Palladium-catalyzed C-P bond formation: Mechanistic studies on the ligand substitution and the reductive elimination. An intramolecular catalysis by the acetate group in PdII complexes

Article abstract of DOI:10.1021/om800641n

Ligand substitution and reductive elimination of the palladium-catalyzed C-P bond forming cross-coupling were investigated in depth. It was found that for PhPd^II(PPh₃)₂X (X = I, Br, Cl) complexes, a step commonly referred to as ligand substitution commenced with coordination of an H-phosphonate diester, followed by its deprotonation to form an equilibrium mixture of penta- and tetracoordinate palladiumphosphonate intermediates, from which reductive elimination of the product (diethyl phenylphosphonate) occurred. For the acetate counterpart, PhPd^II(PPh₃) ₂(OAc), the incorporation of a phosphonate moiety to the complex was preceded by a rate-determining removal of the supporting phosphine ligand, facilitated by an intramolecular catalysis by the acetate group. Both the reaction steps, i.e., formation of palladiumphosphonate intermediates and reductive elimination, were significantly faster for the acetate versus halides containing Pd^II complexes investigated. Similar observations were found to be true also for bidentate ligand complexes [(dppp)Pd ^II(Ph)X]; however, in this instance, a single palladiumphosphonate intermediate, (dppp)Pd^II(Ph)(PO(OEt)₂), could be observed by ³¹P NMR spectroscopy. The synthetic and kinetic studies on the cross-coupling reaction of diethyl H-phosphonate with phenyl halides permitted us to elucidate a crucial catalytic role of an acetate group in Pd^II complexes and to propose two distinctive catalytic cycles, which complemented traditional Pd⁰/Pd^II schemes, for the palladium-mediated C-P bond formation.

Full text of DOI:10.1021/om800641n

5876
Organometallics 2008, 27, 5876–5888
Palladium-Catalyzed C-P Bond Formation: Mechanistic Studies on
the Ligand Substitution and the Reductive Elimination. An
Intramolecular Catalysis by the Acetate Group in PdII Complexes
Marcin Kalek† and Jacek Stawinski*,†,‡
Department of Organic Chemistry, Arrhenius Laboratory, Stockholm UniVersity,
S-106 91 Stockholm, Sweden, and Institute of Bioorganic Chemistry, Polish Academy of Sciences,
Noskowskiego 12/14, 61-704 Poznan, Poland
ReceiVed July 8, 2008
Ligand substitution and reductive elimination of the palladium-catalyzed C-P bond forming cross-
coupling were investigated in depth. It was found that for PhPdII(PPh3)2X (X ) I, Br, Cl) complexes, a
step commonly referred to as ligand substitution commenced with coordination of an H-phosphonate
diester, followed by its deprotonation to form an equilibrium mixture of penta- and tetracoordinate
palladiumphosphonate intermediates, from which reductive elimination of the product (diethyl phe-
nylphosphonate) occurred. For the acetate counterpart, PhPdII(PPh3)2(OAc), the incorporation of a
phosphonate moiety to the complex was preceded by a rate-determining removal of the supporting
phosphine ligand, facilitated by an intramolecular catalysis by the acetate group. Both the reaction steps,
i.e., formation of palladiumphosphonate intermediates and reductive elimination, were significantly faster
for the acetate versus halides containing PdII complexes investigated. Similar observations were found to
be true also for bidentate ligand complexes [(dppp)PdII(Ph)X]; however, in this instance, a single
palladiumphosphonate intermediate, (dppp)PdII(Ph)(PO(OEt)2), could be observed by 31P NMR spec-
troscopy. The synthetic and kinetic studies on the cross-coupling reaction of diethyl H-phosphonate with
phenyl halides permitted us to elucidate a crucial catalytic role of an acetate group in PdII complexes and
to propose two distinctive catalytic cycles, which complemented traditional Pd0/PdII schemes, for the
palladium-mediated C-P bond formation.
nucleophiles follow a three-step catalytic cycle, which for the
example of H-phosphonates, is depicted in Scheme 1.
Introduction
Palladium-catalyzed cross-coupling reactions constitute ex-
tremely useful tools in organic synthesis,1 and their applications
have significantly broadened when, in addition to classical
carbon-carbon bond forming transformations, new variants of
the reaction with heteroatom nucleophiles have been developed.2
Among these, several palladium-catalyzed cross-couplings using
phosphorus nucleophiles3,4 attracted considerable attention due
to their possible applications in the synthesis of modified nucleic
acids5 and chiral phosphine ligands.6
Most of mechanistic studies on catalytic cycles as that shown
in Scheme 1, using experiemental9,10 and computational11
techniques, have been focused on oxidative addition, a common
step of a variety of important cross-coupling reactions. The
reductive elimination steps, in which the C-N,12,13 C-O,12,14
C-S,12,15,16 and C-P17-19 bonds can be produced, received
(5) (a) Abbas, S.; Hayes, C. J. Synlett 1999, 1124–1126. (b) Abbas, S.;
Hayes, C. J.; Worden, S. Tetrahedron Lett. 2000, 41, 3215–3219. (c) Lera,
M.; Hayes, C. J. Org. Lett. 2000, 2, 3873–3875. (d) Abbas, S.; Hayes, C. J.
Tetrahedron Lett. 2000, 41, 4513–4517. (e) Abbas, S.; Bertram, R. D.;
Hayes, C. J. Org. Lett. 2001, 3, 3365–3367. (f) Johansson, T.; Stawinski,
J. Chem. Commun. 2001, 2564–2565. (g) Zmudzka, K.; Johansson, T.;
Wojcik, M.; Janicka, M.; Nowak, M.; Stawinski, J.; Nawrot, B. New
J. Chem. 2003, 27, 1698–1705. (h) Johansson, T.; Stawinski, J. Nucleosides,
Nucleotides Nucleic Acids 2003, 22, 1459–1461. (i) Lave´n, G.; Stawinski,
J. Collect. Symp. Ser. 2005, 7, 195–199.
(6) (a) Glueck, D. S. Synlett 2007, 2627–2634. (b) Gilbertson, S. R.;
Wang, X. F. Tetrahedron Lett. 1996, 37, 6475–6478. (c) Reetz, M. T.; Sell,
T.; Goddard, R. Chimia 2003, 57, 290–292. (d) Reetz, M. T.; Bondarev,
O. Angew. Chem., Int. Ed. 2007, 46, 4523–4526. (e) Trost, B. M.; Vidal,
B.; Thommen, M. Chem.-Eur. J. 1999, 5, 1055–1069.
(7) Shekhar, S.; Ryberg, P.; Hartwig, J. F.; Mathew, J. S.; Blackmond,
D. G.; Strieter, E. R.; Buchwald, L. S. J. Am. Chem. Soc. 2006, 128, 3584–
3591.
(8) In one of the first reports on the C-P bond formation by a
palladium(0)-catalyzed cross-coupling reaction, an incorrect mechanism was
also proposed. See: Hirao, T.; Masunaga, T.; Yamada, N.; Agawa, T. Bull.
Chem. Soc. Jpn. 1982, 55, 909–913.
(9) (a) Amatore, C.; Jutand, A. J. Organomet. Chem. 1999, 576, 254–
278. (b) Amatore, C.; Jutand, A. Acc. Chem. Res. 2000, 33, 314–321.
(10) Casado, A.; Espinet, P. Organometallics 1998, 17, 954–959.
(11) Gooꢀen, L. J.; Koley, D.; Herman, H.; Thiel, W. Chem. Commun.
2004, 2141–2142.
Despite some controversies,7,8 there is a consensus that
palladium-catalyzed cross-coupling reactions with heteroatom
* Corresponding author. Tel: (+46) 0816 2485. Fax: (+46) 0815 4908.
E-mail: js@organ.su.se.
† Stockholm University.
‡ Polish Academy of Sciences.
(1) (a) Metal-Catalyzed Cross-Coupling Reaction; de Meijere, A.;
Diederich, F., Eds.; Wiley-VCH: Weinheim, 1998. (b) Hegedus, L. S.
Transition Metals in the Synthesis of Complex Organic Molecules, 2nd ed.;
University Science Books: Sausalito, CA, 1999.
(2) Prim, D.; Campagne, J. M.; Joseph, D.; Andrioletti, B. Tetrahedron
2002, 58, 2041–2075.
(3) (a) Schwan, A. L. Chem. Soc. ReV. 2004, 33, 218–224. (b) Kovacik,
I.; Wicht, D. K.; Grewal, N. S.; Glueck, D. S.; Incarvito, C. D.; Guzei,
I. A.; Rheingold, A. L. Organometallics 2000, 19, 950–953. (c) Moncarz,
J. R.; Brunker, T. J.; Glueck, D. S.; Sommer, R. D.; Rheingold, A. L. J. Am.
Chem. Soc. 2003, 125, 1180–1181. (d) Bravo-Altamirano, K.; Huang, Z. H.;
Montchamp, J. L. Tetrahedron 2005, 61, 6315–6329. (e) Kazankova, M. A.;
Trostyanskaya, I. G.; Lutsenko, S. V.; Beletskaya, I. P. Tetrahedron Lett.
1999, 40, 569–572.
(4) Deprele, S.; Montchamp, J. L. J. Am. Chem. Soc. 2002, 124, 9386–
9387.
10.1021/om800641n CCC: $40.75
2008 American Chemical Society
Publication on Web 10/07/2008

