Highly curved bowl-shaped fragments of fullerenes: Synthesis, structural analysis, and physical properties

Article abstract of DOI:10.1002/chem.201303357

Highly curved buckybowls 3, 4, and 5 were synthesized from planar precursors, fluoranthenes 8, benzo[k]- fluoranthenes 10 and naphtho[1,2-k]- cyclopenta[cd]fluoranthenes 12, respectively, using straightforward palladium-catalyzed cyclization reactions. Th

Full text of DOI:10.1002/chem.201303357

DOI: 10.1002/chem.201303357
Full Paper
&
Fullerenes
Highly Curved Bowl-Shaped Fragments of Fullerenes:
Synthesis, Structural Analysis, and Physical Properties**
Min-Kuan Chen,[a] Hsin-Ju Hsin,[a] Tsun-Cheng Wu,[a] Bo-Yan Kang,[b] Yen-Wei Lee,[b]
Ming-Yu Kuo,*[b] and Yao-Ting Wu*[a]
Dedicated to Professor Armin de Meijere on the occasion of his 75th birthday
Abstract: Highly curved buckybowls 3, 4, and 5 were syn-
thesized from planar precursors, fluoranthenes 8, benzo[k]-
fluoranthenes 10 and naphtho[1,2-k]-cyclopenta[cd]fluoran-
thenes 12, respectively, using straightforward palladium-cat-
alyzed cyclization reactions. These fluoranthene-based start-
ing materials were easily prepared from 1,8-bis(arylethynyl)-
naphthalenes 6. Both buckybowls 3 and 4 are fragments of
C60, whereas 5 is a unique subunit of C70. The curved struc-
tures were identified by X-ray crystallography, and they are
deep bowls. The maximum p-orbital axis vector (POAV) pyra-
midalization angle in both 3 and 4 is 12.88. Such a high cur-
vature is very rarely obtained. Buckybowls 5 are less curved
than the others because they have a lower density of five-
membered rings, analogous to the tube portion of C70. Cy-
clopentaannulation increases the bowl depths of 3 and 4,
but not the maximum POAV pyramidalization angle. Among
the eight buckybowls studied herein, five form polar crystals.
The bowl-to-bowl inversion dynamics of these buckybowls
can be classified into two types; one has a planar transition
structure, whereas the other has an S-shaped transition
structure. A larger longitudinal length of these buckybowls
corresponds to a stronger preference for the latter. The pho-
tophysical properties of these buckybowls were examined
and compared with those of C60 and C70. Buckybowls 5 have
absorption bands at wavelengths greater than 450 nm,
which are similar to those of C70. The chiral resolution of the
mono-substituted buckybowl 4ac was also studied by using
HPLC with a chiral column.
Introduction
Introduction of five-membered rings into sp2-carbon hexagonal
p-networks enables the formation of aromatic bowls, consis-
tent with Euler’s rule.[1] Corannulene (1)[2] and sumanene (2)[3]
are representative examples of such bowls. These bowl-shaped
compounds (so-called buckybowls or p-bowls) can be extend-
ed to form fullerenes[4] or carbon nanotubes (CNTs).[5] Investiga-
tions of the bowl-shaped subunits of fullerenes may help to
elucidate their chemistry. More importantly, buckybowls are
known for their interesting physical properties, since they are
remarkably effective in the stabilization of neutral radicals[6]
and have (potential) applications as electro-optical organic ma-
terials, such as liquid crystals and organic semiconductors.[7]
Buckybowls have the potential to be utilized as precursors in
the formation of selected CNTs, some forms of which exhibit
excellent electronic conductivity that greatly exceeds that of
copper wire.[8] Owing to the low solubility in common organic
solvents, their pure form cannot be obtained in useful
amounts by either preparation or purification.[9] Hence, the
synthesis of a single form of CNT from a buckybowl is
sought.[10]
[a] M.-K. Chen, H.-J. Hsin, Dr. T.-C. Wu, Prof. Y.-T. Wu
Department of Chemistry, National Cheng Kung University
No. 1 Ta-Hsueh Rd., 70101 Tainan (Taiwan)
E-mail: ytwuchem@mail.ncku.edu.tw
[b] B.-Y. Kang, Y.-W. Lee, Prof. M.-Y. Kuo
Department of Applied Chemistry, National Chi Nan University
No. 1 University Rd., 54561 Puli, Nantou (Taiwan)
E-mail: mykuo@ncnu.edu.tw
Highly curved buckybowls, which are defined herein as p-
bowls being more curved than C60,[11] have a high inherent
strain, so their synthesis is a challenge. They are most com-
monly prepared by a strategy that involves extension of the
backbone of a smaller bowl and/or by the use of high-temper-
[**] Metal-Catalyzed Reactions of Alkynes, Part XV.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201303357.
Chem. Eur. J. 2014, 20, 598 – 608
598
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ature flash vacuum pyrolysis (FVP) as a synthetic tool.[2d,12] The
conditions required for FVP considerably limit the range of
functional groups and potentially cause thermal rearrange-
ment of the molecular framework.[13] Although highly curved
buckybowls have been generated efficiently from corannulene
and sumanene with the help of metal-catalyzed reactions
under mild conditions, the overall lengths of these syntheses
can be quite long, when the many steps required to synthesize
corannulene and sumanene are taken into account.[14] Exam-
ples of such approaches include the preparations of pentain-
denocorannulene[15] and trinaphthosumanene (C42H18).[16]
Unlike these synthetic methods, buckybowls 3aa (R1 =H), 4aa
(R1 =R2 =H),[17] and 5aa (R1 =H)[18] can be directly formed from
easily obtained planar precursors, as determined in our prelimi-
nary studies. Although similar protocols have been utilized to
synthesize less-curved as-indaceno[3,2,1,8,7,6-pqrstuv]pice-
nes[14f] and dibenzocorannulenes[13,14e] by Pd-catalyzed cycliza-
tions, the reaction conditions for the preparation of highly
curved bowls are yet to be systematically examined. Notably,
both 3aa and 4aa are highly curved fragments of C60, whereas
5aa is a distinctive subunit of C70. Fullerene C70 can be consid-
ered to be two C60-like hemispheres that are connected by
a set of ten carbon atoms on the equatorial line. p-Bowl 5aa
contains unique equatorial carbon atoms and maps onto the
tube portion of C70; furthermore, it is also a fragment of nu-
merous higher fullerenes, such as C76, C78,[19] C84,[20] and others.
Following the successful synthesis of these buckybowls, the
same method was extended to prepare more curved cyclopen-
taannulated derivatives and a chiral bowl. The results of stud-
ies of their synthesis, structures, stereochemistry, and physical
properties are presented below.
Scheme 1. Synthesis of precursors for buckybowls.
