Carboxylic acids as promoters for internucleotide-bond formation via condensation of a nucleoside phosphoramidite and a nucleoside: Relationship between the acidity and the activity of the promoter

Article abstract of DOI:10.1016/j.tet.2004.12.017

The relationship between the activity and the acidity of a carboxylic acid as a promoter for condensation of a nucleoside phosphoramidite and a nucleoside was investigated. The investigation revealed that the acid, the pKa value of which in acetonitrile i

Full text of DOI:10.1016/j.tet.2004.12.017

Tetrahedron 61 (2005) 2203–2209
Carboxylic acids as promoters for internucleotide-bond formation
via condensation of a nucleoside phosphoramidite
and a nucleoside: relationship between the acidity
and the activity of the promoter
a
a
a
Yoshihiro Hayakawa,a,b, Toshihiro Iwase, Erkki J. Nurminen, Masaki Tsukamoto
*
and Masanori Kataokac
aGraduate School of Information Science, Nagoya University, Chikusa, Nagoya 464-8601, Japan
bCREST/JST, Nagoya University, Chikusa, Nagoya 464-8601, Japan
cNational Institutes of Natural Science, Research Center for Computational Science, Myodaiji, Okazaki 444-8787, Japan
Received 26 October 2004; accepted 8 December 2004
Available online 19 January 2005
Abstract—The relationship between the activity and the acidity of a carboxylic acid as a promoter for condensation of a nucleoside
phosphoramidite and a nucleoside was investigated. The investigation revealed that the acid, the pKa value of which in acetonitrile is less
than 18, is capable of promoting the condensation reaction, and acid with a pKa value outside of this range does not serve as a promoter. In
carboxylic acids serving as promoters, the ones with higher acidity generally show greater activity. In particular, acids with a pKa value less
than ca. 16 (measured by a potentiometric method), such as trichloroacetic acid (pKaZ10.6), trifluoroacetic acid (pKaZ12.7), dichloroacetic
acid (pKaZ13.2 or 15.8), and 2,4-dinitrobenzoic acid (pKaZ16.1), show high levels of activity higher than that of conventionally used
1H-tetrazole. These carboxylic acids generally serve as excellent promoters for both the liquid-phase and the solid-phase synthesis of
oligonucleotides via phosphoramidite method.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
In nucleotide synthesis conducted via phosphoramidite
method, one of the most important issues has been the
invention of a useful promoter for the condensation of a
nucleoside phosphoramidite and a nucleoside, forming an
internucleotidic phosphityl linkage.1 Previous investi-
gations2 have suggested that condensation using HX as a
promoter can be achieved via mechanism shown in Scheme
1. Thus, the reaction is initiated by protonation of the
phosphoramidite by HX. Then, the protonated phosphor-
amidite A undergoes the nucleophilic attack of XK, which is
generated in the first step, to produce an intermediate B with
nucleophilic to react with the protonated phosphoramidite,
a P–X bond. Finally, this species reacts with a nucleoside to
give a dinucleoside phosphite C. In this process, the rate-
Scheme 1. A suggested mechanism of condensation of a nucleoside
phosphoramidite and a nucleoside using HX as a promoter.
HX becomes a promoter with high reactivity.3
determining step is the condensation of the protonated
phosphoramidite A and XK. Accordingly, when HX is
Carboxylic acids may fulfil these requirements and accord-
ingly are expected to serve as effective promoters for this
sufficiently acidic, thereby giving a high concentration of
reaction.4 Currently, 1H-tetrazole and its derivatives are
conventionally used as the promoters, kinds of which are
the protonated phosphoramidite, and XK is sufficiently
limited. In contrast, there are many alternative carboxylic
acids available. Therefore, in cases when carboxylic acids
are useful, the choice of promoters is greatly increased.
Keywords: Nucleotides; Phosphoramidites; Carboxylic acids.
* Corresponding author. Tel.: C81 52 789 4848; fax: C81 52 789 5646;
e-mail: yoshi@is.nagoya-u.ac.jp
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.12.017

2204
Y. Hayakawa et al. / Tetrahedron 61 (2005) 2203–2209
Table 1. Activity of promotersa
Promoter
pKa value in CH3CNb
Yield of 25 (%)
1 min
5 min
Trichloroacetic acid (1)
2,6-Dihydroxybenzoic acid (2)
Trifluoroacetic acid (3)
Dichloroacetic acid (4)
2,4-Dinitrobenzoic acid (5)
Salicylic acid (6)
3,5-Dinitrobenzoic acid (7)
2,6-Dichlorobenzoic acid (8)
2-Nitrobenzoic acid (9)
4-Nitrobenzoic acid (10)
Benzoic acid (11)
10.6
12.6
12.7
13.2 or 15.8
16.1
16.7
17.0
17.6
18.2
18.7
20.7
20.8
99
ND
82
90
81
ND
55
NDc
2
99
99
99
0
76
74
76
5
3
0
0
55
ND
46
42
ND
ND
ND
ND
15
4-hydroxybenzoic acid (12)
Acetic acid (13)
1H-Tetrazole
22.3
14.5d
14.3d
BIT
97
a Estimated by yield of the product 25 (31P NMR measurement) obtained by 1- and/or 5-min reaction of 17 and 20, followed by TBHP oxidation.
b Unless otherwise stated, the values were determined by the potentiometric method reported in Ref. 7.
c Not determined.
d Data determined by 13C NMR analysis reported in Ref. 6.
