Trimethylsumanene: Enantioselective synthesis, substituent effect on bowl structure, inversion energy, and electron conductivity

450

Bull. Chem. Soc. Jpn. Vol. 85, No. 4, 450-467 (2012)

Selected Papers

Trimethylsumanene: Enantioselective Synthesis, Substituent Eﬀect

on Bowl Structure, Inversion Energy, and Electron Conductivity

Shuhei Higashibayashi,¹Ryoji Tsuruoka,¹Yarasi Soujanya,²Uppula Purushotham,²

G. Narahari Sastry,²Shu Seki,³Takeharu Ishikawa,⁴Shinji Toyota,⁴and Hidehiro Sakurai*¹

¹Research Center for Molecular Scale Nanoscience, Institute for Molecular Science, Myodaiji, Okazaki, Aichi 444-8787

²Molecular Modeling Group, Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500607, India

³Department of Applied Chemistry, Graduate School of Engineering,

Osaka University, 2-1 Yamada-oka, Suita, Osaka 565-0871

⁴Department of Chemistry, Faculty of Science, Okayama University of Science,

1-1 Ridaicho, Kita-ku, Okayama 700-0005

Received October 3, 2011; E-mail: hsakurai@ims.ac.jp

C₃symmetric chiral trimethylsumanene was enantioselectively synthesized through Pd-catalyzed syn-selective

cyclotrimerization of an enantiomerically pure iodonorbornenone, ring-opening/closing oleﬁn metathesis, and oxidative

aromatization where the sp³stereogenic center was transmitted to the bowl chirality. Chiral HPLC analysis/resolution of

the derivatives were also achieved. Based on theoretical calculations, the columnar crystal packing structure of sumanene

and trimethylsumanene was interpreted as due to attractive electrostatic or CH-³ interaction. According to the

experimental and theoretical studies, the bowl depth and inversion energy were found to increase on methylation for

sumanene in contrast to corannulene. Dissimilarities of the eﬀect of methylation on the bowl structure and inversion

energy of sumanene and corannulene were ascribed to diﬀerences in steric repulsion. A double-well potential model was

ﬁtted to the bowl structure-inversion energy correlation of substituted sumanenes, with a small deviation. The eﬀects of

various substituents on the sumanene structure and bowl-inversion energy were analyzed by density functional theory

calculations, and it was shown that the bowl rigidity is controlled by a combination of electronic and steric eﬀects of the

substituents. The electron conductivity of trimethylsumanene was investigated by time-resolved microwave conductivity

method, compared with that of sumanene.

In the wake of the discovery of fullerene, the chemical and

in 2003. Since then, the chemical and physical properties of

sumanenes have been studied, including its bowl structure,^6a

face selectivity,^6abowl-inversion energy,^4,6b,6ccrystal pack-

ing,^6aelectron conductivity,^6dmetal complexes,^6e-6hand optical

properties.⁶ⁱSumanene derivatives, such as heterasumanenes,⁷

have also been studied. With this background knowledge, we

studied ﬁve aspects of C₃symmetric trimethylsumanene (3),

as listed below.

physical properties of buckybowls have attracted a great deal of

interest because of their unique bowl-shaped ³-conjugated

structure.^1-3The science of buckybowls has grown as a result

of the development of practical routes for the synthesis

of compounds such as sumanene (1)⁴and corannulene (2)⁵

(Figure 1). Sumanene (1), which is the smallest C_3vsymmetric

buckybowl, was ﬁrst synthesized by solution-phase method

Direct Selective Synthesis of C₃Symmetric Substituted

Sumanenes. One of the major diﬃculties in studying the

properties of sumanene (1) is that of selective synthesis of its

derivatives. Because sumanene is C_3vsymmetric, the selective

synthesis of C₃symmetric trisubstituted derivatives through

functionalization at either methylene or benzene is diﬃcult

to achieve. Functionalization of the parent compound 1 results

in mixtures of regioisomers, with the desired C₃symmetric

derivatives as minor products that are diﬃcult to separate.^6c,6i

Because of the importance of C₃symmetric trisubstituted

derivatives for studies on the physical properties of sumanene

(1) and its derivatives or for further transformation into ³-

conjugated derivatives, functionalized sumanenes have to be

Figure 1. Sumanenes and corannulene.

Pub li shed on the web March 28, 2012; do i :10.1246 /bcsj.20110286

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

451

prepared by a direct route, rather than by derivatization of 1.

Because gas-phase methods (ﬂash vacuum pyrolysis) cannot

be used in the case of substituted buckybowls, a solution-

phase method is the only available route.¹We will report

full details of a direct selective synthesis of C₃symmetric

trimethylsumanene (3) by a solution-phase route.⁸

of 1 and 3 are analyzed by theoretical calculations. Substituent

eﬀects on the sumanene skeleton are studied with experimental

data, and the eﬀects of various substituents are predicted by

means of calculations from ﬁrst principles.

Electron Conductivity.

Sumanene (1) has a columnar

crystal packing structure in which the molecules are stacked in

a convex-to-concave fashion with all the columns oriented in

the same direction. Because of its stacking structure, sumanene

has a high electron mobility (0.75 cm²V^¹1s^¹1) with a large

anisotropy (9.2 times) along the column axis, as measured by

time-resolved microwave conductivity (TRMC).^6dAlthough

the columnar crystal packing structures of buckybowls must be

correlated with their electronic properties,^1dthese correlations

have not been studied until now. Transfer integrals of LUMO

orbitals in the contacting buckybowls are predicted to play

a crucial role in the eﬀective transport of electrons along

the columnar stacking axes, and unidirectional stacking may

provide a ratchet-type potential for electrons because of the

asymmetric lobes of the LUMO orbitals in the convex-to-

concave sides of the buckybowls. Because trimethylsumanene

(3) was found to possess a columnar crystal structure with

neighboring columns aligned in opposite directions, in contrast

to 1, the eﬀect of column alignment on the electron transport

properties of 3 was studied by the TRMC method.

Chirality Control of Chiral Buckybowls. Because of

their three-dimensional bowl structure, buckybowls can possess

chirality, bowl chirality, in a similar manner to chiral fullerenes

or carbon nanotubes. Bowl chirality can be induced by the ³-

conjugated skeleton itself (as, for example, in hemifullerene),

by the introduction of substituents, or by the replacement of

skeletal carbons by heteroatoms (as in heterabuckybowls).

Although many buckybowls possess this bowl chirality, no

enantioselective synthesis or even a resolution of a racemate

has been reported to date.^9,10Needless to say, the primary

problem of enantioselective synthesis of buckybowls is that of

introducing homochirality; a secondary problem that occurs

with buckybowls with relatively low bowl-inversion energies

is that of racemization through bowl inversion. These problems

must be overcome if we are to achieve an enantioselective

synthesis of a trisubstituted sumanene. In addition to the direct

selective synthesis of trimethylsumanene (3), we have suc-

ceeded in controlling the bowl chirality of 3 by means of

a chiral transmission method. Racemization was inhibited

by performing operations and derivatizations at low temper-

atures.^8aWe also succeeded in resolving racemic trimethyl-

sumanenetrione (4) by means of chiral HPLC.^8bWith the

enantiomerically pure or enantiomerically enriched chiral

buckybowl in hand, the bowl-inversion energy could be easily

determined by measuring the racemization kinetics by means of

circular dichroism spectroscopy, rather than by NMR methods,

which can only be used to determine the bowl-inversion energy

of buckybowls with diastereotopic protons.

Results and Discussion

Enantioselective Synthesis of Trimethylsumanene (3) and

Determination of Its Bowl-Inversion Energy. Our strategy

for the enantioselective synthesis of C₃symmetric trimethyl-

sumanene (3) is based on a conversion involving a syn-selec-

tive cyclotrimerization of an enantiomerically pure norbornene

derivative to give a C₃symmetric syn-tris(norbornadieno)-

benzene derivative, ring-opening/closing metathesis reactions

to form the nonconjugated bowl structure, and aromatization

to the conjugated bowl structure in which the chirality of

the sp³stereogenic center is transformed into bowl chirality

(Scheme 1).^8aIn the ﬁnal chirality-transmission process, the

bowl-inversion energy of (C)-3 must be taken into account,

because inversion from (C)-3 to (A)-3 corresponds to a race-

mization process for the chiral buckybowl.¹²The bowl-

Substituent Eﬀects on the Bowl Structure and the Bowl-

Inversion Energy.

Bowl inversion is one of the unique

properties of buckybowls, and it has been extensively studied

since the determination of the bowl-inversion energy of a

corannulene derivative by Scott and co-workers.^5bThe eﬀects

of substituents on the bowl structure and the correlation

between the bowl structure and the inversion energy of

corannulene (2) with that of many substituted corannulenes

have been studied by Siegel and co-workers.¹¹However, the

corresponding properties of substituted sumanenes have not

been studied, mainly because of limitations on the derivatiza-

tion of these compounds. On the basis of the experimental bowl

structures of sumanene (1) and trimethylsumanene (3) and

studies on the bowl-inversion energies of 1, 3, and 4, we

studied substituent eﬀects on the sumanene skeleton as well

as the structure-inversion energy correlation by means of

combinations of calculated data for these compounds. We also

discuss the diﬀerences between sumanene and corannulene.

Theoretical Calculations to Corroborate the Experi-

mental Observations. We describe a systematic computa-

tional approach involving density functional theory (DFT)-

based calculations to corroborate our experimental results

and to delineate the various factors responsible for controlling

the structure and rigidity of the compounds in question. The

stacking interactions of molecules in columnar crystal structure

¹1

inversion energy of 3 was estimated to be ca. 21 kcal mol

by comparison between the experimental value (¦G^‡= 20.3

kcal mol (303 K)^6b) of 1 and the calculated values (B3LYP/

¹1

6-311+G(d,p): 18.3 and 19.2 kcal mol^¹1) of 1 and 3, respec-

tively. Because this energy corresponds to a half-lifetime for

racemization of about 2 h at 0 °C, the aromatization step must

be carried out at a low temperature. We surmised that the bowl-

shaped structure of 5 might be constructed from the syn-

tris(norbornadieno)benzene derivative 6 by a ring-opening/

closing oleﬁn metathesis approach and that the syn-tris-

(norbornadieno)benzene derivative 6 might be obtained by

syn-selective cyclotrimerization of the chiral halonorbornene

derivative (1S,4S)-7. By adopting this approach, the bowl

chirality of 3 is controlled by the sp³stereogenic center of the

starting norbornene derivative 7.

The C₃symmetric syn-tris(oxonorborneno)benzene 8 was

prepared by Pd-catalyzed syn-selective cyclotrimerization of

enantiomerically pure iodonorbornenone (1S,4S)-7 (Scheme 2),

as developed by our group.^8a,13Methyl groups were introduced

452

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

Scheme 1. Synthetic strategy for enantioselective synthesis of (C)-3.

