Amine-catalyzed direct aldol reactions of hydroxy- and dihydroxyacetone: Biomimetic synthesis of carbohydrates

Article abstract of DOI:10.1021/jo500860g

This article presents comprehensive studies on the application of primary, secondary, and tertiary amines as efficient organocatalysts for the de novo synthesis of ketoses and deoxyketoses. Mimicking the actions of aldolase enzymes, the synthesis of selected carbohydrates was accomplished in aqueous media by using proline- and serine-based organocatalysts. The presented methodology also provides direct access to unnatural l-carbohydrates from the (S)-glyceraldehyde precursor. Determination of the absolute configuration of all obtained sugars was feasible using a methodology consisting of concerted ECD and VCD spectroscopy.

Full text of DOI:10.1021/jo500860g

Subscriber access provided by Otterbein University
Article
Amine-Catalyzed Direct Aldol Reactions of Hydroxy- and
Dihydroxyacetone: Biomimetic Synthesis of Carbohydrates
Oskar Popik, Monika Pasternak-Suder, Katarzyna Le#niak, Magdalena
Jawiczuk, Marcin Górecki, Jadwiga Frelek, and Jacek Mlynarski
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/jo500860g • Publication Date (Web): 19 May 2014
Downloaded from http://pubs.acs.org on May 25, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Amine-Catalyzed Direct Aldol Reactions of Hydroxy- and Dihydroxyacetone:
Biomimetic Synthesis of Carbohydrates
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†
‡
‡
†
Oskar Popik, Monika PasternakꢀSuder, Katarzyna Leśniak, Magdalena Jawiczuk, Marcin
†
†
,†,‡
Górecki, Jadwiga Frelek, and Jacek Mlynarski*
†
Institute of Organic Chemistry, Polish Academy of Sciences, Kasprzaka 44/52, 01ꢀ224
Warsaw, Poland
‡
Faculty of Chemistry, Jagiellonian University, Ingardena 3, 30ꢀ060 Krakow, Poland
Corresponding Author
*
Eꢀmail: jacek.mlynarski@gmail.com.
Table of Contents/Abstract Graphic
Abstract: This article presents comprehensive studies on application of primary, secondary
and tertiary amines as efficient organocatalysts for the de novo synthesis of ketoses and
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
deoxyketoses. Mimicking aldolase enzymes, synthesis of selected carbohydrates was
accomplished in aqueous media by using prolineꢀ and serineꢀbased organocatalysts. Presented
methodology also provides direct access to unnatural Lꢀcarbohydrates from the (S)ꢀ
glyceraldehyde precursor. Determination of the absolute configuration of all obtained sugars
was feasible using methodology consisting of concerted ECD and VCD spectroscopy.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1. Introduction
Dihydroxyacetone (DHA) is one of the most employed donors in nature playing a key role in
many vital biotransformations. Particularly, its phosphorylated form is an intermediate in
glycolysis and gluconeogenesis. Dihydroxyacetone phosphate (DHAP) participates in an
1
enzyme catalyzed aldol reactions en route to carbohydrates. For example, fructose 1,6ꢀ
biphosphate aldolase (FruA) catalyze in vivo the aldol addition of Dꢀglyceraldehyde 3ꢀ
phosphate (DꢀG3P) and dihydroxyacetone phosphate (DHAP) to give Dꢀfructose 1,6ꢀ
biphosphate (Scheme 1). This aldolase controls stereoselective construction of CꢀC bond with
two synꢀconfigured hydroxyꢀsubstituted stereogenic centers. In contrast, construction of antiꢀ
configured ketohexose (Dꢀtagatose) is controlled in nature by tagatose aldolase (TagA,
Scheme 1).
Scheme 1. Biosynthesis of fructose and tagatose by DHAPꢀdependent aldolases.
2
ACS Paragon Plus Environment

Page 3 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
In general, aldolases can catalyze CꢀC bond formation between dihydroxyacetone phosphate
and glyceraldehyde with simultaneous formation of two new stereocenters. As a consequence,
four different ketohexose stereoisomers can be obtained in short and efficient C
3 3
+C strategy
(Scheme 2). Each aldol reaction generates a single product, whose stereochemistry at Cꢀ3 and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cꢀ4 is complementary to the others. In addition to one (R)ꢀconfigured stereogenic center
delivered from glyceraldehyde substrate, these reactions create two new stereogenic centers in
the form of synꢀ (Dꢀfructose and Dꢀsorbose) or antiꢀketohexoses (Dꢀpsicose and Dꢀtagatose)
2ꢀ
(Scheme 2, R = OPO
3
).
Scheme 2. Stereochemical complementarity of DHAPꢀdependent aldolases.
Different aldolases catalyzing the formation of each of those stereoisomers are commercially
available and may be adopted to the straightforward synthesis of natural carbohydrates and
2
their derivatives. In this manner, synthesis of remaining diastereoisomeric sorbose and
psicose molecules can be accomplished by using rhamnulose and fuculose aldolases,
respectively (Scheme 2). All mentioned enzymes have a broad aldehyde specificity while the
need for the application of dihydroxyacetone phosphate constitutes important limitation for
their synthetic use. Only some aldolase antibodies are capable of using αꢀhydroxylated
3
ketones such as hydroxyacetone as the donor.
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Although chemists developed DHAPꢀdependent aldolases into important tools for the
asymmetric synthesis of carbohydrates, similar straightforward transformations promoted by
small organic catalysts remained an elusive goal for a long time. In contrast to the simplicity
of presented attempts, existing methodologies for the synthesis of carbohydrates typically
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
need several synthetic steps and tedious protecting group manipulation. Interestingly,
versatile and simple C +C strategy utilizing DHA donor although most favored by nature is
3
3
still challenging for organic chemists. A promising strategy for the stereoselective application
of hydroxyꢀ and dihydroxyacetone is associated with the rapid development of
5
organocatalytic methods. In analogy to enzymes, organocatalysis allows for the direct
6
catalytic aldol reaction of aldehydes and ketones without the use of preformed enolates. This
concept allows also for the organocatalytic synthesis of some carbohydrates but the stateꢀofꢀ
7
theꢀart in this area is not well recognized especially for unprotected donors.
In 2000 List and Notz described the first enamineꢀbased enantioselective aldol addition of
8
unprotected hydroxyacetone with several enolizable aliphatic aldehydes. Authors
demonstrated that protected (R)ꢀglyceraldehyde reacted with unprotected hydroxyacetone
under (S)ꢀprolineꢀpromoted direct aldol reaction. DꢀTagatose and Dꢀfructose derivatives were
isolated with moderate diastereoselectivities of only (2:1) and poor overall yield (40%,
Scheme 3). Although unselective, the synthesis of the antiꢀ and synꢀ1ꢀdeoxyhexoses provided
an indication of the utility of these types of asymmetric aldol reactions as applied to
carbohydrate synthesis. According to broadly accepted explanation, the Siꢀface of the
hydroxyacetone enamine attacks the Reꢀface of the aldehyde to give the antiꢀproduct (1ꢀ
deoxytagatose). The enamine double bond is presumed to possess a (E)ꢀconfiguration and
9
observed antiꢀstereoselectivity is consistent with chairlike transition state. Further, enamineꢀ
1
0
11
based organocatalytic antiꢀ and synꢀselective direct aldol reaction of unprotected
hydroxyacetone has been demonstrated for aromatic and nonchiral aliphatic aldehydes
4
ACS Paragon Plus Environment

