Quantitative study of interactions between oxygen lone pair and aromatic rings: Substituent effect and the importance of closeness of contact

Article abstract of DOI:10.1021/jo702170j

(Figure Presented) Current models describe aromatic rings as polar groups based on the fact that benzene and hexafluorobenzene are known to have large and permanent quadrupole moments. This report describes a quantitative study of the interactions between oxygen lone pair and aromatic rings. We found that even electron-rich aromatic rings and oxygen lone pairs exhibit attractive interactions. Free energies of interactions are determined using the triptycene scaffold and the equilibrium constants were determined by low-temperature ¹H NMR spectroscopy. An X-ray structure analysis for one of the model compounds confirms the close proximity between the oxygen and the center of the aromatic ring. Theoretical calculations at the MP2/aug-cc-pVTZ level corroborate the experimental results. The origin of attractive interactions was explored by using aromatic rings with a wide range of substituents. The interactions between an oxygen lone pair and an aromatic ring are attractive at van der Waals' distance even with electron-donating substituents. Electron-withdrawing groups increase the strength of the attractive interactions. The results from this study can be only partly rationalized by using the current models of aromatic system. Electrostatic-based models are consistent with the fact that stronger electron-withdrawing groups lead to stronger attractions, but fail to predict or rationalize the fact that weak attractions even exist between electron-rich arenes and oxygen lone pairs. The conclusion from this study is that aromatic rings cannot be treated as a simple quadrupolar functional group at van der Waals' distance. Dispersion forces and local dipole should also be considered.

Full text of DOI:10.1021/jo702170j

Quantitative Study of Interactions between Oxygen Lone Pair and
Aromatic Rings: Substituent Effect and the Importance of
Closeness of Contact
Benjamin W. Gung,*,† Yan Zou,† Zhigang Xu,† Jay C. Amicangelo,*,‡ Daniel G. Irwin,‡
Shengqian Ma,†,§ and Hong-Cai Zhou†,§
Department of Chemistry & Biochemistry, Miami UniVersity, Oxford, Ohio 45056, and School of Science,
Penn State Erie, The Behrend College, Erie, PennsylVania 16563
gungbw@muohio.edu; jca11@psu.edu
ReceiVed October 9, 2007
Current models describe aromatic rings as polar groups based on the fact that benzene and hexafluo-
robenzene are known to have large and permanent quadrupole moments. This report describes a quantitative
study of the interactions between oxygen lone pair and aromatic rings. We found that even electron-rich
aromatic rings and oxygen lone pairs exhibit attractive interactions. Free energies of interactions are
determined using the triptycene scaffold and the equilibrium constants were determined by low-temperature
1H NMR spectroscopy. An X-ray structure analysis for one of the model compounds confirms the close
proximity between the oxygen and the center of the aromatic ring. Theoretical calculations at the MP2/
aug-cc-pVTZ level corroborate the experimental results. The origin of attractive interactions was explored
by using aromatic rings with a wide range of substituents. The interactions between an oxygen lone pair
and an aromatic ring are attractive at van der Waals’ distance even with electron-donating substituents.
Electron-withdrawing groups increase the strength of the attractive interactions. The results from this
study can be only partly rationalized by using the current models of aromatic system. Electrostatic-based
models are consistent with the fact that stronger electron-withdrawing groups lead to stronger attractions,
but fail to predict or rationalize the fact that weak attractions even exist between electron-rich arenes and
oxygen lone pairs. The conclusion from this study is that aromatic rings cannot be treated as a simple
quadrupolar functional group at van der Waals’ distance. Dispersion forces and local dipole should also
be considered.
Introduction
a lp-π interaction between a H2O molecule and a cytosine base
in an RNA pseudoknot.2 Very recently, Sankararamakrishnan
and co-workers have systematically analyzed 500 high-resolu-
tion protein structures (resolution e1.8 Å) and identified 286
examples in which carbonyl oxygen atoms approach the
aromatic centers within a distance of 3.5 Å.3 However, in spite
of increasing recognition of the important role played by weak
In 1995, Egli and Gessner reported the intriguing observation
that an attractive lone pair (lp)-π* interaction stabilizes the
Z-DNA structure, which explains the conformational stability
of Z-DNA at high ionic strength despite poor base pair stacking.1
More recently, a high-resolution X-ray structure analysis shows
† Miami University.
