Biomimetic catalysis with immobilised organometallic ruthenium complexes: Substrate- and regioselective transfer hydrogenation of ketones

Article abstract of DOI:10.1002/1521-3765(20001215)6:24<4604::AID-CHEM4604>3.0.CO;2-Y

Chloro-(η⁶-arene) complexes of ruthenium(II) with N-sulfonyl-1,2-ethylenediamine ligands that have one or two styrene side chains have been synthesised and characterised. The chloro ligand was substituted with a diphenylphosphinato ligand and the resulting organometallic complexes are transition state analogues for the ruthenium-catalysed transfer hydrogenation of benzophenone. Following the protocol of molecular imprinting, these complexes were copolymerised with ethylene glycol dimethacrylate (EGDMA) in the presence of a porogen. The polymers were ground and sieved, and the phosphinato ligand was substituted with a chloro ligand, thus generating a shape-selective cavity in close proximity to the catalytically active metal centre. When tested for their ability to catalyse the reduction of benzophenone, the imprinted polymers showed a significantly higher activity (up to a factor of seven) than control polymers without cavities. Out of a mixture of seven different aromatic and aliphatic ketones, benzophenone was preferentially reduced when the imprinted polymer was used. Furthermore, the specificity of the catalyst for diaryl ketones has been confirmed in a reaction with a bifunctional substrate, 4-acetyl-benzophenone; the diaryl ketone was reduced faster with the imprinted catalyst than the acetyl group. The opposite regioselectivity was observed with the control polymer. Both the activity and the selectivity of the imprinted catalysts are dependent on how the ruthenium complexes are attached to the polymer backbone. A double connection proved to give superior results.

Full text of DOI:10.1002/1521-3765(20001215)6:24<4604::AID-CHEM4604>3.0.CO;2-Y

FULL PAPER
Biomimetic Catalysis with Immobilised Organometallic Ruthenium Complexes:
Substrate- and Regioselective Transfer Hydrogenation of Ketones
Kurt Polborn and Kay Severin*[a]
Abstract: Chloro-ꢀh6-arene) complexes
of rutheniumꢀii) with N-sulfonyl-1,2-eth-
ylenediamine ligands that have one or
two styrene side chains have been syn-
thesised and characterised. The chloro
ligand was substituted with a diphenyl-
phosphinato ligand and the resulting
organometallic complexes are transition
state analogues for the ruthenium-cata-
lysed transfer hydrogenation of benzo-
phenone. Following the protocol of
molecular imprinting, these complexes
were copolymerised with ethylene gly-
col dimethacrylate ꢀEGDMA) in the
presence of a porogen. The polymers
were ground and sieved, and the phos-
phinato ligand was substituted with a
chloro ligand, thus generating a shape-
selective cavity in close proximity to the
catalytically active metal centre. When
tested for their ability to catalyse the
reduction of benzophenone, the imprint-
ed polymers showed a significantly high-
er activity ꢀup to a factor of seven) than
control polymers without cavities. Out
of a mixture of seven different aromatic
and aliphatic ketones, benzophenone
was preferentially reduced when the
imprinted polymer was used. Further-
more, the specificity of the catalyst for
diaryl ketones has been confirmed in a
reaction with a bifunctional substrate,
4-acetyl-benzophenone; the diaryl ke-
tone was reduced faster with the im-
printed catalyst than the acetyl group.
The opposite regioselectivity was ob-
served with the control polymer. Both
the activity and the selectivity of the
imprinted catalysts are dependent on
how the ruthenium complexes are at-
tached to the polymer backbone. A
double connection proved to give supe-
rior results.
Keywords: catalysts
´ hydrogena-
tions ´ molecular imprinting ´ ruthe-
nium ´ transition states
technology for the introduction of specific recognition sites in
highly crosslinked synthetic polymers.[4] Starting with the
pioneering work of Wulff, MI has emerged as a powerful and
simple tool to create ªsmartº polymers with most of the
research being focused on analytical applications.[4] The
strategy of how this technique can be employed to control
the microenvironment of transition metal catalysts is outlined
in Scheme 1. A catalytically active metal complex with a
spectator ligand that has one or more polymerisable side
chains is coordinated to a pseudosubstrate. This pseudosub-
strate should be similar to the real substrateꢀs) in terms of size
and functionality. Ideally, the resulting complex should mimic
the transition state of the catalytic transformation. It is
desirable to choose a pseudosubstrate with a donor group that
has a sufficiently high affinity to the metal complex, since the
isolation and characterisation of the transition state analogue
ꢀTSA) may then be possible. Furthermore, it ensures that the
concentration of the relevant template during polymerisation
is high. Following the standard protocol of MI, the TSA is
subsequently co-polymerised with a large excess of a cross-
linking monomer in the presence of a porogen. Selective
removal of the pseudosubstrate from the resulting polymer
yields an immobilised catalyst with a specific cavity in close
proximity to the catalytically active metal centre.
Introduction
The activity and selectivity of metalloenzymes is determined
decisively by ligands of the second coordination sphere, that
is, by the microenvironment of the catalytically active centre.
The implementation of a structurally defined microenviron-
ment in synthetic organometallic catalysts would be beneficial
for several applications. Due to the specific molecular
recognition between substrate and catalyst, an enhanced
regio, stereo, and substrate selectivity should be possible.
Two major approaches have been attempted in order to
reach this goal. First, large and complex ꢀsupra)molecular
systems have been synthesised, with transition metal catalysts
playing an integral role.[1, 2] These systems are often based on
concave organic host molecules with known binding proper-
ties, such as cyclodextrins.[2] Secondly, transition metal cata-
lysts have been incorporated into porous macromolecular
frameworks, such as zeolites, the pore size of which modulates
the substrate selectivity.[3] A third and much less explored
method is based on molecular imprinting ꢀMI). MI is a
[a] Dr. K. Severin, Dr. K. Polborn
Institut für Anorganische Chemie der Universität, Butenandtstr. 5 ± 13
81377 München ꢀGermany)
The strategy outlined above has been used by Mosbach
et al. to prepare an aldolase-mimicing synthetic polymer[5] and
Fax : ꢀ‡49)89-2180-7866
E-mail: kse@cup.uni-muenchen.de
4604
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4604 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 24

4604±4611
In a preliminary communication, we reported the synthesis
and molecular structure of an organometallic TSA ꢀ7) for the
transfer hydrogenation of benzophenone.[10] To best of our
knowledge, this ꢀh6-arene)RuII complex, with a diphenylphos-
phinato ligand that acts as a pseudosubstrate, represents the
first isolated and structurally characterised organometallic
TSA. The complex was incorporated into a synthetic organic
polymer following the imprinting protocol. The resulting
catalyst was shown to be three times more active than a
control polymer prepared with a complex without pseudo-
substrate ꢀ4) and was specific for benzophenone.[10] A similar
methodology was used to generate an immobilised chiral
rhodiumꢀiii) complex, which catalyses the asymmetric reduc-
tion of aromatic ketones with high enantiomeric excess.[11] We
herein report on the synthesis and spectroscopic properties of
the above-mentioned ruthenium complexes and the corre-
sponding ligands. Furthermore, second-generation catalysts
are described that have two styrene side chains ꢀ6, 8). Even
higher rate enhancements ꢀfactor of seven) were observed
with these catalysts. To demonstrate the specificity of the
imprinted catalyst for benzophenone, a competition experi-
ment with six different cosubstrates was performed. The
regioselectivity of the catalysts was addressed in a reaction
with the bifunctional substrate, 4-acetylbenzophenone.