Page 5 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
whereas the diastereoselective reaction with optically pure glyceraldehyde remained
neglected.
In 2002, Barbas et all. reported for the first time an organocatalyzed cross aldol reaction of
unprotected DHA with acetonide of (R)ꢀglyceraldehyde. The reaction was promoted by (S)ꢀ1ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
(
2ꢀpyrrolidinylmethyl)pyrrolidine in an aqueous phosphate buffer. Authors observed
unselective formation of all four possible stereoisomeric ketohexoses under physiological
conditions by enamine catalysis. Further research also clearly demonstrated that unprotected
1
3
DHA was not an useful substrate for the enamineꢀbased aldol addition (Scheme 3).
Scheme 3. Amineꢀcatalyzed aldol reactions of unprotected hydroxyꢀ and dihydroxyacetone.
In contrast, application of protected DHA molecules has been far more promising in the field
of carbohydrate synthesis. The best results in the direct synthesis of ketohexoses was
described when (S)ꢀproline was used as a catalyst for the reaction of protected DHA
14
derivatives such as 2,2ꢀdimethylꢀ1,3ꢀdioxanꢀ5ꢀone. Since (S)ꢀ and (R)ꢀproline catalysts
13ꢀ15
provided versatile access to antiꢀ1,2ꢀdiols,
they mimic tagatose and fuculose aldolases.
The reaction proceeds through (E)ꢀenamine leading to antiꢀselective aldols.
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
More recently, Luo et al. showed that antiꢀselective reaction of dioxanone with aromatic
aldehydes can be promoted by chiral primary amines, thus confirmed that reaction
stereoselectivity depends on (E)ꢀconfigured enamine formed exclusively from cyclic ketone.
16
Parallel concept of application of primary amineꢀbased organocatalysts to synꢀselective aldol
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
7
reaction of unprotected dihydroxyacetone was presented by Barbas and coꢀworkers. These
synꢀselective reactions of DHA with aromatic aldehydes were carried out in the presence of
tryptophan or threonine derivatives in combination with methyltetrazole.
By using the same principles, Barbas and coꢀworkers reported that primary amino acid
derivatives could catalyze the synꢀaldol reaction of protected DHA with optically pure
glyceraldehyde. OꢀtertꢀButylꢀDꢀthreonine controlled the aldol reaction of tertꢀ
butyldimethylsilyl (TBS)ꢀprotected dihydroxyacetone with (R)ꢀglyceraldehyde to form
1
8
protected Dꢀfructose with high diastereoisomers ratio (98:2). Similarly, OꢀtertꢀbutylꢀLꢀ
threonineꢀbased amide can act as an efficient catalyst for the reaction of TBSꢀprotected DHA
with the acetonide of (R)ꢀglyceraldehyde providing the protected Dꢀsorbose derivative in 86%
1
9
yield and (3:1) syn/anti ratio. Applying water tolerant catalysts mimics Lꢀrhamnulose and Dꢀ
fructose aldolases, respectively.
Despite the fact that many authors claimed to solve the problem of biomimetic reaction of
DHA, presented catalysts were active and selective for protected donors only. Although
stereoselective synthesis of synꢀ and antiꢀaldols from DHA derivatives is well established,
asymmetric reaction of unprotected donors needs further effort, especially for the synthesis of
sugars. To address this problem and since divergent biomimetic access to all ketohexoses by
using enamine organocatalysis and a C ꢀdihydroxyacetoneꢀbased strategy from unprotected
3
donor have been still not possible, we present here our effort toward resolving this
fundamental synthetic problem.
6
ACS Paragon Plus Environment

