ChemComm
Communication
by oxidation–reduction processes between the free ligands and
the metal centers. This point is supported by the fact that
ligands 1–3 have been reported to present very similar oxidation
potentials that are markedly higher than those of the related
phosphole derivative.2b Furthermore, as hypothesized for the
formation of complex 4, the pre-coordination of the CuII metal
center in ligands 1 and 3 is probably necessary to afford
transient CuII complexes that can reduce the free CuII metal
centers supplying CuI ions in the reaction media.8 The effect of
the Lewis basicity of the chalcogen atoms carried by the P-atoms
in 1–3 supports the hypothesis that the pre-coordination of the
CuII centers plays a key role in these reactions. In the case of the
thioxophosphole ligand 1 and the selenophosphole ligand 3,
the central chalcogen atoms being soft Lewis donors can fit the
geometrical constraints imposed by the metal ions and the
overall molecular geometry, driving the first step of the reaction
to the formation of a macrocyclic structure bearing distorted
tetrahedral CuI metal centers. This is supported by the observa-
tion that the chalcogen atoms in 5 or 6 tolerate significant bond
lengths and angle differences within the overall similar macro-
cycle architecture. Conversely, the central oxygen atom of the
oxophosphole ligand 2 is a hard Lewis donor that can probably
not tolerate such structural constraints and does not afford
related stable complexes.
Fig. 4 (a) View of the X-ray crystal structure of the metallacycle 6 (H atoms,
counter-anions and included molecules have been omitted), (b) view highlight-
ing the m-1kN:1,2kSe:2kN coordination mode of the ligand 3 in 6 (the remaining
atoms in the coordination sphere of the CuI atom owing to neighbouring ligands
are shown in white), (c) ‘front’ and ‘side’ views of the Cu6Se6 ring observed in 6.
ligand did not react with Cu(OTf )2 (vide infra). Conversely, as observed
in the case of the synthesis of the derivative 5, the stoichiometric
reaction at room temperature of ligand 3 with Cu(OTf )2 in CH2Cl2
solution resulted instantaneously in a dark red clear solution afford-
ing after purification the derivative 6 as a dark red powder.5 The
derivative 6 is characterized at room temperature in CD2Cl2 by a broad
singlet at +34.4 ppm, shifted to high frequencies relatively to the
signal of the free ligand 3 (d = +40.4 ppm in CD2Cl2) and by a single
set of broad signals in the 1H NMR spectrum that compares well with
those of the free derivative 3. Similarly to what was observed in the
case of the derivative 5, single crystals having different morphologies
(6a, dark red blocks 6b, light red plates) were grown in a moderate
yield from the diffusion of pentane vapors into a CH2Cl2 solution of 6.
In both cases, the crystal structure resolution revealed that derivative 6
is a novel [Cu6(3)6](OTf )6 macrocycle having gross geometric para-
meters closely related to those of the thioxophosphole based macro-
cycle 5 (unit cell of 6a is completed by two [Cu(H2O)5(OTf )](OTf )
complexes, six CH2Cl2 and two H2O molecules while the unit cell of
6b is completed by eight CH2Cl2 and two pentane molecules).
In the derivative 6, ligand 3 acts as an 8-electron m-1kN:1,2kSe:2kN
donor on two CuI metal centers. Therefore, the basic units in 5 and 6
are structurally similar. Nevertheless, the PX and Cu–X (P = Se,
ca. 2.14 Å; Cu–Se, ca. 2.50 Å) bonds are, as expected, longer. The
These results highlight how the subtle balance of the stability of
the CuII and CuI oxidation states in the coordination chemistry of
multitopic ligands bearing different heteroatoms can drive the
formation of complex and original molecular architectures. We
are currently investigating the mechanism of the formal redox
processes occurring along the syntheses of the discrete polymetallic
assemblies 4–6 together with the coordination chemistry of ligands
A and 1–3 with other divalent metal centers.
Notes and references
1 (a) B. Rorabacher, Chem. Rev., 2004, 104, 651; (b) F. Durola and
J.-P. Sauvage, Angew. Chem., Int. Ed., 2007, 46, 3537; (c) S. Durot,
F. Reviriego and J.-P. Sauvage, Dalton Trans., 2010, 39, 10557.
´
2 (a) D. Le Vilain, C. Hay, V. Deborde, L. Toupet and R. Reau, Chem.
Commun., 1999, 345; (b) C. Hay, M. Hissler, C. Fischmeister, J. Rault-
´
Berthelot, L. Toupet, L. Nyulaszi and R. Reau, Chem.–Eur. J., 2001, 7, 4222.
´
T-shape geometry of the chalcogen atoms in 6 is less distorted than in 3 (a) B. Nohra, E. Rodriguez-Sanz, C. Lescop and R. Reau, Chem.–Eur. J.,
´
2008, 14, 3391; (b) Y. Yao, W. Shen, B. Nohra, C. Lescop and R. Reau,
Chem.–Eur. J., 2010, 16, 7143; (c) S. Welsch, C. Lescop, G. Balazs,
R. Reau and M. Scheer, Chem.–Eur. J., 2011, 17, 9130.
5: the Se atoms are kept out of the plane made of the Cu and the
P atoms (from 0.01 Å to 0.78 Å), the (P–Se–Cu) angles (from 87.08(7)1
´
to 95.91(5)1) are markedly smaller while the (Cu–Se–Cu) angles are 4 The chemistry of organophosphorus compounds. Phosphine oxides,
sulphides, selenides and tellurides, ed. F. R. Hartley, John Wiley &
Sons, Chichester, 1992, vol. 2.
5 Modification of the ligand/metal salt ratio afforded the same derivative
similar (from 139.52(4)1 to 175.99(3)1) to those observed in the
derivative 5. Within these macrocycles, a non-planar Cu6Se6 ring
(Fig. 4c, maximal deviation from the mean plane: 6a, 0.51 Å; 6b,
0.52 Å) is formed. This ring is markedly distorted compared to
the ‘6-pointed star’ shaped Cu6S6 ring in 5a,b with opposite
edge to edge Cu–Cu and Se–Se distances ranging from 8.82 Å to
10.29 Å and from 6.70 Å to 9.23 Å respectively. Therefore, the
reaction of the selenoxophosphole N,Se,N ligand 3 affords a
hexametallic CuI macrocycle that is similar to the derivative 5
obtained from the thioxophosphole N,S,N ligand 1.
in a lower yield and non-reacted precursors in excess.
6 Hemilabile behaviour for ligand A has been encountered in bimetallic
CuI-complexes. As a result, 1H NMR spectra bearing broad signals are
observed, with a pseudo-symmetry that is not indicative of the real
molecular symmetry.
7 F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, John
Wiley & Sons, New York, 1988, 5th edn.
8 This assumption is supported by the isolation of a few crystals of a
tricationic monometallic complex based on a copper ion, see part S3
in the ESI† for details.
9 (a) E. D. Blue, T. B. Gunnoe and N. R. Brooks, Angew. Chem., Int. Ed.,
2002, 461, 2571; (b) D. Suresh, M. S. Balakrishna and J. T. Mague,
Dalton Trans., 2008, 3272.
The difference in reactivity of the ligands 1–3 towards a CuII
metal center clearly suggests that these reactions are not driven
c
6160 Chem. Commun., 2013, 49, 6158--6160
This journal is The Royal Society of Chemistry 2013