Remote substituent effects on the stereoselectivity and organocatalytic activity of densely substituted unnatural proline esters in aldol reactions

Article abstract of DOI:10.1002/ejoc.201500160

The organocatalytic activities of highly substituted proline esters obtained through asymmetric [3+2] cycloadditions of azomethine ylides derived from glycine iminoesters have been analyzed by ¹⁹F NMR and through kinetic isotope effects. Kineti

Full text of DOI:10.1002/ejoc.201500160

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
FULL PAPER
DOI: 10.1002/ejoc.201500160
Remote Substituent Effects on the Stereoselectivity and Organocatalytic
Activity of Densely Substituted Unnatural Proline Esters in Aldol Reactions
María de Gracia Retamosa,[a] Abel de Cózar,[a,b] Mirian Sánchez,[a] José I. Miranda,[c]
José M. Sansano,[d] Luis M. Castelló,[d] Carmen Nájera,[d] Ana I. Jiménez,[e]
Francisco J. Sayago,[e] Carlos Cativiela,[e] and Fernando P. Cossío*[a]
Keywords: Organocatalysis / Aldol reactions / Kinetics / Reaction mechanisms / Transition states
The organocatalytic activities of highly substituted proline
esters obtained through asymmetric [3+2] cycloadditions of
azomethine ylides derived from glycine iminoesters have
been analyzed by 19F NMR and through kinetic isotope ef-
fects. Kinetic rate constants have been determined for unnat-
ural proline esters incorporating different substituents. It has
esters yield opposite enantiomers in aldol reactions between
cyclic ketones and aromatic aldehydes. The combined results
reported in this study show subtle and remote effects that
determine the organocatalytic behavior of these synthetic but
readily available amino acid derivatives. These data can be
used as design criteria for the development of new pyrrol-
idine-based organocatalysts.
been found that exo-L and endo-L unnatural proline methyl
well-defined and readily identifiable category. Thus, dif-
ferent organocatalysts have been tested in aldol reactions;[9]
they include natural amino acids[10] such as histidine,[11] ser-
ine,[12] threonine,[12,13] tryptophan,[12,14] alanine,[15] and iso-
leucine.[16] Of these natural amino acids, l-Pro (and its de-
rivatives) has emerged as the most widely used organocata-
lysts for the aldol reaction. Trost and Brindle,[17] in their
comprehensive review published in 2010, reported around
45 l-Pro derivatives that are able to catalyze aldol reactions.
It is relevant to note that in all these cases the stereochemis-
try of the major aldol adducts was found to be the same:
namely, that associated with the (2S,1ЈR) configuration in
the new chiral carbon atoms (Scheme 1).
Introduction
The aldol reaction is one of the most important reactions
for the convergent generation of C–C bonds.[1] This trans-
formation is operative in biochemical processes, type I and
II aldolases being the main representative enzymes.[2] Al-
though it has been known since 1971–1974 that l-proline
(l-Pro) can catalyze intramolecular aldol reactions,[3,4] the
2000 report by Barbas and List[5,6] of intermolecular aldol
reactions catalyzed by l-Pro prompted an impressive up-
surge of work on this reaction.[7] The term organocatalysis,
explicitly mentioned by MacMillan[8] in a paper also pub-
lished in 2000, contributed to placing these reactions in a
[a] Departamento de Química Orgánica I / Kimika Organikoa I
Saila, and Centro de Innovación en Quíimica Avanzada
(ORFEO-CINQA), Facultad de Química / Kimika Fakultatea,
Universidad del País Vasco / Euskal Herriko Unibertsitatea
(UPV/EHU), and Donostia International Physics Center
(DIPC),
P. K. 1072, 20018 San Sebastián - Donostia, Spain
E-mail: fp.cossio@ehu.es
http://www.ehu.eus/eu/web/qbmm/hasiera
[b] IKERBASQUE, Basque Foundation for Science,
48013 Bilbao, Spain
Scheme 1. Stereochemical outcome observed in aldol reactions cat-
alyzed by l-Pro derivatives.
[c] SGIker NMR Facility,
20018 San Sebastián - Donostia, Spain
[d] Departamento de Química Orgánica e Instituto de Síntesis
Orgánica (ISO), and Centro de Innovación en Química
Avanzada (ORFEO-CINQA). Facultad de Ciencias,
Universidad de Alicante,
The mechanism and origin of the stereocontrol of l-Pro-
catalyzed aldol reactions have been studied both experimen-
tally[18] and computationally.[19] The generally accepted
mechanism is shown in Figure 1. According to this catalytic
cycle, the first stage of the reaction involves the formation
of enamine intermediates of type INT1, a process that can
be promoted by acidic protons present in the starting amino
03080 Alicante, Spain
[e] Departamento de Química Orgánica, Instituto de Síntesis
Química y Catálisis Homogénea (ISQCH), CSIC-Universidad
de Zaragoza,
50009 Zaragoza, Spain
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201500160.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
F. P. Cossío et al.
FULL PAPER
acid or in the reaction medium.[20] These intermediates have
To the best of our knowledge, no structure–activity rela-
been characterized spectroscopically[21] and by X-ray dif- tionship studies on the organocatalytic properties of syn-
fraction analysis.[22]
thetic densely substituted l-proline derivatives have been re-
ported to date. These compounds show great promise in the
field and are easily accessible because the 2-alkoxycarb-
onylpyrrolidine scaffold can be formed in a single step
through [3+2] cycloaddition between readily available π-de-
ficient alkenes and azomethine ylides generated in situ.[25]
Recently, we have reported new ferrocenyl-containing pro-
line catalysts that are able to yield both endo- and exo-l-
4-nitroproline esters with high enantiocontrol.[26] We also
showed that the exo cycloadducts are able to catalyze the
aldol reaction efficiently.[26] Here we compare the perform-
ances of l-4-nitroproline esters and acids featuring different
substitution patterns and stereochemical configurations as
organocatalysts in the aldol reaction. The ultimate goal of
this research is to provide design criteria for organocatalysts
based on unnatural proline esters obtained directly through
fully stereocontrolled [3+2] cycloadditions.
Results and Discussion
Figure 1. General catalytic cycle of aldol reactions catalyzed by l-
proline derivatives. Unless otherwise indicated, the possible substit-
uents at the different positions have not been specified.
Preparation of Unnatural Proline Organocatalysts
We prepared different l-4-nitroproline esters through
asymmetric 1,3-dipolar cycloaddition reactions between
different dipolarophiles and azomethine ylides formed in
situ from the corresponding imines (Scheme 3 and
Scheme 4). The reactions were catalyzed by enantiopure 5-
ferrocenyl-4-nitroprolines l-5a and d-5b or phosphoramid-
ite (Sa-R,R)-6, previously developed and tested as catalysts
for reactions of this type by the San Sebastián-Donostia[26]
and Alicante[27] groups, respectively (Scheme 2).
Formation of the new C–C bond takes place during the
second stage of the reaction (Figure 1), giving rise to imin-
ium intermediates of type INT2. Once again, when l-Pro is
used, the carboxy group provides the proton required to
enhance the electrophilicity of a carbonyl compound 2. The
overall arrangement of the transition structures of type
TScc associated with this stage, usually referred to as the
Houk–List model,[23] shows a chair-like shape in which the
Zimmerman–Traxler[24] arrangement is extended to form a
cyclic transition structure involving the carboxy group of
the organocatalyst. Finally, the iminium intermediate of
type INT2 is hydrolyzed to release the aldol 3 and the or-
ganocatalyst 4, thus completing the catalytic cycle (Fig-
ure 1). It is important to note, however, that the Houk–
List model relies on the presence of an acidic residue in the
organocatalyst. When other functional groups such as ester,
amido, amino, etc.[7f,7g] are present instead of the carboxylic
acid, acidic additives or protic substituents are usually re-
quired. Although the origins of stereocontrol in these cases
are far from being fully understood, computational studies
on selected systems[19a] have shown that transition struc-
tures closely related to the Houk–List model are compatible
with the observed stereochemical outcomes.
Scheme 2. Catalysts used in the synthesis of unnatural l-prolines
4.
According to this conceptual framework, the stereo-
chemistry of the reaction is determined by the l configura-
tion of the amino acid and the distal (anti) conformation
of the enamine double bond with respect to the carboxy
By use of glycine iminoesters 7a–f as precursors of azo-
group.[23] Therefore, it is not surprising that l-Pro deriva- methine ylides, nitroalkenes 8a–c as dipolarophiles, and l-
tives yield the same enantiomers in reactions of this kind.
