Amino-substituted O6-benzyl-5-nitrosopyrimidines: Interplay of molecular, molecular-electronic and supramolecular structures

Article abstract of DOI:10.1107/S0108768101021607

The structures of eight 2,4,6-trisubstituted-5-nitrosopyrimidines (one of which crystallizes in two polymorphs) have been determined, including seven O⁶-benzyl derivatives which are potential, or proven, in vitro inhibitors of the human DNA-repair protein O⁶-alkylguanine-DNA-transferase. In the derivatives having an amino substituent at the 4-position, an intramolecular N-H...O hydrogen bond with the nitroso O as an acceptor leads to an overall molecular shape similar to that of substituted purines. There is a marked propensity for these nitroso compounds to crystallize with Z′ = 2. The structure of an analogue with no nitroso group is also reported for comparative purposes. Compounds containing the N-alkyl substituents -NHCH₂COOEt, -NHCH₂CH₂COOEt and -NHCH(CH₂Ph)COOEt, derived from amino acid esters, exhibit a rich variety of conformational behaviour, and in all of the nitroso compounds the bond lengths provide strong evidence for a highly polarized electronic structure. Associated with this polarization is extensive charge-assisted hydrogen bonding between the molecules, leading to supra-molecular aggregation in the form of finite (zero-dimensional) aggregates, chains, molecular ladders, sheets and frameworks.

Full text of DOI:10.1107/S0108768101021607

research papers
Acta Crystallographica Section B
Structural
6
Amino-substituted O -benzyl-5-nitrosopyrimidines:
Science
interplay of molecular, molecular-electronic and
supramolecular structures
ISSN 0108-7681
a,b
Antonio Quesada, Antonio
Received 26 November 2001
The structures of eight 2,4,6-trisubstituted-5-nitrosopyrimi-
dines (one of which crystallizes in two polymorphs) have been
a
a
Accepted 17 December 2001
Marchal, Manuel Melguizo,
6
determined, including seven O -benzyl derivatives which are
potential, or proven, in vitro inhibitors of the human DNA-
a
Manuel Nogueras, Adolfo
a
b,c
S a nchez, John N. Low,
6
repair protein O -alkylguanine-DNA-transferase. In the
b
Debbie Cannon, Dorcas M. M.
d
derivatives having an amino substituent at the 4-position, an
intramolecular NÐHÁ Á ÁO hydrogen bond with the nitroso O
as an acceptor leads to an overall molecular shape similar to
that of substituted purines. There is a marked propensity for
Farrell and Christopher
d
Glidewell *
a
0
Departamento de Qu Âõ mica Inorg a nica y Org a -
these nitroso compounds to crystallize with Z = 2. The
nica, Universidad de Ja e n, 23071 Ja e n, Spain,
structure of an analogue with no nitroso group is also reported
for comparative purposes. Compounds containing the N-alkyl
substituents ÐNHCH COOEt, ÐNHCH CH COOEt and
b
School of Engineering, University of Dundee,
c
Dundee DD1 4HN, Scotland, Department of
2
2
2
Chemistry, University of Aberdeen, Meston
Walk, Old Aberdeen AB24 3UE, Scotland, and
ÐNHCH(CH Ph)COOEt, derived from amino acid esters,
2
d
School of Chemistry, University of St Andrews,
exhibit a rich variety of conformational behaviour, and in all
of the nitroso compounds the bond lengths provide strong
evidence for a highly polarized electronic structure. Asso-
ciated with this polarization is extensive charge-assisted
hydrogen bonding between the molecules, leading to supra-
molecular aggregation in the form of ®nite (zero-dimensional)
aggregates, chains, molecular ladders, sheets and frameworks.
St Andrews KY16 9ST, Scotland
Correspondence e-mail: cg@st-andrews.ac.uk
1
. Introduction
Of all the types of heteroaromatic compounds, the pyrimidines
and their fused bicyclic analogues, the purines and pteridines
are of particular interest on account of their diverse roles in
biological systems. The important role that they play in
biological processes, as nucleic acids constituents and enzyme
cofactors, together with their use as therapeutic and diagnostic
agents, make the preparation of new pyrimidine derivatives a
permanent target for synthetic chemistry. 4-Amino-5-nitro-
sopyrimidines have long been important as intermediates for
the synthesis of bicyclic azaheterocycles (Lister, 1996). More
recently, their 6-alkoxy derivatives have been found to act as
6
inhibitors of the human DNA-repair protein O -alkylguanine-
DNA-transferase (AGT): thus, 2,4-diamino-6-benzyloxy-5-
nitrosopyrimidine (1) has been shown (Chae et al., 1995) to be
a more potent in vitro inhibitor of AGT than the prototype
6
inhibitor O -benzyloxyguanine (2), currently in clinical trials
(Friedman et al., 1998).
The use of newly developed synthetic methodology
Marchal et al., 1998, 2000; Quesada et al., 2000) gives ready
(
6
access to a wide range of O -alkyl-5-nitrosopyrimidines as
potential AGT inhibitors, and we present here a study of the
molecular and supramolecular structures of seven such
compounds (3)±(9), one of which crystallizes in two distinct
monoclinic polymorphs (5A) and (5B), together with a study
of an analogue (10) containing no nitroso substituent. By the
use of in vitro assays performed on mutant E. coli strains
#
2002 International Union of Crystallography
Printed in Great Britain ± all rights reserved
6
ꢀ
300 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
2
were solved by direct methods and re®ned with all data on F .
expressing human AGT (Abril et al., 1999), (4) has already
been shown to exhibit more potent AGT inhibition than the
reference compound (2)
2
2
A weighting scheme based upon P ˆ ‰F ‡ 2F Š=3 was
o
c
employed in order to reduce statistical bias (Wilson, 1976). All
H atoms were located from difference maps and all were fully
ordered: where H atoms of amino groups were found not to lie
.
2
. Experimental
.1. Synthesis
-Amino-4,6-dimethoxypyrimidine
2
2
(purchased
from
Aldrich) was converted into (10) by reaction with sodium
benzylate in benzyl alcohol, and (10) was subsequently
converted into (3)±(9) and (11) using methods previously
described (Marchal et al., 1998, 2000; Quesada et al., 2000).
Crystals suitable for single-crystal X-ray diffraction were
obtained from the following solvents: acetonitrile/ethanol/
water (1/1/1 by volume), (3) as purple blocks, (5) as a mixture
of blue plates [polymorph (5A)] and pink needles [polymorph
(
5B)], (6) as pink plates and (7) as pink plates; isopentanol (3-
methyl-1-butanol) (4) as violet needles; dichloromethane/
methanol (1/1 by volume), (8) as blue blocks; acetonitrile (9)
as green blocks, (10) as colourless blocks and (11) as colourless
plates: considerable dif®culty was experienced in obtaining
(
11) in suitably crystalline form.
2
.2. Data collection, structure solution and refinement
Diffraction data for (3)±(11) were collected at 150 (2) K
using a Nonius Kappa-CCD diffractometer, using graphite-
Ê
monochromated Mo Kꢀ radiation (ꢁ = 0.71073 A). Other
details of cell data, data collection and re®nement are
summarized in Table 1, together with details of the software
employed (Blessing, 1995, 1997; Ferguson, 1999; Nonius, 1997;
Otwinowski & Minor, 1997; Sheldrick, 1997a,b; Spek, 2001).
For (3), (8) and (11) the space group P2 /n was uniquely
1
assigned from the systematic absences; similarly, the space
group P2 /c was uniquely assigned for (9). Two polymorphs
1
were observed for (5) and both are monoclinic: for polymorph
(
5A), the space group P2 /c was uniquely assigned from the
1
systematic absences; for polymorph (5B), the systematic
absences permitted P2 and P2 /m as the possible space
groups; P2 was chosen and subsequently con®rmed by the
1
1
1
successful structure analysis. For (10) the systematic absences
permitted space groups Cc and C2/c; C2/c was chosen and
subsequently con®rmed by the successful structure analysis.
Compounds (4), (6) and (7) are all triclinic; for (4) and (7), the
ꢀ
space group P1 was selected and subsequently con®rmed by
the successful structure analysis, but space group P1 was
chosen for (6) because the molecules are chiral. The structures
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 301

research papers
Table 1
Experimental details.
