Crystal structure of paddle-wheel sandwich-type [Cu2{(CH3)2CO}{μ-Fe(η5-C5H4C N)2}3](BF4)2·(CH3)2CO

The first ferrocenylcarbonitrile copper complex is reported. The structure consists of two CuI ions complexed by ferrocenediyl-1,1′-dicarbonitrile forming a paddle-wheel with two acetone molecules, with one coordinating on top of one trigonal–planar-coordinated copper ion, and the other as a packing solvent.


Chemical context
The electron-transfer properties of the acetylide function have been investigated intensively by using bridging units of the type -C C-M-C C-(M = transition metal), showing moderate electron communication between two redox-active metallocenyl termini in the mixed-valence species (see, for example: Lang et al., 2006;Vives et al., 2006;Jakob et al., 2009;Díez et al., 2008Díez et al., , 2009Osella et al., 1998;Packheiser et al., 2008;Burgun et al., 2013). The nitrile group is isoelectronic with the acetylide function; Bonniard et al. (2011) described how an -N C-C 6 H 4 -C N-linkage between two iron fragments prohibits the electronic interaction between the transition metal atoms, while the isoelectric di(acetylene)phenylene bridge shows a moderate delocalization. In contrast, a weak electron transfer by generation of the mixedvalence species [Ru(N CFc)(NH 3 ) 5 ] 3+ [Fc = Fe( 5 -C 5 H 4 )( 5 -C 5 H 5 )] has been described (Dowling et al., 1981). We recently reported on the synthesis, characterization and electrochemical properties of platinum and copper complexes containing a -C N-M-N C-(M = Cu or Pt) bridging unit between two redox-active ferrocenyl moieties (Strehler et al., 2013(Strehler et al., , 2014 to achieve a direct comparison with the -C C-M-C C-building blocks. In addition, the coordination of ferrocene-1,1 0 -dicarbonitrile towards PtCl 2 resulted in an oligomeric complex (Strehler et al., 2014). In a continuation of this work, we present herein the synthesis and crystal structure of [Cu 2 {(CH 3 ) 2 CO}{-Fe( 5 -C 5 H 4 C N) 2 } 3 ](BF 4 ) 2 Á(CH 3 ) 2 CO, (I). The synthesis of this compound was realized by comproportionation of elementary copper and a copper(II) salt in the presence of 1,1 0 -ferrocenediyl dicarbonitrile. ISSN 2056-9890

Structural commentary
The title compound contains one pentametallic Cu 2 Fe 3 complex molecule in the asymmetric unit consisting of two Cu I ions bridged by three 1,1 0 -ferrocenediyl dicarbonitrile ligands that form a triangular paddle-wheel sandwich-type complex with ironÁ Á Áiron distances ranging from 9.1739 (13) (Fe2Á Á ÁFe3) to 10.0385 (12) Å (Fe1ctdot;Fe2). The complex crystallizes with two BF 4 À counter-ions and two molecules of acetone. One acetone molecule coordinates with its oxygen atom to Cu1 [Cu1-O1 2.375 (2) Å ], leading to an 18 VE complex and an overall distorted trigonal-pyramidal environment. The Cu2 ion exhibits a weak intermolecular 2 , interaction [3.1520 (6) Å ; Table 1, Fig. 1] with two atoms of an adjacent cyclopentadienyl ring, and thus, only a 16 VE complex is present. The deviation from the N 3 plane is increased for Cu1 [0.1883 (16) Å ] as compared to Cu2 [0.0602 (16) Å ] due to a stronger interaction with the axial moiety. The CuÁ Á ÁCu distance [3.3818 (7) Å ] exceeds the sum of the van der Waals radii (AE = 2.80 Å ; Bondi, 1964), indicating that the Cu I ions do not interact with each other.
The two faces of the sandwich-type complex consist of almost coplanar cyclopentadienyl aromatics and central planes formed by three nitrogen atoms that are also almost coplanar towards the C 5 planes. However, one cyclopentadienyl ring of each site deviates from coplanarity (Table 2), which results in a slight bending of the whole complex (Fig. 2). The ferrocenyl cyclopentadienyl moieties virtually exhibit ecliptic conformations [4.5 (2) to 6.4 (2) ], with synperiplanaroriented carbonitrile substituents towards each other. The molecular structure of (I), showing intermolecular 2 , interactions between Cu2 and the C26-C27 bond, and short interactions between C23 and its symmetry-generated equivalent (Table 1), with displacement ellipsoids drawn at the 50% probability level. All H atoms, the BF 4 À ions and the non-coordinating acetone solvent molecule have been omitted for clarity. [Symmetry codes: (A) x À 1, y, z; (B) 1 + x, y, z; (C) Àx, 1 À y, 1 À z.] Table 1 interactions (Å , ) for (I).
The angle is described by calculating the respectivebond relative to the centroid of the involved aromatic C 5 ring.
p defines a plane calculated by the following atom sequence.

