Molybdenum(0) and tungsten(0) catalysts with enhanced reactivity for allylic substitution: Regioselectivity and solvent effects

Article abstract of DOI:10.1039/b100903f

The binuclear Mo(II) and W(II) complexes 28a,b and 29a,b have been developed as pre-catalysts for allylic substitution with β-dicarbonyl nucleophiles. These complexes are reduced in situ to Mo(0) and W(0) catalytic species 30a,b and 31a,b by excess of NaH, employed to generate sodiomalonate nucleophiles, or by DIBAL-H. 1,3-Dioxolane and 1,4-dioxane, when used as solvents, substantially accelerate the reaction. These new catalysts exhibit "traditional" Mo regiochemistry, i.e., the nucleophilic attack occurring preferentially at the more substituted carbon (5 → 9; 37 → 38), unless an additional factor, such as further coordination to another moiety of the allylic electrophile takes part (41), as in the case of the geranyl-type substrates (32 or 33 → 36).

Full text of DOI:10.1039/b100903f

View Article Online / Journal Homepage / Table of Contents for this issue
Molybdenum(0) and tungsten(0) catalysts with enhanced reactivity
for allylic substitution: regioselectivity and solvent effects †
Andrei V. Malkov,‡a,b Ian R. Baxendale,§b Darren J. Mansfield¶c and Pavel Kocˇovský*‡a,b
1
a Department of Chemistry, University of Glasgow, Glasgow, UK G12 8QQ.
E-mail: P.Kocovsky@chem.gla.ac.uk
b Department of Chemistry, University of Leicester, Leicester, UK LE1 7RH
c Aventis CropScience UK Ltd, Chesterford Park, Saffron Walden, UK CB10 1XL
Received (in Cambridge, UK) 24th January 2001, Accepted 19th March 2001
First published as an Advance Article on the web 26th April 2001
The binuclear Mo() and W() complexes 28a,b and 29a,b have been developed as pre-catalysts for allylic
substitution with β-dicarbonyl nucleophiles. These complexes are reduced in situ to Mo(0) and W(0) catalytic
species 30a,b and 31a,b by excess of NaH, employed to generate sodiomalonate nucleophiles, or by DIBAL-H.
1,3-Dioxolane and 1,4-dioxane, when used as solvents, substantially accelerate the reaction. These new catalysts
exhibit “traditional” Mo regiochemistry, i.e., the nucleophilic attack occurring preferentially at the more substituted
carbon (5 → 9; 37 → 38), unless an additional factor, such as further coordination to another moiety of the
allylic electrophile takes part (41), as in the case of the geranyl-type substrates (32 or 33 → 36).
esters, and the electron-rich aromatics and heteroaromatics.8
Introduction
However, in all these reactions Mo() and W() proved to
Palladium(0)-catalysed allylic substitution is a well established
behave purely as mild Lewis acids rather than serving as metal
synthetic method with numerous applications both in academia
templates, so that the original stereochemical information
and industry (Scheme 1).1 As a potential replacement for the
present in the substrate was usually lost and asymmetric
induction would hardly be feasible.8e,f Therefore, we turned our
attention back to Mo(0) complexes and, herein, we report
on our endeavour to enhance their reactivity.
Results and discussion
Scheme 1
The main difference between Pd(0)- and Mo(0)-catalysed allylic
expensive palladium, Group 6 metals (Mo and W) have also
substitution reactions is their regioselectivity toward the β-
been shown to effectively catalyse this reaction,2,3 although their
dicarbonyl-derived nucleophiles, such as NaCH(CO2Me)2:1,2
mode of action may differ from that of Pd.4,5 However, in spite
of the suitable cost, Mo and W are stained with generally lower
while Pd(0) mainly gives the products of substitution at the less
substituted terminus of the intermediate η3-complex 2,1,9
reactivity than Pd. Thus, while the Pd(0)-catalysed reactions
Mo(CO)6 and other Mo(0) and W(0) complexes preferentially
occur in THF at reflux or even at ambient temperature,1 Mo
afford their more substituted isomers, unless the nucleophile
and W catalysts typically require reflux in higher-boiling
is too bulky.2,3 Thus, cinnamyl acetate 5 is converted into the
solvents (e.g., toluene) for several hours.2–4 This striking differ-
“linear” product 12a in the Pd(0)-catalysed reaction (Scheme
ence can be attributed, in part, to the ease of ligand dissoci-
2; Table 1, entry 1),1g whereas W(0) preferentially gives the
ation in the case of Pd catalysts that creates vacancy in the
coordination sphere, e.g., (Ph3P)4Pd
(Ph3P)3Pd + Ph3P. By
contrast, Mo(CO)6, W(CO)6, and their congeners are relatively
stable, so that heating at higher temperature is required.6 Hence,
increasing the reactivity of Mo and W complexes would be
highly desirable.7
The latter goal may seem to have been met by Mo() and
W() complexes with weakly coordinating ligands (TfOϪ, labile
CO, or MeCN), for which we have demonstrated catalytic
allylic substitution at room temperature and a broader spec-
trum of nucleophiles than that for Pd(0)-catalysts, including
simple silyl enol ethers derived from ketones, aldehydes and
† Electronic supplementary information (ESI) available: experimental
procedures for compounds not covered in the Experimental. See http://
www.rsc.org/suppdata/p1/b1/b100903f/
‡ Present address: University of Glasgow.
§ Present address: Chemistry Laboratory, University of Cambridge,
Lensfield Road, Cambridge, UK CB2 1EW
¶ Present address: Departement Synthese, Aventis CropScience SA,
69009 Lyon, France.
Scheme 2
1234
J. Chem. Soc., Perkin Trans. 1, 2001, 1234–1240
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b100903f

View Article Online
Table 1 Allylic substitution of 5 with 7a,b and 13 catalysed by Pd(0), Mo(0) or W(0), and Mo() complexes (Scheme 2)
Entry
Allylic substrate
Nucleophile
Catalyst
Solvent
Products
Ratio
1
2
3
4
5
6
7
5
5
5
5
5
5
5
7a
7a
7b
13
13
7a
7b
(dba)3Pd2
(bipy)W(CO)3(MeCN)
(bipy)W(CO)3(MeCN)
THF
9a + 12aa
9a + 12ab
9b + 12bb
15 + 17c
15 + 17c
9a + 12a
9b + 12b
≤1 : 99
98 : 2
98 : 2
20 : 80
25 : 75
92 : 8
Toluene
Toluene
CH2Cl2
CH2Cl2
Dioxane
Dioxane
22
25
28bd
28bd
80 : 20
a Ref. 1g. b Ref. 3a. c Ref. 8e. d With NaH employed as the base.
