Access to enantioenriched α-amino esters via rhodium-catalyzed 1,4-addition/enantioselective protonation

Article abstract of DOI:10.1021/ja710691p

Conjugate addition of potassium trifluoro(organo)borates 2 to dehydroalanine derivatives 1, mediated by a chiral rhodium catalyst and in situ enantioselective protonation, afforded straightforward access to a variety of protected α-amino esters 3 with high yields and enantiomeric excesses up to 95%. Among the tested chiral ligands and proton sources, Binap, in combination with guaiacol (2-methoxyphenol), an inexpensive and nontoxic phenol, afforded the highest asymmetric inductions. Organostannanes have also shown to participate in this reaction. By a fine-tuning of the ester moiety, and using Difluorophos as chiral ligand, increased levels of enantioselectivity, generally close to 95%, were achieved. Deuterium labeling experiments revealed, and DFT calculation supported, an unusual mechanism involving a hydride transfer from the amido substituent to the α carbon explaining the high levels of enantioselectivity attained in controlling this α chiral center.

Full text of DOI:10.1021/ja710691p

Published on Web 04/09/2008
Access to Enantioenriched r-Amino Esters via
Rhodium-Catalyzed 1,4-Addition/Enantioselective Protonation
Laure Navarre, Rémi Martinez, Jean-Pierre Genet,* and Sylvain Darses*
Laboratoire de Synthèse SélectiVe Organique (UMR 7573, CNRS), Ecole Nationale Supérieure
de Chimie de Paris, 11 rue P&M Curie, 75231 Paris cedex 05, France
Received November 29, 2007; E-mail: sylvain-darses@enscp.fr; jean-pierre-genet@enscp.fr
Abstract: Conjugate addition of potassium trifluoro(organo)borates 2 to dehydroalanine derivatives 1,
mediated by a chiral rhodium catalyst and in situ enantioselective protonation, afforded straightforward
access to a variety of protected R-amino esters 3 with high yields and enantiomeric excesses up to 95%.
Among the tested chiral ligands and proton sources, Binap, in combination with guaiacol (2-methoxyphenol),
an inexpensive and nontoxic phenol, afforded the highest asymmetric inductions. Organostannanes have
also shown to participate in this reaction. By a fine-tuning of the ester moiety, and using Difluorophos as
chiral ligand, increased levels of enantioselectivity, generally close to 95%, were achieved. Deuterium labeling
experiments revealed, and DFT calculation supported, an unusual mechanism involving a hydride transfer
from the amido substituent to the R carbon explaining the high levels of enantioselectivity attained in
controlling this R chiral center.
describe.3,4 On the other hand, very few examples have been
described concerning catalytic enantioselective conjugate ad-
Introduction
Chiral R-amino acids constitute one of the most important
building bocks in organic synthesis, and this structural element
is encountered in many biologically active compounds. Several
methods have been developed to access nonproteogenic amino
acids in the past century, but approaches with stereocontrol of
the R chiral center are more limited. Nevertheless, a variety of
methods, which are quite general, are available for the synthesis
of chiral R-amino acids.1 These include the Strecker and Ugi
condensations, the enzymatic resolution of racemic R-amino
acids, glycine anion functionalization using chiral phase-transfer
catalysis, and the asymmetric hydrogenation of dehydroamino
esters.
An even more straightforward method would consist of a
Michael type addition of an organometallic reagent to a
dehydroalanine derivative with concomitant control of the R
chiral center. However, such an approach has been scarcely
developed in the literature. Conjugate addition of organome-
tallics to R-amino acrylates followed by diastereoselective
protonation of the resulting enolate has been described to some
extent, but the reaction scope was rather limited.2 Similar
reactions, but involving radical conjugate additions, were also
ditions. Sibi et al. reported in 2001 an enantioselective radical
conjugate addition to R-amino acrylates5 and R-methacrylate6
followed by hydrogen atom transfer catalyzed by chiral Lewis
acids. ꢀ-Substituted R-amino acids were also synthesized using
this methodology.7
We recently described preliminary results concerning a new
transition metal catalyzed tandem so-called 1,4-addition/enan-
tioselective protonation of organometallics to dehydroamino
esters that allowed direct preparation of R-amino acid derivatives
in high yields and enantiomeric excesses up to 90%.8 The initial
idea, which initiated this research, was based on the postulated
mechanism of the 1,4-addition of organometallic reagents to
R,ꢀ-unsaturated substrates.9 Several organometallic reagents and
particularly organoboronic acids have been shown to participate
in rhodium-catalyzed Michael additions, allowing the introduc-
tion of a chiral center in the ꢀ position of an electron-
(3) (a) Review: Srikanth, G. S. C.; Castle, S. L. Tetrahedron 2005, 61,
10377–10441.
(4) (a) Beckwith, A. L. J.; Chai, C. L. L. J. Chem. Soc, Chem. Comm.
1990, 1087. (b) Gasanov, R. G.; Il’inskaya, L. V.; Misharin, M. A.;
Maleev, V. I.; Raevski, N. I.; Ikonnikov, N. S.; Orlova, S. A.; Kuzmina,
N. A.; Belokon, Y. N. J. Chem. Soc., Perkin Trans. 1 1994, 3343. (c)
Axon, J. R.; Beckwith, A. L. J. J. Chem. Soc., Chem. Commun. 1995,
549. (d) Jones, R. C. F.; Berthelot, D. J. C.; Iley, J. N. Tetrahedron
2001, 57, 6539. (e) Kabat, M. M. Tetrahedron Lett. 2001, 42, 7521.
(5) Sibi, M. P.; Asano, Y.; Sausker, J. B. Angew. Chem., Int. Ed. 2001,
40, 1293.
(1) (a) For reviews on the preparation of optically active R-amino acids,
see: Williams, R. M. In Synthesis of Optically ActiVe R-Amino Acids;
Baldwin, J. E., Ed.; Organic Chemistry series; Pergamon Press: Oxford,
1989. (b) Williams, R. M.; Hendrix, J. A. Chem. ReV. 1992, 92, 889.
(c) Duthaler, R. O. Tetrahedron 1994, 50, 1539. (d) Seebach, D.; Sting,
A. R.; Hoffmann, M. Angew. Chem., Int. Ed. Engl. 1996, 35, 2708.
(e) Cativiela, C.; Diaz-de-Villegas, M. D. Tetrahedron: Asymmetry
1998, 9, 3417. (f) Maruoka, K.; Ooi, T. Chem. ReV. 2003, 103, 3013.
(g) Najera, C.; Sansano, J. M. Chem. ReV. 2007, 107, 4584.
(2) (a) Cardellicchio, C.; Fiandanese, V.; Marchese, G.; Naso, F.; Ronzini,
L. Tetrahedron Lett. 1985, 26, 4387. (b) Cativiela, C.; Diaz-de-
Villegas, M. D.; Galvez, J. A. Can. J. Chem. 1992, 70, 2325. (c)
Lander, P. A.; Hegedus, L. S. J. Am. Chem. Soc. 1994, 116, 8126. (d)
Cardillo, G.; Gentilucci, L.; Tolomelli, A.; Tomasini, C. Tetrahedron
1999, 55, 6231.
(6) Sibi, M. P.; Sausker, J. B. J. Am. Chem. Soc. 2002, 124, 984.
(7) He, L.; Srikanth, G. S. C.; Castle, S. L. J. Org. Chem. 2005, 70, 8140.
(8) (a) For rhodium-catalyzed tandem 1,4-addition-enantioselective
protonation: Navarre, L.; Darses, S.; Genet, J.-P. Angew. Chem., Int.
Ed. 2004, 43, 719. For a racemic version: (b) Navarre, L.; Darses, S.;
Genet, J.-P. Eur. J. Org. Chem. 2004, 69.
(9) For an investigation of rhodium-catalyzed 1,4-addition mechanism,
see: Hayashi, T.; Takahashi, M.; Takaya, Y.; Ogasawara, M. J. Am.
Chem. Soc. 2002, 124, 5052.
9
10.1021/ja710691p CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 6159–6169 6159

A R T I C L E S
Navarre et al.
