A Unique (3+2) Annulation Reaction between Meldrum's Acid and Nitrones: Mechanistic Insight by ESI-IMS-MS and DFT Studies

Article abstract of DOI:10.1002/chem.201705714

The fragile intermediates of the domino process leading to an isoxazolidin-5-one, triggered by unique reactivity between Meldrum's acid and an N-benzyl nitrone in the presence of a Br?nsted base, were determined thanks to the softness and accuracy of electrospray ionization mass spectrometry coupled to ion mobility spectrometry (ESI-IMS-MS). The combined DFT study shed light on the overall organocatalytic sequence that starts with a stepwise (3+2) annulation reaction that is followed by a decarboxylative protonation sequence encompassing a stereoselective pathway issue.

Full text of DOI:10.1002/chem.201705714

A Journal of
Accepted Article
Title: A Unique (3+2) Annulation Reaction between Meldrum's Acid and
Nitrones: Mechanistic Insight by ESI-IMS-MS and DFT Studies
Authors: Nicolas Lespes, Etienne Pair, Clisy Maganga, Marie Bretier,
Vincent Tognetti, Laurent Joubert, Vincent Levacher, Marie
Hubert-Roux, Carlos Afonso, Corinne Loutelier-Bourhis, and
Jean-François Brière
This manuscript has been accepted after peer review and appears as an
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 published
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.201705714
Link to VoR: http://dx.doi.org/10.1002/chem.201705714
Supported by

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
A Unique (3+2) Annulation Reaction between Meldrum’s Acid and
Nitrones: Mechanistic Insight by ESI-IMS-MS and DFT Studies
Nicolas Lespes,[a] Etienne Pair,[a] Clisy Maganga,[a] Marie Bretier,[a] Vincent Tognetti,*[a] Laurent
Joubert,[a]
Vincent
Levacher,[a]
Marie
Hubert-Roux,[a]
Carlos
Afonso,[a]
Corinne Loutelier-Bourhis,*[a] Jean-François Brière*[a]
Dedication ((optional))
Abstract: The fragile intermediates of the domino process leading to
the overall organocatalytic sequence; a requisite endeavor to
broaden the scope of this useful organocatalytic process and
isoxazolidin-5-one, triggered by the unique reactivity between
Meldrum’s acid and N-Bn nitrone in the presence of a Brønsted base, tackle, eventually, an efficacious asymmetric version.[3a]
were determined thanks to the softness and accuracy of
electrospray ionization mass spectrometry coupled to ion mobility
spectrometry (ESI-IMS-MS). The combined DFT study shed light on
the overall organocatalytic sequence, which starts by a stepwise
(3+2) annulation reaction followed by a decarboxylative protonation
sequence encompassing a stereoselective pathway issue.
Introduction
1,3-dipolar cycloaddition reactions have become an essential
part of organic chemists’ portfolio for the construction of
ubiquitous five-membered ring heterocycles.[1] Since the seminal
investigations of Huisgen and colleagues,[2] a great deal of
research has been carried out, not only to extend the scope of
these useful annulation processes, but also to ascertain the
exact mechanism in action based on either concerted or non-
concerted events. We recently reported a unique dipolarophile-
like behavior of Meldrum’s acid 1 in the presence of nitrone
dipoles 2 allowing a straightforward synthesis of isoxazolidin-5-
ones 3 (Scheme 1),[3] which are useful building blocks for the
elaboration of medicinally relevant β-amino acids or nucleoside
mimics.[4] This novel methodology has highlighted (1) the
practical use of Meldrum’s acid derivatives 1 as a ketene
equivalent or C2 synthon;[5] (2) a facile addition reaction of rather
acidic Meldrum’s acid derivatives 1 (pKa = 4.93 in water for R1 =
H) to nitrone 2 upon organocatalytic Brønsted base (R3N)
Scheme 1. Organocatalytic Synthesis of Isoxazolidinones and Mechanistic
hypotheses to be addressed.
Electrospray ionization mass spectrometry (ESI-MS) along
with direct infusion of liquid solutions, proved to be a useful tool
that allows the “off-line” monitoring of reactions in almost a real-
time mode. The softness of ESI permits the “fishing” of charged
and labile synthetic intermediates.[7],[8] This is a salient feature to
tackle the determination of heat sensitive species such as the
ones putatively encounter from Meldrum's acid 1 (i.e. 4 and 5).
