Boron-based dipolar multicomponent reactions: Simple generation of substituted aziridines, oxazolidines and pyrrolidines

Article abstract of DOI:10.1002/chem.201000349

New multicomponent reactions of aldehydes, isocyanides, trialkylboron reagents and dipolarophiles have been developed as an efficient route to diverse scaffolds, including aziridines, oxazolidines and poly-substituted pyrrolidines. This chemistry, inspired by a report by Hesse in 1965, is simple and involves mild conditions. Computational studies provide a basis to investigate the stereochemical features observed in the formation of oxazolidines and fourcomponent adducts, and permit identification of potential factors that might influence the outcome of the multicomponent reaction. Thus, a rational screening of all the components and reaction parameters is made to examine the manifold mechanistic pathways and establish the practical limits for standard applications. Finally, intramolecular and solid-supported versions of these reactions bring new synthetic possibilities and practical protocols. Overall, the results describe a new family of multicomponent reactions valuable not only for organic reactivity, but also for combinatorial chemistry and drug discovery. 2010 Wiley-VCH Verlag GmbH & Cu. KGaA. Weinheim.

Full text of DOI:10.1002/chem.201000349

DOI: 10.1002/chem.201000349
Boron-Based Dipolar Multicomponent Reactions: Simple Generation of
Substituted Aziridines, Oxazolidines and Pyrrolidines
Nicola Kielland,[a] Federica Catti,[a] Davide Bello,[a] Nicolas Isambert,[a]
Ignacio Soteras,[b] F. Javier Luque,[b] and Rodolfo Lavilla*[a, c]
In memory of Gerhard Hesse
Abstract: New multicomponent reac-
tions of aldehydes, isocyanides, trialkyl-
boron reagents and dipolarophiles have
been developed as an efficient route to
diverse scaffolds, including aziridines,
oxazolidines and poly-substituted pyr-
rolidines. This chemistry, inspired by a
report by Hesse in 1965, is simple and
involves mild conditions. Computation-
al studies provide a basis to investigate
the stereochemical features observed in
the formation of oxazolidines and four-
component adducts, and permit iden-
tification of potential factors that might
influence the outcome of the multicom-
establish the practical limits for stan-
dard applications. Finally, intramolecu-
lar and solid-supported versions of
these reactions bring new synthetic
possibilities and practical protocols.
Overall, the results describe a new
family of multicomponent reactions
valuable not only for organic reactivity,
but also for combinatorial chemistry
and drug discovery.
ponent reaction. Thus,
a
rational
screening of all the components and re-
action parameters is made to examine
the manifold mechanistic pathways and
Keywords: boranes · cycloaddition ·
heterocycles · isocyanides · multi-
component reactions
Introduction
Progress in this area has allowed the use of this methodolo-
gy in target-oriented synthesis, parallelisation protocols and
diversity-oriented synthesis, thus enabling the preparation of
natural products and bioactive compounds.
Multicomponent reactions (MCRs) are domino processes in
which three or more reactants interact to form an adduct in
a single operation. MCRs constitute a fast-growing field of
enormous importance in organic chemistry[1] as they
embody many features of the ideal synthesis, such as atom
and step economy, convergence and structural diversity.[2]
Isocyanides are arguably the most prolific functional
group for MCRs.[3] In recent times new and exciting results
have brought great attention to this field.[4] In pursuing the
development of new MCRs,[5] we considered the interaction
of electrophilic boron reagents with isocyanides as the initial
step to generate reactive intermediates amenable to further
reactivity.[6] A literature search revealed a fascinating pro-
cess described by Hesse and co-workers in 1965, which in-
volves the generation of oxazolidines in useful yields from
the interaction of alkylboranes, aldehydes and isocyanides.[7]
The authors pointed out the intermediacy of an azomethine
ylide, which may undergo a dipolar [3+2] cycloaddition with
a second equivalent of the aldehyde to furnish the final
adduct in a stereoselective manner (Scheme 1). Surprisingly,
almost no references to this work have been reported.[8] The
process, the keystone of which is the reaction between the
isocyanide and the borane,[9] can be reshaped and, through
different trappings of the dipolar intermediate, can provide
a variety of scaffolds, thus exploiting the rich chemistry of
azomethine ylides.[10,11]
[a] N. Kielland, Dr. F. Catti, Dr. D. Bello, Dr. N. Isambert,
Prof. Dr. R. Lavilla
Barcelona Science Park
Baldiri Reixac 10–12, 08028 Barcelona (Spain)
Fax : (+34)934037108
E-mail: rlavilla@pcb.ub.es
[b] Dr. I. Soteras, Prof. Dr. F. J. Luque
Department of Physical Chemistry and
Institute of Biomedicine (IBUB)
Faculty of Pharmacy, University of Barcelona
Av. Diagonal 643, 08028 Barcelona (Spain)
[c] Prof. Dr. R. Lavilla
Laboratory of Organic Chemistry
Faculty of Pharmacy, University of Barcelona
Av. Joan XXIII sn, 08028 Barcelona (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201000349.
7904
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 7904 – 7915

FULL PAPER
to borane 3, which after alkyl migration would yield an imi-
noborane B, which could dimerise and then either undergo
a second ethyl migration[9] or, in the presence of aldehyde 1,
could form an oxazaborolidine intermediate (C) en route to
the azomethine ylide D. Finally, this dipole could efficiently
trap a second equivalent of the aldehyde to yield the oxazo-
lidine 5. Nevertheless, the remarkable bond-forming effi-
ciency[1c] of the MCR could be improved if it allowed incor-
poration of the dipolarophile species 4, which would lead to
the highly substituted pyrrolidine derivatives 6, representing
Scheme 1. Generation of five-membered N-heterocycles from the interac-
tion of isocyanides, aldehydes, alkylboranes and dipolarophiles through
azomethine ylides.
À
À
the formation of one C N and four C C bonds in a single
event.[13]
The main feature of the process deals with a practically
unexplored access route to azomethine ylides, which may
tackle some synthetic niches difficult to reach through
known methodologies. However, the drawbacks of the new
reaction are related to the double incorporation of the alkyl
groups from the borane, and the competition between the
aldehyde and the dipolarophile for trapping the azomethine
ylide. The oxazolidine adduct contains two different moiet-
ies from the starting aldehyde, so its formation can be re-
garded as a chemo-differentiating reaction, a burgeoning
family of processes with great synthetic potential.[14] More-
over, the new MCRs are susceptible to many side reactions,
such as the aforementioned (and undesired) formation of
the isocyanide–borane complexes,[9] the interaction of the
isocyanide with the dipolarophile,[3c,15] and Passerini-type re-
actions promoted by the Lewis acidity of the borane.[16]
As proof of concept, we reacted an equimolecular mixture
of p-chlorobenzaldehyde (1a), p-chlorophenyl isocyanide
(2a), triethylborane (3a) and N-phenylmaleimide (4a) in
THF at room temperature, to obtain the four-component-re-
action (4CR) adduct 6a as a 4:1 mixture of two stereoiso-
mers (the minor isomer being the all-cis epimer) in an over-
all yield of 55% together with lesser amounts (ꢀ10%) of
the corresponding trans-oxazolidine 5a (Scheme 3). The rel-
ative stereochemistry of compounds 5a and 6a was estab-
lished by NMR methods (the NOEs being of diagnostic
value). Incidentally, the structural assignment for the oxazo-
lidine derivative agrees with that reported by Hesse, which
he obtained by using chemical correlation studies.[7]
Herein, we report a new set of MCRs involving alde-
hydes, isocyanides, trialkylboron reagents and dipolarophiles
that is well suited to affording a variety of scaffolds compris-
ing aziridines, oxazolidines and poly-substituted pyrrolidines.
We used computational studies to unravel the mechanistic
details of the stereochemical preferences of the processes.
Lastly, we established the scope and limitations of the new
MCRs, and also devised intramolecular and solid-supported
versions of these reactions.
Results and Discussion
Mechanistic proposal: The suitability of the MCR involving
the interaction of four reactants (aldehydes, isocyanides, tri-
alkylboranes and dipolarophiles) to become a productive
process depends on three critical factors: firstly, the feasibili-
ty of the domino process that links the four chemical inputs
to yield the desired cycloaddition;[12] secondly, the generality
of the process, which is mainly determined by the range of
reactivity for each component; and lastly, the reaction pa-
rameters required for achieving useful transformations
under mild conditions. These issues are particularly chal-
lenging due to the number and nature of the reactive species
presumably engaged in the mechanism, and the manifold
evolutionary pathways they may follow. Thus, as illustrated
in Scheme 2, the process could start with the formation of
an initial adduct A arising from the addition of isocyanide 2
Scheme 3. 4CR leading to pyrrolidine derivative 6a. [a] 4:1 isomeric
ratio, whereby 6a is the major isomer and the minor isomer is the all-cis
compound.
Scheme 2. Mechanistic proposal.
Chem. Eur. J. 2010, 16, 7904 – 7915
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7905

