Intermolecular alkene and alkyne hydroacylation with β-S-substituted aldehydes: Mechanistic insight into the role of a hemilabile P-O-P ligand

Article abstract of DOI:10.1002/chem.200800738

A straightforward to assemble catalytic system for the intermolecular hydroacylation reaction of β-S-substituted aldehydes with activated and unactivated alkenes and alkynes is reported. These catalysts promote the hydroacylation reaction between β-S-substituted aldehydes and challenging substrates, such as internal alkynes and 1-octene. The catalysts are based upon [Rh(cod)(DPEphos)][ClO₄] (DPE-phos = bis(2-diphenylphosphinophenyl) ether, cod = cyclooctadiene) and were designed to make use of the hemilabile capabilities of the DPEphos ligand to stabilise key acyl-hydrido intermediates against reductive decarbonylation, which results in catalyst death. Studies on the stoichiometric addition of aldehyde (either ortho-HCOCH₂CH ₂SMe or ortho-HCOC₆H₄SMe) and methylacrylate to precursor acetone complexes [Rh(acetone)₂(DPEphos)][X] [X = closo-CB₁₁H₆Cl₆ or [BAr₄ ^F] (Ar^F = 3,5-(CF₃)₂C ₆H₃)] reveal the role of the hemilabile DPEphos ligand. The crystal structure of [Rh(acetone)₂(DPEphos)][X] shows a cis-coordinated diphosphine ligand with the oxygen atom of the DPEphos distal from the rhodium. Addition of aldehyde forms the acyl hydride complexes [Rh(DPEphos)(COCH₂CH₂SMe)H][X] or [Rh(DPEphos)(COC ₆H₄SMe)H][X], which have a trans-spanning DPEphos ligand and a coordinated ether group. Compared to analogous complexes prepared with dppe (dppe = 1,2-bis(diphenylphosphino)ethane), these DPEphos complexes show significantly increased resistance towards reductive decarbonylation. The crystal structure of the reductive decarbonylation product [Rh(CO)(DPEphos) (EtSMe)][closo-CB₁₁H₆I₆] is reported. Addition of alkene (methylacrylate) to the acyl-hydrido complexes forms the final complexes [Rh(DPEPhOs)(η¹-MeSC₂H₄- η¹-COC₂H₄CO₂Me)][X] and [Rh-(DPEphos)(η¹-MeSC₆H₄-η¹- COC₂H₄-CO₂Me)][X], which have been identified spectroscopically and by ESIMS/MS. Intermediate species in this transformation have been observed and tentatively characterised as the alkyl-acyl complexes [Rh(CH₂CH₂CO₂Me)-(COC₂H ₄SMe)(DPEphos)][X] and [Rh(CH₂CH₂CO ₂Me)(COC₆H₄SMe)-(DPEphos)][X]. In these complexes, the DPEphos ligand is now cis chelating. A model for the (unobserved) transient alkene complex that would result from addition of alkene to the acyl-hydrido complexes comes from formation of the MeCN adducts [Rh(DPEphos)(MeSC₂H₄CO)H(MeCN)] [X] and [Rh(DPEphos)(MeSC₆-H₄CO)H(MeCN)][X]. Changing the ligand from DPEphos to one with a CH₂ linkage, [Ph₂P(C ₆H₄)]₂CH₂, gave only decomposition on addition of aldehyde to the acetone precursor, which demonstrated the importance of the hemiabile ether group in DPEphos. With [Ph₂P(C ₆H₄)]₂S, the sulfur atom has the opposite effect and binds too strongly to the metal centre to allow access to productive acetone intermediates.

Full text of DOI:10.1002/chem.200800738

FULL PAPER
DOI: 10.1002/chem.200800738
Intermolecular Alkene and Alkyne Hydroacylation with b-S-Substituted
Aldehydes: Mechanistic Insight into the Role of a Hemilabile P–O–P Ligand
Gemma L. Moxham,[a] Helen Randell-Sly,[a] Simon K. Brayshaw,[a]
Andrew S. Weller,*[a, c] and Michael C. Willis*[a, b]
Abstract: A straightforward to assem-
ble catalytic system for the intermolec-
ular hydroacylation reaction of b-S-
substituted aldehydes with activated
and unactivated alkenes and alkynes is
reported. These catalysts promote the
hydroacylation reaction between b-S-
substituted aldehydes and challenging
substrates, such as internal alkynes and
1-octene. The catalysts are based upon
atom of the DPEphos distal from the
rhodium. Addition of aldehyde forms
the acyl hydride complexes [Rh-
MS. Intermediate species in this trans-
formation have been observed and ten-
tatively characterised as the alkyl–acyl
A
complexes
(COC2H4SMe)
[Rh(CH2CH2CO2Me)(COC6H4SMe)-
(DPEphos)][X]. In these complexes,
[Rh(CH2CH2CO2Me)-
R
A
ACHTRE(UNG DPEphos)][X] and
which have a trans-spanning DPEphos
ligand and a coordinated ether group.
Compared to analogous complexes pre-
pared with dppe (dppe=1,2-bis(diphe-
nylphosphino)ethane), these DPEphos
complexes show significantly increased
resistance towards reductive decarbon-
ylation. The crystal structure of the re-
AHCTREUNG
the DPEphos ligand is now cis chelat-
ing. A model for the (unobserved)
transient alkene complex that would
result from addition of alkene to the
acyl–hydrido complexes comes from
formation of the MeCN adducts [Rh-
[Rh
A
(DPEphos)]
A
(DPE-
phos=bis(2-diphenylphosphinophenyl)-
ether, cod=cyclooctadiene) and were
designed to make use of the hemilabile
capabilities of the DPEphos ligand to
stabilise key acyl–hydrido intermedi-
ates against reductive decarbonylation,
which results in catalyst death. Studies
on the stoichiometric addition of alde-
hyde (either ortho-HCOCH2CH2SMe
or ortho-HCOC6H4SMe) and methyl-
ductive
decarbonylation
product
A
ACHTRE(UNG MeCN)]
[Rh(CO)
N
N
ACHTREUNG
CB11H6I6] is reported. Addition of
alkene (methylacrylate) to the acyl–hy-
drido complexes forms the final com-
ACHTREUNG
AHCTREUNG
plexes [Rh
COC2H4CO2Me)][X]
AHCTREUNG
(DPEphos)(h1-MeSC6H4-h1-COC2H4-
A
only decomposition on addition of al-
dehyde to the acetone precursor, which
demonstrated the importance of the
hemiabile ether group in DPEphos.
and
[Rh-
A
CO2Me)][X], which have been identi-
fied spectroscopically and by ESIMS/
plexes
[Rh
(acetone)
U
With [Ph2PACHRTEU(GN C6H4)]2S, the sulfur atom
F
[X=closo-CB11H6Cl6 or [BArF ] ( Ar
has the opposite effect and binds too
strongly to the metal centre to allow
access to productive acetone intermedi-
ates.
= 3,5-(CF3)2C6H3)] reveal the role of
the hemilabile DPEphos ligand. The
Keywords:
heterogeneous catalysis · hydroacy-
lation phosphane complexes
rhodium
decarbonylation
·
crystal structure of [Rh
(DPEphos)][X] shows a cis-coordinat-
ed diphosphine ligand with the oxygen
ACHTREU(GN acetone)2-
·
·
ACHTREUNG
[a] Dr. G. L. Moxham, Dr. H. Randell-Sly, Dr. S. K. Brayshaw,
Dr. A. S. Weller, Dr. M. C. Willis
Department of Chemistry, University of Bath
Bath, BA2 7AY (UK)
[c] Dr. A. S. Weller
Department of Chemistry, University of Oxford
Inorganic Chemistry Laboratories, Oxford, OX1 3QR (UK)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800738. It contains full exper-
imental and characterisation data for all the new complexes and the
Fax : (+44)1865272690
E-mail: andrew.weller@chem.ox.ac.uk
michael.willis@chem.ox.ac.uk
solid-state structure of 1ACTHRE[UNG CbBr6].
[b] Dr. M. C. Willis
Department of Chemistry, University of Oxford
Chemistry Research Laboratories
Oxford, OX1 3TA (UK)
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8383

Introduction
Morehead and Sargent support this mechanism and also
provide insight into the reasons why decarbonylation from
intermediates, such as C is the less-dominant route under
certain conditions of solvent and substrate concentration.[3]
Brookhart has studied intermolecular hydroacylation by
using [M(h5-C5Me4X)(h2-H2C=CHSiMe3)2] complexes (X=
Me, CF3; M=Co[4] Rh[5]) with the reductive elimination of
the ketone again the turnover-limiting step (J!F). He also
showed that judicious functionalisation of the cyclopenta-
dienyl ligand with electron-withdrawing groups can promote
the rate of reductive elimination.[5] The organometallic prod-
ucts of the reductive decarbonylation have been identified,
The transition-metal-catalysed hydroacylation of alkenes
and alkynes with aldehydes is an example of a process that
ꢀ
ꢀ
combines efficient C H activation coupled with C C bond
formation, and yields a product with an attractive ketone
functionality [Eq. (1)].[1] Although it has attracted significant
attention over the last twenty years, the reaction has not en-
joyed widespread synthetic use due to the limited set of sub-
strates that can be used. This restriction is largely due to the
competing reaction to hydroacylation: metal-mediated re-
ductive decarbonylation. This is problematic in that it is an
irreversible process (an alkane is lost) forming metal–car-
bonyl complexes that are generally inactive or at the very
best poor hydroacylation catalysts. A major aspect of the de-
velopment of hydroacylation as a viable synthetic protocol
has been efforts directed to suppress reductive decarbonyla-
tion.
respectively, as cationic [Rh(CO)
[M
(h5-C5Me5)(CO)]2 (M=Co, Rh) in each system. Both are
(dppe)]+ and the dimer
2ACHTREUNG
AHCTREUNG
inactive in the hydroacylation cycle.
The non-productive catalyst decomposition pathway of re-
ductive deacarbonylation requires two processes to occur:
deinsertion of the carbonyl group in the acyl hydride to
form an alkyl or aryl hydrido carbonyl, followed by reduc-
tive elimination of an alkane. Both processes require coordi-
natively unsaturated metal centres. Deinsertion of a carbon-
yl requires a vacant site cis to the acyl (Scheme 2a),[6,7]
whereas reductive elimination of alkanes in neutral or cat-
ionic six coordinate d6 complexes also requires ligand disso-
ciation prior to the rate-determining bond-formation step
(Scheme 2b).[8,9] Brookhart has shown that carbonyl deinser-
tion is reversible in certain hydroacylation systems, and
indeed the metal–carbonyl is a resting state for the catalytic
cycle (I, Scheme 1),[4,5] whereas reversible decarbonylation
has been used to account for deuterium scrambling in inter-
Detailed mechanistic studies of the hydroacylation reac-
tion remain scarce. An early report by Bosnich on intramo-
(dppe)]+”[2] catalyst
lecular hydroacylation by using the “[RhACHTREUNG
(dppe=1,2-bis(diphenylphosphino)ethane) and 4-pentenal
used deuterium-labelling studies to suggest a likely mecha-
nism (Scheme 1), although no intermediates were definitive-
ly characterised spectroscopically. Reductive elimination of
the ketone product was suggested to be the turnover limit-
ing step (D!A, Scheme 1). Very recent calculations by
(dppe)]+” catalys-
molecular hydroacylation using the “[RhACHTREUNG
t.[2a,3,10] Thus, it is the irreversible reductive elimination of
the alkane that results in ultimate catalyst death. Given that
Scheme 1. Abbreviated catalytic cycles for the RhI-catalysed hydroacylation of alkenes. The catalytically inactive organometallic products of decarbony-
lation that is competitive with the hydroacylation cycle are also shown. O.A.=oxidative addition, R.E.=reductive elimination. The turnover-limiting
step in both cycles has been determined to be the reductive elimination of the ketone product. Anions have been omitted from the left-hand cycle.
8384
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8383 – 8397

