Synthesis and reactivity of (RS)-6-chloro-7- or 9-(1,2,3,5-tetrahydro-4,1-benzoxazepin-3-yl)-7H- or 9H-purines bearing a nitrobenzenesulfonyl group on the nitrogen atom

Article abstract of DOI:10.1016/j.tet.2007.03.155

The O,O-acetalic compounds (RS)-3-methoxy-1-[(2)- or (4)-nitrobenzenesulfonyl)]-1,2,3,5-tetrahydro-4,1-benzoxazepine have been studied in the Lewis acid-mediated condensation with 6-chloropurine. 6-Chloropurine leads to the N-7″ aminalic bond in the cycli

Full text of DOI:10.1016/j.tet.2007.03.155

Tetrahedron 63 (2007) 5274–5286
Synthesis and reactivity of (RS)-6-chloro-7- or 9-(1,2,3,5-
tetrahydro-4,1-benzoxazepin-3-yl)-7H- or 9H-purines bearing
a nitrobenzenesulfonyl group on the nitrogen atom
a
Monica Dıaz-Gavilan, Duane Choquesillo-Lazarte, Josefa M. Gonzalez-Perez,
b
b
ꢀ
ꢀ
Miguel A. Gallo, Antonio Espinosa and Joaquın M. Campos
ꢀ
ꢀ
ꢀ
a
a
a,
*
ꢀ
a
´
ꢀ
Departamento de Quꢀımica Farmaceutica y Organica, Facultad de Farmacia, c/ Campus de Cartuja s/n, 18071 Granada, Spain
bDepartamento de Quꢀımica Inorgaꢀnica, Facultad de Farmacia, c/ Campus de Cartuja s/n, 18071 Granada, Spain
Received 21 February 2007; revised 23 March 2007; accepted 26 March 2007
Available online 30 March 2007
Abstract—The O,O-acetalic compounds (RS)-3-methoxy-1-[(2)- or (4)-nitrobenzenesulfonyl)]-1,2,3,5-tetrahydro-4,1-benzoxazepine have
been studied in the Lewis acid-mediated condensation with 6-chloropurine. 6-Chloropurine leads to the N-700 aminalic bond in the cyclic
products and mainly to the N-900 aminalic bond in the acyclic ones. Substitution of the chlorine atom at the 600 position of the purine moiety
is more feasible when the ring is alkylated at N-700 than at N-900. Exchange with a hydroxyl group is performed with water traces in deuterated
dimethylsulfoxide at room temperature in a solvent-mediated process. The exchange with strong nucleophiles (e.g., thiophenol) does not need
further activation.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
agents on MCF-7 human breast cancer cells.1,2 The synthesis
of the O,N-acetals was accomplished from the correspond-
ing O,O-acetals3 in a Lewis acid-mediated process.
In our effort to prepare antiproliferative agents with an ace-
talic function, our group has reported the synthesis and
characterization of 1,2,3,5-tetrahydro-4,1-benzoxazepines
O,O- and pyrimidine O,N-acetals of type I, II and III
(Fig. 1), which turned out to be moderate antiproliferative
Both cyclic and acyclic O,N-acetals (II and III), in which the
pyrimidines are linked to the C-3 atom through their N-100
and N-300 atoms, were products in the reaction between 1-
nitrobenzenesulfonyl-1,2,3,5-tetrahydro-4,1-benzoxazepines
(1 and 21) and the nitrogen bases. After a report on the fac-
tors that control the regioselectivity in this reaction,2 it is our
objective to complete the study with a description of the con-
densation process between the same substrates and a purinic
base. In this paper, new data about this mechanism are re-
ported and the stability of the final compounds is discussed.
6-Chloropurine (6-CP) has been used due to its versatility to
undergo further modifications. The final O,N-acetals were
evaluated as antiproliferative agents against the MCF-7 can-
cerous cell line. Comparison of the biological activities
found for pyrimidine and purine derivatives has given us
further information about the role of the benzoxazepine
moiety and the efficiency of this family of compounds to
act as drugs endowed with a new anticancer mechanism of
action.
R2
R1
N
R1
N
H3CO
N
R1
OMe
R2
O
O
OH
I
II
III
R1: -H, -SO2R
R1: -H, -CH2Ph, -COR,
-COOR, -SO2R
R2: Uracil-1-yl, uracil-3-yl,
5-fluorouracil-1-yl,
5-fluorouracil-3-yl
Figure 1. Structures of the previously described O,O- and O,N-acetalic
compounds type I, II and III.
The Lewis acid-mediated transformation of O,O- into O,N-
acetals has been widely described in bibliography and the
existence is well known of several factors with regard to
the nucleophile, the electrophile and the Lewis acid, that
Keywords: Acetals; Benzoxazepines; Medium ring heterocycles; Purines;
X-ray.
* Corresponding author. Tel.: +34 958 243848; fax: +34 958 243845;
e-mail: jmcampos@ugr.es
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.03.155

M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
5275
R2
R1
N
Cl
N
R1
N
R1
N
N
a)
N
OMe
+
+
N
H
R2
OMe
O
O
OH
4
3
1 R1 = -SO2C6H5-NO2-(2)
2 R1 = -SO2C6H5-NO2-(4)
R2
=
N
N
N
N
X
N
N
N
9H-9-purinyl
X = -Cl, -OH, -SPh
X
N
7H-7-purinyl
Scheme 1. Condensation reactions between O,O-acetals and 6-chloropurine. Reagents and conditions: (a) HMDS (4.0 equiv), TMSCl (4.0 equiv), SnCl4
(4.0 equiv), anhydrous MeCN, 50 ꢀC, more than 48 h.
can influence the velocity and the regioselectivity of this
process.3 Purine alkylation follows different approaches
such as nucleophile attack on halides,4,5 Mitsunobu acti-
vated electrophiles,6,7 or alkylation by modification of the
fact that has not being previously reported in bibliography,
and these compounds may serve as prototypes for the devel-
opment of even more potent structures. Conditions reported
as optimal in the reaction with pyrimidines2 have been used
in the preparation of the purine derivatives: silylating agents
1,1,1,3,3,3-hexamethyldisilazane (HMDS, 4.0 equiv) and
chlorotrimethylsilane (TMSCl, 4.0 equiv), the Lewis acid
tin(IV) tetrachloride (SnCl4, 4.0 equiv), anhydrous MeCN
at 50 ꢀC under an inert atmosphere.
€
Vorbruggen reaction, and mainly gives rise to alkylation at
the N-900 purine atom. Fewer cases of regioselective N-700
alkylation have been reported.8,9 The high number of factors
affecting the regioselectivity makes it difficult to success-
fully predict the position of alkylation3,8–11 and this fact
has caused many authors to build up the purine ring from
the appropriate synthons.12
2. Results and discussion
Our aim is to contribute with further data presenting a case of
N-700 regioselective alkylation of 6-chloropurine, together
with other results obtained in the Lewis acid-mediated alkyl-
ation of 6-chloropurine with the sulfonamides 1 and 2, and
the carbamate 31 (Scheme 1). All the N-900 and N-700 purine
O,N-acetals13 show in vitro antiproliferative activities
against the MCF-7 human breast cancer cell line in the sub-
micromolar range (data not shown). This is an outstanding
2.1. Lewis acid-mediated reaction between (RS)-3-
methoxy-N-nitrobenzenesulfonyl-1,2,3,5-tetrahydro-
4,1-benzoxazepines (1 and 2) and 6-chloropurine
Condensation of the sulfonyl compounds 1 and 2 with 6-
chloropurine produced the cyclic and acyclic O,N-acetals
showed in Table 1.
Table 1. Products of the reaction between the O,O-acetals 1 and 2, and 6-chloropurine (50 ꢀC, 69 h)
R
R
R
R
Cl
N
N
O
S
N
O R2
O
S
N
O
O
S
N
O
O
S
N
O
N
H
N
OMe
OMe
+
+
R2
OMe
O
O
OH
OH
5
3
4
1 R = (2)-NO2
2 R = (4)-NO2
Substrate
R
R2
Products (yield, %)
1
1
1
1
(2)-NO2
(2)-NO2
(2)-NO2
(2)-NO2
6-Chloropurin-9-yl
6-Chloropurin-7-yl
6-Chloropurin-9-yl
6-Chloropurin-7-yl
3a (2)
3b (32)
4a (7)
4b (6)
3a+3b (34)
4a+4b (13)
3a+3b+4a+4b (47)
3c+3d+4c+4d (39)
2
2
2
2
(4)-NO2
(4)-NO2
(4)-NO2
(4)-NO2
6-Chloropurin-9-yl
6-Chloropurin-7-yl
6-Chloropurin-9-yl
6-Chloropurin-7-yl
3c (2)
3d (15)
4c (22)
—
3c+3d (17)
4c+4d (22)
1
2
(2)-NO2
(4)-NO2
—
—
5aa (2)
5bb (7)
5aa (2)
5bb (7)
a
(E+Z) mixture.
E isomer.
b

