Organocatalyzed aldol reaction between pyridine-2-carbaldehydes and α-ketoacids: A straightforward route towards indolizidines and isotetronic acids

Article abstract of DOI:10.1002/chem.201302264

Enantioselective aldol reactions between substituted pyridine carbaldehydes and α-ketoacids were shown to provide isotetronic acids or their corresponding pyridinium salts, depending on the nature of the substituents on the pyridine ring. The pyridinium salts were generated through nucleophilic attack of the pyridine nitrogen atom onto the reactive keto functional group. Moderate-to-good yields of both compounds were typically obtained and high levels of enantioselectivity were observed by using benzimidazole pyrrolidine I as a catalyst. Hydrogenation of the resulting pyridinium salts led to new indolizidines with high ee values and diastereocontrol. X-ray diffraction studies allowed the determination of the relative configuration of the products. Finally, DFT calculations were performed to rationalize the divergent pathway as a function of the pyridine substituents. BIPpity boppity boo: Asymmetric aldol reactions between pyridine carbaldehydes and α-ketoacids were shown to provide isotetronic acids or their corresponding pyridinium salts with high enantioselectivities by using benzimidazole pyrrolidine (BIP) as a catalyst. Hydrogenation of the resulting pyridinium salts led to new indolizidines with high enantioselectivity and diasterocontrol (see scheme). Copyright

Full text of DOI:10.1002/chem.201302264

FULL PAPER
DOI: 10.1002/chem.201302264
Organocatalyzed Aldol Reaction Between Pyridine-2-carbaldehydes and
a-Ketoacids: A Straightforward Route towards Indolizidines and
Isotetronic Acids
Virginie Liautard, Damien Jardel, Clotilde Davies, Muriel Berlande, Thierry Buffeteau,
Dominique Cavagnat, Frꢀdꢀric Robert,* Jean-Marc Vincent,* and Yannick Landais*[a]
Abstract: Enantioselective aldol reac-
tions between substituted pyridine car-
baldehydes and a-ketoacids were
shown to provide isotetronic acids or
their corresponding pyridinium salts,
depending on the nature of the sub-
stituents on the pyridine ring. The pyri-
dinium salts were generated through
nucleophilic attack of the pyridine ni-
trogen atom onto the reactive keto
functional group. Moderate-to-good
yields of both compounds were typical-
ly obtained and high levels of enantio-
selectivity were observed by using ben-
zimidazole pyrrolidine I as a catalyst.
Hydrogenation of the resulting pyridi-
nium salts led to new indolizidines with
high ee values and diastereocontrol.
X-ray diffraction studies allowed the
determination of the relative configura-
tion of the products. Finally, DFT cal-
culations were performed to rationalize
the divergent pathway as a function of
the pyridine substituents.
Keywords: aldol reaction · enantio-
selectivity · heterocycles · indolizi-
dines · organocatalysis
Introduction
to the desired five-membered rings in high yields and with
high enantioselectivities.[4] This work complemented previ-
ous approaches to isotetronic acids by using chiral auxilia-
ries, organometallic catalysis, and proline-catalyzed pyruvate
homo-aldol reactions and represents one of the few exam-
ples of organocatalyzed asymmetric aldol reactions that ex-
ploit 1,2-dicarbonyl compounds as pronucleophiles.[5]
Naturally occurring isotetronic acids exhibit relevant bio-
logical properties,[6] including inhibitory activity of aldose re-
ductase[6a] and antitumor activity against various carcino-
mas.[6b] These simple motifs are also found as key fragments
in more-complex targets, such as erythronolide A.[6d] During
the course of our studies on organocatalyzed aldol reactions,
we made some unexpected observations in the aldol reac-
tions between pyridine-2-carbaldehydes 1 and various a-ke-
toacids 2 (Scheme 1). The long-standing interest of this class
of compounds and recent efforts towards their synthesis[7]
prompted us to provide a full report on these unexpected
results herein.
The organocatalyzed aldol reaction has been at the onset of
the “renaissance” of aminocatalysis[1] and it is still the sub-
ject of intense research activity, with the development of
more-efficient catalysts and their utilization in the total syn-
thesis of natural products.[2] In this context, some years ago,
we reported the development of a readily available benzim-
ACHTUNGTRENNUNGidazole pyrrolidine catalyst (I, “BIP”) that was derived from
l-proline and exhibited remarkable reactivity in various re-
actions, including aldol reactions.[3] Whereas most catalysts
require the use of an excess of one of the reagents (usually
the ketone donor), catalyst I has been shown to catalyze
aldol processes by using an equimolar amount of the ketone
and aldehyde partners. Amongst the various aldol processes
that have been catalyzed by compound I, a straightforward
synthesis of isotetronic acids 3 from a-ketoacids (e.g. 2) and
aromatic or aliphatic aldehydes was developed, thus leading
[a] Dr. V. Liautard, D. Jardel, C. Davies, M. Berlande, Dr. T. Buffeteau,
Dr. D. Cavagnat, Dr. F. Robert, Dr. J.-M. Vincent,
Prof. Dr. Y. Landais
Universitꢀ de Bordeaux
Institut des Sciences Molꢀculaires, UMR-CNRS 5255
351, cours de la liberation, 33405 Talence Cedex (France)
Fax : (+33)540006286
E-mail: f.robert@ism.u-bordeaux1.fr
jm.vincent@ism.u-bordeaux1.fr
y.landais@ism.u-bordeaux1.fr
Supporting information for this article, including the experimental de-
tails, characterization data of the products, spectra of new com-
pounds, computational details, energy values, and Cartesian coordi-
nates of the structures discussed in the text, is available on the
WWW under http://dx.doi.org/10.1002/chem.201302264.
Scheme 1. Aldol reactions between pyridine carbaldehydes
a-ketoacids 2 catalyzed by compound I.
1 and
Chem. Eur. J. 2013, 00, 0 – 0
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

Depending on the nature of the substituents on the pyri-
dine skeleton, the reactions between compounds 1 and pyr-
uvic-acid derivatives 2 were shown to effectively provide the
desired isotetronic acids 3 or, more surprisingly, their corre-
sponding pyridinium salts 4 (Scheme 1). To the best of our
knowledge, this transformation has never been reported
before. Interestingly, complete hydrogenation of the pyridi-
nium salts 4 eventually led to the corresponding non-natural
indolizidines[8] in good yields and with high enantiocontrol
in only two steps from commercially available precursors.
