Introduction

Aromaticity, a foundational concept in organic chemistry, explains the stability of benzene and other molecules with 4n + 2 π-electrons in a planar, conjugated ring. Initially proposed by Ronald Breslow in 1965, the term antiaromaticity refers to a class of organic molecules with 4n πelectrons in a planar, conjugated ring. [1][2][3][4] 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. [5] Antiaromatic compounds are fundamentally interesting because they are extremely reactive, and therefore are nearly impossible to isolate in their unsubstituted form; however, significant work undertaken by various research groups has resulted in a library of aromatic/antiaromatic hybrid structures based on cyclobutadiene, [6,7] pentalene, [8][9][10] and indacene, [11,12] among others.

Our work in this area has been motivated, in part, by the molecular properties associated with antiaromatic systems, such as a small HOMO-LUMO energy gap, concomitant red-shifted absorption, redox amphoterism, and increased conductance. [13] These properties are desirable for applications in organic electronics such as organic field effect transistors (OFETs), organic photovoltaics (OPVs), and organic solar cells (OSCs). [14] Early hydrocarbon-fused sindacenes were incorporated into OFETs as the semiconducting layer and showed respectable hole mobilities up to 7 cm 2 /V • s, [15,16] a result that has motivated our continued exploration of s-indacene derivatives.

s-Indacene can be effectively stabilized through aromatic ring fusion to the outer five-membered rings, and the reactive apical carbons can be kinetically blocked with bulky substituents. A general synthetic method for s-indacene derivatives permits fine-tuning of molecular properties through early-stage (identity of the fused aromatics) or latestage synthetic modification (oxidation). [13] Recent studies have incorporated heterocycles (benzothiophene, benzofuran) as part of the outer aromatic rings, e.g., 1-6 (Figure 1a). [17][18][19] By varying the fusion orientation (Figure 1b) from anti-(where the heteroatom is fused on the opposite side of the scaffold from the apical carbon of the fivemembered ring) to syn-(where the heteroatom is on the same side as the apical carbon), we could further tune molecular properties within a set of regioisomers.

Many of the same properties that make antiaromatic compounds attractive for organic electronics are also associated with donor-acceptor (D/A) molecules. Incorporation of D/A motifs is a popular design strategy in improving aromatic scaffolds for use in organic electronics because it can facilitate intramolecular charge transfer (ICT), [20,21] which is the functional basis for many semiconducting and optoelectronic devices. [22,23] ICT is a fundamental process in photochemistry where upon excitation, electron density flows from a donor group through a covalently linked σ-or π-bridge to an acceptor group. [24] ICT results in molecular properties such as an extension of the absorption spectrum, a decrease in the HOMO-LUMO energy gap, and amphoteric redox behavior. ICT can be tuned by the identity and strength of the donor and acceptor groups, the length of the π-bridge, bond length alternation in the π-bridge, and the ways in which the donor, bridge, and acceptor are connected. [25,26] Several groups have incorporated donor and/or acceptor groups onto antiaromatic scaffolds, but almost always as either a D-A-D or A-D-A triad with the paratropic core as either the acceptor or donor unit. [27][28][29][30][31][32][33] Molecular property tuning is limited to the identity of the donor or acceptor groups, and evidence of ICT has been weak. Recently, the London group appended donor and acceptor groups on two isomers of monobenzopentalene and observed some evidence of donor-acceptor character, although this was not the focus of the work. [34] Das and co-workers were able to desymmetrize an indeno[2,1-c]fluorene (IF) with donor-and acceptor-functionalized R groups, with one derivative showing modest evidence of charge transfer (~10 nm solvatochromic shift), which is noteworthy considering the limited conjugation between the planar core and the orthogonal aryl groups; however, the authors noted the synthesis was difficult. [35] To the best of our knowledge, there are no published examples of antiaromatic molecules with fused donor and acceptor groups.