Scheme 2. Synthesis of trimethylsumanene (3) and trimethylsumanenetrione (4). Reagents and conditions: (a) Pd(OAc)₂5 mol %,

PPh₃10 mol %, Bu₄NOAc 1000 mol %, Na₂CO₃, 4 ¡ MS, 1,4-dioxane, 100 °C, 2 h, 49% (syn/anti = 95:5); (b) NaN(SiMe₃)₂

350 mol %, (2,6-Me₂C₆H₃O)₂P(O)Cl 350 mol %, P(O)(NMe₂)₃350 mol %, THF, ¹80 °C, 75%; (c) Pd(OAc)₂5 mol %, PCy₃¢HBF₄

10 mol %, MeMgI 400 mol %, THF, 40 °C, 2 h, 72%; (d) Grubbs catalyst I 50 mol %, CH₂Cl₂(0.011 M), under ethylene, 40 °C, 6 h,

then Grubbs catalyst II 50 mol %, CH₂Cl₂(0.034 M), 40 °C, 12 h, 24%; (e) Grubbs catalyst II 50 mol %, toluene (0.1 M), (Z)-oct-4-

ene 1000 mol %, sealed tube, 90 °C, 12 h, 26%; (f ) DDQ 600 mol %, CH₂Cl₂, 0 °C, 1 min, 68%; (g) LDA 600 mol %, THF, ¹40 °C,

then (R)-PhC(CF₃)(OMe)COCl ((R)-MTPACl) 1000 mol %, (C)-(8R,13R,18R)-9, 17%; (h) LDA 600 mol %, THF, ¹40 °C, then

Me₃SiCl 1000 mol %, (C)-(8R,13R,18R)-10, 34%; (i) («)-3, NaN(SiMe₃)₂450 mol %, DMF, O₂atmosphere, ¹40 °C, 5 min, («)-4,

26%.

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

453

by Pd-catalyzed cross-coupling of MeMgI with an alkenyl

phosphate prepared from 8 to give the trimethylated syn-

tris(norbornadieno)benzene 6. In the case of the synthesis

of sumanene (1), Grubbs catalyst I worked for both the ring-

opening and ring-closing metathesis reactions of the unsub-

stituted syn-tris(norbornadieno)benzene under an ethylene at-

mosphere, giving hexahydrosumanene. In contrast, treatment of

6 with Grubbs catalyst I under an ethylene atmosphere resulted

in a mixture of ring-opened reaction products exclusively.

Grubbs catalyst II was therefore adopted for the ring-closing

reaction of the mixture, giving the desired hexahydrotrimethyl-

sumanene 5 in 24% yield. Furthermore, Grubbs catalyst II

was eﬀective in the ring-opening metathesis reaction as well

as the ring-closing reaction, which allowed us to use a single

catalyst for the tandem ring-opening/closing metathesis reac-

tion. When ethylene gas was used as the alkene source for the

ring-opening reaction, operations had to be performed under

dilute conditions to dissolve ethylene gas into solution in order

to prevent polymerization of the substrate. After screening

the reaction conditions, we selected liquid (Z)-oct-4-ene (bp

122 °C) as the alkene source to improve solubility and to

prolong the lifetime of the catalyst. Because a higher temper-

ature was necessary for the oleﬁn metathesis reaction to

proceed with (Z)-oct-4-ene, toluene was used as solvent instead

of dichloromethane. By adopting this combination of con-

ditions, we obtained 5 in 26% yield from 6 in one-pot operation

using Grubbs catalyst II and (Z)-oct-4-ene in toluene in a sealed

tube. The ﬁnal aromatization of 5 was carried out with an

excess of 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ)

at 0 °C for a short time (1 min) to prevent racemization of 3.

The resulting reaction mixture was quickly puriﬁed by silica

gel chromatography at a low temperature (below ¹20 °C) to

aﬀord the desired chiral buckybowl 3.^8a

conﬁguration was supported by the comparison between

the observed CD spectrum and the calculated one of (C)-3

by TD-DFT calculation (Supporting Information). In contrast

to the CD spectrum observed at ¹40 °C, a gradual decrease

in the intensity of the CD spectrum was observed at 10 °C

(Figure 2B). By following the decay of the CD spectra at

247 and 282 nm at 10 °C, we determined the rate constant

of the racemization/bowl inversion and the bowl-inversion

¹1

energy of 3 in MeCN to be ¦G^‡= 21.6 kcal mol (283 K) by

¹1

Eyring equation,^8awhich is 1.3 kcal mol higher than that of

¹1

sumanene (1) (¦G^‡= 20.3 kcal mol (303 K)).^6b

Since it is diﬃcult to determine the enantiomeric excess (ee)

of (C)-3 directly due to the racemization derived from the

relatively low bowl-inversion energy, we estimated the ee

of (C)-3 by two methods involving derivatization. Through

the generation of trianion at the three benzylic positions

of sumanene (1) followed by the addition of electrophiles,

substituents can be selectively introduced at the benzylic

positions with exo-orientation¹⁴to the bowl structure of

sumanene as previously reported.^6a,6bBy adopting the deriva-

tization method, the sample of (C)-3, which was kept at low

temperature as mentioned above to avoid racemization, was

subjected to the conditions introducing (S)-MTPA (PhC(CF₃)-

(OMe)CO) groups or SiMe₃groups at the benzylic positions

at low temperature (¹40 °C) (Scheme 2). (S)-MTPA groups or

SiMe₃groups are selectively introduced with exo-orientation¹⁴

without the bowl inversion of (C)-3. Thus, this derivatiza-

tion aﬀords the corresponding derivatives (S)-MTPA-(C)-

(8R,13R,18R)-9 and (C)-(8R,13R,18R)-10 bearing exo-oriented

substituents, respectively (Scheme 2). The absolute conﬁgu-

ration was further supported by the comparison between

the observed CD spectrum and the calculated one of (C)-

(8R,13R,18R)-10 by TD-DFT calculation (Supporting Informa-

tion). The introduction of the substituents induces new stereo-

genic centers at the benzylic positions. As a result, the bowl

inversion cannot cause the racemization but result instead in the

diastereomerization between the exo-orientation and the endo-

orientation of substituents (Scheme 3). The diastereomerization

through the bowl inversion, however, is negligible because the

bulky substituents signiﬁcantly favor the exo-orientation over

the endo-orientation.^6a,14DFT calculation (B3LYP/6-31G(d))

of 10 shows that the exo-orientation of substituents is 20.2

kcal mol^¹1more stable than the endo-orientation. The ¹H NMR

spectra of 9 and 10 do not show the existence of diastereomers

with endo-substituents formed through bowl inversion, either.

Thus, the ratio of (C)-(8R,13R,18R)-9 or 10 from (C)-3 and

(A)-(8S,13S,18S)-9 or 10 from (A)-3 bearing exo-substituents

can be analyzed without racemization or diastereomerization

through the bowl inversion. In the case of 9, the formation of

(S)-MTPA-(C)-(8R,13R,18R)-9 from (C)-3 and (S)-MTPA-(A)-

(8S,13S,18S)-9 from the contaminant (A)-3, which correspond

to the relevant diastereomers, permits the determination of

the diastereomeric excess (de) of 9 to be 90% by means of

¹H NMR analysis.^8aIn contrast, (C)-(8R,13R,18R)-10 and (A)-

(8S,13S,18S)-10 are enantiomers that can be discriminated by

means of chiral HPLC analysis. Indeed, they were resolved

by using Daicel Chiralpak IA with 2-propanol as the eluent

(Figure 3A). The ee of 10 was determined to be 89%, which

agrees with the value for 9. Racemic 10 was also resolved to

The chirality of thus-obtained 3 was conﬁrmed by circular

dichroism (CD) spectroscopy. Because the bowl inversion

causes rapid racemization at room temperature, the CD spec-

trum of 3 was recorded at ¹40 °C. As expected, the CD spec-

trum was observed at ¹40 °C as shown in Figure 2A and the

intensity of the CD spectrum scarcely changed at ¹40 °C. The

absolute conﬁguration of thus-prepared 3 from (1S,4S)-7 is

assigned to be (C)-3 because the bowl structure of 5 is thought

to be maintained during the oxidation step to 3. The absolute

Figure 2. (A) CD spectrum of (C)-3 in MeCN at ¹40 °C

(red line) and UV spectrum of («)-3 in MeCN at rt (blue

line); (B) decay of CD spectra of (C)-3 in MeCN at 10 °C

during 3 h.

454

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

Scheme 3. Relationship of stereoisomers of trimethylsumanene derivatives through bowl inversion.

material was >99%, it was apparent that only a relatively small

amount of racemization had occurred during the aromatization

from 5 to 3 and subsequent puriﬁcation step of 3.

Synthesis and Chiral HPLC Resolution of Trimethyl-

sumanenetrione (4). Chiral HPLC resolution of 10 is possible

because the presence of the sp³stereogenic centers in addition to

the bowl chirality prevents racemization of the compound by

bowl inversion. In contrast, the resolution of a buckybowl with

exclusive bowl chirality, such as trimethylsumanene (3), is

diﬃcult to achieve because of the rapid racemization of the

compound. However, if a buckybowl has a high bowl-inversion

energy, the slow rate of racemization on the HPLC time-scale

might permit the resolution of the racemic buckybowl by means

of chiral HPLC. Separation of racemic chiral buckybowls

might therefore be another eﬀective strategy for obtaining an

enantiomerically pure or an enantiomerically enriched sample

for the determination of the bowl-inversion energy by CD

spectral measurements. («)-Trimethylsumanenetrione (4), pre-

pared by oxidation of («)-3 (Scheme 2), was a candidate for

this process because its bowl-inversion energy was estimated

Figure 3. (A) Chiral HPLC (Daicel Chiralpak IA, 2-

¹1

to be ca. 23.5 kcal mol by comparison between the experi-

propanol, rt) analysis of (C)-10. (B) CD spectra of (C)-

10 and (A)-10 in CHCl₃at rt. (C) Chiral HPLC (Daicel

Chiralpak IA, 2-propanol/hexane = 5/95, 9 °C) analysis

of («)-4. (D) CD spectra of (C)-4 and (A)-4 in MeCN at rt.

¹1

mental value (¦G^‡= 20.3 kcal mol

(303 K))^6bof 1 and

the calculated values (B3LYP/6-311+G(d,p): 18.3 and 21.5

kcal mol^¹1) of 1 and 4, which corresponds to a half-life of about

44 h, even at 10 °C.^8bThis value could be high enough to permit

the isolation of enantiomerically enriched 4 by means of chiral

HPLC. Furthermore, the bowl-inversion energy of 4 cannot be

determined by an NMR technique, because 4 does not contain

give the corresponding enantiomers, and their CD spectra were

recorded (Figure 3B).^8bFrom the results for 9 and 10, the ee of

(C)-3 was estimated to be >89%. Because the ee of the starting

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

455

Figure 4. Molecular packing of buckybowls in crystals.

the diastereotopic protons at the benzylic positions that are

required to examine the bowl-inversion rate by NMR. As we

expected, resolution of 4 was achieved by using Daicel

Chiralpak IA at 9 °C with hexane/2-propanol = 95/5 as the

eluent, giving samples of each of the enantiomerically enriched

forms of 4 (Figures 3C and 3D). By using the resulting

enantiomerically enriched samples of 4, we determined the

rate constant of the racemization/bowl inversion and the

¹1

bowl-inversion energy to be ¦G^‡= 23.4 kcal mol (303 K)

¹1

in MeCN and 23.3 kcal mol (303 K) in CH₂Cl₂by Eyring

equation^8bfrom the time-dependent decay of the intensity of the

CD spectra at 255 nm at 30 °C; these values are higher than the

¹1

corresponding value for 3 (¦G^‡= 21.6 kcal mol (283 K)).^8a

Crystal Packing Structure of Trimethylsumanene.

