Asymmetric organocatalytic thio-diels-alder reactions via trienamine catalysis

Article abstract of DOI:10.1021/ja4007244

We report a new process to access highly enantioenriched sulfur-based heterocycles by an asymmetric catalytic thio-Diels-Alder reaction. Thiocarbonyls are challenging heterodienophiles in enantioselective Diels-Alder reactions, due to their inherent high reactivity and their poor ability to coordinate to chiral catalysts. We successfully circumvented these problems by employing a different strategy, which explores the use of in situ generated catalyst-bound dienes. Synthetically useful dihydrothiopyrans as well as other bi- and tricyclic sulfur-containing heterocycles are formed in high yields and high to excellent selectivities. DFT calculations were performed to examine the mechanism of the developed reaction. Furthermore, a series of synthetic transformations of the optically active sulfur-based heterocycles are presented.

Full text of DOI:10.1021/ja4007244

Article
pubs.acs.org/JACS
Asymmetric Organocatalytic Thio-Diels−Alder Reactions via
Trienamine Catalysis
Hao Jiang, David Cruz Cruz, Yang Li, Vibeke Henriette Lauridsen, and Karl Anker Jørgensen*
Center for Catalysis, Department of Chemistry, Aarhus University, Langelandsgade 140, DK-8000 Aarhus C, Denmark
S
* Supporting Information
ABSTRACT: We report a new process to access highly
enantioenriched sulfur-based heterocycles by an asymmetric
catalytic thio-Diels−Alder reaction. Thiocarbonyls are chal-
lenging heterodienophiles in enantioselective Diels−Alder
reactions, due to their inherent high reactivity and their poor
ability to coordinate to chiral catalysts. We successfully
circumvented these problems by employing a different
strategy, which explores the use of in situ generated catalyst-
bound dienes. Synthetically useful dihydrothiopyrans as well as other bi- and tricyclic sulfur-containing heterocycles are formed in
high yields and high to excellent selectivities. DFT calculations were performed to examine the mechanism of the developed
reaction. Furthermore, a series of synthetic transformations of the optically active sulfur-based heterocycles are presented.
INTRODUCTION
■
The asymmetric hetero-Diels−Alder reactions provide simple
and direct access to a variety of important, chiral, heterocyclic
compounds.1 Driven by the growing demand of these
structures in contemporary synthesis, great efforts have been
devoted to continually develop new and improved reactions
and protocols. One of the central agendas of current research in
asymmetric hetero-Diels−Alder reactions is the elaboration of
efficient and reliable enantioselective catalytic methods, thereby
maximizing the atom economy of the chemical synthesis.2 In
this regard, the research in oxo- and aza-Diels−Alder reactions
has been very fruitful, giving rise to a broad range of
complementary methods to, catalytically, furnish O- or N-
centered heterocycles in high enantioselectivities (Figure 1, top
left).3
On the contrary, the closely related, and equally useful,
catalytic asymmetric thio-Diels−Alder reaction, which leads to
the formation of sulfur-based heterocycles, has been the object
Figure 1. Diels−Alder reactions with different heteroatom centered
of much less attention (Figure 1, top right).4 It is noteworthy
dienophiles and the challenges of the enantioselective thio-Diels−
that several sulfur-based cycloadducts are known to possess
interesting biological activity (Figure 1, bottom).5 For example,
a series of acetyl CoA cholesterol acyltransferase inhibitors was
developed as lead targets for potent oral therapeutic agents
against the onset of atherosclerosis. More importantly, it was
shown that the absolute configuration of the sulfur-containing
heterocycles had a decisive effect on the activities and
bioavailability of these compounds, thus emphasizing the
need for efficient enantioselective thio-Diels−Alder reactions.
Despite the well-known high dienophilicity of thiocarbonyls,
the engagement of sulfur-based dienophiles in asymmetric [4 +
2]-cycloadditions has proven to be difficult. In fact, only few
isolated and elusive attempts have been made toward the
development of general catalytic enantioselective thio-Diels−
Alder reactions to access sulfur heterocycles.4 One of the
Alder reaction. ACAT: acetyl CoA cholesterol acyltransferase.
central challenges lies within the inherent high reactivity of the
thiocarbonyls toward simple dienes, which complicates the
design of general catalytic systems for enantioselective thio-
Diels−Alder reaction.6 Moreover, in contrary to the O- or N-
based analogues, the thiocarbonyl groups are typically
unsuitable for activation (and coordination) by Lewis and
Brønsted acids, thus, significantly reducing the number of
applicable catalysts. The most promising result achieved, up to
date, in this reaction was reported by Gulea et al. in 2010,
Received: January 25, 2013
Published: March 8, 2013
© 2013 American Chemical Society
5200
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207

Journal of the American Chemical Society
Article
where a thiocarbonyl compound was transformed with an
enantioselectivity of up to 82% ee using an organometallic
catalyst (Figure 2, top).4a However, a specific side-chain
the challenges of developing an efficient catalytic enantiose-
lective thio-Diels−Alder reaction. The fact that the two success
criteria of the design plan (chiral catalyst targeting the diene
and in situ generation of the diene) are fulfilled encouraged us
to further evaluate this hypothesis. However, it shall be noted
that aminocatalyzed hetero-Diels−Alder reactions using heter-
oatom-centered dienophiles have only in few cases been
achieved with good enantioselectivity. Moreover, up to date, all
reported trienamine-based organocatalytic Diels−Alder reac-
tions have been limited to the use of carbon-centered
dienophiles. The engagement of a heterodienophile in such
trienamine-mediated reaction pathways has not yet been
successful.
