1,1′-P/O-ferrocenyl ligands in palladium-catalyzed Suzuki coupling of aryl chlorides

Article abstract of DOI:10.1021/om050791j

Several new P/O-functionalized ferrocenyl ligands have been synthesized. They react with Pd₂(dba)₃ to promote the Suzuki cross-couplings of a range of arylboronic acids and aryl chlorides to give the desired biaryl products in high isolated yields. Different oxidative addition products were isolated and crystallographically identified. Their roles in the catalytic pathways are discussed.

Full text of DOI:10.1021/om050791j

Organometallics 2006, 25, 1199-1205
1199
1,1′-P/O-Ferrocenyl Ligands in Palladium-Catalyzed Suzuki
Coupling of Aryl Chlorides
Shihui Teo,† Zhiqiang Weng,‡ and T. S. Andy Hor*,‡
Institute of Chemical and Engineering Sciences, No. 1, Pesek Road, Jurong Island, Singapore 627833, and
Department of Chemistry, National UniVersity of Singapore, 3 Science DriVe 3,
Kent Ridge, Singapore 117543
ReceiVed September 13, 2005
Several new P/O-functionalized ferrocenyl ligands have been synthesized. They react with Pd2(dba)3
to promote the Suzuki cross-couplings of a range of arylboronic acids and aryl chlorides to give the
desired biaryl products in high isolated yields. Different oxidative addition products were isolated and
crystallographically identified. Their roles in the catalytic pathways are discussed.
being among the best known in asymmetric metal-catalyzed
reactions. The vastly different electronic and steric characters
of the P and O donors, coupled with the conformationally
flexible and coordinatively adaptive ferrocene backbone, make
this an attractive system that could demonstrate the intricacies
of ligand hemilability.8 Recently we reported success in the use
of the unsymmetrical [(1′-(di-tert-butylphosphino)ferrocen-1-
yl)methyl]phenylimine9 in supporting Pd(0) in Suzuki coupling10
of aryl chlorides and arylboronic acids. Herein, we focus on
the synthesis of some new 1,1′-phosphine-ether-functionalized
ferrocenyl ligands and their catalytic efficiency toward C-Cl
activation10a,b in Suzuki-Miyaura couplings. Through the
detection or/and isolation of a number of key species in the
catalytic process, we herein also demonstrate the rich underlying
coordination chemistry.
Introduction
Ferrocenyl-based ligands functionalized by different donor
atoms have a rich coordination and catalytic chemistry.1 The
hetero functionalities promote hemilability and coordinative
flexibility of the ligand. They underpin the successes experi-
enced by many P/P′-, P/N-, P/S- and even O/N-donating
ferrocenes.2 The P/O-ferrocenyl ligands remain rare, with a few
notable exceptions, such as [(rac)-PPF-OMe],3 (S)-(R)-bis-
(PPFOMe),4 and aryl-MOPFs,5 which are remarkably active
toward asymmetric Pd- and Ni-catalyzed amination of halides,3a,b
Grignard reactions,3c Suzuki cross-couplings,5a and associated
syntheses.4,5b,c A recent example is found in 1-(diphenylphos-
phino)-1′-methoxylferrocene toward Suzuki reactions of aryl
bromides and arylboronic acids.6 They compare well with the
aromatic-based P/O ligands7 that also showed promise in Suzuki
syntheses7a-d and asymmetric Heck reactions,7e with MOP7f,g
Results and Discussion
* To whom correspondence should be addressed. E-mail:
andyhor@nus.edu.sg.
The combinative use of strongly coordinating phosphine and
weakly basic cyclic (di)ether as pendants on a ferrocene skeletal
backbone is a key to our approach. Adapting known synthetic
methods,11 we synthesized and isolated the novel metalloligands
1a, 1b, and 2 in moderate yields (54-63%) (Scheme 1 and
Figure 1). The advantages of the use of these ligands include
their simple preparations and derivatization at the phosphorus
and cyclic diether sites. To examine the role of hetero func-
tionalities, we compared their activities with those of the related
monophosphine 312 and diphosphine 4 (Figure 1). A survey on
a representative coupling of 4-chloroacetophenone with phen-
ylboronic acid revealed a remarkable advantage of the presence
of the ether pendant (Table 1). The isolated yield of the coupling
† Institute of Chemical and Engineering Sciences.
‡ National University of Singapore.
(1) (a) Togni, A.; Hayashi, T. Ferrocenes: Homogeneous Catalysis,
Organic Synthesis, Materials Science; VCH: Weinheim, Germany, 1995;
Chapter 1. (b) Togni, A.; Haltermann, R. L. Metallocenes; Wiley-VCH:
Weinheim, Germany, 1998. (c) Long, N. J. Metallocenes: An Introduction
to Sandwich Complexes; Blackwell Science: Oxford, U.K., 1998.
(2) (a) Atkinson, R. C. J.; Gibson, V. C.; Long, N. J. Chem. Soc. ReV,
2004, 33, 313. (b) Gan, K. S.; Hor, T. S. A. Ferrocenes: Homogeneous
Catalysis, Organic Synthesis, Materials Science; VCH: Weinheim, Ger-
many, 1995. (c) Togni, A. Angew. Chem., Int. Ed. Engl. 1996, 35, 1475.
(d) Dai, L.-X.; Tu, T.; You, S.-L.; Deng, W.-P.; Hou, X.-L. Acc. Chem.
Res. 2003, 36, 659.
(3) (a) Wolfe, J. P.; Buchwald, S. L. Tetrahedron Lett. 1997, 38, 6359.
(b) Marcoux, J.-F.; Wagaw, S.; Buchwald, S. L. J. Org. Chem. 1997, 62,
1568. (c) Hiyashi, T.; Hayashizaki, K.; Kiyoi, T.; Ito, Y. J. Am. Chem.