Table 1. Phenols as Protonating Agentsa
withdrawing group.10,11 We also shown that potassium organ-
otrifluoroborates participated efficiently in this reaction,12 with
their advantageous stability and ease of preparation compared
to other boron derivatives.13 In such reactions, it is postulated,
and some intermediates have identified, that the mechanistic
cycle involves the generation of an oxa-π-allylmetal intermedi-
ate.14 Thus, one could easily imagine control of the prochiral R
center of an R,R′-disubstituted alkene via a diastereoselective
protonation of the putative rhodium-enolate, leading to an overall
tandem 1,4-addition/enantioselective protonation.15,16
Other groups reported such an approach using water as a
protonating agent, but enantiomeric excesses were limited to
70%. Indeed, Reetz et al.17 were the first to show that, in the
presence of an atropoisomeric binol-based diphosphonite chiral
ligand, the rhodium-catalyzed 1,4-addition of phenylboronic acid
to dehydroamino ester afforded R-amino ester in a modest 77%
ee. Frost et al.18 also described the use of a hindered chiral ligand
in the same reaction, as well as binap ligand in the preparation
of 2-substituted succinic esters,19 but the ee values were still
limited to 72%.
phenol
yield
eec
phenol
yield
eec
C6H5
2-MeC6H4
2-t-BuC6H4
2,6-(t-Bu)2C6H3
2-PhC6H4
81
63
28
63
77
90
88
51
42
0
18
2-MeOC6H4
2-i-PrOC6H4
2-HOC6H4
3-MeOC6H4
4-MeOC6H4
2-MeO-6-IC6H3
2-FC6H4
91
91
4
84
75
0
65
45
0
85
76
72
80
73
83
63
7
51
45
37
44
54
69
8
-8
-20
14
4-ClC6H4
2,4-Cl2C6H3
2,4,6-Cl3C6H2
4-CNC6H4
2-CNC6H4
2-NO2C6H4
2-(AcNH)C6H4
2-(PhCONH)C6H4 18
2-(MeOCO)C6H4
2-(MeCO)C6H4
30
-4
2-CF3C6H4
2-MeSC6H4
3-(MeCO)C6H4
3,5-(MeO)2C6H3
3,4,5-(MeO)3C6H2
P1d
45
57
59
61
62
23b nd
23
3
18
26
16
36
32
P2d
a Reactions conducted using 0.5 mmol of 1a, 2 equiv of 2a, and 1
equiv of phenol with 3 mol% of [Rh(cod)2][PF6], 3.3 mol% of (S)-binap
in toluene at 110–115 °C for 20 h. b Conversion determined by GC.
c Sign plus for the (R) enantiomer. d P1 ) benzo[1,3]dioxol-5-ol and P2
) 7-methoxy-naphthalen-2-ol.
From these results, it appeared that water was not the most
suitable protonating agent. On the contrary, when using other
proton sources such as phenol derivatives, we showed that high
ee values were achieved in the 1,4-addition of organometallics
to dehydroamino esters.8a This concept of 1,4-addition/enanti-
oselective protonation of rhodium enolate using other proton
sources than water was applied to the synthesis of ꢀ2-amino
acids using phthalimide as a protonating agent20 and R,R′-
dibenzyl esters using boric acid.21
We want to report here the scope and limitation of the
rhodium-catalyzed 1,4-addition/enantioselective protonation of
organometallic reagents to dehydroamino esters allowing direct
preparation of R-amino acid derivatives in high yields and
enantiomeric excesses. Moreover, mechanistic studies revealed
a totally unusual reaction mechanism, explaining the efficient
stereocontrol of the R chiral center on such substrates and
resulting in further improvements in the efficiency of the
process.
Catalytic System Optimization. We initially tested the fea-
sibility of the asymmetric 1,4-addition to dehydroamino esters
using methyl N-acetylaminoacrylate (1a) and potassium phe-
nyltrifluoroborate (2a) as model substrates (eq 1). Despite
numerous attempts, we never succeeded in achieving stereose-
lective control of the R center on protonation with water in the
presence of various chiral ligands complexed on a rhodium
catalyst. Whatever the chiral ligand tested, enantioselectivities
were generally lower than 30% and the results were not always
reproducible.22
(10) (a) Sakai, M.; Hayashi, H.; Miyaura, N. Organometallics 1997, 16,
4229. Reviews: (b) Hayashi, T. Synlett 2001, 879. (c) Hayashi, T.;
Yamasaki, K. Chem. ReV. 2003, 103, 2829. (d) Hayashi, T. Bull. Chem.
Soc. Jpn. 2004, 77, 13. (e) Hayashi, T. Pure Appl. Chem. 2004, 76,
465.
(11) (a) Itooka, R.; Iguchi, Y.; Miyaura, N. J. Org. Chem. 2003, 68, 6000.
(b) Hayashi, T.; Ueyama, K.; Tokugana, N.; Yoshida, K. J. Am. Chem.
Soc. 2003, 125, 11508. (c) Ma, Y.; Song, C.; Ma, C.; Sun, Z.; Chai,
Q.; Andrus, M. B. Angew. Chem., Int. Ed. 2003, 42, 5871. (d) Defieber,
C.; Paquin, J.-F.; Serna, S.; Carreira, E. M. Org. Lett. 2004, 6, 3873.
(e) Bocknack, B. M.; Wang, L.-C.; Krische, M. J. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 5421. (f) Shintani, R.; Tsurusaki, A.; Okamoto,
K.; Hayashi, T. Angew. Chem., Int. Ed. 2005, 44, 3909. (g) Duan,
W.-L.; Imazaki, Y.; Shintani, R.; Hayashi, T. Tetrahedron 2007, 63,
8529. (h) Shintani, R.; Ichikawa, Y.; Hayashi, T.; Chen, J.; Nakao,
Y.; Hiyama, T. Org. Lett. 2007, 9, 4643. (i) Helbig, S.; Sauer, S.;
Cramer, N.; Laschat, S.; Baro, A.; Frey, W. AdV. Synth. Catal. 2007,
349, 2331. (j) Tokunaga, N.; Hayashi, T. AdV. Synth. Catal. 2007,
349, 513.
(12) (a) Pucheault, M.; Darses, S.; Genet, J.-P. Tetrahedron Lett. 2002,
43, 6155. (b) Pucheault, M.; Darses, S.; Genet, J.-P. Eur. J. Org. Chem.
2002, 3552. (c) Pucheault, M.; Michaud, V.; Darses, S.; Genet, J.-P.
Tetrahedron Lett. 2004, 45, 4729. (d) Navarre, L.; Darses, S.; Genet,
J.-P. Chem. Commun. 2004, 1108. (e) Pucheault, M.; Darses, S.; Genet,
J.-P. J. Am. Chem. Soc. 2004, 126, 15356. (f) Navarre, L.; Pucheault,
M.; Darses, S.; Genet, J.-P. Tetrahedron Lett. 2005, 46, 4247. (g)
Navarre, L.; Darses, S.; Genet, J.-P. AdV. Synth. Catal. 2006, 348,
317. (h) Martinez, R.; Voica, F.; Genet, J.-P.; Darses, S. Org. Lett.
2007, 9, 3213.
(13) For reviews on potassium trifluoro(organo)borates chemistry, see: (a)
Darses, S.; Genet, J.-P. Eur. J. Org. Chem. 2003, 4313. (b) Molander,
G. A.; Figueroa, R. Aldrichimica Acta 2005, 38, 49. (c) Molander,
G. A.; Ellis, N. Acc. Chem. Res. 2007, 40, 275. (d) Stefani, H. A.;
Cella, R.; Vieira, A. S. Tetrahedron 2007, 63, 3623. (e) Darses, S.;
Genet, J.-P. Chem. ReV. 2008, 108, 288.
(14) Such complexes have been shown to be intermediates in rhodium-
catalyzed 1,4-additons of organometallic reagent (particularly orga-
noboronic acids) to enones; see ref 9.
(15) For a first example of enantioselective protonation of a rhodium enolate
intermediate (18% ee), see: Bergens, S. H.; Bosnich, B. J. Am. Chem.
Soc. 1991, 113, 958.
