Lewis acid/CpRu dual catalysis in the enantioselective decarboxylative allylation of ketone enolates

Article abstract of DOI:10.1039/b910475e

The addition of a Lewis acidic metal triflate salt Mg(OTf)₂ as co-catalyst in the CpRu-catalyzed decarboxylative allylation of in situ-generated ketone enolates allows the reaction to proceed at lower temperature with higher regio- and enantios

Full text of DOI:10.1039/b910475e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Lewis acid/CpRu dual catalysis in the enantioselective decarboxylative
allylation of ketone enolates†
David Linder, Martina Austeri and Je´roˆme Lacour*
Received 28th May 2009, Accepted 18th June 2009
First published as an Advance Article on the web 28th July 2009
DOI: 10.1039/b910475e
The addition of a Lewis acidic metal triflate salt Mg(OTf)2 as co-catalyst in the CpRu-catalyzed
decarboxylative allylation of in situ-generated ketone enolates allows the reaction to proceed at lower
temperature with higher regio- and enantioselectivity. Even so-far-unreactive substrates react.
Table 1 Co-catalyst influence on the CpRu-catalyzed rearrangement of
Introduction
3aa
Catalysts for enantioselective addition of nucleophiles are tra-
ditionally designed to activate the electrophilic reaction partner
through Lewis or Brønsted acid activation.1 Recently, follow-
ing the work of Shibazaki,2 bifunctional catalysts3 or mixtures
of catalysts4 have been successfully developed to achieve the
dual activation of both nucleophilic and electrophilic reaction
partners.5
The catalytic in situ generation of nucleophiles and their
Lewis
Entry Ester acid
participation in the C–C bond-forming processes is also a
topic of sustained interest.6 In this context, the fragmentation
of allyl b-ketoesters and the subsequent condensation of the
resulting unstabilized carbonyl nucleophiles and unsymmetri-
cal allyl–metal fragments still represents a challenge to gen-
erate enantio- and regioselectively nonracemic g,d-unsaturated
Time/hb Conv. (%)c
4a : 5ad
ee (%)e
1
2
3
4
5
6
7
3a
3a
3a
3a
3a
3a
3a
—
LiPF6
KPF6
48
48
48
<3
<3
<3
>97
80
—
—
—
97 : 3
94 : 6
89 : 11
—
—
—
—
84
80
84
—
Mg(OTf)2 24
Zn(OTf)2 54
Cu(OTf)2 54
6
ketones.7 Recently, combinations of a CpRu source (e.g. [CpRu(h -
95
C10H8)][PF6] 1)8 and chiral diimine ligands7f,g were shown to
catalyze this type of reaction with high selectivities, but with
somewhat unsatisfactory reactivity; pyridine mono-oxazolines of
type 2 (Table 1) being nevertheless better ligands than pyridine
imines.9
TfOHf
48
<313
a 1 (2 mol%), ligand 2 (2.4 mol%), THF, 25 ◦C, c(3a) 1 M; the results being
the average of at least two runs. b Reaction time without 1 h induction time
after which is added the substrate followed by the Lewis acid. c Determined
by 1H NMR (400 MHz). d Determined by GC-MS. e Determined by CSP-
GC. 4a is always levorotatory. f 2 mol% were used.
It is this lack of reactivity that we decided to tackle. Mechanisti-
cally, it is considered that, for CpRu allyl complexes, the ionization
of the substrate to form a p-allyl metal complex, separated from the
b-ketocarboxylate moiety, is the rate-determining step (Scheme 1,
grey external cycle).10 So, any process that would render the
b-ketocarboxylate a better leaving group would be beneficial. As
1,3-dicarbonyl moieties are effective bidentate ligands for many
metal ions,11 the addition of a catalytic amount of an additional
Lewis acid was considered.12 It could ease the ionization step,
facilitate the decarboxylation and even possibly organize the
nucleophilic attack. Herein, we report that Mg(OTf)2 is indeed
an effective co-catalyst for the ‘Carroll rearrangement’ affording
lower temperature conditions and overall higher reactivity and
selectivity.
