Optimization of procedures for the syntheses of bisphosphinepalladium(O) precursors for suzuki-miyaura and similar cross-coupling catalysis: Identification of 3:1 coordination compounds in catalyst mixtures containing Pd(0), PCy3, and/or PMeBu2t

Article abstract of DOI:10.1021/om700580d

Although Pd[PCy₃]₂, Pd[PMeBu₂ ^t]₂, Pd[PBu₃^t]₂, and a large number of similar bisphosphinepalladium-(0) compounds are excellent cross-coupling catalysts, most are not readily synthesized and stored, and research involving their utilization generally relies on in situ synthetic processes that proceed at unknown rates and in unknown yields. This paper describes, in part, the development of an efficient synthetic route to the representative catalysts Pd[PCy₃]₂, Pd[PMeBu ₂^t]₂, and Pd[PBu₃^t] ₂ via reactions of Pd(η³-C₃H ₅)(η⁵-C₅H₅) with 2 equiv each of PCy₃, PMeBu₂^t, or PBu₃^t. The procedures, which may be generally useful, are optimized such that known concentrations of the bisphosphine catalysts may be generated quickly, reliably, and under mild conditions. That said, the chemistry of the 2:1 compounds can be surprisingly complex; although Pd[PBu₃^t]₂ shows no inclination to increase its coordination number, Pd-[PCy ₃]₂ and Pd[PMeBu₂^t]₂ react with added PCy₃ or PMeBu₂^t to form 3:1 coordination compounds. Equilibrium constants for dissociation of the compounds PdL₃ to PdL₂ + free L (L = PCy₃, PMeBu ₂^t) were measured over a range of temperatures, and from the resulting linear plots of In K_D vs 1/T, ΔH and ΔS values of 21 kJ mol^-1 and 59 J deg^-1 mol^-1, respectively, were determined for the PCy₃ system, and 23 kJ mol ^-1 and 86 J deg^-1 mol^-1, respectively, for PMeBu₂^t. The enthalpy values provide a rough measure of the Pd-P bond dissociation energies, which are, as expected, considerably lower than both the single Pd(II)-phosphine and the single Pt(0)-phosphine bond energies that have been reported (both sim;55 kJ mol^-1). The significant formation of the catalytically less active 3:1 compounds has serious implications for many of the published catalytic cross-coupling processes that involve catalyst formation via the slow reduction of palladium(II) precursors in the presence of excess phosphine; for many such systems, relatively little of the added palladium may actually be present as the catalytically active bisphosphinepalladium(O) compound.

Full text of DOI:10.1021/om700580d

5230
Organometallics 2007, 26, 5230-5238
Optimization of Procedures for the Syntheses of
Bisphosphinepalladium(0) Precursors for Suzuki-Miyaura and
Similar Cross-Coupling Catalysis: Identification of 3:1
Coordination Compounds in Catalyst Mixtures Containing Pd(0),
PCy3, and/or PMeBut2
Emily A. Mitchell and Michael C. Baird*
Department of Chemistry, Queen’s UniVersity, Kingston, Ontario K7L 3N6, Canada
ReceiVed June 12, 2007
Although Pd[PCy3]2, Pd[PMeBut2]2, Pd[PBut3]2, and a large number of similar bisphosphinepalladium-
(0) compounds are excellent cross-coupling catalysts, most are not readily synthesized and stored, and
research involving their utilization generally relies on in situ synthetic processes that proceed at unknown
rates and in unknown yields. This paper describes, in part, the development of an efficient synthetic
route to the representative catalysts Pd[PCy3]2, Pd[PMeBut2]2, and Pd[PBut3]2 via reactions of Pd(η3-
C3H5)(η5-C5H5) with 2 equiv each of PCy3, PMeBut2, or PBut3. The procedures, which may be generally
useful, are optimized such that known concentrations of the bisphosphine catalysts may be generated
quickly, reliably, and under mild conditions. That said, the chemistry of the 2:1 compounds can be
surprisingly complex; although Pd[PBut3]2 shows no inclination to increase its coordination number, Pd-
[PCy3]2 and Pd[PMeBut2]2 react with added PCy3 or PMeBut2 to form 3:1 coordination compounds.
Equilibrium constants for dissociation of the compounds PdL3 to PdL2 + free L (L ) PCy3, PMeBut2)
were measured over a range of temperatures, and from the resulting linear plots of ln KD vs 1/T, ∆H and
∆S values of 21 kJ mol-1 and 59 J deg-1 mol-1, respectively, were determined for the PCy3 system, and
23 kJ mol-1 and 86 J deg-1 mol-1, respectively, for PMeBut2. The enthalpy values provide a rough
measure of the Pd-P bond dissociation energies, which are, as expected, considerably lower than both
the single Pd(II)-phosphine and the single Pt(0)-phosphine bond energies that have been reported (both
∼55 kJ mol-1). The significant formation of the catalytically less active 3:1 compounds has serious
implications for many of the published catalytic cross-coupling processes that involve catalyst formation
via the slow reduction of palladium(II) precursors in the presence of excess phosphine; for many such
systems, relatively little of the added palladium may actually be present as the catalytically active
bisphosphinepalladium(0) compound.
to coupling of alkyl halides are also being explored. The most
widely accepted catalytic cycle (Scheme 1) typically involves
oxidative addition of ArX to what is believed to be a homoleptic
bisphosphine Pd(0) compound PdL2 (L ) tertiary phosphines),
followed by transmetalation and reductive elimination steps.1
High temperatures and long reaction times are often utilized
to obtain good yields of cross-coupling products, and attempts
to improve reaction efficiencies have included varying the nature
of the phosphines L (sterically demanding phosphines often
seem to be best), the L:Pd ratio, the solvent, the temperature,
added salts and bases, and Y.1 Interestingly, a possibly central
variable that has not generally been considered is the extent to
which the putative catalytic species, PdL2, is actually present
in solution. Bisphosphine Pd(0) compounds not only are
extremely air sensitive but also are prone to expand their
coordination spheres to form species of the types PdL2L′ and
PdL2L′2 (L′ ) CO, alkenes, alkynes, phosphines) and to form
clusters containing Pd-Pd bonds.2 Therefore, because of their
delicate and often unpredictable nature, preformed bisphosphine
compounds are rarely utilized.
Introduction
The discovery and development of palladium-catalyzed
Suzuki-Miyaura cross-coupling reactions have revolutionized
methodologies for the formation of carbon-carbon bonds and
have thereby profoundly changed the strategies for the construc-
tion of many types of complex organic molecules.1 In the general
case, an aryl halide ArX (X ) Cl, Br, I) reacts catalytically
with an arylborate species [Ar′BY3]- (Y ) halide, alkoxide,
etc.) to form the coupled product Ar-Ar′, although extensions
(1) For recent reviews, see: (a) Littke, A. F.; Fu, G. C. Angew. Chem.,
Int. Ed. 2002, 41, 4176. (b) Kotha, S.; Lahiri, K.; Kashinath, D. Tetrahedron
2002, 58, 9633. (c) Handbook of Organopalladium Chemistry for Organic
Synthesis; Negishi, E., de Meijere, Eds.; Wiley: New York, 2002. (d)
Miyaura, N. Top. Curr. Chem. 2002, 219, 11. (e) Beletskaya, I. P.;
Cheprakov, A. V. In ComprehensiVe Coordination Chemistry II; McClev-
erty, J. A., Meyer, T. J., Eds.; Elsevier: Oxford, U.K., 2004; Vol. 9, pp
305-368. (f) Hassan, J.; Se´vignon, M.; Gozzi, C.; Schulz, E.; Lemaire, M.
Chem. ReV. 2002, 102, 1359. (g) Farina, V. AdV. Synth. Catal. 2004, 346,
1553. (h) Bellina, F.; Carpita, A.; Rossi, R. Synthesis 2004, 2419. (i) Tsuji,
J. Palladium Reagents and Catalysts, 2nd ed.; Wiley: New York, 2004.
(j) Cepanec, I. Synthesis of Biaryls; Elsevier: Amsterdam, 2004. (k) Zapf,
A.; Beller, M. Chem. Commun. 2005, 431. (l) Miyaura, N. Metal Catalyzed
Cross-Coupling Reactions, 2nd ed.; de Mejiere, A., Diederich, F., Eds.;
John Wiley & Sons: New York, 2004; pp 41-123. (m) Christmann, U.;
Vilar, R. Angew. Chem., Int. Ed. 2005, 44, 366, and references therein. (n)
Phan, N. T. S.; Van Der Sluys, M.; Jones, C. W. AdV. Synth. Catal. 2006,
348, 609. (o) Hartwig, J. F. Synlett 2006, 1283.
