Dynamic processes in the copper-catalyzed substitution of chiral allylic acetates lending to loss of chiral information

Article abstract of DOI:10.1002/chem.200601684

Copper-catalyzed a-substitution of enantiomerically pure secondary allylic esters with Grignard reagents was studied with the aim to find conditions that give racemic products. It was observed that the degree of chiral transfer is strongly dependent on th

Full text of DOI:10.1002/chem.200601684

DOI: 10.1002/chem.200601684
Dynamic Processes in the Copper-Catalyzed Substitution of Chiral Allylic
Acetates Leading to Loss of Chiral Information
Jakob Norinder and Jan-E. Bäckvall*[a]
Abstract: Copper-catalyzed a-substitu-
tion of enantiomerically pure secon-
dary allylic esters with Grignard re-
agents was studied with the aim to find
conditions that give racemic products.
It was observed that the degree of
chiral transfer is strongly dependent on
the temperature. The loss of chiral in-
formation is consistent with an equili-
bration of the CuIII
prior to product formation. Equilibra-
tion of the reaction intermediates is of
A
importance for a possible development
of a dynamic kinetic asymmetric trans-
formation (DYKAT) process, in which
a chiral catalyst is used to produce an
optically active product from a racemic
substrate, by means of a dynamic equi-
librium of the diastereomeric reaction
intermediates.
Keywords: allylic substitution
chirality copper dynamic
processes · heterogeneous catalysis ·
racemization
·
U
Introduction
2) There is a fast equilibration of the reaction intermedi-
ates, to ensure restoration of the intermediate from
which the desired product enantiomer is formed.
Asymmetric catalysis is currently a highly active area in syn-
thetic organic chemistry involving transition metals, en-
zymes, and organocatalysts.[1] One of the most common ap-
proaches involves the use of a transition metal with a chiral
ligand for the transformation of a prochiral substrate to an
enantiomerically enrichced product. In copper(I) chemistry,
this concept has been successfully exemplified in catalytic
enantioselective 1,4-additions[2] and g-substitution of allylic
electrophiles.[2b,3,4] However, to the best of our knowledge
no examples of dynamic kinetic asymmetric transformation
(DYKAT) reactions based on copper catalysis have been re-
ported.[5] We were interested in investigating the possibility
to perform this type of asymmetric allylic alkylation, in
which a racemic substrate is transformed to an optically
active product (Scheme 1).
Since a dynamic equilibration of the allyl intermediates
prior to product formation is a requirement for a successful
DYKAT, we decided to study this process separately. Hence,
the copper-catalyzed substitution of enantiopure 1 with an
achiral catalyst was studied. With a full dynamic equilibrium
prior to product formation a racemic product will be ob-
tained from enantiopure 1; in this case, 3 and 4 are enantio-
mers.[7]
As CuIII intermediates in allylation reactions are short-
lived, the rate of reductive elimination needs to be retarded.
Previous studies in our group have shown that dialkylcup-
rate, compared to the relatively more electron-deficient
monoalkylcopper species, provides kinetically more stable
CuIII intermediates from reaction with an allylic substrate.[8]
For this reason a catalytic amount of copper in combination
with a fast addition of the Grignard reagent to the reaction
mixture was chosen to ensure formation of dialkylcuprate as
the predominant catalytic species. To achieve reaction con-
ditions that favor an equilibration of the allyl intermediates
prior to product formation, different copper species, sol-
vents, Grignard reagents, and temperatures were examined.
It was found that the enantiomeric excess of the product
varied with the reaction conditions, the reaction tempera-
ture being the most important parameter.
According to the Curtin–Hammet principle, two criteria
must be fulfilled in order to obtain an optically active prod-
uct from a racemic substrate through interconversion of the
stereoisomeric intermediates (Scheme 1):[6]
1) There is a selective formation of product from one of the
diastereomeric intermediates (k5 @k5’).
[a] Dr. J. Norinder, Prof. Dr. J.-E. Bäckvall
Department of Organic Chemistry, Arrhenius Laboratory
Stockholm University, SE-106 91 Stockholm (Sweden)
Fax : (+46)8-154908
E-mail: jeb@organ.su.se
4094
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 4094 – 4102

FULL PAPER
The linear relationship be-
tween temperature and loss of
chiral information, as illustrated
in Figure 1, is in accordance
with a CuIII intermediate in the
copper-catalyzed allylic substi-
tution reactions. Since CuIII is a
d8 electron species, four ligands
are required in order to form a
16-electron square-planar com-
plex. It is likely that the CuIII
intermediate formed from oxi-
dative addition is the 16-elec-
tron
complex
7 L(a)
(Scheme 3). The ligand L may
be a solvent molecule or anoth-
er Lewis base moiety. In order
to form product from 16-elec-
tron complex 7 L(a) via p-allyl
complex 8, dissociation of
ligand L is required. It is likely
Scheme 1. Principle of copper(I)-catalyzed asymmetric allylic alkylation of a racemic substrate (L*=chiral
ligand).
Results and Discussion
that the ligand L dissociates to give the s-allyl transition
state in which the copper starts to interact with the alkene
p-system. Accordingly, the free energy of activation, DGꢀ,
for the conversion of 7 L(a) to 8 will have a positive DSꢀ
due to the loss of the ligand in the transition state. The com-
The allylic substitution reactions were performed by fast ad-
dition of Grignard reagent (within one minute) to a mixture
of substrate (R)-1 and a copper salt in the appropriate sol-
vent (Scheme 2). The reaction is highly regioselective and a-
Table 1. Variation of copper catalyst.[a]
“Cu”
t [h]
Conv [%][b] 5a [%][b] 6 [%][b] ee [%][c]
1
2
3
4
5
Li2CuCl4
CuCN
CuI
CuCl
CuBr·SMe2 3.75
3
90
60
>95
>95
>95
72 (64)
42 (36)
55
79 (67)
80 (72)
<5
22
15
<5
9
74
70
70
69
67
o.n.[d]
o.n.[d]
1.75
Scheme 2. Products formed in the in the allylation reaction.
product 5 was the only regioisomer observed. The alcohol
(R)-6, formed by alkyl anion attack at the carbonyl carbon
atom, was observed in small amounts in most experiments.
[a] Reaction conditions: Allylic acetate and “Cu” were mixed in THF.
After the mixture was cooled to 08C, nBuMgBr was added. [b] Deter-
mined by 1H NMR spectroscopy by using 2-decanol as internal standard.
Yield of isolated product in parentheses. [c] Determined by chiral HPLC.
[d] The reaction was run overnight.
Examination of copper sources: Different copper salts as
catalysts were investigated. Copper halides and CuCN gave
similar results concerning the conservation of chiral infor-
mation (Table 1). However, CuCN and CuI appear to pro-
vide a less reactive cuprate, since they gave lower yield and
more of alcohol 6 (Table 1, entries 2 and 3). Copper chloride
was chosen as the catalyst for further studies, since it afford-
ed a high yield of cross-coupling product 5a in a short reac-
tion time (Table 1, entry 4).
Table 2. Effect of temperature.[a]
T [8C]
t [h]
Conv [%][b]
5a [%][b]
6 [%][b]
ee [%][c]
1
2
3
4
5
6
7
22
0
4
1.75
93
>95
>95
>95
>95
47
74 (62)
79 (67)
93 (70)
93 (67)
74 (67)
25 (22)
3 (2)
12
<5
7
7
8
73
69
51
38
29
15
1
À20
À30
À40
À60
À78
13
12
16
16
24
6
4
12
Influence of reaction temperature: The stereochemical out-
come of the reaction was studied at different temperatures.
It was found that the conservation of chiral information de-
creased with decreasing reaction temperature (Table 2 and
Figure 1).
[a] Reaction conditions: Allylic acetate and CuCl (10 mol%) were mixed
in THF. After the mixture was cooled to the appropriate temperature,
nBuMgBr (1.5 equiv) was added within one minute. [b] Determined by
1H NMR spectroscopy by using 2-decanol as internal standard. Yield of
isolated product in parentheses. [c] Determined by chiral HPLC.
