Efforts Toward Diindenoarene Radicaloids: A Reach for Mixed Electronic Configurations and The Properties They Manifest by Efrain Vidal A dissertation accepted and approved in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Chemistry Dissertation Committee: Darren W. Johnson, Chair Michael M. Haley, Advisor Ramesh Jasti, Core Member Matthew Polizzotto, Institutional Representative University of Oregon Spring 2025 2 © 2025 Efrain Vidal This work is openly licensed via CC BY-NC-ND 4.0. 3 DISSERTATION ABSTRACT Efrain Vidal Doctor of Philosophy in Chemistry Title: Efforts Toward Diindenoarene Radicaloids: A Reach for Mixed Electronic Configurations and The Properties They Manifest Semiconductors are ubiquitous in modern electronic devices, and their market serves an important role in the global economy. Recent advancements have increasingly incorporated organic materials, such as polycyclic conjugated molecules, due to the tunability of their optoelectronic properties. Antiaromatic molecules are a class of organic compounds that inherently have narrow HOMO/LUMO energy gaps, strong low energy absorption and often exhibit ambipolar charge transport behavior. Radicaloids have also garnered attention as promising organic materials because they have those same desirable properties and offer an additional degree of freedom with magnetic susceptibility. This dissertation will highlight key strategies to tune antiaromaticity and diradical character in polycyclic molecules, in addition to discussing refinement of diradicaloid structures and preparation for use in single-molecule junctions. The design of elongated radicaloids and novel photoactive molecules will cap this exposition to inspire continued pursuit of polycyclic conjugated molecules. Chapter I provides perspective on the research accomplished in the Haley lab towards tuning antiaromaticity via arene annulation onto s-indacene and elucidating the electronic parameters dictating the diradical character of indenofluorene analogues. Chapter II explores the differences in molecular properties between alkyne-substituted and alkyne-free diindeno[1,2- b;1′,2′-i]anthracenes. Simplified diindenoanthracene from Chapter II allows for fair comparison between π-expanded indenofluorene derivatives validating the single-molecule conductance 4 studies detailed in Chapter III. The ideas and synthetic efforts contained in Chapter IV capitalize on indenofluorene design principles to establish a few directions for the development of polyradicaloids and novel molecular motors. This dissertation includes previously published and unpublished co-authored material. 5 CURRICULUM VITAE NAME OF AUTHOR: Efrain Vidal GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene University of Southern California, Los Angeles DEGREES AWARDED: Doctor of Philosophy, Chemistry, 2025, University of Oregon Bachelor of Science, Environmental Studies, 2019, University of Southern California AREAS OF SPECIAL INTEREST: Organic Chemistry Polycyclic Antiaromatic Hydrocarbons Open-Shell Character Organic Semiconductors and Topological Insulators Physical Organic Chemistry PROFESSIONAL EXPERIENCE: Graduate Employee, University of Oregon, 2019–2025 Instructor of Record, General Chemistry I, Winter 2024 PUBLICATIONS: Warren, G. I.; Barker, J. E.; Vidal Jr., E.; Haley, M. M. The Interplay Between Antiaromaticity and Diradical Character in Diarenoindacenes and Diindenoarenes. Chem. Sci., in preparation (invited ‘Perspective’ article). Lawson, B.; Vidal Jr., E.; Haley, M. M.; Kamenetska, M. Extreme Anomalous Conductance Enhancement in Neutral Diradical Acene-like Molecular Junctions. ACS Nano 2024, 18, 29059–29066. Vidal Jr., E.; Zakharov, L. N.; Gómez-García, C. J.; Haley, M. M. Probing the Influence of Alkyne Substitution on the Electronic and Magnetic Properties of Diindeno[1,2-b;1’,2’- i]anthracenes. J. Org. Chem. 2024, 89, 14515–14519. 6 Zhang, D.; Beverly, E. J.; Levin, N. E.; Vidal Jr., E.; Matia, Y.; Feakins, S. J. Carbon Isotopic Composition of Plant Waxes, Bulk Organics and Carbonates From Soils of the Serengeti Grasslands. Geochim. Cosmochim. Acta 2021, 311, 316–331. Xiuwei, L.; Feakings, S. J.; Ma, X-F; Anderson, J. D.; Vidal Jr., E.; Blancaflor, E. B. Crop Breeding Has Increased the Productivity and Leaf Wax n-Alkane concentration in a Series of Five Winter Wheat Cultivars Developed Over the Last 60 Years. J. Plant Physiol. 2019, 243, 153056. 7 ACKNOWLEDGMENTS Firstly, I would like to thank my advisor Prof. Michael M. Haley for providing this opportunity to work in his lab during my time in graduate school. Mike let me develop my synthetic skillset while learning all about neat but funky π-systems. The National Science Foundation (CHE-1954389 and CHE-2246964) funded my efforts in the Haley lab, which I appreciate. As for other key faculty who helped shape my scientific development, my committee chair Prof. Darren W. Johnson provided space for me to ask scientific questions and talk about life’s stressors. Prof. Ramesh Jasti, who served as a core member of my committee, pushed me to not only critically think about my research but also the stories I construct to share my work with others. Keeping a distant eye on these antics was Prof. Matt Polizzotto who served as my institutional representative. I also thank the departmental administrative staff for keeping the whole operation in working order. Funnily enough, I started this PhD. program with the attitude “I came to get my PhD., not to make friends”; however, I quickly learned that the people you meet can be the best part of the experience. There is literally never a dull moment with Dr. Arman Garcia. I enjoyed popping into the Pluth lab office and hearing a random screech or your latest rant. I would like to thank Dr. Kaylin G. Fosnacht for all the anime/movie nights, bouldering sessions, and other antics. You also provided sensible perspectives and explanations for all the things I did not really understand about people—truly a great housemate and friend. Dr. Isam D. Moore, you shared many of your random statistics, but your competitive spirit and rivalry with Dr. John N. McNeill were hilarious to witness. Rachel Rocha and Kyra Brakefield, you both also aided in making this period in my life enjoyable during group activities. There are many others I have befriended in the department, and I thank you all for the chats and activities during my stay. 8 The Haley lab, old and new, is a group of people I have spent a lot of time with over the past several years. Saying thank you does not really convey how much I appreciate and cherish our time in the bullpen and around Eugene. My mentors Dr. Joshua E. Barker, Prof. Jeremy P. Bard, and Dr. Justin J. Dressler were the most welcoming group of people when I first spawned in the lab during my rotation and instrumental in getting me through early phases of the program. Dr. Hannah Bates was quick to blow my shell wide-open too. Thank you all for facilitating my transition into graduate school and grounding me right off the bat. Dr. Gabrielle I. Warren, Dr. John N. McNeill, Michael P. Miller, Megan L. Rammer, and Nathan J. Boone…what an incredible group of people to hang out with every day and share the highs and lows of research. My entire family, the people who have been cheering me on throughout my whole life, was constantly supportive and the best group to return to during breaks despite all the happy chaos that comes with them. I do not even know how to express how thankful I am to have them in my corner, but I am always grateful for their existence. Lastly, Joey Crouch has been an incredible addition to my life this past year. Balancing life and research has been challenging, but he has been so supportive and understanding during this final stretch. While we are still in our infancy, I think we have a good thing brewing. 9 To those who believed in me at any step in this educational journey. 10 TABLE OF CONTENTS Chapter Page I. THE INTERPLAY OF ANTIAROMATICITY AND DIRADICAL CHARACTER IN DIARENOINDACENES AND DIINDENOARENES ............................................... 19 1.1 Introduction ...................................................................................................... 19 1.1.1 Synthetic strategies ................................................................................. 26 1.2 Initial Studies ................................................................................................... 30 1.3 Tuning the paratropicity strength of s-indacene by outer arene modification ........................................................................................................... 33 1.4 Computational techniques ................................................................................ 37 1.5 NMR spectroscopy........................................................................................... 40 1.6 Electrochemical and photophysical properties ................................................ 42 1.7 X-ray crystallography ...................................................................................... 44 1.8 Dicyclopenta[b,i]anthracene: the core to achieve strong open-shell character ................................................................................................................. 47 1.8.1 Calculated singlet-triplet energy gaps and electronic parameters ........... 50 1.8.2 SQUID magnetometry ............................................................................ 52 1.9 Caught in the middle: DCN based molecules straddle closed-shell and open-shell structures .............................................................................................. 53 1.9.1 Breaking open diradical character in fluorenofluorenes ......................... 54 1.10 What IF we turned to devices? ....................................................................... 58 Conclusions ............................................................................................................ 61 11 Chapter Page II. PROBING THE INFLUENCE OF ALKYNE SUBSTITUTION ON THE ELECTRONIC AND MAGNETIC PROPERTIES OF DIINDENO[1,2-b;1’,2’-i] ANTHRACENES ........................................................................................................ 62 2.1 Introduction ...................................................................................................... 62 2.2 Results and Discussion .................................................................................... 66 2.3 Conclusions ...................................................................................................... 70 III. EXTREME ANOMALOUS CONDUCTANCE ENHANCEMENT IN NEUTRAL DIRADICAL ACENE-LIKE MOLECULAR JUNCTIONS ...................................... 71 3.1 Introduction ...................................................................................................... 71 3.2 Results and Discussion .................................................................................... 74 3.3 Experimental Methods ..................................................................................... 85 3.4 Conclusions ...................................................................................................... 85 IV. A QUEST FOR EXTENDED SCAFFOLDS AND NEW DIRECTIONS ........... 87 4.1 Introduction ...................................................................................................... 87 4.2 Synthesis Toward Quinoidal Ladders .............................................................. 91 4.3 Synthesis Toward Covalent Dimers................................................................. 93 4.4 Synthesis Toward Molecular Motors ............................................................... 94 4.5 Final Remarks .................................................................................................. 95 APPENDICES ............................................................................................................. 