Vinylic, allylic and homoallylic oxidations of alkenes via π- and σ-organopalladium complexes

Article abstract of DOI:10.1016/S0022-328X(99)00521-5

The stoichiometric and catalytic pathways of oxidative esterification of alkenes via intermediate organopalladium complexes are discussed. The oxidation of propylene, hex-1-ene and cyclohexene by Pd^II acido complexes containing achiral, racemic and chiral carboxylate ligands was first studied in a series of solvents other than acetic acid. Significant changes in the selectivity of the Pd^II-promoted reaction with changes in the solvent nature and ligand chirality were observed. A way to allylic esters based on low-valence Pd nanoclusters provide highly selective oxidation of acyclic alkenes into allylic esters, whereas cycloalkenes undergo mostly redox disproportionation. The role of π-alkene, σ-alkenyl and π-allyl complexes in the mechanism of the alkene oxidative esterification with Pd^II complexes and low-valence Pd clusters is discussed.

Full text of DOI:10.1016/S0022-328X(99)00521-5

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 593–594 (2000) 274–291
Vinylic, allylic and homoallylic oxidations of alkenes via p- and
s-organopalladium complexes
N.Yu. Kozitsyna, M.N. Vargaftik, I.I. Moiseev *
N.S. Kurnako6 Institute of General and Inorganic Chemistry, Russian Academy of Sciences, 31, Leninski Prospect, GSP-1,
Moscow 117907, Russian Federation
Received 8 June 1999; accepted 7 September 1999
Dedicated to Professor Fausto Calderazzo in recognition of his outstanding contribution to inorganic chemistry.
Abstract
The stoichiometric and catalytic pathways of oxidative esterification of alkenes via intermediate organopalladium complexes are
discussed. The oxidation of propylene, hex-1-ene and cyclohexene by PdII acido complexes containing achiral, racemic and chiral
carboxylate ligands was first studied in a series of solvents other than acetic acid. Significant changes in the selectivity of the
PdII-promoted reaction with changes in the solvent nature and ligand chirality were observed. A way to allylic esters based on
low-valence Pd nanoclusters provide highly selective oxidation of acyclic alkenes into allylic esters, whereas cycloalkenes undergo
mostly redox disproportionation. The role of p-alkene, s-alkenyl and p-allyl complexes in the mechanism of the alkene oxidative
esterification with PdII complexes and low-valence Pd clusters is discussed. © 2000 Elsevier Science S.A. All rights reserved.
Keywords: Palladium; p-Complexes; s-Complexes; Alkenes; Oxidative esterification; Regioselectivity; Solvent effects; Ligand effects; Cluster
catalysis; Reaction mechanism
1. Introduction
Organic reactions catalysed by derivatives of Group
VIII metals are still poorly understood notwithstanding
the progress in this field during the last four decades. In
this context, the nature of organopalladium intermedi-
ates of the catalytic alkene oxidations is of great
Scheme 1.
interest.
The ability of Pd(II) carboxylates to form alkene
p-complexes, whose redox decomposition affords
vinylic or allylic esters [1] is used in the synthetic
In this Scheme Ox=1/2O2, CuII, p-benzoquinone or
practice [2–5]. Oxidative esterification of alkenes with
benzoyl peroxide, Red=H2O, Cu (I), hydroquinone or
benzoic acid, respectively.
Reactions shown in Scheme 1 were first observed as
early as in 1960 [5]. During the study of the reactivity of
palladium-based catalysts provides a direct synthetic
way to alkenyl esters. In this reaction, alkenes undergo
replacement of an H atom by an RCOO group in the
vinyl (v), allyl (a) or homoallyl (h) positions of the
alkene molecule (see Scheme 1)):
palladium(II)
p-complexes
toward
nucleophilic
reagents, we have found [5] that the Kharash complex
(p-C2H4·PdCl2)2 [6], which is fairly stable in glacial
acetic acid, readily decomposes when alkaline acetate is
added to its AcOH solution, producing vinyl acetate
and Pd0 according to Eq. (1):
* Corresponding author. Tel.: +7-095-952-1203; fax: +7-095-954-
1279.
E-mail address: iimois@ionchran.rinet.ru (I.I. Moiseev)
0022-328X/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00521-5

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
275
(p-C2H4·PdCl2)2+4NaOAc=CH2
=CHOAc+2Pd0+4NaCl
+2AcOH (1)
Alkeneꢀpalladium(II) p-complexes were found [7] to
be formed by the reaction of Pd(II) chloride with
alkenes in an AcOH solution as shown in Eq. (2):
Scheme 2.
cerning with its regioselectivity and mechanism remain
unsolved till now.
2PdCl2+2C2H4=[p-C2H4·PdCl2]2
(2)
The p-complex does not undergo further reactions [7]
until an alkaline acetate or another nucleophilic agent
are absent from the solution. When NaOAc was added,
the p-complex was immediately converted to Pd0 and
vinyl acetate [5] (see Eq. (1)).
Direct reaction between PdCl2 and ethylene in an
AcOH solution containing alkaline acetate has been
found [5] to produce vinyl acetate (see Eq. (3)) and a
small amounts of ethylidene diacetate (Eq. (3a)).
In this paper, we present a fresh insight into the
mechanism of the alkene oxidative esterification based
on both previous data and our recent experiments on
the solvent and ligand effects in the oxidative esterifica-
tion of alkenes (propylene, cyclohexene and hex-1-ene)
by PdII complexes with achiral and chiral carboxylate
ligands in different solvents other than acetic acid. On
the basis of data for reactions mediated by PdII com-
plexes and Pd nanoclusters, the pathways of formation
of vinylic, allylic and homoallylic esters via p- and
s-organopalladium complexes are discussed.
PdCl2+C2H4+2NaOAc
=CH2ꢁCHOAc+Pd0+2NaCl+HOAc
PdCl2+C2H4+2NaOAc
(3)
2. Reactions with Pd(II) complexes
=CH3CH(OAc)2+Pd0+2NaCl
(3a)
Oxidation of alkenes by PdII in aqueous, alcoholic
and acetic acid solutions (see Scheme 2) has been
studied extensively [1,4].
Palladium (0) formed in reaction (3) was shown to
undergo re-oxidation to PdII by such oxidants (Ox) as
p-benzoquinone or O2+CuII:
The reactions in Scheme 2 can be carried out as
Pd-catalysed processes, being supplemented by re-oxi-
dation of the Pd0 formed (see Eq. (6)). The reaction in
aqueous solution (path ꢀaꢁ) is the basis for the
commercial production of acetaldehyde from ethylene
(Wacker–Hoechst process) [7], and its mechanism have
been studied in most detail. The other two reactions,
those in alcohol (ꢀbꢁ) and AcOH (ꢀcꢁ) solutions,
are also of practical interest [2–4].
Pd0+Ox+2Cl− =PdCl2+Red,
(4)
(Red=hydroquinone or Cu(I)+H2O, respectively),
thus resulting in a catalytic cycle, in which ethylene is
converted into vinyl acetate and where palladium serves
as the catalyst:
(5)
The main mechanistic features of the reactions in
Scheme 1 were first found in early works [5,9,19,22].
Essential idea of the mechanism proposed is that the
oxidation of an alkene molecule by PdII complex occurs
via three stages:
C2H4+AcOH+O2=CH2ꢁCHOAc+H2O
(6)
Under the conditions of stoichiometric (Eq. (3)) or
catalytic (Eqs. (5) or (6)) reactions, propylene is oxi-
dized to isopropenyl acetate as the main reaction
product, along with n-propenyl and allyl acetates [8–
13]. Higher acyclic alkenes C4ꢀC10 are converted to
mixtures of alkenyl acetates, and the product distribu-
tion between vinylic and allylic esters varies in a wide
range depending on the reaction conditions (tempera-
ture, time of reaction, concentrations of the oxidant,
alkaline acetate and water, etc.) [9–18]. In the case of
cyclic alkenes homoallylic esters are also formed [18–
24] (Scheme 1, path h).
(I) formation of alkeneꢀPdII p-complex:
−
PdX24−+C2H4−XX y-C2H4·PdX3−RXO y-C2H4·PdX2OR−
−
−
−X
(7)
(R=OH, OAlk, OAc)
(II) oxymetallation of the alkene CꢁC bond via iso-
merization of the p-complex to s-organopalladium
compound involving inner-sphere [1] or outer-sphere [4]
nucleophilic attack:
p-C2H4·PdX2OR− “XPd2ꢀCH2CH2OR−
(8)
The reactions in Scheme 1 have been studied by
many research groups (see, for instance, recent review
paper [4]). Nevertheless, some important problems con-
(III) redox disproportionation of the s-complex to
Pd0 and the product of alkene oxidation which depends
on the solvent as seen in Eq. (9):

276
N.Y. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
(9)
Studies of the reaction kinetics and equilibria of
Scheme 4.
p-complex formation for C2ꢀC6 alkenes [25–39] have
confirmed the original suggestions.
A rather risky (in early 60s, before R.F. Heck [40])
statement of the involvement of
alcohols and acetic acid). Meanwhile, some problems
are still not solved.
a
s-bonded
(1.) The oxidation of 1-alkenes higher than ethylene
in aqueous and alcohol solutions (paths ꢀaꢁ and
ꢀbꢁ in Scheme 2) produces ketones and ketals, re-
spectively [1,3]. This means, within the framework of
the mechanism under question, that the oxypalladation
stage occurs according to the Markovnikov rule: an
anion adds to the least hydrogenated C atom, and PdII
electrophilic atom adds to another C atom at the
double bond of the alkene molecule (see Scheme 4):
The ketals in Scheme 4 are genetically related to the
carbonyl compounds, and by analogy with this, alkyli-
dene diesters could be expected to form in an AcOH
solution. The question arises: why unsaturated reaction
products, alkenyl esters, are formed instead of alkyli-
dene diesters in an AcOH medium (see Scheme 5)?
(2.) Both reactions, ꢀaꢁ and ꢀbꢁ in Scheme 2,
are highly regioselective. Contrary to these, the reaction
in an AcOH medium (path ꢀcꢁ) is much less selec-
tive, and mixtures of vinylic and allylic esters are com-
monly obtained [3,4,18,22]. The question arises: why is
the reaction in the AcOH/NaOAc system much less
selective than those in aqueous and alcohol solutions?
(3) Oxidative acetoxylation of cyclic and higher
acyclic alkenes in AcOH medium affords, as a rule,
allylic and little or no vinylic esters. In addition, ho-
moallylic esters were often found among the reaction
products [13,21,23]. Meanwhile, homoallylic oxidations
are absent from aqueous and alcoholic solutions
[1,7,18]. The question arises: what is the reason for
homoallylic oxidation in the AcOH/NaOAc system?
(4) Acetic acid plus OAc− anion have mostly been
employed in studies of reaction (1). In fact, the scope of
this reaction was still restricted to oxidative acetoxyla-
tion in an AcOH medium. Studies with other solvents
and carboxylate ligands are scarce (few reports such as
Ref. [42a,b] for the CF3COO−/AcOH and AcOH/
CH2Cl2 systems seem to be rather exceptions). The
question arises: is it possible to perform the alkene
oxidative esterification in other solvents but acetic acid
with other carboxylates but acetate anion?
alkylꢀpalladium(II) derivative into the oxidation has
been confirmed by our experiments [1,41], in which
palladium organyls were generated independently via
the metathesis of palladium chloride with b-halo-, b-
oxy- and b-acetoxyethylmercurials:
XHgCH2CH2OR+PdCl2
“ClPdCH2CH2OR+HgClX
(10)
(X=Cl, Br, OH; R=H, Et, Ac)
The intermediate s-organopalladium compounds are
unstable toward redox decomposition. The products of
their fast transformation are Pd0 and the oxygen-con-
taining products whose nature depends on a solvent. In
an ether or aqueous solution acetaldehyde was formed:
ClPdꢀCH2CH2OH“Pd0+HCl+CH3CHO
(11)
In glacial acetic acid containing NaOAc vinyl acetate
(and traces of ethylidene diacetate) were obtained:
ClPdꢀCH2CH2OAc=Pd0+HCl+CH2ꢁCHOAc
(12)
The same products were found in the direct reaction
between alkene and PdCl2 in these solvents [1,41].
All three reactions in Scheme 2 were rationalized as
involving intermediate formation of the key palladium
organyls Pd-Ch2Ch2OR (where R=H, Et or Ac)
groups. Heterolytic cleavage of the PdꢀC bond of the
intermediate, involving hydride 1,2-shift, possibly as-
sisted by Pd atom interaction, was assumed to give rise
to Pd atoms in a reduced oxidation state and a carboca-
tion preceding the product of alkene oxidation (see
Scheme 3):
The driving force for the key stage shown in Scheme
3 was assumed to be a tendency of the PdII atom to be
reduced by withdrawing an electron pair from the
organic moiety.
At first glance it seems that this general mechanism
gives a plausible explanation for all reactions between
PdII and alkenes in different media (e.g., in water,
Scheme 3.
Scheme 5.