Scheme 1 Mechanistic rationale for a dual-catalysis approach.
Results and discussion
De´partement de Chimie Organique, Universite´ de Gene`ve, Quai Ernest
Ansermet 30, 1211 Gene`ve 4, Switzerland. E-mail: jerome.lacour@unige.ch;
Fax: (+41) 22 379 32 15
Lewis acid selection
† Electronic supplementary information (ESI) available: 1H NMR spectra
of all substrates, ligand and catalyst precursor, and typical experimental
and CSP separation procedures. See DOI: 10.1039/b910475e
To assess the feasibility of this dual-catalysis approach, cinnamyl
ester 3a (Ar = 4-MeO-C6H4) was treated with a preformed
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4057–4061 | 4057
©

View Article Online
Table 2 Optimal stoichiometry determination for the reaction of 3aa
Entry
1
2
[Mg]b
Temp/◦C
Time/hc
Conv. (%)d
4a : 5ae
ee (%)f
1
2
3
4
5
6
7
8
9
2
2
—
2
2
2
2
2
2
2.4
2.4
2.4
—
2.4
2.4
2.4
3.4
4.4
—
—
1
1
1
2
3
1
1
60
25
25
25
25
25
25
25
25
6
48
48
24
24
24
24
24
24
>97
<3
<3
99 : 1
—
—
79
—
—
—
84
83
78
82
82
<3
—
>97
>97
>97
>97
>97
97 : 3
94 : 6
90 : 10
96 : 4
96 : 4
a General conditions: THF, c(3a) 1 M; 1 and 2 in mol%; the results being the average of at least two runs. b Mg(OTf)2, mol%. c Reaction time after the
addition of 3a and Mg(OTf)2 (not including the 1 h mixing time of 1 and 2 at 60 ◦C). d Determined by 1H NMR (400 MHz). e Determined by GC-MS.
f Determined by CSP-GC. 4a is always levorotatory.16
Table 3 Stoichiometry determination for the reaction of 3ba
Entry
1 (mol%)
2 (mol%)
Mg(OTf)2 (mol%)
T/◦C
Time/hb
Conv. (%)c
4b : 5bd
ee (%)e
drd,f
1
2
3
4
5
6
7
8
9
2
2
—
2
2
2
2
2
2
2.4
2.4
2.4
—
2.4
2.4
2.4
3.4
4.4
—
—
1
1
1
2
3
1
1
60
25
25
25
25
25
25
25
25
9
48
48
48
24
24
24
24
24
>97
<3
<3
86 : 14
—
—
77,77
—
—
64 : 36
—
—
<3
—
—
—
>97
>97
>97
>97
>97
81 : 19
72 : 28
70 : 30
85 : 15
91 : 09
87,87
84,82
84,81
86,86
87,88
58 : 42
57 : 43
56 : 44
59 : 41
61 : 39
a General conditions: THF, c(3b) 1 M; the results being the average of at least two runs. b Reaction time after the addition of 3b and Mg(OTf)2 (not
including the 1 h mixing time of 1 and 2 at 60 ◦C). c Determined by 1H NMR (400 MHz). d Determined by GC-MS. e Enantiomeric excess for the syn
and anti stereoisomers of 4b; determined by CSP-GC. No peak separation was achieved for linear 5b. f Ratio between the syn and anti stereoisomers of
product 4b.
combination of CpRu 1 and ligand 2. At room temperature, no
conversion was obtained without a co-catalyst (Table 1, entry 1).
The use of additional cationic alkali metal sources as LiPF6 or
KPF6 (entries 2 and 3) provided no reaction.
