Packing similarities of three isosteric molecules: 4,5-Dichlorophthalic anhydride, 4,5-dibromophthalic anhydride and 5,6-dichlorobenzfurazan 1-oxide, including three polymorphs of 5,6-dichlorobenzfurazan 1-oxide

Article abstract of DOI:10.1107/S0108768199000348

4,5-Dichlorophthalic anhydride (CPA) lies on a twofold axis in space group C2/c; the molecules pack as stacks of two-dimensional sheets. Polymorph A of 5,6-dichlorobenzfurazan 1-oxide (CBF; systematic name 5,6-dichloro-2,1,3-benzoxadiazole 1-oxide) is iso

Full text of DOI:10.1107/S0108768199000348

530
Acta Cryst. (1999). B55, 530±542
Packing similarities of three isosteric molecules: 4,5-dichlorophthalic anhydride, 4,5-dibromo-
phthalic anhydride and 5,6-dichlorobenzfurazan 1-oxide, including three polymorphs of 5,6-
dichlorobenzfurazan 1-oxide
Charles R. Ojala,a William H. Ojala,b Doyle Brittonc* and J. Z. Gougoutasd
aDepartment of Chemistry, Normandale Community College, Bloomington, Minnesota 55431, USA, bDepartment of
Chemistry, University of Saint Thomas, St Paul, Minnesota 55105, USA, cDepartment of Chemistry, University of
Minnesota, Minneapolis, Minnesota 55455, USA, and dThe Bristol±Myers Squibb Pharmaceutical Research Institute,
PO Box 4000, Princeton, New Jersey 08543-4000, USA. E-mail: britton@chemsun.chem.umn.edu
(Received 3 April 1998; accepted 6 January 1999)
Abstract
of the question are discussed at greater length by
Gougoutas (1983). This paper describes the similarities
and differences in a set of related structures.
4,5-Dichlorophthalic anhydride (CPA) lies on a twofold
axis in space group C2/c; the molecules pack as stacks of
two-dimensional sheets. Polymorph A of 5,6-dichloro-
benzfurazan 1-oxide (CBF; systematic name 5,6-
dichloro-2,1,3-benzoxadiazole 1-oxide) is isomorphous
with CPA with disordered CBF molecules being pseudo-
isosteric with the CPA molecule. Polymorph C of CBF
has similar unit-cell dimensions but the arrangement of
the molecules in the cell is different. 4,5-Dibromo-
The original crystal structure of 5,6-dichlorobenzfur-
azan 1-oxide (CBF) showed disordered CBF molecules
situated on a twofold axis in space group C2/c (Britton et
al., 1972). The CBF molecule disordered across a
twofold axis is pseudo-isosteric with 4,5-dichloro-
phthalic anhydride (CPA). Photographic data showed
the two compounds to have similar cell dimensions and
similar intensity patterns, i.e. they are indeed isomor-
phous (Ojala, 1986). We report here the complete
structure of 4,5-dichlorophthalic anhydride (CPA) in
order to look at the details of the isomorphism. In
addition, we report the structure of 4,5-dibromophthalic
anhydride (BPA), which is approximately isosteric with
CPA, for comparison.
As what was originally intended as a secondary part
of this study we began a redetermination of the structure
of CBF. In the course of this, however, we found two
additional polymorphs of CBF. We report here the
redetermined structure of CBF-A at 297 and 173 K to
con®rm that the structure is disordered in C2/c and not
ordered in Cc. We also report the structures of CBF-B
and CBF-C at 297 and 173 K.
Å
phthalic anhydride occurs in space group P1 in a
structure derived from the CPA structure. The four
structures all have similar two-dimensional sheets of
molecules, but differ, in addition to the disorder of the
CBF molecules, in the stacking of successive sheets.
There are three polymorphs of 5,6-dichlorobenzfurazan
1-oxide. In all three the molecules are disordered about
twofold or pseudo-twofold axes. In all three the
molecules form chains involving similar ClÁ Á ÁO contacts
along the direction of the twofold axis. In polymorphs A
and C these chains form virtually identical two-
dimensional sheets with inter-chain CÐHÁ Á ÁO, CÐ
HÁ Á ÁN and ClÁ Á ÁCl contacts, but with different stacking
arrangements between the sheets. In polymorph B the
chains are associated in face-to-face pairs with the same
face-to-face overlap as in A, at an average distance of
Ê
Ê
3.42 A compared to an average distance of 3.38 A in A
(cf. Table 8); there are no two-dimensional sheets.
2. Experimental
2.1. The polymorphs of CBF
CBF was prepared by hypochlorite oxidation of 4,5-
dichloro-2-nitroaniline (Aldrich) and originally recrys-
1. Introduction
The phenomenon of isomorphism (Mitscherlich, 1822) is tallized from methanol. When it was recognized that two
a familiar one, and the idea that, in some cases at least, it polymorphs were present in this sample, CBF was
is traceable to isosterism (having the same size and recrystallized from other solvents as well. Eventually
shape) is also long standing (Langmuir, 1919). For some three polymorphs were found; they are labeled CBF-A,
time (Gougoutas, 1983; Ojala, 1985, 1986) we have been -B and -C in the order from most dense to least dense,
interested in using isosteres as a probe for electronic since this is probably the order from most to least stable
behavior. We feel there is as much interest in a pair of at low temperature (Dunitz, 1995). Crystals of suf®cient
isosteres that do not have isomorphous crystals (and quality for single-crystal diffraction were obtained from
why they do not) as in a pair that do. The general aspects the following solvents (identi®cation of the polymorphs
# 1999 International Union of Crystallography
Printed in Great Britain ± all rights reserved
Acta Crystallographica Section B
ISSN 0108-7681 # 1999

