"Catalysis in a tea bag": Synthesis, catalytic performance and recycling of dendrimer-immobilised bis- and trisoxazoline copper catalysts

Article abstract of DOI:10.1002/chem.200900504

Bis- and trisoxazolines (BOX and trisox), containing a linker unit in the ligand backbone that allows their covalent attachment to carbosilane dendrimers, have been employed as polyfunctional ligands for recyclable Cu" Lewis acid catalysts that were immob

Full text of DOI:10.1002/chem.200900504

DOI: 10.1002/chem.200900504
“Catalysis in a Tea Bag”: Synthesis, Catalytic Performance and Recycling of
Dendrimer-Immobilised Bis- and Trisoxazoline Copper Catalysts
Manuela Gaab,[a] Stꢀphane Bellemin-Laponnaz,*[b] and Lutz H. Gade*[a]
5450
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5450 – 5462

FULL PAPER
Abstract: Bis- and trisoxazolines (BOX
and trisox), containing a linker unit in
the ligand backbone that allows their
covalent attachment to carbosilane
dendrimers, have been employed as
polyfunctional ligands for recyclable
CuII Lewis acid catalysts that were im-
mobilised in a membrane bag. The oxa-
zolines contained an alkynyl unit at-
tached to their backbone that was de-
protonated with LDA or BuLi and
then reacted with the chlorosilyl termi-
ni of zeroth-, first- and second-genera-
tion carbosilane dendrimers in the
presence of TlPF6. The functionalised
dendritic systems were subsequently
separated from excess ligand by way of
dialysis. The general catalytic potential
of these systems was assessed by study-
ing two benchmark reactions, the a-hy-
drazination of a b-keto ester as well as
the Henry reaction of 2-nitrobenzalde-
hyde with nitromethane. For both reac-
tions the bisoxazoline-based catalysts
displayed superior selectivity and, in
particular, catalyst activity. The latter
was interpreted as being due to the
hindered decoordination of the third
oxazoline unit, the key step in the gen-
eration of the active catalyst, in the im-
mobilised trisox–copper complexes.
Solutions of the second-generation
dendrimer catalysts were placed in
membrane bags, fabricated from com-
mercially available dialysis membranes,
with the purpose of catalyst recycling
based on dialysis. Overall, the support-
ed BOX catalyst gave good and highly
reproducible results throughout the
study, whereas the performance of the
trisox dendrimer system decreased mo-
notonically. The reason for the differ-
ent behaviour is the markedly lower
activity of trisox-based catalysts rela-
tive to those based on the BOX ligand.
This necessitated an increased reaction
time for each cycle of the trisox deriva-
tives, resulting in higher levels of cata-
lyst leaching, which was attributed to a
modification of the structure of the
membrane by its exposure to the sol-
vent trifluoroethanol at 408C.
Keywords: bisoxazolines · dendrim-
ers · homogeneous catalysis · mem-
branes · trisoxazolines
Introduction
retention of the catalyst and smooth transport of the reac-
tants and products. Possible interactions of the various com-
pounds/intermediates with the membrane surface have to be
considered. Finally, a membrane that is mechanically, ther-
mally and chemically stable in the required solvent and at
the required temperature should be chosen. This is a major
issue with polymeric membranes that may swell, change
their structure and, therefore, their pore size, leading to in-
creased catalyst leaching.[9]
Driven by a gradient, the substrates/reactants are trans-
ported through the membrane, whereas the catalyst is re-
tained due to its size. The driving force may be a pressure,
concentration or temperature gradient or a difference in
electrical potential. Several types of membrane-based reac-
tors have been reported, some of which require sophisticat-
ed engineering.[6e–h] The simplest approach is based on cata-
lyst trapping within a membrane “bag”. It provides ade-
quate dispersion of the catalyst and guarantees minimal in-
teraction between the catalyst and the polymer, thus allow-
ing the use of relatively labile catalyst systems. This is
certainly the case for the copper(II)-based Lewis acid cata-
lysts employed in this work that are attached to dendrimer
supports and are held within a dialysis membrane “tea
bag”,[10] which may be “dipped” into a reactant solution and
recycled several times.
The exploitation of ultra- and nanofiltration techniques
based on dialysis in the reaction engineering of catalytic
processes was originally developed for biotechnological ap-
plications.[1] Kula, Wandrey and others used continuous-flow
membrane reactors for enzymatic transformations since the
early 1980s,[2] and a simple, practical variation of this ap-
proach for batch reactions was put forward by Whitesides
and co-workers in 1987 who employed membrane bags to
recycle the enzymatic catalyst.[3] The application of this tech-
nique to organometallic homogeneous catalysis has been a
more recent development. Kragl and others made key con-
tributions to the development of continuously operating
membrane reactors for this type of catalytic systems,[4] using,
amongst others, a polymer-supported proline-derivative as a
chiral catalyst.[5] Finally, van Koten and van Leeuwen and
their co-workers first reported the application of this tech-
nique to dendrimer catalysis and led to its establishment in
dendrimer chemistry.[6,7]
These techniques require membranes that are adapted to
the reaction conditions to achieve good catalytic perfor-
mance.[8] The pore size of the membrane should guarantee
[a] M. Gaab, Prof. L. H. Gade
Whereas many examples of immobilised bisoxazolines
have been reported, relatively few contributions to the field
report the use of dendrimers as soluble supports,[11,12] and in
most of these studies the immobilised catalysts only dis-
played moderate activity and selectivity. A notable excep-
tion has been Majoralꢁs and Reiserꢁs series of dendritic aza-
bis(oxazolines), which gave good yields, and good to excel-
lent enantioselectivities at 5 mol% catalyst loading and
were recycled up to three times.[13] In this work, well-estab-
Anorganisch-Chemisches Institut, Universitꢂt Heidelberg
Im Neuenheimer Feld 270, 69120 Heidelberg (Germany)
Fax : (+49)6221-545609
E-mail: lutz.gade@uni-hd.de
[b] Dr. S. Bellemin-Laponnaz
Institut de Chimie UMR7177 CNRS
Universitꢃ de Strasbourg
1, rue Blaise Pascal, 67070 Strasbourg (France)
Fax : (+33)390-245001
E-mail: bellemin@chimie.u-strasbg.fr
Chem. Eur. J. 2009, 15, 5450 – 5462
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5451

L. H. Gade, S. Bellemin-Laponnaz and M. Gaab
lished carbosilane dendrimers[14] are employed that guaran-
tee a minimum interaction of the catalytic centres with the
dendritic core structure.
Based on established procedures, the formation of both
diamide intermediates in the synthesis of the bisoxazolines
proceeded cleanly and almost quantitatively. For (S)-valinol,
this was directly followed by cyclisation using tosyl chloride
to give 1a in 68% yield.[18] The reaction of dimethyl propar-
gylmalonate with (S)-a-phenylglycinol proceeded less clean-
ly; however, treatment with SOCl2 and subsequent cyclisa-
tion with NaOH generated BOX 1b in a reasonable yield
(41%).[19] After lithiation of 1a,b, they were reacted with
Results and Discussion
In this section we first describe the synthesis of bis- and tri-
soxazolines[15] containing a linker unit in the ligand back-
bone that allows their covalent attachment to carbosilane
dendrimers. Our systems of reference are the bisoxazolines
(BOX) A and B as well as the trisoxazoline (trisox) C
(Scheme 1).
MeI to give propargyl-BOXACTHNUTRGNEUGN(iPr) (2a) and propargyl-
BOX(Ph) (2b) in excellent yields over 90%. The replace-
ment of the acidic atom at the bridgehead in 1a,b was re-
quired to avoid competitive reactions in the following im-
mobilisation reaction.
The synthesis of C3 symmetric propargyl-trisoxACTHNUTRGNEUG(N iPr) (3a)
and propargyl-trisox(Ph) (3b) was achieved by the estab-
lished coupling[20] of a lithiated bisoxazoline with a 2-bro-
mooxazoline.[21] The presence of the propargyl moiety neces-
sitated more vigorous conditions as well as longer reaction
times and a higher stoichiometric excess of the bromooxazo-
lines than previously reported for the parent trisox deriva-
tives. Under these conditions 3a was obtained in 69% yield.
In the case of 3b, high temperatures had to be avoided be-
cause 2-bromophenyloxazoline is far less stable than 2-bro-
moisopropyloxazoline. It is especially prone to rearrange
and form a 2-bromoisocyanate.[22] Accordingly, compound
3b could only be isolated in 35% yield.
Scheme 1. Structures of the BOX- and trisox-functionalised ligands used
in this study. A, B and C are reference ligands; G0, G1 and G2=zeroth-,
first-, second-generation dendrimers, respectively; TBDMS=tert-butyldi-
methylsilyl.
Preparation of the oxazoline-functionalised dendrimers: The
parent carbosilane dendrimers {G0}-(SiMe2Cl)4, {G1}-
(SiMe2Cl)8 and {G2}-(SiMe2Cl)16 were synthesised according
to literature procedures,[14e–j] with the aim of a subsequent
nucleophilic substitution at the terminal chlorosilane groups.
To test the potential of ligands 2a,b and 3a,b for this kind
of reaction tBuMe2SiCl was chosen as a model system. After
deprotonation of 2a,b and 3a by LDA or BuLi and reaction
with the chlorosilane, clean and complete product formation
(4a,b and 5) was observed (Scheme 3). However, employing
the same reaction conditions with {G0}-(SiMe2Cl)4, {G1}-
(SiMe2Cl)8 and {G2}-(SiMe2Cl)16 only led to low or moder-
ate conversion and inseparable product mixtures of in part
defective dendritic systems.
The choice of the appropriate linker is based on two prin-
cipal requirements: a relatively inert binding to the support-
ing macromolecule and minimised interference with the cat-
alytic sites.[16] A propargyl function at the bridging position
of the BOX ligand or the apical position of the trisox ligand
appeared to meet these requirements. Moreover, the termi-
nal alkyne subsequently allowed facile linkage to the den-
drimers through deprotonation and reaction with chlorosi-
lane termini.[17]
Synthesis of propargyl-functionalised BOX and trisox deriv-
atives: The synthesis of the new
propargyl-functionalised BOX
derivatives and trisox ligands is
based on a modular strategy
and derives from an intermedi-
ate
bisoxazoline
1a,b
(Scheme 2). (S)-Valinol and
(S)-a-phenylglycinol were used
to introduce groups with differ-
ent steric bulk, and the tethered
ligands 2a,b (BOX derivatives)
and 3a,b (trisox ligands) were
synthesised from commercially
available dimethyl propargyl-
malonate.
Scheme 2. Synthesis of the propargyl-functionalised BOX 2a,b and trisox 3a,b ligands. Reaction conditions: i)
nBuLi, then MeI; ii) tBuLi, then 2-bromooxazoline.
5452
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5450 – 5462

