S-pixyl analogues as photocleavable protecting groups for nucleosides

Article abstract of DOI:10.1021/jo026009w

Several analogues of the 9-phenylthioxanthyl (S-pixyl) photocleavable protecting group have been synthesized, containing substituents on the 9-aryl ring and on the thioxanthyl backbone. Each analogue protected the 5′-hydroxy moiety of thymidine in good to excellent yield. The protected substrates were deprotected in 1:1 water: acetonitrile with irradiation at 300 nm, resulting in recovered thymidine in excellent yield, except for the nitro-substituted analogues which gave substantially lower yields. Substrates with 2,7-dibromo or 3-methoxy substitution on the thioxanthyl backbone were also deprotected efficiently with irradiation at 350 nm. Shorter irradiation times were observed in the less nucleophilic solvent mixture of 1:9 trifluoroethanol: acetonitrile, with no formation of secondary photooxidation products. Photodeprotection with high yields was also achieved in the absence of solvent, with no secondary photoproducts.

Full text of DOI:10.1021/jo026009w

S-P ixyl An a logu es a s P h otoclea va ble P r otectin g Gr ou p s for
Nu cleosid es
Michael P. Coleman and Mary K. Boyd*
Department of Chemistry, Loyola University Chicago, Chicago, Illinois 60626
mboyd@luc.edu
Received J une 3, 2002
Several analogues of the 9-phenylthioxanthyl (S-pixyl) photocleavable protecting group have been
synthesized, containing substituents on the 9-aryl ring and on the thioxanthyl backbone. Each
analogue protected the 5′-hydroxy moiety of thymidine in good to excellent yield. The protected
substrates were deprotected in 1:1 water:acetonitrile with irradiation at 300 nm, resulting in
recovered thymidine in excellent yield, except for the nitro-substituted analogues which gave
substantially lower yields. Substrates with 2,7-dibromo or 3-methoxy substitution on the thioxanthyl
backbone were also deprotected efficiently with irradiation at 350 nm. Shorter irradiation times
were observed in the less nucleophilic solvent mixture of 1:9 trifluoroethanol:acetonitrile, with no
formation of secondary photooxidation products. Photodeprotection with high yields was also
achieved in the absence of solvent, with no secondary photoproducts.
In tr od u ction
protecting groups are effective, alternative chemistries
are worthy of investigation in order to address synthetic
limitations of the technologies which include a subquan-
titative stepwise yield that results in truncated probes
and secondary photoproducts that can be damaging to
the growing DNA strand.1 Nitrobenzyl compounds modi-
fied by the inclusion of a pentadienyl group that traps
the nitroso moiety were developed to address this dif-
ficulty.4 Several photocleavable alcohol protecting groups
that do not employ the o-nitrobenzyl moiety have also
been reported.5
Our previous work demonstrated that 9-phenylxan-
thenol underwent photochemically induced proton trans-
fer from the neutral aqueous solvent to an incipient
hydroxide ion leaving group to generate the xanthyl
cation upon steady-state irradiation.6 We recently re-
ported that primary alcohols protected by the 9-phenyl-
xanthyl moiety as derivatized ethers underwent the same
photochemically initiated carbon-oxygen bond cleavage
to regenerate the substrate alcohol and 9-phenylxan-
thenol in good to excellent yield (Scheme 1).7 The pixyl
group was proven effective in the photodeprotection of a
variety of primary alcohols, including thymidine, extend-
ing its utility to nucleosides. Our investigations then led
us to consider the thioxanthyl analogue which contains
sulfur as the heteroatom within the central xanthyl ring.
9-Phenylthioxanthenol similarly undergoes a photochem-
The increasing demand for biologically significant,
synthetically generated materials immobilized on solid
surfaces is well-known. Advances in photolithography in
conjunction with a combinatorial approach to organic
synthesis on a surface have produced biopolymers in
point-of-use devices with great precision and diversity.1
Of particular interest is the technique of light-directed
synthesis for the synthetic generation of high-density
arrays of polynucleotide probes.2 The synthetic approach
uses nucleoside phosphoramidite monomers with nitro-
benzyl oxycarbonyl or benzoin carbonate photocleavable
protecting groups for the 5′-hydroxyl moiety to generate
short oligomer probes on glass surfaces.3 Although these
(1) (a) Pennisi, E. Science 1998, 282, 814. (b) Service, R. F. Science
1998, 282, 396. (c) Anderson, R. C.; McGall, G.; Lipshultz, R. J . Topics
Curr. Chem. 1998, 194, 117. (d) Burns, M. A.; J ohnson, B. N.;
Brahmasandra, S. N.; Handique, K.; Webster, J . R.; Krishnan M.;
Sammarco, T. S.; Man, P. M.; J ones, D.; Heldsinger, D.; Mastrangelo,
C. H.; Burke, D. T. Science 1998, 282, 484.
(2) (a) Fodor, S. P. A.; Read, J . L.; Pirrung, M. C.; Stryer, L.; Tsi
Lu, A.; Solas D. Science 1991, 251, 767-773. (b) Pease, A. C.; Solas,
D.; Sullivan, E. J .; Cronin, M. T.; Holmes, C. P.; Fodor, S. P. A. Proc.
Natl. Acad. Sci. 1994, 91, 5022. (c) McGall, G.; Labadie, J .; Brock, P.;
Wallraff, G.; Nguyen, T.; Hinsberg, W. Proc. Natl. Acad. Sci. 1996,
93, 13555. (d) McGall, H. G.; Barone, A. D.; Diggleman, M.; Fodor, S.
P. A.; Gentalen, E.; Ngo, N. J . Am. Chem. Soc. 1997, 119, 9, 22, 5082.
(e) McGall, G. H.; Fidanze, J . A. In Advances in Nucleic Acid and
Protein Analyses, Manipulation, and Sequencing; Limbach, P. S.,
Owicki, J . C., Raghavachari, R., Tan, W., Eds.; Proc. SPIE, 2000; Vol.
3926, p 106.
(3) (a) Pirrung, M. C.; Bradley, J . C. J . Org. Chem. 1995, 60, 6270.
(b) McGall, G. H.; Barone, A. D.; Diggelmann, M.; Fodor, S. P. A.;
Gentalen, E.; Ngo, N. J . Am. Chem. Soc. 1997, 119, 5081. (c) Pirrung,
M. C.; Fallon, L. J . Org. Chem. 1998, 63, 241. (d) Pirrung, M. C.; Wang,
L.; Montague-Smith, M. P. Org. Lett. 2001, 3, 1105. (e) Giegrich, H.;
Eisele-Buhler, S.; Hermann, C.; Kvasyuk, E.; Charubala, R.; Pfleiderer,
W. Nucleosides Nucleotides 1998, 17, 1987. (f) Hasan, A.; Stengele,
K.-P.; Giegrich, H.; Cornwell, P.; Isham, K. R.; Sachleben, R. A.;
Pfleiderer, W.; Foote, R. S. Tetrahedron 1997, 53, 4247
(4) Pirrung, M. C.; Lee, Y. R.; Park, K.; Springer, J . B. J . Org. Chem.
1999, 64, 5042.
(5) (a) Bochet, C. G. J . Chem. Soc., Perkin Trans. 1 2002, 125. (b)
Furuta, T.; Hirayama, Y.; Iwamura, M. Org. Lett. 2001, 3, 1809. (c)
J ones, P. B.; Pollastri, M. P.; Porter, N. A. J . Org. Chem. 1996, 61,
9455. (d) Pirrung, M. C.; Lee, Y. R. J . Org. Chem. 1993, 58, 6961. (e)
Porter, N. A.; Bruhnke, J . D. J . Am. Chem. Soc. 1989, 111, 7616.
(6) Wan, P.; Yates, K.; Boyd, M. K. J . Org. Chem. 1985, 50, 2881-
2898.
(7) Misetic, A.; Boyd, M. K. Tetrahedron Lett. 1998, 39, 1653-1656.
