The albatrossenes: Large, cleft-containing, polyphenyl polycyclic aromatic hydrocarbons

Article abstract of DOI:10.1021/ja971261k

The syntheses and X-ray structures of very large polycyclic aromatic compounds containing clefts defined by polyphenylaryl groups are described. The C₂-symmetric 'albatrossenes' 1,3-bis(heptaphenyl-2-naphthyl)benzene (7a) and 1,3-bis(heptaphenyl-1-naphthyl)benzene (12a), as well as brominated derivatives, were synthesized by the addition of tetraphenylbenzyne to the appropriate polyphenyl biscyclopentadienones. The 2-naphthyl isomers have wide, shallow clefts, and the 1-naphthyl isomers have deep, narrow clefts which were observed to change size dramatically in different crystal environments. In a similar way, 1,3,5-tris(pentaphenylphenyl)benzene (19), a D₃-symmetric molecular propeller with a diameter of 21 ?, and 1,3,5-tris(heptaphenyl-2-naphthyl)benzene (22) were prepared by the addition of diphenylacetylene and tetraphenylbenzyne, respectively, to a triscyclopentadienone.

Full text of DOI:10.1021/ja971261k

J. Am. Chem. Soc. 1997, 119, 7291-7302
7291
The Albatrossenes: Large, Cleft-Containing, Polyphenyl
Polycyclic Aromatic Hydrocarbons
Ling Tong, Douglas M. Ho, Nancy J. Vogelaar, Clarence E. Schutt, and
Robert A. Pascal, Jr.*
Department of Chemistry, Princeton UniVersity, Princeton, New Jersey 08544
ReceiVed April 21, 1997X
Abstract: The syntheses and X-ray structures of very large polycyclic aromatic compounds containing clefts defined
by polyphenylaryl groups are described. The C2-symmetric “albatrossenes” 1,3-bis(heptaphenyl-2-naphthyl)benzene
(7a) and 1,3-bis(heptaphenyl-1-naphthyl)benzene (12a), as well as brominated derivatives, were synthesized by the
addition of tetraphenylbenzyne to the appropriate polyphenyl biscyclopentadienones. The 2-naphthyl isomers have
wide, shallow clefts, and the 1-naphthyl isomers have deep, narrow clefts which were observed to change size
dramatically in different crystal environments. In a similar way, 1,3,5-tris(pentaphenylphenyl)benzene (19), a D3-
symmetric molecular propeller with a diameter of 21 Å, and 1,3,5-tris(heptaphenyl-2-naphthyl)benzene (22) were
prepared by the addition of diphenylacetylene and tetraphenylbenzyne, respectively, to a triscyclopentadienone.
Our persistent interest in highly substituted polycyclic
aromatic compounds has been focused chiefly on molecules with
great strain and unusual geometries, and as a result, the syntheses
of these compounds often employ extreme conditions and
proceed in low yield.1 A notable exception is octaphenylnaph-
thalene (1), which may be prepared in 26% yield in three simple
steps from commercially available starting materials.2 Octaphe-
nylnaphthalene is a relatively large molecule (C58H40), and for
many purposes it may be thought of as a rectangular hydro-
carbon slab with approximate dimensions of 14 Å × 12 Å × 6
Å. Compound 1 is extremely stable, and unlike many large
polycyclic aromatic compounds, it has good solubility properties.
Most significantly, the eight phenyl groups are derived from
benzil and 1,3-diphenylacetone, and so numerous modifications
of 1 are accessible from the many readily available derivatives
of its two simple precursors.
Results and Discussion
Syntheses of the Albatrossenes. The syntheses of two-
winged, C2-symmetric albatrossenes are outlined in Scheme 1.
At the outset of this study, the prime targets were the isomeric
1,3-bis(heptaphenyl-2-naphthyl)benzene (7a) and 1,3-bis(hep-
taphenyl-1-naphthyl)benzene (12a). These compounds are
superficially similar in planar drawings, but molecular mechanics
calculations4,5 indicated that their shapes would be quite
different. Compound 7a was predicted to have a wide, shallow
cleft, while that in 12a was expected to be deep and narrow,
with much greater crowding of the phenyl rings near the center
of the molecule.
The key intermediates in the albatrossene syntheses are
compounds in which two tetraphenylcyclopentadienone moieties
share a common phenyl group. The 3-phenyl group is common
in biscyclopentadienone 5a, the precursor of 7a. Compound
5a is a large molecule (C52H34O2), but its preparation requires
only the simplest reactions. 1,3-Benzenediacetic acid (3a) is
converted to the acid chloride and then treated with benzene
and aluminum chloride to give crude 1,3-bis(benzoylmethyl)-
benzene. Without purification, this diketone is oxidized to the
highly crystalline, bright yellow tetraketone 4a, and a double
aldol condensation with 2 equiv of 1,3-diphenylacetone gives
the biscyclopentadienone 5a.
The easy synthesis of 1 and its derivatives permits the
construction of large organic compounds containing clefts or
cavities defined by more than one octaarylnaphthalene. We
recently reported the synthesis and structure of one such
compound, bis(heptaphenyl-2-naphthyl)pyridine (2, “albatros-
sidine”), in which a pyridine nitrogen lies at the bottom of a
large shallow cleft formed by the two heptaphenylnaphthyl
“wings”.3 We now report the syntheses and structures of the
“albatrossenes”, polycyclic aromatic hydrocarbons containing
clefts defined by two polyphenylaryl groups attached to a central
benzene ring, as well as several three-winged analogs, which
are giant molecular propellers.
We were surprised to find that compound 5a was unknown.
The chemistry of biscyclopentadienones was studied in the
1960s by Ogliaruso et al.,6 but for some reason their research
focused on the preparation of para-linked bistetracyclones such
as 13 but not the meta-linked 5a. These authors went on to
synthesize 1,4-bis(pentaphenylphenyl)benzene6,7 (14, the first
albatrossene!) by heating 13 with diphenylacetylene, and they
(4) Molecular mechanics calculations were performed by using the
SYBYL5 force field implemented in the SPARTAN program package
(versions 3.0 and 4.1; Wavefunction, Inc., Irvine, CA).
(5) Clark, M.; Cramer, R. D., III; Van Opdenbosch, N. J. Comput. Chem.
1989, 10, 982-1012.
(6) (a) Ogliaruso, M. A.; Shadoff, L. A.; Becker, E. I. J. Org. Chem.
1963, 28, 2725-2728. (b) Ogliaruso, M. A.; Becker, E. I. J. Org. Chem.
1965, 30, 3354-3360.
(7) The standard systematic name for compound 14 is
2′,3′,5′,6′,2′′′,3′′′,5′′′,6′′′-octaphenyl-1,1′:4′,1′′:4′′,1′′′:4′′′,1′′′′-quinquephe-
nyl, but this is neither shorter nor more illuminating than “1,4-bis-
(pentaphenylphenyl)benzene”, so the latter style of nomenclature will be
employed in this paper.
X Abstract published in AdVance ACS Abstracts, July 15, 1997.
(1) Both extreme geometries and reaction conditions are found in these
two examples: (a) Pascal, R. A., Jr.; McMillan, W. D.; Van Engen, D. J.
Am. Chem. Soc. 1986, 108, 5652-5653. (b) Shibata, K.; Kulkarni, A. A.;
Ho, D. M.; Pascal, R. A., Jr. J. Am. Chem. Soc. 1994, 116, 5983-5984.
(2) Qiao, X.; Padula, M. A.; Ho, D. M.; Vogelaar, N. J.; Schutt, C. E.;
Pascal, R. A., Jr. J. Am. Chem. Soc. 1996, 118, 741-745.
(3) Tong, L.; Ho, D. M.; Vogelaar, N. J.; Schutt, C. E.; Pascal, R. A.,
Jr. Tetrahedron Lett. 1997, 38, 7-10.
S0002-7863(97)01261-4 CCC: $14.00 © 1997 American Chemical Society

7292 J. Am. Chem. Soc., Vol. 119, No. 31, 1997
Tong et al.
Scheme 1
also prepared a variety of molecules containing two para-linked
hexaphenylbenzene groups. However, compound 14 and all
of their other bishexaphenylbenzenes are extended structures
without clefts.
structure determination, compound 7b, which differs from 7a
only in the presence of a bromine atom on the central ring, was
prepared from 5-bromo-1,3-benzenediacetic acid (3b) by an
essentially identical sequence of reactions. All three polyphenyl
aromatics are very stable, crystalline compounds. However, 7a
and 7b are much more soluble in common organic solvents than
6 (we were unable to record the 13C NMR spectrum of 6), and
in general we have found the heptaphenylnaphthyl-substituted
compounds to be more soluble than their pentaphenylphenyl-
substituted counterparts, a desirable characteristic of the former.
The synthesis of albatrossene 12a, in which the naphthyl
wings are attached at C-1 (rather than at C-2 as in 7a), requires
a bistetracyclone in which the 2-phenyl group is shareds
compound 11a. In this sequence (Scheme 1), a mixed double
Claisen condensation of methyl phenylacetate and dimethyl 1,3-
benzenediacetate gave the diketone 10a, and aldol condensation
with 2 equiv of benzil then produced the desired biscyclopen-
tadienone 11a. In order to generate the hindered 12a, compound
11a was treated twice with a slight excess of isoamyl nitrite
and 8, ultimately giving 12a in 5% yield. Again, in order to
facilitate an X-ray structure determination, compound 12b,
which differs from 12a by the addition of two bromine atoms,
was prepared by essentially the same sequence of reactions from
methyl p-bromophenyl acetate (9b). In this case the yield of
the final aryne addition was somewhat higher (14%).
In any event, following the lead of Ogliaruso et al., compound
5a was heated with diphenylacetylene at 300 °C to yield 1,3-
bis(pentaphenylphenyl)benzene (6) in 86% yield, thus demon-
strating 5a’s suitability as a double Diels-Alder diene. We
then diazotized tetraphenylanthranilic acid2 (8) in the presence
of 5a to yield 1,3-bis(heptaphenyl-2-naphthyl)benzene (7a) in
a more modest yield of 12%. However, monoaddition products
are also isolated from this reaction, as well as unreacted 5a,
and these may be treated again with 8 to give substantially more
of the albatrossene 7a. Finally, in order to facilitate an X-ray

