A high-barrier molecular balance for studying face-to-face arene-arene interactions in the solid state and in solution

Article abstract of DOI:10.1002/chem.200900479

An atropisomeric molecular balance was developed to study faceto-face arene-arene interactions. The balance has a large central 1,4,5,8naphthalene diimide surface that forms intramolecular arene-arene interactions with two pendent arms. The balance adopts

Full text of DOI:10.1002/chem.200900479

FULL PAPER
DOI: 10.1002/chem.200900479
A High-Barrier Molecular Balance for Studying Face-to-Face Arene–Arene
Interactions in the Solid State and in Solution
Yong S. Chong, William R. Carroll, William G. Burns, Mark D. Smith, and
Ken D. Shimizu*[a]
Abstract: An atropisomeric molecular
balance was developed to study face-
to-face arene–arene interactions. The
balance has a large central 1,4,5,8-
naphthalene diimide surface that forms
intramolecular arene–arene interac-
tions with two pendent arms. The bal-
ance adopts distinct syn and anti iso-
mers with varying numbers of intramo-
lecular interactions. Thus, the strength
of the arene–arene interaction could be
quantitatively measured by NMR spec-
troscopy from the anti/syn ratios. The
size of the arene arms was easily
varied, which allowed examination of
the relationship between arene size
and strength of the interaction. A non-
linear size dependence was observed in
solution with larger arene arms having
a disproportionately stronger arene–
arene interaction. The intramolecular
arene–arene interactions were also
characterized in the solid state by X-
ray crystallography. These studies were
facilitated by the kinetic stability of the
syn and anti isomers at room tempera-
ture due to the high isomerization bar-
rier (DG=27.0 kcalmolꢀ1). Thus, the
anti isomer could be selectively isolated
and crystallized in its folded conforma-
tion. The X-ray structures confirmed
that the anti isomers formed two strong
intramolecular arene–arene interac-
tions with face-to-face geometries. The
solid-state structure analysis also re-
veals that the rigid framework may
contribute to the observed nonlinear
size trend. The acetate linker is slightly
too long, which selectively destabilizes
the balances with smaller arene arms.
The larger arene arms are able to com-
pensate for the longer linker and form
effective intramolecular arene–arene
interactions.
Keywords: arene–arene interac-
tions · atropisomerism · molecular
torsional balance
·
noncovalent
interactions · pi interactions
Introduction
of similar magnitude such as charge transfer, solvation, elec-
trostatic interactions, and London dispersion forces.[15] Thus,
the dominant term will vary depending upon an areneꢀs sub-
Arene–arene interactions[1,2] play an important role in con-
trolling the structure and reactivity of biological[3] and syn-
thetic molecules[4] and polymers.[5–8] For example, arene–
arene interactions assist in directing the folding and assem-
bly of proteins and polynucleotides.[9] Arene–arene interac-
tions have also been used to assemble supramolecular struc-
tures[10–12] and to control the stereoselectivity of organic
transformations.[13,14] Despite their importance, the factors
that govern arene–arene interactions are not well under-
stood, in contrast to other noncovalent interactions such as
hydrogen bonding. There are two major reasons: the first is
that arene–arene interactions are the sum of multiple forces
stituents,[16–19] size,[20] and solvent environment.[21,22]
A
second difficulty in studying arene–arene interactions is
their geometric promiscuity, which arises from the lack of a
strong electrostatic polarization in most arene surfaces.
Arene surfaces can adopt a variety of geometries with simi-
lar stabilities.[23] The most common geometries are shown in
Figure 1. In addition to the face-to-face and edge-to-face ge-
ometries, there are also a continuum of intermediate geome-
tries.[24–27]
[a] Dr. Y. S. Chong, W. R. Carroll, W. G. Burns, M. D. Smith,
Prof. K. D. Shimizu
Department of Chemistry, University of South Carolina
631 Sumter St., Columbia SC 29208 (USA)
Fax : (+1)803-777-9521
E-mail: shimizu@mail.chem.sc.edu
Figure 1. Three general types of arene–arene geometries.
Chem. Eur. J. 2009, 15, 9117 – 9126
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9117

Thus, we have designed a new molecular balance 1
(Scheme 1) to study face-to-face arene–arene interactions.
The two arene arms of 1 form intramolecular face-to-face
Figure 2. Examples of bimolecular model systems by a) Rebek et al.[32]
and b) Hunter et al.[33,34] for measuring the strength of arene–arene inter-
actions.
of bimolecular arene–arene complexes as the arene–arene
interaction alone is too weak and nondirectional. This intro-
duces the problem of subtracting out the entropic and en-
thalpic contributions of these supporting interactions.[35] The
hydrogen bonds also limit the analysis of these systems to
non-hydrogen bonding organic solvents.
Scheme 1. Schematic representation of molecular balance 1 (R=tert-
amyl), which is designed to measure face-to-face arene–arene interac-
ꢀ
tions. Restricted rotation around the two Caryl Nimide single bonds leads to
room temperature stable syn and anti isomers that contain one and two
The second class of arene–arene model systems are uni-
intramolecular arene–arene interactions, respectively.
AHCTUNGTREGmNNUN olecular systems that measure intramolecular arene–arene
interactions. Unimolecular systems use a rigid covalent
framework to bring two arene surfaces together. Thus, uni-
AHCTUNGTREGmNNUN olecular systems can, in principle, provide better control
over the arene–arene geometry and can be studied in a
wider range of solvent environments. Examples include Sie-
gelꢀs 1,8-biarylnaphthalenes (Figure 3a),[36] Rashkin–Watersꢀ
interactions with a large 1,4,5,8-naphthalene diimide surface.
In the unfolded syn isomer, only one arene arm can form an
intramolecular interaction. In the folded anti isomer, both
arene arms can form intramolecular interactions. Thus, the
anti/syn ratio yields a measure of the arene arm to naphtha-
lene diimide interaction. Molecular balance 1 was efficiently
assembled in two steps. Thus, arene arms of varying size
from benzene to pyrene could be readily incorporated and
their intramolecular interactions measured. We were partic-
ularly interested in how the stability of the arene–arene in-
teraction varied with the size of the arene surface. Recently,
theoretical studies by Grimme have predicted that the
strength of the arene–arene interaction will not scale linear-
ly with arene surface area.[20]
There are a number of advantages in using small-molecule
model systems to study weak, noncovalent forces.[4] For ex-
ample, arene–arene interactions were initially studied in bio-
logical systems such as proteins, peptides, and nucleo-
tides.[9,19,28] However, the contributions of individual arene–
arene interactions were often difficult to accurately extricate
from the myriad of other noncovalent interactions that are
present in biological recognition systems.[29,30] Thus, small-
molecule model systems have been developed to study
arene–arene interactions in simpler and more controlled en-
vironments.[2,31] These model systems can be classified as
either bimolecular or unimolecular systems, where the non-
covalent interactions are inter- or intramolecular. Examples
of bimolecular model systems by Rebek[32] and Hunter[33,34]
and their groups are shown in Figure 2, where the two inter-
acting arene surfaces are on separate molecules. Hunterꢀs bi-
molecular model system, in particular, provided conforma-
tion of the widely used Hunter–Sanders electrostatic model
of the arene–arene interaction.[2] Both systems use support-
ing hydrogen-bonding interactions to ensure the formation
Figure 3. Four unimolecular model systems for studying arene–arene in-
teractions. See text for details.
benzylpyridium,[19] Gellmanꢀs propargylic amides (Fig-
ure 3b),[37] Gungꢀs triptycenes (Figure 3c),[17,38] and Camm-
ers-Goodwinꢀs three-state benzylpyridinium system (Fig-
ure 3d).[39] Each system adopts multiple conformations due
to restricted rotation around a single bond. Measurement of
the rate of isomerization or the equilibrium ratio yields a
measure of the intramolecular arene–arene interaction.