Ligand Substitution from PdII Complexes
Organometallics, Vol. 27, No. 22, 2008 5877
Table 1. Effect of Anionic Additives on the Cross-Coupling of
Diethyl H-Phosphonate with Bromo- and Iodobenzenea,b
Scheme 1. Generic Mechanism for the Palladium-Catalyzed
Cross-Coupling between Aryl Halide and H-Phosphonate
Diesters (G1 ) L, dba, X-, OAc-, solvent; G2 ) X-, OAc-,
solvent)
a Experimental conditions: 0.1 M (EtO)2P(O)H, 1.1 equiv of Ph-X,
1.2 equiv of Et3N, THF, 60 °C, 10 mol % Pd, anions were added as
corresponding n-Bu4N+ salts. b All data from ref. 28.
oxidative addition to transition metals, and the complexes
formed may follow hydrophosphorylation4,23,24 and transfer
hydrogenation4,23,25 reaction pathways. H-Phosphonate deriva-
tives can also act as air-stable HASPO (heteroatom-substituted
secondary phosphine oxides) preligands,26 forming palladium(0)
complexes that are active in oxidative addition of aryl halides
and that can be used as catalysts for other cross-coupling
reactions.26,27 This type of reactivity, however, is highly
undesirable when H-phosphonates themselves are intended to
be substrates in Pd(0)-catalyzed cross-coupling reactions.28
Our recent studies on the influence of palladium sources and
anionic additives on the rate and efficiency of the cross-coupling
between H-phosphonates and aryl halides led to the discovery
of an intriguing phenomenon, namely, that the rate of ligand
substitution was strongly dependent on the leaving group in PdII
complexes and followed the order OAc- . Cl-, Br- > I-.28
It was argued that exceptionally high rates of ligand substitution
for the acetate derivatives could account for a remarkable
acceleration of the overall cross-coupling reactions observed
in the presence of OAc- ions (Table 1).29
less attention, and the studies were confined mainly to complexes
containing bidentate ligands. It is assumed that, similarly to the
palladium-catalyzed C-C bond formation,20,21 in these reactions
a cis arrangement of the eliminated groups secured by bidentate
ligands facilitates carbon-heteroatom bond formation in this
step. Even less information is available for the process of ligand
substitution during cross-coupling reactions with the heteroatom
nucleophiles, with notable exceptions of the work by Jutand et
al.16 on an intimate mechanism of thiol transfer to the PdII center
and that by Hartwig, Blackmon, and Buchwald el al.,7 on a
mechanism of the palladium-catalyzed C-N bond formation.22
There are several interesting practical and mechanistic aspects
of d8-metallophosphonate chemistry, in which H-phosphonate
derivatives play a pivotal role. In addition to acting as
nucleophiles in the cross-coupling reactions, H-phosphonates
(or other species containing the P-H bond) can undergo
(12) Hartwig, J. F. Acc. Chem. Res. 1998, 31, 852–860.
(13) (a) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1995, 117,
4708–4709. (b) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1997, 119,
8232–8245.
(14) (a) Mann, G.; Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, 13109–
13110. (b) Widenhoefer, R. A.; Zhong, H. A.; Buchwald, L. S. J. Am. Chem.
Soc. 1997, 119, 6787–6795. (c) Widenhoefer, R. A.; Buchwald, L. S. J. Am.
Chem. Soc. 1998, 120, 6504–6511. (d) Mann, G.; Incarvito, C.; Rheingold,
A. L.; Hartwig, J. F. J. Am. Chem. Soc. 1999, 121, 3224–3225.
(15) (a) Baranano, D.; Hartwig, J. F. J. Am. Chem. Soc. 1995, 117, 2937–
2938. (b) Mann, G.; Baranano, D.; Hartwig, J. F.; Rheingold, A. L.; Guzei,
I. A. J. Am. Chem. Soc. 1998, 120, 9205–9219.
Inspired by the idea that a ligand substitution step may be
responsible for the overall rate of cross-coupling reactions (Table
1), we set out to perform in-depth investigations on this process,
using as a model reaction the palladium-catalyzed cross-coupling
between aryl halides and diethyl H-phosphonate. In this paper,
we describe our studies on a mechanism of a ligand substitution
in palladium(II) complexes containing monodentate and biden-
tate ligands, as well as reductive eliminations from the palla-
(16) Moreau, X.; Campagne, J. M.; Meyer, G.; Jutand, A. Eur. J. Org.
Chem. 2005, 374, 3749–3760.
(17) (a) Gaumont, A.-C.; Hursthouse, M. B.; Coles, S. J.; Brown, J. M.
Chem. Commun. 1999, 63–64. (b) Kohler, M. C.; Stockland, R. A., Jr.;
Rath, N. P. Organometallics 2006, 25, 5746–5756.
(18) Levine, A. M.; Stockland, R. A.; Clark, R.; Guzei, I. Organome-
tallics 2002, 21, 3278–3284.
(19) Stockland, R. A.; Levine, A. M.; Giovine, M. T.; Guzei, I. A.;
Cannistra, J. C. Organometallics 2004, 23, 647–656.
(20) Gillie, A.; Stille, J. K. J. Am. Chem. Soc. 1980, 102, 4933–4941.
(21) Åkermark, B.; Ljungqvist, A. J. Organomet. Chem. 1979, 182, 59–
75.
(22) (a) Alcazar-Roman, L. M.; Hartwig, J. F.; Rheingold, A. L.; Liable-
Sands, L. M.; Guzei, I. A. J. Am. Chem. Soc. 2000, 122, 4618–4630. (b)
Singh, U. K.; Strieter, E. R.; Blackmond, D. G.; Buchwald, L. S. J. Am.
Chem. Soc. 2002, 124, 14104–14114.
(23) Montchamp, J. L. J. Organomet. Chem. 2005, 690, 2388–2406.
(24) (a) Han, L. B.; Tanaka, M. J. Am. Chem. Soc. 1996, 118, 1571–
1572. (b) Han, L.-B.; Hua, R.; Tanaka, M. Angew. Chem., Int. Ed. 1998,
37, 94–96. (c) Mirzaei, F.; Han, L.-B.; Tanaka, M. Tetrahedron Lett. 2001,
42, 297–299. (d) Han, L. B.; Mirzaei, F.; Zhao, C. Q.; Tanaka, M. J. Am.
Chem. Soc. 2000, 122, 5407–5408. (e) Zhao, C. Q.; Han, L. B.; Tanaka,
M. Organometallics 2000, 19, 4196–4198.
(25) Montchamp, J. L.; Dumond, Y. R. J. Am. Chem. Soc. 2001, 123,
510–511.
(26) Ackermann, L. Synthesis 2006, 1557–1571.
(27) Ackermann, L.; Althammer, A. Org. Lett. 2006, 8, 3457–3460.
(28) Kalek, M.; Stawinski, J. Organometallics 2007, 26, 5840–5847.
(29) OAc- ions also accelerate the oxidative addition to Pd0; however,
their effect on this step of the catalytic cycle is expected to be lower than
that of Cl- and Br-. See also ref 9.

