Stereodivergent Aminocatalytic Synthesis of Z- and E-Trisubstituted Double Bonds from Alkynals

Article abstract of DOI:10.1002/chem.201603437

A highly diastereoselective synthesis of trisubstituted Z- or E-enals, which are important intermediates in organic synthesis, as well as being present in natural products, is described using different alkynals and nucleophiles as starting materials. Diastereocontrol is mainly governed by the appropriate catalyst. Therefore, those reactions controlled by steric effects, such as the J?rgensen–Hayashi's catalyst, give access to E isomers, and those catalysts that facilitate hydrogen bonding, such as tetrazol-pyrrolidine Ley's catalyst, allow the synthesis of Z isomers. A stereochemical model based on DFT calculations is proposed.

Full text of DOI:10.1002/chem.201603437

DOI: 10.1002/chem.201603437
Full Paper
&
Organic Synthesis |Hot Paper|
Stereodivergent Aminocatalytic Synthesis of Z- and
E-Trisubstituted Double Bonds from Alkynals
Leyre Marzo,[a] Javier Luis-Barrera,[a] Rubꢀn Mas-Ballestꢀ,[b] Josꢀ Luis Garcꢁa Ruano,[a] and
Josꢀ Alemꢂn*[a]
Abstract: A highly diastereoselective synthesis of trisubsti-
tuted Z- or E-enals, which are important intermediates in or-
ganic synthesis, as well as being present in natural products,
is described using different alkynals and nucleophiles as
starting materials. Diastereocontrol is mainly governed by
the appropriate catalyst. Therefore, those reactions con-
trolled by steric effects, such as the Jørgensen–Hayashi’s cat-
alyst, give access to E isomers, and those catalysts that facili-
tate hydrogen bonding, such as tetrazol-pyrrolidine Ley’s
catalyst, allow the synthesis of Z isomers. A stereochemical
model based on DFT calculations is proposed.
Introduction
The configuration of double bonds (Z or E) plays a central role
in nature, for example, in fatty acids and also in the biological
properties of different natural and pharmaceutical com-
pounds.[1] For example, (Z)-tamoxifen is an antiestrogen that
inhibits the development and growth of mammary tumors in
rats and is effective in treating estrogen-dependent metastatic
breast cancer in humans.[1b] By contrast, the diastereoisomer
(E)-tamoxifen does not have any clinical uses because it lacks
an antiestrogenic effect (A, Figure 1).[1c] Other examples are the
E- and Z-piperolides that exhibit different antimycotic activity
against Cladosporium cladosporoides.[1e] In addition, the E- and
Z-propenones differ in their COX-2 inhibitory activity.[1j] Further-
more, the stereochemical control of double-bond formation is
critical for the synthesis of heterocycles, such as the lactame
ring synthesis of the anticancer spirodinolinone derivatives B[2]
or the piran-2-one ring C[3] with multiple biological properties
(Figure 1). Therefore, the ability to control the configuration of
a double bond is an extremely important task in the synthetic
design of new drugs containing this structural moiety.
Figure 1. Structure and the importance in the Z/E control.
partial success for one of the two diastereoisomers. Most of
the methods are designed for the addition to nonsubstituted
terminal triple bonds (R2 =H), yielding disubstituted double
bonds (Scheme 1, top). Interestingly, when a terminally substi-
tuted triple bond is used (R2 ¼H), the literature generally
shows two types of additions with some degree of Z/E selectiv-
ity: 1) Michael reaction to alkynones that were reported to
yield the Z adducts,[5] and 2) alkynyl esters that generally pre-
sented E selectivity.[6] Other additions cannot be governed with
any selectivity.[7] This poor control or lack of Z/E selectivity is
related to the scarce stereochemical control in the facial recog-
nition of the intermediary allenic enolate I (Scheme 1, top).
Consequently, the design of Z/E-selective processes is still
a challenge. Interestingly, stereodivergent systems for the syn-
thesis of Z or E double bonds have not been described. Con-
sidering the intermediate I (Scheme 1, top), the selectivity
problem is inherent in the protonation step, which is not con-
trolled. Therefore, most authors have employed different cata-
lytic systems to control, in an intermolecular manner, the pro-
tonation of the enolate intermediate I.
A detailed revision of the literature reveals that one of the
best methods for the synthesis of olefins is the addition of nu-
cleophiles to electron-deficient alkynes, which results in most
cases, in mixtures of Z/E isomers.[4] Different catalytic and non-
catalytic additions have been performed, in some cases with
[a] Dr. L. Marzo, J. Luis-Barrera, Dr. J. L. G. Ruano, Dr. J. Alemꢀn
Departamento de Quꢁmica Orgꢀnica, Universidad Autꢂnoma de Madrid
Cantoblanco, 28049 Madrid (Spain)
To increase the control in this process, it was envisioned that
a catalyst directly attached to the intermediate I (i.e. covalently
bonded) would increase control in the Z/E selectivity
(Scheme 1, bottom) through a protonation step in an intermo-
lecular (steric effect, Scheme 1, left-bottom) and intramolecular
(hydrogen bond, Scheme 1, right-bottom) manner. Therefore,
the use of an alkynal[8] would allow the formation of the imini-
E-mail: jose.aleman@uam.es
Homepage: www.uam.es/jose.aleman
[b] Dr. R. Mas-Ballestꢃ
Departamento de Quꢁmica Inorgꢀnica, Universidad Autꢂnoma de Madrid
Cantoblanco, 28049 Madrid (Spain)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201603437.
Chem. Eur. J. 2016, 22, 1 – 12
1
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Results and Discussion
To check Z/E selectivity in the nucleophilic addition to alkynals,
the reaction between the 1,3 dicarbonyl compound 2A and
acetylenic derivative 1a, in the presence of catalysts 3a and
4a was carried out. The use of 3a (20 mol%) as a prototype of
a steric-shielding catalyst in acetone preferentially led to the E
isomer (Z/E=19:81, Scheme 3, right-side). In contrast, when
the reaction was carried out in the presence of a hydrogen-
bond-type catalyst, such as the prolinol 4a, the Z isomer was
preferentially formed in the same solvent (Z/E=74:26,
Scheme 3, left-side), thus demonstrating our initial hypothesis.