On the other hand, Mg(OTf)2 afforded a clean reaction at room
temperature (entry 4). The products were obtained with a higher
enantiomeric excess but a slightly lower regioselectivity (4a ee 84%,
4a : 5a 97 : 3) than under the classical reaction conditions (60 ◦C,
no co-catalyst, ee 79%, 4a : 5a 99 : 1, Table 2, entry 1). Zn(OTf)2
and Cu(OTf)2 (entries 5 and 6) were also tested. In both cases, the
reactions were noticeably slower (incomplete reaction after 54 h)
and the regio- and/or enantioselectivities were lower. Mg(OTf)2
salt then appeared as the best candidate for further screening.14
provided as references (Tables 2 and 3, entry 1). The results of
the two assays point towards the same conclusions. At 25 ◦C, all
reagents are clearly indispensable to promote the desired reaction
of 3a or 3b: no reactivity is observed with any lack of 1, 2 or
Mg(OTf)2 (entries 2 to 4, both tables). Then, when all three
reagents are present in a 2 : 2.4 : 1 ratio, better ◦enantioselectivities
are obtained at room temperature than at 60 C (Tables 2 and 3
entries 5, 84 and 87% ee for 4a and 4b respectively). Increasing the
amount of Mg(OTf)2 from 1 to 3 mol% has a negative influence
both on the regioselectivity (branched-to-linear ratio, 4 : 5) and
on the enantioselectivity of the reaction (entries 6 to 7). Since
the catalytically active Ru species can only bear a single ligand 2,
the use of an overstoichiometric amount of ligand might lead to
coordination of the free ligand onto to the additional metal salt
and thus modify its Lewis acidity. The influence of the ligand
stoichiometry was thus assessed by increasing the amount of
ligand 2 from 2.4 to 4.4 mol%. Little influence on the reactions
of 3a was observed but results were nicely improved with 3b—
the regioselectivity of the reaction in particular. Unfortunately,
no positive effect was observed on the syn-to-anti ratio for the
diastereoisomers of 4b (Table 3, entries 8 and 9).15
Stochiometry determination
The next step necessary was the determination of the optimal
three-component combination of CpRu 1, ligand 2 and Lewis
acid Mg(OTf)2. Screening was performed with two rather different
substrates 3a and 3b (Tables 2 and 3 respectively). The reactions
◦
of these compounds at 60 C in the absence of a Lewis acid are
4058 | Org. Biomol. Chem., 2009, 7, 4057–4061
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 4 Substrate scope of CpRu-catalyzed rearrangements
Overall, looking at the results of the reactions of 3a–3e, better
enantioselectivity is afforded for the branched products of type 4
under dual-catalysis conditions. We wondered about the generality
of this observation and about the origin of this increase in
enantioselectivity; the better selectivity being possibly a result of
the lowering of the reaction temperature only. Substrate 3h was
therefore selected. This allylic benzoylester is more reactive and
known to react already at 25 ◦C without any Lewis acid additive.9
Significantly, it was observed that the presence of Mg(OTf)2 lowers
the enantioselectivity of the reaction of 3h quite substantially (ee
80 vs. 86%, entry 6, Table 4). Clearly, temperature is not the
only parameter to be considered: the higher reactivity with the
co-catalyst being, at the same temperature, detrimental to the
selectivity.
Entry Ester
1
Time/ha Conv. (%)b 4 : 5c
ee (%)d
3c 24e
>97
>97
>97
98 : 2 83
2
3
3d 24e
3e 24e
93 : 7 82
92 : 8 nd, 88 (57 : 43)f
4
3f 24g
3f 24h
<3
>97
—
—
Mechanistic insight
70 : 30 nd
With these results in hand, preliminary experiments were per-
formed to try to understand the role of the Lewis acid during the
reaction. First, a stoichiometric amount of Mg(OTf)2 was added
to allylic ester 3a and no change was observed in the attenuated
total reflection (ATR)-IR spectrum (nC=O = 1714, 1738 cm-1). This
lack of perturbation suggested an absence of Lewis interaction
between the metal and the substrate at the start of the reaction
and a role for the Mg salt at a later stage of the mechanism.
