High diversity on simple substrates: 1,4-dihalo-2-butenes and other difunctionalized allylic halides for copper-catalyzed SN2′ reactions

Article abstract of DOI:10.1002/chem.200801309

Enantioselective allylic alkylation with an organomagnesium reagent catalyzed by copper thiophene carboxylate (CuTC) was carried out on difunctionalized substrates, such as commercially available 1,4-dichloro-2- butene and 1,4-dibromo-2-butene, and on similar compounds of higher substitution pattern of the olefin for the formation of all-carbon chiral quaternary centers. The high regioselectivity obtained throughout the reactions favored good regiocontrol for the addition of phenyl Grignard reagents. Other difunctionalized substrates (allylic ethers and allylic alcohols) also underwent asymmetric S_N2′ substitution.

Full text of DOI:10.1002/chem.200801309

FULL PAPER
DOI: 10.1002/chem.200801309
High Diversity on Simple Substrates: 1,4-Dihalo-2-butenes and Other
Difunctionalized Allylic Halides for Copper-Catalyzed SN2’ Reactions
Caroline A. Falciola and Alexandre Alexakis*[a]
Abstract:
Enantioselective
allylic
tution pattern of the olefin for the for-
mation of all-carbon chiral quaternary
centers. The high regioselectivity ob-
tained throughout the reactions fa-
vored good regiocontrol for the addi-
tion of phenyl Grignard reagents.
Other difunctionalized substrates (al-
lylic ethers and allylic alcohols) also
A
ACHTUNGTRENNUNG
reagent catalyzed by copper thiophene
carboxylate (CuTC) was carried out on
difunctionalized substrates, such as
commercially available 1,4-dichloro-2-
butene and 1,4-dibromo-2-butene, and
on similar compounds of higher substi-
Keywords: allylic compounds
arylation · asymmetric catalysis ·
copper · nucleophilic substitution
·
underwent
asymmetric
SN2’
AHCTUNGTRENNUNG
substitution.
Introduction
There is a growing industrial and pharmaceutical need for
clean and efficient asymmetric methodologies to access opti-
cally and biologically active compounds. In this context, the
well-documented allylic alkylation is a very useful reaction.
It can be catalyzed by many transition metals with either
soft or hard nucleophiles.[1] Copper usually favors regioselec-
tive SN2’ anti displacements of leaving group with hard nu-
cleophiles, such as organometallic reagents (magnesium,[2]
zinc,[3] or aluminum[4] reagents). Over the past decade, the
literature has flourished with examples of efficient copper-
catalytic SN2’ procedures, which afford nearly perfect enan-
tioselectivities[5] on different types of substrates.
Scheme 1. Asymmetric allylic alkylation of bisfunctionalized 1,4-dihalo-
type substrates.[6]
In this context, we wish to report here our full results
with further studies on higher substituted 1,4-dihalo sub-
strates, with tri- and tetrasubstituted olefins. However, some
important issues remain and have been poorly covered by
the recent literature. The poor regioselectivity of allylic
Versatility in the resulting chiral compounds is a key issue
that has already been addressed by our group in an earlier
report concerning difunctionalized bishalo substrates
(Scheme 1).[6] These are small, simple, commercially avail-
able in both olefinic geometries, cheap, and can be used
without further purification. More importantly, the resulting
chiral synthons offer high versatility for subsequent derivati-
zation, by nucleophilic or electrophilic pathways upon the
remaining monohalide.
ACHTUNGTRENaNUGN rylation has rarely been overcome, with very recent excep-
tions. By taking advantage of the excellent regioselectivity
on such 1,4-dihalo substrates, we also investigated an aryla-
tion protocol.
However, we were initially interested in studying whether
the double-bond geometry of the allylic electrophile had an
influence on the reactivity and the stereochemical outcome
of the reaction. Initially, we considered that the double-
bond geometry would not control the facial approach of the
soft organocopper nucleophile, thus that E- or Z-configured
substrates would afford mirror-image products (Scheme 2).
[a] Dr. C. A. Falciola, Prof. Dr. A. Alexakis
Dꢀpartement de Chimie Organique
Universitꢀ de Genꢁve
Quai E-Ansermet 30, 1211 Genꢁve 4 (Switzerland)
Fax : (+ 41)22-3797-215
Results and Discussion
E-mail: alexandre.alexakis@chiorg.unige.ch
Influence of the double-bond geometry in asymmetric allylic
alkylation of 1,4-dihalo-2-butene: Literature precedents cov-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801309.
Chem. Eur. J. 2008, 14, 10615 – 10627
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10615

cis-configured substrates afforded generally lower enantiose-
lectivities than their trans counterparts with the small library
of ligands tested (L1–L16). An exception was noted with
Scheme 2. Nucleophilic SN2’ approach to E and Z olefins (LG=leaving
group).
ering the subject are scarce. Some examples were produced
by Hoveyda and co-workers, who showed that the reactivity
and enantioselectivities obtained are lower on the cis iso-
mers of aromatic allylic phosphate using either chiral pepti-
dic Schiff bases or diaminocarbene catalysts.[7] Diastereose-
lective procedures also suffer from the cis configuration of
the substrate as was reported by Breit and Breuninger with
the use of their chiral-directing ortho-diphenylphosphanyl-
ferrocene carboxylate (o-DPPF) unit.[8] However, some rare
exceptions to these observations were reported. In an early
case, Calꢃ and co-workers pub-
lished a diastereoselective pro-
cedure that used high excesses
of copper bromide with a slow
addition of a Grignard reagent
(4:1 Cu/Grignard reagent) in
THF.[9] Under these drastic con-
ditions, they observed complete
regio- and stereocontrol over
the allylic substitution on either
Z or E chiral heterocyclic sul-
Scheme 3. CuTC-catalyzed asymmetric allylic substitution of (E)- and (Z)-allylic chlorides 1 and 3.
fides. One of the more recent
exceptions was reported by
Okamoto and co-workers in the
addition of Grignard reagents to various allylic silylated
ethers using chiral diaminocarbene catalysts,[10] which pro-
moted a better asymmetric outcome on the (Z)-allylic sub-
strate. The desymmetrization of meso-cyclic allylic bisphos-
phates is the second example described by Gennari and co-
workers, which proceeds with good enantiomeric excess of
up to 98% ee on cis-type substrates.[11]
Thus far, these rare instances have not permitted conclu-
sions on the influence of the double-bond geometry on the
behavior of the copper-catalyzed asymmetric allylic substitu-
tion reaction to be drawn. Consequently, we decided to
investigate the parameters that govern the chiral outcome
for easily accessible substrates. Prior to our study, previous
examples had been recorded by our group[12c] for the com-
parative allylic substitution of cis- and trans-cinnamyl chlor-
ides (3) and their cyclohexyl analogue (1) under the set of
typical conditions used by our group (Scheme 3).[12] The
ligand L2, which promoted up to 90% ee for the addition of
ethylmagnesium bromide to (Z)-3. More importantly
though, the better ligand, L4, for the substitution of (E)-1
and (E)-3 induced only very poor enantioselectivities on the
Z-configured equivalents (values as high as 30% ee). More-
over, depending on the ligands used, the alkylcopper re-
agent would either recognize the leaving group or the g sub-
stituent. These latter observations inferred that the ligand
dictated the behavior of the copper reagent, forcing the
copper cluster to recognize the cis and trans substrate differ-
ently.
To further comprehend the different behavior of isomers
E and Z, and scope the field of difunctionalized allylic elec-
trophiles, we directed our attention towards a readily acces-
sible substrate: the commercially available 1,4-dichloro-2-
butene (5), which is marketed in both geometrical configu-
rations of the double bond (Scheme 4).
10616
www.chemeurj.org
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10615 – 10627