‡ The Behrend College.
(2) Sarkhel, S.; Rich, A.; Egli, M. J. Am. Chem. Soc. 2003, 125, 8998-
8999.
(3) Jain, A.; Purohit, C. S.; Verma, S.; Sankararamakrishnan, R. J. Phys.
Chem. B 2007, 111, 8680-8683.
§ To whom questions regarding X-ray structure analysis should be addressed.
(1) Egli, M.; Gessner, R. V. ′ Proc. Natl. Acad. Sci. U.S.A. 1995, 92,
180-184.
10.1021/jo702170j CCC: $40.75 © 2008 American Chemical Society
Published on Web 12/15/2007
J. Org. Chem. 2008, 73, 689-693
689

Gung et al.
FIGURE 2. X-ray structure of compound 2 displayed as a space-filling
model. The structure shows a syn conformation with the MOM group
located in close proximity of the pentafluorophenyl ring. The distance
from the methoxy oxygen to the center of the pentafluorophenyl ring
is 3.48 Å.
FIGURE 1. The equilibra among the syn and the anti conformational
isomers of 1,9-disubstituted triptycene derivative 1 (X ) H, Me, F,
CF3).
contact between the C9 arene and the C1 oxygen atom.20,21 Since
the triptycene scaffold provides an otherwise identical environ-
ment for the syn and the anti conformation, the syn:anti ratio
represents the degree of preference for the interactions between
the C1 and the C9 groups. Our previous study involved a C1
methoxy group, which was in contact with the edge of the
rotating arene in the syn conformation.21 To model the lp-π
interaction more closely to that observed in the X-ray structures
reported by Egli and by theoretical predictions,8,11,12 we needed
a model system to place the oxygen lone pair directly above
and near the center of the aromatic ring in the syn conformation.
Through molecular modeling it was found that by replacing the
C1 Me group in compound 1 with a MOM (methoxymethyl)
group, the MeO oxygen lone pairs can be positioned directly at
the center of the C9 arene in the syn conformation.
intermolecular interactions on macromolecular stability,4,5 quan-
titative study of lp-π interactions is rare.6,7 Theoretical studies
show that gas-phase lp-π interaction is favorable with electron-
deficient π-systems.8-14 Currently there is widespread interest
in anion-π interactions since this area of research may lead to
new designs of anion receptor and sensing devices. Experimental
studies are also rapidly emerging.1,2,6,15-18 However, quantitative
study is lacking for the interaction between oxygen lone pairs
and aromatic rings that are not particularly electron-deficient.
In the past few years, we have been studying weak molecular
forces such as π-π interactions using the triptycene scaffold,
which is an excellent model system for studying interactions in
the (1.5 kcal/mol range, Figure 1.19-21
A single crystal was obtained for one of the new series of
triptycene model compounds where the usual spectator 4-piv-
aloyl ester was replaced with a p-iodobenzoate group to increase
crystallinity (2, Figure 2). Despite poor mean bond distances
reported in the X-ray structure analysis for compound (2), the
crystal structure confirms that the syn conformation is the
preferred arrangement, where the methoxy oxygen atom of the
MOM group is situated near the center of the pentafluorophenyl
ring in the syn conformation (2). The distance from the methoxy
oxygen to the center of the pentafluorophenyl ring is 3.481 Å.
Caution should be exercised concerning the precision of the
bond distances since the crystal was not the highest quality.
This is a near van der Waals contact and consistent with most
crystal structures.7 However, the oxygen-to-arene distance is
significantly longer than the calculated equilibrium distance (3.0
Å) for the complex of dimethyl ether-hexafluorobenzene in
the gas phase (see Figure 3), presumably due to the constraint
of the scaffold. The torsional angles of the MOM group are
trans (T) at the first dihedral angle (C1O-CH2O) and gauche
(G) at the second dihedral angel (CH3O-CH2O) in the crystal
structure. Although the conformations of dimethoxymethane
have been studied extensively,22 there appears to be no theoreti-
cal study of the conformations of phenyl methoxymethyl ether.
At least two crystal structures containing the moiety of ArOCH2-
OCH3 have been reported and both show a gauche-gauche
(G,G) conformation.23,24 The T,G conformation in structure 2
Results and Discussion
The triptycene-derived model system allows the bridgehead
(C9 in the triptycene name system) rotating arene to interact
with a C1 group at near van der Waals’ distance in the syn
conformation, Figure 1. In the anti conformation, there is no
(4) Scheiner, S. Theory and Applications of Computational Chemistry:
The First Forty Years; Elsevier: Boston, MA, 2005; pp 831-857.
