Mechanism of palladium complex-catalyzed enantioselective Mannich-type reaction: Characterization of a novel binuclear palladium enolate complex

Article abstract of DOI:10.1021/ja9902827

Studies on the enantioselective addition of enol silyl ethers to imines catalyzed by optically active palladium diaquo complexes 3 or binuclear palladium μ-hydroxo complex 4 are described, with particular focus on the mechanistic aspects. Asymmetric induction in the reaction using [Pd((R)-binap)(H₂O)₂]²⁺(BP₄ ^-)₂ (3a) was quite sensitive to the reaction conditions, suggesting unfavorable effects of HBF₄ generated from 3a in situ. Novel optically active binuclear μ-hydroxo complexes [{Pd((R)-binap)(μ-OH)}₂]²⁺(BF₄ ^-)₂ (4a), [{Pd-((R)-tol-binap)(μ-OH)}₂]²⁺(BF₄ ^-)₂ (4b), [{Pd((R)-binap)(μ-OH)}₂]²⁺(TfO^-)₂ (4c), and [{Pd((R)-tol-binap)(μ-OH)}₂]²⁺(TfO^-) ₂ (4d) were prepared and were found to be better catalysts for the asymmetric Mannich-type reaction. Benzoylalanine derivatives 5 were obtained in excellent chemical and optical yields (up to 90% ee). Mechanistic studies using ¹H NMR and electrospray ionization mass spectrometry indicated that a unique binuclear palladium-sandwiched enolate 12 was involved in the reaction of enol silyl ether 1 with imine 2 catalyzed by 4.

Full text of DOI:10.1021/ja9902827

5450
J. Am. Chem. Soc. 1999, 121, 5450-5458
Mechanism of Palladium Complex-Catalyzed Enantioselective
Mannich-Type Reaction: Characterization of A Novel Binuclear
Palladium Enolate Complex
Akio Fujii, Emiko Hagiwara, and Mikiko Sodeoka*
Contribution from the Sagami Chemical Research Center, Nishi-Ohnuma, Sagamihara,
Kanagawa 229-0012, Japan
ReceiVed January 28, 1999
Abstract: Studies on the enantioselective addition of enol silyl ethers to imines catalyzed by optically active
palladium diaquo complexes 3 or binuclear palladium µ-hydroxo complex 4 are described, with particular
focus on the mechanistic aspects. Asymmetric induction in the reaction using [Pd((R)-binap)(H2O)2]2+(BF4
)
-
2
(3a) was quite sensitive to the reaction conditions, suggesting unfavorable effects of HBF4 generated from 3a
in situ. Novel optically active binuclear µ-hydroxo complexes [{Pd((R)-binap)(µ-OH)}2]2+(BF4-)2 (4a), [{Pd-
((R)-tol-binap)(µ-OH)}2]2+(BF4-)2 (4b), [{Pd((R)-binap)(µ-OH)}2]2+(TfO-)2 (4c), and [{Pd((R)-tol-binap)(µ-
OH)}2]2+(TfO-)2 (4d) were prepared and were found to be better catalysts for the asymmetric Mannich-type
reaction. Benzoylalanine derivatives 5 were obtained in excellent chemical and optical yields (up to 90% ee).
1
Mechanistic studies using H NMR and electrospray ionization mass spectrometry indicated that a unique
binuclear palladium-sandwiched enolate 12 was involved in the reaction of enol silyl ether 1 with imine 2
catalyzed by 4.
Introduction
Scheme 1
Carbon-carbon bond-forming reactions that involve the
addition of resonance-stabilized nucleophiles, such as enols and
enolates, to iminium salts and imines, the so-called Mannich-
type reactions, comprise one of the most important classes of
reactions in organic synthesis.1 A number of methods for the
diastereoselective reaction of imines with enolates of carboxylic
acid derivatives or silyl ketene acetals have been reported, but
examples of enantioselective variants of this type of reaction
are quite limited.2 Recently, we reported the first example of a
Pd(II)-catalyzed enantioselective addition of enol silyl ethers
to imines.3 Reaction of various enol silyl ethers 1 with
iminoacetic acid derivatives 2 using Pd(II) complex 3 or 4 as a
catalyst proceeded smoothly to give optically active acylalanine
derivatives 5 in good chemical and optical yields (Scheme 1).
The optically active 5 made available by this novel reaction
has various potential applications. Substituted or unsubstituted
benzoylalanines obtained by simple deprotection of 5 are known
to be potent inhibitors of kynurenine 3-hydroxylase and
kynureninase, potential drugs for the treatment of neurodegen-
erative disorders that function by controlling the toxic quinolinic
acid concentration in the brain.4 Since the carbonyl group in 5
would allow various modificationsssuch as reduction, alkyla-
tion, alkenylation, ketal formation, halogenation, etc.sthe
acylalanine derivatives 5 are also expected to be important
synthetic intermediates for a wide variety of nonnatural amino
acids. N-Arylated amino acid derivatives themselves would be
quite attractive building blocks for the combinatorial synthesis,
because N-arylation of the amide is expected to cause large
conformational changes in peptide-type molecules. Despite the
recent development of N-arylation methods,5 however, arylation
of the amino groups in optically active functionalized amino
acids without racemization is still a problematic and difficult
(1) For reviews, see: (a) Kleinman, E. F. In ComprehensiVe Organic
Synthesis; Trost, B. M., Fleming, I., Heathcock, C. H., Eds.; Pergamon
Press: New York, 1991; Vol. 2, pp 893-951. (b) Hart, D. J.; Ha, D.-C.
Chem. ReV. 1989, 89, 1447-1465. (c) Arend, M.; Westermann, B.; Risch,
N. Angew. Chem., Int. Ed. 1998, 37, 1044-1070 and references therein.
(2) For reactions using a stoichiometric amount of chiral source, see:
(a) Corey, E. J.; Decicco, C. P.; Newbold, R. C. Tetrahedron Lett. 1991,
32, 5287-5290. (b) Ishihara, K.; Miyata, M.; Hattori, K.; Tada, T.;
Yamamoto, H. J. Am. Chem. Soc. 1994, 116, 10520-10524. (c) Kambara,
T.; Hussein, A.; Fujieda, H.; Iida, A.; Tomioka, K. Teterahedron Lett. 1998,
39, 9055-9058. For reactions using a catalytic amount of chiral source,
see: (d) Fujieda, H.; Kanai, M.; Kambara, T.; Iida, A.; Tomioka, K. J. Am.
Chem. Soc. 1997, 119, 2060-2061. (e) Ishitani, H.; Ueno, M.; Kobayashi,
S. J. Am. Chem. Soc. 1997, 119, 7153-7154. (f) Kobayashi, S.; Ishitani,
H.; Ueno, M. J. Am. Chem. Soc. 1998, 120, 431-432. (g) Yamasaki, S.;
Iida, T.; Shibasaki, M. Tetrahedron Lett. 1999, 40, 307-310.
(4) (a) Giordani, A.; Corti, L.; Cini, M.; Brometti, R.; Marconi, M.;
Veneroni, O.; Speciale, C.; Varasi, M. Recent AdVances in Tryptophan
Research; Plenum Press: New York, 1996; pp 531-534. (b) Giordani, A.;
Pevarello, P.; Cini, M.; Bormetti, R.; Greco, F.; Toma, S.; Speciale, C.;
Varaqsi, M. Bioorg. Med. Chem. Lett. 1998, 8, 2907-2912 and references
therein.
(3) Hagiwara, E.; Fujii, A.; Sodeoka, M. J. Am. Chem. Soc. 1998, 120,
2474-2475.
10.1021/ja9902827 CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/29/1999

A NoVel Binuclear Palladium Enolate Complex
J. Am. Chem. Soc., Vol. 121, No. 23, 1999 5451
process. Thus, the optically active N-aryl-amino acid derivatives
made available by this novel Mannich-type reaction would offer
a good source of diversity for use in a combinatorial library.
After our preliminary report,3 Lectka et al. reported a reaction
of 1 with 2 (R2 ) Et, R3 ) tosyl) using a chiral dicationic
palladium complex, [Pd((R)-binap)(CH3CN)2]2+(ClO4-)2, as a
Lewis acid catalyst.6 Although their results seem similar to ours,
the mechanism of our reaction is quite distinct from the well-
known Lewis acid-catalyzed one. In this paper we describe the
full details of our studies on the Pd(II) complex-catalyzed
enantioselective addition of enol silyl ethers to iminoacetic acid
derivatives, with a particular focus on the reaction mechanism.