10.1021/jo026009w CCC: $22.00 © 2002 American Chemical Society
Published on Web 10/08/2002
J . Org. Chem. 2002, 67, 7641-7648
7641

Coleman and Boyd
SCHEME 1
creased or whether the secondary photooxidation reaction
can be prevented entirely. We also hoped that the S-pixyl
analogues would address some additional consider-
ations: to be effective as a photolabile group at wave-
lengths beyond 300 nm in order to avoid conditions that
could be potentially damaging to oligonucleotides,11 be
less susceptible to secondary photochemistry after depro-
tection, to produce a photolabile protecting group that is
nonsolvent dependent, and further, in the interests of
photolithography, to produce a group that is effective in
the solid-state.1c
We herein report the effect of substitution on the
S-pixyl moiety on the efficiencies and time for deprotec-
tion of several 5′-hydroxyl protected thymidine deriva-
tives. We also report on solvent effects for the most
effective substituted S-pixyl analogues.
ically induced carbon-oxygen bond cleavage to generate
the thioxanthyl cation.8 The incorporation of sulfur offers
the advantage of a significant bathochromic shift in the
absorption spectrum, thus avoiding undesirable nucleo-
side photochemistry that occurs at lower wavelengths for
irradiation.
Balgobin reported that the 9-phenylthioxanthyl (S-
pixyl) group is an effective acid-labile protecting group
for the 5′-hydroxyl moiety of nucleosides.9 9-Phenylthio-
xanthyl (S-pixyl) protected nucleosides underwent acid-
catalyzed hydrolysis in 80% aqueous acetic acid in 90 s.
In addition, the S-pixyl protecting group proved compat-
ible with traditional phosphotriester oligonucleotide syn-
thesis.
Resu lts a n d Discu ssion
Syn t h esis of Su b st it u t ed Th ioxa n t h on e a n d 9-
Ar ylth ioxa n th en ol Der iva tives. The effect of substitu-
tion on both the 9-aryl ring and on the thioxanthyl
backbone was investigated. Two backbone-substituted
thioxanthone derivatives were prepared. Bromination of
thioxanthone in bromine/acetic acid at 100 °C yielded 2,7-
dibromothioxanthone 1, after quenching with bisulfite
and recrystallization from toluene (eq 1).12 Friedel-Crafts
acylation of thiosalicylic acid and anisole with sulfuric
acid and sulfonyl chloride generated 3-methoxythioxan-
thone 2, after recrystallization from toluene (eq 2).
In a preliminary report, we demonstrated that primary
alcohols protected by the S-pixyl moiety as derivatized
ethers underwent the same photoinduced deprotection
reaction as the pixyl-protected alcohols, but at a longer
wavelength for irradiation and with significantly de-
creased irradiation times.10 Further, the S-pixyl group
proved effective toward the protection and photodepro-
tection at 300 nm of all four nucleoside building blocks.
Additionally, use of S-pixyl proved compatible with the
benzoyl and isobutyryl protecting groups for the exocyclic
amino functions of adenosine and guanosine, respectively.
We also showed that the preparative deprotection of
S-pixyl protected nucleosides could be carried out in a
Hanovia photochemical reactor using Pyrex filtered
glassware in protic solvents.
However, upon extended irradiation in thoroughly
degassed aqueous media beyond the time of optimum
deprotection, a secondary photooxidative reaction re-
sulted in the formation of thioxanthone, which appeared
to damage the fidelity of substrate nucleosides upon
prolonged exposure. Control experiments confirmed that
irradiations of thioxanthone in the presence of equimolar
amounts of nucleoside substrates prevented a quantita-
tive recovery of the nucleoside after lengthy irradiation
times. To pursue use of the thioxanthyl group as a viable
protecting group for the 5′-hydroxyl moieties of nucleo-
sides, analogues of S-pixyl are needed to determine
whether the time for optimal deprotection can be de-
Several aryl halide derivatives were reacted with
crushed magnesium turnings in THF to generate Grig-
nard reagents. These arylmagnesium halides were then
reacted with thioxanthone or substituted thioxanthones
1 and 2, followed by weak acid hydrolysis with aqueous
ammonium chloride, to generate the 9-arylthioxanthyl
derivatives 3a -l in good to excellent yield (Scheme 2).
Compounds 3c and 3e required the entrainment
method for formation of the Grignard reagent by slowly
titrating ethylene dibromide into the reaction mixture
containing the aryl halide and magnesium in THF,
followed by the addition of thioxanthone.13 Compounds
3g and 3d were then nitrated in acetic anhydride with
10 mol equiv of nitrated silica gel,14 followed by an
aqueous basic workup to generate 9-(3-methyl-4-nitro-
phenyl)thioxanthenol 4a and 9-(4-methyl-3-nitrophenyl)-
thioxanthenol 4b, after column chromatography (eq 3).
(11) Cadet, J .; Vigny, P. Bioorganic Photochemistry; Morrison, H.,
Ed.; Wiley: New York, 1991; Vol. 1, pp 1-272.
(8) Shukla, D.; Wan, P. J . Photochem. Photobiol. A: Chem. 1994,
79, 55.
(12) Gaffney, P. R. J .; Changsheng, L.; Rao, M. V.; Reese, C. B.;
Ward, J . C. J . Chem. Soc., Perkin Trans. 1 1991, 1355.
(13) Pearson, D. E.; Cowan D.; Beckler, J . D. J . Org. Chem. 1959,
24, 504-509.
(9) Balgobin, N.; Chattopadhyaya, J . B. Chem. Scr. 1992, 19, 143-
144.
(10) Coleman, M. P.; Boyd, M. K. Tetrahedron Lett. 1999, 40, 7911-
7915.
(14) Pasto, D.; J ohnson C.; Miller, M. Experiment & Technique in
Organic Chemistry; Prentice Hall: Upper Saddle River, NJ , 1992.
7642 J . Org. Chem., Vol. 67, No. 22, 2002

Photocleavable Protecting Groups for Nucleosides
SCHEME 2
SCHEME 3
lamps of 300 nm in a Rayonet photolysis reactor. For each
derivative, several irradiations were conducted at ir-
radiation times ranging from 30 s to 60 min in order to
optimize yield. The temperature was held between 10 and
15 °C throughout each experiment. All reactions were
monitored with HPLC. Parallel “dark” reactions were run
against each block of reactions to ensure no thermal
reaction.
Alcohols 3a -l and 4a ,b were purified by recrystallization
or silica gel chromatography to obtain white solids and
were characterized by 1H and 13C NMR spectroscopy and
HRMS exact mass measurement.
Yields for the protection, molar absorptivity at 300 nm
(ꢀ300), photolytic half-life or time for 50% recovery of
thymidine (t1/2), and optimal deprotection yield were
recorded (Table 1). Several analogues showed improved
performance relative to the parent compound, unsubsti-
tuted S-pixyl protected thymidine 5a . With the exception
of the nitro analogues 5g and 5m , all substrates showed
decreased t1/2 values, and each gave deprotection yields
of 90% or greater. S-Pixyl analogues containing electron-
donating methoxy substituents in the thioxanthyl back-
bone or 9-aryl ring para to the 9-position, i.e., substrates
5b, 5d , 5e, and 5j, were most effective in producing a
photosolvolytic reaction in the shortest amount of time.
This result closely parallels the findings of trityl deriva-
tives when they were being developed for acid-labile
nucleoside protection.17 These analogues also produced
clean chromatograms that contained no oxidized thio-
xanthyl photoproducts, in comparison to 5a which pro-
duced significant amounts of thioxanthone upon irradia-
tion. Substrate 5i, with a m-methoxy substituent on the
9-aryl ring, also gave an excellent yield with a short
irradiation time with no thioxanthone produced. How-
ever, substrate 5l which contains a p-methoxy group on
the 9-aryl ring as well as a methoxy substituent on the
thioxanthyl backbone was not stable under the aqueous
reaction conditions, as it underwent a thermal deprotec-
tion reaction in the aqueous acetonitrile solvent.
Syn th esis of Su bstitu ted 5′-S-P ixyl-P r otected 2′-
Deoxyth ym id in e Eth er s. Thymidine was chosen as the
nucleoside substrate for studies of protection and photo-
deprotection in order to test substituent effects of S-pixyl
analogues. Thymidine is well-behaved in photodeprotec-
tion studies, and its behavior with the pixyl and S-pixyl
protecting groups are well-characterized.7,10 Compounds
3a -l and 4a ,b were chlorinated with trimethylsilyl
chloride with a DMSO catalyst15 in methylene chloride,
followed by solvent removal to generate the activated
protecting reagent. These activated groups were then
introduced to substrate thymidine in pyridine/DMAP16
to yield the derivatized 5′-O-[S-pixyl]-2′-deoxythymidine
nucleosides 5a -m (Scheme 3). Note that substrate letters
for compounds 3 and 5 do not necessarily correspond
since the entire set of derivatized thymidines was not
made. Protected nucleosides 5a -m were purified by silica
gel chromatography to give off-white solids in good to
excellent yield (Table 1). Compounds 5a -m were char-
1
acterized by UV/vis, H and 13C NMR spectroscopy, and
HRMS exact mass measurement.