N.Y. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
277
(5) The catalytic cycle of reaction (1) formally in-
volves PdII and Pd0 (see Eqs. (3) and (4)). It seemed
that the reaction can start from either of these palla-
dium species. Nevertheless, the product distribution of
the reaction starting from Pd metal (for example, Pd
black or supported Pd/SiO2 catalyst) is quite different
from that starting from PdII [17,43–47]. The question
arises: what is in fact the valence state and chemical
nature of the Pd species responsible for the stoichiomet-
ric and catalytic oxidative esterification?
In view of these problems, we recently studied the
solvent and ligand effects on the regioselectivity of
oxidative esterification of propylene, hex-1-ene and cy-
clohexene by Pd(II) carboxylates using different achiral,
racemic and chiral carboxylate ligands. The reaction
media were aprotic solvents of different polarities and
solvation abilities: chloroform, dichloromethane, 1,1,2-
trichloroethane, liquid CO2 and THF.
Fig. 1. The structure of the complex Pd3[S(+)-2-EtC*H(Me)COO]6
from the single crystal X-ray diffraction data [48]. Interatomic dis-
2.1. Reacti6e forms of PdII
,
tances (A): Pd(1)…Pd(2) 3.153(4); Pd(2)…Pd(3) 3.125(4);
Pd(1)…Pd(3) 3.128(5); six Pd–O distances are within the range of
In our experiments palladium(II) was introduced into
reaction solutions in the form of neutral salts
Pd3(RCO2)6 (R=Me, Me2CHCH2, (9)-CF3CF2CF2-
OC*F(CF3), S(+)-MeC*H(Et) and (+)-CF3CF2CF2-
OC*F(CF3)). The molecular structure of the first
synthesized Pd(II) salt with chiral (+)-S-2-methylbu-
tyrate anion was established by X-ray diffraction study
of the single crystal [48].
According to the X-ray data [48], the molecule of
PdII (+)-S-2-methylbutyrate is a carboxylate-bridged
cyclic trimer Pd3(m2-S(+)-MeC*H(Et)COO)6 with the
structure (Fig. 1) similar to those of the previously
investigated PdII salts with achiral carboxylate anions
(acetate, propionate and pivalate, see for instance [49–
52]).
,
1.913(37)–2.074(31) A. Angles in the Pd3 triangle (°):
Pd(2)Pd(1)Pd(3) 59.7(1); Pd(1)Pd(3)Pd(2) 60.6(1); Pd(1)Pd(2)Pd(3)
59.8(1).
plexes according to Eqs. (13) and (14) [45,56a–c,57],
and the most of the additional RCOO− anions entered
the composition of the anionic PdII complexes formed:
2[Pd(OAc)2]3+6OAc− X 3[Pd2(OAc)6]2−
[Pd(OAc)2]3+6OAc− X 3[Pd(OAc)4]2−
(13)
(14)
The observed reactions with alkenes are due to the
more reactive anionic complexes [Pd2(OAc)6]2− and
[Pd(OAc)4]2− because trinuclear PdII complexes are
fairly inert kinetically toward alkene oxidations [45,57].
Studies of reaction (1) in low polar organic solvents
turned out to be difficult due to a poor solubility of
alkaline carboxylates. For this reason, we chose bis(t-
2.2. Sol6ent effects
riphenylphosphoranilidene)ammonium
(Ph3P)2N+,
A dramatic increase in the allylic versus vinylic ester-
ification of propylene by the complexes [Pd2(OAc)6]2−
and [Pd(OAc)4]2− was found [55,59] when acetic acid
was changed for CH2Cl2, CHCl3 or CHCl2CH2Cl. The
yield of allyl acetate was at least 99% based on propy-
lene and PdII consumed, whereas only traces of iso-
propenyl and n-propenyl acetates and acetone were
found (Table 1, entries 1–3). The same product compo-
sition was found in catalytic runs with the use of
nitrosobenzene and benzoyl peroxide as oxidants
(Table 1, entries 4, 5). The involvement of Pd0 was
practically ruled out in the presence of these oxidants.
When 5% (vol.) of AcOH was added to the chloro-
carbon solvents, noticeable amounts of isopropenyl and
n-propenyl acetates (and also acetone) appeared in the
reaction products (overall yield of 10%, see Table 1,
entry 6). The decrease in the reaction selectivity with
PNP, carboxylates as the source of additional carboxy-
late anions (see Eqs. (13) and (14)) instead of com-
monly used sodium or potassium salts. PNP salts,
including those with metal complex anions, are well
soluble in aprotic low polar media [53–55].
When the PNP salt of the corresponding carboxylate
anion, (PNP)OCOR, was added to the reaction solu-
tion in a molar ratio of RCOO−:PdII51:2.5, the origi-
nal trinuclear complex Pd3(m2-OCOR)6 was partly
transformed1 into di- and mononuclear anionic com-
1 Solutions with overall concentrations of 0.05 M of Pd3(OCOR)6
(calculated for the PdII mononuclear unit) and 0.01}0.02 M of
(PNP)OCOR− were used in our experiments. Under these condi-
tions, at most 1/3 of original Pd3(OCOR)6 can be converted into
Pd2(OCOR)26− and Pd(OCOR)24−
.

278
N.Y. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
Table 1
Solvent effects in the acetoxylation of propylene by PdII complexes ([Pd(OAc)2]0=0.11 mol l−1, [(PPN)OAc]0=0.04 mol l−1, p(C3H6)=1 atm,
25°C, time of reaction 3 h)
Number Solvent
T (°C)
Reaction products (mol/mol Pd)
Allyl:vinyl ratio
Allyl acetate Isopropenyl n-Propenyl acetate Acetone+propanal
acetate
cis+trans
1
2
3
4
5
6
7
8
CHCl3
CH2Cl2
25
25
25
0.99
0.99
0.99
0.99
0.99
0.92
0.94
0.49
0.01
0.01
0.01
0.01
0.01
0.058
0.01
0.01
–
–
–
\1000
\1000
\1000
\1000
\1000
11
–
–
–
–
–
CHCl2CH2Cl
CH2Cl2+benzoyl peroxide 25
CH2Cl2+nitrosobenzene
CHCl2CH2Cl+5%AcOH
AcOH+0.9 M NaOAc
AcOH+0.9 M NaOAc+
1% H2O
25
25
25
25
–
0.023
0.06 (total vinylic products)
0.51 (total vinylic products)
16
ꢂ1
9
AcOH
25
25
42
46
0.009
0.62
0.76
0.72
0.986
0.005
–
B0.01
1.6
3.1
10
11
12
CO2 liquid a
0.38 (total vinylic products)
0.24 (total vinylic products)
0.28 (total vinylic products)
CO2 fluid b
CO2 fluid+10%AcOH b
2.6
a p(CO2)=90 atm, p(C3H6)=12 atm.
b p(CO2)=90 atm, p(C3H6)=12 atm.
respect to allyl acetate is in line with our (Table 1, entry
8) and literature [43,57] (Table 1, entry 7) data for
reaction (1) in an AcOH solution.
Reaction of cyclohexene with PdII carboxylates in
low polar aprotic solvents is very sensitive toward the
presence of oxidants. When oxidative reagents, which
can re-oxidize Pd0 to PdII (p-benzoquinone, nitrosoben-
zene or benzoyl peroxide) were absent from the reac-
tion solution, the main reaction product was benzene
with the yield at least 98% based on cyclohexene con-
sumed. The esters of cyclohexen-2-ol and cyclohexen-3-
ol were formed in negligible yields (Table 2, entry 2). In
this case cyclohexene undergoes oxidative dehydrogena-
tion (Eq. (16)) and disproportionation (Eq. (17), the
Zelinsky reaction). Both reactions are efficiently
catalysed by highly dispersed Pd metal that is formed
from the very beginning of reaction.
It was found that the oxidation of propylene in liquid
CO2 containing ꢂ1 wt.% H2O produced a mixture of
allylic and vinylic esters in both liquid and superctitical
regions. In fact, this reaction medium is a ꢂ0.5 M
solution of H2CO3 in liquified CO2. As seen in Table 1
(entries 10–12), the product distribution is close to that
in an AcOH medium, and it is little wonder that the
addition of AcOH almost did not change the reaction
selectivity (Table 1, entry 12).
Our experiments showed that vinylic esters also are
entirely absent from the products of the oxidative ace-
toxylation of hex-1-ene in a CHCl3 solution. Allylic,
mainly cis- and trans-hex-2-en-1-ol acetates, and ho-
moallylic hex-3-en-1-ol acetate were found as the oxida-
tion products (Eq. (15))2 [58].
0
Pd
C6H10+2Pd(OCOR)2“C6H6+2Pd0+4RCOOH
(16)
0
Pd
3C6H10“C6H6+2C6H12
(17)
CH2ꢁCH(CH2)3
In the presence of p-benzoquinone, benzoyl peroxide
or nitrosobenzene reactions (16) and (17) are practically
absent, and the esters of cyclohex-2-en-1-ol and cyclo-
hex-3-en-1-ol are the main reaction product in all
aprotic solvents (Table 2, entries 3–13). No vinylic
esters were found in the reaction products.
Thus, on going from protic (CH3COOH, H2CO3/
CO2) to aprotic (CHCl3, CH2Cl2, 1,1,2-trichloroethane,
THF) media a sharp variation in the regioselectivity of
reaction (1) occurs: the vinylic route of alkene esterifi-
cation disappears, the main reaction is allylic esterifica-
tion, and in the case of C6 alkenes a homoallylic route
is present along with the allylic route.
Pd(OAc)
CHCl
2
AcO−CH CHꢁCH CHCH CH (ꢂ70%)
AcO−CH2CH2CHꢁCHCH2CH3 (ꢂ15%)
3
!
2
2
2
3
CH3ꢀꢀꢀꢀꢀꢂ
−AcOH
0
−Pd
(15)
These findings differ from those in early studies
[13,18], where substantial amounts of vinylic esters have
been obtained by the oxidation of 1-alkenes C3ꢀC5 by
Pd(OAc)2 in AcOH solutions containing alkaline
acetates.
2 Other allylic esters, e.g., hex-2-ene-4-ol acetate, were also found as
minor reaction products.