Thus, it was considered of potential benefit to investigate the
utility of carboxylic acids as promoters5 of the condensation
of a nucleoside phosphoramidite and a nucleoside. Further-
more, the present investigation is expected to provide
potential solutions to the following important problem. As
mentioned above, the promoter must be highly acidic.
?A3B2 tlsb=.01w?>However, it is difficult to anticipate
how strong an acid must be in order to serve as an
efficient promoter, because there are no systematic
investigations of the relationship between the acidity and
the activity of the promoter. The reason why there are no
such investigations is that it is difficult to assess the
acidity of the existing promoters such as the tetrazole
derivatives6 in acetonitrile, a solvent generally used for
this type of reaction. In contrast, the pKa values of various
carboxylic acids in acetonitrile have been reported in the
literature.7 Thus, an investigation of the activity of various
kinds of carboxylic acids may provide us valuable
information for solving the problem regarding how high
acidity is required to an acid for serving as a good
promoter. This paper describes the results of an
investigation of these two subjects, that is, determining
the utility of carboxylic acids as promoters for the
formation of internucleotidic phosphite linkages using
the phosphoramidite approach, and describing the relation-
ship between the activity of the promoter and its acidity in
acetonitrile.
phosphite. Consequently, we estimated the activity accord-
ing to the yield of the dinucleoside phosphate 25, which was
quantitatively obtained by oxidizing the dinucleoside
phosphite. The condensation reaction was carried out
using 17 (48 mmol), 20 (40 mmol), and a promoter
(48 mmol)8 in the presence of powdery molecular sieves
3A9 in acetonitrile (1 mL) (25 8C, 1–5 min), and the
resulting phosphite product was converted to 25 by
oxidation with a 1.0 M TBHP/toluene solution.10 As a
control experiment, the synthesis was also carried out in a
similar manner using 1H-tetrazole or benzimidazolium
triflate (BIT)11 as a promoter for the first step. The yield of
25 was determined by 31P NMR analysis. Table 1 lists the
promoters examined, their pKa values in acetnitrile, and
the respective yields of 25.12 The results indicated that the
acidity of the promoter closely correlated with the activity
of the promoter. Roughly, acids with pKa values in
acetonitrile of less than 18.2 (measured by a potentiometric
method) are capable of promoting such condensation, but
acids bearing pKa values outside of this range do not serve
as promoters. Among the acids capable of serving as
promoters, those with smaller pKa values, that is, stronger
acids, generally show higher reactivity, as shown in
Figure 1. Acids with a pKa value lower than ca. 16 [e.g.,
trichloroacetic acid (pKaZ10.6), trifluoroacetic acid
(pKaZ12.7), dichloroacetic acid (pKaZ13.2 or 15.8), and
2,4-dinitrobenzoic acid (pKaZ16.1)] show particularly
favorable activity similar to that of BIT, and a higher
level of activity than that achieved with 1H-tetrazole.13
Surprisingly, no cleavage of the 50-O-DMTr group was
observed in the reactions, using trichloroacetic acid,
trifluoroacetic acid, and dichloroacetic acid, which easily
eliminate the DMTr protector in dichloromethane. Here, in
order to avoid the detritylation, it is crucial to use the acid
promoter in an equal molar amount or less than that of the
phosphoramidite. When the acid promoter was used
excessively, the detritylation took place to considerable
extent. In spite of their low pKa values, 2-hydroxybenzoic
acids, such as 2,6-dihydroxybenzoic acid (pKaZ12.6) and
salicylic acid (pKaZ16.9), did not serve as good promoters.
2. Results and discussion
We selected carboxylic acids 1–13, the acidities of which,
when in acetnitrile, have been reported as shown in Table
1.7 First, we screened these acids in terms of their ability to
promote the condensation of the thymidine 30-phosphor-
amidite 17 and the 50-O-free thymidine 20. Since the
dinucleoside phosphite formed by this condensation reac-
tion is too unstable to undergo various types of analysis, the
activity of the promoter in this series of examinations was
not determined at the stage of the formation of the

Y. Hayakawa et al. / Tetrahedron 61 (2005) 2203–2209
2205
Figure 1. Relationship between pKa and reactivity of the promoter in CH3CN estimated by yield of the product 25 (31P NMR measurement) obtained by 1- or
5-min condensation of 17 and 20, followed by TBHP oxidation. The pKa of 4 was plotted at the average value of the reported ones.
