Electronic properties of 3,3′-dimethyl-5,5'-bis(l,2,4-triazine): Towards design of supramolecular arrangements of N-heterocyclic Cui complexes

Article abstract of DOI:10.1002/chem.200601686

A new efficient and safe synthesis of 3,3′-dimethyl-5,5′-bis- (1,2,4-triazine) is presented. The electron-density distribution and electrostatic properties (charge, electrostatic potential) of this molecule were analyzed. These properties were derived from a high-resolution single-crystal X-ray diffraction experiment at 100 K and compared to the results obtained from ab initio DFT quantum-mechanical calculations. Comparisons of its electrostatic potential features and integrated atomic charges (quantum theory of atoms in molecules, QTAIM) havebeen made with those of related molecules such as bipyrimidine ligands. Two methods were used to derive integrated charges: one is based on the conventional analytical procedure and the second uses a steepest-ascent numerical algorithm. Excellent agreement was obtained between these two methods. Charges and electrostatic potential were used as predictive indices of metal chelation and discussed in the light of complexation abilities of the title compound and related molecules. The crystal structure of a Cu ^I complex of 3,3′-dimethyl-5,5′-bis(1,2,4-triazine) is reported here. In the solid state, this complex forms a three-dimensional multibranch network with open channels in which counterions and solvent molecules are located. This architecture involves both cis and trans isomers of the title compound.

Full text of DOI:10.1002/chem.200601686

DOI: 10.1002/chem.200601686
Electronic Properties of 3,3’-Dimethyl-5,5’-bis(1,2,4-triazine): Towards
I
Design of Supramolecular Arrangements of N-Heterocyclic Cu Complexes
[
a]
[b]
[a]
[c]
Blandine Courcot, Diem Ngan Tran, Bernard Fraisse, FranÅois Bonhomme,
[
b]
[a, c]
Alain Marsura,* and Nour Eddine Ghermani
In memory of Professor Guy Ourisson
Abstract: A new efficient and safe syn-
thesis of 3,3’-dimethyl-5,5’-bis-(1,2,4-tri-
azine) is presented. The electron-densi-
ty distribution and electrostatic proper-
ties (charge, electrostatic potential) of
this molecule were analyzed. These
properties were derived from a high-
resolution single-crystal X-ray diffrac-
tion experiment at 100 K and com-
pared to the results obtained from ab
initio DFT quantum-mechanical calcu-
lations. Comparisons of its electrostatic
potential features and integrated
atomic charges (quantum theory of
atoms in molecules, QTAIM) have
been made with those of related mole-
cules such as bipyrimidine ligands. Two
methods were used to derive integrated
charges: one is based on the conven-
tional analytical procedure and the
second uses a steepest-ascent numeri-
cal algorithm. Excellent agreement was
obtained between these two methods.
Charges and electrostatic potential
were used as predictive indices of
metal chelation and discussed in the
light of complexation abilities of the
title compound and related molecules.
The crystal structure of a Cu complex
of 3,3’-dimethyl-5,5’-bis(1,2,4-triazine)
I
is reported here. In the solid state, this
complex forms
a three-dimensional
multibranch network with open chan-
nels in which counterions and solvent
molecules are located. This architec-
ture involves both cis and trans isomers
of the title compound.
Keywords: ab initio calculations ·
copper
heterocycles
chemistry
·
N
ligands
· nitrogen
A
H
R
U
G
·
supramolecular
AHCTREUNG
Introduction
chemistry, and the search of new ligands for novel self-as-
sembled metallosupramolecular architectures remains of
great interest. Among the nitrogen heterocycles, 2,2’-bipyri-
dine has been widely used as the classical bidendate chelat-
ing heterocyclic ligand, along with its tridendate analogue
Metallosupramolecular chemistry involves the use of combi-
nations of organic ligands and metals for the construction of
both discrete and polymeric aggregates. Nitrogen heterocy-
cles are extensively used as bridging ligands in coordination
[1a–c]
terpyridine.
Besides the pyridine ring, many other het-
erocycles which differ in their electronic and structural prop-
erties are often readily available by synthesis. Surprisingly,
only a few of them have attracted the attention of coordina-
tion chemists. Nevertheless, some of them, for example, pyr-
azine, pyrazole, imidazole, and triazole, have been used as
[
a] Dr. B. Courcot, Dr. B. Fraisse, Prof. N. E. Ghermani
Ecole Centrale Paris
Laboratoire SPMS UMR CNRS 8580 1
Grande Voie des Vignes, 92295 Châtenay-Malabry (France)
[1d–g]
units for metal coordination.
[
b] D. N. Tran, Prof. A. Marsura
Thus, taking in account that varying the heterocycle can
effect dramatic changes in the physicochemical properties of
metal complexes, we have reported original assemblies with
GEVSM, UMR CNRS 7565
Structure et RØactivitØ des Syst›mes MolØculaires Complexes
UniversitØ Henri PoincarØ Nancy-1
[
2a–b]
5
rue A. Lebrun, BP 403, 54250 Nancy (France)
bidendate heterocyclic ligands such as 4,4’-bipyrimidine,
Fax : (+33)383-682-345
E-mail: Alain.Marsura@pharma.uhp-nancy.fr
[3a–b]
[4a–b]
[5]
2
,2’-bipyrazine,
2,2’-bithiazole,
4,4’-bithiazole, and
their bipyridine–bipyrazine, bipyridine–bipyrimidine, and bi-
pyridine–bithiazole heterotopic bis-bidendate combination-
[
c] Dr. F. Bonhomme, Prof. N. E. Ghermani
Laboratoire PPB, UMR CNRS 8612, IFR 141
FacultØ de Pharmacie 5, rue Jean-Baptiste ClØment
[6a–c]
s.
Changes in the electronic structure of the ligands
9
2296 Châtenay-Malabry (France)
proved to have a major effect on complex configuration. For
3414
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3414 –3423

FULL PAPER
[
12]
example, with the ligand combinations, high selectivity for
head-to-tail (H–T) or head-to-head (H–H) orientation of
the ligand strands around the metal centers in the resultant
harmless hydrazine/hydroquinone (1/1) inclusion complex,
which afforded 1 in good yield (60%). Bis-triazine 3 could
be synthesized efficiently in only one step from 1 without
isolation of 3-methyl-1,2,4-triazine (2) and in fairly good
yield (55%). Moreover, acetamidrazone (1) could be ob-
tained quantitatively from acetamidine hydrochloride by
using commercial hydrazine hydrate (N H ·H O) solution,
[6a–c]
crystal structures was observed with different metals.
Such a selectivity has never been obtained before when in-
duced, for example, only by steric hindrance. This was re-
cently investigated by us at a theoretical level that establish-
ed the p-backbonding effect, when effective, to be at the
2
4
2
which contains 32% of water and is totally safe.