The Albatrossenes
J. Am. Chem. Soc., Vol. 119, No. 31, 1997 7293
Figure 1. X-ray structure of compound 7b (from crystal form A). Thermal ellipsoids have been drawn at the 50% probability level, and hydrogens
have been omitted for clarity.
The 1H NMR spectra of the albatrossenes consist entirely of
aromatic proton resonances, but it is notable that all but one of
these compounds show at least one resonance below δ 6.0, and
a few resonances fall at higher field than δ 5.0, suggestive of
aryl protons poised above the centers of nearby aromatic rings.
In addition, the 13C NMR spectra of the heptaphenylnaphthyl-
substituted albatrossenes are exceedingly complex, each with
more than 70 badly overlapped resonances. Only 42 lines would
be expected for each in the fast exchange limit for all single-
bond rotations, so these molecules experience at least some
restricted rotation at room temperature and possibly exist as
more than one discrete conformation on the NMR time scale.
Due to the complexity of the NMR spectra of these molecules,
the acquisition of X-ray crystallographic data is essential for
confident characterization.
X-ray Structures of the Albatrossenes. Ostensibly single
crystals of compounds 7a and 12a were easily obtained, and
X-ray diffraction data were collected, but neither structure could
be solved for several months. For this reason, the brominated
compounds 7b and 12b were prepared, so that the structures
could be solved by heavy atom methods. Two independent
structures of 7b (from two different crystal forms) and one of
12b were thus determined. At that point, the structure of 12a
was solved, which proved to contain two independent molecules
in the asymmetric unit. Thus we have two independent
structures of the albatrossenes 7 and three of the albatrossenes
12, a situation which permits an unusually reliable assessment
of the conformations of these complex molecules.
of a bulk sample. However, the best refinements of the two
resulting crystal structures indicated that the composition of
crystal A was (7b)1(nitrobenzene)3.5(toluene)1 and that of crystal
B was (7b)1(nitrobenzene)1.
The structures of 7b in these two crystals are very similar,
so only the molecular structure of 7b from crystal A, with the
crystallographic numbering scheme, is illustrated in Figure 1.
The molecules adopt a conformation with approximate C2
symmetry, and the two heptaphenylnaphthyl wings are oriented
trans to one another and roughly perpendicular to the central
benzene ring. Both of the perphenylnaphthyl groups have
substantial end-to-end twists8 (crystal A, 23.3° and 19.8°; B,
23.2° and 22.4°). There are many potential rotational degrees
of freedom in 7b, but its conformation is strongly constrained
by interactions of the phenyl groups near the center of the
molecule. Thus the edges of the C(1′)- and C(1′′)-phenyls press
into the faces of the C(3′′)- and C(3′)-phenyls, respectively, such
that the C(10′)- and C(10′′)-hydrogens are positioned only 2.74
and 2.80 Å (for crystal A; for crystal B, 2.69 and 2.73 Å) above
the mean planes of the C(3′′)- and C(3′)-phenyl rings. These
hydrogen atoms are probably responsible for the most highly
1
upfield resonances in the H NMR spectrum of 7b.
The packing of the molecules of 7b in crystals A and B is
also remarkably similar. In each case, the molecules pack cleft-
to-cleft and back-to-back in layers parallel to the ac plane. This
packing is illustrated in Figure 2, which also serves to highlight
the shallow cleft in compound 7b. There is scarcely any visible
difference in the packing when viewed from this perspective.
The principal differences between the structures A and B are
in the separation of adjacent ac layers and the slippage of these
layers relative to one other. The separation of the layers is also
illustrated in Figure 2. In crystal A, the centers of the layers
are separated by 16.7 Å, while in crystal B, the distance is only
13.3 Å. The net result is that in crystal A the layers of 7b are
(a) Compound 7b. Single crystals of 7b were obtained by
concentration of a nitrobenzene-toluene solution. Diffraction
data were collected on two different crystals from the same
crystallization tube. Both crystal A and crystal B were triclinic,
space group P1h, but their unit cells had different dimensions
and volumes. Each crystal contained one molecule of 7b in
the asymmetric unit, but they differed in the number and types
of solvent molecules of crystallization. Because the same tube
produced more than one type of crystal, the amount and identity
of the included solvent could not be checked by NMR analysis
(8) Pascal, R. A., Jr.; McMillan, W. D.; Van Engen, D.; Eason, R. G. J.
Am. Chem. Soc. 1987, 109, 4660-4665; see note 17 for the definition of
“end-to-end twist”.

7294 J. Am. Chem. Soc., Vol. 119, No. 31, 1997
Tong et al.
Figure 2. Two views each of the crystal packing of compound 7b in crystal A (right) and crystal B (left). The top views are perpendicular to the
ac plane; the bottom views are down the c axis. The hydrogen atoms and interstitial solvent molecules have been omitted.
almost entirely separated by solvent molecules, but in crystal
B the layers are in van der Waals contact except for a single
solvent site. It would seem from these two structures that the
nicely interlocked intralayer packing of 7b, which utilizes most
of the surface area of these molecules, provides the driving force
for crystallization, and that the interlayer packing of the edges
is almost incidental.
(b) Compounds 12. Crystals of compound 12b were
obtained by concentration of a CH2Cl2-CH3OH solution. The
crystals were trigonal, space group P3221, Z ) 3; thus each
molecule lies on a special position and possesses crystallographic
C2 symmetry. In addition, several solvent molecules are
included in the asymmetric unit. The structure of 12b, with
the crystallographic numbering scheme, is illustrated in Figure

The Albatrossenes
J. Am. Chem. Soc., Vol. 119, No. 31, 1997 7295
Figure 3. X-ray structure of compound 12b. Thermal ellipsoids have been drawn at the 50% probability level, and hydrogens have been omitted
for clarity.
3 and the cleft is more clearly shown in Figure 5. Once again,
the heptaphenylnaphthyl groups are oriented trans to one another
and roughly perpendicular to the central benzene ring, and the
naphthalenes have an end-to-end twist of 20.1°. The cleft in
compound 12b is much more distinctive than in 7b, and its shape
is a direct result of the stacking of the central benzene ring with
the two flanking, peri phenyl groups on the “bottom” of the
molecule. The centroids of these C(8′)-phenyls are forced to
within 3.2 Å of the mean plane of the central ring, and this
steric conflict forces the two naphthyl groups to bend toward
each other, narrowing and deepening the central cleft. In
addition, the edges of the C(8′)-phenyls are forced into the faces
of the C(2′)-phenyls of the opposite wing, which also promotes
the bending and twisting of the naphthyl groups toward each
other.
The structure of one of the two similar molecules of 12a is
illustrated in Figure 4. The most dramatic difference between
the structures of 12a and 12b is the collapse of the central cleft
of the molecule (Figure 5), even though this requires additional
distortion of the heptaphenylnaphthyl groups. For example, the
four independent naphthyl groups in the structure of 12a have
end-to-end twists of 29.6°, 26.2°, 20.6°, and 35.1°, all of which
are greater than the 20.1° twist seen in 12b. A “Venetian blind”
effect due to rotation of the naphthyl substituents relative to
the central benzene ring (illustrated below) also narrows the
cleft. Thus the dihedral angle between the mean planes of the
naphthyl wings in 12b and the central benzene is 73.0°, which
permits a relatively open cleft [C(4a′)-C(4a*) ) 7.51 Å], but
the corresponding dihedral angles in 12a are 58.2°, 58.4°, 58.7°,
and 52.9°, which closes the cleft considerably [C(4a′)-C(4a′′)
) 6.35 Å].
The cleft of compound 12b has rough dimensions of 3 Å ×
6 Å × 11 Å, and not surprisingly it contains a molecule of
solvent in the crystal, in this case a badly disordered methanol.
The molecules of 12b are packed around the crystallographic
32 screw axes so that successive clefts form a solvent channel
parallel to the c axis, and the included solvent is disordered
along this channel. Interestingly, P3221 is a chiral space group,
so each crystal contains a single enantiomer of 12b.
Crystals of the parent albatrossene 12a were obtained by
concentration of a CHCl3-CH3OH solution, but unlike 12b,
this material contained no solvent of crystallization. The crystals
were triclinic, space group P1h, Z ) 4; thus there are two
independent molecules of 12a in the asymmetric unit. This is
a very large structures220 carbonssso large that conventional
direct methods failed to solve it. However, the structure
eventually yielded to the X-ray structure solution program
CRUNCH,9 and once solved, it refined without difficulty.
Interestingly, these large differences in geometry do not reflect
great differences in energy. Molecular mechanics calculations4
(9) (a) de Gelder, R.; de Graaff, R. A. G.; Schenk, H. Acta Crystallogr.,
Sect. A 1993, 49, 287-293. (b) Beurskens, P. T.; Admiraal, G.; Beurskens,
G.; Bosman, W. P.; de Gelder, R.; Israel, R.; Smits, J. M. M. The DIRDIF-
94 program system. Crystallography Laboratory, University of Nijmegen,
The Netherlands, 1994.