Instead more readily accessible compounds, such as the
relatively poor catalyst Pd(PPh3)4,2d,g,k,l are utilized or the
phosphine to be used is added to solutions of precursor Pd(0)
compounds such as Pd(dba)2 and Pd2(dba)3 (dba ) trans,trans-
10.1021/om700580d CCC: $37.00 © 2007 American Chemical Society
Publication on Web 09/05/2007

Bisphosphinepalladium(0) Precursors for Cross-Coupling Catalysis
Organometallics, Vol. 26, No. 21, 2007 5231
PMePh2 give similar species in DMF at 60 °C.5d Thus reduction
to Pd(0) does seem to occur in some instances although the
species produced have not in fact been identified. Interestingly,
under the particular conditions of the latter study, reduction is
inhibited by bulky substituents on the phosphorus5d such that
reduction of even Pd(OAc)2 by monodentate phosphines be-
comes disfavored when the phosphine cone angle is ∼130°. This
corresponds approximately to the size of PEt3,6 and thus
reduction of Pd(OAc)2 by the bulkier phosphines normally
utilized in Suzuki cross-coupling reactions1 is problematic. For
instance, the commonly used PCy3, PMeBut2, and PBut3 have
cone angles of 170°,6 161°,7 and 182°,6 respectively.
Scheme 1
The question arises then, if one particular palladium(II)-
phosphine cross-coupling catalyst system is found to be more
effective than another, does the difference arise because one
phosphine is intrinsically superior, as is generally claimed or
implied? Or does the difference arise because more of the Pd-
(II) precursor is reduced in the one case than in the other?
Furthermore, in situations where, in spite of the addition of
excess phosphine, reduction of the Pd(II) is far from complete
such that the ratio L:Pd(0) . 1, say 100:1 or 1000:1, might the
major palladium(0) species in solution for some phosphines be
predominantly a less active species PdL3 or PdL4, depending
on the steric requirements of L? The importance of and
difficulties in identifying the true catalyst(s) in the plethora of
processes investigated has been explicitly recognized,1n,5g
although it may be an exaggeration to suggest that the “literature
has grown into a morass, with an endless list of hypothesized
or claimed true catalytic species”.1n
dibenzylideneacetone).3 In the case of Pd(dba)2 and Pd2(dba)3,
it was at one time generally assumed that the dba is completely
displaced or, if not, that it does not impede oxidative addition.
However there has now accrued ample evidence that alkenes
containing electron-withdrawing groups, including dba, bind
very strongly to Pd(0) and are not always fully displaced by
tertiary phosphines.4 Furthermore, for a combination of elec-
tronic and steric reasons, dba is found to actually inhibit many
oxidative addition reactions.4
Catalyst solutions are alternatively generated by adding a
phosphine to a suspension or solution of a Pd(II) compound
such as Pd(OAc)2 or PdCl2, sometimes in the presence of a base,
the assumption made being that the Pd(II) salts are somehow
reduced to Pd(0) compounds.1 Although suggestions of rel-
evance for a number of reducing processes have appeared for a
variety of palladium-based catalytic systems, there seems in fact
to be a paucity of careful studies establishing the usefulness,
general or specific, of any class of reducing agents. Indeed, for
most of the phosphine/palladium catalyst systems used, there
is little or no evidence that substantial reduction is effected either
rapidly or completely.
Indeed, although tertiary phosphines themselves may serve
as reducing agents5 and it has long been known that triarylphos-
phines in DMF reduce Pd(OAc)2 and Pd(OCOCF3)2 via the
corresponding bisphosphine Pd(II) compounds,4a,5b,d,e,f the analo-
gous halo compounds PdX2(PPh3)2 (X ) Cl, Br, I) are stable
with respect to reduction in this way.5a On the other hand,
unidentified but presumed Pd(0) compounds are formed when
Pd(OAc)2 and PdCl2 are treated with PBu3 in benzene and
THF,5c and reactions of Pd(OAc)2 with PBu3, PMe2Ph, and
However, as mentioned above, high temperatures and long
reaction times are often required to obtain good yields in
Suzuki-Miyaura aryl-aryl cross-coupling reactions. If one
hypothesizes that somewhat extreme conditions may be neces-
sary in order that a Pd(0) catalyst species be formed in
catalytically effective concentrations, then a mild and efficient
method to produce bisphosphine Pd(0) compounds unambigu-
ously would permit an accurate assessment of the relative merits
of various phosphines in the catalytic cycle. It might also result
in the realization of much milder reaction conditions for cross-
coupling reactions involving otherwise poorly reducible Pd(II)
precatalysts, a potentially useful development where thermally
sensitive functionality in cross-coupling substrates would pre-
clude the use of high temperatures for prolonged periods of time.
(2) For background references information on Pd(0) chemistry in general,
see: (a) Kudo, K.; Hidai, M.; Uchida, Y. J. Organomet. Chem. 1971, 33,
393. (b) Kudo, K.; Sato, M.; Hadai, M.; Uchida, Y. Bull. Chem. Soc. Jpn.
1973, 46, 2820. (c) Minematsu, H.; Nonaka, Y.; Takahashi, S.; Hagihara,
N. J. Organomet. Chem. 1973, 59, 395. (d) Mednikov, E. G.; Eremenko,
N. K. IzV. Akad. Nauk SSSR 1984, 2781. (e) Immirzi, A.; Musco, A. J.
Chem. Soc., Chem. Commun. 1974, 400. (f) Matsumoto, M.; Yoshioka,
H.; Nakatsu, K.; Yoshida, T.; Otsuka, S. J. Am. Chem. Soc. 1974, 96, 3322.
(g) Mann, B. E.; Musco, A. J. Chem. Soc., Dalton Trans. 1975, 1673. (h)
Otsuka, S.; Yoshida, T.; Matsumoto, M.; Nakatsu, K. J. Am. Chem. Soc.
1976, 98, 5850. (i) Yoshida, T.; Otsuka, S. J. Am. Chem. Soc. 1977, 99,
2134. (j) Ozawa, F.; Ito, T.; Nakamura, Y.; Yamamoto, A. J. Organomet.
Chem. 1979, 168, 375. (k) Negishi, E-I.; Takahashi, T.; Akiyoshi, K. J.
Chem. Soc., Chem. Commun. 1986, 1338. (l) Urata, H.; Suzuki, H.; Moro-
oka, Y.; Ikawa, T. J. Organomet. Chem. 1989, 364, 235. (m) Yoshida, T.;
Otsuka, S. Inorg. Synth. 1990, 28, 113. (n) Krause, J.; Bonrath, W.;
Porschke, K. R. Organometallics 1992, 11, 1158. (o) Tanaka, M. Acta
Crystallogr. C 1992, 48, 739. (p) Kirchoff, J. H.; Netherton, M. R.; Hills,
I. D.; Fu, G. C. J. Am. Chem. Soc. 2002, 124, 13662. (q) Amatore, C.;
Jutand, A.; Meyer, G. Inorg. Chim. Acta 1998, 273, 76. (r) Kuran, W.;
Musco, A. Inorg. Chim. Acta 1975, 12, 187.
(4) (a) Amatore, C.; Jutand, A.; Fouad, K.; M’Barki, M. A.; Mottier, L.
Organometallics 1993, 12, 3168. (b) Jutand, A.; Hii, K. K.; Thornton-Pett,
M.; Brown, J. M. Organometallics 1999, 18, 5367. (c) Amatore, C.; Carre´,
E.; Jutand, A.; Medjour, Y. Organometallics 2002, 21, 4540. (d) Amatore,
C.; Bensalem, S.; Ghalem, S.; Jutand, A.; Medjour, Y. Eur. J. Org. Chem.
2004, 366. (e) Fairlamb, I. J. S.; Kapdi, A. R.; Lee, A. F. Org. Lett. 2004,
6, 4435. (f) Mace´, Y.; Kapdi, A. R.; Fairlamb, I. J. S.; Jutand, A.
Organometallics 2006, 25, 1795. (g) Amatore, C.; Jutand, A. Coord. Chem.
