Patterns of soft C - H...O hydrogen bonding in diaryl sulfones

Article abstract of DOI:10.1107/S0108768100015494

In bis(4-tolyl) sulfone, C₁₄H₁₄O₂S (1), 2,5,4′-trimethyldiphenyl sulfone, C₁₅H₁₆O₂S (2), and 4-chlorodiphenyl sulfone, C₁₂H₉ClO₂S (3), the molecules are lin

Full text of DOI:10.1107/S0108768100015494

research papers
Acta Crystallographica Section B
Structural
Science
Patterns of soft CÐHÁ Á ÁO hydrogen bonding in
diaryl sulfones
ISSN 0108-7681
a
0
Received 14 July 2000
Christopher Glidewell, *
In bis(4-tolyl) sulfone, C H O S (1), 2,5,4 -trimethyldiphenyl
2
1
4
14
b
Accepted 26 October 2000
William T. A. Harrison, John N.
sulfone, C15H O S (2), and 4-chlorodiphenyl sulfone,
16 2
b,c
d
C H ClO S (3), the molecules are linked by soft CÐHÁ Á ÁO
Low, Jamie G. Sime and
12
9
2
b
hydrogen bonds into three different types of one-dimensional
aggregate: simple chains in (1), molecular ladders in (2) and
James L. Wardell
0
chains of fused rings in (3). In each of 3,4-dimethyl-4 -
chlorodiphenyl sulfone, C H ClO S (4), and 2,5-dimethyldi-
a
School of Chemistry, University of St Andrews,
1
4
13
2
St Andrews, Fife KY16 9ST, Scotland,
phenyl sulfone, C H O S (5), the CÐHÁ Á ÁO hydrogen bonds
b
14 14
2
Department of Chemistry, University of
link the molecules into two different types of two-dimensional
sheet, based on a (4,4) net in (4) and a (3,6) net in (5). The
patterns of soft CÐHÁ Á ÁO hydrogen bonds in (1)Ð(5) are
compared with those in other diaryl sulfones, mainly retrieved
from the Cambridge Structural Database, whose substitution
patterns preclude the formation of hard hydrogen bonds.
Observed aggregation modes range from the formation of no
CÐHÁ Á ÁO hydrogen bonds at all, via ®nite (zero-dimensional)
arrays through one-, two- and three-dimensional systems.
Aberdeen, Meston Walk, Old Aberdeen AB24
c
3
UE, Scotland, Department of Applied Physics
and Electronic and Mechanical Engineering,
University of Dundee, Nethergate, Dundee DD1
d
4
HN, Scotland, and 3 West Shore, Pittenweem,
Fife KY10 2NU, Scotland
Correspondence e-mail: cg@st-andrews.ac.uk
1
. Introduction
The SO group in diaryl sulfones is a powerful acceptor of
2
hydrogen bonds (Ferguson et al., 1999). When hard (Braga et
al., 1995) hydrogen-bond donors, such as OH or NH groups,
2
are present the supramolecular aggregation is dominated
between these donors and the SO groups. Thus, for example,
2
in O S(C H OH) (Glidewell & Ferguson, 1996), the mole-
2
6
4
2
cules are linked by OÐHÁ Á ÁO S hydrogen bonds into pair-
4
4
wise-interwoven square nets built from R ꢁ32† rings
(Bernstein et al., 1995); in O S(C H NH ) , a combination of
2 6 4 2 2
NÐHÁ Á ÁO and NÐHÁ Á ÁN hydrogen bonds generates a
continuous three-dimensional framework (Bertolasi et al.,
1
993). When hard hydrogen-bond donors are absent, it may be
expected that CÐHÁ Á ÁO hydrogen bonds will occur instead,
and in the bis-sulfone 2,5-bis-(4-chlorophenylsulfonyl)-1,4-
dimethylbenzene, (ClC H SO ) C H (CH ) , the molecules
6
4
2 2
6
2
3 2
are linked into molecular ladders by CÐHÁ Á ÁO hydrogen
bonds (Wardell et al., 2000).
In the present paper, we report the crystal structures, with
particular emphasis on the supramolecular aggregation, of the
diaryl sulfones (1)±(5), whose molecular structures are all
designed to preclude the possibility of hard hydrogen bonding,
and we compare these with other, related examples (6)±(18),
mostly retrieved from the Cambridge Structural Database
(CSD; Allen & Kennard, 1993). In all of these compounds the
CÐH bonds are the sole possible donors in any hydrogen
bonds, but in several compounds there are potential hydrogen-
bond acceptors other than the SO group. Nonetheless, for
#
2001 International Union of Crystallography
2
Printed in Great Britain ± all rights reserved
very few of these compounds did the original reports make
ꢀ
190 Glidewell et al.
Patterns of hydrogen bonding
Acta Cryst. (2001). B57, 190±200

research papers
any mention of intermolecular aggregation by means of
CÐHÁ Á ÁO hydrogen bonding.
maps and all were included in the re®nements as riding atoms,
Ê Ê
with CÐH 0.93±0.95 A (aromatic), 0.96±0.98 A (methyl). In
(
1), (4) and (5) the methyl groups were all modelled using six
ꢂ
H sites with occupancy 0.5, offset from one another by 60
rotations: the two methyl groups in (2) could each be satis-
factorily modelled with a single set of sites. Supramolecular
analyses were made and the diagrams were prepared with the
aid of PLATON (Spek, 2000).
2. Experimental
2
.1. Syntheses
Samples of (1) and (3) were obtained from Aldrich
Chemical Company. Compound (2) was prepared by reaction
of 4-toluenesulfonyl chloride with excess 1,4-dimethylbenzene
in the presence of AlCl : the product was isolated by chro-
3
matography and recrystallized from ethanol, m.p. 383±385 K
(
literature m.p. 381±385 K; Burton & Praill, 1955); (5) was
prepared by a similar method using benzenesulfonyl chloride
with m.p. 383±385 K (literature m.p. 384 K; Novi et al., 1982).
Compound (4) was prepared by oxidation of the corre-
sponding sul®de (Grant et al., 1987) using 3-chloroperoxy-
benzoic acid: after crystallization from ethanol (4) had m.p.
343±344 K. Crystals of (1)±(5) suitable for single-crystal X-ray
diffraction were grown from solutions in ethanol.
Figs. 1±5 show the molecular units in (1)±(5), and Figs. 6±19
show aspects of the crystal structures of (1)±(18): (6)±(18) are
identi®ed in Table 2. Table 3 gives details of the hydrogen
bonding in (1)±(18) and Table 4 contains details of the short
intermolecular halogenÁ Á Áhalogen contacts in (14) and (15).
