Synthetic approaches to bowl-shaped π-conjugated sumanene and its congeners

Review

1 Introduction

Over a long period of time, polyaromatic hydrocarbons (PAHs) have attracted a tremendous attention of the scientific community because of their diverse potential applications ranging from the chemistry perspective to materials science and technology [1-4]. As we know that bowled (curved) surfaces are universal in nature for example our planets as well as atomic orbitals possess the curvature which generally affects the charge-transport, redox, self-assembly, and optical properties of bowl-shaped π-conjugated systems [5-9]. The synthesis of π-bowls is an extremely challenging job due to the presence of unusual strain in these types of molecules, therefore, the first synthetic breakthrough in this arena came into the picture in the late twentieth century when for the first time corannulene was reported by Barth et al. [10] at the University of Michigan (USA) in 1966 (crystal structure 1971) [11]. On the other hand, since the discovery of fullerene (C₆₀) in 1985 [12] by Sir Harry Kroto and the first synthesis of sumanene in 2003 by Sakurai et al. [13], this field of research is continuously booming because of the developments of advanced synthetic organic tools [14-16]. Among the PAHs, the buckybowls are of significant importance not only because of the presence of unique inherent chirality (bowl-chirality) originating either from the bowl structure itself, e.g., in hemifullerene or by the introduction of substituents (e.g., trimethylsumanene) or heteroatoms (e.g., triazasumanene) into the achiral bowls but also as they are partial structures of carbon nanotubes (CNTs) and fullerenes having bowl inversion activities as well as a tendency of crystal packing [17-20]. Since these chiral buckybowls contain stable convex or concave faces suitable for the generation of chiral molecular recognition sites which can be used for the construction of helical assemblies and have been used to coordinate the metal atom(s) [21-23]. On the other hands, further extension of these molecules leads to homochiral carbon nanotubes which can produce innovative perception in chiral sensing, chiral catalysis, separation techniques and chiral ligands for organocatalysis [24]. Additionally, control on the bowl-to-bowl inversion can not only be useful for enantioselective synthesis of π-bowls but also to produce novel building blocks for molecular switches, chemical machines, molecular motor, ferroelectric memories, molecular devices, and sensory materials etc. Interestingly, electronic switching, thermal transport and thermoelectric properties in addition to the onigiri-type core-shell assemblies have been reported for sumanene and its derivatives. More interestingly, its application in the absorption of small molecules such as NH₃, CO₂, CO, and H₂ using density functional theory (DFT) calculations has also been revealed [25-29].

Therefore, the area of buckybowls particularly sumanene chemistry is gaining more and more pace which can be seen by the inspection of a flow of publications appearing in the literature day-by-day [26]. As far as our knowledge is concerned, this is the first comprehensive review article which covers almost all the unsuccessful as well as successful efforts towards the synthesis of sumanene and its congeners. The ball-and-stick representation, bowl fragments and the chemical structures of corannulene (1) and sumanene (2) are displayed in the Figure 1.

2 Synthesis of sumanene and its derivatives

Around three decades before, it was the year 1993, when first time Mehta and his teammates coined the name sumanene for compound 2 from a Sanskrit word ‘Suman’ which means flower and put their unsuccessful effort towards the synthesis for this architecturally interesting molecule from 1,3,5-tris(bromomethyl)benzene (3) using the flash vacuum pyrolysis (FVP) technique as a key transformation (Scheme 1) [30]. As can be seen from an inspection of Scheme 1, they began their journey with the coupling reaction of 3 with 3-butenylmagnesium bromide in the presence of dilithium tetrachlorocuprate (Li₂CuCl₄) to produce the tripentenylbenzene derivative 4 in 45% yield. Alternatively, they have also prepared the same compound 4 starting from a simple and commercially available 1,3,5-trimethylbenzene (mesitylene) by using n-BuLi, and 4-bromo-1-butene in the presence of N,N,N′,N′-tetramethylethylenediamine (TMEDA) as shown in the Scheme 1. Having the compound 4 in hand, it was subjected to the cyclization in the presence of boron trifluoride to provide the tricyclohexyl-fused benzene derivative which on further dehydrogenation with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) afforded 1,5,9-trimethyltriphenylene (6) in 46% yield. Later on, they first directly tried to convert compound 6 into the expected sumanene (2) by using a cyclodehydrogenation reaction in the presence of Pd/C at 400 °C, surprisingly they obtained only mono-bridged compound 7 in around 70% yield along with some of an unidentified mixture of compounds. In sharp contrast, their attempt under flash vacuum pyrolysis (FVP) at high temperature was also fruitless may be due to the presence of three benzylic sp³ carbons which are unable to endure such harsh reaction conditions. Therefore, next they converted compound 6 into 1,5,9-tribromomethyltriphenylene (8) using N-bromosuccinimide (NBS) in the presence of AIBN in CCl₄. To their surprise, when compound 8 was subjected to FVP at around 850 °C temperature, only mono- and dibridged compounds (7 and 9) were isolated in overall 20% yield in (13:87) ratio. The structure of the dibridged compound 9 was not only confirmed by spectroscopic data but also identified by virtue of the single crystal structure. The reason for their unsuccessful results may be the generation of strain in the sumanene molecule from planar aromatic architecture under the experimental reaction conditions.

In the quest for developing the first successful synthetic method, groundbreaking news came from Sakurai’s laboratory around a decade after Mehta’s first unsuccessful attempt for the synthesis of sumanene (2, Scheme 2) [13]. Their exciting non-pyrolytic synthetic pathway for its construction commenced with an easily accessible compound, namely norbornadiene (10) by involving oxidative aromatization in the presence of DDQ via an intermediate 17, as displayed in Scheme 2. As can be seen from an inspection of Scheme 2, they first performed a single step cyclotrimerization of 10 using n-BuLi and t-BuOK in 1,2-dibromoethane followed by the addition of CuI through an intermediate 11. This procedure provided a very low yield (7%) of a mixture of 12 (syn) and 13 (anti) products. Therefore, they opted an alternative route which involves the formation of organotin compound 14 followed by trimerization in the presence of copper catalyst 15 to yield the trimerized products 12 (syn) and 13 (anti) in respectable yields (Scheme 2). The alkene-bridge exchange of 12 (syn) was accomplished by tandem ring-opening and ring-closing metathesis (ROM–RCM) in the presence of Grubbs’ first generation (G-I) catalyst to generate a C₃-symmetric hexahydrosumanene 17 which on subsequent aromatization using DDQ furnished the desired molecule sumanene (2) in good yield. To their surprise, tandem metathesis for achieving compound 17 from 13 (anti) was fruitless may be because of the endothermic reaction by 37.4 kcal/mol as compared to the exothermic (51.4 kcal/mol) transformation of 12 (syn) to 17, calculated by density functional theory (DFT) calculations.