Chem. Eur. J. 2007, 13, 4094 – 4102
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4095

J.-E. Bäckvall and J. Norinder
therefore examined in order to
find reaction conditions under
which a racemic product could
be obtained in synthetically
useful yields.
Variation of solvents: A few
solvents were tested in the
copper-catalyzed allylic substi-
tution of (R)-1. The best result,
in terms of yield and racemic
product, was obtained with
THF as solvent (Table 3,
entry 1). In contrast, Et2O as
solvent gave 5a with high opti-
Scheme 3. Mechanistic explanation for the temperature-dependent loss of chiral information.
Table 3. Screening of solvents.[a]
Solvent
t [h]
conv [%][b]
5a [%][b]
6 [%][b]
ee [%][c]
1
THF
Et2O
CH3CN
21
18
25
ꢀ95
43
54
85
34
37
15
9
16
21
93
51
2[d]
3
[a] Reaction conditions: Allylic acetate and CuCl (10 mol%) were mixed
in the solvent stated in the Table. After the mixture was cooled to
À408C, nBuMgCl (1.5 equiv) was added. [b] Determined by 1H NMR
spectroscopy. [c] Determined by chiral HPLC. [d] As Grignard reagent
nBuMgBr was used.
cal purity (Table 3, entry 2). The latter reaction was rather
slow indicating that THF is a better coordinating solvent
than Et2O. Coordination by the solvent would provide a
better stabilization of s-allyl intermediate 7 and thereby fa-
cilitate the loss of chiral information. Acetonitrile has good
coordinating abilities and thus was tested as the solvent. Un-
fortunately, this did not increase the loss of chiral informa-
tion (Table 3, entry 3). Other solvents (CH2Cl2 and DMF)
and solvent mixtures (THF/DMA, PhMe/THF and DMSO/
THF) were tested over a range of reaction temperatures.
These experiments were not successful; the reactions were
often sluggish, gave low yields (0–74%) and the ee of the
product was in the range of 24% to 51%.
Figure 1. Product ee as a function of temperature.
peting transformation of 7 L(a) to 7 L(b) on the other hand
has a DSꢀ close to zero, since it only involves a single bond
rotation. Consequently, the formation of 8 from 7 L(a) will
be favored over formation of 7 L(b) at high temperatures,
whereas equilibration between rotamers 7 L(a) and 7 L(b) is
favored over rearrangement to 8 at low temperatures. As a
result conservation of chiral information is favored at high
temperatures, whereas the degree of racemization will in-
crease with decreased temperature.
To investigate if any other reaction mechanism may be
the cause of the formation of racemic product, control ex-
periments without the copper catalyst were performed. No
cross-coupling products were formed, demonstrating that
the alkylation of (R)-1 is catalyzed by copper. In some of
the experiments unreacted substrate (R)-1 and the by-prod-
uct alcohol 6 were recovered from the product mixture and
their enantiomeric purity was analyzed. The ee of the recov-
ered materials was identical to that of the starting material,
confirming that the loss of chiral information does occur in
the copper-mediated alkylation step of the allylic substrate.
Although a racemic product can be obtained at low tem-
perature, the yields under these conditions are too low to be
synthetically useful. Different solvents, organocopper re-
agents, Grignard reagents, additives, and substrates were
Examination of different Grignard reagents: The use of
nBuMgCl gave more loss of chiral information than
nBuMgBr in the copper-catalyzed reaction of (R)-1 to 5
(Table 4, entries 1 and 2). Surprisingly, no loss of chiral in-
formation was observed when aryl Grignard reagents were
used in place of nBuMgBr (Table 4, entries 3 and 4). The
very high conservation of chiral information observed when
employing aryl Grignard reagents could be due to the great-
er bulk of the aryl groups on copper. This would disfavor
the coordination of a plausible ligand to the copper atom in
s-allyl intermediate 7 L(a) and increase the rate of forma-
tion of 8. The bulky phenyl groups would also increase the
barrier for the single-bond rotation that interconverts ro-
tamers 7 L(a) and 7 L(b). Both of these effects would ex-
4096
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 4094 – 4102

Dynamic Transformations
FULL PAPER
Table 4. Variation of the Grignard reagents.[a]
Table 6. Effect of different additives.
Equiv
additive
Additive
t
Conv
[%][a]
5a
6
ee
t [h] Conv [%][b] 5 [%][b] 6 [%][b] ee [%][c]
[h]
[%][a]
[%][a]
[%][b]
RMgX
1[c]
2[c]
3[c]
4[c,d]
5[e]
6
7
8
9
–
–
10
23
7
ꢀ95
75
83
74
66
20
14
32
72
83
66
70
59
66 (48)
7
7
21
60
13
10
13
29
40
20
11
19
85
71
39
22
14
16
29
1
2
3
4
nBuMgBr
nBuMgCl
PhMgCl
16
21
23
ꢀ95
ꢀ95
57
74
85
42
8
15
9
29
21
1.0
1.0
1.0
–
1.0
0.22
0.11
0.11
0.11
1.0
2.5
MgBr2
LiOAc
LiOAc
–
P
P
P
ꢀ99.5
p-MeOPhMgBr 15
44
31 (25) 13
ꢀ99.5
72
26
22
21
16
16
16
22
20
15
n.o.[f]
13
[a] Reaction conditions: Allylic acetate and CuCl were mixed in THF.
G
After the mixture was cooled to À408C, Grignard reagent was added.
G
7
10
12
1
[b] Determined by H NMR spectroscopy. 2-Decanol was used as internal
U
standard for entries 1 and 3. Yield of isolated product in parentheses.
[c] Determined by chiral HPLC.
NEt3
10
11
12
DMAP
DMAP
pyridine
73
58
15
76[g]
33[g]
41[g]
24[g]
10[g]
11[g]
plain why there is no loss of chiral information with aryl
Grignard reagents.[9]
[a] Determined by 1H NMR spectroscopy. Yield of isolated product in
parentheses. [b] Determined by chiral HPLC. [c] nBuMgBr was used.
[d] À608C. [e] Copper acetate was used as catalyst and butylmagnesium
acetate as nucleophile. [f] Not observed. [g] 2-Decanol was used as inter-
nal standard.
Stoichiometric amounts of cuprates: Stoichiometric amounts
of organocuprate were tested in order to improve the rate
of product formation at low temperatures. When one equiv-
alent of Bu2CuMgCl was used at À808C, a low yield of
product with 11% ee was obtained (Table 5, entry 1). The
use of Bu2CuLi as nucleophile at À808C did produce a
nearly racemic product, but the yield was low (Table 5,
entry 2).
spect to loss of chiral information (19% ee) compared to the
use of CuCl as catalyst and BuMgCl as nucleophile
(21% ee) (Table 4, entry 2).
A series of neutrally charged Lewis bases were investigat-
ed in order to find an appropriate ligand to stabilize s-allyl
intermediate 7 (see Scheme 3) and thereby perhaps increase
the loss of chiral information. The rate of product formation
as well as the loss of chiral information decreased when tri-
methyl phosphite was present in the reaction mixture
(Table 6, entries 6–8). Monoalkylcopper reagents are known
to be poor nucleophiles compared to dialkylcuprates, and
reductive elimination is believed to occur faster from allyl
CuIII intermediates obtained from monoalkylcopper reagents
compared to those obtained from dialkylcuprates.[8] Conse-
quently, the decreased amount of cross-coupling product
and the high conservation of chiral information observed in
Table 5. Experiments using stoichiometric Bu2CuM at low temperature.[a]
[b]
M
t [h]
Conv [%][b]
5a [%]
6 [%][b]
ee [%][c]
1
2
MgCl
Li
84
84
32
28
16
12
16
15
11
ꢁ3
[a] Reaction conditions: Allylic acetate was added to a preformed solu-
tion of Bu2CuM in THF. [b] Determined by 1H NMR spectroscopy.
