Controlling selectivity for cycloadditions of nitrones and alkenes tethered by benzimidazoles: Combining experiment and theory

Article abstract of DOI:10.1002/ejoc.200801211

Herein we describe a combined experimental/theoretical study on the effects of substituents on regio- and stereoselectivity in intramolecular 1,3-dipolar cycloadditions of nitrones and alkenes tethered by benzimidazoles. By employing a large substituent a

Full text of DOI:10.1002/ejoc.200801211

FULL PAPER
DOI: 10.1002/ejoc.200801211
Controlling Selectivity for Cycloadditions of Nitrones and Alkenes Tethered by
Benzimidazoles: Combining Experiment and Theory
Liping Meng,[a] Selina C. Wang,[a] James C. Fettinger,[a] Mark J. Kurth,*[a] and
Dean J. Tantillo*[a]
Keywords: Nitrones / Cycloaddition / Quantum chemical calculations / Stereoselectivity / Substituent effects / Transition
states
Herein we describe a combined experimental/theoretical
study on the effects of substituents on regio- and stereoselec-
tivity in intramolecular 1,3-dipolar cycloadditions of nitrones
and alkenes tethered by benzimidazoles. By employing a
large substituent at position R2 or R3, complete selectivity
was achieved for either the fused or bridged cycloadduct,
respectively. In addition, these cycloadducts were formed as
single diastereomers in all of the cycloadditions examined.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
ments. The success of this approach then led to computa-
tion-based predictions of product distributions for ad-
ditional reactions, which were put to the test experimentally.
Ultimately, a synergistic back-and-forth between synthetic
and theoretical chemistry, with each suggesting new experi-
ments for the other, led to a complete picture of the reactiv-
ity for the systems shown in Scheme 1.
Introduction
Dipolar cycloadditions play a central role in modern het-
erocycle synthesis.[1] Many variations in both the dipole and
dipolarophile cycloaddition partners have been explored, as
have inter- and intramolecular reaction manifolds, and the
effects of various promoters. Although nitrone-alkene cy-
cloaddition reactions have been utilized in the synthesis of
many isoxaolidines, including those which form the cores
of potential antifungal and antibacterial agents (e.g., 1),[2,3]
some aspects of this chemistry would benefit from more
detailed understanding. Herein we describe a combined ex-
perimental/theoretical study on the family of intramolecu-
lar nitrone-alkene cycloadditions shown in Scheme 1.[4]
This study originated from a synthetic project aimed at
constructing libraries of conformationally constrained[5]
isoxazolidines with potential biological activity.[6] Quantum
chemical calculations were employed to help rationalize the
regio- and stereoselectivity observed in these initial experi-
Scheme 1.
Results and Discussion
[a] Department of Chemistry, University of California, Davis,
One Shields Avenue, Davis, CA 95616, USA
E-mail: djtantillo@ucdavis.edu
Overview
Experimentally determined product distributions and
computed relative energies of transition-state structures
leading to each product (see Exp. Sect. for computational
mjkurth@ucdavis.edu
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
1578
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1578–1584

Intramolecular Dipolar Cycloadditions
Table 1. Summary of experimental yields[a] and computed transition-state energies[b] for systems with a variety of substitution patterns.[c]
Entry
R1
R2
R3
R4
% Yield of A
% Yield of B TS for A [kcal/mol] TS for B [kcal/mol]
1
2
3
4
5
6
7
8
H
H
H
H
H
H
Me
Ph
tBu
tBu
H
H
H
H
H
H
H
H
H
H
H
tBu
tBu
Me
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
Me
tBu
Ph
Ph
Me
iPr
tBu
cHex
Bn
63
61
69
79
64
64
49
17
0
0
0
0
9
25
37
28
19
31
21
31
69
0
44
96
97
85
97
0
20.3
–
–
22.8
–
22.5
–
–
26.4
–
Ph
17.8
25.8
24.0
30.0
28.3
30.7
23.5
31.5
23.5
23.1
21.2
22.4
21.6
21.3
19.2
25.8
25.2
26.5
23.8
24.2
19.3
27.8
19.3
28.3
25.5
29.7
27.0
25.7
cHex
cHex
cHex
Me
cHex
Me
cHex
Ph
cHex
Me
9
10
11
12
13
14
15
16
17
18
19
H
Me
Me
H
H
H
H
H
3
98
96
97
96
96
0
0
0
0
cHex
cHex
Me
[a] The yields are isolated yields from experiments carried out in ethanol with microwave irradiation at 80 °C for 30 min.[7] [b] B3LYP/6-
31G(d); energies are relative to the lowest energy nitrone conformer found for each system (i.e., various conformations of the nitrone-
and alkene-containing chains were examined).[7] [c] Bold values correspond to the major product observed experimentally or predicted
theoretically.