96 A. SUPPLEMENTARY INFORMATION FOR CHAPTER II ............................ 96 12 Chapter Page B. SUPPLEMENTARY INFORMATION FOR CHAPTER III ........................... 110 C. SUPPLEMENTARY INFORMATION FOR CHAPTER IV .......................... 154 REFERENCES CITED ................................................................................................ 197 13 LIST OF FIGURES Figure Page 1. Figure 1.1. Two popular methods to tame antiaromatic molecules. ..................... 21 2. Figure 1.2. Depiction of frontier molecular orbitals for diradical 5 (two degenerate orbitals) and diradicaloid 6 (two nearly degenerate orbitals). ............. 23 3. Figure 1.3. Isomeric (a) indacenes 7-8 and (b) indenofluorenes 9-13; the indacene core is highlighted in red in the latter examples. The computed values of y0 for the parent hydrocarbons are taken from reference 26. (c) Closed-shell (CS) and open-shell (OS) resonance forms of IF 9. ............................................... 25 4. Figure 1.4. General synthetic strategy for the preparation of indenofluorenes and quinoidal analogues. ........................................................................................ 27 5. Figure 1.5. Generic carbocycle- or heterocycle-fused s-indacene with labelled apical carbons; syn-fusion is highlighted in blue and anti-fusion is highlighted in red. ..................................................................................................................... 29 6. Figure 1.6. Early attempts to alter IF optoelectronic properties. .......................... 31 7. Figure 1.7. Modification of the [1,2-b]IF scaffold follows two main lines of investigation—outer π-extension/ alteration with aromatic carbocycles or heterocycles (diareno-s-indacenes, top) and core π-extension from benzenoid to naphthoquinoid and anthraquinoid (diindenoacenes, bottom). .......................... 32 8. Figure 1.8. Aromatic carbocycle-fused s-indacene derivatives—DNIs 18-20, DPI 21, and DAI 22 ............................................................................................... 33 9. Figure 1.9. (a) Heterocycle-fused s-indacene core antiaromatics grouped by family of heterocycle: thiophenes/ benzothiophenes 23-26 (blue), benzothiophene sulfones 27-28 (yellow), benzofurans 29-30 (orange), naphthothiophenes 31-36 (green), indoles 37-38 (pink), and mixed benzofurans/benzothiophenes 39-42 and their corresponding sulfones 43-46 (purple). (b) syn-Heterocycle-fused s-indacenes in order of increasing antiaromaticity. ...................................................................................................... 35 14 Figure Page 10. Figure 1.10. (a) Example path of the dummy atoms in NICS-XY(1.7)πZZ scan calculations. NICS-XY scans of (b) carbocycle-fused systems (10 and 18-22) and(c) heterocycle-fused systems (25-30 and 37-38) versus that of s-indacene reference molecule 7. (d) NICS-XY scans of only the syn-fused heterocycles in Fig. 9b. All NICS scan plots are ordered from most to least antiaromatic. NICS-XY scans performed at1.7 Å above the ring system using the M11/6-311+G** level of theory. .................................................................... 37 11. Figure 1.11. Molecular structure of syn-IDI 37 (CCDC 2416146; ellipsoids drawn at the 50% probability level and hydrogen atoms are omitted for clarity). ............................................................................................................. 46 12. Figure 1.12. Continuum of core PAAHs from antiaromatic to diradicaloid featuring sequential insertion of 6-membered rings, i.e., benzinterposition. Pro- aromatic p-QDM, 2,6-naphthoquinodimethane (NQD) and 2,6-anthroquino- dimethane (AQD) fragments are highlighted. ....................................................... 47 13. Figure 1.13. CS and OS resonance structures of DIAn 47 depicting transition from singlet to triplet states alongside DIAn homologues 48-52. Pro-aromatic AQDfragments are highlighted.. ........................................................................... 49 14. Figure 1.14. Bleaney-Bowers fits of heating cycle SQUID data for the DIAn series 48-52; traces are shown only up to 400 K for consistency. ......................... 52 15. Figure 1.15. FF homologues 53-61 and 63-64 along with decomposition products 62 and 65. NQD substructures are highlighted while linear and cross conjugation patterns are bolded in heterocycle fused derivatives according to Figure 1.5. .......................................................................................................... 54 16. Figure 1.16. Bleaney-Bowers fits of recorded SQUID data for FF homologues 53-59, 61, and 64 ................................................................................................... 57 17. Figure 1.17. Chemical structures of quinoidal derivatives used in (a) OFET evaluation and (b) single molecule conductance measurements. .......................... 60 18. Figure 2.1. Structures of the known DIAn family 1–5 and new DIAn derivatives 6 and 7 lacking the alkyne groups. The common 2,6- anthroquinoidal/diyl relationship in all molecules is denoted by the bolded bonds. ................................................................................................... 63 15 Figure Page 19. Figure 2.2. (a) Electronic absorption spectra for 1, 6, and 7 in CH2Cl2 at room temperature. (b) CV of 1, 6 and 7 in CH2Cl2 showing two quasi-reversible one-electron reductions and two quasi-reversible one-electron oxidations. Voltammograms are generated according to IUPAC plotting convention; the starting point for each voltammogram is 0.0 V and scans proceed from oxidizing to reducing potentials. ............................................................................ 67 20. Figure 2.3. (a) Crystal structure of DIAn 6 drawn with 50% thermal ellipsoids, hydrogen atoms are omitted for clarity. (b) Comparison of selected bond lengths (Å) of 1 (brown)35 and 6 (blue). ................................................................ 68 21. Figure 2.4. SQUID data of 1, 6, and 7 (empty circles) and the corresponding Bleaney-Bowers fit (solid lines). ........................................................................... 69 22. Figure 3.1. (a) Molecular structures of the BQ and BTQ molecules; calculated y0 values (PUHF) are from ref. 21. (b) Resonance structures converting between the quinone, delocalized, and diradical form of the family of benzo- fused quinoids (BQn) where n = 1-3. Thioanisole linkers (ThA = 2,6-dimethyl -4-(methylthio)phenyl) serve to anchor the molecule to gold electrodes in the junction. ....................................................................................................... 73 23. Figure 3.2. (a) Single-molecule conductance histograms of BQ1-3 in BNP at low bias, –100 mV, in the top panel and high bias, –833 mV for BQ1-2 and –250 mV for BQ3, in the bottom panel. (b) Single-molecule 2D conductance vs displacement histograms of BQ1-3 at low bias in the top three panels and high bias in the bottom three panels. (c) Calculated conductance from Gaussian fits to the peaks in Figure 2a for both low bias (black) and high bias (gray) with exponential fits given by the dashed lines. ............................................................ 76 23. Figure 3.3. (a) Current-Voltage histograms for BQ1-3 from thousands of IV curves taken while holding each molecule between two Au electrodes. Dotted lines correspond to the linear grid lines and the white lines are calculated from the average of gaussian fits of vertical slices of each histogram IV histogram. (b) Fits of the conductance-voltage histograms derived from the current- voltage histograms for BQ1-3 and BTQ1-2........................................................... 78 24. Figure 3.4. Modification to the 1D-SSH Model for polycyclic molecular systems like BQn. (a) Description of the coupling energies (t0, t1, t2, u1, u2, c1, c2) describe by diradical parameter (δ) and scaling parameter (a) used in the modified 1D SSH Hamiltonian. (b) Graphical representation of the energies t1, t2, u1, u2. (c) The calculated band structure for BQ1-3 as a function of δ. .............. 79 16 25. Figure 3.5. Extreme anti-ohmic behavior of acene-like systems is due to quantum interference effects. (a) Calculated transmission at E-EF = 0 eV of BQn and Cn series as a function of diradical character. (b) Transmission curves for BQn (top) and Cn series (bottom) for the case δ = 0. (c) Contribution of constructive quantum interference (CQI, dotted) and destructive quantum interference (DQI, dashed) to the total transmission for BQn and Cn series as a function of δ. ....................................................................................................... 81 26. Figure 3.6. Engineering enhancement in anti-ohmic conductance through synthetic control. Single-molecule conductance histograms of BTQ1-2 (structures shown in Figure 1a) in BNP at –500 mV. Inset: most likely conductance of BTQ1-2 determined from the histograms plotted against radical- radical (R-R) distance results shows inverse decay and a β value of –2.3/Å. ....... 84 27. Figure 4.1. Molecular structures of a) reported tetraradicaloid examples with corresponding y0 and y1 values and b) proposed tetraradicaloids based on AQD and NQD with NOON analysis for 5–9. Functional and basis sets are included as these vary among research groups which could invalidate comparisons. Pro- aromatic units are highlighted in green. ................................................................. 88 28. Figure 4.2. Selection of a) reported TIBS-(ref. 39) and TIPS (ref. 40)- pentacene dimers featuring varying π-spacers and b) target fluorenofluorene (FF) dimers isomers. .............................................................................................. 90 17 LIST OF TABLES Table Page 1. Table 1.1. 1H NMR chemical shift of the core protons of select s-indacene derivatives .............................................................................................................. 41 2. Table 1.2. Electrochemical and photophysical values of select s-indacene derivatives .............................................................................................................. 43 3. Table 1.3. Core C–C bond lengths (Å) of select s-indacene derivatives .............. 45 4. Table 1.4. Electronic Parameters and Singlet-Triplet Energy Gap for DIAn derivatives 47-52 .................................................................................................... 52 5. Table 1.5. Electronic Parameters and Singlet-Triplet Energy Gap for FF derivatives 53-63 .................................................................................................... 55 18 LIST OF SCHEMES Scheme Page 1. Scheme 2.1. Synthesis of diindeno[1,2-b:1’,2’-i]anthracenes 6 and 7. ................. 65 2. Scheme 4.1 Synthesis of a) common asymmetric mono-triflate intermediates and b) π-ladder 4. ................................................................................................... 91 3. Scheme 4.2 Synthesis of bis-IIDBT 5. .................................................................. 92 4. Scheme 4.3 Synthesis of m-dimer 12. ................................................................... 93 5. Scheme 4.4 Alternate synthesis of m-dimer 44. .................................................... 94 6. Scheme 4.5 Synthesis of key intermediates 47 and 50. ......................................... 