516
M. Bollmark et al. / Tetrahedron Letters 43 (2002) 515–518
To elaborate a comprehensive methodology for utiliz-
ing H-phosphonoselenoates as synthetic intermediates
we set out to develop general methods for the prepara-
tion of H-phosphonoselenoate monoesters and for their
efficient conversion into H-phosphonoselenoate
diesters.
protocol has the advantage that selenization occurs
under homogenous conditions and the purification of
nucleoside H-phosphonoselenoate 3 (ca. 80% yield)
does
not
require
additional
reverse-phase
chromatography.
For the synthesis of H-phosphonoselenoate diester 5,
first we tried the reaction conditions developed previ-
ously for their thio congeners.12 To this end, nucleoside
H-phosphonoselenoate 3 and a suitably protected
nucleoside 4 (1.2 equiv.) in acetonitrile–pyridine (4:1,
v/v) were treated with diphenyl phosphorochloridate
As a viable synthetic route to H-phosphonoselenoate
monoesters (e.g. nucleoside 3%-H-phosphonoselenoate 3,
Scheme 1) we considered selenization of phosphinate
intermediates 2, analogously to sulfurization during the
preparation of H-phosphonothioate monoesters.7 Since
alcohols can be efficiently converted in situ into the
corresponding phosphinate esters, this approach might
provide a convenient access to H-phosphonoselenoates
of type 3.
(DPCP, 2.5 equiv.). The reaction was rapid (<5 min, 31
P
NMR), completely chemoselective, and produced the
desired dinucleoside H-phosphonoselenoate diester 5
1
(lP=76.0 and 75.3 ppm, JPH=653 and 657 Hz) practi-
cally quantitatively.13 However, isolation of 5 posed
certain problems. Firstly, excess hydroxylic component
4 was difficult to remove by silica gel chromatography
and, secondly, H-phosphonoselenoate 5 appeared to
hydrolyze partly during work-up. To overcome these
difficulties, we used a slight excess of the nucleotidic
component 3 to ensure that all nucleoside 4 was con-
sumed during the coupling, and to increase the
hydrolytic stability of product 5 during work-up, we
reduced the amount of pyridine used for the reaction.
The generation of nucleoside phosphinate 2 (lP=14.22
1
3
ppm, JPH=572 Hz, JPH=9.8 Hz, td) in pyridine from
the corresponding nucleoside 1 (1.5 equiv.) and triethyl-
ammonium phosphinate in the presence of pivaloyl
chloride (1.5 equiv.) was uneventful and proceeded in
virtually quantitative yield as revealed by 31P NMR
spectroscopy. However, the next step, selenization of
the produced phosphinate 2, required considerable
experimentation to develop the best conditions of sol-
vents and reagents.8 In pyridine, the formation of H-
phosphonoselenoate
3
(lP=48.2 and 47.2 ppm,
The efficacy of these new reaction conditions was
assessed by carrying out the condensation of nucleoside
H-phosphonoselenoate 3 (1.2 equiv.) with nucleosidic
component 4 (1.0 equiv.) in acetonitrile containing pyr-
idine (5 equiv.) using DPCP (2.5 equiv.) (Scheme 2).
The reaction turned out to be only slightly slower than
that when more pyridine was used (completed within 10
min) and produced cleanly, dinucleoside H-phospho-
noselenoate 5. As expected, 5 obtained in this way was
more stable during work-up and could be isolated by
silica gel chromatography in ca. 80% yield (vide infra).
1JPH=564 and 569 Hz, 1JPSe=713 and 713 Hz) was
rather slow (3–4 h) and its amount in the reaction
mixture did not exceed 30–40%. The 31P NMR spectra
usually revealed the generation of significant amounts
of selenophosphonic acid, H-phosphonate mono- and
H-phosphonate diesters (total 60–70%), probably due
to the competing selenization of the phosphinate salt
and high propensity of phosphinates 2 for dispropor-
tionation. Sonication of the reaction mixture shortened
the reaction time to ca. 30 min but without noticeable
changes in product distribution. However, decreasing
the amount of pyridine in the reaction mixtures by
adding various co-solvents (chloroform, acetonitrile,
toluene) tended to suppress some side-product forma-
tion. Unfortunately, the reaction become rather slow
(overnight) and more side-products were formed as the
reaction approached completion. Eventually, the reac-
tion in chloroform in the presence of pyridine (2 equiv.)
and triethylamine (2 equiv.), led to a reasonably fast (4
h) and clean (ca. 80%, 31P NMR) formation of H-phos-
phonoselenoate 3. These conditions, when applied on a
preparative scale, afforded product 3 in 45–60% iso-
lated yield (vide infra).
In conclusion, we have developed efficient and simple
methods for the preparation of nucleoside H-phospho-
noselenoate monoesters and their conversion into
nucleoside H-phosphonoselenoate diesters. Since the
presence of the H-PꢀSe functionality can be exploited
in various oxidative transformations, the easy access to
H-phosphonoselenoates will expand and complement
the present H-phosphonate and H-phosphonothioate
methodologies for the preparation of biologically
important phosphate analogues.14 Further studies are in
progress in this laboratory.
We also assessed other selenizing reagents for the con-
version of phosphinate intermediate 2 into H-phospho-
noselenoate 3. 3H-1,2-Benzothiaselenol-3-one4,9 and
potassium selenocyanate,10 somewhat unexpectedly, did
not produce any H-phosphonoselenoate 3 but mixtures
of the corresponding symmetrical phosphoroselenoates
and H-phosphonate monoesters (31P NMR spec-
Typical protocol for the preparation of nucleoside H-
phosphonoselenoate 3
Method A: Anhydrous nucleoside 1 (1.5 equiv.) and
triethylammonium phosphinate (1.0 mmol) were dis-
solved in chloroform (16 ml) containing pyridine (2
equiv.). Elemental selenium (3.0 equiv.) and pivaloyl
chloride (1.5 equiv.) were then added and, after 10 min,
triethylamine (2.0 equiv.). The mixture was stirred for
ca. 4 h and, after excess selenium was filtered off, the
product was purified by silica gel chromatography
troscopy).
However,
using
triphenylphosphine
selenide11 (TPPSe) in combination with trimethylsilyl
chloride, furnished clean and fast selenization of phos-
phinate intermediate 2 (see Method B below). This