Weak interactions induce asymmetry in the crystal structures of triaryl derivatives of group 14 elements

Article abstract of DOI:10.1039/b105457k

A search of the CSD shows that 42% of compounds of type Ph₃EX (E = group 14 element, X = halogen) possess more than one molecule in the asymmetric unit (Z′ > 1). This phenomenon is rationalised in terms of CH···X hydrogen bonding and confirmed by the preparation of Ph₃GeI. A remarkable (Z′ = 4 structure of Ph₃GeCl is shown to arise from site-selective inclusion of Ph₃GeH impurity.

Full text of DOI:10.1039/b105457k

View Article Online / Journal Homepage / Table of Contents for this issue
Weak interactions induce asymmetry in the crystal structures
of triaryl derivatives of group 14 elements
Paul D. Prince, G. Sean McGrady* and Jonathan W. Steed*
Department of Chemistry, King’s College London, Strand, London, UK WC2R 2LS.
E-mail: jon.steed@kcl.ac.uk
Received (in Montpellier, France) 21st June 2001, Accepted 5th December 2001
First published as an Advance Article on the web
A search of the CSD shows that 42% of compounds of type Ph EX (E ¼ group 14 element, X ¼ halogen)
3
0
possess more than one molecule in the asymmetric unit (Z > 1). This phenomenon is rationalised in terms of
CHÁ Á ÁX hydrogen bonding and confirmed by the preparation of Ph GeI. A remarkable Z ¼ 4 structure of
0
3
3 3
Ph GeCl is shown to arise from site-selective inclusion of Ph GeH impurity.
At the last systematic count, 8.3% of the entries in the
Cambridge Structural Database (CSD) possessed more than
one crystallographically independent molecule, that is more
hydrogen-bonded rings of Ph
hydroxyl groups. Each motif interacts edge-to-face in a non-
crystallographically related fashion to its nearest neighbour.
The germanium analogue is isomorphous: clearly the hydroxyl
functionality dominates the crystal packing in these materials.
There has been much recent interest in crystal packing modes
of triaryl derivatives in general (e.g., complexes of triphenyl-
phosphine), which display reproducible edge-to-face (EF) p–p
3
SiOH units linked by the
1
2
0
than one molecule in the asymmetric unit, with Z values of
1
–6
1=2 , 1and 2 making up 95.3% of all crystal structures.
number of molecules in the asymmetric unit is represented by
The
0
the parameter Z , which is strictly defined as the number of
formula units in the unit cell (Z) divided by the number of
independent general positions. Attempts to isolate the factors
leading to non-crystallographic relationships between two or
1
2,13
interaction
enlarged 6PE (E6PE) and four-fold aryl embrace (4AE).
Calculated intermolecular interaction energies are À14.0 kcal
motifs termed six-fold phenyl embrace (6PE),
1
4
0
more molecules (i.e., Z > 1) have been frustrated by the
extreme complexity of the problem, and by the difficulty in
separating out the various intermolecular packing effects in the
À1
À1
mol per (PAr ) unit for each 6PE and À15.3 kcal mol per
6 6 4 3 3 3 4
3
2
(PAr)
for example.
3
In this study, we have examined compounds with the Ph E
unit in the E6PE in [Cu{P(C H -4-OCH ) } ]ClO ,
0
15,16
solid state. Interestingly, however, high Z values do occur
repeatedly within related series of compounds; notably, the
0
incidence of Z > 1rises significantly (to ca. 40%) for mono-
functionality (E ¼ group 14 element), and have sought to
1,3
alcohols. Similarly, we have recently shown that multiple
hydrogen-bonded structures involving 15-crown-5 often display
establish the presence or absence of a link between the EF
packing mode and Z in the absence of other strongly inter-
acting groups such as the hydroxyl functionality. In particular,
0
0
7
high Z values, although the reasons for this are complex. We
believe that the study of intermolecular interactions in mole-
0
we have chosen to study compounds of type Ph EX
3
cular crystals exhibiting Z > 1is of fundamental interest, since
(X ¼ halogen). Systems of type Ph
3
EH crystallise in a
remarkable series of polymorphs; we have subjected these to a
detailed study, the results of which will be reported elsewhere.
it represents a form of ‘frustration’ in which it is impossible to
accommodate simultaneously the requirements of close pack-
ing and the tendency to indulge in stronger directional inter-
actions such as hydrogen bonding. Hence, the degree to which
non-crystallographic packing occurs should be a function of
the importance of directional interactions in any one case.
