Acetobromomaltose, a new source of carbohydrate radicals. EPR characterisation of maltosyl and 2-deoxymaltos-2-yl radicals and syntheses of tetrasaccharide-like mimics, maltal, 3-α-maltosyl propiononitrile, 1,5-anhydromaltitol and 2-deoxymaltopyranoside

Article abstract of DOI:10.1016/S0040-4020(00)00567-6

The acetoxy-protected maltosyl radical 1, obtained through bromine abstraction from acetobromomaltose (ABM), was studied by means of EPR spectroscopy. At room temperature, only the spectrum of 1 was observed, but at higher temperatures a second radical, t

Full text of DOI:10.1016/S0040-4020(00)00567-6

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 6291±6297
Acetobromomaltose, a New Source of Carbohydrate Radicals.
EPR Characterisation of Maltosyl and 2-Deoxymaltos-2-yl
Radicals and Syntheses of Tetrasaccharide-like Mimics, Maltal,
3
-a-Maltosyl Propiononitrile, 1,5-Anhydromaltitol and
2
-Deoxymaltopyranoside
a
Angelo Alberti, Sabrina Bertini, Maura Comoli, Marco Guerrini, Andrea Mele
b
b
b
c
c,
*
and Elena Vismara
a
I.Co.C.E.A.-CNR, Area della Ricerca, via P. Gobetti 101, I-40129 Bologna, Italy
b
Istituto Giuliana Ronzoni, via Giuseppe Colombo 81, I-20133 Milan, Italy
Dipartimento di Chimica del Politecnico, via Mancinelli 7, I- 20131 Milan, Italy
c
Received 10 February 2000; revised 30 May 2000; accepted 15 June 2000
AbstractÐThe acetoxy-protected maltosyl radical 1, obtained through bromine abstraction from acetobromomaltose (ABM), was studied
by means of EPR spectroscopy. At room temperature, only the spectrum of 1 was observed, but at higher temperatures a second radical, the
acetoxy-protected 2-deoxymaltos-2-yl radical 2, was detected, resulting from migration of an acetoxy group from position 2 to position 1.
Some acetoxy-protected maltose derivatives were prepared from ABM, via different radical pathways involving 1 and 2 as intermediates.
Electroreduction on silver provides tetrasaccharide-like mimics 7 and maltal 8. Photochemical generation of 1, followed by trapping with
tributyltinhydride or with acrylonitrile, leads respectively to 1,5-anhydromaltitol 9 and to 3-a-maltosyl propiononitrile 10. Generation of 2 at
808C by 1!2 isomerisation, followed by trapping with tributyltinhydride, leads to 2-deoxymaltopyranoside 11. q 2000 Elsevier Science
Ltd. All rights reserved.
Introduction
As far as we know, there have been no reports on disacchar-
ide radical based synthetic procedures, nor on the characteri-
sation of these species through an ESR spectroscopic
approach. This prompted us to undertake an ESR investiga-
tion of the acetoxy-protected maltosyl radical 1, originated
via bromine abstraction from acetobromomaltose (ABM),
and of its 2-deoxy isomer 2, deriving from a 2,1 shift of an
acetoxy group. We also endeavoured to investigate the
synthetic potential of these radicals in polysaccharide
chemistry, and in particular in extending the chemistry of
anomeric radicals such as 1 from monosaccharides to more
complex carbohydrates.
Radical reactions have been largely exploited in the synth-
esis and derivatisation of saccharides. In particular, several
1
studies have been carried out on anomeric glycosyl radi-
cals, which have often been used as intermediates to
2
,3
3
synthesise 2-deoxy sugars and C-glycosides; as a matter
of fact, C±C bond radical formation appears to be a suitable
approach to the latter compounds. C-glycosides, which in
4
most cases are not `true' sugars, are stable carbohydrate
analogues that have received much attention as molecules
of potential biological activity and that are also useful for
enzymatic and metabolic studies. More recently, carbohy-
drates with C-interglycosidic bonds were synthesised with
5
the aim of checking their biological activity. Many radicals
Results and Discussion
ESR studies
derived from carbohydrates have also been studied by
means of ESR spectroscopy, the technique of choice for
the study of paramagnetic species.
6
,7
The acetoxy-protected maltosyl radical 1 was generated
inside the cavity of an ESR spectrometer following the
well-established procedure outlined in Scheme 1, i.e.
through bromine abstraction from ABM by tributyltin
centred radicals photogenerated in situ. At room tempera-
ture, only the spectrum of 1 was observed, but as the
Keywords: radicals and radical reactions; carbohydrates; electron spin
resonance; electrochemical reactions.
* Corresponding author. Tel.: 139-02-23993039; fax: 139-02-23993080;
e-mail: vismara@dept.chem.polimi.it
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00567-6

6
292
A. Alberti et al. / Tetrahedron 56 (2000) 6291±6297
Scheme 1.
temperature was raised, a second radical was detected. This
could be identi®ed as radical 2 resulting from migration of
the acetoxy group from position 2 to position 1, a process
already evidenced in the case of radicals from mono-
saccharides.