methods) are shown in Table 1 for the simplest systems the importance of the substituents R1–R4 in controlling the
studied experimentally.[7] Although mixtures of products A A vs. B selectivity. These experiments show that either of
and B were sometimes observed, these products were the polycyclic scaffolds found in products A and B can be
formed as single diastereomers. For nearly all cases (all but obtained to the exclusion of the other if substituents are
Entries 7 and 8), the transition-state structure with the chosen judiciously.
lower computed energy corresponds to the experimentally
observed major product.[8] For Entry 7, no selectivity is pre-
Synthesis of Reactants
dicted computationally, but a slight selectivity is observed
experimentally. For Entry 8, a preference for product A is
predicted computationally, but a preference for B is ob-
served experimentally. Note that these two cases are the two
cases for which the computed transition-state-structure en-
ergies are closest to each other; this shows the limits of our
computational approach. Note also that our calculations
were not biased by knowledge of the experimental results;
in fact, in some cases, the computational prediction pre-
ceded the experiment. Overall, the agreement between
theory and experiment is excellent. The sections that follow
highlight various aspects of the chemistry that is summa-
rized in Table 1.
The requisite 1-substituted allyl-1H-benzo[d]imidazole-2-
carbaldehydes (nitrone precursors) were prepared from o-
diaminobenzene in analogy to our previous work[6] as
shown in Scheme 2. Commercially unavailable allylic bro-
mides or mesylates were prepared according to reported
methods.[9–17] It is worth mentioning that decarbon-
ylation[18] was observed in some cases during conversion
of the acetal to the aldehyde. This side reaction could be
Closely related dipolar cycloadditions have been de-
scribed previously.[4] Of particular note is the seminal work
of Zecchi, Beccalli, Broggini, and co-workers. These au-
thors examined systems with imidazole (R1 = R2 = R3 =
H, R4 = Bn or phenylethyl),[4a] pyrrole or butylpyrrole (R1
= H, Me, Pr, Ph, R2 = H, Me, R3 = H, R4 = Bn, phenyl-
ethyl),[4b] and indole (R1 = Ph, R2 = H, R3 = H, R4
=
H, Bn, phenylethyl) tethers.[4c] Our results on systems with
substitution patterns similar to these further suggest that
the heterocycle used as a tether does not significantly influ-
ence the selectivity of the cycloaddition reactions. Our ex-
periments and quantum chemical calculations involve sys-
tems with a variety of additional substitution patterns – sys-
tems that were explored in a systematic attempt to clarify Scheme 2.
Eur. J. Org. Chem. 2009, 1578–1584
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1579

L. Meng, S. C. Wang, J. C. Fettinger, M. J. Kurth, D. J. Tantillo
FULL PAPER
suppressed to some extent by using more dilute HCl. De- structure verified by X-ray crystallography (Figure 2).[4c,19]
tails on the specific conditions used and yields for these Additional heating at a higher temperature (120 °C, EtOH,
reactions can be found in the Supporting Information.
µW, 45 min) led to the rearrangement of the nitrone func-
tional group to an amide (and to a lesser amount of an
ester; Figure 2).[20,21] Moreover, when the R1 = tBu, R2 =
R3 = H, R4 = cHex system (Table 1, Entry 9) was subjected
to the usual reaction conditions, no cycloadducts were iso-
lated (the importance of the R1 and R4 substituents are
discussed below). NMR spectroscopic data (see Supporting
Information) suggested that a nitrone was formed but again
did not react further, and additional heating at a higher
temperature again led to an amide product (the same is true
for the terminal alkyne analogue of the system shown in
Figure 2; see Supporting Information). Similarly, the R1 =
tBu, R2 = R3 = H, R4 = Me system (Table 1, Entry 10) also
led to a mixture of nitrone isomers (54%) in addition to
product B (see Supporting Information).
