New Molecular Devices: In Search of a Molecular Ratchet

Article abstract of DOI:10.1021/jo9723218

The triptycene-substituted [3]- and [4]helicenes 1 and 2 were examined as possible molecular versions of mechanical ratchets, where the triptycene serves as the ratchet wheel and the hehcenes as pawl and spring. The syntheses of 1 and 2b are described. ¹H NMR was employed to examine rotation around the triptycene/helicene single bond; at 20 °C rotation is frozen for both 1 and 2b but the NMR of 1 revealed a plane of symmetry, indicating that 1 cannot function as a unidirectional ratchet In contrast, NMR revealed that, like a ratchet, triptycvl[4]helicene 2b lacks the symmetry of 1 and has a barrier to rotation of 24.5 kcal/mol, but spin polarization transfer NMR expenments indicated the triptycene in 2b nonetheless rotates equally in both directions. That outcome is rationalized from the standpoint of thermodynamics.

Full text of DOI:10.1021/jo9723218

J . Org. Chem. 1998, 63, 3655-3665
3655
New Molecu la r Devices: In Sea r ch of a Molecu la r Ra tch et
T. Ross Kelly,* J ose´ Pe´rez Sestelo,† and Imanol Tellitu‡
Department of Chemistry, E. F. Merkert Chemistry Center, Boston College,
Chestnut Hill, Massachusetts 02167
Received December 29, 1997
The triptycene-substituted [3]- and [4]helicenes 1 and 2 were examined as possible molecular
versions of mechanical ratchets, where the triptycene serves as the ratchet wheel and the helicenes
as pawl and spring. The syntheses of 1 and 2b are described. 1H NMR was employed to examine
rotation around the triptycene/helicene single bond; at 20 °C rotation is frozen for both 1 and 2b,
but the NMR of 1 revealed a plane of symmetry, indicating that 1 cannot function as a unidirectional
ratchet. In contrast, NMR revealed that, like a ratchet, triptycyl[4]helicene 2b lacks the symmetry
of 1 and has a barrier to rotation of 24.5 kcal/mol, but spin polarization transfer NMR experiments
indicated the triptycene in 2b nonetheless rotates equally in both directions. That outcome is
rationalized from the standpoint of thermodynamics.
In tr od u ction
The dynamics and control of rotational motion about
single bonds has engaged the attention of chemists for
decades. Once-forefront issues such as rotation around
the carbon-carbon bond in ethane have been superseded
over time by more subtle aspects of conformational
isomerism.1 Interest in recent years has been further
fueled by conceptual linkages between macromechanical
issues of rotation and molecular variations on the same
themes, as illustrated by reports of molecular gears,2
brakes,3 and turnstiles.4 A key issue is identifying which
everyday mechanical concepts can be extrapolated to the
molecular scale.
In conjunction with a longer-term goal of devising a
molecule that functions as a motor, this laboratory has
been exploring various aspects of rotary motion at the
molecular level. To that end, we reported in 1994 work
culminating in the development of a molecular brake.3a
More recently, we described in a brief communication5
our examination of molecules that might function as
F igu r e 1. Simple mechanical ratchet: (a) a ratchet wheel;
(b) a pawl; (c) a spring that holds the pawl against the wheel.
molecular ratchets. We now present in greater detail our
efforts to construct molecular ratchets and address some
of the conceptual issues that pertain.
Commonly, ratchets are devices that allow rotary
motion in only one direction; in their simplest form they
consist of three components (Figure 1): (a) a toothed
ratchet wheel; (b) a pawl that prevents unintended
rotation of the ratchet wheel; and (c) a spring that holds
the pawl in place. In seeking to translate the macroscopic
concept of the ratchet to the molecular scale, we chose
compounds 1 and 2 as possible candidates, where the
† Current address: Departamento de Qu´ımica Fundamental
e
Industrial, Facultad Ciencias, Universidad da Corun˜a, 15071 A Co-
run˜a, Spain.
‡ Current address: Departamento de Qu´ımica Orga´nica, Univer-
sidad del Pa´ıs Vasco, 48080-Bilbao, Spain.
(1) (a) For
a leading reference see: Eliel, E. L.; Wilen, S. H.
Stereochemistry of Organic Compounds; Wiley: New York, 1994. (b)
See also: Oki, M. Top. Stereochem. 1983, 14, 1-81 (note especially
Section III).
(2) (a) Iwamura, H.; Mislow, K. Acc. Chem. Res. 1988, 21, 175-182.
(b) Mislow, K. Chemtracts: Org. Chem. 1989, 2, 151-174. (c) Naka-
mura, M.; Oki, M. Bull. Chem. Soc. J pn. 1975, 48, 2106-2111. (d) Oki,
M.; Fukuda, T.; Asakura, M.; Toyota, S. Ibid. 1996, 69, 3267-3272.
(e) Yamamoto, G. Tetrahedron 1990, 46, 2761-2772. (f) Kawada, Y.;
Sakai, H.; Oguri, M.; Koga, G. Tetrahedron Lett. 1994, 35, 139-143.
(g) Stevens, A. M.; Richards, C. J . Ibid. 1997, 38, 7805-7808. (h) Gakh,
A. A.; Sachleben, R. A.; Bryan, J . C. Chem Tech 1997, 27, 26-33.
(3) (a) Kelly, T. R.; Bowyer, M. C.; Bhaskar, K. V.; Bebbington, D.;
Garc´ıa, A.; Lang, F.; Kim, M. H.; J ette, M. P. J . Am. Chem. Soc. 1994,
116, 3657-3658. See the supplementary material appended thereto
for detailed synthetic procedures. See also: (b) Casares, J . A.; Coco,
S.; Espinet, P.; Lin, Y.-S. Organometallics 1995, 14, 3058-3067. (c)
Fanizzi, F. P.; Lanfranchi, M.; Natile, G.; Tiripicchio, A. Inorg. Chem.
1994, 33, 3331-3339.
triptycene6 acts as the wheel and the helicenes7 function
as the pawl and spring. Examination of models, par-
ticularly of 2, indicates a distinctly helical conformation
of the [4]helicene unit; that asymmetry is evident in
Figure 2, which is a stereoview of an electron density
(6) For a review of triptycenes see: Skvarchenko, V. R.; Shalaev,
V. K.; Klabunovskii, E. I. Russ. Chem. Rev. 1974, 43, 951-966.
(7) For reviews of helicene chemistry, see: (a) Martin, R. H. Angew.
Chem., Int. Ed. Engl. 1974, 13, 649-660. (b) Laarhoven, W. H.;
Prinsen, W. J . C. Top. Curr. Chem. 1984, 125, 63-130. (c) Meurer, K.
P.; Vo¨gtle, F. Ibid. 1985, 127, 1-76. Some more recent papers: (d)
Liu, L.; Katz, T. J . Tetrahedron Lett. 1990, 31, 3983-3986. (e)
Willmore, N. D.; Liu, L.; Katz, T. J . Angew. Chem., Int. Ed. Engl. 1992,
31, 1093-1094. (f) J anke, R. H.; Haufe, G.; Wu¨rthwein, E.; Borkent,
J . H. J . Am. Chem. Soc. 1996, 118, 6031-6035.
(4) Bedard, T. C.; Moore, J . S. J . Am. Chem. Soc. 1995, 117, 10662-
10671.
(5) Kelly, T. R.; Tellitu, I.; Sestelo, J . P. Angew. Chem., Int. Ed. Engl.
1997, 36, 1866-1868. For a “Highlight” featuring this paper see:
Davis, A. P. Ibid. 1998, 37, 909-910.
S0022-3263(97)02321-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/13/1998

3656 J . Org. Chem., Vol. 63, No. 11, 1998
Kelly et al.
F igu r e 2. Stereoview of an electron-density surface map of
the lowest energy conformation of 2a .
surface map of the lowest energy conformation of 2a .
Manual rotation of the triptycene in models of 2 forces
the helicene further and further out of planarity until
the energy maximum is reached. That rotation is ac-
complished much more easily with a clockwise instead
of a counterclockwise turning of the triptycene (as viewed
from the left, for the enantiomer of 2 that is shown in
Figure 2), as would be expected of a ratchet.
F igu r e 3. Calculated9a (AM1) energetics for clockwise9b
rotation around the triptycene/[4]helicene bond in 2a . Clock-
wise rotation of the triptycene in 2a corresponds to a left-to-
right progression on the x axis.
phenyl/[4]helicene bond in 3 (X ) H). Laarhoven10 has
previously determined the barrier (∆G‡) experimentally
to be 12.8 ( 0.3 kcal/mol. Our calculated barrier (∆H‡,
not ∆G‡) is 9.6 kcal/mol. Using ∆H‡ values to predict ∆G‡
values is not rigorous, but we hoped it would serve as a
useful guide. If it developed that the calculated ∆H‡
barriers for 1 and 2 also underestimated the actual
barriers (∆G‡), then the barriers might be too high to
quantitate using NMR. In that event, we could use a
pentahelicene pawl as in 4, since the calculated barrier
to rotation is only 16 kcal/mol. On the other hand, if the
actual barriers in 1 and 2 proved substantially lower than
the calculated barriers, then they could be increased by
introduction of substituents on the helicene ortho to the
triptycene, as substituents ortho to the phenyl ring (i.e.,
X * H) in analogues of 3 can increase the barrier to
rotation by several kcal/mol.10
The conformation and chirality of helicenes have
attracted the attention of many researchers over the
years.7 We chose to enlist helicenes as springs because
they have an inherent helicity and because they resist
deformation. Their resistance to being further twisted
is not insurmountable because they can be caused to
racemize, but their tendency to oppose deformation, and
to return to their original geometry, reflects the charac-
teristics of a spring. Several studies8 have shown that
the energy of activation for racemization of helicenes
depends on the size and substituents in the helicene.
P r elim in a r y Stu d ies
A key parameter leading to the decision to investigate
triptycylhelicene systems as possible ratchets was the
barrier to rotation around the triptycene/helicene bond.
We sought molecules with a barrier on the order of 18-
27 kcal/mol in order to facilitate measurement by NMR.
