Long-range bonding/nonbonding interactions: A donor-acceptor resonance studied by dynamic NMR

Article abstract of DOI:10.1021/acs.orglett.5b01152

Long-range bonding interactions were evaluated using variable-temperature NMR spectroscopy and suitable 2′-CH₂X-substituted phenylpyridines (X = Me, NMe₂, OMe, F). It was found that the arylpyridyl rotational barriers were lower when electronegative atoms were bound to the α carbon of the 2′ moiety. This effect was ascribed to a stabilizing interaction in the transition state due to the lone pair of the heterocyclic nitrogen with the α carbon. Computational support for this hypothesis came from CCSD(T)/6-31+G(d) calculations. Steric effects of the X moiety were ruled out by comparison of the rotational barriers of analogous biphenyls.

Full text of DOI:10.1021/acs.orglett.5b01152

Letter
pubs.acs.org/OrgLett
Long-Range Bonding/Nonbonding Interactions: A Donor−Acceptor
Resonance Studied by Dynamic NMR
Renzo Ruzziconi,*,† Susan Lepri,† Federica Buonerba,† Manfred Schlosser,‡,∥ Michele Mancinelli,§
Silvia Ranieri,§ Luca Prati,§ and Andrea Mazzanti*,§
†Department of Chemistry, Biology and Biotechnology, University of Perugia, via Elce di Sotto 10, I-06100 Perugia, Italy
‡Institute of Chemical Sciences, Ecole Polytechnique Fed
́ ́
erale, CH-1015 Lausanne, Switzerland
§Department of Industrial Chemistry “Toso Montanari”, University of Bologna, Viale Risorgimento 4, I-40136 Bologna, Italy
S
* Supporting Information
ABSTRACT: Long-range bonding interactions were evaluated using variable-
temperature NMR spectroscopy and suitable 2′-CH2X-substituted phenylpyridines
(X = Me, NMe2, OMe, F). It was found that the arylpyridyl rotational barriers were
lower when electronegative atoms were bound to the α carbon of the 2′ moiety. This
effect was ascribed to a stabilizing interaction in the transition state due to the lone
pair of the heterocyclic nitrogen with the α carbon. Computational support for this
hypothesis came from CCSD(T)/6-31+G(d) calculations. Steric effects of the X
moiety were ruled out by comparison of the rotational barriers of analogous
biphenyls.
eak chemical interactions play a key role in chemistry.
They are essential for biological systems and they can be
W
employed to design supramolecular assembly and scaffolds.1
Among them, van der Waals dispersive forces,2 hydrogen bonds,3
halogen−halogen,4 and π−π interactions5 are the most studied,
and others have been under investigation in the recent past.6
During our dynamic NMR studies of 2-(2′-alkylphenyl)-
pyridines,7 we came across an apparent discrepancy. While the
CH2 signal of 2-(2-ethylphenyl)pyridine (1) (Table 1) did show
Figure 1. Model compounds for computational studies.
Supporting Information, SI) with a lower experimental aryl−aryl
rotational barrier (4.8 kcal/mol).
Common sense suggests that the methoxymethyl group
should be bigger than a methyl group because the methoxy entity
requires more space than a single hydrogen atom. Such plausible
expectations are corroborated by all standard criteria assessing
steric effects. Charton’s “effective” upsilon parameters,10 derived
from van der Waals radii, place methoxymethyl (υ = 0.63) above
methyl (υ = 0.52) and even between ethyl (υ = 0.56) and propyl
(υ = 0.68). Winstein’s A values11 based on the free energy
differences between axially substituted cyclohexanes and their
equatorial invertomers rank methoxymethyl (A = 1.72) at the
same level as methyl (A = 1.74). Finally, our B scale based on
biphenyls torsional barriers12 puts methoxymethyl (B = 8.6) ex
aequo with ethyl (B = 8.7) and on top of methyl (B = 7.4).
All the more disconcerting it was to find this sound order of
bulkiness reversed when we turned from 2-alkylbiphenyls to 2-
(2′-alkylphenyl)pyridines.
