Direct catalytic asymmetric Mannich-type reactions of γ-butenolides: Effectiveness of Bronsted acid in chiral metal catalysis

Article abstract of DOI:10.1021/ol800756r

(Chemical Equation Presented) Direct catalytic asymmetric Mannich-type reactions of γ-butenolides are described. A chiral Lewis acid/amine base/Bronsted acid combination was used to catalyze a γ-addition of γ-butenolides to N-diphenylphosphinoyl imines, a

Full text of DOI:10.1021/ol800756r

ORGANIC
LETTERS
2008
Vol. 10, No. 11
2319-2322
Direct Catalytic Asymmetric
Mannich-Type Reactions of
γ-Butenolides: Effectiveness of
Brønsted Acid in Chiral Metal Catalysis
Akitake Yamaguchi, Shigeki Matsunaga,* and Masakatsu Shibasaki*
Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo, 7-3-1 Hongo,
Bunkyo-ku, Tokyo 113-0033, Japan
mshibasa@mol.f.u-tokyo.ac.jp; smatsuna@mol.f.u-tokyo.ac.jp
Received April 3, 2008
ABSTRACT
Direct catalytic asymmetric Mannich-type reactions of γ-butenolides are described. A chiral Lewis acid/amine base/Brønsted acid combination
was used to catalyze a γ-addition of γ-butenolides to N-diphenylphosphinoyl imines, affording the products in up to >99% yield, anti/syn )
>97:3, and 84% ee. The use of a catalytic amount of TfOH in addition to La(OTf)3/Me-PyBox/TMEDA was important for improving yield and
stereoselectivity.
Catalytic asymmetric Mannich(-type) reactions of aldehydes,
ketones, esters, and other donors directed toward the
synthesis of ꢀ-amino carbonyl compounds have been inten-
sively investigated over the past decade.1 In contrast,
vinylogous asymmetric Mannich-type reactions affording
useful chiral δ-amino carbonyl compounds are not as well
studied.2–4 γ-Butenolides and related compounds are impor-
tant donors for asymmetric vinylogous Mannich(-type)
reactions, because the γ-butenolide skeleton appears in many
natural products.5 Only a few asymmetric vinylogous Man-
nich-type reactions of preformed siloxyfurans, affording
Mannich-adducts in high enantioselectivity, have been
reported to date.4 A highly enantioselective aza-Friedel-Crafts
reaction with methoxyfuran as a nucleophile has also been
reported, but additional oxidation was required to convert
the aza-Friedel-Crafts product to γ-butenolide.6 There are
no reports on the direct use of γ-butenolides themselves7 in
(1) For reviews, see: (a) Kobayashi, S.; Ueno, M. In ComprehensiVe
Asymmetric Catalysis, Supplement I; Jacobsen, E. N., Pfalz, A., Yamamoto,
H., Eds.; Springer: Berlin, Germany, 2003;Chapter 29.5. (b) Friestad, G. K.;
Mathies, A. K. Tetrahedron 2007, 63, 2541.
(2) For reviews of vinylogous Mannich-type reactions, see: (a) Martin,
S. F. Acc. Chem. Res. 2002, 35, 895. (b) Bur, S. K.; Martin, S. F.
(5) Catalytic asymmetric synthesis of functionalized γ-butenolide skel-
eton: For a review, see: (a) Casiraghi, G.; Zanardi, F.; Appendino, G.; Rassu,
G. Chem. ReV. 2000, 100, 1929. For selected examples, see: (b) Evans,
D. A.; Kozlowski, M. C.; Murry, J. A.; Burgey, C. S.; Campos, K. R.;
Connell, B. T.; Staples, R. J. J. Am. Chem. Soc. 1999, 121, 669. (c) Szlosek,
M.; Figade`re, B. Angew. Chem., Int. Ed. 2000, 39, 1799. (d) Nagao, H.;
Yamane, Y.; Mukaiyama, T. Chem. Lett. 2007, 36, 8. (e) Kitajima, H.; Ito,
K.; Katsuki, T. Tetrahedron 1997, 53, 17015. (f) Desimoni, G.; Faita, G.;
Filippone, S.; Mella, M.; Zampori, M. G.; Zema, M. Tetrahedron 2001,
57, 10203. (g) Brown, S. P.; Goodwin, N. C.; MacMillan, D. W. C. J. Am.
