Aliphatic C-H/π and heteroatom/π interactions in N-aryl-3,4- (9′,10′-dihydroanthracene-9′,10′-diyl)succinimides

Article abstract of DOI:10.1002/ejoc.201402288

A series of N-aryl-3,4-(9′,10′-dihydroanthracene-9′, 10′-diyl)succinimides have been synthesized as model compounds to study the interaction of an aromatic π system with aliphatic H atoms and heteroatoms such as oxygen, sulfur, and fluorine. The steric re

Full text of DOI:10.1002/ejoc.201402288

FULL PAPER
DOI: 10.1002/ejoc.201402288
Aliphatic C–H/π and Heteroatom/π Interactions in N-Aryl-3,4-
(9Ј,10Ј-dihydroanthracene-9Ј,10Ј-diyl)succinimides
Laura Raimondi,[a] Maurizio Benaglia,[a] and Franco Cozzi*[a]
Keywords: Conformation analysis / Pi interactions / Noncovalent interactions / Through-space interactions
A series of N-aryl-3,4-(9Ј,10Ј-dihydroanthracene-9Ј,10Ј-diyl)-
succinimides have been synthesized as model compounds to
study the interaction of an aromatic π system with aliphatic
H atoms and heteroatoms such as oxygen, sulfur, and fluor-
ine. The steric requirements of the rigid scaffold forces the
N-aryl residue and a benzene ring of the dihydroanthracenyl
moiety to adopt an edge-to-face conformation in which one
of the ortho-substituents of the N-aryl group interacts with
the π system. The succinimides were generally obtained as
mixtures of isomers and their ratios were considered as a
function of the substitution pattern at the ortho positions of
the N-aryl residue. The experimentally observed ratios were
in good agreement with those predicted computationally by
DFT calculations at the BMK/cc-pVDZ level of theory. From
this study it was concluded that in these systems the aliphatic
C–H/π interaction is mildly stabilizing and, to a certain ex-
tent, capable of overcoming unfavorable steric effects. On
the other hand, the interaction of a π system with oxygen,
sulfur, or fluorine atoms was destabilizing. When methyl or
ethyl groups were allowed to compete with oxygen and
fluorine for interaction with the benzene ring, C–H/π inter-
actions were consistently observed to preferentially occur de-
spite unfavorable steric effects.
in the ortho position was located either close to[4f] (endo-
isomer, Figure 1) or away from (exo-isomer, Figure 1) the π
electron system, and this conformational preference could
potentially provide a probe with which to detect R/π inter-
actions. Conveniently, even in mono ortho-substituted ad-
ducts, rotation around the N-aryl bond was generally slow
enough to allow the R-endo and R-exo atropisomers to be
observed at room temperature by NMR spectroscopy and
even separated.[4] The R-endo isomer was expected to be
destabilized by both steric factors and R/π repulsive elec-
tronic effects and stabilized only by favorable electronic
R/π interactions, the strength of which can be evaluated
against the steric repulsion.
Introduction
Interactions involving aromatic compounds have been
the subject of extensive theoretical and experimental studies
because of their central role in determining the structures
of supramolecular adducts, organic crystals, and macrobio-
molecules, and in mediating several molecular recognition
events in artificial and biological systems.[1] Among these
interactions, those involving aromatic rings and aliphatic H
atoms or lone-pair-bearing heteroatoms have been studied
experimentally by developing a variety of ad hoc designed
model systems.[2] These have generally featured a rigid scaf-
fold that was used to bring the interacting entities in close
contact and to make experimentally detectable the expected
weak interaction.
N-Substituted 3,4-(9Ј,10Ј-dihydroanthracene-9Ј,10Ј-diyl)-
succinimides have been used as model compounds to study
nonbonding interaction between various N-substituents
and one of the benzene rings of the 9,10-dihydroanthracene
moiety.[3,4] In particular, when the N-substituent was an
aryl residue (as in the structures reported in Figure 1) the
steric requirements of the rigid scaffold and the presence of
the two succinimide oxygen atoms forced the N-aryl group
to adopt a roughly perpendicular, edge-to-face disposition
relative to the benzene ring. Accordingly, the R substituent
Figure 1. Structures of R-endo and R-exo isomers.
In previous studies it was found that when R = Me and
X = H (Figure 1), a 46:54 mixture of Me-endo and Me-exo
adducts was obtained.[3b,4g] Thus, the C–H/π interaction[5]
was favorable enough to partially balance the repulsive
steric effect. Remarkably, the proportion of the Me-endo
adduct increased when the N-aryl residue featured an elec-
tron-withdrawing substituent X in the para position relative
[a] Università di Milano, Dipartimento di Chimica,
19, via Golgi, 20133 Milano, Italy
E-mail: franco.cozzi@unimi.it
http://www.chimica.unimi.it/ecm/home
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201402288.