(16) (a) For enantioselective protonation of enolates see: Duhamel, L.;
Plaquevent, J.-C J. Am. Chem. Soc. 1978, 100, 7415. Reviews: (b)
Duhamel, L.; Duhamel, P.; Launay, J.-C.; Plaquevent, J.-C. Bull. Soc.
Chim. Fr. 1984, 421. (c) Fehr, C. Angew. Chem., Int. Ed. Engl. 1996,
36, 2566. (d) Yanagisawa, A.; Ishihara, K.; Yamamoto, H. Synlett
1997, 411. (e) Eames, J.; Weerasooriya, N. Tetrahedron: Asymmetry
2001, 12, 1. (f) Duhamel, L.; Duhamel, P.; Plaquevent, J.-C.
Tetrahedron: Asymmetry 2004, 15, 3653.
Then, we turned out our attention to the proton source. Among
the various tested proton sources, we were pleased to find that
phenols were particularly suited. For example, using phenol,
an 81% yield of 3aa was obtained in 18% ee (Table 1). In order
(17) Reetz, M. T.; Moulin, D.; Gosberg, A. Org. Lett. 2001, 3, 4083.
(18) Chapman, C. J.; Wadsworth, K. J.; Frost, C. G. J. Organomet. Chem.
2003, 680, 206.
(20) Sibi, M. P.; Tatamidani, H.; Patil, K. Org. Lett. 2005, 7, 2571.
(21) Frost, C. G.; Penrose, S. D.; Lambshead, K.; Raithby, P. R.; Warren,
J. E.; Gleave, R. Org. Lett. 2007, 9, 2119.
(19) Moss, R. J.; Wadsworth, K. J.; Chapman, C. J.; Frost, C. G. Chem.
Commun. 2004, 1984.
(22) See supporting information for some examples of the use of chiral
ligands using water as a proton source.
9
6160 J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008

Access to Enantioenriched R-Amino Esters
A R T I C L E S
to attain useful levels of enantioselectivities, a screening of
phenols was undertaken. Among all the tested phenols, few of
them allowed reaching ee values higher than 60% and the best
ones generally possessed a substituent in the ortho position of
the hydroxyl (Table 1).
From a general point of view, it appeared that the electronic
nature and the steric hindrance of the phenol substituents does
have a major influence on both yields and enantioselectivities.
Phenols bearing an electron-withdrawing substituent on the
aromatic ring (Cl, CN, NO2, COMe, CO2Me) generally afforded
low enantioselectivities in the reaction of N-acetylaminoacrylate
(1a) with potassium phenyltrifluoroborate (2a). Moreover, in
the case of chlorophenol derivatives, increasing the number of
substituents in the ortho and para positions resulted in a reverse,
even moderate, enantioselection (from 18% ee for phenol to
–20% ee for 2,4,6-trichlorophenol), suggesting a change in the
reaction mechanism, and particularly the protonation step. The
presence of potential complexing substituents in the ortho
position of the phenol (NHAc, NHCOPh, CO2Me, COR, F)
seemed to inhibit the reaction, and a low yield and ee were
observed in each case. Moreover, it appeared that increasing
the steric hindrance in the ortho position resulted in an increase
in the enantioselectivity, which reaches 54 and 69% for 2,6-
di-tert-butylphenol and 2-phenylphenol, respectively. Generally
speaking, the highest yields and enantioselectivities were
obtained with electron-rich phenols, bearing an alkoxo sub-
stituent. However, with meta- or para-alkoxo-substituted phe-
nols, moderate enantioselectivities, in the range 50–60%, were
generally achieved, while the yields were acceptable (close to
80%). The highest level of enantioselectivity was attained using
2-methoxyphenol or guaiacol (83% ee).
In order to understand more closely the influence of the
phenols on the enantioselectivities, we wondered if the pKA of
the phenols was influencing the level of enantioselectivities. As
all the pKA values given in the literature are evaluated in a protic
solvent, more generally in water, and given that the reaction is
conducted under aprotic conditions (toluene as solvent), we
evaluated the pKA of several phenols by DFT calculation (Figure
1).23 From these calculations it appeared that there is not a
complete correlation between phenol pKA’s and the measured
enantioselectivities, probably because many other factors can
influence the protonation step, such as steric or electronic factors.
But, the general tendency is that an increase of the pKA results
in an increase of the enantioselectivity; in other words, the less
acidic phenols are the most suitable proton sources in the
reaction. More particularly guaiacol, which gave the highest
enantioselectivity, is the least acidic phenol we have evaluated,
under nonprotic conditions. The influence of the phenol pKA’s
will be discussed in connection with the reaction mechanism
(Vide infra).
a
Figure 1. Influence of phenols pKA on the enantioselectivities. P1 )
benzo[1,3]dioxol-5-ol and P2 ) 7-methoxy-naphthalen-2-ol.
Table 2. Ligand Screening in the 1,4-Addition/Enantioselective
Protonationa
entry
chiral ligand
yieldd
eee
1
2
3
4
(S)-Binap (L1)
91 (89b)
83
83 (89.5b)
(S)-TolBinap (L2)
(S)-MeOBiphep (L3)
(S)-L4
82
48 (82b)
32b
32
58 (84b)
85b
5
(R)-L5
-2
6
7
8
9
(S)-L6
(R)-L7
(R)-L8
(R)-L9
20
7
51b
50b
63
-88.2b
-88.9b
0
10
11
12
13
14
15
16
(R)-(S)-PPF-PCy2 (L10)
(R)-(S)-Cy2PF-PCy2 (L11)
(R)-(S)-PPF-Pt-Bu2 (L12)
(S)-(R)-Cy2PF-PPh2 (L13)
(R)-Binepine (L14)
(R,R)-MeDuphos (L15)
(S)-Synphos (L16)
9
5
7
3
nd
0
nd
2
6
58c
68
0
12b
79b
a Reactions conducted using 0.5 mmol of 1a, 2 equiv of 2a, and 1
equiv of guaiacol with 3 mol% of [Rh(cod)2][PF6], 3.3 mol% of chiral
bidentate ligand in toluene at 110–115 °C for 20 h. b Reactions
conducted with 6.6 mol% of chiral bidentate ligand. c Reactions
conducted with 6.6 mol% of ligand. d Isolated yields. e Determined by
HPLC analysis using Daicel Chiralcel OD-H chiral column, sign plus
for the (R) enantiomer.
low or the absence of chiral induction was observed using
Josiphos chiral ligand derivatives L10–L1324 (entries 10–13),
Binepine L1425 (entry 14), or Duphos ligand L1526 (entry 15)
which have been shown to be useful ligands in asymmetric
hydrogenation of dehydroamino esters. At this point, it is
important to note that the amount of chiral ligand compared to
rhodium does have a notable influence on the stereochemical
outcome of the reaction. Indeed, by using 2.2 equiv of chiral
bidentate ligand instead of 1.1 equiv (compared to Rh catalyst),
we were able to increase the ee by more than 6 points and an
Indeed, given that guaiacol is an inexpensive and nontoxic
phenol, we searched no further for a proton source and
envisioned exploring the influence of the chiral ligand at the
rhodium center (Table 2 and Chart 1).
(24) (a) Togni, A.; Breutel, C.; Schnyder, A.; Spindler, F.; Landert, H.;
Tijani, A. J. Am. Chem. Soc. 1994, 116, 4062. (b) Togni, A. Angew.
Chem., Int. Ed. Engl. 1996, 35, 1475.
Among the tested chiral ligands, only atropoisomeric ligands
allowed reaching high levels of enantioselectivities. Indeed, very
(25) Gladiali, S.; Dore, A.; Fabbri, D.; De Lucchi, O.; Manassero, M.
Tetrahedron: Asymmetry 1994, 5, 511. (b) Junge, K.; Oehme, G.;
Monsees, A.; Riermeier, T.; Dingerdissen, U.; Beller, M. J. Organomet.
Chem. 2003, 675, 91. (c) Junge, K.; Oehme, G.; Monsees, A.;
Riermeier, T.; Dingerdissen, U.; Beller, M. Tetrahedron Lett. 2002,
43, 4977.