This reaction likely proceeded through the stereoselective
addition of an enamine that was formed from the a-ketoacid
2 and catalyst I onto the aldehyde functional group of pyri-
dine 1; subsequent lactonization provided compound 3.[4] Al-
ternatively, the resulting aldol product may evolve through
the attack of the pyridine nitrogen atom onto the electro-
philic a-keto group to provide compound 4. Herein, we de-
scribe the development of this unusual organocatalyzed
catalyst I is given in the Supporting Information). In con-
trast with previous results, the corresponding isotetronic
acid 3a was not observed with this particular aldehyde; in-
stead a pyridinium salt 4a was isolated in various amounts,
depending on the reaction conditions and on the isolation
process. The enantioselectivity of compound 4a could not
be obtained through chiral HPLC and, thus, was evaluated
by measuring the optical rotation of the sample in various
solvents. The optical rotation of enantiomerically pure pyri-
dinium 4a was estimated based on the observation that a
sample of compound 4a in THF (Table 1, entry 9) led to in-
dolizidine 6a, after reduction under PtO2 catalysis[9] and
benzylation of acid 5a, with an ee value of >99%
(Scheme 2). In all cases, after evaporation of the solvent at
À
aldol reaction, which results in the formation of new C C
À
and C N bonds and the generation of up to five stereogenic
centers in only two steps. We discuss the scope and limita-
tions of our method, provide a rationalization of the effects
that favor the formation of compound 4 over compound 3,
supported by DFT calculations, and describe some useful
transformations of aldol products 4. Finally, the configura-
tions of some pyridinium salts, both in the solid state and in
solution, are discussed based on IR and vibrational circular
dichroism (VCD) studies, associated with theoretical calcu-
lations.
Scheme 2. Reduction of compound 4a and X-ray structures of com-
pounds 4a and 5a.
Results and Discussion
room temperature and column chromatography on silica gel
(CH2Cl2/MeOH), compound 4a was obtained as a single
isomer and its structure was unambiguously determined
through X-ray diffraction studies (see below; Scheme 2).
The relative and absolute configurations of zwitterion 5a
were also determined by X-ray diffraction studies. As shown
in Table 1, the best results were obtained in THF, CH2Cl2,
Et2O, and MeOH, with better optical purity in the former
solvent. Interestingly, pyridinium 4a was found to precipi-
tate out of THF, thereby allowing its isolation by simple fil-
tration. The reaction also proceeded in water, although it
provided compound 4a in a mixture of unidentified prod-
ucts, whereas, in 1,4-dioxane, compound 4a was almost race-
mic. No trace of compound 4a could be detected when the
reaction was performed without solvent (Table 1, entry 1).
Small amounts of compound 4a were formed in DMF, but
evaporation of the solvent essentially led to degradation
(Table 1, entry 2). When the reaction was performed in the
absence of a catalyst or with l-proline, no trace amounts of
isotetronic acids or pyridinium salts could be detected. We
previously showed that under acidic conditions, proline
proved to be poorly reactive.[3a] Finally, attempts to extend
this method to a-ketoesters instead of their corresponding
acids were unsuccessful.
Preliminary studies were performed by reacting pyridine
carbaldehyde 1a with a-ketobutyric acid 2a in the presence
of 10 mol% of catalyst I (Table 1; an optimized synthesis of
Table 1. Aldol reaction between pyridine 1a and a-ketobutyric acid 2a.
Entry Solvent
Yield of Optical rotation of
Estimated ee of
4a [%][c]
4a [%][a] 4a (c)[b]
1
2
3
4
5
6
7
8
9
neat
toluene
Et2O
DMF
water
CH2Cl2
0
57
74
0
–
–
33
78
–
62
À11.3 (0.52, MeOH)
À26.9 (0.505, MeOH)
–
34
70
À26.9 (0.52, MeOH)
À27.3 (0.52, MeOH)
0.9 (0.52, MeOH)
À28.2 (0.52, MeOH)
À34.3 (0.56, MeOH)
80
1,4-dioxane 60
À3
MeOH
THF
78
75
82
>99
[a] Yield of isolated compound 4a after column chromatography
(CH2Cl2/MeOH). [b] Reaction was performed at 258C. [c] Enantiomeric
excess of compound 4a: ee=[a]D exp/[a]D pure 4aꢂ100, by using [a]D pure 4a
À34.3 (c=0.51, MeOH) (see text).
=
&
2
&
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Straightforward Route to Indolizidines and Isotetronic Acids
FULL PAPER
With these results in hand, we varied the nature of the
substituents on the pyridine carbaldehyde by using com-
pound 2a as the enamine precursor. The results shown in
Table 2 indicate that the nature of the substituents on pyri-
Table 2. Aldol reactions between pyridines 1a–1k and 2a.
Entry
1
R
H
3
Yield of 3 ee of 3
4
Yield of 4
[%][a,b]
[%][c]
[%][a,b]
Scheme 3. Access to pyridinium salts 4e–4h from pyruvic-acid derivatives
2b–2d and the X-ray structure of pyridinium salt 4e.
1
2
3
1a
–
–
–
–
–
–
–
–
–
–
–
4a 75 (82)
4b 73 (80)
4c 77 (41)
4d 29 (41)
1b 5-Me
1c 2-Me
1d 5-F
4
–
5
6
7
8
9
10
11
1e 2-Br
3e 73 (52)
3 f 42 (20)
3g 69 (72)
3h 84 (82)
93
n.d.
75
89
82
98
90
–
–
–
–
–
–
–
–
–
–
–
–
–
–
of unidentified by-products, which were discarded upon
crystallization. To the best of our knowledge, the reaction
that provided pyridinium 4e represents the first example of
organocatalyzed aldol process that exploits pyruvic acid as a
pronucleophile.[11] The relative configuration of the C7 and
C8 stereocenters in pyridinium salts 4a–4h was assumed
based on the configurations in compounds 4a and 4e that
were obtained by X-ray diffraction studies (Scheme 2 and
Scheme 3). However, it is worthy of note that, in the solid
state, a cis-C7–C9 relative configuration is found in com-
pound 4a, whilst it is trans in compound 4e (see below).