Eur. J. Org. Chem. 2014, 4993–4998
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4993

L. Raimondi, M. Benaglia, F. Cozzi
FULL PAPER
to the nitrogen atom, with the Me-endo/Me-exo ratio be-
coming 70:30 when X was a nitro group.[4d] These results
were explained in terms of a favorable Me/π interaction that
was able to overcome the steric repulsion only if the H
atoms were made “acidic” enough by suitable phenyl ring
substitution with the X group.[4d] No isotope effect was ob-
served on passing from R = CH3 to R = CD3, with the two
adducts showing roughly the same endo/exo ratio.[4g] As for
succinimides bearing R ortho-heterosubstituents on the N-
aryl ring, a MeO group was found to preferentially occupy
the exo position (MeO-exo/MeO-endo isomer ratio
80:20).[4e] Halogen substituents behaved similarly, with
halogen-exo/halogen-endo isomer ratios 80:20 and 88:12 for
R = Cl and R = Br, respectively.[4a]
In a continuation of our efforts towards the development
of model systems to evaluate nonbonding interactions in-
volving aromatic systems,[6–8] and with the aim of obtaining
some new insights into aliphatic C–H/π and heteroatom/π
interactions both individually and in competition with one
another, here we report the synthesis of a series of new N-
aryl-substituted succinimides both ortho-monosubstituted
and ortho,orthoЈ-disubstituted at the aryl ring, and disclose
the results of an experimental and computational study of
their stereodynamic behavior.
Scheme 1. Synthesis and structures of succinimides 2–7.
Table 1. Isolated yields and isomer ratios for succinimides 2–7, and
chemical shifts of diagnostic signals.
[c]
R
Yield [%][a] R endo/exo[b]
Ho
R exo[d]
R endo[e]
2
3
Me
Et
75
73
47:53
56:44
5.43 1.98
5.47 2.40 (CH2),
1.14 (CH3)
0.97
1.09 (CH2),
0.89 (CH3)
0.75 (CH),
0.83 (CH3)
0.69
4
iPr
71
69:31
5.39 2.61 (CH),
1.15 (CH3)
5
6
7
tBu
MeO
MeS
70
80
77
15:85
21:79
17:83
4.94 1.21
5.37 3.64
5.63 2.38
Results and Discussion
3.31
2.12
Monosubstituted N-aryl-3,4-(9Ј,1Ј0-dihydroanthracene-
9Ј,10Ј-diyl)succinimides 2–7 were easily obtained by stirring
an equimolar mixture of anhydride 1 and the corresponding
commercially available anilines at 115 °C in 1,1,2,2-tetra-
chloroethane (TCE) for 18 h (Scheme 1).[4] By monitoring
the reactions by TLC it was evident that substantial
amounts of product were already present after about two
[a] Isolated yield after flash chromatography. [b] Determined by
1H NMR analysis of the crude reaction mixture (300 MHz,
CDCl3). [c] 1H NMR shift of the H atom in the ortho position to
the N-atom in the R-exo isomer. [d] 1H NMR shift of the protons
of the R group in the exo isomer. [e] 1H NMR shift of the protons
of the R group in the endo isomer.
hours. Nevertheless, the reaction times were prolonged to the isopropyl group in adduct 4 were capable of securing
18 h to obtain equilibrium mixtures of the endo- and exo- C–H/π interactions favorable enough to overcome the re-
isomers under identical conditions. The endo/exo ratios (re- pulsive steric effect.[10] The fact that these residues occupied
1
ported in Table 1) were determined by 300 MHz H NMR the endo position preferentially relative to a methyl was ten-
analysis of the crude reaction products in CDCl3. The tatively explained by considering that both substituents
1H NMR spectrum of the R exo-isomers consistently could provide a stabilizing C–H/π interaction more ex-
showed the signal of the N-aryl ortho-H atom as a doublet tended than that provided by a methyl group. This hypothe-
at high field, whereas the signals of the R groups resonated sis was later supported by computational studies (see be-
at lower field for the exo-isomers and at higher field for low). However, steric hindrance largely counteracted at-
the endo-isomers (Table 1). The high field chemical shifts tractive effects in the case of compound 5 in which, not
observed for the signals of the ortho-H atom and the R unexpectedly, the tert-butyl group showed a clear preference
residue of the endo isomers were taken as a clear indication for the exo-position.
