Palladium-catalyzed diastereoselective and enantioselective allylic alkylations of ketone enolates

Article abstract of DOI:10.1002/adsc.200700409

Lithium and magnesium enolates of cyclohexanone undergo palladium-catalyzed allylic alkylations under mild conditions. Diastereoselectivity and enantioselectivity are observed when the diphenyl- and dimethyl-substituted allylic substrates 1a and 1b are re

Full text of DOI:10.1002/adsc.200700409

FULL PAPERS
DOI: 10.1002/adsc.200700409
Palladium-Catalyzed Diastereoselective and Enantioselective
Allylic Alkylations of Ketone Enolates
Manfred Braun,a,* Thorsten Meier,a Frank Laicher,a Panos Meletis,a
and Mesut Fidana
a
Institut für Organische Chemie und Makromolekulare Chemie, Universität Düsseldorf, 40225 Düsseldorf, Germany
Fax : (+49)-211-811-5079; e-mail: braunm@uni-duesseldorf.de
Received: August 22, 2007; Published online: January 11, 2008
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Lithium and magnesium enolates of cyclo- and stereoselectivity. The stereochemical outcome of
hexanone undergo palladium-catalyzed allylic alkyla- the allylic alkylation of cyclohexanone and acetophe-
tions under mild conditions. Diastereoselectivity and none has been investigated by the palladium-cata-
enantioselectivity are observed when the diphenyl- lyzed reaction of their lithium enolates with the cis/
and dimethyl-substituted allylic substrates 1a and 1b trans isomeric alkenes (Z)-18 and (E)-19. It turns out
are reacted with cyclohexanone or ethyl mesityl that the preformed, non-stabilized enolates attack p-
ketone. The lithium enolates of cyclohexanone, cy- allyl-palladium complexes generated in situ from the
clopentanone and a-tetralone lead to the alkylations face opposite to the noble metal thus following the
products 12–14 in an enantioselective manner. Axial- stereochemical pathway of soft, stabilized carban-
ly chiral biphenyl- and binaphthyl-bisphosphanes ions.
provide high enantioselectivity and/or diastereoselec-
tivity. In the case of the lithium enolates, the pres- Keywords: allyl complexes; asymmetric synthesis;
ence of lithium chloride is also crucial to reactivity lithium; nucleophilic additions; stereoselectivity
Introduction
leads to the formation of just one stereogenic center
in the allylic position.[2]
The palladium-catalyzed allylic alkylation, the Tsuji–
If, on the other hand, synthesis is aimed at alkenyl
Trost reaction, has developed into an exceptionally ketones 4 with stereogenic centers in the homoallylic
useful and versatile method for carbon-carbon and (R2 ¼ H, R3 =H) or in both the allylic and the homo-
carbon-heteroatom bond formation.[1] This holds in allylic position (R2 ¼ H, R3 ¼ H) the suitable prochi-
particular for the numerous enantioselective variants ral nucleophiles are preformed, non-stabilized ketone
developed in recent years. In general, the asymmetric enolates 5,[3] as shown in the retrosynthetic Scheme 2.
allylic alkylation follows the route outlined in
In spite of the obvious synthetic utility of this ap-
Scheme 1. Racemic starting material 1 is converted proach, only very few attempts have been made to
into non-racemic product 3 or ent-3 due to chiral li- combine preformed enolates with in situ generated
gands L* at the noble metal that direct the approach allyl-palladium complexes.[4] In recent years, a few
of the nucleophile predominantly to one of the diaste- enantioselective protocols have been elaborated for
reotopic termini of the p-allyl complex 2. As far as the palladium- and iridium-catalyzed allylic alkylation
carbon nucleophiles are concerned, there has been a of ketones, most of them using the corresponding
substantial limitation: For about two decades, almost enol stannanes[5], enol silanes[6a,b] and enamines.[6c] Al-
exclusively “soft”, stabilized carbanions have been though it has been shown that lithium enolates can
used as carbon nucleophiles. Among them, symmetri- also serve as nucleophiles,[5a,c–e] there remained still
cally substituted nucleophiles were applied preferen- substantial limitations inasmuch as most enolate pre-
tially, and the allylation of malonates [Nu=CH- cursors chosen for this purpose contain just one acidic
(CO2R’)2] became a kind of standard protocol of the proton, so that double or multifold alkylation as well
Tsuji–Trost reaction (Scheme 1), which was applied as later racemization can be excluded per se. Thus,
very frequently when the performance of novel li- the enantioselective allylic alkylation of “unrestrict-
gands L* was tested. It is evident that this protocol ed” ketones still remained as an unsolved problem.[7]
Adv. Synth. Catal. 2008, 350, 303 – 314
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
303

Manfred Braun et al.
FULL PAPERS
with a palladium catalyst under concomitant extrusion
of carbon dioxide.[8] Here, we describe the direct ste-
reoselective palladium-catalyzed reaction of allylic
substrates 1a–d with magnesium and lithium enolates
of unrestricted ketones.[9]
Results and Discussion
Cyclohexanone 7 was chosen as the suitable represen-
tative of a ketone that offered the advantage of a
fixed configuration (E) of the enolate 8. Its reactions
with racemic disubstituted allylic substrates 1a and 1b
were investigated first with respect to diastereoselec-
tivity that should be obtained when two adjacent ste-
reogenic centers are formed in alkenyl ketones 9 and
10. Then, the problem of enantioselectivity was tack-
led (Scheme 3).
Scheme 1. “Standard” version of an enantioselective Tsuji–
Trost allylic alkylation, Nu=CH(CO2R’)2.
G
Scheme 3. Diastereoselective and enantioselective palladi-
um-catalyzed allylic alkylation of cyclohexanone.
Thus, the influence of the base, used for the depro-
tonation of the ketone, the ligand at the palladium
and the effect of additives were investigated; the re-
sults are shown in Table 1. In all cases, tris(dibenyzl-
ideneacetone)dipalladium-chloroform
([Pd2(dba)3]·
A
CHCl3)[10] was used as the precursor of the catalyst
that formed upon treatment with the corresponding
phosphane or bisphosphane ligands 11a–k, shown in
Scheme 4.[11] They exhibit a major influence on dia-
stereoselectvity. In all cases, the syn diastereomer
formed predominantly. Thus, all bisphosphane ligands
11b–d were superior to triphenylphosphane (Table 1,
entries 1–5), the ferrocene derivative 11d being the
most efficient (entries 4 and 5). The use of lithium or
potassium hexamethyldisilazide (entries 6 and 7) did
not improve the diastereoselectivity. The additive lith-
ium chloride led to a significant enhancement of the
diastereoselectivity (entry 8). Further improvement
Scheme 2. Retrosynthesis of alkenyl ketones 4 based on ke-
tones 6 and allylic substrates 1a–d via enolates 5 and allyl-
palladium complexes. R1 =alkyl, aryl, R2 =alkyl.
However, a detour has been elaborated recently resulted when, instead of the lithium, the magnesium
based on ketone-derived allyl enol carbonates or allyl enolate was used (entry 9) in the combination with
b-keto esters that liberate an enolate upon treatment the dppf ligand 11d. As aside from diastereoselectiv-
304
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 303 – 314