Difunctionalized Allylic Halides for Cu-Catalyzed SN2’ Reactions
FULL PAPER
Grignard reagent), we are able to control the g-position re-
acting site, according to the Duhamel postulate (the latent
trigonal center concept, Scheme 5).[13] Indeed, owing to dis-
Scheme 4. Comparative Cu-catalyzed addition of RMgX to (E)- versus
(Z)-1,4-dichloro-2-butene (5).
For substrate 5, We first considered the addition of cyclo-
hexylmagnesium chloride catalyzed by copper thiophene
carboxylate (CuTC) (Table 1). The Z isomer invariably af-
Table 1. Comparative CuTC-catalyzed (1 mol%) addition of cyclohexyl-
magnesium chloride on (E)- and (Z)-5 with chiral phosphoramidite li-
gands (1 mol%) (L*=L1–L8; Scheme 4).
Entry[a] Geometry[b] Ligand
R
Conv. [%][c] SN2’/SN2[d] ee [%][e]
Scheme 5. Duhamelꢄs latent trigonal center concept applied to Cu-cata-
lyzed asymmetric allylic alkylation (AAA).
1
2
3
4
5
6
7
8
E
Z
E
Z
E
Z
E
Z
E
Z
E
Z
E
Z
E
Z
6a 100
6a 87
6a 100
6a 95
6a 100
6a 86
6a 100
6a 86
6a 100
6a 92
6a 100
6a 91
6a 100
6a 93
6a 100
6a 71
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
75 (S)
31 (S)
52 (R)
8 (R)
L1 (R,SS)
L2 (S,SS)
L3 (R,SS)
L4 (S,SS)
L5 (RR)
L6 (RR)
L7 (RR)
L8 (SS)
similar electronic and steric properties of the substituents on
the sp2 reacting sites, two types of competing control from
the prostereogenic centers can be characterized, as was
postulated by Duhamel and Plaquevent following their ob-
servations for the enantioselective protonation of prochiral
enolates. By applying Duhamelꢄs concept to the copper-cata-
lyzed allylic substitution (Scheme 5), asymmetric reactions
taking place at the C=C double bond can behave according
to two distinct pathways. In the first, the E olefin will afford
mirror-image products of those formed by the Z-configured
substrate: the face selection of the reacting organocopper
cluster is controlled by the orientation of the b center,
namely neighboring-site control. The substituents on the re-
acting carbon center (g) are seen as one big latent group
and their positioning does not count in the facial approach
of the organocopper nucleophile. However, the nature of
the resulting enantiomer does not necessarily translate from
the geometry of the double bond. Indeed, in some instances,
the same enantiomer will arise from either E or Z configu-
rations in the substrate, and is thus named reaction-site
54 (S)
37 (S)
38 (R)
24 (S)
74 (R)
44 (R)
62 (R)
39 (R)
52 (R)
49 (S)
73 (S)
25 (S)
9
10
11
12
13
14
15
16
[a] Conditions: 5 (1 mmol), CuTC (1 mol%), and L* (1.1 mol%) in
CH2Cl2 (2 mL) at À788C with addition of RMgX in Et2O (1.1 equiv)
over 1 h. [b] Geometry of the double bond. [c]Conversion (Conv.) deter-
mined by GC–MS and 1H NMR spectroscopy. [d] Ratio determined by
GC–MS and 1H NMR spectroscopy. [e] Enantiomeric excess determined
by chiral GC.
forded 6a with lower levels of asymmetric control over the
allylic substitution (up to 49% ee with L7; Table 1,
entry 14). Pleasingly, the totally regiocontrolled allylic dis-
placement of the chloride was observed in every instance of
Table 1. On the other hand, when trans-5 was treated with
slow addition of cyclohexylmagnesium chloride (1.1 equiv)
in a reaction catalyzed by CuTC (1 mol%) and chiral ligand
L5 (1.1 mol%), the product (R)-6a was obtained in 89%
yield and 74% ee (Table 1, entry 9). Furthermore, ligands
L1 and L5 afforded the highest asymmetric induction on
(E)-5 with 75 and 74% ee, respectively (Table 1, entries 1
and 9).
The relative configuration of the products resulting from
the addition of the cyclohexylcopper nucleophile is random
and strongly depends on the structural features of the
ligand, although a major trend favors products of identical
mirror image. The major occurrences suggest that, with a
more sterically demanding nucleophile (i.e., a secondary
AHCTUNGTRENNUNG
ondary Grignard reagent, which is more sterically demand-
ing to the reaction system. With a view to completing this
comparative study, we looked into asymmetric SN2’ substitu-
tion with the aliphatic primary magnesium reagent,
PhCH2CH2MgBr (Table 2). The Z and E substrates of 5 be-
haved as in the previous cases with a net preferred asym-
metric control over the trans isomer. Ligand L1 afforded on
both isomers of 5 a unique SN2’ adduct with enantioselectivi-
ties up to 78% ee for (E)-5 and 39% ee for (Z)-5 (Table 2,
entries 1 and 2).
Interestingly, the different structural motif of the inter-
mediate copper reagent changes the configurational out-
come of the reaction. Instead of reaction-site control, mostly
occurring for the cyclohexyl allylic substitution, the primary
Chem. Eur. J. 2008, 14, 10615 – 10627
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10617

A. Alexakis et al.
Table 2. Comparative CuTC-catalyzed (1 mol%) addition of phenethyl-
magnesium bromide on (E)- and (Z)-5 with chiral phosphoramidite li-
gands (1 mol%) (L*=L1–L6; Scheme 4).
stable in our copper-catalytic system, even at room tempera-
ture over a period of one hour. As the organomagnesium re-
agent was slowly added, there were no significant changes in
either stereoselective parameters to suggest a destabilization
of the substrate in the reaction media and, thus, that any iso-
merization from cis to trans was taking place in the course
of the reaction.
Entry[a] Geometry[b] Ligand
R
Conv. [%][c] SN2’/SN2[d] ee [%][e]
1
2
3
4
5
6
7
8
E
Z
E
Z
E
Z
E
Z
E
Z
E
Z
7a 100
7a 100
7a 100
7a 100
7a 100
7a 100
7a 100
7a 100
7a 100
7a 100
7a 100
7a 100
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
100:0
78 (R)
39 (S)
29 (S)
38 (R)
3 (R)
L1 (R,SS)
L2 (S,SS)
L3 (R,SS)
L4 (S,SS)
L5 (RR)
L6 (RR)
According to the comparative results obtained in the
copper-mediated allylic substitution of the E- and Z-config-
ured substrates, lower enantioselectivities occurring on the
cis-allylic halides can be attributed to possible negative in-
teractions, which result in spatial obstructions of the two re-
acting partners, the chiral organocopper nucleophile and the
cis-prochiral substrate. Schemes 6 and 7 illustrate the two
different stereochemical outcomes for the identical facial
approach of a copper catalyst upon the difunctionalized sub-
strates (Z)-5 or (E)-5, respectively. For the reaction of sub-
strate (Z)-5, one observes that, at the early stage of p–cop-
per(I) complex formation, there is no facial differentiation
hitherto, and that the allylic displacements can give rise to
two enantiomers in the reductive elimination process.
Indeed, due to its inherent symmetry, the crucial asymmetric
differentiation step for the allylic alkylation of the cis-bisha-
lide (5) occurs at a later stage, and relies on the s-allyl spe-
cies formed during the oxidative addition of the metal to
either the g or g’ centers (Scheme 6).
5 (S)
57 (S)
36 (R)
60 (S)
27 (R)
63 (S)
21 (R)
9
10
11
12
[a] Conditions: same as those for Table 1. [b] Geometry of the double
bond. [c] Conversion (Conv.) determined by GC–MS and 1H NMR spec-
troscopy. [d] Ratio determined by GC–MS and 1H NMR spectroscopy.
[e] Enantiomeric excess determined by chiral GC.
phenethyl behaves in a radically different mode, affording
opposite enantiomers for the SN2’-allylic displacement of cis-
versus trans-prochiral electrophiles, no matter which chiral
ligand is used. This implies that, with a linear alkyl reagent,
the incoming organocopper reagent instead targets the leav-
ing group and its surrounding geometry, thus neighboring-
site control. From this line of screening, the tendency of the
enantioselectivity implies that
cis and trans substrates have
opposite affinities for the chiral
ligands. Indeed, it is most obvi-
ous when going from L4 to L6,
for which increasing ee values
for (E)-5 match decreasing ee
values for (Z)-5 (Table 2, en-
tries 7–12).
The obvious conclusion for
such results is that higher enan-
tioselectivities are achieved
with the trans isomer in each
case, rather than with the cis,
and that the regiocontrol to-
wards the g adduct is not affect-
ed in either case. Nevertheless,
a plausible source of poor enan-
tiomeric control could stem
from isomerization of the
double bond by the metal cata-
lyst or by the reaction media.
We therefore monitored the
Z/E ratio (before and during
the reaction) and the evolution
of the enantioselectivity during
the slow addition of cyclo-
Scheme 6. Cu-catalyzed alkylation of (Z)-5.
hexylmagnesium
chloride
(cHexMgCl) to the cis-5 sub-
strate. To our delight, we ob-
served that the Z substrate was Scheme 7. Cu-catalyzed alkylation of (E)-5.
10618
www.chemeurj.org
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10615 – 10627

