Azomethine Ylides from Nitrones: Using Catalytic nBuLi for the Totally Stereoselective Synthesis of trans-2-Alkyl-3-oxazolines

Article abstract of DOI:10.1002/chem.201602159

The cycloaddition of azomethine ylide N-oxides (nitrone ylides) with aldehydes provides 3-oxazolines in a completely stereoselective manner in the presence of a catalytic amount of n-butyllithium. The process involves an initial nucleophilic attack on the aldehyde, followed by intramolecular oxygen addition to the nitrone moiety and lithium-assisted elimination of water, regenerating the catalytic species. Various Li-based catalytic systems are possible and the in situ generated water is required for continuing the catalytic cycle. The best results are observed with 20 mol % of n-butyllithium, whereas the use of stoichiometric amounts inhibit the rate of catalysis. Experimental, spectroscopic, and computational mechanistic studies have provided evidence of lithium-ion catalysis and rationalized several competing catalytic pathways.

Full text of DOI:10.1002/chem.201602159

A Journal of
Accepted Article
Title: Azomethine Ylides from Nitrones: Using Catalytic n-BuLi for the Totally Stereo-
selective Synthesis of trans-2-Alkyl-3-Oxazolines
Authors: Pedro Merino; Tomas Tejero; Ignacio Delso; Veronica Juste-Navarro
This manuscript has been accepted after peer review and the authors have elected
to post their Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by using the Digital
Object Identifier (DOI) given below. The VoR will be published online in Early View as
soon as possible and may be different to this Accepted Article as a result of editing.
Readers should obtain the VoR from the journal website shown below when it is pu-
blished to ensure accuracy of information. The authors are responsible for the content
of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201602159
Link to VoR: http://dx.doi.org/10.1002/chem.201602159
Supported by

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
Azomethine Ylides from Nitrones: Using Catalytic n-BuLi for the
Totally Stereoselective Synthesis of trans-2-Alkyl-3-Oxazolines
Veronica Juste-Navarro,[a] Ignacio Delso,[b] Tomás Tejero[a] and Pedro Merino*[a]
Dedicated to Prof. Miguel Yus for his contribution to the field of organolithium chemistry
Abstract: The cycloaddition of azomethine ylide N-oxides (nitrone
ylides) with aldehydes provides 3-oxazolines in
a complete
stereoselective way in the presence of a catalytic amount of n-
butyllithium. The process involves an initial nucleophilic attack to the
aldehyde followed by intramolecular oxygen addition to the nitrone
moiety and lithium-assisted elimination of water regenerating the
catalytic species. Various Li-based catalytic systems are possible
and the self-generated water is required for continuing the catalytic
cycle. The best results are observed with 20 mol% of n-butyllithium
in contrast to the use of stoichiometric amounts that inhibit the
catalyzed rate. Experimental, spectroscopic and computational
mechanistic studies have provided evidences of lithium ion catalysis
and allowed rationalizing the several competing catalytic processes.
Scheme 1. Reactivity of N-(ethoxycarbonylmethyl) nitrones 1 precursors of
nitrone ylides.
3-Oxazolines are heterocyclic systems of interest,[14] difficult to
be prepared, and no methods are reported for the
stereoselective synthesis of 2,5-disubsutituted derivatives. In
contrast to well-established methods for accessing 2-
oxazolines,[15] the preparation of isomeric 3-oxazolines (2,5-
dihydrooxazoles) have received far less attention.[16] Taking into
account a retrosynthetic analysis considering the fragments
which join to form the heterocyclic nucleus (Figure 1), all these
methods belong to C + OC2N approach. Very recently, Zhong
and co-workers developed an acid-promoted formal [3+2]
cycloaddition between donor-acceptor oxiranes and nitriles,[17]
the only method representing a COC + CN approach.