Interestingly, identical Z/E selectivity was obtained when using
the enantiomers (R or S) of the catalysts 3a and 4a, or their
racemic forms.
Scheme 1. Existing strategies for control in the Z/E selectivity and the pro-
posed new strategy.
Scheme 3. First trials in the Z/E selectivity under aminocatalytic conditions.
um ion that could be attacked by a nucleophile. The nucleo-
phile chosen is a 1,3-dicarbonyl compound (commonly used in
organocatalysis) that would enable the formation of the allena-
mine intermediate II (Scheme 2). Consequently, the control in
the Z and E selectivity would be governed by an aminocatalyst
(hydrogen-bond donor group, IIa, versus bulky group, IIb). Pre-
sumably, the use of a hydrogen-bond donor catalyst would fa-
cilitate the protonation at the same prochiral face at the hy-
drogen-bond group, whereas the use of a bulky group at the
aminocatalyst would provoke a protonation at the opposite
face to this group.
Mechanistic considerations: DFT calculations
To explain the observed Z/E regioselectivities, we addressed
the theoretical investigation of the possible reaction pathways
between the acetylene-containing iminum cationic species de-
rived from catalysts 3a or 4a, and the corresponding nucleo-
phile to generate the product 5Aa (Scheme 3). Geometry opti-
mizations were carried out using the M06 functional with 6-
31G(d,p) basis set,[9] including acetone solvent effects by
means of a SMD continuum model, which is a standard meth-
odology previously used for related aminocatalytic systems.[9a]
Electronic energies were refined by single-point calculations at
M06/6-311+ +G(d,p) also including solvent effects. Energy
values reported are given as increments of Gibbs free energies
considering thermal corrections obtained at the M06/6-
31G(p,d) level. The large dimensions of the calculated systems
prevented the use of more sophisticated methods on both op-
timizations and thermal correction calculations (see the Sup-
porting Information for additional details).
In this work, we report a new method for preparing Z- and
E-trisubstituted olefins starting from alkynals with efficient dia-
stereocontrol under organocatalytic conditions and propose
a mechanistic stereochemical model based on DFT calcula-
tions.
Initial and general considerations
A general picture of the plausible catalytic cycle is shown in
Scheme 4. First, iminium ion formation VI would take place,
which should be attacked by the enol III or enolate IV to form
the allenamine intermediate VII. Such transient species (VII), al-
though undetected, has also been proposed for related sys-
tems.[8] Then, protonation of this allenamine VII would take
place in an intramolecular (Scheme 4, bottom-left) or intermo-
lecular manner (Scheme 4, top-left) to produce the correspond-
Scheme 2. Proposal for the diastereo-divergency for alkynals through amino-
catalysis.
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
2
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
ing trisubstituted alkene VIII, precursor of 5Aa (E or Z). Inter-
estingly, the protonation step was already proposed by Zim-
merman in 1955 as the key step for the Z/E selectivity in the
addition reaction of a,b-unsaturated carbonyl compounds via
an alenol intermediate.[10] In order to rationalize such a mecha-
nistic proposal, each step of this reaction was separately con-
sidered: 1) acid–base equilibrium between the nucleophile and
pyrrolidine catalyst, 2) iminium formation, 3) nucleophilic
attack to the acetylenic species, and 4) protonation of the al-
lenamine intermediate.
Scheme 5. Deprotonation of the enol by the catalyst and formation of the
iminium ion.
cause the attack of the enol III on the iminium ion VI proceeds
through high energetic barriers (between 21.3 and 24.1 kcal
molꢀ1, see Scheme 6 and the Supporting Information). Further-
more, the intramolecular protonation should proceed from the
enol fragment to the acetylene moiety and this mechanism
can only account for the E product and cannot explain the for-
mation of Z products.
Scheme 4. General mechanistic scheme proposed in this work (Q=CPh2OH
or CPh2OTMS).
Iminium formation and enol deprotonation
We initially assessed the thermodynamic feasibility of the two
prior processes: acid–base equilibrium between the nucleo-
phile and pyrrolidine catalyst and iminium formation. The pyr-
rolidine catalyst can act as a base (from 3a or 4a to V,
Scheme 5, top) and the deprotonation of the nucleophile is
feasible with both catalysts 3a and 4a (DG= +0.6 kcalmolꢀ1,
Keq =0.36 and DG=ꢀ1.1 kcalmolꢀ1
, Keq =6.4, respectively),
which is in agreement with the lower nucleophilicity of catalyst
3a. The formation of the iminium ion VI from V and 1a is
clearly favored for catalysts 3a and 4b (Scheme 5, bottom). Re-
markably, formation of each iminium VI implies the consump-
tion of a proton to form a water molecule, which would be
the pyrrolidinium salt V, previously formed by deprotonation
of III by 3a or 4a. Once the iminium ion VI was formed, a nu-
cleophilic attack from the enol III or enolate IV would take
place.
Scheme 6. Enol attack to the iminium ion and intramolecular protonation to
give (E)-5Aa (Q=CPh2OH or CPh2OTMS).
A more plausible addition is related to the attack of the eno-
late IV to the iminium ion VI, leading to thermodynamically fa-
vorable allenamine intermediates with low kinetic barriers
¼
(DG =from 4.2 to 6.1 kcalmolꢀ1, see Figures 4 and 6). The nu-
cleophilic attack of the enolate IV on the iminium ion VI can
proceed in two different orientations and leads to the allena-
mines VII and VII’ under the catalysis of 3a or 4a (Figures 2
and 3, respectively). For each allenamine another plausible ro-
tation in the CꢀN bond can take place, giving two additional
structures VII-rot and VII’-rot.