(5) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int.
Ed. 2003, 42, 1210-1250.
(6) Schneider, H. J.; Werner, F.; Blatter, T. J. Phys. Org. Chem. 1993,
6, 590-594.
(7) Egli, M.; Sarkhel, S. Acc. Chem. Res. 2007, 40, 197-205.
(8) Gallivan, J. P.; Dougherty, D. A. Org. Lett. 1999, 1, 103-105.
(9) Danten, Y.; Tassaing, T.; Besnard, M. J. Phys. Chem. A 1999, 103,
3530-3534.
(10) Scheiner, S.; Kar, T.; Pattanayak, J. J. Am. Chem. Soc. 2002, 124,
13257-13264.
(11) Quinonero, D.; Garau, C.; Rotger, C.; Frontera, A.; Ballester, P.;
Costa, A.; Deya, P. M. Angew. Chem., Int. Ed. 2002, 41, 3389-3392.
(12) Mascal, M.; Armstrong, A.; Bartberger, M. D. J. Am. Chem. Soc.
2002, 124, 6274-6276.
(13) Alkorta, I.; Rozas, I.; Elguero, J. J. Org. Chem. 1997, 62, 4687-
4691.
(14) Raimondi, M.; Calderoni, G.; Famulari, A.; Raimondi, L.; Cozzi,
F. J. Phys. Chem. A 2003, 107, 772-774.
(15) Hettche, F.; Hoffmann, R. W. New J. Chem. 2003, 27, 172-177.
(16) Rosokha, Y. S.; Lindeman, S. V.; Rosokha, S. V.; Kochi, J. K.
Angew. Chem., Int. Ed. 2004, 43, 4650-4652.
(17) Berryman, O. B.; Hof, F.; Hynes, M. J.; Johnson, D. W. Chem.
Commun. 2006, 506-508.
(18) Korenaga, T.; Tanaka, H.; Ema, T.; Sakai, T. J. Fluorine Chem.
2003, 122, 201-205.
(19) Oki, M. Acc. Chem. Res. 1990, 23, 351-356.
(20) Gung, B. W.; Patel, M.; Xue, X. W. J. Org. Chem. 2005, 70, 10532-
10537.
(21) Gung, B. W.; Xue, X. W.; Reich, H. J. J. Org. Chem. 2005, 70,
7232-7237.
(22) Huang, Y. R.; Knippenberg, S.; Hajgato, B.; Francois, J. P.; Deng,
J. K.; Deleuze, M. S. J. Phys. Chem. A 2007, 111, 5879-5897.
(23) Sebhat, I. K.; Tan, Y. L.; Widdowson, D. A.; Wilhelm, R.; White,
A. J. P.; Williams, D. J. Tetrahedron 2000, 56, 6121-6134.
(24) Chen, W. Z.; Zeng, Q. L.; Fang, H.; Gao, Y. X.; Zhao, Y. F. 2006,
E62, o869-o870.
690 J. Org. Chem., Vol. 73, No. 2, 2008

Interactions between Oxygen Lone Pair and Aromatic Rings
FIGURE 3. Ab initio electron density (HF/aug-cc-pVTZ) isosurfaces color-mapped with electrostatic potential for C6H6, C6F6, dimethyl ether, and
the complexes. The isosurface of the arenes shows a circular π system with a bowl-shaped center. In the complex, the oxygen is located at the
center of the arene to achieve maximum contact area in spite of the unfavorable electrostatic interactions in the Me2O-C6H6 complex.
suggests attractive interactions between the methoxy oxygen
and the center aromatic ring because the preferred G,G
conformation in 2 would have placed the methoxy group out
of contact with the C9 aryl group. The crystal was grown
between the interface of CH2Cl2 and hexane and the crystal cell
contains a CH2Cl2 molecule (see the Supporting Information).
One of the CH bonds of the CH2Cl2 may be considered to have
a CH-π interaction with one of the triptycene benzene rings.