5878 Organometallics, Vol. 27, No. 22, 2008
Kalek and Stawinski
Scheme 2. Possible Reaction Pathways for the Ligand
Substitution in PhPdL2X Complexes with an H-Phosphonate
Diester As a Nucleophile (L ) PPh3)
Figure 1. Dependence of the rate of PhPd(PPh3)2I decay on
concentration of (EtO)2P(O)H in THF at 40 °C (initial [PhPd-
(PPh3)2I] ) 5 mM, [Et3N] ) 50 mM).
diumphosphonate intermediates formed. In particular, a mecha-
nistic role of leaving groups in palladium(II) complexes is
addressed in detail, and a plausible explanation, supported by
experimental data, for the dramatic acceleration of the cross-
coupling reactions in the presence of acetate ions is proposed.
Results and Discussion
Figure 2. Dependence of the rate of PhPd(PPh3)2I decay on
concentration and the kind of base in THF at 40 °C (initial
[PhPd(PPh3)2I] ) 5 mM, [(EtO)2P(O)H] ) 50 mM).
Ligand Substitution in PhPdL2X Complexes (X )
halide). Palladium(II) complexes are tetracoordinated, 16-
electron, square-planar species, in which ligand substitution can
occur via two distinct reaction pathways: (i) an associative path,
involving pentacoordinated, 18-electron square-pyramidal and
trigonal-bipyramidal intermediates (or transition states), or (ii)
a dissociative path, which proceeds via tricoordinated, 14-
electron, T-shaped species.30,31
by the replacement of the coordinated solvent molecule by a
phosphonate moiety.30,32 Such a mechanism can be kinetically
indistinguishable from those in path A or C, but stereochemistry
of the palladium phosphonate complexes formed can be
different.
For trans-PhPdII(PPh3)2X (X ) I, Br, Cl) complexes, which
are intermediates implicated in the above-discussed cross-
coupling reactions, two dissociative mechanisms (paths A and
C) and an associative one (path B) (Scheme 2) can be considered
for the ligand substitution with an H-phosphonate nucleophile.
Path A is initiated by a reversible dissociation of the halide
anion to form a 14-electron unsaturated complex, which
undergoes reaction with an H-phosphonate diester in the
presence of base, to produce a trans palladium phosphonate
complex. In path B, an H-phosphonate coordinates to the starting
palladium complex, forming five-coordinate intermediates (iso-
meric square pyramids and trigonal bipyramids), which can
release either the halide or PPh3, leading to the corresponding
phosphonate complexes with different possible geometries (see
further in the text). Finally, in path C, the initial reversible
dissociation of a phosphine ligand is followed by the addition
of an H-phosphonate diester, and after deprotonation, an anionic
cis palladium phosphonate complex is formed. In addition to
these, there is a solvent-assisted substitution pathway possible,
in which solvent molecules act as nucleophiles replacing first
the halide or PPh3 (an associative pathway), and this is followed
To evaluate a possible participation of the above-presented
mechanisms during ligand substitution with H-phosphonate
diesters, we measured rates of decay of PhPd(PPh3)2X com-
plexes (X ) I, Br, Cl) under various experimental conditions,
using 31P NMR spectroscopy. In all kinetic runs the initial
concentration of the palladium complexes was 5 mM. Diethyl
H-phosphonate and triethylamine were used as a model phos-
phorus nucleophile and a base, respectively, and their concentra-
tions were varied from 50 to 150 mM (10- to 30-fold excess
over PdII complexes). To avoid palladium black formation from
low-ligated Pd0 species, generated by reductive elimination of
diethyl phenylphosphonate from PdII complexes, 1 equiv of PPh3
per the Pd complex (5 mM; except for the experiments with
variable amounts of PPh3) was added to the reaction mixtures.
Under such conditions, and with high excess of the H-
phosphonate and a base, first-order kinetics with respect to
PhPdII(PPh3)2X was observed in all instances, with excellent
reproducibility of the measured rate constants.
Figures 1-3 show the observed first-order rate constants for
the decay of complex PhPd(PPh3)2I, as a function of concentra-
tions of (EtO)2P(O)H, the base used, and PPh3, respectively.
As it was apparent from the presented data, the reaction showed
(30) Cross, R. J. AdV. Inorg. Chem. 1989, 34, 219–292.
(31) (a) Romeo, R. Comments Inorg. Chem 1990, 11, 21–57. (b) Cross,
R. Ligand Substitution Reactions of the Square-Planar Molecules; The Royal
Society Chemistry: London, 1985.
(32) (a) Espinet, P.; Echavarren, A. M. Angew. Chem., Int. Ed. 2004,
43, 4704–4734. (b) Crabtree, R. H. The Organometallic Chemistry of the
Transition Metals; Wiley: Hoboken, 2005.

Ligand Substitution from PdII Complexes
Organometallics, Vol. 27, No. 22, 2008 5879
toward a PdII center. It was determined previously that, contrary
to the hard-soft acid base principle prediction, the affinity of
halides to the palladium in PhPd(PPh3)2X complexes decreases
in the order Cl > Br > I.35,36 Thus, if the ligand substitution
would indeed follow path A, the fastest reaction should be
expected for the iodide-ligated complex, followed by the
bromide and chloride derivatives, according to relative concen-
+
trations of free cationic 14-electron species [PhPd(PPh3)2
]
produced from these complexes. However, the data in Table 2
display an opposite order of reactivity, and this, to high
probability, excludes path A as a plausible reaction mechanism.
Analysis of the reaction pathways for the ligand substitution
in PhPd(PPh3)2X complexes (Scheme 2) showed that apparently
only path B was consistent with the observed linear dependence
of the rate on (EtO)2P(O)H concentration, and it correctly
predicted the absence of the influence from PPh3. Also the trend
in reactivity of PdII complexes bearing different halides (Table
2) can be satisfactorily explained on the basis of this mechanism.
The first step of the reaction, a coordination of the H-
phosphonate diester to PhPd(PPh3)2X, should be facilitated by
low electron density at the palladium center, and since this step
is kinetically important, the reaction rates were expected to
follow the electronegativity order of the halogen atoms (Cl >
Br > I) present in these complexes (Table 2).
Figure 3. Dependence of the rate of PhPd(PPh3)2I decay on
concentration of PPh3 in THF at 40 °C (initial [PhPd(PPh3)2I] ) 5
mM, [(EtO)2P(O)H] ) [Et3N] ) 50 mM).
Table 2. Observed First-Order Rate Constants (kobs) of the Decay of
PhPd(PPh3)2X Complexes at 40 °C (initial [PhPd(PPh3)2X] ) 5 mM,
[(EtO)2P(O)H] ) [Et3N] ) 50 mM)
entry
palladium(II) complex
solvent
kobs (min-1
)
1
2
3
4
PhPd(PPh3)2I
PhPd(PPh3)2I
PhPd(PPh3)2Br
PhPd(PPh3)2Cl
THF
toluene
THF
0.0101
0.0138
0.0192
0.0336
THF
The role of base in this reaction (Figure 2) deserves some
comment since deprotonation of the H-phosphonate diester is
a chemically important step. One can envisage two scenarios
for this process (Scheme 3): (i) a base-mediated, rapid pre-
equilibrium between the H-phosphonate and its anion, followed
by attack of the generated phosphorus nucleophile on the
palladium complex (path B1), and (ii) an initial coordination
of the H-phosphonate diester to the PdII complex, followed by
abstraction of the phosphorus-bonded proton by a base (path
B2).
a linear dependence on the H-phosphonate concentration (Figure
1), but was rather insensitive to the amounts of the base or PPh3
used (Figures 2 and 3). Therefore, the experimentally determined
rate law for the ligand substitution in the studied complexes
was as follows:
d[PhPd(PPh3)2X]
-
) kobs[PhPd(PPh3)2X]
(1)
dt
where
The first type of a mechanism, which involves generation of
a phosphonate anion, was the most obvious option, since
phosphorus acid diesters in the H-phosphonate form, cannot act
as phosphorus nucleophiles. The second type of a mechanism,
i.e. coordination of a neutral nucleophile followed by its
deprotonation, however, was postulated for the ligand substitu-
tions involving thiols16 and, in a slightly different kind of
palladium-catalyzed reaction, also for alcohols.37 In the instance
of H-phosphonates derivatives, which do not have a lone
electron pair on the phosphorus atom, the association process
would have to involve either oxygen electron pairs or the PdO
double bond, as tentatively depicted in Scheme 3.
To verify which of the two possible pathways operated in
our reactions, the predicted rate laws for path B1 and path B2
were compared with the experimentally determined one.
For path B1, the following rate expression for the disappear-
ance of the PhPd(PPh3)2X complex can be derived:
kobs ) k[(EtO)2P(O)H]
(2)
One should note that although for the reactions investigated
base is an indispensable reaction component, it does not enter
into the kinetic equation. This, and the fact that using bases of
various strengths33 resulted only in mediocre changes in the
reaction rates (Figure 2), suggested that deprotonation of the
H-phosphonate diester occurred at a kinetically unimportant step
of the reaction (after the rate-determining step; also, Vide infra).
To ascertain further that the P-H bond breaking did not
influence the overall reaction rate, we measured a kinetic isotope
effect, using the deuterated H-phosphonate [(EtO)2P(O)D].34 The
obtained value of kH/kD ) 1.17 ( 0.05 was indeed very small
and excluded the possibility of a significant degree of P-H bond
breaking in the rate-determining step.
The rate constants measured for the complexes bearing
different halide ligands (Table 2) were in perfect agreement with
the previously reported qualitative results.28 Comparable values
of the rate constants for the reaction in toluene and THF (Table
2, entries 1 and 2) indicated no involvement of a solvent-assisted
substitution mechanism, and the lack of inhibitory effect of the
added PPh3 ruled out path C (Scheme 2) as a possible
mechanism.
d[PhPd(PPh3)2X]
-
)
dt
Kk1[(EtO)2P(O)H][base][PhPd(PPh3)2X]
(3)
[baseH+]
To evaluate a possible participation of the second dissociative
pathway (Scheme 2, path A) in the ligand substitution step, one
should take into account the affinity order of halide anions
Although, the assumed rapid equilibrium between the H-
phosphonate and its anion would explain the lack of a primary
(35) Flemming, J. P.; Pilon, M. C.; Borbulevitch, O. Y.; Antipin, M. Y.;
Grushin, V. V. Inorg. Chim. Acta 1998, 280, 87–98.
(36) Amatore, C.; Carre´, E.; Jutand, A.; M’Barki, M. A.; Meyer, G.
Organometallics 1995, 14, 5605–5614.
(37) Ozawa, F.; Kawasaki, N.; Okamoto, H.; Yamamoto, T.; Yamamoto,
A. Organometallics 1987, 6, 1640–1651.
(33) pKa values for the corresponding conjugated acids in water: Et3NH+
10.8, N-Me-morpholineH+ 7.4, pyridineH+ 5.2.
(34) The reactions with deuterated and nondeuterated substrates were
carried out using complex PhPd(PPh3)2I, at [(EtO)2P(O)H] ) [Et3N] ) 50
mM.