The chemistry of azomethine ylides has been extensively
studied in the past.[1] In particular, N-metalated azomethine
ylides have received considerable attention because of their
utility in asymmetric catalytic reactions.[2] We have reported the
use of a novel class of azomethine ylides derived from nitrones
1 (azomethine ylide N-oxides or nitrone ylides) in a tandem
Michael-Mannich reaction with α,β-unsaturated esters to provide
N-hydroxypyrrolidines in excellent yields and complete
stereoselectivity (Scheme 1).[3] The reaction required
stoichiometric amounts of n-BuLi to generate the ylide and the
lithium ion was required for activating the reagents in both
steps.[4]
Organolithium compounds have been extensively employed as
bases in stoichiometric amount for metalation reactions,[5] but
their use as catalysts, other than in polymerization reactions,[6] is
an ongoing research challenge.[7] Although several catalytic
processes involving lithium salts have already been documented
for lithium chloride,[8] lithium bromide,[9] lithium perchlorate[10] and,
in a less extent, lithium hydroxide,[11] to date, there is only one
report dealing with the use of an organic alkali metal compound
in catalytic reactions.[12] In continuation of our work on the
reactivity of nitrone ylides with electrophiles we have
investigated the reaction with aldehydes and found that using
catalytic amounts of n-BuLi, 3-oxazolines are obtained in good
yields and complete selectivity (Scheme 1).[13]
Figure 1. Retrosynthetic approaches for 3-oxazolines
Herein, we report a novel and efficient catalytic stereospecific
synthesis of trans-2-alkyl-5-substituted-3-oxazolines based on
an OC + CNC approach (Figure 1). Various catalytic systems, all
of them based on a lithium salt, are possible. Among them, n-
BuLi provided the best results acting as an efficient pre-catalyst.
To the best of our knowledge, this is the first method that
describes a catalytic synthesis of substituted 3-oxazolines in a
complete stereoselective way and uses aldehydes as starting
materials. A rationale based on DFT computational studies is
provided for explaining both the stereoselectivity and the
catalytic cycles involved in the reaction.
The synthetic approach started by preparing novel C-alkyl
nitrones 1 following the methodology previously reported by
us.[18] Initially we screened reaction conditions developed in our
group for nitrone ylides.[3] Optimization studies are given in Table
1.
[a]
[b]
Mrs. V. Juste-Navarro, Prof. T. Tejero and Prof. P. Merino
Departamento de Síntesis y Estructura de Biomoléculas. Instituto de
Síntesis Química y Catálisis Homogénea (ISQCH). Universidad de
Zaragoza. CSIC. E-50009 Zaragoza. Aragón, Spain.
E-mail: pmerino@unizar.es
Dr. I. Delso
Servicio de Resonancia Magnética Nuclear, Centro de Química y
Materiales de Aragón (CEQMA), Universidad de Zaragoza, CSIC,
Campus San Francisco, 50009 Zaragoza , Spain
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
Table 1. Synthesis of 3-Oxazolines 3. Optimization of Reaction Conditions
The use of stoichiometric amounts of triethylamine and LiBr
(entry 1) led exclusively to trans-3-oxazoline 3a, the elimination
product of the initially expected N-hydroxypyrrolidine.[19] The
reaction can be carried out at 0 ºC and 50º C for longer and
shorter reaction time, respectively, with minimum loss of yield
(entries 2 and 3). On the other hand, both the use of solvents
different from acetonitrile (entries 4 and 5) and lithium salts
different from LiBr (entries 6 and 7) caused evident decreasing
of the yield. Changing the base to DBU (entry 8) or DABCO
(entry 9) improved considerably the yield to more than 90%.