Page 7 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
This article presents first straightforward de novo synthesis of all 1ꢀdeoxyketohexoses of Dꢀ
and Lꢀseries from hydroxyacetone and to date elusive synꢀselective formation of ketohexoses
from unprotected dihydroxyacetone. In addition to the synthetic value, presented
organocatalysts truly mimic aldolases’ stereoselectivity and mode of action by forming
enamine intermediates in homogeneous aqueous solvents.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2. Results and Discussion
2.1. Reaction of hydroxyacetone (HA) promoted by primary and secondary amine-based
organocatalysts. General lack in stereoselective synthesis of carbohydrates from unprotected
donors inspired us for comprehensive research in this field. We initially studied the reaction
of hydroxyacetone (2) and (R)ꢀglyceraldehyde acetonide (3) as a possible general method for
the de novo synthesis of 1ꢀdeoxyketohexoses. Unselective asymmetric aldol reaction of these
substrates may result in the formation of four differently configured 1ꢀdeoxyketohexoses (4ꢀ7,
Scheme 4). To control the reaction stereoselectively enamine should preferentially attack one
of the aldehyde carbonyl group site. Moreover, enamine formation and its reaction with
aldehyde should be faster than possible formation of the product under general base
mechanism. To achieve high enantioselectivity of the aldol product, diastereoselective control
of the CꢀC bond formation by chiral aldehyde should also be restrained.
Scheme 4. Four possible products of the reaction of hydroxyacetone and (R)ꢀglyceraldehyde.
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
Following results described by List, and to investigate application of both enantiomeric
amino acids in various solvents, we started from reꢀexamination of (S)ꢀprolineꢀcontrolled
aldol reaction. Preferential formation of (E)ꢀenamine between hydroxyacetone and proline
molecule should result in selective formation of antiꢀaldol. Indeed, we confirmed observed
reaction selectivity (Table 1, entry 1). Surprisingly, application of enantiomeric (R)ꢀ
configured catalyst resulted in visible drop in reaction yield thus confirming additional
influence of substrate stereochemistry and confirmed possible formation of
matched/mismatched pairs (Table 1, entry 2).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. Direct aldol reaction of hydroxyacetone 2 with (R)ꢀglyceraldehyde 3 promoted by
secondary amineꢀbased catalysts.
a
b
c
Entry Catalyst
Solvent
Yield [%] dr (4:6)
1
2
3
4
5
6
7
8
9
(S)ꢀproline DMF
(R)ꢀproline DMF
(S)ꢀproline DMSO
(S)ꢀproline THF
(S)ꢀproline DMF/H
39
2.5:1
ꢀ
trace
30
2.5:1
2:1
1:1
1:1
2:1
1:7
45
2
2
O (9:1) 7
pyrrolidine DMF/H
O (9:1) 37
d
(S)ꢀproline
pyrrolidine
(S)ꢀserine
(R)ꢀserine
ꢀ
50
60
8
d
ꢀ
e
THF
THF
2:1
e
10
10
1:2
a
Reactions were performed with (R)ꢀ3 (1 mmol), HA 2 (5 mmol), catalyst (20 molꢀ%) in
b
appropriate solvent (2 mL) at RT for 24 h. Total yield of both synꢀstereoisomers. ꢀ
c
1
d
Determined by H NMR and chiral HPLC analysis. Reactions were performed with (R)ꢀ3 (4
e
mmol), HA 2 (2 mmol), catalyst (20 molꢀ%) in neat at RT for 24 h. dr Refers to syn products
(4:6).
8
ACS Paragon Plus Environment

Page 9 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Next, we turned our attention to the reaction in aqueous solvents closely related to biomimetic
concept. In fact, synthetically useful direct aldol reactions in aqueous media are rare and
20
mostly limited to less demanding donors such as cyclohexanone or acetone. According to
our expectation direct aldol reaction of hydroxyacetone in wet DMF was less efficient and
unselective when compared to the reaction in dry solvents (Table 1, entry 5). Proline was an
extremely poor catalyst of the reaction, providing unselective formation of the aldols as
expected. The addition of water to the mixture completely suppressed stereoselective addition
of enamine to aldehyde, probably because of competitive hydrogen bond formation between
carbonyl group of aldehyde and water instead of catalyst. Interestingly, reaction promoted by
pyrrolidine catalyst resulted in better yield (Table 1, entry 6). Such better reactivity of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
pyrrolidine when compared to proline was previously observed in aqueous solvents. Most
intriguing was however, reaction promoted by pyrrolidine without any solvent (Table 1, entry
8). The reaction afforded synꢀdiol i.e. 1ꢀdeoxyꢀDꢀfructose (6) with high level of
stereoselectivity (syn/anti, 7:1) clearly suggesting different reaction mechanism. We assumed
that reaction selectivity resulted from hydroxyacetone enol formation which reacts with
optically pure glyceraldehyde under FelkinꢀAnh model. Proposed reaction mechanism will be
presented and discussed in the next paragraphs.
Learning from these studies, we attempted to the synthesis of carbohydrates. Based on our
2
earlier success in the development of direct aldol reactions catalyzed by C ꢀsymmetrical
2
1
bisamides in water, we decided to apply this methodology to the elusive synthesis of
ketohexoses. First, we decided to test previously developed organocatalytic mimics of the
aldolases for antiꢀ and synꢀselective aldol reaction of hydroxyacetone with optically pure
2
aldehyde. C ꢀSymmetrical bisprolinamides (8, 9) and lipophilic bissiloxyserinamides (10, 11)
readily available from (S,S)ꢀ, (R,R)ꢀdiphenylethylenediamine and amino acids have been used
as enantioselective organocatalysts for direct aldol reaction of unprotected hydroxyacetone
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with nonꢀchiral aldehydes. Now, we screened all four catalysts in the aldol reaction of
hydroxyacetone with glyceraldehyde acetonide (3) (Scheme 5).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 5. Organocatalysts used in this study for biomimetic direct aldol reaction of hydroxyꢀ
(HA) and dihydroxyacetone (DHA).
Our original design for organocatalysts were both their application for the direct aldol
2
2
reactions in water and expected reaction selectivity. Thus we selected prolineꢀbased
catalysts for the antiꢀaldol reaction based on the presumption that (E)ꢀenamine formation
controlled by the catalysts 8 and 9 will result in the formation of antiꢀaldols. In contrast,
design of organocatalysts 10 and 11 for the synꢀaldols was based on the intermediacy of a (Z)ꢀ
enamine in the transition state. According to this concept, we studied direct aldol reaction of
hydroxyacetone (2) with (R)ꢀglyceraldehyde acetonide (3) catalyzed by enantiomeric amides
21
containing secondaryꢀ (8, 9) and primary amine functions (10, 11, Scheme 6). Catalysts
design and unsuccessful application of simple amides as well as other derivatives of amino
21ꢀ23
acids has been tested and discussed in our previous works.
In this paper, we present only
successful application of the most reactive and stereoselective catalysts.
10
ACS Paragon Plus Environment

