Oxorhenium complexes as aldehyde-olefination catalysts

Article abstract of DOI:10.1002/chem.200400296

Several oxorhenium compounds in the formal oxidation states V and VII are examined as catalysts for the aldehyde-olefination starting from diazo compounds, phosphines, and aldehydes. Of these, [ReMeO₂(η ²-alkyne)] complexes provide the simplest catalysts to study, although [ReOCl₃(PPh₃)₂] still remains the most efficient rhenium catalyst for aldehydeolefination described to date. Prior to the reaction with the Re catalysts the phosphine and the diazo compound react to form a phosphazine. No catalytic reaction occurs in cases where no phosphazine formation is observed. The first step of the catalytic cycle involves the formation of a carbene intermediate by the reaction of phosphazine and catalyst under extrusion of phosphine oxide and dinitrogen. In a second step the carbene reacts with aldehyde under olefin formation and catalyst regeneration. Excess of alkyne as well as the presence of ketones slows down the catalytic reaction. The olefination of 4-nitrobenzaldehyde with diazomalonate is possible with these Re catalysts. In contrast, this reaction does not take place either in the classical Wittig fashion from Ph₃P=C(CO₂Et)₂ and aldehyde or by use of all other catalysts for aldehyde olefination reactions reported to date. Catalytic ylide formation from diazo compounds seems therefore not to be the only pathway through which catalytic aldehyde-olefination reactions can proceed.

Full text of DOI:10.1002/chem.200400296

FULL PAPER
Oxorhenium Complexes as Aldehyde-Olefination Catalysts
[
a, b]
[a]
[a]
[b]
Ana M. Santos,
Filipe M. Pedro, Ameya A. Yogalekar, Isabel S. Lucas,
[a]
[
b]
Carlos C. Rom¼o,* and Fritz E. Kühn*
Abstract: Several oxorhenium com-
pounds in the formal oxidation states v
and vii are examined as catalysts for
the aldehyde-olefination starting from
phosphazine formation is observed.
The first step of the catalytic cycle in-
volves the formation of a carbene inter-
mediate by the reaction of phosphazine
and catalyst under extrusion of phos-
down the catalytic reaction. The olefi-
nation of 4-nitrobenzaldehyde with di-
azomalonate is possible with these Re
catalysts. In contrast, this reaction does
not take place either in the classical
Wittig fashion from Ph P=C(CO Et)
2
diazo compounds, phosphines, and al-
2
dehydes. Of these, [ReMeO (h -
phine oxide and dinitrogen. In a
2
3
2
alkyne)] complexes provide the sim-
plest catalysts to study, although [Re-
OCl (PPh ) ] still remains the most ef-
ficient rhenium catalyst for aldehyde-
olefination described to date. Prior to
the reaction with the Re catalysts the
phosphine and the diazo compound
react to form a phosphazine. No cata-
lytic reaction occurs in cases where no
second step the carbene reacts with al-
dehyde under olefin formation and cat-
alyst regeneration. Excess of alkyne as
well as the presence of ketones slows
and aldehyde or by use of all other cat-
alysts for aldehyde olefination reac-
tions reported to date. Catalytic ylide
formation from diazo compounds
seems therefore not to be the only
pathway through which catalytic alde-
hyde-olefination reactions can proceed.
3
3 2
Keywords: aldehydes · homogene-
ous catalysis · olefination · rhenium ·
Wittig reactions
Introduction
In some cases they operate under very mild conditions
room temperature), short reaction times (even belowone
(
The olefination of aldehydes and ketones is a very impor-
tant transformation in organic synthesis. In spite of the exis-
tence of highly efficient and general stoichiometric methods
hour nearly quantitative yields are reached), and usually
high E-selectivities. However, the mechanism of the reac-
tions is still a matter of open debate. The key role of car-
bene complexes is meanwhile well accepted and has been
recently established for certain Re and Fe systems by inde-
[
1,2]
such as the Wittig reaction and its several variations,
a se-
lective catalytic system would be highly desirable to speed
up reactions, shorten experimental multistep procedures,
[6,9]
pendent groups.
However, the situation is still unclear
[3]
and avoid the use of expensive reagents difficult to handle.
The quest for effective catalysts has been developing
with respect to the fate and mode of reaction of these car-
bene complexes. In some cases they are considered to react
with coordinated aldehyde with formation of a metallaoxe-
[4]
slowly since the late 1980s. Presently, several efficient cata-
lytic aldehyde-olefination reactions have been reported
[10]
tane ring, whereas in other cases evidence suggests that
they give rise to free ylides that perform the Wittig reaction
in a classical, uncatalyzed fashion. In other words, it is the
formation of the ylides that is catalyzed, not their reaction
with the aldehydes. Given the variety of metals and systems
studied, it may indeed be the case that both possibilities are
true, depending on the actual catalytic system.
[5,6]
[7,8]
[8a]
[9a–e]
[9f]
based on Re,
Ru,
Rh, Fe,
and Co complexes.
[
a] Dr. A. M. Santos, Dipl.-Chem. F. M. Pedro, MSc. A. A. Yogalekar,
Priv.-Doz. Dr. F. E. Kühn
Lehrstuhl für Anorganisch Chemie der Technischen Universität Mün-
chen
Lichtenbergstrasse 4, 85747 Garching bei München (Germany)
Fax : (+49)89-289-13473
[5,11]
Herrmann and co-workers as well as others
have
shown that [ReMeO ] (MTO) and [ReOCl (PPh ) ] are
3
3
3 2
E-mail: fritz.kuehn@ch.tum.de
good catalysts for aldehyde-olefination according to Equa-
tion (1).
[
b] Dr. A. M. Santos, Dipl.-Chem. I. S. Lucas, Prof. Dr. C. C. Rom¼o
Instituto de Tecnologia Química e Biológica da
Universidade Nova de Lisboa, Quinta do MarquÞs
EAN, Apt 127, 2781–901 Oeiras (Portugal)
Fax : (+351)214-411-277
RHCO þ N CHðCO EtÞ þ PPh ! RHC¼CHðCO EtÞ
2
2
3
2
ð1Þ
þN þ OPPh
E-mail: ccr@itqb.unl.pt
2
3
Chem. Eur. J. 2004, 10, 6313 – 6321
DOI: 10.1002/chem.200400296
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6313

FULL PAPER
They also showed that [Cp*ReO ], [CpReO ] and [(tBu -
Results and Discussion
3
3
2
bipy)ReO ] display a lower activity in aldehyde-olefination.
3
In the case of these trioxo complexes it was also shown that
Survey of oxorhenium complexes for aldehyde-olefination
catalysis
PPh is necessary to reduce the trioxorhenium(vii) species to
3
an active dioxorhenuim(v) species, which in the case of
MTO was isolated as [ReMeO (PR ) ·ReMeO ] (R=Ph,
Cy). In a preliminary communication, we were able to con-
The complexes 1–19 were preliminarily screened as catalysts
in aldehyde-olefination according to Equation (1), with 4-ni-
trobenzaldehyde (4-nba), ethyldiazoacetate (eda), and PPh3.
