Mechanistic insight into the acid-catalyzed isomerization of biomass-derived polysubstituted pyrrolidines: an experimental and DFT study

Article abstract of DOI:10.1039/c6ob02457b

The 1,3-dipolar cycloaddition reactions of azomethine ylides is one of the preferred methods for the synthesis of polysubstituted pyrrolidines. The use of chiral dipolarophiles derived from carbohydrates yields enantiomerically pure pyrrolidines, usually in good to excellent endo selectivities, along with other minor stereoisomers. Recently, we found an unusual isomerization event that allowed the isolation of useful pyrrolidines with relative stereochemistries difficult to obtain otherwise. Although a simple and efficient protocol to promote these transformations was developed, the mechanism was not fully unravelled. Herein, after a combination of experimental, spectroscopic and computational studies (using DFT methods) we propose that this isomerization event takes place through a retro-Mannich//Mannich cascade, via the formation of an iminium ion with E geometry.

Full text of DOI:10.1039/c6ob02457b

View Article Online
View Journal
Organic &
Biomolecular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: G. Gerosa, N.
Grimblat, R. A. Spanevello, A. G. Suarez and A. M. Sarotti, Org. Biomol. Chem., 2016, DOI:
10.1039/C6OB02457B.
This is an Accepted Manuscript, which has been through the
Volume 14 Number
1 7 January 2016 Pages 1–372
Royal Society of Chemistry peer review process and has been
accepted for publication.
Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-0520
COMMUNICATION
Takeharu Haino et al.
Solvent-induced emission of organogels based on tris(phenylisoxazolyl)
benzene
rsc.li/obc

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 1 of 9
View Article Online
DOI: 10.1039/C6OB02457B
Journal Name
ARTICLE
Mechanistic insight into the acid-catalyzed isomerization of
biomass-derived polysubstituted pyrrolidines: an experimental
and DFT study.
Received 00th January 20xx,
Accepted 00th January 20xx
Gabriela G. Gerosa, Nicolás Grimblat, Rolando A. Spanevello , Alejandra G. Suárez, Ariel M.
Sarotti*
DOI: 10.1039/x0xx00000x
www.rsc.org/
The 1,3-dipolar cycloaddition reactions of azomethine ylides is one of the preferred methods for the synthesis of
polysubstituted pyrrolidines. The use of chiral dipolarophiles derived from carbohydrates yields enantiomerically pure
pyrrolidines, usually in good to excellent endo selectivity, along with other minor stereoisomers. Recently, we found an
unusual isomerization event that allowed the isolation of useful pyrrolidines with relative stereochemistries difficult to
obtain otherwise. Although a simple and efficient protocol to promote these transformations was developed, the
mechanism was not fully unraveled. Herein, after a combination of experimental, spectroscopic and computational studies
(using DFT methods) we propose that this isomerization event takes place through a retro-Mannich//Mannich cascade, via
the formation of an iminium ion with E geometry.
ones.7-8 In order to obtain enantiomerically pure pyrrolidines,
this methodology has been extensively studied in its
asymmetric variant by using chiral auxiliaries, ligands and
Introduction
Polysubstituted chiral pyrrolidines are privileged structural
motifs in different branches of organic chemistry, present in a
wide plethora of natural products.1 They are also valuable
scaffolds in medicinal chemistry, since a wide variety of
catalysts.9 Alternatively, the construction of optically active
pyrrolidines have been also achieved starting from chiral
dipolarophiles, being carbohydrates as one of the most
promising

-deficient counterparts.5,6,10 High levels of facial
biologically active compounds share
a five-membered
and regio-selectivity have been observed in the 1,3-dipolar
cycloaddition between sugar-derived enones and azomethine
ylides. Moreover, good to high endo selectivities were
obtained in these reactions, affording the corresponding endo
adduct as the only isolated isomer, or as the major component
in mixtures with other isomeric products that can be obtained
in variable yields, depending on the nature of the dipole,
dipolarophile and the reaction conditions (Scheme 1).5,6,10
nitrogenated heterocyclic ring in their structures.2 For
example, it has been shown that densely substituted prolines
are potent inhibitors of the hepatitis C virus (HCV) polymerase,
emerging as promising agents for the treatment of this
dreadful disease that affects more than 190 million people all
over the world.3 The recent discovery that such molecular
architectures can be also effective chiral organocatalysts
further increased the interest in their synthesis and chemistry
properties.4-6
The 1,3-dipolar cycloaddition reaction between azomethine
ylides and α,-unsaturated carbonyl compounds is a popular
O
1
R2
and efficient strategy for the stereocontrolled synthesis of
pyrrolidines (Scheme 1).7 Among several protocols that have
been developed for this chemical transformation, the use of
stabilized N-metalated azomethine yildes (easily obtained in
situ upon treatment of α-iminoesters with metal salts in basic
media) represents one of the most simple, mild and reliable
R1O
N
Ar
O
O
M
2
H
H
H
R1O2C
R2
R1O2C
R2
R1O2C
R2
O
O
H
H
H
HN
HN
HN
Ar
Ar
Ar
3
4
5
(endo)
(exo)
Scheme
1 Most common diastereoisomers obtained in the 1,3-dipolar
cycloaddition reaction between sugar enones and N-metallated azomethine
ylides.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 2 of 9
ARTICLE
Journal Name
Therefore, unless the endo adduct is the synthetic target, the considered by us in our previous study, aVniewaAdrtdiclietiOonnlinael
DOI: 10.1039/C6OB02457B
application of these protocols can be limited by not providing mechanistic path is possible (Mechanism C), involving an
flexibility to obtain other stereoisomers in high yields.
heterolytic C-N bond cleavage to produce a benzylic
Recently, as part of our ongoing interest in the development of carbocation that could follow a reconnection through the
valuable chemicals from renewable sources,5,6,11 we reported other diastereotopic face to afford the isomerized pyrrolidine.