“branched” isomer 9a (Table 1, entries 2 and 3).2 || On the other
hand, we have recently shown that the Lewis-acidic Mo() and
W() complexes, such as 22 and 25, whose preparation from
Mo(CO)6 is outlined in Scheme 3, tend to exhibit reactivity simi-
lar to that of Pd(0) in most cases (Table 1, entries 4 and 5).8
The latter observation was intriguing and certainly worth
further investigation. After much experimentation, it turned
out that the reaction did occur only when NaCH(CO2Me)2 was
generated (from dimethyl malonate) with a slight excess of
NaH; with larger excess or less then an equivalent of NaH, the
reaction did not occur. Moreover, the regiochemistry of the
substitution switched to that typical for Mo(0), i.e., the
“branched” isomer 9 became the major product when cinnamyl
acetate 5 was employed as the electrophilic partner (Table 1,
entries 6, 7). This behaviour clearly suggests that the initial
Mo() complex 28b was reduced in situ by the excess of NaH to
a reactive Mo(0) species, whose structure can be tentatively
formulated as 30b (L = ligand, such as the solvent). Since 28b is
present in a catalytic amount, only a small quantity of NaH
(presumably equivalent to the amount of 28b) is required to
generate the active species. On the other hand, if a substoi-
chiometric amount of NaH is used to generate NaCH-
(CO2Me)2, there is none left to reduce 28b, whereas its excess
apparently decomposes the catalyst, presumably by further
reduction of the CO groups. In view of a difficult accurate
dosage of very small quantities of NaH, we turned to DIBAL-
H as a substitute and, indeed, were also able to generate the
active species from the precatalyst 28b as reflected by the suc-
cessful reaction of 5 with NaCH(CO2Me)2 and NaCMe-
(CO2Me)2. However, this issue had not been pursued further
since the practicality of handling NaH in the amounts of 50–
100 mg was acceptable. The remaining complexes 28a, 29a, and
29bwerethenfoundtoexhibitbehavioursimilartothatof28b.**
Reactivity of the Mo and W complexes
With the initial results at hand, we set out to probe the catalytic
efficiency of the new Mo and W complexes and their regio-
selectivity in allylic substitution. To this end, we employed
the monoterpene-derived allylic acetates 32 and 33 and the
malonate-type nucleophiles 7a and 7b of different steric bulk.
The reactions were carried out by adding the catalyst (typically
10–20 mol%) and allylic substrate to a preformed solution of
dimethyl sodiomalonate (generated from dimethyl malonate
and NaH) in toluene, followed by a reflux for 24 h. The results
are summarized in Scheme 4 and Table 2. It is pertinent to note
that while the reaction of 33 with 7a, carried out in the presence
of Pd(0), is known to preferentially afford the terminal isomer
36 (Table 2, entry 1),2 Mo(CO)6-catalysed reactions produce its
isomer 34, regardless of whether linalyl or geranyl acetate (32 or
33) is used as starting material (Table 2, entries 2 and 3).2 Inter-
estingly, complex (bipy)Mo(CO)3(MeCN) has been reported to
resemble Pd(0), giving mainly the terminal product 36 (Table 2,
entry 4).2g
Scheme 3
Development of more reactive Mo(0) and W(0) catalysts
We have previously shown that the Lewis-acidic Mo() com-
plexes are deactivated by enolate-type nucleophiles.8 Thus,
while 22 and 23 readily facilitate the reaction of a variety of
silyl enol ethers with allylic substrates, no reaction occurred
with the corresponding enolates. The binuclear complex 25,
prepared in two steps from the commercially available Mo(CO)6
via ligand exchange with MeCN (18 → 24), followed by an
oxidative addition of SnCl4 (24 → 25),8,10 behaved in the same
way (Table 1, entries 4, 5).8e
In order to further explore the reactivity of binuclear cata-
lysts, Mo and W bipyridyl and phenanthroline complexes 28a,b
and 29a,b were prepared by the oxidative addition of SnCl4 to
26a,b and 27a,b, respectively, which in turn were obtained by
heating M(CO)6 (18 and 19, respectively) with the respective
heterocycles. With complexes 28a,b and 29a,b, enolates could
be expected to have the same deactivating effect which, initially,
seemed to be the case as no reaction of 4 or 5 with NaCH-
(CO2Me)2 was observed in the presence of 25. However, with
28b in toluene and the temperature being raised to reflux, the
reaction did slowly occur on some occasions.
Our bipyridyl Mo and W complexes 28a and 29a turned out
to give almost identical results to those reported for 26a (Table
2, entries 5–8). The regioselectivity in favour of terminal substi-
tution was further increased for the sterically more hindered
methylmalonate 7b, with barely any internal product detected
(Table 2, entries 9–12). However, in contrast to Pd(0), the Mo-
|| By contrast, with the bulkier NaC(Me)(CO2Me)2, the Mo(0)-catalysed
reaction preferentially gives the “linear” product 11b.
** Attempted isolation of the reactive complexes 30a,b and 31a,b, gener-
ated in situ from the precatalysts 28a,b and 29a,b, was unsuccessful.
J. Chem. Soc., Perkin Trans. 1, 2001, 1234–1240
1235

View Article Online
Table 2 Reactions of linalyl and geranyl acetates 32 and 33 with malonates 7a,b (Scheme 4)a
Entry
Allylic substrate
Nucleophile
Catalyst
34b
35b
36b
Yield (%)c
1
2
3
4
5
6
7
8
9
33
32
33
33
32
32
33
33
32
32
33
33
7a
7a
7a
7a
7a
7a
7a
7a
7b
7b
7b
7b
Pdd
18e
13
85
85
16
16
15
18
17
1
0
3
3
87
12
12
62
61
63
59
60
77
77
83
82
84g
80h
65h
45h
59
59
58
56
76
18e
[Mo]f
28a
29a
28a
29a
28a
29a
28a
29a
22
23
22
23
23
22
23
16
17
10
11
12
—
1
1
68
25
25
a The reactions were carried out in toluene at reflux for 24 h with 20 mol% of the catalyst unless stated otherwise. b The isomer ratios were determined
by GC of the crude mixtures. c Isolated yield of the mixture. d (Ph3P)4Pd (5 mol%). e 10 mol%. f (bipy)Mo(CO)3(MeCN), 24 h. g Ref. 1. h Ref. 2.
philic attack, which can be attributed to the change in the
solvation (compare entries 5–8 in Table 2 with entries 1–4 in
Table 3).‡‡ Increasing the content of THF in toluene to >30%
had a detrimental effect on the reactivity. Thus, with a 2 : 1
toluene–THF mixture, only 12% conversion was observed
under otherwise identical conditions (reflux for 8 h), which may
be, in part, attributed to the lowering of the boiling point.§§
After the modest success with toluene–THF mixtures, we
searched for solvents with similar relative permittivities (ε) and
boiling points; suitable candidates were identified in 1,3-
dioxolane (C) and 1,4-dioxane (D).¶¶ Using our standard probe
(Scheme 4), we found that these solvents brought about a sub-
stantial acceleration of the reaction (Table 3, entries 5–13).