Using either of these acids as a promoter, the synthesis of 25
from 17 and 20 afforded the desired product in a lower than
10% yield, and gave many undesired products.
reaction may take place with the activated phosphoramidite
to form a sufficient amount of the phosphorous/carboxylic-
mixed anhydride, which corresponds to the reactive species
B shown in Scheme 1. In contrast, the 2-hydroxybenzoic
acid may generate a carboxylate such as 26, which lacks
nucleophilicity, because the anionic carboxylic oxygen
atom is masked by chelation with the hydrogen of the
2-hydroxy group. Therefore, the carboxylate function is not
sufficiently capable of nucleophilically attacking the proto-
nated phosphoramidite to form the reactive phosphorous/
carboxylic-mixed anhydride.14 Indeed, this interpretation of
the results was supported by 31P NMR studies of the
reaction of 17 and 2,4-dinitrobenzoic acid or salicylic acid.
Thus, the reaction of equimolar amounts of 17 (dia-
stereomers; 31P signals: d 148.6 and 148.7 ppm) and
2,4-dinitrobenzoic acid in the absence of a nucleoside in
CD3CN formed the corresponding phosphorous/carboxylic
mixed anhydride with a 31P signal at d 133.5 ppm [cf.
(n-C4H9O)2POCOCH3: d 132 ppm4]. This anhydride was
then instantaneously consumed by the addition of 20 to
quantitatively give the corresponding phosphite, producing
a
31P signal at d 140 ppm. On the other hand, no formation
of the phosphorous/carboxylic-mixed anhydride species
was observed in the reaction using salicylic acid as the
promoter. In this case, three 31P signals appeared at d 123.1,
123.3, and 123.7 ppm due to species other than the
phosphorous/carboxylic-mixed anhydride.
We next selected trichloroacetic acid and 2,4-dinitrobenzoic
acid as the representative carboxylic acids, which showed
high reactivity in the above-mentioned test, in order to
investigate their general utility as promoters for inter-
nucleotide-linkage formation using nucleoside 30-phosphor-
amidites other than 17, such as 14, 15, and 16. In this series
of investigations, the reaction was carried out using
0.12 mmol of a nucleoside phosphoramidite, 0.10 mmol of
a nucleoside, and 0.11 mmol of trichloroacetic acid or
0.12 mmol of 2,4-dinitrobenzoic acid in acetonitrile
(1.0 mL) containing MS 3A, and thus the concentration of
reactants (0.10–0.12 M) was higher than that (40–48 mM)
It remains unclear why the reactions using a nonhydroxy-
substituted benzoic acid or a 2-hydroxybenzoic acid as the
promoter gave such different results from each, but the
following explanation might account for the difference. By
protonation, the nonhydroxy-substituted benzoic acid may
produce a carboxylate with good nucleophilicity, and a

2206
Y. Hayakawa et al. / Tetrahedron 61 (2005) 2203–2209
Table 2. Synthesis of dideoxyribonucleoside phosphates using 1 or 5 as a
promoter
Amidite
Promoter
Product
Yield (%)a
14
14
15
15
16
16
17
17
1
5
1
5
1
5
1
5
22
22
23
23
24
24
25
25
99, 93b
99, 93b
98, 96b
99, 93b
98
99
99, 92b
99, 97b
a Estimated by 31P NMR analysis.
b Isolated yield.
used in the above-mentioned test. The reaction using either
acid as the promoter was completed in 5 min, and the
subsequent oxidation with TBHP for 10 min gave dinucleo-
side phosphates 22–25 in excellent yields as shown in
Table 2. In the preparation of these compounds, no side-
reaction, not even depurination or detritylation, was
detected.
The approach using trichloroacetic acid as the promoter
could be applied to the solid-phase synthesis of oligo-
nucleotides using the phosphormaidites 14, 15, 16,
a0nd 17 as building blocks. For example,
0
5 GpApCpTpCpTpCpTpTpApGpCpTpApApT3 (27) was
prepared starting from the nucleoside 21 attached to a
controlled-pore glass (CPG) on an Applied Biosystems
Model 392 DNA/RNA synthesizer, according to the
procedure shown in Table 3. The synthesis was investigated,
in which the coupling step was carried out by the following
two methods: soon after mixing a 0.1 M solution of a
phosphoramidite and a 0.08 M solution of trichloroacetic
acid, the resulting mixture was delivered over 15 s (method
A) or 30 s (method B) into the reaction column. As control
experiments, we also carried out the synthesis where a
0.35 M (method C) or 0.08 M (method D) solution of
1H-tetrazole for the activation of the phosphoramidite was
used in the coupling step and, after mixing this solution with
a 0.1 M solution of a phosphoramidite, the resulting mixture
was delivered over 30 s into the reaction column. Here,
method C is a current standard method conventionally used,
and was carried out for comparing efficiency of this method
with that of the novel methods using trichloroacetic acid as
the promoter. On the other hand, method D was carried out
to investigate efficiency of 1H-tetrazole as the promoter in
the synthesis using the low concentration (0.08 M) of
solution, which is a suitable one for trichloroacetic acid.