Table 1. Experimental and Calculated Torsional Barriers for
2-Arylpyridines
a
b
compd
X
exptl value
planar TS
8.4
skewed TS
ΔE pl-sk
c
1
2
3
4
5
Me
5.9
6.2
+2.2
d
d
d
H
5.9
5.9
0
OMe
NMe2
F
4.9
6.8
4.0
6.7
6.6
6.9
−1.8
+0.2
−2.9
4.8
a
Calculations at the CCSD(T)/6-31+G(d)//ωB97XD/6-31G(d)
level (energies in kcal/mol). Values in bold indicate the preferred
b
transition state. A positive value indicates a more stable skewed TS.
c
d
See ref 7. For this compound there is only one TS.
decoalescence below −164 °C (ΔG⧧ = 5.9 kcal/mol), in the case
of 2-[2-(methoxymethyl)phenyl]pyridine 3 the coalescence
point was not observed at temperatures as low as −173 °C
(Figure 1). This means the barrier was too small to allow
experimental determination by dynamic NMR measurements.8,9
In order to obtain additional data, we prepared the
dimethylamino analogue 4. When cooled to −170 °C,
compound 4 did show splitting of the CH2 signal (Figure S1,
To find a possible explanation, we resorted to quantum
chemical calculations. The geometries of the stationary points on
the aryl−aryl rotation pathway of 1 and 3 were optimized at the
ωB97XD/6-31G(d) level of theory. Two different transition
Received: April 20, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b01152
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
states (TS) were modeled and optimized, one corresponding to a
N−C1−C1′−C2′ dihedral angle of 180° and the other with the
dihedral set to 0°. The two TS were validated by frequency
analysis, showing a single imaginary frequency corresponding to
the rotation around the aryl−aryl bond. In both cases, the lower
energy transition state corresponds to the crossing of the CH2X
group on the side of the pyridine nitrogen (i.e., N−C1−C1′−C2′
≈ 0°). This is conceivable because of the smaller steric
interference of the lone pair of nitrogen with respect to an
hydrogen atom when the angle is ∼180°.7 Once the aryl−aryl
torsional angle of the TS was modeled, two different dispositions
of the CH2−X moiety are available. The first puts the X fragment
in the same plane of phenyl and pyridine (“planar”
conformation), while the second has the X fragment out of the
plane (“skewed” conformation). Both of the geometries were
optimized and validated as transition states by frequency analysis
and IRC calculations (see the SI). It was found that the energy
barrier for compound 3 (X = OMe) was smaller than that of 1
and 2, while the barrier calculated for the N,N-dimethylamino
compound 4 was in the middle. However, we noted that the
calculated energy barriers were lower than the experimental ones
(see Table S1, SI, for a summary). This outcome could be due to
a wrong estimation of the contribution of the pyridine lone pair
in the transition state.13 For this reason, we calculated the single-
point energies at the CCSD(T)/6-31+G(d)//ωB97XD/6-
31G(d) level (Table 1).
The reversed preference in the TS geometry of 3 with respect
to 1 can be rationalized by considering a stabilizing interaction by
the lone pair of the nitrogen that lowers the TS energy of
compounds 3−5. As long as the nitrogen atom disposes of its free
doublet it may share it with a neighboring electron-deficient
center, provided the correct geometry is achieved. Such kind of
assistance is at the basis of anomeric effects14 and neighboring
group participations15 if the nitrogen donor site and the halide
(or alkoxy) acceptor site are part of geminal or vicinal bonding
patterns. What makes unique the situation met with α-hetero-
substituted 2-arylpyridines is the distance between the
heterocyclic nitrogen atom and the α-carbon, the two centers
being separated by four interposed bonds with no conjugative
effects. Moreover, the amplitude of the barrier-lowering donor−
acceptor interaction is modulated by the nature of the
heteroelement.
In this framework, the long-range interaction manifests itself as
a resonance of bonding and nonbonding limiting structures (A
and B in Figure 3). For reasons of proximity, only the coplanar
transition state of the torsional motion can benefit from this
electron-density leveling interaction.