[7]
origin of the strongly orientational behavior.
Continuing our research on new bis-heterocyclic systems,
we decided to explore the metal-coordination capabilities of
the 3,3’-dimethyl-5,5’-bis(1,2,4-triazine) ligand. We report
here a new synthesis of this ligand and its X-ray structure
and electron-density distribution from both experiment and
theory. As far as we know only one study on the isomeric
Molecular conformation and crystal stacking of 3: An
ORTEP plot of 3,3’-dimethyl-5,5’-bis(1,2,4-triazine) (3) is
given in Figure 1. The molecule is in the s-trans conforma-
[13]
II
bis-3,3’-(5,6-dimethyl-1,2,4-triazine) Fe complexes has been
[8]
reported, whereas nothing has been published on the
metal complexes of the 3,3’-dimethyl-5,5’-bis(1,2,4-triazine)
I
isomer. In the present study, the crystal structure of the Cu
complex of this ligand is reported and is compared to those
of related molecules.
Results and Discussion
Synthesis: Substituted and unsubstituted 5,5’-bis(1,2,4-tria-
zines) have been generally synthesized in one step from the
unsubstituted or substituted 1,2,4-triazine heterocycles by a
simple coupling reaction under strongly basic conditions or
[
13]
[9a–c]
Figure 1. ORTEP plot and atom-numbering scheme of 3. Ellipsoids are
by treatment with potassium cyanide.
The 1,2,4-triazine
at the 50% probability level.
starting materials have been synthesized by intermolecular
[
10a–b]
heterocyclization between amidrazones
or thiosemicar-
[11]
bazides and a,b-dicarbonyl compounds. Here we are in-
terested in 3,3’-dimethyl-5,5’-bis(1,2,4-triazine), which was
obtained by two methods from acetamidrazone (1;
Scheme 1). The authors recommended using anhydrous hy-
tion in the solid state and is not rigorously planar. The dihe-
dral angle C6’-C5’-C5-C6 is 173.63(6)8. The C5ꢀC5’ bond
length of 1.4813(7) Š is slightly shorter than that in 2,2’-di-
methyl-6,6’-diphenyl-4,4’-bipyrimidine
(4;
C4ꢀC4’
[2]
1
.4891(7) Š) at 100 K. The N
heterocycle in 3 exhibits alter-
nately short and long bond
lengths: C5ꢀN4 1.3250(7), N4ꢀ
C3 1.3468(7), C3ꢀN2 1.3380(8),
N2ꢀN1
1.3399(7),
.3280(8), and
N1ꢀC6
C6ꢀC5
1
1
.4056(8) Š. The values ob-
tained for the second triazine
ring are within the estimated
standard deviations. No hydro-
gen bonding occurs in the crys-
tal stacking, which displays par-
allel zigzag arrangements of
molecules (Figure 2).
Scheme 1. Synthesis of 3. i) NH
KOH; K CO /KCN; 408C.
2 2 2 2 2 4
NH /hydroquinone complex/MeOH; 0–58C; ii) NH NH , H O; MgSO ; iii)
2
3
Electron deformation density of
drazine, a very hazardous reagent, for the synthesis of 3-
methyl-1,2,4-triazine (2). To circumvent the dangerous and
difficult preparation of anhydrous hydrazine for obtaining
acetamidrazone, we advantageously used the stable and
3: The presence of three nitrogen atoms in the triazine ring,
two of which are bonded together, is expected to strongly
perturb the electron distribution in this group. One way to
visualize the anisotropy of the electron density is to plot the
Chem. Eur. J. 2007, 13, 3414 –3423
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3415

A. Marsura et al.
[14]
the tetrazole ring of LR-B/081 reported by Destro et al.
Accumulation of electrons in lone pairs on nitrogen atom
are well resolved in these two maps. However, the contrac-
tions of these lone pair electrons are different for the three
nitrogen atoms of the triazine ring in the experimental map
(
Figure 3, top), whereas in the theoretical electron-density
map (Figure 3, bottom), the three nitrogen atoms exhibit an
almost identical contraction of the lone pairs. Slight discrep-
ancies also appear for hydrogen atoms, which display rela-
tively higher polarizations in the experimental electron den-
sity map than in the theoretical one.
Electrostatic potential of 3 and related molecules: In an at-
tempt to relate the electronic properties of N-heterocyclic
compounds to their chelation of metals, we compared the
electrostatic potential obtained from both experiment and
theory. The molecular electrostatic potential of 3 is shown in
Figure 4. As expected, the areas of negative electrostatic po-
Figure 2. Two views of the crystal stacking of 3 along the b axis. Hydro-
gen atoms are omitted for clarity and nitrogen atoms are in blue. Dashed
lines indicate short contacts between adjacent molecules.
electron deformation density, that is, the difference between
the total electron density and that obtained as a superposi-
tion of isolated neutral-atom densities (promolecule).
Figure 3 displays the theoretical and static experimental
Figure 4. Electrostatic potential of 3. Top: from experiment. Bottom:
from theory. Gray and dark red isosurfaces correspond to +0.20 and
ꢀ
1
ꢀ
0.05 eŠ , respectively. The orientation of the molecule is the same as
in Figure 3.
tential (nucleophilic region) surround the N1ꢀN2 bonds due
to the high electronegativity of these atoms. In both maps,
these regions have a kidney shape but exhibit a different ex-
pansion when we compare the results from experiment and
theory. The minima of the negative electrostatic potential
Figure 3. Static experimental electron deformation (top) and theoretical
ꢀ
3
(
bottom) densities of 3. Contour intervals 0.05 eŠ ; negative contours
ꢀ1
Vmin are ꢀ0.174 (experiment) and ꢀ0.149 eŠ (theory),
are dashed in experimental maps and are in red in theoretical map.
both close to N1.
The contraction of the region of negative electrostatic po-
tential is due in part to the contribution of the positive
charges carried by carbon and hydrogen atoms. This is par-
ticularly true for the negative pockets of electrostatic poten-
tial found in the vicinity of the nitrogen atoms N4 and N4’
from the theoretical calculations, which are not compensat-
ed by the positive contribution of H6 and H6’. These fea-
tures are related to the less polarized electron densities of
hydrogen atoms found in theoretical calculations, as men-
electron deformation densities in the plane defined by N4,
C5, and C5’ of 3. These maps exhibit similar features for the
CꢀC, CꢀN, and CꢀH bonds, with an average accumulation
ꢀ
3
of electron density of around 0.6 eŠ , as is generally found
in organic molecules. A good agreement between theory
and experiment is found for the electron-density peak in the
ꢀ
3
N1ꢀN2 bond of about 0.4 to 0.5 eŠ . This value compares
quantitatively well with those obtained for NꢀN bonds in
3416
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3414 –3423

Electronic Properties of 3,3’-Dimethyl-5,5’-bis(1,2,4-triazine)
FULL PAPER
tioned above, but also to the more contracted lone pairs of
nitrogen atoms. For the sake of comparison with related
compounds, we have generated the electrostatic potential
This is the case for experimental charges derived from the
[15]
so-called kappa refinement
tained, for instance, from a Mulliken population analysis.
or theoretical charges ob-
[16]
[2]
for isolated 2,2’-dimethyl-6,6’-diphenyl-4,4’-bipyrimidine
4) and 2,2’-dimethyl-4,4’-bipyrimidine (5). For the latter, no
The charges can also be fitted to the observed electrostatic
potential (ChelpG charges) in order to reproduce this prop-
(
[17]
experimental electron density is available and only DFT cal-
culations were performed. The isosurfaces of the electrostat-
ic potential are depicted in Figure 5. The experimental nega-
erty.
