Di-t-butyl(ferrocenylmethyl)phosphine: Air-stability, structural characterization, coordination chemistry, and application to palladium-catalyzed cross-coupling reactions

Article abstract of DOI:10.1016/j.jorganchem.2004.12.022

Di-t-butyl(ferrocenylmethyl)phosphine (1) has been isolated and structurally characterized. This ligand was found to be reasonably air stable as a solid and it has been shown to possess electron donating ability similar to that of tri-i-propylphosphine. A

Full text of DOI:10.1016/j.jorganchem.2004.12.022

Journal of Organometallic Chemistry 690 (2005) 1478–1486
www.elsevier.com/locate/jorganchem
Di-t-butyl(ferrocenylmethyl)phosphine: air-stability,
structural characterization, coordination chemistry, and
application to palladium-catalyzed cross-coupling reactions
Michael D. Sliger, Grant A. Broker, Scott T. Griffin, Robin D. Rogers,
*
Kevin H. Shaughnessy
Department of Chemistry, The University of Alabama, P.O. Box 870336, Tuscaloosa, AL 35487-0336, USA
Received 12 October 2004; accepted 13 December 2004
Available online 19 January 2005
Abstract
Di-t-butyl(ferrocenylmethyl)phosphine (1) has been isolated and structurally characterized. This ligand was found to be reasonably
air stable as a solid and it has been shown to possess electron donating ability similar to that of tri-i-propylphosphine. A palladium
catalyst bearing this ligand performed room temperature Suzuki–Miyaura coupling reactions with aryl bromides. Modest Heck cou-
pling reactivity with aryl bromides was also observed at 100 ꢀC. Complexation of 1 with Pd2(dba)3 led to formation of (1)2Pd0. Addi-
tion of 4-bromoanisole to solutions containing both 1 and Pd2(dba)3 led to formation of an oxidative addition product when 1:Pd
ratios were 61. With a 2:1 ratio of 1:Pd, monophosphine complex formation and oxidative addition were significantly inhibited.
ꢁ 2004 Elsevier B.V. All rights reserved.
Keywords: Phosphine; Ferrocene; Palladium; Cross-coupling; Suzuki coupling; Heck coupling
1. Introduction
added to the cyclopentadienyl ring a to the phosphine-
bearing carbon to form mono- and polydentate ligands
which are suitable for use in asymmetric catalysis [17,18].
Phosphines have frequently been used over the past
decade in the development of efficient catalysts for vari-
ous cross-coupling reactions. Much of the research focus
has been centered on the use of sterically demanding ter-
tiary alkylphosphine ligands that enhance the rate deter-
mining oxidative addition step. Ligands such as t-Bu3P
[19–22], (o-biphenyl)PR2 [23–25], BuP(1-adamantyl)2
[26] and others [14,15,27] have been used to form cata-
lysts that show high activities at moderate temperatures
(20–100 ꢀC), even with unactivated aryl chlorides. Many
of these ligands oxidize rapidly in air, so they need to be
handled in an inert atmosphere or converted to the cor-
responding phosphonium salt [28]. Ligands with similar
bulk and electron donating ability that show resistance
to oxidation would lead to simpler reaction preparations
and longer catalyst lifetimes.
Ferrocene based phosphines have become increas-
ingly popular for use as ligands in several catalytic pro-
cesses. The electron-rich metallocene helps provide the
electronic and steric characteristics that are often re-
quired to form a highly active catalyst. Palladium and
nickel complexes with 1,10-bis(diphenylphosphino)ferro-
cene (dppf) and similar bidentate ligands have shown
high activity in several types of C–C bond forming
cross-coupling reactions [1–11]. Examples of high activ-
ity in these processes using monophosphine derivatives
have also been reported [8,12–16]. Substituents can be
*
Corresponding author. Tel.: +1 205 348 4435; fax: +1 205 348
9104.
E-mail address: kshaughn@bama.ua.edu (K.H. Shaughnessy).
0022-328X/$ - see front matter ꢁ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.12.022

M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
1479
Henderson [29,30] recently investigated the properties
and coordination abilities of various primary phos-
phines and an arsine bearing ferrocene substituents. De-
spite the fact that these ligand types are commonly very
air-sensitive, it was shown that primary phosphines with
an alkyl linkage between the cyclopentadienyl ligand of
the ferrocene and the phosphorus or arsenic atom were
stable in air for over two years. In comparison, ferroce-
nylphosphine oxidizes in air after a few days. This air
stability was an unexpected result since the aromatic ef-
fects of the cyclopentadienyl ring towards the central
atom should be greatly diminished with the alkyl bridge
present. A crystal structure of ferrocenylethylarsine
shows that the ferrocene points away from the arsenic
atom, which suggests that this air stability is not due
to a weak intramolecular coordination or an indirect
aromatic effect of the ferrocene to the phosphorus atom.
Recently, tertiary ferrocenylmethylphosphines were
synthesized and used as ligands for Suzuki–Miyaura
palladium catalysts [31]. However, these phosphines
possessed an aryl substituent on the methyl bridge in or-
der to maximize steric bulk. Catalytic activity of these
complexes was high using many different aryl bromide
reactants, but the t-butyl-substituted ligand was not
air stable. It is unclear if this decrease in air stability
was due to the introduction of the t-butyl substituents
on phosphorous, the phenyl group on the methylene
bridge, or a combination of the two. Di-t-butyl(ferroce-
nylmethyl)phosphine (1), which lacks the phenyl group
on the methylene bridge, has been reported [32] but little
is known about the properties of this ligand, including
its reactivity towards oxygen. We have taken a closer
look at 1, examining its reactivity and its viability in var-
ious cross-coupling reactions.
Fig. 1. ORTEP drawing of the prism-like crystal of di-t-butyl(ferr-
ocenylmethyl)phosphine (1a), rotomer A. Ellipsoids are drawn at the
50% probability. Hydrogen atoms were placed in idealized positions.
