Design, testing and kinetic analysis of bulky monodentate phosphorus ligands in the mizoroki-heck reaction

Article abstract of DOI:10.1002/ejic.201101271

A series of new monodentate phosphane ligands 2 have been evaluated in the Mizoroki-Heck arylation reaction of iodobenzene and styrene and compared with our previously reported ligands, 1, 3 and 4. The concept of rational ligand design is discussed, and we describe how the performance of this new ligand family could be predicted. Employing our best ligand, 3,3'-di-tert-butyl-5,5'- dimethoxybiphenyl-2,2'-diyl diisopropylphosphoramidite (3b), we explored the scope of the reaction with regards to solvent and the substrate. We also investigated the electronic dependence of the reaction by analysing the relationship between the rate and Hammett constant. Sufficient steric bulk is required to enforce the catalytic reaction to proceed through the mono-coordinated palladium species, thereby increasing its reactivity. The electronic properties determine the concentration of the active species from the monomer dimer equilibrium and their intrinsic reactivity. The cyclic phosphoramidite 3b provides an optimum in both properties within the systems studied, resulting in a rate limiting migratory insertion step. Copyright

Full text of DOI:10.1002/ejic.201101271

FULL PAPER
DOI: 10.1002/ejic.201101271
Design, Testing and Kinetic Analysis of Bulky Monodentate Phosphorus
Ligands in the Mizoroki–Heck Reaction
Deborah L. Dodds,[a] Maarten D. K. Boele,[b] Gino P. F. van Strijdonck,[b]
Johannes G. de Vries,[c] Piet W. N. M. van Leeuwen,*[b] and Paul C. J. Kamer*[a]
Keywords: Homogeneous catalysis / Palladium / C-C coupling / Phosphane ligands / Ligand effects / Reaction mechanisms /
Kinetics
A series of new monodentate phosphane ligands 2 have been
evaluated in the Mizoroki–Heck arylation reaction of iodo-
by analysing the relationship between the rate and Hammett
constant. Sufficient steric bulk is required to enforce the cata-
benzene and styrene and compared with our previously re- lytic reaction to proceed through the mono-coordinated pal-
ported ligands, 1, 3 and 4. The concept of rational ligand de-
sign is discussed, and we describe how the performance of
this new ligand family could be predicted. Employing our
best ligand, 3,3Ј-di-tert-butyl-5,5Ј-dimethoxybiphenyl-2,2Ј-
diyl diisopropylphosphoramidite (3b), we explored the scope
of the reaction with regards to solvent and the substrate. We
also investigated the electronic dependence of the reaction
ladium species, thereby increasing its reactivity. The elec-
tronic properties determine the concentration of the active
species from the monomer dimer equilibrium and their in-
trinsic reactivity. The cyclic phosphoramidite 3b provides an
optimum in both properties within the systems studied, re-
sulting in a rate limiting migratory insertion step.
Introduction
Rational ligand design is usually an iterative process con-
sisting of design, modelling, synthesis, and testing. Many of
these steps will be repeated until eventually (usually with
some serendipity) an optimum is reached. Improvements in
accuracy and accessibility of computer modelling has al-
lowed a closer link between the design and outcome stage,
but lengthy synthetic routes are often the rate-limiting step
of this approach.[7]
In silico catalyst screening and optimisation is an ap-
proach which has received attention in recent years, with
the development of the ligand knowledge base by Fey[8]
(which has recently been expanded to include monodentate
P-donor ligands)[9] and the idea of virtual catalyst libraries
by Rothenberg.[10] Although these computational develop-
ments contribute to enhanced rational ligand design, ligand
synthesis and catalyst testing remain indispensable steps in
the development of new and improved catalysts. Catalyst
development is still hampered by lack of insight in the tran-
sition state of the selectivity-determining step. Furthermore,
in general a catalytic transformation consists of several ele-
mentary steps that will be influenced in different ways by
ligand modifications. Therefore, rational design of ligands,
in particular for enantioselective transformations, is often
beyond reach and “trial and error” methods are still em-
ployed in the development of efficient high-performance
catalysts.
The Mizoroki–Heck reaction, the palladium catalysed
coupling of aryl halides and alkenes was first described by
Mizoroki and Heck independently in the 1970’s and is now
a well established method of C–C bond formation; see
Equation (1).[1,2] Last year (2010) Heck was awarded the
Nobel Prize alongside Suzuki and Negishi for their contri-
bution to palladium-catalysed cross coupling chemistry.[3]
During recent years, enormous progress has been reported
within this area by the development of new procedures and
novel catalysts.[4] Herrmann and Beller set the pace for a
new era of the Mizoroki–Heck (M–H) reaction with the
introduction of high-yielding catalysts based on 1b, al-
though initially their metallated complexes were used.[5] For
a comprehensive overview of the M–H reaction see the
chapter by Jutand.[6]
(1)
[a] Eastchem, School of Chemistry, North Haugh, University of
St Andrews,
St Andrews, Fife, UK
Fax: +44-1334-463808
E-mail: pcjk@st-andrews.ac.uk
[b] University of Amsterdam, Institute of Molecular Chemistry,
Nieuwe Achtergracht 166, 1018 WV Amsterdam,
The Netherlands
E-mail: pvanleeuwen@iciq.es
Here we report a series of monodentate phosphorus li-
gands which have been tested for their catalytic perform-
ance in the M–H reaction. Trends are discussed between
[c] DSM Pharmaceutical Products-Innovative Synthesis &
Catalysis,
P. O. Box 18, 6160 MD Geleen, The Netherlands
1660
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 1660–1671

Bulky Monodentate Phosphorus Ligands
ligand properties and their performance in the M–H reac- abled a virtual ligand library to be determined, from which,
tion. We optimised the system using the best ligand, which promising ligands with regard to TON and TOF could be
in turn provided some insight into the mechanism.
selected.[12]
Phosphoramidites form a unique class of ligands that
have proven their usefulness in several catalytic reac-
tions.[13,14] Their electronic properties, expressed in χ-val-
ues,[15] are in between those of phosphanes and phosphites,
while the substituents on nitrogen enable steering of the
steric bulk in proximity of the phosphorus centre.[16] The
electronic property of phosphoramidites was reflected in
the catalytic results, which were in between those of phos-
phanes and phosphites. Bulky monodentate phosphoramid-
ite ligands also resulted in extremely fast catalysts for (enan-
tioselective) cross coupling reactions.[11,17,18]
Results and Discussion
Based on our previous findings,[11] where we found that
the cyclic phosphoramidite 3b was the most efficient ligand
in our study, we have expanded the family to include a new
series of heterocyclic phosphane ligands 2. The library
(Scheme 1) now comprises phosphorus ligands with struc-
tural variations providing steric and electronic diversity,
which fall into one of the following categories:
· phosphanes (PR3), 1
We selected the phosphacyclic ligands 2 as they can be
easily functionalised at the meta (to phosphorus) position
and also at the heteroatom which will allow the introduc-
tion of steric bulk and adjustment of the angle at the phos-
phorus atom. In addition, the constrained cyclic system
places the phosphorus atom in a unique electronic environ-
ment.[19] Related phosphacycles have found application in
the telomerisation of 1,3-butadiene with methanol.[20] We
anticipated that, based on the steric and electronic proper-
ties of these monodentate ligands, we could predict their
catalytic outcome in the M–H reaction, which we expected
to be between that of phosphanes and phosphoramidites.
· phosphacyclic phosphanes [P(-(C2XC2-)Ph], 2
· phosphites [P(OR)3], 3
· phosphoramidites [P(OR)2NR], 4
Ligand Synthesis
Ligands 2a–c were prepared according to literature meth-
ods via dilithiation of the respective phenyl ethers and sulf-
ides in the presence of TMEDA (tetramethylethylenedi-
amine). Treatment with PhPCl2 followed by an aqueous
workup, purification over silica and recrystallisation results
in the desired monodentate phosphacycles.[21] These ligands
were found to degrade over time, and so for long term stor-
age they were converted into their BH3 adducts. This pro-
cedure was also used as an aid to purify the sulfur analogue
2a, which was hard to purify as the free phosphane. Depro-
tection was achieved in a straightforward manner by re-
fluxing in ethanol or methanol and the resulting borates
were removed in vacuo.[22]
Dihydrophenophosphazines are well known products of
the reaction between dichlorophosphanes and secondary
amines.[23] Current synthetic methods are, however, gen-
erally low yielding (Scheme 2).[24] 2d was obtained as a crys-
talline material in 30% yield from phenyldichlorophos-
phane and diphenylamine via an intramolecular cyclisation
at 220 °C. This route was not amenable to functionalised
phosphazines, and therefore 2e was synthesised by an alter-
native route.
An N-methyl-protected ortho-substituted dibromodi-
phenylamine, prepared from bis[4-(tert-butyl)phenyl]-
amine,[25,26] was dilithiated using tBuLi, which, following
Scheme 1. Monodentate phosphanes 1, phosphacyclic phosphanes
2, phosphites 3 and phosphoramidites 4.
This series of ligands 1, 3, and 4 was also used as a train- treatment with phenyldichlorophosphane, gave the desired
ing set by Rothenberg for in silico ligand design, which en- bulky phosphacycle 2e in 40% yield (Scheme 3).[27]
Eur. J. Inorg. Chem. 2012, 1660–1671
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1661