95 19 CHAPTER I THE INTERPLAY OF ANTIAROMATICITY AND DIRADICAL CHARACTER IN DIARENOINDACENES AND DIINDENOARENES This chapter includes unpublished and co-authored material intended to be submitted as a Perspective article in Chemical Science by Dr. Gabrielle I. Warren, Dr. Joshua E. Barker, Efrain Vidal Jr., and Prof. Michael M. Haley. This manuscript titled The Interplay of Antiaromaticity and Diradical Character in Diarenoindacenes and Diindenoarenes was written by Dr. Gabrielle I. Warren with help from Dr. Joshua E. Barker, Efrain Vidal Jr., and Prof. Michael M. Haley 1.1 Introduction Aromaticity and antiaromaticity are fundamental concepts in organic chemistry that still sustain considerable discussion and research within the community.1 The [4n+2] π-electron rule, proposed by Hückel in 1931,2 continues to be taught and permits quick identification of aromatic species. Breslow and Dewar later postulated the [4n] π-electron rule for systems which they coined the term “antiaromatic”.3 Antiaromaticity is the “opposite” of aromaticity, in that the ring current (in the presence of a magnetic field) is diatropic (stabilizing) for aromatic molecules and paratropic (destabilizing) for antiaromatic molecules. Antiaromatic molecules are of fundamental interest due to their extremely reactive nature, making them nearly impossible to isolate in their unsubstituted form.4 As a result, antiaromatic molecules find pathways to relieve this destabilization. For example, cyclooctatetraene (COT) adopts its well-known, nonplanar tub shape, and cyclobutadiene undergoes rapid dimerization. 20 Polycyclic antiaromatic hydrocarbons (PAAHs), compounds that contain one or more antiaromatic motifs as part of the overall hydrocarbon backbone, tend to be the black sheep of π- conjugated molecules and materials. Unlike the well-studied polycyclic aromatic hydrocarbons (PAHs) that often feature multiple stabilizing, aromatic [4n+2] π-electron pathways,5 inclusion of the antiaromatic [4n] π-electron unit(s) very often results in significant destabilization of the resultant molecule. Nonetheless, interest in PAAHs has seen a strong resurgence over the last 15 years because of the promising optoelectronic properties that are inherent to these compounds, such as small HOMO/LUMO energy gaps, low HOMO and LUMO energy levels, redox amphoterism, and increased conductance; thus, such molecules could potentially find use in device applications such as organic field-effect transistors (OFETs) or organic solar cells (OSCs).6 Fortunately, chemists have devised ways to tame highly reactive antiaromatic molecules. For example, when Hafner attempted in the early 1970s to prepare the parent pentalene 1 (Figure 1.1), an 8 π-electron analogue of COT, the molecule rapidly dimerized into 2 even at –80 °C to alleviate the destabilizing paratropic ring current.7 Bally et al. showed that this process, which is thought to occur via diradical intermediates and thus indicative of the presence of a low-lying triplet state, is reversible upon photolysis of 2 in an Ar matrix, with the photocleavage occurring in two distinct steps–formation first of a diradical then regeneration of 1.8 To obtain an isolable hydrocarbon, Hafner subsequently installed three t-butyl groups onto the pentalene skeleton, e.g., 3, thus kinetically stabilizing the molecule.9 Kinetic stabilization with bulky substituents was a very popular method in the 1970s and 1980s to generate a variety of stable, isolable antiaromatic compounds, including those with cyclobutadiene and indacene cores.4 21 Another widely-used strategy to tame paratropic compounds is to fuse one or more aromatic rings, such as in dibenzopentalene 4, a PAAH first reported in 1912, as such fusion results in a thermodynamically more stable hydrocarbon.10 This stabilization comes with a cost, however, as arene fusion weakens the paratropic ring current significantly, which results in an increased HOMO/LUMO energy gap and higher HOMO and LUMO energy levels compared to the parent antiaromatic hydrocarbon. Nonetheless, Kawase and Takimiya showed in 2010 that derivatized dibenzopentalenes could be used as an active layer in OFETs that showed modest hole mobilities (1.8 x 10–3 cm2 V–1 s–1) as well as be the electron-donor layer in OSCs (PCE = 0.94%), clearly demonstrating the materials potential of PAAHs.11 Very recently, a combination of steric blocking and extended conjugation with pendant aryl groups led to the isolation and thorough characterization of the 1,3,4,6-tetraphenyl derivative of 1.12 Figure 1.1. Two popular methods to tame antiaromatic molecules. Properties associated with antiaromaticity drive motivation to study these systems in their own right, but there is a unique relationship between the antiaromatic nature of a molecule and t-Bu t-Bu t-Bu kinetic stabilization thermodynamic stabilization 1 2 3 4 X 2 –80 °C hn, 20 K Br Et3N 22 irregular electronic configurations that bolsters the merit of investigating such species. A recent theoretical survey of antiaromatic [4n]annulenes conducted by Quintero et al. examined the role of frontier orbitals in a balancing act of stabilization from disjoint orbital character and inter- vs. intra-orbital electron-electron repulsions that favor unpaired electrons to varying degrees.13 For example, computationally forcing COT into a planar conformation permits investigation of this interplay between paired and unpaired configurations. The electronic structure of planar D8h COT featured unpaired spins residing in two degenerate singly occupied molecular orbitals (SOMOs), which was stabilized by the non-bonding and disjoint character of the frontier orbitals. The authors describe D8h COT as having an open-shell (OS) electronic configuration as opposed to one that is closed-shell (CS) based on these results. As this type of electronic stabilization can be extrapolated to other [4n]annulenes and their congeners via perturbation, antiaromatic molecules can be thought of as having some inherent diradical character on account of their frontier orbitals and configurational mixings.13 It is crucial to be clear and consistent regarding the description of diradical-like species. We will use the definition found in the IUPAC “Gold Book” for diradicals, which states that diradicals are “Molecular species having two unpaired electrons, in which at least two different electronic states with different multiplicities [electron-paired (singlet state) or electron-unpaired (triplet state)] can be identified”.14 Pure diradicals are described as molecules in which two electrons occupy two degenerate non-bonding molecular orbitals (e.g., m-xylylene/m- quinodimethane 5, Figure 1.2), while the term diradicaloids is used to identify diradical-like molecules in which the two molecular orbitals are nearly degenerate (e.g., p-xylylene/p- quinodimethane 6). Quantum chemical calculations are indispensable to any study of diradicaloids for determining the effects responsible for changes in diradical character. Standard 23 diradical calculations involve determination of the diradical index value (y0), which represents the occupation of the lowest unoccupied natural orbital (LUNO) according to natural orbital occupation number (NOON) analysis. In this scheme, y0 = 1 represents pure OS character while y0 = 0 represents a pure CS system.15 The magnitude of the value is used as an indicator for the likelihood that a molecule will exhibit diradical-like properties, and compounds with intermediate y0 values are predicted to be diradicaloids. It is important to note that y0 value calculations depend heavily on functional selection and as a result they are only reliably compared when the functional and basis set are the same. This theoretical understanding of novel π-systems provides an additional handle for the organic electronic materials as they share similar properties as CS antiaromatics (e.g., narrow HOMO/LUMO energy gaps and amphoteric redox behavior). Such amphoteric redox behavior is exemplified in bisphenalenyl diradicaloids that display both hole and electron mobilities up to 3 × 10–3 cm2 V–1 s–1.16 Figure 1.2. Depiction of frontier molecular orbitals for diradical 5 (two degenerate orbitals) and diradicaloid 6 (two nearly degenerate orbitals). Installing one six-membered ring between the two five-membered rings of pentalene, a process long-ago called “benzinterposition”,17 affords the tricyclic hydrocarbon known as indacene. Depending upon the symmetric or asymmetric fusion of the five-membered rings on the six-membered ring leads to s-indacene 7 and as-indacene 8 (Figure 1.3a).4,18 As one might SOMOs LUNO HONO Singlet DiradicaloidSinglet Diradical 5 6 (CS) 6 (OS) 24 expect, Hafner and co-workers investigated antiaromatic 7 as well, finding that inclusion of four t-butyl groups kinetically stabilized19 the otherwise highly reactive molecule.20 A number of similarly stabilized s-indacenes have been disclosed over the subsequent ~40 years.21 Very recent work by Tobe and co-workers showed that a series of stable hexaaryl-substituted derivatives of 7 could be readily prepared and their properties modulated by the aryl substituents.22 Interestingly, studies on 8, or simple derivatives thereof, have not been reported in the literature. In fall 2009, the Haley group initiated studies into s-indacene derivatives. Similar to Kawase and others, we elected to include benzo-fusion to help stabilize the paratropic core. Rather than “dibenzoindacene” or “diindenobenzene”, IUPAC nomenclature calls such structures indenofluorenes (IFs), of which there are five possible regioisomers (9-13) depending upon whether the structure contains an s- or as-indacene core, shown in red in Figure 3b. When we started our studies, only two of the five IF isomers had been described: LeBerre reported in 1957 the attempted synthesis of 11,12-diphenyl derivative of 11, which proved to be extremely unstable.23 In 1994 the Swager group disclosed the 5,6,11,12-tetraiodo analogue of 10 that also was prone to rapid decomposition.24 Two years later, Scherf and co-workers claimed the synthesis of 6,12-diphenyl analogue of 10 as a model system for an indenofluorene polymer;25 however, there were no experimental or characterization data in the paper aside from an absorption spectrum lmax of 543 nm to corroborate this structure.§ As of late 2009, there was no definitive evidence of stable, well-characterized indenofluorenes in the literature. The calculated y0 values of the IF regioisomers (Figure 1.3b) offer some insight as to why LeBerre and Swager had difficulty with their systems, as all IF isomers are predicted to possess modest (10, 11, 13) to pronounced (9, 12) diradicaloid character.26 This is especially easy to 25 understand for molecules 9 and 12, as in the open shell form there are three aromatic Clar sextets versus only one sextet in the closed shell form (shown in blue in Figure 1.3c), in addition to the Figure 1.3. Isomeric (a) indacenes 7-8 and (b) indenofluorenes 9-13; the indacene core is highlighted in red in the latter examples. The computed values of y0 for the parent hydrocarbons are taken from reference 26. (c) Closed-shell (CS) and open shell (OS) resonance forms of IF 9. contribution of 5 in the open-shell form of 9 and 12. For compounds 10, 11, and 13, the closed shell form possesses two Clar sextets as well as the presence of an o- or p-quinodimenthane core, which results in an appreciable drop in diradical character. Unfortunately, any regain of aromatic stabilization is offset by the introduction of radical reactivity at the apical carbons of the five- s-indacene as-indacene 2 1 a 7 8 9 10 11 121 2 3 4 5 6 2 1 b 7 8 910 1112 1 2 3 4 5 6 9 indeno[1,2-a]fluorene y0 = 0.80 7 8 10 indeno[1,2-b]fluorene y0 = 0.26 11 indeno[2,1-a]fluorene y0 = 0.29 2 1 a 7 8 9 1011121 2 3 4 5 6 2 1 c 7 8 9 10 11 121 2 3 4 5 6 12 indeno[2,1-b]fluorene y0 = 0.65 13 indeno[2,1-c]fluorene y0 = 0.25 2 1 b 7 8 9 10 11 12 1 2 3 4 5 6 (a) (b) 9 (CS) (c) 9 (OS) 26 membered rings. These positions too can be kinetically stabilized by the introduction of bulky aryl groups (e.g., mesityl) on the apical carbon atoms. Tobe and co-workers prepared 10,12- dimesitylindeno[2,1-b]fluorene in 2013 and confirmed via multiple techniques that the molecule, while possessing a singlet diradicaloid ground state, exhibited significant triplet character