Homogeneous Catalysis
FULL PAPER
L* (L*=2a,b and 3a) as polyfunctional chiral ligand systems
for asymmetric copper(II) Lewis acid catalysis was assessed
by studying two benchmark reactions, the a-hydrazination
of a b-keto ester as well as the Henry reaction of 2-nitroben-
zaldehyde with nitromethane. Both reactions had previously
been studied using various BOX derivatives as stereodirect-
ing ligands.[24,25] Ligands A, B and C (shown above), bearing
only a methyl group at the bridgehead or the apical position,
were employed as reference systems for both of these reac-
tions.
a-Hydrazination of ethyl 2-methylacetoacetate: This reac-
tion has been studied extensively[24] and therefore lent itself
to assess the influence of the dendritic support (and subse-
quently the recycling in a membrane bag, vide infra) on the
catalytic system. The results obtained for the catalytic a-hy-
drazination of ethyl 2-methylacetoacetate are displayed in
Table 1. A remarkably low cata-
Scheme 3. Synthesis of TBDMS-functionalised model systems 4a,b (77%
and 62% yield) and 5 (95% yield) for the dendritic BOX and trisox li-
gands.
lyst loading of 1 mol% was
found to be sufficient and gen-
erally high yields and selectivi-
ties were obtained.
The highest enantioselectivi-
ties were observed for the
BOX(Ph) derivatives—a trend
which had already been noted
earlier.[24] Enantiomeric excess-
es of between 97 and 99%
were obtained with these cata-
lysts, whereas BOX
ACHTUNGTRENN(UNG iPr) and
Scheme 4. Synthesis of the BOX- and trisox-functionalised carbosilane dendrimers G0-L*, G1-L*, G2-L*
(L*=2a,b and 3a).
trisox(iPr) ranged between 90
AHCTUNGTRENNUNG
The sluggish and incomplete conversion could be over-
Table 1. Catalytic asymmetric a-hydrazination of ethyl 2-methylacetoace-
tate with the polyfunctional dendritic ligands and monofunctional refer-
ence systems.
come by addition of one equivalent of TlPF6 per acetylide
unit and the use of an excess of the linker-ligand reagent
(Scheme 4). Applying of this strategy, conversions of around
70% for all dendrimer and ligand combinations were ach-
ieved.
The purification of the new dendritic compounds was car-
ried out by applying van Kotenꢁs strategy of dendrimer iso-
lation by dialysis.[23] The zeroth-generation trisox derivative
G0-3a as well as all first- and second-generation dendrimers
G1-L* and G2-L* (L*=2a,b and 3a) could be purified effi-
ciently and were obtained in pure form in moderate to good
yields (49–77%). However, the zeroth-generation G0-L* de-
rivatives (L* =2a,b) were too small to be efficiently retained
by the membrane pores and therefore had to be purified by
flash column chromatography. As they tend to decompose
slowly on silica, this explains their low yields (28 and 22%).
Characterisation (and assessment of purity) of the oxazo-
line-functionalised dendrimers was provided by 13C and 29Si
NMR spectroscopy, elemental analysis as well as by mass
spectrometry for the lower generations.
Ligand
ee [%]
Ligand
ee [%]
Ligand
ee [%]
C
3a
5
G0-3a
G1-3a
G2-3a
94
96
93
92
95
90
A
2a
4a
G0-2a
G1-2a
G2-2a
94
97
96
93
95
94
B
2b
4b
G0-2b
G1-2b
G2-2b
98
99
99
97
97
98
[a] Structures of the ligands are shown in Scheme 1. Quantitative yields
were obtained for all catalysts.
and 97% ee. There are some notable aspects concerning the
results obtained with BOXACTHNUTRGNENG(U iPr) and trisoxHCATUNGTRENN(UGN iPr) derivatives,
namely an increase of ee values from 94% ee for the cata-
lysts with the parent ligand systems A and C to 97 and 96%
with propargyl-substituted ligands 3a and 2a, respectively.
Similar observations were made for the smaller dendritic
species G0-3a/G0-2a as well as G1-3a/G1-2a. In general,
Comparative studies of copper(II) Lewis acid catalysts. The
general catalytic potential of compounds G0-L*, G1-L*, G2-
Chem. Eur. J. 2009, 15, 5450 – 5462
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5453

L. H. Gade, S. Bellemin-Laponnaz and M. Gaab
enantioselectivities were slightly higher (3–4%) for the bi-
soxazoline with respect to the trisox systems.
Henry reaction of nitromethane and 2-nitrobenzaldehyde:
The Henry reaction was chosen as a complementary test
system to assess the trend observed in the a-hydrazination.
BOX ligands had already been applied successfully in the
reaction of various benzaldehyde derivatives with nitrome-
thane at 5 mol% catalyst loading, yielding products with
good enantioselectivities of around 90%.[25] To prove the
effect of the varying catalyst environments in this study, a
reference reaction was chosen that only gave moderate en-
antiomeric excesses, this allowing both increase and de-
crease in enantioselectivities. Table 2 summarises the reac-
tion conditions as well as the catalytic results for the differ-
ent copper(II) catalysts.
Table 2. Catalytic asymmetric Henry reaction of 2-nitrobenzaldehyde
and nitromethane with the polyfunctional dendritic ligands and mono-
functional reference systems.
Ligand ee
Yield
Ligand ee
Yield
Ligand ee
Yield
[%] [%]
[%] [%]
[%] [%]
C
3a
5
75
67
50
93
32
24
42
58
54
A
2a
4a
74
77
77
95
45
36
86
87
87
B
2b
4b
71
74
74
89
74
65
83
85
82
G0-3a 52
G1-3a 53
G2-3a 53
G0-2a 84
G1-2a 87
G2-2a 83
G0-2b 84
G1-2b 83
G2-2b 81
Figure 1. Conversion curves for the a-hydrazination of ethyl 2-methylace-
toacetate using catalysts with different steric bulk.
[a] Structures of the ligands are shown in Scheme 1.
First of all, we note a significant difference between the
BOX- and trisox-based catalysts. BOX(Ph) and BOX(iPr)
AHCTUNGTRENNUNG
gave similar ee values (71–77% ee) and yields for all mono-
nuclear catalysts. On the other hand, the dendritic BOX de-
rivatives all gave reaction products with significantly higher
enantiomeric excesses (81–87% ee). In contrast, the perfor-
mance of all the trisox-based catalysts proved to be inferior,
both in terms of the activities and enantioselectivities. The
negative trend with respect to the trisox-based catalysts ob-
served for the a-hydrazination discussed above is thus en-
hanced for the Henry reaction.
Comparison of the BOX– and trisox–copper catalysts: To
gain some insight into the different behaviour of the immo-
bilised BOX and trisox ligands for both types of reactions,
the rates of conversion were monitored. For this purpose
the a-hydrazination of ethyl 2-methylacetoacetate was
chosen because it had already been the object of a detailed
kinetic study and proceeded very cleanly.[24c] The corre-
sponding conversion curves are shown in Figure 1. To allow
for a better comparison Figure 2 depicts the time each cata-
lyst required to achieve 90% conversion.
Figure 2. Comparison of the time needed to achieve 90% conversion
with the different catalysts.
Both the BOX-based catalysts and the TBDMS-function-
alised derivatives (TBDMS=tert-butyldimethylsilyl) display
similar rates of conversion, whereas the G2-supported cata-
lysts are markedly less active (Figure 1, top and middle). In
5454
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5450 – 5462

Homogeneous Catalysis
FULL PAPER
contrast, the conversion curves recorded for the three trisox
species reveal different behaviour. Whereas the parent cata-
lyst containing ligand C displays an activity similar to the
BOX-based catalysts, the TBDMS-substituted trisox catalyst
5 is significantly less active, and the conversion is even
slower for the dendritic catalyst G2-3a (Figure 1, bottom
and Figure 2). In summary, whilst the attachment of a linker
at the ligand backbone and the immobilisation only moder-
ately affect the BOX-based catalysts for the two reactions
studied in this work, the effect on the trisox-based catalysts
is dramatic. Attachment of trisox to a second generation car-
bosilane dendrimer increases the reaction times by an order
of magnitude!
This observation may be understood on the basis of the
previously proposed role of the third ligand arm in trisox–
copper(II) Lewis acid catalysts.[24b,c] In solution, an equilibri-
um between k3- and k2-trisox coordinated complexes is
thought to exist, for which the coordination of the third
ligand arm stabilises the resting state but leads to a deactiva-
tion of the copper complexes by reduction of the Lewis
acidity as was shown in a theoretical study on BOX–Cu cat-
alysts.[26] We note though that Reiser et al. have found that
in some cases pentacoordinate CuII species may be effective
as catalysts,[27] but apparently not with strong donors such as
the third oxazoline moiety present in the trisox ligands em-
ployed in this study.
gains in importance. In summary, increasing steric bulk in
the periphery of the immobilised trisox ligands is thought to
result in the negative observed effect on both the catalyst
activity and selectivity. On the other hand, increasing steric
bulk may be beneficial for the enantiodiscrimination in the
case of the more “open” BOX as manifested in an increase
of the ee values by about 10%!
Recycling through dialysis: As indicated above, the dendri-
mer catalysts were developed with the purpose of catalyst
recycling based on dialysis, using membrane bags fabricated
from a commercially available dialysis membrane (Sigma-
Aldrich: benzoylated dialysis tubing, MWCO 2000). As re-
actors we chose simple screw cap vials as depicted in
Figure 3.
Figure 3. General setup for the recycling using the “catalyst in a tea bag”
principle. An enlarged schematic view clarifies the operational principle.
The transformation of the stabilised but inactive resting
state into the active (17 electron CuII) species therefore re-
quires the decoordination of an oxazoline unit, which then
adopts a remote position from the centre, as depicted in
Scheme 5.[24b,c] The resulting vacant coordination site can
Two of the highest generation dendrimers (G2-2b and
G2-3a) were applied to compare the behaviour of BOX and
trisox in the recycling study of the a-hydrazination. The
metalated analogues G2-2b–Cu and G2-3a–Cu of these
dendritic ligands possess molecular weights of around
14500 gmolÀ1, whereas the membrane allows only the mi-
gration of molecules up to 2000 gmolÀ1. On the other hand,
the transport of the substrates and the product in and out of
the membrane bag occurs by diffusion, which was accelerat-
ed by operating at an elevated temperature of 408C, conse-
quently resulting in somewhat lower enantioselectivity. The
substrates migrate into the membrane bag, in which they in-
teract with the catalytically active terminal groups of the
dendrimer and are converted to the product. Consequently,
the latter is enriched in the interior and then passes through
the membrane to the exterior part of the reactor following
the concentration gradient. The practical handling of such a
catalyst “tea bag” is illustrated in Figure 4.
Initially the dendrimer-filled membrane bag was placed
into the vial containing the yellow solution of the substrates,
the colour being due to the diazodicarboxylate (Figure 4a).
After their complete conversion to the product, which is ac-
companied by a decolouration of the solution (Figure 4b),
the bag was transferred into another vial containing fresh
substrates (Figure 4c). For monitoring the rate of decoloura-
tion, reactions times of 10 and 24 h were chosen per cycle
with G2-2b–Cu and G2-3a–Cu, respectively, in order to ach-
Scheme 5. Equilibrium between k3- and k2-trisox-coordinated complexes.
In the k2-coordinated species the third oxazoline adopts a remote posi-
tion from the metal centre.
only then be occupied by the enolate form of ethyl 2-meth-
ylacetoacetate. Enantiodescrimination is thus effected by
similar Cu species in the BOX and trisox systems.
However, decoordination of the third arm and a fast equi-
libration of the k3- and k2-trisox-coordinated complexes re-
quire sufficient space for the intramolecular movement of
one of the oxazoline rings. Introduction of a bulky substitu-
ent (in the form of the linker or even more so of the dendri-
mer) to the apical position hinders this process sterically
and therefore only a fraction of the actually employed
trisox–Cu catalyst will be catalytically active. For immobi-
lised trisox, the catalyst loading is thus effectively reduced
and the background reaction leading to a racemic product
Chem. Eur. J. 2009, 15, 5450 – 5462
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5455