(3)
(4)
(5A)
(5B)
(6)
Crystal data
Chemical formula
Chemical formula
weight
C H N O
5
335.37
C H N O
5
331.34
C H N O
5
345.36
C H N O
5
345.36
C H N O
4
421.45
1
8
17
2
15 17
4
16 19
4
16 19
4
22 23
5
Cell setting, space
group
Monoclinic, P2 =n
Triclinic, P1
Monoclinic, P2 =c
Monoclinic, P21
Triclinic, P1
1
1
Ê
a, b, c (A)
11.677 (2), 13.362 (3),
21.478 (4)
90, 102.58 (3), 90
7.3525 (4), 7.5420 (5),
14.9454 (13)
91.797 (3), 100.467
5.2399 (16), 31.1630
(14), 9.792 (4)
90, 100.200 (12), 90
7.4830 (2), 16.6110
(6), 13.5378 (6)
90, 91.5420 (11), 90
7.9899 (3), 8.4278 (3),
17.0042 (7)
99.9100 (12), 90.5610
(13), 111.4880 (17)
1046.22 (7)
2
1.338
Mo Kꢀ
14 344
ꢀ
, ꢂ, ꢃ ꢁꢂ†
(3), 106.102 (6)
Ê
V (A )
3
3270.7 (11)
8
1.362
Mo Kꢀ
21 450
780.06 (10)
2
1.411
Mo Kꢀ
11 231
1573.7 (8)
4
1.458
Mo Kꢀ
66 201
1682.14 (11)
4
1.364
Mo Kꢀ
8888
Z
D (Mg m
�
3
)
Radiation type
x
No. of re¯ections for
cell parameters
ꢂ
ꢄ range ( )
ꢅ (mm
1.0±27.48
0.093
150 (2)
Block, purple
0.34 Â 0.18 Â 0.18
1.02±27.48
0.105
150 (2)
Needle, violet
0.50 Â 0.20 Â 0.10
1.02±27.48
0.108
150 (2)
Plate, blue
0.20 Â 0.10 Â 0.02
1.0±27.48
0.101
150 (2)
Needle, pink
0.50 Â 0.20 Â 0.18
2.91±27.48
0.095
150 (2)
Plate, pink
0.30 Â 0.20 Â 0.03
�
1
)
Temperature (K)
Crystal form, colour
Crystal size (mm)
Data collection
Diffractometer
Data collection
method
Absorption correc-
tion
Tmin
Kappa-CCD
Kappa-CCD
Kappa-CCD
Kappa-CCD
Kappa-CCD
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans, and ! scans
with ꢆ offsets
Multi-scan
0.9690
0.9834
36 834, 7334, 4449
0.9492
0.9895
10 165, 3418, 1860
0.9788
0.9978
10 470, 2525, 822
0.9513
0.9821
19 365, 2992, 2487
0.9721
0.9972
14 846, 8664, 6321
Tmax
No. of measured,
independent and
observed para-
meters
Criterion for
observed re¯ec-
tions
I > 2ꢇ(I)
I > 2ꢇ(I)
I > 2ꢇ(I)
I > 2ꢇ(I)
I > 2ꢇ(I)
Rint
ꢄmax ( )
Range of h, k, l
0.063
27.48
0.065
27.50
� 9 ! h ! 9
� 9 ! k ! 9
� 19 ! l ! 19
0.065
25.00
� 6 ! h ! 6
� 36 ! k ! 37
� 11 ! l ! 11
0.083
25.00
� 8 ! h ! 8
� 19 ! k ! 19
� 16 ! l ! 16
0.043
27.45
� 10 ! h ! 10
� 10 ! k ! 10
� 22 ! l ! 22
ꢂ
� 14 ! h ! 14
�
�
17 ! k ! 17
27 ! l ! 27
Re®nement
Re®nement on
2
2
2
2
2
F
0.0515, 0.1495, 0.99
F
0.0551, 0.1638, 1.023
F
0.1059, 0.2829, 1.001
F
0.0551, 0.1492, 1.033
F
0.0455, 0.1095, 1.009
2
2
R‰F >2ꢇꢁF †Š,
2
wRꢁF †, S
No. of re¯ections and
parameters used in
re®nement
7334, 451
3418, 218
2525, 227
2992, 453
4686, 561
H-atom treatment
H-atom parameters
constrained
H-atom parameters
constrained
H-atom parameters
constrained
H-atom parameters
constrained
H-atom parameters
constrained
2
2
2
2
2
2
2
2
2
2
Weighting scheme
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
o
o
o
o
o
2
2
2
2
2
(0.0833P) ], where
P = (F + 2F )/3
(0.0778P) ], where
2
(0.0757P) +
2.3684P], where P
(0.1159P) ], where
2
(0.0638P) ], where
2
2
o
2
c
2
c
2
o
2
P = (F + 2F )/3
o
P = (F + 2F )/3
c
P = (F + 2F )/3
o c
2
2
=
0.002
(F + 2F )/3
o c
ꢁ
Á=ꢇ†
0.001
0.301, � 0.406
0.000
0.385, � 0.512
0.004
0.555, � 0.331
0.001
0.195, � 0.246
max
Ê
� 3
)
Áꢈ , Áꢈmin (e A
max
0.231, � 0.219
(9)
(7)
(8)
(10)
(11)
Crystal data
Chemical formula
Chemical formula
weight
Cell setting, space
group
Ê
a, b, c (A)
C H N O
3
C H N O
25 22
426.47
C H N O
25 22
426.47
C H N O
3
307.35
C H N O
24 38:20 12 3:10
544.47
2
25
5
4
4
3
4
3
18 17
2
435.48
Triclinic, P1
8.6595 (2), 16.3567
Monoclinic, P2 =n
Monoclinic, P2 =c
Monoclinic, C =c
Monoclinic, P2 =n
1
1
2
1
14.6856 (5), 7.7726
(2), 18.5721 (8)
13.8074 (3), 15.8175
(5), 10.2119 (5)
32.4902 (9), 7.5034
(2), 12.7804 (4)
18.4687 (8), 8.1067
(3), 19.1880 (6)
(4), 16.9014 (4)
6
ꢀ
302 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
Table 1 (continued)
(7)
(8)
(9)
(10)
(11)
ꢀ
, ꢂ, ꢃ ꢁꢂ†
105.2210 (13),
90, 99.8847 (13), 90
90, 105.7490 (11), 90
90, 98.8790 (16), 90
90, 110.580 (2), 90
1
1
04.6960 (13),
00.3720 (12)
Ê
V (A )
3
2154.95 (9)
4
1.342
Mo Kꢀ
9402
2088.45 (13)
4
1.356
Mo Kꢀ
3748
2146.54 (13)
4
1.320
Mo Kꢀ
10 763
3078.36 (15)
8
1.326
Mo Kꢀ
3482
2689.50 (17)
4
1.345
Mo Kꢀ
4758
Z
D (Mg m
�
3
)
Radiation type
x
No. of re¯ections for
cell parameters
ꢄ range ( )
ꢂ
1.32±27.48
0.094
150 (2)
Plate, pink
0.40 Â 0.40 Â 0.05
1.0±27.47
0.091
150 (2)
Block, blue
0.55 Â 0.30 Â 0.20
1.0±27.48
0.089
150 (2)
Prism, green
0.30 Â 0.24 Â 0.12
2.54±27.47
0.089
150 (2)
Block, colourless
0.40 Â 0.40 Â 0.26
1.31±24.99
0.095
150 (2)
Plate, colourless
0.20 Â 0.20 Â 0.07
�
1
ꢅ (mm
)
Temperature (K)
Crystal form, colour
Crystal size (mm)
Data collection
Diffractometer
Data collection
method
Absorption correc-
tion
Tmin
Kappa-CCD
Kappa-CCD
Kappa-CCD
Kappa-CCD
Kappa-CCD
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans, and ! scans
with ꢆ offsets
Multi-scan
' scans and ! scans
with ꢆ offsets
Multi-scan
0.9632
0.9953
33 749, 9402, 5874
0.9515
0.9820
12 659, 4516, 3296
0.9738
0.9894
18 529, 4857, 3078
0.9654
0.9773
13 174, 3482, 2677
0.9813
0.9934
50 212, 4758, 3200
Tmax
No. of measured,
independent and
observed para-
meters
Criterion for
observed re¯ec-
tions
I > 2ꢇ(I)
I > 2ꢇ(I)
I > 2ꢇ(I)
I > 2ꢇ(I)
I > 2ꢇ(I)
Rint
ꢄmax ( )
Range of h, k, l
0.092
27.48
0.046
27.47
� 19 ! h ! 19
� 10 ! k ! 10
� 23 ! l ! 24
0.075
27.48
� 17 ! h ! 17
� 20 ! k ! 17
� 13 ! l ! 13
0.044
27.47
� 40 ! h ! 42
� 8 ! k ! 9
� 16 ! l ! 15
0.060
25.00
� 21 ! h ! 21
� 8 ! k ! 9
� 22 ! l ! 22
ꢂ
� 11 ! h ! 11
�
�
21 ! k ! 21
20 ! l ! 21
Re®nement
Re®nement on
2
2
2
2
2
F
0.0667, 0.186, 0.99
F
0.0473, 0.1314, 1.065
F
0.0508, 0.1263, 1.011
F
0.0442, 0.115, 1.044
F
0.0761, 0.2362, 1.008
2
2
R‰F >2ꢇꢁF †Š,
2
wRꢁF †, S
No. of re¯ections and
parameters used in
re®nement
9402, 579
4516, 289
4857, 289
3482, 208
4758, 391
H-atom treatment
H-atom parameters
constrained
H-atom parameters
constrained
H-atom parameters
constrained
H-atom parameters
constrained
H-atom parameters
constrained
2
2
2
2
2
2
2
2
2
2
Weighting scheme
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
w = 1/[ꢇ (F ) +
o
o
o
o
o
2
2
2
2
2
(0.1133P) ], where
P = (F + 2F )/3
(0.0762P) ] where
2
(0.0662P) ], where
2
(0.0613P) +
0.8375P], where P
(0.1429P) +
1.8403P], where P
2
o
2
c
2
o
2
c
P = (F + 2F )/3
c
P = (F + 2F )/3
o
2
2
2
2
=
0.001
(F + 2F )/3
= (F + 2F )/3
o c
o
c
ꢁ
Á=ꢇ†
0.000
0.431, � 0.507
0.000
0.261, � 0.329
0.000
0.218, � 0.335
0.000
0.524, � 0.348
max
Ê
� 3
)
Áꢈ , Áꢈmin (e A
max
0.198, � 0.268
in the plane of the adjacent heteroaromatic ring, they were
treated using DFIX commands in SHELXL97, followed by
AFIX; otherwise they were included in the re®nements as
was set by reference to the known (S) con®guration at atoms
C17 and C27. In (11) the coordinates of the two independent
molecules were related by an approximate A-centring rela-
tionship, although a careful search for possible additional
riding atoms with NÐH 0.86±0.94, CÐH 0.95 (aromatic), 0.98
Ê
(
(
CH ), 0.99 (CH ) or 1.00 A (aliphatic CÐH). For (5B) and
3
symmetry showed that P2 /n was indeed the correct space
1
2
6) the re®ned values [� 0.9 (10) and 1.6 (16), respectively] of
group: in particular, hkl re¯ections having (k + l) odd were
present, although generally weak. In both molecules the
nitroso group is disordered over two sets of sites; for each
molecule the site-occupation factors (s.o.f.s) for the two sets of
sites were constrained to sum to unity, but the s.o.f.s for the
two molecules were allowed to re®ne independently. In
addition to the evident disorder of the nitroso groups, there
were indications of orientational disorder in several of the
puckered pyrrolidine rings, although it was possible to model
the Flack parameter (Flack, 1983) were inconclusive (Flack &
Bernardinelli, 2000) and hence for both compounds the
Friedel equivalents were merged before the ®nal re®nements.