RESULTS AND DISCUSSION
■
Screening and Optimization of the Thio-Diels−Alder
Reaction. In order to validate the synthetic design, we chose to
study the [4 + 2]-cycloaddition between dienal 1a and
dithioester 2a as the model reaction (Table 1). Encouragingly,
Table 1. Screening of the Aminocatalytic thio-Diels-Alder
a
Reaction
3
additive
(mol %)
T
t
ee
(%)
b
c
d
entry (mol %)
(°C) (h) conv (%)
dr
1
20
PhCO2H
(20)
rt
1
>95 (85)
94:6
89
2
3
4
5
20
20
20
10
D-MA (20)
NaOAc (20)
None
rt
rt
rt
rt
1
48
48
3
>95
>95
>95
>95
94:6
94:6
94:6
94:6
89
89
89
89
Figure 2. Summary of previous work and the design plan of this work.
PhCO2H
(20)
containing a functional group, which enabled strong bidentate
coordination to the metal center, was shown to be crucial for
the selectivity. Any derivation from this catalyst recognizing
motif in the thio-dienophile led to a significant reduction in the
enantioselectivity of the catalytic enantioselective thio-Diels−
Alder reaction.
6
7
10
5
PhCO2H
(20)
4
4
24
72
>95
94:6
94:6
92
92
PhCO2H
(20)
>95 (89)
a
Reactions were performed with 0.1 mmol 1a and 0.12 mmol 2a in the
b
presence of catalyst 3 and the given additives in CHCl3. Determined
by 1H NMR on the crude reaction mixture. The values in parentheses
The lack of more general and effective protocols triggered us
to seek alternative strategies, which do not rely on catalyst
coordination to thiocarbonyl compounds as a mean to achieve
good enantiocontrol. Instead, we questioned the possibility of
using a chiral catalyst to in situ generate activated diene species
from reactants otherwise inert to the thio-Diels−Alder reaction
(Figure 2). As such, the presence of any nonstereoselective
background reactions can be minimized, and only in its catalyst-
bound state, the diene fragment is available for reaction.
Optimal influence of the chiral catalyst is thereby warranted, as
the active diene is always covalently linked to the catalyst,
which may govern the facial approaches of the dienophile.
In recent years, aminocatalysis has evolved to become a
useful tool to promote enantioselective Diels−Alder reactions.7
In particularly, we have been fascinated by the ability of an
aminocatalyst to in situ generate trienamine species, which
subsequently were able to engage as catalyst-bound dienes in
asymmetric Diels−Alder reactions.8 At the outset of the current
study, we hypothesized that this organocatalytic remote
functionalization strategy might be a viable approach to address
c
1
are yields of isolated product. Determined by H NMR on the crude
d
reaction mixture. Determined by chiral ultraperformance conver-
gence chromatography (UPC2). D-MA = D-mandelic acid.
in the presence of 20 mol % of catalyst 3 and benzoic acid in
CHCl3, the dienal 1a successfully reacted, via a catalyst-bound
trienamine intermediate, to form the corresponding dihydro-
thiopyrane 4a in 85% yield, 94:6 dr and 89% ee (Table 1, entry
1). While the presence of a weak base or no additive led to
longer reaction times, the use of D-mandelic acid resulted in
similar results compared to that of benzoic acid. A small
increase of the enantioselectivity to 92% ee could be achieved
by performing the reaction at lower temperatures (compare
entries 5,6). Furthermore, the catalyst loading could be reduced
to 5 mol %, while maintaining the high yield and selectivities
obtained (entry 7). However, longer reaction times were
required to provide full conversion to the product.
Scope of the Diene Precursors in the Thio-Diels−
Alder Reaction. Next, a series of aldehydes 1 was evaluated as
5201
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207

Journal of the American Chemical Society
Article
diene precursors in the developed thio-Diels−Alder reaction.