5880 Organometallics, Vol. 27, No. 22, 2008
Kalek and Stawinski
Scheme 3. Possible Associative Pathways for Ligand Substitution in PhPdL2X Complexes with an H-Phosphonate Diester as a
Nucleophile (L ) PPh3)
Table 3. Observed First-Order Rate Constants (kobs) of the Decay of
kinetic isotope effect for the reaction investigated, the rate law
(dppp)Pd(Ph)X Complexes at 40 °C in THF (initial [(dppp)Pd(Ph)X]
(3) was inconsistent with the experimental data since it predicted
(i) a first-order dependence on base concentration; (ii) a linear
dependence of the rate on the equilibrium constant K; and (iii)
an inverse dependence of the ligand exchange rates on concen-
tration of the conjugated acid (baseH+).38
) 5 mM, [(EtO)2P(O)H] ) [Et3N] ) 50 mM)
entry
palladium(II) complex
kobs (min-1
)
1
2
3
(dppp)Pd(Ph)I
(dppp)Pd(Ph)Br
(dppp)Pd(Ph)Cl
0.0581
0.1060
0.1697
For path B2 (Scheme 3), application of a steady-state
approximation led to the following rate law:
mol-1 K-1) determined from the Eyring plot39 supported the
associative character of this initial event in the ligand substitution
step.
d[PhPd(PPh3)2X]
-
)
dt
After having established the mechanistic basis for the ligand
exchange for halide-containing complexes with monodentate
ligands [trans-PhPd(PPh3)2X], we investigated the analogous
reaction in the presence of bidentate ligands, which are known
to force cis geometry of PdII complexes. As a model bidentate
ligand, we chose 1,3-bis(diphenylphosphino)propane (dppp), due
to its similar stereoelectronic properties to PPh3.40 In Table 3,
the observed rate constants for the decay of (dppp)Pd(Ph)X
complexes bearing various halide substituents, determined under
the same reaction conditions as those used for the PPh3
complexes, are shown. As apparent from the presented data,
the rate constants for ligand substitution in complexes bearing
this bidentate ligands were noticeably higher, but followed the
same order of reactivity as those for the PhPd(PPh3)2X
complexes (Table 2). These may suggest that also complexes
with bidentate ligands follow an analogous associative mecha-
nistic pathway of ligand substitution as that described above
for their monodentate phosphine counterparts.
Ligand Substitution in PhPdL2(OAc) Complexes. As it was
mentioned in the Introduction, the PhPd(PPh3)2(OAc) complex
exhibited a remarkably high reactivity in ligand substitution
reactions with H-phosphonate nucleophiles.28 This phenomenon
cannot be explained on the basis of the mechanism discussed
in the previous section for the halide complexes using the
electronegativity argument, and thus for the acetate complexes
probably a distinct mechanism was operating. Path A (Scheme
2), which involved the initial dissociation of an anion, seemed
to be unlikely due to similar affinity of OAc- and I- ions to
k1k2[(EtO)2P(O)H][base][PhPd(PPh3)2X]
k-1 + k2[base]
(4)
In this form, the rate expression (4) was inconsistent with
the experimental results (dependence on base concentration).
However, assuming that the proton abstraction step could be
much faster than the collapse of the complex [PhPd(PPh3)2X
+ (EtO)2P(O)H] back into the starting materials (k-1
k2[base]), then consistent with a small kinetic isotope effect
,
observed, eq 4 would simplify to
d[PhPd(PPh3)2X]
-
) k1[(EtO)2P(O)H][PhPd(PPh3)2X] (5)
dt
The resulting rate law (5) is identical to the experimentally
determined one (eq 1) and lends support for path B2, with the
rate-determining step being coordination of the H-phosphonate
diester by the PdII complex, as a possible mechanism for the
studied reaction.
Since the experimentally obtained rate constant k (eqs 1 and
2) corresponded to elementary second-order rate constant k1 for
nucleophilic attack of the H-phosphonate diester on the pal-
ladium center (Scheme 3), it was possible to determine
thermodynamic activation parameters for this step. The highly
negative value of the entropy of activation (∆S‡ ) -72.2 J
(38) The rates of ligand exchange reactions between PhPdL2X and
diethyl H-phosphonate diester were independent of the base concentration
(when used in over stoichiometric amount) and were fairly insensitive to
the base strength (Vide supra). If the reactions were affected by the presence
of the conjugated acids as predicted by eq 3, the first-order decay of
PhPdL2X, would not be observed due to increasing concentration of these
species during the course of the reaction.
(39) See the Supporting Information.
(40) Palladium reduction and oxidative addition processes have also been
studied in detail for complexes with dppp. See: Amatore, C.; Jutand, A.;
Thuilliez, A. Organometallics 2001, 20, 3241–3249.

Ligand Substitution from PdII Complexes
Organometallics, Vol. 27, No. 22, 2008 5881
Scheme 4. Plausible Pathways for the Ligand Substitution in the PhPdL2(OAc) Complex with an H-Phosphonate Diesters as a
Nucleophile (L ) PPh3)
the palladium(II) center.36 One should, however, consider the
other dissociative pathway (path C, Scheme 2), but in a slightly
modified version, namely, with the carbonyl oxygen atom of
the acetate group acting as an intramolecular nucleophilic
catalyst in substitution of one of the phosphine ligands (Scheme
4). The intermediate, bearing a bidentate acetate group, formed
in this reaction step, would then coordinate directly the
H-phosphonate diester (Scheme 4, path B) or react with its anion
(Scheme 4, path A) to give a ligand substitution product. An
intramolecular nucleophilic catalysis by the acetate group would
explain high rate acceleration observed in such reactions, due
to favorable entropy changes, similarly as during neighboring
group participation processes. Feasibility of this reaction
pathway is supported by a number of reports describing metal
complexes bearing bidentate (κ2) acetate or amidate ligands,
which have been observed in solution, solid state, and as reaction
intermediates.41 The involvement of analogous species was also
claimed for certain palladium-catalyzed Heck reactions.9,36
approximation to the reaction sequence in path A, the following
rate law for the decay of the PhPd(PPh3)2(OAc) complex was
obtained:
d[PhPd(PPh3)2(OAc)]
-
)
dt
[(EtO)2P(O)H][base]
k1k2K
[PhPd(PPh3)2(OAc)]
[baseH+]
[baseH+]
(6)
(7)
[(EtO)2P(O)H][base]
k-1[PPh3] + k2K
k-1[PPh3][baseH+]
1
1
)
+
kobs k1k2K[(EtO)2P(O)H][base] k1
Equation 7 correctly predicts the relationships of 1/kobs versus
[PPh3] (Figure 4) and versus 1/[(EtO)2P(O)H] (Figure 5).
However, as for the corresponding halide complexes (Scheme
3, path B1, and eq 3), the existence of the pre-equilibrium
between the H-phosphonate diester and its anion implied that
also a base concentration entered into the rate law (7), and thus
the relationships 1/kobs versus 1/[base], should be identical to
that of 1/kobs versus 1/[(EtO)2P(O)H] (Figure 5). Since the rate
of the reaction investigated was not affected by changes in the
concentration of Et3N (Figure 6), path A in Scheme 4 was
inconsistent with our kinetic data.
To verify a possible intermediacy of bidentate acetate
complexes in ligand substitution involving the palladium acetate
complexes (Scheme 4, paths A and B), we measured rates of
the disappearance of PhPd(PPh3)2(OAc) as a function of PPh3,
(EtO)2P(O)H, and Et3N concentrations. Due to high reaction
rates, the measurements were conducted at 15 °C.
It was found that with an excess of PPh3, (EtO)2P(O)H, and
Et3N, the decay of PhPd(PPh3)2(OAc) followed first-order
kinetics (31P NMR experiments). A plot in Figure 4, 1/kobs versus
concentration of PPh3, showed a linear relationship, with a
positive slope and nonzero y-intercept, which indicated inhibition
of the reaction by the phosphine ligand. The plot of 1/kobs versus
1/[(EtO)2P(O)H] was also linear, with a nonzero y-intercept
(Figure 5), and the reaction was found to be independent of the
base (Et3N) concentration (Figure 6).
To find out which of the reaction pathways in Scheme 4 was
compatible with the experimental kinetic data, we derived rate
laws for path A and path B. After application of a steady-state
(41) (a) Tackett, J. E. Appl. Spectrosc. 1989, 43, 483–489. (b) Harvey,
M.; Baggio, S.; Baggio, R.; Mombru, A. W. Acta Crystallogr. 1999, C55,
308–310. (c) Ikawa, T.; Barder, T. E.; Biscoe, M. R.; Buchwald, L. S. J. Am.
Chem. Soc. 2007, 129, 13001–13007. (d) Fujita, K. i.; Yamashita, M.;
Puschmann, F.; Alvarez-Falcon, M. M.; Incarvito, C. D.; Hartwig, J. F.
J. Am. Chem. Soc. 2006, 128, 9044–9045.
Figure 4. Plot of 1/kobs versus PPh3 concentration for the decay of
PhPd(PPh3)2(OAc) in THF at 15 °C (initial [PhPd(PPh3)2(OAc)]
) 5 mM, [(EtO)2P(O)H] ) [Et3N] ) 50 mM).