The overall precision of the structure determination for the
blue P2 /c polymorph (5B) of (5) is rather low: even at
1
1
50 (2) K the proportion of the re¯ections labelled `observed'
is less than 33% and the fall-off of intensity with ꢄ is such that
ꢂ
data collection was terminated at ꢄ = 25 . For (6) the chirality
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 303

research papers
this satisfactorily only for one of the rings. As well as two
molecules of (11), the structure also contained two partially
occupied water sites containing O1 and O2, respectively,
whose re®ned s.o.f.s are 0.902 (9) and 0.196 (9), respectively. In
one of the pyrrolidine rings one C atom was modelled with
two sites, denoted C143 and C14B and with the s.o.f.s
constrained to sum to unity; these re®ned to 0.796 (7) and
combination of the pseudosymmetry and the multiple
disorder, together with the consistently poor quality of the
diffraction data (Rint = 0.102, probably connected with the
dif®culty experienced in obtaining crystalline material), indi-
cate that the re®nement of (11) is at the margin of tractability.
In an attempt to circumvent these problems by the use of
similar compounds in which the secondary amino components
are not subject to such disorder, the corresponding bis-(1-
piperidino), bis-(4-morpholino) and bis-(dimethylamino)
0.204 (7), respectively; since the C14B and O2 sites cannot
both be occupied in the same asymmetric unit as their contact
Ê
distance would be only ca 2.60 A, it seems likely that there is
correlated disorder here: O2 is present when site C143 is
occupied and absent when site C14B is occupied. Because of
this, and the effectively identical values of the re®ned s.o.f.s for
C14B and O2, these values were ®xed at 0.20 in the ®nal
re®nements, with that of C143 ®xed accordingly at 0.80. The
Figure 3
1
A molecule of (5) in the P2 /c polymorph (5A), showing the atom-
labelling scheme. Displacement ellipsoids are drawn at the 30%
probability level.
Figure 1
The two independent molecules in (3), showing the atom-labelling
scheme. Displacement ellipsoids are drawn at the 30% probability level.
Figure 4
Figure 2
A molecule of (4), showing the atom-labelling scheme. Displacement
ellipsoids are drawn at the 30% probability level.
1
The two independent molecules of (5) in the P2 polymorph (5B),
showing the atom-labelling scheme. Displacement ellipsoids are drawn at
the 30% probability level.
6
ꢀ
304 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
Table 2
Torsional angles ( ) in nitroso compounds involving benzyl substituents.
independent
molecules
differ
ꢂ
signi®cantly (see x3.2 below) and
this alone is suf®cient to preclude
any additional symmetry. Moreover,
(3)
(5B)
(4)
(5A)
n = nil
n = 1
n = 2
n = nil
n = 1
n = 2
(5) crystallizes in two polymorphic
forms, both monoclinic: (5A) in
Nn1ÐCn2ÐNn2ÐCn27
Cn5ÐCn4ÐNn4ÐCn47
Cn4ÐCn5ÐNn5ÐOn5
±
±
±
179.5 (2)²
� 2.1 (3)
±
±
±
171.2 (2) � 178.2 (2)
177.9 (8)² � 179.2 (4)²
179.4 (4)²
1.8 (6)
177.5 (4)
±
126.5 (4)³
� 176.2 (4)
0
2.4 (3) 0.3 (3)
� 1.8 (14)
179.4 (9)
±
� 1.0 (6)
� 174.4 (3)
±
112.8 (5)³
166.7 (3)
space group P2 /c with Z = 1 and
1
Cn5ÐCn6ÐOn6ÐCn67 � 174.6 (2±)
173.5 (2) � 176.8 (2)
0
(
5B) in space group P2 with Z = 2,
1
Cn2ÐNn2ÐCn27ÐCn41
±
±
and none of the three independent
conformers of (5) resembles the
other two (Figs. 3 and 4).
Cn4ÐNn4ÐCn47ÐCn41
Cn6ÐOn6ÐCn67ÐCn61
83.8 (2) � 139.8 (2)
108.2 (2)³ � 175.7 (9)³
165.7 (2) � 172.4 (2)
170.6 (2) 179.0 (8)
(
6)
(7)
(8)
n = nil
(9)
n = nil
n = 1
n = 2
n = 1
n = 2
176.7 (2)
Nn1ÐCn2ÐNn2ÐCn27
Cn5ÐCn4ÐNn4ÐCn47
Cn4ÐCn5ÐNn5ÐOn5
Cn5ÐCn6ÐOn6ÐCn67
Cn2ÐNn2ÐCn27ÐCn41
±
±
� 173.2 (2)
178.6 (2)§
� 178.4 (2)
3.2. Molecular conformations
173.3 (2)² � 170.3 (3)² � 175.3 (2)² � 176.7 (2)²
179.4 (2)
� 1.3 (2)
176.9 (2)
75.4 (2)
95.1 (2) � 179.6 (2)³³
168.2 (2)
� 81.6 (2)
178.6 (2)}
1.7 (2)
1.2 (4)
176.5 (2) � 179.0 (3)
� 1.5 (5)
� 1.5 (3)
179.6 (2)
� 76.1 (3)
80.2 (3)³
� 0.9 (3)
In each of (3)±(9) the nitroso
group is essentially coplanar with
the pyrimidine ring and in (3)±(8)
there is an intramolecular NÐ
HÁ Á ÁO hydrogen bond having the
nitroso O as an acceptor, which
doubtless reinforces this tendency to
coplanarity. There is thus a clearsi-
milarity in overall molecular shape
173.8 (2)
� 178.2 (2)
±
±
76.6 (3) � 167.2 (2)²²
Cn4ÐNn4ÐCn47ÐCn41 � 70.8 (4)³ � 76.9 (4)³
92.3 (2)³
Cn6ÐOn6ÐCn67ÐCn61 � 84.4 (3) 116.3 (3)
172.6 (2) � 164.1 (2)
² Cn5ÐCn4ÐNn4ÐCn7. ³ Cn4ÐNn4ÐCn7ÐCn8. § N1ÐC2ÐO1ÐC27. } C5ÐC4ÐO4ÐC47. ²² C2ÐO1Ð
C27ÐC21. ³³ C4ÐO4ÐC47ÐC41.
Table 3
Torsional angles ( ) involving aliphatic side chains in (5A), (5B) and (7).