L. H. Gade, S. Bellemin-Laponnaz and M. Gaab
decreased monotonically to 14% ee and 38% yield for the
final run. The reason for the different behaviour of the two
dendrimer catalysts is to a large extent the markedly lower
activity of G2-3a compared to G2-2b. This necessitated in-
creased reaction times for each cycle, leading to higher
levels of metal-ion leaching, which were investigated by
atomic absorption spectrometry. This established a loss in
Cu of 2% per cycle of 10 h with G2-2b, whereas 4% of the
initially applied amount of Cu was leached after cycles of
24 h in the catalyses with G2-3a. The level of leaching is
proportional to the reaction time and may result in part
from the substitutional lability of the copper(II)-oxazoline
complexes as well as the modification of the membrane
structure due to its exposure to trifluoroethanol at 408C.
Figure 4. Series showing the course of one recycling: a) dendrimer-filled
membrane bag in the yellow solution of the substrates; b) colourless so-
lution after complete conversion; c) transfer of the bag into another vial;
d) subsequent reaction cycle.
ieve reasonable conversions. The results obtained for seven
successive runs are summarised in Figure 5.
Conclusion
It is notable that the enantioselectivities obtained with
G2-2b–Cu vary only slightly from 82% in the first to
77% ee in the last run. A maximum ee of 88% was reached
during the third and fourth cycle. The corresponding yields
reflect the classical behaviour of membrane systems showing
increase between the first (67%) and the second run (86%),
reflecting the establishment of a stationary state and then
drop gradually to 69% again in the seventh cycle. Overall,
the supported BOX catalyst G2-2b–Cu gave good and
highly reproducible results throughout the study. The trisox
dendrimer system G2-3a–Cu, on the other hand, started out
with a moderate performance (69% ee, 55% yield) which
In this study bis- and trisoxazolines bearing an alkynyl
linker unit have been attached to second-generation carbosi-
lane dendrimers and their copper(II) complexes have been
immobilised in dialysis membrane bags. Only the bisoxazo-
line-based catalysts displayed sufficient activity to allow re-
cycling without significant decrease in activity and selectivi-
ty. This has allowed to effect catalytic conversions by dip-
ping the catalyst-filled dialysis bags into reaction vessels
containing the substrate. Similar to previously reported en-
zymatic systems, the dendrimers provided the basis of a cat-
alytic “tea bag”, which may be conveniently recycled several
times. Ongoing research addresses the problems associated
with the transport phenomena which govern the properties
of this type of system.
Experimental Section
All manipulations were carried out by using standard Schlenk line or
drybox techniques under an atmosphere of argon, unless stated other-
wise. Solvents were pre-dried over activated 4 ꢄ molecular sieves and
heated to reflux over potassium (THF), sodium (toluene) or calcium hy-
dride (CH2Cl2, NEt3) under an argon atmosphere and collected by distil-
lation. 1H, 13C{1H}, 29Si{1H} NMR spectra were recorded on a Bruker
Avance II 400 or a Bruker Avance III 600 spectrometer. 1H and 13C as-
signments were confirmed when necessary by DEPT-135 and two-dimen-
sional H-1H as well as 13C-1H NMR experiments. H and 13C NMR spec-
tra were referenced internally to residual protio-solvent (1H) or solvent
(13C) resonances and are reported relative to tetramethylsilane. 29Si NMR
spectra were referenced externally to tetramethylsilane. Infrared spectra
were prepared as KBr pellets and were recorded on a Varian Excalibur
3100 series FTIR spectrometer. Mass spectra were recorded by the mass
spectrometry service and elemental analyses by the analytical service of
the Chemical Institutes of the University of Heidelberg. Analytical sepa-
ration of enantiomers was provided by high pressure liquid chromatogra-
phy on a Finnigan Surveyor machine. Daicel Chiralcel columns AD-H
and OD-H (250ꢅ4.6 mm, 5 mm) and the corresponding guard cartridge
(10ꢅ4 mm, 5 mm) were used. For gas chromatography a Finnigan Focus
GC apparatus equipped with a capillary column (BPX5, 5% phenyl, poly-
silphenylenesiloxane, nonpolar, 30 mꢅ0.25 mmꢅ0.5 mm) was applied.
Membranes for dialysis were purchased from Sigma-Aldrich (dialysis
tubing, benzoylated, av. flat width 32 mm). (S)-Valinol,[28] (S)-a-phenyl-
glycinol,[28] (S)-4-isopropyloxazoline,[29] (S)-4-phenyloxazoline,[29] (S)-2-
1
1
Figure 5. Results of the catalytic performance obtained in the recycling
study. Seven successive runs with G2-2b–Cu and G2-3a–Cu are present-
ed.
5456
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5450 – 5462

Homogeneous Catalysis
FULL PAPER
bromo-4-isopropyloxazoline,[21] (S)-2-bromo-phenyloxazoline,[21] 1,1,1-
tris[(S)-4-isopropyloxazolin-2-yl]ethane (C)[20] and the chlorosilyl-func-
tionalised carbosilane dendrimers[14e–j] were prepared according to pub-
lished procedures. 2,2-Bis[(S)-4-isopropyloxazolin-2-yl]propane (A) and
2,2-bis[(S)-4-phenyloxazolin-2-yl]propane (B) were obtained by methyla-
tion of the corresponding 2,2-bis(oxazolin-2-yl)ethanes.[18,30] All other
compounds and reagents were purchased from commercial chemical sup-
pliers and used without further purification. The absolute configuration
for the product eeꢁs was determined by HPLC and comparison with the
literature data.[24]
99%). 1H NMR (600.13 MHz, [D4]MeOH, 293 K): d=7.34–7.25 (m,
10H; CHaryl), 5.18 (m, 2H; NCH), 3.86–3.71 (m, 4H; CH2Cl), 3.49 (t, J=
3
ꢀ
ꢀ
7.59 Hz, 1H; CHCH2C CH), 2.76–2.68 (m, 2H; CH2C CH), 2.31 ppm (t,
4
J=2.63 Hz, 1H; CH2C CH); 13C{1H} NMR (150.90 MHz, [D4]MeOH,
ꢀ
293 K): d=170.20/170.19 (NCO), 140.03/139.93 (Cq,aryl), 129.80/129.72/
ꢀ
ꢀ
129.12/129.04/128.00/127.92 (Caryl), 82.29 (C CH), 72.14 (C CH), 56.39/
ꢀ
56.25 (NCH), 54.11 (CHCH2C CH), 47.84/47.65 (CH2Cl), 20.78 ppm
ꢀ
(CH2C CH); IR (KBr): n˜ =3295 (m; NH), 3120–3000 (w; CHaryl), 3000–
2850 (w; CH2), 2121 (w; C C), 1670, 1558 cm (s; CO); MS (FAB): m/z
À1
ꢀ
(%): 417.2 (50) [M+H]+, 367.2 (6) [MÀCH2Cl]+; HRMS (FAB): m/z
calcd for C22H23N2O235Cl2 [M+H]+: 417.1163; found: 417.1136; m/z calcd
for C22H23N2O235Cl37Cl [M+H]+: 419.1108; found: 419.1143; m/z calcd for
C22H23N2O237Cl2 [M+H]+: 421.1078; found: 421.1134.
Preparation of the ligands
N,N’-Bis[(S)-1-(hydroxymethyl)-2-methylpropyl]-2-prop-2-yn-1-ylmalona-
mide (1a’): In a sealed Schlenk tube, l-valinol (7.57 g, 73.4 mmol) and di-
methyl propargylmalonate (5.58 mL, 36.7 mmol) were heated at 1208C
until a solid formed (about 1.5 h). Dry toluene was added to the warm
mixture to generate a suspension that was heated at 1208C for another
2 h. The product was precipitated by addition of hexane at room temper-
ature. Subsequent filtration, another washing with hexane, and removal
of the residual solvent in vacuo yielded the diamide as an off-white