4.8 mT) are constants accounting for spin polarisation and
hyperconjugation, respectively. On the other hand, it has
been recently shown that when the carbon bearing the
b-hydrogen is also linked to an electronegative substistuent
such as ¯uorine or, as in the present case, oxygen, Eq. (1)
should be replaced by the Guerra equation (2):11
8
In principle one might expect that substitution of the
protecting group in position 4 of the sugar with a glucosyl
unit should not affect too much the properties of a radical
where the unpaired electron is located in position 1, and it
may thus be interesting drawing a comparison between the
properties of heptaacetoxymaltosyl and heptaacetoxy-
maltos-2-yl radicals with those of the two tetraacetoxy-
galactosyl and two tetraacetoxyglucos-2-yl radicals 3a/4a
2
aꢀH † ˆ A 1 B £ cos q 1 C £ cos q sin q
ꢀ2†
b
In this equation, besides spin polarisation (Aˆ0.386 mT)
and hyperconjugation (Bˆ3.91 mT), the electronegativity
of the substituent is taken into account through the introduc-
tion of a third element (Cˆ1.264 mT in the case of oxygen).
The lower value of the H splitting in 1 as compared to that
b
8
,9
measured in 3b suggests that in the maltosyl radical this
hydrogen is closer to the plane of the sp carbon than it is
and 3b/4b shown in Scheme 2.
2
in the glucosyl radical. Through the use of Eq. (2) the angle
q between the SOMO and the C -hydrogen bond in 1 is
actually computed as 78.38. In the rearranged radical 2
there are two different hydrogen atoms b to the radical
carbon, that is those in positions 1 and 3. Applying again
the Guerra equation, the dihedral angle q between the
From the data collected in Table 1 it emerges that the spec-
tral parameters exhibited by radicals 1 and 2 correlate well
with those exhibited by radicals 3 and 4, respectively. The
hyper®ne splitting constant of a hydrogen atom b to a p
radical centre depends on the dihedral angle between the
2
singly occupied orbital and the C±H bond, according to
b
1
0
SOMO and the C -hydrogen bond is calculated as 37.98
3
and that between the SOMO and the C -hydrogen bond as
1
the Heller±McConnel equation:
2
6
9.18. The hydrogen in position 3 would therefore have a
aꢀH † ˆ A 1 B £ cos q
ꢀ1†
where q is the dihedral angle, and A (ca. 0.3 mT) and B (ca.
b
marked `axial' character whereas that in position 1 would be
much closer to an `equatorial' situation. It should, however,
Scheme 2.

A. Alberti et al. / Tetrahedron 56 (2000) 6291±6297
6293
Table 1. Hyper®ne spectral parameters for radicals 1±4 (coupling constants
in mT)
likely to re¯ect inaccuracy in determining the relative
amount of radicals 1 and 2. Actually, owing to the poor
signal to noise ratio of the experimental spectra a^15%
Radical
a
a(H )
a(Hb1), a(Hb2
)
a(Hg1), a(Hg2)
error in the resulting E value is possible. On the other
a
1
2
3
3
4
4
1.869
2.174
1.648
1.817
2.160
2.182
0.789
0.344, 0.196
0.086
0.268
0.355
0.059
0.092
hand, the possibility that Eq. (3) overestimates the acti-
vation energy value should not be ruled out a priori despite
the fact that this equation has been tested on a large number
of radical rearrangements. Actually, although we do not
know the appropriate Tmid values for the rearrangement of
the glycosyl radicals, by introducing into Eq. (3) the
1.306, 3.433
2.873
1.245
4.142, 0.707
1.065, 3.753
a
b
a
b
8
21
measured activation energies, i.e. 12.8 and 8.7 kcal mol ,
Tmid values are obtained (285 and 192 K) at which experi-
mentally only the unrearranged radicals 3a and 3b are
observed. On the other hand, it has been claimed that
these rearrangements proceed through a ®ve-membered
cyclic transition state,13 and the higher value of the activa-
tion energy for 1!2 than for 3!4 might re¯ect a greater
rigidity of 1 with respect to 3 and conformational situations
making it more dif®cult for the former than for the latter to
attain the transition state. As will be discussed in the next
section, higher activation energies for the rearrangement
result in lower yields of the 2-deoxyderivative 11.
be borne in mind that the value of the angle q for C ±H is
1
likely to be underestimated to some extent, because Eq. (2)
only accounts for one electronegative substituent whereas
C is in fact bound to two oxygen atoms.
1
Although, as already mentioned, the spectral parameters of
radicals 1 and 2 are consistent with those of radicals 3 and 4,
the situation is less clear as far as the rearrangement 1!2 is
concerned. Actually, in a previous careful study of the
dependence on temperature of the isomerisation of the
radicals obtained via bromine abstraction from acetobromo-
8
galactose and acetobromoglucose, the activation energy for
2
1
this process was determined as 12.8 kcal mol for 3a!4a
Maltose derivatisation
2
1
and 8.7 kcal mol for 3b!4b.