Table 1. Optimization of reaction condition for synthesis of buckybowls
3.[a]
Entry
Fluoranthene
Condition
change
Ratio
3:13
Yield
[%][b]
1
2
3
4
5
6
7
8
8aa
8aa
8aa
8aa
8aa
8aa
8aa
8aa
8aa
8aa
8ba
8ba
8ba
DMF
DMAc
–
24 h
48 h
1408C
1808C
DBU (4 equiv)
DBU (16 equiv)
[PdCl2(PCy3)2] (20 mol%)
–
DMAc
DMF
54:46
78:22
71:29
67:33
71:29
74:26
79:21
67:33
62:38
65:35
36:64
40:60
39:61
26
23
28[c]
21
24
23
19
12
24
20
11
5
Results and Discussion
Synthesis
Fluoranthene derivatives are ideal precursors in the synthesis
of buckybowls 3, 4, and 5 because they all contain a fluoran-
thene segment. Metal-catalyzed annulations of 1,8-bis(arylethy-
9
10
11
12
13
nyl)naphthalenes
6
generate various fluoranthene-based
arenes (Scheme 1).[21] Rh-catalyzed co-cyclotrimerization of
diynes 6 with 2-butyne yielded fluoranthenes 8 in excellent
yields.[14c] Pd-catalyzed annulation of 6 with iodoarenes 9 gave
benzo[k]fluoranthene 10.[22] The low yield of 10ac was unsur-
prising because our earlier studies revealed that substituted o-
diiodoarene exhibits better annulation than substituted mono-
iodoarene. Naphtho[1,2-k]cyclopenta[cd]fluoranthenes 12 were
efficiently prepared by the simple Rh-catalyzed [(2+2)+2] cy-
cloaddition of diynes 6 with acenaphthylene (11 a),[14c] and
were subsequently aromatized by treatment with 2,3-dichloro-
5,6-dicyano-1,4-benzoquinone (DDQ).
18
[a] The reaction was conducted with 8 (0.1 mmol) in NMP (5 mL). The
3:13 ratio was determined by 1H NMR spectroscopy. [b] The yield for
a mixture of isolated 3 and 13. [c] Ref. [17].
Although solvents dimethylformamide (DMF), N-methylpyrroli-
done (NMP) and dimethylacetamide (DMAc) gave a mixture of
3aa and 13aa in similar yields, the amounts of the byproduct
13aa in the latter two solvents declined (entries 1–3 in
Table 1). In general, the reaction temperature, reaction time, or
amounts of DBU used did not significantly influence the 3aa/
13aa ratio, but some conditions drastically reduced their yields
(entries 4–9 in Table 1), such as a temperature of 1808C or the
use of only four equivalents of DBU. Conducting the reaction
at high temperature promoted the formation of dechlorinated
Methylene-bridged buckybowls 3 were generated from fluo-
ranthenes 8 through Pd-catalyzed benzylic and aryl CÀH bond
activations.[23] The cyclization of fluoranthenes 8 produced
a mixture of 3 and 13. The reaction conditions for the cycliza-
tion of fluoranthene 8aa using [PdCl2(PCy3)2] and 1,8-diazabicy-
cloundec-7-ene (DBU) were systematically examined (Table 1).
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
599
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
intermediates. Insufficient DBU resulted in the incomplete con-
version of 8aa. Under the optimal conditions herein, a mixture
of 3aa and 13aa (ratio 71:29) was obtained in 28% yield
(entry 3 in Table 1). Cyclopentaannulated 3ba was similarly
prepared using DMF as a solvent (entry 13 in Table 1). The re-
action was inefficient and 13ba was observed as the major
product, presumably because of the highly curved structure of
3ba. Notably simple chromatography, eluting with a mixture
CH2Cl2/hexane, did not completely separate the compounds 3
from compounds 13. Only very small amounts of them were
obtained in the pure form. Elution with cyclohexane without
the application of external air pressure over several hours grad-
ually eluted 3 from the column, leaving 13 behind.
The reaction conditions that were used in the synthesis of 4
were applied in the preparation of 5, but low yields were ob-
tained (Scheme 2). Since bowls 5 were only weakly soluble in
common organic solvents, some of the material was irreversi-
Buckybowls 4 were prepared by the cyclization of 10 in the
presence of a mixture of DBU and [Pd(PCy3)2Cl2] at 1608C
(Table 2). The concentration of 10 strongly affected the reac-
Scheme 2. Synthesis of buckybowls 5.
bly lost during chromatography. The eluent cyclohexane was
much more effective in purification than the mixture of CH2Cl2/
hexane. Elution with cyclohexane without the application of
external air pressure over several hours gave compounds 5 in
around 10% yields. The soluble residues in the column were
collected by flushing it with CH2Cl2. Small amounts of 5 were
obtained from the collected residues by column chromatogra-
phy and elution with cyclohexane.
Table 2. Preparation of buckybowls 4.[a]
X-ray crystallographic structures
Entry
Starting
material
C
[m]
Product
Yield
[%]
X-ray-quality crystals of new compounds 3ba, 4ba, and 13ba
were obtained by the slow evaporation of the mixed CH2Cl2/
MeOH solvent (Table 3).[25] Single crystals of 3aa were grown
by the diffusion of MeOH into a solution in benzene. The cur-
vature of buckybowls was determined by analyzing the bowl
depth of the corannulene fragment. The bowl depths follow
the order 4>3>5>13, and all significantly exceed that of
corannulene (0.87 ꢂ; Table 4).[26] Bowls 5 are less curved than 3
and 4 owing to the lower density of five-membered rings,
analogous to the tube portion of C70. Cyclopentaannulated
buckybowls 3ba, 4ba, and 5ba have 0.10–0.14 ꢂ larger bowl
depths than their corresponding parent compounds 3aa, 4aa,
and 5aa, respectively. Compound 4ba is the deepest bowl,
with a depth of 1.34 ꢂ. Compounds 4 contain both corannu-
lene and sumanene segments. The bowl depths that were
measured from the sumanene core in 4aa (1.48 ꢂ) and 4ba
(1.50 ꢂ) also exceed that of sumanene (1.11 ꢂ).[27] Cyclopen-
taannulation in 4 insignificantly affects the bowl depth of the
sumanene core. Notably, the geometric calculations at the
B3LYP/cc-pVDZ level are highly consistent with the X-ray struc-
tural data (Table 4 and Tables S3 and S4 in the Supporting In-
formation).
1
2
3
4
5
6
7
10aa
10aa
10aa
10aa
10aa
10ba
10ac
0.10
0.04
0.02
0.01
5ꢁ10À3
0.02
4aa
4aa
4aa
4aa
4aa
4ba
4ac
trace
29
31[b]
14
trace
5
40
0.02
[a] Reaction was conducted with 11 (0.1 mmol) in DMF (5 mL).
[b] Ref. [17].
tion efficiency. The reaction in either a concentrated (0.1m) or
a diluted (5ꢁ10À3 m) solution yielded only a trace of 4aa (en-
tries 1 and 5 in Table 2). At the former concentration, 10aa did
not completely undergo the fourfold cyclization reactions, and
dechlorinated intermediates were formed, whereas a diluted
solution of 10aa caused many of the starting materials to be
retained. Under the optimal conditions herein, compound 4aa
was generated in 31% yield (entry 3 in Table 2). It is notewor-
thy that 4aa was first prepared in a 0.14% yield from 7,12-
bis(2-bromophenyl)benzo[k]fluoranthene
using
FVP
at
11008C.[24] Our protocol is superior to the conventional
method. This synthetic approach was also applied in prepara-
tions of 4ac and 4ba, and they were obtained in 40 and 5%
yields, respectively (entries 6 and 7 in Table 2). Presumably, the
highly curved structure of 4ba was the cause of the unsatisfac-
tory result.
