Microgel-supported oxazaborolidines: Novel catalysts for enantioselective reductions

Article abstract of DOI:10.1016/S0040-4020(00)00072-7

Microgel-bound oxazaborolidines have been prepared and used as catalysts in the enantioselective reduction of prochiral ketones. The preparation of these soluble, crosslinked polymer molecules was accomplished by solution polymerisation. The approach involved the preparation of microgels bearing free boronic acid functionalities and their subsequent conversion to oxazaborolidines. The selectivities of these supported catalysts are in most cases comparable to those achieved with low molecular weight analogues. The advantage of the application of microgel-bound catalysts is their good solubility together with low viscosity of the solution. Reagents of this type can easily be removed by ultrafiltration and the process can be performed in a continuous mode in a membrane reactor. (C) 2000 Published by Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00072-7

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 1693–1699
Microgel-supported Oxazaborolidines: Novel Catalysts for
Enantioselective Reductions
*
C. Schunicht, A. Biffis and G. Wulff
Institute of Organic Chemistry and Macromolecular Chemistry, Heinrich-Heine University Duesseldorf, D-40225 Duesseldorf, Germany
Received 3 January 2000; accepted 25 January 2000
Abstract—Microgel-bound oxazaborolidines have been prepared and used as catalysts in the enantioselective reduction of prochiral
ketones. The preparation of these soluble, crosslinked polymer molecules was accomplished by solution polymerisation. The approach
involved the preparation of microgels bearing free boronic acid functionalities and their subsequent conversion to oxazaborolidines. The
selectivities of these supported catalysts are in most cases comparable to those achieved with low molecular weight analogues. The advantage
of the application of microgel-bound catalysts is their good solubility together with low viscosity of the solution. Reagents of this type can
easily be removed by ultrafiltration and the process can be performed in a continuous mode in a membrane reactor. ᭧ 2000 Published by
Elsevier Science Ltd. All rights reserved.
Introduction
lidine ring is generally unstable under the workup con-
ditions of the reaction mixture, a practical means of
separating the catalyst before workup has to be devised;
an obvious solution to this problem consists of binding the
oxazaborolidine to a heterogeneous support. In the last few
years extensive investigations have been conducted on the
use of oxazaborolidines bound to insoluble polymers.6,8
Following this approach, some researchers succeeded in
preparing heterogenised CBS catalysts which closely
approached the selectivities obtained with the soluble
analogues.8b,c Unfortunately, this was possible only by
using gel-type polymer supports with a very low degree of
crosslinking (1–2%), for which the mechanical properties
under the reaction conditions are very poor. With increasing
degree of crosslinking, diffusional limitations inside the
polymer support slow down the rate of the catalysed reduc-
tion so that the direct, non-selective borane reduction
becomes competitive, causing a marked decrease in the
overall selectivity. These findings suggest that the appli-
cation of soluble polymer supports should bring a decisive
advantage. Such materials can in principle be conveniently
separated from the reaction mixture through precipitation,
ultracentrifugation or ultrafiltration, whereby the use of
membrane reactors9 allows for the application of the cata-
lyst in a continuous mode. In this respect, recent results from
Kragl and coworkers with CBS catalysts supported on linear
polymers appear very encouraging.10
The interest in chiral boron-containing heterocycles as
enantioselective catalysts has been significantly increased
during the last ten years.1–7 Following the pioneering
work of Yamamoto4 with acyloxyboranes as catalysts for
asymmetric reductions and C–C bond formations, as well as
the success of the borane reduction of prochiral ketones,
catalysed by chiral oxazaborolidines (CBS reduction),1,2
quite a number of heterocyclic compounds containing
boron have been investigated. At present, chiral oxazaboro-
lidines appear to be the most versatile catalysts of this kind,
having been successfully employed in the CBS reduction as
well as in a plethora of other reactions of synthetic
interest.3,5–7 In particular, the CBS reduction has paved
the way to the development of novel synthetic strategies
for the asymmetric synthesis of important compound classes
such as prostaglandins or aminoacids, as well as for the
preparation of industrially relevant pharmaceuticals and
natural compounds.5,6
The main disadvantage of the application of oxazaboro-
lidines for asymmetric reductions is the relatively large
amount of catalyst which is required to achieve high
selectivity. At least 1 mol% is needed in the CBS reduction,
and for most of the other reactions the substrate-to-catalyst
ratios are even lower.3,6 Under these conditions, the
efficiency of catalyst recovery and recycling becomes a
critical feature for the economical viability of an industrial
process based on such compounds. Since the oxazaboro-
We present in this paper an approach making use of micro-
gels11—intramolecularly crosslinked polymer molecules
that build stable solutions in suitable solvents—as carriers
of the catalytic functionalities. In comparison to linear
polymers, microgels offer the advantage of a very low
solution viscosity, which simplifies the handling of the
Keywords: microgels; oxazaborolidines; borane; enantioselective reduction.