Page 11 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
OH
O
cat. 8
20 mol-%)
(
O
O
OH
1
-deoxy-D-tagatose (5)
yield 67%, dr (89:11-syn-6)
OH
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
cat. 9
(
20 mol-%)
O
O
OH
O
O
OH
1-deoxy-D-psicose (7)
yield 65%, dr (85:15-syn-6)
+
O
H
O
2
(HA)
(R)-3
0.5 mmol)
OH
O
(
cat. 10
(
(
20 mol-%)
O
O
OH
1
-deoxy-D-sorbose (4)
yield 46%, dr (80:20-syn-6)
OH
O
cat. 11
20 mol-%)
O
O
OH
1
-deoxy-D-fructose (6)
yield 47%, dr (100:0)ꢀ
Scheme 6. Stereoselective synthesis of 1ꢀdeoxyꢀDꢀketohexoses. Reactions were performed
with (R)ꢀ3 (0.5 mmol), catalyst (20 molꢀ%) in a mixture of THF/HA/water (1:9:1, 2 mL) at
RT for 24 h.
2
1
Using optimal conditions elaborated previously for the reaction of nonꢀchiral aldehydes, and
performing reaction at RT, we explored the scope of the reaction in the synthesis of
deoxyketoses. To our delight, each of four selected catalysts perfectly controlled
stereoselective formation of expected sugar with a good yield. First experiments showed that
the reaction controlled by catalyst 8 composed of (S,S)ꢀdiphenylethylenediamine and Lꢀ
proline, resulted in a formation of protected 1ꢀdeoxyꢀDꢀtagtose (5) in a good yield and (89:11)
antiꢀfavored dr when 20 molꢀ% of organocatalyst was used (Scheme 6). Interestingly,
enantiomeric prolineꢀbased catalyst 9 delivered protected 1ꢀdeoxyꢀDꢀpsicose (7) in good yield
and high stereoselectivity (85:15, Scheme 6). Thus, according to our expectation (S)ꢀproline
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
controls enantioselective formation of (4S)ꢀconfigured 3,4ꢀantiꢀaldol while enantiomeric (R)ꢀ
prolineꢀbased catalyst 9 delivered (4R)ꢀconfigured 3,4ꢀantiꢀaldol. Application of the matched
catalyst and aldehyde acceptor was essential for the formation of expected sugars.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Based on the same principles, formation of the (4S)ꢀconfigured 3,4ꢀsynꢀaldol was observed
for the catalyst 10 composed of Lꢀserine. Catalyst 10 provided 1ꢀdeoxyꢀDꢀsorbose (4) with
good chemical yield maintaining essentially the same diastereoselectivity (80:20) compared to
prolineꢀbased catalysts. Finally, reaction controlled by catalyst 11 resulted in exclusive
formation of 1ꢀdeoxyꢀDꢀfructose (6).
It is important to mention that we have not observed any racemisation of optically pure
glyceraldehyde nor hexoses during the reaction, and high enantiomeric excess of all products
have been confirmed by NMR experiments and chiral HPLC analysis. In fact, enantiomeric
excess of the resulting hexoses simply reflects the ee of starting material. Presented
experiments confirm that elaborated methodology provides a practical and direct route to
deoxyketohexoses by using organocatalysts mimicking all four DHAP dependent aldolase
enzymes (Scheme 2).
Encouraged by these results we decided to investigate possible synthesis of Lꢀsugars by using
developed methodology. This would be valuable achievement in the light of rare and tedious
24
synthesis of this family of compounds. To achieve this goal, the reaction of hydroxyacetone
and (S)ꢀglyceraldehyde precursor promoted by the same catalysts have been tested (Scheme
7).
12
ACS Paragon Plus Environment

Page 13 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 7. Stereoselective synthesis of 1ꢀdeoxyꢀLꢀketohexoses. Reactions were performed
with (S)ꢀ12 (0.5 mmol), catalyst (20 molꢀ%) in a mixture of THF/HA/water (1:9:1, 2 mL) at
RT for 24 h.
To our delight, the reactions proceeded smoothly and aldol products (13ꢀ16) were isolated in
good yields and high diastereoselectivities. Presented aminoꢀacid catalysis provides a new
entry to a oneꢀstep de novo synthesis of 1ꢀdeoxyꢀLꢀpsicose (65%), 1ꢀdeoxyꢀLꢀtagatose (68%),
1ꢀdeoxyꢀLꢀfructose (52%) and 1ꢀdeoxyꢀLꢀsorbose (49%). Application of (S)ꢀconfigured
aldehyde additionally confirmed that catalysts control enantioselective formation of Cꢀ4
stereogenic center with additional control of synꢀ or antiꢀconfigured aldols (Scheme 7). This
observation finally proved that the reaction proceeds via enamine formation which, in turn
may react preferentially with one site of the aldehyde molecule.
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The stereochemistry of presented organocatalytic carbohydrate synthesis are in accordance
21
with previously reported principles for prolineꢀ and serine based aldol reactions. In the
crossꢀaldol reaction, the Siꢀface of the (Z)ꢀenamine intermediate formed from catalyst 10 and
HA approaches the Reꢀface of the acceptor aldehyde furnishing the desired (3R,4S)ꢀsynꢀaldol
adduct as illustrated in Scheme 8. High reaction enantioselectivity observed for the reaction of
4ꢀnitrobenzaldehyde (86% ee) confirmed that asymmetric induction is controlled by enamineꢀ
based organocatalyst 10.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 8. Possible structures of transition states leading to synꢀ and antiꢀaldols.
Formation of an enamine in the reaction mixture was previously confirmed by high resolution
MS spectra: signal at m/z 919 matches expected molecular weight. Isotopic pattern of this
25
signal corresponds with calculated pattern of expected enamine structure. Organocatalyst 10
can form hydrogenꢀbondꢀstabilized (Z)ꢀenamine. In contrast, preferential formation of (E)ꢀ
enamine between prolineꢀbased catalys 8 resulted in preferential formation of antiꢀaldols
(Scheme 8).
14
ACS Paragon Plus Environment