ReV. 1998, 178-180, 511. (h) Shekhar, S.; Ryberg, P.; Hartwig, J. F.;
Mathew, J. S.; Blackmond, D. G.; Strieter, E. R.; Buchwald, S. L. J. Am.
Chem. Soc. 2006, 128, 3584. (i) Paul, F.; Patt, J.; Hartwig, J. F.
Organometallics 1995, 14, 3030.
(5) (a) Amatore, C.; Azzabi, M.; Jutand, A. J. Am. Chem. Soc. 1991,
113, 8375. (b) Amatore, C.; Jutand, A.; M’Barki, M. A. Organometallics
1992, 11, 3009. (c) Mandai, T.; Matsumoto, T.; Tsuji, J.; Saito, S.
Tetraherdron Lett. 1993, 34, 2513. (d) Amatore, C.; Carre´, E.; Jutand, A.;
M’Barki, M. A. Organometallics 1995, 14, 1818. (e) Amatore, C.; Carre´,
E.; Jutand, A.; M’Barki, M. A.; Meyer, G. Organometallics 1995, 14, 5605.
(f) Amatore, C.; Jutand, A.; Lemaˆıtre, F.; Ricard, J. L.; Kozuch, S.; Shaik,
S. J. Organomet. Chem. 2004, 689, 3728. (g) See also: Knowles, J. P.;
Whiting, A. Org. Biol. Chem. 2007, 5, 31, who, in a very recent review of
the Heck-Mizoroki cross-coupling reaction, outline the problems inherent
in deducing the origin(s) of many palladium(0) catalyst systems.
(6) Tolman, C. A. Chem. ReV. 1977, 77, 313.
(3) (a) Takahashi, Y.; Ito, T.; Sakai, S.; Ishii, Y. Chem. Commun. 1970,
1065. (b) Ukai. T.; Kawazura, H.; Ishii, Y.; Bonnet, J. J.; Ibers, J. A. J.
Organomet. Chem. 1974, 65, 253. (c) Ito, T.; Hasegawa, S.; Takahashi,
Y.; Ishii, Y. J. Organomet. Chem. 1974, 73, 401. (d) Ishii, Y.; Hasegawa,
S.; Kimura, S.; Itoh, K. J. Organomet. Chem. 1974, 73, 411. (e) Rettig, M.
F.; Maitlis, P. N. M. Inorg. Synth. 1990, 28, 110.
(7) Hills, I. D.; Netherton, M. R.; Fu, G. C. Angew. Chem., Int. Ed. 2003,
42, 5749.

5232 Organometallics, Vol. 26, No. 21, 2007
Mitchell and Baird
Table 1. 31P NMR Data (at 25 °C in toluene-d8 unless
In order to explore these possibilities, we have begun a study
of cross-coupling reaction chemistry in which we have deter-
mined initially the mildest conditions under which useful
compounds of the specific stoichiometry PdL2 may be generated
unambiguously, quantitatively, quickly, and under mild condi-
tions. The method chosen utilizes Pd(η3-C3H5)(η5-C5H5),8 which,
on heating in the absence of potential ligands, reductively
eliminates C5H5-C3H5 (as a mixture of isomers) and deposits
palladium metal.9 It also reacts with phosphines L to give,
following reductive elimination of C5H5-C3H5, Pd(0) com-
pounds of the types PdLn (n ) 2-4);2f,h,m,p,10 σ-allyl compounds
of the type η5-C5H5Pd(η1-C3H5)L and dinuclear species of the
type Pd2L2(µ-C5H5)(µ-C3H5) are often intermediates.11
otherwise indicated)
2J(P-P)
compound
Pd[PCy3]2
δ (free L) δ (PCy3) δ (PMeBut2) δ (PBut3)
(Hz)
9.89
11.4
63.3
39.2
Pd[PCy3]3
25.5a
Pd[PMeBut2]2
Pd[PMeBut2]3
Pd[PBut3]2
41.9
30.3b
84.7
Pd[PCy3][PMeBut2]
Pd[PMeBut2][PBut3]
Pd[PCy3][PBut3]
40.3c
41.4
40.8c
41.8
300
246
241
85.4
85.2
a At -68 °C. b At -88 °C. c In benzene-d6 at 25 °C; calc as an AB spin
system.
Although Pd(η3-C3H5)(η5-C5H5) has been used in this way
previously to generate Pd(0) cross-coupling catalysts,2f,h,m,p,10
a systematic study to determine the optimal conditions for
catalyst generation has not previously been carried out. We have
now determined and herein report the optimum (i.e., minimum
temperature, maximum time required) conditions under which
the representative, catalytically very important1a,b,g,h compounds
PdL2 (L ) PCy3,2e,g PMeBut2,2p PBut32f,m,o) are generated cleanly
and essentially quantitatively.
Complementing these findings, and relevant to understanding
cross-coupling processes under the commonly occurring cir-
cumstances where L:Pd(0) . 1 (see above), we have also
determined the conditions in which significant proportions of
palladium(0) species produced are present as the homoleptic
3:1 compounds PdL3 (L ) PCy3, PMeBut2), in equilibrium with
the corresponding 2:1 species PdL2 (KD of Scheme 2), and the
thermodynamic parameters for dissociation of PdL3 to PdL2 +
L (L ) PCy3, PMeBut2). The heteroleptic species Pd[PCy3]-
[PMeBut2], Pd[PMeBut2][PBut3], Pd[PCy3][PBut3], Pd[PCy3]-
[PMeBut2]2, and Pd[PCy3]2[PMeBut2] are also formed in mixed
phosphine systems. In a following paper,12 we shall describe a
related kinetics study of the oxidative addition of PhBr to Pd-
(PCy3)n (n ) 1, 2, 3), of importance as the first step of the
Suzuki-Miyaura cross-coupling catalytic cycles involving this
very important, representative catalyst system.
deoxygenated by passage through a heated column of BASF copper
catalyst and then dried by passing through a column of 4 Å
molecular sieves. Handling and storage of air-sensitive organome-
tallic compounds were done in an MBraun Labmaster glovebox,
and NMR spectra were recorded using a Bruker AV 600 spec-
trometer. 31P NMR spectra were run at 242.95 MHz and are
referenced with respect to external 85% H3PO4; data are given in
Table 1, although it should be noted that the chemical shifts vary
somewhat with temperature. 1H and 13C NMR data are referenced
to TMS via the residual proton signals of the deuterated solvents.
Toluene-d8 and benzene-d6 (Cambridge Isotope Laboratories, Inc.
or CDN Isotopes) were degassed under vacuum and dried by
passage through a small column of activated alumina before being
stored over 4 Å molecular sieves. The phosphines PCy3, PMeBut2,
and PBut3 were obtained from Strem and used without further
purification. Tetraoctylphosphonium bromide (TOPB) was purchasd
from Aldrich and dried under vacuum at 45 °C prior to use. The
compounds Pd(η3-C3H5)(η5-C5H5),8 Pd[PCy3]2,2m Pd[PMeBut2]2,2m
and Pd[PBut3]22m were prepared according to the literature. While
the 31P resonances of the three bisphosphine compounds were all
sharp singlets at 25 °C and below, the 31P resonances of Pd[PCy3]2
and Pd[PMeBut2]2 broadened significantly in the presence of free
PCy3 or PMeBut2 at 25 °C. The resonances of coordinated PBut3
remained sharp under all circumstances.
In Situ Generation of PdL2 (L ) PCy3, PMeBut2, PBut3). In
a typical reaction, 4.5 mg of Pd(η3-C3H5)(η5-C5H5) (0.021 mmol)
in 0.18 mL of C6D6 was combined with 12 mg of PCy3 (0.043
mmol) in 0.3 mL of C6D6 in an NMR tube, and a 31P NMR
spectrum of the solution was run every 10 min for 2 h at 25 °C.
Similar experiments were carried out for various periods of time
in the temperature range 65-80 °C and using PMeBut2 and PBut3
at various temperatures.
