From adenosine to 3′-deoxyadenosine: Development and scale up

Article abstract of DOI:10.1021/op000209x

A manufacturing process has been developed suitable for the production of 3′-deoxyadenosine (cordycepin, 3′-dA) in 20% yield from adenosine. The chemistry involves conversion of adenosine to isomeric 2′,3′-bromoacetates with isolation of the desired isomer in high purity. Acidic hydrolysis followed by hydrogenolysis afforded product with a purity of ≥99%. Unlike routes reported in the literature, intermediates are isolated as solids, thus avoiding the use of chromatography for isomer separation and final product purification.

Full text of DOI:10.1021/op000209x

Organic Process Research & Development 2000, 4, 601−605
From Adenosine to 3′-Deoxyadenosine: Development and Scale Up
Sayed Aman, D. Jason Anderson,1 Terrence J. Connolly,* Andrew J. Crittall, and Guijun Ji
Raylo Chemicals Inc., A Laporte Fine Chemicals Company, 8045 Argyll Road, Edmonton, Alberta, T6C 4A9 Canada
Abstract:
Results and Discussion
A manufacturing process has been developed suitable for the
production of 3′-deoxyadenosine (cordycepin, 3′-dA) in 20%
yield from adenosine. The chemistry involves conversion of
adenosine to isomeric 2′,3′-bromoacetates with isolation of the
desired isomer in high purity. Acidic hydrolysis followed by
hydrogenolysis afforded product with a purity of g99%. Unlike
routes reported in the literature, intermediates are isolated as
solids, thus avoiding the use of chromatography for isomer
separation and final product purification.
We embarked on a two-prong approach aimed at devel-
oping a scaleable process for the manufacture of 3′-dA.
Central to this approach was the use of 2-acetoxyisobutyryl
bromide (AIB, Mattock’s bromide) for the conversion of
adenosine to the corresponding adenosine bromoacetates 2
and 3. Robins has demonstrated the utility of Mattock’s
bromide as a reagent for the conversion of natural nucleosides
into various derivatives, including deoxy- derivatives.12-16
Several industrial based labs have also published work using
this reagent for the synthesis of deoxynucleosides.17-20 In
developing our route, it was planned that the generated
bromoacetates 2 and 3 could be used as common intermedi-
ates in either of our approaches, as outlined in Scheme 1.
The major concern for both routes was selectivity. To
proceed via the anhydro intermediate 4, we would capitalize
on the convergence of both isomeric bromoacetates to a
common intermediate, but then we would need to address
the regioselectivity of epoxide ring opening. A synthesis
based on reduction then hydrolysis is also prone to certain
pitfalls. As pointed out succinctly by Chattopadhyaya,7 in
approaches based on reduction of the halo-sugars, extensive
purification is required prior to reduction, or a separation of
2′- and 3′-deoxyadenosine must be done at the last stage.
Considering that we were aiming at scaling up to our pilot
plant, a process that involved careful chromatography was
not a suitable candidate.
In Chattopadhyaya’s synthesis of 3′-dA, the isomeric
bromo-acetates were converted to the common 2′-3′-anhy-
droadenosine 4 by treatment with Amberlite (OH-) resin.
Although the remaining steps to his synthesis would be
cumbersome on-scale due to successive protection, reduction,
and deprotection steps and two back-to-back purifications
on silica gel, we believed that his approach provided a good
starting point for process development.
Approach One: Epoxide Reduction. Epoxide 4 was
conveniently prepared from adenosine via a two-step process.
Addition of neat AIB to a suspension of adenosine in
acetonitrile followed by a basic work up afforded a mixture
of bromoacetates 2 and 3. Rapid conversion to epoxide 4
Introduction
Cordycepin (3′-deoxyadenosine, 3′-dA, 1) is a naturally
occurring antibiotic first isolated from the fungus Cordyceps
militaris.2 Being a 3′-deoxy nucleoside, its mode of action
is due to the chain-terminating ability at the 3′-terminus
during RNA synthesis. We have recently become interested
in this compound and decided to embark on a program aimed
at developing a route suitable for the large-scale manufacture
of 3′-dA.