powder (10.62 g, 93%). 1H NMR (600.13 MHz, [D6]DMSO, 293 K): d=
7.59 (d, 3J=9.26 Hz, 1H; NH), 7.52 (d, 3J=9.19 Hz, 1H; NH), 4.64
(brdd, 3J=9.03 Hz, 3J=5.30 Hz, 2H; OH), 3.61 (m, 1H; NCH), 3.54 (m,
4,4-Bis[(S)-4-isopropyloxazolin-2-yl]but-1-yne (1a): Dry NEt3 (12.17 mL,
87.6 mmol) and DMAP (241 mg, 2.0 mmol) were added to a suspension
of dihydroxy diamide 1a’ (6.15 g, 19.7 mmol) in dry dichloromethane
(250 mL). The mixture was cooled to 08C and solid TsCl (7.50 g,
39.3 mmol) was added over a period of 3 h. After warming the resulting
off-white suspension to room temperature, a yellow solution formed that
was stirred for another 4 d. Completion of conversion was achieved by
subsequent heating at 408C for 8 h. The resulting mixture was diluted by
addition of dichloromethane, washed with NH4Claq as well as brine, and
dried over Na2SO4. Filtration and removal of the solvent in vacuo afford-
ed an oily crude product that was purified by column chromatography
(pentane/EtOAc 60:40) yielding the product as a pale yellow oil (3.97 g,
73%). 1H NMR (600.13 MHz, CDCl3, 293 K): d=4.27–4.22 (m, 2H;
CH2O), 4.03–3.94 (m, 4H; CH2O, NCH), 3.65 (t, 3J=7.74 Hz, 1H;
1H; NCH), 3.44–3.26 (m, 5H; CH2OH, CHCH2C CH), 2.73 (t, 4J=
ꢀ
2.59 Hz, 1H; CH2C CH), 2.56 (ddd, 2J=16.66 Hz, 3J=8.62 Hz, 4J=
ꢀ
2.62 Hz, 1H; CH2C CH), 2.46 (ddd, 2J=16.67 Hz, 3J=6.21 Hz, 4J=
ꢀ
ꢀ
2.63 Hz, 1H; CH2C CH), 1.86 (m, 1H; CH
G
(CH3)2), 0.86 (d, 3J=6.83 Hz, 3H; CH
U
CHCH2C CH), 2.86 (ddd, 2J=16.89 Hz, 3J=7.71 Hz, 4J=2.65 Hz, 1H;
ACHTUNGTRENNUNG ACHTUNGTRENNUNG
(CH3)2), 0.81 (d, 3J=6.59 Hz, 3H; CH
ꢀ
CH
6.83 Hz, 3H; CHACHTUNGTRENNUNG
CH2C CH), 2.81 (ddd, 2J=16.88 Hz, 3J=7.80 Hz, 4J=2.66 Hz, 1H;
ꢀ
4
ꢀ
ꢀ
CH2C CH), 1.98 (t, J=2.65 Hz, 1H; CH2C CH), 1.78 (m, 2H; CH-
ꢀ
ꢀ
293 K): d=167.65/167.30 (NCO), 82.22 (C CH), 71.79 (C CH), 61.16/
C
(CH3)2), 0.86 ppm (d, 3J=6.78 Hz,
ACHTUNGTRENNUNG
ꢀ
61.07 (CH2OH), 55.53/55.24 (NCH), 51.70 (CHCH2C CH), 28.22/27.57
AHCTUNGTRENNUNG
ꢀ
(CH
G
(CH3)2), 18.28 (CH2C CH), 17.92/
(CH3)2); IR (KBr): n˜ =3520–3320 (s, br; OH), 3320–3200
ꢀ
17.38 ppm (CH
N
À1
ꢀ
ꢀ
ꢀ
(C CH), 39.08 (CHCH2C CH), 32.28/32.21 (CH
ACHTNUGTRNEN(UGN CH3)2), 19.78 (CH2C
ꢀ
(m; NH), 2980–2870 (m; CH2), 2121 (w; C C), 1634, 1559 cm (s, CO);
MS (FAB): m/z (%): 625.5 (3) [2M+H]+, 313.3 (100) [M+H]+, 295.3
(95) [MÀOH]+; elemental analysis calcd (%) for C16H28N2O4 (312.41): C
61.51, H 9.03, N 8.97; found C 61.25, H 8.92, N 8.95.
CH), 18.56/18.51 (CH(CH3)2), 17.72 ppm (CH(CH3)2); IR (KBr): n˜ =
A
ACHTUNGTRENNUNG
À1
ꢀ
2960–2874 (m; CH2), 2108 (w; C C), 1665 cm (s; C=N); MS (FAB):
m/z (%): 277.2 (100) [M+H]+, 191.1 (4) [M+HÀC5H10O]+; HRMS
(FAB): m/z calcd for C16H25N2O2 [M+H]+: 277.1916; found: 277.1944; el-
emental analysis calcd (%) for C16H24N2O2 (276.37): C 69.53, H 8.75, N
10.14; found C 69.57, H 8.75, N 9.96.
N,N’-Bis[(S)-2-hydroxy-1-phenylethyl]-2-prop-2-yn-1-ylmalonamide
(1b’): In a sealed Schlenk tube, l-a-phenylglycinol (13.77 g, 100.4 mmol)
and dimethyl propargylmalonate (7.63 mL, 50.2 mmol) were heated at
1108C until a solid formed (about 3 h). Then, dry toluene was added to
the warm mixture to generate a suspension that was heated at 1008C for
another 2 h. The product was precipitated by addition of pentane at
room temperature. Subsequent filtration, another washing with pentane,
and removal of the residual solvent in vacuo yielded the diamide as a
4,4-Bis[(S)-4-phenyloxazolin-2-yl]but-1-yne (1b): In air, dichloride 1b’’
(7.39 g, 17.7 mmol) was dissolved in methanolic NaOH (2.13 g,
53.3 mmol in 260 mL) and heated to reflux for 14 h, during which NaCl
precipitated. The solvent was removed in vacuo, and the residue was re-
dissolved in dichloromethane, washed with NH4Claq and dried over
Na2SO4. Filtration and removal of the solvent in vacuo afforded an
orange foam that was purified by column chromatography (pentane/
EtOAc 70:30) yielding the bisoxazoline as an orange, viscous oil (2.61 g,
43%). 1H NMR (600.13 MHz, CDCl3, 293 K): d=7.35–7.26 (m, 10H;
1
white powder (18.20 g, 95%). H NMR (600.13 MHz, [D6]DMSO, 293 K):
d=8.51 (d, 3J=7.93 Hz, 1H; NH), 8.45 (d, 3J=7.87 Hz, 1H; NH), 7.36–
7.21 (m, 10H; CHaryl), 5.04 (s, 2H; OH), 4.84 (m, 2H; NCH), 3.64–3.59
3
(m, 4H; CH2OH), 3.57 (t, J=7.38 Hz, 1H; CHCH2C CH), 2.76 (t, 4J=
ꢀ
3
ꢀ
ꢀ
CHaryl), 5.28 (m, 2H; NCH), 4.71 (m, 2H; CH2O), 4.19 (m, 2H; CH2O),
2.35 Hz, 1H; CH2C CH), 2.53 ppm (d, J=5.76 Hz, 2H; CH2C CH);
13C{1H} NMR (150.90 MHz, [D6]DMSO, 293 K): d=166.99/166.88
(NCO), 140.90/140.68 (Cq, aryl), 127.98/126.69/126.66/126.62 (Caryl), 82.18
3
ꢀ
ꢀ
3.90 (t, J=7.71 Hz, 1H; CHCH2C CH), 3.02 (m, 2H; CH2C CH),
2.10 ppm (dt, 4J=2.50 Hz, J=1.18 Hz, 1H; CH2C CH); 13C{1H} NMR
(150.90 MHz, CDCl3, 293 K): d=164.75/164.67 (NCO), 141.94/141.87 (Cq,
5
ꢀ
ꢀ
ꢀ
(C H), 71.95 (C CH), 64.73/64.42 (CH2OH), 54.94/54.89 (NCH), 51.48
ꢀ
ꢀ
ꢀ
aryl), 128.69/128.68/127.65/126.75/126.65 (Caryl), 80.58 (C CH), 75.55/75.47
(CHCH2C CH), 17.91 (CH2C CH); IR (KBr): n˜ =3520–3340 (m, br;
OH), 3340–3240 (m; NH), 3100–3040 (w; CHaryl), 2990–2860 (w; CH2),
1665, 1541 cmÀ1 (s; CO); MS (FAB): m/z (%): 381.2 (100) [M+H]+;
HRMS (FAB): m/z calcd for C22H25N2O4 [M+H]+: 381.1814; found:
381.1808; elemental analysis calcd (%) for C22H24N2O4 (380.44): C 69.46,
H 6.36, N 7.36; found C 69.03, H 6.43, N 7.28.
ꢀ
ꢀ
(CH2O), 70.49 (C CH), 69.69/69.60 (NCH), 39.14 (CHCH2C CH),
ꢀ
ꢀ
19.97 ppm (CH2C CH); IR (KBr): n˜ =3295 (m; C CH), 3070–3005 (w;
CHaryl), 2980–2860 (m; CH2), 2121 (w; C C), 1664 (s; C=N); MS (FAB):
ꢀ
m/z (%): 345.1 (100) [M+H]+, 225.1 (2) [M+HÀC8H8O]+; HRMS
(FAB): m/z calcd for C22H21N2O2 [M+H]+: 345.1603; found: 345.1600; el-
emental analysis calcd (%) for C22H20N2O2 (344.41): C 76.72, H 5.85, N
8.13; found C 76.55, H 5.90, N 7.99.
N,N’-Bis[(S)-2-chloro-1-phenylethyl]-2-prop-2-yn-1-ylmalonamide (1b’’):
A suspension of dihydroxy diamide 1b’ (6.75 g, 17.7 mmol) in dry toluene
(140 mL) was cooled to 08C and SOCl2 (7.77 mL, 107.1 mmol) was added
slowly. The reaction proceeded at room temperature over night, yielding
a greyish green solution. Removal of the solvent in vacuo gave a yellow
glass, which was redissolved in dichloromethane, washed with KHCO3
(10% w/w in H2O) and dried over Na2SO4. Subsequent filtration and re-
moval of the residual solvent in vacuo yielded the product as a reddish
orange powder of sufficient purity for direct use in the next step (7.39 g,
4,4-Bis[(S)-4-isopropyloxazolin-2-yl]pent-1-yne
(propargyl-BOXACHTUNGTRENNUNG(iPr);
2a): A solution of bisoxazoline 1a (2.40 g, 8.7 mmol) in THF (50 mL)
was cooled to À788C and nBuLi (5.97 mL, 1.6m in hexane) was added.
The resulting bright yellow mixture was stirred at À408C for 30 min prior
to the addition of MeI (1.62 mL, 26.0 mmol). The mixture was slowly
warmed to room temperature over night. After removal of the solvent in
vacuo, the residue was redissolved in dichloromethane, washed with
Chem. Eur. J. 2009, 15, 5450 – 5462
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5457