Soc. 1988, 110, 8153.
(4) Ohmura, H.; Matsuhashi, H.; Tanaka, M.; Kuroboshi, M.; Hiyama,
T.; Hatanka, Y.; Goda, K.-I. J. Organomet. Chem. 1995, 499, 167.
(5) (a) Josen, J. F.; Johannson, M. Org. Lett. 2003, 5(17), 3025. (b)
Pederson, H. L.; Johannsen, M. Chem. Commun. 1999, 2517. (c) Pederson,
H. L.; Johannsen, M. J. Org. Chem. 2002, 67, 7892.
(6) Atkinson, R. C. J.; Gibson, V. C.; Long, N. J.; White, A. J. P.;
Williams, D. J. Organometallics 2004, 23, 2744.
(7) (a) Bei, X.; Turner, H. W.; Weinberg, W. H.; Guram, A. S. J. Org.
Chem, 1999, 64, 6797. (b) Bei, X.; Uno, T.; Norris, J. H.; Turner, W.;
Weinberg, W. H.; Guram, A. S. Organometallics 1999, 18, 1840. (c) Kwong,
F. Y.; Lam, W. H.; Yeung, C. H.; Chan, K. S.; Chan, A. S. C. Chem.
Commun. 2004, 1922. (d) Dai, W. M.; Li, Y.; Zhang, Y.; Lai, K. W.; Wu,
J. Tetrahedron Lett. 2004, 45, 1999. (e) Dai, W. M.; Yeung, K. K. Y.;
Wang, Y. Tetrahedron 2004, 60, 4425. (f) Hayashi, T. Acta Chem. Scad.
1996, 50, 259. (g) Hayashi, T. Acc. Chem. Res, 2000, 33, 354. (h) Dai,
W.-M.; Zhang, Y. Tetrahedron Lett. 2005, 46, 1377.
(8) Braunstein, P.; Naud, F. Angew. Chem., Int. Ed. 2001, 40, 680.
(9) Weng, Z.; Teo, S.; Koh, L. L.; Hor, T. S. A. Organometallics 2004,
23, 4342.
(10) (a) Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 4176.
(b) Miura, M. Angew. Chem., Int. Ed. 2004, 43, 2001. (c) Suzuki, A. J.
Organomet. Chem. 1999, 576, 147. (d) Miyamura, N.; Suzuki, A. Chem.
ReV. 1995, 95, 2457.
(11) (a) Moulton, R. D.; Bard, A. J. Organometallics 1988, 7, 351. (b)
Dong, T.-Y.; Lai, L.-L. J. Organomet. Chem. 1996, 509, 131. (c) Dong,
T.-Y.; Lai, L.-L. Synthesis 1995, 1231.
(12) For ferrocene-based monophosphine ligands in Suzuki cross-
coupling of aryl chlorides, see: (a) Pickett, T. E.; Richards, J. C. Tetrahedron
Lett. 2001, 42, 3767. (b) Liu, S. Y.; Choi, M. J.; Fu, G. C. Chem. Commun.
2001, 2408. (c) Kataoka, N.; Shelby, Q.; Stambuli, J. P.; Hartwig, J. F. J.
Org. Chem. 2002, 67, 5553. (d) Pickett, T. E.; Roca, F. X.; Richards, C. J.
J. Org. Chem. 2003, 68, 2592.
10.1021/om050791j CCC: $33.50 © 2006 American Chemical Society
Publication on Web 02/01/2006

1200 Organometallics, Vol. 25, No. 5, 2006
Teo et al.
Table 1. Ligand Effect on the Suzuki Cross-Coupling
Reaction of 4-Chloroacetophenone with Phenylboronic Acid
Scheme 1. Synthesis of a Phosphine-Cyclic(bi)ether
Ferrocenyl Ligand
entry
ligand (L)
yield (%)b
1
2
3
4
5
1a
1b
2
3
4
84
100
100
tracec
26
dentate) and 4 (bidentate chelate) are much less effective. These
point to the delicate but important balance between coordination
and noncoordination of the ligand as well as saturation and
unsaturation of the metal. The activities of 1 and 2 are
comparable to those of some P/O-ferrocenyl systems such as
aryl-MOPFs, which catalyzed the Suzuki coupling of 4-chlo-
roacetophenone with phenylboronic acid, giving up to 91% yield
at 5 mol % of Pd(OAc)2 and 6 mol % of the ligand under
ambient conditions.5a
a Pd2(dba)3. b Isolated yield. c GC yield; isolated yield of 59% when 5
mol % of Pd(OAc)2 and 6 mol % of L are used.
Ligand 1b is our preferred choice, because it is less air
sensitive than 2 and easier to handle experimentally. A typical
catalytic screening toward a series of aryl substrates using Pd2-
(dba)3 and ligands 1 are conducted in 1,4-dioxane at 70-100
°C in the presence of CsF or Cs2CO3 (Table 2). Use of a base
weaker than CsF, such as Cs2CO3, would generally result in
lower yields. These findings are consistent with published
results.13 The contrast in the activities of 1a and 1b is
exemplified in entries 6 and 12, which revealed that a simple
Figure 1. Different types of ferrocenyl ligands under survey.
product is quantitative for 1b and 2 and satisfactory for 1a,
whereas the controls (3 and 4) gave insignificant products. The
simultaneous presence of a strong coordinating group (phos-
phorus) and a much weaker and restrictive donor (oxygen) gives
the ligands (1 and 2) their hemilability and significantly
increased the cross-coupling product yield, whereas 3 (mono-
Table 2. Pd-Catalyzed Suzuki Coupling of Aryl Chlorides and Aryl Boronic Acids
a Pd2(dba)3. b Pd:ligand ) 1.0:1.1. c Not optimized. d Isolated yields. e p-Methoxylphenylboronic acid is used instead of phenylboronic acid.