A number of methods for the synthesis of 3′-dA exist in
the literature. These routes can be classified as belonging to
one of three types of synthesis: total synthesis from non-
nucleoside derivatives,3-5 metal hydride reduction of the
2′,3′-anhydro derivative,6-8 and reduction of the 3′-halo
derivatives using both noble metal catalysts9,10 and tin
hydrides.11,12 Since we were interested in obtaining samples
that were free from the R-anomer as well as having low levels
of the 2′-deoxy derivative, we focused our attention on the
latter two approaches which would use natural adenosine as
the starting material.
* Author for correspondence. Telephone: (780) 468-6060. Fax: (780) 468-
4784. E-mail: terry.connolly@raylo.laporteplc.com.
(1) University of Alberta Industrial Internship Participant, June 1998-May
1999.
(2) Cunningham, K. G.; Hutchinson, S. A.; Manson, W.; Spring, F. S. J. Chem.
Soc. 1951, 2299.
(3) McDonald, F. E.; Gleason, M. M. Angew. Chem., Int. Ed. Engl. 1995, 34,
350.
(4) McDonald, F. E.; Gleason, M. M. J. Am. Chem. Soc. 1996, 118, 6648.
(5) Ito, Y.; Shibata, T.; Arita, M.; Sawai, H.; Ohno, M. J. Am. Chem. Soc.
1981, 103, 6739.
(6) Hansske, F.; Robins, M. J. Tetrahedron Lett. 1985, 26, 4295.
(7) Bazin, H.; Chattopadhyaya, J. Synthesis 1985, 1108.
(8) Herdewijn, P. Tetrahedron 1989, 45, 6563.
(9) Russell, A. F.; Grenberg, S.; Moffatt, J. G. J. Am. Chem. Soc. 1973, 95,
4025.
(13) Robins, M. J.; Mengel, R.; Jones, R. A. J. Am. Chem. Soc. 1973, 95, 4074.
(14) Robins, M. J.; Jones, R. A.; Mengel, R. J. Am. Chem. Soc. 1976, 98, 8213.
(15) Robins, M. J.; Mengel, R.; Jones, R. A.; Fouron, Y. J. Am. Chem. Soc.
1976, 98, 8204.
(16) Robins, M. J.; Hansske, F.; Low, N. H.; Park, J. I. Tetrahedron Lett. 1984,
25, 367.
(10) Jain, T. C.; Jenkins, I. D.; Russell, A. F.; Verheyden, J. P. H.; Moffatt, J.
G. J. Org. Chem. 1974, 39, 30.
(11) Barton, D. H. R.; Subramanian, R. J. Chem. Soc., Perkin Trans. 1 1977,
1718.
(12) Robins, M. J.; Wilson, J. S.; Madej, D.; Low, N. H.; Hansske, F.; Wnuk,
S. F. J. Org. Chem. 1995, 60, 7902.
(17) Mansuri, M. M.; Starrett, J. E., Jr.; Wos, J. A.; Tortolani, D. R.; Brodfuehrer,
P. R.; Howell, H. G.; Martin, J. C. J. Org. Chem. 1989, 54, 4780.
(18) Bhat, V.; Stocker, E.; Ugarkar, B. G. Synth. Commun. 1992, 22, 1481.
(19) Manchand, P. S.; Belica, P. S.; Holman, M. J.; Huang, T.-N.; Maehr, H.;
Tam, S. Y.-K.; Yang, R. T. J. Org. Chem. 1992, 57, 3473.
(20) He, G.-X.; Bischofberger, N. Tetrahedron Lett. 1995, 36, 6991.
10.1021/op000209x CCC: $19.00 © 2000 American Chemical Society and The Royal Society of Chemistry
Published on Web 10/21/2000
Vol. 4, No. 6, 2000 / Organic Process Research & Development
•
601

Scheme 1
was achieved by switching the solvent to methanol and
adding potassium carbonate. Epoxide 4 was isolated as a
solid, following partial concentration of the stock solution.
