Palladium(II) complexes of new bulky bidentate phosphanes: Active and highly regioselective catalysts for the hydroxycarbonylation of styrene

Article abstract of DOI:10.1002/chem.200901178

The synthesis of four new bulky bidentate phosphines that possess both tert-butyl and trifluoromethylphenyl substituents is described. Symmetric ligands were readily obtained by alkylation of phosphidoboranes of the type Li[P(BH₃)(tBu)(Ar)] with dihaloalkanes. Non-symmetric ligands were prepared from a new stable precursor, tBu₂P(BH₃)(CH ₂)₃Br, that should prove useful for other ligand syntheses. Palladium(II) complexes of the four new ligands were prepared and were characterised by spectroscopic methods, microanalysis and X-ray crystallography. The new [PdCl₂(L)] complexes were evaluated as catalysts for the hydroxycarbonylation of styrene and found to give unprecedented regioselectivity and yields for a diphosphine-based catalyst. A study on promoter effects reveals that the presence of acid and chloride is necessary to achieve such selectivities. It has been proposed in the literature that such conditions result in a new pathway in which styrene is converted into 2-phenethyl chloride, with the latter being the real substrate in the reaction. However, a deuterium labelling study seems to rule out this mechanism, at least under the conditions used herein.

Full text of DOI:10.1002/chem.200901178

DOI: 10.1002/chem.200901178
Palladium(II) Complexes of New Bulky Bidentate Phosphanes: Active and
Highly Regioselective Catalysts for the Hydroxycarbonylation of Styrene
Jamie J. R. Frew,[a] Karen Damian,[a] Hendrik Van Rensburg,[b]
Alexandra M. Z. Slawin,[a] Robert. P. Tooze,[b] and Matthew L. Clarke*[a]
Abstract: The synthesis of four new
bulky bidentate phosphines that pos-
sess both tert-butyl and trifluorome-
thylphenyl substituents is described.
Symmetric ligands were readily ob-
tained by alkylation of phosphidobor-
anes of the type Li[PACHTNUGTRENNUG(BH3)ACHTUNGTRENN(UGN tBu)(Ar)]
with dihaloalkanes. Non-symmetric li-
gands were prepared from a new stable
and were characterised by spectroscop-
ic methods, microanalysis and X-ray
crystallography. The new [PdCl2(L)]
complexes were evaluated as catalysts
for the hydroxycarbonylation of sty-
rene and found to give unprecedented
regioselectivity and yields for a diphos-
phine-based catalyst. A study on pro-
moter effects reveals that the presence
of acid and chloride is necessary to
achieve such selectivities. It has been
proposed in the literature that such
conditions result in a new pathway in
which styrene is converted into 2-phe-
nethyl chloride, with the latter being
the real substrate in the reaction. How-
ever, a deuterium labelling study seems
to rule out this mechanism, at least
under the conditions used herein.
precursor, tBu2PACHTUNGTRENNUNG(BH3)ACHTUNTGERN(NUGN CH2)3Br, that
Keywords: carbonylation · diphos-
phanes · fluorinated phosphanes ·
palladium · reaction mechanisms
should prove useful for other ligand
syntheses. Palladium(II) complexes of
the four new ligands were prepared
Introduction
lution required to separate the two enantiomers.[3] There has
been considerable interest in developing catalysts that can
optimise the branched regioselectivity of the racemic prod-
ucts that arise from vinylarene carbonylation.[4] In addition,
a practical catalytic procedure that is both regio- and enan-
tioselective is particularly desirable. Chiral diphosphines
have been screened for the hydroxycarbonylation and me-
thoxycarbonylation of alkenes, but in the case of arylalkene
carbonylation, these diphosphine-based catalysts give low
turnover frequencies, low enantiomeric excess (ee) and re-
gioselectivity that favours the unwanted linear isomer, often
under very high pressures of >100 bar (Scheme 1).[4,5,6]
Good results have been particularly elusive in hydroxycar-
bonylation, which for some reason seems more challenging
than alkoxycarbonylations.[1,4a,6b] Some monophosphine cata-
lysts do show good activity and the desired branched regio-
selectivity under relatively mild conditions, but in these
À
Hydroxycarbonylation of alkenes is a fundamental C C
bond forming reaction and, providing that regioselectivity
can be controlled, one of the most economical and clean
routes to carboxylic acids.[1] One of the most important ap-
plications of this reaction is to convert vinylarenes into
branched carboxylic acids, such as 2-arylpropanoic acids.
These compounds are useful building blocks, but also form
an important class of commercially applied anti-inflammato-
ry agents.[2] An industrial synthesis produces (S)-Naproxen
by regioselective hydroxycarbonylation, with classical reso-
[a] Dr. J. J. R. Frew, Dr. K. Damian, Prof. A. M. Z. Slawin,
Dr. M. L. Clarke
School of Chemistry, University of St Andrews
EaStCHEM, St Andrews
Fife, KY16 9ST (UK)
Fax : (+44)1334-463-808
E-mail: mc28@st-andrews.ac.uk
[b] Dr. H. V. Rensburg, Dr. R. P. Tooze
SASOL Technology Plc. (UK), Purdie Building
North Haugh, St Andrews
Fife, KY16 9ST (UK)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901178.
Scheme 1. Hydroxycarbonylation of aryl alkenes; effect of ligand coordi-
nation mode on regioselectivity in studies thus far.
10504
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10504 – 10513

FULL PAPER
cases no suitable chiral ligands have been found to induce
high enantioselectivity.[6]
can be directly converted into new secondary phosphine
boranes 3a and 3b (Scheme 2).
Recent studies on the methoxycarbonylation of ethylene
and octenes have revealed that the bulky diphosphine
ligand, 1,2-bis-(di-tertiarybutylphosphino)xylene gives palla-
dium catalysts that are more reactive than simple diphos-
phine–Pd catalysts by several orders of magnitude.[7,8]
A
publication reported while this work was in progress de-
scribes the use of this ligand to promote the formation of
the branched ester with ꢀ9:1 regioselectivity in the related
styrene methoxycarbonylation under a high CO pressure.[8]
Unfortunately, even small changes to either the tert-butyl
groups or the backbone of the phosphine had a very bad
effect on regioselectivity and reaction yield. Nonetheless, in-
vestigating the potential of very bulky (chiral) diphosphines
in styrene hydroxycarbonylation seemed to represent the
best opportunity to reach the long-desired goal of a highly
active regioselective and enantioselective catalyst for
branched selective hydroxycarbonylation.
In our preliminary communication,[9] we briefly described
the synthesis of PdII complexes of two new propyl-bridged
ligands and noted how they give improved regioselectivity
and yield relative to PdII complexes of the related 1,3-bis-
ACHTUNGTRENNUNG(tert-butylphosphino)propane and 1,3-bis(diphenylphosphi-
Scheme 2. Synthesis of new bulky symmetric diphosphane ligands.
no)propane ligands. Herein, we describe these studies in
full, including the synthesis of two new non-symmetric
propyl-bridged ligands, structural and catalytic studies on
the new Pd complexes, a study on promoter effects and deu-
terium labelling studies that shed light on the carbonylating
species in the reaction. It is hoped that these studies will
lead the way to a genuinely practical asymmetric hydroxy-
carbonylation reaction.