Pd-Catalyzed Suzuki Coupling of Aryl Chlorides
Organometallics, Vol. 25, No. 5, 2006 1201
Scheme 2. Different Oxidation Addition Products from the P/O Ligands
Table 3. Selected Bond Lengths and Angles of Complexes
5a, 5b, 7a, and 7b
addition of a phenyl sidearm on the peripheral cyclic ether could
turn a mediocre (47%) to near-quantitative yield toward a stable
substrate such as 2-(chloromethyl)benzene. These somewhat
unexpected differences between 1a and 1b and the atypical
results found in the ortho-substituted aryl chlorides among the
arylboronic acids prompted us to examine further the possible
intermediates formed in these reaction mixtures. Although the
catalytic mechanism of Suzuki-type coupling has been well
researched10a and is outside the scope of our work, further
structural information on the resulting complexes formed from
these potentially hemilabile ligands would eventually lead to
better catalyst design.
5a
5b
7a
7b
Bond Lengths
Pd(1)-I(1)
Pd(1)-P(1)
Pd(1)-C(1)
Pd(1)-P(2)
Pd(1)-C(32)
2.6577(6) 2.6710(6) 2.6535(4)
2.2753(16) 2.2751(16) 2.3638(11) 2.354(2)
2.6452 (13)
2.031(7)
2.046(7)
1.995(11)
2.3415(12)
2.024(4)
Bond Angles
P(1)-Pd(1)-I(1)
P(1)-Pd(1)-C(1)
C(1)-Pd(1)-I(1)
I(1)-Pd-I(1A)
176.62(5) 177.88(5) 90.40(3)
91.34(17) 90.95(17)
86.88(17) 86.97(16)
91.56(7)
88.44(6)
180.00(1)
87.06(19) 86.356(19)
To meet the objective to trap and structurally characterize
species that are mechanistically relevant, we have identified
several oxidative addition products from stoichiometric reactions
(Scheme 2).
Pd(1)-I(1)-Pd(1A) 92.939(19) 93.644(19)
P(1)-Pd(1)-I(1A) 94.62(4)
95.72(4)
C(1)-Pd(1)-I(1A) 173.73(18) 173.32(16)
C(32)-Pd(1)-P(1)
89.88(11)
C(32)-Pd(1)-P(2)
90.41 (11)
177.10 (13)
176.18(4)
All the complexes show oxidative addition of Pd0, with
I-C6F5 giving [I-Pd-C6H5]. Solution NMR analysis suggested
that there is little perturbation at the cyclic ether moiety. X-ray
single-crystal crystallographic analysis of 5a confirmed this and
further revealed a Pd(II) square-planar dimeric structure with a
doubly bridging iodide. Ligand 1a coordinates as a monodentate
phosphine ligand with dangling ether (Figure 2 and Table 3).
This gives rare crystallographic evidence that the oxidative
addition step of a Pd(0)-catalyzed Suzuki coupling does not
necessarily yield the commonly assumed mononuclear species.
If a strongly coordinating chelating ligand (e.g. dppe, dppf, etc.)
is absent or when the supporting monodentate phosphines (e.g.
PR3) are in short supply, the catalyst can be stabilized by a
facile dimerization process. This takes advantage of the basicity
of the halide from the substrate, thereby allowing the metal to
C(32)-Pd(1)-I(1)
P(2)-Pd(1)-P(1)
P(1A)-Pd-P(1)
176.87(13)
gain saturation and minimizing catalyst decomposition in the
absence of a good coordinating solvent. When 1a is replaced
by ligand 1b, the similar dimeric complex 5b is isolated (Figure
3 and Table 3). Such dinuclear formation could be common
among oxidative addition products.
Complex 6 corresponds to the expected mononuclear Pd(II)
oxidative addition product with P/O chelation. The coordination
of ether lacks crystallographic support but is inferred from the
1H NMR shifts of the methylene protons neighboring the oxygen
functions. Complexes 7a and 7b are also mononuclear but
Figure 2. ORTEP representation of 5a (40% probability ellipsoids
Figure 3. ORTEP representation of 5b (40% probability ellipsoids
and hydrogen atoms omitted for clarity).
and hydrogen atoms omitted for clarity).

1202 Organometallics, Vol. 25, No. 5, 2006
Teo et al.
the desired hetero-biaryl product. Addition of free ligand 1 to
complex 5 leads to bridge-cleavage reactions to yield 7 and 8.
In polar solvents, which favor the cleavage of the weak Pd-O
link, 6 is unstable and readily dimerizes to 5. The dinuclear 5
is less reactive than 6 toward reaction with phenylboronic acid;
heating is generally required to facilitate the cross-coupling
reaction. Complex 6 is much more reactive because of the facile
Pd-O cleavage in a weak chelate.
Complexes 7 (7a and 7b) are inactive toward biaryl cross-
coupling in nonpolar solvents (e.g. toluene). Prolonged (over-
night) reflux only yielded C6F5I and biphenyl, which suggests
a competing homocoupling arising from disproportionation and
hydrolytic deboronation.15
Figure 4. ORTEP representation of 7a (40% probability ellipsoids
The above observations on the four different forms of
oxidative addition products (5-8) help us contruct two different
catalytic pathways via the dinuclear (route A) or geometric
isometric (route B) forms (Scheme 3). Since the latter route
stoichiometrically demands a 2-fold presence of ligand, one
could in principle push the catalytic path toward route B from
route A by increasing the ligand dosage. For 1a, doubling the
ligand dosage gives only a slight rise in TON (from ∼1000 to
2000). However, for 1b, there is a ∼20-fold TON increase (from
∼2000 (max) to ∼44 600) for most of the electron-rich
substrates when the ratio 1b:Pd (in Pd2(dba)3) is raised from
1.2:1.0 to 2.4:1.0. This process is probably controlled by the
cis-configured intermediate, 8b. The higher ligand concentration,
however, has little effect on the ortho-substituted electron neutral
substrates. In fact, for o-chlorotoluene, the product yield would
be significantly reduced to ∼10% (with respect to normal
conditions, as in entry 2 in Table 2). This raises the possibility
that different types of substrates would favor different mecha-
nistic pathways.16 The dimerization pathway (route A) goes
through a more electrophilic metal, with only one phosphine,
thus facilitating transmetalation with electron poorer substrates.