Although Chattopadhyaya protected the 5′-hydroxyl group
as a 4-methoxytriphenylmethyl ether prior to LiAlH4-
mediated reduction,7 we chose to skip this on the basis of
Robins’ report of regioselective ring opening of the epoxide
with super-hydride.6 Although LiAlH4 did reduce the anhydro
intermediate 4, it was not selective, and chromatography on
silica gel was required. Other hydride reagents could have
been examined, but the dilute nature of most commercial
solutions of borohydrides, which would limit throughput, and
the possibility of chromatographic purification were consid-
ered as marks against this approach.
Approach Two: Bromoacetate Reduction. While work-
ing on approach one, it was noticed that a thick slurry formed
while concentration of the stock solution of bromoacetates
2 and 3 was being done. Filtration of this slurry and analysis
of the solid revealed that the desired isomer 2 had precipitated
in a purity of g98%. Although the recovery was rather poor
(∼ 30%) with significant losses of the desired isomer to the
mother liquor, we decided to investigate the reduction of 2.
The rather fortuitous precipitation of the desired isomer in
high purity would remove any troublesome purification at
later stages.
It was soon discovered, however, that we could not take
advantage of this serendipitous purification. The bromo-
acetate 2, once precipitated, was quite insoluble in solvents
commonly used for hydrogenolysis with heterogeneous
catalysts. To further complicate matters, once dissolved, the
desired reduction was quite slow with pressures of up to 30
psi hydrogen, and the dominant reduction pathway was
reduction of the glycosyl bond, as evident by the large
amount of adenine detected in the reaction mixtures. On the
basis of the indication of poor conversion of 2 to 5, it was
clear that another strategy was required.
Approach Three: Bromoacetate Hydrolysis and Re-
duction. We reasoned that the poor hydrogenolysis perfor-
mance was the result of two factors: Large groups at the
1′- and 5′-positions that prevented good contact of the
substrate and catalyst and poor solubility in solvents com-
monly used for hydrogenolysis. We speculated that removal
of the 5′-dioxolanyl and 2′-acetyl groups would allow for
greater solubility in polar protic solvents, as well as create
a less congested area for the interaction between the substrate
and catalyst.
Although there were reports in the literature on the
selective hydrolysis of compounds related to bromoaceteate
2 to bromodeoxy derivatives similar to 6 under basic
conditions using ammonia,21 in our hands we managed to
form only epoxide 4. On the basis of these difficulties, acidic
conditions were examined.
After considerable optimization, we found that the hy-
drolysis worked best using methanol as the solvent and
concentrated HCl as the acid source. Rather than using a
catalytic amount of acid, 5 mol equiv were found to be
optimal. This reaction was the most interesting and least
robust step of the process. The reaction proceeds in three
distinct stages that are easily observed by HPLC (Scheme
2). The 5′-dioxolanoyl group of 2 is first converted to an
acetate (7), which is then hydrolyzed to provide 8. The
intermediacy of 7 was determined by an independent
synthesis of 7 from 8. The last step involves hydrolysis of
the remaining 2′-acetate to afford 6, which was isolated as
the hydrochloride salt. This reaction appears to require a fine
balance of acid, water content, and reaction temperature. If
either the acid concentration, water content, or reaction
temperature was too low, then the reaction did not go to
completion, and the isolated product was the hydrochloride
salt of intermediate 8. Increasing the reaction temperature
(21) Dorland, E.; Serafinowski, P. Synthesis 1992, 477.
602
•
Vol. 4, No. 6, 2000 / Organic Process Research & Development

Scheme 2
while keeping the acid concentration and water content low
led to hydrolysis of 8 followed by rapid depurination. A
considerable effort was directed towards developing a robust
process for this stage. Although the reaction time was longer
than desired, increased stir out periods did not have a
negative impact on the product yield or purity, provided the
reaction temperature was maintained between 20 and 25 °C.