A selection of these complexes was further tested under
controlled standardized conditions to allowmeaningful com-
parisons. These latter results for 1–9 are summarized in
Table 1 for the olefination of 4-nitrobenzaldehyde (4-nba)
2
3
2
3
firm the activity of these dioxorhenuim(v) complexes, by
2
using [MeReO (h -RCꢀCR)] as efficient catalysts for olefi-
2
[6a]
nation of activated aldehydes. We also proposed that reac-
tion (1) is initiated by the intermediate formation of a phos-
phazine, Ph PN=NCH(CO Et).
3
2
In the present study we explore the activity of other types
with ethyldiazoacetate and PPh at room temperature, and a
3
VII
V
V
of Re O , Re O , and Re O complexes in this reaction
reaction time of 2 h. TOF values were recorded after a reac-
tion time of 5 min. Details are given in the Experimental
Section.
The trioxorhenium(vii) complexes 4 and 5 showsimilar
yields and TOF values. These are, however, only about half
of those observed for MTO, a coordinatively unsaturated
3
2
and present detailed investigations on the influence of
changing the nature of the reaction components and condi-
tions on catalytic activity, to attempt to establish a better un-
derstanding of the olefination reaction catalyzed by oxorhe-
nium complexes.
6314
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 6313 – 6321

Aldehyde-Olefination Catalysts
6313 – 6321
Table 1. Catalytic results for compounds 1–9 in the olefination of 4-nba
at room temperature; 5 mol% catalyst. The yields obtained with com-
pounds 10–19 are below5% and therefore are considered negligible, the
The prototypical catalyst in the ReX O family is [Re-
3
OCl (PPh ) ], by far the most active Re catalyst for alde-
3
3 2
ꢁ
1
ꢁ1
hyde-olefination. Under appropriate circumstances (reaction
TOFs are below10 molmol
h .
temperature: 508C, catalyst/substrate=1:1000) it can reach
[a]
[b]
ꢁ1 ꢁ1
Catalyst
Yield [%]
TOF [molmol h ]
ꢁ
1
TOF values of 1200h . Replacement of PPh by the biden-
3
1
2
3
4
5
6
7
8
9
80
77
70
41
46
82
14
12
68
150
150
120
70
95
150
30
tate dppe, as in complex (15), totally blocks catalysis sug-
gesting the importance of PPh dissociation as a condition
3
for forming an active species. Also devoid of any catalytic
action in aldehyde-olefination are the complexes [ReO(O-
Me)Cl (PPh ) ] (16), [ReO(OMe)Cl (dppe)] (17: dppe=1,2-
2
3
2
2
3
bis(diphenylphosphino)ethane), [Re(O)Cl {k -B(pz) }] (18),
and [Re(O)(OMe) {k -B(pz) }] (19). The organometallic
2
4
30
120
3
2
4
complexes 7 and 8, bearing bidentate nitrogen donor li-
gands, also showa rather limited catalytic activity. However,
the analogue complex 9, possessing two monodentate tBupy
ligands shows an activity already in the range of the values
for MTO and the MDO derivatives 1–3.
[
a] Yield calculated after a reaction time of 2 h. [b] Calculated after a re-
action time of 5 min.
molecule. The influence of the electronegativity of the chlor-
ine ligand is felt in slightly higher yield/activity counts for Cl
In summary, the crucial factor for an active aldehyde-ole-
fination catalyst on Re basis seems to be high Lewis acidity
(usually Re in higher oxidation states) and either a sterical
unsaturated compound such as MTO or a species, coordinat-
ed with easily removable “protecting ligands” such as the
PPh groups in [ReOCl (PPh ) ], the R C ligands in com-
over CH . More importantly, 6 shows values of yield/activity
3
quite similar to those of MTO. This can be assigned to the
fact that the diazabutadiene ligand is both more labile and a
weaker donor than tBu -bipy. In fact, it is likely that the dia-
2
3
3
3
2
2
2
zabutadiene ligand is at least partially removed from the
metal during the catalytic reaction. Indeed, the reaction
pounds 1–3 or the weak N-donor ligands in compounds 6
and 9. Sterically demanding and strong donor ligands such
[12]
3
of 6 is slowed down by addition of OPPh ; the yield decreas-
as Cp* in [Cp*ReO ] or the {k -B(pz) } ligands in the com-
3
3
4
es to 53%. Exploratory experiments carried out with other
pounds 13, 14, 18, and 19, bidentate aliphatic diimines (as in
11 and 12), bidentate phoshines (as in 15 and 17), or OMe
ligands as in 16 reduce the activity considerably or totally
and are therefore not useful for Re-based catalysts for the
olefination of aldehydes.
diazabutadiene complexes, [(DAB)ReClO ], 11,12 have
3
shown virtually no activity. DAB derivatives with saturated
organic ligands on the nitrogen centers are known to coordi-
nate more strongly to metal atoms in high oxidation
[12]
states. A marginal activity was also observed for the neu-
3
tral complex [ReO {k -B(pz) }] (13; pz=pyrazole ) analogue
The influence of the substrates on the catalytic performance
3
4
2
of [Cp*ReO ]. These results showthat reduced Lewis acidi-
of [ReMeO (h -alkyne)] complexes
3
2
ty and coordinative saturation of the metal centre hamper
the catalytic reaction. Unfortunately, the extreme reactivity
Equation (1) represents a complex multicomponent reaction
involving, at its start, aldehyde, diazo compound, phosphine,
and catalyst precursor. It is, therefore, not easy to predict
howthe reaction starts and what is the influence of each re-
agent in its course. In the following we study the reaction
using catalyst 3 varying the initial reagents to identify their
action and influence in the outcome of the reaction.
of [ClReO ], the limiting reagent in this series, does not
3
allowits use as catalyst.
V
The Re complexes are divided in two groups: ReXO2
and ReX O derivatives.
3
2
The [ReMeO (h -alkyne)] complexes 1–3 perform at
2
levels close to that of MTO. There is a slight dependence on
the nature of the alkyne ligand in the decreasing order of
activity Me C >Et C >Ph C . Excess alkyne slows down
[6a]
The diazo compounds: In our previous study we used 4-
2
2
2
2
2
2
the reaction. In the case of 3 the yield goes down from 70 to
0% when excess Ph C is added. This suggests that the
alkyne ligand has to be displaced during the reaction cycle,
in which case the above reactivity order simply derives from
the increasing stability of the complexes in the order 1<2<
nba as aldehyde, PPh as oxygen abstractor, and eda as stan-
3
5
dard diazo reagent. In this case it was shown that none of
the initial reagents interacts directly with 3. Instead, the for-
mation of phosphazine Ph P=NN=CH(CO Et) from PPh
3
2
2
3
2
and eda precedes and triggers the catalytic reaction. The for-
mation of phosphazine is fast, as can be observed conven-
3
. In fact, 3 can be stored at room temperature for several
3
1
months, whereas 1 and 2 have to be stored belowroom tem-
iently by P NMR spectroscopy (see below). Addition of 3
[13]
perature to avoid decomposition.
to a solution of phosphazine preformed from PPh and eda
3
V
3
Interestingly, both the dimeric Re complex [{Re[k -
leads to a very exothermic reaction and liberation of N . If
2
IV
B(pz) )]O(m-O)} ], (14) and the Re dimer 10 are totally in-
4-nba is present, catalysis is then observed (see belowfor
the characterization of this interaction).