Difunctionalized Allylic Halides for Cu-Catalyzed SN2’ Reactions
FULL PAPER
Conversely, the stereoselectivity of the SN2’ displacement
reaction on the E isomer of 5 is intimately correlated with
the facial approach of the copper(I) cluster, thus to its sub-
sequent d–p* bond formation. Both of the g and g’ oxidative
adducts produce the same enantiomers in the final reductive
elimination step (Scheme 7).
In short, both (E)- and (Z)-1,4-dichloro-2-butene belong
to different symmetry point groups (C2v and C2h, respective-
ly). This implies that, as pictured in Scheme 8, due to the ad-
loading to 2 mol% resulted in an improvement of the enan-
tioselectivity from 74 to 77% ee (Table 3, entries 2 and 4).
Additionally, the copper source could be changed to copper
bromide, which catalyzed the allylic substitution with com-
parable enantioselectivity (Table 3, entry 3). Generally, good
enantioselectivities of up to 85% ee were obtained for the
different additions of primary saturated Grignard reagents
to 5 with biphenol-based ligand L1 (Table 3, entries 7, 11,
and 14). As the ferrocenyl-based bidentate ligand taniaphos
L14 has been used with some success on other types of sub-
strates,[16] it was tested on the allylic compound 5 but
reached only poor enantioselectivities of 10 and 32% ee for
products 9a and 11a, respectively (Table 3, entries 10 and
13).
For comparison, the reactions were carried out on the
trans-1,4-dibromo derivative 8 with the same set of organo-
magnesium reagents, as displayed in Table 4. Although we
Table 3. Asymmetric allylic alkylation of 5 catalyzed by CuTC in CH2Cl2
at À788C (Scheme 9).
g/a[b] ee
[%][c]
Scheme 8. Different enantiotopic faces for trans- and cis-1,4-dihalo-2-
butene.
Entry
R
Product Ligand Conv.
[%][a]
1
2
cHex
cHex
cHex
cHex
PhCH2CH2
nBu
nBu
nBu
nBu
6a
6a
6a
6a
7a
9a
9a
9a
9a
9a
L1 >99 (98) 100:0 75 (S)
L5 >99 (89) 100:0 74 (R)
3[e]
4[f]
5
L5 >99
L5 >99
100:0 75 (R)
100:0 77 (R)
ditional mirror plane cutting the cis substrate at the epicen-
ter of the double bond, the g and g’ prochiral centers
become mirror images of one another, thus producing enan-
tiomorphic adducts. In conclusion, the oxidative addition of
the copper locks the fate of the enantiomeric outcome of
the reaction.
L1 >99 (98) 100:0 78 (R)
L1 >99
100:0 84 (À)
6
7[d]
8[d]
9
L1 >99 (55) 100:0 85 (À)
ent-L4 >99
L5 >99
L14 >99
100:0 81 (À)
100:0 81 (+)
100:0 10 (+)
10[d] nBu
11
12
13
14
15
tBuOCH
tBuOCH
tBuOCH CTHNUGTRENNUNG
T
G
L1 >99 (80) 100:0 85 (À)
trans-1,4-Dichloro- and trans-1,4-dibromo-2-butene—highly
potent, commercially available substrates: As noted in the
above-mentioned study conducted on cis versus trans sub-
strates, the double activation of 1,4-dichloro-2-butene (5),
induced by the two allylic reacting sites, awarded regiospeci-
ficity to the SN2’-allylic displacement reaction. These allylic
substrates are known to undergo clean SN2’ reactions in a
racemic fashion for various cuprates[14] and under copper-
catalyzed addition of Grignard reagents.[15] To confirm the
preliminary results, other Grignard reagents were added to
the trans isomer of the commer-
ent-L4 >99
L14 35
100:0 79 (À)
100:0 32 (+)
L1 >99 (88) 100:0 85 (À)
A
ent-L4 >99
100:0 85 (À)
[a] Conversion (Conv.) determined by GC–MS and 1H NMR spectrosco-
py (in parentheses, yield of isolated product after purification by column
chromatography on silica gel). [b] Ratio determined by GC–MS and
1H NMR spectroscopy. [c] Enantiomeric excess (and configuration or op-
tical rotation in parenthesis) determined by GC on chiral stationary
phase. [d] 3 mol% of CuTC and 3.3 mol% of L*. [e] CuBr used in place
of CuTC. [f] 2 mol% of CuTC and 2.2 mol% of L*.
cially available difunctionalized
allylic substrates, trans-1,4-di-
chloro-2-butene (5) and trans-
1,4-dibromo-2-butene
(8;
Scheme 9). Both substrates af-
forded g adducts regiospecifi-
cally and provided compounds
6–12 with moderate-to-high
enantioselectivities
and 4).
(Tables 3
As we proceeded with an op-
timization of the reaction con-
ditions for the addition of cy-
clohexylmagnesium chloride, a
slight increase of the catalyst Scheme 9. Asymmetric allylic alkylation on trans-1,4-dichloro- (5) and trans-1,4-dibromo-2-butene (8).
Chem. Eur. J. 2008, 14, 10615 – 10627
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10619

A. Alexakis et al.
Table 4. CuTC-catalyzed asymmetric allylic alkylation of 8 with RMgX
Table 5. Copper-catalyzed asymmetric allylic alkylation of 8 with n-dibu-
(Scheme 9).
tylzinc.
Entry
R
Product Ligand Conv.
[%][a]
g/a[b] ee
[%][c]
1
cHex
cHex
cHex
PhCH2CH2
PhCH2CH2
PhCH2CH2
nBu
6b
6b
6b
7b
7b
7b
9b
9b
L1
88
100:0 23 (S)
2[d]
3
L2 >99 (99) 100:0 63 (R)
ent-L3 >99 (83) 100:0 54 (R)
L1 >99 (94) 100:0 83 (À)
Entry CuX/x [mol%] L*/x [mol%] Conv. [%][a] g/a[b] ee [%][c]
4
5
6
7
8
9
10
11
12[d]
13[d]
14
15
L2 >99
ent-L4 >99
L1 >99
100:0 76 (+)
100:0 67 (À)
100:0 86 (À)
1
2
3
4
CuTC/3
CuBr/1
CuBr/1
CuOTf·C6H6/1
L1/6
L1/2
L3/2
L1/2
70
69
73
37
100:0 52 (À)
100:0 48 (À)
100:0
0
nBu
ent-L4
91 (80) 100:0 87 (À)
100:0 40 (À)
tBuOCH
tBuOCH
tBuOCH
2A
2A(CH2)2CH2 11b
2A(CH2)2CH2 11b
(CH2)2 10b
(CH2)2 10b
L1 >99
100:0 88 (À)
N
ent-L4 >99 (77) 100:0 92 (À)
[a]–[c] See Table 3 footnotes.
C
L14 >99
100:0 86 (+)
100:0 89 (À)
T
R
L1
ent-L4
75
T
E
93 (70) 100:0 94 (À)
The impressive regioselectivity observed throughout these
examples can be explained by a stabilization of the inter-
mediate CuIII species formed after the oxidative addition of
the reagent to the allylic substrate. If one refers to the ac-
cepted mechanism for the Cu-catalyzed SN2’ reaction de-
scribed by Bꢅckvall et al.,[17] the imperative for a regiospe-
cific branched product formation is a rapid reductive elimi-
nation to prevent any isomerization through a transitory p–
allyl CuIII species that would support formation of a thermo-
dynamic rearrangement product, the linear SN2 product. In
the presence of a halide in the e position, one can suppose
that this halide gives some electronic density to the empty d
orbitals of the electron-poor metal through p bonding. This
tethered intermediate would prevent loss of regioselectivity
(Scheme 10).
iBu
iBu
12b
12b
L1 >99 (79) 100:0 85 (À)
ent-L4 >99
100:0 84 (À)
[a]–[d] See Table 3 footnotes.
could not reach higher than 63% ee for the addition of the
cyclohexyl moiety with ligand L2 (entry 2), we were pleased
to observe that for all other linear Grignard reagents the di-
bromo substrate 8 afforded higher enantioselectivities than
its dichloro counterpart with either ligands L1 or ent-L4,
again with all reactions undergoing regiospecific g substitu-
tion. When subjecting substrate (E)-8 to a linear Grignard
reagent, such as n-butylmagnesium chloride, the asymmetric
induction was in line with the stereoselectivities obtained
with 1,4-dichloro-2-butene (5). With catalyst loading as low
as 1 mol%, both ligands L1 and ent-L4 promoted the asym-
metric substitution reaction with good selectivity ranging
from 86 to 87% ee (Table 4, entries 7 and 8). We were
pleased to see that under similar conditions, chiral homoal-
lylic bromide adducts with enantioselectivities as high as
92% ee could be achieved with the addition of the function-
alized Grignard reagent (4-tert-butoxybutyl)magnesium bro-
mide with ent-L4 to provide compound (À)-11b (Table 4,
entry 10). Again, ee values higher than 86% could not be
reached using the taniaphos-family ligand L14 (Table 4,
entry 11). Moreover, the terpenic bromocitronellene (À)-
10b was obtained with the best selectivity of this series in
94% ee using a CuTC/ent-L4 loading of 3 mol% (Table 4,
entry 13). Conversely, ligand L1 and ent-L4 induced lower
values of 85% ee with the slightly hindered isobutyl copper
reagent 12b (Table 4, entries 14 and 15). Overall, the dibro-
mo substrates 8 gave (similar or) better results in terms of
enantioselectivity than the dichloro derivatives 5 when the
two phosphoramidite ligands L1 and ent-L4 were used.
To compare the use of different Grignard reagents with
our procedure, we investigated a corresponding commercial-
ly available zinc reagent (n-dibutylzinc), which was tested
with a set of different copper salts, in THF at À408C. This
methodology was less reactive (maximum 73% conversion)
and did not afford as high enantioselectivities (up to 52% ee
with CuTC/L1) (Table 5, entry 1). Nevertheless, this proce-
dure was again exclusively g selective.
Scheme 10. Internal stabilization of the transient copper
the adjacent halide.
ACHTUNGTREN(NGNU III) species by
SN2’ arylation: Throughout the chemical literature the addi-
tion of aryl moieties has been a challenging target in asym-
metric synthetic allylation.[18] Bꢅckvall et al. showed that re-
gioselectivity was a key issue for the Cu-catalyzed addition
of phenylmagnesium bromide on allylic acetates and chlor-
ides, obtaining the nonregioselective a substituent in higher
yields with clean anti displacement of the leaving group.[19]
To counter the preference of the aryl Grignard towards the
a product, addition time of the organomagnesium reagent
and catalyst loading had to be increased dramatically to
obtain a slightly favored regioselective SN2’ adduct. Only re-
cently, Hoveyda and co-workers have reported one of the
rare examples of highly regio- and enantioselective arylation
on very specific vinylsilane substrates.[20]
10620
www.chemeurj.org
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10615 – 10627