Nucleophilic attack
A first possible pathway is the enol formation (III) and a further
attack on the acetylenic iminium intermediate VI, followed by
proton transfer to produce the corresponding trisubstituted
alkene 5Aa (Scheme 6). Such a pathway is not relevant be-
Protonation of these species gives access to the final double
bonds. The protonation source should be the pyrrolidium ion
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
3
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 2. Analysis in the approach of the pyrrolidinum ion to the allenamine intermediate (Q=CPhPhOTMS). Stabilities in kcalmolꢀ1 are shown in brackets.
Figure 3. Analysis in the approach of the pyrrolidinum ion to the allenamine intermediate (Q=CPh2OH). Stabilities in kcalmolꢀ1 are shown in brackets.
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
4
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
(V), which is the more acid proton in the media (Figure 2, top
left, and Figure 3). A detailed analysis in the approach of the
pyrrodinium ion to the allenamines reveals that only two of
the four dispositions can give access to the final double
bonds, since the free energies of these intermediates are very
high or access to the proton source is not adequate (for values
in brackets, see the Supporting Information for more details).
Therefore, the energetic profiles of the plausible pathways (VII
and VII’ allenamine intermediates) for both catalysts 3a and
4a were analyzed.
Energetic profile for catalyst 4a—kinetic control
The energetic profiles shown in this and subsequent sections
start from iminium and enolate ion pairs. Two main steps are
analyzed from both themodynamic and kinetic perspectives:
First the nucleophilic attack of enolate to acetylenic iminium
species to form allenamine intermediate and, second, the pro-
tonation of these allenamines by pyrolidinium cation V accord-
ing the geometries presented in Figures 2 and 3.
For catalyst 4a, the nucleophilic attack of the enolate IV on
the iminium ion VI can proceed in two different orientations
(Figure 4); from TS-VI to form VII or from TS-VI’ to generate
VII’ (nucleophilic addition step). Such processes (TSVI and
TSVI’ Figure 4) are thermodynamically allowed (DGreact =ꢀ14.9
and ꢀ14.8 kcalmolꢀ1) and both products are produced
through a very similar kinetic barrier (4.3 kcalmolꢀ1 for VII and
4.2 kcalmolꢀ1 for VII’, Figure 4).
Figure 4. Energetic profile of the reaction pathway followed using catalyst
4a (Q=CPh2OH). The zero value is established at the VI+IV ion pair.
In this first step there is not a significant differentiation of
the two pathways from either kinetic or thermodynamic points
of view. Therefore, the key to the diastereo-differentiation
should be related to the protonation step of these allene inter-
mediates. An analysis of the protonation step according to the
catalyst 4a is described below. Firstly, a hydrogen bonding be-
tween the two pyrrolidine moieties was found, one as the
proton source and the other in the allenamine fragment (TSVII
and TSVII’, Figure 5). From the allenamine intermediates VII or
VII’, alkene final products are generated by protonation using
pyrrolidinium cation V as the proton source (previously formed
from 4a by deprotonation, see Scheme 5). As a consequence
¼4
of DG being larger than DG¼3, the formation of VIII is fa-
vored under kinetic control, as the protonation orientation is
the determining factor in the Z/E selectivity (Figure 4, top). As
can be deduced from the reaction profile (Figure 4), kinetic
control results from the difference of the activation energies
¼4
(DG ꢀDG¼3 =1.9 kcalmolꢀ1, the pro-Z transition state (TSVII)
and the pro-E transition state (TSVII’), which is in accordance
with the moderate diastereoselectivity that was obtained in
the case of the Z product (74:26 Z/E, Scheme 3). This difference
of energy is even larger when the entropic cost (VII+V and
VII’+V) is considered.
Grey and black traces in the profile shown at Figure 4 are re-
lated to the two steps of two catalytic cycles that could work
and result on formation of Z and E products, respectively. The
corresponding turnover frequency of these two alternative cat-
alytic cycles (grey or black traces) depends on the successive
kinetic barriers, as the energy of these barriers is the factor
Figure 5. Transition states for the protonation step of the allene intermedi-
ate by protonated catalyst 4a. The grey arrow points out the steric hin-
drance region.
that determines the rate at which the catalytic cycle is com-
pleted. In this case, the two calculated barriers for the pro-Z
pathway are 4.2 kcalmolꢀ1 for the nucleophilic attack and
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
5
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
5.8 kcalmolꢀ1 for the protonation step (black trace). For the
pro-E pathway, grey trace, a higher barrier has been found for
the protonation process (7.7 kcalmolꢀ1; see Figure 4, top),
whereas a very similar value has been computed for the barrier
of nucleophilic attack (4.3 kcalmolꢀ1). Therefore, it is clear that
the rate-determining step is the protonation, in which a signifi-
cant difference is found between Z (black trace) and E cycles
(grey trace). In fact, the Z cycle is kinetically preferred over the
E cycle, and, therefore, a higher turnover frequency is expected
for the Z cycle, which should have, as a consequence, a higher
accumulation of Z product. Geometrical analysis of both pro-E
and pro-Z transition states (shown in Figure 5) suggests that
steric hindrance between the nucleophilic fragment and the
pyrrolidinium proton donor make the TSVII’ transition state
(pro-E) less favorable than TSVII transition state (pro-Z).
philic fragment and the pyrrolidinium proton donor in the pro-
E transition state (TSVII’) is, in this case, comparable to the
steric hindrance between the ꢀCPh2OTMS group of the allen-
amine and the nucleophilic fragment in the pro-Z transition
state (TSVII) (see Figure 7). Therefore, in this case, there is not
a kinetic preference (the energy barrier for TSVII’ is only
0.4 kcalmolꢀ1 more stable than the barrier for TSVII, see
Figure 3). Consequently, the preference for the E product
should be understood in the context of thermodynamic con-
trol. Indeed, the E isomer is, in this case, about 6.2 kcalmolꢀ1
more stable that its Z counterpart (vide infra).