of the interaction with the benzene ring of the scaffold due
The study of adducts 6 and 7 provided information about
to the edge-to-face conformation adopted by the N-aryl the interaction between lone-pair-bearing heteroatoms and
group and the dihydroanthracene benzene ring in solution. aromatic systems. This interaction has been the subject of
The adducts were then purified by flash chromatography to a number of studies.[7,11,12,13] In particular, Gung and
obtain the products as high-melting white solids in isolated Amicangelo, studying trypticene-based model compounds
yields ranging from 70 to 80%.[9]
in CDCl3, showed that the methoxy/π interaction was
The data presented in Table 1 show that replacement of weakly favorable when the aromatic system was electron
the methyl group of 2 with either an ethyl or an isopropyl rich (–0.8 to –2.9 kJmol–1) and became even more stabiliz-
group in 3 and 4, respectively, led to a reversal of the ing when the aromatic ring was a pentafluorophenyl group
endo/exo ratio, as the endo-isomer became slightly favored. (up to –5.0 kJmol–1).[11] Motherwell and Aliev, studying a
Thus, the ethyl residue in adduct 3 and even more so dihydroanthracene-based model in CDCl3, observed the
4994
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 4993–4998

Aliphatic C–H/π and Heteroatom/π Interactions
preferential occurrence of sulfur/π over oxygen/π interac- tios, with the only exception being the isopropyl-substituted
tion.[12] The sulfur/π and oxygen/π interaction was also adduct 4, which was formed as a 87:13 mixture of endo and
studied by our group using [3.3]meta(heterocyclo)-para- exo isomers. However, this discrepancy was ascribed to the
cyclophane model compounds in different mixtures of freon fact that the endo isomer of 4 was much less soluble than
solvents, which revealed that the former was less favorable its exo counterpart in refluxing acetic acid and, in the
than the latter as the result of a combination of steric and course of the reaction, precipitated out of the solution con-
electronic effects.[7b]
taining the two equilibrating isomers, thus shifting the equi-
The R-exo/endo isomer ratios observed for adducts 6 librium in its favor. Indeed, a highly enriched sample of the
(79:21) and 7 (83:17) seemed to indicate that, at least in endo isomer (endo/exo isomer ratio 10:1, obtained during
these model systems, the interaction between lone-pair- the chromatographic purification of the crude reaction
bearing heteroatoms and the aromatic system was unfavor- product) kept for 96 h at room temperature in CDCl3 solu-
able. The increase of R-exo/endo ratio observed on passing tion was transformed into a 68:32 endo/exo mixture of iso-
from R = MeO to R = MeS was consistent with an increas- mers, which is a ratio very similar to that obtained by per-
ingly repulsive steric effect[10] acting synergically with the forming the reaction in TCE in which both isomers were
electronic effect. This finding was in agreement with litera- soluble.
ture data[4d] and with our previous observation,[7b] but
contrasted with Motherwell’s and Aliev’s findings, which
Table 3. The endo/exo ratios for compounds 2–4, 6, and 7 in TCE
predicted a sulfur/π interaction more favorable than an oxy-
gen/π interaction because of the higher polarizability of
sulfur.[12]
and AcOH at 115 °C.
R
R-endo/exo ratio in TCE R-endo/exo ratio in AcOH
To gain deeper insight into the behavior of some of these
succinimides, a DFT computational study at the BMK/
cc-pVDZ level of theory[6c,14] was carried out on com-
pounds 2–4, 6, and 7. It was found that the calculations
were able to predict the experimentally observed endo/exo
ratios with excellent qualitative and, in some cases, also
good quantitative agreement (Table 2 and Figure 2). In par-
ticular, the calculation on the ethyl- and isopropyl-substi-
tuted compounds 3 and 4 confirmed the hypothesis of a
C–H/π interaction more extended than that provided by a
methyl group.[15]
2
3
4
6
7
Me
Et
iPr
MeO
MeS
47:53
56:44
69:31
21:79
17:83
46:54
58:42
87:13
20:80
12:88
The synthesis of adducts 4 and 7 was also performed in
toluene heated to reflux, a solvent in which the reactions
proceeded sluggishly (25 and 30% yield, by NMR analysis,
for 4 and 7 respectively). Although these low yields sug-
gested that the observed variations in ratio should be
treated with caution, it is interesting to note that in the case
of adduct 4 (R = iPr), almost equimolecular amounts of
endo and exo isomers was obtained (49:51), whereas in the
case of adduct 7 (R = SMe), the isomer ratio (85:15) was
virtually the same as that observed in TCE and AcOH. It
seemed plausible that, in the case of adduct 4, the aromatic
solvent effectively competed with the benzene ring of the
scaffold for CH/π interaction, whereas it did not have any
effect on the repulsive heteroatom/π interaction present in
adduct 7. Solvent effects do not seem to influence the calcu-
lated ΔE values. We optimized both endo- and exo-3 in ace-
tic acid (ΔE = –1.88 kJ/mol) and in 1,1,2-trichloroethane
(ΔE = –1.88 kJ/mol), with no significant differences with
the in vacuo data (ΔE = –1.72 kJ/mol; see Table 2). Simi-
larly, calculations for endo- and exo-6 in acetic acid (ΔE
= +4.18 kJ/mol) and 1,1,2-trichloroethane (ΔE = +4.06 kJ/
mol) agree with the value calculated in vacuo (ΔE =
+4.90 kJ/mol).