* Corresponding author. Tel.: ϩ49-211-8114760; fax: ϩ49-211-8114788;
e-mail: wulffg@uni-duesseldorf.de
0040–4020/00/$ - see front matter ᭧ 2000 Published by Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00072-7

1694
C. Schunicht et al. / Tetrahedron 56 (2000) 1693–1699
Figure 1. Dependence of the critical monomer concentration (Cm) on the degree of crosslinking.
reaction mixture. Furthermore, the microgel morphology
can be tuned by variation of the degree of crosslinking,
which in turn exerts an influence on the reactivity of the
bound functional groups through size-exclusion or micro-
environmental effects.12,13
monomer, the degree of crosslinking, the solvent and the
polymerisation conditions. We started our investigation by
determining Cm in THF under standard polymerisation
conditions (64ЊC, 4 days) for styrene–divinylbenzene
(DVB) microgels of different degree of crosslinking (Fig.
1). We found that Cm decreased continuously with increasing
amount of crosslinking, as expected from the steric stabili-
sation theory. However, excessive lowering of the degree of
crosslinking also caused a significant decrease in the poly-
merisation yields. We found an optimal value for obtaining
high microgel yields in relatively concentrated solutions at
about 5–10 mol% of crosslinker.
Results and Discussion
Microgels can be prepared using virtually all the known
polymerisation procedures. We chose radical solution poly-
merisation14 which is probably the simplest available
method in microgel synthesis. It does not make use of
surfactants; the stabilisation of the growing microgels
towards macrogelation is accomplished on the basis of the
osmotic repulsion forces generated by the interactions of the
polymer chains attached to the microgels—namely through
steric stabilisation.15 To achieve this, it is necessary to poly-
merise in very diluted solutions with the monomer con-
centration below a critical value (critical monomer
concentration, Cm). Under these conditions, the microgels
which are produced in the course of the polymerisation do
not react intermolecularly to build an insoluble polymer
network (macrogel), but intramolecularly to yield a stable
solution. The value of the Cm is dependent on the type of
Using the polymerisation conditions listed above, we turned
to the synthesis of microgel-bound 1,3,2-oxazaborolidines
(Fig. 2). Two different approaches can be used. The most
direct method consisted of preparing the microgels by
copolymerisation of styryl-functionalised oxazoborolidines
with styrene and DVB,16 but surprisingly this method gave
low selectivities in reductions. We therefore turned to a
second method which consisted of preparing microgels
containing boronic acid moieties followed by the synthesis
of the oxazaborolidine ring in a second step by a polymer-
analogous reaction with chiral aminoalcohols. The advan-
tage of this strategy is the flexibility of the approach: in this
Figure 2.

C. Schunicht et al. / Tetrahedron 56 (2000) 1693–1699
1695
Table 1. Synthesis of polymer-bound oxazaborolidenes (for polymerisation conditions, see Experimental)
Microgel
Functional
monomera
Crosslinking degree
and agent
Aminoalcoholb
Reaction
Content of catalyst
groups (mmol/g)
yield (%)c
M1
1
1
1
1
1
1
1
1
1
2
2
2
5% DVB
DPA
DPA
DPP
DPA
DPA
DPP
DPA
DPP
DPA
DPA
DPA
DPP
52
68
60
70
37
57
57
41
43
18
21
24
0.68
0.84
0.74
0.90
0.50
0.72
0.71
0.52
0.48
0.25
0.29
0.32
M2 (a)
M2 (b)
M3
M4 (a)
M4 (b)
M5 (a)
M5 (b)
M6
10% DVB
10% DVB
2% EDMA
5% EDMA
5% EDMA
10% EDMA
10% EDMA
30% EDMA
5% DVB
M7
M8 (a)
M8 (b)
5% EDMA
5% EDMA
a The content of 1 or 2 in the polymerisation mixture was 17 mol%.
b DPA: (S)-a,a-diphenylalaninol; DPP: (S)-a,a-diphenylprolinol.
c Based on the initial content of 1 or 2.
way different aminoalcohols can be bound to aliquots of the
same microgel backbone, allowing a better comparison of
the selectivity of different chiral modifiers.
initially assessed by considering the intensity of the related
IR absorption band at 1635 cmϪ1, but a more accurate
quantification can be obtained by a procedure involving
the titration of the acetic acid generated by the reaction of
the vinyl groups with mercury(II)-acetate.18 However,
because of the disturbing influence of boronic acid moieties
only unfunctionalised microgels could be used for this
determination.