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
279
2.3. Ligand effects
In order to gain additional information on the nature
of these routs, the ligand chirality effects in the oxidative
esterification of cyclohexene by PdII carboxylates were
studied.. As seen in Table 2, the oxidation of cyclohexene
by PdII complexes containing different carboxylate lig-
ands in aprotic media affords the corresponding allylic
and homoallylic esters, and vinylic oxidation is absent.
The cyclohexene molecule possesses two pairs of
enantiotopic centres in the a- and b-positions relative to
the double bond. Owing to this, reaction (1) could has
afforded optically active cyclohexenyl esters. The ester
molecule containing a chiral carboxylate has two chiral
centres: one C* atom in the a- or b-position of the
cyclohexenyl group and another C* atom in the carboxy-
late group (Scheme 6).
Stereoisomeric composition of the products of cyclo-
hexene esterification by PdII complexes with chiral an-
ions (S(+)-MeC*H(Et)CO−2 and (+)-CF3CF2CF2-
OC*F(CF3)CO−2 in chloroform was investigated by
GC/MS. The capillary column with the universal
stereoselective phase a-Decs failed to satisfactorily sepa-
rate the diastereoisomeric cyclohexenol esters [58]. For
this reason the mixtures of the esters obtained were
preliminarily hydrolyzed by 10% methanolic NaOH with
retention of configuration [60]. It was found by GC/MS
that the stereoisomeric cyclohexene-2-ols obtained after
hydrolysis
are
present
in
practically
equal
concentrations3. Therefore, the mechanism of reaction
(1) is inadequate to utilize the asymmetric induction of
the chiral carboxylate ligands.
Meanwhile, the chirality of carboxylate ligand affects
noticeably the regioselectivity of reaction (1). As seen in
Table 2 (entries 7,8), the allyl: homoallyl ratio increased
more than twofold (from 2.4 to 5.8) on going from the
racemic (9)-Rf*COO− anion to chiral (+)-R*f COO−
anion when cyclohexene was oxidized by the PdII com-
plexes in chloroform. Even more pronounced effect was
found in the reaction involving 2-methylbutyrate ligand.
In this case the allyl: homoallyl ratio increased in more
than one order of value, from 6.5 for racemic to
64.4 for chiral S(+) ligand (Table 2, entries 10,11).
This effect almost disappeared when reaction (1) was
carried out in the more donor solvent, THF. In this
medium the reactions involving 2-methylbutyrate ligand
were found to be low sensitive to the ligand chirality (see
Table 2, entries 12,13).
2.4. Reaction mechanism
On the basis of the above-mentioned data some
mechanistic conclusions can be drawn. A sharp increase
3 The concentrations of stereoisomeric cyclohexene-3-ols were too
low for quantitative analysis by GC/MS.

280
N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
in the selectivity in respect to allylic oxidation of propy-
lene on going from acetic acid to low polar aprotic
solvents (Table 1) hardly can be understood within the
framework of the conventional mechanism (Scheme 7)
with 1,2-oxypalladation in a single stage (ꢀaꢁ),
which has been forwarded at the early studies of the
reaction [1,4,22,57].
According to this mechanism, the oxidation of
propylene should afford isopropenyl acetate, when the
stage ꢀaꢁ in Scheme 74 occurs via the Markovnikov-
addition followed by PdꢀC bond heterolysis (stage ꢀ
bꢁ in Scheme 7). In fact, this reaction product has
been repeatedly observed in AcOH solutions containing
0.1–0.5 mol l−1 additional OAc− ions [1–4,57]. (Note
that in the presence of anions with high elecronegativity
such as NO−3 , 1,2-glycol ester also can be obtained
[62]). When 1,2-oxypalladation occurs via anti-
Markovnikov-route, n-propenyl acetate, the second ob-
servable reaction product of propylene oxidation by
PdII in the AcOH/NaOAc system [1–4,57] should be
expected as shown in Scheme 8.
Meanwhile, the third reaction product, allyl acetate is
commonly formed in the same reaction system, and its
formation cannot be realized within the single-step
1,2-oxypalladation mechanism (Scheme 7).
An alternative mechanistic viewpoint suggests the
reaction proceeds via an intermediate p-allyl complex
[4,9,18,61a,b,63].
Scheme 7.
ion, which is necessary for the transformation of the
palladium h2-alkene to h3-allyl complex (stage ꢀaꢁ
in Scheme 9). Recently Grennberg and Ba¨ckvall [67]
has provided elegant evidence for the formation of the
p-allylpalladium intermediate by the example of reac-
tion between Pd(OAc)2 and 1,2-dideuterocyclohexene in
an AcOH solution. However, the way in which the
p-allyl complex undergoes redox disproportionation
into allyl acetate and Pd0 (Scheme 9, stage ꢀbꢁ) is
still not clear. Direct studies of the reactivity of PdII
p-allyl complexes toward nucleophilic and electrophilic
reagents [9,12,18,32,61] showed that such a dispropor-
tionation occurs in fact but is often a side reaction
accompanied with other transformations of the p-allyl
complex not leading to allylic esters.
At first glance this mechanism looks fairly attractive,
especially because the palladium p-allyl complexes are
well known since 1959, when the first stable complex
[p-C3H5·PdCl]2 has been synthesized independently by
the reactions of PdCl2 with allyl alcohol (Moiseev,
Fedorovskaya and Syrkin [65]) and allyl chloride
(Smidt and Hafner [66]). The involvement of p-allyl
complexes in the mechanism like that shown in Scheme
9 seems to be quite plausible for alkene oxidations in an
AcOH solution with high concentration of OAc− an-
In our experiments ‘free’ OAc− anions were in negli-
gible concentration, ([OAc−] 510−3 mol l−1), being
mostly bound into the complexes [Pd2(OCOR)6]2− and
[Pd(OAc)4]2−. Even minor amounts of OAc− anions
unbound with palladium(II) should exist as tight ionic
pairs with bulky PNP cations, being hardly accessible
for attacking the h2-alkene complex in a low polar
organic medium (CHCl3, CH2Cl2, THF).
In our opinion, an alternative explanation of data
obtained may be proposed by the mechanism, whose
key stage is an intramolecular electrophilic attack of the
PdII atom on the coordinated alkene molecule not
accompanied by the OAc− attack to form a PdII s-
complex (see Scheme 10). The electrophilic attack by
the PdII atom to h2-alkene ligand cannot be assisted
with simultaneous attack of the outer-sphere OAc−
anion (‘nucleophilic assistance’ [1,22,29]) because of the
deficiency of ‘free’ OAc− anions. Most likely that a
palladium s-complex with a carbenium moiety is
formed under these conditions (see Scheme 10, stage
ꢀaꢁ).
Scheme 6.
In the virtual absence of outer-sphere OAc− anions,
the coordinated carbenium group CH3CH+CH2ꢀPd is
expected to react intramolecularly, through a proton
elimination from the CH3 group, possibly assisted by
an additional Pd…H weak interaction (stage ꢀbꢁ in
Scheme 10).
4 The 1,2-oxypalladation stage may occur with the participation of
either coordinated OAc− group as cis-attack shown in Scheme 7 or
as trans-attack by free OAc− anion from a bulk solution [1,4,18,61].
In both cases the Markovnikov-route is preferable.

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
281
Scheme 8.
Our results for the oxidative esterification of cyclo-
hexene in low polar aprotic media (Table 2) agree well
with the mechanism shown in Scheme 10. Note that
homoallyl esters always form along with allylic esters.
Within the frameworks of the ‘allylic’ mechanism, this
fact needs the assumption that reaction (1) is accompa-
nied by the positional isomerization of the double bond
in the allylic esters first formed. However, such an
isomerization is known to be catalysed by Pd metal and
low-valence Pd complexes and is efficiently inhibited by
p-benzoquinone and other oxidants capable of re-oxi-
dizing Pd0 compounds [1,35–38,68,69]. Our experi-
ments showed that such oxidants in fact terminate the
oxidative dehydrogenation and disproportionation of
cyclohexene (Eqs. (15) and (16)), which are also
catalysed by Pd0, but have no effect on the formation
of homoallylic esters. Moreover, the assumed isomer-
ization could produce vinylic esters, which in fact were
absent from all experiments in aprotic media.
Scheme 10.
reaction regioselectivity. In the case when the original
salts (PNP)OCOR*P/ Pd3(OCOR*)6 contain only one
stereoisomer of the chiral carboxylate anion (for in-
stance, (+)-S-2-methylbutyrate, see Scheme 12), the
acido complexes [Pd(OCOR*)4]2− and [Pd2(OC-
OR*)6]2− that react with alkene are built from the
stereochemically identical anions R*COO−. Unlike
this, in experiments with the racemic anions (9)-
R*COO−, the alkene is to react with a set of stereo-
chemically different PdII carboxylate complexes, which
contain both (+)-R*COO− and (−)-R*COO− lig-
ands in various combinations, such as shown, for in-
stance, in structures 3 and 4.
All these data seem to support the mechanism (see
Scheme 10), which is similar to that in Scheme 10 and
includes an inner-sphere attack of the coordinated car-
boxylate anion RCOO− to the CH2 group of the
s-bonded carbenium ligand.
The mechanism in Scheme 11 assumes that the life-
time of the s-complex 1 is long enough to undergo
1,2-hydride shift and displacement of the carbenium
centre to adjacent C atom of the cyclohexane ring
before the redox transformation of the original s-com-
plex 1. The inner-sphere attack of the OAc− ligand to
the CH2 group of the original s-complex results in the
formation of allylic ester (path ꢀaꢁ), whereas a
similar process involving the isomerized s-complex 2
affords homoallylic ester (path ꢀbꢁ).
Our data on the regioselectivity of oxidative acetyla-
tion of hex-1-ene in a chloroform solution (see above)
also well agree with the ‘carbenium’ mechanism (see
Scheme 12).
The data on the oxidation of cyclohexene can be
suggested as an example of the stereocontrol of the
In the stages of p-complex formation, p-s-isomeriza-
tion and 1,2-hydride shift, the reactivity of such com-
plexes seem to be not much different. However, this is
not the case for the following stage of H+ abstraction
by the coordinated R*COO ligand, when the gain in
the competition between paths ꢀaꢁ and ꢀbꢁ
should depend on mutual arrangement of the acido
ligands controlled by their stereoconfiguration.
The above-mentioned discussion allows one to realize
the data for reaction (1) in weakly solvating organic
media like chlorocarbons, which only slightly affect the
inner-sphere complex transformations. In solvents with
higher nucleophilicity like THF, additional factors that
depend on the solvation should be taken into consider-
ation. In this case the H+ ion can be more likely
abstracted from the carbenium ion by the solvent
molecule, and the competition between allylic and
Scheme 9.