Table 2. Experimental and calculated (BMK/cc-pVDZ) isomer ra-
tios for compounds 2–4, 6, and 7.
R
Exp.
Exp.
Calcd.
Calcd.
endo/exo ΔG [kJ/mol] endo/exo ΔE [kJ/mol]
2
3
4
6
7
Me
Et
iPr
MeO
MeS
46:54
58:42
69:31
21:79
17:83
+0.50
–1.05
–2.59
+4.23
+6.48
50:50
63:37
67:33
18:82
13:87
0.00
–1.72
–2.26
+4.90
+6.15
Figure 2. BMK/cc-pVDZ minimized structures of adducts endo-2,
endo-3, and endo-4 (arrows indicate the interacting hydrogen
atoms).
To further investigate the relative importance of C–H/π
and heteroatom/π interactions, compounds 8–11 featuring
To check whether the endo/exo ratios for these molecular different ortho,orthoЈ-substituents on the N-aryl residue
balances were sensitive to the solvent in which the reactions were designed to provide systems capable of adopting these
were performed, the syntheses of compounds 2–4, 6, and 7 interactions in a competitive fashion (Scheme 2). These de-
were carried out under identical reaction conditions rivatives were obtained as described for 2–7 and in compar-
(115 °C, 18 h) in glacial acetic acid (Table 3). The products able yields; the isomer ratios were readily determined by
were obtained in very similar yields and atropisomeric ra- NMR spectroscopy (Table 4).[16]
Eur. J. Org. Chem. 2014, 4993–4998
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4995

L. Raimondi, M. Benaglia, F. Cozzi
FULL PAPER
The validity of this hypothesis was nicely confirmed by
the result obtained with adduct 11, which, in full agreement
with the expectations based on the results collected in
Table 1 and Table 4, was obtained as a 85:15 mixture of
isomers in favor of the ethyl-endo/methoxy-exo compound;
a ratio in reasonable agreement with that predicted by the
DFT calculation (93:7). A synergic effect between the ethyl
C–H/π favorable interaction and the methoxy/π repulsion
could explain the observed strong preference for the ethyl-
endo/methoxy-exo isomer in adduct 11.
As in the case of monosubstituted compounds, the reac-
tions leading to adducts 8–11 were also performed in AcOH
at 115 °C (Table 4). The only significant difference in isomer
ratio was observed in the case of compound 11, which in
AcOH was obtained even more enriched in the ethyl-endo/
methoxy-exo isomer (91:9 vs. 85:15). In this case, we did not
observe any solubility effect that could lead to occasional
enrichment in the preferred isomer of 11 (as previously ob-
served in the case of adduct 4, see above), thus it would be
tempting to rationalize this observation by considering that
the increased polarity of AcOH with respect to that of TCE
could favor the dispersive (hydrophobic) C–H/π interaction.
However, when the reaction was performed in toluene (30%
yield by NMR analysis), a solvent that should favor the
formation of the ethyl-exo isomer (see above the discussion
for compound 4), the isomer ratio remained unaffected
(85:15), and this observation strongly suggested that the ob-
Scheme 2. Structures of succinimides 8–11.
Table 4. Isolated yields and isomer ratios for succinimides 8–11
(chemical shifts of diagnostic signals are reported in the Experi-
mental Section).
R
R1
Yield [%][a]
R1-endo/exo ratio[b]
TCE (115 °C) AcOH (115 °C)
8
9
10
11
Me
F
MeO
MeO
Et
70
75
73
78
60:40
74:26
80:20
85:15
60:40
76:24
80:20
91:9
Me
Me
Et
[a] Isolated yield after flash chromatography; reactions carried out
in TCE. [b] Determined by 1H NMR analysis of the crude reaction
products (300 MHz, CDCl3).
The data presented in Table 4 indicate that, as expected served ratio variations in different solvents should be
on the basis of the results obtained with compounds 2 and treated with caution.
