Chemoselective catalytic conjugate addition of alcohols over amines

Article abstract of DOI:10.1002/anie.201309755

A highly chemoselective conjugate addition of alcohols in the presence of amines is described. The cooperative nature of the catalyst enabled chemoselective activation of alcohols over amines, allowing the conjugate addition to soft Lewis basic α,β-unsaturated nitriles. Divergent transformation of the nitrile functionality highlights the utility of the present catalysis. The cooperative nature of a copper catalyst enabled the highly chemoselective activation of alcohols in the presence of amines and thus the conjugate addition of the hydroxy group to soft Lewis basic α,β-unsaturated nitriles. The presented method proceeds under proton-transfer conditions, reverses the innate reactivity of the OH and NH groups, and does not require protecting groups. dppe=1,2-bis(diphenylphosphino) ethane, MeSal=3-methylsalicylate. Copyright

Full text of DOI:10.1002/anie.201309755

Angewandte
Chemie
DOI: 10.1002/anie.201309755
Synthetic Methods
Chemoselective Catalytic Conjugate Addition of Alcohols over
Amines**
Shuhei Uesugi, Zhao Li, Ryo Yazaki,* and Takashi Ohshima*
Abstract: A highly chemoselective conjugate addition of
alcohols in the presence of amines is described. The cooper-
ative nature of the catalyst enabled chemoselective activation of
alcohols over amines, allowing the conjugate addition to soft
Lewis basic a,b-unsaturated nitriles. Divergent transformation
of the nitrile functionality highlights the utility of the present
catalysis.
Lewis acid catalyst would activate electrophiles to facilitate
subsequent coupling of innately more nucleophilic amines,[7]
a cooperative catalyst comprising soft Lewis acids and hard
Brønsted bases would enable the simultaneous activation of
both soft Lewis basic electrophiles and hard hydroxy groups
through selective deprotonation of the hydroxy group over
the amino group as a result of the difference in acidity,
allowing chemoselective coupling (Scheme 1).[8] Thus, we
selected commercially available acrylonitrile (1a) as a repre-
sentative soft electrophile.[9] Divergent transformation of
a nitrile, which is a masked carboxylic acid functionality, is
beneficial for further elaboration of the product.[10]
C
atalyst-controlled chemoselective reactions offer new
opportunities for minimal reliance on protecting groups,[1]
even in the presence of innately more reactive functional-
ities.[2] Despite the prospects for contributions to both atom
and step economy[3] of catalyst-controlled chemoselective
reactions, progress in this area, especially the reversal of the
innate reactivity of amines and alcohols, has been limited
relative to catalyst-controlled stereo- or regioselective reac-
tions. Some examples of catalyst-controlled chemoselective
reactions of hydroxy groups in the presence of an amino
group were recently reported,[4,5] including our O-selective
acylation.[4e,j] The reaction patterns were, however, highly
limited and remain unexplored. Moreover, the reported
reactions include the inevitable formation of stoichiometric
amounts of unneeded co-products, such as inorganic salts and
alcohols, thus reducing the reaction efficiency. Herein, we
report the first example of a catalyst-controlled chemo-
selective conjugate addition of a hydroxy group in the
presence of an amino group. Conjugate addition of the
hydroxy group was performed under proton-transfer condi-
tions and is applicable to natural product synthesis.[6]
The initial study examined the use of a 1:1 mixture of
alcohol 2a and amine 3a (Table 1).[11] Without the catalyst,
the N adduct 5aa was obtained exclusively in 17% yield at
room temperature in 3 h (Table 1, entry 1). The use of
4 mol% of strong base (nBuLi or LiHMDS) to generate the
more nucleophilic metal alkoxide afforded O adduct 4aa,
Scheme 1. Strategy for reversing innate reactivity.