[c] Determined by chiral HPLC.
the presence of P
ACHTREUNG
cleophile in the allylation reaction is BuCu·P
AHCTREUNG
not a dibutylcuprate. As the enantiomeric excess of the
product was found to be a function of the amount trimethyl
phosphite present in the reaction mixture (Table 6, en-
tries 6–8), we believe that phosphite is reversibly coordinat-
ed to copper and that the relative amount of dibutylcuprate
is reduced accordingly. Consequently, phosphites are not ap-
propriate ligands for the purpose of obtaining racemic prod-
uct. On the other hand 4-dimethylaminopyridine (DMAP)
gave the opposite effect and the presence of 0.11 equivalents
of DMAP, gave product 5a with 14% ee compared to
29% ee in the absence of additives (Table 6, entry 10 versus
entry 1). No further improvement with respect to the loss of
chiral information was observed when the amount of
DMAP was increased from 0.11 to 1.0 equivalents (Table 6,
entry 11). This might be explained by the low solubility of
DMAP in THF at À408C. Pyridine is more soluble in THF
Effect of different additives: Changing from bromide to
chloride in the butylmagnesium halide reagent increased the
loss of chiral information (Table 4, entries 1 and 2). As this
may be an effect of the halide interacting with the reaction
intermediates, the influence of different salt additives was
examined. The presence of magnesium bromide increased
the conservation of chiral information (Table 6, entry 2),
whereas the presence of lithium acetate was found to de-
crease it (Table 6, entry 3). A decreased reaction tempera-
ture (À608C) in combination with the presence of one
equivalent of LiOAc gave after prolonged reaction time
59% yield of 5a with an ee of 11% (Table 6, entry 4).
Halide-free conditions were obtained by using copper ace-
tate as catalyst and butylmagnesium acetate as nucleophile
(Table 6, entry 5). This gave no significant change with re-
Chem. Eur. J. 2007, 13, 4094 – 4102
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4097

J.-E. Bäckvall and J. Norinder
Table 8. Effect of different substrates.[a]
than DMAP and consequently was tested as an analogue to
DMAP. However, neither the yield nor the loss of chiral in-
formation was improved in the presence of pyridine com-
pared to when no additive was used and 5a was obtained in
29%ee (Table 6, entry 12).
Variation of leaving group: It is known from previous stud-
ies on copper-mediated alkylation of allylic esters that oxi-
dative addition is the rate-determining step.[10] Consequently
better leaving groups than acetate were studied in order to
improve the product formation at low temperature. At-
tempts to prepare diethyl phosphate-, diphenylphosphinate-,
and trifluoroacetate esters of alcohol 6 were unsuccessful
and resulted in complex product mixtures in which elimina-
tion product and starting material was observed. Preparation
of the corresponding chloride from enantiopure 6 led to a
quantitative yield of apparently racemic product.[11] The tri-
chloroacetate ester of 6 was successfully prepared, although
this substrate did not give any cross-coupling product in the
reaction with BuMgCl as nucleophile. The electron-deficient
p-CF3-benzoate ester (R)-9 and perfluorobenzoate ester
(R)-10 were prepared from enantiopure alcohol 6. These
esters were more reactive than the original acetate (R)-1, al-
though the conversion at low temperatures was still too low
to give synthetically useful yields (Table 7).
Substrate
Product
Yield
[%][b]
Alcohol
[%][b]
Conv
[%][b]
ee
[%][c]
1
2
(R)-1
5a
14
15
16
74[d] (67)
58[d] (37)
36
8
15
ꢀ95
ꢀ95
ꢀ95
n.o.[g]
29
20[e]
49
(R)-11
(R)-12
(R)-13
3
ꢀ5
4[f]
85[d]
n.o.[g]
ꢀ97[e]
[a] Reaction conditions: Unless otherwise noted allylic acetate and CuCl
were mixed in THF. After the mixture was cooled to À408C, nBuMgBr
was added within one minute. [b] Yield of isolated product unless other-
wise noted. [c] Determined by chiral HPLC. [d] Determined by 1H NMR
spectroscopy, using 2-decanol as internal standard, except entry 4 which
was determined by GC with decane as internal standard. [e] Determined
by derivatization to the corresponding carboxylic acid (2-methylhexanoic
acid (entry 2) and 2-methylnonanoic acid (entry 4)) via oxidative cleav-
age of the double bond using cat. RuCl3/NaIO4
could be separated by GC. [f] The reaction was performed at 08C, using
CuBr·SMe2 as catalyst and HeptMgBr as nucleophile. [g] Not observed.
[14]
whose enantiomers
product 16 with an ee exceeding
97%, (Table 8, entry 4). The a-
product 16 was accompanied by
a small amount (6%) of g-prod-
Table 7. Effect of different leaving groups.[a]
uct 17.
The higher reactivity in com-
bination with the lower chirality transfer observed when
using substrates (R)-1, (R)-11, and (R)-12 compared to
when using substrate (R)-13 strongly indicates that aryl
groups on the allyl terminus provide a stabilization of the
allyl intermediates. This observation is also in accordance
with recent theoretical studies.[12,13]
Substrate
T
[8C]
t
conv
5a
6
ee
A
[%][b]
[%][b]
[%][b]
[%][c]
1[d]
2
3
(R)-1
(R)-9
(R)-10
À70
À75
À75
4
3
3
51[e]
48
45
34[e]
37
45
16[e]
11
10
8
6
n.o.[f]
Preparation of starting materials: The alcohols 6 and 18–20
were obtained from reaction of the appropriate aldehydes
with methylmagnesium chloride in THF. Kinetic resolution
of the racemic alcohols by the use of Candida antarctica
lipase B (CALB) and vinylacetate afforded (R)-1 and (R)-
11–13 in high overall yields and with high eeꢁs (Table 9). The
absolute configuration of the products were assigned accord-
ing to Kazlauskasꢁ rule.[15]
[a] Reaction conditions: Allylic ester and CuCl were mixed in THF.
After the mixture was cooled to the given temperature, nBuMgCl was
added. [b] Determined by 1H NMR spectroscopy. [c] Determined by
chiral HPLC. [d] 10 mol% of CuCl was used. [e] 2-Decanol was used as
internal standard. [f] Not observed.
Variation of the substrate: When various substrates were
tested it was found that the electron-deficient substrate (R)-
11 rendered the highest loss of chiral information and the
electron-rich substrate (R)-12 gave the lowest loss of chiral
information (Table 8, entries 1–3). A plausible explanation
of this observation is that electron-deficient substituents on
the allyl terminus increase the activation energy for cross-
coupling from CuIII allyl complexes.[12] Consequently the
rate of reductive elimination is decreased and the allyl inter-
mediates have more time to equilibrate, prior to product
formation. Substrate (R)-13 was less reactive than substrates
(R)-1, (R)-11, and (R)-12 and no product was obtained at
À408C. Allylic substitution of (R)-13 at 08C gave 85% of a-
Aldehydes 21 and 22 were prepared from the correspond-
ing benzophenone according to
a literature procedure
(Scheme 4).[16] The intermediate iminium salt in the prepara-
tion of 22 was resistant to hydrolysis and stirring over night
in an aqueous solution of ammonium chloride was necessary
to obtain aldehyde 22.
Conclusion
The conservation of chiral information in the copper-cata-
lyzed SN2 substitution of g-disubstituted allylic esters is
4098
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 4094 – 4102

Dynamic Transformations
FULL PAPER
Table 9. Kinetic resolution of alcohols 6 and 18–20.
Experimental Section
General remarks: 1H (400 or 300 MHz) and 13C (100 or 75 MHz) NMR
spectra were recorded on a Varian Mercury spectrometer. Chemical
shifts (d) are reported in ppm, with residual CHCl3 (d=7.26 and
77.16 ppm, respectively) as internal standard. Optical rotations were mea-
sured on a Perkins–Elmer 241 Polarimeter. Chiral chromatography analy-
ses were performed by HPLC or GC. For HPLC, Chiralcel OD-H and
Product
T
[8C]
t
Conv
[%][a]
CALB
ee
Yield
[%][e]
Chiralcel AD columns (0.46 cm
Ø * 25 cm) were used, flow rate
[h]
[mgmmolÀ1
]
[%][b]
0.5 mLminÀ1. For GC a CP-Chirasil-Dex CB column (25 m 1=0.32 mm)
was used, carrier gas; H2, flow rate; 1.8 mLminÀ1. GC program 1: Isother-
mal at 508C for 3 min, then increase 28CminÀ1 up to 908C, then increase
258CminÀ1 up to 2008C and thereafter isothermal at 2008C for 5 min.