Palladium-Catalyzed Diastereoselective and Enantioselective Allylic Alkylations
Table 1. Palladium-catalyzed diastereoselective and enantioselective allylic alkylation of cyclohexanone.
FULL PAPERS
Entry
Substrate
Base/Additive
Ligand
Temperature [8C]
Product
Conversion [%][a]
dr
ee
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1b
LiN
LiN
LiN
LiN
LiN
LiN
U
11a
11b
11c
11d
11d
11d
11d
11d
0
0
0
20
0
0
0
0
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9b, 10b
9b, 10b
45
53:47
81:19
83:17
85:15
87:13
90:10
60:40
92:8
97:3
99:1
99:1
90:10
97:3
-
-
-
-
-
-
-
-
>97
>97
>97
>97
>97
>97
>97
>97
>97
>97 (67)[b]
95
KN
A
9
ClMgN
ClMgN
ClMgN
LiN
LiN
G
11d
0
0
0
-
-
10
11
12
13
G
rac-11e
(R)-11e
rac-11e
(R)-11e
U
99[c]
-
ꢀ15
ꢀ78 to ꢀ8
35[b]
96[d]
[a]
1
Determined from the H NMR spectra of the crude reaction mixture.
Isolated yield.
Determined by HPLC on a chiracel OJ column.
Determined by GC on a Lipodex E column.
[b]
[c]
[d]
Scheme 4. Ligands used for stereoselective allylic alkylations of ketone enolates.
Adv. Synth. Catal. 2008, 350, 303 – 314
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
305

Manfred Braun et al.
FULL PAPERS
ity, enantioselectivity, was targeted in addition, the 11e, a remarkably lower yield was reached with the
search for a chiral phosphane ligand was obvious. optically pure ligand (R)-11e. However, both remark-
Thus racemic BINAP (11e) was tested and led to a able diastereoselectivity was observed and the enan-
diastereomeric ratio of 99:1, when the magnesium tiomeric excess of the isolated product 9b amounted
enolate of cyclohexanone was used (entry 10).
to 96% ee (entry 13) The low yield combined with
Consequently (R)-BINAP (11e) was used as an the fact that recovered starting allylic carbonate 1b
enantiomerically pure ligand to bring about enantio- was enantiomerically enriched suggest the idea of a
selectivity in the allylic alkylation. Indeed, the syn- kinetic resolution in course of the allylic alkylation.[12]
configured diastereomer 9a was obtained not only in
The palladium-catalyzed reaction of the enolate 8
a high diastereoselectivity but also in a remarkable with allyl acetate 1c and allyl methyl carbonate 1d
enantiomeric excess of 99% ee, as determined by was investigated next. As the reaction leads to the
HPLC on a chiral column.
formation of just one stereogenic center, this type of
Most of palladium-catalyzed allylic alkylations with enantioselective allylation of simple ketones appears
various ligands have been investigated using the 1,3- to be rather facile. There is, however, no restriction
diphenyl-disubstitution pattern of the allylic substrate that prevents double allylation and particularly a later
1a. On the contrary, the 1,3-dialkyl-substituted ana- racemization, so that just the simplicity of the cyclo-
logues have been employed by far less frequently. hexanone substrate makes it very challenging.
Thus, racemic carbonate 1b was submitted to the pal-
As the axially chiral bisphosphane ligands had
ladium-catalyzed reaction with the lithium enolate of proved themselves to be most efficient in the diaste-
cyclohexanone. With this substrate, the magnesium reoselective and enantioselective allylic alkylations
enolate reacted in a very sluggish manner, and, in with the substrates 1a and 1b, this type of ligand was
case of the lithium enolate, the additive lithium chlo- also applied in the enantioselective allylation of cyclo-
ride was also essential for reactivity and selectivity. hexanone, shown in Scheme 5. In all cases the lithium
Thus, diastereoselective formation of syn-product 9b enolate 8 was used as the nucleophile. Among the li-
was observed when the reaction was mediated with gands tested, BINAP (11e) provided substantial enan-
racemic BINAP (11e), the diastereomeric ratio tioselectivity. However, the selectivity was strongly in-
amounting to 90:10 (entry 12). Whereas a high degree fluenced by the metal salts used as additives (Table 2,
of conversion was obtained with the racemic ligand entries 1 and 6 vs. entries 2–5 and 7–13). Best results
Scheme 5. Palladium-catalyzed enantioselective allylation of cyclohexanone.
Table 2. Palladium-catalyzed enantioselective allylation of cyclohexanone.
Entry
Substrate
Ligand
Additive
Temperature [8C]
Product
Yield[a]
ee
1
2
3
4
5
6
7
8
1c
1c
1c
1d
1d
1d
1d
1d
1d
1d
1d
1d
1d
(S)-11e
(S)-11e
(S)-11e
(S)-11e
(S)-11e[b]
(R)-11e
(S)-11f
(R)-11g
(R)11h
(S)-11i
(S)-11j
(S)-11k
(R)-11k
-
ꢀ30
ꢀ20
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
(S)-12
(S)-12
(S)-12
(R)-12
(S)-12
-
(R)-12
(R)-12
(R)-12
(S)-12
(S)-12
(S)-12
(R)-12
40
15
53
73
87
32
70
75
97
80
-
94
91
92
93
95
98
>98
ZnBr2
LiCl
LiCl
LiCl
-
LiCl
LiCl
LiCl
LiCl
LiCl
LiCl
LiCl
[c]
[d]
-
68
65
40
71
74
76
72
9
10
11
12
13
[a]
Isolated distilled products.
Catalyst loading: 0.025 mol% [Pd2(dba)3]·CHCl3, 0.1 mol% ligand.
[b]
[c]
[d]
A
Traces.
Not determined.
306
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 303 – 314