5882 Organometallics, Vol. 27, No. 22, 2008
Kalek and Stawinski
Scheme 4), we determined the thermodynamic activation
parameters for the initial step of the reaction, namely, the
dissociation of the phosphine ligand (Scheme 4). To this end,
k1 rate constants were determined at three temperatures (5 and
10 °C in addition to 15 °C; from intercepts of the plots 1/kobs
versus PPh3)39 and plugged into the Eyring equation.39 A highly
positive entropy of activation (∆S‡ ) 53.3 J mol-1 K-1
)
obtained confirmed a dissociative character of this key step of
the reaction.
Finally, we attempted to measure the ligand substitution rate
for a complex containing an acetate group and a bidentate ligand,
(dppp)Pd(Ph)(OAc). Unfortunately, the reaction was too fast
to be followed by means of 31P NMR spectroscopy (completion
in <3 min). Anyway, the very high rate for this reaction
suggested that probably a mechanism analogous to that described
above (Scheme 4, path B) operated also for this bidentate
phosphine complex. Such a mechanism would imply that, at
least at some stages of the reaction, only one phosphorus atom
of the dppp ligand would be engaged in the complexation.
Recent studies on reactivity of dppf palladium(II) complexes
with sulfur nucleophiles16 lend support for such a scenario (also,
Vide infra).
Figure 5. Plot of 1/kobs vs 1/[(EtO)2P(O)H] for the decay of
PhPd(PPh3)2(OAc) in THF at 15 °C (initial [PhPd(PPh3)2(OAc)]
) 5 mM, [PPh3] ) [Et3N] ) 50 mM).
Intermediate Phosphonate PdII Complexes Bearing
PPh3 Ligands and the C-P Bond-Forming Reductive
Elimination Step. The so far discussed results focused on a
ligand substitution step and revealed various ways in which
palladium(II) complexes, formed in the course of oxidative
addition, might react with H-phosphonate diesters, depending
on the anion (Cl-, Br-, I-, AcO-) initially ligated to the PdII
center. However, an important issue that has not been addressed
yet was the structure of palladium phosphonate complexes
formed in ligand substitution steps, from which a reductive
elimination of diethyl phenylphosphonate occurred (Scheme 5).
Figure 6. Dependence of the rate of PhPd(PPh3)2(OAc) decay on
the concentration of Et3N in THF at 15 °C (initial [PhPd-
(PPh3)2(OAc)] ) 5 mM, [(EtO)2P(O)H] ) [PPh3] ) 50 mM).
Scheme 5
Instead, the second mechanistic pathway, namely, path B
(Scheme 4), turned out to be in good agreement with the results
shown in Figures 4-6:
d[PhPd(PPh3)2(OAc)]
-
)
dt
In the instance of halide-containing complexes (Scheme 2,
path B), the possible square-planar palladium phosphonate
products may be neutral (a halide anion departs and two PPh3
ligands remain; two isomers) or anionic species (a halide anion
stays and PPh3 departs; three isomers). For the PdII complex
with an acetate group, the square-planar palladium phosphonate
complex formed (Scheme 4) should be an anionic species with
three different possible configurations. It was also possible that,
before the reductive elimination took place, a ligand exchange
and isomerization of palladium phosphonate complexes oc-
curred. These could be induced by PPh3, halide, or acetate ions
present in the reaction mixture and would add further to the
complexity of the intermediate complexes formed. Some of these
intermediates might be formed preferentially, either on kinetic
or thermodynamic grounds, but only those having phenyl and
phosphonate ligands in cis configuration should be able to
undergo reductive elimination.20,21 The latter issue could be quite
complex for cross-coupling reactions performed in the presence
of monodentate ligands, such as PPh3.
k1k2[(EtO)2P(O)H][PhPd(PPh3)2(OAc)]
k-1[PPh3] + k2[(EtO)2P(O)H]
(8)
(9)
k-1[PPh3]
kobs k1k2[(EtO)2P(O)H] k1
1
1
)
+
For this mechanism, base concentration was not present in
the rate law (8), since the P-H bond breaking occurred after
the rate-determining step and thus was kinetically insignificant.
Equation 9 correctly predicted also the observed linear depend-
encies of 1/kobs versus [PPh3] and versus 1/[(EtO)2P(O)H].
Nonzero y-intercepts for these reactions, which corresponded
to 1/k1, indicated that the rate law (8) could not be simplified
by dropping the k2[(EtO)2P(O)H] term in the denominator. In
mechanistic terms this means that the intermediate bidentate
acetate complex (Scheme 4) reacted with comparable rates both
“backward” with PPh3 and “forward” with the H-phosphonate.
Actually, by plugging into eq 9 the measured values of kobs
,
the known concentrations of the reactants, and k1 (from the
intercept), the k2/k-1 ratio could be calculated. The obtained
value of ∼0.9 showed that the transient bidentate acetate
complex formed was not selective toward different nucleophiles,
as one would expect for a highly reactive intermediate.
To support further the proposed mechanism for the ligand
substitution in the PhPd(PPh3)2(OAc) complex (path B in
To obtain mechanistic insight into the reductive elimination
step, we decided to generate palladium phosphonate intermedi-
ates from PhPd(PPh3)2X (X ) I, Br, Cl, OAc) complexes using
different phosphorus-nucleophiles and to follow both the product
formation and the decay of the palladium phosphonate inter-
mediates by 31P NMR spectroscopy (Schemes 5 and 6).