However, when the reaction was carried out in the presence of
molecular sieves (entry 10) the reaction showed a conversion of
only 10% after 24 h, the starting materials being recovered
unaltered. We studied the possibility of using catalytic amounts
of base and/or lithium salt (entries 11-15) and found that 50
mol% of both DABCO and LiBr led to the same result that
stoichiometric amounts. Unfortunately, decreasing to 20 mol%
led to a considerably loss of yield (30%) and formation of
undesired nitrone 4 as a result of transoximation between the
starting nitrone 1a and aldehyde 2a, through hydrolysis of the
former. In this context, we did not observed cross-reactivity
between 4 and 2a. Next, we consider the possibility of
integrating both base and lithium in the same species and found
that LiOH promotes the reaction in 87% yield (entry 16) although
a 8% of 4 was obtained. Again the presence of molecular sieves
blocked the reaction (entry 17). To increase the solubility of
LiOH, ethanol was tested as a solvent but lower yield was
obtained (entry 18). The use of catalytic amounts led to
increased transoximation or no reaction (entries 19-21). When n-
BuLi was used at low temperature the reaction did not progress
after 24 h (entry 22) but warming slowly the reaction up to
ambient temperature during 12 h the yield increased to 74%
(entry 23). In an attempt of adjusting the reaction conditions we
check the reaction after warming up to -40ºC during 4 h (entry
24). At this time, only a yield 40% was observed and starting
materials were recovered indicating that the reaction had not
finished. On the other hand, the use of substoichiometric
amounts of n-BuLi (entry 25) led to a clean reaction in high
chemical yield under the same conditions, revealing a faster rate
of the reaction. By decreasing n-BuLi to 20 mol%, the same
result was obtained (entry 26), no transoximation being
observed. With lower amounts of n-BuLi the reaction required
more time (entry 27) but still 68% chemical yield was obtained.
In all cases, the trans-isomer was the only observed product.
The structure of compound 3a was confirmed by X-ray
crystallography.[20] At this point, it became evident that the
process was catalytic involving lithium species and that different
mechanisms should operate under stoichiometric and catalytic
conditions.
Base
(eq)
Li salt
(eq)
T
(ºC)
t
Yield
(%)[a]
entry
solvent
(h)
1
Et3N (1.0)
LiBr (1.0)
LiBr (1.0)
LiBr (1.0)
LiBr (1.0)
LiBr (1.0)
CuBr (1.0)
LiOAc (1.0)
LiBr (1.0)
LiBr (1.0)
LiBr (1.0)
LiBr (0.5)
LiBr (0.2)
LiBr (1.0)
LiBr (0.5)
LiBr (0.2)
25
0
24
72
6
MeCN
MeCN
MeCN
CH2Cl2
DMSO
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
EtOH
MeCN
MeCN
MeCN
THF
80
72
2
Et3N (1.0)
3
Et3N (1.0)
50
70
4
Et3N (1.0)
25
24
24
24
24
24
24
24
24
24
24
24
48
12
12
12
12
4 d
4 d
24
12
4
68
5
Et3N (1.0)
25
36
6
Et3N (1.0)
25
15[b]
60
7
Et3N (1.0)
25
8
DBU (1.0)
25
91
9
DABCO (1.0)
DABCO (1.0)
DABCO (1.0)
DABCO (1.0)
DABCO (0.5)
DABCO (0.5)
DABCO (0.2)
25
94
10[b]
11
12
13
14
15
16
17[b]
18
19
20
21
22
23
24
25
26
27
25
10[c]
88
25
25
46
25
90
25
92
25
30(30)
87(8)
10[c]
68
LiOH (1.0)
25
LiOH (1.0)
LiOH (1.0)
LiOH (0.5)
LiOH (0.2)
LiOH (0.1)
n-BuLi (1.0)
n-BuLi (1.0)
n-BuLi (1.0)
n-BuLi (0.5)
n-BuLi (0.2)
n-BuLi (0.1)
25
25
25
80(15)
10(50)
n.r.[c]
<10[c]
74
25
25
-80
rt[d]
-40[d]
-40[d]
-40[d]
-40[d]
THF
THF
40[c]
90
4
THF
90[e]
68
The optimized reaction conditions (20 mol% BuLi) were applied
to various nitrones and aldehydes (Table 2). For the purpose of
comparison the reactions were also carried out with 50 mol%
DABCO / LiBr (Table 1, entry 14) and identical results were
obtained. Thus, both reaction conditions can be equally used.