This assumption was tested with cinnamyl isopropenyl carbonate
6. Importantly, no reaction was obtained under dual-catalysis
reaction conditions whereas a slow but detectable reaction was
obtained under classical conditions (Scheme 2, <15% conv. after
24 h). This result clearly shows the importance of the ketoester
moiety in the activation process despite the lack of evidence in
ATR-IR spectroscopy for its interaction with Mg(OTf)2.
5
6
3g 24g
3g 24h
<3
40
—
—
9 : 91 nd
3h 64i
3h 24e
>97
>97
95 : 5 86
93 : 7 80
a Reaction time without the 1 h induction time after which the substrate and
Mg(OTf)2 are added. b Determined by 1H NMR (400 MHz). c Branched-
to-linear ratio determined by GC-MS. d Determined by CSP-GC or CSP-
HPLC. Products 4c, 4d, 4f and 4h being levorotatory.16 e 1 (2 mol%), 2
(4.4 mol%), Mg(OTf)2 (1 mol%), THF, 24 ◦C, 3: c 1 M. f Diastereomeric
ratio given in brackets. g Classical conditions: 1 (10 mol%), 2 (13 mol%),
THF, 60 ◦C, 3: c 1 M. h Above classical conditions with Mg(OTf)2 (5 mol%)
extra. i 1 (2 mol%), 2 (2.4 mol%), THF, 25 ◦C, 3h: c 2 M.
Substrate scope
With the resulting 2 : 4.4 : 1 stoichiometry of 1 : 2 : Mg(OTf)2 in
hand, the scope of the reaction was studied (Table 4). Two rather
classical substrates were first used, compounds 3c and 3d bearing
different substituents on the aromatic nucleus of the allyl moiety.
As expected, these compounds reacted at 25 ◦C under the new set
of conditions to afford the branched regioisomer with excellent
regioselectivity and ee values of 83 and 82% for 4c and 4d, results
Scheme 2 Reaction of cinnamyl isopropenyl carbonate 6 under dual–
catalysis (top) and classical (bottom) conditions.
◦
better than those at 60 C (ee 77% in both cases).9 In the case
of substrate 3e similar results to 3b were, not too surprisingly,
obtained due to their structural similarities (entry 3).9
Maybe more importantly, experiments with substrates 3f and
3g were also performed. These compounds, which did not afford
any conversion under previous (non co-catalized) conditions at
60 ◦C, now reacted in the presence of Mg(OTf)2 as co-catalyst. For
instance 3f yielded, at 60 ◦C, the corresponding branched product
4f with decent regioselectivity; no conditions being found for the
analytical separation of the enantiomers and the measurement
of the enantiomeric purity. The previously unreactive aliphatic
3g also provided a reaction. The linear product 5g was obtained
predominantly in the crude reaction mixture (4g : 5g, 9 : 91); this
result, however, being in line with previous observations from the
groups of Bruneau and Pregosin with Cp*Ru catalysts.10
Scheme 3 Use of preformed chiral Lewis acidic co-catalyst.
This assertion was verified by looking at the reactivity of chiral
secondary allylic ester 717 (Table 5). Traditionally, this type of
substrate reacts faster than the corresponding primary linear
adducts of type 3 in allylation reactions mediated by CpRu or
Cp*Ru moieties.18,7b,f This higher reactivity is considered to be
linked to a better accessibility of the metal fragment to the
monosubstituted olefin.