On the other hand, the geometric isomerization pathway (route
B) favors more the electron-rich aryl halides via cis-RPdIIXL2
species, which isomerizes and equilibrates readily with the
thermodynamically more stable trans form. The weakened
Pd-X interaction would promote reductive elimination.10b,c,17
When the experimental conditions do not favor route A, which
is preferred by the ortho-substituted substrates, cross-coupling
is thus suppressed.18 There is also a steric influence on the
pathways. When the oxidative addition product is sterically
demanding, the dimeric complex 5 is the more stable form and
is thermodynamically favored.19 Thus, when isolated 5b (1.0
mol % Pd) is added to a mixture of o-chlorotoluene, phenyl-
boronic acid, and CsF, the cross-coupling product yield increases
from 47% (Table 2 (entry 12)) to 55%. This lends support that
route A is operative. These are simplistic generalizations with
parallel pathways and competing reactions such as homo-
coupling, the latter of which is supported by the observed Pd
black and biphenyl. Complex 5b could also turn to 7b upon
and hydrogen atoms omitted for clarity).
Figure 5. ORTEP representation of 7b (40% probability ellipsoids
and hydrogen atoms omitted for clarity).
without ether coordination or chelation. Instead, two units of 1
are coordinated as a unidentate P-only donor.
Complexes 7a and 7b form even under molar equivalence
of Pd:1, although they are best prepared when ligand 1 is in
excess. X-ray single-crystal crystallographic analysis confirmed
a square-planar Pd(II) carrying two ligands of 1 with its dangling
ether unit (Figures 4 and 5 and Table 3). The trans orientation
of the two phosphines has significantly weakened the Pd-P
bond from 2.2753(16) Å in 5a to 2.3638(11) Å in 7a and
2.2751(16) Å in 5b to 2.354(2) Å in 7b. The Pd-C6F5 links in
7a (2.024(4) Å), 5a (2.031(7) Å), and 5b (2.046(7) Å) are
consistent with weak Pd-C bonds.14 The C5‚‚‚C5 ferrocenyl
torsional twists for 7a (124.1 and 84.7°) and 5a (89.2°) are well
short of an ideal anti twist (180°). This suggests that the ether
moiety is not sterically obstructive to the phosphine coordina-
tion. The ferrocenyl twist is facile, which enables the ligand
(1a or 1b) to adapt to either chelating or unidentate mode to
meet the saturation needs of the metal and, hence, catalyst.
There was no 5 or 8 detected (31P NMR) when Pd2(dba)3, 1,
and C6F5I were mixed in C6D6, suggesting that the relative
product formation is solvent dependent. In polar solvents (such
as THF and dioxane) at room temperature, 7 exists in equilib-
rium with 8. Complex 8a could be isolated from a THF mixture
at room temperature. Its catalytic role is demonstrated by its
stoichiometric reaction with phenylboronic acid, which gives
(15) (a) Cammidge, A. N.; Cre´py, K. V. L. Tetrahedron 2004, 60, 4377.
(b) Cammidge, A. N.; Cre´py, K. V. L. J. Org. Chem. 2003, 67, 6832. (c)
Watanabe, T.; Miyaura, N.; Suzuki, A. Synlett 1992, 207.
(16) (a) B-Landeros, F.; Hartwig, J. F. J. Am. Chem. Soc. 2005, 127,
6944. (b) Stambuli, J. P.; Incarvito, C. D.; Bu¨hl, M.; Hartwig, J. F. J. Am.
Chem. Soc. 2004, 126, 1184. (c) Goossen, L. J.; Koley, D.; Hermann, H.
L.; Thiel, W. Organometallics 2005, 24, 2398.
(17) (a) Kastrup, C. J.; Oldfield, S. V.; Rzepa, H. S. Dalton Trans. 2002,
2421. (b) Moore, J. W.; Pearson, R. G. Kinetics and Mechanism; Wiley:
New York, 1981; p 304.
(18) (a) Kotha, S.; Lahiri, K.; Kashinath, D. Tetrahedron 2002, 58, 9633.
(b) Goodson, F. E.; Wallow, T. I.; Novak, B. M. J. Am. Chem. Soc. 1997,
119, 12441.
(19) Echavarren, A. M.; Ca´rdenas, D. J. Metal-Catalyzed Cross-Coupling
Reactions, 2nd ed.; VCH: Weinheim, Germany, 2004; Chapter 1.
(13) See for example: (a) Walker, S. D.; Barder, T. E.; Martinelli, J.
R.; Buchwald, S. L. Angew. Chem., Int. Ed. 2004, 43, 1871. (b) Wolfe, J.
P.; Singer, R. A.; Yang, B. H.; Buchwald, S. L. J. Am. Chem. Soc. 1999,
121, 9550. (c) Littke, A. F.; Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2000,
122, 4020. (d) Stambuli, J. P.; Kuwano, R.; Hartwig, J. F. Angew. Chem.,
Int. Ed. 2002, 41, 4746.
(14) (a) Galardon, E.; Ramdeedul, S.; Brown, J. M.; Cowley, A.; Hii,
K. K.; Jutand, A. Angew. Chem., Int. Ed. 2002, 41, 1760. (b) Nishihara,
Y.; Onodera, H.; Osakada, K. Chem. Commun. 2004, 192.

