Electron-donating 1,1,4,4-tetraarylbutadiene is a representative electrochromic dye, and the same chromophore can be found in 9,10-bis(diarylmethylene)-9,10-dihydrophenanthrene (dibenzo-oQD) albeit in a fixed s-cis geometry. Unlike thermodynamically unstable 7,7,8,8-tetraaryl-o-quinodimethane, spontaneous electrocyclization is prohibited by dibenzo-annulation. Several derivatives of dibenzo-oQD were successfully generated despite highly strained geometry caused by steric hindrance between the bulky diarylmethylene units. Their precursors are phenanthrene-9,10-diyl bis(diarylmethylium) dyes (PDM2+) stabilized by four electron-donating alkoxy groups. The redox pairs of dibenzo-oQD/PDM2+ exhibit vivid change in color upon redox interconversion (electrochromism). Both dibenzo-oQD and PDM2+ adopt a helical conformation, whose configuration is unstable. When a chiral alkoxy substituent is attached on each of the aryl groups, the point chirality is successfully transmitted to helicity of PDM2+. Resulting diastereomeric biasing is the key for dibenzo-oQD/PDM2+ to exhibit the two-way-output response (e.g., UV–vis and CD). In addition, much more strained quinodimethane derivatives, 1,2-bis(diarylmethylene)acenaphthene and 1,16-diaryldibenzo[b,n]perylene, were also generated from the dicationic precursors, demonstrating that the reductive transformation can serve as a useful protocol to generate severely deformed π-conjugated systems.
Recently, much attention has been focused on organic π-electron systems exhibiting electrochromism , by which electrochemical input is transduced into UV–vis spectral output. Advanced features would be attained when the electrolysis induces spectral changes other than UV–vis (Scheme 1) . The basic requirement is that both of the neutral and ionic states are stable enough to assure reversibility of the response. From this point of view, o-quinodimethane (oQD)-type cross-conjugating system is attracting since a new Clar’s sextet is generated upon electron-transfer. Thus, the redox reactions would be facilitated, and the ionic states would be also stabilized. Further stabilization is expected in the ionic states when the proper substituents are attached on the exocyclic carbons. 1,1,4,4-Tetrakis(4-methoxyphenyl)-1,3-butadiene 7a  exhibits electrochromic behavior with vivid change in color, thanks to highly colored bis(methoxyphenyl)carbenium dye in the stable cationic states. By framing the same diene unit into the oQD skeleton, the two terminal chromophores are forced to stay in proximity (Scheme 2). Such geometrical feature would provide a chance to produce novel functions other than ordinary electrochromism. Thus, 7,7,8,8-tetraaryl-o-quinodimethane (Ar4oQD) can serve as a leading motif for developing advanced electrochromic material.
The parent tetraphenyl derivative (Ph4oQD) was successfully generated by Quinkert et al.  as a first member of this severely deformed oQD through photoinduced decarbonylation of 1,1,3,3-tetraphenyl-2-indanone. However, spectroscopic characterization of Ph4oQD could be conducted only at low temperatures because of its facile isomerization to benzocyclobutene or dihydroanthracene derivatives via thermal electrocyclization (Scheme 3). By considering that electrocyclization is also induced by photoirradiation, the photochemical reaction would not be the best method to generate the target molecule. So that, to make Ar4oQD applicable in developing electrochromic systems, there are two major points to be concerned: one is suppression of the spontaneous transformation of Ar4oQD into the cyclized isomers, and the other is development of a non-photochemical method to generate the highly strained Ar4oQD molecules.