Intermolecular Alkene Hydroacylation
FULL PAPER
alkyl aldehydes can be used under mild conditions, several
limitations still remain, the most significant of which is that
alkynes or electron-poor alkenes are needed to achieve
good reactivity. We postulated that by removing decomposi-
tion pathways open to the catalyst, such as reductive decar-
bonylation, longer-lived catalysts would result that would be
able to couple more challenging substrates. Given that many
of the catalytic intermediates in Scheme 1 have open coordi-
nation sites and are electronically unsaturated, we also
postulated that stability towards reductive decarbonylation
might be deliberately engineered into the catalyst by provid-
ing latent stabilisation from a hemilabile ligand (Scheme 3).
This would result in a similar scenario to that observed with
excess substrate but would provide more control without re-
lying on substrate concentration.
Scheme 2. Carbonyl deinsertion (a) and alkane reductive elimination (b).
rate-determining reductive elimination of the product, sub-
strate binding and reductive decarbonylation all require co-
ordinative unsaturation,[9] the challenge is to establish how
productive hydroacylation can be engineered to be favoura-
ble over reductive decarbonylation.
Chelation control[11] in intramolecular hydroacylation re-
actions with 4-pentenals is one answer. This approach was
initially reported by Milstein,[12] Bosnich[2a,13] and James[14]
and then developed into asymmetric variants.[15] Intramolec-
ular hydroacylation has also been reported on other, related,
systems.[16] Scheme 1 shows the catalytic scheme for the Bos-
nich system, in which chelation suppresses reductive decar-
bonylation to such an extent that reasonable catalytic turn-
over may be achieved. However the fact that this reaction
appears specific for the formation of cyclopentanones and is
not intermolecular limits its utility. Chelation control has
also been used for intermolecular hydroacylation, which has
greater synthetic utility. First noted by Suggs[1a] and devel-
oped by others,[17–20] the control arises from constraining the
acyl within a five-membered metallacycle, which suppresses
decarbonylation due to the necessity of this process forming
a strained, four-membered, ring. Decarbonylation is also at-
tenuated by coordinating solvent or substrate that can tem-
porarily block the vacant site needed for decarbonylation.
Miller noted that the addition of excess ethene to the
Scheme 3. Suggested hemilabile-ligand protection for the hydroacylation
catalyst system.
Hemilabile ligands[22] can potentially provide latent coor-
dinate and electronic protection during catalysis.[23,24] Our
approach, within the context of the hydroacylation reaction,
is to use a ligand that would weakly stabilise key coordina-
tively unsaturated intermediates but move away to allow ap-
proach of substrates to the metal centre during the appropri-
ate part(s) of the cycle. The ligand should also be conforma-
tionally flexible enough to allow the metal centre to cycle
through a number of geometries/oxidation states during cat-
alysis (e.g. Scheme 1). A survey of possible ligands identified
DPEphos (DPEphos=bis(2-diphenylphosphinophenyl)eth-
er) as a promising initial candidate as it has been established
to behave in a hemilabile manner and can support a range
of metal geometries.[25,26] In addition to DPEphos stabilising
key unsaturated intermediates, this relatively large bite-
angle ligand might also have an effect in promoting the
turnover limiting reductive elimination step.[25]
This paper details our mechanistic studies into the use of
DPEphos, and related, cationic complexes of RhI in the hy-
droacylation reactions with b-S-substituted aldehydes. Inter-
mediates on the catalytic hydroacylation cycle have been
identified by NMR spectroscopy, ESIMS and X-ray crystal-
lography. The decarbonylation pathways available to some
of these intermediates have also been studied, and the influ-
ence that the hemilabile ligand, the aldehyde and the coun-
terion have on complex stability and catalytic turnover are
all explored. We also report the development of a practica-
ble catalyst for selected, challenging, intermolecular hydroa-
cylation reactions. We have not performed a detailed kinetic
study on these systems and the results presented are a quali-
tative overview of the catalytic cycle. Aspects of this work
have been communicated.[19]
[RhClACHTREUNG(PPh3)3]-catalysed intramolecular hydroacylation of 4-
pentenal increased the yield of desired product,[21a,b] and
similar effects have been noted for intermolecular varia-
tions,[21c] whereas Bosnich found that excess substrate can
retard the reaction while also extending the lifetime of the
“[Rh
A
inhibition and resistance to decarbonylation occur by block-
ing a vacant site. Recent computational work presents a
slightly differing view and suggests that substrate binding to
intermediates such as D (Scheme 1) actually facilitates re-
ductive elimination of the product by preferentially lowering
the barrier to this process relative to that of decarbonyla-
tion.[3] The presence of excess substrate has also been shown
to halt reductive decarbonylation in systems based upon
[Co
Recent work by one of our groups[18,20] has made use of
the Bosnich [Rh(acetone)2A
(dppe)]+ catalyst system com-
ACHTREUNG
(h5-C5Me5)(h2-H2C=CHSiMe3)2].[4]
A
CHTREUNG
bined with chelate control with b-S-substituted aldehydes to
deliver intermolecular hydroacylation reactions. Although
this method has advantages over previous protocols, in that
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8385

A. S. Weller, M. C. Willis et al.
Results and Discussion
Evaluation of precatalysts in a benchmark hydroacylation
reaction: We selected the combination of b-MeS-substituted
propanal (I) and methyl acrylate as the benchmark hydro-
acylation process and used this reaction to evaluate precata-
To explore the role of the hemilabile ligand, we chose to
study the intermolecular hydroacylation reaction between
the activated alkene methyl acrylate and the b-substituted
aldehydes 3-(methylthio)propionaldehyde I or 2-(methyl-
thio)benzaldehyde II to afford the ketones III and IV, re-
spectively [Eq. (2)]. The entry point into our catalytic sys-
lysts [Rh
and [Rh(PCP)
[Rh(dppe)(nbd)]
R
ACHTRE(UGN nbd)]ATHCRE[UNG ClO4], [RhACHRET(UNG PSP)ACHTER(UNG nbd)]ACHTREUNG[ClO4]
A
N
ACHTREUNG
A
N
ACHTREUNG
tems are salts of the general formula [Rh(L2)ACHTRE(UGN nbd)][X]
Table 1. Benchmark catalyst evaluation.[a]
(nbd=norbornadiene) in which L2 is a suitable, hemilabile
ligand and [X]ꢀ is a non-coordinating anion, such as [closo-
CB11H6X6]ꢀ
(X=halogen),[27,28]
[BArF ]ꢀ
(ArF =3,5-
4
(CF3)2C6H3), [PF6]ꢀ or [ClO4]ꢀ. For the purposes of this
Entry
Precatalyst
Conversion [%][b]
t
study, we have principally used the [closo-CB11H6X6]ꢀ (X=
1
2
3
[Rh
[Rh
[Rh
[Rh
C
G
E
100[c]
100
0
0
100
90 min
60 min
48 h
48 h
90 min
Cl, Br, I) and [BArF ]ꢀ anions as these often readily offer
4
G
ACHTREUNG
crystalline materials and can be used interchangeably for
the stoichiometric mechanistic studies. From a practical
viewpoint, the [ClO4]ꢀ anion serves well in catalytic applica-
tions (see sections Evaluation of precatalysts in a bench-
mark hydroacylation reaction and Applications to organic
synthesis—representative examples), although we show
there is a notable counterion effect on the overall rate of
catalysis. The carborane anions all show 11B{1H} NMR spec-
tra that indicate that they are not bound to the metal cen-
tres.[28] The phosphine ligands used for this study are DPE-
phos, PSP and PCP, and were chosen for the variation of
donor properties (O, S or CH2) of the hemilabile ligands.
A
ACHTREUNG
4
G
G
5[d]
[RhCl
N
A
[a] Conditions: aldehyde (1.0 equiv), methyl acrylate (2.0 equiv), catalyst
(5 mol%), acetone, 558C. Catalysts were generated from the correspond-
ing precatalysts by hydrogenation (1 atm, 5 min, acetone). See section en-
titled synthesis of precatalysts for details of their synthesis. [b] Deter-
mined by 1H NMR spectroscopy. [c] The product was obtained as a 4:1
mixture of linear/branched isomers. In all other reactions, the product
was formed exclusively as the linear compound. [d] Catalyst generated
from the simple combination of the three components.
ꢀ ꢀ
ꢀ
ꢀ
ꢀ ꢀ
Replacing
O
with CH2 or
S
groups forms ligands
erated by adding H2, elimination of norborane and forma-
tion of the corresponding acetone adducts, [Rh(L2)-
ACHRTE(UNG acetone)2]ACHTRE[UGN ClO4]. The reaction with the Bosnich catalyst
with similar bite angles but very different hemilabile proper-
ties.
took 90 minutes to achieve 100% conversion (entry 1). Em-
ploying the DPEphos-derived precatalyst achieved the same
conversion after 60 minutes (entry 2). Monitoring the reac-
1
tions by H NMR spectroscopy confirmed this difference in
rate. Reactions employing the thio- (PSP) and methylene-
bridged (PCP) ligands failed to deliver any of the expected
product (entries 3 and 4). The final entry (entry 5) demon-
strated that the simple combi-
nation of [RhCl
ACHTRE(UGN cod)]2, DPE-
phos and Ag[ClO4] generated a
AHCTREUNG
catalyst that achieved complete
conversion in 90 minutes
(entry 5). 31P{1H} NMR spec-
troscopy of this in situ generat-
ed catalyst showed the forma-
tion of [Rh
[ClO4], which on addition of al-
dehyde formed the active species (6a[ClO4], vide infra).
ACHRTUNEG(cod)AHCTRE(UGN DPEphos)]-
AHCTREUNG
We first briefly report the initial screening of hemilabile
ligand sets in a benchmark hydroacylation reaction that
demonstrates the potential of the DPEphos ligand. We then
move on to discuss in detail the synthesis, characterisation
and reactivity of species directly relevant to the catalytic
cycle by stoichiometric addition of aldehyde and alkene to
suitable DPEphos-containing precatalyts. Finally, the utility
of these DPEphos catalysts in challenging hydroacylation re-
actions is demonstrated.
ACHTREUNG
From a practical perspective, the ability to avoid hydrogena-
tion of a precatalyst significantly adds to the utility of the
system (see the section Applications to organic synthesis—
representative examples). After having established that the
DPEphos-derived system offered an advantage in the
benchmark hydroacylation reaction, we moved on to a
series of stoichiometric qualitative mechanistic studies in
order to understand the basis for this reactivity and especial-
ly the role of the hemilabile ligand—if any.
8386
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8383 – 8397