Page 15 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
2.2. Reaction of dihydroxyacetone (DHA) promoted by primary amine-based
organocatalysts. In contrast to unexplored field of direct aldol reaction of hydroxyacetone its
DHA variant has been well recognized although only for protected donors. Particularly
successful development of diastereoꢀ and enantioselective organocatalytic aldol reaction with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
,2ꢀdimethylꢀ1,3ꢀdioxanꢀ5ꢀone as a ketone equivalent was shown as practical tool for the
15
synthesis of antiꢀconfigured carbohydrates mimicking tagatose and fuculose aldolases.
Among screened catalysts, enantiomeric prolines proved to be superior in the construction of
various carbohydrate scaffolds. Less studied synꢀselective direct aldol reaction of silyloxyꢀ
protected DHA provided a direct route to aldol products of the type synthesized with the
18,19
DHAP aldolase enzymes Lꢀrhamnulose 1ꢀphosphate and Dꢀfructose 1,6ꢀdiphosphate.
3 3
To scope the synthesis of ketohexoses from DHA by using direct C +C protocol we tested
2
3
the same catalysts containing serine motifs (10, 11, Scheme 5). In our efforts, we focused
mostly on the synꢀselective reaction as the antiꢀselective variant has been well explored by
other authors. Initial tests for prolineꢀbased catalysts 8 and 9 demonstrated also their
ineffectiveness, especially in terms of stereoselectivity. Having in hand efficient enantiomeric
catalysts 10 and 11 we tested reaction of both enantiomeric glyceraldehyde acetonides (R)ꢀ3
23
and (S)ꢀ12 possibly giving short and elegant entries to Dꢀ and Lꢀhexoses, respectively. As
indicated in Scheme 9 dihydroxyacetone and glyceraldehyde were suitable substrates when
the reaction was conducted in wet DMF at ambient temperature. Application of dry DMF or
increasing of amount of water in the solvent resulted in the decreasing reaction yield and
stereoselectivity.
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 9. Stereoselective synthesis of Dꢀ and Lꢀketohexoses. Reactions were performed with
(R)ꢀ3 or (S)ꢀ12 (1 mmol), DHA (2 mmol of a dimer), catalyst (20 molꢀ%) in a mixture of
DMF/water (9:1, 1 mL) at RT for 24 h.
In the case of optically pure (R)ꢀ and (S)ꢀaldehyde application of appropriate Dꢀ or Lꢀserineꢀ
based catalyst resulted in a clean and selective formation of expected aldols as a result of
formation of a matched pair. In the case of (R)ꢀglyceraldehyde, application of catalyst 11
resulted in a formation of Dꢀfructose in high yield (75%) and dr (95:5). In contrast, catalyst 10
may be used for stereoselective formation of Dꢀsorbose (19) with good diastereoisomers ratio
(7:3). Interestingly, reactions performed in dry DMF were less efficient and delivered aldols
in lower yield (ca. 20%). Starting from (S)ꢀglyceraldehyde, protected Lꢀfructose 20 was
formed under 10ꢀpromoted reaction with high yield (82%) and high stereoselectivity (95:5)
while efficient synthesis of Lꢀsorbose derivative 21 was achieved via the matched chiral pair
between catalyst 11 and (S)ꢀglyceraldehyde substrate (70%, Scheme 9).
16
ACS Paragon Plus Environment

Page 17 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
The plausible transition state for the elaborated synꢀselective aldol reaction of unprotected
DHA is based on the assumption that hydroxyacetone enamine attacks aldehyde
enantioselectively. This mechanism was additionally confirmed via enantioselective direct
aldol reaction of DHA to 4ꢀnitrobenzaldehyde in an aqueous solvent (Scheme 9). An
important feature for synꢀselective transition state leading to (3R,4S)ꢀaldol is the hydrogen
bond supported (Z)ꢀenamine formation presented in Scheme 10.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 10. Possible structure of transition state in synꢀselective aldol reactions promoted by
primary amineꢀbased catalyst 10.
2.3. Reaction of dihydroxyacetone (DHA) promoted by tertiary amines. Previously, we
have shown that addition of hydroxyꢀ and dihydroxyacetone to isopropylidene glyceraldehyde
may result in controlled formation of aldol adducts under enamineꢀcatalyzed aldol reaction. It
was interesting, however, that reaction mixtures was contaminated by the same synꢀaldol, in
all cases. Formation of 1ꢀdeoxyꢀDꢀfructose (Scheme 6) or Dꢀfructose (Scheme 9) was
obviously an alternative reaction pathway leading to the same product despite of the catalyst
used. To solve this interesting problem and venturing into reaction mechanism we
investigated the reaction of dihydroxyacetone with and (R)ꢀglyceraldehyde acetonide (3)
promoted by various achiral amines (Scheme 11). All performed experiments confirmed
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
exclusive formation of synꢀaldols suggesting one common mechanism for all kind of tested
amines.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 11. synꢀSelective aldol reaction of DHA promoted by tertiary amines.
Table 2 shows that reactions promoted by representative examples of tertiary (entries 1ꢀ3),
secondary (entries 4ꢀ5) and even primary amines (entry 6) resulted in preferential formation
of Dꢀfructose from (R)ꢀglyceraldehyde. Similar stereoselectivity favoring formation of natural
fructose was observed for reactions carried out in wet DMF (entry 1). Previously, in Table 1
we also showed that application of pyrrolidine instead of proline for the reaction of
hydroxyacetone in neat switched the reaction selectivity favoring formation of synꢀaldol
(Table 1, entries 7, 8). This tendency is also visible for dihydroxyacetone donor (Table 2,
entry 4).
Table 2. Direct aldol reaction of dihydroxyacetone 1 with (R)ꢀglyceraldehyde 3 promoted by
tertiary amines.
a
b
c
Entry Catalyst
Solvent
DMF/H
Yield [%] dr (18:19)
1
2
3
4
5
6
7
8
DBU
DBU
2
O (9:1) 46
trace
O (9:1) 30
80:20
ꢀ
CHCl
3
triethylamine DMF/H
2
75:25
75:25
75:25
85:15
ꢀ
pyrrolidine
piperidine
DMF/H O (9:1) 26
2
DMF/H O (9:1) 16
2
benzylamine DMF/H
2
O (9:1) 55
O (9:1) trace
48
quinine
quinine
DMF/H
CHCl
2
3
80:20
18
ACS Paragon Plus Environment