the synthesis of chiral pyrrolidines derived from This mechanism was found by Cook to account for the above
levoglucosenone in high yields and endo selectivities (Scheme mentioned cis/trans isomerization of substituted tetrahydro--
2).5 Serendipitously, we observed an epimerization at the carbolines (Scheme 4).15
benzylic (C-7) position of an endo adduct (7) after it had been
accidentally left dissolved in CDCl3 for one month inside the
NMR tube. This unexpected and unprecedented isomerization
event was further optimized using microwave irradiation,
Mechanism A: Retro Mannich / Mannich
O
O
H
O
O
O
O
H
H
H
providing
polysubstituted pyrrolidines with relative stereochemistries
difficult to obtain otherwise (
, Scheme 2).5 For instance,
several precedents can be found in the literature regarding the
isolation of stereomutated isomers (such as , Scheme 1), that
a valuable method for the preparation of
R2
R1
R2
R1
R2
H-A
O
OH
O
R1
H
N
N
N
8
Ar
Ar
H
Ar
H
H
5
Mechanism B: Retro aza-Michael / aza-Michael
O
O
O
were generally interpreted on the basis of the isomerization of
the ylide.4,10,12 However, in those studies the aromatic moiety
of the dipole and the electron withdrawing group of the
dipolarophile are arranged in a syn (endo) orientation,
whereas in our case both groups are anti. Interestingly, the
O
O
O
H
H
H
R2
R1
R2
R
R2
R1
O
O
O
H
H-A
Ar
H
N
H
H2N
N
H
A
H
H
Ar
Ar
isomerized pyrrolidines
8 displayed the highest and most
Mechanism C: Carbocationic intermediate
promising organocatalytic activities in asymmetric Diels-Alder
reactions among the evaluated isomers, affording the desired
products in up to 97% ee.6
O
O
O
O
O
O
H
H
H
R2
R1
R2
R1
R2
R1
O
O
O
H
H
N
H
H2N
N
H
CHCl3
AcOH
MW
O
O
H
H
Ar
Ar
Ar
O
O
O
O
2
H
H
MeO2C
R
MeO2C
R
Scheme 3 Proposed mechanisms for the isomerization process.
O
O
88-97%
87-92%
H
7
H
7
O
HN
HN
Ar
Ar
8
6
7
Encouraged by the novelty of this isomerization process, along
with the promising organocatalytic properties of the
isomerized pirrolidines, we undertook combined experimental,
spectroscopic and theoretical studies to unravel the
mechanism of such interesting transformations.
(endo)
Scheme 2 Synthesis of novel pyrrolidines from levoglucosenone
Although we could succeed in developing efficient protocols to
promote the transformation of into , the mechanism of
7
8
such transformation could not be fully understood. Preliminary
experimental observations allowed us to glimpse and sketch
two mechanistic proposals shown in Scheme 3 (A and B). The
first one involves an acid-catalyzed retro-Mannich reaction
leading to a resonance-stabilized iminium ion, followed by a
Mannich-type closure through the Si face of the cation. This
path has been proposed to explain the isomerization reactions
in spiro oxindole alkaloids (Scheme 4),13 and is mechanistically
related with the retro-Pictet-Spengler process,14 considered by
the Cook group to account for the cis/trans epimerization of
N
CO2Et
Me
N
N
Bn
O
O
N
O
N
H
H
Ar
A
B
C
CO2Et
N
Me
N
N
Bn
O
O
N
O
N
H
H
Ar
tetrahydro 
-carbolines.15 Moreover, it was quoted by Varela
tetrahydro--carbolines
(epimerization via
Mechanism C)
oxindole alkaloids
(epimerization via
Mechanism A)
lobelaline
(epimerization via
Mechanism B)
and co-workers to justify the formation of minor isomers in
related systems.10a The second mechanistic proposal
(Mechanism B) consists in
a retro-aza-Michael process
Scheme 4 Three structurally related isomerization reactions that were proposed
affording a reactive enone intermediate that following a
to occur through the different mechanisms depicted in Scheme 3.
nucleophilic attack of the resulting amine through the Si face
of the
-carbon would lead to the isomerized adduct. Le
Bidegau, Joseph, and co-workers suggested this path to explain
the thermodynamic equilibration of lobelanine, norlobelanine
and related analogues (Scheme 4).16 Finally, though was not
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 9
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
ARTICLE
maximum at 1 eq. From there, a considerable reduction in the
epimerization was observed at higher concentrations of TFA
(entries 5 and 6).
Results and discussion
View Article Online
DOI: 10.1039/C6OB02457B
In our previous study, we developed a microwave-assisted
methodology to promote the epimerization rapid and
efficiently. However, such reaction conditions were not ideal
for mechanistic studies due to the short reaction times and the
difficulty to monitor the progress of the isomerization for
kinetic studies. In an effort to develop new experimental
protocol to achieve our goals, we found that gentle heating of
Table 2 Effect of the acid concentration in the isomerization rate.
% Isomerization
Entry
Eq. TFA
0.0
0.1
0.5
1.0
<1%
53%
57%
62%
40%
30%
1
2
3
4
5
6
a chloroform solution of the corresponding endo adduct
a suitable acid at 70ºC afforded the isomerized product
7
and
in
8
2.0
4.0
more suitable reaction times. As in the original procedure,
total conversion (>99%) was achieved after equilibration for all
systems under study, which required 2 to >48 hours depending
on the nature of the starting material and the reaction
conditions. Next, the effect of the acid strength on the
isomerization rate was first analysed using a fixed-time
procedure, with 7a (R=Me, Ar=Ph) as a model compound. After
2 h of heating at 70ºC, the reactions were quenched by the
addition of NEt3, the solvent was evaporated and the crude
1All reactions were carried out in chloroform; [7a] = 80 mM; T = 70 ºC; 2 hs.