One of the most successful examples of an unimolecular
model system is Wilcoxꢀs molecular torsional balance, which
9118
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9117 – 9126

Face-to-Face Arene–Arene Interactions
FULL PAPER
was designed to study edge-to-face arene–arene interactions
(Scheme 2).[40] This molecular balance is in equilibrium be-
tween folded and unfolded conformers due to restricted ro-
Results and Discussion
Synthesis: One of the most attractive features of molecular
balance 1 is its efficient two-step synthesis (Scheme 3). This
allowed us to rapidly synthesize balances 1a–d with ben-
tation about the Caryl Caryl bond.[41] In the folded conformer,
ꢀ
Scheme 2. Wilcoxꢀs two-state molecular torsional balance for measuring
edge-to-face arene–arene interactions.[42]
the rigid bicyclic framework enforces an intramolecular
edge-to-face arene–arene interaction. In the unfolded con-
former, this edge-to-face interaction is broken. Studies of
Wilcoxꢀs molecular balance have expanded our understand-
ing of the nature of arene–arene interactions, highlighting
the utility of small-molecule model systems. Wilcoxꢀs bal-
ance has also inspired many computational studies leading
to opposing theories on the nature of the arene–arene inter-
actions.[42] Initial computational studies of this system led to
the conclusion that dispersion forces are a much stronger
component of the edge-to-face interactions than was previ-
ously understood.[43] Alternatively, the folding propensities
have been attributed to solvent accessibility and the solvo-
phobic effects.[44,45]
Scheme 3. Synthesis of atropisomeric molecular torsional balances 1a–d
with arene arms of varying size.
zene-, naphthylene-, biphenyl and pyrene arms. First, the
bisACTHNUGRTENUNG(phenol) 2 was prepared as previously described through
the thermal condensation of two equivalents of 2-amino-4-
tert-amylphenol with 1,4,5,8-naphthalene diimide.[50] Bis-
AHCTUNGTREG(NNNU phenol) 2 was alkylated with an a-bromoacylarene (3a–d)
under basic conditions to yield balances 1a–d as a mixture
of syn and anti isomers. The anti/syn ratios of balances 1a–d
were dependent on the anti/syn ratio of the starting materi-
Molecular balance 1 incorporates many of the attractive
features of Wilcoxꢀs balance including a rigid framework
and well-defined folded and unfolded conformations. A key
difference is that balance 1 is designed to study face-to-face
as opposed to edge-to-face arene–arene interactions. In ad-
dition, the analysis of balance 1 is facilitated by the room
temperature stability of the syn and anti isomers.[46–49] In so-
lution, the kinetic stability of the isomers enabled rapid and
accurate measurement of the anti/syn ratios by integrating
al, bisACHTNUTRGNE(NUG phenol) 2, which varied greatly due to the low rota-
tional barrier of 2.[51] Therefore, the crude balances were
heated in tetrachloroethane (TCE) at 808C for 10 h prior to
purification in order to achieve consistent anti/syn ratios. In
the cases of balances 1b and 1c, the syn isomer was isolated
by crystallization. In the case of balance 1d, the syn and anti
isomers were separated by flash chromatography.
The syn and anti isomers were assigned by comparison of
their NMR spectra with anti-enriched samples of 1a–d,
1
the H NMR spectra as the isomers were in slow exchange.
Balance 1 could also be equilibrated in one solvent, and
then the anti/syn ratio measured in another. This was helpful
because each balance could be analyzed in its optimal NMR
solvent. The kinetic stability of the anti and syn isomers was
also helpful in characterizing the arene–arene interactions
of 1 in the solid state. It is often difficult to study intramo-
lecular interactions in the solid state due to the dominance
of intermolecular packing forces.[4] Thus, molecular balances
often prefer to crystallize in an unfolded conformation.[48] In
the case of balance 1, the syn and anti isomers could be sep-
arated and individually crystallized. Of particular interest
were the folded anti isomers of 1 that contained two sym-
metrical arene–arene interactions. Crystallization of anti-1b
through anti-1d and characterization by X-ray crystallogra-
phy confirmed the presence of face-to-face arene–arene in-
teractions.
which were prepared by alkylation of anti-bis
The anti-bis(phenol) 2 was prepared as previously described
by heating neat samples of bis
(phenol) 2 (24 h, 1258C).[51]
The anti-enriched balances maintained their anti/syn ratios
(typically >90% anti isomer) as long as they were not
heated.
ACHTUNGTRENUN(NG phenol) 2.
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
Restricted rotation: In balances 1a–d, restricted rotation
was observed at room temperature. The syn and anti iso-
mers were in slow exchange with distinct peaks in their
1H NMR spectra. An example is shown in Figure 4 for the
biphenyl balance 1c. The a-keto protons (Hb) and the aro-
matic naphthalene diimide protons (Ha) were the most
clearly resolved in [D6]benzene (see Figure 4 for assign-
ment). These protons were singlets at room temperature,
and thus, could be easily and accurately integrated to mea-
Chem. Eur. J. 2009, 15, 9117 – 9126
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9119

K. D. Shimizu et al.
the smallest benzene balance 1a, which were near unity
(45:55). The highest ratios (86:14) were measured for the
largest pyrene balance 1d.
Control studies were carried out to rule out other possible
influences on the anti/syn ratios. First, the effect of intermo-
lecular aggregation was assessed by varying the balance con-
centration from 0.5 mm to 5.0 mm. The anti/syn ratios of bal-
ances 1a–d remained constant over this concentration
range. Second, the influence of the differences in dipole
moment of the syn and anti isomers was examined by equili-
bration of the balances in dimethyl sulfoxide (DMSO), a
more polar solvent. The anti/syn ratios measured in DMSO
were very similar to those obtained in TCE (Table 1). This
suggests that the differences in the molecular dipoles of the
syn and anti isomers does not have a strong influence on the
anti/syn ratios. For comparison, we have studied similar
N,N-diarylimide atropisomeric systems where there are
large differences in the dipole moments of the syn and anti
isomers.[54] In those cases, the anti/syn ratios were very sensi-
tive to solvent polarity. The solvent study also helped to
eliminate the solvophobic effect as a major contributor to
the anti/syn ratios.[55]
Figure 4. 1H NMR spectrum of 1c in [D6]benzene. Distinct signals were
observed for syn- and anti-1c, which were in slow exchange at room tem-
perature.
sure the anti/syn ratios. Not only were the syn and anti iso-
mers stable on the NMR timescale, they also were kinetical-
ly stable. This was evident from analysis of anti-enriched
samples of 1, which maintained steady anti/syn ratios for
days at room temperature.