Page 19 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
9
1
quinidine
CHCl
3
45
46
54
80:20
75:25
75:25
0
1
cinchonine
CHCl3
1
cinchonidine CHCl3
a
Reactions were performed with (R)ꢀ3 (0.5 mmol), DHA 1 (2 mmol as a dimer), catalyst (20
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b
molꢀ%) in CHCl
3
(1 mL) at RT for 24 h or in DMF/water (9:1, 1mL) at RT for 72 h. Total
c
1
yield of synꢀisomers. Determined by H NMR and chiral HPLC analysis.
Formation of synꢀaldols from hydroxyketones is in full accordance with previously published
results however it needs additional comments and should be seen as an important competitive
reaction en route to enamineꢀcontrolled formation of aldols from hydroxyketones. Our
8
predecessors observed also unexpected synꢀselectivity in the reaction of unprotected HA and
1
2
DHA controlled by prolineꢀbased catalysts (Scheme 3). Assuming that amine catalysts can
simply act as a base, formation of enol from hydroxyketone may be an important threat to the
stereoselective addition of expected enamine to aldehyde. Thus formation of protected Dꢀ
fructose most likely goes through general base mechanism instead of enamineꢀcontrolled
reaction. Especially in the case of (S)ꢀ1ꢀ(2ꢀpyrrolidinylmethyl)ꢀpyrrolidine used by Barbas,
formation of fructose may have resulted from competitive enol formation promoted by
12
secondaryꢀprimary amine catalyst (Scheme 3).
However, the use of tertiary amines in the reaction requires clarification. In the mid twentieth
26
century, Gutsche and coꢀworkers demonstrated that in addition to hydroxide ion, tertiary
amines are an effective catalyst for unselective addition of glyceraldehyde and
27
dihydroxyacetone. In 2007, Mahrwald showed examples of tertiaryꢀamine promote synꢀ
28
selective aldol reactions. According to authors, DBUꢀcatalyzed aldol addition of DHA to 3
resulted in unselective formation of synꢀaldols (fructose and sorbose). In contrast, reaction
promoted by cinchonine resulted in extremely high stereoselectivity of the fructose formation.
2
9
In our hand, reaction promoted by either DBU (Table 2, entry 1) or cinchona alkaloids
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(Table 2, entries 7ꢀ11) resulted in the formation of fructose and sorbose with the same level of
stereoselectivity (ca. 4:1) suggesting the same mechanism for both reactions.
Nevertheless, some level of enantioselectivity provided by cinchona alkaloids may also be
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
considered by using alternative reaction mechanism. Previously, we discovered that
30
unmodified cinchona alkaloids work effectively with hydroxyacetone and aromatic
hydroxyketones as highly synꢀselective aldol catalysts providing aldols from nonꢀchiral
31
aldehydes with good enantioselectivities up to 60ꢀ70%. To investigate possible
enantioselective aldol reaction between DHA and nonꢀchiral aldehydes we used 4ꢀ
nitrobenzaldehyde as a case study (Scheme 12). Application of 20 molꢀ% of tertiary amines in
wet DMF at RT resulted in preferential formation of synꢀaldol 22 in good yields ranging 40%
for DBU and 61% for cinchonine catalysts. Reaction yield for more reactive aromatic
aldehyde was far better than in case of glyceraldehyde (Table 2, entry 7). However,
application of cinchonine in aqueous solvent resulted in racemic aldol 22. Further careful
3
optimization revealed that reaction promoted by 20 molꢀ% of cinchonine in CHCl at RT
resulted in formation of aldol 22 in good synꢀselectivity (4.4:1) and poor enantioselectivity
(28% ee). Commenting on the correlation between structure of the catalyst and resulting aldol
it should be noted that application of CN and ‘local pseudoenantiomeric’ QN delivers
expected enantiomeric aldols (Scheme 12).
20
ACS Paragon Plus Environment

Page 21 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 12. synꢀSelective aldol reaction of DHA with 4ꢀnitrobenzaldehyde promoted by
tertiary amines.
Taking into account all observations we postulate that the reaction promoted by tested tertiary
amines proceeds more likely by enol formation and their subsequent stereoselective addition
to optically pure aldehyde. Considering FelkinꢀAnh transition state oxygen at Cꢀ2 of
aldehyde, being electronegative, will lie perpendicular to the carbonyl group in the most
reactive conformer (Scheme 13). This reaction delivers Dꢀfructose as a result of least hindered
direction of attack of enolate on (R)ꢀ3. This explanation supports common formation of
fructose under general base mechanism.
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 13. Possible structures of transition states in the reaction promoted by tertiary amines
(upper) and asymmetric aldol reaction promoted by cinchona alkaloids (lower).
While in all cases reaction of a ketone involves initial deprotonation by the amine catalyst the
high synꢀselectivity observed for cinchona alkaloids in aprotic solvents could be explained by
(Z)ꢀenol formation from the hydroxyketone which then attacks the Si face of the aldehyde.
Reaction at the Re face of the aldehyde is more difficult due to the steric repulsion between
3
1
the two larger substituents.
Reaction of primary, secondary and tertiary amines could be explained in a similar way. The
reaction of glyceraldehyde with inorganic bases in aqueous solution to yield fructose and
2
7
sorbose has been quite wellꢀstudied although it cannot be used for the sugar synthesis
described above because of at least partial racemization of optically pure glyceraldehyde.
22
ACS Paragon Plus Environment

