Phosphoramidite ligands in iridium-catalyzed allvlic substitution

Article abstract of DOI:10.1002/chem.200501180

A new phosphoramidite ligand was used in the iridium-catalyzed allylic substitution reaction. This permitted high regio- and enantioselectivities on a wide variety of substrates and nucleophiles. Because of the stereospecificity of the reaction obtained by using branched substrates, a kinetic resolution reaction was attempted. The origin of the impressive efficiency of this ligand in terms of kinetics was explored in detail, as was the role of the substituent in the ortho-position of the amine moiety.

Full text of DOI:10.1002/chem.200501180

DOI: 10.1002/chem.200501180
Phosphoramidite Ligands in Iridium-Catalyzed Allylic Substitution
Damien Polet,[a] Alexandre Alexakis,*[a] Karine Tissot-Croset,[a]
ClØmence Corminboeuf,[b] and Klaus Ditrich[c]
Abstract: A new phosphoramidite ligand was used in the iridium-catalyzed allylic
substitution reaction. This permitted high regio- and enantioselectivities on a wide
variety of substrates and nucleophiles. Because of the stereospecificity of the reac-
tion obtained by using branched substrates, a kinetic resolution reaction was at-
tempted. The origin of the impressive efficiency of this ligand in terms of kinetics
was explored in detail, as was the role of the substituent in the ortho-position of
the amine moiety.
Keywords: allylic substitution
·
asymmetric catalysis · chiral ligands ·
iridium · phosphoramidite
Introduction
Among the methodologies that synthetic chemists employ
to create chiral centers, allylic substitution is of considerable
importance. The products of this reaction are particularly in-
teresting because they contain a double bond, which is one
Scheme 1.
of the most versatile functional groups in organic synthesis.
Although the most-studied catalytic system is that with
palladium as the metal source, its applicability is limited
mostly to symmetrical substrates in which the enantioselec-
tivity is a result of regioselectivity.[1]
Regarding the nonsymmetrical prochiral substrates
(Scheme 1), the situation is more complex and challenging,
as one has to control both regio- and enantioselectivity. For
this purpose, metals, such as nickel,[2] platinum,[3] copper,[4]
molybdenum,[5] ruthenium,[6] tungsten,[7] rhodium,[8] and iri-
dium,[9,10] have emerged as efficient systems. Although the
regiochemistry of palladium normally favours the linear
product, there is an increasing number of examples favour-
ing the branched regioisomer.[11]
Following the pioneering work of Takeuchi et al.[9,10a–c]
and Helmchen et al.,[10d–g] Hartwig and co-workers achieved
high enantioselectivities by using L1 as ligand in iridium-cat-
alyzed
allylic
amination[10m]
and
etherification[10n]
(Scheme 2). They later identified the active species C1 in
their reaction;[12] it came from an insertion into the C H
ꢀ
bond in the methyl group on the amine part of the ligand
(Scheme 3). Helmchen et al. had previously observed such
ꢀ
an insertion into an aromatic C H bond of triphenylphos-
phite as ligand.[10f]
Here, we describe improved conditions and ligands for al-
lylic alkylation, which result not only in higher regio- and
enantioselectivities, but also in improved kinetics.
[a] D. Polet, A. Alexakis, K. Tissot-Croset
Department of Organic Chemistry, University of Geneva
30, quai Ernest Ansermet, 1211 Geneva 4 (Switzerland)
Fax : (+41)22-379-3215
Results and Discussion
E-mail: alexandre.alexakis@chiorg.unige.ch
[b] C. Corminboeuf
Allylic alkylation reaction: The results obtained by Helm-
chen et al. using L1 were good, but limited in scope.[10g]
Moreover, this catalyst system needed to be improved re-
garding its long reaction time (from 18 hours to 4 days), its
only moderate to good regioselectivities (branched/linear (b/
Department of Physical Chemistry, University of Geneva
30, quai Ernest Ansermet, 1211 Geneva 4 (Switzerland)
[c] K. Ditrich
BASF Aktiengesellschaft GVF/E-A030
Ludwigshafen 67056 (Germany)
3596
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3596 – 3609

FULL PAPER
Scheme 2. Ligands used in this study.
Under the original conditions developed by Helmchen
and co-workers (4% Ir catalyst and 4% ligand) for the cin-
namyl derivative 1a,[10g] ligand L1 did not seem to be opti-
mal (regioselectivity of 70:30 and enantioselectivity of 70%,
Table 1, entry 1), although the introduction of LiCl as an ad-
ditive improved the result remarkably (3a/4a ratio of 91:9
and 86% optical purity, Table 1, entry 2).
Scheme 3.
Knowing that the carbonate derivative 2a was much more
active than 1a in the amination version of the same reac-
tion,[10m] we[15] initially tested 2a under the classical condi-
tions without any additive. The reaction was sluggish, even
at 608C, and led to only low regio- and enantioselectivities
(Table 1, entry 3). The addition of lithium chloride led to an
impressive acceleration of the reaction, together with an im-
provement of the regio- (99:1) and enantioselectivities
(98%) (Table 1, entry 4). The role of LiCl is discussed
below. This result again demonstrates the superiority of the
new ligand L2a relative to L1, both in terms of regio- and
enantioselectivities. We even tested a much lower catalyst
loading (0.5 mol% of Ir catalyst), which did not seem to be
deleterious, as after 4 h at 408C, 83% conversion was ob-
tained together with an ee of 96% and a regioselectivity of
99:1 (Table 1, entry 5). Such a small amount of catalyst
l) ratios 30:70 to 91:9), and its imperfect enantioselectivities
(86% for the classical substrate cinnamyl acetate).
Hartwig et al.[10o,p] and Helmchen et al.[13] recently ach-
ieved improvement by using additives, such as 1,4-
diazabicyclo
A
ACHTREUNG
tion of the catalytic species C1. In contrast, we examined
the potential of structural modifications of L1 to improve
the scope and kinetics of the reaction, as there was a need
for a general and simple catalytic methodology in asymmet-
ric allylic alkylation. To accelerate the reaction, we devel-
oped a new family of phosphoramidite ligands L2a–c, bear-
ing two ortho-methoxy substituents on the amine moiety,
which provided excellent regio- and enantioselectivities in
Cu-catalyzed allylic substitution[14] (Scheme 4).
ꢀ
(0.5%) is impressive for C C
bond formation methodologies.
This represents
a
turnover
number (TON) of 200 and a
turnover frequency (TOF) of
50 at 408C.
Scheme 4.
Chem. Eur. J. 2006, 12, 3596 – 3609
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3597

A. Alexakis et al.
Table 1. Enantioselective Ir-catalyzed allylic alkylation of carbonates.[a]
ortho-position, it was deleteri-
ous for the stereoselectivity
(Table 2, entry 4). An excess of
only 79% was obtained for
this substrate. This was ob-
served previously by Hartwig
et al. in the amination reac-
tion.[10m] With para- and meta-
donor substituents, the dioxo-
lane derivative 2 f led to excel-
lent results (Table 2, entry 5).
Of particular interest is the
substrate bearing a para-chloro
substituent 2g. Its transforma-
tion into 3g is known to give
Entry
Substrate
Ligand
Additive
T [8C]
Time [h]
Yield[b] [%]
3/4[c]
ee[d] [%]
1[e]
2[e]
3
1a
1a
2a
2a
2a
2a
2a
L1
L1
–
LiCl
–
LiCl
LiCl
LiCl
LiCl
25
25
60
25
40
25
25
18
18
22
2
4
14
14
41
98
31
82
70:30
91:9
65:35
99:1
99:1
97:3
70 (R)
86 (R)
44 (R)
98 (R)
96 (R)
89 (S)
93 (R)
L2a
L2a
L2a
L2b
L2c
4[f]
5[g]
6
A
28
81
7
99:1
[a] Molar ratio: substrate/LiCl/NaHC
N
N
[b] Isolated yield; in parenthesis: conversion determined by GC–MS. [c] Determined by 400 MHz 1H NMR
spectroscopy or GC–MS. [d] Determined by chiral SFC (HPLC for entries 1 and 2); in parentheses: absolute
configuration attributed by comparison with published data (ref. [7]). [e] Taken from ref. [10g]. [f] Performed
on
a
1-mmol scale. [g] Performed on
a 4-mmol scale with a molar ratio: substrate/LiCl/NaHC-
A
ACHTREUNG
Comparison of the diaster-
eomeric ligands L2a (aS,SS)
and L2b (aR,SS) clearly shows
a matched/mismatched situa-
tion, with L2b being mis-
matched (entry 6, 89% ee). It
should be stressed that the in-
version of the binaphthol part
of the ligand translates into an
inversion of stereochemistry
on the resulting allylic adduct,
indicating that the binaphthyl
chiral element is dominant. In-
terestingly, the simpler ligand
L2c affords a relatively high ee
(Table 1, entry 7, 93%). The
absolute stereochemistry of
the adduct corresponds to that
obtained with the matched
ligand L2a. Thus, the biphenyl
part of the axially labile ligand
L2c seems to adopt the same
configuration as the (S)-bi-
naphthyl of L2a. This adaptive
behavior is similar to that ob-
served in Cu-catalyzed reac-
tions, in which the induced
atropoisomerism of the bi-
phenyl has the same configura-
tion as the matched diastereo-
meric pair in the binaphthyl-
type ligand.[16]
Scheme 5.
Table 2. Enantioselective Ir-catalyzed allylic alkylation of carbonates with L2a.[a]
Entry
Substrate
R
T [8C]
Time [h]
Yield[b] [%]
3/4[c]
ee[d] [%]
96 (R)
1
2b
25
20
73
99:1
2
3
2c
2d
35
30
24
6
99
58
>99:1
97 (R)
96 (R)
99:1
4
2e
30
6
98
>99:1
79 (R)
5
6
2 f
2g
30
35
44
24
70
90
99:1
97 (R)
>99:1
97 (R)[e]
7
8
9
2h
2i
30
30
30
65
30
65
40
92
65
94:6
80:20
97:3
94 (R)
96 (R)
98 (R)
ꢀ
n-Pr
2j
[a] Molar ratio: substrate/LiCl/NaHC
G
G
[b] Isolated yield. [c] Determined by 400 MHz 1H NMR spectroscopy or GC–MS. [d] Determined by chiral
SFC; in parentheses: absolute configuration. [e] The absolute configuration was determined by comparison
with published data (ref. [7]); all other configurations were attributed by analogy with 3a.
straightforward access to the therapeutically useful GABAB
We next focused our efforts on expanding the scope of
the reaction. We tested different allylic substrates
(Scheme 5), starting with the 2-naphthyl derivative 2b that
regioselectively (99:1) gave the branched isomer in 96% ee
(Table 2, entry 1). We then modulated the electron demand
receptor agonist (R)-baclofen hydrochloride (Scheme 6).[17]
We obtained 3g with an optical purity of 97% without any
trace of the regioisomer 4g (Table 2, entry 6).
ꢀ
of the aromatic ring. With the OMe group in the para- or
meta-positions, substrates 2c and 2d behaved similarly and
gave excellent regioselectivities and enantiomeric excesses
(Table 2, entries 2 and 3). However, for steric or coordina-
tion reasons, if the same substituent as that in 2e was in the
Scheme 6.
3598
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3596 – 3609