We hypothesized that this interesting dependence of the
isomerization rate on the concentration of acid could be
related to the acid-base equilibrium of the system. In order to
have a better understanding of this issue, the effect of TFA in
1
the H chemical shifts of 7a was studied. This approach has
been extensively used to understand the protonation of
related compounds.17 Spectra of 7a in CDCl3 solution were
recorded prior to and after the addition of TFA. Progressive
variations in the chemical shifts were observed upon
increasing the concentration of the acid, and are shown in
Figure 1. With the addition of 0.5 eq. of TFA, the most affected
signals were those assigned to H-7, H-4 and H-3 (in that order),
suggesting that the protonation takes place mostly on the
nitrogen atom, as expected from its higher basicity. The
shifting of the other hydrogen atoms was considerably lower,
consistent from the higher distance from the interaction site.
1
was immediately analysed by H NMR. The different chemical
shifts exhibited by some diagnostic signals in the starting
material (7a) and the isomerized product (8a), allowed the
determination of the isomeric ratio by simple integration of
1
the H NMR spectra of the mixture. Four carboxylic acids with
different pKa values were employed, and the results are
shown in Table 1. Acetic acid (the weaker among the studied
acids), proved to be ineffective at these reaction conditions
since less than 1% of isomerized product was formed after 2 h
of reaction. Significant and systematic increases in the rate
were observed using stronger acids, such as chloroacetic acid
(entry 2), dichloroacetic acid (entry 3), and trifluoroacetic acid
(TFA, entry 4) indicating a clear effect in the acid strength on
the isomerization rate. Considering the good performance
observed with TFA, and given its usefulness in 1H NMR
experiments (vide infra) by not containing additional hydrogen
atoms that could interfere in the analysis of the spectra, this
was the acid of choice for further studies.
6
0.8
0.6
0.4
0.2
0
O
1
O
5
4
H
MeO2C
H3C
3
O
8
H
7
HN
Ph
Table 1 Effect of the acid strength in the isomerization rate.
Entry
Acid
pKa
4.8
2.9
1.3
0.7
% Isomerization
H-1
H-3
H-4
H-5
H-6
H-7
1
2
3
4
CH3CO2H
ClCH2CO2H
Cl2CHCO2H
CF3CO2H
<1%
37%
60%
62%
0.5 eq.
1.0 eq.
2 eq.
4 eq.
Figure 1 Changes in the chemical shifts of selected signals of compound 7a
7a+TFA 7a upon the addition of increasing amounts of TFA.
( =
–
)
1All reactions were carried out in chloroform; [7a] = 80 mM; T = 70ºC; 1
equivalent of acid, 2 hs.
The addition of 1.0 eq. of TFA induced a higher deshielding of
H-7, H-4 and H-3, suggesting a shifting in the equilibrium
towards the protonated species. However, at these conditions
the shifts of other protons away from the nitrogen atom
significantly moved downfield, as was the case of H-1, H-5 and
H-6. Upon the addition of higher amounts of TFA the
deshielding effect of those signals further increased ( up to
0.28 ppm for H-1, 0.16 ppm for H-5 and 0.33 ppm for H-6),
whereas a relatively minor variation in the signals of H-7 and
H-4 was noted. These results could be interpreted assuming
Next, the effect of the concentration of TFA in the
epimerization rate was determined. As indicated in Table 2,
the epimerization was too low in the absence of acid (entry 1),
but significantly increased with the addition of a slight amount
of TFA (entry 2), reinforcing our preliminary observation of the
need of acidic media to promote such transformations. A
modest but systematic increase in the rate was noted using
increasing concentrations of TFA (entries 3 and 4) reaching a
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 4 of 9
ARTICLE
Journal Name
the formation of di-protonated species by additional To gain additional insight into the isomerizationVipewroAcrteicslesO, nwlinee
protonation of the oxygen atoms (mainly those corresponding studied the dependence of the reactionDraOtI:e10o.n10t3h9e/Ce6OleBc0t2ro45n7iBc
to the ketone and 1,6-anhydro groups on the basis of the nature of the aromatic substituent at the C-7 position. Apart
magnitude of the observed  values).
from 7a, we synthesized two isomers bearing a strong electron
According to the data herein presented, a simplified acid-base releasing (7b, Ar=4-OMe-Ph) or withdrawing (7b, Ar=4-NO2-Ph)
equilibrium depicted in Scheme 5 was proposed, involving a groups at the para position of the aromatic ring (Scheme 6).
non-protonated species (7), two mono-protonated species,
with the protonation at the nitrogen (7-NH2) or the carbonyl
oxygen (7-COH), and the di-protonated 7+2H counterpart. The
CDCl3
TFA
60ºC
O
H
O
H
O
O
equilibrium should be shifted towards
7 and 7+2H at low and
H
H
high concentration of TFA, respectively. On the other hand, the
relative proportions of 7-NH2 and its O-protonated tautomer
MeO2C
Me
MeO2C
Me
kobs (s-1) = 5.42x10-5 (7a)
kobs (s-1) = 2.42x10-4 (7b)
kobs (s-1) = 2.89x10-5 (7c)
O
O
HN
7a-c
HN
8a-c
(7-COH) might also depend on their relative energies,
expecting the former to be the most stable form of the mono-
protonated species, prevailing at all ranges of [TFA].
R1
a) R1=H; b) R1=OMe; c) R1=NO2
R1
Scheme 6 Observed rate constants for the isomerizations of 7a-c.
mM; T = 60ºC; [TFA]0 = 80 mM, CDCl3.
