Formation of 3-hexuloses in aldol reactions, analysis of the products as their O-isopropylidene derivatives by GC-MS

Article abstract of DOI:10.1016/j.carres.2004.06.018

A method for analysis of mixtures of 3-hexuloses by gas chromatography mass spectrometry of their di-O-isopropylidene derivatives has been elaborated. The origin of characteristic fragment ions in the mass spectra is suggested on the basis of the spectra of d₁₂ analogues, obtained by acetonation with acetone-d₆ and on MS/MS investigations. The method has been applied to product mixtures from aldol reactions between glycero-tetrulose and glycolaldehyde and between 2-pentuloses and formaldehyde. An interesting result is the formation of ribo-3-hexulose with a high degree of stereoselectivity in alkali catalysed reaction between erythro-2-pentulose and formaldehyde.

Full text of DOI:10.1016/j.carres.2004.06.018

Carbohydrate
RESEARCH
Carbohydrate Research 339 (2004) 2171–2176
Formation of 3-hexuloses in aldol reactions, analysis of the
products as their O-isopropylidene derivatives by GC–MS
Dag Ekeberg and Svein Morgenlie*
Department of Chemistry, Biotechnology and Food Science, Section Chemistry, Agricultural University of Norway,
˚
PO Box5003, N-1432 A s, Norway
Received 30 April 2004; received in revised form 26 May 2004; accepted 26 June 2004
Available online 11 August 2004
Abstract—A method for analysis of mixtures of 3-hexuloses by gas chromatography mass spectrometry of their di-O-isopropylidene
derivatives has been elaborated. The origin of characteristic fragment ions in the mass spectra is suggested on the basis of the spectra
of d12 analogues, obtained by acetonation with acetone-d and on MS/MS investigations. The method has been applied to product
6
mixtures from aldol reactions between glycero-tetrulose and glycolaldehyde and between 2-pentuloses and formaldehyde. An inter-
esting result is the formation of ribo-3-hexulose with a high degree of stereoselectivity in alkali catalysed reaction between erythro-2-
pentulose and formaldehyde.
ꢀ
2004 Elsevier Ltd. All rights reserved.
Keywords: Aldol reactions; 2-Pentuloses; glycero-Tetrulose; 3-Hexuloses; Mass spectrometry; Isopropylidene derivatives; Glycolaldehyde;
Formaldehyde
1
. Introduction
(xylulose) and formaldehyde in 0.1M NaOH solution
gave about equal amounts of the D-xylo-and D-lyxo-3-
hexuloses, isolated as their di-O-isopropylidene deriva-
3
tions such as oxidation of hexitols, aldol reactions
-Hexuloses are found in the product mixtures of reac-
1
2,3
2
tives. When DL-glycero-tetrulose (erythrulose) and
4
and the formose reaction. Reported synthetic methods
glycolaldehyde were used, and strongly basic anion
exchange resin applied as catalyst, acetonation showed
that the DL-xylo-and DL-lyxo-3-hexuloses were formed
in a ratio of about 2:5, and this reaction was used for
preparation of 1,2:3,4-di-O-isopropylidene-b-DL-lyxo-3-
are based on CrO oxidation of the tetraacetates of ethyl-
3
idene, methylene or benzylidene derivatives of hexi-
5
,6
tols, and Br oxidation of 1,2:5,6-di-O-isopropyl-idene
2
7
hexitols. These hexuloses have not received much atten-
3
tion. An important exception is D-arabino-3-hexulose,
which is formed as its 6-phosphate as primary product
in formaldehyde fixation in some methylotrophic bacte-
ria by aldol reaction with ribulose-5-phosphate in the
ribulose monophosphate (RuMP) pathway.
reaction is catalysed by 3-hexulose-phosphate synthase
hexulofuranose.
