balances 1 and 2 provides a direct measure of strength of
the intramolecular CHꢀπ interaction.
CDCl3 (ΔΔG = 0.45ꢀ1.40 kcal/mol). An analogous
analysis with ethoxy balances 1b and 2b yielded a value
of ꢀ1.04 kcal/mol. The magnitude of the CHꢀπ interac-
tion also compares favorably to previous measurements by
Wilcox of ꢀ0.44 kcal/mol for an intramolecular alkyl
CHꢀπ interaction in CDCl3.4
The strengths of the CHꢀπ interactions were measured
by monitoring the folded/unfolded conformational equili-
brium by 1H NMR. Due to restricted rotation around the
CarylꢀNimide single bonds, the folded and unfolded confor-
mations wereinslowexchangeatrt9 and separate peaksfor
the alkoxy groups in each conformation were observed in
Although the CHꢀπ interactions in 1a and 1b were
attractive, the folded conformers were still not the major
conformers (Keq < 1.0). Wehypothesizedthatthis was due
to an opposing repulsive lone pairꢀπ interaction between
the ether oxygen linkers and the arene surfaces. Although
the lone pairꢀπ interaction was considered to be attractive
by many researchers,12 our result showed that it can be
repulsive. To measure the strength of the repulsive inter-
action, control balance 3a was prepared, which lacked an
aromatic surface. Therefore, the Keq of 3a provided a
measure of the intrinsic conformational bias of the N-
arylimide framework in the absence of the attractive
CHꢀπ and the repulsive oxygenꢀarene interaction. As
expected, Keq of 3a was close to unity (0.73). The slight bias
for the unfolded conformer was attributed to differences in
dipole and solvation energy of the conformers. The
ΔΔGfold for 2a and 3a was þ1.22 kcal/mol. This repulsive
oxygenꢀπ interaction was slightly larger than the attrac-
tive CHꢀπ interactions in 1a and 1b, providing an ex-
planation for the overall bias for the unfolded conformers
in both balances.
1
the H NMR spectra. For phenanthrene balances 1aꢀg
that form intramolecular CHꢀπ interactions, large upfield
shiftsof 1.4ꢀ1.6 ppm were observed for the alkoxyprotons
in the folded conformers due to the proximity of the arene
shelf. In contrast, only small upfield shifts were observed
for the folded conformers of control balances 2 (0.1ꢀ
0.3 ppm) and 3 (0.01 ppm).
Table 1. Comparison of the Folded/Unfolded Ratios and ΔGfold
Values for the One-Armed Balances As Measured by 1H NMR
Integration, in CDCl3 at 23 °C
alkoxy-
arm
arene-
shelf
Keq
ΔGfold
balances
[folded]/[unfolded] (kcal/mol)
1a
1b
1c
1d
1e
2a
2b
3a
OMe
OEt
phenanthrene
phenanthrene
phenanthrene
0.46
0.20
0.45
0.94
Oi-Pr
<0.05a
0.13
>1.8
On-Bu phenanthrene
Oc-Hex phenanthrene
1.2
<0.05a
0.09
>1.8
Next, balances 1aꢀe with alkoxy arms (OMe, OEt, Oi-
Pr, On-Bu, and Oc-Hex) of varying lengths and widths
were prepared and studied (Table 1). In general, larger
alkoxy arms appeared to weaken the intramolecular
CHꢀπ interactions. This trend could be explained for
the branched Oi-Pr and Oc-Hex groups in 1c and 1d.
Modeling showed that these secondary alkoxy groups
create significant steric strain in the folded conformation.
One of the two alkyl groups attached to the branch point
was always pressed into the arene shelf. The destabilization
of the balances with the longer linear alkoxy arms 1b (OEt)
and 1d (On-Bu) was more difficult to explain. X-ray and
molecular modeling studies predicted that 1a, 1b, and 1d
should have similar folding energies because (1) they all
formone strong CHꢀπ interaction between the protons on
the carbon attached to the ether oxygen and the phenan-
threne surface and a second weaker CHꢀπ interaction for
the longer OEt and On-Bu balances13 and (2) the more
flexible linear alkoxy groups can adopt conformations that
minimize any destabilizing steric interactions. A possible
explanation was that the observed differences in ΔGfold
were due to differences in conformational entropy (ΔS) of
the alkoxy arms.
OMe
OEt
benzene
benzene
ethene
1.40
0.036
0.73
1.96
0.18
OMe
a For balances 1c and 1e, only the unfolded conformer was observed,
and thus a maximum folded/unfolded ratio of 0.05 was estimated based
on a 1H NMR integration accuracy of (2%.
The folding propensities of the balances 1aꢀe and
control balances 2aꢀb and 3a were measured. Integration
of the peaks for the respective conformers yielded the
folded/unfolded ratios and ΔGfold values (Table 1).10 The
singlets corresponding to two syn-protons on the succini-
mide rings of the balances provided the most accurate
folded/unfolded ratios as they fell in a clear region of the 1H
NMR spectra (4.2ꢀ4.8 ppm) and were well differentiated
in most solvents. The conformers with aromatic shelves
(1aꢀe and 2aꢀb) were assigned by the upfield shifts of the
alkoxy protons in the folded conformers.11
The differences in the folding energies (ΔΔG) of balances
1 and 2 that can and cannot form CHꢀπ interactions,
respectively, provide a measure of the CHꢀπ interactions.
Therefore, the ΔΔG for methoxy balances 1a and 2a yields
an estimate of ꢀ0.95 kcal/mol for the CHꢀπ interaction in
(9) For example, the rotational barrier about the CarylꢀNimide bond
in balance 1a was measured to be 20.5 kcal/mol by VT NMR method
(coalescence temp = 135 °C). This equates to a half-life of 1.4 min at
23 °C.
(10) To verify that aggregation did not also attenuate the folded/
unfolded ratio, the Keq of balance 1a was measured over a wide
concentration range. The folded/unfolded ratio remained constant from
1.9 to 17 mM in CDCl3, confirming that aggregation did not affect the
folded/unfolded ratio.
(12) (a) Schneider, H. J.; Werner, F.; Blatter, T. J. Phys. Org. Chem.
1993, 6, 590–594. (b) Egli, M.; Sarkhel, S. Acc. Chem. Res. 2007, 40, 197–
205. (c) Gung, B. W.; Zou, Y.; Xu, Z. G.; Amicangelo, J. C.; Irwin,
D. G.; Ma, S. Q.; Zhou, H. C. J. Org. Chem. 2008, 73, 689–693. (d)
Korenaga, T.; Tanaka, H.; Ema, T.; Sakai, T. J. Flourine Chem. 2003,
122, 201–205. (e) Annunziata, R.; Benaglia, M.; Cozzi, F.; Mazzanti, A.
Chem.;Eur. J. 2009, 15, 4373–4381. (f) Benaglia, M.; Cozzi, F.;
Mancinelli, M.; Mazzanti, A. Chem.;Eur. J. 2010, 16, 7456–7468.
(13) The second CHꢀπ interaction is near the edge of an arene shelf,
where it is much less electron-rich than the center of the shelf; thus it is
not as strong as the first CHꢀπ contact.
(11) For balance 3a, the conformers were assigned by NOEs between
the methyl ether and vinyl protons in the folded conformer.
4322
Org. Lett., Vol. 13, No. 16, 2011