[7a-c] = 80
0
O
O
H
R1
R
Interestingly, the isomerization rate (as estimated by the
percentage of isomerization observed after 2 h) was strongly
sensitive to the electronic effect of the aromatic substituent at
C-7. While at 2 h the reaction was completed for compound
7b, 38% of isomerization was observed for 7a, and only 20% in
the case of 7c. To have more detailed kinetic information, we
next studied the isomerization of 7a-c using 1H NMR
spectroscopy to monitor the progress of the reactions. Thus,
pure samples of 7a-c were dissolved in CDCl3, the NMR spectra
was recorded prior and after the addition of 1 eq. of TFA. In all
cases, the final concentration of the corresponding pyrrolidine
and TFA was 80 mM. Since the isomerization rate was too slow
at room temperature, the NMR probe was heated to 60ºC and
maintained at that temperature throughout the experiments,
taking the 1H NMR spectra at regular time intervals. In all
cases, the different chemical shifts in the H11 signals exhibited
O
H
O
H
O
H
N
H
7-NH2
O
O
H
[TFA]
[TFA]
O
Ar
H
H
R1
R
H
R1
R
O
O
O
N
H
N
H
Ar
7+2H
Ar
H
H
H
O
7
R1
R
H
N
7-COH
Ar
H
Scheme 5 Proposed acid-base equilibrium.
Returning to the rate reduction at high [TFA], and considering
the acid-base equilibrium indicated in Scheme 5, we
hypothesized that mechanism A (Scheme 3) better fitted the
experimental observations. Thus, if the retro-Mannich
pathway was involved, a rate reduction would be expected at
high concentration of TFA, as this mechanism would not
proceed with the nitrogen protonated. Despite the di-
protonated and mono-protonated forms are in equilibrium,
the concentration of the reactive mono-protonated
intermediate (7-COH) would diminish as the concentration of
TFA increases. Similar trends were observed by Laus and co-
workers in their mechanistic investigation of the isomerization
of spiro oxindole alkaloids (Scheme 4),13 for which suggested a
retro-Mannich//Mannich cascade. On the other hand,
mechanism C should be inconsistent with the reduction of the
rate upon the addition of more than 1 eq. of TFA, as a first-
order kinetics with respect to the concentration of TFA was
observed for the isomerization of trisubstituted-1,2,3,4-
by the starting material
(7a-c) and the corresponding
isomerized product (8a-c) allowed the determination of the
progress of the reaction by integration of those signals in each
1H NMR spectrum. With the integral values we calculated the
concentration of the starting pyrolidine (7a-c), and plotted
them over time. In all cases, the data nicely fitted to
exponential decay functions (R2 = 0.9991, 0.9971 and 0.9996
for 7a, 7b and 7c, respectively), indicating a pseudo-first-order
kinetics for the isomerizations. Figure 2 shows the plot
obtained for 7b as a representative example, the remaining
are given as part of the SI.
tetrahydro--carbolines operating in a related mechanistic
proposal (Scheme 4). Moreover, we explored the
isomerization of 7a using TFA-d at the same reaction
conditions shown in Table 2, and found a minor kinetic isotopic
effect (55% of isomerization with TFA-d vs 62% using TFA),
suggesting that the protonation should not be the rate-
determining step. In addition, the lack of deuterium
incorporation at C-3 provided strong evidence against the
retro-aza-Michael path (Mechanism B, Scheme 3).
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 9
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
ARTICLE
the +N=CH proton of an iminium ion that could be formed
View Article Online
DOI: 10.1039/C6OB02457B
upon the addition of TFA,19 but then we realized that it was
80
60
40
20
0
the signal of the formyl hydrogen of p-anisaldehyde (9b),
benzaldehyde (9a), and p-nitrobenzaldehyde (9c), respectively,
showing a perfect match with those recorded from pure
analytic samples (9.87 ppm for 9b, 10.00 for 9a and 10.16 ppm
for 9c). The formation of 9a, 9b and 9c from 7a, 7b, and 7c,
respectively, was thought to proceed via the hydrolysis of the
corresponding iminium ion Im from the low water content
present in the system (Scheme 7), representing a key
experimental observation in support of mechanism A.
0
10000
20000
Time (s)
30000
CDCl3
TFA
60 ºC
O
H
O
O
O
CHO
Figure 2 Concentration vs time plot obtained for 7b. [7b
TFA]0 = 80 mM, CDCl3.
]
= 80 mM; T = 60ºC;
0
H
H
+
MeO2C
Me
MeO2C
Me
[
O
O
R1
9a-c
HN
10
NH2
The estimated rate constants (k = 5.42x10-5, 2.42x10-4 and
2.89x10-5 s-1 for 7a
7b and 7c, respectively, Scheme 6) nicely
7a-c
O
O
,
R1
agreed the reactivity trends discussed above. The negative
slope in the Hammet plot strongly suggested the formation of
a carbocationic intermediate in direct resonance with the
substituent at the para position,18 reinforcing our observations
against the aza-Michael mechanism. However, despite all the
collected results better fitted with the retro-Mannich path, we
could not completely rule out the cationic mechanism
proposal.
H
H
MeO2C
Me
H2O
OH
N
Im
R1
a) R1=H; b) R1=OMe; c) R1=NO2
Scheme 7 Hydrolysis of the iminium ion Im, key intermediate in the retro-
Mannich//Mannich path.
Detailed analysis of the 1H NMR data collected during the
kinetic measurements provided further evidence to solve this
crossroad. In an effort to identify the reactive intermediate
that could be formed during the isomerization reactions, we
found a very weak signal at ~10 ppm (in ratios from 1:40 to
If this mechanism was correct, then we should have detected
compound 10 from the reaction mixtures as well. However,
the low amount of hydrolyzed products (as estimated from the
intensity of the aldehyde signals) hampered the search in the
1H NMR spectra of the mixtures of
7 and 8 for the signals that
1:256, relative to the averaged signals of
Figure 3.