3, in the ethyl/methyl-substituted adduct 8, the larger ethyl
group occupied the more sterically congested endo position
preferentially over the smaller methyl group by a 60:40 ra-
tio. The agreement with the ratio calculated by DFT
Conclusions
(endo/exo ratio = 53:47) was reasonably good. As suggested
The following conclusions could be drawn from these ex-
before, the possibility of forming a more extended C–H/π periments. In these model systems: (i) the C–H/π interaction
interaction by the ethyl than by the methyl group could be is weak, mildly stabilizing, and capable of balancing (methyl
invoked to explain this observation.
group) or even overcoming (ethyl and isopropyl groups) un-
In methyl/fluorine substituted adduct 9, preferential for- favorable steric effects; (ii) ethyl and isopropyl groups seem
mation of the Me-endo isomer was observed in almost a 3:1 to interact with a π-system more favorably than a methyl
endo/exo ratio (Table 4, entry 2). In principle, the Me-endo group, because they provide a more extended C–H/π inter-
isomer could benefit from the occurrence of a C–H/π inter- action without becoming too sterically demanding; (iii) the
action and the absence of the repulsion between the F atom interaction between a lone-pair-bearing heteroatom (such
and the aromatic system. Given the slight preference for as F, O, and S) and a π system is electronically repulsive
Me-exo isomer formation in adduct 2, it seemed reasonable and sensitive to steric factors (the larger the heteroatom, the
to hypothesize a more significant role for fluorine/π elec- more repulsive the interaction); (iv) in model compounds
tronic repulsion than for C–H/π attraction in determining in which it is possible to selectively adopt a conformation
the isomer ratio observed for compound 9. The small size featuring a C–H/π rather than a lone pair/π interaction, the
of the fluorine atom, which is only marginally larger than conformation featuring the former interaction is largely pre-
a H atom (A = 0.15),[10] seems to support this hypothesis.
ferred over the latter; (v) succinimides such as 8–11 can rep-
The data obtained with the methoxy/methyl-substituted resent an easily accessible and versatile model system with
adduct 10 confirmed that the C–H/π interaction provided which to probe nonbonding interactions of different atoms
by a methyl group was only weakly stabilizing. Indeed, this and groups with a benzene ring.
compound was obtained in exactly the same MeO-exo/endo
Although we are aware that the results obtained with a
ratio observed in the case of adduct 6, as if the methyl given model system are difficult to generalize,[7b] we believe
C–H/π interaction was not capable of adding any effect fav- that these findings can be useful in assembling supramolec-
oring the MeO-exo isomer to the repulsion between the ular architectures, in controlling the conformation of flexi-
oxygen lone-pair and the π-system already present in ad- ble molecules, and in designing and fine-tuning molecular
duct 6.[17]
recognition events, for instance in medicinal chemistry.[1a,1d]
4996
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 4993–4998

Aliphatic C–H/π and Heteroatom/π Interactions
N-(2-Fluoro-6-methylphenyl)-3,4-(9,10-dihydroanthracene-9,10-di-
yl)succinimide (9): Prepared on the same scale as described for 8
from anhydride 1 and 2-fluoro-6-methylaniline in 75% yield. The
product was then purified by flash chromatography (hexane/ethyl
acetate, 75:25).
The conclusions drawn from this study can also help to in-
terpret and explain nonbonding interactions involving aro-
matic compounds in artificial and natural systems.[18]
N-(2-Methoxy-6-methylphenyl)-3,4-(9,10-dihydroanthracene-9,10-
diyl)succinimide (10): Prepared on the same scale as described for
8 from anhydride 1 and 2-methoxy-6-methylaniline in 73% yield.
The product was then purified by flash chromatography (hexane/
ethyl acetate, 75:25).
Experimental Section
General: All commercially available reagents including anhydrous
solvents were used as received. Organic extracts were dried with
sodium sulfate, filtered, and concentrated under vacuum using a
rotary evaporator. Nonvolatile materials were dried under high vac-
uum. Reactions were monitored by thin-layer chromatography on
precoated Merck silica gel 60 F254 plates and visualized either by
UV or by staining with a solution of cerium sulfate (1 g) and
ammonium heptamolybdate tetrahydrate (27 g) in a mixture of
water (469 mL) and concentrated sulfuric acid (31 mL). Flash
chromatography was performed on silica gel 60. NMR spectra were
obtained at 300 MHz for 1H, 75 MHz for 13C, and 282 MHz for
19F NMR with a Bruker 300 spectrometer. All anilines were com-
mercial products, except 2-ethyl-6-methoxyaniline, which was pre-
pared according to a reported procedure,[19] as was anhydride 1.[20]
Adducts 2 and 6 showed spectroscopic data in agreement with re-
ported values.[4]
N-(2-Ethyl-6-methoxylphenyl)-3,4-(9,10-dihydroanthracene-9,10-
diyl)succinimide (11): Prepared on the same scale as described for
8 from anhydride 1 and 2-ethyl-6-methoxyaniline in 78% yield. The
product was then purified by flash chromatography (hexane/ethyl
acetate, 70:30).
Supporting Information (see footnote on the first page of this arti-
1
cle): H and 13C NMR spectra as well as elemental analysis for all
new compounds, and Cartesian coordinates for DFT structures.
Acknowledgments
The authors thank Professor Jay S. Siegel for helpful discussions.