For the rearrangement of enantiopure (S)-7 to occur at room
temperature, as for the linear substrates, the use of the additional
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4057–4061 | 4059
©

View Article Online
Table 5 Enantiospecific CpRu-catalyzed rearrangement of 7a
Entry
L
Lewis acid
Conv. (%)c
4d : 5dd
Diallylated (%)d
ee (%)e
1
2
3
4
5
6
7
2
—
<3
>97
>97
>97
>97
>97
>97
—
—
5
7
7
9
10
16
—
ent-2
ent-2
ent-2
2
Mg(OTf)2
Zn(OTf)2
Cu(OTf)2
Mg(OTf)2
Zn(OTf)2
Cu(OTf)2
93 : 7
91 : 9
71 : 29
89 : 11
86 : 14
51 : 49
(S)-93
(S)-95
(S)-97
(S)-40
(S)-50
(S)-70
2
2
a 1 (2 mol%), ligand (4.4 mol%), THF, 25 ◦C, c(7) 1 M. b Reaction time without the 1 h induction time after which is added the substrate followed by the
1
Lewis acid. c Determined by H NMR (400 MHz). d Sum of diallylated isomeric products determined by GC-MS. e Determined by CSP-GC; absolute
configuration deduced from the optical rotation sign.7b
Lewis acid was necessary (entry 1). Using ent-219 as ligand and
Mg(OTf)2 as Lewis acid, (S)-7 yielded the corresponding ketone
(S)-4d (global retention of configuration) with good regio- and
high enantioselectivity (entry 2, 4d : 5d, 93 : 7, ee 93%). But
contrary to what had been previously observed,7b,f a complex
mixture of diallylated regioisomeric products was obtained along
with 4d and 5d. Once again the use of Zn(OTf)2 or Cu(OTf)2
afforded lower regioselectivity but higher enantioselectivity was
interestingly obtained (entries 3 and 4, up to ee 97% for 4d
with Cu(OTf)2). Using 2 as ligand, a clear mismatched effect
was obtained in all cases with lower enantioselectivity but also
poorer regioselectivity and more diallylated products (entries 5
to 7). Since (S,+)-4d was obtained in all reactions, and this with
both enantiomers 2 and ent-2 of the ligand, it is clear that the
stereogenic information within the substrate (S)-7 remains overall
predominant. This tends to show that a possible isomerization
of the transient p-allyl complex is slow on the timescale of the
reaction. As such, the imperfect stereospecificity of the reaction
is most probably, as hypothesized by Tunge,7b due to the partial
isomerization of branched 7 into linear 3d resulting in a loss of
initial stereochemical information.
ligand (ent-2 and 2, see Table 4, entry 3 for comparison). Only
a slight increase in the branched-to-linear ratio was noticed. The
observed enantioselectivity is only due to the effect of ligand 2.
Conclusions
In conclusion, dual-catalysis with Mg(OTf)2 allows a wider scope
of substrates. The results show that the Lewis acidic metal cation
plays a role in the activation of one of the intermediates prior
to the decarboxylation—and the b-ketoacetate leaving group in
particular.20 In addition, the metal cation influences the nature of
the nucleophilic species since a clear effect on the regioselectivity
of the reaction is observed by varying the nature of the metal or its
ligands. Current studies are pursing the identification of effective
chiral Lewis acidic complexes able to act in synergy with a CpRu-
based catalyst.
Experimental
General remarks
Unless otherwise stated, solvents and chemicals were purchased
and used as received. Deuterated chloroform was filtered through
a plug of basic alumina prior to use. NMR spectra were recorded
On the possibility of using a chiral Lewis acid
1
Finally, in view of the probable role of Mg(OTf)2 as Lewis acid
activator of the ketoester functional group, we entertained the idea
of finding a chiral ligand that would be selective for the magnesium
ion in the presence of 1 and 2. It would generate, in situ, a chiral
Lewis acid specific for the nucleophilic fragment able to reinforce
possibly the enantio- and/or diastereoselectivity of the reactions.
In a very preliminary study, the influence of 2,6-bis((R)-4-phenyl-
4,5-dihydrooxazol-2-yl)pyridine 8 was tested (Scheme 3).