Table 1: Optimization of reaction conditions.[a]
We envisioned that soft Lewis acidic transition metals
would activate soft Lewis basic electrophiles, even in the
presence of hard hydroxy and amino groups. While a simple
[*] S. Uesugi, Z. Li, Dr. R. Yazaki, Prof. Dr. T. Ohshima
Graduate school of Pharmaceutical Sciences
Kyushu University and CREST, JST
Entry
Cat.
Base
Ratio
Yield of
O/N[b]
4aa [%][b]
Maidashi Higashi-ku, Fukuoka, 812-8582 (Japan)
E-mail: yazaki@phar.kyushu-u.ac.jp
1[c]
2
3
4
5
6
7
8
–
–
–
–
<1/99
1.6/1
1.5/1
>20/1
>20/1
>20/1
0.6/1
(17)
6
9
nBuLi
LiHMDS
nBuLi
nBuLi
LiHMDS
–
ohshima@phar.kyushu-u.ac.jp
[**] This work was financially supported by the Grant-in-Aid for Scientific
Research (B), Grant-in-Aid for Research Activity (Start-up), Scientific
Research on Innovative Areas and Platform for Drug Discovery,
Informatics, and Structural Life Science from MEXT, CREST from
JST, Uehara Memorial Foundation, Takeda Science Foundation, and
Kyushu University Interdisciplinary Programs in Education and
Projects in Research Development. Z.L. thanks Otsuka Toshimi
Scholarship Foundation. We are grateful to Dr. Tomofumi Miyamoto
at Kyushu University for HRMS analysis and the use of a polar-
imeter.
CuOAc, dppe
Cu(MeSal), dppe
Cu(MeSal), dppe
Cu(MeSal), dppe
mesitylcopper, dppe
83
84
88
3
–
>20/1
83[d]
[a] Conditions: 1a (1.2 mmol), 2a (1.44 mmol), 3a (1.44 mmol), THF
(1.2 mL). [b] Determined by GC analysis using nonadecane as an internal
standard. Yield of 5aa is shown in parenthesis. [c] At room temperature
for 3 h. [d] Determined by 1H NMR analysis using durene as an internal
standard. dppe=1,2-bis(diphenylphosphino)ethane,
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201309755.
HMDS=1,1,1,3,3,3-hexamethyldisilazane, MeSal=3-methylsalicylate.
Angew. Chem. Int. Ed. 2014, 53, 1611 –1615
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1611

.
Angewandte
Communications
albeit in low yield and low selectivity (Table 1, entries 2 and
3).[12] We next examined the combined use of a transition
metal and Brønsted base. Screening of various metals showed
that poorly soluble copper(I) acetate afforded 4aa in 83%
yield (Table 1, entry 4).[13,14] Commercially available cop-
per(I) 3-methylsalicylate, which showed higher solubility,
delivered 4aa in high yield with high chemoselectivity
(Table 1, entry 5). The less basic LiHMDS also promoted
the conjugate addition of alcohol with the same efficiency
(Table 1, entry 6). Omission of a Brønsted base gave poor
results, suggesting that the combined use of copper(I)
complex with Brønsted base was essential for obtaining
both high yield and high chemoselectivity (Table 1, entry 7).
Using mesitylcopper with a dppe ligand resulted in efficient
catalytic performance, indicating that copper(I) alkoxide is
the actual catalytic species (Table 1, entry 8).[15,16]
With the optimized conditions in hand, we explored the
substrate scope of the catalytic chemoselective conjugate
addition of alcohols (Table 2). Reactions of mixtures of
structurally similar amines 3a–3d and alcohols 2a–2d,
respectively (R3 = R4), gave O adducts 4aa–4ad in high
yield. Furthermore, the N-Boc-protected substrate 2e
afforded product 4ae in high yield. Products 4bb and 4db
were obtained in acceptable yields, when crotononitrile and
methacrylonitrile were employed. The pharmaceutically
active compounds 2 f and 2g, cholesterol (2h), and cincho-
nidine (2i) were successfully converted into conjugate adducts
4af–4ai.