7296 J. Am. Chem. Soc., Vol. 119, No. 31, 1997
Tong et al.
conformation produced by relatively modest perturbations of
critical bond angles and torsional angles. Because of the large
size of the molecule, these distortions translate into very
substantial changes in the size and shape of the cleft. We have
previously observed a similar phenomenon in the crystal
structures of Rebek’s acridine diacid with various guests.10 With
diamine guests (pyrazine and quinoxaline) the cleft in the diacid
opened to as much as 8.36 Å, but with chloride ion bound, the
cleft narrowed to 6.21 Å, in what had previously been thought
to be a relatively rigid molecule.
In the four X-ray structures reported above, there are nine
independent octaphenylnaphthalene substructures. Every one
of these is twisted, with end-to-end twists ranging from 19.8°
to 35.1°, in contrast to the structure of octaphenylnaphthalene
(1) itself, which adopts a Ci symmetric conformation.2 Twisted
naphthyl groups were also observed in the structure of alba-
trossidine 2,3 so it is tempting to speculate that it is the
octaphenylnaphthalene structure which is anomalous! This
suggestion is supported by AM1 calculations11 which indicate
that the twisted, C2 symmetric conformation of 1 is 2.6 kcal/
mol lower in energy than the crystallographically observed Ci
conformation.2 However, the octaphenylnaphthalene groups in
the albatrossenes are interacting with each other, and this may
bias their conformations. A simple derivative of 1 should be
prepared (e.g., with one methyl group added) and its structure
determined, in order to resolve this issue.
Syntheses of Giant Molecular Propellers. It is apparent
from the structures of the albatrossenes described above that it
should be possible to place at least three polyphenylaryl groups
on a single benzene ring, as in compound 19 (Scheme 2).
Conversion of 1,3,5-benzenetriacetic acid12 (15) to the acid
chloride followed by Friedel-Crafts acylation with benzene
gave triketone 16, and this was oxidized to the highly crystalline
hexaketone 17 with DMSO/HBr. No problem was encountered
in the condensation of 17 with diphenylacetone, which afforded
triscyclopentadienone 18 in 77% yield, despite the apparent
steric hindrance in the product. When 18 was heated with
diphenylacetylene at 300 °C, 1,3,5-tris(pentaphenylphenyl)-
benzene (19) was isolated in 35% yield, the high temperature
overcoming any steric barrier to this Diels-Alder reaction. In
addition, a suitably desymmetrized derivative of 19 was required
for dynamic NMR studies, so the hexamethoxy derivative 21
was prepared in 46% yield by heating 18 with bis(4-methox-
yphenyl)acetylene (20).
Figure 4. X-ray structure of compound 12a. In the lower figure,
thermal ellipsoids have been drawn at the 50% probability level, and
for clarity only the ipso carbons of the phenyl substituents have been
included.
Given the synthesis of 19, it seemed likely that a triple
addition of tetraphenylbenzyne to 18 would succeed. Indeed,
diazotization of tetraphenylanthranilic acid (8) in the presence
of 18 gave 1,3,5-tris(heptaphenyl-2-naphthyl)benzene (22) in
15% yield. Thus this enormous hydrocarbon (C162H108) is
prepared in only four steps from 15, which is itself only one
step removed from inexpensive commercial starting materials.12
Interestingly, our search of the literature indicates that compound
22 is the largest hydrocarbon composed entirely of aromatic
rings to have been prepared in pure form.
X-ray Structure of Compound 19. Single crystals of 19
were obtained from a CHCl3-CH3OH solution, and the intitial
X-ray measurements indicated that the crystals were monoclinic,
space group P21/c, with either Z ) 8 or Z ) 12, that is, either
two or three crystallographically independent molecules of 19
in the asymmetric unit. With at least 228 independent carbons
Figure 5. Stereoview of the structures of compounds 12b (above) and
12a (below) which illustrates the collapse of the cleft in compound
12a.
indicate that the “closed-cleft” structure 12a is less than 1.5
kcal/mol higher in energy than the “open-cleft” 12b, so the clefts
in the compounds 12 can easily “breathe” in order to accom-
modate guests of various sizes, or none at all.
Conformational Preferences of the Albatrossenes. In each
of the albatrossene structures the molecule adopts a conforma-
tion of approximate or exact C2 symmetry where the two
polyphenylnaphthyl groups are trans to one another. We have
no direct evidence for the existence of a second, cis conforma-
tion of either 7 or 12, but the cis conformation of 7 is calculated
to be only about 1 kcal/mol higher in energy than the trans-7
and is very likely to be present in solution. On the other hand,
cis-12 is much higher in energy than trans-12, and it is probably
inaccessible. Each of the two observed structures of the
compounds 12 represents a distortion of an ideal ground state
(10) (a) Pascal, R. A., Jr.; Ho, D. M. J. Am. Chem. Soc. 1993, 115, 8507-
8508. (b) Pascal, R. A., Jr.; Ho, D. M. Tetrahedron 1994, 50, 8559-8568.
(11) Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J.
Am. Chem. Soc. 1985, 107, 3902-3909.
(12) Newman, M. S.; Lowrie, H. S. J. Am. Chem. Soc. 1954, 76, 6196-
6197.

The Albatrossenes
J. Am. Chem. Soc., Vol. 119, No. 31, 1997 7297
Scheme 2
and some solvent of crystallization (the crystals decomposed
upon drying), the structure solution was, similar to that of 12a,
well beyond the scope of conventional direct methods. After a
great many trials, the structure was solved by molecular
replacement using the SYBYL geometry. The crystal proved
to contain two very similar molecules of 19 in the asymmetric
unit as well as numerous badly disordered chloroform molecules.
Compound 19 (C114H78) is larger than all but one hydrocarbon
listed in the Cambridge Structural Database,13 hexakis(2,6-di-
tert-butyl-4-biphenylyl)ethane (C122H150),14 but there are no
coordinates deposited for this structure, and its accuracy has
been challenged.15
The structure of one of the molecules of 19 is illustrated in
Figures 6 and 7. Compound 19 is a three-bladed propeller 21
Å in diameter, and it adopts a conformation of approximate D3
symmetry. The blades are tilted with respect to the central
benzene ring with dihedral angles ranging from 55.5° to 60.2°,
and these blades define three broad, chiral clefts on the sides
of the molecule. The faces of the central benzene ring are
completely shielded from the external environment by two
“tents” formed from the interlocking o-phenyl groups on the
blades. The tents enclose two cavities, with a volume of
approximately 15 Å3 each.
Compound 19 is probably the largest D3-symmetric molecular
propeller to have been prepared. Numerous smaller molecular
propellers are known,16 but the vast majority of these compounds
have the structure (Aryl)3X, where X is a single atom. The
static and dynamic stereochemistry of these compounds has been
Figure 6. X-ray structure of compound 19. Thermal ellipsoids have
been drawn at the 30% probability level, and hydrogens have been
omitted for clarity. In the lower drawing all but the ipso carbons of the
peripheral phenyl groups have been omitted.
(13) Allen, F. H.; Kennard, O.; Taylor, R. Acc. Chem. Res. 1983, 16,
146-153.
(14) Stein, M.; Winter, W.; Rieker, A. Angew. Chem., Int. Ed. Engl.
1978, 17, 692-694.
extensively investigated,17 and some, such as perchlorotriph-
enylamine,18 have very high barriers to enantiomerization due
(15) Kahr, B.; Van Engen, D.; Mislow, K. J. Am. Chem. Soc. 1986, 108,
8305-8307.
(16) For a review of helical organic molecules, including molecular
propellers, see: Meurer, K. P.; Vo¨gtle, F. Top. Curr. Chem. 1985, 127,
3-75.
(17) For a thorough discussion of the stereochemistry of molecular
propellers, see: Mislow, K. Acc. Chem. Res. 1976, 9, 26-33.