Following these reaction conditions led to an in-process
adenine content of less than 5%, and e0.1% in the final
product.
With a method in hand that provided the desired 3′-dA
precursor in a purity of 98%, being remarkably free of the
2′-dA precursor, we set out to convert the penultimate
product to 3′-dA. Attempts to carry the hydrochloride salt
of 6 forward through the hydrogenolysis resulted in limited
success. In cases where a reaction did occur, cleavage of
the glycosyl bond and generation of adenine became a
significant pathway. An intensive program was initiated,
aimed at finding reaction conditions that would buffer the
reaction, with the goal of minimizing the amount of adenine
formed during the hydrogenolysis. After considerable ex-
perimentation, it was discovered that that addition of 4-5
mol equiv of sodium acetate to a solution of 6 in 50/50
ethanol/water (w/w) effected smooth hydrogenolysis of the
carbon-bromine bond. Under these conditions, adenine
formation was kept to a minimum (typically e5%) when
palladium on carbon was used as the catalyst. As an added
benefit, the product precipitated from the filtered stock
solution upon standing.
Process Optimization
Having observed that the 3′-bromo-3′-deoxy isomers 6
and 8 precipitated so readily from solution during the
hydrolysis, it was decided to investigate skipping the filtration
at stage one and to carry a stock solution of isomeric
bromoacetates 2 and 3 forward. After the starting adenosine
disappeared (by TLC) the reaction mixture was quenched
and converted to a methanol stock solution. Hydrochloric
acid was then added and the reaction mixture stirred at room
temperature until the starting material 2 and intermediate 7
were each below 1% (by HPLC). It was also convenient,
1
and much faster, to monitor the reaction using H NMR.
The 1′- and 2′-protons of 2 appeared at 6.2 and 5.9 ppm,
respectively. Although there were very minor changes in the
chemical shifts of these protons as the hydrolysis proceeded
from 2 to 8 via 7, significant changes occurred once the 2′-
acetate was removed. Both the 1′ and 2′-protons were shifted
upfield to 5.9 and 4.9 ppm, respectively. From an in-process
control standpoint, it was convenient to observe the region
of the spectrum between 5.9 and 6.2 ppm as it slowly
converged to a single doublet. The reaction solvent did not
interfere with this region of the spectrum; therefore no sample
workup was required. Addition of a few drops of the reaction
mixture to d6-DMSO allowed us to determine when the
reaction was complete, at which time the reaction mixture
was cooled (0-5 °C) and filtered. The isolated solid was
determined to be the desired product 6; however, it was
significantly contaminated with adenine, adenosine, and the
undesired 2′-bromodeoxy isomer. The elevated levels of
adenine and adenosine pointed towards an incomplete
reaction at stage one and caused us to reexamine that stage
of the process.
Although the current process had the advantage of
avoiding chromatographic purification at all stages and
provided the final product in a purity >95%, the overall yield
was only ∼10%. However, the majority of the losses were
encountered at the first stage of the reaction when the pure
bromoacetate 2 was isolated in a yield of ∼30%. It was
believed that improvements at this stage of the reaction would
have a dramatic impact on yield. It was also realized that
this stage was also the weakest point in the process since
the purity of the product isolated set the stage for the purity
of the final product.
Isolation of the major products formed in the reaction of
adenosine and AIB using column chromatography showed
that in addition to the isomeric 5′-dioxolanyl bromoacetates
2 and 3, there was also a mixture of 2′,3′-bromoacetates that
had a free OH group at the 5′-position (8 and its 2′/3′-isomer),
which is consistent with reports in the literature.12,22 Although
(22) Schirmeister-Tichy, H.; Iacono, K. T.; Muto, N. F.; Homan, J. W.;
Suhandolnik, R. J.; Pfleiderer, W. HelV. Chem. Acta 1999, 82, 597.