ꢂ
between
4-amino-5-nitrosopyr-
(
5B)
(7)
n = 1
(5A)
n = nil
imidines (I) on the one hand and
purines (II) on the other.
n = 1
n = 2
126.5 (5)
69.2 (5)
123.9 (5)
n = 2
Cn4ÐNn4ÐCn7ÐCn8
Nn4ÐCn7ÐCn8ÐCn9
Cn7ÐCn8ÐCn9ÐOn91
Cn7ÐCn8ÐCn9ÐOn92
Cn8ÐCn9ÐOn92ÐCn10
Cn9ÐOn92ÐCn10ÐCn11
� 175.7 (9)
175.0 (8)
� 172.2 (9)
9.4 (12)
175.3 (7)
� 175.1 (8)
112.8 (5)
80.2 (3)
168.1 (2)
2.1 (3)
� 177.6 (2)
179.5 (2)
174.4 (2)
92.3 (2)
69.8 (2)
1.7 (3)
� 179.1 (3)
� 179.5 (2)
� 174.7 (2)
In the 6-benzyloxy substituents
the CH2 carbon is also nearly
coplanar with the pyrimidine ring,
with the torsional angle Cn5Ð
Cn6ÐOn6ÐCn67 (Figs. 1±9) always
63.1 (5)
� 179.3 (6)
� 0.6 (6)
� 61.7 (5)
� 175.9 (5)
173.8 (5)
177.6 (5)
� 103.5 (5)
ꢂ
ꢂ
within Æ7 of 180 , and oriented
analogues of (11) have all been examined, but none of these
has been obtained in suitably crystalline form.
Supramolecular analyses were made with the aid of
PLATON (Spek, 2001) and the diagrams were prepared with
the aid of PLATON (Spek, 2001). Figs. 1±10 show the mole-
cular structures of (3)±(11), with the atom-labelling schemes,
and Figs. 11±25 show aspects of the supramolecular structures.
Selected molecular dimensions are presented in Tables 2±6
1
and details of the hydrogen bonding are given in Table 7.
such that the 6-benzyloxy and the 5-nitroso groups are
directed away from one another (Table 2). On the other hand,
the ipso C of the associated aromatic ring is sometimes far
removed from this plane, as indicated by the values of the
torsional angles Cn6ÐOn6ÐCn67ÐCn61 in (6), for example:
however, in (3)±(5), (7) and (8) these torsional angles do not
3
. Results and discussion
3
.1. Crystallization characteristics
Compounds (3)±(7) are all characterized by amino substi-
tuents at positions 2 and 4, with a benzyloxy substituent at
position 6. Compounds (8) and (9) differ from the series (3)±
ꢂ
ꢂ
deviate from 180 by more than 20 . In (10), where there is no
nitroso group, the molecule is almost planar: the C5ÐCnÐ
(
7) in that (8) and (9) both carry benzyloxy substituents at
positions 2 and 4, respectively, in addition to that at position 6.
It is striking that (3), (5)±(7) and (11) all crystallize in forms
having Z = 2: in every case the conformations of the two
ꢂ
OnÐCn7 torsional angles are 177.4 (2) and 2.0 (2) for n = 4
or 6, respectively, and the corresponding CnÐOnÐCn7ÐCn1
0
ꢂ
angles are � 176.0 (2) and 179.2 (2) , respectively.
The most notable aspects of the molecular conformations in
this series are manifested in the behaviour of the aliphatic side
chains in (4)±(7). Compounds (5) and (7) both contain a
1
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: NA0133). Services for accessing these data are described
at the back of the journal.
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 305

research papers
ÐCH CH COOEt substituent at N4: (5) crystallizes in two
2
hydrogen bonds which dominate the molecular packing and
the supramolecular aggregation.
2
0
polymorphs and polymorph (5B) has Z = 2, and likewise (7)
0
crystallizes with Z = 2, so that (5) and (7) between them
provide ®ve independent molecules containing the
ÐCH CH COOEt substituent. In the atom sequence from
2
2
Cn4 to Cn11 (Table 3), there are ®ve independent torsional
angles and from these it is clear that no two molecules have
the same conformation for this substituent. This is most
readily summarized by classifying the torsional angles as
antiperiplanar (ap) and so on: these torsional characteristics
con®rm that all ®ve examples have completely different
0
conformations. Similarly, in (4) and (6) (which has Z = 2), the
three examples of ÐCHRCOOEt side chains [where R = H in
(
4) and PhCH in (6)] all have different conformations (Table
2
4
). It may be concluded that this rich conformational variety is
a consequence of the essentially complete ¯exibility of these
substituents, with very small energy barriers to rotation about
the single NÐC, CÐC and CÐO bonds, and that the precise
folding of the side chains is probably constrained largely by
the intermolecular interactions, particularly the hard
Figure 6
The two independent molecules in (7), showing the atom-labelling
scheme. Displacement ellipsoids are drawn at the 30% probability level.
Figure 5
The two independent molecules in (6), showing the atom-labelling
scheme. Displacement ellipsoids are drawn at the 30% probability level.
Figure 7
A molecule of (8), showing the atom-labelling scheme. Displacement
ellipsoids are drawn at the 30% probability level.
6
ꢀ
306 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
Table 4
Torsional angles ( ) involving aliphatic side chains in (4) and (6).
the O of the nitroso group; in addition, there is serious
competition between electronic and steric factors. Maximum
ꢉ-overlap between, on the one hand, the ring orbitals and, on
the other, the orbitals of the nitroso group and of the planar N
of the pyrrolidine substituents requires all of these orbitals to
be parallel, which in turn requires both planarity of the pyri-
midine ring and coplanarity with this ring of the CÐN O and
CÐNR2 substituent groups. This requirement would,
however, lead to a number of unreasonably short non-
bonded contacts between atoms in the various substituents,
ꢂ
(6)
(4)
n = nil
n = 1
n = 2
Cn4ÐNn4ÐCn7ÐCn8
Nn4ÐCn7ÐCn8ÐOn81
Nn4ÐCn7ÐCn8ÐOn82
Cn7ÐCn8ÐOn82ÐCn9
Cn8ÐOn82ÐCn9ÐCn10
Cn4ÐNn4ÐCn7ÐCn77
Nn4ÐCn7ÐCn77ÐCn71
108.0 (2)
� 70.8 (4)
� 17.0 (4)
166.9 (2)
173.8 (2)
� 173.0 (3)
170.2 (3)
� 68.9 (3)
� 76.9 (4)
� 52.2 (4)
127.6 (3)
� 175.3 (3)
� 87.6 (5)
162.4 (3)
3.9 (4)
� 175.9 (2)
175.4 (2)
� 172.9 (2)
±
±
� 171.6 (3)
in particular between nitroso group and the CH groups of
2
the pyrrolidines. There is some indication that some of the C
atoms in the pyrollidine rings may also be mobile and/or
disordered, but no satisfactory model for this could be
developed: in view of the extensive disorder in this
compound, the bond lengths and the electronic structure
will not be discussed further in x3.3 below.
In both the independent molecules in (11) the pyrimidine
rings are seriously puckered, and the nitroso O atoms (in both
orientations of the disordered nitroso substituent) are not
coplanar with the adjacent CÐCÐN fragments (Table 5). This
may reasonably be ascribed to the in¯uence of the pyrollidine
ring substituents at positions 4 and 6. There is now no
possibility of an intramolecular NÐHÁ Á ÁO hydrogen bond to
3.3. Intermolecular bond lengths
The bond distances in the heteroaromatic rings and their
immediate substituents display a number of patterns. We
consider ®rst those examples, (3)±(7), with amino substituents
Figure 8
A molecule of (9), showing the atom-labelling scheme. Displacement
ellipsoids are drawn at the 30% probability level.
Figure 10
Figure 9
A molecule of (10), showing the atom-labelling scheme. Displacement
ellipsoids are drawn at the 30% probability level.
The two independent molecules in (11), showing the atom-labelling
scheme. Displacement ellipsoids are drawn at the 30% probability level.
For the sake of clarity, only the major components are shown.
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 307

research papers
Table 5
Selected torsional angles ( ) in (11).
length, while bond c is always much shorter than bond b.
Finally, bonds j and k are similar in length, although in simple
neutral compounds where there is no possibility of signi®cant
electronic delocalization these distances normally differ by at
ꢂ
n = 1
11.7 (6)
10.9 (5)
� 25.8 (4)
20.8 (4)
� 1.7 (4)
� 16.3 (5)
143.7 (4)
� 2.0 (6)
167.4 (3)
n = 2
14.0 (5)
1.9 (4)
� 17.8 (4)
20.9 (4)
� 7.7 (4)
� 11.0 (5)
147.0 (3)
� 162.8 (3)
� 0.6 (5)
Nn1ÐCn2ÐNn3ÐCn4
Cn2ÐNn3ÐCn4ÐCn5
Nn3ÐCn4ÐCn5ÐCn6
Cn4ÐCn5ÐCn6ÐNn1
Cn5ÐCn6ÐNn1ÐCn2
Cn6ÐNn1ÐCn2ÐNn3
Nn3ÐCn4ÐCn5ÐNn5A
Cn4ÐCn5ÐNn5AÐOn5A
Cn6ÐCn5ÐNn5AÐOn5A
Ê
least 0.20 A (Talberg, 1977; Schlemper et al., 1986) and the NO
Ê
distance rarely exceeds 1.25 A (Davis et al., 1965; Bauer &
Andreassen, 1972; Talberg, 1977; Schlemper et al., 1986).
These observations taken together con®rm the polarized form
(B) as the dominant contributor to the overall electronic
structure, rather than the classical charge-localized form (A):
Table 6
Selected intramolecular distances for (3)±(10) (A).