Due to small differences in reactivity toward the dithioester 2a,
the reaction conditions were adjusted for each of the substrates
with the aim to obtain the best result with minimum catalyst
loading (Scheme 1). High to excellent selectivities were
the reaction of highly polarized dienes with phosphonodithio-
formate have shown similar site preference.4i A theoretical
study of this system was recently reported.9b However, the
regioselectivity in these studies proved to be substrate and
condition dependent. Moreover, Vedejs and Houk studied the
regioselectivity of Diels−Alder reactions involving thioalde-
hydes using ab initio molecular orbital calculations, suggesting
that substituents have a large effect on the regioselectivity of the
reaction.9a In order to obtain a better understanding of the
reactivity of dithioesters in the developed aminocatalyzed
Diels−Alder reactions, a computational study employing DFT
calculations was undertaken. All calculations were performed at
wB97xd/tzvp10 level of theory and basis set at 298.15 K and 1.0
atm pressure using Gaussian 09.11 The solvent effects were
simulated by the integral equation formalism polarizable
continuum model (IEFPCM, CHCl3).12 All listed energies
are Gibbs free energies. The zero-point energies (ZPEs) and
thermochemical corrections were obtained by frequency
calculations, which also validated all stationary points as either
local minima or first-order saddle points.
Scheme 1. Scope of the Asymmetric Aminocatalytic thio-
Diels-Alder Reaction
In Figure 3, the LUMO and the charge distribution
(chelpg)13 of methyl 2-(methylthio)-2-thioxoacetate 2b are
Figure 3. LUMO and charge distribution (chelpg) of a simple thio-
dienophile 2b.
depicted. We see that the LUMO is only slightly larger on the
sulfur-atom when compared to the carbon-atom of the
thiocarbonyl group. In addition, a relatively small charge
separation of 0.256 is observed across the CS bond, with the
sulfur atom being slightly more negatively charged.
In order to understand the experimentally observed
regioselectivity, the energy profiles of the two aforementioned
reaction pathways (with either the sulfur or the carbon atom as
the electrophilic site) were examined. In Figure 4, the free
energy profile of the observed thio-Diels−Alder reaction
leading to the formation of dihydrothiopyranes 4 is outlined.
The reaction between dithioester 2b and trienamine A was
chosen as the model system. In this reaction, 2b has two
possible ways to approach the trienamine A, namely path I
(ester group endo) or path II (ester group exo). These pathways
should lead to the formation of two diastereomers, CI and CII,
as the final products. TS1-I and TS1-II of the initial
intermolecular reactions were identified as the highest energy
transition states, located at 13.9 and 15.0 kcal/mol. Intrinsic
reaction path (IRC)14 calculations connected TS1-I and TS1-II
to two high-energy zwitterionic intermediates BI and BII. The
final intramolecular cyclization involves very low barriers for
both pathways. Both reactions are exothermic; however, the exo
product CII was identified to be slightly more stable (by 0.6
kcal/mol). Nevertheless, due to the large energetic barriers of
the backward reactions (>20 kcal/mol), the final product
distribution is most likely to be kinetically controlled. In this
achieved in all evaluated cases affording highly enantioenriched
heterocyclic products in high yields. Both alkyl and aryl
substituents were well-tolerated at the γ- and δ-positions of 1 as
demonstrated for the formation of 4a−d in 78−95% yield,
90:10 to >95:5 dr and 85−93% ee. Bi- and tricyclic products,
such as 4e,f, could also be formed in excellent enantioselectiv-
ities. Slightly lower reactivity was observed for trienamines
generated from linear dienals such as hexa- and octadienal,
which is consistent with previous literature.8d Nevertheless, by
increasing the reaction temperature to 40 °C, comparable
results could be achieved (see 4g,h).
Mechanistic Investigations and Rationalization for
the Observed Regio-, Diastereo-, and Enantioselectiv-
ities. While delighted by the obtained results and successful
validation of our hypothesis, we were also intrigued by the
observed selectivities of the thio-Diels−Alder reaction. Of
particular interest to us was the regioselectivity of the reaction
between the unsymmetrical dienes and thiocarbonyl 2a. Our
results demonstrate that such a process proceeds in favor of the
pathway involving an initial attack of the catalyst-bound
trienamine species at the sulfur atom. Previous reports studying
5202
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207

Journal of the American Chemical Society
Article
Figure 4. Energy profile for reaction path I−II: reaction via an initial attack at the sulfur atom. The values are Gibbs free energies in kcal/mol, relative
to A + 2b [wB97xd/tzvp/IEFPCM].
regard, the anticipated endo product CI is favored by 1.0 kcal/
mol over its diastereomer CII. According to the Boltzmann
equation, such an energy difference should correlate to a 85:15
(endo/exo) product distribution, which is in agreement with
experimental observations (79:21 dr). It should also be noted
that interconversion between intermediates BI and BII via a
rotation around the C−S single bond is not possible due to
steric reasons.