The p-orbital axis vector (POAV) pyramidalization angle is an-
other useful metric of the curvature of buckybowls. The values
for planar benzene and C60 are determined to be 0 and 11.68,
respectively.[28] As presented in Figure 1, the POAV pyramidali-
zation angle is highest at the hub carbon atoms of a corannu-
lene core, and follows the order 3ꢀ4>5. All members of the
first two classes of compounds have a maximum POAV pyra-
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
600
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 3. Crystal structure data.
3aa
3ba
4aa
4ba
5aa
5ba
13aa
13ba
formula
Mr
C30H14
374.41
C32H16
400.45
C64H24
792.83
C34H14
422.45
C38H14
470.49
C40H16
496.53
C30H16
376.43
C32H18
402.46
solvent
T [K]
C6H6/MeOH
100(2)
CH2Cl2/MeOH
100(2)
CH2Cl2/MeOH
296(2)
CH2Cl2/MeOH
100(2)
CH2Cl2/MeOH
296(2)
CH2Cl2/MeOH
100(2)
CH2Cl2/MeOH
100(2)
CH2Cl2/MeOH
100(2)
crystal
system
space group
Z
monoclinic
trigonal
orthorhombic
orthorhombic
orthorhombic
orthorhombic
orthorhombic
monoclinic
¯
R3
18
C1c1
4
Cmc21
2
Cmc21
4
Cmc21
4
Pbca
8
Pna21
8
P121/n1
4
unit cell
dimensions
a [ꢂ]
b [ꢂ]
c [ꢂ]
20.4913(14)
10.0252(7)
8.4972(5)
90
18.3650(5)
18.3650(5)
29.5206(11)
90
17.7727(8)
12.6382(6)
8.1349(4)
90
16.897(4)
13.523(4)
8.179(2)
90
18.1315(4)
13.7073(3)
8.37540(10)
90
14.975(3)
16.934(4)
17.587(4)
90
20.405(12)
23.455(14)
7.762(4)
90
13.2867(9)
7.3257(4)
19.8902(14)
90
a [8]
b [8]
g [8]
103.5200(10)
90
1697.20(19)
2.84
90
120
8622.6(5)
4.54
this work
90
90
90
90
1868.9(8)
3.65
this work
90
90
90
90
90
90
3715(4)
6.43
[17]
104.199(2)
90
1876.9(2)
4.30
V [ꢂ3]
R factor [%]
ref.
1827.22(15)
3.28
[17]
2081.57(7)
3.74
[18]
4459.8(17)
4.70
[18]
this work
this work
Table 4. Bowl depth.[a]
X-ray
calcd
Reference
1
0.87
1.19
1.32
1.24 (1.48)
1.34 (1.50)
1.07
1.21, 1.07
1.04[b]
1.19
0.88
1.19
1.31
1.23 (1.49)
1.34 (1.52)
1.07
1.21, 1.07
–
–
–
[26]
[17]
this work
[17], [18]
this work
[18]
[18]
[17]
this work
[27]
3aa
3ba
4aa
4ba
5aa
5ba
13aa
13ba
2
1.11
[a] Bowl depths determined from corannulene core. The bowl depths for
the sumanene segment are shown in brackets. Theoretical studies were
calculated at B3LYP/cc-pVDZ level. [b] There are two molecules in the
asymmetric unit.
midalization angle of approximately 12.88, which exceeds that
of C60. Bowl-shaped molecules seldom have such high values.
To the best of the our knowledge, compounds 3, 4, and pen-
taindenocorannulene (12.78) are the most curved p-bowls that
map onto fullerenes.[29] Cyclopentaannulation does not in-
crease the maximum POAV pyramidalization angle in 3 or 4,
but it does do so in less curved 5 and acecorannulene.[30] This
result may suggest that the maximum POAV pyramidalization
angle reaches its highest value in 3 and 4, and the change in
curvature reflects only on an increase in the bowl depth. This
phenomenon has also been observed in a series of symmetric
corannulene-based molecules (C10)nH10 (n=2–5), whose struc-
tures change from a bowl (n=2, 3) to a tube (n=4, 5).[31] The
corannulene core at the end cap in C50H10 has the largest bowl
depth (calcd 12.2; exptl 12.38, 1.539 ꢂ[10]), but that in C30H10
(calcd 12.48) has the highest POAV pyramidalization angle.
The curvature of buckybowls causes their solid-state packing
to be highly interesting but complex. The stacking order of
buckybowls provides useful information concerning their po-
Figure 1. POAV pyramidalization angles of buckybowls based on X-ray crys-
tallography. The values obtained by averaging the symmetry-related data.
tential applications as organic materials with high electron mo-
bility (organic semiconductors),[7c] piezoelectricity or pyroelec-
tricity,[32] or the ability to generate second harmonics (nonlinear
optoelectronics).[33] The factors that are required to make
bowl-in-bowl stacks and control the orientations of neighbor-
ing columns are not yet well-known.[2e,34] The crystallographic
data clearly reveal intermolecular interactions of buckybowls,
and they are useful for testing theoretical predictions. The in-
termolecular interactions include p–p stacking and CH···p hy-
drogen bonding. In this context, the distance of the p–p stack-
ing is determined by measuring the separation between the
centroids of two aromatic five- or six-membered rings. The in-
termolecular p–p contact of arenes/fullerenes complexes are
typically observed to be in the range 3.5–4.1 ꢂ.[34,35] A theoreti-
cal study has demonstrated that the concave-to-convex
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
601
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
stacked corannulene dimer has its highest binding energy
when two corannulenyl five-membered rings are separated
with an equilibrium distance of 3.64 ꢂ,[36] which somewhat ex-
ceeds the sum of the van der Waals radii of two carbons
(3.40 ꢂ).[37] The strength of the interaction between a CH
group and an aromatic p surface can be determined from the
CH–p contact distance, which is measured as the separation
from the concerned hydrogen atom to the centroid of the
nearest aromatic ring. Alternatively, the extent of CH···p inter-
action (referred to as the CH–Cp contact in context) can also
be gauged by the distance between the hydrogen atom and
a carbon atom in a large aromatic surface, such as the central
carbon of acenaphthylene. The CH–Cp data can be directly
compared to the results that were obtained by analyzing the
separation between the CH group with the fullerene convex
surface. Suezawa et al. estimated a regular value of 2.85Æ
0.13 ꢂ for both sp3CH–Cp and sp2CH–Cp contact distances
based on a systematic analysis of Cambridge Structural Data-
base.[38] This range of values is consistent with the sum of the
van der Waals radii of the corresponding atoms (2.90 ꢂ).[37]
The structure of 3aa was determined to have the polar
monoclinic space group C1c1. The translational element along
the a axis is offset by two steps along the column of the
stacked bowls, and a molecular pair A and B are therefore
formed (Figure 2 and Table 5). Within a molecular pair, the
shortest intermolecular carbon–carbon, p–p, and CH–Cp dis-
tances are 3.36, 3.61, and 2.86 ꢂ, respectively. The latter inter-
action is that of a hydrogen atom on a methylene group with
a carbon atom in the corannulenyl hub ring. The bowl direc-
tions of the molecular pair up the stack are the same, but
bowls slip from side to side with a slipping angle of 58. The
slipping angle is defined as that between the stacking axis and
the normal to the molecule. The interstack p–p and CH–p in-
Table 5. The stacking order parameters of the columnar structures.