The quantum theory of atoms in molecules
[18]
(QTAIM) developed by Bader is based on the partition
of the electron density in atomic basins that allows integrat-
ed properties like atomic charges to be retrieved. Once the
atomic basin is defined through the zero-flux surface of the
gradient of the electron density, volume integration can be
performed to obtain the charge. This method was applied to
the title compound and to the related molecules 4 and 5.
[19]
The program NEWPROP can be used to retrieve the
integrated charges from experimental charge densities. This
calculation is based on a sophisticated analytical determina-
tion of interatomic surfaces. Recently, Henkelman et al. pro-
posed a fast and robust numerical method based on steep-
est-ascent algorithm (BADERWIN program) to integrate
[20]
the electron density on a fine grid. In the steepest-ascent
procedure, each grid point is assigned to the point where the
electron density is maximum (nuclear attractor). As an ex-
ample, Figure 6 displays the integration volumes of the ni-
Figure 5. Electrostatic potential (left) of
4 (top: from experiment;
middle: from theory) and 5 (bottom, from theory). Gray and dark red
Figure 6. Integration volumes of the 1,2,4-triazine nitrogen atoms in 3.
The dotted surface corresponds to the 0.001 eŠ isosurface of the total
electron density. The orientation of the molecule is the same as in
Figure 3.
ꢀ
1
ꢀ3
isosurfaces correspond to +0.20 and ꢀ0.05 eŠ , respectively. Right:
chemical structures of molecules.
tive electrostatic potential surface corresponding to a cutoff
ꢀ
1
of ꢀ0.05 eŠ extends from the phenyl ring to the bipyrimi-
dine group (Figure 5, top) on both sides of 4. From theoreti-
cal calculations, however, this negative region is limited in
the vicinity of the inner (3-position in the pyrimidine ring)
and outer (1-position) nitrogen atoms (Figure 5, center).
When the phenyl rings are absent, the regions of negative
electrostatic potential are found close to the inner and outer
nitrogen atoms, as in 5 (Figure 5, bottom). The minimum
trogen atoms in 3. To compare the atomic charges obtained
from both experiment and theory by a single method, we
decided to use this new approach. As a benchmark, we also
applied the NEWPROP procedure to the experimental elec-
tron density. The results are reported in Table 1. Com-
pounds 4 and 5 have a planar bipyrimidine group and a
center of inversion in the middle of the central C4ꢀC4’
bond. Thus, Table 1 gives the atomic charges corresponding
to half of each molecule. Both programs NEWPROP and
BADERWIN give a difference in charges of symmetry-relat-
ed atoms of around 0.001e. Conversely, as mentioned above,
the molecule of 3 is not planar and the integrations of the
electron density over the entire molecule yield relatively
higher discrepancies in the values of atomic charges for
pseudo-symmetry-related atoms such as N4, N4’ and so on.
The largest difference (0.2e) is obtained for the pair C7, C7’
from NEWPROP calculations, even for a fine grid. The cor-
values of the negative electrostatic potential are V
=
min-exptl
ꢀ1
ꢀ1
0
.233 eŠ and V
=ꢀ0.146 eŠ (both close to N1)
min-DFT
and Vmin-DFT =ꢀ0.159 eŠꢀ (close to N1 from theory) for 4
1
and 5, respectively.
Atomic charges of 3 and related molecules: In general, dif-
ferent types and definitions of atomic charges are used in
the literature depending on the goals and methods. Charges
are not physically observable and are model-dependent.
Chem. Eur. J. 2007, 13, 3414 –3423
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3417

A. Marsura et al.
Table 1. QTAIM charges (in e units) for 3, 4, and 5. C and H atoms of methyl (meth) and phenyl (phe) groups are indicated in the atomic labels.
3
4
5
EXP
EXP
DFT
EXP
EXP
DFT
DFT
NEWPROP
BADERWIN
BADERWIN
NEWPROP
BADERWIN
BADERWIN
BADERWIN
N4
N4’
N1
N1’
N2
N2’
C5
C5’
C6
C6’
C3
C3’
C7 meth
C7’meth
H6
7.825
7.823
7.377
7.363
7.513
7.506
5.606
5.591
5.569
5.592
5.328
5.294
6.528
6.348
0.905
0.898
0.897
0.750
0.864
0.908
0.752
0.852
7.778
7.821
7.397
7.534
7.552
7.604
5.573
5.741
5.486
5.431
5.344
5.307
6.235
6.221
0.902
0.896
0.952
0.740
0.912
0.951
0.784
0.890
8.088
8.022
7.455
7.484
7.665
7.738
5.534
5.661
5.382
5.371
5.056
5.017
5.846
5.809
1.001
1.002
1.033
0.955
0.959
0.967
0.995
1.063
N1
N3
C2
C4
C5
C13meth
C6
C7phe
C8phe
C9phe
C10phe
C11phe
C12phe
H5
H131meth
H132meth
H133meth
H8phe
H9phe
H10phe
H11phe
H12phe
7.929
8.071
5.258
5.590
6.135
6.213
5.661
5.903
6.130
6.126
6.157
6.135
6.133
0.834
0.830
0.831
0.845
0.831
0.830
0.785
0.843
0.841
7.958
8.153
5.334
5.651
6.152
6.106
5.667
5.784
6.190
6.228
6.106
6.119
6.309
0.810
0.906
0.798
0.844
0.831
0.792
0.887
0.818
0.807
8.077
8.149
5.160
5.517
5.988
5.818
5.512
5.934
6.033
6.085
5.916
5.997
6.156
0.954
1.070
0.972
0.988
0.939
0.980
1.074
0.974
0.964
N1
N3
C2
C4
C5
C13meth
C6
8.095
8.142
5.156
5.510
5.979
5.812
5.454
0.993
H6
H5
0.953
1.069
0.972
0.982
H131meth
H132meth
H133 meth
H6’
H71meth
H72meth
H73meth
H71’meth
H72’meth
H73’meth
relation between the experimental atomic charges of 3 and 4
to 3608 at 108 intervals. The barriers for internal rotation
were calculated as the difference between the total energy
of the structure at each torsion angle, without any geometri-
cal optimization, and that of the cis form of the molecule
[19]
[20]
obtained from programs NEWPROP and BADERWIN
is excellent; the statistical factors are R=0.999, RMSD=
.10 for 3 and R=0.999, RMSD=0.07 for 4. To our knowl-
edge, it is the first time that the procedure implemented in
0
[20]
the BADERWIN program has been used for large mole-
cules and the results are very encouraging. Excellent agree-
ment factors (R=0.999 and RMSD=0.02) were also ob-
tained for the DFT theoretical charge sets of 4 and 5 mole-
cules (Table 1). From consideration of electron withdrawal,
the phenyl rings do not seem to modify the atomic charges
in the heteroatomic groups like bipyrimidine. The respective
theoretical electric charges carried by N1 (ꢀ1.077e in 4 and
ꢀ
1.095e in 5) and N3 (ꢀ1.149e in 4 and ꢀ1.142e in 5) are
almost the same in 4 and 5; the inner atom N3 is slightly
more negative. Figure 7 compares the charges obtained from
both experiment and theory for molecules 3 and 4 by using
the BADERWIN program. A very good linear correlation
appears with statistical factors R close to 1 and an RMSD of
0
.2e. The latter can be considered as a global standard devi-
ation for the estimated charges. In 3, the trend observed for
and 5 still remains: the inner nitrogen atom N4 carries the
4
highest negative charge (ꢀ0.8e from experiment and ꢀ1.0e
from theory). From both approaches (see Table 1), N2 (N2’)
displays a slightly more negative charge than N1 (N1’), but
the difference remains in the range of uncertainty (0.2e).