Siegel and Wu classiﬁed the molecular-packing structures of

nine diﬀerent crystalline buckybowls into three types; (A)

columnar convex-to-concave stacks with all columns in a

constant direction, (B) columnar convex-to-concave stacks

with neighboring columns in opposite directions, and (C)

noncolumnar structures (Figure 4; type C is omitted) in their

review.^1eSumanene (1) possesses a columnar packing structure

with unidirectional columns of type A,^6awhere the bowls are

stacked without slipping from side to side. The origin of the

columnar packing structure of sumanene is one of our interests

as well as the substituent eﬀect on the packing. The diﬀerence

of packing structure also must be correlated to the properties in

the solid state such as the electronic properties. A single crystal

of («)-trimethylsumanene (3) for X-ray crystallographic analy-

sis was obtained by crystallization from tetrahydrofuran.

ORTEP drawings of («)-3 are shown in Figure 5. These show

a columnar packing structure in which the molecules are

stacked in a convex-to-concave fashion without slipping from

side to side. In contrast to the columnar crystal packing of 1,

neighboring columns of («)-3 are oriented in opposite direc-

tions, which belongs to type B (Figure 4). (A)-3 and (C)-3 are

stacked alternatively at a distance of 3.96 ¡ with a 46.8° twist

angle in the column; the corresponding values for sumanene (1)

are 3.86 ¡ and 55.5°, respectively.^6aThe twist angles of 1 and 3

result in the stacking of the benzene rings and the cyclo-

pentadiene rings. The twist angles of 1 and 3 can be explained

in terms of a combination of an intermolecular repulsive

interaction and an attractive interaction. The repulsive inter-

action corresponds to the intermolecular steric repulsion

between the endo-H and exo-H atoms of the cyclopentadiene

rings. Stacking with a 0° twist angle would lead to a signiﬁcant

intermolecular repulsion between these hydrogen atoms. The

observed twist angles of 1 and 3 prevent this steric repulsion.

The attractive interaction can be interpreted to be the inter-

molecular electrostatic interactions between benzene rings and

cyclopentadiene rings or the intermolecular CH-³ interactions

Figure 5. (A) ORTEP drawings of X-ray crystallographic

analysis of («)-3, (B) top view, (C) side view, shown at

the 50% probability level.

between endo-H atoms of the cyclopentadiene rings and the

benzene rings. Analysis of the electrostatic potentials (ESP)

generated for the sumanene monomer indicates that the

benzene rings are electron rich and that the cyclopentadiene

rings are electron deﬁcient (Figure 6). The observed twist

angles of 1 and 3 can induce favorable electrostatic interactions

between the electron-rich benzene ring and the electron-

deﬁcient cyclopentadiene ring, although the ESP-based ex-

planation for the intermolecular interaction in crystal is not

conclusive in all cases.¹⁵Another explanation for twisted

stacking involves a possible CH-³ interaction,¹⁶where the

endo-H atom of the cyclopentadiene ring of the lower bowl

points toward the benzene ring of the upper bowl (Figures 7A

and 7B). Since the CH-³ interaction between cyclopentadiene

and benzene has not been reported, the evidence for this kind

of CH-³ interaction is obtained from a model complex of

benzene interacting with cyclopentadiene (Figure 7C). The

interaction energy with counterpoise correction for the ben-

zene-cyclopentadiene complex with interacting distance of

¹1

3.1 ¡ is found to be ¹4.34 kcal mol at the MP2/cc-pVDZ

level. The observed twist angles of 1 and 3 can be correlated to

456

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

Table 1. Experimental and Calculated (B3LYP/6-311+

G(d,p)) Bowl-Inversion Energy (¦E) and the Diﬀerences

(¦¦E) of ¦E of 3 or 4 from ¦E of 1

Experimental

¦E ¦¦E

Calculated

¦¦E

¦E

¹1

/kcal mol

1

3

4

20.3

21.6

23.4

0

+1.3

+3.1

18.3

19.2

21.5

0

+0.9

+3.2

Figure 6. Electrostatic potential (ESP) map plotted onto (A)

concave and (B) convex surfaces of sumanene (1) (the

isosurface at 0.002 au) calculated by using the MP2 density

(MP2/cc-pVDZ). Mapping colors from red (¹3.00 ©

10^¹2au) to blue (3.00 © 10^¹2au).

with 1 and 4, we undertook a theoretical study based on

DFT. It is important to note that the theoretical calculations

are pertinent to properties of a single molecule in the gas

phase. Therefore, discrepancies observed between the theory

and the crystal structure can be attributed, albeit with careful

analysis, to intermolecular interactions resulting from crystal

packing.¹⁷To evaluate the reliability of the calculation method,

we calculated the bowl-inversion energies of 1, 3, and 4 by

optimizing the bowl structure and the ﬂat structure correspond-

ing to the transition state for the bowl inversion at the B3LYP

or M05-2X level of theory with a range of basis sets (6-31G(d),

6-311G(d,p), 6-31++G(d,p), 6-311+G(d), 6-311+G(d,p), and

cc-pVDZ). Among them, the B3LYP/6-311+G(d,p) method

was found to be reliable for the diﬀerences of bowl-inversion

¹1

energies of 3 or 4 from that of 1 (3: +0.9, 4: +3.2 kcal mol

to those of experimental values (3: +1.3, 4: +3.1 kcal mol

)

¹1

(Table 1). Taking the values of the bowl-inversion energies

and the calculation cost into account, we suggest that B3LYP

calculations with the 6-311+G(d,p) basis set are suitable for

modeling of this class of compounds.

To examine the eﬀects of the methyl groups, carbonyl

groups, and intermolecular interactions on the crystal structure

in detail, we analyzed the bond length (r), ³-orbital axis vector

(POAV) pyramidalization angle (º),¹⁸bowl depth (d), and cone

angle (ª) in the calculated and experimental structures of 1, 3,

and 4 (Figure 8).

The calculated and experimental bond lengths (r_caland r_exp

)

for 1, 3, and 4 are shown in Table 2. Instances of diﬀerences

(>0.003 ¡) between the calculated values r_caland experimental

values r_expfor 1, 3, and 4 are presented in Figure 9; with longer

and shorter distances highlighted in blue and red, respectively.

In the calculations, the ﬂank bond (C7-C8) and the rim bond

(C8-C9) of 3 are longer than those of 1 because of the

introduction of the methyl groups (Figure 9A). In contrast,

the diﬀerence (¦r_exp) between the experimental bond lengths

r_exp(3) and r_exp(1) show more complex changes owing to the

additional eﬀects of intermolecular interactions in the crystal,

although the ﬂank bond (C7-C8) remains longer in all cases

(Figure 9C). The diﬀerence ¦r between r_expand r_calfor

1 (Figure 9D) and for 3 (Figure 9E) shows shortening of some

bonds, probably as a result of intermolecular interactions in the

crystal. It is evident from Figures 9D and 9E that the eﬀects

of intermolecular interactions on the bond length of 3 are not

identical to those in 1. Owing to the complex nature of the

changes that emerge from a comparison of the experimental

and the calculated bond lengths for 1 and 3, we will restrict our

discussion to a comparison of the geometries of Figures 9A

Figure 7. CH-³ interactions in (A) crystalline trimethyl-

sumanene (3), (B) crystalline sumanene (1), and (C) a

model complex of benzene interacting with cyclopenta-

diene, obtained by MP2/cc-pVDZ calculation.

the presence of a CH-³ interaction involving the endo-H atom

of the cyclopentadiene ring and the benzene ring. The repulsive

and attractive interactions also cause structural changes of 1

and 3 in the crystal state. The structural changes are discussed

in the next section together with diﬀerences in the crystal

structures of 1 and 3 arising from methylation.

Structural Analysis of Sumanene (1), Trimethyl-

sumanene (3), and Trimethylsumanenetrione (4).

account for the structural diﬀerences observed in the crystals of

1 and 3, and to understand the eﬀects of methylation on the

bowl structure and bowl-inversion energy for 3 in comparison

To

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

457

Figure 9. Bond length diﬀerences between r_cal(B3LYP/

6-311+G(d,p)) and r_expfor 1, 3, and 4. Lengthening and

shortening of bonds (>0.003 ¡) are highlighted by blue

and red colors, respectively. (A) ¦r between r_calfor 3 and

r_calfor 1, (B) ¦r between r_calfor 4 and r_calfor 3, (C) ¦r

between r_expfor 3 and r_expfor 1, (D) ¦r between r_expfor

1 and r_calfor 1, (E) ¦r between r_expfor 3 and r_calfor 3.

Figure 8. Analyzed parameters: bond length (r), POAV (º),

bowl depth (d), and cone angle (ª).

Table 2. Calculated^a)and Experimental Bond Lengths (r_cal

and r_exp, respectively) of 1, 3, and 4

Table 3. Calculated^a)and Experimental ³-Orbital Axis

Vector (POAV) Pyramidalization Angles of 1, 3, and 4

Bond length r_cal/¡

Bond length r_exp/¡

Bond

1

3

4

1

3

POAV º_cal/°

POAV º_exp/°

C1-C2

C1-C6

C6-C7

C1-C10

C7-C8

C9-C10

C8-C9

1.433

1.385

1.397

1.430

1.433

1.384

1.396

1.395

1.404

1.397

1.435

1.552

1.554

1.439

1.376

1.398

1.401

1.391

1.404

1.438

1.542

1.544

1.430

1.381

1.433

1.382

1.390

1.402

1.396

1.428

1.551

1.549

Position

1

3

4

1

3

1.395^b)

1.398^b)

1.430

C1

C2

8.8

9.1

8.6

9.4

8.8

8.7^b)

9.1

8.6

a) B3LYP/6-311+G(d,p). b) Average value between C1 and

C2.

C10-C11 1.552

C11-C12 1.552

1.547

Table 4. Calculated^a)and Experimental Bowl Depths (d_cal

and d_exp) for 1, 3, and 4

a) B3LYP/6-311+G(d,p). b) Average values between r(C6-

C7) and r(C1-C10) and between r(C7-C8) and r(C9-C10),

respectively.

Bowl depth d_cal/¡

Bowl depth d_exp/¡

Position

1

3

4

1

3

C7

C10

C8

C9

C11

0.626

1.143

0.880

0.630

0.628

1.154

1.146

0.891

0.645

0.648

1.195

1.191

0.978

0.620^b)

1.115^b)

0.899

0.622

0.620

1.112

1.102

0.913

and 9C. It is apparent from Figure 9 that, whereas the theo-

retical calculations predict an elongation of the rim and ﬂank

bonds on substitution, crystal packing results in a shortening or

shrinking of the spoke bonds as well as an elongation of the rim

and ﬂank bonds in 1 and 3. A comparison of r_calfor compounds

4 and 3 shows that the ﬂank bond lengths (C10-C11 and C11-

C12) of 4 are shorter than those of 3, owing to the change from

sp³to sp²hybridization at C11 (Figure 9B).

The POAV angles (º) of 1, 3, and 4 are shown in Table 3.

The experimental values º_expfor 1 and 3 are in close agreement

with the calculated values º_cal. However, this consistency

arises from the nature of this parameter, which does not reﬂect

subtle structural changes. Judging from these values, the

methyl group induces a disproportionation only at C1 and C2.