The reaction proceeded smoothly for nitrones 1a-d with both
aliphatic and aromatic aldehydes. In the case of aromatic
aldehydes with electron-donating groups (Table 2, entries 4 and
5) and α,β-unsaturated aldehydes (Table 2, entries, 10, 11 and
17) additional reaction time was needed for achieving good
4
THF
24
THF
[a] Only the trans-isomer was obtained. When compound 4 was also obtained
the yield is given in brackets. [b] The reaction was conducted in the presence
of 4 Å MS. [c] Starting materials were recovered. [d] n-BuLi was added at -
80ºC and after 5 min the reaction was warmed to the stated temperature. [e]
An identical result was obtained by carrying out the reaction at -80ºC to rt

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
chemical yield. With the latter it existed the possibility of 1,4-
addition but it was never observed.
the favorable case of nitrone 4 and aldehyde 2a we observed
the formation of a product in good yield after 48 h (Scheme 2).[22]
However, that product was confirmed to be the corresponding 2-
oxazoline 5 by X-ray crystallography.[20] The formation of 5 is
Table 2. Synthesis of 3-Oxazolines 5. Scope of the reaction.[a]
possibly due to
a migration of the double bond as a
consequence of the stabilization by conjugation with the
aromatic ring.[23]
entry
1
R1
R2
4-NO2C6H4
4-NO2C6H4
Ph
1
2
3
3a
3a
3b
3c
3d
3e
3f
Yield (%)[b]
90 (92)
74 (70)
89 (90)
78 (69)
75 (64)
86 (82)
88 (89)
89 (85)
93 (91)
78 (71)
69 (53)
93 (90)
88 (85)
92 (90)
90 (87)
78 (80)
72 (68)
92 (89)
90 (86)
86 (89)
92 (90)
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Bu
i-Bu
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1b
1c
1c
1c
1c
1d
1d
1a
1a
2a
2a
2b
2c
2d
2e
2f
2[c]
3
4[d]
5[e]
6
4-MeC6H4
4-MeOC6H4
i-Pr
Scheme 2. Exceptional reaction of nitrone 4 with aldehyde 2a.
We also initiate preliminary studies on the chiral version of the
reaction. Unfortunately only negative results were obtained. The
reaction in the presence of 20 mol% (-)-sparteine only 6 %ee.
The use of a chiral lithum alkoxide derived from (R)-BINOL (20
mol%) and n-BuLi (20 mol%) afforded great amounts of
transoximation (30% of 3a and 45% of 4) in a similar way to
LiOH and no chiral induction (10% ee for 3a) was observed.
In our previous work with α,β-unstaurated esters and C-aryl
nitrones like 4 we demonstrated that the reaction was stepwise
by isolating the initial Michael intermediate;[3] further
computational studies[4] corroborated the mechanism providing
the N-hydroxypyrrolidine as the final product of the reaction. On
the contrary, for the reaction between 1a and 2a, which can be
conducted under catalytic conditions, any attempt of isolating
either an intermediate or the N-hydroxypyrrolidine precursor
from 3-oxazoline 3 failed. These observed differences together
with the fact that molecular sieves blocked the reaction,
demonstrating that water is necessary, and the observation that
the reaction is faster with substoichiometric amounts of n-BuLi
prompted us to investigate the mechanism of the reaction
computationally and by NMR spectroscopy.
7
i-Bu
8
cyclopentyl
PhCH2
2g
2h
2i
3g
3h
3i
9
10[d]
11[d]
12
13
14
15
16
17[d]
18
19
20
21
(E)-MeCH=CH
(E)-PhCH=CH
4-NO2C6H4
Ph
2j
3j
2a
2b
2a
2b
2e
2i
3k
3l
cyclohexyl
cyclohexyl
cyclohexyl
cyclohexyl
cyclopentyl
cyclopentyl
i-Pr
4-NO2C6H4
Ph
3m
3n
3o
3p
3q
3r
i-Pr
(E)-MeCH=CH
4-NO2C6H4
Ph
2a
2b
2k
2l
We modelled the reaction between nitrone NI and aldehyde AL
catalyzed by n-BuLi considering the formation of the
corresponding nitrone ylide YL as the first step. Several
approaches are possible for the reaction between ylide YL and
aldehyde AL. A comprehensive study of all the possible
2-piridyl
2-furyl
3s
3t
i-Pr
[a] Reactions were run using nitrones 1 (0.5 mmol scale) and 1.0 eq of
aldehyde in the stated solvent (0.5 M) unless otherwise indicated.
pathways
has
been
carried
out
at
M06-2X/6-
2
311+G(d,p)/PCM=THF level of theory using geometries
optimized at M06-2X/6-31G(d)/PCM=THF level. Further
discussion will only refer to the preferred pathway leading to the
trans-isomer (for the complete computational analysis see
Supporting Information).
