Bonding modes in palladium(II) enolates: Consequences for dynamic behavior and reactivity

Article abstract of DOI:10.1021/om990613o

The behavior of palladium C-bound enolates [Pd(CH₂C(O)CR₃)Cl(PPh₃)₂] (R = H, 1; R = Me, 2) and [Pd(CH₂C(O)CR₃)(PPh₃) ₂(NCMe)](BF₄) (R = H, 5; R = Me, 6) has been studied. Dimeric species with bridging enolate moieties are formed in solution when a coordination site on the metal is made available, either with Pd₂{μ-κ²-C,O-CH₂C(O)CR ₃}₂ or with mixed Pd₂{μ-κ²-C,O-CH₂C(O)CR ₃}(μ-X) (X = Cl, OH) bridges. It is proposed that π back-donation is important to stabilize oxygen bonding. Complexes 1 and 2 undergo exchange between free and coordinated phosphine in solution. Kinetic experiments support an intramolecular associative mechanism which could involve an oxoallyl-like transition state. The reactivity of the complexes has been explored. Some reactions typical of Pd-alkyls have been observed such as insertion of CO to give CR₃C(O)CH₂COOH. Electrophilic attack on oxygen is very important: the hydrolysis of the enolate complexes has been studied and also the reaction with ClSiMe₃ to give silyl enol ethers.

Full text of DOI:10.1021/om990613o

Organometallics 1999, 18, 5571-5576
5571
Bon d in g Mod es in P a lla d iu m (II) En ola tes: Con sequ en ces
for Dyn a m ic Beh a vior a n d Rea ctivity
Ana C. Albe´niz, N. Marta Catalina, Pablo Espinet,* and Roc´ıo Redo´n
Departamento de Quı´mica Inorga´nica, Facultad de Ciencias, Universidad de Valladolid,
Prado de la Magdalena s/ n, 47005 Valladolid, Spain
Received August 2, 1999
The behavior of palladium C-bound enolates [Pd(CH2C(O)CR3)Cl(PPh3)2] (R ) H, 1; R )
Me, 2) and [Pd(CH2C(O)CR3)(PPh3)2(NCMe)](BF4) (R ) H, 5; R ) Me, 6) has been studied.
Dimeric species with bridging enolate moieties are formed in solution when a coordination
site on the metal is made available, either with Pd2{µ-κ2-C,O-CH2C(O)CR3}2 or with mixed
Pd2{µ-κ2-C,O-CH2C(O)CR3}(µ-X) (X ) Cl, OH) bridges. It is proposed that π back-donation
is important to stabilize oxygen bonding. Complexes 1 and 2 undergo exchange between
free and coordinated phosphine in solution. Kinetic experiments support an intramolecular
associative mechanism which could involve an oxoallyl-like transition state. The reactivity
of the complexes has been explored. Some reactions typical of Pd-alkyls have been observed
such as insertion of CO to give CR3C(O)CH2COOH. Electrophilic attack on oxygen is very
important: the hydrolysis of the enolate complexes has been studied and also the reaction
with ClSiMe3 to give silyl enol ethers.
Ch a r t 1
In tr od u ction
Many organic reactions of enol-type substrates are
catalyzed by palladium.1 Examples are the mild syn-
thesis of unsaturated ketones from silyl enol ethers,2
the coupling of in situ generated tin enolates with aryl
or vinyl bromides,3 and the enantioselective aldol4 and
Manich-type reactions.5 In those processes palladium
enolate intermediates are believed to play a fundamen-
tal role. Some Pd-enolate complexes have been synthe-
sized, but they are still scarce. As expected for a late
transition metal, enolates prefer to coordinate to pal-
ladium either through the carbon atom (σ-alkyl type,
A, Chart 1)6 or in the more elusive chelating η3-oxoallyl
fashion (B, Chart 1).2a,4,7 The oxygen-bound type C
(Chart 1), common for early transition metals, is not
easily found for palladium, and only one example has
been proposed.8 Bridging C-O enolates (D, Chart 1)
have also been isolated.6b
C-bound palladium enolates undergo typical reactions
of metal alkyls, such as insertion of isonitriles.6b,g In
addition, the presence of a nucleophilic site (oxygen)
may also make them prone to electrophilic attack, and
in fact they are more susceptible to protonolysis than
unfunctionalized Pd-alkyls.6c
* E-mail: espinet@qi.uva.es. Fax: 34-983-423013.
(1) Trost, B. M.; Verhoeven, T. R. In Comprehensive Organometallic
Chemistry; Wilkinson, G., Stone, F. G. A., Abel, E. W., Eds.; Pergamon
Press: Oxford, 1982; Vol. 8, p 838.
(2) (a) Ito, Y.; Aoyama, H.; Hirao, T.; Mochizuki, A.; Saegusa, T. J .
Am. Chem. Soc. 1979, 101, 494-496. (b) Ito, Y.; Hirao, T.; Saegusa, T.
J . Org. Chem. 1978, 43, 1011-1013.
(3) Farina, V.; Krishnamurthy, V.; Scott, W. J . The Stille Reaction;
Wiley: New York, 1998.
To gain more insight into the behavior of this special
and important type of palladium alkyls, we have studied
in detail two palladium C-enolates synthesized by
oxidative addition of haloketones to [Pd(PPh3)4]. This
is the most convenient way of preparation of these
derivatives and has been used previously.6 The reactions
and solution behavior described in this work show the
importance of the different coordination modes of the
enolate moiety (A, B, and D, Chart 1) and how the
presence of the nucleophilic site makes enolates a
distinct type of palladium alkyl.