Meta substitution is well-known to promote benzylic
heterolytic carbon-oxygen bond cleavage, so the en-
hanced photoreactivity of 5i compared to 5a was not
surprising. Unexpectedly, substrate 5c, with methoxy
substitution at both meta positions on the 9-aryl ring
gave considerably longer t1/2 values compared to other
analogues. m- and p-Methoxy substitution on the 9-aryl
ring in substrate 5e, and p-methoxy substitution in
substrate 5b was reasonably effective. m-Methoxy sub-
stitution on the thioxanthyl backbone, substrate 5j, was
the most effective in producing an S-pixyl analogue with
P h otod ep r otection Efficien cies. Solutions of 4.5-
5.0 × 10-4 M were prepared of compounds 5a -m in 50:
50 (v:v) water/acetonitrile in quartz photolysis tubes, the
highest percentage of water that solubility would allow.
Lower percentages of water resulted in increased forma-
tion of the secondary photoproduct thioxanthone for some
substrates. Reaction mixtures were thoroughly degassed
with argon for 15 min in a quartz photolysis tube prior
to irradiation. The irradiations were conducted using 16
(15) Snyder, D. C, J . Org. Chem. 1995, 60, 2638-2639.
(16) Chattopadhyaya, J . B.; Reese, C. B. J . Chem. Soc., Chem.
Commun. 1978, 639-640.
(17) Caruthers, M. H. Acc. Chem. Res. 1991, 24, 278.
J . Org. Chem, Vol. 67, No. 22, 2002 7643

Coleman and Boyd
% deprotectiona
TABLE 1. P r otection a n d Dep r otection Yield s for Ir r a d ia tion a t 300 n m of S-P ixyl An a logu es
compound
R1
R2
R3
R4
R5
R6
% protection
ꢀ300
t1/2 (min)
5a
5b
5c
5d
5e
5f
5g
5h
5i
H
MeO
H
Me
MeO
F
NO2
H
H
H
H
MeO
Me
H
H
MeO
H
MeO
H
Me
H
MeO
H
H
H
H
H
MeO
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
Br
H
H
H
H
H
H
H
H
H
H
H
MeO
H
MeO
H
H
H
H
H
H
H
H
H
H
H
Br
H
H
89
82
93
90
87
52
79
63
83
89
84
64
72
4560
3839
4849
4005
4259
3079
9007
7805
4690
4619
6876
8592
6617
10.1
4.8
7.3
4.3
5.2
9.3
70
97
93
92
95
94
90
72
92
91
97
94
78
4.2
4.4
3.6
6.6
50
5j
5k
5l
b
5m
NO2
N/Ab
-
a
% Deprotection based on recovery of thymidine in 50:50 (v:v) water/acetonitrile with irradiation at 300 nm. Substrate concentration
for irradiations was 4.5-5.0 × 10-4 M. Comprehensive data for compound 5l was problematic in that it was unstable for extended
b
lengths of time in aqueous media.
TABLE 2. Dep r otection Yield s for Ir r a d ia tion a t 350 n m of S-P ixyl An a logu es
compd
R1
R2
R3
R4
R5
R6
ꢀ320
ꢀ350
t1/2 (min)
% deprotectiona
5g
5j
5k
NO2
H
H
Me
H
H
H
H
H
H
H
Br
H
MeO
H
H
H
Br
N/A
2758
3748
1856
240
51
120
7.3
11.2
75
99
97
a
% Deprotection based on recovery of thymidine in 50:50 (v:v) water/acetonitrile with irradiation at 350 nm. Substrate concentration
for irradiation was 4.5-5.0 × 10-4 M.
enhanced photodeprotection performance. This is pre-
sumably due to more effective conjugation to the 9-carbon
compared to substitution on the 9-aryl ring which is
twisted out of planarity with the thioxanthyl backbone.
The nitrated substrates, 5g and 5m , performed poorly.
Nitro substitution was investigated due to large changes
in the long wavelength absorption of the derivatives.
o-Nitrotoluyl moieties are also the basis for the nitro-
benzyl oxycarbonyl protecting groups,3 and we wished
to compare the same compound structures. Despite the
strong molar absorptivity of these compounds, 5g and 5m
required extensive irradiation times and gave disap-
pointing yields. UV/vis spectra of the reaction mixtures
following prolonged irradiation display a strong absorp-
tion band characteristic of a stable aryl nitroso moiety,
the predominant product after R-hydrogen abstraction
from an excited-state nitro group and subsequent rear-
rangement to a possible nitroso benzyl alcohol.
pointing yields. However, substrates 5j and 5k show
excellent deprotection yields. The t1/2 values for depro-
tection at 350 nm are longer than for irradiation at 300
nm, 7.3 min compared to 3.6 min for 5g, but are still
reasonable for synthetic applications. Compounds 5j and
5k are both substituted directly on the thioxanthyl
backbone which has the largest effect on extending
the absorption wavelength. These findings suggest the
effective use of 9-arylthioxanthyl moieties as photo-
cleavable protecting groups in the synthesis of short
oligonucleotides.
Solven t Effects on P h otod ep r otection Yield s. The
5′-(R-methyl-2-nitropiperonyl)oxycarbonyl or MeNPOC
photocleavable protecting group has been used in the
light-directed synthesis of oligonucleotides. Deprotection
yields with the MeNPOC group exhibited a moderate
dependence on solvent polarity, with the shortest irradia-
tion times obtained in nonpolar solvents or the absence
of solvent.2d
We investigated the effect of solvent on the
Several analogues were selected that exhibited poten-
tial to undergo photodeprotection with irradiation at
350 nm. This longer wavelength irradiation is crucial for
the development of viable photodeprotecting groups for
nucleosides. The criteria for choosing a good analogue
target for longer wavelength irradiation included a short
time for deprotection and a reasonable molar absorptivity
above 300 nm. Substrate 5g was also selected, despite a
long t1/2 value, due to its high molar absorptivity at 350
nm. Solutions of substrates 5g, 5j, and 5k were irradiated
in a Rayonet photochemical reactor using 16 lamps at
350 nm and analyzed via HPLC. Molar absorptivities at
320 and 350 nm, the ꢀ320 and ꢀ350 values respectively,
photolytic half-life t1/2, and optimal deprotection yields
were recorded (Table 2). The parent unsubstituted S-pixyl
moiety 5a was ineffective as a photolabile protecting
group with irradiation at 350 nm due to minimal absorp-
tion at that wavelength. Nitro-substituted substrate 5g
gave poor results at this irradiation wavelength also,
requiring extended irradiation times and giving disap-
photodeprotection of 9-arylthioxanthyl protected thymi-
dine substrates. Protic solvents react with cations in the
order methanol > water . 2,2,2-trifluoroethanol since
solvents of poorer nucleophilicity are less effective in
quenching a ground or excited-state carbocation.18 To
investigate the effects of solvent on the photodeprotection,
compounds 5a and 5j were irradiated under identical
solution-phase conditions as described above, with the
exception that each reaction was carried out in a variety
of solvent conditions: 50/50 v:v methanol-acetonitrile,
10/90 v:v 2,2,2-trifluoroethanol-acetonitrile, and 50/50
v:v water-acetonitrile as a control. Compound 5j was
selected for this study since it was the S-pixyl analogue
with the best photodeprotection yields, short irradiation
times and no photooxidation products.
The time for 50% conversion (t1/2) and overall depro-
tection efficiency were recorded in each solvent (Table
(18) McClelland, R. A. Tetrahedron 1996, 52, 6823-6858.
7644 J . Org. Chem., Vol. 67, No. 22, 2002

Photocleavable Protecting Groups for Nucleosides
TABLE 3. Dep r otection Yield s a n d t1/2 Va lu es in
into an airtight Pyrex vessel and filled with argon.