GC program 2: Isothermal at 1008C for 5 min, then 48CminÀ1 up to
2008C and thereafter isothermal at 2008C for 5 min. Merck silica gel 60
(240–400 mesh) was used for flash chromatography and analytical thin-
layer chromatography (TLC) was performed on Merck precoated silica
gel 60-F254 plates. All experiments were performed by standard Schlenk
techniques under an argon atmosphere. Unless otherwise noted, all mate-
rials were obtained from commercial sources and used without further
purification. Grignard reagents were purchased from Aldrich or prepared
by standard procedures. All Grignard reagents were titrated with 0.100m
HCl(aq), by using bromthymol blue as indicator, prior to use. Diethyl
ether and THF were distilled from sodium benzophenone ketyl radical,
toluene was distilled from sodium, and dichloromethane was distilled
from calcium hydride prior to use. The absolute configurations of prod-
ucts obtained by kinetic resolution were assigned according to Kauzlaus-
kasꢁ rule.[15]
1
2
(R)-1
50
50
RT
RT
51
40
32
43
45
47
32
7.5
7.5
100
7.5
ꢀ99.5
ꢀ99.5
98.1
ꢀ99.5
45.2
45.2
40.5
16.6
(R)-11
(R)-12
(R)-13
3[c]
4[d]
1.75
[a] Determined by 1H NMR spectroscopy, except for entry 4, which was
determined by GC. [b] Determined by chiral HPLC, except for entry 4,
which was determined by chiral GC. [c] A large amount of enzyme and
low temperature was used in order to suppress decomposition of starting
material and product. [d] Diethyl ether was used as solvent. [e] Yield of
isolated product.
Preparation of aldehydes 21 and 22: Aldehydes 21 and 22 were prepared
according to a literature procedure.[16]
Scheme 4. Preparation of aldehydes 21 and 22.
3,3-Bis(p-methoxyphenyl)-2-propenal (22): A solution of TiCl4 in CH2Cl2
(124 mL, 1.0m) was added to a solution of 4,4’-dimethoxybenzophenone
(7.5 g, 31 mmol) in CH2Cl2 (150 mL), cooled on a salt/ice bath. Subse-
quently triethylamine (12.6 g, 124 mmol) was added slowly. After stirring
for 6 h, the reaction mixture was quenched by addition of saturated
NH4Cl(aq) and the resulting mixture was stirred over night. The two
phases were separated and the aqueous phase was extracted twice with
CH2Cl2. The combined organic phases were washed with brine and dried
(Na2SO4). After evaporation of volatiles, a dark oil (4.52 g) was isolated
which, according to 1H NMR consisted of 93% product and 7% starting
material. The crude product was used without further purification. The
NMR-spectral data were in accordance with literature data.[17]
strongly dependent on the reaction temperature. The tem-
perature effect provides strong support for an equilibration
of the transient allyl intermediates, prior to product forma-
tion and was rationalized by an entropy effect. This equili-
bration is a necessary condition for the development of a
copper-catalyzed DYKAT process. The present study has es-
tablished that one enantiomer of the allylic substrate can be
transformed into both enantiomers of the product (race-
mate) by means of such a dynamic equilibrium, showing
that a copper-catalyzed dynamic kinetic asymmetric trans-
formation (DYKAT) is possible.
Synthesis of racemic alcohols 6, 18–19: Alcohols 6, 18, and 19 were pre-
pared according to a literature procedure.[18]
4,4-Bis-(4-fluorophenyl)but-3-en-2-ol (18): The crude product was isolat-
ed in quantitative yield as colorless oil, which was used without further
purification.
When a series of allylic substrates were examined, it was
found that the electron-deficient substrates facilitate loss of
chiral information. This observation is in accordance with
4,4-Bis-(4-methoxyphenyl)but-3-en-2-ol (19): The crude product was pu-
rified by flash chromatography on neutral alumina starting with 1:1 pen-
tane/Et2O and thereafter gradually changing to 100% EtOAc. This gave
76% yield of a yellow oil.
that an electron-deficient allyl ligand on copper(III) is more
A
reluctant to participate in reductive elimination.[13] Hence,
the allyl intermediates will have more time to equilibrate
and for this reason a greater loss of chiral information is ob-
served.
4-Methyl-3-penten-2-ol (20): Compound 20 was prepared according to a
literature procedure.[19]
Synthesis of racemic esters 1 and 11–13: Racemic acetates 1 and 11–13
were prepared from alcohol 6 according to a literature procedure.[18]
It was observed that in the presence of DMAP or LiOAc
the loss of chiral information increased, and the best overall
result considering both yield and loss of chiral information
was obtained at À608C in the presence of LiOAc as an addi-
tive (59% yield, 11% ee).
3,3-Bis-(p-fluoro-phenyl)-1-methyl-allyl acetate (11): Colorless oil.
3,3-Bis-(p-methoxy-phenyl)-1-methyl-allyl acetate (12): White solid, m.p.
below 508C.
4-Methyl-2-pent-3-enyl acetate (13): The crude product was a pale yellow
oil.
Preparation of optically active esters (R)-1, (R)-11, (R)-12,and (R)-13
(R)-(+)-4,4-Diphenyl-2-acetoxy-but-3-ene ((R)-1): Racemic alcohol
6
(5.27 g, 23.5 mmol) and vinyl acetate (8.09 g, 94.0 mmol) were added to
CALB (178 mg) in toluene (120 mL), and the mixture was stirred at
508C for 51 h. The reaction mixture was filtered and volatiles were
evaporated in vacuo. The crude product was purified by flash chromatog-
Chem. Eur. J. 2007, 13, 4094 – 4102
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4099

J.-E. Bäckvall and J. Norinder
raphy first using 9:1 pentane/Et2O to give the acetate and then 7:3 pen-
tane/Et2O to elute the alcohol. The allylic acetate (R)-1 was isolated as a
colorless oil in 2.83 g (45.2%), ꢀ99.5% ee (OD-H column, hexane/
iPrOH 99.9:0.1, tr(R)=27.7, tr(S)=34.6 min), [a]2D2 =+48.8 (c=1.67 in
CHCl3). The NMR spectra were in accordance with the literature.[18]
4.40 (dq, J=9.0, 6.22 Hz, 1H), 3.84 (s, 3H), 3.78 (s, 3H), 1.65 (brs, 1H),
1.33 ppm (d, J=6.2, 3H); 13C NMR (75 MHz, CDCl3): d=159.3, 159.0,
141.9, 134.9, 132.0, 131.0, 130.6, 128.8, 113.7, 113.6, 65.9, 55.39, 55.37,
23.9 ppm; IR (KBr): n˜ =3379, 2971, 1607, 1506, 1243, 1030 cmÀ1; HRMS:
m/z calcd for C18H20O3 [M+]; 284.1412; found: 284.1404.
(S)-(À)-4,4-Diphenyl-2-hydroxybut-3-ene ((S)-6): Alcohol (S)-6 was iso-
lated from the above experiment in 2.87 g (54.4%) as a colorless oil,
73.7% ee (OD-H column, hexane/iPrOH 90:10, tr(S)=12.7, tr(R)=
15.3 min). A subsequent resolution of the residual alcohol under the
same reaction conditions gave, after 93 h, 2.46 g of the alcohol as a color-
less oil (99.5% ee), [a]2D2 =À62.6 (c=1.26 in CHCl3).