The calculated value for the rotational barrier of 2-(2-
ethylphenyl)pyridine (1) benchmarked the reliability of these
calculations. Nevertheless, at the CCSD(T) level of theory the
activation energy for aryl−aryl rotation of 3 was again found to be
smaller than that of 1 and 2 (Table 1). A close examination of the
transition-state geometries showed that the conformation
assumed by the ethyl group in the best TS is substantially
different from that of methoxymethyl. In the first compound, the
methyl group has a dihedral angle of 80° (Me−CH2−C2′−C1′)
with the phenyl plane, whereas the same angle is exactly 180° for
the CH2OMe group (O−CH2−C2′−C1′ dihedral angle). When
the alternative TS (i.e., the methyl at 180° and OMe at ∼75°)
were calculated, their energies were higher than the previous
ones by 2.2 and 1.8 kcal/mol (see Figure 2). This trend was
confirmed by the calculation of the energy barrier to rotation of
2-(o-fluoromethylphenyl)pyridine 5, where the calculated value
was lower than that of 3 and 4 (Table 1). In the case of
compound 4 the two transition states have roughly the same
energy.
Figure 3. Two limiting structures for the long-range interaction.
The resonance extending from the pyridine nitrogen through
the benzylic α-carbon to the heterosubstituent X is accompanied
by a respectable gain of energy. If we assume that all 2′-CH2X
substituents exert the same steric hindrance (see below for the
experimental confirmation), the nonclassical long-range inter-
action lowers the torsional barriers at an extent corresponding to
the stabilization of the transition state, where this interaction can
be effective (Figure 4).16
Figure 4. Schematic representation of the stabilizing interaction.
Although there was no reason to doubt the validity of the
CCSD(T) computational results, we strived for further
experimental confirmation. To make the torsional barrier more
suited to the dynamic-NMR technique,17 a methyl group was
installed in the 3-position of the pyridine ring (compounds 6−10
of Table 2).
For convenience of synthesis, the model compounds 6−8
were prepared from 3,5-dimethylpyridine by a Ziegler-type
addition18 of 2-ethylphenyllithium, [2-(dimethylamino)methyl]-
Figure 2. Two available transition states for aryl−aryl rotation of
compounds 1 and 3. Energies (in kcal/mol) were calculated at the
CCSD(T)/6-31+G(d)//ωB97XD/6-31G(d) level, and they are
relative to the ground states.
B
DOI: 10.1021/acs.orglett.5b01152
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
For all of the compounds containing heteroatoms in the α
position, the calculated transition states showed that the X group
is coplanar with the phenyl ring and points away from the
nitrogen of pyridine. As discussed above, this geometry has the
correct disposition of the two partners for the interaction of the
nitrogen lone pair with the LUMO of C−X. On the contrary, the
lowest energy TS in compounds 6, 11, and 12 was always the
skewed one. When the energy difference between the planar TS
and the skewed TS is evaluated (Table 2), the stabilization of the
planar TS regularly increases with the leaving group ability.
In addition, the results obtained when trying to prepare the
compound bearing a 2′-CH2Br group can be discussed within
this framework. The reaction of 2-bromobenzyl bromide with 2-
pyridineboronic acid led to prompt cyclization to 6H-pyrido[2,1-
a]isoindol-5-inium bromide (B structure of Figure 3). The same
result was found also when [2-(bromomethyl)phenyl]boronic
acid was reacted with 2-bromopyridine or when 2-[2-
(hydroxymethyl)phenyl]-3-methylpyridine was treated with
CBr4 in the Appel conditions.20 The same outcome was obtained
in the attempt to prepare the 2′-CH2Cl compound by the same
synthetic approaches. All of these results are in agreement with a
strong interaction of the pyridine lone pair with the benzylic
carbon, leading to cyclization in the presence of a better leaving
group such as chloride or bromide.