On the other hand, in 2008, Higashibayashi et al. reported the synthesis of first chiral C₃-symmetric trimethylsumanene 28 starting from enantiopure norbornadiene (10) by employing a rational synthetic strategy via the transfer of sp³ chirality of 27 into the bowl chirality of 28 as a key conversion (Scheme 3) [17,31]. In this context, they began with the Pd-catalyzed hydrosilylation reaction using HSiCl₃ at −3 °C in the presence of a chiral phosphine ligand to furnish the hydrosilylated product which on subsequent Tamao–Fleming oxidation provided the exo-diol 18 in an overall good yield with 99% enantiomeric excess (Scheme 3). Furthermore, the diol 18 was converted into the corresponding diketone 19 using pyridinium chlorochromate (PCC) as an oxidizing agent. Interestingly, they have also performed the similar transformation in almost identical yield using Swern oxidation reaction conditions. The diketone 19 was then transformed into the enantiopure iodonorbornanone 22 in three steps which on further regioselective Pd-catalyzed cyclotrimerization furnished the syn-benzocyclotrimer 23 in 55% yield. Next, the methyl substituents were introduced at the position of the carbonyl groups of 23 to generate the methyl-substituted olefin derivative 25 via alkenyl phosphates 24 by means of three-fold cross-coupling reaction with methylmagnesium iodide in THF. Finally, during the tandem ROM–RCM, they noticed that the G-I catalyst provided a mixture of ring-opened products, therefore, the ring-closing step was then carried out using Grubbs’ second generation (G-II) catalyst. Hence, the tandem ROM–RCM step was made successful using both G-I and G-II catalysts in ethylene atmosphere to afford the C₃-symmetric hexahydrotrimethylsumanene 27, which was then converted into the desired compound by aromatization with DDQ. Alternatively, when they used the more active G-II catalyst instead of the G-I and then the G-II catalyst under similar reaction conditions, they directly observed the formation of compound 27 in a single step. They also pointed out that when ethylene gas was used as an alkene source for the ring-opening step, diluted reaction conditions were necessary to dissolve the ethylene in dichloromethane to avoid polymerization. Therefore, after several experimentations, they noticed that liquid Z-oct-4-ene in toluene was found to be a more effective alkene source for the ring-opening reaction in comparison to the gaseous ethylene, as it increases solubility and also improve the lifetime of the catalyst.

2.1 Derivatization at benzylic position

As can be inferred from the architecture of sumanene (2), it is not only gifted with three benzylic positions stapled between the benzene rings of the triphenylene system but also possess the bowl-curvature which enables stereoselective functionalization at these valuable positions. Taking the advantage of these benzylic positions, two years later to the trimethylsumanene synthesis, Sakurai and his co-workers selectively generated mono-, di- and trianions (29–31) using t-BuLi as a base and the formation of these sequential anions were confirmed by ¹H and ¹³C NMR spectroscopy (Scheme 4) [32]. Next, the in situ generated trianion 31 was quenched with an excess of trimethylsilyl chloride (Me₃SiCl) to furnish the only exo-tris(trimethylsilyl) derivative 32 may be because of the unhindered attack from the convex surface selectively in comparison to the hindered concave face (Scheme 4).

Along similar lines, Hirao and his teammates have reported the trideuteriosumanene 33 in a stereoselective manner from the same trianion 31 just by trapping it with CD₃OD/CH₃OD (Scheme 5). In the same paper, they have also reported the hexaallylated sumanene 34 as well as hexa-p-methoxybenzylsumanene 35 by reacting sumanene (2) with allyl bromide and 4-methoxybenzyl chloride (PMB-Cl) in the presence of 30% NaOH_(aq) and tetrabutylammonium bromide (TBAB) in a minimum amount of THF (Scheme 5) [33]. On the other occasion, tribromosumanene 36 has also been synthesized via a radical mechanism from pristine sumanene (2) in the presence of NBS and CCl₄ (Scheme 5) [14].

In another study, the same group has also reported a series of both C₃-symmetric as well as unsymmetric diastereomeric π-conjugated bowl-shaped molecules 37a–f in good to quantitative yields (Scheme 6) [34,35]. These π-conjugated systems were assembled by treating sumanene (2) with different conjugated aromatic aldehydes using 30% aqueous NaOH in the presence of TBAB in THF. Although they have tried several bases for these condensation reactions the aforementioned base was found to be more effective than other bases used (Scheme 6). More interestingly, the sumanene derivative 37b was used as a sensory material for selective recognition of cesium cations in water.

In another event, Hirao’s group has prepared trioxosumanene 40 in the presence of RuCl₃ and t-BuO₂H which could be used as a key building block to generate diverse significant electroactive materials by virtue of nucleophilic addition reactions or by other means. After several experimentations, these workers were able to obtain trioxosumanene 40 in 73% yield (Scheme 7). More interestingly, monooxosumanene 38 was achieved selectively in good yield using phosphotungstic acid and t-BuO₂H as shown in Scheme 7. To their surprise, they were unable to isolate dioxosumanene 39 under similar reaction conditions in reasonable yield. The structures of these compounds were established by spectroscopy means and also in the case of monosumanene 38, they obtained the single crystal structure which was showing almost a similar bowl depth as for the parent sumanene (2). Later on, they transformed trione 40 into the corresponding exo-trimethyl derivative 44 by means of a Grignard reaction with MeMgBr via 1,2-addition (Scheme 7). The imination of monoketosumanene 38 as well as triketosumanene 40 has also been reported by these authors to obtain the corresponding compounds 42 and 43, displayed in Scheme 7 [36-38].

Fascinatingly, the group of Sakurai has revealed the exo-functionalization of trimethylsumanene 28 to produce the products 45a,b stereoselectively, as can be inspected from Scheme 8 [39]. The preparation of the compounds 45a,b were achieved via the generation of trianions at the benzylic positions using LDA followed by the introduction of electrophilic partners such as (S)-MTPA (PhC(CF₃)-(OMe)CO) and SiMe₃ in THF to yield the required compounds in moderate yields (Scheme 8). The trimethylsumanenetrione 46 has also been generated from the corresponding trimethylsumanene 28 by means of nucleophilic oxidation using NaHMDS in the presence of molecular oxygen in DMF as the solvent (Scheme 8).

It has been noticed from the literature that the directly linked π-conjugated systems act as promising electron-accepting materials because of their high LUMO energy level and hence may be gifted to achieve a large open-circuit voltage. To this context, Amaya et al. has developed a novel bissumanenylidene 47 starting from the parent sumanene (2) first by converting it into the monoketosumanene followed by McMurry coupling reaction (Scheme 9). Although, they have attempted various combinations of Ti(IV) and reducing agents but Cp₂TiCl₂ with Zn powder provided the better results compared to the other combinations [14,40]. Additionally, they prepared benzyl–benzyl coupled sumanene dimer 48 by means of oxidative dimerization via sumanenyl monoanion intermediate 29 (Scheme 9) [14].

In 2012, Wu’s team have employed the Pd-catalyzed arylation reaction of sumanene with chlorobenzene and 2-bromo-1,3-dimethylbenzene to afford the aryl-substituted sumanenes 49a (Scheme 10). On the other hand, Higashibayashi et al. has revealed the synthesis of three different mono-substituted sumanene derivatives 49b–d in a stereoselective manner as can be inspected from Scheme 10 [41,42].

In another event, Amaya, Ito, Katoh and Hirao reported a vital building block 53 to extend the π-conjugation bidirectionally through regioselective functionalization (Scheme 11) [43]. To achieve this goal, they commenced with the two-fold Friedel–Crafts alkylation reaction of sumanene (2) with 2,5-dichloro-2,5-dimethylhexane (50) involving AlCl₃ to generate compound 51 which on subsequent oxidation provided triketosumanene 52 (Scheme 11). Finally, compound 52 was reacted with ethylene glycol to afford monocyclic acetal 53 with less hindered keto-group in respectable yield.

Since the introduction of powerful electron-withdrawing groups such as fluorine atom(s) in any material changes its behavior significantly, in this regard, Sakurai’s group installed six fluorine atoms at the benzylic positions of sumanene in a two-step approach to generate hexafluorosumanene 55 in an overall good yield [44]. As can be seen from an inspection of Scheme 12, they first prepared cyclic dithiane 54 from trisumanenone 40 using propanedithiol in the presence of BF₃ since dithianes are outstanding precursors for the installation of fluorine atoms using appropriate fluorinating agents such as (difluoroiodo)benzene derivatives or elemental fluorine. In the second step, when they used para-iodotoluene difluoride the reaction conversion was very low, therefore, when using Olah’s reagent (pyridine hydrofluoride) in the presence of an activating agent, e.g., N-iodosuccinimide (NCS) in a usual glass flask, a mixture of products 55–57 were obtained (Scheme 12). Interestingly, when a polypropylene tube was used instead of glass flask, the expected hexafluorosumanene 55 was obtained in 73% yield.