OJALA, OJALA, BRITTON AND GOUGOUTAS
531
was by unit-cell determination): methanol, a mixture of as previously but with much more assurance. The room-
thick needles (A) elongated along c with faces {110} and temperature data were re®ned with the same restraints
Å
Å Å
{101}, thick plates (A) with faces (110) and (110), and using the low-temperature results as the starting para-
plates (C) with plate faces {001}; 2-propanol, prisms (B) meters and again converged satisfactorily.
with faces {011}, {110} and {010}; ethanol/water, plates The trial structure for CBF-B was obtained by direct
(C) with faces {001} and {110}; ethyl acetate, plates and methods. The same kind of disorder is present as in
sword-shaped (both A) elongated along c with faces CBF-A except that in the CBF-B structure the mole-
Å
{110} and {101} developed to different extents in the two cules do not lie on symmetry elements so that the two
habits; benzene, needles (A) elongated along c; acetone, components of the disorder do not have to be present in
needles (A) elongated along c and prisms (C) with faces equal amounts. Again unrestrained re®nement was
Å
{001} and {111}; chloroform, thick plates (A) with faces unsatisfactory. The restraints used in CBF-A were again
{100}; carbon tetrachloride, thick plates (C) with faces used. In addition, the molecules in the two orientations
{001} and {110}. The needles, prisms and thick plates of the disorder were constrained to have exactly the
were yellow; the thin plates were so thin as to appear same bond lengths and angles. With these restraints
colorless.
re®nement converged satisfactorily. With the same
CBF-A melted at 404±405 K after undergoing a restraints, and using the low-temperature results as the
transition to a second solid phase (turning opaque) at trial structure, re®nement of the room-temperature data
359±366 K. CBF-B melted at 404±406 K after turning also converged satisfactorily.
opaque at 368±376 K. CBF-C melted at 404±405 K. All
The trial structure for CBF-C was obtained by direct
three forms tended to sublime when the temperature methods. The situation is the same as in CBF-A; there
neared 373 K. Since all three forms melted at the same are four molecules in Cc or C2/c with the C2/c alter-
temperature, we interpret these observations as indi- native requiring disorder. Re®nement was carried out in
cating that A transforms to C at 359±366 K, and B the same manner as for CBF-A, with the same restraints,
transforms to C at 368±376 K.
and assuming disorder in C2/c. Re®nement converged
The itemized crystal data, data collection and re®ne- satisfactorily. If re®nement was carried out assuming
ment details for all the polymorphs are given in Table 1. ordered molecules in Cc, R and wR were again more
In each case data were collected ®rst at room than twice as large as for the C2/c re®nement, so again
temperature and then, when re®nement proved to be the latter was accepted as the correct description of the
dif®cult, new data were collected at low temperature. structure. With the same assumption and restraints, and
Data collected on the CAD-4 diffractometer were using the low-temperature results as the trial structure,
processed using CAD-4 Software (Enraf±Nonius, 1989) re®nement of the room-temperature data also
and TEXSAN (Molecular Structure Corporation, 1985); converged satisfactorily.
data collected on the SMART system were processed
For polymorphs A and C, re®nements carried out
using ASTRO and SAINT (Siemens, 1995). The trial assuming that the C, N and Cl atoms were exactly
structures for CBF-B and -C at low temperature were related by the twofold axis converged with R values less
obtained, and all re®nements were carried out, using than 0.05 for the low-temperature data. Similarly, if B
SHELXTL (Sheldrick, 1994). Graphics were prepared was assumed to be ordered, except for disorder in the O-
using SHELXTL and TEXSAN.
atom positions, re®nement converged with R less than
For CBF-A the results from the previous determina- 0.05. However, in all three cases the resulting aniso-
tion were used as the trial structure for the low- tropic displacement ellipsoids were physically unrea-
temperature data. Unrestrained re®nement in C2/c did sonable, so that these partially ordered models were
not lead to reasonable results, so the following restraints rejected in favor of the ones described above.
were included: the CÐC and CÐCl bond lengths related
by the disorder were restrained to be approximately
equal and the atoms involved in these bonds were
restrained to have similar anisotropic displacement
parameters. With these restraints re®nement converged
2.2. The dihalophthalic anhydrides
CPA was obtained from Aldrich Chemical Company
satisfactorily. To con®rm that disorder in C2/c is the and recrystallized from carbon tetrachloride to obtain
correct model, the ®nal result for the C2/c re®nement crystals (needles) suitable for diffraction. The crystals
was used as the starting point for re®nement of an melted at 458±460 K. The same crystal modi®cation was
ordered molecule in Cc; the same restraints were used. also obtained from acetone (prisms and needles),
This re®nement converged with R(F) for the observed benzene (prisms) and chloroform (prisms and plates).
data of 0.073 and wR(F2) for all the data of 0.2395. These The needles were elongated along c. The prisms had
Å
values are more than twice as large as the C2/c values. In faces {100}, {110} and {111}. The plates had plate faces
Å
Å Å
addition, the ®nal difference map showed large peaks (111) and (111), which is surprising in that only part of
for the two O-atom positions that corresponded to the the {111} form was developed and this was not a form
disordered model. Thus, we reach the same conclusion found in the prisms. The alcohols that were used for the

532
PACKING SIMILARITIES OF THREE ISOSTERIC MOLECULES
Table 1. Experimental details
CBF-A
CBF-B
Crystal data
Chemical formula
Chemical formula weight
Cell setting
C6H2Cl2N2O2
205
Monoclinic
C2=c
11.397 (2)
8.871 (2)
7.222 (4)
90
95.12 (3)
90
727.3 (5)
4
1.872
Mo Kꢀ
0.71073
25
C6H2Cl2N2O2
205
Monoclinic
C2=c
11.554 (2)
8.926 (2)
7.310 (6)
90
96.10 (4)
90
749.6 (7)
4
1.816
Mo Kꢀ
0.71073
24
C6H2Cl2N2O2
205
Monoclinic
P21=n
7.0679 (6)
11.7502 (9)
9.0114 (7)
90
98.563 (1)
90
740.05 (10)
4
1.840
Mo Kꢀ
0.71073
3250
C6H2Cl2N2O2
205
Monoclinic
P21=n
7.1739 (2)
11.8351 (2)
9.0639 (1)
90
98.350 (1)
90
761.40 (3)
4
1.788
Mo Kꢀ
0.71073
2148
Space group
Ê
a (A)
Ê
b (A)
Ê
c (A)
ꢀ ꢀꢁ†
ꢁ ꢀꢁ†
ꢂ ꢀꢁ†
3
Ê
V (A )
Z
Dx (Mg m
3
)
Radiation type
Ê
Wavelength (A)
No. of re¯ections for cell
parameters
ꢃ range (ꢁ)
11±22
0.842
173 (2)
Sword-shaped
0.50 Â 0.28 Â 0.12
Yellow
14±23
0.817
297 (2)
Sword-shaped
0.50 Â 0.28 Â 0.12
Yellow
2±25
0.827
173 (2)
Prism
2±25
0.804
297 (2)
Prism
1
ꢄ (mm
)
Temperature (K)
Crystal form
Crystal size (mm)
Crystal color
0.50 Â 0.35 Â 0.30
0.50 Â 0.35 Â 0.30
Yellow
Yellow
Data collection
Diffractometer
Data collection method
Absorption correction
CAD-4
!±2ꢃ scans
CAD-4
!±2ꢃ scans
SMART
! scans
SMART
! scans
SADABS (Sheldrick,
1996)
0.47
0.82
scan (North et al., 1968) scan (North et al., 1968) SADABS (Sheldrick,
1996)
0.74
0.90
2528
Tmin
Tmax
0.74
0.90
1811
0.63
0.81
3699
No. of measured re¯ec-
tions
3751
No. of independent re¯ec- 1267
tions
No. of observed re¯ections 1140
908
1305
1330
780
I > 2ꢅ(I)
1145
I > 2ꢅ(I)
1158
I > 2ꢅ(I)
Criterion for observed
re¯ections
Rint
I > 2ꢅ(I)
0.026
32
0.024
28
0.016
25
0.015
25
ꢃmax (ꢁ)
Range of h, k, l
16 ! h ! 16
15 ! h ! 15
8 ! h ! 8
11 ! k ! 13
10 ! l ! 10
150²
8 ! h ! 8
13 ! k ! 14
9 ! l ! 10
153²
0 ! k ! 13
0 ! k ! 11
10 ! l ! 10
9 ! l ! 9
No. of standard re¯ections
Frequency of standard
re¯ections (min)
3
50
3
50
360
360
Re®nement
Re®nement on
R‰F2>2ꢅꢀF2†Š
wRꢀF2†
F2
F2
F2
F2
0.029
0.076
1.131
1267
0.035
0.085
1.105
908
0.038
0.110
1.044
1305
0.036
0.107
1.687
1330
S
No. of re¯ections used in
re®nement
No. of parameters used
No. of restraints used
H-atom treatment
110
33
110
33
218
113
218
113
Placed in idealized posi-
tions
Placed in idealized posi-
tions
Placed in idealized posi-
tions
Placed in idealized posi-
tions
Weighting scheme
w = 1/[ꢅ2(F2o) + (0.029P)2
+ 0.14P] where
w = 1/[ꢅ2(F2o) + (0.032P)2
+ 0.15P] where
w = 1/[ꢅ2(F2o) + (0.065P)2
+ 0.91P] where
w = 1/[ꢅ2(F2o) + (0.045P)2
+ 0.027P] where
P = (F2o + 2F2c)/3
P = (F2o + 2F2c)/3
P = (F2o + 2F2c)/3
P = (F2o + 2F2c)/3
ꢀÁ=ꢅ†
0.01
0.47
0.35
0.02
0.37
0.21
0.03
0.47
0.50
0.04
0.37
0.31
max
3
Ê
Ê
Áꢆmax (e A
)
3
Áꢆmin (e A
)