Ê
a
b
c
d
e
f
(
3)
4)
1.337 (2) 1.369 (2) 1.306 (2) 1.436 (3) 1.445 (2) 1.339 (2)
.344 (2) 1.366 (2) 1.298 (2) 1.426 (2) 1.454 (2) 1.338 (2)
1.339 (3) 1.368 (3) 1.306 (3) 1.428 (3) 1.436 (3) 1.332 (3)
1
(
(
(
5A) 1.351 (11) 1.368 (10) 1.317 (10) 1.386 (12) 1.427 (11) 1.345 (9)
5B) 1.342 (5) 1.383 (5) 1.304 (5) 1.435 (6) 1.441 (5) 1.336 (5)
1
.333 (5) 1.371 (5) 1.301 (5) 1.427 (5) 1.450 (5) 1.340 (5)
1.344 (4) 1.371 (4) 1.303 (4) 1.433 (4) 1.442 (4) 1.331 (4)
.356 (4) 1.359 (4) 1.303 (4) 1.441 (4) 1.441 (4) 1.324 (4)
1.345 (3) 1.376 (3) 1.298 (3) 1.440 (3) 1.448 (3) 1.330 (3)
.341 (3) 1.378 (3) 1.299 (3) 1.435 (3) 1.446 (3) 1.332 (3)
(
(
6)
7)
1
1
(
(
(
8)
9)
1.317 (2) 1.347 (2) 1.315 (2) 1.425 (2) 1.439 (2) 1.352 (2)
1.349 (2) 1.358 (2) 1.319 (2) 1.425 (2) 1.423 (2) 1.311 (2)
10) 1.355 (2) 1.337 (2) 1.335 (3) 1.378 (2) 1.393 (2) 1.313 (2)
g
h
j
k
l
Á
(
3)
4)
1.322 (2) 1.326 (2)
.319 (2) 1.324 (2)
1.312 (3) 1.329 (3)
1.344 (2) 1.277 (2) 1.339 (2) 0.067 (2)
1.347 (2) 1.275 (2) 1.335 (2) 0.072 (2)
1.352 (3) 1.275 (2) 1.337 (3) 0.077 (3)
1
(
(
(
5A) 1.325 (9) 1.310 (10) 1.369 (9) 1.264 (9) 1.346 (9) 0.105 (9)
5B) 1.314 (5) 1.330 (6)
.327 (5) 1.320 (5)
1.327 (4) 1.334 (4)
.318 (4) 1.334 (4)
1.325 (3) 1.335 (3)
.324 (3) 1.335 (3)
1.346 (5) 1.277 (4) 1.338 (4) 0.069 (5)
1.345 (5) 1.276 (4) 1.340 (4) 0.069 (5)
1.336 (4) 1.289 (3) 1.335 (4) 0.047 (4)
1.332 (4) 1.292 (3) 1.327 (4) 0.040 (4)
1.340 (3) 1.294 (3) 1.336 (3) 0.046 (3)
1.345 (3) 1.288 (3) 1.339 (3) 0.057 (3)
1.358 (2) 1.272 (2) 1.332 (2) 0.086 (2)
1.367 (2) 1.251 (2) 1.331 (2) 0.116 (2)
1
(
(
6)
7)
1
1
at the same time, the alternative formulation (C) can be ruled
out by consideration of the relative magnitudes of the bond
lengths for c and d.
(
(
(
8)
9)
1.334 (2) 1.324 (2)
1.328 (2) 1.320 (2)
10) 1.346 (2) 1.356 (2)
±
±
1.352 (2)
±
By contrast, in (8), where the 2-substituent is a benzyloxy
group rather than an amino group, bond f is signi®cantly
longer than in (3)±(7), while bond h is somewhat shorter: the
previously observed patterns for bonds b and c, and for j and
k, persist. On the other hand, in (9) where the additional
benzyloxy substituent is at the 4-position, bond a is much
longer than usual, while bond f is short: the b/c and j/k bond
pairs show the usual behaviour. Overall these patterns of bond
distances point to the polar forms (D) and (E) as the dominant
contributors to the electronic structures of (8) and (9),
respectively.
Substituents X2 and X4 are as de®ned in Scheme 2; bond g in (8) and bond h in (9) and
10) are CÐO rather than CÐN; the NO substituent is absent in (10).
(
at both the 2- and 4-positions, and then those with a second
benzyloxy substituent at either the 2- or the 4-position.
Firstly, in (3)±(7) the sequence of CÐN bonds labelled a, f,
g and h in Table 6 show very little variation in length between
those which are formally single (exterior to the rings) and
those which might be either double or aromatic delocalized
(
those within the rings). Secondly, bonds b and f are of similar
6
ꢀ
308 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
In (10) there is no nitroso group and hence no possibility of
charge-separated polar forms analogous to (A), (D) and (E):
to that extent (10) acts as a reference point against which to
judge the validity of the deductions made above concerning
Table 7
Hydrogen-bond geometry (A, ).
Ê
ꢂ
DÐHÁ Á ÁA
3)
HÁ Á ÁA DÁ Á ÁA
DÐHÁ Á ÁA Motif Direction
(
(
(
3)±(9). The conformation adopted by the benzyloxy groups in
10) precludes the obvious possibility of twofold rotational
N14ÐH14Á Á ÁO15
N24ÐH24Á Á ÁO25
N12ÐH12BÁ Á ÁN15
N22ÐH22BÁ Á ÁN25
1.97
1.96
2.34
2.12
2.630 (2) 131
2.636 (2) 133
3.177 (2) 166
3.063 (2) 175
S(6)
S(6)
C(7)
C(7)
±
±
i
[010]
[010]
symmetry: it is found (Table 6) that bonds a, c and e are longer
than bonds b, f and d, respectively, indicating beyond question
that the canonical form (F) is the dominant contributor, rather
than form (G). It is, moreover, signi®cant that the exocyclic
bond g is longer in (10) than in any of (3)±(8), where this bond
always forms part of a delocalized ®ve-centre, six-electron
system.
ii
(
4)
N4ÐH4Á Á ÁO5
1.96
2.12
2.623 (2) 131
2.987 (2) 171
S(6)
C(7)
±
[100]
iii
N2ÐH2AÁ Á ÁN5
(
5A)
N4ÐH4Á Á ÁO5
1.93
2.40
2.18
2.603 (9) 133²
3.223 (10) 156²
3.056 (9) 171
S(6)
2
±
³
³
iv
N4ÐH4Á Á ÁO5
R
R ꢁ8†
ꢁ4†
2
v
2
N2ÐH2AÁ Á ÁN3
2
(
5B)
N14ÐH14Á Á ÁO15
1.95
1.93
2.17
2.14
2.628 (4) 133
2.616 (4) 134
3.042 (5) 174
3.023 (4) 176
S(6)
S(6)
C(7)
C(7)
±
±
N24ÐH24Á Á ÁO25
iii
N12ÐH12AÁ Á ÁN15
[100]
[100]
vi
N22ÐH22AÁ Á ÁN25
(
6)
N14ÐH14Á Á ÁO15
N24ÐH24Á Á ÁO25
N22ÐH22BÁ Á ÁN15
N22ÐH22BÁ Á ÁO16
N22ÐH22AÁ Á ÁN25
1.93
1.96
2.48
2.44
2.22
2.17
2.59
2.15
2.46
2.606 (3) 132
2.621 (4) 131
3.302 (4) 155§
3.115 (3) 133§
2.925 (4) 137}
3.043 (4) 167}
3.165 (4) 123²²
2.962 (4) 153²²
3.199 (4) 142
S(6)
S(6)
D
±
±
±
±
D
vi
vi
C(7)
C(8)
C(7)
[100]
[100]
[010]
[010]
N22ÐH22AÁ Á ÁO25
vii
N12ÐH12AÁ Á ÁN15
N12ÐH12AÁ Á ÁO15
vii
C(8)
2
viii
N12ÐH12BÁ Á ÁO26
C ꢁ12† [110]
2
(
7)
N14ÐH14Á Á ÁO15
N24ÐH24Á Á ÁO25
N22ÐH22Á Á ÁO15
N12ÐH12Á Á ÁO25
1.94
1.97
2.07
2.09
2.605 (3) 131
2.626 (2) 131
2.853 (3) 148
2.875 (2) 149
S(6)
S(6)
±
±
±
D
ix
2
2
C ꢁ16† [101]
(
8)
N4ÐH4Á Á ÁO5
1.94
2.40
2.606 (2) 132
3.304 (2) 159
S(6)
C(12) [101]
±
x
ꢀ
C23ÐH23Á Á ÁO5
(
9)
Figure 11
Part of the crystal structure of (8), showing the formation by the CÐ
xi
N2ÐH2Á Á ÁO5
2.22
2.28
3.099 (2) 173³³
3.023 (2) 143³³
C(8)
C(7)
[010]
[010]
xi
N2ÐH2Á Á ÁN5
ꢀ
HÁ Á ÁO hydrogen bonds of a C(12) chain along [101]. The atoms marked
(
10)
xii
2
with an asterisk (*) or hash (#) are at the symmetry positions
‡ x; � y; ‡ z) and ( ‡ x; � y; � ‡ z), respectively.