Difunctionalized Allylic Halides for Cu-Catalyzed SN2’ Reactions
FULL PAPER
30% for the branched adduct
15, parallel to an increase of
the g selectivity to 90%. Fur-
ther decrease in the tempera-
ture of the reaction to À788C
impaired the enantioselective
Scheme 11. Asymmetric allylic arylation of difunctionalized substrates 5 and 8.
outcome, while affording
a
In view of our regioselective successes mentioned above,
we reasoned that the highly regiocontrolling substrates,
namely 1,4-dihalo-2-butenes 5 and 8, may control a prefer-
ential g insertion of the aryl moiety under copper catalysis
(Scheme 11).
Our first results under racemic conditions (Table 6, en-
tries 1 and 2) were very promising and showed in the case of
1,4-dibromo-2-butene (8), a clean and complete regiospecific
Table 7. Copper salt screening for the asymmetric allylic arylation of 8
(Scheme 11).[a]
Entry Cu salt
Ligand T [8C] Conv. [%][b] 15/16[c] ee [%][d]
1
2
3
4
5
6
7
8
9
CuTC
CuBr
CuBr·SMe2
CuACTHNUTRGENUG(N OTf)2
CuCN
CuBr2
CuCl
ent-L6 À50
ent-L6 À50
ent-L6 À50
ent-L6 À50
ent-L6 À50
ent-L6 À50
ent-L6 À50
ent-L6 À50
90
94
91
86
95
94
91
74
91
90:10
87:13
78:22
88:12
97:3
94:6
92:8
95:5
92:8
30 (+)
26 (+)
30 (+)
9 (+)
3 (+)
12 (+)
28 (+)
35 (+)
25 (+)
CuN
Table 6. CuTC-catalyzed allylic arylation of 1,4-dihalo-2-butenes 5 and 8
CuN
(Scheme 11).[a]
[a]–[d] See Table 6 footnotes.
Entry Substrate CuX/x
L*/x
T
Conv.
SN2’/
ee
[mol%]
[mol%] [8C] [%][b]
SN2[c]
[%][d]
1
2
3
4
5
6
7
8
5, Cl
8, Br
8, Br
8, Br
8, Br
8, Br
8, Br
8, Br
8, Br
8, Br
8, Br
8, Br
8, Br
CuTC
CuTC
–
–
À78 >99
À78 >99
80:20[e]
100:0
100:0
100:0
75:25
–
–
0
0
nearly regiospecific SN2’ displacement (Table 6, entry 12).
Completing our study by the screening of copper salts (at
À508C and in the presence of ligand ent-L6) showed that
CuTC/1
CuTC/1
CuTC/3
CuTC/3
CuTC/3
CuTC/5
CuTC/3
CuTC/3
CuTC/5
CuTC/3
CuTC/5
L1/1 À65 >99
L2/1 À65 >99
ent-L3/3 À35
ent-L4/3 À35
L6/3 À35
97
93
97
86
97
98
90
2 (À)
CuTC and CuACTHNUGTRENUNG(OAc)2 were the most efficient, the latter af-
78:22 15 (+)
80:20 22 (À)
fording up to 35% ee and 95% g selectivity (Table 7,
entry 8).
L11/5 À35
L13/3 À35
L14/3 À35
ent-L6/5 À50
76:24
9 (+)
9
36:64 13 (À)
90:10 3 (+)
90:10 30 (+)
99:1
20 (À)
90:10 25 (+)
Other attempts of arylation on the 1,4-dichloro derivative
5 or using other aryl sources (such as solutions of 4-methox-
yphenyl or 4-chlorophenyl Grignard reagents in diethyl
ether, or in situ generated zinc-based organocuprates from
zinc bromide and aryl Grignard reagents) furnished good-
to-high branched-to-linear ratios (up to 94% g selectivity),
without reaching higher than <10% ee. Nevertheless, the di-
functionalized substrate 8 is a promising substrate for a re-
gioselective allylic arylation and further studies should be
undertaken.
10
11
12
13
L6/3 À78 >99
L11/5 À50
90
[a] Conditions: 5 or 8 (1 mmol), CuTC (x mol%), and L* (x mol%) in
CH2Cl2 (2 mL) at À788C with addition of RMgX in Et2O (1.1 equiv)
over 1 h. [b]Conversion (Conv.) determined by GC–MS and 1H NMR
spectroscopy. [c] Ratio determined by GC–MS and 1H NMR spectrosco-
py. [d] Enantiomeric excess determined by chiral GC. [e] Linear SN2
product of arylation isolated was bromide (E)-16.
SN2’ arylation into 15. However, higher substitution patterns
were observed as minor byproducts formed in these reac-
tions, that is, 1,2-diphenylbut-3-ene and (E)-1,4-diphenylbut-
2-ene. These likely result from a second allylic arylation of
the preformed linear product 16, when an excess phenyl-
magnesium reagent is present. Our first attempts to induce
enantioselectivity, using our traditional SN2’ methodology
(CuTC, addition of ArMgX over 1 h) with chiral ligands L1
and L2 (Table 6, entries 3 and 4), produced exclusively the
branched aryl adduct 15 as a racemic mixture. After screen-
ing a variety of chiral phosphorous ligands and N-heterocy-
clic carbenes (NHCs),[21] the biphenol-based ligand L6 pro-
duced the chiral aryl adduct 15 with 22% ee at À358C
(Table 6, entry 7). From that point, we investigated different
temperatures (Table 6) and copper salts (Table 7). When
lowering the temperature to À508C in the presence of L6
(Table 6, entry 11), the enantiomeric excess was increased to
Derivatization of homoallylic chiral halides: The chiral syn-
thons, obtained through the catalyzed asymmetric allylation
of 1,4-dichloro- (5) and 1,4-dibromo-2-butene (8), bear re-
maining tunable functionalities. To further illustrate the syn-
thetic utility of this methodology, we derivatized the chiral
monohalides through electrophilic and nucleophilic
AHCTUNGTRENNUNG
ic means through transformation into an organometallic spe-
cies. To establish the absolute configuration of our chiral ad-
ducts, we attempted the reduction of chloride (+)-7a
(63% ee) into (+)-17 by several means. The Grignard proce-
dure worked best, although conversion was not complete
after 24 h (Scheme 12). Nevertheless, following hydrolysis of
the organomagnesium reagent, compound (+)-17 was ob-
tained with complete retention of its chiroptical information.
Further ozonolysis and reduction afforded (S)-2-methyl-4-
Chem. Eur. J. 2008, 14, 10615 – 10627
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10621

A. Alexakis et al.
Scheme 12. Reduction of chiral homoallylic chloride (+)-7a.
phenylbutan-1-ol (18) with 38% yield, which allowed us to
determine its absolute stereochemistry by optical rotation.
However, a poor reactivity of the chloride can be avoided
through transformation into the corresponding iodide (from
a Finkelstein reaction of the chiral chloride, see Scheme 14
below). Chiral bromides underwent similar treatments. The
formation of a Grignard reagent from the chiral cyclohexyl
adduct (+)-6b with 56% ee and subsequent addition of allyl
bromide led to compound 19 with complete retention of the
chiral information (Scheme 13).[22] Hydrolysis of a sample of
the organomagnesium reagent produced but-3-en-2-ylcyclo-
hexane (20) with 53% ee. Consequently, by correlation with
previous studies, we could assign its absolute configura-
tion.[23] It should be mentioned that the racemic bromide
12b had already been transformed into a Grignard reagent
and used by Jennings-White and Almquist for the synthesis
of analogues of keto-dipeptides.[15]
Scheme 13. Electrophilic derivatization of chiral homoallylic bromides 6b
by using a Grignard reagent.
um intermediate, and once more we did not observe any
loss in the enantiomeric excess (81% ee).
In parallel, organolithium reagents were prepared by
treatment of the chiral homoallylic bromide (+)-11b, or,
better, its iodide counterpart (+)-21, with tert-butyllithium
at À788C (Scheme 14). Subsequent treatment with diphenyl-
disulfide afforded 22, which was further oxidized with hy-
drogen peroxide into the chiral sulfone (+)-23 with com-
plete stereoretention. Chiral aldehyde (+)-24 was obtained
by a dimethylformamide (DMF) quench of the organolithi-
Similar derivatization products could be assessed through
more straightforward nucleophilic procedures. Indeed, the
brominated cyclohexyl adduct, 1-bromobut-3-en-2-ylcyclo-
hexane ((+)-6b) with 51% ee, was transformed into the
hepta-1,6-dien-3-ylcyclohexane (19) in 64% yield with no
loss of enantioselectivity by adding allylmagnesium bromide
(Scheme 15). Previous findings from our research group
showed that product 19 can undergo clean ring-closing
Scheme 14. Transformation of chiral homoallylic bromide (+)-11b by using an organolithium reagent.
10622
www.chemeurj.org
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10615 – 10627