Molecules possessing a barrier to rotation in that range
do not rotate on the NMR time scale at room temperature
but will at temperatures accessible with a variable-
temperature NMR probe. Guided by computer modeling
of barriers to rotation, we chose 1 and 2 as possible
candidates. Calculations9 indicated that the barriers (in
terms of ∆H‡) to rotation in 1 and 2 were 27 and 22 kcal/
mol, respectively. To our minds, those calculated barriers
to rotation were invested with some credibility when the
same calculations were applied to rotation around the
Since the calculated barriers proved, in retrospect, to
be remarkably (to us) accurate in predicting the experi-
mentally observed barriers to rotation in 1-3, mention
of two additional points seems warranted. First, and not
surprisingly, the site of attachment of the triptycene to
the helicene has a major impact on the barrier to rotation,
as would be expected if the helicene is functioning as a
spring and pawl in 2. Thus, in contrast to 2, where the
calculated barrier (∆H‡) to rotation is 22 kcal/mol, the
calculated barrier for the rotation of the helicene in 5 is
only 4 kcal/mol.
(8) (a) Martin, R. H.; Marchant, M. J . Tetrahedron Lett. 1972, 3707-
3708. (b) Martin, R. H.; Marchant, M. J . Tetrahedron 1974, 30, 347-
349. (c) Scheru¨bl, H.; Fritzsche, U.; Mannschreck, A. Chem. Ber. 1984,
117, 336-343.
Second, the barrier to rotation is calculated to be higher
in 1 than in 2. It was counterintuitive to us that a bigger
pawl (as in 2) would lead to a smaller barrier. But since
experiments bear out the calculations, an explanation is
required. We suggest that the barrier to rotation in 2 is
less than in 1 because in 2 the interaction between the
one-ring-longer helicene and the blade of the triptycene
prevents the molecule from relaxing to as stable a
(9) (a) Calculations (AM1) were carried out using the coordinate
drive feature of the Spartan molecular modeling program (Wavefunc-
tion, Inc., Irvine, CA), Version 4.0, on a Silicon Graphics workstation.
(b) The same calculations (Spartan AM1 coordinate drive) give a
significantly different potential energy curve for counterclockwise (120°
f 0° dihedral angle) rotation. The calculations for counterclockwise
rotation give an energy maximum ca. 5 kcal/mol higher than for
clockwise rotation and, in contrast to clockwise rotation, are ac-
companied by inversion of the helicity of the helicene. In contrast,
experimental determinations (vide infra) indicate that helicene inver-
sion is an appreciably higher energy process7 than triptycene rotation.
Given the principle of microscopic reversibility, we attribute the
clockwise/counterclockwise difference to a limitation of the calculational
program, at least as employed by us.
(10) Laarhoven, W. H.; Peters, W. H. M.; Tinnemans, A. H. A.
Tetrahedron 1978, 34, 769-777.

In Search of a Molecular Ratchet
J . Org. Chem., Vol. 63, No. 11, 1998 3657
Sch em e 1
F igu r e 4. Space-filling models of calculated minimum (left)
and maximum energy conformations of 2a .
ground-state conformation (relative to the energy maxi-
mum) as is accessible to 1. The even lower calculated
barrier for rotation in 4 can be accounted for analogously.
Figure 3 presents a plot of the calculated energy of 2a
as a function of rotation around the triptycene/helicene
bond. The asymmetry of the peak is consistent with what
might be expected if 2a rotated unidirectionally. Figure
4 depicts space-filling models corresponding to the mini-
mum and maximum energy conformations of 2a . Note
the difference in the position of the helicene.
On the basis of the foregoing considerations, we decided
to first undertake the synthesis of the 1-triptycyl[4]-
helicenes 2, whose anticipated barrier to rotation seemed
most amenable to assessment by NMR. To study and
compare the theoretical calculations, the synthesis of the
4-triptycylphenanthrene 1 was also carried out.
Sch em e 2
Resu lts a n d Discu ssion
Syn t h et ic Con sid er a t ion s. From retrosynthetic
analysis, we envisioned the synthesis of the triptycene-
helicene systems being based on photocyclization ap-
proaches. To date, carbohelicenes have been mainly
synthesized by using stilbene photocyclizations,11 al-
though lately, alternative synthetic strategies have been
reported.7d-f Relatively few examples of the synthesis of
1-substituted helicenes have been reported, and yields
are generally low.7,10 Nevertheless, Tinnemans and
Laarhoven12 have reported the synthesis of 4-phe-
nylphenanthrene and 1-phenylbenzo[c]phenanthrene (3)-
from 1,4-diarylenynes using an enyne photocyclization.
On the basis of these stilbene and enyne photocyclization
precedents, we considered three different approaches to
the 1-triptycyl[4]helicene system (Scheme 1).
Approach A is based on the photocyclization of the
enyne 6 obtained from Wittig reaction between the ylide
of the phosphonium salt 7 and the propargylic aldehyde
8. The regioselectivity of the photocyclization should be
controlled by electronic effects.12 Approach B entails the
photocyclization of a stilbene (9) where the triptycene
unit has already been installed. Although electronic
effects should favor11 the formation of the benzo[c]-
phenanthrene over the benzo[c]anthracene isomers (13
and 14, Scheme 2), it was nonetheless likely that
predominantsif not exclusivesformation of the undes-
ired 3-substituted benzo[c]phenanthrene regiosomer (12)
instead of the desired (2a ) would occur because of the
steric bulk of the triptycene. It should11 be possible,
however, to suppress cyclization at the undesired site by
blocking it with, e.g., a methyl group as in 9 (Scheme 1),
thereby obtaining 2b, the methylated analogue of 2a . The
stilbene 9 would be formed by Wittig reaction between
the ylide derived from 10 and the triptycene-aldehyde
unit 11, available from the known13 5-bromo-o-tolualde-
hyde via Stille coupling with a 9-stannylanthracene and
subsequent benzyne addition.
Alternatively, approach C was considered in the case
of a stilbene photocyclization failure due to the steric bulk
of the triptycene. In approach C, incorporation of the
triptycene (or a precursor to it) would be achieved at the
end of the synthesis through the 1-bromobenzo[c]phenan-
threne 15. Analogous synthetic strategies were planned
for constructing 4-triptycylphenanthrene 1, although for
this compound nonphotochemical routes were also envis-
aged.
En yn ic Rou te. The synthesis began with a known3a
preparation of triptycene aldehyde 18 starting from the
commercially available anthracene-9-carboxaldehyde 17
(Scheme 3). Carbonyl protection with ethylene glycol,
triptycene generation via benzyne addition, and subse-
quent deprotection afforded the aldehyde 18 in moderate
yield. A Corey-Fuchs reaction14 gave the vinyl dibro-
mide 19, which was transformed through treatment with
n-BuLi at -78 °C and trapping of the resulting alkynyl-
lithium with methyl formate to give the desired propar-
gylic aldehyde 8 in good yield. Treatment of [(3-
(11) Mallory, F. B.; Mallory, C. W. Org. React. 1984, 30, 1-456.
(12) (a) Tinnemans, A. H. A.; Laarhoven, W. H. Tetrahedron Lett.
1973, 817-820. (b) Tinnemans, A. H. A.; Laarhoven, W. H. J . Chem.
Soc., Perkin Trans. 2 1976, 1111-1115. (c) Tinnemans, A. H. A.;
Laarhoven, W. H. Ibid. 1976, 1115-1121.
(13) Stoilov, I.; Watt, D. S.; Goodman, J . P.; Pyrek, J . St. J .
Heterocycl. Chem. 1993, 30, 1645-1651.
(14) Corey, E. J .; Fuchs, P. L. Tetrahedron Lett. 1972, 3769-3772.

3658 J . Org. Chem., Vol. 63, No. 11, 1998
Kelly et al.
Sch em e 3
Sch em e 4
phenanthryl)methyl]triphenylphosphonium bromide15 (7)
with n-BuLi and Wittig reaction of the resulting ylide
with the aldehyde 8 led in 80% yield to the formation of
the enyne 6 in the form of a 2:3 mixture of cis:trans
Sch em e 5
1
isomers as established by H NMR.
Irradiation12 of the enyne 6 in several solvents (ben-
zene, cyclohexane, methanol), under oxidative [I2 (5 to
100 mol %), O2 atmosphere] or nonoxidative (argon,
deoxygenation of the system) conditions, with a mercury
lamp (100 or 450 W) in a quartz apparatus with or
without various filters (Pyrex, Vycor) did not afford, in
any case, the desired photocyclized product. Only re-
duced (alkyne reduction mainly) or oxidized products
were isolated. Since all the enynic photocyclizations
reported10,12 use 1,4-diarylenynes, we are inclined to
attribute the failure of our reactions to the nonaromatic
triptycene unit.16
Stilben ic Rou te. After the foregoing disappointing
results, we abandoned approach A and moved to the
stilbenic route (approach B, Scheme 1). Key steps in this
route are the preparation of the aldehyde 11 and the
stilbene photocyclization (Scheme 4). Starting with
5-bromo-o-tolualdehyde (24, obtained from o-tolualdehyde
by treatment with Br2 and AlCl3),13 we incorporated the
anthracene unit using a Stille17 coupling reaction with
9-(tributylstannyl)anthracene,18 obtaining the aldehyde
23 in moderate yield. Regioselective addition of benzyne,
generated from benzenediazonium 2-carboxylate,19 af-
forded the desired triptycene aldehyde 11 in good yield.
A Wittig reaction between the ylide generated from the
naphthyl phosphonium salt 10,20 and the aldehyde 11
gave a cis/trans mixture (1:1) of stilbenes 9 in excellent
yield. Irradiation21 of the stilbene mixture with a medium-
pressure mercury lamp (100 W) through a Pyrex filter
in dry benzene using I2 (1 equiv) and propylene oxide (100
equiv) over 40 h afforded the desired 1-triptycyl[4]-
helicene 2b in 54% yield. As mentioned previously, the
methyl group in 9 was included for the purpose of
blocking photocyclization at the undesired but otherwise
less hindered position para to the triptycene. Notwith-
standing that precaution, a side product was formed and
identified (by 1H NMR and MS) as compound 12 resulting
from cyclization onto the methyl-bearing carbon of 9 with
loss of the methyl group. This product (12) is generated
in a 1:4 ratio with respect to the desired helicene 2b.
(15) (a) Akiyama, S.; Nakasuji, N.; Nakagama, M. Bull. Chem. Soc.
J pn. 1971, 44, 2231-2236. (b) Mallory, F. B.; Mallory, C. W. J . Org.
Chem. 1983, 48, 526-532.
(16) We note, however, that when we replaced the 9-triptycyl unit
in 6 by an anthracene (21) the results of the attempted photocyclization
were equivalent (eq i). Propargylic aldehyde 20 was prepared from the
corresponding dibromide (38): see the Experimental Section.
