Aryl H-phosphonates. Part 11. Synthetic and 31P NMR studies on the formation of aryl nucleoside H-phosphonates

Article abstract of DOI:10.1039/a907150d

The formation of aryl H-phosphonate diesters from the corresponding nucleoside 3′-H-phosphonates and various phenols was investigated in detail using ³¹P NMR spectroscopy. The major reaction pathways of aryl nucleoside H-phosphonates under diff

Full text of DOI:10.1039/a907150d

View Article Online / Journal Homepage / Table of Contents for this issue
31
Aryl H-phosphonates. Part 11. Synthetic and P NMR studies on
the formation of aryl nucleoside H-phosphonates†
Jacek Cieslak,a Marzena Szymczak,a Malgorzata Wenska,a Jacek Stawinski and
Adam Kraszewski*a
a Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14,
61-704 Poznan, Poland
b Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University,
S-106 91 Stockholm, Sweden
Received (in Lund, Sweden) 1st September 1999, Accepted and transferred from Acta Chem.
Scand. 28th September 1999
The formation of aryl H-phosphonate diesters from the corresponding nucleoside 3Ј-H-phosphonates and various
phenols was investigated in detail using 31P NMR spectroscopy. The major reaction pathways of aryl nucleoside
H-phosphonates under different experimental conditions have been identified and a general method for a clean
and efficient generation of these valuable synthetic intermediates was developed.
The H-phosphonate methodology, due to its efficiency, reli-
ability and experimental simplicity, has emerged in the last
decade as a versatile and powerful approach to the synthesis of
biologically active phosphates and their analogues.1,2
making use of aryl H-phosphonates as intermediates, the
chemistry of this class of compounds needs to be better under-
stood. In this paper we address the most important factors,
particularly those related to the reactivity of the P–H bond in
these compounds, which may affect the formation and stability
of nucleoside aryl H-phosphonates under various experimental
conditions. Since reactivity of aryl H-phosphonate diesters,
especially those of a synthetic value, may exceed that of H-
phosphonic acyl mixed anhydrides, these compounds have to
be viewed in a majority of cases as sensu stricto reactive inter-
mediates. In this context, their clean and efficient in situ
generation rather than the isolation were the major objectives
of our work.14
As part of our studies in this field, we have recently reported
on aryl H-phosphonates as a new type of active H-phosphonate
derivatives,3–8 particularly useful in the synthesis of nucleo-
side phosphoramidates,7 for the aminoalkyl functionalization
of support-bound oligonucleotides,8 or in the preparation of
dinucleoside H-phosphonates via the transesterification
approach.5 We also found that inexpensive, commercially
available diphenyl H-phosphonate can be the reagent of
choice for simple and efficient preparation of nucleoside
H-phosphonate monoesters4 or various synthetically useful
alkyl H-phosphonates.
The main advantages of aryl H-phosphonates as synthetic
intermediates stem from the fact that these compounds, in
principle, possess only one electrophilic centre located on the
phosphorus atom and, in contradistinction to other reactive H-
phosphonate species, e.g. mixed H-phosphono-acyl anhydrides,
their reactivity can be modulated by changing electronic and/
or steric properties of substituents on the aromatic ring of an
aryl moiety. These features significantly widen the synthetic
applications of H-phosphonate methodology by enabling the
syntheses of compounds otherwise difficult to prepare, e.g.
nucleoside H-phosphoramidates.7
Results and discussion
The synthesis of a nucleoside aryl H-phosphonate can be
carried out either by phosphonylation of a suitable protected
nucleoside with an appropriate phosphonylating reagent
bearing an aryl moiety15 or by reacting a nucleoside H-phos-
phonate with an appropriate phenol in the presence of a con-
densing agent.7 We have pursued the latter approach (Scheme 1)
In contradistinction to various dialkyl H-phosphonates,
e.g. dinucleoside H-phosphonate,9,10 glycero alkyl H-phos-
phonates,11 that can be prepared efficiently using the standard
H-phosphonate approach, there are a few inherent problems
associated with the synthesis of some aryl H-phosphonate
diesters. These can be traced back to the highly electrophilic
character of the phosphorus centre in reactive aryl H-
phosphonates which leads to some additional reaction pathways,
not available for dialkyl H-phosphonates. Indeed, our previous
studies have shown, for example, that diphenyl H-phosphonate
undergoes under anhydrous basic conditions a rearrangement12
to triphenyl phosphite and phenyl H-phosphonates (a reaction
not observed for dialkyl H-phosphonates), in addition, the
activation pattern of these compounds can differ from that of
simple alkyl H-phosphonate derivatives.13
Scheme 1
since it alleviates problems of preparing separate phos-
phonylating reagents for each kind of aryl H-phosphonate
derivative and thus, in light of easily accessible H-phosphonate
monoesters, seems to be more convenient and versatile.