7 and 8), as shown in
could be assigned to 10, expected in a much more crowded
region. We foresaw to run the isomerization reactions by
adding increasing amounts of water to facilitate the hydrolysis
reactions, though in all cases we obtained the corresponding
isomerized pyrrolidine 8a-c as the major products, even using
pure water as solvent. Nevertheless, from these experiments
we could isolate a new compound in very low yields (<5%),
whose 1H and 13C NMR data was consistent with those
expected for 10. The presence of impurities that could not be
eliminated by standard chromatographic procedures
complicated the assignment, and this was solved after
acetylation of the mixture, isolating the corresponding
acetamide 11 in sufficiently pure form to be irrefutable
assigned by standard experimental procedures. Thus, the
molecular formula C12H17NO6 was deduced from the HRMS ion
at m/z 294.09578 [M+Na] (calcd. 294.09481). IR absorption
bands at 1738, 1732 and 1661 cm-1 revealed the presence of
three carbonyl groups of ketone, ester and amide,
respectively, that was confirmed by 13C NMR (199.6 ppm,
Figure 3 NMR spectra of 7a after the addition of 1 eq. of TFA, recorded in CDCl3
at 60ºC.
The lack of these signals in the 1H NMR spectra of the
corresponding samples of 7a-c recorded before the addition of
TFA, indicated that the acid should have promoted the
formation of this new compound or intermediate. Moreover,
we found that the chemical shift of these signals moved
downfield when increasing the electron-poor character of the
aromatic moiety (9.88 ppm for 7b, 10.01 ppm for 7a and 10.16
ppm for 7c). Initially, we thought that it could be the signal of
1
172.9 ppm and 170.0 ppm, respectively). Table 3 shows the H
and 13C data collected for compound 11, that were assigned
using COSY, HSQC, HMBC and NOE experiments. The absence
of aromatic signals and those corresponding to C-7/H-7 in the
starting material, the presence of a methylene group at C-3,
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 6 of 9
ARTICLE
Journal Name
along with the fact that 11 was obtained either from 7a
, 7b yield the isomerized product the cyclization must occur
View Article Online
and 7c, represented key observations to support its structural through the Si face of the electrophilDeO. I:T1h0e.1r0e3f9o/rCe6,OBa02C485-7NB
proposal.
single bond rotation should take place to expose the other
diastereotopic face of the iminium ion. In principle, it might be
expected that the pathway involving the iminium with E
geometry (E-path) should be less energetically demanding.
However, there are plenty studies in the recent literature in
which the selective formation of Z-iminium ions have been
postulated.20 Thus, to gain insight into the isomerization event,
Table 1 1H (300 MHz) and 13C (75 MHz) of 11 in CDCl3.
6
O
1
O
5
O
10
H
4
2
3
9
O
O
8
Me
O
we undertook DFT calculations at the
B97XD/6-311+G**
11
HN
11
level of theory, including chloroform as solvent using SMD,21 as
12
13
implemented in Gaussian 09.22 The
B97XD functional was

developed by Head-Gordon and co-workers,23 using a version
of Grimme´s D2 dispersion model, and was successfully
employed in recent theoretical studies.24
Position
H (mult, J, Hz)
C (ppm)
100.8
199.6
32.6
44.3
73.7
68.7
61.5
172.9
52.9
21.0
1
2
3
4
5
6
8
9
10
11
12
13
5.07, s
2.74, dd (17.9, 8.6), 2.62, d (17.9)
3.10, d (8.6)
The paths depicted in Scheme 8 were fully explored for the
three systems under study (7a
, 7b and 7c). For simplicity in the
4.83, bs
presentation, Figure 4 shows the results obtained with 7a
.
4.02-3.97, m
Similar trends were observed for the other two compounds
under study and are given as part of the Supplementary
Material.
According to our calculations, N-protonation is the most stable
form of the pyrrolidine under acidic media, in perfect
agreement with our NMR experiments discussed above.
Regarding the O-protonated pyrrolidine, we found that the
most stable orientation of the NH bond is exo, being 7a-COH-X
3.74, s
1.66, s
170.0
23.5
2.01, s
1.6 kcal/mol lower in energy than 7a-COH-N
.
The
All the experimental evidence discussed above represented
strong evidence in favor of the retro-Mannich//Manich
proposal. However, on closer inspection further questions
arose regarding the path at which the isomerization reactions
take place. As shown in Scheme 8, the configuration at the
nitrogen atom in the starting pyrrolidine dictates the geometry
of the resulting iminium ion following a retro-Mannich ring
opening event. Therefore, the O-protonated pyrrolidine with
the NH bond directed towards the endo face of the molecule
corresponding values computed for 7b and 7c are 1.4 and 1.9
kcal/mol, suggesting that the exo preference is influenced by
the electronic nature of the adjacent aromatic substituent. The
computed exo/endo ratios at 70ºC are 92:8 (7a), 88:12 (7b
)
and 94:6 7c), indicating a clear predominance of the
(
corresponding 7-COH-X species in the equilibrium. In contrast,
the retro-Mannich opening event is much more facilitated
from the endo counterparts 7-COH-N. Thus, the TS that leads
to the E-iminium (7a-TS-E) is 15.6 kcal/mol lower in energy
(7-COH-N) should lead to the E-iminium (7-Im-E), whereas the
than the corresponding 7a-TS-Z
.
Z-iminium (7-Im-Z) might be obtained from the corresponding
7-COH-X pyrrolidine, with the NH bond in exo orientation.