In contrast, the use of trimethylsilyldiazomethane
4
2
active. The formation of a similar dimer in the reaction of
Cp*ReO ] with PPh , namely [{Cp*ReO(m-O)} ] has been
[
3
3
2
held responsible for the lack of aldehyde-olefination reactiv-
N CH(TMS) instead of eda leads to no formation of olefina-
2
[14]
ity of [Cp*ReO ].
Dimerization is not favored for the
tion products, neither for aldehydes nor for ketones with se-
3
[15]
31
[
MeReO ] (MDO) complex.
lected Re complexes (1,3). P NMR spectroscopy also
2
Chem. Eur. J. 2004, 10, 6313 – 6321
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6315

C. C. Rom¼o, F. E. Kühn et al.
FULL PAPER
shows that no phosphazine, Ph P=NN=CH(TMS) is formed
CHCO Et with various aldehydes, but do not catalyze either
3
2
from N CH(TMS) and PPh : the only signal present in the
the reaction of PPh =C(CO Et) with aldehydes or the four-
component reaction of N C(CO Et) with aldehyde and
2 2 2
PPh3.
Taken together, these observations showthat: 1) Phospha-
zine formation from the diazo compound and phosphane
prior to the reaction with the catalyst seems to be an impor-
2
3
3
2
2
31
P NMR spectrum of a 1:1 mixture of these compounds,
[8e]
even after several hours, belongs to free PPh3. Other cata-
[4c,9]
lysts, however, for example, Ru-based systems
generate
aldehyde-olefination products starting from phosphines,
N CH(TMS), and aldehydes, under comparable conditions.
2
We confirmed this by control experiments. These observa-
tions indicate that phosphazine formation is a necessary pre-
requisite for a successful catalytic reaction, regardless of the
fact that phosphazines exist in equilibrium with free diazo
compound and free phosphine.
tant feature for a fast catalytic reaction. 2) The four-compo-
2
nent reaction catalyzed by ReMeO (h -alkyne) is able to
2
form olefins from N C(CO Et) that are not available by the
2
2
2
Wittig reaction of Ph PC(CO Et) with aldehyde nor by the
3
2
2
action of other, usually more active catalysts, such as
[9c–e]
Diazomalonate leads to a lower overall yield of olefin
than eda. With catalyst 3 only 22% is obtained after 24 h.
[Fe(TPP)Cl] that operate via an ylide intermediate.
Therefore, the ReMeO catalysts have the important advan-
2
31
P NMR experiments have shown that in this case the cor-
tage of being able to catalyze aldehyde-olefination reactions
that are not feasible under classical or catalyzed Wittig con-
ditions, that is, from ylide and aldehyde.
responding phosphazine is formed but at a much slower rate
than with eda, even after two days the reaction was not
complete. The fast and quantitative formation of a phospha-
In the four-component reaction of Equation (1) there
must be a catalytic process involved, which allows olefin for-
mation without intermediate ylide formation. The original
assumption of Herrmann and Wang that a rhena-oxethane is
involved as an intermediate for the reaction between alde-
hyde and metal carbene instead of a reaction between metal
carbene and phosphine leading to a phosphorus ylide, which
then would react in a Wittig type reaction seems to be the
simplest explanation that accommodates our findings. Re-
cently Chen et al. also reported experiments pointing to an
intermediary presence of rhenacycles in closely related reac-
zine in the beginning of the catalytic cycle seems decisive
2
for the catalytic performance with [ReMeO (h -alkyne)] cat-
2
alysts. Without formation of phosphazine complexes 1–3 are
seemingly unable to catalyze aldehyde-olefination.
These observations, however, do not rule out the possibili-
ty that concomitant formation of ylides under the reaction
conditions, with or without metal catalysis, may be responsi-
ble for the catalytic activity since they would then react with
the aldehyde either in a metal catalyzed process or in the
classic uncatalyzed Wittig fashion. This possibility has been
clearly observed in other metal-catalyzed olefination reac-
[6c]
tions. Although it has been convincingly shown to operate
in other catalytic aldehyde-olefination systems, a catalytic
reaction yielding an ylide can be by no means the only way
to get to aldehyde-olefination products. Furthermore, for
[6b,9b]
tions according to Equation (1).
There is, however, an important additional set of experi-
ments that demonstrates that in the case of the catalysts ex-
amined here, particularly in the cases of the MDO deriva-
tives 1–3 the catalytic formation of ylides is not the decisive
step of the reaction. No olefin is formed upon reaction of 4-
nitrobenzaldehyde (4-nba) with the ylide Ph P=C(CO Et)
2
the MeReO systems examined in this work we never ob-
2
served the formation of significant amounts of ylides (by
3
1
P NMR spectroscopy) during the course of the catalytic re-
actions.
3
2
in the absence of catalyst. Furthermore, when a catalytic
amount of compound 3 is added to the mixture Ph P=C-
The phosphines: The type of phosphines used for oxygen
abstraction is also a factor of major influence on the catalyt-
ic performance of Equation (1). The reaction of the phos-
phine with the diazo compound is fast in the case of ethyl-
3
(
CO Et) /4-nba, no olefination products are obtained even
2 2
after a reaction time of four days. As described already
above, however, compound 3 catalyzes the reaction of 4-
nba, PPh , and diazomalonate. This reaction is slow, but
diazoacetate (eda) and PPh . In less than 5 min all starting
3
3
reaches good yields after several days. Accordingly, com-
materials are completely consumed. Phosphazine, Ph P=
3
pound 3 does not initiate or catalyze the reaction of Ph P=
NN=CH(CO Et) is the only product formed according to
3
2
3
1
C(CO Et) with 4-nba and an olefination mechanism operat-
P NMR spectroscopy. The chemical shift observed in the
P NMR spectrum, d( P)=23.3 ppm is identical within the
2
2
3
1
31
ing by the catalytic formation of ylides and the further reac-
tion of these ylides with aldehydes (both catalyzed or un-
catalyzed) can be excluded at least for Ph P=C(CO Et) .
error range to the value found in the literature for this phos-
[
17]
1
phazine.
The signal for the a proton of eda at d( H)=
3
2
2
1
Therefore, it is important to point out that olefination of 4-
nba with N C(CO Et) /PPh according to Equation (1) is
4.69 ppm is replaced by a newsignal at d( H)=7.70 ppm, in
full accordance with the literature data for such reactions.
[18]
2
2
2
3
feasible albeit slow, showing an intrinsic advantage of the
catalytic reaction (1) over the classical Wittig reaction.
As already mentioned, in the case of some other alde-
hyde-olefination catalysts the intermediate formation of
The presence of certain amounts of free phosphine and
diazo compound due to equilibrium reactions does not influ-
ence the proposal that the phosphazine is the important spe-
cies for the catalytic reaction. As we have shown
N CH(TMS), which does not react with PPh to form a
[6b,9b]
ylides has been clearly observed.