L. H. Gade, S. Bellemin-Laponnaz and M. Gaab
NH4Claq and dried over Na2SO4. Filtration and removal of the solvent
gave the crude product that was purified by column chromatography
(pentane/EtOAc 60:40) yielding 2a as a pale yellow oil (2.30 g, 91%).
1H NMR (600.13 MHz, CDCl3, 293 K): d=4.23–4.19 (m, 2H; CH2O),
4,4,4-Tris[(S)-4-phenyloxazolin-2-yl]but-1-yne (propargyl-trisox(Ph); 3b):
A solution of bisoxazoline 1b (442 mg, 1.3 mmol) in dry toluene (70 mL)
was cooled to À788C and tBuLi (0.94 mL, 1.7m in hexane) was added;
15 min later (S)-2-bromo-4-phenyloxazoline (406 mg, 1.8 mmol, 54% w/w
in THF) was transferred to the yellow solution, giving rise to an orange-
brown colour. The mixture was warmed to room temperature, concen-
trated to eliminate the hexane originating from tBuLi, and heated at
708C for 10 d. The resulting pale brown solution was evaporated to dry-
ness and the residue was redissolved in dichloromethane (50 mL),
washed with H2O (10 mL) and dried over Na2SO4. Filtration and removal
of the solvent in vacuo yielded a brown oil which was purified by column
chromatography (pentane/EtOAc 40:60, then EtOAc and EtOAc/MeOH
99:1) yielding 3b as an extremely viscous oil, which turned into a glass
upon standing (220 mg, 35%). 1H NMR (600.13 MHz, CDCl3, 293 K):
d=7.36–7.26 (m, 15H; CHaryl), 5.35 (dd, 3J=10.09 Hz, 3J=7.88 Hz, 3H;
2
4
4.03–3.95 (m, 4H; CH2O, NCH), 2.90 (dd, J=16.80 Hz, J=2.64 Hz, 1H;
4
CH2C CH), 2.85 (dd, 2J=16.79 Hz, J=2.63 Hz, 1H; CH2C CH), 1.98
ꢀ
ꢀ
4
ꢀ
(t, J=2.60 Hz, 1H; CH2C CH), 1.80 (m, 2H; CH
G
CCH3), 0.91 (m, 6H; CH
(CH3)2), 0.86 ppm (m, 6H; CH
G
ꢀ
NMR (150.90 MHz, CDCl3, 293 K): d=166.67/166.54 (NCO), 80.02 (C
ꢀ
CH), 71.79/71.54 (NCH), 70.79 (C CH), 70.12/70.09 (CH2O), 41.72
ꢀ
ꢀ
(CCH2C CH), 32.18/32.13 (CH
(CCH3), 18.64/18.47 (CH
(CH3)2), 26.95 (CH2C CH), 21.25
(CH3)2), 17.55/17.39 ppm (CH
(CH3)2); IR
À1
ꢀ
(KBr): n˜ =2960–2874 (m; CH2), 2121 (w; C C), 1661 cm (s; C=N); MS
(FAB): m/z (%): 291.2 (100) [M+H]+, 289.1 (4) [MÀH]+, 247.1 (2)
[MÀC3H7]+, 205.1 (4) [M+HÀC5H10O]+; HRMS (FAB): m/z calcd for
C17H27N2O2 [M+H]+: 291.2072; found: 291.2066; elemental analysis calcd
(%) for C17H26N2O2 (290.40): C 70.31, H 9.02, N 9.65; found C 70.13, H
8.95, N 9.08.
NCH), 4.78 (dd, 2J=10.13 Hz, 3J=8.35 Hz, 3H; CH2O), 4.26 (m, 3H;
4
CH2O), 3.45 (dd, 2J=16.81 Hz, J=2.62 Hz, 1H; CH2C CH), 3.39 (dd,
ꢀ
2J=16.80 Hz, J=2.65 Hz, 1H; CH2C CH), 2.11 ppm (t, 4J=2.62 Hz,
4
ꢀ
1H; CH2C CH); 13C{1H} NMR (150.90 MHz, CDCl3, 293 K): d=164.25
ꢀ
4,4-Bis[(S)-4-phenyloxazolin-2-yl]pent-1-yne (propargyl-BOX(Ph); 2b):
A solution of bisoxazoline 1b (858 mg, 2.5 mmol) in THF (10 mL) was
cooled to À788C and nBuLi (1.70 mL, 1.6m in hexane) was added. The
resulting bright yellow mixture was stirred at À408C for 30 min prior to
adding MeI (0.47 mL, 7.5 mmol). The solution was warmed to room tem-
perature over night. After removal of the solvent in vacuo, the residue
was redissolved in dichloromethane, washed with NH4Claq and dried over
Na2SO4. Filtration and removal of the solvent yielded the crude product
which was purified by column chromatography (pentane/EtOAc 60:40)
ꢀ
(NCO), 141.90 (Cq, aryl), 128.58/127.57/126.93 (Caryl), 79.77 (C CH), 76.04
(CH2O), 71.03 (C CH), 69.64 (NCH), 48.47 (CCH2C CH), 25.68 ppm
(CH2C CH); IR (KBr): n˜ =3291 (m; C CH), 3070–3005 (w; CHaryl),
2980–2850 (m; CH2), 2108 (w; C C), 1663 cm (s; C=N); MS (FAB):
ꢀ
ꢀ
ꢀ
ꢀ
À1
ꢀ
m/z (%): 490.1 (100) [M+H]+, 464.1 (5) [MÀC2H]+, 345.0 (10)
[MÀC2HÀC8H8O]+; HRMS (FAB): m/z calcd for C31H28N3O3 [M+H]+:
490.2130; found: 490.2145; elemental analysis calcd (%) for C31H27N3O3
(489.56)·0.9MeOH: C 73.91, H 5.95, N 8.11; found C 73.96, H 6.00, N
7.85.
yielding 2b as
a colourless, extremely viscous oil (847 mg, 95%).
1H NMR (600.13 MHz, CDCl3, 293 K): d=7.35–7.26 (m, 10H; CHaryl),
5.26 (m, 2H; NCH), 4.70 (m, 2H; CH2O), 4.18 (m, 2H; CH2O), 3.05 (d,
General procedure for the preparation of TBDMS-functionalised ligands:
A solution of the corresponding ligand (0.3–1.3 mmol) in THF (10–
20 mL) was cooled to À788C and LDA, nBuLi or tBuLi (1.1–1.2 equiv)
were added. The resulting mixture was stirred for 30 min and a solution
of TBDMSCl (1.2–1.3 equiv) in THF was added. After warming to room
temperature, it was additionally stirred for 3–4 d. The solvent was re-
moved in vacuo and the residue redissolved in dichloromethane, washed
with NH4Claq and dried over Na2SO4. Filtration and removal of the sol-
vent gave an oily crude product which was purified by column chroma-
tography.
4J=2.48 Hz, 2H; CH2C CH), 2.10 (t, 4J=2.39 Hz, 1H; CH2C CH),
1.81 ppm (s, 3H; CCH3); 13C{1H} NMR (150.90 MHz, CDCl3, 293 K): d=
168.29/168.20 (NCO), 142.13/142.05 (Cq, aryl), 128.69/128.62/127.61/126.84/
ꢀ
ꢀ
ꢀ
ꢀ
126.63 (Caryl), 79.85 (C CH), 75.71/75.63 (CH2O), 71.28 (C CH), 69.74/
ꢀ
69.52 (NCH), 42.08 (CCH3), 27.12 (CH2C CH), 21.41 ppm (CCH3); IR
ꢀ
(KBr): n˜ =3291 (m; C CH), 3070–3000 (w; CHaryl), 3000–2840 (m; CH2),
À1
ꢀ
2121 (w; C C), 1654 cm (s; C=N); MS (FAB): m/z (%): 359.1 (100)
[M+H]+, 357.1 (6) [MÀH]+, 239.0 (5) [M+HÀC8H8O]+; HRMS (FAB):
m/z calcd for C23H23N2O2 [M+H]+: 359.1760; found: 359.1771; elemental
analysis calcd (%) for C23H22N2O2 (358.43): C 77.07, H 6.19, N 7.82;
found C 76.85, H 6.25, N 7.48.
1-[tert-Butyl
ACHTUNGTRENNUNG
yne (TBDMS-propargyl-BOXAHCTNUGTRENNNUG
(600.13 MHz, CDCl3, 293 K): d=4.21 (m, 2H; CH2O), 4.02–3.91 (m, 4H;
CH2O, NCH), 3.00 (d, 2J=16.93 Hz, 1H; CH2C C), 2.84 (d, 2J=
ꢀ
4,4,4-Tris[(S)-4-isopropyloxazolin-2-yl]but-1-yne (propargyl-trisoxACTHNUTRGNE(NUG iPr);
ꢀ
16.93 Hz, 1H; CH2C C), 1.83–1.73 (m, 2H; CH
T
3a): A solution of bisoxazoline 1a (970 mg, 3.5 mmol) in dry toluene
(100 mL) was cooled to À788C and tBuLi (2.27 mL, 1.7m in hexane) was
added; 15 min later (S)-2-bromo-4-isopropyloxazoline (943 mg, 4.9 mmol,
62% w/w in THF) was transferred to the yellow solution, giving rise to
an orange-brown colour. The mixture was warmed to room temperature,
concentrated to eliminate the hexane originating from tBuLi, and heated
at 1008C for 10 d. The resulting deep brown solution was evaporated to
dryness and the residue was redissolved in dichloromethane (100 mL),
washed with H2O (20 mL) and dried over Na2SO4. Filtration and removal
of the solvent in vacuo yielded a brown oil which was purified by column
chromatography (EtOAc/MeOH 97.5/2.5) yielding 3b as a yellow, viscous
oil (934 mg, 69%). 1H NMR (399.89 MHz, CDCl3, 293 K): d=4.27 (dd,
CCH3), 0.93 (d, 3J=6.82 Hz, 3H; CH
U
AHCUTNGERT(GUNNN CH3)3, CHACHTUNTREGN(GNUN CH3)2), 0.85 (m, 6H; CHACTHUNTGRENNUGN
AHCTUNGTRENNUNG
ꢀ
ꢀ
102.99 (C C-TBDMS), 85.27 (C C-TBDMS), 71.88/71.55 (NCH), 70.23/
ꢀ
ꢀ
69.99 (CH2O), 41.88 (CCH2C C), 32.37/32.13 (CH
C), 26.05 (C(CH3)3), 21.29 (CCH3), 18.74/18.54 (CH
(CH(CH3)2), 16.49 (C
(CH3)2); 29Si{1H} NMR
ACHTUNGTRENNUNG
G
ACHTUNGTRENNUNG
E
(CH3)3), À4.57 ppm (Si
ACHTUNGTRENNUNG
(79.44 MHz, CDCl3, 293 K): d=À8.86; IR (KBr): n˜ =2958–2858 (m;
À1
ꢀ
CH2), 2177 (m; C C), 1662 cm (s; C=N); MS (FAB): m/z (%): 405.2
(100) [M+H]+, 361.2 (2) [MÀC3H7]+; HRMS (FAB): m/z calcd for
C23H41N2O2Si [M+H]+: 405.2937; found: 405.2925; elemental analysis
calcd (%) for C23H40N2O2Si (404.66): C 68.27, H 9.96, N 6.92; found C
68.10, H 9.86, N 6.84.
2J=9.48 Hz, 3J=8.07 Hz, 3H; CH2O), 4.07 (m, 3H; CH2O), 4.03–3.98
4
(m, 3H; NCH), 3.24 (dd, 2J=16.69 Hz, J=2.64 Hz, 1H; CH2C CH),
ꢀ
2
4
4
ꢀ
3.14 (dd, J=16.70 Hz, J=2.66 Hz, 1H; CH2C CH), 1.99 (t, J=2.64 Hz,
1-[tert-ButylACTHNUTRGNE(UNG dimethyl)silyl]-4,4-bis[(S)-4-phenyloxazolin-2-yl]pent-1-yne
3
(TBDMS-propargyl-BOX(Ph); 4b): Yield: 62%; 1H NMR (600.13 MHz,
CDCl3, 293 K): d=7.34–7.26 (m, 10H; CHaryl), 5.24 (m, 2H; NCH), 4.68
ꢀ
1H; CH2C CH), 1.80 (m, 3H; CH
(CH3)2), 0.92 (d, J=6.80 Hz, 9H; CH-
(CH3)2), 0.87 ppm (d, 3J=6.77 Hz, 9H; CH
U
(m, 2H; CH2O), 4.19–4.11 (m, 2H; CH2O), 3.17 (d, 2J=16.91 Hz, 1H;
ꢀ
(100.56 MHz, CDCl3, 293 K): d=162.44 (NCO), 80.00 (C CH), 71.66
2
ꢀ
ꢀ
ꢀ
ꢀ
(NCH), 70.28 (CH2O), 70.22 (C CH), 47.92 (CCH2C CH), 32.21 (CH-
CH2C C), 3.01 (d, J=16.90 Hz, 1H; CH2C C), 1.80 (s, 3H; CCH3), 0.94
(s, 9H;
(CH3)2); 13C{1H} NMR
CACHTNGUTREUNN(G CH3)3), 0.10 ppm (s, 6H; SiCAHTUNGTRENNNUG
ꢀ
T
(CH3)2), 17.87/17.71 ppm
(CH3)2); IR (KBr): n˜ =2960–2874 (m; CH2), 2108 (w; C C),
ꢀ
(CH
U
(150.90 MHz, CDCl3, 293 K): d=168.56/168.33 (NCO), 142.26/142.07
1665 cmÀ1 (s; C=N); MS (EI): m/z (%): 387.2 (4) [M]+, 344.1 (100)
[MÀC3H7]+, 275.1 (30) [MÀC6H10NO]+; HRMS (FAB): m/z calcd for
C22H34N3O3 [M+H]+: 388.2600; found: 388.2597; elemental analysis calcd
(%) for C22H33N3O3 (387.52): C 68.19, H 8.58, N 10.84; found C 68.39, H
8.56, N 10.29.
ꢀ
(Cq, aryl), 128.66/127.55/126.72/126.67 (Caryl), 102.64 (C C-TBDMS), 85.81
ꢀ
(C C-TBDMS), 75.63/75.58 (CH2O), 69.71/69.57 (NCH), 42.32 (CCH3),
ꢀ
28.53 (CH2C C), 26.04 (C
(CH3)3), 21.42 (CCH3), 16.51 (C
G
À4.52 ppm (Si
(CH3)2); 29Si{1H} NMR (79.44 MHz, CDCl3, 293 K): d=
À8.61 ppm; IR (KBr): n˜ =3090–3004 (w; CHaryl), 3000–2854 (m; CH2),
5458
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5450 – 5462