Difunctionalized Allylic Halides for Cu-Catalyzed SN2’ Reactions
FULL PAPER
lected ligand. Copper-catalyzed
reactions were carried out with
an array of aliphatic and func-
tionalized Grignard reagents: n-
butyl (a), phenethyl (b), and
tert-butoxybutyl (c) magnesium
bromides. The results are sum-
marized in Table 8. As expect-
ed, when adding the n-butyl
Grignard reagent to 26 in the
presence of ligand L2, product
27a was formed preferentially
(tertiary-to-quaternary
96:4) with 86% ee (Table 8,
entry 2). The minor quaternary
ratio
Scheme 15. Nucleophilic derivatization of chiral homoallylic bromides 6b and 11b.
product 28a (4%) was obtained
metathesis with retention of enantiomeric excess.[24] Con-
versely, adduct (+)-11b with 80% ee was derivatized
through the addition of benzenethiolate in a good yield of
75% to afford compound 22, which was further oxidized to
give the chiral sulfone (+)-23 with an optical purity of
82% ee (Scheme 15).
with 26% ee. Similar behaviors were noted with L2 for the
other Grignard reagents. Chiral compounds 27b and 27c
were obtained with equally good regio- and enantioselectivi-
ties (Table 8, entries 7 and 10), with tertiary selectivities of
94 and 98%, respectively, and ee values of 87% for both ad-
ducts. Under such conditions, the reagent evidently recog-
nizes the substrate as a b-disubstituted allylic electrophile.
However, when using the chiral bis-orthomethoxy phosphor-
Tri- and tetrasubstituted olefins—formation of chiral quater-
nary centers: As we have achieved good enantioselectivities
on the simple b,g-monosubstituted 1,4-dibromo-2-butene
(8), we were interested in testing our methodology on more-
substituted olefinic patterns. Indeed, our group was the first
to disclose a highly regio- and enantioselective methodology
for the allylic substitution of a broad range of b-disubstitut-
ed allylic electrophiles.[8,25] In such cases and with as low as
3 mol% CuTC in combination with binaphthol-based phos-
phoramidite ligand (S,SS)-L2,
ACHTUNGTRENNUNG
remarkable ee values as high as
>99% were reached.
We combined both aspects
into a single allylic electrophile,
the trisubstituted and unsym-
metrical
methylbut-2-ene
(E)-1,4-dibromo-2-
(26;
Scheme 16), which was pre- Scheme 16. Unsymmetrical substrate 26 used in Cu-catalyzed AAA.
pared from isoprene by using
referenced procedures and was
Table 8. Allylic substitution of 26 catalyzed by CuTC (5 mol%) and L* (5.5 mol%) in CH2Cl2 at À788C
(Scheme 16).
used without further purifica-
tion.[26] Interestingly, products
arising from the g or g’ substitu-
tion of this challenging unsym-
metrical dibromide 26 (i.e., pri-
mary versus secondary sites)
would evidently afford different
products, whilst producing in
the latter case an all-carbon
quaternary center. We antici-
pated that such an allylic elec-
trophile would react regioselec-
tively towards the incoming nu-
Entry
R
Ligand
Conv. [%][a]
27/28[a]
ee [%] in 27
ee [%][a] in 28
1
2
3
4
5
6
7
8
nBu (a)
nBu (a)
nBu (a)
nBu (a)
L1
L2
ent-L4
L9
L10
L1
L2
ent-L4
L1
L2
ent-L4
100
100
100
100
100
100
100
100
84
69:31
96:04
82:18
68:32
94:06
75:25
94:06
94:06
72:28
98:02
84:16
4
86
50
2
82
29
87
54
8
56
26
69
61
47
62
6
32
59
32
67
nBu (a)
PhCH2CH2 (b)
PhCH2CH2 (b)
PhCH2CH2 (b)
tBuOCH
tBuOCH
tBuOCH
9
10
11
2A
2A(CH2)2CH2 (c)
2A(CH2)2CH2 (c)
G
100
80
87
56
E
[a] Conversion, product ratio, and enantiomeric excess values of SN2’ adducts were determined by chiral GC
cleophile depending on the se- analysis.
Chem. Eur. J. 2008, 14, 10615 – 10627
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10623

A. Alexakis et al.
as observed for the monosubstituted dibromide 8, the reac-
tion was once more highly SN2’-specific and clean.[27]
To further challenge our system, we attempted the first re-
ported asymmetric Cu-catalyzed alkylation of a difunctional-
ized tetrasubstituted allylic electrophile, namely, (E)-1,4-di-
bromo-2,3-dimethylbut-2-ene (29) (Scheme 17). The latter
(Z)-31 with cyclohexylmagnesium reagent proved less ste-
reoselective with up to 43% ee with L3, in comparison to
the reaction carried out on the (E)-31 isomer, which afford-
ed a clean branched adduct 32 (g/a 100:0) with 82% ee. In
comparison, addition of the smaller ethyl- and methylmag-
nesium bromide to 31 produced (À)-33 and (+)-34 with 87
and 70% ee, respectively. In these last cases, the regioselec-
tivity decreases slightly to 88 and 77% g selectivity for the
ethyl and methyl adducts, respectively.
Another interesting, small, and readily available difunc-
tionalized substrate was thought of. Indeed, an old proce-
dure developed by Colonge and Poilane yields diverse allylic
alcohols by the treatment of unprotected (Z)-4-chlorobut-2-
en-1-ol (35) with organomagnesium reagents.[29] Inspired by
this study, we investigated a similar reaction under the
copper-catalytic SN2’ conditions (Scheme 19).
Scheme 17. Formation of chiral quaternary centers from 29.
was prepared from 2,3-dimethylbuta-1,3-diene and collected
after pentane precipitation as a 90:10 E/Z mixture without
further purification. As illustrated in Table 9, CuTC-cata-
lyzed SN2’ reaction on 29 with a slow addition of the phene-
thylmagnesium reagent (1.1 equiv) over an hour afforded
clean g-substitution product 30 (g/a 100:0). However, enan-
tiomeric excess values for the latter were disappointing, with
the better ligand (R,SS)-L1 affording at the most 30% ee
(Table 9, entry 1).
More than two equivalents of the magnesium reagent
were necessary to obtain a complete conversion of 35
(Table 10, entries 1 and 2), the first leading to deprotonation
and thus in situ protection of the free alcohol. Analogous
SN2’ reactions of 35 were carried out enantioselectively by
the slow addition of excess phenethylmagnesium bromide,
catalyzed by CuTC (3 mol%) and chiral ligands L1–L4
(3.3 mol%). The moderate optical purities of the newly
formed chiral alcohol 36 (44% ee at the most with ent-L4,
Table 10, entry 5) are consistent with the poor results ob-
tained previously on the cis-difunctionalized substrates (see
above). Moreover, the reaction does not proceed with high
Table 9. Allylic substitution of 29 catalyzed by CuTC and chiral phos-
phoramidite ligands L* in CH2Cl2 at À788C (L*=L1–L33; Scheme 17).
regiocontrol, affording at the most 71%
(Table 10, entry 4).
g selectivity
Entry
Product
Ligand
Conv. [%][a]
SN2’/SN2[a]
ee [%][a]
1
2
3
4
5
6
7
8
30
30
30
30
30
30
30
30
L1
L2
ent-L3
ent-L4
L12
L19
L16
L15
100 (60)
100
100
100
100
100
96
100:0
100:0
100:0
100:0
100:0[b]
100:0
100:0
100:0
30 (À)
24 (+)
20 (+)
10 (À)
12 (+)
27 (À)
20 (+)
5 (+)
Conclusion
Difunctionalized allylic substrates are excellent starting ma-
terials for the elaboration of more-complex chiral synthons.
From these studies it is clear that E-configured substrates
afford higher enantioselectivites than their Z counterparts.
Of particular interest is (E)-1,4-dibromo-2-butene, because
it is an inexpensive commercially available compound (even
less expensive than the chloro substrate in the Aldrich cata-
log), and it affords high enantioselectivities. In addition, the
resulting adducts can be further used as electrophilic or nu-
cleophilic reagents.
94
[a] Conversion, product ratio, and enantiomeric excess values of SN2’ ad-
ducts were determined by chiral GC analysis (in parentheses, yield of the
isolated product after purification by flash chromatography over silica
gel). [b] 49% of double-substitution byproducts were observed with a g/a
ratio of 64:36.
Other difunctionalized substrates—formation of chiral ho-
moallylic alcohol derivatives:
Bifunctional allylic substrates,
in the form of protected allylic
alcohols, silyl ethers,[10,28] or
benzyl[16a,28] ethers, have previ-
ously been used in copper-cata-
lyzed SN2’ substitution. Our
trials to alkylate regio- and
enantioselectively the benzyl
ether 31 are summarized in
Scheme 18. Confirming our
prior results on cis-type sub-
strates, the allylic alkylation of Scheme 18. Cu-catalyzed AAA on (E)- and (Z)-31.
10624
www.chemeurj.org
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10615 – 10627