5276
M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
From these results, the following can be observed:
chloropurine when compared to the reaction between 2
[R¼–SO2–C6H4–NO2–(4)] and 6-chloropurine. This ap-
proximation has been previously reported as an important
factor influencing the regioselectivity of 1 and 2 condensa-
tions with pyrimidines.2
(1) Substrates 1 and 2 afford similar total yields of the O,N-
acetalic products, a value that is slightly higher for 1
(47% vs 39%).
(2) In both cases, the formation of cyclic O,N-acetals oc-
curred regioselectively, the purine being linked through
its N-700 atom (3a/3b¼2/32, 3c/3d¼2/15). However, the
O,N-acetalic compound in the acyclic series was selec-
tive only when 2 was the substrate, leading exclusively
to the N-900 compound (4a/4b¼7/6, 4c/4d¼22/0). In no
case was formation of N-300 alkylated purines detected.
(3) The ratio of cyclic/acyclic O,N-acetals was different de-
pending on the substrate of the reaction: from 1 the high-
est yield was found for cyclic products (3a,b/4a,b¼34/
13), while from 2, the acyclic derivatives were the
main products (3c,d/4c,d¼17/22).
According to that previously described cyclic O,N-acetals
could also be formed by cyclization of carbenium ions
type 8 or 92,14 (step C, Scheme 2). Both ions with an N-900
or N-700 alkylated purine (8 or 9, respectively) could be sub-
jected to a stabilizing cession of a negative charge towards
the cationic position exerted by the sulfonyl group. We
have postulated that in ion type 7 one of its oxygen atoms
approximated to the positively charged positions (2.19–
˚
2.28 A). The intramolecular attack of the nucleophile benzyl
hydroxy group at this point could be easiest in the N-700 alky-
lated regioisomers due to the ꢁI effect exerted by the chlo-
rine atom on the acetalic position and the less favoured
delocalization of the positive charge along the purine ring.
In accordance with what was found in the condensation re-
action with pyrimidines, the different cyclic/acyclic and N-
700/N-900 product ratios indicate that the change of the nitro
group on the benzenesulfonyl moiety from ortho to para
can modify the structural characteristics of the reaction in-
termediates in such a way as to influence the behaviour of
the reaction. Our group has previously reported a SN1
mechanism explaining the formation of O,N-acetals from
(RS)-3-methoxy-2,3-dihydro-5H-1,4-benzodioxepines14 or
(RS)-3-methoxy-1,2,3,5-tetrahydro-4,1-benzoxazepines2 and
pyrimidines, in which cyclic (6) and acyclic (7) oxocarb-
enium ions of the substrates are formed (Scheme 2). The
formation of the final cyclic or acyclic compounds depended
on the rate of formation of both ion types (steps A and B), the
affinity of the nucleophile for each ion type and the possibil-
ity of a later cyclization of the acyclic products (step C).
The aminalic bond between the C-3 atom of the seven-mem-
bered moiety and the N atom of the purine fragment depends
on steric and electronic factors. Once silylation of purines
has taken place on the imidazole ring, the equilibrium is set-
tled between the N-700 silyl and the most favoured N-900 silyl
tautomers (Scheme 3).15 Each tautomer reacts with the elec-
trophile or coordinates to the Lewis acid through its free nu-
cleophile position. If there is no coordination to the Lewis
acid, i.e., in the case of electron-poor purines or in a nucleo-
phile solvent such as acetonitrile,3 the main product would
be the N-700 alkylated purine, formed from the prevalent silyl-
ated tautomer N-9 (Scheme 3) (examples can be found from
regioselective N-700 isomers in reactions of nucleoside for-
mation carried out in acetonitrile3,9).
Therefore, any stabilizing effect which can increase the pop-
ulation of one of the intermediate ions facilitated the forma-
tion of the final type of O,N-acetal. It has been observed that
in the case of 6, when the nitro group is ortho, there is an ap-
proximation of the negative oxygen towards the positive
When the formation of s complexes between the Lewis acid
and the purine takes place, the situation becomes more com-
plex. The transformation is possible from the kinetic product
(N-700 alkyl purine) to the thermodynamic one (N-900 alkyl-
purine).3,9,16 The velocity of this process depends on the ac-
tivation energy of the transition state and therefore the ratio
of the N-700/N-900 alkyl products is different depending on the
structure of the starting material. From the results that have
been obtained in our case, this process could take place in
˚
charge on the benzoheterocycle (up to 2.67 A). Therefore
stabilization of the cyclic ions occurred.2 This fact is consis-
tent with the higher yields of cyclic compounds found in the
reaction between 1 [R¼–SO2–C6H4–NO2–(2)] and 6-
R1
N
R1
N
R1
N
R1
N
A
Cl-
OMe
OMe
Nu
O
O
O
O
3
1 R1 = -SO2-C6H4-NO2-(2)
2 R1 = -SO2-C6H4-NO2-(4)
6
[SnCl4]
TMSCl
B
C
R1
N
R1 OMe
N
Cl-
R1
N
Cl-
OMe
SnCl4
Nu
Nu
[SnCl3-OMe]
O-TMS
O-TMS
O-TMS
8 Nu = 6-Chloropurin-9-yl
9 Nu = 6-Chloropurin-7-yl
7
Scheme 2. Proposed mechanism for the Lewis acid-assisted transformation of O,O- into O,N-acetals.

M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
5277
Cl
Cl
N
Cl
E
N
N
N
N
N
N
N
+
N
H
N
N
E
E+ Z-
TMS-Z
N
Si
N-9 Silyl tautomer
N-7 Alkylpurine
N-3 Alkylpurine
Cl
N
(main product)
(not isolated)
N
N
TMS-Cl
N
H
Cl
Cl
Cl
N
Si
N
N
N
N
N
N
+
N
N
N
N
N
E+ Z-
TMS-Z
H
E
E
N-7 Silyl tautomer
N-9 Alkylpurine
(main product)
N-3 Alkylpurine
(not isolated)
Scheme 3. Tautomeric equilibrium of silylated 6-chloropurine and alkylation in the absence of Lewis acid coordination. Nucleophile nitrogen atoms are
indicated within dotted lines. E+Zꢁ: electrophile (oxocarbenium ion) and counterion.
acyclic products but does not seem to happen in the cyclic
ones.
900 transformation, via elimination of the purine ring after
its activation as a leaving group by the coordination of the
Lewis acid to one of its electron-rich positions. This elimina-
tion could happen in an E1 or E2 fashion and it would be eas-
ier for the highest energy N-700 regioisomers than for the
more stable N-900 ones. The formation of olefins would be
justified by the resonance of the final electronic system.
2.2. Other products in the reaction of 1 and 2 with
6-chloropurine
Compounds 6a and 6b were obtained along with the cyclic
and acyclic O,N-acetals in the reaction of 6-chloropurine
with 1 and 2, respectively. Their yields are low (<10%)
but their importance lies in the information that they provide
on the mechanism of the reaction with 6-chloropurine as
none of these olefins have been isolated in the corresponding
reactions with U (uracil) or 5-FU (5-fluorouracil).
According to an E1 mechanism, loss of the purine ring would
lead to oxocarbenium ions, which could either attack nucleo-
phile positions of the purinegiving rise to the formation of the
N-900 O,N-acetals, or eliminate one a proton with the for-
mation of a double bond (Scheme 4). The progress of the cat-
ionic intermediate in oneway or another would depend on the
speed of each process. This mechanism could explain the
mixture of Z/E isomers17 (54/46) that was obtained from 1.
These compounds could have been originated in a process,
which shared the reaction mechanism with the N-700/N-
Stabilization of the oxocarbenium ion
by addition of the purine ring
NO2
NO2
Cl
OOMe
SnCl4
O
S
N
O OMe
O
S
N
N
N
SnCl4
Cl3Sn
N
N
H
Regioisomer N-9'': 4a [(2)-NO2]
4c [(4)-NO2]
Cl
N
b
a
H
a
N
N
Cl
N
OTMS
OTMS
10
Stabilization of the oxocarbenium ion
by an E1 elimination of a hydrogen atom
Acyclic N-7'' O,N-acetals
4b (2)-NO2
b
HCl
4d (4)-NO2 (not isolated)
NO2
NO2
O
S
N
O OMe
O
S
N
O
OMe
+
OH
OH
E-isomer
Z-isomer
5a [(2)-NO2] 54%
5b [(4)-NO2] 0%
46%
100%
Scheme 4. E1 Mechanism proposed for the formation of olefins from acyclic O,N-acetals.

5278
M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
R1
N
In the case of an E2 process, the purine and one a-hydrogen
are eliminated simultaneously and thus would explain the
exclusive formation of the E isomer observed for 2.18 This
coordination of the Lewis acid to the purine ring, leading
to the elimination of the latter, could also be expected to
happen on the cyclic O,N-acetals. However, no cyclic olefins
have been isolated. This result agrees with the apparent ab-
sence of N-700/N-900 transformations observed in cyclic
compounds.
R1
N
N
N
S
N
N
N
O
O
N
N
O
N
H
11 R1 = -SO2-C6H4-NO2-(2)
12 R1 = -SO2-C6H4-NO2-(4)
13: R1 = -SO2-C6H4-NO2-(2)
14: R1 = -SO2-C6H4-NO2-(4)
15: R1 = -H
2.3. Effect of the substituent R1 on the formation of
O,N-acetals derived from 6-chloropurine
Figure 2. Products obtained by substitution of the chlorine atom in the
6-chloropurine moiety.
The following three consequences can be drawn:
importance of this negative inductive effect (ꢁI) when the
benzoxazepine is bound to the N-700 or the N-900 atom of
the purine moiety is reflected in the chemical shift of H-800.
This hydrogen, which appears more shielded than H-200 in
the 1H NMR spectrum (DMSO-d6) of 6-chloropurine (6-CP),
shifts to lower fields in the N-700 or N-900 benzoxazepinyl
purines. The same effect on H-800 can be observed in 7-
benzoxazepinylhypoxanthines when compared to the single
hypoxanthine ring (H) (Table 3).
(1) Compound 1 led to the best yields in the total formation
of O,N-acetals of 6-chloropurine.
(2) The best selectivity in the formation of cyclic com-
pounds was found for substrate 1 (cyclic/acyclic
ratio¼2.61). This may be the consequence of the stabili-
zation of the cyclic cations (67) by approximation of the
electron rich (2)-nitro groups.
(3) The highest selectivity for the N-700 O,N-acetals in cyclic
compounds was obtained for sulfonamide compounds
(N-700/N-900 ratio¼16 for 1 and N-700/N-900 ratio¼7.5 for
2). The highest selectivity for the N-900 O,N-acetals in
acyclic compounds was obtained for substrate 2.
2.6. Change in DMSO at neutral pH
N-700 Benzoxazepinyl-6-chloropurines with R1¼nitrobenz-
enesulfonyl (3b and 3d) resulted unstable in DMSO-d6 at
rt or 4 ꢀC as they underwent substitution of the chlorine
atom to the hydroxy group. Neither the N-900 benzoxazepinyl
purines with R1¼nitrobenzenesulfonyl (3a, 3c) nor the acy-
clic O,N-acetals of N-900 alkyl 6-chloropurine (4a, 4c, 4e) un-
derwent this transformation. Neither did the non-alkylated
6-chloropurine. No conclusion can be drawn for the acyclic
N-70 alkyl-6-chloropurine O,N-acetal (4b), which suffered
a different kind of transformation under these conditions
(vide infra).
2.4. Effect of the nucleophile in the formation of O,N-
acetals from 1 and 2: comparison between 6-chloro-
purine and the pyrimidines uracil (U) and 5-fluorouracil
(5-FU)
Comparing the results that have been obtained in the reac-
tions between 1 and 2 and 6-chloropurine with those that
were previously reported for the same substrates and the pyri-
midines 5-FU and U,2 it can be observed that the reactivity of
6-chloropurine is similar to the one of 5-FU, both affording
lower total yields of the reaction products than those ob-
tained with U in the same experimental conditions (Table 2).
The experimental conditions used are shown in Table 4. A
higher proportion of water in the solvent mixture hastens
the substitution process. The transformation from 6-chloro-
purine to 6-oxopurine occurs faster when R1¼(2)-nitrobenz-
enesulfonyl (3b/11, entries 6, 7 and 8) than when R1¼
(4)-nitrobenzenesulfonyl (3d/12, entries 2, 3 and 4).
2.5. Substitutions by nucleophiles on 6-chloropurine.
Effect of the N-7 and N-9 purine alkylation on the
stability of the 6-chloropurine moiety
Previous evidence exists of the need to activate the 6-chloro-
purine ring to perform effective substitution of chlorine with
nucleophiles.23 We propose a mechanism to explain the
substitution process in DMSO at neutral pH after having ver-
ified the stability of these compounds in other solvents such
as CH2Cl2 and CDCl3. This is based on the formation of in-
termediate aryloxysulfonium salts, which would be formed
by the displacement of the chlorine atom by a DMSO mol-
ecule, as indicated in Scheme 5. Even weak nucleophiles
such as water could attack these salts at the carbon 600 of
the purine ring or at the sulfur atom bearing formal positive
charge. Both possibilities have been previously described in
studies on alcoxysulfonium salts hydrolysis.24 Despite exist-
ing evidence of stable alcoxysulfonium salts that can be iso-
lated, in most cases they have been reported to be unstable
and their hydrolysis can be performed by water traces exist-
ing in solvents.24 Water molecules would thus act as an
oxygen source for oxidation and DMSO would act as a cata-
lyst that is regenerated in the process. This would explain the
Chlorine substitution by nucleophiles takes place more
easily in 7-(benzoxazepin-3-yl)purines than in 9-(benzox-
azepin-3-yl)purines. The chloropurine hydrolysis19,20 and
the chlorine substitution by a thiol group21,22 from chloro-
purine were previously reported. The substitution products
shown in Figure 2 were obtained.
The highest electrophilic character of the C-600 atom in 7-
substituted purines was explained as a consequence of the
inductive effect exerted by the benzoxazepine ring. The
Table 2. Comparison of the reactivity shown by chloropurine versus the
previously reported one for pyrimidines
6-Chloropurine (%)
Ua (%)
5-FUa (%)
Total yields from 1
Total yields from 2
47
39
72
73
53
44
a
See Ref. 2.