Results and Discussion
Pd Aquo Complex-Catalyzed Reaction. That the ketone-
imine addition is relatively more difficult to promote than the
ester/thioester-imine addition may be due to the relatively
higher reactivity of ketones, which may cause undesired side
reactions. Thus, our strategy involved the use of a less
nucleophilic transition metal enolate in place of the highly
nucleophilic metal enolate and/or strong Lewis acid catalyst.
Because we had already reported the catalytic asymmetric
addition of enol silyl ethers to aldehydes with the chiral
Figure 1. Temperature effect on the enantioselectivity in the palladium-
catalyzed Mannich-type reaction. Yields are shown in parentheses.
palladium(II) diaquo complexes, [Pd((R)-binap)(H2O)2]2+
-
(BF4-)2 (3a) or [Pd((R)-tol-binap)(H2O)2]2+(BF4-)2 (3b),7 we
suspected that this chiral palladium catalyst system might be
useful for reactions with imines.
First, based on the best reaction conditions for the aldol
reaction, we tested the reaction of enol silyl ether 1a with the
imine 2 in DMF at 0 °C in the presence of the Pd diaquo
complex 3a (10 mol %); however, only a negligible asymmetric
induction was observed. After extensive experimentation, we
found that 5a could be obtained in 67% ee and 85% yield using
the following rather complicated procedure. A solution of Pd
diaquo complex 3a (10 mol %) and 1a (1.5 equiv) in DMF
was stirred at 25 °C for 1 h, after which time the solution was
warmed to 60 °C, and a solution of imine 2a was added over 4
h using a syringe pump. During the addition of 2a, additional
1a (three times 0.5 equiv at 1-h intervals) was supplied to the
reaction mixture. The whole mixture was stirred at the same
temperature for an additional 2 h. As shown in Figure 1, the
enantiomeric excess of the product (5a) decreased from 67%
to 3% ee as the reaction temperature was lowered from 60 to
25 °C. This temperature-dependent decrease of enantioselectivity
was also observed with methyl ester 2b.
Figure 2.
as dichloromethane and benzotrifluoride, were slower than that
in DMF, and the ee’s of the products were much lower (results
for 5c: in CH2Cl2, 35 °C, 9 h, 24%, 47% ee; in C6H5CF3, 60
°C, 7 h, 36%, 38% ee; in DMF, 60 °C, 7 h, 70%, 62% ee). The
unusual temperature dependence and the sensitivity of the
reaction conditions strongly suggest the existence of an undes-
ired competitive reaction pathway that affords the racemic
product.
The imine concentration should be kept low during the
reaction. When imine 2b was added in one portion, the
enantioselectivity of 5b was drastically decreased. Preincubation
of 1a with the Pd diaquo complex 3a was also critical for the
high asymmetric induction. As the preincubation time was
reduced, the enantiomeric excess decreased significantly (see
Supporting Information). Reactions in less polar solvents, such
Our preliminary mechanistic studies on the aldol reaction7
suggested the formation of Pd enolate 6 from enol silyl ether
1a and Pd aquo complex 3. If the reaction to produce highly
optically active 5 proceeds via the palladium enolate 6, one
plausible route to racemic 5 would be a proton-catalyzed one
(Figure 2). Upon generation of the palladium enolate from 1a
and the diaquo complex 3, an equivalent amount of tetrafluo-
roboric acid should be generated, and this strong protonic acid
could catalyze the unselective addition of 1a to 2. Indeed, HBF4
(5 mol %) prepared from AgBF4 and trimethylsilyl chloride in
DMF (containing a trace amount of water) did catalyze the
reaction of 1a with 2a to give 5a in 76% yield after 3 h at 25
°C. To neutralize the HBF4 generated in situ, various bases were
tested as an additive to the 3a-catalyzed reaction (Table 1). As
expected, various amine bases and also calcium carbonate were
found to be effective to improve the asymmetric induction, and
the ee of the product increased to around 80% (entries 1-6).
In the presence of CaCO3, however, reaction at 25 °C afforded
(5) A limited number of examples of N-arylation of optically active
R-amino acids and intramolecular arylation of R -amino acid ester have
been reported. See: (a) Ma, D.; Yao, H. Tetrahedron: Asymmetry 1996,
7, 3075-3078. (b) Wagaw, S.; Rennels, R. A.; Buchwald, S. L. J. Am.
Chem. Soc. 1997, 119, 8451-8458. (c) Ma, D.; Zhang, Y.; Wu, S.; Tao, F.
J. Am. Chem. Soc. 1998, 120, 12459-12467. For a review on the
N-arylation, see: (d) Hartwig, J. F. Angew. Chem., Int. Ed. 1998, 37, 2047-
2067 and references therein.
(6) (a) Ferraris, D.; Young, B.; Dudding, T.; Lectka, T. J. Am. Chem.
Soc. 1998, 120, 4548-4549. (b) Ferraris, D.; Young, B.; Cox, C.; Drury,
W. J. III; Dudding, T.; Lectka, T. J. Org. Chem. 1998, 63, 6090-6091.
(7) (a) Sodeoka, M.; Ohrai, K.; Shibasaki, M. J. Org. Chem. 1995, 60,
2648-2649. (b) Sodeoka, M.; Tokunoh, R.; Miyazaki, F.; Hagiwara, E.;
Shibasaki, M. Synlett 1997, 463-466. (c) Sodeoka, M.; Shibasaki, M. Pure
Appl. Chem. 1998, 70, 411-414.

5452 J. Am. Chem. Soc., Vol. 121, No. 23, 1999
Fujii et al.
[{Pd((R)-binap)(µ-OH)}2]2+(BF4-)2 (4a), [{Pd((R)-tol-binap)-
(µ-OH)}2]2+(BF4-)2 (4b), [{Pd((R)-binap)(µ-OH)}2]2+(TfO-)2
(4c), or [{Pd((R)-tol-binap)(µ-OH)}2]2+(TfO-)2 (4d), could be
made by the treatment of the diaquo complex 3a, 3b, 3c, or
3d, respectively, with 4-Å molecular sieves in wet acetone.
Furthermore, it was found that these µ-hydroxo complexes could
be obtained directly from the dichloride complexes 8a or 8b
by the treatment of silver salts in the presence of 4-Å molecular
sieves in wet acetone9 or CH2Cl2, but that a large amount of
molecular sieves was required for the complete transformation
to the µ-hydroxo complex. Finally, simple treatment of a solution
of the aquo complexes 3a-d in CH2Cl2 with aqueous NaOH
solution was found to give these µ-hydroxo complexes 4a-d
in excellent yield.
Table 1. Effect of Base in Enantioselective Mannich-Type
Reaction Catalyzed by Aqua Pd Complex 3aa
1a + 2a 10 mol % 3a8 5a
DMF
entry additive (equiv) temp (°C) time (h) yield (%) ee (%)
1
2
3
4
5
6
7b
8
-
60
60
60
60
60
60
25
60
10
12
5
5
6
15
10
12
85
28
57
85
81
84
86
-
67
78
67
80
75
79
40
-
NEt3 (0.13)
iPr2NEt (0.5)
pyridine (0.48)
2,6-lutidine (0.5)
CaCO3 (1.0)
CaCO3 (1.0)
K2CO3 (1.1)
a After preincubation of 3a with 1a at 25 °C for 40 min and then
addition of base, 2a was added by syringe pump over 4 h at 60 °C.
b To the mixture of 3a, 1a, and CaCO3 was added 2a in one portion.
1
The H NMR spectrum of 4d in CDCl3, which contained a
singlet at -3.03 ppm, as is often the case for a µ-hydroxo group,
was clearly distinct from that of 3d.10 Interestingly, treatment
of 3d with an insufficient amount of 4-Å molecular sieves
resulted in production of the binuclear momo-µ-hydroxo species
9. Chemical shifts of the tolyl methyl groups of Tol-BINAP
would thus seem to be a good marker of the ligand environment.