Results
0
One of the pitfalls in analysing high Z structures is that
0
Database survey
mistakes in structure determination can artificially raise Z .
For example, simply assigning the uncommon space group P1
1
7,18
A search of the Cambridge Structural Database (CSD),
Table 1, reveals a total of 24 unique compounds containing a
four-coordinate group 14 central atom, a halogen and three
_
0
instead of the very common P1 immediately doubles Z , and
for this reason structures assigned to space group P1without
an obvious chemical cause (e.g., resolved chiral material) must
0
aryl substituents. Of these, a total of 10 have Z > 1 , a
remarkable 42% of the sample. The sample may be divided
8
be treated with some suspicion. Moreover, particularly for
,9
0
0
systems with very high Z values, a modulated description is
sometimes more appropriate than attempts to solve the
structure based on an enlarged unit cell with multiple inde-
into four classes on the basis of crystal symmetry and Z value:
0
1
(1) trigonal space groups with Z ¼
3
(3 structures); (2) tri-
(1structure, CSD code TPSNCL03);
(3) monoclinic (P or C) or triclinic with Z ¼ 1( 11 structures);
0
1
gonal with Z ¼ 1
3
1
0,11
0
pendent molecules.
In terms of the understanding of
0
intermolecular forces, however, modulated structures are of
equal relevance. Systems that crystallise in common—and, in
particular, centrosymmetric—space groups where there is a
(4) monoclinic P or triclinic, Z ¼ 2 (9 structures).
The trigonal compounds of class 1pack exclusively in an
end-to-end fashion: all comprise systems containing meta-
substituted aryl groups (m-xylyl, m-tolyl and m-anisole), and
are hence less likely to form a 6PE type of interaction. In
contrast, both unique molecules in Ph SnCl (TPSNCL03, class
2) form a perfect 6PE, with one molecule situated on the
0
clear chemical reason for the occurence of a high Z value are
of particular interest. Notable in this respect are compounds of
1
5
type Ph
0
3
EOH (E ¼ Si, Ge); one example is triphenylsilanol,
3
_
Z ¼ 8 (P1), which comprises two motifs of four-membered
DOI: 10.1039/b105457k
New J. Chem., 2002, 26, 457–461
457
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2002

0
+
Table 1 CSD search unique hits along with Z values and space group
HÁ Á ÁI distance of 3.356 A. If the phenyl groups i Vn i et hw i sA rs t ti rc ul ec O- nline
0
ture are replaced by mesityl, the same Z ¼ 2 structure is
0
CSD refcode
Compound
Z
Space group
obtained in two different cases with the inclusion of either
3
enclathrated toluene or CDCl . The p-CH moiety close to the
iodo substituent is replaced by the methyl group of an adjacent
mesityl moiety.