(4) Sodeoka, M.; Ohrai, K.; Shibasaki, M. J . Org. Chem. 1995, 60,
2648-2649.
(5) Fujii, A.; Hagiwara, E.; Sodeoka, M. J . Am. Chem. Soc. 1999,
121, 5450-5458.
(6) (a) Vicente, J .; Abad, J . A.; Chicote, M. T.; Abrisqueta, M. D.;
Lorca, J . A.; Ram´ırez de Arellano, M. C. Organometallics 1998, 17,
1564-1568. (b) Veya, P.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C.
Organometallics 1993, 12, 4899-4907. (c) Suzuki, K.; Yamamoto, H.
Inorg. Chim. Acta 1993, 208, 225-229. (d) Byers, P. K.; Canty, A. J .;
Skelton, B. W.; Traill, P. R.; Watson, A. A.; White, A. H. Organo-
metallics 1992, 11, 3085-3088. (e) Burkhardt, E.; Bergman, R. G.;
Heathcock, C. H. Organometallics 1990, 9, 30-44. (f) Wanat, R. A.;
Collum, D. B. Organometallics 1986, 5, 120-127. (g) Bertani, R.;
Castellani, C. B.; Crociani, B. J . Organomet. Chem. 1984, 269, C15-
C18.
(7) (a) Lemke, F. R.; Kubiak, C. P. J . Organomet. Chem. 1989, 373,
391-400. (b) Yoshimura, N.; Murahashi, S.-I.; Moritani, I. J . Organo-
met. Chem. 1973, 52, C58-C60.
(8) Sodeoka, M.; Tokunoh, R.; Miyazaki, F.; Hagiwara, E.; Shibasaki,
M. Synlett 1997, 463-466.
10.1021/om990613o CCC: $18.00 © 1999 American Chemical Society
Publication on Web 11/25/1999

5572 Organometallics, Vol. 18, No. 26, 1999
Albe´niz et al.
Ta ble 1. F ir st-Or d er Ra te Con sta n ts for th e
Sch em e 1
Exch a n ge of 1 a n d F r ee P P h 3 in CDCl3 a t Differ en t
P h osp h in e Con cen tr a tion sa
PPh3 conc (M)
kobs (1fP) (s-1
)
0.022
0.046
0.067
0.085
0.121
0.127
0.205
1.84 ( 0.06
1.896 ( 0.09
2.76 ( 0.06
2.44 ( 0.04
3.08 ( 0.06
3.59 ( 0.08
4.256 ( 0.016
a
T ) 298 K; concentration of 1, 0.02 M.
ν(CO) band in the IR spectrum of 3 (1654 cm-1), lower
than ν(CO) for complex 1 (1685 cm-1). Similarly 2 also
gives OPPh3 and a complex of stoichiometry [Pd(CH2C-
(O)CMe3)Cl(PPh3)] (4, Scheme 1). In this case the dimer
seems to have chloro bridges in the solid state, since
the presence of a bridging enolate is ruled out by the
high value of ν(CO) (1687 cm-1). 3 and 4 can also be
obtained by addition of a phosphine trap, such as [PdCl2-
(PhCN)2], to a solution of 1 or 2. Complexes 3 and 4 are
fluxional in solution, as discussed below.
Resu lts
Syn th esis a n d Ch a r a cter iza tion of P d -En ola tes.
The oxidative addition of ClCH2C(O)CR3 (R ) H, Me)
to the zerovalent palladium complex [Pd(PPh3)4] in
toluene or THF gives the enolate derivatives [Pd(CH2C-
(O)CR3)Cl(PPh3)2] (R ) H, 1; R ) Me, 2) as air-stable
white solids (Scheme 1).9 Complex 1 has been synthe-
sized previously, and some of its reactions have been
analyzed.6c,g The bromo analogue of complex 2 has also
been described,6f but in either case some features of the
solution behavior and reactivity of the complexes re-
mained unexplored. 1 and 2 are an equilibrium mixture
of cis and trans isomers in solution, the trans being the
major one in both cases, as shown by their 31P{1H} and
1H NMR spectra. Cis and trans isomers had been
detected for 1 before.6c The percentages found in CDCl3
at room temperature are trans-1:cis-1 ) 85:15 and
trans-2:cis-2 ) 94:6. A 1H NOESY experiment on 1
reveals chemical exchange between the methylene
protons of both isomers.
Complexes 1 and 2 reacted with AgBF4 in CH3CN to
give the cationic enolate derivatives [Pd(CH2C(O)CR3)-
(PPh3)2(NCMe)](BF4) (R ) H, 5; R ) Me, 6) (Scheme
1). The IR spectra of 5 and 6 in the solid state show the
characteristic bands of coordinated MeCN (2319 and
2286 cm-1 for 5 and 2317 and 2287 cm-1 for 6) and the
ν(CO) bands at 1680 cm-1 (5) and 1667 cm-1 (6), similar
to the values found for complexes 1 (1685 cm-1) and 2
(1669 cm-1). This indicates that both complexes are
monomeric, in contrast with an analogous cationic
derivative prepared in the same way which crystallizes
as a dimer [Pd2{µ-κ2-C,O-CH2C(O)Ph}2(PPh3)4](BF4)2.6b
In CD3CN solution they show fluxional NMR spectra,
as will be described below.