Cycloaddition Reactions
The intramolecular cycloadditions discussed herein
(Scheme 1 and Table 1) were carried out in ethanol with
microwave irradiation at 80 °C for 30 min, unless otherwise
noted. Note that nitrones were generated in situ by the con-
densation of a given hydroxylamine (HONHR4) with the
carbaldehyde reactant. Preliminary experiments utilized
conventional heating (as was done in most previous studies
on closely related systems[4a–4c]; our systems were refluxed
in ethanol for 24 h; see Supporting Information for details),
but using microwave irradiation reduced the reaction times
while leading to similar yields.[4d] No significant changes in
selectivity were observed between the conventionally heated
and microwave-irradiated reactions. All products were thor-
oughly characterized by NMR and some product assign-
ments were corroborated by X-ray diffractometric analysis
(Figure 1) and/or chemical shift assignments derived from
quantum chemical calculations (see Supporting Information).
Figure 2. Attempted cycloaddition of an alkyne-substituted reac-
tant and X-ray structure of the nitrone product isolated.
Our calculations suggest that the stereochemistry of a
given nitrone (E or Z) is directly translated into which type
of cycloaddition (formation of product A or B) occurs. In
short, for geometric reasons, a (Z)-nitrone leads directly to
the transition-state structure for formation of product A,
while an (E)-nitrone leads directly to the transition-state
structure for formation of product B (in the specific benz-
imidazole-tethered systems we describe herein); i.e., no
transition-state structures with different relative orienta-
tions of the dipole and alkene (endo/exo) were located. Rep-
Figure 1. X-ray structures of several cycloadducts (see Table 1, En-
tries 4, 11, and 15).
Connecting Nitrone and Product Structures
For the cycloadditions described above, nitrones were resentative transition-state structures for the formation of
generated in situ. Consistent with this assumption, when products A and B are shown in Figure 3 (for the R1 = R2
the alkyne (rather than alkene) substituted system shown in = R3 = H, R4 = Me system). The (E) or (Z) stereochemistry
Figure 2 was subjected to the reaction conditions, cycload- of the nitrone precursors is apparent in these transition-
dition did not occur, but a nitrone was isolated and its state structures. For no case was a transition-state structure
1580
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1578–1584

Intramolecular Dipolar Cycloadditions
for formation of product A found with an (E)-like nitrone that hydroxylamine exchange can occur under our reaction
substructure, nor was any transition-state structure for for- conditions (e.g., excess hydroxylamine),[26] which provides
mation of product B found with a (Z)-like nitrone substruc- an avenue for (E)- and (Z)-nitrone interconversion.
ture.
Scheme 4.
The Influence of R1
Comparison of Entries 4, 7, 8, and 9 in Table 1 reveals
the influence of R1 on the relative amounts of products A
and B. The experimental results demonstrate that as the size
of R1 gets larger (from H to Me to Ph, as R2, R3, and
R4 remain constant), the product distribution shifts steadily
Figure 3. Computed [B3LYP/6-31G(d)] transition-state structures
for cycloadditions leading to products A and B with R1 = R2 = R3
from a preponderance of A (see the discussion of R3 below
= H, R4 = Me. Selected distances are shown in Å.
for an explanation for this inherent preference) to a prepon-
derance of B. These cases are the ones for which our calcu-
lations of the transition-state energy deviate most (albeit
slightly) from the experimental results. When R1 = tBu, no
reaction occurs. Steric clashes involving R1 are more signifi-
cant for transition-state structures leading to product A (see
Figure 3) – in transition structures leading to product B,
R1 points away from the rest of the molecule – selectively
destabilizing this structure when R1 is large, allowing prod-
uct B to be produced preferentially.