To secure efficiency and reproducibility of synthetic methods
† Part 10. I. Kers, J. Stawinski and A. Kraszewski, Tetrahedron, 1999, 55,
11579.
J. Chem. Soc., Perkin Trans. 1, 1999, 3327–3331
This journal is © The Royal Society of Chemistry 1999
3327

View Article Online
Table 1 31P NMR dataa of aryl nucleoside H-phosphonates of type 3 and other reaction products
Compound
δ/ppm
1JHP/Hz
3JHP/Hz
Compound
δ/ppm
1JHP/Hz
3JHP/Hz
3a
3b
3c
3d
3e
3f
3.34, 3.50b
4.63
719.3, 721.1
725.8
712.6, 732.6
737.9
737.1
750.8
769.3, 760.6
—
—
—
—
—
—
—
7.4c
7.4d
8.3c
8.3d
8c
8d
125.97, 125.48b
125.69, 126.46b
126.23e
—
—
—
—
—
—
—
—
—
—
—
—
8.3g
9.3g
9.3g
9.3g
9.3g
6.4g
6.5g
6.5g
6.5g
6.5g
6.7g
6.9g
4.44, 4.47b
4.63
8e
8f
8g
124.8, 125.96b
127.19, 127.42b
Ϫ6.44, Ϫ5.56b
Ϫ6.42, Ϫ6.54b
Ϫ6.55, Ϫ6.67b
Ϫ6.45, Ϫ6.63b
Ϫ6.66, Ϫ6.88b
Ϫ6.73, Ϫ6.85b
Ϫ8.02, Ϫ8.29b
e
4.04
4.07
—
—
e
11a
11b
11c
11d
11e
11f
11g
3g
7a
7b
7c
7d
7e
7f
2.44, 2.87b
128.03
128.10
127.63
127.91
128.13
126.45
130.57
8.3d
9.3g
9.3f
9.3f
9.3f
9.3f
8.3f
10.2f
7g
a In pyridine. b Two diastereoisomers. c Two doublets of doublets. d Doublet of doublets. e Broad singlet. f Doublet. g Two doublets.
The presence of an electron-withdrawing aryloxy group
at the phosphonate centre causes the P–H bond in aryl H-
phosphonate diesters to be usually more acidic than that in
dialkyl H-phosphonates and may facilitate conversion of the
tetracoordinated phosphonates into tervalent phosphite forms.
Although 31P NMR spectra of these compounds show the
presence of only the phosphonate form,12,13 one cannot exclude
the possibility that a significantly higher concentration exists
below the detection level of 31P NMR spectroscopy, of the
phosphite form in aryl H-phosphonates. Since most of the
synthetic applications of H-phosphonates take advantage of
the high stability of the H-phosphonate function under the
reaction conditions, changes in the phosphonate–phosphite
equilibria may cause some synthetic complications. Indeed, a
propensity of some aryl H-phosphonates to form tervalent
derivatives of type 7–9 (Chart 1) is consistent with this assump-
tion. To remedy this we tried to pin-point the most important
factors affecting the stability of the phosphonate form of aryl
H-phosphonate diesters. For this purpose the formation and
stability of nucleoside aryl H-phosphonates 3 were investigated
as a function of the aryl moiety, type of condensing agent used,
and basicity of the reaction medium.