(23) Free energies of the phenols GVac(ArOH) and phenates GVac(ArO-)
were calculated at the B3LYP/6–31++G(d,p) level, at 298.15 K, 1
atm, and the results were converted to pKA values in water using the
following linear formula: pKA (ArOH) ) a × pKA (ArOH) + b
H
vac
with a ) 14/[pKA (H2O/HO-)2O- pKA (H3O+/H2O)]; b ) 14/[1 -
vac
vac
{pKA (H2O/HO-)}/{pKA (H3O+/H2O)}] and pKA (ArOH)
)
(26) (a) Burk, M. J. J. Am. Chem. Soc. 1991, 113, 8518. (b) Burk, M. J.;
Feaster, J. E.; Nugent, W. A.; Harlow, R. L. J. Am. Chem. Soc. 1993,
117, 9375.
-log(exp[{Gvac(ArO-) - Gvac(ArOH)}/RT]): see Suppovratcing Informa-
vac
vac
tion.
9
J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008 6161

A R T I C L E S
Navarre et al.
Chart 1. Chiral Ligands in the 1,4-Addition/Enantioselective Protonation
ee of nearly 90% was obtained using Binap27 (entry 1). The
same trend was also observed with MeOBiphep ligand (L3)28
(entry 2).
recycling. As a model of chiral phenol, binol was chosen
because of its ready availability and because it has shown to be
a useful chiral inductor (eq 2).31 Thus, we conducted the reaction
using (R)-Binol as the proton source, in conjunction with either
(R) or (S)-Binap as the chiral ligand in order to avoid match/
mismatch associations. Unfortunately, on the reaction model,
the same low ee values were obtained with both of the
atropoisomeric chiral ligands.
Various substitution patterns were evaluated either on the
backbone of the atropoisomeric ligand or on the phenyl
phosphorus substituents, and it appeared that ligands such as
Binap (entry 1), TolBinap (entry 2), MeOBiphep derivatives29
(entries 3–9), or Synphos30 (entry 16) gave the highest ee on
the model substrates. From these results, it is highly speculative
to find any correlation between the electronic nature of the
atropoisomeric ligand and the level of asymmetric induction
but, in this series, ee generally decreased with increasing electron
density on the chiral backbone (entries 1, 3, 16). Moreover, a
study on MeOBiphep derivatives (entries 3–9) revealed an
interesting steric influence of the phenyl substituents on the
enantioselectivity levels: increasing steric hindrance resulted in
an increase of the ee. Indeed, the 84% ee obtained with
MeOBiphep (entry 3) could be improved to 88.9% when placing
an isopropyl substituent in the 3,5-position (ligand L8, entry
8). The lowest ee obtained with ligand L9 (entry 9), bearing
high steric congestion, may be explained by electronic factors
(presence of a 4-methoxy substituent) or, more probably, by
too much congestion around the ligand, preventing a complete
complexation of the ligand to the rhodium center. This ligand
screening also revealed that electron-rich ligands like Josiphos
or Duphos derivatives did not induce any asymmetric induction
and conversions were generally low.
A screening of other rhodium catalyst precursors32 and even
the preformed Rh-Binap complex did not allow an increase in
reaction enantioselectivities. As in the case of asymmetric 1,4-
addition of potassium trifluoro(organo)borates to Michael ac-
ceptors,12 cationic rhodium complexes were the most suitable
precursors. On the contrary, the reaction medium has a great
influence on either the conversion or enantioselectivity, and
nonpolar solvents like dioxane or toluene were the most adapted.
Finally, it is also important to note that reaction temperature
has a crucial influence. The highest enantiomeric excesses were
generally achieved at 110 ( 10 °C, while outside this range
both yields and enantioselectivities decreased.
Indeed, under these conditions, addition of potassium tri-
fluoro(phenyl)borate to methyl N-acetamido acrylate resulted
in good yields and enantioselectivities close to 90% using Binap
or MeOBiphep derivatives as the chiral ligand.
In order to further improve the enantiomeric excesses we
wondered if protonation by chiral phenol derivatives would give
higher levels of enantioselectivity, keeping in mind their
Scope/Limitations of the Reaction. With optimal conditions
in hand, we evaluated the scope and limitation of this rhodium-
catalyzed tandem 1,4-addition/enantioselective protonation. We
first studied the extension of this reaction to diversely protected
dehydroamino esters not only in order to access differently
protected useful amino esters but also in order to evaluate their
influence on reaction yields and enantioselectivities. All these
dehydroamino esters were easily prepared in high yields from
serine by selective protections and hydroxy elimination (see
experimental section) without intermediate purification steps.
(27) Miyashita, A.; Yasuda, A.; Takaya, H.; Toriumi, K.; Ito, T.; Souchi,
T.; Noyori, R. J. Am. Chem. Soc. 1980, 102, 7932.
(28) Schmid, R.; Forisher, J.; Cereghetti, M.; Schönholzer, P. HelV. Chim.
Acta 1991, 74, 370.
(29) Schmidt, R.; Broger, E. A.; Cereghetti, M.; Crameri, Y.; Foricher, J.;
Lalonde, M.; Müller, R. K.; Scalone, M.; Schoettel, G.; Zutter, U.
Pure Appl. Chem. 1996, 68, 131.
(31) Review: (a) Brunel, J.-M. Chem. ReV 2005, 105, 857. (b) Brunel, J.-
M. Chem. ReV. , ASAP Article (Update 1, DOI: 10.1021/cr078154l).
(32) The following catalyst precursors wee evaluated: [Rh(CH2CH2)2Cl]2
(8% yield, 20% ee), [Rh(cod)2]BF4 (74% yield, 84% ee), [Rh(cod)((R)-
Binap)]BF4 (51% yield, 84% ee), [Rh(cod)OH]2 (15% yield, 18% ee).
(30) (a) Duprat de Paule, S.; Champion, N.; Vidal, V.; Genet, J.-P.; Dellis,
P. (Synkem), WO Patent 03029259, 2003. (b) Duprat de Paule, S.;
Jeulin, S.; Ratovelomanana-Vidal, V.; Genet, J.-P.; Champion, N.;
Deschaux, G.; Dellis, P. Org. Process Res. DeV. 2003, 7, 399.
9
6162 J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008

Access to Enantioenriched R-Amino Esters
A R T I C L E S
Table 3. Scope in the Substitution Patterns on the
substrates are not easily accessible, even using efficient asym-
metric hydrogenation processes.33
Dehydroalaninea
From a practical point of view, dehydroalanine isopropyl
esters appeared to be the most suitable substrates as good yields
and enantioselectivities (ranging from 88 to 93%) were uni-
formly observed. Moreover the sense of the enantioselectivity
is easily controlled: using (S)-binap, products of (R) configu-
ration were obtained.
To further demonstrate the utility the present procedure for
the formation of enantio-enriched arylalanine derivatives, a
larger scale reaction was conducted. Reaction of dehydroamino
ester 1a (5 mmol) with potassium trifluoro(phenyl)borate 2a
furnishes 3aa in 78% yields and 88% enantioselectivity (eq 3).
After a single recrystallization, the enantiomeric excess was
increased to over 99%, and 3aa was obtained in 55% overall
yield.
entry
R1
R2
R3
yieldb
eec
1
2
3
4
5
6
7
8
H
H
H
Ac
Me
Me
Me
Me
Me
i-Pr
i-Pr
t-Bu
(1a)
(1b)
(1c)
(1d)
(1e)
(1f)
91
92
82
91
100
87
89.5 (R)
43
Zd
Bocd
Phthd
COCF3
Ac
Boc
Boc
89.5 (S)e
10
H
H
H
H
15
91 (R)
93 (R)
95 (R)
(1g)
(1h)
76
70
a Reactions conducted using 0.5 mmol of 1, 2 equiv of 2a, and 1
equiv of guaiacol with 3 mol% of [Rh(cod)2][PF6], 6.6 mol% of (S)-Binap
in toluene at 110–115 °C for 20 h. b Isolated yields. c Determined by
HPLC analysis using Daicel Chiralcel OD-H or Chiralpack AS-H chiral
column. d Z: benzyloxycarbonyl. Phth: Phthalimido. Boc: tert-butoxycar-
bonyl. e Using (R)-binap.
All these compounds were generally purified by simple recrys-
tallization or distillation: no chromatographic purifications were
needed.