Palladium-Catalyzed Diastereoselective and Enantioselective Allylic Alkylations
FULL PAPERS
were obtained again with lithium chloride (entries 3–5
The determination of the absolute configuration of
and 7–13). In addition, the use of allyl carbonate 1d the allylation products 12–14 was possible based on
instead of allyl acetate 1c brought about an improve- the comparison of their optical rotations with the data
ment of enantioselectivity (entries 3 and 4). Thus, al- described in the literature.[15] In addition, the CD
lylation occurred in 97% ee, when mediated by (R)- spectrum of (R)-12, displayed a significant positive
BINAP (11e). Remarkably, the reaction smoothly oc- Cotton effect at 285 nm, again in agreement with
curred at ꢀ788C. Under these conditions, the additive known data.[15b] For the allylation products 9 and 10,
lithium chloride was also crucial to reactivity: in the the relative configuration had to be assigned first. In
absence of the salt, no reaction took place (entry 6) at case of the diphenyl derivative 9a that formed as the
that temperature, a result that clearly underlines the major diastereomer, a crystal structure analysis re-
enhanced reactivity of the lithium enolate due to the vealed the syn-configuration.[9a] The absolute configu-
known[13] desaggregation by lithium chloride. As a ration of the product 9a, obtained when the allylation
consequence, racemization of the allylation product was mediated by (R)-BINAP (11e), was determined
12 was widely suppressed. The standard catalyst load- to be (2R,1’S) by the following chemical correlation
ing was 2.5 mol% of [(Pd2(dba)3]·CHCl3 and 10 (Scheme 7). The ozonolysis of 9a led to the keto alde-
G
mol% of the chiral ligand. However, the amount of hyde 15 that was converted without purification into
palladium could be as low as 0.05 mol% and 0.1 the thioacetal 16 by treatment with 1,3-propanedi-
mol% of the ligand, as shown by the allylation that thiol, obtained in 13% yield after a relatively tedious
was mediated by (S)-BINAP (11e) and gave (S)-12 in chromatography. Finally, the desulfurization with
80% ee and 87% isolated yield (entry 5). Under the Raney nickel gave the crude hydrocarbon 17, which
standard conditions that included the presence of lith- showed an optical rotation [a]2D0of +6.8. As the rota-
ium chloride and a reaction temperature of ꢀ788C, tion of (R)-17 has been reported to be ꢀ8.6,[16] the
various axially chiral bisphosphane ligands that are (S)-configuration has to be assigned to the degrada-
commercially available were tested. The results are tion product 17 and – as a consequence of a change in
shown in entries 7–13 of Table 2. As a result of the priorities – the stereogenic a-carbonyl carbon of the
optimization study, the BIPHEP-type ligand (R)- and allylation product 9a is (R)-configured.
(S)-11k provided the highest degrees of enantioselec-
In the case of the allylation products 9b/10b, pre-
tivity and permitted us to obtain enantiomeric (R)- pared previously by different routes,[8a,17] the syn-con-
and (S)-allylcyclohexanone in 98% and >98% ee, re- figuration was assigned to the major diastereomer 9b.
spectively, the chemical yields ranging from 72 to This stereoisomer has been obtained from an addition
76%.[14] It should be mentioned that the source of the of 1-pyrrolidino-1-cyclohexene to CpMo(NO)-
precatalyst [(Pd
A
ence on the enantioselectivity.
plexation with moderate diastereoselectivity.[17b] Nev-
The enantioselective allylation was also studied ertheless, the diastereomers differ in some of the
with cyclopentanone and a-tetralone to give the ole- 1H NMR signals, so that a comparison permits us to
finic ketones 13 and 14 in 64 and 66% ee, respective- assign the syn-configuration to the main product 9b,
ly, when the reaction was mediated with (S)- and (R)- obtained form the palladium-catalyzed reaction. An-
BINAP (11e), as shown in Scheme 6. Particularly in other support for this assignment comes from the
1
the case of allyltetralone 14 a partial racemization comparison of the H NMR spectra of the diastereo-
might be responsible for the relatively low enantio- meric ketones 9b and 10b with those of syn- and anti-
meric excess.
2-(1-methyl-2-propenylcyclohexanone), the chemic
shift of the vinylic 2’-H being characteristic.[8a] The ab-
Scheme 6. Enatioselective allylation of cyclopentanone and a-tetralone.
Adv. Synth. Catal. 2008, 350, 303 – 314
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
307

Manfred Braun et al.
FULL PAPERS
Scheme 7. Assignment of configuration to alkenyl ketone 9a by chemical correlation.
solute configuration of 9a was finally determined to pKa >25.[2c] Cyclohexanone, whose allylation was in-
be (2R,1’S) by CD spectroscopy. Similarly to (R)-allyl- vestigated here is a borderline case in this context in-
cyclohexanone 12, it shows a strong positive Cotton asmuch as its pKa value amounts to 26.4 according to
effect at 290 nm. As this effect is determined princi- Bordwellꢁs scale.[18] Thus, we have studied the stereo-
pally by the stereogenic center at the a-carbonyl chemical outcome of its palladium-mediated allylation
carbon atom, it unambiguously shows that the ke- because, at a glance, it is not predictable whether is
tones (R)-12 and (2R,1’S)-9b are homochiral with re- will react like a “hard” or a “soft” nucleophile. It
spect to that center.
should be pointed out, however, that in early studies
For the reaction of a nucleophile with p-allylpalla- the lithium and the potassium enolates of acetone
dium complexes, two different paths have been dis- have been shown to attack p-allyl-palladium com-
cussed: The nucleophile either approaches the p-allyl plexes from the face opposite to the noble metal.[4a,b]
moiety from the face opposite to the palladium, as As the desaggregation with lithium chloride seems to
outlined in Scheme 1, or it is first coordinated to the play an important role in our approach, the stereo-
transition metal and attacks the p-system from the chemical outcome was investigated with the lithium
same face. It is generally accepted and supported by enolate of cyclohexanone in the presence and in the
manifold evidence that “soft” carbon nucleophiles absence of lithium chloride.
follow the first pathway, whereas “hard” nucleophiles
The enantiomerically and diastereomerically pure
have been shown to react through a precoordination, alkenes (Z)-18 and (E)-19 served as probes to study
thus following the second route.[1,2] Although some- the stereochemical outcome.[19] They are acyclic and
what arbitrarily, a border between the two types of unrestricted, identical in the configuration of their ste-
carbon nucleophiles has been set in such a way that reogenic carbon centers, but they differ in the config-
conjugate bases of carbon acids with pKa <25 have uration of the double bond. Their stereogenic center
been classified as “soft” nucleophiles, whereas the adjacent to the MEM ether group should not be
“hard” species are derived from carbon acids with touched during the allylic substitution (Scheme 8).
Scheme 8. Formation of diastereomeric products 20–23 from palladium-catalyzed reactions of cyclohexanone enolate with
(Z)-18 and (E)-19, respectively.
308
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 303 – 314