Rotational energy barriers: The barriers for internal rota-
tions about the central CꢀC bonds of structurally compara-
ble molecules 3 and 5 in vacuo were calculated at the DFT
level of theory. Total energies as functions of torsion angles
were calculated by scanning these angles from 08 (cis form)
Figure 7. Experimental versus theoretical charges. Top: 3. Bottom: 4.
3418
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3414 –3423

Electronic Properties of 3,3’-Dimethyl-5,5’-bis(1,2,4-triazine)
FULL PAPER
(
torsion angle of 08), taken as reference. Figure 8 shows this
energy difference for 3 and 5. As expected, the trans form
1808) is more stable than the cis form, by about 6.4 and
(
I
Figure 9. Structures of Cu complexes of 4 (left) and 5 (right). Hydrogen
atoms are omitted for clarity. Copper atoms are in red and nitrogen atom
in blue.
Copper complex of 3: Addition of a third nitrogen atom to
the heterocycle to give a triazine should obviously increase
the reactivity towards cations. The similar values of the
atomic charges in 3 (Table 1) and the similar characteristics
of the electrostatic potential (Figure 4) of the outer nitrogen
atoms reinforce this hypothesis. This was experimentally
demonstrated by the structure of the copper complex of 3
Figure 8. DFT energy barriers for internal rotation about central Cꢀ
C
bonds of 3 and 5.
(
6) displayed in Figure 10. In this structure, both trans and
ꢀ
1
7
.5 kcalmol for 3 and 5, respectively. This may be due to
the H-bonding interactions between internal N4(N4’) and
H6(H6’) for 3 and N3(N3’) and H5(H5’) for 5, which stabi-
AHCTREUNG
A
C
H
T
R
E
U
N
G
A
H
R
U
G
ACHTREUNG
lize the system. These nitrogen atoms carry the highest
atomic charges (Table 1), probably because of this H-bond-
ing interaction. For both molecules 3 and 5, two local
minima correspond to torsion angles of 40 and 3208. Howev-
er, the local rotational barrier with respect to the maxima
ꢀ
1
found at 80 and 2808 is higher for 3 (1.43 kcalmol ) than
ꢀ
1
for 5 (1.06 kcalmol ). This means that the energy demand
of molecule 3 to reach its global minimum is higher by
ꢀ
1
0
.4 kcalmol with respect to 5. Furthermore, the latter is
ꢀ
1
more stable in its trans form by 1.1 kcalmol .
I
Figure 10. Crystal structure of the Cu complex of 3,3’-dimethyl-5,5’-bis-
(1,2,4-triazine) 6. Hydrogen atoms are omitted for clarity. Copper cations
and the trans isomer of 3 are shown in ball-and-stick form. The distances
between the copper cations are given in Š. Copper atoms are in red and
nitrogen atoms in blue.
Copper complexes of 4 and 5: The electrostatic properties
derived from both theory and experiment show that the
inner N3 and outer N1 nitrogen atoms of bipyrimidine-
based molecules 4 and 5 have similar charges and topologi-
cal features of the negative electrostatic potential. Accord-
ingly, some years ago, we reported that both nitrogen atoms
cis isomers, in the ratio 1:4, coexist in the solid state. Two
copper cations, labeled Cu1 and Cu2 in Figure 10, were
found in the asymmetric unit, both in tetrahedral coordina-
tion. Each copper cation is connected to four nitrogen
atoms, two of which belong to the same bis-triazine mole-
cule in cis form. For Cu1, the third bonded nitrogen atom is
from another cis ligand, and the fourth from the trans
isomer. However, no trans molecule is connected to Cu2.
Only nitrogen atoms of type N4 (cis isomer, see Figure 1)
and N1 (in both cis and trans ligands) are involved. Coexis-
tence of both cis and trans isomers of ligand 3 agrees to
some extent with the smaller energy barrier between these
two forms reported above. The trans isomer of 3 bridges two
copper cations, as shown in Figure 10 (Cu1ꢀCu1 9.718 Š),
[2]
can participate in metal complexation and we published
I
crystallographic studies on several Cu bipyrimidine com-
[2]
plexes. In complexes, the ligands are in the trans form cor-
responding to the minimum of the rotational energy, as
shown in Figure 8. It was also shown that different architec-
tures can be obtained: for molecule 4, isolated complex
units, involving only the inner N3 nitrogen atoms, were
found in the solid state, mainly due the steric hindrance of
the phenyl groups; for 5, however, in combination with a
molecule of acetonitrile solvent, a polymer complex was ob-
tained. In this case, both N1 and N3 participate in metal
complexation. Figure 9 depicts the solid-state structures of
I
[2]
the Cu complexes obtained from 4 and 5.