Computational: All calculations were performed in the gas phase at
the BMK/cc-pVDZ level of theory (see Cozzi et al.[6c] for a dis-
cussion) with the Gaussian 09 package.[21] All minima were charac-
terized as such by a complete vibrational analysis. The polarizable
continuum model for solvent calculations was used, as im-
plemented as a standard method in the Gaussian package.
[1] For reviews, see: a) E. A. Meyer, R. K. Castellano, F. Dieder-
ich, Angew. Chem. Int. Ed. 2003, 42, 1210–1250; Angew. Chem.
2003, 115, 1244; b) S. L. Cockroft, J. Perkins, C. Zonta, H.
Adams, S. E. Spey, C. M. R. Low, J. G. Vinter, K. R. Lawson,
C. J. Urch, C. A. Hunter, Org. Biomol. Chem. 2007, 5, 1062–
1080; c) H.-J. Schneider, Angew. Chem. Int. Ed. 2009, 48, 3924–
3977; Angew. Chem. 2009, 121, 3982; d) L. M. Salonen, M.
Ellermann, F. Diederich, Angew. Chem. Int. Ed. 2011, 50,
4808–4842; Angew. Chem. 2011, 123, 4908; e) M. L. Waters,
Curr. Opin. Chem. Biol. 2002, 6, 736–741. A recent issue of
Accounts of Chemical Research (Guest Ed.: M. L. Waters) has
been fully dedicated to aromatic interactions, see: Acc. Chem.
Res. 2013, 46, 873–1049. Of the reviews there included, particu-
larly relevant to the present manuscript are: f) J. L. Asensio, A.
Ardá, F. J. Cañada, J. Jimenez-Barbero, Acc. Chem. Res. 2013,
46, 946–954 (on carbohydrate–aromatic CH/π interactions); g)
N. J. Zondio, Acc. Chem. Res. 2013, 46, 1039–1049 (on elec-
tronically tunable CH/π interactions); h) E. H. Krenske, K. N.
Houk, Acc. Chem. Res. 2013, 46, 979–989 (on QM modeling
of aromatic interactions).
N-(2-Ethylphenyl)-3,4-(9,10-dihydroanthracene-9,10-diyl)succin-
imide (3): A stirred solution of anhydride 1 (0.298 g, 1.08 mmol)
and 2-ethylaniline (0.135 mL, 1.08 mmol) in TCE (3 mL) was
heated at 115 °C for 18 h. The solvent was evaporated under vac-
uum, and the residue was analyzed by 1H NMR spectroscopy in
CDCl3 to establish the endo/exo ratio. The crude product was then
purified by flash chromatography (hexane/ethyl acetate, 75:25) to
afford the product in 73% yield.
N-(2-Isopropylphenyl)-3,4-(9,10-dihydroanthracene-9,10-diyl)succin-
imide (4): Obtained in 71% yield by following the procedure de-
scribed for adduct 3. The product was purified by flash chromatog-
raphy (hexane/ethyl acetate, 2:1).
N-(2-tert-Butylphenyl)-3,4-(9,10-dihydroanthracene-9,10-diyl)suc-
cinimide (5): Obtained in 70% yield by following the procedure
described for adduct 3. The product was purified by flash
chromatography (hexane/ethyl acetate, 70:30).
[2] For a review on the use of model systems to quantify nonbond-
ing interactions involving aromatic compounds, see: a) I. K.
Mati, S. L. Cockroft, Chem. Soc. Rev. 2010, 39, 4195–4205. For
recent examples not included in this review, see: b) B. Bhayana,
J. Org. Chem. 2013, 78, 6758–6762; c) K. B. Muchowska, C.
Adam, I. K. Mati, S. L. Cockroft, J. Am. Chem. Soc. 2013, 135,
9976–9979; d) H. Gardarsson, W. B. Schweizer, N. Trapp, F.
Diederich, Chem. Eur. J. 2014, 20, 4608–4616.
N-(2-Methylthiophenyl)-3,4-(9,10-dihydroanthracene-9,10-diyl)suc-
cinimide (7): Obtained in 77% yield by following the procedure
described for adduct 3. The product was purified by flash
chromatography (hexane/ethyl acetate, 80:20).
[3] a) E. Weber, S. Finge, I. Csöregh, J. Org. Chem. 1991, 56, 7281–
7288; b) A. Srivastava, V. Srivastava, S. M. Verma, J. Org.
Chem. 1994, 59, 3560–3563.
N-(2-Ethyl-6-methylphenyl)-3,4-(9,10-dihydroanthracene-9,10-diyl)-
succinimide (8): A stirred solution of anhydride 1 (0.552 g, 2 mmol)
and 2-ethyl-6-methylaniline (0.280 mL, 2 mmol) in TCE (5 mL)
was heated at 115 °C for 18 h. The solvent was evaporated under
vacuum, the solid residue was dissolved in dichloromethane
(25 mL), and this solution was washed with water (25 mL), a satu-
rated aqueous solution of NaHCO3 (25 mL), and brine (25 mL).