This thermodynamic control requires reversibility in the dif-
ferent steps of the catalytic cycle. Once intermediates VII or
VII’ are generated, the reversibility at this step would require
¼1’
overcoming barriers of 18.5 and 19.7 kcalmolꢀ1 (DG
and
DG¼2’; from VII or VII’ to VI or VI’, respectively, to the iminum
ion, Figure 4, top). Considering these values, such reversibility
is plausible especially at higher temperatures and longer reac-
tion times. In addition, overall barriers from VII (TSVII) and VII’
(TSVII’) should include the entropic costs of forming the VII+
Energetic profile for catalyst 3a—thermodynamic control
Although the nucleophilic attack step does not show signifi-
cantly distinct thermodynamic or kinetic features for the Z or E
cycles, a different scenario was found in the protonation step
for the reaction catalyzed by 3a (Figure 6). Due to the steric
crowding, the only orientation of the ꢀCPh2OTMS group of the
allenamine intermediate VII and VII’ should be at the opposite
site of this bulky group in the protonation reagent V (see
TSVII and TSVII’, Figure 6). Due to the relative orientation of
the ꢀCPh2OTMS groups, steric hindrance between the nucleo-
¼1’
V and VII’+V pairs which are not very distinct from DG and
DG¼2’. Furthermore, the scarce availability of the pyrrolidinium
ion is limited by the prior equilibria (acid–base equilibrium be-
tween nucleophile and pyrrolidine catalyst and iminium forma-
tion) making the path from VII or VII’ to TSVII or TSVII’, even
Figure 6. Energetic profile of the reaction pathway followed using catalyst
3a (Q=CPh2OTMS). The zero value is established at the VI+IV ion pair.
Figure 7. Transition states for protonation step of the allene intermediate by
protonated catalyst 3a. The grey arrows point out steric hindrance regions.
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
6
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
slower, and then reversibility from VII to VI or from VII’ to VI’
more probable. From VIII to VII, reversibility is quite plausible
¼3
considering the almost equally low values of DG (7.9 kcal
¼3’
molꢀ1) and DG (7.7 kcalmolꢀ1). In contrast, reversibility from
¼4
VIII’ to VII’ should be less favorable because DG (7.5 kcal
¼4’
molꢀ1) is considerably lower than DG
(13.4 kcalmolꢀ1).
Therefore, the irreversible pathway initiated by dissociation of
the VIII+3a pair and finished by imine hydrolysis is more
probable. Thus, accumulation of the thermodynamically fa-
vored product is expected.
Scheme 7. Effect of ethanol on the diastereoselectivity.
The general thermodynamic preference for the E product
can be understood when considering how the steric interac-
tions in VIII and VIII’ can affect the conjugation in the frag-
ment N=C(H)ꢀC(H)=C(Ph)(Nuc) (Figure 8). When catalyst 4a is
used the dihedral angle H-C-C-H is 1798 for the E product
(VIII’) and 1748 for the Z product (VIII). This loss of planarity
(ideally 1808), and therefore lack of conjugation, is related to
a destabilization of 4.2 kcalmolꢀ1 for the Z product (VIII). This
effect is enhanced by using the more sterically hindered cata-
lyst 3a. In this case, the dihedral angle H-C-C-H is 1728 for the
E product (VIII’) and 1598 for the Z product (VIII). Therefore,
the E product (VIII’) is 6.9 kcalmolꢀ1 more stable than its Z
counterpart.
Scheme 8. Isomerization of (Z)-5Aa in the presence of 3a.
According our mechanistic proposal, kinetic control using
catalyst 4a should be enhanced by lowering the temperature,
whereas thermodynamic control using catalyst 3a should be
enhanced by raising the temperature (Figure 9). Indeed, when
the reaction using catalyst 4a was carried out at 08C (kinetic
control, Figure 9, left) 86% of the Z isomer was obtained,
whereas this was decreased to 60% at 408C. In the case of cat-
alyst 3a (thermodynamic control, Figure 9, right), the reaction
was found to decrease the amount of E product when the re-
action was carried out at 08C (68%) in comparison to the se-
lectivity found at 408C (98% of the E isomer). Therefore, the in-
fluence of the temperature in the kinetic and thermodynamic
control in combination with the structure of the catalysts (3a
and 4a) are the principal factors that determines the observed
Z/E ratio.
Figure 8. A comparison of the thermodynamic stability of isomers E and Z
with 4a catalyst (left) and 3a catalyst (right).
Screening conditions and scope
With these computational studies and the preliminary results
obtained, the reaction conditions for the diastereoselective
synthesis of the E or Z isomers were optimized with a more re-
active nucleophile 2B (Table 1). In the absence of the catalyst,
the reaction did not take place after 48 h in toluene at 158C
(entry 1, Table 1). For the selective preparation of the E isomer,
different catalysts bearing bulky substituents were studied in
toluene at 158C (entries 2–6). The best results in toluene were
obtained with the bulkier ones (3b and 3 f, entries 2 and 6).
The influence of the solvent was then studied using 3b (en-
tries 7–10). EtOH provided the best diastereoselectivity with
3b and 3 f (Z/E=3:97, entries 8 and 14). At this point we
checked whether a decrease in the temperature (0 or ꢀ208C)
would reduce the diastereoselectivity (which would be expect-
ed for a major product resulting from thermodynamic control)
(entries 11–12), and whether a dilution in the reaction condi-
tions had any influence (entry 13). Finally, a decrease in the cat-
alyst loading to 10 or 5 mol% resulted in a lower reactivity and
diastereoselectivity (entries 15 and 16). Interestingly, the results
Experimental proof supporting the mechanistic proposal
According to the calculations mentioned above, the hydrogen
bond under the catalysis of 4a is critical to obtain the Z-enal
5Aa. To test this hypothesis, the reaction was carried out in
ethanol, which is able to break these intermolecular hydrogen
bonds (see Figure 4). The opposite configuration (E) was found
in ethanol due to a steric control taking place (Scheme 7).
To gain a better understanding of the catalyst 3a, we dis-
solved a pure sample of the Z product (obtained by kinetic
control) in toluene or acetone in the presence of catalyst 3a in
order to perform the reaction under thermodynamic control
conditions. As expected, the reaction reached thermodynamic
equilibrium and the isomerization of the E product took place,
which was partial in acetone but complete in toluene as the
solvent (Scheme 8). This experimental evidence confirms that
catalyst 3a promotes thermodynamic control through a reversi-
ble pathway.