Dyn a m ic Beh a vior of th e En ola te P d -Com p lex-
es. The presence of the enolate oxygen leads to several
dynamic processes that affect the derivatives synthe-
sized. Variable-temperature NMR experiments were
performed on CDCl3 solutions of 1 and/or 2. Small
amounts of the cis isomers are present in solution in
each case, but only the major trans derivatives were
1
Complex 1 shows a static H spectrum at room tem-
perature (at 300 MHz) with the characteristic methyl-
3
ene proton signals: a triplet (δ 2.18, J H-P ) 6.7 Hz)
1
studied in detail. Static H NMR spectra are observed
for the trans isomer and a doublet of doublets (δ 2.85,
3J H-P ) 11.6, 5 Hz) for the cis isomer, in agreement with
a previous report. However, complex 2 shows a fluxional
behavior (see below), and the resonance for the meth-
ylene protons, -CH2C(O)CMe3, of the trans isomer
appears as a broad peak (δ 2.35). The corresponding
signal for the cis isomer is a broad doublet of doublets
up to 313 K for complex 1 or 283 K for complex 2 (at
300 MHz). They both show a triplet for the methylene
protons coupled to two equivalent phosphines and a
singlet for the methyl groups. As the temperature is
raised, the methylene triplet loses resolution and even-
tually becomes a broad singlet. The loss of H-P coupling
can be explained by fast decoordination-recoordination
of the phosphine ligands. In fact phosphine exchange
is observed when free PPh3 is added to a solution of
either complex 1 or 2 in CDCl3, as shown by magnetiza-
tion transfer experiments between the 31P signals
corresponding to free and coordinated phosphine. At
least in these conditions PPh3 exchange occurs by an
associative pathway, the usual mechanism of ligand
substitution on Pd(II) complexes, and a linear increase
in kobs is observed when the phosphine concentration
increases (Table 1). The data in Table 1 fit the equation
kobs ) (1.50 ( 0.2) + (13.9 ( 1.8)[PPh3], and the positive
intercept for [PPh3] ) 0 suggests the simultaneous
occurrence of a phosphine-independent pathway.
3
(δ 2.50, J H-P ) 10 Hz, 5 Hz). The 13C NMR spectra
show the characteristic signals for the major trans
isomers in both cases, and the carbonyl resonances
appear at δ 211.2 (trans-1) and δ 220.9 (trans-2).
When a solution of complex 1 in THF was stirred for
1.5 h in the air, the dimeric product [Pd2{µ-κ2-C,O-
CH2C(O)CH3}2Cl2(PPh3)2] (3) and OPPh3 were obtained
(Scheme 1). When the same experiment was carried out
under nitrogen, 1 was recovered unchanged. The pres-
ence of a bridging enolate is shown by the value of the
(9) The organometallic ligand -CH2C(O)CR3 is referred to as enolate
or C-enolate throughout this work. The alternative name ketonyl
(acetonyl when R ) Me) is also used in the literature.

Bonding Modes in Pd(II) Enolates
Organometallics, Vol. 18, No. 26, 1999 5573
Ta ble 2. F ir st-Or d er Ra te Con sta n ts for th e
Exch a n ge of 1 a n d 2 a t Differ en t Com p lex
Con cen tr a tion s a n d Tem p er a tu r es
Ch a r t 2
T (K)
conc 1 (M)
conc 2 (M)
kobs (2f1) (s-1
)
273.2
282.9
283.3
0.024
0.030
0.045 ( 0.003
0.175 ( 0.006
0.201 ( 0.004
0.211 ( 0.005
0.204 ( 0.003
0.440 ( 0.019
0.62 ( 0.02
0.012
0.023
0.096
0.024
0.012
0.028
0.096
0.030
288.3
293.0
298.3
303.5
308.9
0.99 ( 0.04
1.49 ( 0.16
3.1 ( 0.2
In the absence of free phosphine and in a noncoordi-
nating solvent such as CDCl3, the enolate oxygen atom
could play the role of the entering ligand, either in an
intramolecular (η3-oxoallyl transition state) or intermo-
lecular way (enolate bridge transition state). A bridging
chloro ligand could also trigger the substitution. 31P
magnetization transfer experiments using equimolar
solutions of complexes 1 and 2 in CDCl3 were performed,
and the exchange rate between the signals of both
complexes was measured. The exchange rate does not
change with concentration, according to the results
found at 283.3 K for several-fold increase in complex
concentration (Table 2). Rates were also measured in
the temperature range 273-308 K (Table 2), and an
Eyring plot gave the following activation parameters:
1
(3) or two (4) major signals. The assignment of the H
NMR signals for each species at this temperature is
based on their relative intensity and on a heteronuclear
1H-31P inverse correlation. According to the NMR data
the structures depicted in Chart 2 are proposed. Com-
plex 3 at 223 K seems to be a mixture of µ-enol-trans-P
(3a ), µ-enol-µ-Cl-trans-P (3b), and µ-Cl-trans-P (3c), in
a ratio 3a :3b:3c ) 50:30:20, which undergo fast inter-
conversion as the temperature is raised. The chemical
shifts for the 31P{1H} NMR resonances found indicate
that the phosphines are trans to either Cl or O (δ 36.4,
3a ; δ 36.5 and 37, 3b; δ 36.9, 3c) but not to C (expected
δ around 20 ppm) or another phosphine (expected δ
around 25-28 ppm). When complex 3 is dissolved in
CDCl3 at 223 K, the same mixture of isomers is found,
which means that even at this temperature the equi-
librium is established quickly. Complex 4 is a mixture
of two isomers, µ-enol-trans-P (4a ) and µ-Cl-trans-P (4c),
in a ratio 4a :4c ) 24:76 at 223 K (Chart 2). It seems
that a bridging enolate is preferred for 3 versus a chloro
bridge at low temperatures and in the solid state,
whereas the opposite is found for 4.