The remaining class 3 structures exhibit a variety of inter-
actions between molecules, none of which can be classified as
6 5 3 6 4 3
embraces. In the case of (C Cl ) SiCl and (o-C H OMe) SnI,
these take the form of offset face-to-face p–p stacking, while
a small number of long edge-to-face interactions relate the
6 3 2 3
sterically hindered molecules in (o,o-C H Me ) SiF and a
BARNUD
BENQIU
BRTPSN
CIPXUU
FAVYIK
HAKQOZ
HATVIH
HERQUQ
HIBTAN
JIPROP
LEHGEK
NIQZUI
NIRBAR
SEHROM
TICCAJ
TPGEBR
TPSNCL
TPSNCL03
VAFVUT
WEHYUD
WEHZAK
WEHZEO
WEHZEO01Ph
WEHZIS
Ph
Mes
Ph PbBr
(o-Tol) SiF
(C Cl )SiCl
SiF
OMe)
(9-Anthr) SiF
(p-t-BuC
(p-Methiophenyl)SnCl
(o-Me NCH )SiF
(m-Tol) SnCl
(m-Xyl) SnCl
Mes SnF
(m-C OMe)
3
SiCl
2
2
2
1
1
1
1
1
1
1
P2_
1
=b
3
PbBr
P1
P2
1
3
=c
3
P2
P2
P2
P2
P2
1
=c
=n
=c
=n
=n
6
5
1
1
1
1
C
30
H
27
S
3
(o-C
6
H
4
3
SnI
3
6
H
4
)
3
SnCl
C2=c
P2 =c
C2=c
R3
R3c
1
bridged cyclophane (CSD code HAKQOZ). The class 4
0
2
2
C
6
H
4
1
1
structures with Z ¼ 2 display a strong proclivity towards
3
3
1
3
3
another type of embrace motif via a 2 screw axis in which one
1
3
2
P2
R3
1
=n
of the phenyl positions in a 6PE is supplanted by a halogen
substituent (a five-fold phenyl embrace, viz Ph SiCl, Ph SnBr,
1
6
H
4
3
SnCl
3
3
3
Ph
3
Ph
3
Ph
3
GeBr
SnCl
SnCl
2
2
1
1
2
2
2
1
1
P2
P2_
1
=c
=a
3 3
Ph SnCl and Ph GeBr). The non-crystallographic relationship
1
1
between unique molecules in these structures, however,
involves short p-arylCHÁ Á ÁX interactions characteristic of
3
R3_
P1_
P1_
P1_
(C
6
Cl
5
)GeCl
1
9,20
Mes
Mes
3
SnIÁCDCl
3
weak hydrogen bonds
(as in WEHZEO), such as Ph GeBr,
3
+
3
SnIÁPhMe
in which HÁ Á ÁBr is 3.094 A and Ph
3
SnCl with HÁ Á ÁCl equal to
+
Ph
3
SnI
SnI
P1_
2.949 A. Even the sterically bulky mesityl substituents in (o,o-
_
3
P1
P2
0
C PbBr (Z ¼ 2, P1) exhibit a relatively close contact
between a CH proton and the bromo substituent, although an
6
H
3
Me
2
)
3
(p-C
6
H
4
OMe)
3
SnI
1
=c
3
offset face-to-face p-stacking interaction may also play a part
in this case. The only remaining structure, (o,o-C H Me ) -
6
3
2 3
_
crystallographic 3 axis and the other surrounding it on a
general site, as shown in Fig. 1. However, it is not the existence
of the 6PE interaction per se that leads to the presence of two
unique molecules (or fragments), since the pair of molecules
involved in the 6PE are related to one another by crystal-
lographic inversion. The non-crystallographic interaction is
similar to the 6PE, except that the position of two of the
phenyl groups is occupied now by the chloro substituents. The
SnF, adopts a non-crystallographic packing mode related to
TPSNCL03 (class 2), except that the bulky mesityl substituents
prevent the close pairwise approach of the aryl CH unit and
fluoro substituent. The interaction thus resembles an unsym-
metrical four-fold phenyl embrace.
interaction is characterised by two strong aryl CHÁ Á ÁCl inter-
Synthesis and structure
+
actions, with short HÁ Á ÁCl distances of 3.43 and 3.47 A.