5a, d-5b, or (Sa-R,R)-6 as catalysts, endo- and exo-l-4-nitro-
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
Organocatalytic Aldol Activity of Unnatural Proline Esters
proline esters 4aa–ec were obtained (Scheme 3). In all cases,
good yields and high optical purities were achieved.[26]
When R1 was a secondary alkyl group, such as cyclohexyl,
the desired cycloadducts were not formed. However, in the
case of R1 = tert-butyl, exo-l-4fa was obtained with high
enantiomeric excess. When N-methylmaleimide (9d) was
used as the dipolarophile in the presence of d-5b as a cata-
lyst, endo-l-4ad was isolated (Scheme 4). Unfortunately, im-
ine 7h, derived from l-phenylalanine methyl ester, gave poor
enantiomeric excesses in the presence of either l-5a or d-
5b. However, a catalytic amount of phosphoramidite (Sa-
R,R)-6 promoted the formation of exo-l-4ha as the sole
enantiomer (Scheme 4).[27] This catalyst was also effective
in the preparation of exo-l-4ia from imine 7i. Therefore,
our previously reported methodologies,[26,27] provided a
convenient route to cycloadducts endo- and exo-l-4aa–ga
as enantiopure compounds in high yields from easily or
commercially available reactants (7,[28] 8, and 9) in the pres-
ence of readily accessible catalysts (Scheme 2). Such cy-
cloadducts were isolated in useful quantities for subsequent
evaluation as organocatalysts.
Scheme 4. Synthesis of unnatural l-proline esters endo-l-4ad, exo-
l-4ha, exo-l-4ia, and exoЈ-l-4aa. Reagents and conditions:
(a) Cu(CH3CN)4PF6 (3 mol-%), d-5b (3 mol-%), NEt3 (3 mol-%),
THF, r.t.; (b) Cu(OTf)2, (5 mol-%), (Sa-R,R)-6 (5 mol-%), NEt3
(5 mol-%), toluene, r.t.; (c) AgOAc (15 mol-%), NEt3 (15 mol-%),
CH3CN, r.t.; (d) column chromatography on silica gel to isolate
exoЈ-rac-4aa; (e) preparative HPLC resolution on a chiral column
(Chiralpak® IB); (f) 4-bromobenzoyl chloride (2.0 equiv.), DIPEA
(2.6 equiv.), CHCl3, r.t.
With the aim of further increasing the stereochemical di-
versity of the l-4-nitroproline cycloadducts for testing as
organocatalysts, the [3+2] cycloaddition reaction starting
from imine 7a and β-nitrostyrene (8a) was carried out in
the presence of silver acetate and triethylamine. In this case,
a mixture of racemic endo-, exo-, and exoЈ-rac-4aa cycload-
ducts was obtained (Scheme 4), from which the minor iso-
mer exoЈ-rac-4aa was separated by column chromatography
on silica gel. This racemic compound was then subjected to
preparative resolution by HPLC with a chiral column,
which allowed the isolation of exoЈ-l-4aa and exoЈ-d-4aa in
optically pure form (see the Supporting Information). As-
signment of the absolute configurations of the resolved
enantiomers was carried out by transformation of the first-
eluting enantiomer (exoЈ-l-4aa) into the corresponding N-
(4-bromobenzoyl) derivative exoЈ-l-10aa (Scheme 4) and
resolution of the crystalline structure of this compound by
X-ray diffraction analysis.[29]
Scheme 3. Synthesis of unnatural l-proline derivatives endo-l-4aa
and exo-l-4aa–ec. Reagents and conditions: (a) Cu(CH3CN)4PF6
(3 mol-%), NEt3 (5 mol-%), THF; (b) l-5a (3 mol-%), –60 °C; (c) d-
5b (3 mol-%), –20 °C; (d) Cu(OTf)2, (5 mol-%), (Sa-R,R)-6 (5 mol-
%), NEt3 (5 mol-%), toluene, r.t. (e) NaOH, H2O/acetone, r.t.
(f) TFA, CH2Cl2, r.t.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
F. P. Cossío et al.
FULL PAPER
Kinetic Studies
In order to assess the organocatalytic activity of the syn-
thesized unnatural proline esters l-4, we chose as a case
study the aldol reaction between cyclohexanone (1a) and
pentafluorobenzaldehyde (2a). In principle, four isomeric
adducts 3aa can be expected for this reaction (Scheme 5).
Scheme 5. Model aldol reaction to assess the organocatalytic ac-
tivity of cycloadducts 4. Descriptors o, m and p denote ortho, meta
and para positions with respect to the carbaldehyde group.
Preliminary studies[26] showed that 2a is quite electro-
philic, and therefore the progress of the aldol reaction in
the presence of different organocatalysts can be monitored
in relatively short reaction times. In addition, the presence
of fluorine atoms in ortho, meta, and para positions with
respect to the carbaldehyde group (Figure 2) allows moni-
toring of the evolution of the reaction mixtures by 19F
NMR spectroscopy. The different signals were assigned by
Figure 2. (A) 19F NMR spectra obtained for the aldol reaction be-
tween 1a and 2a to form adducts 3aa, catalyzed by exo-l-4aa. The
hollow arrows show the decay of the signals corresponding to 2a.
These decays were used to estimate the kinetic constants. (B) Three
selected spectra showing the evolution of reactants, products, and
iminium intermediate INT2a (Figure 1).
3
4
measuring the JF,F and JF,F coupling constants (see the
Supporting Information). Figure 2 (A) shows the spectra
registered for the reaction between 1a and 2a in the pres-
ence of exo-l-4aa with a catalytic load of 30 mol-% at
298 K. Together with the consumption of the starting alde-
hyde 2a, formation of the major and minor adducts anti-
3aa and syn-3aa, respectively, can be observed. Moreover,
transient formation of iminium intermediates INT2a
(Scheme 1) can be detected under these conditions (Fig-
ure 2).
with the broad signal corresponding to exo-l-4ea (Fig-
ure 3). These observations are compatible with the catalytic
cycle depicted in Figure 1, because both enamine and imin-
ium aldol intermediates were detected, the latter being more
persistent. This suggests that, most likely, the hydrolytic
step (stage III, Figure 1) is kinetically slower than the for-
mation of the nucleophilic enamine.
In order to monitor the presence of enamine intermedi-
With this information, we undertook the determination
ates of type INT1 (Scheme 1) by this method, at least one of the kinetic constants associated with the 1a + 2aǞ3aa
fluorine atom must be present in the catalyst. We therefore reaction in the presence of different organocatalysts. Be-
studied the same reaction in the presence of exo-l-4ea (Fig- cause the previously discussed 19F NMR studies had sug-
ure 3). In this case, aside from the ensemble of signals be- gested a kinetically complex profile, we decided to work un-
tween –140 and –165 ppm associated with the C6F5 moiety der pseudo-first order conditions, with a 1a/2a ratio of 60:1.
(Figure 2), both enamine intermediates INT1b and INT2b The apparent reaction rates were measured according to
were detected between –113 ppm and –115 ppm, together Equation (1).
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
Organocatalytic Aldol Activity of Unnatural Proline Esters
Figure 3. 19F NMR spectra associated with the aldol reaction be-
tween 1a and 2a catalyzed by exo-l-4ea, showing the evolution of
the catalyst, enamine intermediate INT1b, and iminium intermedi-
ate INT2b.
(1)
As internal reference we used the signal corresponding
to trifluoroacetic acid (TFA). For each signal associated
with ortho, meta, and para fluorine atoms of 2a, we moni-
tored the decay of the corresponding integrals (Figure 2, A)
by means of Equation (2)
Figure 4. (A) Levels of conversion over time associated with the 1a
+ 2aǞ3aa transformation in the presence of organocatalysts exo-
l-4aa, endo-l-4aa, exoЈ-l-4aa, and endo-l-4ad. All reactions were
monitored under pseudo-first order conditions at 25 °C by 19F
NMR (see text). (B) Pseudo-first order linear plots obtained for the
organocatalyzed reactions shown in (A). These plots were used to
determine kobs (see text).
(2)
where Iti and I0i are the instant and initial integrals, respec-
tively, of the i-fluorine atom of 2a. The experimental mea-
surements were averaged for the three sets of signals associ-
ated with the ortho, meta, and para fluorine atoms of 2a.