Regardless the initial geometry of the iminium ion, in order to
O
H
O
H
O
O
C8-N
O
O
O
O
bond
H
H
H
H
H
rotation
7-TS-E
8-TS-E
H
H
H
H
R1
R
R1
R
R1
R
R1
R
8
O
O
O
O
(Si,Re)
closure
H
(Re,Re)
opening
N
N
N
N
H
O
H
Ar
H
Ar
Ar
O
H
H
H
O
O
7-COH-N
(NH-endo)
8-COH-X
(NH-exo)
7-Im-E
7-Im-E
Ar
H
H
R1
R
R1
R
O
O
N
H
N
H
H
H
Ar
8-NH2
C8-N
bond
rotation
Ar
7-NH2
O
O
O
O
O
O
O
O
7-TS-Z
8-TS-Z
H
H
H
H
H
H
H
H
R1
R
R1
R
R1
R
R1
R
8
O
O
O
O
(Re,Re)
opening
(Si,Re)
closure
Ar
H
H
N
N
N
N
H
Ar
Ar
Ar
H
H
H
H
7-COH-X
(NH-exo)
8-COH-N
(NH-endo)
7-Im-Z
7-Im-Z
H
Scheme 8 Two different paths for the retro-Mannich//Mannich cascade.
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 7 of 9
View Article Online
DOI: 10.1039/C6OB02457B
Journal Name
ARTICLE
H≠
=

The larger barrier computed for the formation of the Z- destabilizes the corresponding transition structures (

iminium is also well reflected on the relative energies of both 29.8 kcal/mol and 28.2 kcal/mol for 7c-TS-E and 7b-TS-E
geometric isomers. Overall, the formation of 7a-Im-Z is 7.4 respectively, and 33.8 kcal/mol and 31.1 kcal/mol for 8c-TS-E
kcal/mol more endothermic than 7a-Im-E H = 23.6 kcal/mol and 8b-TS-E, respectively). The effect of the electronic nature
,
(
and 16.2 kcal/mol, respectively). Subsequent ring-closing of the substituent at the phenyl ring in the ring-opening and
Mannich-type is also favored in the E-path by 4.5 kcal/mol ring-closing Mannich stages can be also seen in the C3-C7 bond
(
H = 36.5 kcal/mol and 32.0 kcal/mol for 8a-TS-Z and 8a-TS- breaking and forming distances, respectively. In general, the
E, respectively). As a result, our calculations clearly indicate more electron-rich nature of the aromatic ring, the shorter the
that the E-path should predominate, even though the starting C3-C7 distances in the transition structures. For instance, the
pyrrolidine (7-COH-N) is far less stable than the corresponding C3-C7 bond length in the transition structures for the ring-
7-COH-X counterpart.
opening (TS-1-E) are 1.91 Å, 1.86 Å and 1.95 Å for 7a, 7b and
7c, respectively, and 2.00 Å, 1.97 Å and 2.01 Å, respectively,
for the corresponding ring-closing events (TS-2-E).
Moreover, the thermodynamic preference of the isomerized
pyrrolidines experimentally found was also nicely reproduced
by our calculations. Interestingly, the overall exothermicities
computed 7a, 7b and 7c were 2.9 kcal/mol, 3.4 kcal/mol and
2.2 kcal/mol, affording the same relative order found for the
formation of the iminium ions. Thus, according to these
results, the isomerization reactions of electron-rich
compounds
are
benefited
both
kinetic-
and
thermodynamically.
8-TS-E
33.8
7-TS-E
32.0
29.8
31.1
29.0
28.2
Figure 4 Full reaction coordinate diagram of the isomerization of 7a computed at
the
B97XD/6-311+G** level of theory. Relative enthalpies are given in
19.9
kcal/mol. Optimized structures and relevant bond-forming and bond-breaking
7c
distances are also shown.
16.2
13.1
7a
7b
The computational study herein presented also nicely matched
the reactivity trend experimentally found for 7a-c. Figure 5
shows the preferred path (involving the iminium E) for the
three systems under study, with the O-protonated species
omitted for clarity. A good correlation between the energy
barriers and the relative stability of the corresponding iminium
ions was observed. The relative ground-state enthalpies of the
iminium intermediates closely reflected the influence of the
7-Im-E
0.0
-2.2
7-NH2
-2.9
-3.4
8-NH2
Figure 5 Reaction coordinate of the preferred path (E) computed for 7a (in
black), 7b (in grey) and 7c (in red) at the
Relative enthalpies are given in kcal/mol.
B97XD/6-311+G** level of theory.
substituent at the aromatic ring (
kcal/mol and 19.9 kcal/mol computed for 7b

H = 13.1 kcal/mol, 16.2
7a and 7c
Conclusions
,
,
respectively). The effect of the methoxy group to increase
electron density is also responsible for the relative higher
stabilization of the incipient iminium ion, resulting in lower
barriers of activation. Conversely, the strong electron-
withdrawing capacity of the nitro group not only increments
the exothermic character of the resulting iminium ion, but also
Experimental, spectroscopic and computational studies
provided clear evidence for the acid-promoted isomerization
reaction at the benzylic position of chiral pyrrolidines. Three
different mechanistic proposals suggested for related systems
were taken into consideration, and the experimental evidence
indicated that the epimerization takes place though a retro-
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 8 of 9
ARTICLE
Journal Name
Mannich//Mannich cascade. DFT calculations at the time. The reaction was quenched by the addition of NEt3, the
View Article Online

B97XD/6-311+G** level of theory correctly reproduced the solvent was evaporated and the cruDdOeI: 1w0.a1s039im/Cm6OeBd0i2a4t5e7lBy
experimental findings, and provided further insight suggesting analysed by 1H NMR. Further purification by column
both the ring-opening and ring-closure steps should occur chromatography with elution with mixtures of hexanes and
from a key iminium ion with E-configuration.
ethyl acetate furnished pure isomerized compounds 8a-c.
