Managing highly coordinative substrates in asymmetric catalysis: A catalytic asymmetric amination with a lanthanum-based ternary catalyst

Article abstract of DOI:10.1021/ja9052653

Full details of a catalytic asymmetric amination with a lanthanum/amide-based ligand catalyst system are described. A catalyst comprising La(NO₃)₃·6H₂O, (R)-3a and H-D-Val-O^tBu was identified to promote the catalytic asymmetric amination of nonprotected succinimide derivative 1 with as little as 1 mol % catalyst loading. Mechanistic studies by various spectroscopic analyses and several control and kinetic experiments suggested that the catalyst components were in equilibrium between the associated and dissociated forms, and that the reaction likely proceeded through a La(NO₃)₃· 6H₂O/(R)-3a/H-D-Val-O^tBu ternary complex. This catalyst system was also effective for asymmetric amination of N-nonsubstituted α-alkoxycarbonyl amides 7, hitherto unprecedented substrates in asymmetric catalysis, probably due to their attenuated reactivity and difficult stereocontrol, affording the amination products in up to >99% yield and >99% ee. The high catalytic performance and enantiocontrol of the reaction with highly coordinative substrates were achieved by the activation/recognition of the substrates exerted by coordination to lanthanum and hydrogen bonding cooperatively in the transition state.

Full text of DOI:10.1021/ja9052653

Published on Web 09/28/2009
Managing Highly Coordinative Substrates in Asymmetric
Catalysis: A Catalytic Asymmetric Amination with a
Lanthanum-Based Ternary Catalyst
Tomoyuki Mashiko, Naoya Kumagai,* and Masakatsu Shibasaki*
Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo, 113-0033, Japan
Received June 26, 2009; E-mail: mshibasa@mol.f.u-tokyo.ac.jp; nkumagai@mol.f.u-tokyo.ac.jp
Abstract: Full details of a catalytic asymmetric amination with a lanthanum/amide-based ligand catalyst
system are described. A catalyst comprising La(NO3)3•6H2O, (R)-3a and H-D-Val-OtBu was identified to
promote the catalytic asymmetric amination of nonprotected succinimide derivative 1 with as little as 1 mol
% catalyst loading. Mechanistic studies by various spectroscopic analyses and several control and kinetic
experiments suggested that the catalyst components were in equilibrium between the associated and
dissociated forms, and that the reaction likely proceeded through a La(NO3)3•6H2O/(R)-3a/H-D-Val-OtBu
ternary complex. This catalyst system was also effective for asymmetric amination of N-nonsubstituted
R-alkoxycarbonyl amides 7, hitherto unprecedented substrates in asymmetric catalysis, probably due to
their attenuated reactivity and difficult stereocontrol, affording the amination products in up to >99% yield
and >99% ee. The high catalytic performance and enantiocontrol of the reaction with highly coordinative
substrates were achieved by the activation/recognition of the substrates exerted by coordination to lanthanum
and hydrogen bonding cooperatively in the transition state.
bonyl compounds was reported in 1954,5 only recently has this
same transformation been rendered asymmetric using a mag-
Introduction
Nitrogen-containing substructures in which nitrogen is directly
attached to a stereogenic carbon are a common structural motif
in a wide variety of biologically active compounds and
therapeutics.1 Recent advances in asymmetric catalysis have
enabled the enantioselective construction of this class of
substructures. Because of the nucleophilic character of nitrogen,
the majority of strategies involve the enantioselective formation
of a C-N bond with a nitrogen-based nucleophile or CdN bond
formation followed by enantioselective addition to the CdN
bond.2 Alternatively, electrophilic amination using formal
“NH2+” reagents constitutes a complementary methodology,
which allows the installation of a nitrogen functionality adjacent
to a carbonyl group.3 Among the amination reagents used,
azodicarboxylates have gained popularity due to their com-
mercial availability and high electrophilic nature.4 Although the
utility of azodicarboxylates in electrophilic amination of car-
nesium bis(sulfonamide) catalyst.6 Several catalytic asymmetric
electrophilic aminations of carbonyl compounds have been
successively reported with both metal-based catalysts and
organocatalysts.7-9
We were particularly interested in the catalytic asymmetric
amination of nonprotected succinimide derivative 1, which
(4) For a general review of the use of azodicarboxylates in C-N bond-
forming reactions, see: (a) Nair, V.; Biju, A. T.; Mathew, S. C.; Babu,
B. P. Chem. Asian J. 2008, 3, 810.
(5) (a) Huisgen, R.; Jakob, F Justus Liebigs Ann. Chem. 1954, 590, 37.
For electrophilic amination using azodicarboxylates, see: (b) Diels, O
Justus Liebigs Ann. Chem. 1922, 429, 1. (c) Diels, O.; Behncke, H.
Chem. Ber. 1924, 57, 653.
(6) (a) Evans, D. A.; Nelson, S. G. J. Am. Chem. Soc. 1997, 119, 6452.
(b) Evans, D. A.; Johnson, D. S. Org. Lett. 1999, 1, 595.
(7) For reviews of catalytic asymmetric electrophilic amination of carbonyl
compounds, see: (a) Duthaler, R. O. Angew. Chem., Int. Ed. 2003,
42, 975. (b) Greck, C; Drouillat, B.; Thomassigny, C. Eur. J. Org.
Chem. 2004, 1377. (c) Janey, J. M. Angew. Chem., Int. Ed. 2005, 44,
4292. (d) Cativiela, C.; Ordo´n˜ez, M. Tetrahedron: Asymmetry 2009,
20, 1.
(1) (a) Genet, J.-P., Greck, C.; Lavergne, D. In Modern Amination
Methods; Ricci, A., Ed.; Wiley-VCH: Weinheim, 2000; chapter 3. (b)
Krohn, K. In Organic Synthesis Highlights; Wiley-VCH: Weinheim,
1991; pp 45-53.
(8) For selected examples using metal-based catalysts, see: (a) Marigo,
M.; Juhl, K.; Jørgensen, K. A. Angew. Chem., Int. Ed. 2003, 42, 1367.
(b) Ma, S.; Jiao, N.; Zheng, Z.; Ma, Z.; Lu, Z.; Ye, L.; Deng, Y.;
Chen, G. Org. Lett. 2004, 6, 2193. (c) Foltz, C.; Stecker, B.; Marconi,
G.; Bellemin-Laponnaz, S.; Wadepohl, H.; Gabe, L. H. Chem.
Commun. 2005, 5115. (d) Bernardi, L.; Zhuang, W.; Jørgensen, K. A.
J. Am. Chem. Soc. 2005, 127, 5772. (e) Kim, Y. K.; Kim, D. Y.
Tetrahedron Lett. 2006, 47, 4565. (f) Comelles, J.; Pericas, A.;
Moreno-Man˜as, M.; Vallribera, A.; Drudis-Sole´, G.; Lledos, A.;
Parella, T.; Roglans, A.; Garc´ıa-Granda, S.; Roces-Ferna´ndez, L. J.
Org. Chem. 2007, 72, 2077. (g) Hasegawa, Y.; Watanabe, M.; Gridnev,
I. D.; Ikariya, T. J. Am. Chem. Soc. 2008, 130, 2158. (h) Mang, J. Y.;
Kwon, D. G.; Kim, D. Y. J. Fluorine Chem. 2009, 130, 259. (i) Ikariya,
T.; Gridnev, I. D. Chem. Record. 2009, 9, 106.