From the tested amino protecting groups (Table 3, entries
1–5), acetyl and Boc (tert-butoxycarbonyl) gave good yields
of phenylalanine derivatives with equally high ee values. The
Z protecting group (benzyloxycarbonyl) was not compatible with
the reaction conditions, giving a byproduct and low enantiomeric
excess (entry 2). Concerning phthalimido or trifluoroacetyl
(entries 4 and 5), despite a very efficient addition (high yields,
reduced reaction time), we only obtained near racemic adducts.
The influence of the ester moiety was next evaluated using
different Boc and Ac protected amino esters. From these results,
it appeared that increasing the steric hindrance (from methyl to
tert-butyl ester) resulted in improved enantioselectivities. For
example, using the useful Boc protecting group, an ee of 89.5%
obtained with methyl ester (entry 3) was increased to 93% with
an isopropyl derivative (entry 7) and even up to 95% with tert-
butyl ester (entry 8). However, slightly lower yields were
generally achieved using tert-butyl dehydroamino esters. Indeed
a good compromise for achieving either high yields or enanti-
oselectivities appeared with the use of isopropyl derivatives,
all the more so as they are more easily accessible than the tert-
butyl esters.
Extension to Other Organometallic Reagents and Electron-
Deficient r,r′-Disubstituted Alkenes. The described process of
1,4-addition/enantioselective protonation is not limited to potas-
sium trifluoro(organo)borates. We have shown previously that
organosilanes and other organoboranes derivatives (boronic acids
or esters) were not suitable reagents in this reaction because
either no conversion or enantioselectivity was observed.8a On
the other hand organostannanes can be successfully used in this
process (Table 4). Indeed, under identical conditions, addition
of trimethylphenylstannane (4a) to 1a occurred in 89% yield
with a moderate 71% ee.
On that substrate, the ee could be increased up to 88% using
tributylphenylstannane (4b) instead of the trimethyl derivative.
As observed in the case of potassium trifluoro(organo)borates,
a phthalimido (entry 5) or benzyloxycarbonyl (Z, entry 6)
protecting group is not compatible with this reaction, affording
amino esters in low yields and enantiomeric excesses. On the
other hand, with Boc or Ac derivatives, elevated ee values were
generally observed using organostannanes (entries 2–4), whereas
yields were generally slightly lower than those obtained with
the corresponding trifluoroborate salts.
We next evaluated the scope of the reaction, by reacting
different dehydroalanines with potassium aryl- and even alken-
1-yl-trifluoroborates (Scheme 1) under standard conditions. From
these results, it appeared that a great variety of arylalanine
derivatives were readily obtained using this tandem carbometa-
lation/enantioselective protonation process. Interesting levels of
enantioselectivity ranging from 75 to 95% were generally
achieved and, as previously noted, the ee increased with the
steric hindrance of the ester. For example, the addition of
potassium (4-fluorophenyl)trifluoroborate on acetyl and Boc
dehydroalanines (1a and 1c) afforded 3ab and 3cb in 83 and
81% ee, respectively, whereas on isopropyl esters 3f and 3g
the ee were increased to 90 and 94%, respectively, to finally
attain 95% on 3h bearing a tert-butyl ester. Using potassium
alkenyltrifluoroborates 2h, it appeared that alkenyl substituted
alanine derivative 3ah is efficiently produced with good ee and
yield, with some isomerization of the double bond observed,
presumably via rhodium-catalyzed double-bond migration. This
represents an interesting feature of this reaction since these
As reported in the introductive part, this concept of 1,4-
addition/enantioselective protonation was extended to the
preparation of succinic esters,19 ꢀ2-amino acids,20 and R,R′-
dibenzyl esters21 using other proton sources than phenols. We
wondered if our reaction conditions could be extended to other
electron-deficient R,R′-disubstituted alkenes using phenols as
protonating agents. Unfortunately, despite several attempts, all
other tested substrates were not suitable in this reaction. Indeed,
(33) (a) For reviews on asymmetric hydrogenation, see: Noyori, R.
Asymmetric Catalysis In Organic Synthesis; Wiley: New York, 1994.
(b) Genet, J.-P. Reductions in Organic Synthesis: Recent AdVances
and Practical Applications; ACS Symposium Series 641; American
Chemical Society: Washington, DC, 1996; pp 31-51. (c) Burk, M. J.
Chemtracts Org. Chem. 1998, 11, 787. (d) Ohkuma, T.; Kitama, M.;
Noyori, R. In Catalytic Asymmetric Synthesis; Ojima, I., Ed.; VCH:
Weiheim, 2000; Chapter 1.4. (e) Noyori, R.; Ohkuma, T. Angew.
Chem., Int. Ed. 2001, 40, 40. (f) Genet, J.-P. In Reduction of
Functionalized alkenes; Adersson, P. G., Munslow, I., Eds.; Wiley-
VCH, New York, in press.
9
J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008 6163

A R T I C L E S
Navarre et al.
Scheme 1. Scope of the 1,4-Addition/Enantioselective Protonationa
a Reactions conducted using 0.5 mmol of 1, 2 equiv of 2, and 1 equiv of guaiacol with 3 mol % of [Rh(cod)2][PF6], 6.6 mol% of (S)-binap in toluene at
110–115°C for 20 h. b Determined by HPLC analysis using chiral column (see Supporting Information). c Isolated as an inseparable mixture containing 30%
d
of the 3,4-isomer. (S) isomer obtained using (R)-binap as ligand.
Table 4. 1,4-Addition/Enantioselective Protonation Using
hydrogen has been substituted by a methyl, afforded phenyl
alanine 3ia in low and even reversed enantiomeric excess
Phenylstannanesa
(eq 4).
entry
dehydroalanine
PhSnR′3
yieldb
eec
1
2
3
4
5
6
1a
1a
1c
1g
1d
1b
4a
4b
4a
4b
4a
4a
89
82
80
45
61
0
71 (R)
88 (R)
88 (R)
91 (R)
17 (S)
Such low asymmetric induction was generally observed upon
protonation with water. For example, methyl N-acetylami-
noacrylate (1a), under identical conditions, but using water
instead of guaiacol afforded a maximum of 18% ee of opposite
sign compared to phenol protonation.
a Reactions conducted using 0.5 mmol of 1, 2 equiv of 4a or 4b, and
1 equiv of guaiacol with 3 mol% of [Rh(cod)2][PF6], 6.6 mol% of
(S)-Binap in toluene at 110–115 °C for 20 h. b Isolated yields.
c Determined by HPLC analysis using Daicel Chiralcel OD-H or
Chiralpack AS-H chiral column.
Even more intriguing, it appeared that the reaction course
was influenced by the acidity, not only of the phenol (vide supra)
but also of the substrate: reaction of trifluoroacetyl derivative
1e with potassium trifluoro(phenyl)borate afforded phenylalanine
in low ee (Table 3, entry 5). However, the calculated pKA of
this dehydroamino ester (8.05, ACD)34,35 or of the resulting
amino ester (9.7, ACD) are lower by more than one unity than
those of acetamido or Boc derivative (Table 5). The pKA of
this trifluoroacetyl derivative is to compare with acidic phenols
that always resulted in low enantioselectivity: one can
guess that this dehydroamino ester is itself acting as a nonselec-
tive proton source. Indeed, it does seem that the acidity of the
reacting substrate (and even the product) has a great influence
on the stereochemical outcome of the reaction.
itaconic acid derivatives afforded quasi-racemic 1,4-addition
adducts while yields were still elevated. In the same way, on
reaction with potassium trifluoro(phenyl)borate, 5-exomethyl-
enefuran-2-one gave racemic 1,4-adduct in 90% yield. Indeed,
it appeared that our optimized conditions, involving binap (or
other atropoisomeric bidentate phosphanes) as the ligand for
rhodium and guaiacol as the proton source, were only adapted
for dehydroamino ester derivatives. Intrigued by the specificity
of this reaction, we conducted complementary experiments in
order to obtain some insights into the reaction mechanism.