The microgel synthesis was accomplished by the copoly-
merisation of the boronic acids 1 or 2 with different
comonomers. Because of the higher polarity, the p-styryl-
boronic acid was used as its ethylene glycol ester 1. After
copolymerisation, the ester functions were hydrolysed with
water. The microgels were isolated by precipitation.
The number of residual vinyl groups is expected to increase
with increasing amount of crosslinker. Accordingly, in the
case of our microgels, which contain only small amounts of
the crosslinking agent, the number of unreacted vinyl
groups turned out to be rather low. For an unfunctionalised
microgel based on styrene and DVB with a crosslinker
content of 10 mol%, this amount was 0.1 mmol gϪ1. For a
comparable microgel based on MMA and EDMA, the
amount of vinyl groups was hardly detectable.
To investigate the influence of the polymer backbone on
catalyst performance, we prepared styrene–divinylbenzene
(DVB) based microgels as well as ethylene dimethacrylate
(EDMA) and methyl methacrylate (MMA) based microgels.
The polymerisation conditions for EDMA based polymers
could be kept the same as for the polymers based on DVB,
as EDMA microgels in THF exhibit an even higher Cm than
DVB microgels. The molecular weights of the microgels
were determined by membrane osmometry. We found a
strong dependence on the percentage of crosslinker in the
monomer mixture. Molecular weights range from
8.2×104 g molϪ1 for the low crosslinked microgel M1 to
3.2×105 g molϪ1 for the higher crosslinked microgel M6.
The intrinsic viscosities [h] of the prepared microgels in
THF are found to lie between 8 and 15 ml gϪ1. These are
remarkably low values when compared to solutions of
analogous linear polymers, which show an intrinsic vis-
cosity of 70–100 ml gϪ1. The low viscosity of microgel
solutions allows higher concentrations of catalyst and a
much better processability (e.g. ease of ultrafiltration) in
comparison to the linear polymer analogues.
We were able to show that this small amount of vinyl groups
has no effect on the results of the CBS-reduction. For this,
we treated a microgel sample containing pendant boronic
acid moieties with the reagent diimine. By using this selec-
tive reducing agent we nearly quantitatively reduced the
vinyl groups throughout the microgel particles without
affecting the boronic acid moieties. In a direct comparison
with the original, untreated microgel still possessing vinyl
groups we found no difference in the catalytic performance
in the CBS-reduction. Thus, we decided to use all other
microgels without previous reduction of the vinyl groups.
For catalyst preparation, the microgels were condensed with
the aminoalcohols (S)-a,a-diphenylprolinol (DPP) or (S)-
a,a-diphenylalaninol (DPA) to afford the polymer-bound
oxazaborolidines. Analogous structures as depicted in 3
and 4 were obtained, in this case immobilized at a polymer.
Relevant data for the microgel-supported catalysts are
reported in Table 1. The catalysts were employed in the
enantioselective borane reduction of acetophenone or
a-tetralone, with borane–dimethylsulfide complex as the
reducing agent. The catalyst amount was 10 mol% in all
cases. The results are listed in Table 2. Free oxazaborolidine
catalysts were also tested under the same conditions for
comparison (Table 2, 3a–4b). We found that the monomeric
It is known that all microgels possess a certain amount of
pendant vinyl groups stemming from the only partially
reacted crosslinker.11a In principle the presence of such
vinyl groups can have a deleterious effect on the yield and
the enantioselectivity of the CBS-reduction, since the
borane reagent can hydroborate the residual vinyl groups
and this part would therefore be lost for the ketone reduc-
tion. Moreover, it has been reported that hydroboration of
the residual vinyl groups can cause a change in the solubility
of the microgels, which precipitate from the reaction
solution.11a,17 The amount of the pendant vinyl groups was

1696
C. Schunicht et al. / Tetrahedron 56 (2000) 1693–1699
Table 2. CBS reductions with microgel-bound catalysts (M1–M8) and
reference catalysts (3a, 3b, 4a, 4b) (for reduction conditions, see Experi-
mental)
Table 3. Consecutive reduction series with a microgel-bound catalyst
(microgel M3) (substrate: acetophenone, for reduction conditions: see
Experimental)
b
Catalysta
Aminoalcohol
Substrate
Enantiomeric
excessc (%)
Entry
Enantiomeric excessa
(%) [configuration]
[configuration]
1
2
3
4
93 [R]
93 [R]
93 [R]
M1
DPA
DPA
DPP
DPP
DPA
DPA
DPA
DPP
DPP
DPA
DPA
DPP
DPA
DPA
DPA
DPA
DPP
DPP
DPP
DPP
DPP
DPA
DPA
DPA
DPA
Acetophenone
Acetophenone
Acetophenone
a-Tetralone
Acetophenone
Acetophenone
a-Tetralone
Acetophenone
a-Tetralone
Acetophenone
a-Tetralone
Acetophenone
Acetophenone
Acetophenone
Acetophenone
a-Tetralone
Acetophenone
Acetophenone
a-Tetralone
Acetophenone
a-Tetralone
Acetophenone
a-Tetralone
91 [R]
90 [R]
87 [R]
89 [R]
93 [R]
93 [R]
93 [R]
92 [R]
94 [R]
93 [R]
93 [R]
93 [R]
84 [R]
83 [R]
90 [R]
85 [R]
89 [R]
98 [R]
97 [R]
97 [R]
97 [R]
93 [R]
94 [R]
93 [R]
93 [R]
M2 (a)
M2 (b)
M2 (b)
M3
M4 (a)
M4 (a)
M4 (b)
M4 (b)
M5 (a)
M5 (a)
M5 (b)
M6
92–93 [R]
a Determined by HPLC; isolated yield in each case Ͼ95%.
anhydrides could explain the significantly lower yield of
polymer-bound aminoalcohol (Table 1).
These initial findings prompted us to test the microgel cata-
lysts in consecutive reduction series. The reductions were
run in a stirred ultrafiltration reactor with microgel M3 as
catalyst and acetophenone as the substrate. After each
reduction, the microgel solution was purified by repeated
concentration by membrane filtration and dilution with
fresh solvent and directly recycled without isolation.
Separation and recycling of the catalyst was possible for
at least three times in addition to the first reduction with
no or only a slight decrease in the enantiomeric excess
(Table 3).
M7
M8 (a)
M8 (a)
M8 (b)
3a
3a
3b
3b
4a
4a
4b
Acetophenone
a-Tetralone
4b
These results demonstrate the applicability of microgels as
supports for oxazaborolidine catalysts. In particular, the
simple preparation by solution polymerisation and the low
intrinsic viscosity of the microgels are remarkable advan-
tages. Consequently, microgels seem to be suitable candi-
dates as soluble supports for anchoring other catalytic
groups as well, thus providing a practical means of catalyst
recovery while maintaining optimal accessibility of the
functional groups. Especially, the application in a continuous
mode in a membrane reactor seems to be rather promising.
In this case the low viscosity is a particular advantage over
the use of supported linear polymers.
a Catalyst see Table 1.
b Aminoalcohol bound to the boronic acid at the polymer: DPA, (S)-a,a-
diphenylalaninol; DPP, (S)-a,a-diphenylprolinol.
c Determined by HPLC; isolated yield in all cases Ͼ95%.
catalyst 4a and 4b derived from (S)-a,a-diphenylalaninol
exhibit excellent selectivity combined with low cost,
although the use of this aminoalcohol is rather uncommon
in the CBS reduction.
By using microgels based on monomer 1, high enantioselec-
tivities were achieved in all cases. With EDMA as the cross-
linking agent, selectivities were sometimes comparable to
the values for the new monomeric CBS catalyst 4a and close
to those for the widely used monomeric catalyst 3a19 (Table
2). The better performance of the EDMA crosslinked poly-
mers in comparison to the styrene-based analogues may be
due to the higher flexibility of the methacrylate-based
microgel network.
Experimental
General
Melting points were determined on a Bu¨chi 510 melting
point apparatus. Elemental analyses were performed at the
Institute of Pharmaceutical Chemistry, University of
1
Duesseldorf. H NMR (300 MHz) spectra were recorded
The selectivities obtained by the use of microgel catalysts
based on monomer 2 were satisfying as well but on a lower
level compared to the use of monomer 1. This result is
somewhat surprising, because CBS reductions with the
related unbound catalysts 3b and 4b gave excellent selec-
tivities comparable to those achieved with the free catalysts
3a, 4a, respectively (Table 2). Apparently, the higher flexi-
bility of the boronic acid moiety has a disturbing influence
when it is bound to a polymer backbone. The reason for this
probably lies in the possible side reaction of boronic acid
moieties to form the corresponding anhydrides. As a result,
the accessibility of the catalyst sites inside of the microgels
would be reduced because of the higher degree of cross-
linking. In addition, the formation of boronic acid
on a Varian VXR 300 spectrometer. IR spectra were taken
on a Perkin–Elmer 1420 spectrophotometer. Optical rota-
tions were measured on a Perkin–Elmer 241 MC polari-
meter. Ultrafiltration was done in Schleicher & Schuell
SC 300 and Millipore XFUF 04701 stirring cells, using
Celfa CMF-DY-040 ultrafiltration membranes (molecular
weight cut off: 40 000 Dalton). Solution viscosity measure-
ments were performed at 25ЊC in THF with a micro-
Ubbelohde viscosimeter (Schott); the flow times were deter-
mined for three microgel concentrations between 0.05 and
0.02 g/ml. Microgel molecular weights were determined on
a Knauer A0330 membrane osmometer equipped with a
Celfa CMF-DY-040 ultrafiltration membrane; the osmotic
pressure was determined in THF at 25ЊC for four microgel

C. Schunicht et al. / Tetrahedron 56 (2000) 1693–1699
1697
concentrations ranging between 0.02 and 0.002 g/ml. Chiral
HPLC analysis was accomplished with a Daicel Chiralcel
OD column (mobile phase: hexane–i-propanol 95:5; flow
rate: 1 ml/min) with UV detection at 256 nm.
The water formed was removed by azeotropic distillation
using a Dean–Stark trap filled with molecular sieves (4 A).
˚
After 8 h, the solvent was evaporated and the residue was
purified by bulb-to-bulb distillation. The product was stored
under an argon atmosphere. Yield: 55%; bp: 180ЊC
(0,05 mbar); 1H NMR (CDCL3): d (ppm): 1.30 (t, 6H,
3Jˆ7,98 Hz) 2.78 (t, 6H, 3Jˆ7,97 Hz) 7.12–7.32 (m,
15H); IR (KBr) (cmϪ1): 3200(m), 3040(m), 3005(s),
2920(s), 1595(m), 1450(s), 1330(s); MS (EI): found 396
(Mϩ), (CI): found 414 (MNHϩ4 ) calc. 396,22; elemental
analysis: calc. for C24H27B3O3: C 72.81, H 6.87; found: C
73.00, H 7.08
All solvents employed were dried and distilled if necessary
by standard procedures before use. THF was distilled over
calcium hydride, dried over sodium wire and redistilled
immediately before use under an argon atmosphere. Aceto-
phenone and a-tetralone were dried over calcium hydride
and distilled immediately before use under an argon atmos-
phere. Borane–dimethylsulfide complex (2 M in THF,
Aldrich) was used as received. tert-Butyl-peroxy-2-ethyl-
hexanoate (Luperox 26-R᭨) was a gift from Elf Atochem
and was used as received. Commercially available
monomers were dried over calcium hydride before use
and purified by distillation at reduced pressure under a nitro-
gen atmosphere. Pure p-divinylbenzene (p-DVB),20 (S)-1,1-
diphenyl-2-aminopropanol (DPA),21 (S)-1,1-diphenyl-2-
pyrrolidinemethanol (DPP),19 ethylene glycol O-[(4-vinyl-
phenyl)-boronate] 1,22 triphenylboroxine and 2-(4-vinyl-
phenyl)-ethyl boronic acid 223 were prepared according to
literature procedures.
(S)-2-(2-phenylethyl)-5,5-diphenyl-1H,3H-pyrrolo[1,2-c]-
[1,3,2]oxazaborolidine (3b). This catalyst was prepared
from DPP and tri-(2-phenylethyl)-boroxine following the
method described for catalyst 3a. The product is a colorless
liquid, which crystallizes upon standing and can be stored
under an argon atmosphere. Yield: 95%; bp 175ЊC
1
(0.055 mbar); H NMR (CDCl3): d (ppm): 7.52–7.12 (m,
3
15H) 4.32 (dd, 1H, Jˆ5.77, 9.89 Hz) 3.30–3.21 (m, 1H)
3.02–2.94 (m, 1H) 2.89 (t, 2H, 3Jˆ7.96 Hz) 1.76–1.52 (m,
3H) 1.28 (t, 2H, 3Jˆ7.97 Hz) 0.80–0.66 (m, 1H); MS (EI):
found 367 (Mϩ), calc. 367.21; elemental analysis: calc. for
C25H26BNO: C 81.70, H 7.14, N 3.81; found: C 81.34, H
7.24, N 3.63.
(S)-2,5,5-Triphenyl-1H,3H-pyrrolo[1,2-c]-[1,3,2]oxaza-
borolidine (3a). 2.0 g (7.9 mmol) DPP and 0.823 g
(2.6 mmol) triphenylboroxine were heated in 50 ml toluene
under an argon atmosphere. The water formed was removed
by azeotropic distillation using a Dean–Stark trap filled
(S)-2-(2-phenylethyl)-5,5-diphenyl-4-methyl-[1,3,2]oxaza-
borolidine (4b). This catalyst was prepared from DPA and
˚
with molecular sieves (4 A). After 16 h, the solvent was
tri-(2-phenylethyl)-boroxine
following the method
evaporated and the residue was purified by bulb-to-bulb
distillation. The product was stored under an argon atmos-
phere. Yield: 95% (lit.19 98% without distillation); bp 175ЊC
(0.1 mbar); 1H NMR (CDCl3): d (ppm): 7.97–7.93 (m, 2H)
7.63–7.55 (m, 2H) 7.48–7.17 (m, 11H) 4.60 (dd, 1H,
3Jˆ5.77, 9.67 Hz) 3.63–3.55 (m, 1H) 3.40–3.32 (m, 1H)
1.95–1.68 (m, 3H) 1.0–0.85 (m, 1H); IR (KBr) (cmϪ1):
3040(s), 3005(s), 2955(m), 2860(s), 1595(s), 1440(s)
described for catalyst 3b. Yield: 93%; bp 160ЊC
1
(0.055 mbar); H NMR (CDCl3): d (ppm): 7.62–7.16 (m,
3
15H) 4.42–4.36 (q, 1H, Jˆ6.28 Hz) 3.35 (s, 1H) 2.81 (t,
3
3
2H, Jˆ8.17 Hz) 1.27 (t, 2H, Jˆ8.18 Hz) 0.77 (d, 3H,
3Jˆ6.38 Hz); MS (EI): found 341 (Mϩ), calc. 341.20;
elemental analysis: calc. for C23H24BNO: C 80.95, H 7.09,
N 4.10; found: C 80.66, H 7.26, N 4.01.
Polymerisation procedures
(S)-2,5,5-Triphenyl-4-methyl-[1,3,2]oxazaborolidine (4a).
This catalyst was prepared from (DPA) and triphenylborox-
ine following the method described for catalyst 3a. Yield:
Unfunctionalised microgels: For the determination of the
critical monomer concentration (Cm), styrene and p-DVB
were weighed into flasks in the desired proportions. 3%
w/w azobis(isobutyronitrile) (AIBN) was added as initiator;
the mixture was diluted with the required amount of THF to
achieve a specific monomer concentration and nitrogen was
bubbled for some minutes through the solutions to remove
atmospheric oxygen. The flasks were subsequently sealed
and placed in an oven at 63ЊC for four days, after which they
were checked for gelation. The highest monomer concen-
tration which could be employed with a given monomer
mixture without sample gelation was taken as the critical
monomer concentration (Cm).
1
97%; bp 160ЊC (0.1 mbar); H NMR (CDCl3): d (ppm):
7.82–7.13 (m, 15H) 4.57–4.54 (m, 1H) 3.80 (s, 1H) 0.88
3
(d, 3H, Jˆ6.35 Hz); IR (KBr) (cmϪ1): 3400(m), 3220(m),
3020(s), 3005(s), 2960(m), 1595(s), 1440(s), 1015(s),
700(s); elemental analysis: calc. for C21H20BNO: C 80.53,
H 6.44, N 4.47; found: C 80.17, H 6.43 N 4.39
Tri-(2-phenylethyl)-boroxine. The Grignard reagent solu-
tion prepared from 8.5 g (0.35 mol) magnesium and 50 g
(0.27 mol) 2-phenylethyl bromide in 150 ml THF was
cooled to 0ЊC and added to a cooled solution of 62 g
(0.27 mol) tri-n-butylborate in 150 ml THF at Ϫ78ЊC. The
mixture was allowed to warm to room temperature over-
night and was hydrolysed with a saturated aqueous solution
of NH4Cl. After the addition of 100 ml diethyl ether, the
organic phase was isolated and the aqueous phase was
extracted twice with 150 ml of diethylether. After the
evaporation of the solvent, the butanol was evaporated azeo-
tropically under addition of water. The remaining solid was
dried, solved in 150 ml toluene and condensed to the corre-
sponding boroxine by heating under an argon atmosphere.
Some unfunctionalised microgels were isolated after poly-
merisation for the determination of the pendant vinyl
groups. For this, the microgel solutions were concentrated
to about 1/10 of their volume, added dropwise with efficient
stirring to about five times their volume of petroleum ether,
filtered off and vacuum dried to constant weight.
Microgel-bound oxazaborolidenes: A mixture of commer-
cial (DVB) (5–10 mol%; crosslinker content: 80%), styrene

1698
C. Schunicht et al. / Tetrahedron 56 (2000) 1693–1699
and monomer 1 or 2 (17 mol%) was prepared in a round-
bottomed flask; 3% (w/w) tert-butyl-peroxy-2-ethylhex-
anoate was added and the resulting mixture was diluted
with THF to reach a monomer concentration [M] of 5%
(w/w). The solution was degassed three times through
freeze–thaw cycles and polymerised at 64ЊC for four days.
of Hamersma and Snyder.24 To a microgel solution in THF
containing 1 equiv. of pendant vinyl groups, 20 equiv. of
potassium azodicarboxylate were added. 40 equiv. acetic
acid in THF (40% w/w) were added dropwise to this mixture
over 2 days. After complete addition, the microgel was
isolated by precipitation in water/MeOH (50% w/w).
(Caution: this reaction can produce hydrazine. Hydrazine
is toxic and may cause cancer.)
Functionalised microgels based on the comonomers MMA
and EDMA were also prepared by this procedure. After
polymerisation, the microgel solutions were concentrated
to about 1/10 of their volume and, when monomer 1 was
used, treated with a small amount of water for ester clea-
vage. After 15 min, the solutions were added dropwise with
efficient stirring to about five times their volume of a
mixture of water–MeOH (1:1 w/w) containing a small
amount of sodium chloride. The precipitated microgels
were filtered off and dissolved again in THF. Traces of
water were removed with sodium sulfate. The microgels
were then precipitated from petroleum ether, filtered off
and vacuum dried to constant weight.
Catalytic tests
General reduction method for low molecular weight cata-
lysts: In a dried flask, 10 mmol of the borane–dimethyl-
sulfide complex were added to a solution of 1 mmol of
catalyst 1b or 2b in 30 ml dry THF under an argon atmos-
phere. After 30 min, 10 mmol of the ketone (acetophenone
or a-tetralone) dissolved in 3 ml dry THF were added
continuously at room temperature during 30 min. The reac-
tion mixture was stirred for further 12 h, after which it was
quenched with MeOH and concentrated in vacuo. The
product was isolated by bulb-to-bulb distillation; the
enantiomeric excess was determined by HPLC.
For the conversion of the free boronic acid moieties to
oxazaborolidines, the boronic acids were condensed with
the aminoalcohols DPA or DPP. The precipitated microgel
was dissolved in THF containing a small amount of water.
1.5 equiv. of the aminoalcohol of choice were added and the
solution was heated under an argon atmosphere. The water
formed was removed by azeotropic distillation using a
General reduction method for microgel-bound catalysts: A
dried flask was charged with an amount of microgel solution
in THF corresponding to 1 g microgel. The solution was
diluted with dry THF so that a concentration of 1 g microgel
in 45 ml solvent resulted. The further procedure followed
the one employed for the low molecular weight catalysts;
the catalyst amount was maintained at 10 mol%.
˚
Dean–Stark trap filled with molecular sieves (4 A). After
48 h, the microgel solution was purified by membrane filtra-
tion in an ultrafiltration stirring cell by repeated concentra-
tion and dilution with fresh solvent and stored under an
argon atmosphere.
Consecutive reduction series with a microgel-bound cata-
lyst: The reductions were performed under an argon atmos-
phere in an ultrafiltration stirring cell with acetophenone as
the substrate. The procedure for each reduction followed the
one previously described. After each reaction, the microgel
solution was purified by repeated concentration by
membrane filtration and dilution with fresh solvent, after
which the new catalytic run was started. The filtered product
solution was quenched immediately with MeOH and the
ketone was isolated as described above.
Microgel characterisation
Catalyst content: A sample of purified microgel solution
was evaporated to dryness at room temperature. The residue
was vacuum-dried in an oven at 50ЊC to constant weight.
The catalyst content in the microgel was determined by
nitrogen analysis of the residue.
Determination of the pendant vinyl groups by oxymercu-
ration:18 200 mg of precipitated microgel were treated
with 5 ml of a solution of 150 ml dry methanol, 2.5 g
mercury(II)-acetate and 3 drops of acetic acid. After stirring
for 20 h, 0.25 g sodium bromide and 8 drops of a solution of
1% phenolphthaleine in MeOH were dissolved in the
suspension. The mixture was cooled to 0ЊC and titrated
with 0.025 M NaOH in MeOH. The amount of NaOH
solution used was corrected for the amount needed for the
titration of the pure mercury(II)-acetate solution. Another
200 mg of the same microgel were similarly treated with
5 ml of a solution of 150 ml dry methanol and 3 drops of
acetic acid and titrated following the procedure given above.
Again, this blank value was corrected for the amount of
NaOH needed for the titration of the pure solution. From
the difference between the two results, the amount of
pendant vinyl groups in the microgel sample was calculated.
Acknowledgements
This research was supported by the ‘Bundesministerium fu¨r
Bildung, Wissenschaft, Forschung und Technologie’, Bonn,
(reference number 03D0029C6) and by ‘Fonds der
Chemischen Industrie’. A fruitful cooperation with Prof.
H. G. Schmalz, University of Cologne, is gratefully
acknowledged.
References
1. Hirao, A.; Itsuno, S.; Nakahama, S.; Yamazaki, N. J. Chem.
Soc. Chem. Comm. 1981, 315.
2. Corey, E. J.; Bakshi, R. K.; Shibata, S. J. Am. Chem. Soc. 1987,
109, 5551.
3. Wallbaum, S.; Martens, J. Tetrahedron: Asymmetry 1992, 3,
1475.
4. Maruoka, K.; Yamamoto, H. Catalytic Asymmetric Synthesis;
Ojima, I., Ed.; VCH: Weinheim, 1993.
Reduction of the pendant vinyl groups with diimine:
Diimine was generated in situ by acid decomposition of
potassium azodicarboxylate, prepared following the method

C. Schunicht et al. / Tetrahedron 56 (2000) 1693–1699
1699
5. Deloux, L.; Screbnik, M. Chem. Rev. 1993, 93, 763.
6. Corey, E. J.; Helal, C. J. Angew. Chem. Int. Ed. Engl. 1998, 37,
1987.
7. (a) Shimizu, M.; Kamei, M.; Fujisawa, T. Tetrahedron Lett.
1995, 36, 8607. (b) Ostendorf, M.; Romagnoli, R.; Pereiro, I. C.;
Roos, E. C.; Moolenaar, M. J.; Nico Speckamp, W.; Hiemstra, H.
Tetrahedron: Asymmetry 1997, 8, 1773. (c) Kinugasa, M.;
Harada, T.; Oku, A. J. Org. Chem. 1996, 61, 6772. (d)
Kiyooka, S.; Hena, M. A. Tetrahedron: Asymmetry 1996, 7,
2181.
Chem. 1980, 181, 2517. (b) Stacey, K. A.; Weatherhead, R. H.;
Williams, A. Makromol. Chem. 1980, 181, 2529.
13. Ford, W. T.; Lee, J.-J.; Yu, H.; Ackerson, B. J.; Davis, K. A.
Macromol. Symp. 1995, 92, 333.
14. Graham, N. B., UK Patent GB2090264b, 1984; Graham, N. B.;
Mao, J.; Urquhart, A. Angew. Makromol. Chem. 1996, 240, 113;
Graham, N. B.; Cameron, A. Pure Appl. Chem. 1998, 70, 1271.
15. Dispersion Polymerisation in Organic Media, Barrett, K. E. J.
Ed.; Wiley: New York, 1975.
16. Schunicht, C. PhD Thesis, University of Duesseldorf, 1999.
17. Seitz, U. Makromol. Chem. 1977, 178, 1689.
8. (a) Itsuno, S.; Nakano, M.; Ito, K.; Hirao, A.; Owa, M.;
Kanda, N.; Nakahama, S. J. Chem. Soc. Perkin Trans. 1 1985,
2615. (b) Caze, C.; El Moualij, N.; Hodge, P.; Lock, C. J.; Ma, J.
J. Chem. Soc. Perkin Trans. 1 1995, 345. (c) Franot, C.; Stone,
¨
18. (a) Gyenes, I. Titrationen in nichtwaßrigen Medien; Ferdinand
Enke Verlag; 1970, p. 592. (b) Kaess, S. PhD Thesis, University of
Duesseldorf, 1988.
¨
G. B.; Engeli, P.; Spondlin, S.; Waldvogel, E. Tetrahedron:
19. Mathre, D. J.; Jones, T. K.; Xavier, L. C.; Blacklock, T. J.;
Reamer, R. A.; Mohan, J. J.; Turner Jones, E. T.; Hoogsteen, K.;
Baum, M. W.; Grabowski, E. J. J. J. Org. Chem. 1991, 56, 751.
20. (a) Campbell, T. W.; McDonald, R. N. J. Org. Chem. 1959, 24,
1246. (b) Wulff, G.; Akelah, A. Makromol. Chem. 1979, 179, 2647.
21. Dammast, F.; Reißig, H.-U. Chem. Ber. 1993, 126, 2449.
22. Wulff, G.; Hohn, J. Macromolecules 1982, 15, 1255.
23. Wulff, G.; Sarhan, A.; Gimpel, J.; Lohmar, E. Chem. Ber.
1974, 107, 3364.
Asymmetry 1995, 6, 2755.
9. See for example Prazeres, D. M. F.; Cabral, J. M. S. Enzyme
Microb. Technol. 1994, 16, 738.
10. Giffels, G.; Beliczey, J.; Felder, M.; Kragl, U. Tetrahedron:
Asymmetry 1998, 9, 691.
11. For recent reviews, see (a) Funke, W.; Okay, O.; Joos-
Mu¨ller, B. Adv. Polym. Sci. 1998, 136, 139. (b) Antonietti, M.
Macromol. Symp. 1995, 92, 1995.
12. (a) Stacey, K. A.; Weatherhead, R. H.; Williams, A. Makromol.
24. Hamersma, J. W.; Snyder, E. I. J. Org. Chem. 1965, 30, 3985.