Iridium-Catalyzed Allylic Substitution
FULL PAPER
An electron-withdrawing group on the aromatic core of
the substrate was also tested (Table 2, entry 7). Although
very good enantioselectivity was measured for 3h, a signifi-
cant amount of 4h was detected (6%).
bilized alkyl nucleophiles under these conditions
(Scheme 8). Among them, trisubstituted methyl malonate
Alkyl groups as substituents were investigated. Primary
(2i, Table 2, entry 8) and secondary (2j, Table 2, entry 9)
alkyl derivatives gave high enantioselectivites (96 and 98%,
respectively), but lower regioselectivities (80:20 and 97:3, re-
spectively) than aromatic derivatives.
Scheme 8.
Allylic carbonates are usually the substrates of choice,
and allylic acetates are used less often, owing to their lower
reactivity and selectivity.[10m] However, despite requiring a
longer reaction time, the latter (Scheme 7) gave similar re-
sults (Table 3, entry 1) under the conditions employed in
7a required slight heating to lead to full conversion
(Table 4, entry 1). The resulting regio- and enantioselectivi-
ties were as high as those obtained with the simple malo-
nate. Two nucleophiles containing a prostereogenic center,
7b and 7c, were also tested.
Although high regio- and
enantioselectivities were ach-
ieved, no stereocontrol on the
nucleophile occurred, leading
to the corresponding diaster-
eoisomers in
a 1:1 mixture
(Table 4, entries 2 and 3). Of
particular interest was nucleo-
phile 7d, bearing an allyl sub-
stituent (Table 4, entry 4). In
addition to the good regiose-
lectivity obtained, we trans-
formed product 7d into the cy-
clohexene derivative 10 by a
ring-closing metathesis by
using the first generation
Grubbꢁs catalyst[18] (Scheme 9).
The optical purity of 10 was
extremely high (98.7% ee).
Scheme 7.
Table 3. Ir-catalyzed allylic alkylation of acetates with ligand L2a.[a]
Entry
Substrate
R
T [8C]
Time [h]
Yield[b] [%]
79
3/4[c]
ee[d] [%]
97 (R)
1[e]
5a
25
22
99:1
ꢀ
2
3
4
5
5b
5b
5b
5b
n-Pr
n-Pr
n-Pr
n-Pr
25
30
42
55
60
66
15
18
21
87
71
82
91:9
94 (R)
97 (R)
97 (R)
97 (R)
ꢀ
87:13
86:14
82:18
ꢀ
ꢀ
6
6a
25
3
81
>99:1
10 (R)
[a] Molar ratio: substrate/LiCl/NaHC
C
G
Attempts at kinetic resolution:
In view of the intriguing result
obtained by using 6a as sub-
strate (Table 3, entry 6), we
this study. Thus, substrate 5b
needed to be heated slightly
for the reaction to be complet-
ed. This affected the regiose-
lectivity (Table 3, entries 2–5)
rather than the enantioselectiv-
ity. Finally, we also tested a
branched substrate 6a. Howev-
er, the enantioselectivity drop-
ped dramatically, showing that
the isomerization of the Ir p-
complex is a very slow process,
a result also reported by Helm-
chen et al.[10e] and Moberg
et al.[17b]
Table 4. Allylic alkylation with different stabilized nucleophiles.
Entry
Nu
Structure
T [8C]
Time [h]
8/9[a]
Yield[b] [%] (d.r.)
95 (–)
ee[c] [%]
1
7a
40
15
>99:1
97
2
7b
25
3
94:6
79 (51:49)
95[d]
3
4
7c
7d
25
40
5
3
>99:1
68 (53:47)
85 (–)
95[e]
n.d.
94:6[f]
In addition to studying vari-
ous electrophiles, we were in-
terested in testing different sta-
[a] Measured by GC–MS, unless otherwise stated. [b] Isolated yield; in parenthesis: diastereomeric ratio.
[c] Measured by chiral SFC. [d] Measured on the minor diastereoisomer. [e] Measured on the major diaster-
eoisomer. [f] Measured by H NMR spectroscopy.
1
Chem. Eur. J. 2006, 12, 3596 – 3609
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3599

A. Alexakis et al.
Scheme 9.
wondered if it would be possible to use this memory effect
in a kinetic resolution reaction. Treatment of racemic start-
ing material 6a with half an equivalent of sodium dimethyl
malonate gave 3a with 84% ee (R) and the recovered start-
ing material with 52% of the predominant R enantiomer
(Scheme 10). The ratio 6a/2a was 49:51.
Scheme 11.
tion of the nucleophile. The sum of the amounts of (R)-3a
and the rest of (S)-6a is 58%. As the remainder is 42%, it
can be stated that 8% racemization occurred, meaning that
ki ¼
6
0. Therefore, by decreasing the concentration of the nu-
cleophile, we should theoretically be able to get an efficient
dynamic kinetic resolution. This would be reached under
real Curtin–Hammet conditions (see below).
Scheme 10.
From the results of this simple experiment, we have estab-
lished that there are two stereoselection processes in our re-
action: kinetic (important) and thermodynamic (less impor-
tant). The kinetic process is the enantioface preference of
our catalyst (k1 =4.9k2). The thermodynamic one is the equi-
librium of the intermediate, which favours the same face in
terms of stability (ki >kꢀi). An efficient dynamic kinetic res-
olution (DKR) would be theoretically possible by lowering
the concentration of the nucleophile, for example, by using
a weaker base, such as the BSA/KOAc system (Scheme 12).
As already proved by Helmchen et al.,[10f] the reaction
can be seen as a double inversion process, and, hence, as a
formal retention event. Consequently, we can state that the
branched allylic acetate (S)-5a is more reactive than its
enantiomer. This corresponds to the same face-preference
as ligand L2a in the allylic substitution on the nonchiral sub-
strates cinnamyl acetate 5a or carbonate 2a (see above).
The enantioface preference (s) is expressed as a function
of conversion (C) and the enantiomeric excess of the re-
maining starting material (ee) [Eq. (1)]:[19]
s ¼ kS=kR ¼ ln½ð1ꢀCÞð1ꢀeeފ=ln½ð1ꢀCÞð1 þ eeފ ¼ 4:9
ð1Þ
This means that the bottom face is preferred to the upper
face by a factor of 4.9 for the oxidative addition. We assume
that the oxidative addition is an irreversible process. In view
of the ee of the product, we can then state that a racemiza-
tion process occurs during the reaction, because the ee of
the product does not reflect the optical purity of the remain-
ing starting material. Two racemization pathways are possi-
ble:[20] enantioinversion by another complex, or the s–p–s
process (Scheme 11). If the oxidative addition is irreversible
and the nucleophilic attack is stereospecific, then a racemi-
zation process takes place. In the absence of racemization,
the ee of the product should reflect that of the remaining
starting material (52%), which is clearly not the case
(84%).
Scheme 12. Conditions: 1) RT, 16 h, 3 equiv CH2
cat. KOAc; 2) substrate/catalyst ratio =10:1, temperature 658C, reaction
time 16 h, slow addition of 1.1 equiv NaCH(CO2Me)2 over 5 h.
G
AHCTREUNG
Under these classical conditions for palladium catalysis, no
enantiomeric excess was obtained (conditions 1). We tried
(conditions 2) to increase the temperature to 658C and the
amount of catalyst to 10 mol%, and treated the substrate
with 1.1 equivalent of NaCH(CO2Me)2 introduced very
A
The enantioface inversion equilibrium is a first-order
process (ki), whereas nucleophilic attack is a second-order
process (k3[Nu] and k4[Nu]), dependent on the concentra-
slowly (addition time of 5 h) to give the system a chance to
reach the Curtin–Hammett conditions. As a result, we ob-
tained 32% ee, which is not spectacular, but confirmed our
3600
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3596 – 3609

Iridium-Catalyzed Allylic Substitution
FULL PAPER
hypothesis. This result was promising and may be improved
in the future.
the true catalyst. Indeed, this additive is not required for
other nucleophiles, such as amines (see below). Both sys-
tems (RNH2 and NaCH(CO2Me)2/LiCl) give similar regio-
N
Symmetrical substrates: We tested the symmetrical substrate
11 under the conditions described previously (Scheme 13).
and enantioselectivities. Therefore, we turned our attention
to the nature of the alkyl nucleophiles. Two hypotheses
could be formulated: 1) an
Na–Li transmetallation results
in a lithium malonate salt or 2)
LiCl breaks the aggregates of
the sodium salt of the malo-
nate anion. To evaluate these
suppositions, we employed sev-
eral bases, such as lithium hy-
dride, lithium carbonate, and
Scheme 13.
lithium methylate to deproto-
nate
the
nucleophile
Interestingly, although this substrate was more inert than
the other allylic acetates used, it led to 80% conversion and
86% ee after 23 h at 408C. The
(Scheme 15). The results are shown in Table 5. The use of
lithium hydride did not result in any conversion (Table 5,
recovered starting material
gave an enantiomeric excess of
only 8%. This experiment con-
firms the above result obtained
with 6a. The low ee of the re-
Scheme 15.
maining
starting
material
shows that the face selectivity
is not very high; instead, the
Table 5. Evaluation of several bases in the Ir-catalyzed allylic alkylation.[a]
high
enantiodiscrimination
Entry
Base
Additive
T [8C]
Time [h]
Conv.[b] [%]
3/4[c]
ee[d] [%]
occurs on the meso-p-allyl
complex.
1
2
3
LiH
–
–
–
LiCl
LiCl
40
40
40
45
45
22
22
22
4
0
0
0
n.d.
n.d.
n.d.
98:2
99:1
n.d.
n.d.
n.d.
97 (R)
97 (R)
Li2CO3
MeOLi
MeOLi
MeOLi
Allylic
acetate
13a
(Scheme 14) is also a classical
substrate in allylic substitution
and has been extensively stud-
4
100 (73)
95 (78)
5[e]
5
[a] Molar ratio: 2a/base/additive/CH2
N
N
strate. [b] Conversion determined by GC–MS; in parenthesis: isolated yield. [c] Determined by GC–MS.
[d] Determined by chiral SFC; in parentheses: absolute configuration. [e] Catalyst recovered from reaction of
entry 4 and reused.
ied
in
the
Pd-catalyzed
system.[21] However, it led to
no conversion under our con-
ditions. Even the more reac-
tive carbonate congener 13b was inert under the same treat-
ment. Because they feature a defined Z geometry of the
double bond, these substrates do not behave like their iso-
mers, as previously observed by Takeuchi.[9]
entry 1), probably for reasons of solubility. Lithium carbo-
nate seemed equally inefficient in deprotonating the nucleo-
phile (Table 5, entry 2), and lithium methylate, which was
more basic and soluble under our conditions, did not lead to
conversion (Table 5, entry 3). We can, therefore, conclude
that Na–Li metal exchange does not occur. The addition of
LiCl to this latter base gave totally different results (Table 5,
entry 4). After only 4 h at 458C, the conversion was com-
plete and the regio- and enantioselectivities were excellent
(98:2 and 97%, respectively). Thus, LiCl seems to play a
role already observed by Seebach[22] for Li or Na enolates:
the “pure” aggregates of the latter exhibit a different reac-
tivity than the “mixed” aggregates formed with LiX addi-
tives.
Scheme 14.
By employing these simpler conditions, we were also able
to demonstrate that the iridium catalyst was recoverable
without any drawbacks (Table 5, entry 5). The reaction run
with the recycled catalyst required an additional hour to
reach completion. No loss of regio- or enantioselectivity was
Role of LiCl and catalyst recycling: Although the advantage
of adding LiCl in the allylic alkylation reaction is clear, we
have not so far discussed a rationale for its role. An initial
naive hypothesis is that LiCl does not change the nature of
Chem. Eur. J. 2006, 12, 3596 – 3609
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3601