We have selected
2
3
II
II
some of those Ru - and Fe -based aldehyde-olefination cat-
phosphazine under the conditions applied, also fails to start
the catalytic cycle, although—of course—plenty of free PPh3
[16]
[9c–e]
alysts, namely [Cp*RuCl(PR ) ]
and [Fe(TPP)Cl]
TPP=tetra(p-tolyl)porphyrin), and observed that they cat-
alyze (very efficiently) the Wittig reaction of PPh =
3
2
(
and N CH(TMS) is available. When 3 is added to a solution
2
containing the phosphazine in a 1:10 ratio, a rapid reaction
3
6316
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 6313 – 6321

Aldehyde-Olefination Catalysts
6313 – 6321
3
1
ensues with liberation of N and an increase in temperature.
d( P)=20.2 ppm, assigned to the corresponding phospha-
2
3
1
Already after 5 min the P NMR spectrum of the mixture
reveals a significant decrease of the peak corresponding to
the phosphazine as well as the appearance of a new peak at
zine, the signal corresponding to the triethylphosphite being
found at d( P)=139 ppm. However, in this case the reac-
tion is not complete, even after days.
Summarizing the above results it can be again said that
formation of phosphazine is essential to start aldehyde-olefi-
nation in the four-component system eda, PR , 4-nba,
ReO Ln catalyst. Free ylides are only detected in small
amounts at high catalyst concentration in the absence of al-
dehyde.
3
1
31
d( P)=29 ppm (triphenylphosphine oxide). The peak corre-
sponding to the OPPh3 increases with time, limited in
growth by the amount of catalyst present. Small peaks at
3
31
d( P)=18.68 and 17.10 ppm indicate the formation of ylides
Ph P=CH(CO Et) (cisoid and transoid form). In this case,
x
3
2
since the amount of catalyst present is higher than in the
usual catalytic runs and in the previously described experi-
ment, the peak of the phosphazine is significantly reduced
in size and the ylides are formed in a 1:10 ratio to OPPh3.
Therefore, again, the catalytic ylide formation can not be
The carbonyl compounds: Compounds 2 and 3 were tested
as olefination catalysts for 4-nitrobenzaldehyde, 4-bromo-
benzaldehyde, and benzaldehyde using the following reagent
the major reaction with the Re catalysts of type
ratios: aldehyde/PPh /eda/catalyst=1:1.1:1:0.05. In all three
3
2
[
ReMeO (h -alkyne)]. It should be noted that all these reac-
cases compound 2 as the catalyst leads to the best product
yields after 24 h. With 4-bromobenzaldehyde it takes more
time to reach the same yields than with 4-nitrobenzaldehyde
as substrate under the same reaction conditions. After a re-
action time of 24 h, however, in both cases high yields of
70–82% are reached. In the case of benzaldehyde the yields
after 24 h are about 50% both with complex 2 and 3 as the
catalysts. Interestingly, in the case of 4-nitrobenzaldehyde
the product yield increase from 2 to 24 h is only marginal
(5% or less), whereas in the case of 4-bromobenzaldehyde
and benzaldehyde the yield approximately doubles in the
same period of time (see Figure 1).
2
3
1
tions occur quickly and the P NMR spectra do not change
significantly from a reaction of 10 min to 4 h, which is in
good accord with the catalysis results of the catalysts 1–3,
where after 30 min the yield is close to the final yield in
most cases. After the reaction stops, the peaks obtained in
31
the P NMR spectrum, correspond to the following species:
unreacted phosphazine, triphenylphosphine oxide and the
[9a]
ylides, Ph P=CHCO Et, with signals of low intensity. The
3
2
ylides are formed in small amounts either by decomposition
[17]
of the phosphazine or by reaction of PPh with some Re=
3
CH(CO Et) species formed during the course of the reac-
2
tion, that cannot be consumed otherwise in the absence of
aldehydes.
To confirm the latter point, the reaction was also moni-
3
1
1
tored in the presence of an aldehyde by P and H NMR
spectroscopy. The reaction mixture consisted of PPh , eda, 3,
3
4
-nba, 1.1:1:0.07:1, which were added in this order. In the
P NMR spectrum recorded after a reaction time of 5 min
3
1
only one significant signal is present, corresponding to
OPPh . The ylides are not observed even in trace amounts,
3
and within 4 h no further changes are detected in the spec-
trum.
MTO and compounds 3 and 5 were tested by using
P(OEt) as deoxygenating agent in catalytic aldehyde-olefi-
3
nations according to Equation (1) with 4-nba and eda under
the same conditions as described above for PPh . The cata-
3
lytic activity decreases significantly (Table 2). An analogous
Figure 1. Catalytic activity of compounds 2 and 3 with benzaldehyde, 4-
bromobenzaldehyde and 4-nba, respectively. Black columns: yield after
2 h, grey columns: yield after 24 h.
Table 2. Comparison of the catalytic results (olefin yield in %) for MTO
and compounds 3 and 5 in the olefination of 4-nba using two different
oxygen abstraction agents after a reaction time of 2 h.
The reason for this observation in the case of 4-nitroben-
zaldehyde may be related to the fast accumulation of the
Compound
PPh
3
cis/trans
P(OC
2
H
5
)
3
cis/trans
MTO
3
5
77
70
46
10/90
5/95
10/90
17
7
15
22/78
29/71
10/90
by-product OPPh that significantly blocks the reaction at
3
an early stage. In fact, if extra OPPh is added in the begin-
3
ning of the catalysis the yield is reduced significantly, as well
as the reaction rate.
observation was made with [ReOCl (PPh ) ] as the catalyst.
This is very likely due to the fact that phosphazine forma-
The use of the ketones acetophenone, acetone, and benzo-
phenone as substrates instead of the aldehydes does not
lead to significant olefin yields within a reaction time of
24 h with both catalysts 2 and 3. Notably, benzophenone
does not react with the ylide Ph P=CHCO Et even after a
3
3 2
tion is much slower with triethylphosphite than with PPh ,
3
3
1
as can be seen from P NMR measurements.
In fact, reacting triethylphosphite with eda gives rise in
3
2
3
1
the P NMR spectra to the appearance of a newsignal at
reaction time of eight days, whereas, this ylide reacts
Chem. Eur. J. 2004, 10, 6313 – 6321
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6317

C. C. Rom¼o, F. E. Kühn et al.
FULL PAPER
promptly with 4-nitrobenzaldehyde with quantitative forma-
tion of the olefin. It is therefore clear that electron-poor car-
bonyls react faster under the present catalytic olefination
conditions.
stead of 70%). The ratios of the substrate were changed and
runs with 4-nba/PPh /eda/catalyst ratios of 1:2:1:0.05 and
3
1:1.1:2:0.05 were performed exemplary for compound 3. The
catalytic performance was not affected by the presence of
The interaction of different aldehydes with catalyst 3 was
excess of PPh or eda. In the presence of an excess of eda
the formation of diethyl maleate as by-product is observed.
3
1
13
also examined by H and C NMR spectroscopy in CDCl3.