Vol. 4, No. 6, 2000 / Organic Process Research & Development
•
603

Scheme 3
that would remove the high salt load. It was found that
addition of ethyl acetate to the filtered reaction mixture and
partial concentration of the organic layer afforded the product
as a white solid. Analysis of the final product revealed it to
be 99.0% pure, with a 2′-dA and adenine content of 0.1 and
0.2%, respectively, as well as being 0.4% ash by weight.
In conclusion, an efficient process for the conversion of
adenosine to 3′-deoxyadenosine that is amendable to further
scale up has been developed. The yield from adenine is 20%,
and unlike other preparations of 3′-dA, this one is remarkably
free of column chromatography. All intermediates are
isolated as filterable solids, following minimal workups. The
final product is extremely pure and obviates the need for
any further purification.
our reaction was conducted under anhydrous conditions, we
did not see any evidence for the formation of the 2′,3′,5′-
tris(dioxolanon-2-yl) byproduct as suggested by others.12,22
However, we did isolate material that appeared to have two
dioxolanonyl groups present (9). The absence of the tris-
(dioxolanon-2-yl) byproduct indicated that hydrolysis of 9
was likely the source of the observed adenosine. We
speculated that 9 could be an intermediate in the formation
of 2 and 3 and were gratified to see that treatment of 9 with
HBr resulted in conversion to 2, 3, and adenine (Scheme 3).
Since an aqueous work up followed stage one, it was
reasoned that running stage one at a slightly higher temper-
ature (∼30 °C) would result in conversion of intermediate 9
to either product or adenine. If adenine were formed, it would
be conveniently removed by the aqueous work up. By
limiting the amount of 9 carried forward, the adenosine and
adenine generated during the hydrolysis should be controlled.
It was also discovered during the stage one development
that the reaction worked better in a 50/50 (w/w) mixture of
acetonitrile/ethyl acetate, and addition of one back-extraction
of the aqueous layer with ethyl acetate increased the recovery
of the desired product. The additional back-extraction also
increased the adenine content, but incorporation of a brine
wash effectively removed it. Conversion of this acetonitrile/
EtOAc mixture to a methanol stock solution followed by
addition of concentrated HCl afforded, after the appropriate
stir out period, the desired product in a yield of 61% over
the two steps. This compares with a yield of 35-40% if the
intermediate bromoacetate was isolated. Unfortunately, HPLC
analysis of the isolated product showed that it was contami-
nated with 15-20% of the unwanted 2′-bromodeoxy isomer.
The final process development occurred when it was
decided to use the modified reaction conditions discussed
above and isolate the bromoacetate intermediate 2. The
optimized process afforded 2 in a 50% yield, and a purity
of g98%. The isolated product 2 was hydrolyzed to 6 under
the conditions previously described in yields of 70-75% with
no detrimental impact on product purity.
Experimental Section
Adenosine and acetoxyisobutyryl bromide were purchased
from commercial suppliers. All reactions were conducted
under a nitrogen atmosphere. Purity is reported as area
percentage HPLC performed under the following condi-
tions: column, Phenomenex Luna C18(2), 5 µm, 250 × 4.6
mm; solvent, MeCN, phosphate buffer (pH ) 6.0); flow rate,
1.0 mL/min; detection, 260 nm. Melting points are uncor-
rected. 1H NMR and 13C NMR were recorded at 200 and 50
MHz using a Varian Gemini spectrometer or at 300 and 75
MHz on a Bruker AM-300 spectrometer. Electrospray mass
spectra were determined at the University of Alberta.