First, it was verified that 8 is not an efficient ligand in con-
junction with CpRu 1 for the Carroll rearrangement. Indeed, no
reactions occurred in the presence of 8 under classical conditions.
Then, various amounts of 8 (1 to 3 mol%) were premixed with
Mg(OTf)2 (1 mol%) and the mixtures were added to reactions of
3e. In all cases, the reactions performed well but essentially no
influence of 8 was found on the selectivity ratios (4e : 5e, 95 : 5,
dr, 55 : 45, 87% ee), and this with both enantiomers of the pymox
on Brucker ARX-400 and AMX-500 spectrometers. H NMR
chemical shifts are given in ppm relative to Me4Si with solvent
resonances used as internal standards. Data are reported as
follows: chemical shift (d) in ppm on the d scale, multiplicity (s =
singlet, d = doublet, t = triplet, h = septuplet, dd = doublet of
doublet, dt = doublet of triplet, and m = multiplet), coupling
constant (J) in Hz. Electrospray ionization mass spectra were
obtained on a Finnigan SSQ 7000 spectrometer at the Department
of Mass Spectroscopy of the University of Geneva. Optical
rotations were measured by optical rotary dispersion (ORD) on
a JASCO P-1030 polarimeter in a thermostated (20 ◦C) 10.0 cm-
long microcell with high-pressure sodium lamps. Determination
of the enantiomeric purity of compounds was achieved by chiral
stationary phase (CSP) high-performance liquid chromatography
(CSP-HPLC) on an Agilent LC-1100 HPLC equipped with a
binary pump, an autosampler, a column thermostat and a diode
4060 | Org. Biomol. Chem., 2009, 7, 4057–4061
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
array detector, or by CSP-GC on a Hewlett Packard 6890 GC
equipped with an autosampler and a flame ionization detector.
Determination of the branched-to-linear ratios was achieved by
GC-MS on a Agilent 6890 GC-MS using a capillary column HP-5.
ATR-IR spectra were recorded with a Perkin-Elmer 1650 FT-IR
spectrometer using a diamond ATR Golden Gate sampler.
b-Ketoesters 3a, 3c, 3d, 3f and 3g were prepared by DMAP-
catalyzed addition of the corresponding allylic alcohols to
diketene.21 b-Ketoesters 3b, 3e and 3h were prepared using a
transesterification procedure. A scheme showing the preparation
of these starting materials is included in the ESI.†
2 M. Shibasaki, H. Sasai and T. Arai, Angew. Chem., Int. Ed. Engl.,
1997, 36, 1236; M. Shibasaki, M. Kanai and K. Funabashi, Chem.
Commun., 2002, 1989; M. Shibasaki, S. Matsunaga and N. Kumagai,
Synlett, 2008, 1583.
3 J. A. Ma and D. Cahard, Angew. Chem., Int. Ed., 2004, 43, 4566; M.
Shibasaki and Y. Yamamoto, Multimetallic catalysts in organic syn-
thesis, Wiley-VCH, Weinheim, 2004; H. Yamamoto and K. Futatsugi,
Angew. Chem., Int. Ed., 2005, 44, 1924.
4 G. L. Hamilton, E. J. Kang, M. Mba and F. D. Toste, Science, 2007,
317, 496; M. Rueping, A. R. Antonchick and C. Brinkmann, Angew.
Chem., Int. Ed., 2007, 46, 6903.
5 R. Noyori and M. Kitamura, Angew. Chem., Int. Ed. Engl., 1991, 30,
49; M. Kanai, N. Kato, E. Ichikawa and M. Shibasaki, Synlett, 2005,
1491; T. Ikariya, K. Murata and R. Noyori, Org. Biomol. Chem., 2006,
4, 393.