282
N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
tion. The findings [67,71] that addition of strong acids
like MeSO3H favours noticeably the homoallylic oxi-
dation of cyclohexene, seem to agree with the mecha-
nism shown in Scheme 11, because a strong acid
should prevent both 1,2-oxypalladation and the trans-
formation of h2-alkene to h3-allylic complex, favour-
ing the formation of s-bonded carbenium ligand.
3. Reactions with Pd nanoclusters
It was believed for a long time that the same mech-
anism is operative in all of the Pd-catalysed alkene
oxidations, including a vapor-phase process [3] in
which Pd catalyst is introduced into the reaction sys-
tem as the Pd metal black or Pd metal supported on
an inert carrier [17,73,74], due to the ease of the re-
oxidation of Pd0 to PdII via reaction (3)–(4).
However, the hypothesis that PdII is responsible for
the alkene oxidation under the conditions of indus-
trial process with heterogeneous Pd catalyst is in con-
tradiction to the data on the selectivity of oxidative
acetoxylation of propylene (see above).
Scheme 11.
The reaction between PdII and propylene in AcOH
medium produces acetone instead of isopropenyl or
allyl acetate due to competitive reaction (18), when
water content exceeds ꢂ1%:
AcOH+1%H O
2
CH3CHꢁCH2+PdII ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢂ CH3COCH3+Pd0
(18)
Meanwhile, when the Pd metal catalyst and O2 oxi-
dant are used, the allyl pathway of the propylene
acetoxylation predominates in both the liquid- and
vapour-phase reactions, and the allylic versus vinylic
Scheme 12.
homoallylic paths ꢀaꢁ and ꢀbꢁ is almost inde-
pendent of the ligand chirality.
It is noteworthy that the proposed ‘carbenium’
mechanism of alkene oxidation (Schemes 10 and 11)
is in fact closely related to the above-mentioned ‘al-
lylic’ mechanism by Ba¨ckvall and Grennberg [67] (see
Scheme 9). According to numerous NMR observa-
tions (see, for instance, [12,31,72]), h3-allylic com-
plexes readily isomerize to s-allylic complexes in
solutions, and this reaction is normally fast and re-
versible. Hence, Schemes 10 and 11 can be general-
ized by supplementing with the reversible p–s-
isomerization stages (Schemes 13 and 14), thus reflect-
ing the totality of data for the allylic pathway of
alkene oxidation by PdII.
Scheme 13.
Unlike allylic, the homoallyl path of alkene esterifi-
cation has attracted much less attention. Since the
first observations [21,70], the formation of homoallylic
esters was mostly treated as an unessential side reac-
Scheme 14.

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
283
selectivity is only slightly sensitive to a water content
up to 10% and the temperature [17,43,73,74]:
Finely dispersed palladium metal (for instance, as-
prepared Pd black obtained by reduction of PdCl2
or Pd3(OAc)6 with HCOONa or NaBH4) is in-
active in catalysis of reactions (Eqs. (19) and (20)) in
an AcOH solution containing alkaline acetates.
However, the Pd black acquired the catalytic activity,
if it was prepared by reduction of the PdII salts in
the presence of alkaline acetates [17,43,75]. Better cat-
alyst was obtained when the salt Pd3(OAc)6 was re-
duced by H2 in an AcOH solution containing
1,10-phenanthroline (phen) or 4,4%-dipyridine (dipy)
[68].
The phen (dipy)-containing catalysts were isolated
from the reaction solutions as X-ray amorphous
solids, whose nature were characterized by the TEM,
HREM, SAXS, EXAFS, elemental analysis and mag-
netic susceptibility data [68,69,79,80]. These studies
showed that the substance is a colloid-like nanocluster
with the chemical composition of Pd570930L6393
(OAc)190910 (L=phen or dipy). The giant cluster
0
[Pd ]
CH3CHꢁCH2+AcOH+1/2O2 “ CH2ꢁCHCH2OAc+H2O
(19)
Unlike this, allylic versus vinylic selectivity of both
stoichiometric and catalytic oxidation mediated by
PdII has been found to be highly sensitive to the
temperature, the content of acetate salts and water in
the acetic acid solvent: allyl to vinyl ratio varied from
0.05 to \10 depending on these variables
[43,57,68,75]. Only in aprotic solvents of lower polar-
ity such as CH2Cl2, CHCl3, is allyl acetate a prevail-
ing product [55,59].
A similar change in the selectivity was observed in
the oxidative acetoxylation of ethylene. When PdII
complexes such as Pd(OAc)24−, Pd2(OAc)62− were ap-
plied as the oxidants or catalysts, even small additives
(1–3%) of water to the NaOAc/AcOH solvent re-
sulted in a gain of the ‘aqueous’ pathway in the com-
petition between the AcO− and H2O nucleophiles,
producing acetaldehyde instead of vinyl acetate
[46,47,75]. Unlike this, heterogeneous Pd metal cata-
lysts provide vinyl acetate from ethylene and AcOH
with high selectivity, independently of the concentra-
tion of water additives within 0.2–10% [47,75,77] (Eq.
(20)).
molecule consists of
a positively charged metal
,
core of ꢂ25 A in diameter containing 570930
dense-packed (f.c.c. or icosahedral packing) Pd
atoms. The surface metal atoms are surrounded by
coordinated ꢂ63 phen or dipy ligands and
ꢂ
190 outer-sphere OAc− anions, and the latter coun-
terbalance the positive charge of the metal core. Av-
erage formal charge of Pd atoms in the cluster
molecule is equal to 190: 570= +1/3. Within the
framework of the P. Chini concept of cluster ‘magic
numbers’ [81], the structure of the giant cluster was
approximated by idealized formula Pd561L60(OAc)180
(5).
0
CH2ꢁCH2+1/2O2+AcOH[P“d ]CH2ꢁCHOAc+H2O
(20)
Differences in the reactivity of PdII carboxylates
and Pd metal toward alkylarenes were also found
[68,78]. In an AcOH solution of Pd(OAc)2 or
Pd(OCOCF3)2 in CF3COOH, toluene is subjected to
an electrophilic attack on the aromatic ring forming a
mixture of isomeric bis-tolyles, the products of oxida-
tive coupling (Eq. (21)):
The outer-sphere OAc− ligands in cluster 5 can be
readily replaced by another anions. For instance,
complete substitution of OAc− anions, accompanied
by hydrolysis, occurs on treatment of an aqueous so-
lution of 5 with KPF6, forming the cluster with the
idealized formula Pd561phen60O60(PF6)60 (6):
(21)
[Pd561phen60](OAc)180+60PF6− +60H2O
In the presence of high-dispersed palladium metal,
toluene is readily oxidized by O2 in an AcOH solu-
tion to afford benzyl acetate, the product of the oxi-
dative acetoxylation at the alkyl group, in the 95%
yield [47,68,78] (Eq. (22)):
=[Pd561phen60O60](PF6)60+120AcOH+60OAc−
(23)
The composition and structure of cluster 6 and its
analogues with other outer-sphere anions has been in-
vestigated by the same techniques as those used for
cluster 5 [68,80].
(22)
Giant clusters 5 and 6 were found to be efficient
catalysis for various reactions of organic substrates
under mild conditions (293–363 K, 0.1 MPa) [69,79].
For instance, an H atom of alkene molecules can be
readily replaced by the OAc group (oxidative acetoxy-
lation) in acetic acid solutions of clusters 5 and 6:
All these facts imply that the hydrocarbon activa-
tion comes from two different origins: one from PdII
and another from low-valence Pd clusters formed in
the PdII/Pd0 reaction system. This suggestion was
confirmed by experiments with Pd black and palla-
dium nanoclusters [17,68,69,75–78].

284
N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
ethylene is converted to vinyl acetate (Eq. (20)),
propylene to allyl acetate (Eq. (19)) and toluene to
benzyl acetate (Eq. (22)).
The selectivity of reactions (Eqs. (19) and (20))
with respect to the products of oxidative acetoxyla-
tion is at least 95–98%. No decrease in the selectivity
was observed even in an AcOH solution containing
up to 10% of water. The only side reaction observed
with these catalysts is a subsequent oxidation of
alkenyl and benzyl esters to form ethylidene, allyli-
dene and benzylidene diacetates, respectively [69].
Thus, all the aforesaid is in line with suggestion
that giant Pd clusters are the species responsible for
the catalysis of alkenes oxidation in these reaction
system and in fact, these giant clusters open a new
route to the oxidative acetoxylation of alkenes.
Detailed kinetic studies, including the H/D isotope
kinetic effects and inhibition kinetics [68,76,78] en-
abled us to formulate a general mechanism for the Pd
nanocluster-catalysed oxidation of alkenes [68,69,82]
(Scheme 15).
It is noteworthy that oxidative acetoxylation of
ethylene by O2 in the presence of clusters 5 and 6 as
well as Pd black (the latter prepared in an AcOH/
NaOAc solution) smoothly produced vinyl acetate.
Meanwhile, the propylene oxidation in the same reac-
tion systems produced no vinylic esters, and allyl ac-
etate was the sole reaction product at either low or
high concentrations of all reacting species, even when
benzoyl peroxide or Pd(OAc)2 were used as oxidants
instead of O2 [43,68].
Scheme 16.
Scheme 17.
Scheme 18.
ladium (100) surface of palladium crystal has been
obtained by HREELS technique [83].
In the case of propylene, the energy difference be-
tween the carbonꢀcarbon p-bond and the allyl-H
bond is much less than that between a carbonꢀcarbon
p-bond and the vinyl-H bond in ethylene. Thus, the
rate-determining step of the oxidative addition of
propylene molecule to the Pd cluster may include the
cleavage of the CꢀH bond in the coordinated propy-
lene molecule to form the surface-coordinated p-allyl
and hydride groups (see Scheme 17).
The formation of p-allyl group and coordinated hy-
dride in the intermediate cluster complex 8 appears to
favour the allylic route of the reaction in competition
with the vinylic one.
Subsequent reactions of the vinyl or allyl groups
and H atoms coordinated at the surface of the cluster
metal skeleton are assumed to proceed rapidly and
exert no influence on the reaction rate. It is unlikely
that all three molecules (alkene, O2, and AcOH) can
coordinate to the adjacent palladium atoms at the
surface of the cluster core. The alkene molecule is
more probably to bound at one site of the cluster
surface, while the O2 molecule is coordinated to an-
other site, not necessarily next to the site where the
alkene molecule is located. In this situation, electron
transfer from the Pdꢀalkenyl or PdꢀH surface group
to the coordinated O2 molecule can occur though the
metal atom chain, the latter acting as an ‘electron
transfer mediator’ [68,69] (see Scheme 15).
These facts suggested that the rate-determining step
is an oxidative addition (with opening the p-bond) of
a p-coordinated C2H4 molecule to a PdꢀPd group of
the
cluster
to
form
the
s,s-coordinated
…PdꢀCH2CH2ꢀPd… group. Subsequent splitting of
the CꢀH bond with the elimination of a PdꢀH frag-
ment in this group is assumed to be a fast step, fa-
vored by the PdꢁC multiple bond formation in the
intermediate 7 [68,69] (see Scheme 16).
Evidence for the formation of s,s-coordinated
ethylene and p,s-coordinated vinyl groups on the pal-
Scheme 15.