Discussion on the Reaction Pathway. On the Importance
of the N-H Amido Substituent. From the results obtained with
a large variety of substrates (vide supra), it appeared that the
presence of a free N-H bond in the R-position of the Michael
acceptor was essential in order to achieve ahigh level of
enantioselectivity: substrates lacking this NH substituent (ita-
conic esters derivatives) or diprotected dehydroamino esters
(phthalimido) all gave quasi-racemic products. The crucial role
played by the presence of this free N-H substituent was
confirmed by the fact that the reaction with 1i, where the
To obtain further evidence on the reaction pathway, we
initiated deuterium labeling experiments in order to find the
origin of the incoming proton (substrate or phenol). Cross-
experiments were conducted, using either deuterated dehy-
droamino esters or deuterated guaiacol (Scheme 2).
(34) ACDpKa, V. 8.03; Advanced Chemistry Development: Toronto,
Canada, 2007.
(35) The predicted values of pKa were obtained using the ACD/I-Lab
Webservice (ACD/pKa 8.03).
9
6164 J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008

Access to Enantioenriched R-Amino Esters
A R T I C L E S
Table 5. Calculated pKA of Alanine Derivativesa
Scheme 3. Proposed Reaction Mechanism
a pKA values determined at 25 °C and zero ionic strength in aqueous
solution: see ref 35.
Scheme 2. Deuterium Labeling Studies: Origin of the Protonation
a % of deuterium incorporation in the R position determined by 1H NMR.
b 2-Methoxyphenol-d1. c Methyl 2-(acetyldeuterioamino)acrylate.
As expected, reaction of dehydroalanine 1a with potassium
phenyltrifluoroborate (2a), in the presence of 3 mol% [Rh-
(cod)2]BF6 and 6.6 mol% (S)-Binap, using D2O as the proton
source in toluene at 110 °C afforded 100% of R-deuterated near
racemic amino ester 3aa-d in quantitative yield. It is also
important to note that the reaction using water as the proton
source is more than 10 times faster than that using guaiacol.
Using 1 equiv of deuterated guaiacol36 under identical condi-
tions, we were surprised to observe only 28% deuterium
incorporation in the purified product 3aa, with an enantiomeric
excess of 90% as observed before. Low but higher deuterium
incorporation was also observed using 1g as starting material.
Indeed, at this point, it seems that the percentage of deuteration
is dependent on the starting dehydroalanine substrate. We were
all the more surprised by the results obtained starting with
amino-deuterated dehydroalanine 1a-d.
The reaction conducted with 1a-d, but using phenol as the
proton source, afforded 3aa with 41% deuterium incorporation,
a proportion that was even higher than those observed in the
reaction of 1a with guaiacol-d1. In light of these results, it
became clear that the reaction did not involve a direct proto-
nation of a putative rhodium enolate, which should have resulted
in the clean formation of R-deuterated adducts, as observed using
D2O as theproton source.37 On the contrary, it seemed that the
N-H proton of the starting substrate 1 is involved in the reaction
mechanism, all the more so when this proton is absent, as is
substrate 1i (Vide supra), very low yield and enantioselectivity
were achieved.
species IV is formed (Scheme 3). This η1-carbon-bounded
rhodium complex is certainly in equilibrium with η1-oxygen-
bounded rhodium enolate species or oxa-π-allyl9 type enolate.38
This rhodium complex IV can undergo a ꢀ-hydride elimina-
tion to generate a rhodium hydride species V complexed to the
formed imidoester. Generation of such species has also been
evoked in rhodium-catalyzed enantioselective hydrogenation of
R-acetamidocinnamate substrates.39 Hydride transfer to the
imido will generate an amido-rhodium complex VI with
concomitant control of the R center. The first implication of
such a hydride intramolecular transfer and the generation of an
imido species is that the imido must remain complexed to the
rhodium center before the hydride transfer occurs in order to
achieve high enantioselectivity. The amido-rhodium bond is then
protonated by the guaiacol to generate an aryloxo-rhodium
complex I liberating the arylalanine product 3. This aryloxo-
rhodium species is suited for transmetalation with the boron
reagent. Transmetalation of organoboron compounds to an
alkoxo or a hydroxo complex of palladium,40 rhodium,9 or
ruthenium41 has been described, allowing the regeneration of
organo-rhodium species. Indeed, in the absence of phenol, no
reaction occurs using potassium aryltrifluoroborates as an
organometallic partner (Scheme 4). On the other hand, a reaction
using organostannanes occurs even in the absence of added
(38) For references concerning the structure of rhodium-enolates, see: (a)
Slough, G. A.; Bergman, R. G.; Heathcock, C. H. J. Am. Chem. Soc.
1989, 111, 938. (b) Wu, J.; Bergman, R. G J. Am. Chem. Soc. 1989,
111, 7628. (c) Slough, G. A.; Hayashi, R.; Ashbaugh, J. R.; Shamblin,
S. L.; Aukamp, A. M. Organometallics 1994, 13, 890. (d) Refer-
ence 9.
(39) (a) Wiles, J. A.; Bergens, S. H. Organometallics 1998, 17, 2228. (b)
Detllier, C.; Gelbard, G.; Kagan, H. B. J. Am. Chem. Soc. 1978, 100,
7556.
Indeed, we proposed a new reaction mechanism to account
for these experimental results, which is supported by compu-
tational modeling studies (Vide infra). After transmetalation of
potassium organotrifluoroborate to the rhodium(I) complex I
and complexation of the dehydroalanine, an alkyl-rhodium
(40) Review: Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457.
(41) (a) Kakiuchi, F.; Kan, S.; Igi, K.; Chatani, N.; Murai, S. J. Am. Chem.
Soc. 2003, 125, 1698. (b) Kakiuchi, F.; Usui, M.; Ueno, S.; Chatani,
N.; Murai, S. J. Am. Chem. Soc. 2004, 126, 2706. (c) Ueno, S.;
Mizushima, E.; Chatani, N.; Kakiuchi, F. J. Am. Chem. Soc. 2006,
128, 16516.
(36) Mafumé, F.; Hashimoto, Y.; Hashimoto, M.; Kondow, T. J. Phys.
Chem. 1995, 99, 13814.
(37) To a certain extent, some deuterium scrambling is occurring between
dehydroalanine substrate and guaiacol, but this process is slow
compared to the reaction kinetic.
9
J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008 6165

A R T I C L E S
Navarre et al.
Scheme 4. 1,4-Addition/Enantioselective Protonation in the
Absence of Proton Source
Chart 2. Optimized Structures at the B3LYP/BII Level
a Isolated yield after hydrolysis of the reaction mixture.
Figure 2. Model system Amod for DFT calculations.
phenol, proving that the organotin compound can directly
transmetalate imido-rhodium species VI, while it is not the case
with trifluoroborate species.42
Indeed, reaction of phenyltributylstannane (4b) with dehy-
droalanine 1g under optimized conditions, but in the absence
of guaiacol, afforded the expected product 1ga in a 33% yield
and an enantioselectivity of 91%, which is comparable to those
obtained in the presence of guaiacol (Table 4, entry 4).
Apparently, phenol is only useful in the reactions with potassium
organotrifluoroborates, and its role would be the regeneration
of a rhodium species suited for transmetalation with borane
reagents. The overall mechanism implies a 1,4-addition type
reaction of the organometallic reagent followed by an intramo-
lecular hydride transfer from nitrogen to the adjacent carbon. It
is believed that, using more acidic phenols or water as proton
sources, a rhodium enolate species is protonated before the
hydride transfer has occurred and that this protonation is not
diastereoselective.
and C2 carbons are closer in complex Ap compared to structure
A (3.047 and 2.583 Å, respectively), which is favorable for the
carbon-carbon bond formation (Table 6). Thanks to this
favorable rearrangement, the bond formation can occur, with a
comparatively low free activation energy of +8.4 kcal/mol
(TS(Ap-B)), resulting in the formation of a σ-alkyl-rhodium
complex B of type IV (Scheme 3), where the double bond of
the model organometallic part is still complexed to the rhodium
center.43 The carbon-carbon bound formation is a highly
exergonic process (∆G ) -22.8 kcal/mol).
In order to support this postulated mechanism, computational
modeling studies were conducted.