Recent studies focused on novel D/A aromatic topologies suggest electronic interactions between the donor and acceptor groups may be greatly enhanced if they are fused coplanar to the π-conjugated bridge; however, such structures have proven elusive. [36] Herein, we report the synthesis and characterization of a family of unsymmetrical, antiaromatic indacenobenzofuran-benzothiophenes (IBFBTs 7-10, Figure 2). Late-stage oxidation transforms the thiophene into an acceptor motif, furnishing all four D/ A isomers (IBFBTSs 11-14) that, depending on the fusion orientation of the benzofuran donor and benzothiophene-S,S-dioxide acceptor, show varying degrees of antiaromaticity and ICT character. The previously reported sets of symmetric benzoheterocycle-fused s-indacenes (1-6) [17][18][19] provide standards for the impact of each heterocycle and its fusion orientation on the antiaromaticity and molecular properties of the s-indacene scaffold, thus permitting a thorough, rationally designed study of desymmetrization of the paratropic s-indacene core, as well as any interplay between antiaromaticity and ICT. Notably, analogous studies with benzoheterocycles fused to aromatic motifs such as naphthalene (10 π-electrons) or anthracene (14 πelectrons), most comparable to 12 π-electron s-indacene, have not been reported. As Anthony showed in 2004, creating heterocycle-fused acenes is a significant synthetic challenge, as the double Aldol condensation to generate the eventual anthracene core leads to an inseparable syn/anti mixture. [37] While Tykwinski was able to prepare isomerically-pure syn-thiophene [38] and syn-benzothiophene [39] regioisomers, their syntheses were considerably more involved. Importantly, none of the routes are amenable to installing different heterocycles on either side of the aromatic core, thus highlighting the precision synthetic chemistry used to construct 11-14.

Results and Discussion

Molecular design and synthesis. The synthesis of symmetric IFs and their benzoheterocycle analogues (1-6, Figure 1a) typically begins with a Pd-mediated Suzuki cross-coupling of arylboronic acids or arylboronate esters to diethyl 2,5dibromoterephthalate. This modular route is readily adaptable to desymmetrization, as we previously disclosed using diethyl 2-bromo-5-chloroterephthalate (19) with Pd(PPh 3 ) 4 as the catalyst to gain regioselectivity in the initial crosscoupling step with phenylboronic acid. [40] For this current family of molecules, the bromo/chloro reactivity difference in the initial cross-coupling step on 19 was insufficient, as a small amount of double-coupled product was also produced, which proved exceedingly difficult to separate from the desired mono-coupled material. Rather, the revised synthesis (Scheme 1) begins with an aromatic Finkelstein reaction to convert 19 to the more reactive diethyl 2-chloro-5-iodoterephthalate (20). Suzuki cross-coupling with either 3-( 15) or 2-benzothiopheneboronic acid pinacolate ester (16) using third generation Buchwald pre-catalyst yielded chlorides 21 and 22, respectively, with no evidence of double-coupled material. Each chloride was then crosscoupled with 3-benzofuranboronic acid pinacolate ester (17) or 2-benzofuranboronic acid (18) to furnish diesters 23-26 in moderate to good yields. Each diester was subsequently saponified and subjected to Friedel-Crafts acylation to afford the poorly soluble diketones 27-30. Nucleophilic addition of the mesityl units to each dione introduced the protecting groups for the apical carbons, followed by a SnCl 2 -mediated reductive dearomatization to give the four asymmetric IBFBT regioisomers 7-10. Because of the facile decomposition of anti-IDBF 6 to its ring-opened form, [19] we were unsure of the stability of anti/anti-IBFBT 10; however, inclusion of one benzothiophene provides a stabilizing effect. All four parent isomers are stable for months at À20 °C and up to several weeks at 20 °C. IBFBTs 7-10 were subsequently subjected to mCPBA oxidation to yield their sulfone analogues (IBFBTSs) 11-14 in modest yields. All four sulfones are acid sensitive, which is partially responsible for the lowered yields. Similar to anti-IDBF 6, anti/anti-IBFBTS 14 began to decompose upon oxidation and is extremely acid sensitive, as the stabilizing effect of the thiophene in 10 is significantly reduced upon oxidation.

Solid-state structures and bond length alternation analysis. Several groups have utilized asymmetry as a means of creating a dipole moment and exploiting this property to influence solubility, solid-state packing, and polarizability; [41,42] thus, we sought insight into the molecular structures of the asymmetric s-indacene derivatives via single-crystal X-ray diffraction (XRD). Slow evaporation of a CHCl 3 solution of 10 gave deep violet crystals, whereas layering pentanes over CH 2 Cl 2 provided dark blue crystals of 8. [43] The XRD structure of 10 (Figure 3a) revealed positional disordering of the O/S atoms in a ratio of 58/42; no such disordering was observed in 8 (Figure 3b). Whereas 2 packs in a herringbone fashion, the molecules of 10 form a 1D chain (Figure 3c), similar to what was found in 6. The distance between the mean planes of 10 is 3.48 Å, which is slightly shorter than in 6 (3.55 Å). Molecules of 8 also form a 1D chain with a slight curvature of the molecule. The distance between the plane of the six-membered rings on the furan side is 3.57 Å and is 3.33 Å on the thiophene side. The syn/anti-fusion of 8 results in a weak dipole moment toward the side of the molecule with both heteroatoms, affording an anti-parallel packing pattern (Figure 3d).