Palladium-Catalyzed Diastereoselective and Enantioselective Allylic Alkylations
FULL PAPERS
When the reaction of the lithium enolate 8 (M=Li)
of cyclohexanone with (Z)-18 was mediated by
A
N
presence of lithium chloride, a mixture of syn-20 and
anti-21 resulted in a ratio of 79:21. The analogous re-
action with (E)-19, preformed in the absence of lithi-
um chloride, led to the diastereomeric products syn-
22 and anti-23, in a 57:43-mixture. Both mixtures
were separated by column chromatography. The as-
signment of the relative configurations in the syn/anti-
diastereomers 20/21 and 22/23 was possible based on
1
the H NMR spectra. The characteristic differences in
shift and coupling constants were exhibited by the a-
carbonyl, the benzylic and the vinylic protons in each
of the diastereomers. To give an example, the signal
(ddd) of the vinylic proton (indicated in the formulas
20–23) shows up at lower field in the syn-isomers 20
and 22, and at higher field in the anti-diastereomers
21 and 23. This also holds for the diastereomers anti-
10a and syn-9a, the configuration of which was deter-
mined unambiguously by a crystal structure analysis.
The configuration of the stereogenic center at the a-
carbonyl carbon atom was determined by CD spec-
troscopy. Thus, the diastereomers syn-20 and anti-23
showed a negative Cotton effect at 295 nm, indicating
the (S)-configuration in the cyclohexyl ring. Accord-
ingly, anti-21 and syn-22 displayed a positive Cotton
effect at the same wavelength, so that in analogy to
(R)-allylcyclohexanone 12 the (R)-configuration could
be assigned to the a-carbonyl center. This result clear-
ly shows that in the reaction of (Z)-18, the leaving
group has been replaced under inversion of the con-
figuration, whereas, starting from (E)-19, the allylic
substitution has occurred under retention in the allylic
moiety. In both reactions the enantiofacial selectivity
of the enolate was lower than that observed with the
substrates 1a and 1b.
The stereochemical outcome is rationalized as fol-
lows: The substrate (Z)-18 reacted with the nucleo-
phile in an overall inversion due to a p–s–p-intercon-
version,[20] observed already previously with that sub-
strate,[19] as shown in Scheme 9. This is clearly indicat-
ed by the conversion of the (Z)-double bond in the
starting material into (E)-configured products 20 and
21. Thus, the p-complex 24 formed initially from the
substrate (Z)-18 was converted via the s-intermediate
25 and its rotamer 26 into the p-complex 27. This is fi-
nally attacked by the enolate from the face opposite
to the palladium. Thus, the overall inversion results
from two substitutions occurring each under inversion
and an interconversion of the (Z)- into an (E)-double
bond through the p–s–p-mechanism.
Scheme 9. Rationale for the stereochemical outcome in the
reaction of allylic substrates (Z)-18 and (E)-19 with the lithi-
um enolate of cyclohexanone.
The formation of diastereomeric products 22 and a thermodynamically controlled (Z)- to (E)-intercon-
23 from the substrate (E)-19 is plausibly rationalized version. As a result, it has been shown that the lithi-
by a two-fold inversion: First, in the formation of the um enolate of cyclohexanone, despite the presence or
p-complex 28 and, second, by an approach of the nu- absence of lithium chloride, attacks p-allyl-palladium
cleophilic enolate. In this case, there is no “need” for complexes from the face opposite to the noble metal,
Adv. Synth. Catal. 2008, 350, 303 – 314
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
309

Manfred Braun et al.
FULL PAPERS
thus following the stereochemical pathway of “soft” of “soft” nucleophiles in the palladium-catalyzed al-
nucleophiles.
lylic alkylation.
This stereochemical outcome was also proven for
Based thereupon, the diastereoselectivity observed
the lithium enolate of acetophenone, using again the in the reaction of cyclohexanone with the diphenyl-
isomeric alkenes (Z)-18 and (E)-19 as probes. In both and dimethyl-substituted allylic systems 1a and 1b, is
experiments, lithium chloride was used as an additive, rationalized by assuming the “closed” transition state
and dppf (11d) served as the achiral ligand. As shown model 33. It seems to be plausible that in this ap-
in Scheme 10, alkenyl ketones 30 and 32 were ob- proach of the enolate to the p-allyl-palladium com-
tained upon treatment of the lithium enolate 29 with plex, a Coulomb attraction between the negatively
(Z)-18 or (E)-19, respectively. The products formed in charged oxygen of the enolate and the complexed
a regioselective and diastereoselective manner, and allyl cation moiety might be effective (Scheme 11).
the ratio of these isomers surpassed 95:5 in both
cases. They are also pure enantiomers, as their synthe-
sis started from the enantiomerically pure alkenes
(Z)-18 and (E)-19.
Here again, the products 30 and 32 differ with re-
spect to the configuration of the benzylic stereogenic
center: the former diastereomer results from an over-
all inversion, the latter from an overall retention.
Analogously to the case of cyclohexanone, the forma-
tion of the diastereomer 30 is explained by an inver-
sion during the formation of the p-allylpalladium
complex, a p–s–p-conversion, which leads to (Z)- to
(E)-isomerization, and a final inversion occurring
during the attack of the nucleophile to the p-allyl-pal-
Scheme 11. Transition-state model for the diastereoselective
allylic alkylation of cyclohexanone enolate 8.
ladium complex (cf. Scheme 9). The overall retention,
on the other hand, that is observed in the formation
In order to support this model, the enolate 35 of
of the diastereomer 32 from (E)-19 is rationalized as ethyl mesityl ketone 34 was submitted to a reaction
the result of a double inversion. The configurations at with allyl acetate 1a following the above protocol and
the stereogenic centers formed upon the allylic alkyla- using the achiral ligand dppf (11d). The treatment of
tion were proved by ozonolysis of the alkenes 30 and the ketone 34 with lithium diisopropylamide is
32 which gave the enantiomeric aldehydes (R)- and known[3] to give the (E)-configured enolate 35 in a
(S)-31, respectively. As their optical rotation is highly selective manner. This was proven by a deter-
known,[21] a comparison permitted us to assign the ab- mination of the E:Z ratio that amounted to 96:4 and
solute configuration to them and, as a consequence, changed only slightly (to 92:8) under the influence of
to determine the relative configuration to the allyla- [Pd2(dba)3]·CHCl3, shown by quenching with chloro-
G
tion products 30 and 32. Thus, it has been demonstrat- trimethylsilane. As both the enolate 35 and the depro-
ed that the lithium enolates of both acetophenone tonated cyclohexanone 8 have E-configuration, the
and cyclohexanone follow the stereochemical pathway predominant formation of syn-product 36 was expect-
Scheme 10. Formation od diastereomeric products 30 and 32 from the palladium-catalyzed allylation of acetophenone with
(Z)-18 and (E)-19 through the lithium enolate 29.
310
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 303 – 314

