ) State a Energy (cm -1 ) Mechanism b 34917 34918.1 4F 5/2 9878 1+1 34935 34934.7 2P 3/2 10579 1+1 34934.7 4P 1/2 20237 1+1 34936.3 4F 5/2 5785 1+1 34956 34956.9 2D 5/2 12219 1+1 132 Table C.1 Continued this work Ref. (11) this work transition transition Initial Initial State Ionization hv (cm -1 ) hv (cm -1 ) State a Energy (cm -1 ) Mechanism b 34974 c 35001 35002.5 2P 3/2 10579 1+1 35008 c 35014 35014.9 4P 5/2 16103 1+1 35016.2 4P 3/2 18547 1+1 35026 c 35036 c 35043 c 35046 35047.0 4F 9/2 2835 1+2 35068 35068.9 2F 5/2 19061 1+1 35073 35075.0 2G 9/2 23506 1+1 35088 35084.6 2H 11/2 19593 1+1 35092 35091.1 4F 9/2 0 1+2 35097 c 35103 c 35107 c 35116 c 35129 35129.1 2F 7/2 13088 1+1 35129.1 2G 9/2 23506 1+1 35146 c 35150 c 35165 c 35171 c 35183 35084.6 4F 7/2 7107 1+1 35190 35190.3 4F 3/2 11831 1+1 35193 c 35197 c 35208 35208.6 4F 7/2 6324 1+1 35220 35220.6 4F 3/2 4079 1+1 35221.7 2F 7/2 13088 1+1 35228 c 35233 c 133 Table C.1 Continued this work Ref. (11) this work transition transition Initial Initial State Ionization hv (cm -1 ) hv (cm -1 ) State a Energy (cm -1 ) Mechanism b 35244 35244.4 4F 5/2 9878 1+1 35256 35255.9 4F 3/2 4079 1+1 35259.7 4P 3/2 16565 1+1 35264 35265.2 4P 5/2 12952 1+1 35270 35269.6 2F 5/2 19061 1+1 35278 c 35291 c 35300 c 35308 c 35319 35318.7 4F 5/2 9878 1+1 35334 35333.6 4F 9/2 2835 1+1 35340 35340.5 2D 5/2 12219 1+1 35370 35371.1 2F 7/2 13088 1+1 35383 35383.0 4F 3/2 11831 1+1 35391 35391.9 4F 5/2 9878 1+1 35406 35405.1 4F 9/2 2835 1+1 35423 35421.1 4F 9/2 0 1+2 35428 35428.8 4P 3/2 16565 1+1 35431 c 35439 35438.0 2D 3/2 22110 1+1 35454 35453.0 2H 11/2 19593 1+1 35499 35496.7 4P 3/2 16565 1+1 35499.6 4P 5/2 12952 1+1 35508 35507.2 4P 5/2 12952 1+1 35515 35515.8 2F 5/2 19061 1+1 35526 35525.5 2P 3/2 10579 1+1 35531 35531.2 2H 11/2 19593 1+1 35552 35551.8 2F 7/2 13088 1+1 35583 c 35589 35589.1 2F 5/2 19061 1+1 35601 c 35617 35616.9 2D 5/2 12219 1+1 134 Table C. 1 Continued this work Ref. (11) this work transition transition Initial Initial State Ionization hv (cm -1 ) hv (cm -1 ) State a Energy (cm -1 ) Mechanism b 35631 35632.9 4P 1/2 16681 1+1 35637 35637.3 2D 3/2 22110 1+1 35645 c 35662 35660.8 2F 5/2 19061 1+1 35669 c 35682 c 35685 c 35689 c 35704 35703.8 4F 5/2 9878 1+1 35717 35716.5 4F 3/2 11831 1+1 35717.6 4P 1/2 16681 1+1 35722 35721.7 4P 5/2 16103 1+1 c 6F 7/2 62018 1+1 35727 35725.1 3G 7/2 26365 1+1 35744 35743.6 4F 9/2 2835 1+1 a For details on state assignments see Ref. (11). b Total number of photons required for ionization. See Sect. 4.5.4. c Unassigned transition. 135 Appendices References 1. Harms, J.; Toennies, J. P.; Dalfovo, F. Phys. Rev. B 1998, 58, 3341. 2. Buchenau, H.; Knuth, E. L.; Northby, J.; Toennies, J. P.; Winkler, C. J. Chem. Phys. 1990, 92, 6875. 3. Zadorozhnyy, A. Energy Transfer Pathways for NO 2 -Rare Gas Complexes in Helium Droplets. Master Thesis, University of Southern California, Los Angles, CA, 2009. 4. Callicoat, B. E.; Forde, K.; Ruchti, T.; Jung, L.; Janda, K. C.; Halberstadt, N. J. Chem. Phys. 1998, 108, 9371. 5. Hartmann, M.; Miller, R. E.; Toennies, J. P.; Vilesov, A. J. Science 1996, 272, 1631. 6. Lewerenz, M.; Schilling, B.; Toennies, J. P. J. Chem. Phys. 1995, 102, 8191. 7. Toennies, J. P.; Vilesov, A. F. Angew. Chem. Int. Ed. 2004, 43, 2622. 8. Ramsey, N. F. Molecular Beams; Oxford University Press: New York, NY, 1985. 9. Moore, J. H.; Davis, C. C.; Coplan, M. A. Building Scientific Apparatus, 3 rd ed.; Perseus Books: Cambridge, MA, 2003. 10. Mark, T. D.; Dunn, G. H. Electron Impact Ionization; Springer-Verlag / Wien: New York, NY, 1985. 11. Van Kleef, T. A. M. Physica 1957, 23, 843.