Conditions A: 20 mol% BuLi, THF, -80º C to -40ºC, 4 h. Conditions B: 50
mol% DABCO, 50 mol% LiBr, MeCN, rt, 24 h. [b] Isolated yield after
purification by column chromatography; values corresponding to conditions A.
Isolated yields corresponding to conditions B are given between brackets Only
the trans-isomer was obtained. [c] Reaction was run in 6.0 mmol scale (1.04 g
of nitrone). [d] Reaction was run for 36 h. [e] Reaction was run for 48 h..
Once the ylide is formed, the aldehyde coordinates the lithium
atom to form the starting complex SC in which both reagents are
activated. The catalytic cycle, illustrated in Scheme 3, starts
from SC which is the most stable intermediate in the cycle. The
most favorable TS1 leads to IN1 with a barrier of ΔG = 5.0
kcal/mol which resulted to be the rate-limiting step. Further
exchange of a solvent molecule is favored by 14.6 kcal/mol
giving rise to IN2 which evolves through the preferred Re attack
Although C-alkyl nitrones afforded good reactivity, there
still remained limitations on the substrate. When C-aryl nitrones
were used we did not observe any reactivity, illustrating the well-
known fact that C-aryl nitrones are less reactive than C-alkyl
nitrones.[21] The lack of reactivity was also observed for
heteroaryl nitrones (C-(2-pyridyl) and C-(2-furylyl)) and α,β-
unsaturated nitrone derived from cinnamalehyde. Indeed, only in

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
(TS2 located 1.3 kcal/mol below the ground state) to IN3.
Exchange of coordination in this intermediate formed IN4 (7.3
kcal/mol more stable) which can be transformed into IN6 through
a 1,3-H shift.
The prototropy observed between IN4 and IN6 can take place
through an intramolecular transition state (located 10.8 kcal/mol
above the ground state, not shown in Scheme 3; see SI) or the
more stable TS3 (located 14.2 kcal/mol below the ground state)
after formation of IN5, involving a water molecule generated in
the last step of the catalytic cycle. Under catalytic conditions
only the first round of the cycle takes place through the above
mentioned intramolecular transition state higher in energy, at
which time water is produced and the catalytic cycle carries on
through IN5 and TS3.
In agreement with this hypothesis, the use of molecular sieves
should reduce drastically the rate of the reaction as indeed, it is
observed experimentally. When equimolar amounts of base/LiBr
are used, the reaction is completed in one round and in this case,
the prototropy in IN4 can be facilitated by the protonated base,
also through a bimolecular process. In the case of using
equimolar amounts of n-BuLi, transformation of IN4 should
account through the intramolecular transition state (1,3-H shift),
higher in energy (see SI). This justifies the reduced rate of the
reaction observed with 1.0 eq of n-BuLi. After formation of IN6
from IN5, the reaction proceeds through TS4 located at -21.6
kcal/mol below the ground state to form intermediate IN7 which
after releasing a solvent molecule provides IN8.
Scheme 3.. Catalytic cycle for Li-catalyzed synthesis of 3-oxazolines

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
Figure 2. Energy surface for the reaction between 1 and 2 catalyzed by n-BuLi. The most stable route corresponding to the formation of trans-3 is shown

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
Figure 2 illustrates the energy profile for the formation of the only
observed trans-3-oxazoline PR and completion of the catalytic
cycle. The driving force of the catalytic cycle illustrated in
Scheme 2 is determined by the regeneration of SC from IN8
(through IN9, TS5 and IN10) with concomitant release of the
final product and two molecules of water (actually only one is
produced per round). This final catalyst-turnover step involves a
favored energy (ΔG = -15.8 kcal/mol) for the reaction of 1 (NI)
with 2 (AL).