Moving to classes 3 and 4, it appears again that the existence
of a 6PE interaction is not a primary factor in determining
class 3 or 4 behaviour. Thus, the two structures belonging to
The database survey reveals that within this class of com-
0
pounds, structures with Z > 1are more common for Ph
3
EX
(X ¼ Cl, Br or I) species, suggesting that the observed CHÁ Á ÁX
class 3 in space group C2=c, specifically (o-C
6
H
4
CH
SnCl, each exhibit 6PE structures with
SnI (CSD code WEHZEO01), which
2
NMe
2
)
3
-
interactions may be important. Accordingly, we decided to
t
Bu )
SiF and (p-C
0
6
H
4
3
3 3 3
examine Ph GeF, Ph GeCl and Ph GeI, the structures of
Z ¼ 1. Similarly, Ph
3
which were absent from the CSD. These compounds are
commercially available, however, for reasons of economy they
were prepared by reaction of the appropriate halide with either
_
crystallises in space group P1, exhibits a 6PE structure with
Z ¼ 1. Interestingly, however, this complex is polymorphic,
0
_
0
forming a second P1 structure with Z ¼ 2 (WEHZEO), which
Ph
3
GeBr or Ph
The X-ray crystal structure
3
GeH (see Experimental).
of Ph GeI (1) proved also to
contains two independent molecules, each of which forms
6PE interactions via a crystallographic inversion operation.
The interaction between the two independent molecules is
_
3
0
exhibit Z ¼ 2 (space group P1). Both unique molecules form
full 6PE structures via crystallographic inversion. The inter-
action between the independent molecules consists of CHÁ Á ÁI
hydrogen bonds and a 4PE motif combined with a single EF
characterised by a single short p-arylCHÁ Á ÁI contact, with an
interaction, isomorphous with Ph SnI (WEHZEO); Fig. 2. The
3
fluoride, Ph
analogous to those previously discussed, with Z ¼ 1. The
3
GeF (2), exhibits a rather different structure, not
0
packing comprises slightly offset 6PE interactions (Fig. 3)
between pairs of molecules with GeÁ Á ÁGe distances of 5.75 A.
+
Each molecular pair interacts with surrounding molecules via
two CHÁ Á ÁF hydrogen bonds with CÁ Á ÁF distances of 3.28 and
+
3
.32 A.
In our studies, Ph GeCl (3) was found to crystallise in the
common monoclinic space group P2 =c with a total of four
3
1
0
independent molecules (i.e., Z ¼ 4, Z ¼ 16). The structure
closely resembles the bromide analogue (CSD code TPGEBR),
and the unit cell parameters are likewise similar to the fivefold
embrace found in Ph
0
3
SiCl (BARNUD). Both TPGEBR and
BARNUD exhibit Z ¼ 2. In Ph GeCl the length of the b axis
3
0
is approximately doubled, resulting in Z ¼ 4, meaning that all
four molecules of the two crystallographically unrelated 6PE
pairs are unique. This, in turn, seems to be linked to the
absence of inversion symmetry for one 6PE pair in particular,
3
Fig. 1 The structure of crystalline Ph SnCl, showing the central six-
fold phenyl embrace pair surrounded by three other similar (crystal-
lographically unique) dimers.
4
58
New J. Chem., 2002, 26, 457–461

View Article Online
Fig. 2 The two crystallographically independent molecules in
Ph GeI. The bromide analogue exhibits two similar independent six-
3
0
fold phenyl embrace motifs with one half of each pair unique (Z ¼ 2).
+
+
Selected intermolecular distances and angles: C(4)Á Á ÁI(2) 4.051 A,
+
+
ꢀ
ꢀ
H(4)Á Á ÁI(2) 3.28 A, C(4)–H(4)Á Á ÁI(2) 140 , C(10)Á Á ÁI(1) 3.869 A,
H(10)Á Á ÁI(1) 3.33 A, C(10)–H(10)Á Á ÁI(1) 118 .
centred on Ge(1) and Ge(2). The GeÁ Á ÁGe distance in this 6PE
+
+
pair is very short at 5.427 A (Fig. 4), compared with 5.533 A
+
in the Ge(3)Á Á ÁGe(4) pair, 5.565 A in Ph SnCl and even lon-
3
Fig. 4 (a) Two independent six-fold embrace pairs in the asymmetric
0
ger distances in the bromide analogue. Most interestingly, it is
the Ge(1)Á Á ÁGe(2) pair that forms the least number of aryl
CHÁ Á ÁCl interactions (only one appreciably short one—see
figure caption), suggesting that in the absence of CHÁ Á ÁCl
bonds the 6PE interaction becomes stronger and slightly less
symmetrical (the lowered symmetry arising from slight tor-
sional twists in the aryl rings).
unit of Ph GeCl doped with Ph
is highly compressed, resulting in loss of inversion symmetry. (b) Plane
3
3
GeH (Z ¼ 4). The Ge(1)Á Á ÁGe(2) pair
view of the Ge(1)Á Á ÁGe(2) six-fold phenyl embrace. Selected inter-
+
molecular distances and angles: C(57)Á Á ÁCl(1) 3.665 A, H(57)Á Á ÁCl(1)
+
+
ꢀ
3
2
.01 A, C(57)–H(57)Á Á ÁCl(1) 130 ; C(3)Á Á ÁCl(3) 3.479 A, H(3)Á Á ÁCl(3)
+ +
ꢀ
.91 A, C(57)–H(57)Á Á ÁCl(1) 120 ; C(33)Á Á ÁCl(4) 3.730 A,
+
+
ꢀ
H(33)Á Á ÁCl(4) 3.04 A, C(33)–H(33)Á Á ÁCl(4) 138 ; C(45)Á Á ÁCl(4) 3.684 A,
+
ꢀ
H(45)Á Á ÁCl(4) 3.07 A, C(45)–H(45)Á Á ÁCl(4) 127 .
During the course of this study the structure of a commer-
2
1
3
cial sample of Ph GeCl was reported by Tiekink et al. This
independent work produced a structure closely related to that
of 3, except that the crystallographic b axis is halved, resulting
in the more common Z ¼ 2, structure 4. We were readily able
to reproduce this structure from a variety of solvents (toluene,
chloroform, dichloromethane) using a sample of Ph GeCl
3
0
purchased from Aldrich. However, numerous independent
syntheses and crystallisations of Ph GeCl prepared from
3
0
Ph GeH reliably gave the Z ¼ 4 structure, compound 3. We
considered the possibility that the b axis in compound 3 was
erroneously doubled, resulting in absent data for k
The data was examined using the program Layer; however,
data for both even and odd k values were present and strong,
demonstrating that the doubled unit cell is indeed the correct
one.
3
6¼ 2n on hkl.
2
2
A reduction in temperature can cause phase transitions to
,23
7
lower symmetry structures.
0
Accordingly, a crystal of 4
(
Z ¼ 2) was cooled on the diffractometer to 120 K (the tem-
perature of our study of the structure of 3). No doubling in
unit cell volume was observed, although the possibility of a
kinetically slow phase change cannot be ruled out. An alter-
native explanation for the occurrence of the two structures is
3
the inclusion on crystallisation of small amounts of Ph GeH
1
into crystalline 3 as a result of the synthetic route adopted. H
0
NMR analysis of the crystalline Z ¼ 4 structure did indeed
reveal the presence of significant amounts of hydride, while
close examination of the anisotropic displacement parameters
for the four independent chloro substituents shows them to be
slightly larger than might be expected at 120 K, particularly
in the case of Cl(3) and Cl(4). Moreover, the crystallo-
graphically determined Ge–Cl distances in 3, which range from
2.1377(16) to 2.1734(14) A, are markedly shorter than those
+
Fig. 3 The offset six-fold phenyl embrace between two crystallo-
graphically identical molecules in Ph GeF.
+
21
obtained from the commercial sample [2.184(2)–2.191(2) A ].
3
New J. Chem., 2002, 26, 457–461
459

1
The shortest distance is to Cl(3), which also exhibits the largest
anisotropic displacement parameters. It is likely, therefore,
that Ph GeH acts as a crystal-growth modifier and is actually
3
À202.2. H NMR (CDCl
3
, d) 7.28–7.38 (m, 9H V) ,ie 7w . 5A 1r –t i 7c l. e5 5O nline
26
ꢀ
ꢀ
(m, 6H). M.p. 77–78 C (lit. mp 76.6 C ). Anal. calcd: C,
66.94; H 4.68%. Found: C 67.00; H, 4.81%.
included upon crystallisation in small quantities (up to ca. 25%
from analysis of the chloride ellipsoids), resulting in an inter-
0
Ph GeCl (3). Triphenylgermane (0.075 g, 0.2 mmol) was
3
esting variation of the Z ¼ 2 structure; a phenomenon remi-
3
refluxed in CHCl (25 cm ) for 90 h. The resulting crystalline
2
4
niscent of the now discredited ‘‘bond stretch isomerism’’.