The isomerization barrier was measured by heating anti-
enriched 1 at 808C in [D2]TCE and monitoring the change
in the anti/syn ratios over the course of several hours by
1H NMR. The isomerization kinetics were consistent with a
From the anti/syn ratios, the relative energies DG of the
intramolecular arene–arene interactions can be quantitative-
ly estimated from Equation (1):
first order process. A plot of ln
G
ACHTUNGTREN(NUGN R+1)] versus time
yielded a straight line, where R and Re are the anti/syn
ratios at time t and at equilibrium, respectively.[52] From the
slope of the plot, a first-order rate constant of 2.37ꢂ10ꢀ4 sꢀ1
was measured for balance 1b, which equates to an isomeri-
zation barrier of 27.0 kcalmolꢀ1 calculated with the Eyring
equation.[53] The Eyring equation was also used to estimate
a half-life of 55 days at 258C, which is consistent with the
observation that the isomers were stable for extended peri-
ods of time at room temperature. However, the barrier was
also low enough that the equilibrium anti/syn ratio could be
achieved with gentle heating. To measure the equilibrium
anti/syn ratios, the balances were heated for 10 h at 808C.
These conditions were chosen because they corresponded to
more than ten half-lives and thus, ensured that the balance
had reached equilibrium.
DG ¼ ꢀRTln ð½antiꢁ=½synꢁÞ
ð1Þ
These are shown in Table 1 in parentheses. A plot of the
energy of the interaction versus the arene surface area as
measured by the number of carbons atoms was nonlinear
(Figure 5). The larger arene balances had disproportionately
anti/syn ratios: The equilibrium anti/syn ratios provide a
measure of the strength of the intramolecular arene–arene
interactions in balance 1. This was evident from the increas-
ing equilibrium anti/syn ratios with increasing arene surface
area (Table 1). In TCE, the lowest ratios were measured for
Table 1. The equilibrium anti/syn ratios and DG (kcalmolꢀ1) of balances
1a–d in DMSO, TCE, and neat. The anti/syn ratios were measured by in-
tegration of the H NMR spectra after equilibration of the balances in so-
Figure 5. Plot of the ꢀDG (kcalmolꢀ1) of the arene–arene interactions in
solution for balances 1a–d versus size of the arene surfaces as measured
by the number of carbons.
1
lution and in the solid states (808C, 10 h).
Solvent
TCE
1a
1b
1c
1d
45:55
T
53:47
T
61:39
T
86:14
ACHTUNGTRENNUNG
stronger arene–arene interactions than the smaller arene
balances. This trend is consistent with recent computational
studies by Grimme that predicted a similar nonlinear rela-
tionship between the magnitude of the arene–arene interac-
tion and the arene size.[20] The origin of this trend was attrib-
DMSO
neat
N
N
N
ACHTUNGTRENNUNG
T
T
T
ACHTUNGTRENNUNG
9120
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9117 – 9126

Face-to-Face Arene–Arene Interactions
FULL PAPER
uted to the enhanced dispersion terms of larger, more polar-
izable arene surfaces.
highest anti/syn ratios were observed for the larger arene
surfaces (1c and 1d). However, the changes in ratio were
very abrupt as the anti/syn ratios were either about 50:50 or
99:1. This suggests that a combination of intramolecular and
intermolecular (packing effects) interactions were operative
in the solid state.
However, the overall magnitude of the arene–arene inter-
actions in balance 1 was significantly smaller than expected.
For example, 1a with a benzene surface had a very small
DG of folding, and 1b with the naphthalene surface has a
DG of about 0.1 kcalmolꢀ1. For comparison, with other
Crystallographic studies allowed us to measure and com-
pare the geometries and orientations of the intramolecular
arene–arene interactions in balances 1b–d. These analyses
were facilitated by the room temperature stability of the syn
and anti isomers. Studies of conformationally flexible mole-
cules have often been hampered by the inability to selective-
ly crystallize the folded conformer of interest, as intermolec-
ular interactions often dominate in the solid state. However,
in the case of balance 1, the folded anti isomer is kinetically
stable and thus, could be separated from the syn isomer and
selectively crystallized. X-ray quality crystals of anti isomers
of balances with naphthalene (1b), biphenyl (1c), and
pyrene arms (1d) were obtained. In each case, the expected
intramolecular arene–arene interactions were observed
(Figure 6). Both arene arms form symmetrical face-to-face
model
systems
benzene–benzene
interactions
of
ꢂ1.0 kcalmolꢀ1 were measured.[38,40,48] There are a number
of possible reasons for the low estimates of the arene–arene
interactions in 1. The most important is the elevated equili-
bration temperatures. Balance 1 was heated to 808C in
order to reach equilibrium, and thus, the DG values in
Table 1 are a measure of the arene–arene interaction at
these elevated temperatures. In contrast, the higher values
for the arene–arene interactions reported for other balance
systems were all measured at room temperature or lower
temperatures.[38,40,48] Balance 1 could be heated at a lower
temperature for a longer period of time. However, this did
not change the observed anti/syn ratios. The reason is that
the anti/syn ratios correspond to equilibrium ratios at the
ꢀ
temperature at which the Caryl Nimide bonds stopped rotating
and not to the temperature the balance was heated to. A
second possible reason is that the arene arms can form at-
tractive interactions in the syn conformer. This would
reduce the difference in stability between the isomers and
lead to an apparently smaller arene–arene interaction. The
possibility of repulsive interactions between the arene arms
in the syn isomer was also considered. However, repulsive
interactions were deemed unlikely due to the ability of the
syn isomer to avoid unfavorable interactions because of the
flexibility of the ether linkers. Finally, both interacting arene
surfaces in the balances are electron poor, and the Hunter–
Sanders electrostatic arene–arene model predicts that these
electrostatically similar arene surfaces will form weaker
arene–arene interactions.[56] The 1,4,5,8-naphthalene diimide
surface is highly electron deficient due to the four imide car-
bonyl substituents. Likewise, the arene arms of balance 1
are electron poor due to their acetyl linkers.
Solid-state analysis: Next, balance 1 was characterized in
the solid state. These studies established that the intramolec-
ular arene–arene interactions were maintained in the solid
state. This allowed us to precisely characterize the face-to-
face geometry of the arene surfaces by using X-ray crystal-
lography. First, the equilibrium anti/syn ratios were mea-
sured in the solid state (Table 1). The isomerization barriers
in the solid state were similar to the isomerization barriers
in solution. Thus, the neat balances were equilibrated under
similar conditions (10 h at 808C). The solids were cooled to
room temperature and then dissolved in a NMR solvent and
their anti/syn ratios were measured by integration of the
NMR spectra. The measured anti/syn ratios were qualita-
tively similar to those obtained in solution, suggesting that
the intramolecular arene–arene interactions were main-
tained in the solid state. The lowest anti/syn ratios were ob-
served for the smaller arene surfaces (1a and 1b), and the
Figure 6. X-ray crystal structures of anti-1b, 1c, and 1d in ORTEP (left)
and space filling depictions (right). The hydrogens are not shown in the
ORTEP figures for viewing clarity.
arene–arene interactions above and below the central naph-
thalene diimide surface. The inability to crystallize the ben-
zene balance 1a maybe due to the weak intramolecular
arene–arene interactions in 1a.