To explore the chemistry of the sumanene and its congeners, quite recently the same group has also prepared sumanene-based carbene 60 starting from monosumanenone 38 by reacting it with hydrazine hydrate to provide the corresponding hydrazone 58 which on further oxidation with MnO₂ followed by irradiation using an LED lamp at ca. 365 nm for 1 min in 2-methyltetrahydrofuran (mTHF) at 77 K gave 60 (Scheme 13) [45]. It was noticed from both experimental as well as theoretical studies that the ground state of the prepared carbene 60 is a triplet which was confirmed by ESR as well as density functional theory (DFT) calculations. As we are aware that if the carbene formed is a singlet then a C–H bond inserted product is predominating whereas if the dimer is the major product along with the minor C–H bond inserted product then the triplet carbene is generated. During their study, they obtained the C–H inserted product 61 in 65% along with the dimer 62 in 18% yield with cyclohexane, confirming that the reaction precedes via the singlet state, albeit it was a triplet species in the ground-state. They further reported the applicability of the carbene precursor 59 by reacting it with different thiocarbonyl compounds in a Barton–Kellogg coupling manner to yield the corresponding olefins. To this context, the carbonyl species were first transformed into the thiocarbonyl systems in the presence of Lawesson’s reagent which were then reacted with diazosumanene 59 to furnish the corresponding alkene systems. It has been noticed from the literature that the construction of these types of sterically hindered tetrasubstituted alkenes is really difficult through McMurry coupling reaction or by other means (Scheme 14).

2.2 Functionalization at benzene ring(s) bay positions

In another event, the Sakuria group has also reported hydroxysumanene 68 by means of a Baeyer–Villiger oxidation reaction of acyl- or formylsumanene derivatives in the presence of m-chloroperbenzoic acid (m-CPBA) followed by acid-catalyzed solvolysis in 10% HCl/MeOH (Scheme 15) [46]. Interestingly, it was found from both theoretically as well as experimentally that compound 68 has a deeper bowl architecture and also a higher bowl inversion energy as compared to the sumanene (2), measured by NMR studies. In this report, these workers further revealed the electronic state of 68 by means of electrochemical and UV-absorption measurements.

Recently, they have also reported the functionalized sumanene derivatives 73–75 at the convex side of the internal carbon commencing from the bromination of the hydroxysumanene 68 using NBS and subsequent nucleophilic substitution reaction [47,48]. To their surprise, when they used molecular bromine instead of NBS, they obtained o-bromohydroxysumanene 72 instead of 71. More interestingly, it was also noticed that by treating 71 with trifluoroacetic acid (TFA), it instantly transformed into sumanene derivative 72, suggesting that 71 is unstable under acidic conditions. The acid-catalyzed aromatization can be explained via the migration of the bromide ion followed by the aromatization of bromodienone compound 71 to the stable o-bromohydroxysumanene 72. Having compound 71 in hand, next it was treated with silver acetate in H₂O/THF (1:1) to furnish compound 74. On the other hand, alcoholysis of the same compound 71 with methanol provided methoxysumanenone 73 in an overall good yield (Scheme 16). From their studies they observed that compound 74 is unstable to both heat as well as light, may be due to instability of the internal C–O bond. In this regard, very recently they transformed this unstable compound 74 into the stable derivatives 75 having a strong C–C bond through the nucleophilic substitution reaction with phenol and anisole in the presence of trifluoromethanesulfonic acid (TfOH) with total stereoinversion. This suggests that the nucleophile attacks occur from the concave face of the π-bowl [48]. Although, for the nucleophilic substitution reaction with phenol derivatives, they have tried several reaction conditions including the amount of acid as well as phenols. After several experimentations, they found that 30 equivalents of phenol and 1 equivalent of TfOH at 0 °C provided the best results. As can be inspected from Scheme 16, in this report, they have also developed an alternate route for the synthesis of compound 74 involving only one step instead of the previously reported two step approach. The possible mechanism for the nucleophilic substitution at the internal carbon is displayed in Scheme 17.

Amaya et al. in 2009 have revealed the synthesis of monobromosumanene 82 from sumanene (2) by treating it with pyridinium perbromide as displayed in Scheme 18 [49]. In this report, the authors have exposed the anisotropic electron transport properties of the needle-like single crystal of sumanene derivative 82 by means of time-resolved microwave conductivity technique. On the other hand, four years later, Sakurai and his teammates have reported the selective synthesis of diverse monosubstituted sumanene derivatives by employing electrophilic aromatic substitution reactions as they are the most trustworthy strategies for direct functionalization of aromatic scaffolds [50]. As can be inspected from Scheme 18, monoiodosumanene 79 was obtained by gold-catalyzed iodination in the presence of N-iodosuccinimide (NIS). The mononitrosumanene 80 was achieved through the nitration using trifluoroacetyl nitrate which was in situ generated from conc. HNO₃ and trifluoroacetic anhydride. On the other hand, formylation and acetylation were performed under microwave conditions at 130 °C using triflic anhydride and DMF or dimethylacetamide (DMA) to deliver the corresponding monoformylsumanene 66 and monoacetylsumanene 69 in good yields. The monobenzoylsumanene 81 was prepared from benzoyl chloride in the presence of triflic acid. Moreover, disubstituted sumanene derivatives 83–85 were prepared under similar reaction conditions just by increasing the amount of the reagents and easily separable regioisomers were obtained in moderate-to-good overall yields (Scheme 19). In similar fashion, trisubstituted sumanene derivatives were also prepared as shown in Scheme 19.

The monochlorosumanene 88 was also reported by Amaya and Hirao in three steps from the sumanene (2) using a classic nitration reaction followed by the reduction of the nitro functionality into the amine group and subsequently performing a Sandmeyer reaction as shown in Scheme 20 [14]. Additionally, they have reported the hydrogenation of sumanene to generate compound 89 having one benzene ring intact in the framework using Pd/C in the presence of hydrogen gas in toluene at room temperature, confirmed by mass spectral data as well as NMR spectroscopy (Scheme 20) [14].

As can be inspected from Scheme 21, the dimer 90 of sumanene (2) was obtained by two different routes, first by employing a one-pot borylation as well as a Suzuki–Miyaura cross-coupling reaction of bromosumanene 82 under microwave reaction conditions. Whereas another route involves the Ni-catalyzed aryl–aryl homo-coupling between the two molecules of iodosumanene 79 [51,52]. Moreover, the bissumanenyl 92, which is thought to be chiral because of the two connected asymmetric bowls, as well as atropisomerism, was constructed by Hirao and his two group members, namely Amaya and Kobayashi, starting from the same bromo derivative 82. First it was converted into ethynylsumanene 91 using a Sonogashira-coupling which on subsequent desilylation and Glaser-coupling reaction yielded 92 (Scheme 21).

As can be seen from the literature, pyrene-based compounds remain among the highest beautiful and fascinating classes of molecules because of their extremely characteristic (“fingerprint”) optical absorption as well as emission behavior. Keeping in mind the uniqueness of the pyrene moiety, an appealing sumanene derivative 93, namely as pyrenylsumanene, was constructed by the groups of Higashibayashi and Sakurai which display both herringbone and columnar crystal packing (Scheme 22) [53]. To synthesize this architecturally interesting molecule, they began with the monoiodosumanene 79, which was prepared via an alternate route using 6,6’-diiodo-2,2’-dimethoxy-1,1’-binaphthol in the presence of Sc(OTf)₃. The Suzuki-coupling reaction between 1-pyreneboronic acid and iodosumanene 79 furnished the desired compound 93 in 84% yield (Scheme 22).

