244
H.-W. Shih et al. / Tetrahedron Letters 53 (2012) 243–246
Table 2
OAc
O
AcO
AcO
AcO
Examination of reaction factors (solvent and time)a
O
P
+
Yieldb (%)
O
OH
OH
Entry
Solvent
T (°C)
Time (min)
HO
2
1
2
3
4
5
6
7
8c
Toluene
Acetone
1,4-Dioxane
THF
Acetonitrile
DMF
100
100
100
100
100
100
90
60
60
60
60
60
60
15
60
13
7
6
17
28
45
89
11
α:β = 10:1
1
2
OAc
O
AcO
AcO
AcO
conditions
in Table 3
O
O
O P
OH
Pyridine
Pyridine (80 W)
2
90
a
3
Reagents and conditions: 1 (0.4 mmol), 2 (1.5 equiv) and CCl3CN (2 mL) in the
desired solvent (entries 1–7) was irradiated at 120 W in a sealed tube.
b
Scheme 1. Model reaction used for the screening of coupling conditions.
Isolated yield.
The reaction was irradiated at 80 W.
c
Table 1
Examination of coupling conditions for the preparation of glycophospholipid 3a
This encouraging observation promoted us to further
investigate various solvent systems, and the results were shown
in Table 2. Polar aprotic solvents such as DMF, pyridine and aceto-
nitrile (Table 2, entries 5–7) gave better yields than less polar ones.
These results can be rationalized by higher dielectric constants of
solvents and easy charge separation of the phosphate anion from
its counter ion in polar solvents. Notably, pyridine was the best
Entry
Reagents and conditions
T (°C)
Time
Yieldb (%)
1
2
3
4
5
6
7
8
TPSCl/py.
DCC/py.
80
80
80
90
80
80
90
90
>72 h
>72 h
>72 h
48 h
60 min
60 min
60 min
15 min
64
9
5
PPh3/DIAD/py.
CCl3CN/py.
TPSCl/py. (
PPh3/DIAD/py. (
DCC/py. ( w 120 W)
CCl3CN/py. ( w 120 W)
47
41
l
w 120 W)
c
lw 120 W)
—
solvent in our study to prepare glycophsopholipid
3 under
l
13
89
microwave irradiation (120 W, 90 °C, 15 min, Table 2, entry 7). In
contrast, changing irradiation power from 120 to 80 W, a signifi-
cant decrease in reaction yield was obtained (Table 2, entry 7 vs 8).
To demonstrate the efficiency and the scope of the microwave-
assisted synthesis, various sugar phosphates and alcohols were
chosen to investigate the reaction feasibility to form glycophosp-
holipids or phosphate diesters using CCl3CN as an activating re-
agent in weak basic conditions (pyridine). As shown in Table 3,
the reaction’s progress was strongly dependent on the electronic
or steric property of the alcohol. For example, farnesol (4, entry
2) was converted to the corresponding glycophospholipid 5 at
80 W within 15 min, but solanesol (6, C45H73OH, entry 3) required
a longer time (60 min) to consume all the starting materials. Pre-
sumably, this long-chain lipid in 6 did not make the reaction mix-
ture homogeneous to perform a uniform heating pattern. Besides,
uridine-50-monophosphate glucopyranoside (12) could also be pre-
pared by microwave-assisted conjugation of 8 and 11 in 71% with-
in 15 min (Table 3, entry 5). Notably, when an alcohol such as 14
containing a labile tetrachlorophthalic (TCP) moiety was coupled
with a phosphate (entry 6), the reaction conditions should be
milder (70 °C, 60 W) in order to obtain a reasonable yield (62%, en-
try 6). Secondary alcohols such as benzyl 2-hydroxypropanoate
(16) and 1-phenylpropan-2-ol (29) were also applied to success-
fully generate desired adducts 17 and 30, respectively (entries 7
and 12). Excitingly, diverse phosphonate monoesters could also
be smoothly prepared via this method (Table 3, entries 9–11). As
we know, during the preparation of glycophospholipids via a
typical glycosylation reaction from activated sugars and lipid-
phosphates, the neighboring participation could affect the anomer-
ic selectivity. Thus, more synthetic steps might be required to
prepare specific sugar building blocks in order to control the de-
sired anomeric selectivity.12 In contrast, the fixed ratio of anomers
in our starting glycophosphates such as 1, 8, and 13 would directly
convert to the corresponding glycophospholipid products without
changing their anomeric ratio (see in Table 3).
l
a
Reagents and conditions: 1 (0.4 mmol), 2 (1.5 equiv) and coupling reagents
(3.0 equiv) in pyridine.
b
Isolated yield.
Not detected.
c
times required (>72 h).1,3 These poor results urged us to develop a
more practical method for the preparation of glycophospholipids.
Recently, Kalek and Stawinski have reported the synthesis of
mono- and di-arylphosphinic acids via microwave-assisted palla-
dium-catalyzed cross coupling.4 Their work inspired us to evaluate
microwave conditions to prepare our target molecules. Theoreti-
cally, a phosphorus–oxygen bond formation by the coupling of
phosphates or phosphonates to alcohols might be suitable under
microwave irradiation because these molecules can absorb the
microwave energy more efficiently through the ionic conduction
mechanism.5 Over the last two decades, microwave-assisted or-
ganic synthesis (MAOS) has become a powerful technique to boost
the reaction rates and to reduce the reaction time with an
improvement in the yield and quality of the product.5 However,
to the best of our knowledge, a microwave-assisted phosphorus
oxygen bond formation in a one-pot manner has not been studied
yet. Herein, we report the development of simple and convenient
method for the synthesis of structurally diverse glycophospholi-
pids, phosphate diesters, and phosphonate monoesters via micro-
wave irradiation with specific coupling conditions.
Our investigation was started from evaluating the efficiency of
coupling reagents for the assembly of mannopyranosidyl mono-
phosphate (1)6 with 3,7,11-trimethyl-dodecan-1-ol (2) under ther-
mal and microwave-assisted conditions (Scheme 1 and Table 1).
As shown in Table 1, under thermal conditions (entries 1–4),
better yields were given by using TPSCl and CCl3CN as coupling re-
agents instead of DCC or PPh3/DIAD but a longer time (>48 h) was
required in all reactions. With the assistance of microwave irradi-
ation, the yields were still too low (<13%) and messy results were
showed in either DCC-mediated conditions or Mitsunobu reaction
(entries 6 and 7).3,7 Fortunately, when CCl3CN-mediated condi-
tions8 were applied with pyridine as the solvent, the yield was dra-
matically improved to 89% within 15 min at 120 W (Table 1, entry
8). Presumably, the mechanism of this one-pot synthesis could be
proposed to generate an iminophosphate as a key intermediate.8
Next, D-arabinofuranosyl-1-monophosphorylsolanesol (31, DPA
analogue), a key natural product-like compound for mycobacte-
rium cell wall biosynthesis, was synthesized to demonstrate the
efficiency of our methodology (Scheme 2). The TBS protected
D
-arabinose-monophosphate 1810 was chosen as our starting
material to conjugate with a polyisoprenol, solanesol 6, at 90 °C
for 60 min under 80 W irradiation, followed by global desilylation