(R)-(+)-3,3-Bis-(p-fluorophenyl)-1-methylallyl acetate ((R)-11): Racemic
18 (6.51 g, 25.0 mmol) and vinyl acetate (8.61 g, 100 mmol) were added
to CALB (188 mg) in toluene (125 mL), and the mixture was stirred at
508C for 40 h. The reaction mixture was filtered and the volatiles were
evaporated in vacuo. The crude product was purified by flash chromatog-
raphy first with 9:1 pentane/Et2O to give the acetate and then with 2:1
pentane/Et2O to elute the alcohol. The allylic acetate (R)-11 was isolated
in 3.49 g (45.2%) as a colorless oil, ꢀ99.5% ee (AD column, i-hexane/
iPrOH 99.9:0.1, tr(R)=23.8 tr(S)=26.6 min), [a]2D2 =+51.2 (c=1.98 in
CHCl3); 1H NMR (300 MHz, CDCl3): d=7.23–7.15 (m, 4H), 7.12–7.04
(m, 2H), 7.00–6.92 (m, 2H), 6.00 (d, J=9.0 Hz, 1H), 5.37 (dq, J=9.0,
6.4 Hz, 1H), 2.02 (s, 3H), 1.34 ppm (d, J=6.4 Hz, 3H). 13C NMR
(R)-(+)-4-Methyl-2-pent-3-enyl acetate ((R)-(+)-13): Racemic alcohol 20
(3.4 g, 34 mmol) and vinyl acetate (2.93 g, 34 mmol) were added to
CALB (255 mg) in Et2O (75 mL). The mixture was stirred at RT and the
progress of the reaction was followed by GC. After 1.75 h (32% conver-
sion), the reaction mixture was filtered and the mixture of volatile prod-
ucts was concentrated in vacuo. The crude product was purified by prepa-
rative HPLC, using a gradient eluent (pentane/Et2O 10:1 to Et2O 100%).
The allylic acetate (R)-13 was isolated in 0.817 g (16.6%) as a yellow oil,
ee ꢀ99.5% (GC, tr(R)=24.0, tr(S)=24.3 min, GC Program 1), [a]D22
=
+31.5 (c=1.71 in CHCl3); 1H NMR (400 MHz, CDCl3): d=5.58 (dq,
J=8.9, 6.4 Hz, 1H), 5.16 (dhept, J=8.9, 1.4 Hz, 1H), 2.01 (s, 3H), 1.72
(d, J=1.4 Hz, 3H), 1.71 (d, J=1.4 Hz, 3H), 1.25 ppm (d, J=6.4 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=170.7, 136.5, 125.1, 68.4, 25.9, 21.6, 21.1,
18.5 ppm; IR (KBr): n˜ =2978, 1733, 1449, 1371, 1243 cmÀ1; HRMS: m/z
calcd for C8H14O2 [M+]: 142.0994; found: 142.0967. The alcohol (S)-20
was isolated in 1.52 g (44.6%) as a yellow oil, ee=39.3% (GC, tr(R)=
24.7, tr(S)=25.2 min, GC Program 1).
Preparation of allylic benzoates (S)-9 and (S)-10
(100 MHz, CDCl3): d=170.0, 162.6 (d, J
(C,F)=247.0 Hz), 141.8, 137.5 (d, J(C,F)=3.3 Hz), 134.7 (d, J
3.4 Hz), 131.2 (d, J(C,F)=3.3 Hz), 129.1 (d, J(C,F)=7.9 Hz), 128.4, 115.5
(d, J(C,F)=21.6 Hz), 115.1 (d, J(C,F)=21.5 Hz), 69.3, 21.2, 21.0 ppm;
ACHTREUNG
(S)-(+)-3,3-Diphenyl-1-methyl-allyl p-CF3-benzoate ((S)-(+)-9): Alcohol
(S)-6 (99.5% ee) (536 mg, 2.39 mmol) and pyridine (1.5 mL) were mixed
in CH2Cl2 (3.5 mL). The reaction flask was put in a salt/ice bath and p-
CF3-benzoyl chloride (0.70 g, 3.4 mmol) was added. The cooling bath was
removed and the reaction mixture was stirred for 5 h. Thereafter
ꢂ1.5 mL of water was added followed by Et2O. The organic phase was
washed twice with brine. The combined aqueous phases were extracted
with Et2O and the combined organic phases were washed with 1m
NaOH(aq) (3”20 mL), followed by drying with Na2SO4 and evaporation
of volatiles in vacuo. Purification through a short column of silica (pen-
tane/Et2O 4:1) afforded 958 mg (101%) of (S)-9 as a yellow oil. An ana-
lytical sample was isolated by semipreparative HPLC (Kromasil 250”
A
G
ACHTREUNG
U
ACHTREUNG
G
U
19F NMR (376 MHz, CDCl3): d=À114.6 (m, 1F), À114.8 ppm (m, 1F);
IR (KBr): n˜ =3047, 2983, 2935, 1739, 1733, 1603, 1515, 1506, 1236,
834 cmÀ1
;
HRMS: m/z calcd for C18H16F2O2 [M+]: 302.1118; found:
302.1120.
(S)-(À)-4,4-Bis(4-fluorophenyl)but-3-en-2-ol ((S)-(À)-18): Alcohol (S)-18
was isolated from the above experiment in 3.52 g (53.0%) as a colorless
oil, 81.9% ee (AD column, i-hexane/iPrOH 95:5, tr(S)=21.1, tr(R)=
22.7 min).
A subsequent resolution gave 2.96 g of the alcohol in
ꢀ99.5% ee, [a]2D2 =À35.0 (c=1.98 in CHCl3); 1H NMR (300 MHz,
CDCl3): d=7.25–6.90 (m, 8H), 6.01 (d, J=9.1 Hz, 1H), 4.34 (dq, J=9.1,
6.26 Hz, 1H), 1.87 (brs, 1H), 1.33 ppm (d, J=6.3 Hz, 3H); 13C NMR
20 mm 100 SIL 5 mm column, 10 mL flow, pentane/Et2O 40:1). [a]D26
=
1
+75.6 (c=1.19 in CHCl3); H NMR (300 MHz, CDCl3): d=8.14 (dm, J=
8.4 Hz, 2 H), 7.69 (dm, J=8.4 Hz, 2 H), 7.43–7.20 (m, 10 H), 6.18 (d, J=
8.9 Hz, 1H), 5.65 (dq, J=8.9, 6.3 Hz, 1H), 1.49 ppm (d, J=6.3, 3H).
13C NMR (100 MHz, CDCl3): d=164.5, 144.7, 141.4, 139.0, 134.4 (q, J-
(75 MHz, CDCl3): d=162.6 (d, J
247.0 Hz), 140.7, 137.9 (d, J(C,F) =3.3 Hz), 135.1 (d, J
132.6 (d, J(C,F) =1.6 Hz), 131.4 (d, J(C,F)=8.0 Hz), 129.2 (d, J
8.1 Hz), 115.5 (d, J
C
G
U
ACHTREUNG
G
U
(C,F)=33.0 Hz), 134.1, 130.1, 129.6, 128.6, 128.4, 128.0, 127.8, 127.6,
E
A
125.4, (q, J
N
N
19F NMR (376 MHz, CDCl3): d=À63.4 ppm (s, 3F); IR (KBr): n˜ =3059,
2982, 1732, 1324, 1270, 1168, 863 ppm; HRMS: m/z calcd for C24H19F3O2
[M+]: 396.1337; found: 396.1338.
(m, 1F); IR (KBr): n˜ =3394, 2974, 2928, 1603, 1510, 1231, 835 cmÀ1
HRMS: m/z calcd for C16H14F2O [M+]: 260.1013; found: 260.1016.