Parallel to the development of the successful route
described above, we also examined (Scheme 5) the
approach C where key steps are the synthesis of the
1-bromobenzo[c]phenanthrene 15 and the attachment of
the triptycene unit to it. A Wittig reaction between the
ylide derived from 1020 and the aldehyde 2413 led to the
formation of stilbene 16 as a 1:2.3 mixture of cis/trans
isomers in excellent yield. Irradiation of the mixture
with a medium-pressure mercury lamp (100 W) through
a Pyrex filter for 20 h produced the desired 1-bromo[4]-
(17) Stille, J . K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508-524.
(18) Prepared3a from 9-bromoanthracene by metalation with n-BuLi
and reaction of the resulting aryllithium with tributyltin chloride (98%
yield).
(19) Logullo, F. M.; Seitz, A. H.; Friedman, L. Organic Syntheses;
1973; Collect. Vol. 5, pp 54-59.
(20) Geerts, J . P.; Martin, R. H. Bull. Soc. Chim. Belg. 1960, 60,
563-569.
(21) See the Experimental Section for general conditions of irradia-
tion. See also: Lui, L.; Yang, B.; Katz, T. J .; Poindexter, M. K. J . Org.
Chem. 1991, 56, 3769-3775.

In Search of a Molecular Ratchet
J . Org. Chem., Vol. 63, No. 11, 1998 3659
Sch em e 6
Sch em e 7
helicene 15 in good yield. Literature precedents22 indi-
cate that bromine substituents on benzene rings normally
direct photocyclizations away from their ortho positions.
In our case, however, the methyl group functions as a
blocking agent and directs the reaction ortho to the
bromine. Palladium-catalyzed coupling of 15 with 9-(tribu-
tylstannyl)anthracene18 gave the desired coupled product
25 as part of a complex mixture of products.23 Unfortu-
nately, benzyne addition to that 25-containing mixture,
under the same experimental conditions used for the
synthesis of 11, never afforded any of the desired
1-triptycyl[4]helicene 2b, even though by then we had
an authentic sample of 2b (from approach B) to use as
an aid in assaying the outcome of the reaction.
conditions15 led, unfortunately, to the photocyclization
product with loss of the bromide (29b).24
Given the foregoing results, we undertook the prepara-
tion of the 4-substituted phenanthrene using a nonpho-
tochemical route (Scheme 7). The known ketone 3025 was
transformed into the enol triflate 31 in satisfactory yield.
Palladium-catalyzed coupling of 31 with 9-(tributylstan-
nyl)anthracene afforded 32. Dehydrogenation26 of 32
with DDQ afforded 33, but reaction of the latter with
benzyne failed to give any detectable amount of the
desired 1. The major isolated product, obtained in 43%
yield, is tentatively assigned structure 34 and results
from the addition of two molecules of benzyne to 33.
Apparently, 1 is so strained that it is more reactive than
33 toward benzyne. When the order of reaction of 32
with DDQ and benzyne was reversed, the problem was
solved, the benzyne adding only across the 9,10 positions
of the anthracene. The desired 4-triptycylphenanthrene
(1) was thus obtained from 30 in a three-pot sequence in
23% overall yield.
In summary, using the convergent (four steps, 20%
overall yield) stilbene photocyclization route summarized
in Scheme 4, the 1-triptycyl[4]helicene 2b was synthe-
sized.
Syn th esis of 4-(9-Tr ip tycyl)p h en a n th r en e (1). On
the basis of the experience obtained from the synthesis
of the [4]helicene 2b, in undertaking the synthesis of 4-(9-
triptycyl)phenanthrene (1), we initially used the stilbenic
strategy (approach B) directly (Scheme 6). Following
standard procedures we prepared the stilbene 27 as a
1:1 cis/trans mixture of isomers. Surprisingly, irradiation
of the mixture under the same conditions effective for
cyclizing 9 and 16 did not afford any of the desired
cyclized product; instead, only cis/trans isomerization was
observed. Further studies using different filters, oxida-
tive conditions, lamps (100 or 450 W), and/or solvents
failed to result in formation of any of the desired product.
1
NMR Da ta An a lysis. The H NMR spectra of 1 and
2b indicate that at room temperature rotation around
the triptycene/helicene bond is frozen in both compounds
but that the topology of the two is distinctly different. In
1
particular, the H NMR spectrum of 1 reveals a plane of
symmetry since two (but not all three) of the rings of the
triptycene unit are equivalent. The phenanthrene com-
ponent of 1 is therefore either planar or exists as a
rapidly racemizing mixture of two helicene enantiomers.
Even at 160 °C, the upper temperature limit of the NMR
These unexpected results led us to investigate alterna-
tive routes. One of them involved the introduction of the
triptycene unit after construction of a 4-substituted
phenanthrene (such as 29a ). The synthesis of 29a , from
28, was expected to be complicated by the above-noted22
propensity of a bromo substituent to direct photocycliza-
tion away from the ortho position, but as before, an extra
methyl group was expected to overcome that difficulty.
A Wittig reaction between the phosphonium salt 26 and
the aldehyde 24 delivered the stilbene 28 (cis/trans
mixture) in good yield. Irradiation under standard
1
probe, H NMR spin polarization transfer experiments
(see below) reveal no detectable rotation within two
seconds of polarization, which indicates a barrier (∆G‡)
to rotation around the triptycene/phenanthrene bond in
excess of 27 kcal/mol.
The 1H NMR room-temperature spectrum of 2b reflects
an absence of symmetry, where all three rings of the
triptycene are nonequivalent. At 160 °C, but not at lower
temperatures, however, some peak-broadening of trip-
(22) (a) Sudhakar, A.; Katz, T. J . Tetrahedron Lett. 1986, 27, 2231-
2234. (b) Sudhakar, A.; Katz, T. J .; Yang, B.-W. J . Am. Chem. Soc.
1986, 108, 2790-2791. (c) Liu, T.; Katz, T. J . Tetrahedron Lett. 1991,
32, 6831-6834.
(24) Haworth, R. D. J . Chem. Soc. 1932, 1125-1133.
(25) Huggenberg, S.; Hesse, M. Helv. Chem. Acta 1980, 63, 2295-
2301.
(26) Tashiro, M.; Yamato, T.; Kobayashi, K.; Arimura,T. J . Org.
Chem. 1987, 52, 3196-3199.
(23) Compound 25 was identified by MS; its 1H NMR spectrum
suggests a mixture of two compounds.

3660 J . Org. Chem., Vol. 63, No. 11, 1998
Kelly et al.
protons. In the latter case, the polarization that moves
should be transferred equally to the remaining two
resonances.
In the event, spin polarization of one (see 35) of the
three Hx protons in 2b was achieved using a selective
180°-pulse-delay-observe sequence. As is evident from
Figure 5, the triptycene rotates clockwise and counter-
clockwise to the same extent. Control experiments
polarizing the spin of each of the other two Hx protons
gave equivalent results.
The spin polarization transfer results demonstrate
unequivocally that rotation occurs equally in both direc-
tions. That is, 2b () 35) does not behave as a ratchet.
Three points about the spin polarization transfer NMR
experiment merit being addressed explicitly. First,
although we discuss the interpretation of the results in
terms of a single enantiomer of 2b, the experiment was
actually conducted on racemic material. Since enanti-
omers in an achiral environment give identical NMR
spectra, carrying out the measurement on a single
enantiomer would not have increased the information
generated by the experiment and, therefore, would not
have warranted the investment of effort required to
develop a resolution of 2b.
Second, since the spin polarization is transferred
equally to both of the other two protons regardless of
whether Ha, Hb, or Hc in 35 is polarized, it was not
necessary to make specific assignments of the three
relevant peaks at δ 7.6, 7.1, and 6.9. And even if
polarization had transferred to only one peak, neither
resolution of enantiomers nor specific peak assignments
would have been necessary to distinguish between uni-
directional rotation and racemization of the helicene.
Consider, for example, 35 () 2b) and its enantiomer ent-
35 () ent-2b). If unidirectional rotation of the helicene
is occurring, then regardless of whether one polarizes
(separately) Ha, Hb, or Hc, the spin will always transfer,
and the transfer will not be degenerate; i.e., in no case
will polarization transfer to two other peaks to a similar
extent, and no peak will be the primary recipient of
polarization transfer from more than one hydrogen. The
possibility of unidirectional rotation is contradicted by
the data in Figure 5.
F igu r e 5. Results of spin polarization transfer experiment
at 160 °C (calibrated temperature). The resonances for Ha, Hb,
and Hc (see 35) appear at δ 7.6, 7.1, and 6.9 ppm (not
necessarily respectively). The spin of the proton resonating at
δ 7.6 was polarized, and transfer of that polarization was
monitored over time.
tycene (but not helicene) resonances is observed, indicat-
ing somewhat less retarded rotation. Extrapolation,
based on analogy to our earlier work,3a suggested a
coalescence temperature of about 220 °C, which corre-
sponds to a ∆G‡ of approximately 25 kcal/mol, a value
that is corroborated by the studies described below and
which is reassuringly close to the calculated barrier (∆H‡,
not ∆G‡) of 22 kcal/mol.
Given the above, it was of interest to establish the
direction of rotation of the triptycene in 2b. Preparation
of isotopically labeled rotamers of 2b should be possible
but would require extensive synthetic, rotamer-separa-
tion, and structure-determination efforts. Fortunately,
the spin polarization transfer NMR technique27 affords
the same information at a small fraction of the effort. In
short, if one has a system that is conformationally mobile,
but that mobility is slow on the NMR time scale, then
one can polarize the spin of a slowly conformationally
mobile atom, wait an appropriate time, and assay where
(if anywhere) that polarization has moved to.