Next, the alternative and experimentally not observed
cycloaddition reactions involving an initial attack at the carbon
center was investigated (Figure 5). The transition states TS-III
and TS-IV involving the two possible approaches (ester group
endo or exo, respectively) of 2b to A were identified. IRC
calculations show a distinct shoulder on the path to the
products D. However, no intermediate could be located,
suggesting an asynchronous concerted mechanism. In terms of
thermodynamic stability, the products D are favored over their
regioisomers C. However, the calculated transition states (TS-
III and TS-IV) leading to these cycloadducts (DIII and DIV)
were found to be >5.5 kcal/mol higher in energy than TS-I and
TS-II. This large discrepancy in activation barriers is in good
agreement with experimental results, in which the regioisomers
arising from path III−IV were not detected.15
The transition-state structures for the initial intermolecular
reactions for path I−IV are depicted in Figure 6. Important
bond lengths in the dithioester (CS, C−C, CO) as well as
the distances of the forming bonds have been compared
(numbers in italic). In TS1-I and TS1-II, the initial attack of the
trienamine on the sulfur atom leads to increasing conjugation in
the thio-dienophile, which is shown by the shortening of the
C−C bonds, while the CO bonds are lengthened (compared
Figure 5. Energy profile for reaction path III−IV: reaction via an initial
attack at the carbon atom. The values are Gibbs free energies in kcal/
mol, relative to A + 2b [wB97xd/tzvp/IEFPCM].
Figure 6. Transition-state geometries for path I−IV. The numbers in
italic are bond distances in angstroms. The nonitalic values are free
energies in kcal/mol, relative to A + 2b [wB97xd/tzvp/IEFPCM].
5203
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207

Journal of the American Chemical Society
Article
to the distances in the parent 2b). The negative charge, which is
built up, in the transition state is stabilized by delocalization. In
contrast, the intermolecular C−C bond formation in TS-III
and TS-IV localizes the negative charge solely on the sulfur
atom. Compared to TS1-I and TS1-II, the C−C and CS
bond distances in TS-III and TS-IV are longer by 0.05 and 0.03
Å, respectively. Evidently, the charge buildup in TS-III and TS-
IV is much less stabilized, resulting in the apparent higher
activation energies in these transition states.
By computationally investigating the reaction of trienamine A
with three different thio-dienophiles (E, F and G) via the two
alternating pathways (initial S- or C-attack), the influence of the
substituents on the “S versus C” regioselectivity was examined
(Figure 7). In the case of compound E, in which the
(an ester and a SMe group) warrants complete selectivity for
the experimentally observed reaction pathway.
The obtained high enantioselectivities of the thio-Diels−
Alder reaction reflect the ability of the chiral TMS-protected
diphenylprolinol to provide facial differentiation when engaged
as the catalyst-bound trienamine intermediate. The approaches
of dienophile 2b to both faces of this trienamine (formed from
1a and 3) were studied. The structures were optimized
employing wB97xd and the 6-31G(d) basis set in CHCl3
(IEFPCM) (Figure 8). Activation energies were obtained
Figure 7. Transition-state geometries for the reaction of trienamine A
and thiocarbonyl E/F/G via two possible pathways. The numbers in
italic are bond distances in angstroms. The nonitalic values are free
energies in kcal/mol, relative to A + E/F/G [wB97xd/tzvp/IEFPCM].
Figure 8. Transition-state geometries for the favored and the
disfavored approaches of 2b to the catalyst-bound trienamine
intermediate. The values in italic are bond distances in angstroms.
The nonitalic numbers are free energies in kcal/mol [wB97xd/cc-
pVDZ/IEFPCM//wB97xd/6-31G(d)/IEFPCM].
thiocarbonyl is only flanked by hydrogen atoms, an initial
attack of A at the carbon atom is strongly favored (ΔΔG‡ = 7.4
kcal/mol). Substituting one of the hydrogen atoms with a SMe
group increases the relative preference of the S-attack pathway,
although TS-FC‑attack is still 4.7 kcal/mol lower in energy than
TS-FS‑attack. The incorporation of the CO2Me-substituent
proved to have a decisive effect on the regioselectivity. When
the approach of G to the trienamine intermediate A was
studied, reaction via the S-attack pathway was found to be the
lowest energy pathway. Presumably, a thiocarbonyl compound
carrying an adjacent ester group (such as G and 2b) can be
regarded as the sulfur equivalent of a Michael acceptor.
Similarly, the attack of a nucleophilic species on the
electrophilic sulfur atom in these thiocarbonyls may resemble
a 1,4-addition, in which the electron-withdrawing ester group
assists in delocalizing the negative charge buildup in the
transition state (and product). In summary, electronic factors
are responsible of the regioselectivity of the trienamine
mediated thio-Diels−Alder reaction. The ester group plays a
major role in directing the reaction toward an initial attack of
the trienamine at the sulfur atom, while the SMe group
provides a minor contribution. The effects of the substituents
are not additive; however, the combination of the two groups
with wB97xd/cc-pVDZ16 single points.17 The favored approach
of 2b to the trienamine species results in an activation energy,
which is 1.4 kcal/mol lower than that of the disfavored
approach (TS1-F vs TS1-DF). This energy difference
corresponds to a calculated enantioseletivity of 83% ee, which
is in good agreement with the experimental value (79% ee).
