Towards more chemically robust polymer-supported chiral catalysts: α,α-diphenyl-L-prolinol based catalysts for the reduction of prochiral ketones with borane

Article abstract of DOI:10.1039/b306294p

α,α-Diphenyl-L-prolinol derivatives with para-bromo substituants in either one or both of the phenyl rings are easily bound to crosslinked polystyrene beads containing phenylboronic acid residues by Suzuki couplings. By using extended reaction periods boronic acid residues that do not take part in the couplings are simply lost by hydrolysis. The polymer-supported (PS) α,α-diphenyl-L-prolinols were used to catalyse reductions of several prochiral ketones with borane in tetrahydrofuran at 22 °C. The expected secondary alcohols were obtained in high chemical yields and ees were generally in the range 79-97 %. One PS catalyst was recycled 14 times without loss of stereochemical performance.

Full text of DOI:10.1039/b306294p

View Article Online / Journal Homepage / Table of Contents for this issue
Towards more chemically robust polymer-supported chiral
catalysts: ꢀ,ꢀ–diphenyl-L-prolinol based catalysts for the reduction
of prochiral ketones with borane
Roger J. Kell, Philip Hodge,* Peter Snedden and David Watson
Department of Chemistry, University of Manchester, Oxford Road, Manchester, UK M13 9PL
Received 3rd June 2003, Accepted 22nd July 2003
First published as an Advance Article on the web 13th August 2003
α,α–Diphenyl--prolinol derivatives with para-bromo substituents in either one or both of the phenyl rings are easily
bound to crosslinked polystyrene beads containing phenylboronic acid residues by Suzuki couplings. By using
extended reaction periods boronic acid residues that do not take part in the couplings are simply lost by hydrolysis.
The polymer-supported (PS) α,α–diphenyl--prolinols were used to catalyse reductions of several prochiral ketones
with borane in tetrahydrofuran at 22 ЊC. The expected secondary alcohols were obtained in high chemical yields and
ees were generally in the range 79–97 %. One PS catalyst was recycled 14 times without loss of stereochemical
performance.
Introduction
catalyst recovery, and/or allowing reductions to be achieved in
6,7
flow systems. In these studies three approaches have been
used to bind the catalyst moieties to the polymer support. In
one α,α–diphenyl--prolinol (1) was reacted with beads bearing
The reduction of prochiral ketones with borane (reaction 1) in
the presence of chiral β-amino-alcohols, first introduced by
1
Itsuno et al., has since been developed into a powerful
8
benzenesulfonyl chloride residues. The catalyst moieties were
2
,3
method for asymmetric synthesis. The reductions are actually
catalysed by the oxazaborolidines formed in situ by reaction of
the β-amino-alcohols with borane. B-Substituted oxazaborol-
idines can also serve as catalysts and these can be be prepared
by reacting β-amino-alcohols with boronic acids. In terms of
the percentage enantiomeric excesses (%ees) of the secondary
alcohol products, one of the most effective β-amino-alcohols is
α,α–diphenyl--prolinol 1. This reacts with borane to give
oxazaborolidine 2, and with boronic acids to give oxaza-
borolidines 3.
then bound via the N-atom as part of a sulfonamide link. In a
second approach the β-amino-alcohol was reacted with a PS
2
9–12
boronic acid.
The catalyst moieties were then bound to the
support via the boron atom of the oxazaborolidine groups. In
the third approach, used by Wandrey et al. to prepare a soluble
2
polysiloxane for applications in a membrane reaction sys-
13,14
tem,
α,α–diphenyl-4-hydroxy--prolinol was bound via the
4,5
4
-hydroxyl group. So far, however, the α,α–diphenyl--prolinol
moiety does not appear to have been bound to a polymer
support via the phenyl residues, the approach which is the
most likely not to result in interference with the catalyst “active
site”. The work described in this paper uses this approach.
Other PS catalysts used for achieving reaction 1 are PS α,
(
1)
15
α–diphenyl--tyrosinol and certain PS chiral binaphthol
16
derivatives.
An increasingly important approach to organic synthesis
involves the use of appropriate combinations of PS reagents,
catalysts and scavengers to prepare organic compounds in solu-
17,18
tion, the PS–RCS approach.
This approach has, for example,
been used recently to achieve the synthesis of the natural prod-
19
uct (ϩ)-plicamine, and for the parallel synthesis of libraries of
20
bicyclo[2.2.2]octanes. To aid further progress in this general
area there is now a need to develop new and/or improved PS
catalysts which, as far as possible, have the following features.
(
i) They are easy to prepare with a substantial loading
Ideally they could be prepared by simply linking pre-formed
catalyst moieties to supports via chemically robust linkages
under conditions where the catalyst moieties do not need to
carry protecting groups that subsequently need to be removed.
(
ii) They are sufficiently physically robust to withstand many
reaction cycles
Magnetic stirring generally physically degrades beads. Putting
them in a “T-bag” in an apparatus that keeps the beads from
the grinding action that takes place between a magnetic stirrer
bar and the bottom of the reaction vessel is one possible solu-
Various polymer-supported (PS) α,α–diphenyl--prolinols,
or oxazaborolidine derivatives, have been prepared and studied
with the aim of simplifying reaction procedures, facilitating
21
tion. Another is to carry out the reaction in a flow system,
7
preferably with the polymer in the form of a monolith.
3
238
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 2 3 8 – 3 2 4 3
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3

View Article Online
(
iii) They are sufficiently chemically robust to withstand many
By GC the mol ratio was 77 : 23. No attempt was made to
reaction cycles
separate out the α,α–diphenyl--prolinol (1) from the bromo
compounds 7, because only the latter can take part in the
Suzuki coupling.
Even a small loss of active sites per reaction cycle is significant
over many cycles. Ester, acetal or benzyl ether linkages, for
example, used to bind the catalyst moieties to the support may
not always be sufficiently robust to withstand many reaction
cycles. There is also the serious, but so far scarcely investi-
Preparation of PS Catalysts A and B
PS boronic acid 4 was prepared via the direct lithiation of 1%
22
32
gated, question as to the long term stability of the catalyst
moieties themselves.
Using this method
crosslinked gel-type polystyrene beads.
rather than the more frequently used bromine–lithium
exchange ensures the beads contain no bromo residues which
might subsequently take part in Suzuki reactions leading to
(
iv) If the PS catalysts bring about asymmetric synthesis, then
Ϫ1
they should afford at least the same %ees as the corresponding
crosslinking. The final beads contained 2.21 mmol g of resi-
low-molecular-weight catalysts
due 4. A suspension of the beads was then reacted with
α,α–di(4-bromophenyl)--prolinol (6) and a mixture of 2 M
sodium carbonate, 1,2-dimethoxyethane and tetrakistriphenyl-
phosphinepalladium[0] at 80–85 ЊC for 4 days. This afforded PS
Catalyst A, see Scheme 1a, that by elemental analysis contained
To achieve this requires that the reactants in solution have
easy access to the catalytic sites, that there are no deleterious
interactions between the active sites, that there are no signifi-
cant microenvironmental effects, and that the “catalytic sites”
have not been altered by the method of attachment to the
Ϫ1
Ϫ1
1
.11 mmol g of nitrogen, 0.55 mmol g of bromine and
no detectable amount of boron. The former corresponds to a
23,24
support.
Ϫ1
loading of 1.11 mmol g of α,α–diphenyl--prolinol residues.
Achieving all these objectives in one system is clearly very
demanding. This paper reports progress towards meeting these
objectives. It describes the facile synthesis of PS α,α–diphenyl-
Based on the bromine analysis, 50% of these were present as
residues 8 and 50% as residues 9. The ratio of “singly bound” to
“
doubly bound” residues is consistent with the results of earlier

-prolinol derivatives where the catalyst moieties are linked to
33,34
studies.
the support through the phenyl rings via robust linkages and the
use of these catalysts to achieve the asymmetric reduction of
various prochiral ketones by borane, reaction 1, through up to
PS Catalyst B was prepared in order to determine whether
having the α,α–diphenyl--prolinol residues entirely “singly
bound” leads to better stereochemical results. If it does then
serious consideration could be given to synthesising pure
samples of one or both of the diastereoisomeric monobromo
isomers 7. Reaction of Product 1 with the PS boronic acid 4,
under similar conditions to those used to prepare PS Catalyst
A, gave PS Catalyst B: see Scheme 1b. By elemental analysis it
1
%
4 reaction cycles with little or no change in the % ee. The better
ees obtained are comparable, or only slightly less, to those
obtained using α,α–diphenyl--prolinol (1). It should be noted
that achieving excellent stereochemical results in these reduc-
tions is particularly challenging because the PS catalysts have to
10
compete with uncatalysed reductions giving racemic products.
Ϫ1
contained 0.99 mmol g of nitrogen, but no boron. Thus the
loading of “singly bound” α,α–diphenyl--prolinol residues 10
Ϫ1
was 0.99 mmol g .
Results and discussion
The initial aim of the present project was to prepare bromo
derivatives of α,α–diphenyl--prolinol. These could then be
attached to crosslinked polystyrene beads bearing boronic acid
residues 4 by Suzuki reactions. There are several reasons for
using the Suzuki reaction here. First, it is very tolerant of
The use of PS Catalysts A and B to achieve reaction 1
9,10
Similarly to our earlier studies,
the examples of reaction 1
investigated were carried out for 20 h in THF at 22 ЊC under
nitrogen with a ketone to borane ratio of 1.00 to 0.70. For the
reactions using the PS catalysts the reaction vessel was a tube
and after an appropriate amount of catalyst had been placed in
the tube it was sealed with a septum cap: see Fig. 1. All solutions
were then added or removed by syringe. In this way once a
reaction was complete the unquenched catalyst in the tube
could be stored or could be reused easily. One charge of PS
catalyst was used for a series of reactions: see below and Table
1. Each experiment was carried out in duplicate. For each pair
of experiments the %ees were essentially the same, i.e. within
25
functional groups, so that it is not necessary to protect the
functionalities present in the catalyst moieties. Second, the reac-
tion conditions do not require the use of rigorously dried sol-
vents. Thirdly, the catalyst moieties become attached through
26
very strong links which also serves as small rigid “spacers”.
Fourthly, by using an extended reaction time the boronic acid
residues 4 that do not take part in the Suzuki coupling are
simply hydrolysed to leave hydrogen atoms. Thus, the final prod-
uct does not have any residues other than the catalyst residues.
This approach to preparing functional polymers has been used
previously to prepare polymers with thiophene, phenol, amine
27
or phosphine residues, and PS catalysts containing N-methyl-
α,α–diphenyl--prolinol residues which catalyse the reactions
28
of dialkylzincs with aldehydes.
Synthesis of ꢀ,ꢀ–diphenyl-L-prolinol (1) and bromo derivatives 6
and 7
α,α–Diphenyl--prolinol (1) was prepared by the reaction of
2
9
α-N-carbonic anhydride 5 with phenylmagnesium bromide.
30
As expected this synthesis proceeded without any racemis-
ation problems. Repeating the synthesis but with 4-bromo-
phenylmagnesium bromide as the Grignard reagent gave
α,α–di(4-bromophenyl)--prolinol (6). To obtain a mono-
bromo derivative, a mixture of phenylmagnesium bromide and
31
4
-bromophenylmagnesium bromide (mol ratio 60 : 40) was
reacted with α-N-carbonic anhydride 5. As intended, the prod-
uct, Product 1, was a mixture of α,α–diphenyl--prolinol (1)
and the epimeric α,α–phenyl–4-bromophenyl--prolinols (7).
Fig. 1 Reaction vessel arrangement that allowed the PS catalysts to be
reused or stored without removal from the tube.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 2 3 8 – 3 2 4 3
3239

View Article Online
a
Table 1 Reduction of various prochiral ketones by a borane–dimethyl sulfide complex in the presence of various α,α-diphenyl--prolinol
derivatives
%ee Using:
Catalyst A
10 mol%
Catalyst B
30 mol%
α,α-diphenyl--prolinol
5 mol%
Entry
Ketone
20 mol%
30 mol%
1
2
3
4
5
6
7
8
Acetophenone
Propiophenone
2-Chloroacetophenone
4-Chloroacetophenone
α-Acetonaphthone
β-Acetonaphthone
α-Tetralone
72
70
88
83
93
87
92
85
54
84
79
97
90
99
94
99
92
89
61
80
b
c
Acetophenone
94
96
a
Reactions carried out in THF at 22 ЊC using a ketone to borane mol ratio of 1.00 : 0.70. Reactions were carried out in duplicate. The %ees were
b
c
repeatable within 2%. Chemical yields were >95%. After being used for all the 14 preceding experiments using 30 mol% of Catalyst A. After being
used for all the 8 preceding experiments using 30 mol% of Catalyst B.
Scheme 1 Preparation of PS Catalysts A and B and the catalytic moieties each catalyst contains.
2
>
%. Yields of recovered materials were >95% and of these
95% were the chiral alcohols. As in our previous studies,
in part because the “doubly bound” residue 8 correspond to
approximately 5% crosslinking in addition to the 1% originally
present in the beads. Such a high total percentage crosslinking
would seriously reduce the swelling properties of the beads and
9,10,35
the %ees were determined by GC over a chiral stationary phase.
To determine what mol% of PS Catalyst A was appropriate
to use in order that the PS catalyst could compete successfully
with the uncatalysed reaction 1 leading to racemate, three tubes
were set up for reactions using 10, 20 and 30 mol% of catalyst
and they were used for a series of reductions. It is evident from
the results summarised in Table 1, entries 1 and 2, that the %ees
improved as more catalyst was used. Since 5 mol% of the sol-
uble catalyst 1 is generally sufficient to obtain the optimal %ees,
this result suggests that a substantial fraction of the supported
catalyst residues, perhaps >80%, are not sufficiently readily
accessible for the catalysed reaction to compete successfully
with the uncatalysed reaction. Such a high percentage may arise
23
hence site accessability.
A range of other ketones were then reduced using 30 mol%
of PS Catalyst A under the standard conditions: see Table 1,
entries 3–7. The %ees ranged from 54%–92%. Various ketones
were also reduced using 30 mol% of PS Catalyst B under the
standard conditions: see Table 1, entries 1–7. With this catalyst
the ees were 3%–7% higher than with PS Catalyst A indicating
that it is better to avoid “doubly bound” sites and just have the
“singly bound” sites 10. In most cases the ees were now only
2%–4% less than those obtained with the soluble catalyst 1.
This suggests that binding the α,α–diphenyl--prolinol residues
3
240
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 2 3 8 – 3 2 4 3

View Article Online
to the support via one phenyl group does not significantly inter-
fere with the “active site” of the catalyst moieties.
instrument. Elemental analyses were made in house: for C, H
and N analyses a Carlo Erba 1108 Analyser was used: for
bromide analyses silver nitrate titrations were carried out using
a Metrohm 686 Titroprocessor: for boron analyses a Horizon
ICP Elemental Analyser was used. Optical rotations were
measured using an Optical Activity Ltd AA-100 Digital Polar-
imeter with a cell of path length 10 cm and are reported in units
It was hoped that the PS catalysts would recycle well. The
most extensively recycled was a sample of PS Catalyst A. This
was used twice for each of the 30 mol% reactions summarised
in Table 1, entries 1–7, i.e. a total of 14 times. It was then used
again at 30 mol% to catalyse the reduction of acetophenone: see
Table 1, entry 8. It is evident that the %ee obtained is, within
experimental error, the same as in the initial experiment indi-
cating that PS Catalyst A recycles well. Similarly all the experi-
ments carried out in duplicate with PS Catalyst B summarised
in entries 1, 2, 4 and 5 were carried out with just one sample
of the catalyst, i.e. a total of 8 experiments, and on reuse the
3
Ϫ1
Ϫ1
of deg cm
g
dm . GC analyses were carried out using
a Carlo Erba 4000 Chromatograph equipped with a flame-
ionisation detector and a 25 m capilliary column (0.32 mm
diameter) packed with WT COT FUS SIL (12 µ particles)
supporting the chiral species cyclodextrin-β-2,3,6-M-19.
%
ee obtained with acetophenone (see entry 8) was essentially
Preparation of ꢀ,ꢀ–diphenyl-L-prolinol (1)
unchanged from the first reduction. This suggests that the
catalytic residues are sufficiently chemically stable to withstand
many reaction cycles.
(
a) Preparation of the N-carbonic anhydride 5. A solution
of the N-carbonic anhydride 5 in THF was prepared from -
proline (5.76 g, 50 mmol) and bis(trichloromethyl) carbonate
After PS Catalyst A had been used for 16 reactions it was
recovered and washed (with acid, base and organic solvents)
before being used for a further 4 reactions. It was then washed
again and analysed. Overall the weight fell to 57% of its origin-
al value, an average loss per reaction of <3%. However the
nitrogen analysis was essentially the same as for the original
catalyst. This suggests that the catalyst groups are firmly bound
and that the losses were mainly, if not entirely, due to physical
attrition.
(
“triphosgene”) (4.95 g, 50 mmol) using the procedure of Daly
29
and Poche’. The solution was used immediately.
(
b) Reaction of the N-carbonic anhydride 5 with phenyl-
magnesium bromide. A Grignard reagent was prepared from
bromobenzene (23.55 g, 150 mmol) and magnesium (3.45 g, 150
mmol) in THF and then reacted with the anhydride 5 and the
product isolated using the procedures described in detail by
30
Mathre et al. Recrystallisation of the crude product from hex-
ane gave compound 1 (5.58 g, 44% based on proline) as white
30
Conclusions
crystals mp 78–80 ЊC (lit., 79–79.5 ЊC); δ 1.45–1.90 (m; 5H; 2 ×
H3, 2 × H4 and NH), 2.85–3.10 (m; 2H; 2 × H5), 4.25 (t, J =
PS Catalysts A and B are easily prepared by linking bromo
derivatives of α,α–diphenyl--prolinol to a polymer support
containing boronic acid residues 4 using Suzuki reactions. By
using extended reaction times boronic acid residues that do
not take part in Suzuki coupling are lost by hydrolysis. In PS
Catalyst A half of the prolinol moieties were bound through
one phenyl group and half through both phenyl groups, whilst
in PS Catalyst B all the prolinol moieties were bound through
just one phenyl group. The catalysts were used to achieve reduc-
tions of several prochiral ketones with borane in tetrahydro-
furan at 22 ЊC. The expected alcohols were obtained in high
chemical yields. The PS catalysts gave good stereochemical
results and recycled well. The best %ees were obtained when PS
Catalyst B, in which all the catalyst moieties are “singly
bound”, was used at 30 mol%. In the four cases where compar-
isons could be made with the results obtained using 5 mol% of
α,α–diphenyl--prolinol, the ees obtained with the supported
catalyst were only 2%–4% lower than those obtained with the
soluble catalyst. PS Catalyst A was shown to suffer physical
attrition on repeated use but little or no loss of catalytic sites.
Thus, most of the objectives set out above appear to have been
met with this catalyst system and the next stage of development
is to prepare a specific monobromo isomer of 7 and attach it to
a polymer which can be used in a flow system. In such a system
the polymer is unlikely to suffer significant mechanical damage
7
.5 MHz; 1H; H2), 4.55 (br s; 1H; OH), 7.10–7.4 (m; 6H;
aromatic protons) and 7.5–7.7 ppm (m; 4H; aromatic protons);
20
D
30
[α]
Ϫ53.5 (c 0.30, methanol) [lit., Ϫ54.3 (c 0.26, methanol)].
Preparation of ꢀ,ꢀ–di(4-bromophenyl)-L-prolinol (6)
A Grignard reagent was prepared from 1,4-dibromobenzene
(
35.40 g, 150 mmol) and magnesium (3.54 g, 154 mmol) in dry
31
THF. It was reacted with anhydride 5 and the product isolated
30
using the procedure described in detail by Mathre et al. This
gave the crude product 6 (4.75 g, 23% yield based on the pro-
line) as a pale amber gum. This was purified by flash chromato-
graphy over silica gel. Elution with acetone–hexane (1 : 1) gave
the mixed acetal 11, mp 44–46 ЊC; δ 1.16 (s; 3H; CH ), 1.53–1.65
3
(
m; 2H; 2 × H4), 1.69 (s; 3H; CH ), 1.80–2.15 (m; 2H; 2 × H3),
3
2
7
.81 (m; 2H; 2 × H5), 4.54 (dd, J = 3.4 and 7.5 Hz; 1H; H2),
.21 (m; 2H; aromatic protons), 7.35 (m; 2H; aromatic protons)
and 7.46 ppm (m; 2H; aromatic protons). MS (ES) 450, 452 and
ϩ
4
54 (intensites 1 : 2 : 1) corresponding to [M ϩ H] for dibromo
79 79 81 81
products 11 with two Br, one Br and one Br, and two Br.
7
on repeated reuse.
Experimental
Organic extracts were dried over magnesium sulfate. Solid
samples were dried in a vacuum oven at 1.0 mm of Hg. The
THF solution of the borane–dimethyl sulfide complex was pur-
chased from the Aldrich Chemical Company. Infrared spectra
were recorded using a Perkin Elmer 1720 instrument: solid
samples were measured as potassium bromide discs and liquid
The acetal was decomposed by treatment with a mixture of
ether (40 ml), methanol (60 ml) and hydrochloric acid (40 ml, 1
M) at 20 ЊC. Dilution of the mixture after 72 h and extraction
with ether and recovery gave α,α–di(4-bromophenyl)--prolinol
1
20
D
samples as thin films between sodium chloride plates. H NMR
(6) as a white solid, mp 84–85 ЊC; [α]
Ϫ43.4 (c 0.14, chloro-
spectra were recorded for solutions in deuteriated chloroform
on a Unity Inova 300 MHz NMR spectrometer using TMS as
an internal standard. Electrospray (ES) mass spectra (MS) were
obtained using a MicroMass Platform instrument. Chemical
ionisation (CI) MS were obtained using a Fisons VG Trio 2000
form). νmax 3368 (br O–H and N–H), 3083–3028 (aromatic
C–H) and 2973–2869 cm (aliphatic C–H); δ 1.50–1.85 (m;
Ϫ1
5H; 2 × H3, 2 × H4, NH), 2.92–3.12 (m; 2H; 2 × H5), 4.20 (t;
J = 7.6 Hz; 1H; H2), 4.40–5.00 (br s; 1H; OH) and 7.20–7.60
ppm (m; 8H; aromatic protons). C H Br NO requires C, 49.7;
17
17
2
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 2 3 8 – 3 2 4 3
3241

View Article Online
H, 4.2; N, 3.4 and Br, 38.9%. Found C, 50.1; H, 4.3; N, 3.4; and
Br, 38.8%.
1.80 mmol) was then added dropwise, again using a syringe,
then the tube was shaken at 22 ЊC for 18 h. Acetophenone (308
mg, 2.57 mmol) in THF (4.0 ml) was added by syringe over 1 h.
It was found helpful to stir the mixture magnetically during this
addition. The reaction was allowed to continue for 20 h at 22 ЊC
then the organic layer was syringed off. The beads were washed
with THF (4 × 5.0 ml). The combined organic solutions were
added to hydrochloric acid (2.0 ml of 2 M) and distilled
water (100 ml). The aqueous solution was extracted with ether
Preparation of a mixture of ꢀ,ꢀ–diphenyl-L-prolinol (1) and the
diastereoisomers of ꢀ,ꢀ–(4-bromophenyl)phenyl-L-prolinol
(
7)–(Product 1)
A mixture of Grignard reagents was prepared from bromo-
31
benzene (13.24 g, 84.3 mmol), 1,4-dibromobenzene (13.3 g,
6.2 mmol) and magnesium (3.54 g, 154 mmol) in THF. This
5
(
3 × 5.0 ml), the combined extracts washed with aqueous
was reacted with anhydride 5 using the procedure referred to
above. The product was a pale yellow oil (Product 1) (3.98 g,
sodium carbonate (5 ml of 0.5 M) and dried. Evaporation of
the solvent gave the crude product as a pale yellow oil (301 mg,
3
2% yield based on the proline). MS (CI) 331 and 333 due to
9,10,35
1
9
6%). As in previous studies,
a H NMR spectrum was
ϩ
79
81
[
M] for monobromo products 7 with Br and Br, and 253
due to unbrominated product. By elemental analysis it had
.8% Br. The NMR spectrum was very similar to that of com-
recorded and a GC run to determine both the %ee and the
chemical yield (i.e. percentage of ketone in the recovered prod-
uct converted into the desired alcohol). The GC instrument was
equipped with a flame-ionisation detector and a 25m capilliary
column (0.32 mm diameter) packed with WT COT FUS SIL
2
pounds 1 and 6. Attempts to achieve crystallization or signifi-
cant resolution of the diastereoisomers by flash chromato-
graphy failed. GC analysis indicated it consisted of compound
(
2
12 µ particles) supporting the chiral species cyclodextrin-β-
,3,6-M-19 and it was calibrated using mixtures of enantiomers
1
and the diastereoisomers 7 in the ratio 77 : 12 : 11.
9
of known composition. The PS catalyst in the tube was washed
with dry THF (2 × 5 ml) then used for the next reaction. Each
reaction was carried out in duplicate.
Preparation of crosslinked polystyrene beads containing residues
4
Polystyrene beads (gel-type; 1% crosslinked; 200–400 mesh)
were purchased from Phase Separations Ltd. Direct lithiation
of the beads in dry cyclohexane then reaction of the lithiated
product with trimethyl borate followed by hydrolysis, as
Recovery and analysis of PS Catalyst A
As indicated in Table 1, the original charge of PS Catalyst A
(599 mg) was used consecutively for 14 reactions without being
removed from the reaction tube. In between each reaction it was
simply washed with THF (2 ×). The duplicate 30 mol% runs of
the acetophenone reduction summarised in entry 8 were then
carried out. The sample was subsequently removed from the
tube and washed successively with THF, THF–2 M HCl, THF–
ammonium hydroxide, THF and then methanol and dried. The
recovered catalyst (354 mg) was then used for four more reac-
tions before being washed as above. By elemental analysis the
recovered catalyst (341 mg) contained 1.54% nitrogen (origin-
ally 1.55%) and bromine 3.49% (4.34%).
32
described in detail by Farrall and Fréchet, gave beads contain-
Ϫ1
ing 2.39% B, corresponding to 2.21 mmol g of residues 4. The
infrared spectrum (KBr disc) showed the expected bands at
Ϫ1
1
380–1310 (B–O) and 1240–620 (B–C) cm .
Preparation of PS Catalysts A and B
a) PS Catalyst A. A mixture of polystyrene beads contain-
(
ing boronic acid groups 4 (1.00 g, 2.21 mmol), α,α–di(4-bromo-
phenyl)--prolinol (6) (1.8 g), 2 M sodium carbonate (2.5 ml,
5
.0 mmol) and 1,2-dimethoxyethane (25 ml) was stirred under
argon for 15 min. Tetrakistriphenylphosphinepalladium[0]
248 mg, 0.215 mmol) was added and the mixture stirred and
(
General procedure for the reduction of ketones with borane in the
presence of compound 1
heated at 80–85 ЊC for 4 days. At the end of this period the
beads were filtered off, washed successively on the filter with
The following procedure is typical of the reductions summar-
ised in Table 1 using catalyst 1.
1
,2-dimethoxyethane (25 ml), 1,2-dimethoxyethane–water (25
ml), and ethyl acetate (30 ml), and dried. The product (1.34 g),
Ϫ1
PS Catalyst A, had 1.55% N, corresponding to 1.11 mmol g
of α,α–diphenyl--prolinol residues of both types 8 and 9,
Table 1, entry 1. Catalyst 1 (26 mg, 0.10 mmol, 5 mol%) and
a magnetic stirrer bar were placed in a round-bottomed tube
(100 mm × 20 mm). The tube was sealed with a septum cap and
nitrogen passed through the tube via syringe needles. Dry THF
4
0
.42% Br, and 0.00% of B. The bromine analysis corresponds to
Ϫ1
.55 mmol g of bromine. This indicates that PS Catalyst A
Ϫ1
contained 0.55 mmol g each of residues 8 and 9.
(
3 ml) was syringed into the tube. A solution of borane–
(
b) PS Catalyst B. This catalyst was prepared similarly to
dimethylsufide complex in THF (0.18 ml, 10 M, 1.80 mmol)
was added dropwise, again using a syringe, then the mixture
was stirred at 20 ЊC for 18 h. Acetophenone (308 mg, 2.57
mmol) in THF (4.0 ml) was added by syringe over 1 h. After 4 h
at 22 ЊC the organic layer was added to hydrochloric acid
PS Catalyst A but using Product 1 (1.8 g) in place of
the α,α–di(4-bromophenyl)--prolinol (6). The product, PS
Catalyst B, had 1.38% of N, corresponding to 0.99 mmol g of
residues 10, and 0.0% B.
Ϫ1
(
2.0 ml of 2 M) and distilled water (100 ml). The product was
General procedure for reduction of ketones with borane in the
presence of PS catalysts
extracted with ether and analysed as in the experiment with PS
Catalyst A described above.
The following is typical of the procedure used for all the reac-
tions summarised in Table 1 employing 10, 20 or 30 mol% of PS
catalyst with respect to the ketones.
Acknowledgements
We thank the EPSRC for a Ph.D. studentship (RJK).
Table 1, entry 1, using 30 mol% of catalyst. PS Catalyst A
599 mg, 0.54 mmol) and a small magnetic stirrer bar were
(
References
placed in a round-bottomed tube (100 mm × 20 mm) sealed
with a septum cap and the tube was mounted in a shaker:
see Fig. 1. Nitrogen was passed through the tube via syringe
needles. Dry THF (3 ml) was syringed into the tube and the
mixture was left for 15 min for the beads to swell. A 10 M
solution of borane–dimethylsufide complex in THF (0.18 ml,
1
2
3
A. Hirao, S. Itsuno, S. Nakahama and N. Yamazaki, J. Chem. Soc.,
Chem. Commun., 1981, 315–317.
E. J. Corey and C. J. Helal, Angew. Chem., Int. Ed., 1998, 37, 1986–
2
012.
S. Itsuno, Organic Reactions: Vol. 52, John Wiley, New York, 1998,
pp 395–576.
3
242
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 2 3 8 – 3 2 4 3

View Article Online
4
5
6
E. J. Corey, R. K. Bakshi and S. Shibata, J. Am. Chem. Soc., 1987,
09, 5551–5553.
E. J. Corey, R. K. Bakshi, S. Shibata, C.-P. Chen and V. K. Singh,
J. Am. Chem. Soc., 1987, 109, 7925–7926.
19 I. R. Baxendale, S. V. Ley and C. Piutti, Angew. Chem., Int. Ed.,
1
2002, 41, 2194–2197.
20 S. V. Ley and A. Massi, J. Chem. Soc., Perkin Trans. 1, 2001, 3645–
3654.
21 P. J. Comina, A. K. Beck and D. Seebach, Org. Process Res. Dev.,
1998, 2, 18–26.
22 P. Hodge, D. W. L. Sung and P. W. Stratford, J. Chem. Soc.,
Perkin Trans. 1, 1999, 2333–2342.
23 P. Hodge, Chem. Soc. Rev., 1997, 26, 417–424.
24 P. Hodge, in Innovation and Perspectives in Solid Phase Synthesis,
ed. R. Epton, SPCC (UK) Ltd., Birmingham, 1990, pp 273–292.
25 N. Miyaura and A. Suzuki, Chem. Rev., 1995, 95, 2457–2483.
26 M. S. Chiles and P. C. Reeves, Tetrahedron Lett., 1979, 31, 3367–
3369.
27 R. J. Kell, P. Hodge, M. Nisar and R. T. Williams, J. Chem. Soc.,
Perkin Trans. 1, 2001, 3403–3408.
28 R. J. Kell, P. Hodge, M. Nisar and D. Watson, Bioorg. Med. Chem.
Lett., 2002, 12, 1803–1807.
29 W. H. Daly and D. Poche’, Tetrahedron Lett., 1988, 29, 5859–5862.
30 D. J. Mathre, T. K. Jones, L. C. Xavier, T. J. Blacklock,
R. A. Reamer, J. J. Mohan, E. T. T. Jones, K. Hoogsteen,
M. W. Baum and E. J. J. Grabowski, J. Org. Chem., 1991, 56, 751–
762.
31 G. P. Schiemenz, Org. Synth., Coll. Vol. V, 1973, 496–499.
32 M. J. Farrall and J. M. J. Fréchet, J. Org. Chem., 1976, 41, 3877–
3882.
S. Itsuno, K. Ito, T. Maruyama, N. Kanda, A. Hirao and
S. Nakahama, Bull. Chem. Soc. Jpn, 1986, 59, 3329–3331.
P. Hodge, Curr. Opin. Biol. Chem., 2003, 7, 362–373.
7
8
J.-B. Hu, G. Zhao and Z.-D. Ding, Angew. Chem., Int. Ed., 2001, 40,
1
109–1111.
9
C. Caze, N. El Moualij, P. Hodge, C. J. Lock and J. Ma, J. Chem.
Soc., Perkin Trans. 1, 1995, 345–349.
1
1
1
1
1
0 C. Caze, N. El Moualij, P. Hodge and C. J. Lock, Polymer, 1995, 36,
6
21–629.
1 C. Franot, G. B. Stone, P. Engeli, C. Spondlin and E. Waldvogel,
Tetrahedron: Asymmetry, 1995, 6, 2755–2766.
2 C. Schunicht, A. Biffis and G. Wulff, Tetrahedron, 2000, 56, 1693–
1
699.
3 M. Felder, G. Giffels and C. Wandrey, Tetrahedron: Asymmetry,
997, 8, 1975–1977.
4 For a recent review of membrane-based reaction systems see:
H. P. Dijkstra, G. P. M. van Klink and G. van Koten, Acc. Chem.
Res., 2002, 35, 798–810.
5 S. Itsuno, M. Nakano, K. Ito, A. Hirao, M. Owa, N. Kanda and
S. Nakahama, J. Chem. Soc., Perkin Trans. 1, 1985, 2615–2619.
6 W.-S. Huang, Q.-S. Hu and L. Pu, J. Org. Chem., 1999, 64, 7940–
1
1
1
1
1
7
956.
7 A. Kirschning, H. Monenschein and R. Wittenberg, Angew. Chem.,
Int. Ed., 2001, 40, 650–679.
8 S. V. Ley, I. R. Baxendale, R. N. Bream, P. S. Jackson, A. G. Leach,
D. A. Longbottom, M. Nesi, J. S. Scott, R. I. Storer and S. J. Taylor,
J. Chem. Soc., Perkin Trans. 1, 2000, 3815–4195.
33 P. Hodge and J. Waterhouse, J. Chem. Soc., Perkin Trans. 1, 1983,
2319–2323.
34 P. Hodge and E. Khoshdel, Reactive Polymers, 1985, 3, 143–150.
35 D. W. L. Sung, P. Hodge and P. W. Stratford, J. Chem. Soc., Perkin
Trans. 1, 1999, 1463–1472.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 2 3 8 – 3 2 4 3
3243