P. W. N. M. van Leeuwen, P. C. J. Kamer et al.
FULL PAPER
apparent in the other series too, notably in the phos-
phoramidite series, where the conversion is 18%, 100% and
6%, respectively, for ligands 3a, 3b and 3c (entries 10–12).
Table 1. Mizoroki–Heck reaction of iodobenzene with styrene em-
ploying ligands 1–4 at 80 °C.[a]
Entry
Ligand
Conv. [%][b] 5 [%] 6 [%]
7 [%][c]
1
2
3
4
5
6
7
8
none
1a
1b
1c
2a
2b
2c
2d
2e
3a
3b
3c
7
6
12
1
90
90
90
89
83
88
89
90
X
90
90
88
90
90
85
9
9
9
10
13
9
1
1
1
1
0
Ͻ 1
Ͻ 1
Ͻ 1
X
1
1
1
2
Scheme 2. Synthesis of 2d. (a) PhPCl2, 220 °C, 2 h; (i) PhPCl2,
– HCl; (ii) intramolecular rearrangement (dearomatisation); (iii)
rearomatisation; (iv) intramolecular cyclisation, – HCl.
26
35
36
1
18
100
6
8
48
6
9
10
X
9
9
11
9
9
10
11
12
13
14
15
4a
4b
4c
1
1
1
9
14
[a] Conditions: see Exp. Section. [b] Based on iodobenzene, mea-
sured after 45 min. [c] Less than 1% biphenyl was formed.
It is interesting to note that the new phosphacyclic li-
gands 2 did fit the data as we predicted; their electronic
properties placed them between phosphanes and phos-
phoramidites, which is reflected in their catalytic perform-
ance (entries 5–9). The best phosphacyclic ligand, 2d, out-
performed phosphane 1b, but did not compete with the
phosphoramidite 3b or phosphite 4b.
Scheme 3. Synthesis of 2e. (a) Br2, AcOH, 5 °C; (b) NaH, DME,
Δ; (c) MeI; (d) tBuLi, pentane/diethyl ether, 0 °C; (e) PhPCl2, 0 °C.
Changing the heteroatom in the heterocycle has an im-
pact on its catalytic performance. If the sulfur and oxygen
analogues are compared, 2a and 2b, it is immediately appar-
ent that there is a large difference in reactivity as conver-
sions of 2% and 26% were obtained (entries 5 and 6). The
nitrogen and oxygen ligands, 2c and 2d behave in a similar
manner with yields of 35% and 36%, which suggests that
steric influences play a major role. The N-protected ligand
(2e) then showed a reduced conversion of 1%. By introduc-
ing methyl groups at the meta (to P) position, the conver-
sion was greatly improved and it performed better than the
acyclic analogue, triphenylphosphane. The ligands lacking
a substituent, 2a and 2b, performed poorly (entries 5 and
6). The steric influence was increased further by incorporat-
ing a tBu group (2e), and in a similar manner to the other
systems tested it was found that the conversion decreased
again (entry 9). It should be noted, however, that if the reac-
tion time (45 min) is lengthened to run overnight, 2e also
Catalyst Screening
Each ligand was tested in the M–H reaction for the syn-
thesis of E-stilbene (5) from styrene and iodobenzene as a
model reaction; see Equation (2). Side products which are
detectable by GC are Z-stilbene (7), 1,2-diphenylethylene
(6) and biphenyl (8). The same ratio is usually observed of
90:10 E-stilbene/1,2-diphenylethylene, with ca. 1% biphenyl
and a trace of Z-stilbene.
(2)
The results of the ligand screening are given in Table 1. gives full conversion (in line with the other ligands in the
A clear trend is observable; as the steric bulk increases series). This suggests that there may be an induction period,
(from left to right across the series, see Scheme 1), the con- possibly due to the increased steric bulk of this ligand; this
version also increases, with maxima present at the optimum will be investigated further. Another point to consider is
steric bulk. Comparing the phosphane series 1 (entries 2– that the heteroatom may interfere with catalysis, as they
4), and the optimum conversion is achieved with ligand 1b. have the potential to compete with phosphorus for coordi-
This is an important point as it is thought to be related to nation to the palladium centre. This could be another
the palladium species that forms during the catalytic cycle. reason why much lower reactivity was observed for the N-
Bulkier ligands (with larger Tolman cone angles)[15] tend to protected ligand 2e than for the unprotected 2d.
form a mono ligated species, whereas less sterically hindered
The best performance ligand was chosen for further stud-
ligands can form more highly ligated species. This trend is ies, in this case phosphoramidite 3b. To ensure that the
1662
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 1660–1671

Bulky Monodentate Phosphorus Ligands
same complex was present at the start of each reaction it Table 2. Solvent effect on the M–H reaction of iodobenzene with
styrene catalysed by 9 at 80 °C.[a]
was used as a pre prepared complex.
[b]
Entry Solvent
TOFini
Yield [%] [c] 5 [%]
7 [%]
1
2
3
4
5
6
7
8
9
CH3CN
NMP
DMF
315
605
400
340
760
1900
685
85
96
89
90
22
61
58
39
11
90
92
92
92
92
91
92
93
92
Ͻ 0.5
Ͻ 0.5
Ͻ 0.5
Ͻ 0.5
Ͻ 0.5
Ͻ 0.5
Ͻ 0.5
Ͻ 0.5
Ͻ 0.5
Complex Synthesis and Characterisation
Reaction of Pd(dba)2 with two equivalents of ligand 3b
in the presence of an excess of bromobenzene in toluene
resulted in the formation of complex 9 in good yield; see
Equation (3). The formation of analogous palladium di-
mers containing monodentate phosphanes have been con-
firmed by X-ray crystallography.[28]
DMA
EtOAc
toluene[d]
toluene/NMP[e]
iPrOH
480
CH3CN/H2O[f]
n.d.[g]
[a] Conditions: iodobenzene/styrene/triethylamine
=
1:1.1:1.1,
0.25 mol-% of 9, see the Exp. Section. [b] Initial turn-over fre-
quency, determined after 5 min [molmol–1 h–1]. [c] Based on all
products, determined after 3 h by GLC. [d] 0.125 mol-% of 9 used.
[e] 1:1 ratio (v/v). [f] 10:1 ratio (v/v). [g] Not determined.
(3)
The elemental composition of 9, together with the inte-
gration in the H NMR spectrum, suggests that it consists
The solvent used for the Mizoroki–Heck coupling be-
tween iodobenzene and styrene has an important influence
on the outcome of the reaction. Good conversions can be
obtained when polar, non-protic solvents are employed, as
illustrated for acetonitrile, NMP, DMF and DMA (N,N-
dimethylacetamide), see entries 1–4. The initial reaction
rates within this series are of the same order of magnitude.
The application of ethyl acetate results in low yields. Appar-
ently, the coordinating ability of the solvent plays an impor-
tant role in the catalyst stability and therefore the total
turn-over number. When toluene, a non polar solvent, is
used (entry 6), a huge increase in reaction rate is observed.
This rate improvement is accompanied by a moderate yield
and the reaction was performed at lower catalyst concentra-
tion for accurate measurement of the turn-over frequency.
In addition to the shift of the equilibrium to monomeric
species at lower catalyst concentration, this indicates that
the less coordinating ability the solvent has, the faster the
resulting catalyst system is. Possibly, competition between
the solvent and substrates occurs when the solvent can co-
ordinate, thereby rendering the catalyst less active, but more
stable. It is not clear at which stage of the catalytic cycle
this effect is most important. Attempts to combine high ac-
tivity and stability by employing mixtures of toluene and
NMP were not successful (entry 7). The use of polar protic
solvents also resulted in low yields, especially in the pres-
ence of water (10% v/v in acetonitrile), which had a large
detrimental effect on the stability of the catalyst. Appar-
ently some of the intermediate complexes are not stable
towards protic species. The observed selectivities in all cata-
lytic runs are nearly the same. This is in contrast to other
systems, containing two coordinated phosphorus ligands or
cyclopalladated catalysts, where increasing polarity of the
solvent was found to have a negative effect on the selectivity
towards the (E)-alkene product.[32–35]
1
of a palladium centre surrounded by one aryl, one bromo
and one phosphoramidite ligand. Osmometric molecular
mass determination in dichloromethane indicated that com-
pound 9 mainly exists as a dimeric species in solution (Mw
= 1400Ϯ200). Recent developments in solid state EXAFS
data-analysis methods, which make use of the so-called dif-
ference file technique,[29] allow a reliable separation of the
different contributions resulting in a proper analysis. EX-
AFS data of excellent quality was obtained for compound
9 and suggests a dimeric structure.
The 31P NMR spectrum of 9 showed only one broad
resonance at δ = 118.3 (CDCl3) at room temperature. This
suggests the existence of a fast equilibrium, presumably be-
tween the trans and cis isomers of complex 9. Similar equi-
libria have been observed for other dimeric palladium com-
plexes containing monodentate ligands.[30,31] Upon cooling
to 240 K decoalescence occurred, resulting in two signals at
δ = 119.8 and 117.5 in an approximate 4:3 ratio. Based on
steric arguments and in analogy with the literature,[30,31] we
propose that the major species has a trans geometry with
respect to the two bulky phosphoramidite ligands. In aceto-
nitrile, only one sharp singlet was observed at δ = 122.3 in
a temperature range of 255–335 K. Apparently the cis/trans
isomerisation is faster in solutions containing polar sol-
vents. Possibly, the ability of acetonitrile to coordinate to
palladium plays a role herein. Complex 9 shows good sta-
bility in the solid state, but in solutions of non-polar sol-
vents the compound decomposes slowly to give palladium
metal and several unidentified phosphorus-containing
products. In polar, coordinating solvents like DMF (N,N-
dimethylformamide) or NMP (N-methylpyrrolidinone), this
decomposition process is slower.
We also investigated the scope of the aromatic and ole-
finic substrates using the bulky phosphoramidite complex 9
Scope of the Catalyst
To explore the potential of the new catalyst system 9 in as a catalyst. The results hereof are summarised in Table 3.
the M–H reaction, we varied several of the reaction param- The phosphoramidite-based catalyst results in good con-
eters. The influence of the solvent on the catalyst perform- versions when aryl iodides are applied, especially in combi-
ance is represented in Table 2.
nation with electron-poor alkenes. Thus, the reaction be-
Eur. J. Inorg. Chem. 2012, 1660–1671
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1663