OJALA, OJALA, BRITTON AND GOUGOUTAS
533
Table 1 (cont.)
CBF-A
CBF-B
Extinction method
SHELXL (Sheldrick,
1994)
0.039 (3)
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
SHELXL (Sheldrick,
1994)
0.055 (4)
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
None
None
Ð
Extinction coef®cient
Source of atomic scat-
tering factors
Ð
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
CBF-C
CPA
BPA
Crystal data
Chemical formula
Chemical formula weight
Cell setting
C6H2Cl2N2O2
205
Monoclinic
C2=c
7.2233 (1)
9.0549 (3)
11.5700 (4)
90
95.315 (2)
90
753.50 (4)
4
1.807
Mo Kꢀ
0.71073
1656
C6H2Cl2N2O2
205
Monoclinic
C2=c
7.293 (5)
9.061 (6)
11.618 (2)
90
96.12 (3)
90
763.4 (12)
4
1.784
Mo Kꢀ
0.71073
25
C8H2Cl2O3
217
Monoclinic
C2=c
11.987 (4)
8.994 (3)
7.423 (4)
90
92.91 (4)
90
799.2 (6)
4
1.803
Mo Kꢀ
0.71073
49
C8H2Br2O3
305.92
Triclinic
C1
12.923 (3)
9.209 (2)
7.6280 (15)
105.08 (3)
99.20 (3)
81.29 (3)
859.5 (3)
4
Space group
Ê
a (A)
Ê
b (A)
Ê
c (A)
ꢀ ꢀꢁ†
ꢁ ꢀꢁ†
ꢂ ꢀꢁ†
3
Ê
V (A )
Z
Dx (Mg m
3
)
Radiation type
2.364
Mo Kꢀ
0.71073
25
Ê
Wavelength (A)
No. of re¯ections for cell
parameters
ꢃ range (ꢁ)
2±25
0.812
173 (2)
Thick plate
0.50 Â 0.45 Â 0.20
Yellow
14±22
0.802
297 (2)
Thin plate
0.45 Â 0.25 Â 0.01
Colorless
11±29
0.774
297 (2)
Thick needle
0.45 Â 0.25 Â 0.15
Colorless
13±22
9.4
297 (2)
Plate
1
ꢄ (mm
)
Temperature (K)
Crystal form
Crystal size (mm)
Crystal color
0.35 Â 0.20 Â 0.07
Colorless
Data collection
Diffractometer
Data collection method
Absorption correction
SMART
! scans
SADABS (Sheldrick,
CAD-4
!±2ꢃ scans
CAD-4
! scans
CAD-4
!±2ꢃ scans
scan (North et al., 1968) scan (North et al., 1968) scan (North et al., 1968)
1996)
0.67
0.85
Tmin
Tmax
0.82
0.99
1588
0.76
0.89
2336
0.23
0.52
4009
No. of measured re¯ec-
tions
3292
No. of independent re¯ec- 659
tions
No. of observed re¯ections 594
747
1171
2058
472
I > 2ꢅ(I)
1041
I > 2ꢅ(I)
1284
I > 2ꢅ(I)
Criterion for observed
re¯ections
Rint
I > 2ꢅ(I)
0.024
25.05
0.079
26
0.015
30
0.050
28
ꢃmax (ꢁ)
Range of h, k, l
8 ! h ! 8
8 ! h ! 8
16 ! h ! 16
16 ! h ! 16
10 ! k ! 7
13 ! l ! 13
0 ! k ! 11
0 ! k ! 12
12 ! k ! 12
14 ! l ! 14
10 ! l ! 10
10 ! l ! 10
No. of standard re¯ections 150²
Frequency of standard
re¯ections (min)
3
60
3
50
3
50
360
Re®nement
Re®nement on
R‰F2>2ꢅꢀF2†Š
wRꢀF2†
F2
F2
F2
F2
0.027
0.073
1.116
659
0.038
0.136
0.825
747
0.029
0.085
1.063
1171
0.050
0.139
1.028
2057
S
No. of re¯ections used in
re®nement

534
PACKING SIMILARITIES OF THREE ISOSTERIC MOLECULES
Table 1 (cont.)
CBF-C
CPA
BPA
No. of parameters used
No. of restraints used
H-atom treatment
109
33
109
33
61
0
119
0
Placed in idealized posi-
tions
Placed in idealized posi-
tions
Not re®ned, placed in
idealized positions
w = 1/[ꢅ2(F2o) + (0.036P)2
+ 0.23P] where
P = (F2o + 2F2c)/3
0.01
Placed in idealized posi-
tions
w = 1/[ꢅ2(F2o) + (0.064P)2
+ 0.18P] where
P = (F2o + 2Fc2)/3
0.01
Weighting scheme
w = 1/[ꢅ2(F2o) + (0.043P)2
+ 0.199P] where
w = 1/[ꢅ2(F2o) + (0.055P)2
+ 1.03P] where
P = (F2o + 2Fc2)/3
P = (F2o + 2Fc2)/3
ꢀÁ=ꢅ†
0.02
0.18
0.21
None
0.01
0.18
0.15
None
max
3
Ê
Ê
Áꢆmax (e A
)
0.43
0.18
0.80
0.74
3
Áꢆmin (e A
Extinction method
)
SHELXL (Sheldrick,
1994)
0.146 (7)
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
SHELXL (Sheldrick,
1994)
0.0047 (9)
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
Extinction coef®cient
Source of atomic scat-
tering factors
Ð
Ð
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
International Tables for
Crystallography (1992,
Vol. C, Tables 4.2.6.8
and 6.1.1.4)
² The ®rst N re¯ections were recollected at the end of the data collection.
recrystallization of CBF could not be used since they (normal) twin with 180ꢁ rotation of the interface plane
reacted with CPA to form esters.
BPA was prepared by oxidizing 4,5-dibromoxylene to
(001).
Crystal data, data collection and re®nement details
4,5-dibromophthalic acid with permanganate and then are given in Table 1. The same programs were used as
dehydrating the acid with thionyl chloride. Crystals above, plus MITHRIL (Gilmore, 1984) and DIRDIF
suitable for diffraction were obtained by recrystalliza- (Beurskens, 1984) were used to obtain a trial structure
tion from carbon tetrachloride (plates). The crystals for BPA.
melted at 493±495 K. The same crystal modi®cation was
The CPA structure was straightforward; the trial
also obtained from acetone and chloroform. The faces structure was taken from the CBF-A results. The BPA
were not determined on the crystal used for the struc- structure was solved by direct methods in space group
Å
Å
ture determination. Other crystals, examined later, were P1, but it is reported throughout the paper in C1 for
all twinned. DIRAX (Duisenberg, 1992) was used to sort better comparison with the CPA structure. The matrix
out the twinning. The crystal was a perpendicular
1 1
1 1
Å
2,2,0/0,0,1/2,2,0 will convert the C1 cell to the conven-
Fig. 1. The structures of CPA (left)
and BPA (right). Displacement
ellipsoids are shown at the 50%
probability level. H atoms are
shown with arbitrary radii. CPA
lies on a crystallographic twofold
axis; BPA does not.