When the components were reacted in a 2:1 and 1:1 ratio
no change in the chemical shifts could be observed since the
deviations from the signals of the isolated compounds
always stayed significantly below 0.1 ppm. Addition of a
ketone, such as acetophenone, however, leads to a shift of
Influence of the solvent, catalyst amount and reaction tem-
perature: When the aldehyde-olefination with catalyst 3 is
conducted in toluene instead of THF the olefin yield de-
creases from 70% to 52% after 2 h. The results in CH Cl ,
2
2
1
both the Re-bound methyl group (from d( H)=2.61 ppm in
CHCl , and CH CN were 26, 24, and 38% respectively after
3 3
1
the free molecule to d( H)=2.49 ppm in the mixture) and
2 h. In this case the same amounts of azine and olefin are
formed. The cis/trans selectivity also decreases to 10:90 in
the case of CHCl and CH CN. When the catalyst charge is
1
the CH3 protons of the ketone from d( H)=2.55 ppm to
1
d( H)=2.51 ppm. To examine this behavior in more detail,
3
3
we reacted MTO and polymer-bound triphenylphosphine,
the usual method of preparation of the alkyne adducts of
MDO 1–3. However, instead of offering an alkyne as reac-
tion partner we added—under the same conditions—aceto-
phenone. The reaction mixture turned yellowand the NMR
signals of both the ketone and the Re-bound methyl group
changed as observed in the way described above. Unfortu-
nately, we were unable to isolate the reaction product due
to its sensitivity to both temperature and moisture. Also,
when acetophenone is added to a catalytic aldehyde-olefina-
decreased to 2 mol% (instead of 5 mol%) the olefin yield
(cis and trans) obtained decreases to 55%, for 1 mol% to
40%, and for 0.5 mol% to 20% yield. Azine is the main by-
product. Catalytic runs were performed with catalyst 3 at 0,
20, 50, and 678C in THF. At 08C the yield reaches 65%
after 2 h only slightly belowthe yield achieved at room tem-
perature. At 508C the yield reaches after 30 min the same
value obtained after 2 h at room temperature. The TOF (de-
termined after a reaction time of 5 min) increases from
ꢁ
1
ꢁ1
about 120h at 208C to about 170h at 508C. A further
raise in the reaction temperature, however, does not acceler-
ate the reaction significantly, the TOF staying below200
tion run (4-nba, eda, PPh as the substrates) with compound
3
3
as catalyst in an acetophenone/catalyst ratio of 1:1, the re-
ꢁ
1
action rate and yield decrease somewhat (yield obtained
after 2 h: 65%). However, after 24 h the same yield is ach-
ieved as in the absence of ketone. These results suggest that
1h .
Catalyst–phosphazine interaction: As shown above, the cat-
alysts 1–3 interact strongly with phosphazine but seemingly
not with the other components of the catalytic reaction mix-
ture or with ylides. Upon reaction of 1–3 with phosphazine,
N and OPPh are liberated and a Re-carbene complex is
V
ketones are able to coordinate to the Re center of
2
[
ReMeO (h -alkyne)] in competition with the substrates of
2
catalysis, namely the phosphazine, thus blocking the catalyt-
ic site.
2
3
formed. A terminal oxygen atom of the catalyst is abstracted
1
7
Substrate ratios: In the blank runs performed in the absence
of catalyst no olefination products were formed for all
tested aldehydes. In all cases azine, RC H CH=NN=
under formation of phosphine oxide as observed by O-
1
7
NMR spectroscopy from a reaction with O-labeled 3* pub-
[6a]
lished previously.
6
4
CH(CO Et) was the main product. The formation of diethyl
In the case of catalyst 3 at ꢁ308C, the carbene species
2
13
maleate or diethyl fumarate was not observed in a detecta-
ble amount (in the catalytic runs these products were also
not detected). The formation of azine results from a com-
peting reaction between the initially formed phosphazine
and the aldehyde, which is not influenced by the presence of
the catalyst.
Since this reaction is slower than the reaction between the
catalyst and phosphazine, if the catalyst is present in suffi-
cient amount, the reaction between the phosphazine and the
aldehyde does not occur to a significant extent, and the
azine is a minor side product. The order by which the cataly-
sis components of Equation (1) are added, can be varied
without negative effects on the catalytic performance, pro-
vided the eda and the aldehyde are not added in the ab-
sence of catalyst (see above). This means, either all compo-
nents can be mixed and eda added as the last reaction com-
ponent, or both the catalyst and the aldehyde can be added
after mixing all other reaction components. This latter pro-
cedure in the case of compound 3 even leads to slightly
higher yields after a reaction time of two hours (78% in-
was characterized by the signals at d( C)=323 ppm and
1
Re=CꢁH at d( H)=16.32 ppm both in accordance with the
values reported in the literature for related complexes, such
[19]
as [RuClCp(=CHCO Et)(PPh )].
The corresponding sig-
2
3
nals of the carbene complexes formed from compounds 1
and 2 are, as expected, identical within the measurement
V
error. This Re carbene complex is observable in situ only
at lowtemperatures (see Scheme 1 for its possible formula).
At room temperature and above it decomposes quickly if no
aldehyde is available for the continuation of the catalytic
cycle. In fact, when the carbene species, formed from cata-
lyst and phosphazine at lowtemperature, is allowed to react
with aldehyde, fast formation of olefin and the re-formation
of the catalyst can be observed by NMR spectroscopy.
The involvement of Re carbene complexes in the catalytic
aldehyde-olefination has independently been confirmed by
Chen et al. for Re O -based, cationic catalysts in the gas
2
7
phase by mass spectrometry, although in their systems a cat-
alytic ylide formation (following the carbene formation)
[6b]
seems to take place.
6318
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 6313 – 6321

Aldehyde-Olefination Catalysts
6313 – 6321
ReMeO species able to re-enter the catalytic cycle. It can
2
not be excluded that alternatively a reaction of a Re-coordi-
nated triphenylphosphine in IIIa leads to the formation of
[6b,c]
an ylide, as has been suggested by Chen et al.
and that
the ylide then reacts with an aldehyde in a classical Wittig
reaction. This possibility, however, is very unlikely to occur
with diazo malonate derived compounds, since otherwise no
products other than ylides would be obtained. Furthermore,
we did not observe large amounts of ylides, even in the ab-
sence of aldehydes, which would be necessary, however, if
the catalyzed reaction would be a ylide formation. There-
fore, formation of the metallacycle V is, in our view, the sim-
plest way to accommodate the fact that olefination of 4-nba
is still possible with N C(CO Et) . Since a phosphazine is
2
2
2
formed from PPh and N C(CO Et) , the reaction just fol-
3
2
2
2
lows the ordinary course albeit more slowly since this phos-
phazine is more difficult to form, and other electronic or
steric factors may also play a role. If the reaction were to
proceed by ylide formation, it would not take place since
Ph P=C(CO Et) does not react with aldehydes, even 4-nba.
3
2
2
Scheme 1. Catalytic cycle for aldehyde-olefination with [ReMeO
plexes (L=alkyne, PPh ) with PPh and N CR R .