(R)-(+)-3,3-Bis(p-methoxyphenyl)-1-methylallyl acetate ((R)-12): Race-
mic alcohol 19 (3.67 g, 12.9 mmol) and vinyl acetate (4.44 g, 51.6 mmol)
were added to CALB (1.29 g) in toluene (60 mL), and the mixture was
stirred at RT for 30 h. The reaction mixture was filtered and the volatiles
were evaporated in vacuo. The crude product was purified by flash chro-
matography on basic alumina using 4:1 pentane/EtOAc to give the ace-
tate and thereafter the eluent was gradually changed to 100% EtOAc to
elute the alcohol. The allylic acetate (R)-12 was isolated in 1.70 g
(40.5%) as a colorless oil, 98.1% ee (AD column, hexane/iPrOH 95:5,
tr(R)=14.9, tr(S)=17.5 min). [a]2D2 =+40.8 (c=1.63 in CHCl3). 1H NMR
(300 MHz, CDCl3): d=7.13 (dm, J=8.6 Hz, 2 H), 7.10 (dm, J=8.6 Hz, 2
H), 6.90 (dm, J=8.8 Hz, 2 H), 6.80 (dm, J=8.8 Hz, 2 H), 5.91 (d, J=
9.1 Hz, 1H), 5.39 (dq, J=9.1, 6.3 Hz, 1H), 3.83 (s, 3H), 3.79 (s, 3H), 2.01
(s, 3H), 1.32 ppm (d, J=6.3, 3H); 13C NMR (75 MHz, CDCl3): d=170.2,
159.5, 159.1, 143.2, 134.6, 131.6, 130.8, 128.8, 126.2, 113.8, 113.6, 69.9,
55.4, 55.3, 21.5, 21.3 ppm; IR (KBr): n˜ =2963, 1730, 1606, 1511, 1237,
1048, 1033, 838 cmÀ1; HRMS: m/z calcd for C20H22O4 [M+]: 326.1518;
found: 326.1511.
(S)-3,3-Diphenyl-1-methyl-allyl perfluorobenzoate ((S)-10): Alcohol (S)-6
(99.5% ee) (528 mg, 2.35 mmol), pyridine (0.305 mL, 3.77 mmol), and
DMAP (36 mg, 0.29 mmol) were mixed in CH2Cl2 (19 mL). At À608C
perfluorobenzoyl chloride (0.42 mL, 2.9 mmol) was added. The reaction
mixture was stirred for 7 h, while the temperature was allowed to reach
08C. While still kept in cooling bath, water was added. The aqueous
phase was extracted with Et2O (2”10 mL). The combined organic phases
were washed twice with NaHCO3(aq) and subsequently twice with water.
After drying over Na2SO4, volatiles were evaporated in vacou, which
gave 899 mg (91%) of crude product as a pale yellow solid. The crude
product was used without further purification. 1H NMR (400 MHz,
CDCl3): d=7.45–7.33 (m, 3 H), 7.32–7.18 (m, 7 H), 6.11 (d, J=9.1 Hz,
1H), 5.64 (dq, J=9.1, 6.4 Hz, 1H), 1.49 ppm (d, J=6.4 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=158.2, 145.6, 145.4 (dm,
256.0 Hz), 143.2 (dm, (C,F)=258.7 Hz), 141.3, 138.9, 137.9 (dm, J-
J
A
J
ACHTREUNG
A
73.0, 21.2 ppm; 19F NMR (376 MHz, CDCl3): d=À139.2 (m, 2F), À149.8
(tt J=20.8, 4.2 Hz, 1F), À161.0 ppm (m, 2F); IR (KBr): n˜ =3084, 2992,
1737, 1505, 1233, 992, 766, 702 cmÀ1; HRMS: m/z calcd for C23H15F5O2
[M+]: 418.0992; found: 418.1001.
(S)-4,4-Bis(4-methoxyphenyl)but-3-en-2-ol ((S)-19): The alcohol (S)-19
was isolated from the above experiment in 1.51 g (42.0%) as a yellow oil,
56.7% ee (AD column, hexane/iPrOH 90:10, tr(S)=34.6, tr(R)=
39.8 min). 1H NMR (300 MHz, CDCl3): d=7.21–7.15 (m, 2H), 7.14–7.08
(m, 2H), 6.94–6.88 (m, 2H), 6.85–6.78 (m, 2H), 5.95 (d, J=9.0 Hz, 1H),
General procedure for the copper(I)-catalyzed allylic substitution reac-
tions: The allylic acetate (R)-1 (123.8 mg 0.465 mmol, >99.5% ee) and
CuCl (4.6 mg, 0.046 mmol) were mixed in THF (4 mL) and cooled to
4100
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 4094 – 4102

Dynamic Transformations
FULL PAPER
À408C. A solution of nBuMgBr in THF (1.06 mL, 0.66m) was added
over a period of one minute and the reaction mixture was stirred at the
appropriate temperature. After 16 h the reaction mixture was quenched
by addition of HCl(aq) (2m, 10 mL) followed by addition of 2-decanol
(51.3 mg, 0.324 mmol) as internal standard. The aqueous phase was ex-
tracted with Et2O (3”10 mL) and the combined organic phases were
washed with H2O (10 mL), followed by drying with Na2SO4 and evapora-
tion of volatiles in vacuo. The formation of 5a was determined to be
74%, together with 8% of alcohol 6, and the conversion to >95%, ac-
(R)-1,1-Diphenyl-3-(p-methoxyphenyl)butene ((R)-5c): The yield of (R)-
5c was determined to 31% and the amount of 6 to 13%, by 1H NMR
analysis of the crude product. Column chromatography (pentane/Et2O
50:1) gave 54.5 mg (25%) of (R)-5c as a colorless solid (m.p.=106.4–
108.18C) in ꢀ99.5% ee (AD column, i-hexane:iPrOH 99.9:0.1, tr(S)=
15.7, tr(R)=18.8 min), [a]2D5 =+107 (c=1.29 in CHCl3); 1H NMR
(100 MHz, CDCl3): d=7.45–7.32 (m, 3H), d=7.30–7.19 (m, 7H), 7.15
(dm, J=8.9 Hz, 2H), 6.86 (dm, J=8.9 Hz, 2H), 6.21 (d, J=10.4 Hz, 1H),
3.81 (s, 3H), 3.58 (dq, J=10.4, 6.9 Hz, 1H) 1.39 ppm (d, J=6.9 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=158.0, 142.6, 140.3, 140.0, 138.5, 134.6,
130.0, 128.4, 128.2, 128.0, 127.4, 127.2, 127.1, 114.1, 55.4, 38.6, 22.6 ppm;
IR (KBr): n˜ =2957, 1609, 1511, 1249, 1233, 1177, 1035, 833 cmÀ1; HRMS:
m/z calcd for C23H22O [M+]; 314.1671; found: 314.1672.
1
cording H NMR analysis of the crude product. Column chromatography
(100% pentane) afforded 82.5 mg (74%) of 5a as
a colorless oil,
29.3% ee (OD-H column, 100% i-hexane, tr(R)=9.9, tr(S)=10.5 min).
Preparation of BuMgOAc: A solution of nBu2Mg in heptane (1.88 mL,
1.0m) was added to a solution of acetic acid (113 mg, 1.88 mmol) in THF
(3 mL) at À408C. The reaction mixture was stirred for 25 min at À408C
and used immediately thereafter.
(R)-1,1-Di(p-fluorophenyl)-3-methylheptene ((R)-14): The conversion
was determined to be >95%, the yield of (R)-14 to be 58%, and the
1
amount of alcohol 18 to be 15%, by H NMR analysis of the crude prod-
uct. Column chromatography (pentane 100%) gave 112.1 mg (37%) of
(R)-14 as a colorless oil. The optical purity was analyzed by GC of the 2-
methyl-hexanoic acid (20% ee), obtained by oxidative cleavage of the
olefin.[14] 1H NMR (300 MHz, CDCl3): d=7.22–7.01 (m, 6H), 7.00–6.89
(m, 2H), 5.80 (d, J=10.2 Hz, 1H), 2.34–2.14 (m, 1H), 1.42–1.08 (m, 6H),
1.01 (d, J=6.6 Hz, 3H), 0.85 ppm (t, J=7.0 Hz, 3H); 13C NMR (75 MHz,
Allylic substitution of (R)-1 mediated by stoichiometric amounts of
Bu2CuMgCl: nBuMgCl in THF (0.87 mL 2.3m) added to a suspension of
CuCl (49.5 mg 0.500 mmol) in THF (25 mL) at À208C was. The mixture
was stirred at the given temperature for a half an hour. After cooling the
reaction mixture to À808C, the allylic acetate (R)-1 (133 mg 0.500 mmol,
>99.5% ee) in THF (5 mL) was added. After 3.5 days HCl(aq) (2m, 3 mL)
was added slowly to the reaction mixture, while still in the cooling bath.