Intermolecular Alkene Hydroacylation
FULL PAPER
Synthesis of precatalysts: For the stepwise mechanistic
study, well-defined and pure precursor complexes were re-
quired. So although from a practical viewpoint (see the sec-
tion Applications to organic synthesis—representative ex-
amples) catalysts are best prepared in situ from [RhCl-
C
G
[BF4] has been reported previously.[29] The
(DPEphos)]ACHTREUNG
AHCTREUNG
AHCTREUNG
ACHTREUNG
AHCTREUNG
A
displays the expected square-planar motif, with no close
ꢀ
Rh···O
contacts
(Rh O
3.523 Š) (see the Supporting
Information). Interestingly, the
1H NMR spectrum of 2
ACHTREUNG[CbBr6]
shows the inequivalent methyl-
ene protons of the PCP ligand
at two very different chemical
shifts, d=6.24 and 4.33 ppm,
the former shifted downfield
compared with the free ligand
(d=4.45 ppm). Similar chemi-
cal-shift differences were noted
in the related complex [PdCl2-
AHCTRE(UNG PCP)], and while the solid-
state structure shows a relative-
ꢀ
ly close C H
A
Scheme 4. Salts 1[X]–3[X].
Table 2. Crystal structure and refinement data.
1
A
3
[CbCl6]
4
N
6b
N
7
G
8aACHTER[UNG CbCl6]
formula
Mw
1434.92
1102.06
1169.88
1698.37
1362.03
1197.09
specimen 0.48”0.30”0.10
[mm]
0.45”0.35”0.20
0.30”0.20”0.20
0.35”0.20”0.10
0.45”0.30”0.08
0.28”0.23”0.20
crystal
system
space
group
a [Š]
b [Š]
c [Š]
a [8]
monoclinic
orthorhombic
triclinic
monoclinic
monoclinic
monoclinic
¯
P21/c
Pbca
P1
P21/c
C2/c
C2/c
13.2823(2)
31.7248(4)
14.2156(2)
90
117.2678(7)
90
5324.49(13)
4
5.020
20.3691(1)
18.8427(1)
25.3130(1)
90
90
90
9715.35(8)
8
0.829
13.6680(1)
13.7350(1)
17.8597(1)
92.8916(3)
106.7497(3)
116.7593(3)
2802.33(3)
2
14.1224(2)
32.9909(3)
14.4927(2)
90
95.8867(5)
90
6716.70(15)
4
3.954
25.1370(4)
13.2248(2)
30.6433(5)
90
90.2907(7)
90
10186.7(3)
8
5.181
35.0748(3)
11.3221(1)
32.2252(3)
90
110.2554(7)
90
12005.9(2)
8
0.683
b [8]
g [8]
V [Š3]
Z
m [mmꢀ1
1calcd
]
0.688
1.387
1.790
1.507
1.680
1.776
1.326
[gcmꢀ3
]
2qmax [8]
57.3
66.2
57.0
54.9
55.2
52.6
Ntotal
54777
201967
54423
65148
39213
54980
collected
Nindependent 12480 (0.0632)
18428 (0.0585)
14132
14013 (0.0451)
11933
15013 (0.0648)
10469
10330 (0.0708)
8287
11996 (0.0680)
9621
(Rint
)
N
9867
(I>2s(I))
R1
(I>2s(I))
R1
0.0554
0.0761
0.1327
0.1431
0.0301
0.0321
0.0471
0.0585
0.0566
0.0496
0.0428
0.0832
0.0772
0.0768
ACHTREUNG
0.0727
0.0757
0.1055
0.1415
0.1302
ACHTREUNG
0.0831
0.0804
0.1184
0.1515
0.1394
ACHTREUNG
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8387

A. S. Weller, M. C. Willis et al.
tance,
a
ring-current-induced
chemical shift was suggested as
the reason for the downfield
shift.[30] An alternative explana-
tion comes from recent work
from the Bercaw group who
have classified such interactions
as pre-agostic with support from
Scheme 5. Generation of acetone adducts. Identity of [X] follows that shown in Scheme 4.
structural and computational studies.[31] It is likely that such
an interaction is also present in 2[CbBr6].
By contrast with the other nbd adducts, 3
downfield-shifted resonance in the 31P{1H} NMR spectrum
with reduced coupling to 103Rh, d=57.5 ppm (d, J
(RhP)=
AHCTREUNG
AHCTRE[UGN CbCl6] shows a
AHCTREUNG
125 Hz). Both these observations suggest coordination of
the sulfur atom to form a five-coordinate 18-electron com-
plex with a tridentate PSP ligand.[32] The solid-state structure
confirms this (Figure 1) and shows a pseudo-trigonal bipyra-
midal coordinated RhI centre, with a close rhodium–sulfur
Figure 2. Molecular structure of 4ACHTER[UNG CbCl6]. Selected bond lengths [Š]:
ꢀ
ꢀ
ꢀ
ꢀ
Rh O1 2.154(2), Rh O2 2.126(4), Rh O3 3.562(2), Rh P1 2.2019(5),
ꢀ
Rh P2 2.1978(6); selected bond angles [8]: O2-Rh-O1 83.30(12), P2-Rh-
P1 96.23(2), O2-Rh-P1 94.42(11), O1-Rh-P2 85.52(5), O1-Rh-P1
174.31(5), O2-Rh-P2 167.70(12). Sum of angles around Rh, 359.478. The
acetone ligands are disordered and only the major component is shown.
3.562(2) Š) and mutually cis-phosphines. 31P{1H} NMR spec-
troscopy shows resonances at d=41.8 ppm (d, J
209 Hz) for 4[CbCl6] and d=41.5 ppm (d, J(RhP)=202 Hz)
for 5[CbBr6]. The endo and exo protons for the CH2 linker
in 5[CbBr6] are now observed as being almost isochronous
at ꢁd=4.0 ppm in the H NMR spectrum, which suggests
ACHTREUNG(RhP)=
A
ACHTREUNG
AHCTREUNG
AHCTREUNG
1
[BArF ] is now
that the C H···Rh interaction present in 2ACHTREUNG
4
ꢀ
absent, or at least attenuated significantly. Electrospray ioni-
sation mass spectrometry (ESIMS) of these acetone adducts
shows the expected peaks due to the parent cation and loss
Figure 1. Molecular structure of the cationic portion of 3ACHTRE[UNG CbCl6]. Ellip-
soids are drawn at the 50% probability level. The hydrogen atoms and
ꢀ
anion have been omitted for clarity. Selected bond lengths [Š]: Rh1 C5
of one and two acetone molecules. By contrast, salt 3ACHTREUNG[CbCl6]
ꢀ
ꢀ
ꢀ
ꢀ
2.167(2), Rh1 C4 2.192(2), Rh1 C1 2.218(2), Rh1 C2 2.230(2), Rh1 S1
does not react with hydrogen (even at ꢁ100 bar H2), which
suggests that the sulfur atom is very strongly bound, the 18-
electron configuration retained and the oxidative addition
of H2 disfavoured. This lack of reactivity correlates with the
inactivity of this complex in the benchmark hydroacylation
reaction (Table 1) as presumably the active acetone adduct
is not formed.
ꢀ
ꢀ
ꢀ
ꢀ
2.2955(5), Rh1 P1 2.3487(5), Rh1 P2 2.3489(5), C1 C2 1.387(2), C4 C5
1.415(2); selected bond angles [8]: P1-Rh1-P2 103.38(2), S1-Rh1-P1
85.82(2), S1-Rh1-P2 86.73(2), C4-Rh1-S1 95.25(5), C2-Rh1-S1 156.64(5).
distance (Rh1–S1 2.2955(5) Š). The alkene C4/C5 unit and
the two phosphines lie in the trigonal plane. 18-electron,
RhI, complexes with p-accepting ligands have been reported
previously.[33]
Addition of H2 to acetone solutions of these nbd adducts
was anticipated to remove the strained diene and generate
the desired precatalysts, analogous to the Bonsich cata-
lyst.[34] For salts 1[X] and 2[X] this is the case and the ace-
Synthesis of acyl–hydrido intermediates: With suitable pre-
cursor RhI complexes in hand, we next investigated the ad-
dition of aldehyde, accepted to be the first step in the cata-
lytic cycle for hydroacylation.[2a,3] Addition of either alde-
hyde I or II to acetone solutions of 4[X] resulted in the im-
mediate (less than 5 minutes) formation of new acyl hydrido
tone adducts [Rh(L2)
A
species, [Rh(COCH2CH2SMe)
[Rh(COC6H4SMe)(DPEphos)H][X] 6b[X], in quantitative
yield as determined by NMR spectroscopy (Scheme 6). Ad-
dition of I or II to 5[CbCl6], which offers no hemilabile pro-
ACHTRE(UNG DPEphos)H][X] 6a[X] and
tively, can be generated in quantitative yield as determined
by NMR spectroscopy (Scheme 5). The reaction is immedi-
ate by NMR spectroscopy. A solid-state structure has been
ACHTREUNG
AHCTREUNG
obtained for 4
A
tection from the CH2 group, immediately (less than 5 mi-
nutes) gave a number of unidentified species and no acyl–
tion of the ether linkage with the metal centre (Rh···O3
8388
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8383 – 8397