3 3 2
2
L] com-
II
II
Moreover, Ru and Fe catalysts known to operate by ylide
formation are not active catalysts for olefination with
N C(CO Et) . Therefore, this catalytic system, although not
1
2
2
2
2
The catalytic cycle for aldehyde-olefination with
ReMeO L] complexes: Based upon the experimental re-
sults presented before, a catalytic cycle describing the reac-
tion of Equation (1) can be postulated as depicted in
being as active as some other aldehyde-olefination catalyst
systems, is useful for the olefination of aldehydes that are in-
capable of reacting with ylides, thus circumventing the lack
of reactivity of electron-poor ylides. In fact, this catalyst
presents one of the key characteristics desirable in avoiding
the drawbacks of the classical Wittig reaction: the total ab-
sence of basic, carbanionic intermediates, like ylides.
In spite of its interest, this system is, unfortunately, still in-
capable of promoting ketone-olefination and of activating
electron-rich diazo compounds, for example, the synthetical-
[
2
Scheme 1 taking as examples, catalysts 1–3, PPh , eda and 4-
3
nba as reaction partners. An important point is the observa-
tion that we only obtained olefination of aldehydes in the
cases where the phosphine reacts with the diazo compound
to form a phosphazine. Besides, catalysts 1–3 do not interact
directly with aldehydes, diazo compounds, ylides, or PPh3,
whereas a rapid reaction is observed between these and
ly useful N CH(TMS). Ketones were found to compete for
2
phosphazine.
the catalytic intermediates, slowing down but not blocking
aldehyde-olefination. This suggests that they do not disrupt
the catalytic cycle by deactivating its most reactive species,
namely the carbene III. Therefore, it seems likely that they
do not have the electronic characteristics necessary to either
form metallacycle V or lead to its productive (clockwise
2
Therefore, the first step is the interaction of [ReMeO (h -
2
1
2
alkyne)] with the phosphazine Ph P=NꢁN=CR R ] leading
3
to the species I or Ia. This reaction is most probably accom-
panied by displacement of the alkyne ligand, since this one
retards the reaction rate when present in excess. Admitting
total displacement of the alkyne, and formation of I, the
next step depicts the intramolecular arrangement of II that
sense) disruption. Further studies are necessary in order to
V
eventually tune the reactivity of the Re O derivatives for
2
precedes Ph PO elimination.
A
six-membered ring is
enlarging their scope of applications in olefination reactions.
3
formed in which the positively charged P atom faces the
negatively charged O ligand. The ensuing liberation of
Ph PO will lead also to N liberation and formation of the
Conclusion
3
2
carbene complex III. Again, this coordinatively unsaturated
species may be stabilized by any ligand present in the reac-
tion mixture (PPh , OPPh , alkyne) leading to the formation
Several oxorhenium compounds act as catalysts for the alde-
hyde-olefination starting from diazo compounds, phos-
phines, and aldehydes. Steric and electronic reasons allow
3
3
13
of IIIa. C NMR spectroscopy performed in a test reaction
that was forced to stop at this stage due to absence of alde-
hyde, showed a signal for the carbene-C atom in the range
expected for this kind of species. In a catalytic reaction,
the selection of “ideal” catalysts quite easily. Among the
2
Re-based catalysts [ReMeO (h -alkyne)] complexes provide
2
the simplest catalysts to study, although they are not the
most active ones. The first step of the catalytic cycle involves
the formation of a carbene intermediate by the reaction of
preformed phosphazine and catalyst under extrusion of
phosphine oxide and dinitrogen. In a second step the car-
bene reacts with aldehyde under olefin formation and cata-
lyst regeneration. Formation of significant amounts of ylides
3
however, aldehyde R CHO is present in large amounts and
may replace any of the L ligands in IIIa. This leads to com-
plex IV. Coordination will impart increased electrophilicity
to the carbonyl-carbon atom facilitating the formation of
the metalla-oxetane ring V. Splitting of this ring liberates
the olefin, while reforming the starting complex or simply a
Chem. Eur. J. 2004, 10, 6313 – 6321
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6319

C. C. Rom¼o, F. E. Kühn et al.
FULL PAPER
as intermediates is not observed. Furthermore, the olefina-
tion of diazomalonate can be catalyzed, whereas the corre-
sponding ylide does not react in a Wittig reaction in the
presence of the same Re compound. It can therefore be as-
sumed that the catalytic formation of ylides does not play a
dominant role in the catalytic aldehyde-olefination with the
Catalytic ketone-olefination: Dry acetone (15.82 g, 0.272 mol), PPh
3
(
1.88 g, 7.2 mmol), compound 3 (0.148 g, 0.36 mmol), and eda (0.904 g,
7
.92 mmol) were refluxed at 1008C for 48 h. The reaction mixture was
distilled until no more acetone came out and the residue was treated
with dry hexane and cooled. The solution was filtered and concentrated
to yield a crude oil. The oil was chromatographed in silica gel over
hexane, and the product was obtained from the required fractions as a
colorless oil.
ReMeO type catalysts. This does, of course, not exclude
2
Catalytic aldehyde-olefination usingdiazomalonate
:
4-nba (0.5 g,
that other catalysts lead to olefin formation by an entirely
or partially different pathway in which ylide intermediates
play a prominent role.
3
.3 mmol), PPh (0.95 g, 3.6 mmol), fluorene (0.4 g, internal standard),
3
compound 3 (0.068 g, 0.166 mmol), and ethyl diazomalonate (0.739 g,
3.97 mmol), were dissolved in dry THF (20 mL) and allowed to react at
room temperature.
Experimental Section
All reactions were carried out under an argon atmosphere by using stan-
dard Schlenk techniques. Solvents were dried by conventional methods
and distilled under nitrogen before use. All compounds were purchased
Acknowledgement
The authors are greatly indebted to the Fonds der Chemischen Industrie,
FSE and the Fundażo para a CiÞncia e Tecnologia (FCT) projects
PRAXIS XXI/P/QUI/10047/1998 and POCTI/37726/QUI/2001 for finan-
cial support including a grant to I.S.L. A.M.S. is grateful to the FCT and
the Alexander von Humboldt Foundation for postdoctoral research fel-
lowships, F.M.P. for a Ph.D. grant from the Gulbenkian Foundation and
the FCT, and A.A.Y. to the German Institute of Science and Technology
[
20]
from Aldrich unless stated otherwise. The ligand p-tolyl-DAB,
com-
[Re-
[
13]
[21]
[22]
[23]
[13]
[24]
pounds 1–3,
4,
5,
7ꢁ9,
10,
3 3 2],
[ReOCl (PPh )
[
25]
[26]
[27]
[25]
[28]
[25]
[27]
[27]
3
OCl (dppe)], 13, 14, 15, 16, 17, 18, and 19 were syn-
thesized according to published procedures. Complexes 13–19 were a
kind offer from Dr. Isabel Santos and Dr. António Paulo from the Insti-
tuto de Tecnologia Nuclear, SacavØm, Portugal. Elemental analysis were
obtained at the ITQB by Conceiżo Almeida. IR spectra were recorded
on a Perkin-Elmer FT-IR spectrometer by using KBr pellets as the IR
(
GIST) for financial support. The authors are indebted to Dr. I. Santos
and Dr. A. Paulo for a generous gift of samples.