R ea ct ion s w it h Un sa t u r a t ed Su b st r a t es a n d
Electr op h iles. As a particular type of Pd-alkyls, inser-
tion of some unsaturated substrates into the Pd-C bond
of enolates can be anticipated. However 1 or 2 do not
react with MeOOC-CtC-COOMe, and neither the
neutral nor the cationic derivatives undergo insertion
of alkenes into the Pd-C bond. In contrast, CO reacts
with 1 or 2 in CDCl3 to give the carboxylic acid
CH3C(O)CH2COOH and acetone or CMe3C(O)CH2-
COOH and pinacolone, respectively, plus a dimeric
palladium derivative 7 (Scheme 2). 7 can be isolated as
an orange solid in high yield when CO is bubbled
through a solution of 1 or 2 in THF. It has been
previously synthesized by comproportionation of Pd(0)
and Pd(II) complexes, and in our case it may have been
formed in a similar way, as depicted in Scheme 2.10
∆Hq ) 77 ( 4 kJ mol-1; ∆Sq ) 15 ( 13 J K-1 mol-1
.
Although subjected to a large error, the value of ∆Sq is
significantly small, and these results rule out an
intermolecular mechanism for phosphine exchange. An
intramolecular coordination of the enolate oxygen in an
η3-oxoallyl fashion would account for the small ∆Sq
value and the concentration-independent rate (eq 1).
This oxygen-triggered substitution may account for the
phosphine-independent contribution observed in the
exchange in the presence of phosphine.
Complexes 5 and 6 also show fluxional NMR spectra
1
in CD3CN solution: broad H NMR signals, the meth-
ylene resonances appearing as broad singlets. Slow
exchange spectra are obtained at about 243 K, and a
triplet is observed for the CH2C(O)CR3 protons. The lack
of H-P coupling at room temperature can be attributed
to PPh3 decoordination-recoordination, the same proc-
ess observed for the neutral precursors 1 and 2. How-
ever, since acetonitrile is a coordinating solvent, we
cannot rule out that the ligand exchange is solvent
assisted in this case with little involvement of the
enolate oxygen.
The dimeric complexes 3 and 4 also display a dynamic
behavior in solution which can be attributed to rapid
interconversion of different isomers. Their 1H NMR
spectra at room temperature show broad signals for both
the methylene and the methyl protons and only one
broad resonance in the 31P{1H} NMR spectra (3, δ 36.2,
4 δ 36.7). The 31P resonances split at 223 K into four
On the other hand, the oxygen center in the enolate
moiety is susceptible to attack by electrophiles, and
(10) Bender, R.; Braunstein, P.; Tiripicchio, A.; Tiripicchio-Camellini,
M. J . Chem. Soc., Chem. Commun. 1984, 42-43.

5574 Organometallics, Vol. 18, No. 26, 1999
Albe´niz et al.
Sch em e 2
F igu r e 1. 31P{1H} NMR for the mixture of 5 (*), 8 ([),
and 9 (2) in CDCl3, (a) before and (b) after adding CD3CN.
indeed hydrolysis reactions are observed in noncoordi-
nating solvents. Complexes 1 and 2 slowly decompose
in CDCl3 by protonolysis of the enolate moiety to give
the corresponding ketone and [PdCl2(PPh3)2]. The same
products are observed when HCl(aq) is added to 1 or 2
in chloroform. Solutions of the complexes in rigorously
dried chlorinated solvents can be kept for longer periods
of time if protected from light. The cationic complexes
5 and 6 are stable in acetonitrile solution, but when
dissolved in CDCl3, they undergo hydrolysis reactions
immediately. When 5 is dissolved in CDCl3, a mixture
of 5, the hydroxo derivative [Pd(OH)(PPh3)2(NCMe)]-
(BF4) (8), and a dimer [Pd2{µ-κ2-C,O-CH2C(O)CH3}(µ-
OH)(PPh3)4](BF4)2 (9) is found (eqs 2 and 3). A solution
of 6 in CDCl3 contains complexes 6, 8, and the enolate-
bridged [Pd2{µ-κ2-C,O-CH2C(O)CMe3}2(PPh3)4] (BF4)2
(10) (eqs 2 and 4). The addition of H2O increases the
protonolysis of the enolate derivatives 1 or 2 in the same
way (water acting as a proton source) would produce a
putative neutral hydroxo derivative, which in chlori-
nated solvents may give [PdCl2(PPh3)2], and this is the
only palladium complex observed.
Complexes 1 and 2 react with electrophiles other than
proton, and attack on oxygen is observed. Addition of
ClSiMe3 to solutions of 1 or 2 gives the silyl enol ethers
CH2dC(OSiMe3)-CR3 (11, R ) H; 12, R ) Me) and
[PdCl2(PPh3)2] (eq 5).