Electrosynthesis. Electroreduction of acetoxy-protected
glycosyl bromides on silver was found to be a useful
When we tried to undertake a similar study of the 1!2
process, we found that the quality of the EPR spectra of
the maltosyl radical 1 was `reasonably good' at room
temperature, but became much poorer as the temperature
was raised in order to induce the radical translocation.
Indeed we only succeeded in obtaining a `decent' (see
Fig. 1) spectrum showing the presence of radicals 1 and 2
in nearly equal amount at ca. 508C, and careful study of the
process proved impossible.
1
4
approach to the synthesis of C-disaccharide-like mimics.
For example acetobromoglucose (ABG) provided a mixture
of a,a-, a,b-, and b,b-biglucosyl derivatives, along with
9
,14
1
4
signi®cant amounts of 3,4,6-tri-O-acetyl-glucal. Based on
several pieces of evidence we justi®ed the formation of the
former products through dimerisation of radical 3b, as
outlined in Eqs. (4) and (5):
2
2
ABG 1 e ! 3b 1 Br
ꢀ4†
ꢀ5†
On the other hand, in a recent report Walton suggested the
possibility of determining the activation energy for a
rearrangement, E , through Eq. (3) given the so-called
2
3b ! biglucosyl derivatives
r
mid-point temperature (Tmid), de®ned as the temperature at
which the rearranged and unrearranged species have equal
concentration.
While a single electron-transfer followed by loss of a
bromide anion accounts well for the formation of the
dimers, a two electrons cleavage of the C±Br bond and a
subsequent very fast elimination of an acetate anion, justify
the formation of the unsaturated sugar.
1
2
21
Er=kcal mol ˆ 0:044 £ Tmid 1 0:22
ꢀ3†
Substituting in the above equation 323 for T leads to an
mid
Electroreduction of ABM leads to results quite similar to
those obtained with ABG. Thus, controlled potential
2
1
activation energy value E ˆ14.43 kcal mol , a value some
r
2
±4 kcal mol2 larger than those reported for the isomeri-
sation of the glycosyl radicals. The observed discrepancy is
1
(
21.28 V vs SCE) electrolysis of ABM in acetonitrile at a
silver electrode afforded the bimaltosyl derivatives 7a and
b with an overall 40% yield, together with maltal 8 (25%).
8
7
The process is outlined in Scheme 3: once 1 is formed in a
fashion similar to that indicated in Eq. (4) for ABG, its
subsequent dimerisation leads to the a,a (7a) and a,b
(
7b) diastereomeric bimaltosyl derivatives (7b/7aˆ1.5).
This is in contrast with the reduction of ABG, from which
the three diastereoisomeric biglucosyl derivatives were
obtained in a statistical ratio.
9
Incidentally, we also note that ABG displays at a silver
cathode one irreversible peak at E ˆ21.28 V vs SCE.
p
Diagram 1. Dependence of yields of rearranged sugars 11, 12, 13 on
activation energy.
Actually, electroreduction of the ABG and ABM C±Br
bonds on silver does not seem to be in¯uenced by the

6
294
A. Alberti et al. / Tetrahedron 56 (2000) 6291±6297
Scheme 3.
substitution of the OAc group in position 4 of the glucose
with an O-glucosyl unit: nevertheless, the difference in
the diastereoselectivity allows us envisage the possibility
of some stereochemical control of this electrochemical
dimerisation.
compound 11 (56%). As outlined in Scheme 5, the isolated
products re¯ect the reactivity of radical 1, generated from
ABM by bromine abstraction brought about by tributyltin
radical. In agreement with the ESR results reported in the
preceding section, the product distribution can be inter-
preted on the basis of the partial isomerisation (cis-selective
migration of an acetoxy group at high temperature) of 1 to 2,
which in turn is trapped by tinhydride affording 11. This
approach had already been used to synthesise 2-deoxy-
galactopyranoside 12 and 2-deoxyglucopyranoside 13 via
Despite the low yields, ABM electroreduction shows some
interesting aspects as it is a one-pot chemoselective deriva-
tisation of a commercially available product and provides,
under clean and mild conditions, new structures such as the
bimaltosyl derivatives 7 along with more common
compounds like maltal 8, a member of the family of glycals
1
3
3a!4a and 3b!4b isomerisations, respectively.
1
5
which are important chiral building blocks.
It seems worthwhile to note the existence of an inverse
relationship between the activation energies characterising
the rearrangements of 3a (12.87 kcal mol ), 3b (8.7 kcal -
2
1 8
Photochemical syntheses. Photolysis of a THF solution of
tributyltin hydride and ABM led to the formation of 1 via
bromine abstraction by tributyltin radical. 1 was quantita-
tively converted to maltitol 9 by the tributyltinhydride
2
1 8
21
mol ) and 1 (14.43 kcal mol ) with the yields (12,
1
3
13
71%; 13, 92%; 11, 56%). The qualitative correlation
shown in Diagram 1 is probably too good to be true, but it
nevertheless gives us con®dence in the estimated activation
energy for the 1!2 rearrangement, which does not seems to
exceed the expected value.