M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
5279
Table 3. Chemical shifts (ppm) for H-800 and H-200 in 6-chloropurine (6-CP), hypoxanthine (H) and derived O,N-acetals (1H NMR, DMSO-d6, 300 MHz, rt)
6-CPa
3c
3d
3ab
3b
4c
4cc
4a
4ac
Ha
12
11
H-200
H-800
8.80
8.73
8.90
8.95
8.95
9.04
8.81
8.89
8.86
9.00
8.77
8.90
8.72
8.85
8.90
8.99
8.86
8.99
8.03
8.17
8.02
8.45
8.04
8.55
Effect of the alkylation on N-900 and N-700 products.
NMR 400 MHz.
a
b
It refers to the minor conformer.
See Ref. 18.
c
Table 4. Experimental conditions for 6-CP chlorine substitution in DMSO at neutral pH
Entry
Substrate (concentration)
Product (yield, %)
Reagentsc,d
Time
Temperature
1
2
3
4
5
6
7
8
3c (27 mM)
3d (27 mM)
3d (10 mM)
3d (10 mM)
3a (27 mM)
3b (27 mM)
3b (10 mM)
3b (10 mM)
3e (27 mM)
3f (27 mM)
4a (38 mM)
4b (38 mM)
4c (38 mM)
4d (12 mM)
6-CP (10 mM)
Unchanged, 100a
14, 100a
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (0.71 M)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (0.71 M)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (16 mM)
DMSO-d6, H2O (0.71 M)
30 days
30 days
6 days
6 days
30 days
7 days
5 days
44 h
10 min
10 min
30 days
30 days
30 days
10 min
7 days
4 ꢀC
4 ꢀC
rt
Unchanged, not determinedb
Unchanged, not determinedb
Unchanged, 100a
13, 100a
rt
4 ꢀC
4 ꢀC
rt
13, not determinedb
13, not determinedb
Unchanged, 100a
Unchanged, 100a
Unchanged, 100
1, >90
Unchanged, 100
Unchanged, 100
Unchanged, not determinedb
rt
9
60 ꢀC
60 ꢀC
4 ꢀC
4 ꢀC
4 ꢀC
95 ꢀC
rt
10
11
12
13
14
15
Quantitative control by 1H NMR.
Qualitative control by TLC: in entries 7 and 8, the time indicates the moment when the spot corresponding to the substitution product becomes visible to the
a
b
naked eye.
A minimum concentration of DMSO (5% v/v) was always added due to solubility reasons.
No substitution took place in CDCl3 at 4 ꢀC or rt.
c
d
H
H
O
Cl
O
DMSO
HCl
S
N
S
O
S
O
O
N
OH
N
N
N
N
N
N
N
HN
N
N
Aryloxysulfonium chloride
6-Oxopurine
derivative
S
O
Cl
N
Cl
N
N
N
N
DMSO
HCl
N
N
N
H
H
O
6-Chloropurine
derivative
Cl
S
H
O
Cl
S
N
O
N
O
N
H
N
N
N
N
N
Aryloxysulfonium chloride
Scheme 5. DMSO-mediated substitution of chlorine.
hastening of the substitution process with the increase of the
proportion of water in the solvent mixture.
2.7. Change Cl/SPh (from 6-chloropurines to
6-phenylthiopurines 13–15)
The acyclic O,N-acetal 4b (Table 5, entry 12) underwent vir-
tually quantitative transformation to the initial compound 1,
after remaining in DMSO-d6 (0.03% H2O/D2O) at 4 ꢀC for
one month. The resonance signals corresponding to 1 can
be observed by 1H NMR spectroscopy one week after disso-
lution of 4b. The formation of 1 by attack of the benzyl hy-
droxy group on the acetalic position would be favoured in
the polar solvent DMSO and could be aided by protonation
of the purine ring after releasing hydrochloric acid during
the formation of 6-oxopurine (Scheme 6).
N-700 Alkyl-6-chloropurines 3b and 3d underwent sub-
stitution of chlorine after treatment with thiophenol and
Table 5. Results of the treatment of 3b and 3d with PhSH/K2CO3/40 min/rt
Substrate
Product (yield, %)
Equiv of PhSH
3b
3d
15 (89)
14 (36)
15 (37)
13 (61)
2.0
0.8
3b
0.8

5280
M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
NO2
O OMe
NO2
O OMe
NO2
O
S
N
O
S
N
O
S
N
O
N
N
N
N
X
N
X
N
OMe
N
N
O
OH
OH
4b
1
Scheme 6. Cyclization of 4b in DMSO.
potassium carbonate in DMF at rt.7 This process was fav-
oured against the removal of the nitrobenzenesulfonyl
group, which was performed in the same conditions.
X-ray diffraction on 3b crystals confirmed the N-700 aminalic
bond (Fig. 4) thus ratifying the spectroscopic identification
of the N-700/N-900 regioisomers.
Table 5 shows the results obtained when 3b and 3d were
treated with different amounts of the mentioned reagents.
A selective substitution of the chlorine atom versus the
(2)-nitrobenzenesulfonyl removal can be observed when
3b is treated with 0.8 equiv of thiophenol. However, both
processes compete when (4)-nitrobenzenesulfonyl group is
at position 1. The easier removal of the (4)-nitrobenzene-
sulfonyl group than the (2)-nitrobenzenesulfonyl one could
obey to steric reasons as the mechanism described by
Fukuyama et al.25 proposes an intermediate Meisenheimer
complex, as a consequence of the thiophenol attack at the
adjacent position to the sulfonyl function.
As was observed for pyrimidine derivatives,2 the 1H and 13
C
NMR spectra of the acyclic purine O,N-acetals with sulfon-
amide functions (4a, 4b and 4c) appear as a mixture of two
structures with the following distribution: 4a, 53/47; 4b, 50/
50; 4c, 57/43. The N-1-tertiary sulfonamide adopts certain
orientations, nonconvertible between each other, in such
a way that two conformers may be observed. In all the acy-
clic O,N-acetals the benzylic and the R1N–CH2– hydrogen
atoms appear as diastereotopic.
2.9. X-ray diffraction analysis on 3b
The crystal structure of 3b (Fig. 4) revealed the following
features: (a) The seven-membered ring of benzoxazepine
Due to the nucleophile strength, the chlorine substitution by
thiophenol needed no further activation of the purine ring.
2.8. Spectroscopic analysis of purine O,N-acetals
The complete unambiguous assignment of the proton and
carbon atoms was confirmed by the heteronuclear multi-
quantum correlations HMQC (on 3a, 3b, 3d, 4a, 4c, 11
and 12) and heteronuclear multi-bond correlations HMBC
(on 3a–c), in DMSO-d6.
Compounds containing N-700 alkyl purines have been distin-
guished from those with N-900 alkyl purines from different in-
teractions that the aminalic hydrogen (see Fig. 3) establishes
with the purine ring in each case. N-700 Alkyl purines pre-
sented a three-bond interaction between the aminalic hydro-
gen atoms and the quaternary C-500 atoms. N-900 Alkyl
purines presented a three-bond interaction between the
aminalic hydrogen atoms and the quaternary C-400 atoms.
Figure 4. ORTEP drawing of 3b at 50% probability.
R1
R1
N
9
6
2
8'
'
'
8'
'
'
9'
'
9a N
1
7'
'
3
8
7
3
N
N
N
N
4'
N
4'
'
3''
'
5a
X
O
O
5'
N
5
4'
NO2
X
N
6'
2'
'
3'
2'
N1'
5'
'
6'
R1
=
OMe
R1 OMe
R1
N
6
3
8'
'
8'
'
7'
'
1
N
O
S1'
O
5
4
N
N
N
N 9''
N
5''
4'
N
'
2
4''
N 3''
X
X
6''
OH
N
OH
1'' 2''
N-9 Alkyl purine derivatives
N-7 Alkyl purine derivatives
Figure 3. Numbering of cyclic and acyclic purine O,N-acetals.