The monomeric complex [Pd((R)-tol-binap)(H2O)2]2+(TfO-)2
(3d) showed two singlets at each of 2.05 and 2.42 ppm. One of
the singlets at 2.42 ppm moves to a higher field in the spectrum
of the binuclear complex [{Pd((R)-tol-binap)(µ-OH)}2]2+(TfO-)2
(4d) (1.96, 2.11 ppm), probably because of the anisotropic effect
of the tolyl group of another ligand. In addition to these aquo
and µ-hydroxo complexes, mononuclear PdCl2((R)-tol-binap)
complex 8b and binuclear µ-chloro complex prepared by
treatment of 8b with 1 equiv of silver triflate in CH2Cl2 also
show a similar pattern of chemical shifts (2.00, 2.38 and 2.02,
2.06 ppm, respectively, in CDCl3).11 In contrast to these highly
symmetric complexes, the binuclear mono-µ-hydroxo complex
9 shows four singlets at each of 1.82, 2.06, 2.14, and 2.24 ppm,
suggesting less symmetrical structure.
Scheme 2
Electrospray ionization mass spectrometry (ESI-MS) analysis
clearly supports the above-determined structure of the mono-
nuclear complex 3d and the binuclear complexes 4d and 9.12
A
the product with lower ee (entry 7). Addition of more soluble
inorganic bases, such as potassium carbonate and sodium
bicarbonate, accelerated decomposition of the imine, and no
product was obtained (entry 8).
large peak at m/z 933 corresponding to the [Pd((R)-tol-binap)-
(TfO)]+ fragment was observed in the spectrum of the mono-
nuclear complex 3d. In contrast, a small peak at m/z 1752
corresponding to a monocationic fragment [{Pd((R)-tol-binap)-
(µ-OH)}2(TfO)]+, and a large peak at m/z 802 corresponding
to a dicationic fragment [{Pd((R)-tol-binap)(µ-OH)}2]2+, were
observed in the spectrum of the binuclear complex 4d (Figure
5c). The spectrum of the complex 9 shows peaks corresponding
to [{Pd((R)-tol-binap)}2(µ-OH)(OH)(H2O)(TfO)]+ (m/z 1772),
[Pd((R)-tol-binap)(OH)(H2O)]+ (m/z 821), and [{Pd((R)-tol-
binap)}2(µ-OH)(OH)(H2O)]2+ (m/z 811). These data strongly
indicate that ESI-MS is a powerful tool for analyzing these
cationic palladium complexes.
Preparation of Binuclear µ-Hydroxo Pd Complex. To
overcome these unfavorable protonic acid effects, we next
attempted to prepare a mono-hydroxo complex 7, which is
expected to give a palladium enolate without undesired forma-
tion of HBF4. It has been reported that simple treatment of a
palladium dichloride complex of 1,2-bis(diphenylphosphino)-
propane (DPPP) or 1,1′-bis(diphenylphosphino)ferrocene (DPPF)
with silver salt in wet solvent affords a corresponding binuclear
µ-hydroxo complex that is a dimeric form of the mono-hydroxo
complex.8 However, it has also been demonstrated that simple
treatment of PdCl2((R)-binap) (8a) and PdCl2((R)-tol-binap) (8b)
with silver tetrafluoroborate in wet acetone gave the aquo
complexes 3a and 3b in excellent yield.7b Steric bulkiness of
optically active ligands may prevent automatic formation of the
binuclear complex under these conditions (Scheme 2). We
therefore tried to remove 1 equiv of HBF4 from the diaquo
complex. After examination of a variety of possibilities, we
determined that either of the binuclear µ-hydroxo complexes,
(9) Prolonged reaction in acetone with a large amount of 4-Å molecular
sieves caused byproduction of acetone dimer.
(10) Similar binuclear µ-hydroxo Pd complexes of dppp and dppf have
been reported to show the OH signal at -2.3 and -2.32 ppm, respectively,
1
in the H NMR.8a,b
(11) Chemical shifts of the methyl groups of the monochloride complex
in DMF-d7 were 2.07 and 2.45 ppm, suggesting dissociation of the binuclear
µ-chloro complex to the corresponding mononuclear complex similar to
10 in the coordinating solvent.
(12) For a review on ESI-MS of metal complexes, see: (a) Colton, R.;
D’Agostino, A.; Traeger, J. C. Mass Spectrom. ReV. 1995, 36, 1718-1719.
For examples of the use of ESI-MS to observe organometallic reaction
intermediates, see: (b) Aliprantis, A. O.; Canary, J. W. J. Am. Chem. Soc
1994, 116, 6985-6986. (c) Feichtinger, D.; Plattner, D. A. Angew. Chem.,
Int. Ed. Engl. 1997, 36, 1718-1719. (d) Feichtinger, D.; Plattner, D. A.;
Chen, P. J. Am. Chem. Soc. 1998, 120, 7125-7126 and references therein.
(8) (a) Pisano, C.; Consiglio, G.; Sironi, A.; Moret, M. J. J. Chem. Soc.,
Chem. Commun. 1991, 421-423. (b) Longato, B.; Pilloni, G.; Valle, G.
Corain, B. Inorg. Chem. 1988, 27, 956-958. (c) Bushnell, G. W.; Dixon,
K. R.; Hunter, R. G.; McFarland, J. J. Can. J. Chem. 1972, 50, 3694-
3699.

A NoVel Binuclear Palladium Enolate Complex
J. Am. Chem. Soc., Vol. 121, No. 23, 1999 5453
Scheme 3
isopropyl glyoxylate and anisidine in DMF in the presence of
4-Å molecular sieves, and then enol silyl ether 1a (1.5 equiv)
and the catalyst 4b (5 mol %) were added to this mixture. The
desired product 5a was obtained in 74% chemical yield and
83% ee, and only a small amount (5%) of the aldol product
was obtained under these conditions. Since the catalyst 4 can
also catalyze the aldol reaction,13 and anisidine may cause
deactivation of the catalyst, a clean formation of imine using
molecular sieves was critical for achieving high chemical and
optical yields.
Figure 3. ORTEP diagram of the µ-hydroxy palladium complex of
(R)-Tol-BINAP (4d). The TfO- anions and hydrogens are omitted for
clarity.
Table 2. Enantioselective Mannich-Type Reaction of 2a Catalyzed
by Binuclear µ-OH Pd Complex 4a
catalyst
1a + 2a
8 5a
solvent
Reaction Mechanism. Enolate complexes of transition metals
have received considerable attention in recent years. Several
transition metal complexes with either carbon-bound (A),
oxygen-bound (B), or oxo π-allylic-type (C) enolate ligands have
been structurally characterized (Scheme 4).14 In addition to our
studies on asymmetric aldol7 and Mannich-type3 reactions, it
has been proposed that palladium enolate was involved in
several synthetically useful reactions.15 The number of well-
characterized palladium enolate complexes is, however, rather
limited.16 We have reported O-bound palladium enolate forma-
tion by the reaction of 1a with the palladium complex 10
timeb
B
catalyst
entry (mol %) solvent
A
C
yield (%) ee (%)
1
2
3
4
5
6
7
8
9
4b (3) DMF
4b (3) DMF
4b (3) DMF
4a (3) DMF
30 min 4 hc
2 hc
73
82
80
87
45
75
59
60
79
95
80
76
80
81
83
53
81
82
67
83
90
89
15 min 4 hc
1 hc
60 min 2 min 24 h
45 min 2 min 24 h
4b (3) CH2Cl2 60 min 4 h
119 h
3 h
4b (3) DMI
4b (3) TMU
4b (3) THF
4b (3) NMP
4b (5) DMF
4d (5) DMF
45 min 4 h
40 min 5 min 21.5 h
40 min 4 h 112 h
40 min 5 min 15 h
40 min 5 min 18.5 h
40 min 5 min 18 h
10
11
(13) Fujii, A.; Sodeoka, M., unpublished results. Reaction of 1a with
benzaldehyde catalyzed by 4d (1 mol %) in wet DMF at 23 °C for 7 h
gave the aldol product in 71% yield and 72% ee.
a Reaction was carried out at 25 °C. b A, preincubation time; B, time
for addition of imine 2a; C, additional reaction time. c Reaction was
carried out at 60 °C.
(14) For a review, see: (a) Mekelburger, H. B.; Wilcox, C. S. In
ComprehensiVe Organic Synthesis; Trost, B. M., Fleming, I., Heathcock,
C. H., Eds.; Pergamon Press: New York, 1991; Vol. 2, pp 99-131 and
references therein. For recent selected examples of middle and late transition
metal enolate complexes, see the following references. W and Mo: (b)
Burkhardt, E. R.; Doney, J. J.; Slough, G. A.; Stack, J. M.; Heathcock, C.