Thus, with the more reactive 32 and the less hindered malonate
7a, the reaction times could be shortened to 6 or even 3 hours,
and the regioselectivity was substantially improved in favour of
the “terminal” attack (Table 3, entries 5–11). The catalyst load
could be decreased from 20 to 10 mol% without any effect on
the reactivity (compare entries 5 vs. 6 and 8 in Table 3) and the
reaction proceeded acceptably well even with as little as 5 mol%
of the catalyst (Table 3, entries 7, 9–11). While 32 produced
∼2 : 1 E–Z mixtures (35–36), its allylic isomer 33 was much
more selective in favour of 36, demonstrating the conservation
of the original information present in the substrate (Table 3,
entries 10 and 11).12 Similar trends were observed for the less
reactive nucleophile 7b (Table 3, entries 12 and 13). Interest-
ingly, comparison of 1,3-dioxolane (C) with 1,4-dioxane (D)
shows that, with these coordinating solvents, the reaction tem-
perature plays a minor role: the difference in boiling point here
is 27 ЊC ¶¶ and yet the reactions were essentially unaffected. By
contrast, a switch from toluene to benzene (a difference of
30 ЊC in bp) as the solvent results in stopping the reaction
almost completely.2e
Scheme 4
and W-catalysed reactions exhibited a higher proportion of the
(Z)-isomer 35 (Table 2, entries 4–12), with the exception of
Mo(CO)6 (Table 2, entries 2 and 3). While the reactivity of 32
and 33 toward 7a did not dramatically differ, substantial differ-
ences were experienced with the more bulky nucleophile 7b. In
this case, good yields were obtained with the tertiary acetate 32
(Table 2, entries 9 and 10), whereas poor conversions were
attained with the inherently less reactive primary acetate 33
(Table 2, entries 11 and 12).
Solvent effects
In his early studies, Trost and Lautens demonstrated the sensi-
tivity of Mo(0)-catalysed reactions to solvents.2e Thus, DMF
and diglyme poisoned the catalyst, presumably via coordin-
ation, whereas toluene seemed to be the reasonable com-
promise. A standard procedure using the latter solvent requires
that the nucleophile be added to a suspension of NaH, followed
by a period of reflux, which generates a gelatinous mixture
of the solvent and the salt, to which the other ingredients are
added. This protocol was later superseded by the utilization
of enolates derived from the reaction of malonate with N,O-
bis(trimethylsilyl)acetamide (BSA),11 although this protocol
has been shown to give somewhat different product ratios and
stereochemistry to those of the corresponding enolates.2,11
Clearly, the problems associated with the sodiomalonate-type
nucleophiles are those of the non-homogeneity and viscosity of
the reaction mixture. To address this issue, we conducted a ser-
ies of experiments in toluene containing 20–25% of THF (Table
3, entries 1–4).†† This modification resulted in a three-fold
acceleration as shown by quantitative conversions of 32 with
either 7a or 7b within 8 h (vs. 24 h in pure toluene). The product
ratios have been slightly altered in favour of the terminal nucleo-
Regioselectivity
To further address the regioselectivity issue (vide supra), we
employed cinnamyl-type substrates 4 and 5 (Scheme 2, Table 4).
In all cases, 4 exhibited a strong preference for attack at the
allylic terminus proximal to the phenyl ring, regardless of
whether malonate 7a (Table 4, entries 1–9) or the bulkier meth-
ylmalonate 7b (entries 10–17) was employed as the nucleophile.
†† Other combinations of the two solvents proved inferior, often result-
ing in very little reaction or extended reaction times.
‡‡ For computation of solvation effects in Pd(0)-catalysed allylic substi-
tution, see ref. 9.
§§ Using a larger excess of the nucleophile had little effect on the overall
conversion.
¶¶ Toluene: bp 110 ЊC, ε = 2.379; THF: bp 66 ЊC, ε = 7.58; 1,4-dioxane:
bp 102–103 ЊC, ε = 2.209; 1,3-dioxolane, bp 74–75 ЊC. The value for ε of
the latter solvent can be approximated by the value for CH3CH(OEt)2
(ε = 3.80). The data have been taken from ref. 20.
1236
J. Chem. Soc., Perkin Trans. 1, 2001, 1234–1240

View Article Online
Table 3 Solvent effects in the reactions of linalyl and geranyl acetates 32 and 33 with malonates 7a,b (Scheme 4)a
Entry
Allylic substrate
Nucleophile
Catalyst (mol%)
Solventb
Time/h
34c
35c
36c
Yield (%)d
1
2
3
4
5
6
7
8
9
10
11
12
13
32
32
32
32
32
32
32
32
32
33
33
32
32
7a
7a
7b
7b
7a
7a
7a
7a
7a
7a
7a
7b
7b
28a (20)
29a (20)
28a (20)
29a (20)
28a (20)
28a (10)
28a (5)
28a (10)
28a (5)
28a (5)
28a (5)
A
A
B
B
C
C
C
D
D
C
D
C
D
8
8
8
8
5
5
9
11
3
2
2
4
3
—
1
30
31
17
16
24
24
29
27
30
1
65
64
74
73
73
73
69
69
67
99
90
80
75
60
60
62
62
75
75
47
70
50
51
50
92
88
6
6
4e
3
3e
4e
3e
5
9
11
14
28a (20)
28a (20)
9
11
5
a The reactions were carried out at reflux (see the General procedure). b A = toluene–THF (3 : 1); B = toluene–THF (4 : 1); C = 1,3-dioxolane;
D = 1,4-dioxane. c The isomer ratios were determined by GC of the crude mixtures. d Isolated yield of the mixture. e No further reaction was observed
on prolonged time.