This is further confirmed by the depressed melting point of the
3
product (0.076 g) was found to be a mixture of Ph GeCl (45%
3
1
conversion, from H NMR) and Ph GeH. Slow evaporation of
ꢀ
ꢀ
sample at 105–108 C (cf. 114–115 C for the commercial
sample). It is remarkable, however, that one 6PE pair within
the asymmetric unit is more susceptible to replacement by
triphenyl germane than the other, suggesting the importance of
CHÁ Á ÁClGe and CHÁ Á ÁHGe interactions in controlling the
supramolecular architecture during crystal growth.
3
a CHCl solution of this product gave colourless crystals of
3
predominantly the chloride although contaminated with ca.
, d): 5.71(s, 1H), 7.3–7.7
1
0% of the hydride. H NMR (CDCl
2
3
À1
(
intensity=%): 340 (1), 338 (1), 336 (0.7), 305 (14), 303 (11), 228
m, 27H). IR: n(Ge–H) 2034 cm . MS (EI): m=z (relative
ꢀ
ꢀ
27
100). M.p. 105–108 C (lit. mp 114–115 C ).
(
Conclusions
Crystallography
Two common types of intermolecular interactions are found in
Crystals were mounted using a fast setting epoxy resin on the
end of a glass fibre and cooled on the diffractometer. All
crystallographic measurements were carried out with a Nonius
KappaCCD diffractometer equipped with graphite mono-
compounds of type Ph EX, namely CHÁ Á ÁX hydrogen bonding
3
and EF interactions. The p–p interactions are generally com-
patible with crystallographic symmetry, whereas the presence
of significant CHÁ Á ÁX bonding may be a factor leading to an
ꢀ
0
chromated Mo-Ka radiation using f rotations with 2 frames
and a detector-to-crystal distance of 30 mm. Integration was
increased Z value, particularly in cases where it is necessary to
simultaneously satisfy the EF and CHÁ Á ÁX bonding require-
ments. The structure of 3 shows that small quantities of
impurities, either present in the mother liquor or doped within
2
8
carried out by the program DENZO-SMN. Data sets were
corrected for Lorentz and polarisation effects and for the
2
8
0
effects of absorption using the program Scalepack. Struc-
tures were solved using the direct methods option of SHELXS-
the crystal, may have a significant effect on the Z value. This
observation has wider implications for the occurrence of
polymorphism in general.
2
9a
9
least squares refinement and difference Fourier synthesis
(
7
and developed using conventional alternating cycles of
2
9b
30
with the aid of XSeed. All non-hydrogen
SHELXL-97)
atoms were refined anisotropically, whilst hydrogen atoms
were fixed in idealised positions and allowed to ride. Hydrogen
atom thermal parameters were tied to those of the atom to
which they were attached. All calculations were carried out
either on a Silicon Graphics Indy workstation or an IBM-PC
compatible personal computer.
Experimental
Instrumental
Mass spectra were recorded at King’s College London on a
Jeol AX505W spectrometer in EI mode in a thioglycerol
matrix. NMR spectra were recorded on a Bruker ARX-360
spectrometer operating at 360.1MHz. IR spectra were recor-
ded in the form of nujol mulls on a PE Paragon 100 FTIR
spectrometer. Microanalyses were performed at The Uni-
versity of North London.