Page 23 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
2.4. Determination of the structure of sugars. Confirmation of the structures of resulting
sugars was not trivial and constitutes additional problem. Among all presented deoxyhexoses
only structures of protected 1ꢀdeoxyꢀDꢀpsicose and 1ꢀdeoxyꢀDꢀtagatose have been previously
described in the literature but presented low resolution NMR (80 MHz) spectra were not
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
2
useful for unambiguous determination of their complex structures. Determination of the
structures of protected fructose also required more attention especially in the light of difficult
12
distinction of two diastereoisomeric synꢀaldols formed in the reaction. Simply deprotection
and cyclisation resulted in a complex mixture of anomers existing in both furanose and
pyranose forms. To conclude this subject and undoubtedly confirm absolute configuration of
aldols we decided to use CD techniques allowing for insight into structures of unmodified
sugars.
To determine the absolute configuration (AC) of aldols (diols) 4ꢀ7 and 18ꢀ21 the soꢀcalled in
situ methodology of electronic circular dichroism spectroscopy (ECD) with dimolybdenum
3
3
tetraacetate acting as auxiliary chromophore has been used. Recently, this methodology has
gained increasingly widespread application in solving stereochemical problems of transparent
molecules. This can be evidenced by its escalated use for threeꢀdimensional structure
3
4
determination of 1,2ꢀdiols. Certainly, the evidenced escalation of the in situ method is due to
its simplicity consisting of nothing less than mixing a chiral ligand with an achiral auxiliary
chromophore and recording the spectra. The exchange of ligand(s) results in transferring the
chirality of ligand to the chiral complex, which is formed in solution. Application of the in
situ method involves linking a positive/negative sign of the Cotton effects (CEs) occurring in
the 300ꢀ400 nm spectral range in the spectra of resultant complexes with the positive/negative
OꢀCꢀCꢀO torsion angle of the diol unit. This relationship, called the helicity rule, allows
3
3,35
assignment of AC of vicꢀdiols with confidence through this rule.
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In the current case, however, the in situ methodology had to be modified due to strongly
overlapping carbonyl bands with the bands of resultant Mo ꢀcomplexes at around 300 nm.
2
Workaround the issue was made by subtracting from the complex spectrum of free ligand
recorded under the same measurement conditions. The ECD data for diols 4ꢀ7 and 18ꢀ21
resulting after such treatment are collected in Table 3.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
Table 3. Difference ECD data for the inꢀsitu formed Mo ꢀcomplexes of vicꢀdiols 4ꢀ7 and 18ꢀ
*
21. Values are given as ꢀε′(nm).
ECD ꢀε (λmax
)
comp. band I
+0.08
band II
band III
band IV
band V
ꢀ
A
+
B
+
4
5
6
7
1
1
2
2
+0.96
(312.5)
ꢀ0.05 (353.5) +0.22
(395.5)
(
280.0)
b
+0.26
+
ꢀ0.38 (344.5) +0.19
(428.5)
ꢀ0.07 (542.5) +/ꢀ ꢀ
(276.0)
b
+1.06
ꢀ1.69 (312.0)
+
ꢀ0.78 (371.0)
ꢀ
ꢀ
ꢀ
(277.5)
a
ꢀ0.08 (274.0) +0.54
ꢀ
+0.14
(378.5)
ꢀ0.06 (429.5) +/ꢀ +
(312.0)
a
8
9
0
1
+0.20
277.5)
ꢀ
+0.28
(315.0)
ꢀ0.82 (372.5) +0.31
(459.0)
ꢀ
ꢀ
(
a
ꢀ0.46 (262.0) +0.08
ꢀ
+0.02
(355.5)
ꢀ0.13 (421.5)
+
+
ꢀ
+
+
ꢀ
(308.0)
b
ꢀ0.22 (277.0)
+0.09
+
ꢀ0.15 (319.0) +0.63
373.0)
ꢀ0.25 (453.5)
(
ꢀ0.13 (307.0) +0.19
(336.0)
ꢀ0.27 (382.0) +0.13
(462.5)
(273.0)
A – predicted sign of the O–C–C–O torsion angle. B – sign of the O–C–C–O torsion angle
a
b
*
from ECD; Positive minimum. Negative minimum. Since the real complex structure as well
as the concentration of the chiral complex formed in solution is not known the ECD data are
presented as the artificial ꢀε′ values which are calculated in the usual way as ꢀε′ = ꢀA/c×d,
where c is the molar concentration of the chiral ligand, assuming 100%complexation.
24
ACS Paragon Plus Environment

Page 25 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Based on the positive signs of CEs at ~320 and ~380 nm the (3R,4R) AC was assigned to
diols 4, 19, and 20. Again in accordance with helicity rule and negative signs of CEs in the
same spectral region (3S,4S) AC was assigned to diols 6, 18, and 21. Sets of ECD spectra for
all diols are presented in the Supporting Information material.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A relatively more complex situation is observed for erythro diols 5 and 7. In their case, there
2
are two possible conformations of the diol unit after ligation to the Mo ꢀcore. This is because,
in contrast to the threo diols, in erythro 1,2ꢀdiols two OꢀCꢀCꢀR groups cannot adopt an
antiperiplanar conformation simultaneously. This leads to two possible arrangements of the
diol unit characterized by opposite signs of decisive CEs for the same absolute configuration
3
6
as shown in Scheme14, chart B. However, based on the previous statement, that flexible
erythro vicꢀdiols fall also under the helicity rule and follow signs of crucial CEs in the spectra
of their Mo ꢀcomplexes the (3S,4R) and (3R,4S) AC was assigned to the diols 5 and 7,
2
respectively.
In an effort to corroborate the conclusions made on the basis of the in situ dimolybdenum
method, the vibrational circular dichroism (VCD) spectra were performed. This was
particularly important for erythro diols as calculated Gibbs free energy difference between
conformers was relatively small. Thus, ECD spectroscopy alone is not sufficient for a reliable
AC determination here and independent confirmation of the results significantly increases the
confidence level of assignment.
The experimental and population weighted IR and VCD spectra for diol 7 are summarized in
Scheme 14 charts C,D. As one can see, the overall agreement of the predicted and
experimental spectra for this diol is excellent, and the confidence level of (3R,4S) AC
TM
37
assignment is equal 100% according to the CompareVOA program. The VCD results for
remaining diols collected in the Supporting Information section positively verify the absolute
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
configuration preꢀassignment made by means of the in situ dimolybdenum method. Thus, one
can conclude that the assignment of absolute configuration of diols under study was done in a
reliable manner.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
Scheme 14. A: ECD data of diol 7: navy blue line – chiral Mo ꢀcomplex, green line – free
ligand, purple line – difference spectrum (complex spectrum minus free ligand spectrum); B:
Two possible arrangements of the erythro 1,2ꢀdiol unit in the chiral complex formed after
complexation with the Mo2ꢀdimer; C: VCD data of diol 7: navy blue line – experimental
spectrum, purple line – population weighted spectrum for (3R,4S,5R)ꢀisomer, gray line –
noise; D: IR data of diol 7: navy blue line – experimental spectrum, purple line – population
weighted spectrum for (3R,4S,5R)ꢀisomer.
3. Conclusion
In summary, we disclosed short and elegant de novo synthesis of deoxyketoses and ketoses by
amino acidꢀcatalyzed bond forming reactions with hydroxyacetone and dihydroxyacetone,
respectively. Prolineꢀ and serineꢀbased amides proved to be excellent catalysts in term of
yield and stereoselectivity in the construction of various carbohydrates in aqueous solvents.
Presented enamineꢀbased C +C methodology provided direct entry the synthesis of various
3
3
26
ACS Paragon Plus Environment