Slipping
angle q [8][a]
A···A
[ꢂ]
3aa
4aa
4ba
5aa
13aa
13ba
5.0
31.8
32.1
20.0
17.5
29.2
8.50
8.14
8.18
8.38
7.76
7.33
[a] The slipping angle is defined as that between the stacking axis and
the normal to the molecule.
teractions are insignificant. Neighboring columnar stacks are
oriented with the same bowl direction, forming polar crystals.
Buckybowl 3ba crystallizes with the centrosymmetric trigo-
¯
nal R3 space group. In the crystalline state, the 3ba molecules
do not form columnar stacks. Instead, every set of six mole-
cules constitutes a propeller-like structure (Figure 3). Within
Figure 3. Crystal packing in 3ba: a) concave/convex interaction, b) propeller-
like stack, and c) concave/concave interaction. In (b) and (c), only carbon
atoms are shown for clarity.
this propeller structure, the acenaphthenyl moiety of 3ba can
be treated as a propeller blade and two neighboring propeller
blades are arranged one behind the other. Along the resulting
C3 propeller axis, three of the six ethylene-bridged moieties are
oriented toward the front, and the other three point toward
the back. If the three ethylene-bridged groups on each side
are regarded as the components of an “unconnected cyclohex-
ane”, then all of the “unconnected distances” are equal, with
a value of 3.82 ꢂ. The intermolecular interactions are complex.
Each molecule is observed to exhibit the concave/convex (cc/
cv) and the convex/convex (cv/cv) interactions with its nearest
neighbors. In the cc/cv interaction, the effective p–p overlap-
ping is small. The shortest CH–p contact distance is deter-
mined to be 2.65 ꢂ. In the cv/cv interaction with neighboring
molecules, the shortest p–p and CH–p contact distances are
observed to be 3.70 and 2.78 ꢂ, respectively.
Figure 2. Crystal packing in 3aa along: a) the a axis, and b) the c axis; only
the carbon atoms are shown for clarity. c) Fragment of the extended 2D
stack.
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
602
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The structure of 4aa was solved to have the polar ortho-
rhombic space group Cmc21. All bowls form bowl-in-bowl
stacks and all columnar stacks are arranged with a single bowl
direction (Figure 4). The translational element along the a axis
Figure 5. Crystal packing in 4ba along: a) the a axis, and b) the c axis; only
carbon atoms are shown for clarity. c) Fragment of the extended 2D stack.
Figure 4. Crystal packing in 4aa along: a) the a axis, and b) the c axis; only
carbon atoms are shown for clarity. c) Fragment of the extended 2D stack.
moves two steps along the column of stacked bowls. The
bowls slip from side to side with a slipping angle of 31.88, pro-
ducing a small effective p–p overlap. The shortest carbon–
carbon contact distance and CH–p separation within a molecu-
lar pair in a columnar stack are observed to be 3.35 and
3.06 ꢂ, respectively. The interstack p–p interaction is insignifi-
cant; the shortest interstack CH–Cp contact distance is deter-
mined to be 2.80 ꢂ.
Buckybowl 4ba crystallizes in the polar orthorhombic space
group Cmc21. The translational element along the a axis shifts
two steps along the column of stacked bowls, forming a molec-
ular pair A and B (Figure 5 and Table 5). The stacked molecular
pair have the same bowl directions, but the bowls slip from
side to side with a slipping angle of 32.18. The shortest
carbon–carbon contact distance within a molecular pair is
3.24 ꢂ, but the p–p and the CH···p interactions are negligible.
An observed significant CH–Cp contact distance (2.85 ꢂ) corre-
sponds to a separation between a hydrogen atom on an ethyl-
ene group in a molecule A and a carbon atom of the corannu-
lenyl hub ring in the closest molecule A within a columnar
stack. The bowl directions are the same in all neighboring col-
umnar stacks. The interstack p–p interaction is insignificant;
the shortest interstack CH–Cp contact distance is determined
to be 2.81 ꢂ.
Figure 6. Crystal packing in 5aa along: a) the a axis, and b) the c axis; only
carbon atoms are shown for clarity. c) Fragment of the extended 2D stack.
directions in a molecular pair up the stack are identical, but
the bowls slip from side to side with a slipping angle of 20.08
(Table 5). Neighboring columnar stacks have the same bowl di-
rection. The interstack p–p interaction is unimportant; the in-
terstack CH–Cp contact distances start from 2.91 ꢂ.
Molecules of 5ba pack in a centrosymmetric crystal form
with the orthorhombic space group Pbca. Buckybowl 5ba has
a complex molecular packing. All eight of the molecules in
a unit cell orient differently (Figure 7). All molecules form
“dimer” pairs through the interactions of the protons on the
ethylene bridge with the p surface, forming the CH–Cp contact
distance of 2.62 ꢂ. No significant p–p interaction occurs within
the “dimer”. In addition, each molecule has strong intermolec-
The structure of 5aa was determined to exhibit the polar or-
thorhombic space group Cmc21. The translational element
along the a axis is offset by two steps up the column of
stacked bowls (Figure 6). Within a molecular pair in a columnar
stack, the shortest carbon–carbon and p–p contact distances
are determined to be 3.20 and 3.82 ꢂ, respectively. The bowl
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
603
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
3.45 and 3.86 ꢂ, respectively. The CH···p interaction between
a methylene proton and a carbon atom in the corannulenyl
hub ring is the strongest, with a value of 2.80 ꢂ. The bowl di-
rections of each molecular pair up the stack are the same, but
the bowls slip from side to side with a slipping angle of 17.58
(Table 5). Neighboring columnar stacks have the same bowl di-
rection. The interstack p–p interaction is insignificant; the
shortest interstack CH–Cp contact distance is found to be
2.70 ꢂ.
The structure of 13ba was determined to exhibit the centro-
symmetric monoclinic space group P121/n1. All molecules form
bowl-in-bowl stacks, but each columnar stack has one neigh-
bor with the same bowl direction and one with the opposite
bowl direction (Figure 9). Within a columnar stack, the bowls
Figure 7. Crystal packing in 5ba: a) the unit cell, and b) the “dimer” struc-
ture. Only the carbon atoms are shown for clarity.
ular interactions with other nearest neighbors. The shortest p–
p and CH–Cp contact distances are measured to be 3.52 and
2.67 ꢂ, respectively.
The structure of 13aa was determined to exhibit the polar
orthorhombic space group Pna21, with half of the molecules in
the asymmetric unit. The two kinds of molecules have very
similar structural data, including bond lengths, bond angles,
bowl depths and POAV angles (see the Supporting Informa-
tion). The translational element along the a axis moves two
steps up the column of the stacked bowls (Figure 8). The
shortest carbon–carbon contact distance and p–p separation
in a molecular pair in a columnar stack are determined to be
Figure 9. Crystal packing in 13ba: a) along the b axis, b) interstack convex/
convex interaction, and c) fragment of the extended 2D stack. Only the
carbon atoms are shown for clarity.
slip from side to side with a slipping angle of 29.28 (Table 5),
and the p–p interaction is negligible. The shortest CH–p con-
tact distance within a packed column is determined to be
2.72 ꢂ. The interstack p–p interaction is unimportant. A signifi-
cant interstack CH–Cp contact separation is observed to be
2.81 ꢂ, which is the distance between a methylene proton and
a carbon atom of the corannulenyl hub ring.