Recently, the coupling of ion mobility spectrometry (IMS)
technique with MS not only afforded a new dimension for the
separation of closely related structures within challenging
complex mixtures,[9],[10],[11] but provided novel opportunities for
the determination of three-dimensional structures.[12] In IMS, ions
are separated as a function of their charge, shape and size
through their collision cross-section (CCS, Ω). Experimental
CCS values can be determined and compared to those
calculated from theoretical structures. Then, ESI-IMS-MS
coupled with quantum chemical calculations emerges as a
powerful modern strategy to get insights into complex reaction
pathways,[13] although its application to mechanistic elucidation
conditions despite the poor nucleophilic nature of the anion
5c]
intermediate
4
(step I);[6,
and (3) a domino reaction
encompassing likely a formal (3+2) annulation-fragmentation-
decarboxylation-protonation sequence towards the formation of
product 3 (steps I-III). Nevertheless, the veracity of these
assumptions still has to be addressed in order to shed light on
[a]
Dr. N. Lespes,+ Dr. E. Pair,+ C. Maganga, M. Bretier, Dr. V. Tognetti,
Prof. L. Joubert, Dr. V. Levacher, Dr. M. Hubert-Roux, Prof. C.
Afonso, Dr. C. Loutelier-Bourhis, Dr. J.-F. Brière
Normandie Univ, INSA Rouen, UNIROUEN, CNRS, COBRA, 76000
Rouen, France
endeavors
of
organocatalytic
processes
remains
underexploited.[7],[10],[14]
E-mail: corinne.loutelier@univ-rouen.fr
jean-francois.briere@insa-rouen.fr
We are pleased to report a mechanistic investigation of the
(3+2) annulation reaction between Meldrum’s acid 1a (R1 = H)
and a N-Bn nitrone 2a (R2 = Bn and R3 = Ph) allowing an
unprecedented insight into this useful albeit complex
organocatalyzed domino process thanks to the powerfulness of
vincent.tognetti@univ-rouen.fr
These authors contributed equally to this work.
[+]
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
the combined real-time reaction monitoring by ESI-IMS-MS and
DFT computational study.
species with (1) an important role of the cation likely as
hydrogen bonding donor species and (2) no correlation to the
nucleophilic character of the 1,3-dienone type nucleophile.
At the onset of the mechanistic investigation, we attempted
to follow the reaction in the presence of Hünig base (20 mol%) in
an NMR tube with toluene-d8 as solvent. It turned out that no
intermediate was detected at 40°C demonstrating a rapid
process in action. Nevertheless, a major intermediate was
observed at 20°C with a relative proportion of 10%, with respect
to nitrone 2a and product 3a (see supporting information).
Despite a much slower process at 20°C the isoxazolidinone 3a
accumulated with up to 78% NMR yield after 24 hours. Then, the
reaction conditions for the ESI-MS investigation were set.
Results and Discussion
Preliminary results
The specific reactivity of Meldrum’s acid 1a with N-benzyl
nitrone dipole 2a is depicted in Table 1. In the presence of 20
mol% of Hünig base at 40 °C in toluene, a conversion of 68%
1
into isoxazolidinone 3a was measured by H NMR after 2 hours
of reaction (entry 1).[3a] Actually, the reaction was completed
within 18 hours to give product 3a in 86% isolated yield (entry 2).
The reaction can proceed in other solvents but the faster rate
was observed in toluene (entry 2).[3a] On the contrary, hardly no
formation of product 3a was observed in the absence of
Brønsted base promoter after 18 hours (entry 3). However, 48%
of nitrone 2a was consumed as estimated by 1H NMR of the
crude mixture (see supporting information).
Mechanistic investigation by ESI-MS/MS
The first step was the analysis of individual chemical species,
namely reagents 1a, iPr2EtN and 2a together with the final
product 3a, by ESI-MS (Scheme 2a-b).
Table 1. Preliminary Investigations.[a]
Entry
Precursor
Catalyst (equiv), time
iPr2EtN (0.2), 2 h
iPr2EtN (0.2), 18 h
(0), 18 h
3a (%)[b]
1
2
3
4
5
6
1a
1a
1a
6
68 (64)[c]
(86)[c] (35)[d]
(-)
0
iPr2EtN (0.2), 2 h
TBD (0.2), 2 h
6
0
1a
nBu4NBr (0.2), K2CO3 (1), 2 h
32[e]
[a] Meldrum’s acid 1a or diethylmalonate 6 (1.1 equiv), N-benzylnitrone 2a
(1 equiv), Catalyst (0.2 equiv), toluene (0.1 M), or 18 hours. [b]
2
Conversion determined by 1H NMR of the crude product with respect to the
remaining nitrone 2a. [c] Isolated yield after silica gel column
chromatography. [d] Performed in acetonitrile as solvent. [e] 20% of
conversion without nBu4NBr. TBD: triazabicyclo[4.4.0]dec-5-ene.