Computational methods
Conformational searches for the reactants, the transition
structures (TSs) and the products were run to locate the global
minima employing the
Experimental Section
Experimental methods

B97XD functional,23 coupled with the
6-311+G(d,p) basis set. Initially, a large number of geometries
were generated using the conformational search module
of Hyperchem with the MM+ force field.26 The selected
All reagents and solvents were used directly as purchased or
purified according to standard procedures. Analytical thin layer
chromatography was carried out using commercial silica gel
plates and visualization was effected with short wavelength UV
light (254 nm) and a p-anysaldehyde solution (2.5 mL of p-
anysaldehyde + 2.5 mL of H2SO4 + 0.25 mL of AcOH + 95 mL of
EtOH) with subsequent heating. Column chromatography was
performed with silica gel 60 H, slurry packed, run under low
pressure of nitrogen using mixtures of hexane and ethyl
structures were then optimized at the
B97XD/6-311+G**
level of theory, using Gaussian 09.22 The geometries for all
structures were fully optimized in chloroform as solvent using
SMD.21 The reported thermochemical properties include zero-
point energies (ZPEs) without scaling and were calculated at 1
atm, and the corresponding temperature for each reaction.
Normal mode analysis was used to confirm the nature of
the stationary points and to evaluate the thermochemical
properties. All transition structures were confirmed to have
only one imaginary frequency corresponding to the formation
of the expected bonds. Intrinsic reaction coordinate (IRCs)
calculations were run to verify the connectivity between
reactants, TSs and products.
1
acetate. NMR spectra were recorded at 300 MHz for H, and
75 MHz for 13C with CDCl3 as solvent and (CH3)4Si (1H) or CDCl3
(13C, 76.9 ppm) as internal standards. Chemical shifts are
reported in delta (δ) units in parts per million (ppm) and
splitting patterns are designated as s, singlet; d, doublet; t,
triplet; q, quartet; m, multiplet and br, broad. Coupling
constants are recorded in Hertz (Hz). Isomeric ratios were
1
determined by H NMR analysis. The structure of the products
were determined by a combination of spectroscopic methods
such as IR, 1D and 2D NMR (including NOE, DEPT, COSY, HSQC
Acknowledgements
and HMBC experiments) and HRMS. Infrared spectra were This research was supported by UNR (BIO 316), ANPCyT (PICT-
recorded using sodium chloride plates pellets. Absorbance 2012-0970), and CONICET (PIP 11220130100660CO). G.G.G.
frequencies are recorded in reciprocal centimeters (cm–1). High and N.G. thank CONICET for the award of a fellowship.
resolution mass spectra (HRMS) were obtained on a TOF-Q LC-
MS spectrometer. Optical rotations were determined using a
digital polarimeter in 100 mm cells and the sodium D line (589
References
nm) at room temperature in the solvent and concentration
indicated.
Levoglucosenone (
The
melting
points
are
uncorrected.
1
2
(a) J. R. Lewis, Nat. Prod. Rep. 2001, 18, 95. (b) D.
O’Hagan, Nat. Prod. Rep. 2000, 17, 435.
1
) was prepared from our microwave-
(a) P. V. Fish, M. D. Andrews, M. J. Fray, A. Stobie, F.
Wakenhut, G. A. Whitlock, Bioorg. Med. Chem. Lett.
2009, 19, 2829. (b) C. V. Galliford, K. A. Scheidt, Angew.
Chem., Int. Ed. 2007, 46, 8748. (c) F. Bellina, R. Rossi,
Tetrahedron 2006, 62, 7213. (d) M. G. Moloney, P. C.
Trippier, M. Yaqoob, Z. Wang, Curr. Drug Discovery
assisted pyrolysis of acid-pretreated microcrystalline
cellulose.25
General Procedure for the synthesis of 7a-c 5,6
.
To a solution of
1
(126 mg, 1 mmol) and the corresponding α-iminoester 11
(1.5 mmol), prepared by condensation of alanine methyl ester
hydrochloride and benzaldehyde (9a), p-anisaldehyde (9b), or
p-nitrobenzaldehyde
successively added AgOAc (50 mg, 0.3 mmol) and DBU (45
0.3 mmol). The mixture was stirred for 1 hour at room
temperature under argon atmosphere and in the absence of
light (flask covered in aluminum foil). After filtration through
Celite, the filtrate was concentrated to dryness under reduced
pressure. The crudes were purified by column chromatography
with elution with mixtures of hexanes and ethyl acetate to
afford pure compounds 7a (97%), 7b (86%) and 7c (63%).
Technol. 2004,
1
, 181. (e) S. R. Hussaini, M. G. Moloney,
, 1838. (f) W. H. Pearson,
Org. Biomol. Chem. 2003,
1
(
9c), in dry MeCN (10 mL) was
L,
Pure Appl. Chem. 2002, 74, 1339.
C. Nájera, J. M. Sansano, Org. Biomol. Chem. 2009,
4567.

3
4
5
6
7
7,
E. Conde, D. Bello, A. de Cozar, M. Sanchez,; M. A.
Vazquez, F. P. Cossio, Chem. Sci. 2012, , 1486.
3
A. M. Sarotti, R. A. Spanevello, A. G. Suárez, G. A.
Echeverría, O. E. Piro, Org. Lett. 2012, 14, 2556.
G. G. Gerosa, R. A. Spanevello, A. G. Suárez, A. M.
Sarotti, J. Org. Chem. 2015, 80, 7626.
(a) L. M. Stanley, M. P. Sibi, Chem. Rev. 2008, 108
,
2887; (b) H. Pellissier, Tetrahedron 2007, 63, 3235; (c)
G. Pandey,; P. Banerjee, S. R. Gadre, Chem. Rev. 2006,
106, 4484; (d) I. Coldham, R. Hufton, Chem. Rev. 2005,
105, 2765.