Method A and method B gave similar results and the target
oligonucleotide 27 was produced in an 88% overall yield (a
Figure 2. HPLC profile of the crude product of 27. (a) Prepared by method
A. (b) Prepared by method B. (c) Prepared by method C. (d) Prepared by
method D. Conditions: column, COSMOSIL 5C18-AR-II [4.6 (diameter)
mm!25 (height) cm]; flow rate, 1.0 mL/min; detection, 254 nm; eluent
and gradient, [AZa 100 mM triethylammonium acetate buffer solution in
H2O, BZa 20:80 mixture of H2O and acetonitrile, 0–60 min with a linear
gradient from A 100% to A 70%/B 30%]; temperature, 40 8C.
99% average coupling yield) by both these two methods.
As shown in Figure 2, HPLC analysis indicated that the
purity of the product was satisfactorily high, even in the
crude form. These results were similar to that obtained by
method C and much better than that obtained by method D
Table 3. Reaction sequence of the solid-phase synthesis
Step
Operation
Reagent(s)
Time (s)
1
2
3
4
5
6
7
8
9
Washing
Detritylation
Washing
Coupling
Washing
Capping
Washing
Oxidation
Washing
CH3CN
24
77
27
15–30
14
15
14
30
58
3% CCl3COOH/CH2Cl2
CH3CN
0.1 M amidite/CH3CNC0.08 M CCl3COOH/CH3CN
CH3CN
Ac2O/2,6-lutidine/THF (1:1:8)CN-methylimidazole/THF
CH3CN
1.0 M t-C4H9OOH/toluene
CH3CN

Y. Hayakawa et al. / Tetrahedron 61 (2005) 2203–2209
2207
(see Fig. 2) to indicate that trichloroacetic acid has higher
reactivity than that of 1H-tetrazole in the solid-phase
synthesis.
Model 392 DNA/RNA synthesizer using empty synthesis
columns (TWIST from Glen Research).
4.2. Materials and solvents
3. Conclusion
Acetonitrile and CD3CN were distilled from CaH2. Powdery
molecular sieves 3A (MS 3A) were used after drying the
commercially supplied sieves (Nacalai Tesque) at 200 8C
for 12 h. Other organic solvents were used after simple
distillation of the commercially supplied solvents. Acetic
acid purchased from Kishida was distilled from CrO3 in
the presence of acetic anhydride. Trichloroacetic acid
supplied from Kishida was dried under reduced pressure
(1–3 mm Hg) for O2 h before use. 2,6-Dihydroxybenzoic
acid (Aldrich), trifluoroacetic acid (Kishida), dichloroacetic
acid (Tokyo Kasei), 2,4-dinitrobenzoic acid (ACROS),
salicylic acid (Kishida), 3,5-dinitrobenzoic acid (Tokyo
Kasei), 2,6-dichlorobenzoic acid (Tokyo Kasei), 2-nitro-
benzoic acid (Tokyo Kasei), 4-nitrobenzoic acid (Tokyo
Kasei), 4-hydroxybenzoic acid (Kishida), benzoic acid
(Nacalai Tesque), 1H-tetrazole (Wako), N6-(benzoyl)-50-
O-(p,p0-dimethoxytrityl)-20-deoxyadenosine 30-(2-cyano-
ethyl N,N-diisopropylphosphoramidite) (14) (Applied Bio-
systems Inc. or G0len Research), N4-(benzoyl)-50-O-(p,p0-
dimethoxytrityl)-2 -deoxycytidine 30-(2-cyanoethyl N,N-
diisopropylphosphoramidite) (15) (Applied Biosystems
Inc. or Glen Research), N2-(iso0butyryl)-50-O-(p,p0-
dimethoxytrityl)-20-deoxyguanosine 3 -(2-cyanoethyl N,N-
diisopropylphosphoram0idite) (16) (Applied Biosystems Inc.
or Glen Research), 5 -O-(p,p0-dimethoxytrityl)thymidine
30-(2-cyanoethyl N,N-diisopropylphosphoramidite) (17)
(Applied Biosystems Inc. or Glen Research), and the
thymidine derivative 21 attached to CPG (Applied Bio-
systems Inc.), were commercially supplied. The 30-O-
In this study, we disclosed that various carboxylic acids
serve as useful promoters for the formation of internucleo-
tide linkage in acetonitrile via phosphoramidite approach.