at room temperature.27 Haley and co-workers found in 2017 that the 7,12-dimesityl analogue of 9, predicted to be a ground state triplet, was so reactive that the molecule had to be analyzed as a dilute, degassed solution. While the spectral data implicated formation of the [1,2-a]IF, confirmation of its successful synthesis came about with the x-ray crystal structure of the dianion reduction product, confirming the correct skeletal connectivity.26 Until 2024 (vide infra), these were the only examples of the [1,2-a]IF and [2,1-b]IF isomers, so they will not be discussed further. The Tobe and Haley groups reported the syntheses of stable, isolable derivatives of the [2,1-a]IF28 and [2,1-c]IF29 regioisomers in 2011 and 2013, respectively, of which the spectral (NMR, UV-Vis) and structural (x-ray) data support the dominance of the closed-shell structure. Both isomers have been revisited by the Haley30 and Das31 groups in recent years; still, the number of [2,1-a] and [2,1-c]IFs is rather limited, so these too will not be discussed further. By far, the [1,2-b]IF scaffold, and its structurally-related, heterocycle-fused s-indacene congeners (vide infra), will be the main focus of the antiaromaticity studies described in this article. We refer the reader to several reviews with more information on the other IF regioisomers.18,32 1.1.1 Synthetic strategies Our studies of PAAHs began in fall 2009 as a rotation project for a first-year graduate student. We repeated the synthesis of the 5,6,11,12-tetraiodo analogue of 10 via double 27 transannular cyclization of octadehydrodibenzo[12]annulene;24 however, all attempts at Sonogashira cross-coupling degraded the IF starting material. Rather, we introduced the (triisopropylsilyl)ethynyl (TIPSethynyl) groups as two independent steps. Gratifyingly, this revised strategy worked, leading to the first well-characterized and stable indeno[1,2-b]fluorene derivatives (14);33 however, this route suffered from three major pitfalls: (1) the transannular cyclization afforded no more than a hundred milligrams of tetraiodo material per run, and often much less; (2) the yields for preparing the octadehydrodibenzo[12]annulene were also low;24 and (3) the [12]annulene was prone to violent decomposition, as we (re)discovered.34 An improved synthetic route for IF synthesis was developed based on work of Deuschel35 and Wang.36 This modular and scalable strategy has been and continues to be our main method for preparing a wide variety of IFs and related quinoidal analogues from strongly Figure 1.4. General synthetic strategy for the preparation of indenofluorenes and quinoidal analogues. 28 diradical to strongly antiaromatic and spanning a variety of carbocycle- and heterocycle-fused rings. The variations of the current syntheses, termed “inside-out” and outside-in”, share three key steps: a Suzuki cross-coupling, Friedel-Crafts acylation or alkylation, and finally an oxidative or reductive dearomatization of the central arene. The “inside-out” route starts with the desired core, which possesses two halides (or pseudohalides) and either two esters or two aldehydes. The core is then Suzuki cross-coupled to outer arenes, which are typically functionalized with a boronic acid or boronate ester. If the resultant para-substituted core has two ester groups, saponification followed by Friedel-Crafts acylation yields the poorly soluble diketone precursor to the desired IF derivative (route I). Nucleophilic addition of bulky aryl or ethynyl groups, either by lithiate addition or Grignard addition, gives the penultimate diol precursor. Subsequent reductive dearomatization using SnCl2 yields the desired IF (Figure 1.4, top entry). If the Suzuki cross-coupling furnished a para-substituted core with two aldehydes (route II), nucleophilic addition of the bulky aryl groups (by lithiate addition or Grignard addition) followed by Friedel-Crafts alkylation via the resultant secondary alcohol gives the dihydro precursor. Finally, oxidative dearomatization with DDQ yields the desired IF (Figure 1.4, second entry). The “outside-in” method follows the same steps as the “inside-out” but starts with the functional handles in reversed positions. This method only requires the core to have para- substituted boronic esters/acids or while the outer arene cross-coupling partner bears the carbonyl functionality. After the Suzuki cross-coupling, the two “outside-in” routes parallel the “inside-out” methods – nucleophilic addition of bulky aryl groups, Friedel-Crafts alkylation, and oxidative dearomatization with DDQ (Figure 1.4, third entry) or Friedel-Crafts acylation, nucleophilic addition of the aryl groups, and reductive dearomatization using SnCl2 (Figure 1.4, 29 last entry). While the “outside-in” route does lead predominantly to the formation of the [1,2- b]IF core, the Friedel-Crafts reaction can “close” the wrong way to produce a very small amount of the [2,1-a]IF isomer.30b Given the proper functionalization on the core and corresponding coupling partners, these two general methods provide ready access to a wide range of indenofluorene and related quinoidal topologies, allowing us to explore both core π-extension and outer arene modification. Figure 1.5. Generic carbocycle- or heterocycle-fused s-indacene with labelled apical carbons; syn-fusion is highlighted in blue and anti-fusion is highlighted in red. As the new synthetic strategies were employed for increasingly complex systems, new naming practices became important to adopt. When fusing a simple benzene to s-indacene (as in the parent IFs), there are only two possible isomers (e.g., [1,2-b]IF vs. [2,1-b]IF). Extending the parent [1,2-b]IF by one benzene on each side introduces three possible isomers, which were named linear-, syn-, and anti-, according to the direction of the angular naphthalene fusion.37 In the case of fusing heterocycles to an s-indacene core, there are two symmetric orientations possible. In one case, the heteroatom is on the same side as the apical carbon of the 5-membered core ring (syn-) and the other case the heteroatom is on the opposite side as the apical carbon (anti-) (Figure 1.5).38 This naming convention is utilized throughout the remainder of this article. 30 1.2 Initial studies Admittedly, when we first started working on the 20 π-electron indenofluorenes, we assumed their high degree of conjugation might mean their electronic properties could be similar to the well-known 22 π-electron pentacenes,39 yet the IFs did not possess the Achilles’ heel that leads to acene degradation, namely the reactive s-cis diene orientation of the double bonds. In many ways, our naïveté proved to be a blessing in disguise. Aside from this one (wrong) assumption of an indenofluorene/pentacene correlation, the slate was completely blank given the dearth of known indenofluorenes. After our initial work on 14 showed that the [12]annulene transannular cyclization route was not scalable as well as potentially hazardous,33,34 we switched to a variation of “inside-out” route I reported by Wang36 where the central piece was commercially available 2,5-dibromo-p- xylene. After the Suzuki cross-coupling, we oxidized the benzylic methyl groups with basic KMnO4 and then completed IF synthesis as outlined at the top of Figure 4. While this route did provide a small library of stable donor- or acceptor-substituted [1,2-b]IFs 15 (Figure 1.6),40 it too had drawbacks as some of the R substituents were not tolerant of the harsh KMnO4 oxidation. In addition, this study illustrated the need to examine the potential targets computationally before attempting their syntheses. Had we done so, we would have found that substitution on the 2- and 8-positions is undesirable as these carbons have minimal HOMO/LUMO electron density, relegating any influence to be through weaker inductive effects. Instead, the calculations suggested we should functionalize the 6- and 12-positions, which led to the series of [1,2-b]IFs 16.41 This proved much more successful, as we could vary the UV-Vis λmax by ~50 nm, whereas for 15 this variation of λmax was a very modest 16 nm. Importantly, this latter study also demonstrated that we could prepare single-crystal OFETs with a [1,2-b]IF as the active layer that 31 showed ambipolar transport, our first demonstration of the materials chemistry potential of the IFs,41 but more on that below. Figure 1.6. Early attempts to alter IF optoelectronic properties. One other area we examined briefly was to exchange the central benzene ring with either a thiophene or selenophene, leading to [2,1-c]IF analogues 17.42 While this structural variation altered the optoelectronic properties by ~25 nm compared to the [2,1-c]IF, the λmax were hypsochromically shifted much to our surprise. Furthermore, the difference between S and Se was minimal (1 nm), so further efforts on this pathway were abandoned. Rather than inclusion of the heterocycle within the antiaromatic core, we soon discovered that that fusion of heterocycles to the core was the trick to generate highly antiaromatic structures (vide infra). As noted earlier, the indenofluorenes sit at a unique intersection of antiaromaticity and diradical character. Modification of the indenofluorene scaffold has followed two main directions: core π-extension and outer π-extension (Figure 1.7). Consistent with earlier comput- Ar Ar i-Pr3Si Sii-Pr3 R R i-Pr3Si Sii-Pr3 R R R R Sii-Pr3 Sii-Pr3 14, R = H, Dec 15, R = H, F, Cl, Br, Me, Ph, 2-(5-BuC4H2S) 4-CF3C6H4, 3,5-(CF3)2C6H3 16, Ar = Ph, 4-FC6H4, 4-ClC6H4, 4-BrC6H4, 4-MeC6H4, 4-MeOC6H4 2,4,6-Me3C6H2, 2,4,6-(MeO)3C6H2 4-CF3C6H4, 3,5-(CF3)2C6H3, C6F5 X i-Pr3Si Sii-Pr3 17, X = S, Se 32 ational studies on a series of dicyclopenta-fused acenes,43 we will show that expanding the quinoidal indenofluorene core by successive “benzinterposition”17 from one benzenoid ring ([1,2-b]IF 10) to two or three benzenoid rings (diindenoacenes), while significantly weakening core paratropicity, leads to strong diradical character in the latter systems.44 [1,2-b]IFs that possess an s-indacene core generally do not have strong diradical character; however, they do Figure 1.7. Modification of the [1,2-b]IF scaffold follows two main lines of investigation—outer π-extension/ alteration with aromatic carbocycles or heterocycles (diareno-s-indacenes, top) and core π-extension from benzenoid to naphthoquinoid and anthraquinoid (diindenoacenes, bottom). maintain a higher degree of antiaromaticity in the core. Concurrent exploration of the antiaromaticity and diradical character of these quinoidal systems has led to modification of the outer π-system to modulate the antiaromaticity within the s-indacene motif and thus translation of these findings about the outer arenes to π-expanded cores furnishes systems with tunable diradical character, thus highlighting the antiaromaticity/diradical character interrelationship.13 Modifications to rationally alter antiaromaticity will be presented first, followed by methods [1,2-b]IF (10) core π-extension outer π-extension Ar Ar diindenoarene (n = 1-2) Ar Ar Ar Ar n diareno-s-indacene 33 used to assess antiaromaticity and other inherent properties of these molecules. The use of different outer arenes to increase diradical character will be discussed next, and finally some device applications that take advantage of the antiaromaticity/diradical interrelationship will be described. 1.3 Tuning the paratropicity strength of the s-indacene core by outer arene modification While antiaromatic s-indacene derivatives have a potential role in the future of organic electronics, for this potential to be realized, we needed to better understand how to alter the strength of the antiaromatic s-indacene ring current in a rational manner, especially through fusion of different arenes to the outside. Our first test of this hypothesis in 2014 fused thiophene (indacenodithiophene, IDT) and benzothiophene (indacenodibenzothiophene, IDBT) to 7.38 While the initial idea was more curiosity driven and not motivated originally by increasing the Figure 1.8. Aromatic carbocycle-fused s-indacene derivatives—DNIs 18-20, DPI 21, and DAI 22. 