OJALA, OJALA, BRITTON AND GOUGOUTAS
535
Table 2. Fractional atomic coordinates and equivalent
Ê 2
Table 2 (cont.)
isotropic displacement parameters (A )
x
y
z
Ueq
Ueq ˆ ꢀ1=3†ÆiÆjUijaiajai:aj:
CBF-C 173 K
x
y
z
Ueq
C1
C2
C3
C4
C5
C6
N1
N2
O1
O2
Cl4
Cl5
0.5311 (13)
0.2759 (6)
0.2843 (7)
0.4281 (16)
0.5510 (18)
0.5284 (18)
0.3957 (15)
0.1343 (5)
0.1483 (5)
0.0635 (4)
0.0517 (3)
0.7204 (9)
0.6932 (9)
0.7869 (9)
0.6770 (7)
0.6298 (10)
0.6922 (9)
0.8080 (9)
0.8566 (10)
0.8159 (3)
0.6335 (4)
0.8995 (3)
0.7181 (2)
0.6282 (6)
0.8783 (6)
0.028 (2)
0.028 (2)
CPA
Cl5
O1
O2
C7
C1
C5
C6
0.4322 (14)
0.3963 (12)
0.4539 (16)
0.5525 (16)
0.5955 (12)
0.5552 (5)
0.3926 (5)
0.6282 (4)
0.4665 (4)
0.4124 (10)
0.6101 (10)
0.62077 (3)
0.66550 (9)
1/2
0.58546 (11) 0.0907 (2)
0.55276 (9) 0.24651 (14)
0.55436 (10) 0.50873 (14)
0.67439 (4)
0.03635 (13)
0.0016 (2)
0.33897 (5)
0.4008 (2)
1/4
0.3254 (2)
0.2933 (2)
0.2914 (2)
0.3361 (2)
0.0508 (2)
0.0646 (4)
0.0565 (4)
0.0461 (3)
0.0378 (3)
0.0384 (3)
0.0403 (3)
0.0322 (14)
0.0286 (15)
0.0296 (15)
0.0328 (14)
0.0384 (8)
0.0410 (8)
0.0513 (7)
0.0431 (7)
0.0368 (7)
0.0372 (7)
0.60860 (9)
0.3765 (2)
BPA
Br4
Br5
O1
O2
O3
C7
C8
C1
C2
C3
0.37379 (6)
0.62425 (6)
0.6722 (5)
0.5069 (4)
0.3416 (5)
0.5912 (6)
0.4208 (6)
0.5553 (5)
0.4529 (5)
0.3955 (5)
0.4498 (5)
0.5523 (5)
0.6090 (5)
0.70771 (8)
0.69818 (9)
0.0542 (6)
0.0238 (5)
0.0645 (6)
0.1096 (8)
0.1150 (8)
0.2655 (8)
0.2683 (8)
0.3988 (8)
0.5276 (7)
0.5250 (7)
0.3925 (7)
0.16374 (12) 0.0508 (3)
0.35590 (12) 0.0518 (3)
² Site occupancy 0.622 (4). ³ Site occupancy 0.378 (4).
0.3695 (9)
0.2304 (8)
0.0907 (9)
0.3068 (11)
0.1613 (11)
0.2842 (9)
0.1995 (10)
0.1619 (9)
0.2125 (9)
0.2963 (10)
0.3352 (9)
0.063 (2)
0.0509 (13)
0.061 (2)
0.044 (2)
0.044 (2)
Ê
Å
tional P1 cell [a = 7.344 (3), b = 7.632 (3), c = 8.485 (3) A,
ꢀ = 105.25 (3), ꢁ = 109.26 (4), ꢂ = 91.26 (3)ꢁ].
0.0378 (14)
0.0399 (15)
0.0368 (14)
0.0377 (14)
0.0382 (14)
0.0356 (13)
C4
C5
C6
3. Discussion
3.1. General
CBF-A 173 K
C1
C2
C3
C4
C5
C6
N1
N2
O1
O2
Cl4
Cl5
0.5438 (5)
0.2394 (5)
0.2469 (5)
0.3893 (12)
0.5146 (12)
0.5003 (12)
0.3655 (11)
0.0959 (2)
0.1079 (3)
0.0266 (2)
0.2850 (10)
0.1901 (10)
0.1527 (12)
0.2069 (15)
0.2997 (14)
0.3403 (12)
0.3104 (3)
0.1540 (4)
0.3865 (3)
0.2272 (3)
0.1563 (9)
0.3490 (9)
0.0196 (6)
0.0216 (7)
0.0227 (9)
0.0202 (8)
0.0209 (8)
0.0238 (10)
0.0247 (4)
0.0272 (4)
0.0335 (4)
0.0299 (4)
0.0276 (5)
0.0257 (4)
The atom labeling and anisotropic displacement
ellipsoids for CPA and BPA are shown in Fig. 1. The
®nal coordinates are given in Table 2.² Bond lengths and
angles are given in Table 3.
0.4298 (4)
0.3798 (7)
0.4430 (9)
0.5623 (8)
0.6137 (7)
0.5718 (2)
0.3874 (2)
0.6566 (2)
The bond lengths and angles are normal. There is
good agreement between the chemically equivalent
distances and angles in BPA, which in turn agree with
the distances and angles in CPA. Both molecules deviate
signi®cantly from planarity. In both cases the benzene
rings are close to planarity but the Cl and Br atoms are
bent out of the plane in opposite directions by 0.02±
0.47343 (13) 0.0101 (2)
0.6869 (7)
0.6699 (6)
0.3794 (6)
0.6330 (5)
CBF-B 173 K
C1²
C2²
C3²
C4²
C5²
C6²
Ê
0.07 A. The anhydride C atoms are bent out of the plane
0.1896 (3)
0.4137 (4)
0.3091 (3)
0.2584 (5)
0.3157 (6)
0.4240 (6)
0.4745 (5)
0.4407 (4)
0.2686 (4)
0.5204 (3)
0.3488 (3)
0.2592 (5)
0.4910 (5)
0.3440 (5)
0.4509 (4)
0.5019 (6)
0.4425 (8)
0.3319 (8)
0.2812 (6)
0.3173 (4)
0.4931 (4)
0.2363 (4)
0.4117 (4)
0.4991 (10)
0.2623 (9)
0.2518 (6)
0.2555 (6)
0.1212 (7)
0.0075 (7)
0.0108 (7)
0.1173 (7)
0.3937 (6)
0.3948 (6)
0.4562 (4)
0.4852 (4)
0.1750 (5)
0.1826 (4)
0.2617 (7)
0.2462 (7)
0.1032 (9)
0.0144 (8)
0.0021 (8)
0.1388 (8)
0.4078 (6)
0.3775 (7)
0.4847 (5)
0.4822 (5)
0.1915 (7)
0.1566 (6)
0.0242 (12)
0.0250 (11)
0.0273 (10)
0.0242 (8)
0.0211 (8)
0.0243 (10)
0.0343 (11)
in a similar fashion and to about the same extent.
Rudman (1971) found similar non-planarity in tetra-
chlorophthalic anhydride.
The displacement ellipsoids and the atom labeling for
the low-temperature structures of CBF are shown in Fig.
0.0951 (3)
0.0561 (5)
0.1142 (6)
0.2121 (6)
0.2510 (5)
0.2067 (4)
0.0517 (4)
0.2753 (4)
0.1195 (3)
0.0718 (7)
0.2816 (7)
0.1242 (4)
0.2116 (4)
0.2549 (7)
0.2079 (9)
0.1173 (9)
0.0744 (7)
0.1008 (5)
0.2445 (6)
0.0338 (6)
0.1769 (5)
N1²
N2²
0.0363 (11) 2, where the disorder is shown also. The ®nal positional
0.0460 (10)
0.0401 (10)
0.0368 (7)
0.0241 (5)
0.0247 (13)
0.0237 (14)
0.0233 (12)
0.0206 (9)
0.0240 (9)
0.0274 (12)
O1²
O2²
Cl4²
Cl5²
C21³
C22³
C23³
C24³
C25³
C26³
N21³
N22³
O21³
O22³
parameters are given in Table 2. Bond lengths and
angles are given in Table 4.
As can be seen in Fig. 2, the relationship between the
two orientations of the disordered molecules is similar in
all three polymorphs. In A and C the two orientations
are related by a crystallographic twofold axis and are
equally populated. In B there is no crystallographic
symmetry relating the two orientations; they are related
0.0338 (14) by a rotation of 177.6ꢁ around an axis that is 1.3ꢁ away
0.0340 (14)
0.0381 (15)
0.036 (2)
0.039 (2)
0.0410 (14)
from being parallel to the c axis. This can be described as
² Supplementary data for this paper are available from the IUCr
electronic archives (Reference: BK0061). Services for accessing these
data are described at the back of the journal.
Cl24³ 0.2563 (12)
Cl25³ 0.0625 (13)