One must often rely on computationally-optimized geometries in the absence of XRD data, such as this study (Table S2), making the choice of DFT functional critically important (see experimental versus computational comparison in Table S1). [44] Bond length alternation (BLA) is defined as the difference between the average lengths of

Angewandte

Chemie adjacent C=C and CÀC bonds and is an important parameter in assessing both antiaromaticity and ICT. Antiaromatic molecules are often characterized as highly bond alternant because of the instability of a delocalized structure. While pronounced BLA is a characteristic of antiaromatic molecules, it is important to remember that the degree of bond alternation corresponds to the depth of the energy well of the valence tautomers/the size of the activation barrier (E a ). [45,46] The size of E a is proportional to the degree of pseudo-Jahn-Teller-distortion (PJTD), which depends on a multitude of factors including the antiaromaticity of the delocalized state, planarity, symmetry of the molecule itself, aromaticity of the stabilizing aromatic groups, and in this particular case, charge transfer. [47][48][49] As seen in Table 1 and Figure S7, IBFBTs 7-9 are highly bond alternant, typical of the heterocycle-fused s-indacenes, while 10 is decidedly less so. IBFBTSs 11 and 13 both exhibit a significant decrease in bond length alternation upon oxidation, while 12 sees only a slight decrease. Compound 14 is more than twice as bond alternant as its unoxidized analogue 10, likely due to the strong ICT present in 14. Higher BLA values (~0.1 Å) trend with stronger charge transfer and more quinoidal species. [50,51] Perhaps the most meaningful information we can gather from the calculated bond lengths is the fact that all four IBFBTSs (11-14) are "bond flipped" (Table S2 and Figure S7), i.e., the bond alternation in 11-14 is opposite the BLA pattern observed across the family of aryl-and heteroaryl-fused s-indacenes (e.g., 1-2 and 5-10, among others). The only prior "bond flipped" exceptions were IDBTSs 3 and 4. [18] Typically, one of the bond localized valence tautomers of an antiaromatic molecule is in a lower energy structure than the other (due to the PJTD known to occur in 4n π-electron antiaromatic molecules, caused by the singly degenerate to doubly degenerate H!L transition), [5,47] and is further from the delocalized state energetically, "less antiaromatic". It is noteworthy that this effect is consistent across the symmetric and asymmetric series of sulfones (3, 4, 11-14), and with no other s-indacene derivatives. A plausible explanation for this phenomena comes from the reduced aromaticity of the benzothiophene-S,S-dioxide relative to thiophene, a concept that has been shown to decrease the E a between the two valence tautomers in antiaromatic molecules. [48] This in turn facilitates flipping of the bond alternation pattern within the indacene core to afford a molecule with less pronounced, i.e., reduced, antiaromaticity. [18] 1 H NMR Spectroscopy. The relative strengths of the paratropic ring current in antiaromatic molecules can be assessed by monitoring the upfield shift of key protons in the 1 H NMR spectra. Although the identity of the R groups on the apical carbons can affect the shift of the protons on the central six-membered ring of the s-indacene core, we can compare the 1 H NMR data of 7-14 with the symmetric IDBF, IDBT, and IDBTS derivatives (1-6) since all 14 compounds bear mesityl or "mesityl-like" 4-t-butyl-2,6dimethylphenyl substitution.