Condensations of H-phosphonate 1 with phenols 2 in pyridine in
the presence of pivaloyl chloride
Initially, we attempted the synthesis of various nucleoside
aryl H-phosphonates 3 under standard condensation con-
ditions that have been shown to give clean and quantitative
formation of dinucleoside H-phosphonate diesters.16,17 To this
end nucleoside H-phosphonate 1 (Scheme 1) was reacted with
various phenols 2 (1.2 molar equiv.) in the presence of pivaloyl
chloride (4, 3.0 molar equiv.) in neat pyridine. 31P NMR spectra
of the reaction mixtures revealed that the efficiency of the for-
Chart 1
mation of the desired product 3 varied significantly, depending
on the kind of phenol used. Aryl H-phosphonates 3 bearing
electron-donating groups on the aromatic ring (2,4,6-trimethyl-
and 4-methylphenyl derivatives 3a and 3b) were formed readily
(ca. 3 min) and virtually quantitatively. In the reactions of
unsubstituted phenol 2c or 4-chlorophenol 2d the formation
of the corresponding aryl H-phosphonate 3c or 3d was less clean
(ca. 92% and 85% of the desired product, respectively), and the
reaction mixtures were contaminated by side products which
resonated in the chemical shift range of tervalent phosphorus
derivatives (species of type 7 and 8, Chart 1 and also Table 1).
This tendency was even more pronounced for more acidic
phenols investigated (2,4-dichloro- 2e, 4-nitro- 2f and 2,4,6-
trichloro- 2g derivatives). In these instances, the corresponding
aryl H-phosphonates 3 were the minor reaction products
(3e, 47%; 3f, 6%; 3g, 12%), formed with various proportions of
tervalent species 7, 8 and 9. Upon standing, compositions
of these reaction mixtures underwent further changes, and
additional side products, 11 and 12, started to form (see
below).
The product distribution in the above reactions deserves
some comment since it sheds light on general trends in the
reactivity of aryl H-phosphonate diesters. It is likely that due to
lower nucleophilicity the acidic phenols (e.g. 2d–g) will couple
slower than alkanols or less acidic phenols (e.g. 2a–c) and thus
the primary activation product of 1, the mixed phosphonic
carboxylic anhydride 10,18 may undergo further activation with
pivaloyl chloride under the reaction conditions to produce
bispivaloyl phosphite 9.18 This is an unfavourable situation
since the latter species can react with the phenol present giving
a mixture of side products, pivaloyl aryl 8 and bisaryl phosphite
7. One should note that aryl pivaloyl phosphites 8 can also be
3328
J. Chem. Soc., Perkin Trans. 1, 1999, 3327–3331

View Article Online
formed in a subsequent reaction of the produced aryl H-
phosphonates 3 and pivaloyl chloride, and this is probably a low
energy pathway for 3 with electron-withdrawing substituents
on the phenyl ring, e.g. 3d–g. For these reasons, the amounts
of tervalent side products 7, 8 and 9 and their relative
ratio19 correlated well with acidity of the phenol used for the
condensation.
similar extent in the pivaloyl chloride promoted condensations
discussed above. However, due to different activation path-
ways of H-phosphonates by pivaloyl chloride18 and those
by chlorophosphates,23–25 the disproportionation of aryl H-
phosphonates 3 could be clearly detected only in the latter
instances.
Although the formation of 3a and 3b was a clean reaction,
these compounds were not completely stable under the reaction
conditions and upon standing underwent a slow reaction with
excess pivaloyl chloride to produce P-acylated products 11.
The formation of 11 most likely involves intermediacy of 8
(produced from 3), but since 8a and 8b are probably formed
slowly but undergo fast P-acylation, we did not observe these
intermediates using 31P NMR spectroscopy. However, for aryl
H-phosphonates 3d,e, the formation of the corresponding 8
was significantly faster, and since their P-acylation seemed to be
rather slow (electronic effects), we observed in these instances
the accumulation of 8d,e and then their gradual transformation
to 11d,e. For the most acidic phenols investigated (e.g. 4-
nitrophenol, 2,4,6-trichlorophenol), significant formation of
pivaloyl phosphite 9 and its higher reactivity in the P-acylation
than those of 8f or 8g led to the formation of the P-pivaloylated
species 12, rather than 11f or 11g.