Scheme 2. Proposed structures of ions detected by ESI-MS.
The Hünig base gave an abundant signal corresponding to
the ammonium ion 7" instead of ammonium ion 7 signal while
protonated nitrone 8 and isoxazolidinone 9 were low-abundant
species in the positive ion mode. When the final product was
analyzed in the negative ion mode, the deprotonated
isoxazolidinone 3a (ion 12 in Scheme 2c) was not detected,
even in the presence of the Hünig base (20 mol%). The
deprotonated Meldrum’s acid 4a could be detected in the
negative ion mode although some precautions were required.
Indeed, the detection of the resulting enolate (m/z 143)
Despite the help of Hünig base (entry 4) or the strongly basic
TBD guanidine (entry 5), the more nucleophilic anion of
diethylmalonate 6 did not react with nitrone 2a and starting
materials were recovered. Interestingly, the reaction turned out
to be sluggish and not clean upon Phase Transfer Catalysis
(PTC) conditions, which may involve a quaternary ammonium
(R4N+) ion-pair with the anion 4a of Meldrum’s acid 1a (entry 6).
Obviously, a tertiary ammonium salt (R3N+-H) flanked to the
Meldrum’s acid anion 4a might be the catalytically effective
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
necessitates soft source conditions in order to prevent any
fragmentation (see supporting information). Accordingly, the
increase of either source cone voltage or source temperature
induced the known enolate fragmentation with acetone neutral
loss leading to m/z 85, the acyl-ketene species 10. This is a
well-known reactivity pathway of Meldrum’s acid 1a which has to
be taken into account in our investigation.[15]
intermediates 11 (m/z 354), 5a (m/z 296) and 12 (m/z 252)
appeared as early as 10 minutes, increased with time in a first
stage and then decreased. This evolution was confirmed by
replicate and a semi-quantitative monitoring of the reaction by
infusion of the reaction media at various time intervals in a flow-
injection approach using heptadecanoic acid as an internal
standard (Figure 2). The relative abundances of 12 and
especially carboxylate 5a followed a Gaussian curve. Meanwhile,
the enolate 4a, the first intermediate in our mechanistic proposal,
steadily decreased. The ionic intermediate 11, hypothesized as
one out of three possible isomers 11A-C at that stage, was a
low-abundance species along the reaction but was clearly
identified by high-resolution mass spectrometry (vide infra).
Thanks to the softness of ESI-MS no fragmentation of 4a into
the acyl ketene product 10 was detected ruling out a ketene
based mechanism.[16]
To pursue the structural identification of 12, 11 and 5a
intermediates, tandem mass spectrometry (MS/MS) experiments
using collision induced dissociation (CID) activation process
were carried out (Figure 3). The product ions observed were
comfortingly consistent with the supposed structures. Indeed,
the ion at m/z 252 (12) dissociates into ions either at m/z 208
(14) or m/z 147 (13) which can respectively be explained by
elimination of CO2 or C7H7N (Figure 3A). While (11) can
eliminate one CO2 molecule, m/z 296 dissociates with
successive elimination of two CO2 molecules (leading to m/z 252
and m/z 208, Figure 3B) which is possible for structure 5a. Then,
m/z 354 (11) and m/z 296 (5a) first dissociate into m/z 252 and
then into some common product ions, m/z 208 (14) and m/z 147
(13), which were previously stated for m/z 252 (11). Common
fragmentation patterns indicate that both m/z 252 product ions
observed in Figure 3B and 3C correspond to structure (12).
Then, the dissociation of m/z 354 (11) into m/z 252 involves
elimination of both CO2 and C3H6O in a concerted manner
(Figure 3C) which is possible when fragmentation of a six-
membered ring via a rearrangement pathway occurs.
Figure 1. Negative ion ESI mass spectra after 10 min (A), 45 min (B) and 90
min (C) of reaction recorded on the QIT instrument. *Contaminant. Inset show
a zoom of m/z 350-357 range
100
4a
11
5a
12
80
60
40
20
0
2.84
252
Ph
100
208
N
147
Ph
O
A
- CO
2
O
208
- C H N
t (ms)
2
3
7
7
147
140
116
100
Ph
234
220
0
60
O
180
260
300
296
340
380
420
m /z
m /z
m /z
100
0
200
400
600
3.01
N
O
Ph
252
Reaction time (min)
B
252
- CO
2
- CO
2
O
O
Figure 2. Evolution with time of ion relative intensity of 4a, 5a, 11 and 12 from
the ESI-QIT analysis.