Intermolecular Alkene Hydroacylation
FULL PAPER
tent with solution NMR spec-
troscopic data (vide infra). The
ꢀ
ꢀ
Rh O
distance
(Rh O2
2.248(3) Š) is ca. 0.1 Š longer
than that observed in [Rh-
ACHRTE(UNG xantphos)CO]AHCTRE[UNG BF4]
(2.126(3) Š; xantphos=9,9-di-
methyl-4,5-bis(diphenylphosphi-
no)xanthene),[24] and related
Scheme 6. Formation of new acyl–hydrido species 6a[X] and 6b[X]; [X]ꢀ =[closo-CB11H6Cl6]ꢀ.
compounds.[35] The Rh Cacyl dis-
ꢀ
tance, 1.969(5) Š, is similar to
ꢀ
hydrido complexes. In this case, possible C H activation of
the bridging methylene group may also play a role in the de-
composition pathway.[30] This rapid decomposition correlates
with the poor performance of this complex in the bench-
mark reaction (Table 1).
other reported RhIII-acyl bond lengths, for example that
found in [Rh
on the sulfur atom is canted away from lying in the Rh S
ꢀ ꢀ
ACHTRE(UNG PiPr3)2HACHERT(UNG COCH3)]ACHTRENUG
ꢀ
ꢀ
Cacyl
O
along with the twist of the DPEphos ligand backbone,
makes the phosphine atoms inequivalent in the solid-state.
cis-Acyl–hydrido complexes formed from addition of an al-
dehyde to a low-valent rhodium centre are known[1a,9,12,36]
and in some cases have been crystallographically character-
ised.[7,37] However, as far as we are aware, very few have ac-
tually been shown to be active in the hydroacylation reac-
tion.[1a,12]
The anions [closo-CB11H6Br6]ꢀ, [closo-CB11H6Cl6]ꢀ, and
[BArF ]ꢀ were used interchangeably with no effect on the
4
yield or identity of the cationic portion of the products, and
the NMR spectroscopic data is discussed for 6a
though the molecular structure from an X-ray diffraction ex-
periment was obtained for 6b[CbBr6] (Figure 3). We were
ACHTRE[UNG CbCl6], al-
AHCTREUNG
unable to obtain a crystal structure for 6a[X] (with a variety
of anions) because of the reductive decarbonylation of the
acyl ligand over the timescale of the crystallisation (days,
In solution at room temperature, complexes 6a/b
show similar sets of peaks for the phosphine and hydride li-
gands in their NMR spectra and only those for 6a[CbCl6]
are discussed in detail. The hydride signal is observed as a
(RhH)=23, J(PH)=
ACHTREUNG[CbCl6]
vide infra). The structure of 6b
A
ACHTREUNG
coodination of the DPEphos ligand, with the (located) hy-
dride ligand trans to sulfur atom and the acyl ligand trans to
the ether, which now occupies a coordination site on the
metal centre. The rhodium is best described as being in the
RhIII oxidation state, whereas the ligand orientations are in
accord with trans-influence arguments and are also consis-
doublet of triplets at d=ꢀ8.75 ppm (J
ACHTREUNG
ꢁ1 Hz), and one slightly broadened signal is observed for
the SMe group at d=1.56 ppm. The small coupling to 31P
suggests a mutual cis-orientation of hydride and phosphine
ligands. A H-1H COSY spectrum at 298 K revealed a four-
1
bond correlation between the hydride signal and the S-
methyl group, which suggests that the trans orientation of
the hydride and thioether observed in the solid-state is re-
tained in solution. The 13C{1H} NMR spectrum displays the
acyl carbon atom as a doublet at d=223.2 (JACHTREU(NG RhC)=34 Hz),
with the expected coupling to the cis phosphines presumably
small and unresolved. This chemical shift and coupling con-
stant are similar to those reported for related rhodium
acyl complexes, such as [Rh
(d=212.4 ppm,
(RhC)=28 Hz)[38]
[BArF ] (d=233.2 ppm,
JACHTREUNG
G
G
ACHTREUNG
J
A
ACHTREUNG
C5Me5)(C(O)C6H4CH3)]
29 Hz).[39]
A
4
A single resonance is observed in the 31P{1H} NMR spec-
trum at room temperature that shows coupling to 103Rh.
This is at odds with the solid-state structure, which shows in-
equivalent phosphorus environments. Progressive cooling to
180 K resolves this signal into a tightly-coupled ABX dou-
blet of doublets, showing large 31P–31P coupling (J
125, J(PP)=305 Hz) for 6a[CbCl6]. Such a large coupling
ACHTREUNG(RhP)=
AHCTREUNG
indicates trans-orientated phosphines—as found in the solid-
state structure. The hydride signal does not change signifi-
cantly on cooling, and only one SMe environment is ob-
served. The fluxional process that makes equivalent the
phosphorus atoms at room temperature cannot involve
Figure 3. Molecular structure of the cationic portion of 6bACHTRE[UNG CbBr6]. Ellip-
soids are drawn at the 50% probability level. The anion and hydrogen
ꢀ
atoms other than Rh H are omitted for clarity. Selected bond lengths
ꢀ
ꢀ
ꢀ
ꢀ
[Š]: Rh H(10) 1.47(4), Rh O2 2.248(3), Rh P1 2.293(1), Rh P2
ꢀ
ꢀ
2.303(1), Rh S 2.428(1), Rh C7 1.969(5); selected bond angles [8]: P1-
Rh-P2 155.93(5), C7-Rh-S 84.88(14), H(10)-Rh-S 169.7(15), C7-Rh-O2
177.3(2) C8-S-Rh-O2 torsion ꢀ48.4(2)
breaking of the Rh P bond (103Rh coupling is observed at
ꢀ
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8389

A. S. Weller, M. C. Willis et al.
ꢀ
298 K) or the Rh S bond (a correlation is observed between
hydride and SMe). Line-shape analysis affords DG (298 K)
°A
for this process, and comparison shows that the order for
ease of reorganisation for 6b
[CbCl6] ((32ꢂ2.3) kJmolꢀ1) is
less than 6a
CTHREUNG
AHCTREUNG
[CbCl6] ((42ꢂ2.7) kJmolꢀ1. The thiobenzalde-
hyde II thus results in a lower barrier to reorganisation by
ꢁ10 kJmolꢀ1 relative to I). DS° values are close to zero
(less the (5ꢂ3) JKꢀ1 molꢀ1). Two plausible mechanisms for
ꢀ
this reorganisation involve breaking the Rh O bond to
access a conformationally flexible five coordinate intermedi-
ate or inversion at the sulfur atom, which if coupled with a
twist of the phosphine backbone would render the phospho-
rus atoms equivalent. Inversion at the sulfur atom is a well-
documented process[40,41] and reported barriers are similar to
those determined here. The available data does not allow
Figure 4. Molecular structure of the cationic portion of 7aACHTRE[UNG CbI6] Ellip-
soids are drawn at the 50% probability level. The anion and hydrogen
ꢀ
the discrimination of which process is occurring: Rh O
cleavage or inversion at the sulfur atom. Although, as is
demonstrated later, the DPEphos ligand can be displaced by
ꢀ
atoms have been omitted for clarity. Selected bond lengths [Š]: Rh
ꢀ
ꢀ
ꢀ
ꢀ
C(38) 1.869(4), Rh S1 2.3694(11), Rh P1 2.387(1), Rh P2 2.309(1), Rh
O2 3.525(3); selected bond angles [8]: C(38)-Rh-S1 88.13(14) P2-Rh-P1
98.64(4), C(38)-Rh-P2 90.87(14), S1-Rh-P1 82.27(4).
ꢀ
excess MeCN, which shows that the Rh O bond can be
broken, we cannot envisage a plausible low-energy pathway
that would retain coupling between the hydride and SMe
(as is observed at 298 K) and invoke five-coordinate inter-
mediates that interconvert the methyl group from one side
to the other. Inversion at the sulfur atom would pass
through a planar, sp2-like at S intermediate. Such an inter-
mediate would be stabilised by the aryl group present in
ligand II, whereas stabilising (p–d)p conjugation would also
solid-state structure: two 31P environments are observed at
d=20.7 and 15.9 ppm that show cis 31P–31P coupling
(J(PP)=30 Hz) as well as coupling to 103Rh. The lower-field
signal shows a smaller 103Rh–31P coupling constant (150 vs.
125 Hz), which suggests it is due to the phosphorus trans to
the thioether.[42] Signals due to the thioether group are ob-
be favoured by weakening, although not necessarily break-
[41]
1
ꢀ
ing, of the Rh O bond to the hemilabile ligand.
served in the H NMR spectrum. At room temperature, the
31P{1H} NMR spectrum shows a broad, frequency-averaged,
signal that suggests an exchange process with reversible de-
coordination of the, now monodentate, thioether being the
most likely mechanism. The ESIMS spectrum shows a peak
at m/z=699.06 that corresponds to the parent ion minus
EtSMe, which further suggests a weakly bound thioether
Decarbonylation of the acyl–hydrido complexes: Given that
suppressing reductive decarbonylation is central to making a
longer-lived catalyst, the stability of the new acyl–hydrido
complexes in acetone was studied. Leaving a sample of 6a-
ACHTREUNG[CbCl6] or 6bACHTREUNG[CbCl6] in acetone for seven days at room tem-
perature caused new compounds to be formed (Scheme 7).
These were characterised by NMR spectroscopy and ESIMS
group. Data for 7bACHTRE[UNG CbCl6] are broadly similar and suggest a
closely related structure.
to be [Rh(CO)
[CbCl6]) and
CB11H6Cl6] (7b[CbCl6]). The molecular structure of 7a-
[CbI6] (the iodo carborane was used to obtain crystals suita-
ble for the diffraction experiment) is shown in Figure 4 and
demonstrates a square-planar RhI centre with a cis-DPE-
phos ligand, a carbonyl and methylethylthioether ligand.
The last two ligands arise from reductive decarbonylation of
the acyl hydride in the starting material. The 1H and
31P{1H} NMR spectra at 220 K are in full accord with the
A
G
This reductive decarbonylation was monitored over time
U
G
E
by NMR spectroscopy. Solutions of 6a
ACHTRE[UGN CbCl6] and 6b-
G
ACHTREUNG
G
hyde I or II to the acetone adduct 4ACHTRE[UNG CbCl6], were kept at
1
298 K and monitored periodically by 31P{1H} and H NMR
spectroscopy over a period of one week. The disappearance
1
of the hydride peak of the starting materials in the H NMR
spectrum was monitored and plotted (Figure 5). The
31P{1H} NMR spectrum shows similar time-dependant pro-
files, with the smooth conversion of 6
observed for both complexes. By way of comparison, the
Bonsich system [Rh(acetone)2A
(dppe)]+ (as the [CbBr6]ꢀ
ACHRTUNEG[CbCl6] into 7ACHTREUNG[CbCl6]
A
CHTREUNG
salt) was also studied by adding aldehyde I or II to acetone
solutions of this precatalyst.
Complexes 6a
bonylate to give 7a
lowing first-order kinetics. Complex 6aAHCTREUNG
A
ACHTREUNG
A
ACHTREUNG
lates significantly faster than 6b
A
vs. (1.2ꢂ0.1”10ꢀ6) sꢀ1, respectively), the aryl backbone in
Scheme 7. Formation of products 7a[X] and 7b[X].
6bACHTRE[UNG CbCl6] clearly inhibiting decarbonylation, as is to be ex-
8390
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8383 – 8397