2. Results and discussion
Fig. 2. ORTEP drawings of one of the two molecules of 1 in the
asymmetric unit of the flattened rod shaped crystal of di-t-butyl(ferr-
ocenylmethyl)phosphine (1b). The structure was inverted to have the
same orientation as 1a. Ellipsoids are drawn at the 50% probability.
Hydrogen atoms were placed in idealized positions.
The synthesis of 1 was performed using a previously
reported method, but the procedure needed to be ex-
tended since the authors did not fully isolate or charac-
terize the ligand (Eq. (1)) [32]. Recrystallization of the
crude product from methanol led to orange crystals with
either a prismatic (1a, Fig. 1) or flattened rod shape (1b,
Fig. 2). To investigate these differences, X-ray crystal
structures of both crystal types were acquired. The crys-
tallographic data is presented in Table 1. Selected bond
lengths and angles for both crystals 1a and 1b are shown
in Table 2. ORTEP drawings of the resultant spectra are
presented in Figs. 1 and 2.
Crystal 1a resolved into a structure containing only
one phosphine in the asymmetric unit, although two dif-
ferent rotomers of one of the t-Bu groups were observed.
Crystal 1b had two molecules of 1 in the asymmetric
unit, which had similar, although not identical, bond
lengths and angles. Structures 1a and 1b were very sim-
ilar with only small variations in the corresponding
bond lengths and angles of these polymorphs. DSC
studies of both crystal types showed a sharp melting
point at 67.4 ꢀC, so there appears to be little difference
in the physical properties of the two polymorphs.
NMe2
P(t-Bu)2
Fe
+
HP(t-Bu)2
Fe
The Tolman electronic parameter, m, for this ligand
was determined spectroscopically from (1)Ni(CO)3,
which was prepared in CH2Cl2 [33]. The symmetric
AcOH
1
ð1Þ

1480
M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
Table 1
X-ray crystallographic data
Compound
1a
1b(1,2)
Formula
C19H29FeP
Orange, prismatic
Monoclinic
P21
C19H29FeP
Color, shape
Crystal system
Space group
Unit cell
Orange, flattened rod
Monoclinic
P21/c
˚
a (A)
5.992(4)
11.410(7)
13.198(8)
90
28.310(7)
11.188(3)
11.591(3)
90
˚
b (A)
˚
c (A)
a (ꢀ)
b (ꢀ)
c (ꢀ)
95.644(11)
90
101.729(4)
90
3
˚
V (A )
897.9(9)
2
1.273
3594.6(15)
8
1.272
Z
Density calcd. (g/cm3)
Crystal size (mm)
k (Mo Ka) (mmꢀ1
F(0 0 0)
T (K)
0.36 · 0.24 · 0.14
0.921
368
0.41 · 0.22 · 0.14
0.921
1472
)
173(2)
1.55, 23.31
4097
173(2)
0.73, 23.30
15896
hmin, max (ꢀ)
Reflections collected
Independent reflections, Rint
Goodness-or-fit on F2a
Final R indices [I > 2r(I)], R1b, w2
R indices (all data)
2527, 0.0324
0.979
0.0357, 0.0832
5138, 0.0202
1.269
c
0.0680, 0.1630
0.0726, 0.1645
na
0.0405, 0.0846
0.07(2)
Absolute structure parameter
P
2
GOF ¼ f ½wðF 2o ꢀ F 2c Þ ꢁ=ðn ꢀ pÞg1=2, where n is the number of data and p is the number of parameters refined.
a
b
c
P
P
R = iFo|ꢀ|Fci/ |Fo|.
P
P
2
2
1=2
wR2 ¼ f ½wðF 2o ꢀ F c2Þ ꢁ= ðwðF 2oÞ ꢁg . Mo Ka radiation (k = 0.71073 A).
˚
Table 2
63 h heating at 100 ꢀC (2060, 1987 and 1976 cmꢀ1 in tol-
uene) [34]. A small amount of residual Ni(CO)4 was also
detected by FTIR spectroscopy (2046 cmꢀ1) which indi-
cates a degree of substitution of 0.94 and an experimen-
tal ligand cone angle of approximately 186ꢀ. This
effective steric bulk is slightly greater than that seen
for t-Bu3P, suggesting that the ferrocenylmethyl group
has a significant steric effect.
˚
Selected bond lengths (A) and angles (ꢀ)
a
1a
1ba
1b0
Bond lengths
P–C11
C11–C9
Fe–C9
P–C12
P–C16
1.868(4)
1.492(6)
2.048(4)
1.876(4)
1.895(4)
1.883(7)
1.502(9)
2.056(6)
1.888(7)
1.892(7)
1.870(7)
1.512(9)
2.054(6)
1.899(7)
1.905(7)
Bond angles
Despite the presence of two tertiary butyl groups on
the central phosphorus atom, 1 shows significant air sta-
bility. It was initially unknown if this characteristic, seen
in analogous primary (ferrocenylalkyl)phosphines,
would be retained due to the addition of two highly elec-
tron donating groups to the phosphorus atom [29,30].
After 54 days in the solid-state, 31P NMR spectroscopy
showed approximately 20% conversion of the ligand
(34.1 ppm in THF) to a compound that appears to be
di-t-butyl(ferrocenylmethyl)phosphine oxide (57.1 ppm
in THF). The reasons for the low oxidation levels are
still unclear since X-ray crystallography does not indi-
cate any direct or indirect interaction between the ferro-
cene and the phosphorus atom. However, if a THF
solution of the ligand is exposed to air, complete con-
sumption of the ligand to primarily form the phosphine
oxide occurs within 21 h. Thus, 1 can be manipulated as
a solid in air for ease of use, but all solutions should be
handled in an inert atmosphere.