A. Alexakis et al.
nucleophiles. With ligand L2a,
the results were as high or
even slightly higher than those
obtained by Hartwig et al. with
L1.
Comparison with ligands
L2b (Table 5, entry 4) and L2c
(Table 5, entry 5) allows ap-
proximately the same observa-
tions to be made as for the al-
kylation reaction. Ligand L2b
was the mismatched diastereo-
meric pair (47% ee) and led to
the
opposite
enantiomer,
whereas the flexible biphenyl
derivative L2c gave relatively
high ee (92%), in favour of the
same enantiomer as for the
matched ligand L2a.
Scheme 16. Catalyst recycling technique.
observed. Scheme 16 shows the procedure used to recycle
the catalyst (see Experimental Section). This method al-
lowed us to recover 65% of the catalyst without the need to
take any precautions regarding oxygen or water.
Mechanistic considerations: Apparently, the iridium catalyst
resulting from ligand L2a was not only more enantioselec-
tive, but also showed impressively higher kinetics than the
one obtained by using L1.[14,15] To explain the higher effi-
ciency of L2a over L1, we initially postulated a P–O hemila-
bile character, although this would lead to a seven-mem-
bered metallo-ring C2 (Scheme 18) and the exclusion of the
chloride anion from the first coordination sphere.
Allylic amination reaction: In addition to the previous nu-
cleophiles, we tested primary amines under the conditions
described previously by Hartwig et al.[10m] (Scheme 17). The
results are presented in Table 6.
Scheme 18.
Scheme 17.
Indeed, it was legitimate to consider L2a as a bidentate
ligand, which would allow a stronger positive character to
enhance the oxidative addition to the substrate.[23] Prelimi-
nary calculations (see Experimental Section) on the related
ligand L4 support such an assumption (Figure 1). Interest-
ingly, the square-planar arrangement favoured by d8 transi-
tion metals, such as IrI, is not distorted by a bidentate mode
of coordination of the ligand L4. However, we have not
been able to observe any coordination of the ortho-methoxy
Table 6. Enantioselective Ir-catalyzed allylic amination.[a]
Entry
Amine
Ligand
Conv.[b] [%]
16/17[c,e]
ee[d–f] [%]
1
2
3
4
5
15a
15b
15c
15a
15a
L2a
L2a
L2a
L2b
L2c
98 (88)
>99 (91)
>99 (89)
62
98:2 [99:1]
99:1 [n.d.]
98:2 [98:2]
50:50
97 (S) [95]
97 (+) [97]
98 (S) [96]
47 (R)
92
99:1
92 (S)
[a] Molar ratio: 2a/15/[{Ir(cod)Cl}2]/L=1:1.3:0.01:0.02; scale: 1 mmol
G
1
group by H or 31P NMR spectroscopy or to get any crystals
substrate. [b] In parentheses: isolated yields after silica gel chromatogra-
phy. [c] Determined by 1H NMR analysis of the crude reaction mixtures.
[d] Determined by chiral SFC analysis of the corresponding acetamide.
[e] In square brackets, Hartwigꢁs results with L1 (ref. [10m]). [f] In paren-
theses, absolute configuration.
suitable for X-ray crystallography.
Although these calculations were promising for our hy-
pothesis, experimental proof was not obtained. To confirm
or deny our initial hypothesis, we focused on generating spe-
cific structural changes in the amine part of this family of
phosphoramidite ligands (Table 7).[24]
By using L2a as the chiral ligand (Table 6, entries 1–3),
we were able to get high enantio- and regioselectivities in
favour of the branched products 16. Enantioselectivities re-
mained constantly high (97–98%), irrespective of the amine
15 used, and these reflected the values obtained with alkyl
The secondary amines were prepared very easily by con-
densation of the corresponding enantiopure primary amine
[25]
and ketone
(Scheme 19). Notably, amine 25 bearing a
mesityl group could not be synthesized if starting from a-
3602
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3596 – 3609

Iridium-Catalyzed Allylic Substitution
FULL PAPER
Me-benzylamine and the corresponding ketone. Under the
conditions used for all other amines, no formation of imine
occurred, probably for steric reasons. However, the inverted
strategy was successful, as seen in entry 8.
We evaluated these new ligands in the iridium-catalyzed
allylic alkylation reaction by using sodium dimethyl malo-
nate as the nucleophile. The kinetic data is presented in
Figure 2. As shown by Hartwig et al.,[10n] in the amination
process L1 gave only low conversion during the first two
hours of the reaction. Because it bears a methoxy substitu-
ent in the para-position, ligand L3 was used to indicate a
possible electronic effect of the OMe group. This did not
seem to be the case, as the kinetics of L3 were much slower
than those of L2a, and gave a slope similar to that of L1.
Next, we investigated the influence of a sole ortho-me-
thoxy substituent, as the calculations for this ligand (L4)
have already been performed.
Figure 1. Density functional theory model of L4.
Table 7. Synthesis of different secondary amines and the corresponding (S)-BINOL-based ligands.
The reaction showed a rate
significantly higher than that
for L1 and L3, but lower than
that for L2a. To confirm a pos-
sible coordination role of the
methoxy substituent, we re-
placed it by a simple methyl
group in the ortho-position,
as in L5. This ligand afforded
a kinetic slope similar to that
of L4, which suggests that
our first postulated model
Entry
Ar1
Ar2
Amine[a] (d.r.)
Yield of amine[b] [%]
Ligand
Yield of L [%]
1
2
3
4
5
6
7
8
o-MeOC6H4
p-MeOC6H4
Ph
o-MeOC6H4
p-MeOC6H4
o-MeOC6H4
o-CH3C6H4
o-CH3C6H4
1-naphthyl
xylyl
18 (82:18)
19 (89:11)
20 (82:18)
21 (92:8)
22 (94:6)
23 (94:6)
24 (88:12)
25 (98:2)
59
L2a
L3
L4
L5
L6
L7
L8
L9
94
32
34[d]
82
63
74
67
52
47
55[c]
64
Ph
o-CH3C6H4
1-naphthyl
Ph
64
34
25[d]
25[d]
mesityl
Ph
[a] In parenthesis, diastereomeric ratio measured by GC–MS. [b] Isolated yield after recrystallization of the
corresponding hydrochloric salt, unless otherwise stated. [c] Purification by silica gel chromatography.
[d] Yield not optimized.
Scheme 19.
Figure 2. Kinetics of Ir-catalyzed allylic alkylation with sodium dimethyl malonate, using catalysts with the corresponding ligands L.
Chem. Eur. J. 2006, 12, 3596 – 3609
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3603

A. Alexakis et al.
(Ir–OMe coordination) might be wrong. To confirm this, we
synthesized the C2-symmetrical ligand L6, bearing two
ortho-methyl substituents. This ligand showed even more
spectacular kinetics than L2a, enabling completion of the
reaction in less than two hours at room temperature! For
comparison, we studied the congener featuring the 1-naph-
thyl moiety L7 used by Hartwig et al.[10o,p] This gave a
slower reaction rate than that observed for L2a and L6.
Reaction using the ligand featuring a xylyl group, L8,
showed a kinetic slope similar to that for L1. However, the
mesityl group in L9 seemed to be deleterious both in terms
of kinetics and selectivity, giving the lowest ee (40%) and
the poorest regioselectivity (90:10) of the series.
tion, gave even better results. Such steric effects are well
known with phosphorus ligands (for example, Tol-BINAP
versus BINAP; Tol=toluene, BINAP=2,2’-bis(diphenyl-
phosphino)-1,1’-binaphthyl),[26] in which the ligand cone
angle q plays a crucial role.[27]
Furthermore, the position at which steric hindrance
occurs is of critical importance, as testified by the experi-
ments with ligands L8 and L9. Indeed, neither two Me
groups in the meta-position (as in L8) nor two Me groups in
the ortho-position of the same aromatic core could improve
the kinetics of the reaction. An h2 effect of the aryl group[28]
neighboring the amine part cannot be excluded; further
studies are currently in progress.
A similar study was conducted on the allylic amination
(Figure 3).[24] Again, the slowest reaction was obtained with
the ligand bearing a mesityl group, L9. Ligands that resulted
in gradual slopes were those bearing either no substituent,
L1, a xylyl group, L8, or a methoxy group in the para-posi-
tion, L3, as in the alkylation reaction. The two pseudo-C2-
symmetrical ligands, L4 and L5, gave more impressive kinet-
ic slopes than that of L1; L5 afforded a catalyst as efficient
as that with L2a. Ligand L6 showed impressive kinetic re-
sults, facilitating a faster reaction than that with L2a. Once
again, L7 was less efficient in terms of kinetics, although it
surpassed L2a.
Concerning the enantio- or regioselectivities, all the li-
gands gave results within the same range (up to >99:1, 92–
99% ee, respectively, Figure 4), except for L9, which we ex-
cluded from the figure (40% ee in alkylation and 0% ee in
amination reaction). This suggests that only the kinetics of
the reaction are dramatically affected by the steric bulk of
the amine moiety of the ligand. Nevertheless, ligand L6
These two kinetic studies reveal that our initial postula-
tion was wrong. Indeed, L6, bearing two methyl groups in
the ortho-position of the amine part of the ligand, gave even
more spectacular results than our, until now, best catalyst
derived from L2a. That suggests that there is a more subtle
explanation for the effect of the substituents in the ortho-po-
sition, as an electronic or even a coordination effect does
not seem to explain the improvement of the kinetics.
We can, however, postulate from this study that the ortho-
methoxy group plays a steric role, because ligand L6, bear-
ing no oxygen, but rather a methyl group at the same posi-
Figure 4. Recapitulative comparison of the ee values (%) resulting from
the Ir-catalyzed allylic alkylation with sodium dimethyl malonate and the
amination with benzylamine, with ligands L1–L8.
Figure 3. Kinetics of Ir-catalyzed allylic amination with benzylamine, using catalysts with the corresponding ligands L.
3604
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3596 – 3609