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
7
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Table 1. Optimization in the reaction of nucleophile 2B and aldehyde 1a
in the presence of catalyst 3.[a]
Entry Cat [mol%] Solvent t [h] T [8C] Conv. [%][b] Z/E [%][c]
1
2
3
4
5
6
7
8
–
Tol.
Tol.
Tol.
Tol.
Tol.
Tol.
CH3CN
EtOH
THF
CH2Cl2
EtOH
EtOH
EtOH[d]
EtOH
EtOH
EtOH
48
16
16
16
16
17
24
24
24
24
20
20
18
16
20
20
15
15
15
15
15
15
20
20
20
20
0
-20
15
20
15
15
–
>98
>98
>98
>98
>98
>98
>98
>98
>98
>98
90
–
3b (20%)
3c (20%)
3d (20%)
3e (20%)
3 f (20%)
3b (20%)
3b (20%)
3b (20%)
3b (20%)
3b (20%)
3b (20%)
3b (20%)
3 f (20%)
3b (10%)
3b (5%)
10:90
12:88
44:56
46:54
6:94
7:93
3:97
8:92
12:88
8:92
20:80
3:97
3:97
6:94
7:93
9
10
11
12
13
14
15
16
>98
>98
94
60
[a] All reactions were performed using 0.1 mmol of 2B and 0.2 mmol of
1
1a, in 0.2 mL of solvent. [b] Conversion was determined by H NMR. [c] Z/
E ratio was determined by 1H NMR. [d] Diluted up to 0.1m. Tol.=toluene
Table 2. Scope of the nucleophiles under optimal conditions for the dia-
stereoselective preparation of the E isomers starting from 1a.[a,b]
Figure 9. Reactions under 3a and 4a catalysts under different temperatures.
obtained in the reaction of 1a with 2A were identical to those
using (R)-3b, (S)-3b, or (ꢁ)-3b as the catalyst. Therefore, the
cheaper catalyst (S)-3b was employed in further experi-
ments.[11] From the results in Table 1, we chose the conditions
of entries 8 and 14 as optimal for achieving the E isomers of
the trisubstituted olefins with the best diastereoselectivity.
Under these conditions, the reactions of 1a with nucleophiles
(2A–G) were studied. The results are depicted in Table 2.
The nucleophiles 2B and 2C yielded (E)-5Ba and (E)-5Ca, re-
spectively, as the unique diastereoisomers (Table 2), but the
use of toluene as the solvent was required in the second case.
Similarly, diastereoselectivity was complete in reactions with
2E and 2A, whereas 5Da and 5Fa were obtained in toluene
at a 15:85 and 25:75 mixture, respectively (Table 2). This result
could not be improved by using 3 f as the catalyst. Finally, the
Meldrum acid derivative 2G yielded the trisubstituted olefin
(E)-5Ga (72%) as a unique diastereoisomer.
[a] All the reactions were performed using 0.1 mmol of 2 and 0.2 mmol
of 1 in 0.2 mL EtOH at 158C. [b] Z/E ratio was determined by 1H NMR.
[c] This reaction was carried out in 0.2 mL of toluene.
At this point, the scope of the reaction with different alky-
nals was studied (Table 3). We found that under the conditions
of entry 8 in Table 1, very good results were achieved from b-
aryl alkynals, which only yielded the E isomer when electron-
donating substituents were present in the aryl group (5Bb and
5Bc) and highly selective E/Z ratios when electron-withdrawing
and ortho-substituents were used (5Bd and 5Be). In the alkyl
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
8
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
groups, slightly worse results were found. Using 1g (R=n-
pentyl) and 1h (R=cyclohexyl) produced 6:94 (61% yield) and
14:86 (73% yield) using 3 f as the catalyst. However, 1i did not
react (probably due to the strong steric restrictions of the terti-
ary alkyl groups), whereas a lower stereoselectivity was ob-
tained from the b-cyclohexenyl enal 1 f, resulting in a 31:69
mixture of isomers under the catalysis of 3 f (Table 3).
cleophiles (2A–F) to form the (Z)-5 isomers, under the condi-
tions of entry 14 of Table 4, was investigated (Table 5).
b-Ketoesters 2B–D were stereoselectively transformed in
high yields into isomer mixtures, in which Z was clearly pre-
dominant (diastereomeric ratios better than 91:9). The azlac-
tone 2E evolved into a modest 28:72 ratio of 5EA (83%
yield)[13] and b-diketones 2A and 2F were produced in yields
of 77:23 (61% yield) and 90:10 (96% yield) mixtures of Z/E iso-
mers, respectively (Table 5).
Table 3. Scope of alkynals under optimal conditions for the diastereose-
lective preparation of the E isomers starting from 2B.[a,b]
Table 4. Optimization of the reaction of 1a with 2B to give (Z)-5.[a]
Entry Cat [mol%] Solvent t [h] T [8C] Conv. [%][b]
Z/E [%]]c]
1
2
3
4
5
6
7
8
4a (20%)
4b (20%)
4c (20%)
4d (20%)
4e (20%)
4d (20%)
4d (20%)
4d (20%)
4d (20%)
4d (20%)
4d (20%)
4a (20%)
4e (20%)
4d (20%)
Tol.
Tol.
Tol.
Tol.
Tol.
Tol.
EtOH
DCE
CH2Cl2
CHCl3
THF
THF
THF
THF[d]
16
16
16
16
16
16
20
20
20
20
20
20
20
18
15
15
15
15
15
0
15
15
15
15
15
18
18
15
63
76:24
–
–
no reaction
no reaction
>98
59
53
no reaction
>98
>98
>98
>98
63:37
83:27
77:23
[a] All reactions were performed with 0.1 mmol of 2, 0.2 mmol of 1, and
20 mol% of catalyst 3b in 0.2 mL of EtOH at 158C. [b] Z/E ratio was deter-
mined by 1H NMR. [c] The Wittig reaction product was isolated. [d] Ob-
tained using 3 f as the catalyst.