(Scheme 4). Addition of excess acrylonitrile led to the iso-
lation of 3-a-maltosyl propiononitrile 10 (69%) together
1
b
with 9 (20%). Owing to the anomeric effect, this reaction
is totally diastereoselective.
Synthesis of 2-deoxymaltopyranoside 11. Heating a tolu-
ene solution of ABM containing tributyltin hydride and a
catalytic amount of AIBN [azobis(isobutyronitrile)] at
Conclusion
The radical 1 is generated from ABM both via electro-
chemical and chemical pathways and used to prepare
353 K resulted in the formation of maltitol 9 (34%) and of
Scheme 4.

A. Alberti et al. / Tetrahedron 56 (2000) 6291±6297
6295
Scheme 5.
Figure 1. EPR spectrum observed at 508C upon photolysis of a deoxygenated benzene solution of ABM containing some hexabutylditin. The spectrum of
radical 1 consists of the starred lines ( ), whereas the low-®eld and high-®eld doublets (#) are part of the spectrum of radical 2.
p
different compounds like 7±11. The presence of the
O-glucosyl substituent in position 4 does not in¯uence
signi®cantly the chemistry of this radical centred on the
anomeric position of a glucose unit, except for its rear-
rangement to 2. In agreement with the glycosyl radicals
(Sigma) in 2:3 (v/v) mixture of acetonitrile and aqueous
0.1% (v/v) tri¯uoroacetic acid (TFA). The concentration
of the carbohydrate solutions was kept close to 100 pmol/
ml, corresponding to 0.1 mM. The samples were prepared
by mixing 0.5 mL of matrix solution with 0.5 ml of analyte
solution directly on the sample slide and allowing evapo-
3a and 3b, 1 undergoes the acetoxy group migration from
position 2 to position 1, but EPR study and synthetic results
indicate that it is characterised by a higher activation
energy.
ration of the solvent under a gentle N stream.
2
EPR spectra were recorded at room temperature with a
Bruker ER200D spectrometer, equipped with an NMR
gaussmeter for ®eld calibration and a frequency counter
for the determination of g-factors that were corrected with
respect to that of perylene radical cation in conc. sulphuric
acid. Typical samples, consisting of argon-purged benzene
Experimental
General methods
2
3
solutions of ABM (ca. 10 M) and some hexabutylditin,
were photolysed with the un®ltred light from a Hanovia
high-pressure 1 kW mercury lamp.
1 13
H and C NMR were recorded on Bruker AC 300P and
AMX 500 spectrometers at 303 K. Chemical shifts are
expressed in ppm down®eld from TMS; assignments were
made through COSY and HMQC experiments. To identify
and con®rm the multiplicities of the proton signals the
coupling constants have been re®ned using the windaisy
Bruker program.
An Amel 553 potentiostat was used to drive electrolysis
under potentiostatic control. A Rayonet RPR 208 (Southern
New England Ultraviolet Company) and RLU-2537 A
lamps (Hg low pressure, 120 watts, lˆ254 nm) were used
for photochemical reactions. Flash-chromatography was
performed using silica gel 60 (230±400 mesh, Merck).
Observed rotations at the Na-D line were obtained at 208C
using a Propol Dr Kernchen polarimeter. Matrix Assisted
Laser/Desorption IonizationÐTime of Flight mass spectra
(
MALDI-TOF MS) were acquired on a Shimadzu±Kratos
General procedures
Kompact Maldi II instrument operating in the linear mode
with an accelerating potential of 20 kV and equipped with a
Preparative electrolysis. Controlled potential electrolysis
was carried out in acetonitrile in a two-compartment cell
UV pulsed laser (N , lˆ337 nm, 100 mJ per shot, 3 ns
2
pulse width of a single laser shot) and time delayed extrac-
tion. All the spectra were acquired by scanning the sample
spot along the x axis and averaging 100 laser shots. Mass
calibration was achieved using the matrix and g-cyclo-
dextrin as internal references. A 65 mM solution of the
matrixes were prepared by dissolving 2,5-dihydroxybenzoic
acid (DHB) or a-cyano-4-hydroxycinnamic acid (CHCA)
at E ˆ21.4 V. The cathode consisted of a silver plate,
p
cathodic solution 0.04 M ABM, 0.1 M tetraethylammonium
perchlorate (TeaP) as supporting electrolyte. A silver plate
in ACN-saturated tetraethylammonium bromide was used
as anode to provide a harmless counter-reaction (namely,
2
Ag1Br!AgBr1e ). A glass diaphragm was used to
separate anodic and cathodic compartments. ABM was

6
296
A. Alberti et al. / Tetrahedron 56 (2000) 6291±6297
exhaustively electrolysed. 7a, 7b and 8 were isolated from
the catholyte by ¯ash chromatography on silica gel, after
salt removal. 8 was eluted with hexane±ethyl acetate (1:1),
while a more polar mixture (hexane±ethyl acetate, 3:7) was
used for 7a and 7b. 7±8 were identi®ed on the basis of their
J9±10ˆ9.6 Hz, H-9), 6.01 (1H, dd, J ˆ1.1 Hz, H-1).