Subsequently, most of the THF was evaporated and the aqueous phase
was extracted with Et2O (3”10 mL) and subsequently the combined or-
ganic phases were washed with H2O (10 mL), followed by drying with
Na2SO4 and evaporation of volatiles in vacuo. The formation of 5a was
determined to 16% together with 16% of alcohol (R)-6 according to
1H NMR analysis of the crude product. The optical purity of 5a (11% ee)
was determined by chiral HPLC.
CDCl3): d=162.2 (d, J
138.9 (m), 138.3, 136.9, 136.2 (d, J
205.3 Hz), 130.0 (d, J(C,F)=205.6 Hz), 115.3 (d, J
(d, J
(C,F)=21.3 Hz), 37.5, 33.9, 29.9, 22.9, 21.5, 14.2 ppm; 19F NMR
A
ACHTREUNG
A
ACHTREUNG
A
ACHTREUNG
AHCTREUNG
(376 MHz, CDCl3): d=À116.0 (m, 1F), À116.5 ppm (m, 1F); IR (KBr):
n˜ =2959, 2927, 2871, 2856, 1603, 1512, 1506, 1232, 1157, 835 cmÀ1
HRMS: m/z calcd for C22H20F2 [M+]: 300.1690; found: 300.1688.
;
(R)-1,1-Di(p-methoxyphenyl)-3-methylheptene ((R)-15): Column chro-
matography (pentane/Et2O 19:1) gave 54.5 mg (36%) of (R)-15 as a col-
orless oil in 49% ee (AD column, hexane:iPrOH 99.9:0.1, tr(R)=15.0,
tr(S)=16.3 min). 1H NMR (300 MHz, CDCl3): d=7.15 (dm, J=8.9 Hz, 2
H), 7.08 (dm, J=8.9 Hz, 2 H), 6.89 (dm, J=9.0 Hz, 2 H), 6.79 (dm, J=
9.0 Hz, 2 H), 5.72 (d, J=10.2 Hz, 1H), 3.84 (s, 3H), 3.78 (s, 3H), 2.34–
2.20 (m, 1H), 1.36–1.09 (m, 6H), 0.99 (d, J=6.5 Hz, 3H), 0.84 ppm (t,
J=7.1 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=158.7, 158.5, 139.2,
136.0, 134.9, 133.3, 131.0, 128.3, 113.6, 113.5, 55.4, 55.3, 37.6, 33.8, 29.9,
22.9, 21.7, 14.2 ppm; IR (KBr): n˜ =2956, 2926, 1607, 1286, 1243, 1037,
830, 806 cmÀ1; HRMS: m/z calcd for C22H28O2 [M+]: 324.2089; found:
324.2078.
Allylic substitution of (R)-1 mediated by stoichiometric amounts of
Bu2CuLi: nBuLi in hexanes (1.25 mL 1.6m) was added to a suspension of
CuCl (99.0 mg 1.00 mmol) in THF (6 mL) at À208C. The mixture was
stirred at the given temperature for 15 min. After cooling the reaction
mixture to À808C, the allylic acetate (R)-1 (133 mg 0.500 mmol,
>99.5% ee) in THF (2 mL) was added. After 3.5 days HCl(aq) (2m, 3 mL)
was added slowly to the reaction mixture, while still in the cooling bath.
Thereafter ꢂ10 mL of water was added and the aqueous phase was ex-
tracted with Et2O (3”10 mL) and the combined organic phases were
washed with H2O (10 mL), followed by drying with Na2SO4 and evapora-
tion of volatiles in vacuo. The formation of 5a was determined to be
12% together with 15% of alcohol (R)-6 according to 1H NMR analysis
of the crude product. The optical purity of 5a (ꢁ 3% ee) was determined
by chiral HPLC.
(R)-2,4-Dimethyl-2-undecene ((R)-16): According to GC, using decane
as internal standard, 85% of a-product (R)-16 was obtained together
with 6% of the corresponding g-product (17). The optical purity of (R)-
16 was determined by transformation to 2-methyl-nonanoic acid
(ꢀ97% ee), obtained by oxidative cleavage of olefin,[14] and subsequent
analysis by chiral GC. 1H NMR (400 MHz, CDCl3): d=4.87 (d, J=
9.3 Hz, 1 H), 2.34–2.23 (m, 1 H), 1.68 (d, J=1.4 Hz, 3 H), 1.60 (d, J=
1.4 Hz, 3 H), 1.35–1.10 (m, 12 H), 0.89 (d, J=6.7 Hz, 3 H), 0.88 ppm (t,
J=6.9 Hz, 3 H).
Cross-coupling products 5a–c and 14–16: As reference compounds for
chiral chromatography racemic 5a, 5b (colorless oils), 5c (colorless
solid), 14, 15 (colorless oils) and 16 were prepared from the appropriate
racemic allylic acetates.
(R)-(À)-1,1-Diphenyl-3-methylheptene ((R)-5a): 1H NMR (300 MHz,
CDCl3): d=7.41–7.15 (m, 10H), 5.87 (d, J=10.3 Hz, 1H), 2.36–2.20 (m,
1H), 1.39–1.10 (m, 6H) 1.01 (d, J=6.6 Hz, 3H) 0.84 ppm (t, J=7.1 Hz,
3H); 13C NMR (75 MHz, CDCl3): d=142.9, 140.8, 140.2 (2 C), 136.6,
130.0, 128.3, 128.2, 127.2, 126.9, 37.5, 33.9, 29.9, 22.9, 21.6, 14.2 ppm; opti-
4,4-Dimethyl-2-undecene (17): Only those peaks that could be distin-
guished in a 1H NMR spectrum of a mixture of 16 and 17 are reported.
1H NMR (400 MHz, CDCl3): d=5.37 (d, J=15.5 Hz, 1 H), 5.29 (dq, J=
15.3, 5.9 Hz, 1 H), 1.65 (d, J=5.9 Hz, 3 H), 0.94 ppm (s, 6 H).
cal rotation was estimated for a sample of 5a with an ee of 92.5%; [a]D24
=
À85.2 (c=1.03 in CHCl3); IR (KBr): n˜ =2957, 2924, 1599, 1495, 1445,
762, 701 cmÀ1; HRMS: m/z calcd for C20H24 [M+]: 264.1878; found:
264.1873.
Determination of optical purity by oxidation to the corresponding car-
boxylic acids: Racemic 2-methylhexanoic acid and 2-methylnonanoic acid
were prepared for GC-references. The characterization of 2-methylnona-
noic acid will be reported elsewhere.[20]
(S)-(+)-1,1,3-Triphenylbutene ((S)-5b): Column chromatography (100%
pentane) gave a colorless oil, which formed a colorless solid on standing
(m.p.=59.8–63.98C). The optical purity (ꢀ99.5% ee) was determined by
chiral HPLC (OD-H column, 100% i-hexane tr(R)=21.7, tr(S)=
2-Methylhexanoic acid: The Sharpless procedure[14] was followed: cross-
coupling product 14 (38.1 mg, 0.127 mmol) and NaIO4 (128 mg,
0.598 mmol) were mixed in a 1:1 mixture of CCl4 and CH3CN (2 mL). A
solution of RuCl3·H2O (2.6 mg, 12.5 mmol) in H2O (1.5 mL) was added to
the flask and the reaction mixture was vigorously stirred. The reaction
was completed within 3 h, according to TLC. Thereafter, Et2O (10 mL)
and NaHCO3(aq) (10 mL) were added. The organic phase was subsequent-
ly extracted with NaHCO3(aq) (3”10 mL). The aqueous phases were com-
bined and concentrated HCl(aq) was carefully added until the pH reached
2. The product was extracted from the aqueous phase with CH2Cl2 (3”
1
23.7 min), [a]2D2 =+80.6 (c=1.83 in CHCl3); H NMR (400 MHz, CDCl3):
d=7.43–7.13 (m, 15H), 6.21 (d, J=10.4 Hz, 1H), 3.60 (dq, J=10.4,
6.8 Hz, 1H), 1.38 ppm (d, J=6.8 Hz, 3H); 13C NMR (100 MHz, CDCl3):
d=146.3, 142.5, 140.3, 140.2, 134.3, 130.0, 128.6, 128.4, 128.2, 127.4, 127.2,
127.13, 127.08, 126.1, 39.4, 22.5 ppm: IR (KBr): n˜ =2960, 2920, 1493,
1443, 767, 754, 699 cmÀ1; HRMS: m/z calcd for C22H20 [M+]: 284.1565;
found: 284.1562.