536
PACKING SIMILARITIES OF THREE ISOSTERIC MOLECULES
ꢁ
ꢁ
Ê
Table 3. Selected geometric parameters (A, )
Ê
Table 4. Bond lengths (A) and angles ( ) in CBF at low
temperature
CPA
BPA
Average values weighted to allow for the standard uncertainties in the
individual values.
Cl4(Br4)ÐC4
Cl5(Br5)ÐC5
C1ÐC7
C2ÐC8
C7ÐO1
1.888 (6)
1.885 (7)
1.486 (9)
1.476 (10)
1.179 (9)
1.403 (9)
1.408 (8)
1.179 (9)
1.376 (9)
1.385 (9)
1.405 (9)
1.376 (9)
1.401 (9)
1.386 (9)
1.718 (1)
1.471 (2)
CBF-A
CBF-B
CBF-C
Average
C1ÐC2
C2ÐC3
C3ÐC4
C4ÐC5
C5ÐC6
C6ÐC1
C1ÐN1
C2ÐN2
N1ÐO1
N1ÐO2
N2ÐO2
C4ÐCl4
C5ÐCl5
1.415 (5)
1.402 (9)
1.363 (12)
1.467 (4)
1.352 (12)
1.411 (8)
1.321 (5)
1.341 (5)
1.233 (3)
1.441 (3)
1.378 (3)
1.717 (11)
1.729 (10)
1.397 (5)
1.413 (7)
1.351 (8)
1.447 (4)
1.361 (8)
1.418 (7)
1.340 (7)
1.335 (7)
1.230 (5)
1.440 (5)
1.377 (5)
1.716 (7)
1.740 (6)
1.403 (7)
1.427 (10)
1.370 (15)
1.472 (6)
1.350 (15)
1.405 (9)
1.332 (8)
1.350 (9)
1.237 (5)
1.454 (5)
1.383 (5)
1.717 (14)
1.732 (14)
1.405 (3)
1.413 (5)
1.357 (6)
1.460 (3)
1.357 (6)
1.412 (5)
1.328 (4)
1.341 (4)
1.233 (2)
1.444 (2)
1.379 (2)
1.716 (5)
1.736 (5)
1.191 (2)
1.395 (2)
C7ÐO2
C8ÐO2
C8ÐO3
C1ÐC2(C10)
C2ÐC3
C3ÐC4
1.389 (2)
C4(C50)ÐC5
C5ÐC6
C6ÐC1
1.412 (3)
1.388 (2)
1.376 (2)
C6ÐC1ÐC2(C10)
C1ÐC2ÐC3
C2ÐC3ÐC4
C3ÐC4ÐC5
C4(C50)ÐC5ÐC6
C5ÐC6ÐC1
C2(C10)ÐC1ÐC7
C1ÐC2ÐC8
C1ÐC7ÐO2
121.83 (7)
122.4 (6)
122.4 (7)
115.4 (6)
122.3 (6)
121.6 (6)
115.9 (6)
108.0 (6)
108.5 (6)
106.8 (6)
106.8 (6)
110.0 (5)
121.2 (6)
120.9 (7)
121.1 (5)
121.3 (5)
C6ÐC1ÐC2
C1ÐC2ÐC3
C2ÐC3ÐC4
C3ÐC4ÐC5
C4ÐC5ÐC6
C5ÐC6ÐC1
C2ÐC1ÐN1
C1ÐC2ÐN2
C1ÐN1ÐO2
C2ÐN2ÐO2
N1ÐO2ÐN2
O1ÐN1ÐO2
C5ÐC4ÐCl4
C4ÐC5ÐCl5
124.8 (7)
118.4 (6)
119.0 (8)
120.4 (11)
122.8 (11)
114.7 (7)
108.2 (4)
110.5 (4)
106.3 (3)
105.8 (2)
109.1 (2)
118.2 (2)
122.1 (10)
114.5 (10)
123.2 (6)
120.3 (6)
116.9 (5)
122.7 (7)
121.4 (6)
115.5 (5)
107.2 (5)
112.2 (6)
106.1 (4)
105.1 (4)
109.3 (3)
118.2 (4)
117.9 (6)
119.1 (6)
126.3 (9)
117.1 (9)
120.3 (11)
117.6 (17)
125.2 (17)
113.4 (11)
108.9 (5)
111.2 (5)
105.2 (4)
105.0 (4)
109.8 (3)
117.8 (3)
124.5 (16)
112.5 (16)
124.4 (4)
118.9 (4)
117.9 (4)
121.6 (6)
122.0 (5)
115.0 (5)
108.1 (3)
111.1 (3)
106.0 (2)
105.6 (2)
109.3 (1)
118.1 (2)
119.5 (5)
117.9 (5)
120.99 (7)
117.15 (10)
107.68 (7)
107.34 (11)
C2ÐC8ÐO2
C7ÐO2ÐC8(C70)
O1ÐC7ÐO2
O3ÐC8ÐO2
109.9 (2)
120.69 (14)
C5ÐC4(C50)ÐCl4(Br4)
C4(C50)ÐC5ÐCl5(Br5)
119.8 (2)
rotation around a pseudo-twofold axis. The distribution
between the two sites has the more populated orienta-
tion with an occupancy of 0.622 (4) at 173 K and
0.610 (3) at 298 K for the crystal reported in Table 1. The
closeness of these two values suggests that the disorder
might be frozen in, and that these two values are the
same within experimental error. To test this, complete
data sets were collected at the same two temperatures
on a different crystal. Using the same constraints, these
re®ned to the same results within experimental error. In
Fig. 2. Polymorphs A, B and C of CBF. Displacement ellipsoids are shown at the 50% probability level and H atoms are shown as spheres of
arbitrary radii. The two orientations for each polymorph are shown in the correct juxtaposition. In A and C the molecules are located on a
twofold axis and present in equal amounts. In B the shaded molecule is the predominant orientation (about 60%).