This investigation also revealed that the ability of a
carboxylic acid (excluding a hydroxy-substituted carboxylic
acid) to serve as a promoter depends on its pKa value in
acetonitrile. Before the present study was conducted, it was
assumed that strong acids such as trichloroacetic acid and
trifluoroacetic acid could not be used as promoters, because
the conjugate base, that is, the trichloroacetate or trifluoro-
acetate anion, would not exhibit sufficient nucleophilicity
required for the promoter. However, this was fortunately not
the case.15 The present results indicated that, in considering
the ability of carboxylic acids to serve as promoters, concern
about the nucleophilicity of the corresponding carboxylate
anions is unwarranted. In the case of carboxylic acids,
acidity is the main factor controlling the promoter ability. In
other words, it is possible to predict the ability of a
carboxylic acid to serve as a promoter, provided its pKa
value is known. In general, acids with low pKa value, that is,
strong acids, have greater promoter activity, and acids with
the pKa values lower than 16 (measured by a potentiometric
method) exhibit the preferred reactivity. The result that
trichloroacetic acid, with the pKa value of 10.6 in
acetonitrile, served as an excellent promoter suggested
that organosulfonic acids16 or organophosphoric acids
bearing similar acidity to that of trichloroacetic acid might
also be used as promoters in the synthesis of nucleotides,
according to a phosphoramidite strategy. Accordingly,
investigation of the utility of organic acids other than
carboxylic acids is ongoing in our laboratory. Results
related to this topic will be reported in separate papers.
protected 20-deoxyribonucleosides, 18,17 19,17 and 20,18
a
1.0 M tert-butyl hydroperoxide/toluene solution,19 and
BIT11b were described in the literature or prepared by
previously reported methods.
4.3. Screening of promoter reactivity
4. Experimental
4.1. General
A mixture of the phosphoramidite 17 (35.8 mg, 48 mmol),
the nucleoside 20 (14.2 mg, 40 mmol), and MS 3A (30 mg)
in dry acetonitrile (0.8 mL) was stirred for 30 min. To this
mixture was added an acetonitrile (0.2 mL) solution of a
promoter (48 mmol). After 1 and/or 5 min, the reaction was
quenched by the addition of diisopropylethylamine (16 mL,
96 mmol) and a 1.0 M toluene solution of TBHP (0.08 mL,
0.08 mmol); stirring was then continued for 10 min. The
insoluble material was removed by filtration. The mixture
was concentrated, and the resulting residue was dissolved in
CDCl3 (1 mL). An aliquot of the solution was subjected to
31P NMR analysis to measure the product yield.
Unless otherwise stated, the NMR spectra were obtained in
CDCl3 on a JEOL JNM-400 or ECA-500 instrument. The
1H and 31P NMR chemical shifts are described as values in
ppm relative to (CH3)4Si and 85% H3PO4, respectively.
ESI-TOF HRMS spectra were measured using a Mariner
spectrometer. HPLC analysis was carried out using a
COSMOSIL 5C18-AR-II column [Nacalai Tesque, 4.6
(diameter) mm!25 (height) cm] on a Waters 2695
Separations Module chromatograph with a Waters 2996
Photodiode Array Detector. Nacalai Tesque silica gel 60
(neutrality, 75 mm) was used for column chromatography.
Unless otherwise stated, reactions were carried out at
ambient temperature. Reactions requiring anhydrous con-
ditions were carried out under an argon atmosphere in 10 or
20 mL Schlenk tubes dried by heating at 400 8C under
reduced pressure (1–3 mm Hg). Trityl assay was carried out
using a JASCO V-550 UV spectrometer. Solid-phase
syntheses were conducted on an Applied Biosystems
4.4. A typical procedure for the liquid-phase synthesis of
dideoxyribonucleoside phosphates
A mixture of the phosphoramidite 17 (89.2 mg, 0.12 mmol),
the nucleoside 20 (35.6 mg, 0.10 mmol), and MS 3A
(60 mg) in dry acetonitrile (0.8 mL) was stirred for
30 min. An acetonitrile (0.2 mL) solution of trichloroacetic
acid (18.0 mg, 0.11 mmol) or 2,4-dinitrobenzoic acid
(25.6 mg, 0.12 mmol) was then added to the mixture,

2208
Y. Hayakawa et al. / Tetrahedron 61 (2005) 2203–2209
which was stirred for an additional 5 min. To this mixture, a
1.0 M toluene solution of TBHP (0.24 mL, 0.24 mmol) was
added, and stirring was continued for 10 min. The insoluble
material was filtered off. Concentration of the filtrate
afforded a crude product, which was purified by column
chromatography (eluent, 20:1 CH2Cl2/CH3OH) to afford 25
(986 mg, 97% yield): 31P NMR (162 MHz) d K3.33;
HRMS (ESIC) calcd for C50H62N5O14PSiNa (MCNaC)
1038.37, found 1038.37. The 1H NMR spectrum (400 MHz)
was identical to those of an authentic sample.11b Dideoxy-
ribonucleoside phosphates 22,20 23,20 and 2420 were
COSMOSIL 5C18-AR-II [4.6 (diameter) mm!25 (height)
cm]; flow rate, 1.0 mL/min; detection, 254 nm; eluent and
gradient, [AZa 1.0 mM ammonium acetate buffer solution
in H2O, BZa 0.2 mM ammonium acetate buffer solution
in a 20:80 mixture of H2O and acetonitrile, 0–90 min with
a linear gradient from A 100% to A 70%/B 30%];
temperature, 40 8C. In both cases of 18 and 19, only a
peak due to the starting material was detected, indicating
that 18 and 19 underwent no decomposition under the above
reaction condition.