Irradiations were carried out from 20 to 90min at 300
nm, after which the plates were washed with 10 mL of
50/50 water/acetonitrile and the washings analyzed by
HPLC. Three key points were monitored: the recovery
of thymidine over time as determined by HPLC analysis,
a clean reaction that would not produce excessive by-
products, and the absence of any T-T dithymidine
lesions that are known to occur upon exposure to irradia-
tions below 280 nm.8 Of the three investigated, only 5′-
O-[3-methoxy-9-phenylthioxanthyl]-2′-deoxythymidine
5j proved effective in photodeprotection under these
conditions, which produced an optimal deprotection yield
of 93% with no observation of secondary photoproducts.
We note, however, the addition of a protic solvent (water
or methanol) was necessary after irradiation in order to
recover the substrate nucleoside. Compounds 5g and 5k
needed prolonged irradiation times and produced complex
product mixtures.
Com p etitive Solven ts
compound
MeOHa
H2Ob
CF3CH2OHc
5a (t1/2, % yield)
5j (t1/2, % yield)
21.1, 91
17.5, 95
10.1, 97
3.6, 97
6.7, 94
2.0, 95
a
b
50% (v:v) in acetonitrile. 50% (v:v) in acteonitrile. c 10% (v:v)
in acetonitrile. All photolysis reactions were done at concentrations
ranging from 4-5.3 × 10-4 M using 16 300 nm lamps. Reactions
were quantitated by HPLC based on the recovery of substrate
thymidine, t1/2 values are in minutes and % yields are based on
best overall percent recoveries.
3). Shorter irradiation times for substrate 5j compared
to 5a were observed in each solvent. In addition, the t1/2
data for 5a and 5j show a significant solvent effect. The
t1/2 values decrease in the order methanol > water >
trifluoroethanol. The trend in the data can be evaluated
in terms of solvent parameters, although caution should
taken since solvent mixtures were used (50:50 methanol:
acetonitrile, 50:50 water:acetonitrile and 10:90 trifluoro-
ethanol:acetonitrile), and solvent parameters are not
available for those solvent compositions. Instead we
considered the trend in the t1/2 values using solvent
parameters for pure methanol, water and trifluoroetha-
nol. Only limited data are available for acetonitrile so
we could not attempt to account for the percent acetoni-
trile in the solvent mixtures. As the t1/2 values decrease,
the solvent nucleophilicity (NT and N′OTs) decreases,19
while solvent acidity and hydrogen bond donating ability
(R) increase.20,21 Solvent polarity (ET(30)) and solvent
ionizing ability (Y and YOTs) did not follow an increasing
or decreasing trend for these solvents.22 These interesting
trends will be explored in future work.
The trifluoroethanol-acetonitrile solvent mixture of-
fered a benefit in addition to lower t1/2 values. The
thioxanthyl cation produced in the photodeprotection
reaction was highly stable in this solvent mixture, and
allowed the extent of deprotection to be estimated
spectroscopically, with observation of the 9-phenylthiox-
anthyl cation at 394 nm and the 3-methoxy-9-phenyl-
thioxanthyl cation at 397 nm. In addition, this solvent
also decreased the amount of thioxanthone produced from
irradiation of 5a to negligible amounts. No thioxanthone
was produced from 5j in any solvent. The use of methanol/
acetonitrile also resulted in minimal thioxanthone forma-
tion from 5a compared to irradiation in water/acetoni-
trile. Substrates 5a and 5j were also irradiated in
anhydrous acetonitrile, which resulted in no deprotection
after 60 min irradiation time.
Con clu sion
Several 9-arylthioxanthyl derivatives have been proven
effective in the photodeprotection of the 5′ hydroxyl
functionality of thymidine under a variety of conditions.
Relative to the unsubstituted parent derivative (S-pixyl),
electron-donating methoxy and methyl substituents in
the position para to the 9-carbon on either the 9-aryl ring
or thioxanthyl backbone were most effective in achieving
near quantitative deprotection in the shortest amount
of time while producing the least amount of byprod-
ucts. Further, irradiations in 10% 2,2,2-trifluoroethanol
produced a highly stabilized carbocation, negligible by-
products, and decreased the irradiation times relative
to other solvents. The derivatives were also stable ther-
mally under this solvent condition. The use of a methoxy
substituent within the xanthyl backbone (i.e., compound
5j) combined with a protic, nonnucleophilic solvent
system thus produced the most effective reaction reported
herein.
Although no correlation between molar absorptivity
and an overall effectiveness as a photocleavable protect-
ing group was clearly observed among the derivatives,
the incorporation of substituents on the thioxanthyl ring
produced the greatest increase in molar absorptivity and
was directly related to a decreased time for irradiation.
The results reported herein now permit the use of
9-arylthioxanthyl moieties to be used as photocleavable
protecting groups in the synthesis of short oligonucleo-
tides with conventional solid-phase protocols.
Ir r a d ia tion in th e Absen ce of Solven t. Effective
solid-state deprotection has significant advantages in
photolithographic techniques.1c Solutions (3 mL, 6.64 M)
of compounds 5g, 5j, and 5k in 5% water/acetonitrile
were coated onto glass microscope slides of 2 × 4 cm to
produce a thin film of approximately 2.5 mmol/cm2, which
was calculated based on the volume and concentration
of the solution applied to the surface of known area. After
complete evaporation of solvent, the plates were placed
Exp er im en ta l Section
Gen er a l Meth od s. NMR spectra were obtained at 300 Hz
and 400 Hz for 1H, and 75 Hz and 91 Hz for 13C. 1H NMR
signals are reported in parts per million (δ) from TMS as an
internal standard. 13C NMR chemical shifts (δ) were reported
in reference to the 77.23 ppm peak for CDCl3, 39.45 ppm peak
for DMSO, and 23.21 peak for acetone, respectively. HPLC
analyses were conducted using a Zorbax column (XDB C18, 5
µm, 4.6 X 150 mm), with detection at 254 nm and a xanthene
internal standard, with 0.2% TFA (buffered to pH5.6)/ACN
gradient, 0.75 mL/min. Mass spectrometry was provided by
the Washington University Mass Spectrometry Resource, an
NIH Research Resource (Grant No. P41RR00954). Melting
points are uncorrected. All photochemical reactions were
(19) Kevill, D. N.; Anderson, S. W. J . Org. Chem. 1991, 56, 1845.
(20) CRC Handbook of Chemistry and Physics, 81st ed.; Lide, D.
R., Ed.; pp 816-818.
(21) Kamlet, M. J .; Abboud, J .-L. M.; Abraham, M. H.; Taft, R. W.
J . Org. Chem. 1983, 48, 2877.
(22) Smith, M. B.; March, J . March’s Advanced Organic Chemistry,
5th ed.; Wiley: New York, 2001; p 3.
J . Org. Chem, Vol. 67, No. 22, 2002 7645

Coleman and Boyd
conducted in a Rayonet RPR-100 reactor using 16 300 nm or
350 nm mercury lamps. Reaction mixtures were in quartz
tubes, degassed with argon, and inserted into a rotating
carousel. The reactor was air cooled with fans in a cold room
at 5 °C. Solid-state photolysis was conducted in a specially
prepared, Pyrex filtered vessel that held six 1.5 × 4 cm
microscope slides in a carousel. Pyridine, acetonitrile and
methylene chloride were distilled from calcium chloride.
Methanol was stored over molecular sieves. THF was distilled
from sodium-benzophenone ketyl. NMR grade 2,2,2-trifluo-
roethanol was purchased from a commercial source and used
without further purification.
2,7-Dibr om oth ioxa n th en -9-on e (1).12 Bromine (10 mL)
was added dropwise over 20 min to a solution of thioxanthone
(5.0 g, 0.024 mol) in acetic acid (45 mL). The solution was
stirred continuously and heated at reflux temperature for 20
h, after which half of the solvent volume was removed by
distillation. The reaction mixture was then allowed to cool and
poured over crushed ice. The resulting yellow precipitate was
collected by filtration and washed with saturated sodium
bicarbonate, a 20% aqueous solution of sodium bisulfite, and
finally with water. The solid was then dried under vacuum
and recrystallized twice from toluene to afford 4.8 g (54%) of
compound 1 as a bright yellow solid. mp 263-266 °C; 1H NMR
(400 MHz, CDCl3) δ 7.43 (2H, d), 7.78 (2H, d), 8.76 (2H, s);
13C NMR (100 MHz, CDCl3) δ 120.86, 127.47, 127.85, 130.32,
132.80, 135.81, 177.84. Exact mass (EI, M+) Calcd for C13H6-
Br2OS 367.8506, found 367.8508.