Among the eight buckybowls investigated herein, 3aa, 4aa,
4ba, 5aa, and 13aa all form polar crystals, as verified by their
space groups (Table 3). Interestingly, 4aa, 4ba, and 5aa all
crystallize with the orthorhombic space group Cmc21, and they
have similar packing patterns. Buckybowl 5aa is less curved
than 4aa and 4ba, and has a larger p-surface. The structural
characteristics may cause 5aa to have greater p–p surface
overlaps than the other two bowls and a smaller slipping
angle with the respect to the bowl-stacking axis. Both 3aa and
13aa are methylene-bridged bowls, and the methylene proton
provides strong intermolecular CH···p interactions within a col-
umnar stack. Unlike the polar columnar packing in 4aa/4ba,
the additional ethylene bridge in 3ba, 5ba, and 13ba critically
Figure 8. Crystal packing in 13aa along: a) the a axis, and b) the c axis; only
carbon atoms are shown for clarity. c) Fragment of the extended 2D stack.
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
604
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
affects molecular packing, such that it differs from that of the
parent compounds 3aa, 5aa, and 13aa, respectively. The
former three bowls crystallize with centrosymmetric space
groups as apolar crystals, but in different molecular stacks.
Bowl 13ba exhibits apolar columnar packing, whereas 5ba
and 3ba form dimers and cyclic hexamers, respectively. Al-
though nearest neighbors are known critically to influence the
packing, further investigations must be performed to identify
the factors that govern this effect.
Table 6. Inversion dynamics.[a]
Inversion barrier
Inversion
type
Reference
[kcalmolÀ1
]
°
°
DG1
DG2
1
ca. 10
–
–
–
I
I
I
II
II
II
II
II
I
[39], [42]
[17]
this work
[18]
this work
[18]
[18]
[40]
[14c]
[41]
[3a], [43]
3aa
3ba
4aa
4ba
5aa
5ba
14
56.2
84.4
134.3
169.8
116.3
148.4
73.8[b]
26.4
124.3
135.1
79.8
84.3
47.7[b]
–
Inversion dynamics
15
16
6.96[c]
–
–
I
I
Based on our preliminary investigations, the inversion dynam-
ics of buckybowls 3–5 can be classified to two types: Type I in-
volves a planar transition structure, and II involves an S-shaped
(nonplanar) transition structure (Scheme 3).[18] These transition
2
ca. 20
[a] Theoretical studies of were calculated at B3LYP/cc-pVDZ level, unless
otherwise mentioned. [b] Calculated at the B3LYP/6-311G** level. [c] Cal-
culated at the B3LYP/6-31G* level.
Measurements of the inversion barrier of dideutero-substi-
tuted 3aa reveal that it exceeds 40 kcalmolÀ1 [17]
Accordingly,
.
the inversion barrier of buckybowls 3–5 was analyzed theoreti-
cally using DFT calculations at the B3LYP/cc-pVDZ level
(Table 6). Unlike corannulene with an inversion barrier of ap-
proximately 10 kcalmolÀ1[39,42] and sumanene with a barrier of
around 20 kcalmolÀ1 [3,43]
these buckybowls have very high in-
,
version barriers (>50 kcalmolÀ1), and so they can be regarded
as static bowls at room temperature. A previous study of the
relationship between the structure and energy of corannulene-
based buckybowls has demonstrated that a deeper bowl corre-
sponds to a higher inversion barrier.[39] However, the order of
inversion barriers, 4>5>3, is inconsistent with the order of
bowl depths (Table 4). Importantly, the relationship between
bowl depth and inversion barrier described above, is based on
simple corannulene derivatives that undergo inversion process
via a planar transition structure. Although the studied mole-
cules herein are grouped into two types, depending by their
Scheme 3. Inversion dynamics of buckybowls.
states were verified to exhibit only one imaginary frequency. inversion route, the obtained relationship between bowl depth
The longitudinal length of these buckybowls critically influen- and inversion barrier is reasonable. Within a compound class,
ces their inversion dynamics: a larger length corresponds to peri-annelation increases the inversion barrier (Table 6).
a greater preference for route II. For example, bowl-to-bowl in-
versions of corannulene (1),[2e,39] sumanene (2),[3] and 3 pro-
ceed via a planar transition state (route I in Scheme 3). The in-
Photophysical properties
version of semibuckminsterfullerene (14),[40] 4, and 5 through The photophysical properties of buckybowls 3–5 in CH2Cl2
inversion route II is suggested to have a lower inversion barrier (10 mm) at room temperature were investigated and compared
than that through route I (Table 6). The inversion mechanisms with those of fullerenes C60 and C70 (Table 7 and the Support-
of “softer” 5aa[18] and “harder” 4aa (see the Supporting Infor- ing Information). Within compound classes 3 and 5, the effect
mation) were confirmed using the pseudo-intrinsic reaction co- of the ethylene bridge on the photophysical properties seems
ordinate (pseudo-IRC), and the results thus obtained support relatively unimportant because they have very similar spectra.
their bowl-to-bowl inversion via a nonplanar transition struc- Both the photoabsorption and photoluminescence of cyclo-
ture. Importantly, the large longitudinal length of a buckybowl pentaannulated 4ba, unlike those of 4aa, are significantly red-
is not the only cause of inversion through route II. Compounds shifted. Absorption bands of bowls 3 and 4 are below 450 nm,
15[14c] and 16[41] have longer longitudinal lengths than those of whereas 5 have absorption bands at greater than 450 nm. No-
14 and 4aa, respectively. However, the former two buckybowls tably, an absorption band in the region of 450–500 nm is im-
are less “condensed” and their inversions proceed through portant to distinguish C70 from C60.[44] Based on this finding,
route I. Only a “condensed” buckybowl with a large longitudi- the photoabsorption properties of 3/4 and 5 are similar to
nal length undergoes inversion through route II.
those of C60 and C70, respectively. The photoluminescence of
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
605
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
exhibits “bowl chirality”, arising from the three-dimensional ge-
ometry. Attempts were made herein to produce enantiopure
forms from the racemates of 4ac by chiral resolution. The
chiral resolution of 4ac was performed with HPLC by using
a DAICEL CHIRALPAK IA chiral column. A mixed eluent system
composed of methanol and 2-propanol (1:1) gave the best re-
sults, clearly yielding two well-resolved peaks (see the Support-
ing Information). Owing to its very low solubility in the eluent
system, the chiral resolution of 4ac in useful amounts is im-
practical.