M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
5281
adopts a chair-like conformation, (b) the flat ring of purine
lies in an equatorial orientation with its six-membered ring
orientated towards face a, (c) the 2-nitrobenzenesulfonyl
moiety situates in axial disposition, towards the face a of
the molecule with its benzene ring tilted to the benzoxaze-
pine ring and (d) the H-2b atom is axial while H-2a is
equatorial.
groups of the molecule, the 6-chloropurine moiety is the
most suitable to be involved in p/p-stacking interactions.
3. Conclusions
The O,O-acetalic compounds (RS)-3-methoxy-1-(2-nitro-
benzenesulfonyl)-1,2,3,5-tetrahydro-4,1-benzoxazepine, 1,
and (RS)-3-methoxy-1-(4-nitrobenzene sulfonyl)-1,2,3,5-
tetrahydro-4,1-benzoxazepine, 2, have been studied in the
Lewis acid-mediated condensation with 6-chloropurine.
When (RS)-3-methoxy-4,1-benzoxazepine O,O-acetals 1
and 2 are condensed with 6-chloropurine in the presence
of tin(IV) chloride in acetonitrile, different ratios of the final
cyclic and acyclic O,N-acetals are obtained depending on the
starting compound. The presence of the starting compounds
of electron-rich groups that can approximate to, and there-
fore stabilize, the cationic position of the cyclic intermediate
carbenium ions [e.g., (2)-NO2 in 1] favours the formation of
the final cyclic O,N-acetals. 6-Chloropurine leads to an N-700
aminalic bond in the cyclic products and mainly to the N-900
aminalic bond in the acyclic ones. Substitution of the chlo-
rine atom at the 600 position of purine is more feasible
when the ring is alkylated at the N-700 position than at the
N-900 one. Exchange with a hydroxyl group is performed
with water traces in deuterated dimethylsulfoxide at rt in
a solvent-mediated process. Furthermore, this process takes
place in acetonitrile at 50 ꢀC in the presence of a Lewis acid.
The exchange with strong nucleophiles (e.g., thiophenol)
does not need further activation. Chlorine substitution has
been observed in cyclic O,N-acetals with the N-700 aminalic
bond. Coordination of the Lewis acid to the nitrogen bases
could cause N-700/N-900 isomerization, formation of olefins
5a and 5b and a Cl/OH exchange.
This conformation seems to be the most favourable to enable
the building of a 3D-net with the relevant contribution of p/
p-stacking and CH–p interactions, involving the purine
moiety. The p/p interaction stacks two purine moieties in
a nearly antiparallel fashion (dihedral angle between mean
planes a¼0.46ꢀ, slipping angles of approximately 21ꢀ and
˚
an interplanar distance of 3.28 A). This interaction builds
up pairs of molecules in the crystals (Fig. 5a).
In addition (Fig. 5b), the six-membered ring of each purine
(with centroid hereafter named as Cg) acts as acceptor of
a C–H/p interaction with the C-28–H-28 bond (H donor)
˚
of the nitrobenzene ring. The distance H-28–Cg is 2.95 A
and the angle C-28–H-28–Cg is 119ꢀ. Thus each purine
ring is involved in a p/p-stacking interaction by one side
and in a C–H/p interaction by the other one. The crystal
structure of 3b suggests that among the three aromatic
4. Experimental
4.1. General
The general methods were the same as those previously de-
scribed.1,2 The HMBC spectra26 were measured using
a pulse sequence optimized for 10 Hz (inter-pulse delay
for the evolution of long-range couplings: 50 ms) and
a 5:3:4 gradient combination. In this way, direct responses
(1J couplings) were not completely removed. The HMQC
spectra27 (inv4gs in the standard Bruker software) resulted
from 256ꢂ1024 data matrix size with 16–64 scans per t1 de-
pending on the sample concentration and inter-pulse delay of
3.2 ms and a 5:3:4 gradient combination. For numbering of
compounds, see Figure 3.
4.2. General procedure for the preparation of 6-chloro-
purine O,N-acetals from O,O-acetalic compounds
A suspension of O,O-acetalic compound (1.0 equiv) and
6-chloropurine (2.5 equiv) in anhydrous MeCN (5 mL/
mmol) was prepared under argon and cooled to 0 ꢀC. At
this temperature, TMSCl (4.0 equiv), HMDS (4.0 equiv)
and SnCl4 (commercial solution 1.0 M in CH2Cl2) were
added subsequently. The temperature was allowed to rise
to 10 ꢀC before heating up to 50 ꢀC under argon for more
than 48 h. The reactions were quenched by cooling
Figure 5. (a) Antiparallel p/p-stacking interaction between the 6-chloro-
purine moieties of a pair of molecules in the crystal of 3b (dashed lines
link the centroids of the rings). (b) Additional contribution of C–H/p inter-
actions (dotted lines) cooperates with the p/p-stacking to build up the 3D
framework molecular crystal of 3b.

5282
M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
(ice/water bath) and the addition of distilled water. The pH
was fixed to 7–8 with NaHCO3 (10%) and the aqueous
phases were extracted with EtOAc and CH2Cl2. The organic
layers were dried (Na2SO4), filtered and evaporated.
1.2/10 (CH3)2CO/CH2Cl2; Rf (EtOAc/hexane 3/1): 0.24;
white solid; mp 220.0–221.0 ꢀC; yield 32%. 1H NMR
(DMSO-d6, 300 MHz): d (ppm) 9.00 (s, 1H, H-800), 8.86
(s, 1H, H-200), 8.12–8.08 (m, 2H, H-30, H-600), 8.00 (ddd,
0
0
0
0
0
0
J4 –3 ¼J4 –5 ¼7.7 Hz, J4 –6 ¼1.5 Hz, 1H, H-4 ), 7.91 (ddd,
4.2.1. Reaction of 11 with 6-chloropurine: 3a, 3b, 4a, 4b,
5a. The general method was followed and the reaction was
kept at 50 ꢀC for 69 h.
J5 –4 ¼J5 –6 ¼7.7 Hz, J5 –3 ¼1.4 Hz, 1H, H-5 ), 7.53 (m,
1H, H-6), 7.44–7.38 (m, 2H, H-8, H-8), 7.19 (m, 1H, H-9),
6.34 (dd, J3b–2a¼10.0 Hz, J3b–2b¼1.8 Hz, 1H, H-3b),
4.92 (d, Jgem benzylic H¼13.8 Hz, 1H, benzylic H), 4.83 (dd,
0
0
0
0
0
0
0
4.2.1.1. N-(2-Hydroxymethylphenyl)-N-(2-methoxy-
vinyl)-2-nitrobenzenesulfonamide 5a. Elution by flash
chromatography, 1/2 EtOAc/hexane or CH2Cl2; Rf
(MeOH/CH2Cl2 0.5/10): 0.56; yellow solid; yield 2% [ace-
tone-d6, rt, isomer Z (54%), isomer E (46%)]. 1H NMR (ace-
tone-d6, 300 MHz): d (ppm) 7.98–7.75 (m, 4H, 4H), 7.70
(dd, J1¼7.8 Hz, J2¼not det., 1H), 7.63 (dd, J1¼7.7 Hz,
J2¼not det., 1H), 7.43 (ddd, J1¼J2¼7.6 Hz, J3¼1.3 Hz,
1H or 1H), 7.36 (ddd, J1¼J2¼7.5 Hz, J3¼1.3 Hz, 1H or
1H), 7.22 (ddd, J1¼J2¼7.6 Hz, J3¼1.6 Hz, 1H or 1H),
7.16 (ddd, J1¼J2¼7.7 Hz, J3¼1.7 Hz, 1H or 1H), 7.03 (dd,
J1¼7.9 Hz, J2¼1.3 Hz, 1H), 6.90 (dd, J1¼7.9 Hz,
Jgem
¼15.0 Hz, J2b–3b¼1.9 Hz, 1H, H-2b), 4.74 (d,
2b–2a
Jgem benzylic H¼13.8 Hz, 1H, benzylic H), 4.27 (dd, Jgem 2a–2b
¼
15.0 Hz, J2a–3b¼10.1 Hz, 1H, H-2a). 13C NMR (DMSO-d6,
75 MHz): d (ppm) 161.4 (C-400), 152.2 (CH-200), 148.0 (CH-
800), 147.2 (C-20), 142.3 (C-600), 138.5 (C-10a), 137.2 (C-5a),
135.5 (CH-40), 133.2 (CH-50), 132.0 (C-10), 130.3 and 130.1
(CH-60, CH-6), 129.5 (CH-8), 128.8 (CH-7), 128.1 (CH-9),
125.1 (CH-30), 121.4 (C-500), 84.5 (CH-3), 69.6 (benzylic
CH2), 53.1 (CH2-2) (proton and carbon atom shifts, and
the nature of the N-900 aminalic bond were confirmed by
HMBC and HMQC studies in DMSO-d6). HR LSIMS
calcd for C20H15ClN6O5SNa (M+Na)+ 509.0411, found
509.0411. Anal. Calcd for C20H15ClN6O5S: C, 49.34; H,
3.11; N, 17.26; S, 6.59. Found: C, 49.69; H, 3.27; N,
17.50; S, 6.39.
J2¼1.2 Hz, 1H), 6.36 (d, JNCH CH ¼ 11:3 Hz, 1H, NCH2 or
2
aminalic CH), 6.22 (d, JNCH CH ¼ 11:4 Hz, 1H, NCH2 or
2
aminalic CH), 5.95 (d, JNCH CH ¼ 4:5 Hz, 1H, NCH2 or ami-
2
nalic CH), 5.61 (d, JNCH CH ¼ 4:5 Hz, 1H, NCH2 or aminalic
2
CH), 4.76 and 4.71 (2s, 2 benzylic H atoms, 2 benzylic H
atoms), 3.58 (s, 3H, Me), 3.43 (s, 3H, Me). HR LSIMS calcd
for C16H16N2O6SNa (M+Na)+ 387.0627, found 387.0627.
Anal. Calcd for C16H16N2O6S: C, 52.74; H, 4.43; N, 7.69;
S, 8.80. Found: C, 52.84; H, 4.76; N, 7.47; S, 8.57.
4.2.1.4.
(RS)-6-Chloro-9-{2-[N-(2-hydroxymethyl-
phenyl)-2-nitrobenzenesulfonamide]-1-methoxyethyl}-
9H-purine 4a. Elution by flash chromatography, 3/1 EtOAc/
hexane or 1.3/10 (CH3)2CO/CH2Cl2; Rf (MeOH/CH2Cl2 0.5/
10): 0.53; white solid; mp 89.3–90.0 ꢀC; yield 7% [DMSO-
d6, rt, isomer A (53%), isomer B (47%)]. 1H NMR (DMSO-
d6, 300 MHz): d (ppm) 8.99 (2s, 2H, H-800, H-800), 8.90 (s,
1H, H-200 or H-200), 8.86 (s, 1H, H-200 or H-200), 8.13–8.02
(m, 4H, H-40, H-40, 2H-30 or 2H-50), 7.93–7.84 (m, 2H,
4.2.1.2. (RS)-6-Chloro-9-[1-(2-nitrobenzenesulfonyl)-
1,2,3,5-tetrahydro-4,1-benzoxazepin-3-yl]-9H-purine 3a.
Elution by flash chromatography, 1/1 EtOAc/hexane or
CH2Cl2; Rf (MeOH/CH2Cl2 0.5/10): 0.63; white solid; mp
0
0
2H-30 or 2H-50), 7.77 (d, J6 –5 ¼7.2 Hz, 1H, H-6 or H-6 ),
7.69–7.61 (m, 3H, H-60 or H-60, H-6, H-6), 7.55–7.45 (m,
2H, H-5, H-5), 7.33 (ddd, J4–3¼J4–5¼7.6 Hz, J4–6¼1.3 Hz,
1H, H-4), 7.20 (ddd, J4–3¼J4–5¼7.6 Hz, J4–6¼1.2 Hz, 1H,
0
0
1
123–125 ꢀC; yield 2%. H NMR (DMSO-d6, 400 MHz):
d (ppm) 8.89 (s, 1H, H-800), 8.81 (s, 01H, H-200), 8.12 (dd,
0
0
0
0
J3 –4 ¼7.9 Hz, J3 –5 ¼0.6 Hz, 1H, H-3 ), 8.08 (d, J¼7.2 Hz,
1H, H-60), 8.01 (ddd, J4 –3 ¼J4 –5 ¼7.7 Hz, J4 –6 ¼0.9 Hz,
H-4), 7.11 (d, J3–4¼7.7 Hz, 1H, H-3), 6.81 (d, J3–4
¼
0
0
0
0
0
0
1H, H-40), 7.92 (t, J5 –4 ¼J5 –6 ¼7.7 Hz, 1H, H-5 ), 7.55
7.5 Hz, 1H, H-3), 5.98 (t, JNCH CH ¼ 6:2 Hz, 1H, H-9),
0
0
0
0
0
2
(dd, J7–8¼6.8 Hz, J7–9¼2.0 Hz, 1H, H-6), 7.42 (m, 1H,
5.89 (t, JNCH CH ¼ 6:0 Hz, 1H, aminalic H), 5.25 (bb),
2
H-7), 7.40–7.38 (m, 1H, H-8), 7.08 (dd, J9–8¼7.1 Hz, J9–7
¼
4.78–4.69 (m, 2H, H-8, NCH2), 4.63–4.46 (m, 4H, H-7,
benzylic H, H-8, NCH2), 4.33 (d, Jgem benzylic H¼14.0 Hz,
1.7 Hz, 1H, H-9), 6.18 (dd, J3b–2a¼10.0 Hz, J3b–2b¼1.5 Hz,
1H, H-3b), 4.94 (d, Jgem benzylic H¼13.8 Hz, 1H, benzylic H),
4.86 (d, Jgem benzylic H¼13.7 Hz, 1H, benzylic H), 4.61 (dd,
Jgem 2b–2a¼15.0 Hz, J2b–3b¼1.6 Hz, 1H, H-2b), 4.33 (dd,
Jgem 2a–2b¼15.1 Hz, J2a–3b¼10.2 Hz, 1H, H-2a). 13C NMR
(DMSO-d6, 100 MHz): d (ppm) 152.0 (CH-200), 151.3 (C-
400), 149.5 (C-600), 147.3 (C-20), 145.6 (CH-800), 138.7 (C-
10a), 137.5 (C-5a), 135.6 (CH-40), 133.2 (CH-50), 132.0
(C-10), 130.9 (C-500), 130.5 (CH-60), 130.3 (CH-6), 129.6
(CH-8), 128.8 (CH-7), 127.7 (CH-9), 125.2 (CH-30), 84.0
(CH-3), 70.1 (benzylic CH2), 53.2 (CH2-2) (proton and car-
bon atom shifts, and the nature of the N-900 aminalic bond
were confirmed by HMBC and HMQC studies in DMSO-
d6). HR LSIMS calcd for C20H15ClN6O5SNa (M+Na)+
509.0411, found 509.0412. Anal. Calcd for C20H15ClN6O5S:
C, 49.34; H, 3.11; N, 17.26; S, 6.59. Found: C, 49.64; H,
3.06; N, 17.43; S, 6.67.
1H, benzylic H or benzylic H), 4.25 (d, Jgem benzylic H
¼
14.1 Hz, 1H, benzylic H or benzylic H), 3.33 (s, 6H, H-
10, Me). 13C NMR (DMSO-d6, 75 MHz): d (ppm) 151.8
(C-400, C-400), 151.7 (CH-200, CH-200), 149.3 (C-600, C-600),
147.4 (C-20, C-20), 145.9 and 145.8 (CH-800, CH-800),
142.4 (C-1), 141.9 (C-1), 135.3 (CH-40, CH-40), 134.5 (C-
2), 134.2 (C-2), 132.1 and 132.0 (2CH-30 or 2CH-50),
131.2 and 131.1 (C-500, C-500), 130.7 (CH-60, CH-60),
129.6 and 129.2 (C-10, C-10), 129.4 (CH-3), 129.1 and
129.0 (CH-3, CH-5, CH-5), 128.6 (CH-6), 128.4 (CH-
6), 127.5 (CH-4), 127.2 (CH-4), 124.1 (2CH-30 or 2CH-
50), 85.1 (aminalic CH), 84.2 (aminalic CH), 58.3 (CH2-7,
benzylic CH2), 56.2 (OMe, OMe), 53.8 and 53.2 (CH2-8,
NCH2) (proton and carbon atom shifts, and the nature of
the N-900 aminalic bond were confirmed by HMBC and
HMQC studies in DMSO-d6). HR LSIMS calcd for
C21H19ClN6O6SNa (M+Na)+ 541.0673, found 541.0673.
Anal. Calcd for C21H19ClN6O6S: C, 48.60; H, 3.69;
N, 16.19; S, 6.18. Found: C, 48.75; H, 3.67; N, 16.00; S,
6.03.
4.2.1.3. (RS)-6-Chloro-7-[1-(2-nitrobenzenesulfonyl)-
1,2,3,5-tetrahydro-4,1-benzoxazepin-3-yl]-7H-purine 3b.
Elution by flash chromatography, 3/1 EtOAc/hexane or