Discu ssion
Three different coordination modes to palladium have
been shown to be important for the enolate ligands used
in this work. Stable complexes display a C-bound
enolate when there is only one coordination site on
palladium. When a second coordination position is
available, a bridging enolate is preferred to an oxoallyl
chelating form, both in the neutral and cationic deriva-
tives, and the stability of the bridge is finely tuned by
the nature of the R group in the -CH2(CO)R moiety.
We observe that the preference for the dimeric enolate
t
bridge follows the trend Ph > Me > Bu as shown by
the distribution of enolate versus Cl bridges in 3 and 4
and the isolation of cationic complexes from an aceto-
nitrile solution which are monomeric C-bound Pd eno-
lates in this work (R ) Me, tBu) or dimeric C-O
bridging Pd enolates for R ) Ph.6b Since the stability of
the bridge decreases as the R group becomes a stronger
donor, a π back-bonding interaction may be important
in this bonding mode, which seems to be supported by
amount of 8 and 9 in the first case, or 8 in the second,
as expected from the equilibrium in eq 2. Addition of
CD3CN produces bridge splitting and the disappearance
of the dimeric derivatives, leaving 5/8 or 6/8 as the only
species present (Figure 1 and eqs 3 and 4).
Cationic hydroxo derivatives (8, 9) are the straight-
forward products of the hydrolysis of 5 and 6. The

Bonding Modes in Pd(II) Enolates
Organometallics, Vol. 18, No. 26, 1999 5575
analyzer. Solvents were dried following standard procedures
and distilled before use. Haloketones were purchased from
Aldrich Chemical Co. and used without further purification.
[Pd(PPh3)4] was prepared as described elsewhere.13
the important decrease in ν(CO) upon bridge formation.
A third bonding mode, chelating oxoallyl, is important
in the fast intramolecular phosphine exchange observed
for complexes 1 and 2. The rates observed parallel the
basicity of the enolate oxygen R ) tBu > Me, as
determined for the parent ketones.11 Good σ donor
properties for the entering ligand (O) seems to be a
major factor that promotes the attainment of the transi-
tion state in the process. These apparently contradictory
results show that the factors that favor the two C,O-
enolate bonding modes (oxoallyl or bridging) are intrin-
sically different. The bridging mode is thermodynami-
cally preferred, at least in the complexes described here,
and is met in the isolated compounds. Nevertheless a
chelating oxoallyl-like mode, with an incipient Pd-O
bond (where π back-bonding could still be unimportant),
seems to be a key transition state or intermediate in
some reactions of the palladium enolates.
Syn th esis of [P d(CH2C(O)CH3)Cl(P P h 3)2] (1). The prepa-
ration in the literature was slightly modified.6c To a slurry of
Pd(PPh3)4 (2 g, 1.73 mmol) in toluene under a nitrogen
atmosphere was added ClCH2C(O)CH3 (0.1516 mL, 0.19
mmol). After 2 h a solution was formed, and it was stirred for
one more hour, whereupon a white solid (1) appeared. It was
filtered, washed with toluene, and air-dried: 0.93 g, 74% yield
(mixture of trans-1:cis-1 in a ratio 85:15 in CDCl3 solution).
1: Anal. Calcd for C39H35ClOP2Pd: C, 64.74; H, 4.87.
Found: C, 64.42; H, 4.84. IR, ν(CdO) 1685 cm-1, ν(Pd-Cl) 262
cm-1. 1H NMR (300 MHz, δ, CDCl3): trans-1, 7-7.9 (m, 30 H,
Ph), 2.18 (t, 2 H, CH2-, 3J H-P ) 6.7 Hz), 1.3 (s, 3 H, Me); cis-1,
7-7.9 (m, Ph), 2.85 (dd, 2 H, CH2- J ) 4.5, 11 Hz), 2.35 (s,
3H, Me). 31P{1H} NMR (121 MHz, δ, CDCl3): trans-1, 28.3 (s);
2
2
cis-1, 21.3 (d, 1P, J P-P ) 34 Hz), 38.1 (d, 1P, J P-P ) 34 Hz).
13C{1H} NMR (74.5 MHz, δ, CDCl3): trans-1, 211.2 (s, CdO),
30.7 (s, Me), 32.8 (s, CH2), 127-135 (Ph).