1±3
Anal. calcd for C H O : C, 51.43; H, 5.71. Found: C,
32 15
2
4
49.28; H, 5.63.
Synthesis of maltitol (9)
1
13
mass and H and C NMR parameters.
A solution of ABM (0.3 M) and tributyltinhydride (0.37 M)
in anhydrous THF under argon atmosphere was photolysed
for 6 h. THF was evaporated at reduced pressure. The resi-
due was dissolved in acetonitrile. The acetonitrile phase was
washed with pentane, to remove tin derivatives. Acetonitrile
was then evaporated and the residue ¯ash-chromatographed
on silica-gel with hexane±ethyl acetate (3:7) to give 9,
0
compound was isolated as a white solid, mp 82±848C;
1
-a-(a -Maltosyl)-1,5-anhydro-maltitol
(7a).
The
2
0
[
[
a] ˆ166 (c 0.2, CHCl ); MALDI-MS: m/zˆ1260.8
D
3
1 1
1
M1Na] , 1276.7 [M1K] ; H NMR (500 MHz, C D , J
6 6
values in Hz) d: 1.64, 1.65, 1.69, 1.73, 1.85, 1.90, 1.96, 2.06
(
H-4), 4.26 (1H, m, H-5), 4.27 (1H, dd, H-6b), 4.33 (1H, ddd,
42H, 8s, COCH ), 3.78 (1H, dd, J ˆ4.7 Hz, J ˆ7.1 Hz,
3
3±4
4±5
1
which was identi®ed on the basis of its mass and H and
13
C NMR parameters.
0
J4
(
4
(
0
0
ˆ10.0 Hz, J
0
0
ˆ4.0 Hz, J
0 0
ˆ2.6 Hz, H-5 ), 4.36
±5
5 ±6 a
5 ±6 b
0
0
1H, dd, J
0
0
ˆ12.4 Hz, H-6 b), 4.41 (1H, dd, H-6 a),
6
a±6 b
.44 (1H, d, J1±2ˆ4.2 Hz, H-1), 4.49 (1H, dd, H-6a), 5,11
1,5-Anhydromaltitol (9). This compound was isolated as a
0
20
D
1H, dd, J1
J2±3ˆ6.0 Hz, H-2), 5.37 (1H, dd, J
0
0
ˆ3.8 Hz, J
0
0
ˆ10.4 Hz, H-2 ), 5.23 (1H, dd,
white solid, mp 120±1238C; [a] ˆ116.8 (c 0.6, CHCl );
±2
2 ±3
3
1
0
1H, dd, H-3), 5.51 (1H, d, H-1 ), 5.73 (1H, dd, H-3 );
1
1
0
0
ˆ9.5 Hz, H-4 ), 5.47
MALDI-MS: m/zˆ643.4 [M1Na] , 659.4 [M1K] ; H
3 ±4
0
NMR (125 MHz, C D ) d: 61.8 (C-6 ), 62.6 (C-6), 68.8
0
13
(
C
NMR (500 MHz, C D , J values in Hz) d: 1.60, 1.66,
6
6
0
C-4 ), 69.0 (C-1, C-5 ), 69.3 (C-2), 70.1 (C-3 ), 70.7
6
6
1.68, 1.74, 1.79, 1.81, 1.94 (21H, 7s, COCH ), 2.83 (1H,
3
0
0
0
0
C-2 ), 71.4 (C-3), 72.8 (C-5), 74.7 (C-4), 97.4 (C-1 ).