_
À1
GeIPh (1). C H GeI, M 430.79 g mol , P1, a ¼ 9.5925(4),
3
18 15
+
ꢀ
ꢀ
b ¼ 9.6743(3), c ¼ 18.3085(8) A, a ¼ 83.959(3) , b ¼ 78.664(2) ,
+
3
ꢀ
g ¼ 77.060(2) , U ¼ 1620.26(11) A , Z¼ 4, 5696 unique data
ꢀ
2
2
(2y 4 50 ), 361parameters, R [F > 2s(F )]¼ 0.033, wR (all
1
2
data) ¼ 0.082.
GeFPh3 (2). C H GeF, M 322.89 g mol , P2 =c, a ¼
Preparations
À1
18
15
1
Ph
3
GeI (1). Sodium iodide (1.56 g, 10.4 mmol) was added
+
ꢀ
1
0.3444(8), b ¼ 11.7723(7), c ¼ 13.0869(11) A, b ¼ 110.328(4) ,
+
ꢀ
3
to acetone (20 cm ) with stirring according to the published
3
U ¼ 1494.43(19) A , Z ¼ 4, 2621unique data (2 y 4 50 ), 242
2
procedure. Triphenylgermanium bromide (0.9 g, 2.3 mmol)
5
2
2
parameters, R
1
[F > 2s(F )]¼ 0.0619, wR
2
(all data) ¼ 0.1714.
3
was dissolved in acetone (30 cm ), added to the sodium iodide
ꢀ
acetone mixture and refluxed overnight. Cold hexane (0 C,
1
3
00 cm ) was then added and the mixture was filtered to
À1
GeClPh
(3). C18
H
15ClGe, M 339.34 g mol , P2
=c,
ꢀ
3
1
+
remove excess NaI and NaBr. The solvent was removed by
evaporation to give the product as a yellow white powder. The
product was recrystallised from diethyl ether. Yield 0.60 g, 1.4
a ¼ 18.637(4), b ¼ 18.317(4), c ¼ 18.129(4) A, b ¼ 98.38(3) ,
+
3
ꢀ
U ¼ 6122(2) A , Z ¼ 16, 12 004 unique data (2y 4 52 ), 722
2
2
parameters, R
1
[F > 2s(F )] ¼ 0.055, wR (all data) ¼ 0.131.
2
1
mmol, 61%. H NMR (CDCl
3
, d) 7.62–7.59 (m, 2H), 7.44–
.41 (m, 3H). M.p. 154–156 C (lit. mp 157 C ). Anal calcd: C,
0.18; H, 3.51%. Found: C, 50.25; H, 3.58%.
CCDC reference numbers 179372–179374. See http:==www.
rsc.org=suppdata=nj=b1=b105457k= for crystallographic data
in CIF or other electronic format.
ꢀ
ꢀ
26
7
5
Ph
mmol) was dissolved in CH
to a stirred solution of (NMe
3
GeF (2). Triphenylgermanium bromide (0.92 g, 2.6
3
2
Cl
)F (0.3 g, 3.0 mmol) in CH
15 cm ), resulting in the immediate formation of a white
2
(15 cm ) and added dropwise
Acknowledgements
4
2
Cl
2
3
(
Particular thanks are due to Dr L. J. Barbour (University of
Missouri at Columbia) for his superb XSeed software,
Dr D. A. Armitage (KCL) for advice and inspiration and to
Dr E. Tiekink (University of Adelaide, Australia) for the
crystallographic data for 4 and helpful discussions. We also
thank the EPSRC for a studentship (to P. D. P.), King’s
College London for the provision of X-ray crystallographic
facilities and NATO for a collaborative grant.
precipitate. The mixture was refluxed for 2 h after which time
3
distilled water (30 cm ) was added. The organic layer was
separated and the aqueous portion was washed with 3 ꢁ 20
3
cm portions of CH
2
Cl
2
. The organic portions were combined
and the solvent removed under vacuum. The final product was
obtained from recrystallisation from petroleum spirit (bp 60–
ꢀ
19
8
0 C). Yield 0.77 g, 2.4 mmol, 92%. F NMR (CDCl , d)
3
4
60
New J. Chem., 2002, 26, 457–461

1
1
8
9
Cambridge Structural Database, April 2001 release, V2 i3e 3w 2 A1 8rt ei cn l ter i Oe s n. line
G. R. Desiraju and T. Steiner, T he Weak Hydrogen Bond, Oxford
University Press=IUPAC, Oxford, UK, 1999.