[e]
–
76:24
77:23
80:20
90:10
–
9
10
11
12
13
14
no reaction
>98
>98
For the optimization of the conditions providing Z isomers
as the major components of the reaction mixtures, we also
chose the reaction between 1a and 2B in toluene at 158C
(Table 4). Firstly, we studied the influence of different catalysts
bearing acid protons (entries 1–5, Table 4). The reaction did
not take place in the presence of 4b and 4c (entries 2 and 3,
Table 4), whereas 4a (entry 1), 4d (entry 4), and 4e (entry 5)
were found to be efficient catalysts, yielding mixtures of geo-
metric isomers (with Z being predominant) with moderate se-
lectivity. As the reactivity was higher with 4d (the only one
that provided full conversion after 16 h) it was initially chosen
for the rest of the optimization experiments. A decrease in the
reaction temperature to 08C resulted in a higher diastereose-
lectivity (as expected for a major product coming from kinetic
control) but with a lower conversion (entry 6, Table 4). Different
solvents were studied under the catalysis of 4d at 158C (en-
tries 7–11). The reaction did not take place in EtOH (entry 7),
whereas the conversion was complete in aprotic solvents (en-
tries 8–11), with THF giving the best Z/E ratio (90:10, entry 11).
As 4a and 4e had shown higher stereoselectivity control than
4d in toluene, their reactions were also studied in THF, but the
results were unsuccessful (compare entries 12 and 13 with 11).
Finally, a dilution to 0.1m in THF at 158C improved the results
at entry 11, providing a full conversion and a Z/E ratio of 92:8
(entry 14).[12] These conditions were chosen as the most appro-
priate for the diastereoselective synthesis of the Z-trisubstitut-
ed olefins. The scope of the reaction of 1a with different nu-
86:14
92:8
[a] All reactions were performed on a 0.1 mmol scale of 2B in 0.2 mL sol-
vent. [b] Conversion was determined by 1H NMR. [c] Z/E ratio was deter-
mined by 1H NMR. [d] Diluted to 0.1m. [e] Catalyst was not completely
soluble.
Table 5. Scope of the nucleophiles under optimal conditions for the dia-
stereoselective preparation of the Z isomers starting from 1a.[a,b]
[a] All reactions were performed with 0.1 mmol of 2 and 0.2 mmol of 1 in
1
1 mL THF at 158C. [b] Z/E ratio was determined by H NMR. [c] Isolated as
Wittig’s reaction product.
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
9
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
The results obtained in the reaction of 1a with the Mel-
drum’s derivative 2G (Scheme 9, top) under conditions favor-
ing the formation of the Z isomers merit special comment. In-
stead of the expected compound (Z)-5Ga, pyrone 6[14] was ex-
clusively obtained in a high yield, as a consequence of the re-
arrangement exerted by the aldehyde. This is an important
result taking into account that pyran-2-ones, such as C
(Figure 1) are important biological targets. The rearrangement
indicated in Scheme 9 could not take place from the olefin of
the E configuration because the spatial arrangement of the
formyl group would not be appropriated. Therefore, com-
pound (E)-5Ga could easily be isolated under the conditions in
Table 2. Another interesting example that highlights the impor-
tance of the configuration of the double bond is outlined at
the bottom of Scheme 9. Thus, only the Z configuration of
5Ba is able to have the appropriated configuration to cyclize
and give the spyro-compound 7, which is presented in several
natural products.[15]
Table 6. Scope of alkynals under optimal conditions for the diastereose-
lective preparation of the E isomers starting from 2B.[a,b]
[a] All reactions were performed with 0.1 mmol 2 and 0.2 mmol 1 in 1 mL
1
THF at 158C. [b] Z/E ratio was determined by H NMR. [c] The correspond-
ing Wittig reaction product was isolated.
Scheme 10. Derivatization of (E)-5Aa and X-ray structure of compound 8.
Conclusion
Scheme 9. Reaction of 2G with 1a under catalysis of 4d and reduction of
(Z)-5Ba.
In conclusion, we have presented here a highly diastereoselec-
tive synthesis of trisubstituted Z or E enals by using different
alkynals and nucleophiles as starting materials. The diastereo-
control is mainly governed by the catalyst used. Thus, reac-
tions controlled by steric effects, like Jørgensen–Hayashi’s cata-
lyst, give access to E isomers through thermodynamic control,
whereas catalysts such as tetrazol-pyrrolidine Ley’s catalyst
allow the synthesis of Z isomers, preferably obtained by means
of kinetic control. The combination of experimental and theo-
retical work points towards the protonation of allenamine in-
termediates as the key step in the case of the kinetic products
Z, whereas in the case of the E products the thermodynamic
control shifts the equilibrium to the more stable isomer. Al-
though protonation steps are generally considered as low bar-
rier nonselective processes, in the kinetic control, the forma-
tion of the putative pyrrolidinium species as protonation
agents allows the generation of distinct kinetic barriers that, in
part, depend on the different steric interactions generated in
the protonation process. Otherwise, diastereoselectivity arises
from thermodynamic control due to distinct steric repulsions
on pro-E and pro-Z iminium intermediates. The reaction allows
the synthesis of a large variety of trisubstituted-enals from dif-
ferent alkynals and 1,3-dicarbonyl derivatives, achieving in all
the cases from moderate to excellent diastereoselectivities. In
addition, by choosing the appropriate catalyst, the synthesis of
The scope of the reaction of 2B with different alkynals (1a–
i) to form the Z-5 isomers, under the conditions in entry 14 of
Table 4, was also investigated (Table 6). b-Aryl alkynals usually
provided high yields and very good selectivity, regardless the
electronic character and the position of the aryl substituents.
