Stereospecific fluorination of 1,3,5-tri-O-benzoyl-α-D-ribofuranose-2-sulfonate esters: Preparation of a versatile intermediate for synthesis of 2′-[ 18F ]-fluoro-arabinonucleosides

Article abstract of DOI:10.1016/S0022-1139(00)00307-9

A detailed investigation on fluorination of 1,3,5-tri-O-benzoyl-α-D-ribofuranose-2-sulphonate esters is reported. Various combinations of sulfonate esters, fluorinating agents and solvents were evaluated in this study. Organic ammonium fluoride, in particular n-Bu₄NF, was found to be better fluorinating agent than inorganic fluoride, and 1,3,5-tri-O-benzoyl-α-D-ribofuranose-2-trifluoromethylsulphonate ester appeared to be the best substrate. The developed method is suitable for stereospecific (arabino) incorporation of radiofluorine (¹⁸F) into the sugar moiety.

Full text of DOI:10.1016/S0022-1139(00)00307-9

Journal of Fluorine Chemistry 106 (2000) 87±91
Stereospeci®c ¯uorination of 1,3,5-tri-O-benzoyl-a-D-ribofuranose-
2-sulfonate esters: preparation of a versatile intermediate for
synthesis of 20-[18F]-¯uoro-arabinonucleosides
Mian M. Alauddina, Peter S. Contia,*,1, Thomas Mathewb, John D. Fissekisa,
G.K. Surya Prakashb, Kyoichi A. Watanabec,2
aPET Imaging Science Center, University of Southern California, Los Angles, CA 90033, USA
bLoker Hydrocarbon Research Institute, University of Southern California, Los Angeles, CA 90033, USA
cLaboratory of Organic Chemistry, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
Accepted 24 May 2000
Abstract
A detailed investigation on ¯uorination of 1,3,5-tri-O-benzoyl-a-D-ribofuranose-2-sulphonate esters is reported. Various combinations of
sulfonate esters, ¯uorinating agents and solvents were evaluated in this study. Organic ammonium ¯uoride, in particular n-Bu4NF, was
found to be better ¯uorinating agent than inorganic ¯uoride, and 1,3,5-tri-O-benzoyl-a-D-ribofuranose-2-tri¯uoromethylsulphonate ester
appeared to be the best substrate. The developed method is suitable for stereospeci®c (arabino) incorporation of radio¯uorine (18F) into the
sugar moiety. # 2000 Elsevier Science S.A. All rights reserved.
Keywords: Fluorination; n-Bu4NF; Nucleophilic radio¯uorine
1. Introduction
moiety in a pyrimidine nucleoside has not been possible
[1,6±8]. Alternatively, the incorporation of ¯uorine in the
arabino con®guration at C-2 of the sugar followed by
coupling with the pyrimidine base has been successful
[1,9±12]. The common methods for the stereospeci®c (ara-
bino) incorporation of ¯uorine involve use of KHF2 or
Et3NÁ3HF as ¯uorinating agent in excess (6 eq) [11,13],
and require long reaction times (6 h). These methods are not,
however, satisfactory for incorporation of radio¯uorine.
Pyrimidine-1-(20-deoxy-20-¯uoro-b-D-arabinofuranosyl)-
nucleosides and their radiolabeled derivatives are potentially
important compounds [1±5]. 20-Fluoro-5-[11C-methyl]-1-b-
D-arabinofuranosyluracil (11C-FMAU) 1a (Fig. 1) is being
studied as a marker for cell proliferation by positron emis-
sion tomography (PET) [3,4]. In addition, radiolabeled 20-
¯uoro-5-iodo-1-b-D-arabino-furanosyluracil (FIAU) 1b is
considered to be a useful agent for imaging the expression
of transfected genes in vivo [5]. 11C-FMAU 1a was ®rst
synthesized in our laboratory [3], in quantities suitable for
animal and patient studies. However, the short half-life of
C-11 (t1/2ˆ20 min) limits the clinical application of that
compound. Therefore, labeling with the longer-lived PET
isotope ¯uorine-18 (t1/2ˆ110 min) in the glycone would be
more advantageous.
The direct, stereospeci®c (arabino) introduction of ¯uor-
ine by an SN2 substitution at the 20-position of the furanosyl
* Corresponding author. Tel.: ‡1-323-442-5940; fax: ‡1-323-442-5778.
E-mail address: pconti@hsc.usu.edu (P.S. Conti).