3-Meth oxyth ioxa n th en -9-on e (2). To concentrated sul-
furic acid (30 mL) was added sulfonyl chloride dropwise (2.4
mL) until the effervescence of hydrogen gas ceased. Thiosali-
cylic acid (4.0 g, 0.027 mol) was added slowly over 10 min to
generate a deep red solution. Anisole (8.5 g, 0.079 mol) was
then added dropwise to the initially orange solution slowly over
a 30 min period. The reaction mixture was heated to 60 °C
and allowed to stir for 3 h. The reaction mixture was then
cooled to room temperature and poured over ice. The resulting
orange precipitate was filtered and washed with water. The
solid was then collected and repeatedly washed with saturated
sodium bicarbonate and filtered in order to remove a major
byproduct, dithiosalicylic acid. The final solid was then rinsed
with water and then cold methanol and finally dried under
vacuum to afford 2.91 g (44%) of compound 2 as a pale yellow
solid. mp 115-118 °C; 1H NMR (400 MHz, CDCl3) δ 3.94 (3H,
s), 7.26 (1H, m), 7.48-7.59 (4H, m), 8.08 (1H, d), 8.63 (1H,
dd); 13C NMR (100 MHz, CDCl3) δ 55.91, 110.60, 122.93,
126.17, 126.27, 127.23, 128.51, 129.03, 130.07, 130.47, 132.20,
137.70, 158.61, 179.85. Exact mass (EI, M+) Calcd for C14H10O2S
242.0402, found 242.0389.
Syn th esis of Su bstitu ted 9-Ar ylth ioxa n th yl-9-ols. Ap-
propriate aryl bromides or chlorides (0.0534 mol) were added
dropwise over 30 min to refluxing THF containing Mg turnings
(0.0521 mol) before ethylene dibromide (1.68 mmol, 50 µL) was
added all at once. The solution was allowed to stir at reflux
temperature until all the Mg was consumed, at which point a
solution of thioxanthone (0.0241 mol) in THF was added
dropwise over 15min. Every 45 min aliquots of the reaction
mixture were taken and added to a vial containing THF and
aqueous NH4Cl. The organic layer was then separated and
subjected to TLC (10% hexane in methylene chloride) and then
stained with 15% sulfuric acid, which produced a deep red,
purple, or violet spot as an indication of product formation.
After 3 h the reaction mixture was allowed to cool, and 50 mL
of saturated aqueous ammonium chloride was added and
allowed to stir for 30 min or until any Mg salts were digested.
The mixture was then poured into a separatory funnel,
extracted three times with 50 mL of methylene chloride each.
The combined organic layers were then washed with 150 mL
of water and dried over MgSO4, and the solvent was evapo-
rated under reduced pressure to give a brown oil. The crude
product was then purified by flash chromatography (10%
hexane in methylene chloride) on silica gel and/or recrystal-
lized from toluene to give the target compound. Special
considerations are listed as indicated.
9-P h en ylth ioxa n th en ol (3a ). White powder (85%) mp
115-116 °C (lit. 103-107 °C);23 1H NMR (400 MHz, CDCl3) δ
2.78 (1H, s, D2O ex.), 6.96-7.41 (11H, m), 8.00 (2H, d); 13C
NMR (100 MHz, CDCl3) δ 126.15, 126.43, 126.63, 126.87,
127.24, 127.71, 128.01, 131.59, 139.98, 143.27, 198.08. Exact
mass (EI, M+) Calcd for C19H14OS 290.0765, found 290.0762.
9-(4-Meth oxyp h en yl)th ioxa n th en ol (3b). White powder
(76%) mp 105-109 °C; 1H NMR (300 MHz, CDCl3) δ 2.93 (1H,
br, D2O ex.), 3.51 (3H, s), 6.75 (4H, dd), 7.19-7.53 (6H, m),
7.97 (2H, d); 13C NMR (75 MHz, CDCl3) δ 55.23, 112.43, 125.73,
126.07, 126.58, 127.91, 127.47, 128.04, 131.81, 135.71, 140.02,
158.43. Exact mass (EI, M+) Calcd for C20H16O2S 320.0871,
found 320.0870.
9-(3,5-Dim eth oxyp h en yl)th ioxa n th en ol (3c). Following
the entrainment method,13 1 mol equiv of ethylene dibromide
was titrated into the THF solution containing the aryl halide
and magnesium metal over 36 h. The red brown solution was
then allowed to stir at reflux temperature for an additional
24 h at which time the reaction was executed in the usual
manner. Off-white powder (27%) mp 120-121 °C; 1H NMR
(300 MHz, CDCl3) δ 2.83 (1H, s, D2O ex.), 3.61 (6H, s), 6.18
(2H, d), 6.24 (1H, t), 7.21-7.63 (6H, m), 7.94 (2H, d); 13C NMR
(75 MHz, CDCl3) δ 55.49, 99.44, 105.93, 125.83, 126.48, 126.51,
126.63, 127.39, 129.89, 131.45, 132.28. Exact mass (EI, M+)
Calcd for C21H18O3S 350.0977, found 350.0981.
9-(4-Meth ylp h en yl)th ioxa n th en ol (3d ). White powder
(81%) mp 164-166 °C (lit. 161-163 °C);19 1H NMR (400 MHz,
CDCl3) δ 2.24 (3H, s), 2.79 (1H, s, D2O ex.), 6.81 (4H, q), 7.23-
7.49 (6H, m), 8.01 (2H, d); 13C NMR (100 MHz, CDCl3) δ 21.25,
126.20, 126.67, 126.86, 127.12, 127.42, 128.98, 131.77, 137.76,
140.35, 140.68. Exact mass (EI, M+) Calcd for C20H16OS
304.0922, found 304.0928.
9-(3,4-Dim eth oxyp h en yl)th ioxa n th en ol (3e). Following
the entrainment method,13 1 mol equiv of ethylene dibromide
(6.05 g, 0.0331mol) was titrated into the THF solution contain-
ing the aryl halide and magnesium metal over 10 h. The brown
solution was then allowed to stir at reflux temperature for an
additional 2 h at which time the reaction was executed in the
usual manner. Off-white powder (51%) mp 119-124 °C; 1H
NMR (300 MHz, CDCl3) δ 3.04 (1H, s, D2O ex.), 3.57 (3H, s),
3.69 (3H, s), 6.39 (1H, d), 6.57 (2H, d), 7.16-7.43 (6H, m), 7.91
(2H, d); 13C NMR (75 MHz, CDCl3) δ 55.61, 55.85, 109.94,
110.20, 110.82, 111.12, 119.07, 125.81, 126.31, 126.63, 127.09,
131.32, 135.73, 139.98, 148.21. Exact mass (EI, M+) Calcd for
C
21H18O3S 350.0977, found 350.0967.
9-(4-F lu or op h en yl)th ioxa n th en ol (3f). White powder
(79%); mp 134-135 °C (lit. 140-142 °C);19 1H NMR (300 MHz,
CDCl3) δ 2.90 (1H,s), 6.70-7.39 (10H, m), 7.92 (2H,d); 13C NMR
(100 MHz, CDCl3) δ 77.01, 115.05, 115.26, 126.29, 126.92,
127.01, 127.79, 129.15, 129.23, 131.76, 140.12, 161.85, 163.59.
Exact mass (EI, M+) Calcd for C19H13FOS 308.0671, found
308.0672.
9-(3-Meth ylp h en yl)th ioxa n th en ol (3g). Off-white solid
(88%); mp 103-105 °C; 1H NMR (400 MHz, CDCl3) δ 2.18 (3H,
s), 2.79 (1H, s), 6.7-7.42 (10H, m), 7.98 (2H, d); 13C NMR (100
MHz, CDCl3) δ 21.79, 77.23, 124.23, 126.36, 126.67, 126.84,
127.44, 127.75 128.08, 128.81, 131.74, 137.91, 140.23, 143.55.