P. W. N. M. van Leeuwen, P. C. J. Kamer et al.
FULL PAPER
Table 3. M–H reaction with different substrates using 9 as the cata- “leakage cycle” is observed.[40] It is important to note that
lyst at 80 °C.[a]
these results leave open the possibility of a base-assisted
elimination mechanism. The stability of the catalyst is sig-
nificantly lower when aryl bromides are applied, but the
addition of an extra equivalent of ligand to the palladium
catalyst, thereby increasing its stability, results in good
yields when the more reactive n-butyl acrylate is used as the
alkene (entries 9, 10). At lower ligand concentrations the
use of less reactive aryl bromides might lead to the forma-
tion of palladium nanoparticles and eventually Pd metal
deposition. As expected when the oxidative addition is rela-
tively slow, the use of electron-rich aryl bromides such as
para-bromoanisole leads to only moderate conversion (en-
try 11). Electronically activated 1-bromonaphthalene gives
smooth conversion, while aryl chlorides are not reactive.
Probably the activation of the C–Cl bond does not take
place under the mild conditions employed here using a rela-
tively electron-deficient palladium catalyst like 9. Beller re-
ported the use of even more strongly electron-withdrawing
phosphane ligands in the M–H reaction of aryl chlorides,
but high temperatures (160 °C) and large amounts of ligand
were required to achieve good yields.[41] Electron-rich alkyl-
phosphanes have been shown to be better candidates for
the activation of aryl chlorides under mild conditions.[42–45]
Entry Aryl halide
Alkene
Solvent
Yield
[%][b,c]
1
PhI
PhI
2-MeC6H4I
PhOTf
PhI
n-butyl acrylate
n-butyl acrylate
n-butyl acrylate
n-butyl acrylate
styrene
styrene
styrene
styrene
styrene
n-butyl acrylate
NMP
NMP
NMP
NMP
CH3CN
CH3CN
NMP
NMP
NMP
NMP
NMP
NMP
Ͼ 99
95
Ͼ 99
Ͻ 1
85
83
0
0
5
89
30
2[d]
3[e]
4
5
6[f]
7[g]
8[h]
9[e]
10[e]
11[e]
12[e]
PhI
PhI
PhI
PhBr
PhBr
4-MeOC6H4Br n-butyl acrylate
1-naphthyl
bromide
n-butyl acrylate
Ͼ 99
13[e,i] PhCl
n-butyl acrylate
NMP
0
[a] Conditions as described for Table 2. [b] Based on conversion of
aryl halide, determined after 16 h. [c] Selectivity for stilbene in all
cases approx. as in Table 2. [d] Reaction run at room temperature.
[e] Reaction run with one equivalent of ligand (3b) added. [f]
20 equiv. (to Pd) of +N(nBu)4Br– added. [g] Pyridine applied as the
base. [h] KOAc applied as the base. [i] Reaction run at 110 °C using
N,N-diisopropylethylamine as the base.
tween iodobenzene and n-butylacrylate resulted in quantita-
tive conversion to the expected cinnamyl ester (entry 1). In
accordance with general observations,[36] the selectivity
towards the (E)-alkene product is very high (Ͼ 99.5%)
Electronic Dependence
In accordance with present mechanistic insights[4] and
when acrylates are employed. At 80 °C, this reaction was our studies (vide supra and ref.[11]) oxidative addition is not
completed within five minutes, which corresponds to an rate limiting for reactive substrates. This indicates that al-
average turn-over frequency exceeding 4800 molmol–1 h–1. kene addition or migratory insertion might determine the
When the reaction was run at room temperature, longer re- overall rate of the catalytic reaction. To distinguish between
action times were required, resulting in high yields after 16 a rate-limiting step involving the alkene coordination or the
hours (entry 2). Increasing the steric bulk of the aryl iodide subsequent insertion, we tested the influence of electroni-
by introducing an ortho-substituent did not affect the con- cally modified aryl halides and styrene derivatives on the
version (entry 3).
outcome of the catalytic reaction. Therefore, we determined
Aryl triflates are not compatible with the catalyst system, the initial reaction rate of para-substituted styrenes 10 (p-
see entry 4. The reason for this is unclear, but probably RЈC6H4CHCH2; RЈ = OMe, Me, H, CF3) and para-substi-
the lack of stabilizing halides results in catalyst deactivation tuted iodobenzenes 11 (p-RC6H4I; R = Cl, H, Me, OMe)
under the applied conditions. Attempts to isolate cationic using phosphoramidite complex 9 as the catalyst in NMP
complexes by reaction of 9 with halide-abstracting reagents at 50 °C.
(e.g. AgOTf, OTf = trifluoromethanesulfonate) in acetoni-
Application of styrene derivatives with increasing elec-
trile resulted in the rapid formation of palladium metal. As tron density on the olefinic double bond results in a slower
illustrated in Table 3, the application of acetonitrile as the reaction. Thus, the observed order of reactivity is p-OMe-
solvent results in slightly lower yields. The addition of tetra- 10 Ͻ p-Me-10 Ͻ 10 ϽϽ p-CF3-10. The Hammett plot (Fig-
alkylammonium salts does not have a beneficial effect on ure 1) shows an approximately linear correlation between
the conversion or the selectivity (entry 6). This is in contrast the Hammett σ-parameter and the relative reaction rates,
to other studies, where it was reported that the presence with r = 0.9. As commonly found in M–H reactions, the
of halide salts can accelerate the reaction, caused by the use of more electron-rich alkenes results in a decrease in the
formation of anionic palladium species.[37–39] Changing the regioselectivity towards the (E)-alkene product in favour of
nature of the base by using pyridine or inorganic salts, e.g. the 1,1-substituted product, but the effect is small within
KOAc, does not result in any conversion (entries 7 and 8). the series studied [(E)/(1,1) = 88:12 for p-MeO-10].[46–48]
Both the strength of the base and its solubility (in the case
In addition, the use of iodobenzenes containing electron-
of salts) can play a role herein. Also the coordinating prop- withdrawing substituents also results in lower reaction rates,
erties of pyridine and precipitation of KI can have a detri- exhibiting the order p-Cl-11 Ͻ 11 Ͻ p-Me-11 Ͻ p-OMe-
mental influence on catalyst activity. No evidence for the 11. It is obvious that electron rich aryl iodides give faster
operation of a mechanism in which the catalyst itself acts as conversion although no linear Hammett plot is obtained in
the base, resulting in sub-stoichiometric turnover, so-called this case (Figure 2).
1664
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 1660–1671

Bulky Monodentate Phosphorus Ligands
dium complexes.[50] They found that styrenes bearing elec-
tron-withdrawing substituents accelerated the migratory in-
sertion, caused by destabilisation of the ground-state of the
preceding (η2-styrene)palladium complex. Similar results
were found in theoretical studies performed by Svensson et
al.[53] In earlier studies, Cabri reported the higher reactivity
of para-substituted aryl triflates containing electron-donat-
ing groups in the M–H reaction with vinyl butyl ether using
Pd(OAc)2/dppp as the catalyst.[54] Recently, Hartwig re-
ported kinetic studies which showed that electron rich li-
gand free anionic palladium complexes gave faster alkene
insertion, which he attributed to steric factors.[55] Our re-
sults indicate that increased nucleophilicity of the aryl li-
gand is also important. The fact that the electronic depen-
dence of the iodobenzenes in our study does not yield a
linear Hammett correlation can be explained by assuming
that changing the electronic properties of the aryl halide
does not only change the rate of insertion, but also the equi-
librium value of the monomer/dimer equilibrium. There-
fore, both the amount of active catalyst and its intrinsic
reactivity is affected by the electron-density in the aryl hal-
ide substrate. In addition, it is conceivable that the oxidative
addition step becomes involved in the overall reaction rate
equation when using electron-rich aryl halides.
Figure 1. Hammett plot of the M–H reaction of iodobenzene with
para-substituted styrene derivatives (10) using 9 as the catalyst, at
50 °C in NMP (σ-values were taken from reference[49]).
Kinetic Studies
Despite its importance, very few detailed kinetic studies
on the M–H reaction have been reported that include the
entire catalytic process.[56] Recent breakthroughs in Reac-
tion Progress Kinetic Analysis by Blackmond should allow
for more efficient and accurate kinetic studies.[57,58]
To obtain further insight into the mechanistic details of
the M–H reaction of iodobenzene with styrene and with
butyl acrylate catalysed by 9, kinetic studies have been per-
formed.[16,17] Using the initial rate method, the rate Equa-
tion (4) was observed.
Figure 2. Hammett plot of the M–H reaction of para-substituted
iodobenzene derivatives (11) with styrene using 9 as the catalyst, at
50 °C in NMP (σ-values were taken from reference[46]).
rate = k[Pd]1/2[PhI]0[alkene]1[Et3N]0
(4)
When using palladium concentrations below 4.0 mm for
2 and conversions that do not exceed 70%, no decomposi-
If the alkene coordination is the slower step in the cata- tion of the catalyst was observed, resulting in well-defined
lytic cycle, an increasing electron-density on the alkene moi- pseudo first order kinetics during a large part of the reac-
ety enhances the overall reaction rate. This is because a tion. This is illustrated by the linear correlation between the
more electron-rich alkene is better capable of coordinating logarithm of the conversion and the reaction time (Fig-
to the electron-deficient palladium centre through the well- ure 3).
known donation/back-donation mechanism of π-elec-
These kinetic results have been rationalised as follows.
trons.[50,51] This is not in accordance with the electronic de- The zero-order dependence in iodobenzene indicates a cata-
pendence of the reaction rate observed. This implies that lytic reaction in which the oxidative addition is fast under
the rate of the migratory insertion is involved in the overall the conditions employed. This is in contrast to earlier re-
rate equation.
ports that the oxidative addition is the rate-limiting step in
It has been suggested that the insertion process is best the M–H reaction.[59] Assuming that the elimination/re-
described as a concerted process, in which the actual charge duction step is base-assisted (vide infra), the independence
separation is very low.[36] This assumption is supported by of the observed reaction rate on the concentration of trieth-
theoretical calculations.[51,52] Brookhart and co-workers re- ylamine implies that this elementary reaction is also fast in
ported a systematic study on the electronic effects in the β- our model system. The observed first-order rate-depen-
alkyl migratory insertion reaction of styrene in methylpalla- dence on the alkene concentration clearly shows that
Eur. J. Inorg. Chem. 2012, 1660–1671
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1665