C9–C11–P
11–C9–Fe
112.8(3)
126.5(3)
102.6(2)
100.6(2)
110.0(2)
65.4
112.4(5)
127.4(5)
102.8(3)
99.5(3)
110.6(3)
65.5
112.4(4)
127.6(4)
100.1(3)
102.3(3)
110.2(3)
66.8
C
C11–P–C12
C11–P–C16
C12–P–C16
C8–C9–C11–P
a
Second molecule of 1 in the asymmetric unit of 1b (ORTEP not
shown in Fig. 2).
carbonyl stretch was observed at 2059.1 cmꢀ1, which is
close to the observed value for i-Pr3P (2059.2 cmꢀ1). Li-
gand 1 is a weaker electron donor than t-Bu3P
(2056.1 cmꢀ1) [33], which suggests that Pd complexes
of 1 might be expected to show lowered effectiveness
in oxidative addition reactions with less reactive aryl ha-
lides. Spectroscopic cone angle determination through
complexation with Ni(CO)4 in toluene indicated that
only one carbonyl could be exchanged with 1 after

M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
1481
2.1. Suzuki–Miyaura coupling
Table 4 shows that the Suzuki coupling yields are af-
fected by the steric and electronic properties of the aryl-
boronic acid. Use of 4-fluorophenylboronic acid gave
modest yields at room temperature and at 44 ꢀC. Reac-
tions with boronic acids bearing electron-donating
groups showed high reactivity as expected. It does ap-
pear that steric effects can lower reaction yields, as evi-
denced when 2-methylboronic acid was used as a
reactant. Low yields were seen by GC for the coupling
of 4-chlorotoluene and phenylboronic acid at room tem-
perature. A modest yield was obtained for this coupling
at 64 ꢀC, although the yield was significantly lower than
that achieved with 4-bromotoluene at room
temperature.
An in situ catalyst derived from 1 and Pd2(dba)3 gave
good activity for Suzuki–Miyaura couplings of aryl bro-
mides after conditions were optimized (Eq. (2)). Initial
studies were performed using phenylboronic acid (3a)
and either 4-bromotoluene or 4-bromoanisole as reac-
tants in THF at room temperature (Table 3). Nearly
complete reactions were normally detected in 2 h, but
most reactions were allowed to progress for 20–24 h to
maximize the conversion. Cs2CO3 proved to be a more
effective base than KF for the initially tested aryl bro-
mides, so it was used preferentially in subsequent reac-
tions. Pd loadings could be lowered to 2 mol%
without any detrimental effect. More electron deficient
aryl bromides such as 4-bromobenzaldehyde (2c) and
4-bromoacetophenone (2d) coupled well using the
optimal reaction conditions determined above. More
sterically demanding aryl bromides such as 4-bromo-3-
methylanisole (2e) led to a decrease in yield.
2.2. Heck coupling
As anticipated, Heck couplings with the 1/Pd2(dba)3
catalyst system were not as facile as the Suzuki–Miya-
ura couplings. Styrene proved to be an ideal reactant
as GC spectra of the reaction solutions showed
Pd2(dba)3
1
R
R'
+
R
R'
base
ð2Þ
X
B(OH)2
THF, r.t.
Table 4
Suzuki coupling with 1/Pd2(dba)3 under optimized conditionsa
2
3
4
Aryl halide
Boronic acid
T (ꢀC) Time (h) Yield (%)b
Table 3
Suzuki coupling of aryl bromides with 3a using 1/Pd2(dba)3
2a
rt
24
66(52)
F
B(OH)2
Aryl bromide
Mol% Pda
Base
KF
Yield (%)b
3b
4.0
61
2a
2a
3b
44
rt
20
24
(30)
95(84)
MeO
Me
Br
Me
B(OH)2
2a
2b
3c
3.9
KF
100
Br
2c
2a
3c
rt
rt
18
24
(54)
70
CH3
2a
2a
2a
2b
2b
4.1c
3.9
2.0
2.1
0.9
KF
60
100
B(OH)2
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
3d
100(83)
100(83)
54
2a
2a
3d
44
rt
20
24
(67)
(46)
MeO
B(OH)2
O
H
2.0
Cs2CO3
100(86)
3e
Br
2c
2b
2c
2d
3e
3e
3e
rt
rt
rt
rt
24
24
24
17
(85)
(90)
(78)
6c
O
2.3
Cs2CO3
(72)
Br
Me
Cl
B(OH)2
Me
2d
2f
3a
1.8
Cs2CO3
(33)
2f
3a
64
21
22d
CH3
Br
a
Pd2(dba)3 (2 mol %), 1, (2 mol %), and Cs2CO3 (3 equiv.) in THF.
Yields determined by gas chromatography unless in parenthesis,
MeO
b
2e
which signifies isolated yield. Spectroscopic data for isolated com-
pounds were consistent with literature data. Significant amounts of
unreacted aryl bromide were present in low conversion results.
a
1:1 1:Pd ratio.
Yields determined by gas chromatography unless in parenthesis,
b
c
which signifies isolated yield.
c
Pd(OAc)2 used as catalyst precursor.
4 mol% of Pd2(dba)3 and 3 equiv. KF used.
4 mol% of Pd2(dba)3 used. 1,4-dioxane used as solvent.
d

1482
M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
Table 5
Heck couplings using 1/Pd2(dba)3 at 80–85 ꢀCa
Table 6
Heck couplings using 1/Pd2(dba)3 at 95–100 ꢀCa
Baseb
Mol% Pd
P:Pd ratio
Yield (%)c
Aryl halide
Mol% Pd
5.6
P/Pd ratio
0.72
Yield (%)b
100
Cs2CO3
Cy2NMed
Cy2NMe
Cy2NMe
NBu3
2.8
3.7
4.2
3.6
4.0
3.8
1.04
0.58
1.04
1.49
0.71
0.86
0
22
55
4
Br
2g
2g
4.3
5.3
0.98
0.76
(47)
57 (33)c
40
37
CH3
Ph2NEt
a
Br
2b (1 equiv), Styrene (2 equiv), Pd2(dba)3, 1, 79–85 ꢀC, 17–23 h,
dioxane.