Iridium-Catalyzed Allylic Substitution
FULL PAPER
sodium b-ketoester solutions were prepared as follows: NaH (dispersion
in mineral oil, 50%, 0.049 g, 1 mmol) was rinsed with n-pentane (3”
5 mL), followed by dry THF (2”5 mL), and then suspended in dry THF
(2 mL). Neat malonate or b-ketoester (1 mmol) was added dropwise
under nitrogen and stirring at RT to give a colorless solution, which was
immediately used.
gave the best enantioselectivities for both reactions, 98.3
and 99.1%, respectively.
Conclusion
Computational calculation remarks: Geometry optimization and elec-
tronic structure calculation for the ligand L4 were performed by using
density functional theory as implemented in the Gaussian 03 program.[30]
The PW91 functional was used for treating both the exchange and corre-
lation effects. The relativistic ECP and its associate basis set, SBKJC
VDZ ECP, were employed to describe the iridium metal center, and all
others atoms were described by the 6–31G* Pople type basis. A frequen-
cy calculation was performed on the optimized structure to ensure it was
a minimum on the potential-energy surface.
Ligand L2a appears to be an extremely efficient ligand for
iridium-catalyzed allylic alkylation. In conjunction with [{Ir-
A
system that permitted the use of a wide range of electro-
philes (allylic carbonates or acetates) or nucleophiles (sec-
ondary or tertiary) with high enantio- and regioselectivities.
This general methodology could be extended to the devel-
opment by ring-closing metathesis of other interesting prod-
General procedure for the preparation of new secondaryamines : A mix-
ture of ketone (9 mmol), amine (9 mmol), and titanium(iv) isopropoxide
(8 mL, 27 mol) was stirred at RT for 20 min. The mixture was then hy-
drogenated at 1 atm with 10% palladium on charcoal (180 mg, 2 mol%)
under vigorous stirring at RT. The reaction course was monitored by
GC–MS. Upon complete conversion, the reaction mixture was treated
with an aqueous solution of 1m sodium hydroxide. After stirring for
10 min, the solution was extracted five times with ethyl acetate. The com-
bined organic layers were filtered over celite, dried over sodium sulfate,
and evaporated under reduced pressure. After dissolving the resulting oil
in diethyl ether (10 mL), an aqueous solution of 37% HCl was added
dropwise until a slightly acidic medium was reached, to afford the crude
hydrochloric acid salt of the amine. The latter was recrystallized by slow
evaporation of an ethyl acetate/methanol saturated solution followed by
desalinification by treatment with an aqueous solution of 1m NaOH, ex-
traction with dichloromethane, and drying over sodium sulfate for analy-
sis.
ꢀ
ucts, such as cyclic molecules. Similarly, C N bonds could
be created by using several amines as nucleophiles.
The results also provide mechanistic insight into the role
of the ortho-methoxy substituent in terms of kinetics. Our
first, and most obvious, hypothesis was a coordination role,
which was later confirmed by theoretical calculations. Nev-
ertheless, we designed several other ligands that revealed
that neither a coordination nor an electronic role could pro-
vide an explanation. Instead, we found a steric effect, as the
ligand bearing only methyl groups, L6, gave the most im-
pressive kinetics. We are studying this hypothesis further.
A
Experimental Section
general procedure described. Yield: 25%; [a]2D0 =ꢀ150.7 (c=1.25 in
CHCl3); 1H NMR (400 MHz, CDCl3): d=7.35–7.24 (m, 5H), 6.90 (s,
1H), 6.84 (s, 2H), 3.55 (q, J=6.8 Hz, 1H), 3.46 (q, J=6.8 Hz, 1H), 2.33
(s, 6H), 1.80–1.50 (brs, 1H), 1.30 (d, J=7.8 Hz, 3H), 1.28 ppm (d, J=
7.8 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=137.9, 128.5, 128.4, 126.8,
124.5, 55.2, 55.1, 25.0, 24.9, 21.4 ppm; MS(EI): m/z (%): 239 (15), 238
(79), 148 (10), 133 (98), 120 (37), 117 (10), 105 (100); HRMS(ES): m/z
calcd for C17H20N [MꢀHꢀCH3]+: 238.1596; found: 238.1596.
General remarks: 1H (400 MHz) and 13C (100 MHz) NMR spectra were
recorded in CDCl3 by using a Bruker 400F NMR spectrometer, and
chemical shifts (d) are given in ppm relative to residual CHCl3. Multiplic-
ity is indicated as follows: s (singlet), d (doublet), t (triplet), q (quartet),
m (multiplet), dd (doublet of doublets), dt (doublet of triplets), ddd
(doublet of doublet of doublets), qd (quartet of doublets), brs (broad sin-
glet). Coupling constants are reported in Hertz (Hz). The evolution of re-
actions was monitored by GC–MS(EI mode) with an HP6890. Optical
rotations were recorded by using a Perkin–Elmer 241 polarimeter at
208C in a 10-cm cell in the stated solvent; [a]D values are given in
10ꢀ1 degcm2 gꢀ1 (concentration c is given as g per 100 mL). Enantiomeric
excesses were determined by performing chiral supercritical-fluid chro-
matography (SFC) with a Berger SFC and the stated column. Gradient
programs are described as follows: initial methanol concentration (%),
initial time (min), percent gradient of methanol (%minꢀ1), final metha-
nol concentration (%); retention times (RT) are given in min.
A
the general procedure described. Yield: 25%; [a]2D0 =ꢀ129.1 (c=1.24 in
CHCl3); 1H NMR (400 MHz, CDCl3): d=7.27–7.20 (m, 5H), 6.83 (brs,
2H), 4.07 (q, J=7.0 Hz, 1H), 3.51 (q, J=6.8 Hz, 1H), 2.61 (brs, 3H),
2.31 (s, 3H), 1.76 (brs, 4H), 1.38 (d, J=7.0 Hz, 3H), 1.35 ppm (d, J=
6.6 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=135.5, 128.4, 126.8, 55.6,
51.0, 25.3, 20.8, 20.7 ppm; MS(EI): m/z (%): 267 (7) [M]+, 252 (36), 146
(100), 131 (17), 105 (76); HRMS(ES): m/z calcd for C19H25N [MꢀH]+:
267.1987; found: 267.1986.
General procedure for the preparation of new phosphoramidite ligands:
The hydrochloric acid salt of the amine (3.33 mmol) was added neat in
one portion to a stirred mixture of Et3N (16.67 mmol, 2.3 mL) and PCl3
(3.33 mmol, 0.29 mL) at 08C in dichloromethane (6 mL) under nitrogen
atmosphere. The reaction mixture was stirred for 3–5 h at RT until com-
plete disappearance of the PCl3 (monitored by 31P NMR spectroscopy).
Neat (S)-binaphthol (3.33 mmol, 0.95 g) was added in one portion to the
reaction mixture at 08C and the suspension was stirred at RT overnight.
The reaction was quenched with water, dried over sodium sulfate, and
purified by flash chromatography on silica gel (cyclohexane/ethyl acetate
or pentane/dichloromethane) affording the ligand as white foam.
Enantiomeric excesses were in some cases determined by chiral GC
measurement by using either an HP6890 (H2 as vector gas) or HP6850
(H2 as vector gas) with the stated column. Temperature programs are de-
scribed as follows: initial temperature (8C), initial time (min), tempera-
ture gradient (8Cminꢀ1), final temperature (8C); retention times (RT) are
given in min. Flash chromatography was performed by using silica gel
32–63 mm, 60 Š. THF and dichloromethane were dried by filtration over
alumina (activated at 3508C under nitrogen atmosphere for 12 h). Allylic
carbonates and acetates were synthesized by using known experimental
procedures.[29]
All chiral primary amines were obtained from BASF. Chiral secondary
amines 18–23 were prepared according to the procedure described previ-
ously.[25]
O,O’-(S)-1,1’-Dinaphthyl-2,2’-diyl-N,N’-(S,S)-1-phenylethyl-1-xylylethyl-
phosphoramidite (L8): Synthesized according to the general procedure
described. Yield: 67%; [a]2D0 =+5.4 (c=1.03 in CHCl3); 1H NMR
(400 MHz, CDCl3): d=8.06 (d, J=8.6 Hz, 1H), 7.97 (d, J=8.1 Hz, 1H),
7.86 (d, J=7.8 Hz, 1H), 7.78 (d, J=8.8 Hz, 1H), 7.63 (d, J=8.6 Hz, 1H),
7.48–7.23 (m, 12H), 6.86 (s, 1H), 6.80 (s, 2H), 4.42 (m, 2H), 2.27 (s, 6H),
1.74 (d, J=7.1 Hz, 3H), 1.71 ppm (d, J=6.8 Hz, 3H); 13C NMR
All chiral ligands L1–L7 were prepared according to the literature proce-
dure.[14, 15] Lithium chloride was dried for 24 h at 808C prior to use. [{Ir-
ACHTREUNG
Chem. Eur. J. 2006, 12, 3596 – 3609
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3605