Chem. Soc. 2003, 125, 1192. See also ref 7.
Tetrahedron 2001, 57, 3221
.
(3) Catalytic asymmetric γ-selective (vinylogous) Mannich-type reac-
tions: For direct reactions see: (a) Liu, T.-Y.; Cui, H.-L.; Long, J.; Li, B.-
J.; Wu, Y.; Ding, L.-S.; Chen, Y.-C. J. Am. Chem. Soc. 2007, 129, 1878.
(b) Niess, B.; Jørgensen, K. A. Chem. Commun. 2007, 1620. For
Mukaiyama-type reaction see: (c) Sickert, M.; Schneider, C. Angew. Chem.,
Int. Ed. 2008, 47, 3631
.
(4) For asymmetric vinylogous Mannich-type reactions of siloxyfurans,
see: (a) Martin, S. F.; Lopez, O. D. Tetrahedron Lett. 1999, 40, 8949. (b)
Carswell, E. L.; Snapper, M. L.; Hoveyda, A. H. Angew. Chem., Int. Ed.
(6) Uraguchi, D.; Sorimachi, K.; Terada, M. J. Am. Chem. Soc. 2004,
126, 11804.
2006, 45, 7230
.
10.1021/ol800756r CCC: $40.75
Published on Web 05/07/2008
2008 American Chemical Society

catalytic asymmetric Mannich(-type) reactions,8 which is a
potentially more straightforward and atom-economical pro-
cess. Here we describe the utility of a chiral Lewis acid/
amine base/Brønsted acid combination in direct catalytic
asymmetric Mannich-type reactions of γ-butenolides 1 and
N-diphenylphosphinoyl imines 2, affording the products in
up to >99% yield, anti/syn ) >97:3, and 84% ee. The use
of a catalytic amount of TfOH in addition to La(OTf)3/Me-
PyBox/TMEDA (Figure 1) was important for improving
yield and stereoselectivity.
phinoyl imine15 2a. The optimization studies are summarized
in Tables 1 and 2. The combined use of La(OTf)3/Me-PyBox/
Table 1. Conditions Screening
amine
(mol %)
time yield anti/ % eeb
entry (S,S)-PyBox
(h)
(%)
syna (anti)
1
2
3
4
5
6
7
8
Me-PyBox
Me-PyBox
Me-PyBox
Me-PyBox
Me-PyBox
Me-PyBox
Et3N (20)
Et3N (40)
none
pyridine (20)
DBU (20)
TMEDA (20)
35
35
24
24
35
35
39
35
53 81:19
62 82:18
NR
NR
0
1
10
56 93:7
17 70:30
60:40
66
11
2
i-Pr-PyBox TMEDA (20)
Ph-PyBox TMEDA (20)
Figure 1. Catalyst components and structures of γ-butenolide
donors 1 and N-diphenylphosphinoyl imines 2.
8
a Determined by crude 1H NMR analysis. b Determined by chiral HPLC
analysis.
Although the pKa values of γ-butenolides are low enough
for in situ generation of dienolate,9 other obstacles must be
overcome to achieve the desired Mannich-type reactions of
imines with γ-butenolide donors. (1) Both γ-butenolides
themselves and the γ-butenolide units in Mannich-adducts
have high reactivity as electrophiles,10 and serve as conjugate
addition acceptors to give undesirable side products. There-
fore, selective activation of γ-butenolides as donors and
imines as electrophiles is required. (2) The enantioselectivity,
diastereoselectivity, and R/γ-selectivity of dienolate must be
controlled.
In the initial screening, various chiral metal aryloxide
catalysts developed for direct Mannich-type reactions of other
donors in our group11 produced unsatisfactory results.