2h
b
1.3–1.4 equiv relative to 2b.
Yields determined by gas chromatography.
2.4 equiv Cy2NMe.
c
5.1
0.78
86 (70)
d
Me
Br
2b
2b
2b
2b
2b
5.5
2.9
2.0
5.2
1.34
0.90
0.88
0.78
74
75
minimal side product concentrations (Eq. (3), Table 5).
The application of Cs2CO3 led to no product forma-
tion, possibly due to the limited solubility of this base
in 1,4-dioxane. Trialkylamines removed this complica-
tion and allowed coupling reactions to occur. Cy2NMe
gave higher yields than NBu3 or Ph2NEt and was used
in subsequent studies. Initial catalyst optimizations
were performed between 80 and 85 ꢀC using 4-bromo-
toluene and styrene as reactants. It was found that
reaction yields were significantly lowered if the catalyst
loading was dropped below 4 mol% and if the ratio be-
tween Pd and 1 was not near 1. Table 6 shows that ele-
vation of the reaction temperature to around 100 ꢀC
and an increase in catalyst loading to 5.0–5.5 mol% sig-
nificantly increased product yields. Yield decreases sim-
ilar to those observed in lower temperature studies
were also seen with variations of catalyst loading and
1:Pd ratios. Electron deficient aryl bromides led to sim-
ilar isolated yields, indicating that catalyst efficiency
has reached a plateau. Electron rich aryl bromide 2a
gave a lower yield of the coupled product compared
to less electron rich aryl halides. As anticipated, reac-
tion attempts with aryl chloride analogs led to virtually
no product formation.
22
(54)
MeO
Br
2a
5.8
0.77
(77)
O
H
Br
Br
Cl
2c
2d
2f
O
5.7
5.1
0.73
0.78
(65)
Me
trace
Me
a
CyNMe2 (1.3 equiv.), styrene (2 equiv.), Pd2(dba)3, 1, 16.5–20.7 h,
95–99 ꢀC, dioxane.
Yields determined by gas chromatography unless in parenthesis,
b
which signifies isolated yield. Spectroscopic data for isolated com-
pounds was consistent with literature data.
c
Isolated material contaminated with ꢂ10% (mol/mol) 2h.
spectra acquired over the same time frame showed
downfield shifts for the CH2 bridge protons
(0.15 ppm) and for the protons bound to the 2 and
5 positions of the substituted Cp group (0.54 ppm).
A downfield shift of 0.164 ppm for the t-Bu group
proton signal was also seen along with a change to
an apparent triplet pattern (3JP–H = 5.91 Hz). A trip-
let pattern is indicative of virtual coupling of two
phosphorous nuclei coordinated to a metal center
with a P–M–P bond angle of 180ꢀ [35]. Non-coordi-
Pd2(dba)3
1
base
R
R
ð3Þ
+
X
Ph
dioxane
80-99 C
Ph
˚
2
5
2.3. Complexation studies
To help understand the nature of the catalytic spe-
cies in solution, THF and THF-d8 solutions contain-
ing both Pd2(dba)3 and 1 at various concentration
ratios were prepared and analyzed periodically by
1H and 31P NMR spectroscopy. In all cases, 31P
NMR spectra showed a strong signal at 59.5 ppm
that appeared quickly and remained the most intense
signal through three days of observations. 1H NMR
nated phosphine, with
a
31P NMR signal at
34.1 ppm, was only detected when the 1:Pd ratio ap-
proached 2:1. The observed stoichiometry dependence
along with the observed virtual coupling pattern are
consistent with an (1)2Pd0 complex. The fact that
the solution color remained dark purple, indicative
of residual Pd2(dba)3, until a 2:1 ratio of 1:Pd was
reached also supports this conclusion.

M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
1483
OMe
As time progressed, a new set of resonances began
to appear in NMR spectra for solutions with 0.48
t-Bu
t-Bu
P
Br
Pd Pd
Br
Fe
and 1.06 ratios of 1:Pd. For these solutions, both a
1
Fe
broad H NMR signal at 1.03 ppm and a 31P NMR
0.5
P
t-Bu
t-Bu
signal at 64.3 ppm began to appear in 30–60 min.
These signals gradually grew in intensity over the next
three days along with new weak resonances in the Cp
region. In contrast, these signals did not appear in
the 2.00 1:Pd ratio solution spectra even after three
2a
excess
6
or
t-Bu
MeO
ð4Þ
1 + Pd2(dba)3
1:Pd = 1:1
THF
t-Bu
Pd
P
Fe
Br
1
7
days. A H NMR spectrum obtained at 268 K resolved
MeO
the broad 1.03 ppm signal into two equal intensity dou-
blets at 1.073 and 0.912 ppm, while the broad ferro-
cenyl resonance resolved to a single set of sharp
peaks. Each doublet integrated for nine protons rela-
tive to the new set of ferrocene resonances, indicating
that the t-butyl groups are diastereotopic in this com-
plex. The observation of doublets for the t-butyl
groups, rather than triplets shows that there is no P–
P coupling in this complex.
Placing 5.6 equivalents of 4-bromoanisole into a 0.92
ratio of 1:Pd solution in THF-d8 initially resulted in a
31P NMR spectrum that mirrored what was seen in
experiments without aryl halide present. However, a
new signal at 53.7 ppm along with a shoulder signal at
52.8 ppm became dominant in less than 2.5 h. All other
31P NMR signals were no longer present after 8.25 h.
The resultant product was isolated and washed with
hexane to remove excess dibenzylidene acetone. Isolated
yields were poor since the palladium product is also
somewhat soluble in hexane.
Replication of this experiment using a 1.9 ratio of 1:Pd
along with 3.0 equivalents of 4-bromoanisole in THF-d8
showed only a small amount of oxidative addition prod-
uct formation after 21 h stirring at room temperature.