(2) For selected reviews on catalytic asymmetric addition to CdN double
bond, see: (a) Friestad, G. K.; Mathies, A. K. Tetrahedron 2007, 63,
2541. (b) Weinreb, S. M.; Orr, R. K. Synthesis 2005, 1205. (c)
Marques, M. M. B. Angew. Chem., Int. Ed. 2006, 45, 348. (d) Ting,
A.; Schaus, S. E. Eur. J. Org. Chem. 2007, 5797. (e) 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. For a general review on
catalytic asymmetric Mannich reactions, see: (f) Kobayashi, S.; Ueno,
M. In ComprehensiVe Asymmetric Catalysis, Supplement 1; Jacobsen,
E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer: Berlin, 2004; pp 143-
150 (chap 29.5).
(3) For reviews, see: (a) Genet, J. P.; Greck, C. Synlett 1997, 741. (b)
Erdik, E. Tetrahedron 2004, 60, 8747.
9
14990 J. AM. CHEM. SOC. 2009, 131, 14990–14999
10.1021/ja9052653 CCC: $40.75  2009 American Chemical Society

Highly Coordinative Substrates in Asymmetric Catalysis
A R T I C L E S
modes displayed by 1.13 To address this issue, we initiated
studies aimed at developing a catalytic asymmetric amination
of 1.14,15 We recently achieved a concise enantioselective
synthesis of (-)-2, in which the catalytic asymmetric amination
of 1 with a lanthanum/amide-based ligand catalyst was a key
step.14,16,17 The amination reaction was scalable, cost-effective,
and the investigations of an industrial application are
underway.14b This article describes the full details of the catalytic
asymmetric amination with the La/amide-based ligand catalyst
system. Based on analyses of the privileged functional group
arrangement of the highly coordinative substrates, the substrate
scope was successfully expanded to other nonprotected sub-
strates, such as N-nonsubstituted R-alkoxycarbonyl amides.
Mechanistic studies suggested that the three components of the
second-generation catalyst were in a dynamic equilibrium and
an associated ternary complex was involved in the transition
state.
Scheme 1. Concise Enantioselective Synthesis of (-)-2
allows for an efficient enantioselective synthesis of AS-
3201((-)-2, ranirestat), a highly potent aldose reductase inhibitor
(Scheme 1). (-)-2 has attracted particular attention because of
its effectiveness for treating diabetic complications via oral
administration.10 Diabetes is one of the most serious health
concerns throughout the world, and the number of patients with
diabetes is rapidly increasing.11 The main focus in the treatment
of diabetes is on decreasing blood glucose levels, and there are
few effective therapeutics for the treatment of late-stage
pathologic conditions that seriously compromise patients’ quality
of life.12 Clinical trials on (-)-2 specifically for diabetic
neuropathy are ongoing, and therapeutic development is highly
anticipated. In this context, an efficient route for the enantiose-
lective synthesis of (-)-2 is in high demand to fulfill a
prospective future supply. Catalytic asymmetric amination of
succinimide derivative 1 emerged as a primary option toward
this end. The Lewis basicity of nonprotected 1, however,
interferes with the efficient enantiodifferentiation by Lewis
acidic chiral catalysts, likely because of the multiple coordination
Results and Discussion
Development of the First-Generation Amination Catalyst.
Our initial effort was devoted to the development of a new class
of catalysts for catalytic asymmetric amination of nonprotected,
highly coordinative succinimide derivative 1. To avoid ligand
dissociation and/or a multiple coordination pattern of 1, we
initially focused on an asymmetric catalyst comprising a rare
earth metal (RE)/amide-based ligand, where the high affinity
of the amide functionality to RE would avoid ligand dissociation
by 1. Taking high coordination number of REs into account, 1
could coordinate to REs surrounded by amide-based ligands and
the hydrogen bonding provided by the ligand could specify the
coordination mode of 1. Three types of amide-based ligands
bearing a phenol functionality for coordination to RE, salicylic
acid-type (S)-3, biphenol-type (S)-4, and hydroxamic acid-type
(S)-5, were examined in the catalytic asymmetric amination of
1 and di-t-butyl azodicarboxylate (Scheme 2). The catalyst was
prepared by mixing the ligand and La(OiPr)3 in a 2:1 ratio. The
catalyst prepared from ligands (S)-3a-d or (S)-4 efficiently
promoted the desired reaction at -40 °C to give (+)-6, whereas
the catalyst prepared from (S)-5 failed the reaction, presumably
because the strongly coordinative hydroxamic acid induced the
formation of a catalytically inactive aggregate. The reaction with
(S)-3a gave the highest enantioselectivity, and further investiga-
tion revealed that lanthanum worked best among the REs tested,
(9) For selected examples using organocatalysts, see: (a) Sabby, S.; Bella,
M.; Jørgensen, K. A. J. Am. Chem. Soc. 2004, 126, 8120. (b) Pihko,
P. M.; Pohjakallio, A. Synlett 2004, 2115. (c) Liu, X.; Li, X.; Deng,
L. Org. Lett. 2005, 7, 167. (d) Xu, X.; Yabuta, T.; Yuan, P.; Takemoto,
Y. Synlett 2006, 137. (e) Terada, M.; Nakano, M.; Ube, H. J. Am.
Chem. Soc. 2006, 128, 16044. (f) Liu, Y.; Melgar-Fernandez, R.;
Juaristi, E. J. Org. Chem. 2007, 72, 1522. (g) Jung, S. H.; Kim, D. Y.
Tetrahedron Lett. 2008, 49, 5527. (h) He, R.; Wang, X.; Hashimoto,
T.; Maruoka, K. Angew. Chem., Int. Ed. 2008, 47, 9466. (i) Lan, Q.;
Wang, X.; He, R.; Ding, C.; Maruoka, K. Tetrahedron Lett. 2009, 50,
3280. (j) Liu, X.; Sun, B.; Deng, L. Synlett 2009, 10, 1685. (k) Cheng,
L.; Liu, L.; Wang, D.; Chen, Y.-J. Org. Lett. 2009, 11, 3874.
(10) (a) Negoro, T.; Murata, M.; Ueda, S.; Fujitani, B.; Ono, Y.; Kuromiya,
A.; Komiya, M.; Suzuki, K.; Matsumoto, J.-I. J. Med. Chem. 1998,
41, 4118. (b) Kurono, M.; Fujiwara, I.; Yoshida, K. Biochemistry 2001,
40, 8216. (c) Bril, V.; Buchanan, R. A. Diabetes Care 2004, 27, 2369.
(d) Kurono, M.; Fujii, A.; Murata, M.; Fujitani, B.; Negoro, T.
Biochem. Pharmacol. 2006, 71, 338. (e) Giannoukakis, N. Curr. Opin.
InVestig. Drugs 2006, 7, 916. (f) Matsumoto, T.; Ono, Y.; Kurono,
M.; Kuromiya, A.; Nakamura, K.; Bril, V. J. Pharmacol. Sci. 2008,
107, 231.
(13) Representative metal catalysts (Cu-BOX, LaPyBox) and organocata-
lysts (chiral urea catalyst, cinchona alkaloids) afforded the desired
product in less than 30% ee (see ref 14a).
(14) (a) Mashiko, T.; Hara, K.; Tanaka, D.; Fujiwara, Y.; Kumagai, N.;
Shibasaki, M. J. Am. Chem. Soc. 2007, 129, 11342. (b) Mashiko, T.;
Kumagai, N.; Shibasaki, M. Org. Lett. 2008, 10, 2725.
(11) (a) Nath, D.; Heemels, M.-T.; Anson, L. Nature 2006, 444, 839. and
the following review articles. (b) International Diabetes Federation
(IDF), Diabetes Atlas, 3rd Edition, Dec, 2006; http://www.iotf.org/
diabetes.asp.