Figure 4 (A) shows the conversion of 1a and 2a into 3aa in
the presence of organocatalysts exo-l-4aa, endo-l-4aa, and 4aa results in a significantly lower kinetic constant (Table 1,
exoЈ-l-4aa. Figure 4 (B) shows the linear plot obtained after Entry 4).
application of Equation (2). Similar excellent correlations
In order to confirm this result with an endo-cycloadduct
were obtained in the remaining cases, thus showing pseudo- possessing cis substituents at C3 and C4, we measured the
first order behavior of the studied reactions (see the Sup- kobs value for endo-l-4ad and also observed a relatively low
porting Information). The kinetic constants calculated from kinetic constant (Table 1, Entry 5). Therefore, we conclude
treatment of NMR spectroscopic data by application of that a cis relationship between the substituents at C3 and
Equation (2) are collected in Table 1.
C4 results in a lower organocatalytic activity. It is also note-
Our results indicate that, for the different l-4aa stereoiso- worthy that exo-l-4aa retains a significantly high kobs value
mers tested, the catalytic activity order is exo Ͼ endo ϾϾ with a catalytic load of 5 mol-% (Table 1, Entry 2).
exoЈ. Therefore, although the configurations of the C2 and
In order to assess the role of TFA in general or specific
C5 positions contiguous to the N1 active site of the pyrrol- co-catalytic schemes[30] we measured the values of kobs for
idine ring are identical in the three cases, the stereochemis- the 1a + 2aǞ3aa reaction catalyzed by exo-l-4aa in the
try at the distal C3 and C4 atoms has a considerable effect presence of different TFA concentrations. The results are
on the rate of the aldol reaction. In particular, the relative collected in Table 2, in which we have included the kobs
cis configuration of the phenyl and nitro groups in exoЈ-l- value in the absence of acid. This value (Table 2, Entry 1)
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
F. P. Cossío et al.
FULL PAPER
Table 1. Measured catalytic constants (kobs) for cycloadducts exo- 2a). For higher values, the measured kobs values evolve
l-4aa, endo-l-4aa, exoЈ-l-4aa, and endo-l-4ad in the 1a + 2aǞ3aa
towards a saturation profile.
reaction.[a,b]
We reasoned that the behavior reported in Table 2 could
be described by Equation (3).
A[TFA]
kobs = k0 +
(3)
B + [TFA]
In Equation (3), k0 is the kobs value for [TFA] = 0
(Table 2, Entry 1), and A and B are two constants that can
be determined from the experimental data reported in
Table 2. For low values of [TFA] (i.e., for [TFA] ϽϽ B),
Equation (3) can be approximated as kobs – k0 ≈ (A/B)-
[TFA], thus resulting in linear behavior. On the other hand,
if [TFA] ϾϾ B, then Equation (3) approaches to a zero-
slope profile in the form kobs – k0 ≈ A, which would be
associated with a maximum kobs – k0 value.
A double reciprocal treatment[31] of the data in Table 2
revealed a good linear regression (r2 = 0.996) with A =
47.40ϫ10–4 and B = 95.10, as can be seen in Figure 5. This
result confirms that the hypothesis proposed in Equa-
tion (3) is reasonable for the concentrations tested in our
experiments. A possible interpretation of these results is
that, despite the complex, multistep mechanism required for
the catalysis of these aldol reactions, at catalytic concentra-
tions the proton transfer to the oxygen atom of the alde-
hyde is involved in the rate-limiting step. At higher concen-
trations of TFA, protonation of the aldehyde occurs prior
to the step associated with the formation of the C–C bond,
the protonated aldehyde being the effective electrophile.
[a] Pseudo-first order constants calculated by application of Equa-
tion (2) with a 1a/2a ratio of 60:1. [b] All reactions were monitored
by 19F NMR at 25 °C.
was found to be two to three orders of magnitude lower
than those observed when TFA was added (Table 2, En-
tries 4–11). Other carboxylic acids such as acetic acid and
4-nitrobenzoic acid promoted smaller enhancements of the
kobs values. Therefore, TFA proved to be a convenient acidic
additive for this reaction. In addition, according to our ex-
periments with variable concentrations of TFA, a roughly
linear behavior of kobs with respect to [TFA] is observed for
acid concentrations within the catalytic range (i.e., close to
or lower than 30 mol-% with respect to the limiting reactant
Table 2. Pseudo-first order experimentally measured[a,b] (kobs) ki-
netic constants measured for the 1a + 2aǞ3aa reaction catalyzed
by exo-l-4aa with different acids and concentrations.
Entry
Acid
Mol-%
kobs (ϫ10–4 s–1)
Figure 5. Double reciprocal treatment of the pseudo-first order
constants (kobs) reported in Table 2.
1
2
3
4
5
6
7
8
none
AcOH
PNB[c]
TFA
TFA
TFA
TFA
TFA
TFA
TFA
TFA
TFA
0
0.0429 (Ϯ0.0004)
0.551 (Ϯ0.016)
0.388 (Ϯ0.003)
6.37 (Ϯ0.14)
30
30
15
20
25
30
50
60
75
100
150
Kinetic isotope effects (KIEs) were measured for the 1a
+ 2aǞ3aa reaction catalyzed by exo-l-4aa. The results are
collected in Table 3. When the reaction was performed in
the presence of α,α,αЈ,αЈ-tetradeuterated cyclohexanone 1a-
d4, a KIE of 2.68 was observed (Table 3, Entry 1). This
value is between the KIE reported for the nornicotin-cata-
lyzed aldol reaction between acetone and 4-nitrobenzalde-
hyde, for which Janda et al.[32] measured a KIE of 3.0 in the
presence of D2O, and the KIE ≈ 2 found by Blackmond
et al.[33] for the l-Pro-catalyzed reaction between [D6]acet-
one and 3-chlorobenzaldehyde.
8.60 (Ϯ0.34)
10.00 (Ϯ0.29)
11.29 (Ϯ0.30)
15.73 (Ϯ0.35)
18.44 (Ϯ0.14)
21.88 (Ϯ0.19)
25.53 (Ϯ0.76)
26.75 (Ϯ0.06)
9
10
11
12
[a] Pseudo-first order constants calculated by application of Equa-
tion (2) with a 1a/2a ratio of 60:1. [b] All reactions were monitored
by 19F NMR at 25 °C. [c] PNB: p-nitrobenzoic acid.
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
Organocatalytic Aldol Activity of Unnatural Proline Esters
Table 3. Kinetic isotope experiments conducted[a] for the 1a + 5), the measured ee values were slightly lower than those
2aǞ3aa reaction catalyzed by cycloadduct exo-l-4aa.
obtained with TFA, thus confirming the superior perform-
ance of this last additive, not only in terms of kinetic accel-
eration (vide supra), but also in terms of stereocontrol.
Cycloadducts endo-l-4aa and endo-l-4ad gave good syn/anti
diastereocontrol and chemical yields (Table 4, Entries 6, 7,
and 9). Surprisingly, the sense of chiral induction of endo-
l-4aa, exoЈ-l-4aa, and endo-l-4ad was found to be the op-
posite of that observed for exo-l-4aa. This unexpected re-
sult is discussed later on. Finally, it is worth noting that
exoЈ-l-4aa and endo-l-4ad, with cis substituents at C3 and
[b]
C4, give poor enantiocontrol, the former also showing a
low syn/anti diastereoselectivity (Table 4, Entries 8 and 9).
Therefore, these two compounds are not convenient in
terms of organocatalytic activity (vide supra) and ste-
reocontrol.
Entry
X
Y
Z
kobs (ϫ10–4 s–1) kH/kD
1
2
3
4
5
D
H
H
H
D
H
D
H
D
D
H
H
D
D
D
4.21 (Ϯ0.29)
7.61 (Ϯ0.22)
11.59 (Ϯ0.22)
7.36 (Ϯ0.51)
3.16 (Ϯ0.24)
2.68 (Ϯ0.20)
1.48 (Ϯ0.06)
0.97 (Ϯ0.03)
1.53 (Ϯ0.11)
3.57 (Ϯ0.29)
Table 4. Stereocontrol and chemical yields observed[a] in the 1a +
2aǞ3aa reaction catalyzed by cycloadducts exo-l-4aa, endo-l-4aa,
exoЈ-l-4aa, and endo-l-4ad in the presence of TFA (30 mol-%).
[a] All reactions were conducted in neat 1a at 25 °C with a 1a/2a
ratio of 60:1. Reference value: kH = 11.29ϫ10–4 (Ϯ0.30) (see
Table 1). [b] Errors were calculated from standard deviations of the
rate constants according to ref.[34]
Entry Cat.