The reaction with 1i did not work (as has been observed in
other publications[8]), whereas 1g (R=n-pentyl) and 1h (R=cy-
clohexyl) resulted in 57:43 (56% yield) and 80:20 (37% yield)
mixtures, with the Z isomer being predominant, thereby sug-
gesting some positive influence of the size of the substituent
on the stereoselectivity, which contrasts with the tendency
shown in Table 3. Finally, the b-cyclohexenyl alkynal 1 f pro-
duced (Z)-5Bf as the only isomer with an excellent yield (89%),
which also contrasts with the scarce stereoselective evolution
of this alkynal under the conditions in Table 3.
The configurational assignment of the double bonds present
in compound 5 was unequivocally established by NMR and X-
ray diffraction studies. The configuration (E) of the double
bond was unequivocally assigned by X-ray diffraction studies
of compound 8,[16] synthesized from (E)-5Aa by Wittig reaction
with the appropriated phosphorous ylide (Scheme 10).[17]
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
10
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
2015, 54, 9954; Angew. Chem. 2015, 127, 10092; d) Y. Hasegawa, I. D.
Gridnev, T. Ikariya, Angew. Chem. Int. Ed. 2010, 49, 8157; e) G. Kang, Q.
Wu, M. Liu, Q. Xu, Z. Chen, W. Chen, Y. Luo, W. Ye, J. Jiang, H. Wu, Adv.
Synth. Catal. 2013, 355, 3131.
the Z or E enals can be achieved, making this methodology at-
tractive for the synthesis of trisubstituted double bonds.
[7] a) M. Bella, K. A. Jørgensen, J. Am. Chem. Soc. 2004, 126, 5672; b) H. E.
Zimmermann, A. Pushechnikov, Eur. J. Org. Chem. 2006, 3491; c) Q. Lan,
X. Wang, S. Shirakawa, K. Maruoka, Org. Process Res. Dev. 2010, 14, 684;
d) T. Misaki, N. Jin, K. Kawano, T. Sugimura, Chem. Lett. 2012, 41, 1675.
[8] For selected papers on the aminocatalytic functionalization of alkynals,
see: a) X. Zhang, S. Zhang, W. Wang, Angew. Chem. Int. Ed. 2010, 49,
1481; Angew. Chem. 2010, 122, 1523; b) J. Alemꢂn, A. NfflÇez, L. Marzo,
V. Marcos, C. Alvarado, J. L. Garcꢁa Ruano, Chem. Eur. J. 2010, 16, 9453;
c) J. Alemꢂn, C. Alvarado, V. Marcos, A. NfflÇez, J. L. Garcꢁa Ruano, Synthe-
sis 2011, 12, 1840; d) J. Alemꢂn, A. Fraile, L. Marzo, J. L. Garcꢁa Ruano, C.
Izquierdo, S. Dꢁaz-Tendero, Adv. Synth. Catal. 2012, 354, 1665; e) X. Cai,
C. Wang, J. Sun, Adv. Synth. Catal. 2012, 354, 359; f) X. Zhang, X. Song,
H. Li, S. Zhang, X. Chen, X. Yu, W. Wang, Angew. Chem. Int. Ed. 2012, 51,
7282; Angew. Chem. 2012, 124, 7394.
[9] For aminocatalytic systems, see: a) K. S. Halskov, B. S. Donslund, B.
Matos-Paz, K. A. Jørgensen, Acc. Chem. Res. 2016, 49, 974; for noncova-
lent interactions in organocatalysis, see: b) S. E. Wheeler, T. J. Seguin, Y.
Guan, A. C. Doney, Acc. Chem. Res. 2016, 49, 1061; c) D. M. Walden,
O. M. Ogba, R. C. Johnston, P. H.-Y. Cheong, Acc. Chem. Res. 2016, 49,
1279.
Experimental Section
Experimental procedures, complete screening tables, NMR spectra
of all new compounds, SFC chromatograms, and computational
data are reported in the Supporting Information.
Acknowledgements
Financial support from the Spanish Government (CTQ2015-
64561-R, CTQ-2012-37420-C02-02 and CTQ-2012-35957) is
gratefully acknowledged. J.A. thanks the MICINN for the
‘Ramon y Cajal’ contract. We gratefully acknowledge computa-
tional time provided by the CCC (UAM). We thank P. SanLazaro
for the initial experiments in this work.
Keywords: alkynals · aminocatalysis · double bonds · enals ·
protonation
[10] H. E. Zimmerman, J. Org. Chem. 1955, 20, 549.
[11] We have analyzed the enantioselectivity obtained in the created chiral
centers with both types of catalysts (steric and hydrogen-bond types).
However, the diastereoisomer (E or Z) obtained as the major one in
each case exhibited a low enantiomeric excess (<20%), whereas it was
much better (ee>60%) for the geometric isomers obtained as the
minor in the mixture. Additionally, the catalysts affording better Z/E ste-
reocontrol were not the same as those providing better ee. Therefore,
we have not paid attention to the enantioselectivity in the rest of the
work.
[1] For compounds with different biological properties depending on the
double bond stereochemistry, see: a) N. Redwane, H. B. Lazrek, J. L. Bar-
ascut, J. L. Imbach, J. Balzarini, M. Witvrouw, E. De Clerq, Nucleosides Nu-
cleotides Nucleic Acids 2001, 20, 1439; b) D. W. Robertson, J. A. Katzenel-
lenbogen, J. Org. Chem. 1982, 47, 2387; c) C. Arellano, B. Allal, A.
Goubaa, H. Rochꢀ, E. Chatelut, J. Pharm. Biomed. Anal. 2014, 100, 254;
d) S. Nanda, A. I. Scott, Tetrahedron: Asymmetry 2004, 15, 963; e) J. H. G.
Lago, T. M. Tanizaki, M. C. M. Young, E. F. Guimar¼es, M. J. Kato, J. Braz.
Chem. Soc. 2005, 16, 153; f) Y. Tang, R. Muthyala, R. Vince, Bioorg. Med.