Intermolecular Alkene Hydroacylation
FULL PAPER
G
AHCTRE[UNG CbCl6]) as an equilibrium mixture of three isomers (a, b
and c) and the starting materials (Scheme 8, inset table).[43]
Excess MeCN (ꢁ100 equivalents) pushes the equilibrium
completely over to the products.
The isomers of 8ACTHERNGU[CbCl6] and 9ACHTRE[UGN CbCl6] have been identi-
fied by 1H and 31P NMR spectroscopy as those shown in
Scheme 8. At room temperature, the MeCN adducts show
as broad peaks in the 31P{1H} NMR spectrum, whereas 6-
AHCTREUNG
AHCTREUNG
are resolved, and the ratio of the adducts to starting materi-
al changes in favour of the MeCN complexes (Scheme 8).
Warming restores the original concentrations.
31P–1H coupling observed for the hydride signals of isomers
8c[CbCl6] and 9c
[CbCl6] (J(PH)= ꢁ160 Hz), which appear
as a well-resolved doublet of doublet of doublets for 9c-
[CbCl6], indicate a hydride trans to one phosphine and cis to
the other, whereas those for 8a/b[CbCl6] and 9a/b[CbCl6]
A large
A
ACHTREUNG
AHCTREUNG
A
ACHTREUNG
show only smaller cis coupling to 31P. The 31P{1H} NMR
spectrum demonstrates that the phosphine ligands in all the
isomers lie cis to one another as indicated by the two differ-
ent environments observed for each isomer, which also
show mutual cis 31P–31P coupling. Each pair of phosphorus
Figure 5. a) Decomposition of the acyl–hydride intermediates over one
week. Concentrations calculated from NMR spectroscopic integrals by
using an internal standard. b) Plot of ln([Rh
ACHTRE(UNG acyl)hydride]t/[Rh-
ACHTREUNG
^
tion of the acyl–hydride complexes of the DPEphos ligand. : 6aACHTRE[UNG CbCl6];
&
: 6bACHTREUNG[CbCl6].
environments shows one large and one small J
pling (for example, 157/65, 150/66 and 155/65 for 9a
9b[CbCl6] and 9c[CbCl6], respectively), which suggests
structures in which one phosphorus atom is opposite a high
trans-influence ligand (acyl or hydride: low J(RhP)), where-
as the other is trans to the thioether or MeCN (larger J-
(RhP)). Only three peaks are observed for the SMe groups
ACHTREUNG
ACHTREUNG
pected for a more rigid system.[1] For the dppe system, a
complex mixture is formed on addition of aldehyde to [Rh-
A
ACHTREUNG
A
(dppe)]
A
ACHTREUNG
finitively assign signals due to an acyl–hydride species.
ESIMS of this mixture indicated that the product of reduc-
AHCTREUNG
tive decarbonylation [Rh(CO)
G
G
between d=1.8 and 1.4 ppm in the 1H NMR spectrum,
which is accounted for by a coincidence of resonances for
two of the species. Owing to the complexity of the system,
13C{1H} NMR spectroscopy was not useful in determining
the structures of the products. Integrating over all the iso-
formed. Noteworthy is that acyl–hydrido intermediates were
not observed in the original report of addition of 4-pentenal
[ClO4].[2a]
to [RhACHTREUNG(acetone)2ACHTREUNG(dppe)]ACHTREUGN
Displacement of the hemilabile
oxygen ligand by using MeCN:
As we show in the next section,
addition of alkene to 6ACHTRE[UNG CbCl6]
rapidly affords the ultimate
products of hydroacylation,
which means that the simple
alkene adducts are not ob-
served. We have modelled the
coordination of alkene to the
metal centre by addition of ace-
tonitrile. Adding ten equiva-
lents of MeCN to acetone solu-
tions of 6aACHTREUNG[CbCl6] or 6bACHRTE[UGN CbCl6]
resulted in displacement of the
bound oxygen atom and the
formation of the adducts [Rh-
ACHTREUNG(DPEphos)(MeSC2H4CO)H-
Scheme 8. At 298 K, the signals due to 9a and 9b isomers overlap in both the 31P{1H} and 1H NMR spectra.
ACHTREUNG
ACHTREUNG
[Rh- 8=aldehyde I, 9=aldehyde II.
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8391

A. S. Weller, M. C. Willis et al.
mers of the MeCN adducts in the hydride region of the
1H NMR spectrum at various temperatures indicates equilib-
rium thermodynamics from a van’t Hoff plot (lnKeq versus
1/T), and DG0 derived from this represents a weighed aver-
age of the free-energy terms for the all isomers. With this
caveat, the data indicate that the DPEphos ligand is just as
easy to replace in either complex (DH0 =ꢁ(ꢀ30ꢂ
0.5) kJmolꢀ1) and the process shows an overall large nega-
tive change in entropy (DS0 =ꢁ(ꢀ90ꢂ3) JKꢀ1 molꢀ1), con-
sistent with the addition of a molecule of solvent MeCN in
(3.592(2) Š). The trans influence of the acyl group is demon-
ꢀ
ꢀ
strated by the differing Rh P bond lengths: Rh P1
ꢀ
2.5080(9), Rh P2 2.305(1) Š.
Complexes 8
A
[CbCl6] represent models for
alkene coordination in the catalytic cycle with MeCN re-
placing the alkene, and in particular the isomers B and C
(Scheme 8), which have a cis-disposition of the incoming
ligand with respect to the hydride group, are perfectly setup
for subsequent insertion into alkene (e.g. structure C in
Scheme 1). cis!trans!cis-folding of the DPEphos ligand as
seen here (5!6!8) has been noted before in the alcholysis
of acyl–palladium complexes bearing this ligand.[25]
the product. DG0
(<5 kJmolꢀ1) in favour of the adducts. The adducts 8
and 9[CbCl6] do not undergo decarbonylation, as anticipated
ACHTREUNG
AHCTREUNG
ACHTREUNG
for complexes in which a vacant site is blocked by a MeCN
ligand.
Addition of alkene—closing the catalytic cycle: In the pre-
ceding sections, it has been demonstrated that addition of al-
Single crystals of 8a
A
dehyde to the well-defined precatalysts (e.g. 4
sults in oxidative addition to form acyl hydrides (e.g. 6-
[CbCl6]) in which the hemilabile phosphine ligand stabilises
ACHTRE[UNG CbCl6]) re-
pentane to an acetone solution of the mixture of isomers,
and the result of the X-ray structure determination is shown
in Figure 6. Dissolution of these crystals gave the equilibri-
AHCTREUNG
the metal centre towards reductive decarbonylation. MeCN,
ꢀ
acting as a model for an incoming alkene, displaces the Rh
O bond and forces the phosphines to lie cis. Addition of
alkene to complexes 6ACHTRE[UGN CbCl6] should now close the catalytic
cycle.
Addition of methylacrylate to acetone solutions of 6a-
AHCTREUNG
1
complex that was short lived (t = = ꢁ10 min) and evolved to
2
give the product complex 11aACHTRE[UNG CbCl6] (Scheme 9), which was
characterised by NMR spectroscopy and ESIMS/MS. For
1
6b
A
2
and has been spectroscopically characterised as 10b
ACHTREUNG[CbCl6],
which then evolved to give 11b[CbCl6]. The transformation
AHCTREUNG
retains mass-balance (Figure 7) and no other complexes of
significant concentration are observed by 31P{1H} NMR
spectroscopy or ESIMS. The final products are discussed
Figure 6. Molecular structure of the cationic portion of 8aACHTRE[UNG CbCl6]. Ellip-
soids are drawn at the 50% probability level. The anion and hydrogen
first, which were characterised as [Rh
MeSC2H4-h1-COC2H4CO2Me)]
[CbCl6] (11a
[Rh
(DPEphos)(h1-MeSC6H4-h1-COC2H4CO2Me)]
(11b[CbCl6]).
The room temperature 31P{1H} NMR spectrum of 11a-
ACHTREUNG
ꢀ
atoms other than Rh H are omitted for clarity. Selected bond lengths
A
ACHTREUNG
ꢀ
ꢀ
ꢀ
ꢀ
[Š]: Rh P1 2.5080(9), Rh P2 2.305(1), Rh H
U
U
ACHTREUNG
ꢀ
ꢀ
ꢀ
2.129(3), Rh O1 2.039(4), Rh S 2.372(1), Rh O2 3.592(2); selected
bond angles [8]: P1-Rh-P2 100.37(3), C-Rh-S, 82.7(1),
AHCTREUNG
HACHTRE(GNU 100)-Rh-N
179.0(2).
AHCTREUNG
um mixture observed before. The solid-state structure shows
an octahedral RhIII centre with the DPEphos ligand now
folded (P1-Rh-P2 100.37(3)8) so that the phosphines lie cis
to one another and the (located) hydride lies trans to the
acetonitrile ligand. There is no close Rh···O contact
(JACHTRE(UNG RhP)=220 and 173 Hz, respectively). We assign the
Scheme 9. Addition of methylacylate to 6a
A
ACHRTUNEG[CbCl6] and then final product 16aACHTRENUG
A
G
A
ACHTREUNG
8392
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8383 – 8397