Computational Modeling Studies on the Reaction Mecha-
nism. DFT calculations of the catalytic reaction potential energy
surface were undertaken for a simple achiral model at the
B3LYP/BII level. The transmetalation step will not be addressed
as well as complexation of the starting substrate and decom-
plexation of the final product. Model complex Amod (Figure 2)
was chosen to mimic the structural and electronic features while
minimizing computational time. Model Amod, which corresponds
to the starting dehydroalanine-catalyst adduct III (Scheme 3),
consists of a neutral rhodium complex with two cis ligated PH3
ligands, coordinated substrate 1a and a vinyl substituent to
mimic the aromatic ring of the potassium aryltrifluoroborate
precursor.
In complex B, the N-H bond of the amide is above the plane,
with a Rh-H distance of 2.911 Å. From this minimum, we
Among the different structures of type A found on the
potential energy surface (minima), two of them A and Ap, which
are in rapid equilibrium, were susceptible to further involvement
in carbon-carbon bond formation (Chart 2 and Figure 3). In
the more stable complex A, the carbon-carbon double bond is
perpendicular to the rhodium-phosphanes plane whereas in Ap
the double bong is parallel. These two minima are interconnected
Via the transition state TS(A-Ap), corresponding to the rotation
of the dehydroalanine moiety along the C4-Rh bond. The C3
(42) Transmetalation of organoboronic acids to silylamido-rhodium species
have been described: Zhao, P.; Incarvito, C. D.; Hartwig, J. F. J. Am.
Chem. Soc. 2007, 129, 1876.
Figure 3. Free energy (kcal/mol) surface for the 1,4-addition/protonation
at the B3LYP/BII level.
9
6166 J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008

Access to Enantioenriched R-Amino Esters
A R T I C L E S
Table 6. Selected Bond Length (in Å) of the Optimized Rhodium Structures
Rh-C2
Rh-C3
Rh-C4
Rh-N
Rh-H
C2-C3
C4-N
N-H
C4-H
A
2.030
2.054
2.095
2.160
2.249
2.239
2.491
2.155
2.218
2.226
2.313
2.795
2.763
2.887
2.304
2.280
2.281
2.152
2.172
2.162
2.187
3.131
3.025
3.051
2.984
2.985
2.151
2.065
3.028
2.925
2.916
2.933
2.911
1.638
1.538 1.364
3.047
2.693
2.583
1.970
1.519
TS(A-Ap)
Ap
TS(Ap-B)
B
TS(B-C1)
C1
1.488
2.426
2.812
TS(C1-C2)
C2
TS(C2-C3)
C3
TS(C3-D1)
D1
D2
2.529
3.374
2.878
3.156
3.188
2.174
2.178
2.175
2.212
2.676
3.027
2.070
2.053
2.082
2.083
2.011
2.061
1.529
1.563
1.585
1.598
2.281
3.652
1.368
1.358
1.351
1.353
1.437
1.477
2.485
2.994
2.978
2.634
2.744
2.390
2.283
1.839
1.123
1.101
3.042
2.160
2.557
2.944
Chart 3. Difluorphos Ligand
found a transition state TS(B-C1) associated with the ꢀ-hydride
elimination of the amide hydrogen resulting in the formation
of a rhodium hydride species C1 of type V (Scheme 3). The
imaginary frequency of this transition state is associated with
the hydride migration from the nitrogen to the rhodium atom,
starting from an η3-(C-N-H)Rh complex whose structure is
very close to TS(B-C1). However, at this level of theory, we
could not find, despite several attempts, any intermediate
between B and TS(B-C1). The overall activation energy of this
ꢀ-elimination process is ∆G‡ ) +27.8 kcal/mol, representing
the rate-determining step of the reaction, but this energy is
compatible with the reaction temperature (110–115 °C). Rhod-
ium hydride C1 has a distorted trigonal bipyramidal structure,
the hydride ligand and the CdC bond being in pseudoequatorial
positions. After a series of favorable isomerizations, the hydride,
complexed at the rhodium center in C1, is transferred to carbon
C4 leading to the formation of the amido-rhodium D1.
The rhodium hydride C1 is first isomerized to the slightly
more stable complex C2 Via a transition state TS(C1-C2) of
low activation energy (1.4 kcal/mol). In this process the acetyl
substituent on the nitrogen is moving from below the rhodium-
phosphane plane in C1 to above the plane in C2, facilitating
the following hydride migration from the equatorial to axial
position, adapted to the hydride transfer. The hydride movement
takes place Via the transition state TS(C2-C3) (∆G‡ ) +9.7
kcal/mol) to give the square planar complex C3. This process
is endergonic by 9.3 kcal/mol, but in this complex the hydride
is in close contact with the C4 carbon atom (the distance between
C4 and H is 2.283 Å). The imaginary frequency of this transition
state TS(C2-C3) only corresponds to the movement of the
hydride from the axial to equatorial position. Indeed, at this
level of theory, the transformation of C2 to C3 first involves a
decomplexation of the carbon-carbon double bound followed
by the hydride transfer. But once again, we did not localize
any other intermediate between C2 and C3.
any transition state or intermediate connecting D1 to D2 since
they did not involve any bound transformation, but simple ligand
exchange at the rhodium center.
Indeed, the DFT calculations, on this simple model, support
our proposed reaction mechanism involving a ꢀ-hydride elimi-
nation and, thus, an overall process involving an hydride transfer
from the nitrogen atom of the amido substituent to the adjacent
carbon C4. The high activation energy of the rate-determining
ꢀ-hydride elimination step is compatible with a reaction
temperature of 110–115 °C necessary for the reaction to proceed.
Difluorphos Ligand in 1,4-Addition/Enantioselective Proto-
nation. Thanks to a better understanding of the reaction
mechanism, we wondered if the 1,4-addition/enantioselective
protonation reaction could not be further improved in terms of
catalytic activity and/or enantioselectivity based on simple
propositions. First in order to achieve high enantioselectivity,
and because of this internal hydride transfer, it is important that
the generated imide in complex C1 remains complexed to the
rhodium center until the formation of the carbon-hydrogen
bond in D1. Second, based on the putative mechanism, the
overall process could be favored by facilitating the ꢀ-hydride
elimination step. We simply reasoned that employing a more
electron-deficient chiral ligand would create a more electron-
deficient rhodium center, not only favoring the ꢀ-elimination
step44 but also preventing decomplexation of the intermediate
imide.
Among the atropoisomeric ligands described in the literature,
difluorphos45 (L17, Chart 3) appeared to fulfill the previous
requirement: the presence of a difluoromethylene on the biaryl
backbone resulting in a slight, but measurable, increase in
π-accepting properties of the ligand.46 Moreover, as mentioned
earlier, in a related process, Sibi et al. have shown that the
Starting from the rhodium-hydride complex C3, we were able
to localize a transition state TS(C3-D1) connecting C3 to the
amido-rhodium D1. In this step, the hydride is transferred from
the rhodium to the carbon atom C4 giving D1 presenting an
agostic bond with the newly formed C-H bond. Overall, at
this stage the hydride transfer process is reversible, but a more
stable amido-rhodium species D2 (type VI, Scheme 3) was
found in the potential energy surface. We did not try to localize
(44) (a) Pruett, R. L.; Smith, J. A. J. Org. Chem. 1969, 34, 327. (b) Unruh,
J. D.; Christenson, J. R. J. Mol. Catal. 1982, 14, 19. (c) Schultz, M. J.;
Adler, R. S.; Zierkiewicz, W.; Privalov, T.; Sigman, M. S. J. Am.
Chem. Soc. 2005, 127, 8499.
(43) On related DFT calculations of other rhodium-catalyzed processes,
starting from an aryl-rhodium complex, the structure of type B is also
presenting an η2-complexation to the arene moiety: Darses, S.
Unpublished results.
(45) Jeulin, S.; Duprat de Paule, S.; Ratovelomanana-Vidal, V.; Genet, J.-
P.; Champion, N.; Dellis, P. Proc. Natl. Acad. Sci. U.S.A. 2004, 101,
5799.
9
J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008 6167

A R T I C L E S
Navarre et al.