In prior studies, we ranked relative ring current strength by comparing the chemical shift of the two-proton core singlet on similarly substituted molecules (e.g., syn-fused vs anti-fused). Because of the IBFBT isomers are unsymmetrical, the two core protons now appear as individual singlets in the NMR spectra (Figure 4). The midpoint between the two core peaks for IBFBTs 7-10 and IBFBTSs 11-14 falls approximately halfway between each of the two symmetric analogues 1-6 (Figure S32). The trend in increasingly upfield-shifted peaks is supported by the Nucleus Independent Chemical Shift (NICS) calculations, [52,53] where we observe increasing NICS values (see below) of the center ring (7 > 8 > 9 > 10). The core proton peaks of IDBTS isomers 11-14 shift further upfield for isomers with antifused sulfones (12 and 14) and downfield for the isomers with syn-fused sulfones (11 and 13), consistent with the observed behavior in symmetric IDBTSs 3 and 4. [18] Notable differences between the parent isomers and their corresponding sulfones include (a) the two core proton peaks in 8 and 10 converge to a singlet (2H) for 12 (5.64) and 14 (5.99 ppm), and (b) the difference between core peaks in 7 and 9 increases by 0.23 and 0.08 ppm in 11 and 13, respectively, all data in CD 2 Cl 2 . It should be noted that the core protons of 8 and 10 do split into two singlets in other NMR solvents of varying polarity.

NICS-XY Scans. The magnetic property associated with antiaromaticity is the strength of the paratropic ring current present in an applied magnetic field. The compound with the most antiaromatic character of a set should have the strongest paratropic ring current and therefore the most upfield shifted core protons and most positive NICS value. When comparing NICS values and experimental H NMR shifts, we typically focus on the NICS value of the center 6membered ring because that is the ring with the core protons and is the center of the paratropic ring current; however, averaging the NICS values of the whole s-indacene core (rings 3-5) trends the same way (Table 2). The trends in the structural and magnetic properties for IBFBTs 7-10 experimental [17,19] and computational work, [54,55] which showed that syn-fusion results in increased paratropicity of the s-indacene compared to anti-fusion of the same heterocycle for π-donors (see page S23 for more details). We expected syn/anti-8 to be more paratropic than anti/syn-9 because benzofuran-fused 5 is significantly more paratropic than benzothiophene-fused 1, and 10 to be the least paratropic with both heterocycles anti-fused. The relative antiaromaticities of 7-10 are illustrated by the NICS-XY plots [56][57][58] (Figure 5a) and corroborate our expectations, with all but IBFBT 10 more paratropic than s-indacene itself. [54] Due to the overlapping electronic effects of antiaromaticity and ICT through the s-indacene core, interpretation of the NICS plots for 11-14 (Figure 5b) is less straightforward. Based on prior studies, we expected 12 to be the most antiaromatic since IDBTS 3 and 4 show a reversed trend where the anti-fused isomer is more antiaromatic than the syn isomer, and 13 to be the least antiaromatic. IBFBTS 11 and 14 were expected to fall in between 12 and 13, but the Table 2: Comparison of the NICS values (ppm) of the rings based on their optimized M11/6-311 + G** geometries and experimental 1 H NMR core proton shifts (ppm).

Ring [a,b] (ppm) Core Avg. Core Proton Shift [c] cmpd 1 2 3 4 5 6 7 (ppm) (ppm) s-Indacene 15.1 13.2 15.1 14.5 6.59 [d] syn-IDBT 1 À15.0 À4.2 16.3 14.3 16.3 À4.2 À15.0 15.6 6.06 [e] anti-IDBT 2 À15.5 À4.0 14.1 11.7 14.1 À4.0 À15.5 13.3 6.11 [e] syn-IDBTS 3 À15.0 0.8 10.0 8.5 10.0 0.8 À15.0 9.5 6.91 [f ] anti-IDBTS 4 À15.7 0.4 13.3 14.0 13.3 0.4 À15.7 13.5 6.01 [g] syn-IDBF 5 À14.4 À4.0 19.6 18.2 19.6 À4.0 À14.4 19.1 5.60 [h] anti-IDBF 6 À14.8 À3.7 16.4 13.1 16.4 À3.7 À14.8 15.3 6.14 [h] syn/syn-7 À14.5 À4.3 17.7 16.2 18.0 À4.3 À15.0 17.3 5.85 syn/anti-8 À15.4 À3.5 18.3 14.6 15.0 À4.9 À14.6 15.9 5.87 anti/syn-9 À15.1 À4.6 14.2 13.8 18.8 À3.7 À14.8 15.6 6.05 anti/anti-10 À15.5 À3.8 14.4 12.4 16.1 À3.7 À14.9 14.3 6.10 syn/syn-11 À11.8 2.1 16.6 13.9 14.9 0.3 À15.7 15.1 6.17 syn/anti-12 À16.0 0.3 16.1 17.9 20.6 3.4 À11.4 18.2 5.64 anti/syn-13 À12.9 1.8 15.0 9.5 9.3 0.5 À15.1 11.2 6.50 anti/anti-14 À15.4 0.7 9.8 12.1 18.4 2.7 À12.6 13.4 5.99