In contrast, the structural changes in 4 are suﬃciently large to

aﬀect the value of º. The introduction of a carbonyl group

increases the POAV angle as well as causing a disproportio-

nation at C1 and C2.

a) B3LYP/6-311+G(d,p). b) Average value between d (C7)

and d (C10) and between d (C8) and d (C9), respectively.

The bowl depths (d) of 1, 3, and 4 and the diﬀerences in

these values ¦d are shown in Tables 4 and 5. The diﬀerence

¦d_calbetween d_calof 3 and that of 1 suggests that the

introduction of methyl groups increases the bowl depth at all

positions. On the other hand, the experimental values d_expat C8

and C9 in the crystal of 3 are shallower than those of 1,

whereas d_expat C11 of 3 is deeper than that of 1. The diﬀerence

¦d between the experimental d_expand the calculated d_cal

suggests that the eﬀect of intermolecular interactions on the

bowl depths d_expof 3 is similar to that of 1, which is deeper

at C11 and shallower at other positions in the crystal. The

458

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

Table 5. Diﬀerences^a)¦d among Bowl Depths d_cal(B3LYP/6-311+G(d,p)) and d_expof 1, 3, and 4

¦d_cal/¡

¦d_exp/¡

¦d = d_exp¹ d_cal/¡

Position

d (3) ¹ d (1)

d (4) ¹ d (3)

d (3) ¹ d (1)

1

3

C7

C10

C8

C9

C11

0.004

0.002

0.011

0.002

0.011

0.015

0.020

0.041

0.045

0.087

0.002

0.000

¹0.003

¹0.013

0.014

¹0.006

¹0.028

0.019

¹0.008

¹0.042

¹0.044

0.022

a) Longer and shorter changes (>0.003 ¡) are highlighted in blue and red, respectively.

Table 6. Calculated^a)and Experimental Cone Angles (ª_caland ª_exp) of 1, 3, and 4

Cone angle ª_cal/°

Cone angle ª_exp/°

Position

1

3

4

1

3

C7

C10

C8

C9

C11

76.72

71.33

75.06

76.64

76.65

71.26

71.27

74.85

76.35

76.16

70.65

70.52

73.12

76.83^b)

71.82^b)

74.67

76.78

76.83

71.96

72.01

74.43

a) B3LYP/6-311+G(d,p). b) Average value between ª (C7) and ª (C10) and between ª (C8) and

ª (C9), respectively.

Table 7. Diﬀerences ¦ª between Cone Angles ª_cal(B3LYP/6-311+G(d,p)) and ª_expof 1, 3, and 4^a)

¦ª_cal/°

ª (4) ¹ ª (3)

¦ª_exp/°

¦ª = ª_exp¹ ª_cal/°

Position

ª (3) ¹ ª (1)

1

3

C7

C10

C8

C9

C11

¹0.08

¹0.07

¹0.06

¹0.21

¹0.29

¹0.49

¹0.61

¹0.75

¹1.73

¹0.05

0.00

0.14

0.19

¹0.24

0.11

0.49

¹0.39

0.14

0.18

0.70

0.74

¹0.42

a) Narrower and wider changes are highlighted in blue and red, respectively.

magnitude of the eﬀect is signiﬁcant at C8, C9, and C11 in both

1 and 3. However, the diﬀerences in depth at C8 and C9 of

3 (¹0.042 and ¹0.044, respectively) are much bigger than the

corresponding values in 1 (¹0.028 and ¹0.028, respectively).

The pronounced eﬀects of intermolecular interactions on the

shallowness at C8 and C9 of 3 appear to overwhelm the eﬀects

of introducing of the methyl group (0.011 and 0.002 at C8 and

C9, respectively, from ¦d_calbetween d_calof 3 and d_calof 1),

resulting in a shallower change ¦d_expbetween 3 and 1 in the

crystal (¹0.003 and ¹0.013, respectively). With regard to the

comparison between the value d_calfor 4 and that of 3, the

introduction of the carbonyl groups induces markedly deeper

bowl depth at every position.

From the analysis of the bowl depth, one can easily obtain a

clear picture of the eﬀect on the sumanene skeleton of methyl

groups, which increase the depth at every position, as well as

the eﬀects of intermolecular interactions in the crystal, which

cause an increase in depth at C11 and a reduction in depth

at other positions. However, the changes in bond length

(Figure 9) are not directly correlated to the changes in bowl

depth (Table 5), except that longer bond lengths at C7-C8 and

C8-C9 in 2 can increase the calculated bowl depth at C8. Thus,

the change in bowl depth is the result of a change in the angle

of the bowl shape. Since the POAV angle is not a suitable

parameter for identifying changes in the bowl shape, the cone

angle ª (Figure 8) was introduced to identify the critical points

for structural changes in the bowl shape. The cone angles ª at

C7, C10, C8, C9, and C11 in 1, 3, and 4, and the diﬀerences in

these angles (¦ª) are shown in Tables 6 and 7. We also found a

good correlation between the changes in the bowl depth and the

changes in the cone angle, as evident from Tables 5 and 7. On

the introduction of the methyl groups, the optimized geometries

of 1 and 3 exhibited a narrower calculated cone angle ª_calat

every position, resulting in a deeper bowl depth. In the crystal

of 3, the angles ª_expat C8 and C9 were narrower and the angle

ª_expat C11 was wider than the corresponding angles in 1.

Intermolecular interactions in the crystal therefore result in a

narrower cone angle ª_expat C11 and wider cone angles ª_expat

other positions, producing the observed changes in bowl depth

listed in Table 5. The magnitudes of the changes are consistent

with one another.

The observed eﬀects of intermolecular interactions can be

interpreted in terms of repulsive and attractive interactions

in the crystal packing of 1 and 3 (Figure 10). The repulsive

interaction corresponds to intermolecular steric repulsion

between the benzene rings and the exo-H atoms of the cyclo-

pentadiene rings, which induces a narrower cone angle ª_expat

C11 and wider cone angles ª_expat C8 and C9. The attractive

interaction corresponds to the CH-³ interaction between the

benzene rings and the endo-H atoms of the cyclopentadiene

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

459

Table 8. Calculated (B3LYP/6-311+G(d,p)) Structural Pa-

rameters for Corannulene (2) and Its Methylated Derivative

11, and Diﬀerences between the Two Compounds

Figure 10. Repulsive (red) and attractive (blue) interactions

of (A) 1 and (B) 3 in the crystal packing.

Diﬀerence

Parameter

Position

2

11

(11) ¹ (2)

Bond length/¡ r (C6-C7)

r (C7-C8)

1.446

1.386

0.889

74.73

1.458

1.390

0.883

0.879

74.78

74.86

0.012

0.004

¹0.006

¹0.010

0.05

rings, as discussed in the previous section (Figure 7). The

combination of repulsive and attractive interactions induces

the observed structural changes in 1 and 3 (Figure 10). The

repulsive interaction appears to be enhanced by the steric

hindrance of the methyl group, which results in a bigger change

in angle ª_expat C8 and C9 in 3 than in 1.

Bowl depth/¡

d (C7)

d (C8)

ª (C7)

ª (C8)

Cone angle/°

0.13

Substituent Eﬀects: Sumanene versus Corannulene.

A

comparative analysis of several parameters for trimethyl-

sumanene (3) with those of sumanene (1) showed that the

introduction of methyl groups causes the cone angle of the

bowl shape to be narrower and the bowl depths to be deeper, in

addition to producing an increase in the lengths of the C7-C8

and C8-C9 bonds in sumanene. This ﬁnding is surprising,

because Siegel and co-workers reported that the bowl shape

of corannulene (2) becomes shallower on the introduction of

substituents.¹¹To compare the two molecules, we calculated

the optimized bowl structures of corannulene (2) and methyl-

corannulene (11) at the B3LYP/6-311+G(d,p) level of theory.

Selected structural parameters for 2 and 11 and the diﬀerences

between them are shown in Table 8. The C6-C7 and C7-C8

bonds in 11 are longer than the corresponding bond in 2, in a

similar manner to the case of trimethylsumanene (3). However,

the bowl depth is shallower and the cone angle is wider in the

methylated derivative, in contrast to the case of sumanene.

Similar structural changes have been reported for corannulene

substituted with various other groups. The eﬀect of these sub-

stituents has been explained by an increase in steric repulsion

through peri interactions of the substituents. In general, the

direction of the structural change to a deeper or shallower bowl

depth is governed by two factors: steric repulsion and elec-

tronic eﬀects. To elucidate the diﬀerence between sumanene

and corannulene, we made a thorough analysis of these two

factors.

analysis of the calculated ﬂat and bowl structures of methyl-

sumanene and methylcorannulene (B3LYP/6-311+G(d,p))

showed that the ﬂat structure is stabilized to a greater degree

in both sumanene (0.33 kcal mol^¹1) and corannulene (0.50

kcal mol^¹1). Thus, the electronic eﬀect of the methyl group

leads to a shallower structure, which is consistent with the

observed structural change of methylcorannulene (11) but not

that of trimethylsumanene (3).

Next, we analyzed the diﬀerence in the steric eﬀects of

methyl groups in sumanene and corannulene. In the optimized

bowl and ﬂat structures of both 3 and 11 (B3LYP/6-311+

G(d,p)), one hydrogen atom of the methyl group is oriented

toward the ortho-hydrogen of the ring structure, as shown in

Table 9. The major steric repulsions are those between the

hydrogen of the methyl group and the ortho-hydrogen (H_22a

and H₉in sumanene; H_21aand H₈in corannulene) and between

the carbon of the methyl group and the ortho-carbon (C₂₂and

C₂₁in sumanene, C₂₁and C₂₀in corannulene). Their distances

(r₁(ortho-H), r₂(ortho-C)) and the diﬀerences (¦r₁, ¦r₂) from

van der Waals radii (r₀) are shown in Table 9.¹⁹In both

sumanene and corannulene, the distance (r₁(ortho-H) between

the methyl hydrogen and the ortho-hydrogen is shorter in

the ﬂat structure than in the bowl structure, and is within the

van der Waals radius (¦r₁< 0), whereas the distance (r₂

(ortho-C)) between the methyl carbon and the ortho-carbon is

shorter in the bowl structure than in the ﬂat structure within the

van der Waals radius (¦r₂< 0). Thus, steric repulsion against

the ortho-hydrogen is increased in the ﬂat structures, thereby

favoring a deeper bowl structure. On the other hand, steric

repulsion against the ortho-carbon is increased in the bowl

structure, tending to favor a shallower bowl structure. This

means that the direction of the structural change is determined

by the balance between these two opposing steric eﬀects. In the

case of 3, steric repulsion of the ortho-hydrogen has a greater

eﬀect than that of the ortho-carbon because of the wider bay

angle (C8-C7-C21) arising from the fusion of the six- and ﬁve-

First, we compared the electronic eﬀects of the methyl group

on sumanene and corannulene. The electronic eﬀect of the

substituent stabilizes the core structure through conjugation

between the substituent and the core structure. The bowl shape

is governed by the diﬀerence in the stabilization eﬀect between

the bowl structure and the ﬂat structure corresponding to the

transition state for the bowl inversion. We assumed that the

electronic stabilization eﬀect is bigger in the ﬂat structure than

in the bowl structure, because the conjugation is maximized by

overlapping of orbitals in the ﬂat structure. Indeed, an NBO

460

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

Table 9. Steric Repulsions in Calculated (B3LYP/6-311+

G(d,p)) Trimethylsumanene (3) and Methylcorannulene

(11) and the Internuclear Distances in These Compounds

3

11

Position

Bowl

Flat

Bowl

Flat

r₁(ortho-H)^a)

r₂(ortho-C)^a)

2.335

3.292

2.268

3.509

2.310

3.154

2.287

3.298

Figure 11. Double-well potential model E = ax⁴¹ bx²of

bowl inversion for the sumanene structure. a, b: coef-

ﬁcients, ¦E: bowl-inversion energy.