We envisaged that dibenzo-annulation would destabilize the cyclized isomers in terms of the Clar’s sextet as well as steric repulsion, leading us to design 9,10-bis(diarylmethylene)-9,10-dihydrophenanthrenes (1) as the first isolable members of Ar4oQD . Two of the four aryl rings in neutral 1 are forced to overlap in proximity resulting in its helical geometry, whereas the steric repulsion can be released by twisting the diarylmethylium units around the exocyclic bonds against the planarized phenanthrene core in the corresponding dication 22+ (Scheme 4). So, phenanthrene-9,10-diyl bis(diarylmethylium)s 22+ were selected as the precursors of dibenzo-oQD derivatives 1. Electron transfer (i.e., 2e-reduction) but not photoreaction would be used as the final key step of the transformation. We have been successfully generated the highly strained hexaphenylethane-type compounds by the reductive transformation of the dicationic precursors . So that, it is highly likely that the severely deformed p-conjugated systems would be also produced by the similar approach. The carbocationic species are generally labile due to their high electrophilicity. However, they can be stabilized by incorporation of suitable electron-donating group, such as an alkoxy group. Once the dicationic precursors such as 22+ could be stabilized, they can work as promising component to realize electrochromism with vivid change color since the donating groups can cause a red-shift of strong absorptions of the dications.
In the synthetic scheme based on reductive generation of 1 from dications 22+, an effective route to the dication is another important issue besides the final key transformation. Similar to the oxidative dimerization studied in the preparation of p-extended TTF derivatives , dications 22+ with four alkoxy groups were successfully obtained from the much less-hindered biphenyl donors 8 (Scheme 5). They were in turn facilely prepared by the reactions of α,α′-dilithio-2,2′-dimethylbiphenyl 9 and various diarylketones 10 having different alkoxy groups. The dication salts thus generated by “oxidative cylization” exhibit a characteristic strong absorption band in the visible region with the absorption maximum around 500 nm (log ε ≈ 5), as expected.
When dication salt 2a2+(BF4–)2 with four methoxy groups was treated with excess Zn powder, the deep purple color disappeared rapidly. Desired dibenzo-oQD 1a was generated and isolated as stable yellow cubes in 92 % yield. Its thermodynamic stability was demonstrated by quantitative recovery after refluxing for 24 h in toluene. There were no signs of electrocyclization to its isomers. The detailed structural features of the first isolable derivative of Ar4oQD were revealed by a low-temperature X-ray analysis. The most striking feature is the large torsion angle of 63.4(6) ° for the s-cis diene unit. Such deformation is surely induced to avoid the anomalous proximity of the two inner aryl groups, which are overlapped in a face-to-face manner with the closest C–C contact of 3.19 Å (sum of the vdW radii. 3.40 Å) (Fig. 1).
Not only the electron-donating methoxy groups but also the forced π–π interaction seem to be responsible for the raised HOMO level of dibenzo-oQDs 1. Thus, methoxy-substituted derivative 1a undergoes one-wave two-electron oxidation at +0.78 V vs. SCE in MeCN. The reduction potential of the corresponding dication 2a2+ was observed in the far cathodic region (+0.28 V). Such a large shift of redox peaks as well as a one-wave 2e oxidation process are commonly observed in “dynamic redox pairs”  undergoing drastic structural changes upon electron transfer. When an MeCN solution of 1a was subjected to a constant current electrolysis, a deep purple color characteristic to 2a2+ developed gradually. A continuous change in UV–vis spectrum was observed as shown in Fig. 2. The presence of several isosbestic points indicates clean conversion as well as negligible steady-state concentration of the intermediary cation-radical species. The latter is favorable in terms of high reversibility of electrochromic response. In this way, we could demonstrate that newly designed dibenzo-oQDs 1 were endowed with thermodynamic stability, and thus they can serve as a new entry into the electrochromic systems with vivid change in color and bistability .
Since the redox active chromophore is framed in an s-cis manner in the oQD skeleton, both neutral (1) and dicationic (22+) species are sterically congested molecules with the chiral element of helicity. If the electrochemical interconversion of optically resolved redox pair 1/22+ could be conducted without loss of enantio purity, the pair can work as the less-explored electrochiroptical material , by which the electrochemical input is transduced not only into UV–vis but also chiroptical outputs such as circular dichroism (CD) (Scheme 1b). However, a VT-NMR analysis indicated that the enantiomers of 1/22+ are interconverting easily even at room temperature.