The organic phase was dried and concentrated under vacuum to
afford a pale-gray solid that was analyzed by 1H NMR spec-
troscopy to establish the endo/exo ratio. The product was then puri-
fied by flash chromatography (hexane/ethyl acetate, 60:40).
[4] a) S. M. Verma, N. B. Singh, Aust. J. Chem. 1976, 29, 295–300;
b) S. Mahanti, S. M. Verma, Indian J. Chem. Sect. B: Org.
Chem. Incl. Med. Chem. 1982, 21, 1098–1101; c) A. K. Verma,
S. Mahanti, S. M. Verma, Indian J. Chem. Sect. B: Org. Chem.
Incl. Med. Chem. 1989, 28, 457–461; d) K. Kishikawa, K. Yosh-
izaki, S. Kohmoto, M. Yamamoto, K. Yamaguchi, K. Yamada,
J. Chem. Soc. Perkin Trans. 1 1997, 1233–1239; e) G. Gross-
mann, M. J. Potrzebowski, S. Olejniczak, N. E. Ziolkowska,
G. D. Bujacz, W. Cieselski, W. Prezdo, V. Nazarov, V. Golovko,
New J. Chem. 2003, 27, 1095–1101; f) W. R. Carroll, P. Pellec-
chia, K. D. Shimizu, Org. Lett. 2008, 10, 3547–3550; g) C.
Eur. J. Org. Chem. 2014, 4993–4998
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4997

L. Raimondi, M. Benaglia, F. Cozzi
FULL PAPER
Zhao, R. M. Parish, M. D. Smith, P. J. Pellecchia, C. D. Sher-
rill, K. D. Shimizu, J. Am. Chem. Soc. 2012, 134, 14306–14309;
in this article the crystallographic distance between the arene
plane and the H atom of a methyl group pointing towards the
benzene ring was reported to be 2.61 Å.
For a review on C-H/π interactions, see: a) M. Nishio, Y. Ume-
zawa, K. Honda, S. Tsuboyama, H. Suezawa, CrystEngComm
2009, 11, 1757–1788; and references cited therein. It was shown
that this interaction is dominated by dispersion and C-H/π con-
tacts cannot be considered hydrogen bonds, see: b) S. Tsuzuki,
A. Fujii, Phys. Chem. Chem. Phys. 2008, 10, 2584–2594.
For a review on arene–arene interactions, see: a) F. Cozzi, J. S.
Siegel, Pure Appl. Chem. 1995, 67, 683–689; for articles not
included in this review, see: b) F. Cozzi, R. Annunziata, M.
Benaglia, M. Cinquini, L. Raimondi, K. K. Baldridge, J. S. Sie-
gel, Org. Biomol. Chem. 2003, 1, 157–162; c) F. Cozzi, R. An-
nunziata, M. Benaglia, K. K. Baldridge, G. Auirre, J. Estrada,
Y. Sritana-Anant, J. S. Siegel, Phys. Chem. Chem. Phys. 2008,
10, 2686–2694; d) J. L. Xia, S. H. Liu, F. Cozzi, M. Mancinelli,
A. Mazzanti, Chem. Eur. J. 2012, 18, 3611–3620.
For heteroarene–arene interactions, see: a) R. Annunziata, M.
Benaglia, F. Cozzi, A. Mazzanti, Chem. Eur. J. 2009, 15, 4373–
4381; b) M. Benaglia, F. Cozzi, M. Mancinelli, A. Mazzanti,
Chem. Eur. J. 2010, 16, 7456–7468.
able results, in agreement with a previous work of ours (see
ref.[6c]).
[15] One of the referees, very correctly, pointed out that in Table 2
we should compare the experimental ΔG values with the calcu-
lated values, since we fully characterized vibrationally all opti-
mized structures. However, calculated ΔG values were found to
be less reliable than ΔE values. This is probably due to the
harmonic approximation used in the vibrational analysis, an
approximation that fails with low-energy modes of vibration.
Indeed, comparison of low-energy frequencies for the endo and
exo pairs show systematic differences, probably due to the dif-
ferent steric hindrance of the endo and exo regions. Thus, the
anharmonic contribution is probably different in the two atrop-
isomers, and this error affects the calculated ΔG values.