Compounds 3a–3h and 7 have spectral characteristics identical
to those already described in the literature.7a,f
6 B. M. Trost, J. Org. Chem., 2004, 69, 5813; J. T. Mohr and B. M. Stoltz,
Chem.–Asian J., 2007, 2, 1476; Z. Lu and S. Ma, Angew. Chem., Int.
Ed., 2008, 47, 258 and references therein.
Cinnamyl prop-1-en-2-yl carbonate (6). Cinnamyl prop-1-
en-2-yl carbonate (6) was synthesised following a literature
procedure22 and obtained as a colorless oil. 1H NMR (500 MHz,
CDCl3) d: 7.50–7.30 (m, 5H), 6.80 (d, 3J(H–H) = 16 Hz, 1H), 6.40
7 (a) E. C. Burger and J. A. Tunge, Org. Lett., 2004, 6, 2603; (b) E. C.
Burger and J. A. Tunge, Chem. Commun., 2005, 2835; (c) T. Graening
and J. F. Hartwig, J. Am. Chem. Soc., 2005, 127, 17192; (d) H. He,
X. J. Zheng, U. Yi, L. X. Dai and S. L. You, Org. Lett., 2007, 9, 4339;
(e) D. J. Weix and J. F. Hartwig, J. Am. Chem. Soc., 2007, 129, 7720;
(f) S. Constant, S. Tortoioli, J. Muller and J. Lacour, Angew. Chem., Int.
Ed., 2007, 46, 2082; (g) S. Constant, S. Tortoioli, J. Muller, D. Linder,
F. Buron and J. Lacour, Angew. Chem., Int. Ed., 2007, 46, 8979.
8 E. P. Ku¨ndig and F. R. Monnier, Adv. Synth. Catal., 2004, 346, 901; A.
Labonne, L. Zani, L. Hintermann and C. Bolm, J. Org. Chem., 2007,
72, 5704.
3
3
2
(dt, J(H–H) = 16 Hz, J(H–H) = 6.5 Hz, 1H), 4.93 (d, J(H–
3
4
H) = 1.5 Hz, 1H), 4.88 (dd, J(H–H) = 7 Hz, J(H–H) = 1 Hz,
2H), 4.82–4.76 (m, 1H), 2.07 (d, 2J(H–H) = 1 Hz, 3 H). 13C NMR
(125 MHz, CDCl3) d: 153.0 (1C), 152.9 (1C), 136.0 (1C), 135.2
(1C), 128.7 (2C), 128.3 (1C), 126.8 (2C), 122.1 (1C), 102.0(1C),
68.7 (1C), 19.2(1C). MS (EI): 218 (1), 174 (3), 117 (100), 115 (25).
ATR-IR (neat): 1747, 1672, 1447, 1378, 1269, 1198 966, 931, 741,
691 cm-1. Low-resolution ESI-MS: 218 (1%), 117 (100%). Please
see the ESI for the GC-MS trace and 1H and 13C NMR spectra.†
9 D. Linder, F. Buron, S. Constant and J. Lacour, Eur. J. Org. Chem.,
2008, 5778.
10 For leading references on Cp*Ru-catalyzed allylic substitutions see: C.
Bruneau, J. L. Renaud and B. Demerseman, Chem.–Eur. J., 2006, 12,
5178; R. Hermatschweiler, I. Fernandez, P. S. Pregosin and F. Breher,
Organometallics, 2006, 25, 1440; B. M. Trost, P. L. Fraisse and Z. T.
Ball, Angew. Chem., Int. Ed., 2002, 41, 1059 and references therein.
11 C. Pettinari, F. Marchetti, A. Drozdov, J. A. McCleverty and T. J.
Meyer, in [beta]-Diketones and Related Ligands, Oxford, 2003,
p. 97.