The relative orientations of the interacting arene surfaces
in 1b thru 1d are shown in Figure 7. The arene arms are
shown in black and the naphthalene diimide surfaces are
shown in gray. The individual rings of the arene arm are la-
Chem. Eur. J. 2009, 15, 9117 – 9126
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9121

K. D. Shimizu et al.
arene–arene geometries.[21,59–61] First, the HOMOs of the
electron-rich arene arms were calculated, and their interac-
tions with the LUMOs of the electron-poor 1,4,5,8-naphtha-
lene diimide surface were examined (Figure 8). When the
Figure 7. Top view of the intramolecular arene–arene interactions in the
X-ray crystal structures of anti-1b, anti-1c, and anti-1d. The 1,4,5,8-naph-
thalene diimide surface is shown in gray and the top arene arm is shown
in black. The second arene arm is not shown because its orientation is
identical to the first arene arm. In the case of anti-1d, the biphenyl arene
arm from an adjacent molecule is shown in black.
beled A thru D. Only one arene arm is shown for each bal-
ance as the two arms are symmetrically positioned above
and below the central naphthalene diimide surface. The
naphthalene and pyrene arms (1b and 1d) show a strong
face-to-face interaction having almost maximal surface over-
lap with the naphthalene diimide surface. The relative posi-
tioning of the naphthalene and pyrene surfaces is almost
identical. The A and B rings of the pyrene in 1d are almost
super-imposable with the corresponding A and B rings of
the naphthalene in 1b. The arene–arene geometry in 1b is
also virtually identical to that of a crystal structure reported
by Hamilton et al. of the same two surfaces.[57] This suggests
that the acetyl linker in balance 1 holds the arene surfaces
together with a minimum of steric strain.
The influences of intermolecular packing interactions are
more evident in the X-ray structure of balance anti-1c. The
solid-state structure of anti-1c still maintains a symmetrical
folded structure with the two biphenyl arms stacked above
and below the central naphthalene diimide surface. Howev-
er, one biphenyl arm was positioned over the edge and not
the center of the naphthalene diimide. This is due to the in-
fluence of intermolecular interactions with a second biphen-
yl arm from an adjacent molecule (shown in gray), which is
stacked over the opposing edge of the naphthalene diimide
surface (Figure 7). Thus, each naphthalene diimide surface
in 1c interacts with four arene arms. The two biphenyl arms
also form intermolecular edge-to-face arene–arene interac-
tions with each other. Although both intra- and intermolec-
ular arene–arene interactions are present in the solid-state
structure of 1c, this is probably not representative of the so-
lution structure. Dilution studies confirmed that intermolec-
ular interactions do not influence the anti/syn ratios in solu-
tion. Overall, the solid-state structures of balances 1b–d
demonstrate that intermolecular arene–arene interactions
between the arene arms and the central naphthalene diimide
surface are reasonable structures and provide valuable in-
sight into the possible geometries of these interactions.
To study the origins of the solid-state arene–arene geome-
tries, the HOMOs/LUMOs and electrostatic potentials of
the respective arene surfaces were calculated by using
SPARTAN[58] at a semi-empirical level (AM1). Both of
these analyses have been used to rationalized observed
Figure 8. The calculated HOMOs of the arene arms 1b (top left), 1c
(bottom left), 1d (top right), and the LUMO of the 1,4,5,8-naphthalene
diime (bottom right). Calculations were preformed by using SPARTAN
at the semi-empirical level (AM1).[58]
HOMOs and LUMOs are fixed in the orientations observed
in the crystal structures of 1c and 1d, they show poor phase
complementarity. For example, 1b and 1d have similar ori-
entations of their arene arms in the solid state. However,
the HOMOs of the A and B rings of the naphthalene and
pyrene arms of 1b and 1d are neither in phase with each
other nor are they complementary with the LUMO of the
naphthalene diimide surface.
The observed arene–arene orientations in 1b, 1c, and 1d
show better correlation with the electrostatic surface poten-
tials of the arene surfaces (Figure 9). The electron-rich re-
gions are highlighted in red and electron-poor regions are
highlighted in blue. The most electron-rich region of the
Figure 9. Electrostatic potential surfaces of the arene arms of 1b, 1c, and
1d and the central 1,4,5,8-naphthalene diimide surface (bottom right) cal-
culated by using SPARTAN at a semi-empirical level (AM1).[58] The elec-
tron-rich regions are highlighted in red and electron-poor regions are
highlighted in blue.
9122
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9117 – 9126

Face-to-Face Arene–Arene Interactions
FULL PAPER
three arene arms is the center of each surface. The most
electron-poor regions of the naphthalene diimide are over
shows the influence of the tether length (Table 2). The dis-
tances for rings adjacent to the tether are slightly beyond
the optimal stacking distances of 3.5 ꢃ (Figure 10). Rings
ꢀ
the carbonyl carbons and the aromatic C H groups. The
arene–arene orientations observed in balances 1b–d appear
to yield optimal overlap between the electron-rich regions
of the arene arms and the electron-poor regions of the naph-
thalene diimide surface. For example, the folded structure of
anti-1d maximizes the interaction between the electron-rich
center of the pyrene surface and the electron-poor naphtha-
lene diimide surface. Similarly, the off-set geometry in bi-
phenyl balance 1c places the electron-rich centers of the bi-
phenyl surfaces over the electron-poor edges of the naph-
thalene diimide surfaces. This correlation between geometry
and electrostatic complementarity is consistent with Iverson
et al.ꢀs analysis of the geometry of the arene–arene complex
of 1,4,5,8-naphthalene diimide and naphthalene surfaces.[21]
Comparison of the potential surfaces of the arene arms
also provided an electrostatic basis for the stronger intramo-
lecular interactions in pyrene balance 1d. First, the pyrene
surface is significantly more electron-rich than either the
naphthalene or biphenyl surfaces, leading to stronger elec-
trostatic or dispersion interactions with the electron-poor
naphthalene diimide surface. This is evident from the large
red regions of the 1-acetylpyrene in comparison to the 2-
acetylnaphthalene or 4-acetylbiphenyl potential surfaces
(Figure 9). The lower ionization potential of the pyrene arm
appears to be primarily due to its larger, more polarizable
surface.[20] A second factor that differentiates the pyrene
arm from the other arene arms is the influence of the elec-
tron-withdrawing acyl tether. In the case of the naphthalene
and biphenyl surfaces, molecular modeling predicts that the
acetyl group will be nearly coplanar with the arene surface.