7298 J. Am. Chem. Soc., Vol. 119, No. 31, 1997
Tong et al.
and C1 (one trans naphthyl) conformers, and their interconver-
sion is slow. When the 13C NMR spectrum of 22 was recorded
at 140 °C in 1,1,2,2-tetrachloroethane-d2, some resonances
coalesced, particularly in a region near δ 130, but the spectrum
still contained at least 60 sharp lines. This indicates that rotation
of the heptaphenyl-naphthyl groups remains slow at 140 °C and
that the barrier to rotation is above 20 kcal/mol; however, we
have been unable to achieve a chromatographic separation of
the diastereomeric C3 and C1 isomers, so the barrier is likely to
be below 26 kcal/mol.
Figure 7. Stereoview of the X-ray structure of compound 19.
Conclusions
to the high degree of distortion that must accompany the rotation
of the propeller blades about the aryl-X bond.16-18 The X-ray
structure of 19 shows that its central o-phenyl groups are tightly
packed, but their geometries are not unusual. However,
substantial distortions would be required to permit the inter-
conversion of enantiomers of 19, even though the blades are
much further apart in an (aryl)3benzene propeller than in the
classical (aryl)3X propellers. Resolved enantiomers might show
exceptional optical activity, so we investigated the barriers to
rotation in compound 19 and its derivatives.
Dynamic NMR Studies. The 13C NMR spectrum of 19
contains only 18 sharp resonances (for 114 carbons). So simple
a spectrum can result only if 19 has time-averaged D3h symmetry
or D3 symmetry with rapid rotation of the peripheral phenyl
groups. A large barrier might exist for rotation about the central
benzene-pentaphenylphenyl bonds in 19, but because of the
the high symmetry of 19, it would not be evident in the NMR
spectrum. In order to provide an unambiguous NMR “handle”
for this rotation in 19 the hexamethoxy derivative 21 was
prepared. Four conformers of compound 21 are likely: two
enantiomeric pairs of C3 and C1 diastereomers. The enantiomers
may interconvert by three 60° rotations about the central
benzene-pentaarylphenyl bonds, but interconversion of the C3
and C1 diastereomers requires a full 180° rotation about one of
these bonds.17 If all conformers are of approximately equal
energy, then statistical considerations ensure that there will be
a 1:3 ratio of C3 and C1 conformers.
The results presented herein show that very large aromatic
hydrocarbons with well-defined conformations need not be
difficult to prepare. Polyphenyl polycyclic aromatic compounds
have excellent solubility properties, and they are very stable.
Moreover, the polyphenylaryl groups have large surface areas
which can be used to create molecular clefts, and it is not
necessary to restrict these structures to simple hydrocarbons.
Their highly modular syntheses use only simple reactions that
are tolerant of many functional groups and permit the introduc-
tion of a wide variety of substituents on the peripheral phenyls.
The albatrossenes are akin to several polycyclic aromatic
systems of current interest. Most closely related are the C2-
symmetric 2,6-dinaphthylbenzoic acids that Siegel and co-
workers have proposed for use as chiral complexing agents or
catalysts.21 Compound 12 also calls to mind Zimmerman’s
“molecular tweezers”,22 which are forged from anthracene
subunits. The albatrossenes share the merits of these smaller
compounds but have a much greater surface area for interaction
with various guests. The three-bladed propellers 19 and 22, in
addition to their kinship with many smaller propeller struc-
tures,16,17 are reminiscent of aromatic dendrimers,23,24 but the
propellers are more densely substituted. The present chemistry
is simple enough that true dendrimers might be prepared with
similar reactions. For example, the triscyclopentadienone 18
could be treated with 3 equiv of a bisalkyne, and the remaining
alkynes in the product could be treated with more 18, etc., to
create an ever-branching structure. Since only Diels-Alder
reactions would be used in such a dendrimer synthesis, a great
variety of functional groups might be easily incorporated.
Thus the chemistry of polyphenyl polycyclic aromatic
compounds should be rich, and the preparation of highly
functionalized versions of these molecules is now in progress.
The 600 MHz 1H NMR spectrum of 21 at room temperature
shows five methyl resonances: an upfield group of three very
close resonances in a 1:2:1 ratio and a downfield group of two
resonances, separated by 1.6 Hz, in a 2:2 ratio. One of the
latter pair is slightly broader and is probably due to the
superposition of two separate resonances. This is the expected
pattern if the 60° rotations are fast at room temperature, so the
resolution of compound 19 into pure enantiomers will be
impossible at normal temperatures. A variable temperature
NMR experiment conducted at 500 MHz in toluene-d8 showed
coalescence of the downfield resonances at 60 °C, corresponding
Experimental Section
An “aqueous workup” is the washing of an organic extract with
aqueous Na2CO3 or NaHCO3, and then water, followed by drying over
Na2SO4.
5-Bromo-1,3-benzenediacetic acid (3b). A solution of 1,3-bis-
(bromomethyl)-5-bromobenzene25 (11.0 g, 32 mmol) and KCN (7.6 g,
117 mmol) in methanol (73 mL) and H2O (22 mL) was heated to reflux
for 5 h. After cooling, ether was added, and after an aqueous workup,
q
to a ∆Gc of 18.9 kcal/mol for the 180° rotation of the
pentaarylphenyl groups.19,20 Thus all possible single-bond
rotations appear to accessible for compound 19 and 21 on the
laboratory time scale at room temperature.
In contrast to that of 19, the 13C NMR spectrum of 22 contains
more than 70 badly overlapping lines; 40 resonances would be
expected in the fast exchange limit of all single bond rotations,
so there is also restricted rotation in this compound. It is
probable that 22 exists as a mixture of C3 (all cis naphthyls)
(21) Chen, C.-T.; Chadha, R.; Siegel, J. S.; Hardcastle, K. Tetrahedron
Lett. 1995, 36, 8403-8406.
(22) (a) Zimmerman, S. C.; VanZyl, C. M.; Hamilton, G. S. J. Am. Chem.
Soc. 1989, 111, 1373-1381. (b) Zimmerman, S. C.; Zeng, Z.; Wu, W.;
Reichert, D. E. J. Am. Chem. Soc. 1991, 113, 183-196. (c) Zimmerman,
S. C.; Wu, W.; Zeng, Z. J. Am. Chem. Soc. 1991, 113, 196-201.
(23) (a) Zhang, J.; Moore, J. S.; Xu, Z.; Aguirre, R. A. J. Am. Chem.
Soc. 1992, 114, 2273-2274. (b) Xu, Z.; Lee, S.; Moore, J. S. J. Am. Chem.
Soc. 1992, 114, 8730-8732. (c) Wu, Z.; Moore, J. S. Angew. Chem. Int.
Ed. Engl. 1993, 32, 1354-1357. (d) Zhang, J.; Pesak, D. J.; Ludwick, J.
L.; Moore, J. S. J. Am. Chem. Soc. 1994, 116, 4227-4239. (e) Bedard, T.
C.; Moore, J. S. J. Am. Chem. Soc. 1995, 17, 10662-10671.
(24) Tour, J. M. Chem. ReV. 1996, 96, 537-553.
(18) Hayes, K. S.; Nagumo, M.; Blount, J. F.; Mislow, K. J. Am. Chem.
Soc. 1980, 102, 2773-2776.
(19) For ∆ν ) 1.3 Hz in the absence of exchange at 500 MHz, the
Gutowsky-Holm approximation16 gives kc ) 2.9 s-1 for coalescence at
60 °C (333 K); if a transmission coefficient of 1 is assumed for the Eyring
equation, ∆Gc ) 18.9 kcal/mol.16
q
(20) Sandstrom, J. Dynamic NMR Spectroscopy; Academic: New York,
1982; pp 77-123.
(25) Hogberg, H.-E.; Wennerstrom, O. Acta Chem. Scand. 1982, B36,
661-667.