It was envisaged that the borane protection would facili-
tate deprotonation–alkylation with bisACTHNUTRGNE(UNG halide) electrophiles
without the problems of over-alkylation or E2 elimination
and would ease isolation because phosphine boranes are
generally stable to air, water and chromatography. For the
synthesis with secondary borane 2a, this strategy was com-
plicated by BH3 decomplexation problems, both in the de-
protonation step and during isolation, which necessitated
the direct isolation of the free phosphine because the phos-
phine borane of 1a was not stable. However, under the opti-
mised conditions, it proved possible to reproducibly obtain
the new symmetric diphosphine, rac/meso-1a, in good yields.
The para-substituted diphosphine diborane did not suffer
from this problem, and was stable enough for chromatogra-
phy under non-inert conditions, which allowed separation of
the meso and rac forms of the diphosphine diborane. Re-
Results and Discussion
Many research groups specialise in the study of the subtle
interplay between ligand stereo-electronic effects and cata-
lytic activity. Furthermore, bulky phosphines have become
increasingly popular in carbonylation catalysis.[7–10] It was,
therefore, surprising that examples of very bulky diphos-
phines in which the phosphorus is substituted by two bulky
groups with differing electronic properties were essentially
unstudied. We therefore investigated the synthesis of di-
phosphines of type 1. The starting materials for this type of
phosphine were not known. Fortunately, we found that
chlorophosphines 2a and 2b can be readily synthesised from
commercially available tBuPCl2 and that crude 2a and 2b
[11]
moval of the borane protecting group by using HBF4 gave
either diastereomer of diphosphine 1b in high purity.
To compare the catalytic performance of Pd complexes of
the symmetric diphosphines with non-symmetric ligands that
have PACHTUNGETNRU(NGN tBu)2 and PACHUTNGTERNNUG(tBu)AHCUTNGERTG(NUNN o/p-ACHTNUGRTEN(NUGN CF3)C6H4) groups, we investi-
gated the preparation of non-symmetric ligands 6a and 6b.
Although it was possible to synthesise the non-symmetric
structure by sequential displacements of BrACHTNUGTRNEG(UN CH2)3Cl with
the phosphine borane anions, a much better strategy is
shown in Scheme 3. The key intermediate in this synthesis,
4, is readily obtained by addition of Li[PACHTNUTRGENNUG(tBu)2ACHTUNTGRENN(UNG BH3)] to an
excess of 1,3-dibromopropane, and is stable to chromatogra-
phy and long-term storage in air. This new electrophilic
phosphine borane reagent should prove useful in the synthe-
sis of a wide range of ligand structures. Compound 4 reacted
Chem. Eur. J. 2009, 15, 10504 – 10513
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10505

M. L. Clarke et al.
diastereomer of the chelate complex, [PdACTHNUTRGENUG(N 1a)Cl2] (rac-7a)
could be obtained in pure form after several recrystallisa-
tions. Ligand rac-1b reacts cleanly and instantaneously to
give only the desired species, rac-7b, which was also isolated
in pure form after recrystallisation. The Pd complexes of the
non-symmetric ligands were obtained most conveniently
from [Pd
plex of 1,3-bis(tert-butylphosphino)propane was most con-
veniently prepared in 59% yield from [Pd2A(dba)3] and HCl
(PhCN)2] gave a mixture
2ACHTUNGTERN(NUNG dba)3] and HCl in moderate yield. The Pd com-
CTHUNGTRENNUNG
because direct reaction with [PdCl
of products.
2ACHTUNGTRENNUNG
The X-ray crystal structures of all five complexes were de-
termined (Figures 1, 2 and 3) and some of the more striking
structural features merit some discussion here. In the X-ray
structure of complex 9 (Figure 2), the unit cell contains two
distinct molecules. The arrangement of ligands around Pd(1)
is distorted from planar. The other molecule, containing
Pd(21), is severely distorted from planar. The bite angles are
almost identical for both molecules (97.91(8)8 and 97.73(8)8)
and are wider than for [PdCl2ACTHNUTRGNE(UNG dppp)] (90.58(5)8; dppp=1,3-
Scheme 3. Synthesis of bulky non-symmetric diphosphane ligands.
bis(diphenylphosphino)propane).[12] To compensate for this,
the Cl-Pd-Cl angles are significantly smaller that those
found in the dppp analogue (85.46(7)8 and 86.17(8)8 com-
pared with 90.78(5)8), which is likely to be an effect of the
more sterically demanding phosphorus substituents. Com-
plex rac-7a (Figure 1, top) has many structural features and
smoothly with the anions of 3a and 3b to give borane com-
plexes 5a and 5b. Once again, the ortho-substituted ligand
spontaneously lost one of the BH3 groups. The monoborane
5a and diborane 5b could be decomplexed by using HBF4
to give the desired non-symmetric diphosphines in high
yield. With the four new diphosphines in hand, we turned
our attention to synthesising their PdII complexes and com-
paring their structural features and catalytic properties.
Palladium(II) complexes of the four ligands and, as a con-
trol, 1,3-bis(tert-butylphosphino)propane were prepared
(Scheme 4). The coordination of more bulky ligand meso/
rac-1a is somewhat complicated by the appearance of side
products that give very broad NMR spectra, but a single rac
measurements that lie between those of [PdCl
[PdCl2A(dtbpp)] (dtbpp=1,3-bis(di-tert-butylphosphino)pro-
pane), as may be expected for this hybrid alkylaryl-substitut-
2ACHTUNGTRENN(UNG dppp)] and
CTHUNGTRENNUNG
AHCTUNGTRENNUNG
ed molecule. The geometry of the ligand around Pd is dis-
torted. One of the aromatic groups is oriented so the CF3
group is accommodated below the centre of the six-mem-
bered ring of the chelate. The other is oriented so the CF3
group points away from the metal centre. As a result of this,
the ortho protons on the aromatic ring are pushed towards
the palladium centre, with a short distance (2.64(1) ꢁ) possi-
bly indicating some weak electrostatic interaction (these
protons are also shifted significantly downfield compared
with the free ligand).[13] Thus the significant steric demand
of the ortho-trifluoromethylphenyl grouping is caused by the
bulky aromatic ring being forced nearer to the Pd centre
and the chloride ligands.
This effect is seen again in non-symmetric ortho-CF3-sub-
stituted complex rac-8a (Figure 3, top), but no similar effect
can be seen in either of the para-CF3 substituted phosphine
complexes, rac-7b and rac-8b (Figure 1, bottom and
Figure 3, bottom). The P-Pd-P angles for all four new com-
plexes are all significantly larger than those found for
[PdCl
2A
[PdCl2ACHTUNGTRN(ENUNG dtbpp)]. The structures reveal that all five ligands
are very bulky with steric demand increasing in the follow-
ing order: 7b !8b<7aꢀ8aꢀ9 (see the Supporting Infor-
mation for further data).