R. Lavilla et al.
Computational analysis: Quantum chemistry computations
were performed to gain insight into the stereochemical pref-
erences of the chemical processes involved in these MCRs.
To this end, a series of model compounds that includes
methylborane, phenyl isocyanide, benzaldehyde and N-phe-
nylmaleimide was considered in calculations, and the rela-
tive free energies between chemical reactants, products and
transition states (TSs) were determined at the MP2/6-311G-
ACHTUNGTRENNUNG(d,p)+[CCSD-MP2/6-31G(d)] level (see the Experimental
and Computational Section).
Figure 2. Representation of the cis and trans forms of the azomethine
ylide D.
Formation of azomethine ylide D: According to the reaction
mechanism (Scheme 2), the iminoborane intermediate B,
generated after alkyl migration in the initial adduct A, in
the presence of aldehyde 1 yields the azomethine ylide D
via the oxazaborolidine intermediate C. For the model com-
pounds considered in the present computations, formation
of oxazaborolidine C is highly favourable (À17.9 kcalmolÀ1),
but the exothermicity of this process is counterbalanced by
the energy required to open the oxazaborolidine ring to
yield the azomethine ylide D. As a result, the ylide D is
slightly stabilised relative to the separated reactants (inter-
mediate B and benzaldehyde; see Figure 1). The orientation
of the two benzene rings in D defines two configurations de-
noted here as D-cis and D-trans, which mimic the W- and S-
type configurations of ylides (Figure 2). The D-trans species
is marginally favoured (0.4 kcalmolÀ1) compared to the D-
cis form. Accordingly, both species can a priori participate
in the reaction with either another molecule of aldehyde to
yield oxazolidine 5 or with dipolarophile 4 to generate
adduct 6.
leading to trans-5 is destabilised by 7.0 kcalmolÀ1 relative to
the separate reactants, whereas the TS yielding cis-5 is fur-
ther destabilised by 2.6 kcalmolÀ1 (Figure 1). Such a differ-
ence can be attributed to two factors. First, the more
eclipsed approach of the reactants in the TS cis-5, as noted
À
in the Newman projection along the forming C C bond (see
Figure 3). Second, the larger steric repulsion due to the ap-
proach of the benzene rings in the TS cis-5, as noted in the
short contact between the hydrogen atoms of the benzene
rings (2.31 ꢁ; see Figure 3). On the other hand, when the cy-
cloaddition takes place starting from the ylide D-trans, there
is an additional destabilisation (by around 2 kcalmolÀ1) in
the TSs leading to oxazolidine 5, which reflects the in-
creased eclipsed character of the incoming reactants (note
also the short H···H contact, 2.27 ꢁ, found in TS trans-5; see
Figure 3). Overall, the results indicate that the stereoselec-
tive formation of trans-5 (Scheme 3) can be attributed to the
kinetic preference associated with the trans cycloaddition
pathway originating from the reaction of benzaldehyde with
the D-cis form of the azomethine ylide.
Formation of oxazolidine 5: The azomethine ylide D can
react with benzaldehyde to yield oxazolidine 5, which can
exist in two stereoisomeric arrangements (trans and cis, see
Scheme 3). Formation of 5 is predicted to be very favoura-
ble (around À42 kcalmolÀ1; see Figure 1), trans-5 being fav-
oured by 2.4 kcalmolÀ1 relative to cis-5. When the [3+2] cy-
cloaddition involves the azomethine ylide D-cis, the TS
Inspection of the geometries of the most favoured TSs re-
veals the asynchronicity of the bond formation in the cyclo-
addition reaction leading to trans-5 (i.e. the lengths of the
À
À
forming C C and C O bonds are 2.15 and 2.42 ꢁ, respec-
À
tively), which is in contrast with the similar lengths (C C:
Figure 1. Diagram of the free energy differences [kcalmolÀ1] for the chemical species involved in the formation of oxazolidine 5 and pyrrolidine 6.
7906
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 7904 – 7915

Boron-Based Dipolar Multicomponent Reactions
FULL PAPER
ble than the cis-6 adduct (Figure 1). Remarkably, the TS
leading to trans-6 is favoured by 4.7 kcalmolÀ1 compared to
that associated with the formation of cis-6. In the two TSs,
there is a steric contact between the N-phenyl ring in the
ylide and the N-phenylmaleimide moiety (distances of 2.31
and 2.45 ꢁ in TS trans-6 and TS cis-6, respectively; see
Figure 4). However, the more eclipsed approach of the reac-
tants in TS cis-6 explains the larger stability of the TS trans-
6.
Figure 4. Representation of the TSs leading to trans and cis forms of pyr-
rolidine 6 (distances are in ꢁ).
Overall, the results point out that the stereochemical pref-
erences in the formation of oxazolidine 5 and pyrrolidine 6
are dictated by the kinetically more favoured TS, which cor-
responds to the one leading to trans-5 and trans-6, respec-
tively. Notably, the latter is predicted to be 1.5 kcalmolÀ1
more favourable than the former, which agrees qualitatively
Figure 3. Representation of the TSs leading to trans and cis forms of oxa-
zolidine 5 formed from (top) D-cis and (bottom) D-trans forms of the
azomethine ylide. Newman projections viewed from along the forming
À
C(benzaldehyde) CACHTUNGTRENNUNG(ylide) bond are also shown (distances and torsional
angles are in ꢁ and degrees, respectively).
À
2.23 ꢁ; C O: 2.25 ꢁ) found in the TS leading to cis-5.
with the larger amount found for pyrrolidine 6 (see
These trends, which differ from the negligible asynchronicity
found for the corresponding [3+2] cycloaddition between
benzaldehyde and nitrile-containing carbon ylides reported
in previous studies,[17] most likely reflect the influence of the
substituents attached to the ylide D and the steric effects
arising from the aromatic ring bound to the nitrogen atom
of the ylide.
Scheme 3). Nevertheless, the preceding discussion also sug-
gests that the precise outcome of the MCR may be very sen-
sitive to the nature of the substituents attached to both iso-
cyanide and aldehyde, the differential stabilisation played
by the solvent on the [3+2] cycloaddition, and the chemical
features of the dipolarophile, which might affect the relative
stability of the cis and trans species of the ylide D and the
competitive reaction of the ylide with either the aldehyde or
the dipolarophile.
Formation of pyrrolidine 6: The azomethine ylide D may in-
teract with N-phenylmaleimide to afford the desired pyrroli-
dine 6, which can display two possible stereoisomeric ar-
rangements (trans, cis) for the C-phenyl ring coming from
intermediate D with respect to the maleimide ring, the
trans-6 adduct being the major component (ratio 4:1) of the
mixture (see Scheme 3). The interaction between the ylide
and the dipolarophile is conditioned by the presence of the
bulky benzene rings, which thus only enables the approach
of the reactants through a reactive pathway where the ben-
zene rings adopt an anti orientation. Accordingly, the reac-
tion of N-phenylmaleimide with the D-cis ylide exclusively
affords trans-6, whereas reaction with D-trans yields cis-6.
Computations point out that the formation of pyrrolidine 6
is highly exothermic (around 55 kcalmolÀ1), and the pyrroli-
dine trans-6 is predicted to be 0.7 kcalmolÀ1 more favoura-
Reaction scope: Upon surveying the reaction conditions, we
found that the best solvents were diethyl ether (Et2O) and
THF; no reaction occurred in toluene, dichloromethane,
DMF, CH3CN or MeOH. Surprisingly, water (50% in THF)
is well tolerated and it is even possible to run the MCR in
micellar systems (H2O/sodium dodecyl sulfate). The reaction
proceeds within a useful temperature range: from À20 to
+208C. At lower temperatures, there is no noticeable reac-
tivity, whereas at higher values, the starting materials severe-
ly decompose (the formation of isocyanide–borane adducts
is presumably favoured). Therefore, all the following experi-
ments were run at room temperature, after pre-mixing the
four components at 08C. Microwave irradiation seems to in-
crease the formation of oxazolidines, consequently lowering
Chem. Eur. J. 2010, 16, 7904 – 7915
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7907