M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
5283
4.2.1.5.
(RS)-6-Chloro-7-{2-[N-(2-hydroxymethyl-
the nature of the N-900 aminalic bond were confirmed by
HMBC and HMQC studies in DMSO-d6). HR LSIMS calcd
for C20H15ClN6O5SNa (M+Na)+ 509.0411, found 509.0412.
Anal. Calcd for C20H15ClN6O5S: C, 49.34; H, 3.11; N,
17.26; S, 6.59. Found: C, 49.09; H, 3.17; N, 17.38; S, 6.69.
phenyl)-2-nitrobenzenesulfonamide]-1-methoxyethyl}-
7H-purine 4b. Elution by flash chromatography, 1/0 EtOAc/
hexane or 1/3 (CH3)2CO/CH2Cl2; Rf (EtOAc/hexane 0.5/
10): 0.48; white solid; mp 88.0–89.0 ꢀC; yield 6% [CDCl3,
1
rt, isomer A (50%), isomer B (50%)]. H NMR (CDCl3,
300 MHz): d (ppm) 8.83 and 8.80 (2s, 2H), 8.52 and 8.47
(2s, 2H), 7.73–7.55 (m, 6H), 7.51–7.34 (m, 6H), 7.17 (ddd,
J1¼J2¼7.7 Hz, J3¼1.7 Hz, 1H), 7.13 (ddd, J1¼J2¼7.7 Hz,
J3¼1.6 Hz, 1H), 6.84 (pt, J¼7.3 Hz, 2H), 6.25 (dd,
JNCH CH ¼ 7:3 Hz, JNCH CH ¼ 4:9 Hz, 1H, aminalic H),
4.2.2.3. (RS)-6-Chloro-7-[1-(4-nitrobenzenesulfonyl)-
1,2,3,5-tetrahydro-4,1-benzoxazepin-3-yl]-7H-purine 3d.
Elution by flash chromatography, 1.5/1 EtOAc/hexane; Rf
(EtOAc/hexane 2/1): 0.35; white solid; mp 219.0–220.0 ꢀC;
1
yield 15%. H NMR (DMSO-d6, 300 MHz): d (ppm) 9.04
0
2
2
0
0
0
6.03 (dd, JNCH CH ¼ 8:4 Hz, JNCH CH ¼ 3:6 Hz, 1H, ami-
(s, 1H, H-800), 8.95 (s, 1H, H-200), 8.52 (d, J3 –2 ¼J5 –6
¼
2
2
8.9 Hz, 2H, H-30,50), 8.26 (d, J2 –3 ¼J6 –5 ¼8.9 Hz, 2H,
H-20,60), 7.58–7.45 (m, 3H, H-6, H-7, H-8), 7.38 (dd,
J1¼7.4 Hz, J2¼1.4 Hz, 1H, H-9), 6.42 (dd, J3b–2a¼10.0 Hz,
J3b–2b¼1.7 Hz, 1H, H-3b), 5.00 (dd, Jgem 2b–2a¼14.9 Hz,
J2b–3b¼1.9 Hz, 1H, H-2b), 4.93 (d, Jgem benzylic H¼14.0 Hz,
1H, benzylic H), 4.61 (d, Jgem benzylic H¼13.9 Hz, 1H, benz-
ylic H), 4.23 (dd, Jgem 2b–2a¼14.9 Hz, J2a–3b¼10.1 Hz, 1H,
H-2a). 13C NMR (DMSO-d6, 75 MHz): d (ppm) 161.4
(C-400), 152.2 (CH-200), 150.2 (C-40), 148.0 (CH-800), 145.4
(C-10), 142.4 (C-600), 138.7 (C-10a), 137.0 (C-6), 130.0
(CH-6), 129.6 (CH-8), 128.6 (CH-20, CH-60), 127.7 (CH-9),
125.1 (CH-30, CH-50), 121.5 (C-500), 84.2 (CH-3), 69.3
(CH2-5), 52.6 (CH2-2) (proton and carbon atom shifts, and
the nature of the N-700 aminalic bond were confirmed by
HMBC and HMQC studies in DMSO-d6). HR LSIMS calcd
for C20H15ClN6O5SNa (M+Na)+ 509.0411, found 509.0410.
Anal. Calcd for C20H15ClN6O5S: C, 49.34; H, 3.11; N,
17.26; S, 6.59. Found: C, 49.46; H, 3.12; N, 17.56, S, 6.40.
0
0
0
0
nalic H), 4.76 and 4.61 (2dsystem AB, Jgem benzylic H, part A
12.5 Hz, Jgem benzylic H, part B¼12.4 Hz, 2H, 2H benzylic
¼
or 2H benzylic), 4.61 (dd, Jgem NCH ¼ 14:7 Hz,
2
JNCH CH ¼ 8:3 Hz, 1H, NCH2), 4.51 and 4.46 (2dsystem AB
,
2
Jgem benzylic H, part A¼12.1 Hz, 2H, H-7 or benzylic H), 4.46
(dd, JCHCH N ¼ 4:8 Hz, 1H, NCH2), 4.28 (dd,
2
Jgem NCH ¼ 15:0 Hz, JNCH CH ¼ 7:6 Hz, 1H, NCH2), 4.01
2
2
(dd, Jgem NCH ¼ 14:6 Hz, JNCH CH ¼ 3:6 Hz, 1H, NCH2),
2
2
3.44 (s, 3H, Me or Me), 3.37 (s, 3H, Me or Me). HR LSIMS
calcd for C21H19ClN6O6SNa (M+Na)+ 541.0673, found
541.0674. Anal. Calcd for C21H19N6O6SCl$0$26H2O: C,
48.17; H, 3.76; N, 16.05; S, 6.12. Found: C, 48.31; H,
3.42; N, 15.65; S, 5.74.
4.2.2. Reaction of 21 with 6-chloropurine: 3c, 3d, 4c, 5b.
The general method was followed and the reaction was
maintained at 50 ꢀC for 67 h.
4.2.2.1. N-(2-Hydroxymethylphenyl)-N-(2-methoxy-
vinyl)-4-nitrobenzenesulfonamide 5b. Elution by flash
chromatography, 1/1.5 EtOAc/hexane or 0.15/10
(CH3)2CO/CH2Cl2; Rf (EtOAc/hexane 2/1): 0.61; yellow
4.2.2.4.
(RS)-6-Chloro-9-{2-[N-(2-hydroxymethyl-
phenyl)-4-nitrobenzenesulfonamide]-1-methoxyethyl}-
9H-purine 4c. Elution by flash chromatography, 1.5/1
EtOAc/hexane; Rf (EtOAc/hexane 2/1): 0.24; white solid;
mp 135.4–136.4 ꢀC; yield 22% [DMSO-d6, rt, isomer A
1
liquid; yield 7%. H NMR (CDCl3, 300 MHz): d (ppm)
0
0
0
0
0
0
0
0
8.39 (d, J3 –2 ¼J5 –6 ¼80.0 0Hz, 2H, H-3 ,5 ), 7.92 (d, J2 –3
(m, 1H), 7.17–7.13 (m, 1H), 6.42–6.38 (m, 1H), 6.35 (d,
¼
1
0
0
J6 –5 ¼8.0 Hz, 2H, H-2 ,6 ), 7.68–7.64 (m, 1H), 7.44–7.40
(57%), isomer B (43%)]. H NMR (DMSO-d6, 300 MHz):
d (ppm) 8.90 (s, 1H, H-800), 8.85 (s, 1H, H-800), 8.77 (s, 1H,
H-200), 8.72 (s, 1H, H-200), 8.41 (d, J3 –2 ¼J5 –6
¼
0
0
0
0
JNCH CHw12 Hz, 1H, NCH2 or aminalic H), 6.04 (d, J8–9
2
8.9 Hz, 2H, H-30,50), 8.37 (d, J3 –2 ¼J5 –6 ¼8.9 Hz, 2H, H-
0
0
0
0
w 12 Hz, 1H, NCH2 or aminalic H), 4.75 (s, 2H, benzylic
H), 3.55 (s, 3H, Me). HR LSIMS calcd for C16H16N2O6SNa
(M+Na)+ 387.0627, found 387.0627. Anal. Calcd for
C16H16N2O6S: C, 52.74; H, 4.43; N, 7.69; S, 8.80. Found:
C, 52.88; H, 4.54; N, 7.71; S, 8.54.
30,50), 7.91 (d, J2 –3 ¼J6 –5 ¼8.9 Hz, 2H, H-2 ,6 ), 7.84 (d,
0
0
0
0
0
0
0
0
0
0
0
0
J2 –3 ¼J6 –5 ¼8.9 Hz, 2H, H-2 ,6 ), 7.59 (d, J6–5¼7.6 Hz,
1H, H-6), 7.54 (d, J6–5¼6.8 Hz, 1H, H-6), 7.42–7.34 (m,
2H, H-5, H-5), 7.15 (dt, J1¼7.7 Hz, J2¼1.3 Hz, 2H, H-4,
H-4), 6.72 (d, J3–4¼7.3 Hz, 1H, H-3), 6.56 (d, J3–4
¼
4.2.2.2. (RS)-6-Chloro-9-[1-(4-nitrobenzenesulfonyl)-
1,2,3,5-tetrahydro-4,1-benzoxazepin-3-yl]-9H-purine 3c.
Elution by flash chromatography, 1/2 EtOAc/hexane or 2/
10 (CH3)2CO/CH2Cl2; Rf (EtOAc/hexane 2/1): 0.51; light
7.2 Hz, 1H, H-3), 5.79 (t, JNCH CH ¼ 6:2 Hz, 1H, aminalic
2
H), 5.69 (t, JNCH CH ¼ 6:1 Hz, 1H, aminalic H), 5.15 (m),
2
4.58 (dd, Jgem NCH ¼ 14:3 Hz, JNCH CH ¼ 6:7 Hz, 1H,
2
2
NCH2 or NCH2), 4.47–4.39 [m, 6H, 2H-7, 2 benzylic H,
NCH2 (NCH2 and/or NCH2)], 4.22 (dd, Jgem NCH
1
yellow solid; yield 2%. H NMR (DMSO-d6, 300 MHz):
d (ppm) 8.95 (s, 1H, H-800), 8.90 (s, 1H, H-200), 8.52 (d,
¼
2
14:3 Hz, JNCH CH ¼ 5:5 Hz, 1H, NCH2 or NCH2), 3.16 (s,
2
J3 –2 ¼J5 –6 ¼9.0 Hz, 2H, H-3 ,5 ), 8.27 (d, J2 –3 ¼J6 –5
¼
3H, Me), 3.12 (s, 3H, Me). 13C NMR (DMSO-d6, 75 MHz):
d (ppm) 151.9 and 151.9–151.7 (C-400, C-400), 151.8 and
151.7 (CH-200, CH-200), 150.1 and 150.0 (C-40, C-40), 149.3
and 148.8 (C-600, C-600), 146.0 and 145.7 (CH-800, CH-800),
142.8, 142.8, 142.7 and 142.5 (C-10, C-10, C-1, C-1), 135.7
and 134.9 (C-2, C-2), 132.0 and 131.0 (C-500, C-500), 129.4,
129.1 and 128.8 (CH-20, CH-60, CH-20, CH-60, CH-5, CH-
5), 128.2 (CH-6, CH-6), 128.0 and 127.8 (CH-3, CH-3),
127.4 (CH-4, CH-4), 124.5 and 124.5 (CH-30, CH-30, CH-
50, CH-50), 84.8 and 83.9 (aminalic CH, aminalic CH-9),
58.6 (benzylic CH2, benzylic CH2), 56.1 (Me, Me), 53.3
(CH2-8, NCH2) (proton and carbon atom shifts, and the
0
0
0
0
0
0
0
0
0
0
9.0 Hz, 2H, H-20,60), 7.60–7.28 (m, 4H, H-6, H-7, H-8, H-
9), 6.26 (dd, J3b–2a¼10.2 Hz, J3b–2b¼1.8 Hz, 1H, H-3b),
4.96 (d, Jgem benzylic H¼12.0 Hz, 1H, benzylic H), 4.83–
4.73 (m, 2H, H-2b, benzylic H), 4.27 (dd, Jgem
¼
2b–2a
15.0 Hz, J2a–3b¼9.9 Hz, 1H, H-2a). 13C NMR (DMSO-d6,
75 MHz): d (ppm) 152.0 (CH-200), 151.2 (C-400), 150.2 (C-
40), 149.4 (C-600), 145.5 (CH-800), 145.4 (C-10), 139.0 (C-
10a), 136.9 (C-6), 131.0 (C-500), 130.1 (CH-6), 129.6,
128.5 and 127.2 (CH-7, CH-8, CH-9), 128.6 (CH-20, CH-
60), 125.1 (CH-30, CH-50), 83.7 (CH-3), 69.9 (benzylic
CH2), 52.6 (CH2-2) (proton and carbon atom shifts, and