Thus, we planned to adopt another approach based on rapid interconversion of helical sense by attaching a chiral handle on each of the aryl groups (Scheme 6). When the intramolecular transmission of the point chirality to helicity occurs effectively , there would be a chance to generate helically biased redox pair, that would give strong enough chiroptical signals to be used as an output of molecular response systems. As for the chiral handle on each of the aryl groups, (R)-sec-butyl and (R)-2-octyl were selected for their accessibility.
Dibenzo-oQDs 1b,c with the chiral handles were prepared from the corresponding (R,R)-4,4′-dialkoxybenzophenones 10b,c via dications 2b,c2+. Although 1b,c with asymmetric centers exist as mixtures of diastereomers with an opposite sense of helicity, they were treated as single entities due to facile interconversion of the diastereomers. The 1H NMR spectra of diastereomeric donors were barely distinguishable, but the presence of both diastereomers was unambiguously confirmed by X-ray analysis of 1b composed of an equal amount of (P,R,R,R,R)- and (M,R,R,R,R)-diastereomers. The presence of both diastereomers in the crystal lattice is a rare observation, and it means that the diastereomeric excess (de) of 1b is 0 % in the crystalline state.
We also succeeded in analyzing X-ray structure of the corresponding dication salt 2b2+(BF4–)2. Different from neutral 1b, only one of the two diastereomers with P-helicity is present in the crystal. Thus, dication 2b2+ is 100 % de in the crystalline state. The CD spectrum measured on a KBr disk exhibits a negative couplet around 550 nm, which likely comes from the exciton coupling of two diarylmethylium dye units arranged in a P-helical fashion. Albeit less complete, preference for (P,R,R,R,R)-diastereomer is expected also in solution. In fact, a CH2Cl2 solution exhibits similar bisignated Cotton effects with large amplitude [λext 591 nm (Δε –23), 520 (+17)]. Although the helicity inversion between (P,R,R,R,R)- and (M,R,R,R,R)-2b2+ is too fast even at low temperature to determine the de value by 1H NMR spectroscopy, the observed CD signal is large enough to use as an additional output against the electrochemical input. A large couplet was also observed for 2c2+ with (R)-2-octyl groups [589 (–18), 519 (+15)].
When the electrolyses of dibenzo-oQD 1b,c with asymmetric centers on the aryl groups were conducted, the electrochemical input was transduced not only into UV–vis but also into CD spectral changes (Fig. 3) due to strong CD signaling thanks to successful transmission of point chiralities to helicity in 2b,c2+. This is the successful demonstration of two-way-output molecular response by the dibenzo-oQD 1 and the corresponding dication 22+. It should be noted that the monocation with the same chromophore never shows the strong CD couplet: the Cotton effects of (R,R)-bis(4-sec-butoxyphenyl)phenylcarbenium are very weak [508 (–1.3)]. Thus, the exciton coupling of two chromophores closely arranged in a chiral environment is quite effective in attaining strong CD signals .
Successful reductive generation and isolation of dibenzo-oQDs 1 from the dicationic precursors 22+ prompted us to study other strained/deformed quinodimethane derivatives, such as 1,2-bis(diarylmethylene)acenaphthene 3. As in 1 with the dihydrophenanthrene skeleton, 3 also suffers severe steric hindrance since the s-cis diene unit is framed in the ring system. The higher planarity of acenaphthene skeleton than dihydrophenanthrene may also cause larger strain in 3 than in 1.