2
.2. Data collection, structure solution and refinement
Details of cell data, data collection and re®nement are
summarized in Table 1 (Blessing, 1995, 1997; Bruker, 1997;
Ferguson, 1999; Nonius, 1997; Otwinowski & Minor, 1997;
1
Sheldrick, 1997a,b; Spek, 2000). For (1)±(5) the space groups
3
. Results and discussion
were uniquely assigned from the systematic absences: P2 /c
1
3.1. Isomorphous and isomeric relationships
for (1) and (3), P2 /n for (2), (4) and (5). The structures were
1
all solved by direct methods using SHELXS97 (Sheldrick,
1
Compounds (6) and (7) are also isomorphous, as are (8) and
9); however, for neither of these pairs was the isomorphism
2
997a) and re®ned with all data on F using SHELXL97
(
(
Sheldrick, 1997b). A weighting scheme based upon
2
noted in the latter of the two structure reports. As noted in the
original report (Sime & Woodhouse, 1974b), (13)±(15) are
isomorphous, with the molecules lying across twofold rotation
axes. Of the compounds newly studied here, (3) is isomor-
phous with (6) and (7). However, despite the isomorphous
nature of (8) and (9), (1) and (13) are not isomorphous. These
2
c
P ˆ ‰F ‡ 2F Š=3 was employed in order to reduce statistical
o
bias (Wilson, 1976). All H atoms were located from difference
1
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: BM0036). Services for accessing these data are described
at the back of the journal.
ꢀ
Acta Cryst. (2001). B57, 190±200
Glidewell et al.
Patterns of hydrogen bonding 191

research papers
Table 1
Experimental details.
(1)
(2)
(3)
(4)
(5)
Crystal data
Chemical formula
Chemical formula
weight
C H O S
14
246.31
C H O S
2
260.34
C H ClO S
2
252.7
C H ClO S
2
280.75
C H O S
2
246.31
1
4
2
15 16
12
9
14 13
14 14
Cell setting, space
group
Monoclinic, P2 =c
Monoclinic, P2 =n
Monoclinic, P2 =c
Monoclinic, P2 =n
Monoclinic, P2 =n
1
1
1
1
1
Ê
a, b, c (A)
14.1773 (7), 7.7437 (4),
2
8.3374 (6), 14.9785 (8),
10.9064 (6)
98.241 (3)
1347.95 (14)
4
1.283
Mo Kꢁ
12 210
13.239 (3), 7.4521 (15),
12.9682 (14), 7.5293 (12),
7.8879 (3), 10.9788 (4),
15.2489 (4)
103.525 (2)
1283.93 (8)
4
1.274
Mo Kꢁ
14 179
4.0161 (12)
11.737 (2)
93.17 (3)
1156.3 (4)
4
1.452
Mo Kꢁ
3848
14.4627 (17)
110.625 (7)
1321.6 (3)
4
1.411
Mo Kꢁ
9255
ꢂ
ꢀ
V (A )
ꢁ †
101.205 (1)
2586.3 (2)
8
1.265
Mo Kꢁ
15 872
Ê
3
Z
D (Mg m
�
3
)
Radiation type
x
No. of re¯ections for
cell parameters
ꢂ
ꢂ range ( )
ꢃ (mm
1.46±25.03
0.237
306 (2)
Needle, colourless
0.56 Â 0.18 Â 0.13
2.32±27.39
0.231
150 (2)
Plate, colourless
0.2 Â 0.2 Â 0.025
1.54±25.0
0.491
307 (2)
Block, yellow
0.40 Â 0.28 Â 0.24
1.81±25.85
0.437
150 (2)
Block, colourless
0.10 Â 0.08 Â 0.05
2.30±27.32
0.239
150 (2)
Plate, colourless
0.20 Â 0.15 Â 0.05
�
1
)
Temperature (K)
Crystal form, colour
Crystal size (mm)
Data collection
Diffractometer
Bruker SMART
000CCD
Kappa-CCD
Bruker SMART
1000CCD
Kappa-CCD
Kappa-CCD
1
Data collection
method
' and ! scans with ꢄ
offsets
' and ! scans with ꢄ
offsets
' and ! scans with ꢄ
offsets
' and ! scans with ꢄ
offsets
' and ! scans with ꢄ
offsets
Absorption correction
Tmin
Tmax
Multi-scan
0.879
0.970
Multi-scan
0.929
0.969
Multi-scan
0.828
0.891
Multi-scan
0.958
0.979
Multi-scan
0.954
0.988
No. of measured,
independent and
observed para-
meters
15 872, 4554, 2746
12 210, 3060, 2172
7000, 2030, 1527
9255, 2545, 1332
14 179, 2895, 1990
Criterion for observed
re¯ections
Rint
ꢂmax ( )
Range of h, k, l
I > 2ꢅ(I)
I > 2ꢅ(I)
I > 2ꢅ(I)
I > 2ꢅ(I)
I > 2ꢅ(I)
0.023
25.03
0.037
27.39
� 10 ! h ! 10
� 18 ! k ! 19
� 14 ! l ! 14
0.020
25.00
� 15 ! h ! 15
� 8 ! k ! 8
� 8 ! l ! 13
0.053
25.85
� 15 ! h ! 15
� 9 ! k ! 9
� 17 ! l ! 17
0.038
27.32
� 10 ! h ! 10
� 14 ! k ! 14
� 19 ! l ! 19
ꢂ
� 16 ! h ! 16
�
�
9 ! k ! 9
28 ! l ! 22
Re®nement
Re®nement on
2
2
2
2
2
F
0.0399, 0.1009, 0.928
F
0.0466, 0.1173, 1.044
F
0.0386, 0.1076, 1.041
F
0.0591, 0.1575, 0.93
F
0.047, 0.1224, 1.044
2
2
R‰F >2ꢅꢁF †Š,
2
wRꢁF †, S
No. of re¯ections and
parameters used in
re®nement
4554, 311
3060, 166
2030, 145
2545, 165
2895, 156
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/[ꢅ (F ) +
w = 1/[ꢅ (F ) +
w = 1/[ꢅ (F ) +
w = 1/[ꢅ (F ) +
w = 1/[ꢅ (F ) +
o
2
o
2
o
2
o
2
o
2
(0.0511P) ], where
P = (F o + 2F )/3
(0.0551P) +
0.1788P], where
(0.0550P) +
0.2889P], where
(0.0760P) ], where
2
(0.0642P) +
0.0204P], where
2
2
c
2
c
P = (F o + 2F )/3
2
2
c
2
2
c
2
2
P = (F o + 2F )/3
0.005
0.24, � 0.438
P = (F o + 2F )/3
0.001
0.239, � 0.396
P = (F + 2F )/3
o c
ꢁ
Á=ꢅ†
0.000
0.142, � 0.23
0.003
0.342, � 0.283
0.004
0.258, � 0.422
max
Ê
� 3
)
Áꢆ , Áꢆmin (e A
max
Computer programs used: SMART (Bruker, 1997), SAINT (Bruker, 1997), SHELXS97 (Sheldrick, 1997a), SHELXL97 (Sheldrick, 1997b), SHELXL97 and WordPerfect macro
PRPKAPPA (Ferguson, 1999), Kappa-CCD server software (Nonius, 1997), DENZO (Otwinowski & Minor, 1997).
latter two compounds crystallize in different space groups, (1)
0
(see xx3.2.1 and 3.2.3), the structure of (15) contains CÐ
HÁ Á ÁO hydrogen bonds, while the structures of (13) and (14)
do not. Likewise (see x3.2.3) the hydrogen-bonding behaviour
of (3) differs from that of (6) and (7). Similar comments also
apply to other types of intermolecular interaction (see x3.2.7).
More substantial differences in hydrogen-bonding beha-
viour will be noted amongst pairs of isomeric, but not
in P2 /c with Z = 2 and (13) in the I2/a setting of C2/c with
1
0
Z = 0.5.
Even in compounds which are isomorphous, subtle differ-
ences in interatomic distances can mean that one compound of
an isomorphous pair may exhibit signi®cant soft CÐHÁ Á ÁO
hydrogen bonds, while the other does not. Thus, for example
ꢀ
192 Glidewell et al.
Patterns of hydrogen bonding
Acta Cryst. (2001). B57, 190±200

research papers
isomorphous compounds. Thus, (1) and (5) are isomeric, as are
compounds (2), (8) and (10), and (4) and (9). No two
compounds within any of these isomeric series behave in the
same manner (see x3.2).
symmetrically disposed; and (16)±(18) all contain nitro groups,
which are themselves potentially good hydrogen-bond
acceptors.
As Cotton et al. (1997) have noted, the criteria for the
acceptance of non-covalently bonded contacts as hydrogen
bonds have been undergoing a continuous relaxation in recent
years. It is therefore appropriate to de®ne the acceptance
criteria adopted here for the CÐHÁ Á ÁO hydrogen bonds which
are considered to be structurally signi®cant. The CÐH
distances in (1)±(5), determined here, are all derived from
X-ray diffraction data using the well established SHELXL97
values for riding CÐH atoms (Sheldrick, 1997b). Conse-
quently, since the structures of compounds (6)±(18) were all
also derived from X-ray diffraction data, the CÐHÁ Á ÁO
parameters in these compounds have all been derived from
the atomic coordinates deposited in the CSD: no attempt has
been made to convert the CÐH distances into the so-called
neutron-normalized distances. It is worthwhile to point out
here that neither X-ray nor neutron diffraction can provide
correct coordinates for an `H atom': the two types of diffrac-
tion data provide coordinates for different components of
atoms, which in the case of H are far from coincident with
neither set of coordinates being more, or less, `correct' than
the other (Cotton & Luck, 1989). The acceptance criteria for
CÐHÁ Á ÁO hydrogen bonds adopted here refer speci®cally to
structural parameters derived from X-ray diffraction data. The
CÐHÁ Á ÁO interactions in Table 3 all involve aromatic CÐH
bonds as donors and all are subject to the reasonably, but not
excessively conservative criteria of an upper limit (within
3
.2. Supramolecular structures
In (1)±(5) and (6)±(18) (Table 2), the CÐHÁ Á ÁO hydrogen
bonding exhibits a variety of patterns: in some compounds,
there are no CÐHÁ Á ÁO hydrogen bonds at all, while in others
these interactions lead to aggregates which can be zero-
dimensional (i.e. ®nite aggregates) or one-, two- or three-
dimensional, and it is convenient to treat each of these
variants in turn. Of the compounds (6)±(18) retrieved from the
CSD, (6) is the unsubstituted parent diphenyl sulfone; (7)±(11)
all contain methyl substituents, together with chloro in the
case of (9); (12)±(15) contain only halogen substituents,
Figure 1
The two independent molecules in (1), showing the atom-labelling
schemes: (a) molecule 1; (b) molecule 2. Displacement ellipsoids are
drawn at the 30% probability level and only one set of H atom sites is
depicted for each methyl group.
Figure 3
The asymmetric unit in (3), showing the atom-labelling scheme.
Displacement ellipsoids are drawn at the 30% probability level.
Figure 4
Figure 2
The asymmetric unit in (2), showing the atom-labelling scheme.
Displacement ellipsoids are drawn at the 30% probability level.
The asymmetric unit in (4), showing the atom-labelling scheme.
Displacement ellipsoids are drawn at the 30% probability level and only
one set of H atom sites is depicted for each methyl group.
ꢀ
Acta Cryst. (2001). B57, 190±200
Glidewell et al.
Patterns of hydrogen bonding 193

research papers
Ê
experimental uncertainty) of 3.50 A on the CÁ Á ÁO distance,
actions. The structure of (11) has been reported only from
photographic data and no H atoms were either located or
included in the re®nements; however, since there are no
ꢂ
combined with a lower limit on the CÐHÁ Á ÁO angle of 130
Ê
and an upper limit on the HÁ Á ÁO distance of 2.60 A. No
Ê
account has been taken of interactions involving methyl CÐH
bonds which fall within these geometric criteria, ®rstly because
methyl CÐH bonds are less acidic than aromatic CÐH bonds
and, more importantly, because methyl groups are likely to be
undergoing extensive libration, if not nearly free rotation
about their proximal CÐC bonds (Riddell & Rogerson, 1996,
intermolecular CÁ Á ÁO distances less than 3.50 A, it may
reasonably be inferred that there are no intermolecular CÐ
HÁ Á ÁO hydrogen bonds. In (8), (9) and (11), therefore, the
crystal structures contain isolated molecules. There are no CÐ
HÁ Á ÁO hydrogen bonds in the isomorphous pair of (13) and
(14), although in (14) there are short BrÁ Á ÁBr contacts (see
x3.2.7).
1
997). Any short contacts involving such groups are thus
unlikely to have real structural signi®cance.
.2.1. Structures with no intermolecular hydrogen bonds
within the acceptance criteria. Compounds (8) and (9) are
isomorphous, and the molecular volume in (8) exceeds that in
3.2.2. Finite (zero-dimensional) hydrogen bonding. There
is a single type of CÐHÁ Á ÁO hydrogen bond in (12): only a
single CÐH bond, in one ring only, acts as a hydrogen-bond
donor and only one of the O atoms acts as a hydrogen-bond
3
(
methyl and chloro substituents. In neither compound are there
any CÐHÁ Á ÁO hydrogen bonds or any ꢇÁ Á Áꢇ stacking inter-
9) by only ca 1.4%, re¯ecting the isosteric nature of the
acceptor. The CÐHÁ Á ÁO hydrogen bonding in (12) generates
2
2
a ®nite motif, a dimer containing a centrosymmetric R ꢁ12†
Figure 5
The asymmetric unit in (5), showing the atom-labelling scheme.
Displacement ellipsoids are drawn at the 30% probability level and only
one set of H atom sites is depicted for each methyl group.
Figure 7
Part of the crystal structure of (1), showing the formation of two
independent chains running parallel to [010]. Atoms are depicted as in
0
Fig. 6. The atoms marked with a prime ( ) or a hash (#) are both at the
symmetry position (x; 1 ‡ y; z).
Figure 6
Part of the crystal structure of (12), showing the formation of a
Figure 8
Part of the crystal structure of (2), showing the formation of a molecular
2
centrosymmetric R ꢁ10† ring. For the sake of clarity H atoms not involved
ladder running parallel to [100]. Atoms are depicted as in Fig. 6. The
0
2
in the hydrogen bonding motif shown are omitted. The atom marked with
a prime ( ) is at the symmetry position (� 1 � x; 1 � y; � z).
atoms marked with a prime ( ) or a star (*) are at the symmetry positions
(1 � x; 1 � y; � z) and (� 1 ‡ x; y; z), respectively.
0
ꢀ
194 Glidewell et al.
Patterns of hydrogen bonding
Acta Cryst. (2001). B57, 190±200

research papers
Table 2
Sulfones retrieved from the CSD.
Atom C4A in the molecule at (x; y; z) acts as a donor to O2 in
the molecule at (� 1 ‡ x; y; z), thus generating by translation a
C(7) chain running parallel to the [100] direction. Four such
chains run through each unit cell and these are linked in pairs
by a further type of hydrogen bond. Atom C2 at (x; y; z) acts
Compound
No.
Space
group
0
CSD code
DPSULO
NOFWEK P2
Z
Reference
(
(
(
(
(
(
(
(
(
(
(
(
(
6)
7)
P2
1
/c
/c
1
1
1
1
2
1
1
Sime & Woodhouse (1974a)
Jeyakanthan et al. (1998)
Jeyakanthan et al. (1998)
Jeyakanthan & Velmurugan (1997)
Chawdhury (1976)
1
as a donor to O2 at (1 � x; 1 � y; � z), thus giving rise to a
8)
9)
NOFWAG C2/c
RIWRUK C2/c
TMPSUL10 Pn2
2
2
centrosymmetric R ꢁ10† ring motif. The combination of these
rings with the translational chains generates a molecular
ladder parallel to [100] in which an antiparallel pair of C(7)
10)
11)
12)
13)
14)
15)
16)
17)
18)
1
a²
Pbca
P2 /c
HMPSUL
DFPSLO
Chawdhury & Hargreaves (1971)
Sime & Woodhouse (1974a)
1
2
2
chains play the role of the uprights and the R ꢁ10† rings act as
CLPSUL10 I2/a
BPHSUL01 I2/a
0.5 Sime & Woodhouse (1974b)
0.5 Sime & Woodhouse (1974b)
0.5 Sime & Woodhouse (1974b)
IPSULO
TEMCET
NFPHSO
NOFVUZ
I2/a
P2
P2
P2
1
1
1
1
1
1
Ellena et al. (1996)
Sime & Woodhouse (1974a)
Jeyakanthan et al. (1998)
/c
/n
²
1
Non-standard setting of Pna2 .
ring (Fig. 6), but there are no CÐHÁ Á ÁF hydrogen bonds.
Hence, the hydrogen-bonded aggregation is strictly ®nite and
zero-dimensional.
3
.2.3. One-dimensional hydrogen bonding. The crystal
structures of (1)±(3) all contain CÐHÁ Á ÁO hydrogen bonds
and in all three compounds the hydrogen bonds lead to the
development of one-dimensional aggregation. However, all
three compounds exhibit different one-dimensional motifs,
simple chains in (1), molecular ladders in (2) and chains of
fused rings in (3). In (1) there are two molecules in the
asymmetric unit (Fig. 1): each type of molecule forms a single
type of CÐHÁ Á ÁO hydrogen bond and the molecules are
linked into C(6) chains, generated by translation and running
parallel to the [010] direction (Fig. 7, Table 3). There are thus
eight chains running through each unit cell, four of each type,
but there are neither hydrogen bonds nor aromatic ꢇÁ Á Áꢇ
stacking interactions between adjacent chains.
Figure 10
Part of the crystal structure of (6), showing a C(6) chain running parallel
to [010]. Atoms are depicted as in Fig. 6. The atom marked with a prime
(0) is at the symmetry position (x; � 1 ‡ y; z).
In (2) the molecules (Fig. 2) are linked by two types of
CÐHÁ Á ÁO hydrogen bond (Table 3) into molecular ladders.
Figure 9
Part of the crystal structure of (3), showing the formation of a chain of
Figure 11
Part of the crystal structure of (10), showing one of the zigzag C(7) chains
2
fused R ꢁ16† rings running parallel to [010]. Atoms are depicted as in Fig.
3
0
6
. The atoms marked with a prime ( ) or a hash (#) are at the symmetry
positions (� x; � ‡ y; � z) and (x; � 1 ‡ y; z), respectively.
running parallel to [100]. Atoms are depicted as in Fig. 6. The atom
1
1
2
1
2
0
1
2
marked with a prime ( ) is at the symmetry position (� ‡ x; y; � z).
2
ꢀ
Acta Cryst. (2001). B57, 190±200
Glidewell et al.
Patterns of hydrogen bonding 195

research papers
Table 3
Hydrogen-bond geometry for (1)±(18) (A, ).
donor to O1 at (x; � 1 ‡ y; z), thus
generating by translation a C(6) chain
parallel to [010]. Atom C3 in the
molecule at (x; y; z) also acts as a
donor to an O1 atom, this time in the
Ê
ꢂ
The atom-numbering for (6)±(18) follows the original papers (Table 2).
HÁ Á ÁO
CÁ Á ÁO
CÐHÁ Á ÁO
Motif
Direction
(
1)
1
1
molecule at (� x; � ‡ y; � z): C3 in
i
2
2
C15AÐH15AÁ Á ÁO2A
2.44
2.58
3.354 (3)
3.317 (3)
166
137
C(6) translation
C(6) translation
[010]
[010]
i
this molecule in turn acts as a donor to
O1 at (x; � 1 ‡ y; z), thus generating a
C(6) spiral parallel to [010]. The
C5BÐH5BÁ Á ÁO2B
(
2)
C2ÐH2Á Á ÁO2
C4AÐH4AÁ Á ÁO2
ii
2
2.59
2.43
3.306 (2)
3.292 (2)
134
154
R ꢁ20†
2
iii
combination of these two C(6) motifs
2
C(7) translation
[100]
produces a chain of fused R ꢁ16† rings
3
(
3)
C3ÐH3Á Á ÁO1
C5AÐH5AÁ Á ÁO1
running parallel to [010] (Fig. 9). Two
of these chains run through each unit
cell and there are aromatic ꢇÁ Á Áꢇ
stacking interactions between adja-
cent chains. The ring C1±C6 at
iv
2.58
2.53
3.500 (3)
3.326 (4)
169
143
C(6) spiral (2
C(6) translation
1
axis)
[010]
[010]
v
(
4)
vi
C3ÐH3Á Á ÁO2
2.52
2.45
3.416 (5)
3.319 (6)
158
151
C(6) zigzag (n glide)
C(6) translation
[101]
[010]
.
..
C5AÐH5A O1
v
(
x; y; z) is a component of a chain in
(
5)
the domain � 0.14 < z < 0.64; this ring
forms a ꢇÁ Á Áꢇ stacking interaction
with the corresponding ring at
(
chain in the domain 0.36 < z < 1.14.
iii
C3ÐH3Á Á ÁO2
C4ÐH4Á Á ÁO1
C6ÐH6Á Á ÁO1
2.41
2.55
2.46
3.317 (3)
3.392 (3)
3.309 (3)
161
147
149
C(6) translation
[100]
[010]
[010]
vii
C(7) spiral (2
C(5) spiral (2
1
1
axis)
axis)
viii
� x; � y; 1 � z), which lies in the
(
6)
v
C11ÐH11Á Á ÁO1
2.51
2.48
3.398 (4)
3.376 (4)
163
159
C(6) translation
C(6) translation
[010]
[010]
The symmetry-related ring at
1
(
7)
1
(
� x; � ‡ y; � z) in the original
v
2
2
C5ÐH5Á Á ÁO1
chain forms a ꢇÁ Á Áꢇ stacking interac-
(
10)
tion with the corresponding ring at
1
ix
C4ÐH4Á Á ÁO1
2.52
2.43
3.354 (4)
3.330 (4)
135
148
C(7) zigzag (a glide)
C(7) zigzag (a glide)
[100]
[100]
1
(
x; � y; � ‡ z), which is a compo-
0
0
0
x
2
2
C4 ÐH4 Á Á ÁO1
nent of a chain in the domain ±0.64 < z
< 0.14. Hence, each chain is linked by
aromatic ꢇÁ Á Áꢇ stacking interactions
to its two immediate neighbours in the
[001] direction.
(
12)
xi
2
C12ÐH12Á Á ÁO2
2.37
2.51
2.48
3.364 (5)
3.441 (9)
3.251 (5)
163
144
141
R ꢁ12†
2
(
15)
xii
2
2
C5ÐH5Á Á ÁO1
R ꢁ12† chain of rings
[001]
[010]
One-dimensional
bonding motifs are also observed in
6), (7), (10), (15) and (16), but, with
hydrogen-
(
16)
xiii
xiv
C6ÐH6Á Á ÁO1
1
C(5) spiral (2 axis)
(
(
17)
the exception of the isomorphous pair
(6) and (7), all of these one-dimen-
sional motifs are different. In (6) and
C6ÐH6Á Á ÁO4
2.36
2.53
2.57
3.250 (6)
3.255 (6)
3.246 (6)
158
135
132
C(7) zigzag (c glide)
C(6) translation
C(5) spiral (2 axis)
1
[001]
[010]
[010]
v
C11ÐH11Á Á ÁO1
C12ÐH12Á Á ÁO2
iv
(
7), simple C(6) chains analogous to
(
18)
0
0
xv
2
those in (1) are generated by transla-
tion (Fig. 10), although there is only a
single type of chain in each structure.
In (10), by contrast, each of the two
independent types of molecule in the
structure forms a C(7) chain gener-
ated by the action of a glide plane
C3 ÐH3 Á Á ÁO3
2.51
2.48
2.52
3.361 (3)
3.423 (4)
3.341 (3)
135
151
138
R ꢁ12†
2
xvi
ꢀ
[101]
[010]
C5ÐH5Á Á ÁO3
C(11) zigzag (n glide)
C(5) spiral (2 axis)
0
0
viii
C5 ÐH5 Á Á ÁO3
1
1
1
Symmetry codes: (i) x; 1 ‡ y; z; (ii) 1 � x; 1 � y; � z; (iii) � 1 ‡ x; y; z; (iv) � x; � ‡ y; � z; (v) x; � 1 ‡ y; z;
2
2
1
1
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
1
2
1
2
(
(
(
vi) � 2 ‡ x; � y; � ‡ z; (vii) � � x; � ‡ y; � z; (viii) � x; � ‡ y; � z; (ix) � ‡ x; y; � z; (x) ‡ x; y; � z;
1 1 1 1 1
2 2 2 2 2
xvi) ‡ x; � y; � ‡ z.
2
2
xi) � 1 � x; 1 � y; � z; (xii) x; � y; ‡ z; (xiii) � x; ‡ y; 1 � z; (xiv) x; � y; � ‡ z; (xv) � x; 1 � y; � z;
1
2
1
2
1
2
(
Fig. 11): the chains formed by the two
2
1 1
2 2
the rungs. These R ꢁ10† rings are centred at (n + , , 0) (n = zero
independent molecules do not interact. The chain in (16) (Fig.
12), is based on a C(5) motif and is generated by the action of a
21 screw axis. Thus, the aromatic rings in this group of
compounds can be used for the formation of CÐHÁ Á ÁO
hydrogen bonds, CÐH bonds which can be ortho, as in (16),
2
4
4
or integer), while between the rungs there are R ꢁ20† rings
1
2
centred at (n, , 0) (n = zero or integer; Fig. 8). There are two
such ladders running through each unit cell, related to one
another by the glide planes: one lies largely in the domain
1
3
1
1
4
< y < , and the other in the domain � < y < , so that there
are no close interactions between neighbouring ladders.
meta, as in (6) and (7), or para, as in (10), to the SO unit, and
2
4
4
4
the chain motifs can be generated by translation, or by the
action of screw axes or glide planes.
The one-dimensional motif in (15) differs from the simple
chains in (6), (7), (10) and (16), largely because the molecules
There are two types of CÐHÁ Á ÁO hydrogen bond in (3)
(
Table 3), which link the molecules (Fig. 3) into a chain of
fused rings. Atom C5A in the molecule at (x; y; z) acts as a
ꢀ
196 Glidewell et al.
Patterns of hydrogen bonding
Acta Cryst. (2001). B57, 190±200

research papers
in (15) have twofold rotational symmetry, so that the forma-
tion of a single type of CÐHÁ Á ÁO hydrogen bond means that
each molecule acts as a double donor of hydrogen bonds and
Table 4
Geometry of CÐXÁ Á ÁX fragments in (14) and (15) (A, ).
0
Ê
ꢂ
0
0
0
X
X
XÁ Á ÁX
CÐXÁ Á ÁX
as a double acceptor. The resulting structure (Fig. 13) is a
2
(
Br
Br
14)
chain of centrosymmetric rings, each of R ꢁ12† type and fused
i
2
Br
Br
3.65
3.65
91
162
ii
in spiro fashion at the S atoms.
3
.2.4. Two-dimensional hydrogen bonding. Two-dimen-
(
I
I
15)
sional aggregation is observed in the methyl-substituted
compounds (4) and (5), as well as in the nitro compound (17).
In (4) there are two types of intermolecular CÐHÁ Á ÁO
hydrogen bonds and each molecule is thus linked to four
others by means of such bonds: the donors are two CÐH
i
I
3.90
3.90
104
172
ii
I
1
2
1
2
1
2
1
2
Symmetry codes: (i) � x; � ‡ y; � z; (ii) � x; ‡ y; � z.
bonds meta to SO and each of the two sulfone O atoms acts as
2
acceptor (Table 3). As in (4), (6) and (7), one motif is a C(6)
chain generated by translation: atom C3 at (x; y; z) acts as a
donor to O2 at (� 1 ‡ x; y; z) (Fig. 15). In addition, C4 at
a single acceptor. Atom C5A at (x; y; z) acts as a donor to O1
at (x; � 1 ‡ y; z), thus forming by translation a C(6) chain
running parallel to [010] (Fig. 14): in this respect, the beha-
viour of (4) resembles that of (6) and (7) (see x3.2.3; Fig. 10).
In addition, however, C3 in the other aryl ring of the molecule
1
2
1
2
1
2
(x; y; z) acts as a donor to O1 at (� � x; � ‡ y; � z), thus
1
4
1
4
generating a C(7) spiral around the 2 axis along (� , y, ).
1
Similarly, C6 at (x; y; z) acts as a donor to O1 at
1
1
1
1
1
1
at (x; y; z) acts as a donor to O2 at (� ‡ x; � y; � ‡ z),
( � x; � ‡ y; � z), thus generating a C(5) spiral around the
1 1
1
2
2
2
2
2
2
1
2
1
2
1
2
while C3 at (� ‡ x; � y; � ‡ z) in turn acts as a donor to
2 axis along ( , y, ): this latter motif mirrors the one-
4 4
O2 at (� 1 ‡ x; y; � 1 ‡ z): this hydrogen bond thus forms a
second C(6) chain motif, generated by the action of the glide
plane at y = 0.25 and running parallel to [101]. The combi-
dimensional aggregation in (16) (see x3.2.3; Fig. 12). The
combination of the two spiral motifs parallel to [010] and the
translational motif along [100] generates a (3,6) net (Batten &
Robson, 1998) lying parallel to (001) and containing a checker-
nation of these two C(6) motifs leads to a square (4,4) net
4
2
3
(
Batten & Robson, 1998) built from a single type of R ꢁ24†
board pattern of R ꢁ10† and R ꢁ14† rings (Fig. 15).
4
3
3
ꢀ
ring and lying parallel to (101) (Fig. 14). There are two such
nets passing through each unit cell, but they are not inter-
woven.
By contrast with (16) (x3.2.3), where the nitro groups do not
participate in the intermolecular hydrogen bonding, in (17)
one of the four nitro O atoms acts as an acceptor in addition to
the two sulfone O atoms. As in (5), molecules of (17) act as
triple donors and triple acceptors: each molecule of (17) is
thus linked to six others but, as in (5) the resulting supramo-
The structure of (5) contains three types of intermolecular
CÐHÁ Á ÁO hydrogen bond, but the supramolecular aggrega-
tion is still only two-dimensional, even though each molecule
is linked to six others. All of the CÐH donors are in the
unsubstituted aryl ring and both O atoms acts as hydrogen-
bond acceptors: O1 is a single acceptor and O2 is a double
Figure 12
Part of the crystal structure of (16), showing one of the spiral C(5) chains
Figure 13
Part of the crystal structure of (15), showing a chain of centrosymmetric
2
running parallel to [010]. Atoms are depicted as in Fig. 6. The atom
0
R ꢁ12† rings running parallel to [001]. Atoms are depicted as in Fig. 6. The
2
1
2
0
1
2
1
2
marked with a prime ( ) is at the symmetry position (� x; ‡ y; 1 � z).
atom marked with a prime ( ) is at the symmetry position (x; � y; ‡ z).
ꢀ
Acta Cryst. (2001). B57, 190±200
Glidewell et al.
Patterns of hydrogen bonding 197

research papers
lecular structure is again only two-dimensional. There are two
parallel motifs involving sulfone O atoms as acceptors; one is
the now-familiar C(6) translational chain and the other is
3.2.5. Three-dimensional hydrogen bonding. Compound
(18) differs from all of those previously discussed in that the
intermolecular CÐHÁ Á ÁO hydrogen bonds do not involve the
spiral lying around a 2 screw axis (Fig. 16). Orthogonal to
1
SO
2
group at all. There are three types of CÐHÁ Á ÁO hydrogen
these is the third motif, which involves a nitro O atom as an
acceptor, and this is a C(7) zigzag chain generated by a glide
plane (Fig. 17). The combination of these motifs generates a
continuous two-dimensional sheet.
bond present, all involving the nitro O atoms as acceptors. Two
2
of these serve to generate (4,4) nets built from R ꢁ20† and
4
4
4
R ꢁ32† rings, arranged in a checker-board fashion (Fig. 18),
while the third type joins adjacent sheets to form a continuous
three-dimensional framework.
3.2.6. General comments on hydrogen bonding. When a
crystal structure contains a single molecular species which acts
as a single donor and as a single acceptor of hydrogen bonds,
chain formation is the necessary consequence and no supra-
molecular aggregation of higher dimensionality is possible.
The aggregation of simple diaryl sulfones, where the nature of
the substituents precludes the formation of hard hydrogen
bonds of OÐHÁ Á ÁO, OÐHÁ Á ÁN or NÐHÁ Á ÁO types, is domi-
nated by chain-forming motifs (Table 3). Depending upon
whether CÐH bonds at the 2-, 3- or 4-positions are involved as
hydrogen-bond donors to the sulfone O atoms as acceptors,
the chain-forming motifs are of C(5), C(6) or C(7) types. For
the compounds considered in this paper, the C(5) chains are
all generated by the action of 2 screw axes, C(6) are mostly
1
Figure 14
Part of the crystal structure of (4), showing the C(6) chains running
generated by translation and most of the small number of C(7)
motifs are generated by glide planes. Against this, it must be
noted that in diphenyl sulfoxide (DPHSOX01; Yatsenko et al.,
1986) there are C(5) chains generated by glide planes. Longer
chain-forming motifs are possible in the presence of other
acceptors, such as nitro groups, as observed in (18) where a
C(11) motif can be identi®ed.
4
parallel to [010] and [101], which generate a (4,4) net of R ꢁ24† rings
4
ꢀ
parallel to (101). Atoms are depicted as in Fig. 6. The atoms marked with
0
1
a prime ( ) or a star (*) are at the symmetry positions (x; � 1 ‡ y; z) and
1
2
1
2
(
sake of clarity.
� ‡ x; � y; � ‡ z), respectively. Methyl groups are omitted for the
2
If there are chain motifs in more than one direction, as in
4), (5), (17) and (18), the supramolecular aggregation of
(
higher dimensionality results: a two-dimensional structure
generated by hydrogen bonding involving a single molecular
Figure 15
Part of the crystal structure of (5), showing the formation of a sheet
parallel to (001) built from a C(6) translational chain running parallel to
100], and C(5) and C(7) spirals running parallel to [010]. Atoms are
[
depicted as in Fig. 6. The atoms marked with a prime ( ), a star (*) or a
Figure 16
Part of the crystal structure of (17), showing the C(5) and C(6) chains
0
hash (#) are at the symmetry positions (� 1 ‡ x; y; z),
running parallel to [010]. Atoms are depicted as in Fig. 6. The atoms
0
1
2
1
2
1
2
1
2
1
1
2
(
groups are omitted for the sake of clarity.
� x; � ‡ y; � z) and (� � x; � 2 ‡ y; � z), respectively. Methyl
marked with a prime ( ) or a star (*) are at the symmetry positions
1
1
(x; � 1 ‡ y; z) and (� x; � ‡ y; � z), respectively.
2
2
ꢀ
198 Glidewell et al.
Patterns of hydrogen bonding
Acta Cryst. (2001). B57, 190±200

research papers
species requires at least twofold donor and acceptor beha-
viour, while three-dimensional aggregation can be generated
by either twofold or threefold donor/acceptor behaviour. Two-
fold donor/acceptor behaviour can lead to two-dimensional
plane nets or to three-dimensional diamondoid structures.
The formation of ®nite (zero-dimensional) cyclic aggregates
3.3. Molecular conformations and dimensions
For diaryl sulfones, the commonest conformation for the
C ±SO ±C skeleton is that of exact or approximate C2
6
2
6
symmetry, with each aryl ring almost coplanar with a CÐSÐO
fragment (Ferguson et al., 1999). With the exception of (1) and
(10), all of the compounds considered in this paper exhibit this
conformation, while (13)±(15) exhibit exact C symmetry.
requires the formation of paired CÐHÁ Á ÁO hydrogen bonds
2
2
2
and such R ꢁn† motifs are usually generated around centres of
Although the simplest interpretation is in terms of the polarity
of the aromatic CÐH bonds ortho to the SO unit, in which
inversion. It is striking that although all of the compounds
considered in this paper, except for (10) and (16), crystallize in
centrosymmetric space groups, such cyclic motifs are rather
uncommon: only in (12) is a genuinely zero-dimensional
aggregate formed.
2
both O atoms are involved, one with each ring (Ferguson et al.,
999), the C conformation persists even in (11), although the
2
1
rings are displaced from coplanarity with the CÐSÐO frag-
ments. Both molecules in (1) have approximate C symmetry,
s
The absence of any CÐHÁ Á ÁO hydrogen bonds in (8) and
with close contacts between CÐH bonds from both rings and
a single O atom. It is interesting to note that in diphenyl
sulfoxide, CÐH bonds from both rings form close contacts
with the single O atom, giving a conformation close to Cs
symmetry. In both molecules of (10), the unsubstituted ring is
almost coplanar with an adjacent CÐSÐO plane, but the
(
9) is unexpected and somewhat puzzling, in comparison to
the behaviour of (2) and (10), which are isomeric with (8), and
that of (4), which is isomeric with (9). On the other hand, the
differing behaviour of (1) and (5), in which the supramolecular
aggregation is one- and two-dimensional, respectively, shows
that isomeric pairs differing only marginally in the placing of
the methyl substituents can behave quite differently.
ꢂ
substituted ring is displaced by ca 25 , so that the overall
3
.2.7. Non-bonded halogenÁ Á Áhalogen contacts. In the
original report on the structures of (13)±(15) it was noted
Sime & Woodhouse, 1974b) that there were intermolecular
(
XÁ Á ÁX contacts signi®cantly shorter than the sum of the van
der Waals radii (Bondi, 1964) when X = Br or I. The CÐXÁ Á ÁX
geometries for (14) and (15) (Table 4) are admirably consis-
tent with the generalizations proposed (Ramasubbu et al.,
1
986) from the results of database analysis, namely that in
structures where the XÁ Á ÁX distances are signi®cantly less than
the van der Waals sum, the observed CÐXÁ Á ÁX angles are
ꢂ
ꢂ
clustered around 180 and around 90 . In systems of the type
CÐXÁ Á ÁXÐC one of the CÐXÁ Á ÁX angles is generally close
ꢂ
ꢂ
to 180 and the other to 90 (Table 4, Fig. 19). In (3), (4), (9)
and (13) there are no intermolecular ClÁ Á ÁCl contacts which
are signi®cantly shorter than the sum of the van der Waals
radii.