We have now extended the investigation of aldol reac-
tions of small sugars to the reaction of formaldehyde
with D-erythro-2-pentulose (ribulose). This seemed of
particular interest in the light of the earlier mentioned
8
–10
This
8
–11
formaldehyde fixation in methylotrophic bacteria.
We have also included in our investigation the alkali cat-
(
dephosphorylation.
HPS), and the free hexulose may be obtained on
1
1
alysed aldol reaction between DL-glycero-tetrulose and
2
+
In continuation of our work with 3-hexuloses, we are
interested in their formation by aldol reactions of small
sugars. It was shown previously that D-threo-2-pentulose
glycolaldehyde, with and without the presence of Sr
2
+
2+
ions. It has been found previously that Ca and Sr
ions strongly influence the product composition in alkali
1
2
catalysed triose aldol reactions.
A method for analysis of the products in these reac-
tions was needed. Our experience with GC–MS of
*
Corresponding author. Tel.: +47-6494-7714; fax: +47-6494-7720;
e-mail: svein.morgenlie@nlh.no
0
008-6215/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2004.06.018

2172
D. Ekeberg, S. Morgenlie / Carbohydrate Research 339 (2004) 2171–2176
CH3
CH3
O
O
CH2
C
CH2
C
O
O
CH3
CH3
H
H
O
O
OH
OH
O
O
O
CH3
O
CH3
CH3
CH3
1
2
CH3
CH3
OH
O
O
O
O
H
C
O
C
CH2
H
O
O
O
CH3
O
O
CH2OH
CH3
CH3
CH3
CH3
CH3
3
4
CH3
CH3
O
CH2
CH3
O
OH
H
O
C
H
O
CH3
CH3
CH3
O
O
O
O
C
C
O
O
O
H
O
O
O
CH2OH
OH
O
CH2
CH3
CH3
CH3
CH3
CH3
CH3
7
6
5
Scheme 1.
1
3,14
O-isopropylidene derivatives of monosaccharides
ence in the chromatographic mobilities of 1 and 2
(Table 2) eliminates this as a problem in analysis of
the 3-hexuloses. An explanation of this difference, ob-
served on Silica gel as well as in GC, was suggested in
a previous paper in this series based on the ability of 1
to form intramolecular hydrogen bond in a favourable
suggested this as an attractive method. The usefulness
of O-isopropylidene derivatives for identification of
monosaccharides by mass spectrometry was first shown
by DeJongh and Biemann. The advantage of these
compounds over acetates and trimethylsilyl ethers is
due to the fact that they frequently are structural, and
not merely configurational, isomers. This method has
previously been found useful for analysis of 2-hexuloses
1
5
2
conformation. The 3-hexuloses with ribo and arabino
configurations give more than one di-O-isopropylidene
derivative. From ribo-3-hexulose are obtained two deriv-
atives 3 and 4 in relative proportions about 6:1. The
1
4
and 2-pentuloses. In a previous paper, some EI-MS
data were reported for di-O-isopropylidene derivatives
7
of 3-hexuloses without discussion of the spectra. In
6
,7
major product 3 is known to be the 1,2:4,5-di-O-isop-
ropylidene-ribo-3-hexulofuranose, whereas 4 is the
2,3:4,5-di-O-isopropylidene-b-furanose derivative. From
arabino-3-hexulose are obtained three derivatives, 5, 6
and 7 in relative proportions 65%:30%:5%, respectively.
the present paper, the origin of some important frag-
ments in the spectra is suggested, the possibility to dis-
criminate between the 3-hexuloses is investigated, and
the application of the GC–MS method to the product
mixtures from aldol reactions is described.
6
,7
They are known to be 2,3:4,5-di-O-isopropylidene-b-
arabino-3-hexulofuranose, 1,2:4,5-di-O-isopropylidene-
arabino-3-hexulofuranose and the 1,2:3,4-di-O-isopro-
pylidene-a-furanose derivative, respectively.