Additional studies revealed that a La(OTf)3/PyBox/amine
base combination12–14 was promising for the direct Mannich-
type reaction of γ-crotonolactone (1a) and N-diphenylphos-
Et3N in 1,2-dichloroethane (DCE) at -20 °C afforded
γ-addition product 3aa in 53% yield, albeit in poor enanti-
oselectivity and modest diastereoselectivity (Table 1, entries
1 and 2).16 No reaction proceeded without an amine base
(entry 3) or with pyridine (entry 4). On the other hand, imine
2a was consumed completely with DBU, but desired product
3aa was not obtained. Complex mixtures of side products
(12) Rare earth-PyBox complexes in asymmetric catalysis: For a review,
see: (a) Desimoni, G.; Faita, G.; Quadrelli, P. Chem. ReV. 2003, 103, 3119.
(b) For selected examples, see: Keith, J. M.; Jacobsen, E. N. Org. Lett.
2004, 6, 153. (c) Sammis, G. M.; Danjo, H.; Jacobsen, E. N. J. Am. Chem.
Soc. 2004, 126, 9928. (d) Evans, D. A.; Fandrick, K. R.; Song, H.-J.; Scheidt,
K. A.; Xu, R. J. Am. Chem. Soc. 2007, 129, 10029. (e) Evans, D. A.; Song,
H.-J.; Fandrick, K. R. Org. Lett. 2006, 8, 3351. (f) Aspinall, H. C.; Bickley,
J. F.; Greeves, N.; Kelly, R. V.; Smith, P. M. Organometallics 2005, 24,
3458. (g) Comelles, J.; Pericas, A.; Moreno-Manas, M.; Vallribera, A.;
Drudis-Sole, G.; Lledos, A.; Parella, T.; Roglans, A.; Garcia-Granda, S.;
Roces-Fernandez, L. J. Org. Chem. 2007, 72, 2077. (h) Desimoni, G.; Faita,
G.; Toscanini, M.; Boiocchi, M. Chem. Eur. J. 2007, 13, 9478.
(7) Direct use of γ-butenolides in racemic aldol reactions was reported:
(a) Franck, X.; Figade`re, B. Tetrahedron Lett. 2002, 43, 1449. (b) Sarma,
K. D.; Zhang, J.; Curran, T. T. J. Org. Chem. 2007, 72, 3311.
(8) For recent reviews on direct catalytic Mannich(-type) reactions
affording ꢀ-amino carbonyl compounds, see: (a) Marques, M. M. B. Angew.
Chem., Int. Ed. 2006, 45, 348. (b) Shibasaki, M.; Matsunaga, S. J.
Organomet. Chem. 2006, 691, 2089. (c) Verkade, J. M. M.; van Hemert,
L. J. C.; Quaedflieg, P. J. L. M.; Rutjes, F. P. J. T. Chem. Soc. ReV. 2008,
37, 29. (d) Ting, A.; Schaus, S. E. Eur. J. Org. Chem. 2007, 5797.
(9) The pKa value of γ-crotonolactone (1a) was estimated to be 17.7 in
DMSO. Calculations were performed by using the B3LYP level of the
density functional theory. See the Supporting Information for details.
(10) For example, treatment of 1a with La(OTf)3, Me-PyBox, and Et3N
in the absence of imine 2a gave oligomers of 1a.
(13) Combination of metal triflate and amine base in asymmetric
catalysis: For reviews, see: (a) Kobayashi, S.; Sugiura, M.; Kitagawa, H.;
Lam, W. W.-L. Chem. ReV. 2002, 102, 2227. (b) Kanemasa, S.; Ito, K.
Eur. J. Org. Chem. 2004, 4741. For selected examples, see: (c) Itoh, K.;
Kanemasa, S. J. Am. Chem. Soc. 2002, 124, 13394. (d) Itoh, K.; Hasegawa,
M.; Tanaka, J.; Kanemasa, S. Org. Lett. 2005, 7, 979. (e) Itoh, K.;
Oderaotoshi, Y.; Kanemasa, S. Tetrahedron: Asymmetry 2003, 14, 635. (f)
Frantz, D. E.; Fa¨ssler, R.; Carreira, E. M. J. Am. Chem. Soc. 2000, 122,
1806. (g) Anand, N. K.; Carreira, E. M. J. Am. Chem. Soc. 2001, 123,
9687. (h) Takita, R.; Yakura, K.; Ohshima, T.; Shibasaki, M. J. Am. Chem.