The synthesis of the precursors to model compounds 3a-k
generally follows reported procedures and the installation of
the MOM group gave the desired compounds.21 The preparation
of the compound with X ) NO2 (3a) employs an alternate
method (see the Supporting Information for details). The
substituent X on the center arene varies from electron-withdraw-
ing to electron-donating groups. Low-temperature NMR spec-
troscopy enabled us to study the population of the syn and anti
conformations in chloroform solution. Distinct conformational
group. The general trend for the interaction strength based on
the data in Table 1 is consistent with electrostatic forces playing
a dominant role, i.e., the more electron-deficient the aromatic
ring is, the stronger the attraction. The strongest interaction is
observed with the pentafluorophenyl group (entry 11, Table 1).
For the monosubstituted aromatic rings, a Hammett plot did
not give a good correlation between the substituent constants
(σpara or σmeta) and the free energies obtained.25 The Hammett
σ parameters are known to not necessarily correlate with the
π-electron density on the aromatic rings. This point has been
discussed by both Dougherty and Sherrill.26,27 The reason for
this has been attributed to how the Hammett parameters were
determined. They were determined from the equilibrium con-
stants for the dissociation of substituted benzoic acids, therefore,
there is no reason to assume that they necessarily correlate with
the π-electron density in the reactants for those dissociations.26,27
However, if the substituents are broadly arranged into
electron-withdrawing and electron-donating groups, the data in
Table 1 are consistent with the expectations based on electro-
static interactions. The syn:anti ratios for the compounds with
electron-withdrawing substituents (i.e., NO2, CN, and CF3, and
the halogens) are greater than that of the standard compound
(3g, X ) H).25 The syn:anti ratios for the compounds with
electron-donating groups (CH3, CH3O, and Me2N) are smaller
than that of 3g. Interestingly, despite their known capacity to
donate electrons to aromatic rings, weak interactions were
observed even for the compounds with electron-donating groups
(3h, 3i, 3j). Higher than statistical syn:anti ratios are observed
for all three model compounds indicating the presence of
attractive interactions (entries 8-10, Table 1). To probe
alternative CH-π interactions, we prepared three model com-
pounds with a C(1) propyloxy group (entries 3, 7, and 8 in Table
1). Despite the C(9) arene in these three compounds bearing
three different substituents (CF3, H, CH3), all three exhibit
weaker interactions than their corresponding C(1) MOM and
MeO analogues. The propyloxy group and the MOM group are
similar in steric bulk. However, the MOM derivatives exhibit
higher than twice the syn:anti ratio as the propyloxy derivative.
Furthermore, CH-π interactions should prefer electron-rich C(9)
aromatic rings. The opposite is observed. Therefore any CH-π
attraction may have been countered by steric interactions, which
explains why the PrO is less favorable than the corresponding
MeO derivatives in assuming the syn conformation. Taking the
experimental evidence as a whole including the X-ray structure,
1
isomers were observed at low temperatures by H NMR. The
CH2Ar protons become diastereotopic in the syn conformation
and display an AB quartet while they are enantiomeric in the
anti conformation and appear as a singlet. The CH2 group
between two oxygen atoms in the MOM group also displays
the same distinct patterns in the syn and anti conformational
isomers.
Integration of the relative peak area from the singlet and the
AB quartet of the benzyl or MOM CH2 protons provided the
syn:anti isomer ratio for each compound. An average value from
the integration of both CH2 groups is recorded. The conforma-
tional freedom is rather limited for the rotating arene in either
the anti or the syn isomers in the triptycene model system.21
The MOM group is free to rotate in the anti conformation, but
its rotational freedom is limited in the syn conformation due to
the lp-π interaction. Thus the syn:anti ratio is effectively the
equilibrium constant between the syn and the anti conformations.
Since statistically two syn and one anti conformer are expected,
the equilibrium constant, Keq, equals a 1:2 syn:anti ratio. The
experimentally determined syn:anti ratios in CDCl3 and the free
energies derived from these ratios are shown in Table 1.
All model compounds studied showed attractive interactions
between the C9 substituted aromatic ring and the C1 MOM
(25) Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165-195.
(26) Mecozzi, S.; West, A. P.; Dougherty, D. A. Proc. Natl. Acad. Sci.
U.S.A. 1996, 93, 10566-10571.
(27) Sinnokrot, M. O.; Sherrill, C. D. J. Am. Chem. Soc. 2004, 126,
7690-7697.
J. Org. Chem, Vol. 73, No. 2, 2008 691

Gung et al.