Having established this general reactivity pattern of 3 in
pyridine, we attempted to suppress the formation of side
products 7–12 by optimizing the reaction conditions. By using
less phenol 2 (1.1 molar equiv.) and reducing the amount of
pivaloyl chloride to 1.1 equiv we were able to generate rapidly
(less than 3 min) and cleanly aryl H-phosphonates 3a–d. For the
reaction of 1 with more acidic phenols (2e–g) we also observed an
increase in the yield of the desired product (3e, 60%; 3f 26%; 3g,
22%), but these improvements were hardly satisfactory.20
Condensations of H-phosphonate 1 with phenols 2 in methylene
chloride
Although aryl nucleosides H-phosphonates 3a–d could be
produced rather cleanly in pyridine in the presence of chloro-
phosphates 5 or 6 as condensing agents, those bearing aryl
moieties with strong electron-withdrawing substituents (3e–g)
underwent a significant disproportionation under the reaction
conditions.
To suppress this reaction pathway, we carried out the con-
densation of nucleoside H-phosphonate 1 with phenols 2
(1.1 molar equiv.) in methylene chloride using variable amounts
of pyridine. Since 3 molar equiv. of pyridine relative to the
condensing agent 4, 5 or 6 secured stability of the 5Ј-O-
dimethoxytrityl group, we used this amount of base in all
experiments in methylene chloride.
Pivaloyl chloride 4 (1.1 molar equiv.) proved to be a useful
condensing agent (coupling time ca. 5 min) under these condi-
tions but only in the synthesis of aryl nucleoside phosphonates
3a–d. For more acidic phenol derivatives 3e–g, the synthesis
was less efficient (3e, ca. 90%; 3f, 76%; 3g 72%) although the
products did not undergo disproportionation. The observed
low yields of the generation of 3 were due to the formation of
phosphonic pivalic mixed anhydride 10 (ca. 10% in the instance
of 3e, 24% for 3f and 28% for 3g; 31P NMR) probably occurring
via an attack of the pivaloate anion on electrophilic phosphorus
centres in 3e–g. In agreement with these, the ratio 3e–g:10
varied with addition to the reaction mixtures of the appropriate
phenol or pivalic acid, thus indicating the existence of a real
equilibrium between these two species.
Condensations of H-phosphonate 1 with phenols 2 in pyridine in
the presence of chlorophosphates 5 and 6
To eliminate the formation of side products with the pivaloyl
moiety (8–12), we attempted to use chlorophosphates as con-
densing agents, which are also known to efficiently promote
the formation of H-phosphonate diesters.21,22 Indeed, using
diphenyl phosphorochloridate 6 (1.2 molar equiv.) or 2-chloro-
All problems experienced in the previous syntheses of aryl
nucleoside H-phosphonates 3 have been alleviated by using
diphenyl phosphorochloridate 6 (1.1 equiv.) in methylene
chloride in the presence of pyridine (3 molar equiv.). The
coupling of nucleoside H-phosphonate 1 with all investigated
phenols 2a–g was clean, relatively fast (ca. 20 min),26 and the
produced aryl H-phosphonates 3a–g did not undergo any
detectable changes within a few hours (31P NMR spectroscopy).
In light of these results, we considered this protocol as a general
procedure for the in situ formation of aryl H-phosphonates
3a–g. For less acidic phenol derivatives, however, other reac-
tion conditions (e.g. condensation in pyridine) can also be used,
if so desired.
In conclusion, we have found that the most important factors
in the synthesis of nucleoside aryl H-phosphonates from
nucleoside H-phosphonates 1 and the corresponding phenols
2 are (i) acidity (pKa) of the phenols; (ii) the nature of the
coupling agent used for the condensation, and (iii) basicity of
the reaction medium. These factors affect the yields of the
condensation to produce 3, the extent of the subsequent
reactions of the product 3 with the condensing agent and the
rate of disproportionation of 3, and thus ultimately determine
the efficiency of generating aryl H-phosphonates as synthetic
intermediates. Taking all these factors into account we have
designed a synthetic protocol consisting of almost equimolar
amounts of all reactants (nucleoside H-phosphonate mono-
ester, the appropriate phenol and diphenyl phosphoro-
chloridate) in methylene chloride containing a few molar
equivalents of pyridine. These reaction conditions provided
stability for the acid-sensitive protecting group (e.g. the dimeth-
oxytrityl function) and secured a rapid and clean formation
of nucleoside aryl H-phosphonates 3, even those bearing aryl
moieties with strongly electron-withdrawing substituents.