208
234
116
2
3
4
t (ms)
147
140
- C H N
7
7
91
0
60
x2
100
180
220
260
300
340
380
420
3.56
354
100
Next, the reaction monitoring was carried out in the presence
of Hünig base as catalyst (Scheme 2c, Figure 1). The obtained
diluted reaction mixtures were thus analyzed at various time
intervals (10 minutes, 45 minutes, 1h30, 2h15, 3h45, 6h and
21h) in both negative and positive ion modes. The positive ESI
Ph
Ph
C
N
O
O
- C H O
3 6
O
- CO
2
O
252
- C H N
7
7
252
O
3
4
t (ms)
208
220
147
140
0
60
100
180
260
300
340
380
420
mass spectra only showed the ammonium ion iPr2EtNH+
7
together with the protonated nitrone 8 and isoxazolidinone 9
whereby their abundances increased along the reaction time
(data not shown). In the negative ion mode, the anionic
Figure 3. Negative ion ESI-MS/MS spectra of 12 (A), 5a (B) and 11 (C)
precursor ions recorded with the Q-IMS-TOF instrument. The right insets show
the ion mobility spectra of the corresponding precursor ions while the left
insets display their fragmentation patterns.
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
It is remarkable that transient and likely enolate species 12
(m/z 252) was detected in the reaction media given the expected
fast protonation event of this basic last intermediate of the
domino sequence (vide infra). At that stage, we cannot rule out
that part of the m/z 252 (12) is originated from some
fragmentation of m/z 296, namely carboxylate 5a, inside the
mass spectrometer.
Eventually, the elemental composition of the detected
species from MS and MS/MS experiments was determined by
accurate mass measurements using the Q-IMS-TOF mass
spectrometer and, pleasingly, matched with the supposed
intermediates 4a, 5a, 11 and 12 (Table 2).
kcal/mol). Interestingly, this ring-opening reaction leads first to
an alcoholate intermediate Int3 showing that the formal
fragmentation event occurs in two steps. Accordingly, an
acetone molecule can be exothermically eliminated from Int3 by
the relevant C-O bond breaking at a low energetic cost (6.4
kcal/mol), leading to the relatively stable carboxylate
intermediate Int4. The activation barrier TS4 (26.0 kcal/mol) for
the subsequent decarboxylation reaction providing intermediate
Int5 is markedly higher than those for the nucleophilic addition
and cyclization steps (Int1 to Int2 to Int3).[17] Finally, the rather
unstable enolate Int5 (or 12 in MS) rapidly evolves towards the
much more stable isoxazolidin-5-one product 3a, which is
obtained by the C-protonation and regeneration of the amine-
catalyst, as described in detail in the supplementary information.
These computational results are consistent with the MS study,
which showed the accumulation of the carboxylate intermediate
5a in situ (Figure 2). At first glance, the close energy between
TS1 and TS2, namely the addition and cyclisation steps of the
(3+2) annulation sequence, may account for the modest
enantiomeric excesses previously obtained on this system thus
far, due to some equilibration events of the first enantio-relevant
addition step.[3a]
Table 2. Accurate mass measurements, elemental composition of ions.
Ion
Experimental
Calculated
M
Elemental
ions
(m/z)
(m/z)
(m/z)
(ppm)
composition
143
252
296
143.0341
252.1021
296.0930
143.0349
252.1030
296.0928
4.9
3.6
C6H7O4
4a
12
5a
C16H14NO2
C17H14NO4
-0.7
354
147
208
354.1332
147.0444
208.1126
354.1347
147.0451
208.1132
4.2
4.8
2.9
C20H20NO5
C9H7O2
11
13
14
However, one should not forget that mass spectrometry only
analyzes charged species in gas phase (after desolvation of the
ions by ESI from the reaction media diluted in acetonitrile),
whereas the previous reaction pathway was studied in the
C15H14
N
presence of
a
protonated trimethylamine counterion in
Computational study
acetonitrile solution, to be close to the experimental conditions of
the organic synthesis. While, the release of acetone and
decarboxylation were found to be almost unaffected by the
presence of the counteranion (see SI for an exhaustive
description), the addition and cyclization steps appeared to be
strongly influenced by the presence of the protonated
trimethylamine either in gas phase or solvent as shown in
Scheme 4. Indeed, as revealed by intrinsic reaction coordinate
(IRC) calculations in the absence of Me3N+H, the initial two
stages process (addition and cyclization sequence – previously
Int1 to Int3 with Me3N) occur in the same step leading directly to
Int7 (or Int9 in gas phase), through a concerted (3+2)-
cycloaddition.[18] Accordingly, the ion of m/z 354 seen by ESI-MS
likely refers to cyclic-structure 11C (see Scheme 2). Besides, the
associated activation barrier is considerably higher (more than
twice than that for the catalyzed version in solution), proving the
fundamental role of the R3N species to facilitate and influence
the first steps of this domino process.
