Crystal structure of catena-poly[[(μ-6-{[bis(pyridin-2-ylmethyl)amino]methyl}pyridine-2-carboxylato)copper(II)] perchlorate acetonitrile monosolvate]

The crystal structure of the complex CuII-6-carboxylato-2-(pyridylmethyl)-bis(pyridin-2-ylmethyl)amine is reported and compared to similar structures in the literature. The title compound is observed to form extended chains in the solid state, in contrast to the literature examples discussed.

The crystal structure of the title compound, {[Cu(C 19 H 17 N 4 O 2 )]ClO 4 ÁC 2 H 3 N} n , is reported and compared to similar structures in the literature. The compound crystallizes in the monoclinic space group P2 1 . The unit cell contains one complex molecule in addition to perchlorate as the counter-ion and solvent (acetonitrile). The crystal packing evinces extended chains whereby the carboxylate moiety on the 6-carboxylato-2-(pyridylmethyl)bis(pyridin-2ylmethyl)amine ligand bridges between two different copper centers in adjacent molecules. This packing arrangement for the title compound appears to be unique when compared to allied structures in the literature. The perchlorate anion showed signs of disorder and its oxygen atoms were modelled over two sets of partially occupied sites, the occupancy of which was competitively refined to 0.564 (12)/0.436 (12). The crystal studied was refined as a two-component inversion twin.

Chemical context
A key part of the natural nitrogen cycle is the reduction of nitrite to nitric oxide by denitrifying bacteria. The coppercontaining nitrite reductases are one of the classes of enzymes that undertake this reduction reaction (Maia & Moura, 2014). Previous studies have shown that the active site of these enzymes consists of a Cu center coordinated by three N-donor ligands, with the coordination environment being completed by either water or nitrite, depending on the progress of the catalytic cycle (Godden et al., 1991). This realization has spawned a large number of studies examining the use of copper centers held by multi-dentate N-donor ligands as mimics of the active sites of the copper-containing nitrite reductases (Wasser et al., 2002;Timmons & Symes, 2015). In this context, there has been interest from ourselves (Cioncoloni et al., 2018) and others (Komeda et al., 1995;Nagao et al., 1996;Orain et al., 2013) in employing copper complexes based on the tetradentate ligand tris(2-methylpyridyl)amine (TMPA) as electrocatalysts for the reduction of nitrite to nitric oxide, mimicking some of the activity of the copper nitrite reductases.
In the course of our previous study (Cioncoloni et al., 2018), we reported two TMPA-based copper complexes that were electrocatalysts for nitrite reduction to NO: [Cu(OH 2 )(TMPA-CO 2 )] + [where one of the TMPA pyridines bears a carboxylate unit that can coordinate to the Cu center and in which the ligand can be named as 6-carboxylato-2-(pyridylmethyl)bis-(pyridin-2-ylmethyl)amine] and the methylated analogue of this complex, [Cu(OH 2 )(TMPA-CO 2 Me)] 2+ . The acid-bearing ISSN 2056-9890 complex was found to be the more active of the two electrocatalysts. However, although we were able to report the crystal structure of the methyl ester complex after recrystallization from acetonitrile solution, and also the structure of the acid-bearing complex when coordinated to nitrite ([Cu(NO 2 )(TMPA-CO 2 )]), we were not able to obtain similar data for [Cu(TMPA-CO 2 )] + when not bound to nitrite. This was a source of considerable frustration to us at the time, as the structure of the acid-bearing complex prior to nitrite addition constituted a 'missing link' in our characterization of this suite of compounds.
Herein, we report that we have now indeed been able to obtain crystals of the acid-bearing complex in the absence of nitrite ([Cu(TMPA-CO 2 )] + ) by vapour diffusion of diethyl ether into an acetonitrile solution of [Cu(OH 2 )(TMPA-CO 2 )] + . In this manuscript, we report this crystal structure and compare it to the structures of [Cu(NO 2 )(TMPA-CO 2 )], [Cu(NCCH 3 )(TMPA-CO 2 Me)] 2+ , and other allied structures from the literature.