The Albatrossenes
J. Am. Chem. Soc., Vol. 119, No. 31, 1997 7299
concentration gave a brown solid. This solid (the dicyano derivative)
was heated in a solution of concentrated H2SO4 (60 mL), water (60
mL), and acetic acid (50 mL) at reflux overnight. The reaction mixture
was concentrated on a rotary evaporator and extracted with ether, and
after an aqueous workup, concentration gave the crude solid diacid,
which was recrystallized once from benzene-methanol to give pure
3b (3.4 g, 12.5 mmol, 39%), mp 176-177 °C: 1H NMR (DMSO-d6)
δ 3.58 (s, 4 H), 7.15 (s, 1 H), 7.37 (s, 2 H), 12.45 (br s, 1 H); 13C
NMR (DMSO-d6) δ 39.9, 121.1, 129.7, 130.5, 137.5, 172.2 (6 of 6
expected resonances observed); MS, m/z 272 (M+ [79Br], 40), 227 (M
- CO2H, 44), 199 (54), 182 (100); exact mass 271.9661, calcd for
C10H979BrO4 271.9685.
1,3-Bis(phenylglyoxalyl)benzene (4a). A mixture of 1,3-benzene-
diacetic acid (2.0 g, 10 mmol), thionyl chloride (7 mL), and benzene
(27 mL) was heated at reflux for 1 h. The excess thionyl chloride and
benzene were distilled away, and then benzene (40 mL) was added.
While cooling in an ice bath, AlCl3 (3.4 g, 25 mmol) was added slowly.
The ice bath was then removed and the temperature was maintained at
50 °C overnight. The resulting mixture was poured onto a mixture of
ice (15 g) and concentrated HCl (5 mL), and after an aqueous workup,
concentration gave crude 1,3-bis(benzoylmethyl)benzene (1.5 g). This
material was heated at reflux overnight with a mixture of 48% HBr (5
mL) and DMSO (80 mL). Concentrated HCl (20 mL) and water (200
mL) were added, and the mixture was heated for another 1.5 h. After
cooling to room temperature, the mixture was extracted three times
with ether, and the combined organics were then washed four times
with water in order to remove residual DMSO. After drying over Na2-
SO4, concentration gave tetraketone 4a as a light yellow solid (1.47 g,
4.3 mmol, 43%), mp 84-86 °C (lit.26 mp 98-99.5 °C): 1H NMR
(CDCl3) δ 7.54 (t, J ) 8 Hz, 4 H), 7.69 (m, 3 H), 7.99 (dd, J ) 8 Hz,
1.5 Hz, 4 H), 8.25 (dd, J ) 8 Hz, 1.5 Hz, 2 H), 8.60 (t, J ) 1.5 Hz,
1 H); 13C NMR (CDCl3) δ 129.3, 130.0, 130.2, 130.9, 132.7, 133.9,
135.4, 135.7, 192.8, 193.5 (10 of 10 expected resonances observed);
MS, m/z 342 (M+, 2), 324 (M - H2O, 4), 237 (M - C6H5CO, 20),
105 (C6H5CO+, 100).
5-Bromo-1,3-bis(phenylglyoxalyl)benzene (4b). Diacid 3b (1.5 g,
5.5 mmol) was carried through the procedure described for the
preparation of 4a, to give compound 4b (1.73 g, 4.1 mmol, 75%). An
analytical sample was obtained by recrystallization from CHCl3-
ethanol, mp 128-129.5 °C: 1H NMR (CDCl3) δ 7.55 (t, J ) 8 Hz, 4
H), 7.71 (t, J ) 8 Hz, 2 H), 7.98 (dd, J ) 8 Hz, 1.5 Hz, 4 H), 8.37 (d,
J ) 1.5 Hz, 2 H), 8.50 (t, J ) 1.5 Hz, 1 H); 13C NMR (CDCl3) δ
124.0, 129.1, 130.1, 132.2, 135.2, 135.4, 137.9, 190.9, 192.3 (9 of 10
expected resonances observed); MS, m/z 422 (M+ [81Br], 0.1), 420 (M+
[79Br], 0.1), 317 (M - C6H5CO, 4), 315 (M - C6H5CO, 4), 105 (C6H5-
CO+, 100); exact mass 419.9988 (calcd for C22H1379BrO4 419.9998.
1,3-Bis(3-oxo-2,4,5-triphenylcyclopenta-1,4-dienyl)benzene (5a).
A 40% solution of Triton B in methanol (0.23 mL, 0.5 mmol) was
added to a solution of 1,3-diphenylacetone (0.55 g, 2.5 mmol) and
compound 4a (0.35 g, 1.0 mmol) in refluxing butanol (20 mL), and
the heating was continued for 1 h. At this time another equal portion
of 40% Triton B solution was added, and heating was continued for
an additional 2.5 h. After cooling in an ice bath, a dark solid was
collected by filtration and washed with cold ethanol, to leave deep
purple 5a (0.40 g, 0.58 mmol, 58%), mp 308-311 °C: 1H NMR
(CDCl3) δ 6.63 (s, 1 H), 6.81 (t, J ) 7.5 Hz, 6 H), 7.02-7.27 (m, 27
H); 13C NMR (CDCl3) δ 125.2, 125.8, 127.7, 127.9, 128.1, 128.2, 128.4,
128.8, 129.1, 129.2, 129.3, 129.4, 130.2, 130.4, 130.5, 130.7, 133.0,
134.2, 154.0, 154.3, 200.3 (21 of 21 expected resonances observed);
MS, m/z 690 (M+, 70), 662 (M - CO, 30), 456 (33), 278 (100); exact
mass 690.2567, calcd for C52H34O2 690.2559.
h. After cooling, the mixture was slurried with acetone, and the
resulting crystalline solid was collected to give pure compound 6 (76
mg, 0.077 mmol, 86%) which exhibited a single component upon TLC
analysis (Rf 0.46, silica gel GF, 9:1 hexanes-ethyl acetate), mp > 350
°C: 1H NMR (CDCl3) δ 6.15 (d, J ) 7 Hz, 4 H), 6.26 (m, 3 H), 6.67-
7.00 (m, 47 H); FAB MS, m/z 991 (M+ [13C1], 100), 938 (6), 926 (10),
914 (6), 898 (8), 854 (13), 837 (9), 807 (20).
1,3-Bis(heptaphenyl-2-naphthyl)benzene (7a). Biscyclopentadi-
enone 5a (0.15 g, 0.22 mmol) in 1,2-dichloroethane (5 mL) was heated
to reflux under argon. A solution of isoamyl nitrite (0.1 mL) in 1,2-
dichloroethane (7 mL) was added, followed by the addition of 3,4,5,6-
tetraphenylanthranilic acid2 (0.20 g, 0.45 mmol) in 1,2-dichloroethane
(7 mL) over 0.5 h. Heating was continued for 3 h; then ethanol (4
mL) and 1 N NaOH (6 mL) were added to terminate the reaction.
Chloroform was added, and after an aqueous workup, the resulting
material was chromatographed on a column of silica gel (solvent, 98:2
hexanes-ethyl acetate), and the fractions containing 7a, which exhibited
Rf 0.12 upon TLC analysis (silica gel GF, same solvent), were combined
and concentrated to give pure 7a as a colorless solid (37 mg, 0.026
mmol, 12%), mp 237-239 °C: 1H NMR (CDCl3) δ 4.99 (d, J ) 7
Hz, 1 H), 5.82 (m, 2 H), 5.91 (m, 3 H), 6.15 (d, J ) 7 Hz, 2 H), 6.23
(m, 3 H), 6.32-6.92 (m, 63 H); 13C NMR (CDCl3) >70 resonances
observed; 42 are expected in the fast exchange limit of all single-bond
rotations; FAB MS, m/z 1395 (M+ [13C1], 100), 1318 (M - C6H5, 10).
5-Bromo-1,3-bis(heptaphenyl-2-naphthyl)benzene (7b). Biscy-
clopentadienone 5b (0.13 g, 0.17 mmol) was carried through the
procedure described for the preparation of 7a, to give compound 7b as
a colorless solid (25 mg, 0.017 mmol, 10%), mp 170-173 °C: 1H
NMR (CDCl3) δ 5.02 (br s, 1 H), 5.83 (br s, 1 H), 5.86 (br s, 1 H),
5.96 (d, J ) 1.5 Hz, 1 H), 6.01 (d, J ) 1.5 Hz, 1 H), 6.13-6.98 (m,
68 H); 13C NMR (CDCl3) >80 resonances observed; 42 are expected
in the fast exchange limit; FAB MS, m/z 1475 (M+ [13C181Br], 88),
1396 (19), 1126 (24), 917 (31), 776 (63), 307 (100).
1,3-Bis(3-phenyl-2-oxopropyl)benzene (10a). Ether (40 mL) and
n-butyllithium (8.0 mL, 2.5 M solution in hexanes) were added under
argon to a 3-necked round-bottomed flask and cooled in an ice bath.
Diisoproplamine (2.1 mL) was added, and the solution was stirred for
30 min. At this time, methyl phenylacetate (2.0 mL) was added
dropwise, and the mixture was stirred for 15 min. Dimethyl 1,3-
benzenediacetate (1.47 g, 6.6 mmol) was added dropwise, and
precipitation was immediately observed. The ice bath was removed,
and the mixture was stirred overnight. The reaction mixture was poured
into 1 N HCl (100 mL), and the layers were separated. The aqueous
layer was extracted twice with ether, and the combined organic layers
were washed with water and dried over Na
2SO4. The solvent was
removed, and the residue was heated with 6 N HCl (15 mL) and acetic
acid (100 mL) at reflux for 7 h. The solvent was removed again and
the residue extracted with ether, and after an aqueous workup, the
material was chromatographed on silica gel (solvent, 9:1 hexanes-
ethyl acetate). The fractions containing the desired product, which
exhibits Rf 0.11 upon TLC analysis (same solvent), were combined
and concentrated to give pure 10a (285 mg, 0.83 mmol, 13% yield),
mp 54-55 °C: 1H NMR (CDCl3) δ 3.55 (s, 4 H), 3.57 (s, 4 H), 6.78
(t, J ) 1.5 Hz, 1 H), 6.91 (dd, J ) 7.5 Hz, 1.5 Hz, 2 H), 7.01 (dd, J
) 7.5 Hz, 2 Hz, 4 H), 7.08-7.21 (m, 7 H); 13C NMR (CDCl3) δ 48.8,
49.2, 127.0, 128.2, 128.7, 128.9, 129.5, 130.7, 133.9, 134.4, 205.4 (11
of 11 expected resonances observed); MS, m/z 342 (M+, 19), 223 (M
- C6H5CH2O, 24), 91 (C6H5CH2+, 100); exact mass 342.1621, calcd
for C24H22O2 342.1620.
1,3-Bis[3-(4-bromophenyl)-2-oxopropyl]benzene (10b). Methyl
p-bromophenylacetate (9b, 2.5 g, 11 mmol) was carried through the
same procedure described for the preparation of 10a to give compound
10b (527 mg, 1.05 mmol, 9%), mp 129.5-131 °C: 1H NMR (CDCl3)
δ 3.67 (s, 4 H), 3.69 (s, 4 H), 6.90 (s, 1 H), 6.99 (d, J ) 8 Hz, 4 H),
7.05 (d, J ) 8 Hz, 2 H), 7.25 (t, J ) 8 Hz, 1 H), 7.42 (d, J ) 8 Hz,
4 H); 13C NMR (CDCl3) δ 48.6, 49.3, 48.6, 121.4, 129.4, 130.9, 131.4,
132.0, 133.0, 134.4, 204.8 (11 of 11 expected peaks observed); MS,
m/z 500 (M+ [79Br81Br], 19), 303 and 301 (M - BrC66H44CH22CO,
95), 169 (100); exact mass 499.9817, calcd for C24H2079Br81BrO2
499.9810.
5-Bromo-1,3-bis(3-oxo-2,4,5-triphenylcyclopenta-1,4-dienyl)ben-
zene (5b). Tetraketone 4b (0.52 g, 1.24 mmol) was carried through
the procedure described for the preparation of 5a, to give deep purple
5b (0.14g, 0.18 mmol 15%), mp 125.5-128 °C: 1H NMR (CDCl3) δ
6.5-7.5 (m); MS, m/z 770 (M+ [81Br], 100) 742 (M - CO, 33), 178
(75).
1,3-Bis(pentaphenylphenyl)benzene (6). Compound 5a (63 mg,
0.09 mmol) and diphenylacetylene (380 mg, 2.1 mmol) were mixed in
a screw-capped Pyrex tube and heated to 300 °C in a metal bath for 2
1,3-Bis(2-oxo-3,4,5-triphenylcyclopenta-1,4-dienyl)benzene (11a).
A 40% solution of Triton B in methanol (0.4 mL, 0.8 mmol) was added
(26) Wrasidlo, W.; Augl, J. M. J. Polym. Sci. A 1969, 1, 3393-3405.