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
285
Thus, all the aforesaid is in line with suggestion
that giant Pd clusters are the species responsible for
the catalysis of alkene oxidation in these reaction sys-
tem and in fact, these giant clusters open a new route
to the oxidative acetoxylation of alkenes.
sively to allyl alcohol, acrolein, and acrylic acid
(Table 3):
C3H6+O2“CH2ꢁCHCH2OH“CH2ꢁCHCHO
“CH2ꢁCHCOOH
Pd cluster-catalysed oxidation of alkenes drastically
differs from the oxidation by PdII even in several
cases that have been observed in aqueous and alcohol
solutions. The following two examples are worthy of
notice.
Only traces of acetone were found in the reaction
products, in contrast to the reaction mediated by PdII
in which acetone was nearly a sole product [1,5,7].
In alcoholic (MeOH, EtOH) solutions acidified with
1.6 M H2SO4, allylic oxidation of propylene was also
found to occur, affording acrylic esters along with
acrolein and acrylic acid (Table 3). In parallel to the
main reactions, the oxidation of the alcohols to the
corresponding esters (MeOH to HCOOMe and EtOH
to CH3COOEt) took place (see below). In the absence
of added acid, the oxidation of alcohols dominated,
and the ethylene and propylene remained almost
unreacted.
(1) 1-Alkenes branched at the double bond
(isobutylene, 2-methylbut-1-ene, etc.) cannot be oxi-
dized by PdII in water via the Markovnikov 1,2-oxy-
palladation route (Scheme 7). Aldehydes are formed
only as trace by-products due to the anti-
Markovnikov route [1,18]. Meanwhile, such alkenes
are oxidizable by PdCl2 in aqueous solution affording
the allylic derivatives [7,63,64] (see Scheme 18)).
No PdII p-allyl complexes were found in this reac-
tive system, and the reaction mechanism was rather
obscure within the common oxidation schemes involv-
ing PdII [18]. On the other hand, the reactions in
Scheme 15 are quite explicable in the framework of
the cluster-catalysed mechanism (Scheme 16), where
low-valence Pd nanocluster functions the catalyst and
PdII is the oxidant.
(2) The oxidation of alkenes by palladium (II) is
known to be strongly suppressed by donor ligands,
even by such bulky as PPh3 and phen, due to strong
complexation with PdII, in contrast to the reactions
catalysed by giant Pd clusters, whose rates are unaf-
fected by these ligands [1,68]. It is also known that
the rates of alkene oxidations by PdII in aqueous so-
lution are inversely proportional to the H+ concen-
tration, and the reaction is quenched by addition of
1–2 M mineral acid [1,39]. Quite another phenomena
were observed when giant clusters 5 and 6 were em-
ployed as catalysts: the oxidation of alkenes did not
occur in neutral aqueous solutions. The cluster-
catalysed reaction started only after the addition of
strong acids (e.g., HClO4, H2SO4) [84], and its rate
was maximum at the acid concentration, which would
be sufficient to stop the reaction of alkenes with PdII.
The products of the alkene oxidations mediated by
PdII and the giant clusters are quite different. At 323
K and 0.1 MPa of 1:1 alkene: O2 gas mixture
ethylene was oxidized in 0.36–1.67 M aqueous solu-
tion of H2SO4 containing (2.3–4.6) 10−4 M of cluster
5 successively to acetaldehyde and acetic acid [84]:
4. Experimental
Solvents (n-pentane, benzene, diethyl ether, chloro-
form, dichloromethane, trichloroethane, acetic acid
and THF) of ‘reagent grade’ were purified by stan-
dard methods [85]. Cyclohexene and hex-1-ene of
‘reagent grade’ were distilled over sodium metal and
,
stored over molecular sieve 4 A. Propylene of ‘special
puruty grade’ in cylinder was purchased from
Moscow Petrol Refining factory. Trade carbon diox-
ide in cylinder was purified by re-condensation and
stored in an
1 l cylinder. p-Benzoquinone was
purified by vacuum sublimation. Palladium(II) chlo-
ride, sodium borohydride and isovaleric acid were of
‘chemically pure’ grade. (S)-(+)-2-methylbutyric and
(9)-2-methylbutyric acids, bis(triphenyl-phospho-
ranilidene)ammonium chloride (‘reagent grade’,
Fluka) were used. Palladium giant clusters with the
idealized
formula
Pd561phen60(OAc)180
and
Pd561phen60O60(PF6)60 were prepared according to the
published procedures [68]. (Heptapropoxy)perfluoro-
propionic flouroanhydride and (S)-(−)-1-phenylethy-
lamine of ‘reagent grade’ were provided by the
Kirovo–Chepetsk Chemical Industries (Russia) and
Zeeland Chemicals (USA), respectively.
4.1. Analysis methods
C2H4+O2“CH3CHO“CH3COOH
Element C,H,N-microanalysis of the complexes was
performed on an automatic C,H,N-analysator ‘Carlo-
Erba Strumentazione’, Italy. NMR spectra were
recorded on Bruker WP-200 spectrometer, IR spectra
were recorded on a spectrometer Specord M80 (Carl
Note, that acetaldehyde is scarcely oxidized by PdII
to acetic acid under these conditions [1,7,18]. A more
pronounced difference was observed for the cluster-
catalysed oxidation of propylene: in acidic aqueous
solution of cluster 5 propylene was converted succes-

286
N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
Table 3
Oxidation of propylene in acidic solutions of cluster 5 (50°C, 1 atm of C3H6: O2 (1:1) gas mixture, time of reaction 1 h) [84]
Solvent
H2O
C3H6:O2
1:1
[5] 10−4
M
[H2SO4] M
1.67
Initial rate of propylene consumption
10−3 mol (l min)−1
Reaction products (%)
2.34
4.04
5.25
CH2ꢁCHCH2OH (38)
CH2ꢁCHCHO (14)
CH2ꢁCHCOOH (20)
MeCOMe (4)
Me2CHOH (11) a
CH2ꢁCHCHO (4)
CH2ꢁCHCOOH (40)
CH2ꢁCHCOOMe (40)
MeCOMe (4)
MeOH
2:3
1.61
1.61
HCOOMe (2)
a Isopropyl alcohol formed due to the acid-catalysed propylene hydration, independently of the presence of Pd cluster.
Zeiss, Iena) in pellets with KBr. GC/MS analyses of the
4.4. (+)-(Heptapropoxy)perfluoropropionic acid
reaction products were performed on an Automass 150
instrument (Delsi Nermag, France, capillary columns
(+)-(Heptapropoxy)perfluoropropionic acid was ob-
with OV-1P/ a-Decs deposited phases), GLC analyses
tained by the separation of diastereoisomeric (S)-(−)-1-
were conducted on a chromatograph Varian 3600, USA,
phenylethylamides according to the known method
(universal capillary column with the OV-1 stationary
[89,90].
phase).
4.5. Palladium(II) acetate
4.2. X-Ray diffraction study of palladium
S(+)-2-methylbutyrate [48]
Palladium(II) acetate was obtained by the oxi-
dation of as-prepared Pd black (obtained from PdCl2
and NaBH4) by concentrated HNO3 in glacial AcOH by
the previously reported method [91] and puri-
fied from the traces of nitrates and nitrito complexes by
refluxing in glacial AcOH with as-prepared Pd
black.
Experimental reflection set was recorded on a four-cy-
cle automatic diffractometer Siemens R3/PC (uMo–Ka,
,
u=0.71074
A,
T= −120°C).
The
para-
meters of elementary cell were measured and refined
from 24 equivalent reflections with 2qB26–28°.
The structure was solved by direct method and refined
in full-matrix anisotropic approximation for all non-hy-
drogen atoms. The H atoms of the EtC*HMe groups
were generated geometrically and refined in a rider
model It was found that the C atoms of the EtC*HMe
groups have rather high temperature factors due to some
irregularity of these fragments. For this reason, refine-
ment was performed for the atoms in intermediate
positions with the multiplicity of 1. The correction for
absorption (m=13,35 cm−1) was introduced according
to the program [86]. Calculations were performed by the
program complex SHELXTL PLUS (PC version) [87]. The
cell parameters, atomic coordinates, bond lengths and
angles are given in Ref. [48].
4.6. Palladium(II) (S)-(+)-2-methylbutyrate
A total of 1 g (10 mmol) of (S)-(+)-2-methylbutyric
acid in 5 ml of benzene was added to a solution of 0.5
g of palladium(II) acetate (2.23 mmol) in 30 ml of
benzene and refluxed under mixing during 2 h. The dark
precipitate formed after cooling was filtered off, the
mother liquid was evaporated on a rotary evaporator
until a viscous orange oil was formed. The oil was
dissolved in pentane, the excess of the acid was washed
off with water (3–4 times by 200 ml), and the oil was
dried over CaCl2 and then over a molecular sieve. Excess
pentane was evaporated on a rotary evaporator, and the
residue was dried in vacuum. Yield 0.52 g (75% based on
Pd). According to GC/MS data, the substance obtained
as a viscous orange oil contains ꢂ1% of an admixture
of (S)-(+)-2-methylbutyric acid. The sample was used
in experiments on cyclohexene oxidation in the form of
solutions in chloroform or THF without additional
4.3. (9)-(Heptapropoxy)perfluoropropionic acid
(9)-(Heptapropoxy)perfluoropropionic acid was pre-
pared by hydrolysis of the corresponding fluoroan-
hydride in water followed by fractionation of the
aqueous solution by a modified method [88].
purification. IR spectrum (Vaseline oil), wCO cm−1
:

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
287
1
1637, 1425. H-NMR spectrum (CDCl3, l ppm, J Hz):
1H-NMR spectrum (CDCl3, l ppm): 1.95 (s, 3H,
3
3
0.62 (t, 3H, J=7.36); 0.83 (d, 3H, J=7.01); 1.41 (m,
OAc−); 7.3–7.7 (m, 30H, Ph).
2H); 2.2 (m, 1H).
4.11. Bis(triphenylphosphoranilidene)ammonium iso6aler-
ate
4.7. Palladium(II) (9)-2-methylbutyrate and iso6alerate
A suspension of ZnO (0.10 g, 1.23 mmol) in isova-
leric acid (0.25 g, 2.45 mmol) and water (0.3 ml) was
stirred at a room temperature for 24 h till complete
dissolution of ink oxide. The solution was added to a
saturated at ꢂ50°C aqueous solution of (Ph3P)2NCl
(0.705 g, 90% with respect to ink isovalerate). The
reaction mixture was heated at 60°C for 0.5 h to
form a pale rose oil, which was washed with water
(3–4 times by 50 ml) and extracted with chloroform.
Palladium(II) (9)-2-methylbutyrate and isovalerate
were prepared by the same method. Yields based on Pd
are 77–72%, respectively.
4.8. Palladium(II) (+)-(heptafluoropropoxy)-
perfluoropropionate
The acid (+)-CF3CF2CF2OCF(CF3)CO2H (1.63 g,
4.9 mmol, 10% excess) was added under stirring to a
solution of Pd(OAc)2 (0.5 g, 2.23 mmol) in 25 ml of
benzene, heated at 60°C for 2 h and filtered. The
residue was washed with diethyl ether, and the dark
brown solution was evaporated to dryness on a ro-
tary evaporator. The oil obtained was dried over
KOH during 3 days and purified by crystallization
from THF, precipitating the palladium carboxylate by
hexane. Yield 1.02 g (60% based on Pd). Found (%):
C, 19.01; Pd, 13.58. C12F22O6Pd. Calc. (%): C, 18.85;
Pd, 13.92. 19F-NMR spectrum (acetone d6, l ppm, J
Hz): −81.0 (t, 3F, CF3–CF2–, 3J=6.6); −82.1 (d,
3F, CF3–CFO–, 3J=2.5); −85.7 (m, 2F, –CF2–
CF3); −126.2 (m, 1F, CF3–CFO–);−129.3 (m, 2F,
–O–CF2–CF2–).
,
The organic layer was dried with molecular sieve 4A.
After filtration and evaporation of the solution to 1/3
of the volume, diethyl ether was added to the begin-
ning of precipitation, and the product was crystallized
in 0.5–1 h. The substance was recrystallized from a
chloroform–diethyl ether mixture. Yield 0.47 g (62%
based on (Ph3P)2NCl). Found (%): C, 76.42; H, 6.20;
N, 2.15. C41H39NO2P2. Calc. (%): C, 76.97; H, 6.15;
N, 2.19. IR spectrum (KBr), wCO cm−1: 1605, 1425.
1H-NMR spectrum (CDCl3, l ppm, J Hz): 0.65 (d,
3
6H, J=6.2); 1.25 (m, 1H); 2.02 (m, 2H); 7.68−7.4
(m, 41H).
4.12. Bis(triphenylphosphoranilidene)ammonium (+)-
(heptafluoropropoxy)perfluoropropionate
4.9. Palladium(II) (9)-(heptafluoropropoxy)-
perfluoropropionate
A suspension of ZnO (0.125 g, 1.52 mmol) and
(+)-CF3CF2CF2OCF(CF3)COOH (1.02 g, 3.07
mmol) was stirred at a room temperature till com-
plete dissolution of ZnO (ꢂ12 h). The reaction mix-
ture was added together with 5 ml of H2O to a warm
(ꢂ50°C) solution of (Ph3P)2NCl (1.57 g, 1 equiv.
with respect to a 90% yield of Zn(RFCOO)2). The
mixture was heated for 0.5 h at 60°C and then was
worked up by the same procedure as that for bis-
Palladium(II) (9)-(heptafluoropropoxy)perfluoro-
propionate was prepared by the same method. Yield
1.1 g (65% based on Pd).
4.10. Bis(triphenylphosphoranilidene)ammonium acetate
Bis(triphenylphosphoranilidene)ammonium acetate
was prepared by the modified method [53]. To a
solution of (Ph3P)2NCl (1 g, 1.75 mmol) in 40 ml
of CHCl3 a solution of AgOAc (0.44 g, 2.62 mmol,
20% excess) in 30 ml of CHCl3 was added. The reac-
tion mixture was stirred during 1 h and the AgCl
precipitated was filtered off. The solution was evapo-
rated on a rotary evaporator to a volume of 10 ml
and added diethyl ether to the beginning if cry-
stallization. The main amount of PNP acetate was
first precipitated as a colorless oil, which crystal-
lizes on storing in a fridge for 10–12 h. The sub-
stance was purified by recrystallization from CHCl3–
(triphenylphosphoranilidene)ammonium
isovalerate
(see above). Yield 0.93 g (85% based on (Ph3P)2NCl).
Found (%): C, 58.56; H, 3.5; N, 1.73.
C42H30NF11O3P2. Calc. (%): C, 58.14; H, 3.49; N,
1.61. 19F-NMR spectrum (CDCl3, l ppm, J Hz): −
84.3 (t, 3F, CF3–CF2–, 3J=7.43); −84.7 (d, 3F,
CF3–CFO–, 3J=3.3); −86.0 (m, 2F, –CF2–CF3);
−126.8 (m, 1F, CF3–CFO–); −132.7 (m, 2F, –O–
CF2–CF2–).
4.13. Bis(triphenylphosphoranilidene)ammonium (S)-
(+)-2-methylbutyrate
diethyl ether. Yield 0.94
(Ph3P)2NCl). Found (%): C, 76.43; H, 5.61; N, 2.37.
C39H33NO2P2. Calc. (%): C, 76.37; H, 5.57; N, 2.34.
g
(90% based on
A mixture of (S)-(+)-2-methylbutyric acid (0.2 g,
1.96 mmol), NaOH (0.078 g, 1.96 mmol) and water
(0.5 ml) was stirred till complete dissolution of

288
N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
NaOH. To the solution was added
a
saturated
spheric with argon. The gaseous and liquid reaction
aqueous solution of AgO3 (0.33 g, 1.95 mmol) aci-
dified with five drops of (S)-(+)-2-methylbutyric
acid. The Ag[EtC*H(Me)]COO precipitated was
filtered off, washed with water, methanol and hexane
and dried in a vacuum dessicator over alkali for 24 h.
To a solution of (Ph3P)2NCl (0.907 g, 1.56 mmol) in
chloroform was added a chloroform solution of the
silver carboxylate. The AgCl precipitated was filtered
off, the mother liquid was evaporated on a rotary
evaporator till 1/3 of the initial volume, and added
diethyl ether to the beginning of precipitation. After
0.5–1 h, the crystalline precipitate was filtered off and
recrystallized from a chloroform–diethyl ether mix-
ture. Yield 0.82 g (81% based on (Ph3P)2NCl). Found
(%): C, 76.63; H, 6.19; N, 2.17. C41H39NO2P2. Calc.
(%): C, 76.97; H, 6.15; N, 2.19. IR spectrum (KBr),
wCO cm−1: 1616, 1425. 1H-NMR spectrum (CDCl3,
l ppm, J Hz): 0.83 (t, 3H, 3J=7.35); 1.15 (d, 3H,
3J=7.1); 1.75 (m, 2H); 2.35 (m, 1H); 7.68–7.4 (m,
30H).
products were analyzed by GLC and GC/MS.
4.15.2. Procedure B (liquefied CO2 sol6ent)
The procedure was similar to the previous one,
with the difference connected with the use of liquefied
CO2. After loading the solid and liquid reaction com-
ponents into the autoclave (and additionally 3 glass
balls of 3 mm diameter for more efficient mixing), the
autoclave was cooled with liquid nitrogen. Propylene
(72 ml) and CO2 (540 ml), both dosed from a gas
volumeter at 1 atm and room temperature, were se-
quentially condensed into the autoclave from a vol-
umeter, the valve was closed, and the reaction
mixture was stored in a water thermostat for 3 h at
25°C (liquid CO2) or 45°C (supercritical (fluid) CO2)
under periodical manual shaking. The total pressure
was measured by a pressure gauge at 25°C; the par-
tial pressures and the state of CO2 (liquid or fluid) at
the reaction temperature were calculated from the
PVT data [92]. When the reaction ceased, the auto-
clave was cooled by liquid nitrogen and the valve was
open to an evacuated (0.1 Torr) round-bottom 2.5 l
flask connected with a manometer. The autoclave was
slowly heated to a room temperature, the gas pres-
sure was measured, then adjusted to atmospheric with
argon and the reaction products were analyzed as
shown above. The results are shown in Table 1.
4.14. Bis(triphenylphosphoranilidene)ammonium (S)-
(9)-2-methylbutyrate
Bis(triphenylphosphoranilidene)ammonium (S)-(9)-
2-methylbutyrate was prepared by the same procedure
as bis(triphenylphosphoranilidene)ammonium (S)-
(+)-2-methylbutyrate (see above). Yield 0.80 g (79%
based on (Ph3P)2NCl). Found (%): C, 76.57; H, 6.21;
N, 2.22. C41H39NO2P2. Calculated (%): C, 76.97; H,
6.15; N, 2.19. IR spectrum (KBr), wCO cm−1: 1615,
4.16. Oxidation of cyclohexene and hex-1-ene by palla-
dium(II) carboxylates [59]
1
1425. H-NMR spectrum (CDCl3, l ppm, J Hz): 0.85
(t, 3H, 3J=7.37); 1.1 (d, 3H, 3J=7.09); 1.77 (m,
PNP carboxylate (0.042 mmol) and p-benzoquinone
(0.223 mmol) were placed into a two-necked flask.
The flask was twice evacuated and filled with argon.
Palladium carboxylate (0.111 mmol) dissolved in 1.5–
2.0 ml of a solvent and alkene (0.223 mmol) were
added in an argon flow. The flask was carefully evac-
uated with cooling and filled with argon. The reaction
mixture was stirred for 10–12 h, and the reaction
products were analyzed by GLC and GC/MS. The
results are shown in Table 2.
2H); 2.38 (m, 1H); 7.66–7.5 (m, 30H).
4.15. Oxidation of propylene by palladium(II) acetate
complexes [55]
4.15.1. Procedure A (liquid sol6ents)
The reaction was conducted in a small stainless
steel autoclave of 7.2 ml free volume. PNP acetate
(0.042 mmol), nitrosobenzene or p-benzoquinone
(0.223 mmol) and Pd(OAc)2 (0.111 mmol) dissolved
in 1.5 ml of a solvent (chloroform, dichloromethane,
AcOH, etc.) were placed into the autoclave and evac-
uated (5 Torr) under cooling to 0°C. The reaction
vessel was heated to 25°C (water bath) and filled with
propylene (6 ml at 25°C and 1 atm (manometric
monitoring), ꢂ0.25 mmol). The reaction mixture was
stored in a water thermostat for 3 h at 25°C under
periodical manual shaking. Then the autoclave valve
was open to an evacuated (0.1 Torr) 10 ml round-
bottom flask connected with a manometer, the gas
pressure was measured and then adjusted to atmo-
4.17. Oxidation of propylene by O2 in acidic (aqueous
or methanolic) solutions of giant Pd cluster 5 [84]
The reaction was conducted and the products were
worked up similarly to the previous procedure using
cluster 5 (4 mg, corresponding to 0.027 mmol of Pd)
as the catalyst. A propylene–O2 (1:1) gas mixture was
introduced in the reaction flask thermostated at 50°C
from a volumetric burette at 1 atm during 1 h. The
results are shown in Table 3.