[16] The isolation of nonequilibrium isomeric mixtures of adducts
8–10 by flash chromatographic separation of the isomers al-
lowed the energy barriers to isomer interconversion to be deter-
mined by heating the samples in [D6]DMSO and monitoring
the isomerization process by 1H NMR analysis. As expected
on the basis of purely steric considerations, adduct 8 showed
the highest barrier (ΔG϶ = 144.0 kJmol–1 at 123 °C) followed
by compound 10 (ΔG϶ = 132.6 kJmol–1 at 125 °C), and 9 (ΔG϶
= 123.8 kJmol–1 at 92 °C). These values were in agreement with
those reported in the literature for similar adducts,[4] and con-
firmed the sensitivity of our model systems to steric effects.
[17] It is worth mentioning that compound 10 was also obtained in
lower yields (about 40%) and identical isomer ratio by stirring
overnight an equimolar mixture of anthracene and N-2-meth-
oxy-6-methylphenylmaleimide in o-xylene heated to reflux,
thus showing that the isomer ratio was not affected by the
method of preparation and confirming that the products were
indeed obtained as equilibrium mixtures.
[5]
[6]
[7]
[8]
[9]
For cation–arene interactions, see: K. K. Baldridge, F. Cozzi,
J. S. Siegel, Angew. Chem. Int. Ed. 2012, 51, 2903–2906; Angew.
Chem. 2012, 124, 2957.
When differently enriched samples of the nonequilibrium iso-
meric mixtures of adducts 4 and 6, isolated by flash chromatog-
raphy, were heated at 115 °C for 18 h in TCE endo/exo mixtures
having the same composition of the crude products were ob-
tained. These experiments, as suggested by one of the referees,
allowed us to establish that the mixture of atropisomers ob-
tained from the synthesis were indeed equilibrium mixtures and
confirmed the validity of NMR analysis of the crude products
as a method to establish their endo/exo ratios.
[18] For a recent example of the occurrence of aliphatic C-H/π in-
teractions between carbohydrates and aromatic systems, see: a)
A. G. Santana, E. Jimenez-Moreno, A. M. Gomez, F. Corzana,
C. Gonzalez, G. Jimenez-Oses, J. Jimenez-Barbero, J. L. Asen-
sio, J. Am. Chem. Soc. 2013, 135, 3347–3350. For a recent study
on C-H/π interactions, see: N. A. Seifert, D. P. Zaleski, C.
Perez, J. L. Neill, B. H. Pate, M. Vallejo-Lopez, A. Lesarri, E. J.
Cocinero, F. Castano, I. Kleiner, Angew. Chem. Int. Ed. 2014,
53, 3210–3213.
[10] A values: methyl, 1.70; ethyl, 1.75; isopropyl, 2.15; methoxy,
0.60; methylthio, 0.70 (J. A. Hirsch, Top. Stereochem. 1967, 1,
199–222). B values: methyl, 7.4; ethyl, 8.7; isopropyl, 11.1;
methoxy, 5.6 (R. Ruzziconi, S. Spizzichino, L. Lunazzi, A.
Mazzanti, M. Schlosser, Chem. Eur. J. 2009, 15, 2645–2652).
[11] a) B. W. Gung, Y. Zou, Z. Xu, J. C. Amicangelo, D. G. Irwin,
S. Ma, H.-C. Zhou, J. Org. Chem. 2008, 73, 689–693; b) J. C.
Amicangelo, B. W. Gung, D. G. Irwin, N. C. Romano, Phys.
Chem. Chem. Phys. 2008, 10, 2695–2705.
[19] L. Quaranta, O. Corminboeuf, P. Renaud, Org. Lett. 2002, 4,
39–41.
[20] W. E. Bachmann, M. C. Kloetzel, J. Am. Chem. Soc. 1938, 60,
481–485.
[21] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starov-
erov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Ad-
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
revision D.01, Gaussian, Inc., Wallingford CT, 2009.
Received: March 20, 2014
[12] W. B. Motherwell, J. Moise, A. E. Aliev, M. Nic, S. J. Coles,
P. H. Horton, M. B. Hursthouse, G. Chessari, C. A. Hunter,
J. G. Vinter, Angew. Chem. Int. Ed. 2007, 46, 7823–7826; An-
gew. Chem. 2007, 119, 7969.
[13] For a review, see: M. Egli, S. Sarkhel, Acc. Chem. Res. 2007,
40, 197–205.
[14] It is beyond the scope of this work to discuss in depth the
reliabilities of different DFT methods in treating noncovalent
interactions (see for example Krenske and Houk,[1h] and R.
Peverati, K. K. Baldridge, J. Chem. Theory Comput. 2009, 5,
2772–2786). Despite all their drawbacks, some DFT function-
als provide a qualitatively reliable evaluation of π-π and CH-π
interactions. We tested for our purposes some of the most com-
mon functionals (the most popular B3LYP, B2PLYP, M06-2X
and BMK) with different basis sets [from 6–31G(d,p) to
cc-pVDZ]; in our case, BMK/cc-pVDZ data gave the most reli-
Published Online: July 11, 2014
4998
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 4993–4998