2
. Results and discussion
The EI mass spectra of di-O-isopropylidene deriva-
tives of the common hexoses have been thoroughly de-
6
In accordance with a report by Angyal and Evans, and
15
scribed by DeJongh and Biemann. The molecular ion
2
,3,7
confirmed in previous papers in this series,
and lyxo-3-hexuloses gave the 1,2:3,4-di-O-isopropylid-
ene-b-furanose derivatives 1 and 2, respectively, as single
products (see Scheme 1 for structures 1–7). These com-
pounds are diastereomers, and their EI mass spectra are,
as a consequence, almost identical (Table 1). The differ-
the xylo-
is usually not observed, the highest mass peak is ob-
served at m/z 245, due to the loss of a methyl radical
from the molecular ion. The loss of acetone, acetic acid
or both from this ion leads to fragments with m/z 187,
185 or 127, respectively. In addition, high intensity
peaks at m/z 59 (protonated acetone) and 43 (acetylium
ion) are observed. Suggested fragmentations leading to
the more characteristic ions observed in the spectra of
1 and 2 are shown for compound 1 in Figure 1. The sug-
ꢀ
ꢀ
Only one enantiomer, D or L, of each hexulose and the derivatives is
shown throughout this paper.

D. Ekeberg, S. Morgenlie / Carbohydrate Research 339 (2004) 2171–2176
2173
Table 1. Mass spectral data for the main di-O-isopropylidene-3-
hexuloses
5) (Table 1). The high intensity m/z 200 fragment, due
to a C-3–O-6 and a C-4–C-5 cleavage, and the fragment
ion at m/z 142, resulting from loss of acetone from the
m/z 200 fragment, are not observed in the spectra of
the ribo and arabino derivatives. In addition, peaks at
m/z 101 from C-2–C-3 cleavage of high abundance are
observed in the spectra of 1 and 2. A fragment ion with
m/z 159, also resulting from C-2–C-3 cleavage, is
observed in the spectrum of 3 as well as in those of 1
and 2. However, in the spectrum of 3, the m/z 101 frag-
ment ion is much less abundant than in the spectra of 1
and 2, whereas the m/z 159 fragment is of high abun-
dance. An explanation could be that the 159 fragment
ion from 3 is more easily stabilised than those from 1
and 2 due to the possibility to achieve planarity within
the group of atoms involved in the resonance stabilisa-
tion of the fragment ion (Fig. 2). It is, however, observed
an other route, which may be more important, resulting
in a m/z 159 fragment with the same elemental composi-
tion. This route is confirmed by a product ion mass spec-
trum of a m/z 217 fragment ion as precursor. This
fragment is observed in the spectrum in relatively low
abundance, and could be represented as shown in Figure
a
m/z
Relative intensity (%)
1
2
3
5
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
9
9
8
8
7
7
6
5
5
5
4
45 (254)
100
79
3
40
1
100
43 (255)
29 (241)
27 (236)
17 (229)
02 (208)
00 (212)
87 (190)
85 (191)
71 (177)
69 (172)
59 (165)
47 (153)
43 (146)
42 (148)
31 (134)
27 (127,128)
15
4
88
3
7
6
5
59
18
27
35
14
21
6
8
3
8
20
4
10
76
8
89
68
13
6
12
30
10
4
8
5
11
18
10
14
12
35
37
9
39
8
14 (120)
13 (116)
01 (107)
9
25
20
7
6
64
13
6
7
61
12
8
28
7
12
24
52
10
11
34
27
93
38
34
88
7
5
15
2
. Loss of acetone from the m/z 217 fragment leads to a
5 (88,91)
3
16
14
22
24
12
73
16
37
90
14
12
21
23
10
78
19
31
m/z 159 fragment. A very small, but significant peak is
observed at m/z 243 in the spectrum of 3 due to loss
of the OH-3 as a radical from the molecular ion. This
is characteristic in spectra of derivatives with the ano-
3
7
13
1
9
9 (65)
7
100
17
15
meric hydroxyl group free. In the spectrum of com-
pound 5, a characteristic peak of high abundance, not
observed in the spectrum of 3, and of very low intensity
in the spectra of 1 and 2, is seen at m/z 227, due to the
5
3 (46)
19
75
100
a
The figures in parentheses refer to m/z for the d12-analogues.