[a] Throughout this manuscript we have adopted the convention where ring 2 is always the furan ring and ring 6 is always the thiophene or thiophene-S,S-dioxide ring for unsymmetrical structures 7-10 and 11-14, respectively, as illustrated in Figure 5. [c] Experimental core proton shift (ring 4) in CD 2 Cl 2 ; mesityl R group unless otherwise noted. In the case of two core proton signals, the value reported is the average. [d] Compound 1 f in reference [59].

[e] Reference [17].

[f ] Reference [18]; NMR in CDCl 3 and 4-t-butyl-2,6-dimethylphenyl R group.

[g] Reference [18], NMR in CDCl 3 .

[h] Reference [19].

strengths of the effect of the syn-furan vs. the antibenzothiophene-S,S-dioxide, with the potential for ICT in 14, made the exact order more difficult to predict. The NICS align with predicted relative paratropicities based on previously synthesized symmetric analogues 1, 2, 5, and 6. We anticipated syn/syn isomer 7 to be the most antiaromatic based on prior data as well as 1 H NMR shifts in general corroborate this predicted order; however, they do not perfectly agree with each other. Ring 4 (R4) of 11 is more positive than R4 of 14 in the NICS calculations, but the core proton signals for 14 are significantly more upfield than for 11 (Table 2). NICS scans, when interpreted through the assumption that the absolute value of each ring center linearly correlates with relative electron density (i.e., the larger the value (ppm), the greater the electron density), illustrate the relative ICT character strength across the four sulfones (Figures 5c-5d). The R 2 values for 12 and 14 are much higher than 11 or 13, indicating a much more linear distribution of charge in the two isomers that exhibit charge transfer. The slope of the core (R3!R5) indicates strength of the charge transfer, and trends with decreasing transition dipole moments (11 > 13 > 12 > 14, Table S3) and observed solvatochromic shift (see below). While this trend does not align perfectly with the trend in increasing NICS values (13 < 11 < 14 < 12), or decreasing 1 H NMR ppm (13 > 11 > 14 > 12). All three trends rank 12 and 14 as first and second greatest antiaromatic character, in alignment with trends in ICT character (observed and predicted) across the four isomers (Figure S30). IBFBTs 12 and 14 both display a positive slope from R3!R5, due to the pull of the anti-fused benzothiophenesulfone acceptor (R6) from the furan donor (R2). The slope of 14 is greater because the furan is antifused (better donation, less antiaromatic R3 vs syn-fused in 12, larger R3 value; see resonance structures in Figure 7). Regardless, the similar values of R5!R7 of 12 and 14 imply anti-fusion of the benzothiophene-S,S-dioxide acceptor is the driving force behind the ICT in these molecules. This is corroborated by DFT generated molecular orbitals, which show clear orbital overlap from the s-indacene core to the sulfone moiety in the LUMO of 12 and 14, but not 11 and 13 (Figures S12-S15). Compounds 11 and 13 both display a negative slope from R3!R5, implying the furan can still donate into the core, but the syn-fused sulfones do not withdraw. The steeper negative slope from R3!R5 of 13 is explained by the stronger donation of the anti-fused furan relative to the syn-fusion in 11. Additional NICS comparison plots of each isomer pre-and post-oxidation can be found in Figures S24-S27.