Scheme 1. Strategy to generate an immobilised catalyst with a specific
cavity in close proximity to the active center: a) coordination of the catalyst
to
a ligand which mimics the substrate; b) polymerisation of the
pseudosubstrate/catalyst complex in the presence of a porogen and a large
amount of cross-linker; c) selective removal of the pseudo-substrate.
by Lemaire et al. to produce a catalytically active polymer for
the asymmetric reduction of ketones.[6] Both groups utilised
classical coordination compounds, a CoII pyridine complex or
a RhI amine complex, respectively, which were prepared in
situ prior to polymerisation and were not further character-
ised. Preliminary studies about molecular imprinting with
Results and Discussion
Synthesis of the ruthenium complexes: The synthesis of the
ligands employed in this study is outlined in Scheme 3.
Treatment of ethylene diamine with 4-vinyl-benzenesulfonyl
titanium catalysts have been presented by the group of
[7]
Â
Gagne.
The transfer hydrogenation of ketones catalysed by half-
sandwich complexes of rutheniumꢀii) has recently received
much attention, because highly active and enantioselective
catalysts can be obtained when certain amine-based ligands
are used.[8] For reactions of this kind, a six-membered cyclic
transition structure, A, was proposed by Noyori ꢀScheme 2).[8b]
Computational studies have recently provided additional
Scheme 3. a) H2NCH2CH2NH2, CH2Cl2; b) 4-vinylbenzaldehyde, 4 Š,
CH2Cl2; c) NaBH4, MeOH; d) HCl/Et2O.
chloride gives monosubstituted ligand 1. The introduction of a
second styrene group is achieved by condensation of 1 with
4-vinylbenzaldehyde. Since the presence of NH moieties was
shown to be crucial for successful catalytic transformations,[8b]
imine 2 was subsequently reduced with NaBH4 to give
disubstituted ligand 3.
Scheme 2. A: the proposed transition structure for the transfer hydro-
genation of aromatic ketones catalysed by ꢀh6-arene)RuII complexes; B:
phosphinato complexes as transition state analogues. R ˆ alkyl, aryl.
evidence for such a geometry.[9] We suggested phosphinato
complexes of the general formula B to imitate this structure
with a pseudosubstrate ruthenium complex.
A racemic mixture of ruthenium complexes 4 and 5 was
prepared by treating [{ꢀh6-arene)RuCl2}2] ꢀarene ˆ p-cymene,
Chem. Eur. J. 2000, 6, No. 24
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4605 $ 17.50+.50/0
4605

K. Severin, K. Polborn
FULL PAPER
C6Me6) with the sodium salt of
amine ligand 1. Both complexes
are chiral, with the ruthenium
atom as the stereogenic centre
and a ligand that acts as an
anionic N,N'-chelate.
Similarly to 4 and 5, the ruthenium complex 6 was obtained
by reaction of [{ꢀp-cymene)RuCl2}2] with the anion of amine
ligand 3. The complex has two stereogenic centres, the amine
1
The H NMR spectrum of 4
in CDCl3 at room temperature
shows a broad signal at d ˆ 1.2
for the diasterotopic methyl
groups of the cymene ligand and a series of broad signals
between d ˆ 5 and 6 for the diasterotopic CH protons of the
cymene ligand. These signals sharpen to some extent upon
heating the sample to 558C. The underlying dynamic process
is the epimerisation of the chiral metal centre. When the
spectrum is recorded in a mixture of CDCl3 and CD3OD, this
process is fast relative to the NMR timescale and a well-
resolved spectrum is obtained. This is in accordance with
other studies of chiral ruthenium half-sandwich complexes,
which have shown that the rate of epimerisation strongly
depends on the polarity of the solvent.[12] Interestingly,
compounds 4 and 5 are both soluble in water. It is most likely
that the chloro ligand in this medium is replaced by a water
molecule to form a cationic solvato complex as was suggested
for other RuII half-sandwich complexes.[13] Unfortunately,
complex 5 has a poor solubility in organic solvents, such as
dichloromethane and chloroform, and therefore was not used
for molecular imprinting studies.
nitrogen atom and the metal atom, and the formation of two
diastereoisomers was thus expected. The H NMR spectrum
1
of 6 ꢀCDCl3, 558C), however, shows only one set of signals.
Below that temperature, broad signals point to dynamic
processes. These results suggest that epimerisation of one of
the stereogenic centres is fast on the NMR timescale, which is
in agreement with the results obtained for 8 ꢀsee discussion
below).
The structure of one enantiomer of 5 in the crystal state is
shown in Figure 1. The geometry around the ruthenium atom
can be described as pseudo-octahedral and the h6-arene ring
The chloro ligand in 4 and 6 can be substituted with a
diphenylphosphinato ligand ꢀthe pseudosubstrate) with the
use of the silver salt of diphenylphosphinic acid. No reaction
was observed with the sodium salt. In perfect agreement with
the postulated transition structure ꢀScheme 2), the phos-
phinato ligand in 7 acts as a monodentate ligand with an
additional hydrogen bond to the adjacent amine group. This
was established by the result of a single-crystal analysis of 7.[10]
A similar geometry is assumed for complex 8. As for
compound 6, only one set of signals is observed for complex
8 although two stereogenic centres are present ꢀCDCl3,
238C). At temperatures below À508C, however, two isomers
can be detected by 31P NMR spectroscopy, the relative ratio of
which is 8:1 ꢀFigure 2). These results show that the formation
of 8 is highly diastereoselective and that epimerisation
processes occur within the NMR timescale.
Figure 1. Molecular structure of 5 in the crystal. Selected bond lengths
À
À
À
[pm] and angles [8]: Ru1 N1 212.7ꢀ6), Ru1 N2 213.1ꢀ8), Ru1 Cl1 241.7ꢀ2),
À
S1 N2 157.7ꢀ7); N1-Ru1-N2 78.0ꢀ3), N1-Ru1-Cl1 83.6ꢀ2), N2-Ru1-Cl1
89.3ꢀ2).
and other ligands form a ªpiano stoolº configuration. Due to
the restraints of the N,N'-chelate, the N1-Ru-N2 angle ꢀ78.08)
is smaller than the N1-Ru-Cl ꢀ83.68) and the N2-Ru-Cl angles
ꢀ89.38). The Ru ± Cl and Ru ± N bond lengths are comparable
to those found in related ruthenium complexes.[14]
Preparation of the polymeric catalysts: The complexes 4 and
6 ± 8 were copolymerised with ethylene glycol dimethacrylate
ꢀRu:EGDMA ˆ 1:99) in the presence of CHCl3, a solvent
which acts as a porogen and helps to dissolve the ruthenium
complexes. The combination of EGDMA and CHCl3 has
4606
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4606 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 24

Molecular Imprinting
4604±4611
Table 1. Elemental analyses of the polymers.
Polymer
C [%] H [%] Ru [%][a] P[%][a]
P4
P6
P7
P8
59.60 7.23
59.78 7.13
59.59 7.15
59.73 7.13
0.41
0.41
0.37
0.37
0.37
±
±
ꢀ 0.02
ꢀ 0.02
0.11
EGDMA ´ 0.17H2O ´ 0.77 Â 10À2 8 ꢀcalcd) 59.75 7.14
[a] Determined with ICE-AES.
Figure 2. 31P NMR spectrum of 8 at 23 and À558C ꢀCDCl3).
sition ꢀTable 1). The only difference is the microenvironment
of the ruthenium complexes. The same is true for P7 and P4.
previously been shown to be well suited for imprinting
studies.[15] To initiate the reaction, 2,2'-azobisꢀ4-methoxy-2,4-
dimethylvaleronitrile) ꢀV-70) was employed. In contrast to
2,2'-azobisisobutyronitrile ꢀAIBN), the thermal initiator nor-
mally used in molecular imprinting, V-70 allows the polymer-
isation to be carried out under mild conditions ꢀ358C) and
thus the incorporation of sensitive organometallic complexes
is possible.
After polymerisation, the orange polymers were ground
and sieved to obtain uniform particle sizes ꢀ25 ± 100 mm). The
polymers were then treated with a solution of [BnNEt3]Cl in
methanol. During this procedure, the phosphinato ligands are
removed and replaced by chloro ligands, thereby generating a
form-selective cavity ꢀScheme 4).[16] Elemental analyses with
Transfer hydrogenation: Transfer hydrogenations of aromatic
ketones with ꢀh6-arene)RuII complexes that contain N-sul-
fonyl-1,2-ethylenediamine ligands can be performed with
2-propanol or formic acid as the reducing agent.[8] With the
latter hydrogen donor, the reaction is quasi-irreversible
ꢀgeneration of CO2) and 100% conversion can be achieved
in theory. Furthermore, the reactions can be performed in air
without the need for an inert atmosphere. When tested for
their ability to catalyse the reduction of benzophenone with
azeotropic HCO2H/NEt3,[17] all polymers ꢀP4, P6 ± P8) dis-
played a high activity: with 1 mol% catalyst at 708C, initial
turnover frequencies ꢀTOF) of more than 10 hÀ1 were
determined. This indicates good incorporation[18] and acces-
sibility of the ruthenium complexes in the porous organic
matrix. Quantification of the catalytic activity revealed that
the imprinted polymers were significantly more active than
the control polymers: for P7 and P4, a difference in the initial
rates of a factor of three was observed.[10] Under identical
conditions, the difference in activity between P8 and P6 was
even higher ꢀmore than a factor of five, Figure 3). This rate
Figure 3. Product formation as
hydrogenation of benzophenone in the presence of catalyst P8 ꢀ ) or P6
a
function of time for the transfer
Scheme 4. Immobilisation of ruthenium complexes. a) EGDMA, CHCl3,
V-70, 358C, 20 h, then 658C, 4 h; b) [BnNEt3]Cl in MeOH ꢀ0.1m).
&
&
ꢀ ). The data represent averaged values from two independent experi-
ments; the errors are less than 0.4%. a) azeotropic HCO2H/NEt3, CH3CN,
Ru catalyst, 708C ꢀbenzophenone:Ru ˆ 100:1).
ICP-AES have indicated that at least 80% of the template is
released by this procedure ꢀTable 1). Polymers P4 and P6 ±
P8 were then washed extensively with methanol and dried in
vacuo. It is important to note that the imprinted catalyst P8
and the control catalyst P6 have the same chemical compo-
enhancement is remarkable considering the fact that the
imprinted and the control polymer contain the same amount
of ruthenium complexes.
Chem. Eur. J. 2000, 6, No. 24
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4607 $ 17.50+.50/0
4607

K. Severin, K. Polborn
FULL PAPER
The improved relative performance of imprinted catalyst
P8 is most likely the result of the reduced flexibility of the
second-generation ruthenium complexes within the polymeric
matrix. Since they are connected by two styrene side chains to
the polymer backbone, the catalytically active metal centre is
fixed and thus more precisely positioned opposite to the
cavity.
With the new polymers P8 and P6, we also carried out the
reduction of benzophenone with the use of 2-propanol as the
reducing agent. Both the observed rate enhancement ꢀfactor
of 6.5) as well as the absolute rates are slightly higher than
with formic acid.
The form-selective cavity obtained by molecular imprinting
is not only expected to affect the activity, but also the
substrate selectivity of the catalyst. Indeed, competition
experiments with equal amounts of benzophenone and a
second ketone with the polymeric catalysts P7 and P4 have
shown that molecular imprinting with a diphenylphosphinato
ligand enhances the specificity for benzophenone.[10] We
performed a more complicated experiment with the second-
generation catalysts. Equal amounts of seven different
ketones ꢀS1 ± S7), one of which was benzophenone, were
reduced in the presence of imprinted polymer P8 or control
polymer P6. The product distribution after 20 minutes is
shown in Figure 4. With the control catalyst P6 only low
conversion with no significant preference for either substrate
was observed. The imprinted catalyst P8, on the other hand,
was not only significantly more active, but also specific for
benzophenone: diphenylmethanol is clearly the dominant
product in the reaction mixture. The relative increase in yields
for all substrates are quite informative ꢀFigure 4). A factor of
6.6 is observed for benzophenone. The value drops to 3.1 or
3.2 for the aromatic ketones, tetralone and acetophenone,
respectively. This shows that the catalyst can differentiate
between structurally very similar substrates, such as aceto-
phenone and benzophenone. As expected, for reactions with
aliphatic substrates like 2-adamantanone, a relatively low
increase in yield is observed ꢀfactor of 2.1).
We were also interested whether imprinted catalyst P8
would preferentially reduce a diarylketo group within a
multifunctional substrate. To address this question, we
synthesised 4-acetylbenzophenone. This molecule contains
both a diaryl-keto and an acetyl group. The results obtained
with this substrate and the MIP catalyst P8 or the control
catalyst P6 with formic acid as the reducing agent are
summarised in Scheme 5. After one hour, 33% of the
Scheme 5. Transfer hydrogenation of 4-acetylbenzophenone in the pres-
ence of catalyst P8 and P6. a) azeotropic HCO2H/NEt3, CH3CN, 708C.
The product distribution after 1 h is reported ꢀthe values in brackets
correspond to the reaction with catalyst P6).
substrate was reduced when control catalyst P6 was used. The
product distribution reveals that the sterically less crowded
acetyl group was reduced almost twice as fast as the diaryl-
keto group. Again, the MIP catalyst P8 was more active: after
one hour, 74% of 4-acetylbenzophenone had reacted under
identical conditions. More importantly, the selectivity was
reversed; the dominant product was now diarylmethanol
ꢀScheme 5). These results show that the regioselectivity can
also be controlled when a TSA is used in combination with
molecular imprinting.[19] The difference in selectivity in this
experiment would probably have been even higher, if the
correct pseudosubstrate, that is, ꢀ4-acetylphenyl)phenylphos-
phinate, had been used as a template.
Figure 4. Transfer hydrogenation of a mixture of seven different ketones
ꢀS1 ± S7) in the presence of catalyst P8 or P6. Product distribution after
20 min reveals that the MIP catalyst P8 ꢀdark grey) is significantly more
active than the control catalyst P6 ꢀlight grey). Furthermore, P8 shows a
pronounced selectivity for benzophenone. The values for the relative
increase in yields are reported below.
Conclusion
We have shown that structurally defined organometallic
transition state analogues in combination with an imprinting
technique can be used to generate immobilised catalysts with
4608
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4608 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 24

Molecular Imprinting
4604±4611
d ˆ 43.90, 60.04 ꢀCH2), 115.59, 117.37, 126.53, 126.81, 127.48, 128.56, 135.03,
a shape-selective cavity in close proximity to the catalytically
active centre. As a result, these catalysts show a significantly
enhanced activity. Furthermore, they are specific and, thus,
substrate- and regio-selective transformations can be per-
formed. Both the activity and the selectivity are dependent on
how the metal complexes are attached to the polymer
backbone: a rigid connection by two styrene side chains was
shown to be superior.
The manipulation of the microenvironment of organo-
metallic complexes often requires an elaborate molecular
design. In our approach, specific cavities are formed by self-
assembly of small molecules around an organometallic TSA
with subsequent covalent capture.[20] The advantage of this
method is its generality and simplicity. We are currently
applying this technique to other transition metal-catalysed
reactions.
À
ˆ
ˆ
135.45, 136.28, 139.03, 140.29, 141.81 ꢀC6H4 CH CH2), 163.15 ꢀCH N);
elemental analysis calcd ꢀ%) for C19H20N2O2S ꢀ340.44): C 67.03, H 5.92, N
8.23; found C 67.15, H 5.87, N 8.23.
[C8H7CH2NH2CH2CH2NHSO2C8H7]Cl 63): NaBH4 ꢀ50 mg, 1.32 mmol)
was added to a solution of 2 ꢀ400 mg, 1.18 mmol) in methanol ꢀ10 mL).
After stirring for 15 min at room temperature, the solution was heated
under reflux for another 15 min. The solvent was removed in vacuo. The
residue was extracted with diethyl ether ꢀ100 mL), washed three times with
water and dried over Na2SO4. After evaporation of the solvent, the product
was redissolved in diethyl ether ꢀ40 mL ) and precipitated as the HCl
adduct with a few milliliters of HCl-saturated ethyl acetate to give a white
powder ꢀ380 mg). Yield: 61%; m.p. 181 ± 1838C; 1H NMR ꢀ270 MHz,
CDCl3): d ˆ 2.65 ± 2.69 ꢀm, 2H; CH2), 2.98 ± 3.02 ꢀm, 2H; CH2), 3.62 ꢀs, 2H;
3
CH2), 5.23 ꢀdd, 2J ˆ 0.8 Hz, J ˆ 10.9 Hz, 1H; CH CH2), 5.41 ꢀd, 3J ˆ
ˆ
2
3
ˆ
ˆ
10.9 Hz, 1H; CH CH2), 5.72 ꢀd, J ˆ 0.8 Hz, J ˆ 17.4 Hz, 1H; CH CH2),
3
ˆ
ˆ
5.85 ꢀd, J ˆ 17.6 Hz, 1H; CH CH2), 6.64 ± 6.77 ꢀm, 2H; CH CH2), 7.15 ꢀd,
3J ˆ 8.2 Hz, 2H; C6H4), 7.32 ꢀd, 3J ˆ 8.3 Hz, 2H; C6H4), 7.47 ꢀd, 3J ˆ 8.6 Hz,
2H; C6H4), 7.79 ꢀd, J ˆ 8.4 Hz, 2H; C6H4); 13C NMR ꢀ101 MHz, CDCl3):
3
d ˆ 42.51, 47.51, 52.93 ꢀCH2), 113.76, 117.38, 126.39, 126.81, 127.47, 128.34,
À
ˆ
135.45, 136.57, 138.72, 139.45, 141.77 ꢀC6H4 CH CH2); elemental analysis
calcd ꢀ%) for C19H23ClN2O2S ꢀ378.91): C 60.23, H 6.12, N 7.39; found C
59.21, H 6.10, N 7.15.
Experimental Section
[6p-Cymene)Ru{NH2CH2CH2NSO2C8H7}Cl] 64): A solution of NaOMe
ꢀ1.0 mmol) in methanol ꢀ2m) was added to a solution of 1 ꢀ226 mg,
1.0 mmol) in dichloromethane/methanol ꢀ1:1; 15 mL). The resulting
solution was slowly added ꢀ15 min) to a solution of [{ꢀp-cymene)RuCl2}2]
ꢀ306 mg, 0.5 mmol) in dichloromethane ꢀ20 mL). After 3 h, the solvent was
removed in vacuo. The residue was stirred with dichloromethane/diethyl
ether ꢀ2:1; 30 mL). After filtration, the solution was reduced to a volume of
2 mL in vacuo. Addition of hexane ꢀ15 mL), evaporation of the solvent and
drying in vacuo yielded an orange powder ꢀ457 mg). Yield: 89%; m.p. 192 ±
1948C; 1H NMR ꢀ400 MHz, CDCl3/CD3OD, 4:1, 558C): d ˆ 1.11 ꢀd, 3J ˆ
7 Hz, 6H; CHꢀCH3)2), 2.01 ꢀs, 3H; CH3, cymene), 2.10 ꢀbrm, 4H; NCH2),
General: The synthesis of all ruthenium complexes as well as the transfer
hydrogenation with 2-propanol was performed under an atmosphere of dry
dinitrogen and by using standard Schlenk techniques. Solvents were freshly
distilled over an appropriate drying agent and stored under dinitrogen prior
to usage. EGDMA was washed with NaOH ꢀ1m) and saturated NaCl
solution and dried with Na2SO4. After filtration, the monomer was distilled
under reduced pressure. The azo initiator V-70 was a gift from Wako
Chemicals. 4-Acetylbenzophenone was prepared by oxidation of commer-
cially available 4-ethylbenzophenone with KMnO4/CuSO4 ´ ꢀH2O)n
ꢀCH2Cl2, 350 h) and recrystallised from hot hexane. 4-Vinylbenzene-
sulfonyl chloride,[21] 4-vinylbenzaldehyde,[22] [{ꢀp-cymene)RuCl2}2][23] and
[{ꢀC6Me6)RuCl2}2][23] were prepared according to literature procedures. The
1H; 13C and 31P NMR spectra were recorded on a JEOL EX 400 or a
GSX 270 spectrometer. All spectra were recorded at room temperature
unless stated otherwise. The GC analysis was performed with a Varian 3800
spectrometer with a CP-Cyclodextrin-B-2,3,6-M-19 column ꢀ50 m).
3
3
ˆ
2.68 ꢀsept, J ˆ 7 Hz, 1H; CHꢀCH3)2), 5.15 ꢀd, J ˆ 11 Hz, 1H; CH CH2),
5.33 ꢀbrd, 3J ˆ 5 Hz, 2H; CH, cymene), 5.48 ꢀd, 3J ˆ 5 Hz, 2H; CH,
3
3
ˆ
cymene), 5.64 ꢀd, J ˆ 18 Hz, 1H; CH CH2), 6.55 ꢀdd, J ˆ 11, 18 Hz, 1H;
CH CH2), 7.26 ꢀd, 3J ˆ 8 Hz, 2H; C6H4), 7.67 ꢀd, 3J ˆ 8 Hz, 2H; C6H4);
ˆ
13C NMR ꢀ101 MHz, CDCl3/CD3OD, 4:1, 508C): d ˆ 18.31, 22.17 ꢀCH3,
cymene), 30.54 ꢀCH, cymene), 46.86, 48.52 ꢀNCH2), 81.50 ꢀbr, CH,
cymene), 96.40, 102.20 ꢀC, cymene), 115.31, 125.82, 127.49, 135.99, 139.61,
H2NCH2CH2NHSO2C8H7 61): A solution of 4-vinylbenzenesulfonyl chlor-
ide ꢀ5 mL) in dichloromethane ꢀ15 mL) was slowly added ꢀ1 h) to a solution
of ethane-1,2-diamine ꢀ5 mL) in dichloromethane ꢀ10 mL). The yellow
suspension was stirred for 2 h, then dichloromethane ꢀ200 mL), H2O
ꢀ100 mL) and HCl/H2O ꢀ2m, 200 mL) were added. After separation of the
layers, the water phase was washed twice with dichloromethane, filtered
and then neutralised with KOH ꢀ2m, pH 9 ± 10). The mixture was extracted
with dichloromethane ꢀ3 Â 150 mL). After drying over Na2SO4, filtration
and evaporation of the solvent, a slightly yellow powder was obtained
ꢀ3.8 g). Yield: 68%; m.p. 127 ± 1288C; 1H NMR ꢀ270 MHz, CDCl3): d ˆ
À
ˆ
142.17 ꢀC6H4 CH CH2); elemental analysis calcd ꢀ%) for C20H27ClN2O4-
RuS ´ H2O ꢀ514.04): C 46.73, H 5.69, N 5.45; found C 47.03, H 5.36, N 5.32.
[6C6Me6)Ru{NH2CH2CH2NSO2C8H7}Cl] 65): The synthesis of
5 was
performed analogous to that of 4. Dichloromethane/diethyl ether ꢀ6:1,
90 mL) was used to extract the product. Yield: 78%; orange powder; m.p.
159 ± 1628C; 1H NMR ꢀ400 MHz, CDCl3/CD3OD, 4:1): d ˆ 1.96 ꢀs, 6H;
CH3), 1.99 ꢀt, 3J ˆ 5 Hz, 2H; NCH2), 2.25 ꢀt, 3J ˆ 5 Hz, 2H; NCH2), 5.09 ꢀd,
3
3
3
ˆ
ˆ
J ˆ 11 Hz, 1H; CH CH2), 5.58 ꢀd, J ˆ 18 Hz, 1H; CH CH2), 6.49 ꢀdd,
J ˆ 11, 18 Hz, 1H; CH CH2), 7.17 ꢀd, 3J ˆ 8 Hz, 2H; C6H4), 7.68 ꢀd, 3J ˆ
ˆ
8 Hz, 2H; C6H4); 13C NMR ꢀ101 MHz, CDCl3/CD3OD, 4:1): d ˆ 15.72
ꢀCH3), 44.86, 49.38 ꢀNCH2), 91.10 ꢀC6Me6) 115.20, 125.64, 128.07, 135.89,
2.17 ꢀbrm, 2H; NH2), 2.80 ꢀm, 2H; CH2), 2.97 ꢀm, 2H; CH2), 5.42 ꢀd, 3J ˆ
3
11 Hz, 1H;; CH CH2), 5.87 ꢀd, J ˆ 17 Hz, 1H; CH CH2), 6.75 ꢀdd, 3J ˆ
ˆ
ˆ
3
3
À
ˆ
ˆ
139.31, 142.56 ꢀC6H4 CH CH2); elemental analysis calcd ꢀ%) for
C22H31ClN2O4RuS ´ 0.25CH2Cl2 ꢀ545.32): C 49.01, H 5.82, N 5.14; found C
48.95, H 5.99, N 5.05.
11, 17 Hz, 2H; CH CH2), 7.52 ꢀd, J ˆ 8 Hz, 2H; C6H4), 7.81 ꢀd, J ˆ 8 Hz,
2H; C6H4); 13C NMR ꢀ68 MHz, CDCl3): d ˆ 40.89, 45.15 ꢀCH2), 116.30,
À
ˆ
126.52, 127.05, 135.38, 139.31, 141.88 ꢀC6H4 CH CH2); elemental analysis
calcd ꢀ%) for C10H14N2O2S ´ 0.25H2O ꢀ230.80): C 52.04, H 6.33, N 12.14;
found C 52.34, H 6.30, N 12.12.
[6p-Cymene)Ru{NH6CH2C8H7)CH2CH2NSO2C8H7}Cl] 66): A solution of
NaOMe ꢀ2.0 mmol) in methanol ꢀ2m) was added to a suspension of 3
ꢀ378 mg, 1.0 mmol) in dichloromethane ꢀ10 mL). The resulting solution was
slowly added ꢀ15 min) to a solution of [{ꢀp-cymene)RuCl2}2] ꢀ306 mg,
0.5 mmol) in dichloromethane ꢀ10 mL). After 15 h, the mixture was filtered
and the solvent removed in vacuo. The product was purified by flash
chromatography ꢀchloroform/methanol 9:1, silica gel) and crystallised by
slow diffusion of hexane into a solution of 6 in dichloromethane. Yield:
68%; m.p. 203 ± 2058C ꢀdecomp); 1H NMR ꢀ270 MHz, CDCl3): d ˆ 1.30 ꢀd,
3J ˆ 7 Hz, 3H; CHꢀCH3)2), 1.34 ꢀd, 3J ˆ 7 Hz, 3H; CHꢀCH3)2), 2.11 ± 2.45
ꢀm, 3H; CH2), 2.22 ꢀs, 3H; CH3, cymene), 2.94 ± 3.07 ꢀm, 2H; CH2,
CHꢀCH3)2), 3.82 ꢀbrs, NH), 4.13 ꢀdd, 2J ˆ 13 Hz, 3J ˆ 10 Hz, 1H; NCH2Ar),
4.25 ꢀdd, 2J ˆ 13 Hz, 3J ˆ 4 Hz, 1H; NCH2Ar), 5.23 ± 5.82 ꢀm, 8H; CH-
C8H7CH ˆ NCH2CH2NHSO2C8H7 62): A solution of 1 ꢀ1.0 g, 4.42 mmol)
and 4-vinylbenzaldehyde ꢀ584 mL, 4.42 mmol) in dichloromethane ꢀ30 mL)
was stirred at room temperature in the presence of molecular sieves ꢀ4 Š).
After 24 h, the solids were removed by filtration. The solvent was
evaporated in vacuo and the remaining oil was stirred in diethyl ether
ꢀ60 mL) to give a white precipitate ꢀ590 mg). A second fraction of
analytically clean product was obtained by cooling the mother liquor to
À208C ꢀ330 mg). Yield: 61%; m.p. 101 ± 1038C; 1H NMR ꢀ270 MHz,
CDCl3): d ˆ 3.30 ± 3.31 ꢀm, 2H; CH2), 3.64 ꢀt, 3J ˆ 6 Hz, 2H; CH2), 5.30 ꢀm,
3
1H; NH), 5.32 ꢀd, J ˆ 11 Hz, 1H; CH CH2), 5.40 ꢀd, 3J ˆ 11 Hz, 1H;
ˆ
3
CH CH2), 5.81 ꢀd, J ˆ 18 Hz, 1H; CH CH2), 5.84 ꢀd, 3J ˆ 18 Hz, 1H;
ˆ
ˆ
3
3
ˆ
ˆ
ˆ
ˆ
CH CH2), 6.71 ꢀdd, J ˆ 11, 18 Hz, 2H; CH CH2), 7.40 ꢀd, J ˆ 8 Hz, 2H;
cymene, CH CH2), 6.62 ± 6.74 ꢀm, 2H; CH CH2), 7.24 ± 7.43 ꢀm, 6H;
C6H4), 7.78 ꢀd, 3J ˆ 8 Hz, 2H; C6H4); 13C NMR ꢀ68 MHz, CD3OD): d ˆ
18.96, 21.94, 23.03 ꢀCH3), 30.97 ꢀCHꢀCH3)2), 48.25, 54.93, 61.15 ꢀCH2),
C6H4), 7.45 ꢀd, 3J ˆ 8 Hz, 2H; C6H4), 7.57 ꢀd, 3J ˆ 8 Hz, 2H; C6H4), 7.79 ꢀd,
J ˆ 8 Hz, 2H; C6H4), 8.12 ꢀs, 1H; CH N); 13C NMR ꢀ101 MHz, CDCl3):
3
ˆ
Chem. Eur. J. 2000, 6, No. 24
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4609 $ 17.50+.50/0
4609

K. Severin, K. Polborn
FULL PAPER
79.60, 79.92, 82.67, 83.46 ꢀCH, cymene), 96.49, 102.60 ꢀC, cymene), 114.97,
125.73, 126.96, 128.01, 128.64, 135.31, 136.02, 136.45, 138.24, 139.46, 142.46
ruthenium ꢀÆ0.01%) and phosphorous ꢀÆ0.01%) were measured by
inductively coupled plasma atomic emission spectroscopy ꢀICP-AES) with
a VARIAN Vista ICP-AES. The C/H analysis ꢀÆ0.30%) was carried out
with a ELEMENTAR Vario EL. N2 adsorption isotherms were measured
on a Sorptomatic 1800 ꢀCarlo Erba), which was controlled by the Mile-
stone 200 program. For the polymers, a BETꢀBET ˆ Brunauer± Emmet ±
À
ˆ
ꢀC6H4 CH CH2); elemental analysis calcd ꢀ%) for C29H35ClN2O2RuS
ꢀ612.19): C 56.90, H 5.76, N 4.58; found C 56.55, H 5.77, N 4.37.
[6p-Cymene)Ru{NH2CH2CH2NSO2C8H7}6O2PPh2)] 67): A suspension of 4
ꢀ103 mg, 0.20 mmol) and [AgO2PPh2] ꢀ72 mg, 0.22 mmol) in dichloro-
methane ꢀ10 mL) was stirred for 72 h in the dark. After filtration, the
solvent was reduced to 2 mL in vacuo. Addition of hexane ꢀ15 mL),
evaporation of the solvent and drying in vacuo gave 128 mg of an orange
powder. Yield: 91%; m.p. 190 ± 1928C ꢀdecomp); 1H NMR ꢀ270 MHz,
2
Teller) inner surface area of 89 m gÀ1 was calculated from the experimental
data in the range of relative pressures of 0.05 ꢀ p/p0 ꢀ 0.35 ꢀp0 ˆ N2
saturation pressure at liquid N2 temperature). Prior to the measurements,
samples were evacuated at 808C for 4 h. The apparent dry density of the
polymers was determined to be 0.74 gmLÀ1. The swelling, given as the
volume of swollen polymer per volume of dry polymer, was determined to
be 1.64 in acetonitrile.
3
CDCl3/CD3OD, 4:1, 558C): d ˆ 1.00 ꢀd, J ˆ 7 Hz, 6H; CHꢀCH3)2), 1.67 ꢀs,
3H; CH3, cymene), 2.07 ± 2.09 ꢀm, 2H; NCH2), 2.26 ꢀsept, 3J ˆ 7 Hz,
CHꢀCH3)2), 2.29 ± 2.36 ꢀm, 2H; NCH2), 5.20 ꢀd, 3J ˆ 11 Hz, 1H; CH ˆ
CH2), 5.29 ꢀbrm, 2H; CH, cymene), 5.39 ± 5.41 ꢀbrm, 2H; CH, cymene),
Transfer hydrogenation of benzophenone
3
3
ˆ
ˆ
5.66 ꢀd, J ˆ 18 Hz, 1H; CH CH2), 6.60 ꢀdd, J ˆ 11, 18 Hz, 1H; CH CH2),
Method A: A suspension of the polymer ꢀ20.2 mg, 1 mmol Ru) in a mixture
of azeotropic HCO2H/NEt3 ꢀ500 mL) and CH3CN ꢀ400 mL) was stirred for
30 min at 708C ꢀto calculate the amount of catalyst, a quantitative
incorporation of the complexes into the polymer was assumed). The
reaction was started by addition of benzophenone in CH3CN ꢀ100 mL, 1m).
A sample ꢀ100 mL) was removed every five minutes from the reaction
vessel, quenched with CH3CN/CH3CO2H ꢀ3:1, 400 mL), filtered and
analysed by capillary GC. Reactions with the imprinted polymer P7 ꢀP8)
and the control polymer P4 ꢀP6) were performed simultaneously.
7.13 ± 7.42 ꢀm, 12H; C6H4, C6H5), 7.73 ꢀd, J ˆ 6 Hz, 2H; C6H4); 13C NMR
3
ꢀ68 MHz, CDCl3/CD3OD, 4:1, 558C): d ˆ 17.57, 21.90 ꢀCH3, cymene), 30.16
ꢀCH, cymene), 45.77, 48.96 ꢀNCH2), 79.81, 81.38 ꢀCH, cymene), 95.14,
100.47 ꢀC, cymene), 115.44, 125.99, 127.09, 127.71, 127.90, 130.33, 130.92,
131.06, 135.89, 136.69, 138.66, 139.85, 143.40 ꢀC6H4 CH CH2, Ph); 31P
NMR ꢀ109 MHz, CDCl3/CD3OD, 4:1, 109 MHz): d ˆ 30.17; elemental
analysis calcd ꢀ%) for C32H37N2O4PRuS ´ 0.33CH2Cl2 ꢀ706.07): C 55.00, H
5.38, N 3.97; found C 55.37, H 5.74, N 4.02.
À
ˆ
Method B: A suspension of the polymer ꢀ20.2 mg, 1 mmol Ru) in a solution
of KOH ꢀ1.2 mmol) in 2-propanol ꢀ950 mL) was stirred for 15 min at 708C
under an atmosphere of dinitrogen. The reaction was started by addition of
benzophenone in 2-propanol ꢀ50 mL, 2M) and analysed as described above.
[6p-Cymene)Ru{NH6CH2C8H7)CH2CH2NSO2C8H7}6O2PPh2)] 68):
T
he
synthesis was performed analogous to that of 7. Yield: 87%; orange
powder; m.p. 146 ± 1508C ꢀdecomp); 1H NMR ꢀ270 MHz, CDCl3): d ˆ 1.18
ꢀd, 3J ˆ 7 Hz, 3H; CHꢀCH3)2), 1.24 ꢀd, 3J ˆ 7 Hz, 3H; CHꢀCH3)2), 1.27 ꢀm,
2H; CH2), 1.78 ꢀs, 3H; CH3, cymene), 1.81 ± 1.99 ꢀm, 1H; CH2), 2.09 ± 2.14
ꢀm, 1H; CH2), 2.56 ꢀm, 1H; CHꢀCH3)2), 2.96 ± 3.02 ꢀm, 1H; CH2), 3.96 ±
Transfer hydrogenation of ketones S1 ± S7: A suspension of the polymer
ꢀ20.2 mg, 1 mmol Ru) in a mixture of azeotropic HCO2H/NEt3 ꢀ500 mL) and
CH3CN ꢀ400 mL) was stirred for 30 min at 708C. The reaction was started
by addition of S1 ± S7 in CH3CN ꢀ100 mL, 0.2m each). After 20 min, a
sample ꢀ200 mL) was removed from the reaction vessel, quenched with
CH3CN/CH3CO2H ꢀ2:1, 300 mL), filtered and analysed by capillary GC.
ˆ
4.05 ꢀm, 1H; CH2), 5.24 ± 5.92 ꢀm, 8H; CH-cymene, CH CH2), 6.63 ± 6.78
ꢀm, 2H; CH CH2), 7.20 ± 7.57 ꢀm, 16H; C6H4), 7.87ꢀd, 3J ˆ 8 Hz, 2H;
ˆ
C6H4); 13C NMR ꢀ68 MHz, CDCl3): d ˆ 19.01, 22.09, 23.32 ꢀCH3), 30.69
ꢀCHꢀCH3)2), 48.44, 53.73, 61.49 ꢀCH2), 80.38, 80.72, 81.03, 81.42 ꢀCH,
cymene), 96.77, 99.09 ꢀC, cymene), 114.86 ± 143.86 ꢀPC6H5,
Transfer hydrogenation of 4-acetylbenzophenone: A suspension of the
polymer ꢀ20.2 mg, 1 mmol Ru) in a mixture of azeotropic HCO2H/NEt3
ꢀ500 mL) and CH3CN ꢀ500 mL) was stirred for 30 min at 708C. The reaction
was started by addition of 4-acetylbenzophenone in CH3CN ꢀ100 mL, 1m).
After 1 h, the suspension was diluted with dichloromethane ꢀ10 mL) and
filtered immediately, and the solvent removed in vacuo. The residue was
dissolved in CDCl3 and analysed by 1H NMR. The product distribution was
determined by integration of the signals for the methyl groups.
À
ˆ
C6H4 CH CH2); elemental analysis calcd ꢀ%) for C41H45N2O4PRuS
ꢀ793.92): C 62.03, H 5.71, N 3.53; found C 62.36, H 6.00, N 3.37.
X-ray structure analysis of 5: Crystals of 5 were obtained by slow diffusion
of hexane into
a solution of 5 in dichloromethane. Enraf-Nonius
diffractometer, MoKa radiation, l ˆ 0.71073 Š, crystal size 0.53  0.47
Â0.17 mm, Tˆ 22ꢀ2)8C, orange crystal, monoclinic, space group P21/c,
a ˆ 12.416ꢀ2), b ˆ 13.399ꢀ2), c ˆ 16.611ꢀ3) Š, Vˆ 2761.2ꢀ8) Š3, Z ˆ 4,
1calcd ˆ 1.548 MgmÀ3, m ˆ 1.054 mmÀ1. Data collection: w from 2.45 to
23.97, À14 ꢀ h ꢀ 14, À15 ꢀ k ꢀ 0, À18 ꢀ l ꢀ 0, 4484 reflections collected,
4316 independent reflections ꢀRint ˆ 0.0143), semiempirical absorption
correction from psi-scans, max./min. transmission 0.9992/0.7506, R1 ˆ
0.0598, wR2 ˆ 0.1592 [I >> 2sꢀI)], GOFꢀF 2) ˆ 1.013, largest difference
peak 0.651 eŠÀ3, largest difference hole À1.125 eŠÀ3. The structure was
solved with direct methods ꢀSHELXS-86). A riding model was employed
for the hydrogen atoms. The vinyl groups as well as one methylene group
are disordered and split positions with restraints were utilised. Restraints
were also employed for the solvent molecule. Crystallographic data
ꢀexcluding structure factors) for the structure reported in this paper have
been deposited with the Cambridge Crystallographic Data Centre as
supplementary publication no. CCDC-139602. Copies of the data can be
obtained free of charge on application to CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK ꢀfax: ꢀ‡44)1223-336-033; e-mail: deposit@
ccdc.cam.ac.uk).
Acknowledgements
K.S. thanks Prof. Dr. W. Beck ꢀLMU München) for his generous support,
Prof. Dr. H. Knözinger and F. Anderle for their help with N2 adsorption
measurements and E. Karaghiosoff and R. Fetouaki for their technical
assistance. Financial funding from the Bayerischer Habilitations-Förder-
preis ꢀK.S.) is gratefully acknowledged.
[1] a) A. J. Kirby, Angew. Chem. 1996, 108, 770 ± 790; Angew. Chem. Int.
Ed. 1996, 35, 706 ± 724; b) Y. Murakami, J. Kikuchi, Y. Hisaeda, O.
Hayashida, Chem. Rev. 1996, 96, 721 ± 758; c) M. C. Feiters in
Comprehensive Supramolecular Chemistry, Vol. 10 ꢀEds.: J.-M. Lehn,
D. N. Reinhoudt, J. C. Atwood, J. E. Davies, D. D. Macnicol, F.
Vögtle), pp. 267 ± 360.
[2] a) M. T. Reetz, S. R. Waldvogel, Angew. Chem. 1997, 109, 870 ± 873;
Angew. Chem. Int. Ed. Engl. 1997, 36, 865 ± 867; b) M. T. Reetz, C.
Frömbgen, Synthesis, 1999, 9, 1555 ± 1557; c) R. Breslow, S. D. Dong,
Chem. Rev. 1998, 98, 1997 ± 2011.
[3] a) J. M. Thomas, Angew. Chem. 1999, 111, 3800 ± 3843; Angew. Chem.
Int. Ed. 1999, 38, 3588 ± 3628; b) E. J. Creyghton, R. S. Downing, J.
Mol. Cat. A 1998, 134, 47 ± 61.
[4] a) B. Sellergren, Angew. Chem. 2000, 112, 1071 ± 1078; Angew. Chem.
Int. Ed. 2000, 39, 1031 ± 1837; b) K. Haupt, K. Mosbach, Trends
Biotech. 1998, 16, 468 ± 475; c) G. Wulff, Angew. Chem. 1995, 107,
1958 ± 1979; Angew. Chem. Int. Ed. Engl. 1995, 34, 1812 ± 1832;
d) M. E. Davies, A. Katz, W. R. Ahmad, Chem. Mater. 1996, 8, 1820 ±
Preparation of the polymers: A freshly prepared mixture of V-70 ꢀ15 mg)
and EGDMA ꢀ934 mL, 4.95 mmol) was added to a solution of 4 ꢀ25 mg,
50 mmol) in CHCl3 ꢀ400 mL) in a 4 mL screw-cap vial. After sonication
ꢀ1 min) and degassing with dinitrogen ꢀ2 min), the vial was closed and
tempered for 20 h at 358C, and then heated for 4 h at 658C. The resulting
polymer was ground, wet-sieved ꢀparticle size: 25 ± 100 mm), treated with a
solution of [BnNEt3]Cl in MeOH ꢀ40 mL, 0.1m, 30 min), washed with
MeOH ꢀ4 Â 40 mL) and dried in vacuo. Polymers containing complex 6, 7
or 8 were prepared in an analogous fashion.
Characterisation of the polymers: To determine the content of ruthenium
and phosphorous in the polymers, the polymers ꢀ10 ± 40 mg) were dissolved
in a mixture of HNO3 ꢀconc., 0.5 mL) and HCl ꢀconc., 0.5 mL) by heating
to 1208C for 4 h. After dilution with H2O/HNO3 ꢀ4%), the amounts of
4610
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4610 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 24

Molecular Imprinting
4604±4611
[15] a) J. Matsui, Y. Miyoshi, O. Doblhoff-Dier, T. Takeushi, Anal. Chem.
1995, 67, 4404 ± 4408; b) B. Sellergren, K. J. Shea, J. Chromatogr. 1993,
635, 31 ± 49; c) G. Vlatakis, L. I. Andersson, R. Müller, K. Mosbach,
Nature 1993, 361, 645 ± 647.
[16] Addition of [BnNEt3]Cl to a solution of 7 or 8 in methanol leads to a
quantitative exchange of the phosphinato ligand. This was established
by in situ NMR experiments ꢀ1H, 31P).
[17] K. Narita, M. Sekiya, Chem. Pharm. Bull. 1977, 25, 135 ± 140.
[18] According to ICP-AES measurements ꢀTable 1) between 77 and 83%
of the ruthenium complexes have been incorporated.
[19] Imprinted polymers have been used for the regioselective modifica-
tion of steroids: a) C. Alexander, C. R. Smith, M. J. Whitcombe, E. N.
Vulfson, J. Am. Chem. Soc. 1999, 121, 6640 ± 6651; b) S. E. Byström, A.
Börje, B. Akermark, J. Am. Chem. Soc. 1993, 115, 2081 ± 2083.
[20] Strategies to generate catalytic systems with the help of transition
state analogues have been discussed by Sanders: a) P. A. Brady,
J. K. M. Sanders, Chem. Soc. Rev. 1997, 26, 327 ± 336; b) J. K. M.
Sanders, Chem. Eur. J. 1998, 4, 1378 ± 1383.
[21] H. Kamogawa, A. Kanzawa, M. Kadoya, T. Naito, M. Nanasawa, Bull.
Chem. Soc. Jpn. 1983, 56, 762 ± 765.
[22] J. Ren, K. Sakakibara, M. Hirota, Bull. Chem. Soc. Jpn. 1993, 66,
1897 ± 1902.
1839; e) Molecular and Ionic Recognition with Imprinted Polymers
ꢀEds.: R. A. Bartsch, M. Maeda), Oxford University Press, New York,
1998.
[5] J. Matsui, I. A. Nicholls, I. Karube, K. Mosbach, J. Org. Chem. 1996,
61, 5414 ± 5417.
[6] a) P. Gamez, B. Dunjic, C. Pinel, M. Lemaire, Tetrahedron Lett. 1995,
36, 8779 ± 8782; b) F. Locatelli, P. Gamez, M. Lemaire, J. Mol. Catal. A
1998, 135, 89 ± 98.
Â
[7] B. P. Santora, A. O. Larsen, M. R. Gagne, Organometallics 1998, 17,
3138 ± 3140.
[8] a) M. J. Palmer, M. Wills, Tetrahedron: Asymmetry 1999, 10, 2045 ±
2061; b) R. Noyori, S. Hashiguchi, Acc. Chem. Res. 1997, 30, 97 ± 102.
[9] a) D. A. Alonso, P. Brandt, S. J. M. Nordin, P. G. Andersson, J. Am.
Chem. Soc. 1999, 121, 9580 ± 9588; b) M. Yamakawa, H. Ito, R.
Noyori, J. Am. Chem. Soc. 2000, 122, 1466 ± 1478.
[10] K. Polborn, K. Severin, Chem. Commun. 1999, 2481 ± 2482.
[11] K. Polborn, K. Severin, Eur. J. Inorg. Chem. 2000, 1687 ± 1692.
[12] a) H. Brunner, Angew. Chem. 1999, 111, 1248 ± 1263; Angew. Chem.
Int. Ed. 1999, 38, 1194 ± 1208; b) D. L. Davies, J. Fawcett, R. Krafczyk,
D. R. Russell, J. Organomet. Chem. 1997, 545/546, 581 ± 585.
[13] a) G. Capper, L. C. Carter, D. L. Davies, J. Fawcett, D. R. Russell, J.
Chem. Soc. Dalton Trans. 1996, 1399 ± 1403; b) L. Carter, D. L. Davies,
J. Fawcett, D. R. Russell, Polyhedron 1993, 12, 1599 ± 1602; c) D. F.
Dersnah, M. C. Baird, J. Organomet. Chem. 1977, 127, C55 ± C58.
[14] a) K.-J. Haak, S. Hashiguchi, A. Fujii, T. Ikariya, R. Noyori, Angew.
Chem. 1997, 109, 297 ± 300; Angew. Chem. Int. Ed. Engl. 1997, 36, 285 ±
288; b) S. Hashiguchi, A. Fujii, J. Takehara, T. Ikariya, R. Noyori, J.
Am. Chem. Soc. 1995, 117, 7562 ± 7563.
[23] M. A. Bennett, T.-N. Huang, T. W. Matheson, A. K. Smith, Inorg.
Synth. 1982, 21, 74 ± 78.
Received: January 27, 2000
Revised version: August 17, 2000 [F2264]
Chem. Eur. J. 2000, 6, No. 24
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0624-4611 $ 17.50+.50/0
4611