This conformation maximizes the electron-withdrawing in-
fluence of the acyl group and reduces the electron-donating
ability of these smaller arene surfaces. However, in the case
of the pyrene surface, the acetyl group is twisted out of con-
jugation with the arene surface due to the steric interactions
between the carbonyl oxygen and the pyrene C10 proton.
This deplanarization reduces the electron-withdrawing effect
of the acetyl tether making the pyrene arm considerably
more electron-rich than the other arene arms. The twisted
configuration of the pyrene arm was confirmed by examina-
tion of the X-ray structures. In pyrene balance anti-1d, the
acetyl group was twisted 39.38 out of the arene plane. In
contrast, the acetyl groups in anti-1b and anti-1c were twist-
ed only 23.48 and 26.38 out of plane.
Table 2. Centroid to plane distances [ꢃ] between the individual rings of
the arene arms and the 1,4,5,8-naphthalene diimide surface in the crystal
structures of balances 1b, 1c, and 1d. The rings of the arene arms are as-
signed as A, B, C, and D as shown in Figure 7.
Ring
A
B
C
D
1b
1c
3.67
3.46
–
–
–
–
4.16
3.31
4.37[a]
3.79
5.21[b]
3.55
1d
3.78
3.93
[a] Ring A’. [b] Ring B’.
further from the tether have
shorter distances (<3.5 ꢃ) and
therefore form stronger arene–
arene interactions. For example,
the ring attached to the tether
(A ring) in naphthalene balance
1b is 3.67 ꢃ from the plane of
the naphthalene diimide sur-
face, and the ring furthest from
the tether (B ring) has a dis-
tance of 3.46 ꢃ. Thus, the
larger arene arms can form
Figure 10. A representation of
the correlation between arene
arm length and the observed
distance to the plane of the
1,4,5,8-naphthalene diimide.
stronger arene–arene interactions because the rings further
from the tether can form more effective arene–arene inter-
actions. This may also explain why the overall magnitudes of
the arene–arene interactions were relatively small compared
to other balance systems. For example, balance 1a with only
a single ring (A ring) cannot form effective intramolecular
arene–arene interactions. This may explain the inability to
crystallize anti-1a as it would not prefer a folded structure.
Conclusions
In summary, we have successfully synthesized four new mo-
lecular balances for studying face-to-face arene–arene inter-
actions. Due to their efficient syntheses, balances with arene
arms of varying size (benzene, naphthalene, biphenyl, and
pyrene) could be readily prepared and studied. The balances
adopt distinct syn and anti isomers that have differing num-
bers of intramolecular arene–arene interactions. Therefore,
the anti/syn ratio provides a measure of the intramolecular
arene–arene interactions. A unique feature of these balances
was the kinetic stability of their syn and anti isomers at
room temperature. This stability facilitated the accurate
measurement of the strength of the arene–arene interactions
in solution and also the conformational analysis of the
arene–arene interaction in the solid state. First, the syn and
anti isomers were in slow exchange and thus, the anti/syn
ratio could be readily measured by integration of the
1H NMR spectra. Second, the stability of the isomers en-
Arene–arene distances: Finally, the intramolecular arene–
arene distances in the crystal structures provide an addition-
al explanation for the stronger arene–arene interaction ob-
served for pyrene balance 1d. This analysis suggests that the
acyl tether is slightly too long, which destabilizes the folding
of the smaller arene balances. The effects are subtle and dif-
ficult to see in the side views of the balances shown in
Figure 6. However, comparison of the intramolecular cen-
ACHTUNGTRENNUNG
Chem. Eur. J. 2009, 15, 9117 – 9126
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9123

K. D. Shimizu et al.
ture for 6 h. The reaction mixture was added to a solution of 1n HCl
(25 mL) resulting in the formation of a precipitate. The precipitate was
separated through filtration and dried in vacuo to give crude compound
1a. The crude product was heated in TCE at 808C for 10 h, yielding 1a
as a light brown solid (130 mg, 84%) with an anti/syn ratio of 45:55.
1H NMR (300 MHz, CDCl3, 258C, TMS): d=8.75 (s, 4H-anti), 8.74 (s,
4H-syn), 7.80–7.74 (m, 4H-anti and 4H-syn), 7.45–7.16 (m, 10H-anti and
10H-syn), 7.05 (d, J=8.8 Hz, 2H-syn), 7.03 (d, J=8.8 Hz, 2H-anti), 5.18
(s, 4H-anti and 4H-syn), 1.66 (q, J=7.4 Hz, 4H-anti and 4H-syn), 1.32 (s,
12H-anti), 1.31 (s, 12H-syn), 0.77 ppm (t, J=7.4 Hz, 6H-anti and 6H-
syn); 13C NMR (125 MHz,CDCl3, 258C, TMS): d=194.1, 162.7, 151.2,
143.8, 134.5, 133.4, 131.1, 131.0, 128.5, 128.4, 128.1, 128.0, 127.8, 127.3,
126.9, 123.6, 133.4, 72.1, 72.0, 37.6, 37.0, 28.4, 9.2 ppm; IR (liquid film):
n˜ =697, 734, 767, 825, 982, 1127, 1198, 1224, 1249, 1346, 1448, 1508, 1582,
1604, 1679, 1716, 2963, 3058 cmꢀ1; HRMS (ESI) m/z: calcd for [M+H]+:
C52H46N2O8 826.3254; found 826.3268.
ACHTUNGTRENNUNGabled the selective crystallization of anti-1, which main-
tained its folded structure in the solid state. Thus, the distan-
ces and orientations of the arene surfaces could be precisely
analyzed by X-ray crystallography.
A strong nonlinear free-energy trend was observed in so-
lution. Larger arene surfaces (pyrene: 1d, and biphenyl: 1c)
formed disproportionately stronger arene–arene interactions
than smaller surfaces (benzene: 1a, and naphthyl: 1b).
These results are in agreement with the recent calculations,
which predict that larger arene surfaces form stronger
arene–arene interactions due to their greater polarizabili-
ty.[20] However, a more careful structural and computational
analysis identified additional contributing factors. Specifical-
ly, the acyl tether connecting the interacting arene surfaces
imposes conformational constraints and electrostatic influ-
ences that contribute to the observed nonlinear trends. This
study demonstrates that even in simple model systems, the
study of weak, noncovalent interactions is still very challeng-
ing.[43]
Naphthalene balance 1b:
A solution of bisACHTUNREGTG(NNUN phenol) 2 (400 mg,
0.677 mmol), potassium carbonate (934 mg, 6.77 mmol), and a-bromo-
2’acetonaphthone (361 mg, 1.42 mmol) in dry DMF (15 mL) was stirred
for 12 h. The reaction mixture was added to a solution of 1n HCl
(25 mL) resulting in the formation of a precipitate. The precipitate was
separated through filtration and dried in vacuo to give crude compound
1b. The crude product was heated in TCE at 808C for 10 h, yielding 1b
as a light brown solid (535 mg, 85%) with an anti/syn ratio of 53:47.