1
prepared in a similar manner. H NMR, 31P NMR, and
4.6. Solid-phase synthesis of
0
0
HRMS (ESIC) spectral data of these compounds were as
follows.
5 GpApCpTpCpTpCpTpTpApGpCpTpApApT3 (27)
Chain elongation of the oligodeoxyribonucleotide was
carried out on a 0.2 mmol scale according to the reaction
cycle shown in Table 3. The resulting fully protected
oligonucleotide was exposed to concentrated ammonia
at 25 8C for 2 h and then 55 8C for 12 h to afford the
target compound: HRMS (ESIK) of 27 calcd for
C156H196N54O96P15 (MK3HC) 1608.96, found 1608.93.
1
4.4.1. Compound 22. H NMR (400 MHz) d 0.10 (s, 6H),
0.90 (s, 9H), 1.91 (s, 3H), 2.18–2.29 (m, 2H), 2.65–2.85 (m,
3H), 3.10–3.15 (m, 1H), 3.41–3.50 (m, 2H), 3.78, 3.79 (2 s,
6H), 4.01–4.46 (m, 7H), 5.30–5.34 (m, 1H), 6.14, 6.22 (2 t,
1H, JZ6.8 Hz), 6.49–6.53 (m, 1H), 6.78–6.83 (m, 4H),
7.22–7.31 (m, 9H), 7.37–7.39 (m, 2H), 7.52–7.55 (m, 2H),
7.59–7.63 (m, 1H), 8.04, 8.06 (2 br, 2H), 8.14, 8.15 (2 s,
1H), 8.72, 8.75 (2 s, 1H), 9.16, 9.18 (2 s, 1H); 31P NMR
(162 MHz) d K3.15, K3.22; HRMS (ESIC) calcd for
C57H66N8O13PSi (MCHC) 1129.43, found 1129.34.
4.7. NMR analysis of the reaction of 17 and 20 using 5 as
a promoter
The phosphoramidite 17 (32.2 mg, 43 mmol) was weighed
into a dry NMR tube equipped with a Young’s tap and
dissolved in CD3CN (0.3 mL). To this sample, a solution of
the carboxylic acid 5 (9.3 mg, 44 mmol) in CD3CN (0.3 mL)
was added with a syringe under an argon stream. The NMR
4.4.2. Compound 23. 1H NMR (400 MHz) d 0.08, 0.09 (2 s,
6H), 0.89, 0.89 (2 s, 9H), 1.87, 1.90 (2 s, 3H), 2.14–2.41 (m,
3H), 2.67 (q, 1H, JZ6 Hz), 2.76 (t, 1H, JZ6 Hz), 2.94–2.99
(m, 1H), 3.43–3.53 (m, 2H), 3.79–3.80 (4 s, 6H), 3.98–4.03
(m, 1H), 4.11–4.31 (m, 4H), 4.37–4.42 (m, 2H), 5.13–5.18
(m, 1H), 6.19, 6.23 (2 t, 1H, JZ6.4 Hz), 6.27–6.30 (m, 1H),
6.85–6.88 (m, 4H), 7.22–7.42 (m, 11H), 7.48–7.52 (m, 2H),
7.57–7.62 (m, 1H), 7.92–7.94 (m, 2H), 8.11–8.13(m, 1H),
9.23 (br, 1H), 9.25 (br, 1H); 31P NMR (162 MHz) d K3.11,
K3.26; HRMS (ESIC) calcd for C56H66N6O14PSi (MC
HC) 1105.41, found 1105.44.
1
tube was sealed by the Young’s tap, and the H and 31P
NMR spectra were recorded. After the measurements were
carried out, a CD3CN (0.3 mL) solution of the nucleoside 20
(13.0 mg, 37 mmol) was introduced into the tube in a similar
manner for subsequent NMR analysis. The experiment was
conducted in a similar manner with trichloroacetic acid or
salicylic acid.
1
4.4.3. Compound 24. H NMR (400 MHz) d 0.10 (s, 6H),
0.90 (s, 9H), 1.16–1.29 (m, 6H), 1.99, 2.05 (2 s, 3H), 2.27–
2.42 (m, 2H), 2.64–2.82 (m, 5H), 3.30–3.42 (m, 2H), 3.73
(q, 1H, JZ7.2 Hz), 3.79, 3.79 (2 s, 6H), 3.91–4.00 (m, 1H),
4.18–4.39 (m, 5H), 4.56–4.63 (m, 1H), 5.23–5.31 (m, 1H),
5.83–6.11 (m, 1H), 6.14–6.30 (m, 1H), 6.81–6.84 (m, 4H),
7.21–7.39 (m, 9H), 7.73, 7.73 (2 s, 1H), 9.02, 9.84 (2 s, 1H),
10.06, 10.54 (2 s, 1H), 12.21, 12.33 (2 s, 1H); 31P NMR
(162 MHz) K2.34, K3.04; HRMS (ESIC) calcd for
C54H68N8O14PSi (M C HC) 1111.44, found 1111.51.