R. Lavilla et al.
the amount of the 4CR adduct. The use of a set of additives
including Lewis acids, bases, ligands and dehydrating agents
under the usual conditions or gave rise to complex mixtures
(Table 1, entries 9–12). As revealed in Table 1, the best
yields were obtained with electron-rich benzene rings, which
may reflect the need for nucleophilicity at both the C-termi-
nal atom of the isocyanide and at the N atom of intermedi-
ate B (Scheme 2).[19] The stability of the resulting azome-
thine ylide may also be important for the viability of the
MCRs, as it would presumably restrict the process to the
presence of aromatic rings in the dipole.
such as scandium
ACHTUNGTRENNUNG
triflate, Sc(OTf)3), MgACHTGNUTRENNUNG
G
(DIPEA), 2-chloropyridine[18] or molecular sieves (4 ꢁ), re-
spectively, precludes the reaction. In contrast, the chemistry
proceeds normally with galvinoxyl and with limited expo-
sure to oxygenated atmospheres (no effect on the yields and
selectivity), which seems to rule out a radical process.
Isocyanide range: We determined the range of reactivity for
each component by using comparable sets of reactants and
reaction conditions. Starting with isocyanides 2, we found
that only the aromatic derivatives were productive under
the mild conditions tested. For example, p-methoxyphenyl
isocyanide (2b) was efficiently incorporated into oxazolidine
5b and the 4CR adduct 6b (Table 1, entries 1 and 2, respec-
tively). As usual, we obtained epimeric mixtures (ꢀ7:3) of
the 4CR adducts, the major stereoisomers being depicted in
Table 1. p-Tolyl (2c) and phenyl isocyanides (2d) afforded
the expected adducts in somewhat lower yields (Table 1, en-
tries 3 and 4). Halo-substituted aromatic derivatives also re-
acted in a convenient manner (Table 1, entries 5–7), whereas
the deactivated 4-ethoxycarbonyl derivative (2 f, Table 1,
entry 8) was considerably less reactive and the 3- and 4CR
adducts could only be detected in minute amounts. Cyclo-
hexyl (2g), cyclohexenyl (2h), benzyl (2i) and deactivated
isocyanides, such as tosylmethyl isocyanide (2j), were inert
Aldehyde range: Regarding the aldehydes (Table 2), we
found that aromatic derivatives afforded the best conver-
sions in both three-component reactions (3CRs; oxazoli-
dines 5) and 4CRs (pyrrolidines 6). p-Chlorobenzaldehyde
(1a, Table 2, entries 1 and 2) yielded the corresponding ad-
ducts in good amounts, and methyl 4-formylbenzoate (1b,
Table 2, entries 3 and 4) was also productive. Benzaldehyde
(1c, Table 2, entry 5) and even the electron-rich p-methoxy-
benzaldehyde (1d, Table 2, entry 6) afforded the corre-
sponding MCR adducts, although in lower yield in the latter
case. Among the heteroaromatic aldehydes, furfural (1e,
Table 2, entry 7) yielded the corresponding adduct 6g,
whereas pyridine-3-carbaldehyde (1 f, Table 2, entry 8) gave
predominantly the oxazolidine 5h in low yield; the 4CR
adduct was only detected under these conditions (general
Table 2. Range of aldehydes 1.
Table 1. Range of isocyanides 2.
Entry[a]
Procedure
R
Adduct
Yield[b]
1
2
3
4
A
B
A
B
B
B
B
A
B
B
B
4-Cl-C6H4- (1a)
4-Cl-C6H4- (1a)
4-MeO2C-C6H4- (1b)
4-MeO2C-C6H4- (1b)
C6H5- (6e)
4-MeO-C6H4- (1d)
2-furyl- (1e)
3-pyridyl- (1 f)
5b
6b[c]
5g
6d
6e
6 f/7a
6g
5h
6h
6i[g]
6j
85
68
50
24
55
25/4
20
62[f]
40
71
63
–
Entry[a]
Procedure
R
Adduct
Yield[b]
1
2
3
A
B
A
A
A
B
A
B
A
A/B
A
A
4-MeO-C6H4- (2b)
4-MeO-C6H4- (2b)
4-Me-C6H5- (2c)
C6H5- (2d)
4-Cl-C6H4- (2a)
4-Cl-C6H4- (2a)
4-F-C6H4- (2e)
4-EtO2C-C6H4- (2 f)
cyclohexyl- (2g)
1-cyclohexenyl- (2h)
benzyl- (2i)
5b
6b[c]
5c
5d
5a
6a
5e
5 f/6c
–
85
68
67
22
65
55
36
traces
–
5
6[d,e]
7[d]
8[d]
9
4[d]
5
CH2=CH- (1g)
Ph-CH=CH- (1h)
6[e]
7
10[d]
11
12
13
ꢁ
C5H11-C C- (1i)
8[f]
9
A/B
A
(Me)3C- (1j)
EtO2C- (1k)
–
5i
7
10
11
12
–
–
–
–
–
–
[a] The reactions were performed following the standard procedure A or
B to form respectively 5 or 6; p-methoxyisocyanide (2b) was used as the
aromatic isocyanide input if not stated otherwise. [b] Overall yield of iso-
lated product. [c] Adducts 6 were generated as ꢀ7:3 diastereomeric mix-
tures, from which the major epimer was isolated and characterised. The
major epimer correspond to the depicted relative stereochemistry. [d] p-
Chlorophenyl isocyanide (2a) was used as the aromatic isocyanide input.
[e] Aziridine 7a was also isolated in this experiment. [f] Following proce-
dure B, the expected 4CR adduct was detected in minute amounts,
whereas the oxazolidine 5h adduct was obtained in 15% yield. [g] Ob-
tained as a 7:3 mixture of stereoisomers, from which the minor adduct
was isolated and characterised.
tosylmethyl- (2j)
[a] The reactions were performed following the standard procedure A or
B to form respectively 5 or 6 with p-chlorobenzaldehyde 1a as the aro-
matic aldehyde input. [b] Overall yield of isolated product. [c] Adducts 6
were generated as ꢀ7:3 diastereomeric mixtures, from which the major
epimer was isolated and characterised. The major epimer corresponds to
the depicted relative stereochemistry. [d] The reaction was performed
with BBu3 (3b). [e] In this experiment, oxazolidine 5a was also generated
(10%). [f] This experiment was attempted with methyl-4-formylbenzoate
1b as the aldehyde component.
7908
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 7904 – 7915