34 antiaromaticity from the parent [1,2-b]IF, we found via NICS calculations that these heterocycle- fused systems restored the paratropic ring current of the s-indacene motif close to that calculated for the parent hydrocarbon 7. This initially surprising result suggested that we needed to do a much deeper dive to elucidate the effects of outer arene alteration. Of course, this begs the question—what exactly are we “tuning”? The main effect is varying the HOMO/LUMO energy gap, with a secondary effect of altering the exact HOMO and LUMO energy levels (vide infra). Our first intentional foray into tuning the paratropicity of the s-indacene core was achieved by creating a series of molecules with fused aromatic carbocycles such as naphthalene (dinaphthoindacenes (DNI) 18-20, Figure 1.8), phenanthrene (diphenanthroindacene (DPI) 21), and anthracene (dianthracenoindacene (DAI) 22).37,45 While 18, 19, and 21 showed core paratropicities between those of [1,2-b]IF and the IDTs/IDBTs, both 20 and 22 indicated weaker core paratropicity than [1,2-b]IF.37 From these studies, a simple “bond order” rationalization suggested that greater double bond character of the fused bond (e.g., the bond order of the fused bond in 21 is 1.8) resulted in increased antiaromaticity of the s-indacene core, whereas less double bond character of the fused bond (e.g., bond order of fused bond in 22 is 1.2) resulted in decreased antiaromaticity. Interestingly, DAI 22 is the only fluorescent indacene derivative to date.45 The linear 2,3-anthracene fusion severely diminished the paratropicity of the core of 22, resulting in an increase of the S0S1 energy gap and thus deactivating the low-barrier conical intersection observed in all other s-indacene derivatives.46 Nonetheless, this rudimentary “bond order” rationalization only works for hydrocarbon-fused s-indacenes. In the realm of hetero- cycle-fused s-indacenes, the heteroatom plays an important role in affecting the antiaromaticity of the core, and this nuance is not captured simply by the bond order of the fused ring. 35 Figure 1.9. (a) Heterocycle-fused s-indacene core antiaromatics grouped by family of heterocycle: thiophenes/ benzothiophenes 23-26 (blue), benzothiophene sulfones 27-28 (yellow), benzofurans 29-30 (orange), naphthothiophenes 31-36 (green), indoles 37-38 (pink), and mixed benzofurans/benzothiophenes 39-42 and their corresponding sulfones 43-46 (purple). (b) syn- Heterocycle-fused s-indacenes in order of increasing antiaromaticity. 36 After initially fusing thiophene (23-24) and benzothiophene (25-26) rings to s-indacene and observing an increase in antiaromaticity from the parent [1,2-b]IF,38,47 we became interested in further manipulating the antiaromaticity of the s-indacene core. Whereas fusion of arenes to 7 affects the degree of decreased paratropicity, fusion of heterocycles restores the antiaromaticity of 7 and has the potential to increase beyond the level of unsubstituted s-indacene itself. To further explore the effect of heterocycle fusion, we examined three different directions: oxidation of the benzothiophene-fused s-indacenes to sulfones (indacenodibenzothiophenesulfone (IDBTS) 27-28),48 further π-extending the outer arenes from thiophenes and benzothiophenes to naphthothiophenes (indacenodinaphthothiophene (IDNT) 31-36),49 and changing the heteroatom from sulfur to oxygen (indacenodibenzofuran (IDBF) 29-30)50 or nitrogen (indacenodiindole (IDI) 37-38).51 With the modular synthetic routes available (Figure 1.4), if the desired outer arene cross-coupling partner could be accessed, the “inside-out” method yielded the desired isomers. The ultimate examples of this precision synthetic method were showcased in the preparation of unsymmetrical, regioisomeric indacenobenzofuranbenzothiophenes (IBFBTs) 39- 42 and their corresponding oxidized donor/acceptor IBFBTS counterparts 43-46 via selective Suzuki cross-coupling to dimethyl 2-chloro-5-iodoterephthalate,52 whereas such specific, highly defined molecular geometries in acenes and related structures would be difficult to produce given current preparative methods (e.g., regio-random Aldol condensation).39 While a majority of these s-indacenes were stable and could be thoroughly characterized, a few either were short- lived (30, 46) or afforded something other than the desired product (38). 37 1.4 Computational techniques When evaluating the antiaromaticity of s-indacene derivatives, we employ a variety of experimental and computational techniques. Our most powerful and useful computational tool is NICS-XY scans.53 This convenient and information-rich method plots the NICS values of dummy atoms placed in a set path across the compound of interest (Figure 1.10a), making it visually simple to identify trends in aromaticity and antiaromaticity across a series of molecules (Figures 1.10b-10d). Using NICS-XY scans, students in the Haley lab identified several new substituted s-indacene targets with pronounced paratropicity (e.g., 29-38), as they were predicted to be more antiaromatic than our benchmark, unsubstituted C2h s-indacene 7.49-51 Figure 1.10. (a) Example path of the dummy atoms in NICS-XY(1.7)πZZ scan calculations. NICS-XY scans of (b) carbocycle-fused systems (10 and 18-22) and (c) heterocycle-fused systems (25-30 and 37-38) versus that of s-indacene reference molecule 7. (d) NICS-XY scans of only the syn-fused heterocycles in Fig. 9b. All NICS scan plots are ordered from most to least antiaromatic. NICS-XY scans performed at 1.7 Å above the ring system using the M11/6- 311+G** level of theory. Figure 1.10 includes NICS-XY scans for selected carbocycle- (10 and 18-22) and 38 heterocycle-fused compounds (25-30 and 37-38) with benchmark 7. It should be noted that these are recently recalculated/newly published NICS scans based on M11-optimized geometries,54 as Wu and co-workers showed that NICS scans based on B3LYP-optimized geometries, as we originally published, can often provide values that overestimate delocalization in the s-indacene core.55 It is also important to note that one should not focus solely on the absolute NICS values but rather on the trends in data that are all calculated with the same parameters. For the carbocycle series (Figure 1.10b), s-indacene 7 is the most antiaromatic followed by phenanthro- fused 21, the syn/anti-DNI isomers (18/19), the parent [1,2-b]IF (10), and then the linear- analogues (20/22), supporting the simple “bond order” rationalization for paratropicity strength. Whereas the differences between syn- and anti-fusion are minimal in the carbocycles, the differences are much more pronounced in the heterocycle series (Figure 1.10c). In general, syn- fused aromatic heterocycle isomers are more antiaromatic than the anti-fused isomers, with the syn-indole (37), -benzofuran (29) and -benzothiophene (25) all more paratropic than benchmark 7 and anti-indole (38), -benzofuran (30) and -benzothiophene (26) possessing roughly comparable paratropicity to 7. Interestingly, the nonaromatic benzothiophene sulfone-fused derivatives 27 and 28 are calculated to be less antiaromatic than 7 and are the only heterocycle- fused s-indacenes that reverse the trend of increased antiaromaticity for syn-fusion, i.e., anti- IDBTS 28 is more paratropic than syn-IDBTS 27. While Figures 1.10b-10d show an appreciable ~20 ppm variance in the NICS πZZ values of the s-indacene core, the corresponding values of the external rings are essentially insensitive to bond fusion/bond order. In the carbocyclic series, the range of NICSπZZ values for the ring directly fused to s-indacene is small (–10 to –12 ppm), and even smaller (–15 to –16 ppm) for the terminal ring of the DNIs. Whereas the values of the inner ring show the ‘conflict’ between 39 opposing diatropic and paratropic ring currents, the outer ring is more isolated and thus the values are more typical of benzene/naphthalene. Exchanging the fused carbocyclic ring for a heterocyclic ring significantly negates this ‘conflict’, as the now-weakly diatropic heterocycles (NICSπZZ values of –2 to –4 ppm) or atropic sulfone 27 (+1 ppm) act as a spacer separating the strongly diatropic outer benzenes and the highly paratropic s-indacene. While our studies had shown that extending the π-system of s-indacene through heterocycle or benzoheterocycle annelation leads to modification of the antiaromatic character of the core, we did not know whether this was a result extension of the π-system or the presence/ location of the heteroatom. To understand these differences, we examined a series of test molecules to elucidate the specific effects of the heteroatom and its substitution position, the heterocycle, and the further extension of π-conjugation.56 Computationally this was accompli- shed using NICS calculations,57 current density maps,58 and NICS2BC,59 all of which painted a consistent picture. Firstly, we found that the heteroatom itself can have a dramatic effect on the antiaromaticity of the core: placing uncyclized heteroatoms at the 1,5-positions (i.e., the anti- isomers) affords the greatest extent of antiaromaticity alleviation, leading in the case of the NH2- substituted s-indacene to almost complete loss of antiaromatic character. Second, we established that this effect is highly site-specific. Using resonance structures, we rationalized this site- sensitivity by showing that the anti-position enables the substituent to interact with the core via resonance while the syn-position does not. Third, by comparing the two types of isomers we found that the extension of the π-system by heterocycle annelation does indeed alleviate the antiaromaticity of the core, although it is a weaker effect than the resonative heteroatom effect. Finally, we demonstrated that this effect is dampened by the addition of another ring, i.e., benz- annelation. Overall, this study showed that the reduction in antiaromaticity in the anti-isomers is 40 primarily a heteroatom effect, which is weakened by cyclization and further benzannelation. In contrast, for the syn-isomers (Figure 1.9b) the heteroatom effect is weak and the majority of antiaromaticity alleviation is achieved by the neighboring ring effect; however, further benz- annelation cancels this, allowing the s-indacene to regain (and even surpass) its full antiaromaticity, as illustrated in Figure 1.10d.56a 1.5 NMR spectroscopy 1H NMR spectroscopy is a routine part of characterizing new s-indacene derivatives and provides an additional way to compare the paratropicity of similar systems. The protons on the six-membered ring of the s-indacene core are shifted upfield relative to the degree of paratropicity—the more antiaromatic molecules show a greater upfield shift of the core protons. When the bulky aryl groups are consistent, e.g., the mesityls in Figures 8 and 9a, comparisons between different systems become possible. This is a powerful tool for several reasons: NMR spectroscopy is a magnetic-based method, making it easy to compare with magnetic criteria based computational methods such as NICS, and it can be compared across a wider range of systems than cyclic voltammetry (CV) or UV-Vis spectroscopy, for example. In most cases, comparison of the chemical shift of the core protons matches the order predicted by the NICS-XY scans. Table 1 shows the collected values for s-indacene core IF analogues. NICS-XY scans predict syn-fusion of the sulfur heteroatom to increase antiaromaticity through π-extension (IDT 23IDBT 25IDNT 35), and this matches with the further upfield chemical shift of the core proton: IDT 23 (6.06 ppm) > IDBT 25 (6.02 ppm) > IDNT 35 (5.99 ppm). While the ordering within the six IDNT regioisomers does not follow the NICS-XY scans perfectly, the differences are small and other effects sometimes dominate.49b 41 The oxidized IDBTSs reverse in trend and the anti-fusion has the more upfield chemical shift (28, 6.03 ppm) than the syn-fusion (27, 6.91 ppm).48 Through various modifications to the sulfur- based heterocycle-fused IFs, we notice a significant upfield shift from the parent IF and smaller changes within the family. Consistent with the computational studies,56a changing the heteroatom from sulfur to oxygen or nitrogen leads to a dramatic upfield shift of the core proton to 5.60 ppm for syn-IDBF 2950 and 5.39 ppm for syn-IDI 37.51 From these trends we can conclude that large changes to antiaromaticity can be made through heteroatom alteration, and fine tuning anti- aromaticity is done through fusion orientation (syn-/anti-), π-extension (IDT, IDBT, IDNT), and electronics (thiophene vs. thiophene sulfone). Finally, the core proton shifts of the unsymmetric IBFBT systems can be thought of as an “average” of their two constituent pieces, e.g., the core protons of syn,syn-IBFBT (5.90, 5.80 ppm) are roughly the average of those in syn-IDBT 25 (6.02 ppm) and syn-IDBF 29 (5.60 ppm). Table 1.1. 