NMR experiments using 0.2 eq of n-BuLi recording the complete
reaction in real time showed the presence of the final product 3a
without any addition of a proton source, confirming that an N-
hydroxypyrrolidine could not be a real intermediate.
However, when the reaction mixture was maintained at -80ºC
both nitrone and aldehyde are consumed immediately giving rise
to signals that are in agreement with the formation of IN4 (see
SI). This intermediate demonstrated to be stable at -80ºC for 4h,
the formation of 3a being observed after warming at -40ºC.
Since the use of 1.0 eq of n-BuLi does not allow the formation of
water in the catalytic cycle an alternative -and more energetic-
path have to be present, as indeed it is (see SI). By using 0.2 eq
of n-BuLi, warming at -40ºC is enough for completing the
reaction in 4 h. demonstrating the faster pathway illustrated in
Scheme 3 in which assistance of a water molecule render lower
in energy the corresponding TS3 (in the alternative path where
water is not involved, alternative TS3' is, indeed, 5.8 kcal/mol
higher in energy). Thus, both computational and spectroscopic
studies are in complete agreement with the experimental
observations.
In summary, a lithium-catalyzed route towards 3-oxazolines
starting from novel azomethine ylide N-oxides (nitrone ylides)
and aldehydes has been developed. Actually, lithium contributes
to facilitate elimination of water, which serves to push the
catalytic cycle forward. The combination of a base and a lithium
salt or the use of n-BuLi serve as a pre-catalyst to initiate the
catalytic cycle which is self-sufficient until starting materials are
consumed. The only species formally regenerated in the
catalytic cycle are the lithium atom, which, of course, must be
solvated/coordinated along the whole cycle and the water, which
might be considered to have an autocatalytic role. These results
demonstrate that a lithium ion can act as a catalyst confirming
previous computational results of Saa and Capo.[8e] Notably, an
excess of water or the use of lithium salts like LiOH promote
undesired nitrone hydrolysis leading to a transoximation making
the use of catalytic amounts of n-BuLi a unique system for the
progress of the reaction (using DABCO/LiBr is also possible but
below 50 mol% transoximation is also observed). The same
problem arises for the asymmetric version of the reaction when
a chiral lithium alkoxide is used as a base. Both spectroscopic
and computational studies are in agreement with the
experimental observations showing
a faster reaction with
substoichiometric amounts of n-BuLi than with equimolar
amounts. Also, those studies explain the necessity of water for
the progress of the reaction at low temperature. Further studies
directed to other lithium-catalyzed reactions with nitrone ylides
including cataytyic asymmetric versions of the reaction are
currently in progress in our laboratories.
[4]
[5]
P. Merino, T. Tejero, A. Diez Martinez, J. Org. Chem. 2014, 79, 2189-
2202.
(a) C. Strohmann, V. H. Gessner, J. Am. Chem. Soc. 2007, 129, 8952-
8953. (b) Z. Xi, Acc. Chem. Res. 2010, 43, 1342-1351. (c) A. C. Jones,
A. W. Sanders, M. J. Bevan, H. J. Reich, J. Am. Chem. Soc. 2007, 129,
3492-3493.
Acknowledgements
This work was supported by the Spanish Ministerio de
[6]
[7]
[8]
A. F. Halasa, D. N. Schulz, D. P. Tate, V. D. Mochel, Adv. Organomet.
Chem. 1980, 18, 55-97.
Economía
y
Competitividad (MINECO) (project number
CTQ2013-44367-C2-1-P), by the Fondos Europeos para el
Desarrollo Regional (FEDER) and the Gobierno de Aragón
(Zaragoza, Spain, Bioorganic Chemistry Group, E-10). We thank
Prof. J. M. Saá (University of Balearic Islands, Spain) for reading
the manuscript. The authors acknowledge the Institute of
Biocomputation and Physics of Complex Systems (BIFI) at the
University of Zaragoza for computer time at clusters Terminus
and Memento. V. J.-N. thanks MINECO for a pre-doctoral grant
(FPI program).