T
Cat. load anti/syn[b] Yield[c]
ee[d]
[°C]
[mol-%]
[%]
[%]
1[e]
2
exo-l-4aa
25
25
–15
25
25
25
–15
25
25
30
5
95:05
95:05
99:01
97:03
96:04
96:04
99:01
42:58
92:08
73
81
70
73
69
83
80
75
85
89[d]
89
92
81
78
–81
–84
–24
–47
In the presence of N-deuterated exo-l-4aa, a significant
normal KIE was detected (Table 3, Entry 2). When the
acidic position of TFA was deuterated (Table 3, Entry 3)
we were unable to measure any noticeable KIE within the
experimental error. In line with these two results, deutera-
tion of the acidic positions of both exo-l-4aa and TFA re-
sulted in a normal KIE similar to that found when only the
NH-site of exo-l-4aa was deuterated (Table 3, Entries 2 and
4). Finally, deuteration at the four α-positions of 1a and of
the acidic sites in exo-l-4aa and TFA resulted in a KIE of
3.57 (Table 3, Entry 5), a value larger than that measured
in the presence of 1a-d4.
3[e]
4[f]
5[g]
6
30
30
30
30
30
30
30
endo-l-4aa
7
8
9
exoЈ-l-4aa
endo-l-4ad
[a] Determined on crude reaction mixtures after Ͼ99% conversion.
[b] Determined by 1H or 19F NMR on crude reaction mixtures.
[c] Isolated yields of anti and syn aldols after workup. [d] Deter-
mined by HPLC with chiral stationary phases and computed as ee
= 100ϫ[(2R,1ЈS) – (2S,1ЈR)]/[(2R,1ЈS)+(2S,1ЈR)]. [e] Results re-
ported in ref.[26] [f] Acetic acid was used as additive. [g] 4-Nitro-
benzoic acid was used as additive.
In summary, from the experiments reported in Table 3
we conclude that the observed KIE for the deuterated nu-
cleophile is the outcome of a tradeoff between the large
We also investigated the effect of including para-substitu-
normal primary KIE associated with the formation of the ents on the 3,5-diphenyl groups of exo-l-4aa. To this end,
enamine intermediate (Figure 1, stage I), the inverse sec- the behavior of cycloadducts exo-l-4ea–dc was analyzed.
ondary KIE expected for the C–C bond-forming step (Fig- The results are reported in Table 5. All these compounds
ure 1, stage II), and the normal primary KIE associated provided satisfactory chemical yields and anti/syn diastereo-
with the hydrolysis of the iminium aldol adduct (Figure 1, selectivities. Interestingly enough, these effects do not seem
stage III), in the presence of enolization equilibria under to be additive, because neither two fluorine atoms (Table 5,
general acid catalysis conditions within the 30 mol-% range Entry 5) nor one methoxy and one fluorine substituent
(see Table 2). We can also conclude that stage II, associated (Table 5, Entry 6) produced kobs values higher than those
with the formation of the C–C bond and hence with the obtained with the monosubstituted cycloadducts. Despite
stereocontrol of the aldol reaction, cannot be considered the modest acceleration observed with respect to exo-l-4aa,
the rate-limiting step.
The diastereo- and enantioselectivities observed for the slightly lower than those found for the parent catalyst.
1a + 2aǞ3aa reaction catalyzed by the three l-4aa cyclo- We next explored other substitution patterns in cyclo-
the measured enantiomeric excesses were found to be
adducts were also measured, and the results are included in adducts exo-l-4. In Table 6 we report the results obtained
Table 4. In the case of exo-l-4aa, an excellent anti selectivity with respect to l-Pro. We observed that the free amino acid
was observed. A catalytic load of 5 mol-% was not detri- exo-l-4Јaa, obtained by hydrolysis of methyl ester exo-l-
mental in terms of stereocontrol and chemical yield 4aa (Scheme 3), retains the anti/syn diastereoselectivity and
(Table 4, Entries 1 and 2). In addition, the enantiomeric ex- the sense of enantioselection, although with a slightly lower
cess (ee) was slightly better at –15 °C (Table 4, Entry 3). enantiomeric excess and chemical yield (Table 6, Entry 3).
When acetic acid and 4-nitrobenzoic acids were used as ad- In the case of endo-l-4Јaa (Table 6, Entry 3), the ee has the
ditives in the presence of exo-l-4aa (Table 4, Entries 4 and same sign as observed with the naturally occurring l-Pro
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
F. P. Cossío et al.
FULL PAPER
Table 5. Pseudo-first[a] order kinetic constants (kobs), chemical
yields,[b] and enantiomeric excesses (ee)[c] observed in reaction 1a
+ 2aǞ3aa catalyzed by compounds exo-l-4ea–dc.
Table 6. Stereocontrol and chemical yields observed[a] in the 1a +
2aǞ3aa reaction catalyzed by l-Pro and by substituted exo-l-4
cycloadducts (30 mol-%) in the presence of TFA (30 mol-%).
Entry Cat.
R1
R2
Yield [%] ee [%] kobs (ϫ10–4 s–1)
1
2
3
4
5
6
exo-l-4ea
F
H
OMe
H
F
H
F
H
OMe
F
80
72
68
78
88
79
78
83
77
87
81
76
6.23 (Ϯ0.19)
10.18 (Ϯ0.36)
11.28 (Ϯ0.35)
14.74 (Ϯ0.78)
5.59 (Ϯ0.08)
8.64 (Ϯ0.35)
exo-l-4ac
exo-l-4da
exo-l-4ab
exo-l-4ec
exo-l-4dc
OMe
F
[a] All reactions were monitored at 25 °C with a 1a/2a ratio of 60:1
until conversion Ͼ99%. In each case the anti/syn ratio was about
95:5. [b] Isolated yields of 3aa. [c] Enantiomeric excesses corre-
sponding to the major anti diastereomer were determined by HPLC
and computed as ee = 100ϫ[(2R,1ЈS) – (2S,1ЈR)]/[(2R,1ЈS) +
(2S,1ЈR)].
(Table 6, Entry 1). This change in the sense of chiral induc-
tion is discussed in more detail below. The presence of an
alkoxy group bulkier than a methoxy group (compound
exo-l-4ba, Table 6, Entry 4) also resulted in a lower ee and
chemical yield. This detrimental effect was more pro-
nounced in the case of a 2-naphthyl group at C5 (com-
pound exo-l-4ca, Table 6, Entry 5). Finally, we found that
either a quaternary group at C2 or a tert-butyl group at C5
(compounds exo-l-4ha, exo-l-4ia, and exo-l-4fa) resulted
in an almost complete loss of organocatalytic activity
(Table 6, Entries 6–9). Therefore, the substitution patterns
included in Entries 4–9 of Table 6 should be avoided in the
development of other catalysts based on this scaffold.
We also tested the stereoselectivity and catalytic effi-
ciency of selected compounds 4 in the absence of any acid
additive. The results are reported in Table 7 and indicate
that under these conditions the reaction times required for
esters exo- and endo-l-4aa are much longer (Entries 1 and
2). In addition, the ee values are considerably lower, al-
though the sense of induction was that obtained in the pres-
ence of TFA or other acids. In the case of acids exo-l-4Јaa
and endo-l-4Јaa (Entries 3 and 4) the reaction times were
considerably shorter than those observed with the corre-
sponding esters. In addition, the ee values were also lower
than those achieved with the corresponding esters in the
presence of TFA (Table 6, Entries 2 and 3). However, it is
noteworthy that the sense of chiral induction of these com-
[a] All reactions were monitored for 24 h, with a 1a/2a ratio of 60:1.
[b] Isolated yields of 3aa. [c] Enantiomeric excesses corresponding
to the anti diastereomer were determined by HPLC on crude reac-
tion mixtures and computed as ee = 100ϫ[(2R,1ЈS) – (2S,1ЈR)]/
[(2R,1ЈS) + (2S,1ЈR)]. [d] Observed levels of conversion after 48 h.
pounds is the same as observed for the corresponding esters most convenient reaction conditions correspond to the un-
in the presence of TFA. These results confirmed that the natural proline esters 4 with TFA (30 mol-%), in terms of
8
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
Organocatalytic Aldol Activity of Unnatural Proline Esters
stereocontrol, reaction rate, and preparative simplicity of 2aǞ3aa reaction, with 92% ee at –15 °C. In contrast, endo-
the synthetic organocatalysts.
l-4aa yields (2S,1ЈR)-3aa as the major adduct, with 84% ee
under the same reaction conditions. The latter adduct is
the enantiomer obtained in the presence of natural l-Pro
(Table 6, Entry 1). In order to explore the persistence of this
enantiodivergent behavior for exo-l-4aa and endo-l-4aa, we
studied different aldol reactions involving ketones 1a–c and
aldehydes 2a–f (Scheme 6). The results obtained are col-
lected in Table 7.