In Situ Generation of Pd[PCy3][PMeBut2]. A solution of 4.4
mg of Pd[PMeBut2]2 (0.010 mmol) in 0.5 mL of benzene-d6 was
added to a solution of 7.2 mg of Pd[PCy3]2 (0.011 mmol) in 0.5
mL of benzene-d6 at 21 °C, and a 31P NMR spectrum was obtained
at 25 °C. In a complementary experiment, 10 mg of Pd[PCy3]2
(0.015 mmol) in 0.7 mL of toluene-d8 was treated with 2.5 mg of
PMeBut2 (0.015 mmol) in 0.3 mL of toluene-d8 at 21 °C, and
31P NMR spectra were obtained in the temperature range 25 to
-93 °C. Similarly 27 mg of Pd[PCy3]2 (0.04 mmol) in 0.7 mL of
toluene-d8 was treated with 13 mg of PMeBut2 (0.082 mmol) in
0.016 mL of toluene-d8 at 21 °C, and a 31P-31P NMR correlation
spectrum was obtained at -88 °C.
Scheme 2
Experimental Section
All syntheses were carried out under a dry, deoxygenated argon
atmosphere using standard Schlenk line techniques. Argon was
(8) Tatsuno, Y.; Yoshida, T.; Otsuka, S. Inorg. Synth. 1990, 28, 342.
(9) (a) Liang, C.; Xia, W.; Soltani-Ahmadi, H.; Schlu¨ter, O.; Fischer, R.
A.; Muhler, M. Chem. Comm. 2005, 282. (b) Niklewski, A.; Strunskus, T.;
Witte, G.; Wo¨ll, C. Chem. Mater. 2005, 17, 861. (c) Xia, W.; Schlu¨ter, O.
F.-K.; Liang, C.; van den Berg, M. W. E.; Guraya, M.; Muhler, M. Catal.
Today 2005, 102-103, 34.
(10) (a) Galardon, E.; Ramdeehul, S.; Brown, J. M.; Cowley, A.; Hii,
K. K.; Jutand, A. Angew. Chem. Int. Ed. 2002, 41, 1760. (b) Leoni, P.
Organometallics 1993, 12, 2432. (c) Stauffer, S. R.; Beare, N. A.; Stambuli,
J. P.; Hartwig, J. F. J. Am. Chem. Soc. 2001, 123, 4641. (d) Matsumoto,
T.; Kasai, T.; Tatsumi, K. Chem. Lett. 2002, 346. (e) Barrios-Landeros, F.;
Hartwig, J. H. J. Am. Chem. Soc. 2005, 127, 6944. (f) Stambuli, J. P.; Buhl,
M.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 9346. (g) Grotjahn, D. B.;
Gong, Y.; Zakharov, L.; Golen, J. A.; Rheingold, A. L. J. Am. Chem. Soc.
2006, 128, 438. (h) Mann, G.; Shelby, Q.; Roy, A. H.; Hartwig, J. F.
Organometallics 2003, 22, 2775. (i) Netherton, M. R.; Fu, G. C. Angew.
Chem., Int. Ed. 2002, 41, 3910.
(11) (a) Werner, H.; Ku¨hn, A.; Tune, D. J. Chem. Ber. 1977, 110, 1763.
(b) Ku¨hn, A.; Werner, H. J. Organomet. Chem. 1979, 179, 421. (c) Werner,
H.; Ku¨hn, A.; Burschka, C. Chem. Ber. 1980, 113, 2291. (d) Werner, H.
Angew. Chem., Int. Ed. Engl. 1977, 16, 1. (e) Werner, H. AdV. Organomet.
Chem. 1981, 19, 155.
(12) Mitchell, E. A.; Jessop, P. G.; Baird, M. C. Manuscript in
preparation.
In Situ Generation of Pd[PCy3][PBut3]. A solution of 7.3 mg
of PBut3 (0.036 mmol) in 0.18 mL of toluene-d8 was added to 10
mg of PCy3 (0.036 mmol) in 0.25 mL of toluene-d8 in an NMR
tube. The sample was preheated to 77 °C, and 7.6 mg of Pd(η3-
C3H5)(η5-C5H5) (0.036 mmol) in 0.25 mL of toluene-d8 was added.
The sample was heated for 1 h at 77 °C, then cooled to 21 °C, and
a 31P NMR spectrum was then obtained at 21 °C. Alternatively 7.0
mg of Pd[PCy3]2 (0.011 mmol) in 0.5 mL of toluene-d8 was added
to 7.3 mg of Pd[PBut3]2 (0.014 mmol) in 0.2 mL of toluene-d8 at
21 °C, and a 31P NMR spectrum was then obtained.

Bisphosphinepalladium(0) Precursors for Cross-Coupling Catalysis
Organometallics, Vol. 26, No. 21, 2007 5233
Figure 1. Reaction sequence in the formation of PdL2.
In Situ Generation of Pd[PMeBut2][PBut3]. A solution of 71
mg of Pd[PMeBut2]2 (0.17 mmol) and 89 mg of Pd[PBut3]2 (0.17
mmol) was made up in 0.65 mL of toluene-d8 at 21 °C, and a 31P
NMR spectrum was obtained at 25 °C.
Determination of KD for PdL3 (L ) PCy3, PMeBut2). In a
typical experiment, 6.1 mg of Pd[PCy3]2 (9.8 × 10-3 mmol) in 0.5
mL of toluene-d8 was added to 2.5 mg of PCy3 (9.0 × 10-3 mmol)
in 0.1 mL of toluene-d8. The resulting solution was diluted with
an additional 0.2 mL of toluene-d8, a glass capillary containing a
toluene-d8 solution of tetraoctylphosphonium bromide (TOPB)
(0.046 M, 3.0 × 10-3 mmol) was added ([Pd(PCy3)2]o ) [PCy3]
) 0.011 M), and 31P inverse-gated NMR spectra were obtained
between -85 and -68 °C with d1 ) 45 s. Similar experiments
were carried out using PMeBut2.
Results and Discussion
Syntheses of the Homoleptic Compounds PdL2 (L ) PCy3,
PMeBut2, PBut3). A small number of Pd(0) compounds PdL2
have been prepared previously via reactions of Pd(η3-C3H5)-
(η5-C5H5) with tertiary phosphines,2f,h,m,p,10 but there have
apparently been no attempts to ascertain the conditions under
which 2:1 compounds are best prepared. We have therefore
monitored by 31P NMR spectroscopy the reactions of Pd(η3-
C3H5)(η5-C5H5) with 2 equiv each of the representative phos-
phines PCy3, PMeBut2, and PBut3, all of great relevance in
the context of catalytic Suzuki-Miyaura cross-coupling
reactions.2p,10h,i,13,14 In this way we have determined the optimal
conditions (lowest temperature for quantitative conversion
in minimal time) under which specifically the compounds
Pd[PCy3]2, Pd[PMeBut2]2, and Pd[PBut3]2 are obtained in
toluene-d8.
Figure 2. Stacked plots of 31P NMR spectra for reaction of Pd-
(η3-C3H5)(η5-C5H5) with 2 equiv of PCy3 (a) at 25 °C and (b) at
65 °C.
ished. Concurrently a third broad resonance, attributable to the
η5-C5H5 ligand of an intermediate η1-allyl species, appeared at
δ 6.28. As reported previously for other phosphines,11c it seems
that Pd(η3-C3H5)(η5-C5H5) reacts with PCy3 reversibly to
produce the intermediate η1-allyl species, which in turn forms
irreversibly the dinuclear species Pd2(PCy3)2(µ-C3H5)(µ-C5H5).
1
While H resonances anticipated for an η1-allyl intermediate
Conversion of Pd(η3-C3H5)(η5-C5H5) to the 2:1 compounds
is expected to proceed as in Figure 1, via the types of η1-allyl
and dinuclear intermediates shown.11 Reactions of Pd(η3-C3H5)-
(η5-C5H5) with 2 equiv each of PCy3 or PMeBut2 at 25 °C
(Figure 2a for L ) PCy3) were found to proceed slowly to
produce intermediate species exhibiting 31P resonances at δ 27.8
and 35.4, respectively, attributable to the corresponding di-
nuclear species,11 in addition to extremely broad resonances at
δ 54.6 and 61.0, respectively, attributable to the η1-allyl inter-
mediates.11c After a day at this temperature, there were observed
no 31P resonances at δ 39.4 or 41.0, the chemical shifts of the
2:1 compounds Pd[PCy3]2 and Pd[PMeBut2]2, respectively.2p,15
were not observed at 25 °C, such species are fluxional and rapid
exchange processes would undoubtedly result in extensive
1
broadening. In addition to the above, new H resonances also
emerged over the course of the reaction in the regions δ ∼6-
6.5, ∼5.7, ∼4.9, and ∼3; these are attributed to the products of
reductive elimination, C5H5-C3H5, formed as a mixture of
isomers.11
Similarly, the Pd(η3-C3H5)(η5-C5H5)/PMeBut2 system reacted
in toluene-d8 at 25 °C to produce a new species with a resonance
at δ 6.02, consistent with the coordinated η5-C5H5 ring of a
dinuclear species. Two virtual triplets appeared at δ 1.08 (6.3
Hz splitting) and 1.00 (6 Hz splitting) and are attributed
respectively to the But and PMe groups on the mutually trans
phosphines of Pd2(PMeBut2)2(µ-C3H5)(µ-C5H5). In addition,
resonances of the anticipated reductive elimination products
C5H5-C3H5 appeared as above.