Figure 2
Packing of molecular layers in the crystal structure of (I), with displacement ellipsoids drawn at the 30% probability level. All H atoms have been omitted for clarity. The disorder of one of the counter-anions is not shown.

Supramolecular features
Besides the already noted intermolecular interaction between Cu2 and the mid-point of the C26-C27 bond,interactions are present in the crystal packing between the C23 atom and its symmetry-generated equivalent [3.167 (6) Å ; Table 1]. All other interactions occur almost perpendicular to the involved C 5 ring [ C 5 Á Á ÁC23, 92.2 (2) ; C 5 Á Á ÁCu2, 93.23 (1) ; Table 1]. Compound (I) forms a layer-type structure parallel to (111) (Fig. 2), in which the coordinating acetone molecule is part of the overlaying layer. The second acetone molecule is present in each layer and does not exhibit any notable intermolecular interactions. The distances between two layers are in the range of the above-mentioned interactions.
Regarding nitriles as donating molecules, a tris(benzonitrilo)copper(I) perchlorate complex (Bowmaker et al., 2004) has been reported, exhibiting a similar trigonal-planar coordination environment including the counter-ion acting as one axial ligand with a similar Cu-O distance of 2.404 (4)

Synthesis and crystallization
Ferrocene-1,1 0 -dicarbonitrile was prepared according to a published procedure (Strehler et al., 2014). Synthesis of [Cu 2 {(CH 3 ) 2 CO}{-Fe( 5 -C 5 H 4 C N) 2 } 3 ](BF 4 ) 2 Á(CH 3 ) 2 CO: Copper powder (6 mg, 0.09 mmol), Cu(BF 4 ) 2 Á5H 2 O (12.5 mg, 0.05 mmol) and ferrocene-1,1 0 -dicarbonitrile (50.0 mg, 0.20 mmol) were stirred in 5 ml of dichloromethane at room temperature overnight. The resulting orange precipitate was filtered off using zeolite and washed several times with 20 ml of dichloromethane until the filtrate was colorless. The residue was taken up in acetone and this solution was evaporated to dryness using a rotary evaporator affording (I) as an orange solid. The evaporation was stopped before dryness, small orange crystals of (I) suitable for X-ray crystal structure analysis could be isolated. On further drying, the crystals decomposed due to evaporation of acetone from the crystal.

Refinement details
Crystal data, data collection and structure refinement details are summarized in Table 3   calculated positions and constrained to ride on their parent atoms with U iso (H) = 1.2U eq (C) and a C-H distance of 0.93 Å for aromatic and U iso (H) = 1.5U eq (C) and a C-H distance of 0.96 Å for methyl H atoms. The F atoms of one of the two BF 4 À ions were refined as equally disordered over two sets of sites using DFIX [B-F 1.38 (2) Å ] and DANG [F-F 2.25 (4) Å ] instructions. Since some anisotropic displacement ellipsoids were rather elongated, DELU/SIMU/ISOR restraints were also applied (McArdle, 1995;Sheldrick, 2008).

(Acetone-κO)tris(µ-ferrocene-1,1′-dicarbonitrile-κ 2 N:N′)dicopper(I) bis
where P = (F o 2 + 2F c 2 )/3 (Δ/σ) max = 0.001 Δρ max = 0.59 e Å −3 Δρ min = −0.49 e Å −3 Special details Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å 2 )
x y z U iso */U eq Occ.