Boron-Based Dipolar Multicomponent Reactions
FULL PAPER
Table 3. Set of boron reagents.
procedure B). However, this oxazolidine was conveniently
generated (62%) when the 3CR protocol (general proce-
ACHTUNGTRENNUNGdure A) was used, which probably means that this electro-
philic aldehyde overrides maleimide as the dipolarophile
input, thereby preventing the 4CR process. Interestingly,
formyl groups linked to non-aromatic sp2- or sp-hybridised
carbon atoms were also reactive. Thus, acrolein (1g), cinna-
maldehyde (1h) and oct-2-ynal (1i) readily afforded the ex-
pected MCR adducts (Table 2, entries 9–11).[20] In contrast,
pivalaldehyde (1j) was not reactive (Table 2, entry 12) and
ethyl glyoxylate (1k) gave low yields of the corresponding
oxazolidine (5i, Table 2, entry 13). These results clearly indi-
cate that electron-withdrawing substituents in aromatic rings
seem to favour the MCRs, by rendering the carbonyl group
more electrophilic and stabilising the anionic site in the
dipole. Interestingly, strong electron-donor substituents,
such as the p-MeO group, also provided the 4CR adduct
albeit in lower yields. In this experiment the aziridine 7a
(Table 2, entry 6) was also isolated whereas the correspond-
ing oxazolidine was not detected, probably due to the low
reactivity of this aldehyde as dipolarophile in [3+2] cycload-
ditions. Confirming this result, the 2-allyloxy derivative 1l
and 2,4-dimethoxybenzaldehyde (1m) afforded aziridines
7b and 7c, respectively (Scheme 4), most likely by electro-
cyclisation of the azomethine ylide intermediate.[21,22] Over-
all, the reactivity pattern of aldehydes seems to parallel that
of the carbonyl precursors of the azomethine ylides generat-
ed by conventional methods (e.g. condensation, decarboxy-
lation, prototropy, carbene insertion); the aromatic deriva-
tives are the most widely used for preparative purposes.[10]
Entry[a]
Procedure
R
Adduct
Yield[b]
1
2
A
A
B
B
B
B
Et (3a)
Bu (3b)
H (3c)
C6H5- (3d)
9-benzyl-BBN (3e)
F (3 f)
5a
5j
6k[d]
–
–
–
65
40
3
–
–
3[c]
4
5
6
–
[a] The reactions were performed following the standard procedures A
or B to form respectively adducts 5 or 6 with p-chlorobenzaldehyde as
the aromatic aldehyde input and p-chlorophenyl isocyanide as the isocya-
nide input if not stated otherwise. [b] Overall yield of isolated product.
[c] p-Methoxyphenyl isocyanide was used in this experiment. [d] Adduct
6k was generated as a ꢀ4:1 diastereomeric mixture.
BBN: 9-borabicycloACTHNUGRTNEUNG[3.3.1]nonane) and BF3·Et2O (3 f) led to
complex mixtures (Table 3, entries 4–6). Whether these
MCRs can be improved with more suitable reaction condi-
tions, namely by minimising unproductive side reactions, re-
mains to be determined.
Dipolarophile range: Regarding the dipolarophile
4
(Table 4), we obtained the highest yields of the 4CR adducts
using N-phenylmaleimide (4a, Table 4, entry 1), which is the
reference compound in many dipolar cycloadditions. It
should be noted that azomethine ylides preferentially react
with electron-deficient dipolarophiles.[25] As such, dimethyl
acetylenedicarboxylate (4b, Table 4, entry 2), dimethyl fu-
marate (4c, Table 4, entry 3) and fumaronitrile (4d, Table 4,
entry 4) were found to be reactive in our system. The 4CR
adducts were isolated (although the adduct from acetylene
dicarboxylate was somewhat unstable), and in all applicable
cases the stereochemistry of the dipolarophile was con-
served (7:3 mixtures of isomers in Table 4, entries 1, 3 and
4). Methyl acrylate (4 f, Table 4, entry 5) afforded the 4CR
adduct 6o (24%, Figure 5) together with oxazolidine 5b as
the major compound (76%). The structure of 6o was unam-
biguously assigned by diagnostic correlations from NMR ex-
periments (1H, 13C, COSY, HSQC, HMBC, NOESY). In this
case, the non-symmetrically substituted dipolarophile be-
haves in a particular manner, as it not only affords a cis ste-
reochemistry, but seemingly alters the regiochemistry
(taking into account the expected polarity in this type of
process). Literature precedents on cycloadditions involving
acrylates and azomethine ylide intermediates generally
report low selectivity profiles, and in most cases the predom-
inant regiochemistry is opposite to that observed in our
system.[26] Diethyl azodicarboxylate (4e, Table 4, entry 6)
also reacted conveniently but again the slight instability of
the adduct prevented its full characterisation. The assay
Scheme 4. Aziridine formation from o-alkoxybenzaldehydes.
Boron reagent range: The borane component is the main
limitation in this MCR: we only found acceptable reactivity
by using the commercially available ethyl (3a) and butyl
(3b) boranes (Table 3, entries 1 and 2). However, borane
itself afforded minute amounts of the expected adduct (as a
somewhat unstable compound), together with the benzylic
alcohol from the reduction of the aldehyde (Table 3,
entry 3).[23] Triphenyl boron (3d),[24] 9-benzyl-BBN (3e; 9-
Chem. Eur. J. 2010, 16, 7904 – 7915
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7909

R. Lavilla et al.
Table 4. Set of dipolarophile reagents 4.
(1b) were used as reaction partners. However, no produc-
tive reaction was detected under the mild conditions used,
and only the corresponding oxazolidines were formed. As
noted above, the complexity of the mechanism promotes
competition between two species (the aldehyde and the
nominal dipolarophile) for trapping the dipole, which limits
the use of the less reactive dipolarophiles. Moreover, the
need to perform the processes at room temperature (cyclo-
additions with azomethine ylides are typically conducted at
much higher temperatures) to avoid side reactions diminish-
es the feasibility of some combinations, particularly those in-
volving the less reactive species.
Entry
A=B
Adduct
Yield[a]
1
N-phenylmaleimide (4a)
6b[b]
6l[b]
6m[d]
6n[b]
6o/5a
6p
–
5b
5b
–
85
2
3
4
MeO2C-C C-CO2Me (4b)
dimethyl fumarate (4c)
fumaronitrile (4d)
49[c]
26
ꢁ
20[e]
24/76
34
5
methyl acrylate (4 f)
Intramolecular MCRs: We also explored the possibility of
promoting intramolecular MCRs by taking advantage of the
structural diversity provided by this methodology and the
extraordinary success recently achieved in this field.[10c,f,27]
Again, the mild conditions demanded use of electron-defi-
cient olefin moieties (for instance, the deactivated o-allyl-
6
7
8
9
EtO2C-N=N-CO2Et (4e)
1,4-naphthoquinone (4g)
ethyl 3-coumarincarboxylate (4h)
diethyl benzylidenemalonate (4i)
norbornene (4j)
traces
58
35
–
–
10[f]
11[f]
2,3-dihydrofuran (4k)
–
[a] Overall yield of isolated product. [b] Adducts 6 were generated as
ꢀ7:3 diastereomeric mixtures, from which the major epimer was isolated
and characterised. The major epimer corresponds to the depicted relative
stereochemistry. [c] Unreacted aldehyde was recovered (50%). [d] Ob-
tained as an inseparable 6:4 mixture of stereoisomers. [e] Unreacted al-
dehyde was recovered (78%). [f] In these entries, 4-ethoxycarbonylphen-
yl isocyanide (2 f) and methyl 4-formylbenzoate (1b) were used as the
isocyanide and aldehyde components, respectively.
ACHUTNGRENoNUG xyACHUTNGTRENbNGUN enzaldehyde (1l) failed to give any pyrrolidine adduct,
and afforded the aziridine 7b instead; see Scheme 4).
Hence, aldehyde 1n[28] was chosen as an alternative sub-
strate (Scheme 5). Rewardingly, the MCR yielded the chro-
menopyrrolidine adducts (6q and 6q’) in an overall yield of
45%, as a 2:1 mixture of stereoisomers displaying trans and
cis ring fusions, respectively.
Figure 5. 4CR adduct 6o. Diagnostic correlations in the NOESY spec-
trum.
Scheme 5. Intramolecular MCR.
Recent precedents to access these heterocyclic systems
used different methodologies to generate the azomethine
ylide.[29] However, the method presented here seems to be
the most straightforward for the introduction of two alkyl
substituents at position 2 of the tricyclic scaffold.[30]
with 1,4-naphthoquinone (4g, Table 4, entry 7) afforded a
complex mixture in which the expected adduct together
with its oxidation product were each detected in minute
amounts.
On the other hand, when more sterically hindered dipo-
larophiles, such as ethyl 3-coumarincarboxylate (4h, Table 4,
entry 8) and diethyl benzylidenemalonate (4i, Table 4,
entry 9) were used, the oxazolidine 5b was the only isolated
product. Two examples of inverse electron-demanding cyclo-
addition were attempted by using non-activated (electron-
rich) dipolarophiles, namely norbornene (4j, Table 4,
entry 10) and 2,3-dihydrofuran (4k, Table 4, entry 11).[25] In
both cases, to adjust the polarity of the process, 4-ethoxycar-
bonylphenyl isocyanide (2 f) and methyl 4-formylbenzoate
Solid-supported MCRs: Many side reactions do take place
in these MCR processes and the undesired by-products, to-
gether with remaining starting materials (usually aldehyde
and dipolarophile), can produce very complex mixtures, as
noted in the HPLC profile of the crude organic extract from
the experiment leading to 6 f (see the Supporting Informa-
tion). Therefore, the purification of the expected adducts is
extremely challenging, especially in cases with low yields.
Solid-phase versions of MCRs[31] might be a practical and ef-
7910
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 7904 – 7915