The direct translation of nitrone stereochemistry into
product structure, coupled with the strong correspondence
between experimentally observed selectivity and predicted
relative energies of transition-state structures[22] (vide supra)
suggests that the cycloaddition reactions occur under Cur-
tin–Hammett conditions,[23] i.e. the nitrone precursors in-
terconvert rapidly under the reaction conditions, allowing
the product distributions to reflect the relative energies of
the competing transition-state structures. What is the
mechanism by which nitrone isomerization takes place? Our
calculations on a small model system (Scheme 3) suggest
that simple rotation about the C=N bond (we were unable
to locate any transition-state structure for such a process) or
formation of a diradical or oxaziridine intermediate under
thermal conditions (barriers for each of Ͼ50 kcal/mol) are
unlikely (photochemical generation of such species cannot
be definitively ruled out for our reaction conditions).[20,21,24]
Scheme 4 shows an experiment that suggests another pos-
sible pathway for isomerization. A nitrone with R4 = Me
but not bearing an alkene was synthesized (Z:E, 3:1 in [D6]-
DMSO; Z:E, 1.1:1 in [D6]benzene)[25a] and subjected to the
usual reaction conditions for cycloaddition in the presence
of a hydroxylamine with a different R4 group (cHex). Un-
der these conditions, the nitrone bearing the cHex group
was formed in 80% yield as the (Z)-nitrone,[25b] suggesting
The Influence of R2
Comparison of Entries 4 and 11, where the only differ-
ence is the identity of R2, shows that the size of R2 has a
dramatic effect on the relative proportions of products A
and B. When R2 is small (Entry 4), product A predominates
(this is also true for Entries 1, 2, 3, 5 and 6), but when R2
is large (Entry 11; see also Entry 12), only product B is
observed. This effect is mirrored by the computed differ-
ences in transition-state-structure energies. Again, this ap-
pears to be a steric effect that selectively destabilizes transi-
tion-state structures leading to A (see Figure 3). When both
R1 and R2 are larger than H, even when they are as small
as methyl groups (Entries 13 and 14), product B again pre-
dominates.
The Influence of R3
The importance of R3 is revealed through a comparison
of Entries 4, 15, 17, and 18 (where R1, R2, and R4 remain
constant). Formation of product A is inherently preferred
Scheme 3.
Eur. J. Org. Chem. 2009, 1578–1584
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1581

L. Meng, S. C. Wang, J. C. Fettinger, M. J. Kurth, D. J. Tantillo
FULL PAPER
over formation of product B (e.g., see Entry 4, where R3 =
H). Increasing the size of R3 (Entries 15, 17, and 18), how-
ever, leads to complete selectivity in favor of product A.
This size effect is also manifested in the computed differ-
ences in transition-state-structure energies. As shown in
Figure 3, R3 is roughly eclipsed with a C–H bond at the
other end of the forming C–C bond in transition-state
structures leading to B, an unfavourable steric situation that
is not present in transition-state structures leading to A.[27]
In that the effects of having substituents at positions R2 and
R3 on the cycloaddition selectivity are opposite, we decided
to examine a system with substituents at both positions,
here directly connected to each other – the cyclo-
pentene shown in Scheme 5. For this system, the influence
of neither position dominates, leading to a mixture of both
cycloaddition products, in accord with the predicted
difference of transition-state-structure energies of only
0.1 kcal/mol. These products were produced in low yield,
however, with unreacted nitrone being formed as the major
product.
Scheme 6.
11 vs. 12 and 13 vs. 14 for A and Entries 15 vs. 16 and 18
vs. 19 for B). These experiments show that the selectivity is
affected only slightly by changing R4, even for cases with
substituents at R1, R2, or R3.
Conclusions
We have demonstrated the utility of microwave-
assisted[31] intramolecular nitrone-alkene dipolar cycload-
ditions for the synthesis of two polyheterocyclic scaffolds
that position their arrays of functional groups in well-de-
fined but different orientations. A combination of experi-
ment and theory was used to reveal general trends that can
be used to control and predict the relative amounts of each
type of product for these and closely related systems.[4] By
changing the substituents on the dipolarophile, it is possible
to achieve complete selectivity for either product: large sub-
stituents at position R3 favor product A and large substitu-
ents at position R2 favor product B.
Experimental Section
General: See ref.[7]
Representative Procedure for Cycloaddition: Reaction of 1-Allyl-1H-
benzimidazole-2-carbaldehyde with N-Cyclohexylhydroxylamine
(Entry 4): In a pyrex test tube, 1-allyl-1H-benzimidazole-2-carbal-
dehyde (74 mg, 0.4 mmol), triethylamine (0.06 mL, 0.44 mmol),
and N-cyclohexylhydroxylamine hydrochloride (67 mg, 0.44 mmol)
in absolute ethanol (5 mL) was submitted to microwave irradiation
at 80 °C for 30 min. The solvent was removed under vacuum and
the residue extracted with chloroform, washed with water and dried
with sodium sulfate. After evaporation of the solvent, the resultant
solid was purified by column chromatography on SiO2 with the
eluent of ethyl acetate and ethanol (100% to 90% ethyl acetate in
gradient) to obtain two products. The first fraction gave product B
and the second fraction contained product A.