Table 2: Chemoselective conjugate addition of alcohols (2) over amines (3).[a]
Entry
1
2
3
4 (yield [%])
1
2
1a
1a
2a
2b
3a
3b
4aa (86)
4ab (87)
3
1a
2c
3c
4ac (86)
4
1a
1a
1a
2d
2c
2e
3d
3e
3b
4ad (76)
4ac (88)
4ae (83)
5
6[b]
7[b,c,d]
8[b,d,e]
9[b,d,e]
1b
1c
1d
2b
2b
2b
3b
3b
3b
4bb (63)
4cb (78)
4db (61)
10[b,f]
1a
1a
2 f
3b
3b
4af (80)
4ag (84)
11[b,f]
2g
12[b]
1a
1a
2h
2i
3b
3b
4ah (54)
4ai (79)
13[b,c,f]
[a] Conditions: 1 (1.0 equiv), 2 (1.2 equiv), 3 (1.2 equiv), THF (1.0m). Yields of isolated products are given. In all cases, the O/N ratio (measured by
1H NMR analysis of the crude reaction mixture) was >20/1. [b] nBuLi was used instead of LiHMDS. [c] 2 equivalents of 1 were used. [d] Reaction was
performed at room temperature. [e] 5 equivalents of 1 were used. [f] DMF was used as solvent.[10] “O”=O adduct, “N”=N adduct.
1612
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1611 –1615

Angewandte
Chemie
We next performed the reaction using amino alcohols 6
(Table 3). A survey showed that DMF was an optimal solvent
for amino alcohols, and O adducts 7 were obtained with high
selectivity. The length of the alkyl chain did not affect the
chemoselectivity (7aa, 7ab). In preceding chemoselective
catalytic reactions, b-amino alcohols could not be used as
substrates.[4] In contrast, exclusive O-selective addition of l-
phenylalaninol (6d) was observed with 0.5 mol% catalyst. An
aniline NH group did not affect either the yield or chemo-
selectivity (7af, 7ag).[17] In the presence of a primary amine in
a position, secondary alcohols selectively reacted with 1a to
afford the O adduct (7ah). The process could be applied to
amino alcohols with a primary alcohol and a highly nucleo-
philic secondary amine, although the chemoselectivity
decreased slightly (7ai). The pharmaceutically active com-
pounds 6j and 6k) proved to be good substrates. High
chemoselectivities were observed using compounds 6l and
6m, but the yields were low as a result of poor solubility.
Chemoselective conjugate addition was also possible with
substrates containing amide functionalities with acidic NH
protons without loss of the stereoinformation at the a posi-
tion of the amide, based on 1H NMR analysis (7ao, 7ap). The
tolerance of esters was demonstrated in a highly chemo-
selective conversion of H-Ser-Phe-OtBu (7aq).
possible using Wilkinsonꢀs catalyst and acetoaldoxime afford-
ing primary amide 13.[19] Employment of sodium azide
produced tetrazole 14.[20]
Table 3: Substrate scope of amino alcohols (6).[a]
7aa[b,c]
7ab
7ac[c]
O/N= >20/1
68% yield
O/N= >20/1
79% yield
O/N= >20/1
88% yield
7ad
7ad
7ae
O/N= >20/1
89% yield
(0.5 mol% cat.)