Exact mass (EI, M+) Calcd for C20H16OS 304.0922, found
304.0910.
2-Ch lor o-9-p h en ylth ioxa n th en ol (3h ). After formation of
the Grignard reagent, the solution was added dropwise via Ar
partial pressure through a glass cannula tube into a solution
of 2-chlorothioxanthone in THF over a 30 min period. White
1
powder (83%); mp 182-183 °C; H NMR (300 MHz, CDCl3) δ
2.84 (1H, s, D2O ex.), 6.95 (2H, m), 7.16-7.43 (10H, m), 8.02
(2H, q); 13C NMR (100 MHz, CDCl3) δ 77.98, 126.30, 126.55,
126.95, 126.98, 127.08, 127.61, 127.74, 128.08, 128.29, 128.43,
(23) Bedlek, J . M.; Valentino, M. R.; Boyd, M. K. J . Photochem.
Photobiol. A 1996, 94, 7-13.
7646 J . Org. Chem., Vol. 67, No. 22, 2002

Photocleavable Protecting Groups for Nucleosides
130.44, 131.56, 132.95, 139.65, 142.13, 142.88. Exact mass (EI,
M+) Calcd for C19H13ClOS 324.0376, found 324.0373.
126.75, 127.00, 128.08, 131.00, 131.11, 133.60, 138.86, 148.17,
149.13. Exact mass (EI, M+) Calcd for C20H15NO3S 349.0773,
found 349.0771.
9-(3-Meth oxyp h en yl)th ioxa n th en ol (3i). White powder
(70%); mp 116-118 °C; 1H NMR (400 MHz, CDCl3) δ 2.79 (1H,
br, D2O ex.), 3.65 (3H, s), 6.56 (2H, m), 6.69 (1H, m), 7.05-
7.43 (10H, m), 7.98 (2H, m); 13C NMR (100 MHz, CDCl3) δ
55.24, 77.23, 112.79, 113.55, 119.57, 126.40, 126.67, 126.78,
127.52, 129.22, 131.69, 139.99, 145.43, 159.42. Exact mass (EI,
M+) Calcd for C20H16O2S 320.0871, found 320.0867.
3-Meth oxy-9-p h en ylth ioxa n th en ol (3j). White powder
(49%); mp 117-120 °C; 1H NMR (400 MHz, CDCl3) δ 2.96 (1H,
s, D2O ex.), 3.80 (3H, s), 6.82 (1H, q), 6.96 (2H, m), 7.13-7.40
(5H, m), 7.61 (2H, d), 8.00 (2H, q); 13C NMR (100 MHz, CDCl3)
δ 55.79, 77.50, 111.86, 113.89, 123.06, 126.13, 126.58, 127.02,
127.24, 127.50, 128.01, 128.11, 128.32, 132.64, 140.22, 141.98,
143.00, 159.22. Exact mass (EI, M+) Calcd for C20H16O2S
320.0871, found 320.0864.
2,7-Dibr om o-9-p h en ylth ioxa n th en ol (3k ). After forma-
tion of the Grignard reagent, the solution was added dropwise
via Ar partial pressure through a glass cannula tube into a
solution of 2,7-dibromothioxanthone (1) in THF over a 30 min
period. Off-white powder (66%); mp 244-245 °C, 1H NMR (400
MHz, CDCl3) δ 2.84 (1H, s), 6.96 (2H, m), 7.18 (2h, m), 7.42-
7.55 (5H, m), 8.28 (2H, d); 13C NMR (100 MHz, CDCl3) δ 77.06,
121.21, 127.80, 128.64, 128.90, 129.16, 130.50, 131.30, 131.41,
143.61. Exact mass (EI, M+) Calcd for C19H12Br2OS 445.8976,
found 445.8971.
3-Meth oxy-9-(4-m eth oxyp h en yl)th ioxa n th en ol (3l). Off-
white powder (59%); mp 99-104 °C; 1H NMR (400 MHz,
CDCl3) δ 2.92 (1H, s, D2O ex.), 3.67 (3H, s), 3.77 (3H, s), 6.64-
6.84 (4H, m), 7.23-7.41 (3H, m), 7.63 (2H, d), 8.01 (2H, q);
13C NMR (100 MHz, CDCl3) δ 55.30, 55.71, 68.98, 111.74,
113.53, 113.68, 122.99, 125.94, 126.50, 126.94, 127.33, 127.94,
128.60, 132.53, 135.31, 140.47, 142.26, 159.13, 159.18. Exact
mass (EI, M+) Calcd for C21H18O3S 350.0977, found 350.0972.
Syn t h esis of 5′-O-[9-P h en ylt h ioxa n t h yl]-2′-d eoxy-
th ym id in e Eth er s. Trimethyl silyl chloride (TMS-Cl) (6 mL)
was added dropwise over 10 min to the appropriate 9-arylthio-
xanthen-9-ol (1.480 mmol) dissolved in 10 mL of methylene
chloride and 200 µL of dimethyl sulfoxide. After 12 h the
solvent was removed under reduced pressure to generate the
appropriate 9-chloro-9-arylthioxanthene15 in nearly quantita-
tive yields that were then used in situ without further
purification. The red, purple, or violet solid was then dissolved
in 7 mL of pyridine and added dropwise to thymidine (1.345
mmol) and DMAP (0.041 mmol) in 10 mL of pyridine. TLC on
a silica plate indicated that the reaction was complete within
7-12 h. The reaction was quenched with 10 mL of 25%
aqueous Na2CO3, extracted three times with 30 mL of
methylene chloride and dried over magnesium sulfate. Evapo-
ration of the organic layer gave a faintly colored glassy oil that
was coevaporated with 20 mL of toluene to produce a foam.
The foam was then dissolved in methylene chloride and
subjected to silica gel chromatography using methylene chloride/
2-propanol or methylene chloride/acetone eluents. Sorbents for
chromatography were previously treated with a 1% TEA/
methylene chloride solution and allowed to stand for 30 min.
The solution was washed from the silica, rinsed with meth-
ylene chloride and then allowed to air-dry for 4 h. All
compounds were isolated as off-white powders.
5′-O-[9-P h en ylt h ioxa n t h yl]-2′-d eoxyt h ym id in e (5a ).
1
(79%) mp (dec) 137-145 °C. H NMR (300 MHz, DMSO-d6) δ
1.45 (s, 3H); 2.35 (m, 2H); 3.24 (dd, 2H); 3.95 (d, 1H); 4.41 (br,
1H),); 5.40 (t, 1H); 6.23 (t, 1H); 7.20-7.77 (m, 13H); 8.28 (s,
1H); 11.40 (s, 1H); 13C NMR (75 MHz, DMSO-d6) δ 12.90,
62.24, 70.46, 79.92, 82.0, 82.8, 109.98, 124.8, 125.5, 126.7,
127.0, 128.9, 128.2, 128.82, 129.1, 129.3, 130.0, 133.8, 133.9,
136.1, 148.1, 150.5, 163.9; Exact mass (EI, M+) Calcd for
9-(3-Nit r o-4-m et h ylp h en yl)t h ioxa n t h en ol (4a ). 9-(4-
Methylphenyl)thioxanthenol (3d ) (3.22 g, 0.0106 mol) was
azeotroped into 30 mL of acetic anhydride over 1 h. After the
compound was completely dissolved, 4.05 g of nitrated silica
gel14 was added slowly over 20 min while shaking the mixture.
Once all the nitrated silica gel was added, the solution was
allowed to heat near reflux temperature without stirring. After
3 h the solution was allowed to cool, the silica gel was removed
by filtration and the solution was poured into a separatory
funnel, followed by the addition of 100 mL of water and 100
mL of methylene chloride. After separation of the organic layer,
the aqueous layer was extracted with an additional 100 mL
of methylene chloride. The two organic layers were combined,
and the solution was added to 75 mL of saturated aqueous
K2CO3 and vigorously stirred for 1 h in order to deprotect any
9-thioxanthyl carbonate ester that may have formed. The
solution was then poured back into a separatory funnel,
washed an additional time with 75 mL of saturated aqueous
K2CO3, and then finally with two 50 mL portions of distilled
water. The organic layer was then evaporated to dryness under
reduced pressure to yield bright yellow crystals. The crude
product was then purified by preparative TLC (silica, 30/70
ethyl acetate/methylene chloride) to give a pale yellow solid
(2.56 g, 69%). mp 181-183 °C; FTIR cm-1 1580, 1375; 1H NMR
(400 MHz, acetone-d6) δ 2.45 (3H, s), 3.04 (1H, s, D2O ex.),
C
29H26N2O5S 514.1565, found 514.1562.