(15) Catalytic asymmetric amination of N-Boc protected 1 was reported
in refs 9h and 9i.
(16) Other examples of asymmetric catalysis using rare earth metal/amide-
based ligand catalysts, see: (a) Nishida, A.; Yamanaka, M.; Nakagawa,
M. Tetrahedron Lett. 1999, 40, 1555. (b) Nitabaru, T.; Kumagai, N.;
Shibasaki, M. Tetrahedron Lett. 2008, 49, 272. (c) Sudo, Y.; Shirasaki,
D.; Harada, S.; Nishida, A. J. Am. Chem. Soc. 2008, 130, 12588. (d)
Nojiri, A.; Kumagai, N.; Shibasaki, M. J. Am. Chem. Soc. 2008, 130,
5630. (e) Nojiri, A.; Kumagai, N.; Shibasaki, M. J. Am. Chem. Soc.
2009, 131, 3779.
(12) For selected examples of the studies toward the treatment of diabetic
complications: Selective PKC ꢀ2 inhibitor for the treatment of diabetic
neuropathy, see: (a) Inoguchi, T.; Battan, R.; Handler, E.; Sportsman,
J. R.; Heath, W.; King, G. L. Proc. Natl. Acad. Sci. U.S.A. 1992, 89,
11059. (b) Ishii, H.; Jirousek, M. R.; Koya, D.; Takagi, C.; Xia, P.;
Clermont, A.; Bursell, S.-E.; Kern, T. S.; Ballas, L. M.; Heath, W. F.;
Stramm, L. E.; Feener, E. P.; King, G. L. Science 1996, 272, 728. (c)
Tanaka, M.; Sagawa, S.; Hoshi, J.-I.; Shimoma, F.; Yasue, K.;
Ubukata, M.; Ikemoto, T.; Hase, Y.; Takahashi, M.; Sasase, T.; Ueda,
N.; Matsushita, M.; Inaba, T. Bioorg. Med. Chem. 2006, 14, 5781,
and references cited therein. Inhibition of vascular endothelial growth
factor (VEGF) for prevention of diabetic microvascular complications,
see: (d) Gemma, T.; Rosangela, L.; Gabriella, M.; Gianpaolo, Z. Am. J.
CardioVasc. Drugs 2007, 7, 393.
(17) For recent selected examples of small peptide-based asymmetric
catalysis, see: (a) Zhao, Y.; Rodrigo, J.; Hoveyda, A. H.; Snapper,
M. C. Nature 2006, 443, 67. (b) Lewis, C. A.; Chiu, A.; Kubryk, M.;
Balsells, J.; Pollard, D.; Esser, C. K.; Murry, J.; Reamer, R. A.; Hansen,
K. B.; Miller, S. J. J. Am. Chem. Soc. 2006, 128, 16454. For reviews,
see: (c) Hoveyda, A. H.; Hird, A. W.; Kacprzynski, M. A. Chem.
Commun. 2004, 1779. (d) Blank, J. T.; Miller, S. J. Biopolymers (Pept.
Sci.) 2006, 84, 38.
9
J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009 14991

A R T I C L E S
Mashiko et al.
Scheme 2. Catalytic Asymmetric Amination of La/Amide-Based
Ligand Catalyst
Table 1. Screening of Several La Saltsa
entry
La salt
x
additive (30 mol %)
time (h)
ee (%)
1
2c
3
LaCl3
LaCl3
LaBr3
LaBr3
LaI3
La(OTf)3
LaCl3
LaCl3
La(NO3)3 ·6H2O
La(NO3)3 ·6H2O
10
10
10
-
0.5
0.5
4
0.75
1
1
4.5
10
4
3
-2
29
83
23
14
11
82
67
68
58
-
-
4d
5
2.5
-
10
10
10
10
10
-
6
7
8
-
nBu4N(NO3)
NH4(NO3)
-
-
9
10d
2.5
a 1: 0.2 mmol, BocNdNBoc: 0.24 mmol. b Determined by 1H NMR
analysis with Bn2O as an internal standard. c LaCl3 and KOtBu were
stirred at 50 °C for 30 min. d 1: 0.4 mmol, BocNdNBoc: 0.48 mmol.
Scheme 3. First-Generation Amination Catalyst
and (-)-6 was obtained in a nearly racemic form (entry 1).
Stirring the LaCl3/KOtBu mixture at 50 °C before adding the
(R)-3a marginally improved the enantioselectivity (entry 2).
Among other lanthanum halides tested, LaBr3 provided remark-
able enantioselectivity (83% ee), but the enantioselectivity was
dependent on catalyst loading and less reproducible, possibly
due to its poor solubility (entries 3,4). The use of La(OTf)3 as
a soluble lanthanum salt hardly improved the enantioselectivity
(entry 6). Based on the fact that nitrate anions enhance the
solubility of lanthanide complex,20 nBu4N(NO3) was added to
the reaction using LaCl3, which significantly improved the
enantioselectivity to 82% ee (entry 1 vs entry 7). NH4NO3 was
also effective for catalytic efficiency (entry 8). Eventually,
La(NO3)3•6H2O was identified as the best lanthanum source for
our purpose due to its relatively high solubility in AcOEt,
stability to moisture, commercial availability in large quantities,
and low cost ($443/500 g, Aldrich),19 affording the amination
product in 95% yield and 68% ee and the catalyst loading can
be reduced to 2.5 mol % (entries 9,10).21
Because additional base can promote the background reaction
leading to racemic product and/or be directly involved in the
transition state, the character of the base was expected to have
a significant impact on enantioselectivity. Thus, we next
examined other bases in combination with La(NO3)3•6H2O
(Table 2). We anticipated that amine bases located in near
proximity to La3+ by coordination and worked cooperatively
with La(NO3)3•6H2O and ligand (R)-3a,22 preventing the
undesirable background reaction promoted by the base itself.
Indeed, generally higher enantioselectivity was observed in the
reaction using various amine bases. Of particular note is that
the structure of the amines, including their stereochemistry, was
a determining factor for enantioselectivity in the reaction,
and CHCl3 solvent in the presence of N,N-dimethylacetamide
(DMA) was optimum, establishing a first-generation catalyst.14a
With 2 mol % catalyst, the asymmetric amination proceeded
smoothly at 0 °C under air to furnish amination product (-)-6
in >99% yield and 92% ee (Scheme 3). (R)-3a was readily
synthesized from Boc-D-valine in high yield via a 5-pot sequence
without chromatographic purification. Compound (-)-6 was
successfully converted to (-)-2 in an optically pure form after
a 5-step transformation and a single recrystallization.10a,14a,18
Development of the Second-Generation Catalyst. Although
the first-generation catalyst provided the amination product in
92% ee under air, several concerns remained in regard to the
industrial application; (1) La(OiPr)3 is costly ($224/3 g, Ald-
rich)19 and not available in bulk quantities; (2) La(OiPr)3 is
unstable to moisture and must be handled in an inert atmosphere;
(3) fluctuation in both reactivity and enantioselectivity was
occasionally observed depending on the production lot of
La(OiPr)3; and (4) the use of halogenated solvent (CHCl3) is
undesired. To address these issues, we began to search for a
more appropriate lanthanum source (Table 1). In contrast to the
basic nature of La(OiPr)3, the readily available lanthanum salts
shown in Table 1 are neutral and required an additional base to
promote the amination of 1. The amination reaction was
performed with a catalyst prepared from La salt/(R)-3a/KOtBu
in a 1:2:3 ratio using AcOEt as the solvent, which is less toxic
than a halogenated solvent and more favorable for industrial
implementation. When 10 mol % of LaCl3 was used, the reaction
mixture was heterogeneous due to the poor solubility of LaCl3
(20) (a) Hayano, T.; Sakaguchi, T.; Furuno, H.; Ohba, M.; Okawa, H.;
Inanaga, J. Chem. Lett. 2003, 32, 608. (b) Furuno, H.; Hayano, T.;
Kambara, T.; Sugimoto, Y.; Hanamoto, T.; Tanaka, Y.; Jin, Y. Z.;
Kagawa, T.; Inanaga, J. Tetrahedron 2003, 59, 10509.