5,5-dimethyl-2-oxo-1,3,2-dioxaphosphinane
5
(2.5 molar
equiv.) in pyridine, the formation of 3a–d from 1 and the corre-
sponding phenol 2 was practically quantitative and complete in
less than 5 min. We observed, however, that in contradistinction
to 3a and 3b, aryl H-phosphonate 3d underwent a slow decom-
position in pyridine yielding equimolar amounts of diaryl
nucleoside phosphite 7d and the starting material 1 (ca. 13%
after 2 h). Compounds 7d and 1 can be formed via dispro-
portionation of 3d under anhydrous basic conditions, a
reaction previously observed for diphenyl H-phosphonate12
and thus it seems that phosphite–phosphonate dispropor-
tionation can be a general reaction pathway accessible for
reactive aryl H-phosphonate diesters. Since the rate of the
disproportionation should grow with increasing electron-
withdrawing character of an aryl moiety, it was anticipated that
nucleoside aryl H-phosphonates 3f and 3g should undergo this
reaction significantly faster than compounds 3a–d.
The product distribution in the reaction of nucleoside H-
phosphonate 1 with more acidic phenols (2e–g) in the presence
of chlorophosphate 5 or 6 in pyridine seemed to support this
assumption. The yields of the desired products 3 indeed
decreased with the increasing acidity of the phenol used (after
5 min ca. 66% for 3e, 42% for 3f to 33% for 3g), while the
amount of the disproportionation products, 1 and 7, increased
along the series [ratio (1 ϩ 7):3 ca. 1:2 for 3e, 1.5:1 for 3f, and
2:1 for 3g; 31P NMR spectroscopy].
Since disproportionation of aryl H-phosphonates 3 does not
involve a condensing agent it probably also occurred to a
J. Chem. Soc., Perkin Trans. 1, 1999, 3327–3331
3329

View Article Online
1.44 (3H, 2s), 2.30, 2.32 (3H, 2s), 2.43, 2.62 (2H, 2m), 3.38 (2H,
Experimental
m), 3.77 (6H, s), 6.40 (1H, m), 5.33 (1H, m), 6.40 (1H, m), 6.82,
6.85 (4H, 2d, J = 7.0 Hz), 6.96–7.41 (13H, m), 7.07 and 7.08
(1H, 2d, 1JHP = 725.6, 723.8 Hz), 7.48 (1H, m), 8.86 (1H, br m,
exch. D2O) (multiplicity of some signals due to the presence of
P diastereoisomers); for the 31P NMR data, see Table 1. Anal.
C38H39N2O9P: C, 65.32; H, 5.63; N, 4.01. Found: C, 65.22; H,
5.67; N, 3.94%.
1H and 31P NMR spectra were recorded on a 300 MHz
spectrometer. The 31P NMR experiments were carried out in
5 mm tubes using 0.1 mol LϪ1 concentration of phosphorus-
containing compounds. For column chromatography, Kieselgel
60 Merck was used. The amount of water in solvents was
measured with Karl Fisher coulometric titration. Methylene
dichloride was dried over P2O5, distilled and kept over
molecular sieves 4 Å until the amount of water was less than
10 ppm. Pyridine was stored over molecular sieves 4 Å until
the amount of water was below 20 ppm. Triethylamine was
distilled and stored over CaH2. 5Ј-O-Dimethoxytritylthymidine
3Ј-H-phosphonate (triethylammonium salt) 14 and 2-chloro-
5,5-dimethyl-2-oxo-1,3,2-dioxaphosphinane 527 were prepared
according to published procedures. Diphenyl phosphoro-
chloridate and phenols 2a–d were commercial grade from
Aldrich. Pivaloyl chloride (Aldrich) was distilled and stored in a
refrigerator.
Phenyl 5Ј-O-dimethoxytritylthymidin-3Ј-yl phosphonate (3c).
1
Yield 64%; H NMR δH (in ppm, CD2Cl2) 1.43 and 1.44 (3H,
2s), 2.43, 2.62 (2H, 2m), 3.85 (2H, m), 3.77 (6H, s), 4.22 (1H,
m), 5.33 (1H, m), 6.39, 6.42 (1H, m), 6.81, 6.84 (4H, 2d, J = 7.0
Hz), 7.09, 7.12 (1H, 2d, 1JHP = 727.67 and 725.58 Hz), 7.09–7.41
(14H, m), 7.47 (1H, m), 8.83 (1H, br s exch. D2O); for the
31P NMR data, see Table 1. Anal. C37H37N2O9P: C, 64.91; H,
5.45; N, 4.09. Found: C, 64.86; H, 5.48; N, 4.03%.