N2ÐH2AÁ Á ÁN1
2.13
2.27
3.002 (2) 170
3.013 (2) 141
R ꢁ8†
±
[001]
2
1
2
1
2
1
2
1
2
1
2
1
2
xiii
(�
N2ÐH2BÁ Á ÁO4
C(6)
(
11)
N12ÐH12BÁ Á ÁN21
2.30
2.24
2.33
2.02
1.79
2.29
1.77
2.24
3.163 (4) 166
3.107 (4) 169
3.059 (4) 140
2.870 (4) 163
2.779 (4) 171}}
3.184 (5) 149}}
2.769 (4) 176³³³
3.109 (5) 145³³³
D
D
§§
§§
±
±
²²²
±
N22ÐH22BÁ Á ÁN11
xiv
2
N12ÐH12AÁ Á ÁO15A
xv
R ꢁ16†
2
N22ÐH22AÁ Á ÁO1
D
D
O1ÐH11Á Á ÁO15A
xv
O1ÐH11Á Á ÁN15A
D
xvi
4
O1ÐH12Á Á ÁO25A
R ꢁ20†
±
±
4
xvi
2
O1ÐH12Á Á ÁN25A
R ꢁ18†
2
3
2
1
2
1
2
5
2
1
2
1
2
Symmetry codes: (i) � x; � ‡ y; � z; (ii) � x; ‡ y; � z; (iii) 1 ‡ x; y; z; (iv)
�
�
x; 1 � y; � z; (v) 2 � x; 1 � y; 1 � z; (vi) � 1 ‡ x; y; z; (vii) x; � 1 ‡ y; z; (viii)
1
1
1
1
3
1 ‡ x; � 1 ‡ y; z; (ix) 1 ‡ x; y; 1 ‡ z; (x) � ‡ x; � y; ‡ z; (xi) 1 � x; � ‡ y; � z;
2
2
2
(xiv)
2
2
1
2
1
2
(
1
2
xii)
�
1 � x; y; � � z;
(xiii)
x; 1 � y; � ‡ z;
� x; 2 � y; � z; (xv)
system: sum
1
2
1
2
x; � ‡ y; � z; (xvi) 1 � x; 2 � y; � z. ² Three-centre NÐHÁ Á Á(O)
2
ꢂ
2
2
2
2
2
2
of angles at H4, 359 . ³ Secondary network C ꢁ12† [R ꢁ4†][R ꢁ8†] parallel to
[
201]. § Three-centre NÐHÁ Á Á(N,O) system: sum of angles at H22B,
ꢂ
ꢂ
ꢂ
ꢂ
3
3
3
3
53 . } Three-centre NÐHÁ Á Á(N,O) system: sum of angles at H22A,
38 . ²² Three-centre NÐHÁ Á Á(N,O) system: sum of angles at H12A,
06 . ³³ Three-centre NÐHÁ Á Á(N,O) system: sum of angles at H2,
2
48 . §§ Secondary network R ꢁ8†. }} Three-centre NÐHÁ Á Á(N,O) system: sum of
2
ꢂ
3
3
angles at H2, 354 . ²²² Secondary network C ꢁ10† parallel to [010]. ³³³ Three-
ꢂ
centre NÐHÁ Á Á(N,O) system: sum of angles at H2, 355 .
It is interesting to note that the general pattern of the bond
lengths in the 5-nitrosopyrimidines (3)±(9) reported here,
although highly unusual, is in fact entirely consistent with the
pattern observed (Low et al., 2000) in N-(6-amino-3,4-dihydro-
Figure 12
Part of the crystal structure of (4), showing the formation of a C(7) chain
along [100]. The atoms marked with an asterisk (*) or hash (#) are at the
symmetry positions (1 ‡ x; y; z) and (� 1 ‡ x; y; z), respectively.
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 309

research papers
3
-methyl-5-nitroso-4-oxopyrimidin-2-yl) derivatives (12) of a
3.4. Supramolecular aggregation via hydrogen bonding
.4.1. Hard hydrogen bonds form finite, zero-dimensional
3
aggregates. In (8) (Fig. 7) the sole hard hydrogen bond present
is the intramolecular NÐHÁ Á ÁO hydrogen bond forming the
usual S(6) motif. There are no other hard hydrogen-bond
donors in the molecule: however, uniquely in this series, and
possibly associated with the absence of hard intermolecular
hydrogen bonds, there is a signi®cant soft CÐHÁ Á ÁO hydrogen
bond (Table 7). Aromatic C23 at (x; y; z) acts as a hydrogen-
1
1
1
bond donor to nitroso O5 at (� ‡ x; � y; ‡ z), thus
2
2
2
ꢀ
producing a C(12) chain running parallel to the [101] direction
and generated by the n-glide plane at y = 0.25 (Fig. 11). Thus,
although the hard hydrogen bond is only intramolecular, the
soft hydrogen bond generates a one-dimensional structure.
3.4.2. Hard hydrogen bonds generate one-dimensional
structures.
range of amino acids, which, apart from carrying an alkyl
group on N rather than on O, closely resemble the amino-acid
derivatives (4)±(6).
That similar bond polarization occurs in both types of
compound con®rms the deductions made both here and
earlier (Low et al., 2000) that the enol-imidic [Ð
Simple chains. Compounds (4) and (9) (Figs. 2 and 8) both
0
ꢀ
crystallize with Z = 1, in space groups P1 and P2 /c, respec-
1
N
C(OCH Ph)Ð] and amidic [Ð(CH )NÐC( O)Ð] frag-
3
tively. Compound (4) contains the common S(6) motif
generated by the intramolecular NÐHÁ Á ÁO hydrogen bond
and, in addition, N2 at (x; y; z) acts as a hydrogen-bond donor
2
ments do not participate signi®cantly in the bond-polarization
process.
Figure 13
Part of the crystal structure of (9), showing the formation of a C(7)/C(8)
Figure 14
2
2
Part of the crystal structure of (7), showing the formation of a C ꢁ16†
chain along [010]. The atoms marked with an asterisk (*) or hash (#) are
3
chain along [101]. The atoms marked with an asterisk (*) or hash (#) are
at the symmetry positions (1 ‡ x; y; 1 ‡ z) and (� 1 ‡ x; y; � 1 ‡ z),
respectively.
1
3
2
1
2
at the symmetry positions (1 � x; � 2 ‡ y; � z) and (1 � x; ‡ y; � z),
2
respectively.
6
ꢀ
310 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
to nitroso N5 at (1 ‡ x; y; z), thus generating by translation a
C(7) chain running parallel to [100] (Fig. 12). The chain
formation in (9) involves the nitroso O, rather than N, as an
acceptor: N2 at (x; y; z) acts as a donor to O5 at
Paired chains. Compound (3) and polymorph (5B) of (5)
0
crystallize with Z = 2 in space groups P2 /n and P2 , respec-
1
1
tively, and in both structures each independent molecule
forms a C(7) chain. Thus, in (3) molecules of type 1 (Fig. 1)
1
3
1
3
(
1 � x; � ‡ y; � z), while N2 at (1 � x; � ‡ y; � z) in
form a C(7) chain parallel to [010] generated by the 2 axis
1
2
2
2
2
3
1
turn acts as a donor to O5 at (x; � 1 ‡ y; z), thus producing a
along ( , y, ), while type 2 molecules form an antiparallel
4 2
1
4
C(8) chain parallel to [010] generated by the 2 screw axis
1
chain generated by the 2 axis along (1.25, y, ): N12 and N22
1
1
3
along ( , y, ) (Fig. 13).
at (x; y; z) act as hydrogen bond donors, respectively, to
2
4
0
ꢀ
3
1
1
By contrast, (7) (Fig. 6) has Z = 2 in space group P1, but the
chain formation is very similar to that in (9), in that the two
independent molecules are linked by two NÐHÁ Á ÁO hydrogen
bonds each having a nitroso O atom as an acceptor. Within the
asymmetric unit (Fig. 6) N22 acts as a hydrogen-bond donor to
nitroso N15 and N25 at ( � x; � ‡ y; � z) and
2
2
2
5 1 1
( � x; ‡ y; � z) (Fig. 15).
2
2
2
The chain formation in (5B) is very similar, except only that
the chains are generated by translation along [100]: N12 and
N22 at (x; y; z) (Fig. 4) act as hydrogen-bond donors, respec-
tively, to N15 at (1 ‡ x; y; z) and to N25 at (� 1 ‡ x; y; z) (Fig.
16). There are no hydrogen bonds between the two types of
chain in either (3) or (5B).
A chain of rings. There are two intermolecular hydrogen
bonds in polymorph (5A) of (5) (Fig. 3), in addition to the
intramolecular NÐHÁ Á ÁO hydrogen bond having nitroso O as
an acceptor. This intramolecular bond is, in fact, one compo-
nent of a planar three-centre system, in which N4 is the
hydrogen-bond donor and where the two acceptors are nitroso
O5 within the same molecule at (x; y; z), and in the molecule
at (� x; 1 � y; � z). The net result is the formation of the fused
tricyclic hydrogen-bond system, in which the central ring is a
O15: in similar fashion, N12 at (x; y; z) acts as a donor to O25
2
2
at (1 ‡ x; y; 1 ‡ z), thus generating by translation a C ꢁ16†
chain running parallel to [101] (Fig. 14). In each of (4), (7) and
(
9) there are two chains running through each unit cell, related
to one another by the centres of inversion and hence anti-
parallel.