General Procedure for the isomerization of 7a-c
. The
corresponding pyrrolidine 7a-c (0.1 mmol) was placed in a
Hach tube and was dissolved in chloroform (1.25 mL). After the
addition of the appropriate amount of acid, the tube was
sealed and heated to 70ºC for the corresponding period of
8
C. Nájera, J. M. Sansano, Curr. Org. Chem. 2003, 7,
1105.
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 9
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
a) C. Nájera, J. M. Sansano, Angew. Chem. Int. Ed. 2005,
ARTICLE
9
View Article Online
44, 6272. b) J. Adrio, J. C. Carretero. Chem. Comm.
2011, 47, 6784. c) J. Adrio, J. C. Carretero. Chem.
Comm. 2014, 50, 12434.
DOI: 10.1039/C6OB02457B
10 (a) G. A. Oliveira Udry, E. Repetto, O. Varela. J. Org.
Chem. 2014, 79, 4992. (b) G. A. Oliveira Udry, E.
Repetto, D. R. Vega, O. Varela. J. Org. Chem. 2016, 81,
4179. c) G. Bashiardes, C. Cano, B. Mauzé, Synlett 2005,
4
, 587.
11 (a) M. B. Comba, A. G. Suárez, A. M. Sarotti, M. I.
Mangione, R. A. Spanevello, E. D. V. Giordano, Org.
Lett. 2016¸ 18, 1748. (b) V. Corne, M. C. Botta, E.D.V.
Giordano, G.F. Giri, D.F. Llompart, H.D. Biava, A.M.
Sarotti, M.I. Mangione, E.G. Mata, A.G. Suárez, R.A.
Spanevello, Pure & Appl. Chem. 2013, 85, 1683.
12 (a) J. L. García Ruano, A. Tito, M. T. Peromingo, J. Org.
Chem. 2002, 67, 981. (b) J. Casas, R. Grigg, C. Nájera, J.
M. Sansano, Eur. J. Org. Chem. 2001, 1971. (c) O.
Tsuge, S. Kanemasa, M. Yoshioka, J. Org. Chem. 1988,
53, 1384.
13 (a) E. Wenkert, J. H. Udelhofen, N. K. Bhattacharyya, J.
Am. Chem. Soc. 1959, 81, 3763. (b) G. Cravotto, G.
Battista Giovenzana, T. Pilati, M. Sisti, G. Palmisano, J.
Org. Chem. 2001, 66, 8447. (c) G. Laus, D. Brossner, G.
Senn,; Wurst, K. J. Chem. Soc., Perkin Trans. 2 1996,
1931. (d) G. Laus, J. Chem. Soc., Perkin Trans. 2 1998,
315.
14 E. D. Cox, J. M. Cook, Chem. Rev. 1995, 95, 1797.
15 M. L. Van Linn, J. M. Cook, J. Org. Chem. 2010, 75
,
3587.
16 Z. Amara, G. Bernadat, P.–E. Venot, P. Retailleau, C.
Troufflard, E. Drège, F. Le Bideau, D. Joseph, Org.
Biomol. Chem. 2014, 12, 9797.
17 (a) P. Przybylski, K. Pyta, J. Czupryniak, B. Wicher, M.
Gdaniec, T. Ossowski, B. Brzezinski, Org. Biomol. Chem.,
2010,
8
, 5511. (b) B. Sliter, J. Morgan, A. Greenberg, J.
2770. (c) V. Vimalraj, K.
Org. Chem. 2011, 76
,
Pandiarajan, Magn. Reson. Chem. 2011, 49, 682.
18 (a) L. P. Hammett. Chem. Rev. 1935, 17, 125. (b) L. P.
Hammett. J. Am. Chem. Soc. 1937, 59, 96.
19 J. E. Johnson, N. M. Morales, A. M. Gorczyca, D. D.
Dolliver, M. A. McAllister. J. Org. Chem. 2001, 66, 7979.
20 (a) L. Belding, S. Zaretsky, B. H. Rotstein, A. K. Yudin, T.
Duddin. J. Org. Chem. 2014, 79, 9465. (b) M. H. Haindl,
J. Hioe, R. M. Gschwind. J. Am. Chem. Soc. 2015, 137
,
12842. (c) D. Seebach, R. Gilmour, U. Groselj, G.
Deniau, C. Sparr, M.-O. Ebert, A. K. Beck. Helv. Chim.
Acta 2010, 93, 603. (d) L. E. Overman, W. C. Trenkle.
Isr. J. Chem. 1997, 37, 23. (e) M. Lukowski, K. Jacobs, P.
Hsueh, H. A. Lindsay, M. C. Milletti. Tetrahedron 2009,
65, 10311.
21 A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys.
Chem. B, 2009, 113, 6378.
22 M. J. Frisch, et. al. Gaussian 09, Revision B.01,
Gaussian, Inc.: Wallingford CT, 2009.
23 J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys.,
2008, 10, 6615.
24 (a) M. Hamzehloueian, Y. Sarrafib, Z. Aghaeib. RSC Adv.,
2015,
5, 76368. (b) L. A. Kondacs, M. V. Pilipecz, Z.
Mucsi, B. Balázs, T. Gáti, M. Nyerges, A. Dancsó, P.
Nemes. Eur. J. Org. Chem. 2015, 6872. (c) C. Sobhi, A.
K. Nacereddine, A. Djerourou, M. Ríos‐Gutiérrez, L. R.
Domingo. J. Phys. Org. Chem. 2016, in press. DOI
10.1002/poc.3637.
25 A. M. Sarotti, R. A. Spanevello, A. G. Suárez. Green
Chem. 2007, 9, 1139.
26 Hyperchem Professional Release 7.52, Hypercube, Inc.,
2005.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins