A General Asymmetric Formal Synthesis of Aza-Baylis–Hillman Type Products under Bifunctional Catalysis

Article abstract of DOI:10.1002/chem.201705218

A new organocatalytic strategy for the synthesis of enantioenriched aza-Baylis–Hillman type products via a frustrated vinylogous reaction is presented. This process proceeds under mild conditions with good yields, completed Z/E selectivity and excellent e

Full text of DOI:10.1002/chem.201705218

DOI: 10.1002/chem.201705218
Communication
&
Bifunctional Catalysis
A General Asymmetric Formal Synthesis of Aza-Baylis–Hillman
Type Products under Bifunctional Catalysis
Marꢀa Frꢀas,[a] Ana Cristina Carrasco,[a] Alberto Fraile,[a, b] and Josꢁ Alemꢂn*[a, b]
Abstract: A new organocatalytic strategy for the synthesis
of enantioenriched aza-Baylis–Hillman type products via a
frustrated vinylogous reaction is presented. This process
proceeds under mild conditions with good yields, com-
pleted Z/E selectivity and excellent enantioselectivities.
Moreover, easy derivatizations of the final products led to
important building blocks of organic synthesis such as
1,3-aminoalcohols and Lewis base catalysts.
Carbon–carbon bond formation is one of the most important
reactions in organic chemistry, especially the Csp2ÀCsp3 bonds.
In this area, the asymmetric aza-Baylis–Hillman reaction (aza-
BHR),[1] represents the most straightforward methodology for
the synthesis of chiral allylic amines, which have been used as
starting materials or as building blocks for the synthesis of dif-
ferent pharmaceuticals and natural products.[2] Even though
Scheme 1. Known aza-BHR and the present work.
this is a well-known reaction, only a few examples in the asym-
metric field have been reported and most of these are related
to the use of non-substituted double bonds and ketones and
esters as EWGs. Since the pioneering asymmetric aza-Baylis–
Hillman reaction published by Shi and co-workers using tosyli-
mines,[3] excellent and brilliant asymmetric organocatalyzed ex-
amples have been shown by Masson and Zhu,[4a–c] Hatakeya-
ma,[4d] Sasai,[4e,f] and Jacobsen,[4g,h] as well as asymmetric metal
catalysis reported in the works of Shibata[4i] and Shibasaki,[4j]
amongst others.[4k,l] All these examples have shown that the re-
action can only take place with non-substituted double bonds
and mostly with ketone and esters (e.g. acrylates or vinylmeth-
yl ketones; top, Scheme 1). The lack of reactivity of the mono-
b-substituted and b,b-disubstituted double bonds makes the
synthesis of these enantioenriched tri- and tetra-substituted
double bonds with different electron-withdrawing groups in
the Baylis–Hillman reaction difficult. In addition, most of these
examples have employed aryl-tosylimines as the starting mate-
rial, in which alkyl imines or ketimines were unreactive, or led
to moderate enantioselectivities.[4]
More recently, List and Carretero’s groups have shown ele-
gant organocatalytic and metal-catalyzed methods, respective-
ly, to functionalize the 1,5-positions of silyldienolate derivatives
through
a
vinylogous
Mukaiyama–Mannich
reaction
(Scheme 1, middle).[5] These remarkable examples showed that
final aldehydes and esters can be selectivity funcionalized at
the 1,5-positions from moderate to good enantioselectivies.
The orbital coefficients and electrophilic susceptibility are
mainly responsible for this reactivity[6] provoking the observed
1,5-nucleophilic attack. Very recently our group has shown that
1,5- can be easily switched to 1,3-funcionalization using bifunc-
tional catalysis.[7] This provokes a dramatic change in the regio-
selectivity, from the 1,5 to the 1,3-functionalization. This varia-
tion enables the 1,3-addition of silyl-dienol ethers to nitroal-
kenes for the synthesis of tri- and tetra-substituted double
bonds in Rauhut–Currier type products. It would be highly de-
sirable if a catalyst could change to the 1,3-selectivity, and the
double bond of the intermediate obtained could be isomer-
ized, leading to Baylis–Hillman type products, which are excel-
lent building blocks for the synthesis of complex molecules
(bottom, Scheme 1). In this work, for the first time the addition
of silyl-dienol ethers to imines, catalyzed by bifunctional cata-
lysts to obtain any kind of aza-Baylis–Hillman products with
high ee values is achieved. In addition, a rational mechanistic
pathway based on mechanistic experiments is presented.
[a] M. Frꢀas, A. C. Carrasco, Dr. A. Fraile, Dr. J. Alemꢁn
Organic Chemistry Department, Mꢂdulo 1
Universidad Autꢂnoma de Madrid
Madrid-28049 (Spain)
E-mail: jose.aleman@uam.es
Homepage: www.uam.es/jose.aleman
[b] Dr. A. Fraile, Dr. J. Alemꢁn
Institute for Advanced Research in Chemical Sciences (IAdChem)
Universidad Autꢂnoma de Madrid
28049 Madrid (Spain)
Supporting information for this article can be found under:
https://doi.org/10.1002/chem.201705218.
Chem. Eur. J. 2017, 23, 1 – 6
1
ꢃ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
Firstly, we studied the reaction of the silyl-dienol ether 1a
with tosylimine 2a, and different thiourea and squaramide bi-
functional catalysts 4a–d (Table 1). All the catalysts 4a–d
showed full conversion in the presence of 0.4 equivalents of
water. However, Takemoto’s catalysts 4c showed the best
enantioselectivity (entry 3). Then, different solvents were stud-
ied (entries 5–8), decreasing the reactivity and the enantiose-
lectivity compared to dichloromethane using catalyst 4c.
Table 2. Different imines for the aza-BHR.[a]
Table 1. Optimization of reaction conditions and catalyst for the aza-
BHR.[a]
Entry
Imine Water [equiv] ee [%][b] Time [h] Conversion [%][c]
1
2
3
4
5
6
7
8
2a
2b
2c
2d
2e
2 f
2g
2g
2g
2g
2g
2g
0.4
0.4
0.4
0.4
0.4
0.4
0.4
1.5
3
50
53
n.d.
12
48
48
48
48
48
48
48
48
24 (48)
24 (48)
24
48
48
100
100
22
61
46
–
n.r.
67
>99
>99
>99
>99
>99
n.d.
22 (51)[d]
58 (100)[d]
100
100
7
9
10
11[e]
12
4.5
4.5
–
Entry
Catalyst (20 mol%)
Solvent
ee [%][b]
Conversion [%][c]
[a] All the reactions were performed in 0.1 mmol scale in 1.0 mL solvent.
[b] Determined by SFC chromatography. [c] Determined by H NMR analy-
sis of the crude mixture after 24 or 48 hours. [d] Conversion after 48 h in
brackets. [e] 10 mol% of catalyst 4c.
1
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4c
4c
4c
4c
DCM
DCM
DCM
DCM
HFB
DCE
MeCN
THF
20
46
50
28
42
44
30
49
100
100
100
100
46
70
55
86
fluence of different quantities of water (entries 8–10). We de-
termined that 4.5 equivalents of water were needed to obtain
a full conversion after 24 hours. When a water free reaction
was carried out only 7% conversion was obtained (entry 12)
and the use of 1.5 and 3 equivalents did not afford full conver-
sions after 24 h (entries 8 and 9). The amount of catalyst was
reduced to 10 mol% with the same result but with a slightly
longer reaction time of 48 h (entry 11).
[a] All the reactions were performed in 0.1 mmol scale in 1.0 mL solvent.
[b] Determined by SFC chromatography. [c] Determined by H NMR analy-
sis of the crude mixture after 48 hours. HFB=hexafluorobenzene, DCE=
1,2-dichloroethane.
1
In view of these moderate results, we investigated different
imines 2a–g (Table 2). When pyridylsulfonylimine 2b[8] was
Under these optimized conditions the scope of the reaction
using different imines 2 (Table 3) and silyl reagents 1
used,
a
moderate enantioselectivity was found (entry 2),
(Schemes 2 and 3) was carried out. Different electron-donating
groups (p-Me, p-MeO 2h, 2i) or electron-withdrawing groups
(p-CF3, 2j) worked with excellent enantioselectivies (all exam-
ples; >99%ee). Heteroaromatic groups such as thienyl or furyl
moieties also afforded aza-BHR products with good to excel-
lent enantioselectivities (95–99%ee, 2k,l). Bromo (2m), and
the bulkier substituents such as 2n and 2o were also tolerated
under these conditions, providing 3m, 3n and 3o with good
ee values. The reaction was scale up (1.2 mmol) with the
bromo derivative 3m with a similar yield and enantioselectivity
(result between brackets). A double bond such as imine 2p or
an alkyl imine such as 2q also led to the corresponding ad-
ducts (3p and 3q) under these catalytic conditions, which are
difficult to obtain with the standard aza-BHR, with a slightly
lower Z/E selectivity (dr=94:6 and 93:7). In addition, the more
challenging ketimine 2r was also studied and afforded the
whereas the other bulkier imines such as 2c and 2d gave a
lower conversion and worse enantioselectivities (entries 3 and
4). The use of a sulfonyl group with a electron-donanting
group such as 2e did not improve the results (entry 5) and
Boc-imine 2 f did not afford the desired product (entry 6). All
these imines (2a–2e) have a moderate capability of forming a
hydrogen bond with Takemoto’s catalyst 4c because the sulfo-
nyl group weakens the availability of the lone pair on the ni-
trogen (C=N:). Therefore, we hypothesized that a more hydro-
gen coordinated imine such as 2g[9] would increase the enan-
tiomeric excess. Pleasantly, the reaction of imine 2g led to the
final product 3g with the highest enantiomeric excess
(>99%ee, entry 7) with a moderate conversion of 67% due to
the lower reactivity of this imine compared with the sulfonyl-
imines. In order to increase the conversion, we studied the in-
&
&
Chem. Eur. J. 2017, 23, 1 – 6
www.chemeurj.org
2
ꢃ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
Table 3. Scope of the aza-BHR with different imines 2g–r.[a]
Scheme 3. Different substitution at the silyl-dienol ethers for the synthesis
of tri and tetrasubstituted aza-BHR type products.
In addition, the substitution at the 4 and 5-positions of the
silyl-enolether (1e and 1 f) led to the b,b-disubstituted 3v and
b-monosubstituted 3w adducts, respectively in good yields
and enantioselectivies (Scheme 3, equations a and b). It is re-
markable that it is not possible to obtain these enantio-en-
riched adducts using other methods described in the litera-
ture.
Finally, we carried out different transformations of adducts
obtained to synthesize privileged compounds. Thus, the ben-
zothiazole group could be easily removed after reduction of
the aldehyde 3g and subsequent hydrolysis in basic condi-
tions, giving the 1,3-aminoalcohol 5g (equation a, Scheme 4),
[a] Conditions:
1 (0.3 mmol), 2 (0.1 mmol), 4c (10 mol%) and H2O
(4.5 equiv) in DCM (0.3 mL) at RT after 48 h. [b] Reaction carried out at
1.2 mmol scale.
Scheme 4. Derivatization of tri- and tetra-substituted aza-BHR type products
(BT=benzothiazole).
which can be used as precursor for a large number of pharma-
ceutical products.[2] Moreover, our methodology also provided
the possibility of synthetize privileged structures such as cata-
lysts 7v and 7m that have been previously used as Lewis
super-bases (equation b, Scheme 4).[10] The procedure started
with the reduction of 3v and 3m to give 6v and 6m, which
after cyclization afforded the catalysts 7v and 7m. The abso-
lute configuration of Baylis–Hillman products 3 were assigned
by correlation with known compounds in the literature (6v
and 7v)[10b] and were determined as 2S and 3R, whereas the
configuration of the double bond was determined as E by
n.O.e. NMR experiments (see Supporting Information).
Scheme 2. Synthesis of ester, ketone and amide aza-BH type products.
optically enriched quaternary center product 3r in a good
yield and enantioselectivity.
The reaction also tolerated different groups at the a-position
to the TMSO group at the silyl-dienol ether 1b–d (Scheme 2).
Therefore, the phenyl group led to the ketone 3s with an ex-
cellent enantioselectivity (equation a), whereas the use of eno-
late 1c afforded the amide 3t in an excellent yield and enan-
tioselectivity (equation b) which cannot be activated under the
standard aza-Baylis–Hillman reaction conditions due to the low
electrophilic character of the double bond. In a similar manner,
the silyl reagent 1d reacted with the imine 2g to give the
ester 3u with excellent ee and yield (equation c).
Based on our previous calculations[7] and the additional ex-
perimental data obtained in this work (see below), a proposed
mechanism is outlined in Scheme 5. A water molecule can
easily attack the Si center (I) and further evolution of this
Chem. Eur. J. 2017, 23, 1 – 6
www.chemeurj.org
3
ꢃ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
stituted by one with a stronger electronegativity such as
oxygen (benzo[d]oxazole 3x), the resultant enantioselectivity is
also lower (ee=87%). These last two results indicate that the
key coordination hydrogen bond is the lone pair of the iminic
nitrogen (C=N:), indicating that the more plausible mechanism
is Pꢂpai’s model, in which the dienolate is strongly stabilized
by the thiourea moiety[13] and the imine is coordinated to the
ammonium ion group (see IV). In addition, the different size of
the tosyl and benzothiazole group could also have an impor-
tant role. Then, after the addition (CÀC bond formation) and
isomerization of the double bond gives the final aza-Baylis–
Hillman products with high enantioselectivity.
In conclusion, a new organocatalytic strategy for the synthe-
sis of enantioenriched aza-Baylis–Hillman type products via a
frustrated vinylogous reaction is presented. This process pro-
ceeds under mild conditions with good yields and excellent
enantioselectivities. The reaction tolerates a large numer of dif-
ferent imines, and the synthesis of tri- and tetra-substituted
aza-Baylis–Hillman type products. Moreover, easy derivatiza-
tions of the final products led to important building blocks in
organic synthesis such as 1,3-aminoalcohols and Lewis super-
base catalysts. A mechanism for this reaction based on the ex-
perimental data has been proposed.
Acknowledgements
Scheme 5. Different substitution at the benzothiazole core to test the influ-
ence on the enantioselectivity of the reaction (BT=benzothiazole).
M. F. would like to express thanks to the UAM for a predoctoral
fellowship-UAM. Financial support from the Spanish Govern-
ment (CTQ2015-64561-R) is also gratefully acknowledged.
system implies firstly a proton transfer from the water mole-
cule to the amine nitrogen, followed by a nucleophilic attack
of the resulting hydroxide to the silicon atom. After the hydrol-
ysis step, two intermediates before the coordination to the
imine can be postulated (II and III). The hydrogen bond forma-
tion with the imine 2 leads to the intermediates IV or V. Inter-
mediate V is based on Takemoto’s model,[11] in which the coor-
dination with the electrophile, in this case the imine 2a, is
taking place through the thiourea moiety. By contrast, the in-
termediate IV is based on the well-known Pꢂpai’s model.[12]
Such pre-organization is characterized by the coordination of
the electrophile through the ammonium salt, whereas the nu-
cleophile can be strongly stabilized by the thiourea group. In
order to differentiate between these two models (Takemoto’s
and Papai)[11,12] different imines 2 with different aromatic resi-
dues at the benzothiazole group were synthesized (bottom,
Scheme 5).
Conflict of interest
The authors declare no conflict of interest.
Keywords: asymmetric synthesis · Baylis–Hillman reaction ·
bifunctional catalysis · silyl-enol ethers · thiourea
[1] For specific review on the asymmetric aza-Morita Baylis–Hillman, see:
a) F.-L. Hu, M. Shi, Org. Chem. Front. 2014, 1, 587; b) H. Pellissier, Tetrahe-
dron 2017, 73, 2831.
[2] For general reviews on the aza-Morita Baylis Hillman, see: a) D. Basa-
vaiah, T. Satyanarayana, A. J. Rao, Chem. Rev. 2003, 103, 811; b) Y. Wei,
M. Shi, Chem. Rev. 2013, 113, 6659; c) Y. L. Shi, M. Shi, Eur. J. Org. Chem.
2007, 2905; d) D. Basavaiah, B. S. Reddy, S. S. Badsara, Chem. Rev. 2010,
110, 5447; for a related work, see: e) N. Utsumi, H. Zhang, F. Tanaka, C. F.
Barbas III, Angew. Chem. Int. Ed. 2007, 46, 1878; Angew. Chem. 2007,
119, 1910.
[3] For a pioneering work, see: a) M. Shi, Y.-M. Xu, Angew. Chem. Int. Ed.
2002, 41, 4507; Angew. Chem. 2002, 114, 4689; For other works of the
group, see: b) M. Shi, L.-H. Chen, C.-Q. Li, J. Am. Chem. Soc. 2005, 127,
3790; c) M. Shi, Y.-M. Xu, Y.-L. Shi, Chem. Eur. J. 2005, 11, 1794; d) M. Shi,
L.-H. Chen, W.-D. Teng, Adv. Synth. Catal. 2005, 347, 1781.
[4] a) N. Abermil, G. Masson, J. Zhu, J. Am. Chem. Soc. 2008, 130, 12596;
b) N. Abermil, G. Masson, J. Zhu, Org. Lett. 2009, 11, 4648; c) N. Abermil,
G. Masson, J. Zhu, Adv. Synth. Catal. 2010, 352, 656; d) S. Kawahara, A.
Nakano, T. Esumi, Y. Iwabuchi, S. Hatakeyama, Org. Lett. 2003, 5, 3103;
e) K. Matsui, S. Takizawa, H. Sasai, J. Am. Chem. Soc. 2005, 127, 3680;
f) S. Takizawa, N. Inoue, S. Hirata, H. Sasai, Angew. Chem. Int. Ed. 2010,
49, 9725; Angew. Chem. 2010, 122, 9919; g) I. T. Raheem, E. N. Jacobsen,
Adv. Synth. Catal. 2005, 347, 1701; h) M. S. Taylor, E. N. Jacobsen, Angew.
The use of different benzothiazole imines with EWG and
EDGs were not found to have any influence of the enantiose-
lectivity, and in all the cases high enantioselectivities, >99%
for products 3y, 3z and 3aa were obtained. Therefore, the co-
ordination of the benzothiazolinic nitrogen to the catalyst is
implausible since the influence of these EWG and EDGs has a
scarce influence in this para-positon. Conversely, the use of
imines with a strong electron-withdrawing character such as
the tosyl group, provoke a dramatic decrease in the enantiose-
lecitivy (3a, ee=50%). In addition, when a sulfur atom is sub-
&
&
Chem. Eur. J. 2017, 23, 1 – 6
www.chemeurj.org
4
ꢃ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
Chem. Int. Ed. 2006, 45, 1520; Angew. Chem. 2006, 118, 1550; i) K.
Hyodo, S. Nakamura, N. Shibata, Angew. Chem. Int. Ed. 2012, 51, 10337;
Angew. Chem. 2012, 124, 10483; j) T. Yukawa, B. Seelig, Y. Xu, H. Morimo-
to, S. Matsunaga, A. Berkessel, M. Shibasaki, J. Am. Chem. Soc. 2010,
132, 11988; k) Y. L. Shi, M. Shi, Adv. Synth. Catal. 2007, 349, 2129; l) R.
Gausepohl, P. Buskens, J. Kleinen, A. Bruckmann, C. W. Lehmann, J. Klan-
kermayer, W. Leitner, Angew. Chem. Int. Ed. 2006, 45, 3555; Angew.