A. Alexakis et al.
(100 MHz, CDCl3): d=149.9, 143.4, 142.7, 137.1, 133.0, 132.8, 131.4,
130.5, 130.4, 129.6, 128.3, 128.1, 127.7, 127.3, 127.2, 126.7, 126.1, 125.9,
124.8, 124.3, 122.6, 122.5, 121.2, 54.7, 54.6, 54.4, 26.6, 23.3, 22.9, 22.7,
21.4 ppm; 31P NMR (162 MHz, CDCl3): d=151.3 ppm.
129.0, 116.3, 114.0, 57.5, 55.2, 52.5, 48.9 ppm; MS(EI): m/z (%): 278 (14),
219 (35), 147 (100), 115 (13), 91 (16).
Dimethyl (R)-2-[1-(3-methoxyphenyl)allyl] malonate (3d): Product pre-
pared from substrate 2d and sodium dimethyl malonate according to the
general allylic alkylation procedure described. Yield: 58%. Ee was meas-
ured by chiral SFC analysis with a Regis (R,R) WHELK-O column (1%
MeOH, flow rate 2 mLminꢀ1, pressure 130 bars, 08C); RT: 5.95 (+), 6.54
O,O’-(S)-1,1’-Dinaphthyl-2,2’-diyl-N,N’-(S,S)-1-phenylethyl-1-mesityl-
ethylphosphoramidite (L9): Synthesized according to the general proce-
dure described. Yield: 52%; [a]2D0 =+300.8 (c=1.11 in CHCl3); 1H NMR
(400 MHz, CDCl3): d=8.01 (d, J=8.8 Hz, 1H), 7.92 (t, J=8.6 Hz, 2H),
7.87 (d, J=7.3 Hz, 1H), 7.56 (d, J=8.8 Hz, 1H), 7.54 (d, J=8.8 Hz, 1H),
7.43–6.94 (m, 11H), 6.55 (brs, 2H), 4.88 (m, 1H), 4.69 (m, 1H), 2.35
(brs, 6H), 2.14 (s, 3H), 1.74 (m, 3H), 1.61 ppm (d, J=7.3 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=150.7, 149.7, 136.5, 136.4, 135.9, 132.9,
131.5, 130.5, 130.4, 129.5, 128.4, 128.1, 127.8, 127.2, 127.1, 126.2, 126.1,
126.0, 124.8, 124.4, 122.6, 122.4, 121.3, 22.2, 21.9, 20.8, 20.6 ppm;
31P NMR (162 MHz, CDCl3): d=144.2 ppm.
1
(ꢀ); [a]2D0 =+28.9 (c=1.08 in CHCl3) for 96% ee; H NMR (400.13 MHz,
CDCl3): d=7.25 (td, J=7.8, 1.0 Hz, 1H), 6.85 (d, J=7.8 Hz, 1H), 6.81–
6.78 (m, 2H), 6.00 (ddd, J=17.2, 10.1, 8.3 Hz, 1H), 5.17 (dt, J=16.9,
1.3 Hz, 1H), 5.12 (d, J=10.4 Hz, 1H), 4.12 (dd, J=10.8, 8.3 Hz, 1H),
3.90 (d, J=10.8 Hz, 1H), 3.82 (s, 3H), 3.77 (s, 3H), 3.56 ppm (s, 3H);
13C NMR (100.59 MHz, CDCl3): d=168.2, 159.7, 141.6, 137.7, 129.7,
120.1, 116.7, 113.8, 112.4, 57.3, 55.2, 52.6, 52.5, 49.8 ppm; MS(EI): m/z
(%): 278 (28), 219 (100), 203 (10), 187 (38), 173 (10), 159 (40), 147 (86),
132 (12), 115 (34), 91 (35), 77 (13), 59 (12); elemental analysis: calcd (%)
for C15H18O5: C 64.74, H 6.52; found: C 64.49, H 6.48.
General procedure for the Ir-catalyzed allylic alkylation: In a flame-dried
Schlenk tube, lithium chloride (0.021 g, 0.5 mmol), chiral ligand
Dimethyl (R)-2-[1-(2-methoxyphenyl)allyl] malonate (3e): Product pre-
pared from substrate 2e and sodium dimethyl malonate according to the
general allylic alkylation procedure described. Yield: 98%. Ee was meas-
ured by chiral SFC analysis with a Regis (R,R) WHELK-O column (1%
MeOH, flow rate 2 mLminꢀ1, pressure 130 bars, 08C); RT: 5.67 (+), 6.30
(0.022 mmol), and [{Ir(cod)Cl}2] (0.0067 g, 0.01 mmol) were dissolved in
N
THF (0.5 mL) at RT. The resulting orange solution was stirred for 20 min
and treated with substrate (0.5 mmol) as well as a freshly prepared
sodium malonate or ketoester solution (1 mmol in 2.0 mL THF) at RT.
The reaction mixture was stirred at the indicated temperature until com-
plete disappearance of the starting material. The mixture was hydrolyzed
with water, extracted with diethyl ether, and dried over magnesium sul-
fate. GC–MSanalysis of the crude mixture indicated the ratio of re-
gioisomers. The product was concentrated in vacuo and subjected to
chromatography on silica gel (n-pentane/diethyl ether 7:3) to afford the
alkylated adduct as a colorless oil.
1
(ꢀ); [a]2D0 =+31.1 (c=1.09 in CHCl3) for 79% ee; H NMR (400.13 MHz,
CDCl3): d=7.20 (td, J=7.6, 1.8 Hz, 1H), 7.16 (dd, J=7.6, 1.5 Hz, 1H),
6.90–6.84 (m, 2H), 6.14 (ddd, J=17.0, 10.1, 8.6 Hz, 1H), 5.12 (dt, J=
17.2, 1.3 H, 1H z), 5.04 (dd, J=10.1, 0.8 Hz, 1H), 4.33 (dd, J=10.6,
8.6 Hz, 1H), 4.18 (d, J=10.6 Hz, 1H), 3.85 (s, 3H), 3.72 (s, 3H),
3.49 ppm (s, 3H); 13C NMR (100.59 MHz, CDCl3) 168.7, 168.3, 157.1,
136.9, 129.5, 128.3, 128.1, 120.7, 116.8, 111.1, 55.5, 55.4, 52.4, 52.3,
46.2 ppm; MS(EI): m/z (%): 278 (10), 219 (53), 187 (16), 159 (12), 147
(100), 131 (11), 115 (17), 91 (26); elemental analysis calcd (%) for
C15H18O5: C 64.74, H 6.52; found: C 64.46, H 6.51.
Dimethyl (R)-2-(1-phenylallyl) malonate (3a):[7] Product prepared from
substrate 2a and sodium dimethyl malonate according to the general al-
lylic alkylation procedure described. Yield: 82%. The absolute configura-
tion was determined by comparison with published data.[7] Ee was meas-
ured by chiral SFC analysis with a Regis (R,R) WHELK-O column (2%
Dimethyl (R)-2-(1-benzo[1,3]dioxol-5-yl-allyl) malonate (3 f): Product
R
prepared from substrate 2 f and sodium dimethyl malonate according to
the general allylic alkylation procedure described. Yield: 70%. Ee was
measured by chiral SFC analysis with a Regis (R,R) WHELK-O column
(1% MeOH during 6 min, then 1%minꢀ1, flow rate 2 mLminꢀ1, pressure
130 bars, 308C); RT: 6.02 (+), 6.77 (ꢀ); [a]2D0 =+22.0 (c=1.13 in CHCl3)
, , pressure
MeOH for 2 min, then 1%minꢀ1 flow rate 2 mLminꢀ1
130 bars, 08C); RT: 3.05 (R), 3.29 (S); [a]2D0 =+34.1 (c=1.08 in CHCl3)
for 98% ee; 1H NMR (400.13 MHz, CDCl3): d=7.40–7.23 (m, 5H), 6.03
(ddd, J=18.2, 10.1, 8.1 Hz, 1H), 5.15 (d, J=18.0 Hz, 1H), 5.12 (d, J=
10.6 Hz, 1H), 4.15 (dd, J=10.8, 8.1 Hz, 1H), 3.90 (d, J=11.1 Hz, 1H),
3.78 (s, 3H), 3.53 ppm (s, 3H); 13C NMR (100.59 MHz, CDCl3): d=
168.2, 167.8, 139.9, 137.8, 128.7, 127.9, 127.2, 116.7, 57.4, 52.6, 52.4,
49.8 ppm; MS(EI): m/z (%): 189 (98), 157 (19), 129 (43), 117 (100), 91
(17), 77 (10).
1
for 97% ee; H NMR (400.13 MHz, CDCl3): d=7.74–7.66 (m, 3H), 5.98–
5.89 (m, 3H), 5.10 (d, J=15.4 Hz, 1H), 5.07 (d, J=8.8 Hz, 1H), 4.03 (dd,
J=10.9, 8.1 Hz, 1H), 3.79 (d, J=10.9 Hz, 1H), 3.73 (s, 3H), 3.54 ppm (s,
3H); 13C NMR (100.59 MHz, CDCl3): d=168.3, 167.9, 139.9, 137.7, 137.5,
133.5, 132.6, 128.4, 127.9, 127.7, 126.7, 126.2, 126.1, 125.9, 116.9, 57.3,
52.7, 52.6, 52.5, 49.8 ppm; MS(EI): m/z (%): 298 (24), 239 (58), 207 (13),
179 (29), 167 (100), 152 (28); HRMS(EI) calcd for [ M+Na]+ 315.0845,
found 315.0834.
Dimethyl (R)-2-[1-(4-chlorophenyl)allyl] malonate (3g):[11c] Product pre-
pared from substrate 2g and sodium dimethyl malonate according to the
general allylic alkylation procedure described. Yield: 90%. Absolute con-
figuration was determined by comparison with published data.[7] Ee was
measured by chiral SFC analysis with a Regis (R,R) WHELK-O column
(2% MeOH during 2 min, then 1%minꢀ1, flow rate 2 mLminꢀ1, pressure
130 bars, 308C); RT: 3.30 (+), 3.82 (ꢀ); [a]2D0 =+40.9 (c=1.00 in CHCl3)
for 97% ee; 1H NMR (400.13 MHz, CDCl3): d=7.30 (d, J=8.3 Hz, 2H),
7.20 (d, J=8.3 Hz, 2H), 5.99 (ddd, J=16.9, 10.4, 8.1 Hz, 1H), 5.16 (d, J=
7.1 Hz, 1H), 5.12 (d, J=0.8 Hz, 1H), 4.13 (dd, J=10.9, 8.1 Hz, 1H), 3.85
(d, J=10.8 Hz, 1H), 3.78 (s, 3H), 3.56 ppm (s, 3H); 13C NMR
(100.59 MHz, CDCl3): d=168.0, 167.7, 138.5, 137.3, 133.0, 129.4, 128.8,
117.1, 57.2, 52.7, 52.6, 49.0 ppm; MS(EI): m/z (%): 223 (100), 191 (17),
163 (23), 151 (85), 128 (19), 115 (53), 59 (15).
Dimethyl (R)-2-(1-naphthalen-2-yl-allyl) malonate (3b):[11a] Product pre-
pared from substrate 2b and sodium dimethyl malonate according to the
general allylic alkylation procedure described. Yield: 73%. Ee was meas-
ured by chiral SFC analysis with a chiralcel OB-H column (2% MeOH
for 2 min, then 1%minꢀ1, flow rate 2 mLminꢀ1, pressure 200 bars, 308C);
RT: 7.43 (+), 8.05 (ꢀ); [a]D20 =+39.3 (c=0.925 in CHCl3) for 96% ee;
1H NMR (400.13 MHz, CDCl3): d=7.81–7.79 (m, 3H), 7.69 (s, 1H), 7.47–
7.44 (m, 2H), 7.37 (dd, J=8.6, 1.5 Hz, 1H), 6.08 (ddd, J=17.2, 10.4,
8.1 Hz, 1H), 5.18 (d, J=17.2 Hz, 1H), 5.04 (d, J=10.1 Hz, 1H), 4.30 (dd,
J=9.6, 8.4 Hz, 1H), 4.01 (d, J=11.1 Hz, 1H), 3.77 (s, 3H), 3.46 ppm (s,
3H); 13C NMR (100.59 MHz, CDCl3): d=168.3, 167.9, 137.7, 137.5,
133.5, 132.6, 128.4, 127.9, 127.7, 126.7, 126.2, 126.1, 125.9, 116.9, 57.3,
52.7, 52.5, 49.8 ppm; MS(EI): m/z (%): 298 (24), 239 (58), 207 (13), 179
(29), 167 (100), 152 (28).
Dimethyl (R)-2-[1-(4-methoxyphenyl)allyl] malonate (3c):[11c] Product
prepared from substrate 2c and sodium dimethyl malonate according to
the general allylic alkylation procedure described. Yield: 99%. Ee was
measured by chiral SFC analysis with a Regis (R,R) WHELK-O column
(2% MeOH for 2 min, then 1%minꢀ1, flow rate 2 mLminꢀ1, pressure
130 bars, 308C); RT: 4.99 (+), 5.61 (ꢀ); [a]2D0 =+23.3 (c=1.10 in CHCl3)
for 97% ee; 1H NMR (400.13 MHz, CDCl3): d=7.14 (d, J=8.6 Hz, 2H),
6.85 (d, J=8.6 Hz, 2H), 5.97 (ddd, J=17.1, 10.1, 7.8 Hz, 1H), 5.09 (d, J=
15.4 H z, 1H), 5.06 (d, J=8.6 Hz, 1H), 4.06 (dd, J=10.9, 8.3 Hz, 1H),
3.82 (d, J=10.9 Hz, 1H), 3.77 (s, 3H), 3.73 (s, 3H), 3.50 ppm (s, 3H);
13C NMR (100.59 MHz, CDCl3): d=168.3, 167.9, 158.6, 138.1, 131.9,
Dimethyl (R)-2-[1-(4-trifluoromethylphenyl)allyl] malonate (3h):[11c]
Product prepared from substrate 2h and sodium dimethyl malonate ac-
cording to the general allylic alkylation procedure described. Yield:
40%. Ee was measured by chiral GC analysis with a Chiraldex LIPO-
DEX E column (60-0-1-170-5); RT: 59.09 (ꢀ), 59.68 (+); [a]2D0 =+40.1
(c=0.81 in CHCl3) for 94% ee; 1H NMR (400.13 MHz, CDCl3): d=7.57
(d, J=8.1 Hz, 2H), 7.36 (d, J=8.3 Hz, 2H), 5.97 (ddd, J=16.9, 10.1,
8.1 Hz, 1H), 5.16 (d, J=7.6 Hz, 1H), 5.12 (s, 1H), 4.19 (dd, J=10.9,
3606
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3596 – 3609