Page 27 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Dꢀ and Lꢀketohexoses. Its synthetic potential have been demonstrated in the direct biomimetic
synthesis of a series of ketohexoses and deoxyketohexoses.
In this study (S)ꢀprolineꢀbased catalyst act as an organocatalytic mimic of tagatose aldolase,
whereas its (R)ꢀconfigured enantiomer can be regarded as a mimic of fuculose aldolase.
Application of enantiomeric serineꢀbased organocatalysts delivered the product with the
stereoselectivity similar to that controlled by rhamnulose and fructose aldolases in nature.
These enamineꢀbased synꢀ and antiꢀaldol additions represent a very easy and elegant approach
to ketohexoses with good yields and high degree of stereoselectivity. Based on the elaborated
methodology four deoxyketohexoses: Dꢀdeoxyfructose, Dꢀdeoxysorbose, Dꢀdeoxypsicose and
Dꢀdeoxytagatose have been obtained from hydroxyacetone and (R)ꢀglyceraldehyde in an
asymmetric manner. The same set of catalysts allows for the formation of Lꢀdeoxyhexoses
from hydroxyacetone and (S)ꢀglyceraldehyde.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Furthermore, stereoselective addition of dihydroxyacetone to optically pure glyceraldehyde
has been demonstrated to operate under primaryꢀ, secondaryꢀ and tertiary amine control. This
reaction cannot be compared with the aldol addition catalyzed by amino acids with regards to
the reaction mechanism although it results in a stereoselective formation of synꢀconfigured Dꢀ
fructose as a result of the FelkinꢀAnh mechanism. Enantioselective version of the reaction
promoted by cinchona alkaloids seems to be also possible, and we believe that the procedure
described above will influence the development of asymmetric synthesis and will lead to the
discovery of more selective tertiary amineꢀbased organocatalysts.
Structures of all ketohexoses have been unambiguously established by using CD
spectroscopy. The results of combined in situ method and VCD spectra analysis of all aldols
allowed one to assign threeꢀdimensional molecular structure of all deoxyhexoses and hexoses
with great certainty. Thus, the methodology consisting of concerted ECD and VCD
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
spectroscopy to solve structural and stereochemical problems has demonstrated its
possibilities and excellent effectiveness in the case of this class of compounds, for the first
time.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4. Experimental Section
General Information. All starting materials and reagents were obtained from commercial
sources and used as received unless otherwise noted. All solvents used were freshly distilled
prior to use. Optical rotations were measured at room temperature with a polarimeter. Highꢀ
resolution mass spectra were acquired using electrospray ionization mode with a timeꢀofꢀ
flight detector. Infrared (IR) spectra were recorded on a Fourier transform infrared (FTꢀIR)
1
spectrometer as either a thin film on a NaCl plate (film) or as an KBr pellet (KBr). H NMR
spectra were recorded on spectrometers operating at 300, 400, 500 and 600 MHz in CDCl
3
or
acetoneꢀd . Data were reported as follows: chemical shifts in parts per million (ppm) from
6
tetramethylsilane as an internal standard, integration, multiplicity (s = singlet, d = doublet, t =
triplet, q = quartet, dd = doubleꢀdoublet, m = multiplet, br = broad), coupling constants (in
13
Hz), and assignment. C NMR spectra were measured at 75, 100, 125 or 150 MHz with
complete proton decoupling. Chemical shifts were reported in ppm from the residual solvent
as an internal standard. Reactions were controlled using TLC on silica [aluꢀplates (0.2 mm)].
Plates were visualized with UV light (254 nm) and by treatment with: ethanolic pꢀ
anisaldehyde with sulfuric and glacial acetic acid followed by heating, aqueous cerium(IV)
sulfate solution with molybdic and sulfuric acid followed by heating or ethanolic ninhydrin
solution followed by heating. All organic solutions were dried over anhydrous sodium sulfate.
Reaction products were purified by flash chromatography using silica gel 60 (240ꢀ400 mesh).
HPLC analysis were performed on HPLC system equipped with chiral stationary phase
columns, detection at 254 nm.
28
ACS Paragon Plus Environment

Page 29 of 41
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
(R)ꢀGlyceraldehyde acetonide ((R)ꢀ3) was prepared from Dꢀmannitol according to procedure
38
found in the literature. (S)ꢀGlyceraldehyde acetonide ((S)ꢀ12) was prepared from Lꢀascorbic
39
acid according to procedure found in the literature.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
21,23,40
Synthesis and spectroscopic data for organocatalysts 8ꢀ11 were described previously.
General Methods for CD measurement and calculations. The ECD measurements of in
situ formed chiral Mo
2
ꢀcomplexes of diols (1.2–3 mg of ligand, ca. 0.003 M) with stock
complex [Mo (O CCH ) ] (2.8ꢀ4 mg, ca. 0.002 M) were dissolved in DMSO (2.5ꢀ5 mL) so
2
2
3 4
that the molar ratio of the stock complex to ligand was about (1.5:1), in general. The
measurement parameters 0.5 nm/step with an integration time of 0.25 s over the range 245–
8
00 nm with 200 nm/min scan speed. For difference spectra the ECD measurements of diols
with concentration ca. 0.001 M were acquired at room temperature in DMSO (for UVꢀ
Spectroscopy) on a spectropolarimeters and were collected with the same parameters as for
chiral complexes. UV–Vis spectra were measured in DMSO (for UVꢀSpectroscopy). The
VCD and IR spectra of compounds 4ꢀ7 and 18ꢀ21 were measured by VCD spectrometer at a
–1
resolution of 4 cm using CD
3
CN. FTꢀVCD spectrometer was equipped with dual sources
cell with a
and dual PEM’s technology. Solutions (0.25ꢀ0.39 M) were measured in a BaF
2
path length of 102 ꢁm assembled in rotating holder (14 s/cycle) at room temperature. The
–1
ZnSe photo elastic modulator of the instrument was set to 1400 cm . To improve the S/N
ratio, the spectra were measured between 6 and 24 h. Baseline correction was achieved by
3
subtracting the spectrum of a reference CD CN obtained under the same conditions.
41
Computational details. The conformational search was performed with ComputeVOA
using MMF94 force field within 5 kcal/mol energy ranges. The further optimization was
carried out at DFT level using metaꢀhybrid B3LYP functional and the augꢀccꢀpVDZ basis set.
42
The polarizable continuum model (PCM) implemented in the Gaussian09 was applied for
29
ACS Paragon Plus Environment