Table 2. Experimental and Calculated Torsional Barriers for
Compounds 6−18
a
b
compd
R1
R2
X
exptl barrier calcd barrier ΔE pl-sk
6
Me
Me
Me
Me
Me
Et
Me
Me
Me
H
Me
13.1
12.8
11.1
10.9
8.8
13.1
14.7
12.4
12.2
10.5
2.3
0.0
7
NMe2
OMe
OMe
F
8
−1.4
−1.4
−3.4
9
10
11
12
13
14
15
16
17
18
H
Et
H
13.7
14.4
14.7
12.6
15.9
15.5
15.1
15.1
Et
Et
Me
Et
Et
NMe2
OMe
Me
Et
Et
18.9
18.8
17.7
17.9
2.1
3.1
2.4
1.5
NMe2
OMe
F
a
Calculations at the CCSD(T)/6-31+G(d)//ωB97XD/6-31G(d)
b
level (energies in kcal/mol). A positive value indicates a more stable
skewed TS.
The key role played by the pyridine lone pair in the
stabilization of the coplanar transition state can be further
confirmed by comparing the rotational barriers of the analogous
biphenyls 15−18 where a methoxy group was installed in the
ortho position to tune the aryl−aryl rotational barriers to values
similar to those of compounds 6−10.21 Calculations suggested
that in this series of compounds the rotational barrier were very
similar on varying the X substituent. Between the two feasible
transition states, the preferred one corresponds to a C2−C1−
C1′−C2′ dihedral angle close to 180°, where the steric interaction
between the 2-methoxy and 2′-CH2X moieties is minimized.
Within this series of compounds, the preferred conformation of
the CH2X group in the best transition state was always the
skewed one. Biphenyls 15−18 were easily prepared by Suzuki−
Miyaura coupling (see the SI for details), and the experimental
rotational barriers were indeed found to be clustered in a very
limited range (less than 1 kcal/mol, see Table 2), thus confirming
that all of the CH2X groups have almost identical steric
contributions.
phenyllithium and (2-(methoxymethyl)phenyl)lithium,19 fol-
lowed by spontaneous lithium hydride elimination at ambient
temperature. Compound 10 (X = F) could not be prepared this
way because the intramolecular nucleophilic attack by the lithium
amide intermediate at the fluorinated benzylic carbon led to the
cyclized product (see below). Nevertheless, compound 10 was
obtained by Suzuki−Miyaura cross-coupling between 2-bromo-
3-methylpyridine and [(2-(fluoromethyl)phenyl]boronic acid.
To check whether the presence of the 5-methyl group on the
pyridine ring could modify the TS geometry and the rotational
barrier by inductive electronic effect we prepared compound 9.
As proven by the identity of the rotational barriers of compounds
8 and 9, the methyl in position 5 of pyridine has no effect on the
torsional barrier.
An evaluation of the contribution of hydrogen (X = H) was
also desirable, but 2′,3,5-trimethyl-2-phenylpyridine lacks
diastereotopic nuclei and is hence unsuitable. To fill this gap,
another set of compounds (11−14, Table 2) was prepared by
adding the suitable aryllithium onto the 2-position of 3,5-
diethylpyridine followed by thermal rearomatization.
The long-range effect described above evokes short-range
“nonclassical” electronic interactions such as “negative hyper-
conjugation”22 and homoaromaticity.23 We further recognize a
All of the torsional barriers were derived by line-shape
simulation of the signal of CH2X that splits into an AB system
when the aryl−aryl rotation is slow on the NMR time scale.
Spectra were acquired in CD2Cl2 or in CDFCl2 at a field of 14.4 T
close relationship to the Dunitz−Burgi trajectory continuum24
̈
and Mayr’s nucleofugality parameters.25 The concept of long-
range resonance interactions raises new questions about
intramolecular interactive effects, and further investigations are
currently underway.
1
(600 MHz for H, see Figures S2−S14, SI, for the spectra and
simulations at different temperatures).