1H NMR (300 MHz, CDCl3, 258C): d=8.07 (s, 4H-anti and 4H-syn), 7.77
(dd, J=8.5, 1.6 Hz, 2H-syn), 7.71 (dd, J=8.5, 1.6 Hz, 2H-anti), 7.58 (d,
J=2.5 Hz, 2H-syn), 7.42 (d, J=2.5 Hz, 2H-anti), 7.36–7.01 (m, 14H-anti
and 14H-syn), 6.76 (d, J=8.8 Hz, 2H-syn), 6.72 (d, J=8.8 Hz, 2H-anti),
4.83 (s, 4H-syn), 4.70 (s, 4H-anti), 1.65 (q, J=7.4 Hz, 4H-syn), 1.57 (q,
J=7.4 Hz, 4H-anti), 1.31 (s, 12H-syn), 1.22 (s, 12H-anti), 0.85 (t, J=
7.4 Hz, 6H-syn), 0.79 ppm (t, J=7.4 Hz, 6H-anti); 13C NMR
(125 MHz,CDCl3, 258C): d=195.0, 194.8, 162.4, 151.3, 151.1, 143.9, 135.2,
135.0, 132.2, 132.1, 132.0, 131.8, 130.6, 130.5, 130.3, 129.3, 128.8, 128.1,
128.0, 127.8, 127.6, 127.4, 126.9, 126.5, 126.3, 126.1, 123.9, 123.8, 123.5,
113.7, 113.5, 77.3, 77.0, 76.7, 72.9, 72.8, 37.6, 37.6, 37.1, 37.0, 28.4, 9.2, 9.1;
FTIR (film): n˜ =2963, 2927, 1714, 1678, 1626, 1582, 1508, 1469, 1446,
1346, 1250, 1191, 1125, 979, 822, 769, 757, 475 cmꢀ1; HRMS (ESI) m/z:
calcd for [M+H]+: C60H50N2O8: 926.3567; found: 926.3585; elemental
analysis calcd (%) for C60H50N2O8·H2O: C 76.25, H 5.55, N 2.96. found:
C 76.15, H 5.40, N 2.94.
Experimental Section
General methods: 1H NMR and 13C NMR spectra were obtained by
using a Varian Mercury 300 MHz NMR spectrometer and a Varian Mer-
cury 400 MHz NMR spectrometer. Peak areas were determined by using
the standard deconvolution routine in Varian VNMR 6.1B software.
Visual inspection of the difference mode spectra (simulation/data) was
used to test fit. Infrared spectra were performed on a Perkin–Elmer 1600
series FT-IR spectrometer by using sodium chloride plates. Mass spectra
were performed on a Finnigan 4521C system by using a GC probe for
volatile samples and a solid probe for nonvolatile samples. High resolu-
tion mass spectroscopy (HRMS) was performed on a VG 705Q spec-
trometer. All reagents were purchased from either Aldrich or Acros and
used as received. Elemental analysis was preformed by Atlantic Micro-
labs, Inc. The crystallographic data were collected on a Bruker SMART
APEX CCD-based diffractometer system (MoKa radiation, l=
0.71073 ꢃ). CCDC-720864 (anti-1b), CCDC-720863 (anti-1c), and
CCDC-720865 (anti-1d) contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif
Biphenyl balance 1c: A solution of bisACTHNUGTRNEUNG(phenol) 2 (250 mg, 0.423 mmol),
potassium carbonate (584 mg, 4.23 mmol), and 2-bromo-4-phenylaceto-
phenone (237 mg, 0.846 mmol) in dry DMF (10 mL) was stirred at room
temperature for 12 h. The reaction mixture was added to a solution of 1n
HCl (25 mL) resulting in the formation of a precipitate. The precipitate
was separated through filtration and dried in vacuo to give crude 1c. The
crude product was heated in TCE at 808C for 10 h, yielding 1c as a light
brown solid (410 mg, 99%) with an anti/syn ratio of 61:39. 1H NMR
(300 MHz, CDCl3, 258C): d=8.64 (s, 4H-anti), 8.63 (s, 4H-syn), 7.87 (dd,
J= 8.5, 2.7 Hz, 4H-anti and 4H-syn), 7.60 (d, J=8.2 Hz, 4H-anti and
4H-syn), 7.55–7.35 (m, 14H-anti and 14H-syn), 7.18 (d, J=8.8 Hz, 2H-
anti and 2H-syn), 5.44 (s, 4H-syn), 5.43 (s, 4H-anti), 1.64 (q, J=7.1 Hz,
4H-anti), 1.61 (q, J=6.6 Hz, 4H-syn), 1.29 (s, 12H-anti), 1.26 (s, 12H-
syn), 0.72 (t, J=7.1 Hz, 6H-anti), 0.70 ppm (t, J=7.4 Hz, 6H-syn);
13C NMR (125 MHz, CDCl3, 258C): d=194.3, 162.4, 151.1, 145.5, 143.8,
139.1, 133.3, 130.9, 129.0, 128.9, 128.5, 128.1, 127.8, 126.9, 126.8, 126.7,
123.5, 113.4, 72.3, 37.6, 37.0, 28.4, 9.2; IR (film): n˜ =2965, 2357, 1716,
Quantification of the anti/syn-ratio: The 1H NMR spectra of balances
1a–d in [D2]TCE showed two singlets in the d=4–6 ppm region. These
were assigned to the methylene protons of the acetyl linker of the syn
and anti isomers. The more upfield singlet was assigned to the anti
isomer by comparison with the 1H NMR of anti-enriched samples. The
relative magnitude of each singlet was quantified by integration by using
the peak fitting commands in the Varian VNMRJ software.
BisACHTUNGTRENNUNG(phenol) 2: anti/syn-BisACHTUNRTGEGUNN(N phenol) 2 and anti-enriched bisACHTNUGTNER(NUGN phenol) 2
were prepared as previously described.[50,51] In brief, a solution of 1,4,5,8-
naphthalenetetracarboxylic dianhydride (250 mg, 0.932 mmol) and
2-amino-4-tert-amylphenol (345 mg, 3.16 mmol) in glacial acetic acid
(8 mL) was heated at reflux for 2.5 h. The resulting light brown precipi-
tate was separated through filtration and dried in vacuo to give anti/syn-2
1683, 1604, 1582, 1506, 1448, 1346, 1249, 1210, 766, 730, 698, 668 cmꢀ1
;
HRMS (ESI) m/z: calcd for [M+H]+: C64H54N2O8: 978.3880; found
978.3842; elemental analysis calcd (%) for C64H54N2O8·H2O: C 77.09, H
5.66, N 2.81; found C 77.30, H 5.71, N 2.86.
(540 mg, 99%). anti-Enriched bisACTHNUTRGENUGN(phenol) 2 was prepared by heating a
neat sample of anti/syn-2 at 1108C for 10 h. The sample was cooled to
room temperature. H NMR ([D6]acetone) verified that the samples were
Pyrene balance 1d: A solution of bisACTHNUGRTNEUNG(phenol) 2 (100 mg, 0.169 mmol),
1
potassium carbonate (234 mg, 1.69 mmol), and 1-bromoacetylpyrene
(112 mg, 0.349 mmol) in dry DMF (5 mL) was stirred at room tempera-
ture for 2 h. The reaction mixture was added to a solution of 1n HCl
(25 mL) resulting in the formation of a precipitate. The precipitate was
separated through filtration and dried in vacuo to give crude compound
1c as a red solid (165 mg, 91%) with an anti/syn ratio of 86:14. The
>95% anti-2. The anti-enriched samples were stable at room tempera-
ture in the solid state but isomerized rapidly in solution (t1/2 ꢂ13 min).