Chem. Eur. J. 2007, 13, 3414 –3423
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3419

A. Marsura et al.
exclusively through its outer N1, N1’ atoms. The nitrogen
atoms corresponding to N2 of both isomers are never con-
nected to any metal atom. Note that the absolute minimum
of the electrostatic potential was found in the vicinity of N1
ꢀ
1
ꢀ1
(
Vmin-exp =ꢀ0.174 eŠ and Vmin-DFT =ꢀ0.149 eŠ ), and this
result agrees with the structure of the present copper com-
plex. Each cis ligand of 3 is, in turn, bonded to three metal
ions through N4, N4’ (Cu1), N1 (Cu2), and N1’ (Cu2’). The
CuꢀN bond lengths and angles are reported in Table 2. The
Table 2. Selected bond lengths [Š] and angles [8] in the copper complex
[
a]
of 3.
Bond lengths [Š]
Bond angles [8]
N1’A-Cu1-N1’C
N1’A-Cu1-N4B
N1’C-Cu1-N4B
N1’-Cu1-N4’B
Cu1ꢀN1’A
Cu1ꢀN1C
Cu1ꢀN4B
Cu1ꢀN4’B
1.993(8)
2.018(9)
2.041(9)
2.076(8)
110.9(3)
128.2(3)
111.7(4)
119.6(3)
99.4(3)
N1C-Cu1-N4’B
N4B-Cu1-N4’B
N1’B#1-Cu2-N1B#2
N1’B#1-Cu2-N4A
N1B#2-Cu2-N4A
N1’B#1-Cu2-N4’A
N1B#2-Cu2-N4’A
N4A-Cu2-N4’A
80.5(3)
97.2(3)
Cu2ꢀN1’B#1
Cu2ꢀN1B#2
Cu2ꢀN4A
1.987(8)
1.991(9)
1.991(9)
2.055(8)
124.0(3)
117.9(4)
112.2(3)
127.2(3)
80.8(3)
Cu2ꢀN4’A
[
a] A, B, C correspond to the three molecules found in the asymmetric
1
1
unit. Symmetry operations: #1: ꢀ1+x, y, z; #2: ꢀ1+x, =
2
ꢀy, =
2
+z
Figure 11. Crystal architectures of the copper complexes of 5 (top, copper
and acetonitrile atoms are shown in ball-and-stick form) and 6 (bottom,
copper cations and atoms of ligands 3 in trans form are shown in ball-
and-stick form). Hydrogen atoms are omitted for clarity. Copper atoms
are in red and nitrogen atoms in blue.
bond lengths are very similar to those reported for the com-
plexes of 4 and 5.
[2]
Crystal architectures of the copper complexes of 3 and 5:
Figure 11 depicts the crystal arrangements of the copper
complexes of 3 and 5. The bipyrimidine ligands 5 chelate
the copper cations through two inner and one outer nitrogen
atoms plus one from the acetonitrile solvent molecule. One
outer nitrogen atom of the bipyrimidine molecule is never
connected to any metal center. This gives rise to 1D zigzag
chains in the solid state (Figure 11, top). These chains are
isolated and parallel to the b axis in the crystal lattice. The
distance between two adjacent copper cations is 6.12 Š. This
is shorter than the Cu1ꢀCu2 distances of respectively 6.62,
6.75, and 6.77 Š (Figure 10) found in the copper complex of
3
, due to the para positions of the metal bonded nitrogen
Figure 12. Left: View of channels in the structure of the copper complex
atoms (N1, N4). In the bottom part of Figure 11 (view down
a axis) one particular view of the structure obtained for the
copper complex of molecule 3 is shown. In this partial repre-
sentation, each ligand 3 is linked to three metal atoms. This
gives a cross-linked structure consisting of undulating chains
of 3,3’-dimethyl-5,5’-bis(1,2,4-triazine) 6. Copper tetrahedra are dis-
played. Right: the distribution of PF6 counterions (in space-filling form)
in the empty spaces. Hydrogen atoms are omitted for clarity.
ꢀ
[21]
ꢀ
(MOFs) reviewed recently by Fletcher et al.
The PF6
(
or branches) of alternating Cu1 and Cu2 atoms. The chains
counterions and solvent molecules occupy these channels.
These anions are stabilized by multiple weak F···H hydrogen
bonds with H atoms of the methyl groups of ligands 3.
are interconnected via molecules 3 in trans form bonded ex-
clusively to Cu1 cations (see Figure 10). In fact, both Cu1
and Cu2 atoms can be considered as the nodes of branches
and this leads to a 3D polymer structure of the copper com-
plex of 3. Contrary to the structure of the copper complex
of 5, complexation of 3 gives rise to an novel multibranch
network with open channels parallel to [101] (Figure 12),
which resembles those of metal–organic frameworks
Conclusion
We have reported a new, efficient, and safer synthesis of
3,3’-dimethyl-5,5’-bis(1,2,4-triazine) that avoids the use of
3420
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3414 –3423

Electronic Properties of 3,3’-Dimethyl-5,5’-bis(1,2,4-triazine)
FULL PAPER
potentially explosive reagents. Its crystal structure, establish-
ed for the first time, shows that the molecule adopts an s-
trans conformation in the solid state. Preliminary results on
its complexation properties with metals were also reported,
and spontaneous formation of a novel crystal network with
Method B: KOH (2.3 g, 41.04 mmol) and glyoxal (5.5 mL, 47.8 mmol)
were added to a solution of acetamidrazone (5.0 g, 45.6 mmol) in water
(
30 mL). The mixture turned bright yellow after 15 min. K
2 3
CO (0.457 g,
4
.56 mmol) was added. The reaction mixture was stirred for 3 h and then
warmed to 408C. Then solid KCN (2.0 g, 30.8 mmol) was added, the
color turned red and some precipitate appeared. The temperature was
maintained for a further 1 h. After the stopper was opened, the reaction
mixture turned black. After evaporation of water, the black residue was
extracted by continuous stirring with diethyl ether (3.0 L). The organic
I
Cu cations was demonstrated. In crystallization processes,
kinetic factors are of significant importance. Nevertheless,
we found a very good agreement between the complexation
ability and the electronic features (from experiment or
theory) of the isolated 3,3’-dimethyl-5,5’-bis(1,2,4-triazine)
ligand without any kinetic or thermodynamic considerations.
This shows that an intrinsic molecular property like the elec-
trostatic potential is, to some extent, a good predictive index
for metal chelation.
phase was dried over anhydrous MgSO
afford 3 as an orange crude solid, which was crystallized from CHCl
hexane to afford pure bright yellow crystals (2.34 g, 12.4 mmol, 55%).
4
, filtered, and evaporated to
3
/
1
H NMR (400 MHz, CDCl
3
, 258C): d=10.08 (s, 2H; H-6), 2.97 ppm (s,
1
3
CH
44.7 (C-6), 24.2 ppm (CH
3
); C NMR (100 MHz, CDCl , 258C): d=167.8 (C-3), 151.1 (C-5),
3
1
(
(
3
); IR (KBr): n˜ =3036 (CH arom.), 2926
ꢀ1
CH
3
), 1537 (C=N), 1500 cm (C=C); UV/Vis (MeOH): lmax (e)=360
ꢀ1
3
ꢀ1
9000), 290 nm (25000 mol m cm ); MS (70 eV): m/z (%): 188.2 (25)
+
+
+
[
M] , 160 (25) [MꢀN
for C (188): C 51.06, H 4.25, N 44.68; found: C 51.11, H 4.15, N
4.21.