Table 4 The reactions of acyclic acetates 4–6 with malonates 7a,b (Scheme 2)a
Entry
Allylic substrate
Nucleophile
Catalyst (mol%)
Solventb
Time/h
Products
Ratioc
Yield (%)d
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
4
7a
7a
7a
7a
7a
7a
7a
7a
7a
7b
7b
7b
7b
7b
7b
7b
7b
7a
7a
7a
7a
7b
7b
7b
7b
7a
7a
7a
7a
28a (20)
28a (20)e
28a (20)
28b (20)
28b (20)
29a (20)
29a (20)
29b (20)
29b (20)
28a (20)e
28a (20)
28b (20)
28b (20)
29a (20)
29a (20)e
29b (20)
29b (20)
28a (20)
28b (10)
28b (10)
29b (10)
28a (10)
28b (10)
28b (10)
29a (10)
28a (10)
28b (10)
29a (10)
29b (10)
C
C
D
C
D
C
D
C
D
C
D
C
D
C
D
C
D
D
C
D
C
D
C
D
D
C
C
C
C
24
4
8a + 11a
8a + 11a
8a + 11a
8a + 11a
8a + 11a
8a + 11a
8a + 11a
8a + 11a
8a + 11a
8b + 11b
8b + 11b
8b + 11b
8b + 11b
8b + 11b
8b + 11b
8b + 11b
8b + 11b
8a + 11af
9a + 12a
9a + 12a
9a + 12a
9b + 12b
9b + 12b
9b + 12b
9b + 12b
10
83 : 17
84 : 16
82 : 18
89 : 11
90 : 10
83 : 17
90 : 10
89 : 11
90 : 10
88 : 12
88 : 12
94 : 6
93 : 7
88 : 12
89 : 11
93 : 7
93 : 7
83 : 17
93 : 7
92 : 8
95 : 5
76 : 24
80 : 20
80 : 20
76 : 24
—
81
84
82
89
94
89
81
80
84
89
85
90
98
87
88
85
87
78
68
74
79
75
87
88
61
95
94
93
91
4
4
1.5
1.5
1.5
24
1.5
1.5
1.5
4
1.5
1.5
1.5
1.5
4
1.5
1.5
1.5
7
7
7
7
7
7
7
0.5
0.5
0.5
0.5
4
4
4
4
4
4
4
4
4
4
4
4
4
4
(R)-4
5
5
5
5
5
5
5
6
6
6
6
10
—
—
—
10
10
a The reactions were carried out at reflux (see the General procedure). b C = 1,3-dioxolane; D = 1,4-dioxane. c The isomer ratios were determined by
GC of the crude mixtures. d Isolated yield of the mixture. e The catalyst recrystallised prior to use. f Both products were racemic.
Cinnamyl acetate 5 followed essentially the same trend but
was slightly more discriminating between 7a and 7b (compare
entries 19–21 with 22–25 in Table 4). All the complexes
28a,b and 29a,b exhibited similar reactivity; differences in the
reaction times required for completion were, apparently, mainly
associated with the reaction temperature (i.e., with the boil-
ing point of the solvent—vide supra). However, the purity of
the pre-catalyst proved to be crucial. Thus, recrystallised 28a
was much more reactive than its crude counterpart (compare
entries 1 vs. 2 and 6 vs. 7 in Table 4). The bisphenyl substrate 6
proved more reactive than the other cinnamyl derivatives, being
almost quantitatively converted into 10 in 30 min (Table 4,
entries 26–29).
Trost has shown that attack at the more substituted terminus
is characteristic for Mo(0) catalysts2 (as in the case of 5), our
monoterpene models seem to be out of line, raising the follow-
ing questions. (1) Is it the steric hindrance exercised by the
additional methyl group in 32 and 33 that swings the reaction
toward the primary (vs. tertiary) terminus, or is it the electronic
effect that plays the decisive role in the cinnamyl model 5?
(2) Does the additional double bond at the far end of the
molecule of the terpenes 32 and 33 play any role, e.g., by chelat-
ing the metal?13 To address these issues, we employed the
truncated prenyl-type acetate 37 as a mimic of linalyl acetate
32 lacking the additional double bond (Scheme 5). On reac-
tion with NaCH(CO2Me)2, carried out in the presence of
28a, acetate 37 produced a 50 : 50 mixture of the regio-
isomers 38 and 39 (55%), which represents a dramatic enhance-
ment of the internal product as compared to the geranyl system
(Tables 2 and 3). Similarly, NaC(Me)(CO2Me)2 also produced
The striking difference in regioselectivity observed between
the monoterpenes 32 or 33 (attack at the less substituted
terminus) and the cinnamyl model 5 (attack at the more sub-
stituted terminus of the allylic system) is intriguing. While
J. Chem. Soc., Perkin Trans. 1, 2001, 1234–1240
1237

View Article Online
inlet and the lowest temperature enabling evaporation. All
products were dried under high vacuum before the recording
the yield; the term yield refers to isolated product(s) showing a
single spot on TLC. The identity of the known compounds
prepared by different methods was checked by comparison of
their NMR, IR, and MS and by their TLC behaviour. All solv-
ents for the reactions were of reagent grade and were dried and
distilled under nitrogen immediately prior to use as follows:
tetrahydrofuran (THF) from sodium–benzophenone; dichloro-
methane from calcium hydride (40 Mesh); diethyl ether was
provisionally dried over sodium wire followed by distillation
from lithium aluminium hydride; acetonitrile was dried by
stirring over phosphorus pentaoxide (10% by weight) for 24 h
followed by distillation onto potassium carbonate and a second
distillation to remove acidic impurities.
Scheme 5
substantially more of the internal isomer (38–39 = 34 : 66; 75%)
than its linalyl congener, although this shift was not as strong as
in the latter case, presumably owing to the inherent bias of the
more bulky nucleophile to seek the less hindered site of attack.
Thus, the above results strongly indicate that the marked
preference for terminal attack in the case of the geranyl system
is, indeed, associated with the chelation of the transient species
General procedure for the allylic substitution reactions catalysed
by complexes 28a–29b
by the double bond.13 Since the η3–η1 (40
can be expected for the latter chelate (Scheme 6), the preferential
attack on the terminal carbon can be understood in terms of
the η1-species 41.
41) equilibrium
To a stirred suspension of sodium hydride (4 mmol) in 1,3-
dioxolane or 1,4-dioxane (6 mL) was added dropwise a solution
of the nucleophile (4 mmol) in the appropriate solvent (2 mL).
To the resulting clear solution was added catalyst (10–20 mol%)
followed by a solution of the allylic substrate (2 mmol) in the
corresponding solvent (2 mL). The reaction was heated at reflux
until TLC analysis indicated disappearance of the starting
material or until no further reaction was detected after 24 h.
The reaction mixture was diluted with ether (10 mL), absorbed
onto silica (∼2.5 g) and the product was purified by flash
chromatography. The amount of silica, the polarity of the
eluent, and the size of the column used for each separation were
varied with respect to the number and relative polarities of the
products. For details and yields see Tables 2–4.