Difunctionalized Allylic Halides for Cu-Catalyzed SN2’ Reactions
FULL PAPER
Scheme 19. Asymmetric copper-catalyzed SN2’ reaction on unprotected allylic alcohol (Z)-35.
residue was purified by means of flash column chromatography. Gas
chromatography on a chiral stationary phase showed the enantiomeric
excess of the SN2’ product.
Table 10. CuTC-catalyzed allylic alkylation on (Z)-4-chlorobut-2-enol
(35) with phenethylmagnesium bromide (Scheme 19).
Entry
Equiv. RMgX
Ligand
Conv. [%][a]
36/37[a]
ee [%][b]
(S)-(1-Chlorobut-3-en-2-yl)benzene (13): Rf =0.88 (silica gel, pentane);
1H NMR (400 MHz, CDCl3): d=7.37–7.32 (m, 2H), 7.29–7.23 (m, 3H),
1[c]
2[c]
3
4
5
1.2
2.5
2.5
2.5
2.5
rac-L6
rac-L6
L1
L2
ent-L4
<50
n.d.[d]
53:47
57:43
71:29
65:35
–
–
34
26
44
100 (14)
100 (32)
100 (31)
100 (15)
1
2
3
2
6.02 (ddd, J=7.3 Hz, J=10.4 Hz, J=17.4 Hz, 1H), 5.23 (d, J=10.3 Hz,
1H), 5.17 (d, 3J=17.2 Hz, 1H), 3.79 (d, J=2.0 Hz, 1H), 3.77 (s, 1H),
3.70–3.65 ppm (m, 1H); 13C NMR (100 MHz, CDCl3): d=140.8, 138.1,
128.8, 128.8, 127.9, 127.9, 127.3, 117.2, 52.0, 47.9 ppm; the enantiomeric
excess was measured by chiral GC with a Hydrodex B6-TBDM, hydro-
gen flow (program: 70–0–1–170); tR =38.30 (À), 40.98 (+).
[a] Conversion and product ratios determined by 1H NMR spectroscopy
(in parentheses, yield of isolated product after purification by flash chro-
matography on silica gel). [b] Enantiomeric excess of branched adduct
determined by chiral GC analysis. [c] The racemic equivalent of phos-
phoramidite ligand L6 was used. [d] Not determined.
(+)-(1-Bromobut-3-en-2-yl)benzene (15): Rf =0.79 (silica gel, pentane);
[a]2D2 =+12.0 (c=0.28, CHCl3) for 35% ee; 1H NMR (400 MHz, CDCl3):
d=7.38–7.34 (m, 2H), 7.30–7.29 (m, 1H), 7.28–7.22 (m, 2H), 6.03 (ddd,
1J=10.4 Hz, 2J=7.3 Hz, 3J=17.4 Hz, 1H), 5.22 (dd, 1J=10.4 Hz, 4J=
3
4
1.0 Hz, 1H), 5.17 (dd, J=17.2 Hz, J=1.3 Hz, 1H), 3.71 (m, 1H), 3.65 (s,
1H), 3.63 ppm (d, J=1.8 Hz, 1H); 13C NMR (100 MHz, CDCl3): d=
141.3, 138.6, 128.9, 128.9, 127.8, 127.8, 127.4, 117.2, 51.9, 36.5 ppm; IR
(neat): n˜ =3063 (w), 2959 (w), 2923 (w), 1873 (w), 1640 (w), 1601 (w),
1493 (m), 1453 (m), 1416 (w), 1258 (w), 1220 (m), 1074 (w), 989 (m), 920
(s), 762 (m), 748 (m), 698 (s), 650 cmÀ1 (m); MS (EI mode): m/z (%): 212
(21), 210 (21), 154 (17), 131 (22), 118 (20), 117 (100), 116 (14), 115 (42),
91 (30), 77 (16), 51 (19); HRMS (EI mode): m/z calcd for C10H10Br:
210.0044; found: 210.0044; the enantiomeric excess was measured by
chiral GC with a Chirasil-Dex CB, helium flow (program: 70–0–1–170);
tR =47.55 (À), 48.33 (+).
Experimental Section
General remarks: 1H (400 MHz) and 13C (100 MHz) NMR spectra were
recorded on a Bruker 400F NMR spectrometer in CDCl3 unless other-
wise stated, and chemical shifts (d) are given in ppm relative to residual
CHCl3. Multiplicity is indicated as follows: s (singlet), d (doublet), t (trip-
let), q (quartet), m (multiplet), dd (doublet of doublet), and dt (doublet
of triplet). IR spectra were recorded on a Perkin-Elmer FTIR spectrome-
ter. The evolution of reaction was followed by means of TLC and GC–
MS (EI mode) on an HP6890 instrument. Optical rotations were record-
ed on a Perkin-Elmer 241 polarimeter at 208C in a 10 cm cell in the
stated solvent; [a]D values are given in 10À1 degcm2 gÀ1 (concentration c
is given as g/100 mL). Enantiomeric excesses were determined by means
of chiral GC measurements either on a HP6890 (H2 as vector gas) or
HP6850 (H2 or He as vector gas) instrument with the stated column.
Temperature programs are described as follows: initial temperature
[8C]—initial time [min]—temperature gradient [8CminÀ1]—final temper-
ature [8C]; retention times (tR) are given in min. In some cases, enantio-
meric excess values were determined by means of chiral supercritical
fluid chromatography (SFC) measurements on a Berger SFC with the
stated column. Gradient programs are described as follows: initial metha-
nol concentration [%]—initial time [min]—percentage gradient of metha-
nol [%minÀ1]—final methanol concentration [%]. Flash chromatography
was performed by using silica gel 32–63 mm, 60 ꢆ. THF, diethyl ether,
and dichloromethane were dried by filtration over alumina (activated at
3508C under a nitrogen atmosphere for 12 h). Copper(I) thiophenecar-
boxylate (CuTC) was purchased from Frontier Scientific.
(+)-(S)-(3-Methylpent-4-enyl)benzene (17): [a]2D2 =+1.4 (c=0.9 in
CHCl3) for 63% ee; 1H NMR (400 MHz, CDCl3): d=7.32 (m, 5H), 5.77
(ddd, 1J=7.6 Hz, 2J=10.1 Hz, 3J=17.4 Hz, 1H), 5.03 (d, J=17.2 Hz,
1H), 5.00 (d, J=7.3 Hz, 1H), 2.71–2.57 (m, 2H), 2.24–2.17 (m, 1H),
1.68–1.62 (m, 1H), 1.06 ppm (d, J=6.56 Hz, 3H); 13C NMR (100 MHz,
CDCl3): d=144.6, 142.9, 128.5, 128.4, 125.8, 113.1, 38.6, 37.6, 33.7,
20.4 ppm; MS (EI mode): m/z (%): 160 (14), 145 (19), 131 (9), 117 (11),
104 (100), 91 (96), 77 (13), 65 (14), 55 (7); the enantiomeric excess was
measured by chiral GC with a Hydrodex-B-3P column, hydrogen flow
(program: 70–0–1–170); tR =23.48 (+), 24.28 (À).
(À)-(S)-2-Methyl-4-phenylbutan-1-ol (18): A solution of olefin (+)-13
(0.86 mmol) in dry CH2Cl2 (30 mL) was cooled to À788C, and ozone was
passed through until a persisting blue color appeared (ꢀ10 min). After
completion of the reaction, the excess ozone was removed by purging
with O2 and N2. After cooling the system with an ice bath, sodium boro-
hydride (1.72 mmol) and methanol (5 mL) were added to the reaction.
The resulting mixture was permitted to warm to RT and was stirred over-
night. More NaBH4 (1.72 mmol) was added at 08C, while allowing the
temperature to rise to RT, to decompose the ozonide. After 7 d, the reac-
tion was hydrolyzed with H2O (15 mL), extracted with Et2O, and dried
over Na2SO4. The crude product was then subjected to column chroma-
tography on silica gel (pentane/Et2O 75:25). The alcohol was obtained as
a slightly yellow oil (0.33 mmol, 38% yield). [a]2D2 =À10.2 (c=0.9 in
CHCl3) for 63% ee; (ref. [30]: [a]2D2 =+20.0 (c=1.5 in CHCl3) for 99% ee
(R)); 1H NMR (400 MHz, CDCl3): d=7.29–7.26 (m, 2H), 7.20–7.16 (m,
3H), 3.51 (dd, 1J=38.1 Hz, 2J=5.8 Hz, 1H), 3.51 (dd, 2J=5.8 Hz, 3J=
17.2 Hz, 1H), 2.75–2.56 (m, 2H), 1.81–1.63 (m, 2H), 1.49–1.38 (m, 2H),
0.99 ppm (d, J=6.8 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=128.5,
125.9, 68.3, 35.5, 35.1, 33.4, 16.6 ppm.
Typical procedure for the enantioselective copper-catalyzed allylic substi-
tution with Grignard reagents: CuTC (1 mol%) and chiral ligand
(1.1 mol%) were charged in a dried Schlenk tube, under an inert gas,
and suspended in dichloromethane (2 mL). The mixture was stirred at
room temperature for 30 min, followed by the addition of the allylic
halide (1 mmol) at room temperature before cooling the mixture to
À788C in an ethanol/dry ice bath. The Grignard (3m in diethyl ether,
1.2 equiv) diluted in CH2Cl2 (0.6 mL) was added over 60 min by using a
syringe pump. Upon completion of the addition, the reaction mixture
was left a further 4 h at À788C. The reaction was quenched by addition
of aqueous HCl (1n, 2 mL) and then Et2O (10 mL). The aqueous phase
was separated and further extracted with Et2O (3ꢇ3 mL). The combined
organic fractions were washed with brine (5 mL), dried over anhydrous
magnesium sulfate, filtered, and concentrated under vacuum. The oily
(À)-[3-(Bromomethyl)-4-methylpent-4-enyl]benzene (27b): Rf =0.95
(silica gel, pentane); [a]2D2 =À10.3 (c=1.1 in CHCl3) for 62% ee (contain-
Chem. Eur. J. 2008, 14, 10615 – 10627
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10625