P. W. N. M. van Leeuwen, P. C. J. Kamer et al.
FULL PAPER
Scheme 4. Postulated catalytic cycle of the Mizoroki–Heck reaction
using bulky monodentate phosphorus ligands.
Figure 3. Logarithmic plot of the conversion of iodobenzene vs.
time in a typical experiment using 9 as the catalyst in the reaction
with styrene in CH3CN at 65 °C.
bon bond takes place to form a complex with a syn arrange-
styrene is involved in the turn-over limiting step of the cycle. ment between palladium and the hydrogen atom that is
This implies that either the alkene coordination to palla- eliminated. This β-H elimination yields the product and a
dium or the subsequent insertion of the alkene into the Pd– coordinatively unsaturated palladium(0) species is formed
aryl bond is slow. Since the former process is likely to be by reaction with the external base that can activate a new
reversible,[60] these two steps cannot be distinguished based aryl halide molecule. Recent kinetic methods also include
on the present kinetic results only. The observed linear cor- catalyst deactivation processes resulting in an even better
relation between the reaction rate and the square root of agreement between model and experiment.[16,58]
the catalyst concentration is indicative of an equilibrium be-
Our previous studies on ligand effects[17] combined with
tween a dimeric form of the complex and the monomeric this kinetic study show that optimal catalyst performance is
isomer. This monomer is the catalytically active species. obtained by a subtle balance between the steric and elec-
When it is assumed that the equilibrium is shifted mainly tronic properties of the ligands employed. When the steric
towards the side of the dimer, and that the rate of equilibra- bulk is large enough, mono-coordination is enforced, since
tion is much higher than the rate of reaction of the mono- the steric crowding does not allow the coordination of a
mer with the alkene, the overall reaction rate shows a half second ligand. This renders the metal centre coordinatively
order dependence on the catalyst concentration. NMR unsaturated, resulting in rapid overall reaction. Brown and
spectroscopy (vide supra), shows that the dimer is the main co-workers studied the influence of the steric bulk of phos-
species present and that the cis/trans isomerisation (that phanes on the oxidative addition reaction.[62] They found,
might proceed through a monomeric complex) is fast. Pfaltz similar to the earlier work of Hartwig,[63] that different re-
and Blackmond reported kinetic studies using palladacycles action pathways exist, yielding mononuclear and/or dimeric
as the catalyst.[56b,56c] Their results confirmed our conclu- species, and that the exact outcome of the reaction exhibits
sions, i.e. the oxidative addition product acts as a dimeric extreme sensitivity to the steric environment of the metal
resting state from which the monomeric complex is formed centre.
in small quantities and reacts relatively slowly with the
Within our model studies, ligand 3b apparently possesses
alkene. Westerhuis et al. reported the presence of an appar- the optimal steric and electronic ligand properties for fast
ent initiation period using 9 in the M–H coupling of n-butyl reaction. The use of bulkier ligands (e.g. 3c) retards the re-
acrylate and iodobenzene using real-time near infrared action, while the use of smaller ligands like 3a also results
spectroscopy which is also in agreement of this theory.[61] in slower catalytic reaction.[11,17] When an excess (more
The traditional catalytic cycle of the M–H reaction should than two equivalents) of the relatively small phosphoramid-
therefore be adjusted for the systems applied in our studies ite 3a is reacted with Pd(dba)2 in acetonitrile, a singlet for
as depicted in Scheme 4.
the free ligand together with two doublets (δ = 161.9 and
2
In this mechanism, the oxidative addition takes place at 163.9, JP,P = 84 Hz) are observed. We assign this to a
a coordinatively unsaturated palladium intermediate. The L2Pd(dba) complex, in which two phosphoramidite ligands
resulting monomeric Pd(aryl)(halide) complex is in rapid are coordinated to the metal centre. The formation of sim-
equilibrium with the corresponding dimer. After slow coor- ilar species is well known for non-bulky monodentate triar-
dination/insertion of the alkene to the mononuclear species, ylphosphane ligands.[64] Addition of iodobenzene to this
the mechanism can proceed through the commonly ac- L2Pd(dba) complex resulted in slow reaction, with the for-
cepted reaction sequence. Rotation around the carbon-car- mation of several unidentified species. In contrast, after the
1666
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 1660–1671

Bulky Monodentate Phosphorus Ligands
gen. Benzene, toluene and dioxane were distilled from sodium, di-
ethyl ether, tetrahydrofuran and dimethoxyethane were distilled
from sodium/benzophenone and hexane and pentane were distilled
from sodium/benzophenone/triglyme. Dichloromethane, ethyl acet-
ate and acetonitrile were distilled from CaH2. DMF was vacuum-
distilled from CaH2. NMP and DMA were degassed by a freeze-
thaw cycle and stored on molsieves (4 Å) for at least a week. Trieth-
ylamine was distilled from LiAlH4. Alkenes were purified prior to
use by filtration through a plug of basic alumina (150 mesh, Ald-
rich Chemical Co.). Chemicals were purchased from Aldrich and
reaction of an excess of 3b with Pd(dba)2 under similar con-
ditions, no signals could be observed in the 31P NMR spec-
trum. Probably, the resonances are too broad to be ob-
served at room temperature. The resulting complex decom-
poses slowly in solution to palladium metal, with the con-
comitant formation of uncoordinated ligand and several
unidentified phosphorus-containing species. Addition of
iodobenzene to the (3b)Pd0 complex results in the rapid for-
mation of the anticipated iodide-bridged palladium(II) di-
mer with a structure analogous to 9, as indicated by the Acros Chimica. Ligands 1, 3,[13] 4,[66] and complex [Pd(3b)(Ph)(μ-
Br)]2 (9)[30] were prepared according to literature methods. NMR
spectra were recorded on a Bruker AMX 300, a Varian Mercury
300, a Bruker DRX 300, a Bruker Avance 300 or a Bruker Avance
II 400. 31P and 13C spectra were measured 1H decoupled (unless
stated otherwise). Deuterated solvents were first degassed and then
vacuum transferred from a drying agent. CD2Cl2 and CD3CN were
distilled from CaH2. TMS was used as an external standard for 1H
and 13C NMR spectroscopy and H3PO4 for 31P NMR spec-
troscopy. Molecular weight determinations were performed on a
Hewlett–Packard vapour pressure osmometer model 301A, using
appearance of a signal at δ = 119.4 in the 31P NMR spec-
trum. This demonstrates the importance of the steric bulk
of the ligand in enforcing mono-coordination and thereby
resulting in fast oxidative addition.
Since the palladium complex containing the bulky tris(o-
tolyl)phosphane (1b) was shown to react at a lower rate[17]
than the complex modified with the bulky phosphoramidite
3b, we conclude that also the electronic properties of the
ligand are important for a fast reaction. Since the oxidative
addition is not the slow step of the cycle, the use of ligands benzil as the reference compound, a Micromass GCT or a Micro-
mass LCT. Elemental analyses were performed on a Hereaus Ele-
mentar Vario EL, a Carlo Erba CHNS analyser or by Stephen
Boyer at the London Metropolitan University. GC measurements
were performed on a Shimadzu GC-17A apparatus [split/splitless,
equipped with a F.I.D. detector and a BPX35 column (internal dia-
meter of 0.22 mm, film thickness 0.25 mm, carrier gas 70 kPa He)]
and an Interscience HR GC Mega 2 apparatus (J&W Scientific,
DB-1 column, 30 m/inner diameter 0.32 mm/film thickness 3.0 mm,
carrier gas 70 kPa He, F.I.D. detector). GC/MS measurements (E.I.
detection) were performed on a HP 5890/5971 apparatus, equipped
with a ZB-5 column (5% cross-linked phenyl polysiloxane) with a
0.25 mm internal diameter and 0.25 mm film thickness.
having larger basicity does not result in faster overall reac-
tion. The observed ligand trend suggests that the alkene
coordination or insertion is faster when the palladium cen-
tre is less electron-rich. When phosphites are employed, the
reaction is less efficient again,[11,17] which might be caused
by the fact that the oxidative addition in this case has be-
come a slower step.
Conclusions
The results presented demonstrate that not only the ba-
sicity and the steric properties of the ligands but also the
solvent, the temperature and additives have a large influ-
ence on the catalytic performance of the catalytic systems
used. Studies have been reported in which these factors
were shown to be of decisive importance.[65] This implies
that ligand effects change upon variation of the conditions
and nature of the model reaction. Extrapolation of the re-
sults obtained in catalytic studies should therefore be done
with caution. Nevertheless, we have demonstrated that
bulky phosphoramidite-based palladium catalysts in the
M–H reaction show very high activity and important mech-
anistic insight has been gained.
Our kinetic investigations have led us to the conclusion
that oxidative addition is not rate limiting. This is rational-
ised by the zero order dependence in iodobenzene which
suggests rapid oxidative addition (also determined by 31P
NMR). The rate of the base assisted elimination step is in-
dependent of the concentration of base, also implying a fast
reaction. The first order rate dependence of styrene suggests
involvement in the rate-limiting step in the form of alkene
coordination or insertion. The electronic substrate effects
are more in line with a rate limiting migratory insertion.
Preparation of Complex 9: This compound, [Pd(3b)(Ph)(μ-Br)]2,
was prepared according to a literature method.[28] 1H NMR
(300 MHz, CDCl3): δ = 6.94 (m, 4 H, Ar-H), 6.81 (m, 4 H, Ar-H),
6.62 (m, 10 H, Ar-H), 4.45 [m, 4 H, NCH(CH3)2], 3.81 (s, 12 H,
CH3O), 1.41 [s, 36 H, C(CH3)3], 1.25 [s, 12 H, NCH(CH3)2], 1.23
[s, 12 H, NCH(CH3)2] ppm. 31P NMR (121.5 MHz, CDCl3): δ =
118.3 (s) ppm. 13C NMR (75.5 MHz, CDCl3): δ = 155.0, 143.0 (d,
2JCP = 14 Hz), 141.9, 135.2, 131.3, 126.7, 122.7, 117.1, 114.5, 113.2,
2
55.4, 50.0, 49.9 (d, JCP
= 18 Hz), 35.6, 31.7, 25.0 ppm.
C68H94Br2N2O4P2Pd2 (1438.06): calcd. C 54.38, N 1.86, H 6.31;
found C 54.86, N 1.69, H 6.38.
10-Phenyl-10H-dibenzo[b,e][1,4]thiaphosphinine (2a):[21] Diphenyl
sulfide was aziotropically dried with toluene (3ϫ2 cm3) and dis-
solved in 1:1 diethyl ether/hexane (50 cm3) to which TMEDA
(3.575 cm3, 2.789 g, 24 mmol) was added. The solution was cooled
to 0 °C and nBuLi (2.5 m in hexanes, 9.6 cm3, 24 mmol) added
dropwise via syringe. The mixture was warmed to room temp.
slowly overnight. Dilithiation was confirmed by GC–MS, the solu-
tion cooled to –78 °C and a solution of PhPCl2 (2.035 cm3, 2.685 g,
15 mmol) in hexane (5 cm3) was added dropwise and the reaction
warmed to room temp. slowly overnight. Degassed water (30 cm3)
and diethyl ether (20 cm3) were added to aid separation and the
organics extracted with diethyl ether (3ϫ10 cm3), washed with
brine and dried with MgSO4 (oven dried and degassed). The sol-
vent was removed to give a yellow oil, crude yield (1.681 g, 58%).
The title compound was then purified by conversion to the BH3
adduct, and then deprotected as required. 31P NMR (121.5 MHz,
CDCl3): δ = –19.1 ppm. 1H NMR (300 MHz, CDCl3): δ = 7.45 (m,
2 H, Ar-H), 7.32 (m, 5 H, Ar-H), 7.15 (m, 6 H, Ar-H) ppm. 13C
Experimental Section
General Remarks: All reactions were carried out using standard
Schlenk techniques under an atmosphere of purified argon or nitro-
Eur. J. Inorg. Chem. 2012, 1660–1671
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1667