With the pure pre-catalysts in hand, we turned our atten-
tion to the hydroxycarbonylation of styrene. Palladium com-
plexes of the new ligands were compared against the com-
plexes other C3 diphosphines dppp and dtbpp (complex 9).
Scheme 4. Synthesis of palladium complexes used as pre-catalysts.
10506
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10504 – 10513

Palladium(II) Complexes of Bulky Bidentate Phosphanes
FULL PAPER
palladium catalysts under simi-
lar conditions is found in
Table 1, entries 1–7. It is strik-
ing that the catalytic activity
and regioselectivity towards
the branched isomer is sub-
stantially improved relative to
the two control systems utilis-
ing any of the new catalysts.
Both the para- and ortho-sub-
stituted ligands give good-to-
excellent regioselectivity and
improved yields relative to the
control ligands. A pronounced
feature is the effect of excess
ligand on these systems, which
can be examined by comparing
the results in Table 1, en-
tries 8–15. The best yields for
the catalysts derived from
ligand 1a were obtained in the
presence of excess ligand,
whereas adding excess of li-
gands 1b or 6b to their Pd
complexes almost completely
inhibited the reaction. There is
a
more modest inhibitory
effect for ligand 6a. The most
likely explanation for this is
the formation of a catalytically
inactive stable bischelate spe-
cies (Pd0 or Pd2+) that is only
formed for the less sterically
demanding ligands. The cata-
lyst systems function under
Figure 1. Molecular structures of rac-7a (top) and rac-7b (bottom).
milder conditions than are
often employed for diphos-
phine systems (1508C is often used in the literature stud-
ies).[4a] If the initial pressure is reduced to 30 bar (Table 1, en-
tries 21–24), improved results are obtained, particularly with
regard to regioselectivity, which is almost complete for cata-
lyst 8b. On the other hand, further reducing the pressure to
10 bar had a serious negative effect on both yield and regio-
selectivity for all ligands (Table 1, entries 17–20). Reducing
the temperature to 1008C at a CO pressure of 50 bar has a
negative effect on catalyst 8b (Table 1, entries 24 and 28),
but gives improved performance for the symmetrical diphos-
phine systems (Table 1, entries 8 and 10 vs. 25 and 26). The
optimised conditions from Table 1 use either pre-catalyst 7b
or 8b (Table 1, entries 26 and 24, respectively) under slightly
different conditions, and are, to the best of our knowledge,
the highest regioselectivity in a high yielding reaction under
the mildest conditions for a diphosphine-based catalyst. It is
clear that the two most active systems are derived from the
less bulky diphosphines that do not present their aryl ring
into the coordination sphere of palladium, although this fea-
ture has contrastingly been found to beneficial in other car-
Figure 2. Molecular structure of one of the independent molecules ob-
served within the asymmetric unit of complex 9.
Dppp catalysts have previously been reported to preferen-
tially give the linear acid under more forcing conditions
than we employed here. The conditions used herein are
more typical to those reported in studies that used mono-
phosphine catalysts and also include the LiCl and TsOH
promoters that have been found to be important in other
studies (Scheme 1 and Table 1).[14] The comparison of the six
Chem. Eur. J. 2009, 15, 10504 – 10513
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10507

M. L. Clarke et al.
nethyl chloride from styrene; Scheme 5). Our overall con-
clusions on reading all the published work by Chaudari and
co-workers is that the secondary alkyl chloride is certainly
present in the reactions, and a good number of their obser-
Scheme 5. Key differences between the hydride mechanism and the phe-
nethyl chloride mechanism.
vations point towards it being the active species, although
none of the experiments have proved this. Their most exten-
sive study on this additive effect also notes that the
carbonylACTHNUTRGENUGaN tion of the secondary alkyl chloride does not take
place with the same level of activity and selectivity unless
the H+/ClÀ promoter system is present and that vinylarenes
are also detected under these conditions. This suggests that
the role of the promoter may be more complicated, or
leaves open the possibility that the alkyl chloride is not the
carbonylating species because alkyl chloride could continu-
ously eliminate vinylarene and the HCl promoter at low
concentrations that are then carbonylated by the hydride
mechanism, which explains why alkyl chloride but not vinyl-
arenes can be carbonylated without acid present. Perhaps
surprisingly, the use of this specific promoter system has not
been studied with Pd complexes of bidentate ligands. The
most relevant report describes how the use of HCl as a pro-
moter can improve branched regioselectivity for Pd/PPh3
Figure 3. Top: Molecular structure of one of the three independent mole-
cules observed in the asymmetric unit of the X-ray structure of complex
rac-8a. Bottom: Molecular structure of complex rac-8b.
bonylation reactions.[13] Other hydroxycarbonylation experi-
ments (not shown) revealed that acceptable yields could still
be realised by using 0.5% catalyst, but the lower loadings
that would be required for industrial application give low
yields. Nonetheless, the results are a significant improve-
ment for diphosphine-based systems and open the possibility
that further development of bulky diphosphines could result
in practical catalysts for this long-desired transformation. It
is worth noting that the new ligands are chiral, although we
have not succeeded in isolating them in an enantiomerically
pure form. It is hoped that future studies by the catalysis
community into chiral bulky diphosphines may result in
useful enantioselective catalysts.
An interesting series of papers by Chaudari and co-work-
ers[14] report that for achiral Pd–monophosphine complexes,
improved branched regioselectivity can be realised by using
a specific combination of promoters, that is, LiCl and p-tolu-
ene sulfonic acid. It is known that the nature of the promot-
ers can have quite dramatic effects on this reaction, but this
group point out that the LiCl/ArSO3H combination may in-
volve a different and more regioselective reaction pathway
that proceeds by oxidative addition of the secondary alkyl
chloride that can be formed during the catalytic reaction by
systems, but has essentially no effect on
a Pd/dppb
system.[15] In this case, the role of HCl was proposed (but
not proven) to be in the formation of a mono-ligated Pd in-
termediate, which was more regioselective than Pd com-
plexes containing two phosphorus ligands. It is also possible
that the chloride conditions could have a third unknown
effect; for example, slowing the coordination and subse-
quent insertion of CO for the linear Pd–alkyl to a greater
degree than the branched isomer.
A study of various promoters was therefore carried out
(Table 2) by using the best symmetric catalyst, 7b. This
study does indeed show that the presence of acid and chlo-
ride is required to obtain high yields and regioselectivity. p-
Toluenesulfonic acid alone promotes the reaction reasonably
well, but reverses the regioselectivity towards the linear
acid. The results obtained with HCl are relatively similar to
those obtained with TsOH/LiCl (98.5 vs. 97.4% regioselec-
tivity), although productivity and regioselectivity is some-
what reduced if HCl and LiCl are used.
Most experiments cannot distinguish between the phe-
nethyl chloride and hydride mechanisms because the vinyl-
arene and alkyl chloride are in equilibrium (and the protons
at the =CH2/CH3 position can exchange during this process).
À
Markovnikov addition of H Cl to the vinylarene (e.g., phe-
10508
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10504 – 10513

Palladium(II) Complexes of Bulky Bidentate Phosphanes
FULL PAPER
Table 1. Hydroxycarbonylation of styrene by using the new palladium pre-catalysts.[a]
small amount of deuterated
acid, would be expected to be
the main product in the early
stages of the reaction, arising
from the deuterated styrene
formed from the elimination of
the deuterated (1-chloroethyl)-
benzene.