+
loss of water from M ÀCH , whereas a unique frag-
3
gestions are based on the known fragmentation of 2-
4,15
ment for 5 is observed at m/z 171, arising from loss of
acetone from the fragment ion m/z 229, which is ob-
served in surprisingly low abundance in the spectrum
(Fig. 3). A peak at m/z 147, seen in the spectra of 3
and 5, is obviously an analogue of the m/z 117 ion,
which is characteristic of a 1,2-O-isopropylidene group
1
hexulose derivatives
of d12 analogues, obtained by derivatisation with ace-
and is supported by the spectra
tone-d . For some of the fragments, product/precursor
6
ion techniques have also been applied. Suggested frag-
mentation in the mass spectrum of 3, which is almost
identical to that of its diastereomer 6, is shown in Figure
1
5,16
14
in derivatives of 2-hexuloses
and 2-pentuloses con-
2
, and the possible origin of the most characteristic frag-
taining the C-1 to C-3 part of the molecule, as well as
protonated O-6 (Figs. 2 and 3). Finally, a fragment
ion with m/z 114 is observed in the spectra of 3 and 5,
not observed in the spectra of 1 and 2, containing the
C-4 to C-6 part of the molecule, being characteristic of
3-hexulose derivatives with a 4,5-O-isopropylidene
ments in the spectrum of 5 is shown in Figure 3, the
spectrum is for the same reason similar to that of 4.
The mass spectra of the xylo (1) and lyxo (2) deriva-
tives are readily differentiated from those of the main
derivatives of the ribo-and arabino-3-hexuloses (3 and
Table 2. Chromatographic data for the main di-O-isopropylidene derivatives of the hexuloses
a
Hexulose
di-O-Isopropylidene-acetal
Retention time/min
Molar responses
xylo-3-Hexulose
ribo-3-Hexulose
arabino-3-Hexulose
lyxo-3-Hexulose
arabino-2-Hexulose
xylo-2-Hexulose
1,2:3,4
1,2:4,5
2,3:4,5
1,2:3,4
2,3:4,5
2,3:4,6
10.3
11.0
11.6
13.5
13.3
14.1
1.00
0.85
0.65
1.00
0.95
0.80
a
Relative peak areas of the main acetal of each hexulose.

2174
D. Ekeberg, S. Morgenlie / Carbohydrate Research 339 (2004) 2171–2176
O
CH2
CH3
CH3
O
C
H
m/z 101
O
O
CH2
CH3
CH3
O
O
CH2
CH3
CH3
O
O
CH2
CH3
CH3
C
H
C
H
C
H
O
C
O
O
-
CH3
CH3
OH
OH
m/z 142
O
O
O
O
CH3
O
CH3
O
CH3
m/z 200 CH3
m/z 260 CH3
m/z 245
1
2
-H O
O
C
-
CH3
CH3
O
O
CH2
CH3
CH3
C
H
O
O
OH
m/z 187
O
O
O
O
CH3
CH3
CH3
-
H2O
m/z 227
O
C
m/z 159
-
CH3
CH3
m/z 169
Figure 1. Suggested origin of characteristic fragment ions in the mass spectrum of 1.