The reference compounds used for the identification of
some of the reaction products or intermediates were obtained
as follows: diaryl nucleoside phosphites 7a–g, by reaction of
H-phosphonate 1 in pyridine with pivaloyl chloride (3 equiv.)
followed by the addition of the appropriate phenol 2a–g
(4 equiv.); aryl pivaloyl nucleoside phosphites 8c–g, by reac-
tion of the appropriate aryl H-phosphonate 3c–g with
pivaloyl chloride (3 equiv.); bispivaloyl nucleoside phosphite 9,
by pivaloylation of H-phosphonate 1;18 aryl nucleoside pival-
oylphosphonates 11, by condensation of P-pivaloylated nucleo-
side phosphonates28 with appropriate phenol and diphenyl
chlorophosphate.
Acknowledgements
We are indebted to Professor Maciej Wiewiórowski and to
Professor Per J. Garegg for their interest and helpful discussions.
Financial support from the State Committee for Scientific
Research, Republic of Poland [3 T09A 118 14], the Swedish
Natural Science Research Council, and the Swedish Found-
ation for Strategic Research is gratefully acknowledged.
References
1 A. Kers, I. Kers, A. Kraszewski, M. Sobkowski, T. Szabó, M.
Thelin, R. Zain and J. Stawinski, Nucleosides Nucleotides, 1996,
15, 361.
General procedure for the formation of aryl nucleoside phos-
phonates 3a–g
2 J. Stawinski, in Handbook of Organophosphorus Chemistry, ed.
R. Engel, Marcel Dekker, New York, 1992, pp. 377–434.
3 M. Sobkowski, J. Stawinski, A. Sobkowska and A. Kraszewski,
J. Chem. Soc., Perkin Trans. 1, 1994, 1803.
4 J. Jankowska, M. Sobkowski, J. Stawinski and A. Kraszewski,
Tetrahedron Lett., 1994, 35, 3355.
5 J. Cieslak, M. Sobkowski, A. Kraszewski and J. Stawinski, Tetra-
hedron Lett., 1996, 37, 4561.
6 A. Kers, J. Stawinski, L. Dembkowski and A. Kraszewski, Tetra-
hedron, 1997, 53, 12691.
7 A. Sobkowska, M. Sobkowski, J. Cieslak, A. Kraszewski, I. Kers
and J. Stawinski, J. Org. Chem., 1997, 62, 4791.
8 M. Sobkowski, A. Kraszewski and J. Stawinski, Nucleosides Nucleo-
tides, 1998, 17, 253.
Nucleoside H-phosphonate 1 (1 molar equiv.) and appropriate
phenol 2a–g (1.1 molar equiv.) were rendered anhydrous by
repeated evaporation of the added excess of pyridine. The
residue was dissolved in methylene chloride (1 mL per 0.1 mmol
of 1) containing pyridine (3 molar equiv.) and treated with
diphenyl chlorophosphate (1.1 molar equiv.). In all investigated
instances the reaction was completed after ca. 20 min (31P
NMR) affording 3a–g as the only nucleotidic material. Thus the
aryl nucleoside H-phosphonates 3a–g produced could be used
for further reactions.
9 P. J. Garegg, I. Lindh, T. Regberg, J. Stawinski, R. Strömberg and
C. Henrichson, Tetrahedron Lett., 1986, 27, 4051.
10 B. C. Froehler and M. D. Matteucci, Tetrahedron Lett., 1986, 27,
469.
11 I. Lindh and J. Stawinski, J. Org. Chem., 1989, 54, 1338.
12 A. Kers, I. Kers, J. Stawinski, M. Sobkowski and A. Kraszewski,
Tetrahedron, 1996, 52, 9931.
13 A. Kers, I. Kers, A. Kraszewski and J. Stawinski, Collect. Czech.
Chem. Commun. (Special Issue), 1996, 61, S246.
14 Some less reactive aryl nucleoside H-phosphonates were isolated
and characterized. See the Experimental section.
15 V. Ozola, C. B. Reese and Q. L. Song, Tetrahedron Lett., 1996, 37,
8621.