5′-[2,5,5-Trimethyl-1,3-dioxolan-4-on-2-yl]-3′-bromo-
3′-deoxy-2′-O-acetyl Adenosine (2). Adenosine (0.3 kg, 1.12
mol) was suspended in equal amounts of acetonitrile and
ethyl acetate (1.30 kg each) in a 5 L reactor equipped with
a water-cooled condenser, nitrogen line, addition funnel,
thermocouple probe, and overhead agitator. The pot tem-
perature was adjusted to 25-30 °C, and Mattock’s bromide
(0.66 kg, 3.16 mol, 2.8 equiv) was added at a rate keeping
the temperature below 35 °C (Caution: Exotherm). When
the addition was complete, the reaction mixture was agitated
at 30 °C until complete by HPLC (typically 5-6 h). The
reaction mixture, which was still a slurry, was cooled to room
temperature and was quenched into a solution of potassium
bicarbonate (0.5 kg) and water (1.05 L) (Caution: CO2
evolved). The mixture was allowed to agitate for 1 h, and
then the lower aqueous layer was separated and extracted
with ethyl acetate (0.75 kg). The combined organic layers
were washed with a 10% brine solution (450 mL), and the
aqueous layer was discarded. The remaining organic layer
was concentrated and then coevaporated with ethyl acetate
(100 mL). The total dissolved solids (TDS) was adjusted to
30-35% by the addition or removal of ethyl acetate. Heptane
(0.9 kg) was added to the slurry and the mixture was cooled
to 0-5 °C for ∼1 h. The slurry was then filtered and the
solids pulled dry, then transferred to a vacuum oven and dried
overnight at 40-50 °C. The product was obtained as a white
solid in a yield of 52% (0.29 kg) and a purity of 98.5%
(HPLC). 1H NMR (CDCl3, 200 MHz) 1.45 (s, 3H), 1.50 (s,
3H), 1.74 (s, 3H), 2.11 (s, 3H), 3.86 (m, 2H), 4.50 (m, 1H),
4.94 (m, 1H), 5.92 (m, 1H), 6.14 (d, J ) 3 Hz), 7.40 (br,
2H), 8.12 (s, 1H), 8.22 (s, 1H) ppm; mp 163-165 °C (lit.
171 °C)9
Having improved the yield to the penultimate intermedi-
ate, the next stage was to carry the isolated solid forward
through the hydrogenolysis and purify the final product.
Upon further scale up of the hydrogenolysis, it was necessary
to increase the pressure and temperature to 50 psi and 50
°C, respectively, to get good conversions of 6 to 1. These
changes in the reaction conditions did not adversely affect
the adenine content.
With conditions that allowed for smooth hydrogenolysis,
the remaining challenge was to develop an isolation method
604
•
Vol. 4, No. 6, 2000 / Organic Process Research & Development

3′-Bromo-3′-deoxyadenosine Hydrochloride (6). Bro-
moacetate 2 (220 g, 0.44 mol) was suspended in methanol
(880 g) in a 2 L reactor. Concentrated HCl (216.8 g, 2.2
mol, 5 equiv) was charged slowly, keeping the temperature
below 20 °C (Caution: exotherm). The temperature was then
maintained at 20-25 °C, and the slurry was agitated until
the reaction was complete, as monitored by NMR or HPLC.
Typically this reaction required 2-4 days to go to comple-
tion. The slurry was then cooled to 0-5 °C for ∼1 h and
then filtered. After drying in a vacuum oven at 40 °C for 16
h, the product was isolated as a white solid (113.02 g, 70.1%)
in a purity of 97.8% (HPLC). 1H NMR (DMSO, 300 MHz)
3.72 (dd, J ) 12.5, 6.0 Hz, 1H), 3.81 (dd, J ) 12.5, 6.0 Hz,
1H), 4.40 (q, J ) 6.0 Hz 1H), 4.59 (dd, J ) 6.0, 3.5 Hz,
1H), 4.78 (t, J ) 3.5 Hz, 1H), 5.94 (d, J ) 3.5 Hz, 1H),
8.54 (s, 1H), 8.62 (s, 1H) ppm; 13C NMR (75 MHz, DMSO)
52.56, 61.64, 79.94, 80.35, 88.41, 117.64, 140.82, 144.41,
147.25, 149.59 ppm; IR (KBr) 3200 (br), 1695 cm-1; mp.
172 °C (dec); EIMS (m/z,%) 330.0 (M+, 100), 332 (M+ +
2, 99).