(21) Recent examples for the use of lanthanide nitrates for asymmetric
catalysis, see: (a) Hamada, T.; Manabe, K.; Ishikawa, S.; Nagayama,
S.; Shiro, M.; Kobayashi, S. J. Am. Chem. Soc. 2003, 125, 2989, and
ref 20b. Lanthanide nitrates generally exhibit much less catalytic
efficiency compared with corresponding triflates. For example, see:
(b) Kano, S.; Nakano, H.; Kojima, M.; Baba, N.; Nakajima, K. Inorg.
Chim. Acta 2003, 349, 6.
(18) For an alternative enantioselective synthetic approach toward AS-3201,
see: Watanabe, T.; Kawabata, T. Heterocycles 2008, 76, 1593.
(19) Sigma-Aldrich Handbook of Fine Chemicals 2009-2010; Sigma-
Aldrich: Milwaukee, WI, 2008.
9
14992 J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009

Highly Coordinative Substrates in Asymmetric Catalysis
A R T I C L E S
Table 2. Catalytic Asymmetric Amination of 1 with La(NO3)3/
(R)-3a/base Ternary Catalysta
Scheme 4. One-Hundred Gram Scale Demonstration of the
Second-Generation Catalytic Asymmetric Amination
Catalytic Asymmetric Amination of N-Nonsubstituted
r-Alkoxycarbonyl Amides. Optically active N-nonsubstituted
R-amino-R-alkoxycarbonyl amides constitute useful chiral build-
ing blocks for the synthesis of biologically active compounds.
Several heterocycles have been synthesized from this class of
compounds.25 Catalytic asymmetric amination of N-nonsubsti-
tuted R-alkoxycarbonyl amides 7 allows for efficient access to
this class of compounds in an enantiomerically enriched form,
however, to the best of our knowledge, there are no catalytic
asymmetric transformations utilizing N-nonsubstituted R-alkox-
ycarbonyl amides 7 as substrates.26,27 The expected difficulties
with a catalytic asymmetric reaction of 7 are: (1) a relatively
high pKa of the R-C-H proton compared with other representa-
tive active methylene compounds;28 (2) competitive deproto-
nation of the R-C-H proton and kinetically more labile amide
N-H protons;29 and (3) a multiple coordination pattern of 7
leading to poor enantioselectivity. In this context, we anticipated
that the lanthanum/amide-based ligand catalytic system de-
scribed above would be suitable for these substrates, in which
coordination to La and hydrogen bonding would cooperatively
activate the substrates and control the stereochemical course of
the reaction. Indeed, the succinimide derivative 1 contains the
substructure of R-alkoxycarbonyl amide 7, and the substandard
enantioselectivity observed in the reaction with N-Me protected
derivative 8 suggested that the catalyst would recognize the
nonprotected R-alkoxycarbonyl amide 7 including the trans-
amide N-H proton as a privileged structure (Scheme 5).
Therefore, we applied the second-generation ternary catalytic
system to the catalytic asymmetric amination of 7. Initially,
R-phenyl-R-ethoxycarbonyl amide 7a was selected as a repre-
sentative substrate and subjected to the second-generation
amination conditions with di-tert-butyl azodicarboxylate (Scheme
6). Using 10 mol % of catalyst, the desired amination product
entry
x
y
base
z
time
yieldb (%)
ee (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
5
-
-
18 h
71
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
92
0
58
79
52
70
35
89
91
80
91
-1
13
90
5
5
5
5
5
5
5
5
KOtBu
Et3N
7.5 180 min
7.5 120 min
7.5 120 min
7.5 60 min
7.5 330 min
Et2NH
tBuNH2
2,6-lutidine
H-D-Phe-OtBu 7.5 20 min
H-D-Val-OtBu 7.5 20 min
H-L-Val-OtBu
7.5 90 min
2.5 2.5 H-D-Val-OtBu 7.5 20 min
2.5
0
0
5
H-D-Val-OtBu 7.5 120 min
H-D-Val-OtBu 7.5 24 h
0.5 0.5 H-D-Val-OtBu 1.5 1 h
>99
a 1: 0.4 mmol, BocN)NBoc: 0.48 mmol. b Determined by 1H NMR
analysis with Bn2O as an internal standard.
implying that the amines are involved in the transition state.
The reaction with simple achiral amine bases (1°, 2°, 3°
alkylamines, 2,6-lutidine) completed using 2.5 mol % of catalyst,
but the enantioselectivity remained unsatisfactory (entries 3-6).
Assuming that 2 molecules of (R)-3a to La3+ are needed to
construct a favorable transition state in the first-generation
catalyst using La(OiPr)3, amines with the substructure of ligand
(R)-3a were examined. As shown in entries 7-9, the reaction
with R-amino acid tBu esters as additional amines significantly
improved the enantioselectivity. In particular, H-D-Val-OtBu
exhibited the best performance in terms of both catalytic activity
and enantioselectivity (entry 8, 91% ee).23 Intriguingly, the use
of an antipode, H-L-Val-OtBu, resulted in inferior enantiose-
lectivity compared with H-D-Val-OtBu (entry 9), indicating that
additional amines are at work under an asymmetric environment.
In this La/(R)-3a/amine ternary catalyst system, the ratio of La/
(R)-3a was reduced to 1:1 without any detrimental effect (entries
8 vs10). The attempted reaction in the absence of either (R)-3a
or La(NO3)3•6H2O gave nearly racemic product 6, (entries
11,12), verifying the significance of the ternary catalyst system
of (R)-3/La(NO3)3•6H2O/amine to exert high catalytic efficiency.
The catalyst loading can be reduced to as little as 0.5 mol %
and the reaction at 0 °C maintained high enantioselectivity (entry
13). The second-generation catalytic asymmetric amination was
successfully performed on a 100-g scale with 1 mol % of catalyst
prepared from La(NO3)3•xH2O,24 affording the desired product
in 96% yield (2 steps) and 91% ee after removal of Boc groups
under acidic conditions (Scheme 4).14b
(25) (a) Hermann, P.; Eduard, E.; Hans-Dietrich, S.; Andreas, W.; Kurt,
P. J. Heterocyclic Chem. 2000, 37, 839. (b) Lyubomir, R. D.;
Wolfgang, F.; Ivanov, I. C. Synlett 2004, 1584. (c) Li, S.; Wang, S.
J. Heterocyclic Chem. 2008, 45, 1875. (d) Bambang, P.; Robert, S. K.;
Thomas, P. C. Synlett 1992, 231.
(26) For catalytic asymmetric fluorination of N-unsubstituted R-tert-
butoxycarbonyl lactams, see: (a) Suzuki, T.; Oto, T.; Hamashima, Y.;
Sodeoka, M. J. Org. Chem. 2007, 72, 246. For a partially successful
1,4-reduction of unsaturated primary amide and the following enan-
tioselective protonation by asymmetric catalysts (1 example, 50% ee),
see: (b) Ohtsuka, Y.; Ikeno, T.; Yamada, T. Tetrahedron: Asymmetry
2003, 14, 967. For a review of enzymatic kinetic resolution of primary
amides, see: (c) Wang, M.-X Top. Catal. 2005, 35, 117, and references
cited therein.