OJALA, OJALA, BRITTON AND GOUGOUTAS
537
particular the occupancies were 0.624 (2) at 173 K and with adjacent ribbons oriented randomly. All three of
0.611 (3) at 298 K. We conclude that the difference in these arrangements would give the same diffraction
the occupancy between the two temperatures, although pattern. Welberry has described the use of diffuse
small, is real. It should be mentioned that data for the scattering to obtain information about the short-range
®rst crystal were collected at room temperature ®rst and scattering (see Khanna & Welberry, 1990, and refer-
low temperature next, while this order was reversed for ences therein). Such data would be necessary to say
the data collection on the second crystal. If we assume more about the details of the disorder.
that there are no cooperative effects, then the enthalpy
In view of the disorder, it is to be expected that the
and entropy changes can be estimated in the usual way bond lengths and angles have larger-than-usual errors.
(Sim, 1990). For the reaction in which the major The results in Table 4 show that the constrained
component converts to the minor one, using the re®nement does give reasonable values, values which
averages of the occupancies quoted above, ÁH = agree, where comparison is possible, with those found in
1
+0.18 (6) kJ mol and ÁS = 3.1 (3) J K 1 mol 1. In benzfurazan 1-oxide (Britton & Olson, 1979) and 5-
any case, the molecules are similarly disordered in all methylbenzfurazan 1-oxide (Britton & Noland, 1972).
three polymorphs, and there is no polymorph, discov- Since the errors are large, we will not discuss these
ered so far, in which the molecules are ordered.
values at length, except to note that the difference
It should be noted that the details of the disorder between the two CÐCl bond lengths seems to be real.
cannot be determined from the Bragg diffraction data.
To illustrate, three simple models of the disorder are:
3.2. Packing
(i) each molecule could have a random orientation
unaffected by the orientations of adjacent molecules;
(ii) in each ribbon (see x3.2) all the molecules could
have the same orientation with adjacent ribbons
oriented randomly; (iii) in each ribbon successive
molecules could alternate between the two orientations
3.2.1. Packing of CPA. We will start the discussion
with the packing of CPA, since this is what Megaw
(1973) calls the aristotype, the simplest and most
Å
Fig. 4. The packing of BPA. Below: view perpendicular to the (101)
Å
Fig. 3. The packing of CPA. Below: view perpendicular to the (101)
plane showing a two-dimensional layer of molecules. Two molecules
each from the next layers on either side are shown in outline.
Above: view along the b axis showing the same molecules as the
lower view.
plane showing a two-dimensional layer of molecules. Two molecules
from the next layer are shown in outline. Above: view along the b
axis showing the same molecules as the lower view.

538
PACKING SIMILARITIES OF THREE ISOSTERIC MOLECULES
symmetric of the ®ve packings we will discuss here. A the van der Waals distances suggested by Pauling (1960)
view of the packing is shown in Fig. 3. The molecules and by Bondi (1964). The packings of BPA, CBF-A,
Å
form two-dimensional layers parallel to the (101) plane; CBF-B and CBF-C are shown, respectively, in Figs. 4, 5,
the molecular twofold axis coincides with the crystal- 6 and 7. All of the comparisons will be made using the
lographic twofold axis. In the b direction adjacent room-temperature distances since the structures of CPA
molecules interact to form ribbons through ClÁ Á ÁO and BPA were determined only at room temperature.
contacts that are slightly longer than the usual van der
3.2.2. Contacts along the ribbons. All ®ve of the
Waals distance. In the direction at right angles to b structures have ribbons of molecules along the twofold
adjacent ribbons of molecules interact through HÁ Á ÁO or pseudo-twofold axes with halogen atoms in one
contacts that are slightly shorter than the distance molecule in contact with O atoms in the adjacent
expected from the usual van der Waals radii of H and O, molecule. This is the one feature that all ®ve structures
and which might be described as CÐHÁ Á ÁO hydrogen have in common. The XÁ Á ÁO distances and CÐXÁ Á ÁO
bonds. In addition there are ClÁ Á ÁCl and ClÁ Á ÁO inter- angles are given in Table 5. For the CBF polymorphs, the
actions between molecules in adjacent ribbons, although two short distances between molecules with identical
the distances here are larger than the usual van der orientations with respect to the disorder are given ®rst,
Waals distances. The layers stack to form the three- followed by the two short distances between molecules
dimensional structure, again with no unusually short with opposite orientations. If the distances are compared
distances between the layers.
with the usual van der Waals distances (ClÁ Á ÁO: Pauling
Ê Ê Ê
We will compare the packing in the ®ve structures 3.20 A, Bondi 3.27 A; BrÁ Á ÁO: Pauling 3.35 A, Bondi
Ê
reported here; ®rst, with respect to the contacts along 3.31 A), the distances in CPA and BPA are slightly
the ribbons of molecules parallel to the twofold or longer than usual and the shortest distances in all the
pseudo-twofold axes; second, with respect to the CBF polymorphs are the usual length or slightly shorter.
contacts at the edges of these ribbons; third, with respect There are, however, two more ClÁ Á ÁO contacts in CPA
to the face-to-face contacts between these ribbons. In and this may be the controlling factor since the melting
each case the contact distances will be compared with point of CPA is about 50 K higher than that of CBF.
Allen et al. (1997) surveyed the Cambridge Crystal-
lographic Database for ClÁ Á ÁO distances for O atoms in
nitro groups. It is reasonable to expect similar interac-
tions with O atoms in benzfurazan oxide groups, and this
Å
Fig. 5. The packing of CBF-A. Below: view perpendicular to the (101)
Fig. 6. The packing of CBF-B. Below: view perpendicular to the (011)
plane showing a two-dimensional layer of molecules. Two molecules
from the next layer are shown in outline. Above: view along the c
axis showing the same molecules as the lower view.
plane showing a two-dimensional layer of molecules. Two molecules
from the next layer are shown in outline. Above: view along the b
axis showing the same molecules as the lower view.