H.; Bergman, R. G. Pure Appl. Chem. 1988, 60, 1-6. (c) Faller, J. W.;
Ma, Y. Organometallics 1993, 12, 1927-1930. Re: (d) Stack, J. G.;
Simpson, R. D.; Hollander, F. J.; Bergman, R. G.; Heathcock, C. H. J. Am.
Chem. Soc. 1990, 112, 2716-2729. Ru: (e) Rasley, B. T.; Rapta, M.;
Kulawiec, R. J. Organometallics 1996, 15, 2852-2854. (f) Hartwig, J. F.;
Bergman, R. G.; Andersen, R. A. Organometallics 1991, 10, 3326-3344.
Rh and Ir: (g) Milstein, D. J. Am. Chem. Soc. 1982, 104, 5227-5228. (h)
Aoyama, Y.; Yoshida, T.; Ogoshi, H. Tetrahedron Lett. 1985, 26, 6107-
6108. (i) Slough, G. A.; Bergman, R. G.; Heathcock, C. H. J. Am. Chem.
Soc. 1989, 111, 938-949. (j) Milstein, D.; Calabrese, J. C. J. Am. Chem.
Soc. 1982, 104, 3773-3774. Pt: (k) Cairns, M. A.; Dixon, K. R.; Smith,
M. A. R. J. Organomet. Chem. 1977, 135, C33-C34. (l) Arnold, D. P.;
Bennett, M. A. J. Organomet. Chem. 1980, 199, 119-135. (m) Almeida,
J. F.; Pidcock, A. J. Organomet. Chem. 1981, 209, 415-423. (n) Fujimura,
O. J. Am. Chem. Soc. 1998, 120, 10032-10039. Cu: (o) Kru¨ger, J.; Carreira,
E. M. J. Am. Chem. Soc. 1998, 120, 837-838 and references therein.
(15) (a) Ito, Y.; Hirano, T.; Saegusa, T. J. Org. Chem. 1978, 43, 1011-
1013. (b) Ito, Y.; Aoyama, H.; Hirano, T.; Mochizuki, A.; Saegusa, T. J.
Am. Chem. Soc. 1979, 101, 494-496. (c) Tsuji, J.; Mandai, T. Acc. Chem.
Res. 1987, 20, 140-145. (d) Nokami, J.; Mandai, T.; Watanabe, H.;
Ohyama, H.; Tsuji, J. J. Am. Chem. Soc. 1989, 111, 4126-4127. (e) Nokami,
J.; Watanabe, H.; Mandai, T.; Kawada, M.; Tsuji, J. Tetrahedron Lett. 1989,
30, 4829-4832. (f) Masuyama, Y.; Sakai, T.; Kato, T.; Kurusu, Y. Bull.
Chem. Soc. Jpn. 1994, 67, 2265-2272. (g) Kuwajima, I.; Urabe, H. J. Am.
Chem. Soc. 1982, 104, 6831-6833. (h) Palucki, M.; Buchwald, S. L. J.
Am. Chem. Soc. 1997, 119, 11108-11109. (i) Hamann, B. C.; Hartwig, J.
F. J. Am. Chem. Soc. 1997, 119, 12382-12383. (j) Åhman, J.; Wolfe, J.
P.; Troutman, M. V.; Palucki, M.; Buchwald, S. L. J. Am. Chem. Soc. 1998,
120, 1918-1919. (k) Sugiura, M.; Nakai, T. Angew. Chem., Int. Ed. Engl.
1997, 36, 2366-2368 and references therein.
Finally, the binuclear µ-hydroxo structure of 4d was con-
firmed by X-ray crystallography. As shown in Figure 3, the
crystal structure is slightly distorted from the complete sym-
metric structure to avoid steric interaction of the tolyl groups.
µ-Hydroxo Complex-Catalyzed Reactions. Using the novel
chiral complex 4b (3 mol %) as a catalyst, we first tested the
reaction using the procedure described above (Table 2, entry
1). As expected, 5a was obtained in 73% yield and 76% ee
without addition of any base. It was also found that the
preincubation, slow addition, and warming to 60 °C were no
longer necessary (entries 2 and 3). Catalysts 4a and 4b did not
show much difference in reactivity and selectivity for these
substrates (entries 3 and 4). The solvent effects were also
reexamined, and amide or urea solvents were found to be
suitable (entries 5-9). Finally, 95% chemical yield and 90%
ee were achieved by the reaction of 1a with 2a at 25 °C using
5 mol % of the catalyst 4b (entry 10). Since decomposition of
1a was slow under these conditions, repeated addition of 1a
during the reaction was no longer necessary to complete the
reaction. Similar results were also obtained using the triflate
complex 4d (entry 11). Optically pure 5a was obtained by the
simple recrystallization of the products with more than 80%
ee.
Next, we tried a one-pot, three-component coupling reaction
(Scheme 3). The imine 2a was prepared in situ from the

5454 J. Am. Chem. Soc., Vol. 121, No. 23, 1999
Fujii et al.
Scheme 4
prepared from 8a and 1 equiv of silver tetrafluoroborate (Scheme
1
4).7b This palladium O-enolate 6a was characterized in the H
NMR spectrum in DMF-d7 as having three singlets at 4.41, 4.68,
and 9.54 ppm. There were several possibilities for the ligand L
of the complex. The possibility of the chloride anion has already
been ruled out7b on the basis of the observation of the same
three singlets after reaction of the aqua complex 3a with 1a.
Since the hydroxy proton of the transition metal-OH complex
usually appears at a very high field, as mentioned above,10,17 it
is not likely that the signal at 9.54 ppm represents Pd-OH.
Silanol complex (L ) Me3SiOH) is also an unlikely candidate,
because exactly the same three singlets were observed when
the same NMR experiment was carried out using the corre-
sponding diphenylmethylsilyl enol ether and 10.18 Thus, we
assumed that the ligand L in the observed species was water
(6aw), and that only one of the two protons of the water ligand
which was stabilized by intramolecular hydrogen bonding to
the enolate oxygen was observed as a singlet, while the other
might rapidly exchange with water molecules in the solvent.
Since this hydrogen-bonding proton is expected to have a
character intermediate between those of the coordinating water
and enol, the chemical shift at an unusually low field would be
reasonable.
To determine whether the reaction using the µ-hydroxo
complex proceeds via the same O-bound palladium enolate, we
next undertook NMR experiments using 4c. To a solution of
the µ-hydroxo complex 4c (0.01 mmol) in DMF-d7 was added
0.01 mmol of the enol silyl ether 1a (methylene singlets at 4.51
and 5.11 ppm) at room temperature. No rapid formation of the
three singlets corresponding to 6cw was observed. Instead, as
the singlet at -1.70 ppm of the µ-OH proton of 4c slowly
decreased (Figure 4b), three new signals, a doublet of doublets
Figure 4. 1H NMR spectra: (a) µ-hydroxy palladium complex 4c;
(b) reaction mixture at 40 min after the addition of 1a (1 equiv to 4c)
to 4c in DMF-d7; (c) reaction mixture at 40 min after the addition of
excess 1a to the mixture in (b).
at 4.05 ppm, a multiplet at 3.35 ppm, and one broad singlet at
1.26 ppm, appeared (Figure 4b), indicating formation of a novel
complex. Unlike 6cw generated from 3c in DMF-d7, this novel
complex was found to be fairly stable for several hours at room
temperature. Furthermore, upon addition of excess 1a (3 equiv
to 4c) to this mixture, signals of 6cw were observed (Figure
4c), suggesting that this complex can be converted to 6cw.
To obtain additional information about the structure of this
novel complex, D2O was added to the mixture. Two signals at
4.05 and 3.35 ppm survived, but rapid disappearance of the
broad singlet at 1.26 ppm was also observed, suggesting that
this third peak should be assigned to the exchangeable proton.