P. W. N. M. van Leeuwen, P. C. J. Kamer et al.
FULL PAPER
NMR (100 MHz, CDCl3): δ = 136.8 (d, JCP = 10 Hz), 136.4 (d,
JCP = 13 Hz), 135.2 (d, JCP = 21 Hz), 134.9 (d, JCP = 12 Hz), 132.3
(d, JCP = 22 Hz), 129.8, 128.9 (d, JCP = 8 Hz), 128.4, 128.2 (d, JCP
= 2 Hz), 126.9 (d, JCP = 7 Hz) ppm.
crude batch (2.040 g, 6.703 mmol, combined weight) and dissolved
in THF (50 cm3) and cooled to 0 °C. BH3·DMS (2 m in THF,
5.66 cm3, 11.3 mmol) was added drop wise. The reaction was com-
plete after 1 h, and the solvent removed in vacuo. The white solid
was recrystallised from acetonitrile (1.0821 g, 51%, based on pro-
tection step, 32% overall yield). 31P NMR (162 MHz, CDCl3): δ =
–14.8 (m) ppm. 1H NMR (400 MHz, CDCl3): δ = 7.54 (m, 4 H,
10-Phenyl-10H-dibenzo[b,e][1,4]thiaphosphinine·BH3 (2a·BH3): To a
solution of 10-phenyl-10H-dibenzo[b,e][1,4]thiaphosphinine (2a)
(1.681 g, 5.75 mmol) in THF (50 cm3) at 0 °C, BH3·DMS (2 m solu-
tion in THF, 11.50 cm3, 6.325 mmol) was added dropwise. After
2.5 h, 31P{1H} NMR confirmed conversion to the BH3 adduct. The
solvent was removed in vacuo, and crystals were obtained from
recrystallysation from toluene. Analytically pure samples could be
obtained by repeated recrystallisations; yield (0.192 g, 11%). 31P
NMR (162 MHz, CDCl3): δ = 4.5 (br. m) ppm. 1H NMR
(400 MHz, CDCl3): δ = 8.30 (m, 2 H, Ar-H), 7.60 (m, 2 H, Ar-H),
7.51 (m, 2 H, Ar-H), 7.35 (m, 1 H, Ar-H), 7.25 (m, 2 H, Ar-H),
7.16 (m, 2 H, Ar-H), 1.31 (3 H, br. m, BH3) ppm. 13C NMR
3
4
Ar-H), 7.34 (m, 5 H, Ar-H), 7.18 (dd, 2 H, JHH = 8.5, JHH
=
1
3.6 Hz, Ar-H), 2.3 (s, 6 H, CH3), 1.27 (3 H, bq, JHB = 95 Hz,
BH3) ppm. 13C NMR (75.5 MHz, CDCl3): δ = 153.8, 134.5 (d, JCP
= 2 Hz), 134.4, 133.0, 132.7 (d, JCP = 13 Hz), 132.2 (d, JCP
=
11 Hz), 131.5 (d, JCP = 2 Hz), 129.2 (d, JCP = 10 Hz), 118.4 (d, JCP
= 4 Hz), 110.2 (d, JCP = 61 Hz), 21.1 ppm. 11B NMR (128.4 MHz,
CDCl3): δ = –39.64 (br. s) ppm. MS(ESI+) m/z 314.95 [M – 3 H].
C20H20BOP (318.16): calcd. C 70.50, H 6.34; found C 73.00, H
6.64.
(100 MHz, CDCl3): δ = 139.3, 134.7 (d, JCP = 16 Hz), 131.5 (d, 2,8-Dimethyl-10-phenyl-10H-phenoxaphosphinine (2c): 2,8-Di-
JCP = 1 Hz), 131.3 (d, JCP = 10 Hz), 130.7 (d, JCP = 2 Hz), 128.5
(d, JCP = 11 Hz), 128.1 (d, JCP = 4 Hz), 127.5 (d, JCP = 12 Hz),
124.4, 123.8 ppm. 11B NMR (128.4 MHz, CDCl3): δ = 39.6 (m)
ppm. MS (ESI+): m/z 329 [M + Na], 635 (2M + Na). C18H16BPS
(306.17): calcd. C 70.61, H 5.19; found C 70.69, H 5.27.
methyl-10-phenyl-10H-phenoxaphosphinine·BH3 (2c·BH3) was dis-
solved in ethanol and heated to reflux for 2–3 h. Once the gas evol-
ution had ceased and the deprotection was deemed complete by
31P{1H} NMR, the solvent was removed in vacuo. The residue was
then azeotropically dried with 3 aliquots of toluene and finally
dried in vacuo at 60 °C. 31P NMR (121 MHz, CDCl3): δ = –54.4
10-Phenyl-10H-phenoxaphosphinine (2b):[21] Diphenyl ether
(1.702 g, 10 mmol) was azeotropically dried with toluene
(3ϫ3 cm3) and dissolved in 1:1 hexane/diethyl ether (50 cm3) to
which TMEDA (3.575 cm3, 2.789 g, 24 mmol) was added. The
solution was cooled to 0 °C and nBuLi (2.4 m in hexanes, 10 cm3,
24 mmol) was added dropwise via syringe. The mixture was
warmed to room temp. slowly overnight. Dilithiation was con-
firmed by GC–MS, the solution cooled to –78 °C and a solution
1
ppm. H NMR (400 MHz, CDCl3): δ = 7.28 (m, 2 H, Ar-H), 7.23
3
(m, 2 H, Ar-H), 7.18 (m, 3 H, Ar-H), 7.13 (ddd, 2 H, JHH = 8.3,
4
3
4JHH = 2.2, JHH = 0.6 Hz, Ar-H), 7.05 (d, 2 H, JHH = 8.3 Hz,
Ar-H), 2.29 (s, 6 H, CH3) ppm. 13C NMR (100 MHz, CDCl3): δ =
153.4, 140.7 (d, JCP = 22 Hz), 135.0 (d, JCP = 35 Hz), 132.9 (d, JCP
= 11 Hz), 132.0 (d, JCP = 20 Hz), 131.7, 128.6, 128.5 (d, JCP
7 Hz), 117.9 (d, JCP = 3 Hz), 117.5, 20.6 ppm.
=
of PhPCl2 (2.035 cm3, 2.685 g, 15 mmol) in hexane (5 cm3) was 2,8-Dimethyl-10-phenyl-5,10-dihydrophenophosphazinine
(2d):[21]
added drop wise and the reaction warmed to room temp. slowly
overnight. Degassed water (20 cm3) was added and the organics
extracted with diethyl ether (3ϫ10 cm3), washed with brine
(10 cm3) and then dried with MgSO4 (oven dried and degassed).
The solvent was removed in vacuo to give a yellow oil. Purification
over silica using hexanes/ethyl acetate (60:40) followed by
recrystallisation from ethanol and THF at 5 °C produced off white
(pTol)2NH (2.48 g, 12.56 mmol) was azeotropically dried with tolu-
ene (3ϫ5 cm3), and dried in vacuo. This was then dissolved in
diethyl ether (30 cm3), and PhPCl2 (2.04 cm3, 2.70 g, 15.07 mmol)
was added. On stirring, the solution turned orange with the forma-
tion of a white precipitate. The mixture was heated to 60–80 °C
and was the ether distilled off to leave an orange oil. The flask was
then fitted with a reflux condenser, and the mixture heated to
150 °C, with the evolution of HCl. After 30 min, the temperature
crystals. Further washing with pentane gave a white solid (0.414 g,
1
15%). 31P NMR (121.47 MHz, CDCl3): δ = –54.3 ppm. H NMR was increased to 200 °C for 2 h. Once cooled to room temperature,
3
3
(300 MHz, CDCl3): δ = 7.52 (ddd, 2 H, JHP = 10.4, JHH = 7.5,
dichloromethane (100 cm3) was added, followed by 1 m NaOH
3
3
4JHH = 1.7 Hz, Ar-H), 7.38 (ddd, 2 H, JHH = 8.3, JHH = 7.2, solution (95 cm3). The organic layer was removed via cannula, and
4JHH = 1.7 Hz, Ar-H), 7.20 (m, 7 H, Ar-H), 7.12 (2 H, tt, JHH
=
the aqueous layer extracted with dichloromethane (3ϫ30 cm3) and
the combined organic layers dried with magnesium sulfate. The
pale yellow solution was then filtered through a silica column, and
reduced in volume. Hexane was added, and a white precipitate be-
gan to form. On standing at 5 °C overnight, a white crystalline
solid was produced, which was filtered via cannula and dried in
vacuo. Further product could be obtained from reducing the fil-
trate and cooling. In order to remove any oxide, the compound was
dissolved in dichloromethane, and passed through a second silica
column, and recrystallised from dichloromethane/hexane; yield
(1.17 g, 31%). 31P NMR (162 MHz, CDCl3): δ = –45.1 ppm. 1H
3.7, JHH = 1.4 Hz, Ar-H) ppm. 13C NMR (100 MHz, CD2Cl2): δ
= 155.7, 140.9 (d, JCP = 21 Hz), 135.3 (d, JCP = 35 Hz), 132.0 (d,
JCP = 20 Hz), 131.5, 129.0, 128.9 (d, JCP = 7 Hz), 124.2 (d, JCP
11 Hz), 118.9 (d, JCP = 4 Hz), 118.2 ppm.
=
2,8-Dimethyl-10-phenyl-10H-phenoxaphosphinine·BH3
(2c·BH3):
Di-p-tolyl ether was azeotropically dried with toluene (3ϫ2 cm3)
and dissolved in hexane/diethyl ether, 1:1 (30 cm3) to which
TMEDA was added. The solution was cooled to 0 °C and nBuLi
(2.3 m in hexanes, 5.26 cm3, 12.11 mmol) added drop wise via sy-
ringe. The mixture was allowed to slowly warm to room temp.,
overnight. Dilithiation was confirmed by GC–MS. The solution
was cooled to –78 °C and a solution of PhPCl2 (1.03 cm3, 1.32 g,
7.56 mmol) in hexane (3 cm3) was added drop wise. The mixture
was allowed to slowly warm to room temp. overnight. Degassed
water (15 cm3) was added, the organics extracted with diethyl ether
(3ϫ10 cm3), washed with brine and dried with MgSO4 (oven dried
and degassed). The solvent was removed in vacuo to give a yellow
oil. Purification over silica with 40% ethyl acetate in hexane gave
a white oil. The crude phosphane 2c was combined with another
3
4
NMR (300 MHz, CD2Cl2): δ = 7.31 (dd, 2 H, JHP = 11.7, JHH
=
=
3
1.4 Hz, Ar-H), 7.06–6.95 (m, 7 H, Ar-H), 6.65 (d, 2 H, JHH
8.2 Hz, ArH), 6.41 (s, 1 H, NH), 2.22 (s, 6 H, CH3) ppm. 13C NMR
(100 MHz, CD2Cl2): δ = 141.2 (d, JCP = 23 Hz), 140.3 (d, JCP
=
2 Hz), 134.8 (d, JCP = 39 Hz), 130.5, 129.9 (d, JCP = 18 Hz), 129.4
(d, JCP = 13 Hz), 127.4 (d, JCP = 6 Hz), 127.0, 114.6 (d, JCP
=
3 Hz), 114.0, 19.5 ppm.
Bis(2-bromo-4-tert-butylphenyl)amine:[25] Bis[4-(tert-butyl)phenyl]-
amine (10 g, 35.53 mmol) was dissolved in acetic acid (150 cm3)
1668
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 1660–1671