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
289
5. Conclusions
why p-allylic complexes have been observed by NMR
in the reaction solutions during formation of allylic
esters (see for example, Ref. [67]). Meanwhile, a role of
the s- and p-allylic species in the mechanism of reduc-
tive elimination needs additional studies.
Under the conditions of asynchronous electrophilic
and nucleophilic attacks in low-polar solvents, carboxy-
late anions exhibit lower nucleophilicity toward PdII-co-
ordinated alkene molecules as compared to those of
H2O and ROH molecules in aqueous and alcohol solu-
tions, correspondingly. Some other explanation should
be invoked to account for allylic oxidations in these
media.
Our results showed that alkene oxidation by PdII
carboxylate complexes in aprotic media most likely
does not involve 1,2-oxymetallation of the p-coordi-
nated alkene. Instead of this, two key stages seem to
occur in succession:
(1) isomerization of the alkene–PdII p-complex to a
s-pallado-substituted carbenium ion:
Palladium chemistry makes available such an alterna-
tive way for allylic oxidation. New reaction channel is
opened when low-valence palladium clusters are used as
the catalysts for alkene oxidation [68,69,82]. This path-
way provide advances toward better selectivity and
performance under milder conditions.
A shift of the Pd atom toward one of the alkene C
atoms is separated in time from a nucleophilic attack
on the carbenium centre formed at the adjacent C
atom;
(2) the attack of a nucleophilic reagent on the C+
atom of the s-carbenium complex:
In many cases a Pd cluster is in fact a precursor of an
actual catalyst, but experimental data on the transfor-
mations of starting forms of cluster catalysts are still
scare. On the other hand, when starting from a
mononuclear PdII complex, palladium cluster or clus-
ter-like material (e.g., giant Pd cluster or colloidal
palladium) can form in the course of reaction, being the
actual species responsible for the catalytic reaction.
Speculations based on the idea of mechanisms in-
volving the oxidative addition of a substrate to a cluster
molecule with the cleavage of the metal–metal bond
can be used to rationalize the experimental data avail-
able. In catalysis of redox reactions, metal clusters can
serve a function of an electron transfer mediator.
Either inner- or outer-sphere RCOO− anion may be
an attacking group at this stage. Trans-attack by free
OAc− anion could be expected in the AcOH/NaOAc
medium. (Noteworthy, a synchronous 1,2-oxypallada-
tion of ethylene in aqueous solution has been found to
proceed as a trans-attack [93]). When free carboxylate
anions are deficient, this stage may occur as a cis-attack
by the coordinated RCOO− ligand.
The asynchronism of electrophilic and nucleophilic
attacks provides a means for allylic and homoallylic
esterification. If the s-carbenium complex is a fairly
long-lived species, it has time to undergo a reversible
1,2-hydride shift (see Schemes 10 and 11) before the
nucleophilic attack to occur. The nucleophilic attack
might result in either addition of RCOO− group to the
carbenium centre or proton abstraction from the adja-
cent C atom. The absence of vinylic esters in the
products of oxidation of propylene, hex-1-ene and cyclo-
hexene gives evidence that proton abstraction to form
the palladium s-allylic or s-homoallylic complex fol-
lowed by the reductive elimination of Pd0 and RCOO−
anion is in fact the main reaction path, which affords
allylic and homoallylic esters.
This mechanistic conception allows one to eliminate
the controversy as to whether p-allylic complexes par-
ticipate in the allylic oxidation by PdII or not. As seen
in Schemes 10 and 11, the reactions under question
involve s-allylic intermediates, which are commonly
subjected to fast s X p-isomerization. This can explain
Acknowledgements
The work was financially supported by the Russian
Foundation for Basic Research (projects Nos. 95-03-
10746, 96-03-33581, 96-15-97577 and 99-03-32291), the
Program of the Government of the Russian Federation
for support the leading Russian scientific schools (grant
96-15-97577), and the program Integration of the Min-
istry of Science and Technologies of the Russian Feder-
ation (grant S/O 225-2.1). The Zeeland Chemicals, Inc.
(USA) and the Kirovo–Chepetsk Chemical Industries
(Russia) are gratefully acknowledged for kind granting
L
-(−)-a-phenylethylamine and (9)-2-(heptafluoropro-
poxy)-perflouropropionic acid fluoride, respectively.
Authors are grateful to S.E. Nefedov and I.L. Ere-
menko for X-ray structural study of complex 1, S.G.
Sakharov for recording and interpreting NMR spectra
and I.P. Stolarov, A.E. Gekhman, M.V. Martens, A.A.
Bukharkina and O.V. Lykova for help in experiments.

290
N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
[30] A.P. Belov, M.N. Vargaftik, I.I. Moiseev, Izv. Akad. Nauk
References
SSSR. Ser. Khim. (Russ. Chem. Bull.) (1965) 1521 (in Russ.).
[31] I.I. Moiseev, A.P. Belov, G. Yu. Pek, Zh. Neorg. Khim. (Russ.
Inorg. Chem.) 10 (1965) 336 (in Russ).
[32] A.P. Belov, I.I. Moiseev, N.G. Satsko, Ya. K. Syrkin, Izv. Akad.
Nauk SSSR. Ser. Khim. (1971) 265 [Russ. Chem. Bull. 20 (1971)
209 (Engl. Transl.)].
[33] R. Jira, J. Sedlmeier, J. Smidt, Justus Liebigs Ann. Chem. 693
(1966) 99.
[34] I.I. Moiseev, M.N. Vargaftik, S.V. Pestrikov, O.G. Levanda,
Doklady Akad. Nauk SSSR (Russ. Reports Acad. Sci.) 171
(1966) 1365 (in Russ.).
[35] S.V. Pestrikov, I.I. Moiseev, Izv.Akad.Nauk SSSR. Ser.Khim.
(1965) 349[Russ. Chem. Bull. (1965) 324 (Engl. Transl.)].
[36] S.V. Pestrikov, I.I. Moiseev, T.N.Romanova, Zh. Neorg. Khim.
(Russ. Inorg. Chem.) 10 (1965) 2203 (in Russ).
[37] S.V. Pestrikov, I.I. Moiseev, B.A. Tsvilikhovskaya, Zh. Neorg.
Khim. (Russ. Inorg. Chem.) 11 (1966) 1742 (in Russ).
[38] S.V. Pestrikov, I.I. Moiseev, L.M. Sverzh, Zh. Neorg. Khim.
(Russ. Inorg. Chem.) 11 (1966) 2081 (in Russ).
[1] I.I. Moiseev, p-Complexes in Liquid-Phase Olefin Oxidation,
Nauka, Moscow, 1970 (in Russ.).
[2] J. Tsuji, Palladium Reagents and Catalysts, Wiley, Chichester,
1995.
[3] R. Jira, in: B. Cornils, W.A. Herrmann (Eds.), Applied Homoge-
neous Catalysis with Organometallic Compounds, vol. 1, VCH,
Weinheim, 1996, p. 374.
[4] H. Grennberg, J.-E. Ba¨ckvall, in: M. Beller, C. Bolm (Eds.),
Transition Metals for Organic Synthesis, Wiley–VCH, Wein-
heim, 1998, p. 200.
[5] I.I. Moiseev, M.N. Vargaftik, Ya. K. Syrkin, Doklady Akad.
Nauk SSSR (Russ. Doklady Chemistry) 130 (1960) 820; ibid. 133
(1960) 377 (in Russ.).
[6] M.S. Kharash, R.C. Seyler, F.R. Mayo, J. Am. Chem. Soc. 60
(1938) 882.
[7] J. Smidt, W. Hafner, R. Jira, I. Sedlmeier, R. Ro¨ttinger, H.
Kojer, Angew.Chem. 71 (1959) 176.
[8] M.N. Vargaftik, I.I. Moiseev, Ya. K. Syrkin, V.V. Yakshin, Izv.
Akad. Nauk SSSR. Ser. Khim. (Russ. Chem. Bull.) (1962) 930
(in Russ.).
[9] I.I. Moiseev, A.P. Belov, Ya. K. Syrkin, Izv. Akad. Nauk SSSR.
Ser. Khim. (Russ. Chem. Bull.) (1963) 1527 (in Russ.).
[10] E.W. Stern, Proc. Chem. Soc. (London) (1963) 111.
[11] E.W. Stern, M.L.Spector, Proc. Chem. Soc. (London) (1961)
370.
[12] A.P. Belov, G. Yu. Pek, I.I. Moiseev Izv. Akad. Nauk SSSR.
Ser. Khim. (Russ. Chem. Bull.) (1965) 2204 (in Russ.) [Russ.
Chem. Bull., Engl. Ed. (1965) 2170].
[13] W. Kitching, Z. Rappoport, S. Winstein, W.G. Yong, J. Am.
Chem. Soc. 88 (1966) 2054.
[14] (a) K. Bergstad, Ph. D. Thesis, University of Uppsala, 1998, p.
13. (b) H. Grennberg, K.Bergstad, J.-E. Ba¨ckvall, J. Mol. Catal.
A 113 (1996) 355.
[15] P.M. Henry, J. Org. Chem. 32 (1967) 2575.
[16] R.G. Brown, J.M. Davidson, J. Chem. Soc., Chem. Commun.
(1972) 642.
[17] S.F. Politanski, M.N. Vargaftik, A.M. Shkitov, I.P. Stolarov, I.I.
Moiseev, O.M. Nefedov, Izv. Akad. Nauk. SSSR. Ser. Khim.
(Russ. Chem. Bull.) (1979) 1913 (in Russ.).
[18] R. Jira, in: E. Becker, M. Tsutsui (Eds.), Organometallic Reac-
tions, vol. 3, Wiley, New York, 1972, p. 44.
[39] I.I. Moiseev, O.G. Levanda, M.N. Vargaftik, J. Am. Chem. Soc.
96 (1974) 1003.
[40] R.F. Heck, J. Am. Chem. Soc. 90 (1968) 5518, 5526, 5531, 5535,
5538, 5542.
[41] I.I. Moiseev, M.N. Vargaftik, Doklady Akad. Nauk SSSR
(Russ. Reports Acad. Sci.) 166 (1966) 370 (in Russ.).
[42] (a) J.E. McMurry, P.Kocovsky, Tetrahedron Lett. 25 (1984)
,
4187. (b) B. Akemark, E.M. Larsson, J.D. Oslob, J. Org. Chem.
59 (1994) 5729.
[43] I.P. Stolarov, M.N. Vargaftik, O.M. Nefedov, I.I. Moiseev, Izv.
Akad. Nauk SSSR. Ser. Khim. (1983) 1455 [Russ. Chem. Bull.
32 (1983) 1327 (Engl. Transl.)].
[44] S. Nakamura, T. Yasui, J. Catal. 23 (1971) 315.
[45] S. Winstein, J. McCaskie, H.-B. Lee, P.M. Henry, J. Am.
Chem.Soc. 98 (1976) 6913.
[46] I.I. Moiseev, in: M.E. Vol’pin, Chemistry Reviews (Soviet Scien-
tific Reviews, Section B), Vol. 2, Harwood, New York, 1982, p.
951.
[47] I.I. Moiseev, in: A.E. Shilov (Ed.), Fundamental research in
Homogeneous Catalysis, Gordon and Breach, New York, 1986,
p. 12.
[48] N.Yu. Kozitsyna, M.V. Martens, I.P. Stolarov, S.E. Nefedov,
M.N. Vargaftik, I.L. Eremenko, I.I. Moiseev, Zh. Neorg. Khim.
44 (1999) 1915 [(Russ. Inorg. Chem. (1999), 11 (Engl. Transl.)].
[49] F.A. Cotton, S. Han, Rev. Chim. Miner. 20 (1983) 496; 22
(1985) 277.
[50] A.S. Batsanov, G.A. Timko, Yu. T. Struchkov, N.V. Gerbeleu,
K.M. Indrichan, G.A. Popovich, Izv.Akad.Nauk SSSR.
Ser.Khim. (1987) 697 [Russ. Chem. Bull. 36 (1987) 636 (Engl.
Transl.)].
[51] N.N. Lyalina, S.V. Dargina, A.N. Sobolev, T.M. Buslaeva, I.P.
Romm, Koordinat. Khimiya 19 (1993) 57 [Russ. Coordin.
Chem. 19 (1993) 50 (Engl. Transl.)].
[52] D.P. Bancroft, F.A. Cotton, L.R. Favello, W. Schwotzer, Poly-
hedron 7 (1988) 615.
[53] A. Martinsen, J. Songstadt, Acta Chem. Scand. A 31 (1977) 645.
[54] M.N. Vargaftik, Yu. T. Struchkov, A.I. Yanovsky, P.M. Maitlis,
Mendeleev Commun. (1993) 247.
[55] N. Yu. Kozitsyna, M.N. Vargaftik, A.E. Gekhman, I.I. Moiseev,
Doklady Akad. Nauk. 346 (1996) 486 [(Doklady Chemistry 346
(1996) 29 (Engl. Transl.)].
[56] (a) R.J. Pandey, P.M. Henry, Canad. J. Chem. 52 (1974) 1241;
53 (1975) 1833. (b) D.D. Kragten, Ph. D. Thesis, Eindhoven,
Technische Universiteit Eindhoven, 1999, p. 33; (c) D. Kragten,
R.A. van Santen, M.K. Crawford, W.D. Provine, J.J. Lerou,
Inorg. Chem. 38 (1999) 331.
[19] C.B. Anderson, S. Winstein, J. Org. Chem. 28 (1963) 605.
[20] M. Green, R.N. Haszeldin, J. Lindley, J.Organomet. Chem. 6
(1966) 107.
[21] R.G. Brown, J.M. Davidson, J. Chem. Soc. A (1971) 1321.
[22] P.M. Henry, Acc. Chem. Res. 6 (1973) 16.
[23] R.G. Brown, R.V. Chaudhari, J.M. Davidson, J. Chem. Soc. D
(1977) 183.
,
[24] S. Hansson, A. Heumann, T. Rein, B. Akemark, J. Org. Chem.
55 (1990) 975.
[25] M.N. Vargaftik, I.I. Moiseev, Ya. K. Syrkin, Doklady Akad.
Nauk SSSR (Russ. Reports Acad. Sci.) 139 (1961) 1396 (in
Russ.).
[26] M.N. Vargaftik, I.I. Moiseev, Ya. K. Syrkin, Doklady Akad.
Nauk SSSR (Russ. Reports Acad. Sci.) 147 (1962) 399 (in
Russ.).
[27] I.I. Moiseev, M.N. Vargaftik, Ya. K. Syrkin, Doklady Akad.
Nauk SSSR (Russ. Reports Acad. Sci.) 153 (1963) 140 (in
Russ.).
[28] M.N. Vargaftik, I.I. Moiseev, Ya. K. Syrkin, Doklady Akad.
Nauk SSSR (Russ. Reports Acad. Sci.) 152 (1963) 147 (in
Russ.).
[29] P.M. Henry, J. Am. Chem. Soc., 86 (1964) 3246; ibid. 88 (1966)
1595.
[57] P.M. Henry, Palladium-Catalyzed Oxidation of Hydrocarbons,
D. Reidel, Dordrecht, 1980, p. 103.