Table 7. Stereocontrol and chemical yields observed[a] in the 1a +
2aǞ3aa reaction catalyzed by substituted exo- and endo-l-4aa,
exo-l-4Јaa, and endo-l-4Јaa cycloadducts without acidic additives.
Scheme 6. Aldol reactions between ketones 1a–c and aldehydes 2a–
f catalyzed by endo-l-4aa and exo-l-4aa. Reagents and conditions:
(a) organocatalyst (30 mol-%), TFA (30 mol-%), –15 or +25 °C,
24 h (see Table 7).
[a] All reactions were monitored at 25 °C with a 1a/2a ratio of 60:1
until 99% conversion. [b] Isolated yields of 3aa. [c] Enantiomeric
excesses corresponding to the anti diastereomer were determined
by HPLC on crude reaction mixtures and computed as ee =
100ϫ[(2R,1ЈS) – (2S,1ЈR)]/[(2R,1ЈS) + (2S,1ЈR)].
Consistent behavior was observed for reactions between
cyclohexanone (1a) and aldehydes 2a–f (Tables 4 and 8).
Thus, in the presence of organocatalyst exo-l-4aa, aldol ad-
ducts (2R,1ЈS)-3aa–af were obtained with ee values of
about 90%. In comparison, use of endo-l-4aa provided the
Stereoselectivity of Organocatalysts exo-L-4aa and endo-L-
4aa
In Table 4 we reported that exo-l-4aa yields the aldol enantiomeric adducts (2S,1ЈR)-3aa–af as major isomers,
adduct (2R,1ЈS)-3aa as the major stereoisomer of the 1a + with slightly lower enantiomeric excesses. Cyclopentanone
Table 8. Stereocontrol and chemical yields observed[a,b] in the 1a–c + 2a–fǞ3ab–ca reactions (Scheme 6) catalyzed by cycloadducts exo-
l-4aa and endo-l-4aa.
Entry
n
R
Cat.
Aldol
anti/syn[c]
Yield [%]
ee[d] [%]
1[a]
2
2
2
2
2
2
2
2
2
2
1
1
3
3
4-NO2C6H4
4-NO2C6H4
4-CF3C6H4
4-CF3C6H4
3-MeC6H4
3-MeC6H4
3-NO2C6H4
3-NO2C6H4
3,5-F2C6H3
3,5-F2C6H3
C6F5
exo-l-4aa
endo-l-4aa
exo-l-4aa
endo-l-4aa
exo-l-4aa
endo-l-4aa
exo-l-4aa
endo-l-4aa
exo-l-4aa
endo-l-4aa
exo-l-4aa
endo-l-4aa
exo-l-4aa
endo-l-4aa
(2R,1ЈS)-3ab
(2S,1ЈR)-3ab
(2R,1ЈS)-3ac
(2S,1ЈR)-3ac
(2R,1ЈS)-3ad
(2S,1ЈR)-3ad
(2R,1ЈS)-3ae
(2S,1ЈR)-3ae
(2R,1ЈS)-3af
(2S,1ЈR)-3af
(2R,1ЈS)-3ba
(2S,1ЈR)-3ba
(2R,1ЈS)-3ca
(2S,1ЈR)-3ca
85:15
88:12
86:14
85:15
85:15
83:17
80:20
80:20
85:15
83:17
82:18
89:11
99:01
99:01
81[e]
84[e]
79[e]
83[e]
70[e]
75[e]
60[e]
65[e]
55[e]
67[e]
65[e]
45[e]
40[f]
84[f]
92
–90
94
–70
92
–66
94
–64
86
–74
15
–10
59
2[a]
3[a]
4[a]
5[a]
6[a]
7[a]
8[a]
9[a]
10[a]
11[b]
12[b]
13[b]
14[b]
C6F5
C6F5
C6F5
–35
[a] Reactions carried out at –15 °C. [b] Reactions carried out at 25 °C. [c] Ratios determined by 1H NMR examination of the crude
reaction mixtures. [d] Enantiomeric excesses corresponding to the major anti aldol adducts determined by HPLC with a chiral stationary
phase and computed as ee = 100ϫ[(2R,1ЈS) – (2S,1ЈR)]/[(2R,1ЈS) + (2S,1ЈR)]. [e] Yields of isolated aldol adducts after conversion Ͼ 99%.
[f] Yields of isolated aldol adducts after 48 h with conversion of 96%.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
9

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
F. P. Cossío et al.
FULL PAPER
(1b) was more reactive than 1a, but the enantiocontrol was
found to be much lower (Table 8, Entries 11 and 12). The
stereoselectivity observed with cycloheptanone (1c) was
also lower than that found with 1a, although the enantio-
control was somewhat higher than that obtained for 1b
(Table 8, Entries 13 and 14). Therefore, cyclohexanone ap-
pears to represent a good compromise between reactivity
and enantiocontrol in these reactions.
From these results it is clear that the stereochemical be-
havior of exo-l-4aa–ec (see also Table 5 and Table 6) is ex-
ceptional and completely different from that observed for
l-Pro itself and other organocatalysts such as endo-l-4aa,
exoЈ-l-4aa, and endo-l-4ad. In order to understand the
reasons for this unexpected outcome, we decided to under-
take a computational study of the model aldol reaction 1a
+ 2aǞ3aa in the presence of exo-l-4aa and endo-l-4aa.
According to the accepted model for this reaction, the ste-
reochemically relevant step is C–C bond formation from
the enamine intermediates of type INT1 and aldehydes 2.
In the case of the 1a + 2aǞ3aa reaction, the enamine inter-
mediate can exist in two families of conformers INT1a and
INT1Јa depending on the s-cis or s-trans disposition,
respectively, between the double bond of the enamine moi-
ety and the methoxycarbonyl group (Scheme 7).
If we define the dihedral angle ω as ω = Ca-Nb-Cc-Cd
(Figure 6), s-cis conformations will be obtained for ω values
between 0 and Ϯ75 deg., whereas s-trans conformations can
be grouped for ωε(Ϯ95, Ϯ180). MM3 molecular dynamics
(MD) simulations (Figure 6) show different behavior of in-
termediates exo-INT1a and endo-INT1a, although in both
cases both conformations are significantly populated. Both
MD and Monte Carlo simulations show a remarkable con-
formational rigidity in the pyrrolidine ring, with three sub-
stituents occupying equatorial positions and one being iso-
clinal. In contrast, the cyclohexenyl moiety is much more
Scheme 7. (A) “Like” reaction path, and (B) “unlike” reaction path
associated with the C–C bond-forming step of the aldol reaction
between cyclohexanone and pentafluorobenzaldehyde (2a) in the
presence of organocatalyst exo-l-4aa.
flexible. At the M06–2X(PCM)/6–31+G**//B3LYP(PCM)/ these saddle points along different topologies at the
6–31G*+ΔZPVE level, exo-INT1a was calculated to be B3LYP(PCM = cyclohexanone)/6–31G* level of theory met
0.4 kcalmol–1 less stable than the exo-INT1Јa conformer, with no success in the case of the interaction of 2a with
whereas the reverse relative stability was found for the endo exo-INT1a through the nitro group (paths a, aЈ, b, and bЈ,
analogues (see Figure S5 in the Supporting Information). Scheme 7) when X = OMe and Y = nothing, H+. For Y =
Therefore, from these analyses, no general kinetic stereo- H3O+, saddle points associated with interaction patterns a
selectivity trends can be envisaged for the aldol reaction and c through the nitro group were located, but were found
in the presence of organocatalysts exo-l-4aa and endo-l- to lie ca. 4 kcalmol–1 above their analogues exo-TS1 and
4aa.
exo-TS4 (see Figure S6 in the Supporting Information), in
If we consider the interaction pathways of the intermedi- which the interaction between exo-INT1a and 2a takes
ate exo-INT1a, derived from exo-l-4aa with aldehyde 2a, in place through the methoxycarbonyl group (Scheme 7, inter-
principle these interactions can occur through both the action modes b and d). Therefore, we will restrict the follow-
nitro and methoxycarbonyl (X = Me) or carboxy (X = H) ing discussion to transition structures involving X = H, Me
groups, as depicted in Scheme 7. In addition, depending and Y = H+ through the interaction patterns b, bЈ, d, and
upon the reaction conditions, the elementary step leading dЈ shown in Scheme 7.