Then, the pyridinium ring was fully hydrogenated as
above, under PtO2 catalysis,[9] thereby affording the corre-
sponding indolizidines 5a–5d (Scheme 4). Esterification of
1 f 2-CHO
1g 2-p-CNPh
1h 2-OMe
1i 2-pyrrolidinyl 3i 55 (14)
1j 3-Br
3j 75 (90)
3k 63 (50)
1k 3-CN
[a] Yield of isolated product after column chromatography (see the Sup-
porting Information). [b] Yield of isolated compound as a racemate (re-
action performed with (+/-)-I) is given in parentheses. [c] Enantiomeric
excesses were measured by chiral HPLC (Chiralpak AD-H and IA col-
umns). n.d.=not determined.
dine 1 has a strong influence on the outcome of the reaction.
The presence of hydrogen, methyl, or fluorine groups at the
C5 position (1a, 1b, and 1d; Table 2, entries 1, 2, and 4) ex-
clusively afforded the corresponding pyridinium compounds
(4a, 4b, and 4d, respectively), in generally good yields as a
single diastereomer for both the enantiomeric and racemic
series. Changing the position of the Me substituent (to the
C2 position) as in compound 1c also led to pyridinium 4c
(Table 2, entry 3). In contrast, other electron-withdrawing
substituents (Table 2, entry 5–7) and also electron-donating
substituents (Table 2, entries 8 and 9) exclusively led to their
corresponding isotetronic acids in good yields and high
enantioselectivities (measured or estimated by chiral
HPLC). As above, the enantiomeric excesses for pyridini-
ums 4a–4d could not be directly determined by HPLC anal-
ysis and, thus, were estimated after further functionalization
(see below). Similar to compound 4a, the relative configura-
tions of compounds 4b and 4c were determined by X-ray
diffraction (see the Supporting Information), after further
elaboration of the pyridine ring (see below).
Scheme 4. Access to indolizidines 6a–6d from pyridinium salts 4 and the
X-ray structure of pyridinium salt 5b.
Having established the reactivity of a-ketoacid 2a in the
presence of various pyridine carbaldehydes, we studied the
effects of the nature of a-ketoacid substituents on the for-
mation of pyridinium salts 4. The reactions of pyridine car-
baldehydes 1a and 1b with a-ketoacids 2b–2d led to salts
4e–4h in moderate-to-good yields with high levels of dia-
stereocontrol (except for compounds 4 f and 4h;
Scheme 3).[10] Moderate yields resulted from the formation
the products with TMSCHN2 furnished their corresponding
methyl esters in good yields, but their enantiomeric excesses
could not be measured by chiral HPLC. Better results were
obtained by performing the direct benzylation of com-
pounds 5a–5d after hydrogenation of compounds 4a–4c
Chem. Eur. J. 2013, 00, 0 – 0
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

F. Robert, J.-M. Vincent, Y. Landais et al.
and 4e, thus affording UV-active benzyl esters 6a–6d in
good yields (Scheme 4). Enantiomeric excesses, as measured
by chiral HPLC, showed that compounds 6a–6d were essen-
tially pure. The relative configuration of acid 5b, which was
obtained through the catalytic reduction of compound 4b,
was determined by X-ray diffraction analysis, which allowed
us to assign similar configurations for benzyl esters 6a–6d.
Crystallographic studies on compounds 4a and 4e (con-
firmed by X-ray analysis of compounds 5a and 5b) showed
that the secondary alcohol at the C7 position in pyridinium
salts 4a–4h possessed an R configuration, in good agree-
ment with that in isotetronic acids that were previously pre-
pared through BIP-I-catalyzed aldol reaction between a va-
riety of aldehydes and a-ketoacids.[4] Consequently, we can
assume that isotetronic acids 3e–3k bear a similar R config-
uration. The aldol process provides a second stereogenic
center at the C8 position (e.g., intermediate B), which, ac-
cording to transition state A, should possess an S configura-
tion (Scheme 5). This stereochemistry is lost through enoli-
dithionite,[14] or Hantzsch esters[15] failed, whilst hydrogena-
tion by using [IrACTHNUTRGNEUNG
(cod)2Cl2]/H2 (cod=1,5-cyclooctadiene)[16]
led to complete hydrogenation. Most interestingly, the treat-
ment of pyridinium salts 4a,b with H2O2 in MeOH led to
esters 7a,b in excellent yields after conversion of their cor-
responding acids with TMSCHN2 (Scheme 6).[17] The forma-
Scheme 6. Further transformations of pyridinium salts 4a–4c.
tion of compounds 7a,b through oxidative decarboxylation
likely involves the oxidation of the carboxylic acid into a
peracid, followed by a Baeyer–Villiger-type process.[18] Ten-
tative allylation of the pyridinium salts in the presence of al-
lylindium[19] in water did not give the expected allyldihydro-
pyridine, but rather afforded the highly substituted butyro-
lactones (8a,b), which resulted from nucleophilic addition
onto the a-keto group, with good diastereocontrol and high
enantioselectivity, albeit in modest yields.[20] Finally, the
enantiomeric excesses of compounds 7a,b and 8a,b con-
firmed the high enantiopurity of pyridinium salts 4a,b, as es-
timated above.