In line with the ESI-IMS-MS study, density functional theory
(DFT) calculations were performed in order to investigate the
possible reaction pathways and intermediates for the synthesis
of isoxazolidin-5-one 3a (Scheme 3). The first step consists in
the formation of the noncovalent adduct Int1 between the
enolate 4a and the nitrone 2a, which displays a hydrogen-
bonding interaction with the protonated trimethylamine. This
interaction likely enhances the electrophilic character of nitrone
2a and promotes the subsequent nucleophilic addition reaction
of Meldrum acid enolate 4a. The formation of the C-C bond
takes place through transition state TS1 and the protonation of
nitrone 2a, which occur at the same single step (with an overall
activation barrier equal to 14.6 kcal/mol), affording intermediate
Int2. This compound then undergoes a cyclization by the
formation of a C-O bond (see dashed line in TS2) associated to
a second proton transfer from the hydroxylamine moiety to
trimethylamine, along with a moderate activation barrier (11.9
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
Scheme 3. Standard Gibbs energy profile for the formation of isoxazolidin-5-one 3a in acetonitrile through the cis-carboxylate intermediate 5a-cis (Oxygen atoms
in red, nitrogen in blue, carbon in grey, hydrogen in white).
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
While collision cross-section (CCS) values can be obtained
from DFT optimized structures, used as input in the open source
software program MOBCAL,[19],[20] the experimental CCS can be
determined from the ESI-IMS-MS data (Table 3). The
comparison and the fitting of experimental and calculated CCS
values should permit to validate the theoretically determined
structures. Among the three different models used to calculate
CCS with MOBCAL, the trajectory method (TM), which is
considered as the most reliable approach for small organic
molecules was used to calculate the CCS values of the
intermediates. Meanwhile, the experimental CCS values were
determined using the measured drift time of the ions in
optimized IMS conditions and the calibration method described
in the literature (in He) using dextran oligomer ions as reference
ions to correct the non-uniform electric field of the TWIM
cell.[21],[22] The CCS values of the intermediates 11C, 5a and 12
are reported in Table 3. Pleasingly, these experimental CCS
values match with the calculated CCS obtained from the DFT
studies which turned out to be the structures of intermediates
5a-cis, 11C and 12 (Scheme 2 versus Scheme 3).
Scheme 4. Standard Gibbs energy profile for the whole cycloaddition process
in various environments. (a) In acetonitrile without Me3N (3D structure
displayed is similar to intermediate 11C). (b) In gas phase process without
Me3N. (c) In acetonitrile with Me3N as a base catalyst. (Oxygen atoms in red,
nitrogen in blue, carbon in grey, hydrogen in white).
Table 3. Experimental TWCCSHe and calculated CCS of 11C, 5a and 12
11C
5a-cis
12
TWCCSHe (Å2)
119 ± 2
119 ± 6
107 ± 2
107 ± 6
101 ± 2
99 ± 6
One can also investigate stereochemistry issues by the
same theoretical protocol (Scheme 5 versus Scheme 3).
Subsequent to the first addition reaction step, the cyclization
event may occur on one of the two diastereotopic carbonyl
functional groups of the 2,2-dimethyl-1,3-dioxan-4,6-dione
moiety leading to the formation of either cis (Int3) or trans-
intermediates (Int12). To this aim, the two competing pathways
are compared in Scheme 5. Noteworthy, it was found that the
two intermediates, Int2 (cis pathway) and Int11 (trans pathway),
are energetically close, and can be easily interconverted (the
activation barrier being equal to 6 kcal/mol) and are thus in
equilibrium. This is no longer the case at the following step,
since the Gibbs energies for Int3 and Int12 significantly differ.
The low stabilization of this last compound strengthens the
reversibility of the trans path (reverse activation barrier equal to
4.5 kcal/mol). One can notice that intermediate Int3 (cis
pathway) resulted from an early ring opening process (C-O bond
breaking), on the contrary to the cyclic quaternary alcoholate
Int12 obtained in the trans pathway (Scheme 5). That might
account why Int12 is more prone to reverse back to the Int11,
en route to the cis path. Importantly, a major intermediate was
seen during a preliminary investigation of this reaction by 1H
NMR reaction which was supposed to be carboxylate 5a with
regard to the ESI-MS kinetic analysis (see Figure 3). The
coupling constant of 12.2 Hz (doublet) is related to a cis junction
of the isoxazolidinone ring of 5a-cis which is consistent with the
preferred cis-pathway (see Scheme 3).