Boron-Based Dipolar Multicomponent Reactions
FULL PAPER
ficient way to facilitate removal of undesired products and
minimise the impact of side reactions. For instance, support-
ing the aldehyde component on a resin would secure its role
as the azomethine ylide precursor, and would prevent it
from acting as a dipolarophile. Remarkable work in the area
has shown the possibility of conducting dipolar cycloaddi-
tions in the solid phase.[32] Hence, we decided to link the al-
dehyde to a suitable resin, generate the azomethine ylide
and quench the reaction in situ with the dipolarophile, or
promote the selective formation of the aziridine and oxazo-
lidine systems. In this way, only the desired products would
remain attached to the resin, ideally to furnish pure samples
after the cleavage. Also, it is conceivable to generate the
linked dipolar intermediate, to elute all other compounds
and, in a sequential manner, promote the cycloaddition step
without interferences.
ficiently removes the unreacted aldehyde, and affords the
pure MCR adducts.
We then promoted the formation of “mixed oxazolidines”
in a 4CR using two different aldehydes. A solution-phase
approach would be unproductive, because at least four com-
pounds would be generated (two from homo-combinations
plus two from hetero-couplings). In an initial assay, we first
linked 4-hydroxybenzaldehyde to a 2-chlorotrityl chloride
resin, and then reacted it with borane 3a, p-bromobenzalde-
hyde (1q) and isocyanide 2b.
After the usual sequence of elutions, cleavage and a final
treatment with a solid-supported aldehyde scavenger, we ob-
tained a nearly equimolecular mixture of the two hetero-
coupled oxazolidines 5k and 5k’ which were not separated
(Scheme 7). This suggested that two azomethine ylides had
been generated: one from the aldehyde in solution, trapped
by the linked aldehyde, and vice versa. To solve this prob-
lem, we performed the reaction in a sequential manner, and
the protocol was modified to include washings after the gen-
eration of the linked dipole D, then we added the solution
of aldehyde 1q, and let the reaction proceed (Scheme 7). In
this way, we managed to selectively prepare the mixed oxa-
zolidine 5k (10%, unoptimised). The low yield may be at-
tributable to the manipulation of the azomethine ylide. Re-
markably, this methodology allows the selective generation
of solid-supported azomethine ylides and their interactions
with dipolarophiles in a sequential manner. The high purity
of the crude reaction mixture obtained in this experiment
compared to that obtained with previous solution-phase pro-
cesses is noteworthy (see the Supporting Information). In-
terestingly, this approach distinguishes between two pseudo-
equivalent partners in chemo-differentiating ABB’-type pro-
cesses, and consequently can significantly improve the syn-
thetic applications of these useful procedures.[14,34]
To explore these possibilities, we first linked 4-formylben-
zoic acid to hydroxymethyl polystyrene resin, and after-
wards we added the borane 3a and isocyanide 2b and
stirred the resin at room temperature (Scheme 6). After
A first attempt to perform a 4CR using the solid-support-
ed aldehyde 1o (Scheme 6), p-methoxyphenyl isocyanide
(2b), N-phenylmaleimide (4a) and triethylborane (3a) af-
forded a complex mixture, probably due to the degradation
of the desired adduct during the cleavage with sodium meth-
oxide (based on mass spectrometry evidence). We then at-
tempted the same reaction with the aldehyde-loaded resin
1p (Scheme 7) and, although we did detect the expected
4CR adduct, the overall yield was very low. We then decid-
ed to use a solid-supported isocyanide[35] to avoid the dimer-
isation process of the intermediate B (see Scheme 2). The 2-
chlorotrityl resin was loaded with N-(3-hydroxyphenyl)for-
mamide and the resulting polymer was subjected to dehy-
dration by using the Hulme protocol, to yield the solid-sup-
ported isocyanide 2k.[35a] The loading of the solid-supported
formamide was then evaluated with a TFA cleavage proto-
col. In a parallel experiment, the cleavage of the solid-sup-
ported isocyanide 2k afforded the same starting formamide.
Thus, we roughly estimated the loading of 2k to be about
0.4 mmolgÀ1, which suggests that the isocyanide or the for-
mamide is partially cleaved during the dehydration process.
A mixture of this isocyanide-functionalised resin and p-
chlorobenzaldehyde (1a) was added to a solution of triethyl-
Scheme 6. Solid-supported versions of the MCRs.
standard washings and a mild sodium methoxide cleavage,
the corresponding aziridine 7d was obtained in 50% yield.
Remarkably, this compound cannot be formed in solution,
as the aldehyde efficiently traps the dipole once generated,
which leads to the corresponding oxazolidine 5g. In another
experiment, we specifically promoted the latter reaction,
which took place smoothly upon interaction with the alde-
hyde component 1b. The resulting product was identical in
all aspects to that previously obtained in the solution phase.
The yields were calculated upon the loading of the aldehyde
onto the resin. In spite of using excesses of the non-support-
ed components, we could not accomplish quantitative trans-
formations, which probably reflects some practical limita-
tions in reaching sterically crowded regions in the solid sup-
port. However, the use of a solid-supported scavenger[33] ef-
Chem. Eur. J. 2010, 16, 7904 – 7915
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7911

R. Lavilla et al.
Scheme 8. MCRs using solid-supported isocyanide. [a] 4:1 isomeric ratio,
whereby 6r is the major isomer and the minor isomer is the all-cis com-
pound.
ane (3a), p-chlorobenzaldehyde (1a) and N-phenylmalei-
mide (4a) in a 4CR; standard TFA cleavage afforded the
expected adduct 6r (60%) in a stereoselective manner.[36]
Thus, we have conducted all our new MCRs in the solid
phase, thereby showing the feasibility of this strategy to fa-
cilitate access to pure compounds.
Conclusion
Based on an initial report by Hesse and co-workers, we
have designed and implemented a family of MCRs involving
the interaction of isocyanides, aldehydes, boranes and dipo-
larophiles to yield pyrrolidine, aziridine and oxazolidine ad-
ducts in a single step, with high atom economy and bond-
forming efficiency. These scaffolds have relevant presence in
bioactive compounds. Moreover, oxazolidines are also found
in pro-drugs as protected forms of 1,2-aminoalcohols.[37] In-
cidentally, this approach can also stereoselectively furnish
these compounds after the customary hydrolysis of the oxa-
zolidine precursors. Calculations performed for model com-
pounds indicate that the stereochemical preferences in the
reaction outcome are dictated by the TSs that are more fav-
oured kinetically, which correspond to those leading to trans
stereochemistries in the cycloaddition of the azomethine
ylide D-cis with either the aldehyde or dipolarophile. How-
ever, these studies also show that the precise evolution of
the MCRs may be subtly modulated by factors such as the
nature of the substituents attached to both the isocyanide
and the aldehyde units, or the chemical features of the dipo-
larophile. This has allowed a rationalisation of the manifold
mechanistic pathways and established the practical limits for
standard reactions. Furthermore, intramolecular and solid-
supported versions of these processes have been developed
Scheme 7. Solid-phase approaches to mixed oxazolidines. TFA: trifluoro-
acetic acid.
borane in THF and stirred for 48 h (Scheme 8). After cleav-
age with 1% TFA–dichloromethane, the pure oxazolidine 5l
was obtained in 91% yield. In another experiment, the
solid-supported isocyanide 2k was reacted with triethylbor-
7912
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 7904 – 7915