Pd-Catalyzed Suzuki Coupling of Aryl Chlorides
Organometallics, Vol. 25, No. 5, 2006 1203
Scheme 3. Different Catalytic Pathways of Heterocoupling via a Dinuclear Intermediate (Route A) and Mononuclear
Geometric Isomers (Route B)
1′-(2,5-dioxacyclopentyl)ferrocene were prepared by modifications
of literature procedures.11 Pd2(dba)3 was prepared according to
literature procedures.20 All other reagents and chemicals were
obtained commercially and were used as received.
ligand addition, thus crossing route A to route B. Realistically,
both pathways are operative and their dominance is governed
sensitively by the choice of ligands and substrates, among others.
2. 1-(Dicyclohexylphosphino)-1′-(1-phenyl-2,5-dioxacyclo-
pentyl)ferrocene (1a). To a solution of 0.205 g of 1-bromo-1′-(1-
phenyl-2,5-dioxacyclopentyl)ferrocene (0.496 mmol) in dry THF
was added n-BuLi (1.56 M, 0.32 mL, 0.50 mmol) dropwise at -78
°C, and the resultant solution was stirred for 1 h. Chlorodicyclo-
hexylphosphine (0.1 mL, 0.53 mmol) was added and the resultant
solution stirred and warmed to room temperature for 2 h. The
resultant solution was filtered through Celite and the filtrate
concentrated and chromatographed on silica gel under an argon
atmosphere to give 0.184 g of 1a as a yellow solid (yield: 0.302
mmol, 61%). 1H NMR (C6D6): δ 7.80-7.22 (m, 5 H, Ph), 4.35 (t,
J ) 1.6 Hz, 2 H, Cp), 4.29 (t, J ) 2.0 Hz, 2 H, Cp), 4.22 (quart,
J ) 1.6 Hz, 2H, Cp), 4.13 (t, J ) 1.6 Hz, 2 H, Cp), 3.73-3.43 (m,
4 H, OCH2CH2O), 2.00 - 1.16 (m, 22 H, Cy). 31P NMR (C6D6):
δ -8.35. 13C NMR (C6D6): δ 129.01, 128.68, 128.36 (Ph), 110.26
(PhCO), 73.50, 73.35, 72.02, 71.99, 70.97, 69.57 (Cp), 65.61
(OCH2CH2O), 34.91, 34.68, 31.54, 31.51, 31.35, 28.48, 28.40,
28.35, 28.28, 28.27, 27.53, 27.50 (PCy). MS (FAB): m/z 531 ([M
+ H])+, 530 (M+). Anal. Calcd for C31H39FePO2: C, 70.18; H,
7.36. Found: C, 69.99; H, 7.37.
3. 1-(Dicyclohexylphosphino)-1′-(2,5-dioxacyclopentyl)ferro-
cene (1b). A 0.608 g portion of 1-bromo-1′-(2,5-dioxacyclopentyl)-
ferrocene (1.81 mmol) was used, similar to the procedure for 1a,
giving 1b as a yellow powder (yield: 0.522 g, 1.15 mmol, 63%).
1H NMR (C6D6): δ 5.85 (s, 1 H, HCOO), 4.43 (t, J ) 1.6 Hz, 2
H, Cp), 4.33 (t, J ) 2.0 Hz, 2H, Cp), 4.27 (quart, J ) 1.6 Hz, 2 H,
Cp), 4.12 (t, J ) 1.6 Hz, 2 H, Cp), 3.62-3.47 (m, 4H, OCH2-
CH2O), 1.88-1.13 (m, 22 H, Cy). 31P NMR (C6D6): δ -8.10. MS
(FAB): m/z 455 ([M + H])+, 454 (M)+. 13C NMR (C6D6): δ 103.82
(HCO), 73.42, 73.27, 71.56, 71.05, 69.18 (Cp), 65.86 (OCH2CH2O),
34.84, 34.66, 31.54, 31.50, 31.49, 31.35, 28.51, 28.37, 28.27, 27.53
(PCy). Anal. Calcd for C25H35FePO2: C, 66.07; H, 7.70. Found:
C, 66.05; H, 7.69.
Conclusion
We have offered structural evidence that oxidative addition
of Pd(0) in Suzuki coupling could proceed to yield several
isolable products that could demonstrate geometric and nucle-
arity isomeric properties. They proceed to give the desirable
cross-coupling products, the efficiency and effectiveness of
which depend largely on the ligand design. Use of hemilabile
P/O-ferrocenyl ligands potentially promotes a facile switchable
protection and deprotection mechanism at the active metal
center. It also adds stability to the metal through a direct
dimerization process. The flexible coordination behavior inher-
ent in a hybrid ligand system thus offers the metal several
options to gain stability without compromising on its vital
reactivitity. This is not achievable in a standard diphosphine
system. These ideas should allow us to design better catalysts
by the use of intelligent hemilabile ligands that could support
different structural forms of the catalyst. When the catalyst can
switch easily among different active forms, it should in principle
cater to a range of different guest substrates and reaction
conditions and, if needed, different catalytic pathways. Our
research is driven by these desirable outcomes.