5284
M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
nature of the N-900 aminalic bond were confirmed by HMBC
and HMQC studies in DMSO-d6). HR LSIMS calcd for
C21H19ClN6O6SNa (M+Na)+ 541.0673, found 541.0673.
Anal. Calcd for C21H19ClN6O6S: C, 48.60; H, 3.69; N,
16.19; S, 6.18. Found: C, 48.79; H, 3.57; N, 15.81; S, 6.15.
100), 8.45 (s, 1H, H-800), 8.41 (d, J3 –2 ¼J5 –6 ¼8.8 Hz, 2H,
0
0
0
0
H-30,50), 8.15 (d, J2 –3 ¼J6 –5 ¼8.8 Hz, 2H, H-2 ,6 ), 8.02
0
0
0
0
0
0
(s, 1H, H-200), 7.46–7.36 (m, 3H, H-6, H-8, H-9), 7.27 (dd,
J9–8¼not det., J9–7¼1.8 Hz, 1H, H-9), 6.32 (d, J3–2
¼
8.7 Hz, 1H, H-3b), 4.76 (d, Jgem benzylic H¼13.9 Hz, 1H,
benzylic H), 4.68 (dd, Jgem 2b–2a¼not det., J2b–3b¼not det.,
1H, H-2b), 4.29 (d, Jgem benzylic H¼13.9 Hz, 1H, benzylic
4.3. Substitutions on 7-alkyl 6-chloropurines: formation
of 7-alkyl 6-oxopurines 11 and 12
H), 4.10 (dd, Jgem
¼14.9 Hz, J2a–3b¼10.3 Hz, 1H,
2a–2b
H-2a). 13C NMR (DMSO-d6, 75 MHz): d (ppm) 157.0
(C-400), 154.0 (C-600), 150.1 (C-40), 145.5 (C-10), 145.1
(CH-200), 142.0 (CH-800), 138.5 (C-10a), 137.0 (C-5a),
130.0 (CH-6), 129.4 (CH-8), 128.6 (CH-20, CH-60), 128.5
(CH-7), 127.9 (CH-9), 125.0 (CH-30, CH-50), 114.0 (C-500),
84.3 (CH-3), 69.4 (CH2-5), 53.2 (CH2-2) (proton and carbon
atom shifts, and the nature of the N-700 aminalic bond were
confirmed by HMBC and HMQC studies in DMSO-d6).
HR LSIMS calcd for C20H16N6O6SNa (M+Na)+ 491.0750,
found 491.0750. Anal. Calcd for C20H16N6O6S: C, 51.28;
H, 3.44; N, 17.94; S, 6.85. Found: C, 51.35; H, 3.70; N,
17.89; S, 6.66.
Solutions A (27 mM of 3b in DMSO-d6 with 16 mM of H2O/
D2O) and B (27 mM of 3d in DMSO-d6 with 16 mM of H2O/
D2O) were prepared and kept at 4 ꢀC. Periodic registration of
1H NMR spectra was performed to control any changes that
took place. In solution A, a mixture of compounds 11 (more
than 50%) and 3b could be observed after 3 days, and after 1
week from the preparation of the solution 11 was the only
existing compound. In solution B, no traces of 12 could be
observed after one week and it was necessary one month
for the complete transformation of 3d into 12.
Solutions C containing 10 mM 3b in DMSO-d6 (1 mL,
16 mM D2O/H2O), solution D containing 10 mM 3b in
DMSO-d6 (50 mL, 16 mM D2O/H2O) (0.95 mL), solution
E containing 10 mM 3d in DMSO-d6 (1 mL, 16 mM D2O)
and solution F containing 10 mM 3d in DMSO-d6 (50 mL,
16 mM D2O/H2O) (0.95 mL) were prepared, maintained at
rt and controlled in a qualitative manner by TLC. Compound
3b disappeared in favour of 11 in solution D after 44 h but 5
days were necessary for its disappearance from solution C.
However, 3d was the only compound noticeable in solutions
E and F after 6 days from the preparation.
4.4. Substitutions on 7-alkyl-6-chloropurines: formation
of 7-alkyl 6-phenylthiopurines
4.4.1. (RS)-7-[1-(2-Nitrobenzenesulfonyl)-1,2,3,5-tetra-
hydro-4,1-benzoxazepin-3-yl]-6-phenylthio-7H-purine
13. K2CO3 (1.6 equiv) and PhSH (0.8 equiv) were added at rt
to a solution of 3b (1.0 equiv) in DMF (5 mL/mmol). The
mixture was stirred at rt for 40 min. The reaction was halted
by the addition of EtOAc and water. The aqueous layer was
extracted with EtOAc and the combined organic layers were
dried (Na2SO4), filtered and evaporated. Purification of 15
was carried out by flash chromatography with gradient elu-
tion using EtOAc/hexane mixtures (0/1/2/1). Compound
15: Rf (EtOAc): 0.34; white solid; mp 131.9–132.3 ꢀC; yield
61%. 1H NMR (CDCl3, 300 MHz): d (ppm) 8.77 (s, 1H, H-
800 or H-200), 8.23 (s, 1H, H-200 or H-800), 7.92 (dd, J1¼7.4 Hz,
J2¼not det., 1H), 7.79–7.75 (m, 2H), 7.70–7.59 (m, 3H),
7.49–7.43 (m, 4H), 7.39 (ddd, J1¼J2¼7.4 Hz, J3¼1.1 Hz,
1H), 7.28 (ddd, J1¼J2¼7.6 Hz, J3¼1.8 Hz, 1H), 7.00 (d,
J¼7.7 Hz, 1H), 6.59 (dd, J3b–2a¼10.0 Hz, J3b–2b¼2.0 Hz,
1H, H-3b), 5.20 (d, Jgem benzylic H¼13.6 Hz, 1H, benzylic
4.3.1. (RS)-7-[1-(2-Nitrobenzenesulfonyl)-1,2,3,5-tetra-
hydro-4,1-benzoxazepin-3-yl]-1,7-dihydropurin-6-one
11. Rf (MeOH/CH2Cl2 0.5/10): 0.39; white solid; mp 215.0–
217.0 ꢀC. 1H NMR (DMSO-d6, 400 MHz): d (ppm) 8.55 (s,
1H, H-800), 8.09–8.05 (m, 2H, H-30, H-60), 8.04 (s, 1H, H-
200), 7.97 (ddd, J4 –3 ¼J4 –5 ¼7.7 Hz, J4 –6 ¼1.2 Hz, 1H,
0
0
0
0
0
0
H-40), 7.88 (ddd, J5 –4 ¼J5 –6 ¼7.7 Hz, J5 –3 ¼1.2 Hz, 1H,
0
0
0
0
0
0
H-50), 7.49 (dd, J6–7¼7.1 Hz, J6–8¼1.9 Hz, 1H, H-6), 7.41–
7.33 (m, 2H, H-8, H-8), 7.04 (dd, J9–8¼7.2 Hz, J9–7
1.8 Hz, 1H, H-9), 6.35 (dd, J3b–2a¼10.1 Hz, J3b–2b
¼
¼
1.6 Hz, 1H, H-3b), 4.85 (d, Jgem benzylic H¼13.8 Hz, 1H, benz-
ylic H), 4.64 (d, Jgem benzylic H¼13.7 Hz, 1H, benzylic H),
4.53 (dd, Jgem 2b–2a¼14.8 Hz, J2b–3b¼1.6 Hz, 1H, H-2b),
4.24 (dd, Jgem 2a–2b¼14.9 Hz, J2a–3b¼10.2 Hz, 1H, H-2a).
13C NMR (DMSO-d6, 75 MHz): d (ppm) 156.7 (C-400),
153.8 (C-600), 147.2 (C-20), 145.3 (CH-200), 142.1 (CH-800),
138.5 (C-10a), 137.6 (C-6), 135.4 (CH-40), 133.1 (CH-50),
132.1 (C-10), 130.4 (CH-60), 130.1 (CH-6), 129.4 (CH-8),
128.7 (CH-7), 127.8 (CH-9), 125.0 (CH-30), 114.1 (C-500),
85.0 (CH-3), 69.7 (CH2-5), 53.5 (CH2-2) (proton and carbon
atom shifts, and the nature of the N-700 aminalic bond were
confirmed by HMBC and HMQC studies in DMSO-d6).
HR LSIMS calcd for C20H16N6O6NaS (M+Na)+ 491.0750,
found 491.0749. Anal. Calcd for C20H16N6O6S: C, 51.28;
H, 3.44; N, 17.94; S, 6.85. Found: C, 51.42; H, 3.61; N,
17.90; S, 6,74.
H), 4.91 (dd, Jgem
¼14.6 Hz, J2b–3b¼2.0 Hz, 1H,
2b–2a
H-2b), 4.90 (d, Jgem benzylic H¼13.6 Hz, 1H, benzylic H),
3.82 (dd, Jgem 2a–2b¼14.7 Hz, J2a–3b¼10.0 Hz, 1H, H-2a).
13C NMR (CDCl3, 75 MHz): d (ppm) 159.2 (C-400), 153.3
(C-600), 153.2 (CH-200 or CH-800), 148.0 (C-20), 144.2 (CH-
800 or CH-200), 138.6, 138.4, 133.6, 127.2, 122.2, 135.5
(2C, PhS), 134.7, 132.2, 132.2, 130.4, 129.8, 129.7, 129.5,
128.4 and 124.9 (CH-aromatics), 86.8 (CH-3), 71.2 (benz-
ylic CH2), 54.8 (CH2-2). HR LSIMS calcd for
C26H20N6O5S2Na (M+Na)+ 583.0834, found 583.0833.
Anal. Calcd for C26H20N6O5S2: C, 55.70; H, 3.60; N,
14.99; S, 11.44. Found: C, 55.51; H, 3.81; N, 15.21; S, 11.59.
4.4.2. (RS)-7-[1-(4-Nitrobenzenesulfonyl)-1,2,3,5-tetra-
hydro-4,1-benzoxazepin-3-yl]-6-phenylthio-7H-purine
14 and (RS)-6-phenylthio-7-(1,2,3,5-tetrahydro-4,1-benz-
oxazepin-3-yl)-7H-purine 15. Compound 15 was prepared
from 3d following the procedure described for 13, using
0.8 equiv of PhSH and 1.6 equiv of K2CO3 for each equiva-
lent of 3d. After 40 min the reaction was halted as explained
for the preparation of 13. The purification of the products
4.3.2. (RS)-7-[1-(4-Nitrobenzenesulfonyl)-1,2,3,5-tetra-
hydro-4,1-benzoxazepin-3-yl]-1,7-dihydropurin-6-one
1
12. Rf (EtOAc): 0.2; white solid; mp 171.8–172.6 ꢀC. H
NMR (DMSO-d6, 300 MHz): d (ppm) 12.50 (s, 1H, NH-