1
31
13
matrix. H, P, and C NMR spectra were obtained by using a 400 MHz
Bruker Avance DPX-400 and a 300 MHz Bruker CPX 300 spectrometer.
3
1
8
5% H
ments.
Preparation of [ReClO
All these complexes were prepared in a similar fashion described in
detail for DAB=4-MeC N=CHCH=NC Me-4) (6)
Preparation of 6: A solution of Re (0.400 g, 0.826 mmol) in THF
15 mL) was treated with Me SiCl (0.210 mL, 1.652 mmol, d=0.856).
3 4
PO was used as an internal standard for the P NMR measure-
[
1] a) G. Wittig, G Geissler, Liebigs Ann. Chem. 1953, 44, 580; b) G.
Wittig, U. Schçllkopf, Chem. Ber. 1954, 87, 1318; c) G. Wittig, M.
Schlosser, Tetrahedron 1962, 1023.
3
(DAB)] complexes
[2] a) B. E. Maryanoff, A. B. Reiz, Chem. Rev. 1989, 89, 863; b) S. E.
Kelly in Comprehensive Organic Synthesis, Vol. 1 (Eds.: B. M. Trost,
I. Fleming, S. L. Schreiber), Pergamon, Oxford, 1991, p. 729;
c) A. R. Maguire in Comprehensive Organic Functional Group
Transformations, Vol. 1 (Eds.: A. R. Katritzky, O. Meth-Con, C. W.
Rees), Pergamon, NewYork, 1995, p. 589; d) I. Gosney, D. Lloyd in
Comprehensive Organic Functional Group Transformations, Vol. 1,
(Eds.: A. R. Katritzky, O. Meth-Con, C. W. Rees), Pergamon, New
York, 1995, p. 719.
[3] See for example: a) R. R. Schrock, J. Am. Chem. Soc. 1976, 98,
5399; b) F. N. Tebbe, G. W. Parshall, G. S. Reddy, J. Am. Chem. Soc.
1978, 100, 3611; c) S. H. Pine, R. Zahler, D. A. Evans, R. H. Grubbs,
J. Am. Chem. Soc. 1980, 102, 3270; d) K. A. Brown-Wensley, S. L.
Buchwald, L. Cannizzo, L. Clawson, S. Ho, D. Meinhardt, J. R.
Stille, D. Straus, R. H. Grubbs, Pure Appl. Chem. 1983, 55, 1733;
e) F. W. Hartner, J. Schwarz, S. M. Clift, J. Am. Chem. Soc. 1983,
6
H
4
6 4
H
2
O
7
(
3
Two equivalents of the ligand were then added (0.389 g, 1.652 mmol).
The orange precipitate that formed immediately was filtered from the
red mother liquor, washed with diethyl ether and dried under vacuum.
Yield 63%. IR selected (KBr pellets): n˜ =1597 m, 1500 s, 947 s, 922 vs,
ꢁ
1
1
9
(
10 s, 896 vs, 858 s, 814 cm s; HNMR (CDCl
s, 2H), 7.65–7.12 (m, 8H), 2.41–2.24 ppm (m, 6H); elemental analysis
calcd (%) for C16 ClRe (505.978): C 37.98, H 3.19, N 5.54; found:
C 37.78, H, 3.34, N 5.45.
Complexes 11–13: These complexes were prepared in a similar fashion
by using the corresponding ligands, respectively Me CN=CHCH=NCMe
and (C 11)N=CHCH=N(C
Data for compound 11: IR selected (KBr pellets): n˜ =3080 m, 2980 vs,
3
, 300 MHz, RT): d=8.12
16 2 3
H N O
3
3
6
H
6
H11).
1
05, 640; f) T. Kauffmann, B. Ennen, Sander, R. Wieschollek,
Angew. Chem. 1983, 95, 237; Angew. Chem. Int. Ed. Engl. 1983, 22,
44; g) A. Aguero, J. Kress, J. A. Osborn, J. Chem. Soc. Chem.
ꢁ
1
1
2
3
928 s, 2893 s, 983 s, 947 s, 923 vs, 902 vs, 877 cm s; HNMR (CDCl
00 MHz, RT): d=7.51 (s, 2H), 1.19–1.69 ppm (m, 18H); elemental anal-
ClRe (437.94): C 27.43, H 4.60, N, 6.40;
found: C 27.44, H, 5.00, N 6.39.
Data for compound 12: IR selected (KBr pellets): d=3203 m, 2939 vs,
3
,
2
ysis calcd (%) for C10
20 2 3
H N O
Commun. 1986, 531; h) R. H. Grubbs, L. R. Gillom, J. Am. Chem.
Soc. 1986, 108, 733; i) S. H. Pine, Org. React. 1993, 43, 1; j) S. Matsu-
bara, Y. Otake, Y. Hashimoto, K. Utimoto, Chem. Lett. 1999, 747;
k) T. Takeda, K. Shimane, K. Ito, N. Saeki, A. Tsubouchi, Chem.
Commun. 2002, 1974; l) T. Ito, Y. Ishino, T. Mizuno, A. Ishikawa, J.
Kobayashi, Synlett 2002, 2116; m) M. G. Edwards, J. M. J. Williams,
Angew. Chem. 2002, 114, 4934; Angew. Chem. Int. Ed. 2002, 41,
4740; n) D. K. Barma, A. Kundu, H. Zhanh, C. Mioskowski, J. R.
Falck, J. Am. Chem. Soc. 2003, 125, 3218.
ꢁ
1
1
2
858 s, 935 s, 922 vs, 908 s, 898 cm vs; HNMR (CDCl
d=7.72–7.09 (m, 2H), 2.24–1.26 ppm (m, 22H); elemental analysis calcd
%) for C14 ClRe (490.02): C 34.32, H 4.94, N 5.72; found: C
4.45, H 5.05, N 6.01.
Catalytic aldehyde-olefination usingdiazoacetate : 4-Nitrobenzaldehyde
0.6 g, 3.9 mmol), PPh (1.12 g, 4.3 mmol), fluorene (0.4 g, internal stan-
dard), 5 mol% (if not indicated otherwise) of compounds 1–19 and eda
0.40 mL, 3.9 mmol) were dissolved in dry THF (20 mL) and allowed to
3
, 300 MHz, RT):
(
3
24 2 3
H N O
(
3
[4] a) J. A. Smegal, I. K. Meier, J. Schwartz, J. Am. Chem. Soc. 1986,
108, 1322; b) X. Lu, H. Fang, Z. Ni, J. Organomet. Chem. 1989, 373,
77; c) for a recent reviewsee: F. E. Kühn, A. M. Santos, Mini-Rev.
Org. Chem. 2004, 1, 55.
[5] a) W. A. Herrmann, M. Wang, Angew. Chem. 1991, 103, 1709;
Angew. Chem. Int. Ed. Engl. 1991, 30, 1641; b) W. A. Herrmann,
P. W. Roesky, M. Wang, W. Scherer, Organometallics 1994, 13, 4531;
c) W. A. Herrmann in Applied Homogeneous Catalysis with Organo-
metallic Compounds, Vol. 3 (Eds.: B. Cornils, W. A. Herrmann), 2nd
ed., Wiley-VCH, Weinheim, 2002, pp. 1078–1086.