O/N= >20/1
86% yield
(gram scale)
O/N= >20/1
75% yield
7af[d]
7ag[d]
7ah
The usefulness of the present chemoselective catalysis was
demonstrated by further elaboration of the nitrile function-
ality (Scheme 2). Product 7ad, derived from a chiral amino
alcohol, was transformed into ethyl ester 9 under Pinner
reaction conditions. This ester was converted into amide 10 by
coupling with Z-Val-OH with no stereochemical erosion,
O/N= >20/1
61% yield
O/N= >20/1
74% yield
O/N/N+O=15.8/1.0/
1.3
61% yield
1
based on H NMR analysis, and seven-membered lactam 11
through intramolecular amidation. The nitrile functionality of
7ad could be readily reduced with NiCl2/NaBH4, providing
diamine 12.[18] The hydration of the nitrile functionality was
7ai
O/N/
7aj[d]
O/N=20/1
7ak[d]
O/N= >20/1
56% yield
O+N=13.7/1.0/ 63% yield
1.2
82% yield
7al[d,e,f]
7am[d]
7an
O/N=20/1
34% yield
O/N+N/
O+N+N=8.2/1.2/1.0 83% yield
O/N=20/1
35% yield
7ao
7ap
7aq
O/N=20/1
70% yield
O/N=20/1
72% yield
O/N=20/1
58% yield
Scheme 2. Transformation of the product. Reagents and conditions:
a) TMSCl, EtOH, 508C, 13 h, 72%; b) Z-Val-OH, EDC, HOBt, NMM,
DMF, RT, 24 h, 92%, d.r.= >20/1; c) 1. 1n NaOH, MeOH, RT, 2 h;
2. EDC, HOBt, NMM, DMF, RT, 12 h, 65% over two steps; d) CbzCl,
Et3N, CH2Cl2, RT, 1 h, 87%; e) NiCl2 (cat.), NaBH4, MeOH, 08C!RT,
Boc2O, 73%; f) Boc2O, Et3N, CH2Cl2, RT, 0.5 h, 86%; g) [RhCl(PPh3)3]
[a] Conditions: 1a (1.0 equiv), 6 (1.2 equiv), DMF (1.0m). Yields of
isolated products are given. O/N ratios were measured by H NMR
1
analysis of the crude reaction mixture. [b] Isolated as the N-Boc-
protected product. Yield over two steps is given. [c] LiHMDS was used
instead of nBuLi. [d] Reaction was performed at room temperature. [e] A
mixture of DMF/DMSO=1/1 was used as a solvent. [f] 2 equivalents of
1a were used. DMF=N,N-dimethylformamide, “N+N”=N,N-dialky-
lated product, “O+N”=O,N-dialkylated product,
=
(cat.), CH3CH NOH, toluene, reflux, 1.5 h, 97%; h) NaN3, Et3NHCl,
DMF, 1208C, 24 h, 86%. Boc=tert-butyloxycarbonyl, HOBt=1-hydrox-
ybenzotriazole, EDC=1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride, NMM=N-methylmorpholine, TMS=trimethylsilyl.
“O+N+N”=O,N,N-trialkylated product.[11]
Angew. Chem. Int. Ed. 2014, 53, 1611 –1615
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1613

.
Angewandte
Communications
The present catalysis turned out to be highly chemo-
selective, not only for nucleophiles, but also for electrophiles.
In the presence of relatively hard electrophile tert-butyl
acrylate (15) as an additional electrophile, selective conjugate
addition of alcohol to soft acrylonitrile was achieved over the
three other possible reaction pathways, while a mixture of
N adducts derived from both 1a and 15 were obtained
without catalyst (Scheme 3).[11] In a similar way, amino
alcohol 6d selectively reacted with 1a over 15 to give 7ad.
innately more nucleophilic amines. It is particularly note-
worthy that a variety of b-amino alcohols, including pharma-
ceutically active compounds and peptides, are applicable to
the present catalysis. Further studies to elucidate the precise
reaction mechanism and the application of this unique
catalytic chemoselective activation of functionalities with
innately low reactivity in other reactions are in progress.
Received: November 9, 2013
Published online: January 22, 2014
Keywords: amino alcohols · chemoselective catalysis ·
.
cooperative effects · proton transfer · synthetic methods
[1] a) R. W. Hoffmann, Synthesis 2006, 3531; b) P. S. Baran, T. J.
Maimone, J. M. Richter, Nature 2007, 446, 404; c) I. S. Young,
P. S. Baran, Nat. Chem. 2009, 1, 193.
[2] For accounts on chemoselective reactions: a) B. M. Trost,
Science 1983, 219, 245; b) R. A. Shenvi, D. P. OꢀMalley, P. S.