Scheme 6
Dimethyl (1-phenylbut-2-en-1-yl)malonate 8a. Obtained as
a mixture with 11a; for experimental details see General
procedure and Table 4: H NMR δ 7.21–7.03 (5 H, m, arom),
Conclusions
In conclusion, we have demonstrated that the binuclear Mo()
and W() complexes 28a,b and 29a,b can serve as pre-catalysts
in allylic substitution. These complexes are reduced in situ to
Mo(0) and W(0) species by a slight excess of NaH (equivalent to
the amount of the catalyst), employed to deprotonate malonate
nucleophiles. 1,3-Dioxolane and 1,4-dioxane have been shown
to be the solvents of choice as they substantially accelerate the
reaction and render it homogeneous (in contrast to toluene,
the traditional solvent in this area). These complexes exhibit
“traditional” Mo regiochemistry, i.e., the nucleophilic attack
occurring preferentially at the more substituted carbon. The
exception to this rule, observed for the geranyl system, has
been rationalized via coordination of the metal to the distal
1
5.5–5.3 (2 H, m, CH᎐CH), 3.91 (1 H, dd, J = 11, 7 Hz, 1-H), 3.63
᎐
(1 H, d, J = 11 Hz, CH(CO2Me)2), 3.58 (3 H, s, OMe), 3.34 (3 H,
s, OMe), 1.53 (3 H, d, J = 6 Hz, 3-Me); 13C NMR δ 168.1
(C), 167.7 (C), 140.6 (C), 139.3 (CH), 128.3 (2 × CH), 127.6
(2 × CH), 127.4 (CH), 126.7 (CH), 57.7 (CH), 52.2 (CH3), 52.0
(CH3), 48.8 (CH), 17.7 (CH3), in accordance with the
literature.2,3,14
Dimethyl methyl(1-phenylbut-2-en-1-yl)malonate 8b. Obtained
as a mixture with 11b; for experimental details see General
1
procedure and Table 4: H NMR δ 7.38–7.15 (5 H, m, arom),
5.95 (1 H, dd, J = 15.8, 8 Hz, 2-H), 5.55 (1 H, dq, J = 15.8, 6 Hz,
3-H), 4.09 (1 H, d, J = 8 Hz, 1-H), 3.68 (3 H, s, OMe), 3.58 (3 H,
s, OMe), 1.67 (3 H, d, J = 6 Hz, 3-Me), 1.43 (3 H, s, Me);
13C NMR δ 171.4 (C), 171.2 (C), 139.8 (C), 129.4 (2 × CH),
129.2 (CH), 128.6 (CH), 128.0 (2 × CH), 126.8 (CH), 59.0 (C),
53.5 (CH), 52.2 (CH3), 52.1 (CH3), 18.1 (CH3), 18.0 (CH3), in
accordance with the literature.2,3,14
C᎐C bond in the substrates which leads to a distortion of the
᎐
complex geometry and, consequently, to the change of the
regiochemical outcome (41).
Experimental
General
Dimethyl (1-phenylprop-2-en-1-yl)malonate 9a. Obtained as a
mixture with 12a; for experimental details see General pro-
cedure and Table 4: 1H NMR δ 7.21–7.03 (5 H, m, arom), 5.90
(1 H, m, 2-H), 5.01 (1 H, d, Jtrans = 14 Hz, 3-Ha), 4.96 (1 H, d,
Jcis = 6 Hz, 3-Hb), 4.01 (1 H, dd, J = 11, 6 Hz, 1-H), 3.74 (1 H, d,
J = 11 Hz, CH(CO2Me)2), 3.60 (3 H, s, OMe), 3.35 (3 H, s,
OMe); 13C NMR δ 167.9 (C), 167.6 (C), 139.8 (C), 137.6 (CH),
128.4 (2 × CH), 127.9 (2 × CH), 126.9 (CH), 116.3 (CH2), 57.1
(CH), 52.3 (CH3), 52.1 (CH3), 49.5 (CH), in accordance with
the literature.2,3,14,15
Melting points were determined on a Kofler block and remain
uncorrected. Optical rotations were measured at 20 ЊC. The
NMR spectra were recorded in CDCl3 or DMSO-d6, 1H at 250
MHz and 13C at 62.9 MHz with reference to the solvent signals
of chloroform-d1 (δ 7.27, 1H; δ 77.0, 13C) and dimethyl
1
sulfoxide-d6 (δ 2.62, H; δ 39.7, 13C) as internal standards; the
suffixes “a” and “b” are used to distinguish between two dia-
stereotopic protons or equivalent groups. Coupling constants
were determined by first order approximation. Various 2D-
techniques and DEPT experiments were used to establish a
compound’s structure and to assign the signals. The IR spectra
were recorded for a thin film between KBr plates or using the
“Golden-Gate” technique. The mass spectra (EI, ES or CI)
were measured on a dual sector mass spectrometer using direct
Dimethyl methyl(1-phenylprop-2-en-1-yl)malonate 9b. Obtained
as a mixture with 12b; for experimental details see General
1
procedure and Table 4: H NMR δ 7.36–7.15 (5 H, m, arom),
1238
J. Chem. Soc., Perkin Trans. 1, 2001, 1234–1240

View Article Online
6.32 (1 H, m, 2-H), 5.17 (1 H, m, 3-Ha), 5.06 (1 H, m, 3-Hb),
4.13 (1 H, d, J = 9 Hz, 1-H), 3.70 (3 H, s, OMe), 3.60 (3 H, s,
OMe), 1.46 (3 H, s, Me); 13C NMR δ 171.3 (C), 171.2 (C), 139.0
(C), 137.0 (CH), 129.4 (2 × CH), 128.1 (2 × CH), 127.0 (CH),
117.6 (CH2), 58.8 (C), 54.5 (CH), 52.4 (CH3), 52.3 (CH3), 18.3
(CH3), in accordance with the literature.15
(2 × CH3), 42.1 (C), 39.0 (CH2), 31.8 (CH2), 25.6 (CH3), 23.3
(CH3), 19.6 (CH3), in accordance with the literature.2,13,19
Methyl 2-methoxycarbonyl-2,3,7-trimethyl-3-vinyloct-6-
enoate 34b. We were unable to obtain a pure sample of 34b from
the reaction of 32 with 7b (General procedure and Tables 2
1
and 3) due to its low concentration, although the H NMR
Dimethyl (1,3-diphenylprop-2-en-1-yl)malonate 10a. 1H NMR
δ 7.35–7.13 (10 H, m, 2 × arom), 6.48 (1 H, d, J = 15.7 Hz,
spectrum of the crude product mixture showed characteristic
1
signals compatible with the structure of 34b: H NMR 5.96
(2 H, m, CH ᎐C), 3.32 (6 H, s, 2 × OMe).