The signals at 4.05 and 3.35 ppm, indicating apparent coupling
with the phosphorus atoms of the ligand, suggest the π-coor-
dination of enolate to the palladium. Since chemical shifts of
protons of the oxo-π-allyl complex 11 prepared by the reaction
of cationic complex 10 with the potassium enolate of
acetophenone7a were completely different (dd at 3.98 ppm and
m at 2.25-2.44 ppm), this novel complex would have η2-
coordination to the Pd atom. Finally, the structure of this
complex was taken to be a unique binuclear palladium-
sandwiched enolate 12c on the basis of the results of ESI-MS
analysis (Figure 5b). In addition to the signals derived from
the original µ-hydroxo complex 4c (m/z 746 for [{Pd((R)-binap)-
(µ-OH)}2]2+, 1643 for [{Pd((R)-binap)(µ-OH)}2(TfO)]+), three
signals corresponding to the cationic fragments [{Pd((R)-
binap)}2(µ-O-C(C6H5)dCH2)(µ-OH)]2+ (m/z 797), [Pd((R)-
binap)(O-C(C6H5)dCH2)]+ (m/z 849), and [{Pd((R)-binap)}2-
(µ-O-C(C6H5)dCH2)(µ-OH)(TfO)]+ (m/z 1744) were observed.
There was good agreement between the experimental and
calculated isotopic distributions of each ion. In this complex
12c, the unstable mononuclear O-bound palladium enolate
would be stabilized by the intramolecular η2-coordination to
the second palladium atom. Floriani et al. reported a symmetrical
(16) (a) Yoshimura, N.; Murahashi, S.-I.; Moritani, I. J. Organomet.
Chem. 1973, 52, C58-C60. (b) Ito, Y.; Nakatsuka, M.; Kise, N.; Saegusa,
T. Tetrahedron Lett. 1980, 21, 2873-2876. (c) Bertani, R.; Castellani, C.
B.; Crociani, B. J. Organomet. Chem. 1984, 269, C15-C18. (d) Wanat, R.
A.; Collum, D. B. Organometallics 1986, 5, 120-127. (e) Burkhardt, E.
R.; Bergman, R. G.; Heathcock, C. H. Organometallics 1990, 9, 30-44.
(f) Veya, P.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. Organometallics 1993,
12, 4899-4907. (g) Vicente, J.; Abad, J. A.; Chicote, M. T.; Abrisqueta,
M. D. Lorca, J. A.; de Arellano, M. C. R. Organometallics 1998, 17, 1564-
1568 and references therein.
(17) For a review of hydroxo complexes of transition metals, see: (a)
Bryndza, H. E.; Tam, W. Chem. ReV. 1988, 88, 1163-1188 and references
therein. For hydroxo palladium complexes, see: (b) Yoshida, T.; Okano,
T. Otsuka, S. J. Chem. Soc., Dalton Trans. 1976, 993-999. (c) Grushin,
V. V.; Alper, H. Organometallics 1993, 12, 1890-1901. (d) Grushin, V.
V.; Alper, H. Organometallics 1996, 15, 5242-5245 and references therein.
(18) Fujii, A.; Sodeoka, M., unpublished results.

A NoVel Binuclear Palladium Enolate Complex
J. Am. Chem. Soc., Vol. 121, No. 23, 1999 5455
at much higher field (2.67 and 2.97 ppm) than those of 12c,
indicating that their binding mode was distinct from that of 12c.
Upon addition of excess 1a, the intensity of the signal at m/z
849, which corresponded to the mononuclear enolate, was
significantly increased in ESI-MS (Figure 5c).19 These observa-
tions suggest that 6 generated from 12 under catalytic conditions
would be an active intermediate for the reaction with imine.
The same type of binuclear complex formation (12d) was
observed in the experiments using the Tol-BINAP complex 4d.
Three similar signals (at 3.94, 3.33, and 0.92 ppm) were
observed in 1H NMR, and a similar fragmentation pattern (m/z
853, 903, and 1855) was observed in ESI-MS.
Finally, 12c was found to react with imine 2a. The imine 2a
(0.004 mmol) was added to the solution of 12c prepared from
1a (0.01 mmol) and 4c (0.01 mmol) in DMF-d7 after confirma-
tion of complete consumption of the enol silyl ether 1a (after 5
h) by the 1H NMR. The complex 12c was slowly reacted with
2a to give the optically active product 5a (93% ee) in 79% yield.
This fact clearly indicated that the highly optically active product
was formed from this novel complex 12c, and not from the
reaction of the enol silyl ether 1a with imine simply activated
by the palladium complex.
We next tried to isolate a mononuclear palladium enolate
complex stabilized by intramolecular coordination. To ac-
complish this, the µ-hydroxo complex 4c was mixed with the
enol silyl ether of 2-acetyl pyridine 13 (1 equiv to Pd, methylene
singlets at 4.59 and 5.67 ppm) in DMF-d7. In this case, slow
formation of two signals, ddd at 3.25 ppm (2J(H-H) ) 8.7 Hz,
3J(P-H, trans) ) 13.2 Hz, 3J(P-H, cis) ) 1.5 Hz) and ddd at
2.31 ppm (2J(H-H) ) 8.7 Hz, 3J(P-H, trans) ) 8.7 Hz,
3J(P-H, cis) ) 4.7 Hz), was observed, and this result was
consistent with the formation of C-bound enolate 14c (Figure
6a). The chemical shifts and the coupling pattern were similar
to those observed in known Pd and Pt carbon-bound eno-
lates,16,20 but not to those in O-bound enolates. The same
reaction was also carried out using the Tol-BINAP complex,
and a clean conversion of the two singlets of the tolyl methyl
group in 4d to four singlets at 2.47, 2.32, 2.09, and 2.04 ppm
indicated an almost quantitative formation of 14d and supported
the unsymmetrical monomeric structure. As shown in Figure
6b, ESI-MS analysis also supported the monomeric enolate
structure 14. It is likely that the mononuclear palladium enolate
of acetophenone exists in rapid equilibrium between the C-bound
and O-bound forms. And only the water-coordinated form 6w,
stabilized by intramolecular hydrogen bonding, was observed
1
as sharp singlets in H NMR. In the case of the enolate of
2-acetylpyridine, the C-bound form would be stabilized by the
strong intramolecular coordination of the nitrogen to palladium.
However, no reaction of these C-bound palladium enolates in
DMF-d7 with the imine 2a or benzaldehyde was observed up
to 50 °C. This low reactivity strongly suggests that reaction of
the palladium enolate with imine (or aldehyde) requires a vacant
coordination site on the palladium atom.
Possible reaction pathways using the µ-hydroxo complex are
summarized in Scheme 5. First, the sandwiched enolate 12
would be formed from 4 by the reaction with 1a, probably via
the partially dissociated form 15. Then, dissociation of 12 would
afford the mononuclear palladium O-enolate complex 6 and the
Figure 5. ESI mass spectra with observed (upper) and calculated
(lower) isotope distribution for the peaks: (a) µ-hydroxo palladium
complex 4c; (b) reaction mixture at 40 min after the addition of 1a (1
equiv to 4c) to 4c in DMF-d7; (c) reaction mixture at 40 min after the
addition of excess 1a to the mixture in (b).
(19) As shown in Figure 8c, the signal around m/z 745 showed a different
pattern from that in Figure 8a, which was good agreement with the calculated
isotope distribution for the monocationic fragment, [Pd((R)-binap)(OH)]+.
This fact indicated that the mononuclear complex 7 was formed in the
reaction mixture.
(20) (a) Tsutui, M.; Ori, M.; Francis, J. J. Am. Chem. Soc. 1972, 94,
1414-1415. (b) Hillis, J.; Tsutui, M. J. Am. Chem. Soc. 1973, 95, 7907-
7908.
binuclear palladium complex in which two enolate anions bridge
two (Ph3P)2Pd fragments through the carbon and oxygen
atoms.16f But the methylene protons of this complex appeared

5456 J. Am. Chem. Soc., Vol. 121, No. 23, 1999
Fujii et al.
Scheme 6
than that from the µ-hydroxo complex 7, although decomposi-
tion was also much faster. It is possible that palladium enolate
6 would be generated from 3 via a different mechanism. Tsutui
et al. reported a stepwise preparation of a C-bound platinum
enolate 18 via π-complex of the enol silyl ether followed by
formation of the enol complex by acid hydrolysis.20 It is likely
that the palladium enolate 6 is formed from 3 via such a pathway
(Scheme 6). In the case of palladium, however, the enol or the
enolate complex would be very unstable in the presence of
protonic acid and water.
Lectka et al. reported that reaction of 1a with 2d (R2 ) Et,
R3 ) tosyl) in CH2Cl2 at -80 °C using a dicationic palladium
complex, [Pd((R)-binap)(CH3CN)2]2+(ClO4-)2 (21), gave 5d (R2
) Et, R3 ) tosyl) of 80% ee.6 Because we wondered if their
reaction also proceeded via a mechanism similar to that reported
here, we tested the tosylimine substrate 2d using our catalysts.