1 Present address: Suite 101, 2250 Alcazar st. Los Angeles, CA 90033,
USA.
2 Present address: Pharmasset, Inc., 1860 Montreal Road, Tucker, GA
30084, USA.
Fig. 1. Structure of FMAU (1a) and FIAU (1b).
0022-1139/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 0 0 ) 0 0 3 0 7 - 9

88
M.M. Alauddin et al. / Journal of Fluorine Chemistry 106 (2000) 87±91
¯uoride (n-Bu4NF) were prepared from their hydroxy-
compounds by neutralization with aqueous hydro¯uoric
acid (48% HF) to pH 7.00. The aqueous solution was
dried by azeotropic evaporation of water with aceto-
nitrile before the ¯uorination reaction. Triethylamine tri-
(hydrogen ¯uoride) (Et3NÁ3HF) was prepared by adding
cold, anhydrous HF into Et3N at low temperature (dry ice
acetone).
2.3. 2-Deoxy-2-fluoro-1,3,5-tri-O-benzoyl-a-D-
arabinofuranose 3
Fig. 2. General methodology of fluorination reactions.
Earlier, we developed a method suitable for stereospeci®c
(arabino) incorporation of radioactive ¯uorine in the 2-
position of the protected sugar [14,15]. The chemical yield
was variable, but could be as high as 30%. However, the
radiochemical yield was quite low (2±6%). In our continuing
effort towards the synthesis of 18F-FMAU, we have inves-
tigated reactions of 1,3,5-tri-O-benzoyl-a-D-ribofuranose-2-
sulphonate esters 2a±c with several ¯uorinating agents,
under a variety of experimental conditions (Fig. 2). Here,
we report a procedure for stereospeci®c (arabino) incorpora-
tion of radio¯uorine into the protected sugar, much superior
to the currently available method in terms of simplicity and
radiochemical yield.
All ¯uorination reactions were performed in a similar
manner; a representative procedure is described here. Each
precursor (¯uorosulfonate, methylsulfonate or tri¯uoro-
methylsulfonate, 10±12 mg) was dissolved in either EtOAc
or MeCN (0.5 ml), and the solution was added to the dry
¯uorination reagent under argon in a V-vial. The vial was
heated in a heating block at 71±728C for 20 min, when TLC
(12% acetone in hexane) showed no starting material
remained. The reaction mixture was cooled to room tem-
perature, solvent evaporated, and puri®ed by HPLC (C18
semi-prep. column, 70% MeCN/H2O, ¯ow 4.5 ml/min). The
appropriate fraction was collected and evaporated to dryness
to provide the pure product. 1H NMR: dˆ8.05±8.15 (m, 6H,
aromatic), 7.55±7.64 (m, 3H, aromatic), 7.39±7.48 (m, 6H,
aromatic), 6.74 (d, 1H, C1H, Jˆ8.6 Hz), 5.61 (dd, 1H, C3H,
J ˆ19.4 and 2.9 Hz), 5.37 (d, 1H, C2H, Jˆ48.6 Hz), 4.68±
4.79 (m, 3H, C4H and C5H). 19F NMR: dˆ 190.8. MS: 465
(M‡1, 10), 391 (35), 343 (75), 242 (100).
2. Experimental
2.1. Reagents and instrumentation
All reagents and solvents were purchased from Aldrich
Chemical Co. (Milwaukee, WI), and used without further
puri®cation. 2-O-(Fluorosulfonyl)-1,3,5-tri-O-benzoyl-a-D-
ribofuranose 2a, 2-O-(methylsulfonyl)-1,3,5-tri-O-benzoyl-
a-D-ribofuranose 2b and 2-O-(tri¯uoromethylsulfonyl)-
1,3,5-tri-O-benzoyl-a-D-ribofuranose 2c were prepared fol-
lowing literature methods [11,13].
Proton and 19F NMR spectra were recorded on a Brucker
360 or 500 MHz spectrometer in chloroform-D3 using tetra-
methylsilane as an internal reference, and trichloro¯uoro-
methane as an external reference, respectively. Mass spectra
were obtained on a Finnigan 400 mass spectrometer at the
University of Minnesota using ammonia chemical ionization
technique. High performance liquid chromatography
(HPLC) was performed on a Waters Associates system using
a 510 pump, UV detector (Isco) operated at 254 nm, and a
radioactivity detector with single-channel analyzer (Tech-
nical Associates, Canoga Park, CA) using semi-preparative
or analytical C18 reverse phase columns (Alltech). A MeCN/
H2O solvent system (70% MeCN) was used.