Since activity in full catalytic studies was significantly de-
creased under comparable 1:Pd ratios, a decreased level
of reactivity was anticipated. The slow oxidative addition
of (1)2Pd0 is consistent with the slow rates of oxidative
addition of phenyl iodide to (t-Bu3P)2Pd reported by
Jutand [39] and Hartwig [38]. Fu [21] has shown that
(t-Bu3P)2Pd0 is not a competent Suzuki coupling catalyst
unless an equivalent of is Pd2(dba)3 added to encourage
monophosphine complex formation.
3. Conclusions
Di-t-butyl(ferrocenylmethyl)phosphine has been
shown to possess several properties that will aid in
cross-coupling efficiencies in large scale reaction proce-
dures. The significant resistance towards oxidation in
air as a solid has been retained from the primary phos-
phine analog. This air-stability allows for benchtop
reaction mixture preparations in air. The strong electron
donating ability and large effective cone angle lead to
catalyst complexes which show good activity for room
temperature Suzuki–Miyaura couplings with aryl bro-
mides and moderate yields in Heck couplings at elevated
temperatures.
Stoichiometric complexation studies show that 1
forms the expected (1)2Pd0 complex when mixed with
Pd2(dba)3 independent of the 1:Pd ratio in analogy to
FuÕs observations with (t-Bu)3P [21]. With L:Pd ratios
61, (1)2Pd0 is partially converted to a new species that
is fluxional on the NMR time scale at room tempera-
ture. Key spectroscopic signatures of this new species in-
clude the inequivalence of the t-butyl groups and
doublets rather than triplet patterns for the t-butyl res-
onances. This observed inequivalence indicates that
there is an asymmetry in the coordination environment
around the palladium center in this new complex that re-
sults in the t-butyl groups being diastereotopic. The
doublet pattern for the t-butyl resonances indicates they
are coupled to a single phosphorous nuclei, thus there is
likely only one phosphine in the palladium coordination
The isolated oxidative addition product was fully
1
characterized by 31P, H, and 13C NMR spectroscopy.
The 1H NMR spectrum showed a 1:1 ratio for the
integrated area of the anisole and ligand resonances.
The t-butyl resonance of the oxidation addition prod-
1
uct was a doublet in the H NMR spectrum (3JP–H
=
13.56 Hz), which is inconsistent with the AXX0 pattern
expected for a trans-(1)2Pd(C6H4-p-OMe)Br structure.
These results indicate that the product has a stoichi-
ometry of (1)Pd(C6H4-p-OMe)Br. Based on other iso-
lated aryl halide complexes of sterically demanding
phosphines,
a
dimeric halide bridged structure,
[(1)Pd(l-Br)(C6H4-p-OMe)]2 (6), could be proposed
(Eq. (4)). These types of products have been observed
upon oxidative addition to (o-tol3P)2Pd, for example
[36]. Recently, Hartwig [37,38] reported the isolation
of 3-coordinate, T-shaped LPd(Ar)X products from
the oxidative addition of aryl halides to Pd0 complexes
of sterically demanding, electron rich alkylphosphines,
such as t-Bu3P. Given the similar steric and electronic
properties of 1 and t-Bu3P, formation of a three-coor-
dinate product (7) is also possible. We cannot rule
out either structure 6 or 7 based on our spectro-
scopic data. Unfortunately, attempts to grow crystals
of the oxidative addition product (6/7) were
unsuccessful.

1484
M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
environment [40]. Based on these observations, we have
tentatively identified this new species as a monophos-
phine palladium(0)(dba) complex, (1)Pd(dba)n (8).
Although we cannot confirm coordination of dba by
spectroscopic means, the trans-dibenzylideneacetone li-
gand provides an asymmetric coordination environment
that is typically fluxional at room temperature [41,42].
L2Pd(dba) (L = tertiary phosphine) complexes have
been previously reported [36,43–47], but LPd(dba)n
complexes of monodentate phosphines have not previ-
ously been reported to our knowledge.
Observations of monophosphine Pd0 complexes are
uncommon unless other stabilizing ligands are present
[48–50]. However, a number of authors have speculated
that dissociation of a phosphine from a L2Pd0 resting
state must occur to give a highly reactive monophos-
phine complex as the catalytically active species when
bulky phosphines are used [21,36,49,51]. Involvement
of a monophosphine-palladium(0) active species is con-
sistent with our catalytic and stoichiometric results ob-
tained with mixtures of 1/Pd2(dba)3. Optimal catalytic
activity is observed with 1:Pd ratios near 1, while the
activity significantly decreases at higher ratios. Simi-
larly, oxidative addition occurs efficiently to give a pal-
ladium aryl halide complex (6 or 7) with a 1:1 ratio of
1:Pd, but was significantly retarded at higher ratios.
The apparent observation of monophoshine complex 8
at low 1:Pd ratios suggests that 1 may be more readily
dissociated than other sterically demanding alkylphos-
phines, such as (t-Bu)3P.
Although ligand 1 appears to promote formation of
monophosphine/Pd0 species and undergoes oxidative
addition with an electron rich aryl bromide at room
temperature, it gives only moderately active catalysts
when compared catalysts derived from t-Bu3P [20,21]
or other electron rich ferrocenyl ligands [10,13–15].
The modest activity of catalysts derived from 1 may
indicate that while oxidative addition occurs efficiently,
other steps in the catalytic cycle, such as transmetalla-
tion from boron or migratory insertion of alkenes, are
inefficient. There are several approaches that can be ta-
ken to potentially enhance the ligand electronic and ste-
ric factors further in effort to improve catalytic activity.
Introduction of more sterically demanding adamantyl
substituents on the phosphorous might be expected to
improve catalytic activity, particularly toward less reac-
tive aryl chloride substrates. Functionalization of the
ferrocene cyclopentadienyl rings with electron donating
groups may also improve catalysts in this fashion as
well, leading to more active catalysts.
chemistry of this ligand may provide important insights
for future ligand design.