Experimental Section
1. General Considerations. All reactions were carried out using
conventional Schlenk techniques under an inert atmosphere of
nitrogen or under argon in an M.Braun Labmaster 130 inert gas
system and dry deoxygenated solvents. Dry THF and hexanes were
obtained by distillation from Na/benzophenone. Dry dichlo-
romethane was distilled on CaH2. Nuclear magnetic resonance
spectra were recorded on a Bruker ACF 300 MHz FT NMR
spectrometer operating at 300.14 MHz for 1H, 75.43 MHz for 13C,
and 121.49 MHz for 31P. Solvent peaks are used as an internal
4. 1-(Di-tert-butylphosphino)-1′-(2,5-dioxacyclopentyl)ferro-
cene (2). A 0.860 g portion of 1-bromo-1′-(2,5-dioxacyclopentyl)-
ferrocene (2.56 mmol) reacted with chlorodi-tert-butylphosphine
(0.43 mL, 2.56 mmol), similar to the procedure for 1a, to give 2 as
an orange-red oil (yield: 0.553 g, 54%). 1H NMR (C6D6): δ 5.79
(s, 1 H, HCOO), 4.37 (t, J ) 2.0 Hz, 2 H, Cp), 4.34 (t, J ) 1.6 Hz,
1
reference relative to TMS for H and 13C chemical shifts (ppm);
31P chemical shifts are relative to an 85% H3PO4 external reference.
Coupling constants are given in hertz. The following abbreviations
are used: s, singlet; d, doublet; t, triplet; m, multiplet. Mass spectra
were obtained on a Finnigan Mat 95XL-T spectrometer. Elemental
analyses were performed by the in-house microanalytical laboratory.
1-Bromo-1′-(1-phenyl-2,5-dioxacyclopentyl)ferrocene and 1-bromo-
(20) Ukai, T.; Kawazura, H.; Ishii, Y.; Bonnet, J. J.; Ibers, J. A. J.
Organomet. Chem. 1974, 65, 253.

1204 Organometallics, Vol. 25, No. 5, 2006
Teo et al.
2 H, Cp), 4.29 (quart, J ) 1.6 Hz, 2 H, Cp), 4.09 (t, J ) 1.6 Hz,
2 H, Cp), 3.62-3.47 (m, 4 H, OCH2CH2O), 1.23,1.19 (d, 9H,
CCH3). 31P NMR (C6D6): δ 27.5. 13C NMR (C6D6): δ 103.72
(PhCO), 74.84, 74.69, 71.57, 71.32, 69.85, 69.24 (Cp), 65.87
(OCH2CH2O), 33.61, 33.20, 31.86, 31.69, (PtBu). MS (FAB): m/z
403 ([M + H]+), 402 (M)+.
2.27-1.10 (m, 44 H, Cy). 31P NMR (C6D6): δ 25.5. MS (FAB):
m/z 1460 (M+). This compound was inevitably contaminated by
dba, and hence, a satisfactory elemental analysis could not be
obtained.
7.2. trans-[(1b)2Pd(I)(C6F5)] (7b). Ligand 1b (23 mg, 0.048
mmol) reacted with Pd2(dba)3 and C6F5I according to a procedure
1
similar to that for 7a. A yellow oil of 7b was obtained. H NMR
5.1. [(1a)Pd(I)(C6F5)]2 (5a). Ligand 1a (30 mg, 0.0633 mmol),
Pd2(dba)3 (29 mg, 0.0317 mmol), and C6F5I (28 mg, 0.0945 mmol)
were dissolved in THF, and the mixture was stirred for 5 h. The
reaction mixture was filtered through a layer of Celite, and the
solvent was removed. The dark brown oil was washed with hexane
until a brown solid, 5a, was left. This residual solid was redissolved
in CH2Cl2 and overlayered with hexane to obtain dark brown
(C6D6): δ 5.46 (s, 1 H, HCOO), 4.15 (s, 4 H, Cp), 3.92 (s, 4 H,
Cp), 3.84 (t, J ) 1.62 Hz, 4 H, Cp), 3.75 (s, 4 H, Cp), 3.52-3.39
(m, 8 H, OCH2CH2O), 2.16-1.16 (m, 44 H, Cy). 31P NMR
(C6D6): δ 25.2. MS (FAB): m/z 1460 (M+). This compound was
inevitably contaminated by dba, and hence, a satisfactory elemental
analysis could not be obtained.
1
8. cis-[(1a)2Pd(I)(C6F5)] (8a). Ligand 1a (20 mg, 0.042 mmol),
Pd2(dba)3 (11 mg, 0.012 mmol), and C6F5I (28 mg, 0.0945 mmol)
were mixed and stirred at room temperature in THF for 1 h. The
resultant orange solution was filtered to remove the unreacted Pd2-
(dba)3. The orange filtrate was evaporated, dried, and redissolved
in a minimum amount of Et2O and left in the refrigerator (ca. -20
°C) overnight. The mixture was filtered, and the solid that was
collected was dried and washed with a minimum amount of hexane
to obtain the crude product 8a as a red-orange oil (yield: ca. 20.1
crystals of 5a (yield: 23.5 mg, 40%). H NMR (C6D6): δ 7.82-
6.90 (m, Ph), 4.00 (s, 4 H, Cp), 3.94 (s, 4 H, Cp), 3.78 (s, 8 H,
Cp), 3.60-3.54 (m, 4 H, OCH2CH2O), 3.33-3.31 (m, 4 H, OCH2-
CH2O), 1.32-0.97 (m, 44 H, Cy). 31P NMR (C6D6): δ 45.2. MS
(FAB): m/z 1861 (M+), 803 ([1/2M - I]+). Anal. Calcd for C74H78-
Fe2P2O4F10I2‚3CH2Cl2: C, 43.67; H, 3.97. Found: C, 43.23; H,
4.07. X-ray-quality single crystals of 6a were grown from a CH2-
Cl2/hexane solution.