Palladium-Catalyzed Diastereoselective and Enantioselective Allylic Alkylations
FULL PAPERS
ed at the expense of anti-37. Surprisingly however, Conclusions
the latter diastereomer was found to be the major
product, and the syn-36:anti-37 ratio was determined It has been shown, in summary, that palladium-cata-
to be 10:90 (Scheme 12). The structure of the main lyzed allylic alkylations are feasible with preformed,
diastereomer 37 was assigned based on a crystal struc- non-stabilized lithium and magnesium enolates of
ture analysis.[7a]
simple ketones. Not only diastereoselective but also
highly enantioselective protocols have been elaborat-
ed. In the case of lithium enolates, the presence of
lithium chloride permits us to perform the allylation
at low temperatures due to an enhancement of reac-
tivity by desaggregation of the enolate structures.
Enantioselectivity is provided by using axially chiral
bisphosphane ligands of the BINAP and BIPHEP
types. Thus, it becomes more and more evident that
tools are now available that permit to combine allyl-
palladium and preformed enolate chemistry.
Experimental Section
Characterization data are available in a supporting informa-
tion file. [Pd2(dba)3]·CHCl3 was purchased from Strem
G
Chemicals, Inc. Allyl acetate 1c and methyl allyl carbonate
1d are commercially available. (E)-1,3-Diphenylprop-2-enyl
acetate (1a) was prepared according to ref.[22]
(E)-Methyl (Pent-3-en-2-yl) Carbonate (1b)[23]
A 500-mL three-necked flask was equipped with a dropping
funnel, closed with a septum, a magnetic stirrer, a septum,
an inlet thermometer and a connection to the combined ni-
trogen/vacuum line. The air in the flask was replaced by ni-
trogen and a 3M solution of CH3MgCl in THF (100 mL,
300 mmol) was added through a cannula. The solution was
diluted with THF (125 mL) and cooled to 08C. Under stir-
ring, (E)-crotonaldehyde (18.0 g, 255 mmol) was added
through the dropping funnel at such a rate that the tempera-
ture did not exceed 38C. The dropping funnel was then
rinsed with THF (5 mL) that was added to the reaction mix-
ture. After stirring at room temperature for 2 h, the solution
was cooled to ꢀ788C. Methyl chloroformate (24 mL, 29 g,
312 mmol) was added through the dropping funnel while
keeping the temperature below ꢀ708C. After rinsing of the
dropping funnel with THF (5 mL), the mixture was allowed
to warm up to room temperature within 18 h under stirring.
It was poured into a buffer solution (150 mL, pH 7). The or-
ganic layer was separated and the aqueous phase was ex-
tracted three times with CH2Cl2. The combined organic
layers were dried with MgSO4 and concentrated in a rotary
evaporator. The residue was purified by distillation under
Scheme 12. Diastereoselective allylic alkylation of ethyl me-
sityl ketone 34 and transition-state models.
It seems to be possible that a torsion of the mesityl
group with respect to the enolate plane leads to an
unfavorable steric interaction between the allyl
moiety and one of the ortho methyl groups in the reduced pressure to give colorless 1b; yield: 29.7 g (81%).
The spectroscopic data are in accodance with those descri-
cyclic transition-state model 38. The alternative open
transition-state model 39 could avoid this type of in-
teraction so that it becomes at least less unfavorable
than the reaction pathway through 38. When the ally-
lation of the enolate 35 with diphenylallyl acetate 1a
was mediated with (R)-BINAP (11e), the syn-diaste-
reomer 36 formed in 88% ee, as shown by NMR mea-
surement in the presence of the chiral shift reagent
berd in the literature.[24]
A
phenyl-2-penten-1-yl} Acetate [(Z)-18]
To a solution of (1R,2Z,4S)-4[(2-methoxyethoxy)methoxy]-
1-phenyl-2-penten-1-ol[19] (5.4 g, 20.3 mmol) in CH2Cl2
(10 mL), stirred at 08C under nitrogen, was added an ice-
cold mixture of triethylamine (2.8 g, 28 mmol), acetic anhy-
Eu(hfc)3.
A
Adv. Synth. Catal. 2008, 350, 303 – 314
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
311