Chem. 2006, 14, 5866; g) V. Devreux, J. Wiesner, H. Jomaa, J. V. Eycken,
S. V. Calenbergh, Bioorg. Med. Chem. 2007, 17, 4920; h) S. Zhou, M. N.
Prichard, J. Zemlicka, Tetrahedron 2007, 63, 9406; i) B. Modzelewska-Ba-
nachiewicz, B. Michalec, T. Kaminska, L. Mazur, A. E. Koziol, J. Banachie-
wicz, M. Ucherek, M. Kandefer-Szerszen, Monatsh. Chem. 2009, 140, 439;
j) S. Arfaie, A. Zarghi, Eur. J. Med. Chem. 2010, 45, 4013; k) L. Filippelli,
C. O. Rossi, N. A. Uccella, Colloids Surf. B 2011, 82, 13; l) M. Nagaki, T.
Ichijo, R. Kobashi, Y. Yagihashi, T. Musashi, J. Kawalami, N. Ohya, T. Goth,
H. Sagami, J. Mol. Catal. B 2012, 80, 1.
[12] The decrease of the catalyst loading to 10 or 5 mol% resulted in a de-
crease in the reactivity and the diastereoselectivity.
[13] Reaction of 2E under conditions of Table 2 gave a 28:72 mixture of Z/E
isomers. This is the only case in which these conditions did not pro-
duce the Z isomer as the major component of the reaction mixture.
[14] It is known that Meldrum ’s derivatives bearing a formyl group at the
appropriated position can undergo rearrangements resulting in a six-
membered ring, see ref. [3e].
[15] See, for example, a) H. Newman, R. B. Angier, J. Org. Chem. 1966, 31,
1456; b) H. Newman, A. Howard, R. B. Angier, J. Org. Chem. 1966, 31,
1462; c) K. C. Nicolaou, T. Montagnon, G. Vassilikogiannakis, C. J. N. Ma-
thison, J. Am. Chem. Soc. 2005, 127, 8872; d) H.-S. Yeom, Y. Lee, J.
Jeong, E. So, S. Hwang, J.-E. Lee, S. S. Lee, S. Shin, Angew. Chem. Int. Ed.
2010, 49, 1611; Angew. Chem. 2010, 122, 1655; e) E. Li, Y. Huang, L.
Liang, P. Xie, Org. Lett. 2013, 15, 3138.
[2] J. Zhang, Z. Zhang, US-2010/0210675A1.
[3] a) R. Vleggaar, Pure Appl. Chem. 1986, 58, 239; b) L. Leite, D. Jansone, M.
Veveris, H. Cirule, Y. Popelis, G. Melikyan, A. Avetisyan, E. Lukevics, Eur. J.
Med. Chem. 1999, 34, 859; c) I. J. S. Fairlamb, L. R. Marrison, J. M. Dickin-
son, F.-J. Lu, J. P. Schmidt, Bioorg. Med. Chem. 2004, 12, 4285; d) A.
Ripka, G. Shapiro, R. Chesworth, WO 2009/158393A1; e) A. Arcadi, S.
Cacchi, F. Marinelli, P. Pace, Synlett 1993, 10, 743.
[4] For reviews on the use of triple bonds in organocatalysis, see: a) A.
Fraile, A. Parra, M. Tortosa, J. Alemꢂn, Tetrahedron 2014, 70, 9145; b) R.
Salvio, M. Moliterno, M. Bella, Asian J. Org. Chem. 2014, 3, 340.
[5] For Z selectivity, see: a) Z. Wang, Z. Chen, S. Bai, W. Li, X. Liu, L. Lin, X.
Feng, Angew. Chem. Int. Ed. 2012, 51, 2776; Angew. Chem. 2012, 124,
2830; b) Z. Zhang, X. Liu, Z. Wang, X. Zhao, L. Lin, X. Feng, Tetrahedron
Lett. 2014, 55, 3797; c) Z. Wang, Z. Zhang, Q. Yao, X. Liu, Y. Cai, L. Lin, X.
Feng, Chem. Eur. J. 2013, 19, 8591; d) T. Misaki, K. Kawano, T. Sugimura,
J. Am. Chem. Soc. 2011, 133, 5695; e) Z. Chen, M. Furutachi, Y. Kato, S.
Matsunaga, M. Shibasaki, Angew. Chem. Int. Ed. 2009, 48, 2218; Angew.
Chem. 2009, 121, 2252.
[16] The structure of
8 was determined by X-ray crystal analysis:
CCDC 1043200 (8) contains the supplementary crystallographic data for
this paper. These data are provided free of charge by The Cambridge
Crystallographic Data Centre.
[17] Firstly, we observed that the d values of the aldehydic proton of the
compounds obtained as the major product under the conditions in
Tables 2 and 3, are lower than those resulting from the conditions in
Tables 5 and 6. In the case of the olefinic protons, d values of com-
pounds in Tables 2 and 3 are higher than those of Tables 5 and 6. This
allowed us to confirm that the compounds in Tables 2 and 3 have the
same configuration and were different to the compounds in Tables 5
and 6.
[6] For E selectivity, see: a) X. Wang, M. Kitamura, K. Maruoka, J. Am. Chem.
Soc. 2007, 129, 1038; b) Q. Lan, X. Wang, K. Maruoka, Tetrahedron Lett.
2007, 48, 4675; c) D. Uraguchi, K. Yamada, T. Ooi, Angew. Chem. Int. Ed.
Received: July 20, 2016
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
11
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Organic Synthesis
L. Marzo, J. Luis-Barrera, R. Mas-Ballestꢃ,
J. L. G. Ruano, J. Alemꢀn*
&& – &&
Stereodivergent Aminocatalytic
Synthesis of Z- and E-Trisubstituted
Double Bonds from Alkynals
Diastereoselectivity: A highly diastereo-
selective synthesis of trisubstituted Z- or
E-enals, which are important intermedi-
ates in organic synthesis, as well as
scribed using different alkynals and nu-
cleophiles as starting materials (see
scheme). Diastereocontrol is mainly gov-
erned by the appropriate catalyst (hy-
drogen bonding versus steric effects).
being present in natural products, is de-
&
&
Chem. Eur. J. 2016, 22, 1 – 12
www.chemeurj.org
12
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!