Entry
Pd pre-catalyst ([mol%])
[PdCl2A(dppp)] (2)
9 (2)
7a (2)
[Pd2A(dba3)] (2)
7b (2)
8a (2)
8b (2)
7a (1)
7a (1)
7b (1)
7b (1)
8a (1)
8a (1)
8b (1)
8b (1)
9 (1)
7a (1)
7b (1)
8a (1)
8b (1)
7a (1)
7b (1)
8a (1)
8b (1)
7a (1)
7b (1)
8a (1)
8b (1)
Ligand [mol%]
t [h]
T [8C]
Pressure [bar]
b/l
4.8
3.0
9.6
8.0
14.7
9.4
23.2
13.2
11.3
24.9
1.1
27.1
9.3
52.3
0.2
15.2
8.3
3.7
1.5
4.8
30.8
43.0
25.5
75.5
56.3
66.7
13.3
23.0
Yield[b] [%]
1
2
3
4
5
6
7
8
G
G
0
0
0
2
2
2
2
2
2
2
130
130
130
130
130
130
130
120
120
120
120
120
120
120
120
120
120
120
120
120
120
120
120
120
100
100
100
100
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
10
10
10
10
30
30
30
30
50
50
50
50
2
18
40
95
82
48
42
36
86
82
1
83
62
71
44
20
12
21
7
15
72
81
87
84
82
83
71
42
R
U
6 (1a)
0
0
0
0
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
The
hydroxycarbonylation
9
3 (1a)
0
3 (1b)
reaction was carried out by
using an equimolar mixture of
deuterated phenethyl chloride
([D]-10) and styrene (see
Table 3) by using [PdCl2-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
0
3
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
AHCTUNTGREG(NUNN PPh3)2] as the catalyst. The
more traditional triphenylphos-
phine-based catalyst was se-
lected because although we did
observe small amounts of phe-
nethyl chloride during some of
the carbonylations described in
Table 1, phenethyl chloride
was found to be a poor sub-
strate for the new catalysts.
Using the triphenylphosphane-
based system and a shortened
reaction times of 6 H; we eval-
uated the degree of deutera-
tion in the product acid at
moderate conversion at 1008C.
[a] Reactions were carried by using CO at the initial pressures (RT) given in butanone and water (2.5 equiv),
LiCl (20 mol%), TosOH (20 mol%) for the time described. A magnetic stirring bar was used. The b/l ratio
(b=branched isomer, 2-phenylpropanoic acid; l=linear isomer, 3-phenylpropanoic acid) was determined by
1H NMR spectroscopic analysis. [b] Yields are by weight after purification by acid-base extraction via the car-
boxylate salt of the product to remove catalyst and other organic impurities leaving the two isomers after re-
protonation.
Table 2. Effect of different promoters on the yield and regioselectivity of
Table 3. Hydroxycarbonylation labelling studies.[a]
the hydroxycarbonylation of styrene catalysed by pre-catalyst 7b.[a]
Entry
Promoter 1
Promoter 2
b/l[b]
Yield[c] [%]
1
2
3
4
5
TsOH
TsOH
none
HCl
LiCl
none
LiCl
LiCl
none
66.7
0.2
0.4
21.7
37.0
83
52
18
59
73
HCl
Entry Promoter 1 Promoter 2 b/l
deuteration in Yield [%]
product [%]
[a] Conditions: Pre-catalyst 7b (1 mol%), 50 bar initial pressure CO,
H2O (2.5 equiv), promoter (20 mol% each), 1008C, butanone, 16 h.
[b] The b/l ratios were determined by using 1H NMR spectroscopy.
[c] Yields are after purification by extraction (see Table 1, footnotes).
1
2
3
4
TsOH
TsOH
none
LiCl
none
LiCl
none
46.0
1.2
0[b]
0
26
18
n.d.[c]
20.0
0
0
trace
18
none
[a] Conditions were as shown. Reactions were analysed by using GCMS
under conditions that separate regio- and isotopomers, and by 1H and
13C NMR spectroscopic analysis of the product. [b] 0% refers to no deut-
eration detected in the product by GCMS, MS or NMR analysis.
[c] n.d.=not determined due to low yield.
However, if the phenethyl chloride mechanism were in oper-
ation, it can be envisaged that when a mixture of equimolar
amounts of styrene and deuterium-labelled phenethyl chlo-
ride ([D]-10) is carbonylated, then during the early stages of
the reaction the product would primarily arise from the la-
belled secondary alkyl chloride and thus produce mainly la-
belled acid. Even if the Markovnikov addition was extreme-
ly fast and there was an unprecedented and large negative
secondary kinetic isotope effect, the product acid should cer-
tainly contain at least as much labelled material as unla-
belled. On the other hand, if the hydride mechanism oc-
curred, the non-deuterated form of the acid, with possibly a
The three different combinations of the two promoters
TsOH and LiCl were investigated (Table 3, entries 1–3). The
reaction mixtures were analysed by GCMS and 1H NMR
spectroscopy, with the product acid also isolated at the end
of the six-hour experiments. This procedure led us to identi-
fy the presence of only small amounts of deuterated styrene
Chem. Eur. J. 2009, 15, 10504 – 10513
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10509

M. L. Clarke et al.
when the reaction was run in the presence of TsOH only
(Table 3, entry 2). The product formed did not show any of
the deuterium label, which led us to the conclusion that sty-
rene, not the deuterated alkyl chloride, was the species
being carbonylated under these conditions. Finally, we also
carried out the hydroxycarbonylation of styrene with no
promoters other than an equimolar amount of secondary
alkyl chloride. Hydroxycarbonylation of styrene did take
place in the presence of the deuterated phenethyl chloride
promoter (Table 3, entry 4; 18% yield of mainly branched,
completely undeuterated acid after 6 h). In this experiment,
small amounts of deuterated styrene were also detected
alongside much larger amounts of styrene, which presuma-
bly means that phenethyl chloride (which is the main com-
ponent in the product mixture) eliminates the true promot-
er, HCl and [D]styrene in small amounts during this reac-
tion. These experiments are therefore highly consistent with
the proposal that phenethyl chloride and styrene are in equi-
librium under the reaction conditions, but with only the sty-
rene being carbonylated. Our conditions are similar but not
identical to previous work on alkyl chloride carbonylation;
therefore, we suggest that it is still conceivable that pheneth-
yl chloride might act as the carbonylating substrate under
some different conditions, despite the envisaged high energy
for this oxidative addition that would also be strongly re-
tarded by carbon monoxide, but it is not in operation in any
of the reactions described here. In any case, the only mecha-
nism that seems to fit all the experimental observations and
is likely to represent a viable catalytic pathway is the hy-
dride mechanism.
selectivity. Studies along these lines are currently underway
in our laboratory.