At this point, several questions remained unanswered,
that is, 1) the difference in reactivity between pyridine car-
baldehydes 1a–1d and 1e–1k; 2) the behavior of pyridinium
salts 4c and 4d, which were prone to afford isotetronic acids
after standing for a short period of time in solution (whereas
compound 4b was indefinitely stable for at least one year!);
and 3) the observation that compound 4a crystallized as its
7-OH/9-OH cis isomer, whereas pyridinium 4e crystallized
as its 7-OH/9-OH trans isomer (Scheme 2 and Scheme 3).[21]
To answer these questions and to gain a better insight into
the mechanism of pyridinium formation and the nature of
the equilibria that are involved in the process, DFT compu-
tational studies were performed (see the Experimental Sec-
tion and the Supporting Information). Starting from ketoa-
cid 9, which was formed from the organocatalyzed aldoliza-
tion reaction, two different pathways were envisaged. The
pyridine nitrogen center could attack the ketone group,
thereby affording the pyridinium salts (4) as cis or trans iso-
mers, depending on which diastereotopic face of the ketone
was attacked. Alternatively, the alcohol function could react
with the carboxylic acid group, thereby affording compound
Scheme 5. Transition-state models for the aldol and hydrogenation
processes.
zation during the subsequent lactonization step into isote-
tronic acids 3e–3k, but it is retained upon the formation of
pyridinium salts 4. Based on transition state model A,[4,12]
we can conclude that the cyclization of aldol adduct B into
pyridinium salts 4a–4h occurs with retention of the configu-
ration at the C8 position. Finally, the relative configuration
of the stereocenters at the C2/C5 and C6 positions on the
six-membered ring in compounds 6a–6d is fully consistent
with the PtO2-catalyzed hydrogenation step occurring from
the top face (e.g., model C), that is, anti to the OH and R3
substituents on the five-membered ring. Similarly, the ster-
eochemistry at the C9 position is fixed during the hydroge-
nation from the same face of an iminium intermediate.
Disappointingly, efforts to partially reduce the pyridinium
ring in compounds 4 by using borohydrides (NaBH3CN),[13]
&
4
&
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Straightforward Route to Indolizidines and Isotetronic Acids
FULL PAPER
10 and, finally, the isotetronic acid 3, after dehydration and
keto/enol isomerization (Scheme 7).
Computation of the energies of all of the species, as well
as the transition states, in the reaction of pyridine-2-carbal-
hydes, four of which led to the pyridinium salts (1a–1d),
whilst the other four afforded isotetronic acids (1e, 1h, 1j,
and 1k). A clear picture emerges when comparing the three
particular cases shown in Figure 2. For aldehyde 1b, the ex-
Scheme 7. Equilibria and species that are involved in the reaction.
dehyde 1a led to the following results. Whereas in the gas
phase, pyridinium salts were much higher in energy than iso-
tetronic acids, calculations in THF (the optimal solvent ex-
perimentally) by using IEFPCM as the continuum solvation
model revealed that pyridinium trans isomer 4atrans was
almost isoenergetic with the isotetronic acid and should be
the compound present in solution (Figure 1).[22] The energy
of the transition state between compound 9 and the pyridini-
um salts was very low (3–7 kcalmolÀ1); thus, the equilibrium
should be very fast. In contrast, the much-higher calculated
energy barrier (TStet) for the conversion of the free acid 9
into intermediate 10 was in good agreement with the sole
formation of the pyridinium salt from this aldehyde.
Figure 2. Comparison of the energy profiles [kcalmolÀ1] at the M06-2X/6-
31+GACTHUNRTGNEUNG(d,p) level of theory in THF for aldehydes 1b, 1d, and 1e.
clusive formation of the pyridinium salt 4btrans may be ex-
plained by its greater stability compared to that of its iso-
A
energies of pyridinium salt 4dtrans, which was obtained from
aldehyde 1d, revealed that isomer 4dtrans was the kinetic
product, thus explaining its strong tendency to isomerize
into the thermodynamically more stable isotetronic 3d upon
standing in solution. Finally, with aldehyde 1e, the free acid
form 9 was more stable than the 2-pyridinium salt 4 (8.3
versus 6.1 kcalmolÀ1); thus, the compound has no option
other than forming the isotetronic acid 3e.
The energies of the 2-pyridinium salts, the free acid, and
the isotetronic acid were computed for eight different alde-
When comparing the values of eight aldehydes (Table 3),
the difference between the two sets of aldehydes is clear
and is in good agreement with the experimental observa-
tions. Although the isotetronic acids are always the more-
stable compounds (except for compound 1b, for which pyri-
dinium salt 4b is more stable), a clear difference can be
seen in terms of the energy difference between the isote-
tronic acids and the pyridinium salts (DG4transÀ3). The energy
difference between the aldehydes that afford pyridinium
Table 3. Differences in Gibbs free energy (in kcalmolÀ1) between the
species that were involved in the reactions of eight different aldehydes.
1
DG3À9
DG4transÀ9
DG4cisÀ9
DG4transÀ3
1a
1b
1c
1d
1e
1h
1j
À7.2
À9.6
À7.4
À9.8
À6.1
À7
À6.9
À12.4
À5.5
À7.3
2.2
À3.1
À3.3
À1.1
À2.7
À7.1
À1.1
À2.3
5.9
1.9
1.0
2.8
0.3
À2.8
1.9
2.5
8.3
3.9
3.5
6.1
À6.8
À7.2
Figure 1. Energy profile [kcalmolÀ1] of the reaction pathway at the M06-
2X/6-31+GACHTUNGTRENNUNG(d,p) level of theory in THF for aldehyde 1a.
1k
Chem. Eur. J. 2013, 00, 0 – 0
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

F. Robert, J.-M. Vincent, Y. Landais et al.
salts is small, whilst it is much larger between those that
afford isotetronic acids, thus confirming the experimentally
observed difference in reactivity. The energy difference be-
tra for isomers 4btrans and 4bcis should allow us to assess
which diastereoisomer is predominant in solution.
IR and VCD spectra of pyridinium 4b were recorded in
CD3CN and are shown in Figure 3a,b, respectively. Calcu-
lated IR and VCD spectra of isomers 4btrans and 4bcis at the
CAM-B3LYP/aug-cc-pVTZ level of theory are also shown
in Figure 3a,b, respectively. It is clear that the experimental
IR spectrum is well reproduced by the one calculated for
the 4btrans isomer, whereas strong divergences are observed
for the 4bcis isomer. It is noteworthy that a frequency shift is
observed for the stretching nC=O vibrations of the carboxy-
late groups due to intermolecular interaction with the sol-
vent and that the intense band calculated at 1403 cmÀ1 for
the C-O-H bending appears as a very broad band with weak
intensity in the experimental spectrum. The VCD spectrum
confirmed the presence of the trans isomer in CD3CN solu-
tion, because the calculated bands for the 4btrans isomer cor-
related and agreed in sign for most of the vibrational modes
that were observed in the experimental spectrum. Finally,
the IR and VCD spectra were recorded in CD3OD, because
this solvent was used to crystallize compound 4a. DFT cal-
culations were performed on the deuterated hydroxyl form
of pyridinium 4b to take into account the H/D exchange.
