A R T I C L E S
Yamada et al.
equiv) were added neat to a 50 mL round-bottom flask. The reaction
was allowed to stir for 10 min and then checked by 31P NMR for
completion. Triethyl phosphite has a chemical shift of δ 138 ppm,
whereas the products had chemical shifts of approximately δ -4 ppm.
There was no need for purification as all reactions went to completion
and no side products were detected. The model compounds were
analyzed by 31P NMR, 13C NMR, and mass spectroscopy (see below).
Diethyl 2-Cyanoethyl Phosphonoformate (4). 6.7 g (0.05 mol, 1
equiv) of 2-cyanoethyl chloroformate was reacted with 8.6 mL (1 equiv)
of triethyl phosphite in a 50 mL round-bottom flask. Because the density
of the 2-cyanoethyl chloroformate was not known, the liquid was simply
weighed into a tared flask. The reaction was very rapid at room
temperature, creating heat and emitting bubbles of chloroethane. After
10 min, the reaction was checked by 31P NMR, and no starting material
was present. Yield: 95%, 11.2 g. 13C NMR (CDCl3): δ 164.8, 116.6,
64.5, 59.8, 17.4, 15.8. 31P NMR: δ -5.0. Electron impact mass
spectrometry gave a molecular ion of 236.0 m/e.
Diethyl Allyl Phosphonoformate (5). 10 mL (0.09 mol, 1 equiv)
of allyl chloroformate was reacted with 16 mL (1 equiv) of triethyl
phosphite in a 50 mL round-bottom flask. The reaction was very rapid
at room temperature, creating heat and emitting bubbles of chloroethane.
After 10 min, the reaction was checked by 31P NMR, and no starting
material was present. Yield: 95%, 18.9 g. 13C NMR (CDCl3): δ 165.3,
130.9, 119.8, 66.3, 64.5, 15.3. 31P NMR: δ -4.2. FAB mass
spectrometry gave a molecular ion of 223.0 m/e.
Diethyl Methyl Phosphonoformate (6). 4 mL (0.05 mol, 1 equiv)
of methyl chloroformate was reacted with 8.9 mL (1 equiv) of triethyl
phosphite in a 50 mL round-bottom flask. The reaction was very rapid
at room temperature, creating heat and emitting bubbles of chloroethane.
After 10 min, the reaction was checked by 31P NMR, and no starting
material was present. Yield: 95%, 9.3 g. 13C NMR (CDCl3): δ 165.7,
64.2, 52.1, 15.9. 31P NMR: δ -4.1. FAB mass spectrometry gave a
molecular ion of 197.0 m/e.
Diethyl â-(Diphenylmethylsilyl)ethyl Phosphonoformate (7). 8.9
g (0.03 mol, 1 equiv) of methyldiphenylsilylethyl chloroformate was
reacted with 5 mL (1 equiv) of triethyl phosphite in a 50 mL round-
bottom flask. Because the density of the DPSE chloroformate was not
known, the liquid was simply weighed into a tared flask. The reaction
was very rapid at room temperature, creating heat and emitting bubbles
of chloroethane. After 10 min, the reaction was checked by 31P NMR,
and no starting material was present. Yield: 95% 11.6 g. 13C NMR
(CDCl3): δ 165.7, 135.4, 134.4, 129.7, 128.2, 64.5, 64.4, 16.4, 15.5,
-4.0. 31P NMR: δ -4.3. Electron impact mass spectrometry gave a
molecular ion of 406 m/e.
Synthesis of Formic Acid, [Bis(N,N-diisopropylamino)phos-
phino]-â-(diphenylmethylsilyl)ethyl ester (12). The synthesis of the
phosphine was a three-step reaction. First, bis(N,N-diisopropylamino)-
chlorophosphine (26 g, 0.1 mol) was dissolved in dry THF in a 1 L
round-bottom flask. The solution was placed in an ice bath and cooled
to 0 °C. A 100 mL bottle of 1 M lithium aluminum hydride (0.1 mol)
in tetrahydrofuran (Sigma-Aldrich, Co) was transferred directly by
cannula into the flask containing the bis(N,N-diisopropylamino)-
chlorophosphine. The reaction was stirred for 10 min and then checked
by 31P NMR (δ ) 41.44 ppm). The ice bath was removed, and a heating
mantle along with a Friederich’s condenser was placed on the flask.