M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
5285
was carried out by flash chromatography [gradient elution
with (CH3)2CO/CH2Cl2 mixtures (0/1/1/0)] and com-
pounds 14 and 15 were obtained. Compound 14: elution
by flash chromatography, (CH3)2CO/CH2Cl2 2/10; Rf
(EtOAc): 0.57; white solid; mp 116.8–117.5 ꢀC; yield
using SHELXL-97.31 Positional and anisotropic atomic dis-
placement parameters were refined for all non-hydrogen
atoms. Hydrogen atoms were inserted at calculated positions
and constrained with isotropic thermal parameters. Criteria
of a satisfactory complete analysis were the ratios of rms
shift to standard deviation less than 0.001 and no significant
features in final difference maps. Molecular graphics and
geometrical calculations were obtained from PLATON32
and SHELXTL.33 Relevant crystal data: formula
C20H15ClN6O5S, formula weight 486.89, T¼100(2) K, crys-
tal system monoclinic, space group P21/n, unit cell dimen-
1
36%. H NMR (CDCl3, 300 MHz): d (ppm) 8.79 (s, 1H,
H-800 or H-200), 8.34 (d, J¼8.9 Hz, 2H, H-30,50), 8.19 (s,
1H, H-200 or H-800), 8.02 (d, J¼8.9 Hz, 2H, H-20,60), 7.66–
7.63 (m, 2H), 7.54–7.50 (m, 3H), 7.45–7.33 (m, 4H), 6.50
(dd, J3b–2a¼10.0 Hz, J3b–2b¼2.0 Hz, 1H, H-3b), 4.92 (dd,
Jgem
Jgem
¼14.5 Hz, J2b–3b¼2.0 Hz, 1H, H-2b), 4.78 (d,
¼13.8 Hz, 1H, benzylic H), 4.63 (d,
2b–2a
benzylic
˚
sions a¼7.9549(6), b¼14.6821(10), c¼17.0430(12) A and
H
Jgem benzylic H¼13.7 Hz, 1H, benzylic H), 3.73 (dd, Jgem 2a–2b
¼
a¼90ꢀ, b¼90.6720(10)ꢀ and g¼90ꢀ, Z¼4, D¼1.625 Mg/
m3, m(MoKa)¼0.348 mmꢁ1, measured/unique reflections
12,114/4508 (R(int) 0.0233), refined parameters 298, final
R1 (I>2s(I))¼0.0505 and wR2¼0.1229, GOF¼1.125.
CCDC reference number 625270. Copies of the data can
be obtained, free of charge, on application to CCDC, 12
Union Road, Cambridge, CB2 1EZ, UK [fax: +44 1223
336033 or e-mail: deposit@ccd.cam.ac.uk].
14.6 Hz, J2a–3b¼10.0 Hz, 1H, H-2a). HR LSIMS calcd for
C26H20N6O5S2Na (M+Na)+ 583.0834, found 583.0834.
Anal. Calcd for C26H20N6O5S2: C, 55.70; H, 3.60; N,
14.99; S, 11.44. Found: C, 55.79; H, 3.91; N, 15.01; S, 11.63.
Compound 15: elution by flash chromatography, (CH3)2CO/
CH2Cl2 1/1; yield, 37%. Rf (EtOAc): 0.17; white solid; mp
1
90.0–91.3 ꢀC; H NMR (CDCl3, 300 MHz): d (ppm) 8.76
(s, 1H, H-200 or H-800), 8.56 (s, 1H, H-200 or H-800), 7.65–
7.59 (m, 2H, Ph–S), 7.50–7.44 (m, 3H, Ph–S), 7.25–7.20
(m, 2H), 6.97 (ddd, J1¼J2¼not det., J3¼not det., 1H), 6.89
Acknowledgements
This study was supported by the Instituto de Salud Carlos III
(d, J¼7.6 Hz, 1H), 6.45 (dd, J3b–2a¼6.9 Hz, J3b–2b
¼
ꢀ
(Fondo de Investigacion Sanitaria) through the project no.
PI041206, and by the Junta de Andalucıa through the Excel-
2.1 Hz, 1H, H-3b), 4.92 (d, Jgem benzylic H¼14.2 Hz, 1H, ben-
zylic H), 4.88 (d, Jgem benzylic H¼14.1 Hz, 1H, benzylic H),
4.16 (s, 1H, NH-1), 3.95 (d, J¼13.2 Hz, 1H, H-2b), 3.48
(dd, Jgem 2a–2b¼13.4 Hz, J2a–3b¼6.9 Hz, 1H, H-2a). 13C
NMR (CDCl3, 75 MHz): d (ppm) 159.2 (C-400), 153.0 (CH-
200 or CH-800), 152.7, 148.3 (C-600, C-11), 145.6 (CH-800 or
CH-200), 135.4 (2CH, PhS), 129.9, 129.6, 129.6, 129.4,
121.8 and 119.0 (CH-aromatics), 128.4, 127.0, 122.0, 87.1
(CH-3), 70.5 (CH2-5), 53.5 (CH2-2). HR LSIMS calcd for
C20H17N5OSNa (M+Na)+ 398.1052, found 398.1051. Anal.
Calcd for C20H17N5OS: C, 63.98; H, 4.54; N, 18.65; S,
8.54. Found: C, 63.65; H, 4.61; N, 18.91; S, 8.64.
ꢀ
lence Research Project no. 00636.
References and notes
ꢀ
1. Dıaz-Gavilan, M.; Rodrıguez-Serrano, F.; Gomez-Vidal, J. A.;
Marchal, J. A.; Aranega, A.; Gallo, M. A.; Espinosa, A.;
ꢀ
ꢀ
ꢀ
ꢀ
Campos, J. M. Tetrahedron 2004, 60, 11547.
ꢀ
ꢀ
ꢀ
2. Dıaz-Gavilan, M.; Gomez-Vidal, J. A.; Entrena, A.; Gallo,
M. A.; Espinosa, A.; Campos, J. M. J. Org. Chem. 2006, 71,
1043.
3. Vorbr€uggen, H.; Ruh-Pohlenz, C. Synthesis of Nucleosides.
Organic Reactions (New York); Paquette, L. A., Ed.; John
Wiley and Sons: New York, NY, 2000; Vol. 55, pp 1–630.
4. Hasan, A.; Srivstava, P. J. Med. Chem. 1992, 35, 1435.
5. Brik, A.; Wu, C. Y.; Best, M. D.; Wong, C. H. Bioorg. Med.
Chem. 2005, 13, 4622.
6. Yamada, K.; Sakata, S.; Yoshimura, Y. J. Org. Chem. 1998, 63,
6891.
7. Peterson, M. L.; Vince, R. J. Med. Chem. 1991, 34, 2787.
4.4.3. (RS)-6-Phenylthio-7-(1,2,3,5-tetrahydro-4,1-ben-
zoxazepin-3-yl)-7H-purine 15. The reaction from 3b was
performed as described for the preparation of 13, using
2.0 equiv of PhSH and 3 equiv of K2CO3. After 40 min of
stirring at rt the reaction was halted as described for the prep-
aration of 13. The only product 15 was purified by flash chro-
matography [gradient elution using EtOAc/hexane (0/1/1/
0)]: yield 89%.
ꢀ
8. Hockova, D.; Budesinsky, M.; Marek, R.; Marek, J.; Holy, A.
Eur. J. Org. Chem. 1999, 2675.
4.5. Crystallization and X-ray diffraction on 3b
9. Garner, P.; Ramakanth, S. J. Org. Chem. 1988, 53, 1294. N-7
Alkylated regioisomers were obtained from silylated purines
using SnCl4 and MeCN at rt, while the N-9 alkylated products
were obtained with the same purines, but using TMSOTf and
refluxing 1,2-dichloroethane as solvent.
10. Okumura, K.; Oine, T.; Yamada, Y.; Tomie, M.; Adachi, T.;
Natura, T.; Kawazu, M.; Mizoguchi, T.; Inoue, I. J. Org.
Chem. 1971, 36, 1573.
ꢀ
11. Marek, R.; Brus, J.; Tousek, J.; Kovacs, L.; Hockova, D. Magn.
Reson. Chem. 2002, 40, 353.
12. Holy, A.; Dvorakova, H.; Jindrich, J.; Masojidkova, M.;
Budesinsky, M.; Balzarini, J.; Andrei, G.; De Clercq, E.
J. Med. Chem. 1996, 39, 4073.
Suitable crystals of 3b were obtained after dissolving this
compound in the minimum volume of dichloromethane
and the addition of a slightly larger volume of chloroform.
A vial with a screw top allowed the slow evaporation of
the solvents at 4 ꢀC, to the formation of colourless crystals.
A crystal of 3b was mounted on a glass fibre and used for
data collection. Crystal data were collected at 100(2) K, us-
ing a Bruker SMART CCD 1000 diffractometer. Graphite
˚
monochromated Mo Ka radiation (l¼0.71073 A) was
used throughout. The data were processed with SAINT28
and corrected for absorption using SADABS (transmissions
factors: 0.9465–0.8537).29 The structure was solved by
direct methods using the programme SHELXS-9730 and
refined by full-matrix least-squares techniques against F2
13. For the numbering of the compounds, the atoms of the benzox-
azepine are tagged with numbers without primes, the atoms of