5′-O-[9-(4-Meth oxyp h en yl)th ioxa n th yl]-2′-d eoxyth ym i-
d in e (5b). (82%) mp (dec) 149-150 °C. 1H NMR (300 MHz,
DMSO-d6) δ 1.4 (s, 3H); 2.22 (m, 2H); 3.25 (q, 2H); 3.62 (s,
3H); 4.40 (m, 1H, D2O ex.); 4.95 (d, 1H); 5.39 (d, 1H); 6.23 (t,
1H); 6.80 (d, 2H); 7.18 (d, 1H); 7.20-7.68 (m, 8H); 11.25 (s,
1H, D2O ex.).13C NMR (75 MHz, DMSO-d6) δ 12.9, 55.0, 62.0,
70.1, 79.9, 83.8, 85.5, 109.8, 113.7, 125.4, 126.25, 126.6, 127.7,
128.0, 128.7, 128.9, 129.3, 130.1, 134.2, 134.5, 135.7, 140.0,
150.1, 158.0, 163.8. Exact mass (FAB, M + Na+) Calcd for
C
30H28N2O6SNa 567.1566, found 567.1564.
5′-O-[9-(3,5-Dim eth oxyph en yl)th ioxan th yl]-2′-deoxyth y-
m id in e (5c). (93%) mp (dec) 169-172 °C. 1H NMR (400 MHz,
DMSO-d6) δ 1.40 (s, 3H); 2.25 (m, 2H); 3.20-3.33 (m, 2H); 3.63
(s, 3H); 3.95 (d, 1H); 4.39 (br, 1H, D2O ex.); 5.40 (m, 1H); 6.25
(t, 1H); 6.34 (s, 1H); 6.40 (d, 2H); 7.22-7.62 (m, 9H); 11.35 (s,
1H, D2O ex.). 13C NMR (100 MHz, DMSO-d6) δ 11.30, 59.8,
63.88, 70.38, 79.24, 83.58, 85.06, 97.28, 103.44, 109.94, 125.11,
125.19, 126.38, 127.76, 127.98, 128.63, 128.71, 128.98, 129.63,
133.29, 133.44, 135.26, 150.11, 159.99, 163.38. Exact mass
(FAB, M + Na+) Calcd. for C31H30N2O7SNa 597.1671, found
597.1648.
5′-O-[9-(4-Met h ylp h en yl)t h ioxa n t h yl]-2′-d eoxyt h ym i-
d in e (5d ). (90%) mp (dec) 165-167 °C. 1H NMR (300 MHz,
DMSO-d6) δ 1.40 (s, 3H); 2.15 (m, 2H); 3.35 (m, 2H); 3.95 (dd,
1H); 4.35 (m, 1H); 6.25 (t, 1H); 7.0-7.65 (m, 12H). 13C NMR
(100 MHz, DMSO-d6) δ 12.0, 20.5, 64.05, 70.5, 79.5, 84.0, 85.5,
109.5, 124.9, 125.3, 125.4, 126.5, 127.8, 128.0, 128.7, 128.8,
129.0, 129.3, 130.0, 133.9, 134.2, 135.3, 136.1, 145.0, 151.0,
164.5. Exact mass (FAB, M + Na+) Calcd for C30H28N2O5SNa
551.1617, found 551.1641.
7.10 (2H, m), 7.28-7.43 (6H, m), 7.67 (1H, s), 7.95 (2H, q); 13
C
NMR (100 MHz, acetone-d6) δ 20.25, 76.17, 123.11, 126.22,
126.89, 127.09, 128.08, 131.30, 131.44, 132.56, 133.00, 139.08,
143.24, 149.02. Exact mass (EI, M+) Calcd for C20H15NO3S
349.0773, found 349.0773.
9-(3-Met h yl-4-n it r op h en yl)t h ioxa n t h en ol (4b ). 9-(3-
Methylphenyl)thioxanthenol (3g) was subjected to the same
conditions as in 4a followed by column chromatography to yield
an off-white solid (75%); mp 172-174 °C; 1H NMR (300 MHz,
acetone-d6) δ 2.41 (3H, s), 2.87 (1H, br), 6.96-7.04 (2H, m),
7.27-7.41 (6H, m), 7.69 (1H, d), 7.89 (2H, q); 13C NMR (100
MHz, acetone-d6) δ 20.96, 76.14, 124.64, 125.37, 126.37, 126.56,
5′-O-[9-(3,4-Dim eth oxyph en yl)th ioxan th yl]-2′-deoxyth y-
m id in e (5e). (87%) mp (dec) 160-161 °C. 1H NMR (400 MHz,
DMSO-d6) δ 1.38 (s, 3H); 2.24 (m, 2H); 3.59 (s, 3H); 3.66 (s,
3H); 3.96 (dd, 1H); 4.23 (m, 2H); 5.40 (d, 1H); 6.25 (t, 1H);
6.56-7.66 (m, 12H); 11.35 (s, 1H). 13C NMR (100 MHz, DMSO-
J . Org. Chem, Vol. 67, No. 22, 2002 7647

Coleman and Boyd
d6) δ 11.53, 54.88, 55.27, 55.47, 64.19, 70.77, 79.67, 83.82,
85.40, 108.97, 109.59, 111.54, 117.95, 125.44, 125.89, 126.30,
126.55, 126.59, 127.06, 127.79, 128.07, 128.65, 128.86, 129.30,
130.12, 134.10, 134.36, 135.60, 140.10, 147.76, 148.24, 150.35,
163.60. Exact mass (FAB, M + Na+) Calcd for C31H30N2O7-
SNa 597.1671. Found 597.1668.
114.36, 114.60, 115.53, 121.45, 122.50, 125.77, 125.85, 125.89,
126.96, 126.38, 126.43, 126.78, 126.82, 127.45, 127.48, 127.87,
128.09, 128.43, 128.49, 129.03, 129.73, 131.24, 131.86, 134.22,
135.44, 135.76, 136.00, 147.13, 150.68, 158.51, 158.73, 163.96,
164.00. Exact mass (FAB, M + Na+) Calcd for C30H28N2O6-
SNa 567.1566, found 567.1574.
5′-O-[9-(4-F lu or op h en yl)t h ioxa n t h yl]-2′-d eoxyt h ym i-
d in e (5f). (52%) mp (dec) 154-156 °C. 1H NMR (400 MHz,
DMSO-d6) δ 1.47 (s, 3H); 2.22 (m, 2H); 3.26 (m, 2H); 3.95 (m,
1H); 4.38 (m, 1H); 6.24 (t, 1H); 7.08-7.62 (m, 13H); 11.35 (br,
1H). 13C NMR (100 MHz, DMSO-d6) δ 11.4, 64.18, 70.46, 79.40,
83.71, 85.11, 109.54, 114.92, 115.13, 125.47, 125.52, 126.72,
126.93, 127.02, 128.02, 128.25, 128.79, 128.89, 129.31, 129.95,
133.64, 133.80, 135.68, 144.18, 150.34, 159.62, 162.05, 163.58.
Exact mass (FAB, M + Na+) Calcd for C29H25FN2O5SNa
555.1366, found 555.1375.
5′-O-[2,7-Dib r om o-9-p h en ylt h ioxa n t h yl]-2′-d eoxyt h y-
m id in e (5k ). (84%) mp (dec) 180-181 °C. 1H NMR (400 MHz,
DMSO-d6) δ 1.43 (s, 3H); 2.23 (m, 2H, J ) 6.59); 3.28-3.42 (t,
1H); 3.93 (t, 1H); 4.44 (q, 1H); 5.46 (d, 1H); 6.28 (t, J ) 6.59,
1H); 7.22-7.71 (m, 12H); 11.20 (s, 1H). 13C NMR (100 MHz,
DMSO-d6) δ 11.74, 64.53, 69.97, 70.07, 79.26, 83.26, 84.64,
109.91, 119.52, 124.61, 126.28, 127.59, 127.63, 127.69, 128.36,
128.43, 128.69, 129.02, 131.09, 131.22, 131.28, 135.73, 135.80,
135.90, 147.05, 150.38, 150.46, 163.55. Exact mass (FAB, M
+ Na+
) Calcd for C29H24Br(81)N2O5SNa 694.965, found 694.964.