¦r₁= r₁¹ r₀(H-H)^b)¹0.065 ¹0.132

¦r₂= r₂¹ r₀(C-C)^b)¹0.108

0.109

¹0.090 ¹0.113

¹0.246 ¹0.102

a) r₁(ortho-H): distance between H22a and H9 in 3 or H21a

and H8 in 11; r₂(ortho-C): distance between C22 and C21 in 3

or C21 and C20 in 11. b) r₀(H-H), r₀(C-C) are, respectively,

the H-H and C-C distances calculated from the van der Waals

radii with H = 1.20 ¡ and C = 1.70 ¡.¹⁹¦r₁and ¦r₂are,

respectively, the diﬀerence between r₁(ortho-H) and r₀(H-H)

or between r₂(ortho-C) and r₀(C-C). Between the value (r or

¦r) of the bowl structure and that of the ﬂat structure, shorter

r and more negative ¦r are highlighted by colors, where blue

color or red color favors shallower bowl or deeper bowl,

respectively. The values shown in boldface result in more

signiﬁcant steric repulsions between the blue and red values.

membered rings. Thus, the steric eﬀect of the methyl group in

sumanene results in a deeper bowl structure. In contrast, the

steric repulsion of the ortho-carbon in methylcorannulene (11)

is more signiﬁcant, owing to the narrower bay angle (C7-C6-

C20) of the fully six-membered-ring structure. The steric eﬀect

of the methyl group therefore favors a shallower bowl structure

in corannulene.

Figure 12. Correlation between the structure and the bowl-

inversion energy for

a double-well potential model

¦E = ax⁴with the sumanene structure. a: coeﬃcient.

The cooperative electronic and steric eﬀects of a methyl

substitute thus favor a shallower bowl structure in corannulene.

The shallower structural change in methylcorannulene (11) is

consistent with the lower bowl-inversion energy of 11. In the

case of sumanene, although the two eﬀects of the methyl group

are in opposition to one another, the steric eﬀect is greater than

the electronic eﬀect, leading to a deeper bowl structure on

introduction of methyl substituents. The structural changes in 3

are matched by a higher bowl-inversion-energy change.

Structure-Bowl-Inversion Energy Correlation. Siegel

and co-workers investigated bowl depths and bowl-inver-

sion energies of many substituted corannulenes.¹¹From their

experimental and calculated results, they proposed a correla-

tion between the bowl depth and the inversion energy of

corannulenes in the form of a double-well potential equa-

tion E = ax⁴¹ bx², where E is the energy, x is the reac-

tion coordinate (bowl depth), and a and b are coeﬃcients

(Figure 11). The bowl-inversion energy ¦E is given by the

inversion and E₁and x₁are the energy and reaction coordinate,

respectively, for a bowl structure with a minimum energy

(Figure 12). The experimental and calculated values for

corannulenes are well ﬁtted by this model. Corannulene tends

to adopt a shallower bowl depth and a lower bowl-inversion

energy on the introduction of substituents other than cyclic

substituents. With the experimental and calculated bowl struc-

tures and bowl-inversion energies of sumanene (1), trimethyl-

sumanene (3), and trimethylsumanenetrione (4) in hand, we

analyzed their correlation by the double-well potential model

E = ax⁴¹ bx²and ¦E = ax⁴(Figures 11 and 12, respectively).

For this analysis, we selected the cosine of the cone angle

ª as the reaction coordinate x, and we used three cone angles

independently (the average at C8 and C9 (C8-C9), the average

at C7 and C10 (C7-C10), and angle at C11). Plots of the

calculated cos ª_calat (A) C8-C9, (B) C7-C10, and (C) C11, and

the experimental ¦E_expfor compounds 1, 3, and 4 are shown

in Figure 13, as well as the plots of the calculated cos ª_caland

¦E_cal, and the experimental cos ª_expand ¦E_exp. At the energy-

minimum point, the coeﬃcients a and b are given by the

4

equation ¦E = E₀¹ E₁= ax₁, where E₀is the energy for the

ﬂat structure corresponding to the transition state for the bowl

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

461

Figure 13. Plot of the cone angle (cos ª) against the bowl-inversion energy (¦E). (A) Average ª at C8 and C9, (B) average ª at

C7 and C10, (C) ª at C11. Circle dot: calcd cos ª_cal(B3LYP/6-311+G(d,p)), exptl ¦E_expwith ¦E = a cos⁴ª curve; triangle dot:

calcd cos ª_cal, calcd ¦E_cal; square dot: exptl cos ª_exp, exptl ¦E_exp, of sumanene (1) (blue), trimethylsumanene (3) (red), and

trimethylsumanenetrione (4) (green).

Table 10. Coeﬃcients a and b from the Experimental Bowl-

Inversion Energy ¦E_expand the Calculated^a)Cone Angle

(cos ª_cal) at (A) C8-C9, (B) C7-C10, and (C) C11

of cos ª_expfor both 1 and 3 are signiﬁcantly enlarged. However,

the order for 1 and 3 is not diﬀerent from that of the calculated

cos ª_cal, because the magnitudes of the shifts in 1 and 3 are

similar. This analysis suggests that the calculated value of

cos ª_calat C7-C10 can be supported by a comparison with the

experimental value cos ª_expin the crystal in discussing the

structure-bowl-inversion energy correlation, because these

cone angles are less aﬀected by intermolecular interactions in

the crystal.

Coeﬃcients a of 1 and 3 show good ﬁts to one another at

every position with small deviations (Figure 13 and Table 10).

In detail, there are small deviations at C8-C9 and C7-C10,

whereas there is almost no deviation at C11. In general, a

deviation in coeﬃcient a is caused by a deviation in the

reaction coordinate (structure) and/or a deviation in the bowl-

inversion energy from the values predicted by the original

E = ax⁴¹ bx²equation. A deviation in the structure arises

from a partial change in structure caused by the presence of a

substituent. The deviation in bowl-inversion energy is caused

by a signiﬁcant diﬀerence in the electronic stabilization eﬀects

between the bowl structure and the ﬂat structure. Because, in

the case of 3, a small deviation is observed at C8-C9 and

C7-C10, but not at C11, the deviation of coeﬃcient a for

3 arises mainly from a deviation in the reaction coordinate

resulting from the small structural deviations at C8-C9 and

at C7-C10. If the deviation in the bowl-inversion energy had

the greater inﬂuence, deviations of a similar magnitude would

be observed at all positions.

Position^b)

Coeﬃcient^c)

1

3

4

(A) C8-C9

a

b

a

b

a

b

1.93

3.96

7.29

3.96

4.60

6.11

2.03

4.19

7.60

4.19

4.63

6.32

1.92

4.24

7.35

8.30

3.29

5.55

(B) C7-C10

(C) C11

a) B3LYP/6-311+G(d,p). b) C8-C9: average at C8 and C9,

C7-C10: average at C7 and C10. c) a: ©10³; b: ©10².

equations a = ¦E/x⁴, b = 2ax², respectively. By using the

calculated cos ª_caland the experimental ¦E_expfor 1, 3, and 4,

the coeﬃcients a and b were determined as listed in Table 10,

and the curve for ¦E = a cos⁴ª is shown in Figure 13.

The diﬀerence between the calculated value ¦E_caland the

experimental value ¦E_expshows that ¦E_calfor 1, 3, and 4 is

underestimated with almost similar degree (1.9-2.4 kcal mol

¹1

)

(Figure 13; see also Table 1). As already discussed, the

experimental cone angles ª of 1 and 3 show some diﬀerences

owing to intermolecular interactions in the crystal (Figure 13;

see also Tables 6 and 7). At C8 and C9, the experimental

values of cos ª_expfor 1 and 3 are signiﬁcantly shifted to smaller

values in comparison with the calculated value cos ª_cal.

Because the magnitude of the change in 3 is much greater

than that in 1, the experimental value cos ª_expof 3 is smaller

than cos ª_expfor 1, whereas the calculated value cos ª_calfor 3 is

bigger than cos ª_calfor 1. On the other hand, the shift at C7 and

C10 is small and of a similar magnitude in both 1 and 3. Thus,

the experimental values cos ª_expat C7 and C10 show a better ﬁt

to the calculated values cos ª_cal. At C11, the experimental value

Coeﬃcient b determines the reaction coordinate with a

minimum energy on the ¦E = ax⁴curve. Because the deviation

of coeﬃcient a for 3 from that of 1 is relatively small,

coeﬃcient b is more responsible for the diﬀerence between

1 and 3. The steric repulsion between the methyl group and the

ortho-hydrogen destabilized the ﬂat structure more than the

462

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

Figure 14. Plot of the calculated cone angle (cos ª_cal) versus the bowl-inversion energy (¦E) of trisubstituted sumanenes (B3LYP/

6-311+G(d,p)), together with the ¦E = a cos⁴ª curve of 1 (blue line). (A) Average ª at C8 and C9, (B) average ª at C7 and C10,

(C) ª at C11.

bowl structure, which leads to a deeper bowl and a higher value

of ¦E.

Figure 14 shows plots of the calculated cone angles (cos ª_cal) at

C8-C9, C7-C10, and C11 and of the calculated bowl-inversion

energies ¦E_calfor 1, 3, and 12-18, together with the ¦E =

a cos⁴ª curve of 1. Interestingly, the substituent eﬀect varies

with the nature of the substituent, aﬀording deeper or shallower

bowl structures and higher or lower bowl inversion energies. At

every position, the plots for compounds 12-18 show a good ﬁt

with the ¦E = a cos⁴ª curve of 1. Trihydroxysumanene (12) is

predicted to possess a bigger cos ª_caland a higher ¦E_calthan 1

or 3. Trihalosumanenes 13, 14, and 15 show similar values of

cos ª_caland ¦E_calto 1. 16 with cyano groups, 17 with carboxy

groups, and 18 with formyl groups show a smaller value of

cos ª_caland a lower value of ¦E_calthan 1. These substituent

eﬀects on the structure and on the bowl-inversion energy can

be explained in terms of a combination of electronic and steric

eﬀects of the substituents. Trihydroxysumanene (12) shows a

shortened distance (r₁(ortho-H)) between the hydroxy group

and the ortho-hydrogen in the ﬂat structure, in a similar manner

to 3 (Figure 15 and Table 11), leading to a deeper bowl and a

higher value of ¦E. Halogens of 13, 14, and 15 do not induce

signiﬁcant diﬀerences between the steric repulsions in the bowl

structures and those in the ﬂat structures. Although the cyano

group of 16 does not induce a signiﬁcant diﬀerence in steric

repulsion, the ﬂat structure is stabilized to a greater degree by

electronic eﬀects through ³-conjugation, leading to a shallower

bowl and a lower ¦E. The optimized bowl and ﬂat structures of

17 show that the direction of carboxy group is opposite in the

bowl and the ﬂat structures (Figure 15 and Table 11). Never-

theless, the ﬂat structure is considered to be more stabilized

by the electronic eﬀect through the ³-conjugation, in a similar

manner to 16, leading to a shallower bowl and a lower ¦E.