Experimental Investigation on the Influence of
Substituents and Extended Reaction Scope. The obtained
computational insights on substituent effects in the regiose-
lectivity of the reaction were experimentally validated by
employing methyl benzodithioate as thiodienophile in the
developed organocatalytic thio-Diels−Alder reaction. Exchang-
ing the ester group with the phenyl ring completely shut down
the reaction, hence confirming the importance of the EWG in
2. Furthermore, the influence of substituent size (in 2) on the
reactivity and selectivity of the thio-Diels−Alder reaction was
examined (Table 2). It was anticipated that the gained
knowledge could lead to further development of this reaction,
i.e., improving the reaction diastereo- and enantioselectivity.
Several thiodienophiles carrying different EWG and R groups
5204
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207

Journal of the American Chemical Society
Article
Table 2. Experimental Variation of Substituents on the
Thiodienohile and the Influence of the Stereoselectivities
Scheme 2. Extended Scope of the Asymmetric
Aminocatalytic Thio-Diels-Alder Reaction
a
b
c
d
entry
2
EWG
R
yield (%)
dr
ee (%)
1
2
3
4
5
2a
2c
2d
2e
2f
CO2Bn
CO2Bn
CO2Bn
CO2iPr
COPh
Et
Me
Bn
Et
4a −93
4i − 73
4j − 73
4k − 96
4l − 87
94:6
92:8
89
84
90:10
97:3
87
94
Et
65:35
6/53
a
Reactions were performed with 0.1 mmol 1a, 0.12 mmol 2, 0.02
b
mmol 3, and 0.02 mmol PhCO2H in CHCl3. Yields of isolated
c
1
products. Determined by H NMR on the crude reaction mixture.
d
Determined by chiral UPC2.
were tested and the results are summarized in Table 2. Altering
the size of R substituent on the thioester affected the
enantioselectivity of the reaction (compare entries 1−3). This
effect is anticipated, since the R group clashes with the OTMS
group of the catalyst when the thiocarbonyl approaches from
the disfavored face (see TS1-DF). By exchanging the ethyl
group with a smaller methyl group, a 5% loss of
enantioselectivity was observed, while substituting with a larger
Bn group only had minor effects. To our delight, further
improvement of the selectivities could be achieved by
employing a more bulky iPr-ester as the EWG. As result, the
cycloadduct 4k was formed in 97:3 dr and 94% ee at rt.
Surprisingly, by exchanging the ester moiety on the dithioester
with an phenyl ketone a disappointing 65:35 dr was obtained.
More importantly, the formation of 4l proceeded with almost
no (or poor) enantioselectivities for the diastereomeric pair.
This dramatic decrease in selectivities can presumably be
rationalized by the conformation adopted by 2f. In order to
minimize steric clashes between the Ar−H with the sulfur
atoms, the CO and SO groups are forced to be oriented
close to perpendicularity. The resulting O−C−C−S dihedral
angle is significantly different from those estimated for 2a−e,
which may influence the approach of 2f to the catalyst-bound
trienamine. Finally, the optimized dithioester 2e was evaluated
under standard reaction conditions with several dienals. In all
except one case (4q), a general trend of similar or improved
selectivities was observed (compared to the use of 2a). The
results are summarized in Scheme 2.
Figure 9. X-ray structure of compound 4k.
compound 4p as starting material, the tetracyclic compound 5b
could be afforded (Scheme 3, eq II). Alternatively, an
intramolecular lactonization can be promoted by treating the
corresponding alcohol of 4r with p-TSA. Noticeably, the
sterically hindered iPr-esters is readily transesterified at 50 °C
in CHCl3 to give the six-membered fused lactone product 6 in
59% yield over two steps (Scheme 3, eq III). The endocyclic
double bonds in the obtained sulfur-based cycloadducts are
useful handles for further functionalizations, as demonstrated
for the formation of diol 7 by osmium-catalyzed dihydrox-
ylation in 71% yield and >20:1 dr.
In order to access a broader scope of enantioenriched
thioheterocycles, other strategies to generate catalyst-bound
dienes for the desired asymmetric thio-Diels−Alder reactions
were explored. Recently, our laboratory revealed that cyclic
dienals, such as 1i, were able to selectively react via catalyst-
bound cross-trienamine species. We envisioned that this
complementary aminocatalyzed cross-trienamine pathway
might be compatible with the thio-Diels−Alder reaction. To
our delight, such a strategy can be realized providing a facile
route to sulfur-containing bridged bicycles. This was
exemplified by the cycloaddition of 1i with 2b to form
compound 8 in 72% yield, 72:28 dr, and >96% ee (Scheme 4).
It should be noted that the addition of water had a decisive
effect on the conversion of the reaction. Lower substrate
The absolute configuration of the dihydrothiopyrane 4k was
unambiguously established to be (2S,3R) by X-ray crystallog-
raphy (Figure 9).18 The configurations of the remaining
products 4 were assigned by analogy.