Benzene balance 1a: A solution of bisACTHNUGTRNEUNG(phenol) 2 (110 mg, 0.186 mmol),
potassium carbonate (257 mg, 1.86 mmol) and 2-bromoacetophenone
(79.4 mg, 0.391 mmol) in dry DMF (5 mL) was stirred at room tempera-
9124
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9117 – 9126

Face-to-Face Arene–Arene Interactions
FULL PAPER
crude products were purified and separated from each other by flash
[11] M. Nishio, CrystEngComm 2004, 6, 130–158.
chromatography on silica gel (EtOAc/Hex/CH2Cl2 1:4:5) to give syn-1d
and anti-1d as red solids. syn-1d: H NMR (400 MHz, CDCl3, 258C): d=
[12] F. Klꢄrner, B. Kahlert, Acc. Chem. Res. 2003, 36, 919–932.
[13] G. Jones, Tetrahedron 2001, 57, 7999–8016.
1
8.56 (d, J=8.5 Hz, 2H), 8.19 (d, J=7.5 Hz, 2H), 8.06 (d, J=6.8 Hz, 2H),
8.04 (d, J=8.8 Hz, 2H), 7.98 (d, J=7.9 Hz, 2H), 7.97 (t, J=7.8 Hz, 2H),
7.90 (d, J=9.5 Hz, 2H), 7.75 (s, 4H), 7.47–7.43 (m, 4H), 7.33 (d, J=
8.1 Hz, 2H), 7.15 (d, J=8.8 Hz, 2H), 7.08 (d, J=2.4 Hz, 2H), 5.44 (s,
4H), 1.61 (q, J=7.4 Hz, 4H), 1.26 (s, 12H), 0.73 ppm (t, J=7.4 Hz, 6H);
13C NMR (125 MHz,CDCl3, 258C): d=199.6, 162.0, 151.2, 143.7, 133.3,
130.7, 130.1, 129.7, 129.6, 129.5, 128.1, 127.8, 126.8, 126.7, 126.6, 126.2,
125.9, 125.7, 124.2, 123.6, 123.3, 73.5, 37.5, 37.0, 28.4, 9.1 ppm; IR (film):
n˜ =2961, 2928, 2923, 2916, 2857, 2852, 2849, 2369, 2362, 2358, 2340, 1712,
[14] J. Garcꢅa Ruano, J. Aleman, I. Alonso, A. Parra, V. Marcos, J.
Aguirre, Chem. Eur. J. 2007, 13, 6179–6195.
[15] S. Tsuzuki, T. Uchimaru, Curr. Org. Chem. 2006, 10, 745–762.
[16] S. Arnstein, C. Sherrill, Phys. Chem. Chem. Phys. 2008, 10, 2646–
2655.
[17] B. Gung, X. Xue, Y. Zou, J. Org. Chem. 2007, 72, 2469–2475.
[18] S. Cockroft, J. Perkins, C. Zonta, H. Adams, S. Spey, C. M. R. Low,
J. Vinter, K. Lawson, C. J. Urch, C. Hunter, Org. Biomol. Chem.
2007, 5, 1062–1080.
[19] M. Rashkin, M. Waters, J. Am. Chem. Soc. 2002, 124, 1860–1861.
[20] S. Grimme, Angew. Chem. 2008, 120, 3478–3483; Angew. Chem. Int.
Ed. 2008, 47, 3430–3434.
[21] M. Cubberley, B. Iverson, J. Am. Chem. Soc. 2001, 123, 7560–7563.
[22] D. Smithrud, T. Wyman, F. Diederich, J. Am. Chem. Soc. 1991, 113,
5420–5426.
[23] S. K. Burley, G. A. Petsko, J. Am. Chem. Soc. 1988, 110, 7995.
[24] T. Blundell, J. Singh, J. Thornton, S. Burley, G. Petsko, Science 1986,
234, 1005–1005.
[25] W. Liu, Y. Jiang, J. Org. Chem. 2008, 73, 1124–1127.
[26] S. Tsuzuki, K. Honda, T. Uchimaru, M. Mikami, K. Tanabe, J. Am.
Chem. Soc. 2002, 124, 104–112.
[27] M. Sinnokrot, C. Sherrill, J. Phys. Chem. A 2006, 110, 10656–10668.
[28] M. S. Searle, D. Williams, Nucleic Acids Res. 1993, 21, 2051–2056.
[29] D. Williams, E. Stephens, D. OꢀBrien, M. Zhou, Angew. Chem.2004,
116, 6596–6616; Angew. Chem. Int. Ed. 2004, 43, 6760–6782.
[30] S. Cockroft, C. Hunter, Chem. Soc. Rev. 2007, 36, 172–188.
[31] R. M. Hughes, M. L. Waters, Curr. Opin. Struct. Biol. 2006, 16, 514–
524.
[32] J. Rebek, Jr., Angew. Chem. 1990, 102, 261–272; Angew. Chem. Int.
Ed. Engl. 1990, 29, 245–255.
[33] H. Adams, K. Harris, G. Hembury, C. A. Hunter, D. Livingstone,
J. F. Mccabe, Chem. Commun. 1996, 2531–2532.
[34] H. Adams, F. Carver, C. Hunter, J. Morales, E. Seward, Angew.
Chem. 1996, 108, 1628–1631; Angew. Chem. Int. Ed. Engl. 1996, 35,
1542–1544.
[35] D. Williams, M. Westwell, Chem. Soc. Rev. 1998, 27, 57–63.
[36] F. Cozzi, R. Annunziata, M. Benaglia, M. Cinquini, L. Raimondi, K.
Baldridge, J. Siegel, Org. Biomol. Chem. 2003, 1, 157–162.
[37] R. R. Gardner, S. L. Mckay, S. H. Gellman, Org. Lett. 2000, 2, 2335–
2338.