Boron-Based Dipolar Multicomponent Reactions
FULL PAPER
were dried (Na2SO4) and filtered. The solvent was removed under re-
duced pressure.
and, in the latter cases, the controlled environment achieved
using linked substrates allows selective reactions, and facili-
tates purification. Our future work will follow along these
lines, namely, to increase the reactivity range of some com-
ponents (particularly the boranes)[38] and to expand the syn-
thetic usefulness and scaffold variability for this complex
family of MCRs.
Cleavage B: 2-Chlorotrityl polystyrene resin: The resin-bound product
was treated with TFA (1%) in dichloromethane (5ꢂ3 mLꢂ1 min). The
mixture was dropped directly onto an extraction funnel with an aqueous
saturated NaHCO3 solution (10 mL). After extraction with dichlorome-
thane (3ꢂ10 mL), the combined organic extracts were dried (Na2SO4)
and filtered. The solvent was removed under reduced pressure.
Computational studies: Full geometry optimisations were performed with
the B3LYP[39] functional and the 6-31G(d)[40] basis set. The nature of the
stationary points was verified by inspection of vibrational frequencies
within the harmonic oscillator approximation. Intrinsic reaction coordi-
nate calculations were carried out to check the connection between TSs
and minimum-energy structures.[41] Single-point calculations were carried
Experimental and Computational Section
General procedure A: A solution of isocyanide (1 mmol, 1 equiv) and al-
dehyde (2 mmol, 1 equiv) in THF (2 mL) was added slowly to a solution
of trialkylborane in THF (1m, 1.2 mmol, 1.2 equiv) at 08C. After 3 min of
stirring at this temperature, the ice bath was removed and the reaction
mixture was stirred for 48 h at room temperature. An aqueous saturated
solution of Na2CO3 (10 mL) was added and the mixture was extracted
with dichloromethane (2ꢂ5 mL). The combined organic extracts were
dried (Na2SO4) and filtered. The solvent was removed under reduced
pressure and the residue was purified by column chromatography (SiO2,
hexanes/ethyl acetate) to yield the oxazolidine 5.
out at the MP2/6-311GACTHNGUTERNNUG
(d,p)[42] level to determine the relative energies.
Higher-order electron correlation effects were estimated from coupled-
cluster singles and doubles (CCSD) calculations with the 6-31G(d) basis
set. The final estimate of the electronic energies was determined by com-
bining the MP2/aug-cc-pVDZ energy with the difference between the en-
ergies calculated at both CCSD and MP2 levels with the 6-31G(d) basis
(denoted in the text as MP2/6-311GACTHNUTRGENUGN(d,p)+[(CCSD-MP2)/6-31G(d)]. Fi-
nally, zero-point, thermal and entropic corrections evaluated within the
framework of harmonic oscillator rigid-rotor models at 1 atm and 298 K
using the B3LYP/6-31G(d) vibrational frequencies (scaled by a factor of
0.96)[43] were added to estimate the free energy differences in the gas
phase. Calculations were performed with Gaussian 03.[44]
General procedure B: A solution of isocyanide (1 mmol, 1 equiv) and al-
dehyde (1 mmol, 1 equiv) in THF (2 mL) was added slowly to a solution
of trialkylborane in THF (1m, 1.2 mmol, 1.2 equiv) at À108C. Subse-
quently the dipolarophile was added to the mixture. After 3 min of stir-
ring at this temperature, the cooling bath was removed and the reaction
was stirred for 48 h at room temperature. An aqueous saturated solution
of Na2CO3 (10 mL) was added and the mixture was extracted with di-
chloromethane (2ꢂ5 mL). The combined organic extracts were dried
(Na2SO4) and filtered. The solvent was removed under reduced pressure
and the residue was purified by column chromatography (SiO2, hexanes/
ethyl acetate) to yield the adduct 6.
Acknowledgements
We thank the DGICYT (Spain, project CTQ 2009-07758, and SAF2008-
05595), Generalitat de Catalunya (SGR 2009-249), Almirall SA and
Grupo Ferrer (Barcelona) for financial support. N.K. thanks DGICYT
for a PhD fellowship.
General procedure C: For the synthesis of aziridines in the solution
phase, the stoichiometry of general procedure A was modified (aldehyde,
1 mmol, 1 equiv). After chromatographic purification, the corresponding
aziridine 7 was obtained.
[1] For overviews, see: a) Multicomponent Reactions (Eds.: J. Zhu, H.
Bienyamꢃ), Wiley-VCH, Weinheim, 2005; b) Domino Reactions in
Organic Synthesis (Eds.: L. F. Tietze, G. Brasche, K. M. Gericke),
Wiley-VCH, Weinheim, 2006. For reviews, see: c) L. F. Tietze,
Chem. Rev. 1996, 96, 115–136; d) H. Bienaymꢃ, C. Hulme, G.
Oddon, P. Schmitt, Chem. Eur. J. 2000, 6, 3321–3329; e) R. V. A.
Orru, M. de Greef, Synthesis 2003, 1471–1499; f) D. J. Ramꢄn, M.
Yus, Angew. Chem. 2005, 117, 1628–1661; Angew. Chem. Int. Ed.
2005, 44, 1602–1634; g) N. Isambert, R. Lavilla, Chem. Eur. J. 2008,
14, 8444–8454; h) J. D. Sunderhaus, S. F. Martin, Chem. Eur. J. 2009,
15, 1300–1308; i) B. B. Toure, D. G. Hall, Chem. Rev. 2009, 109,
4439–4486.
[2] For reviews, see: a) H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew.
Chem. 2001, 113, 2056–2075; Angew. Chem. Int. Ed. 2001, 40, 2004–
2021; b) B. M. Trost, Acc. Chem. Res. 2002, 35, 695–705; c) M. D.
Burke, S. L. Schreiber, Angew. Chem. 2003, 115, 48–60; Angew.
Chem. Int. Ed. 2004, 43, 46–58; d) P. A. Wender, V. A. Verma, T. J.
Paxton, T. H. Pillow, Acc. Chem. Res. 2008, 41, 40–49; e) I. S.
Young, P. S. Baran, Nat. Chem. Biol. 2009, 1, 193–205.
General procedure D (solid-phase approach, supported aldehyde): A so-
lution of isocyanide (1.39 mmol, 8 equiv) and aldehyde (8 equiv) in THF
(3 mL) was added to the solid-supported aldehyde (200 mg, loading
0.82 mmolgÀ1
, 174 mmol). The resulting suspension was added to a
borane solution in THF (1m, 3.48 mmol, 20 equiv) in a falcon tube at
08C. The bath was removed and the reaction mixture was stirred for 48 h
at room temperature. The mixture was then transferred to a solid-phase
syringe, and washed with dichloromethane (3ꢂ5 mL). The cleavage was
made depending on the nature of the resin. The unreacted aldehyde was
scavenged with
a suspension of p-toluenesulfonylhydrazide polymer-
bound in dichloromethane. In this way, oxazolidines 5 and aziridines 7
(without adding the aldehyde in solution) were obtained.
General procedure E (solid-phase approach, supported isocyanide): A so-
lution of aldehyde (1.39 mmol, 17 equiv) in THF (3 mL) was added to
the solid-supported isocyanide (200 mg, 3-isocyanophenylether 2-chloro-
trityl polymer-bound, loading 0.4 mmolgÀ1, 80 mmol). The resulting sus-
pension was added to
a borane solution in THF (1m, 3.48 mmol,
20 equiv) in a falcon tube at 08C. The bath was removed and the reaction
mixture was stirred for 48 h at room temperature. The mixture was then
transferred to a solid-phase syringe, and washed with dichloromethane
(3ꢂ5 mL). The cleavage was made through cleavage procedure B (see
below). In this way, oxazolidines 5 and the four-component adducts 6
(adding the corresponding dipolarophile, 25 equiv) were obtained.
[3] For reviews, see: a) A. Dçmling, I. Ugi, Angew. Chem. 2000, 112,
3300–3344; Angew. Chem. Int. Ed. 2000, 39, 3168–3210; b) J. Zhu,
Eur. J. Org. Chem. 2003, 1133–1144; c) V. Nair, C. Rajesh, A. U.
Vinod, S. Bindu, A. R. Sreekanth, J. S. Mathen, L. Balagopal, Acc.
Chem. Res. 2003, 36, 899–907; d) A. Dçmling, Chem. Rev. 2006,
106, 17–89; e) L. El Kaim, L. Grimaud, Tetrahedron 2009, 65, 2153–
2171; f) B. Ganem, Acc. Chem. Res. 2009, 42, 463–472.
[4] For recent examples in isocyanide chemistry, see: a) A. Dos Santos,
L. El Kaim, L. Grimaud, C. Ronsseray, Chem. Commun. 2009,
3907–3909; b) H. Mihara, Y. Xu, N. E. Shepherd, S. Matsunaga, M.
Shibasaki, J. Am. Chem. Soc. 2009, 131, 8384–8385; c) M. C. Pir-
rung, S. Ghorai, T. R. Ibarra-Rivera, J. Org. Chem. 2009, 74, 4110–
Cleavage A: Hydroxymethyl polystyrene resin: The resin-bound product
was suspended in THF (2 mL). A solution of NaOMe in MeOH (0.5m,
1 mmol, 5.7 equiv) was then added and the reaction mixture was stirred
for 24 h. Afterwards, the solid was removed by filtration washing with di-
chloromethane (2 mL). H2O (2 mL) was added and the mixture was ex-
tracted with dichloromethane (2ꢂ5 mL). The combined organic extracts
Chem. Eur. J. 2010, 16, 7904 – 7915
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7913