(27) Diastereoselective reactions using N-nonsubstituted R-alkoxycarbonyl
amides as substrates, see: Kozlowski, M. C.; DiVirgilio, E. S.;
Malolanarasimhan, K.; Mulrooney, C. A. Tetrahedron: Asymmetry
2005, 16, 3599. Catalytic transformation utilizing N-nonsubstituted
R-alkoxycarbonyl amides, see: Zhang, J.; Sarma, K. D.; Curran, T. T.;
Belmont, D. T.; Davidson, J. G. J. Org. Chem. 2005, 70, 5890.
(28) pKa of R-C-H proton of 7a is 21.3, which was calculated at the B3LYP/
6-31G+(d,p) level. For details, see Supporting Information.
(29) pKa of cis- and trans-amide N-H protons of 7a is 24.0, and 27.8,
respectively, which was calculated at the B3LYP/6-31G+(d,p) level.
For details, see supporting information. Whereas N-H protons are
less acidic thermodynamically, they are expected to be kinetically more
labile, compromising effective deprotonation of R-C-H proton.
(22) Selected examples of rare earth metal asymmetric catalyst showing
enhanced stereoselectivity in the presence of amine, see: (a) Kobayashi,
S.; Ishitani, H. J. Am. Chem. Soc. 1994, 116, 4083. (b) Kobayashi,
S.; Ishitani, H.; Hachiya, I.; Araki, M. Tetrahedron 1994, 50, 11623.
(c) Kobayashi, S.; Kawamura, M. J. Am. Chem. Soc. 1998, 120, 5840.
(d) Furuno, H.; Hanamoto, T.; Sugimoto, Y.; Inanaga, J. Org. Lett.
2000, 2, 49. (e) Fukuzawa, S.-I.; Komuro, Y.; Nakanao, N.; Obata, S.
Tetrahedron Lett. 2003, 44, 3671. See also ref 16a.
(23) When LaBr3 was used instead of La(NO3)3•6H2O, the product (-)-6
was obtained in 97% yield and 72% ee after 13 h at 0 °C.
(24) La(NO3)3•xH2O (99.9% purity, $186.5/500 g, Aldrich (ref 19)) is less
expensive than La(NO3)3•6H2O.
9
J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009 14993

A R T I C L E S
Mashiko et al.
Scheme 5. Expected Privileged Structure for Catalytic Asymmetric
Amination with La(NO3)3/(R)-3a/base Ternary Catalyst
Scheme 7. Transformation of the Amination Product (+)-12 to
Hydantoin (-)-16a
a Reagents and conditions: (a) HCl (g), toluene, 0 °C, 1 h, 98%; (b)
Raney-Ni, H2, EtOH/AcOH ) 3/2, rt, 3 h, >99%; (c) (i) p-nitrophenyl
chloroformate, CH3CN, NaHCO3, rt, 1 h; (ii) H2O, rt, 1.5 h, 83%.
Scheme 6. Catalytic Asymmetric Amination of Nonprotected
R-Ethoxycarbonyl Amide 7a and Its N-Alkylated Derivatives 10,11
probably because 7d was reluctant to form a planar enolate due
to steric or electrostatic repulsion of 2-fluoro substituent (entry
4). Notably, even a substrate with a strongly coordinative pyridyl
group exhibited high enantioselectivity (entry 11). This would
be attributed to the preferential coordination of 7k to the catalyst
through the privileged substructure shown in Scheme 5. The
enantioselectivity of the asymmetric amination of R-alkyl
substrates 7l was not as high as that of aromatic and heteroaro-
matic substrates, providing the product 12l in 98% yield and
83% ee (entry 14). The obtained amination product (+)-12a
was successfully converted to a hydantoin derivative (-)-16
(Scheme 7).30 The compound (+)-12a (>96% ee) was treated
with concentrated hydrochloric acid to remove the Boc group,
followed by N-N bond cleavage by Raney nickel to afford the
R, R-disubstituted R-amino acid derivative (+)-15 in 98% yield
in 2 steps. Subsequent reaction with p-nitophenyl chloroformate
buffered with NaHCO3 afforded hydantoin derivative (-)-16
in 83% yield,31 the absolute configuration of which was
determined by X-ray crystallographic analysis.32 The present
protocol provides an efficient new entry to access enantio-
enriched R,R-disubstituted amino acid derivatives and hydan-
toins bearing a stereogenic tetrasubstituted carbon, which are
important structural motifs in medicinal chemistry.33
Table 3. Catalytic Asymmetric Amination of Nonprotected
R-Alkoxycarbonyl Amides 7 with the Second-Generation Catalysta
R-alkoxycarbonyl amide 7
entry
R1
R2
product
time (h)
yieldb (%)
ee (%)
1
2
3
4
5
6
7
8
9
Ph
Ph
4-FC6H4
2-FC6H4
4-CF3C6H4
4-NO2C6H4
Et
7a
12a
12b
12c
12d
12e
12f
12g
12h
12i
48
24
14
48
14
10
48
12
18
15
12
45
83
92
99
45
94
>99
88
92
>99
>99
>99
94
>99
99
99
99
>99
>99
>99
83
Me 7b
Me 7c
Et
Me 7e
Et
7d
7f
7g
7h
7i
7j
7k
7l
4-MeOC6H4 Et
4-MeC6H4
2-naphthyl
3-thienyl
3-pyridyl
Me
Et
Et
Et
Et
Et
Mechanistic Studies of the La(NO3)3/(R)-3a/H-D-Val-OtBu
Ternary Amination Catalyst. The utility of the La(NO3)3/(R)-
3a/H-D-Val-OtBu ternary catalytic system for highly coordina-
tive substrates led us to examine the mechanistic details. ESI
MS spectra of the first-generation catalyst (La(OiPr)3/3a ) 1/2
>99
>99
70
10
11
12
12j
12k
12l
98
a 7: 0.2 mmol, BocN dNBoc: 0.24 mmol. b Isolated yield.
12a bearing a tetrasubstituted stereogenic carbon was obtained
with nearly perfect enantioselectivity, although the reaction was
sluggish, likely due to the higher pKa of the R-C-H proton
than that of 1. In contrast to 7a, the amination reaction hardly
proceeded with trans-N-Me 10 or the N,N-dimethyl derivative
11, indicating that hydrogen bonding through trans-N-H of
7a was crucial for the recognition/activation of substrates
(Scheme 5). The substrate scope of the catalytic asymmetric
amination of 7 is summarized in Table 3. Methyl ester analog
7b exhibited higher reactivity to afford 12b in 99% yield and
>99% ee after 24 h (entry 2). Generally excellent enantiose-
lectivity was observed and the electronic properties of the
aromatic ring at the R-position were responsible for the reaction
rate. Substrates bearing an electron-withdrawing group resulted
in a rapid reaction (entries 3,5,6), whereas the reaction proceeded
sluggishly with 7g, which has an electron-donating methoxy
group (entry 7). 7d showed unexpectedly low reactivity,
(30) For the rich chemistry and property of hydantoins, see: (a) Ware, E.
Chem. ReV. 1950, 46, 403. For a leading examples of the racemic
synthesis of hydantoins, see: (b) Rizzi, J. P.; Schnur, R. C.; Hutson,
N. J.; Kraus, K. G.; Kelbaugh, P. R. J. Med. Chem. 1989, 32, 1208.
(c) Kim, D.; et al. Bioorg. Med. Chem. Lett. 2001, 11, 3099. (d)
Sheppeck, J. E.; Gilmore, J. L.; Tebben, A.; Xue, C.-B.; Liu, R.-Q.;
Decicco, C. P.; Duan, J. J.-W. Bioorg. Med. Chem. Lett. 2007, 17,
2769. (e) Zhao, B.; Du, H.; Shi, Y. J. Am. Chem. Soc. 2008, 130,
7220.