2.4. 1,3,5-Tri-O-benzoyl-a-D-arabinofuranose 4
This compound was essentially the major product in most
of the ¯uorination reactions. It was isolated during HPLC
puri®cation of the crude material in 44±60% yield, and
1
characterized by H NMR spectroscopy and mass spectro-
metry. 1H NMR: dˆ8.04±8.14 (m, 6H, aromatic), 7.60±7.66
(m, 3H, aromatic), 7.41±7.49 (m, 6H, aromatic), 6.75 (s, 1H,
C1H), 5.82 (s, 1H, C2H), 5.68 (d, 1H, C3H, J ˆ3.5 Hz), 4.81±
4.84 (m, 1H, C4H), 4.71±4.75 (m, 2H, C5H). MS: 461 (M 1,
5), 445 (100), 242 (85), 201 (70).
2.5. 2-Acetyl-1,3,5-tri-O-benzoyl-a-D-arabinofuranose 5
The precursor (2a, 7.7 mg) was dissolved in EtOAc
(0.5 ml) in a V-vial under argon. Acetic acid (1.2 eq) and
n-Bu4NF (1.2 eq) were added and the reaction mixture was
heated in a heating block at 71±728C for 95 min. The
reaction mixture was cooled to room temperature, solvent
evaporated, and puri®ed by HPLC. The desired product was
isolated in 75% yield. 1H NMR: dˆ8.07±8.12 (m, 6H,
aromatic), 7.55±7.64 (m, 3H, aromatic), 7.41±7.47 (m,
6H, aromatic), 6.58 (s, 1H, C1H), 5.58 (s, 1H, C2H), 5.52
(d, 1H, C3H, Jˆ3.4 Hz), 4.71±4.77 (m, 3H, C4H and C5H),
2.09 (s, 3H, acetate). MS: 522, M‡NH4 (42), 383 (100), 343
2.2. Fluorinating reagents
Tetraethylammonium ¯uoride (Et4NF), tetrapropylam-
monium ¯uoride (n-Pr4NF) and tetrabutylammonium

M.M. Alauddin et al. / Journal of Fluorine Chemistry 106 (2000) 87±91
89
(28), 104 (65). Exact mass calculated for C28H28NO9
(M‡NH4) 522.1766, found 522.1766.
2-deoxy-2-¯uoro-1,3,5-tri-O-benzoyl-a-D-arabinofuranose
3 by ¯uorination of the ¯uorosulfonate 2a. Most of these
experiments were performed in EtOAc, which has been
reported to be the best solvent for this reaction [13]. In
addition, MeCN and DMF were also used in some experi-
ments for comparison. Fig. 2 represents the general meth-
odology of such ¯uorination reactions.
2.6. 2-Deoxy-2-[18F]-fluoro-1,3,5-tri-O-benzoyl-a-D-
arabinofuranose 3
[18F]-Fluoride was reacted with n-Bu4NHCO3 and eva-
porated to dryness azeotropicaly with MeCN (1.2 ml). To
the dry residue, n-Bu4N18F, a solution of 2a or 2c (5±7 mg)
in ethyl acetate (0.5 ml) was added, and the reaction mixture
was heated at 71±728C for 20 min. The crude reaction
mixture was passed through a sep-pack cartridge (silica
gel), and eluted with EtOAc. After evaporation of the solvent,
the residue was dissolved in the HPLC solvent (70% MeCN/
H2O), and chromatographed using a C18 semi-prep column.
The appropriate fraction containing the radioactive species
was isolated and evaporated to dryness. The radiochemical
yields were 32±41% for 2c and 3±6% for 2a, respectively,
both in >99% radiochemical purity.