2 4 3
] , 91 (70) [C HN ] ; elemental analysis calcd (%)
8
8 6
H N
4
Experimental and Computational Section
Crystallization, data collection, and refinement of 3: Yellow crystals of 3
were grown from acetone. A crystal of good quality was chosen for high-
resolution X-ray diffraction. The data were collected at 100.0(1) K on a
Bruker-SMART CCD diffractometer using graphite-monochromated
MoKa radiation. Cooling to 100 K was achieved with an N2 gas stream
(Oxford Cryosystems). The area detector surface was placed 4.02 cm
from the crystal sample. The diffraction data were collected at different
detector positions: 2q=ꢀ60, ꢀ25, 0, +45, +758. The data spots were re-
corded as w scans (Dw=0.208) to reconstruct accurate three-dimensional
diffracted intensity profiles. According to the q dependence of the dif-
fracted intensities, the chosen exposure times were respectively 60, 30,
20, 45, and 90 s per frame for the detector positions given above. The
General procedures: Solvents were purified by standard literature meth-
ods and reagents are used as received from Aldrich. Structures of all
1
13
compounds were assigned by H and C NMR spectra recorded on a
Bruker DRX 400 spectrometer. FTIR spectra were recorded on a Bruker
Vector 22 spectrometer. UV/Vis spectra were recorded on a Safas
2
Uvmc . Mass spectra were recorded on a Trio 1000 spectrometer in EI
mode. Elemental analyses were determined on a Thermofinnigan Flash
EA1112 analyzer.
Hydrazine–hydroquinone complex was synthesized by the literature
[
12]
method.
Acetamidrazone (1): method A: A solution of acetamidine hydrochloride
8.0 g, 84.4 mmol) in MeOH was cooled in an ice bath (0–58C), hydra-
maximum reciprocal resolution reached for this data set is (sinq/l)max
=
ꢀ
1
1
.10 Š . Lorentzian polarization correction and the integration of the
(
diffracted intensities were performed with the SAINT software pack-
zine–hydroquinone complex (12.0 g, 84.4 mmol, 1 equiv) added, and the
mixture stirred for 4 h affording an intense red solution. The solution was
then concentrated in vacuo, and the residue crystallized from n-propanol
at 08C overnight to give pure white crystals of acetamidrazone (5.61 g,
[
22]
age.
An empirical absorption correction was applied using
[
22]
[23]
SADABS.
Finally, SORTAV
was used for sorting and averaging
equivalent and redundant data of high-resolution diffraction experiments.
Table 3 lists the crystallographic data.
6
0%).
method B: A mixture of hydrazine hydrate solution in water (5.0 g,
Structure and density refinements of 3: The crystal structure of 3 was
[
24]
0
1
5
.1 mol) and anhydrous MgSO
5 min. Then acetamidine hydrochloride (9.5 g, 0.1 mol) was added at 0–
8C. The resulting reaction mixture was stirred for a further 4 h at the
4
(3.0 g) was stirred under argon for
solved and refined using WINGX software package.
Starting from
these structural parameters, the conventional and electron density refine-
ments were carried out using MOLLY program based on the Hansen–
[
25]
same temperature. The solution turned from pink to red. After filtration
and evaporation of the solvent, pure acetamidrazone (11.0 g, 0.1 mol)
was obtained quantitatively and used as is for the next step. IR (KBr):
Coppens multipole model. The frozen-core and valence spherical den-
[
26]
sities are calculated from the Hartree–Fock free-atom wave functions.
ꢀ
1
In this study, the x
chosen to be equal to 3.0, 3.8 and n
for C and N atoms, respectively; x
l
exponents (in bohr ) of the radial functions were
ꢀ
1
ꢀ1
n˜ =3356–3169 cm
(NH), 1697 cm
(C=NH); ESIMS: m/z: 74.4
l
=2, 2, 3 up to octupole level (l=3)
+
ꢀ1
[
M+H] .
l
=2.26 bohr and n
l
=1 (dipole level,
[
27]
l=1) for the hydrogen atoms. All the multipole parameters were ob-
tained by least-squares fitting to the observed X-ray diffraction structure
amplitudes F. Before the electron density refinement, the atomic posi-
tions and anisotropic thermal displacements for C and N were estimated
3
-Methyl-1,2,4-triazine (2): method A: A mixture of acetamidrazone
(3.83 g, 45.6 mmol, 1 equiv) in diethyl
ether (80 mL) was stirred at room temperature for 15 min. Glyoxal
7.85 mL, 68.4 mmol, 1.5 equiv) was added to the mixture. The solution
(
5.0 g, 45.6 mmol) and NaHCO
3
(
ꢀ
1
from high-order data (sinq/lꢁ0.8 Š ). The CꢀH distances were con-
became yellow and the mixture was stirred for a further 3 h. Then a satu-
rated solution of NaCl (20 mL) was added, the aqueous phase was ex-
tracted with diethyl ether, and the organic phases were combined, dried
strained to the values observed by neutron diffraction (CaromaticꢀH=
1
.08 Š, CmethylꢀH=1.07 Š). All these structural and thermal parameters
were relaxed in the last cycles of refinements. Figure 13 displays the map
of residual electron density obtained after the multipole refinements. In
this map, the absolute residues of the electron density do not exceed
over anhydrous MgSO
4
, and filtered. The filtrate was concentrated under
reduced pressure to afford a crude yellow mobile and volatile liquid
1
(
(
3.43 g, 79%). H NMR (400 MHz, CDCl
3
, 258C): d=9.12 (s, 1H), 8.55
+
ꢀ3
s, 1H), 2.91 ppm (s, 3H). ESIMS: m/z: 96.4 [M+H] .
0.20 eŠ , and this attests to good convergence of the refinements. The
experimental errors in the electron density are hs (D1)i =0.045 eŠ
2
1/2
ꢀ3
3
7
,3’-Dimethyl-5,5’-bis(1,2,4-triazine) (3): method A: Solid KCN (0.5 g,
.7 mmol) was added to a solution of 2 (0.20 g, 2.1 mmol) in water
2
res
1/2
ꢀ3 [28,29]
and hs
i
=0.063 eŠ .