Table 7. Photophysical properties of buckybowls.[a]
[b]
Compound
labs [nm]
lem [nm]
3aa
3ba
4aa
4ba
5aa
5ba
C60
287, 337, 374 (sh), 405 (sh)
228, 268, 293, 336, 378 (sh), 400 (sh)
238, 288, 297, 308, 354 (sh)
252 (sh), 274, 300, 314, 370 (sh)
362, 421, 436 (sh), 461 (sh), 502 (sh)
345, 362, 416, 439, 466, 502 (sh)
336, 407
416, 436, 463 (sh)
414, 437, 462 (sh)
438, 466, 493 (sh)
467, 550 (sh)
528, 560 (sh)
528, 560 (sh)
C70
336, 366, 383, 470, 551 (sh)
[a] Samples (concentration ꢀ10À5 m) measured in CH2Cl2; sh=shoulder.
Wavelengths shown in italics represent excitation wavelengths.
Conclusion
A simple method for synthesizing highly curved buckybowls
was developed. These products can be directly generated from
easily obtained planar precursors under mild reaction condi-
tions. The bowl-shaped molecules thus formed exhibit a large
bowl depth and a high bowl-to-bowl inversion barrier. These
p-bowls can be used to study
C60 and C70 is known to be very weak because of the highly ef-
ficient intersystem crossing,[45] whereas 3 and 5 both exhibit
strong fluorescence at approximately 416/436 and 528 nm, re-
spectively. The photoluminescence of 4 is very weak.
bowl-to-bowl inversion dynam-
ics because they have planar or
S-shaped transition structures. It
is noteworthy that compounds
4 should be suitable starting
materials for constructing the
smallest
corannulene-based
carbon nanotube (C40H10).[31] The
electron mobility of polar bucky-
bowl crystals and the chiral aux-
iliary-assisted resolution of buck-
ybowls are currently being ex-
amined.
Figure 10. HOMO and LUMO distributions of compounds 3–5.
Figure 10 demonstrates that ethylene bridges of 3ba and
5ba participate similarly in establishing their HOMOs and
LUMOs, explaining why the photophysical properties of 3ba
and 5ba are similar to those of their parent compounds. How-
ever, the ethylene bridge in 4ba influences the HOMO more
strongly than it does the LUMO, increasing the potential of the
HOMO more than it does that of the LUMO (Table S1 in the
Supporting Information). The consequently reduced band gap
is responsible for the redshifting of the UV and PL spectra of
4ba relative to those of 4aa. Like C70, bowls 5 have longer-
wavelength absorptions than 3 and 4, although 3, 4 and 5 can
all map onto C70. The extent of p-conjugation of both HOMOs
and LUMOs of the three compound classes follows the order
5>3/4 (Figure 10). Accordingly, bowls 5 have the smaller band
gap. Time-dependent density functional theory (TD-DFT) calcu-
lations demonstrate that the strongest vertical excitation wave-
lengths for 3aa, 4aa, and 5aa are 306, 345, and 439 nm, re-
spectively (Table S2 in the Supporting Information).
Experimental Section
General procedures
1H and 13C NMR: Bruker 300 (300 and 75 MHz), 400 (400 and
100 MHz) and 500 (500 and 125 MHz). MS: High-resolution mass
spectra (HRMS) were obtained on Finnigan MAT-95XL high-resolu-
tion mass spectrometer. X-ray crystal structure determination:
Single-crystal X-ray diffraction was performed on a Bruker APEX
DUO at 100(2) K. Data were collected and processed by using
APEX II 4 K CCD detector. Melting points were determined with
a Bꢃchi melting point apparatus B545 and are uncorrected. UV
spectra were recorded with VARIAN CARY 50 Probe. Photolumines-
cence experiments were accomplished with Jasco FP-6300.
Procedure for the preparation of buckybowls
The procedures for preparing buckybowls 3/4[17] and 5[18] have
been described previously. The Supporting Information provides
analytic data concerning new compounds generated herein.
The high inversion barriers of 3–5 cannot be measured
using common NMR techniques, such as variable-temperature
and 2D EXSY methods, owing to instrumental limitations. Since
bowl inversion of an enantiopure bowl corresponds to the rac-
emization process, the racemization at high temperature may Acknowledgements
help to confirm experimentally the value of the inversion barri-
er. Numerous chiral buckybowls have been synthesized and ex-
amined.[46] Mono-substituted 4ac is a chiral buckybowl, which
This work was supported by the National Science Council of
Taiwan (NSC 101-2628-M-006-002-MY3 and NSC 99-2113-M-
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
606
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[13] H. A. Reisch, M. S. Bratcher, L. T. Scott, Org. Lett. 2000, 2, 1427.
[14] Several less-curved buckybowls have been synthesized by metal-cata-
lyzed reactions. For recent examples, see: a) H.-I. Chang, H.-T. Huang, C.-
H. Huang, M.-Y. Kuo, Y.-T. Wu, Chem. Commun. 2010, 46, 7241; b) A. C.
Whalley, K. N. Plunkett, A. A. Gorodetsky, C. L. Schenck, C.-Y. Chiu, M. L.
Steigerwald, C. Nuckolls, Chem. Sci. 2011, 2, 132; c) Y.-T. Wu, T. Hayama,
K. K. Baldridge, A. Linden, J. S. Siegel, J. Am. Chem. Soc. 2006, 128, 6870;
d) H. A. Wegner, H. Reisch, K. Rauch, A. Demeter, K. A. Zachariasse, A. de
Meijere, L. T. Scott, J. Org. Chem. 2006, 71, 9080; e) Z. Marcinow, A.
Sygula, A. Ellern, P. W. Rabideau, Org. Lett. 2001, 3, 3527; f) L. Wang, P. B.
Shevlin, Org. Lett. 2000, 2, 3703. For other nonpyrolysis coupling tech-
niques, see: g) K. T. Rim, M. Siaj, S. Xiao, M. Myers, V. D. Carpentier, L.
Liu, C. Su, M. L. Steigerwald, M. S. Hybertsen, P. H. McBreen, G. W. Flynn,
C. Nuckolls, Angew. Chem. 2007, 119, 8037; Angew. Chem. Int. Ed. 2007,
46, 7891; h) O. Allemann, S. Duttwyler, P. Romanato, K. K. Baldridge, J. S.
Siegel, Science 2011, 332, 574; i) K. Y. Amsharov, M. A. Kabdulov, M.
Jansen, Angew. Chem. 2012, 124, 4672; Angew. Chem. Int. Ed. 2012, 51,
4594. For a saddle-shaped molecule based on corannulene, see: j) K. Ka-
wasumi, Q. Zhang, Y. Segawa, L. T. Scott, K. Itami, Nat. Chem. 2013, 5,
739.
260-007-MY3). We also thank Prof. S.-L. Wang and Ms. P.-L.
Chen (National Tsing Hua University, Taiwan) for the X-ray
structure analyses, and the National Center for High-perfor-
mance Computing for providing computational resources.
Keywords: fluoranthene · fullerene · palladium · rhodium · X-
ray diffraction
[1] A. Beck, M. N. Bleicher, D. W. Crowe, in Excursions into Mathematics,
Worth, New York, 1969.