In contrast, the steric eﬀect causes an opposite eﬀect to that

in 3 or 12, because of the rotation. Unlike 3 or 12, the distance

to ortho-hydrogen (r₁(ortho-H)) is shortened in the bowl

structure compared with the ﬂat structure because the angle

C8-C22-O22a is wider than the angle C8-C22-O22b. Thus,

the steric repulsion against the ortho-hydrogen is increased

in the bowl structure. The steric repulsion against the ortho-

carbon (r₂(ortho-C)) is increased in the bowl structure owing

to the shorter distance, in a similar manner to that in 3 or 12. As

a result, both steric repulsions lead to a shallower bowl. The

cooperative electronic and steric eﬀects result in a shallower

In contrast to the diﬀerence between 1 and 3, the coeﬃcient

a for 4 is smaller at every position than the corresponding value

for 3. Whereas the deviations of coeﬃcient a at C8-C9 and

C7-C10 are small, that at C11 is quite large. The signiﬁcant

deviation at C11 reﬂects the signiﬁcant change in cos ª at

C11 caused by the marked structural change from sp³to sp²

hybridization at this position. From a diﬀerent point of view,

the good ﬁt between the ¦E = a cos⁴ª curves at C8-C9 and

C7-C10 for 4 with those of 1 and 3 is surprising, in that the

marked structural change at C11 does not appear to induce a

large deviation at these positions. Furthermore, the sumanene

derivatives 3 and 4, which are substituted at diﬀerent positions,

can still share the ¦E = a cos⁴ª curve for 1 at C8-C9 or C7-

C10 with little deviation.

The thorough analysis of the structure-bowl inversion

correlation for 1, 3, and 4 based on the E = ax⁴¹ bx²equation

suggests a suitable reaction coordinate for structure-energy

correlation analysis with a uniﬁed equation. The reaction

coordinate close to an introduced substituent is aﬀected to a

greater degree by the structural deviation, causing a signiﬁcant

deviation from the original ¦E = ax⁴equation. A reaction

coordinate distant from the introduced substituent is more

suitable, because of the minimal deviation from the original

¦E = ax⁴equation. In the case of substituted sumanenes, an

analysis at C7-C10 is favorable in this sense.

After our thorough analysis of 1, 3, and 4, our next aim was

to focus on the eﬀects of other substituents on the structure and

the bowl-inversion energy. Siegel and co-workers investigated

the bowl structure and bowl-inversion energy of corannulenes

with various substituents and substitution patterns,¹¹and they

reported that acyclic substituents lead to a shallower bowl and

to a lower bowl-inversion energy in corannulene. However, the

eﬀects of individual substituents were not clear in their study

because of mixing of various substituents and substitution

patterns. Because the reliability of the B3LYP/6-311+G(d,p)

method for calculating the structure and bowl-inversion energy

was well demonstrated in the cases of 1 and 3, we continued to

use this method to predict the substituent eﬀects and structure-

bowl inversion energy correlations for the C₃symmetric

sumanenes 12-18 containing other substituents (Figure 14).

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

463

bowl and a lower ¦E, even compared with 16. In the case of

18, the electronic eﬀects of formyl groups lead to a shallower

bowl and a lower ¦E. In contrast, the steric eﬀects seem to

cause a deeper bowl and a higher value of ¦E because of the

shortened distance (r₁(ortho-H)) between the formyl group and

the ortho-hydrogen in the ﬂat structure in a similar manner to

3 and 12 (Figure 15 and Table 11). In spite of the conﬂicting

steric and electronic eﬀects, 18 shows a much shallower bowl

and a lower ¦E (Figure 14). Although the quantitative analysis

of the magnitude of the steric and electronic eﬀects is diﬃcult

at this point, a possible explanation is that the stronger

electronic eﬀects overwhelm the weaker steric eﬀects. The

relatively smaller diﬀerence between the ¦r₁(¹0.210) at the

bowl structure and that (¹0.236) at the ﬂat structure (Table 11),

compared with those of 3 and 12, could support the weaker

steric eﬀects leading to deeper bowl structure.

From this analysis, we can conclude that the substituent

eﬀect on the bowl structure and the inversion energy vary

depending on the nature of the substituent. The structure-bowl-

inversion energy correlation of sumanenes with a variety of

substituents can be ﬁtted to a double-well potential model with

a uniﬁed equation. The eﬀects of the substituents can be

explained in terms of a combination of their electronic and

steric eﬀects, with the conformation of the substituents playing

an important role.

Columnar Crystal Packing and Electron Conductivity.

As mentioned in the section of crystal packing of trimethyl-

sumanene (3), the molecular-packing structures of crystalline

buckybowls are classiﬁed into three types; (A) columnar

convex-to-concave stacks with all columns in a constant direc-

tion, (B) columnar convex-to-concave stacks with neighboring

columns in opposite directions, and (C) noncolumnar structures

(Figure 4). Although the diﬀerence of packing structure must

be strongly correlated to the electronic properties, the correla-

tions have not been studied until now. Sumanene (1) possesses

a columnar packing structure with unidirectional columns of

type A,^6aand it has been reported to possess a high electron

mobility (0.75 cm²V^¹1s^¹1) and a large anisotropy (9.2 times)

along the columnar axis, as measured by the TRMC method.^6d

Trimethylsumanene (3) has a similar columnar packing struc-

ture to 1, but with neighboring columns oriented in opposite

directions, and is therefore of type B. Sumanenes 1 and 3 are

therefore ideal examples for studying the eﬀects of the column

alignment on the electron-transport properties, especially the

eﬀects on the anisotropic mobility of electrons of ratchet-type

potentials along the stacking axes.

Figure 15. Interactions of peripheral atoms cause steric

repulsion in the calculated structures (B3LYP/6-311+

G(d,p)).

The charge carrier mobility in 1 and 3 were also examined

by means of electrodeless TRMC measurements in the present

Table 11. Nuclear Distances to Cause Steric Repulsion in the Calculated Structures (B3LYP/6-311+G(d,p))

12

17

18

Position

Bowl

Flat

Bowl

Flat

Bowl

Flat

r₁(ortho-H)^a)

r₂(ortho-C)^a)

2.227

3.143

2.178

3.377

2.353

3.010

2.435

3.072

2.510

2.890

2.484

3.092

¦r₁= r₁¹ r₀(X-H)^b)

¦r₂= r₂¹ r₀(Y-C)^b)

¹0.173

¹0.077

¹0.222

0.157

¹0.367

¹0.210

¹0.285

¹0.148

¹0.210

¹0.010

¹0.236

0.192

a) r₁: distance between H22a and H9 in 12, O22a and H9 in bowl-17 and 18, O22b and H9 in ﬂat-17;

r₂: distance between O22 and C21 in 12, O22b and C21 in bowl-17, O22a and C21 in ﬂat-17, H22b and C21

in 18. b) r₀(X-H) and r₀(Y-C) are, respectively, the H-H (X = H) and O-C (Y = O) distances for 12, the O-

H (X = O) and O-C (Y = O) distances for 17, and the O-H (X = O) and H-C (Y = H) distances for 18

calculated from the van der Waals radii with H = 1.20 ¡, C = 1.70 ¡, O = 1.52 ¡.¹⁹¦r₁and ¦r₂are,

respectively, the diﬀerence between r₁(ortho-H) and r₀(X-H) or between r₂(ortho-C) and r₀(Y-C). Between

the value (r or ¦r) of the bowl structure and that of the ﬂat structure, shorter r and more negative ¦r are

highlighted by colors, where blue color or red color favors shallower bowl or deeper bowl, respectively. The

values shown in boldface result in more signiﬁcant steric repulsions between the blue and red values.

464

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

carriers.²⁰The value of º is essential for determining the value

of the mobility. The transient optical absorption spectra were

recorded with identical polycrystalline ﬁlms. Figure 16C

shows the transient optical absorption spectrum of 3, in which

two distinct absorption bands are observed at about 430 and

580 nm, respectively. To assign these absorption bands, the

spectrum of the radical anion of 3 was also recorded by a

radiolytic method in a solution, as shown in Figure 16D. The

spectrum was measured in a solution of 3 in methyltetra-

hydrofuran (MTHF), where the following radiolytic reaction

occurs:²¹

MTHF ! MTHF^•þþ e^ꢀ

ð2Þ

ð3Þ

sol

e^ꢀ_solþ 3 ! 3^•ꢀ

From the radiolytic dose applied to the system and the rate

constant for the reaction of the solvated electron (e _sol) with 3,

¹

the yield of negative ion was estimated quantitatively at the

time the spectrum was recorded to give the molar extinction

Figure 16. (A) Observed conductivity transients in com-

pounds 1 (red) and 3 (blue-violet). The transients were

recorded at room temperature with excitation at 355 nm at

¹

¹1

mol^¹1dm³cm (at 430 nm).²²

The transient absorption spectrum observed under photo-

excitation (Figure 16C) is almost identical to that shown in

Figure 16D; the radical anions of 3 are therefore the major

intermediates present 5 ¯s after exposure to the UV pulse. This

also suggests that the major charge carriers are electrons on 3.

¹2

10¹⁶photons cm (blue). (B) Plot of the maximum values

of conductivity against the excitation photon density.

(C) Transient absorption spectrum of a polycrystalline ﬁlm

of 3 (average thickness 2.5 ¯m) with excitation at 355 nm

by 9 © 10¹⁶photons cm^¹2. The spectrum was recorded 5 ¯s

after the pulse exposure. (D) Transient absorption spectrum

of radical anions of 3, recorded at 100 ns after irradiation of

a 30 mM solution of 3 in methyltetrahydrofuran by a 130-

Gy electron pulse. The solution (10 cm³) was deaerated by

bubbling argon gas for 10 min before recording the spectra.

¹

The value of ¾ gives an estimate of º as µ0.003 from the

photoexcitation spectrum, leading to a mobility in 3 of ® =

¹1

0.2 cm²V^¹1s^¹1. Anistropic mobility value of 0.75 cm²V^¹1s

in a sumanene single crystal leads to the isotropic value of

0.25 cm²V^¹1s^¹1with an assumption of the random orientation

of the crystallographic axes of the polycrystalline state,

suggesting striking agreement to the values in the present

case. On the basis of the crystallographic analysis of 1 and 3,

diﬀerences in the respective stacking distances (¦d < 0.1 ¡)

and the rotation angles (¦ª) are minimal, leading to equivalent

local motion of the electrons along the stacking axes in the

present case. Derived value of the anisotropic electron mobil-

ity; ® µ 0.6 cm²V^¹1s^¹1also gives the estimated spatial size of

the local motion of carriers as 15 nm from the ternover period

of microwave (110 ps), Q value of the microwave cavity (Q =

2500), and the electric ﬁeld strength E µ 10¹V cm^¹1. This is

suggestive that the electron-hole pairs lead respective free

charge carriers when the negative and positive charges are

apart at the interval longer than few tens nm. The yield of

photocarrier generation was markedly less in 3 (º µ 0.003)

than in 1 (º µ 0.06).^6dThe unidirectional columnar packing in

1 produces a ratchet-type potential along the stacking axes,

which leads to asymmetric potential barriers for transport of

electrons and holes along the axes. This may be the cause of the

20-fold higher value of º in 1 upon excitation at 355 nm and it

also suggests that the unique stacking structure of the present

buckybowl systems leads to unique anisotropic photoelectric

properties when they are in their crystalline state.

study. Figure 16A shows the conductivity transients observed

for polycrystalline samples of both 1 and 3 upon excitation by

Nd:YAG laser pulses (FWHM µ 6 ns) at 355 nm. The observed

values of maximum conductivity are also plotted against the

excitation photon density in Figure 16B. The kinetic traces

for 3 in Figure 16A overlap in the range of excitation photon

unsubstituted sumanene single crystal, second-order kinetics

play a crucial role in the recombination processes of photo-

generated electrons and holes in an isolated column of

sumanene with ³-stacking,^6dhowever, unlike the case of 1,

the trimethyl derivative 3 shows relatively little contribution

from second-order recombination processes. This view is also

supported by the almost constant values of º∑®, as shown in

Figure 16B.