(
15 mL) and stirred at 408C. The solution turned rapidly red then black
Computational methods: The experimental atomic coordinates of 3 (this
[
2]
and an intensely colored precipitate was immediate formed. The suspen-
sion was stirred for a further 1.5 h and then extracted with diethyl ether
study) and 4 were used for the theoretical calculations. For 5, the
phenyl groups in 4 were replaced by hydrogen atoms 1.08 Š from the at-
tached carbon atom. Ab initio single-molecule calculations were per-
with DFT B3LYP/6-31G++(d,p) basis
For comparison with experimental electron deformation density
maps [D1(r)=1molecule(r)ꢀ1promolecule(r)], the promolecule electron density
(
250 mL). The ether phase was dried over anhydrous MgSO
rated. The residue was purified by column chromatography on silica gel
Et O/hexane 7/3) to afford 3 (0.08 g, 0.4 mmol, 40%) as an orange crys-
4
and evapo-
[
30]
formed with Gaussian03
set.
[
31,32]
(
2
talline powder.
Chem. Eur. J. 2007, 13, 3414 –3423
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3421

A. Marsura et al.
Table 3. Crystallographic data and refinement details of 3 and its copper
complex 6.
3
6
empirical formula
C
8
H
8
N
6
Cu
(C
887.55
monoclinic
P2 /c
11.6531(7)
21.9875(13)
15.3104(9)
110.029(2)
3685.6(4)
1.60
2
A
H
E
N
8
H
8
N
6
5/2 6 2
) (PF ) -
A
T
E
N
[
a]
A
H
R
U
G
3
6
x
formula weight
crystal system
space group
a [Š]
b [Š]
c [Š]
188.2
monoclinic
P2 /c
5.5158(1)
11.6940(2)
9.9613(1)
100.765(1)
860.09(2)
1.45
1
1
b [8]
V [Š ]
3
ꢀ
3
1
calcd [gcm ]
Z
4
4
T [K]
l [Š]
m [mm
100(1)
0.71073
0.100
100(1)
0.71073
1.34
ꢀ1
]
3
crystal size [mm ]
(000)
0.20ꢁ0.08ꢁ0.06 0.20ꢁ0.15ꢁ0.08
392
37598
8962
1.10
3.6
F
A
C
H
T
R
E
U
N
G
1764
24545
9214
0.69
reflections collected
unique reflections
ꢀ
1
A
C
H
T
R
E
U
N
G
(sinq/l)max [Š
]
R
int [%]
10.6
Figure 13. Map of residual electron density of 3 calculated with all data
ꢀ3
after multipole refinement. Contour intervals 0.05 eŠ ; negative con-
tours are dashed.
Spherical refinement
2
2
R1
A
C
H
T
R
E
U
N
G
[F >2s(F )] [%]
5.12
15.7, 0.991
4909
11.93
31.3, 1.03
4095
2
wR2(F ) [%], S
reflections used
[22]
al hemisphere data-collection procedure was used. The crystal struc-
2
2
[
F >2s(F )]
ture of the complex was solved and refined with the WINGX software
number of parameters
159
0.000
0.60, ꢀ0.47
409
0.001
2.32, ꢀ2.77
[24]
package. The hydrogen atoms were placed at idealized positions and
A
C
H
T
R
E
U
N
G
(d/s)max
refined by using the riding-atom method. Highly disordered acetone sol-
ꢀ
3
D1max, D1min [eŠ
]
ꢀ
vent molecules and PF
6
counterions were observed in the crystal lattice
channels. During the structure refinements, the number of acetone sol-
vent molecules was difficult to estimate. Moreover, the peaks located in
the void from the difference Fourier map could not be properly refined.
Multipole refinement
reflections used
3316
2
2
[
F >3s(F )]
R[F] [%]
[F] [%]
GOF
Instead,
a disordered solvent correction was applied using the
2.74
2.27
0.83
[35]
[36]
SQUEEZE procedure in the PLATON program. The final residual
factors were wR2=0.313 and R1=0.119. Crystallographic measurement
and refinement details are given in Table 3. CCDC-628144 and CCDC-
R
w
6
28145 contain the supplementary crystallographic data for the structures
[
3 6
a] The number x of C H O solvent molecules could not be quantified.
of 3 and 6 reported in this paper. These data can be obtained free of
charge from the Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data request/cif.
(
superimposition of spherically symmetrical densities of isolated atoms)
was computed from the same basis set as for the calculation of the molec-
ular density in order to minimize the errors due to the basis-set depend-
ency.
Electrostatic potential: The electrostatic potential is the most important
property that can reveal the interaction of a chemical system with other
species. The highly positive electrostatic potential generated by the
nuclei is compensated by the negative contribution of the surrounding
electrons. Once the electron density is obtained, the calculation and plot-
ting of the electrostatic potential is a convenient and instantaneous way
to reveal the electrophilic and nucleophilic characters of a chemical
system. Furthermore, in many cases, the interaction energy of molecular
systems is dominated by the electrostatic part. This makes the electrostat-
ic potential a predictive property of particular importance for the quan-
tification of the chemical reactivity of molecules. In the present study, the
electrostatic potential was generated on a 3D grid around the molecules
Acknowledgements
The financial support of the CNRS, UniversitØ Henri PoincarØ-Nancy-1,
Ecole Centrale Paris, and UniversitØ Paris XI is gratefully acknowledged.
[
1] a) G. R. Newkome, A.K. Patri, E. Holder, U. S. Schubert, Eur. J.
Org. Chem. 2004, 235–254; b) C. Kaes, A. Katz, M. W. Hosseini,
Annu. Rev. Biophys. Biophys. Chem. Chem. Rev. 2000, 100, 3553–
590; c) H. Hofmeier, U. S. Schubert, Chem. Soc. Rev. 2004, 33,
73–399; d) B. Machura, M. Jaworska, R. Kruszynski, Polyhedron
005, 24, 2, 267–279; e) P. Govindaswamy, Y. A. Mozharivskyj,
M. R. Kollipara, J. Org. Chem. 2004, 69 3265–3274; f) U. Arnold, O.
Walter, M. Dçring, Inorg. Chim. Acta 2006, 359, 1, 327–333; g) Y.-
Q. Huang, B. Ding, H.-L. Gao, P. Cheng, D.-Z. Liao, S.-P. Yan, Z.-
H. Jiang, J. Mol. Struct. 2005, 743, 1–3, 201–207.
[2] a) F. Bodar-Houillon, T. Humbert, A. Marsura, J.-B. Regnouf de
Vains, O. Dusausoy, N. Bouhmaida, N. E. Ghermani, Y. Dusausoy,
Inorg. Chem. 1995, 34, 5205–5209; b) N. E. Ghermani, N. Bouhmai-
da, C. Lecomte, A. L. Papet, A. Marsura, J. Phys. Chem. 1996, 100,
6287–6292.
3
3
2
[
33]
by using the ELECTROS program and the corresponding routine im-
[
30]
plemented in the Gaussian03 package. The MOLEKEL graphic soft-
[
34]
ware was used for visualizing the electrostatic potential.