Soc. 2005, 127, 13760. (i) Juhl, K.; Gathergood, N.; Jørgensen, K. A. Chem.
Commun. 2000, 2211. (j) Gathergood, N.; Juhl, K.; Poulsen, T. B.; Thordrup,
K.; Jørgensen, K. A. Org. Biomol. Chem. 2004, 2, 1077. (k) Kjærsgaard,
A.; Jørgensen, K. A. Org. Biomol. Chem. 2005, 3, 804. (l) Palomo, C.;
Oiarbide, M.; Halder, R.; Laso, A.; Lo´pez, R. Angew. Chem., Int. Ed. 2006,
45, 117. (m) Tur, F.; Saa, J. M. Org. Lett. 2007, 9, 5079.
(11) (a) Morimoto, H.; Lu, G.; Aoyama, N.; Matsunaga, S.; Shibasaki,
M. J. Am. Chem. Soc. 2007, 129, 9588. (b) Matsunaga, S.; Yoshida, T.;
Morimoto, H.; Kumagai, N.; Shibasaki, M. J. Am. Chem. Soc. 2004, 126,
8777. (c) Harada, S.; Handa, S.; Matsunaga, S.; Shibasaki, M. Angew.
Chem., Int. Ed. 2005, 44, 4365. (d) Yamaguchi, A.; Aoyama, N.; Matsunaga,
S.; Shibasaki, M. Org. Lett. 2007, 9, 3387. (e) Handa, S.; Gnanadesikan,
V.; Matsunaga, S.; Shibasaki, M. J. Am. Chem. Soc. 2007, 129, 4900. (f)
Chen, Z.; Morimoto, H.; Matsunaga, S.; Shibasaki, M. J. Am. Chem. Soc.
2008, 130, 2170.
(14) Other metal triflates gave less satisfactory results. No reaction
proceeded with Cu(OTf)2, In(OTf)3, and Zn(OTf)2. Some of other rare earth
metal triflates promoted the reaction of imine 2a with donor 1a, but
enantioselectivity was lower than La(OTf)3 or comparable to La(OTf)3.
(15) Weinreb, A. M.; Orr, R. K. Synthesis 2005, 1205.
(16) Stereochemistry was determined by conversion to known com-
pounds. See the Supporting Information for details.
2320
Org. Lett., Vol. 10, No. 11, 2008

Table 2. Conditions Screening
Table 3. Direct Catalytic Asymmetric Mannich-Type Reactions
of γ-Butenolides 1 and Imines 2
Brønsted acid Me-PyBox time yield anti/ % eeb
entry
(mol %)
(mol %)
(h)
(%)
syna (anti)
time yielda
(h) (%)
anti/ % eec
synb (anti)
entry
R1
2
1
3
1
2
3
4
5
6
none
10
10
10
10
10
10
15
10
35
35
35
35
39
24
72
39
56 93:7
16 76:24
66
40
65
68
79
85
84
PMPOHc (10)
AcOH (10)
MsOH (10)
TfOH (10)
TfOH (15)
TfOH (10)
TfOH (10)
1
2
3
4
5
6
Ph
2b 1a 72
2a 1a 72
2c 1a 69
2d 1a 69
2e 1a 45
2f 1a 50
83 3ba 96:4
89 3aa 96:4
85 3ca
80 3da 96:4
99 3ea
>99 3fa
84 3ga 86:14 78
86 3ha 82:18 72
83
84
79
83
83
84
81 52:48
43 88:12
86 96:4
22 97:3
89 96:4
4-MeC6H4-
2-MeC6H4-
4-MeOC6H4-
4-ClC6H4-
2-thienyl
91:9
95:5
97:3
7
8d
NR
7d (CH3)2CHCH2- 2g 1a 31
a Determined by crude 1H NMR analysis. b Determined by chiral HPLC
8d CH3(CH2)3-
2h 1a 43
analysis. c 4-MeOC6H4OH. d Without La(OTf)3.