Chem. Eur. J. 2007, 13, 4094 – 4102
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4101

J.-E. Bäckvall and J. Norinder
10 mL). The combined organic phases were dried over Na2SO4 and the
product was concentrated in vacuo, giving 14.6 mg (88%) of crude prod-
uct as a colorless oil, 20% ee (GC, tr(S)=10.1, tr(R)=10.6 min, GC pro-
gram 2). The 1H NMR and 13C NMR spectral data were in accordance
with literature data.[21]
[5] For related studies on palladium-catalyzed substitution reactions of
allylic substrates see: a) P. R. Auburn, P. B. Mackenzie, B. Bosnich,
J. Am. Chem. Soc. 1985, 107, 2033; b) P. B. Mackenzie, J. Whelan, B.
Bosnich, J. Am. Chem. Soc. 1985, 107, 2046; c) B. M. Trost, R. C.
Bunt, R. C. Lemoine, T. L. Calkins, J. Am. Chem. Soc. 2000, 122,
5968; d) B. M. Trost, Chem. Pharm. Bull. 2002, 50, 1.
[6] a) J. I. Seeman, Chem. Rev. 1983, 83, 83; b) R. E. Gawley, J. AubØ, J.
Baldwin, Principles of Asymmetric Synthesis (Eds.: J. E. Baldwin,
P. D. Magnus), Elsevier Science, 1995.
[7] In this case k2 =k-2, k3 =k4, kÀ3 =kÀ4, and k5 =k’5. Accordingly, the
objective is to find reaction conditions under which k2 and kÀ3 @k5.
[8] a) E. S. M. Persson, J.-E. Bäckvall, Acta Chem. Scand. 1995, 49, 899;
b) J.-E. Bäckvall, M. SellØn, B. Grant, J. Am. Chem. Soc. 1990, 112,
6615; c) J.-E. Bäckvall, M. SellØn, J. Chem. Soc. Chem. Commun.
1987, 827; d) A. S. E. Karlstrçm, J. E. Bäckvall, Chem. Eur. J. 2001,
7, 1981–1989.
[9] Attempts to use more sterically hindered alkyl Grignard reagents
such as cyclohexylmagnesium bromide or isopropylmagnesium bro-
mide gave complex product mixtures.
2-Methylnonanoic acid: An analytical sample of a 93:7 mixture of a-
product (R)-16 and g-product 17 were oxidised, according to the Sharp-
less procedure.[14] The optical purity (ꢀ97% ee) of the (R)-2-methylnona-
noic acid obtained from (R)-16 was determined by chiral GC (tr(S)=
23.1, tr(R)=23.5 min, GC program 2).
Acknowledgements
Financial support from the Swedish Research Council is gratefully ac-
knowledged.
[10] L. H. Goering, S. S. Kantner, E. P. Seitz, J. Org. Chem. 1985, 50,
5495.
[1] a) J. Halpern, B. M. Trost, Proc. Natl. Acad. Sci. USA 2004, 101,
5347; b) Catalytic Asymmetric Synthesis, 2nd ed. (Ed.: I. Ojima),
Wiley-VCH, Weinheim , 2000; c) Comprehensive Asymmetric Catal-
ysis (Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamoto), Springer,
Berlin, 1999; d) R. Noyori in Asymmetric Catalysis in Organic Syn-
thesis, Wiley, New York, 1994; e) O. Pàmies, J. E. Bäckvall, Chem.
Rev. 2003, 103, 3247; f) K. Faber, Biotransformations in Organic
Chemistry, 5th ed, Springer: Berlin, 2004; g) P. I. Dalko, L. Moisan,
Angew. Chem. 2004, 116, 5248; Angew. Chem. Int. Ed. 2004, 43,
5138.
[11] Optical rotation of this product was close to zero. In the absence of
a reference value, the product was considered practically racemic.
[12] This has been suggested from theoretical studies on the reductive
elimination from allyl CuIII complexes. M. Yamanaka, S. Kato, E.
Nakamura, J. Am. Chem. Soc. 2004, 126, 6287.
[13] J. Norinder, J.-E. Bäckvall, N. Yoshikai, E. Nakamura, Organometal-
lics 2006, 25, 2129.
[14] P. H. J. Carlsen, T. Katsuki, V. S. Martin, K. B. Sharpless, J. Org.
Chem. 1981, 46, 3936.
[2] a) B. L. Feringa, R. Naasz, R. Imbos, A. L. Arnold, In Modern Or-
ganocopper Chemistry, (Ed.: N. Krause), Wiley-VCH, Weinhem
(Germany), 2002, p. 224; b) S. Woodward, Angew. Chem. 2005, 117,
5696; Angew. Chem. Int. Ed. 2005, 44, 5560; c) A. Alexakis, C. Ben-
heim, Eur. J. Org. Chem. 2002, 3221; d) N. Krause, A. Hoffmann
Rçder, Synthesis 2001, 171.
[3] a) A. S. E. Karlstrçm, J.-E. Bäckvall, In Modern Organocopper
Chemistry (Ed.: N. Krause), Wiley-VCH, Weinhem (Germany),
2002, p. 259; b) A. Alexakis, C. Malan, L. Lea, K. Tissot-Croset, D.
Polet, C. Falciola, Chimia, 2006, 60, 124; c) H. Yorimitsu, K.
Oshima, Angew. Chem. 2006, 118, 4509; Angew. Chem. Int. Ed.
2006, 45, 4435.
[4] For some recent selected references on Cu-catalyzed enantioselec-
tive 1,4-additions and g-substitution of allylic electrophiles see: a) F.
López, S. R. Harutyunyan, A. Meetsma, A. J. Minnaard, B. L. Ferin-
ga, Angew. Chem. 2005, 117, 2812; Angew. Chem. Int. Ed. 2005, 44,
2752; b) J. J. Van Veldhuizen, J. E. Campbell, R. E. Giudici, A. H.
Hoveyda, J. Am. Chem. Soc. 2005, 127, 6877; c) C. A. Falciola, K,
Tissot-Croset, A. Alexakis, Angew. Chem. 2006, 118, 6141; Angew.
Chem. Int. Ed. 2006, 45, 5995.
[15] a) M. Cygler, P. Grochulski, R. J. Kazlauskas, J. D. Schrag, F. Bou-
thillier, B. Rubin, A. N. Serriqi, A. K. Gupta, J. Am. Chem. Soc.
1994, 116, 3180; b) R. J. Kazlauskas, A. N. E. Weissfloch, A. T. Rap-
paport, L. A. Cuccia, J. Org. Chem. 1991, 56, 2656.
[16] P. Bharathi, M. Periasami, Org. Lett. 1999, 1, 857.
[17] M. Nyerges, A. Arany, I. Fejes, P. W. Groundwater, W. Zhang, D.
Bendell, R. J. Anderson, L. Tꢂke, Tetrahedron 2002, 58, 989.
[18] G. J. Dawson, J. M. J. Williams, Tetrahedron: Asymmetry 1995, 6,
2535.
[19] J. L. Stavinoha, P. S. Mariano, A. Leone-Bay, R. Swanson, C. Brack-
en, J. Am. Chem. Soc. 1981, 103, 3148.
[20] J. Norinder, K. Bogꢃr, L. Kanupp, J.-E. Bäckvall, unpublished re-
sults.
[21] J. L. Belelie, J. M. Chong, J. Org. Chem. 2001, 66, 5552.
Received: November 24, 2006
Published online: February 19, 2007
4102
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 4094 – 4102