Bulky Monodentate Phosphorus Ligands
and cooled to 0 °C. To this a solution of bromine (3.64 cm3,
11.357 g, 71.06 mmol) in acetic acid (20 cm3) was added drop wise
via a dropping funnel which was washed with further acetic acid
(80 cm3) and allowed to stir overnight. Dilute sodium bisulfite (1%,
250 cm3) was added and the resultant white solid was filtered and
recrystallised from DCM/EtOH to give white crystals (14.153 g,
91%). 1H NMR (300 MHz, CDCl3): δ = 7.61 (dd, 2 H,4JHH = 0.9,
4JHH = 1.5 Hz, Ar-H), 7.26 (m, 4 H, Ar-H), 6.31 (s, 1 H, NH), 1.34
(s, 18 H, CH3) ppm. 13C NMR (100 MHz, CDCl3): δ = 145.7,
137.8, 130.0, 125.1, 117.6, 113.9, 34.3, 31.3 ppm. MS(EI+) m/z 439.
C20H25Br2N (439.23): calcd. C 54.69, N 3.19, H 5.74; found C
54.76, N 3.11, H 5.74; m.p. 159 °C (from hexane).
solution was then filtered through silica, and the solvent removed
in vacuo. White crystals were obtained by recrystallisation from
hexane (0.37 g, 0.93 mmol, 42%). 31P NMR (162 MHz, CDCl3): δ
= –43.8 ppm. 1H NMR (400 MHz, CDCl3): δ = 7.75 (dd, 2 H, 3JHP
4
3
4
= 12.7, JHH = 2.5 Hz, Ar-H), 7.43 (dd, 2 H, JHH = 8.6, JHH
=
3
2.5 Hz, Ar-H), 7.11 (m, 3 H, Ar-H), [4 H, 6.97 (d, JHH = 8.9 Hz,
Ar-H) + 6.95 (m, Ar-H)], 3.37 (s, 3 H, NCH3), 1.36 [s, 18 H,
C(CH3)3] ppm. 13C NMR (100 MHz, CDCl3): δ = 145.2, 143.7 (d,
JCP = 13 Hz), 140.1 (d, JCP = 18 Hz), 132.5 (d, JCP = 42 Hz), 130.4
(d, JCP = 17 Hz), 127.9 (d, JCP = 5.5 Hz), 127.4, 127.2, 120.1, 113.9,
35.9, 34.1, 31.5 ppm.
Catalysis General Procedure: A Schlenk vessel was charged with
complex 2 (0.0094 g, 6.25 mmol). Subsequently, solvent (10.0 cm3),
decane (0.50 cm3, 2.56 mmol) as internal standard for GLC, iodo-
benzene (0.56 cm3, 5.0 mmol) and triethylamine (0.76 cm3,
5.5 mmol) were added. The reaction mixture was heated to 65 °C
and styrene (0.63 cm3, 5.5 mmol) was added. Samples were taken
at regular intervals, diluted with dichloromethane and washed with
dilute aqueous HCl. The organic phase was dried with MgSO4 and
analysed by GLC.
Bis(2-bromo-4-tert-butylphenyl)-N-methylamine:[26] Bis(2-bromo-4-
tert-butylphenyl)amine (4.39 g, 10 mmol) was dried azeotropically
with dioxane (10 cm3) and then dissolved in dimethoxyethane
(40 cm3), and triethylamine (13.94 cm3, 10.12 g, 100 mmol) added.
In a separate flask, sodium hydride (60% dispersion in mineral oil,
0.88 g, 22 mmol) was washed with hexane (3ϫ20 cm3), dried in
vacuo and then suspended in dimethoxyethane (10 cm3). The amine
solution was transferred via cannula to the cooled (0 °C) sodium
hydride suspension. This was then stirred for 3–4 h, before adding
methyliodide (0.93 cm3, 2.13 g, 15 mmol) dropwise via a microsy-
ringe. The mixture was left to stir overnight. The reaction was mon-
itored by tlc (silica, hexane, vanillin stain). Once complete saturated
ammonium chloride solution (60 cm3) was added, followed by so-
dium hydroxide [5% (w/w), 75 cm3]. The organics were then ex-
tracted with dichloromethane (3ϫ150 cm3), washed with brine,
dried with MgSO4, and then the solvent removed in vacuo. Purifi-
cation by column chromatography (silica, hexane) gave white crys-
tals (3.53 g, 7.80 mmol, 78%). 1H NMR (400 MHz, CDCl3): δ =
Screening of Phosphacyclic Ligands 2: The ligands were screened
using a Greenhouse parallel synthesiser fitted with reflux heads
allowing each reaction to be carried out in a tube with a microstir-
rer bar, under reflux and an inert atmosphere. In a typical experi-
ment, the ligand and metal precursor were weighed into the dried
test tube and placed in the parallel synthesiser which was sealed
and placed under an inert atmosphere with repeated vacuum argon
cycles. Solvents and liquid reagents were added though individual
injection ports above each reactor fitted with a butyl rubber sepum
mat. The greenhouse was then placed on a standard hotplate and
heated to 80 °C with stirring. Once the temperature had stabilised
triethylamine was added via the injection port to initiate the reac-
tion. Samples were taken after 45 min, diluted with diethyl ether,
filtered through a small silica plug and analysed by GLC. Reactions
were conducted in triplicate to ensure reproducibility.
4
7.48 (d, 2 H, JHH = 2.28 Hz, Ar-H), 7.16 (dd, 2 H,3JHH = 8.40,
3
4JHH = 2.28 Hz, Ar-H), 6.85 (d, 2 H, JHH = 8.40 Hz, Ar-H), 3.12
(s, 3 H, NCH3), 1.22 [s, 18 H, C(CH3)3] ppm. 13C NMR (75 MHz,
CDCl3): δ = 148.0, 146.2, 131.3, 125.0, 123.2, 120.0, 41.4, 34.3, 31.3
ppm. MS(ESI+) m/z 476 (H+) + Na. C21H27Br2N (453.26): C
55.65, N 3.09, H 6.00, found C 56.02, N 3.15, H 6.12.
Kinetic Studies General Procedure: In a representative procedure, a
Schlenk vessel was charged with complex 9. Subsequently, acetoni-
trile (to a total volume of 10 cm3), decane (0.5 cm3, 2.56 mmol) as
internal standard for GLC, iodobenzene and triethylamine were
added. The reaction mixture was heated to 65 °C and styrene
(0.63 cm3, 5.5 mmol) was added. Samples were taken at regular in-
tervals, diluted with dichloromethane and washed with dilute HCl
solution. The organic phase was dried with MgSO4 and analysed
by GLC.
2,8-Di-tert-butyl-10-phenyl-5,10-dihydrophenophosphazinine (2e):[27]
Bis(2-bromo-4-tert-butylphenyl)-N-methylamine (1 g, 2.21 mmol)
was azeotropically dried with dioxane (10 cm3), and dissolved in a
mixture of diethyl ether (2 cm3) and n-pentane (8 cm3). tBuLi (1.6 m
in pentane, 6.89 cm3, 11.03 mmol) was added to n-pentane
(3.1 cm3) and cooled to 0 °C. The amine solution was then added
drop wise via a Teflon cannula, and left to stir for 2 h. The mixture
was then cooled to 0 °C and PhPCl2 (0.30 cm3, 0.40 g, 2.21 mmol)
added drop wise, and the mixture left to stir overnight. At this stage
31P{1H} NMR spectroscopy showed 98% conversion. Degassed
water (7 cm3) was added, and the organic layer separated via can-
nula. The aqueous layer was then extracted with dichloromethane
(3ϫ10 cm3) and the combined organics dried with MgSO4. The
XAFS Measurements on Complex 9: Palladium K-edge (24350 eV)
EXAFS spectra were measured at the European Synchrotron Radi-
ation Facility (ESRF) in Grenoble, France, Beamline 29, using a
Si(111) double crystal monochromator. The monochromator was
Figure 4. Fourier Transforms of raw data (solid line) and R-space fit (dotted line) (3.2 Ͻ k Ͻ 17 Å–1 and 1.0 Ͻ R Ͻ 3.0 Å) of 9 in solid
state at room temp. (a) k0-weighted, (b) k3-weighted.
Eur. J. Inorg. Chem. 2012, 1660–1671
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1669