Synthetic Transformations of the Chiral Products. The
obtained highly enantioenriched thioheterocycles readily
engage in straightforward synthetic manipulations, which
allow the access to more complex sulfur-containing chiral
molecules. Two cyclization strategies affording products of
different ring sizes are demonstrated in Scheme 3. Upon
reduction of the aldehyde functionality in 4r to the
corresponding alcohol, the subsequent addition of N-
iodosuccinimide triggered a rapid O-cyclization to furnish the
bicycle 5a in 92% ee as a single diastereomer (Scheme 3, eq I).
Using the same setup and by employing the indole-based
5205
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207

Journal of the American Chemical Society
Article
explained based on the findings of our computational study.
Finally, several synthetic transformations of the optically active
sulfur-based heterocycles are presented affording complex
sulfur-containing chiral molecules.
Scheme 3. Synthetic Transformations of the
Enantioenriched Sulfur-Based Heterocycles
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures, analytical data, computational details,
NMR spectra, chiral UPC2 traces, and complete ref 11. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
kaj@chem.au.dk
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was made possible by grants from Aarhus University,
Carlsberg Foundation and FNU. D.C.C. acknowledges
■
́
CONACYT (Mexico) for a postdoctoral fellowship. The
authors thank Dr. Jacob Overgaard for performing the X-ray
analysis. Thanks to Dr. R. L. Davis for the fruitful discussions
and for help and suggestions regarding the computational
section.
REFERENCES
■
Scheme 4. Asymmetric thio-Diels-Alder reaction via a in-situ
generated cross-conjugated trienamine
(1) For selected reviews for the application of hetero-Diels−Alder
reactions in total synthesis, see (a) Schmidt, R. R. Acc. Chem. Res.
1986, 19, 250. (b) Boger, D. L.; Winreb, S. N. Hetero Diels-Alder
Methodology in Organic Synthesis; Academic Press: San Diego, 1987.
(c) Trost, B.; Fleming, I. Comprehensive Organic Synthesis; Pergamon
Press: Oxford, U.K., 1991; Vol. 2, p 661. (c) Nicolaou, K. C.; Snyder,
S. A.; Montagnon, T.; Vassilikogiannakis, G. Angew. Chem., Int. Ed.
2002, 41, 1668. (d) Osborn, H. M. I.; Coisson, D. Mini-Rev. Org.
Chem. 2004, 1, 41.
(2) For selected general on catalytic asymmetric hetero-Diels−Alder
reactions, see: (a) Jørgensen, K. A. Angew. Chem., Int. Ed. 2000, 39,
3558. (b) Lin, G.-Q.; Li, Y.-M.; Chan, A. S. C. Principles and
Applications of Asymmetric Synthesis; Wiley: NY, 2001, 290. (c) Corey,
E. J. Angew. Chem., Int. Ed. 2002, 41, 1650. (d) Kouznetsov, V. V.
Tetrahedron 2009, 65, 2721. (e) Pellissier, H. Tetrahedron 2009, 65,
2839. (f) Girling, P. R.; Kiyoi, T.; Whiting, A. Org. Biomol. Chem.
2011, 9, 3105.
(3) For selected examples catalytic enantioselective normal-electron
demand oxo-Diels−Alder reactions, see: (a) Thompson, C. F.;
Jamison, T. F.; Jacobsen, E. N. J. Am. Chem. Soc. 2000, 122, 10482.
(b) Doyle, M. P.; Phillips, I. M.; Hu, W. J. Am. Chem. Soc. 2001, 123,
5366. (c) Long, J.; Hu, J.; Shen, X.; Ji, B.; Ding, K. J. Am. Chem. Soc.
2002, 124, 10. (d) Yamashita, Y.; Saito, S.; Ishitani, H.; Kobayashi, S. J.
Am. Chem. Soc. 2003, 125, 3793. (e) Unni, A. K.; Takenaka, N.;
Yamamoto, H.; Rawal, V. H. J. Am. Chem. Soc. 2005, 127, 1336.
(f) Lin, L.; Fan, Q.; Qin, B.; Feng, X. J. Org. Chem. 2006, 71, 4141.
(g) Yu, Z.; Liu, X.; Dong, Z.; Xie, M.; Feng, X. Angew. Chem., Int. Ed.
2008, 47, 1308. (h) Jensen, K. H.; Sigman, M. S. Angew. Chem., Int. Ed.
2007, 46, 4748. For aza-DA, see: (i) Ishihara, K.; Miyata, M.; Hattori,
K.; Tada, T.; Yamamoto, H. J. Am. Chem. Soc. 1994, 116, 10520.
(j) Kobayashi, S.; Kusakabe, K.; Komiyama, S.; Ishitani, H. J. Org.
turnover was observed in the absence of additional water,
presumably due to the collapse of a possible zwitterionic
species on the iminium ion leading to catalyst inhibition.