Surprisingly, however, no asymmetric induction was observed
for the substrate 2d using the catalyst 3 or 4 in DMF or CH2-
Cl2 at any temperature. The highly activated imine 2d reacted
with 1a in DMF at 23 °C even in the absence of any catalyst to
give the racemic product in good yield. Rigorous anhydrous
conditions and low temperature for the reaction using 21 in CH2-
Cl2 were required for the high asymmetric induction. Although
the structures of the complexes 3 and 21 are quite similar, it is
reasonable that the reaction using aquo complex 3 in CH2Cl2
at -78 °C afforded racemic 2d, since the coordinated water
can be a potential proton source, causing nonselective reaction
of the highly activated substrate. Reaction of 1a with 2a using
the dicationic complex, such as 21 in CH2Cl2 at -78 °C, gave
5a, but only in low chemical and optical yields. Next, to
determine the difference between the reaction mechanisms
involved in polar (e.g., DMF) and nonpolar solvents, several
1H NMR experiments were carried out. In contrast to the
observations described above for DMF-d7, no Pd enolate
formation from 3 or 4 was observed in CDCl3. Though no
change in the spectrum of imine 2a was observed upon mixing
with 3 or 4 in DMF-d7, some broadening of the peaks of 2a
was observed in CDCl3 upon mixing with 1 equiv of the aquo
complex 3d, suggesting strong coordination of the imine to
palladium in the nonpolar solvent. Based on these observations,
we agree that the reaction using 21 in nonpolar solvent would
proceed via a simple Lewis acid mechanism. It is quite
interesting that these two similar Pd(II)-catalyzed reactions
proceed via completely different reaction mechanisms according
to the nature of the catalyst, solvent, and substrate.
Figure 6. 1H NMR spectra (a) and ESI mass spectra (b) of palladium
C-enolate with observed (upper) and calculated (lower) isotope distribu-
tions for the peak.
Scheme 5
mononuclear hydroxo palladium complex 7. In the presence of
excess enol silyl ether 1a, the hydroxo complex 7 would be
further converted to the enolate complex 6. Alternatively, a
certain amount of enolate complex 6 might be decomposed by
water to give acetophenone and 7. Coordination of imine to
the palladium (16) would then activate both enolate and imine,
and C-C bond formation would occur through the highly
ordered transition state to give 17. Hydrolysis of 17 would give
the highly optically active product 5a and regenerate 7.
Palladium enolate formation from the aquo complex was faster
Conclusion
We have found that highly optically active acylalanine
derivatives (up to 90% ee) can be obtained by the enantiose-
lective Mannich-type reaction of enol silyl ethers 1 with N-aryl-
iminoacetic acid esters 2 catalyzed by chiral palladium(II)
complexes. Preliminary studies using the known palladium (II)

A NoVel Binuclear Palladium Enolate Complex
J. Am. Chem. Soc., Vol. 121, No. 23, 1999 5457
and free H2O), 6.65 (br s, 2H), 6.71 (d, J ) 8.6 Hz, 2H), 7.22 (dd, J
) 7.5, 7.5 Hz, 2H), 7.38-7.70 (m, 22H); [R]20D +357.8 (c 0.25, CHCl3).
[Pd((R)-binap)(H2O)2]2+(TfO-)2 (3c): yellow powder; mp 125 °C
diaquo complex 3 indicated undesirable effects of protonic acid
generated in situ, but this problem was overcome by develop-
ment of the novel chiral binuclear µ-hydroxo palladium(II)
complex 4. Mechanistic studies using NMR and ESI-MS
indicate that this reaction proceeds via optically active palladium
enolate. The unique binuclear palladium-sandwiched enolate 12
was observed in the reaction of the µ-hydroxo palladium
complex 4 with the enol silyl ether 1a. To the best of our
knowledge, this is the first example of characterization of chiral
binuclear O- and π-bound palladium enolate complexes pos-
sessing high reactivity as nucleophiles.
1
dec; IR (KBr) 3200, 1290, 1170, 700 cm-1; H NMR (CDCl3, 200
MHz) δ 3.46 (br s, >6H, coordinated and free H2O), 6.69 (d, J ) 8.7
Hz, 2H), 6.78-7.06 (m, 6H), 7.21 (dd, J ) 7.5, 7.5 Hz, 2H), 7.45-
7.87 (m, 22H); [R]20 +365.3 (c 0.38, CHCl3).
D
Representative Procedure for the Reaction Using the Aquo
Complexes. To a solution of 3a (9.6 mg, 0.010 mmol) in DMF (1.0
mL, distilled from CaH2) was added 1a (31 µL, 0.15 mmol) at 25 °C
under an atmosphere of argon. After being stirred at 25 °C for 1 h, the
mixture was then warmed to 60 °C. To this mixture was added a
solution of imine 2a (22.0 mg, 0.10 mmol) in DMF (0.5 mL) by syringe
pump over 4 h. During the addition of 2a, 1a (0.5 equiv) was added
three times at 1-h intervals. The whole reaction mixture was stirred at
60 °C for an additional 2 h, diluted with ether, and filtered through a
short silica gel column. The solution was poured into water and
extracted with ether (2 × 5 mL). The combined organic layers were
washed with water and brine, dried over Na2SO4, and concentrated.
The residue was purified by preparative silica gel thin-layer chroma-
tography (AcOEt/hexane ) 1/4) to afford 5a (26.7 mg, 85%). The ee
of 5a was determined to be 67% by HPLC analysis using Daicel
Chiralcel OG (hexane:iPrOH ) 9:1). Similarly, the reactions were
carried out at various temperatures (Figure 1), with various preincu-
bation times (Supporting Information), or in the presence of bases (Table
1).
Experimental Section
General. All melting points were determined with a Yamato MP-
21 microscale melting point apparatus and are uncorrected. Infrared
(IR) spectra were recorded on a JASCO FT/IR-5300 infrared spec-
trometer. NMR spectra were measured on a Brucker AC-200 spec-
1
trometer, operating at 200 MHz, for H NMR. Chemical shifts were
reported in the δ scale relative to tetramethylsilane (TMS). Optical
rotation was measured on a Horiba SEPA-200 automatic digital
polarimeter. Low-resolution mass spectra were taken with a Hitachi
RMU-6MG spectrometer, and high-resolution mass spectra were
obtained on a Hitachi M-80A spectrometer. In general, reactions were
carried out in dry solvents under argon atmosphere, unless otherwise
mentioned. N,N-Dimethylformamide (DMF) and dichloromethane were
distilled from calcium hydride.
N-(4-Methoxyphenyl)benzoylalanine Isopropyl Ester (5a): color-
less needles; mp 113.0-113.5 °C (100% ee); IR (KBr) 3370, 2355,
2340, 1725, 1680, 1515, 1280, 1220, 1195, 1110, 820 cm-1; 1H NMR
(CDCl3) δ 1.14 (d, J ) 6.3 Hz, 3H), 1.19 (d, J ) 6.3 Hz, 3H), 3.50 (d,
J ) 5.4 Hz, 2H), 3.73 (s, 3H), 4.18 (br s, 1H), 4.50 (br t, J ) 5.4 Hz,
1H), 5.03 (sept, J ) 6.3 Hz, 1H), 6.68 (d, J ) 9.1 Hz, 2H), 6.77 (d, J
) 9.1 Hz, 2H), 7.41-7.61 (m, 3H), 7.94 (d, J ) 7.0 Hz, 2H); MS (EI)
m/z 341 (M+); HRMS m/z calcd for C20H23NO4 341.1625, found
Preparation of Imines. Iminoacetic acid esters 2a-c were prepared
according to the reported procedure.21 Namely, for 2a, to a stirred
solution of freshly distilled isopropyl glyoxylate22 (568 mg, 4.9 mmol)
in CH2Cl2 (10 mL) was added at 0 °C a solution of p-anisidine (570
mg, 4.6 mmol) in CH2Cl2 (10 mL). After 15 min, MgSO4 (3.14 g) was
added, and stirring was continued for 1 h. After completion of the
reaction, MgSO4 was filtered off with suction under an atmosphere of
argon, and the solvent was removed in vacuo to give 2a as a pale yellow
oil in quantitative yield. The imine 2a can be further purified by
distillation (Kugelrohr distillation at 110-115 °C, 0.1 mmHg).