TABLE 1. Substituent Effect for Lone Pair-Arene Interactionsa
syn:anti ratio (-50 °C),
∆Gantifsyn in CDCl3
for C1 MOMO,
kcal/mol
entry
compd
Ar
C1 MOMO
C1 MeO
C1 n-PrO
1
2
3
4
5
6
7
8
9
3a
3b
3c
3d
3e
3f
3g
3h
3i
C6H4NO2
C6H4CN
C6H4CF3
C6H4Br
C6H4Cl
C6H4F
8.4
9.4
10.1
7.8
7.1
7.5
6.8
4.8
5.2
7.2c
8.5b
5.5b
4.7c
3.5b
3.9c
3.3b
2.3b
2.4b
2.0c
>25:1c
-0.64 ( 0.05
-0.69
4.1c
-0.70
-0.59
-0.57
-0.59
C6H5
2.5c
1.7c
-0.54
C6H4Me
C6H4OMe
C6H4NMe2
C6F5
-0.40
-0.38
10
11
3j
3k
3.1
>25:1
-0.19
<-1.2
a The errors are estimated at (0.05 kcal/mol from an average of two runs. b Reference 17.21 c This work.
1H NMR data, and the results from the propyloxy derivatives,
the observed weak interactions between the bridgehead arene
and the MOM group do indeed originate from lp-π interactions.
Several models have been developed, which depict aromatic
rings as polar groups for studies related to π-π interactions,
and cation-π interactions.26,28,29 The fact that benzene and
hexafluorobenzene are known to have large and permanent
quadrupole moments30,31 supports these descriptions. Other
studies on the binding of cations and anions by aromatic rings
with quadrupole moments of the same polarity indicate the
importance of polarizability of the aromatic rings.32,33 Electro-
static potential surfaces of aromatic rings have also become a
convenient tool to describe the polarity of the arenes.26 The
surface of hexafluorobenzene has positive potential and the
surface of benzene has negative potential. The large attractive
interaction between the pentafluorobenzyl group and the MOM
group observed for compound 3k (entry 11, Table 1) is
consistent with the argument based on the polarity and the large
positive quadrupole moment of hexafluorobenzene.
However, an aromatic ring similar to benzene should have
negative potential and show repulsive interactions with an
oxygen lone pair. In contrast, the current study shows that a
phenyl ring and a MOM group exhibit considerable attractive
interaction (syn:anti ratio: 6.8, entry 7, Table 1). Electron-
donating groups should increase the electron density of the
aromatic ring and therefore cause the arene surface to have a
higher negative potential than benzene. Although compounds
3h-j (where X ) Me, MeO, and NMe2) did exhibit smaller
attractive interactions, they still showed greater than statistical
syn:anti ratios, indicating weak attractiVe interactions.
or the electrostatic potential surface of the arenes. The attractive
interactions with a MOM group are universally greater than that
with a methoxy group for the model compounds studied. With
a MOM group a lone pair of the outer (or second) oxygen is
located at the center of the arene (model series 3) while with a
MeO group the oxygen lone pair is located at the edge of the
arene (model series 1). The data in Table 1 clearly indicate that
the center approach of oxygen-to-arene is more favorable. Based
either on the quadrupole moment of the arenes or on examina-
tion of the color-coded electrostatic potential surface of the
arenes,26,34 one would not be able to predict these observations
because the edges of the aromatic rings should have more
positive potential and the center of the arene should have more
negative potential.28,34 Hence the edge of the arenes should be
more attractive toward an oxygen lone pair and the center of
the arenes should be repulsive toward an oxygen lone pair. The
data in Table 1 show that all interactions with a MOM group
are about twice as large as that with a MeO group except for
the arenes with a strong EWG (NO2 and CN), and the
pentafluorophenyl group. This appears to indicate that the
interactions between arenes with EDG and less strong EWG
and an oxygen atom may originate less from electrostatic forces
and more from dispersion forces. However, no correlation was
found between the free energy of interaction and the polariz-
ability of the substituted benzyl groups.35