O
OH
O
OH
H
CH2
C
H
O
O
C
C
O
CH2
O
O
O
-CH3
CH3
CH2
O
O
O
O
CH3
m/z 159
CH3
CH3
CH3
CH3
CH3
CH3CH3
CH3
CH3
m/z 217
m/z 114
m/z 260
3
O
OH
O
H
CH CH2
O
O
CH
O
O
CH2
O
O
CH3 CH3
CH3
H
O
C
O
CH2
O
CH3
CH3
O
CH3
m/z 147
H
m/z 243
O
O
CH3CH3
m/z 101
CH3
CH3
m/z 159
Figure 2. Suggested origin of characteristic fragment ions in the mass spectrum of 3.
group. The MS data are shown in Table 1. The main di-
O-isopropylidene derivatives of the 3-hexuloses are well
separated under the applied GC conditions (Table 2).
This, in combination with the MS data, allows ready
discrimination between the 3-hexuloses.
lysed
reaction
between
glycero-tetrulose
and
glycolaldehyde has a marked influence on the product
composition. The relative amount of lyxo-3-hexulose is
reduced, whereas that of the arabino isomer is more than
doubled. In addition, substantial amounts of arabino-2-
hexulose (fructose) and, in particular, xylo-2-hexulose
(sorbose) are observed in the product mixture. The in-
crease in the amounts of 2-hexuloses formed in the pres-
ence of Sr ions must be a result of stabilisation as
alkaline earth ion complexes. The 2-hexuloses are sec-
ondary products, they are not observed during the initial
phase of the reaction. The change in product composi-
The results of the aldol reactions are shown in Table
. It is seen that the application of strongly basic ion ex-
3
change resins as catalysts in the reaction between glyc-
ero-tetrulose and glycolaldehyde leads to a higher
stereoselectivity than observed in the alkali catalysed
reaction. This is in accordance with previous observa-
2
+
1
4,17,18
tions during triose aldol reactions
reaction between dihydroxyacetone and glycolalde-
and the aldol
2
+
tion in the presence of Sr ions is not surprising, the
same phenomenon was observed during alkali catalysed
1
9
hyde. The presence of Sr ions during the alkali cata-
2+

D. Ekeberg, S. Morgenlie / Carbohydrate Research 339 (2004) 2171–2176
2175
CH3
CH3
O
CH2
O
O
CH3
O
CH3
O
O
O
C
O
O
C
C
H
-CH3
CH3
CH2
H
m/z 171
O
O
O
O
HO
CH3
CH3
CH3
CH3
CH3
m/z 114
CH3
m/z 229
m/z 260
5
CH3
O
O
CH3
H
O
O
CH3
C
O
O
H
CH2
O
O
C
HO
H
CH2
HO
CH3
CH3
-H2O
m/z 147
m/z 245
CH3
O
O
O
C
CH2
O
O
CH3
CH3
m/z 227
Figure 3. Suggested origin of characteristic fragment ions in the mass spectrum of 5.
Table 3. Relative proportions of hexuloses in the reaction mixtures after the aldol reactions between DL-glycero-tetrulose and glycolaldehyde (Panel
A), and between formaldehyde and 2-pentuloses in 0.05M NaOH solution (Panel B)
Catalyst
Reaction time (min)
Products (% of total hexuloses)
xylo-3-
Hexulose
ribo-3-
Hexulose
arabino-3-
Hexulose
lyxo-3-
Hexulose
arabino-2-
Hexulose
xylo-2-
Hexulose
Panel A
À
Amberlite, IRA-400 (OH ) resin
À
12
12
50
50
21
26
29
25
6
6
7
2
52
58
44
12
5
2
2
5
9
6
Dowex, 1 · 16 (OH ) resin
NaOH, 0.02M
11
11
8
25
6
22
NaOH, 0.02M, SrCl
2
2M
Panel B
erythro-2-Pentulose
threo-2-Pentulose
30
30
––
46
84
––
16
––
––
54
1
2
triose aldol reactions. Attempts to use anion exchange
resin as a catalyst in reactions between 2-pentuloses and
formaldehyde were unsuccessful. An unexpected result
was, on the other hand, observed on alkali catalysed al-
dol reaction between erythro-2-pentulose and formalde-
hyde. Whereas the hexulose phosphate synthase
catalysed formaldehyde fixation gives the arabino-3-hex-
ulose-6-phosphate, the ribo isomer is strongly favoured
in the alkali catalysed reaction. The stereoselectivity
here is high, unlike the situation in the analogous reac-
tion with threo-2-pentulose. In addition to hexuloses,
the product mixtures from these reactions also contain
aldoses. Erythrose and threose are, as expected, formed
by self aldolisation of glycolaldehyde, and small
amounts of aldopentoses are occasionally produced by
isomerisation of the 2-pentuloses. Based on some of
the presented results, work is in progress to develop
alternative procedures for preparation of 3-hexuloses.