1
Supporting these conclusions, the H NMR spectrum of the
Pd(η3-C3H5)(η5-C5H5)/PCy3 reaction mixture in toluene-d8 at
25 °C exhibited a broadened η5-C5H5 singlet of Pd(η3-C3H5)-
(η5-C5H5) at δ 5.84 and a new sharp singlet at δ 6.03, attributed
to the dinuclear species Pd2(PCy3)2(µ-C3H5)(µ-C5H5). The
resonance at δ 6.03 gained in intensity over the course of the
reaction and exhibited an HSQC correlation to a 13C resonance
at δ 89.9, consistent with a coordinated η5-C5H5 ligand in a
At 65 °C (Figure 2b for L ) PCy3), the dinuclear species
Pd2L2(µ-C3H5)(µ-C5H5) (L ) PCy3, PMeBut2) were slowly
consumed to produce the anticipated 2:1 products Pd[PCy3]2
(δ 39.415) and Pd[PMeBut2]2 (δ 41.02p). In contrast, and of note,
at 77 °C the 2:1 compounds Pd[PCy3]2 (δ 39.5) and Pd-
[PMeBut2]2 (δ 43.9) were both formed quantitatively within 1
h. In both cases, the anticipated resonances for the reductive
elimination products C5H5-C3H5 were also observed in the 1H
NMR spectra, while for the PMeBut2 system broad singlets at
δ 1.26 and 1.03 appeared and are attributed to the But and PMe
groups, respectively, of Pd[PMeBut2]2.
1
dinuclear species,11b while the H resonance at δ 5.84 dimin-
(13) Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 1998, 37, 3387.
(14) (a) Netherton, M. R.; Dai, C.; Neuschutz, K.; Fu, G. C. J. Am. Chem.
Soc. 2001, 123, 10099. (b) Kirchhoff, J. H.; Dai, C.; Fu, G. C. Angew.
Chem., Int. Ed. 2002, 41, 1945.
(15) Grushin, V. V.; Benismon, C.; Alper, H. Inorg. Chem. 1994, 33,
4804.

5234 Organometallics, Vol. 26, No. 21, 2007
Mitchell and Baird
In contrast, we find that PBut3 generates neither η1-allyl nor
dinuclear intermediate species at 25 °C in toluene-d8, as has
been noted elsewhere;11e only Pd[PBut3]2 is formed in a reaction
1
that is complete within 1 h at 25 °C. The 31P and H NMR
spectra (25 °C) exhibit resonances respectively at δ 86.0 (δ 85.2
in C6D616) and at δ 1.49 (virtual triplet, 5.6 Hz splitting) (δ
1.53, 5.7 Hz splitting in C6D616). The compound Pd[PBut3]2 is
also formed quantitatively in about 30 min at 77 °C (31P: δ
88.4; 1H: δ 1.49, 5.6 Hz splitting). It seems likely in this system
that an η1-allyl species is formed initially but that the steric
bulk of the PBut3 ligand induces rapid reductive elimination of
C5H5-C3H5. The anticipated resonances of the latter are
Figure 3. Stacked plots of low-temperature 31P inverse gated NMR
spectra for reaction of Pd[PCy3]2 with 1 equiv of PCy3 (242.95
MHz).
1
observed in the H NMR spectra of the reaction mixtures at
both 25 and 77 °C.
Homoleptic Compounds PdL3; Equilibria between PdL2
and PdL3. The analysis described above inevitably involved
on occasion the addition of slight excesses or deficiencies of
the phosphines because of the small amounts of materials being
reacted, and we became aware of other apparent Pd(0) species
that were also being formed. In some cases there was observed
exchange broadening that implicated species other than the 2:1
compounds being investigated, while low-temperature spectra
sometimes exhibited weak resonances that could not be ac-
counted for.
There is certainly precedent for 3:1 compounds PdL3 exist-
ing in solution in equilibrium with the corresponding 2:1 com-
pounds,2g,17 as in Scheme 2, but these usually involve sterically
undemanding ligands and perhaps would not be expected to be
very important in the chemistry of the three relatively large
phosphine ligands of interest here. On the other hand, there is
considerable speculation in the literature concerning possible
equilibria between 2:1 and 1:1 compounds and the possible role-
(s) of monophosphine compounds PdL in cross-coupling
reactions,1m,18 and it was decided to investigate further the
palladium(0) systems being formed. Therefore, for all three
phosphines, experiments were carried out in which 1 equiv of
the phosphine was added to a solution of the PdL2 compound
in toluene-d8 and the 31P NMR spectra were obtained at various
temperatures.
Figure 4. Plot of ln KD as a function of 1/T for the dissociation of
Pd[PCy3]3 (R2 ) 0.988) (* point from ref 2g).
hence to obtain thermodynamic data for this system in toluene-
d8. To this end, the concentrations of all three species in a 1:1
mixture of PCy3 and Pd[PCy3]2 were determined via integrations
in the temperature range -68 to -85 °C, mass balances being
ensured by relating all intensity measurements to the constant
intensity of a very narrow 31P resonance of a solution of tetra-
octylphosphonium bromide (TOPB) standard in toluene-d8 and
contained in a glass capillary inserted into the NMR tube.
A linear plot of ln KD vs 1/T was obtained (Figure 4), from
which values of ∆H and ∆S were found to be 21 kJ mol-1 and
59 J deg-1 mol-1, respectively. Extrapolation of this plot
indicates that KD is 0.3 at 25 °C and 1.5 at 100 °C. Thus, for
situations posited in the Introduction in which excess PCy3 has
been added to a Pd(II) precursor and the ratio of PCy3:Pd(0) in
an attempted cross-coupling reaction might be 100:1, the
proportion of the total palladium present as Pd[PCy3]3 could
be 84% at 25 °C and 48% at 100 °C. On the other hand, if the
ratio of PCy3:Pd(0) was, say, 1000:1, as might be the case in
the very early stages of any cross-coupling reaction involving
Pd(II) precatalysts and excess PCy3, the proportion of the total
palladium present as Pd[PCy3]3 at 25 and 100 °C could be 98%
and 90%, respectively.
Although the 31P NMR spectrum of pure Pd[PCy3]2 at 25 °C
exhibits a sharp resonance at δ 39.2 (Table 1), the 31P NMR
spectrum of a solution containing equimolar amounts of Pd-
[PCy3]2 and PCy3 at this temperature exhibits only broadened
resonances at δ 39.7 and 9.44, respectively, very close to the
chemical shifts of Pd[PCy3]2 and PCy3 alone and hence
attributed to these compounds undergoing exchange. On cooling
the solution (Figure 3), the two resonances sharpen somewhat
and shift slightly, and at -56 °C a new broad resonance appears
at δ 26.3. At -68 °C the latter is clearly apparent and of
intensity comparable with that of the 2:1 compound (Figure 3).
The resonance is not observed in low-temperature 31P NMR
spectra containing just Pd[PCy3]2, and therefore, consistent with
a previous report based on relatively low field spectra (36.43
MHz),2g it is attributed to the 3:1 compound Pd[PCy3]3 rather
than the 1:1 analogue Pd[PCy3].
Similar 31P ΝΜR experiments involving 1:1 mixtures of Pd-
[PMeBut2]2 and PMeBut2 in toluene-d8 resulted in similar
observations, with a resonance attributable to Pd[PMeBut2]3
being observed at δ ∼30.3 over a range of temperatures. A linear
plot of ln KD vs 1/T was again obtained, from where ∆H and
∆S values 23 kJ mol-1 and ∆S ) 86 J deg-1 mol-1, respectively,
were obtained. Extrapolation to 25 and 100 °C yielded KD values
at these temperatures of 3 and 19, respectively. Therefore a
100:1 mixture of PMeBut2 and Pd[MeBut2]2 under the conditions
decribed above could contain 32% of the total palladium as Pd-
(MeBut2)3 at 25 °C and 7% at the higher temperature. Similarly,
if the PMeBut2:Pd(0) ratio was 1000:1, then the total palladium
present as Pd(MeBut2)3 at 25 and 100 °C could be 83% and
42% respectively.