Iridium-Catalyzed Allylic Substitution
FULL PAPER
8.6 Hz, 1H), 3.90 (d, J=11.1 Hz, 1H), 3.76 (s, 3H), 3.53 ppm (s, 3H);
13C NMR (100.59 MHz, CDCl3): d=167.8, 167.5, 144.1, 136.9, 129.6,
129.2, 128.3, 125.6–125.5 (q, J=3.3 Hz, 1C), 125.4, 117.5, 56.9, 52.7, 52.5,
49.3 ppm; MS(EI): m/z (%): 260 (10), 257 (100), 224 (20), 197 (26), 185
(98), 177 (11), 165 (31), 151 (10), 128 (11), 115 (48), 59 (37).
Dimethyl (R)-2-(1-propylallyl) malonate (3i):[32] Product prepared from
substrate 2i and sodium dimethyl malonate according to the general al-
lylic alkylation procedure described. Yield: 82%. Ee was measured by
chiral GC analysis with a Chiraldex G-TA column (70-0-0.5-110-15-170);
RT: 29.51 (+), 30.41 (ꢀ); [a]2D0 =+2.3 (c=1.12 in CHCl3) for 97% ee;
1H NMR (400.13 MHz, CDCl3): d=5.66 (dt, J=16.9, 10.1 Hz, 1H), 5.11
(m, 2H), 3.76 (s, 3H), 3.71 (s, 3H), 3.40 (d, J=9.1 Hz, 1H), 2.79 (qd, J=
9.3, 3.0 Hz, 1H), 1.45–1.21 (m, 4H), 0.91 ppm (t, J=6.8 Hz, 3H);
13C NMR (100.59 MHz, CDCl3) 168.8, 168.6, 138.1, 117.4, 57.0, 52.4, 52.2,
44.1, 34.5, 20.2, 13.8 ppm; MS(EI) m/z (%): 171 (41), 155 (100), 151
(32), 139 (66), 132 (69), 126 (30), 123 (29), 113 (49), 100 (48), 97 (10), 95
(36), 81 (50), 67 (41), 59 (51), 55 (83).
Dimethyl (S)-2-(1-cyclohexylallyl) malonate (3j):[31] Product prepared
from substrate 2j and sodium dimethyl malonate according to the gener-
al allylic alkylation procedure described. Yield: 65%. It was impossible
to remove the 6% of starting material that remained. Ee was measured
by chiral GC analysis with a Chiraldex G-TA column (80-0-1-170-5); RT:
44.67 (ꢀ), 45.00 (+); [a]2D0 =ꢀ0.9 (c=0.98 in CHCl3) for 98% ee and 6%
starting material; 1H NMR (400.13 MHz, CDCl3): d=5.73 (dt, J=16.9,
10.1 Hz, 1H), 5.08 (dd, J=10.1, 2.0 Hz, 1H), 5.03 (dd, J=17.2, 1.5 Hz,
1H), 3.73 (s, 3H), 3.68 (s, 3H), 3.63 (d, J=9.1 Hz, 1H), 2.60–2.69 (m,
1H), 1.72–0.88 ppm (m, 10H); 13C NMR (100.59 MHz, CDCl3): d=169.1,
168.8, 135.8, 118.2, 54.2, 52.4, 52.2, 50.1, 39.1, 28.6, 26.3 ppm; MS(EI): m/
z (%): 195 (11), 172 (20), 140 (11), 133 (14), 122 (100), 113 (72), 107 (29),
101 (20), 93 (26), 81 (90), 77 (13), 67 (52), 59 (31), 55 (75).
then 2%minꢀ1, flow rate 2 mLminꢀ1, pressure 175 bars, 308C); RT: 2.67
(Maj, S), 3.05 (Maj, R); [a]D20 =+27.1 (c=1.29 in CHCl3) for 95% ee;
1H NMR (400.13 MHz, CDCl3): d=7.34–7.18 (m, 5H; Maj+Min), 6.03–
5.90 (m, 1H; Maj+Min), 5.16–5.09 (m, 2H; Maj+Min), 4.16 (dd, J=
11.9, 8.3 Hz, 1H; Min), 4.06 (dd, J=15.2, 11.4 Hz, 1H; Maj), 3.76 (s, 3H;
Maj), 3.49 (s, 3H; Min), 2.32 (s, 3H; Min), 2.01 ppm (s, 3H; Maj);
13C NMR (100.59 MHz, CDCl3): d=201.7 (Min), 201.5 (Maj), 168.4
(Maj), 168.1 (Min), 140.1 (Maj), 139.8 (Min), 138.2 (Maj), 137.8 (Min),
128.9 (Maj), 128.7 (Min), 128.0 (Maj), 127.9 (Min), 127.2 (Maj), 127.1
(Min), 116.8 (Maj), 116.3 (Min), 65.2 (Min), 64.8 (Maj), 52.6 (Maj), 52.4
(Min), 49.6 (Min), 49.4 (Maj), 30.2 (Min), 29.8 ppm (Maj); MS(EI): m/z
(%): 214 (39), 189 (48), 173 (25), 157 (35), 129 (35), 117 (100); HRMS
(ESI) calcd for C14H17O3 [M+H]+ 233.1172, found 233.1155.
Diethyl (S)-2-allyl-2-(1-phenylallyl) malonate (8d):[18] Product prepared
from substrate 2a and nucleophile 7d according to the general allylic al-
kylation procedure described. Following chromatography, the product
still contained some allyldiethyl malonate, which was removed by kugel-
rohr distillation (808C, 0.1 mmHg) and afforded 0.134 g of a colorless oil
(85%). Ee was measured by chiral GC analysis of the cyclized metathesis
product 11; [a]2D0 =+55.3 (c=1.4 in CHCl3); 1H NMR (400.13 MHz,
CDCl3): d=7.27–7.20 (m, 5H), 6.46 (ddd, J=17.0, 10.2, 8.3 Hz, 1H),
5.87–5.76 (m, 1H), 5.16–5.02 (m, 4H), 4.25 (q, J=7.3 Hz, 2H), 4.19–4.16
(m, 2H), 4.05 (d, J=8.4 Hz, 1H), 2.63 (ddt, J=14.1, 6.6, 1.2 Hz, 1H),
2.45 (dd, J=14.2, 8.1 Hz, 1H), 1.30 ppm (t, J=7.1 Hz, 3H); 13C NMR
(100.59 MHz, CDCl3): d= 170.3, 170.1, 139.2, 138.0, 133.4, 129.4, 128.3,
127.2, 118.6, 117.1, 62.4, 61.2, 54.3, 39.4, 14.1, 14.0 ppm; MS(EI): m/z
(%): 316 (2) [M]+, 229 (14), 117 (100), 115 (29), 91 (21); HRMS(ESI)
calcd for C19H25O4 [M+H]+ 317.1747, found 317.1761.
Diethyl (S)-2-phenyl-cyclopent-3-ene-1,1-dicarboxylate (10):[18] Product
8d (0.25 mmol, 0.079 g) was dissolved in dichloromethane (5 mL) and
treated with Grubbꢁs first generation catalyst (0.0125 mmol, 0.010 g) for
1.5 h at RT. After solvent removal in vacuo, the crude product was sub-
jected to chromatography (SiO2, pentane/diethyl ether 7:3) to afford
0.069 g (96%) of 10 as a white product. M.p. 75–778C. Ee was measured
by chiral GC analysis with a Chiraldex BTA column (80-0-1-170-10); RT:
68.97 (R), 69.34 (S); [a]2D0 =ꢀ336.3 (c=0.82 in CHCl3) for 98.8% ee;
1H NMR (400.13 MHz, CDCl3): d=7.24–7.18 (m, 5H), 5.87–5.84 (m,
1H), 5.71–5.68 (m, 1H), 4.87 (d, J=1.3 Hz, 1H), 4.32–4.14 (m, 2H), 3.68
(dq, J=10.8, 7.1 Hz, 1H), 3.50 (dq, J=17.4, 2.3 Hz, 1H), 3.39 (dq, J=
10.6, 7.4 Hz, 1H), 2.78 (dd, J=17.7, 2.5 Hz, 1H), 1.25 ppm (t, J=7.1 Hz,
3H); 13C NMR (100.59 MHz, CDCl3): d=172.2, 169.6, 139.2, 132.3, 129.2,
128.6, 128.0, 127.2, 64.9, 61.6, 61.1, 56.9, 40.7, 14.1, 13.5 ppm; MS(EI):
m/z (%): 288 (21) [M]+, 242 (10), 214 (100), 197 (25), 169 (32), 155 (19),
141 (90), 128 (20), 115 (29); HRMS(ESI) calcd for C 17H21O4 [M+H]+
289.1434, found 289.1414.
Dimethyl (S)-2-(1,3-diphenylallyl) malonate (12):[33] Product prepared
from substrate 11 and sodium dimethyl malonate according to the gener-
al allylic alkylation procedure described. Yield 60%. Ee was measured
by chiral SFC analysis with a Regis (R,R) WHELK-O column (2%
MeOH during 2 min, then 1%minꢀ1, flow rate 2 mLminꢀ1, pressure
130 bars, 308C); RT: 2.83 (S), 3.13 (R); [a]2D0 =ꢀ11.2 (c=1.04 in CHCl3)
for 86% ee, lit. data: ꢀ22.4 (c=1.8 in CHCl3);[30] 1H NMR (400.13 MHz,
CDCl3): d=7.35–7.20 (m, 10H), 6.50 (d, J=15.8 Hz, 1H), 6.35 (dd, J=
15.8, 8.8 Hz, 1H), 4.28 (dd, J=10.8, 9.8 Hz, 1H), 3.97 (d, J=11.0 Hz,
1H), 3.72 (s, 3H), 3.53 ppm (s, 3H); 13C NMR (100.59 MHz, CDCl3): d=
168.2, 167.8, 140.1, 136.8, 131.8, 129.1, 128.7, 128.5, 127.8, 127.6, 127.2,
126.4, 57.6, 52.6, 52.4, 49.2 ppm; MS(EI): m/z (%): 324 (8) [M]+, 292 (8),
232 (12), 205 (81), 193 (86), 178 (26), 165 (13), 128 (13), 115 (100).
Dimethyl (S)-2-methyl-2-(1-phenylallyl) malonate (8a): Product prepared
from substrate 2a and nucleophile 7a according to the general allylic al-
kylation procedure described. Yield: 95%. Ee was measured by chiral
SFC analysis with
a Daicel OJ column (0% MeOH, flow rate
2 mLminꢀ1, pressure 200 bars, 458C); RT: 2.76 (S), 3.06 (R); [a]2D0 =+39.3
(c=1.4 in CHCl3) for 97% ee; 1H NMR (400.13 MHz, CDCl3): d=7.30–
7.2 (m, 5H), 6.32 (ddd, J=16.9, 10.2, 8.6 Hz, 1H), 5.12 (m, 2H), 4.15 (d,
J=8.6 Hz, 1H), 3.71 (s, 3H), 3.62 (s, 3H), 1.43 ppm (s, 3H); 13C NMR
(100.59 MHz, CDCl3): d=171.5, 171.3, 139.1, 136.9, 129.6, 128.2, 127.2,
117.8, 58.9, 54.6, 52.5, 52.4, 18.5 ppm; MS(EI): m/z (%): 262 (2) [M]+,
203 (30), 170 (10), 117 (100); HRMS(ESI) calcd for C 15H19O4 [M+H]+
263.1277; found: 263.1298.
Methyl (S)-2-oxo-1-(1-phenylallyl) cyclohexanecarboxylate (8b): Product
prepared from substrate 2a and nucleophile 7b according to the general
allylic alkylation procedure described. Yield: 79% as a mixture of the
two unseparable diastereoisomers in a 51:49 ratio. “Maj” refers to the
signals of the majority diastereoisomer, and “Min” to the signals of the
minor one. Ee was measured by chiral SFC analysis with a Daicel OJ
column (0% MeOH, flow rate 2 mLminꢀ1, pressure 175 bars, 308C); RT:
2.93 (Min, R), 3.39 (Min, S); [a]2D0 =+32.7 (c=0.87 in CHCl3) for 95%
ee; 1H NMR (400.13 MHz, CDCl3): d=7.42–7.21 (m, 5H; Maj+Min),
6.40–6.25 (m, 1H; Maj+Min), 5.16–5.08 (m, 2H; Maj+Min), 4.18 (d, J=
8.3 Hz, 1H; Maj+Min), 4.02 (d, J=9.4 Hz, 1H; Min), 3.62 (s, 3H; Min),
3.56 (s, 3H; Maj), 2.61–2.36 (m, 3H; Maj+Min), 2.01–1.56 ppm (m, 5H;
Maj+Min); 13C NMR (100.59 MHz, CDCl3): d=206.5 (Min), 206.3
(Maj), 171.2 (Min), 171.1 (Maj), 139.8 (Min), 139.5 (Maj), 137.2 (Min),
130.1 (Min), 129.8 (Maj), 128.1 (Maj), 128.0 (Min), 127.0 (Min), 126.8
(Maj), 117.5 (Maj), 117.4 (Min), 65.8 (Maj), 65.7 (Min), 54.2 (Min), 53.4
(Maj), 52.2 (Min), 52.1 (Maj), 42.0 (Maj), 41.9 (Min), 34.6 (Min), 33.0
(Maj), 27.0 (Min), 26.7 (Maj), 22.6 (Maj), 22.6 ppm (Min); MS(EI): m/z
(%): 272 (1) [M]+, 254 (18), 195 (16), 117 (100); HRMS(ESI) calcd for
C17H21O3 [M+H]+ 273.1485, found 273.1504.
Procedure for recycling of the iridium catalyst: After the normal course
of the reaction, the crude mixture was treated with 10 mL of a 1:1 mix-
ture of diethyl ether and pentane. The resulting solution was stored at
RT for about 8 h. An orange precipitate formed and settled to the
bottom of the solution. Filtration gave an orange powder that was rinsed
with diethyl ether/pentane (1:1, 3”). The catalyst was dried in vacuo and
could be used for another run.
Methyl (R)-2-acetyl-3-phenyl-pent-4-enoate (8c): Product prepared from
substrate 2a and nucleophile 7c according to the general allylic alkyla-
tion procedure described. Yield: 68% as a mixture of the two inseparable
diastereoisomers in a 53:47 ratio. “Maj” refers to the signals of the major-
ity diastereoisomer, and “Min” to the signals of the minor one. Ee was
measured by chiral SFC analysis with a Daicel OJ column (2% MeOH,
General procedure for the iridium-catalyzed enantioselective allylic ami-
nation: [{Ir
A
solved in THF (0.5 mL) in a 3-mL test tube under argon. A small mag-
Chem. Eur. J. 2006, 12, 3596 – 3609
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3607