Among the bowl-shaped π-conjugated molecules, sumanene is more attractive as chemical modifications can be easily performed at both benzylic as well as peripheral aromatic carbons. Although, a handful of reports are available in the literature related to the selective functionalization at the benzylic carbons of the sumanene. In contrast, very few reports could be found for selective functionalization at the peripheral carbons because of difficulties arose during their synthesis. In this regard, Toda et al. in 2017 have successfully synthesized 2,3,5,6,8,9-hexaarylsumanene derivatives 95a–h utilizing a Suzuki–Miyaura cross-coupling reaction as the critical step [54]. Towards this goal, they started with the hexabromination of sumanene using bromine and iron powder in nitrobenzene to provide the desired compound 94 in 61% yield. Having the bromosumanene 94 in hand, it was then subjected to the Suzuki coupling with several arylboronic acids in the presence of Pd(PPh₃)₄ and K₂CO₃ in THF/water to furnish the required hexaarylated sumanenes in 20–85% yields (Scheme 23).

In the same year, the groups of Hisaki, Sato and Sakurai have reported a hydrogen-bonded 2D sumanene buckybowl framework having 4,4’-dicarboxy-o-terphenyl groups in the periphery. Thy employed the Suzuki coupling as the crucial step (Scheme 24) [55]. In contrast, very recently, Sakurai’s group has described the synthesis of 2,3,5,6,8,9-hexakis(phenoxycarbonyl)sumanene (97) from hexabromosumanene 94 as depicted in Scheme 24 [56]. Because of the low solubility of 94 in common organic solvents, they first tried the esterification reaction with soluble monoiodosumanene 79 using Pd(OAc)₂ under CO atmosphere and obtained the desired sumanene methyl ester in very low yield. Surprisingly, when they used hexabromosumanene 94 under similar reaction conditions, no product formation was observed. Therefore, they used a Pd-catalyzed carbonylative esterification in the presence of phenyl formate which in situ generated CO and phenol to provide the required product 97 in 73% yield (Scheme 24).

In an independent work reported in 2017, Fukushima and co-workers detailed the synthesis of first liquid-crystalline sumanene derivatives 99a–d accomplished by the incorporation of six thioalkyl groups in the peripheral aromatic positions through aromatic nucleophilic substitution reaction of hexabromosumanene 94 with thioalkoxide (Scheme 25) [57]. As can be inspected from Scheme 25, an excess of sodium thioalkoxide in 1,3-dimethyl-2-imidazolidinone 98 reacted with 94 at 100 °C to generate the required sumanene derivatives 99a–c.

In another report, Sakurai’s group in situ generated sumanyne and treated it with different dienes (101a–d) in a Diels–Alder (DA) manner to produce the corresponding DA adducts 102a–d in moderate yields as shown in Table 1 [58]. o-Bromohydroxysumanene 72 was converted to o-hydroxysumanenyl borate 100 by means of a Pd-catalyzed Miyaura-borylation reaction. Compound 100 was then subjected to triflation and subsequent treatment with CsF afforded sumanyne which on further reaction with dienes furnished the DA-adducts 102a–d.

In 2012, another type of functionalized sumanene, namely tris(ethoxycarbonylethenyl)sumanene 108 was prepared by Sakurai's group through ROM–RCM as well as Horner–Wadsworth–Emmons (HWE) reactions as key transformations (Scheme 26) [59]. To this context, their journey started from syn-tris(oxonorborneno)benzene 23, obtained via palladium-catalyzed cyclotrimerization of iodonorbornene. The C₃-symmetric compound 23 was then converted into the methylene groups containing compound 103 using Wittig reaction which on further treatment with dimethyldioxirane (DMDO) gave the triepoxy compound 104 in decent yield. Next, the methoxymethyl groups were installed by means of ring-opening reaction of 104 with LDA followed by protection of thus-created hydroxy groups in the presence of NaH/MeI to afford compound 105. Compound 105 was then transformed into 106 by virtue of ROM–RCM which on subsequent DDQ oxidation followed by HWE reaction provided the desired compound 108 (Scheme 26).

In a really brilliant manner, Kasprzak and Sakurai in 2019 have created a sumanene–ferrocene dual system for selective recognition of cesium cations by means of site-selective cation–π interaction (Scheme 27) [60]. In this study, they used already discussed formylsumanene 66 and aminosumanene 114 prepared from nitrosumanene through the reduction of 80 with Pd/C/H₂. As can be seen from Scheme 27, these attractive building blocks 66 and 114 were then converted into the ferrocene–sumanene conjugates by employing reductive amination reaction for the formation of compounds 110 and 116. Whereas compounds 111, 113, and 117 were prepared by means of condensation reaction as depicted in Scheme 27.

On the other hand, Lentz’s group has reported the synthesis of sumanenylferrocene 118 as well as 1,1’-disumanenylferrocene 119 through Pd-catalyzed Negishi coupling of iodosumanene (79) with ferrocenylzinc chloride and 1,1’-bis(chlorozincio)ferrocene (Scheme 28) [9]. These synthesized molecules were confirmed utilizing spectroscopic techniques as well as by virtue of X-ray crystallographic analysis.

To further advance the sumanene chemistry, in 2018, Sakurai and co-workers reported diosmylation and also synthesized the phenylboronate ester 121 of sumanene (2, Scheme 29) [24]. The osmylation of sumanene was carried out by OsO₄ in pyridine to yield the diadduct 120. Although, they tried to obtain the monoadduct by decreasing the amount of OsO₄ but all the efforts produced only diadducts along with unreacted sumanene (2). Later on, to confirm the configuration, they attempted to transform it into the tetrol through the decomplexation of the osmate ester using Na₂S₂O₃ and t-BuOH/H₂O solution, but they received unsuccessful results may be due to its unstable behavior. Pleasingly, they got the fruitful result by converting 120 into phenylboronate ester 121 as can be noted from an inspection of Scheme 29.

2.3 Complexation of sumanene and its derivatives

Since coordination is one of the vital modes in metal binding and a range of π-conjugated planar systems having ηⁿ-binding to metals have been reported. In contrast, buckybowls have multiple coordination sites for instance the positions available in the polycyclic architecture and also because of the presence of concave or convex faces. Therefore, in recent years, the coordination of bowl-shaped molecules with the transition metals is of fundamental interest in the area of π-bowls chemistry since the first details of the metal complex of C₆₀. In this context, Hirao’s group has reported the first example of Fe(η⁶-sumanene) complex 123 as an endo-coordinated complex at the concave side. The complex 123a was prepared through the metalation of sumanene involving ligand exchange with a cyclopentadienyl (Cp) group of ferrocene by heating in the presence of Al powder and AlCl₃ under solvent-free conditions followed by a counter-anion exchange with hexafluorophosphate (Scheme 30) [61]. Two years later to this report, the same group has also reported the synthesis of complexes 123b and 123c under almost similar reaction conditions [62].

On the other occasion, they have also prepared the tri- and mononuclear sumanenyl zirconocene complexes 124 and 125, respectively, as displayed in Scheme 31 [63]. The mono-zirconocene complex 125 was arranged by treating sumanene (2) with 1 equivalent of n-BuLi in THF-d₈ to generate the monoanion at the benzylic position which on further treatment with CpZrCl₃ provided the mono-zirconocene complex 125a as a brown solid. Alternatively, when the monoanion was reacted with Cp*ZrCl₃ under similar reaction conditions, the complex 125b was achieved as a red solid (Scheme 32). The trinuclear sumanenyl zirconocene 124 was prepared by treating sumanene (2) with 3.1 equivalents t-BuLi in THF-d₈ followed by quenching with Cp*ZrCl₃ to furnish the blackish-green solution. Since the complex 124 was found to be unstable in THF-d₈, therefore, immediately it was replaced by toluene-d₈ which was further replaced by CD₂Cl₂ to remove THF-d₈ completely.