Once again, good agreement was obtained between the ex-
perimental IR and VCD spectra and those calculated for
the 4btrans isomer (see the Supporting Information), thus re-
vealing its presence in CD3OD solution. Thus, theoretical
analysis of the reaction is valid and the isolation of the cis
isomer by recrystallization is most likely due to a shift in the
equilibrium during the crystallization process.
tween the free acids 9 and the isotetronic acids 3 (DG3À9
)
were within the same range for all of the aldehydes that
were considered; the difference in reactivity was due to sta-
bilization/destabilization of the pyridinium forms by the pyr-
idine substituents. The destabilization of the pyridinium
salts was striking in terms of the energy difference between
the free acid and the cis/trans pyridinium (DG4cisÀ9,
DG4transÀ9), for which marked differences between the eight
aldehydes were observed.
Thus, the calculations and experimental results strongly
suggest that the pyridinium salt that is present in solution is
the trans isomer. Interestingly, single-crystal X-ray diffrac-
tion established a cis-C7–C9 relative configuration in com-
pound 4a. To determine which form was present in solution,
IR and VCD studies were performed to determine the rela-
tive and absolute configurations of compound 4b in solu-
tion. These spectroscopic methods, in combination with
DFT calculations, are well-established methods for the de-
termination of the absolute configuration and solution con-
formation of chiral molecules.[23] Thus, a comparison of the
experimental IR and VCD spectra with the calculated spec-
Conclusion
In summary, asymmetric aldol reactions between pyridine
carbaldehydes and a-keto acids catalyzed by benzimidazole
pyrrolidine I lead to aldol adducts that cyclized through two
different pathways, depending on the substitution pattern of
the pyridine nucleus, thereby affording either pyridinium
salts or isotetronic acids. Further functionalization of the
pyridinium salts 4 was shown to provide access to valuable
building blocks in their enantiomerically pure form, includ-
ing indolizidines, aldol products, and butyrolactones. Ration-
alization of both pathways was provided through DFT stud-
ies, which correlated well with experimental observations.
For example, the formation of pyridinium salt 4b as a
unique compound was explained by its greater stability com-
pared to that of its isotetronic form. In contrast, other less-
stable pyridinium salts, such as compounds 4c and 4d,
tended to rearrange into their corresponding isotetronic
acids, a feature that was also well-illustrated by the DFT
calculations.
Figure 3. Comparison of the experimental a) IR and b) VCD spectra in
CD3CN with those calculated at the CAM-B3LYP/aug-cc-pVTZ DFT
level of theory for diastereomers 4btrans and 4bcis
.
&
6
&
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Straightforward Route to Indolizidines and Isotetronic Acids
FULL PAPER
d1), 5.39 (d, J=9.6 Hz, 0.1H; d2), 5.28–5.11 (m, 2H), 3.08 (dq, J=7.0,
9.7 Hz, 0.1H; d2), 2.85 (dq, J=6.8, 9.7 Hz, 0.9H; d1), 2.64 (ddt, J=1.2,
6.9, 14.0 Hz, 1H), 2.57 (ddt, J=1.07, 7.8, 14.0 Hz, 1H), 2.45 (s, 0.3H; d2),
2.44 (s, 2.7H; d1), 1.11 (d, J=7.2 Hz, 0.3H; d2), 1.01 ppm (d, J=6.8 Hz,
2.7H; d1); 13C NMR (100 MHz, CD3OD): d=178.8 (Cq), 154.1 (Cq),
148.03 (Cq; d1), 147.97 (Cq; d2), 140.6 (CH; d2), 140.5 (CH; d1), 135.0
(Cq), 133.5 (CH; d1), 133.5 (CH; d2), 125.3 (CH; d2), 125.2 (CH; d1),
119.7 (CH2), 82.7 (CH; d1), 81.4 (CH; d2), 77.2 (C9), 47.3 (CH; d2), 44.0
(CH; d1), 40.5 (CH2; d1), 38.5 (CH2; d2), 18.3 (CH3; d1), 18.2 (CH3; d2),
9.7 (CH3; d2), 8.4 ppm (CH3; d1); IR (ATR): n˜ =3441, 2981, 2950,
Experimental Section
Preparation of compound 4a: To a solution of 2-ketobutyric acid (2a,
441 mg, 4.3 mmol, 1 equiv) in THF (13 mL) were added 2-pyridine car-
boxaldehyde (1a, 409 mg, 4.3 mmol, 1 equiv) and catalyst I (10 mol%).
The reaction mixture was stirred at RT for 3 days under a N2 atmosphere.
Filtration of the precipitate afforded the desired compound (4a) as a
white solid (661 mg, 73% yield). Single crystals suitable for X-ray diffrac-
tion analysis were grown by the slow diffusion of Et2O into a solution of
compound 4a in MeOH. Rf =0.23 (CH2Cl2/MeOH, 97:3); [a]2D5 =À34.3
1
1734 cmÀ1
; HPLC (Chiralpak IA column; CH2Cl2/n-hexane/EtOH,
(c=0.56, MeOH); m.p. 1388C; H NMR (300 MHz, CD3OD): d=8.83 (d,
70:29:1; flow rate: 1.0 mLminÀ1; l=265 nm): t=5.3 (minor diastereom-
er), 10.0 min (major diastereomer); HRMS (ESI): m/z calcd for
C14H17NO3: 270.11061 [M+Na]+; found: 270.1102.
J=6.0 Hz, 1H), 8.58 (dt, J=0.9, 7.8 Hz, 1H), 8.17 (d, J=7.8 Hz, 1H),
8.04 (t, J=6.6 Hz, 1H), 5.47 (br s, 1H), 2.99 (br s, 1H), 1.15 ppm (d, J=
6.0 Hz, 3H); 13C NMR (50 MHz, CD3OD): d=170.4 (Cq), 158.7 (CqAr),
147.9 (CHAr), 137.8 (CHAr), 128.6 (CHAr), 125.9 (CHAr), 104.2 (Cq), 73.5
(CH), 46.6 (CH), 7.9 ppm (CH3); IR (ATR): n˜ =3255, 2754, 1631, 1213,
X-ray crystallography data: CCDC-943762 (4a), CCDC-943765 (4e),
CCDC-943763 (5a), and CCDC-943764 (5b) contain the supplementary
crystallographic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
1038 cmÀ1
; HRMS (ESI): m/z calcd for C10H11NO4Na: 232.0580
[M+Na]+; found: 232.0585.
