2492
K. E. Elson et al. / Tetrahedron Letters 45 (2004) 2491–2493
Table 1. Examples of useful synthetic transformations using polymer-supported triphenylphosphine ditriflate 1
,
Entrya b Substrate
Nucleophile
Product
Yield (%) (mp °C)
Lit. mp (°C)ref
1
4-NO2–C6H4–CH2–OH
4-NO2–C6H4–COOH
4-Nitrobenzyl 4-nitrobenzo-
ate
95 (165–167)
1686
2
4-CH3–C6H4–COOH
4-Cl–C6H4–CH2–OH
4-NO2–C6H4–COOH
Z-Gly-Phe-OH
4-CH3–C6H4–COOH
CH3–C(O)–SH
4-Toluic anhydride
4-Chlorobenzyl thioacetate
N-Benzyl 4-nitrobenzamide
95 (87–89)
91 (oil)
937
3
––2
4
5c
C6H5–CH2–NH2
H-Val-OMeÆHCl
4-MeO–C6H4–OH
96 (140–143)
65 (95–97)
88 (88–90)
141.5–1438
985
889
L,L-Z-Gly-Phe-Val-OMe
6
4-NO2–C6H4–CH2–OH
O-(4-Nitrobenzyl)-4-
methoxyphenol
(E)-Stilbene oxide
7
meso-
––
85 (64–66)
65–6710
C6H5CH(OH)CH(OH)C6H5
4-Cl–C6H4–CH2–OH
C6H5–C(O)–NH2
Cyclohexanol
d
8
NaN3
––
4-Chlorobenzyl azide
Benzonitrile
89 (oil)
88 (oil)
85 (49–51)
––2
––
47–4811
9
10
4-NO2–C6H4–COOHe Cyclohexyl 4-nitrobenzoate
a Reaction conditions: polymer-supported triphenylphosphine oxide (1.35 equiv), triflic anhydride (1.0 equiv), substrate (1.0 equiv), nucleophile
(1.0 equiv), diisopropylethylamine (3.5 equiv), DCM (10 mL).
b Reaction times: entries 1, 4 (2 h, rt); entries 2, 3, 5–8, 10 (overnight, rt); entry 9 (overnight, reflux).
c Excess of diisopropylethylamine (5.5 equiv) used.
d Added as a suspension in DMF.
e Active ester generated using DMAP (1.0 equiv).
salt was preformed by addition of 4-nitrobenzoic acid
to 1 in DCM. This order of addition ensured that
the primary amine did not react competitively with
In conclusion, the polymer-supported triphenylphos-
phine oxide was obtained by oxidation of the corre-
sponding commercially available phosphine with
hydrogen peroxide. The novel dehydrating reagent,
polymer-supported triphenylphosphine ditriflate 1, was
prepared by subsequent addition of triflic anhydride.
The potential of 1 as a general dehydrating reagent/
activating agent was displayed by the synthesis of esters
from primary and secondary alcohols, an anhydride,
thioacetate, amide, tripeptide, ether, epoxide, azide and
nitrile. The beauty of the polymer-supported triphenyl-
phosphine ditriflate 1 lies in the fact that the main
by-product, the phosphine oxide, remains attached to
the polymer-support. All products can be obtained
cleanly following filtration of the polymer beads and a
sodium hydrogen carbonate wash of the filtrate to
remove the diisopropylethylammonium triflate. An
additional advantage of this polymeric reagent 1 is that
after reaction, the polymer is again obtained as its oxide,
ready for reconversion to 1. The recovered polymer
could be recycled at least three times without loss of
efficiency. Thus, 1 is an effective dehydrating reagent,
which avoids the use of azodicarboxylates and chro-
matography to remove the phosphine oxide. Further
investigations to examine the scope and limitations of
this reagent are currently in progress.
1
to form the (unreactive) polymer-supported
aminophosphonium triflate. Subsequent addition of
benzylamine and diisopropylethylamine formed N-benz-
yl 4-nitrobenzamide in good yield (96%) after 2 h at
room temperature.4 After recycling the recovered poly-
mer three times, the yield of N-benzyl 4-nitrobenzamide
dropped from 96% to 92%. The use of 1 for peptide
bond formation was also explored. The extent of rac-
emisation was examined using Arteunisꢀ test5 (the cou-
pling of Z-Gly-Phe-OH and Val-OMeÆHCl). Addition of
Z-Gly-Phe-OH to a mixture of 1 and diisopropylethyl-
amine in DCM, followed by stirring at room tempera-
ture for 2 h formed the activated acid. Subsequent
treatment with Val-OMeÆHCl (in the presence of the
racemisation-suppressing agent 1-hydroxybenzotriazole,
HOBT) formed the desired tripeptide Z-Gly-Phe-Val-
OMe after stirring at room temperature overnight as a
single epimer (L,L) (none of the D,L-epimer was
1
observed by H NMR) in reasonable (65%) yield. Rac-
emisation occurred in the absence of HOBT. Ether
formation using 1 was examined. Addition of 4-meth-
oxyphenol and diisopropylethylamine to a mixture of
4-nitrobenzyl alcohol and 1 gave O-(4-nitrobenzyl)-4-
methoxyphenol in high yield (88%). In other examples,
(E)-stilbene oxide was synthesised in good yield (85%)
from 1, meso-hydrobenzoin and diisopropylethylamine.
Treatment of a solution of 1 and 4-chlorobenzyl alcohol
with a suspension of sodium azide in dimethylform-
amide (DMF) and diisopropylethylamine gave 4-chlo-
robenzyl azide in high yield (89%). Benzonitrile was
generated in good yield (88%) by refluxing 1 with
benzamide and diisopropylethylamine overnight.
Finally, the esterification of a secondary alcohol with
1 was investigated. Conversion of cyclohexanol to its
corresponding 4-nitrobenzoate ester was achieved in
good yield with 1 but only in the presence of an
activating agent such as 4-dimethylaminopyridine
(DMAP) or HOBT.
Supplementary material
A
31P gelphase NMR spectrum of polymer-supported
triphenylphosphine ditriflate 1 is available.
Acknowledgements
We gratefully acknowledge financial support for this
work from Griffith University, Natural Product Dis-
covery and AstraZeneca.