Furthermore, they reported the ruthenium sumanene complex [CpRu(η⁶-sumanene)]PF₆ 126 along with its bowl-to-bowl inversion and anticipated it to be more flexible in comparison to the iron analogue 123a, may be due to the longer C–Ru bond (Scheme 32) [64]. The complex 126 was prepared in a similar manner as its iron analogue was prepared.

Quite recently, Yakiyama, Hasegawa, and Sakurai reported two sumanene-based porous coordination networks 127 (spherical tetramer units) and 128 (belt-like trimer units) through complexation of the hexapyridylsumanene 95d with Cd(NO₃)₂·4H₂O as well as Zn(NO₃)₂·6H₂O, (Scheme 33) [65]. The C₃-symmetric sumanene derivative 95d was assembled through Suzuki reaction of hexabromosumanene with 4-pyridylboronic acid as detailed in the above section. Having the hexapyridylsumanene 95d in hand, it was dissolved in MeOH/CHCl₃ solution and then the MeOH solution of Zn(NO₃)₂·6H₂O was diffused into it for several days to afford colorless crystals of 128. Along similar lines the crystals of 127 were achieved by following the same procedure.

On the other hand, the groups of Amaya and Hirao have reported the synthesis of sumanenyl mono- and trihafnocene complexes 129 and 130, respectively, along with the catalytic activity of 129 for hydroethylation of allylbenzene through carboalumination (Scheme 34) [66]. Complex 129 was prepared by treating pristine sumanene (2) with n-BuLi to generate the monoanion which on further treatment with Cp*HfCl₃ provided the required monohafnocene complex 129. Alternatively, to obtain the trianion species, they subjected 2 with an excess of t-BuLi and then quenched it with Cp*HfCl₃ as displayed in Scheme 34. Furthermore, they revealed for the first time the catalytic activity of sumanene metallocene for the hydroethylation reaction of unactivated allylbenzene by treating it with AlEt₃ at room temperature to furnish compound 132 (Scheme 34).

Since polyanilines are interesting π-conjugated polymers that contain characteristic conductivity by acid doping, environmental stability, optical and redox properties. Therefore, bridging of oligoaniline as a spacer between the functional units can lead toward the generation of versatile functional organic materials. On the basis of this concept, in recent years, a handful of appealing molecules and materials have been synthesized and due to which nanotechnology as well as supramolecular chemistry are continuously attracting the recent attention of researchers worldwide. To this context, Hirao’s team has reported the synthesis of sumanenemonoone imine compounds 134 and 135 along with the Pd(II) complex 136 formed in a stepwise coordination manner investigated by a titration experiment (Scheme 35) [67]. As can be inspected from Scheme 35, compound 134 was prepared by condensing sumanenone 38 with the amino-system 133 in refluxing toluene. The synthesized compound 134 was then treated with Ag₂O in THF to afford the corresponding quinonediimine 135 in respectable yield. Furthermore, the stepwise coordination of the imino functionality of compound 135 to the palladium(II) was carried out in the presence of PdCl₂(MeCN)₂ in a stepwise manner, confirmed by UV–vis spectroscopic technique.

The cyclopentadienyl (Cp) ligand has its own identity in the field of organometallic chemistry as a plethora of transition metal complexes contain this moiety in their structures. In contrast, not much is reported related to the η⁵-coordinated alkali metal complexes though the first report came around 120 years ago. To this context, in 2015, Hirao and Petrukhina’s group have reported the first alkali metal–sumanene complex K₇(C₂₁H₁₀²⁻)₂(C₂₁H₉³⁻)·8THF (137) having di- and tripotassiumsumanenide [68]. As can be inspected from Scheme 36, this complex was prepared by treating the sumanene molecule with an excess of potassium metal. After stirring the reaction for 56 h, the suspension was filtered off and the filtrate was layered in hexanes at 10 °C to yield the beautiful dark-red blocks of the self-assembly. This was confirmed by single crystal structure and it was noticed that all 15 carbon atoms of the three Cp-types of the ring were interacted with the metal atoms and the six K atoms were sandwiched between the convex faces of two sumanenyl trianions.

On the other hand, two years later to this report, in a really dazzling manner, the groups of Rogachev, Hirao, and Petrukhina reported a novel organometallic sandwich supramolecular complex [Na⁺(18-crown-6)(THF)₂][Cs(C₂1H₁₁^‒)₂]^‒ (138) encapsulating the cesium cation between the sumanenyl anions in a concave manner [69]. To this context, they first prepared a monoanion of sumanene 2 by reacting it with sodium in THF at room temperature and after stirring for 2 hours the cesium was added and stirred for another 8 hours. The reaction mixture was then filtered off and the filtrate was layered with 18-crown-6 in hexane at 10 °C for two days to afford the required complex whose structure was confirmed by X-ray crystallography (Scheme 37).

3 Synthesis of heterosumanene derivatives

Doping the backbone of buckybowls such as sumanene and corannulene with heteroatoms (N, P, O, S, Si, and Te) is a promising tactic to modulate the geometrical structure (e.g., bowl depth and bowl-to-bowl inversion energy) and also the physiochemical properties. Therefore, after synthesizing these molecules, scientists turned their attention towards the construction of heterobuckybowls, as heteroatoms could tune the electronic properties of these molecules. Notably, it has been found that the introduction of heteroatom(s) in the periphery of these architectures generally decrease the depth of bowl, may be because of an increase in the carbon–heteroatom bond length. For this reason, these heterobuckybowls possess less strained energy due to shallower bowl structure as compared to the pristine buckybowls containing all-carbon atoms in their frameworks. In sharp contrast, deeper bowl structure (1.30 Å) was observed in the case of triazasumanene as compared to the pristine sumanene (1.11Å) since the C–N bond length (1.47 Å) is shorter compared to the C–C bond (1.54 Å). Due to this, triazasumanene hold high strain energy, hence difficulty in its synthesis and also high reactivity of nitrogen present in it. To date, in most of the heterosumanene substitution has been reported at the benzylic positions except in the case of triazasumanene. In this section we will discuss the developments in heterosumanenes.

3.1 Synthesis of sulfur-doped sumanenes

Although, the interest in the sumanene chemistry arise in 1993 when for the first time Mehta and co-workers reported the unsuccessful attempt towards the construction of this beautifully simple yet much valuable π-conjugated buckybowl. Four years later, McGlinchey’s group has tried to synthesize it by means of an organometallic precursor but they also did not make available the breakthrough for the research community [70,71]. On the other hand, four years before to the Mehta’s report, Klemm and co-workers attempted to synthesize trithiasumanene 151 [72] from triphenylene 139 but they could obtain merely mono- and di-bridged systems and no required product was observed (Scheme 38) [73]. However, the group of Otsubo in 1999 provides the breakthrough by successfully synthesizing the first member of the sumanene family, namely trithiasumanene 151 through tribenzannulation of benzotrithiophene 143 by means of a flash-vacuum pyrolysis process as a vital transformation (Scheme 39) [72,74]. As can be observed from an inspection of Scheme 39, their journey for its synthesis commenced from easily assessable 1,2,3,4,5,6-hexakis(bromomethyl)benzene (142) to generate benzotrithiophene 143 using sodium sulfide followed by DDQ oxidation. Having benzotrithiophene 143 in hand, it was then subjected to bromination using NBS in DMF to produce two isomeric tribromo derivatives 144 and 145 which were subsequently transformed into the trimethylsilylethynyl derivatives 147 and 148 by employing the Sonogashira coupling reaction. Next, compounds 147 and 148 were reacted with HCl in CH₃COOH to provide the tris(chlorovinyl) derivatives 149 and 150 which were then subjected to flash vacuum pyrolysis to afford the desired trithiasumanene 151 (Scheme 39).