to the formation of the C–C bond can involve different Y-
species. Thus, under non-acidic conditions Y can be H2O saddle points exo-TS1–4 [M06–2X(PCM = cyclohexan-
(formed as a consequence of the generation of exo-INT1a one)/6-31+G**//B3LYP(PCM cyclohexanone)/6-31G*
The chief geometric features and relative energies of
=
in situ) or nothing. Similarly, reaction conditions including level of theory] are shown in Figure 7. These transition
acidic additives can give rise to structures in which Y = H+ structures show Houk–List geometries[23] in which three
or H3O+ (Scheme 7). The possible combinations generate a substituents of the pyrrolidine ring occupy equatorial posi-
very complex situation in which many alternative transition tions in half-chair conformations and the methoxycarbonyl
structures can be present. However, intensive searching of group, in general constrained by the hydrogen bond with
10
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
Organocatalytic Aldol Activity of Unnatural Proline Esters
Figure 6. Molecular dynamics plots for (A) exo-INT1a, and (C) endo-INT1a, computed with use of the MM3 forcefield. The Ca-Nb-Cc-
Cd dihedral angle is denoted ω. Areas denoted s-cis include structures for which ω Յ Ϯ75 deg. Monte Carlo conformational analyses
calculated with MM3 for (B) exo-INT1a, and (D) endo-INT1a. Ball and stick representations correspond to the lowest-energy conformers,
and the wire structures include the 20 most stable conformers within 2.6 and 2.9 kcalmol–1, respectively.
the protonated aldehyde, occupies an isoclinal position lower ee for exo-l-4Јaa in the absence of TFA (Table 7, En-
(Figure 7). The Bürgi–Dunitz angles defined by Cc, Cb, and try 4).
the OH group vary from 100.5 to 106.9°. The dihedral angle
The transition structures endo-TS1–4 associated with the
φ defined as φ = Ca-Cb-Cc-Cd (Figure 8) shows values close 1a + 2aǞ3aa reaction catalyzed by endo-l-4aa (Figure 8)
to +100° in exo-TS2 and –100° in exo-TS3. This dihedral show non-optimal geometric features in all cases. Thus,
angle is closer to 180° in exo-TS1 and exo-TS4. The former endo-TS1–3 each show one axial and two isoclinal groups
saddle point, which is the transition structure of lowest en- in the pyrrolidine ring, and only endo-TS4 is able to accom-
ergy, shows an s-cis conformation of the enamine moiety modate four equatorial substituents, but at the cost of a
and an antiperiplanar conformation of this group with re- certain steric congestion between the C6F5 group and the
spect to the C6F5 substituent, with φ = –174.1°, thus mini- equatorial phenyl group at C5 of the organocatalyst (Fig-
mizing the steric repulsion generated by the 5-phenyl group ure 8). This tradeoff between unfavorable features might be
of exo-l-4aa. These features make exo-TS1 the least ener- responsible for the slightly lower catalytic activity observed
getic transition structure associated with the C–C bond- for endo-l-4aa (Figure 4 and Table 1, Entry 3). The reverse
forming step of the 1a + 2aǞ3aa reaction in the presence enantioselectivity observed for this catalyst with respect to
of organocatalyst exo-l-4aa. This transition structure leads exo-l-4aa is also compatible with the relative energies
to anti-aldol (2R,1ЈS)-3aa, in good agreement with the ex- shown in Figure 8. Saddle point endo-TS4 has suitable
perimental results. The calculated anti/syn ratio is Ͼ99:01, Bürgi–Dunitz and φ angles, the four substituents of the
and the computed ee for the anti isomers in the presence of pyrrolidine ring being in equatorial positions. In addition,
exo-l-4aa is Ͼ99%. Both results are in qualitative agree- this transition structure can develop a strong hydrogen
ment with the experimentally determined values of 95:05 bond between the OH group being formed and the meth-
and 89%, respectively (Table 4, Entries 1 and 2).
oxycarbonyl group. However, the s-trans conformation of
It is worth noting that when X = H and Y = none, the the cyclohexenyl moiety imposes a close proximity between
corresponding Houk–List saddle point exo-TS1 is the C6F5 group of the electrophile and the 5-phenyl group
0.8 kcalmol–1 lower in energy than its exo-TS4 analogue of endo-l-4aa. This unfavorable steric congestion does not
(see Figure S7 in the Supporting Information), a result overcome the other favorable features of endo-TS4, thus
pleasingly in agreement with the experimentally observed making this stationary point, which leads to anti aldol
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
11

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
F. P. Cossío et al.
FULL PAPER
Figure 8. Chief features of transition structures endo-TS1–4 fully
optimized at the B3LYP(PCM)/6-31G* level of theory. See the cap-
tion of Figure 7 for further details.
Figure 7. Chief features of transition structures exo-TS1–4 fully op-
timized at the B3LYP(PCM)/6–31G* level of theory. Bond lengths
and angles are given in Å and deg., respectively. The dihedral angle
defined by Ca, Cb, Cc, and Cd atoms is denoted φ. Numbers in
parentheses correspond to relative energies (in kcalmol–1)
computed at the M06-2X(PCM)/6-31+G**//B3LYP(PCM)/6-
31G*+ΔZPVE level. Equatorial and isoclinal positions are denoted
eq and iso, respectively.
Conclusions
From the experimental and computational data reported
in this study the following conclusions can be drawn:
(i) densely substituted unnatural proline esters, readily ac-
cessible through asymmetric [3+2] cycloadditions, can be
used as efficient organocatalysts for aldol reactions, and
methyl esters in the presence of TFA are superior to the
corresponding acids in terms of accessibility, reaction rate,
(2S,1ЈR)-3aa, the lowest-energy saddle point of this series. and stereocontrol, (ii) in these systems, substituents at posi-
The calculated anti/syn ratio at 298 K is 97:03 and the ee tions C3 and C4 of the pyrrolidine ring must be trans to
for the anti aldols is Ͼ99%. Once again, these results are in each other, (iii) quaternary centers at C2 destroy the cata-
qualitative agreement with the experimental data at 298 K lytic activity, (iv) aryl groups at C5 are preferable to alkyl
[96:04 and 81% (absolute value), respectively (Table 4, En- groups, (v) bulky alkoxycarbonyl groups at C2 must be
try 6)].
avoided, (vi) the sense of induction of the organocatalysts
These combined experimental and computational data depends on several factors (maximum number of equatorial
indicate that densely substituted proline-based organocata- substituents in the pyrrolidine ring, intramolecular
lysts can show unexpected organocatalytic properties. Thus, hydrogen bonds, gauche disposition between the N–C=C
aside from the geometric and stereoelectronic effects im- and C=O bonds in the transition structure, and potential
posed by the Houk–List geometries, subtle remote effects steric clash between the substituent at C5 and the substitu-
stemming from conformational preferences in the pyrrol- ent of the aldehyde), and (vii) according to the balance of
idinone ring can result in unique asymmetric induction ef- these factors exo-l-proline esters preferentially yield anti
fects, not available for organocatalysts based on natural l- (2R,1ЈS) aldols, whereas the endo organocatalysts yield the
amino acids such as l-Pro.
corresponding (2S,1ЈR) aldols as major products.
12
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
Organocatalytic Aldol Activity of Unnatural Proline Esters
and the DIPC for generous allocation of computational and analyt-
ical resources. Technical and human support provided by SGIker
(UPV/EHU, MINECO, GV/EJ, ESF) is gratefully acknowledged.
Experimental Section
Determination of Pseudo-First Order Kinetic Constants: 2,3,4,5,6-
Pentafluorobenzaldehyde (2a) (0.1 mmol, 20 mg, 1 equiv.) and the
corresponding organocatalyst 4 (0.03 mmol, 0.3 equiv.) were dis-
solved in neat, freshly distilled (MS, 3 Å) cyclohexanone (1a,
0.63 mL, 6 mmol, 60 equiv.) at room temperature. This mixture was
[1] a) R. Mahrwald (Ed.), Modern Methods in Stereoselective Aldol
Reactions, Wiley-VCH, Weinheim, Germany, 2011; b) R.
Mahrwald (Ed.), Modern Aldol Reactions, Wiley-VCH, New
York, vol. 1 and 2.
transferred to
a NMR tube, and then trifluoroacetic acid
(0.03 mmol, 2.3 μL, 0.3 equiv. in standard experiments) was added.
The resulting mixtures were immediately introduced into the NMR
probe. Samples were stabilized at 298 K. The kinetic studies were
performed by monitoring the three 19F NMR signals of free 2a,
together with the other signals also associated with the C6F5 group,
obtained with inverse gate 1H-decoupling. Trifluoroacetic acid (δ =
–76 ppm) was used as internal reference. NMR measurements were
carried out at 376.40 MHz with a Bruker Advance 400 NMR spec-
trometer, equipped with a BBOF probe incorporating z-gradients.