Homogeneous Catalysis
FULL PAPER
À1
ꢀ
2175 (m; C C), 1668, 1664 cm (s; C=N); MS (FAB): m/z (%): 473.2
(100) [M+H]+, 471.2 (3) [MÀH]+, 353.1 (1) [M+HÀC8H8O]+, 319.1 (4)
[MÀC9H17Si]+; HRMS (FAB): m/z calcd for C29H37N2O2Si [M+H]+:
473.2624; found: 473.2693; elemental analysis calcd (%) for
C29H36N2O2Si (472.69): C 73.69, H 7.68, N 5.93; found C 73.28, H 7.61, N
5.92.
16H; CH2O), 4.04–3.92 (m, 32H; CH2O, NCH), 2.95 (d, 2J=17.06 Hz,
2
ꢀ
ꢀ
8H; CH2C C), 2.84 (d, J=17.00 Hz, 8H; CH2C C), 1.83–1.75 (m, 16H;
CH
3J=6.68 Hz, 24H; CH
0.86 (d, 3J=6.79 Hz, 24H; CH
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
N
ACHTUNGTRENNUNG
ACHNUTERTGGNN(UN CH3)2), 0.66–0.60 (m, 16H; Hb), 0.58–0.51 (m, 32H; Hd, Hf, Hh), 0.08 (s,
48H; Ha), À0.08 ppm (s, 12H; He); 13C{1H} NMR (150.90 MHz, CDCl3,
1-[tert-ButylACHTUNGTRENNUNG(dimethyl)silyl]-4,4,4-tris[(S)-4-isopropyloxazolin-2-yl]but-1-
yne (TBDMS-propargyl-trisoxACHTUNGTRNEUNG
(iPr); 5): Yield: 96%; 1H NMR
ꢀ
ꢀ
293 K): d=166.69/166.64 (NCO), 102.73 (C C-Si), 86.59 (C C-Si), 71.77/
71.48 (NCH), 70.16/70.01 (CH2O), 41.83 (CCH3), 32.15/32.12 (CH-
(399.89 MHz, CDCl3, 293 K): d=4.25 (dd, 2J=9.37, 3J=8.39 Hz, 3H;
CH2O), 4.05 (m, 3H; CH2O), 3.97 (ddd, 3J=9.60 Hz, 3J=7.34 Hz, 3J=
ꢀ
(CH3)2), 28.29 (CH2C C), 21.30 (CCH3), 20.98 (Cb), 18.75/18.48 (CH-
6.10 Hz, 3H; NCH), 3.31 (d, J=16.88 Hz, 1H; CH2C C), 3.15 (d, 2J=
2
ꢀ
A
N
(Ca), À5.07 ppm (Ce); 29Si{1H} NMR (79.44 MHz, CDCl3, 293 K): d=1.04
ꢀ
16.81 Hz, 1H; CH2C C), 1.78 (m, 3H; CH
N
9H; CH
(CH3)2), 0.90 (s, 9H; C
A
(Si
(CH2)3CH3), 0.42 (Si
E
ACHTUGNTRENUN(NG CH3)2); IR (KBr): n˜ =
À1
ꢀ
(CH3)2), 0.04 (s, 3H; Si
R
ACHTUNGTRENNUNG
2960, 2915, 2875 (s; CH2), 2177 (w; C C), 1655 cm (s; C=N); MS
(MALDI): m/z (%): 3485 (76) [M+H]+, 3231 (24) [M+HÀC14H24N2O2]+
, 3136 (31) [M+HÀC19H31N2O2Si]+; HRMS (MALDI): m/z calcd for
ꢀ
NMR (100.56 MHz, CDCl3, 293 K): d=162.56 (NCO), 102.84 (C C-
TBDMS), 84.29 (C C-TBDMS), 71.88 (NCH), 70.60 (CH2O), 48.05
ꢀ
C
192H333N16O16Si13 [M+H]+: 3483.2731; found: 3483.2798.
ꢀ
ꢀ
(CCH2C C), 32.43 (CH
(CH3)2), 26.88 (CH2C C), 26.07 C
N
{G2}-[propargyl-BOX
(iPr)]16
E
Yield:
66%;
1H NMR
(CH
(CH3)2), 17.91 (CH
(CH3)2), 16.57 (C
G
ACHTUNGTRENNUNG
29Si{1H} NMR (79.44 MHz, CDCl3, 293 K): d=À9.09; IR (KBr): n˜ =
(600.13 MHz, CDCl3, 293 K): d=4.19 (m, 32H; CH2O), 4.03–3.92 (m,
À1
ꢀ
2961–2858 (m; CH2), 2180 (w; C C), 1677 cm (s; C=N); MS (FAB):
2
64H; CH2O, NCH), 2.96 (d, J=16.92 Hz, 16H; CH2C C), 2.84 (d, 2J=
ꢀ
m/z (%): 502.3 (100) [M+H]+, 348.3 (4) [MÀC9H17Si]+; HRMS (FAB):
m/z calcd for C28H48N3O3Si [M+H]+: 502.3465; found: 502.3445; elemen-
tal analysis calcd (%) for C28H47N3O3Si (501.78): C 67.02, H 9.44, N 8.37;
found C 67.01, H 9.26, N 8.19.
ꢀ
16.87 Hz, 16H; CH2C C), 1.84–1.74 (m, 32H; CH
A
CCH3), 1.40–1.31 (m, 56H; Hc, Hg, Hk), 0.93 (d, 3J=6.79 Hz, 48H; CH-
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
General procedure for the preparation of the functionalised dendrimers:
The appropriate amount of ligand (1.1–1.8 equiv with respect to each
chlorosilyl function of the carbosilane dendrimer) was dissolved in dry
toluene (3 mL), cooled to À788C and LDA (1.0 equiv with respect to the
amount of ligand, 1.8m in THF/heptane/ethylbenzene) was added slowly.
The resulting solution was stirred at À408C for 30 min. A solution of the
corresponding chlorosilyl-functionalised dendrimer (10–70 mmol) in dry
toluene (2 mL) and solid TlPF6 (1.1–1.6 equiv per chlorosilyl function)
were added. The mixture was warmed to room temperature over night
and stirred for several days. After removal of the solvent in vacuo, the
residue was redissolved in dichloromethane, washed with H2O (1–2 mL),
and dried over Na2SO4. The solvent was again removed, the residue re-
dissolved in MeOH, and filtered through a biochemical filter (0.45 mm
pore size). This step was repeated once or twice. {G0}-[propargyl-BOX-
32H; Hb), 0.59–0.49 (m, 80H; Hd, Hf, Hh, Hj, Hl), 0.08 (s, 96H; Ha),
À0.08 (s, 24H; He), À0.09 ppm (s, 12H; Hi); 13C{1H} NMR (150.90 MHz,
ꢀ
ꢀ
CDCl3, 293 K): d=166.78/166.63 (NCO), 102.70 (C C-Si), 86.57 (C C-
Si), 71.77/71.46 (NCH), 70.14/69.98 (CH2O), 41.81 (CCH3), 32.25/32.13
ꢀ
(CH
(CH
G
G
(CH3)2), À1.55 (Ca), À5.05 ppm (Ce, Ci); 29Si{1H} NMR (79.44 MHz,
CDCl3, 293 K): d=1.09/0.85 (Si
(SiACHTUNGTRNEU(GN CH3)2); IR (KBr): 2959, 2913, 2874 (s; CH2), 2177 (w; C C),
1662 cmÀ1 (s; C=N); elemental analysis calcd (%) for C400H700N32O32Si29
(7284.51): C 65.95, H 9.69, N 6.15; found C 65.26, H 9.25, N 5.85.
{G0}-[propargyl-BOX(Ph)]4
(G0-2b):
Yield:
22%;
1H NMR
(600.13 MHz, CDCl3, 293 K): d=7.36–7.26 (m, 40H; CHaryl), 5.25 (m,
8H; NCH), 4.68 (m, 8H; CH2O), 4.14 (m, 8H; CH2O), 3.13 (d, 2J=
ACHTUNGTRENNUNG(iPr)]4 and {G0}-[propargyl-BOX(Ph)]4 were then purified by flash filtra-
2
ꢀ
ꢀ
16.90 Hz, 4H; CH2C C), 3.03 (d, J=16.90 Hz, 4H; CH2 C), 1.80 (s,
12H; CCH3), 1.38 (m, 8H; Hc), 0.69 (m, 8H; Hb), 0.59 (m, 8H; Hd),
0.14 ppm (s, 24H; Ha); 13C{1H} NMR (150.90 MHz, CDCl3, 293 K): d=
168.44/168.38 (NCO), 142.23/142.07 (Cq, aryl), 128.65/127.54/126.75/126.62
tion through silica gel (pentane/EtOAc 60:40, later EtOAc/MeOH
97.5:2.5 to 95:5). All other derivatives were purified by dialysis: the
membrane was washed with dichloromethane (3ꢅ) and filled with a con-
centrated solution of the crude product in dichloromethane. The resulting
bag was placed into pure dichloromethane (200 mL) and gently stirred
for 8–16 h. Then the exterior solvent was replaced. This was repeated up
to four times, depending on the amount of residual free ligand and other
impurities in the sample.
ꢀ
ꢀ
(Caryl), 102.44 (C C-Si), 87.15 (C C-Si), 75.63/75.58 (CH2O), 69.72/69.50
ꢀ
(NCH), 42.26 (CCH3), 28.50 (CH2C C), 21.43 (CCH3), 21.09 (Cb), 18.38
(Cc), 17.06 (Cd), À1.52 ppm (Ca); 29Si{1H} NMR (79.44 MHz, CDCl3,
293 K): d=0.89 (Si
(CH2)4), À17.68 ppm (Si
ACHTUGNTRENUN(GN CH3)2); IR (KBr): n˜ =3086,
ꢀ
3063, 3030 (w; CHaryl), 2960, 2915, 2875 (s; CH2), 2177 (m; C C),
1662 cmÀ1 (s; C=N); MS (MALDI): m/z (%): 1858 (73) [M+H]+, 1618
(5) [M+HÀC16H16O2]+; HRMS (MALDI): m/z calcd for C112H133N8O8Si5
[M+H]+: 1857.9087; found: 1857.9106.
{G0}-[propargyl-BOX
(iPr)]4
E
Yield:
28%;
1H NMR
(600.13 MHz, CDCl3, 293 K): d=4.20 (dd, 2J=9.47 Hz, 3J=8.12 Hz, 4H;
CH2O), 4.18 (dd, 2J=9.47 Hz, 3J=8.17 Hz, 4H; CH2O), 4.02–3.98 (m,
2
ꢀ
8H; CH2O), 3.98–3.93 (m, 8H; NCH), 2.96 (d, J=16.92 Hz, 4H; CH2C
{G1}-[propargyl-BOX(Ph)]8
(G1-2b):
Yield:
52%;
1H NMR
2
C), 2.84 (d, J=16.92 Hz, 4H; CH2C C), 1.79 (m, 8H; CH
(CH3)2), 1.61
3
(600.13 MHz, CDCl3, 293 K): d=7.35–7.25 (m, 80H; CHaryl), 5.23 (m,
(s, 12H; CCH3), 1.34 (m, 8H; Hc), 0.93 (d, J=6.81 Hz, 12H; CH
(CH3)2),
16H; NCH), 4.67 (m, 16H; CH2O), 4.13 (m, 16H; CH2O), 3.12 (d, 2J=
ACTHNUTRGNEUNG
0.90 (d, 3J=6.82 Hz, 12H; CH(CH3)2), 0.87 (d, 3J=6.78 Hz, 12H; CH-
2
ACHTUNGTRENNUNG ACHTUNTGRNE(NUNG CH3)2), 0.64 (m, 8H; Hb), 0.56
(CH3)2), 0.85 (d, 3J=6.79 Hz, 12H; CH
ꢀ
ꢀ
16.88 Hz, 8H; CH2C C), 3.02 (d, J=16.85 Hz, 8H; CH2C C), 1.79 (s,
24H; CCH3), 1.38 (m, 24H; Hc, Hg), 0.68 (m, 16H; Hb), 0.63–0.49 (m,
32H; Hd, Hf, Hh), 0.13 (s, 48H; Ha), À0.07 ppm (s, 12H; He); 13C{1H}
NMR (150.90 MHz, CDCl3, 293 K): d=168.33 (NCO), 142.33/142.11
(m, 8H; Hd), 0.081 (s, 12H; Ha), 0.080 ppm (s, 12H; Ha); 13C{1H} NMR
ꢀ
(150.90 MHz, CDCl3, 293 K): d=166.75/166.60 (NCO), 102.67 (C C-Si),
ꢀ
86.56 (C C-Si), 71.75/71.44 (NCH), 70.11/69.96 (CH2O), 41.79 (CCH3),
ꢀ
ꢀ
ꢀ
32.27/32.10 (CH
(CH3)2), 28.23 (CH2C C), 21.25 (CCH3), 21.06 (Cb),
(Cq,aryl), 128.66/127.55/126.76/126.63 (Caryl), 102.36 (C C-Si), 87.12 (C C-
18.70/18.43 (CH
A
(CH3)2), 17.02 (Cd),
ꢀ
Si), 75.63 (CH2O), 70.41 (NCH), 42.41 (CCH3), 28.47 (CH2C C), 21.44
À1.63/-1.65 ppm (Ca); 29Si{1H} NMR (79.44 MHz, CDCl3, 293 K): d=0.73
(CCH3), 21.09 (Cb), 18.63 (Cd, Cf, Ch), 18.28 (Cc, Cg), À1.46 (Ca),
À4.93 ppm (Ce); 29Si{1H} NMR (79.44 MHz, CDCl3, 293 K): d=1.05 (Si-
(Si
(CH2)4), À17.93 ppm (Si
ACHTUGNERTN(NUNG CH3)2); IR (KBr): n˜ =2960, 2915, 2875 (s;
À1
ꢀ
CH2), 2177 (w; C C), 1662 cm (s; C=N); MS (FAB): m/z (%): 1856
A
N
ACHTUGNTRENUN(GN CH3)2); IR (KBr): n˜ =3086,
(17) [M+H]+, 1196 (5) [MÀC22H37N2O2Si]+, 807 (2) [MÀC44H74N4O
Si2]+
4ACHTUNGTRENNUNG
ꢀ
3063, 3030 (w; CHaryl), 2956, 2913, 2874 (s; CH2), 2177 (m; C C),
1655 cmÀ1 (s; C=N); MS (MALDI-TOF): m/z (%): 4051 (100) [M+Na]+,
4029 (71) [M+H]+, 3707 (37) [MÀC20H19N2O2]+, 3028 (28)
[M+HÀC60H75N4O4Si3]+; MS (MALDI-TOF): m/z (%): 2038 (98)
,
389 (24) [C22H37N2O2Si]+; HRMS (MALDI): m/z calcd for
C88H149N8O8Si5 [M+H]+: 1586.0339; found: 1586.0330,.
{G1}-[propargyl-BOX(iPr)]8 (G1-2a): Yield:
(600.13 MHz, CDCl3, 293 K): d=4.19 (dd, 2J=18.97 Hz, 3J=9.27 Hz,
A
R
77%;
1H NMR
[M+2Na]2+
.
Chem. Eur. J. 2009, 15, 5450 – 5462
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5459