9e Ph
2b 1b 62
82 3bb >97:3
68
a Isolated yield after column chromatography. b Determined by crude
1H NMR analysis. c Determined by chiral HPLC analysis. d 2.5 equiv of
1a was used. e Reaction was performed at rt. 20 mol % of La(OTf)3, (S,S)-
Me-PyBox, TfOH, and 40 mol % of TMEDA were used.
were observed, possibly due to undesirable conjugate addi-
tions to γ-butenolide units (entry 5). TMEDA was the most
promising in terms of diastereo- and enantioselectivity (anti/
syn ) 93:7 and 66% ee), giving 3aa in 56% yield (entry 6).
PyBox ligands with bulkier substituents than those in the
Me-PyBox, such as i-Pr-PyBox and Ph-PyBox, afforded 3aa
in low yield and stereoselectivity (entries 7 and 8).
aryl imines 2a-e with various substituents on the aromatic
ring, as well as heteroaryl imine 2f, gave Mannich adducts
in >99-80% yield, 97:3-91:9 anti-selectivity, and 84-79%
ee (entries 1-6). Imine 2c with ortho-substituent 3ca
somewhat diminished both the diastereo- and enantioselec-
tivity (anti/syn ) 91:9, 79% ee, entry 3). It is noteworthy
that readily isomerizable alkyl imines 2g and 2h17 were also
applicable, giving 3ga and 3ha in 84% and 86% yield,
respectively. Enantio- and diastereoselectivity, however, both
had room for improvement (entries 7 and 8). A Mannich-
type reaction of 3-methyl-(5H)-furan-2-one (1b) was per-
formed at room temperature because no reaction proceeded
at -20 or 0 °C (entry 9). With donor 1b, 20 mol % of
La(OTf)3 was required for good yield and moderate enan-
tioselectivity (entry 9).
At the beginning of the initial screening, the reactions had
low reproducibility. For example, the conditions of entry 5
in Table 1 afforded 3aa in variable yield (38-76%) and
enantioselectivity (44-79% ee). Detailed studies to overcome
the reproducibility problem revealed that a small amount of
TfOH contaminated in La(OTf)3 was important. Therefore,
La(OTf)3 was carefully flame-dried under reduced pressure
to remove TfOH prior to use in all experiments described in
this paper to obtain good reproducibility, and the accurate
amount of Brønsted acid was added in Table 2. The effects
of additional Brønsted acid are shown in Table 2. The
addition of 10 mol % of 4-MeOC6H4OH to a mixture of
La(OTf)3/Me-PyBox/TMEDA ) 1:1:2 (La ) 10 mol %) had
adverse effects on both reactivity and stereoselectivity (entry
2). An AcOH additive afforded 3aa in 81% yield, but the
diastereoselectivity was low (anti/syn ) 52:48, entry 3).
MsOH diminished the yield and diastereoselectivity, although
the enantioselectivity was comparable to that of entry 1 (entry
4). Among Brønsted acids investigated, TfOH was the best,
giving 86% yield and 79% ee (entry 5). Increasing TfOH
loading to 15 mol % gave 3aa in 85% ee, albeit in poor
yield (entry 6). 15 mol % of Me-PyBox in the presence of
10 mol % of TfOH gave 3aa in 89% yield, anti/syn ) 96:4,
and 84% ee (entry 7). No reaction proceeded in the absence
of La(OTf)3, implying that neither Me-PyBox/TfOH nor
TMEDA/TfOH is an effective catalyst species in the present
system (entry 8).
In the present reaction, the catalytic amount of TfOH was
key to achieving the optimized enantioselectivity. To glean
1
insight on the role of TfOH, we obtained H NMR spectra
of the La(OTf)3/Me-PyBox/TMEDA mixture at a ratio of
1:1:2 in CDCl3 with and without TfOH (Figure 2). In the
absence of TfOH, metal-free Me-PyBox was the major
constituent and Me-PyBox complexed with La metal was
the minor constituent (Figure 2A), suggesting that
(tmeda)2La(OTf)3 would be the major La species. In the
presence of TfOH, the ratio of metal-free Me-PyBox/metal-
bound Me-PyBox changed, and Me-PyBox complexed with
La metal was observed as the major species (Figure 2B).