OJALA, OJALA, BRITTON AND GOUGOUTAS
539
ꢁ
Ê
Table 5. Distances (A) and angles ( ) in the ClÁ Á ÁO and BrÁ Á ÁO contacts
173 K
297 K
173 K
297 K
CPA
Cl5Á Á ÁO1i
Ð
Ð
3.33
3.33
C5ÐCl5Á Á ÁO1i
C5ÐCl5Á Á ÁO2i
Ð
Ð
161
122
Cl5Á Á ÁO2i
BPA
Br4Á Á ÁO3i
Br4Á Á ÁO2i
Br5Á Á ÁO1i
Br5Á Á ÁO2i
Ð
Ð
Ð
Ð
3.43
3.49
3.40
3.44
C4ÐBr4Á Á ÁO3i
C4ÐBr4Á Á ÁO2i
C5ÐBr5Á Á ÁO1i
C5ÐBr5Á Á ÁO2i
Ð
Ð
Ð
Ð
156
119
159
120
CBF polymorph A
Cl4Á Á ÁO2i
3.09
3.19
3.05
3.28
3.10
3.26
3.07
3.37
C4ÐCl4Á Á ÁO2i
C5ÐCl5Á Á ÁO1i
C4ÐCl4Á Á ÁO1ii
C5ÐCl5Á Á ÁO2ii
131
156
162
127
133
155
162
127
Cl5Á Á ÁO1i
Cl4Á Á ÁO1ii
Cl5Á Á ÁO2ii
CBF polymorph B
Cl4Á Á ÁO2iii
3.21
3.28
3.13
3.35
3.24
3.28
3.17
3.35
3.20
3.36
3.15
3.47
3.24
3.38
3.18
3.47
C4ÐCl4Á Á ÁO2iii
135
152
161
127
133
155
159
130
138
150
164
126
136
151
162
128
Cl5Á Á ÁO1iii
C5ÐCl5Á Á ÁO1iii
Cl4Á Á ÁO21iii
Cl5Á Á ÁO22iii
Cl24Á Á ÁO22iii
Cl25Á Á ÁO21iii
Cl24Á Á ÁO1iii
Cl25Á Á ÁO2iii
C4ÐCl4Á Á ÁO21iii
C5ÐCl5Á Á ÁO22iii
C24ÐCl24Á Á ÁO22iii
C25ÐCl25Á Á ÁO21iii
C24ÐCl24Á Á ÁO1iii
C25ÐCl25Á Á ÁO2iii
CBF polymorph C
Cl4Á Á ÁO2i
3.19
3.36
3.13
3.46
3.20
3.41
3.16
3.52
C4ÐCl4Á Á ÁO2i
C5ÐCl5Á Á ÁO1i
C4ÐCl4Á Á ÁO1ii
C5ÐCl5Á Á ÁO2ii
134
154
160
129
138
151
164
127
Cl5Á Á ÁO1i
Cl4Á Á ÁO1ii
Cl5Á Á ÁO2ii
Symmetry codes: (i) x, 1 ‡ y, z; (ii) 1 x, 1 ‡ y, 12 z; (iii) x, y, 1 ‡ z. For contacts with (i) both molecules have the same orientation, for contacts
with (ii) they have opposite orientations with respect to the twofold axis.
is what we ®nd. In A the distances are toward the lower the usual van der Waals distances (HÁ Á ÁO: Pauling
Ê
Ê
Ê
end of the range of values found by Allen et al., while in 2.60 A, Bondi 2.72 A; HÁ Á ÁN: Pauling 2.70 A, Bondi
Ê
B and C they are closer to the average values. The trends 2.75 A), it can be seen that in every structure there are
in the distances from A to B to C are consistent with the HÁ Á ÁO distances that are shorter than usual. However,
lengths of the axes in the repeat directions of the the HÁ Á ÁN distances that occur in all the CBF structures
Ê
Ê
ribbons: in A b = 8.871 A, in B c = 9.011 A, in C b = do not show any shortening even though they are as well
Ê
9.055 A. It should also be noted that in all three poly- aligned as the HÁ Á ÁO distances.
morphs the shortest ClÁ Á ÁO distances, which presumably
Inspection of Figs. 3±7 suggests that XÁ Á ÁX and
correspond to the strongest interactions, change the XÁ Á ÁO, where X is Cl or Br, contacts between adjacent
least on increasing the temperature. ribbons should also be considered. Table 7 shows the
3.2.3. Contacts at the edges of the ribbons. The exis- XÁ Á ÁX distances. These appear to be normal van der
tence of CÐHÁ Á ÁX (X = N or O) hydrogen bonds has Waals distances or slightly larger (ClÁ Á ÁCl: Pauling
Ê
Ê
Ê
been the subject of some controversy, but the consensus 3.60 A, Bondi 3.52 A; BrÁ Á ÁBr: Pauling 3.90 A, Bondi
Ê
Ê
is that they are real (see, for example, Taylor & 3.70 A). The XÁ Á ÁO distances (not given) are all 0.2 A or
Kennard, 1982; Berkovitch-Yellin & Leiserowitz, 1984; more larger then the van der Waals distances.
Desiraju, 1991). They are present in CPA and BPA with
3.2.4. Face-to-face contacts between the ribbons. In all
the graph-set motif R22(10) (Etter, 1990; Bernstein et al., of the structures except CBF-B the ribbons aggregate
1995) and in all three polymorphs of CBF with the into two-dimensional layers that are approximately
graph-set motif R22(9). The motif occurs on both sides of planar. In CBF-B the layers are considerably puckered.
the molecule in CPA, BPA, and CBF-A and -C and on Table 8 gives the metrical details. Approximately
one side in CBF-B. On the other side in CBF-B, there parallel molecules in adjacent layers make partial face-
are similar contacts, but the CÐHÁ Á ÁX angles are much to-face contacts. As can be seen in Figs. 3±7, it is in these
farther from the ideal of 180ꢁ. Details of these contacts contacts that the greatest differences between the
are given in Table 6. If the distances are compared with structures occur.

540
PACKING SIMILARITIES OF THREE ISOSTERIC MOLECULES
ꢁ
Ê
Ê
Table 6. Distances (A) and angles ( ) in the CÐHÁ Á ÁO Table 7. Distances (A) in the ClÁ Á ÁCl and BrÁ Á ÁBr
and CÐHÁ Á ÁN contacts at room temperature
contacts at room temperature
HÁ Á ÁX
Symmetry code
for N or O
Symmetry code for second Cl/Br
distance
Angle
CPA
Cl5Á Á ÁCl5
3
2
CPA
C6ÐH6Á Á ÁO1
4.05
x, 32 y, 1
z
z
3
2
2.46
156
x, 12 y, 1
z
z
BPA
Br4Á Á ÁBr4
1
BPA
C3ÐH3Á Á ÁO3
3.85
3.72
x, 32 y,
x, 32 y, 1
z
z
2
1
3
2.46
2.45
166
159
x, 12 y,
z
z
Br5Á Á ÁBr5
2
2
3
C6ÐH6Á Á ÁO1
x, 12 y, 1
2
CBF-A
1
2
CBF-A
Cl4Á Á ÁCl4
Cl4Á Á ÁCl5
Cl5Á Á ÁCl5
3.79
3.67
3.57
x, 32 y,
1
2
1
2
1
2
1
2
1
2
z
C3ÐH3Á Á ÁO1
C3ÐH3Á Á ÁN2
C6ÐH6Á Á ÁO1
C6ÐH6Á Á ÁN2
2.28
2.72
2.51
2.83
161
163
148
164
‡ x,
y,
‡ z
‡ x, 32 y,
‡ z
1
2
3
3
2
x, 12 y,
x, 32 y, 1
x, 12 y, 1
z
2
1
‡ x, 12 y, 12 ‡ z
CBF-B
2
1
2
1
2
1
1
2
1
Cl4Á Á ÁCl5
Cl4Á Á ÁCl24
Cl5Á Á ÁCl25
Cl24Á Á ÁCl25
3.57
3.62
3.49
3.53
x,
x,
‡ y, 12
‡ y, 12
z
z
CBF-B
2
1
2
1
1
2
C6ÐH6Á Á ÁO21
C6ÐH6Á Á ÁN2
C23ÐH23Á Á ÁO21
C23ÐH23Á Á ÁN2
C3ÐH3Á Á ÁO1
C3ÐH3Á Á ÁN22
C26ÐH26Á Á ÁO1
C26ÐH26Á Á ÁN22
2.67
2.84
2.29
2.57
2.67
3.03
2.77
3.02
159
171
174
166
113
112
120
122
x, 12 ‡ y,
z
z
z
z
z
z
z
z
x, 12 ‡ y, 12
z
z
2
1
2
1
x, 12 ‡ y,
x, 12 ‡ y, 12
2
2
1
1
2
x, 12 ‡ y,
2
1
1
2
x, 12 ‡ y,
CBF-C
2
1
1
2
1
2
1
1
2
1
2
1
1
2
x,
x,
x,
x,
‡ y,
‡ y,
‡ y,
‡ y,
Cl4Á Á ÁCl4
Cl4Á Á ÁCl5
Cl5Á Á ÁCl5
3.73
3.56
3.41
x, 32 y, 1
‡ x, 32 y,
x, 32 y, 2
z
2
1
2
1
2
1
2
1
2
1
2
‡ z
3
2
z
2
2
1
1
2
2
CBF-C
virtually the same. The two-dimensional unit cell, which
was rectangular in CPA, has a cell angle of 90.11ꢁ in
BPA.
1
2
1
2
1
2
z
z
C3ÐH3Á Á ÁO1
C3ÐH3Á Á ÁN2
C6ÐH6Á Á ÁO1
C6ÐH6Á Á ÁN2
2.27
2.74
2.41
2.77
160
153
165
172
‡ x,
y,
‡ z
1
2
3
x, 12 y, 1
x, 12 y, 2
2
‡ x, 12 y, 12 ‡ z
1
2
In CPA a Cl atom in one molecule lies over the six-
membered ring in the next; the other Cl atom lies under
a six-membered ring on the opposite side. This
arrangement is duplicated in CBF-A. In BPA one Br
atom is similarly next to a six-membered ring, but the
other is next to a ®ve-membered ring. In CBF-C both Cl
atoms are next to ®ve-membered rings. In CBF-B a
given molecule has a parallel neighbor on only one side,
but, as can be seen in Fig. 8, the overlap on this side is
virtually identical to that in CBF-A.
3.2.5. Summary. As stated above CPA is the aristotype
of this series. The CPA molecule has its twofold axis
coincident with the twofold axis of the space group. The
molecules form two-dimensional layers with rectangular
symmetry. Face-to-face molecules in adjacent layers are
antiparallel to each other.
Again following Megaw (1973), the BPA structure is a
hettotype (less symmetric variation) of the CPA struc-
ture. Again the molecules form two-dimensional layers
with the molecular twofold axis now aligned with a
pseudo-twofold axis in the layer. In this structure,
however, there is no crystallographic twofold axis. The
face-to-face interactions on opposite sides of the mole-
cule are no longer equivalent and the molecular over-
laps on the two sides are different. Thus, the interlayer
contacts have changed considerably, leading to the lower
symmetry for BPA, but the packing within the layers is
Å
Fig. 7. The packing of CBF-C. Below: view perpendicular to the (101)
plane showing a two-dimensional layer of molecules. Two molecules
from the next layer are shown in outline. Above: view along the b
axis showing the same molecules as the lower view.