Manfred Braun et al.
FULL PAPERS
dride (2.8 g, 28 mmol), 4-(dimethylamino)pyridine (10 mg,
0.08 mmol) and CH2Cl2 (30 mL). The solution was stirred at
room temperature for 23 h and poured into water. The or-
ganic layer was separated and the aqueous phase was ex-
tracted three times with CH2Cl2. The combined organic
layers were washed with 2N NaOH solution and with water
and dried with MgSO4. The solvent was removed in a rotary
evaporator and then in oil-pump vacuum and the residue
was purified by distillation under reduced pressure to give
colorless, oily (Z)-18; yield: 5.647 g (90.3%)
dried with MgSO4 and evaporated at 408C and a pressure
that did not fall below 10 mbar. The flask was connected by
glass tubes with two subsequent traps that were cooled to
08C and ꢀ1968C, respectively. When the flask containing
the crude product was heated to 508C at 0.07 mbar, pure
product was collected in the 08C trap. Yield of (S)-12:
0.497 g (72%); 98% ee according to chiral GC.
According to this procedure, the following compounds
were prepared:
(2R,1’S,2’E)-2-(1-Methyl-but-2-enyl)cyclohexanone (9b):
Obtained from cyclohexanone (7) and methyl pentenyl car-
bonate (1b); the reaction was mediated by the ligand (R)-
11e; yield: 35%, diastereomeric ratio 9b:10b=97:3; 96% ee.
A
penten-1-yl} Acetate [(E)-19]: Also obtained according to
this procedure from (1R,2E,4S)-4-[(methoxyethoxy)me-
thoxy]-1-phenyl-2-penten-1-ol[19]; yield: 56.7%.
1
The H and 13C NMR data measured in CDCl3 are in ac-
cordance with those described in the literature[8a] for race-
mic 9b and 10b, respectively.
(2R,1’S,2’E)-2-(1,3-Diphenylallyl)cyclohexanone (9a)
(S)-2-Allylcyclopentanone (13): Obtained from cyclopen-
tanone and allyl methyl carbonate (1d); the reaction was
Through the magnesium enolate ofcyclohexanone: Under
nitrogen, diisopropylamine (0.15 mL, 1.1 mmol) dissolved in
1 mL of dry THF was allowed to react with 1.1 mmol of
methylmagnesium chloride (0.37 mL of 3M solution in
THF) at ꢀ168C. After 30 min, cyclohexanone (0.11 mL,
1.1 mmol) dissolved in 1 mL of dry THF was added drop-
wise, and stirring was continued for 30 min at ꢀ168C. The
enolate thus generated was injected into a solution of 1a
1
mediated by (S)-11e; yield: 76%. The H NMR data are in
accordance with those described in the literature.[25]
(R)-2-Allyl-1-tetralone (14): Obtained from a-tetralone
and allyl methyl carbonate (1d); the reaction was mediated
by (R)-11e. In the work-up procedure, 10% citric acid was
used instead of the buffer solution. Yield: 85%. The
1H NMR data are in accordance with those described in the
literature.[15e]
(252 mg, 1 mmol), [Pd2(dba)3]·CHCl3 (25.9 mg, 25 mmol) and
U
(R)-BINAP 11e (62 mg, 100 mmol) in 2 mL of dry THF.
After stirring for 16 h at 08C, the mixture was hydrolyzed
by addition of phosphate-buffer (pH 7) and extracted three
times with CH2Cl2 (50 mL each). The combined organic
layers were washed with brine, dried with MgSO4 and con-
centrated in a rotary evaporator to give the crude product;
yield: 285 mg (98%).
(2S,1’R,2’E,4’S)- and (2R,1’R,2’E,4’S)-2-{4-[(2-Methoxy-
ethoxy)methoxy]-2-phenylpent-2-enyl}cyclohexanone (syn-
20 and anti-21): Obtained from cyclohexanone (7) and (Z)-
18; the reaction was performed through the lithium enolate
in the presence of lithium chloride and mediated with the
ligand 11d, quantitative conversion. The diastereomeric
ratio of syn-20 and anti-21 amounted to 79:21, according to
GC-MS. Diastereomerically pure samples were obtained by
column chromatography on silica gel (CHCl3/ethyl acetate,
5:1).
(2R,1’S,2’E,4’S)- and (2S,1’S,2’E,4’S)-2-{4-[(2-Methoxy-
ethoxy)methoxy]-1-phenylpent-2-enyl}cyclohexanone (syn-
22 and anti-23): Obtained from cyclohexanone (7) and (E)-
19; the reaction was performed through the lithium enolate
without an additive and mediated with the ligand 11d; quan-
titative conversion. The diastereomeric ratio of syn-22 and
anti-23 was determined to be 57:43. Diastereomerically pure
samples were obtained by column chromatography on silica
gel (CHCl3/ethyl acetate, 5:1).
(S)-2-Allylcyclohexanone (12); Typical Procedure for
the Allylic Alkylation of Lithium Enolates in the
Presence of Lithium Chloride
A 100-mL two-necked flask was equipped with a magnetic
stirrer and charged with [Pd2
(dba)3].CHCl3 (25.9 mg; 25
E
mmol), (S)-11k (65.8 mg, 101 mmol), and LiCl (0.51 g,
12 mmol). The flask was closed with a septum, connected to
a combined nitrogen/vacuum line, evacuated for 5 h at 258C
and filled with nitrogen. A solution of 1d (0.581 g, 0.57 mL,
5.0 mmol) in THF (13 mL) was added by syringe. The deep
purple mixture was stirred at 258C for approximately 0.5 to
1 h. In the course of this, the color changed to yellow. A
100-mL two-necked flask was equipped with a magnetic stir-
rer, a resistance thermometer, a connection to the combined
nitrogen/vacuum line and a septum. The air in the flask was
replaced by nitrogen, and diisopropylamine (0.7 mL,
5.0 mmol) and THF (10 mL) were injected. After cooling to
ꢀ788C, a 1.6M solution of n-BuLi in hexane (3.1 mL,
5.0 mmol) was added, whereby the temperature was kept
below ꢀ708C. After stirring at 08C for 30 min, it was cooled
again to ꢀ788C and treated with distilled, degassed cyclo-
hexanone (0.52 mL, 5.0 mmol) in 5 mL of THF. After stir-
ring at 08C for 30 min, the solution was cooled to ꢀ788C
and added by syringe to the first flask. After stirring at
ꢀ788C for 40 h, the mixture was poured into 100 mL of
phosphate buffer (pH 7) and extracted four times with
CH2Cl2 (50 mL each). The combined organic layers were
A
hepten-1-one (30): Obtained from acetophenone and (Z)-
18; the reaction was performed through the lithium enolate
in the presence of lithium chloride and mediated with the
ligand 11d; quantitative conversion. The NMR spectra did
not show the presence of diastereomers or regioisomers.
The crude product was purified by column chromatography
on silica gel (CHCl3/ethyl acetate, 5:1) to give pure 30;
yield: 83%.
A
hepten-1-one (32): Obtained from acetophenone and (E)-
19; the reaction was performed through the lithium enolate
in the presence of lithium chloride and mediated with the
ligand 11d. Conversion 95%. The NMR spectra and the GC
did not show the presence of diastereomers and or re-
gioisomers to an extent of more than 2%. The crude prod-
312
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 303 – 314