Preparation of compound 3e: To a solution of 2-ketobutyric acid (2a,
103 mg, 1 mmol, 1 equiv) in CH2Cl2 (3 mL) were added 6-Bromo-2-pyri-
IR and VCD measurements: IR and VCD spectra were recorded on a
ThermoNicolet Nexus 670 FTIR spectrometer that was equipped with a
VCD optical bench.[24] The light beam was focused onto the sample by
using a BaF2 lens (focal length: 191 mm), passed through an optical filter
(which depended on the spectroscopic range to be studied), a BaF2 wire-
grid polarizer (Specac), and a ZnSe photoelastic modulator (Hinds In-
struments, Type II/ZS50). Then, the light was focused by using a ZnSe
lens (focal length: 38.1 mm) onto a HgCdTe detector (1ꢂ1 mm2, Thermo-
Nicolet, MCTA* E6032). IR and VCD spectra were recorded at a resolu-
tion of 4 cmÀ1, by co-adding 50 scans and 24000 scans (acquisition time:
dine carboxaldehyde (1e, 103 mg, 1 mmol, 1 equiv) and catalyst
I
(10 mol%). The reaction mixture was stirred at RT for 3 days under a N2
atmosphere. After concentration in vacuo, purification by column chro-
matography on silica gel (CH2Cl2/MeOH, 98:2) afforded the desired com-
pound (3e) as a yellow oil (239 mg, 89% yield). Rf =0.2 (CH2Cl2/MeOH,
98:2); [a]2D5 =+161 (c=0.84, MeOH); 1H NMR (300 MHz, CD3OD): d=
7.74 (dd, J=7.8, 7.5 Hz, 1H), 7.59 (dd, J=0.6, 7.8 Hz, 1H), 7.36 (dd, J=
0.6, 7.8 Hz, 1H), 5.73 (br d, J=1.5 Hz, 1H), 1.78 ppm (d, J=1.2 Hz, 3H);
13C NMR (150 MHz, CD3OD): d=171.4 (Cq), 158.2 (CqAr), 142.8 (CqAr),
141.4 (CHAr), 139.6 (Cq), 131.4 (Cq), 129.6 (CHAr), 121.6 (CHAr), 83.0
(CH), 9.5 ppm (CH3); IR (ATR): n˜ =1749, 1695, 1206 cmÀ1; HRMS
(ESI): m/z calcd for C10H8BrNO3Na: 291.95852 [M+Na]+; found:
291.9586; HPLC (Chiralpak IA column; n-hexane/iPrOH, 90:10; flow
rate: 1.0 mLminÀ1; l=265 nm): t=11.1 (ent-1), t=12.1 min (ent-2).
8 h), respectively. The sample was held in
a fixed-path-length cell
(100 mm) with BaF2 windows. IR and VCD spectra of compound 4b were
recorded in CD3CN and CD3OD at a concentration of 75 mm. Baseline
corrections of the VCD spectra were performed by subtracting the bare
VCD spectra of the solvents. In all experiments, the photoelastic modula-
tor was adjusted for maximum efficiency at 1400 cmÀ1. Calculations were
performed by using the standard ThermoNicolet software, with Happ
and Genzel apodization, de-Haseth phase-correction, and a zero-filling
factor of one. Calibration spectra were recorded by using a birefringent
plate (CdSe) and a second BaF2 wire-grid polarizer, according to a litera-
ture procedure.[25] Finally, the solvent absorption was subtracted in the
presented IR spectra.
Preparation of compound 6a: To a solution of pyridinium 4a (66.5 mg,
0.32 mmol, 1 equiv) in MeOH (1 mL) was added PtO2 (5 mol%) and the
reaction mixture was pressurized with H2 gas (balloon). The reaction
mixture was stirred at RT for 24 h. After filtration over celite and con-
centration in vacuo, the crude residue was dissolved in DMF (0.5 mL).
Cesium carbonate (0.22 mmol, 1.1 equiv) and benzyl bromide
(0.20 mmol, 1 equiv) were added and the reaction mixture was stirred
overnight. The product was extracted with Et2O and the organic layer
washed with water (3ꢂ25 mL) and dried over sodium sulfate. Concentra-
tion in vacuo, followed by purification by column chromatography on
silica gel (CH2Cl2/MeOH, 99:1), afforded compound 6a as a yellow oil
(35 mg, 45% yield). Rf =0.25 (CH2Cl2/MeOH, 98:2); [a]2D5 =+38 (c=0.95,
MeOH); 1H NMR (300 MHz, CDCl3): d=7.44–7.23 (m, 5H), 5.15 (ABq,
J=12.3 Hz, 2H), 3.67 (dt, J=3.0 Hz, 1H), 3.26 (d, J=10.5 Hz, 1H), 2.94
(br d, J=10.5 Hz, 1H), 2.52–2.11 (m, 4H), 1.91–1.44 (m, 5H), 1.31–1.07
(m, 1H), 0.95 ppm (d, J=7.5 Hz, 3H); 13C NMR (75 MHz, CDCl3): d=
172.5 (Cq), 134.7 (CqAr), 127.6 (CHAr), 127.4 (CHAr), 127.3 (CHAr), 74.3
(CH), 67.7 (CH), 66.2 (CH), 65.5 (CH2), 50.7 (CH2), 38.6 (CH), 24.6
(CH2), 24.5 (CH2), 22.7 (CH2), 9.3 ppm (CH3); IR (ATR): n˜ =3480, 2937,
Computational details: The geometry optimizations, vibrational frequen-
cies, absorptions, and VCD intensities were calculated by using the Gaus-
sian 09 program[26] at the Computer Center “Pꢃle Modꢀlisation” of the
Institut des Sciences Molꢀculaires (University Bordeaux I). To determine
the energy profiles of the reactions, the M06–2X exchange-correlation
functional[27] was used with a standard double-z 6–31+G
ACHTNUTRGNEUNG(d,p) basis set.