On the other hand, Shao and his co-workers synthesized trichalcogenasumanene 155 and tris(S,S‐dioxides)trithiasumanene 156 by virtue of non-pyrolytic approach as displayed in Scheme 40 [23,75-77]. The sumanene derivative 155 was obtained in two steps first by treating the commercially available triphenylene derivative 152 with n-BuLi in the presence of tetramethylethylenediamine (TMEDA) to produce the hexalithiated intermediate which on further treatment with sulfur powder generated the compound 153 having 1,2-dithiin rings as well as one thiophene ring. Later, desulfurizion was accomplished by copper nanopowder to provide the ring-contracted desired product 155 in 30% yield along with a minor amount of 154 which was further converted into the required product under similar reaction conditions (Scheme 40). Moreover, they regioselectively converted 155 into the sumanene-based trisulfone derivative 156 in the presence of hydrogen peroxide in AcOH as detailed in the Scheme 40.

Liu et al. has revealed the functionalization of a sulfur-doped sumanene by means of perbromination followed by nucleophilic substitution as depicted in Scheme 41 [78]. They first performed the hexabromination using Br₂ and iron powder in PhNO₂. The brominated derivative 157 was then converted to the hexathiolated trithiasumanenes 158a–c by substitution reaction (Scheme 41). The structures of these functionalized heterosumanenes were confirmed by spectroscopy as well as crystallography means.

3.2 Synthesis of triselenasumanene

The first synthesis of triselenasumanene 160 was also performed in the laboratory of Shao’s group as depicted in Scheme 42 [23]. The sumanene derivative 160 was prepared from the triphenylene system 152 by treating it with n-BuLi in the presence of TMEDA followed by quenching the hexaanionic species with selenium powder to afford 159 containing one 1,2-diselenin ring and two selenophene rings. Compound 159 was later subjected to deselenation in the presence of copper nanopowder (80–100 nm grain size) to furnish the expected compound 160 in quantitative yield (Scheme 42).

3.3 Preparation of tritellurasumanene derivatives

On the other occasion, tritellurasumanene derivatives 162–165 have been prepared by Shao’s team starting from triphenylene skeletons by means of an ultrasound-assisted one-pot procedure (Scheme 43) [79,80]. As can be noted from Scheme 43, the hexalithiation of 152 was accomplished using butyllithium in the presence of TMEDA which on further treatment with tellurium powder afforded the desired product 162 in 30% yield with 60% starting material recovery. Along similar lines, they also assembled compound 163 in 25% yield as displayed in Scheme 43. Furthermore, when they subjected the tritellurasumanenes 162 and 163 with Br₂ in CH₂Cl₂, they obtained the tris(Te,Te-dibromo)tritellurasumanenes 164 and 165 as covalent adducts in quantitative yield (Scheme 43).

3.4 Preparation of sulfur, selenium and tellurium-doped sumanenes

The pyrazine-fused trichalcogenasumanenes have been assembled via the transformation of 1,2-dibutoxybenzene unit into the corresponding ortho-quinones by means of FeCl₃ in CH₂Cl₂/MeCN at room temperature (Scheme 44) [79-81]. Interestingly, it was observed that 1,2-dibutoxy groups could be selectively transformed to the ortho-quinones in the presence of FeCl₃. From the experimentations, they noticed that heteroatoms also play a significant role in this reaction for instance with sulfur and selenium-doped systems, ortho-quinones 166 and 167 were obtained (Scheme 44). In sharp contrast, the tellurium-doped sumanene provided the covalent adduct 171 with Cl atoms attached onto the Te atoms, confirmed by an X-ray analysis (Scheme 44). Having the quinone derivatives 166 and 167 in hands, they were next subjected to the condensation reaction with a variety of aryl-1,2-diamines 168 in the presence of AcOH to generate the corresponding pyrazine-fused sumanene networks 169a–f and 170a–f in low-to-good yields (Scheme 44).

As can be inspected from Scheme 45, oxidation of 155 and 160 with both oxone (potassium peroxymonosulfate) and H₂O₂ afforded the one-ring-opened product 177. Interestingly, when these compounds were treated with tert-butyl hydroperoxide (TBHP), two flanking benzene rings were found to be cleaved to afford 174 and 175 in impressive yields (Scheme 45) [82-84]. Moreover, they observed radical cations 172 and 173 formation when the same compounds were reacted with Br₂ or 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F₄-TCNQ). On the other hand, when compounds 178 and 179 were reacted with oxidizing agents such as oxone and H₂O₂, selectively one benzene ring-cleaved products 180 and 181 were isolated (Scheme 46) [85]. Having the ring-opened products in hands, they were then subjected to the functional group transformation (molecular-surgery-type functionalization) by treating them with NaOH in EtOH/H₂O (10:1) under reflux conditions to first convert esters functionality into the corresponding carboxylic acid derivatives. As displayed in Scheme 47, these acid derivatives were then transformed into acid anhydrides which on further treatment with different aromatic amines afforded a variety of polyheterocylic compounds [82,86]. Furthermore, Shao’s group has also assembled the diimide-based heterocycles as depicted in Scheme 48 [83].

Quite recently, Tan et al. has disclosed a simple yet effective strategy for scalable synthesis of pristine sulfur, selenium and tellurium-doped sumanenes starting from 1,2-dichloro-3-nitrobenzene 200 by employing the Ullmann coupling and Sandmeyer reaction as key transformations (Scheme 49) [18,87,88]. The target was achieved starting from a three-fold Ullmann coupling reaction of 200 to produce the C₃-symmetric trinitrotriphenylene 201. Having compound 201 in hand, the nitro groups were further converted into the corresponding amino groups via reduction in the presence of Pd/C under hydrogen gas atmosphere to furnish compound 202 in almost quantitative yield. Next, compound 202 was treated with KI/NaNO₂ under Sandmeyer reaction conditions to afford the triiodo compound 203 which on subsequent reaction with mCPBA and triflic acid (TfOH) provided the iodine-doped sumanene 204 in 68% yield. Finally, compound 204 was converted to the required tritellurasumanene 205 through the reaction with Te powder in the presence of tetrabutylammonium bromide (TBAB) and 2-picoline in DMSO as shown in Scheme 49. Interestingly, as can be pointed out from Scheme 49, pristine trithiasumanene 151 and triselenasumanene 206 were also obtained from the common iodine-doped sumanene building block 204. Furthermore, the precursor 204 was converted to the 1,4,5,8,9,12-hexaiodotriphenylene 208 via ring-opening with KI in the presence of a copper/diamine catalyst (Scheme 50). Gratefully, compound 208 was also used as building block for the construction of heterasumaneness 151, 206 and 209 as depicted in Scheme 50.

3.5 Synthesis of trisilasumanene systems

The synthesis of trisilasumanene 214 from triphenylene derivative 152 has been achieved by Furukawa, Kobayashi and Kawashima using the intramolecular sila-Friedel–Crafts reaction as a crucial step (Scheme 51) [89-92]. In this context, they embarked with the bromination of triphenylene derivative 152 using bromine in CH₂Cl₂ to produce two isomeric tribromotriphenylene derivatives 210 and 211 as depicted in Scheme 51. The doubly cyclized monobromo derivative 212 was obtained in two steps from a mixture of tribromotriphenylenes, first by installing two Ph₂HSi groups into it through the lithiation using butyllithium and subsequent addition of Ph₂SiCl₂ followed by reduction with LiAlH₄. Next, the compounds containing Ph₂Si groups were treated with Ph₃CB(C₆F₅)₄ to generate silicenium ionic intermediates which on further two-fold intramolecular sila-Friedel–Crafts cyclization reaction furnished compound 212. Lastly, by repeating the same steps under similar reaction conditions, the desired trisilasumanene 214 was obtained along with a desilylated product 215 (Scheme 51).

On the other hand, two years later of the Kawashima’s report, Saito’s revealed the synthesis of novel trisilasumanenes 218 and 219 as shown in Scheme 52 [93]. Their exploration of these attractive architectures commenced with the formation of dilithiotriphenylene using butyllithium and TMEDA and subsequent reaction with dichlorodimethylsilane to deliver the monocyclized compound 216 in 50% yield (Scheme 52). For the formation of another ring, compound 216 was further treated with n-BuLi in the presence of TMEDA to generate the dianion which on subsequent quenching with Me₃SiCl afforded compound 217 in 46% yield. The final step for the construction of the first trisilasumanene 218 containing no functional groups on the benzene ring system involve the bridging of the remaining bay positions of 217 which was achieved by treating it with n-BuLi followed by the addition of Me₂SiCl₂ to provide title compound 218 in low yield. Furthermore, by reacting compound 218 with butyllithium yielded the required novel C₃-symmetric hexabutyltrisilasumanene 219 in 31% yield (Scheme 52).

3.6 Construction of novel trigermasumanene

One year later, Saito and co-workers have prepared the trigermasumanene 223 having no substituents on the benzene rings through the repetitive lithiation starting from the triphenylene system 139 followed by the insertion of germanium functionalities as displayed in Scheme 53 [94]. As discussed above for the construction of trisilasumanene 218, herein they have used a similar strategy involving identical steps to achieve the desired target 223 as depicted in Scheme 53.

3.7 An attempt towards the synthesis of tristannasumanene

To further advance the chemistry of heterasumanenes, Saito’s group has also put their unsuccessful effort for the preparation of tristannasumanene by employing the similar tactic as discussed above for trisila- and trigermasumanene derivatives [94]. As can be inspected from Scheme 54, the monocyclic compound 225 was achieved in 19% yield which on further treatment with butyllithium and TMEDA followed by quenching with Me₂SnCl₂ provided 227 in just 9% yield along with 139, 225, 226 as shown in Scheme 54. In this procedure, they noticed the transmetalation between tin and lithium atoms. Therefore, because of the cleavage of the Sn–C bonds competitively during the lithiation steps at the bay regions, it’s difficult to obtain the final sumanene derivative 228. Hence, they suggested for an alternative route to assemble the tristannasumanene derivative.

3.8 Synthesis of phosphorous-doped sumanene derivatives

In 2017, the groups of Furukawa, Tada, Fujii and Saito have reported the synthesis of triphosphasumanene trisulfide 232 from 2,3,6,7,10,11-hexaethoxytriphenylene (161), prepared by oxidative coupling of 229 (Scheme 55) [95]. The hexalithiated intermediate 230 was obtained by treating the compound 161 with excess of n-BuLi under hexane reflux conditions. The intermediated 230 was then reacted with dichlorophenylphosphine and subsequent addition of elemental sulfur to furnish the compounds (syn)-232 and (anti)-232 along with triphenylenodiphosphole disulfides (syn)-231 and (anti)-231 (Scheme 55).

3.9 Heterasumanenes containing different heteroatom functionalities

To advance the heterasumanenes chemistry possessing value-added functionalities, doping sumanene with both chalcogen and phosphorus seems to be much promising. To this context, in 2019, Wang et al. have prepared the heterasumanenes 234a–c having chalcogens (S, Se, and Te) and phosphorus atoms at the benzylic positions of the sumanene derivatives starting from trichalcogenasumanenes in a single step (Scheme 56) [96]. In this report, they skillfully opened-up one of the rings of trichalcogenasumanenes by means of butyllithium to produce the dilithiated intermediates 233a–c which on subsequent treatment with dichlorophenylphosphine (PhPCl₂) followed by the addition elemental sulfur produced the required sumanene derivatives 234a–c in moderate-to-excellent yields. Surprisingly, it was observed that when compound 162 reacted with butyllithium at 60 °C, two of the rings were found to be opened. Therefore, they carried out the reaction at lower temperature to open only one of the rings to generate 233c which was further transformed into the heterosumanene 234c as can be inspected from Scheme 56.

In another experiment, Saito and his co-workers disclosed heterasumanenes 237 possessing three dissimilar heteroatom functionalities as shown in Scheme 57 [97,98]. Their journey towards this goal stems from the triphenylene skeleton 139 to produce the triphenyleno[1,12-bcd]thiophene 140 which on further reaction with butyllithium and Me₃SiCl generated the bis(trimethylsilyl) derivative 235. Having compound 235 in hand, it was then transformed into 236 by lithiation and subsequent quenching with Me₂SiCl₂. The final step was to arch the residual bay positions by virtue of heteroatom involving the same sequence of reactions under similar reaction conditions (Scheme 57).

Since trichalcogenasumanene and their congeners are gifted with rich and fascinating chemistry, as regioselective functionalization of these attractive molecules could be easily operated. The chalcogen atom(s) containing molecules particularly having S, Se, and Te atoms in their structures play a crucial role in determining the optical properties, molecular geometry (bowl-to-planar), and chemical reactivities. Having this in mind, very recently a series of trichalcogenasumanene derivatives by means of benzylic carbon replacement with two kinds of chalcogen atoms has been exposed from Shao’s laboratory (Schemes 58–60) [85]. The synthesis of sumanene derivatives 240 and 179 began with the same starting material 2,3,6,7,10,11-hexabutoxytriphenylene (HBT) 152 by treating it with butyllithium and dimethyl disulfide (DMDS) to produce the triphenylene derivative 238 in excellent yield. The compound 238 was further converted into the building block 239 by reacting it with iodine in refluxing CHCl₃ which on subsequent treatment with n-BuLi followed by the addition of tellurium powder provided the required compound 240. Alternatively, after lithiation with butyllithium and subsequent addition of selenium powder, compound 239 provided 241 and 242 in 21% and 46% yields, respectively. These compounds 241 and 242 were then converted into the desired sumanene derivative 179 upon treatment with copper nanopowder as shown in Scheme 58. Furthermore, as can be seen from an inspection of Scheme 59 and Scheme 60, these authors have used almost similar repetitive steps to assemble the other sumanene derivative such as 245, 248, 252, and 178 starting from HBT (152).

To advance the chemical space of heterasumanenes, in 2018, the groups of Furukawa and Saito have reported the synthesis of spiro-heterasumanenes possessing the elements of group 14 (Si, Ge, and Sn) as bridging atoms (Scheme 61) [99]. Towards this mission, they started with the lithiation of hexasubstituted triphenylene derivatives 152 and 161 in the presence of butyllithium and subsequently adding the sulfur/iodine to generate the compounds 255 and 256 in good yields. The compounds 255 and 256 were then treated with copper powder for desulfurization and further subjected to lithiation using butyllithium to generate dilithiated intermediates 233a and 257. Having these valuable intermediates 233a and 257 in hands, they were then treated with heteroatom reagents (MCl₄, M = Si, Ge, Sn or SiHCl₃) to produce the corresponding spiro-sumanene derivatives 264–269 in moderate yields. Moreover, the non-spiro-type heterasumanenes 258–263 were also assembled by reacting with heteroatom reagents (Ph₂XCl₂, X = Si, Ge, Sn) in low-to-good yield as shown in Scheme 61.

To further exploit the chemistry of sumanenes, Shao and co-workers have constructed several sumanene-type hetero-polycycles and they further noticed that the planar 273 and 275 could be easily transformed into the bowl-shaped molecules via the chelatation of Zn²⁺ ion with the bipyridyl unit at room temperature through the formation of a five-membered ring system (Scheme 62) [100]. Noticeably, they observed that this coordinated system displays totally different optical properties compared to the parent molecules. To achieve the target, they commenced with the compounds 250/255 by reacting them with tert-butyl nitrite (TBN) to provide the corresponding mononitro compounds in good yields. Having these nitro compounds in hands, they were next subjected to the reduction using Zn/AcOH to furnish the amino group-containing compounds 270 and 271. These compounds were then treated with pyridine-2-aldehyde and benzaldehyde in a Pictet–Spengler fashion to yield the pyridine-based sumanene-type molecules 272–275. Interestingly, they noticed that when the compounds 273 and 275 were treated with trifluoroacetic acid (TFA), the protonation occurs only at the hanging pyridine ring system. Whereas in excess of TFA both the pyridine rings of the bipyridyl system got protonated which on further neutralization with triethylamine (TEA) produced the deprotonated compounds back (Scheme 62). Moreover, when these compounds were titrated with ZnCl₂, the coordination was occurred with zinc cations forming the five-membered chelate as displayed in Scheme 62.