5′-O-[9-(3-Meth yl-4-n itr op h en yl)th ioxa n th yl]-2′-d eoxy-
th ym id in e (5g). (79%) mp (dec) 164-166 °C. 1H NMR (400
MHz, CDCl3) δ 1.72 (s, 3H); 2.18 (s, 2H); 2.49 (s, 3H); 3.22 (br,
1H, D2O ex.); 3.36 (m, 2H); 4.13 (q, 1H); 4.43 (d, 1H); 6.37 (t,
1H); 7.10-7.45 (m, 11H), 7.85 (d, 1H), 9.60 (br, 1H, D2O ex.);
13C NMR (100 MHz, DMSO-d6) δ 11.88, 20.01, 64.39, 70.34,
79.12, 83.84, 84.97, 109.50, 123.95, 124.84, 125.49, 125.63,
126.83, 126.89, 128.36, 128.44, 128.60, 129.22, 129.28, 129.41,
129.81, 132.51, 132.58, 132.96, 135.71, 147.38, 150.35, 153.17,
163.62. Exact mass (FAB, M + Na+) Calcd for C30H27N3O7-
SNa 596.1467, found 596.1469.
5′-O-[3-Meth oxy-9-(4-m eth oxyp h en yl)th ioxa n th yl]-2′-
d eoxyth ym id in e (5l). Isolated as a mixture of isomers (84%).
mp (dec) 163-164 °C. H NMR (400 MHz, DMSO-d6) δ 1.34-
1
1.43 (d, 3H); 2.09 (s, 3H); 2.19-2.42 (m, 2H); 3.29-3.52 (m,
2H); 3.67 (s, 3H); 3.99 (m, 1H, D2O ex.); 4.41-4.58 (dq, 2H);
6.31 (dt, 1H); 6.82-7.74 (m, 10H); 8.06 (d, 2H); 11.25 (d, 1H,
D2O ex.). 13C NMR (100 MHz, DMSO-d6) δ 11.49, 11.60, 54.80,
55.00, 55.03, 63.57, 64.47, 70.45, 70.48, 79.96, 80.01, 83.56,
83.68, 85.23, 109.70, 109.75, 112.88, 113.08, 113.51, 113.63,
114.73, 115.38, 120.38, 121.24, 124.99, 125.59, 125.91, 126.33,
126.45, 126.68, 126.73, 127.59, 127.82, 128.22, 128.26, 128.53,
129.73, 130.54, 131.54, 133.15, 134.30, 135.68, 135.76, 135.83,
136.18, 138.94, 150.40, 158.10, 158.13, 158.19, 158.24, 163.57,
163.63, 167.63. Exact mass (FAB, M + Na+) Calcd for
5′-O-[2-Ch lor o-9-p h en ylt h ioxa n t h yl]-2′-d eoxyt h ym i-
d in e (5h ). Isolated as a mixture of isomers (63%). mp (dec)
1
174-176 °C. H NMR (400 MHz, CDCl3) δ 1.59 (d, 3H); 2.38
(m, 2H); 3.10 (m, 2H); 3.37 (m, 2H); 4.09 (s, 1H); 4.57 (s, 1H),
6.40 (q, 1H); 7.13-7.53 (m, 13H); 9.35 (d, 1H). 13C NMR (100
MHz, CDCL3) δ 12.27, 40.95, 40.99, 64.33, 64.36, 72.46, 72.54,
80.87, 80.90, 85.19, 85.24, 86.03, 86.13, 111.54, 111.63, 125.70,
125.75, 125.80, 126.79, 126.89, 126.99, 127.63, 127.68, 128.19,
128.30, 128.36, 128.52, 128.56, 128.59, 129.26, 129.38, 129.62,
129.75, 129.88, 130.79, 132.18, 132.29, 133.57, 133.87, 135.73,
135.81, 135.96, 136.46, 146.97, 147.29, 150.78, 164.13. Exact
mass (FAB, M + Na+) Calcd for C29H25ClN2O5SNa 571.107,
found 571.109.
C
31H30N2O7SNa 597.1671, found 597.1683.
5′-O-[9-(3-Nitr o-4-m eth ylp h en yl)th ioxa n th yl]-2′-d eoxy-
1
th ym id in e (5m ). (64%) mp (dec) 175-177 °C. H NMR (400
MHz, CDCl3) δ 1.54 (s, 3H); 2.21 (m, 2H); 2.43 (s, 3H); 2.51
(br, 1H); 3.24 (m, 2H); 3.96 (q, 1H); 4.30 (m, 1H); 5.39 (d, 1H,
D2O ex.); 6.21 (t, 1H); 7.21-7.55 (m, 11H); 7.89 (d, 1H); 11.36
(s, 1H, D2O ex.). 13C NMR (100 MHz, CDCl3) δ 12.61, 21.13,
40.92, 64.54, 72.42, 80.26, 85.90, 111.37, 124.19, 125.02,
125.76, 125.81, 126.77, 126.87, 128.57, 128.68, 129.81, 129.88,
129.99, 130.82, 131.01, 132.63, 132.93, 133.96, 135.53, 147.87,
150.67, 153.47, 164.03. Exact mass (FAB, M + Li+) Calcd for
5′-O-[9-(3-Meth oxyp h en yl)th ioxa n th yl]-2′-d eoxyth ym i-
d in e (5i). (83%) mp (dec) 138-141 °C. 1H NMR (400 MHz,
DMSO-d6) δ 1.46 (s, 3H); 2.25 (m, 2H); 3.23 (m, 2H); 3.63 (s,
3H); 3.91 (s, 1H); 4.38 (s, 1H); 5.39 (s, 1H); 6.24 (t, 1H); 6.76
(q, 2H); 6.85 (s, 1H); 7.14-7.61 (m, 10H), 11.35 (s, 1H). 13C
NMR (100 MHz, DMSO-d6) δ 11.63, 54.91, 64.64, 70.64, 79.52,
83.76, 85.24, 109.63, 111.32, 117.41, 125.36, 125.42, 126.61,
127.96, 128.18, 128.91, 129.00, 129.26, 129.45, 129.92, 133.64,
133.80, 135.55, 150.27, 150.35, 158.96, 163.61. Exact mass
(FAB, M + Na+) Calcd for C30H28N2O6SNa 567.1566, found
567.1564.
5′-O-[3-Meth oxy-9-p h en ylth ioxa n th yl]-2′-d eoxyth ym i-
d in e (5j). Isolated as a mixture of isomers (89%) mp (dec)
171-173 °C. 1H NMR (400 MHz, DMSO-d6) δ 1.50 (d, 3H);
2.37 (m, 2H); 2.96 (br, 1H); 3.32-3.59 (m, 2H); 3.65 (d, 3H);
4.07 (q, 1H); 4.63 (m, 1H), 6.43 (dt, 1H); 6.80-6.90 (m, 1H);
7.01-7.65 (m, 12H); 9.20 (s, 1H). 13C NMR (100 MHz, DMSO-
d6) δ 12.13, 41.04, 41.07, 55.41, 55.54, 63.89, 64.07, 72.45,
72.61, 81.37, 81.39, 84.92, 85.05, 86.16, 86.21, 111.50, 111.60,
C
30H27N3O7SLi 580.173, found 580.173.
Ack n ow led gm en t is made to the donors of the
Petroleum Research Fund, administered by the Ameri-
can Chemical Society, and the Department of Education
GAANN program for support of this research. We also
thank the NSF-REU program (CHE-9619709) and the
Washington University Mass Spectrometry Resource,
Chemistry wing (Grant No. P41RR00954). We thank
Candace Ponsaran for her assistance in the preparation
of some substrates.
Su p p or tin g In for m a tion Ava ila ble: 13C NMR spectra of
1, 2, 3b-c, 3e, 3g-l, 4a -b, 5a -m . This material is available
free of charge via the Internet at http://pubs.acs.org.
J O026009W
7648 J . Org. Chem., Vol. 67, No. 22, 2002