7300 J. Am. Chem. Soc., Vol. 119, No. 31, 1997
Tong et al.
to a solution of compound 10a (0.35 g, 1.0 mmol) and benzil (0.53 g,
2.5 mmol) in boiling butanol (15 mL), and the solution was heated at
reflux for 1 h. At this time another equal portion of 40% Triton B
was added, and heating was continued for an additional 2.5 h. The
reaction mixture was then concentrated to leave a purple residue. This
material was chromatographed on silica gel (solvent, 95:5 hexanes-
ethyl acetate) to give a purple solid which proved to be the monocy-
clopentadienone. This material (125 mg) was heated with benzil (41
mg, 0.2 mmol) in ethanol (10 mL) at reflux. A solution of 10% KOH
in ethanol (0.25 mL) was added, and the solution was heated at reflux
for 3 h. After cooling in the freezer, compound 11a precipitated as
deep purple crystals (92 mg, 0.13 mmol 13% yield), mp 284-286 °C:
1H NMR (CDCl3) δ 6.80 (m, 7 H), 6.94-7.12 (m, 27 H); 13C NMR
(CDCl3) δ 125.40, 125.43, 127.5, 128.1, 128.6, 129.4, 129.5, 130.3,
130.86, 130.93, 133.1, 133.3, 154.5, 154.7, 200.2 (15 of 21 expected
peaks observed); FAB MS, m/z 690 (M+, 28), 307 (100).
mp 101-103 °C: 1H NMR (CDCl3) δ 4.23 (s, 6 H), 7.08 (s, 3 H),
7.34-7.44 (m, 6 H), 7.53 (m, 3 H), 7.95 (d, J ) 8 Hz, 1.5 Hz, 6 H);
13C NMR (CDCl3) δ 45.5, 128.7, 128.8, 129.5, 133.3, 135.4, 136.7,
197.5 (8 of 8 expected resonances observed); MS, m/z 432 (M+, 11),
313 (M - C6H5COCH2, 3), 105 (C6H5CO+, 100); exact mass 432.1716,
calcd for C30H24O3 432.1725.
1,3,5-Tris(phenylglyoxalyl)benzene (17). Compound 16 (2.0 g, 4.6
mmol) was mixed with 48% HBr (7 mL) and DMSO (120 mL) and
heat overnight at reflux. Concentrated HCl (36 mL) and water (200
mL) were added, and the mixture was heated for another 1.5 h. After
cooling to room temperature, the mixture was extracted with CHCl3
three times, and the combined organics were then washed four times
with water in order to remove residual DMSO. After drying over Na2-
SO4, concentration gave crude hexaketone 17 as a brown solid (1.74
g), but one recrystallization from CHCl3-benzene gave pure 17 (1.25
g, 2.64 mmol, 57%), mp 188-190 °C: 1H NMR (CDCl3) δ 7.55 (t, J
) 8 Hz, 6 H), 7.71 (tt, J ) 8 Hz, 1 Hz, 3 H), 8.00 (dd, J ) 8 Hz, 1 Hz,
6 H), 8.84 (s, 3 H); 13C NMR (CDCl3) δ 129.4, 130.4, 132.4, 134.8,
135.7, 191.2, 192.4 (7 of 8 expected resonances observed); MS, m/z
474 (M+, 4), 456 (M - CO, 5), 369 (M - C6H5CO, 100); exact mass
474.1105, calcd for C30H18O6 474.1103.
1,3,5-Tris(3-oxo-2,4,5-triphenylcyclopenta-1,4-dienyl)benzene (18).
A 40% solution of Triton B in methanol (0.4 mL, 0.8 mmol) was added
to a solution of 1,3-diphenylacetone (0.92 g, 4.0 mmol) and compound
17 (0.50 g, 1.05 mmol) in boiling butanol (20 mL), and the solution
was heated at reflux for 1 h. At this time another equal portion of
40% Triton B was added, and heating was continued for an additional
2 h. After cooling in an ice bath, a dark solid was collected by filtration
and washed with cold ethanol, leaving deep purple compound 18 (0.81
g, 0.81 mmol, 77%), mp 300-303 °C: 1H NMR (CDCl3) δ 6.46 (s, 3
H), 6.72 (m, 6 H), 7.03-7.38 (m, 39 H); 13C NMR (CDCl3) δ 125.3,
126.2, 127.8, 128.09, 128.15, 128.18, 128.4, 128.9, 129.1, 129.3, 130.2,
130.3, 130.5, 130.6, 133.0, 134.8, 153.0, 154.0, 199.9 (19 of 19 expected
resonances observed); FAB MS, m/z 998 (M+ [13C1], 100), 575 (36).
1,3,5-Tris(pentaphenylphenyl)benzene (19). Compound 18 (100
mg, 0.10 mmol) and diphenylacetylene (636 mg, 3.57 mmol) were
mixed in a screw-capped Pyrex tube and heated to 300 °C in a metal
bath for 2 h. After cooling, the mixture was taken up in acetone, and
crude 19 precipitated as light green crystals (67 mg). This material
was subjected to preparative TLC (silica gel GF, 9:1 hexanes-ethyl
acetate) to give pure compound 19 (51 mg, 0.035 mmol, 35% yield)
as a pale yellow solid, mp 242-244 °C: 1H NMR (CDCl3) δ 5.68 (d,
J ) 7 Hz, 6 H), 6.25 (s, 3 H), 6.43 (m, 6 H), 6.63-6.76 (m, 45 H),
6.88 (m, 6 H), 7.01-7.09 (m, 12 H); 13C NMR (CDCl3) δ 124.6, 124.7,
124.8, 126.1, 126.2, 127.3, 131.1, 131.2, 131.5, 131.8, 134.2, 136.6,
137.5, 139.3, 139.9, 140.5, 140.7, 140.9 (18 of 18 expected resonances
observed); FAB MS, m/z 1448 (M+ [13C2], 45), 1314 (20), 896 (22),
850 (32).
Bis(4-methoxyphenyl)acetylene (20). 4,4′-Dimethoxybenzil (513
mg, 1.9 mmol) and trimethyl phosphite were placed in a screw-capped
Pyrex tube, and this material was heated at 215 °C overnight under
argon. After cooling to room temperature, a light yellow precipitate
formed which was collected by suction filtration. Recrystallization of
this material from ethanol gave pure 20 as light yellow needles (202
mg, 0.85 mmol, 45% yield), mp 133-135 °C (lit.27 mp 140-146 °C);
1H NMR (CDCl3) δ 3.82 (s, 6 H), 6.87 and 7.45 (AA′BB′ system, 8
H); 13C NMR (CDCl3) δ 55.8, 88.1, 114.1, 115.8, 133.0, 159.5 (6 of 6
expected resonances observed).
1,3,5-Tris{[3,4-bis(4-methoxyphenyl)-2,5,6-triphenyl]phenyl}-
benzene (21). Triscyclopentadienone 18 (90 mg, 0.090 mmol) and
compound 20 were mixed in a screw-capped Pyrex tube and heated to
300 °C in a metal bath for 2 h. After cooling, the mixture was subjected
to silica gel column chromatography (4:1 hexanes-ethyl acetate, then
2:1 hexanes-ethyl acetate), and the fractions containing compound 21
were combined and crystallized from CHCl3-methanol to give pure
21 (67 mg, 0.041 mmol, 46% yield), mp 224.5-227.5 °C: 1H NMR
(CDCl3) δ 3.56 (3 close s’s, 9 H), 3.59 (2 close s’s, 9 H), 5.67 (m, 5
H), 6.20 (m, 3 H), 6.30 (m, 13 H), 6.40 (m, 3 H), 6.55 (m, 9 H), 6.70
(m, 19 H), 6.88 (m, 7 H), 7.03 (m, 13 H); FAB MS, m/z 1630 (M +
2H [13C2], 100).
1,3-Bis[2-oxo-3-(4-bromophenyl)-4,5-diphenylcyclopenta-1,4-di-
enyl]benzene (11b). A 40% solution of Triton B in methanol (0.3
mL, 0.6 mmol) was added to a solution of compound 10b (0.30 g, 0.6
mmol) and benzil (0.31 g, 1.5 mmol) in boiling butanol (10 mL), and
the solution was heated at reflux for 1 h. At this time another equal
portion of 40% Triton B was added, and heating was continued for an
additional 2 h. After cooling in an ice bath, a dark solid was collected
by filtration and washed with cold ethanol, leaving deep purple
compound 11b (0.35 g, 0.41 mmol, 69%), mp 273.0-276.5 °C: 1H
NMR (CDCl3) δ 6.87 (t, J ) 7 Hz, 8 H), 7.01-7.35 (m, 24 H); 13C
NMR (CDCl3) δ 122.0, 124.3, 125.6, 128.0, 128.2, 128.4, 128.8, 128.9,
129.0, 129.3, 129.5, 130.0, 130.8, 131.4, 131.9, 132.2, 132.9, 133.1,
154.8, 155.2, 199.8 (21 of 21 expected peaks observed); MS, m/z 848
(M+ [79Br81Br], 97), 820 (M - CO, 15), 358 (20), 356 (21), 278 (54),
98 (100).
1,3-Bis(heptaphenyl-1-naphthyl)benzene (12a). Biscyclopentadi-
enone 11a (92 mg, 0.13 mmol) in 1,2-dichloroethane (5 mL) was heated
to reflux under argon. A solution of isoamyl nitrite (0.07 mL) in 1,2-
dichloroethane (5 mL) was added, followed by the slow addition of
3,4,5,6-tetraphenylanthranilic acid (8, 132 mg, 0.30 mmol) in 1,2-
dichloroethane (5 mL) over 0.5 h, and heating was continued for 8 h.
The treatment with isoamyl nitrite and 8 was then repeated, and heating
was continued overnight. Ethanol (4 mL) and 1 N NaOH (6 mL) were
added to terminate the reaction. CHCl3 was then added, and after an
aqueous workup, the resulting material was chromatographed on a
column of silica gel (solvent, 95:5 hexanes-ethyl acetate). The
fractions containing the desired product, which exhibited Rf 0.14 upon
TLC analysis (silica gel GF, 2.5% ethyl acetate in hexanes), were
combined and concentrated to give pure 12a as a pale yellow solid (9
mg, 0.006 mmol, 5%); mp >350 °C (darkens at 293 °C). 1H NMR
(CDCl3) δ 4.97 (br s, 1 H), 5.28 (br s, 1 H), 5.38 (t, J ) 8 Hz, 1 H),
5.80 (d, J ) 8 Hz, 2 H), 6.11 (br s, 2 H), 6.22 (s, 1 H), 6.32 (m, 4 H),
6.41 (br s, 1 H), 6.51-7.25 (m, 62 H). FAB MS, m/z 1395 (M+ [13C1],
100), 1318 (M + H - C6H5, 10), 659 ([M + H - C6H5]2+, 27).
1,3-Bis[2,3,5,6,7,8-hexaphenyl-4-(4-bromophenyl)-1-naphthyl]-
benzene (12b). Biscyclopentadienone 11b (250 mg, 0.29 mmol) was
carried through the same procedure used for the preparation of 12a, to
give compound 12b as a pale yellow solid (63 mg, 0.041 mmol, 14%),
mp >350 °C: 1H NMR (CDCl3) δ 4.91 (br s, 1 H), 5.27 (br s, 1 H),
5.37 (t, J ) 8 Hz, 1 H), 5.80 (dd, J ) 8 Hz, 1 Hz, 2 H), 6.09 (br d, J
) 6 Hz, 2 H), 6.19 (s, 1 H), 6.21-7.17 (m, 64 H); 13C NMR (CDCl3)
δ 119.1, 125.0-134.5 (broad, overlapping resonances), 135.8, 137.0,
137.4, 138.0, 139.1, 139.2, 139.91, 139.98, 140.03, 141.2, 140.4, 140.6,
81
140.7, 140.8, 141.3, 141.9, 142.0; FAB MS, m/z 1553 (M+ [13C1
Br79Br], 100).
-
1,3,5-Tris(benzoylmethyl)benzene (16). 1,3,5-Benzenetriacetic
acid12 (15, 1.5 g, 5.95 mmol), thionyl chloride (5 mL), and benzene
(15 mL) were heated to reflux for 1.5 h. The excess thionyl chloride
and benzene were distilled away, and then benzene (40 mL) was added.
While cooling in an ice bath, AlCl3 (3.4 g, 25 mmol) was added slowly.
The ice bath was then removed, and the temperature was maintained
at 50 °C overnight. The resulting mixture was poured onto a mixture
of ice (20 g) and concentrated HCl (5 mL), and after an aqueous
workup, concentration to dryness and recrystallization of the residue
from CHCl3 gave light yellow crystals of 16 (2.14 g, 4.95 mmol, 83%),
(27) Newman, M. S.; Reid, D. E. J. Org. Chem. 1958, 23, 665-666.