(31) Yamaguchi, J.; Harada, M.; Kondo, T.; Noda, T.; Suyama, T. Chem.
Lett. 2003, 32, 372.
(32) For details, see Supporting Information.
(33) (a) Comber, R. N.; Reynolds, R. C.; Friedrich, J. D.; Manguikian,
R. A.; Buckheit, R. D. Jr.; Truss, J. W.; Shannon, W. M.; Secrist,
J. A. J. Med. Chem. 1992, 35, 3567. (b) Brouillette, W. J.; Jestkov,
V. P.; Brown, M. L.; Akhtar, M. S.; DeLorey, T. M.; Brown, G. B.
J. Med. Chem. 1994, 37, 3289. (c) Stilz, H. U.; Guba, W.; Jablonka,
B.; Just, M.; Klingler, O.; Ko¨nig, W.; Wehner, V.; Zoller, G. J. Med.
Chem. 2001, 44, 1158. (d) Muccioli, G. G.; Fazio, N.; Scriba, G. K. E.;
Poppitz, W.; Cannata, F.; Poupaert, J. H.; Wouters, J.; Lambert, D. M.
J. Med. Chem. 2006, 49, 417.
9
14994 J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009

Highly Coordinative Substrates in Asymmetric Catalysis
A R T I C L E S
Figure 1. ESI MS analysis of La catalyst solution in positive ion mode. (a) La(OiPr)3/3a ) 1/2; (b) La(OiPr)3/3a ) 1/2 with DMA (N,N-dimethylacetamide);
(c) La(NO3)3•6H2O/(R)-3a/H-D-Val-OtBu ) 1/1/3.
with or without DMA) and the second-generation catalyst
(La(NO3)3/(R)-3a/H-D-Val-OtBu ) 1/1/3) are shown in Figure
1. In contrast to the oligomeric La/3a complexes observed in
the MS spectrum of the first-generation catalyst in the absence
of DMA (Figure 1a), a peak derived from monomeric lanthanum
species became dominant by the addition of DMA to the catalyst
solution (Figure 1b). Based on the higher catalytic efficiency
and reproducibility in the presence of DMA,14a we assume that
the monometallic lanthanum species was involved in the desired
enantioselective pathway. This assumption was further supported
by the MS analysis of the second-generation catalyst, which
displayed generally higher catalytic activity and reproducibility
compared to the first-generation catalyst. In the MS spectrum
of the second-generation catalyst solution prepared from La(NO3)3/
(R)-3a/H-D-Val-OtBu ) 1/1/3 (Figure 1c), only monometallic
lanthanum complexes appeared, probably because highly co-
ordinative and bidentate nitrate anions circumvented the forma-
tion of unfavorable multimetallic aggregates. Together with
partially associated La complexes, the peaks derived from a
ternary complex of [La(NO3)3/(R)-3a/H-D-Val-OtBu]+ (m/z )
638) and [La(NO3)3/(R)-3a/H-D-Val-OtBu/AcOEt]+ (m/z ) 726)
appeared, supporting the notion that the additional amine base
H-D-Val-OtBu coordinates to lanthanum and is involved in the
transition state.34 This observation is consistent with the fact
that the enantioselectivity of the amination was dependent on
the three-dimensional structure of the additional amine bases
(Table 2). In the MS analysis of the first-generation catalyst
with DMA and the second-generation catalyst, we observed no
peaks derived from a multimetallic lanthanum complex. Pre-
venting undesired lanthanum aggregation would be key to the
high catalyst turnover and high reproducibility.
To gain further insight into the actual active species, we
conducted several 1H NMR experiments. The 1H NMR spectrum
of La(NO3)3/(R)-3a ) 1:1 in THF-d8 was essentially identical
to that of (R)-3a itself, indicating that no appreciable association
occurred between neutral La(NO3)3 and (R)-3a in the absence
of a base.35 Indeed, any attempts of crystallization from
La(NO3)3/(R)-3a ) 1:1 in AcOEt resulted in the recovery of
the crystal of La(NO3)3•xH2O. In the presence of H-D-Val-OtBu
(3 equivalents to La), substantial peak broadening was observed
in 1H NMR spectrum of the ternary catalyst solution, indicating
that complexation occurred in the presence of additional base
and that the assembly of La3+, (R)-3a, and H-D-Val-OtBu were
in equilibrium within NMR time scale.32 The negligible change
in the CD spectrum of the ternary catalyst solution from that of
(R)-3a itself suggested that any possible complexation was in
equilibrium and the dissociated state was dominant.32 Any
(35) Only the peaks derived from [(R)-3a + H+]+ and [((R)-3a)2 + H+]+
appeared predominantly in ESI MS spectrum of La(NO3)3/(R)-3a )
1:1 solution. Catalytic asymmetric amination of 7a proceeded in THF
solvent to give the product 12a in comparable yield and enantiose-
lectivity (10 mol % of catalyst, 0 °C, 45 h, 91% yield, 96% ee). For
details of NMR and ESI MS analysis, see Supporting Information.
(34) The peaks derived from the ternary complex were observed even in
the presence of substrate 7c. See Supporting Information.
9
J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009 14995

A R T I C L E S
Mashiko et al.
Table 4. Reexamination of Base Effect in the Second-generation
Catalytic Asymmetric Amination of 7ca
the enantioselection as an additional chiral entity of the ternary
catalyst. Remarkably, the reaction rate and enantioselectivity
recovered with mixed-base conditions (H-D-Val-OtBu/Et3N,
H-D-Val-OtBu/Cs2CO3), exhibiting a comparable reaction profile
and enantioselectivity as that obtained under the standard
conditions using H-D-Val-OtBu as the sole base (Table 4, entries
1 vs 4, 1 vs 5, Figure 2, open squares and open circles). This
observation further supported the above assumption and implied
a dual role of H-D-Val-OtBu; (1) the deprotonation of (R)-3a
to facilitate the association of La3+ and (R)-3a; and (2) the
coordination to La3+ as an additional ligand working coopera-
tively with La3+ to recognize/activate substrates (Scheme 8a).38
In the mixed-base system, Et3N or Cs2CO3 worked solely to
deprotonate (R)-3a to shift the equilibrium to form a greater
fraction of the La/(R)-3a complex than that formed in the
standard conditions (La(NO3)3/(R)-3a/H-D-Val-OtBu ) 1/1/3);39
subsequently H-D-Val-OtBu coordinated to La/(R)-3a to form
a ternary complex, enhancing the reaction rate and enantiose-
lection (Scheme 8b).40 Taking advantage of the preferential
coordination of H-D-Val-OtBu, the catalyst loading was suc-
cessfully reduced to as little as 0.5 mol % with the aid of
additional base (Table 5). This mixed-base catalytic system
highlighted the fact that neither Et3N nor the La/(R)-3a complex
worked as an efficient catalyst and the ternary catalyst including
H-D-Val-OtBu was involved in the highly enantioselective
reaction pathway. Et3N worked solely to deprotonate (R)-3a,
thereby increasing the molar fraction of La/(R)-3a, leading to a
greater molar fraction of the ternary complex in combination
with H-D-Val-OtBu.40 The mixed-base conditions were ap-
plicable to several R-alkoxycarbonyl amides 7, affording product
12 in excellent yield and comparable enantioselectivity with
0.5-3 mol % of catalyst loading.41
base (y mol %)
vobs
(mM min-1
time yieldc ee
entry
x
H-D-Val-OtBu Et3N Cs2CO3
keyb
)
(h)
(%)
(%)
1
2
3
4
5
4
4
4
4
4
12
-
-
4
-
12
-
8
-
-
12
-
8
1.16
0.24
0.10
1.35
1.27
b
9
2
0
O
10
10
10
85 96
75 90
53 88
10 >99 98
10 >99 98
4
-
a 7c: 0.4 mmol, BocNdNBoc: 0.48 mmol. b Key in Figure 2.