A polarizable continuum model (IEFPCM), as implemented in Gaussi-
an 09, was used to describe the medium (MeOH, THF, MeCN). All of
the structures were optimized and frequency calculations were performed
to ensure that there were no imaginary frequencies in the local minima
and only one imaginary frequency in the transition states. The connectivi-
ty between the stationary points was established by intrinsic reaction co-
ordinate (IRC) calculations. Additional calculations of the optimized ge-
ometries of diastereomers 4bcis and 4btrans were performed by using den-
sity functional theory level with the CAM-B3LYP functional[28] and the
aug-cc-pVTZ basis set. Vibrational frequencies and IR and VCD intensi-
ties were calculated at the same level of theory for the isolated molecule
in vacuo, by utilizing the magnetic-field-perturbation method with gauge-
invariant atomic orbitals.[29] Calculations were performed with OH and
OD hydroxyl groups for comparison with the IR and VCD spectra in
CD3CN and CD3OD, respectively. For comparison to the experimental
data, the calculated frequencies were scaled by a factor of 0.965 and the
calculated intensities were converted into Lorentzian bands with a half-
1733 cmÀ1
;
HRMS (ESI): m/z calcd for C17H23NO3Na [M+Na]+
290.17562; found: 290.1757; HPLC (Chiralpak IA column; heptane/
EtOH, 95:5; flow rate: 1.0 mLminÀ1; l=210 nm): t=9.97 min.
Preparation of compound 8b: To a solution of compound 4b (75 mg,
0.34 mmol, 1 equiv) in water (3.3 mL) were added allyl bromide (47 mL,
0.5 mmol, 1.5 equiv) and indium (58 mg, 0.5 mmol, 1.5 equiv). The reac-
tion mixture was stirred at RT for 4 h. Then, the milky white reaction
mixture was concentrated in vacuo. Extraction with CH2Cl2 and washing
of the organic layer with a saturated aqueous solution of NH4Cl, followed
by purification by column chromatography on silica gel (CH2Cl2/MeOH,
99:1) afforded the desired compound (8b) as a colorless oil (32.4 mg,
24% yield, d.r. 9:1). Rf =0.2 (CH2Cl2/MeOH, 98:2); 1H NMR (600 MHz,
CDCl3): d=8.43 (dd, J=1.2, 4.8 Hz, 1H), 7.70 (ddd, J=0.4, 1.2, 8 Hz,
1H), 7.31 (dd, J=4.8, 7.6 Hz, 1H), 6.07–5.95 (m, 0.1H; diastereomer 2
(d2)), 5.94–5.82 (m, 0.9H; diastereomer 1 (d1)), 5.50 (d, J=9.6 Hz, 0.9H;
width of 7 cmÀ1
.
Chem. Eur. J. 2013, 00, 0 – 0
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

F. Robert, J.-M. Vincent, Y. Landais et al.
[13] M. Mehmandoust, C. Marazano, R. Singh, B. Gillet, M. Cesario,
J. L. Fourrey, B. C. Das, Tetrahedron Lett. 1988, 29, 4423.
[14] V. Carelli, F. Liberatore, L. Scipione, B. Di Rienzo, S. Tortorella,
Tetrahedron 2005, 61, 10331.
[15] A. Bruckmann, M. A. Pena, C. Bolm, Synlett 2008, 900.
[16] a) J. A. Bull, J. J. Mousseau, G. Pelletier, A. B. Charrette, Chem.
Rev. 2012, 112, 2642; b) C. Y. Legault, A. B. Charette, J. Am. Chem.
Soc. 2005, 127, 8966.
Acknowledgements
We thank the CNRS and the MENRT for financial support. We grateful-
ly acknowledge Dr. Laurent Chabaud (ICSN, Gif s/Yvette) and Antoine
Jacquet (ISM, u-Bordeaux) for performing the chiral HPLC measure-
ments and Dr. Brice Kaufmann (IECB, u-Bordeaux) for performing the
X-ray diffraction studies.
[17] a) D. Seyferth, H. Menzel, A. W. Dow, T. C. Flood, J. Am. Chem.
Soc. 1968, 90, 1080; b) D. Seyferth, H. Menzel, A. W. Dow, T. C.
Flood, J. Organomet. Chem. 1972, 44, 279.
[18] a) C. W. Jefford, A. F. Boschung, A. B. M. Bolsman, R. M. Moriarty,
B. Melnick, J. Am. Chem. Soc. 1976, 98, 1017; b) L. Ju, A. R. Lip-
pert, J. W. Bode, J. Am. Chem. Soc. 2008, 130, 4253; c) A. R. Lip-
pert, K. R. Keshari, J. Kurhanewicz, C. J. Chang, J. Am. Chem. Soc.
2011, 133, 3776.
[19] P. C. Lobben, L. A. Paquette, J. Org. Chem. 1998, 63, 6990.
[20] The conversion of compounds 4a,b into compounds 8a,b was com-
plete, as shown by 1H NMR spectroscopy. The moderate yields
should be attributed to the high polarity and solubility of com-
pounds 8a,b in water, which resulted in the loss of material.
[21] Cis and trans isomers of pyridinium salts 4 refer to the relative con-
figuration between the OH substituents at C7 and C9 positions. This
terminology is used throughout the discussion.
[22] Additional calculations (data not shown) in MeOH and CH2Cl2 with
the same basis set and solvation model gave similar results.
[23] a) T. B. Freedman, X. Cao, R. K. Dukor, L. A. Nafie, Chirality 2003,
15, 743; b) T. Buffeteau, L. Ducasse, A. Brizard, I. Huc, R. Oda, J.
Phys. Chem. A 2004, 108, 4080; c) T. Buffeteau, D. Cavagnat, A.
Bouchet, T. Brotin, J. Phys. Chem. A 2007, 111, 1045.
[1] a) For
a historical perspective on aminocatalysis, see: B. List,
Angew. Chem. 2010, 122, 1774; Angew. Chem. Int. Ed. 2010, 49,
1730; b) B. List, R. A. Lerner, C. F. Barbas III, J. Am. Chem. Soc.
2000, 122, 2395.