We speculate that La(OTf)3/Me-PyBox/TMEDA ) 1:1:1
would be the active species18,19 in the present reaction, and
(17) For synthesis of alkyl N-diphenylphosphinoyl imines, see: (a) Trost,
B. M.; Jaratjaroonphong, J.; Reutrakul, V. J. Am. Chem. Soc. 2006, 128,
2778. (b) Yamaguchi, A.; Matsunaga, S.; Shibasaki, M. Tetrahedron Lett.
2006, 47, 3985.
Substrate scopes and limitations were investigated under
the optimized reaction conditions (Table 3). Nonisomerizable
Org. Lett., Vol. 10, No. 11, 2008
2321

Figure 3. The postulated catalytic cycle of the Mannich-type
reaction.
activated by Lewis acidic (pybox)(tmeda)La(OTf)3 A to give
La-dienolate B and TMEDA/(TfOH)2. B would react with
imine 2 to give D via transition state C to minimize steric
repulsion. D would be protonated by TMEDA/(TfOH)2 to
give product 3 and regenerate complex A.
In conclusion, we developed direct catalytic asymmetric
Mannich-type reactions of γ-butenolides using a chiral Lewis
acid/amine base/Brønsted acid combination. La(OTf)3/Me-
PyBox/TMEDA/TfOH afforded Mannich adducts 3 in up to
>99% yield, anti/syn ) >97:3, and 84% ee. The addition
of TfOH was important for achieving good yield and
stereoselectivity. Further investigations on extending Lewis
acid/amine base/Brønsted acid combinations to other catalytic
enantioselective reactions are ongoing.
Figure 2.
1H NMR spectra of La(OTf)3/Me-PyBox/TMEDA
mixture (A) without TfOH and (B) with TfOH [3.95-4.80 and
7.75-8.20 ppm regions].
that TfOH would support the generation of the desirable
active species. In the reaction without TfOH, a racemic
reaction promoted by the predominant (tmeda)2La(OTf)3
complex20 competed with the enantioselective reaction
promoted by a minor chiral La complex, resulting in only
moderate yield and enantioselectivity (Table 2, entry 1).
Reducing (tmeda)2La(OTf)3 by the addition of TfOH im-
proved both the yield and enantioselectivity (Table 2, entry
5). 15 mol % of Me-PyBox gave better enantioselectivity
than 10 mol % of Me-PyBox (Table 2, entry 7 vs entry 5),
possibly due to the decreased amount of undesirable
(tmeda)2La(OTf)3 in the presence of a slight excess of Me-
PyBox. (Table 2, entry 7).
Acknowledgment. This work was supported by Grant-
in-Aid for Specially Promoted Research, and the Sumitomo
Foundation (for S.M.). A.Y. thanks the JSPS Fellowship for
Young Scientists.
Supporting Information Available: Experimental pro-
cedures, spectral data of products, NMR charts of catalyst
mixtures, and determination of stereochemistry of products.
This material is available free of charge via the Internet at
http://pubs.acs.org.
The postulated catalytic cycle is shown in Figure 3.
TMEDA/TfOH would deprotonate γ-butenolide 1, which is
OL800756R
(18) We speculate that the active catalyst may contain TMEDA as a
ligand, because enantioselectivity strongly depended on amines in Table
1.
(20) La(OTf)3/TMEDA ) 1:2 (La ) 10 mol %, without Me-PyBox)
with and without 10 mol % of TfOH catalyzed the Mannich-type reaction
of imine 2a and donor 1a, affording 3aa in 15% yield (35 h, anti/syn )
84:16, without TfOH) and 18% yield (35 h, anti/syn ) 94:6, with TfOH).
(19) La(OTf)3/Me-PyBox/TMEDA ) 1:1:1 without TfOH (La ) 10 mol
%) gave 3aa in 56% yield, anti/syn ) 93:7, and 73% ee from 2a and 1a.
Thus, the use of TfOH is essential to obtain optimum results.
2322
Org. Lett., Vol. 10, No. 11, 2008