Ligand Substitution from PdII Complexes
Organometallics, Vol. 27, No. 22, 2008 5883
Scheme 6. Rapid in Situ Generation of the Equilibrating
PdII Phosphonate Complexes via the Reaction of PhPd-
(PPh3)2X with (EtO)2PONa and the Following Reductive
Elimination (L ) PPh3)
Analysis of the cross-coupling reactions involving phosphorus
nucleophiles by 31P NMR spectroscopy was facilitated by the
fact that phosphine ligands present in different complexes, and
also the reactants and the products, could be observed simul-
taneously. In Figures 7-10, the filled squares represent the first-
order decays of PhPd(PPh3)2X complexes in the presence of
excess (EtO)2P(O)H and Et3N, and the filled diamonds in the
same figures represent the amounts of the reductive elimination
product [(EtO)2P(O)Ph)] formed.
Figure 9. Reaction of PhPd(PPh3)2Cl with (EtO)2P(O)H in THF at
40 °C (initial [PhPd(PPh3)2Cl] ) 5 mM, [(EtO)2P(O)H] ) [Et3N]
) 50 mM).
As apparent from the data for PhPd(PPh3)2Br and
PhPd(PPh3)2Cl complexes (Figures 8 and 9, respectively), the
reductive elimination product formation lagged behind the
disappearance of the starting palladium complexes; that is,
the amount of diethyl phenylphosphonate formed after a given
time was less than expected from the substrate consumption.
What was important, in the 31P NMR spectra of these reaction
mixtures, there were no other signals than those from
PhPd(PPh3)2X, (EtO)2P(O)H, (EtO)2P(O)Ph, and PPh3-ligated
Pd0 species. The “missing” amount of the cross-coupling product
(calculated as the difference of the integrals of the (EtO)2P(O)Ph
and PhPd(PPh3)2X signals), which should correspond to
Figure 10. Reaction of PhPd(PPh3)2(OAc) with (EtO)2P(O)H in
THF at 15 °C (initial [PhPd(PPh3)2(OAc)] ) 5 mM, [(EtO)2P(O)H]
) [Et3N] ) 50 mM).
“intermediates” in Scheme 5, is shown in the graphs as open
triangles. The fact that these intermediates, although present in
a relatively large concentration, were not detectable by the 31
P
NMR spectroscopy could be due to their involvement in a
complex equilibria system and extensive splitting pattern of the
resonances (due to coupling of nonequivalent phosphorus atoms
of phosphine and phosphonate ligands), which could ultimately
lead to broadening of the signals into the baseline. A similar
phenomenon, ascribed to a dynamic solution behavior of square-
planar palladium complexes, was already reported for PdII
phosphonate species containing triarylphosphines.19 Interest-
ingly, this effect of accumulation of equilibrating palladium
phosphonate intermediates was observed mainly for PhPd(PPh3)2-
Br and PhPd(PPh3)2Cl complexes (Figures 8, 9) and to a smaller
extent for PhPd(PPh3)2I (Figure 7).
Figure 7. Reaction of PhPd(PPh3)2I with (EtO)2P(O)H in THF at
40 °C (initial [PhPd(PPh3)2I] ) 5 mM, [(EtO)2P(O)H] ) [Et3N] )
50 mM).
Due to expected difficulties in dealing with a series of
sequential reactions that could lead to complex kinetic expres-
sions, we decided to study the reductive elimination from an
equilibrium mixture of palladium phosphonates, rapidly gener-
ated from PhPd(PPh3)2X and a powerful phosphorus nucleo-
phile, namely, sodium diethyl phosphite [(EtO)2PONa]. Such a
technique has been used for studying other reductive elimination
processes.13-15,17,19
To this end, to a solution of PhPd(PPh3)2I in THF containing
PPh3 (the presence of PPh3 was necessary to avoid Pd black
precipitation, Vide supra) was added 0.9 equiv of (EtO)2PONa,
and progress of the reaction was followed by 31P NMR
spectroscopy. The first spectrum recorded (<1 min) revealed
the presence of only a sharp, residual signal of PhPd(PPh3)2I
(∼10%), while the resonances due to (EtO)2PONa and PPh3
Figure 8. Reaction of PhPd(PPh3)2Br with (EtO)2P(O)H in THF
at 40 °C (initial [PhPd(PPh3)2Br] ) 5 mM, [(EtO)2P(O)H] ) [Et3N]
) 50 mM).

5884 Organometallics, Vol. 27, No. 22, 2008
Kalek and Stawinski
served lag in the product formation when X ) Cl and Br. This
was consistent with the fact that rates of coordination of
(EtO)2P(O)H by PhPd(PPh3)2X (Scheme 5, the first step)
decreased in the order X ) Cl > Br > I (Table 2), while those
of the reductive elimination (Scheme 5, the second step)
decreased in an opposite direction (I > Br > Cl). For X ) Cl
or Br, the intermediate PdII phosphonate complexes were quickly
formed, but collapsed slowly to the product, whereas for X )
I, formation of the intermediate complexes was slower than the
subsequent reductive elimination step.
According to theoretical studies, the concerted reductive
elimination is symmetry allowed and can occur from both
square-planar (sq)21 and trigonal bipyramid (tbp)43 complexes,
provided that the eliminated ligands occupy the adjacent
positions (cis in sq and equatorial-apical in tbp complexes).
Assuming that the phenyl group and a phosphonate ligand
cannot dissociate from palladium(II) complexes, the postulated
equilibria of palladium phosphonate species were thus due to
exchange of halide anions, PPh3, and/or spatial rearrangements
of the ligands. There are 5 sq and 11 tbp reasonable (neutral or
monoanionic) structures possible, interconverting via 13 transient
square-pyramids.39 Nine of them (3 sq and 6 tbp) fulfill the
geometrical requirements for a reductive elimination; however,
due to stereoelectronic effects, these structures may differ
significantly in their ability to undergo reductive elimination.
From electronic and structural points of view, more suscep-
tible to reductive elimination are complexes with low electron
density and bulky spectator ligands present.44 In light of this,
pentacoordinate palladium complexes containing halides should
exhibit rather low reactivity in reductive elimination, due to both
net negative charge present and small steric demands for
spherical halide anions. However, if the halide anion is expelled,
the resulting complexes, ligated only by phenyl, a phosphonate,
and 2 or 3 PPh3 molecules (2 sq and 3 tbp structures), should
be more susceptible to reductive elimination. We believe that
the observed trend in reactivity of PdII phosphonate complexes
in reductive elimination as a function of the halide present, i.e.,
I > Br > Cl, originated from the different affinity of these anions
toward the PdII center. A weakly bound I- was expected to be
more easily replaced by PPh3 than a strongly bound Cl-, and
in the presence of excess phosphine ligand, the equilibrium
should be shifted toward the neutral, more sterically hindered
species, from which reductive elimination should be facilitated.
Finally, there was a case of a remarkably fast reductive
elimination when the acetate group was present in the precursor
complex [PhPd(PPh3)2(OAc); Table 4, entry 7]. Since this
reaction was more than an order of magnitude faster than that
for the most reactive halide derivative [PhPd(PPh3)2I; Table 4,
entries 1 and 4] it seemed that a distinct mechanism might
operate for the acetate-containing palladium complexes. For
PhPd(PPh3)2(OAc), the addition of (EtO)2PONa triggered a rapid
reaction, with only a small lag in the diethyl phenylphosphonate
formation. This indicated that both steps of the reaction in
Scheme 6, i.e., the reaction of PhPd(PPh3)2(OAc) with the
phosphorus nucleophile and the reductive elimination from the
intermediate palladium phosphonate complexes formed, were
fast.
Figure 11. Generation of the equilibrating PdII phosphonate
complexes by the addition of (EtO)2PONa to a solution of
PhPd(PPh3)2I and the following reductive elimination of
(EtO)2P(O)Ph in THF at 40 °C (initial [PhPd(PPh3)2I] ) 5 mM).
Table 4. Observed First-Order Rate Constants (kobs) of the
Reductive Elimination of (EtO)2P(O)Ph from the Mixtures of
Equilibrating PdII Phosphonate Complexes, Generated from
Different PhPd(PPh3)2X Precursors and (EtO)2PONa, in THF at
40 °C (initial [PhPd(PPh3)2X] ) 5 mM)
entry PhPd(PPh3)2X precursor and amount of PPh3 added kobs (min-1
)
1
2
3
4
5
6
7
PhPd(PPh3)2I + PPh3
0.0141
0.0077
0.0074
0.0346
0.0291
0.0121
0.7412
PhPd(PPh3)2Br + PPh3
PhPd(PPh3)2Cl + PPh3
PhPd(PPh3)2I + 10PPh3
PhPd(PPh3)2Br + 10PPh3
PhPd(PPh3)2Cl + 10PPh3
PhPd(PPh3)2(OAc) + PPh3
disappeared completely (Scheme 6).42 Such an appearance of
the 31P NMR spectrum suggested that the expected equilibrium
mixture of palladium phosphonate complexes was, indeed,
rapidly formed and that free PPh3 [but not PhPd(PPh3)2I] was
engaged into this equilibria system. After ca. 10 min, a signal
of the expected reductive elimination product, (EtO)2P(O)Ph,
started to emerge, together with a broad peak from PPh3 ligated
to Pd0 species. The formation of (EtO)2P(O)Ph displayed first-
order kinetics (Figure 11), and the corresponding rate constant
could be determined (Table 4, entry 1). A similar behavior was
observed for the other PhPd(PPh3)2X complexes (X ) Br, Cl,
and OAc),39 and the rate constants for a reductive elimination
from these complexes are listed in Table 4. Also, the effect of
added PPh3 on rates of the reductive eliminations was investi-
gated (Table 4).
The data in Table 4 revealed two important features of the
reductive elimination process. First, the rate of a reductive
elimination depends on the anion present in the precursor
complex, and second, the addition of PPh3 accelerates the
reaction. These findings suggested that a reductive elimination
could occur from pentacoordinate intermediates (probably, tbp)
containing a halide anion or that the departure of a halide anion
to form a square-planar complex, from which the elimination
occurs, was the rate-determining step for the process.
The trend of reactivity found for the halides (Table 4, entries
1-3) explained the postulated accumulation of the intermediate
palladium phosphonate complexes during reactions of
PhPd(PPh3)2X with (EtO)2P(O)H (Figures 7-9) and the ob-
The high reactivity of PhPd(PPh3)2(OAc) in ligand substitu-
tion was discussed above, and it was assigned to generation of
(42) A rapid reaction of (EtO)2PONa indicates that the H-phosphonate
anion can efficiently act a nucleophile in the ligand substitution at the PdII
center. This means that path B1 in Scheme 3 is a possible mechanism for
the reaction. However, no involvement of this mechanism when Et3N,
N-methylmorpholine, and pyridine were used as bases shows that these bases
generate only negligible amount of the H-phosphonate anion, and thus under
standard reaction conditions apparently only path B2 is operating.
(43) Braterman, P. S.; Cross, R. J.; Young, G. B. J. Chem. Soc., Dalton
Trans. 1976, 1306–1310.
(44) Zuidema, E.; van Leeuwen, P. W. N. M.; Bo, C. Organometallics
2005, 24, 3703–3710.
(45) Amatore, C.; Carre, E.; Jutand, A.; M’Barki, M. A. Organometallics
1995, 14, 1818–1826.