Table 7. Binap versus Difluorphos in 1,4-Addition/Enantioselective
phenol derivatives, and particularly nontoxic and inexpensive
guaiacol, as the proton source was crucial in order to achieve
good levels of enantioselectivity and for the reproducibility of
the reaction. This proton source may be easily recovered and
have completely suppressed competitive reduction of the
organometallic reagent that is usually observed in these rhodium-
catalyzed reactions. Applied to dehydroamino esters, this
reaction allowed access to diversely substituted R-amino esters
in high yields and good enantiomeric excesses. By fine-tuning
of the ester moiety, and using difluorophos as a chiral ligand,
increased levels of enantioselectivity, generally close to 95%,
were achieved. Deuterium labeling experiments revealed, and
DFT calculation supported, an unusual mechanism involving a
hydride transfer from the amido substituent to the R carbon
explaining the high levels of enantioselectivity achieved in
controlling this R chiral center.
Protonationa
yieldb (ee)
entry
product 3
L1
L17
1
2
3
4
5
6
7
3aa
3cb
3cc
3cd
3ga
3gb
3fc
89 (89.5)
84 (81)c
72 (85)
79 (86)c
76 (93)
44 (95)
73 (91)
86 (92)
97 (94)c
77 (95)
97 (94)c
97 (95)
97 (95)
99 (91)
a Reactions conducted using 0.5 mmol of 1, 2 equiv of 2, and 1 equiv
of guaiacol with 3 mol% of [Rh(cod)2][PF6], 6.6 mol% of (S)-L1 or
(S)-L17 in toluene at 110–115 °C for 20 h. b Isolated yields. c (S) isomer
obtained using (R)-L1 or (R)-L17.
Experimental Section
Computational Details. All calculations were performed with
Gaussian 03.47 All intermediate and transition-state geometries were
optimized using density functional theory (DFT), with Becke’s
three-parameter functional (B3)48 and Lee, Yang, and Parr (LYP)
correlation energies.49 The double-ꢁ quality, Hay and Wadt
LANL2DZ basis set for the valence and penultimate shells, with
effective core potentials at rhodium50 and phosphorus,51 and a
Dunning/Huzinaga full double-ꢁ basis52 on C, H, N, and O, was
used for initial studies. This set of basis functions and effective
core potentials is referred to as BI. From these optimized structures,
geometry optimizations were performed using the more sophisti-
cated basis sets and effective core potentials from the Stuttgart
group: these computations used effective core potentials for
rhodium53 replacing 28 core–electrons and a valence basis set with
the following contraction scheme: (311111/22111/411). For all other
atoms a 6-31G(d,p) basis set was utilized.54 This basis set will be
referred to as BII throughout this work.
Vibration frequency calculations were performed on all of the
optimized structures with the same method to provide free energies
at 298.15 K. The optimized minima and the transition structures
have been confirmed by harmonic vibration frequency calculations
as well as with IRC calculations for some transitions states. The
calculated relative electronic energies have been corrected with zero-
point energies (ZPEs).
Typical Procedure for the Asymmetric 1,4-Addition to Dehy-
droalanine Derivatives. A septum-capped vial equipped with a
magnetic stirring bar was charged with dehydroamino ester 1a (0.5
mmol, 71.6 mg), potassium (4-bromophenyl)trifluoroborate 2e (2
equiv, 1 mmol, 263 mg), [Rh(cod)2][PF6] (3 mol%, 6.9 mg), and
(R)-binap (6.6 mol%, 20.6 mg). The vial was closed and evacuated
under vacuum and placed under an argon atmosphere. Degassed
toluene (2 mL) and distilled guaiacol (1 equiv, 0.5 mmol, 55 µL)
were added, and the mixture was stirred in a preheated oil bath at
110–115 °C for 20 h. After cooling the vessel to room temperature,
the reaction mixture was purified by silica gel chromatography
difluorphos ligand was the most suitable in the synthesis of ꢀ2-
amino acids Via 1,4-addition, enantioselective protonation.20
Indeed, we envisioned evaluating this ligand in the rhodium-
catalyzed addition of potassium organotrifluoroborates to de-
hydroalanine derivatives. As a test ligand for comparison, we
also evaluated the Binap ligand on the same substrates. Under
previously optimized conditions, e.g., using [Rh(cod)2]BF6 as
a catalyst precursor, 2 equiv of an atropoisomeric ligand
compared to rhodium, and guaiacol as the proton source, some
potassium aryltrifluoroborates were allowed to react with
dehydroalanine derivatives (Table 7).
We were pleased to find that the difluorphos ligand compared
favorably and even surpassed the Binap ligand, not only in terms
of reactivity but also in terms of enantioselectivity. In the
reactions conducted with dehydroalanines possessing a methyl
ester (entries 1–4), comparable or higher yields were generally
observed, but, and more importantly, the enantiomeric excesses
obtained were always higher than those using binap ligand (up
to 10 point were gained in the formation of 3cc, entry 3). We
also evaluated the benefit of the difluorphos ligand on isopropyl
ester dehydroalanine derivatives (entries 5–7). As noted previ-
ously, and because of steric hindrance, higher enantioselectivities
were generally achieved with these substrates compared to the
methyl esters. As expected, the enantioselectivities observed
with this ligand were in the same range as those obtained with
binap ligand. However, higher yields (and generally quantitative
yields) were achieved with difluorphos (entries 5–7). More
particularly, phenylalanine 3gb was obtained in 97% yield,
whereas by using the binap ligand only a moderate yield of
44% was achieved.
Indeed, from these preliminary results, it appeared that
difluorphos is a highly suitable ligand in this reaction, allowing
production of phenylalanines in quantitative yields and with
enantioselectivities close to 95%.
(47) Pople, J. A. et al. GAUSSIAN 03, ReVision B.05; Gaussian, Inc.:
Pittsburgh, PA, 2003.
(48) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
(49) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(50) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
(51) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 285.
(52) Dunning, T. H.; Hay, P. J. In Modern Theoretical Chemistry; Schaefer,
H. F., III, Ed.; Plenum: New York, 1976; Vol. 3, p 1.
(53) Andrae, D.; Häussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor.
Chim. Acta 1990, 77, 123.
(54) (a) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54,
724. (b) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972,
56, 2257. (c) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973,
28, 213. (d) Hariharan, P. C.; Pople, J. A. Mol. Phys. 1974, 27, 209.
(e) Gordon, M. S. Chem. Phys. Lett. 1980, 76, 163.
Conclusion
We have shown that the concept of rhodium-catalyzed tandem
carbometalation/enantioselective protonation was highly suited
for introduction of an organic substituent in the ꢀ-position of
an electron-deficient alkene with concomitant control of the
chirality of the R center. In this tandem process, the use of
(46) Jeulin, S.; Duprat de Paule, S.; Ratovelomanana-Vidal, V.; Genet, J.-
P.; Champion, N.; Dellis, P. Angew. Chem., Int. Ed. 2004, 43, 320.
9
6168 J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008

Access to Enantioenriched R-Amino Esters
A R T I C L E S
(ethyl acetate/cyclohexane 3:2) to afford 3ae as a white solid (113
mg, 75% yield). [R]D25) +79 (c ) 1.03, CHCl3). HPLC (Daicel
Chiralcel OJ, hexane/propan-2-ol 90/10, 0.5 mL/min, 215 nm): tR
) 25.3 (S) and 29.2 min (R).
Note Added after ASAP Publication. The graphics for Chart
2 and Table 7 were incorrect due to a production error in the version
published on the Web April 9, 2008. The corrected version was
reposted on April 12, 2008.
Acknowledgment. This work was supported by the Centre
National de la Recherche Scientifique (CNRS). L.N. and R.M.
thank the Ministère de l’Education et de la Recherche for a
grant. We are grateful to Dr. R. Schmidt and Dr. M. A.
Scalone (Hoffmann-La-Roche) for the generous gift of
MeOBiphep chiral ligand and derivatives. We are also
grateful to Dr. H. U. Blaser (Solvias) and Prof. S. Gladiali
for the generous gift of chiral Josiphos and Binepine ligands,
respectively.
Supporting Information Available: Experimental section for
the preparation and the description of the compounds, and
the optimized structures (phenols and rhodium) with total
energies of the species in the proposed pathways. This
information is available free of charge via the Internet at
http://pubs.acs.org.
JA710691P
9
J. AM. CHEM. SOC. VOL. 130, NO. 19, 2008 6169