(
(
ddd, J5±6bˆ2.8 Hz, J ˆ2.8 Hz, J ˆ9.9 Hz, H-5), 2.89
5±6a
4±5
0
(1H, t, J1
J1b±2ˆ5.7 Hz, H-1b), 3.82 (1H, t, J ˆ9.2 Hz, H-4), 4.07
and J ˆ10.8 Hz, H-1a), 3.69 (1H, dd,
1a±2
a±1b
4±3
0
pound was isolated as a white solid, mp 100±1018C;
1-a-(b -Maltosyl)-1,5-anhydro-maltitol (7b). The com-
(1H, dd, J6 ˆ4.4 Hz, J
a±5
ˆ12.3 Hz, H-6a), 4.21 (1H, dt,
6a±6b
0
0 0
J5
0
0
ˆ9.7 Hz, H-5 ), 4.33±4.36 (2H, m, H-6 a, H-6 b), 4.44
-4
2
D
0
[
[
a] ˆ181 (c 0.35, CHCl ); MALDI-MS: m/zˆ1261.2
(1H, dd, J ˆ3.9 Hz, H-6b), 4.90 (1H, ddd, J ˆ9.8 Hz,
3
6b±5
2±3
1
M1Na] , 1277.0 [M1K] ; the four rings are named for
1
0
H-2), 5.09 (1H, dd, J2
0
0
0
ˆ4.0 Hz, J
0 0
2 ±3
ˆ10.5 Hz, H-2 ),
±1
1
0
(1H, d, H-1 ), 5.84 (1H, t, H-3); C NMR (125 MHz,
0
convenience A±B±C±D: H NMR (500 MHz, C D , J
6
5.29 (1H, t, J
0
ˆ9.0 Hz, H-4 ), 5.39 (1H, t, H-3 ), 5.55
6
4 ±3
0
CDCl ) d: 61.5 (C-6 ). 63.0 (C-6), 66.0 (C-1), 68.0
13
values in Hz) d: 1.60±1.90 (30H, 8s, COCH ), 1.99, 2.02,
3
0
(C-4 ,C-5 ), 69.5 (C-3 ), 70.0 (C-2,C-2 ), 72.5 (C-4), 76.0
2
.15 (12H, 4s, COCH ), 2.97 (1H, ddd, J
3
ˆ2.5 Hz,
5C±6Cb
3
0
(C-3), 76.5 (C-5), 95.0 (C-1 ). Anal. calcd for C H O : C,
0
0
0
J5C±6Caˆ7.4 Hz, H-5C), 3.63 (1H, ddd, J
ˆ9.6 Hz,
4C±5C
0
50.32; H, 5.81. Found: C, 48.90; H, 5.63.
H-4C), 3.67 (1H, dd, J
1C), 3.74 (1H, m, H-4B), 4.08 (1H, m, H-6Ca), 4.13 (1H,
ˆ5.2 Hz, J
ˆ9.6 Hz, H-
1
B±1C
1C±2C
26 36 17
ddd, J1B±2Bˆ4.9 Hz, H-1B), 4.21 (1H, dd, J
ˆ
6Ca±6Cb
1
1.9 Hz, H-6Cb), 4.23 (1H, m, H-5D), 4.28 (1H, m,
Synthesis of 3-a-maltosyl propiononitrile (10)
H-5B), 4.29 (3H, m, H-6Aa, H-6Da, H-6Ba), 4.37 (1H, m,
H-5A), 4.42 (3H, m, H-6Bb, H-6Ab, H-6Db), 5.07 (2H, dd
A solution of ABM (0.07 M), tributyltinhydride (0.1 M) and
acrilonitrile (0.7 M) in anhydrous THF under argon atmos-
phere was photolysed for 6 h. Same work-up as described
for 9. The residue was ¯ash-chromatographed with hexane±
ethyl acetate (1:1) to give 9 and 10; 10 was identi®ed on the
H-2A, H-2D, J2A±3Aˆ10.4 Hz, J
ˆ10.6 Hz), 5.15 (1H,
2D±3D
dd, J2C±3Cˆ9.0 Hz, H-2C), 5.29 (1H, m, H-4D), 5.33 (1H,
dd, J3C±4Cˆ8.1 Hz, H-3C), 5.36 (2H, m, J ˆ6.0 Hz,
2B±3B
H-4A, H-2B), 5.39 (1H, d, J
(
ˆ4.1 Hz, H-1D), 5.52
1
D±2D
1
13
1H, d, J1A±2Aˆ4.1 Hz, H-1A), 5.70 (1H, dd, H-3B), 5.74
basis of its mass and H and C NMR parameters.
(
1H, dd, J3A±4Aˆ9.5 Hz, H-3A), 5.82 (1H, dd, J
ˆ
3D±4D
1
3
.5 Hz, H-3D); C NMR (125 MHz, C D ) d: 61.9
9
3-a-Maltosyl propiononitrile (10). This compound was
6
6
2
D
0
(
(
(
(
(
(
C-6B, C-6A, C-6D), 63.9 (C-6C), 68.7 (C-4D), 68.9
C-4A, C-5A, C-2B), 69.4 (C-5D), 69.9 (C-3D), 70.1
C-3A), 70.5 (C-1B), 70.7 (C-2A), 70.7 (C-2D), 70.9
C-3B), 71.1 (C-2C), 73.1 (C-5B), 73.9 (C-4C), 74.4
C-4B), 76.3 (C-5C), 76.6 (C-3C), 77.3 (C-1C), 96.3
C-1D), 97.2 (C-1A).