R. Lavilla et al.
4117; d) F. Sha, X. Huang, Angew. Chem. 2009, 121, 3510–3513;
Angew. Chem. Int. Ed. 2009, 48, 3458–3461; e) X. Li, Y. Yuan, C.
Kan, S. J. Danishefsky, J. Am. Chem. Soc. 2008, 130, 13225–13227;
f) K. V. Luzyanin, A. G. Tskhovrebov, M. F. C. Guedes da Silva, M.
Haukka, A. J. L. Pombeiro, V. Y. Kukushkin, Chem. Eur. J. 2009, 15,
5969–5978; g) A. V. Lygin, O. V. Larionov, V. S. Korotkov, A. De
Meijere, Chem. Eur. J. 2009, 15, 227–236; h) T. Yue, M-X. Wang,
D.-X. Wang, J. Zhu, Angew. Chem. 2008, 120, 9596–9599; Angew.
Chem. Int. Ed. 2008, 47, 9454–9457; i) L. A. Wessjohann, D. G.
Rivera, O. E. Vercillo, Chem. Rev. 2009, 109, 796–814; j) J. Vicente,
I. Saura-Llamas, J.-A. Garcia-Lopez, D. Bautista, Organometallics
2009, 28, 448–464.
k) H.-S. Yeom, J.-E. Lee, S. Shin, Angew. Chem. 2008, 120, 7148–
7151; Angew. Chem. Int. Ed. 2008, 47, 7040–7043.
[12] Incidentally, Hesse reported an attempt to perform such a reaction,
but the expected adduct could not be isolated. See reference [7a].
[13] For alternative approaches to the synthesis of five-membered rings
based on related [3+2] cycloadditions, see: a) D. Bonne, L. Salat, J.
Dulcere, J. Rodriguez, Org. Lett. 2008, 10, 5409–5412; b) A. R. Sia-
maki, B. A. Arndtsen, J. Am. Chem. Soc. 2006, 128, 6050–6051;
c) A. DeAngelis, M. T. Taylor, M. J. Fox, J. Am. Chem. Soc. 2009,
131, 1101–1105.
[14] For a recent review, see: a) D. Tejedor, F. Garcꢅa-Tellado, Chem.
Soc. Rev. 2007, 36, 484–491. For recent results, see: b) D. Tejedor,
A. Sꢇntos-Expꢄsito, F. Garcꢅa-Tellado, Chem. Eur. J. 2007, 13, 1201–
1209; c) D. Tejedor, S. Lopez-Tosco, J. Gonzalez-Platas, F. Garcia-
Tellado, Chem. Eur. J. 2009, 15, 838–842.
[5] For results from our group, see: a) J. L. Dꢅaz, M. Miguel, R. Lavilla,
J. Org. Chem. 2004, 69, 3550–3553; b) N. A. O. Williams, C.
Masdeu, J. L. Dꢅaz, R. Lavilla, Org. Lett. 2006, 8, 5789–5792; c) C.
Masdeu, E. Gꢄmez, N. A. O. Williams, R. Lavilla, Angew. Chem.
2007, 119, 3103–3106; Angew. Chem. Int. Ed. 2007, 46, 3043–3046;
d) M. J. Arꢃvalo, N. Kielland, C. Masdeu, M. Miguel, N. Isambert,
R. Lavilla, Eur. J. Org. Chem. 2009, 617–625.
[15] a) V. Nair, A. U. Vinod, R. S. Menon, V. Shanti, R. L. Varma, S. Viji,
S. Mathew, R. Srinivas, Tetrahedron 2003, 59, 10279–10286; b) A.
Alizadeh, S. Rostamnia, N. Zoreh, Q. Oskueyan, Synlett 2007, 1610–
1612.
[6] For representative examples of the participation of boron reagents
in MCRs, see: a) N. A. Petasis in Multicomponent Reactions (Eds.: J.
Zhu, H. Bienyamꢃ) Wiley-VCH, Weinheim, 2005, Chapter 7,
pp. 199–222; b) S. J. Patel, J. F. Jamison, Angew. Chem. 2003, 115,
1402–1405; Angew. Chem. Int. Ed. 2003, 42, 1364–1367; c) N. Chris-
tinat, R. Scopelliti, K. Severin, Angew. Chem. 2008, 120, 1874–1878;
Angew. Chem. Int. Ed. 2008, 47, 1848–1852; d) X. Zhang, R. C.
Larock, Org. Lett. 2003, 5, 2293–2296.
[7] a) H. Witte, W. Gulden, G. Hesse, Liebigs Ann. Chem. 1968, 716, 1–
10; b) For the preliminary communication, see: G. Hesse, H. Witte,
W. Gulden, Angew. Chem. 1965, 77, 591; Angew. Chem. Int. Ed.
Engl. 1965, 4, 569.
[16] For recent results, see: a) S. Wang, M. X. Wang, D. X. Wang, J. Zhu,
Eur. J. Org. Chem. 2007, 4076–4080.
[17] a) G. Bentabed-Ababsa, A. Derdour, T. Roisnel, J. A. Sꢇez, L. R.
Domingo, F. Mongin, Org. Biomol. Chem. 2008, 6, 3144–3157; b) G.
Bentabed-Ababsa, A. Derdour, T. Roisnel, J. A. Sꢇez, P. Pꢃrez, E.
Chamorro, L. R. Domingo, F. Mongin, J. Org. Chem. 2009, 74,
2120–2133.
[18] For the beneficial use of 2-chloropyridine to stabilise electrophilic
intermediates, see: J. W. Medley, M. Movassaghi, J. Org. Chem.
2009, 74, 1341–1344.
[19] V. V. Tumanov, A. A. Tishkov, H. Mayr, Angew. Chem. 2007, 119,
3633–3636; Angew. Chem. Int. Ed. 2007, 46, 3563–3566.
[8] To the best of our knowledge, only one reference citing this work
(dealing with metal coordination aspects) has appeared to date in
SciFinder: J. L. Schneider, V. G. Young, Jr., W. B. Tolman, Inorg.
Chem. 2001, 40, 165–168.
[9] a) G. Hesse, H. Witte, Angew. Chem. 1963, 75, 791–792; Angew.
Chem. Int. Ed. Engl. 1965, 2, 617; b) J. Casanova, Jr., H. R. Kiefer,
J. Org. Chem. 1969, 34, 2579–2583; c) M. Tamm, T. Lꢆgger, F. E.
Hahn, Organometallics 1996, 15, 1251–1256. Also see: d) H. Witte,
P. Mischke, G. Hesse, Liebigs Ann. Chem. 1969, 722, 21–28; e) J. F.
Stevens, Jr., J. W. Bevan, R. F. Curl, Jr., R. A. Geanangel, M. Grace
Hu, J. Am. Chem. Soc. 1977, 99, 1442–1444.
[10] For reviews on azomethine ylides, see: a) W. Eberbach, Sci. Synth.
2004, 27, 441–498; b) C. Nꢇjera, J. M. Sansano, Curr. Org. Chem.
2003, 7, 1105–1150; c) I. Coldham, R. Hufton, Chem. Rev. 2005,
105, 2765–2809; d) G. Pandey, P. Banerjee, S. R. Gadre, Chem. Rev.
2006, 106, 4484–4517; e) H. Pellissier, Tetrahedron 2007, 63, 3235–
3285. Also see: f) V. Nair, T. D. Suja, Tetrahedron 2007, 63, 12247–
12275; g) L. M. Stanley, M. P. Sibi, Chem. Rev. 2008, 108, 2887–
2902; h) T. M. V. D. Pinho e Melo, Eur. J. Org. Chem. 2006, 2873–
2888.
[11] For recent and relevant results on azomethine ylides, see: a) R. E.
Looper, M. T. C. Runnegar, R. M. Williams, Tetrahedron 2006, 62,
4549–4562; b) C. Nꢇjera, M. de Gracia Retamosa, J. M. Sansano,
Angew. Chem. 2008, 120, 6144–6147; Angew. Chem. Int. Ed. 2008,
47, 6055–6058; c) H. A. Dondas, C. W. G. Fishwick, X. Gai, R.
Grigg, C. Kilner, N. Dumrongchai, B. Kongkathip, N. Kongkathip,
C. Polysuk, V. Sridharan, Angew. Chem. 2005, 117, 7742–7746;
Angew. Chem. Int. Ed. 2005, 44, 7570–7574; d) M.-Z. Wang, H.-W.
Xu, Y. Liu, M.-K. Wong, Adv. Synth. Catal. 2006, 348, 2391–2396;
e) C. V. Galliford, M. A. Beenen, S. T. Nguyen, K. A. Scheidt, Org.
Lett. 2003, 5, 3487–3490; f) I. Coldham, S. Jana, L. Watson, C. D.
Pilgram, Tetrahedron Lett. 2008, 49, 5408–5410; g) A. F. Khlebnikov,
M. S. Novikov, P. P. Petrovskii, J. Magull, A. Ringe, Org. Lett. 2009,
11, 979–982; h) P. J. S. Gomes, C. M. Nunes, A. A. C. C. Pais,
T. M. V. D. Pinho e Melo, L. G. Arnaut, Tetrahedron Lett. 2006, 47,
5475–5479; i) P. Garner, U. H. Kaniskan, J. Org. Chem. 2005, 70,
10868–10871; j) C. Nꢇjera, M. de Gracia Retamosa, J. M. Sansano,
A. de Cozar, F. P. Cossio, Eur. J. Org. Chem. 2009, 2385–2400;
[20] For a recent example of [3+2] cycloadditions with cinnamaldehydes,
see: B.-C. Hong, K.-L. Liu, C.-W. Tsai, J.-H Liao, Tetrahedron Lett.
2008, 49, 5480–5483.
[21] For recent reviews on aziridine chemistry, see: a) J. B. Sweeney,
Chem. Soc. Rev. 2002, 31, 247–258; b) A. Padwa, S. S. Murphree,
ARKIVOC 2006, 3, 6–33.
[22] For relevant examples of aziridine synthesis through azomethine
ylides, see: a) A. L. Williams, J. N. Johnston, J. Am. Chem. Soc.
2004, 126, 1612–1613; b) X. Zhang, M. Yan, D. Huang, Org.
Biomol. Chem. 2009, 7, 187–192. For the use of aziridines in cyclo-
additions, see: c) P. Dauban, G. Malik, Angew. Chem. 2009, 121,
9188–9191; Angew. Chem. Int. Ed. 2009, 48, 9026–9029, and refer-
ences therein.
[23] For the interaction of diborane with isocyanides, see: S. Bredasola,
F. Rossetto, G. Piosi, Tetrahedron Lett. 1965, 6, 4775–4778.
[24] For the reaction of triphenylboron with benzhydryl isocyanide lead-
ing to the generation of the corresponding nitrile ylide adduct, see:
G. Bittner, H. Witte, G. Hesse, Liebigs Ann. Chem. 1968, 713, 1–11.
[25] For the recently reported inverse demand reactions with azomethine
ylides, see: a) P. D. Pohlhaus, R. K. Bowman, J. S. Johnson, J. Am.
Chem. Soc. 2004, 126, 2294–2295; b) P. D. Pohlhaus, R. K. Bowman,
J. S. Johnson, J. Am. Chem. Soc. 2006, 128, 6520.
[26] For examples of cycloadditions with azomethine ylides and acrylates,
see: a) P. DeShong, D. A. Kell, D. R. Sidler, J. Org. Chem. 1985, 50,
2309–2315; b) P. Garner, F. Arya, W.-B. Ho, J. Org. Chem. 1990, 55,
412–414; c) M. T. Epperson, D. Y. Gin, Angew. Chem. 2002, 114,
1856–1858; Angew. Chem. Int. Ed. 2002, 41, 1778–1780; d) A. S.
Gothelf, K. V. Gothelf, R. G. Hazell, K. A. Jorgensen, Angew.
Chem. 2002, 114, 4410–4412; Angew. Chem. Int. Ed. 2002, 41, 4236–
4238; e) W. H. Pearson, P. Stoy, Y. Mi, J. Org. Chem. 2004, 69, 1919–
1939; f) C.-J. Wang, G. Liang, Z.-Y. Xue, F. Gao, J. Am. Chem. Soc.
2008, 130, 17250–17251. See also refs. [11i,h].
[27] For recent examples of intramolecular azomethine ylide cycloaddi-
tions, see: a) G. Bashiardes, I. Safir, A. S. Mohamed, F. Barbot, J.
Laduranty, Org. Lett. 2003, 5, 4915–4918; b) I. Coldham, A. J. M.
Burrell, L. E. White, H. Adams, N. Oram, Angew. Chem. 2007, 119,
6271–6274; Angew. Chem. Int. Ed. 2007, 46, 6159–6162; c) F. Lev-
esque, G. Belanger, Org. Lett. 2008, 10, 4939–4942.
7914
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 7904 – 7915