3.10 Synthesis of chiral nitrogen-doped sumanene and its congeners

To explore the diversity of heteroatom-doped sumanene derivatives, Tan, Higashibayashi, Karanjit, and Sakurai reported the first enantioselective synthesis of triazasumanenes, being more curved and displaying a larger bowl depth than pristine sumanene (2), confirmed by X-ray analysis (Scheme 63) [20,101]. The synthetic strategy for triazasumanenes began with the palladium-catalyzed cyclotrimerization of enantiopure (1S,4R)-282 to yield the C₃-symmetric compound 283 which on hydrolysis and further condensation afforded the non-conjugated lactam 284 (Scheme 63). Since several unsuccessful experiments were performed to directly convert 284 into the desired aromatic bowl-shaped triazasumanene. Therefore, an alternative route was chosen which involve first conversion of the lactam 284 to the thioimidate 285 using Lawesson’s reagent, and then deprotection of PMB followed by methylation using MeI/K₂CO₃. Furthermore, dehydrogenation of 285 was successfully carried out using Ph₃CBF₄ and 2,6-di-tert-butylpyridine (DTBMP) to afford the aromatized compound 286. Since a single crystal of the heterosumanene 286 could not be obtained, they transformed it into the corresponding sulfone derivative 287 whose X-ray study was successful carried out. On the other hand, Sakurai and co-workers have also reported the pristine triazasumanene 288 by performing the desulfurization using poly(methylhydrosiloxane) (PMHS) or Et₂MeSiH in the presence of Pd₂(dba)₃/tris(2-furyl)phosphine (TFP) along with copper(I) thiophene-2-carboxylate (CuTC) (Scheme 63) [20].

In another paper, Sakurai’s group has also reported a series of C₃-symmetric chiral triaryltriazasumanene derivatives 290a–f from chiral tris(methylthio)triazasumanene 286 using diverse boronic acids employing a palladium-catalyzed cross-coupling reaction (Scheme 64) [102]. The coupling reaction was performed between different boronic acids and sumanene derivatives 286 using Pd₂(dba)₃, CuTC and TFP in THF at 50 °C. They have assembled a variety of triazasumanene derivatives using this wonderful strategy in decent yields (Scheme 64).

4 Synthesis of higher-order bowl-shaped sumanene molecules

In addition to C₆₀ and carbon nanotubes (CNTs), the buckybowls are also considered as an interesting class of materials. Although, the chemistry of the corannulene scaffold is highly explored but the sumanene series is still immature due to synthetic difficulties because of the presence of three pentagonal and four hexagonal rings causing high strain. The deeper buckybowls are of significant interest as they possess more similar properties to those of CNTs and fullerenes. These interesting π-bowls are not only expected to be the building block for the construction of fullerene and carbon CNT skeletons but are also useful in liquid crystals and organic semiconductors. The sumanene scaffold is of much interest compared to corannulene not only because it has a large bowl depth but also due to the presence of available benzylic positions for further functionalization. Since these bowl-shaped molecules could also be used for the preparation of fullerenes in a bottom-up approach. Therefore, in recent years, a significant attention of the research community is directed towards the construction of higher-order bowl-shaped architectures. In this context, Hirao and his co-workers reported the synthesis of mono, di- and trinaphthosumanenes using benzannulation reactions (Scheme 65 and Scheme 66) [103]. Towards this mission, the first investigation was carried out by utilizing the benzannulation reaction on momobromosumanene 82 to afford the corresponding mononaphtosumanene 293 by means of a Suzuki-coupling reaction with 2-formylphenylboronic acid 291 followed by an intramolecular condensation reaction to afford the required sumanene derivative 293 in excellent yield (Scheme 65). Furthermore, to obtain the di- and trinaphthosumanenes, they stem from the bromination of pristine sumanene (2) in the presence of molecular bromine to generate an inseparable mixture of dibromosumanenes 294 and 295 and tribromosumanenes 296 and 297. The Pd-catalyzed Suzuki–Miyaura cross-coupling reaction of this mixture with boronic acid 291 provided the corresponding separable compounds 298–301 in overall good yields. The compounds 298 and 299 on two-fold benzannulation in the presence of KHMDS afforded a mixture of 302 and 303 in a 1:1.2 ratio. In contrast, only compound 301 underwent three-fold benzannulation reaction to afford the trinaphthosumanene 304 as depicted in Scheme 66.

On the other hand, in a truly brilliant way, Hirao’s team has reported the synthesis of the hemifullerene skeleton in just two steps starting from sumanene 2 by involving the regioselective intramolecular oxidative cyclization as a critical step (Scheme 67) [104]. To achieve their target, they first performed the condensation reaction of sumanene (2) with benzophenone derivatives 305 and 309 utilizing t-BuOK to give 306 and 310 in 77% and 85% yields, respectively. The dehydrogenative oxidative cyclization of 306 using DDQ in the presence of Sc(OTf)₃ afforded the C₃-symmetric molecule 307 along with 308 in 94% yield with 1:1.5 ratio. Whereas compound 310 solely produced the hemifullerene 311 in excellent yield under similar reaction conditions (Scheme 67).

More strikingly, the Sakurai group has reported the synthesis of a buckybowl fragment of C₇₀ from a C₆₀ sumanene fragment through the ring expansion and annulation reactions in three steps including a Wagner–Meerwein rearrangement to transform the five-membered ring to a six-membered ring as a key transformation (Scheme 68) [105,106]. Their synthetic plan to this goal started with the formation of benzylic carbanion using butyllithium followed by the reaction with a range of aromatic aldehydes to generate the arylsumanyl alcohols 312a–e in 96–99% yields. The Wagner–Meerwein rearrangement was then performed by treating these alcohols 312a–e with stoichiometric amounts of p-TsOH under toluene reflux conditions to afford the corresponding benzopyrene derivatives 313a–e in 89–99% yields. Finally, the required C₇₀ fragment of buckybowls 314a–c were achieved through the cyclization reaction of o-brominated derivatives 313a–c using Pd(PPh)₂Cl₂ and DBU under microwave reaction conditions as depicted in Scheme 68.

Conclusion

The discovery of the pristine sumanene inspired a new wave of investigations on heteroatom-doped sumanene derivatives, and significant developments have been made in the past one and half decades. More interestingly, in recent years, the chemistry of sumanene is continuously attracting tremendous interest of the research community because of their outstanding physiochemical properties as well as potential applications spanning from organometallics, organic chemistry to the supramolecular chemistry and materials science. Although, the chemistry of coronnulene systems is already matured but sumanene, one of the most beautiful and fascinating classes of buckybowl architectures is yet to be explored to higher level. Therefore, to provide readers a quick overview of where the field has been, where it stands now, and where it might be going in near future, herein we have comprehensively summarized all the available synthetic strategies towards the construction of sumanene and its congeners since its invention to hitherto. It seems that the future of sumanene is very bright and the coming few decades would be the era of buckybowl architectures because a range of interesting properties of these systems are yet to be unmasked. To our perception, the next stage for the chemistry of sumanene is to study the crystals engineering such as the polarity of the crystals, use of sumanene derivatives as catalysts and also to prepare novel sumanene-based electro-active functional materials. Therefore, we believe that still more efficient and practical synthetic methods involving user-friendly chemicals under operationally simple reaction conditions are of pressing need.