2
1
2
rather unusual R ꢁ4† motif centred at (0, , 0) (Fig. 17). In
2
addition, N2 in the molecule at (x; y; z) acts as a hydrogen-
bond donor, via H2A, to N3 at (2 � x; 1 � y; 1 � z), thus
2
generating a second centrosymmetric ring, this time of R ꢁ8†
2
1
2
1
2
type, centred at (1, , ) (Fig. 17). Propagation of these two
2
2
2
2
2
centrosymmetric motifs generates a C ꢁ12† [R ꢁ4†] [R ꢁ8†]
2
chain of rings (Bernstein et al., 1995) running parallel to the
201] direction (Fig. 17). It is interesting to note that H2B,
[
Figure 15
Stereoview of part of the crystal structure of (1), showing the formation of
antiparallel C(7) chains along [010].
Figure 17
Part of the crystal structure of (5A), showing the formation of a chain of
rings along [201]. The atoms marked with an asterisk (*), hash (#) or
dollar sign ($) are at the symmetry positions (2 � x; 1 � y; 1 � z),
(� x; 1 � y; � z) and (2 ‡ x; y; 1 ‡ z), respectively.
Figure 16
Stereoview of part of the crystal structure of (5B), showing the formation
of antiparallel C(7) chains along [100].
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 311

research papers
bonded to N2, plays no part in the hydrogen bonding and that
2
in the molecule at (x; y; z) acts as a hydrogen-bond donor, via
1
the R ꢁ4† motif is apparently preferred to the formation of an
H2A, to ring N1 in the molecule at (1 � x; y; � � z), while
2
2
1
2
NÐHÁ Á ÁO hydrogen bond involving H2B.
N2 at (1 � x; y; � � z) in turn acts as a donor to N1 at
2
2
A molecular ladder. Compound (10) is unique amongst the
compounds studied here in that the pyrimidine ring carries no
nitroso group (Fig. 9). There are just two intermolecular
hydrogen bonds, one each of NÐHÁ Á ÁN and NÐHÁ Á ÁO types,
and the combined effect of these when propagated in space
group C2/c is the generation of a molecular ladder. Amino N2
(x; y; z), thus producing an R ꢁ8† ring generated by the
1
2
1
4
twofold rotation axis along ( , y, � ) (Fig. 18). The same N2 at
(x; y; z) also acts as a hydrogen-bond donor, this time via H2B,
1
2
1
2
to O4 at (x; 1 � y; � ‡ z), while N2 at (x; 1 � y; � ‡ z) acts
as a donor to O4 at (x; y; � 1 ‡ z), thus producing a C(6) chain
running parallel to [001] and generated by the c-glide plane at
y = 0.5 (Fig. 19).
The combination of these two motifs generates a molecular
ladder, in which two antiparallel C(6) chains, related to one
another by the twofold rotation axes, form the uprights and
2
2
pairs of molecules linked by the R ꢁ8† rings act as the rungs;
the scheme below emphasizes the interplay of hydrogen-
bonding and benzyl conformation.
Figure 18
Part of the crystal structure of (10) showing the formation of a dimer lying
across a twofold rotation axis. The atoms marked with an asterisk (*) are
1
at the symmetry position (1 � x; y; � 2 � z).
4
2
Between these rungs there are R ꢁ16† rings: the R ꢁ8† rings lie
4
2
1
2
1
4
across the rotation axes along ( , y, n/2 + ) (n = zero or
1
4
integer) and the R ꢁ16† rings lie across inversion centres at ( ,
4
2
1
2
1
,
n/2 + ) (n = zero or integer). There are two such ladders
2
passing through each unit cell, one in the domain 0.26 < x <
.74 and the other in the domain 0.76 < x < 1.24, but there are
no hydrogen bonds between neighbouring ladders.
.4.3. Hard hydrogen bonds generate a two-dimensional
0
3
structure. The crystal structure of (6) (Fig. 5), which crystal-
0
lizes in space group P1 with Z = 2, is characterized by a large
number of hydrogen bonds, both two-centred and three-
centred, which link the two independent molecules into a two-
dimensional network. Each of the molecules exhibits the usual
S(6) intramolecular NÐHÁ Á ÁO hydrogen bond with the
nitroso O as an acceptor. It is striking that the two inter-
Figure 19
Part of the crystal structure of (10) showing the formation of a C(6) chain
parallel to [001]. The atoms marked with an asterisk (*), hash (#) or dollar
1
1
sign ($) are at the symmetry positions (x; 1 � y; ‡ z), (x; 1 � y; � ‡ z)
2
2
and (x; y; 1 ‡ z), respectively.
6
ꢀ
312 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
molecular hydrogen bonds formed by the amino group in the
type 1 molecule are both two-centre NÐHÁ Á ÁO interactions,
whereas each NÐH bond of the amino group in the type 2
molecule participates in a three-centre system with one N and
one O acceptor in each case.
Finally, in a further motif which involves both types of
molecule, N12 in the type 1 molecule at (x; y; z) acts as a
donor, via H12B, to O26 in the type 2 molecule at
2
(1 ‡ x; 1 ‡ y; z), thus generating a C ꢁ12† chain parallel to
2
[110] (Fig. 22). The effect of this motif is to subdivide the
8
4
4
6
6
The type 1 molecules are linked into a C(8) chain parallel to
010] via a two-centre hydrogen bond, while the type 2
R ꢁ44† rings of the (001) sheet into R ꢁ16† and R ꢁ34† sectors,
8
[
and thereby to strengthen the sheet.
molecules are linked into a C(8) chain parallel to [100] via a
three-centre hydrogen bond. Amino N12 in the type 1 mole-
cule at (x; y; z) acts as a hydrogen-bond donor, via H12A, to
O15 and weakly to N15 in the type 1 molecule at (x; � 1 ‡ y; z)
(
Fig. 20). In the type 2 molecules, amino N22 at (x; y; z) acts as
a hydrogen-bond donor, via H22A, to both N25 and O25 at
(
� 1 ‡ x; y; z): of the two individual components of this three-
centre system, that involving O25 is much more nearly linear
and counting via O25 leads to a C(8) motif (Fig. 21).
The two sets of chains, parallel to [010] and [100], are linked
into a continuous two-dimensional network by a three-centre
hydrogen bond which links the two independent molecules
within the asymmetric unit (Fig. 5), where amino N22 in the
type 2 molecule acts as a donor, via H22B, to both N15 and
O16 in the type 1 molecule. In this three-centre system, the
interaction with the N acceptor is the more nearly linear, in
contrast to that formed by H22A. The combination in this
manner of the two types of chain generates a sheet parallel to
8
(
the pathway is counted via the O acceptor.
001) built from R ꢁ44† rings, where in all three-centre systems
8
Figure 21
Part of the crystal structure of (6) showing the formation of a chain along
100] comprising type 2 molecules only. For the sake of clarity, H atoms
bonded to C are omitted. The atoms marked with an asterisk (*) or hash
[
(#) are at the symmetry positions (1 ‡ x; y; z) and (� 1 ‡ x; y; z),
respectively.
Figure 22
Stereoview of part of the crystal structure of (6) showing the formation of
2
2
a C ꢁ12† chain along [110] involving both types of molecule. For the sake
of clarity, H atoms bonded to C are omitted.
Figure 20
Part of the crystal structure of (6) showing the formation of a chain along
[010] comprising type 1 molecules only. For the sake of clarity, H atoms
bonded to C are omitted. The atoms marked with an asterisk (*) or hash
Figure 23
Part of the crystal structure of (11) showing the formation of a
2
2
centrosymmetric six-molecule aggregate containing R ꢁ8† and R ꢁ16†
2
2
rings. For the sake of clarity, only the major components are shown and H
atoms bonded to C are omitted. The atoms marked with an asterisk (*)
are at the symmetry position (� x; 2 � y; � z).
(
#) are at the symmetry positions (x; 1 ‡ y; z) and (x; � 1 ‡ y; z),
respectively.
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 313

research papers
3
.4.4. Hard hydrogen bonds generate a three-dimensional
As noted above (x2.2), several analogues of (11) containing
different dialkylamino substituents have been examined in an
attempt to ®nd an analogue whose structure is not bedevilled
by multiple disorder, but none has been obtained in suitably
crystalline form; this particular series seems particularly
reluctant to form appropriately crystalline solids. Thus, to
date, the structure of (11) is the sole representative of this
class.
structure. In (11) both of the nitroso groups are independently
disordered each over two sets of sites (see x2.2), but only the
major sites will be considered in the following analysis of the
hydrogen bonding. Similarly, the minor-occupancy water
molecule will not be considered further in this context. Subject
to these constraints, the major components are linked into a
three-dimensional framework.
Within the asymmetric unit of (11), N12 and N22 act as
hydrogen-bond donors, via H12B and H22B, respectively, to
3.4.5. General comments on the hydrogen bonding. The
hydrogen bonding observed in the nitroso (3)±(9) (Table 7,
Figs. 11±17 and 20±22) is striking for the distribution of the
hydrogen-bond donors and acceptors, as well as for the
frequent occurrence of nearly planar three-centre DÐ
2
N11 and N21, giving an R ꢁ8† motif (Fig. 23) having approx-
2
imate but not precise twofold rotational symmetry. In addi-
tion, N12 at (x; y; z) acts as a hydrogen-bond donor, via H12A,
2
2
to O15A at (� x; 2 � y; � z), thus generating an R ꢁ16† ring
HÁ Á Á(A) systems in (5A), (6), (9) and (11). While the donors
2
centred at (0, 1, 0), while N22 at (x; y; z) acts as a donor, via
H22A, to O1 also at (x; y; z). In this manner a centrosym-
metric six-molecule aggregate containing three rings is formed
are exclusively the amino N substituents at positions 2- and/or
4- of the pyrimidine ring, the dominant acceptors are the N
and/or O atoms of the nitroso groups. Only in polymorph
(5A), where a ring N acts as an acceptor (Fig. 17), and in (6),
where the O atoms at position 6 act as acceptors in both of the
independent molecules, are any other acceptors involved. This
general trend is entirely consistent with the polarized form
adopted by all the nitroso derivatives (3)±(9) and hence most
of the hard hydrogen bonds noted in Table 7 are charge
assisted (Gilli et al., 1994). Since few of the intermolecular
hydrogen bonds have short DÁ Á ÁA distances, it may well be
that in the absence of the charge assistance induced by the
polarized electronic structures, the intermolecular hydrogen
bonding would be considerably less extensive.
(
Fig. 23).
The water molecule containing O1 acts as a hydrogen-bond
donor in two independent three-centre systems (Table 7): O1
at (x; y; z) acts as a donor, via H11, to both O15A and N15A at
1
2
1
2
1
2
3
(
� x; � ‡ y; � z), forming a C ꢁ10† chain (counting via
3
O15A, as this has the stronger interaction) running parallel to
010] (Fig. 24); O1 at (x; y; z) also acts as a donor via H12 to
[
both O25A and N25A at (1 � x; 2 � y; � z), thus generating
4
an R ꢁ20† ring (again counting via O, as this has the stronger
4
1
interaction) centred at ( , 1, 0) (Fig. 25). These two hydrogen
bonds link each six-molecule aggregate (Fig. 23) to six others:
2
3
thus, the aggregate centred at (0, 1, 0) is linked via the C ꢁ10†
3
1
2
1
2
1
2
1
2
3
4
1
2
1
2
1
2
1
2
chain motif to those centred at ( , , ), ( , , ), (� , , � ) and
4
. Conclusions
1
2
3
2
1
2
(
interaction generates a (101) sheet. Similarly, the aggregate
� , , � ) so that propagation by the space group of this
ꢀ
The structures reported here for nitrosopyrimidines (3)±(9) all
exhibit strong polarization of the electronic structures, as
4
4
centred at (0, 1, 0) is linked via the R ꢁ20† ring motif to those
centred at (1, 1, 0) and (� 1, 1, 0), thus generating a chain of
2
2
2
2
4
4
fused R ꢁ8†, R ꢁ16† and R ꢁ20† rings parallel to [100]. Propa-
gation of these interactions and the combination of the [100]
ꢀ
chain with the (101) sheet generates a continuous three-
dimensional framework.
Figure 24
Figure 25
Part of the crystal structure of (11) showing the formation of a
3
3
Part of the crystal structure of (11) showing the formation of a C ꢁ10†
4
chain along [010]. For the sake of clarity, only the major components are
shown and H atoms bonded to C are omitted. The atoms marked with an
asterisk (*) or hash (#) are at the symmetry positions
centrosymmetric four-molecule aggregate containing an R ꢁ20† ring.
4
For the sake of clarity, only the major components are shown and H
atoms bonded to C are omitted. The atoms marked with an asterisk (*)
are at the symmetry position (1 � x; 2 � y; � z).
1
2
1
2
1
2
(
� x; � ‡ y; � z) and (x; � 1 ‡ y; z), respectively.
6
ꢀ
314 Quesada et al.
O -Benzyl-5-nitrosopyrimidines
Acta Cryst. (2002). B58, 300±315

research papers
found earlier (Low et al., 2000) for a series of related dihy-
dropyrimidinones. Although the intramolecular distances are
unusual in comparison with those in simple nitroso-arenes and
in analogous pyrimidines carrying no nitroso substituent, a
simple polarization model is consistent with all the present
observations. At the same time, in (11), although the structure
is heavily disordered, it is clear that the presence of two
secondary amino substituents adjacent to the nitroso group is
suf®cient to cause severe non-planarity, with a consequent
reduction in electronic delocalization.
Flack, H. D. & Bernardinelli, G. (2000). J. Appl. Cryst. 33, 1143±
1148.
Friedman, H. S., Kokkinakis, D. M., Pluda, J., Friedman, A. H.,
Cokgor, I., Haglund, M. M., Ashley, D. M., Rich, J. N., Dolan, M. E.,
Pegg, A. E., Moschel, R. C., McLendon, R. E., Kerby, T., Herdon,
J. E., Bigner, D. D. & Scold, S. C. (1998). J. Clin. Oncol. 16, 3570±
3575.
Gilli, P., Bertolasi, V., Ferretti, V. & Gilli, G. (1994). J. Am. Chem. Soc.
116, 909±915.
Lister, J. H. (1996). The Purines, Suppl. 1; The Chemistry of
Heterocyclic Compounds, edited by E. C. Taylor, Vol. 54, pp. 44±
50. New York: John Wiley and Sons.
Low, J. N., L o pez, M. D., Arranz Mascar o s, P., Cobo Domingo, J.,
Godino, M. L., L o pez Garz o n, R., Guti e rrez, M. D., Melguizo, M.,
Ferguson, G. & Glidewell, C. (2000). Acta Cryst. B56, 882±
X-ray data were collected at the EPSRC X-ray Crystal-
lographic Service, University of Southampton, England, using
a Nonius Kappa-CCD diffractometer: the authors thank the
staff for all their help and advice. AQ thanks the Ministry of
Education (Spain) for funding in support of his Semester of
study in the University of Dundee, and JNL thanks NCR Self-
Service, Dundee, for grants which have provided computing
facilities for this work.
892.
Marchal, A., S a nchez, A., Nogueras, M. & Melguizo, M. (1998). 18th
European Colloquium on Heterocyclic Chemistry, Rouen, France.
Abstract B-42.
Marchal, A., S a nchez, A., Nogueras, M. & Melguizo, M. (2000). 7th
Ibn Sina International Conference on Pure and Applied Hetero-
cyclic Chemistry, Alexandr õÂ a, Egypt. Abstract P-214.
Nonius (1997). Kappa-CCD Server Software. Windows 3.11 Version,
Nonius B.V., Delft, The Netherlands.
Otwinowski, Z. & Minor, W. (1997). Methods Enzymol. 276, 307±
326.
Quesada, A., Marchal, A., S a nchez, A., Nogueras, M. & Melguizo, M.
(2000). XIXth European Colloquium on Heterocyclic Chemistry,
Aveiro, Portugal. Abstracts, p. 227.
References
Abril, N., Luque-Romero, F. L., Christians, F. C., Encell, L. P., Loeb,
L. A. & Pueyo, C. (1999). Carcinogenesis, 20, 2089±2094.
Bauer, S. H. & Andreassen, A. L. (1972). J. Phys. Chem. 76, 3099±
3
108.
Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995). Angew.
Chem. Int. Ed. Engl. 34, 1555±1573.
Blessing, R. H. (1995). Acta Cryst. A51, 33±37.
Blessing, R. H. (1997). J. Appl. Cryst. 30, 421±426.
Chae, M. Y., McDougall, M. G., Dolan, M. E., Swenn, K., Pegg, A. E.
Schlemper, E. O., Murmann, R. K. & Hussain, M. S. (1986). Acta
Cryst. C42, 1739±1743.
Sheldrick, G. M. (1997a). SHELXL97. University of G oÈ ttingen,
Germany.
Sheldrick, G. M. (1997b). SHELXS97. University of G oÈ ttingen,
Germany.
&
Moschel, R. C. (1995). J. Med. Chem. 38, 359±365.
Davis, M. I., Boggs, J. E., Coffey, D. & Hanson, H. P. (1965). J. Phys.
Chem. 69, 3727±3730.
Spek, A. L. (2001). PLATON. Version of May 2001. University of
Utrecht, The Netherlands.
Ferguson, G. (1999). PRPKAPPA. University of Guelph, Canada.
Flack, H. D. (1983). Acta Cryst. A39, 876±881.
Talberg, H. J. (1977). Acta Chem. Scand. A, 31, 485±491.
Wilson, A. J. C. (1976). Acta Cryst. A32, 994±996.
6
ꢀ
Acta Cryst. (2002). B58, 300±315
Quesada et al.
O -Benzyl-5-nitrosopyrimidines 315