The Albatrossenes
J. Am. Chem. Soc., Vol. 119, No. 31, 1997 7301
1,3,5-Tris(heptaphenyl-2-naphthyl)benzene (22). Triscyclopen-
tadienone 18 (0.15 g, 0.15 mmol) in 1,2-dichloroethane (5 mL) was
heated to gentle reflux under argon. A solution of isoamyl nitrite (0.12
mL) in 1,2-dichloroethane (7 mL) was added, followed by the slow
addition of 3,4,5,6-tetraphenylanthranilic acid (0.22 g, 0.50 mmol) in
1,2-dichloroethane (7 mL) over 0.5 h. Heating was continued for 8 h,
and then ethanol (4 mL) and 1 N NaOH (6 mL) were added to terminate
the reaction. CHCl3 was added, and after an aqueous workup, the
resulting material was chromatographed on a column of silica gel
(solvent, 97:3 hexanes-ethyl acetate). Fractions containing the desired
product, which gives Rf 0.45 on TLC (silica gel GF, 9:1 hexanes-
ethyl acetate), were combined and concentrated to give pure 22 as a
pale yellow solid (46 mg, 0.022 mmol, 15%), mp >350 °C: 1H NMR
(CDCl3) δ 4.99 (t, J ) 1 Hz, 1 H), 5.25 (s, 1 H), 5.27 (t, J ) 1 Hz, 1
H), 5.49 (m, 4 H), 5.86 (m, 10 H), 6.21-6.97 (m, 91 H); 13C NMR
(CDCl3) >70 resonances observed; 40 are expected in the fast exchange
limit; FAB MS, m/z 2054 (M+ [13C2], 100), 1977 (M - C6H5, 5), 1171
(28).
account for the solvent electron density. A total electron count of 141.6
e in a total volume of 495.8 Å3 was found in a single-potential solvent
area, consistent with the presence of two nitrobenzenes (64 e) per unit
cell (141.6 e/2 ) 70.8 e) and thus a 1:1 5b:C6H5NO2 ratio. The
SQUEEZE-processed data were used for all subsequent cycles of
refinement. Due to the weak data, only the bromine atom was refined
anisotropically; carbon atoms were refined isotropically and H atoms
were included with a riding model [C-H ) 0.93 Å, U(H) ) 1.2U(C)].
The refinement converged to R(F) ) 0.079, wR(F2) ) 0.133, and S )
1.09 for 2819 reflections with F > 4σ(F), and R(F) ) 0.241, wR(F2)
) 0.189, and S ) 0.68 for 11,263 unique reflections (three reflections
were suppressed) and 450 variables. Full details are given in the
Supporting Information.
X-ray Crystallographic Analysis of Compound 12a. A colorless
prism of 12a, 0.04 mm × 0.12 mm × 0.22 mm in size and mounted
on a glass fiber, was used for X-ray measurements. Crystal data:
C110H74; triclinic, space group P1h; a ) 12.43 (1) Å, b ) 25.51 (2) Å,
c ) 27.74 (5) Å, R ) 114.89 (7)°, â ) 100.08 (9)°, γ ) 90.61 (8)°, V
) 7820 (16) Å3, Z ) 4, Dcalcd ) 1.185 g/cm3. Intensity data were
collected and processed in a manner similar to that described for
compound 7b, crystal A, yielding a total of 11 880 unique reflections.
The structure was solved by direct methods (CRUNCH9) and refined
by full-matrix least-squares on F2 (SHELXL-93). All non-H atoms
were refined with anisotropic displacement coefficients, and H atoms
were included with a riding model [C-H ) 0.93 Å, U(H) ) 1.2U(C)].
The refinement converged to R(F) ) 0.065, wR(F2) ) 0.183, and S )
1.08 for 10 821 reflections with F > 4σ(F), and R(F) ) 0.070, wR(F2)
) 0.194, and S ) 1.05 for 11 875 unique reflections (five reflections
were suppressed) and 1981 variables. Full details are given in the
Supporting Information.
X-ray Crystallographic Analysis of Compound 7b. Crystal A.
A colorless needle was cut to 0.08 mm × 0.09 mm × 0.18 mm and
sealed in a glass capillary for use in X-ray measurements. Crystal
data: C110H73Br‚3.5C6H5NO2‚C6H5CH3; triclinic, space group P1h; a
) 16.95 (3) Å, b ) 18.22 (3) Å, c ) 20.32 (2) Å, R ) 79.39 (8)°, â
) 72.11 (10)°, γ ) 66.73 (11)°, V ) 5472 (12) Å3, Z ) 2, Dcalcd
)
1.212 g/cm3. Intensity data was collected out to 2θ ) 93° by using
Cu KR radiation (λ ) 1.541 84 Å) at 298 K on a Rigaku R-AXIS IIC
image plate system equipped with a rotating anode and double-focusing
mirrors. A total of 175 frames of data were collected with 5° of
oscillation per frame. A total of 32 618 reflections were indexed,
integrated, and corrected for Lorentz and polarization effects (using
the program DENZO28,29 ), and then the reflections were scaled and
merged (SCALEPACK29) to give 9015 unique reflections (Rint ) 0.060).
The structure was solved by heavy atom methods (SHELXTL30 ) and
refined by full-matrix least-squares on F2 (SHELXL-9331 ). For the
molecule of 7b, all non-H atoms were refined with anisotropic
displacement coefficients, and H atoms were included with a riding
model [U(H) ) 1.2U(C)]. For the solvent molecules (5 sites: 1, 1
NB; 2, 1 NB; 3, 0.5 NB; 4, 0.5/0.5 NB/toluene; 5 0.5/0.5 NB/toluene),
all non-H atoms were assigned isotropic displacement coefficients, H
atoms were included with a riding model [U(H) ) 1.2U(C) or
1.5U(Cmethyl)], rings were modeled as regular hexagons and refined as
rigid groups, and nitro groups were included with restraints but were
free to rotate. The refinement converged to R(F) ) 0.115, wR(F2) )
0.308, and S ) 1.08 for 7509 reflections with F > 4σ(F), and R(F) )
0.125, wR(F2) ) 0.326, and S ) 1.04 for 9015 unique reflections, 1135
variables, and 11 restraints. Full details are given in the Supporting
Information.
X-ray Crystallographic Analysis of Compound 12b. A colorless
prism of 12b, 0.18 mm × 0.22 mm × 0.25 mm in size, was sealed in
a glass capillary for use in X-ray measurements. Crystal data: C110H72-
Br2‚2CH2Cl2‚4CH3OH; trigonal, space group P3221; a ) b ) 17.493
(2) Å, c ) 26.676 (4) Å, V ) 7069 (2) Å3, Z ) 3, Dcalcd ) 1.305
g/cm3. Intensity measurements were made with 3° e 2θ e 45° using
Mo KR radiation at 298 K on a Siemens P4 diffractometer. A total of
6594 reflections were measured, of which 6090 were unique (Rint
)
0.067). The structure was solved by heavy atom methods (SHELXTL)
and refined by full-matrix least-squares on F2 (SHELXL-93). The
presence of several disordered solvent molecules was evident, and the
refinement of a discrete-atom solvent model converged to R(F) ) 0.077,
but it was abandoned in favor of a solvent-free model due to uncertainty
in the identity of the included solvent. The SQUEEZE/BYPASS
procedure described above was again employed to account for the
solvent electron density, and the SQUEEZE-processed data were used
for all subsequent cycles of refinement. Due to the weak data, only
the bromine was refined anisotropically; carbon atoms were refined
isotropically, and H atoms were included with a riding model [C-H
) 0.93 Å, U(H) ) 1.2U(C)]. The refinement converged to R(F) )
0.055, wR(F2) ) 0.092, and S ) 1.01 for 1555 reflections with F >
4σ(F), and R(F) ) 0.176, wR(F2) ) 0.116, and S ) 0.61 for 6090
unique reflections and 230 variables. Full details are given in the
Supporting Information.
Crystal B. A colorless needle, cut to 0.05 mm × 0.05 mm × 0.22
mm and mounted on a glass fiber, was used for X-ray measurements.
Crystal data: C110H73Br‚C6H5NO2; triclinic, space group P1h; a ) 16.922
(2) Å, b ) 13.627 (2) Å, c ) 20.291 (2) Å, R ) 77.741 (9)°, â )
72.291 (8)°, γ ) 85.612 (8)°, V ) 4355 (1) Å3, Z ) 2, Dcalcd ) 1.218
g/cm3. Intensity measurements were made with 3° e 2θ e 45° using
graphite-monochromated Mo KR radiation (λ ) 0.710 73 Å) at 230 K
on a Siemens P4 diffractometer. A total of 11 843 reflections were
measured, of which 11,266 were unique (Rint ) 0.071). The structure
was solved by heavy atom methods (SHELXTL) and refined by full-
matrix least-squares on F2 (SHELXL-93). The presence of a disordered
solvent molecule in the lattice was evident, but this electron density
was not well modeled with discrete atoms. Therefore, the SQUEEZE/
BYPASS32 procedure implemented in PLATON-9633 was used to
X-ray Structure of Compound 19. A colorless plate of 19, 0.25
mm × 0.38 mm × 0.50 mm in size, was sealed in a glass capillary for
use in X-ray measurements. Crystal data: C114H78‚2.5CHCl3; mono-
clinic, space group P21/c; a ) 24.80 (3) Å, b ) 27.15 (2) Å, c )
33.26 (4) Å, â ) 98.77 (11)°, V ) 22,133 (44) Å3, Z ) 8, Dcalcd
)
1.048 g/cm3. Intensity data were collected and processed in a manner
similar to that described for compound 7b, crystal A, yielding a total
of 18 850 unique reflections (Rint ) 0.092). The structure was solved
by molecular replacement (PATSEE34 ) and refined by full-matrix least-
squares on F2 (SHELXL-93). In addition to the two independent
molecules of 19 eight disordered chloroform sites were identified in
the asymmetric unit. Most of these exhibited partial occupancy, and
refinement led to a model with an overall formulation for the
asymmetric unit of (19)2(CHCl3)5.6. However, this model refined to
(28) Otwinowski, Z. Oscillation Data Reduction Program. In Proceedings
of the CCP4 Study Weekend, Data Collection and Processing; compiled
by Sawyer, L., Isaacs, N., Bailey, S.; SERC Daresbury Laboratory: England,
1993; pp 56-62.
(29) Minor, W. XDISPLAYF Program. Purdue University, USA, 1993.
(30) Sheldrick, G. M. SHELXTL, Version 4.2. Siemens Analytical X-ray
Instruments, Madison, WI, 1991.
(31) Sheldrick, G. M. SHELXL-93. Program for the Refinement of
Crystal Structures. University of Gottingen, Germany, 1993.
(32) Van der Sluis, P.; Spek, A. L. Acta Crystallogr., Sect. A 1990, 46,
194-201.
(34) Egert, E.; Sheldrick, G. M. Acta Crystallogr., Sect. A 1985, 41, 262-
268.
(33) Spek, A. L. Acta Crystallogr., Sect. A 1990, 46, C34.