(
react at room temperature. For benzaldehyde and 4-bromobenzaldehyde
the same molar amounts (if not indicated otherwise) were used. Samples
were taken after the first 5 min and then every 30 min for 2 h. The con-
version of aldehyde (4-nitrobenzaldehyde, benzaldehyde, and 4-bromo-
benzaldehyde) and the formation of ethyl-4-nitrocinnamate, ethylcinna-
mate, and ethyl-4-bromocinnamate were monitored by GC and calculat-
2
ed from calibration curves (r =0.999, internal standard fluorene) record-
ed prior to the reaction course.
6320
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 6313 – 6321

Aldehyde-Olefination Catalysts
6313 – 6321
[
6] a) A. M. Santos, C. C. Rom¼o, F. E. Kühn, J. Am. Chem. Soc. 2003,
25, 2414; b) X. Zhang, P. Chen, Chem. Eur. J. 2003, 9, 1852; c) X. P.
Chen, X. Y. Zhang, P. Chen, Angew. Chem. 2003, 115, 3928; Angew.
Chem. Int. Ed. 2003, 42, 3798.
Herdtweck, J. Organomet. Chem. 1988, 355, 297; d) K. P. Gable,
J. J. J. Juliette, M. A. Gartman, Organometallics 1995, 14, 3138;
e) K. P. Gable, J. J. J. Juliette, C. Li, S. P. Nolan, Organometallics
1996, 15, 5250; f) W. A. Herrmann, R. Serrano, H. Bock, Angew.
Chem. 1984, 96, 364; Angew. Chem. Int. Ed. Engl. 1984, 23, 383;
g) F. E. Kühn, W. A. Herrmann, R. Hahn, M. Elison, J. Blümel, E.
Herdtweck, Organometallics 1994, 13, 1601.
1
[
7] O. Fujimura, T. Honma, Tettrahedron Lett. 1998, 39, 625.
8] a) H. Lebel, V. Paquet, C. Proulx, Angew. Chem. 2001, 113, 2971;
Angew. Chem. Int. Ed. 2001, 40, 2887; b) G. A. Grasa, Z. Moore,
K. L. Martin, E. D. Stevens, S. P. Nolan, V. Paquet, H. Lebel, J. Or-
ganomet. Chem. 2002, 658, 126; c) H. Lebel, V. Paquet, Org. Lett.
[
[15] P. Gisdakis, S. Antonczak, N. Rçsch, Organometallics 1999, 18, 5044.
[16] W. Baratta, E. Herdtweck, W. A. Herrmann, P. Rigo, J. Schwarz, Or-
ganometallics 2002, 21, 2101.
2
002, 4, 1671; d) H. Lebel, V. Paquet, Organometallics 2004, 23,
1
187; e) F. E. Kühn, A. A. Jogalekar, A. M. Santos, F. M. Pedro, P.
[17] T. A. Albright, W. J. Freeman, E. E. Schweizer, J. Org. Chem. 1976,
41, 2716.
Rigo, W. Baratta, J. Catal. 2004, 227, 253.
[
9] a) G. A. Mirafzal, G. Cheng, L. K. Woo, J. Am. Chem. Soc. 2002,
[18] H. J. Bestmann, F. M. Soliman, K. Geibel, J. Organomet. Chem.
1980, 192, 177.
1
24, 176; b) G. Cheng, G. A. Mirafzal, L. K. Woo, Organometallics
2
003, 22, 1468; c) Y. Chen, L. Huang, M. A. Ranade, X. P. Zhang, J.
[19] W. Baratta, A. Del Zotto, P. Rigo, Organometallics 1999, 18, 5091.
[20] R. Colton, I.B Tomkins, Aust. J. Chem. 1965, 18, 447.
[21] P. Ferreira, W. M. Xue, E. Bencze, E. Herdtweck, F. E. Kühn, Inorg.
Chem. 2001, 40, 5834.
[22] W. A. Herrmann, F. E. Kühn, C. C. Rom¼o, M. Kleine, J. Mink,
Chem. Ber. 1994, 127.
Org. Chem. 2003, 68, 3714; d) Y. Chen, L. Huang, X. P. Zhang, Org.
Lett. 2003, 5, 2493; e) Y. Chen, L. Huang, X. P. Zhang, J. Org.
Chem. 2003, 68, 5925; f) M. Y. Lee, Y. Chen, X. P. Zhang, Organo-
metallics 2003, 22, 4905; g) W. Sun, F. E. Kühn, Tetrahedron Lett.
2
004, 45, 7415.
[
10] a) C. C. Rom¼o, F. E. Kühn, W. A. Herrmann, Chem. Rev. 1997, 97,
[23] W. A. Herrmann, M. U. Rauch, P. W. Roesky, J. Organomet. Chem.
1996, 511, 299.
3
197; b) W. A. Herrmann, F. E. Kühn, Acc. Chem. Res. 1997, 30,
1
69.
[24] G. W. Parshall, Inorg. Synth. 1977, 17, 110.
[
[
11] B. E. Ledford, E. M. Carreira, Tetrahedron Lett. 1997, 38, 8125.
12] a) F. E. Kühn, A. M. Santos, A. D. Lopes, I. S. GonÅalves, E. Herdt-
weck, C. C. Rom¼o, J. Mol. Catal. A 2000, 164, 25; b) F. E. Kühn,
A. D. Lopes, A. M. Santos, E. Herdtweck, J. J. Haider, C. C.
Romq¼o, J. Mol. Catal. A 2000, 151, 147.
[25] M. Freni, D. Giusto, R. Pierfrancesco, Gazz. Chim. Ital. 1967, 97,
833.
[26] A. Domingos, J. MarÅalo, A. Paulo, A. P. de Matos, I. Santos, Inorg.
Chem. 1993, 32, 5114.
[27] A. Paulo, A. Domingos, J. MarÅalo, A. P. de Matos, I. Santos, Inorg.
Chem. 1995, 34, 2113.
[28] J. E. Fergusson, Coord. Chem. Rev. 1966, 1, 459.
[
13] W. A. Herrmann, J. K. Felixberger, J. G Kuchler, E. Herdtweck, Z.
NaturForsch. B 1990, 45, 876.
[
14] a) W. A. Herrmann, R. Serrano, U. Küsthardt, M. L. Ziegler, E Gug-
golz, T. Zahn, Angew. Chem. 1984, 96, 498; Angew. Chem. Int. Ed.
Engl. 1984, 23, 515; b) W. A. Herrmann, R. Serrano, U. Küsthardt,
E. Guggolz, B. Nuber, M. Ziegler, J. Organomet. Chem. 1985, 287,
Received: March 25, 2004
Revised: August 23, 2004
3
29; c) W. A. Herrmann, M. Flçel, J. Kulpe, J. K. Felixberger, E.
Published online: November 3, 2004
Chem. Eur. J. 2004, 10, 6313 – 6321
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6321