P. W. N. M. van Leeuwen, P. C. J. Kamer et al.
FULL PAPER
[11] 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,
7, 1073–1076.
detuned to 50% intensity to avoid effects of higher harmonics pres-
ent in the X-ray beam. The measurements were done in the trans-
mission mode using optimised ion chambers as detectors. To de-
crease noise scans were made in k-space and three scans were col-
lected for each sample.
[12] E. Burrello, G. Rothenberg, Adv. Synth. Catal. 2003, 345,
1334–1340.
[13] A. van Rooy, D. Burgers, P. C. J. Kamer, P. W. N. M.
van Leeuwen, Recl. Trav. Chim. Pays-Bas 1996, 115, 492–498.
[14] a) J. F. Hartwig, L. M. Stanley, Acc. Chem. Res. 2010, 43,
1461–1475; b) J. F. Teichert, B. L. Feringa, Angew. Chem. 2010,
122, 2538; Angew. Chem. Int. Ed. 2010, 49, 2486–2528; c) L.
Eberhardt, D. Armspach, J. Harrowfield, D. Matt, Chem. Soc.
Rev. 2008, 37, 839–864; d) J.-H. Xie, Q.-L. Zhou, Acc. Chem.
Res. 2008, 41, 581–593; e) A. J. Minnaard, B. L. Feringa, L.
Lefort, J. G. de Vries, Acc. Chem. Res. 2007, 40, 1267–1277; f)
B. L. Feringa, Acc. Chem. Res. 2000, 33, 346–353.
[15] C. A. Tolman, Chem. Rev. 1977, 77, 313–348.
Theoretical reference data were generated in the commercially
available program XDAP[67] using FEFF8.0[68] Recent develop-
ments in EXAFS data-analysis methods, which make use of the so-
called difference file technique,[29] allow the reliable separation of
the different contributions resulting in a proper analysis (Figure 4).
The results of the EXAFS data-analysis are given in Table 4. The
total fits are of good quality in all weightings applied as can be
concluded from the low variances found between both the imagi-
nary and absolute parts of the Fourier transforms of the fits and
the spectra.[29]
[16] G. Rothenberg, S. Cruz, G. P. F. van Strijdonck, H. Hoefsloot,
Adv. Synth. Catal. 2004, 346, 467–473.
[17] M. D. K. Boele, PhD Thesis, University of Amsterdam, The
Netherlands, 2002.
[18] M. D. K. Boele, P. C. J. Kamer, M. Lutz, A. L. Spek, J. G.
de Vries, P. W. N. M. van Leeuwen, G. P. F. van Strijdonck,
Chem. Eur. J. 2004, 10, 6232–6246.
[19] M. F. Haddow, A. J. Middleton, A. G. Orpen, P. G. Pringle, R.
Papp, Dalton Trans. 2009, 202–209.
[20] M. J.-L. Tschan, E. J. García-Suárez, Z. Freixa, H. Launay, H.
Hagen, J. Benet-Buchholz, P. W. N. M. van Leeuwen, J. Am.
Chem. Soc. 2010, 132, 6463–6473.
[21] a) L. A. van der Veen, P. C. J. Kamer, P. W. N. M.
van Leeuwen, Organometallics 1999, 18, 4765–4777; b) L. A.
van der Veen,PhD Thesis, University of Amsterdam,1999.
[22] M. Van Overschelde, E. Vervecken, S. G. Modha, S. Cogen, E.
Van der Eycken, J. Van der Eycken, Tetrahedron 2009, 65,
6410–6415.
[23] a) E. Goudriaan, Final year project report,University of Am-
sterdam, 2001; b) P. C. Crofts, R. Higginbottom, M. J.
Tajbakhsh-Jadidi, Chem. Res. (S) 1988, 294.
Table 4. EXAFS data analysis of compound 9.[a]
Ab-ScPair[b]
N
R [Å]
σ2 [Å–2]
E0 [eV]
Pd–C
Pd–P
Pd–Br
Pd–Pd
1.1
0.9
1.8
0.8
2.02
2.24
2.53
2.83
0.004
0.001
0.005
0.002
7.37
2.63
9.72
–12.2
[a] Nindp = 28. Fit: r space, 3.2 Ͻ k Ͻ 17.0; k0-weighted Var. Im.
= 0.15, Var. Abs. 0.087, k3-weighted Var. Im. = 0.87, Var. Abs. =
0.35. All parameters iteratively refined. [b] Ab = absorber; Sc =
scatterer. Var. Im. and Var. Abs. are the variances of the fit of the
Imaginary and Absolute part, respectively.
Acknowledgments
The research described in this paper was financially supported by
the Engineering and Physical Sciences Research Council (EPSRC)
(grant number EP/E022154/1; to D. L. D.), DSM Research (to
M. D. K. B.) and the Netherlands Ministry of Economic Affairs
(grant number EETK 97107; to G. P. F. v. S. The authors would
like to thank Dr. M. Tromp (Technische Universität München,
Germany) for providing the EXAFS data, and E. Goudriaan and
M. Kunzi for their work on ligand synthesis.
[24] L. D. Freedman, H. S. Freeman, Chem. Rev. 1987, 87, 289–306.
[25] H. Gilman, E. A. Zuech, J. Org. Chem. 1961, 26, 3481–3484.
[26] a) E. Merisor, U. Beifuss, Tetrahedron Lett. 2007, 48, 8383–
¸
8387; b) K. Mohri, K. Suzuki, M. Usui, K. Isobe, Y. Tsuda,
Chem. Pharm. Bull. 1995, 43, 159–161.
[27] T. Agou, J. Kobayashi, T. Kawashima, Org. Lett. 2006, 8, 2241–
2244.
[28] a) F. Paul, J. Patt, J. F. Hartwig, Organometallics 1995, 14,
3030–3039; b) F. Barrios-Landeros, B. P. Carrow, J. F. Hartwig,
J. Am. Chem. Soc. 2009, 131, 8141–8154.
[29] D. C. Koningsberger, B. L. Mojet, G. E. Van Dorssen, D. E.
Ramaker, Top. Catal. 2000, 10, 143–155.
[1] T. Mizoroki, K. Mori, A. Ozaki, Bull. Chem. Soc. Jpn. 1971,
44, 581.
[2] R. F. Heck, J. P. Nolley, J. Org. Chem. 1972, 37, 2320–2322.
[3] X. F. Wu, P. Anbarasan, H. Neumann, M. Beller, Angew.
Chem. 2010, 122, 138; Angew. Chem. Int. Ed. 2010, 49, 9047–
9050.
[4] a) J. P. Knowles, A. Whiting, Org. Biomol. Chem. 2007, 5, 31–
44; b) M. García-Melchor, G. Ujaque, F. Maseras, A. Lledós,
in: Catalysis by Metal Complexes, Phosphorus Compounds
(Eds.: M. Peruzzini, L. Gonsalvi), Springer, Dordrecht, The
Netherlands, 2011, vol. 37, pp. 57–84.
[5] W. A. Herrmann, C. Brossmer, K. Oefele, C.-P. Reisinger, T.
Priermeier, M. Beller, H. Fischer, Angew. Chem. 1995, 107,
1989; Angew. Chem. Int. Ed. Engl. 1995, 34, 1844–1848.
[6] A. Jutand, in: The Mizoroki–Heck Reaction (Ed.: M. Oestre-
ich), John Wiley & Sons, Chichester, United Kingdom, 2009,
p. 1–50.
[7] J. A. Gillespie, D. L. Dodds, P. C. J. Kamer, Dalton Trans. 2010,
39, 2751–2764.
[8] N. Fey, Dalton Trans. 2010, 39, 296–310.
[9] J. Jover, N. Fey, J. N. Harvey, G. C. Lloyd-Jones, A. G. Orpen,
G. J. J. Owen-Smith, Organometallics 2010, 29, 6245–6258.
[10] A. G. Maldonado, G. Rothenberg, Chem. Soc. Rev. 2010, 39,
1891–1902.
[30]
F. Paul, J. Patt, J. F. Hartwig, Organometallics 1995, 14, 3030–
3039.
[31]
[32]
V. V. Grushin, H. Alper, Organometallics 1993, 12, 1890–1901.
W. Cabri, I. Candiani, A. Bedeschi, R. Santi, J. Org. Chem.
1990, 55, 3654–3655.
[33]
[34]
[35]
[36]
[37]
[38]
M. Ludwig, S. Stromberg, M. Svensson, B. Akermark, Organo-
metallics 1999, 18, 970–975.
K. S. A. Vallin, M. Larhed, A. Hallberg, J. Org. Chem. 2001,
66, 4340–4343.
K. Maeda, E. Farrington, E. Galardon, B. John, J. Brown, Adv.
Synth. Catal. 2002, 344, 104–109.
I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100, 3009–
3066.
W. A. Herrmann, C. Brossmer, C.-P. Reisinger, T. H. Riermeier,
K. Oefele, M. Beller, Chem. Eur. J. 1997, 3, 1357–1364.
C. Amatore, A. Jutand, A. Suarez, J. Am. Chem. Soc. 1993,
115, 9531–9541.
[39]
[40]
C. Amatore, A. Jutand, Acc. Chem. Res. 2000, 33, 314–321.
a) Y. Ben-David, M. Portnoy, M. Gozin, D. Milstein, Organo-
metallics 1992, 11, 1995–1996; b) M. Portnoy, Y. Ben-David, I.
Rousso, D. Milstein, Organometallics 1994, 13, 3465–3479; c)
1670
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 1660–1671