Scheme 5.
(1R*,4S*)-2-Cyclohexyl-1,2,4,5-tetrahydro-1,4-methano[1,2,5]oxa-
diazepino[5,4-a]benzimidazole (Entry 4, Product A): Off-white solid
The Influence of R4
(90 mg, 79%); m.p. 224–225 °C. IR (neat): νmax = 3058, 2925, 2852,
˜
As shown in Table 1, most of our experiments involved
systems where R4 = cHex. We did, however, explore the
influence of this group. In particular, the identity of R4 was
varied for systems where R1 = R2 = R3 = H (Entries 1–6).
The observed ratios of products A and B differed only
slightly as R4 was changed. As expected, when R4 = H,
no cycloaddition was observed and the oxime product was
isolated instead (Scheme 6; see Supporting Information for
1
1618, 1522, 1479, 1450, 1418, 1275, 1226, 1170, 776, 737 cm–1. H
NMR (600 MHz, CDCl3, 23 °C): δ = 7.740 (br. s, 1 H), 7.253 (br.
s, 3 H), 4.903 (br. s, 1 H), 4.751 (br. s, 1 H), 4.138 (br. d, 1 H, J =
12 Hz), 3.950 (br. d, 1 H, J = 9.0 Hz), 2.889 (br. s, 1 H), 2.306 (br.
d, 1 H, J = 8.4 Hz), 2.232 (br. s, 1 H), 1.105–2.000 (m, 10 H) ppm.
13C NMR (150 MHz, CDCl3, 23 °C): δ = 150.6, 143.0, 134.6, 123.1,
122.5, 120.2, 109.3, 71.1, 63.0, 55.7, 50.8, 36.2, 31.7, 30.3, 26.0,
24.6, 24.2 ppm. ESI MS: m/z = 284.12 [M + H]+. Purity was deter-
details). This is clearly the result of N-deprotonation, which mined to be 100% by HPLC analysis on the basis of absorportion
is not possible for the other systems.[28]
at 214 nm.
Several other comparisons were also made to see whether
(3aR*,10bR*)-1-Cyclohexyl-1,3a,4,10b-tetrahydro-3H-isoxazolo-
[3Ј,4Ј:3,4]pyrrolo[1,2-a]benzimidazole (Entry 4, Product B): Off-
or not changing R4 from cHex affects the robustness of the
selectivity for some of the most selective cases (e.g. Entries white solid (22 mg, 19%); m.p. 200 °C (dec.). IR (neat): ν
=
˜
max
1582
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1578–1584

Intramolecular Dipolar Cycloadditions
2765–2810; b) V. Nair, T. D. Suja, Tetrahedron 2007, 63, 12247–
12275; c) A. Padwa, S. K. Bur, Tetrahedron 2007, 63, 5341–
5378; d) D. St. C. Black, R. F. Crozier, V. C. Davis, Synthesis
1975, 205–221.
R. R. K. Kumar, H. Mallesha, S. K. Rangappe, Arch. Pharm.
(Weinheim, Ger.) 2003, 336, 159–164.
3053, 2934, 2856, 1620, 1526, 1487, 1456, 1420, 1329, 1220, 1177,
930, 764, 743 cm–1; 1H NMR (300 MHz, CDCl3, 23 °C): δ = 7.730–
7.771 (m, 1 H), 7.180–7.295 (m, 3 H), 4.903 (d, 1 H, J = 7.5 Hz),
4.127–4.224 (m, 2 H), 3.907–4.029 (m, 2 H), 3.838 (dd, 1 H, J =
9 Hz, 3 Hz), 2.820 (br. s, 1 H), 1.200–2.168 (m, 10 H) ppm. 13C
NMR (75 MHz, CDCl3, 23 °C): δ = 158.1, 149.3, 132.1, 122.6,
122.4, 120.7, 110.0, 72.1, 63.4, 62.4, 49.8, 47.8, 31.6, 30.1, 26.1,
24.8, 24.7 ppm. ESI-MS: m/z = 284.12 [M + H]+. Purity was deter-
mined to be 97% by HPLC analysis on the basis of absorportion
at 214 nm.