3. Experimental
3.1. General methods
GC was performed with a Shimadzu GC-14B gas chro-
matograph, equipped with an open tubular fused silica
column, 25 m · 0.32 mm ID, wall coated with CP-SIL
43 CB, programmed at 6ꢁ/min from 90 to 225ꢁC. For
mass spectrometry, a JEOL JMX-DX 303 double focus-
ing mass spectrometer with EB geometry was used,
equipped with a 10kV post acceleration conversion dy-
node detector and an electron ionisation––chemical ion-
isation (EI–CI) ion source, operated in EI mode at 70eV
and an ion source temperature at 180ꢁC. MS/MS inves-
tigations were performed with an AutoSpec Ultima 2000
(Micromass ltd) mass spectrometer with EBE geometry.
A MIKEscan technique was used, the sample was intro-
duced into the ion source by direct insertion probe (DIP)

2176
D. Ekeberg, S. Morgenlie / Carbohydrate Research 339 (2004) 2171–2176
and ionised by electron ionisation at 70eV. The ion
source temperature was 200ꢁC. A Carlo Erba high res-
olution gas chromatograph (HRGC) was applied for
the GC–MS combination.
sin was added to fill the solution volume. The mixture
was kept at room temperature for 12min under nitro-
gen, 20% aqueous acetic acid was then added. After fil-
tration, the resin was washed with more acetic acid
solution and finally with water. The combined filtrates
and washings were evaporated, and the residue was trea-
ted with acetone or acetone-d –H SO as described
3
.2. Materials
6
2
4
2
D and L-eythro-2-Pentulose, D-threo-2-pentulose, DL-
0
21
above and the solution subjected to GC or GC–MS.
3
glycero-tetrulose DL-lyxo-, D-arabino-, L-xylo-, and
3
7
7
7
D-ribo-3-hexulose were prepared according to reported
methods.
3.6. Alkali catalysed reaction of DL-glycero-tetrulose with
glycolaldehyde
3
.3. GC–MS analysis
To DL-glycero-tetrulose (15mg) in water (5mL) were
added a solution of glycolaldehyde (12mg) in water
(3mL) and then 0.04M NaOH solution (8mL) with or
without SrCl Æ6H O (8.5g). After 50min under nitrogen
The sugars were identified by GC–MS of their isopropy-
,13,14,19
lidene derivatives on the basis of previous work.
7
2
2
The relative proportions of the hexuloses in the product
mixtures was calculated from the peak areas of the
derivatives. For the hexuloses giving more than one
derivative, the molar response of the main acetal was
stipulated from the proportion formed of this acetal
on acetonation of the hexulose.
at room temperature, the solution without SrCl was
2
+
neutralised with Dowex 50 W (H ) resin, filtered and
evaporated under reduced pressure. The solution with
À
SrCl was deionised with IRA 400 (HCO ) and Dowex
2
3
+
50 W (H ) resins, filtered and evaporated. The residues
were treated with acetone––H SO as described above
and subjected to GC and GC–MS.