[2] a) W. E. Barth, R. G. Lawton, J. Am. Chem. Soc. 1966, 88, 380. Reviews for
corannulene-type buckybowls, see: b) Fragments of Fullerenes and
Carbon Nanotubes: Designed Synthesis, Unusual Reactions, and Coordina-
tion Chemistry (Eds.: M. A. Petrukhina, L. T. Scott), Wiley, Hoboken, 2012;
c) A. Sygula, Eur. J. Org. Chem. 2011, 1611; d) V. M. Tsefrikas, L. T. Scott,
Chem. Rev. 2006, 106, 4868; e) Y.-T. Wu, J. S. Siegel, Chem. Rev. 2006,
106, 4843; f) A. Sygula, P. W. Rabideau, in Carbon-Rich Compounds: From
Molecules to Materials (Eds.: M. Haley, R. Tykwinski), Wiley-VCH, Wein-
heim, 2006, p.529; g) L. T. Scott, Angew. Chem. 2004, 116, 5102; Angew.
Chem. Int. Ed. 2004, 43, 4994; h) L. T. Scott, H. E. Bronstein, D. V. Preda,
R. B. M. Ansems, M. S. Bratcher, S. Hagen, Pure Appl. Chem. 1999, 71,
209; i) G. Mehta, H. S. P. Rao, Tetrahedron 1998, 54, 13325; j) L. T. Scott,
Pure Appl. Chem. 1996, 68, 291; k) P. W. Rabideau, A. Sygula, Acc. Chem.
Res. 1996, 29, 235; l) T. J. Seiders, J. S. Siegel, Chem. Br. 1995, 31, 307.
For a kilogram-scale synthesis of corannulene, see: m) A. M. Butterfield,
B. Gilomen, J. S. Siegel, Org. Process Res. Dev. 2012, 16, 664.
[15] a) B. D. Steinberg, E. A. Jackson, A. S. Filatov, A. Wakamiya, M. A. Petru-
khina, L. T. Scott, J. Am. Chem. Soc. 2009, 131, 10537; b) E. A. Jackson,
B. D. Steinberg, M. Bancu, A. Wakamiya, L. T. Scott, J. Am. Chem. Soc.
2007, 129, 484.
[16] T. Amaya, T. Nakata, T. Hirao, J. Am. Chem. Soc. 2009, 131, 10810.
[17] T.-C. Wu, H.-J. Hsin, M.-Y. Kuo, C.-H. Li, Y.-T. Wu, J. Am. Chem. Soc. 2011,
133, 16319.
[18] T.-C. Wu, M.-K. Chen, Y.-W. Lee, M.-Y. Kuo, Y.-T. Wu, Angew. Chem. 2013,
125, 1327; Angew. Chem. Int. Ed. 2013, 52, 1289.
[3] a) H. Sakurai, T. Daiko, T. Hirao, Science 2003, 301, 1878. Review for su-
manene-type buckybowls: b) S. Higashibayashi, H. Sakurai, Chem. Lett.
2011, 40, 122; c) T. Amaya, T. Hirao, Chem. Commun. 2011, 47, 10524.
[4] a) H. W. Kroto, J. R. Heath, S. C. O’Brien, R. F. Curl, R. E. Smalley, Nature
1985, 318, 162. For reviews on fullerenes, see: b) A. L. Balch, M. M. Olm-
stead, Chem. Rev. 1998, 98, 2123; c) Fullerenes: Principles and Applica-
tions (Eds.: F. Langa, J.-F. Nierengarten), RSC, Cambridge, 2007; d) C.
Thilgen, F. Diederich, Chem. Rev. 2006, 106, 5049; e) A. Hirsch, M. Bret-
treich, Fullerenes: Chemistry and Reactions, Wiley-VCH, Weinheim, 2005;
f) Fullerenes and Related Structures, in Top. Curr. Chem. Vol. 99 (Ed.: A.
Hirsch), Springer, Berlin, 1999.
[19] X-ray structures of [C76(D2)·Ag] and [C78(C2v)·Cu]: L. Epple, K. Y. Amsharov,
M. Jansen, Fullerenes Nanotubes Carbon Nanostruct. 2009, 17, 67.
[20] X-ray structure of [C84(Cs)·Ag] complex: L. Epple, K. Amsharov, K. Simeo-
nov, I. Dix, M. Jansen, Chem. Commun. 2008, 5610.
[21] T.-C. Wu, Y.-T. Wu, Synlett 2011, 741.
[22] Y.-H. Kung, Y.-S. Cheng, C.-C. Tai, W.-S. Liu, C.-C. Shin, C.-C. Ma, Y.-C. Tsai,
T.-C. Wu, M.-Y. Kuo, Y.-T. Wu, Chem. Eur. J. 2010, 16, 5909.
[23] C.-C. Hsiao, Y.-K. Lin, C.-J. Liu, T.-C. Wu, Y.-T. Wu, Adv. Synth. Catal. 2010,
352, 3267.
[24] M. D. Clayton, P. W. Rabideau, Tetrahedron Lett. 1997, 38, 741.
[25] CCDC 938415 (3aa), 938416 (3ba), 835497 (4aa), 938417 (4ba), 901335
(5aa), 901336 (5ba), 835496 (13aa) and 938418 (13ba) contain the
supplementary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[26] a) M. A. Petrukhina, K. W. Andreini, J. Mack, L. T. Scott, J. Org. Chem.
2005, 70, 5713; b) J. C. Hanson, C. E. Nordman, Acta Crystallogr. Sect. B
1976, 32, 1147.
[5] S. Iijima, Nature 1991, 354, 56.
[6] a) Y. Morita, S. Nishida, T. Kobayashi, K. Fukui, K. Sato, D. Shiomi, T. Takui,
K. Nakasuji, Org. Lett. 2004, 6, 1397; b) Y. Morita, A. Ueda, S. Nishida, K.
Fukui, T. Ise, D. Shiomi, K. Sato, T. Takui, K. Nakasuji, Angew. Chem. 2008,
120, 2065; Angew. Chem. Int. Ed. 2008, 47, 2035; c) A. Ueda, S. Nishida,
K. Fukui, T. Ise, D. Shiomi, K. Sato, T. Takui, K. Nakasuji, Y. Morita, Angew.
Chem. 2010, 122, 1722; Angew. Chem. Int. Ed. 2010, 49, 1678; d) A.
Ueda, K. Ogasawara, S. Nishida, T. Ise, T. Yoshino, S. Nakazawa, K. Sato, T.
Takui, K. Nakasuji, Y. Morita, Angew. Chem. 2010, 122, 6477; Angew.
Chem. Int. Ed. 2010, 49, 6333.
[27] H. Sakurai, T. Daiko, H. Sakane, T. Amaya, T. Hirao, J. Am. Chem. Soc.
2005, 127, 11580.
[28] a) R. C. Haddon, L. T. Scott, Pure Appl. Chem. 1986, 58, 137; b) R. C.
Haddon, Acc. Chem. Res. 1988, 21, 243; c) R. C. Haddon, J. Am. Chem.
Soc. 1990, 112, 3385; d) R. C. Haddon, Science 1993, 261, 1545.
[29] Tetrabenzo[4]circulene (quadrannulene) should be the most curved aro-
matic bowl, with the POAV pyramidalization angle of 178. For details,
[7] a) M. Mattarella, J. M. Haberl, J. Ruokolainen, E. M. Landau, R. Mezzenga,
J. S. Siegel, Chem. Commun. 2013, 49, 7204; b) D. Miyajima, K. Tashiro, F.
Araoka, H. Takezoe, J. Kim, K. Kato, M. Takata, T. Aida, J. Am. Chem. Soc.