L. H. Gade, S. Bellemin-Laponnaz and M. Gaab
{G2}-[propargyl-BOX(Ph)]16
G
Yield:
49%;
1H NMR
ing ligand or functionalised dendrimer (respective amount) in MeOH
(1.00 mL) were prepared under air. Successive aliquots of both homoge-
neous solutions (containing 1.5 mmol of [CuACTHNUTRGENUG(N OTf)2] and 1.8 mmol of the
(600.13 MHz, CDCl3, 293 K): d=7.36–7.18 (m, 160H; CHaryl), 5.22 (m,
32H; NCH), 4.65 (m, 32H; CH2O), 4.11 (m, 32H; CH2O), 3.11 (d, 2J=
ꢀ
ꢀ
16.73 Hz, 16H; CH2C C), 3.00 (d, J=16.69 Hz, 16H; CH2C C), 1.77 (s,
48H; CCH3), 1.44–1.31 (m, 56H; Hc, Hg, Hk), 0.71–0.63 (m, 32H; Hb),
0.63–0.44 (m, 80H; Hd, Hf, Hh, Hj, Hl), 0.12 (s, 96H; Ha), À0.09 ppm (s,
36H; He, Hi); 13C{1H} NMR (150.90 MHz, CDCl3, 293 K): d=168.40
(NCO), 142.22/142.06 (Cq,aryl), 128.65/127.55/126.75/126.62 (Caryl), 102.45
corresponding ligand species) were taken and reacted for 1 h to obtain
the catalyst for each run. The resulting turquoise/light-green solution was
evaporated to dryness (30 min) and the complex redissolved in trifluoroe-
thanol (1.00 mL). Ethyl 2-methylacetoacetate (21.5 mL, 0.15 mmol) was
added to the solution prior to cooling to 08C and addition of a precooled
solution of dibenzylazodicarboxylate (54.7 mg, 0.18 mmol) in trifluoroe-
thanol (0.50 mL). After 16 h at 08C, the products were isolated by flash
column chromatography (pentane/EtOAc 80:20). The enantioselectivity
of the product was determined by HPLC, using a Daicel Chiralpak AD-
H column (hexane/iPrOH 90:10, 82 bar, 10 mL, 0.95 mLminÀ1, detection
ꢀ
ꢀ
(C C-Si), 87.13 (C C-Si), 75.64/75.59 (CH2O), 69.71/69.50 (NCH), 42.26
ꢀ
(CCH3), 28.50 (CH2C C), 21.45 (CCH3), 20.96 (Cb), 19.03 (Cd, Cf, Ch, Cj,
Cl), 18.35 (Cc, Cg, Ck), À1.46 (Ca), À5.03 ppm (Ce, Ci); 29Si{1H} NMR
(79.44 MHz, CDCl3, 293 K): d=1.03/0.82 (SiACTHUNGRTENNUG(CH2)3CH3), 0.40 (SiACHTUNGTREN(NUNG CH2)4),
À17.65 ppm (Si
ACHTUNGTRENNUNG(CH3)2); IR (KBr): n˜ =3086, 3062, 3031 (w; CHaryl), 2956,
À1
2913, 2876 (s; CH2), 2177 (m; C C), 1655 cm (s; C=N); elemental anal-
ysis calcd (%) for C496H636N32O32Si29 (8373.03): C 71.15, H 7.66, N 5.35;
found C 70.06, H 7.21, N 5.19.
213 nm, 225 nm, 254 nm,
tRACHTGUNTRENNU(G maj)=33.0 min, tRCAHTUNGTREN(NUNG min)=36.3 min). All
given values were determined as average of three corroborating runs.
Monitoring of conversion curves: Catalyses were conducted as described
above. The progress of the reaction was monitored by measuring the dis-
appearance of ethyl 2-methylacetoacetate by GC, using methyl hexa-
{G0}-[propargyl-trisox
N
Yield:
33%;
1H NMR
(600.13 MHz, CDCl3, 293 K): d=4.23 (m, 12H; CH2O), 4.05 (m, 12H;
CH2O), 3.99 (td, 3J=9.10 Hz, 3J=6.95 Hz, 12H; NCH), 3.30 (d, 2J=
noate as internal standard. The following GC method was applied: Tinj
=
ꢀ
ꢀ
16.72 Hz, 4H; CH2C C), 3.07 (d, J=16.77 Hz, 4H; CH2C C), 1.79 (m,
2008C, Tdet =2508C, 20 mLminÀ1 He flow, splitless, temperature program:
408C, 1 min, 258CminÀ1 up to 2708C, 2708C, 2 min; tR (methyl hexa-
noate)=4.3 min, tR (substrate)=4.9 min.
12H; CHACHTUNGTRENNUNG
(CH3)2), 1.30 (m, 8H; Hc), 0.91 (d, 3J=6.62 Hz, 36H; CH-
ACHTUNGTRENNUNG ACHTUNTGRNE(NUNG CH3)2), 0.62 (m, 8H; Hb), 0.52
(CH3)2), 0.86 (d, 3J=6.61 Hz, 36H; CH
(m, 8H; Hd), 0.05 (s, 12H; Ha), 0.04 ppm (s, 12H; Ha); 13C{1H} NMR
ꢀ
General procedure for the catalytic Henry reaction of nitromethane and
2-nitrobenzaldehyde: Stock solutions of [CuACTHNGUETRNNU(G OAc)2] hydrate (9 mg) in
MeOH (1.50 mL) and of the corresponding ligand or functionalised den-
drimer (respective amount) in MeOH (1.00 mL) were prepared under
air. Successive aliquots of both homogeneous solutions (containing
(150.90 MHz, CDCl3, 293 K): d=162.44 (NCO), 102.50 (C C-Si), 85.57
ꢀ
ꢀ
(C C-Si), 71.80 (NCH), 70.55 (CH2O), 47.98 (CCH2C C), 32.40 (CH-
ꢀ
A
U
(CHACHTUNGTRENNUNG
CDCl3, 293 K): d=0.73 (Si
(CH2)4), À18.29 ppm (Si
ACHTUNGTRENNUNG
À1
ꢀ
n˜ =2961, 2919, 2877 (s; CH2), 2181 (m; C C), 1669, 1666 cm (s; C=N);
MS (MALDI): m/z (%): 1974 (39) [M+H]+, 1863 (75)
[M+HÀC6H10NO]+; HRMS (MALDI): m/z calcd for C108H177N12O12Si5
[M+H]+: 1974.2450; found: 1974.2461.
1.5 mmol of [CuACHTNUGTRENUNG(OAc)2] hydrate and 1.8 mmol of the corresponding ligand
species) were taken and reacted for 1 h to obtain the catalyst for each
run. The resulting green solution was evaporated to dryness (30 min) and
the complex redissolved in trifluoroethanol/iPrOH 2:1 (1.00 mL) prior to
the addition of nitromethane (410 mL) and a solution of 2-nitrobenzalde-
hyde (22.7 mg, 0.15 mmol) in trifluoroethanol/iPrOH 2:1 (0.50 mL). The
resulting brownish solution was stirred at 238C for 3 d. Then, the mixture
was filtered through a small pad of silica gel to remove the catalyst. The
solvent was removed in vacuo and the product was isolated by flash
column chromatography (pentane/EtOAc 80:20). The enantioselectivity
of the product was determined by HPLC, using a Daicel Chiralpak OD-
H column (hexane/iPrOH 90:10, 64 bar, 10 mL, 0.8 mLminÀ1, detection
{G1}-[propargyl-trisox(iPr)]8
E
Yield:
54%;
1H NMR
(600.13 MHz, CDCl3, 293 K): d=4.21 (m, 24H; CH2O), 4.02 (m, 24H;
CH2O), 3.95 (td, 3J=9.37 Hz, 3J=6.85 Hz, 24H; NCH), 3.28 (d, 2J=
2
ꢀ
ꢀ
16.73 Hz, 8H; CH2C C), 3.05 (d, J=16.72 Hz, 8H; CH2C C), 1.74 (m,
3
24H; CHACHTUNGTRENNUNG(CH3)2), 1.29 (m, 24H; Hc, Hg), 0.88 (d, J=6.73 Hz, 72H; CH-
ACHTUNGTRENNUNG ACHTUNTGREN(NUNG CH3)2), 0.59 (m, 16H; Hb), 0.49
(CH3)2), 0.83 (d, 3J=6.72 Hz, 72H; CH
(m, 32H; Hd, Hf, Hh), 0.03 (s, 48H; Ha), 0.01 ppm (s, 12H; He); 13C{1H}
ꢀ
NMR (150.90 MHz, CDCl3, 293 K): d=162.48 (NCO), 102.54 (C C-Si),
85.58 (C C-Si), 71.82 (NCH), 70.58 (CH2O), 48.01 (CCH2C C), 32.43
ꢀ
ꢀ
225 nm, 254 nm, tRACHTNUGRTNENUG(min)=17.8 min, tRACHTUNGTNER(NUGN maj)=19.6 min). All given values
were determined as average of at least two corroborating runs.
ꢀ
(CH
G
U
Cf, Ch), 18.24 (Cc, Cg), 17.96 (CH
General procedure for the catalytic recycling using a dialysis membrane:
Membrane pieces (Sigma-Aldrich: benzoylated dialysis tubing, avg. flat
width 32 mm) of identical length (6.5 cm) were washed with trifluoroe-
thanol (4ꢅ). Stock solutions of [CuACHTNUGTRENUNG(OTf)2] (10.9 mg) and G2-2b or G2-
3a (respective amount) in MeOH (1.00 mL) were prepared. Successive
aliquots of both homogeneous solutions (containing 1.5 mmol of [Cu-
29Si{1H} NMR (79.44 MHz, CDCl3, 293 K): d=1.05 (Si
(Si
ACHTUNGTRENNUNG
(CH2)4), À18.22 ppm (Si
ACHTUNGTRENNUNG
CH2), 2182 (m; C C), 1665, 1660 cm (s; C=N); MS (MALDI-TOF): m/
z (%): 4327 (<1) [M+C2H8O2]+, 3201 (<1) [MÀC58H97N6O6Si3]+; ele-
mental analysis calcd (%) for C232H388N24O24Si13 (4262.82): C 65.37, H
9.17, N 7.89; found C 65.04, H 9.02, N 7.34.
AHCTUNGRTEG(NNNU OTf)2] and 0.11 mmol of the corresponding dendrimer) were taken and
{G2}-[propargyl-Trisox(iPr)]16
E
Yield:
50%;
1H NMR
reacted for 1 h to obtain the catalyst for each run. The solvent was re-
moved in vacuo (30 min). Then, the residue was redissolved in trifluoroe-
thanol (5 mL) and transferred into the membrane. The resulting bag was
placed into a screw cap vial and gently stirred in pure trifluoroethanol
(10 mL) at 408C for 10 h and subsequently transferred into a screw cap
(600.13 MHz, CDCl3, 293 K): d=4.25 (m, 48H; CH2O), 4.06 (m, 48H;
CH2O), 3.98 (td, 3J=9.09 Hz, 3J=6.86 Hz, 48H; NCH), 3.31 (d, 2J=
2
ꢀ
ꢀ
16.76 Hz, 16H; CH2C C), 3.09 (d, J=16.83 Hz, 16H; CH2C C), 1.82–
1.74 (m, 48H; CH
ACHTUNGTRENNUNG
6.74 Hz, 144H; CH
N
ACHTUNGTRENNUNG
vial containing
a solution of ethyl 2-methylacetoacetate (21.5 mL,
0.65–0.59 (m, 32H; Hb), 0.57–0.46 (m, 80H; Hd, Hf, Hh, Hj, Hl), 0.06 (s,
96H; Ha), À0.09 (s, 24H; He), À0.10 ppm (s, 12H; Hi); 13C{1H} NMR
0.15 mmol) and dibenzylazodicarboxylate (54.7 mg, 0.18 mmol) in tri-
fluoroethanol (10 mL). The reaction was allowed to proceed for the ap-
propriate time (10 h with G2-2b, 24 h with G2-3a) at 408C and then, the
catalyst bag was directly transferred into another vial containing a fresh
solution of substrates for the next run whereas the solution from the pre-
ceding run was concentrated in vacuo. Isolation of the product was pro-
vided by flash column chromatography (pentane/EtOAc 80:20). The
enantioselectivity of the product was determined by HPLC, using a
Daicel Chiralpak AD-H column (hexane/iPrOH 90:10, 82 bar, 10 mL,
ꢀ
(150.90 MHz, CDCl3, 293 K): d=162.45 (NCO), 102.51 (C C-Si), 85.54
ꢀ
ꢀ
(C C-Si), 71.81 (NCH), 70.56 (CH2O), 47.97 (CCH2C C), 32.43 (CH-
ꢀ
(CH3)2), 26.74 (CH2C C), 21.02 (Cb), 18.62 (CH
(CH3)2), 18.51 (Cd, Cf,
Ch, Cj, Cl), 18.24 (Cc, Cg, Ck), 17.98 (CH
(CH3)2), À1.59 (Ca), À5.03 (Ce),
À5.04 ppm (Ci); 29Si{1H} NMR (79.44 MHz, CDCl3, 293 K): d=1.03/0.82
(Si
(CH2)3CH3), 0.40 (Si
E
ACHTUGNTRENUN(NG CH3)2); IR (KBr): n˜ =
À1
ꢀ
2961, 2916, 2877 (s; CH2), 2182 (m; C C), 1669, 1665 cm (s; C=N); ele-
mental analysis calcd (%) for C480H812N48O48Si29 (8838.36): C 65.23, H
9.26, N 7.61; found C 64.26, H 9.02, N 7.65.
0.95 mLminÀ1, detection 213 nm, 225 nm, 254 nm, t
(min)=36.3 min). Values given for each run were determined as average
of at least five corroborating experiments.
RACHTUNGTREN(NGNU maj)=33.0 min, tR-
AHCTUNGTRENNUNG
General procedure for the catalytic a-hydrazination of ethyl 2-methylace-
toacetate: Stock solutions of [CuACHTUNTRGNE(UNG OTf)2] (10.9 mg) and of the correspond-
5460
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5450 – 5462

