Poly(Ethylene Glycol)-Supported Proline: A Versatile Catalyst for the Enantioselective Aldol and Iminoaldol Reactions

Article abstract of DOI:10.1002/1615-4169(200207)344:5<533::AID-ADSC533>3.0.CO;2-Y

(2S,4R)-4-Hydroxyproline has been anchored to the monomethyl ether of poly(ethylene glycol), M_w 5000, by means of a succinate spacer to afford a soluble, polymer-supported catalyst (PEG-Pro) for enantioselective aldol and iminoaldol condensation reactions. This organic catalyst can be considered as a minimalistic version of a type I aldolase enzyme, with the polymer chain replacing the enzyme's peptide backbone, and the proline residue acting as the enzyme's active site. In the presence of PEG-Pro (0.25-0.35 mol equiv.), acetone reacted with enolizable and non-enolizable aldehydes and imines to afford β-ketols and β-aminoketones in good yield and high enantiomeric excess (ee), comparable to those obtained using non-supported proline derivatives as the catalysts. Extension of the PEG-Pro-promoted condensation to hydroxyacetone as the aldol donor opened an access to synthetically relevant anti-α,β-dihydroxyketones and syn-α-hydroxy-β-aminoketones, that were obtained in moderate to good yields, and good to high diastereo- and enantioselectivity. Exploiting its solubility properties, the PEG-Pro catalyst was easily recovered and recycled to promote all of the above-mentioned reactions, that occurred in slowly diminishing yields but virtually unchanged ee's.

Full text of DOI:10.1002/1615-4169(200207)344:5<533::AID-ADSC533>3.0.CO;2-Y

FULL PAPERS
Poly(Ethylene Glycol)-Supported Proline: A Versatile Catalyst
for the Enantioselective Aldol and Iminoaldol Reactions
a,
a
a,
a
Maurizio Benaglia, * Mauro Cinquini, Franco Cozzi, * Alessandra Puglisi,
b
Giuseppe Celentano
CNR-ISTM and Dipartimento di Chimica Organica e Industriale, Universit a¡ degli Studi di Milano, via Golgi 19,
a
2
0133 Milano, Italy
Fax: ( ‡ 39) 02-5835-4159, e-mail: franco.cozzi@unimi.it
Istituto di Chimica Organica, Facolt a¡ di Farmacia, Universit a¡ degli Studi di Milano, via Venezian 21, 20133 Milano, Italy
b
Received: December 27, 2001; Accepted: February 22, 2002
Abstract: (2S,4R)-4-Hydroxyproline has been anch- line derivatives as the catalysts. Extension of the
ored to the monomethyl ether of poly(ethylene PEG-Pro-promoted condensation to hydroxyacetone
glycol), MW 5000, by means of a succinate spacer to as the aldol donor opened an access to synthetically
afford a soluble, polymer-supported catalyst (PEG- relevant anti-a,b-dihydroxyketones and syn-a-hy-
Pro) for enantioselective aldol and iminoaldol con- droxy-b-aminoketones, that were obtained in moder-
densation reactions. This organic catalyst can be ate to good yields, and good to high diastereo- and
considered as a minimalistic version of a type I enantioselectivity. Exploiting its solubility properties,
aldolase enzyme, with the polymer chain replacing the PEG-Pro catalyst was easily recovered and
the enzyme×s peptide backbone, and the proline recycled to promote all of the above-mentioned
residue acting as the enzyme×s active site. In the reactions, that occurred in slowly diminishing yields
presence of PEG-Pro (0.25 ± 0.35 mol equiv.), acetone but virtually unchanged ee×s.
reacted with enolizable and non-enolizable aldehydes
and imines to afford b-ketols and b-aminoketones in Keywords: aldol reactions; asymmetric catalysis; imi-
good yield and high enantiomeric excess (ee), com- noaldol reactions; soluble polymers; supported catal-
parable to those obtained using non-supported pro- ysis
Introduction
tially represent a very useful catalyst for the following
reasons. First, the solubility profile of PEG would
Proline has long been known to catalyze enantioselec- secure PEG-Pro to be soluble in many organic solvents
[
12]
[
1]
tive Robinson annulation processes. On this basis and insoluble in a few other solvents, thus allowing us to
there has recently been a renewed interest in a broader operate under homogeneous (and best-performing)
use of this amino acid for the biomimetic approach to catalysis conditions, and to easily recover and recycle
[
2±4]
catalytic asymmetric C-C bond-forming reactions.
In the catalyst as if working under heterogeneous condi-
[
2]
the context of aldol condensations, proline plays the tions. Secondly, being a supported organic catalyst,
role of an oversimplified version of type I aldolase PEG-Pro would not suffer from metal leaching,[
10a]
a
enzymes,[ promoting the activation of an aldol donor major problem associated with the use and the recycling
by formation of a ketone enamine and its enantioselec- of metal-based, polymer-supported catalytic systems,
tive condensation with an aldol acceptor (an aldehyde or which often require addition of fresh metal species to
an imine). The use of proline represents an ™all-organic∫ fully restore the original catalytic activity.[13] Finally,
5]
(
i.e., metal-free) alternative to the metal-based catalysts PEG-Pro can be considered a less dissimilar model than
[
6,7]
for enantioselective aldol-type processes.
proline itself of a type I aldolase enzyme, with the
In the course of a project aimed at the immobilization polymer backbone roughly acting as the peptide skel-
[
8]
[9]
[14]
of chiral ligands and catalysts on poly(ethylene eton and proline acting as the catalytic site.
glycols) (PEGs),[ we were attracted by the possibility
10]
Here, we report that the immobilization of (2S,4R)-4-
of using a modified PEG of MW 5000 (PEG5000) as hydroxyproline on PEG5000 monomethyl ether (MeO-
[
11]
support for chiral organic catalysts. We hypothesized PEG) afforded an efficient and recyclable catalyst for
that a PEG-supported proline (PEG-Pro) could poten- enantioselective aldol and iminoaldol reactions.[15]
Adv. Synth. Catal. 2002, 344, No. 5
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002 1615-4150/02/34405-533 ± 542 $ 17.50+.50/0
533

FULL PAPERS
Maurizio Benaglia et al.
Aldol Condensations
Results and Discussion
The aldol condensation between acetone (68 mol
equiv.) and 4-nitrobenzaldehyde 5a carried out at
room temperature (23 ± 25 8C) in the presence of
Synthesis of the Catalysts
Reaction of MeOPEG monosuccinate 1[10b] with diiso- catalyst 2 (0.3 mol equiv.) was selected to establish the
propylcarbodiimide (DIC, 2.2 mol equiv.) and (2S,4R)- best reaction conditions (Scheme 2 and Table 1, En-
4
-hydroxyproline (2.0 mol equiv.) at 140 8C for 20 h in tries 1 ± 8). These involved the use of an aprotic dipolar
the absence of solvent (Scheme 1) afforded PEG-Pro 2 solvent (with DMF performing better than DMSO or
in 85 ± 95% yield (see the Experimental Section for the acetonitrile, Entries 2, 4, 8), and of a reaction time of at
determination of the yields of reactions that involve least 48 h (Entries 1 vs. 2, and 3 vs. 4). Under the best
polymer-supported substrates and the purity criteria for conditions (Entry 4) aldol (R)-6a was obtained in 68%
polymer-supported products). Lower yields (35 ± 40%) isolated yield and 77% enantiomeric excess (ee), as
were observed by running the reaction for 40 h in determined by HPLC analysis on a chiral stationary
refluxing 1,2-dichloroethane using a four-fold excess of phase. The absolute configuration was determined by
[
2a,2e]
DIC and 4-hydroxyproline with respect to MeOPEG.
comparison of the sign of optical rotation.
Less
Formation of product 2 was suggested by 300 MHz polar solvents (Entries 5 ± 7) performed poorly both in
H NMR spectroscopy which showed that the coinci- terms of chemical yield and enantioselectivity. Remark-
1
dent signals of the succinic methylene groups of 1 (a ably, and quite surprisingly, the use of acetone as the
singlet at 2.60 ppm) became two distinct triplets (at 2.66 reaction solvent (Entry 7) gave an aldol of the opposite
and 2. 44 ppm) in 2. In a control experiment, it was absolute configuration.
shown that a mechanical mixture of 1 and 4-hydroxypro-
A comparison of these results with those obtained
line gave a singlet for the same protons. Configurational working with non-supported proline derivatives at the
[
2a]
stability of 4-hydroxyproline under the reaction con- same catalyst loading (0.3 ± 0.4 mol equiv.) showed
ditions required for the synthesis of 2 was demonstrated that the use of PEG-Pro achieved similar yield and
when a sample of the amino acid showed an unchanged enantioselectivity. Among the non-supported catalysts,
optical rotation after heating at 150 8C for 40 h. In order the best term of comparison is perhaps represented by
to obtain a catalyst with a double loading of catalytic (2S,4R)-4-acetoxyproline, which features an ester link-
sites, bis-succinate 3 was prepared from unprotected age at C-4 as in 2. Working with this catalyst in DMSO at
PEG4600 and transformed into the bis-proline derivative room temperature, aldol 6a was obtained in 70% yield
4
in 87% yield.
and 74% ee; proline (68% yield, 76% ee), and trans-4-
hydroxyproline (85% yield, 78% ee) performed sim-
[
2a]
ilarly. With our catalyst, the use of DMSO rather than
DMF led to marginally higher yield but lower ee
(
Entries 1 vs. 3 and 2 vs. 4).
Extension of this methodology to other aldehydes
5b± d , Entries 9 ± 12) was also possible. The corre-
O
O
(
OH
a
O
O
O
COOH
O
O
N
1
2
H
O
OH O
a
+
R-CHO
O
O
Me
Me
R
Me
O
b
HO
OH
5a - d
6a - d
O
O
O
O
3
O
O
O
Me
Me
Me
a
O
O
O
O
+
HOOC
COOH
O
O
O
O
N
H
N
H
O
4
7
=
=
MeO-(CH2CH2O)n-CH2CH2, n = ca. 110
CH2CH2O-(CH2CH2O)n-CH2CH2, n = ca. 100
a, R = 4-NO2Ph; b, R = Ph; c, R = 4-BrPh; d, R = c-C6H11.
Reagents and conditions. a: 2 (0.3 mol equiv.) or 4 (0.15 mol equiv.),
Reagents and conditions. a: DIC (2.2 mol equiv.), (2S,4R)-4-hydroxyproline
solvent, room temperature, 20 - 130 h.
(
2.0 mol equiv.), 140 °C, 20 h; b: DIC (4.4 mol equiv.), (2S,4R)-4-hydroxy-
proline (4.0 mol equiv.), 140 °C, 20 h.
Scheme 2. Synthesis of b-hydroxyketones 6a ± d and ketone
7.
Scheme 1. Synthesis of PEG-supported catalysts 2 and 4.
534
Adv. Synth. Catal. 2002, 344, 533 ± 542

Poly(Ethylene Glycol)-Supported Proline: Catalyst for Aldol and Iminoaldol Reactions
FULL PAPERS
Table 1. Catalytic enantioselective synthesis of b-hydroxyketones 6a ± d.
Entry
Aldehyde
Solvent/Time [h]
Product
Yield [%][a]
ee [%][b]
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
5a
5a
5a
5a
5a
5a
5a
5a
5b
5c
5d
5d
DMSO/20
DMSO/48
DMF/20
(R)-6a
(R)-6a
(R)-6a
(R)-6a
6a
36
73
35
68
8
60
62
67
77
DMF/48
CH Cl /48
undetermined
undetermined
21
41
59
2
2
Toluene/48
Acetone/48
6a
15
23
50
45
55
33
81
67
63
58
51
77
60
(S)-6a
(R)-6a
(R)-6b
(R)-6c
(R)-6d
(R)-6d
(R)-6a
(R)-6a
(R)-6a
(R)-6a
(R)-6d
(R)-6a
CH CN/48
3
DMF/60
DMF/60
DMF/60
DMSO/130
DMSO/24
DMF/48
DMF/48
DMF/48
1
1
1
1
1
1
1
1
1
63
93
ꢀ 98
[
c]
d]
e]
f]
5a
74
5a[
5a[
5a[
5d
77
77
75
96
[
g]
DMSO/130
DMSO/24
5a[
h]
74
[
a]
Isolated yields. For each aldehyde the highest yields were average values of at least duplicate experiments run on different
scales. The variations in yield were ꢁ4%.
[
b]
As determined by HPLC on a chiral stationary phase. The ee values for the highest yielding reactions were average values of
at least duplicate experiments. The variations in ee were ꢁ2%.
[
[
[
[
[
[
c]
d]
e]
f]
With catalyst 4.
Carried out with a catalyst sample recycled after use in Entry 4.
Carried out with a catalyst sample recycled after use in Entries 4 and 14.
Carried out with a catalyst sample recycled after use in Entries 4, 14, and 15.
Carried out with a catalyst sample recycled after use in Entry 10.
Carried out with a sample of catalyst 4 recycled after use in Entry 13.
g]
h]
sponding aldols 6b± d were obtained in yields and ee diethyl ether and filtration (70 ± 80% yield), could be
comparable to those reported using the non-supported re-used at least three times (Entries 14 ± 16) with some
catalysts.[ The result obtained with the a-branched loss of chemical efficiency but almost no effect on the
cyclohexanecarboxyaldehyde 6d (Entry 12) was par- enantioselectivity. Indeed, while the chemical yield of
ticularly relevant both in terms of yield (81%) and aldol 6a showed a slow but steady decrease of about 5%
enantioselectivity (ee > 98%). Although in this case for each cycle (from 68%, 1st cycle, to 51%, 4th cycle),
the use of DMSO and a long reaction time was the ee remained practically unchanged at the > 75%
necessary, the efficiency of the process was superior to level. A single recycling of catalyst 4 showed the same
2a]
[
2e]
[17]
that of the corresponding proline-catalyzed reaction.
trend (Entries 13 and 18).
Finally, a sample of 2
The condensation between acetone and aldehyde 5a employed for the synthesis of aldol 6c could be re-used
Entry 13) was used to test the bis-proline substituted to promote the formation of compound 6d with virtually
(
catalyst 4. At a 15% molar concentration this compound unchanged stereoselectivity (Entries 10 and 17). These
was shown to behave virtually as catalyst 2, promoting recovery/recycling experiments can be compared to
[
2e]
the synthesis of aldol 6a in 67% yield and 74% ee those performed on non-supported proline. Proline
DMSO, 24 h). This result seems to indicate that the two recovery was achieved by two different approaches:
catalytic sites∫ of 4 act independently from, and do not simply running the aldol condensation in chloroform, a
(
™
interfere with, one another. It also shows that an solvent in which the catalyst is barely soluble and thus
increase of the catalyst ™loading∫ (i.e., of the number easily recovered by filtration, or by performing the
of active groups per gram of polymer) is a viable reaction with a sample of proline non-covalently
protocol to reduce the weight amount of these high M
polymer-supported reagents.
immobilized on a silica gel column. However, both
modifications clearly depressed the efficiency of the
w
[
16]
The possibility of catalyst recovery and recycling was process, and it was concluded that the results did not
studied using the synthesis of 6a and the experiments justify these approaches toward catalyst recycling.[
2e]
reported in Entries 14 ± 18. The data showed that
From a more general point of view, it is worth
catalyst 2, readily recovered by precipitation with mentioning that the effect exerted by recycling a PEG-
Adv. Synth. Catal. 2002, 344, 533 ± 542
535

FULL PAPERS
Maurizio Benaglia et al.
supported chiral catalyst is different from that exerted
by recycling a PEG-supported chiral ligand,[ a process
that, over three cycles, led to less stereoselective
reactions (from 96% ee, 1st cycle, to 88% ee, 3rd cycle)
that occurred in almost unchanged chemical yield (from
PMP
PMP
O
O
N
NH O
8]
a
+
Me
Me
Me
Me
O2N
4-NO2Ph
Me
8
9
PMP
NH
O
a
95 to 93%).
+
R-CHO + PMPNH2
R
Me
5
a, 10a,b
9, 11a,b
Synthesis of the Wieland±Mischler Ketone
As mentioned in the introduction, the original proce-
dure in which proline was recognized as an enantiose-
lective catalyst was the Eder±Sauer±Wiechert±Hajos
synthesis of bicyclic ketones, including the Wieland±
Mischler ketone 7, a useful starting material for natural
PMP = 4-MeOPh; 10a and 11a, R = i-Pr; 10b and 11b, R = i-Bu.
Reagents and conditions. a: 2 (0.3 mol equiv.) , DMSO, room temperature,
[
1]
product synthesis. When the PEG-Pro catalyst 2
0.35 mol equiv.) was employed in the Robinson annu-
24 - 72 h.
(
Scheme 3. Synthesis of b-aminoketones 9, 11a, and 11b.
lation between 2-methyl-1,3-cyclohexanedione and 3-
propen-2-one (1.5 mol equiv.) at room temperature,
ketone (S)-7 was obtained in up to 55% yield and 75%
A comparison of these results with those obtained
[3]
ee (Scheme 2). Optimum conditions involved the use of under similar conditions with non-supported proline is
DMSO for a reaction time of 90 h, whereas the use of difficult, since the reported syntheses of compound 9
DMF led to lower yield (37%) and ee (66%). An involved samples of imine 8 generated in situ (see
attempt to improve the yield of the reaction in DMSO below). It must be noted, however, that the use of the
by prolonging the reaction time to 6 days failed, afford- ortho-methoxy isomer of 8, led to the corresponding
ing 7 in 37% yield and 72% ee. At its best, the PEG-Pro- adduct in 50% yield and 40% ee. The reported results
catalyzed reaction nicely compares with the proline- for the proline-catalyzed three-component syntheses of
[
3b]
[
3]
promoted process (same catalyst loading, DMSO as
9
differ slightly. List, working with 0.35 mol equiv. of
solvent, 89 h at room temperature) that afforded (S)-7 in catalyst (DMSO, 48 h), claimed a 50% yield for a
[
2c]
[3a]
49% yield and 76% ee.
product having a 94% ee;
Barbas, working with
0
.2 mol equiv. of catalyst (DMSO, 24 h), isolated com-
[
3b]
pound 9 in 52% yield and 89% ee. In both cases the
yields and the ee×s were lower than those observed with
catalyst 2 in the two-component reaction. In addition,
Iminoaldol Reactions
To the best of our knowledge [10a] a single report in the undisclosed amounts of the aldol product 6a and of the
literature has described the use of a polymer-supported corresponding a,b-unsaturated ketone were also ob-
catalyst for the enantioselective iminoaldol reac- tained in the proline-catalyzed three-component proce-
tion.[
18,19]
In this example a BINAP ligand immobilized dure.
[3]
on polystyrene was used for the Pd-catalyzed addition of
When PEG-Pro 2 (0.3 mol equiv.) was used to
the trimethylsilyl enol ether of acetophenone to a catalyzed the three-component synthesis of b-amino-
glyoxylate-derived imine, that occurred in 95% yield ketone 9 (DMSO, 48 h, RT; Entry 6), a 50:50 mixture of
and 81% ee. After catalyst recovery and regeneration by this product and aldol 6a was obtained in only 20%
a basic treatment a second, less efficient cycle was combined yield. With the aim of speeding up the imine
possible (72% yield, 71% ee). Since proline and its formation, 4 ä molecular sieves were added to the
analogues have been shown to promote the enantiose- reaction mixture. This increased the 9/6a ratio to 80:20
lective iminoaldol condensation between acetone and but exerted a marginal effect on the yield (36%,
pre-formed or in situ generated imines,[ catalyst 2 was combined). Slow addition of the aldehyde to the other
2e,3]
tested in this type of process.
The reaction of acetone (36 mol equiv.) with imine8 in 27% combined yield). These results suggested that the
reagents was not beneficial either (9/6a ratio ˆ 60:40;
the presence of 0.3 mol equiv. of 2 at room temperature iminoaldol reaction catalyzed by 2 is slower than that
(
Scheme 3) afforded b-aminoketone (S)-9 in yields and catalyzed by proline, and therefore the aldol process
ee that depended on the reaction time and the sol- remains an available pathway in the PEG-Pro-promot-
vent (Table 2, Entries 1 ± 5). Best results (81% yield and ed three-component reaction.
9
6% ee, Entry 5) were obtained in DMSO with a
b-Ketols were also observed as separable by-products
reaction time of 72 h, whereas shorter condensations (10 ± 20%) when the three-component condensation
(
Entries 2 ± 4) or the use of DMF (Entry 1) were less was extended to imines derived from aliphatic alde-
hydes 10a and 10b (Entries 9 ± 14, Table 2).[ The
20]
satisfactory.
536
Adv. Synth. Catal. 2002, 344, 533 ± 542

Poly(Ethylene Glycol)-Supported Proline: Catalyst for Aldol and Iminoaldol Reactions
FULL PAPERS
Table 2. Catalytic enantioselective synthesis of b-aminoketones 9, 11a, 11b.
Entry
Imine [a]
Time [h]
Product
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
8
8
8
8
12
12
24
48
72
48
48
48
24
48
72
24
48
72
72
72
9
9
18[d]
29
43
57
81
10 ‡ 10
29 ‡ 7
16 ‡ 11
41
35
38
undetermined
undetermined
60
93
(S)-9
(S)-9
(S)-9
9 ‡ 6a
9 ‡ 6a
9 ‡ 6a
(S)-11a
(S)-11a
(S)-11a
(S)-11b
(S)-11b
(S)-11b
(S)-9
(S)-9
8
96
5a ‡ PMPNH
5a ‡ PMPNH
5a ‡ PMPNH
10a ‡ PMPNH
10a ‡ PMPNH
10a ‡ PMPNH
10b ‡ PMPNH
10b ‡ PMPNH
10b ‡ PMPNH
undetermined
undetermined
undetermined
2
2
2
2
2
2
2
2
2
78
73
55
83
83
40
96
97
1
1
1
1
1
1
1
45
51
37
71
[
[
e]
f]
8
8
64
[
[
a]
In the reactions of Entries 6 ± 14 the imines were generated in situ.
b]
Isolated yields for reactions in DMSO (unless otherwise stated). Yields of Entries 6 ± 8 are for iminoaldol 9 ‡ ketol 6a.
Yields of Entries 9 ± 14 are for pure iminoaldols. For each imine the highest yields were average values of at least duplicate
experiments run on different scales. The variations in yield were ꢁ 4%.
[
c]
As determined by HPLC on a chiral stationary phase or by comparison of optical rotation values. The ee values for the
Ih ni g Dh eMs tF y. ielding reactions were average values of at least duplicate experiments. The variations in ee were ꢁ 2%.
[
[
[
d]
e]
f]
Carried out with a catalyst sample recycled after use in Entry 5.
Carried out with a catalyst sample recycled after use in Entries 5 and 15.
outcome of these reactions, carried out under conditions Reactions with Hydroxyacetone as Aldol Donor
similar to those employed for the synthesis of 9 (0.3 mol
equiv. of 2, DMSO, RT), strongly depended on the The extension of the proline-catalyzed enantioselective
[
2b,2e]
[3a]
reaction time. Best enantioselectivities were observed aldol
after 24 h for both b-aminoketones 11a (R ˆ i-Pr, 78% tone provided a convenient alternative to Sharpless×
and aminohydroxyla-
and iminoaldol reactions
to hydroxyace-
[
21]
ee, Entry 9) and 11b (R ˆ i-Bu, 83% ee, Entry 12). The asymmetric dihydroxylation
[
22]
chemical yields were moderate and decreased for tion protocols for the synthesis of 2,3-dihydroxyke-
prolonged reaction time, as did the enantioselectivities tones and 2-hydroxy-3-aminoketones, respectively. In
(
Entries 11 and 14). In order to show that a process this context, the possibility of obtaining anti-2,3-dihy-
[
2b,2e]
involving b-elimination of 4-methoxyaniline from the b- droxyketones,
a class of compounds not available in
aminoketone followed by a non-stereoselective conju- this configuration via the asymmetric dihydroxylation
[
23]
gate addition of the latter to the resulting unsaturated reaction, was particularly appealing.
ketone could account for the lower ee observed at long
We were pleased to find that the supported catalyst
reaction time, compound 11a (78% ee) was stirred for PEG-Pro 2 efficiently promoted also these synthetically
24 h at RTin DMSO in the presence of 0.3 mol equiv. of
relevant reactions in good yields and enantioselectiv-
PEG-Pro 2. From this reaction the product was recov- ities (Scheme 4 and Table 3). Condensation of hydroxy-
ered in 87% yield virtually in the racemic form, in acetone (66 mol equiv.) with cyclohexanecarboxyalde-
agreement with the proposed racemization mechanism. hyde carried out in DMF in the presence of 0.25 mol
Finally, it is worth mentioning that the results obtained equiv. of catalyst 2 (48 h, RT) afforded adduct (3S,4S)-12
in the PEG-Pro-catalyzed three-component synthesis of in 48% yield, > 20:1 anti-diastereoselectivity, and 96%
11a nicely compare to those obtained under similar ee. The use of DMSO as solvent (60 h, RT) resulted in
conditions with proline as catalyst (56% yield, 70% identical stereoselectivities and 45% yield. The latter
ee).[
3a]
was slightly lower than that observed in the proline-
Catalyst recovery and recycling, tested in the synthesis catalyzed reaction carried out under identical condi-
[
2b,24]
of (S)-9 (Entries 5, 15, and 16, Table 2), showed a trend tions (60%).
similar to that observed in the aldol condensation The condensations between hydroxyacetone (37 mol
reaction, with a constant, slow decrease of chemical equiv.) and imines 8 and 13 catalyzed by 0.35 mol equiv.
yield (from 81%, 1st cycle, to 64%, 3rd cycle), and no of PEG-Pro 2 were also investigated (Scheme 4 and
erosion of the enantioselectivity (ꢀ96%).
Table 3). Best conditions were established exploiting
Adv. Synth. Catal. 2002, 344, 533 ± 542
537

FULL PAPERS
Maurizio Benaglia et al.
Table 3. Catalytic enantioselective synthesis of a,b-dihydroxyketone 12 and a-hydroxy-b-aminoketones 14 and 15.
Entry
Acceptor
Time [h]
Product
Yield [%][a]
anti/syn[b]
ee [%][c]
1
2
3
4
5
6
7
8
9
5d
5d
48
60
60
60
48
20
48
72
72
72
(3S,4S)-12
(3S,4S)-12
(3S,4S)-12
(3S,4S)-12
(3S,4R)-14
(3S,4R)-14
(3S,4R)-14
(3S,4R)-14
(3S,4R)-14
(3S,4R)-15
48[d]
45
37
28
21[
33
73
71
69
70
20/1
>20/1
>20/1
>20/1
1/4
96
96
96
96
80
76
83
94[
92
93[
[
e]
5d
5d[
8
f]
d]
8
8
1/5.6
1/8
1/9
1/9
1/5.6
g]
8
8
[
h]
j]
1
0
13
[
a]
Isolated yields for reactions in DMSO (unless otherwise stated). For products 12 and 14 the highest yields were average
values of at least duplicate experiments run on different scales. The variations in yield were ꢁ4%.
[
[
b]
1
As determined by 300 MHz H NMR analysis of the crude reaction products.
c]
Reported values refer to the major isomer and were determined by HPLC on a chiral stationary phase. Highest ee values for
1
2 and 14 were average values of at least duplicate experiments. The variations in ee were ꢁ2%.
[
[
[
[
[
[
d]
e]
f]
In DMF.
Carried out with a catalyst sample recycled after use in Entry 2.
Carried out with a catalyst sample recycled after use in Entries 2 and 3.
The ee of the anti-isomer was 40%.
Carried out with a catalyst sample recycled after use in Entry 8.
The ee of the anti-isomer was 40%.
g]
h]
j]
OH
O
signed, on the basis of the reasonable assumption that
the stereocenters at C-3 in 12 and 14 should be the same,
and by analogywith the inversion of configurationat C-4
observed on passing from aldol 6a to iminoaldol 9a. a-
Hydroxy-b-aminoketone 15 (Entry 9) was also obtained
in almost identical yield (70%) and ee (93%), but lower
diastereoselectivity (syn:anti ˆ 5.6:1).
O
O
a
Me
+
5d
Me
OH
OH
12
R
R
N
NH O
a
+
Me
Ar
Ar
Me
OH
In order to compare the efficiency of the supported
catalyst to that of non-supported proline, the synthesis
of 14 was repeated using the latter as the catalyst. The
product was obtained in the same yield (71%) and
diastereoselection (syn:anti ratio 9:1), and a marginally
higher ee (96%) after 20 h reaction in DMSO at room
temperature, showing that also in this condensation
PEG-Pro 2 is a chemically less efficient catalyst than
OH
4, 15
8
, 13
1
O2N
O
H
H
b
Me
14
16
PMPN
O
O
[
25]
proline.
Recovery and recycling of catalyst 2 in the reactions in
which hydroxyacetone acted as the aldol donor were
studied both in the case of the synthesis of a,b-
dihydroxyketone 12 (Entries 3 and 4, Table 3) and in
that of a-hydroxy-b-aminoketone 14 (Entry 8, Table 3).
The results confirmed the trend emerged in the previous
recycling experiments (see above), showing that catalyst
recycling was indeed possible and allowed us to perform
PMP = 4-MeOPh; 8 and 14, Ar = 4-NO2Ph, R = PMP;
3 and 15, Ar = Ph, R = 4-ClPh.
1
Reagents and conditions. a: 2 (0.25 - 0.35 mol equiv.) , DMF or DMSO,
room temperature, 20 - 72 h. b: COCl2, TEA,
DMAP, 0 °C to RT, 1 h.
Scheme 4. Synthesis of a,b-dihydroxyketone 12 and of a-
hydroxy-b-aminoketones 14 and 15.
the synthesis of a-hydroxy-b-aminoketone 14 (En- equally stereoselective reactions occurring in slowly
tries 5 ± 8), that, after reaction in DMSO for 72 h, was decreasing chemical yields.
obtained in 71% yield as a 9:1 mixture of diastereoisom-
ers. The relative configuration at the stereocenters of the
major isomer was determined as syn by its conversion Conclusion
into the corresponding trans-oxazolidinone 16 (20%
phosgene in toluene, TEA, DMAP, 0 8C to RT, 1 h, 55% In conclusion, succinic acid-modified poly(ethylene
yield). HPLC analysis showed a 94% ee for adduct 14 to glycol)s were shown to be suitable polymers for the
which the (3S,4R) configuration was tentatively as- immobilization of (2S,4R)-4-hydroxyproline to afford
538
Adv. Synth. Catal. 2002, 344, 533 ± 542

Poly(Ethylene Glycol)-Supported Proline: Catalyst for Aldol and Iminoaldol Reactions
FULL PAPERS
soluble, polymer-supported catalysts for the an enan- DIC (0.204 mL, 1.32 mmol) was added dropwise. After 20 h
stirring at 140 8C, the thick mixture was cooled at RT, dissolved
tioselective version of the aldol and iminoaldol reac-
in CH Cl (5 mL), and poured dropwise in Et O (200 mL). The
tions. Both acetone and hydroxyacetone could be
successfully employed as the aldol donors in these
processes. The PEG-Pro-catalyzed condensations gave
synthetically relevant products in yields and ee com-
parable to those observed in the reactions promoted by
non-supported proline derivatives. The immobilized
2
2
2
precipitated pale brown solid was filtered, washed first with Et O
2
(
150 mL) and then with ice-cold absolute EtOH (150 mL), and
dried under vacuum to give 2; yield: 2.67 ± 2.93 g (85 ± 95%,
1
0
5
.52 ± 0.57 mmol, loading 0.191 meq/g); H NMR: d ˆ 5.58 ±
3
.67 (m, 1H, HC-4 of proline), 4.25 [t, J(H,H) ˆ 4.4 Hz, 2H,
PEG-CH OCO], 4.00 ± 4.17 (m, 1H, HC-2 of proline), 3.30
2
3
catalysts were recovered and recycled to perform (s, 3H, CH O), 2.66 [t, J(H,H) ˆ 6.8 Hz, 2H, CH COO-
3
2
reactions occurring in slowly diminishing yields but proline), 2.58 ± 2.70 (m, 2H, H C-3 of proline), 2.44 [t,
2
3
virtually unchanged enantioselectivities. These results
expand the scope of organic catalysts in the field of
enantioselective synthesis, making the search for these
reagents more and more appealing.
J(H,H) ˆ 6.8 Hz, 2H CH
2 2
CH COO-proline).
Starting from compound 3 (3 g, 0.63 mmol, loading
.211 meq/g) and following the same procedure, catalyst 4
0
was isolated; yield: 2.74 g (87%, 0.548 mmol, loading 0.2 meq/
1
g); H NMR (D O): d ˆ 4.59-4.65 (m, 2H, HC-4 of proline),
2
3
4
.25 ± 4.35 [dd, J(H,H) ˆ 8.1, 2.1 Hz, 2H, HC-2 of proline),
3
4
.22 [t, J(H,H) ˆ 4.5 Hz, 4H, PEG-CH CH OCO], 3.90 [t,
2
2
3
J(H,H) ˆ 4.7 Hz, 4H, PEG-CH CH OCO], 3.43 [dd,
2
2
Experimental Section
3
3
3
3
3
3
J(H,H) ˆ 13.0, 3.8 Hz, 2H, HC-5 of proline], 3.31 [dd,
J(H,H) ˆ 13.0, 1.0 Hz, 2H, HC-5 of proline], 2.64 [t,
J(H,H) ˆ 6.8 Hz, 4H, CH COO-proline], 2.46 22 [t,
General
2
J(H,H) ˆ 6.8 Hz, 4H, CH CH COO-proline], 2.38 [ddd,
2
2
All PEG samples, with the exception of the recovered catalysts,
were melted at 90 8C under vacuum for 30 min before use to
remove traces of moisture. After reaction, PEG-supported
product purification involved evaporation of the reaction
solvent in vacuum and addition of the residue dissolved in a few
J(H,H) ˆ 14.0, 8.0, 1.0 Hz, 2H, HC-3 of proline], 2.11 [ddd,
J(H,H) ˆ 14.0, 8.0, 2.1 Hz, 2H, HC-3 of proline].
±1
mL of CH Cl to diethyl ether (50 mL g of polymer), which
2
2
General Procedure for the Aldol Condensations
was stirred and cooled at 0 8C. After 20 ± 30 min stirring at 0 8C,
the obtained suspension was filtered through a sintered glass
filter, and the solid repeatedly washed on the filter with diethyl
The synthesis of 6a is illustrative of the procedure. To a stirred
solution of catalyst 2 (0.160 g, 0.0306 mmol), previously dried
under vacuum at 90 8C, in dry DMF (2 mL) and acetone
1
ether (up to 100 mL/g of polymer, overall). H NMR spectra
(
0.5 mL, 6.8 mmol), 4-nitrobenzaldehyde 5a (0.0151 g,
were recorded at 300 MHz as solutions in CDCl at 25 8C; d
3
0
.1 mmol) was added in one portion. The mixture was stirred
were in ppm downfield from TMS; IR spectra were recorded
on thin film or as solutions in CH Cl .
at RTfor 48 h and then poured in Et 2O (100 mL). The
precipitated 2 was filtered off, and the solid was washed with
2
2
Et O (50 mL). Average recovery of catalyst ranged from 70 to
2
8
0% (0.110 to 0.130 g) when the reaction was run on the above-
Yield and Purity Determination of PEG-Supported
Compounds
mentioned scale, and increased to ꢀ 90% when the reaction
was run on a doubled or a larger scale. The filtrate was washed
twice with water, and the Et O phase was dried and concen-
trated under vacuum. The residue was purified by flash
The yields of the PEG-supported compounds were determined
by weight with the assumption that the Mw of the PEG
fragment is 5000 Da (for compound 2) or 4600 Da (for
compound 4). The Mw actually ranged from 4500 to 5500
2
chromatography with a 1:1 hexanes:Et O mixture as eluant
2
to give the product; yield: 0.010 g (68%). The ee was
determined by HPLC (Daicel Chiralpack AS, hexane/i-
(
4200 to 5000). The indicated yields were for pure compounds.
1
PrOH ˆ 70:30, flow rate 0.8 mL/min, l ˆ 270): t : 13.28 min
The purity of these compounds was determined by H NMR
analysis in CDCl3 at 300 MHz with pre-saturation of the
methylene signals of the polymer centered at d ˆ 3.63. In
recording the NMR spectra, a relaxation time of 6 s and an
acquisition time of 4 s were used to ensure complete relaxation
and accuracy of the integration. The relaxation delay was
R
2
3
(
major) and 15.94 min (minor). [a] : ‡ 55.3 (c 0.2, CHCl ) for
D
3
[2e] 23
a sample of 77% ee (Lit.: [a] :‡46.2, c 1.0, CHCl for a
D
3
sample of 76% ee).
Aldols 6a ± d,[
2a,2e]
and diol 12[2b] were known compounds.
They had spectral data in agreement with those reported. Their
absolute configuration was assigned on the basis of comparison
of optical rotation sign. 6b (Daicel Chiralpack OJ, hexane/i-
selected after T measurements. The integration of the signals
1
of the PEG CH OCH fragment at d ˆ 3.30 and 3.36 were used
2
3
PrOH ˆ 70:30, flow rate 0.8 mL/min, l ˆ 270): t : 4.92 min
as internal standard. The estimated integration error was Æ
R
2
3
(
minor) and 7.35 min (major). [a] : ‡ 42.0 (c 0.1, CHCl ) for a
5%.
D
3
[2a] 23
sample of 59% ee (Lit.: [a] : ‡ 41.8, c 1.0, CHCl for a
D
3
sample of 60% ee). 6c (Daicel Chiralpack AS, hexane/i-
PrOH ˆ 85:15, flow rate 0.8 mL/min, l ˆ 270): t : 5.55 min
R
Synthesis of the Catalysts
2
3
(
minor) and 7.93 min (major). [a] : ‡ 41.0 (c 0.1, CHCl ) for a
D
3
To a mixture of 1 (3.0 g, 0.6 mmol, loading 0.196 meq/g),
previously dried under vacuum at 90 8C, and (2S,4R)-4-
hydroxyproline (0.156 g, 1.2 mmol) kept at 140 8C under N2,
sample of 63% ee. 6d (Daicel Chiralpack OJ, hexane/i-
PrOH ˆ 70:30, flow rate 0.8 mL/min, l ˆ 270): t : 3.64 min
R
2
3
(minor) and 4.17 min (major). [a] : ‡ 45.5 (c 0.6, CHCl ) for a
D
3
Adv. Synth. Catal. 2002, 344, 533 ± 542
539

FULL PAPERS
Maurizio Benaglia et al.
sample of 3 98% ee (Lit.:[2e] [a] : ‡ 41.5, c 1.0, CHCl for a
23
catalyst (0.3 mol equiv.) and 4-methoxyaniline (1.1 mol equiv.)
in DMSO at room temperature. The products were isolated as
described above for 9 and were considered to be likewise
configurationally unstable. Their ee×s were determined by
HPLC (Daicel Chiralpack OD, hexane/i-PrOH ˆ 85:15, flow
D
3
sample of 85% ee). 12 (Daicel Chiralpack AS, hexane/i-
PrOH ˆ 85:15, flow rate 0.8 mL/min, l ˆ 270): t : 6.84 min
R
2
3
(
minor) and 7.85 min (major). [a] : ‡ 82.2 (c 0.2, CHCl ) for a
D
3
[2b] 23
sample of 96% ee (Lit.: [a] : ‡ 83.0, c 1.0, CHCl for a
D
3
sample of the anti-isomer of > 20:1 diastereoisomeric purity
rate 0.8 mL/min, l ˆ 270). 11a: t : 6.22 min (minor) and
R
having a 99% ee). For this compound, the diastereoisomeric
ratio was determined by H NMR analysis of the crude
reaction product (the spectrum of the undetected syn-isomer
7.64 min (major). 11b: t : 6.00 min (minor) and 8.90 min
R
1
1
(major). These products had H NMR data in agreement
with those reported.[ A sample of 11a for which a 68% ee was
3a]
has been reported [ ).
2b]
determined had [a] : ‡ 8.1 (c 0.2, CHCl ). A sample of 11b for
23
D
3
2
3
which an 82% ee was determined had [a] : ‡ 7.5 (c 0.15,
D
CHCl ). No optical rotation has previously been reported for
3
these compounds.
Synthesis of the Wieland±Mischler Ketone (S)-7
To a stirred solution of catalyst 2 (0.366 g, 0.070 mmol),
previously dried under vacuum at 90 8C, in dry DMSO
(
2 mL), 2-methyl-1,3-cyclohexanedione (0.0252 g, 0.20 mmol)
Synthesis of the a-Hydroxy-b-aminoketones 14 and 15
and freshly distilled 3-buten-2-one (0.025 mL, 0.30 mmol)
were added, and the mixture was stirred at room temperature
for 90 h. The mixture was then poured in Et O (100 mL). The
precipitated 2 was filtered off, and the solid was washed with
Et O (50 mL). The filtrate was washed twice with water, and
the Et O phase was dried and concentrated under vacuum. The
residue was purified by flash chromatography with a 1:1
hexanes:Et O mixture as eluant to give the product; yield:
0
These compounds were prepared by condensing hydroxyace-
tone with imines 8 and 13 in the presence of 2 following the
procedure described above for the synthesis of compound 9.
Yields and diastereoisomeric ratios were reported in Table 3.
2
2
1
The diastereoisomeric ratios were determined by H NMR
2
analysis of the crude reaction products. The products were
purified by flash chromatography with a 3:7 hexanes:Et O
2
2
1
mixture as eluant.
.020 g (55%), the H NMR data were in agreement with those
[
26a]
23
[26b]
23
2-Hydroxy-3-(N-4-methoxyphenyl)amino-4-(4-nitrophenyl)-
-butanone 14: The pure syn-isomer was a low-melting
reported;
1
[a] : ‡77.0 (c 0.2, toluene) (Lit.:
[a] : ‡94.0, c
D
D
2
.0, toluene for a sample of 96% ee). HPLC analysis (Daicel
1
3
material. H NMR: d ˆ 8.22 [d, J(H,H) ˆ 8.7 Hz, 2H, H ortho
Chiralpack OD, hexane/i-PrOH ˆ 90:10, flow rate 0.8 mL/min,
3
to nitro group], 7.57 [d, J(H,H) ˆ 8.7 Hz, 2H, H meta to nitro
l ˆ 270): t : 18.90 min (major) and 20.42 min (minor).
R
3
group], 6.70 [d, J(H,H) ˆ 8.9 Hz, 2H, H meta to methoxy
3
group], 6.48 [d, J(H,H) ˆ 8.9 Hz, 2H, H ortho to methoxy
3
group], 5.03 [d, J(H,H) ˆ 2.3 Hz, 1H, HC-3], 4.45 [d,
Synthesis of b-Aminoketone 9
3
J(H,H) ˆ 2.3 Hz, 1H, HC-4], 4.33 and 3.93 (2 brs, 2H, OH
To a stirred solution of catalyst 2 (0.160 g, 0.0306 mmol),
previously dried under vacuum at 90 8C, in dry DMSO (2 mL)
and acetone (0.5 mL, 6.8 mmol), N-4-methoxyphenyl-4-nitro-
benzaldimine 8 (0.0256 g, 0.1 mmol) was added in one portion.
The mixture was stirred at RT for 72 h and then poured in Et O
(
and NH), 3.70 (s, 3H, methoxy group), 2.40 (s, 3H, methyl
1
3
group); C NMR: d ˆ 206.4, 153.0,148.0, 147.4, 139.2, 128.2,
124.0, 115.3, 115.1, 80.1, 58.8, 55.7, 25.0. IR (film liquid): n ˆ
±1
23
3375,1715, 1345 cm . A sample of 94% ee had [a]
: ±12.1 (c
D
0.27, CHCl ); elemental analysis: calcd. (%) for C17H N O
2
3
1
8
2
5
100 mL). The precipitated 2 was filtered off, and the solid was
(330.3): C 61.81, H 5.49, N, 8.48; found: C 61.59, H 5.40, N, 8.64.
1
washed with Et O (50 mL). Average recovery of catalyst
The pure anti-isomer was a thick oil. H NMR: d ˆ 8.17 [d,
2
3
ranged from 70 to 80% (0.110 to 0.130 g) when the reaction was
run on the above mentioned scale, and increased to ꢀ 90%
when the reaction was run on a doubled or a larger scale. The
filtrate was washed twice with water, and the Et O phase was
dried and concentrated under vacuum. The residue was purified
J(H,H) ˆ 8.8 Hz, 2H, H ortho to nitro group], 7.49 [d,
3
J(H,H) ˆ 8.8 Hz, 2H, H meta to nitro group], 6.73 [d,
3
J(H,H) ˆ 9.0 Hz, 2H, H meta to methoxy group], 6.56 [d,
3
J(H,H) ˆ 8.9 Hz, 2H, H ortho to methoxy group], 4.91 [d,
2
3
3
J(H,H) ˆ 3.3 Hz, 1H, HC-3], 4.75 [d, J(H,H) ˆ 3.3 Hz, 1H,
by flash chromatography with a 1:1 hexanes:Et O mixture as
HC-4], 3.83 and 3.60 (2 brs, 2H, OH and NH), 3.71 (s, 3H,
2
eluant to give the product; yield: 0.025 g (81%). Since this
compound was known to racemize on standing as such,[ its ee
was determined immediately after isolation of the product by
HPLC (Daicel Chiralpack AD, hexane/i-PrOH ˆ 60:40, flow
methoxy group), 2.26 (s, 3H, methyl group). A sample of 40%
3a]
23
ee had [a]
: ±11.5 (c 0.29, CHCl
). The ee×s were determined
3
D
by HPLC (Daicel Chiralpack OD, hexane/i-PrOH ˆ 85:15,
: 11.90 min (major)
: 9.09 min (minor) and
flow rate 0.8 mL/min, l ˆ 270). Syn-14: t
R
rate 0.8 mL/min, l ˆ 270): t : 9.70 min (minor) and 11.45 min
and 13.90 min (minor). Anti-14: t
R
R
1
(
minor). The product had H NMR data in agreement with
10.70 min (major).
[3a]
those reported.
A sample for which an 84% ee was
3-(N-4-Chlorophenyl)amino-2-hydroxy-4-phenyl-2-buta-
2
3
determined had [a] : ‡ 13.8 (c 0.14, CHCl ); no optical
none 15: This compound, a thick oil, was isolated as a mixture
D
3
1
rotation has previously been reported for this compound.
of diastereoisomers. For the syn-isomer: H NMR: d ˆ 7.20 ±
3
7
.30 (m, 5H, Ph), 6.97 [d, J(H,H) ˆ 8.8 Hz, 2H, H meta to
3
chloro], 6.38 [d, J(H,H) ˆ 8.7 Hz, 2H, H ortho to chloro], 4.83
3
(
2
brs, 1H, HC-3), 4.50 (brs, 1H, OH or NH), 4.36 [d, J(H,H) ˆ
Synthesis of b-Aminoketones 11a and 11b
.4 Hz, 1H, HC-4], 3.60 (brs, 1H, OH or NH), 2.25 (s, 3H,
1
3
These products were obtained by the three-component version
of the iminoaldol reaction (see above), that involved addition
of the aldehyde (1 mol equiv.) to the stirred mixture of the
methyl); C NMR: d ˆ 207.1, 144.7, 138.7, 129.1, 128.7, 126.8,
122.8,115.1, 114.9, 80.6, 58.4, 25.2. For the anti-isomer:
1
3
H NMR: d ˆ 7.20 ± 7.30 (m, 5H, Ph); 7.00 [d, J(H,H) ˆ
540
Adv. Synth. Catal. 2002, 344, 533 ± 542

Poly(Ethylene Glycol)-Supported Proline: Catalyst for Aldol and Iminoaldol Reactions
FULL PAPERS
3
8
.7 Hz, 2H, H meta to chloro], 6.44 [d, J(H,H) ˆ 8.8 Hz, 2H, H
[2] For recent applications to aldol and related reactions see:
a) B. List, R. A. Lerner, C. F. Barbas III, J. Am. Chem.
Soc. 2000, 122, 2395 ± 2396; b) W. Notz, B. List, J. Am.
Chem. Soc. 2000, 122, 7386 ± 7387; c) T. Bui, C. F. Barbas
III, Tetrahedron Lett. 2000, 41, 6951 ± 6954; d) B. List, P.
Pojarliev, C. Castello, Org. Lett. 2001, 3, 573 ± 575; e) K.
Sakthivel, W. Notz, T. Bui, C. F. Barbas III, J. Am. Chem.
Soc. 2001, 123, 5260 ± 5267.
ortho to chloro], 4.75 (brs, 1H, HC-3), 4.68 (brs, 1H, HC-4), 2.10
±1
(
s, 3H, methyl). IR (film liquid): n ˆ 3380, 1715 cm ; elemental
analysis calcd. (%) for C H ClNO (288.8): C 66.55, H 5.24, N,
1
6
15
2
4
.85; found: C 66.33, H 5.07, N, 4.99. The ee×s were determined
by HPLC (Daicel Chiralpack OD-H, hexane/i-PrOH ˆ 85:15,
flow rate 0.8 mL/min, l ˆ 270). Syn-15: t : 10.60 min (minor)
R
and 23.10 min (major). Anti-15: t : 9.10 min (major) and
R
1
1.35 min (minor).
[
3] For recent applications to iminoaldol reactions see: a) B.
List, J. Am. Chem. Soc. 2000, 122, 9336 ± 9337; b) W.
Notz, K. Sakthivel, T. Bui, G. Zhong, C. F. Barbas III,
Tetrahedron Lett. 2001, 42, 199 ± 201.
(
4R,5S)-5-Methylcarbonyl-3-(4-methoxyphenyl)-4-(4-
nitrophenyl)-2-oxazolidinone 16
[
4] For recent applications to conjugated addition reactions
see: a) M. Yamaguchi, T. Shiraishi, M. Hirama, J. Org.
Chem. 1996, 61, 3520 ± 3530; b) M. Yamaguchi, Y.
Igarashi, R. S. Reddy, T. Shraishi, M. Hirama, Tetrahe-
dron 1997, 53, 11223 ± 11236; c) S. Hanessian, V. Pham,
Org. Lett. 2000, 2, 2975 ± 2978; d) B. List, P. Pojarliev,
H. J. Martin, Org. Lett. 2001, 3, 2423 ± 2425.
To a cooled (0 8C), stirred solution of compound syn-14
0.130 g, 0.394 mmol) in dry toluene (5 mL), triethylamine
0.164 mL, 1.18 mmol), 4-dimethylaminopyridine (0.040 g,
.33 mmol), and a 20% solution of phosgene in toluene
0.227 mL, 0.44 mmol) were added in this order. After 1 h
stirring at 0 8C, water was added to the bright yellow solution.
The organic phase was separated, and the aqueous phase was
extracted twice with CH Cl . The combined organic phases
were dried over sodium sulfate and filtered, and the filtrate was
concentrated under vacuum to give the crude product. This was
purified by flash chromatography with a 7:3 hexanes:ethyl
acetate mixture as eluant. The product was obtained as a pale
yellow solid; yield: 55%; mp 260 ± 262 8C. Longer reaction
(
(
0
(
[
[
5] C.-H. Wong, R. L. Halcomb, Y. Ichikawa, T. Kajimoto,
Angew. Chem. Int. Ed. Engl. 1995, 34, 412 ± 432.
6] For recent reviews see: a) E. M. Carreira, in Compre-
hensive Asymmetric Catalysis, Vol. 3, chapter 29.1, (Eds.:
E. N. Jacobsen, A. Pfaltz, H. Yamamoto), Springer-
Verlag, Heidelberg, 1999; b) S. G. Nelson, Tetrahedron
Asymmetry 1998, 9, 357 ± 389.
2
2
1
times led to extensive product decomposition. H NMR: d ˆ
3
[7] For leading references to chiral Lewis-base catalyzed
enantioselective aldol reactions see: S. E. Denmark,
R. A. Stavenger, J. Am. Chem. Soc. 2000, 122, 8837 ±
8847 and references cited therein.
[8] R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi, M.
Pitillo, J. Org. Chem. 2001, 66, 3160 ± 3166.
8
.23 [d, J(H,H) ˆ 8.0 Hz, 2H, H ortho to nitro group], 7.55 [d,
3
J(H,H) ˆ 8.8 Hz, 2H, H meta to nitro group], 7.27 [d,
J(H,H) ˆ 9.2 Hz, 2H, H meta to methoxy group], 6.82 [d,
J(H,H) ˆ 9.0 Hz, 2H, H ortho to methoxy group], 5.64 [d,
3
3
3
3
J(H,H) ˆ 5.2 Hz, 1H, HC-5], 4.58 [d, J(H,H) ˆ 5.2 Hz, 1H,
HC-4], 3.75 (s, 3H, methoxy group), 2.48 (s, 3H, methyl group );
13
C NMR: d ˆ 204.3, 157.0, 154.0,148.0, 144.9, 127.8, 124.8,
[9] R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi, G.
Tocco, Org. Lett. 2000, 2, 1737 ± 1739.
2
3
1
23.1, 115.8, 114.5, 82.6, 61.8, 55.5, 27.0; [a] : ±15.7 (c 0.7, CH
D
2
±1
Cl ); IR (CHCl ): n ˆ 1764, 1727, 1387 cm ; elemental analysis
2
3
[10] For an excellent review on the immobilization of ligands
and catalysts on polymer supports, see: a) Chiral Catalyst
Immobilization and Recycling, (Eds.: D. E. De Vos,
I. F. J. Vankelecom, P. A. Jacobs), Wiley-VCH, Wein-
heim, 2000; for reports concerning the use of PEG as
support in this context see: b) C. Bolm, A. Gerlach,
Angew. Chem. Int. Ed. Engl. 1997, 36, 741 ± 743; c) H.
Han, K. D. Janda, Angew. Chem. Int. Ed. Engl. 1997, 36,
calcd. (%) for C H N O (356.3): C 60.67, H 4.53, N, 7.86;
18
16
2
6
found: C 60.53, H 4.44, N, 7.90.
Acknowledgements
Financial support fromCNR and MIUR ± Progetto Nazionale
Stereoselezione in Sintesi Organica. Metodologie ed Applica-
zioni is gratefully acknowledged. We thank Prof. Rita Annun-
ziata for valuable NMR assistance.
1
731 ± 1733; d) M. Glos, O. Reiser, Org. Lett. 2000, 2,
2
045 ± 2048.
[
11] For an excellent review on asymmetric organic catalysis
see: P. I. Dalko, L. Moisan Angew. Chem. Int. Ed. Engl.
2
001, 40, 3726 ± 3748.
12] D. J. Gravert, K. D. Janda, Chem. Rev. 1997, 97, 489 ±
09.
13] For instance, both soluble PEG-supported- (Ref. ) and
insoluble polystyrene-supported-bisoxazolines/Cu(II)
[
[
5
References and Notes
[8]
[
1] a) U. Eder, G. Sauer, R. Wiechert, Angew. Chem. Int. Ed.
Engl. 1971, 10, 496 ± 497; b) Z. G. Hajos, D. R. Parrish, J.
Org. Chem. 1974, 39, 1615 ± 1621; c) P. A. Grieco, N.
Fukamiya, M. Miyashita, J. Chem. Soc. Chem. Commun.
complexes, see: S. Orlandi, A. Mandoli, D. Pini, P.
Salvadori, Angew. Chem. Int. Ed. 2001, 40, 2519 ± 2521,
showed considerable decrease of catalytic efficiency
upon recovery and recycling.
1
976, 573 ± 575. For discussions on the reaction mecha-
nism see: d) C. Agami, Bull. Chem. Soc.Fr. 1988, 3, 499 ± [14] Obviously, PEG cannot be expected to exert the stereo-
07; e) D. Rajagopal, M. S. Moni, S. Subramanian, S.
directing effect of a peptide chain.
Swaminathan, Tetrahedron Asymmetry 1999, 10, 1631 ± [15] For a preliminary account of the part of this work
5
1
634.
concerning the PEG-Pro-catalyzed aldol condensation,
Adv. Synth. Catal. 2002, 344, 533 ± 542
541

FULL PAPERS
Maurizio Benaglia et al.
see: M. Benaglia, G. Celentano, F. Cozzi, Adv. Synth.
Catal. 2001, 343, 171 ± 173.
16] For different solutions to the problem of loading
[21] Review: D. J. Berrisford, C. Bolm, K. B. Sharpless,
Angew. Chem. Int. Ed. Engl. 1995, 34, 1059 ± 1070.
[22] Review: P. O×Brien, Angew. Chem. Int. Ed. Engl. 1999,
38, 326 ± 329.
[
expansion of reagents supported on high M PEG, see:
w
[
[
23] P. J. Wash, K. B. Sharpless, Synlett 1993, 605 ± 606.
24] When aromatic aldehydes were employed in this reac-
tion, both the anti selectivity and the ee dropped to less
useful levels. For instance, in the case of 2-chrobenzal-
dehyde a 60:40 mixture of anti and syn isomers was
obtained in 71% yield (DMSO, 60 h, RT). The anti
isomer had 55% ee; the syn isomer had 50% ee. Very
similar results were observed using non-supported pro-
a) M. Benaglia, R. Annunziata, M. Cinquini, F. Cozzi, S.
Ressel, J. Org. Chem. 1998, 63, 8628 ± 8629; b) N. N.
Reed, K. D. Janda, Org. Lett. 2000, 2, 1311 ± 1313.
17] The highly hygroscopic nature of PEG can be one of the
factors that account for this behavior. In this hypothesis,
the observed lower yields could arise from the fact that a
wet sample (and thus a lesser amount) of catalyst was
actually recycled. Other hypotheses, such as hydrolysis of
the ester linkages and/or degradation of the polymer
backbone were ruled out by checking by NMR the
integrity of the recovered catalyst. A precise evaluation
of the water content of the catalyst by NMR was not
possible.
[
[2b,2e]
line as catalyst (see Refs.
).
25] List (Ref.[ ) has reported that the three-component
reaction between hydroxyacetone, 2-methylpropanal,
and 4-methoxyaniline afforded (57% yield) a 17:1
mixture of syn and anti isomers, the syn isomer having
a 65% ee and the same configuration indicated for
adducts 14 and 15. Attempts to use catalyst 2 in this
reaction led to the products in very low yields (10 ±
2b]
[
[
[
18] A. Fujii, M. Sodeoka, Tetrahedron Lett. 1999, 40, 8011 ±
8
014.
19] For an example of the use of polymer supported catalysts
in the asymmetric Strecker reaction, see: M. S. Sigman,
E. N. Jacobsen, J. Am. Chem. Soc. 1998, 120, 4901 ± 4902.
20] Pre-formed aliphatic aldimines were not employed in the
iminoaldol reaction because of their intrinsic instability.
1
5%).
[26] a) E. Wada, J. Funakoshi, S. Kanemasa, Bull. Chem. Soc.
Jpn. 1992, 65, 2456 ± 2464; b) G. Zhong, T. Hoffmann,
R. A. Lerner, S. Danishefsky, C. F. Barbas, III, J. Am.
Chem. Soc. 1997, 119, 8131 ± 8132.
[
542
Adv. Synth. Catal. 2002, 344, 533 ± 542