Experimental Section
General experimental procedures and instrumentation: Routine NMR
spectroscopic data was recorded by using either a Bruker Avance 300 (1H
at 300 MHz, 13C at 75 MHz, 19F at 282 MHz, 31P at 121 MHz) or a Bruker
Avance II 400 (1H at 400 MHz, 13C at 100 MHz, 19F at 376 MHz, 31P at
1
161 MHz). H and 13C NMR spectra were referenced to external tetrame-
thylsilane, 19F NMR spectra were referenced to external trichlorofluoro-
methane, and 31P NMR spectra were referenced to external phosphoric
acid. Chemical shifts are expressed in parts per million. Chemical ionisa-
tion mass spectroscopy and electron ionisation mass spectroscopy were
performed by using a Micromass GCT spectrometer. Electrospray mass
spectroscopy was performed by using a Micromass LCT spectrometer.
All were operated by Mrs Caroline Horsburgh. Microanalysis was carried
out for C and H by using an EA 1110 CHNS CE Instruments elemental
analyser by Mrs S. Williamson. Infrared absorbsion spectra were record-
ed by using a Perkin–Elmer GX-FTIR System spectrometer. Thin-layer
chromatography was performed by using 0.20 mm layers of silica gel sup-
ported on plastic sheets (Macherey–Nagel, Polygram Sil G/UV254) or by
using 0.20 mm layers of aluminium oxide supported on plastic sheets
(Merck aluminium oxide F254). Preparative chromatography was per-
formed by using Davasil silica gel 35–70 mesh. Dry degassed diethyl
ether, petroleum ether, THF and toluene were obtained from an Innova-
tive Technologies Puresolve 400 solvent still. Other solvents were bought
and used as received without further purification other than degassing by
either purging with nitrogen or repeated freeze/thaw cycles under
vacuum. Organic solutions were dried by standing over anhydrous
sodium sulfate and evaporated either under reduced pressure on a rotary
evaporator or under reduced pressure with manual agitation. tert-Butyldi-
chlorophosphine was obtained from the Aldrich Chemical Company and
used as received. All manipulations were carried out under a nitrogen at-
mosphere unless otherwise stated. Full experimental details are available
in the supporting information.
Conclusions
Di-tert-butyl(3-bromopropyl)phosphine borane (4): Di-tert-butylphosphi-
noborane (3.00 g, 18.75 mmol) was dissolved in THF (60 mL) with stir-
ring and cooled to À788C. n-Butyllithium (7.6 cm3, 19 mmol as a 2.5m so-
lution in hexane) was then added dropwise before the solution was al-
lowed to warm to RT for complete conversion to the anion. The solution
was again cooled to À788C and 1,3-dibromopropane (8 mL, 78.4 mmol)
was rapidly added to avoid disubstitution. The temperature was main-
tained overnight until the dry ice was exhausted, at which point the sol-
vent and excess 1,3-dibromopropane were removed under vacuum. The
residue was partitioned between CH2Cl2/water and the organic phase was
washed twice with water before being dried over MgSO4, filtered and the
solvent removed under vacuum. Flash chromatography (10% diethyl
ether in hexane) gave the product in the second fraction as a white crys-
talline solid (4.15 g, 14.6 mmol, 78% yield). An analytical sample was
In summary, the synthesis of several new building blocks for
bulky phosphine synthesis have been reported and shown to
readily form a set of four structurally related bulky diphos-
phines. Palladium(II) complexes of these phosphines were
prepared and structurally characterised, which shows that
the para-trifluoromethylphenyl-substituted ligands are less
bulky than the known dtbpp system. The palladium com-
plexes were studied in the hydroxycarbonylation of styrene
and found to give extremely high regioselectivities and high
yields of the branched acid. Although all the catalysts show
significantly improved regioselectivity and massively im-
proved yields relative to simple diphosphines (e.g., dppp),
there is a clear trend in which the slightly less bulky diphos-
phines with para-trifluoromethylphenyl groups give im-
proved results relative to the ortho-substituted systems. The
role of the LiCl and TsOH promoters has also been investi-
gated; the presence of H+ and ClÀ is critical to the success-
ful results obtained and labelling studies seem to rule out
the phenethyl chloride mechanism. It is hoped that this
work can pave the way towards achieving an effective asym-
metric hydroxycarbonylation at low catalyst loadings be-
cause very bulky enantiopure diphosphines may offer the
desired combination of activity, regioselectivity and enantio-
prepared by recrystallisation from hexane at low temperature (À788C).
3
1H NMR (300 MHz; CDCl3): d=1.2 (d, J
(H,P)=12.5 Hz, 18H; 6
E
1.55 (m, 2H; CH2-P), 2.0 (m, 2H; C-CH2-C), 3.3 ppm (td, 3J
ACHTUNGTRENNUNG
4
1
7 Hz, J
1.2 (18H; 6
(t, 3J
ACHUTNGTREN(UNG H,P)=1 Hz, 2H; CH2-Br); HACHTNUGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
R
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
;
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
10510
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10504 – 10513

Palladium(II) Complexes of Bulky Bidentate Phosphanes
FULL PAPER
methane (10 mL) and stirred at RT, then [PdCl
G
13 Hz, ArC p-CF3),128.46 (m, ArC o-CF3), 39.05 (d, 2J
R
2
A
U
(CH3)), 38.78 (d, 2J
ACHUTGTNRENNUG CAHTUNGTRENNUNG
0.433 mmol) was added before the reaction was stirred overnight. The re-
action was then reduced to near dryness by removal of solvent under
vacuum, and then hexane was slowly added under atmospheric condi-
tions to form a precipitate that was isolated by filtration and washed with
more hexane. The precipitate was then redissolved and precipitated sev-
eral times to give the product as a orange powder (70 mg, 0.102 mmol,
23.5% yield). A sample suitable for X-ray analysis was prepared by slow
diffusion of hexane into a solution of the product in dichloromethane.
This showed the product to be present as the rac isomer. 1H NMR
A
N
ACHUTGTNERNNG(U CH3)), 31.32 (s, CACTHUNGTRENNNUG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
P(CACHTNUTGRNEUNG
(CH3)3)Ar); IR (KBr): n˜ =2900–3000, 2325, 1300 cmÀ1; ES+ MS: m/z
calcd for [MÀCl]+: 560.11; found: 560.15; elemental analysis calcd (%)
for C22H37Cl2F3P2Pd: C 44.20, H 6.24; found: C 44.45, H 6.47.
