1910
F. Orsini, E. M. Lucci / Tetrahedron Letters 46 (2005) 1909–1911
Table 1. Sm(II)-catalysed Reformatsky-type reactions
Halogen compound
EWG
Electrophile
Addition product (%)
Pinacolic coupling product (%)
R
Y
R
R
R
X
EWG
R'
R'
R'
R'
YH
YH YH
''
Y = O, NR
tert-Butyl-a-bromo acetate
tert-Butyl a-bromo acetate
tert-Butyl a-bromo acetate
tert-Butyl a-bromo acetate
tert-Butyl a-bromo acetate
tert-Butyl a-bromo acetate
Benzyl a-bromo acetate
a-Bromo acetophenone
a-Bromo acetonitrile
Diethyl (a-iodomethyl) phosphonate
Diethyl (a-iodomethyl) phosphonate
Diethyl (a-iodomethyl) phosphonate
C6H5CHO
(C6H5)2CO
75a
77
10
—
—
—
60.5
—
—
—
—
—
—
—
CH3(CH2)5C(O)CH3
C6H5CH@CHCHO
C6H5CH@NCH2C6H5
2,3,5-Tri-O-benzyl-D-arabino-1,4-lactone
C6H5CHO
83
91a
24
85a,b d.r. = 2/1
90a
38
C6H5CHO
C6H5CHO
65
C6H5CHO
(CH3)3CHO
(CH3)2CHCHO
35c
33
35
a Yields determined by NMR analysis.
b A diastereoisomeric mixture was obtained.
c Diethyl methyl phosphonate was obtained (25%).
The addition to the carbon–nitrogen bond present in the
benzylimine of benzaldehyde was, in contrast, unsatis-
factory: the main product was 1,2-diphenyl-1,2-diami-
This substoichiometric protocol was also tested for the
reaction of diethyl (a-iodomethyl)phosphonate with
various aldehydes: the addition product, b-hydrox-
yphosphonate, was formed, but only in modest yields
(33–35%) and was accompanied by appreciable amounts
of diethyl methylphosphonate (20–25%). Furthermore
the reaction required, to proceed, the addition of an
equimolar amount of trimethylsilyl chloride/bromide.
noethane (meso and racemate), arising from
pinacolic coupling reaction.
a
Once optimised, the protocol was applied to the lactone
of 2,3,5-tri-O-benzyl–b-D-arabinofuranose, as model
compound for aldonolactones, useful precursors of C-
glycosides and sugar-fused b-hydroxy and b-amino
acids.8 The reaction proceeded in very good yield
(85%) and afforded the two diastereoisomeric products
in 2/1 diastereoisomeric ratio.
In a typical procedure, a solution of the carbonyl com-
pound (1 mmol) and the halo compound (1 mmol) in
tetrahydrofuran (2.0 mL) was added portion-wise at
room temperature and under vigorous stirring to a mix-
ture of samarium iodide (0.1 M solution in THF,
1.0 mL) and activated (with 1,2-dibromoethane) magne-
sium turnings (0.15 g) in THF (1.0 mL). The course of
the reaction was monitored by observing the colour of
the solution which turned blue to yellow and then back
to blue for several times during the addition of the or-
ganic reagents.9 At the end of the reaction, monitored
by thin layer chromatography (silica; eluting with n-hex-
ane/ethylacetate), the reaction mixture was filtered. The
remaining magnesium turnings were washed with tetra-
hydrofuran and reused. The organic phase was diluted
with ethyl acetate and poured into crushed ice. The
aqueous phase was extracted three times with ethyl ace-
tate (3 · 6 mL). The combined organic extracts were
dried (Na2SO4) and the solvent removed under reduced
pressure, to afford a crude material, which was purified
by column chromatography on silica gel.
To investigate whether an organomagnesium com-
pound might be involved in the carbon–carbon bond
formation, in some cases parallel experiments were
performed (tert-butyl a-bromo acetate with trans-
cinnamaldehyde or with benzaldehyde; benzyl
a-bromoacetate with cinnamaldehyde or with 2,3,5-tri-
O-benzyl-D-arabino-1,4-lactone). The bromoester and
the electrophile were added at room temperature to
two reaction flasks: one containing a substoichiometric
amount of SmI2 and magnesium in tetrahydrofuran,
the other only magnesium in tetrahydrofuran. The
amount of bromoester, electrophile, magnesium was
the same in the parallel experiments and so were the
other experimental parameters (molarity, speed of
addition of the reagents, stirring and time of reaction).
In the magnesium-only experiments, appreciable
amounts of reagents were recovered unchanged. Fur-
ther experiments evidenced that longer times and/or
heating were required for the Mg-only reactions, which
were, however, less reproducible and strongly depen-
dent on the experimental parameters, most of all mag-
nesium activation. An added advantage, when Sm(II) is
present, is the colour of the solution that continuously
changes during the addition of the organic reagents
and that allows a visualisation of the course of the
reaction.
In conclusion we have developed a new protocol for the
synthesis of b-hydroxyesters and b-hydroxynitriles via
samarium-mediated Reformatsky-type reactions, with
some advantages with respect to the corresponding stoi-
chiometric reactions: reduction of the samarium and the
solvent (tetrahydrofuran) amounts; greater reproduc-
ibility; easier monitoring of the course of the reaction
and timely visualisation of the end point, with conse-
quent minimisation of the side products. The results ob-