Synthesis and isolation of aryl nucleoside H-phosphonates 3a–c
The solution of aryl nucleoside H-phosphonates 3a–c (pro-
duced as above, 0.1 mmol in 1 mL of CH2Cl2) was diluted
with methylene dichloride (5 mL) and washed with saturated
aqueous NaHCO3 (3 × 5 mL). The organic phase was
separated, dried (Na2SO4 anhydrous), and concentrated to an
oil. The residue was purified by silica gel column chromato-
graphy using ethyl acetate as an eluent. Fractions containing
pure products 3a–c were collected, evaporated and freeze-dried
from benzene to afford white amorphous powders.
16 J. Stawinski, R. Strömberg and R. Zain, Tetrahedron Lett., 1992, 33,
2,4,6-Trimethylphenyl
5Ј-O-dimethoxytritylthymidin-3Ј-yl
1
3185.
phosphonate (3a). Yield 74%; H NMR δH (in ppm, CD2Cl2)
1.41 and 1.43 (3H, 2s), 2.21, 2.47 (9H, 2s), 2.44, 2.60 (2H, 2m),
3.38 (2H, m), 3.77 (6H, s), 4.20 (1H, m), 5.32 (1H, m), 6.38,
6.41 (1H, 2m), 6.81, 6.84 (4H, 2d, J = 7.0 Hz), 7.13 and 7.17
17 H. Almer, J. Stawinski, R. Strömberg and M. Thelin, J. Org. Chem.,
1992, 57, 6163.
18 P. J. Garegg, T. Regberg, J. Stawinski and R. Strömberg, Nucleosides
Nucleotides, 1987, 6, 655.
19 In the reactions of 1 with 2c and 2d, no bispivaloyl phosphite 9 was
present in the reaction mixtures and the major side products were
those corresponding to diaryl phosphites 7 and aryl pivaloyl
phosphites 8. For more acidic phenols, 2e–g, the amount of
9 increased along the series, and for a highly acidic penta-
chlorophenyl, it was the major reaction product (no formation of
nucleoside pentachlorophenyl H-phosphonate was observed in this
instance).
1
(1H, 2d, JHP = 722.1, 720.2 Hz), 6.87–7.50 (11H, m), 8.78
(1H, br s, exch. D2O) (multiplicity of some signals due to the
presence of P diastereoisomers); for the 31P NMR data, see
Table 1. Anal. C40H43N2O9P: C, 66.11; H, 5.96; N, 3.85. Found:
C, 66.16; H, 6.01; N, 3.83%.
4-Methylphenyl 5Ј-O-dimethoxytritylthymidin-3Ј-yl phosphon-
20 In these reactions, the unreacted starting material 1 and the side
1
ate (3b). Yield 71%; H NMR δH (in ppm, CD2Cl2) 1.43 and
products 7 ϩ 8 ϩ 9 amounted to 19% and 21% (for 2e), 40% and
3330
J. Chem. Soc., Perkin Trans. 1, 1999, 3327–3331

View Article Online
34% (for 2f), 41% and 38% (for 2g), respectively (the composition of
the reaction mixtures after ca. 5 min).
21 P. J. Garegg, T. Regberg, J. Stawinski and R. Strömberg, Chem. Scr.,
1985, 25, 280.
22 R. Strömberg and J. Stawinski, Nucleic Acids Symp. Ser., 1987, 18,
185.
23 J. Stawinski, R. Strömberg, T. Szabó and E. Westman, Nucleosides
Nucleotides, 1989, 8, 1029.
25 P. J. Garegg, J. Stawinski and R. Strömberg, J. Org. Chem., 1987, 52,
284.
26 The coupling promoted by rather unreactive chlorophosphate 5
(2.5 molar equiv.) was significantly slower, and was complete after
ca. 5 h.
27 R. L. McConnell and H. W. Coover, J. Org. Chem., 1959, 24, 630.
28 A. Kume, M. Fujii, M. Sekine and T. Hata, J. Org. Chem., 1984, 49,
2139.
24 J. Stawinski, R. Strömberg, M. Thelin and E. Westman, Nucleosides
Nucleotides, 1988, 7, 601.
Paper 9/07150D
J. Chem. Soc., Perkin Trans. 1, 1999, 3327–3331
3331

View Article Online