Intermolecular Alkene Hydroacylation
FULL PAPER
alkene in solution. Similar spectra are obtained for 11b-
AHCTREUNG
ACHTREUNG
AHCTREUNG
A
ACHTREUNG
demonstrates turnover of the catalyst.
After having established the identity of 11a/bCAHTRE[UGN CbCl6], the
identity of the intermediate species 10a/b
investigated. For 10a[CbCl6], this intermediate is only rela-
tively short lived, but for 10b[CbCl6] it survives long enough
ACHTRE[UNG CbCl6] was then
AHCTREUNG
AHCTREUNG
for successful characterisation by NMR spectroscopy and
ESIMS/MS, presumably a consequence of the constraints
imposed by the aryl ring in II. These data suggest a sensible
formulation for the intermediate species as [Rh-
Figure 7. Time/concentration plot for the addition of methylacrylate to
~
&
^
6b
A
G
A
N
A
ACHTRE[UGN CbCl6] (10a-
signal with the larger coupling to a phosphorus trans to the
A
ACHTREUNG
bound ketone, which would be expected to be less-strongly
A
ACHTREUNG
bound to the Rh and thus result in a larger J
A
ACHTREUNG
constant. Similar RhP couplings are observed for complexes
in which h1-ketones are located trans to phosphines or phos-
phines trans to thioethers,[42] and the chemical shifts and
is reached after approximately 15 minutes of reaction time
(Figure 7). ESIMS at this time shows a strong-intensity ion
at m/z=879.2 that corresponds to the formula suggested for
J
A
10b
Interrogation of the m/z=879.2 peak by MS/MS resulted in
the loss of methylacrylate (m/z=86.1) to give 6b[CbCl6]
(m/z=793.1), presumably by facile b-elimination in the
spectrometer under MS/MS conditions (Figure 8). This
clearly differentiates this intermediate from the final prod-
ACHRTUNEG[CbCl6], as well as peaks due to 6bACHTRE[UNG CbCl6] (m/z=793.1).
AHCTREUNG
the DPEphos ligand is not bound through the oxygen atom.
ACHTREUNG
We discount binding through two h1-carbonyls on the basis
of the difference in JACHTRE(UGN RhP) coupling and that SMe would
find a better match with soft RhI than the ester carbonyl.
1
The H NMR spectrum shows a featureless hydride region,
uct 11bAHCTREU[NG CbCl6] that has the same mass but fragments by loss
which demonstrates loss of a hydride ligand, in addition to
signals due to the product MeSC2H4COC2H4CO2Me in the
aliphatic region. ESIMS displays a parent ion at m/z=
831.13 that loses MeSC2H4COC2H4CO2Me under MS/MS
conditions (m/z=190.1) consistent with the suggested for-
mulation. The same NMR and ESIMS spectra are recreated
of MeSC2H6COC2H4CO2Me on MS/MS (m/z=238.1).
1
Following the reaction by 31P{1H} and H NMR spectros-
copy reveals additional data allowing for the tentative iden-
tification of the intermediates, which are discussed for 10b-
AHCTRE[UNG CbCl6]. This complex is fluxional at room temperature, dis-
1
playing broad signals in the H and 31P{1H} NMR spectra, in
addition to sharper signals due to 6b[CbCl6] and 11b-
[CbCl6]. The observation of separate signals for 10b[CbCl6]
and 6b[CbCl6] shows that they are not in rapid equilibrium
by adding MeSC2H4COC2H4CO2Me to 4
A
ACHTREUNG
In this case, the molecule is static at room temperature,
showing sharp signals in the 31P{1H} NMR spectrum. The
broadness observed in the sample in which alkene is added
A
ACHTREUNG
AHCTREUNG
on the NMR timescale with each other. Cooling a sample to
1
to 6a
A
200 K gave a complicated H NMR spectrum in the aliphatic
Figure 8. ESI (a,b) and ESIMS/MS (c,d) data for the sequential addition of aldehyde II to 4CAHTREU[GN CbCl6].
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8393

A. S. Weller, M. C. Willis et al.
region, with overlapping signals. In the hydride region, only
a diminished signal due to 6b[CbCl6] was observed, which
demonstrated that the intermediate species does not have a
hydride ligand, and suggested that hydride insertion into the
alkene has taken place. In the 31P{1H} NMR spectrum, six
doublets of doublets were observed, in addition to the signal
The complete hydroacylation catalytic cycle from observable
intermediates: Although a rigorous kinetic study is outside
the scope of this paper, the reactivity and new complexes re-
ported that arise from stoichiometric addition of substrates
to precatalyst species allows a catalytic cycle to be presented
based upon observed species. Scheme 10 presents the cycle
and shows the flexible role of the hemilabile ligand in ac-
commodating the various metal electron counts, oxidation
states and resulting geometries. This cycle complements sim-
ilar schemes outlined by Bosnich,[2a] which have been com-
putationally studied by Morehead and Sargent[3] for the in-
tramolecular hydroacylation of 4-pentenals, and intermolec-
ular hydroacylation reactions by using [M(h5-C5Me4X)(h2-
H2C=CHSiMe3)2] catalysts (X=Me, CH3; M=Co,[4] Rh[5])
as reported by Brookhart.
AHCTREUNG
for 6bACHTREUNG[CbCl6], which we assign to three isomers of 10b-
ACHTREUNG[CbCl6] in which the alkyl, acyl and thioether ligands sit in
different, mutually cis, positions around an octahedral RhIII
centre. These isomers are observed in the ratio 5:3:2, facili-
tating the pairing of resonances. All the pairs show cis J(PP)
coupling (26–32 Hz]). For two of these pairs, there is one
large and one small JACHTRE(UNG RhP) coupling (for example, 174 and
79 Hz) and the third shows two small values (ꢁ75 Hz). This
is reminiscent of the couplings observed in the MeCN ad-
ducts (e.g. 8b
either opposite a high trans influence ligand (alkyl or acyl:
small J(RhP)) or the thioether group (larger J(RhP)). The
ACHTREUNG[CbCl6]) and suggests phosphines that are
Applications to organic synthesis—representative examples:
Employing the DPEphos catalyst in the benchmark hydro-
acylation reaction indicated a modest advantage over the
Bosnich dppe-derived catalyst (see the section Evaluation of
precatalysts in a benchmark hydroacylation reaction). How-
ever, this benchmark reaction is achievable with a number
of catalysts and the test of whether the DPEphos-derived
system had any real advantages in terms of reactivity would
be to employ it in more challenging hydroacylation reac-
tions. Accordingly, the DPEphos catalyst was used in a
series of hydroacylation reactions between b-S-substituted
aldehydes I, II, V and VI (Table 3), the two alkenes methyl
acrylate and octene, and a terminal and internal alkyne. The
G
ACHTREUNG
DPEphos oxygen atom is suggested to provide stabilisation
to the, formally, 16-electron RhIII centre, but as these iso-
ꢀ
mers interconvert at room temperature, any Rh O interac-
tion is likely to be weak, allowing access to a five-coordi-
nate, fluxional intermediate. Reductive elimination from
10bACHTREUNG[CbCl6] to form 11bACHRTE[UGN CbCl6] would also require a five-co-
ordinate species.[8,9] Alkyl-acyl species have been reported
previously, but as found here, they often undergo reductive
elimination to give the corresponding ketones.[44]
Scheme 10. Catalytic cycle, decarbonylation products and acetontrile adducts for the hydroacylation reaction by using 4ACHTRE[UGN CbCl6], aldehyde I or II and
methylacrylate.
8394
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8383 – 8397

Intermolecular Alkene Hydroacylation
FULL PAPER
catalyst used in these reactions was generated from the
combinations of coupling partners delivered good yields of
the hydroacylation adducts. In all cases, the yields achieved
matched, or surpassed, those obtained with the dppe-con-
taining catalyst.[20] Particularly worthy of note are the reac-
tions of all four aldehydes with the simple alkene octene
(entries 4, 8, 12 and 16); use of the dppe-derived catalyst in
these reactions delivered very
simple combination of [RhCl
ACHTRE(UNG cod)]2, DPEphos and Ag-
ACHTREUNG
to the cationic fragment from a pragmatic view, considering
the lack of general availability of the alternative carborane
derived systems. The results are presented in Table 3; all
Table 3. Scope of the new catalyst system.[a]
poor yields (<5%) of hydroa-
cylation adducts, at best. For
example, the reactions of both
aldehydes V and VI with octene
deliver none of the hydroacyla-
tion adducts with this catalyst.
No doubt key in this marked
performance gain with these
challenging substrates com-
pared with the dppe ligated
system is the longevity of the
catalyst with regard to reduc-
tive decarbonylation. Although
we suggest the overall rate of
reaction might not be signifi-
cantly faster with the DPEphos
systems (c.f. the broadly similar
reaction times observed in the
benchmark reaction) it is the
catalyst robustness that allows
continued turnover over a long
period of time (up to 48 h).
Consistent with this, at the end
of catalysis under the bench-
mark conditions, addition of II
to the reaction mixture regener-
Entry
Aldehyde
Alkene/alkyne
Product
t [h]
Yield [%][b]
74
1
1
2
3
4
2
16
24
82
80
67
5
6
7
8
16
16
24
48
81
97
73
61
9
4
4
75
83
81
89
ates 6bACHTRE[UNG CbCl6] as the only ob-
servable organometallic prod-
uct, which demonstrates that
post catalysis the catalyst is
available for further turnovers
and has not decomposed signifi-
cantly.
10
11
12
24
48
Counterion effects: Although
the stoichiometric additions of
aldehydes and alkenes to the
precursor complexes show no
significant difference on chang-
ing the counterion, a compari-
son of the performance of vari-
ous precursor complexes re-
veals that there is an apprecia-
ble counterion effect (Table 4)
in the benchmark catalysis.
Counterion effects in catalysis
are well documented and can
result from a combination of
enhanced metal vacant site
availability and attenuation of
13
14
15
16
2
2
86
91
81
73
24
24
[a] Conditions: aldehyde (1.0 equiv), alkene or alkyne (2.0 equiv), [RhCl
(5 mol.%), Ag[ClO4] (5 mol.%), acetone, 558C. [b] Isolated yields of single isomers.
ACHTRE(UNG cod)]2 (2.5 mol.%), DPEphos
ACHTREUNG
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8395