A. Alexakis et al.
netic stirring bar was added, the test tube was capped with a septum, and
the mixture was stirred for 10 min at RT. Amine 15 (1.3 mmol) and cin-
namyl methyl carbonate 2a (0.18 mL, 0.98 mmol) were added to the re-
action mixture by using a syringe. The reaction mixture was stirred at RT
until the starting material had disappeared completely. The solvent was
removed in vacuo. GC–MSanalysis of the crude mixture indicated the
ratio of regioisomers 16/17. The mixture was then purified by flash
column chromatography on silica gel (cyclohexane/ethyl acetate 8:2) to
give the secondary amine as a white oil. A small amount of the product
was converted to the corresponding acetamide by treatment with acetic
anhydride in diethyl ether for 5 min. The acetamide was washed with
water and filtered on silica gel. Chiral SFC analysis indicated the enantio-
meric excess.
(S)-N-(1-Phenyl-2-propenyl)benzylamine (16a):[10m] Product prepared
from substrate 2a and amine 15a according to the general allylic amina-
tion procedure described. Yield: 88%. The absolute configuration was
determined by comparison of the optical rotation with literature data.[10m]
Ee was measured by chiral SFC analysis of the acetamide with a chiralcel
OB-H column (2% MeOH, flow rate 2 mLminꢀ1); RT: 5.99 (S), 7.00 (R);
[a]2D0 =+6.2 (c=1.12 in CHCl3) for 97% ee; 1H NMR (400.13 MHz,
CDCl3): d=7.30–7.22 (m, 10H), 5.99 (ddd, J=17.3, 10.2, 7.1 Hz, 1H),
5.26 (d, J=17.1 Hz, 1H), 5.17 (d, J=10.2 Hz, 1H), 4.27 (d, J=7.2 Hz,
1H), 3.81 (d of AB pattern, J=13.4 Hz, 1H), 3.75 (d of AB pattern, J=
13.4 Hz, 1H), 1.72 ppm (brs, 1H); 13C NMR (100.59 MHz, CDCl3): d=
142.8, 141.0, 140.5, 128.6, 128.4, 128.2, 127.4, 127.3, 126.9, 115.2, 65.2,
51.3 ppm; MS(EI): m/z (%): 223 (10) [M]+, 196 (38), 146 (22), 132 (56),
117 (35), 115 (17), 106 (13); HRMS(ESI +) calcd for C16H18N [M+H]+
224.1433, found 224.1438.
1999, pp. 833–886; c) P. J. Guiry, C. P. Saunders, Adv. Synth. Catal.
2004, 346, 497–537.
[2] H. Bricout, J.-F. Carpentier, A. Mortreux, Tetrahedron Lett. 1996,
37, 6105–6108.
[3] J. M. Brown, J. E. MacIntyre, J. Chem. Soc. Perkin Trans. 2 1985,
961–970.
[4] a) A. S. E. Karlstrçm, J.-E. Bäckvall in Modern Organocopper
Chemistry (Ed.: N. Krause), Wiley-VCH, Weinheim, 2002, pp. 259–
288; b) A. Alexakis, K. Croset, Org. Lett. 2002, 4, 4147–4149;
c) M. A. Kacprzynski, A. H. Hoveyda, J. Am. Chem. Soc. 2004, 126,
10676–10681; d) A. W. Van Zijl, L. A. Arnold, A. J. Minnaard, B. L.
Feringa, Adv. Synth. Catal. 2004, 346, 413–420.
[5] a) O. Belda, C. Moberg, Acc. Chem. Res. 2004, 37, 159–167;
b) B. M. Trost, I. Hachiya, J. Am. Chem. Soc. 1998, 120, 1104–1105.
[6] B. M. Trost, P. L. Fraisse, Z. T. Ball, Angew. Chem. 2002, 114, 1101–
1103; Angew. Chem. Int. Ed. 2002, 41, 1059–1061.
[7] G. C. Lloyd-Jones, A. Pfaltz, Angew. Chem. 1995, 107, 534–536;
Angew. Chem. Int. Ed. Engl. 1995, 34, 462–464.
[8] T. Hayashi, A. Okada, T. Suzuka, M. Kawatsura, Org. Lett. 2003, 5,
1713–1715.
[9] For a review on Ir, see: R. Takeuchi, Synlett 2002, 1954–1965.
[10] a) R. Takeuchi, M. Kashio, Angew. Chem. 1997, 109, 268–270;
Angew. Chem. Int. Ed. Engl. 1997, 36, 263–265; b) R. Takeuchi, M.
Kashio, J. Am. Chem. Soc. 1998, 120, 8647–8655; c) R. Takeuchi, K.
Tanabe, Angew. Chem. 2000, 112, 2051–2054; Angew. Chem. Int.
Ed. 2000, 39, 1975–1978; d) J. P. Janssen, G. Helmchen, Tetrahedron
Lett. 1997, 38, 8025–8026; e) B. Bartels, G. Helmchen, Chem.
Commun. 1999, 741–742; f) B. Bartels, C. Garcia-Yebra, F. Roming-
er, G. Helmchen, Eur. J. Inorg. Chem. 2002, 2569–2586; g) B. Bar-
tels, G. Helmchen, C. Garcia-Yebra, Eur. J. Org. Chem. 2003, 1097–
1103; h) K. Fuji, N. Kinoshita, T. Kawabata, K. Tanaka, Chem.
Commun. 1999, 2289–2290; i) T. Kanayama, K. Yoshida, H. Miyabe,
T. Kimachi, Y. Takemoto, J. Org. Chem. 2003, 68, 6197–6201; j) R.
Takeuchi, N. Ue, K. Tanabe, K. Yamashita, N. Shiga, J. Am. Chem.
Soc. 2001, 123, 9525–9534; k) T. Kanayama, K. Yoshida, H. Miyabe,
Y. Takemoto, Angew. Chem. 2003, 115, 2100–2102; Angew. Chem.
Int. Ed. 2003, 42, 2054–2056; l) H. Miyabe, K. Yoshida, M. Yamau-
chi, Y. Takemoto, J. Org. Chem. 2005, 70, 2148–2153; m) T.
Ohmura, J. F. Hartwig, J. Am. Chem. Soc. 2002, 124, 15164–15165;
n) F. Lopez, T. Ohmura, J. F. Hartwig, J. Am. Chem. Soc. 2003, 125,
3426–3427; o) C. Shu, J. F. Hartwig, Angew. Chem. 2004, 116, 4898–
4901; Angew. Chem. Int. Ed. 2004, 43, 4794–4797; p) C. Shu, A.
Leitner, J. F. Hartwig, Angew. Chem. 2004, 116, 4901–4904; Angew.
Chem. Int. Ed. 2004, 43, 4797–4800; q) C. Fischer, C. Defieber, T.
Suzuki, E. M. Carreira, J. Am. Chem. Soc. 2004, 126, 1628–1629.
[11] a) R. PrØtôt, A. Pfaltz, Angew. Chem. 1998, 110, 337–339; Angew.
Chem. Int. Ed. 1998, 37, 323–325; b) F. Glorius, M. Neuburger, A.
Pfaltz, Helv. Chim. Acta 2001, 84, 3178–3196; c) S.-L. You, X.-Z.
Zhu, Y.-M. Luo, X.-L. Hou, L.-X. Dai, J. Am. Chem. Soc. 2001, 123,
7471–7472; d) X.-L. Hou, N. Sun, Org. Lett. 2004, 6, 4399–4401;
e) O. Pamies, M. Dieguez, C. Claver, J. Am. Chem. Soc. 2005, 127,
3646–3647.
(S)-N-(1-Phenyl-2-propenyl)allylamine (16b):[10m] Product prepared from
substrate 2a and amine 15b according to the general allylic amination
procedure described. Yield: 91%. The ee was measured by chiral SFC
analysis of the acetamide with a chiralcel OB-H column (1% MeOH,
flow rate 2 mLminꢀ1); RT: 3.92 (+), 4.53 (ꢀ); [a]2D0 =+15.1 (c=1.23 in
CHCl3) for 97% ee; 1H NMR (400.13 MHz, CDCl3): d=7.45–7.26 (m,
5H), 5.91–6.02 (m, 2H), 5.29–5.13 (m, 4H), 4.27 (d, J=7.1 Hz, 1H),
3.29–3.19 (m, 2H), 1.50–1.30 ppm (brs, 1H); 13C NMR (100.59 MHz,
CDCl3): d=142.7, 140.9, 136.8, 128.6, 127.3, 127.3, 116.0, 115.2, 65.2,
49.9 ppm; MS(EI):
m/z (%): 173 (7) [M]+, 146 (100), 132 (48), 117 (74),
G
104 (16); HRMS(ESI +) calcd for C12H16N [M+H]+ 174.1277, found
174.1275.
(S)-N-(1-Phenyl-2-propenyl)-n-hexylamine (16c):[10m] Product prepared
from substrate 2a and amine 15c according to the general allylic amina-
tion procedure described. Yield: 89%. Ee was measured by chiral SFC
analysis of the acetamide with a chiralcel OD-H column (3% MeOH,
flow rate 1.8 mLminꢀ1); RT: 5.50 (+), 6.23 (ꢀ); [a]2D0 =+20.6 (c=1.21 in
CHCl3) for 98% ee; 1H NMR (400.13 MHz, CDCl3): d=7.40–7.26 (m,
5H), 5.98 (ddd, J=17.2, 8.6, 7.0 Hz, 1H), 5.25 (dt, J=17.2, 1.3 Hz, 1H),
5.14 (dt, J=10.4, 1.6, 1.3 Hz, 1H), 4.21 (d, J=7.0 Hz, 1H), 2.65–2.50 (m,
2H), 1.55–1.28 (m, 9H), 0.92 ppm (t, J=7.1 Hz, 3H); 13C NMR
(100.59 MHz, CDCl3): 143.1, 141.3, 128.5, 127.3, 127.1, 114.8, 66.3, 47.8,
31.8, 30.2, 27.1, 22.7, 14.1 ppm; MS(EI): m/z (%): 217 (3) [M]+, 190 (21),
146 (29), 132 (17), 117 (100), 115 (20); HRMS(ESI +) calcd for C15H24N
[M+H]+ 218.1903, found 218.1918.
[12] C. A. Kiener, C. Shu, C. Incarvito, J. F. Hartwig, J. Am. Chem. Soc.
2003, 125, 14272–14273.
[13] G. Lipowsky, N. Miller, G. Helmchen, Angew. Chem. 2004, 116,
4695–4698; Angew. Chem. Int. Ed. 2004, 43, 4595–4597.
[14] K. Tissot-Croset, D. Polet, A. Alexakis, Angew. Chem. 2004, 116,
2480–2482; Angew. Chem. Int. Ed. 2004, 43, 2426–2428.
[15] A. Alexakis, D. Polet, Org. Lett. 2004, 6, 3529–3532.
[16] A. Alexakis, S. Rosset, J. Allamand, S. March, F. Guillen, C. Ben-
haim, Synlett 2001, 1375–1378.
[17] a) E. J. Corey, F.-Y. Zhang, Org. Lett. 2000, 2, 4257–4259; b) O.
Belda, S. Lundgren, C. Moberg, Org. Lett. 2003, 5, 2275–2278.
[18] a) This product was synthesized in a racemic mixture by: P. A.
Evans, L. J. Kennedy, Tetrahedron Lett. 2001, 42, 7015–7018; b) The
chiral version of the same reaction for the dimethyl derivative was
recently reported by Helmchen in: S. Streiff, C. Welter, M. Schelw-
ies, G. Lipowsky, N. Miller, G. Helmchen, Chem. Commun. 2005,
2957–2959.
Acknowledgements
The authors wish to thank StØphane Rosset, Marco Tuveri, Krystel Le-
moine, and Xavier Rathgeb for analytical, synthetic, and experimental
help, as well as the Swiss National Research Foundation (grant No.
200020-105368) for financial support.
[1] a) B. M. Trost, C. Lee in Catalytic Asymmetric Synthesis, 2nd ed.
(Ed.: I. Ojima), Wiley-VCH, New York, 2000, pp. 593–649; b) A.
Pfaltz, M. Lautens in Comprehensive Asymmetric Catalysis, Vol. 2
(Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamoto), Springer, Berlin,
3608
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3596 – 3609