Determined by H NMR with Bn2O as an internal standard.
c
1
attempts to crystallize the second-generation catalyst mixture
La(NO3)3/(R)-3a/H-D-Val-OtBu ) 1:1:3 gave rise to the exclu-
sive crystallization of the H-D-Val-OtBu•HNO3 salt.36
On the basis of the NMR and ESI MS analyses, the additional
amine base would deprotonate the phenolic proton of (R)-3a,
facilitating the complexation of La(NO3)3 and (R)-3a in equi-
librium. The correlation of the observed enantioselectivity of
the amination and the three-dimensional structure of the amine
bases (Table 2) clearly indicated that the function of additional
base is not only deprotonation, but also the enhancement of
enantioselection through participation in the transition state.
Although the catalyst components were in equilibrium between
the associated and dissociated states, a La(NO3)3/(R)-3a/H-D-
Val-OtBu ternary complex, which was indeed observed in MS
analysis in Figure 1c, was anticipated to be catalytically much
more active than the dissociated or partially complexed state.
To examine this assumption, a kinetic profile of the asymmetric
amination of R-alkoxycarbonyl amides 7c with amine bases (H-
D-Val-OtBu or Et3N), inorganic base (Cs2CO3), and mixed-base
(H-D-Val-OtBu/Et3N or H-D-Val-OtBu/Cs2CO3) conditions was
developed (Table 4, Figure 2).37 The initial rate of the reaction
with H-D-Val-OtBu (V ) 1.16 mM min-1, solid circles) was
The asymmetric amination was then performed with partially
resolved ligand (R)-3a. NEt3 was used as an additional base
instead of H-D-Val-OtBu because the combination of (S)-3a and
H-D-Val-OtBu is a mismatched pair (Table 1, entry 9) and the
interpretation of the data was complicated. The reactions were
conducted with 2.5 mol % of the second-generation catalyst
and completed after 90 min of stirring at room temperature. As
(38) As La/H-D-Val-OtBu complex was observed in ESI MS (Figure 1c),
an alternative pathway to form the ternary complex through the
coordination of H-D-Val-OtBu to La3+ to give La/H-D-Val-OtBu and
subsequent coordination of (R)-3a is also feasible. The catalytic
asymmetric amination reaction of 7c in the absence of (R)-3a was
very sluggish (room temperature, 25 h, 18% yield) and the enanti-
oselectivity of the product was poor (4% ee), indicating that La/H-
D-Val-OtBu was as poor a catalyst as La/(R)-3a.
(39) The mixed-base conditions exhibited higher reaction rate as shown in
Table 4, entries 1, 4, and 5. This observation is likely explained by
assuming that H-D-Val-OtBu was less effective for the deprotonation
of (R)-3a due to extensive coordination to La3+ as shown in ESI MS
spectrum of La(NO3)3/(R)-3a/H-D-Val-OtBu ) 1/1/3 (Figure 1c).
(40) In ESI MS spectrum of the mixed-base catalyst solution (La/(R)-3a/
H-D-Val-OtBu/Et3N ) 1/1/1/2), the peaks derived from the ternary
complexes including La, (R)-3a, and H-D-Val-OtBu were observed,
whereas no peaks derived from the complexes containing Et3N were
identified. See Supporting Information for details.
substantially faster than that with Et3N (V ) 0.24 mM min-1
,
solid triangles) or Cs2CO3 (V ) 0.10 mM min-1, solid squares).
Moreover, the enantioselectivity in the reaction with H-D-Val-
OtBu was higher than that with Et3N or Cs2CO3 (Table 4, entries
1-3, Figure 2). This would be due to the fact that amine bases
coordinated to the La/(R)-3a complex to form the ternary
complex in equilibrium, which recognized/activated the sub-
strates by harnessing a cooperative noncovalent interaction of
coordination to La3+ and hydrogen bonding, leading to a higher
reaction rate and enantioselectivity. The advantage of H-D-Val-
OtBu over Et3N for the reaction rate would be ascribed to smaller
steric factor and possible bidentate coordination through an ester
carbonyl group, facilitating the coordination of H-D-Val-OtBu
to La3+ to form the ternary catalyst in equilibrium, a tentative
actual active catalyst. The coordinated H-D-Val-OtBu reinforced
(41) In the standard reaction conditions using H-D-Val-OtBu as the sole
base, 3 mol equiv of H-D-Val-OtBu was required for high catalytic
efficiency and enantioselectivity, likely because 2 mol equiv of H-D-
Val-OtBu was consumed for the deprotonation of (R)-3a (although
protonation/deprotonation was in equilibrium) and the apparent amount
of H-D-Val-OtBu was 1 mol equiv to La(NO3)3. In the mixed-base
conditions, 3 mol equiv of Et3N was expected to deprotonate (R)-3a,
thereby 1 mol equiv of H-D-Val-OtBu appeared to be sufficient. Three
mol equiv of H-D-Val-OtBu, however, were still required to assure
the high catalytic efficiency and enantioselectivity. Thus, 3 mol equiv
of H-D-Val-OtBu to La(NO3)3 were employed to form the ternary
complex irrespective of its protonation.
(36) The H-D-ValOtBu deprotonated (R)-3a to form H-D-Val-OtBu•HNO3
salt. Although the deprotonation/protonation would be in equilibrium,
H-D-Val-OtBu•HNO3 was highly crystalline and preferencially crystal-
lized.
(37) A kinetic study using substrate 1 provided unreliable data due to a
background reaction under basic or acidic conditions at the work-up
stage. The data on the kinetic study of 7c is detailed in the Supporting
Information.
9
14996 J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009

Highly Coordinative Substrates in Asymmetric Catalysis
A R T I C L E S
Figure 2. Kinetic profiles of the asymmetric amination of 7c promoted by La(NO3)3•6H2O/(R)-3a/base ) 1/1/3 catalyst. The reactions were run with [7c]
) 100 mM, [BocN)NBoc] ) 120 mM, [La(NO3)3•6H2O] ) 2.5 mM, [(R)-3a] ) 2.5 mM, and [base] ) 7.5 mM (4 mol % of catalyst loading based on La).
b: base ) H-D-Val-OtBu (corresponds to Table 4, entry 1), 9: base ) Et3N (entry 2), 2: base ) Cs2CO3 (entry 3), 0: base ) H-D-Val-OtBu/Et3N (1/2)
(entry 4), O: base ) H-D-Val-OtBu/Cs2CO3 (1/2) (entry 5).