As in the case of 1, 3a with four methoxy groups was successfully generated from acenaphthylene-1,2-diylbis(diarylmethylium) 4a2+, which in turn was obtained from 1,8-bis(2,2-diarylethenyl)naphthalene 11a by oxidative cyclization (Scheme 7). According to the X-ray analysis, 3a suffers much severer steric repulsion as indicated by the twisting of exocyclic double bonds of 23.5°. The closest C–C contact between the two facing aryl groups is 3.22 Å (Fig. 4). Such a deformation would be responsible for raising HOMO of 3a. So, 3a undergoes 2e-oxidation at the lower oxidation potential [+0.61 V vs. SCE in MeCN] than in 1a. Due to much closer contact of two cationic chromophores, 4a2+ is more electrophilic and labile than 2a2+. However, clean electrochromic behavior was observed as shown in Fig. 5 by using polar but less nucleophilic (CF3)2CHOH as a solvent. The above results indicate that, once successfully generated, the sterically hindered redox pairs have a chance to serve as molecular response systems. Usefulness of the present synthetic approach based on the redox reactions is demonstrated when we found facile degradation of 3a under photoirradiation conditions via electrocyclization (Scheme 8) .
Further challenge was made based on the reductive transformation of dicationic precursors into highly strained electron donors. 1,16-Diaryldibenzo[b,n]perylene 5a was selected as a next target. In this compound, the two aryls groups at the bay region of perylene are further forced to proximity by buttressing effects of the two extra fused-benzene rings. Although the dicationic precursor 6a2+ may suffer from the similar degree of steric repulsion to 5a, we found that oxidative cyclization of 1,1′-biphenanthryl derivative 12a proceeded smoothly to give tetramethoxy dicationic dye 6a2+ as a stable (SbCl6–)2 salt (Scheme 9). In this scheme, biphenyl donor 8a was again used to prepare 12a by photocyclization.
Upon reduction of deep purple powder of the dication salt with Zn in THF at 0 °C, dibenzoperylene 5a was isolated as yellow crystals, whose X-ray analysis revealed largely twisted π-condensed skeleton. The torsion angle of 42.1° for C1-C16b-C16a-C16 shows the severe strain around the bay region, which is induced by two aryl groups facing to each other with the closest C–C contact of 3.00 Å (Fig. 6). The lower electron-donating properties of 5a (Eox: +1.03 V vs. SCE in MeCN) than 1a or 3a could be accounted for by less conjugation of two methoxyphenyl groups with the polycondensed π-framework. Due to slow degradation of 5a to 13a via electrocyclization even at room temperature (Scheme 10), we have not succeeded in obtaining reversible electrochromic behavior. Its reversible electrochemical interconversion with 6a2+ could be realized, if possibly, by conducting its electrolysis at a low temperature .
Photochemical transformation is a powerful tool in obtaining highly strained molecules from the less-hindered precursors, and thus has been used to generate structurally novel aromatic compounds for years . As an attractive alternative, this work proposes usefulness of the redox transformation of dicationic precursors. This protocol is especially effective when the target molecules undergo degradation under irradiation . By attaching electron-donating substituents such as alkoxy groups, the precursors could be stabilized enough to be used as key synthones. At the same time, stability of the dicationic dyes enables reversible redox interconversion between the target molecules and the dications, thus allowing their use as redox-based molecular response systems (e.g., electrochromic systems). In general, preparation of helically deformed quinodimethanes and related polycyclic condensed hydrocarbons  are challenging, however, a new protocol presented here would facilitate their studies for further pursuing highly functionalized materials (e.g., multi-output molecular response systems).
This work was supported by the Cooperative Research Program of “Network Joint Research Center for Materials and Devices” and by Grant-in Aid for Scientific Research on Innovative Areas: “Organic Synthesis Based on Reaction Integration” (No. 2105) from MEXT, Japan. TS thanks JSPS Grant-in-Aid for Challenging Exploratory Research on “Maximum Function on Minimum Skeleton (MFMS)” (No. 25620050).