1H NMR chemical shift of the core protons of select s-indacene derivatives cmpd δ (ppm), (solvent) cmpd δ (ppm), (solvent) 7 (Ar = Mes)a 6.51 (CDCl3) 16 (Ar = Mes) 6.86 (CDCl3) 18 6.68 (CD2Cl2) 19 6.62 (CD2Cl2) 20 7.14 (CDCl3) 21 6.95 (CDCl3) 22 7.19 (CDCl3) — — 23 6.06 (CDCl3) 24 6.05 (CDCl3) 25 6.02 (CDCl3) 26 6.07 (CDCl3) 27 6.91 (CDCl3) 28 6.03 (CDCl3) 29 5.60 (CD2Cl2) 30 6.14 (CD2Cl2) 35 5.99 (CDCl3) 36 6.09 (CDCl3) 37 5.39 (CD2Cl2) — — 39 5.85 (CD2Cl2) b 40 5.87 (CD2Cl2) b 41 6.05 (CD2Cl2) b 42 6.10 (CD2Cl2) b 43 6.17 (CD2Cl2) b 44 5.64 (CD2Cl2) 45 6.50 (CD2Cl2) b 46 5.99 (CD2Cl2) aReference 22. bAverage of two peaks. 42 1.6 Electrochemical and photophysical properties The electrochemical and photophysical properties of the diarenoindacenes are also indicative of their antiaromaticity. While they possess HOMO-LUMO energy gaps and thus low energy absorptions similar to acenes, the diarenoindacenes are characterized by considerably lower HOMO and LUMO energy levels, typical of antiaromatic molecules. Additionally, these systems are redox active and usually have two reduction and one or two oxidation events. Unlike NMR spectroscopy, it is more difficult to draw meaningful comparisons of the photophysical and electrochemical properties across a wide range of derivatives. The HOMO-LUMO gap is affected by both antiaromaticity and π-extension, making it difficult to deconvolute their combined impact. Similarly, direct comparison of photophysical properties across families of diarenoindacenes is complicated by the same factors, resulting in red-shifted absorptions; however, both CV and UV-Vis provide important information about the system and its antiaromaticity, especially the ability to tune the HOMO-LUMO energy gap. Table 1.2 contains the HOMO and LUMO energy levels, HOMO-LUMO energy gaps, and λmax of the low energy absorption of a variety of diarenoindacenes. In general, we see that the more antiaromatic systems (25, 29, and 35) have a smaller HOMO-LUMO energy gap and more red-shifted absorption, especially when comparing the syn- and anti-isomers of a particular heterocycle. The broad class of diarenoindacenes typically has HOMO and LUMO energy levels in the range of –5.5 to –5.8 and –3.7 to –4.1 eV, respectively, with HOMO-LUMO energy gaps in the range of 1.4-2.0 eV. Interestingly, oxidation of the benzothiophenes to their corresponding sulfones (27/28/43) further drops the LUMO energy level (–4.4 to –4.5 eV). The low LUMO energy levels reflect the high electron affinity of the diarenoindacene family, which stand in 43 Table 1.2. Electrochemical and photophysical values of select s-indacene derivatives cmpd EHOMO (eV) ELUMO (eV) Egap (eV) λmax (nm) 16 (Ar = Mes) –5.34 –2.92 2.42 516 18 –5.73 –3.72 2.01 578 19 –5.59 –3.68 1.91 549 20 — — — 543 21 –5.73 –3.87 1.86 622 22 — — — 615 (em 664) 25 –5.54 –3.93 1.61 626 26 –5.52 –3.81 1.71 618 27 –6.28 –4.51 1.77 624 28 –6.12 –4.46 1.67 587 29 –5.55 –3.97 1.58 642 30 — — — 584 35 –5.49 –4.08 1.41 697 36 –5.43 –3.83 1.60 665 39 –5.67 –4.08 1.59 630 43 –5.84 –4.44 1.40 670 contrast to acenes, which typically have much lower electron affinities unless heavily appended with electron-withdrawing groups. These molecules’ pronounced electron affinities make intuitive sense—as they are successively reduced, the compounds regain an aromatic configuration and thus aromatic stabilization. Much like fullerenes, the presence of the five- membered rings containing all-sp2 hybridized carbons makes this class compounds inherently electron-accepting. The diarenoindacenes are highly colored, covering a broad spectrum with colors varying from orange (23) to red-orange (16a) to purple (27) to blue (36), which are reflected in the electronic absorption spectra of the molecules. Compared to 16 (516 nm, Table 1.2), the λmax of the low energy band of the carbocycle-fused indacenes progressively shifts to the red—543 (20), 549 (19), 578 (18), and 622 nm (21), decreasing the optical gap from ca. 2.3 to 1.9 eV.37 Although not shown in Table 1.2, replacement of the mesityls with (triisopropylsilyl)ethynyl 44 groups typically shifts the low-energy λmax further into the red by roughly 50-75 nm (e.g., low- energy λmax for TIPSethynyl-substituted 16 and 21 are 568 and 692 nm, respectively). This difference is attributable to the degree of the conjugation of the quinoidal core with apical substituents: whereas the alkynes are fully conjugated to the π-electron-rich cores, the mesityls are in partial π-electronic communication only due to their near orthogonality (>75° dihedral) to the diarenoindacene backbone. Despite being highly colored and π-electron-rich systems, the diarenoindacenes are surprisingly non-fluorescent due to a non-radiative decay pathway,46 as was also observed in earlier studies of s-indacene derivatives.60 Transient absorption spectroscopy of derivatives of 10 and 13 were shown to possess extremely short S1 lifetimes of ~10-15 ps.46 Interestingly, studies on tetrakis(t-butyl) analogue of core structure 7 showed it to have a lifetime of 18 ps for the S1 to ground state relaxation.60a Such extremely short lifetimes explain why these molecules are non- emissive, since fluorescence (typically occurring with lifetimes of greater than 10–9 s) is not a competitive process at this time scale. Quantum chemical calculations indicate this non- emissiveness to be the result of an easily accessible potential energy surface crossing between the first excited singlet state (S1) and ground electronic state (S0), i.e., a conical intersection.46,60 Similar to 7, this process allows efficient internal conversion to the ground state, deactivating fluorescence and thus yielding ‘dark’ molecules. As noted earlier, the only exception to this case is DAI 19—with its antiaromaticity turned ‘off’ by 2,3-anthraceno-fusion, the S0 and S1 states in 19 separate enough in energy that fluorescence, albeit weak, is restored.45 1.7 X-ray crystallography X-ray crystallography and its use in determining bond lengths has been an important 45 factor in understanding our s-indacene derivatives. We find that when fusing arenes or hetero- arenes to an s-indacene core, we observe a bond localized structure similar to C2h symmetric s- indacene 7. Applying X-ray crystallography to understand the change in antiaromaticity in the IF scaffold focuses on the bond length alternation in the s-indacene core. Generally, we observe that the structures with a greater degree of antiaromaticity have more bond length alternation, e.g., the short bonds are shorter and the long bonds are longer as antiaromaticity increases. While this metric provides support for ranking the antiaromaticity of IF derivatives, it also has some pitfalls. Table 1.3. Core C–C bond lengths (Å) of select s-indacene derivatives cmpd a b c d e f 16 (Ar = Mes) 1.413 1.469 1.356 1.433 1.380 1.471 18 1.401 1.470 1.367 1.431 1.377 1.461 22a 1.450 1.447 1.368 1.418 1.378 1.452 23 1.384 1.461 1.363 1.418 1.398 1.447 24 1.389 1.452 1.360 1.431 1.388 1.460 25 1.391 1.457 1.371 1.421 1.407 1.441 26 1.393 1.437 1.377 1.412 1.409 1.435 27a 1.442 1.397 1.415 1.371 1.467 1.376 28 1.437 1.384 1.420 1.370 1.464 1.375 29 1.374 1.451 1.371 1.418 1.418 1.432 30 1.390 1.412 1.386 1.394 1.435 1.419 35 1.373 1.468 1.368 1.422 1.389 1.438 36a 1.397 1.433 1.376 1.406 1.413 1.429 37a 1.382 1.458 1.363 1.429 1.394 1.448 aAverage of two independent molecules in the crystal lattice. In addition to the complex relationship between antiaromaticity and bond lengths, crystal packing forces are known to affect bond lengths and make bond length analysis difficult,22 a b c d e f 46 however, as well as some antiaromatic compounds can also be delocalized.61 Table 1.3 lists the core bond lengths of mesityl-substituted 16, 18, 22-30, and 35-37. All the structures exhibit the para-quinoidal motif of closed shell 6 within the indacene core, with bonds c and e possessing more double-bond character clearly evident, e.g., syn-IDI 37 in Figure 1.11 (left). For example, double bond c in the central six-membered ring varies from 1.356 to 1.386 Å. Bond e between the core and the apical carbon bearing the mesityl groups varies from 1.371 to 1.380 Å in the hydrocarbon structures, yet lengthens to 1.389−1.435 Å in the hetero-cycle-fused systems, which begins to hint at the contribution of a biradical resonance form such as open-shell 6. The outer benzenes exhibit archetypical bond lengths of 1.38−1.40 Å, with the fused bond at ~1.41 Å. A large majority of the diarenoindacene structures are well-bracketed by these values, whether aryl- or alkynyl-substituted on the apical carbons. In addition, nearly all the molecules are planar species, with root-mean-square deviations from the average molecular plane of 0.04 Å or less. Interestingly, we find that sulfone-fused s-indacenes 27 and 28 exhibit a “flipped” core in which the double bonds are exocyclic to the heterocycle. Figure 1.11. Molecular structure of syn-IDI 37 (CCDC 2416146; ellipsoids drawn at the 50% probability level and hydrogen atoms are omitted for clarity). 