A few reactions undergone by the complexes prepared
parallel the reactivity of Pd-alkyls. Thus insertion of CO
into the Pd-C(enol) bond is observed, and â-ketoacids
are obtained, in contrast with a previous report on the
palladium enolate analogue derived from acetophenone.6b
Since phosphine substitution by CO is needed for
insertion, the higher reactivity of the complexes used
here seems to be a consequence of the easier phosphine
decoordination promoted by the more strongly donating
Complex 2 was obtained following a similar procedure but
using tetrahydrofurane as solvent and 1.5 h reaction time. A
white solid was obtained (61% yield, mixture of trans-2:cis-2
in a ratio 94:6 in CDCl3 solution). Anal. Calcd for C42H41ClOP2-
Pd: C, 65.89; H, 5.40. Found: C, 65.49; H, 5.54. IR, ν(CdO)
1
1669 cm-1, ν(Pd-Cl) 284 cm-1. H NMR (300 MHz, δ, CDCl3,
293 K): trans-2, 7.4-7.75 (m, 30 H, Ph), 2.35 (bs, 2 H, CH2-),
0.22 (bs, 9 H, 3Me), cis-2, 7.4-7.75 (m, Ph), 2.50 (dd, 2H, CH2-,
3J H-P ) 10.5 Hz, 5.4 Hz), 1.06 (s, 9H, 3Me). 1H NMR (300 MHz,
t
3
enolate when R ) Me, Bu than when R ) Ph. Also,
δ, CDCl3, 243 K): trans-2, 2.32 (t, J H-P ) 7.6 Hz, CH2, 2 H),
0.13 (s, 9 H, 3Me), 7.4-7.75 (m, 30 H, Ph). 31P{1H} NMR (121
MHz, δ, CDCl3, 293 K): trans-2, 28.9 (bs), cis-2, 19.9 (d, 1P,
2J P-P ) 34 Hz), 39.9 (d, 1P, 2J P-P ) 34 Hz). 13C{1H} NMR (74.5
MHz, δ, CDCl3, 243 K): 26.7 (s, 3 CH3), 29.1 (s, CH2-), 44.2
(s, CMe3), 220.87 (s, CdO), 128.1-135.1 (Ph).
reactions of 1 and other Pd-enolates with isonitriles
have been described elsewhere, and insertion into the
Pd-C bond has been observed.6b,g However, one impor-
tant difference with Pd-alkyls concerns the easy cleav-
age of the Pd-C(enolate) bond. Hydrolysis of the enolate
ligand occurs for 1 and 2 and more easily for the cationic
derivatives 5 and 6 in noncoordinating solvents. The
cleavage probably occurs by intramolecular deprotona-
tion of a coordinated water molecule by the enolate
oxygen, since H2O coordination seems to be a crucial
step, as shown by the factors that favor the reaction:
noncoordinating solvents and ligands that can be easily
substituted (NCMe).12 Other electrophiles also attack
the enolate oxygen with cleavage of the Pd-C bond. The
reaction of 1 or 2 with SiClMe3 affords, by clean
trimethylsilyl attack on the O, the corresponding silyl
enol ethers, and this is relevant to the Pd-catalyzed
syntheses that use silyl enol ethers, since it is the
reverse reaction of the first step in the catalytic cycle.2,4,5
Syn th esis of [P d 2(µ-CH2C(O)CH3)2Cl2(P P h 3)2] (3). Com-
plex 1 (0.1 g, 0.138 mmol) was dissolved in tetrahydrofurane
(20 mL), and the solution was stirred for 1.5 h in the air. The
solution was evaporated to ca. 5 mL, and Et2O (15 mL) was
added. A light yellow solid appeared, which was filtered,
washed with THF (2 mL) and then Et2O (2 × 2 mL), and air-
dried: 0.048 g, 75% yield. Anal. Calcd for C42H40Cl2O2P2Pd2:
C, 54.68; H, 4.37. Found: C, 54.24; H, 4.33. IR, ν(CdO) 1654
1
cm-1, ν(Pd-Cl) 279 cm-1. H NMR (300 MHz, δ, CDCl3, 293
K): 7.4-7.7 (m, Ph), 2.25 (bs, 2 H, CH2-), 1.95 (bs, 3 H, CH3).
31P{1H} NMR (121 MHz, δ, CDCl3, 293 K): 36.2 (s). 13C{1H}
NMR (74.5 MHz, δ, CDCl3, 293 K): 31.6 (s, CH2-), 31.9 (s,
CH3), 211.2 (s, CdO), 128.3-134.8 (m, Ph). 1H NMR (300 MHz,
δ, CDCl3, 223 K): 7.3-7.8 (m, Ph), 2.25 (bs, CH2-, 3c), 2.19
(bs, CH2-, 3a ), 2.14 (bs, CH2-, 3b), 1.84 (s, CH3, 3a , 3c), 1.74
(s, CH3, 3b ), 1.69 (s, CH3, 3b ). 31P{1H} NMR (121 MHz, δ,
CDCl3, 223 K): 36.38 (s, 3a ), 36.45 (s, 3b), 37.1 (s, 3b), 36.9
(s, 3c).
Exp er im en ta l Section
Complex 4 was prepared in a similar way (50% yield). Anal.
Calcd for C48H52Cl2O2P2Pd2: C, 57.27; H, 5.21. Found: C,
Gen er a l P r oced u r es. C, H, and N elemental analyses were
performed on a Perkin-Elmer 240 microanalyzer. 1H, 13C, and
31P NMR spectra were recorded on Bruker AC-300 and ARX-
300 spectrometers. Chemical shifts (in δ units, ppm) were
referenced to TMS for 1H and 13C and to H3PO4 for 31P. The
spectral data were recorded at 293 K unless otherwise noted.
IR spectra were recorded using Nujol mulls on a Perkin-Elmer
883 spectrophotometer. Organic products were analyzed using
a HP-5890 gas chromatograph connected to a HP-5988 mass
spectrometer at an ionizing voltage of 70 eV and a quadrupole
1
56.94; H, 5.07. IR, ν(CdO) 1687 cm-1, ν(Pd-Cl) 275 cm-1. H
NMR (300 MHz, δ, CDCl3, 293 K): 7.4-7.8 (m, Ph), 1.80 (bs,
2 H, CH2-), 1.30 (bs, 3 H, CH3). 31P{1H} NMR (121 MHz, δ,
CDCl3, 293 K): 36.7 (s). 1H NMR (300 MHz, δ, CDCl3, 223 K):
7.3-7.8 (m, Ph), 1.79 (b, CH2-, 4a ), 1.58 (b, CH2-, 4c), 1.42
(bs, CH3, 4a ), 1.20 (s, CH3, 4c). 31P{1H} NMR (121 MHz, δ,
CDCl3, 223 K): 38.1 (s, 4c), 37.4 (s, 4a ).