[2]
[3]
[4]
B. G. Mullen, R. T. Decory, T. J. Mitchell, D. S. Allen, R. C.
Kinsolving, J. Med. Chem. 1988, 31, 2008–2014.
Closely related systems: a) with tether imidazole, R1 = R2
=
R3 = H, R4 = Bn or phenylethyl: E. Beccalli, G. Broggini, A.
Contini, I. De Marchi, G. Zecchi, C. Zoni, Tetrahedron: Asym-
metry 2004, 15, 3181–3187 (gas phase MP2/6-311++G**//HF/
6-31G** and MP2/6-31+G**//HF/6-31G** calculations were
reported for two systems in this paper) and D. Basso, G. Brog-
gini, D. Passarella, T. Pilati, A. Terraneo, G. Zecchi, Tetrahe-
dron 2002, 58, 4445–4450 [consistent with our results (see
Table 1), product A predominated for the two examples re-
ported therein without additional allylic substituents]; b) with
tether pyrrole or butylpyrrole, R1 = H, Me, Pr, Ph, R2 = H,
Me, R3 = H, R4 = Bn, phenylethyl: G. Broggini, C. La Rosa,
T. Pilati, A. Terraneo, G. Zecchi, Tetrahedron 2001, 57, 8323–
8332; A. Arnone, G. Broggini, D. Passarella, A. Terraneo, G.
Zecchi, J. Org. Chem. 1998, 63, 9279–9284 [consistent with our
results (see Table 1), the six examples reported therein with R2
and/or R3 larger than H led to product B as the major product,
while two of the three examples with R1 = R2 = R3 = H led to
Computations: GAUSSIAN03 was employed for all calculations.[32]
All geometries were optimized at the B3LYP/6-31G(d) level of
theory.[33] Structures for the system with R1 = R2 = R3 = H, R4 =
Me were also optimized at the B3LYP/6-31+G(d,p) level of theory
and no significant differences from the lower level of theory were
observed (see Supporting Information for details). All stationary
points were characterized as minima or transition structures by an-
alyzing their vibrational frequencies (minima had only real fre-
quencies and transition-state structures had one imaginary fre-
quency). In select cases, intrinsic reaction coordinate (IRC) calcula-
tions were used to further characterize the identity of transition
structures (see Supporting Information for details).[34] All reported
energies include zero-point energy corrections from frequency cal-
culations, scaled by 0.9806.[35] Transition-state energies in Table 1
are all relative to the lowest energy nitrone conformer found for a
given system (see Supporting Information for structures). Due to
the complexities associated with conformer distributions for nitro-
nes, as well as the fact that these species are generated in situ experi-
mentally, activation barriers are not discussed in detail in the text.
Structures for the system with R1 = R2 = R3 = H, R4 = Me were
also optimized at the CPCM(UAKS)-B3LYP/6-31+G(d,p)[36] level
of theory to assess the effects of solvation (ethanol and chloroform
were tested) and only minor differences from the gas-phase calcula-
tions were observed (see Supporting Information for details).
Structural drawings in Figure 3 were produced using Ball &
Stick.[37]
product A as the major product]; c) with tether indole, R1
=
Ph, R2 = H, R3 = H, R4 = H, Bn, phenylethyl: E. M. Beccalli,
G. Broggini, C. La Rosa, D. Passarella, T. Pilati, A. Terraneo,
G. Zecchi, J. Org. Chem. 2000, 65, 8924–8932 [consistent with
our results (see Table 1), for the two cases reported therein with
R1 = R2 = R3 = H, product A was the major product, and for
the two cases with R1 = Ph, product B was the major product].
Note that with a 3-methylindole tether, products of type B were
reported for R1 = R2 = R3 = H; see: D. St. C. Black, D. C.
Craig, R. B. Deb-Das, N. Kumar, Aust. J. Chem. 1993, 46, 603–
622; P. J. Bhuyan, R. C. Boruah, J. S. Sandhu, Tetrahedron Lett.
1989, 30, 1421–1422; d) For a system with a longer (aniline-
derived) tether, see: G. Broggini, F. Colombo, I. De Marchi, S.
Galli, M. Martinelli, G. Zecchi, Tetrahedron: Asymmetry 2007,
18, 1495–1501; e) For systems with cyclohexenes directly at-
tached to the N of a pyrrole or indole tether, see: E. M.
Beccalli, G. Broggini, A. Farina, L. Malpezzi, A. Terraneo, G.