isolated as a colourless oil; [a] ˆ174.4 (c 0.4, CHCl );
3
1
1
MALDI-MS: m/zˆ696.6 [M1Na] ; H NMR (500 MHz,
C D , J values in Hz) d: 1.65, 1.67, 1.68, 1.78, 1.82, 1.95
6
6
0
(21H, 6s, COCH ), 1.11 (1H, m, H-3 ), 1.50 (1H, m,
3
0
J3±4ˆ4.8 Hz, H-3), 1.65 (2H, m, H-2, H-2 ), 3.52 (1H,
ddd, J7 ˆ6.8 Hz, J
0
ˆ3.3 Hz, J ˆ6.2 Hz, H-8), 3.61
8±9
±8
8±9
(
m, H-9 ), 4.19 (1H, m, H-9), 4.28 (1H, m, H-14, J
1H, dd, J ˆ6.52 Hz, H-7), 3.73 (1H, m, H-4), 4.14 (1H,
6
±7
2
0
0
10.3 Hz), 4.34 (2H, m, H-15, H-15 ), 4.82 (1H, dd, J
Maltal (8). This was isolated as a white foam; [a] ˆ172.2
ˆ
ˆ
D
13±14
1
0
(
[
c 0.2, CHCl ); MALDI-MS: m/zˆ583.1 [M1Na] , 599.0
3
4±5
1
1
M1K] ; H NMR (500 MHz, C D , J values in Hz) d:
6
6
4.6 Hz, J5±6ˆ6.9 Hz, H-5), 5.05 (1H, dd, J
10±11
ˆ3.9 Hz,
1.74, 1.76, 1.85, 1.88 (18H, 4s, COCH ), 3.75 (1H, q,
3
J11±12ˆ10.5 Hz, H-11), 5.25 (1H, dd, J ˆ6.5 Hz, H-6),
6±7
H-5), 4.02 (1H, t, J4±5 and J4±3ˆ8.2 Hz, H-4), 4.20 (1H,
5.31 (1H, dd, J
(1H, d, J10±11ˆ3.9 Hz, H-10), 5.80 (1H, dd, J
ˆ9.4 Hz, J
ˆ10.3 Hz, H-13), 5.43
12±13
12±13
13±14
m, H-11), 4.20 (2H, dd, H-6a, H-6b), 4.31 (1H, dd,
J12a±11ˆ4.1 Hz, J
ˆ9.4 Hz,
6 6
1
3
ˆ12.4 Hz, H-12a), 4.36 (1H, dd,
J11±12ˆ10.5 Hz, H-12); C NMR (125 MHz, C D ) d:
12a±12b
12b±12a
J12b±11ˆ2.7 Hz, J
ˆ12.4 Hz, H-12b), 4.60 (1H, dd,
19.5±21.2 (±OCH ), 13.2 (C-2), 23.5 (C-3), 61.8 (C-15),
3
J2±1ˆ6.02 Hz,
J
ˆ3.09 Hz, H-2), 4.99 (1H, dd,
62.3 (C-9), 68.7 (C-13), 69.0 (C-14), 69.6 (C-4), 69.7
(C-5), 69.8 (C-12), 70.6 (C-11), 71.02 (C-8), 71.4 (C-6),
73.8 (C-7), 97.0 (C-10), 119 (C-1, CN). Anal. calcd for
2±3
J8±9ˆ10.4 Hz, J ˆ3.9 Hz, H-8), 5.34 (1H, m, H-3), 5.35
8±7
(
1H, t, H-10), 5.652 (1H, d, H-7), 5.834 (1H, dd,

A. Alberti et al. / Tetrahedron 56 (2000) 6291±6297
6297
C H NO : C, 51.71; H, 5.79; N, 2.08. Found: C, 53.20; H,
2
B. Appl. Chem. 1988, 60, 1655±1658. (c) Vismara, E.; Donna, A.;
Minisci, F.; Naggi, A.; Pastori, N.; Torri, G. J. Org. Chem. 1993,
9
39
17
5.97; N, 2.01.
5
8, 959±963. (d) Cipolla, L.; Guerrini, M.; Nicotra, F.; Torri, G.;
Vismara, E. Chem. Commun. 1997, 1617±1618.
. Descotes, G. Carbohydrates; Kodansha: Tokyo, 1992; pp 89±
Synthesis of 2-deoxymaltopyanoside (11)
2
3.5 ml of a solution of tributyltinhydride (0.3 M) and AIBN
96.
(
0.06 M) in anhydrous toluene was slowly dropped (10 ml/
3. Levy, D. E.; Tang, C. The Chemistry of C-glycosides,
Tetrahedron Organic Chemistry Series, Pergamon: Oxford, 1995;
Vol. 13.
h) in 13 ml of a stirred solution of ABM (0.078 M) in
anhydrous toluene at 354 K. At the end of the dropping,
the reaction was further stirred for 30 min at 354 K. After
evaporation of toluene, same work-up as described for 9.
The residue was ¯ash-cromatographed with hexane±ethyl
acetate (1:1) to give 9 and 11; 11 was identi®ed on the basis
4. Giese, B. Radicals in Organic Syntheses: Formation of C±C
Bonds, Pergamon: Oxford, 1986.
5. Petitou, M.; Herault, J.-P.; Lormeau, J.-C.; Helmboldt, A.;
Mallet, J.-M.; Sinay, P.; Herbert, J.-M. Bioorg. Med. Chem.