P r ep a r a tion of [P d (CH2C(O)CH3)(P P h 3)2(NCMe)](BF 4)
(5). Complex 1 (0.15 g, 0.207 mmol) was added to a solution
of AgBF4 (0.044 g, 0.227 mmol) in acetonitrile (20 mL). The
mixture was stirred for 1.5 h in the dark, and then the
suspension was filtered through Celite. The pale yellow
solution was evaporated to ca. 2 mL, and Et2O (10 mL) was
(11) Campbell, H. J .; Edward, J . T. Can. J . Chem. 1960, 38, 2109.
(12) Protonation of the enolate oxygen could take place either in a
σ-ketonyl form or in an intermediate π-coordinated enolate to give a
vinyl alcohol complex (see: Hillis, J .; Francis, J .; Ori, M.; Tsutsui, M.
J . Am. Chem. Soc. 1974, 96, 4800-4804). We have no experimental
evidence to favor one pathway or the other.
(13) Coulson, D. R. in Inorg. Synth. 1990, 28, 107-109.

5576 Organometallics, Vol. 18, No. 26, 1999
Albe´niz et al.
added, yielding a pale yellow solid, which was filtered washed
with Et2O (2 × 10 mL) and air-dried: 0.15 g, 90% yield. Anal.
Calcd for C41H38BF4NOP2Pd: C, 60.35; H, 4.69; N, 1.71.
Found: C, 60.41; H, 5.02; N, 1.62. IR, 2319 and 2286 cm-1
22.4 (bs). 13C{1H} NMR (74.5 MHz, δ, CDCl3, under a CO
atmosphere): 210.4 (s, CdO), 128.0-135.1 (Ph).
CO was bubbled through solutions of 1 or 2 in CDCl3 for 5
min. The yellow solutions turned orange, and 7 and the
corresponding ketoacids, CR3C(O)CH2COOH, were observed
by NMR.
(NCMe), ν(CdO) 1680 cm-1 1H NMR (300 MHz, δ, CD3CN,
.
293 K): 7.2-7.8 (m, Ph, 30 H), 2.30 (bs, 2 H, CH2), 1.20 (bs, 3
1
H, CH3). H NMR (300 MHz, δ, CD3CN, 238 K): 7.1-7.7 (m,
R ) H: 1H NMR (300 MHz, δ, CDCl3), 12.1 (bs, 1 H, COOH),
3.50 (s, 2 H, CH2-), 2.17 (s, 3 H, CH3).
3
Ph, 30 H), 2.24 (t, 2 H, J H-P ) 8 Hz, CH2), 0.90 (s, 3 H, CH3).
31P{1H} NMR (121 MHz, δ, CD3CN, 293 K): 28.0 (s). 13C{1H}
NMR (74.5 MHz, δ, CD3CN, 238 K): 211.8 (s, CdO), 129.0-
135.1 (Ph), 32.2 (s, CH2), 30.0 (s, CH3).
R ) Me: 1H NMR (300 MHz, δ, CDCl3), 12.3 (bs, 1 H,
COOH), 3.58 (s, 2 H, CH2-), 1.17 (s, 9 H, CMe3).
Rea ction of 1 w ith SiClMe3. To a solution of 1 (0.0156 g,
0.022 mmol) in CDCl3 (0.6 mL) was added SiClMe3 (0.003 mL,
0.023 mmol). [PdCl2(PPh3)2] and 11 appeared immediately as
Complex 6 was prepared following the same procedure.
Anal. Calcd for C44H44BF4NOP2Pd: C, 61.59; H, 5.17; N, 1.63.
Found: C, 61.18; H, 5.26; N, 1.58. IR: 2317 and 2287 cm-1
1
shown by H and 31P NMR.
(NCMe), ν(CdO) 1667 cm-1 1H NMR (300 MHz, δ, CD3CN,
.
11: 1H NMR (300 MHz, δ, CDCl3): 4.06 (m, 1 H, H1), 4.05
(bs, 1 H, H1′), 1.78 (d, J ) 1 Hz, 3 H, Me3), 0.21 (s, 9 H, SiMe3).
Compound 12 was obtained in a similar way: 1H NMR (300
MHz, δ, CDCl3): 4.09 (d, J ) 1 Hz, 1 H, H1), 3.93 (bs, 1 H,
H1′), 1.05 (s, 9 H, 3 Me3), 0.21 (s, 9 H, SiMe3).
293 K): 7.2-7.9 (m, 30 H, Ph), 2.55 (bs, 2 H, CH2), 0.20 (bs, 9
H, CMe3). 1H NMR (300 MHz, δ, CD3CN, 243 K): 7.2-7.8 (m,
3
Ph, 30 H), 2.54 (t, 2 H, J H-P ) 7.8 Hz, CH2), 0.08 (s, 9 H,
CMe3). 31P{1H} NMR (121 MHz, δ, CD3CN, 293 K): 28.3 (bs).