FID files were obtained with a spectral window of 240 ppm and
transformed with 65536 points. All spectra were recorded by ac-
cumulating 16 acquisitions, with recycling delays of 1 s. Measure-
ments were performed directly on the reaction mixtures in the
NMR tubes every 5 min. Relative concentrations of 2a at different
times were quantified by integration and statistical treatment of the
three sets of 19F signals. Kinetic constants reported in Tables 1, 2,
3, and 5 were obtained from the NMR experiments and application
of Equation (2). All KIE experiments reported in Table 3 were car-
ried out in triplicate and subjected to further statistical treatment.
[2] a) W.-D. Fessner, A. Schneider, H. Held, G. Sinerius, C. Walter,
M. Hixon, J. V. Schloss, Angew. Chem. Int. Ed. Engl. 1996, 35,
2219–2221; Angew. Chem. 1996, 108, 2366–2369; b) L. J.
Whalen, C.-H. Wong, Aldrichim. Acta 2006, 39, 63–71; c) T. D.
Machajewski, C.-H. Wong, Angew. Chem. Int. Ed. 2000, 39,
1352–1375; Angew. Chem. 2000, 112, 1406–1430; d) D. J. Kuo,
I. A. Rose, Biochemistry 1985, 24, 3947–3955; e) A. Heine,
L. J. G. Luz, C.-H. Wong, I. A. Wilson, J. Mol. Biol. 2004, 343,
1019–1030.
[3] U. Eder, G. Sauer, R. Wiechert, Angew. Chem. Int. Ed. Engl.
1971, 10, 496–497; Angew. Chem. 1971, 83, 492–492.
[4] a) Z. G. Hajos, D. R. Parrish, German Patent DE2120623; b)
Z. G. Hajos, D. R. Parrish, J. Org. Chem. 1974, 39, 1615–1621.
[5] a) B. List, R. A. Lerner, C. F. Barbas III, J. Am. Chem. Soc.
2000, 122, 2395–2396; b) For the extension to an l-Pro-cata-
lyzed Robinson reaction, see: T. Bui, C. F. Barbas III, Tetrahe-
dron Lett. 2000, 41, 6951–6954; c) See also: C. F. Barbas III,
Angew. Chem. Int. Ed. 2008, 47, 42–47; Angew. Chem. 2008,
120, 44–50.
[6] B. List, P. Pojarliev, C. Castello, Org. Lett. 2001, 3, 573–575.
[7] a) W. Notz, F. Tanaka, C. F. Barbas III, Acc. Chem. Res. 2004,
37, 580–591; b) B. List, J. W. Yang, Science 2006, 313, 1584–
1586; c) S. Mukherjee, J. W. Yang, S. Hoffmann, B. List, Chem.
Rev. 2007, 107, 5471–5569; d) G. Guillena, C. Nájera, D. J.
Ramón, Tetrahedron: Asymmetry 2007, 18, 2249–2293; e) M. J.
Gaunt, C. C. C. Johansson, A. McNally, N. T. Vo, Drug Dis-
covery Today 2007, 12, 8–27; f) A. Dondoni, A. Massi, Angew.
Chem. Int. Ed. 2008, 47, 4638–4660; Angew. Chem. 2008, 120,
4716–4739; g) P. Melchiorre, M. Marigo, A. Carlone, G. Bar-
toli, Angew. Chem. Int. Ed. 2008, 47, 6138–6171; Angew. Chem.
2008, 120, 6232–6235; h) S. Bertelsen, K. A. Jørgensen, Chem.
Soc. Rev. 2009, 38, 2178–2189; i) M. Nielsen, D. Worgull, T.
Zweifel, B. Gschwend, S. Bertelsen, K. A. Jørgensen, Chem.
Commun. 2010, 46, 632–649; j) J. G. Hernández, E. Juaristi,
Chem. Commun. 2012, 48, 5396–5409.
Molecular Dynamics: The MD simulations of intermediates INT1a
and INT1Јa were carried out with use of the MM3 forcefield[35]
as implemented in the MacroModel package.[36] Calculations were
performed with SHAKE[37] to constrain the C–H bonds. The tem-
perature was set up to 298 K. The system was equilibrated for 1 ns
with time steps of 1 fs. The production run was started from this
point and lasted another 600 nanoseconds with time steps of 1 fs.
In all cases we observed that during the production period the en-
ergy and temperature of the whole system were equilibrated. Dur-
ing the production run, the coordinates of 500 structures were
saved.
DFT Calculations: Density functional theory calculations were per-
formed by use of the Gaussian 09 suite of programs.[38] Full geome-
try optimizations and harmonic analyses were carried out by use
of the B3LYP hybrid functional[39] and the 6-31G* basis set. Sol-
vent effects were computed by the PCM method[40] and with cyclo-
hexanone as model solvent. Single-point energies were computed
by use of the M06-2X functional[41] and the 6-31+G** basis set.
Free energies were computed at 298.15 K.
[8] a) K. A. Ahrendt, C. J. Borths, D. W. C. MacMillan, J. Am.
Chem. Soc. 2000, 122, 4243–4244; b) D. W. C. MacMillan, Na-
ture 2008, 455, 304–308.
[9] a) S. G. Zlotin, A. S. Kucherenko, I. P. Beletskaya, Russ. Chem.
Rev. 2009, 78, 737–784; b) H. Wennemers, Chimia 2010, 64,
864–865; c) U. Scheffer, R. Mahrwald, Synlett 2011, 1660–
1667; d) V. Bisai, A. Bisai, V. K. Singh, Tetrahedron 2012, 68,
4541–4580.
[10] a) Y. Hayashi, T. Itoh, N. Nagae, M. Ohkubo, H. Ishikawa,
Synlett 2008, 1565–1570; b) M. Amedjkouh, Tetrahedron:
Asymmetry 2007, 18, 390–395; c) M. Majewski, I. Niewczas,
N. Palyam, Synlett 2006, 2387–2390; d) A. Córdova, W. Zou,
P. Dziedzic, I. Ibrahem, E. Reyes, Y. Xu, Chem. Eur. J. 2006,
12, 5383–5397.
[11] a) Y. Lam, K. N. Houk, U. Scheffler, R. Mahrwald, J. Am.
Chem. Soc. 2012, 134, 6286–6295; b) M. Markert, U. Scheffler,
R. Mahrwald, J. Am. Chem. Soc. 2009, 131, 16642–16643; c)
U. Scheffler, R. Mahrwald, J. Org. Chem. 2012, 77, 2310–2330.
[12] S. S. V. Ramasastry, H. Zhang, F. Tanaka, C. F. Barbas III, J.
Am. Chem. Soc. 2007, 129, 288–289.
[13] S. S. V. Ramasastry, K. Albertshofer, N. Utsumi, C. F.
Barbas III, Org. Lett. 2008, 10, 1621–1624.
[14] a) S. S. V. Ramasastry, K. Albertshofer, N. Utsumi, F. Tanaka,
C. F. Barbas III, Angew. Chem. Int. Ed. 2007, 46, 5572–5575;
Angew. Chem. 2007, 119, 5668–5671; b) Z. Jiang, H. Yang, X.
Han, J. Luo, M. W. Wong, Y. Lu, Org. Biomol. Chem. 2010, 8,
1368–1377.
Acknowledgments
Financial support was provided by the Spanish Ministerio de Eco-
nomía y Competitividad (MINECO) and the Fondo Europeo de
Desarrollo Regional (FEDER) (projects CTQ2010-16959/BQU,
CTQ2012-35535,
CTQ2013-40855-R,
CTQ2007-62771/BQU,
CTQ2010-20387, CTQ2010-17436, and Consolider-Ingenio
CSD2007-00006), the University of the Basque Country (UPV/
EHU, UFI11/22 QOSYC), the Basque Government (GV/EJ, grant
IT-324-07), the Generalitat Valenciana-FEDER (PROMETEO/
2009/039), the Gobierno de Aragón-FSE (research group E40), and
the University of Alicante. M. d. G. R. thanks the Donostia Inter-
national Physics Center (DIPC) for a postdoctoral contract. M. S.
and L. C. gratefully thank MINECO for a contract funding their
PhD projects. The authors thank the SGI/IZO-SGIker UPV/EHU
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
13

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
F. P. Cossío et al.
FULL PAPER
[15] A. Bassan, W. Zou, E. Reyes, F. Himo, A. Córdova, Angew.