1-(Di-tert-butylphosphino)-3-{tert-butyl
phosphino}propane palladium dichloride (rac-8b)
Synthesis with [PdCl2A(PhCN)2]: 1-(Di-tert-butylphosphino)-3-{tert-butyl-
[para-(trifluoromethyl)phenyl]phosphino}propane (63 mg, 0.15 mmol)
was dissolved in CH2Cl2 (5 cm3) and [PdCl
2A(PhCN)2] (57 mg, 0.49 mmol)
ACHTUNGTRNE[NUNG para-(trifluoromethyl)phenyl]-
(300 MHz; CDCl3): d=1.4 (d, 3J
2.4 (m, 6H; 6CHH), 7.55 (m, 4H; 4ArH), 7.8 (m, 2H; 2ArH o-CF3),
9.05 ppm (brm, 2H; 2ArH o-P); 1H{31P} NMR (300 MHz; CDCl3): d=
1.4 (s, 18H; 6(CH3)), 1.95, 2.15, 2.4 (m, 6H; 6CHH), 7.55 (m, 4H;
ACHUTNGTREN(NUG H,P)=16 Hz, 18H; 6ACHTUNTGREN(NUNG CH3)), 1.95, 2.15,
CTHUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
CTHUNGTRENNUNG
4ArH), 7.8 (m, 2H; 2ArH o-CF3), 9.05 ppm (brm, 2H; 2ArH o-P);
19F NMR (282 MHz; CDCl3): d=À53.2 ppm (s); 31P{1H} NMR
(121 MHz; CD2Cl2): d=33.7 ppm (s); IR (KBr) n˜ =2900–3000, 2200,
1300 cmÀ1; ES+ MS: m/z calcd for [MÀCl]+: 651.06; found: 651.08 (100);
elemental analysis calcd (%) for C25H32Cl2F6P2Pd: C 43.78, H 4.70;
found: C 43.86, H 4.42.
was added. The solvent was reduced under vacuum and hexane was
slowly added to form a precipitate, which was filtered off and then redis-
solved in the minimum amount of CH2Cl2 and precipitated again, washed
with further hexane and dried under vacuum to yield the title compound
as an orange powder (47 mg, 0.078 mmol, 52% yield). A sample suitable
for X-ray analysis was prepared by slow diffusion of hexane into a solu-
1,3-Bis{tert-butyl
dium dichloride (rac-7b): rac-1,3-Bis{tert-butyl
phenyl]phosphino}propane (248 mg, 0.49 mmol) was dissolved in CH2Cl2
(10 mL) and [PdCl2A(PhCN)2] (188 mg, 0.49 mmol) was added. There fol-
ACHTUNGTRENNUNG
1
tion of the product in CH2Cl2. H NMR (400 MHz; CD3CN): d=8.35 (m,
AHCTUNGTRENNUNG
2H; ArH o-P), 7.71 (m, 2H; ArH o-CF3), 2.25 (m, 1H; CHH), 2.02 (m,
1H; CHH), 1.78 (m, 1H; CHH), 1.60 (m, 1H; CHH), 1.46 (d, 3J
(H,P)=
CHTUNGTRENNUNG
3
14.7 Hz, 9H; P(C
(CH3)3)Ar), 1.32 (d, 3J
1.20 ppm (m, 1H; CHH); 1H
AHCTUNGTRENNUNG
G
E
E
lowed an almost instant colour change from deep red to yellow. Hexane
was added and a precipitate formed that was filtered, washed with more
hexane and dried under vacuum to give the title compound as a bright
yellow crystalline powder (325 mg, 0.475 mmol, 97% yield). 1H NMR
G
(H,P)=14.3 Hz, 9H; P(C
U
ACHTUNGTRENNUNG
8.33 (m, 2H; ArH o-P), 7.73–7.70 (m, 2H; ArH o-CF3), 2.25 (m, 1H;
CHH), 2.02 (m, 1H; CHH), 1.78 (m, 1H; CHH), 1.60 (m, 1H; CHH),
(400 MHz; CD2Cl2): d=1.28 (d, 3J
m, 4H; CHH), 2.05 (2ꢂm, 2H; CHH), 7.71 (d, JACHTUNGTRENNNUG
ACHUTGTNREN(UNG H,P)=18 Hz, 18H; 6ACTHUNGTRENNNUG
3
1.46 (s, 9H; P(C
9H; P(C(CH3)3)(C
(101 MHz, CD3CN): d=136.00 (d, 2J
(dq, 3J(C,P)=10.2 Hz, 3J(C,F)=3.7 Hz, ArC o-CF3), 30.60 (d, 2J
ACHTUNGTRENNUNG
A
ACHTUTNGRENNU(G CH3)3)), 1.40 (s, 9H; P(CACHTUNGTRNE(NUGN CH3)3)Ar), 1.32 (s,
3
3
A
ACHTUNGTRENNUNG
o-CF3), 8.22 ppm (dd, J
A
ACHTUNGTRENNUNG(H,H)=8 Hz, JACHTNUGTRENNUNG
{31P} NMR (400 MHz; CD2Cl2): d=1.28 (s, 18H; 6
4H; CHH), 2.05 (2ꢂm, 2H; CHH), 7.70 (d, J
CF3), 8.21 ppm (d, 3J
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
3
A
G
ACHTUNGTRENNUNG
2
2.9 Hz, P(C
G
A
ACHTUGTNRENG(NU CH3)3)(CACHTUNGTRENNUNG
AHCTUNGTRENNUNG
29.49 (d, 2J
A
E
N
ACHTUNGTRENNUNG
3
3
A
U
24.9; J
G
ACHUTNGRENNUG CAHTUNGTRENNUNG
(C,P)=21.8, 1J
A
U
À63.64 ppm (s); 31P{H} NMR (162 MHz, CD3CN): d=41.42 (s),
31.21 ppm (s); IR (KBr): n˜ =3050, 2900–3000, 2325, 1300 cmÀ1; ES+ MS:
m/z calcd for [MÀCl]+: 561.10; found: 561.14; elemental analysis calcd
(%) for C22H37Cl2F3P2Pd: C 44.20, H 6.24; found: C 44.76, H 6.38.
A
ACHTUNGTRENNUNG
R
ACHTUNGTRENNUNG
CD2Cl2): d=28.10 ppm (s); IR (KBr): n˜ =2900–3000, 2250, 1350 cmÀ1
;
ES+ MS: m/z calcd for [MÀCl]+: 649.06; found: 649.15; elemental analy-
sis calcd (%) for C25H32Cl2F6P2Pd: C 43.78, H 4.70; found: C 43.82, H
4.45.
Synthesis with [Pd
2ACHUGTTRENNGU(dba)3]: [Pd2ACHTUNGTRNE(NUNG dba)3] (104 mg, 0.11 mmol) was added to
a stirred solution of 1-(di-tert-butylphosphino)-3-{tert-butylAHCTNUGTRENNNUG
[para-(trifluor-
omethyl)phenyl]phosphino}propane (96 mg, 0.22 mmol) in THF (10 mL).
After 1 hour, the dark orange solution was filtered and the solvent re-
moved under vacuum to give an orange oil. This was redissolved in dieth-
yl ether, then HCl (2.5 equiv) in diethyl ether was slowly added to form
the product as a yellow precipitate. After filtration the product was redis-
solved in CH2Cl2 and precipitated by the addition of hexane (49 mg,
38% yield). Crystals suitable for X-ray crystallography were obtained by
slow diffusion of hexane into a solution of the product in CH2Cl2.