In the present case, where the experimentally deter-
mined barrier (∆G‡) to rotation is approximately 24.5
kcal/mol,28 then at 160 °C a single rotamer has a half-
life of about 0.17 s. Thus, if one selectively polarizes one
of the three Ha, Hb, and Hc (see 35 in Figure 5) protons
andsafter appropriate time delayssassays for the loca-
tion of that polarization, a clear-cut distinction between
predominantly unidirectional rotation and bidirectional
rotation is available. In the former case, a disproportion-
ate share of the polarization that has moved should be
transferred to a resonance for only one of the two other
Third, since separate polarization of the resonances at
7.6, 7.1, and 6.9 is transferred equally, in each instance,
to the other two resonances, one can also distinguish
between uni/bidirectional rotation and helicene racem-
ization. If the mechanism of polarization transfer were
racemization of the helicene then, for example, polariza-
tion of the peak for Hb in 35 would transfer to the peak
for Hs in ent-35 (which is magnetically equivalent to Hc
in 35) and vice versa, but polarization would not transfer
to or from Ha/Hr. The experimental results require that
the rate of bidirectional rotation is, at minimum, sub-
stantially faster than that of helicene racemization. If
the rate of helicene racemization were similar to or faster
than the rate of bidirectional rotation, then the polariza-
(27) For leading references to the spin polarization transfer method,
see: (a) Dahlquist, F. W.; Longmur, K. J .; Du Vernet, R. B. J . Magn.
Reson. 1975, 17, 406-410. (b) Frim, R.; Zilber, G.; Rabinovitz, M. J .
Chem. Soc., Chem. Commun. 1991, 1202-1203. (c) Abdourazak, A. H.;
Sygula, A.; Rabideau, P. W. J . Am. Chem. Soc. 1993, 115, 3010-3011.
For a more general review of the application of 2D NMR spectroscopy
to the kinetics of exchange processes, see: Perrin, C. L.; Dwyer, T. J .
Chem. Rev. 1990, 90, 935-967.
(28) The value of ∆G‡ is calculated from the half-life obtained from
the data in Figure 5. See, for example, Scudder, P. H. Electron Flow
in Organic Chemistry; Wiley: New York, 1992; pp 38-40.

In Search of a Molecular Ratchet
J . Org. Chem., Vol. 63, No. 11, 1998 3661
Because of the bombardments of gas molecules on the
vane, the vane oscillates and jiggles. All we have to do
is to hook onto the other end of the axle a wheel which
can turn only one waysthe ratchet and pawl. Then
when the shaft tries to jiggle one way, it will not turn,
and when it jiggles the other, it will turn. Then the
wheel will slowly turn, and perhaps we might even tie
a flea onto a string hanging from a drum on the shaft,
and lift the flea! Now let us ask if this is possible.
According to Carnot’s hypothesis, it is impossible. But
if we just look at it, we see, prima facie, that it seems
quite possible.
tion of, for instance, Hb would be transferred faster to
Hs than it leaks out to Ha/Hr, which is not observed.
The present experiment avoids the complications Laar-
hoven faced10 when seeking to determine the barrier to
rotation around the phenyl/[4]helicene bond in 3 (X )
H). NMR measurements of 3 (X ) H) indicated that the
resonances for Hx and Hy (see 36) coalesce at -4 °C, but
coalescence does not distinguish between equivalency
arising from rotation around the phenyl/[4]helicene bond
or racemization of the helicene. Laarhoven had to
include a -CH2OAc unit in 3 to differentiate between the
rotation and racemization mechanisms, where the per-
sistence of an AB quartet for the methylene protons Hq
and Hr in 36 under conditions where Hx and Hy become
equivalent requires that rotation is occurring and ex-
cludes the helicene racemization mechanism.
If one imagines mentally taking Feynman’s device
(Figure 6), shrinking the axle length to zero and fusing
the vanes to the ratchet wheel to give the assembly in
Figure 7, then 2b is the functional equivalent of that
assembly, except that in 2b the triptycene functions
simultaneously as the wheel and the vanes, and the
helicene serves double duty as the pawl and the spring.
The finding that 2b rotates equally in both directions is
contrary to the Feynman scenario excerpted above. But
as Feynman recognized, his argument, while tantalizing,
is specious. For Feynman’s device, or our 2b, to rotate
unidirectionally without an input of energy would require
the violation of the Second Law of Thermodynamics.
Th er m od yn a m ic Issu es. For those who are fully
aware of the more subtle consequences of the Second Law
of Thermodynamics, the finding that the triptycene in
2b rotates equally in both directions will come as no
surprise. But for many, especially those who have
manipulated models of 2, the finding is counterintuitive.
And for those who have been seduced by the models, the
substantial asymmetry of the energy curve (Figure 3) for
rotation around the triptycene/[4]helicene bond reinforces
the expectation that rotation in one direction will be more
facile than in the other. But Figure 3 contains a
cautionary hint: the two energy minima flanking the
energy maximum are equal in energy. More specifically,
the expectation of a preferred direction of rotation
overlooks an elementary consequence of energy diagrams,
which requires that, unlike mountain climbing, it is only
the height of the summit, not the steepness of the ascent,
that matters. Put another way, the principle of micro-
scopic reversibility rules.
F igu r e 7. Machine of Figure 6 with axle length reduced to
zero and vanes fused to wheel.
Con clu sion s. The synthesis of triptycyl[4]helicene 2b
successfully incorporates into a single molecule the
essential components of a simple ratchet: the asymmetric
combination of a ratchet wheel, a pawl, and a spring.
Tantalizing as models of 2b are, the experimental
demonstration that 2b rotates bidirectionally rather than
unidirectionally illustrates the perils of extrapolating
from macroscopic to molecular scales. It follows that
molecular units such as 2 cannot be used to induce
unidirectional rotation in an isothermal environment.
Nonetheless, their ability to modulate the barrier to free
rotation should make them useful components of more
complex systems such as molecular motors and machines.
The finding that 2b does not function as a ratchet is
reminiscent of the somewhat whimsical ratchet and pawl
conundrum posed in 1962 by Nobel Laureate physicist
Richard Feynman29 and reiterated below.
Let us try to invent a device that will violate the Second
Law of Thermodynamics, that is, a gadget that will
generate work from a heat reservoir with everything at
the same temperature. Let us say we have a box of gas
at a certain temperature, and inside there is an axle
with vanes in it. (See Figure [6] but take T1 ) T2 ) T,
say).
Exp er im en ta l Section
Gen er a l Ma ter ia ls a n d Meth od s. Unless otherwise
stated, reactions were typically conducted under an argon
atmosphere using dry, freshly distilled solvents under anhy-
drous conditions employing oven-dried glassware. Tetrahy-
drofuran (THF) and benzene were distilled from sodium/
benzophenone; dichloromethane (CH2Cl2), triethylamine (NEt3),
and pyridine (Py) from calcium hydride (CaH2). Anhydrous
dioxane was purchased from Aldrich and used directly. Methyl
formate was distilled from CaH2 before use. All other reagents
were purchased from Aldrich and used directly. For reactions
(29) Feynman, R. P.; Leighton, R. B.; Sands, M. The Feynman
Lectures on Physics; Addison-Wesley: Reading, MA, 1963; Vol. 1,
Chapter 46.
F igu r e 6. Ratchet and pawl machine (reproduced from ref
29).

3662 J . Org. Chem., Vol. 63, No. 11, 1998
Kelly et al.
where a component was added by cannula, the total volume
of solvent is given. Usually the compound was dissolved in
80% of that volume and the flask was then rinsed with the
remaining 20% of fresh solvent.
saturated aqueous NH4Cl (2 mL). The resulting solution was
poured into a separatory funnel with Et2O (50 mL) and washed
with saturated aqueous NH4Cl (30 mL). The organic layer was
dried (MgSO4) and filtered, and the solvent was evaporated
on a rotary evaporator. The crude residue was purified by
flash column chromatography (5-15% Et2O/hexanes), afford-
ing aldehyde 8 (270 mg, 68%) as a white solid: mp 263-265
°C (crystallized from Et2O/hexanes); 1H NMR (CDCl3, 400
MHz) δ 9.73 (s, 1 H), 7.66 (m, 3 H), 7.42 (m, 3 H), 7.07 (m, 6
H), 5.45 (s, 1 H); 13C NMR (CDCl3, 100 MHz) δ 177.6, 144.0,
142.6, 126.1, 125.3, 123.7, 122.0, 91.3, 90.9, 53.1, 53.0; IR
(NaCl) 2219, 1671, 1451, 752 cm-1; LRMS (EI) m/z 230 (100),
202 (65); HRMS (EI) calcd for C23H14O (M)+ 306.1045, found
Flash column chromatography was conducted according to
the procedure of Still et al.30 on silica gel purchased from
Baxter Scientific Products (S/P brand silica gel 60 A, 230-
400 mesh). 1H NMR spectra were recorded on 300, 400, or
500 MHz spectrometers; chemical shifts are reported in ppm
downfield from tetramethylsilane and are referenced to CHCl3
as internal standard. 13C NMR spectra were recorded at 100
MHz; chemical shifts are reported in ppm downfield from
tetramethylsilane and are referenced to CDCl3 as internal
standard. Infrared spectra were recorded on an FT-IR spec-
trometer in CH2Cl2 solution. Silica gel GF plates were used
for analytical TLC and visualized using short- and longwave
UV lamps. Melting points are uncorrected. Elemental analy-
ses were performed by Robertson Microlit Labs Inc., Madison,
NJ . Low-resolution mass spectra (LRMS) and high-resolution
mass spectra (HRMS) were performed by the Mass Spectrom-
etry Laboratory, School of Chemical Sciences, University of
Illinois, Urbana, IL.
306.1048. Anal. Calcd for
Found: C, 89.89; H, 4.89.
C23H14O: C, 90.17; H, 4.61.
1-(3-P h en a n th r yl)-4-(9-tr ip tycyl)bu t-1-en -3-yn e (6). To
a
cooled suspension of [(3-phenanthryl)methyl]triphenyl-
phosphonium bromide (7)15 (223 mg, 0.42 mmol) in THF at
-78 °C was added dropwise a solution of 2.3 M n-BuLi in
hexanes (182 µL, 0.42 mmol). The brown-red mixture was left
warming to room temperature over a 1 h period, stirred for
30 min at room temperature, and cooled again to -78 °C. Then
a solution of the aldehyde 8 (120 mg, 0.35 mmol) in THF (5
mL) was added over a 5 min period via cannula. The cooling
bath was removed, and after 20 min the reaction mixture was
quenched with saturated aqueous NH4Cl (2 mL), poured into
a separatory funnel with Et2O (60 mL) and washed with
saturated aqueous NH4Cl (40 mL). The organic layer was
dried (MgSO4) and filtered, and the solvent was evaporated
in vacuo. The crude product was purified by flash column
chromatography (3% Et2O/hexanes), affording the enyne 6 (150
mg, 90%) as a mixture of cis/trans isomers (1:1.5). The mixture
was dissolved in Et2O (6 mL) and the trans isomer separated
by precipitation. Trans isomer: 1H NMR (CDCl3, 400 MHz)
δ 8.79 (s, 1 H), 8.78 (d, J ) 8.0 Hz, 1 H), 7.94 (d, J ) 8.4 Hz,
2 H), 7.87 (m, 4 H), 7.79 (s, 1 H), 7.72 (m, 1 H), 7.65 (m, 2 H),
7.44 (d, J ) 6.8 Hz, 4 H), 7.12 (m, 6 H), 6.87 (d, J ) 16.4 Hz,
1 H), 5.48 (s, 1 H); 13C NMR (CDCl3, 100 MHz) δ 144.6, 144.5,
142.7, 134.4, 132.4, 132.4, 130.5, 130.3, 129.2, 128.8, 127.6,
126.9, 126.8, 126.6, 125.7, 125.3, 123.7, 123.5, 122.6, 122.0,
108.3, 92.0, 86.4, 53.4; LRMS (EI) m/z 480 (M+, 100), 265 (40),
252 (35); HRMS (EI) calcd for C38H24 480.1874, found 480.1874;
UV (MeOH) λmax 290, 310, 340 nm.