Table 1 represents the results of ¯uorination with various
substrates, ¯uorinating agents, solvents and time at 71±
728C. We have investigated each of three substrates against
various ¯uorinating agents. When reacted with n-Bu4NF, the
tri¯ate 2c and ¯uorosulfonate 2a produced 3 in 30±40 and
12±15% yield, respectively. In contrast, the mesylate 2b did
not produce any product. n-Bu4NF in either EtOAc or MeCN
led to successful ¯uorination of 2a and 2c, although MeCN
was optimal for isolation of the pure product. Fluorinations
with either tetra-alkyl ammonium ¯uorides resulted in rela-
tively higher yield (12±16%) in EtOAc. In MeCN, the yields
were slightly lower (10%) in the reactions of Et4NF and
n-Pr4NF, but this difference was not signi®cant in the case
of n-Bu4NF. The small difference of yields (Table 1) in the
reactions of tetra-alkyl ammonium ¯uorides with these
solvents, may be due to the relative solubility of the reagents,
and thus the availability of the ¯uoride ion in solution. It has
been suggested, that the solvent plays an important role in
reactions of the ¯uoride ion [16]. Our results also suggest
that choice of solvent is important, with EtOAc appearing to
be a better solvent for organic ammonium ¯uorides in these
¯uorination reactions in terms of yields. EtOAc was also
3. Results and discussion
The precursors, ¯uorosulfonate 2a, mesylate 2b and
tri¯ate 2c were prepared following literature methods in
high yields [11,13]. Like other investigators, we were not
successful in preparing the tosylate 2d by the reaction
of tosyl chloride on 1,3,5-tri-O-benzoyl ribofuranose [11].
We carried out extensive studies on the preparation of
Table 1
Fluorination reactions of substrates with various fluorinating agents and solvents at 71±728C
Substrate
Fluorinating agent
Solvent
Time
Yield (%)
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2a
2b
2c
2c
2c
2c
2c
2c
(Et)4NF (1.2 eq)
EtOAc
EtOAc/THF
MeCN
EtOAc
MeCN
EtOAc
MeCN
EtOAc
EtOAc
MeCN
MeCN
DMF
20 min
20 min
20 min
20 min
20 min
20 min
20 min
45 min
96 min
45 min
75 min
75 min
45 min
2.5 h
13
(Et)4NF (1.2 eq)
9.0
(Et)4NF (1.2 eq)
10.5
12.5
10.5
15.7
14.5
5.0
(nPr)4NF (1.2 eq)
(nPr)4NF (1.2 eq)
(nBu)4NF (1.2 eq)
(nBu)4NF (1.2 eq)
CsF (10.0 eq)
CsF (10.0 eq)
10.5
12.3
17.5
5.3
CsF (3.4 eq)
CsF (3.4 eq)
CsF (3.3 eq)
CsF (8.7 eq)
DMF
14.0
56.0
60.5
35.5
30
(Et)3NÁ3HF/(Et)3N (6 eq)
(Et)3NÁ3HF/(Et)3N (6 eq)
(Et)3NÁ3HF/(Et)3N (6 eq)
(nBu)4NF‡Et3NÁ3HF
(nBu)4N(F-18)F‡Et3NÁ3HF
(nBu)4NF (1.2 eq)
(nBu)4NF (1.2 eq)
(nBu)4NF (1.2 eq)
(nBu)4NF (1.2 eq)
(nBu)4N18F (trace)
(nBu)4N18F (trace)
(nBu)4N18F (trace)
EtOAc
EtOAc
MeCN
EtOAc
EtOAc
EtOAc
EtOAc
EtOAc
EtOAc
EtOAc
EtOAc
EtOAc
4 h
4 h
30 min
30 min
20 min
20 min
20 min
20 min
20 min
20 min
20 min
6a
no product
31
34
30
32a
41a
35a
a Radiochemical yield.

90
M.M. Alauddin et al. / Journal of Fluorine Chemistry 106 (2000) 87±91
found to be much better solvent in the reactions with the
other organic ¯uoride Et3NÁ3HF/Et3N, but the reaction time
was much longer. Cesium ¯uoride (CsF), in the presence of a
crown ether, produced higher yields in MeCN than in
EtOAc. A highly polar aprotic solvent, DMF, however,
did not produce a high yield of product with CsF unless
the amount of CsF was increased to more than two-fold
compared to that used in MeCN.
As an electron donor it forms strong hydrogen bonded
complex with acetic acid producing acetate nucleophile,
which competes with ¯uoride as a nucleophile and produces
the acetoxy compound. In order to verify the formation of
the compound 5, separate experiments were performed with
n-Bu4NF (1.2 eq) in the presence of same equivalent acetic
acid. The major product in these experiments was 5 in 70±
75% yields.