S. Matsunaga, in Sci. Synth., Knowl. Updates, Vol. 2010/4 (Ed.: M.
Yus), George Thieme, Stuttgart, 2011, pp.209-222.
(a) L. Gupta, A. C. Hoepker, K. J. Singh, D. B. Collum, J. Org. Chem.
2009, 74, 2231-2233. (b) Z. Fang, G.-C. Zhou, S.-L. Zheng, G.-L. He,
J.-L. Li, L. He, D. Bei, J. Mol. Catal. A Chem. 2007, 274, 16-23. (c) N.
Kurono, M. Yamaguchi, K. Suzuki, T. Ohkuma, J. Org. Chem. 2005, 70,
6530-6532. (d) G. Sabitha, B. V. S. Reddy, R. S. Babu, J. S. Yadav,
Chem. Lett. 1998, 773-774. (e) M. Capo, J. M. Saa, J. Am. Chem. Soc.
2004, 126, 16738-16739.
[9]
(a) R. Patel, V. P. Srivastava, L. D. S. Yadav, Adv. Synth. Catal. 2010,
352, 1610-1614. (b) A. Saini, S. Kumar, J. S. Sandhu, Synlett 2006,
1928-1932. (c) A. K. Chakraborti, S. Rudrawar, A. Kondaskar, Eur. J.
Org. Chem. 2004, 3597-3600. (d) C. A. Loeschorn, M. Nakajima, P. J.
McCloskey, J. P. Anselme, J. Org. Chem. 1983, 48, 4407-4410.
Keywords: Azomethine ylides • Nitrones • 3-Oxazolines •
Lithium Catalysis • DFT calculations
[10] (a) A. Guy, L. Serva, Synlett 1994, 647-648. (b) G. Desimoni, G. Faita,
P. P. Righetti, G. Tacconi, Tetrahedron 1991, 47, 8399-8406. (c) W. J.
Kinart, C. M. Kinart, R. Oszczeda, Q. T. Tran, Catal. Lett. 2005, 103,
185-189.
[1]
[2]
W. Eberbach, in Science of Synthesis, Vol. 27 (Eds.: D. Bellus, A.
Padwa), George Thieme, Stuttgart, 2004, pp. 441-498.
(a) C. Najera, J. M. Sansano, Angew. Chem., Int. Ed. 2005, 44, 6272-
6276. (b) G. Pandey, P. Banerjee, S. R. Gadre, Chem. Rev. 2006, 106,
4484-4517. (c) C. Nájera, J. M. Sansano, Top. Heterocycl. Chem. 2008,
12, 117-145. (d) J. Adrio, J. C. Carretero, Chem. Commun. 2011, 47,
6784-6794. (e) R. Narayan, M. Potowski, Z.-J. Jia, A. P. Antonchick, H.
Waldmann, Acc. Chem. Res. 2014, 47, 1296-1310.
[11] (a) S. A. Miller, N. E. Leadbeater, RSC Adv. 2015, 5, 93248-93251. (b)
R. Wolfenden, F. Zhao, J. Am. Chem. Soc. 2004, 126, 8646-8647.
[12] C. Liu, Y. Zhang, Q. Qian, D. Yuan, Y. Yao, Org. Lett. 2014, 16, 6172-
6175.
[13] The reaction between classical N-silver azomethine ylides and
aldehydes has been investigated by Somfai and co-workers (see: B.
Seashore-Ludlow, S. Torssell, P. Somfai, Eur. J. Org. Chem. 2010,
[3]
P. Merino, T. Tejero, A. Diez-Martinez, Z. Gultekin, Eur. J. Org. Chem.
2011, 6567-6573.

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
3927-3933.) but oxazolidines were obtained and either low
diastereoselectivities were obtained with benzaldimines or a weak acid
catalyst was required with benzophenone-derived imines. The
presence of the oxygen atom in ylides derived from 1 is crucial for the
obtention of 3-oxazolines.
[18] A. Diez-Martinez, Z. Gultekin, I. Delso, T. Tejero, P. Merino, Synthesis
2010, 678-688.