3′-Deoxyadenosine (1). A solution of bromide 6 (100 g,
0.27 mol) and sodium acetate (111.91 g, 1.37 mol, 5 equiv)
in ethanol (500 g) and distilled water (500 g) was prepared
and filtered through a glass fiber filter paper to remove a
small amount of insoluble material. Palladium on carbon
(5%, 50% water wet) was charged to the solution, and the
slurry was charged to a 5 L Parr glass-lined pressure vessel.
The reactor was flushed with nitrogen three times, then
charged to 30 psi hydrogen. The reaction mixture was
agitated at 250 rpm and 25 °C for 44 h, after which time,
in-process analysis had showed that the rate of the reaction
had slowed dramatically. The pressure was then increased
to 50 psi, and after another 4 h, the rate had not changed
significantly. The temperature was then increased to 50 °C,
and agitation continued until the reaction was complete
(another 21 h). The atmosphere inside the reactor was then
converted from hydrogen to nitrogen, and the contents were
discharged. The catalyst was removed by filtration over a
Celite bed, blanketed with nitrogen (Caution: Dry catalyst
is pyrophoric), and the clear yellow stock solution was
extracted with ethyl acetate (1L). The organic layer was
concentrated under reduced pressure to a volume of 180-
200 mL (45 °C bath temperature) and methanol (150 mL)
was charged. The slurry was cooled to 5 °C and agitation
continued ∼ 1 h, and then the slurry was filtered and dried
to afford the product as a white solid (36.0 g, 52.5%), which
was 99.0% pure by HPLC and had an ash content of 0.4 wt
%. Another crop of product could be isolated from the mother
liquor, which increased the yield to over 60%; however, the
purity was lower (96% by HPLC), and the ash content was
1
also higher (3.5%). H NMR (d6-DMSO, 200 MHz) 1.95
(m, 1H), 2.21 (m, 1H), 3.60 (m, 2H), 4.30 (m, 1H), 4.55
(m, 1H), 5.82 (s, 1H), 7.22 (s, 1H), 8.15 (s, 1H), 8.28 (s,
9
1H) ppm; mp 224-226 °C (lit. 224 °C).
3′-Bromo-3′-deoxy-2′,5′-di(O-acetyl) Adenosine (7). 1H
NMR (CDCl3, 200 MHz) 2.12 (s, 3H), 2.20 (s, 3H), 4.50
(m, 4H), 5.65 (br, 2H), 5.72 (s, 1H), 6.24 (d, J ) 3.0 Hz),
8.29 (s, 1H), 8.35 (s, 1H) ppm; 13C NMR (50 MHz, CDCl3)
20.66, 20.75, 49.29, 64.91, 78.50, 83.13, 87.66, 119.58,
138.52, 149.73, 153.41, 155.7, 169.14, 170.43 ppm; mp
163-165 °C (lit. 166-167 °C).23
3′-Bromo-3′-deoxy-2′-O-acetyl Adenosine Hydrochlo-
1
ride (8). H NMR (200 MHz, DMSO) 2.05 (s, 3H), 3.80
(m, 2H), 4.40 (q, J ) 6.0 Hz), 4.50 (br, 4H), 4.90 (t, J )
6.0 Hz), 5.90 (t, J ) 6.0 Hz), 6.21 (d, J ) 6.0 Hz), 8.5 (s,
1H), 8.62 (s, 1H) ppm; 13C NMR (50 MHz, DMSO) 20.45,
49.19, 62.24, 81.31, 81.43, 86.64, 118.51, 141.28, 145.98,
148.16, 150.87, 169.42 ppm; IR (KBr) 1210, 1690, 1730,
2700 (br), 3000 (br) cm-1; EIMS (m/z, %) 372.0 (M+ + H,
100%), 374.0 (M+ + 2 + H, 99%); mp 160 °C (dec).
Received for review May 5, 2000
OP000209X
(23) Kondo, K.; Adachi, T.; Inoue, I. J. Org. Chem. 1977, 42, 3967.
Vol. 4, No. 6, 2000 / Organic Process Research & Development
•
605