PhCH=), 6.32 (1 H, dd, J = 15.7, 8.5 Hz, PhCH᎐CH), 4.27 (1
᎐
2
᎐
H, dd, J = 11.0, 8.5 Hz, PhCH), 3.96 (1 H, d, J = 11.0 Hz,
CH(CO2Me)2), 3.69 (3 H, s, OMe), 3.50 (3 H, s, OMe); 13C
NMR δ 168.1 (C), 167.7 (C), 140.1 (C), 136.8 (C), 131.8 (CH),
129.1 (CH), 128.6 (CH), 128.4 (CH), 127.8 (CH), 127.5 (CH),
127.1 (CH), 126.3 (CH), 57.6, 52.4 (CH3), 52.3 (CH3), 49.1; IR
Methyl (Z)-2-methoxycarbonyl-5,9-dimethyldeca-4,8-dienoate
35a. Prepared according to General procedure and Tables 2 and
3; separation from its isomers 34a and 36a was accomplished by
preparative HPLC as above (Rt = 30.29 min): 1H NMR δ 5.05 (2
H, m, 2 × C᎐CH), 3.72 (6 H, s, 2 × OMe), 3.35 (1 H, t, J = 7 Hz,
᎐
(neat) ν 2951, 1747, 1601, 1493, 1431, 1319, 1261, 1173, 1142,
2-CH), 2.60 (2 H, m, 3-CH2), 2.05 (4 H, m, 6- and 7-CH2),
1.66 (3 H, s, Me), 1.64 (3 H, s, Me), 1.60 (3 H, s, Me); 13C NMR
δ 169.5 (2 × C), 138.7 (C), 131.8 (C), 124.0 (CH), 120.1 (CH),
52.3 (2 × CH3), 52.1 (CH), 31.8 (CH2), 27.3 (CH2), 26.4 (CH2),
25.6 (CH3), 23.4 (CH3), 17.6 (CH3), in accordance with the
literature.2,13,19
+
ؒ
968, 744 cmϪ1; MS (EI) m/z (%) 324 (29, M ), 264 (11), 232
(17), 205 (87), 193 (100), 178 (20), 128 (10), 115 (73), 91 (30), 77
(9); HRMS (EI) 324.13616 (C20H20O4 requires 324.13623), in
accordance with the literature.16
Dimethyl (4-phenylbut-3-en-2-yl)malonate 11a. Obtained as a
mixture with 8a; for experimental details see General procedure
and Table 4: 1H NMR δ 7.25–7.00 (5 H, m, arom), 6.30 (1 H, d,
J = 16 Hz, 4-H), 6.01 (1 H, dd, J = 16, 8 Hz, 3-H), 3.60 (3 H, s,
OMe), 3.52 (3 H, s, OMe), 3.28 (1 H, d, J = 8 Hz,
CH(CO2Me)2), 2.98 (1 H, m, 1-H), 1.06 (3 H, d, J = 7.4 Hz,
Me), in accordance with the literature.2,3,14,18
Methyl (Z)-2-methoxycarbonyl-2,5,9-trimethyldeca-4,8-di-
enoate 35b. Prepared according to General procedure and
Tables 2 and 3; separation from its isomer 36b was accom-
plished by preparative HPLC on a Dynamax 60 Å column
(C18, 250 × 41.4 mm id) using a 30 : 70 H2O–MeCN mixture,
flow rate 50 mL minϪ1 at 1.58 kpsi, detection by UV at 230 nm.
Analysis of the mixture was performed on a Dynamax 60 Å
column (C18, 250 × 4.6 mm 8 µm id) using a 40 : 60 H2O–
MeCN mixture, flow rate 1 mL minϪ1 at 2.02 kpsi, detection by
UV at 230 nm; 35b was the most polar fraction (Rt = 29.51
Dimethyl methyl(4-phenylbut-3-en-2-yl)malonate 11b. Obtained
as a mixture with 8b; for experimental details see General pro-
cedure and Table 4: 1H NMR δ 7.25–7.00 (5 H, m, arom), 6.45
(1 H, d, J = 15 Hz, 4-H), 6.13 (1 H, dd, J = 15, 8 Hz, 3-H), 3.72
(6 H, s, 2 × OMe), 3.47 (1 H, m, 1-H), 1.44 (3 H, s, Me), 1.16 (3
H, d, J = 8 Hz, 1-Me), in accordance with the literature.2,3,14,18
min): 1H NMR δ 5.10 (2 H, m, 2 × C᎐CH), 3.71 (6 H, s,
᎐
2 × OMe), 2.60 (2 H, d, J = 7.6 Hz, 3-CH2), 2.04 (4 H, m, 6- and
7-CH2), 1.68 (3 H, s, Me), 1.61 (3 H, s, Me), 1.59 (3 H, s, Me),
1.38 (3 H, s, Me); 13C NMR δ 172.6 (2 × C), 139.2 (C), 131.3
(C), 123.9 (CH), 117.9 (CH), 53.8 (C), 52.2 (2 × CH3), 39.8
(CH2), 33.9 (CH2), 26.4 (CH2), 25.5 (CH3), 19.5 (CH3), 17.5
(CH3), 16.0 (CH3); IR (neat) ν 2975, 2940, 1725, 1464, 1452,
Dimethyl (3-phenylprop-2-en-1-yl)malonate 12a. Obtained as
a mixture with 9a; for experimental details see General pro-
cedure and Table 4: 1H NMR δ 7.35–7.20 (5 H, m, Ar), 6.48 (1
H, d, J = 16 Hz, PhCH᎐), 6.15 (1 H, dt, J = 16, 6 Hz, CH᎐), 3.55
᎐
᎐
1435, 1380, 1250, 1170, 1110, 938, 809 cmϪ1
.
(1 H, d, J = 7 Hz, CH(CO2Me)2), 3.51 (6 H, s, 2 × OMe), 2.82
(2 H, dd, J = 7, 6 Hz, CH2); 13C NMR δ 169.0 (2 × C), 140.3
(CH), 134.1 (C), 132.7 (CH), 128.3 (2 × CH), 127.2 (CH), 126.0
(2 × CH), 125.2 (CH), 52.1 (CH), 51.5 (2 × CH3), 32.1 (CH2), in
accordance with the literature.15
Methyl (E)-2-methoxycarbonyl-5,9-dimethyldeca-4,8-dienoate
36a. Prepared according to General procedure and Tables 2 and
3 as a mixture with 34a and 35a, which were separated by pre-
1
parative HPLC as above, (Rt = 29.66 min): H NMR δ 5.05
(2 H, m, 2 × C᎐CH), 3.72 (6 H, s, 2 × OMe), 3.35 (1 H, t, J =
᎐
Dimethyl
methyl(3-phenylprop-2-en-1-yl)malonate
12b.