1-(Di-tert-butylphosphino){3-(tert-butyl
phosphino}propane palladium dichloride (rac-8a): [Pd
0.23 mmol) was added to a stirred solution of 1-(di-tert-butylphosphi-
no){3-(tert-butyl[ortho-(trifluoromethyl)phenyl]phosphino}propane
(196 mg, 0.46 mmol) in THF (10 mL). After 1 hour, the dark orange solu-
tion was filtered and the solvent removed under vacum to give an orange
oil. This was redissolved in diethyl ether, then HCl (2.5 equiv) in diethyl
ether was slowly added to form the product as a yellow precipitate. After
filtration, the product was redissolved in CH2Cl2 and precipitated by the
addition of hexane (49 mg, 0.08 mmol, 18% yield). Crystals suitable for
X-ray crystallography were obtained by slow diffusion of hexane into a
solution of the product in CH2Cl2. 1H NMR (400 MHz; CD2Cl2): d=1.1
ACHTUNGTRENNUNG
CTHUNGTRENNUNG
ACHTUNGTRENNUNG
Example hydroxycarbonylation procedure: Lithium chloride (8.4 mg,
0.20 mmol), para-toluenesulfonic acid (34.4 mg, 0.20 mmol), rac-7a
(6.8 mg, 0.01 mmol) and 1,3-bis{tert-butylACTHNUGTRNEUNG[ortho-(trifluoromethyl)phenyl]-
phosphino}propane (15 mg, 0.03 mmol) were weighed into a Biotage 0.5–
2 mL microwave vial with a stirring bar, sealed with a crimp cap and
placed under an inert atmosphere. Styrene (114 mL, 1 mmol), water
(45 mL, 2.5 mmol) and degassed butan-2-one (1.5 mL) were added before
the cap was pierced with a capillary tube. The reaction tube was then
placed in an autoclave that was purged with CO three times and then
pressurized to 50 bar, then placed in a heating jacket and heated to
1108C. After 16 h the autoclave was cooled to RT and the pressure
slowly released. The solvent was then removed from the reaction mixture
and the residue dissolved in toluene and filtered to remove precipitate.
The filtrate was then extracted three times with saturated aqueous
NaHCO3 and the combined extracts acidified with conc. HCl until the
mixture remained decidedly acidic as tested by litmus paper. This was
then extracted three times with CH2Cl2, then the combined extracts were
(d, 3J
1.48 (m, 1H; CHH) 1.5 (d, 3J
CHH), 2.15 (m, 1H; CHH), 2.2 (m, 1H; CHH), 2.3 (m, 1H; CHH), 2.75
(m, 1H; CHH), 7.55 (t, 3J(H,H)=8 Hz, 1H; ArCH), 7.65 (t, 3J
(H,H)=
(H,P)=3 Hz, 1H; ArC o-
(H,H)=8.5 Hz, 1H; ArH o-P);
{31P} NMR (400 MHz; CD2Cl2): d=1.1 (s, 9H; (CH3)3), 1.35 (s, 9H;
G
ACHTUNGTREN(UNGN H,P)=16 Hz, 9H; (CH3)3),
AHCTUNGTRENNUNG
G
ACHTUNGTRENNUNG
3
4
8 Hz, 1H; ArCH), 7.75 (dd, J
CF3), 9.30 ppm (dd, 3J(H,P)=16 Hz, 3J
1H
ACHUTGTNREN(NUG H,H)=10 Hz, JACHTNUGTRENNUNG
G
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
(CH3)3), 1.48 (m, 1H; CHH) 1.5 (s, 9H; (CH3)3), 1.85 (m, 1H; CHH),
2.15 (m, 1H; CHH), 2.2 (m, 1H; CHH), 2.3 (m, 1H; CHH), 2.75 (m,
1H; CHH), 7.55 (t, 3J
1H; ArCH), 7.75 (d, 3J
(H,H)=8.5 Hz, 1H; ArH o-P); 13C{H} NMR (75 MHz; CD2Cl2): d=
ACHTUNGTRENNUNG ACHTUNGTRENNUNG
(H,H)=8 Hz, 1H; ArCH), 7.65 (t, 3J
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
144.39 (d, JACHTUNGTRENNUNG ACHTUNGTNER(NUGN C,P)=
(C,P)=22 Hz, ArC o-P), 133.69 (s, ArC) 133.34 (d, 3J
2
Chem. Eur. J. 2009, 15, 10504 – 10513
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10511

M. L. Clarke et al.
Table 4. Crystal data for compounds 7a, 7b, 8a, 8b, and 9.
Compound
7a
7b
8a
8b
9
formula
T [K]
C25H32Cl2F6P2Pd
93(2)
C25H31Cl2F6P2Pd·CH2Cl2
93(2)
C22H37Cl2F3P2Pd
93(2)
C22H37Cl2F3P2Pd
93(2)
C19H42Cl2P2Pd
125(2)
l [ꢁ]
0.71073
monoclinic
P21/n
8.8506(13)
34.101(5)
9.9169(13)
–
0.71073
monoclinic
P21/c
9.806(2)
15.204(3)
22.223(5)
–
0.71073
monoclinic
P21/c
15.0481(18)
14.4736(17)
33.621(4)
–
0.71073
orthorhombic
P2(1)2(1)2(1)
11.533(3)
14.130(3)
18.630(4)
–
0.71073
monoclinic
Cc
11.920(5)
28.733(11)
14.923(7)
–
crystal system
space group
a [ꢀ]
b [ꢀ]
c [ꢀ]
a [8]
b [8]
g [8]
113.885(4)
–
2736.7(7)
4
94.379(5)
–
3303.4(13)
4
98.383(6)
–
7244.4(15)
12
–
–
113.114(7)
–
4701(3)
8
volume [ꢁ3]
Z
3036.1(12)
4
1.308
1calcd [mgmÀ3
]
1.664
1.619
1.644
1.441
reflns collected
15475
18493
43470
19300
20580
indep. reflns
4773
1.045
0.0340/0.0721
0.0389/0.747
5765
1.114
0.1105/0.2783
0.1142/0.2804
13010
1.154
0.0959/0.1976
0.1157/0.2108
3051
0.918
0.1035/0.2552
0.1123/0.2622
8515
1.154
0.0513/0.0851
0.0730/0.0941
absorption coefficient [mmÀ1
R1/wR2 [I>2s(I)]
R1/wR2 (all data)
]
dried over MgSO4, filtered and the solvent was removed to give a viscous
product, a mixture of 2-phenylpropanoic and 3-phenylpropanoic acids.
2-Phenylpropionic acid: H1 NMR (400 MHz; CDCl3): d=1.45 (d, J=
7.2 Hz, 3H; CH3), 3.7 (q, J=7.2 Hz, 1H; CH), 7.2 ppm (m, 5H; ArH).
3-Phenylpropanoic acid: H1 NMR (400 MHz; CDCl3) d=2.6 (t, J=
7.7 Hz, 2H; CO-CH2), 2.9 (t, J=7.7 Hz, 2H; Ar-CH2) 7.2 (5H; m, ArH).
[4] a) I. del Rio, N. Ruiz, C. Claver, L. A. van der Veen, P. W. N. M. van
Leeuwen, J. Mol. Catal. A: Chem. 2000, 161, 39; b) S. Jayasree, A.
Seayad, R. V. Chaudhari, Chem. Commun. 2000, 1239; c) A. Seayad,
S. Jayasree, R. V. Chaudhari, Org. Lett. 1999, 1, 459; d) A. Ionescu,
G. Laurenczy, O. F. Wendt, Dalton Trans. 2006, 3934; e) M. S. Goed-
heijt, J. N. H. Reek, P. C. J. Kamer, P. W. N. M. Van Leeuwen, Chem.