Baran, Acc. Chem. Res. 2009, 42, 530; c) N. A. Afagh, A. K.
Yudin, Angew. Chem. 2010, 122, 270; Angew. Chem. Int. Ed.
2010, 49, 262; d) J. Mahatthananchai, A. M. Dumas, J. W. Bode,
Angew. Chem. 2012, 124, 11114; Angew. Chem. Int. Ed. 2012, 51,
10954.
Scheme 3. Competitive conjugate addition of amino alcohol to acryl-
[3] a) B. M. Trost, Science 1991, 254, 1471; b) P. A. Wender, V. A.
Verma, T. J. Paxton, T. H. Pillow, Acc. Chem. Res. 2008, 41, 40.
[4] Arylation: a) G. E. Job, S. L. Buchwald, Org. Lett. 2002, 4, 3703;
b) A. Shafir, P. A. Lichtor, S. L. Buchwald, J. Am. Chem. Soc.
2007, 129, 3490; c) D. Maiti, S. L. Buchwald, J. Am. Chem. Soc.
2009, 131, 17423; Transesterification: d) M.-H. Lin, T. V. Rajan-
Babu, Org. Lett. 2000, 2, 997; e) T. Ohshima, T. Iwasaki, Y.
Maegawa, A. Yoshiyama, K. Mashima, J. Am. Chem. Soc. 2008,
130, 2944; f) S. De Sarkar, S. Grimme, A. Studer, J. Am. Chem.
Soc. 2010, 132, 1190; g) M. Hatano, Y. Furuya, T. Shimmura, K.
Moriyama, S. Kamiya, T. Maki, K. Ishihara, Org. Lett. 2011, 13,
426; h) M. Hatano, K. Ishihara, Chem. Commun. 2013, 49, 1983;
i) R. C. Samanta, S. De Sarkar, R. Frçhlich, S. Grimme, A.
Studer, Chem. Sci. 2013, 4, 2177; j) Y. Hayashi, S. Santoro, Y.
Azuma, F. Himo, T. Ohshima, K. Mashima, J. Am. Chem. Soc.
2013, 135, 6192.
[5] Recently reported catalyst-controlled C-O versus C-N allylic
functionalization: I. I. Strambeanu, M. C. White, J. Am. Chem.
Soc. 2013, 135, 12032.
[6] For reviews on conjugate addition of alcohols: a) C. F. Nising, S.
Brꢁse, Chem. Soc. Rev. 2008, 37, 1218; b) C. F. Nising, S. Brꢁse,
Chem. Soc. Rev. 2012, 41, 988.
[7] For reviews on conjugate addition of amines: a) M. Liu, M. P.
Sibi, Tetrahedron 2002, 58, 7991; b) T. C. Tobias, J.-Q. Yu, J. B.
Spencer, Chem. Eur. J. 2004, 10, 484, and references therein.
[8] For reviews on Lewis acid/Brønsted base cooperative catalysis:
a) M. Shibasaki, N. Yoshikawa, Chem. Rev. 2002, 102, 2187; b) N.
Kumagai, M. Shibasaki, Angew. Chem. 2011, 123, 4856; Angew.
Chem. Int. Ed. 2011, 50, 4760.
onitrile.
A series of control experiments were conducted to gain
preliminary mechanistic insight.[11] Although O adducts 4
underwent retro-reaction, N adducts 5 were almost com-
pletely recovered under the present catalysis, suggesting that
the N adduct did not undergo retro-reaction and the kineti-
cally favorable O adduct was obtained under optimal con-
ditions. The plausible catalytic cycle is depicted in Scheme 4.