1678, 1594, 1582, 1510, 1448, 1443, 1344, 1248, 1105, 846, 768, 669 cmꢀ1
;
HRMS (ESI) m/z: calcd for [M+H]+: C72H54N2O8: 1074.3880; found
1074.3906; elemental analysis calcd (%) for C72H56N2O8: C 80.28, H 5.24,
N 2.60; found C 79.87, H 5.15, N 2.31. anti-1d: 1H NMR (400 MHz,
CDCl3, 258C): d=8.02 (d, J=9.2 Hz, 2H), 7.92 (d, J=7.1 Hz, 2H), 7.89
(d, J=7.1 Hz, 2H), 7.82 (t, J=7.4 Hz, 2H), 7.75 (d, J=7.9 Hz, 2H), 7.70
(d, J=9.3 Hz, 2H), 7.66 (d, J=8.8 Hz, 2H), 7.54 (dd, J= 8.6, 2.2 Hz,
2H), 7.47 (d, J=2.2 Hz, 2H), 7.27 (d, J=8.1 Hz, 2H), 7.25 (d, J=7.1 Hz,
2H), 7.17 (s, 4H), 7.12 (d, J=8.8 Hz, 2H), 5.12 (s, 4H), 1.81 (q, J=
7.3 Hz, 4H), 1.48 (s, 12H), 0.91 ppm (t, J=7.3 Hz, 6H); 13C NMR
(125 MHz,CDCl3, 258C): d=201.1, 161.4, 150.6, 143.6, 132.1, 131.2, 129.9,
129.7, 128.8, 128.5, 128.3, 128.2, 128.1, 127.3, 126.7, 126.4, 126.1, 126.0,
124.1, 123.8, 123.5, 123.2, 123.0, 112.9, 73.5, 37.7, 37.3, 30.9, 29.7, 28.6,
9.3 ppm; IR (film): n˜ =2961, 2928, 2923, 2916, 2857, 2852, 2849, 2369,
2362, 2358, 2340, 1712, 1678, 1594, 1582, 1510, 1448, 1443, 1344, 1248,
1105, 846, 768, 669 cmꢀ1
;
HRMS (ESI) m/z: calcd for [M+H]+:
C72H54N2O8: 1074.3880; found 1074.3906; elemental analysis calcd (%)
for C72H56N2O8: C 80.28, H 5.24, N 2.60; found C 79.87, H 5.15, N 2.31.
Measurement of the isomerization barrier: The isomerization barrier was
calculated from the experimentally measured isomerization rate constant
as previously described by using the Eyring equation.[53] The rate of iso-
merization was determined for balance system 1b. A sample of anti-en-
riched 1b (>90% anti) was heated at 808C in [D2]TCE and the anti/syn
ratio was monitored by integration of the a-keto protons of the acetyl
tether. The first-order rate constant was calculated from the slope of the
plot of ln
[(RꢀRe)/
ACHTUNGTRENNUNG(R+1)] versus time., where R and Re are the anti/syn
ratios at time t and at equilibrium, respectively.[52]
[38] B. Gung, X. Xue, H. Reich, J. Org. Chem. 2005, 70, 3641–3644.
[39] M. Sindkhedkar, H. Mulla, A. Cammers-Goodwin, J. Am. Chem.
Soc. 2000, 122, 9271–9277.
[40] S. Paliwal, S. Geib, C. S. Wilcox, J. Am. Chem. Soc. 1994, 116, 4497–
4498.
Acknowledgements
The authors would like to acknowledge the support of the National Sci-
ence Foundation (grant 0616442).
[41] J. Adrian, C. Wilcox, J. Am. Chem. Soc. 1992, 114, 1398–1403.
[42] E. Kim, S. Paliwal, C. S. Wilcox, J. Am. Chem. Soc. 1998, 120,
11192–11193.
[43] J. Ribas, E. Cubero, F. Luque, M. Orozco, J. Org. Chem. 2002, 67,
7057–7065.
[1] E. Meyer, R. Castellano, F. Diederich, Angew. Chem. 2003, 115,
1244–1287; Angew. Chem. Int. Ed. 2003, 42, 1210–1250.
[2] C. Hunter, K. Lawson, J. Perkins, C. Urch, J. Chem. Soc. Perkin
Trans. 2 2001, 651–669.
[44] K. Nakamura, K. N. Houk, Org. Lett. 1999, 1, 2049–2051.
[45] S. Cockroft, C. Hunter, Chem. Commun. 2006, 3806–3808.
[46] D. P. Curran, S. Geib, N. Demello, Tetrahedron 1999, 55, 5681–5704.
[47] S. M. Verma, N. B. Singh, Aust. J. Chem. 1976, 29, 295–300.
[48] W. R. Carroll, P. Pellechia, K. D. Shimizu, Org. Lett. 2008, 10, 3547–
3550.
[3] P. Chakrabarti, R. Bhattacharyya, Prog. Biophys. Mol. Biol. 2007,
95, 83–137.
[4] W. Jennings, B. Farrell, J. Malone, Acc. Chem. Res. 2001, 34, 885–
894.
[5] H. Colquhoun, Z. Zhu, C. Cardin, Y. Gan, M. Drew, J. Am. Chem.
Soc. 2007, 129, 16163–16174.
[49] C. Degenhardt, J. Lavin, M. Smith, K. Shimizu, Org. Lett. 2005, 7,
4079–4081.
[6] F. Ilhan, M. Gray, K. Blanchette, V. M. Rotello, Macromolecules
1999, 32, 6159–6162.
[50] Y. S. Chong, M. D. Smith, K. D. Shimizu, J. Am. Chem. Soc. 2001,
123, 7463–7464.
[51] G. Rushton, W. G. Burns, J. M. Lavin, Y. S. Chong, P. Pellechia, K.
Shimizu, J. Chem. Educ. 2007, 84, 1499–1501.
[52] W. Stewart, T. Siddall, Chem. Rev. 1970, 70, 517–551.
[53] J. Lavin, K. Shimizu, Chem. Commun. 2007, 228–230.
[54] D.-S. Choi, Y. S. Chong, D. Whitehead, K. D. Shimizu, Org. Lett.
2001, 3, 3757–3760.
[7] U. H. F. Bunz, V. Enkelmann, L. Kloppenburg, D. Jones, K. D. Shi-
mizu, J. B. Claridge, H. C. Zur Loye, G. Lieser, Chem. Mater. 1999,
11, 1416–1424.
[8] S. Lahiri, J. L. Thompson, J. S. Moore, J. Am. Chem. Soc. 2000, 122,
11315–11319.
[9] S. Burley, G. Petsko, Science 1985, 229, 23–28.
[10] D. Venkataraman, S. Lee, J. Zhang, J. S. Moore, Nature 1994, 371,
591.
Chem. Eur. J. 2009, 15, 9117 – 9126
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9125

K. D. Shimizu et al.
[55] D. Smithrud, E. Sanford, I. Chao, S. Ferguson, D. Carcanague, J.
Evanseck, K. Houk, F. Diederich, Pure Appl. Chem. 1990, 62, 2227–
2236.
[56] C. A. Hunter, J. Sanders, J. Am. Chem. Soc. 1990, 112, 5525–5534.
[57] D. Hamilton, J. Davies, L. Prodi, J. Sanders, Chem. Eur. J. 1998, 4,
608–620.
[59] S. Kato, T. Matsumoto, K. Ideta, T. Shimasaki, K. Goto, T. Shinmyo-
zu, J. Org. Chem. 2006, 71, 4723–4733.
[60] A. S. Cooray, K. De Silva, J. Mol. Struct. 2004, 678, 223–231.
[61] A. Muehldorf, D. Vanengen, J. C. Warner, A. Hamilton, J. Am.
Chem. Soc. 1988, 110, 6561–6562.
Received: February 20, 2009
[58] SPARTAN ’04, Wavefunction Inc., http://www.wavefunction.com.
Published online: July 31, 2009
9126
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9117 – 9126