Ligand Substitution from PdII Complexes
Organometallics, Vol. 27, No. 22, 2008 5885
Scheme 7. Putative Penta- and Tetracoordinate Palladium
Phosphonate Species Formed from the PhPd(PPh3)2(OAc)
Complexa
Scheme 8. In Situ Generation of (dppp)Pd(Ph)(PO(EtO)2)
Complex via the Reaction of (dppp)Pd(Ph)X (X ) I, Br, Cl)
with (EtO)2PONa and the Following Reductive Elimination
Table 5. Observed First-Order Rate Constants (kobs) for the
Reductive Elimination of (EtO)2P(O)Ph from the Palladium
Phosphonate Intermediate, Generated from (dppp)Pd(Ph)X and
(EtO)2PONa, in THF at 40 °C (initial [(dppp)Pd(Ph)X] ) 5 mM)
entry (dppp)Pd(Ph)X precursor and amount of PPh3 added kobs (min-1
)
1
2
3
4
5
(dppp)Pd(Ph)I + PPh3
(dppp)Pd(Ph)Br + PPh3
(dppp)Pd(Ph)Cl + PPh3
(dppp)Pd(Ph)I + 10PPh3
(dppp)Pd(Ph)(OAc) + PPh3
0.0568
0.0499
0.0500
0.0497
0.4013
tation. Two tbp’s from which the reductive elimination is
permitted are marked with dashed rectangles in Scheme 7. Due
to steric demands for a bulky PPh3, this ligand should show
preference for an equatorial position in tbp, and thus formation
of tbp species with an apical-equatorial arrangement of the
groups to be eliminated should be energetically favored. Third,
tbp’s bearing κ2-acetate in the equatorial-apical position are
expected to be distorted, and this may facilitate their collapse,
with expulsion of the reductive elimination product.
Intermediate Phosphonate PdII Complexes Bearing
dppp Ligand and the C-P Bond-Forming Reductive
Elimination Step. To clarify certain aspects of the reductive
elimination steps studied, we conducted additional experiments
on complexes bearing a bidentate ligand (dppp). In this instance,
in contradistinction to the monodentate complexes of type
PhPd(PPh3)2X, treatment of (dppp)Pd(Ph)X (X ) I, Br, Cl)
complexes with (EtO)2PONa provided palladium phosphonate
species with a well-defined 31P NMR characteristic (three groups
of resonances in the region of chemical shifts of ca. 0, 4, and
82 ppm). This complex was formed irrespective of the halide
present in the parent compound (Scheme 8), and on the basis
of the chemical shift values and the splitting pattern of the
signals, its structure was assigned as (dppp)Pd(Ph)(PO(EtO)2).39
a
Complexes from which reductive elimination is allowed are marked
with dashed rectangles (charges are omitted for clarity).
a highly reactive κ2-acetate species (Scheme 4). However, the
enhanced reactivity observed in the reductive elimination step
of the acetate-containing complexes was less obvious, and we
ascribed it to some unique features of the generated palladium
phosphonate intermediates bearing an acetate group. Since 31
P
NMR spectroscopy did not provide any insight into chemical
structures of these intermediates, we could only speculate that
coordination of (EtO)2PONa to a κ2-acetate species should result
in formation of negatively charged, pentacoordinate (tbp)
intermediates, in which the κ2-acetate group might span the
equatorial and the apical position of tbp (Scheme 7). Since it
was rather unlikely that a κ2-ligand could be expelled from the
complex, we considered two scenarios for the reductive elimina-
tion step. In one of them, from the pool of generated pentaco-
ordinate intermediates, four square-planar palladium phospho-
nates could be formed (Scheme 7). A reductive elimination of
the product (diethyl phenylphosphonate), with generation of
LPd0(OAc)-45 or κ2-Pd0(OAc)- species, might occur from three
of these interconverting (via pentacoordinate species) complexes.
In the second scenario (Scheme 7), reductive elimination of
diethyl phenylphosphonate might take place directly from
pentacoordinate palladium phosphonate intermediates. In our
opinion, reductive elimination from pentacoordinate intermedi-
ates containing an acetate group is a viable option on three
counts. First, due to expected positional preference, the κ2-
acetate group should occupy a equatorial-apical position in tbp,
and thus the number of tbp’s formed, should be limited to three.
Second, from the initial sp complex bearing a phosphonate group
in the vertex position, two tbp’s (both with a κ2-acetate group
spanning the equatorial and the apical position) could be formed.
One of them would have the groups to be eliminated in a
favorable position for reductive elimination, an equatorial-apical
position, while from the other tbp, such an arrangement of the
ligands could be attained after one pseudorotation, using either
the phenyl or the phosphonate moiety as a pivot of pseudoro-
Apparently, the entropically driven chelating effect of the
dppp ligand made this PdII phosphonate complex by far less
fluxional compared to complexes with monodentate PPh3
ligands, and formation of this complex was in agreement with
the earlier reports on analogous PdII phosphonate complexes
bearing bidentate ligands.17-19,46
For the (dppp)Pd(Ph)(PO(EtO)2) complex, generated from
different precursors, a reductive elimination of diethyl phe-
nylphosphonate was investigated, and the first-order rate
constants were determined. These are listed in Table 5, together
with the rate constant for the reaction when a 10 M excess of
PPh3 was used (Table 5, entry 4).
For all the precursor complexes bearing different halides,
similar rate constants were obtained. Since the addition of extra
PPh3 did not influence significantly the rate of this reaction, it
(46) Stockland, R. A.; Maher, D. L.; Anderson, G. K.; Rath, N. P.
Polyhedron 1999, 18, 1067–1075.

5886 Organometallics, Vol. 27, No. 22, 2008
Kalek and Stawinski
Scheme 9. Proposed Catalytic Cycle for the Palladium-Mediated Cross-Coupling of H-Phosphonates with Aryl Halidesa
a
Structures to the left and to the right refer to PPh3 or a bidentate ligand, respectively. Complexes’ charges are omitted for clarity.
seemed that the reductive elimination probably occurred directly
Mechanisms for the Palladium-Catalyzed C-P Bond
Formation. The above mechanistic studies on ligand substitu-
tion and reductive elimination, appended with the existing
knowledge about oxidative addition, lend themselves to a more
detailed picture of two catalytic cycles for the palladium-
mediated C-P bond formation, which are summarized in
Schemes 9 and 10.
Scheme 9 depicts a catalytic cycle for the reaction in which
halide anions are involved. This is by far the most common
situation for cross-coupling reactions using aryl halides, since
even in the absence of external halide additives, these anions
are released during the course of the reaction and can interact
with various palladium species at different stages of the cycle.
from this cis-configured (dppp)Pd(Ph)(PO(EtO)2) complex.
As to the acetate-containing complex (dppp)Pd(Ph)(OAc),
also in this instance it behaved differently form the halide
complexes of type (dppp)Pd(Ph)X. First, it reacted an order of
magnitude faster than the other bidentate complexes investigated
(Table 5, entry 5), and second, the expected palladium phos-
phonate (dppp)Pd(Ph)(PO(EtO)2) could not be detected by 31
P
NMR spectroscopy. Instead, upon addition of (EtO)2PONa to
the reaction mixture, the signal from the (dppp)Pd(Ph)(OAc)
complex disappeared immediately, and a fast formation of the
reductive elimination product [(EtO)2P(O)Ph] was observed.
This phenomenon appeared to be similar to that observed for
the reaction of PhPd(PPh3)2X complexes with (EtO)2PONa and
pointed to the formation of an equilibrium mixture of palladium
phosphonate complexes.
Although structures of such complexes remained cryptic, one
can speculate that the acetate group in (dppp)Pd(Ph)(OAc) can
open the six-membered ring system of the dppp ligand and form
a highly reactive bidentate (κ2) acetate species, analogous to
that shown in Scheme 4. This could coordinate the phosphorus
nucleophile (EtO)2PONa and generate a mixture of pentacoor-
dinate palladium phosphonate intermediates. However, in
contrast to the halide-containing complexes, which can expel a
halide anion and collapse to (dppp)Pd(Ph)(PO(EtO)2), com-
plexes with κ2-acetate ligands probably could not eliminate the
acetate group and thus remained highly fluxional. Since the
presence of a κ2-acetate ligand in pentacoordinate complexes
will confer significant distortions to tbp geometry, the reactivities
of such complexes are expected to be high, and a reductive
elimination may thus occur directly from these species (Vide
supra).
The first step, oxidative addition of an aryl halide to Pd0
catalyst A, results in formation of arylpalladium complexes B.
Reactivity of the palladium catalyst A in this step depends on
the halide anions ligated to this species (X1 in Scheme 8) and
decreases in the order Cl > Br > I. Additional ligation of PPh3
or dba (G1) lowers the reactivity of catalytic species A.9,47 The
next step, a reaction of arylpalladium B with a nucleophile,
usually referred to as a ligand substitution, in the case of
phosphorus nucleophiles is better described as a nucleophile
coordination step. This affords pentacoordinate species C, which
has to undergo deprotonation by a base to form penta- or four-
coordinated palladium phosphonate complexes D. This pathway
is followed by H-phosphonate diesters and related phosphorus
nucleophiles that do not have a lone electron pair on the
phosphorus atom. The reactivity of PdII complexes B, irrespec-
tive of denticity of the supporting ligand used (mono- or
bidentate), follows the same trend as that for the oxidative
addition; namely, the reaction is fastest when X2 ) Cl, followed
by Br and I. These two steps of the cycle are responsible for
the overall acceleration of the cross-coupling reaction when
chlorides or bromides are added to the reaction mixture (Table
1, entries 2, 3, 5, and 6).
Irrespective of the mechanistic details, which remain to be
clarified, an important feature of the acetate group as a ligand
in palladium complexes is its ability to form bidentate (κ2)
acetate species, both for square-planar and pentacoordinate
complexes, which may have enhanced reactivity in ligand
substitution and reductive elimination reactions.
(47) Amatore, C.; Jutand, A. Coord. Chem. ReV. 1998, 178-180, 511–
528.