The combined use of copper(I) complex with Brønsted base
generated the actual catalytic species copper(I) alkoxide
(Table 1, entry 8).[11] The nucleophilic addition of copper
alkoxide to acrylonitrile (1a) then proceeded to afford the C-
or N-bound copper intermediate.[21–23] Subsequent protona-
tion occurred with the OH proton to give amino ether 7 with
the concomitant generation of copper(I) alkoxide.
In conclusion, we developed a highly chemoselective
catalytic conjugate addition of alcohols in the presence of
[9] F. F. Fleming, Q. Wang, Chem. Rev. 2003, 103, 2035.
[10] Utility of nitrile functionality in pharmaceuticals: F. F. Fleming,
L. Yao, P. C. Ravikumer, L. Funk, B. C. Shook, J. Med. Chem.
2010, 53, 7902.
[11] See the Supporting Information for details.
[12] Attempted Brønsted base catalysis using NaHMDS or KHMDS
was complicated by competitive polymerization of acrylonitrile
although conjugate addition of alcohol was promoted.
[13] Copper catalyzed conjugate addition of alcohols in the absence
of amines: a) C. Munro-Leighton, E. D. Blue, T. B. Gunnoe, J.
Am. Chem. Soc. 2006, 128, 1446; b) C. Munro-Leighton, S. A.
Delp, E. D. Blue, T. B. Gunnoe, Organometallics 2007, 26, 1483;
Scheme 4. Plausible catalytic cycle.
1614
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1611 –1615

Angewandte
Chemie
c) F. Wang, H. Yang, H, Fu, Z. Pei, Chem. Commun. 2013, 49,
517.
[20] For the utility of tetrazole as a carboxylic acid bioisostere, see:
N. A. Meanwell, J. Med. Chem. 2011, 54, 2529.
[14] Structure of copper(I)/dppe complex, see: N. Vijayashree, A. G.
Samuelson, M. Nethaji, Curr. Sci. 1993, 65, 57.
[21] C- and N-bound transition metal cyanocarbanion, see; T. Naota,
A. Tannna, S. Kamuro, M. Hieda, K. Ogata, S.-I. Murahashi, H.
Takaya, Chem. Eur. J. 2008, 14, 2482, and references therein.
[22] It is unclear whether simultaneous activation of acrylonitrile and
alcohol is operative because intramolecular conjugate addition
of copper(I) alkoxide coordinated by acrylonitrile in an end-on
fashion would be sterically unlikely. Coordination in an end-on
fashion for nitrile functionality, see: V. Y. Kukushkin, A. J. L.
Pombeiro, Chem. Rev. 2002, 102, 1771. For further discussion,
see Supporting Information.
[15] For a preparation of copper alkoxide from mesitylcopper and
alcohol, see: a) M. Hꢂkansson, C. Lopes, S. Jagner, Organo-
metallics 1998, 17, 210; b) M. Hꢂkansson, C. Lopes, S. Jagner,
Inorg. Chim. Acta 2000, 304, 178; c) M. Stollenz, F. Meyer,
Organometallics 2012, 31, 7708, and references therein.
[16] The possibility of nucleophilic addition of lithium alkoxide to
acrylonitrile activated by copper(I) salt cannot be ruled out.
[17] Aniline shows a comparable nucleophilicity to primary alkyl
amines, see: F. Brotzel, Y. C. Chu, H. Mayr, J. Org. Chem. 2007,
72, 3679.
[23] A simultaneous activation of acrylonitrile and alcohol was
proposed, see; a) C. S. Yi, S. Y. Yun, Z. He, Organometallics
2003, 22, 3031; b) A. B. Salah, C. Offenstein, D. Zargarian,
Organometallics 2011, 30, 5352.
[18] S. Caddick, D. B. Judd, A. K. de K. Lewis, M. T. Reich, M. R. V.
Williams, Tetrahedron 2003, 59, 5417.
[19] J. Lee, M. Kim, S. Chang, H.-Y. Lee, Org. Lett. 2009, 11, 5598.
Angew. Chem. Int. Ed. 2014, 53, 1611 –1615
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1615