The conductivity (¦·) was determined from the transient

microwave absorption in the TRMC measurements, and is

given as a function of the changes in the reﬂected microwave

power (¦P/P_r) from a microwave cavity containing the

sample;

X

1 ꢀP_r

ꢀ· ¼ eºN

® ¼

ð1Þ

A P_r

Summary

where A is a sensitivity factor, e is the elementary charge on the

electron, º is the quantum eﬃciency for photocarrier genera-

tion, N is the number of photons absorbed by the samples, and

∑® is the sum of the mobilities for negative and positive

We have developed a selective and direct approach for the

synthesis of C₃symmetric trisubstituted sumanenes from C₃

symmetric functionalized syn-tris(norborneno)benzene precur-

sor. By using an enantiopure precursor with an sp³stereogenic

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

465

center and by realizing chiral transmission to bowl chirality

at low temperatures, we achieved the ﬁrst enantioselective

synthesis of the chiral C₃symmetric buckybowl trimethyl-

sumanene (3). Bowl inversion, corresponding to racemization

of the chiral buckybowl, was frozen by the introduction of

substituents to generate sp³stereogenic centers or controlled

by functionalization to increase the bowl inversion energy;

this permitted chiral HPLC to be used to analyze and to re-

solve trimethyltris(trimethylsilyl)sumanene (10) and trimethyl-

sumanenetrione (4). The bowl-inversion energies of 3 and 4

were determined by means of time-dependent CD spectral

measurements on the enantiomerically enriched samples of

3 and 4. This strategy, by the introduction of other substituents

and further derivatization, could permit a versatile synthesis of

a range of C₃symmetric substituted sumanenes for elucidation

of their physical properties. The resulting chiral buckybowls

can also be used in studies on asymmetric molecular recog-

nition through bowl chirality, as well to control the chirality of

fullerenes or carbon nanotubes.

buckybowl systems leads to unique anisotropic photoelectric

properties when the substances are in a crystalline phase.

Directional control of the anisotropic transport of electrons in

the columnar structures might be realized by design of the

molecular stacking to allow for convex-to-concave bowl

inversion in the solid state.

Experimental

Single X-ray Crystallographic Analysis.

The single-

crystal X-ray crystallographic analysis of («)-trimethyl-

sumanene (3) was performed on a Rigaku Mercury-CCD

(Mo K¡, - = 0.71070 ¡, T = 173 K, 2ª_max= 55.0°). Data

were collected and processed by using CrystalClear software

(Rigaku).²³The structure was solved by direct methods

(SIR2008²⁴) and expanded by Fourier techniques. Reﬁnements

were performed by the full-matrix least-squares methods.

Hydrogen atoms of methyl groups were calculated and reﬁned

as riding atoms. All calculations were performed using the

CrystalStructure software package (Rigaku).²⁵

We found that 3 has a columnar crystal packing structure in

which (A)-3 and (C)-3 are stacked alternately in a convex-to-

concave fashion with 46.8° twist angle and with neighboring

columns oriented in opposite directions. Theoretical calcula-

tions indicate that an attractive interaction derived from an

electrostatic interaction or from CH-³ interaction plays an

important role in favoring this type of stacking.

Crystallographic data for («)-3: C₂₄H₁₈, brown block

P2₁/a (No. 14), a = 7.905(6), b = 17.708(12), c = 11.434(9)

¹3

¡, ¢ = 104.010(9)°, V = 1553.0(19) ¡³, μ_calcd= 1.310 g cm

,

Z = 4, R₁= 0.0570, and wR₂= 0.1506 for 2518 reﬂections

[I > 2·(I)], 253 reﬁned parameters.

Crystallographic data have been deposited with Cambridge

Crystallographic Data Centre: Deposition number CCDC-

850477. Copies of the data can be obtained free of charge

via http://www.ccdc.cam.ac.uk/conts/retr i ev i ng.htm l (or from

the Cambridge Crystallographic Data Centre, 12, Union Road,

Cambridge, CB2 1EZ, U.K.; Fax: +44 1223 336033; e-mail:

deposit@ccdc.cam.ac.uk).

A detailed structural analysis of sumanene (1) and its

trimethyl derivative 3, based on the calculated and experimen-

tal data, revealed that the presence of the methyl group results

in a deeper bowl, as well as having an eﬀect on the inter-

molecular interactions in the columnar crystal packing. The

diﬀerence of the eﬀect of methyl group between sumanene

and corannulene was thoroughly investigated in terms of both

electronic and steric eﬀects, and it was found that a diﬀerence

in steric repulsion is responsible for the diﬀerence in structure.

The double-well potential model was ﬁtted to a bowl structure-

inversion energy correlation for 1, 3, and the corresponding

trione 4, with a small deviation. A suitable reaction coordinate

was identiﬁed for analyzing the correlation with a uniﬁed

equation. The eﬀects of other substituents on the structure and

bowl-inversion energy of sumanene were predicted by means

of a reliable calculation method. The eﬀects of substituents on

the depth of the bowl and the inversion energy can be explained

in terms of a combination of electronic and steric eﬀects, if

the conformation of the substituents is taken into account. The

knowledge on the substituent eﬀects obtained in this study is

useful to control a bowl shape and/or a bowl inversion of a

buckybowl.

The electron-transport properties of 3 with a bidirectional

columnar crystal packing was examined by the TRMC method,

and compared with those of 1, which has a unidirectional

columnar packing. Crystals of 3 exhibit an electron mobility

(0.2 cm²V^¹1s^¹1) that is equivalent to that of 1. However, the

yield of photocarrier was markedly lower in 3 (º µ 0.003) than

in 1 (º µ 0.06), possibly because the unidirectional columnar

packing in 1 produces a ratchet-type potential along the

stacking axes and hence an asymmetric potential barrier for

transport of electrons and holes along these axes. We also

suggest that the unique stacking structure of the present

Computational Methodology. All the theoretical calcu-

lations were performed by using the Gaussian 03 or 09 program

package.²⁶The CD spectra of (C)-3 and (C)-(8R,13R,18R)-10

were calculated by TD-DFT calculation method at the B3LYP/

6-311+G(d,p) level of theory for the optimized structure of

3 and 10 at the B3LYP/6-31G(d) level of theory. The opti-

mization of 10 with exo- or endo-oriented substituents was

carried out at the B3LYP/6-31G(d) level of theory. The energy

diﬀerence between the exo and the endo was calculated from

the zero-point corrected energies. The optimization of the bowl

and ﬂat structures of 1, 3, and 4 was carried out at the B3LYP

or M05-2X level of theory with a range of basis sets (6-31G(d),

6-311G(d,p), 6-31++G(d,p), 6-311+G(d), 6-311+G(d,p), and

cc-pVDZ), and those of 12-18 were calculated at the B3LYP/

6-311+G(d,p) level of theory. The ﬂat structure has been

characterized as the transition state with one imaginary

frequency corresponding to the bowl inversion for these

sumanenes. The bowl-inversion energies were calculated from

the diﬀerence of the zero-point corrected energies of the bowl

structure and the ﬂat structure. The interaction energy of

benzene-cyclopentadiene complex and natural bond orbital

(NBO) analysis of methylsumanene and methylcorannulene

were carried out at the MP2/cc-pVDZ level of theory.

Transient Optical Spectroscopy.

Compounds were

¹3

dissolved at 30 mmol dm in methyltetrahydrofuran (spectro-

scopic grade, Dojin Chemical Co., Ltd.). The solutions were

placed in quartz cells with a 2-cm optical path, deaerated by

466

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

Chiral Buckybowl: Trimethylsumanene

M. S. Bratcher, J. Am. Chem. Soc. 1992 , 114 , 1920. c) A.

Borchardt, A. Fuchicello, K. V. Kilway, K. K. Baldridge, J. S.

Siegel, J. Am. Chem. Soc. 1992 , 114 , 1921. d) L. T. Scott, P.-C.

Cheng, M. M. Hashemi, M. S. Bratcher, D. T. Meyer, H. B.

Warren, J. Am. Chem. Soc. 1997 , 119 , 10963. e) A. Sygula, P. W.

Rabideau, J. Am. Chem. Soc. 1999 , 121 , 7800. f ) T. J. Seiders,

E. L. Elliott, G. H. Grube, J. S. Siegel, J. Am. Chem. Soc. 1999 ,

121 , 7804. g) A. Sygula, P. W. Rabideau, J. Am. Chem. Soc. 2000 ,

122 , 6323.

bubbling argon for 15 min, and irradiated by an electron pulse

from a linear accelerator (Osaka University). Other details of

the transient spectroscopy have been described elsewhere.²²

typical instrument function was µ8 ns.

A

Support for this work was generously provided by MEXT,

JST, the Tokuyama Foundation, the Ishikawa Carbon Founda-

tion, the Asahi Glass Foundation, and the Sumitomo Founda-

tion. We thank Ms. Sachiko Nakano for her technical support.

6

a) H. Sakurai, T. Daiko, H. Sakane, T. Amaya, T. Hirao,

J. Am. Chem. Soc. 2005 , 127 , 11580. b) T. Amaya, H. Sakane, T.

Muneishi, T. Hirao, Chem. Commun. 2008 , 765. c) T. Amaya, T.

Nakata, T. Hirao, J. Am. Chem. Soc. 2009 , 131 , 10810. d) T.

Amaya, S. Seki, T. Moriuchi, K. Nakamoto, T. Nakata, H. Sakane,

A. Saeki, S. Tagawa, T. Hirao, J. Am. Chem. Soc. 2009 , 131 , 408.

e) T. Amaya, H. Sakane, T. Hirao, Angew. Chem., Int. Ed. 2007 ,

46 , 8376. f ) H. Sakane, T. Amaya, T. Moriuchi, T. Hirao, Angew.

Chem., Int. Ed. 2009 , 48 , 1640. g) T. Amaya, W.-Z. Wang, H.

Sakane, T. Moriuchi, T. Hirao, Angew. Chem., Int. Ed. 2010 , 49 ,

403. h) P. Zanello, S. Fedi, F. F. de Biani, G. Giorgi, T. Amaya, H.