4. Experimental
Standard Schlenk and drybox techniques were used for
all reactions unless noted [52,53]. All solvents used were
purified and dried before use following procedures found
1
in Perrin and Armarego [54]. H and 31P NMR spectra
were obtained on either a Bruker AX360 or a Bruker
AX500 spectrometer. H and 13C NMR spectra are re-
1
ported as positive parts per million (ppm) downfield from
a tetramethylsilane (TMS) standard. 31P NMR spectra
are referenced from an external H3PO4 standard. Mass
spectra were acquired on a VG Autospec high resolution
mass spectrometer. A Varian CP-3800 gas chromato-
graph with a Varian CP-8410 autoinjector was used for
all gas chromatography studies. Elemental analyses were
performed by Atlantic Microlabs Inc., Norcross, GA.
X-ray crystallography data collection was performed
using a Siemens SMART diffractometer with a CCD
area detector, using graphite monochromated Mo Kc
radiation. The SHELXTL software, version 5, was used
for solutions and refinement [55]. Absorption correc-
tions were made with SADABS [56]. ORTEP drawings
were made with SHELXTL [55] and ORTEP-3 [57,58].
In each structure, the atoms were readily located with
high thermal motion observed for the alkyl substituents
on the phosphorus atom. Once the gross structural fea-
tures of each molecule were in place, it became apparent
that these substituents were disordered in compound 1a,
and to a lesser extent in compound 1b. The disorder ap-
pears to arise from two different rotomers of the t-butyl
groups. In compound 1a, t-butyl group C(12) was re-
solved into two rotomers, while t-butyl group C(16),
although with high thermal motion, could not be resolved
into two orientations. In compound 1b, high thermal mo-
tion was observed in the methyl substituents, most nota-
bly C(13), but the disorder could not be satisfactorily
resolved. The disordered atoms were refined in alternate
least-squares cycles, each with 50% occupancy. Com-
pound 1b was a pseudo-enantiomer of 1a. The structure
was inverted so that 1a and 1b have the same orientation.
Complexations of Ni(CO)4 with 1 were performed as
directed by Tolman [33,34]. N,N-dimethylaminomethyl-
ferrocene, di-t-butylphosphine, nickel carbonyl and
tris(dibenzylideneacetone)dipalladium were acquired
from Strem Chemicals. All bases, boronic acids, aryl ha-
lides, olefins and solvents were purchased from either
Aldrich Chemicals or Acros Organics.
While the catalytic utility of 1 may be limited, this li-
gand may prove useful in mechanistic studies of this
important class of catalytic reactions. The unusual coor-
dination chemistry at low L:Pd ratios warrants further
exploration. Additional study of the coordination chem-
istry of 1 to more fully understand the coordination
4.1. Di-t-butyl(ferrocenylmethyl)phosphine (1)
The procedure reported by Dunina et al. [32] was
adapted to form the crude product. Di-t-butylphosphine

M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
1485
(2.99 mL, 16.2 mmol) was mixed with N,N-dimethylam-
inomethylferrocene (1.409 mL, 7.117 mmol) in acetic
acid (18.9 mL). The solution was heated at 95 ꢀC for
90 h to ensure reaction completion. Acetic acid and ex-
cess di-t-butylphosphine were removed by vacuum evap-
oration. The product was extracted from the crude solid
with hexane and was recrystallized in methanol
50 mL round bottomed flasks. The appropriate aryl ha-
lide (1 equiv., usually 4 · 10ꢀ4 or 4 · 10ꢀ3 mol) was
added via syringe to a mixture of base (3 equiv.), boro-
nic acid (1.2–1.3 equiv.) and the catalyst components in
solvent (1.0 or 10.0 mL depending on aryl halide
amount). Usually a 1:1 phosphine to Pd ratio was em-
ployed with a catalyst loading of 0.9–4.1% in compari-
son with the aryl halide. The reaction mixtures were
actively stirred at room temperature unless otherwise
stated. Product yields were determined by gas chroma-
tography analysis or by isolating the product through
hexane extraction followed by column chromatography
using silica gel and an appropriate solvent determined
by thin layer chromatography.
1
(1.6567 g, 4.8121 mmol, 68% yield). H NMR (d6-ben-
zene): d 4.24 (m, 2H, subst. Cp), 4.03 (s, 5H, Cp), 3.93
3
(t, JP–H@3JH–H = 1.85 Hz, 2H, subst. Cp), 2.66 (d,
2JP–H = 1.85 Hz, 2H, –CH2–), 1.09 (d, 3JP–H
=
10.48 Hz, 18H, t-Bu). 13C NMR (d6-benzene): d 70.4
(s), 69.5 (s), 67.8 (s), 31.8 (d, JP–C = 15.70 Hz), 30.4 (s),
23.3 (d, JP–C = 25.70 Hz). 31P NMR (d6-benzene): d
33.3 (br). EI-MS: 344 m/z (M + ), 287 m/z (M + -t-Bu),
231 m/z (M + -2 t-Bu+H), 199 m/z (FcCH2), 186 m/z
(FcH), 121 m/z (CpFe). Anal. Calc. for C19H29 FeP:
C, 66.29; H, 8.49; Found: C, 66.24; H, 8.55%.
4.5. Heck coupling reactions
Solutions were prepared in nitrogen-filled vials or
flasks. Reactant concentrations are given in Tables 5
and 6. Small scale reactions were normally performed
using 2 · 10ꢀ4 mol of aryl halide and 0.5 mL of solvent.
Large scale reactions normally involved 2 · 10ꢀ3 mol of
aryl halide in 5.0 mL of solvent. Product yields were
determined by gas chromatography or by direct isola-
tion of the resultant olefin through silica gel based col-
umn chromatography.