Calculated CCS (Å2)
Conclusions
This work highlights the usefulness of the combined ESI-
IMS-MS and DFT investigation in order to get insights into
complex organocatalytic domino processes involving rather
labile intermediates. Upon the direct infusion of a solution of
Meldrum’s acid 4a, nitrone 2a and a Brønsted base, an original
sequence leading to isoxazolidinone architecture 3a was probed
for the first time. The soft ESI-MS techniques allowed the “off-
line” detection of negatively charged intermediates and IMS-MS
permitted the comparison between experimental/theoretical
collision cross-section values. With regard to the ESI-MS
outcome, the DFT study revealed a sequence triggered by a
stepwise (3+2) annulation reaction between Meldrum’s acid 4a
and nitrone 2a favored by the Brønsted base organocatalyst
(R3N) upon hydrogen bonding. The cyclisation steps turned out
to occur along a stereoselective pathway giving rise to a cis-
carboxylate transient species which undergoes a base-promoted
decarboxylation (highest energy transition step) followed by a
facile protonation sequence. The similarity between the energies
of the addition and cyclization steps might explain the low
selectivity obtained for the enantioselective version thus far.
Based on this knowledge, barely addressed in the chemistry of
Meldrum’s acid,[13] the scope of this unique cycloaddition-based
reactivity with nitrone dipoles is currently under investigation to
develop a more broadly applicable methodology and tackle an
enantioselective version.
Mechanistic probe by ESI-IMS-MS and DFT
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
Scheme 5. Standard Gibbs energy profile for the formation of isoxazolidin-5-one 3a either (a) through the cis-carboxylate intermediate 5a-cis or (d) through trans-
carboxylate intermediate 5a-trans. Only transition states of the trans pathway are depicted. (Oxygen atoms in red, nitrogen in blue, carbon in grey, hydrogen in
white).
intermediates could be detected at soft source conditions (from 30°C to
50°C desolvation gas temperature) with the ESI source of the QIT
instrument to avoid fragmentation of the most fragile species. Some in-
Experimental Section
source fragmentation was observed with the Q-TOF for the fragile
intermediate ion 4a (see SI).
The reaction monitoring was performed by infusing reaction media at
different time intervals into the ESI source of the mass spectrometer.
The TWIM cell is operating with a non-uniform electric field that prevents
direct CCS determination. To estimate TWIM CCS, a calibration with
To cross-validate and ensure our results, with regard to the heat
reference compounds of known CCS is required.[21] In negative ion
sensitivity of transient species, two mass spectrometers were used; (1) a
mode, various CCS calibrants have been recently described including
quadrupole ion trap (HCT Ultra ETD II, Bruker) and (2) a quadrupole-ion
dextran or fatty acid ions whose CCS have been determined in He or N2
mobility-time of flight (Synapt G2 HDMS, Waters), both equipped with
drift gases.[22],[21d] The extracted ion mobility spectra were fitted using
ESI source. The QIT instrument is equipped with a LC pump-injector
Origin Pro 9.1 software (OriginLab, Northampton, MA, USA).
system allowing flow-injection analysis (FIA) and was used for semi-
quantitative monitoring. The Q-IMS-TOF is equipped with a travelling
wave ion mobility (TWIM) cell, previously described,[23] and was used for
Computational details
accurate mass measurements, MS/MS and IMS-MS experiments.
MS/MS experiments involved selection of the precursor ion with the
Gaussian 09 program in order to gain insight into the structures of gas
quadrupole mass analyzer and CID in the transfer cell using argon as
collision gas. Sample solutions were infused into the ESI (Q-IMS-TOF)
source using a Cole-Palmer syringe pump. Note that ions of the studied
Density Functional Theory (DFT) calculations were carried out with
phase ions and into the reaction mechanism in a solvent medium.[24] The
(dispersion-corrected range-separated hybrid) ωB97X-D exchange-
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
correlation functional was used in conjunction with the 6-311++G(d,p)
triple-ζ basis set for all atoms.[25] Solvent effects were implicitly described
by the latest implementation of the IEF polarizable continuum model
(IEF-PCM).[26] Vibrational analyses were carried out to determine the
nature of the stationary points, and additional intrinsic reaction coordinate
(IRC) calculations were run to confirm the connection between reactants
and products.[27] The MOBCAL software was used in complement to
generate CCS values from these DFT optimized structures using default
Lennard-Jones parameters and computed APT atomic charges.[19],[20]
[12] V. D'Atri, M. Porrini, F. Rosu, V. Gabelica, J. Mass Spectrum. 2015, 50,
711.