A. S. Weller, M. C. Willis et al.
Table 4. Counterion influence on benchmark reaction (see Table 1).[a]
[1] a) J. W. Suggs, J. Am. Chem. Soc. 1978, 100, 640–641; b) K. Sakai, J.
Ide, 0. Oda, N. Nakamura, Tetrahedron Lett. 1972, 1287-1290.
Entry
Catalyst
Conversion [%][b]
t
1
2
3
4
5
[Rh
[Rh
[Rh
[Rh
[Rh
C
G
A
100
100
100
100
100
90 min
60 min
4 h
75 min
30 min
[2] a) D. P. Fairlie, B. Bosnich, Organometallics 1988, 7, 946–954;
b) This catalyst has been formulated as existing mainly as monomer-
G
ACHTREUNG
G
ACHTREUNG
ic [Rh
ly associated ion-pair [Rh
arene bridged dimer [Rh(dppe)] CHTREUNG
ACHTREU(GN acetone)2AHCERT(UNG dppe)]ACHTRENUG
T
G
]
A
ACHTREUNG
4
ACHTREUNG
C
[closo-CB11H6Cl6]
[Rh(DPEphos)(nbd)]
[closo-CB11H6Br6]
catalyst is strongly argued to be a monomeric substrate adduct, and
rates of catalysis for the hydroacylation of 4-pentenal are similar for
all three solvents, which suggests a common active species.
[3] I. F. D. Hyatt, H. K. Anderson, A. T. Morehead, A. L. Sargent, Or-
ganometallics 2008, 27, 135–147.
6
A
G
100
45 min
[a] Conditions: aldehyde (1.0 equiv), methyl acrylate (2.0 equiv), catalyst
(5 mol.%), acetone, 558C. Catalyst generated by addition of H2 (1 atm,
5 min, acetone). [b] Determined by H NMR spectroscopy.
1
[4] C. P. Lenges, P. S. White, M. Brookhart, J. Am. Chem. Soc. 1998,
120, 6965–6979.
[5] A. H. Roy, C. P. Lenges, M. Brookhart, J. Am. Chem. Soc. 2007, 129,
2082–2093.
[6] R. H. Crabtree, The Organometallic Chemistry of the Transition
Metals, Wiley, Hoboken, New Jersey, 2005.
[7] R. Goikhman, D. Milstein, Angew. Chem. 2001, 113, 1153–1156;
Angew. Chem. Int. Ed. 2001, 40, 1119–1122.
[8] a) U. Fekl, K. I. Goldberg, J. Am. Chem. Soc. 2002, 124, 6804–6805;
b) J. Procelewska, A. Zahl, G. Liehr, R. van Eldik, N. A. Smythe,
B. S. Williams, K. I. Goldberg, Inorg. Chem. 2005, 44, 7732–7742;
c) A. T. Luedtke, K. I. Goldberg, Inorg. Chem. 2007, 46, 8496–8498.
[9] D. Milstein, Acc. Chem. Res. 1984, 17, 221–226.
decomposition pathways by the anion.[45] Under bench-top
conditions, the carborane anions perform the best (entries 5
and 6), as measured by time taken for the reaction to reach
completion, with [BArF ]ꢀ performing less well (entry 4),
4
and [PF6]ꢀ significantly poorer (entry 3). This could well be
due to trace amounts of water present in the acetone that
have been shown to attenuate the activity of [BArF ]ꢀ and,
4
more so, [PF6]ꢀ salts in hydrogenation reactions with cation-
ic Group 9 metals.[28,46] Overall, salts of [ClO4]ꢀ represent a
reasonable balance of convenience and overall rate
(entry 2).
[10] B. Bosnich, Acc. Chem. Res. 1998, 31, 667–674.
[11] C. H. Jun, E. A. Jo, J. W. Park, Eur. J. Org. Chem. 2007, 1869–1881.
[12] D. Milstein, J. Chem. Soc. Chem. Commun. 1982, 1357–1358.
[13] D. P. Fairlie, B. Bosnich, Organometallics 1988, 7, 936–945.
[14] B. R. James, C. G. Young, J. Chem. Soc. Chem. Commun. 1983,
1215–1216.
[15] a) R. W. Barnhart, D. A. McMorran, B. Bosnich, Chem. Commun.
1997, 589–590; b) R. W. Barnhart, D. A. McMorran, B. Bosnich,
Inorg. Chim. Acta 1997, 263, 1–7; c) R. W. Barnhart, B. Bosnich, Or-
ganometallics 1995, 14, 4343–4348; d) R. W. Barnhart, X. Wang, P.
Noheda, S. H. Bergens, J. Whelan, B. Bosnich, J. Am. Chem. Soc.
1994, 116, 1821–1830.
[16] a) K. Tanaka, G. C. Fu, J. Am. Chem. Soc. 2001, 123, 11492–11493;
b) K. Tanaka, G. C. Fu, Angew. Chem. 2002, 114, 1677–1679;
Angew. Chem. Int. Ed. 2002, 41, 1607–1609; c) H. D. Bendorf, C. M.
Colella, E. C. Dixon, M. Marchetti, A. N. Matukonis, J. D. Mussel-
man, T. A. Tiley Tetrahedron Letters 2002, 43, 7031–7034.
[17] a) C.-H. Jun, Y.-G. Lim, Tetrahedron Lett. 1995, 36, 3357–3360;
b) C.-H. Jun, H. Lee, J.-B. Hong, J. Org. Chem. 1997, 62, 1200–1201;
c) C.-H. Jun, J.-B. Hong, D.-Y. Lee, Synlett 1999, 1–12; C.-H. Jun, J.-
B. Hong, Y.-H. Kim, K.-Y. Chung, Angew. Chem. 2000, 112, 3582–
3584; Angew. Chem. Int. Ed. 2000, 39, 3440–3442; d) C.-H. Jun, D.-
Y. Lee, H. Lee, J.-B. Hong, Angew. Chem. 2000, 112, 3214–3216;
Angew. Chem. Int. Ed. 2000, 39, 3070–3072; e) C.-H. Jun, H. Lee,
S.-G. Lim, J. Am. Chem. Soc. 2001, 123, 751–752; f) C. H. Jun, J. H.
Chung, D. Y. Lee, A. Loupy, S. Chatti, Tetrahedron Lett. 2001, 42,
4803–4805; g) C. H. Jun, H. Lee, S. G. Lim, J. Am. Chem. Soc. 2001,
123, 751–752; h) C.-H. Jun, H. Lee, J.-B. Hong, B.-I. Kwon, Angew.
Chem. 2002, 114, 2250–2251; Angew. Chem. Int. Ed. 2002, 41, 2146–
2147; i) C.-H. Jun, C. W. Moon, D.-Y. Lee, Chem. Eur. J. 2002, 8,
2422–2428; j) C.-H. Jun, C. W. Moon, H. Lee, D.-Y. Lee, J. Mol.
Catal. A: Chem. 2002, 189, 145–156; k) C.-H. Jun, C. W. Moon, S.-
G. Lim, H. Lee, Org. Lett. 2002, 4, 1595–1597; l) C.-H. Jun, J. H.
Lee, Pure Appl. Chem. 2004, 76, 577–587; m) Y. Taura, M. Tanaka,
X. M. Wu, K. Funakoshi, K. Sakai, Tetrahedron 1991, 47, 4879–
4888; n) M. Tanaka, M. Imai, M. Fujio, E. Sakamoto, M. Takahashi,
Y. Eto-Kato, X. M. Wu, K. Funakoshi, K. Sakai, H. Suemune, J.
Org. Chem. 2000, 65, 5806–5816; o) M. Tanaka, M. Takahashi, E.
Sakamoto, M. Imai, A. Matsui, M. Fujio, K. Funakoshi, K. Sakai, H.
Suemune, Tetrahedron 2001, 57, 1197–1204; p) M. Tanaka, M. Imai,
Y. Yamamoto, K. Tanaka, M. Shimowatari, S. Nagumo, N. Kawa-
hara, H. Suemune, Org. Lett. 2003, 5, 1365–1367; q) M. Tanaka, K.
Sakai, H. Suemune, Current Organic Chemistry 2003, 7, 353–367;
Conclusions
We have developed a practicable catalyst for the hydroacy-
lation reaction of b-S-substituted aldehydes with activated
and unactivated alkenes and alkynes. The catalyst system
uses a hemilabile DPEphos ligand to stabilise key acyl–hy-
drido intermediates against unproductive reductive decar-
bonylation. This leads to a longer-lived catalyst that is able
to couple relatively unreactive alkenes, such as 1-octene,
with the aldehyde. Stoichiometeric studies on sequential ad-
dition of aldehyde and alkene to precatalyst systems shows
that DPEphos provides not only stabilisation towards reduc-
tive decarbonylation through reversible coordination of the
oxygen ligand, but can also adjust its coordination geometry
(cis, trans, cis). This not only reflects the requirements of the
rhodium centre throughout the catalytic cycle but also
allows coordinated groups (hydride, acyl and alkene) to
adopt mutually cis orientations necessary for the key inser-
tion and reductive elimination steps. The flexible and hemi-
labile nature of the ligand is thus important, and these
design features could well be significant in the realisation of
future catalyst systems that will be able to couple any alde-
hyde with any alkene—the ultimate goal in the area of hy-
droacylation chemistry.
Acknowledgement
The authors wish to thank the EPSRC for financial support of this work
and R. Dallanegra, N. Genta and Dr J. Osbourne for experimental assis-
tance.
8396
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8383 – 8397