2009, 131, 44; c) 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.
´
see: Bharat, R. Bhola, T. Bally, A. Valente, M. K. Cyranski, L. Dobrzycki,
S. M. Spain, P. Rempała, M. R. Chin, B. T. King, Angew. Chem. 2010, 122,
409; Angew. Chem. Int. Ed. 2010, 49, 399.
[30] The crystal structure of acecorannulene has not yet been determined.
Theoretical DFT studies calculated at Becke3LYP/3-21G level indicated
the maximum POAV angle the bowl depth are 11.48 and 1.114 ꢂ, re-
spectively. For details, see ref. [14e].
[8] P. G. Collins, P. Avouris, Sci. Am. 2000, 283, 62.
[9] a) D. M. Guldi, N. Martin, Carbon Nanotubes and Related Structures: Syn-
thesis Characterization, Functionalization, and Applications, Wiley-VCH,
Weinheim, 2010; b) N. Komatsu, F. Wang, Materials 2010, 3, 3818.
[10] L. T. Scott, E. A. Jackson, Q. Zhang, B. D. Steinberg, M. Bancu, B. Li, J.
Am. Chem. Soc. 2012, 134, 107.
[31] K. K. Baldridge, J. S. Siegel, Theor. Chem. Acc. 1997, 97, 67.
[32] a) F. C. Krebs, P. S. Larsen, J. Larsen, C. S. Jacobsen, C. Boutton, N.
Thorup, J. Am. Chem. Soc. 1997, 119, 1208; b) G. K. H. Madsen, F. C.
Krebs, B. Lebech, F. K. Larsen, Chem. Eur. J. 2000, 6, 1797. For a review,
see: c) D. Y. Curtin, I. C. Paul, Chem. Rev. 1981, 81, 525.
[33] C. Bosshard, R. Spreiter, U. Meier, I. Liakatas, M. Bçsch, M. Jꢄger, S. Man-
etta, S. Follonier, P. Gꢃnter, in Crystal Engineering: From Molecules and
Crystals to Materials (Eds.: D. Braga, F. Grepioni, A. G. Orpen), Kluwer,
Dordrecht, 1999, p.261.
[11] The curvature of C60 and a highly curved buckybowl is compared by
the maximum POAV (p-orbital axis vector) pyramidalization angle, and
their values are 11.6 and larger than 128, respectively.
[12] Synthesis of circumtrindene from decacyclene is one such example. For
details, see: a) R. B. M. Ansems, L. T. Scott, J. Am. Chem. Soc. 2000, 122,
2719; b) D. M. Forkey, S. Attar, B. C. Noll, R. Koerner, M. M. Olmstead,
A. L. Balch, J. Am. Chem. Soc. 1997, 119, 5766. The crystal structure of
circumtrindene has been determined. The maximum POAV pyramidali-
zation angle is about 12.28.
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
607
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[34] a) S. Filatov, L. T. Scott, M. A. Petrukhina, Cryst. Growth Des. 2010, 10,
4607; b) L. Zoppi, L. Martin-Samos, K. K. Baldridge, J. Am. Chem. Soc.
2011, 133, 14002.
[35] a) P. D. W. Boyd, C. A. Reed, Acc. Chem. Res. 2005, 38, 235; b) J. L.
Atwood, L. J. Barbour, C. L. Raston, Cryst. Growth Des. 2002, 2, 3; c) H.
Matsubara, A. Hasegawa, K. Shiwaku, K. Asano, M. Uno, S. Takahashi, K.
Yamamoto, Chem. Lett. 1998, 923; d) A. L. Balch, M. M. Olmstead, Coord.
Chem. Rev. 1999, 185–186, 601.
7891; b) P. U. Biedermann, S. Pogodin, I. Agranat, J. Org. Chem. 1999, 64,
3655.
[41] a) T. C. Dinadayalane, U. D. Priyakumar, G. N. Sastry, J. Mol. Struct. THEO-
CHEM 2001, 543, 1; b) T. C. Dinadayalane, G. N. Sastry, J. Org. Chem.
2002, 67, 4605.
[42] a) L. T. Scott, M. M. Hashemi, M. S. Bratcher, J. Am. Chem. Soc. 1992, 114,
1920; b) A. Sygula, A. H. Abdourazak, P. W. Rabideau, J. Am. Chem. Soc.
1996, 118, 339.
[36] a) A. Sygula, S. Saebø, Int. J. Quantum Chem. 2009, 109, 65. For other
theoretical studies on the effects of curvature on p–p interactions, see:
b) T. Janowski, P. Pulay, A. A. S. Karunarathna, A. Sygula, S. Saebø, Chem.
Phys. Lett. 2011, 512, 155; c) M. R. Kennedy, L. A. Burns, C. D. Sherrill, J.
Phys. Chem. A 2012, 116, 11920; d) D. Vijay, H. Sakurai, G. N. Sastry, Int. J.
Quantum Chem. 2011, 111, 1893. For a review, see: e) T. Kawase, H.
Kurata, Chem. Rev. 2006, 106, 5250.
[37] L. Pauling, The Nature of the Chemical Bond, 3rd ed., Cornell University
Press, Ithaca, 1960.
[38] H. Suezawa, T. Yoshida, S. Ishihara, Y. Umezawa, M. Nishio, CrystEng-
Comm 2003, 5, 514.
[43] T. Amaya, H. Sakane, T. Muneishi, T. Hirao, Chem. Commun. 2008, 765.
[44] a) H. Ajie, M. M. Alvarez, S. J. Anz, R. D. Beck, F. Diederich, K. Fostiropou-
los, D. R. Huffman, W. Krꢄtschmer, Y. Rubin, K. E. Schriver, D. Sensharma,
R. L. Whetten, J. Phys. Chem. 1990, 94, 8630; b) J. Catalꢅn, J. Elguero, J.
Am. Chem. Soc. 1993, 115, 9249.
[45] a) M. N. Berberan-Santos, J. M. M. Garcia, J. Am. Chem. Soc. 1996, 118,
9391; b) S. Nascimento, C. Baleizꢅo, M. N. Berberan-Santos, Springer Ser.
Fluoresc. 2008, 4, 151.
[46] Selected examples for chiral buckybowls: a) S. Higashibayashi, H. Sakur-
ai, J. Am. Chem. Soc. 2008, 130, 8592; b) Q. Tan, S. Higashibayashi, S.
Karanjit, H. Sakurai, Nat. Commun. 2012, 3, 891.
[39] T. J. Seiders, K. K. Baldridge, G. H. Grube, J. S. Siegel, J. Am. Chem. Soc.
2001, 123, 517.
[40] The inversion of 14 was suggested to proceed via a non-planar inter-
mediate. For details, see: a) P. W. Rabideau, A. H. Abdourazak, H. E.
Folsom, Z. Marcinow, A. Sygula, R. Sygula, J. Am. Chem. Soc. 1994, 116,
Received: August 27, 2013
Published online on December 5, 2013
Chem. Eur. J. 2014, 20, 598 – 608
www.chemeurj.org
608
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim