Additive-controlled regioselective direct asymmetric aldol reaction of hydroxyacetone and aldehyde

Article abstract of DOI:10.1016/j.tetasy.2013.03.018

A structurally simple dipeptide derivative 1b prepared from l-proline and l-valine has been developed for the direct asymmetric aldol reaction of hydroxyacetone and various aldehydes with moderate to high yields and high enantioselectivities. More importa

Full text of DOI:10.1016/j.tetasy.2013.03.018

Tetrahedron: Asymmetry 24 (2013) 533–542
Contents lists available at SciVerse ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Additive-controlled regioselective direct asymmetric aldol reaction of
hydroxyacetone and aldehyde
⇑
Ling-yan Liu, Bing Wang, Yunna Zhu, Wei-xing Chang, Jing Li
State Key Laboratory of Elemento-Organic Chemistry, Nankai University, Weijin Road, Nankai District, Tianjin 300071, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 February 2013
Accepted 11 March 2013
A structurally simple dipeptide derivative 1b prepared from L-proline and L-valine has been developed for
the direct asymmetric aldol reaction of hydroxyacetone and various aldehydes with moderate to high
yields and high enantioselectivities. More importantly, this regioselective reaction could be easily
regulated by changing the additives in the presence of the same organocatalyst 1b, to afford the normal
1,2-diol adducts and the disfavoured 1,4-diol products, respectively, in a highly regioselective fashion. A
possible reaction mechanism has also been proposed.
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
as an interesting synthetic challenge (Fig. 1).3 So far, high levels
of enantioselectivity have been achieved for both the syn and anti
Since the first report of the proline-catalyzed direct asymmetric
aldol reaction,1 many organocatalysts have been designed and syn-
thesized in order to improve the catalyst reactivity, the enantiose-
lectivity, and the substrate scope.2 Since then, the organocatalysis
reaction has been rapidly developed over the last few decades. In
this field, the use of hydroxyacetone as a donor in organocatalyzed
direct aldol reactions, which generate polyoxygenated compounds,
chiral 1,2-diols or 1,4-diols, which are common structural motifs
found in a vast array of natural and biologically active molecules,
has been the subject of increasing research and is still regarded
products by utilizing chiral catalysts derived from primary and sec-
ondary amines including multipeptides or prolinamides.4 Despite
these important advances, some limitations still exist. For example,
aldol reaction donors are limited to the protected hydroxy- and
dihydroxy-acetones. The development of stereoselective organo-
catalysts for direct aldol reactions with unprotected hydroxyke-
tones still remains an important goal because of enantio- and
chemoselectivity issues. In addition, it should be noted that this
reaction between an hydroxyacetone and an aldehyde usually
gives a 1,2-diol product and hardly any 1,4-diol compound which
OH
O
OH
OH
OH
O
HO
OH
HO
OH
HO
HO
H
OH
O
F
H
O
O
O
O
O
amphidinolide H
dexamethasone
Figure 1. 1,2-Diol or 1,4-diol unit in natural products.
brassinolide
⇑
Corresponding author. Tel.: +86 22 2350 6051/1410.
E-mail addresses: liulingyan@nankai.edu.cn (L. Liu), lijing@nankai.edu.cn (J. Li).
0957-4166/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetasy.2013.03.018

534
L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
Ph
Ph
O
O
O
COOEt
COOEt
H
N
OMe
N
H
N
H
N
H
N
H
N
H
O
O
OH
Ph
A
B
Figure 2. Two organocataysts used to catalyze the reaction of hydroxy-acetone and aldehyde to afford the disfavored 1,4-diol product in Gong’s reports.
is the disfavored product. That is why that there are so few reports
on the construction of the adverse 1,4-diol product.
except for chiral catalysts 1a and 1g (Table 1, entries 2–6). In the
case of organocatalyst 1a, only the single 1,2-diol product was
obtained with high yield and moderate enantioselectivity (yield
89%, ee 50%) (entry 1). When using the chiral small organic mole-
cule 1g with a terminal hydroxy group as the catalyst, this reaction
became sluggish and produced only a trace amount of the 1,2-diol
product (yield <10%) (entry 7). This demonstrated that the terminal
hydroxy group of the prolinethioamide was disfavored for this
reaction in our protocol. Overall, this reaction generally gave the
usual 1,2-diol adduct 4a as the major product and simultaneously
generated the minor 1,4-diol compound 5a. In addition, we
observed that prolinethioamide catalysts exhibited better enanti-
oselectivities than those of the corresponding prolinamide (entries
1 and 2, entries 5 and 6). With respect to the reaction yield and
enantioselectivity, we finally chose organocatalyst 1b as the best
candidate.
To the best of our knowledge, only Gong et al. have reported on
the formation of this unfavorable product up to date. They used a
small peptide (Fig. 2, A) as an efficient catalyst for the asymmetric
direct aldol reactions of hydroxyacetone with aldehydes to afford
the chiral 1,4-diols, which are the disfavored products in similar
aldol reactions catalyzed by either aldolases or L-proline, that were
obtained in high yields and enantioselectivities in aqueous media.5
They also developed another more efficient organocatalyst (Fig. 2,
B), prepared from (2R,3R)-diethyl 2-amino-3-hydroxy succinate
and L-proline, which showed superior stereo- and regio-control
for the reaction compared to the small peptides. In the presence
of 20–30 mol % of this organocatalyst B, excellent enantioselectiv-
ities ranging from 91% to 99% and high regioselectivities were
observed for the otherwise disfavored products (1,4-diols) from
the direct aldol reactions of hydroxyacetone with a broad range
of aldehydes.6
With the optimal chiral small organic molecule catalyst in hand,
we then performed screening experiments regarding solvents in
order to further improve the regioselectivity and enantioselectivity
of this asymmetric aldol reaction.
Our research group also reported on the asymmetric aldol reac-
tion of acetone or cyclohexanone and aldehydes catalyzed by the
prolinethioamides derived from
L-proline and
L
-valine.7 On the
As shown in Table 2, this reaction could afford the 1,2-diol 4a as
the major product with high enantioselectivity in DMSO (Table 2,
entry 1). When changing to the non-coordinating solvent CH2Cl2
or DMF, the yield of the reaction product 4a was increased to
90% or 85%, but the enantioselectivity slightly decreased (entries
2 and 3). Moreover, expanding to the other aromatic aldehydes
in a solvent of CH2Cl2 or DMF, both the reaction regioselectivity
and enantioselectivity decreased sharply. Using strongly polar sol-
vents, such as CH3CN and CH3NO2, especially for CH3NO2, afforded
the 1,2-diol product 4a with largely improved yields 95%, although
with decreased enantioselectivities (entries 4 and 5). In the case of
a non-polar solvent, such as toluene, the 1,2-diol product could be
obtained with moderate enantioselectivity and good regioselectiv-
ity (Table 2, entry 6). This reaction could give the 1,2-diol product
with high (86%) yield but very low enantioselectivity (ee 20%)
under neat conditions (entry 7). Based on Gong’s report on the
regioselectivity of the aldol reaction of hydroxyacetone and an
basis of Gong’s research results, we reasoned that our prolinethioa-
mides may also act as efficient catalysts for the asymmetric direct
aldol reaction of hydroxyacetone and aldehydes. Thus, in order to
explore the application of these prolinethioamide catalysts further,
we herein report on the aldol reaction donor to encompass
hydroxyacetone.
2. Results and discussion
The direct aldol reaction of hydroxyacetone with 4-nitrobenzal-
dehyde was first selected as a model for catalyst screening and
evaluation. A series of chiral prolinamide catalysts were synthe-
sized and examined in this reaction (Fig. 3).7
The results are listed in Table 1. It was found that both aldol
adducts 4a and 5a could be generated in the presence of all of
the prolinamides 1a–1g in DMSO, with benzoic acid as an additive,
HO
COOMe
TBSO
O
S
S
S
N
H
COOMe
N
H
N
H
COOMe
N
H
CH2OOMe
N
H
N
H
N
H
N
H
1c
1d
1a
1b
Ph
Ph
S
O
S
Ph
Ph
N
H
N
H
COOMe
N
H
COOMe
N
H
N
N
H
OH
H
1g
1e
1f
Figure 3. A series of prolinamide organocatalysts.

L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
535
Table 1
Direct asymmetric aldol reaction of hydroxyacetone with 4-nitrobenzaldehyde catalyzed by prolinamide organocatalystsa
OH
O
OH
O
O
CHO
20 mol%
OH
organocatalysts
OH
+
+
OH
20 mol% PhCOOH
DMSO, 0 oC
O2N
O2N
O2N
4a
2
3a
5a
1,2-diol product
1,4-diol product
Entry
Catalysts
Product 4a
anti:sync
Product 5a
Total yieldb (%)
Yieldb (%)
eed (%)
Yieldb (%)
eed (%)
1
2
3
4
5
6
7
1a
1b
1c
1d
1e
1f
89
67
75
62
86
7:3
8:1
2:8
5:1
7:3
8:2
—
50
88
33
11
41
83
—
—
—
89
88
91
87
95
21
16
25
9
15
—
94
90
82
73
92
—
70
<10
85
<10
1g
a
Procedure: In a tube, the DMSO solvent (0.5 mL), 20mol % of organocatalyst, 20 mol % of benzoic acid and hydroxyacetone (111 mg, 1.5 mmol, 6 equiv) were added
sequentially. The reaction mixture was cooled to 0 °C, and then the 4-nitrobenzaldehyde substrate (0.25 mmol, 1 equiv) was added to this reaction system which was
followed by TLC. After completion, saturated aqueous NH4Cl was added to quench the reaction, which was then extracted with ethyl acetate three times (3 ꢀ 10 mL), dried
over anhydrous magnesium sulfate, filtered, concentrated and the residue was chromatographied by silica gel to give the pure diol product.
b
Isolated yield.
c
The diastereoselectivity value was determined by 1H NMR spectroscopy.
d
The ee was determined by HPLC using a chiral stationary phase (Daicel Chiralpak AD-H or AS-H).
Table 2
The solvent effect on the direct asymmetric aldol reaction of hydroxyacetone with 4-nitrobenzaldehyde catalyzed by the dipeptide derivative 1b
20 mol%
S
N
H
COOMe
OH
O
OH
O
N
H
OH
O
CHO
1b
+
OH
+
OH
O2N
O2N
20 mol% PhCOOH
Solv, 0 oC
O2N
4a
1,2-diol product
5a
2
3a
1,4-diol product
Entry
Solvent
Product 4a
anti:synb
Product 5a
Yielda (%)
Total yielda (%)
Yielda (%)
eec (%)
eec (%)
1
2
3
4
5
6
7
8
DMSO
CH2Cl2
DMF
CH3CN
CH3NO2
Toluene
Neat
67
90
85
87
95
75
86
<10
<10
81
8:1
19:1
8:2
7:3
8:2
8:1
8:2
—
88
81
83
50
38
55
20
—
21
—
—
—
—
20
—
—
—
—
—
—
94
—
—
—
—
85
—
—
—
—
—
—
88
90
85
87
95
95
86
<10
<10
81
H2O
9d
10d
11
12d
THF:H2O
DMSO:H2O
[Bmim]BF4
[Bmim]BF4:H2O
—
—
8:2
7:3
7:3
50
44
30
76
10
76
10
a
b
c
Isolated yield.
The diastereoselectivity was determined by 1H NMR spectroscopy.
The ee was determined by HPLC using a chiral stationary phase (Daicel Chiralpak AD-H or AS-H).
The ratio is 1:1 (0.25 mL + 0.25 mL).
d
aldehyde controlled by water in the presence of multipeptide
derivatives, we also attempted to employ water or a THF/H2O
mixture as the solvent under the standard reaction conditions.
Consequently, this reaction gave completely different results from
those of Gong’s work. The reaction catalyzed by our dipeptide
derivative hardly ran in aqueous media or THF/H2O mixture sol-
vents with only trace amounts of 1,2-diol product (entries 8 and
9). The reason for this remains unclear. To further confirm whether

536
L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
the water did play a role in the reaction of hydroxyacetone and
4-nitrobenzaldehyde, we examined the DMSO/H2O combination
system. Consequently, we observed that this reaction proceeded
smoothly and afforded the corresponding 1,2-diol adduct with
increased yield and decreased enantioselectivity compared to
when using pure DMSO as solvent (entry 10). In addition, when
employing the hydrophilic ionic liquid BmimBF4, which possesses
a strong polarity and is environmentally benign, the 1,2-diol
adduct could be generated in high yield, but the enantioselectivity
was lower (entry 11). However, this reaction gave the 1,2-diol
product with a sharply decreased yield and enantioselectivity
when carried out in a solvent system of BmimBF4 and water (entry
12). All of the above results indicated that water was unsuitable for
this reaction using our methodology.
The catalyst loading and reaction temperature usually have an
important influence on the asymmetric reaction yield and enanti-
oselectivity. Thus, we adjusted the amount of organocatalyst 1b
under the standard procedure. When increasing the amount of
prolinethioamide 1b from 20 mol % to 30 mol %, the 1,2-diol was
obtained with excellent yield (95%), although the enantioselectivi-
ty was slightly decreased (Table 3, entries 1 and 2). Decreasing the
organocatalyst loading to 10 mol % or 5 mol %, led to decreasing
yields and sharply decreased enantioselectivities (entries 3 and
4). The reaction temperature was also tested. Good regioselectivity
was produced with the single 1,2-diol compound being obtained
with up to 90% yield but moderate enantioselectivity at room
temperature (entry 5). Generally, lowering the temperature is
favorable for improving the reaction enantioselectivity. However,
no improved results were observed at ꢁ20 °C (entry 6). In view
of these results, we adopted the following reaction condition as
the optimal protocol: 20 mol % amount of catalyst 1b, 0 °C and
DMSO solvent.
in the presence or absence of benzoic acid with a slightly decreased
enantioselectivity in the latter case (entries 1 and 2). When using
2-hydroxy benzoic acid as the additive, only 1,2-diol product 4a
was obtained with 86% yield and 89% enantiomeric excess (Table 4,
entry 3). However, the results were completely different with the
major product 1,4-diol 5a (yield 50%, ee 90%) and minor 1,2-diol
4a (yield 27%, ee 65%) in the case of 4-hydroxybenzoic acid (entry
4). Further adjusting of the acidity of the additives, such as phthalic
acid, phenol, 2-hydroxyphenol or even chiral (R)-BINOL and so on,
did not improve on the results obtained (Table 4, entries 5–8).
When using a stronger acid, such as CF3COOH or CH3COOH, the
reaction hardly proceeded. This reaction regioselectivity was thor-
oughly reversed and afforded most of the 1,4-diol 5a with up to
93% ee and only trace amounts of the usual 1,2-diol 4a (entry 9)
when employing neutral Al2O3 as the auxiliary. This observation
encouraged us to further fine tune the Al2O3. Unfortunately, adopt-
ing activated Al2O3 treated by water, this reaction gave similar
results to those of unactivated Al2O3 (entry 10). Considering the
possible effect of the whole reaction system pKa value on the reac-
tion, we attempted to use basic Al2O3. However, while the regiose-
lectivity decreased slightly, the enantioselectivity of 1,4-diol
increased to 98% (entry 11). This indicated that the acid–base prop-
erty of the Al2O3 had a certain influence on the reaction regioselec-
tivity. On the other hand, we assumed that the steric hinderance of
Al2O3 may also have a dominant role on the reaction regioselectiv-
ity due to the larger mesh lattice structure of Al2O3. With these
factors in mind, other metal oxides, such as SiO2, ZnO, MgO and
CaO and so on, were also examined in the reaction of hydroxyace-
tone and 4-nitrobenzaldehyde under the standard procedure. All
reactions proceeded smoothly but afforded the products with
low regioselectivity and low enantioselectivity (entries 12–15). In
addition, we also attempted to introduce other bulky additives,
such as b-cyclodextrin, diatomite and Cr2O3 and so on; the reaction
produced the major disfavored 1,4-diol and minor 1,2-diol. The
1,4/1,2-diol ratios were 2:1, 1.2:1 and 1.3:1, respectively. This fur-
ther confirmed our hypothesis that the steric hinderance of the
additive plays a key role.
Although benzoic acid is usually used as an additive in asym-
metric aldol reactions which can accelerate the formation of an
imine or transformation to an enamine via the interaction of the
catalyst and ketone, the fine tuning of the additive may have an
important impact on the reaction yield and enantioselectivity.
Hence, additives were also investigated seriously. As shown in
Table 4, this reaction gave similar regioselectivities and total yields
Based on the above results, we concluded that this reaction
could be selectively triggered by only changing the additives from
Table 3
Screening of the catalyst loading and temperature employed in the direct asymmetric aldol reaction of hydroxyacetone with 4-nitrobenzaldehyde
S
N
H
COOMe
OH
O
OH
O
N
H
OH
O
CHO
1b
+
OH
+
OH
O2N
O2N
20 mol% PhCOOH
DMSO, temp.
O2N
3a
4a
2
5a
1,2-diol product
1,4-diol product
Entry
Temp (°C)
The amount of catalyst (mol %)
Product 4a
anti:synb
Product 5a
Total yielda (%)
Yielda (%)
eec (%)
Yielda (%)
eec (%)
1
2
3
4
5
6
0
0
0
0
rt
30
20
10
5
20
20
95
67
65
38
90
66
9:1
8:1
4:1
2:1
8:1
8:1
77
88
56
33
64
85
—
—
95
88
76
38
90
81
21
11
—
—
15
94
90
—
—
94
ꢁ20
a
b
c
Isolated yield.
The diastereoselectivity was determined by 1H NMR spectroscopy.
The ee was determined by HPLC using a chiral stationary phase (Daicel Chiralpak AD-H or AS-H).

L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
537
Table 4
The different additives employed in the direct asymmetric aldol reaction of hydroxyacetone with 4-nitrobenzaldehyde
20 mol%
S
N
H
COOMe
OH
O
OH
O
N
H
OH
O
CHO
+
1b
OH
+
OH
O2N
O2N
20 mol% additive
DMSO, 0 oC
O2N
4a
1,2-diol product
5a
2
3a
1,4-diol product
Entry
Additive
Product 4a
anti:synb
Product 5a
Total yielda (%)
Yielda (%)
eec (%)
Yielda (%)
eec (%)
1
2
3
4
5
6
7
8
—
65
67
86
27
65
27
45
62
6
12
38
55
37
47
20
8:2
8:1
8:1
7:1
8:1
1:1
1:1
8:2
9:1
8:1
8:2
4:1
6:1
7:1
1:9
79
88
89
65
67
50
47
50
63
30
43
56
67
7
22
21
—
90
94
—
87
88
86
77
94
57
99
87
73
76
88
99
100
72
41
PhCOOH
2-OH-C6H4COOH
4-OH-C6H4COOH
Phthalic acid
PhOH
2-OH-C6H4OH
(R)-BINOL
Al2O3
Al2O3 + 5
Al2O3 (basic)
SiO2
50
29
30
54
25
67
64
50
44
63
25
21
90
90
91
90
86
93
92
98
90
85
89
93
9
10
11
12
13
14
15
l
L H2O
ZnO
MgO
CaO
10
a
b
c
Isolated yield.
The diastereoselectivity was determined by 1H NMR spectroscopy.
The ee was determined by HPLC using a chiral stationary phase (Daicel Chiralpak AD-H or AS-H).
O
OH
O
OH
O
Al2O3
(20 mol%)
2-HO-C6H4COOH
(20 mol%)
OH
20 mol% 1b
20 mol% 1b
OH
OH
DMSO, 0 oC
O2N
O2N
DMSO, 0 oC
CHO
5a
4a
O2N
86% yield
67% yield
93% ee
anti:syn=8:1
89% ee
S
1b
N
H
COOMe
Cat.
N
H
Scheme 1. A regioselective aldol reaction of hydroxyacetone and 4-nitrobenzaledhyde.
2-hydroxybenzoic acid to aluminum oxide and catalyzed by the
same organocatalyst 1b at 0 °C in DMSO solvent (Scheme 1).
To better understand the scope and limitations of our catalyst
systems, some aromatic aldehydes were chosen as aldol acceptors
in the presence of salicylic acid or Al2O3, respectively (Tables 5 and
6). As shown in Table 5, it can be seen that the favored 1,2-diols 4
were mainly obtained with high yields and high diastereoselectiv-
ities as well as excellent enantioselectivities when using the
salicylic acid as an additive. Upon closer inspection of these data,
we observed that the aromatic aldehydes with strong electron-
withdrawing substituents were better substrates than their unsub-
stituted counterpart in terms of reactivity and stereoselectivity
(Table 5, entries 1–8 vs 9). For example, for the 4-nitro, 2-trifluoro-
methyl and 4-cyanobenzaldehyde substrates, only single 1,2-diols
products were obtained (entries 1, 5 and 6) under the standard
procedure. Furthermore, excellent enantioselectivity was obtained
(up to 98% for the 2-trifluoromethyl benzaldehyde). In comparison,
for the other substituted benzaldehydes, there were small amount
of 1,4-diols obtained except for the major 1,2-diols (Table 5, entries
2–4 and 7–9).
Similarly, adopting Al2O3 as an auxiliary, the above substrates
were also examined in the asymmetric aldol reaction with hydrox-

538
L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
Table 5
The asymmetric aldol reaction of hydroxyacetone and various aldehydes catalyzed by a prolinethioamide with salicylic acid as an additivea
20 mol%
S
N
H
COOMe
OH
O
N
H
OH
O
O
CHO
OH
1b
R
R
+
R
20 mol% 2-OH-C6H4COOH
DMSO, 0 oC-rt.
OH
4 (major)
OH
5 (minor)
3
2
Entry
–R
Product 4 (major)
Product 5 (minor)
Products (4+5)
Yieldb (%)
86
anti:sync
eed (%)
Yieldb (%)
Total yieldb (%)
1
2
3
4
5
6
7
8
9
4-NO2
2-NO2
3-NO2
4-CF3
2-CF3
4-CN
2,6-Cl2
4-Br
8:1
8:1
7:1
8:1
7:1
8:1
8:1
7:1
7:1
89
90
92
87
98
80
97
95
96
—
15
10
14
—
86
96
87
96
76
86
96
87
66
81
77
82
76
86
81
77
51
—
15
10
15
—
a
Procedure: In a tube, the DMSO solvent (0.5 mL), 20mol % organocatalyst 1b, 20mol % salicylic acid and hydroxyacetone (111 mg, 1.5 mmol, 6 equiv) were added
sequentially. The reaction mixture was cooled to 0 °C, and then the aromatic aldehyde substrates (0.25 mmol, 1 equiv) were added to this reaction system which was
followed by TLC. After completion, saturated aqueous NH4Cl was added to quench the reaction, which was extracted with ethyl acetate three times (3 ꢀ 10 mL), dried over
anhydrous magnesium sulfate, filtered, concentrated and the residue was chromatographied on silica gel to give the pure diol product.
b
Isolated yield.
c
The diastereoselectivity was determined by 1H NMR spectroscopy.
d
The ee was determined by HPLC using a chiral stationary phase (Daicel Chiralpak AD-H or AS-H).
Table 6
The asymmetric aldol reaction of hydroxyacetone and various aldehydes catalyzed by a prolinethioamide in the presence of aluminum oxidea
20 mol%
S
N
H
COOMe
OH
O
OH
O
N
H
O
CHO
1b
OH
+
R
+
R
R
20 mol% Al2O3
DMSO, 0 oC-rt
OH
OH
5
2
(major)
4 (minor)
3
Entry
–R
Product 5 (major)
Product 4 (minor)
Products (4+5)
Yieldb (%)
eec (%)
Yieldb (%)
Total yieldb (%)
1
2
3
4
5
6
7
8
9
4-NO2
2-NO2
3-NO2
2-CF3
4-CF3
4-CN
2,6-Cl2
4-Br
67
59
76
71
55
65
59
50
41
93
87
85
92
89
94
95
88
96
6
23
15
18
37
11
23
25
22
73
82
91
89
92
76
82
75
63
—
a
Procedure: In a tube, the DMSO solvent (0.5 mL), 20 mol % organocatalyst 1b, 20mol % Al2O3 and hydroxyacetone (111 mg, 1.5 mmol, 6 equiv) were added sequentially.
The reaction mixture was cooled to 0 °C, and then the aromatic aldehyde substrates (0.25 mmol, 1 equiv) were added to this reaction system which was followed by TLC. After
completion, saturated aqueous NH4Cl was added to quench the reaction, which was extracted with ethyl acetate three times (3 ꢀ 10 mL), dried over anhydrous magnesium
sulfate, filtered, concentrated and the residue was chromatographied by silica gel to give the pure diol product.
b
Isolated yield.
c
The ee was determined by HPLC using a chiral stationary phase (Daicel Chiralpak AD-H or AS-H).
yacetone under the optimal protocol. As a result, we obtained the
main disfavored 1,4-diols with high or excellent enantioselectivi-
ties but with moderate or high yields in all substrates. Compared
to the salicylic acid additive, this kind of reaction regioselectivity
was relatively poor. All reactions afforded a certain amount of
the favored 1,2-diols. With regard to the nitro-substituted benzal-

L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
539
Figure 4. The 1H NMR spectrum in DMSO-d6 (A: hydroxyacetone, B: organocatalyst 1b, C: hydroxyacetone + organocatalyst + benzoic acid (1:1:1), D: hydroxyace-
tone + organocatalyst + Al2O3 (1:1:1)).
dehydes, 2-nitrobenzaldehyde resulted in the lowest yield and
regioselectivity compared to 3-nitro and 4-nitrobenzaldehyde
(entries 2 vs 1 and 3). In addition, the reaction gave the approxi-
mate yields of 1,4-diol (yield 55%) and 1,2-diol (yield 37%) when
using the 4-trifluoromethyl benzaldehyde (Table 6, entry 5). For
the 2-CF3–C6H4CHO, the reaction yield and regioselectivity were
better than values obtained for the corresponding 4-substituted
substrate (entries 4 and 5). In terms of unsubstituted aldehyde, a
poorly reactive benzaldehyde, although the enantiomeric excess
was high (up to 96%), the reaction yield was lower than any
electron-withdrawing substituted benzaldehydes with only 41%
for the 1,4-diol and 22% for the 1,2-diol (entry 9).
From the results summarized in Tables 5 and 6, we found that
the reaction methodology had some drawbacks at present. For
instance, the substrates were limited to electron-withdrawing
substituted aromatic aldehydes. For the poorly reactive substrates,
the reaction results, including the regioselectivity and yield as well
as the enantioselectivity were relatively poor. To the best of our
knowledge, current reports on the asymmetric aldol reaction of
hydroxyacetone for the construction of this disfavored 1,4-diol
product are generally restricted to aromatic aldehydes with elec-
tron-withdrawing substituents.5,6 In the case of aluminum oxide
as an additive, the reaction regioselectivity was only moderate
and needs to be further improved upon. Despite the existing limi-
tations, this protocol has advantages, such as the regioselectivity
that can be easily regulated by changing the additives to afford
the usual 1,2-diol or 1,4-diol. Most importantly, the organocatalyst
employed is structurally simple and readily available from the
We attempted to determine the mechanism of the direct asym-
metric aldol reaction of hydroxyacetone and aldehydes, using
several methods. For instance, by using 1H NMR spectroscopy in
two cases (using the salicylic acid and aluminum oxide as addi-
tives) (Fig. 4), we observed that the chemical shift of the ring CH2
(a) adjacent to the pyrrolidinic N atom was downfield shifted to
the place of a0. Moreover, the integral and shape of peak b (CH)
close to the pyrrolidine N atom were changed significantly to the
broad peak b0 in the presence of salicylic acid. This phenomenon
can be ascribed to a strong intermolecular hydrogen-bond interac-
tion between the enol-enamine and the salicylic acid. However, for
the Al2O3 additive, no obvious changes were observed in the 1H
NMR. This may be due to the Al2O3 insolubility in the DMSO-d6.
We then looked at other MS spectra, but no significant results
were obtained. Thus, based on the experimental results and previ-
ous relevant reports,5,8 we proposed a possible reaction mecha-
nism as shown in Scheme 2.
At first, the interaction of organocatalyst 1b and hydroxyace-
tone led to the formation of an imine cation which simultaneously
tautomerizes with the enamine intermediate. The acid additive
should be involved in the iminium–enamine equilibrium in the
asymmetric aldol reaction catalyzed by L-proline or its derivatives,
and thus accelerate the reaction process and affect the enantiose-
lectivity.9 Hence, in the presence of the salicylic acid, the imine cat-
ion preferentially adopts the dominant configuration A to form the
enol-enamine C due to its strong hydrogen bonding interaction
with the prolinethioamide 1b, thus affording the usual 1,2-diol
products. Comparatively, when using Al2O3 as an additive, because
of its steric hinderance, it can affect the imine cation via B to then
give enamine D, which was predicted to be a slightly more favor-
able than the enol-enamine,5 to finally give 1,4-diols. In view of
natural amino acids L-proline and L-valine. Moreover, the reaction
conditions were mild without harsh anhydrous and anaerobic
environments. Hence, the operation is also very simple.

540
L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
S
O
S
OMe
HO
N
H
OMe
N
N
H
O
N
H
O
HO
Al2O3
o-HOC6H4COOH
S
S
OMe
Al3+ Al3+
N
H
O2-
O2-
O2-
OMe
N
N
O
N
H
O
H
O
H
O
B
O
O
H
H
O
A
S
OMe
N
H
N
O
S
OMe
N
OH
D
N
H
O
RCHO
O
OH
C
RCHO
O
OH
*
OH
R
OH
*
1,4-diol product
*
R
OH
1,2-diol product
Scheme 2. A possible mechanism of the regioselective aldol reaction of hydroxylacetone and aromatic aldehyde.
Gong’s report regarding the DFT for the reaction, the difference be-
tween the enol-enamine and enamine is so slight that the revers-
ible transformation between these two isomers results in the
competitive formation of 1,2-diols and 1,4-diols.
developed herein is the first example and of high synthetic
potential.
4. Experimental
4.1. General
3. Conclusion
In conclusion, we have developed a novel, structurally simple
dipeptide derivative 1b, which was readily prepared from
commercially available and inexpensive L-proline and L-valine,
for the asymmetric direct aldol reactions of hydroxyacetone and
aromatic aldehydes. Good yields and diastereoselectivities as well
as high enantioselectivities have been achieved with 20 mol % cat-
alyst loading. The reaction regioselectivity could be easily regu-
lated by altering the additives from salicylic acid to aluminum
oxide and afforded the usual 1,2-diols and 1,4-diols. The method
1H NMR and 13C NMR spectra were recorded on Bruker AV 300
or Varian mercury Plus 400 using CDCl3 as the solvent and TMS as
the internal standard unless otherwise noted; chemical shifts are
given in (ppm) and coupling constants (J) in Hz. The enantiomeric
excess was determined by HPLC using Chiralpak AS-H and AD-H
column (Daicel Chemical Ind., Ltd) with n-hexane and i-propanol
as eluents. Acetone was distilled from K2CO3 prior to use; all other
solvents and commercially available reagents were used as
received without further purification unless otherwise stated.

L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
541
Analytical thin layer chromatography (TLC) was performed using
Acknowledgments
Merck 60 F254 precoated silica gel plate. Subsequent to elution,
the plates were visualized using UV radiation (254 nm) on a
Spectroline Model ENF-24061/F 254 nm.
We are grateful for the financial support from the National Nat-
ural Science Foundation of China (21142009) and Tianjin Natural
Science Foundation of China (12JCQNJC03500) for this work.
4.2. Procedure for the preparation of the organocatalyst
References
To a stirred solution of N-Boc-L-proline (1.25 g, 5.81 mmol) in
1. (a) Hajos, Z. G.; Parrish, D. R. J. Org. Chem. 1974, 39, 1615–1621; (b) Eder, U.;
Sauer, R.; Wiechert, R. Angew. Chem., Int. Ed. 1971, 10, 496–497; (c) List, B.;
Lerner, R. A.; Barbas, C. F., III J. Am. Chem. Soc. 2000, 122, 2395–2396; (d)
Sakthivel, K.; Notz, W.; Bui, T.; Barbas, C. F., III J. Am. Chem. Soc. 2001, 123, 5260–
5267.
CH2Cl2, Et3N (1.2 mL, 8.71 mmol) was added slowly at 0 °C. After
the solution was stirred for 10 min, isobutyl chloroformate
(875
at 0 °C. After the solution was stirred for an additional 20 min at
0 °C, Et3N (90 L, 6.67 mmol) and -valine methyl ester hydrochlo-
lL, 6.67 mmol) was added dropwise to the reaction mixture
2. (a) Dondoni, A.; Massi, A. Angew. Chem., Int. Ed. 2008, 47, 4638–4660; (b) Raj, M.;
Parashari, G. S.; Singha, V. K. Adv. Synth. Catal. 2009, 351, 1284–1288; (c)
Calderon, F.; Fernandez, R.; Sanchez, F.; Fernandez-Mayoralas, A. Adv. Synth.
Catal. 2005, 347, 1395–1403; (d) Tzeng, Z.-H.; Chen, H.-Y.; Reddy, R. J.; Huang,
C.-T.; Chen, K.-M. Tetrahedron 2009, 65, 2879–2888; (e) Da, C.-S.; Che, L.-P.; Guo,
Q.-P.; Wu, F.-C.; Ma, X.; Jia, Y.-N. J. Org. Chem. 2009, 74, 2541–2546; (f)
Paradowska, J.; Rogozinska, M.; Mlynarski, J. Tetrahedron Lett. 2009, 50, 1639–
1641; (g) Yan, J.-C.; Wang, L. Synthesis 2008, 2065–2072; (h) Guillena, G.; Hita,
M. D. C.; Najera, C.; Viozquez, S. F. Tetrahedron: Asymmetry 2007, 18, 2300–2304;
(i) Xu, X.-Y.; Wang, Y.-Z.; Gong, L.-Z. Org. Lett. 2007, 9, 4247–4249; (j) Cordova,
A.; Zou, W.-B.; Dziedzic, P.; Ibrahem, I.; Reyes, E.; Xu, Y.-M. Chem. Eur. J. 2006, 12,
5383–5397; (k) Hayashi, Y.; Sumiya, T.; Takahashi, J.; Gotoh, H.; Urushima, T.;
Shoji, M. Angew. Chem. 2006, 118, 972. Angew. Chem., Int. Ed. 2006, 45, 958–961;
(l) Dziedzic, P.; Zou, W.-B.; Hafren, J.; Cordova, A. Org. Biomol. Chem. 2006, 4, 38–
40; (m) Samanta, S.; Liu, J.-Y.; Dodda, R.; Zhao, C.-G. Org. Lett. 2005, 7, 5321–
5323; (n) Zou, W.-B.; Ibrahem, I.; Dziedzic, P.; Sunden, H.; Cordova, A. Chem.
Commun. 2005, 39, 4946–4948; (o) Cordova, A.; Notz, W.; Barbas, C. F., III Chem.
Commun. 2002, 24, 3024–3025; (p) Notz, W.; List, B. J. Am. Chem. Soc. 2000, 122,
7386–7387; (q) Gruttadauria, M.; Giacalone, F.; Noto, R. Catalytic Methods
Asymmetric Synth. 2011, 83–175; (r) Kano, T.; Song, S.; Kubota, Y. Angew. Chem.,
Int. Ed. 2012, 51, 1191–1194; (s) Giacalone, F.; Gruttadauria, M.; Agrigento, P.
Chem. Soc. Rev. 2012, 41, 2406–2447; (t) Rambo, R. S.; Schneider, R. S.
Tetrahedron: Asymmetry 2010, 21, 2254–2257.
3. (a) Kobayashi, J.; Ishibashi, M. Chem. Rev. 1993, 93, 1753–1769; (b) Chakraborty,
T.; Das, S. Curr. Med. Chem.: Anti Cancer Agents 2001, 1, 131–149; (c) Tsuda, M.;
Kubota, T.; Sakuma, Y.; Kobayashi, J. Chem. Pharm. Bull. 2001, 49, 1366–1367; (d)
Kobayashi, J.; Shimbo, K.; Kubota, T.; Tsuda, M. Pure Appl. Chem. 2003, 75, 337–
342; (e) Kobayashi, J.; Tsuda, M. Nat. Prod. Rep. 2004, 21, 77–93; (f) Suri, J. T.;
Ramachary, D. B.; Barbas, C. F., III Org. Lett. 2005, 7, 1383–1385; (g) Font, D.;
Jimeno, C.; Pericas, M. A. Org. Lett. 2006, 8, 4653–4655; (h) Kumar, A.; Singh, S.;
Kumar, V. Org. Biomol. Chem. 2011, 9, 2731–2742; (i) Sarkar, D.; Harman, K.;
Ghosh, S. Tetrahedron: Asymmetry 2011, 22, 1051; (j) Liu, C.; Dou, X.; Lu, Y. Org.
Lett. 2011, 13, 5248–5251; (k) Wu, X.; Ma, Z.; Ye, Z. Adv. Synth. Catal. 2009, 351,
158–162; (l) Paradowska, J. Tetrahedron Lett. 2009, 50, 1639–1641.
4. (a) List, B.; Pojarliev, P.; Castello, C. Org. Lett. 2001, 3, 573–575; (b) Benaglia, M.;
Cenlentano, G.; Cozzi, F. Adv. Synth. Catal. 2001, 343, 171–173; (c) Maggiotti, V.;
Resmini, M.; Gouverneur, V. Angew. Chem. 2002, 114, 1054–1056. Angew. Chem.,
Int. Ed. 2002, 41, 1012–1016; (d) Liu, H.; Peng, L.; Zhang, T.; Li, Y. New J. Chem.
2003, 27, 1159–1160; (e) Magiotti, V.; Bahmanyar, S.; Reiter, M.; Resmini, M.;
Houk, K. N.; Gouverneur, V. Tetrahedron 2004, 60, 619–632; (f) Pan, Q.; Zou, B.;
Wang, Y.; Ma, D. Org. Lett. 2004, 6, 1009–1012; (g) Storer, R. I.; MacMillan, D. W.
C. Tetrahedron 2004, 60, 7705–7714; (h) Guillena, G.; Hita, M. C. M.; Najera, C.
Tetrahedron: Asymmetry 2006, 17, 1027–1031; (i) Gu, Q.; Wang, X. F.; Wang, L.;
Wu, X.-Y.; Zhou, Q.-L. Tetrahedron: Asymmetry 2006, 17, 1537–1540; (j)
Calderon, F.; Doyaguez, E. G.; Fernandez-Mayoralas, A. J. Org. Chem. 2006, 71,
6258–6261; (k) Ramasastry, S. S. V.; Zhang, H.; Tanaka, F.; Barbas, C. F., III J. Am.
Chem. Soc. 2007, 129, 288–289; (l) Wu, X.-Y.; Jiang, Z.-Q.; Shen, H.-M.; Lu, Y.-X.
Adv. Synth. Catal. 2007, 349, 812–816; (m) Guillena, G.; Hita, M. del C. M.; Najera,
C. Tetrahedron: Asymmetry 2007, 18, 1272–1277; (n) Luo, S.-Z.; Xu, H.; Zhang, L.;
Li, J.-Y.; Cheng, J.-P. Org. Lett. 2008, 10, 653–656; (o) Guillena, G.; Hita, M. D. C.;
Najera, C.; Viozquez, S. F. J. Org. Chem. 2008, 73, 5933–5943; (p) Angelici, G.;
Falgiani, A.; Luppi, G.; Kaptein, B.; Broxterman, Q. B.; Tomasini, C. Synth.
Commun. 2008, 38, 1137–1146; (q) Ramasastry, S. S. V.; Albertshofer, K.; Utsumi,
N.; Tanaka, F.; Barbas, C. F., III; Kucherenko, A. S.; Siyutkin, D. E.; Muraviev, V. O.;
Struchkova, M. I.; Zlotin, S. G. Mendeleev Commun. 2007, 17, 277–278; (r) Wu, X.-
Y.; Ma, Z.-X.; Ye, Z.-Q.; Qian, S.; Zhao, G. Adv. Synth. Catal. 2009, 351, 158–162; (s)
Tang, Z.; Jiang, F.; Cui, X.; Gong, L.-Z.; Mi, A.-Q.; Jiang, Y.-Z.; Wu, Y.-D. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 5755–5760; (t) Notz, W.; Tanaka, F. J.; Barbas, C. F., III
Acc. Chem. Res. 2004, 37, 580–591; (u) Wennemers, H. Chem. Commun. 2011, 47,
12036–12041; (v) Nugent, T. C.; Umar, M. N.; Bibi, A. Org. Biomol. Chem. 2010, 8,
4085–4089; (w) Ban-Caballero, A.; Guillena, G.; Najera, C. Green Chem. 2012, 12,
1599–1606; (x) Li, J.; Luo, S.; Cheng, J. P. J. Org. Chem. 2009, 74, 1747–1750.
5. Tang, Z.; Yang, Z.-H.; Cun, L.-F.; Mi, A.-Q.; Jiang, Y.-Z.; Gong, L.-Z. Org. Lett. 2004,
6, 2285–2287.
l
L
ride (1.10 g, 6.67 mmol) were added to the reaction mixture. The
reaction mixture was then warmed up to room temperature. After
3 h, TLC analysis indicated complete consumption of the starting
material. The reaction was quenched by the addition of 1 M HCl
and diluted with CH2Cl2.The organic phase was washed with
H2O, brine and dried over Na2SO4, filtered, and then concentrated
under reduced pressure. Purification was accomplished by
recrystallization (EtOAc/n-hexane) to give compound (S)-tert-
butyl-2-((S)-1-methoxy-3-methyl-1-oxo-butan-2-ylcarbamoyl)pyr-
rolidine-1-carboxylate (1.66 g, yield 87%, white crystal). Mp
68–70 °C; ½a 2D0
ꢂ
¼ ꢁ102:0 (c 1.0, EtOH); 1H NMR (400 MHz, CDCl3):
d 0.91 (d, 3H, J = 7.0 Hz), 0.94 (d, 3H, J = 6.9 Hz), 1.46 (s, 9H),
1.71–1.80 (m, 3H), 1.85–1.99 (m, 2H), 2.10–2.25 (m, 1H),
2.37–3.47 (m, 2H), 3.72 (s, 3H), 4.36–4.44 (m, 2H).
A mixture of the above compound (657 mg, 2 mmol) and
Lawesson’s reagent (410 mg, 1 mmol) in dry THF was stirred for
2 h at room temperature and then refluxed for 0.5 h under argon
atmosphere. After removal of the solvent under reduced pressure,
the resulting residue was purified by flash chromatography
(hexane/EtOAc) to afford compound (S)-tert-butyl 2-((S)-1-meth-
oxy-3-methyl-1-oxobutan-2-ylcarbamothioyl)pyrrolidine-1-car-
boxylate (448 mg, yield 65%, white solid). 1H NMR (400 MHz,
CDCl3): d 0.92–0.99 (m, 6H), 1.44 (s, 9H), 1.72–1.75 (m, 2H),
1.86–1.87 (m, 2H), 2.01 (br s, 1H), 2.29–2.34 (m, 1H), 3.45–3.50
(m, 2H), 3.75 (s, 3H), 4.68–4.70 (m, 1H), 5.06–5.08 (m, 1H).
The above compound (482 mg, 1.4 mmol) was dissolved in dry
CH2Cl2 (2.8 mL) and then TFA (1.4 mL) and Et3SiH (0.55 mL) were
added. After 2 h, the solvent was removed, diluted with CH2Cl2
and washed with saturated NaHCO3, dried over Na2SO4, filtered,
and then concentrated under reduced pressure to give catalyst
1b (266 mg, yield 78%, yellow oil). ½a D20
¼ ꢁ53:6 (c 1.0, CH2Cl2);
ꢂ
1H NMR (400 MHz, CDCl3): d 0.94–1.00 (dd, 6H, J = 6.9, 10.9 Hz),
1.65–1.78 (m, 2H), 2.00–2.04 (m, 2H), 2.35–2.38 (m, 2H), 3.01–
3.09 (m, 2H), 3.75 (s, 3H), 4.21–4.26 (dd, 1H, J = 3.8, 9.0 Hz),
5.12–5.14 (m, 1H), 10.36 (br s, 1H); 13C NMR (100 MHz, CDCl3): d
205.9, 171.2, 67.9, 61.8, 52.3, 47.5, 34.7, 31.2, 26.1, 18.9, 18.5;
HR-MS (ESI) calcd for
245.1322.
C11H20N2O2S (M+1), 245.1318 found:
4.3. Typical procedure for the aldol reaction of the aldehyde
with hydroxyacetone catalyzed by an organocatalyst
To a mixture of organocatalyst 1b (0.05 mmol) and additives
(0.05 mmol) in DMSO (0.2 mL), hydroxyacetone (111 mg,
1.5 mmol, 6 equiv) was added. After the corresponding mixture
was stirred for 5 min, the aldehyde (0.25 mmol) was added at
0 °C. After 2–4 days, TLC analysis indicated the complete consump-
tion of the starting material, after which the reaction mixture was
quenched with a saturated NH4Cl solution, extracted with EtOAc
and dried over Na2SO4. The crude product was purified by flash sil-
ica gel chromatography (hexane/EtOAc) to afford the aldol
products.
6. (a) Chen, X.-H.; Luo, S.-W.; Tang, Z.; Cun, L.-F.; Mi, A.-Q.; Jiang, Y.-Z.; Gong, L.-Z.
Chem. Eur. J. 2007, 13, 689–701; (b) Wu, F.-C.; Da, C.-S.; Du, Z.-X.; Guo, Q.-P.; Li,
W.-P.; Yi, L.; Jia, Y.-N.; Ma, X. J. Org. Chem. 2009, 74, 4812–4818.
7. (a) Wang, B.; Chen, G. H.; Liu, L. Y.; Chang, W. X.; Li, J. Adv. Synth. Catal. 2009, 351,
2441–2448; (b) Wang, B.; Liu, X. W.; Liu, L. Y.; Chang, W. X.; Li, J. Eur. J. Org.
Chem. 2010, 5951–5954.

542
L. Liu et al. / Tetrahedron: Asymmetry 24 (2013) 533–542
8. (a) Ramasastry, S. S. V.; Zhang, H. L.; Tanaka, F. J.; Barbas, C. F., III J. Am. Chem. Soc.
2007, 129, 288–289; (b) Zhu, M.-K.; Xu, X.-Y.; Gong, L.-Z. Adv. Synth. Catal. 2008,
350, 1390–1396; (c) Allemann, C.; Um, J. M.; Houk, K. N. J. Mol. Catal. A: Chem.
2010, 324, 31–38; (d) Zotova, N.; Broadbelt, L. J.; Armstrong, A.; Blackmond, D. G.
Bioorg. Med. Chem. Lett. 2009, 19, 3934–3937; (e) Schmid, M. B.; Zeitler, K.;
Gschwind, R. M. Angew. Chem., Int. Ed. 2010, 49, 4997–5003; (f) Rankin, K. N.;
Gauld, J. W.; Boyd, R. J. J. Phys. Chem. A 2002, 106, 5155–5159; (g) Bahmanyar, S.;
Houk, K. N. Org. Lett. 2003, 5, 1249–1251; (h) Schmid, M. B.; Zeitler, K.;
Gschwind, R. M. J. Am. Chem. Soc. 2011, 133, 7065–7074.
9. Gryko, D.; Zimnicka, M.; Lipin´ ski, R. J. Org. Chem. 2007, 72, 964–970.