[13] Few DFT studies have addressed the understanding of the Meldrum's
acid chemical transformation especially the fragmentation event; see
reprensative examples: a) S. Perreault, M. Poirier, P. Léveillé, O. René,
P. Joly, Y. Dory, C. Spino, J. Org. Chem. 2008, 73, 7457; b)
Representative DFT studies to Meldrum's acid based reactions, see: F.
A. Leibfarth, M. Wolffs, L. M. Campos, K. Delany, N. Treat, M. J. Kade,
B. Moon, C. J. Hawker, Chem. Sci. 2012, 3, 766.
[14] For a representative example of IM-MS/DFT Study getting insight into
organometallic mechanism, see: G.-J. Cheng, P. Chen, T.-Y. Sun, X.
Zhang, J.-Q. Yu, Y.-D. Wu, Chem.-Eur. J. 2015, 21, 11180.
Acknowledgements
[15] For a review on acylketenes from Meldrum's acid, see K. P. Reber, S.
D. Tilley, E. J. Sorensen, Chem. Soc. Rev. 2009, 38, 3022.
This work has been partially supported by INSA Rouen, Rouen
University, CNRS, EFRD and Labex SynOrg (ANR-11-LABX-
0029) and region Normandie (CRUNCh network). This research
was supported by the French National Research Agency (ANR)
as part of the ANR-16-CE07-0011-01 project OMaChem.
[16] L. S. Santos, B. A. DaSilveira Neto, C. S. Consorti, C. H. Pavam, W. P.
Almeida, F. Coelho, J. Dupont, M. N. Eberlin, J. Phys. Org. Chem. 2006,
19, 731.
[17] We also computed the activation barrier for this rate-determining step in
toluene as solvent. It was found to be slightly lower (24.7 kcal/mol), in
agreement with our experimental data.
[18] E. Falkowska, V. Tognetti, L. Joubert, P. Jubault, J.-P. Bouillon, X.
Pannecoucke, RSC Advances 2015, 5, 6864.
Keywords: Meldrum’s acid • organocatalysis • density functional
calculations • mass spectrometry • ion mobility
[19] A. A. Shvartsburg, M. F. Jarrold, Chem. Phys. Lett. 1996, 261, 86.
[20] M. F. Mesleh, J. M. Hunter, A. A. Shvartsburg, G. C. Schatz, M. F.
Jarrold, J. Phys. Chem. 1996, 100, 16082.
[1]
For recent reviews on dipolar cycloadditions, see: a) T. Hashimoto, K.
Maruoka, Chem. Rev. 2015, 115, 5366; b) M. S. Singh, S. Chowdhury,
S. Koley, Tetrahedron 2016, 72, 1603.
[21] a) B. T. Ruotolo, J. L. Benesch, A. M. Sandercock, S. J. Hyung, C. V.
Robinson, Nat. Protoc. 2008, 3, 1139; b) D. Smith, T. Knapman, I.
Campuzano, R. W Malham, J. Berryman, S. E Radford, A. Ashcroft,
Eur J Mass Spectrom (Chichester, Eng) 2009, 15, 113; c) M. F. Bush, I.
D. Campuzano, C. V. Robinson, Anal. Chem. 2012, 84, 7124; d) J. G.
Forsythe, A. S. Petrov, C. A. Walker, S. J. Allen, J. S. Pellissier, M. F.
Bush, N. Hud, F. Fernandez, Analyst 2015, 6853.
[2]
[3]
a) R. Huisgen, Angew. Chem. Int. Ed. 1963, 2, 565; b) R. Huisgen,
Angew. Chem. Int. Ed. 1963, 2, 633.
a) S. Postikova, T. Tite, V. Levacher, J.-F. Brière, Adv. Synth. Catal.
2013, 355, 2513; b) T. Tite, M. Sabbah, V. Levacher, J.-F. Brière,
Chem. Commun. 2013, 49, 11569.
[22] a) J. Hofmann, W. B. Struwe, C. A. Scarff, J. H. Scrivens, D. J. Harvey,
K. Pagel, Anal. Chem. 2014, 86, 10789; b) F. Zhang, S. Guo, M. Zhang,
Z. Zhang, Y. Guo, J. Mass Spectrum. 2015, 50, 906.
[4]
[5]
For reviews on isoxazolidinones and derivatives, see: a) M. Berthet, T.