Bulky Monodentate Phosphorus Ligands
M. Portnoy, Y. Ben-David, D. Milstein, Organometallics 1993,
12, 4734–4735.
M. Beller, A. Zapf, Synlett 1998, 792–793.
A. F. Littke, G. C. Fu, J. Am. Chem. Soc. 2001, 123, 6989–
7000.
A. F. Littke, G. C. Fu, J. Org. Chem. 1999, 64, 10–11.
J. P. Stambuli, S. R. Stauffer, K. H. Shaughnessy, J. F. Hartwig,
J. Am. Chem. Soc. 2001, 123, 2677–2678.
R. Stürmer, Angew. Chem. 1999, 111, 3509; Angew. Chem. Int.
Ed. 1999, 38, 3307–3308.
W. Cabri, I. Candiani, Acc. Chem. Res. 1995, 28, 2–7.
G. D. Daves, A. Hallberg, Chem. Rev. 1989, 89, 1433–1445.
2005, 127, 12054–12065; e) A. Sud, R. M. Deshpande, R. V.
Chaudari, J. Mol. Catal. A 2007, 270, 144–152.
[57] a) D. G. Blackmond, Angew. Chem. 2005, 117, 4374; Angew.
Chem. Int. Ed. 2005, 44, 4302–4320.
[58] J. S. Mathew, M. Klussmann, H. Iwamura, F. Valera, A. Fu-
tran, E. A. C. Emanuelsson, D. G. Blackmond, J. Org. Chem.
2006, 71, 4711–4722.
[41]
[42]
[43]
[44]
[59] a) V. P. W. Böhm, W. A. Herrmann, Chem. Eur. J. 2001, 7,
4191–4197; b) W. Cabri, I. Candiani, A. Bedeschi, J. Org.
Chem. 1990, 55, 3654–3655.
[45]
[60] I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100, 3009–
[46]
[47]
[48]
3066.
[61] S. C. Cruz, P. J. Aarnoutse, G. Rothenberg, J. A. Westerhuis,
A. K. Smilde, A. Bliek, Phys. Chem. Chem. Phys. 2003, 5,
4455–4460.
P. Fristrup, S. Le Quement, D. Tanner, P.-O. Norrby, Organo-
metallics 2004, 23, 6160–6165.
[49] J. March, in: Advanced Organic Chemistry, 3rd ed., John
Wiley & Sons, New York, 1985.
[50] F. C. Rix, M. Brookhart, P. S. White, J. Am. Chem. Soc. 1996,
118, 2436–2448.
[62] E. Galardon, S. Ramdeehul, J. M. Brown, A. Cowley, K. K.
Hii, A. Jutand, Angew. Chem. 2002, 114, 1838; Angew. Chem.
Int. Ed. 2002, 41, 1760–1763.
[63] J. F. Hartwig, F. Paul, J. Am. Chem. Soc. 1995, 117, 5373–5374.
[64] a) C. Amatore, A. Jutand, J. Organomet. Chem. 1999, 576, 254–
278; b) C. Amatore, A. Jutand, G. Meyer, Inorg. Chim. Acta
1998, 273, 76–84; c) W. A. Herrmann, W. R. Thiel, C.
Brossmer, K. Öfele, T. Priermeier, W. Scherer, J. Organomet.
Chem. 1993, 461, 51–60.
[51] H. von Schenck, B. Akermark, M. Svensson, Organometallics
2002, 21, 2248–2253.
[52] K. Albert, P. Gisdakis, N. Rosch, Organometallics 1998, 17,
1608–1616.
[53] H. von Schenck, S. Strömberg, K. Zetterberg, M. Ludwig, B.
Akermark, M. Svensson, Organometallics 2001, 20, 2813–2819.
[54] W. Cabri, I. Candiani, A. Bedeschi, S. Penco, R. Santi, J. Org.
Chem. 1992, 57, 1481–1486.
[55] B. P. Carrow, J. F. Hartwig, J. Am. Chem. Soc. 2010, 132, 79–
81.
[56] a) M. Casey, J. Lawless, C. Shirran, Polyhedron 2000, 19, 517–
520; b) T. Rosner, J. Le Bars, A. Pfaltz, D. G. Blackmond, J.
Am. Chem. Soc. 2001, 123, 1848–1855; c) T. Rosner, A. Pfaltz,
D. G. Blackmond, J. Am. Chem. Soc. 2001, 123, 4621–4622; d)
C. S. Consorti, F. R. Flores, J. Dupont, J. Am. Chem. Soc.
[65] V. P. W. Böhm, W. A. Herrmann, Chem. Eur. J. 2000, 6, 1017–
1025.
[66] A. van Rooy, E. N. Orij, P. C. J. Kamer, P. W. M. N.
van Leeuwen, Organometallics 1995, 14, 34–43.
[67] M. Vaarkamp, J. C. Linders, D. C. Koningsberger, Phys. B
1995, 208–209, 159–160.
[68] A. L. Ankudinov, B. Ravle, J. J. Rehr, D. S. Conradson, Phys.
Rev. B 1998, 58, 7565–7576.
Received: November 14, 2011
Published Online: February 21, 2012
Eur. J. Inorg. Chem. 2012, 1660–1671
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1671