 (a) P. M. S. Monk, R. J. Mortimer, D. R. Rosseinsky. Electrochromism: Fundamentals and Applications, VCH, Weinheim (1995); (b) P. M. S. Monk, R. J. Mortimer, D. R. Rosseinsky. Electrochromism and Electrochromic Devices, Cambridge University Press, Cambridge (2007). Search in Google Scholar
 T. Suzuki, E. Ohta, H. Kawai, K. Fujiwara, T. Fukushima. Synlett (Account) 851 (2007). Search in Google Scholar
 S. Hünig, M. Kammer, H. Wenner, F. Barbosa, G. Gesheidt, I. F. Perepichka, P. Bäuerle, A. Emge, K. Peters. Chem. —Eur. J.6, 2618 (2000). Search in Google Scholar
 G. Quinkert, W. -W. Wiersdorff, M. Finke, K. Opitz, F.-G. von der Haar. Chem. Ber. 101, 2302 (1968). Search in Google Scholar
 S. Iwashita, E. Ohta, H. Higuchi, H. Kawai, K. Fujiwara, K. Ono, M. Takenaka, T. Suzuki. Chem. Commun. 2076 (2004). Search in Google Scholar
 (a) A. Ohta, Y. Yamashita. J. Chem. Soc., Chem. Commun. 1761 (1995); (b) Y. Yamashita, M. Tomura, S. Tanaka, K. Imaeda. Synth. Metal102, 1730 (1999); (c) A. J. Moore, M. R. Bryce, P. J. Skabara, A. S. Datsanov, L. M. Goldenberg, J. A. K. Howard. J. Chem. Soc., Perkin Trans. 1 3443 (1997). Search in Google Scholar
 (a) J. Nishida, T. Suzuki, M. Ohkita, T. Tsuji. Angew. Chem., Int. Ed.40, 3251 (2001); (b) H. Higuchi, E. Ohta, H. Kawai, K. Fujiwara, T. Tsuji, T. Suzuki. J. Org. Chem.68, 6605 (2003); (c) T. Suzuki, Y. Ishigaki, T. Iwai, H. Kawai, K. Fujiwara, H. Ikeda, Y. Kano, K. Mizuno. Chem. —Eur. J.15, 9434 (2009). Search in Google Scholar
 (a) T. Suzuki, T. Iwai, E. Ohta, H. Kawai, K. Fujiwara. Tetrahedron Lett.48, 3599 (2007); (b) T. Suzuki, S. Tanaka, H. Kawai, K. Fujiwara. Chem. Asian J.2, 171 (2007); (c) T. Suzuki, K. Ohta, T. Nehira, H. Higuchi, E. Ohta, H. Kawai, K. Fujiwara. Tetrahedron Lett.49, 772 (2008). Search in Google Scholar
 T. Suzuki, Y. Sakano, T. Iwai, S. Iwashita, Y. Miura, R. Katoono, H. Kawai, K. Fujiwara, Y. Tsuji, T. Fukushima. Chem. —Eur. J.19, 117 (2013). Search in Google Scholar
 T. Suzuki, S. Iwashita, T. Yoshino, H. Kawai, K. Fujiwara, M. Ohkita, T. Tsuji, K. Ono, M. Takenaka. Tetrahedron Lett.47, 467 (2006). Search in Google Scholar
 S. Iwashita, T. Suzuki. J. Synth. Org. Chem. Jpn. 64, 958 (2006). Search in Google Scholar
 (a) B. Halton. Advances in Strained and Interesting Organic Molecules, Vol. 7, JAI Press, Stanford, CT (1999); (b) R. Gleiter, H. Hopf. Modern Cyclophane Chemistry, Wiley-VCH, Weinheim (2004). Search in Google Scholar
 (a) T. Kawase, T. Fujiwara, C. Kitamura, A. Konishi, Y. Hirao, K. Matsumoto, H. Kurata, T. Kubo, S. Shinamura, H. Mori, E. Miyazaki, K. Takimiya. Angew. Chem., Int. Ed.49, 7728 (2010); (b) D. T. Chase, B. D. Ross, S. P. McClintock, L. N. Zakharov, M. M. Haley. Angew. Chem., Int. Ed.50, 1127 (2011); (c) A. Shimizu, Y. Tobe. Angew. Chem., Int. Ed.50, 6906 (2011). Search in Google Scholar
©2014 by IUPAC & De Gruyter Berlin/Boston