4.2. Palladium(di-t-butyl(ferrocenylmethyl)phosphine)
(p-methoxyphenyl)bromide (6/7)
Pd2(dba)3 (24.4 mg, 26.6 lmol) and 1 (26.8 mg,
77.8 lmol) were dissolved in 2.5 mL THF. An excess
of 4-bromoanisole (34.0 lL, 0.272 mmol) was added
and the solution was stirred for two days in a sealed
flask. The yellow solution was decanted from traces of
palladium black and vacuum evaporated to yield an
oil. The product was washed with four aliquots of hex-
anes to yield a small amount of yellow paste. 1H
NMR (CDCl3): d 7.27 (d, J = 8.02 Hz, 2H, Ph), 6.57
(d, J = 8.63 Hz, 2H, Ph), 4.72 (m, 2H, subst. Cp), 4.14
(m, 2H, subst. Cp), 4.02 (s, 5H, Cp), 3.65 (s, 3H,
5. Supporting information available
Crystallographic data for the structural analysis has
been deposited with the Cambridge Crystallographic
Data Centre, CCDC Nos. 252459 and 252460 for struc-
tures 1a and 1b. These data may be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, by
emailing data_request@ccdc.cam.ac.uk, or by contacting
the Cambridge Crystallographic Data Center, 12 Union
Road, Cambridge CB2 1EZ, UK, Fax. +44 1223 336033.
2
OMe), 3.08 (d, JP–H = 8.63 Hz, 2H, –CH2–), 1.26 (d,
3JP–H = 13.56 Hz, 18H, t-Bu). 13C NMR (CD2Cl2): d
157.1 (s), 137.4 (s), 134.0 (s), 113.5 (s), 83.6 (s), 72.0
(s), 69.8 (s), 68.0 (s), 55.6 (s), 38.0 (s), 31.2 (s), 31.0 (s).
31P NMR (CDCl3): d 53.4 (br).
4.3. Palladium(di-t-butyl(ferrocenylmethyl)phosphine)
Ln complex (8)
Acknowledgements
Pd2(dba)3 (12.3 mg, 13.4 lmol) and 1 (0.48–2.00
equiv.) were dissolved in 1.00 mL d8-THF. Portions of
each solution were placed in NMR tubes, which were
sealed with septa and were examined by NMR spectros-
The National Science Foundation (CHE-0124255)
provided financial support for this work. Valuable assis-
tance was provided by Dr. Ken Belmore with the NMR
studies and Marc Klingshirn with the DSC studies.
1
copy over several days. H NMR (d8-THF, 268 K): d
4.22 (br, 2H, subst. Cp), 3.93 (br, 5H, Cp), 3.80 (br,
3
2H, subst. Cp), 3.23 (br, 2H, –CH2–), 1.07 (d, JP–H
=
References
3
11.23 Hz, 9H, t-Bu), 0.91 (d, JP–H = 12.82 Hz, 9H,
t-Bu). 31P NMR (d8-THF): d 64.3 (br).
[1] T. Hayashi, M. Konishi, Y. Kobori, M. Kumada, T. Higuchi, K.
Hirotsu, J. Am. Chem. Soc. 106 (1984) 158–163.
[2] M. Ogasawara, K. Yoshida, T. Hayashi, Organometallics 19
(2000) 1567–1571.
4.4. Suzuki–Miyuara coupling reactions
[3] T. Oh-e, N. Miyaura, A. Suzuki, Synlett (1990) 221–223.
[4] T. Oh-e, N. Miyaura, A. Suzuki, J. Org. Chem. 58 (1993) 2201–
2208.
Solutions were prepared in nitrogen using either one
dram vials capped with PTFE lined silicone septa or

1486
M.D. Sliger et al. / Journal of Organometallic Chemistry 690 (2005) 1478–1486
[5] V. Percec, J.-Y. Bae, D.H. Hill, J. Org. Chem. 60 (1995) 1060–
1065.
[33] C.A. Tolman, Chem. Rev. 77 (1977) 313–348.
[34] C.A. Tolman, J. Am. Chem. Soc. 92 (1970) 2956–2965.
[35] J.M. Jenkins, B.L. Shaw, Proc. Chem. Soc. (1963) 279.
[36] F. Paul, J. Patt, J.F. Hartwig, Organometallics 14 (1995) 3030–
3039.
[6] M. Ueda, A. Saitoh, S. Oh-Tani, N. Miyaura, Tetrahedron 54
(1998) 13079–13086.
[7] S. Saito, S. Oh-tani, N. Miyaura, J. Org. Chem. 62 (1997) 8024–
8030.
[37] J.P. Stambuli, M. Buhl, J.F. Hartwig, J. Am. Chem. Soc. 124
¨
(2002) 9346–9347.
[8] A.L. Boyes, I.R. Butler, S.C. Quayle, Tetrahedron Lett. 39 (1998)
7763–7766.
[38] J.P. Stambuli, C.D. Incarvito, M. Buhl, J.F. Hartwig, J. Am.
¨
Chem. Soc. 126 (2004) 1184–1194.
[9] A.M. Echavarren, J.K. Stille, J. Am. Chem. Soc. 110 (1988) 1557–
1565.
[39] E. Galardon, S. Ramdeehul, J.M. Brown, A. Cowley, K.K. Hii,
A. Jutand, Angew. Chem., Int. Ed. 41 (2002) 1760–1763.
[40] Coupling with a second phosphorous center also would not be
observed if two phosphines were coordinated to the palladium
center in a cis fashion, but this seems unlikely given the large steric
demand of 1.
[10] T.J. Colacot, Platinum Metals Rev. 45 (2001) 22–30.
[11] T.J. Colacot, H.A. Shea, Org. Lett. 6 (2004) 3731–3734.
[12] N. Kataoka, Q. Shelby, J.P. Stambuli, J.F. Hartwig, J. Org.
Chem. 67 (2002) 5553–5566.
[13] T.E. Pickett, F.X. Roca, C.J. Richards, J. Org. Chem. 68 (2003)
2592–2899.
[41] C. Amatore, A. Fuxa, A. Jutand, Chem. Eur. J. 6 (2000) 1474–
1482.