Crystallization and structure determination of copper complex of 3: Crys-
tals of the complex [(3,3’-dimethyl-5,5’-bis(1,2,4-triazine))5/2Cu (PF (6)
2
]
A
H
R
U
G
6
)
2
were obtained after crystallization by slow evaporation of a solution in
acetone. A suitable platelet-shaped, dark-brown crystal was glued to the
top of a glass fiber for diffraction measurements at 100 K, collected on a
Bruker-SMART CCD diffractometer with MoKa radiation. A convention-
3422
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3414 –3423

Electronic Properties of 3,3’-Dimethyl-5,5’-bis(1,2,4-triazine)
FULL PAPER
[
3] a) F. Bodar-Houillon, Y. Elissami, A. Marsura, N. E. Ghermani, E.
Espinosa, N. Bouhmaida, A. Thalal, Eur. J. Org. Chem. 1999, 1427–
[21] A. J. Fletcher, K. M. Thomas, M. J. Rosseinsky, J. Solid State Chem.
2005, 178, 8, 2491–2510.
1
440; b) J.-B. Regnouf de Vains, J.-M. Lehn, N. E. Ghermani, O. Du-
[22] Bruker SMART (Version 5.054), SAINT (Version 6.36A) and
SADABS (Version 2.05), Bruker AXS Inc., Madison, Wisconsin,
USA, 1998.
sausoy, Y. Dusausoy, A.-L. Papet, A. Marsura, P. Friant, J. L. Rivail,
New J. Chem. 1994, 18, 701–708.
[
4] a) S. Pellet-Rostaing, J.-B. Regnouf de Vains, R. Lamartine, P. Meal-
lier, S. Guitonneau, B. Fenet, Helv. Chim. Acta 1997, 80, 1229–1243;
b) K. Krause, R. A. Krause, S. Lamtruong, J. Coord. Chem. 1988, 19,
[23] R. H. Blessing, J. Appl. Crystallogr. 1997, 30, 421–426.
[24] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[25] N. K. Hansen, P. Coppens, Acta Crystallogr. Sect. A 1978, 34, 909–
921.
9
1–99.
[
[
5] J. Mathieu, PhD thesis, University of Nancy (France), 2003.
6] a) J. Mathieu, B. Fraisse, D. Lacour, N. Ghermani, F. Montaigne, A.
Marsura, Eur. J. Inorg. Chem. 2006, 133–136; b) J. Mathieu, N. E.
Ghermani, N. Bouhmaida, B. Fenet, A. Marsura, Eur. J. Inorg.
Chem. 2004, 5338–5346; c) F. Bodar-Houillon, A. Marsura, Supra-
mol. Chem. 1998, 9, 191–198.
[26] E. Clementi, D. L. Raimondi, J. Chem. Phys. 1963, 41, 2686–2689.
[27] E. Clementi, C. Roetti in Atomic Data and Nuclear Data Tables,
Vol. 44, Academic, New York, 1974, pp. 177–178.
[28] D. W. J. Cruickshank, Acta Crystallogr. 1949, 2, 65–82.
[29] B. Rees, Acta Crystallogr. Sect. A 1976, 32, 483–488.
[30] Gaussian03, RevisionB.04, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr.,
T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J.
Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford,
CT, 2004.
[
7] K. Roussel, A. Cartier, A. Marsura, Chem. Phys. Lett. 2003, 367,
4
63–467.
[
[
8] R. Jenssen, R. T. Pflaum, Anal. Chim. Acta 1965, 32, 235–244.
9] a) D. K. Krass, T, -K. Chen, W. W. Paudler, J. Heterocycl. Chem.
1
973, 10, 343–345; b) D. K. Krass, W. W. Paudler, J. Heterocycl.
Chem. 1974, 11, 43–44; c) W. W. Paudler, R. E. Moser, N. M. Pol-
lack, US 4105434, 1978.
[
10] a) H. Neunhoeffer, H. Hennig, H. W. Frꢂhauf, M. Mutterer, Tetrahe-
dron Lett. 1969, 10, 3147–3150; b) H. Neunhoeffer, F. Weischedel,
Liebigs Ann. Chem. 1971, 749, 16–23.
11] W. W. Paudler, T.-K. Chen, J. Heterocycl. Chem. 1970, 7, 767–771.
12] F. Toda, S. Hyoda, K. Okada, K. Hirotsu, J. Chem. Soc. Chem.
Commun. 1995, 1531–1532.
13] M. N. Burnett, C. K. Johnson, ORTEPIII, Oak Ridge Thermal Ellip-
soid Plot Program for Crystal Structure Illustrations, Oak Ridge Na-
tional Laboratory Report ORNL-6895, Oak Ridge, Tenessee, USA,
996.
14] R. Destro, R. Soave, M. Barzaghi, L. Lo Presti, Chem. Eur. J. 2005,
1, 4621–4634.
15] P. Coppens, T. N. Guru, P. Leung, E. D. Stevens, P. Becker, Y. W.
Yang, Acta Crystallogr. Sect. A 1979, 35, 63–72.
16] R. S. Mulliken, J. Chem. Phys. 1955, 23, 10, 1833–1840.
17] C. M Breneman, K. B. Wiberg, J. Comput. Chem. 1990, 11, 361–373.
18] R. F. W. Bader, Atoms in Molecules: A Quantum Theory, Oxford
University Press, New York, 1990.
19] M. Souhassou, Atomic Properties from Experimental Electron Den-
sities: Program Newprop-Int (LCM3B Internal Report, UniversitØ
Henri PoincarØ, Nancy 1, France), 19th European Crystallographic
Meeting, Nancy, France, 25–31 August 2000, Abstract No. S2-m2-
p2, p. 195.
[
[
[
1
[
[
[31] A. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
1
[32] C. Lee, W. Yang, R. Parr, Phys. Rev. B 1988, 37, 785–789.
[33] N. E. Ghermani, N. Bouhmaida, C. Lecomte, ELECTROS, STAT-
DENS, FIELD+: Computer program to calculate electrostatic
properties from high resolution X-ray diffraction, Internal report
UMR CNRS 7036, UniversitØ Henri PoincarØ, Nancy 1, France and
UMR CNRS 8612, UniversitØ Paris XI, France, 2001.
[34] S. Portmann, H. P. Lꢂthi, MOLEKEL: An Interactive Molecular
Graphics Tool, Chimia 2000, 54, 766–770.
[35] P. van der Sluis, A. L. Spek, Acta Crystallogr. Sect. A 1990, 46, 194–
201.
[
[
[
[
[36] A. L. Spek, J. Appl. Crystallogr. 2003, 36, 7–13.
[
20] G. Henkelman, A. Arnaldsson, H. Jonsson, Comput. Mater. Sci.
Received: November 24, 2006
2
006, 36, 354–360.
Published online: February 7, 2007
Chem. Eur. J. 2007, 13, 3414 –3423
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3423