CONCLUSIONS
■
In conclusion, we have reported a new route to access highly
enantioenriched sulfur-based heterocycles by asymmetric
catalysis. Thiocarbonyls have been challenging substrates in
enantioselective Diels−Alder reactions, due to their inherent
high reactivity and their lack of ability to coordinate to chiral
catalysts. We successfully circumvented these problems by
employing a different strategy, which is based on the use of in
situ generated catalyst-bound dienes. In the presence of a chiral
aminocatalyst, otherwise inert substrates are transformed into
active reactants, which readily undergo stereocontrolled thio-
Diels−Alder reactions with the reactive thiocarbonyls. Syntheti-
cally useful dihydrothiopyrans as well as other bi- and tricyclic
sulfur-containing heterocycles are formed in high yields and
high to excellent selectivities with up to 97% ee. DFT
calculations were performed to examine the mechanism of
the developed reaction, which is believed to proceed via a
stepwise mechanism involving zwitterionic intermediates. The
regio- and the diastereoselectivity of the reaction were also
́
Chem. 1999, 64, 4220. (k) Mancheno, O. G.; Arrayas, R. G.; Carretero,
̃
J. C. J. Am. Chem. Soc. 2004, 126, 456. (l) Mandai, H.; Mandai, K.;
Snapper, M. L.; Hoveyda, A. H. J. Am. Chem. Soc. 2008, 130, 17961.
́ ́
(m) Sunden, H.; Ibrahem, I.; Eriksson, L.; Cordova, A. Angew. Chem.,
Int. Ed. 2005, 44, 4877. (n) Rueping, M.; Azap, C. Angew. Chem., Int.
Ed. 2006, 45, 7832. (o) Tambar, U. K.; Lee, S. K.; Leighton, J. L. J. Am.
Chem. Soc. 2010, 132, 10248.
5206
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207

Journal of the American Chemical Society
Article
(4) For examples of catalytic enantioselective thio-Diels−Alder
reactions with thiodienophiles, see: (a) Dentel, H.; Chataigner, I.; Le
Cavelier, F.; Gulea, M. Tetrahedron Lett. 2010, 51, 6014. For
thiodienes, see: (b) Saito, T.; Takekawa, K.; Nishimura, J.; Kawamura,
M. J. Chem. Soc., Perkin Trans. 1 1997, 2957. (c) Saito, T.; Takekawa,
K.; Takahashi, T. Chem. Commun. 1999, 1001. For diastereoselective
examples incorporating thiodienophiles, see: (d) Friedrich, J. D. J. Org.
Chem. 1987, 52, 2442. (e) Vedejs, E.; Stults, J. S.; Wilde, R. G. J. Am.
Chem. Soc. 1988, 5452. (f) Takahashi, T.; Kurose, N.; Koizumi, T.
Heterocycles 1993, 36, 1601. (g) Timoshenko, V. M.; Siry, S. A.;
Rozhenko, A. B.; Shermolovich, Y. G. J. Fluorine Chem. 2010, 131, 172.
(h) Dentel, H.; Chataigner, I.; Lohier, J.-F..; Gulea, M. Tetrahedron
2012, 68, 2326. For nonasymmetric synthesis (with catalyst), see:
(i) Heras, M.; Gulea, M.; Masson, S.; Philouze, C. Eur. J. Org. Chem.
2004, 160. (j) Bastin, R.; Albadri, H.; Gaumont, A. G.; Gulea, M. Org.
Lett. 2006, 8, 1033. (k) Mikhailichenko, S. S.; Bouillon, J. P.; Besson,
T.; Shermolovich, Y. G. Tetrahedron Lett. 2010, 51, 990.
(10) (a) Chai, J.-D.; Head-Gorden, M. Phys. Chem. Chem. Phys. 2008,
10, 6615. (b) Schaefer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994,
100, 5829.
(11) Frisch, M. J.; et al. Gaussian 09, revision A.02, Gaussian, Inc.:
Wallingford, CT, 2009.
(12) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999.
(13) Breneman, C. M.; Wiberg, K. B. J. Comput. Chem. 1990, 11, 361.
(14) (a) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523.
(b) Fukui, K. Acc. Chem. Res. 1981, 14, 363.
(15) In previous reports, a mixture of products, arising from both the
Cattack and the Sattack pathways has been observed; see ref 4i.
(16) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007.
(17) Other methods and basis set were also evaluated; however, none
of these provided energies which were in better agreement with the
experimental results. See Supporting Information for details.
(18) See Supporting Information for details. CCDC 909665 contains
the supplementary crystallographic data.
(5) Wilde, R. G.; et al. Bioorg. Med. Chem. 1996, 4, 1493.
(6) For examples of uncatalyzed thio-Diels−Alder reactions at room
temperature, see (a) Laduron, F.; Nyns, C.; Janousek, Z.; Viehe, H. G.
J. Prakt. Chem. 1997, 339, 697. (b) El-Sayed, I.; Ali, O. M.; Fischer, A.;
Senning, A. Heteroatom Chem. 2003, 14, 170. (c) Heras, M.; Gulea,
M.; Masson, S. Chem. Commun. 2001, 611 see also: refs 4g and i.
(7) For selected reviews on aminocatalysis, see: (a) Mukherjee, S.;
Yang, J. W.; Hoffmann, S.; List, B. Chem. Rev. 2007, 107, 5413.