Chem. Int. Ed. 2005, 44, 7028–7032; Angew. Chem. 2005, 117,
7190–7194.
[16] K. Rohr, R. Mahrwald, Org. Lett. 2012, 14, 2180–2183.
[17] B. M. Trost, C. S. Brindle, Chem. Soc. Rev. 2010, 39, 1600–
1632.
[18] a) B. List, L. Hoang, H. J. Martin, Proc. Natl. Acad. Sci. USA
2004, 101, 5839–5842; b) M. B. Schmid, K. Zeitler, R. M.
Gschwind, Angew. Chem. Int. Ed. 2010, 49, 4997–5003; Angew.
Chem. 2010, 122, 5117–5123; c) H. Zhu, F. R. Clemente, K. N.
A. Linden, F. P. Cossío, J. Am. Chem. Soc. 2000, 122, 6078–
6092; c) M. Ayerbe, A. Arrieta, F. P. Cossío, A. Linden, J. Org.
Chem. 1998, 63, 1795–1825.
[29] CCDC-965082 (for exoЈ-l-10aa) contains 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. The absolute
configuration of exoЈ-l-10aa was unambiguously assigned by
considering the Flack parameter value: 0.008(6); see the Sup-
porting Information for details.
Houk, M. P. Meyer, J. Am. Chem. Soc. 2009, 131, 1632–1633; [30] E. V. Anslyn, D. A. Dougherty, Modern Physical Organic
d) D. Seebach, A. K. Beck, D. M. Badine, M. Limbach, A.
Eschenmoser, A. M. Treasurywala, R. Hobi, W. Prikoszovich,
B. Linder, Helv. Chim. Acta 2007, 90, 425–471.
Chemistry, University Science Books, Sausalito, CA, 2006,
p. 507–523.
[31] H. Lineweaver, D. Burk, J. Am. Chem. Soc. 1934, 56, 658–666.
[19] a) P. H.-Y. Cheong, C. Legaut, J. M. Um, N. Çelebi-Ölçum, [32] C. J. Rogers, T. J. Dickerson, K. D. Janda, Tetrahedron 2006,
K. N. Houk, Chem. Rev. 2011, 111, 5042–5137; b) A. K.
Sharma, R. B. Sunoj, Angew. Chem. Int. Ed. 2010, 49, 6373– [33] N. Zotova, L. J. Broadbelt, A. Armstrong, D. G. Blackmond,
62, 352–356.
6377; Angew. Chem. 2010, 122, 6517–6521; c) M. P. Patil, R. B.
Sunoj, J. Org. Chem. 2007, 72, 8202–8215; d) F. R. Clemente,
K. N. Houk, J. Am. Chem. Soc. 2005, 127, 11294–11302; e) C.
Allemann, R. Gordillo, F. R. Clemente, P. H.-Y. Cheong, K. N.
Houk, Acc. Chem. Res. 2004, 37, 558–569; f) F. R. Clemente,
K. N. Houk, Angew. Chem. Int. Ed. 2004, 43, 5766–5768; An-
gew. Chem. 2004, 116, 5890–5892; g) S. Bahmanyar, K. N.
Houk, Org. Lett. 2003, 5, 1249–1251; h) L. Hoang, S. Bahman-
yar, K. N. Houk, B. List, J. Am. Chem. Soc. 2003, 125, 16–17;
i) K. N. Rankin, J. W. Gauld, R. J. Boyd, J. Phys. Chem. A
2002, 106, 5155–5159.
Bioorg. Med. Chem. Lett. 2009, 19, 3934–3937.
[34] D. C. Harris, Experimental Error in Quantitative Chemical
Analysis, 8th ed., W. H. Freeman, New York, 2010, p. 51–67.
[35] N. L. L. Allinger, Molecular Structure. Understanding Steric
and Electronic Effects from Molecular Mechanics, Wiley, Ho-
boken, N.J. 2010.
[36] a) MacroModel, version 9.9, Schrodinger LLC, New York, NY,
2012; b) F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Lis-
kamp, M. Lipton, C. Caulfield, G. Chang, T. Hendrickson,
W. C. Still, J. Comput. Chem. 1990, 11, 440–467.
[37] J. P. Rickaert, G. Ciccotti, H. J. C. Berendsen, J. Comp. Physiol.
1977, 23, 327–341.
[20] In the absence of acidic protons in the reaction media, l-pro-
line methyl ester yields racemic aldol adducts, see: A. Bøgevic,
N. Kumaragurubaran, K. A. Jørgensen, Chem. Commun. 2002,
620–621.
[21] a) M. B. Schmid, K. Zeitler, R. M. Gschwind, J. Org. Chem.
2011, 76, 3005–3015; b) C. Marquez, J. O. Metzger, Chem.
Commun. 2006, 1539–1541.
[38] 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. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, 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. Starov-
erov, 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. Ad-
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, 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 09, re-
vision B.01, Gaussian, Inc., Wallingford CT, 2009.
[22] D. A. Bock, C. W. Lehman, B. List, Proc. Natl. Acad. Sci. USA
2010, 107, 20636–20641.
[23] S. Bahmanyar, K. N. Houk, H. J. Martin, B. List, J. Am. Chem.
Soc. 2003, 125, 2475–2479.
[24] H. E. Zimmerman, M. D. Traxler, J. Am. Chem. Soc. 1957, 79,
1920–1923.
[25] a) A. de Cózar, F. P. Cossío, Phys. Chem. Chem. Phys. 2011,
13, 10858–10868; b) C. Nájera, J. M. Sansano, Chem. Rev.
2007, 107, 4584–4671; c) G. Pandeym, P. Banerjee, S. R. Gadre,
Chem. Rev. 2006, 106, 4484–4517; d) C. Nájera, J. M. Sansano,
Angew. Chem. Int. Ed. 2005, 44, 6272–6276; Angew. Chem.
2005, 117, 6428; e) C. Nájera, M. G. Retamosa, J. M. Sansano,
Angew. Chem. Int. Ed. 2008, 47, 6055–6058; Angew. Chem.
2008, 120, 6144–6147.
[39] a) C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–
789; b) A. D. Becke, J. Chem. Phys. 1993, 98, 1372–1377; c)
A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[26] E. Conde, D. Bello, A. de Cózar, M. Sánchez, M. A. Vázquez,
F. P. Cossío, Chem. Sci. 2012, 3, 1486–1491.
[40] a) R. Cammi, B. Mennucci, J. Tomasi, J. Phys. Chem. A 2000,
104, 5631–5637; b) J. Tomasi, B. Mennucci, R. Cammi, Chem.
Rev. 2005, 105, 2999–3093.
[41] a) Y. Zhao, D. G. Truhlar, J. Chem. Theory Comput. 2008, 4,
1849–1868; b) Y. Zhao, D. G. Truhlar, Acc. Chem. Res. 2008,
41, 157–167; c) Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008,
120, 215–241.
[27] L. M. Castelló, C. Nájera, J. M. Sansano, O. Larrañaga, A.
de Cózar, F. P. Cossío, Org. Lett. 2013, 15, 2902–2905.
[28] a) A. Arrieta, D. Otaegui, A. Zubia, F. P. Cossío, A. Díaz-
Ortiz, A. de la Hoz, M. A. Herrero, P. Prieto, C. Foces-Foces,
J. L. Pizarro, M. I. Arriortua, J. Org. Chem. 2007, 72, 4313–
4322; b) S. Vivanco, B. Lecea, A. Arrieta, P. Prieto, I. Morao,
Received: February 2, 2015
14
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O50160
/KAP1
Date: 02-03-15 15:17:10
Pages: 15
Organocatalytic Aldol Activity of Unnatural Proline Esters
Design of Organocatalysts
Densely substituted enantiopure proline es-
ters obtained through [3+2] cycloadditions
catalyze aldol reactions. These synthetic or-
ganocatalysts produce different enantiom-
eric aldol adducts depending on the stereo-
chemical dispositions of the substituents at
distal positions with respect to the catalytic
site. Design criteria are proposed for this
kind of organocatalysts.
M. d. G. Retamosa, A. de Cózar,
M. Sánchez, J. I. Miranda, J. M. Sansano,
L. M. Castelló, C. Nájera, A. I. Jiménez,
F. J. Sayago, C. Cativiela,
F. P. Cossío* .................................... 1–15
Remote Substituent Effects on the Stereo-
selectivity and Organocatalytic Activity of
Densely Substituted Unnatural Proline Es-
ters in Aldol Reactions
Keywords: Organocatalysis / Aldol reac-
tions / Kinetics / Reaction mechanisms /
Transition states
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
15