[7] M. Frias, R. Mas-Ballestꢁ, S. Arias, C. Alvarado, J. Alemꢂn, J. Am. Chem.
Soc. 2017, 139, 672.
[8] See for example: a) S. Nakamura, H. Nakashima, H. Sugimoto, H. Sano,
M. Hattori, N. Shibata, T. Toru, Chem. Eur. J. 2008, 14, 2145; b) J. Esqui-
vias, R. Gꢄmez Arrayꢂs, J. C. Carretero, J. Org. Chem. 2005, 70, 7451.
[9] H.-X. He, W. Yang, D.-M. Dua, Adv. Synth. Catal. 2013, 355, 1137.
[10] a) L. C. Morrill, T. Lebl, A. M. Z. Slawina, A. D. Smith, Chem. Sci. 2012, 3,
2088; b) L. C. Morrill, J. Douglas, T. Lebl, A. M. Z. Slawin, D. J. Fox, A. D.
Smith, Chem. Sci. 2013, 4, 4146.
ˇ
´
Chem. 2006, 118, 3635; m) S. Cꢀhalovꢂ, P. Dziedzic, A. Cꢄrdova, J. Vesely,
Adv. Synth. Catal. 2011, 353, 1906.
[5] a) A. Salvador Gonzꢂlez, R. Gꢄmez-Arrayꢂs, M. Rodrꢀguez Rivero, J. C.
Carretero, Org. Lett. 2008, 10, 4335; b) Q. Wang, M. van Gemmeren, B.
List, Angew. Chem. Int. Ed. 2014, 53, 13592; Angew. Chem. 2014, 126,
13810; for a Mannich reaction catalyzed by bifunctional thiourea cataly-
sis, see: c) D. Zhao, D. Yang, Y. Wang, Y. Wang, L. Wang, L. Maoa, R.
Wang, Chem. Sci. 2011, 2, 1918.
[6] For review of catalytic vinylogous aldol reactions, see: a) S. Denmark,
E. J. R. Heemstra, G. L. Beutner, Angew. Chem. Int. Ed. 2005, 44, 4682;
Angew. Chem. 2005, 117, 4760; b) S. V. Pansare, E. K. Paul, Chem. Eur. J.
2011, 17, 8770; c) V. Bisai, Synthesis 2012, 44, 1453; For Mannich reac-
tions, see: d) C. Schneider, M. Sickert in Chiral amine synthesis: Methods,
developments and applications (Eds.: T. C. Hugent), Wiley, Weinheim,
2010, pp. 157–177.
[11] T. Okino, Y. Hoashi, Y. Takemoto, J. Am. Chem. Soc. 2003, 125, 12672.
[12] A. Hamza, G. Schubert, T. Sꢄos, I. Pꢂpai, J. Am. Chem. Soc. 2006, 128,
13151.
[13] In our previous studies (Ref. [7]), we calculated both models (Takemoto
and Pꢂpai’s mechanisms) in the reaction of silyl-dienol ethers with ni-
troalkenes. However, according to the obtained energy values, we de-
scribed that both models were plausible and only a difference of nearly
5 kcalmolÀ1 between the energy of these two models were found.
Manuscript received: November 3, 2017
Accepted manuscript online: November 28, 2017
Version of record online: && &&, 0000
Chem. Eur. J. 2017, 23, 1 – 6
www.chemeurj.org
5
ꢃ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
COMMUNICATION
&
Bifunctional Catalysis
M. Frꢀas, A. C. Carrasco, A. Fraile,
J. Alemꢁn*
&& – &&
A General Asymmetric Formal
Synthesis of Aza-Baylis–Hillman Type
Products under Bifunctional Catalysis
With frustration to success: A new or-
ganocatalytic strategy for the synthesis
of enantioenriched aza-Baylis–Hillman
type products via a frustrated vinylo-
gous reaction is presented.
&
&
Chem. Eur. J. 2017, 23, 1 – 6
www.chemeurj.org
6
ꢃ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!