Sakane, T. Hirao, Da l ton Trans. 2009 , 9192. i) T. Amaya, K. Mori,

H.-L. Wu, S. Ishida, J. Nakamura, K. Murata, T. Hirao, Chem.

Commun. 2007 , 1902.

Supporting Information

Stereodescriptor system for chiral buckybowls, experimental

procedures, characterization data for all new compounds,

determination of bowl inversion energy of 3 and 4 by CD

spectra, chiral HPLC analysis of 10 and 4, CD spectra of 3

and 10 by TD-DFT calculation, and H and ¹³C NMR spectra.

1

This material is available free of charge on the web at http://

www.csj.jp/journa l s/bcsj/.

References

1

For recent reviews: a) S. Higashibayashi, H. Sakurai,

Chem. Lett. 2011 , 40 , 122. b) T. Amaya, T. Hirao, Chem. Commun.

2011 , 47 , 10524. c) A. Sygula, Eu r . J . Org. Chem. 2011 , 1611.

d) A. S. Filatov, M. A. Petrukhina, Coord. Chem. Rev. 2010 , 254 ,

2234. e) Y.-T. Wu, J. S. Siegel, Chem. Rev. 2006 , 106 , 4843.

f ) V. M. Tsefrikas, L. T. Scott, Chem. Rev. 2006 , 106 , 4868. g) A.

Sygula, P. W. Rabideau, in Carbon-Rich Compounds: From

Molecules to Materials, ed. by M. M. Haley, R. R. Tykwinski,

Wiley-VCH, Weinheim, 2006, pp. 529-565. do i :10.1002 /

3527607994.ch12.

7

a) K. Imamura, K. Takimiya, T. Otsubo, Y. Aso, Chem.

Commun. 1999 , 1859. b) S. Furukawa, J. Kobayashi, T.

Kawashima, J. Am. Chem. Soc. 2009 , 131 , 14192. c) S.

Furukawa, J. Kobayashi, T. Kawashima, Da l ton Trans. 2010 , 39 ,

9329. d) M. Saito, T. Tanikawa, T. Tajima, J. D. Guo, S. Nagase,

Tetrahedron Lett. 2010 , 51 , 672. e) T. Tanikawa, M. Saito, J. D.

Guo, S. Nagase, Org. B i omo l . Chem. 2011 , 9 , 1731.

8

a) S. Higashibayashi, H. Sakurai, J. Am. Chem. Soc. 2008 ,

130 , 8592. b) R. Tsuruoka, S. Higashibayashi, T. Ishikawa, S.

Toyota, H. Sakurai, Chem. Lett. 2010 , 39 , 646.

2

For recent studies: a) A. V. Zabula, S. N. Spisak, A. S.

Filatov, A. Yu. Rogachev, M. A. Petrukhina, Angew. Chem., Int.

Ed. 2011 , 50 , 2971. b) D. Bandera, K. K. Baldridge, A. Linden, R.

Dorta, J. S. Siegel, Angew. Chem., Int. Ed. 2011 , 50 , 865. c) A.

Ueda, K. Ogasawara, S. Nishida, T. Ise, T. Yoshino, S. Nakazawa,

K. Sato, T. Takui, K. Nakasuji, Y. Morita, Angew. Chem., Int. Ed.

2010 , 49 , 6333. d) D. Eisenberg, E. A. Jackson, J. M. Quimby,

L. T. Scott, R. Shenhar, Angew. Chem., Int. Ed. 2010 , 49 , 7538.

e) B. D. Steinberg, E. A. Jackson, A. S. Filatov, A. Wakamiya,

M. A. Petrukhina, L. T. Scott, J. Am. Chem. Soc. 2009 , 131 , 10537.

f ) T. Bauert, L. Merz, D. Bandera, M. Parschau, J. S. Siegel, K.-H.

Ernst, J. Am. Chem. Soc. 2009 , 131 , 3460. g) D. Miyajima, K.

Tashiro, F. Araoka, H. Takezoe, J. Kim, K. Kato, M. Takata, T.

Aida, J. Am. Chem. Soc. 2009 , 131 , 44. h) L. Merz, M. Parschau,

L. Zoppi, K. K. Baldridge, J. S. Siegel, K.-H. Ernst, Angew. Chem.,

Int. Ed. 2009 , 48 , 1966.

9

For racemic synthesis of C₃symmetric chiral hemi-

fullerene: a) A. H. Abdourazak, Z. Marcinow, A. Sygula, R.

Sygula, P. W. Rabideau, J. Am. Chem. Soc. 1995 , 117 , 6410. b) S.

Hagen, M. S. Bratcher, M. S. Erickson, G. Zimmermann, L. T.

Scott, Angew. Chem., Int. Ed. Eng l . 1997 , 36 , 406. For racemic

synthesis of corannulene derivatives, see Ref. 1 and following

reference: c) Y.-T. Wu, T. Hayama, K. K. Baldridge, A. Linden,

J. S. Siegel, J. Am. Chem. Soc. 2006 , 128 , 6870.

10 Siegel and co-workers recently reported dynamic enantio-

selective complexation of racemic pentamethylcorannulene with

enantiopure metal complexes under equilibrium through bowl

inversion. See Ref. 2b.

11 T. J. Seiders, K. K. Baldridge, G. H. Grube, J. S. Siegel,

J. Am. Chem. Soc. 2001 , 123 , 517.

12 The stereodescriptor system, (C) and (A), for the absolute

conﬁguration of chiral buckybowls in this article is based on

fullerene nomenclature. See Supporting Information and following

reference: a) C. Thilgen, F. Diederich, Chem. Rev. 2006 , 106 ,

5049. Another stereodescriptor system was proposed in the

following reference: b) M. A. Petrukhina, K. W. Andreini, L.

Peng, L. T. Scott, Angew. Chem., Int. Ed. 2004 , 43 , 5477.

13 a) S. Higashibayashi, H. Sakurai, Chem. Lett. 2007 , 36 , 18.

b) A. F. G. M. Reza, S. Higashibayashi, H. Sakurai, Chem.®As i an

J. 2009 , 4 , 1329. c) S. Higashibayashi, A. F. G. M. Reza, H.

Sakurai, J. Org. Chem. 2010 , 75 , 4626.

3

For recent theoretical studies: a) D. Vijay, H. Sakurai, G. N.

Sastry, Int. J. Quantum Chem. 2011 , 111 , 1893. b) J. R. Green,

R. C. Dunbar, J. Phys. Chem. A 2011 , 115 , 4968. c) L. Zoppi, A.

Garcia, K. K. Baldridge, J. Phys. Chem. A 2010 , 114 , 8864.

d) G. N. Sastry, THEOCHEM 2006 , 771 , 141. e) U. D.

Priyakumar, M. Punnagai, G. P. K. Mohan, G. N. Sastry,

Tetrahedron 2004 , 60 , 3037. f ) G. N. Sastry, U. D. Priyakumar,

J. Chem. Soc., Perk i n Trans. 2 2001 , 30. g) U. D. Priyakumar,

G. N. Sastry, J. Phys. Chem. A 2001 , 105 , 4488. h) G. N. Sastry,

H. S. P. Rao, P. Bednarek, U. D. Priyakumar, Chem. Commun.

2000 , 843.

14 The relative conﬁguration of a substituent at the benzylic

position of sumanenes to the bowl structure is expressed by exo or

endo. The exo or endo is deﬁned to be a substituent directing to the

convex face or the concave face of the bowl structure, respectively.

4

5

H. Sakurai, T. Daiko, T. Hirao, Sc i ence 2003 , 301 , 1878.

a) L. T. Scott, M. M. Hashemi, D. T. Meyer, H. B. Warren,

J. Am. Chem. Soc. 1991 , 113 , 7082. b) L. T. Scott, M. M. Hashemi,

S. Higashibayashi et al.

Bull. Chem. Soc. Jpn. Vol. 85, No. 4 (2012)

467

15 a) M. A. Spackman, J. J. McKinnon, D. Jayatilaka,

CrystEngComm 2008 , 10 , 377. b) S. E. Wheeler, J. Am. Chem.

Soc. 2011 , 133 , 10262.

16 O. Takahashi, Y. Kohno, S. Iwasaki, K. Saito, M. Iwaoka,

S. Tomoda, Y. Umezawa, S. Tsuboyama, M. Nishio, Bu ll . Chem.

Soc. Jpn. 2001 , 74 , 2421.

17 a) A. S. Reddy, G. M. Sastry, G. N. Sastry, Prote i ns 2007 ,

67 , 1179. b) A. S. Reddy, H. Zipse, G. N. Sastry, J. Phys. Chem. B

2007 , 111 , 11546. c) J. S. Rao, H. Zipse, G. N. Sastry, J. Phys.

Chem. B 2009 , 113 , 7225.

18 R. C. Haddon, J. Am. Chem. Soc. 1987 , 109 , 1676.

19 A. Bondi, J. Phys. Chem. 1964 , 68 , 441.

20 a) A. Acharya, S. Seki, A. Saeki, Y. Koizumi, S. Tagawa,

Chem. Phys. Lett. 2005 , 404 , 356. b) A. Acharya, S. Seki, Y.

Koizumi, A. Saeki, S. Tagawa, J. Phys. Chem. B 2005 , 109 , 20174.

21 a) S. Seki, Y. Kunimi, K. Nishida, Y. Yoshida, S. Tagawa,

J. Phys. Chem. B 2001 , 105 , 900. b) S. Seki, Y. Koizumi, T.

Kawaguchi, H. Habara, S. Tagawa, J. Am. Chem. Soc. 2004 , 126 ,

3521. c) K. Okamoto, S. Seki, S. Tagawa, J. Phys. Chem. A 2006 ,

110 , 8073.

Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,

G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.

Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,

H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,

V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,

O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,

P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.

Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.

Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,

J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Cliﬀord, J.

Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.

Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,

C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B.

Johnson, W. Chen, M. W. Wong, C. Gonzalez, J. A. Pople,

Gaussian 03, Revision C.02, Gaussian, Inc., Wallingford CT, 2004.

b) M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,

M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.

Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.

Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg,

M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida,

T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A.

Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd,

E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.

Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S.

Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,

J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.

Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.

Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G.

Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S.

Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz,

J. Cioslowski, D. J. Fox, Gaussian 09, Revision A.1, Gaussian,

Inc., Wallingford CT, 2009.

22 S. Seki, Y. Yoshida, S. Tagawa, K. Asai, Macromo l ecu l es

1999 , 32 , 1080.

23 CrystalClear: Crystal Structure Analysis Package, Rigaku

and Rigaku/MSC, The Woodlands, TX 77381, USA, 2007.

24 R. Caliandro, B. Carrozzini, G. L. Cascarano, L. De Caro,

C. Giacovazzo, A. Mazzone, D. Siliqi, J. App l . Crysta ll ogr. 2008 ,

41 , 548.

25 CrystalStructure: Crystal Structure Analysis Package,

Rigaku and Rigaku/MSC, The Woodlands, TX 77381, USA,

2007.

26 a) M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.

Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T.

Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J.

Article Doi

Trimethylsumanene: Enantioselective synthesis, substituent effect on bowl structure, inversion energy, and electron conductivity

DOI: 10.1246/bcsj.20110286

Source and publish data:

Authors:

Article abstract of DOI:10.1246/bcsj.20110286

Full text of DOI:10.1246/bcsj.20110286

Products guided by the article

R&D Labs maybe for 1040140-91-7

Relevant to this article

Hot Product