CpRu-Catalyzed Carroll rearrangement—dual-catalysis procedure
In a typical procedure, in a 2 mL screw-cap vial equipped with
6
a magnetic stirring bar, [CpRu(h -naphthalene)][PF6] 1 (5.3 mg,
12 Acid Catalysis in Modern Organic Synthesis, ed. H. Yamamoto and
0.012 mmol, 2 mol%) and ligand 2 (8.1 mg, 0.033 mmol, 4.4 mol%)
were dissolved in 0.6 mL dry THF. The vial was flushed with
argon and capped. After heating for 1 h at 60 ◦C, the vial
was cooled to room temperature (~25 ◦C) and allyl b-ketoester
3a (150 mg, 0.6 mmol) was added in one portion followed by
Mg(OTf)2 (2.0 mg, 0.006 mmol, 1 mol%) and the vial was flushed
again with argon and stirred at room temperature for 24 h. The
cooled reaction mixture was diluted with 1.5 mL of ether–pentane
(60 : 40). After precipitation, the metal salts were filtered off on
a short SiO2 column (0.5 cm ¥ 4 cm, elution ether–pentane, 60 :
40); the solvents were then evaporated under reduced pressure to
afford the crude reaction mixture (4a + 5a) as a pale yellow oil
which was analysed (1H NMR, GC-MS, CSP-GC, ORD) without
further purification. Compounds 4a–4f, 5g and 4h have already
been reported in the literature and all general data collected fit the
previous description.7,23
K. Ishihara, Wiley-VCH, Weinheim, Germany, 2008.
13 This shows that the observed reactivity of the metal triflate salts is not
due to a possible Brønsted acid contamination.
14 No glove box manipulation of reagents needed.
15 These diastereoisomers result from the presence of a novel stereogenic
center a to the ketone carbonyl moiety.
16 Determined from the crude reaction mixture in CDCl3 (c ~ 1).
17 (S)-7 was prepared by olefination of commercially available enan-
tiopure styrene oxide using the protocol developed by Mioskowski
et al. and then subjected to an esterification with diketene (ee >
99%, CSP-HPLC). For the olefination, see: L. Alcaraz, J. J. Harnett,
C. Mioskowski, J. P. Martel, T. Legall, D. S. Shin and J. R. Falck,
Tetrahedron Lett., 1994, 35, 5449.
18 R. Hermatschweiler, I. Fernandez, F. Breher, P. S. Pregosin, L. F. Veiros
and M. J. Calhorda, Angew. Chem., Int. Ed., 2005, 44, 4397.
19 ent-2 and 2 are the ligands of (3S,8R) and (3R,8S) configuration
respectively.
20 For the use of Mg cations in the catalytic activation processes see: Y.
Kiyotsuka, H. P. Acharya, Y. Katayama, T. Hyodo and Y. Kobayashi,
Org. Lett., 2008, 10, 1719; R. W. Coscia and T. H. Lambert, J. Am.
Chem. Soc., 2009, 131, 2496 and references therein.
21 I. Collado, C. Pedregal, A. Mazon, J. F. Espinosa, J. Blanco-Urgoiti,
D. D. Schoepp, R. A. Wright, B. G. Johnson and A. E. Kingston,
J. Med. Chem., 2002, 45, 3619.
22 T. Nishimata, Y. Sato and M. Mori, J. Org. Chem., 2004, 69,
1837.
23 M. Fujii, K. Nakamura, S. Yasui, S. Oka and A. Ohno, Bull. Chem.
Soc. Jpn., 1987, 60, 2423; M. Morita, S. Sakaguchi and Y. Ishii, J. Org.
Chem., 2006, 71, 6285.
Notes and references
1 E. N. Jacobsen, A. Pfaltz and H. Yamamoto, Comprehensive asym-
metric catalysis, Springer, Berlin, New York, 1999; I. Ojima, Catalytic
asymmetric synthesis, Wiley-VCH, Weinheim, 2000; E. N. Jacobsen,
A. Pfaltz and H. Yamamoto, Comprehensive asymmetric catalysis.
Supplement, Springer, Berlin, New York, 2004.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4057–4061 | 4061
©