(b) MacMillan, D. W. C. Nature 2008, 455, 304. (c) Barbas, C. F., III
Angew. Chem., Int. Ed. 2008, 47, 42. (d) Melchiorre, P.; Marigo, M.;
Carlone, A.; Bartoli, G. Angew. Chem., Int. Ed. 2008, 47, 6138.
(e) Dondoni, A.; Massi, A. Angew. Chem., Int. Ed. 2008, 47, 4638.
(f) Ramachary, D. B.; Reddy, Y V. Eur. J. Org. Chem. 2012, 865.
(g) Jensen, K. L.; Dickmeiss, G.; Jiang, H.; Albrecht, Ł.; Jørgensen, K.
A. Acc. Chem. Res. 2012, 45, 248. (h) Arceo, E.; Melchiorre, P. Angew.
Chem., Int. Ed. 2012, 51, 5290. (i) Li, J.-L.; Liu, T.-Y.; Chen, Y.-C. Acc.
Chem. Res. 2012, 45, 1491.
(8) For reviews on trienamine-mediated reactions, see: (a) Kumar, I.;
Ramaraju, P.; Mir, N. A. Org. Biomol. Chem. 2013, 11, 709. For
examples of aminocatalytic reactions via trienamine or related
intermediates, see: (b) Jia, Z.-J.; Jiang, H.; Li, J.-L.; Gschwend, B.;
Li, Q.-Z.; Ying, X.; Grouleff, J.; Chen, Y.-C.; Jørgensen, K. A. J. Am.
Chem. Soc. 2011, 133, 5053. (c) Jiang, H.; Gschwend, B.; Albrecht, Ł.;
Hansen, S. G.; Jørgensen, K. A. Chem.Eur. J. 2011, 17, 9093. (d) Jia,
Z.-J.; Zhou, Q.; Zhou, Q.-Q.; Chen, P.-Q.; Chen, Y.-C. Angew. Chem.,
Int. Ed. 2011, 50, 8638. (e) Liu, Y.; Nappi, M.; Arceo, E.; Vera, S.;
Melchiorre, P. J. Am. Chem. Soc. 2011, 133, 15212. (f) Xiong, X.-F.;
Zhou, Q.; Gu, J.; Dong, L.; Liu, T.-Y.; Chen, Y.-C. Angew. Chem., Int.
Ed. 2012, 51, 4401. (g) Liu, Y.; Nappi, M.; Escudero-Adan, E. C.;
Melchiorre, P. Org. Lett. 2012, 14, 1310. (h) Halskov, K. S.; Johansen,
T. K.; Davis, R. L.; Steurer, M.; Jensen, F.; Jørgensen, K. A. J. Am.
Chem. Soc. 2012, 134, 12943. (i) Albrecht, Ł.; Acosta, F. C.; Fraile, A.;
Albrecht, A.; Christensen, J.; Jørgensen, K. A. Angew. Chem., Int. Ed.
2012, 51, 9088. (j) Jiang, H.; Rodriguez-Escrich, C.; Johansen, T. K.;
Davis, R. L.; Jørgensen, K. A. Angew. Chem., Int. Ed. 2012, 51, 10271.
(k) Ma, C.; Jia, Z.-J.; Liu, J.-X.; Zhou, Q.-Q.; Dong, L.; Chen, Y.-C.
Angew. Chem., Int. Ed. 2013, 52, 948.
(9) (a) Vedejs, E.; Perry, D. A.; Houk, K. N.; Rondan, N. G. J. Am.
Chem. Soc. 1983, 105, 6999. (b) Wang, Y.; Zeng, X.-L.; Shen, S.-J.
Chinese J. Struct. Chem. 2008, 27, 513. For theoretical studies of thio-
Diels−Alder reactions between butadiene and thiocarbonyls, see:
(c) Barone, V.; Arnaud, R. Chem. Phys. Lett. 1996, 251, 393.
(d) Barone, V.; Arnaud, R.; Chavant, P. Y.; Vallee
1996, 61, 5121. For selected reviews on theoretical and kinetic
studies of the Diels−Alder reaction, see: (e) Houk, K. N.; Gonzalez, J.;
Li, Y. Acc. Chem. Res. 1995, 28, 81. (f) Kumar, A. Chem. Rev. 2001,
101, 1. For a review on theoretical studies on organicatalytic
reactions, see: (g) Cheong, P. H.-Y.; Legault, C. Y.; Um, J. M.; C
́
, Y. J. Org. Chem.
́
̧
elebi-
̈
Olcu
̧
m, N.; Houk, K. N. Chem. Rev. 2011, 111, 5042. After of the
̈
submission of this work, the following computational work has been
published: (h) Dieckmann, A.; Breugst, M.; Houk, K. N. J. Am. Chem.
Soc. 2013, 135, 3237.
5207
dx.doi.org/10.1021/ja4007244 | J. Am. Chem. Soc. 2013, 135, 5200−5207