7302 J. Am. Chem. Soc., Vol. 119, No. 31, 1997
Tong et al.
only R(F) ) 0.190, so the SQUEEZE/BYPASS procedure implemented
in PLATON-96 was once again employed to account for the solvent
electron density. A total electron count of 1041 e in a volume of 7870
Å3 was found for a single large solvent region in the unit cell (35.6%
of the cell volume). This electron count corresponds to 18 CHCl3’s,
which would yield a formulation for the asymmetric unit of (19)2-
Acknowledgment. This work was supported by NSF Grant
CHE-9408295 (to R.A.P.) and NIH Grant GM44038 (to C.E.S.).
We thank Prof. F. M. Hughson for use of an X-ray diffracto-
meter.
(CHCl3)4.5
. [(19)2(CHCl3)5 ) C114H78‚2.5CHCl3 has been used for
Supporting Information Available: Crystal structure re-
ports for compounds 7b, 12a, 12b, and 19, which include full
experimental details, tables of atomic coordinates, bond dis-
tances, bond angles, and thermal parameters, and selected figures
(162 pages). See any current masthead page for ordering and
Internet access instructions.
calculation of the density of the crystal.] SQUEEZE-processed data
were employed for all subsequent refinements. The carbon atoms were
refined anisotropically with hydrogens riding [C-H ) 0.93 Å, U(H)
) 1.2U(C)], and the refinement converged to R(F) ) 0.116, wR(F2)
) 0.298, and S ) 1.10 for 14,059 reflections with F > 4σ(F), and
R(F) ) 0.127, wR(F2) ) 0.319, and S ) 1.00 for 18,850 unique
reflections and 2054 parameters. Full details are given in the Supporting
Information.
JA971261K