Abstract (if available)

Abstract Tris(2-phenylpyridine)iridium (Ir(ppy)3) is a popular green phosphor used in organic light emitting diodes (OLEDs). In these applications, strong spin-orbit coupling induced by the iridium atom is responsible for this molecule’s high electroluminescent quantum efficiency. A comprehensive understanding of the electronic properties of Ir(ppy)3 is desirable in this setting. Such knowledge may lead to the smart design of OLED phosphorescent dopants. To achieve this goal, helium droplet isolation and multiphoton ionization (MPI) spectroscopies were applied. The helium droplet study led to important findings regarding the helium droplet capture process: helium droplets containing, on average, 8000 helium atoms are unable to efficiently capture vibrationally excited Ir(ppy)3. Gas-phase MPI revealed competing ionization and dissociation channels. Nevertheless, the parent Ir(ppy)3+ ion was isolated and the ionization energy was determined to be < 6.55 eV. This work is complimented by time-dependent density functional theory calculations.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A multifaceted investigation of tris(2-phenylpyridine)iridium

PDF

Energy transfer pathways for NO2-rare gas complexes in helium nanodroplets

PDF

Infrared spectroscopy of carbocations in helium droplets

PDF

Location, ionization and shielding of impurities in helium nanodroplets

PDF

Infrared and Raman spectrosopy of molecules and molecular aggregates in helium droplets

PDF

Rovibrational spectroscopy in helium droplets

PDF

Atomic and molecular clusters at ultra-low temperature

PDF

Electronic structure of ionized non-covalent dimers: methods development and applications

PDF

Photophysical properties of luminescent iridium and coinage metal complexes

PDF

The effect of microhydration on ionization energy and proton transfer in nucleobases: analysis and method development

PDF

Studies of molecular orientation using iridium phosphors and integration of corannulene into organic light emitting diodes (OLEDs)

PDF

Electronic structure and spectroscopy of excited and open-shell species

PDF

Spectroscopy of hydrogen and water clusters in helium droplets

PDF

Synthesis and photophysical study of phosphorescent hetero-cyclometalated organometallic complexes involving phosphino-carbon ligands

PDF

Photoelectron and ion imaging investigations of spectroscopy, photoionization, and photodissociation dynamics of diazomethane and diazirine

PDF

Hydrogen transfer reactions catalyzed by iridium and ruthenium complexes

PDF

Photoexcitation and nonradiative relaxation in molecular systems: methodology, optical properties, and dynamics

PDF

Magnetization study of two dimensional helium three

PDF

Imaging the state-specific vibrational predissociation of hydrogen bonded complexes

PDF

New electronic structure methods for electronically excited and open-shell species within the equation-of-motion coupled-cluster framework

Asset Metadata

Creator Nemirow, Christopher M. (author)

Core Title Multiphoton ionization of tTris(2-phenylpyridine)Iridium

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Chemistry

Publication Date 06/13/2011

Defense Date 04/29/2011

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

Tag helium droplets,ionization,ionization energy,Ir(ppy)3,Multiphoton,OAI-PMH Harvest,OLED

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wittig, Curt (committee chair), Haas, Stephan (committee member), Krylov, Anna I. (committee member)

Creator Email cmnemirow@gmail.com,cmnemirow@hotmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c127-615915

Unique identifier UC1368915

Identifier usctheses-c127-615915 (legacy record id)

Legacy Identifier etd-NemirowChr-21.pdf

Dmrecord 615915

Document Type Dissertation

Rights Nemirow, Christopher M.

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

Repository Email cisadmin@lib.usc.edu