To understand the nature of the lp-π interaction, we have
carried out molecular orbital calculations for benzene-dimethyl
ether, and hexafluorobenzene-dimethyl ether complexes, Figure
3. At the MP2/aug-cc-pVTZ level of theory, the complex
formation energy is -1.09 kcal/mol for C6H6-OMe2 and -4.59
kcal/mol for C6F6-OMe2 with optimized oxygen-to-arene plane
distances of 3.2 and 3.0 Å, respectively. The corresponding
calculations were also carried out at the Hartree-Fock level of
theory. The C6H6-OMe2 dimer has no energy minimum at the
Hartree-Fock level. If the intermolecular distance between
benzene and dimethyl ether is fixed at 3.2 Å, the interaction
energy is repulsive (∼3 kcal/mol). The C6F6-OMe2 has a
complex formation energy of -1.2 kcal/mol at the Hartree-
Fock level. Since the difference between calculations at the
Hartree-Fock level and at the MP2 level is correlation energy,
which is primarily made up of the dispersion energy, our
computational results indicate that electrostatic forces are
attractive in the complex of C6F6-OMe2, but are repulsive in
The data in the Table lead to the conclusion that any repulsive
interactions are outweighed by attractive ones when aromatic
rings and oxygen lone pairs are placed into close contact at
their van der Waals distance. Although electron-deficient arenes
interact with oxygen lone pairs more strongly, electron-rich
arenes also show weak attractive interactions. The observation
of the attractive interactions with the electron-rich arenes cannot
be rationalized by considering either the quadrupole moments
(28) Hunter, C. A.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112,
5525-5534.
(29) Muller-Dethlefs, K.; Hobza, P. Chem. ReV. 2000, 100, 143-167.
(30) Dennis, G. R.; Ritchie, G. L. D. J. Phys. Chem. 1991, 95, 656-
660.
(31) Williams, J. H. Acc. Chem. Res. 1993, 26, 593-598.
(32) Garau, C.; Frontera, A.; Quinonero, D.; Ballester, P.; Costa, A.;
Deya, P. M. ChemPhysChem 2003, 4, 1344-1348.
(33) Clements, A.; Lewis, M. J. Phys. Chem. A 2006, 110, 12705-12710.
(34) Ma, J. C.; Dougherty, D. A. Chem. ReV. 1997, 97, 1303-1324.
(35) Lide, D. R. Handbook of Chemistry and Physics, 86th ed.; CRC
Press: Ann Arbor, MI, 2005.
692 J. Org. Chem., Vol. 73, No. 2, 2008

Interactions between Oxygen Lone Pair and Aromatic Rings
the complex of C6H6-OMe2. Thus the attractive interactions
observed in the dimers with electron-rich aromatic rings must
be due to dispersion forces.
from the C-X bond in the aromatic rings might play a role as
well. The substituent constants normally used for constructing
the Hammett plot become inadequate when London dispersion
forces become a dominant factor. To gain further insight on
lp-π interactions, an additional systematic computational study
and a study involving heterocyclic arenes are underway in our
laboratories.
The calculated complex formation energies for these two
complexes corroborate our experimental measurements although
a solvent effect has not been considered. The ab initio electron
density (HF/aug-cc-pVTZ) isosurfaces color-mapped with elec-
trostatic potential for C6H6, C6F6, dimethyl ether, and the
complexes are displayed in Figure 3. As we pointed out
recently,36 the total electron density isosurfaces reveal that the
aromatic π system has a doughnut-shaped surface with the center
depressed. The preferred center-approach of the oxygen lone
pair toward an aromatic ring may be explained by considering
the shape-sensitive close-contact hypothesis.36 The oxygen of
dimethyl ether fits into the depressed center area of the arene
to achieve maximum overlap. The fact that the electron clouds
interpenetrate, and the electrostatic penetration term is usually
attractive provide a rational for the favorable center-approach
of the lone pair-π interaction.27
Thus, in spite of the negative potentials of both the oxygen
atom and the benzene ring, the preferred contact point is such
that the closeness of contact and maximum overlap are achieved.