47 1.8 Dicyclopenta[b,i]anthracene: the core to achieve strong OS Character After establishing a large body of work on indenofluorenes with multiple strategies for external acene and heterocycle fusion on s-indacene to tune optoelectronic properties and paratropicity, the next logical progression was internal π-expansion featuring dicyclopenta- [b,g]naphthalene (DCN) and dicyclopenta[b,i]anthracene (DCA) cores, e.g. sequential benzinterposition, to investigate further changes in optoelectronic properties. However, the relationship between antiaromaticity and diradical character discussed above sees a greater shift towards OS character proportional to size and can be thought of as a continuum from antiaromaticity to diradical character (Figure 1.12).13,43c,62 While unintended at the time, the Haley lab would find itself entering the arena of stable diradicaloids on account of the elongated quinoidal conjugation within the pro-aromatic DCN and DCA cores. Admittedly, core expansion did begin with fluoreno[4,3-c]fluorene ([4,3-c]FF)63, but discussion of the FF homologues will be described later as they uniquely straddle CS and OS behavior. The focus here will instead be on DCA-based diindeno[1,2-b/1’,2’-i]anthracene (DIAn) homologues as we now know they are strong diradicaloids and serve as our upper limit to OS character in our lab so far. Figure 1.12. Continuum of core PAAHs from antiaromatic to diradicaloid featuring sequential insertion of 6-membered rings, i.e., benzinterposition. Pro-aromatic p-QDM, 2,6- naphthoquinodimethane (NQD) and 2,6-anthroquinodimethane (AQD) fragments are highlighted. Antiaromatic Closed-Shell Open-Shell Nonaromatic 48 Pursuit of DIAn 47 (Figure 1.13) had begun in 2012 motivated by anticipation of enhanced optoelectronic properties, but the project remained slow going until then graduate student, Gabriel Rudebusch, completed the synthesis in 2015 following our “outside-in” I strategy.64 To his credit he rationalized that inclusion of both the bulky mesityl and TIPSethynyl groups in the starting material would impede the Friedel-Crafts alkylation at the favored 1/5- positions of the anthracene core because of steric clash, but rather close onto to the less favored 3/7-positions, which worked brilliantly. Treatment with DDQ afforded deep blue crystals of 47. Interestingly, the peaks in the proton NMR spectrum at room temperature were slightly broadened, so Rudebusch performed variable-temperature 1H NMR (VT-NMR) experiments up to 150 °C to qualitatively probe a singlet to triplet state transition. Upon warming, the aromatic peaks in the series of spectra gradually broadened as the paramagnetic triplet state is populated. By 150 °C, all the peaks were in the baseline save those of the nondeuterated NMR solvent, and all the peaks returned to their former shape and height upon cooling to room temperature, offering the first evidence of having a genuine diradicaloid. In the electronic absorption of 47, a major low energy absorbance at 690 nm was recorded with a peculiar shoulder starting near 725 nm and tailing to 900 nm. This peculiar profile hints towards OS character as shoulders extending into the near-infrared (NIR) region have been attributed to the doubly-excited configuration in diradicaloids.65 The system also exhibited textbook redox amphoterism via cyclic voltammetry as two quasi-reversible single-electron oxidations and reductions were recorded, affording a particularly narrow Egap of 1.45 eV, which is a much smaller value compared to its relative [1,2- b]IF cousin (2.22 eV). A single crystal of DIAn suitable for X-ray diffraction was then grown and analyzed. Of particular importance was the 1.406 Å bond length between the apical carbon 49 and the anthracene core, suggesting this bond was displaying some single bond character reflecting the OS resonance structure. Other elongated double bond lengths in the crystal structure suggested there was substantial contribution from the OS form in the ground state canonical structure. These findings frame DIAn as a stable singlet diradicaloid, especially when paired with its y0 value of 0.623. Figure 1.13. CS and OS resonance structures of DIAn 47 depicting transition from singlet to triplet states alongside DIAn homologues 48-52. Pro-aromatic AQD fragments are highlighted. Inspired by our work with IFs, naphthalene fusion to the periphery of the DIAn scaffold affords three dibenzodiindenoanthracene (DBDIAns) isomers 48-51 that mirror the linear-, syn- and anti-fusions seen in IFs 18-20. With such π-extension, the diradicaloid character sequentially strengthens from 48-50 covering a y0 range of 0.623-0.711 for the hydrocarbon DIAns (Table 1.4).66 Benzothiophene fusion in 51 pushes OS character even further (y0 = 0.812) and effectively caps our suite of diradicaloids as the upper bound. Each analogue exhibited Mes Mes Sii-Pr3 Sii-Pr3 47 52 Mes Mes Sii-Pr3 Sii-Pr3 48 Mes Mes Sii-Pr3 Sii-Pr3 50 Mes Mes Sii-Pr3 Sii-Pr3 49 Mes Mes Sii-Pr3 Sii-Pr3 51 S S Mes Mes Sii-Pr3 Sii-Pr3 Mes Mes Sii-Pr3 Sii-Pr3 Mes* Mes Me Me Me MeMe Mes t-Bu MeMe Mes* * 50 prominent diradicaloid signatures in VT 1H NMR and electronic absorption experiments with spin-state transitions qualitatively observed in NMR signal broadening along with double- excitonic features in the NIR region. Elongated double bonds between the apical carbon and anthracene in crystal structures (1.391-1.434 Å) also supported substantial OS character. Further structural adjustment resulted in DIAn 52. The structural design of 47 included TIPSethynyl groups pendant to the DCA backbone on account of directing the correct closure during the Friedel-Crafts alkylation step of “outside-in” I, additional solubility, and thermodynamic stabilization along the core. It was suspected that the ethynyl groups were electronically non-innocent due to a y0 value shifting to 0.615.66 Investigation of 52 revealed a very similar UV/Vis profile and X-ray structure to 47, so optoelectronic properties were minimally impacted by loss of alkyne substitution.67 Even so, the new y0 suggests there is a change to the diradicaloid character. This computational and experimental evidence begins to paint a detailed image of our diradicaloids; however, the finer details regarding changes in spin- state transitions between DIAn derivatives were still unclear. To probe deeper into the OS nature of our DIAns, it was necessary for our collaborators to dig computationally into the details. 1.8.1 Calculated singlet-triplet energy gaps and electronic parameters The energy gap between the singlet state and triplet state (ΔEST) in diradicaloids is an important property that can be calculated and then unambiguously measured experimentally to quantify diradical character. Calculations using the two-electron/two-stie model were carried out by our collaborators in the Nakano group to probe the parameters contributing to ΔEST. Equations 1 and 2 describe the model’s determination of ΔEST and its relationship to y, where a and b represent the two electrons in localized natural orbitals, with several key terms: U is 51 defined as the difference between onsite and intersite Coulomb repulsions, tab is the transfer integral, and Kab as the direct exchange integral. Within this framework, the balance between U and tab is detailed and rationalizes underlying structure-property relationships in this series of diradicaloids. (1) ∆𝐸𝑆𝑇 = 𝑈 2 [1 − 1 √𝑦(2 − 𝑦) ] + 2𝐾𝑎𝑏 (2) 𝑦 = 1 − 1 √1 + ( 𝑈 4𝑡𝑎𝑏 ) 2 Analysis of DIAn 47 provided a calculated ΔEST value of –4.7 kcal mol–1 and serves as a reference point for the remaining derivatives (Table 1.4). In the case of 48-50, tab appreciably decreases corresponding to an increase in OS character in accordance with Eq 2. U only slightly changes between these three isomers highlighting the major contribution of external delocalization to tab. DIAn 51 has the smallest tab of this set and a larger U attributed to the electron rich S-atom enhancing diradical character with both parameters. Despite the trend in increasing y0, predicted ΔEST values hold 48 and 50 as the largest and smallest gaps, respectively, in this set of diradicaloids. DIAn 52 increases in its calculated ΔEST to –5.4 kcal mol–1 with slightly higher Coulombic repulsions and transfer integral values relative to 47. TIPSethynyl substitution there-fore assists in stabilizing the triplet state in all DIAns to some degree. This computational analysis illustrates the balance between U and tab on both diradical character and ΔEST, where they trend proportionally together. However, careful consideration of heterocycle fusion is needed as in the case of 51. 52 Table 1.4. Electronic Parameters and Singlet-Triplet Energy Gap for DIAn derivatives 47-52 cmpd U/2 (eV)a,b tab (eV)a,b y0 a,c ΔESTcalc (kcal mol–1)a,d ΔESTexpt (kcal mol– 1) DIAn 47 1.435 0.916 0.623 –4.71 –4.2 linear-DBDIAn 48 1.348 0.905 0.638 –5.09 –4.8 syn-DBDIAn 49 1.378 0.818 0.686 –4.16 –3.8 anti-DBDIAn 50 1.377 0.781 0.711 –3.45 –3.2 DBTDIAn 51 1.572 0.774 0.815 –3.41 –3.4 DIAn 52 1.463 0.957 0.615 –5.42 –4.6 aGeometries optimized at R- and UB3LYP/6-311G* levels. bEstimated at CASCI(2,2)6-311G* level using (tuned-)LC-RBLYP MOs (denoted as tuned-LC-RBLYP-CASCI(2,2)/6-311G*). cCalculated at the PUHF/6-311G* level. dAdiabatic ΔEST value calculated at the spin-flip noncollinear (SF-NC)-TDDFT PBE5050/6-311G* level along with R- or UB3LYP/6-311G* zero-point vibrational energy correction for each spin state. 1.8.2 SQUID magnetometry A crucial part of the puzzle is experimental determination of ΔEST by superconducting quantum interference device (SQUID) magnetometry. This technique was initially employed on 47 by the Gómez-García group from the University of Valencia for experimental measure of ΔEST where small magnetic responses were recorded while heating powder samples to 400 K. Having fit the collected data to the Bleaney-Bowers equation (Figure 1.14),68 the experimental ΔEST for 47 was –4.2 kcal mol–1, in reasonable agreement with its predicted value. Similar examination of 52 shows a change of +0.4 kcal mol–1, agreeing with the predicted increase of the ΔEST compared to 47. SQUID magnetometry performed on 48-51 revealed ΔEST values (Table 1.4) that are in reasonable agreement with calculated estimates when fit to the Bleaney-Bowers equation and confirm the ordering of ΔEST in DIAns 47-52 where 48 features the largest spin-state energy gap and 50 has the smallest gap. Our diradicaloids exhibit the power of logical structural alterations in fine tuning magnetic behavior (Figure 1.14) with greater radical delocalization (smaller tab) and minimal 53 Coulombic repulsions (smaller U) favoring small ΔEST values. This level of control over spin- paired and spin-unpaired states is desirable for magnetic material applications for on-demand transitions. Figure 1.14. Bleaney-Bowers fits of heating cycle SQUID data for the DIAn series 48-52; traces are shown only up to 400 K for consistency. 1.9 Caught in the middle: DCN-based molecules straddle closed- and open-shell structures Recalling the continuum from antiaromaticity to diradical character, the sections in this perspective have so far illustrated our group’s mastery over the two extreme ends where IFs only exhibit CS behavior and DIAns exhibit prominent OS behavior. As suggested by its placement in the continuum, the NQD unit within the DCN core provides the basis for moderate diradicaloids and would help elucidate the cross-over from CS to OS characteristics. The aforementioned [4,3-c]FF had a y0 value of 0.377 and no observable indication of an OS electronic structure.63,69 Efforts for in depth investigations were then placed on [3,2-b]FF 53 (Figure 1.15) as it was predicted to have 54 greater OS character (y0 = 0.434).69,70 Optoelectronic properties were indeed enhanced due to longer conjugation with a redshifted low energy absorption and a very weak low energy shoulder teasing a doubly-excited configuration; however, the lack of peak broadening in the proton NMR spectrum and distinct bond alternation within the quinoidal core led us to erroneously believe that 53 was CS when first reported in 2017. Figure 1.15. FF homologues 53-61 and 63-64 along with decomposition products 62 and 65. NQD substructures are highlighted while linear and cross conjugation patterns are bolded in heterocycle fused derivatives according to Figure 1.5. 