A. Alexakis et al.
ing 3% of 28b with 7% ee); 1H NMR (400 MHz, CDCl3): d=7.31–7.27
(m, 2H), 7.21–7.17 (m, 3H), 4.99 (d, J=1.3 Hz, 1H), 4.86 (s, 1H), 3.40
(d, J=6.6 Hz, 2H), 2.67–2.43 (m, 3H), 1.96–1.87 (m, 1H), 1.72 (m, 1H),
1.72 ppm (s, 3H); 13C NMR (100 MHz, CDCl3): d=144.3, 142.0, 128.5,
128.5, 126.0, 114.3, 48.9, 36.6, 33.5, 33.4, 18.9 ppm; MS (EI mode): m/z
(%): 210 (2), 174 (2), 117 (7), 105 (42), 104 (100), 92 (10), 91 (58), 77 (8),
69 (17), 67 (6), 65 (10); the enantiomeric excess was measured by chiral
GC with a Hydrodex B6-TBDM column, hydrogen flow (program: 100–
0–1–170); tR =46.89 (+), 47.4 (À) [for 28b: the enantiomeric excess was
measured by chiral GC with a Chirasil Dex CB, helium flow (program:
100–0–1–170); tR =51.98 (major), 52.3 (minor)].
meric excess was measured by chiral HPLC with an OD-H column (5 mL,
99.9:0.1 nHex/iPrOH, 0.5 mLminÀ1, 158C); tR =16.68 (+), 18.41 (À).
(+)-2-Phenethylbut-3-en-1-ol (36): Rf =0.36 (silica gel, pentane/Et2O
1
75:25); H NMR (400 MHz, CDCl3): d=7.30–7.26 (m, 2H), 7.20–7.17 (m,
3H), 5.65 (ddd, 1J=8.8 Hz, 2J=10.1 Hz, 3J=17.2 Hz, 1H), 5.24 (dd, 1J=
1.5 Hz, 2J=10.1 Hz, 1H), 5.19 (dd, 1J=1.0 Hz, 2J=17.2 Hz, 1H), 3.61–
3.43 (m, 2H), 2.74–2.53 (m, 2H), 2.32–2.23 (m, 1H), 1.80–1.54 (m, 2H),
1.50 ppm (brs, 1H); 13C NMR (100 MHz, CDCl3): d=142.3, 139.7, 128.5,
128.5, 125.9, 118.2, 65.7, 46.7, 33.4, 32.6 ppm; the enantiomeric excess was
measured by chiral GC with a Chirasil-Dex CB column, helium flow
(program: 80–0–1–170); tR =61.51, 62.42.
(À)-3-(Bromomethyl)-7-tert-butoxy-2-methylhept-1-ene (27c): Rf =0.46
(silica gel, pentane/Et2O 97.5:2.5); [a]2D2 =À8.0 (c=1.13 in CHCl3) for
87% ee (containing 2% of 28c with 37% ee); 1H NMR (400 MHz,
CDCl3): d=4.89 (s, 1H), 4.78 (s, 1H), 3.38 (d, J=1.3 Hz, 1H), 3.36 (d,
J=2.3 Hz, 1H), 3.32 (t, J=6.6 Hz, 2H), 2.45–2.38 (m, 1H), 1.66 (s, 3H),
1.58–1.22 (m, 6H), 1.17 ppm (s, 9H); 13C NMR (100 MHz, CDCl3): d=
144.8, 113.6, 72.6, 61.4, 49.4, 36.8, 31.6, 30.6, 27.7, 24.0, 18.9 ppm; MS (EI
mode): m/z (%): 263 (5), 261 (5), 141 (6), 123 (29), 96 (10), 81 (15), 69
(26), 67 (12), 59 (22), 57 (100), 55 (15); the enantiomeric excess was mea-
sured by chiral GC with a Hydrodex B6-TBDM column, hydrogen flow
(program: 100–0–1–170); tR =37.29 (+), 37.64 (À) [for 28c: the enantio-
meric excess was measured by chiral GC with a Hydrodex B3P, hydrogen
flow (program: 100–60–1–110–30–20–170); tR =90.00 (major), 90.86
(minor)].
Acknowledgements
The authors thank the Swiss National Research Foundation (grant no.
200020-113332) and COST action D40 (SER contract no. C07.0097) for
financial support, as well as BASF for the generous gift of chiral amines,
Solvias for the ferrocenyl ligands, Laetitia Palais for the SimplePhos li-
gands,[31] Stephane Rosset for analytical help, and Mariona Canto Valve-
rdu for experimental help.
(À)-[3-(Bromomethyl)-3,4-dimethylpent-4-enyl]benzene (30): [a]2D2 =À3.0
(c=1.03 in CHCl3) for 30% ee; 1H NMR (300 MHz, CDCl3): d=7.31–
7.26 (m, 2H), 7.21–7.16 (m, 3H), 5.04 (t, J=1.2 Hz, 1H), 4.86 (s, 1H),
3.49 (d, J=10.1 Hz, 1H), 3.41 (d, J=10.1 Hz, 1H), 2.56–2.38 (m, 2H),
1.80 (s, 3H), 1.81–1.63 (m, 2H), 1.26 ppm (s, 3H); 13C NMR (CDCl3,
75 MHz): d=146.8, 142.5, 128.6 (2ꢇ), 128.4 (2ꢇ), 126.0, 113.7, 43.8, 43.6,
31.2, 22.9, 19.5 ppm; MS (EI mode): m/z (%): 268 (1), 266 (1), 164 (39),
162 (40), 131 (13), 117 (11), 105 (79), 104 (97), 92 (22), 91 (88), 84 (17),
83 (100), 82 (13), 81 (21), 79 (14), 77 (13), 69 (14), 67 (15), 65 (12), 55
(23); the enantiomeric excess was measured by chiral GC with a Hydro-
dex B6-TBDM, hydrogen flow (program: 120–0–1–170); tR =33.98 (+),
34.32 (À).
(+)-1-[(2-Cyclohexylbut-3-enyloxy)methyl]benzene (32): Rf =0.79 (silica
gel, pentane); [a]2D8 =+5.9 (c=1.00 in CHCl3) for 36% ee; 1H NMR
(400 MHz, CDCl3): d=7.38–7.28 (m, 5H), 5.71 (ddd, J=9.1, 10.3,
17.0 Hz, 1H), 5.07 (d, J=10.3 Hz, 1H), 5.03 (d, J=17.7 Hz, 1H), 4.51 (s,
2H), 3.47 (m, 2H), 2.22–2.15 (m, 1H), 1.72–1.61 (m, 5H), 1.28–0.87 ppm
(m, 5H); 13C NMR (100 MHz, CDCl3): d=139.0, 128.4, 128.4, 127.7,
127.6, 127.6, 116.3, 73.1, 71.8, 50.0, 38.6, 31.3, 29.6, 26.8, 26.7 ppm; MS
(EI mode): m/z (%): 244 (4), 171 (1), 161 (2), 153 (5), 135 (11), 123 (7),
104 (9), 91 (100), 81 (44), 67 (18), 55 (15); the enantiomeric excess was
measured by chiral SFC with an OD-H column (program: 0–5–1–15;
200 bar, 2 mLminÀ1, 308C); tR =5.85 (À), 6.69 (+).
(À)-(R)-1-[(2-Ethylbut-3-enyloxy)methyl]benzene (33): Rf =0.5 (silica
gel, pentane/Et2O 98:2); (ref. [16a]) [a]2D8 =+19.0 (c=1.1 in CHCl3) for
94% ee on ent-33); 1H NMR (400 MHz, CDCl3): d=7.34–7.33 (m, 4H),
7.30–7.26 (m, 1H), 5.66 (ddd, 1J=8.3 Hz, 2J=9.8 Hz, 3J=17.9 Hz, 1H),
5.10–5.06 (m, 2H), 4.52 (s, 2H), 3.40 (d, J=6.3 Hz, 2H), 2.31–2.22 (m,
1H), 1.62–1.52 (m, 1H), 1.34–1.23 (m, 1H), 0.88 ppm (t, J=7.3 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=140.2, 138.8, 128.5, 127.7, 127.6, 115.6,
73.7, 73.1, 45.9, 24.2, 11.6 ppm; MS (EI mode): m/z (%): 190 (5), 189 (4),
104 (6), 92 (11), 91 (100), 65 (13); the enantiomeric excess was measured
by chiral HPLC with an OD-H column (5 mL, 99.8:0.2 nHex/iPrOH,
1 mLminÀ1, 408C); tR =5.66 (À), 6.14 (+).
(+)-(R)-1-[(2-Methylbut-3-enyloxy)methyl]benzene (34): Rf =0.85 (silica
gel, pentane); (ref. [16a]: [a]2D8 =À6.0 (c=1.1 in CHCl3) for 92% ee on
ent-34); 1H NMR (400 MHz, CDCl3): d=7.35–7.27 (m, 5H), 5.81 (ddd,
1J=17.2 Hz, 2J=10.4 Hz, 3J=6.8 Hz, 1H), 5.08 (dt, 1J=17.4 Hz, 4J=
1.8 Hz, 1H), 5.02 (d, 2J=10.3 Hz, 1H), 4.53 (s, 2H), 3.35 (dd, J=6.6,
39.6 Hz, 1H), 3.35 (dd, J=6.6, 21.5 Hz, 1H), 2.51 (m, J=6.6 Hz, 1H),
1.04 ppm (d, J=6.8 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=141.5,
138.6, 128.5, 127.7, 127.6, 114.2, 75.2, 73.1, 38.0, 16.8 ppm; MS (EI mode):
m/z (%): 176 (5), 175 (5), 92 (12), 91 (100), 65 (15), 55 (9); the enantio-
[1] a) B. M. Trost, C. Lee, in Catalytic Asymmetric Synthesis, 2nd ed.
(Ed.: I. Ojima), Wiley, New York, 2000, p. 593; b) A. Pfaltz, M.
Lautens, in Comprehensive Asymmetric Catalysis I–III (Eds.: E. N.
Jacobsen, A. Pfaltz, H. Yamamoto), Springer, Berlin, 1999, p. 833;
for reviews of asymmetric allylic alkylation with various metals, see:
c) Z. Lu, S. Ma, Angew. Chem. 2008, 120, 264–303; Angew. Chem.
Int. Ed. 2008, 47, 258–297; d) H. Miyabe, Y. Takemoto, Synlett 2005,
1641–1655; e) B. M. Trost, J. Org. Chem. 2004, 69, 5813–5837;
f) B. M. Trost, M. L. Crawley, Chem. Rev. 2003, 103, 2921–2943;
g) R. Takeuchi, Synlett 2002, 1954–1965.
[2] a) M. van Klaveren, E. S. M. Persson, A. del Villar, D. M. Grove, J.-
E. Bꢅckvall, G. van Koten, Tetrahedron Lett. 1995, 36, 3059–3062;
b) A. Alexakis, C. Malan, L. Lea, C. Benhaim, X. Fournioux, Synlett
2001, 927–930.
[3] a) F. Dꢈbner, P. Knochel, Angew. Chem. 1999, 111, 391–393; b) H.
Malda, A. W. van Zijl, L. A. Arnold, B. L. Feringa, Org. Lett. 2001,
3, 1169–1171; c) C. A. Luchaco-Cullis, H. Mizutani, K. E. Murphy,
A. H. Hoveyda, Angew. Chem. 2001, 113, 1504–1508; Angew.
Chem. Int. Ed. 2001, 40, 1456–1460; d) C. Borner, P. J. Goldsmith, S.
Woodward, J. Gimeno, S. Gladiali, D. Ramazzotti, Chem. Commun.
2000, 2433–2434.
[4] D. G. Gillingham, A. H. Hoveyda, Angew. Chem. 2007, 119, 3934–
3938; Angew. Chem. Int. Ed. 2007, 46, 3860–3864.
[5] For recent reviews of Cu-catalyzed AAA reactions, see: a) C. A.
Falciola, A. Alexakis, Eur. J. Org. Chem. 2008, 3765–3780; b) A.
Alexakis, C. Malan, L. Lea, K. Tissot-Croset, D. Polet, C. Falciola,
Chimia 2006, 60, 124–130; c) H. Yorimitsu, K. Oshima, Angew.
Chem. 2005, 117, 4509–4513; Angew. Chem. Int. Ed. 2005, 44, 4435–
4439; d) A. Kar, N. P. Argade, Synthesis 2005, 2995–3022; e) S. A. E.
Karlstrçm, J.-E. Bꢅckvall, in Modern Organocopper Chemistry (Ed.:
N. Krause), Wiley-VCH, Weinheim, 2002.
[6] C. A. Falciola, A. Alexakis, Angew. Chem. 2007, 119, 2673–2676;
Angew. Chem. Int. Ed. 2007, 46, 2619–2622.
[7] a) J. J. Van Veldhuizen, J. E. Campbell, R. E. Giudici, A. H. Hovey-
da, J. Am. Chem. Soc. 2005, 127, 6877–6882; b) M. A. Kacprzynski,
A. H. Hoveyda, J. Am. Chem. Soc. 2004, 126, 10676–10681.
[8] B. Breit, D. Breuninger, Synthesis 2005, 147–157.
[9] V. Calꢃ, A. Nacci, V. Fiandanese, Tetrahedron 1996, 52, 10799–
10810.
[10] a) S. Tominaga, Y. Oi, T. Kato, D. K. An, S. Okamoto, Tetrahedron
Lett. 2004, 45, 5585–5588; b) S. Okamoto, S. Tominaga, N. Saino, K.
Kase, K. Shimoda, J. Organomet. Chem. 2005, 690, 6001–6007.
[11] a) U. Piarulli, P. Daubos, C. Claverie, M. Roux, C. Gennari, Angew.
Chem. 2003, 115, 244–246; Angew. Chem. Int. Ed. 2003, 42, 234–
10626
www.chemeurj.org
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10615 – 10627