The results collected in Table 2 convey an unequivocal
message. Electronegative substituents such as oxygen and
fluorine substantially lower the activation energy for the rotation
around the aryl-pyridyl axis, while the effect of a methyl instead of
an hydrogen has a tiny effect on the barrier (compound 11 vs
12). The effect of the nitrogen substitution on the α carbon is less
straightforward (compound 6 vs 7 and 12 vs 13). In these cases,
the lone pair of the sp3 nitrogen is highly nucleophilic, and it can
be oriented in the ground state in such a way to develop a
stabilizing interaction with the electron-poor carbon in position 2
of pyridine.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthetic details, spectroscopic data, and variable-temperature
1
NMR spectra for compounds 4 and 6−18. H and 13C NMR
spectra of compounds 4 and 6−18. Computational details of
compounds 1−10 and 15−18. The Supporting Information is
available free of charge on the ACS Publications website at DOI:
10.1021/acs.orglett.5b01152.
C
DOI: 10.1021/acs.orglett.5b01152
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
found to be negligible by the calculations of the following isodesmic
reaction (see the SI for details).
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: renzo.ruzziconi@unipg.it.
*E-mail: andrea.mazzanti@unibo.it.
Author Contributions
(17) Casarini, D.; Lunazzi, L.; Mazzanti, A. Eur. J. Org. Chem. 2010,
2035−2053.
(18) (a) Ziegler, K.; Zeiser, K. Ber. Dtsch. Chem. Ges. 1930, 63, 1847−
1851. (b) Ziegler, K.; Zeiser, K. Justus Liebigs Ann. Chem. 1931, 485,
174−192.
The manuscript was written through contributions of all authors
Notes
(19) Reich, H. J.; Goldenberg, W. S.; Sanders, A. W.; Jantzi, K. L.;
Tzschucke, C. C. J. Am. Chem. Soc. 2003, 125, 3509−3521.
(20) (a) Appel, R. Angew. Chem., Int. Ed. 1975, 14, 801−811.
(b) Denton, R. M.; Ana, J.; Adenirana, B. Chem. Commun. 2010, 46,
3025−3027.
(21) A 2-methyl group raised the barrier to about 20 kcal/mol.
(22) Allinger, N. L. Molecular Structure; John Wiley & Sons, Inc.:
Hoboken, 2010; pp 162−188.
(23) (a) Griffiths, P. R.; Pivonka, D. E.; Williams, R. V. Chem.Eur. J.
2011, 17, 9193−9199. (b) Williams, R. V. Chem. Rev. 2001, 101, 1185−
1204.
The authors declare no competing financial interest.
∥Deceased June 26th, 2013.
ACKNOWLEDGMENTS
■
The University of Bologna (RFO funds 2012) and Molecular
Discovery Ltd (Perugia) are gratefully acknowledged for
financial support. ALCHEMY Fine Chemicals & Research
(Bologna, www.alchemy.it) is gratefully acknowledged for a
generous gift of chemicals
(24) (a) Burgi, H. B.; Dunitz, J. D. Acc. Chem. Res. 1983, 16, 153−161.
̈
REFERENCES
(b) Jana, D. F.; Wodrich, M. D.; Corminboeuf, C. J. Org. Chem. 2012, 77,
2548−2552.
■
(1) Ward, M. D.; Raithby, P. R. Chem. Soc. Rev. 2013, 42, 1619−1636
and references cited therein.
(25) (a) Streidl, N.; Denegri, B.; Kronja, O.; Mayr, H. Acc. Chem. Res.
2010, 43, 1537−1549. (b) Mayr, H.; Ammer, J.; Baidya, M.; Maji, B.;
Nigst, T. A.; Ofial, A. R.; Singer, T. J. Am. Chem. Soc. 2015, 137, 2580−
2599.
(2) (a) Hunter, C. A. Angew. Chem., Int. Ed. 2004, 43, 5310−5324.
(b) Ehrlich, S.; Bettinger, H. F.; Grimme, S. Angew. Chem., Int. Ed. 2013,
41, 10892−10895. (c) Yang, L.; Adam, C.; Nichol, G. S.; Cockroft, S. L.