Palladium-Catalyzed Diastereoselective and Enantioselective Allylic Alkylations
FULL PAPERS
uct was purified by column chromatography on silica gel
(CHCl3/ethyl acetate, 5:1) to give pure 32; yield: 25%.
anti-2-Methyl-3,5-diphenyl-1-(2,4,6-trimethylphenyl)-4-
penten-1-one (36): Obtained from ketone 34 by deprotona-
tion with lithium diisopropylamide and subsequent reaction
with 1a. The reaction was performed without an additive
and mediated with the ligand 11d; quantitative conversion.
The diastereomeric ratio of anti-37 and syn-36 was deter-
mined to amount to 90:10. The major product anti-37 was
obtained as a pure diastereomer upon column chromatogra-
phy on silica gel (n-hexane/ethyl acetate, 10:1); yield: 85%.
measured by Dr. habil. Wolfgang Schrader, Max-Planck-In-
stitut für Kohlenforschung, Mülheim/Ruhr. Generous gifts of
ꢀ- and (S)-11k by LANXESS Deutschland GmbH are
kindly acknowledged.
References
[1] a) B. M. Trost, Tetrahedron 1977, 33, 2615; b) J. Tsuji,
Organic Synthesis with Palladium Compounds, Spring-
er, New York, 1980; c) B. M. Trost, Acc. Chem. Res.
1980, 13, 385; d) B. M. Trost, Pure Appl. Chem. 1981,
53, 2357; e) J. Tsuji, Pure Appl. Chem. 1982, 54, 197;
f) A. Godleski, in: Comprehensive Organic Synthesis
Vol. 4, (Ed.: B. M. Trost), Pergamon, Oxford, 1991,
p 585.
[2] For reviews, see: a) O. Reiser, Angew. Chem. 1993, 105,
576; Angew. Chem. Int. Ed. Engl. 1993, 32, 547;
b) J. M. J. Williams, Synlett 1996, 705; c) B. M. Trost,
D. L. Van Vranken, Chem. Rev. 1996, 96, 395; d) G.
Helmchen, J. Organomet. Chem. 1999, 576, 203; e) G.
Helmchen, A. Pfaltz, Acc. Chem. Res. 2000, 33, 336;
f) B. M. Trost, M. L. Crawley, Chem. Rev. 2003, 103,
2921; g) B. M. Trost, J. Org. Chem. 2004, 69, 5813.
[3] C. H. Heathcock, in: Modern Synthetic Methods 1992,
(Ed.: R. Scheffold), VHCA, VCH, Basel, Weinheim,
1992, p 1; and references cited therein.
(1-Cyclohexylethyl)benzene (17)
By means of a frit, a stream of ozone in oxygen was passed
through a solution of the alkene (2R,1’S)-9a (389 mg,
1.34 mmol) (99% ee) in CH2Cl2 (40 mL) at ꢀ788C until the
blue color persisted. Then, the solution was treated with a
stream of oxygen and nitrogen. Dimethyl sulfide (831 mg,
13.4 mmol) was added, the mixture was allowed to reach
room temperature overnight and poured into a buffered
aqueous solution (pH 7). The organic layer was separated
and the aqueous layer was extracted with three 50 mL por-
tions of CH2Cl2. The combined organic layers were dried
with MgSO4, the solvent was removed in a rotary evapora-
tor, and the residue was exposed to oil pump vacuum to
give crude oily (2-oxocyclohexyl)phenyl-acetaldehyde 15;
yield: 210 mg (72%).
[4] a) J.-C. Fiaud, J.-L. Malleron, Chem. Soc., Chem.
Commun. 1981, 1159; b) B. Akermark, A. J. Jutand,
Organomet. Chem. 1981, 217, C41; c) E. Negishi, H.
Matsushita, S. Chatterjee, R. A. John, J. Org. Chem.
1982, 47, 3188; d) B. M. Trost, E. Keinan, Tetrahedron
Lett. 1980, 21, 2591; e) B. M. Trost, C. R. Self, J. Org.
Chem. 1984, 49, 468; f) U. Kazmaier, Curr. Org. Chem.
2003, 7, 317; g) U. Kazmaier, F. L. Zumpe, Angew.
Chem. 1999, 111, 1572; Angew. Chem. Int. Ed. 1999, 38,
1468; h) T. D. Weiß, G. Helmchen, U. Kazmaier, Chem.
Commun. 2002, 1270.
[5] a) B. M. Trost, G. M. Schroeder, J. Am. Chem. Soc.
1999, 121, 6759; b) B. M. Trost, G. M. Schroeder, Chem.
Eur. J. 2005, 11, 174; c) S.-L. You, X.-L. Hou, L.-X.
Dai, X.-Z. Zhu, Org. Lett. 2001, 3, 149; d) X.-X. Yan,
C.-G. Liang, Y. Zhang, W. Hong, B.-X. Cao, L.-X. Dai,
X.-L. Hou, Angew. Chem. 2005, 117, 6702; Angew.
Chem. Int. Ed. 2005, 44, 6544; e) W.-H. Zheng, B. H.
Zheng, Y. Zhang, X.-L. Hou, J. Am. Chem. Soc. 2007,
129, 7718.
Under nitrogen,
a solution of crude 15 (210 mg,
0.97 mmol) in CH2Cl2 (5 mL) was stirred in a 25-mL flask,
equipped with a septum, a magnetic stirrer, an inlet ther-
mometer and a connection to the combined nitrogen/
vacuum line at 08C. 1,3-Propanediol (0.18 mL, 2.5 mmol)
was injected and, thereafter, boron trifluoride-etherate
(0.12 mL, 1.0 mmol) was added dropwise by syringe. The
mixture was allowed to warm up to room temperature and
stirred for 1 h. After the addition of 1N aqueous NaOH, the
mixture was extracted three times with CH2Cl2. The com-
bined organic layers were washed with water, dried with
MgSO4 and concentrated in a rotary evaporator. The crude
product was purified by two-fold column chromatography
on silica gel to give bis-thioacetal 16; yield: 50 mg (13%).
A suspension of Raney nickel in water (10 mL) was de-
canted, methanol was added and the supertenant liquid was
decanted. This procedure was repeated five times. To the re-
maining suspension, a solution of 16 (50 mg, 0.13 mmol) in
methanol (40 mL) was added. The mixture was refluxed for
10 h and thereafter filtered through celite 577, which was
rinsed with CH2Cl2. The combined filtrates were washed
with water and dried with MgSO4. Evaporation of the sol-
vent gave crude 17 (yield: 23 mg, 94%) as a yellowish oil,
whose NMR spectroscopic data were in accordance with
those described in the literature,[16]
[6] a) T. Gaening, J. F. Hartwig, J. Am. Chem. Soc. 2005,
127, 17192; b) E. Bꢂlanger, K. Cantin, O. Messe, M.
Tremblay, J.-F. Paquin, J. Am. Chem. Soc. 2007, 129,
1034; c) D. J. Weix, J. F. Hartwig, J. Am. Chem. Soc.
2007, 129, 7720.
[7] For recent reviews, see: a) M. Braun, T. Meier, Synlett
2006, 661; b) M. Braun, T. Meier, Angew. Chem. 2006,
118, 7106; Angew. Chem. Int. Ed. 2006, 45, 6952.
[8] a) E. C. Burger, J. A. Tunge, Org. Lett. 2004, 6, 4113;
b) D. C. Behenna, B. M. Stoltz, J. Am. Chem. Soc. 2004,
126, 15044; c) B. M. Trost, J. Xu, J. Am. Chem. Soc.
2005, 127, 2846; d) B. M. Trost, J. Xu, J. Am. Chem.
Soc. 2005, 127, 17180; e) B. M. Trost, R. N. Bream, J.
Xu, Angew. Chem. 2006, 118, 3181; Angew. Chem. Int.
Ed. 2006, 45, 3109; f) J. T. Mohr, D. C. Behenna, A. M.
Acknowledgements
This work was supported by the Deutsche Forschungsgemein-
schaft (Br 604–13/1–2 and 16/1). We would like to thank Pro-
fessor Dr. Jçrg Pietruszka, Institut für Bioorganische
Chemie, Forschungszentrum Jülich for chiral HPLC and GC
measurements. High resolution mass spectra were kindly
Adv. Synth. Catal. 2008, 350, 303 – 314
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
313