Since the 31P resonances of all three compounds PCy3, Pd-
[PCy3]2, and Pd[PCy3]3 are observable over a range of temper-
atures, the opportunity was presented to measure the equilibrium
constant KD (see Scheme 2) as a function of temperature and
(16) Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2001, 123, 2719.
(17) Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M.
AdVanced Inorganic Chemistry, 6th ed.; John Wiley and Sons: New York,
1999; p 1065.
(18) (a) Ahlquist, M.; Norrby, P. Organometallics 2007, 26, 550. (b)
Ahlquist, M.; Fristrup, P. Tanner, D.; Norrby, P. Organometallics 2006,
25, 2066.

Bisphosphinepalladium(0) Precursors for Cross-Coupling Catalysis
Organometallics, Vol. 26, No. 21, 2007 5235
Figure 5. 31P NMR spectrum (25 °C) of the reaction of Pd(η3-C3H5)(η5-C5H5) with 1 equiv of PCy3 and 1 equiv of PBut3.
In contrast, 31P ΝΜR spectra of a 1:1 solution of PBut3 and
Pd[PBut3]2 in toluene-d8 exhibit two unbroadened resonances
for the two compounds over the entire temperature range 0 to
-73 °C, implying that any exchange must be slow on the NMR
time scale. There is thus no evidence for the formation of Pd-
[PBut3]3, reasonable in view of the much larger cone angle of
this ligand.6 The great steric requirements of this ligand have
been noted elsewhere10f,13 as being of significance in their effect
on cross-coupling reactions using Pd[PBut3]2.
compound of the type Pd[PR3][PR′3]2 is expected to exhibit the
spectrum of an AX2 or AB2 spin system, and compounds of
the type Pd[PR3]2[PR′3]2 or Pd[PR3][PR′3]3 are expected to
exhibit the spectra of A2X2 (A2B2) or AX3 (AB3) spin systems,
respectively.20 As we had wondered in any case if heteroleptic
2:1 compounds would form as readily and in the same way as
the homoleptic compounds described abovessuch species have
not been studied as cross-coupling catalystsswe carried out
experiments designed to generate heteroleptic compounds in
which different phosphines were bound to palladium.
The ∆H values presented above, 21 and 23 kJ mol-1 for
respectively the PCy3 and PMeBut2 systems, presumably provide
reasonable approximations of the Pd-P bond dissociation
energies. If so, then the Pd(0)-phosphine bond energies of these
two 3:1 coordination compounds are considerably lower than
are those determined for many other types of phosphine-
transition metal complexes.19a They are also, as would be
anticipated, significantly lower than the only available bond
dissociation energy of a Pd(II)-phosphine compound (54 kJ
To this end, Pd(η3-C3H5)(η5-C5H5) was reacted with 1 equiv
each of PCy3 and PBut3 for 1 h at 77 °C in toluene-d8, and the
resulting 31P NMR spectrum, run at 25 °C after brief cooling,
is shown in Figure 5. As can be seen, in addition to the
resonances of Pd[PCy3]2 (δ 40.1, exchange broadened), Pd-
[PBut3]2 (δ 86.0, sharp), unreacted PBut3 (δ 63.3), and a minor
amount of impurity (δ 46.3), the spectrum also exhibits an AX
pattern with chemical shifts δ 41.4 and 85.2 and coupling
mol-1 19b
and the Pt-PCy3 bond dissociation energy of the Pt-
)
2
constant J(P-P) of 241 Hz (∆ν/J ) 0.182). An essentially
(0) compound Pt[PCy3]3 (55 kJ mol-1).19c
identical 31P NMR spectrum is obtained on reaction of equimolar
amounts of Pd[PBut3]2 with Pd[PCy3]2 in toluene-d8, and a 31P-
31P COSY correlation experiment at 25 °C confirms the mutual
coupling of the two doublets and hence supports their assign-
ment to the new heteroleptic compound Pd[PCy3][PBut3].
Syntheses of the 2:1 Heteroleptic Compounds PdLL′ (L,
L′ ) PCy3, PMeBut2, PBut3); Identification of the Corre-
sponding 3:1 Compounds. As indicated above, in situ iden-
tification of the 2:1 compounds Pd[PCy3]2, Pd[PMeBut2]2, and
Pd[PBut3]2 was readily made on the basis of comparisons with
spectra of authentic samples in addition to previously published
spectroscopic data for these compounds.2m,p,15,16 Our identifica-
tion of the 3:1 homoleptic species formed in situ on addition of
excess PCy3 or PMeBut2 to solutions of Pd[PCy3]2 or Pd-
[PMeBut2]2, respectively, seems therefore eminently reasonable,
but there is a possibility that the resonances of these two
compounds were for some reason severely exchange broadened
and therefore unobservable. In this case, since the 31P nuclei
are expected to remain chemically equivalent regardless of the
number of ligands coordinated to the palladium, the resonances
observed could conceivably be attributable to the 4:1 adducts,
and thus identification of the compounds formed is to this point
ambiguous.
Formation of Pd[PCy3][PBut3] directly from the two 2:1
homoleptic precursors shows that ligand exchange in this system
is facile, although exchange broadening is observed for only
Pd[PCy3]2, not for Pd[PCy3][PBut3] or Pd[PBut3]2. Similar
results were noted above where the 31P resonance of Pd[PCy3]2
broadened in the presence of free PCy3, while that of Pd[PBut3]2
remained sharp in the presence of free PBut3. Exchange
processes are probably associative for compounds PdL2, with
the low coordination number of two,2g and it is likely that
compounds containing relatively bulky PBut3 are too sterically
hindered to participate readily.
A very similar result is obtained on reaction of equimolar
amounts of Pd[PMeBut2]2 and Pd[PBut3]2 in toluene-d8. Again
an AX spin system with chemical shifts δ 85.4 and 41.8 and
2J(P-P) 246 Hz, attributable to the heteroleptic compound Pd-
[PMeBut2][PBut3] results, and again the new compound is in
equilibrium with its precursors Pd[PMeBut2]2 (δ 41.7, exchange
broadened) and Pd[PBut3]2 (δ 85.3). As above, a 31P-31P COSY
correlation spectrum of Pd[PMeBut2][PBut3] confirms the as-
signments.
Unambiguous determinations of coordination numbers could
be made via syntheses of mixed phosphine heteroleptic com-
pounds, since the spin-spin coupling patterns observed in the
31P NMR spectra would provide information concerning the
nature of any new species formed. Thus a bis-ligated compound
of the type Pd[PR3][PR′3] is expected to exhibit the spectrum
of an AX or AB spin system, depending on the ratio ∆ν/J,20
a
The third of the heteroleptic systems involve PCy3 and
PMeBut2, both with cone angles significantly smaller than that
of PBut3 (see above). Complicating matters, while the 31P
chemical shifts of coordinated PBut3 generally lie ∼20-40 ppm
downfield from the chemical shifts of coordinated PCy3 and
PMeBut2 (Table 1), giving rise to first-order 31P NMR spectra,
(19) (a) Dias, P. B.; Minas de Piedade, M. E.; Simo˜es, J. A. M. Coord.
Chem. ReV. 1994, 135/136, 737. (b) Sen, A.; Chen, J.; Vetter, W. M.;
Whittle, R. R. J. Am. Chem. Soc. 1987, 109, 148. (c) Immiri, A.; Musco,
A.; Mann, B. E. Inorg. Chim. Acta 1977, 21, L37.
(20) Bovey, F. A. Nuclear Magnetic Resonance Spectroscopy; Academic
Press: New York, 1969.

5236 Organometallics, Vol. 26, No. 21, 2007
Mitchell and Baird
achievable, many better resolved features can be recognized
(Figure 7e), although the limiting spectrum may not yet have
been reached. Prominent in the spectrum at -93 °C are singlets
at δ 7.52 and 7.95, which are assigned respectively to free
PMeBut2 and PCy3 shifted somewhat from the room-temperature
chemical shifts (Table 1) by temperature effects. There are also
strong singlets at δ 25.5 and 30.3, which are readily assigned
to the 3:1 homoleptic compounds Pd[PCy3]3 and Pd[PMeBut2]3,
respectively (see above).