Intermolecular Alkene Hydroacylation
FULL PAPER
r) K. Tanaka, M. Tanaka, H. Suemune, Tetrahedron Lett. 2005, 46,
6053–6056; s) K. Tanaka, Y. Shibata, T. Suda, Y. Hagiwara, M.
Hirano, Org. Lett. 2007, 9, 1215–1218; t) M. Imai, M. Tanaka, S.
Nagumo, N. Kawahara, H. Suemune, J. Org. Chem. 2007, 72, 2543–
2546; u) Y. Oonishi, M. Mori, Y. Sato, Synthesis 2007, 2323–2336.
[18] a) M. C. Willis, S. Sapmaz, Chem. Commun. 2001, 2558–2559;
b) M. C. Willis, S. J. McNally, P. J. Beswick, Angew. Chem. 2004, 116,
344–347; Angew. Chem. Int. Ed. 2004, 43, 340–343; c) M. C. Willis,
H. E. Randell-Sly, R. L. Woodward, G. S. Currie, Org. Lett. 2005, 7,
2249–2251.
[31] a) J. C. Lewis, J. Wu, R. G. Bergman, J. A. Ellman, Organometallics
2005, 24, 5737–5746; b) Y. Zhang, J. C. Lewis, R. G. Bergman, J. A.
Ellman, E. Oldfield, Organometallics 2006, 25, 3515–3519.
[32] P. S. Pregosin, R. W. Kunz, NMR of transition metal phosphine com-
plexes 1979, Springer-Verlag, Berlin
[33] a) J. A. Long, T. B. Marder, P. E. Behnken, M. F. Hawthorne, J. Am.
Chem. Soc. 1984, 106, 2979–2989; b) I. V. Kourkine, C. S. Slone,
C. A. Mirkin, L. M. Liable-Sands, A. L. Rheingold, Inorg. Chem.
1999, 38, 2758–2759.
[34] R. R. Schrock, J. A. Osborn, J. Am. Chem. Soc. 1976, 98, 2134–
2143.
[35] N. W. Alcock, J. M. Brown, J. C. Jeffery, J. Chem. Soc. Dalton Trans.
1976, 583–588.
[19] G. L. Moxham, H. E. Randell-Sly, S. K. Brayshaw, R. L. Woodward,
A. S. Weller, M. C. Willis, Angew. Chem. 2006, 118, 7780–7784;
Angew. Chem. Int. Ed. 2006, 45, 7618–7622.
[20] M. C. Willis, H. E. Randell-Sly, R. L. Woodward, S. J. McNally, G. S.
Currie, J. Org. Chem. 2006, 71, 5291–5297.
[36] K. Wang, T. J. Emge, A. S. Goldman, C. Li, S. P. Nolan, Organome-
tallics 1995, 14, 4929–4936.
[21] a) C. F. Lochow, R. G. Miller, J. Am. Chem. Soc. 1976, 98, 1281–
1283; b) R. E. Campbell, C. F. Lochow, K. P. Vora, R. G. Miller, J.
Am. Chem. Soc. 1980, 102, 5824–5830; c) T. B. Marder, D. C. Roe,
D. Milstein, Organometallics 1988, 7, 1451–1453.
[22] P. Braunstein, F. Naud, Angew. Chem. 2001, 113, 702–722; Angew.
Chem. Int. Ed. 2001, 40, 680–699.
[37] a) D. Milstein, W. C. Fultz, J. C. Calabrese, J. Am. Chem. Soc. 1986,
108, 1336–1338; b) C. Bianchini, A. Meli, M. Peruzzini, F. Vizza, F.
Bachechi, Organometallics 1991, 10, 820–823; c) M. A. Garralda, R.
Hernandez, L. Ibarlucea, E. Pinilla, M. R. Torres, M. Zarandona,
Organometallics 2007, 26, 1031–1038.
[38] Y. S. Varshavsky, T. G. Cherkasova, N. A. Buzina, L. S. Bresler, J.
Organomet. Chem. 1994, 464, 239–245.
[39] B. K. Corkey, F. L. Taw, R. G. Bergman, M. Brookhart, Polyhedron
2004, 23, 2943–2954.
[40] a) E. W. Abel, S. K. Bhargava, K. G. Orrell, Prog. Inorg. Chem.
1984, 32, 1–118; b) E. W. Abel, G. W. Farrow, K. G. Orrell, J. Chem.
Soc. Dalton Trans. 1976, 1160–1163; c) E. W. Abel, M. Booth, K. G.
Orrell, J. Chem. Soc. Dalton Trans. 1980, 1582–1592; d) F. Qian, M.
Ferrer, M. L. H. Green, P. Mountford, V. S. B. Mtetwa, K. Prout, J.
Chem. Soc. Dalton Trans. 1991, 1397–1406; e) J. R. Dilworth,
C. A. M. von Beckh, S. I. Pascu, Dalton Trans. 2005, 2151–2161.
[41] X. P. Shan, J. H. Espenson, Organometallics 2003, 22, 1250–1254,
and references therein.
[42] a) S. Sjovall, C. Andersson, O. F. Wendt, Organometallics 2001, 20,
4919–4926; b) A. M. Brown, M. V. Ovchinnikov, C. A. Mirkin,
Angew. Chem. 2005, 117, 4279–4281; Angew. Chem. Int. Ed. 2005,
44, 4207–4209; c) Y. M. Jeon, J. Heo, A. M. Brown, C. A. Mirkin,
Organometallics 2006, 25, 2729–2732.
[23] a) J. C. Jeffery, T. B. Rauchfuss, P. A. Tucker, Inorg. Chem. 1980, 19,
3306–3316; b) J. W. Faller, N. Sarantopoulos, Organometallics 2004,
23, 2008–2014; c) G. Chen, W. H. Lam, W. S. Fok, H. W. Lee, F. Y.
Kwong, Chem. Asian J. 2007, 2, 306–313; d) D. B. Grotjahn, Y.
Gong, L. Zakharov, J. A. Golen, A. L. Rheingold, J. Am. Chem.
Soc. 2006, 128, 438–453; e) S. Chikkali, D. Gudat, M. Niemeyer,
Chem. Commun. 2007, 981–983; f) H. Werner, Dalton Trans. 2003,
3829–3837; g) N. C. Gianneschi, M. S. Masar, C. A. Mirkin, Acc.
Chem. Res. 2005, 38, 825–837; h) P. A. Ulmann, A. M. Brown, M. V.
Ovchinnikov, C. A. Mirkin, A. G. DiPasquale, A. L. Rheingold,
Chem. Eur. J. 2007, 13, 4529–4534; i) T. E. Barder, S. D. Walker,
J. R. Martinelli, S. L. Buchwald, J. Am. Chem. Soc. 2005, 127, 4685–
4696; j) Z. Q. Weng, S. H. Teo, T. S. A. Hor, Acc. Chem. Res. 2007,
40, 676–684; k) M. H. Chisholm, J. C. Gallucci, G. Yaman, Chem.
Commun. 2006, 1872–1874.
[24] A. J. Sandee, L. A. van der Veen, J. N. H. Reek, P. C. J. Kamer, M.
Lutz, A. L. Spek, P. van Leeuwen, Angew. Chem. 1999, 111, 3428–
3432; Angew. Chem. Int. Ed. 1999, 38, 3231–3235.
[43] In our initial communication (reference [19]), we did not report the
[25] P. W. N. M. van Leeuwen, M. A. Zuideveld, B. H. G. Swennenhuis,
Z. Freixa, P. C. J. Kamer, K. Goubitz, J. Fraanje, M. Lutz, A. L.
Spek, J. Am. Chem. Soc. 2003, 125, 5523–5539.
[26] R. Venkateswaran, J. T. Mague, M. S. Balakrishna, Inorg. Chem.
2007, 46, 809–817.
[27] a) C. A. Reed, Acc. Chem. Res. 1998, 31, 133–139; b) N. J. Patmore,
C. Hague, J. H. Cotgreave, M. F. Mahon, C. G. Frost, A. S. Weller,
Chem. Eur. J. 2002, 8, 2088–2098; c) A. Rifat, N. J. Patmore, M. F.
Mahon, A. S. Weller, Organometallics 2002, 21, 2856–2865.
[28] G. L. Moxham, T. M. Douglas, S. K. Brayshaw, G. Kociok-Kohn, J. P.
Lowe, A. S. Weller, Dalton Trans. 2006, 5492–5505.
minor isomer 8cACHTRE[UNG CbCl6] present in 5% relative concentration at
289 K.
[44] a) C. P. Casey, D. M. Scheck, J. Am. Chem. Soc. 1980, 102, 2723–
2728; b) J. W. Suggs, M. J. Wovkulich, S. D. Cox, Organometallics
1985, 4, 1101–1107; c) K. I. Goldberg, R. G. Bergman, Organometal-
lics 1987, 6, 430–432.
[45] a) Y. X. Chen, T. J. Marks, Chem. Rev. 2000, 100, 1391–1434; b) A.
Macchioni, Chem. Rev. 2005, 105, 2039–2073; c) P. S. Pregosin,
P. G. A. Kumar, I. Fernꢁndez, Chem. Rev. 2005, 105, 2977–2998.
[46] S. P. Smidt, N. Zimmermann, M. Studer, A. Pfaltz, Chem. Eur. J.
2004, 10, 4685–4693.
[29] M. Utsunomiya, R. Kuwano, M. Kawatsura, J. F. Hartwig, J. Am.
Chem. Soc. 2003, 125, 5608–5609.
[30] W. Lesueur, E. Solari, C. Floriani, A. ChiesiVilla, C. Rizzoli, Inorg.
Chem. 1997, 36, 3354–3362.
Received: April 17, 2008
Published online: July 30, 2008
Chem. Eur. J. 2008, 14, 8383 – 8397
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8397