[2] a) G. Guillena in Modern Methods in Stereoselective Aldol Reactions
(Ed.: R. Mahrwald), Wiley-VCH, Weinheim, 2013, Chapter 3,
pp. 155–268, and references therein; b) L. M. Geary, P. G. Hultin,
Tetrahedron: Asymmetry 2009, 20, 131; c) B. M. Trost, C. S. Brindle,
Chem. Soc. Rev. 2010, 39, 1600; d) G. Guillena, D. J. Ramon, C.
Najera in Enantioselective Organocatalyzed Reactions II (Ed.: R.
Mahrwald), Springer, Dordrecht, 2011, pp. 245–342; e) X.-W. Wang,
Y. Wang, J. Jia in Science of Synthesis, Vol. 1 (Eds.: B. List, K. Mar-
uoka), Georg Thieme Verlag, Stuttgart, 2011, pp. 1–34; f) S. M. Yli-
niemelꢄ-Sipari, A. Piisola, P. M. Pihko in Science of Synthesis, Vol. 1
(Eds.: B. List, K. Maruoka), Georg Thieme Verlag, Stuttgart, 2011,
pp. 35–72.
[3] a) E. Lacoste, E. Vaique, M. Berlande, I. Pianet, J.-M. Vincent, Y.
Landais, Eur. J. Org. Chem. 2007, 167; b) E. Lacoste, Y. Landais, K.
Schenk, J.-B. Verlhac, J.-M. Vincent, Tetrahedron Lett. 2004, 45,
8035.
[4] J.-M. Vincent, C. Margottin, M. Berlande, D. Cavagnat, T. Buffe-
teau, Y. Landais, Chem. Commun. 2007, 4782.
[5] a) J. Bigorra, J. Font, O. C. de Echaguen, R. M. Ortuno, Tetrahedron
1993, 49, 6717; b) D. Enders, H. Dyker, F. R. Leusink, Chem. Eur. J.
1998, 4, 311; c) N. Gathergood, K. Juhl, T. B. Poulsen, K. Thordrup,
K. A. Jorgensen, Org. Biomol. Chem. 2004, 2, 1077; d) P. Dambruo-
so, A. Massi, A. Dondoni, Org. Lett. 2005, 7, 4657; e) W. Raimondi,
D. Bonne, J. Rodriguez, Angew. Chem. 2012, 124, 40; Angew. Chem.
Int. Ed. 2012, 51, 40; f) W. Raimondi, D. Bonne, J. Rodriguez,
Chem. Commun. 2012, 48, 6763.
[6] a) T. Namiki, M. Nishikawa, Y. Itoh, I. Uchida, M. Hashimoto, J.
Antibiot. 1987, 40, 1400; b) M. Suzuki, Y. Hosaka, H. Matsushima,
T. Goto, T. Kitamura, K. Kawabe, Cancer Lett. 1999, 138, 121;
c) A. J. Pallenberg, J. D. White, Tetrahedron Lett. 1986, 27, 5591;
d) G. Stork, S. D. Rychnovsky, J. Am. Chem. Soc. 1987, 109, 1564.
[7] B. Zhang, Z. Jiang, X. Zhou, S. Lu, J. Li, Y. Liu, C. Li, Angew.
Chem. Int. Ed. 2012, 51, 13159.
[24] T. Buffeteau, F. Lagugnꢀ-Labarthet, C. Sourisseau, Appl. Spectrosc.
2005, 59, 732.
[25] L. A. Nafie, D. W. Vidrine in Fourier Transform Infrared Spectrosco-
py, Vol. 3 (Eds.: J. R. Ferraro, L. J. Basile), Academic Press, New
York, 1982, pp. 83–123.
[26] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, 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. Son-
nenberg, 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. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. 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, O.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian,
Inc., Wallingford CT, 2009.
[8] For selected recent syntheses of indolizidines, see: a) G. Lapointe,
K. Schenk, P. Renaud, Chem. Eur. J. 2011, 17, 3207; b) S. V. Pronin,
G. Tabor, D. J. Jansen, R. A. Shenvi, J. Am. Chem. Soc. 2012, 134,
2012; c) X.-G. Hu, B. Bartholomew, R. J. Nash, F. X. Wilson,
G. W. J. Fleet, S. Nakagawa, A. Kato, Y.-M. Jia, R. van Well, C.-Y.
Yu, Org. Lett. 2010, 12, 2562.
[9] D. Giomi, R. Alfini, A. Micoli, E. Calamai, C. Faggi, A. Brandi, J.
Org. Chem. 2011, 76, 9536.
[10] Compounds 4 f and 4h were obtained as a mixture of two diaster-
eomers at the C9 position.
[27] a) Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215; b) Y.
Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 119, 525; c) Y. Zhao,
D. G. Truhlar, Acc. Chem. Res. 2008, 41, 157.
[28] T. Yanai, D. Tew, N. Handy, Chem. Phys. Lett. 2004, 393, 51.
[29] J. R. Cheeseman, M. J. Frisch, F. J. Delvin, P. J. Stephens, Chem.
Phys. Lett. 1996, 252, 211.
[11] M. J. Walters, E. J. Toone, Nat. Nanotechnol. Nat. Protocols 2007, 2,
1825.
Received: June 14, 2013
[12] S. Bahmanyar, K. N. Houk, J. Am. Chem. Soc. 2001, 123, 12911.
Published online: && &&, 0000
&
8
&
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Straightforward Route to Indolizidines and Isotetronic Acids
FULL PAPER
Organocatalysis
V. Liautard, D. Jardel, C. Davies,
M. Berlande, T. Buffeteau,
D. Cavagnat, F. Robert,*
J.-M. Vincent,* Y. Landais* &&&& — &&&&
BIPpity boppity boo: Asymmetric
ACHTUNGTRENNiUNG midazole pyrrolidine (BIP) as a cata-
aldol reactions between pyridine carb-
ACHTUNGTRENNUNGaldehydes and a-ketoacids were shown
to provide isotetronic acids or their
corresponding pyridinium salts with
high enantioselectivities by using benz-
lyst. Hydrogenation of the resulting
pyridinium salts led to new indol-
ACHTUNGTRENNUNG
Organocatalyzed Aldol Reaction
Between Pyridine-2-carbaldehydes and
a-Ketoacids: A Straightforward Route
towards Indolizidines and Isotetronic
Acids
Chem. Eur. J. 2013, 00, 0 – 0
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!