Figure 6. 31P NMR spectrum at 25 °C of a solution containing
equimolar quantities of Pd[PCy3]2 and Pd[PMeBut2]2.
Much weaker resonances present in the spectrum at -93 °C
include several in the region δ 39-40; these are consistent with
the anticipated presence of Pd[PCy3]2, Pd[PMeBut2]2, and Pd-
[PCy3][PMeBut2], although the chemical shifts do not coincide
exactly with those listed in Table 1 because of temperature
effects. Also observed are two very broad resonances at δ ∼23.3
and ∼21.6, two less broadened 1:2:1 triplets at δ ∼32.2 and
∼34.4, and an extremely broad resonance at δ ∼35, which is
partially obscured by the low-field line of the triplet at δ ∼34.4.
None of these resonances appears in spectra of solutions
containing only PCy3 or PMeBut2, and thus they suggest the
presence of various of the possible heteroleptic compounds Pd-
[PCy3][PMeBut2]2, Pd[PCy3]2[PMeBut2], Pd[PCy3]2[PMeBut2]2,
Pd[PCy3][PMeBut2]3, and/or Pd[PCy3]3[PMeBut2]. Our initial
and somewhat surprising interpretation of the pair of 1:2:1
triplets at δ ∼32.2 and ∼34.4 was that they comprise an A2X2
spin system and thus are to be attributed to the 4:1 compound
Pd[PCy3]2[PMeBut2]2.
the 31P chemical shifts of Pd[PCy3]2 and Pd[PMeBut2]2 differ
by only ∼2 ppm. This (∼486 Hz at a probe frequency of ∼243
2
MHz) is comparable in magnitude to the values of J(P-P)
discussed above for Pd[PCy3][PBut3], and thus we anticipated
that a second-order 31P spectrum would be observed20 for the
compound Pd[PCy3][PMeBut2], for example.
The Pd(0)/PCy3/PMeBut2 system turned out to be much more
complex and interesting than anticipated, and it has been
investigated in rather more detail than were the two systems
involving PBut3. In the first experiment carried out, a 31P NMR
spectrum of a toluene-d8 solution containing equimolar amounts
of Pd[PMeBut2]2 and Pd[PCy3]2 was run at 25 °C, and as is
apparent from Figure 6, there are observed four resonances of
comparable intensity at δ 41.8, 40.6, 40.5, and 39.3, very similar
to the 31P chemical shifts of the corresponding homoleptic 2:1
2
compounds. The two apparent J(P-P) separations (i.e., the
differences in Hz between the resonances at δ 41.8 and 40.6,
and between those at δ 40.5 and 39.3) were identically 300 Hz
In order to obtain better resolved 1D spectra and satisfactory
2D 31P-31P correlation spectra, the NMR experiment was
repeated utilizing a higher concentration of Pd[PCy3]2 (0.025
mmol in 0.5 mL of toluene-d8) and 2 equiv of PMeBut2. The
spectra obtained at the various temperatures are largely similar
to those shown in Figure 7, although the exchange broadening
at 25 °C was so severe in the presence of the larger excess of
PMeBut2 that the two resonances at δ ∼14 and ∼42 were barely
distinguishable from the baseline. A spectrum run at -88 °C
(Figure 8) is similar to that shown in Figure 7e except that the
resonance of free PMeBut2 at δ 7.71 is as expected relatively
more intense than that of free PCy3 (δ 8.14). More importantly,
the triplets at δ ∼32.2 and ∼34.3 are now much more intense
than the resonances of the 2:1 compounds in the region δ 39-
40, consistent with their being attributed to a species of higher
coordination number. In addition they are sufficiently well
resolved that the splittings can be reasonably accurately
measured: ∼90.7 Hz for that centered at δ ∼34.3 and ∼101.6
Hz for that centered at δ ∼32.2. Furthermore, for both, the center
lines are also found to be broader than the corresponding outside
lines, and thus the two resonance patterns are not coupled 1:2:1
triplets but are more likely pairs of overlapping doublets!
2
and thus somewhat greater than the values of J(P-P) found
above for Pd[PMeBut2][PBut3] and Pd[PCy3][PBut3]. Thus ∆ν
would indeed appear to be comparable to 2J(P-P) for the
presumed species Pd[PCy3][PMeBut2] in solution. However, if
the two very closely spaced central lines at δ 40.6 and 40.5 are
taken as the two center peaks of an AB quartet, then the two
outside peaks should exhibit much lower intensities than are
observed20 and it seems very likely that most of the intensity
of each of the resonances at δ 41.8 and 39.3 is a result of the
presence in the solution of significant concentrations of the two
homoleptic compounds Pd[PMeBut2]2 and Pd[PCy3]2, respec-
tively. Strengthening this conclusion, a 31P-31P COSY NMR
experiment shows that the central peaks at δ 40.6 and 40.5
correlate with resonances at δ 41.8 and 39.3, respectively, hidden
under the main resonances having these chemical shifts.
Furthermore the chemical shifts of the two “outside” resonances
are essentially identical to those of Pd[PMeBut2]2 and Pd[PCy3]2,
respectively, and we note that, in Figure 1, the chemical shifts
of the two “outside” resonances of the quartet in the spectrum
of Pd[PCy3][PBut3] are very similar to those of the two
homoleptic compounds Pd[PBut3]2 and Pd[PCy3]2.
Consistent with our findings for the PdPMeBut2/PBut3 and
PdPCy3/PBut3 systems, it is not surprising to find significant
amounts of the two homoleptic compounds present in equilib-
rium with the heteroleptic analogue. The 31P chemical shifts of
the AB spin system do not, of course, coincide with any of the
four observed resonances, but are calculated to be δ 40.3 and
40.8 (Table 1).20
In the second experiment carried out to better understand this
system, a solution containing Pd[PCy3]2 (0.015 mmol) and 1
equiv of PMeBut2 was prepared in 0.5 mL of toluene-d8 at
21 °C, and 31P NMR spectra were obtained over the temperature
range +25 to -93 °C (Figure 7). As can be seen, the 31P NMR
spectrum at 25 °C exhibits only two very broad resonances at
δ ∼14 and ∼42. On cooling to -40 °C, these narrow somewhat
and shift upfield very slightly and, by -63 °C, both have begun
to decoalesce (Figure 7c). At -93 °C, the lowest temperature
The two broad resonances at δ ∼23.6 and ∼21.7 have also
gained in intensity relative to the resonances at δ 39-40, but
while the resonance at δ ∼23.6 remains featureless, that at δ
∼21.7 has become a broadened apparent triplet, with splittings
of ∼90.4 Hz, and it also appears to be comprised of overlapping
doublets. The resonances at δ ∼32.2, ∼34.3, ∼23.6, and ∼21.7
are all of comparable intensities. The extremely broad resonance
at δ ∼35 remains barely visible.
A
31P-31P correlation experiment was carried out on this
sample at -88 °C (Figure 9), and several features are notable.
For instance the resonances in the region δ 39-40 do not
correlate with any resonances at higher field, consistent with
their assignment solely to the homo- and heteroleptic 2:1
compounds. Of note, the apparent double doublets at δ ∼32.2,
∼34.3, and ∼21.7 are all mutually coupled, implying that they

Bisphosphinepalladium(0) Precursors for Cross-Coupling Catalysis
Organometallics, Vol. 26, No. 21, 2007 5237
Figure 7. Stacked plots of 31P NMR spectra of the reaction of Pd[PCy3]2 with 1 equiv of PMeBut2 at (a) 25 °C, (b) -40 °C, (c) -63 °C,
(d) -83 °C, and (e) -93 °C.
[PCy3][PMeBut2]2, with the multiplets at δ ∼32.2 and ∼34.3
attributed to two PMeBut2 rendered nonequivalent because they
are “frozen” at this low temperature in different rotational
conformations relative to the PCy3. In support of this suggestion,
we note that the room-temperature 31P NMR spectrum of trans-
PdCl2(PMeBut2)2 consists of two singlets of unequal intensity
because of the presence of two different rotamers in solution.21
The two resonances only coalesce at higher temperatures,
consistent with the chlorine ligands presenting a significant steric
barrier to rotation about the Pd-P bond. It is thus clearly
reasonable to suppose that the two PMeBut2 ligands in Pd[PCy3]-
[PMeBut2]2 assume different orientations in the most stable
conformation, with significant barriers to interconversion.