Gen er a l P r oced u r e for P h otocycliza tion s.21 The pho-
toreactor, similar to Ace Glass catalog no. 7840, was a
cylindrical glass vessel with an immersion well connected
through a standard taper 60/40 ground glass joint. The vessel
was flat-bottomed to allow a magnetic stirring bar to rotate.
The immersion well was a double-walled quartz tube cooled
by water containing a 100- or 450-W quartz Hg vapor lamp
and, if necessary, a filter. The molarity of the solutions with
respect to the starting stilbene or enyne was kept in the 10-3
M range. In stilbene photocyclizations, propylene oxide (PO)
was used21 to scavenge HI, and argon was bubbled through
the stirred solution of stilbene and iodine for 20-30 min to
deoxygenate the solution; the lamp was then turned on. The
argon flow was maintained throughout the procedure, and the
photoirradiation was monitored by 1H NMR (aliquots taken
via the angle joint, evaporating the solvent, and dissolving the
residue in CDCl3).
9-(2,2-Dibr om oeth en yl)tr ip tycen e (19). To a mixture of
zinc dust (1.60 g, 24.7 mmol) and PPh3 (6.47 g, 24.7 mmol) in
CH2Cl2 (40 mL), cooled at 0 °C, was added CBr4 (8.17 g, 24.7
mmol) in portions over 15 min. The mixture was allowed to
reach rt and stirred for 20 min, and then dry pyridine (4 mL)
was added. The brown reaction mixture was left stirring for
an additional 1 h, and a solution of the aldehyde 183a (1.73 g,
6.17 mmol) in CH2Cl2 (6 mL) was added via cannula dropwise
during 10 min. After 24 h, the reaction mixture was poured
onto a silica gel (16 cm × 3 cm) chromatography column and
the organic products were eluted with Et2O (200 mL). The
eluate was then evaporated under reduced pressure and the
residue further purified by flash column chromatography (5%
Et2O/hexanes). The bromide 19 was obtained as a solid (1.48
g, 56%): mp 218-220 °C (crystallized from Et2O/hexanes) and
the starting aldehyde 18 was also recovered (0.80 g, 30%); 1H
NMR (CDCl3, 400 MHz) δ 8.11 (s, 1 H), 7.47 (m, 3 H), 7.41 (m,
3 H), 7.04 (m, 6 H), 5.40 (s, 1 H); 13C NMR (CDCl3, 100 MHz)
δ 145.4, 142.4, 132.4, 125.7, 124.8, 124.2, 123.3, 96.6, 58.4, 54.4;
IR (NaCl) 3062, 1608, 1464, 733 cm-1; LRMS (EI, 70 eV) m/z
437.9 (M+, 45), 278.1 (100), 252 (15), 138 (30); HRMS (EI) calcd
for C22H14Br2 435.9466, found 435.9462. Anal. Calcd for
1-(9-An th r a cen yl)-2,2-d ibr om oeth en e (38).
To a mixture of zinc dust (4.73 g, 72.8 mmol) and PPh3 (19.0
g, 72.8 mmol) in CH2Cl2 (45 mL) at 0 °C was added CBr4 (24.1
g, 72.8 mmol) in portions over a 20 min period (exothermic
reaction). The mixture was stirred for 20 min at room
temperature, and dry pyridine (4 mL) was added dropwise via
syringe. The brown reaction mixture was left stirring for 1 h,
and a solution of commercial 9-anthraldehyde (5.00 g, 6.17
mmol) in CH2Cl2 (12 mL) was added via cannula over a 10
min period. After 6 h of stirring, the reaction mixture was
poured onto a chromatography column containing silica gel
(10 cm × 5 cm), and the organic products were eluted with
Et2O (400 mL). The solvents were removed under reduced
pressure, and the crude product was further purified by flash
column chromatography (5% Et2O/hexanes). Dibromide 38
(7.8 g, 86%) was obtained as a yellow solid: mp 116-117 °C
(crystallized from Et2O/hexanes); 1H NMR (CDCl3, 400 MHz)
δ 8.47 (s, 1 H), 8.12 (d, J ) 8.4 Hz, 2 H), 8.10 (s, 1 H), 8.04 (d,
J ) 8.4 Hz, 2 H), 7.60 (dt, J ) 8.2, 1.2 Hz, 2 H), 7.54 (dt, J )
8.2, 1.2 Hz, 2 H); 13C NMR (CDCl3, 100 MHz) δ 135.7, 131.4,
130.2, 129.0, 128.8, 128.1, 126.5, 125.6, 125.5, 95.5; IR (NaCl)
3044, 778 cm-1. Anal. Calcd for C16H10Br2: C, 53.08; H, 2.78;
Br, 44.14. Found: C, 53.10; H, 2.67; Br, 44.04.
C
22H14Br2: C, 60.31; H, 3.22. Found: C, 60.47; H, 2.97.
3-(9-Tr ip tycyl)p r op yn a l (8). To a solution of the vinyl
dibromide 19 (515 mg, 1.17 mmol) in THF (10 mL) cooled at
-78 °C was added a 2.3 M solution of n-BuLi in hexanes (1.22
mL, 2.8 mmol) dropwise via syringe over a 10 min period.
The resulting light-yellow solution was stirred for 20 min at
-78 °C and then transferred via cannula over a 10 min period
to a cooled (-78 °C) solution of methyl formate (360 µL, 5.85
mmol) in THF (3 mL). The reaction mixture was slowly
warmed to room temperature over 2 h, stirred at room
temperature for an additional 4 h, and then quenched with
3-(9-An th r a cen yl)p r op yn a l (20). To a solution of 1-(9-
anthracenyl)-2,2-dibromoethene (38, 1.50 g, 4.0 mmol) in THF
(30) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923-
2925.

In Search of a Molecular Ratchet
J . Org. Chem., Vol. 63, No. 11, 1998 3663
(20 mL) cooled at -78 °C was added a 2.5 M solution of n-BuLi
in hexanes (3.84 mL, 9.60 mmol) over a 10 min period via
syringe. The resulting green mixture was stirred for 15 min
at -78 °C and then transferred via cannula over a 5 min period
to a solution of distilled methyl formate (495 µL, 8.0 mmol,
2.0 equiv) in THF (5 mL) cooled to -78 °C. The reaction
mixture was slowly (1 h) warmed to room temperature, and
stirred for an additional 1 h at room temperature, and then
quenched with saturated aqueous NH4Cl (2 mL), poured into
a separatory funnel containing Et2O (60 mL), and washed with
saturated aqueous NH4Cl (30 mL). The organic layer was
dried (MgSO4) and filtered, and the solvent was evaporated
in vacuo. The crude residue was purified by flash column
chromatography (5-15% Et2O/hexanes), affording the alde-
hyde 20 (650 mg, 82%) as yellow, wooly needles: mp 147-
149 °C (crystallized from AcOEt/hexanes); 1H NMR (CDCl3,
400 MHz) δ 9.74 (s, 1 H), 8.62 (s, 1 H), 8.56 (d, J ) 8.8 Hz, 2
H), 8.07 (d, J ) 8.4 Hz, 2 H), 7.57 (dt, J ) 6.8, 1.2 Hz, 2 H),
7.68 (dt, J ) 6.8, 1.2 Hz, 2 H); 13C NMR (CDCl3, 100 MHz) δ
176.7, 134.7, 132.3, 131.2, 129.3, 128.5, 126.4, 126.3, 112.8,
100.0, 92.8; IR (NaCl) 2168, 1653, 729 cm-1; LRMS (EI) m/z
230 (M+, 100), 202 (67); HRMS (EI) calcd for C17H10O 230.0730,
found 230.0731. Anal. Calcd for C17H10O: C, 88.67; H, 4.38.
Found: C, 88.46; H, 4.33.
2-Meth yl-5-(9-tr ip tycyl)ben za ld eh yd e (11). To a reflux-
ing solution of 5-(9-anthracenyl)-2-methylbenzaldehyde (23)
(600 mg, 2.03 mmol) in dioxane (20 mL) in a two-necked round-
bottomed flask was added a previously prepared suspension
of benzenediazonium-2-carboxylate19 (16.2 mmol, 8.0 equiv) in
dioxane (10 mL) portionwise over a 2 h period via Pasteur pipet
(CAUTION: when dry, benzenediazonium-2-carboxylate deto-
nates violently on being scraped or heated). The reaction was
1
monitored by H NMR by removing aliquots, evaporating the
dioxane, and dissolving the residue in CDCl3. The reaction
mixture was cooled to room temperature, and the solvent was
removed under reduced pressure. The black crude product
was purified by flash column chromatography (5% Et2O/
hexanes), affording the triptycene 11 (460 mg, 62%) as a white
crystalline solid: mp 236 °C (crystallized from Et2O/hexanes);
1H NMR (CDCl3, 400 MHz) δ 10.38 (s, 1 H), 8.61 (d, J ) 2 Hz,
1 H), 8.22 (dd, J ) 8.0, 2.0 Hz, 1 H), 7.57 (d, J ) 8.0 Hz, 2 H),
7.45 (d, J ) 7.2 Hz, 3 H), 7.20 (d, J ) 7.2 Hz, 3 H), 7.02 (d, J
) 7.6 Hz, 3 H), 6.96 (t, J ) 7.6 Hz, 3 H), 2.84 (s, 3 H); 13C
NMR (CDCl3, 100 MHz) δ 193.3, 147.0, 146.6, 139.9, 137.1,
135.8, 135.2, 134.7, 132.4, 125.6, 125.0, 124.4, 124.1, 60.0, 55.4,
19.8; IR (NaCl) 3065, 1695, 1456, 749 cm-1; LRMS (EI, 70 eV)
m/z 372 (M+, 50), 252 (69), 117 (71), 116 (64), 115 (80); HRMS
(EI) calcd for C28H20O 372.1514, found 372.1514. Anal. Calcd
for C28H20O: C, 90.29; H, 5.41. Found: C, 90.06; H, 5.58.