Optoelectronic properties. All asymmetric parent IBFBTs 7-10 and their corresponding sulfones 11-14 possess strong absorptions in the range of 300-400 nm and lowenergy absorptions between ~600-700 nm (Figure 6). According to gas-phase TD-DFT calculations (Figures S8-S11, Table S3), the low energy peaks of 7-10, 12 and 14 are primarily attributed to the HOMO-1!LUMO transition, as expected due to the prototypical forbidden HOMO! LUMO transition in antiaromatic molecules. [58] Oxidation of the syn-fused benzothiophene in 11 and 13 (and 3) results in a significant drop in antiaromatic character, causing the HOMO and HOMO-1 to invert (Table S4 Figures S16, S17), allowing the HOMO-LUMO transition to occur. This is a Angewandte Chemie known occurance in weakly antiaromatic molecules. [60] The difference between the parent and sulfone of the same isomer is the low energy absorption band, where a significant bathochromic red-shift (30-40 nm) is observed for all sulfones, indicative of D/A character imparted by the oxidation of the thiophene ring. Molecules 7 and 8 exhibit the largest (40 nm) shift from 630 nm to 670 nm for 11 and 12, followed by 10 at 605 nm to 14 at 644 nm (39 nm shift). IBFBT 9 (615 nm) shifts by 30 nm to its corresponding sulfone 13 (645 nm). In contrast, the symmetric IDBTSs 3 (624 nm) and 4 (587 nm) are hypsochromic blue-shifted relative to their respective unoxidized analogues 1 (626 nm) and 2 (618 nm), [18] which further confirms that the redshifted absorption in 11-14 is due to ICT across the sindacene core and not merely to the presence of the sulfone moiety. Finally, the molar absorptivities of 11 and 13 are considerably larger than those of 12 and 14, which is consistent with the calculated transition dipole moments (Table S3), indicating much stronger charge transfer in 12 and 14. Plots of the frontier molecular orbitals (Figures S12-S15) illustrate the spatial similarities between the isomers that exhibit charge transfer (12, 14) vs. isomers that do not (11, 13).

As noted earlier, the aromatic analogues of 11-14 that possess naphthalene or anthracene as the core motif are unknown. Nonetheless, we computationally investigated both the syn-and anti-isomers of a linearly benzofuran/ benzothiophene-fused naphthalene pre-and post-oxidation and note some striking differences. Whereas the calculated HOMO-LUMO energy gap in general decreases upon oxidation to 11-14, this gap increases significantly (~0.3 eV) in the naphthalene model compounds (Tables S5-S6), which is reflected in the predicted 30-40 nm hypsochromic shift of the lowest energy optical transition. In addition, the calculated frontier molecular orbital plots (Figures S21-S22) show little to no evidence of ICT. Whereas the antiaromatic s-indacene core facilitates ICT, an aromatic naphthalene would seem to impede this behavior.

ICT often results in a bathochromic shift in more polar solvents due to the stabilization of the more polar chargetransfer/zwitterionic state. This can be seen clearly for both isomers with anti-fused benzothiophene-S,S-dioxides (12 and 14, Figure 7), which show significant 29 and 41 nm shifts, respectively, from hexanes to DMF. IBFBTSs 11 and 13 both exhibit more modest bathochromic shifts of ~10 nm in the same solvents. IBFBTS 11 has significant antiaromatic character but does not exhibit strong ICT. This can be explained by the zwitterionic resonance structure (Figure 7a) in which the fused benzene on the benzothiophene-S,S-dioxide would have to sacrifice its aromaticity. Conversely, IBFBTS 14 is the most solvatochromic, showing a 25 nm shift just changing from hexanes to DCM, likely due to the stability of the pro-aromatic charge transfer resonance structure (Figure 7d).

Electrochemistry. To assess the redox properties of these molecules, cyclic voltammetry (CV) was carried out in CH 2 Cl 2 with anhydrous Bu 4 NPF 6 . Solution-state CV measurements of the reduction and oxidation potentials provide approximate experimental values for the ionization potential (IP) and electron affinity (EA) of a molecule, of which the difference is referred to as the fundamental gap (E fund ), which is known to trend linearly with HOMO-LUMO gaps in PAHs. For the sake of simplicity and consistency, we will refer to these measurements as the HOMO-LUMO gap (E gap ). [61][62][63] Comparison between the TD-DFT predicted HOMO-LUMO gap and electrochemical HOMO-LUMO gap for each compound can found in Table 3.

All four IBFBTs (7-10) display amphoteric redox properties typical of antiaromatic molecules, with potentials for each event at approximately the average value of their respective symmetric counterparts (Table S7). Consistent with previous findings, decreasing reduction potentials trend with increasing antiaromaticity. [12] All four isomers (7-10) have a reversible first oxidation and first reduction. The isomer with the strongest antiaromatic character, 7, is the only molecule without a second oxidation in the DCM solvent window, whereas the isomer with the least antiaromatic character, 9, is the only one of the set without a second reduction in the DCM solvent window.