Alternatively, of course, the two PMeBut2 ligands may be
magnetically nonequivalent (diastereotopic) by virtue of the
Figure 8. 31P NMR spectra of the reaction of Pd[PCy3]2 with 2
equiv of PMeBut2 at -88 °C.
chirality of the coordinated PCy3.22
Assuming, as seems reasonable, that the magnitudes of 2J(P-
P) in these palladium(0) compounds are governed largely by
the Fermi contact term,23 the decrease in 2J(P-P) from 486 Hz
for Pd[PCy3][PMeBut2] to ∼96 Hz for Pd[PCy3][PMeBut2]2 is
consistent with the change from the linear sp-hybridized
structure to the trigonal sp2 analogue and supports our suggestion
of an increase in the coordination number. Weakening of the
Pd-P σ bonds because of greater steric crowding in the 3:1
compound should also result in a decrease in 2J(P-P).23
Observation of a near first-order spectral pattern between the
resonances at δ ∼32.2 and ∼34.3 arises from the fact that ∆ν
≈ 500 Hz and J ≈ 90 Hz and therefore ∆ν/J ≈ 5.6, much larger
than was found to be the case with Pd[PCy3][PMeBut2].
By process of elimination, the very broad resonance at δ
∼23.6 is assigned to the PCy3 ligands of the compound Pd-
[PCy3]2[PMeBut2], with the PMeBut2 resonance being the very
broad band at δ ∼35. These resonances appear not to be
correlated, but would not expected to be in view of their extreme
breadth.
(21) (a) Mitchell, E. A.; Baird, M. C. Unpublished results. For similar
results with the platinum analogue, see: (b) Luck, L. A.; Elcesser, W. L.;
Hubbard, J. L.; Bushweller, C. H. Magn. Reson. Chem. 1989, 27, 488.
(c) Dimeglio, C. M.; Luck, L. A.; Rithner, C. D.; Rheingold, A. L.;
Elcesser, W. L.; Hubbard, J. L.; Bushweller, C. H. J. Phys. Chem. 1990,
94, 6255.
(22) Bowmaker, G. A.; Brown, C. L.; Hart, R. D.; Healy, P. C.; Rickard,
C. E. F.; White, A. H. J. Chem. Soc., Dalton Trans. 1999, 881.
(23) (a) Pregosin, P. S.; Kinz, R. W. In NMR Basic Principles and
Progress; Diehl, P., Fluck, E., Kosfeld, R., Eds.; Springer-Verlag: New
York, 1979; Vol 16: 31P and 13C NMR of Transition Metal Phosphine
Complexes, Chapter B. (b) Harris, R. K. Nuclear Magnetic Resonance
Spectroscopy; Pitman: London, 1983; Chapter 8B.
Figure 9. 31P-31P correlation spectrum of a solution containing
Pd[PCy3]2 and 2 equiv of PMeBut2 at -88 °C.
comprise an AMX spin system and are to be attributed to one
of the three-coordinate compounds Pd[PCy3]2[PMeBut2] or Pd-
[PCy3][PMeBut2]2.
Since the 31P resonance of Pd[PMeBut2]3 lies downfield from
and is sharper than that of Pd[PCy3]3 (Figure 8), it seems likely
that the compound giving rise to the AMX spin system is Pd-

5238 Organometallics, Vol. 26, No. 21, 2007
Mitchell and Baird
largely of an unknown nature, proceeding at unknown rates and
resulting in unknown yields. Relative catalytic activities are often
deduced on the basis of yields of cross-coupling products and
then rationalized on the basis of steric and/or electronic
properties of the phosphines used. We initiated this research
with the notion that relative catalytic activities may rather reflect
the relative rates of reduction of palladium(II), but now we must
realize also that reactivities may vary because of differing
degrees to which the initially very small amounts of generated
palladium(0) species may be present as tris-phosphine com-
pounds rather than the generally assumed bisphosphine species.
Summary and Conclusions
Although reactions of Pd(η3-C3H5)(η5-C5H5) with 2 equiv
each of PCy3, PMeBut2, or PBut3 are slow at room temperature
and produce η1-allyl and dinuclear intermediates in the cases
of PCy3 and PMeBut2, reactions of all three ligands at 77 °C
proceed cleanly and at very useful rates to give Pd[PCy3]2 and
Pd[PMeBut2]2 within 1 h and Pd[PBut3]2 within 30 min. While
application of this methodology to syntheses of analogous Pd-
(0) compounds of other phosphines would ideally be preceded
by a brief 31P NMR spectroscopic assessment similar to those
described above, the known lability of Pd(η3-C3H5)(η5-C5H5)9
suggests that 1 h at 77 °C will normally suffice. The procedure
outlined here appears to provide the only reliable method
available to date for generating known concentrations of
bisphosphine-palladium(0) compounds quickly, reliably, and
under mild conditions.
Interestingly, and consistent with its greater steric bulk, Pd-
[PBut3]2 is quite disinclined to increase its coordination number,
its 31P NMR resonance being completely unaffected by added
PCy3, PMeBut2, or PBut3. In contrast, the 31P NMR resonances
of Pd[PCy3]2 and Pd[PMeBut2]2 are both broadened significantly
by added PCy3 or PMeBut2, and both ligands form 3:1
coordination compounds. Indeed, the homoleptic compounds
Pd[PCy3]3 and Pd[PMeBut2]3 are the dominant species in the
presence of excess phosphines at low temperatures, and even
the heteroleptic compounds Pd[PCy3]2[PMeBut2] and Pd[PCy3]-
[PMeBut2]2 appear to exist.
For example, given the aforementioned temperature variation
of the equilibrium constant for dissociation of Pd[PCy3]3 to
Pd[PCy3]2 + PCy3, the proportion of palladium(0) present at
25 °C as the tris complex if the PCy3:Pd ratio lies in the range
100:1 to 1000:1, as it must during the early stages of a cross-
coupling reaction involving the use of excess phosphine and
Pd(II) precatalysts, varies between 84% and 98%. Even at
100 °C, in the range of temperatures often used for Suzuki-
Miyaura cross-coupling reactions,1 the proportion of palladium-
(0) present as Pd[PCy3]3 varies between 48% and 90%. For
PMeBut2 under the same conditions, the calculated proportions
present as the tris species are lower but still in the range 32-
83% at 25 °C and 7-42% at 100 °C. Thus the fraction of the
total added palladium present as the tris complexes in catalytic
solutions can be quite significant.
The relative tendency of a particular phosphine to form the
more highly coordinated, presumably much less reactive species
may be a determining factor of the overall efficiency of a cross-
coupling process, one that should be considered during the
optimization of a catalytic cross-coupling protocol. In a paper
to follow, we shall describe a kinetics study involving the
oxidative addition reactions of bromobenzene to the palladium-
(0) compounds Pd[PCy3], Pd[PCy3]2, and Pd[PCy3]3.12
Equilibrium constants for dissociation of the compounds PdL3
to PdL2 and free L (L ) PCy3, PMeBut2) were measured over
a range of temperatures, and from the resulting linear plots of
ln KD vs 1/T, ∆H, and ∆S values of 21 kJ mol-1 and 59 J deg-1
mol-1, respectively, were found for the PCy3 system and 23 kJ
mol-1 and 86 J deg-1 mol-1, respectively, for PMeBut2. The
enthalpy values are presumably, to a first approximation at least,
a measure of the Pd-P bond dissociation energies. If so, then
the Pd(0)-phosphine bond energies are, as expected, consider-
ably lower than are those reported previously for both a Pd-
Acknowledgment. We are grateful to the Natural Sciences
and Engineering Research Council of Canada (Graduate Schol-
arships to E.A.M., Discovery Grant to M.C.B.) and the Canadian
Foundation for Innovation for financial assistance and Johnson
Matthey PLC for a generous loan of PdCl2. We also thank Dr.
F. Sauriol for assistance with the NMR experiments.
(II)-PPh3 bond (54 kJ mol-1 19b
)
and a Pt(0)-PCy3 bond (55
kJ mol-1).19c
The ready formation of the 3:1 compounds has serious
implications for cross-coupling processes involving bisphos-
phine-palladium(0) catalysts. As outlined in the Introduction,
such catalysts are most often presumed to be generated via
reductions of palladium(II) precursors in processes that are
OM700580D