Acknowledgements
This work was supported in part by Grants-in-Aid for
Scientific Research (Nos. 16011223 and 16350086) from
the Ministry of Education, Science, Sports and Culture of
Japan, and Mitsui Chemicals, Inc.
References and notes
4.5. Stability of purine nucleosides 18 and 19 to
trichloroacetic acid in acetonitrile
1. Beaucage, S. L.; Iyer, R. P. Tetrahedron 1993, 49, 6123–6194
and references cited therein.
A mixture of the nucleoside 18 (45.2 mg, 0.10 mmol) or 19
(46.9 mg, 0.10 mmol) and MS 3A (60 mg) in dry aceto-
nitrile (0.8 mL) was stirred for 30 min. An acetonitrile
(0.2 mL) solution of trichloroacetic acid (18 mg,
0.11 mmol) was then added to the mixture, which was
stirred for an additional 30 min. To this mixture were added
a NaHCO3-saturated aqueous solution (ca. 2 mL) and
dichloromethane (ca. 3 mL), and stirring was continued
for 10 min. The lower organic layer was separated, dried
over sodium sulfate, concentrated, and then subjected to
HPLC analysis under the following conditions: column,
2. For the mechanism of the condensation reaction of a nucleo-
side phosphoramidite and a nucleoside using 1H-tetrazole as
HX, see: (a) Dahl, B. H.; Nielsen, J.; Dahl, O. Nucleic Acids
Res. 1987, 15, 1729–1743. (b) Barone, A. D.; Tang, J.-Y.;
Caruthers, M. H. Nucleic Acids Res. 1984, 12, 4051–4061.
(c) Berner, S.; Mu¨hlegger, K.; Seliger, H. Nucleic Acids Res.
1989, 17, 853–864. For related works, see: (d) Nurminen, E. J.;
¨
Mattinen, J. K.; Lonnberg, H. J. Chem. Soc., Perkin Trans. 2
1998, 1621–1628. (e) Nurminen, E. J.; Mattinen, J. K.;
¨
Lonnberg, H. J. Chem. Soc., Perkin Trans. 2 1999, 2551–2556.

Y. Hayakawa et al. / Tetrahedron 61 (2005) 2203–2209
2209
¨
(f) Nurminen, E. J.; Mattinen, J. K.; Lonnberg, H. J. Chem.
Soc., Perkin Trans. 2 2000, 2238–2240.
these pKa values, we can account for the reactivity of BIT
observed in this study; however the reactivity of 1H-tetrazole
appears to be too low, because 2,4-dinitrobenzoic acid, which
has a higher pKa value, that is, lower acidity, showed much
higher reactivity. This discrepant result might be explained as
follows. The pKa values of the carboxylic acids and those
of 1H-tetrazole and BIT were determined by fundamentally
different methods, that is, a potentiometric method was used
for the carboxylic acids, and a 13C NMR method was
employed for 1H-tetrazole and BIT; pKa values determined
by such different methods might not be consistent. As a
consequence, it might be irrelevant to discuss the relative
reactivity of carboxylic acids and 1H-tetrazole on the basis of
their pKa values reported to date.
3. Hayakawa, Y. Bull. Chem. Soc. Jpn. 2001, 74, 1547–1565.
4. There is one example of the formation of a trialkyl phosphite
via phosphoramidite method using a carboxylic acid as a
promoter: Evdakov, V. P.; Beketov, V. P.; Svergun, V. I.
J. Gen. Chem. USSR 1973, 43, 51–55.
5. Carboxylic acids as the promoters would overcome drawbacks
of tetrazole promoters, including expensiveness, explosive-
ness, and harmfulness to health.
6. To the best of our knowledge, with regard to the experi-
mentally obtained pKa values of tetrazole derivatives in
acetonitrile, that of 1H-tetrazole has only been reported in:
¨
Nurminen, E.; Lonberg, H. J. Phys. Org. Chem. 2004, 17,
1–17.
14. The following explanation provides another conceivable
reason. The carboxylate anion generated from 2-hydroxy-
benzoic acid has a hydroxy group that competes with the
carboxylate anion in the reaction with the activated phosphor-
amidite. Accordingly, two types of reactive species, namely, a
phosphorous/carboxylic-mixed anhydride and a phosphityl
phenoxide, may be generated. Between these two species, the
former may react with the 50-O-free nucleoside to give the
desired dinucleoside phosphite, but the latter may be inactive
under these reaction conditions, that is, may remain stable
without any changes. Accordingly, the subsequent oxidation
would give a complex mixture of a variety of compounds.