1.9.1 Breaking open diradical character in fluorenofluorenes Strategies for developing diradical character in FFs ultimately followed our theme of external arene modification to push its homologues into the intermediate diradicaloid regime. Angular dibenzo-fusion was adapted to the FF scaffold since π-extension influenced para- tropicity and OS character in previous studies. It was hypothesized that large changes to tab in FFs would boost diradical character as it did in the DIAn series. Computational analysis Mes* Mes Mes* Mes 54 Mes* Mes 56 Mes* Mes 55 Mes* Mes 58, X = S 60, X = SO2 63, X = O X X 53 Mes* Mes 59, X = S 61, X = SO2 64, X = O X X Mes* Mes 57 Mes* Mes 65 HO OH O OMes* Mes 62 SN S N O O O O * * * * * * * * * 55 supported this hypothesis as increasing y0 values evolved over linear-, syn-, and anti-fusion as tab decreased (Table 1.5).70 Further π-extension in 57 by way of phenanthrene fusion pushed the extent of FF OS character even further within the hydrocarbon series (y0 = 0.629). Table 1.5. Electronic Parameters and Singlet-Triplet Energy Gap for FF derivatives 53-63 cmpd U/2 (eV)a,b tab (eV)a,b y0 a,c ΔESTcalc (kcal mol–1)a,d ΔESTexpt (kcal mol– 1) FF 53 1.446 1.163 0.49 –10.25 –9.3 linear-DBFF 54 1.315 1.143 0.51 –10.03 –9.6 syn-DBFF 55 1.352 1.055 0.56 –9.63 –8.7 anti-DBFF 56 1.336 1.012 0.60 –8.61 –7.8 TBFF 57 1.349 0.96 0.63 –7.90 –7.6 anti-IIDBT 58 1.563 1.031 0.613 –8.77 –8.2 syn-IIDBT 59 1.404 0.905 0.658 –8.06 –7.2 anti-IIDBTS 60d –– –– 0.601 –9.65 –– syn-IIDBTS 61d –– –– 0.652 –8.29 –6.5 anti-IIDBF 63 1.677 1.058 0.623 –9.10 –– syn-IIDBF 64 1.38 0.865 0.682 –7.68 –6.0 aGeometries optimized at the R- and UB3LYP/6-311G* levels. bEstimated at the CASCI(2,2)6-311G* level using the (tuned-)LC-RBLYP MOs (denoted as tuned-LC-RBLYP-CASCI(2,2)/6-311G*). cCalculated at the PUHF/6-311G* level. dAdiabatic ΔEST value calculated at the spin-flip noncollinear (SF-NC)-TDDFT PBE5050/6-311G* level along with R- or UB3LYP/6-311G* zero-point vibrational energy correction for each spin state. dElectronic parameters were not calculated for these entries. In collaboration with the Kato group at the University of Shiga Prefecture, FFs 54-57 were synthesized following the “inside-out” II strategy. Characterization of this new set of dibenzofluorenofluorenes (DBFFs) confirmed appreciable diradical character in 56 and 57 with peak broadening in the 1H NMR spectra at elevated temperatures (~90 °C) and double exciton state signatures in the NIR region. syn-DBFF 55 showed weak indications of OS character requiring ~150 °C to see minimal 1H signal broadening and 54 exhibited a CS structure like 53. 56 Fortunately, the SQUID magnetometer had been recently upgraded to support temperatures up to 800 K; thus, experimental data for the weaker diradicaloids could be obtained.§§ The spin-state dynamics of these FFs illustrate stepwise decrease in ΔEST from 54→53→55-57 (Figure 1.16). Like in the DIAn series, linear 54 has a larger ΔEST compared to parent 53 due to the slight increase in U corresponding to a greater electronic screening between the peripheral arene moiety and the core. It is known that heterocycle fusion in 51 dramatically affects the electronic parameter U contributing to elevated diradical character while widening ΔEST relative to 50 on account of Pauli repulsions disfavoring the unpaired spin-state. Fusion of DCN with benzothiophene, like in 51, can result in an anti-relationship between the S-atom and apical carbon of the core’s five- membered ring. The syn-relationship can also be attained when reversing the fusion mode of benzothiophene onto DCN. As such, electron-rich indenoindenodibenzothiophenes (IIDBTs) 58 and 59 were prepared by Justin Dressler and Joshua Barker, respectively.71. These isomers feature different conjugation patterns between the radical center and S-atom where the anti-isomer has a linear pattern, and the syn-isomer has a cross pattern (Figure 1.5). Following these conjugation patterns, there is less interaction between the unpaired electrons and S-atoms in 59 as they come within two atoms from each other while they can neighbor each other in 58. Applying the same late-stage modification used to prepare sulfones 27 and 28 should give the IIDBTS analogues, yielding changes to their OS characteristics as we know the S-atom influences both y0 and ΔEST. Oxidation of 59 furnished sulfone 61, whose calculated y0 and ΔEST values show that it retains the diradical character of 59 yet exhibits an earlier spin-state transition than 60.48 Unfortunately, oxidation of 58 did not yielded sulfone 60 but rather the nitrogen-insertion product 62 (X-ray).48 57 Figure 1.16. Bleaney-Bowers fits of recorded SQUID data for FF homologues 53-59, 61, and 64. Relating these structural considerations to U and tab, both parameters are greater in 58 than they are in 59, as listed in Table 1.5, leading to increased OS character and a slightly narrower singlet-triplet gap for the syn-isomer. This is supported by elongation of the bond between the apical carbon and the core (1.419 and 1.421 Å, respectively) and red shifting of low energy absorption features from anti- to syn-fusion. The key evidence lies in SQUID results confirming the narrowed ΔEST in 59, relative to 58. When comparing these heterocycle-fused isomers to the hydrocarbon series, 56 and 57 sit between 58 and 59 in the range of ΔEST values for FF derivatives. Ordering of these FFs mirrors the ordering of DIAns 49-51 highlighting the consistent effect of electron repulsions with linear conjugation between the radical carbon and S- atom across multiple analogues. 59 capitalizes on the balance between π delocalization and coulombic repulsions and offers a design strategy to enhance y0 and ΔEST through syn-fusion. The experimentally determined ΔEST of 61 unexpectedly subverts that of 53 despite the calculated value being greater by 0.13 kcal mol–1. This was a sensible discovery since oxidation 58 of these S-atoms imparts electron deficiency further decreasing electronic repulsions near the cross-conjugated pathway. Exploration of heteroatom effects on diradical character in the FF scaffold continued with the inclusion of benzofurans as the O-atoms would provide even greater electronegativity which we know enhanced molecular properties in the IF congeners 29-30.50 Quantum chemical calculations on indenoindenodibenzofurans (IIDBFs) 63-64 painted these regioisomers as worthwhile targets on account of increased OS character compared to their benzothiophene counterparts. Calculated ΔEST values suggested syn-IIDBF 64 to be more magnetically interesting between the two, as we would now expect given the behavior of 59 compared to 58. Unfortunately, all attempts to prepare 63 instead afforded ring-opened product 65, analogous to the case of 30, where ring-opened material predominated.50 syn-IIDBF 64 was remarkably stable and its high diradical character with SQUID data revealed that 64 has the smallest ΔEST of the FF homologues at –6.0 kcal mol–1. Such results exemplify the profound effect electronegativity has on optoelectronic and magnetic properties in these highly conjugated systems. Taking a look at the ground covered by our forays into diradicaloids shows a wide window of the ΔEST landscape filled by our lab comprising a total range of 6.4 kcal mol–1 across the DIAn and FF families, from as low as –3.2 to as high as –9.6 kcal mol–1. Such efforts effectively demonstrate our mastery of tuning OS character and ΔEST through our judicious molecular designs. 1.10 What IF we turned to devices? Having produced such a large library of diarenoindacenes and diindenoarenes, we have covered a wide range of frontier orbital energy levels, low energy absorptions, and redox potentials through precise structural modification. The CS nature of IF derivatives allows 59 (trialkylsilyl)ethynyl groups to installed on their apical carbons providing favorable solid-state packing and solution processability compared to Mes and Mes* groups. These properties are desirable for semiconductor applications, as was mentioned several times throughout this perspective, so how do these molecules perform in device active layers? 1. Single crystals of 16 (R = C6F5) Weak and unbalanced ambipolar transport in an OFET constructed using single crystals of perfluorophenyl substituted [1,2-b]IF had been reported by our group with hole and electron mobilities of 7 × 10–4 and 3 × 10–3 cm2 V–1 s–1, respectively.41 2. Another instance of ambipolar behavior was observed in an OFET containing 47 having a hole mobility of 2 × 10–3 cm2 V–1 s–1 and an electron mobility of 4 × 10–3 cm2 V–1 s–1(for additional fabrication and experimental details, we direct readers to appropriate references for each OFET example).64 The balanced transport seen here is ideal for effective ambipolar transport, but performance is limited due to 1D stacking of 47. 3. TIPS anti-IDBT also exhibited hole-carrier transport in an OFET (0.44 cm2 V–1 s–1).47 Shortly after our work, a group in China showed that TIPS anti-IDBT was also ambipolar, with hole mobilities up to of 0.64 cm2 V–1 s–1 and an electron mobilities up to 0.34 cm2 V–1 s– 1.72 4. Best OFETs were TIPSethynyl substituted linear- and anti-DNI regioisomers were successfully spin-coated and annealed while retaining bricklayer packing.73 Evaluation of charge-carrier transport revealed average hole mobilities of 1.04 cm2 V–1 s–1 and 4.72 cm2 V–1 s–1 for TIPS linear-DNI and TIPS anti-DNI, respectively. Engineering the solid-state packing of OS molecules with simpler groups remains limited by the reactivity at the radical centers as was demonstrated during hydrogen/deuterium abstraction experiments with 12.24 60 Another method to probe transport properties that circumvents poor packing is scanning tunneling microscope break junction (STMBJ) technique.74 Key molecular handles needed to bind to Au electrodes in this technique are Lewis bases, such as thioethers in our case, which were incorporated into a select group of our molecules (Figure 1.17b) to undergo room temperature single molecule conductance experiments. In collaboration with the Kamanetska group at Boston University, hydrocarbons 67-68 were found to have moderate to high conductance (G0 = 10–4–10–2) as molecular length increased following anti-ohmic decay.75 Similar behavior was observed in 69 and 70. Notably, 70 behaved similarly to 68 rather than 67 despite its DCN core indicating greater constructive quantum interference (CQI) with diradicaloids having persistent singlet spin-states. Enhanced CQI is possible due to the electronic configurations in these strong open-shell molecules bringing their frontier orbitals closer to the Figure 1.17. Chemical structures of quinoidal derivatives used in (a) OFET evaluation and (b) single molecule conductance measurements. ThA ThA R R 20'R R 19' ThA ThA 70 S S 67 ThA ThA 69 S S ThA ThA 68ThA ThA 66 (b) (a) SMe Me ThA R R 26' S SR' R' 16 F FF R' = FF R = Sii-Pr3 Me 61 Fermi energy of the gold electrodes in the single-molecule junctions. See Chapter III for an in- depth theoretical discussion on quantum interference effects. 1.11 Conclusions Over the last 13 years, the Haley lab has contributed quite the library of molecules to antiaromaticity and diradicaloid literature. There have been several synthetic routes presented in our work and their modularity provides a powerful tool to logically extend conjugation in the core and periphery. Heterocycle fusion adds additional possibilities synthetic targets as we have shown with benzothiophene, benzofuran, and indole. While we have certainly made our mark with these achievements, it is no secret that there are many others in the antiaromatic/ diradicaloid arena. This field remains highly active with major contenders such as the Tobe group producing fluoreno[2,3-b]fluorene76 and indeno[2,1-b]fluorene,27 the Stepien group presenting diindeno[1,2-a:2′,1′-i]phenanthrene77 and twisted alkene oligoradicaloids,78 the Wu group with periacenes79 and zethrenes,80 the Chi group reporting nanographene fragments81 and heteroacenes,82 and many others.83 New minds are also making impacts with even larger conjugated systems incorporating multiple s- and as-indacene moieties like those from the Das group.84,85 We hope that our contributions offer inspiration to emerging groups and motivation to continue research in antiaromatic, OS molecules, and their device performance as the desire for novel organic electronic materials persists. 62 CHAPTER II PROBING THE INFLUENCE OF ALKYNE SUBSTITUTION ON THE ELECTRONIC AND MAGNETIC PROPERTIES OF DIINDENO[1,2-B;1’,2’-I]ANTHRACENES This chapter includes previously published and co-authored material from Vidal Jr., E.; Zakharov, L. N.; Gómez-García, C.; and Haley, M. M. Probing the Influence of Alkyne Substitution on the Electronic And Magnetic Properties of Diindeno[1,2-b;1’,2’-i]Anthracenes. J. Org. Chem. 2024, 89, 14515-14519. This manuscript was written by Efrain Vidal Jr. with assistance from Prof. Michael M. Haley. The project in this chapter was conceived by Prof. Michael M. Haley. The experimental work in this chapter was performed by Efrain Vidal Jr with help from Dr. Lev N. Zakharov (X-ray crystallograp