Iridium-Catalyzed Allylic Substitution
FULL PAPER
[19] E. L. Eliel S. H. Wilen, Stereochemistry of Organic Compounds,
Wiley, New York, 1994, p. 396.
[20] G. Consiglio, R. M. Waymouth, Chem. Rev. 1989, 89, 257–276.
[21] B. M. Trost, M. L. Crawley, Chem. Rev. 2003, 103, 2921–2943.
[22] D. Seebach, Angew. Chem. 1988, 100, 1685–1715; Angew. Chem.
Int. Ed. Engl. 1988, 27, 1624–1654.
[23] This argument was used to justify a reaction-acceleration effect in
polar solvents. See ref. [9].
[24] D. Polet, A. Alexakis, Org. Lett. 2005, 7, 1621–1624.
[25] A. Alexakis, S. Gille, F. Prian, S. Rosset, K. Ditrich, Tetrahedron
Lett. 2004, 45, 1449–1451.
son, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Ha-
segawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M.
Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma,
G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dap-
prich, A. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D.
Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G.
Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashen-
ko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W.
Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, J. A. Pople,
Gaussian 03, Revision C.02; Gaussian, Inc., Wallingford, CT (USA)
2004.
[26] a) W. J. Tang, X. M. Zhang, Chem. Rev. 2003, 103, 3029–3069; b) M.
McCarthy, P. J. Guiry, Tetrahedron 2001, 57, 3809–3844.
[27] C. A. Tolman, Chem. Rev. 1977, 77, 313–348.
[28] G. C. Lloyd-Jones, S. C. Stephen, M. Murray, C. P. Butts, S. Vyskocil,
P. Kocovsky, Chem. Eur. J. 2000, 6, 4348–4357.
[31] P. A. Evans, J. D. Nelson, Tetrahedron Lett. 1998, 39, 1725–1728.
[32] R. Takeuchi, M. Kashio, J. Am. Chem. Soc. 1998, 120, 8647–8655.
[33] J. Sprinz, G. Helmchen, Tetrahedron Lett. 1993, 34, 1769–1772.
[29] J. Lehmann, G. C. Lloyd-Jones, Tetrahedron 1995, 51, 8863–8874.
[30] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N.
Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Peters-
Received: September 26, 2005
Published online: February 2, 2006
Chem. Eur. J. 2006, 12, 3596 – 3609
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3609