All reactions using the ¯uorosulfonate ester 2a produced
low yield of the desired product 3 within 20±30 min except
in one instance when two reagents were used. Combination
of n-Bu4NF and Et3NÁ3HF (3:2) appeared to produce the
higher yield within shorter reaction time (30 min). The
chemical yield could be improved from 15 to 30% by
manipulation of the reagent composition, with the radio-
chemical yield improved proportionally from 3 to 6%. This
persistent low radiochemical yield was due to an undesirable
competitive exchange, between radioactive ¯uoride and the
covalently bound ¯uorine on the leaving group. The nucleo-
philic radio¯uorine (¯uoride) is consumed by displacing the
isotopic ¯uorine on the sulfonate ester leaving group (Fig. 3)
and thus becomes unavailable for the desired substitution
reaction. The net result of this facile exchange is a signi®cant
reduction in radiochemical yield. Although, 2a is known to
be stable [13], the sulfur ¯uorine bond have been reported to
be very labile [17]. Therefore, it is likely that the radioactive
¯uoride in solution is equilibrated fast with the non-radio-
active ¯uoride before the apparently slower displacement of
the sulfonate ester is completed.
Of the three substrates studied 2a, 2b, and 2c, only 2a and
2c produced the desired product. The failure of the mesylate
2b to yield 3 may be due to an electronic effect; i.e. the
mesylate may not be a good leaving group in this system.
When ¯uorination of the tri¯ate 2c was performed using
radioactive ¯uoride the radiochemical yield was 32±41%. In
this reaction, the radioactive ¯uoride was consumed in the
substitution reaction producing high yield of the desired
compound. An HPLC chromatogram of radiolabeled ¯uor-
osugar 3 is represented in Fig. 4.
Earlier studies demonstrated that the reaction of sulfonate
esters 2b and 2c with n-Bu4NF resulted, either in an
elimination product or in the recovery of the starting mate-
rial [11]. These authors explained their results on the basis of
an earlier suggestion indicating that nucleophilic displace-
ment at C2 creates an unfavorable alignment of dipoles in the
transition state, particularly when the substituent at C1 is the
a-D-anomer [19]. Our results, however, showed the absence
of any elimination product or starting material, and instead
two products, 2-¯uoro-and 2-hydroxy-arabinose derivatives
produced by nucleophilic substitution (Fig. 2). These results,
therefore, suggest that the ¯uorosulfonate and tri¯uor-
methylsulfonate esters of 1,3,5-tri-O-benzoyl-2-ribose do
undergo nucleophilic substitution with ¯uoride, acetate
and hydroxide ions, while mesylate is inactive in this
system. Furthermore, nucleophilic substitution of tri¯uor-
omethanesulfonate ester at the C2 position of a xylo sugar
has also been reported [20]. Since chemical yield in ¯uor-
ination reaction of sulfonate esters with n-But4NF in EtOAc
and in MeCN are comparable, use of MeCN might be
advantageous to avoid the formation of the acetate 5 which
Fig. 3. Exchange reaction between radiofluorine and isotopic stable
fluorine.
All ¯uorinating agents, except Et3NÁ3HF/Et3N, produced
a common by-product (45±60%), which was characterized
as 1,3,5-tri-O-benzoyl-1-a-D-arabinofuranose 4. This pro-
duct was produced by the competitive nucleophilic substitu-
tion of the sulfonate esters with a hydroxide ion from either
an excess of tetra-alkyl ammonium hydroxide or traces of
water present in the reaction mixture. In the reactions with
Et3NÁ3HF/Et3N, this by-product was not formed, since the
overall reaction was under acidic conditions.
A second by-product was formed in small amounts when
the reactions were performed in EtOAc. It was identi®ed as
2-acetyl-1,3,5,-tri-O-benzoyl-1-a-D-arabinofuranose 5. The
formation of 5 may be explained by the nucleophilic sub-
stitution of the sulfonate ester by trace of acetate ion present
in EtOAc. Fluoride ion is known to be a strong base, a potent
hydrogen bond electron donor, and a weak nucleophile [18].
Fig. 4. HPLC chromatogram of [F-18]-labeled-2-fluoro-1,3,5-tri-O-ben-
zoyl-a-D-arabinofuranose.

M.M. Alauddin et al. / Journal of Fluorine Chemistry 106 (2000) 87±91
91
[2] J.J. Fox, K.A. Watanabe, T.C. Chou, R.F. Schinazi, K.E. Soike, I.
Fourel, O. Hantz, C. Trepo. in: N. F. Taylor (Ed.), Fluorinated
Carbodydrates, Am. Chem. Soc., 1988, p. 176.
is, rather, dif®cult to separate from the desired product
during HPLC puri®cation.