Nat. Chem. 2013, 5, 1006−1010. (d) Mati, I. K.; Cockroft, S. L. Chem.
Soc. Rev. 2010, 39, 4195−4205.
(3) (a) Emenike, B. U.; Carroll, W. R.; Roberts, J. D. J. Org. Chem.
2013, 78, 2005−2011. (b) Desiraju, G. R. Acc. Chem. Res. 1996, 29,
441−449. (c) Steiner, T. Angew. Chem., Int. Ed. 2002, 41, 48−76.
(4) (a) Pilati, T.; Cozzi, F. CrystEngComm 2011, 13, 4549−4556.
(b) Metrangolo, P.; Neukirch, H.; Pilati, T.; Resnati, G. Acc. Chem. Res.
2005, 38, 386−395.
(5) (a) Hohenstein, E. G.; Duan, J.; Sherrill, C. D. J. Am. Chem. Soc.
2011, 133, 13244−13247. (b) Watt, M.; Hardebeck, L. K. E.;
Kirkpatrick, C. C.; Lewis, M. J. Am. Chem. Soc. 2011, 133, 3854−3862.
(6) (a) Chiarucci, M.; Ciogli, A.; Mancinelli, M.; Ranieri, S.; Mazzanti,
A. Angew. Chem., Int. Ed. 2014, 53, 5405−5409. (b) Mati, I. K.; Adam,
K.; Cockroft, S. L. Chem. Sci. 2013, 4, 3965−3972.
(7) Mazzanti, A.; Lunazzi, L.; Lepri, S.; Ruzziconi, R.; Schlosser, M.
Eur. J. Org. Chem. 2011, 6725−6731.
(8) A small chemical shift difference between the diastereotopic signals
could prevent the detection of this energy barrier. However, large shift
differences were observed for compounds 1 and 4.
(9) In the case of o-tolylpiridine, a diastereotopic probe had been
installed on the 4-position of pyridine. See ref 7
(10) Charton, M. Top. Curr. Chem. 1983, 114, 57−91.
(11) (a) Winstein, S.; Holness, N. J. J. Am. Chem. Soc. 1955, 77, 5562−
5578. (b) Hirsch, J. A. Top. Stereochem. 1967, 1, 199−222. (c) Eliel, E.
L.; Wilen, S. H. Stereochemistry of Organic Compounds; Wiley: New York,
1994; pp 690−700. (d) Kleinpeter, E.; Taddei, F. Chem.Eur. J. 2003,
9, 1360−1368.
(12) (a) Ruzziconi, R.; Spizzichino, S.; Lunazzi, L.; Mazzanti, A.;
Schlosser, M. Chem.Eur. J. 2009, 15, 2645−2652. (b) Lunazzi, L.;
Mancinelli, M.; Mazzanti, A.; Lepri, S.; Ruzziconi, R.; Schlosser, M. Org.
Biomol. Chem. 2012, 10, 1847−1855.
(13) Dahlgren, M. K.; Schyman, P.; Tirado-Rives, J.; Jorgensen, W. L. J.
Chem. Inf. Model. 2013, 53, 1191−1199 and references cited therein.
(14) (a) Eliel, E. L. Angew. Chem., Int. Ed. Engl. 1972, 11, 739−750.
(b) Bauerfeldt, G. F.; Cardozo, M.; Pereira, M. S.; da Silva, C. O. Org.
Biomol. Chem. 2013, 11, 299−308. (c) Mo, Y. Nat. Chem. 2010, 2, 666−
671.
(15) (a) Yamada, K.; Sakata, S.; Yoshimura, Y. J. Org. Chem. 1998, 63,
6891−6899. (b) Cram, D. J. J. Am. Chem. Soc. 1952, 74, 2129−2137.
(c) Winstein, S.; Lucas, H. J. J. Am. Chem. Soc. 1939, 61, 1576−1581.
(16) In principle, ground-state effects due to the different X
substituents cannot be excluded. However, this contribution was
D
DOI: 10.1021/acs.orglett.5b01152
Org. Lett. XXXX, XXX, XXX−XXX