7 Hz, CH(CO2Me)2), 2.61 (2 H, m, 3-CH2), 1.96 (4 H, m, 6- and
7-CH2), 1.67 (3 H, s, Me), 1.63 (3 H, s, Me), 1.58 (3 H, s, Me);
13C NMR δ 169.5 (2 × C), 138.6 (C), 131.4 (C), 124.0 (CH),
119.4 (CH), 52.3 (2 × CH3), 51.9 (CH), 39.6 (CH2), 27.5 (CH2),
26.5 (CH2), 25.6 (CH3), 17.6 (CH3), 16.0 (CH3), in accordance
with the literature.2,13,19
Obtained as a mixture with 9b; for experimental details see
General procedure and Table 4: 1H NMR δ 7.36–7.18 (5 H, m,
arom), 6.42 (1 H, d, J = 16 Hz, 3-H), 6.09 (1 H, dq, J = 16, 8 Hz,
2-H), 3.73 (6 H, s, 2 × OMe), 2.78 (2 H, d, J = 8 Hz, 1-CH2),
1.48 (3 H, s, Me); 13C NMR δ 172.2 (2 × C), 136.9 (C), 134.0
(CH), 128.4 (2 × CH), 127.3 (2 × CH), 126.1 (CH), 124.1 (CH),
53.9 (C), 52.2 (2 × CH3), 39.4 (CH2), 20.0 (CH3), in accordance
with the literature.15
Methyl
(E)-2-methoxycarbonyl-2,5,9-trimethyldeca-4,8-
dienoate 36b. Prepared according to General procedure and
Tables 2 and 3 as a mixture with 35b, which was separated by
preparative HPLC; 36b was the least polar fraction (Rt = 28.32
Methyl 2-methoxycarbonyl-3,7-dimethyl-3-vinyloct-6-enoate
34a. Prepared according to General procedure and Tables 2 and
3, separation from the regioisomers 35a and 36a was accom-
plished by preparative HPLC on a Dynamax 60 Å column
(C18, 250 × 41.4 mm id) using a 40 : 60 H2O–MeCN mixture,
flow rate 50 mL minϪ1 at 1.59 kpsi, detection by UV at 230 nm.
Analysis of the mixture was performed on a Dynamax 60 Å
column (C18, 250 × 4.6 mm, 8 µm id) using a 45 : 55 H2O–
MeCN mixture, flow rate 1 mL minϪ1 at 2.30 kpsi, detection by
1
min): H NMR δ 5.10 (1 H, m, 8-H), 4.98 (1 H, t, J = 7.3 Hz,
4-H), 3.73 (6 H, s, 2 × OMe), 2.60 (2 H, d, J = 7.3 Hz, 3-CH2),
2.05 (4 H, m, 6- and 7-CH2), 1.70 (3 H, s, Me), 1.68 (3 H, s, Me),
1.61 (3 H, s, Me), 1.38 (3 H, s, Me); 13C NMR δ 172.6 (2 × C),
139.3 (C), 131.6 (C), 124.0 (CH), 118.4 (CH), 53.7 (C), 52.3
(2 × CH3), 33.8 (CH2), 31.9 (CH2), 26.5 (CH2), 25.6 (CH3), 23.5
(CH3), 19.8 (CH3), 17.6 (CH3); IR (neat) ν 2979, 2960, 1728,
1465, 1460, 1435, 1409, 1380, 1280, 1195, 1120, 942 cmϪ1
.
1
UV at 230 nm (Rt = 31.37 min): H NMR δ 6.01 (1 H, dd,
J = 17, 11 Hz, CH᎐CH ), 5.15–5.00 (3 H, m, Me C᎐CH,
᎐
᎐
2
2
Acknowledgements
C᎐CH ), 3.73 (6 H, s, 2 × OMe), 3.46 (1 H, s, CH(CO Me) ),
᎐
2
2
2
1.70–1.56 (4 H, m, 2 × CH2), 1.62 (3 H, s, Me), 1.58 (3 H, s,
Me), 1.21 (3 H, s, Me); 13C NMR δ 168.2 (2 × C), 131.5 (C),
124.0 (CH), 120.1 (CH), 113.8 (CH2), 59.9 (CH), 52.0
We thank the EPSRC for grants No. GR/H 92067, GR/K07140
and GR/L41448 and EPSRC and Aventis CropScience UK Ltd
for a CASE award to I. R. B.
J. Chem. Soc., Perkin Trans. 1, 2001, 1234–1240
1239

View Article Online
(d) A. V. Malkov, S. L. Davis, W. L. Mitchell and P. Kocˇovský,
Tetrahedron Lett., 1997, 38, 4899; (e) A. V. Malkov, I. R. Baxendale,
D. J. Mansfield and P. Kocˇovský, J. Org. Chem., 1999, 64, 2737; ( f )
A. V. Malkov, S. L. Davis, I. R. Baxendale, W. L. Mitchell and P.
Kocˇovský, J. Org. Chem., 1999, 64, 2751. For catalysis of ene and
Prins reactions by Mo() and W(), see: (g) P. Kocˇovský, G. Ahmed,
References
1 For reviews, see: (a) B. M. Trost, Tetrahedron, 1977, 33, 371; (b)
B. M. Trost, Acc. Chem. Res., 1980, 13, 385; (c) J. Tsuji, Tetrahedron,
1986, 42, 4361; (d) J. Tsuji, Synthesis, 1990, 739; (e) C. G. Frost,
J. Howarth and J. M. J. Williams, Tetrahedron Asymmetry, 1992, 3,
1089; ( f ) B. M. Trost and D. L. Van Vranken, Chem. Rev., 1996, 96,
395. For the Pd(0) regioselectivity, see, e.g. (g) G. Giambastriani and
G. Poli, J. Org. Chem., 1998, 63, 9608.
ˇ
J. Srogl, A. V. Malkov and J. Steele, J. Org. Chem., 1999, 64, 2765.
For Mo()-catalyzed allylic substitution, see: (h) A. V. Malkov, P.
Spoor, V. Vinader and P. Kocˇovský, J. Org. Chem., 1999, 64, 5308.
9 For the use of computational methods to rationalize the regio-
selectivity in Pd(0)-catalysed reactions, see: H. Hagelin, B. Åkermark
and P.-O. Norrby, Chem. Eur. J., 1999, 3, 902.
2 Mo: (a) B. M. Trost and M. Lautens, J. Am. Chem. Soc., 1982, 104,
5543; (b) B. M. Trost and M. Lautens, J. Am. Chem. Soc., 1983, 105,
3343; (c) B. M. Trost and M. Lautens, Organometallics, 1983, 2,
1687; (d) B. M. Trost, M. Lautens and B. Peterson, Tetrahedron
Lett., 1983, 24, 4525; (e) B. M. Trost and M. Lautens, J. Am. Chem.