1-[2-Met h yl-5-(9-t r ip t ycyl)p h en yl]-2-(2-n a p h t h yl)et h -
en e (9). To a solution of [(2-naphthyl)methyl]triphenylphos-
phonium bromide20 (315 mg, 0.64 mmol) in THF (20 mL) cooled
at -78 °C, was added via syringe a 2.3 M solution of n-BuLi
in hexanes (280 µL, 0.64 mmol) dropwise over a 10 min period.
The reaction mixture was allowed to warm to room temper-
ature by removing the bath and stirred at room temperature
for 30 min. The red ylide was then cooled to -78 °C, and a
solution of the aldehyde 11 (203 mg, 0.54 mmol) in THF (5
mL) was added via cannula dropwise over a 5 min period. The
cooling bath was removed, and the reaction mixture was left
stirring for an additional 1 h at room temperature and then
quenched with saturated aqueous NH4Cl (2 mL). The reaction
mixture was poured into a separatory funnel with Et2O (60
mL) and washed with saturated aqueous NH4Cl (40 mL). The
organic layer was dried (MgSO4) and filtered, and the solvent
was evaporated in vacuo. The resulting crude residue was
purified by flash column chromatography (10% Et2O/hexanes)
to give stilbene 9 (230 mg, 85%) as a mixture of cis/trans
isomers (1:1): 1H NMR of mixture (CDCl3, 400 MHz) δ 8.44
(s), 7.99 (d, J ) 8.0 Hz), 7.92 (dd, J ) 8.2 Hz), 7.86 (m), 7.84
(s), 7.78 (d, J ) 1.6 Hz), 7.77 (m), 7.70 (d, J ) 8.0 Hz), 7.55 (d,
J ) 8.0 Hz), 7.51 (d, J ) 8.0 Hz), 7.45 (m), 7.47 (d, J ) 7.6
Hz), 7.40 (dd, J ) 8.4, 1.6 Hz), 7.39 (d, J ) 7.2 Hz), 7.32 (d, J
) 6.8 Hz), 7.19 (d, J ) 16.4 Hz), 7.02 (dt, J ) 8.4, 1.6 Hz),
6.94 (s), 6.43 (dt, J ) 12.8, 8.0 Hz), 5.51 (s), 5.31 (s), 2.72 (s),
2.60 (s); IR (NaCl) 3053, 2944, 1456, 747 cm-1; LRMS (EI) m/z
496 (M+, 87), 252 (M - Tript, 100); HRMS (EI) calcd for C39H28
496.2192, found 496.2191; UV (benzene) λmax 283, 321 nm.
Anal. Calcd for C39H28: C, 94.32; H, 5.68. Found: C, 93.95;
H, 5.47.
4-Meth yl-1-tr ip tycylben zo[c]p h en a n th r en e (2b). To a
solution of stilbene 9 (193 mg, 0.39 mmol) in benzene (250 mL)
in the photoreactor under an argon atmosphere were added
I2 (99 mg, 0.39 mmol) and propylene oxide (1 mL). The
mixture was irradiated (100 W mercury lamp) through a Pyrex
filter over 40 h, progress of the reaction being monitored by
1H NMR. The solvent was evaporated, and the crude residue
was purified by flash column chromatography (2 to 5% Et2O/
hexanes). The benzo[c]phenanthrene 2b was further purified
by crystallization (Et2O/hexanes), obtaining transparent crys-
tals (104 mg, 54%): mp 308-310 °C; 1H NMR (CDCl3, 300
MHz) δ 8.43 (m, 1 H), 8.33 (d, J ) 8.4 Hz, 1 H), 8.23 (d, J )
7.2 Hz, 1 H), 7.96 (d, J ) 7.5 Hz, 1 H), 7.87 (t, J ) 8.4 Hz, 2
H), 7.76 (d, J ) 8.1 Hz, 1 H), 7.75 (d, J ) 8.4 Hz, 1 H), 7.57
(m, 1 H), 7.35 (m, 3 H), 7.05 (dd, J ) 7.2, 1.5 Hz, 1 H), 6.88
(dd, J ) 7.6, 1.2 Hz, 1 H), 6.82 (t, J ) 7.8 Hz, 1 H), 6.61 (dq,
J ) 7.5, 1.8 Hz, 2 H), 6.48 (d, J ) 8 Hz, 1 H), 6.38 (dt, J ) 7.2,
1.2 Hz, 1 H), 5.92 (dt, J ) 7.8, 1.2 Hz, 1 H), 5.87 (d, J ) 7.8
Hz, 1 H), 5.79 (dt, J ) 7.8, 1.2 Hz, 1 H), 5.10 (s, 1 H), 3.00 (s,
1-(9-An th r acen yl)-4-(3-ph en an th r yl)bu t-1-en -3-yn e (21).
To a cooled solution of [(3-phenanthryl)methyl]triphenyl-
phosphonium bromide (7)15 (239 mg, 0.45 mmol) in THF at
-78 °C was added a 2.3 M solution of n-BuLi in hexanes (195
µL, 0.45 mmol). The cooling bath was removed, and the
brown-red mixture was left stirring for 30 min at room
temperature. The reaction mixture was cooled again to -78
°C, and a solution of the aldehyde 20 (100 mg, 0.431 mmol) in
THF (9 mL) was added dropwise (5 min) via cannula. The
reaction mixture was allowed to warm to room temperature
and after 20 min quenched with saturated aqueous NH4Cl (2
mL). The resulting solution was poured into a separatory
funnel with Et2O (60 mL) and washed with saturated aqueous
NH4Cl (50 mL). The organic layer was dried (MgSO4) and
filtered, and the solvent was removed in vacuo. The crude
product was purified by flash column chromatography (3%
Et2O/hexanes), affording the enyne 21 (150 mg, 86%) as a
mixture of cis/trans isomers (1.8:1): 1H NMR of mixture
(CDCl3, 400 MHz) δ 9.45 (s), 8.75 (br s), 8.66 (d, J ) 8.8 Hz),
8.50 (d, J ) 8.4 Hz), 8.45 (d, J ) 8.0 Hz), 8.30 (d, J ) 8.4 Hz),
8.02 (m), 7.91 (m), 7.84 (d, J ) 8.0 Hz), 7.70 (m), 7.76 (s), 7.65-
7.37 (m), 7.16 (t, J ) 8.1 Hz), 7.11 (d, J ) 9.6 Hz), 6.93 (d, J
) 16.0 Hz), 6.41 (d, J ) 12.0 Hz); HRMS (EI) calcd for C32H20
404.1562, found 404.1565; UV (hexanes) λmax 285, 340, 425 nm.
5-(9-An th r a cen yl)-2-m eth ylben za ld eh yd e (23). A solu-
tion of 5-bromo-2-methylbenzaldehyde13 (1.0 g, 5 mmol) in dry
dioxane (6 mL) was placed into a Schlenk tube under argon,
and a solution of 9-(tributylstannyl)anthracene18 (4.84 g, 10.4
mmol) in dioxane (4 mL) and (PPh3)2PdCl2 (175 mg, 5%) were
added. The tube was closed, and the mixture was heated at
130 °C for 60 h with magnetic stirring. Then the reaction
mixture was cooled to room temperature and filtered through
a short pad of silica gel, washing the solids with Et2O (100
mL). The combined filtrate and washes were evaporated, and
the crude product was purified by flash column chromatogra-
phy (3-10% AcOEt/hexanes). The coupling product 23 was
obtained as a light yellow solid (930 mg, 62%): mp 147-148
1
°C (crystallized from CH2Cl2/hexanes); H NMR (CDCl3, 400
MHz) δ 10.38 (s, 1 H), 8.58 (s, 1 H), 8.11 (d, J ) 5.6 Hz, 2 H),
7.92 (d, J ) 1.6 Hz, 1 H), 7.65 (d, J ) 8.8 Hz, 2 H), 7.61 (dd,
J ) 8.8, 1.6 Hz, 1 H), 7.55 (d, J ) 8.0 Hz, 1 H), 7.51 (t, J ) 6.0
Hz, 2 H), 7.41 (t, J ) 8.0 Hz, 2 H), 2.90 (s, 3 H); 13C NMR
(CDCl3, 100 MHz) δ 193.1, 140.0, 137.2, 136.6, 135.4, 135.3,
134.4, 132.3, 131.5, 130.3, 128.7, 127.3, 126.5, 125.9, 125.4,
19.9; IR (NaCl) 3049, 1696, 740 cm-1; LRMS (EI) m/z 296.2
(M+, 100), 252 (60); HRMS (EI) calcd for C22H16O 296.1201,
found 296.1201. Anal. Calcd for C22H16O: C, 89.16; H, 5.44.
Found: C, 88.94; H, 5.34.

3664 J . Org. Chem., Vol. 63, No. 11, 1998
Kelly et al.
was purified by flash column chromatography (3% Et2O/
hexanes) to give the bromobenzo[c]phenanthrene 15 (150 mg,
50%) as a white solid: mp 163-165 °C (crystallized from Et2O/
hexanes); 1H NMR (CDCl3, 300 MHz) δ 8.30 (m, 1 H), 7.99
(m, 3 H), 7.86 (d, J ) 8 Hz, 1 H), 7.81 (d, J ) 8.0 Hz, 2 H),
7.60 (m, 2 H), 7.35 (d, J ) 8.0 Hz, 1 H), 2.79 (s, 3 H); 13C NMR
(CDCl3, 100 MHz) δ 134.7, 134.0, 132.5, 132.3, 132.1, 131.9,
129.0, 128.9, 128.7, 128.0, 127.5, 127.4, 127.1, 126.1, 125.6,
124.8, 122.7, 119.5, 19.9; LRMS (EI) 322 (M+, 65), 320 (65),
241 (100), 226 (90); HRMS (EI) calcd for C19H13Br 320.0200,
found 320.0200. Anal. Calcd for C19H13Br: C, 71.05; H, 4.08;
Br, 24.88. Found: C, 70.93; H, 3.82; Br, 24.73.