Zecchi, Eur. J. Org. Chem. 2002, 2080–2086; in these cases, only
products of type B were observed; f) For systems with inverted
pyrrole tethers [i.e., CHR–C=N+(O–)Bn attached to N and a
vinyl group attached to carbon], see: E. Borsini, G. Broggini,
A. Contini, G. Zecchi, Eur. J. Org. Chem. 2008, 2808–2816.
M. A. Marx, A. L. Grillot, C. T. Louer, K. A. Beaver, P. A.
Bartlett, J. Am. Chem. Soc. 1997, 119, 6153–6167.
Supporting Information (see also the footnote on the first page of
this article): Experimental details and characterization data for car-
baldehydes, representative cycloadducts, nitrones, oximes, amides,
and esters, as well as details on calculations, including full Gaussian
citation, coordinates and energies.
CCDC-701193 (for Table 1, Entry 11, product B), -701194 (for
Table 1, Entry 4, product B), -701195 (for Table 1, Entry 4, product
A), -701196 (for compound 37 in Supporting Information), and
-708168 (for Table 1, Entry 15, product A) contain the supplemen-
tary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[5]
[6]
a) For our work on related cycloadditions using azomethine
ylides, see: L. Meng, J. C. Fettinger, M. J. Kurth, Org. Lett.
2007, 9, 5055–5058; b) A 36 compound collection of the hetero-
cycles reported in this manuscript has been added to the NIH
Molecular Libraries Small Molecule Repository (MLSMR) for
high-throughput biological screening.
[7]
[8]
Experimental and computational details can be found in the
Supporting Information.
This is true even though the computed energies are for gas
phase calculations. Tests on selected systems indicated that sol-
vation (modeled using a continuum approach) has only small
effects on the relative energies of the transition-state structures.
See Supporting Information for details.
M. Schelper, A. de Meijere, Eur. J. Org. Chem. 2005, 582–592.
T. Konoike, T. Hayashi, Y. Araki, Tetrahedron: Asymmetry
1994, 5, 1559–1566.
C. Jimeno, M. Pasto, A. Riera, M. A. Pericas, J. Org. Chem.
2003, 68, 3130–3138.
Acknowledgments
We gratefully acknowledge the University of California, Davis and
the National Science Foundation (NSF) (CAREER program, com-
puter time from the Pittsburgh Supercomputer Center, and CHE-
0614756), and the National Institutes of Health (NIH)
(GM076151) for support of this work. NMR spectrometers used
were partially funded by the NSF (CHE-0443516 and CHE-
9808183).
[9]
[10]
[11]
[12]
[1] For leading references on intramolecular dipolar cycload-
I. S. Kim, G. R. Dong, Y. H. Jung, J. Org. Chem. 2007, 72,
5424–5426.
ditions, see: a) I. Coldham, R. Hufton, Chem. Rev. 2005, 105,
Eur. J. Org. Chem. 2009, 1578–1584
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1583

L. Meng, S. C. Wang, J. C. Fettinger, M. J. Kurth, D. J. Tantillo
FULL PAPER
[13] J. H. Rigby, S. V. Cuisiat, J. Org. Chem. 1993, 58, 6286–6291.
[14] W. G. Dauben, J. M. Cogen, G. A. Ganzer, V. Behar, J. Am.
Chem. Soc. 1991, 113, 5817–5824.
[26] Note that reaction of the nitrone with 1-pentyne did not lead
to any cycloadduct, consistent with the results described earlier
for systems with tethered alkynes.
[27] This is an example of “torsional steering”. For leading refer-
ences, see: R. G. Iafe, K. N. Houk, Org. Lett. 2006, 8, 3469–
3472.
[28] The pure oxime was subsequently heated in toluene under a
nitrogen atmosphere in a sealed tube at 180 °C for 8 h,[29,30]
but no desired cycloadduct was obtained.
[15] S. Yu, C. Rabalakos, W. D. Mitchell, W. D. Wulff, Org. Lett.
2005, 7, 367–369.
[16] D. D. Kim, S. J. Lee, P. Beak, J. Org. Chem. 2005, 70, 5376–
5386.
[17] A. Chatterjee, D. K. Maiti, P. K. Bhattacharya, Org. Lett.
2003, 5, 3967–3969.
[18] I. Antonini, G. Cristalli, P. Franchetti, M. Grifantini, U. Gul-
ini, S. Martelli, J. Heterocycl. Chem. 1978, 15, 1201–1203.