References
1
2
3
4
5
C. P. Brock and L. L. Duncan, Chem. Mater., 1994, 6, 1307.
C. P. Brock and J. D. Dunitz, Chem. Mater., 1994, 6, 1118.
C. P. Brock, J. Res. Natl. Inst. Stand. Technol., 1996, 101, 321.
A. J. C. Wilson, Acta Crystallogr., Sect. A, 1993, 49, 795.
N. Padmaja, S. Ramakumar and M. A. Viswamitra, Acta Crys-
tallogr., Sect. A, 1990, 46, 730.
20 G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford
University Press, Oxford, UK, 1997.
21D. Dakternieks, A. E. K. Lim, B. Zobel and E. R. T. Tiekink,
Main Group Met. Chem., 2000, 23, 789.
22 L. J. Barbour, Layer–A program for the Display and Analysis of
Crystallographic Reflection Data, University of Missouri, Columbia,
MO, USA, 1995.
23 J. D. Dunitz, G. Fillippini and A. Gavezzotti, Helv. Chim. Acta,
2000, 83, 2317.
24 (a) V. C. Gibson and M. McPartlin, J. Chem. Soc., Dalton T rans.,
1992, 947; (b) G. Parkin, Chem. Rev., 1993, 93, 887.
25 R. Fuchs and H. Gilman, J. Org. Chem., 1958, 23, 911.
26 C. A. Kraus and L. S. Foster, J. Am. Chem. Soc., 1927, 49, 457.
27 Aldrich Handbook of Fine Chemicals, Aldrich Chemical Co.,
Gillingham, UK, 2000–01, p. 1747
28 Z. Otwinowski and W. Minor, in Methods in Enzymology, ed.
C. W. Carter and R. M. Sweet, Academic Press, London, 1996,
vol. 276, pp. 307–326.
6
7
T. Steiner, Acta Crystallogr., Sect. B, 2000, 56, 673.
J. W. Steed, E. Sakellariou, P. C. Junk and M. Smith, Chem. Eur.
J., 2000, 7, 1240.
8
H. Jacobsen, H. W. Schmalle, A. Messmer and H. Berke, Inorg.
Chim. Acta, 2000, 306, 153.
9
0
R. E. Marsh, Acta Crystallogr., Sect. B, 1999, 55, 931.
M. Meyer, W. A. Paciorek, K. J. Schenk, G. Chapuis and
W. Depmeier, Acta Crystallogr., Sect. B, 1994, 50, 333.
W. Paciorek, V. B. Gaillard, K. Schenk and G. Chapuis, Acta
Crystallogr., Sect. A, 1996, 52, 349.
C. A. Hunter and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112,
5525.
H. Takahashi, S. Tsuboyama, Y. Umezawa, K. Honda and M.
Nishiro, Tetrahedron, 2000, 56, 6185.
1
1
1
1
1
2
3
29 (a) G. M. Sheldrick, SHELXS-97, University of Go
Germany, 1997; (b) G. M. Sheldrick, SHELXL-97, University of
Gottingen, Germany, 1997.
¨
ttingen,
1
1
4
5
C. Horn, M. Scudder and I. Dance, Cryst. Eng. Comm., 2001, 2.
M. Scudder and I. Dance, J. Chem. Soc., Dalton T rans., 2000,
¨
2909.
M. Scudder and I. Dance, Cryst. Eng. Comm., 2001, 12.
F. H. Allen and O. Kennard, Chem. Des. Autom. News, 1993, 8, 31 .
30 L. J. Barbour, XSeed, A Program for the Manipulation and Display
of Crystallographic Models, University of Missouri, Columbia,
MO, USA, 1999.
1
1
6
7
New J. Chem., 2002, 26, 457–461
461