[19] Apparently, compound 3a might be formed through dehydration of the
corresponding N-hydroxypyrrolidine which was the initially expected
product in a similar way to the reaction between nitrone ylides and
unsaturated esters (see ref. 3). However, such hydroxylamine is never
formed (see below)
[14] (a) Z. Xiao, J. R. Lu, J. Agric. Food Chem. 2014, 62, 6487-6497. (b) T.-
T. Zeng, J. Xuan, W. Ding, K. Wang, L.-Q. Lu, W.-J. Xiao, Org. Lett.
2015, 17, 4070-4073.
[20] The authors have deposited the atomic coordinates for these structures
with the Cambridge Crystallographic Data Centre. Deposition numbers
are as follows: 3a, CCDC 1444429; 3f, CCDC 1444430. The
coordinates can be obtained on request from the Director, Cambridge
Crystallographi Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK.
For ORTEP graphics see Supporting Information.
[15] (a) T. G. Gant, A. I. Meyers, Tetrahedron 1994, 50, 2297-2360. (b) K.
Kempe, M. Lobert, R. Hoogenboom, U. S. Schubert, J. Comb. Chem.
2009, 11, 274-280. (c) M. Brandstätter, F. Roth, N. W. Luedtke, J. Org.
Chem. 2015, 80, 40-51.
[16] (a) V. Yeh, R. Iyengar, in Comprehensive Heterocyclic Chemistry III,
Vol. 4 (Eds.: A. R. Katritzky, C. A. Ramsden, E. F. V. Scriven, R. J. K.
Taylor), Elsevier, Oxford, 2008, pp. 487-543. (b) A. Padwa, J.
Smolanoff, S. I. Wetmore, Jr., J. Org. Chem. 1973, 38, 1333-1340. (c)
A. Padwa, P. H. J. Carlsen, J. Am. Chem. Soc. 1976, 98, 2006-2008.
(d) A. Padwa, J. K. Rasmussen, A. Tremper, J. Am. Chem. Soc. 1976,
98, 2605-2614. (e) M. C. M. Sa, A. Kascheres, J. Org. Chem. 1996, 61,
3749-3752. (f) S. Favreau, L. Lizzani-Cuvelier, M. Loiseau, E. Dunach,
R. Fellous, Tetrahedron Lett. 2000, 41, 9787-9790. (g) K. Murai, Y.
Takahara, T. Matsushita, H. Komatsu, H. Fujioka, Org. Lett. 2010, 12,
3456-3459. (h) R. Chakraborty, V. Franz, G. Bez, D. Vasadia, C. Popuri,
C.-G. Zhao, Org. Lett. 2005, 7, 4145-4148.
[21] (a) P. Merino, in Science of Synthesis, Vol. 27 (Eds.: D. Bellus, A.
Padwa), George Thieme, Stuttgart, 2004, pp. 511-580. (b) P. Merino, in
Sci. Synth., Knowl. Updates, Vol. 2010/4 (Ed.: E. Schaumann), George
Thieme, Stuttgart, 2011, pp. 325-403.
[22] Any other combination of nitrone and aldehyde gave no reaction,
proving that both the aldehyde (for the first step) and the nitrone (for the
second step) must be activated by the presence of electronwithdrawing
groups .
[23] In addition to conjugation as a driving force it should be note that 2-
oxazolines are more stable than 3-oxazolines
[17] (a) H. Zhou, X. Zeng, L. Ding, Y. Xie, G. Zhong, Org. Lett. 2015, 17,
2385-2387. (b) H. Zhou, X. Zeng, Y. Xie, G. Zhong, Synlett 2015, 26,
1693-1696.

Chemistry - A European Journal
10.1002/chem.201602159
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
The reaction between nitrone ylides
and aldehydes in the presence of
catalytic amounts of n-BuLi provides
3-oxazolines with complete trans-
selectivity. DFT calculations allow
rationalizing the role of self-generated
water for the progress of the catalytic
cycle
V. Juste-Navarro, I. Delso, T. Tejero and
P. Merino*
Page No. – Page No.
Azomethine Ylides from Nitrones:
Using Catalytic n-BuLi for the Totally
Stereoselective Synthesis of trans-2-
Alkyl-3-Oxazolines