5.2. [(1b)Pd(I)(C6F5)]2 (5b). Ligand 1b (29 mg, 0.062 mmol)
reacted with Pd2(dba)3 and C6F5I, similarly to the procedure
described for 5a. A dark brown solid of 5b was obtained (yield: 8
1
mg). H NMR (C6D6): δ 7.81-6.88 (m), 4.80 (s, 4 H, Cp), 4.69
(s, 4 H, Cp), 4.45 (s, 4 H, Cp), 4.39 (m, 4 H, Cp), 3.72-3.68 and
3.44-3.40 (m, 8 H, OCH2CH2O), 1.79-1.23 (m, 44H, Cy). 31P
NMR (C6D6): δ 30.0. MS (FAB): m/z 1460 (M+). A satisfactory
elemental analysis could not be obtained on the oily product.
9. NMR Studies. Experiments were performed at room tem-
perature with 0.1 mmol of Pd2(dba)3 and 0.1 mmol of ligand in
various solvents: C6D6, toluene, THF, and dioxane. After addition
of the aryl iodide, the spectra of the mixture (in a solvent such as
C6D6) were recorded at regular time intervals until no further
changes were observed. Phenylboronic acid (1.3 equiv) and Cs2-
CO3 (1.5 equiv) were added to the mixture and monitored by 31P
NMR spectroscopy at regular intervals until there were no further
changes or conversions.
1
mg, 15%). H NMR (C6D6): δ 5.39 (s, 1H, HCOO), 4.10 (t, J )
2.0 Hz, 4H, Cp), 3.88 (s, 4 H, Cp), 3.76 (s, 8 H, Cp), 3.55-3.35
(m, 8 H, OCH2CH2O), 2.67-1.14 (m, 44 H, Cy). 31P NMR
(C6D6): δ 45.6. MS (FAB): m/z 727 ([1/2M - I]+). X-ray-quality
single crystals of 5b were grown from a CH2Cl2/hexane solution.
6.1. [(1a)Pd(I)(C6F5)] (6a). Ligand 1a (30 mg, 0.0633 mmol),
Pd2(dba)3 (29 mg, 0.0317 mmol), and C6F5I (28 mg, 0.0945 mmol)
were dissolved in THF, and the mixture was stirred for 1 h. The
reaction mixture was filtered through a layer of Celite, and the
solvent was removed. The orange oil was recrystallized in Et2O to
obtain a red solid of 6a and traces of dba (yield: ca. 29.8 mg,
30%). 1H NMR (C6D6): δ 7.80-7.03 (m), 4.48 (s, 2 H, Cp), 4.59
(s, 2 H, Cp), 4.44 (s, 2 H, Cp), 4.35 (s, 2 H, Cp), 3.68-3.41 (m,
4 H, OCH2CH2O), 2.14-1.12 (m, 44H, Cy). 31P NMR (C6D6): δ
20.2. MS (FAB): m/z 931 ([M + H]-), 803 ([M - I]+). This
compound was inevitably contaminated by dba, and hence, a
satisfactory elemental analysis could not be obtained.
10. General Procedure for Coupling Reactions of Aryl
Chlorides with Boronic Acids. A typical procedure is given for
the reaction represented in entry 2 of Table 2. Ligand 1a (5 mg,
0.0110 mmol, 1.1 mol %), Pd2(dba)3 (5 mg, 0.005 mmol, 1 mol %
Pd), phenylboronic acid (146 mg, 1.2 mmol), and CsF (2.3 mmol)
were introduced into a flask filled with N2 gas. 4-Chloroacetophenyl
(154 mg, 130 µl, 1 mmol) was injected into the flask via a
microsyringe, followed by addition of 1,4-dioxane. The mixture
was stirred at 90 °C for 16 h, under an N2 atmosphere. The solvent
was then removed under reduced pressure, followed by addition
of H2O (10 mL) and Et2O (10 mL). Extraction was repeated twice
with Et2O. The organic layer was collected and evaporated in vacuo
to dryness. The crude product was purified by column chromatog-
raphy on silica with hexanes/ethyl acetate as eluent to give 196
mg of 4-acetylbiphenyl as a solid, verified by GC/MS spectroscopy.
11. Crystal Structure Determinations. The crystals of 5a, 5b,
7a, and 7b were mounted on quartz fibers and X-ray data collected
on a Bruker AXS APEX diffractometer, equipped with a CCD
detector at -50 °C, using Mo KR radiation (λ 0.710 73 Å). The
data were corrected for Lorentz and polarization effects with the
SMART suite of programs21 and for absorption effects with
SADABS.22 Structure solution and refinement were carried out with
the SHELXTL suite of programs.23 The structures were solved by
direct methods to locate the heavy atoms, followed by difference
maps for the light non-hydrogen atoms. For 5a, in the asymmetric
unit there was half of the dimer and one dichloromethane. The
6.2. [(1b)Pd(I)(C6F5)] (6b). Ligand 1b (45.6 mg, 0.101 mmol),
Pd2(dba)3 (45 mg, 0.050 mmol), and C6F5I (50 mg, 0.170 mmol)
were dissolved in THF, and the mixture was stirred for 1 h. The
reaction mixture was filtered through a layer of Celite, and the
solvent was removed. The dark brown oil was washed copiously
with hexane to give a solid residue of 5b. The washings were
collected and evaporated to dryness to obtain a reddish brown solid,
which was recrystallized in Et2O to obtain the red solid 6b and
traces of dba (yield: ca. 63.7 mg, 70%). 1H NMR (C6D6): δ 5.78
(s, 1 H, HCOO), 4.70 (s, 2 H, Cp), 4.54 (m, 4 H, Cp), 4.42 (s, 2
H, Cp), 3.73-3.41 (m, 4 H, OCH2CH2O), 2.00-1.12 (m, 22H, Cy).
31P NMR (C6D6): δ 20.2. MS (FAB): m/z 727 ([M - I]+). This
compound was inevitably contaminated by dba, and hence, a
satisfactory elemental analysis could not be obtained.