Isopropyl N-(4-Methoxyphenylimino)acetate (2a): IR (neat) 1740,
341.1613; [R]20 +30.4 (c 0.54, CHCl3) (100% ee); HPLC analysis,
D
Daicel Chiralcel OG, hexane:iPrOH ) 9:1, 0.5 mL/min, 42 min (major,
S-enantiomer), 48 min (minor, R-enantiomer). The following conditions
are recommended for a recent lot of Chiralcel OG, because the
stationary phase of Chiralcel OG had recently been changed: hexane:
EtOH:Et2NH ) 9:1:0.1, 0.7 mL/min, 34 min (major), 36 min (minor).
N-(4-Methoxyphenyl)benzoylalanine Methyl Ester (5b): pale
yellow solid (65% ee); IR (KBr) 3358, 2361, 1742, 1680, 1514, 1238,
1
1590, 1290, 1165, 1107, 840 cm-1; H NMR (CDCl3) δ 1.38 (d, J )
6.3 Hz, 6H), 3.83 (s, 3H), 5.26 (sept, J ) 6.3 Hz, 1H), 6.92 (d, J ) 9.0
Hz, 2H), 7.35 (d, J ) 9.0 Hz, 2H), 7.91 (s, 1H); MS(EI) m/z 221(M+),
134 (bp); HRMS m/z calcd for C12H15NO3 221.1051, found 221.1057;
GC analysis, SE-30 on uniport B, 30%, 1 m, He 50 mL/min, 100 °C,
2 min, then 100-240 °C (16 °C/min), retention time ) 11.1 min.
The imines 2b and 2c were prepared in a similar manner from the
corresponding glyoxylate ester and amine. Since imines 2b and 2c were
unstable, these imines were used for the ketone-imine addition without
further purification.
1
1194, 822 cm-1; H NMR (CDCl3) δ 3.55 (d, J ) 5.3 Hz, 2H), 3.72
(s, 3H), 3.73 (s, 3H), 4.55 (br t, J ) 5.3 Hz, 1H), 6.68 (d, J ) 9.1 Hz,
2H), 6.78 (d, J ) 9.1 Hz, 2H), 7.41-7.63 (m, 3H), 7.94 (d, J ) 7.0
Hz, 2H); MS (EI) m/z 313 (M+); HRMS m/z calcd for C18H19NO4
313.1312, found 313.1304; [R]20 +26.0 (c 0.52, CHCl3) (65% ee);
D
HPLC analysis, Daicel Chiralpak AD, hexane:iPrOH ) 9:1, 1.0 mL/
min, 41 min (major), 45 min (minor).
Methyl N-(4-Methoxyphenylimino)acetate (2b): 1H NMR (CDCl3)
δ 3.84 (s, 3H), 3.95 (s, 3H), 6.94 (d, J ) 9.0 Hz, 2H), 7.37 (d, J ) 9.0
Hz, 2H), 7.96 (s, 1H); MS(EI) m/z 193 (M+), 134 (bp); GC analysis,
same conditions as for 2a, retention time ) 10.4 min.
N-(4-Phenyl)benzoylalanine Isopropyl Ester (5c): colorless oil;
1
IR (neat) 3387, 1732, 1684, 1603, 1508, 1215, 1107, 752 cm-1; H
NMR (CDCl3) δ 1.15 (d, J ) 6.3 Hz, 3H), 1.21 (d, J ) 6.3 Hz, 3H),
3.55 (d, J ) 5.3 Hz, 2H), 4.58 (br t, J ) 5.3 Hz, 1H), 5.05 (sept, J )
6.3 Hz, 1H), 6.66-6.79 (m, 3H), 7.14-7.24 (m, 2H), 7.41-7.63 (m,
3H), 7.94 (d, J ) 7.0 Hz, 2H); MS (EI) m/z 311 (M+); HRMS m/z
calcd for C19H21NO3 311.1520, found 311.1512; [R]20D +24.9 (c 0.76,
CHCl3) (64% ee); HPLC analysis, Daicel Chiralpak AD, hexane:iPrOH
) 9:1, 1.0 mL/min, 19 min (major), 21 min (minor).
Reaction Catalyzed by Tetrafluoroboric Acid. To a solution of
chlorotrimethylsilane (63 µL, 0.50 mmol) in DMF (5 mL) was added
AgBF4 (97 mg, 0.5 mmol). The mixture was stirred at 25 °C for 2 h,
and the supernatant solution was used as 0.1 M HBF4 solution. To a
solution of imine 2a (22 mg, 0.1 mmol) in DMF (0.5 mL) were added
1a (31 µL, 0.15 mmol) and 0.1 M HBF4 solution (50 µL, 0.005 mmol).
The mixture was stirred at 25 °C for 3 h, diluted with water, and
extracted with Et2O. The organic extracts were dried over Na2SO4 and
concentrated. The residue was purified by preparative thin-layer
chromatography (AcOEt/hexane ) 1/4) to afford 5a (25.9 mg, 76%).
Preparation of Complexes 4a-d. The µ-hydroxo complexes 4a-d
can be prepared either from 3a-d, or 8a and 8b. The representative
procedures are as follows.
Isopropyl N-(4-Phenylimino)acetate (2c): 1H NMR (CDCl3) δ 1.40
(d, J ) 6.3 Hz, 6H), 5.26 (sept, J ) 6.3 Hz, 1H), 7.10-7.50 (m, 5H),
7.89 (s, 1H); MS(EI) m/z 191 (M+), 104 (bp); GC analysis, same
conditions as for 2a, retention time ) 9.0 min.
Preparation of Palladium Aqua Complexes. Preparation of 3a and
3b was already reported.7b Palladium complexes 3c and 3d were also
prepared by a similar method.
[Pd((R)-tol-binap)(H2O)2]2+(TfO-)2 (3d). To a solution of 8b (110
mg, 0.128 mmol) in acetone (0.5% v/v H2O) was added AgOTf (66
mg, 0.256 mmol) at 25 °C. The reaction mixture was stirred at 25 °C
for 5 h. The precipitated AgCl was filtered off on Celite, and the filtrate
was evaporated under reduced pressure. The product was recrystallized
from CH2Cl2-ether to afford 3d (96.0 mg, 71%): yellow powder; mp
135 °C dec; IR (KBr) 3250, 1280, 1030, 630 cm-1; 1H NMR (CDCl3,
200 MHz) δ 2.05 (s, 6H), 2.42 (s, 6H), 3.75 (br s, >4H, coordinated
(21) Tietze, L. F.; Bratz, M. Synthesis 1989, 439-442.
(22) Kelly, T. R.; Schmid, T. E.; Haggerty, J. G. Synthesis 1972, 544-
545.

5458 J. Am. Chem. Soc., Vol. 121, No. 23, 1999
Fujii et al.
Preparation of [{Pd((R)-binap)(µ-OH)}2]2+(BF4-)2 (4a) from 8a.
To a solution of 8a (100 mg, 0.125 mmol) in wet CH2Cl2 (10 mL)
were added 4-Å molecular sieves (1.5 g) and AgBF4 (48 mg, 0.25
mmol). The mixture was stirred at 25 °C for 30 h and filtered through
a Celite pad. The filtrate was concentrated to give 4a (86 mg, 79%):
reddish orange prisms; mp 217 °C dec; IR (CHCl3) 3580, 3050, 1430,
needle was maintained at 2000-3000 V, and the orifice was at 20-60
V. The scan range was from m/z 200 to 3000. The mass spectra of 3d,
4d, and 9 were shown in Figure 5.
Representative Procedure for Table 2. To a solution of 4b (8.9
mg, 0.005 mmol) in DMF (1.0 mL) was added 1a (41 µL, 0.20 mmol)
at 25 °C under an atmosphere of argon. After the mixture was stirred
at 25 °C for 40 min, a solution of imine 2a (22.0 mg, 0.10 mmol) in
DMF (0.5 mL) was added. The whole reaction mixture was stirred at
25 °C for overnight, then diluted with ether and filtered through a short
silica gel column. The solution was poured into water and extracted
with ether (2 × 5 mL). The combined organic layers were washed
with water and brine, dried over Na2SO4, and concentrated. The residue
was purified by preparative silica gel thin-layer chromatography
(AcOEt/hexane ) 1/4) to afford 5a (32.3 mg, 95%). The ee of 5a was
determined to be 90% by HPLC analysis using Daicel Chiralcel OG
(hexane:iPrOH ) 9:1). This sample (28 mg) was recrystallized from
ethyl acetate-hexane to give an optically pure sample (21 mg, 75%
recovery).