However, this possibility was excluded by the 31P NMR
analysis of the reaction of 17 and salicylic acid in the absence
of the nucleoside 20, which showed only scant formation of
both the phosphorous/carboxylic-mixed anhydride and phos-
phityl phenoxide intermediates.
7. Izutsu, K. Acid–Base Dissociation Constants in Dipolar
Aprotic Solvents: International Union of Pure and Applied
Chemistry Chemical Data Series No. 35; Blackwell Scientific:
Oxford, 1990; pp 17–35, and references cited therein.
8. In order to clearly observe a difference in the reactivity of
the promoter, the reaction had to be carried out using a low
(40–48 mM) concentration. At this concentration, the use of
17, 20, and a promoter at a molar ratio of 1.2:1.0:1.2 gave the
highest product yield shown in Table 1. When the reaction was
carried out using 17, 20, and a promoter in equimolar amounts,
the product yield was slightly lower. For example, in the
reaction using trichloroacetic acid as a promoter for 3 min,
the yield of 25 was 84%.
9. Hayakawa, Y.; Hirata, A.; Sugimoto, J.; Kawai, R.; Sakakura,
A.; Kataoka, M. Tetrahedron 2001, 57, 8823–8826.
10. Hayakawa, Y.; Uchiyama, M.; Noyori, R. Tetrahedron Lett.
1986, 27, 4191–4194.
15. In order to determine whether the reaction of a phosphor-
amidite using trichloroacetic acid as an activator will form the
corresponding phosphorous/carboxylic-mixed anhydride, we
carried out 31P NMR studies of the reaction of 17 and
trichloroacetic acid; however, in this case, no evidence
supporting the formation of the mixed acid anhydride was
obtained. Therefore, in cases involving such a strong acid as a
promoter, the condensation of a nucleoside 30-phosphorami-
dite and a 50-O-free nucleoside might not be achieved via
phosphorous/carboxylic-mixed anhydride as the reactive
intermediate, although the condensation might be achieved
by direct attack of a nucleoside on the protonated phosphor-
amidite. The detailed mechanism of the condensation of a
nucleoside phosphoramidite and a nucleoside using a strong
carboxylic acid as a promoter is now under investigation.
16. For example, the pKa value of methanesulfonic acid in
acetonitrile is ca. 10, which is similar to that of trichloroacetic
acid (see, Ref. 7).
11. (a) Hayakawa, Y.; Kawai, R.; Hirata, A.; Sugimoto, J.;
Kataoka, M.; Sakakura, A.; Hirose, M.; Noyori, R. J. Am.
Chem. Soc. 2001, 123, 8165–8176. (b) Hayakawa, Y.;
Kataoka, M.; Noyori, R. J. Org. Chem. 1996, 61, 7996–7997.
12. In the case using trichloroacetic acid as the promoter, reaction
conditions including the amount of the promoter and
concentration of the reactants affected the product yield.
When the reaction was carried out using trichloroacetic acid in
an equimolar amount to the phosphoramidite in a 40–48 mM
solution, the 1-min reaction gave the highest product yield.
The reaction for a longer period, for example, 5 min, led to
some (ca. 5%) decomposition of the product to decrease the
yield. In contrast, when the reaction was conducted using
trichloroacetic acid in the amount less equivalent (ca.
0.9 equiv) to the phosphoramidite in a 0.10–0.12 M solution
(see Table 2), no decomposition was observed after 5 min. The
activity of the promoter decreased in solvents other than
acetonitrile. For example, 2,4-dinitrobenzoic acid did not
successfully carry out the condensation of 17 and 20 in 5 min
in dichloromethane, toluene, THF, or DMF. In these cases,
when the reaction was stopped at 5 min and then the resulting
product was subjected to TBHP oxidation, 25 was obtained in
only 81 (dichloromethane), 66 (toluene), 12 (THF), and 9%
(DMF) yields, respectively.
17. (a) Kierzek, R.; Ito, H.; Bhatt, R.; Itakura, K. Tetrahedron Lett.
1981, 22, 3761–3764. (b) Saha, A. K.; Schairer, W.;
Waychunas, C.; Prasad, C. V. C. Tetrahedron Lett. 1993, 34,
6017–6020.
18. Ogilvie, K. K. Can. J. Chem. 1973, 51, 3799–3807.
19. Hill, J. G.; Rossiter, B. E.; Sharpless, K. B. J. Org. Chem.
1983, 48, 3607–3608.
13. The pKa values of 1H-tetrazole and BIT in acetonitrile are
reported in Ref. 6 to be 14.5 and 14.3, respectively. Based on
20. Kataoka, M.; Hattori, A.; Okino, S.; Hyodo, M.; Asano, M.;
Kawai, R.; Hayakawa, Y. Org. Lett. 2001, 3, 815–818.