Homogeneous Catalysis
FULL PAPER
[11] For reviews on dendrimer catalysis with chiral catalysts, see: a) Y.
Ribourdouille, G. D. Engel, L. H. Gade, C. R. Chimie 2003, 6, 1087;
b) J. Kassube, L. H. Gade, Top. Organomet. Chem. 2006, 20, 61;
c) A.-M. Caminade, P. Servin, J.-P. Majoral, Chem. Soc. Rev. 2008,
37, 56.
[12] Examples: a) H.-F. Chow, C. C. Mak, J. Org. Chem. 1997, 62, 5116;
b) C. C. Mak, H.-F. Chow, Macromolecules 1997, 30, 1228; c) M.
Malkoch, K. Hallman, S. Lutsenko, A. Hult, E. Malmstrçm, C.
Moberg, J. Org. Chem. 2002, 67, 8197; d) B.-Y. Yang, X.-M. Chen,
G.-J. Deng, Y.-L. Zhang, Q.-H. Fan, Tetrahedron Lett. 2003, 44,
3535.
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft (SFB 623) for funding,
the Deutsche Telekom Stiftung for a pre-doctoral fellowship (to M.G.),
the Deutsch–Franzçsische Hochschule (UFA) and the European Doctor-
al College (CDE) for support.
[1] a) U. Kragl, T. Dwars, Trends Biotechnol. 2001, 19, 442; b) H. P.
Dijkstra, G. P. M. van Klink, G. van Koten, Acc. Chem. Res. 2002,
35, 798; c) C. Mꢆller, M. G. Nijkamp, D. Vogt, Eur. J. Inorg. Chem.
2005, 4011; d) A. V. Gaikwad, V. Boffa, J. E. ten Elshof, G. Rothen-
berg, Angew. Chem. 2008, 120, 5487; Angew. Chem. Int. Ed. 2008,
47, 5407.
[13] A. Gissibl, C. Padiꢃ, M. Hager, F. Jaroschik, R. Rasappan, E.
Cuevas-YaÇez, C.-O. Turrin, A.-M. Caminade, J.-P. Majoral, O.
Reiser, Org. Lett. 2007, 9, 2895.
[14] General review: a) H. Frey, C. Schlenk, Top. Curr. Chem. 2000, 210,
69; for reviews on transition-metal-containing carbosilane dendrim-
ers, see: b) O. Rossell, M. Seco, I. Angurell, C. R. Chimie 2003, 6,
803; c) J. N. H. Reek, D. de Groot, G. E. Oosteroom, P. C. J. Kamer,
P. W. N. M. van Leeuwen, C. R. Chimie 2003, 6, 1061; early key ref-
erences: d) N. Hadjichristidis, A. Guyot, L. J. Fetters, Macromole-
cules 1978, 11, 668; e) A. W. van der Made, P. W. N. M. van Leeu-
wen, J. Chem. Soc. Chem. Commun. 1992, 1400; f) A. W. van der
Made, P. W. N. M. van Leeuwen, J. C. de Wilde, R. A. C. Brandes,
Adv. Mater. 1993, 5, 466; g) L.-L. Zhan, J. Roovers, Macromolecules
1993, 26, 963; h) J. Roovers, P. M. Toporowski, L.-L. Zhan, Polym.
Prepr. Am. Chem. Soc. Div. Polym. Chem. 1992, 23, 182; i) A. M.
Muzafarov, O. B. Gorbatsevich, E. A. Rebrov, G. M. Ignat’eva, T. B.
Myakushev, A. F. Bulkin, V. S. Papkov, Polym. Sci. Ser. A 1993, 35,
1575; j) D. Seyferth, D. Y. Son, A. L. Rheingold, R. L. Ostrander,
Organometallics 1994, 13, 2682.
[2] a) M. R. Kula, C. Wandrey, Methods Enzymol. 1987, 136, 9; for
early work, see: b) E. Flaschel, C. Wandrey, DECHEMA Monogr.
1979, 337; c) C. Wandrey, R. Wichmann, A. F. Bꢆckmann, M. R.
Kula, Prepr. Eur. Congr. Biotechnol. 1979, 44; d) A. S. Jandel, H.
Hustedt, C. Wandrey, Eur. J. Appl. Microbiol. Biotechnol. 1982, 15,
59; for a recent overview, see also: e) A. Liese, K. Seelbach, C. Wan-
drey, Industrial Biostransformations, Wiley-VCH, Weinheim, 2000.
[3] M. D. Bednarski, H. K. Chenault, E. S. Simon, G. M. Whitesides, J.
Am. Chem. Soc. 1987, 109, 1283.
[4] U. Kragl, D. Vasic-Racki, C. Wandrey, Chem. Ing. Tech. 1992, 64,
499.
[5] a) U. Kragl, D. Gygax, O. Ghisalba, C. Wandrey, Angew. Chem.
1991, 103, 854; Angew. Chem. Int. Ed. Engl. 1991, 30, 827; b) U.
Kragl, A. Gçdde, C. Wandrey, Tetrahedron: Asymmetry 1993, 4,
1193; c) U. Kragl, C. Dreisbach, Angew. Chem. 1996, 108, 684;
Angew. Chem. Int. Ed. Engl. 1996, 35, 642.
[15] For reviews of chiral oxazoline ligands in asymmetric catalysis, see:
a) F. Fache, E. Schulz, M. Lorraine Tommasino, M. Lemaire, Chem.
Rev. 2000, 100, 2159; b) M. Gꢇmez, G. Muller, M. Rocamora,
Coord. Chem. Rev. 1999, 193, 769; c) A. K. Ghosh, P. Mathivanan, J.
Cappiello, Tetrahedron: Asymmetry 1998, 9, 1; d) H. A. McManus,
P. J. Guiry, Chem. Rev. 2004, 104, 4151; e) S. Dagorne, S. Bellemin-
Laponnaz, A. Maisse-FranÅois, Eur. J. Inorg. Chem. 2007, 913.
[16] a) L.-X. Dai, Angew. Chem. 2004, 116, 5846; Angew. Chem. Int. Ed.
2004, 43, 5726; b) N. E. Leadbeater, M. Marco, Chem. Rev. 2002,
102, 3217; c) P. McMorn, G. J. Hutchings, Chem. Soc. Rev. 2004, 33,
108.
[17] For the previous fixation of alkynyl units to carbosilane dendrimers,
see: a) C. Kim, M. Kim, J. Organomet. Chem. 1998, 563, 43; b) C.
Kim, I. Jung, J. Organomet. Chem. 1999, 588, 8; c) C. Kim, I. Jung,
J. Organomet. Chem. 2000, 599, 208; d) C. Kim, S. Son, J. Organo-
met. Chem. 2000, 599, 123; e) M. B. Meder, I. Haller, L. H. Gade,
Dalton Trans. 2005, 1403.
[18] J. Bourguignon, U. Bremberg, G. Dupas, K. Hallman, L. Hagberg,
L. Hortala, V. Levacher, S. Lutsenko, E. Macedo, C. Moberg, G.
Quꢃguiner, F. Rahm, Tetrahedron 2003, 59, 9583.
[19] M. Peer, J. C. de Jong, M. Kiefer, T. Langer, H. Rieck, H. Schell, P.
Sennhenn, J. Sprinz, H. Steinhagen, B. Wiese, G. Helmchen, Tetrahe-
dron 1996, 52, 7547.
[20] a) S. Bellemin-Laponnaz, L. H. Gade, Chem. Commun. 2002, 1286;
b) S. Bellemin-Laponnaz, L. H. Gade, Angew. Chem. 2002, 114,
3623; Angew. Chem. Int. Ed. 2002, 41, 3473; Review: c) L. H. Gade,
S. Bellemin-Laponnaz, Chem. Eur. J. 2008, 14, 4142.
[6] See for example: a) J. W. Knapen, A. W. van der Made, J. C. Wilde,
P. W. N. M. van Leeuwen, P. Wijkens, D. M. Grove, G. van Koten,
Nature 1994, 372, 659; b) N. J. Hovestad, E. B. Eggeling, H. J. Heid-
bꢆchel, J. T. B. H. Jastrzebski, U. Kragl, W. Keim, D. Vogt, G.
van Koten, Angew. Chem. 1999, 111, 1763; Angew. Chem. Int. Ed.
1999, 38, 1655; c) A. W. Kleij, R. A. Gossage, J. T. B. H. Jastrzebski,
J. Boersma, G. van Koten, Angew. Chem. 2000, 112, 179; Angew.
Chem. Int. Ed. 2000, 39, 176; d) A. W. Kleij, R. A. Gossage, R. J. M.
Klein Gebbink, N. Brinkmann, E. J. Reijerse, U. Kragl, M. Lutz,
A. L. Spek, G. van Koten, J. Am. Chem. Soc. 2000, 122, 12112;
e) A. W. Kleij, R. van de Coevering, R. J. M. Klein Gebbink, A.-M.
Noordman, A. L. Spek, G. van Koten, Chem. Eur. J. 2001, 7, 181;
f) H. P. Dijkstra, C. A. Kruithof, N. Ronde, R. van de Coevering,
D. J. Ramꢇn, D. Vogt, G. P. M. van Klink, G. van Koten, J. Org.
Chem. 2003, 68, 675; g) H. P. Dijkstra, N. Ronde, G. P. M. van Klink,
D. Vogt, G. van Koten, Adv. Synth. Catal. 2003, 345, 364; h) E. L. V.
Goetheer, A. W. Verkerk, L. J. P. van den Broeke, E. de Wolf, B.-J.
Deelman, G. van Koten, J. T. F. Keurentjes, J. Catal. 2003, 219, 126.
[7] a) G. E. Oosterom, J. N. H. Reek, P. C. J. Kamer, P. W. N. M. van
Leeuwen, Angew. Chem. 2001, 113, 1878; Angew. Chem. Int. Ed.
2001, 40, 1828; b) D. de Groot, B. F. M. de Waal, J. N. H. Reek,
A. P. H. J. Schenning, P. C. J. Kamer, E. W. Meijer, P. W. N. M. van
Leeuwen, J. Am. Chem. Soc. 2001, 123, 8453; c) R. van Heerbeek,
P. C. J. Kamer, P. W. N. M. van Leeuwen, J. N. H. Reek, Chem. Rev.
2002, 102, 3717; for a review, see: d) Q.-H. Fan, Y. M. Li, A. S. C.
Chan, Chem. Rev. 2002, 102, 3385.
[8] I. F. J. Vankelecom, Chem. Rev. 2002, 102, 3779.
[21] A. I. Meyers, K. A. Novachek, Tetrahedron Lett. 1996, 37, 1747.
[22] C. Foltz, S. Bellemin-Laponnaz, M. Enders, H. Wadepohl, L. H.
Gade, Org. Lett. 2008, 10, 305.
[23] Examples: a) R. van de Coevering, A. P. Alfers, J. D. Meeldijk, E.
Martꢈnez-Viviente, P. S. Pregosin, R. J. M. Klein Gebbink, G.
van Koten, J. Am. Chem. Soc. 2006, 128, 12700; b) J. Le Nꢉtre, J. J.
Firet, L. A. J. M. Sliedregt, B. J. van Steen, G. van Koten, R. J. M.
Klein Gebbink, Org. Lett. 2005, 7, 363.
[24] a) M. Marigo, K. Juhl, K. A. Jørgensen, Angew. Chem. 2003, 115,
1405; Angew. Chem. Int. Ed. 2003, 42, 1367; b) C. Foltz, B. Stecker,
G. Marconi, H. Wadepohl, S. Bellemin-Laponnaz, L. H. Gade,
Chem. Commun. 2005, 5115; c) C. Foltz, B. Stecker, G. Marconi, S.
[9] a) M. an der Heiden, H. Plenio, Chem. Eur. J. 2004, 10, 1789; b) L.
Greiner, S. Laue, A. Liese, C. Wandrey, Chem. Eur. J. 2006, 12,
1818; c) K. de Smet, A. Pleysier, I. F. J. Vankelecom, P. A. Jacobs,
Chem. Eur. J. 2003, 9, 334.
[10] The term “tea-bag catalyst” has been first coined by J. M. Thomas
for certain types of heterogeneous catalyst systems: a) J. M. Thomas,
Philos. Trans. R. Soc. A 1990, 333, 173; the concept of “catalyst in a
tea bag” has been discussed in: b) A. Berger, R. J. M. Klein Geb-
bink, G. van Koten, Top. Organomet. Chem. 2006, 20, 1; the strategy
of manipulating polymer-immobilised molecules in permeable bags
was developed for solid phase syntheses of peptides: c) R. A.
Houghten, Proc. Natl. Acad. Sci. USA 1985, 82, 5131.
Chem. Eur. J. 2009, 15, 5450 – 5462
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5461

L. H. Gade, S. Bellemin-Laponnaz and M. Gaab
Bellemin-Laponnaz, H. Wadepohl, L. H. Gade, Chem. Eur. J. 2007,
13, 9912.
[25] D. A. Evans, D. Seidel, M. Rueping, H. W. Lam, J. T. Shaw, C. W.
Downey, J. Am. Chem. Soc. 2003, 125, 12692.
[28] A. Abiko, S. Masamune, Tetrahedron Lett. 1992, 33, 5517.
[29] W. R. Leonard, J. L. Romine, A. I. Meyers, J. Org. Chem. 1991, 56,
1961.
[30] S. E. Denmark, C. E. Stiff, J. Org. Chem. 2000, 65, 5875.
[26] J. Thorhauge, M. Roberson, R. G. Hazell, K. A. Jørgensen, Chem.
Eur. J. 2002, 8, 1888.
[27] a) A. Schꢂtz, R. Rasappan, M. Hager, A. Gissibl, O. Reiser, Chem.
Eur. J. 2008, 14, 7259; Review: b) R. Rasappan, D. Laventine, O.
Reiser, Coord. Chem. Rev. 2008, 252, 702.
Received: February 24, 2009
Published online: April 22, 2009
5462
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5450 – 5462