F igu r e 8. Pulse sequence for spin polarization transfer
experiment.
1-Meth yl-2-styr yl-4-tr ip tycylben zen e (27). To a solution
of commercial benzyltriphenylphosphonium chloride (112 mg,
0.29 mmol) in THF (8 mL) at room temperature was added a
25 wt % solution of NaOMe (66 µL, 0.29 mmol) in MeOH
dropwise over a 10 min period via syringe. The red solution
was stirred for 1 h, and a solution of the aldehyde 11 (112
mg, 0.29 mmol) in THF (5 mL) was added via cannula over a
10 min period. The reaction mixture was stirred for 15 min,
quenched with saturated aqueous NH4Cl (2 mL), poured into
a separatory funnel with Et2O (60 mL), and washed with
saturated aqueous NH4Cl (20 mL). The organic layer was
dried (MgSO4) and filtered, and the solvent was evaporated
under reduced pressure. Purification of the residue by flash
column chromatography (10% AcOEt/hexanes) afforded stil-
bene 27 (95 mg, 76%) as a mixture of cis/trans isomers (1:1):
1H NMR of the mixture (CDCl3, 400 MHz) δ 8.35 and 7.81 (d,
J ) 2.4 Hz, 1 H), 7.88 and 7.85 (dd, J ) 11, 2.4 Hz, 1 H), 7.54-
7.43 (m, 4 H), 7.34 (d, J ) 8.0 Hz, 3 H), 7.25 (d, J ) 8.0 Hz, 3
H), 7.05-6.9 (m, 7 H), 6.85-6.70 (m, 3 H), 5.44 and 5.33 (s, 1
H), 2.62 and 2.52 (s, 3 H); HRMS (EI) calcd for C35H26
446.2000, found 446.2001.
3 H); 13C NMR (CDCl3, 100 MHz) δ 151.9, 150.4, 147.0, 145.2,
143.8, 143.6, 137.3, 134.9, 134.4, 133.2, 131.5, 130.9, 130.8,
130.4, 129.9, 129.7, 127.9, 127.6, 126.4, 126.2, 126.1, 126.0,
125.9, 125.6, 125.5, 124.6, 124.2, 124.0, 124.0, 123.9, 123.9,
123.7, 123.5, 123.1, 122.7, 122.2, 65.8, 55.8, 20.6; LRMS (EI)
494 (M+, 87), 252 (- Tript, 100); HRMS (EI) calcd for C39H26
494.2030, found 494.2034. Anal. Calcd for C39H26: C, 94.70;
H, 5.30. Found: C, 94.80; H, 5.13.
Sp in P ola r iza tion Tr a n sfer Exp er im en t on 2b.
A
Varian Unity 300 spectrometer was used with temperature
control. The resonances for Ha, Hb, and Hc were identified by
NOE’s from the bridgehead proton, whose chemical shift (δ
5.10) is unique. 1H T1 values were measured using the
inversion recovery method. T1 values for the protons of
interest (Ha, Hb, and Hc) are about 2 s. The exchange rate is
faster than the relaxation rate but still slow enough to give
resolved spectral lines. Therefore, a spin polarization transfer
experiment can be used to measure the exchange rate. The
pulse sequence is shown in Figure 8. A 10 s delay was used
before the first pulse. The initial 180° pulse with a duration
of 0.013 s is a selective pulse that inverts the Ha, Hb, or Hc
resonance; it is followed by different delay times and then by
a 90° detection pulse (17.5 µs). The data were acquired in
DMSO at 160 °C. Each spectrum shown in Figure 5 represents
the difference between the application of a 0.013 s, 180°
selective pulse and a normal pulse sequence.
1-(2-Meth yl-5-br om oph en yl)-2-(2-n a p h th yl)eth en e (16).
To a solution of (2-naphthylmethyl)triphenylphosphonium
bromide20 (800 mg, 1.65 mmol) cooled at -78 °C was added a
2.30 M solution of n-BuLi in hexanes (1.65 mmol, 720 µL) over
a 10 min period via syringe. The red solution was allowed to
warm to room temperature by removing the cooling bath and
stirred for an additional 30 min; then the resulting ylide was
cooled again to -78 °C, and a solution of the aldehyde 24 (300
mg, 1.51 mmol) in THF (5 mL) was added via cannula over a
5 min period. The mixture was warmed to room temperature
over 15 min and after an additional 1 h quenched with
saturated aqueous NH4Cl (1.5 mL). The reaction mixture was
poured into a separatory funnel with Et2O (80 mL) and washed
with saturated aqueous NH4Cl (30 mL). The organic layer was
dried (MgSO4) and filtered, and the solvent was evaporated
under reduced pressure. The crude product was purified by
flash column chromatography (10% Et2O/hexanes) to give the
bromostilbene 16 (395 mg, 83%) as a mixture of cis/trans
isomers (1:2.3): 1H NMR of mixture (CDCl3, 300 MHz) δ 7.88
(m), 7.84 (d, J ) 2.0 Hz), 7.80-7.73 (m), 7.67 (s), 7.63 (d, J )
8.8 Hz), 7.56-7.50 (m), 7.49-7.44 (m), 7.40 (s), 7.38-7.33 (m),
7.22 (dd, J ) 8.4, 1.2 Hz), 7.16 (s), 7.12 (d, J ) 8.4 Hz), 6.85
(d, AB system, J ) 12.0 Hz), 6.67 (d, AB system, J ) 12.0
Hz), 2.43 (s), 2.21 (s); LRMS (EI) 324 (M+, 100), 322 (95), 228
(45); HRMS (EI) calcd for C19H15Br 322.0357, found 322.0357.
Anal. Calcd for C19H15Br: C, 70.60; H, 4.68. Found: C, 70.55;
H, 4.51.
4-Br om o-1-m eth yl-2-styr ylben zen e (28). To a solution
of commercial benzyltriphenylphosphonium bromide (1.12 g,
2.59 mmol) in THF (25 mL) at room temperature was added
a 25 wt % solution of NaOMe (0.69 mL, 3.0 mmol) in MeOH
dropwise over a 10 min period via syringe. The orange solution
was stirred for 45 min, and a solution of the aldehyde 24 (430
mg, 2.16 mmol) in THF (15 mL) was added via cannula over
a 10 min period. The reaction mixture was stirred overnight,
quenched with saturated aqueous NH4Cl (2 mL), poured into
a separatory funnel with Et2O (60 mL), and washed with
saturated aqueous NH4Cl (40 mL). The organic layer was
dried (MgSO4) and filtered and the solvent evaporated under
reduced pressure. Purification of the residue by flash column
chromatography (hexanes) afforded stilbene 28 (356 mg, 61%)
as a mixture of cis/trans isomers (1:1): 1H NMR of the mixture
(CDCl3, 400 MHz) δ 7.72 (d, J ) 2.8 Hz), 7.53 (d, J ) 7.6 Hz),
7.39 (t, J ) 7.3 Hz), 7.27 (m), 7.18 (m), 7.05 (m), 6.99 (d, J )
16 Hz), 6.65 (d, AB system, J ) 12.0 Hz), 6.55 (d, AB system,
J ) 12 Hz), 2.37 (s), 2.11 (s); 13C NMR (CDCl3, 100 MHz) δ
139.2, 138.4, 137.1, 136.4, 135.0, 134.6, 131.9, 131.7, 131.6,
131.4, 131.2, 130.2, 130.1, 128.8, 128.7, 128.2, 128.1, 128.0,
127.9, 127.8, 127.4, 127.3, 127.1, 126.9, 126.8, 126.7, 125.1,
119.8, 119.1; HRMS (EI) calcd for C15H13Br 272.0200, found
272.0163.
1-Meth ylp h en a n th r en e (29b). A solution of stilbene 28
(120 mg, 0.44 mmol) and I2 (11 mg, 0.04 mmol) in benzene
(220 mL) was prepared in the photoreactor under an argon
atmosphere, and propylene oxide (3 mL) was added following
standard procedures. The mixture was irradiated with a
medium-pressure mercury lamp (100 W) through a Pyrex filter
during 12 h (monitored by GC-MS), and no reaction was
observed. The filter was replaced with a Vycor filter, and the
reaction mixture was irradiated for 7 h. The solvent was
evaporated in vacuo and the crude product purified by flash
column chromatography (hexanes) to give the phenanthrene
29b24 (25 mg, 21%), recovering part of the starting stilbene
28 (35 mg, 30%): 1H NMR of 29b (CDCl3, 300 MHz) δ 8.70 (d,
J ) 10.8 Hz, 1 H), 8.58 (d, J ) 10.8 Hz, 1 H), 7.95 (d, J ) 12.0
Hz, 1 H), 7.89 (dd, J ) 12.0, 2.0 Hz, 1 H), 7.78 (d, J ) 12.0 Hz,
1 H), 7.60 (m, 3 H), 7.74 (d, J ) 7.6 Hz, 1 H), 2.76 (s, 3 H);
LRMS (EI) 192 (M+, 100), 191 (58), 189 (36), 165 (43).
1-Br om o-4-m eth ylben zo[c]p h en a n th r en e (15). A solu-
tion of stilbene 16 (350 mg, 1.04 mmol) and I2 (289 mg, 1.14
mmol) in benzene (700 mL) was prepared in the photoreactor
under an argon atmosphere, and propylene oxide (2.5 mL) was
added following standard procedures. The mixture was ir-
radiated with a medium-pressure mercury lamp (100 W)
through a Pyrex filter during 20 h (monitored by GC-MS).
The solvent was evaporated in vacuo, and the crude product

In Search of a Molecular Ratchet
J . Org. Chem., Vol. 63, No. 11, 1998 3665
4-[(Tr iflu or om eth an esu lfon yl)oxy]-1,2-dih ydr oph en an -
th r en e (31). To a solution of phenanthrone 3025 (750 mg, 3.82
mmol) and 2,6-lutidine (890 µL, 7.65 mmol) in CH2Cl2 (20 mL)
was added at 0 °C trifluoromethanesulfonic anhydride (1.30
mL, 6.68 mmol) dropwise over a 5 min period via syringe. The
reaction mixture was allowed to reach rt and stirred overnight.