One-Pot Reaction. To a solution of freshly distilled isopropyl
glyoxylate (23 mg, 0.20 mmol) in DMF (2 mL) were added p-anisidine
(26.0 mg, 0.20 mmol) and 4-Å molecular sieves (100 mg). After this
mixture was stirred at 25 °C for 3 h, 1a (60 µL, 0.30 mmol) and 4d
(19.0 mg, 10.0 µmol) were added. The whole reaction mixture was
stirred at 25 °C for 40 h, diluted with ether, and filtered through a
short silica gel column. The filtrate was poured into water and extracted
with ether. The combined organic layers were washed with water and
brine, dried over Na2SO4, and concentrated. The residue was purified
by preparative silica gel thin-layer chromatography (AcOEt/hexane )
1/4) to afford 5a (51.0 mg, 74%). The ee of 5a was determined to be
83% by HPLC analysis using Daicel Chiralcel OG.
1
1310, 1050, 810, 740 cm-1; H NMR (CDCl3, 200 MHz) δ -2.90 (s,
2H), 6.55 (d, J ) 8.4 Hz, 4H), 6.60-7.75 (m, 60H); [R]20D +735.6 (c
0.32 CHCl3).
Preparation of [{Pd((R)-tol-binap)(µ-OH)}2]2+(BF4-)2 (4b) from
3b Using 4-Å Molecular Sieves. To a solution of 3b (50 mg, 0.05
mmol) in acetone (5 mL) was added dried 4-Å molecular sieves
(powder, 300 mg). The mixture was stirred at 25 °C for 3 h. After
filtration through a Celite pad, the filtrate was concentrated to give 4b
(42 mg, 88%): reddish orange prisms; mp 247 °C dec; IR (KBr) 3480,
1
1260, 1030, 810, 510 cm-1; H NMR (CDCl3) δ -3.03 (s, 2H), 1.96
(s, 12H), 2.11 (s, 12H), 6.53 (d, J ) 8.6 Hz, 4H), 6.60-7.75 (m, 52H);
[R]20 +590.3 (c 0.2, CHCl3).
D
[{Pd((R)-tol-binap)(µ-OH)}2]2+(TfO-)2 (4c): reddish orange prisms;
mp 195 °C dec; IR (KBr) 3500, 3060, 1430, 1260, 1030, 740, cm-1
;
1H NMR (CDCl3, 200 MHz) δ -2.86 (s, 2H), 6.56 (d, J ) 8.4 Hz,
4H), 6.80-7.67 (m, 60H); [R]20 +683.6 (c 0.5 CHCl3).
D
Preparation of [{Pd((R)-tol-binap)(µ-OH)}2]2+(TfO-)2 (4d) from
3d Using NaOH. To a solution of 3d (235 mg, 0.21 mmol) in CH2Cl2
(3.0 mL) were added water (1.0 mL) and 0.1 N NaOH (2.1 mL). The
mixture was stirred at 25 °C for 4 h. The organic phase was separated,
washed with water, and dried over Na2SO4. The solvent was removed
in vacuo to give 4d (173 mg, 87%) as a red-orange solid. Recrystal-
lization from CH2Cl2-Et2O afforded yellow crystals: mp 228 °C dec;
IR (CHCl3) 3620, 2950, 1270, 1010, 740 cm-1; 1H NMR (CDCl3, 200
MHz) δ -3.00 (s, 2H), 1.96 (s, 12H), 2.11 (s, 12H), 6.53 (d, J ) 8.6
NMR and ESI-MS Experiment for the Reaction of 4c and 1a.
To an NMR tube containing a solution of palladium complex 4c (18.0
mg, 0.01 mmol) in DMF-d7 (0.6 mL) was added enol silyl ether 1a
Hz, 4H), 6.60-7.75 (m, 52H); [R]20 +628.6 (c 0.3 CHCl3).
D
Single-Crystal X-ray Analysis Data for 4d. The molecular structure
for 4d determined by single-crystal X-ray analysis is shown in Figure
3. A yellow prism crystal of C98H82O8P4S2F6Pd2, having approximate
dimensions of 0.40 × 0.20 × 0.12 mm, was mounted on a glass fiber.
All measurements were made on a Rigaku RAXIS-II imaging plate
area detector with graphite-monochromated Mo KR radiation. A single
crystal of 4d was orthorhombic, having a space group P212121, a )
14.577(6) Å, b ) 16.488(5) Å, c ) 41.32(2) Å, V ) 9931(14) Å3, Z )
4. The final R factor was 0.104 for 6892 measured reflections. Selected
distances (Å) and angles (deg) are as follow: Pd(1)-P(1) 2.242(5),
Pd(1)-P(2) 2.231(7), Pd(1)-O(1) 2.04(2), Pd(1)-O(2) 2.12(1), Pd-
(2)-P(3) 2.219(6), Pd(2)-P(4) 2.243(5), Pd(2)-O(1) 2.07(1), Pd(2)-
O(2) 2.11(1), P(1)-Pd(1)-P(2) 93.1(2), P(1)-Pd(1)-O(2) 169.6(4),
P(2)-Pd(1)-O(1) 174.1(6), O(1)-Pd(1)-O(2) 80.0(6), P(3)-Pd(2)-
P(4) 93.0(2), P(3)-Pd(2)-O(2) 170.2(4), P(4)-Pd(2)-O(1) 172.1(6),
O(1)-Pd(2)-O(2) 79.5(5). Details associated with data collection, data
reduction, and structure solution and refinement are given in the
Supporting Information.
1
(2.1 µL, 0.01 mmol) at 25 °C. After 40 min at 25 °C, H NMR was
measured (Figure 4b), and a 5-µL sample was withdrawn from the
solution using a microsyringe, diluted with THF (5 mL), and analyzed
by ESI-MS as described above (Figure 5b). To the reaction mixture
were added 0.5 equiv (1.0 µL, 0.005 mmol) and 3.0 equiv (3.2 µL,
1
0.015 mmol) of 1a after 40 min and 1.5 h. After 30 min at 25 °C, H
NMR (Figure 4c) and ESI-MS (Figure 5c) were measured according
to the same manner as above.
NMR and ESI-MS Experiment for the Reaction of 4c and 13.
To an NMR tube containing a solution of palladium complex 4c (18.0
mg, 0.01 mmol) in DMF-d7 (0.6 mL) was added enol silyl ether 14
1
(2.3 mg, 0.012 mmol). After 10 h, H NMR was measured (Figure
6a), and a 5-µL sample was withdrawn from the solution using a
microsyringe, diluted with wet acetone (5 mL), and analyzed by ESI-
MS as described above (Figure 6b).
Acknowledgment. We thank Professor Masako Nakagawa
(Chiba University) for the X-ray structural analysis. We also
thank Dr. Kazuo Kamemura (JST) for technical support for the
ESI-MS analysis.
Preparation of 9. To a solution of 3d (35 mg, 0.031 mmol) in
acetone was added 4-Å molecular sieves (40 mg). The mixture was
stirred at 25 °C for 5 h. After filtration through a Celite pad, the filtrate
was concentrated to give 9 (24.1 mg, 81%): orange powder; mp 191
°C dec; IR (KBr) 3480, 1260, 815, cm-1; 1H NMR (200 MHz) δ -3.07
(s, 1H), 1.82 (s, 6H), 2.06 (s, 6H), 2.14 (s, 6H), 2.24 (s, 6H), 6.56 (d,
Supporting Information Available: Effects of preincuba-
tion time on the asymmetric induction, 1H NMR spectra of 2a,
3c, 3d, 4a, 4d, and 9, ESI-MS data of 3d, 4d, and 9, X-ray
J ) 8.6 Hz, 4H), 6.66 (d, J ) 8.6 Hz, 4H), 7.01-7.75 (m, 48H); [R]20
+731.3 (c 0.23 CHCl3).
D
1
Mass Spectrometric Investigation of Isolated Palladium Com-
plexes 3d, 4d, and 9 by Electrospray Ionization Mass Spectrometry
(ESI-MS). For mass spectra of palladium complexes, a 10-20 µM
solution of 3d, 4d, or 9 in THF was infused into a PE SCIEX API 300
mass spectrometer at 0.5 µL/min using a syringe pump. The ionspray
structural information on 4d, and H NMR and ESI-MS data
on the reactions of 4d with 1a and 13 (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA9902827