Boron-Based Dipolar Multicomponent Reactions
FULL PAPER
[28] For the synthesis of the aldehyde 1n, see: I. Kim, S. G. Kim, J. Choi,
G. H. Lee, Tetrahedron 2008, 64, 664–671.
ChemMedChem 2008, 3, 302–315; e) (oxazolidine scaffold) R. B.
Walker, V. N. Dholakia, K. L. Brasfield, R. Bakhtiar, Gen. Pharma-
col. 1998, 30, 725–731.
[29] For previous access to this system, see: a) J. Pospisil, M. Potacek,
Eur. J. Org. Chem. 2004, 710–716; b) A. F. Khlebnikov, M. S. Novi-
kov, A. A. Bespokoev, R. R. Kostikov, J. Kopf, Z. A. Starikova,
M. Y. Antipin, Russ. J. Org. Chem. 2005, 41, 922–932, and referen-
ces therein.
[30] For reports on the biological relevance of this scaffold, see: a) T.
Dubuffet, A. Newman-Tancredi, D. Cussac, V. Audinot, A. Loutz,
M. J. Millan, G. Lavielle, Bioorg. Med. Chem. Lett. 1999, 9, 2059–
2064; b) M. L. Bolognesi, M. Bartolini, A. Cavalli, V. Andrisano, M.
Rosini, A. Minarini, C. Melchiorre, J. Med. Chem. 2004, 47, 5945–
5952.
[38] In this respect, and following a sharp suggestion from a referee, we
explored the in situ generation of borane reactants, the MCR to be
performed in tandem. In this way, tributylborane (3b) was prepared
by interaction of nBuLi with BF3·Et2O in THF, and afterwards the
addition of aromatic aldehyde 1a and isocyanide 2a, following gen-
eral procedure A, successfully afforded oxazolidine 5j (30% unopti-
mised yield). This result opens interesting synthetic possibilities on
the use of a broader set of borane components, and is currently
under study in our laboratory.
[39] a) A. B. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) A. B. Becke,
Phys. Rev. A 1988, 38, 3098–3100; c) C. Lee, W. Yang, R. G. Parr,
Phys. Rev. B 1988, 37, 785–789.
[31] For a review, see: L. Banfi, L.; G. Guanti, R. Riva, A. Basso, Curr.
Opin. Drug Discovery Dev. 2007, 10, 704–714.
[32] For a recent review, see: a) K. Harju, J. Yli-Kauhaluoma, Mol. Di-
versity 2005, 9, 187–207. Also see: b) M. Albers, T. Meyer in Hand-
book of Combinatorial Chemistry, Vol. 1 (Eds.: K. C. Nicolaou, R.
Hanko, W. Hartwig), Wiley-VCH, Weinheim, 2002, Chapter 16,
pp. 440–469.
[33] S. V. Ley, I. R. Baxendale, R. N. Bream, P. S. Jackson, A. G. Leach,
D. A. Longbottom, M. Nesi, J. S. Scott, R. I. Storer, S. J. Taylor,
Perkin 1 2000, 3815–4195.
[40] V. Rassolov, J. A. Pople, M. Ratner, P. C. Redfern, L. A. Curtiss, J.
Comput. Chem. 2001, 22, 976–984.
[41] a) C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154–2161;
b) C. Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523–5527.
[42] R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem. Phys.
1980, 72, 650–654.
[43] http://cccbdb.nist.gov/.
[44] Gaussian 03, Revision E.01, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[34] To our knowledge, this is the first application of solid-phase tech-
niques to chemo-differentiating ABB’ processes.
[35] For precedents on the use of solid-supported isocyanides, see: a) C.
Hulme, J. Peng, G. Morton, J. M. Salvino, T. Herpin, R. Labaudi-
niere, Tetrahedron Lett. 1998, 39, 7227–7230; b) B. A. Kulkami, A.
Ganesan, Tetrahedron Lett. 1999, 40, 5633–5636. Also see
refs. [3a,d].
[36] In this experiment the ratio of dipolarophile (4a) and aldehyde (1a)
was adjusted to 25:2 to avoid the competitive process leading to ox-
azolidine 5l.
[37] The 4CR, the oxazolidine and the aziridine scaffolds have relevant
biological applications. For example, see: a) (4CR scaffold) J. Olsen,
P. Seiler, B. Wagner, H. Fisher, T. Tshopp, U. Obst-Sander, D. W.
Banner, M. Kansy, K. Mꢆller, F. Diederich, Org. Biomol. Chem.
2004, 2, 1339–1352; b) H. Agirbas, S. Guner, F. Budak, S. Keceli, F.
Kandemirli, N. Shvets, V. Kovalishyn, A. Dimoglo, Bioorg. Med.
Chem. 2007, 15, 2322–2333; c) J. W. Sadownik, D. Philp, Angew.
Chem. 2008, 120, 10113–10118; Angew. Chem. Int. Ed. 2008, 47,
9965–9970; d) (aziridine scaffold) B. Degel, P. Staib, S. Rohrer, J.
Sheiber, E. Martina, C. Bꢆchold, J. Morschhꢈuser, T. Schirmeister,
Received: February 9, 2010
Published online: May 27, 2010
Chem. Eur. J. 2010, 16, 7904 – 7915
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7915