OJALA, OJALA, BRITTON AND GOUGOUTAS
541
Table 8. Distances between planes at room temperature
Angle between
molecule
and
layer (ꢁ)
Perpendicular
distance
between
Distance
between
layers (A)
Layer
parallel to
Ê
Ê
molecules (A)
Å
CPA
BPA
CBF-A
CBF-B
CBF-C
101
3.23
3.35
3.23
3.65
3.23
5.9
5.9
5.7
27.2²
4.0
3.40
3.45, 3.55
3.38
3.42²
3.44
Å
101
Å
101
011
Å
101
² Based on the mean plane of the two disordered molecules.
The CBF-A structure is a hettotype of a different sort. the face-to-face interactions occur only on one side of
It is a true isomorph of CPA, but the CBF molecule does the molecule. However, the remaining edge-to-edge and
not contain a twofold axis and disorder about a crys- face-to-face contacts are virtually identical to those in
tallographic twofold axis is required to make the mole- CBF-A.
cules
pseudo-isosteric.
Similar
intermolecular
The structure of CBF-C at ®rst glance might be
interactions occur in both structures along the ribbons regarded as isomorphous with CBF-A; they have very
and at the sides of the ribbons, although both the similar unit-cell dimensions and the same space group.
numbers and the types change. The face-to-face contacts However, the cell dimensions do not correspond without
between molecules in adjacent layers are quite similar to exchanging the a and c directions in CBF-C. If this is
those in CPA.
performed, the space group becomes A2/a, a nonstan-
The CBF-B structure is the most different of the ®ve. dard setting of C2/c. Nevertheless, CBF-C is still a
It can be regarded as the CBF-A structure with the variation of the CBF-A structure. The two dimensional
layers puckered to the extent that they are no longer layer in CBF-C is virtually identical to that in CBF-A.
close to planar. The end-to-end contacts in the ribbons What is different is that the adjacent layers have been
stay the same, but the crystallographic twofold axis is displaced so that the face-to-face molecules now have
replaced by a pseudo-twofold axis. Half of the edge-to- parallel rather than antiparallel orientations.
edge contacts of the ribbons are changed drastically and
After accepting the disorder in the CBF molecules as
making them pseudo-isosteric with CPA and BPA, we
might summarize the overall relationships by noting that
CPA, CBF-A, BPA and CBF-C are all isostructural in
their two-dimensional layers, but only the ®rst two are
isostructural in three dimensions.
We thank Dr Victor G. Young Jr, Director of the
University of Minnesota X-ray Diffraction Laboratory,
for help in using the SMART system and the SHELXTL
programs, and for determining the twin plane in BPA.
References
Allen, F. H., Lommerse, J. P. M., Hoy, V. J., Howard, J. A. K. &
Desiraju, G. R. (1997). Acta Cryst. B53, 1006±1016.
Berkovitch-Yellin, Z. & Leiserowitz, L. (1984). Acta Cryst.
B40, 159±165.
Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995).
Angew. Chem. Int. Ed. Eng. 34, 1555±1573.
Beurskens, P. T. (1984). DIRDIF. Direct Methods for
Difference Structures ± an Automatic Procedure for Phase
Extension and Re®nement of Difference Structure Factors.
Technical Report 1984/1. Crystallography Laboratory, Toer-
nooiveld, 6525 ED Nijmegen, The Netherlands.
Bondi, A. (1964). J. Phys. Chem. 68, 441±451.
Fig. 8. Left: CBF-A. The four molecules in the upper right corner of
Fig. 5 viewed from a direction perpendicular to the plane of the
molecules. Right: CBF-B. Four molecules from two parallel ribbons
are viewed from a direction perpendicular to the plane of the
molecules.
Britton, D., Konnert, J., Hamer, J. & Trefonas, L. M. (1972).
Acta Cryst. B28, 1123±1126.

542
PACKING SIMILARITIES OF THREE ISOSTERIC MOLECULES
Britton, D. & Noland, W. E. (1972). Acta Cryst. B28, 1116± North, A. C. T., Phillips, D. C. & Mathews, F. S. (1968). Acta
1121. Cryst. A24, 351±359.
Britton, D. & Olson, J. M. (1979). Acta Cryst. B35, 3076±3078. Ojala, C. R. (1985). Crystallographic Aspects of some Isosteric
Desiraju, G. R. (1991). Acc. Chem. Res. 24, 290±296.
Duisenberg, A. J. M. (1992). J. Appl. Cryst. 25, 92±96.
Dunitz, J. D. (1995). Acta Cryst. B51, 619±631.
Enraf±Nonius (1989). CAD-4 Software. Enraf±Nonius, Delft,
The Netherlands.
Molecules. MS thesis, University of Minnesota, USA.
Ojala, W. H. (1986). Solid-State Structures and Properties of
some Organic Strict Isosteres. PhD thesis, University of
Minnesota, USA.
Pauling, L. (1960). The Nature of the Chemical Bond, 3rd ed, p.
260. Ithaca: Cornell University Press.
Rudman, R. (1971). Acta Cryst. B27, 262±269.
Etter, M. C. (1990). Acc. Chem. Res. 23, 120±126.
Gilmore, C. J. (1984). J. Appl. Cryst. 17, 42±46.
Gougoutas, J. Z. (1983). Structural Aspects of Organic Sheldrick, G. M. (1994). SHELXTL. Version 5.0. Siemens
Peroxides. In The Chemistry of Peroxides, edited by S.
Analytical X-ray Instruments Inc., Madison, Wisconsin,
USA.
Khanna, R. & Welberry, T. R. (1990). Acta Cryst. A46, 975± Sheldrick, G. M. (1996). SADABS. Program for Empirical
Patai, pp. 375±415. New York: John Wiley and Sons.
È
979.
Absorption Correction. University of Gottingen, Germany.
Langmuir, I. (1919). J. Am. Chem. Soc. 41, 1543±1559.
Megaw, H. D. (1973). Crystal Structures: A Working Approach,
p. 216. W. B. Saunders Co: Philadelphia.
Mitscherlich, E. (1822). Ann. Chim. Phys. 19, 350±419.
Siemens (1995). ASTRO and SAINT. Data Collection and
Processing Software for the SMART System. Siemens
Analytical X-ray Instruments, Inc., Madison, Wisconsin,
USA.
Molecular Structure Corporation (1985). TEXSAN. TEXRAY Sim, G. A. (1990). Acta Cryst. B46, 676±682.
Structure Analysis Package. MSC, 3200 Research Forest Taylor, R. & Kennard, O. (1982). J. Am. Chem. Soc. 104, 5063±
Drive, The Woodlands, TX 77381, USA.
5070.