Ligand Substitution from PdII Complexes
Organometallics, Vol. 27, No. 22, 2008 5887
Scheme 10. Proposed Catalytic Cycle for the Palladium-Mediated Cross-Coupling of H-Phosphonates with Aryl Halides,
Involving Acetate As a Liganda
a
Structures to the left and to the right refer to PPh3 or a bidentate ligand, respectively. Complexes’ charges are omitted for clarity.
A nucleophile coordination step, which produces pentacoor-
dinate complex C, is usually slow compared to the subsequent
deprotonation, and thus the strength of a base used for the
reaction and its concentration practically do not significantly
affect the rates of the palladium phosphonates D formation.48
Importantly, the whole incorporation process of a phosphonate
ligand into complexes D is not influenced by PPh3 ligand
concentration.
For phosphorus nucleophiles that do not require deprotonation
[e.g., (EtO)2PONa], direct nucleophilic attack on PdII complexes
B is possible.49 This reaction is very fast and produces palladium
phosphonates D. Now, depending on the denticity of the
supporting ligand used, there are two scenarios. For monodentate
ligand PPh3, the PdII phosphonates D are involved in a complex
equilibria system of penta- and tetracoordinate species, powered
by the halide anions and PPh3 present. Reductive elimination
of the product Ar-P(O)(OR)2 from such an equilibrating
mixture of palladium phosphonate complexes is highest if I-
ions are engaged in the equilibria and somewhat slower for Br-
and Cl-. The added PPh3 accelerates further this reductive
elimination.
at different stages of a cross-coupling reaction, and this is shown
in Scheme 10.
The acetates can be brought to the reaction mixture together
with the palladium source [e.g., Pd(OAc)2] or may be introduced
as an external additive. Due to rather weak affinity of the acetate
ion to Pd0 (Cl- > Br- > OAc- > I-) or PdII (Cl- > Br-
>
OAc- ∼ I-) centers, the halides released during the course of
a cross-coupling reaction may effectively compete with acetate
ions, and ultimately, the reaction may switch to the “halide
cycle” (Scheme 9). This is usually manifested in a gradual
slowing of the cross-coupling reaction. However, if the reaction
is carried out in the presence of externally added acetate anions
(e.g., n-Bu4NOAc), a remarkable shortening in the reaction time
is observed (Table 1, entry 7).
In this acetate type of catalytic cycle, the oxidative addition
of aryl halides to Pd0 catalyst A bearing an acetate group is
slower than a reaction with a palladium catalyst with coordinated
halides (Scheme 9) and is also inhibited by the added PPh3.50
The observed overall acceleration of cross-coupling reactions
upon addition of acetate anions is due to mechanistic changes
in steps leading from arylpalladium B to palladium phosphonate
E complexes (collectively referred to as a ligand substitution)
and in the reductive elimination step.
The presence of an acetate group in PdII complexes B
(Scheme 10), formed in oxidative addition, opens a new
mechanistic pathway for the reaction with H-phosphonate
diesters. Namely, the acetate may act as an internal nucleophile
and expel the phosphine ligand, producing highly reactive
intermediates of type C, with a κ2-acetate ligand. From this
species, after coordination of an H-phosphonate diester to
produce a pentacoordinate complex D, and its subsequent
deprotonation, an equilibrium mixture of palladium phospho-
nates E is formed. Since κ2-acetate complexes C and arylpal-
ladium B are in equilibrium, and a nucleophile coordination to
For bidentate phosphines used as supporting ligands, less
fluxional, tetracoordinate PdII phosphonate complexes D, with
defined cis configuration of an aryl and a phosphonate group,
are formed. Rates of reductive elimination from such complexes
are not affected by the kind and concentration of the halide
anions present, nor the added PPh3.
The above catalytic cycle is modified to large extent when
the acetate ion is involved in complexation of palladium species
(48) Base is, however, an indispensable reactant component and has to
be used in stoichiometric amounts in cross-coupling reactions involving
H-phosphonate diesters.
(49) This pathway is not observed for cross-coupling reactions of dialkyl
H-phosphonate diesters in the presence of a base, due to negligible
concentration of the corresponding H-phosphonate anions under the reaction
conditions. However, for more acidic H-phosphonate or H-phosphinate
derivatives, this pathway cannot be excluded.
(50) Jutand, A.; Mosleh, A. Organometallics 1995, 14, 1810–1817.

5888 Organometallics, Vol. 27, No. 22, 2008
Kalek and Stawinski
C is apparently a rate-determining step (cf. Scheme 4), the
generation of palladium phosphonate complexes E is inhibited
by the added PPh3. Due to high reactivity of κ2-acetate
complexes C, the overall incorporation of a phosphonate moiety
to form palladium phosphonates E is very fast compared to that
in the halide type of catalytic cycle (Scheme 9). For phosphorus
nucleophiles, which do not require prior deprotonation, an
alternative, direct path from arylpalladium complexes B to
palladium phosphonates E is also available. Irrespective of the
reaction path involved and denticity of the supporting ligands
used, palladium phosphonates E in this instance are formed as
an equilibrium mixture of penta- and tetracoordinate species.
The reductive elimination from such palladium phosphonate
complexes is extremely fast, most likely due to the ability of
the acetate group to form highly reactive, pentacoordinate
palladium κ2-acetate species.
for PhPd(PPh3)2X and involved the intermediacy of highly
reactive κ2-acetate complexes.
Rates of a reductive elimination of diethyl phenylphosphonate
from palladium phosphonate intermediates depended strongly
on the kind of arylpalladium complexes used for their genera-
tion, and for PhPd(PPh3)2(OAc), it was an order of magnitude
higher than that observed for the halide-containing complexes.
These findings, together with kinetic experiments, enabled
us to clarify some important mechanistic features of an acetate
group present in palladium(II) complexes and to propose two
catalytic cycles for the formation of the C-P bond via a
palladium-mediated cross-coupling of aryl halides with H-
phosphonate diesters, depending on whether halides versus
acetates were dominant ligands.
Studies on synthetic applications of a remarkable rate
enhancement of the C-P bond formation by acetate ions are in
progress in our laboratory.
Conclusions
Acknowledgment. Financial support of the Swedish
National Research Council is gratefully acknowledged.
In summary, we investigated in detail the ligand substitution
and the reductive elimination steps of the palladium-catalyzed
cross-coupling reaction of aryl halides with diethyl H-phospho-
nate under various experimental conditions. It was found that
what was usually referred to as a ligand substitution step for
PhPd(PPh3)2X (X ) Cl, Br, I) complexes involved a coordina-
tion of an H-phosphonate diester in a rate-determining step,
followed by deprotonation to form an equilibrium mixture of
penta- and tetracoordinated palladium phosphonate intermediates.
Generation of palladium phosphonate intermediates from
PhPd(PPh3)2(OAc) was significantly faster than that observed
Supporting Information Available: Experimental synthetic
procedures, protocols for the kinetic measurements, raw kinetic data,
the Eyring plots and a graph used to obtain k1 values, full
isomerization scheme for phosphonate complexes with PPh3 as
ligand, and 31P NMR spectra of in situ generated (dppp)Pd(Ph)-
(PO(EtO)2) complex. This material is available free of charge via
the Internet at http://pubs.acs.org.
OM800641N