5286
M. Dꢀıaz-Gavilaꢀn et al. / Tetrahedron 63 (2007) 5274–5286
the R1 group are numbered with primes (0), while the purine
22. Pathak, A. K.; Pathak, V.; Seitz, L. E.; Suling, W. J.; Reynolds,
R. C. J. Med. Chem. 2004, 47, 273.
23. Keck, J. H., Jr.; Simpson, R. A.; Wong, J. L. J. Org. Chem.
1978, 43, 2587.
24. The 6-chloropurine activation as an ammonium salts using
NEt3 allows the nucleophile attack at the position 6, in DMF,
see: Gurvich, V.; Kim, H. Y.; Hodge, R. P.; Harris, C. M.;
Harris, T. M. Abstracts of Papers, 218th ACS National
Meeting, New Orleans, 1999.
25. Fukuyama, T.; Jow, C. K.; Cheung, M. Tetrahedron Lett. 1995,
36, 6373.
26. Willker, W.; Leibfritz, D.; Kerssebaum, R.; Bermel, W. Magn.
Reson. Chem. 1993, 31, 287.
bases are numbered with double primes (00).
14. (a) Saniger, E.; Campos, J. M.; Entrena, A.; Marchal, J. A.;
ꢀ
ꢀ
ꢀ
Suarez, I.; Aranega, A.; Choquecillo, D.; Niclos, J.; Gallo,
M. A.; Espinosa, A. Tetrahedron 2003, 59, 5457; (b) Saniger,
E.; Campos, J. M.; Entrena, A.; Marchal, J. A.; Boulaiz, H.;
ꢀ
Aranega, A.; Gallo, M. A.; Espinosa, A. Tetrahedron 2003,
59, 8017.
15. Studies about stability of each tautomer of 6-chloropurine in
DMSO-d6 have been developed at 303 and 213 K, and preva-
ꢀ
lence of the 9H tautomer has been found, see: Seckarova, P.;
Marek, R.; Malinakova, K.; Kolehmainen, E.; Hockova, D.;
ꢀ
ꢀ
ꢀ
Hocek, M.; Sklenar, V. Tetrahedron Lett. 2004, 45, 6259.
16. Boryski, J.; Manikowski, A. Nucleosides Nucleotides 1999, 18,
1057.
17. Isomer’s identification has been made from the value of the
coupling constant between the olefinic hydrogens in each struc-
ture.
18. For examples of E2 syn eliminations of hydrogen atoms a- to
protonated purines, see: Rodgers, M. T.; Campbell, S.;
Marzluff, E. M.; Beauchamp, J. L. Int. J. Mass Spectrom. Ion
Processes 1995, 148, 1.
27. Hurd, R. E.; John, B. K. J. Magn. Reson. 1991, 91, 648.
28. Bruker. SMART and SAINT. Area Detector Control and
Integration Software; Bruker Analytical X-ray Instruments:
Madison, WI, 1997.
29. Sheldrick, G. M. SADABS. Program for Empirical Absorption
€
Correction of Area Detector Data; University of Gottingen:
€
Gottingen, Germany, 1997.
30. Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
31. Sheldrick, G. M. SHELXL-97. Program for the Refinement of
€
Crystal Structures; University of Gottingen: Gottingen,
€
19. Girgis, N. S.; Cottam, H. B.; Larson, S. B.; Robins, R. K.
J. Heterocycl. Chem. 1987, 24, 821.
20. Horhota, A. T.; Szostak, J. W.; McLaughlin, L. W. Org. Lett.
2006, 8, 5345.
Germany, 1997.
32. Spek, A. L. PLATON. A Multipurpose Crystallographic Tool;
Utrecht University: Utrecht, 2003.
21. Koppel, H. C.; O’Brien, D. E.; Robins, R. K. J. Org. Chem.
1959, 24, 259.
33. Bruker. SHELXTL: Integrated System for the Determination of
Crystal Structures; Bruker AXS: Madison, WI, 2000.