Sodium pieces (4.6 g, 0.2 mol) were added carefully to the flask. The
solution was allowed to reflux in tetrahydrofuran with stirring. After 2
h, the reaction was removed from the heating mantle and cooled to
room temperature. â-(Diphenylmethylsilyl)ethyl chloroformate (121.6
g, 0.4 mol) was dissolved in tetrahydrofuran in a 1 L round-bottom
flask. The phosphine reaction was cannulated into the chloroformate
solution, leaving behind unreacted sodium. (Unreacted sodium was
neutralized with 2-propanol.) The reaction was stirred for 1 h at room
temperature and then checked for completeness by 31P NMR. If
complete, the tetrahydrofuran was removed by rotary evaporation. The
resulting viscous oil was extracted three times with anhydrous hexanes.
potentially useful in various biochemical systems. They form
specific duplexes with complementary RNA, are stable toward
exo- and endonucleases, and activate RNase H1. Further
research is needed to test the value of these compounds toward
activation of RNase H1 in cells and to determine if phospho-
noformate DNA, in a manner similar to thiophosphonoacetate
DNA,13 transports into cells in the absence of cationic lipids.
Experimental Section
1
General Procedures. H NMR spectra were recorded on Varian
500 MHz and Varian 400 MHz spectrometers with tetramethylsilane
as an internal reference. 31P NMR spectra were recorded on a Varian
400 MHz spectrometer using an external capillary containing 85%
H3PO4 in D2O as a reference. Downfield chemical shifts were recorded
as positive values for 31P NMR. The University of Colorado Central
Analytical Laboratories performed ESI, EI, FAB, and accurate mass
spectroscopy analysis. Reverse phase (Zorbax 300SB C-18 column,
Agilent Technologies, Palo Alto, CA) chromatography was performed
on an Agilent Technologies Model 1100 HPLC. Solid phase DNA
synthesis was accomplished using an ABI model 394 automated DNA
synthesizer (Applied Biosystems, Foster City, CA) modified for the
synthesis cycle shown in Scheme 6. All reagents, columns, standard
2′-deoxynucleotide-3′-phosphoramidites, and 2′-deoxythymidine-3′-H-
phosphonate monomer (19a) were purchased from Glen Research
(Sterling, VA).
Unless otherwise noted, materials were obtained from commercial
suppliers and used without further purification. Anhydrous solvents were
purchased from Sigma-Aldrich Co. (Milwaukee, WI). Protected 2′-
deoxynucleosides (compounds 13a-d) were purchased from Chem-
Genes Corporation (Wilmington, MA). Medium-pressure, preparative,
column chromatography was performed using 230-450 mesh silica
gel from Sorbent Technologies (Atlanta, GA). Thin layer chromatog-
raphy was performed on aluminum-backed silica gel 60 F254 plates from
EM Sciences (Gibbstown, NJ).
Synthesis of Diphenylmethylsilylethyl Chloroformate. In a fume
hood, 500 mL of a 20% solution of phosgene (0.94 mol, 1 equiv) in
toluene was added into a 2 L round-bottom flask with a stir bar.36 The
flask was placed into an ice bath and cooled to 0 °C. In a flat-bottomed
500 mL flask, 214 mL of 2-(methyldiphenylsilyl)ethanol (0.94, 1 equiv)
was diluted into toluene, slowly cannulated into the phosgene solution,
and allowed to stir overnight, resulting in a greenish/brown clear
solution. Using a Teflon high vacuum pump, the reaction was
evaporated to produce a greenish/brown oil. The exit hose of the pump
was placed in a large flask of water to neutralize the excess phosgene.
No further purification was necessary. Typical yields were 95-97%.
1H NMR (CDCl3): δ 7.16-7.49 (m, aromatic, 10 H), δ 4.40 (t, 2 H),
δ 1.68 (t, 2 H), δ 0.61 (s, 1H). Electron impact mass spectrometry
gave a molecular ion of 304 m/e.
Synthesis of 2-Cyanoethyl Chloroformate. In a fume hood, 500
mL of a 20% solution of phosgene (0.94 mol, 1 equiv) in toluene was
added into a 2 L round-bottom flask with a stir bar. The flask was
placed into an ice bath and cooled to 0 °C. In a flat-bottomed 500 mL
flask, 23.4 g of 3-hydroxypropionitrile (0.33 equiv) was diluted into
toluene, slowly cannulated into the phosgene solution, and allowed to
stir overnight. The reaction was evaporated to an oil as described above.
1H NMR (CDCl3): δ 4.44 (t, 2 H), δ 2.8 (t, 2 H). Electron impact
mass spectrometry gave a molecular ion of 133 m/e.
General Procedure for Arbuzov Reactions. Model compounds
were synthesized using the following general procedures. All chloro-
formates were commercially available except the 2-cyanoethyl chlo-
roformate and the methyldiphenylsilylethyl chloroformate (syntheses
described above). Triethyl phosphite (1 equiv) and chloroformate (1
(36) Phosgene is a toxic and flammable gas and should be used with appropriate
safety precautions (see the Sigma-Aldrich Material Safety Data Sheet for
appropriate handling and safety instructions).
9
5258 J. AM. CHEM. SOC. VOL. 128, NO. 15, 2006