Scheme 8. Proposed Dual Role of Additional Amine Base in the
Second-Generation Amination Reaction
Table 5. Second-Generation Catalytic Asymmetric Amination of 7
with H-D-Val-OtBu/Et3N Mixed-Base Systema
R-alkoxycarbonyl amide 7
temp
(°C)
time yieldb
ee
(%)
entry
R1
R2
x
product
(h)
(%)
1
Ph
4-FC6H4
4-FC6H4
4-CF3C6H4 Et
4-NO2C6H4 Et
3-thienyl
Et
Me 7c
Me 7c 0.5
7a
3
1
rt
rt
rt
rt
0
12a
12c
12c
12e
12f
12j
48
99
98
98
93
96
>99
93
2c
3d
4
18 >99
48
13
13 >99
13 >99
92e
90
7e
7f
7j
1
3
3
5
6
Et
rt
a 7: 0.2 mmol, BocNdNBoc: 0.24 mmol. b Isolated yield. c 0.4 mmol,
BocNdNBoc: 0.48 mmol. d 0.8 mmol, BocNdNBoc: 0.96 mmol.
e
1
Determined by H NMR with Bn2O as an internal standard.
group features of amide-based ligand 3a were anticipated to be
of prime importance for the association to form the ternary
complex that engaged in the reaction. To probe this point, ester
analogs of the ligand 3a were synthesized and evaluated in the
second-generation amination reaction of 1. Both catechol analog
(S)-17 and salicyl ester analog (S)-18 promoted the amination
reaction of 1, albeit with poor enantioselectivity, indicating that
the amide functionality of (R)-3a was crucial for the formation
of the ternary complex and/or recognition of 1. Stronger
coordination of a more Lewis basic amide carbonyl to La3+ and/
or hydrogen bonding of the N-H proton would be assumed to
explain the preference of 3a over 17 and 18 (Scheme 9). On
the basis of the experimental results obtained above, a plausible
catalytic cycle of the second-generation amination reaction is
outlined in Figure 4. NMR and CD spectral analyses suggested
that the catalyst components, La(NO3)3•6H2O, (R)-3a, and H-D-
Val-OtBu in a 1:1:3 ratio, were in equilibrium between the
associated and dissociated states in the reaction mixture with
dissociated form predominating. ESI MS analysis and the linear
relationship between enantiopurity of the catalyst and product
enantioselectivity were in accord with the assumption that the
catalytic asymmetric amination likely involves a monomeric
shown in Figure 3, a linear relationship was observed between
the enantiopurity of (R)-3a and the enantioselectivity of (-)-
6.42 Subsequent kinetic studies of the second-generation catalyst
using 7c as the substrate indicated that the reaction was
dependent on the initial concentration of the catalyst compo-
nents. From these observations, the most likely scenario is that
the coordination of (R)-3a to La3+ was induced by a base and
the association was in equilibrium. The coordination of H-D-
Val-OtBu to the monomeric La/(R)-3a complex furnished the
ternary complex, which allowed the enantioselective pathway
to proceed, likely through the monomeric lanthanum complex
furnished with (R)-3a and H-D-Val-OtBu.38
Because the catalyst components were in equilibrium between
the associated and dissociated states, the structural and functional
(42) Girard, C.; Kagan, H. B. Angew Chem., Int. Ed. 1998, 37, 2922. (b)
Blackmond, D. G. Acc. Chem. Res. 2000, 33, 402. (c) Satyanarayana,
T.; Abraham, S.; Kagan, H. B. Angew. Chem., Int. Ed. 2009, 48, 456.
9
J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009 14997

A R T I C L E S
Mashiko et al.
Figure 3. Plot of % ee of product (-)-6 versus % ee of ligand (R)-3a in
the second-generation catalytic asymmetric amination using Et3N as a base.
Scheme 9. Second-generation Catalytic Asymmetric Amination of
1 Using Ester Ligand (S)-17 or (S)-18
Figure 4. Proposed catalytic cycle for second-generation catalytic asym-
metric amination through ternary complex.
is intriguing that a metallic catalyst complex was in equilibrium
between associated and dissociated forms while it exhibited an
excellent enantioselection. Association of catalyst components
and highly coordinative substrates through cooperative metal-
coordination and hydrogen bonding enabled the employment
of highly coordinative substrates in asymmetric catalysis.
Conclusion
In summary, a catalytic asymmetric amination of nonprotected
succinimide derivative 1 and R-alkoxycarbonyl amides 7 was
developed with a lanthanum/amide-based ligand catalyst. These
two classes of substrates share the same privileged structure
likely recognized by the present catalytic system. The absence
of precedents in the literature using these highly coordinative
substrates in enantioselective catalysis highlights the utility of
the lanthanum/amide-based ligand catalyst. Mechanistic studies
using MS, NMR, CD analyses, and Eyring plot as well as several
control and kinetic experiments suggested that the catalyst
components were in equilibrium between the associated and
dissociated state, and a ternary complex of La/(R)-3a/H-D-Val-
OtBu was likely involved. Cooperative coordination to La3+ and
hydrogen-bonding interactions were key to constructing the
ternary complex and specific recognition/activation of highly
coordinative nonprotected substrates 1 and 7, affording the
amination product in a highly enantioselective manner.
lanthanum complex such as A. The beneficial effects of
salicylamide carbonyl over salicyl ester in the ligand substructure
(Scheme 9) might be due to the higher Lewis basicity for
coordination to La3+ in a bidentate fashion. Subsequently, 7
was deprotonated by the complex A or H-D-Val-OtBu, and the
following coordination to La3+ together with di-tert-butyl
azodicarboxylate led to the transition structure B. Hydrogen
bonding between (R)-3a and an enolate generated from 7, amide
NH (aminophenol side)/enolate oxygen, and the phenolic
oxygen/imide N-H, may assist in determining the enantiofacial
selection of the enolate.43 The absolute stereochemistry of H-D-
Val-OtBu located in near proximity of La3+ would contribute
to direct azodicarboxylate approaching from Re-face of the
enolate. An Eyring plot of the reaction provided large negative
entropy of activation ∆Sq ) -2.68 × 102 J mol-1 K-1, further
supporting the notion that the association of catalyst components
to form the ternary catalyst and an ordered transition state would
be operative.44 The nucleophilic addition from B followed by
the protonation via H+(H-D-Val-OtBu)NO3- afforded the desired
product 12 and regenerated the monomeric lanthanum complex
A, which was again in equilibrium with the dissociated state. It
Acknowledgment. Financially support was provided by Grant-
in-Aid for Scientific Research (S) (for M. S.) and Grant-in-Aid for
Scientific Research on Innovative Areas (for N.K.) from JSPS and
(44) (a) Eyring, H. J. Chem. Phys. 1935, 3, 107. (b) Evans, M. G.; Polanyi,
M. Trans. Faraday Soc. 1935, 31, 875. Usage of Eyring plot in
asymmetric catalysis, see: (c) Josephsohn, N. S.; Kuntz, K. W.;
Snapper, M. L.; Hoveyda, A. H. J. Am. Chem. Soc. 2001, 123, 11594.
For experimental details, see Supporting Information.
(43) Bidentate coordination through two phenolic oxygens of (R)-3a was
topologically unlikely.
9
14998 J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009

Highly Coordinative Substrates in Asymmetric Catalysis
A R T I C L E S
MEXT. T.M. thanks JSPS for predoctoral fellowship. Drs. S.
Furusho and A. Sato at JASCO international, and Dr. K. Konuma
at JEOL Co. Ltd. are gratefully acknowledged for technical
assistance in ESI MS analysis. We are grateful to Dr. M. Shiro at
Rigaku Corporation for X-ray crystallographic analysis of com-
pound (-)-16. We thank Prof. T. Ohwada and Dr. S. Uchiyama
for assistance with the measurement of CD spectra of the catalysts.
Supporting Information Available: Experimental details and
characterization of new compounds. Complete ref 30c. This
material is available free of charge via the Internet at http://
pubs.acs.org.
JA9052653
9
J. AM. CHEM. SOC. VOL. 131, NO. 41, 2009 14999