2
4
3
.4. Reaction of 2-pentuloses with formaldehyde
To the 2-pentulose (20mg) in water (4.5mL) were added
5% formaldehyde solution (0.5mL) and 0.1M NaOH
5mL). The solution was flushed with nitrogen and kept
References
3
(
1. Decker, P.; Schweer, H. Carbohydr. Res. 1982, 107, 1–6.
2. Morgenlie, S. Acta. Chem. Scand. B 1982, 36, 725–727.
. Morgenlie, S. Acta. Chem. Scand. B 1987, 41, 745–748.
. Mizuno, T.; Weiss, A. H. Adv. Carbohydr. Chem. Biochem.
at room temperature for 30min under nitrogen. After
+
neutralisation with Dowex 50 W (H ) ion exchange re-
3
4
sin, filtering of the solution and evaporation of the sol-
vent under reduced pressure, the residue was dissolved
+
in MeOH. Dowex 50 W (H ) was added to the solution
1
974, 29, 173–227.
5
6
. Angyal, S. J.; James, K. Chem. Commun. 1970, 320–321.
. Angyal, S. J.; Evans, M. E. Aust. J. Chem. 1972, 25,
1495–1512.
. Ekeberg, D.; Morgenlie, S.; Stenstrøm, Y. Carbohydr. Res.
001, 335, 141–146.
. Kemp, M. B. Biochem. J. 1974, 139, 129–134.
. Strøm, T.; Ferenci, T.; Quayle, J. R. Biochem. J. 1974, 144,
4
10. Colby, J.; Zatman, L. J. Biochem. J. 1975, 148, 513–520.
and the mixture was heated at reflux temperature under
stirring for 2h to remove excess formaldehyde as its di-
methyl acetal, which was continuously distilled off dur-
ing the reaction period together with about one third
of the MeOH. After cooling, the solution was filtered
and the solvent evaporated. The residue was heated in
7
2
8
9
65–476.
1
1. Brockamp, H.-P.; Steigel, A.; Kula, M.-R. Liebigs. Ann.
Chem., 1993, 621–624.
0
methyl glycosides. The solution was then cooled, neu-
tralised with Amberlite IRA 400 (HCO3 ) ion exchange
.1M H SO (10mL) at 65ꢁC for 2h to hydrolyse
2
4
12. Morgenlie, S. Acta. Chem. Scand. B 1988, 42, 546–549.
13. Morgenlie, S. Carbohydr. Res. 1975, 41, 285–289.
14. Morgenlie, S. Carbohydr. Res. 1980, 80, 215–222.
À
resin, filtered and the solvent was evaporated. The resi-
due was stirred for 90min with 2% H SO in acetone or
acetone-d6 (5mL). After neutralisation with solid
15. DeJongh, D. C.; Biemann, K. J. Am. Chem. Soc. 1964, 86,
7–74.
6. Brady, R. F., Jr. Adv. Carbohydr. Chem. Biochem. 1971,
6, 197–278.
2
4
6
1
1
NaHCO and filtration of the solution, it was subjected
3
2
to GC and GC–MS.
7. Gutsche, C. D.; Redmore, D.; Buriks, R. S.; Nowotny, K.;
Grassner, H.; Armbruster, C. W. J. Am. Chem. Soc. 1967,
89, 1235–1245.
3
.5. Anion exchange resin catalysed reaction of
1
1
2
8. Morgenlie, S. Carbohydr. Res. 1982, 107, 137–141.
9. Morgenlie, S. Carbohydr. Res. 1984, 132, 330–334.
0. Ekeberg, D.; Morgenlie, S.; Stenstrøm, Y. Carbohydr. Res.
DL-glycero-tetrulose with glycolaldehyde
To DL-glycero-tetrulose (10mg) in water (2mL) was
added a solution of glycolaldehyde (8mg) in water
2
002, 337, 779–786.
2
1. Hough, L.; Theobald, R. S. Methods Carbohydr. Chem.
1962, 1, 94–98.
(
2mL). Enough freshly regenerated anion exchange re-