It appears that at the van der Waals contact distance, the criteria
of closeness of contact override the polarity consideration for
aromatic rings. As pointed out by both Reisse and Dougherty,
the representation of the electronic distribution of a molecule
as a multipole expansion is valid only at large interaction
distances.34,37 Because of the distance requirement, cation-π
interaction cannot be quantitatively modeled as just an ion-
quadrupole interaction when van der Waals contact distance is
involved.34 For the same reason, it is improper to view the
arene-oxygen interaction as just a quadrupole-dipole interac-
tion at the van der Waals distance especially when the arenes
are not symmetric. The additional polarizability of the aromatic
ring makes it a special functional group.32,33 Although it is
convenient to describe aromatic rings as quadrupolar groups,
this study has shown the circumstances where this is not entirely
accurate. Other forces such as dispersion and local dipole should
come into play at the van der Waals distance. It has been
recognized that dispersion forces play an important role in π-π
interactions.5,27 It is reasonable to consider the attractive
interactions between the electron-rich arenes and oxygen lone
pair are mainly due to London dispersion forces. Consideration
of dispersion forces allows the understanding of both the weak
interactions between electron-rich arenes and oxygen lone pairs
and the center-approach of the oxygen toward arenes. In
addition, the fact that substituents with strong inductive electron-
withdrawing ability, such as CF3 and MeO groups, produced
higher than expected syn:anti ratios indicates that local dipoles
Experimental Section
Representative Procedure for the Preparation of the Model
Compounds: 9-(4-Nitrobenzyl)-1-(methoxymethoxy)-4-pivaloy-
loxytriptycene (3a). 9-(4-Nitrobenzyl)-1-hydroxy-4-pivaloyloxy
triptycene (5) (0.25mmol) was dissolved in 2 mL of dry CH2Cl2 at
0 °C. DIPEA (0.44 mL, 2.5 mmol) and chloromethyl methyl ether
(0.095 mL, 1.25 mmol) were added dropwise at 0 °C under N2
atmosphere. The reaction mixture was stirred at room temperature
for 4 h, then 10 mL of CH2Cl2 was added to dilute the reaction
mixture, which was washed with 1 N HCl three times, followed
by water. The organic layer was dried over MgSO4. Solvent was
removed and the residue was purified by column chromatography
to afford 9-(4-nitrobenzyl)-1-(methoxymethoxy)-4-pivaloyloxytrip-
1
tycene (3a). Yellow solid; mp 238-239 °C; H NMR (300 MHz,
CDCl3) δ 8.06 (2H, br), 7.92-7.31 (4H, br), 7.24-6.89 (6H, m),
6.80-6.70 (2H, q), 5.43 (1H, s), 5.42-3.80 (4H, br), 3.07 (3H, s),
1.52 (9H); 13C NMR (75 MHz, CDCl3) δ 176.8, 151.8, 148.8, 145.7,
140.5, 140.2, 130.1, 125.4, 124.8, 123.8, 122.7, 120.4, 113.7, 85.5,
55.8, 52.2, 48.7, 39.4, 35.3, 27.5; HRMS calcd for C34H31NO6 +
Na 572.2049, found 572.2051.
Variable-Temperature NMR Experimental Procedure The 1H
NMR spectra were recorded on a 300 MHz instrument with a
variable-temperature probe. A 0.05 M solution of the sample in a
deuterated solvent such as chloroform was placed in a high-quality
NMR tube. All samples were degassed using a needle-to-bubble
nitrogen through the sample for ∼1 min. The NMR tube was then
capped with a cap and sealed with Parafilm. The sample tube was
placed into the NMR probe and the air line to the probe was
replaced with a liquid nitrogen transfer line. The desired temperature
was set on the variable-temperature unit and the sample was allowed
to equilibrate for 10-15 min at each set temperature. Then the 1H
NMR spectrum at each temperature was recorded. The ratios of
rotamers were obtained through the integrations of selected peaks.
Acknowledgment. Acknowledgment is made to the donors
of the Petroleum Research Fund (PRF#40361-AC1), adminis-
tered by the American Chemical Society. B.W.G. is grateful
for support from the National Institutes of Health (GM069441).
J.C.A. acknowledges Penn State Erie, The Behrend College for
the funds to purchase the Gaussian and GaussView software.
Supporting Information Available: Experimental procedures
including low-temperature NMR spectroscopy, X-ray structure
analysis, and computational details including the absolute energies
and optimized geometries. This material is available free of charge
via the Internet at http://pubs.acs.org.
(36) Gung, B. W.; Amicangelo, J. C. J. Org. Chem. 2006, 71, 9261-
9270.
(37) Luhmer, M.; Bartik, K.; Dejaegere, A.; Bovy, P.; Reisse, J. Bull.
Soc. Chim. Fr. 1994, 131, 603-606.
JO702170J
J. Org. Chem, Vol. 73, No. 2, 2008 693