The solvent was evaporated under reduced pressure, and the
resulting crude product was purified by flash column chroma-
tography (hexanes) to give 31 (960 mg, 77%) as a white foam:
1H NMR (CDCl3, 400 MHz) δ 8.39 (d, J ) 8.8 Hz, 1 H), 7.77
(d, J ) 8.0 Hz, 1 H), 7.72 (d, J ) 8.0 Hz, 1 H), 7.53-7.41 (m,
2 H), 7.27 (d, J ) 8.0 Hz, 1 H), 6.23 (t, J ) 5.2 Hz, 1 H), 2.90-
2.84 (m, 2 H), 2.37-2.31 (m, 2 H); 13C NMR (CDCl3, 100 MHz)
δ 147.8, 137.6, 133.4, 129.9, 128.8, 128.1, 126.9, 125.8, 125.4,
125.0, 124.6, 123.3, 121.0, 120.1, 116.9, 113.8, 29.2, 22.1; LRMS
(EI, 70 eV) m/z 329 ([M + H]+, 10), 328 (M+, 55), 167 (100),
165 (54); HRMS (EI) calcd for C15H11O3SF3 328.0379, found
328.0379.
Hz, 2 H), 7.35-7.33 (m, 1 H), 7.21 (d, J ) 8.8 Hz, 1 H), 7.17 (t,
J ) 7.3 Hz, 1 H), 7.00-6.96 (m, 2 H), 6.89-6.86 (m, 2 H), 6.79
(d, J ) 7.6 Hz, 2 H), 6.59-6.55 (m, 2 H), 6.38-6.34 (m, 1 H),
5.46 (s, 1 H); 13C NMR (CDCl3, 100 MHz) δ 150.2, 148.7, 145.3,
144.1, 134.6, 134.0, 132.2, 131.7, 131.6, 131.5, 130.6, 128.9,
127.5, 127.1, 126.9, 125.9, 125.3, 124.7, 124.5, 124.4, 123.5,
123.3, 122.0, 63.5, 55.2; IR (film) 3055, 2948, 1451, 746 cm-1
;
LRMS (EI, 70 eV) m/z 431 ([M + H]+, 45), 430 (M+, 100), 253
(21), 252 (30); HRMS (EI) calcd for C34H22 430.1721, found
430.1721.
4-(9-An th r a cen yl)p h en a n th r en e (33). A solution of di-
hydrophenanthrene 32 (160 mg, 0.45 mmol) and DDQ (153
mg, 0.67 mmol) in benzene (15 mL) was heated at reflux for 6
h. After cooling, the solvent was evaporated under vacuum,
and the residue was purified by flash column chromatography
(5% EtOAc/hexanes) to afford the phenanthrene 33 (142 mg,
89%) as an amorphous solid that was crystallized from
hexanes: mp 139-140 °C; 1H NMR (CDCl3, 400 MHz) δ 8.64
(s, 1 H), 8.12-8.10 (m, 3 H), 7.94 (d, J ) 9.0 Hz, 1 H), 7.81 (d,
J ) 9.0 Hz, 1 H), 7.78 (d, J ) 8.0 Hz, 1 H), 7.74 (t, J ) 7.6 Hz,
1 H), 7.48-7.42 (m, 5 H), 7.27-7.17 (m, 3 H), 7.09 (d, J ) 8.8
Hz, 1 H), 6.66 (t, J ) 7.8 Hz, 1 H); 13C NMR (CDCl3, 100 MHz)
δ 139.3, 136.2, 133.6, 133.1, 132.4, 131.7, 130.3, 129.9, 129.3,
128.4, 128.3, 127.9, 127.8, 126.7, 126.4, 126.0, 125.9, 125.8,
4-(9-An th r a cen yl)-1,2-d ih yd r op h en a n th r en e (32).
A
mixture of vinyl triflate 31 (1.33 g, 4.04 mmol), 9-(tributyl-
stannyl)anthracene (2.83 g, 6.05 mmol), anhydrous LiCl (850
mg, 20.2 mmol), Pd(PPh3)4 (934 mg, 0.800 mmol), and dioxane
(15 mL) was placed into a sealable tube. After several cycles
of evacuation and addition of a N2 atmosphere, the tube was
sealed, and the solution was heated at 120 °C for 3 days. The
solvent was removed under reduced pressure, and the corre-
sponding crude reaction product was purified by flash column
chromatography (hexanes), affording 32 (791 mg, 55%) as a
yellow solid: mp 154-155 °C (needles crystallized from EtOAc/
hexanes); 1H NMR (CDCl3, 400 MHz) δ 8.46 (s, 1H), 8.04-
8.00 (m, 4H), 7.74 (d, J ) 8.4 Hz, 1 H), 7.65 (d, J ) 8.0 Hz, 1
H), 7.53 (d, J ) 8.0 Hz, 1 H), 7.43-7.25 (m, 4 H), 7.10-7.00
(m, 1 H), 6.92 (d, J ) 9.2 Hz, 1 H), 6.56-6.50 (m, 1 H), 6.41-
6.37 (m, 1 H), 3.22-3.17 (m, 2 H), 2.62-2.56 (m, 2 H); 13C NMR
(CDCl3, 100 MHz) δ 139.0, 135.9, 135.7, 135.4, 133.5, 132.8,
131.7, 129.8, 129.7, 128.5, 128.4, 128.0, 126.8, 126.5, 126.3,
125.5, 125.4, 125.1, 124.6, 124.2, 30.5, 23.1; IR (neat) 3056,
2931-2829, 1437, 810, 737 cm-1; LRMS (EI, 70 eV) m/z 357
([M + H]+, 29), 356 (M+, 100); HRMS (EI) calcd for C28H20
356.1565, found 356.1565. Anal. Calcd for C28H20: C, 94.35;
H, 5.65. Found: C, 94.14; H, 5.61.
4-(9-Tr ip tycyl)p h en a n th r en e (1). To a refluxing solution
of dihydrophenanthrene 32 (210 mg, 0.59 mmol) in dioxane
(15 mL) in a two-neck round-bottomed flask was added a
previously prepared19 suspension of benzenediazonium-2-
carboxylate (4.8 mmol, 8 equiv) in dioxane portionwise during
2 h via Pasteur pipet (CAUTION: when dry, benzenediazo-
nium-2-carboxylate detonates violently on being scraped or
heated). The reaction was monitored by 1H NMR by removing
aliquots, evaporating the dioxane, and dissolving the residue
in CDCl3. The reaction was allowed to cool to room temper-
ature and the solvent was evaporated under reduced pressure.
The crude product was purified by flash column chromatog-
raphy (10% EtOAc/hexanes), affording a pale yellow oil that
was immediately dissolved in benzene (10 mL). Then, DDQ
(163 mg, 0.72 mmol) was added, and the reaction mixture was
heated at reflux for 10 h. The solvent was evaporated in a
rotary evaporator and the crude residue was purified by flash
column chromatography (20% CH2Cl2/hexanes) to give the
desired triptycylphenanthrene 1 (127 mg, 51%) as a white
solid: mp 316-317 °C (crystallized from CH2Cl2/hexanes); 1H
NMR (CDCl3, 400 MHz) δ 8.38 (d, J ) 8.0 Hz, 1 H), 8.06 (d, J
) 8.0 Hz, 1 H), 7.87-7.80 (m, 2 H), 7.75 (d, J ) 8.4 Hz, 1 H),
7.72 (d, J ) 8.0 Hz, 1 H), 7.56-7.54 (m, 1 H), 7.50 (d, J ) 7.6
125.7, 125.4; IR (film) 3047, 1441, 1294, 907, 827, 753 cm-1
;
HRMS (EI) calcd for C28H18 354.1405, found 354.1408. Anal.
Calcd for C28H18: C, 94.88; H, 5.12. Found: C, 94.87; H, 4.99.
7,12-Eth en o-12-(9-tr iptycyl)-3H-ben z[a ]an th r acen e (34).
To a refluxing solution of phenanthrene 33 (105 mg, 0.30
mmol) in dioxane (25 mL), in a two-neck round-bottomed flask,
was added a previously prepared19 suspension of benzenedia-
zonium-2-carboxylate (2.4 mmol, 8 equiv) in dioxane portion-
wise during 2 h via Pasteur pipet (CAUTION: when dry,
benzenediazonium-2-carboxylate detonates violently on being
1
scraped or heated). The reaction was monitored by H NMR
by removing aliquots, evaporating the dioxane, and dissolving
the residue in CDCl3. The mixture was cooled to room
temperature and the solvent evaporated under reduced pres-
sure. The corresponding crude product was purified by flash
column chromatography (5% EtOAc/hexanes) to give 65 mg
(43%) of 34 (tentative structure assignment) as a pale yellow
oil, which was crystallized from a mixture of Et2O/hexanes as
a yellowish solid: mp 328-329 °C; 1H NMR (CDCl3, 400 MHz)
δ 7.75-7.41 (m, 8 H), 7.28-7.24 (m, 1 H), 7.13-6.91 (m, 8 H),
6.69-6.48 (m, 6 H), 6.24-6.18 (m, 1 H), 5.60 (s, 1 H), 5.53
(dd, J ) 8.2, 2.2 Hz, 1 H); 13C NMR (CDCl3, 100 MHz) δ 148.2,
148.1, 146.9, 145.7, 145.6, 145.5, 145.2, 145.1, 144.2, 143.5,
143.0, 136.3, 132.7, 132.0, 130.5, 130.2, 129.7, 129.1, 128.4,
127.7, 125.6, 125.3, 124.6, 124.4, 124.0, 123.5, 123.3, 123.2,
123.1, 123.0, 122.9, 122.8, 122.6, 122.3, 121.6, 60.9, 59.7, 55.3,
54.4; IR (film) 3050-3011, 2955-2854, 1457, 916, 740 cm-1
;
LRMS (EI, 70 eV) m/z 507 ([M + H]+, 3), 506 (M+, 7), 86 (34),
85 (56), 84 (57), 83 (100); HRMS (EI) calcd for C40H26 506.2034,
found 506.2034.
Ack n ow led gm en t. J .P.S. acknowledges Xunta de
Galicia and University of Santiago de Compostela
(Spain) for a Postdoctoral Fellowship. We thank Dr.
Chun Zhou for assistance with the spin polarization
transfer measurements and Dr. Matthew Bratcher and
Prof. W. H. Laarhoven7,10,12 for valuable exchanges of
information.
J O9723218