[19] A. Basak, S. C. Ghosh, Tetrahedron Lett. 2005, 46, 7385–7388.
[20] M. J. Haddadin, J. P. Freeman, Chem. Heterocycl. Compd.
1985, 42, 283–350.
[21] M. Lamchen, in Mechanisms of Molecular Migrations (Ed.:
B. S. Thyagarajan), Interscience Publishers, New York, 1968,
vol. 1, pp. 1–60.
[22] Note also that no correlation was observed between the com-
puted preference of a given nitrone for the E or Z geometry
and the experimentally observed preference for a particular cy-
cloaddition product. See Supporting Information for details on
computed nitrone structures. Note also that for cases where
nitrone products were observed (i.e., cycloaddition did not oc-
cur), mixtures of geometric isomers were observed when R4 =
Me but only one geometric isomer (of unknown stereochemis-
try) was observed when R4 = cHex.
[29] A. Hassner, R. Maurya, O. Friedman, H. E. Gottlieb, A.
Padwa, D. Austin, J. Org. Chem. 1993, 58, 4539–4546.
[30] U. Chiacchio, G. Buemi, F. Casuscelli, A. Procopio, A. Rescif-
ina, R. Romeo, Tetrahedron 1994, 50, 5503–5514.
[31] For leading references on microwave heating, see: a) C. O.
Kappe, Chem. Soc. Rev. 2008, 37, 1127–1139; b) C. O. Kappe,
D. Dallinger, Nat. Rev. Drug Discovery 2006, 5, 51–63; c) C. O.
Kappe, Angew. Chem. Int. Ed. 2004, 43, 6250–6284.
[32] Gaussian 03, revision B.04, Gaussian, Inc., Pittsburgh, PA,
2004 (see Supporting Information for the full citation including
all author names).
[33] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) A. D.
Becke, J. Chem. Phys. 1993, 98, 1372–1377; c) C. Lee, W. Yang,
R. G. Parr, Phys. Rev. B: Solid State 1988, 37, 785–789; d) P. J.
Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys.
Chem. 1994, 98, 11623–11627; e) The reliability of the B3LYP/
6-31G(d) method for modeling dipolar cycloaddition reactions
has been assessed previously. See, for example: D. H. Ess, K. N.
Houk, J. Phys. Chem. A 2005, 109, 9542–9553; P. Merino, J.
Revuelta, T. Tejero, U. Chiacchio, A. Rescifina, G. Romeo, Tet-
rahedron 2003, 59, 3581–3592.
[34] a) C. Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523–
5527; b) K. Fukui, Acc. Chem. Res. 1981, 14, 363–368.
[35] A. P. Scott, L. Radom, J. Phys. Chem. 1996, 100, 16502–16513.
[36] a) V. Barone, M. J. Cossi, J. Phys. Chem. A 1998, 102, 1995–
2001; b) B. Barone, M. Cossi, J. Tomasi, J. Comput. Chem.
1998, 19, 404–417; c) Y. Takano, K. N. Houk, J. Chem. Theory
Comput. 2005, 1, 70–77.
[23] For a review, see: J. I. Seeman, Chem. Rev. 1983, 83, 83–134.
[24] a) Solvation single point calculations [CPCM(UAKS)-B3LYP/
6-31G(d) using water] were also carried out, and these barriers
do not change significantly. See Supporting Information for
details; b) We are aware that the presence of the benzimidazole
ring could significantly lower these barriers, but we have as yet
been unable to locate analogous transition-state structures
using larger models that include benzimidazole rings.
1
[25] a) Solvent induced shifts in the H NMR spectra[25c] of these
nitrones were used to make these Z/E assignments (data pre-
sented in the Supporting Information); b) Since Z/E assign-
ment for this compound based on solvent induced shifts in the
1H NMR spectra[25c] was ambiguous, this Z assignment is
based on steric considerations[25c] (data presented in the Sup-
porting Information); c) H. G. Aurich, M. Franzke, H. P. Kes-
selheim, Tetrahedron 1992, 48, 663–668.
[37] N. Müller, A. Falk, Ball & Stick V.3.7.6, molecular graphics
application for MacOS computers, Johannes Kepler University,
Linz, 2000.
Received: December 4, 2008
Published Online: March 4, 2009
1584
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1578–1584