N.Yu. Kozitsyna et al. / Journal of Organometallic Chemistry 593–594 (2000) 274–291
291
[76] I.P. Stolarov, M.N. Vargaftik, I.I. Moiseev, Kinet.Katal. (Russ.
Kinetics and Catalysis) (1987) 1359 (in Russ.).
[77] M.K. Starchevski, M.N. Vargaftik, I.I. Moiseev, Izv.Akad.Nauk
SSSR. Ser.Khim. (Russ. Chem. Bull.) (1979) 242 (in Russ.).
[78] V.P. Zagorodnikov, M.N. Vargaftik, Kinet.Katal. (Russ. Kinet-
ics and Catalysis) 27 (1986) 851 (in Russ.).
[79] M.N. Vargaftik, V.P. Zagorodnikov, I.P. Stolarov, I.I. Moiseev,
V.A. Likholobov, D.I. Kochubey, A.L. Chuvilin, V.I.
Zaikovsky, K.I. Zamaraev, G.I. Timofeeva, J. Chem. Soc.,
Chem. Commun. (1985) 937.
[80] M.N. Vargaftik, I.I. Moiseev, D.I. Kochubey, K.I. Zamaraev,
Faraday Discuss. 92 (1991) 13.
[58] N.Yu. Kozitsyna, A.A. Bukharkina, unpublished results.
[59] N.Yu. Kozitsyna, M.V. Martens, I.P. Stolarov, M.N. Vargaftik,
A.E. Gekhman, I.I. Moiseev, Izv.Akad.Nauk. Ser.Khim. (1999)
9, in press [Russ. Chem. Bull. (1999), 9 (Engl. Transl.)].
[60] R.S. Atkinson, Stereoselective Synthesis, Wiley, Chichester,
1995, p. 79.
[61] J.E. Ba¨ckvall, Advances in Metal-Organic Chemistry, JAI Press,
Greenwich, 1989, p. 135.
[62] M.G. Volkonsky, V.A. Likholobov, Yu. I. Ermakov, Kinet.
Katal. (Russ. Kinetics and Catalysis) 24 (1983), 347, 578 (in
Russ.).
[63] R. Hu¨ttel, H.Christ, Chem. Ber. 96 (1963) 3101; ibid., 97 (1964)
1439.
[64] R. Jira, Tetrahedron Lett. (1971) 1225.
[65] I.I. Moiseev, E.A. Fedorovskaya, Ya. K. Syrkin, Zh. Neorg.
Khim. (Russ. Inorg. Chem.) 4 (1959) 2641 (in Russ).
[66] J. Smidt, W. Hafner, Angew. Chem. 71 (1959) 284.
[67] H. Grennberg, J.-E. Ba¨ckvall, Chem. Eur. J. 4 (1998) 1083.
[68] M.N. Vargaftik, V.P. Zagorodnikov, I.P. Stolarov, I.I. Moiseev,
D.I. Kochubey, V.A. Likholobov, A.L. Chuvilin, K.I. Zama-
raev, J. Mol. Catal. 53 (1989) 315.
[81] (a) P. Chini, Gazz. Chim. Ital. 109 (1979) 225. (b) P. Chini, J.
Organometal. Chem. 200 (1980) 37.
[82] I.I. Moiseev, M.N. Vargaftik, in: J.A. Thomas, K.I. Zamaraev
(Eds.), Perspectives in Catalysis, Blackwell, Oxford, 1992, p. 91.
[83] E.M. Stuve, R.J. Madix, J. Phys. Chem. 89 (1985) 105.
[84] P.I. Pasichnyk, M.K. Starchevsky, Yu. A. Pazdersky, V.P. Za-
gorodnikov, M.N. Vargaftik, I.I. Moiseev, Mendeleev Commun.
(1994) 1.
[85] D.D. Perrin, W.L.F. Armarego, Purification of Laboratory
Chemicals, 3rd, Pergamon, Oxford, 1988.
[86] N. Walker, D. Stuart, Acta Crystallogr. 39 (1983) 158.
[87] G.M. Sheldrick, Crystallographic Computering 3: Data Collec-
tion, Structure Determination, Proteins, and Databases, New
York, 1985, p. 175.
[88] USA Patent No. 3.250.808 (cl. 260-535), 1966; Chem. Abstrs., 65
(1966), 13554b.
[89] H. Kawa, N. Ishikawa, Chem. Lett. (1980) 843.
[90] .H. Kawa, F. Yamaguchi, N. Ishikawa, Chem. Lett. (1982) 745.
[91] T.A. Stephenson, S.M. Morehouse, A.R. Powell, J.P. Heffer, G.
Wilkinson, J. Chem. Soc. (1965) 3632.
[92] N.B. Vargaftik, Handbook of physical properties of liquids and
gases. Pure substances and mixtures, 2nd edn, Hemisphere,
Washington, 1975, pp. 167–313.
[69] I.I. Moiseev, M.N. Vargaftik, in: R.D. Adams, F.A. Cotton
(Eds.), Catalysis by Di- and Polynuclear Metal Complexes,
Wiley–VCH, New York, 1998, p. 395.
[70] S. Wolfe, P.G.C. Campbell, J. Am. Chem. Soc. 93 (1971) 1497–
1499.
,
[71] B. Akemark, S. Hanson, T. Rein, J. Va˚gberg, A. Heumann, J.E.
Ba¨kvall, J. Organomet. Chem. 369 (1985) 433.
[72] P.M. Maitlis, The Organic Chemistry of Palladium, Academic
Press, New York, 1971, p. V1.
[73] K. Fujimoto, T. Kunugi, J. Jap. Petrol. Inst. 17 (1974) 739.
[74] K. Fujimoto, H. Takeda, Sh. Yokoyama, T. Kunugi, Nippon
Kagaku Kaishi (1974) 1599.
[75] M.N. Vargaftik, V.P. Zagorodnikov, I.I. Moiseev, Kinet.Katal.
(Russ. Kinetics and Catalysis) 22 (1981) 951(in Russ.).
,
[93] J.E. Ba¨ckvall, B. Akemark, S.O. Ljunggren, J. Am. Chem. Soc.
101 (1979) 2411.
.
.