[14] K.H. Shaughnessy, P. Kim, J.F. Hartwig, J. Am. Chem. Soc. 121
(1999) 2123–2132.
´
´
´
[42] F.A. Jalon, B.R. Manzano, F. Gomez-de la Torre, A.M. Lopez-
Agenjo, A.M. Rogr´ıguez, W. Weissensteiner, T. Sturm, J. Mahia,
M. Maestro, J. Chem. Soc., Dalton Trans. (2001) 2417–2424.
[15] J.P. Stambuli, S.R. Stauffer, K.H. Shaughnessy, J.F. Hartwig, J.
Am. Chem. Soc. 123 (2001) 2677–2678.
¨
[43] W.A. Herrmann, W.R. Thiel, C. Broßmer, K. Ofele, T. Prierme-
[16] M. Shen, G. Li, B.Z. Lu, A. Hossain, F. Roschangar, V. Farina,
C.H. Sennayake, Org. Lett. 6 (2004) 4129–4132.
[17] T.J. Colacot, Chem. Rev. 103 (2003) 3101–3118.
[18] A. Togni, in: A. Togni, R.L. Halterman (Eds.), Metallocenes:
Synthesis, Reactivity, Applications, vol. 2, Wiley-VCH, New
York, 1998; pp. 685–721.
ier, W. Scherer, J. Organomet. Chem. 461 (1993) 51–60.
[44] A.G. Orpen, P.G. Pringle, M.B. Smith, K. Worboys, J. Organo-
met. Chem. 550 (1998) 255–266.
[45] C. Amatore, A. Jutand, Coord. Chem. Rev. 178–180 (1998) 511–
528.
[19] A.F. Littke, G.C. Fu, Angew. Chem. Int. Ed. 38 (1999) 2411–
2413.
[46] A.D. Burrows, N. Choi, M. McPartlin, D.M.P. Mingos, S.V.
Tarlton, R. Vilar, J. Organomet. Chem. 573 (1999) 313–322.
[47] G. Adjabeng, T. Brenstrum, C.S. Frampton, A.J. Robertson, J.
Hillhouse, J. McNulty, A. Capretta, J. Org. Chem. 69 (2004)
5082–5086.
[20] A.F. Littke, G.C. Fu, J. Org. Chem. 64 (1999) 10–11.
[21] A.F. Littke, C. Dai, G.C. Fu, J. Am. Chem. Soc. 122 (2000) 4020–
4028.
[22] J.F. Hartwig, M. Kawatsura, S.I. Hauck, K.H. Shaughnessy,
L.M. Alcazar-Roman, J. Org. Chem. 64 (1999) 5575–5580.
[23] J.P. Wolfe, S.L. Buchwald, Angew. Chem. Int. Ed. 38 (1999)
2413–2416.
[48] J. Krause, G. Cestaric, K.-J. Haack, K. Seevogel, W. Storm, K.-
R. Po¨rschke, J. Am. Chem. Soc. 121 (1999) 9807–9823.
[49] X. Bei, H.W. Turner, W.H. Weinberg, A.S. Guram, J.L. Peterson,
J. Org. Chem. 64 (1999) 6797–6803.
[24] J.P. Wolfe, R.A. Singer, B.H. Yang, S.L. Buchwald, J. Am.
Chem. Soc. 121 (1999) 9550–9561.
[50] M.G. Andreu, A. Zapf, M. Beller, Chem. Commun. (2000) 2475–
2476.
[25] J.P. Wolfe, H. Tomori, J.P. Sadighi, J. Yin, S.L. Buchwald, J.
Org. Chem. 65 (2000) 1158–1174.
[51] A. Aranyos, D.W. Old, A. Kiyomori, J.P. Wolfe, J.P. Sadighi,
S.L. Buchwald, J. Am. Chem. Soc. 121 (1999) 4369–4378.
[52] D.F. Shriver, M.A. Drezdon, The Manipulation of Air-Sensitive
Compounds, second ed., John Wiley and Sons, New York, 1986.
[53] A.L. Wayda, M.Y. Darensborg (Eds.), Experimental Organome-
tallic Chemistry; ACS Symposium Series #357; American Chem-
ical Society: Washington D.C., 1987.
[26] A. Zapf, A. Ehrentraut, M. Beller, Angew. Chem. Int. Ed. 39
(2000) 4153–4155.
[27] G.P.F. van Strijdonck, M.D.K. Boele, P.C.J. Kamer, J.G. de
Vries, P.W.N.M. van Leeuwen, Eur. J. Inorg. Chem. (1999) 1073–
1076.
[28] M.R. Netherton, G.C. Fu, Org. Lett. 3 (2001) 4295–4298.
[29] N.J. Goodwin, W. Henderson, B.K. Nicholson, J. Fawcett, D.R.
Russell, J. Chem. Soc., Dalton Trans. 1 (1999) 1785–1793.
[30] W. Henderson, S.R. Alley, J. Organomet. Chem. 656 (2002) 120–
128.
[54] W.L.F. Armarego, D.D. Perrin, Purification of Laboratory
Chemicals, fourth ed., Butterworth-Heinemann, 1996.
[55] G.M. Sheldrick, SHELXTL, Version 5.05, Siemens Analytical X-
Ray Instruments, Inc., 1996.
[56] G.M. Sheldrick, Program for Semiempirical Absorption Correla-
tion of Area Detector Data, University of Go¨ttingen, Germany,
1996.
[31] Z.-Y. Tang, Y. Lu, Q.-S. Hu, Org. Lett. 5 (2003) 297–300.
[32] V.V. Dunina, O.N. Gorunova, M.V. Livantsov, Y.K. Grishin,
L.G. KuzÕmina, N.A. Kataeva, A.V. Churakov, Tetrahedron:
Asymm. 11 (2000) 3967–3984.
[57] M. Burkitt, Meth. Enzymol. 234 (1994) 66.
[58] J. Farrugia, Appl. Crystallogr. 30 (1997) 565–566.