7.1. trans-[(1a)2Pd(I)(C6F5)] (7a). Ligand 1a (24 mg, 0.045
mmol), Pd2(dba)3 (11 mg, 0.012 mmol), and C6F5I (14 mg, 0.047
mmol) were dissolved in a minimum amount of toluene, and the
mixture was stirred for 12 h. The toluene solution was then
concentrated. The mixture was then left to stand in the freezer at
-30 °C overnight. The resulting orange oil was then filtered through
Celite. The orange filtrate was pumped, dried, and washed with
hexane. The orange oil was recrystallized in CH2Cl2 and layered
over hexane to obtain orange-red crystals of 7a and yellow crystals
(21) SMART, version 5.1; Bruker Analytical X-ray Systems, Madison,
WI, 2000.
(22) Sheldrick, G. M. SADABS, a Program for Empirical Absorption
Correction; University of Go¨ttingen, Go¨ttingen, Germany, 2000.
(23) SHELXTL, version 5.03; Bruker Analytical X-ray Systems,
Madison, WI, 1997.
1
of dba (yield: 28 mg, 80%). H NMR (C6D6): δ 7.74-7.06 (m,
10 H, Ph), 4.30 (s, 4 H, Cp), 3.99 (s, 4 H, Cp), 3.91 (s, 4 H, Cp),
3.89 (s, 4 H, Cp), 3.67-3.66, 3.45-3.40 (m, 8 H, OCH2CH2O),

Pd-Catalyzed Suzuki Coupling of Aryl Chlorides
Organometallics, Vol. 25, No. 5, 2006 1205
Table 4. Crystal Data and Structure Refinement Parameters for Complexes 5a, 5b, 7a and 7ba
5a‚2CH2Cl2
5b‚2CH2Cl2
7a‚C6H14
7b
empirical formula
Mr
C37H39F5FeIO2PPd‚CH2Cl2
1015.73
C31H35F5FeIO2PPd‚CH2Cl2
1879.27
C68H78F5Fe2IO4P2Pd‚C6H14
1547.42
C56H70F5Fe2IO4P2Pd
1309.06
temp, K
223(2)
293(2)
223 (2)
243(2)
cryst color
dark brown
0.30 × 0.20 × 0.10
monoclinic
P21/c
11.9657(8)
16.0604(11)
20.2335(14)
90
98.876(2)
90
3841.8(5)
4
brown
orange
orange
cryst size, mm3
0.28 × 0.08 × 0.06
0.14 × 0.08 × 0.06
0.08 × 0.06 × 0.03
monoclinic
C2/c
17.9960(15)
19.5414(15)
17.0394(15)
90
114.753
90
5441.6(8)
4
cryst system
space group
a, Å
b, Å
triclinic
triclinic
P1h
P1h
10.6130(5)
12.0109(6)
13.3195(7)
90.4090(10)
95.3170(10)
94.3280(10
1685.56(15)
1
11.4379(4)
17.5926(6)
18.4657(7)
75.9770(10)
76.1800(10)
79.4350(10)
3469.4(2)
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
2
density, Mg/m3
abs coeff, mm-1
F(000)
1.756
1.888
2016
1.851
2.143
928
1.481
1.221
1588
1.598
1.541
2656
θ range for data collection, deg
1.63-27.50
-11 e h e 15,
-20 e k e 15,
-26 e l e 26
26 142
1.93-25.00
-12 e h e 12,
-14 e k e 14,
-15 e l e 15
18 450
1.84-27.50
-14 e h e 14,
-22 e k e 22,
-23 e l e 23
36 447
1.62-25.00
-21 e h e 21,
-23 e k e 18,
-20 e l e 15
15 772
index ranges
no. of rflns collected
no. of indep rflns
8828 (Rint ) 0.0499)
0.8337 and 0.6013
8828/24/460
R1 ) 0.0651,
wR2 ) 0.1683
R1 ) 0.0994,
wR2 ) 0.1901
1.045
5941 (Rint ) 0.0447)
0.8822 and 0.5853
5941/42/432
R1 ) 0.0512,
wR2 ) 0.1238
R1 ) 0.0656,
wR2 ) 0.1305
1.063
15 871 (Rint ) 0.0546)
0.9303 and 0.8476
15 871/0/804
R1 ) 0.0607,
wR2 ) 0.1150
R1 ) 0.0956,
wR2 ) 0.1274
1.021
4788 (Rint ) 0.1337)
0.9552 and 0.8866
4788/0/323
R1 ) 0.0771,
wR2 ) 0.1332
R1 ) 0.1521,
wR2 ) 0.1555
1.010
max and min transmissn
no. of data/restraints/params
final R indices (I >2σ(I))a,b
R indices (all data)
goodness of fit on F2 c
largest diff peak and hole, e Å-3
1.959 and -0.820
1.167 and -0.778
0.957 and -0.478
0.854 and -0.974
2
2
2
2
2
a R1 ) ||Fo| - |Fc||/|Fo|. b wR2 ) [(∑w(Fo - Fc )2)/(∑w(Fo )2)]1/2
.
c GOF ) [(∑w(Fo - Fc )2)/(n - p)]1/2
.
whole dimer was generated through the center of symmetry. For
7a, the asymmetric unit contains one molecule of the title compound
and one hexane. For 7b, the asymmetric unit consists of half of
the title molecule. Further details of the crystallographic information
are provided in the Supporting Information. Crystal data and
structure refinement parameters for 5a, 5b, 7a, and 7b are given in
Table 4.
and Engineering Sciences (Singapore) (Grant No. 012-101-0035)
and the National University of Singapore for support. We are
indebted to L. L. Koh and G. K. Tan for assistance in the
crystalographic analysis.
Supporting Information Available: CIF files giving crystal
data for complexes 5a, 5b, 7a, and 7b. This material is available
free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. We acknowledge the Agency for Science,
Technology & Research (Singapore), the Institute of Chemical
OM050791J