Electrochemical data for 11-13 (Figure 8) display an increase in the oxidation and reduction potentials for all isomers post oxidation, a decrease in the electrochemical gap for 11 and 12 (relative to 7 and 8, respectively), and an increase in the electrochemical gap for 13 relative to 9. Interestingly, the second oxidation event present for both 7 and 8 disappears post oxidation but is maintained for 9. It is also worth noting the marked difference between 11-12 and IDBTSs 3-4. Whereas both the oxidation and reduction potentials increase by the same amount for 3 and 4, affording an E gap similar to that found in 1 and 2, [17,18] the reduction potentials of 11 and 12 increase twice as much as the oxidation potentials relative to 7 and 8 resulting in the observed smaller E gap . Finally, the non-Faradaic capacitance for 11 and 12 is much larger than their unoxidized counterparts, implying an increased ability for the molecules to carry charge in their oxidized forms. As already noted, 14 was too unstable for accurate electrochemical measurement.

Antiaromaticity. The relative degree of antiaromatic character of a molecule is generally evaluated using three criteria: structural, magnetic, and energetic. The primary structural property associated with antiaromaticity is BLA. Compounds 7, 8, 9, 12, and 14 are strongly bond alternant, typical of antiaromatic molecules. The trends in relative antiaromatic character for the set of parent isomers are straightforward and as expected. The trend in proton NMR data, NICS scans, optical gap, electrochemical gap, and calculated HOMO-LUMO gaps agree within the margin of error for each measurement. As predicted, 7 is the most antiaromatic, then 8, followed by 9, and then 10, by all measures. Based on NICS alone, 12 is the only isomer that shows an increase in antiaromatic character upon oxidation, while NMR implies 12 and 14 both do. Absorption spectroscopy reveals a red shift for all four isomers postoxidation, which can be attributed to the introduction of D/ A character in the molecule, but obscures analysis of relative antiaromaticities. Electrochemical gaps decrease for 11 and 12 relative to 7 and 8 but increase slightly for 13 relative to 9. The exact trend in relative antiaromaticities for the sulfones is less clear due to the effect ICT has on half the isomers. IBFBTS 12 is clearly the most antiaromatic, and 13 the least. Molecules 11 and 14 fall in the middle, but the strong ICT in 14 and lack thereof in 11 make a fair comparison between the two difficult.

Conclusions

Four asymmetric benzofuran-benzothiophene-fused s-indacenes 7-10 have been synthesized and were further modified via a late-stage oxidation to afford the corresponding antiaromatic p-bridged donor-acceptor molecules 11-14. The IBFBT parent compounds display a fine-tuned control of the molecular properties of s-indacene molecules, as by all measures they are an average of their previously reported symmetric counterparts. Two new fusion orientations (syn/

Table 3: Electrochemical Values. [a] cmpd E red2 (V)

E gap (eV) [b] E gap (eV) [c] 7 À1.29 À0.61 0.98 -1.59 1.60 8 À1.36 À0.75 0.89 1.61 1.63 1.67 9 -À0.88 0.73 1.40 1.61 1.65 10 À1.76 À0.83 0.62 1.10 1.69 1.62 11 À1.25 À0.24 1.16 -1.40 1.39 12 À1.15 À0.32 1.08 -1.43 1.37 13 -À0.53 1.12 1.49 1.65 1.59 14 -----1.65

Angewandte

Chemie anti and anti/syn) are possible with this synthetic approach, which result in anti-parallel packing in the solid state. Upon oxidation, two the four isomers exhibit moderate intramolecular charge transfer (12 and 14). This is significant because benzofuran is not a strong donor. The fact that evidence of ICT is observed in these molecules speaks to the conductivity of the s-indacene core, which is an area of active investigation. [64][65][66][67] Somewhat surprising, such structures with different benzoheterocycles fused to aromatic motifs such as naphthalene (10 π-electrons) or anthracene (14 π-electrons), most comparable to 12 π-electron s-indacene, are unknown. TD-DFT calculations on such hypothetical acene analogues suggest that ICT would likely be absent in these molecules (Table S5-S6 and Figures S21-S22). While IBFBTS 12 appears to be more antiaromatic than 14 by all comparative measures, we believe 14 exhibits the strongest charge transfer because of its pro-aromatic resonance form. While it is difficult to define the correlation between antiaromaticity and charge transfer, based on the most reliable indication of each, the two isomers with the furthest upfield core protons exhibit the most dramatic solvatochromic shifts by a significant margin, in both measures.