Difunctionalized Allylic Halides for Cu-Catalyzed SN2’ Reactions
FULL PAPER
236; b) U. Piarulli, C. Claverie, P. Daubos, C. Gennari, A. J. Min-
naard, B. L. Feringa, Org. Lett. 2003, 5, 4493–4496.
[12] a) K. Tissot-Croset, D. Polet, A. Alexakis, Angew. Chem. 2004, 116,
2480–2482; Angew. Chem. Int. Ed. 2004, 43, 2426–2428; b) K.
Tissot-Croset, A. Alexakis, Tetrahedron Lett. 2004, 45, 7375–7378;
c) Karine Tissot-Croset, Ph.D. Thesis no. 3634, University of Geneva
(Switzerland), 2005.
[13] a) P. Duhamel, Bull. Soc. Chim. Fr. 1996, 133, 457–459; b) L. Duha-
mel, P. Duhamel, J.-C. Plaquevent, Tetrahedron: Asymmetry 2004,
15, 3653–3691.
[21] C. A. Falciola, Ph.D. Thesis no. 3896, University of Geneva (Switzer-
land), 2007.
[22] The low yield was attributed to the small scale of the reaction
(1.5 mmol) and to the fact that the Grignard was not used immedi-
ately after preparation, which resulted in some b-elimination prod-
uct being afforded.
[23] S. E. Denmark, L. K. Marble, J. Org. Chem. 1990, 55, 1984–1986.
[24] A. Alexakis, K. Croset, Org. Lett. 2002, 4, 4147–4149.
[25] a) C. A. Falciola, K. Tissot-Croset, A. Alexakis, Angew. Chem. 2006,
118, 6141–6144; Angew. Chem. Int. Ed. 2006, 45, 5995–5998;
b) C. A. Falciola, K. Tissot-Croset, H. Reyneri, A. Alexakis, Adv.
Synth. Catal. 2008, 350, 1090–1100.
[14] a) M. J. Dunn, R. F. W. Jackson, J. Pietruszka, D. Turner, J. Org.
Chem. 1995, 60, 2210–2215; b) M. Yus, J. Gomis, Eur. J. Org. Chem.
2003, 2043–2048; c) R. Ortiz, M. Yus, Tetrahedron 2005, 61, 1699–
1707.
[26] N. Doi, S. Seko, K. Kimura, T. Takahashi, US Patent 20020107422,
2002.
[15] C. Jennings-White, R. G. Almquist, Tetrahedron Lett. 1982, 23,
2533–2534.
[16] a) F. Lopez, A. W. Van Zijl, A. J. Minnaard, B. L. Feringa, Chem.
Commun. 2006, 409–411; b) K. Geurts, S. P. Fletcher, B. L. Feringa,
J. Am. Chem. Soc. 2006, 128, 15572–15573.
[27] No byproducts resulting from further substitution were observed.
[28] A. W. van Zijl, F. Lopez, A. J. Minnaard, B. L. Feringa, J. Org.
Chem. 2007, 72, 2558–2563.
[29] a) J. Colonge, G. Poilane, Bull. Soc. Chim. Fr. 1955, 953–955;
b) D. A. Thomas, W. K. Warburton, J. Chem. Soc. 1965, 2988–2990.
[30] Z. Huang, Z. Tan, T. Novak, G. Zhu, E. Negishi, Adv. Synth. Catal.
2007, 349, 539–545.
[17] E. S. M. Persson, M. van Klaveren, D. M. Grove, J. E. Bꢅckvall, G.
van Koten, Chem. Eur. J. 1995, 1, 351–359.
[18] C. C. Tseng, S. D. Paisley, H. L. Goering, J. Org. Chem. 1986, 51,
2884–2891.
[19] J.-E. Bꢅckvall, E. S. M. Persson, A. Bombrun, J. Org. Chem. 1994,
59, 4126–4130.
[31] L. Palais, I. S. Mikhel, C. Bournaud, L. Micouin, C. A. Falciola, M.
Vuagnoux-dꢄAugustin, S. Rosset, G. Bernardinelli, A. Alexakis,
Angew. Chem. 2007, 119, 7606–7609; Angew. Chem. Int. Ed. 2007,
46, 7462–7465.
[20] M. A. Kacprzynski, T. L. May, S. A. Kazane, A. H. Hoveyda,
Angew. Chem. 2007, 119, 4638–4642; Angew. Chem. Int. Ed. 2007,
46, 4554–4558.
Received: June 30, 2008
Published online: October 16, 2008
Chem. Eur. J. 2008, 14, 10615 – 10627
ꢂ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10627