[3] P.S. Conti, M.M. Alauddin, J.D. Fissekis, B. Schmall, K.A.
Watanabe, Nucl. Med. Biol. 22 (1995) 783.
4. Conclusion
[4] P.S. Conti, J.H. Anderson, D.F. Wong, J.R. Bading, M.M. Alauddin,
H.T. Ravert, W.B. Mathews, J.L. Musachio, R.F. Dannals, U.
Scheffel, J. Nucl. Med. 37 (1996) 107P.
We have investigated, ¯uorination of 1,3,5-tri-O-benzoyl-
a-D-ribofuranose-2-sulfonyl esters with several ¯uorides in
various solvents, and developed a suitable method for
incorporation of radio¯uorine into the protected sugar.
The 2-O-tri¯uoromethyl-sulfonyl-1,3,5-tri-O-benzoyl-a-D-
ribofuranose was found to be the best substrate. Tetrabuty-
lammonium ¯uoride in either EtOAc or MeCN led to
successful ¯uorination, although MeCN is optimal. For
radiolabeling with 18F, the tri¯ate is the optimal substrate,
which consistently produces the higher radiochemical yield
in the series studied. The 2-¯uoro-1,3,5-tri-O-benzoyl-a-D-
arabinofuranose 3, successfully labeled with 18F, will be a
potential intermediate for synthesis of 18F labeled nucleo-
side analogues for PET imaging.
[5] J.G. Tjuvajev, G. Stockhammar, R. Desai, H. Uehara, K.A.
Watanabe, B. Gansbacher, R.G. Blasberg, Cancer Res. 55 (1995)
6126.
[6] K.W. Pankiewicz, B. Nawrot, H. Gadler, R.W. Price, K.A. Watanabe,
J. Med. Chem. 30 (1987) 2314.
[7] K.W. Pankiewicz, K.A. Watanabe, Chem. Pharm. Bull. 35 (1987)
4494.
[8] L.C. Cheng, T.L. Su, K.W. Pankiewicz, K.A. Watanabe, Nucleosides
Nucleotides 8 (1989) 1179.
[9] K.A. Watanabe, T.L. Su, R.S. Klein, K.C. Chu, A. Mastuda, M.W.
Chun, C. Lopez, J.J. Fox, J. Med. Chem. 26 (1983) 152.
[10] K.A. Watanabe, T.L. Su, R.S. Klein, U. Riechman, N. Greenberg, C.
Lopez, J.J. Fox, J. Med. Chem. 27 (1984) 91.
[11] C.H. Tann, P.R. Brodfuehrer, S.P. Brundidge, C. Sapino, H.G.
Howell, J. Org. Chem. 50 (1985) 3644.
[12] H.G. Howell, P.R. Brodfuehrer, S.P. Brundidge, D.A. Beningni, C.
Sapino Jr., J. Org. Chem. 53 (1988) 85.
[13] T.S. Chou, L.M. Becke, J.C. O'Toole, M.A. Carr, B.E. Parker,
Tetrahedron Lett. 37 (1996) 17.
Acknowledgements
[14] M.M. Alauddin, P.S. Conti, T. Mathew, H. Abrahamian, J.D. Fissekis,
G.K.S. Prakash, K.A. Watanabe, J. Nucl. Med. 40 (1999) 311P.
[15] M.M. Alauddin, P.S. Conti, T. Mathew, H. Abrahamian, G.K.S.
Prakash, K.A. Watanabe, Anticancer Res. 18 (1998) 4991.
[16] M.M. Alauddin, J.M. Miller, J.H. Clark, Can. J. Chem. 62 (1984)
263.
The authors wish to thank Mr. Allan D. Kershaw for
assisting with the NMR spectra. This work was supported by
the National Cancer Institute Grant CA 72896.
[17] P.J. Durrant, B. Durrent, Introduction to Advanced Inorganic
Chemistry, 2nd Edition, Wiley, New York, 1970, p. 857.
[18] J.H. Clark, Chem. Rev. 80 (1980) 429.
References
[19] A.C. Richardson, Carbohydr. Res. 10 (1969) 395.
[20] T. Haradahira, A. Kato, M. Maeda, Y. Kanazawa, M. Yamada, Y.
Tori, Y. Ichiya, K. Masuda, J. Label. Compd. Radiopharm. (1995)
552.
[1] U. Reichman, K.A. Watanabe, J.J. Fox, Carbohydr. Res. 42 (1975)
233.