Cheviet, G. Dujardin, I. Parrot, J. Martinez, Chem. Rev. 2016, 116,
15235; b) J. Annibaletto, S. Oudeyer, V. Levacher, J.-F. Brière,
Synthesis 2017, 49, 2117.
[23] a) K. Giles, S. D. Pringle, K. R. Worthington, D. Little, J. L. Wildgoose,
R. H. Bateman, Rapid Commun. Mass Spectrom. 2004, 18, 2401; b) K.
Giles, J. P. Williams, I. Campuzano, Rapid Commun. Mass Spectrom.
2011, 25, 1559.
For review on Meldrum's acid, see: a) A. S. Ivanov, Chem. Soc. Rev.
2008, 37, 789; b) A. M. Dumas, E. Fillion, Acc. Chem. Res. 2010, 43,
440; c) E. Pair, T. Cadart, V. Levacher, J.-F. Brière, ChemCatChem
2016, 8, 1882.
[24] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H.
Nakatsuji, M. Caricato, X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G.
Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T.
Vreven, J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro, M.J. Bearpark, J.
Heyd, E.N. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A.P. Rendell, J.C. Burant, S.S. Iyengar, J.
Tomasi, M. Cossi, N. Rega, N.J. Millam, M. Klene, J.E. Knox, 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, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P.
Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö. Farkas, J.B.
Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox. Gaussian 09 rev. D01,
Gaussian, Inc., Wallingford, CT, USA (2009).
[6]
[7]
Meldrum's acid is 106 less reactive than the corresponding ester of
malonate: O. Kaumanns, H. Mayr, J. Org. Chem. 2008, 73, 2738.
a) S. R. Wilson, J. Perez, A. Pasternak, J. Am. Chem. Soc. 1993, 115,
1994; b) L. S. Santos, Eur. J. Org. Chem. 2008, 235; c) K. L. Vikse, J. S.
McIndoe, Pure Appl. Chem. 2015, 87.
[8]
[9]
L. S. Santos, C. H. Pavam, W. P. Almeida, F. Coelho, M. N. Eberlin,
Angew. Chem. Int. Ed. 2004, 43, 4330; Angew. Chem. 2004, 116, 4430.
K. M. Roscioli, X. Zhang, S. X. Li, G. H. Goetz, G. Cheng, Z. Zhang, W.
F. Siems, H. H. Hill, Int. J. Mass Spectrom. 2013, 336, 27.
[10] For a recent investigation of an organocatalytic reaction by ESI-IM-MS,
see: J.-T. Zhang, H.-Y. Wang, X. Zhang, F. Zhang, Y.-L. Guo, Catal.
Sci. Technol. 2016, 6, 6637.
[11] a) A. Revesz, D. Schroder, T. A. Rokob, M. Havlik, B. Dolensky, Angew.
Chem. Int. Ed. 2011, 50, 2401; Angew. Chem. 2011, 123, 2449; b) V.
Domalain, V. Tognetti, M. Hubert-Roux, C. M. Lange, L. Joubert, J.
Baudoux, J. Rouden, C. Afonso, J. Am. Soc. Mass Spectrom. 2013, 24,
1437; c) . C. Lapthorn, F. Pullen, B. Z. Chowdhry, Mass Spectrom. Rev.
2013, 32, 43.
[25] J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys. 2008, 10, 6615.
[26] G. Scalmani, M. J. Frisch, J. Phys. Chem. 2010, 132, 114110.
[27] K. Fukui, Acc. Chem. Res. 1981, 14, 363.
This article is protected by copyright. All rights reserved.

10.1002/chem.201705714
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 2:
FULL PAPER
Nicolas Lespes, Etienne Pair] Clisy
Maganga, Marie Bretier, Vincent
Tognetti,* Laurent Joubert, Vincent
Levacher, Marie Hubert-Roux, Carlos
Afonso, Corinne Loutelier-Bourhis,*
Jean-François Brière*
Page No. – Page No.
The fragile intermediates of the domino process leading to isoxazolidin-5-one,
triggered by the unique reactivity between Meldrum’s acid and N-Bn nitrone in the
presence of a Brønsted base, were determined thanks to the off-line ESI-IMS-MS
technic. The combined DFT study shed light on the overall organocatalytic pathway
which starts by a stepwise (3+2) annulation reaction followed by a decarboxylative
protonation sequence
A Unique (3+2) Annulation Reaction
between Meldrum’s Acid and
Nitrones: Mechanistic Insight by ESI-
IMS-MS and DFT Studies
This article is protected by copyright. All rights reserved.