1998, 6, 1509±1516.
1
13
of its mass and H and C NMR parameters.
6
. Fischer, H.; Paul, H. In Magnetic Properties of Free Radicals;
2
-Deoxymaltopyranoside (11). This compound was
Fisher, H., Hellwege, K.-H., Eds.; Landolt-B oÈ rnstein Group II,
Vol. 9, Part b; Springer: Heidelberg, 1977; Neugebauer, F. A. In
Magnetic Properties of Free Radicals; Fisher, H., Ed.; Landolt-
B oÈ rnstein Group II, Vol. 17, Part b; Springer: Heidelberg, 1987.
7. Tabner, B. J. Electron Spin ResonanceÐA Specialist Periodical
Report; The Chemical Society: London, 1986; Vol. 10A; Tabner,
B. J. Electron Spin ResonanceÐA Specialist Periodical Report;
The Chemical Society: London, 1988; Vol. 11A; Tabner, B. J.
Electron Spin ResonanceÐA Specialist Periodical Report; The
Chemical Society: London, 1990; Vol. 12A; Alberti, A.; Hudson,
A. Electron Spin ResonanceÐA Specialist Periodical Report; The
Chemical Society: London, 1986; Vol. 15.
2
0
isolated as a white foam; [a] ˆ1145 (c 0.5, CHCl );
D
3
1
1
1
MALDI-MS: m/zˆ643.1 [M1Na] , 659.4 [M1K] ; H
NMR (500 MHz, CDCl , J values in Hz) d: 1.70 (1H,
3
ddd, H-2a), 1.9±2.1 (21H, m, COCH ), 2.20 (1H, ddd,
3
J2b±1ˆ1.7 Hz, J ˆ5.3 Hz, J
ˆ13.5 Hz, H-2b), 3.90
2b±3
0
2b±2a
0
(
(
2H, m, H-4, H-5 ), 4.0 (2H, m, H-5, H-6 a), 4.15±4.2
2H, m, H-6a, H-6 b), 4.40 (1H, dd, J6 ˆ2.6 Hz,
0
b±5
J6a±6bˆ12.2 Hz, H-6b), 4.70 (1H, dd, J
0
0
0
ˆ4.0 Hz,
0
2 ±1
0
J2
0
0
ˆ9.9 Hz, H-2 ), 5.0 (1H, t, J
0
0
and J
0
ˆ9.9 Hz,
±3
0
4 -3
4 ±5
H-4 ), 5.15 (1H, dd, H-3), 5.3±5.40 (1H, t, H-3 ), 5.50
0
13
(
(
(
(
1H, d, H-1 ), 6.10 (1H, dd, H-1, J ˆ3.6 Hz); C NMR
0
3
1
±2a
125 MHz, CDCl ) d: 33.0 (C-2), 61.0 (C-6 ), 62.5
8. Korth, H.-G.; Sustmann, R.; Groninger, K. S.; Leisung, M.;
È
Giese, B. J. Org. Chem. 1988, 53, 4364±4369.
0
0
0
0
C-6),68.0 (C-4 , C-5 ), 69.0 (C-3 ), 70.0 (C-2 , C-5), 71.5
0
C-3), 72.5 (C-4), 90.0 (C-1), 95.5 (C-1 ). Anal. calcd for
9. Alberti, A.; Della Bona, M. A.; Macciantelli, D.; Pelizzoni, F.;
Sello, G.; Torri, G.; Vismara, E. Tetrahedron 1996, 42, 10241±
C H O : C, 50.32; H, 5.81. Found: C, 50.27; H, 5.83.
2
6
36 17
1
1
1
1
0248.
0. Heller, C.; McConnell, H. M. J. Phys. Chem. 1960, 32, 1535±
539.
1. Guerra, M. Chem. Phys. Lett. 1987, 139, 463±469; Guerra, M.
Acknowledgements
The authors are grateful to Marc Robert (Laboratoire
d'Electrochimie Mol e culaire de l'Universit e Denis Diderot,
Paris, France) for the voltammetries and E determinations.
They are also indebted to MURST-CNR (Rome) for
®
J. Am. Chem. Soc. 1992, 114, 2077±2085; Guerra, M. Pure Appl.
Chem. 1995, 67, 797±804.
p
1
1
2. Walton, J. C. J. Chem. Soc., Perkin Trans. 2 1998, 603±605.
3. Giese, B.; Gilges, S.; Gr oÈ ninger, K. S.; Lamberth, C.; Witzel,
nancial support.
T. Liebigs Ann. Chem. 1988, 615±617.
4. Guerrini, M.; Mussini, P.; Rondinini, S.; Torri, G.; Vismara, E.
Chem. Commun. 1998, 1575±1576.
5. Spencer, R. P.; Cavallaro, C. L.; Schwartz, J. J. Org. Chem.
1999, 64, 3987±3995.
1
References
. (a) Sinay, P. Pure Appl. Chem. 1998, 70, 1495±1499. (b) Giese,
1
1