Soc., 1987, 109, 1469; ( f ) B. M. Trost and M. Lautens, Tetrahedron,
1987, 43, 4817; (g) B. M. Trost and C. A. Merlic, J. Am. Chem. Soc.,
1990, 112, 9590; (h) B. M Trost and C. A. Merlic, J. Org. Chem.,
1990, 55, 1127; (i) B. M. Trost and I. Hachiya, J. Am. Chem. Soc.,
1998, 120, 1104; (j) B. M. Trost, S. Hildbrand and K. Dogra, J. Am.
Chem. Soc., 1999, 121, 10416; (k) L. Eriksson, M. P. T. Sjögren and
B. Åkermark, Acta Crystallogr., Sect. C, 1996, 52, 77; (l) M. P. T.
Sjögren, H. Frisell and B. Åkermark, Organometallics, 1997, 16, 942.
3 W: (a) B. M. Trost and M.-H. Hung, J. Am. Chem. Soc., 1983, 105,
7757; (b) B. M. Trost and M.-H. Hung, J. Am. Chem. Soc., 1984,
106, 6837; (c) G. C. Lloyd-Jones and A. Z. Pfaltz, Angew. Chem., Int.
Ed. Engl., 1995, 34, 462; (d) G. C. Lloyd-Jones and A. Pfaltz,
Z. Naturforsch., Teil B, 1995, 50, 361; (e) J. Lehmann and G. C.
Lloyd-Jones, Tetrahedron, 1995, 51, 8863; ( f ) H. Frisell and B.
Åkermark, Organometallics, 1995, 14, 561.
4 D. Dvorˇák, I. Starý and P. Kocˇovský, J. Am. Chem. Soc., 1995, 117,
6130.
5 (a) J. W. Faller and D. Linebarrier, Organometallics, 1988, 7, 1670;
(b) Y. D. Ward, L. A. Villanueva, G. D. Allred and L. S. Liebeskind,
J. Am. Chem. Soc., 1996, 118, 897.
6 L. S. Hegedus, Transition Metals in the Synthesis of Complex
Organic Molecules, University Science Books, Mill Valley, CA, 1994.
7 Replacing some of the carbonyls in Mo(CO)6 and W(CO)6 by other
ligands, such as MeCN, DMF, etc. has been shown to slightly
accelerate the reaction.2,5b See also: (a) A. J. Pearson and E.
Schoffers, Organometallics, 1997, 16, 5365. Another acceleration
has been observed with pyridine as the ligand for stoichiometric
reactions: (b) A. Kuhl, J. A. Christopher, L. J. Farrugia and P. J.
Kocien´ ski, Synlett, 2000, 1765; (c) A. Kuhl, J. A. Christopher, L. J.
Farrugia and P. J. Kocien´ ski, Acta Crystallogr., Sect. C, 2000, 56,
510.
10 (a) P. K. Baker and A. Bury, J. Organomet. Chem., 1989, 359, 189;
(b) P. K. Baker and A. Quinlan, Inorg. Chim. Acta, 1989, 162, 179;
(c) D. Miguel, J. A. Perez-Martinez and S. Garcia-Granda,
Polyhedron, 1991, 10, 1717; (d) G. Barrado, D. Miguel, J. A.
Perez-Martinez and V. Riera, J. Organomet. Chem., 1993, 463, 127;
(e) M. Cano, M. Panizo, J. A. Campo, J. Tornero and N. Menendez,
J. Organomet. Chem., 1993, 463, 121.
11 (a) B. M. Trost and A. Brandi, J. Org. Chem., 1984, 49, 4811; (b) J. A.
Marshall, R. C. Andews and L. Lebioda, J. Org. Chem., 1987, 52,
2378; (c) B. M. Trost and J. M. Tour, J. Org. Chem., 1989, 54, 484;
(d) T. Mino, W. Imiya and M. Yamashita, Synlett, 1997, 5, 583; (e)
T. Morimoto, K. Tachibana and K. Achiwa, Synlett, 1997, 7, 783.
12 For these “memory effects” in Pd(0)-catalysed allylic substitution,
ˇ
see: G. C. Lloyd-Jones, S. C. Stephen, M. Murray, C. P. Butts, S.
Vyskocˇil and P. Kocˇovský, Chem. Eur. J., 2000, 6, 4348 and refs. cited
therein.
13 Chelation by the terminal double bond has been detected for the
corresponding η3-Pd-complex by NMR spectroscopy: B. Åkermark
and A. Vitagliano, Organometallics, 1985, 4, 1275.
14 (a) T. Hayashi, A. Yamamoto and T. Hagihara, J. Org. Chem., 1986,
51, 723; (b) R. G. Hayter, J. Organomet. Chem., 1968, 13, 1.
15 B. M. Trost, G. B. Tometzki and M.-H. Hung, J. Am. Chem. Soc.,
1987, 109, 2176.
16 (a) N. Yamaguchi, T. Shima, T. Yamagishi and M. Hids,
Tetrahedron: Asymmetry, 1991, 2, 663; (b) A. Togni, Tetrahedron:
Asymmetry, 1991, 2, 683.
17 I. Starý, I. G. Stará and P. Kocˇovský, Tetrahedron, 1994, 50, 529.
18 (a) S.-W. Zhang, T.-A. Mitsudo, T. Kondo and Y. Watanabe,
J. Organomet. Chem., 1993, 450, 197; (b) J. M. Brown, D. I. Hulmes
and P. Guiry, Tetrahedron, 1994, 50, 4493; (c) T. Doi, A. Yanagisawa,
M. Miyazawa and K. Yamamoto, Tetrahedron: Asymmetry, 1995, 6,
389; (d) C. Goux, D. Lhoste and D. Sinou, Tetrahedron Lett., 1994,
34, 10321.
ˇ
8 (a) H. Dvorˇáková, D. Dvorˇák, J. Srogl and P. Kocˇovský, Tetrahedron
19 B. M. Trost and T. R. Verhoeven, J. Am. Chem. Soc., 1980, 102,
4730.
20 J. A. Riddick, W. B. Bunger and T. K. Sakano, Techniques of
Chemistry, Vol. 2, Organic Solvents, 4th edn., J. Wiley & Sons, New
York, 1986.
Lett., 1995, 36, 6351; (b) A. P. Abbott, A. V. Malkov, N.
Zimmermann, J. B. Raynor, G. Ahmed, J. Steele and P. Kocˇovský,
Organometallics, 1997, 16, 3690; (c) A. V. Malkov, I. R. Baxendale,
D. J. Mansfield and P. Kocˇovský, Tetrahedron Lett., 1997, 38, 4895;
1240
J. Chem. Soc., Perkin Trans. 1, 2001, 1234–1240