Manfred Braun et al.
FULL PAPERS
Harned, B. M. Stoltz, Angew. Chem. 2005, 117, 7084;
Angew. Chem. Int. Ed. 2005, 44, 6924; g) J. A. Tunge,
E. C. Burger, Eur. J. Org. Chem. 2005, 1715; h) S. Tru-
deau, J. P. Morken, Tetrahedron 2006, 62, 11470; i) H.
He, X.-J. Zheng, Y. Li, L.-X. Dai, S.-L. You, Org. Lett.
2007, 9, 4339; j) for a recent review, see: S.-L. You, L.-
X. Dai, Angew. Chem. 2006, 118, 5372; Angew. Chem.
Int. Ed. 2006, 45, 5246.
[14] So far, most procedures for enantioselective allylic al-
kylations of ketones rely on the use of azaenolates with
chiral auxiliary groups, cf: D. Enders, in: Asymmetric
Synthesis, Part B, Vol. 2 (Ed.: J. D. Morrison) Academ-
ic Press, New York, 1984, Chap. 4, and cited references
therein. For the few catalytic variants, see: refs.[8,15e]
[15] a) D. Enders, H. Eichenauer, Angew. Chem. 1976, 88,
579; Angew. Chem. Int. Ed. Engl. 1976, 15, 549; b) A. I.
Meyers, D. R. Williams, G. W. Erickson, S. White, M.
Druelinger, J. Am. Chem. Soc. 1981, 103, 3081; c) S.
Hashimoto, Y. Miyazaki, S. Ikegami, Synth. Commun.
1992, 22, 2717; d) A. Yanagisawa, T. Kikuchi, T. Kuri-
bayashi, H. Yamamoto, Tetrahedron 1998, 54, 10253;
e) M. Imai, A. Hagihara, H. Kawasaki, K. Manabe, K.
Koga, Tetrahedron 2000, 56, 179.
[9] Preliminary communications: a) M. Braun, F. Laicher,
T. Meier, Angew. Chem. 2000, 112, 3637; Angew. Chem.
Int. Ed. 2000, 39, 3494; b) M. Braun, T. Meier, Synlett
2005, 2968.
[10] T. Ukai, H. Kawazura, Y. Ishii, J. J. Bonnet, J. A. Ibers,
J. Organomet. Chem. 1974, 65, 253.
[11] For the preparation of 11b, see: S. Hillebrand, J. Bruck-
mann, C. Krüger, M. W. Haenel, Tetrahedron Lett.
1995, 36, 75. All other ligands shown in Scheme 4 are
commercially available and were purchased.
[12] Cf: a) S. Ramdeehul, P. Dierkes, R. Aguado, P. C. J.
Kamer, P. W. N. M. van Leeuwen, J. A. Osborn, Angew.
Chem. 1998, 110, 3302; Angew. Chem. Int. Ed. 1998,
37, 3118; b) G. C. Lloyd-Jones, S. C. Stephen, Chem.
Commun. 1998, 2321; c) B. J. Lussem, H.-J. Gais, J.
Org. Chem. 2004, 69, 4041.
[16] S. J. Blarer, W. B. Schweizer, D. Seebach, Helv. Chim.
Acta 1982, 65, 1637.
[17] a) O. Takazawa, K. Kogami, K. Hayashi, Bull. Chem.
Soc. Jap. 1984, 57, 1876; b) J. W. Faller, C. Lambert,
Tetrahedron 1985, 41, 5755.
[18] F. G. Bordwell, Acc. Chem. Res. 1988, 21, 456.
[19] M. Braun, C. Unger, K. Opdenbusch, Eur. J. Org.
Chem. 1998, 2389.
[20] T. Hayashi, A. Yamamoto, T. Hagihara, J. Org. Chem.
1986, 51, 723.
[13] For a review, see: D. Seebach, Angew. Chem. 1988, 100,
1685; Angew. Chem. Int. Ed. Engl. 1988, 27, 1624. The
lithium salt might also influence the aggragation of the
metal catalyst, cf: G. C. Lloyd-Jones, S. C. Stephen,
I. J. S. Fairlamb, A. Martorell, B. Dominguez, P. M.
Tomlin, M. Murray, J. M. Fernandez, J. C. Jeffery, T.
Riis-Johannessen, T. Guerziz, Pure Appl. Chem. 2004,
76, 589.
[21] J.-M. Lassaletta, R. Fernꢃndez, E. Martꢄn-Zamora, E.
Dꢄez, J. Am. Chem. Soc. 1996, 118, 7002.
[22] M. H. Nomura, Bull. Soc. Chim. Fr. 1925, 37, 1245.
[23] Cf.: V. Grignard, Chem. Zentralblatt 1901, 2, 622.
[24] H.-J. Gais, T. Jagusch, N. Spalthoff, F. Gerhards, M.
Frank, G. Raabe, Chem. Eur. J. 2003, 9, 4202.
[25] T. Hirao, T. Fujii, Y. Ohshiro, Tetrahedron 1994, 50,
10207.
314
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 303 – 314