13C{1H} NMR (74.5 MHz, δ, CD3CN, 293 K): 219.4 (s, CdO),
127.8-135.3 (Ph), 44.8 (s, CMe3), 27.6 (s, CH2), 26.4 (s, CMe3).
When complex 5 was dissolved in CDCl3, a mixture of
complexes 5, 8, and 9 was formed as shown by NMR. The
addition of water to the mixture changed the ratio of products
from 5:8:9 ) 1.4:1:1.9 (referred to Pd) to 5:8:9 ) 0.2:1:0.4.
Complex 6 gives a mixture of 6, 8, and 10 in CDCl3.
Kin etic Mea su r em en ts. NMR tubes (5 mm) were charged
with the appropriate amount of complexes 1 and 2 (or complex
1 and PPh3), and CDCl3 was added to a total volume of 0.6
mL to give solutions of the concentrations collected in Tables
1 and 2. The samples were placed in a probe provided with a
B-VT-2000 temperature control unit. The temperature was
calibrated measuring the difference between the chemical
shifts of the MeOH signals at each temperature.14 Kinetics
were carried out by magnetization transfer experiments, with
selective inversion of the 31P resonance of complex 1 using a
90°-D1-90°-t-90°-D2 sequence, where D1 ) 1/2∆ν, ∆ν is the
separation in Hz between both signals, t is the magnetization
transfer delay, and D2 is the relaxation delay. Values of 90°
pulses and D1 were determined at each temperature. After
excitation, the signal areas of both signals (1 and 2, or 1 and
PPh3) were measured and processed to obtain the values of
5: 1H NMR (300 MHz, δ, CDCl3): 6.8-7.9 (m, Ph), 2.30 (t,
3
2 H, J H-P ) 8 Hz, CH2), 1.40 (bs, 3 H, CH3CN), 1.00 (s, 3 H,
CH3). 31P{1H} NMR (121 MHz, δ, CDCl3): 27.1 (s).
6: 1H NMR (300 MHz, δ, CDCl3,): 7.0-8.0 (m, Ph), 2.57 (t,
3
2 H, J H-P ) 7.8 Hz, CH2), 0.18 (s, 9 H, CMe3), 1.34 (s, 3 H,
CH3CN). 31P{1H} NMR (121 MHz, δ, CDCl3): 27.4 (s).
8: 1H NMR (300 MHz, δ, CDCl3): 6.8-7.9 (m, Ph), 0.9 (bs,
3 H, CH3CN), -2.3 (bs, OH). 31P{1H} NMR (121 MHz, δ,
CDCl3): 33.9 (s).
9: 1H NMR (300 MHz, δ, CDCl3): 6.8-7.9 (m, Ph), 4.1 (dd,
k
obs, as was reported before.15 Values of kobs(1fPPh3) reported
in Table 1 were calculated from the experimental kobs(PPh3f1)
2
3
1 H, J H-H ) 11 Hz, J H-P ) 4.5 Hz, CH2), 2.8 (m, 1 H, CH2),
1.2 (s, 3 H, CH3), -0.4 (bs, OH). 31P{1H} NMR (121 MHz, δ,
CDCl3): 36.9 (dd, 1 P, J ) 23, 7 Hz), 32.6 (dd, 1 P, J ) 18, 7
Hz), 30.9 (d, 1 P, J ) 18 Hz), 28.5 (d, 1 P, J ) 23 Hz).
10: 1H NMR (300 MHz, δ, CDCl3): 7.0-8.0 (m, Ph), 3.08
using the equilibrium equation [1]kobs(1fPPh3) ) [PPh3]kobs
-
(PPh3f1). Errors were calculated as reported before.16
Ack n ow led gm en t. We thank the Spanish Ministry
of Education and Science (grant no. PB-96-0363) and
the J unta de Castilla y Leo´n (grant no.VA18/97) for
financial support.
3
(d, 2 H, J H-P ) 6.5 Hz, CH2), 1.09 (s, 9 H, CMe3). 31P{1H}
NMR (121 MHz, δ, CDCl3): 33.9 (d, 2 P, J ) 36.8 Hz), 21.2 (d,
2 P, J ) 36.8 Hz).
Rea ction s w ith CO. Syn th esis of [P d 2(µ-CO)Cl2(P P h 3)3]
(7).10 CO was bubbled through a suspension of 1 (0.150 g, 0.207
mmol) in THF. The light yellow suspension turned into an
orange solution. After 5 min an orange solid precipitated,
which was filtered and air-dried: 0.1 g, 84% yield. Anal. Calcd
for C55H45Cl2OP3Pd2: C, 60.13; H, 4.13. Found: C, 59.75; H,
OM990613O
(14) Van Geet, A. L. Anal. Chem. 1970, 42, 679-680.
(15) Albe´niz, A. C.; Casado, A. L.; Espinet, P. Inorg. Chem. 1999,
38, 2510-2515.
(16) (a) Casado, A. L.; Casares, J . A.; Espinet, P. Organometallics
1997, 16, 5730-5736. (b) Kirkup, L. Experimental Methods; Wiley:
Brisbane, 1994.
4.36. IR, ν(CdO) 1864 cm-1
.
1H NMR (300 MHz, δ, CDCl3):
7.15-7.8 (m, 45 H, Ph). 31P{1H} NMR (121 MHz, δ, CDCl3):