Figure 18
2
4
Part of the crystal structure of (18), showing a (4,4) net built from R ꢁ20†
4
4
and R ꢁ32† rings. Atoms are depicted as in Fig. 6. The atoms marked with
1
2
1
2
1
2
a star (*) or a hash (#) are at symmetry positions ( ‡ x; � y; � ‡ z)
1
2
1
2
1
2
0
and ( ‡ x; � y; � ‡ z), respectively: the atom H5 is designated as
such in the original report. Methyl groups are omitted for the sake of
clarity.
Figure 19
Figure 17
Part of the crystal structure of (17), showing the C(7) chain running
Part of the crystal structure of (15), showing the geometry of the CÐIÁ Á ÁI
contacts. For the sake of clarity H atoms are omitted. The atoms marked
with a prime ( ) or a star (*) are at the symmetry positions
0
parallel to [001]. Atoms are depicted as in Fig. 6. The atom marked with a
1
0
1
2
1
1
1
1
prime ( ) is at the symmetry position (x; � y; � ‡ z).
(� x; � ‡ y; � z) and (� x; ‡ y; � z), respectively.
2
2
2
2
2
ꢀ
Acta Cryst. (2001). B57, 190±200
Glidewell et al.
Patterns of hydrogen bonding 199

research papers
skeletal symmetry is only C . The bond lengths and angles in
1
Chawdhury, S. A. (1976). Acta Cryst. B32, 1069±1073.
Chawdhury, S. A. & Hargreaves, A. (1971). Acta Cryst. B27, 548±552.
Cotton, F. A., Daniels, L. M., Jordan, G. T. & Murillo, C. A. (1997).
Chem. Commun. pp. 1673±1674.
(
1)±(5) show no unusual features.
Cotton, F. A. & Luck, R. L. (1989). Inorg. Chem. 28, 3210±3213.
Ellena, J., Punte, G. & Nudelman, N. S. (1996). Acta Cryst. C52, 2929±
2932.
Ferguson, G. (1999). PRPKAPPA: a WordPerfect5.1 Macro to
Formulate and Polish CIF Format Files from the SHELXL97
Re®nement of Kappa-CCD Data. University of Guelph, Canada.
Ferguson, G., Glidewell, C., Gregson, R. M. & Lavender, E. S. (1999).
Acta Cryst. B55, 573±590.
4
. Concluding comments
In the diaryl sulfones (1)±(5), whose structures are reported
here, the substituent types preclude the formation of hard
hydrogen bonds. All of (1)±(5) exhibit soft CÐHÁ Á ÁO
hydrogen bonds and the dominant mode of supramolecular
aggregation is chain formation. For the related diaryl sulfones
(
6)±(18) retrieved from the CSD, chain formation is again the
Glidewell, C. & Ferguson, G. (1996). Acta Cryst. C52, 2528±2530.
Grant, D. W., Hogg, D. R. & Wardell, J. L. (1987). J. Chem. Res. (S),
pp. 392±393.
Jeyakanthan, J. & Velmurugan, D. (1997). Acta Cryst. C53, 1344±1345.
Jeyakanthan, J., Velmurugan, D., Panneerselvam, K., Soriano-Garcia,
M., Perumai, S. & Chandrasekaran, R. (1998). Acta Cryst. C54,
630±633.
dominant mode of aggregation, with a variety of motifs
observed. Where the molecules utilize more than one CÐH
bond, the resulting supramolecular aggregation may be zero-,
one-, two- or three-dimensional, but the constituent one-
dimensional substructures can exhibit a rich diversity of form.
Nonius (1997). Kappa-CCD Server Software, Windows 3.11 Version.
Nonius BV, Delft, The Netherlands.
Novi, M., Dell'Erba, C., Garbarino, G. & Sancassan, F. (1982). J. Org.
Chem. 47, 2292±2298.
Otwinowski, Z. & Minor, W. (1997). Methods Enzymol. 276, 307±326.
Ramasubbu, N., Parthasarathy, R. & Murray-Rust, P. (1986). J. Am.
Chem. Soc. 108, 4308±4314.
Diffraction data for (2), (4) and (5) were collected via the
EPSRC X-ray Crystallographic Service at the University of
Southampton: we thank Professor M. B. Hursthouse and his
staff for their assistance.
Riddell, F. G. & Rogerson, M. (1996). J. Chem. Soc. Perkin Trans. 2,
pp. 493±504.
Riddell, F. G. & Rogerson, M. (1997). J. Chem. Soc. Perkin Trans. 2,
pp. 249±255.
References
Allen, F. H. & Kennard, O. (1993). Chem. Des. Autom. News, 8, 31±
3
7.
Batten, S. R. & Robson, R. (1998). Angew. Chem. Int. Ed. 37, 1460±
494.
Sheldrick, G. M. (1997a). SHELXS97. University of G oÈ ttingen,
Germany.
Sheldrick, G. M. (1997b). SHELXL97. University of G oÈ ttingen,
Germany.
Sime, J. G. & Woodhouse, D. I. (1974a). J. Cryst. Mol. Struct. 4, 269±
285.
Sime, J. G. & Woodhouse, D. I. (1974b). J. Cryst. Mol. Struct. 4, 287±
303.
1
Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995). Angew.
Chem. Int. Ed. Engl. 34, 1555±1573.
Bertolasi, V., Ferretti, V., Gilli, P. & de Benedetti, P. G. (1993). J.
Chem. Soc. Perkin Trans 2, pp. 213±219.
Blessing, R. H. (1995). Acta Cryst. A51, 33±37.
Blessing, R. H. (1997). J. Appl. Cryst. 30, 421±426.
Bondi, A. (1964). J. Phys. Chem. 68, 441±451.
Braga, D., Grepioni, F., Biradha, K., Pedireddi, V. R. & Desiraju, G. R.
Spek, A. L. (2000). PLATON, May 2000 Version. University of
Utrecht, The Netherlands.
Wardell, J. L., Low, J. N. & Glidewell, C. (2000). Acta Cryst. C56, 862±
864.
(1995). J. Am. Chem. Soc. 117, 3156±3166.
Bruker (1997). SADABS, SMART, SAINT. Bruker AXS Inc.,
Madison, Wisconsin, USA.
Burton, H. & Praill, P. F. G. (1955). J. Chem. Soc. pp. 887±889.
Wilson, A. J. C. (1976). Acta Cryst. A32, 994±996.
Yatsenko, A. V., Medvedev, S. V., Tursina, A. I. & Aslanov, L. A.
(1986). Zh. Obshch. Khim. 56, 2330±2339.
ꢀ
200 Glidewell et al.
Patterns of hydrogen bonding
Acta Cryst. (2001). B57, 190±200