Commun. 1998, 2431; f) interesting observations have recently been
made in the related methoxycarbonylation of styrene in which 3,5-
bis(trifluoromethyl)phenyl-substituted diphosphines gave greater
branched regioselectivity than phenyl phosphines (enantioselectivity
up to 30%), see: E. Guiu, M. Caporali, B. Munoz, C. Muller, M.
Lutz, A. L. Spek, C. Claver, P. W. N. M. Van Leeuwen, Organometal-
lics 2006, 25, 3102; g) some diphosphines have given ee values of up
to 86% in the related methoxycarbonylation of styrene but regiose-
lectivity favours the linear isomer, see: C. Goddard, A. Ruiz, C.
Claver, Helv. Chim. Acta 2006, 89, 1610.
X-ray crystallography
Crystallographic data: X-ray diffraction studies for 7a, 7b, 8a and 8b
were performed at 93 K by using a Rigaku MM007/Mercury diffractome-
ter (confocal optics MoKa radiation). Compound 9 was studied at 125 K
by using a Rigaku SCX Mini (graphite monochromator, MoKa radiation).
Intensity data were collected by using w steps accumulating area detector
frames spanning at least a hemisphere of reciprocal space for all struc-
tures (data were integrated by using CrystalClear[16]). All data were cor-
rected for Lorentz, polarisation and long-term intensity fluctuations. Ab-
sorption effects were corrected on the basis of multiple equivalent reflec-
tions. Structures were solved by direct methods and refined by full-
matrix least-squares against F2 (SHELXTL[17]) Hydrogen atoms were as-
signed riding isotropic displacement parameters and constrained to ideal-
ised geometries. Crystal data for 7a, 7b, 8a, 8b and 9 are shown in
Table 4. CCDC-728877 (7a), CCDC-728878 (7b), CCDC-728879 (8a),
CCDC-728880 (8b) and CCDC-728881 (9) contain the supplementary
crystallographic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[5] A chiral phosphonic acid additive was communicated to give very
good ee values in hydroxycarbonylation, but unfortunately yields
were moderate and over 10% catalyst was required, see: H. Alper,
N. Hamel, J. Am. Chem. Soc. 1990, 112, 2803.
[6] a) Monodentate phosphetanes in a related methoxycarbonylation
give a b/l ratio of 98:2 (97% yield), but an ee of only 12%, see: B.
Munoz, A. Marinetti, A. Ruiz, S. Castillon, C. Claver, Inorg. Chem.
Commun. 2005, 8, 1113; b) monophosphines in a related methoxy-
carbonylation can give an ee of 86% but a b/l ratio of only ꢀ1:1
(17% conversion), see: S. Oi, M. Nomura, T. Aiko, Y. Inoue, J. Mol.
Cat. A, 1997, 115, 289; c) a monophosphine in a related methoxycar-
bonylation gives an ee of 52% and a b/l ratio of 94:6, see: G. Com-
etti, G. P. Chiousoli, J. Organomet. Chem. 1982, 236, C31. There are
very few reports of asymmetric hydroxycarbonylation of aryl al-
kenes, see refs. [1] and [4a].
Acknowledgements
[7] a) W. Clegg, G. R. Eastham, M. R. J. Elsegood, R. P. Tooze, X. L.
Wang, K. Whiston, Chem. Commun. 1999, 1877; b) C. Jimenez Ro-
driguez, D. F. Foster, G. R. Eastham, D. J. Cole-Hamilton, Chem.
Commun. 2004, 1720.
[8] H. Ooka, T. Inoue, S. Itsuno, M. Tanaka, Chem. Commun. 2005,
1173.
The authors wish to thank SASOL(UK) and EPSRC for funding, Mr
Peter Pogorzelec, Mrs Malanya Smith, Dr Tomas Lebl, Mrs Caroline
Horsborough and Sylvia Williamson for technical assistance, and the
EPSRC National Mass Spectrometry Service. Chirotech Technology Ltd.
and Dr Chris Cobley are also thanked for their interest in this work and
financial support of K.D. as part of another project.
[9] J. J. R. Frew, M. L. Clarke, U. Mayer, H. van Rensburgh, R. P.
Tooze, Dalton Trans. 2008, 1976.
[10] For other selected examples of bulky phosphorus ligands that give
unusual selectivity and rates in carbonylation, see: a) R. I. Pugh, E.
Drent, P. G. Pringle, Chem. Commun. 2001, 1476; b) M. L. Clarke,
G. J. Roff, Chem. Eur. J. 2006, 12, 7978; c) M. L. Clarke, G. J. Roff,
Green Chem. 2007, 9, 792; d) B. Breit, R. Winde, T. Mackewitz, R.
Paciello, K. Harms, Chem. Eur. J. 2001, 7, 3106; e) A. van Rooy,
[1] I. del Rio, C. Claver, P. W. N. M. van Leeuwen, Eur. J. Inorg. Chem.
2001, 2719.
[2] J-P Rieu, A. Boucherle, H. Cousse, G. Mouzin, Tetrahedron 1986,
42, 4095.
[3] For a review of industrial uses of Pd catalysts, including Naproxen
synthesis, see: A. Zapf, M. Beller, Top. Catal., 2002, 19, 101.
10512
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10504 – 10513

Palladium(II) Complexes of Bulky Bidentate Phosphanes
FULL PAPER
E. N. Orij, P. C. J. Kamer, P. W. N. M. van Leeuwen, Organometallics
1995, 14, 34–43; f) B. Breit E. Fuchs, Chem. Commun. 2004, 694–
695; g) C. J. Rodriguez, D. F. Foster, G. R. Eastham, D. J. Cole-Ham-
ilton, Chem. Commun. 2004,1720.
1067; c) S. Jayasree, A. Seayad, R. V. Chaudhari, Chem. Commun.
2000, 1239; d) S. Jayasree, A. Seayad, R. V. Chaudhari, Catal. Lett.
1999, 61, 99.
[15] I. del Rio, N. Ruiz, C. Claver, Inorg. Chem. Commun. 2000, 3, 166.
[16] CrystalClear version 1.36, Rigaku Corporation, Tokyo, 2004.
[17] G. M. Sheldrick, Acta Crystallogr. Sect. A, 2008, 64, 112
[11] L. McKinstry, T. Livinghouse, Tetrahedron 1994, 50, 9319.
[12] W. L. Steffen, G. J. Palenik, Inorg. Chem. 1976, 15, 2432.
[13] C. Bianchini, A. Meli, W. Oberhauser, C. Claver, E. J. Garcia
Suarez, Eur. J. Inorg. Chem. 2007, 2702.
[14] a) A. Seayad, S. Jayasree, R. V. Chaudhari, Org. Lett. 1999, 1, 459;
b) S. Jayasree, A. Seayad, R. V. Chaudhari, Chem. Commun. 1999,
Received: May 5, 2009
Published online: September 8, 2009
Chem. Eur. J. 2009, 15, 10504 – 10513
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10513