Structural commentary
The Cu ion in [Cu(TMPA-CO 2 )] + is in a hexa-coordinated environment where the two oxygen atoms of the carboxylate group are coordinated by two symmetry-related Cu centers, thus forming extended chains (see Figs. 1 and 2). This suggests that the carboxylate group is deprotonated. The presence of a perchlorate anion in the unit cell, coupled with the hexacoordinate geometry, suggests that the copper is in the +2 oxidation state. The geometry around the Cu center is considerably distorted from octahedral, with all the vertices showing significant deviations from their positions in the platonic scenario as evinced by the following bond angles: N1-Cu1-O2 = 150.1 (2) , O1 i -Cu1-N2 = 161.0 (2) and N4-Cu1-N3 = 158.7 (2) [symmetry code: (i) Àx + 1, y À 1 2 , Àz].
In terms of bond lengths, the Cu-N1 bond length is long [2.296 (6) Å ], consistent with N1 being a tertiary amine, rather than a pyridine nitrogen. The Cu-N pyridyl lengths are all significantly shorter than this Cu-N alkyl interaction, covering the range 2.000 (5)-2.076 (6) Å . There is an intriguing disparity in the lengths of the two Cu-oxygen interactions, with Cu-O1 i being very short [1.965 (5) Å ] in comparison to the Cu-O2 bond length [2.387 (5) Å ]. Coupled with the long Cu-N1 bond, this implies that the N1-Cu1-O2 axis is displaying the Jahn-Teller elongation characteristic of many octahedral Cu II complexes. Indeed, this Jahn-Teller effect manifests along the same axis as we observed previously for the related complex [Cu(NO 2 )(TMPA-CO 2 )] (Cioncoloni et al., 2018).

Figure 1
View showing the asymmetric unit and atom labelling. Displacement ellipsoids are drawn at 50% probability level. The minor disorder component of the perchlorate ion and H atoms are omitted for clarity.

Supramolecular features
The crystal structure shows that the carboxylate moiety on the modified TMPA ligand bridges between two different copper centers, with the result that extended chains of complexes form (see above for a discussion of these coordination bonds). Between the chains, there is some evidence for weak hydrogen bond-type interactions between the perchlorate oxygen atoms and the protons on the -CH 2 -units in the TMPA ligands of complexes in neighbouring chains, and/or between the perchlorate anions and protons on the pyridine rings of some of the TMPA ligands (see Table 2). In this sense, the perchlorate ions can be considered as bridging between adjacent chains of complexes. Meanwhile, the acetonitrile molecule forms a hydrogen bond between its nitrogen and a C-H proton on one of the aromatic rings of the ligand, whilst simultaneously engaging in a hydrogen-bond-like interaction involving one of its methyl protons and the oxygen on a nearby perchlorate anion. However, there would appear to be no direct supramolecular interactions between adjacent chains.

Comparison with related structures
Kojima and co-workers have previously reported complexes of this same (TMPA-CO 2 ) À ligand with Ru III and Cr III (Kojima et al., 2010;Kotani et al., 2015). These complexes all display a distorted octahedral geometry in which the metals coordinate to the three pyridyl nitrogen atoms and the alkyl nitrogen in the TMPA backbone as well as to one of the carboxylate oxygen atoms. However, in these examples the sixth coordination site is occupied by either chloride or BF 4 À (which binds through an F atom) and so the complexes do not form extended chains in the solid state (the carboxylate does not bridge between adjacent metal centers). Likewise, Lonnon et al. (2003)  À interactions tend to be long, and the pyridine bearing the carboxylate unit is found to be trans to this chloride or BF 4 À ligand (and so it tends to display the shortest metal-N pyridyl length in the complex). In our case, however, it is precisely this Cu-N pyridyl bond between the Cu atom and the pyridine that bears the carboxylate substituent which is the longest of the three Cu-N pyridyl bonds. A similar elongation of the Cu-N pyridyl bond for pyridines bearing substituents adjacent to the N-donor has been observed previously by Tanaka and co-workers for the complex [CuCl(TMPA-Me)] + , where the TMPA pyridine bearing the methyl group exhibited a significantly longer Cu-N interaction than that found for the unsubstituted pyridines (2.337 vs. 1.99 Å ; Nagao et al., 1996). We also noted an analogous elongation in our previously reported crystal structure of the complex [Cu(NCCH 3 )(TMPA-CO 2 Me)] 2+ (Cioncoloni et al., 2018). One possible cause of this bond-elongation effect could be steric crowding brought about by the close proximity of the various substituents to the N-donor atom (Symes & Wilson, 2018). In support of this hypothesis, long Cu-N interactions of a similar nature also manifest in certain Cu II -tris(2methylpyridyl)amine complexes (where all the pyridines bear substituents next to the N-donors) reported by Reinaud and co-workers (Izzet et al., 2007).
Suzuki and co-workers have reported two related Cu II complexes with carboxylate-substituted TMPA-like ligands where the carboxylate group binds to the Cu II centre (Hayashi et al., 2002;Mizuno et al., 2006), but for both of these structures the metal is only five-coordinate. The same authors have also published the crystal structure of an allied six-coordinate Ni II complex containing an NiÁ Á ÁO carboxylate interaction (2.084 Å ) and where the sixth ligand is water (Shiren et al., 2000). Again, these complexes appear to exist as discrete ions in the solid state and the formation of extended chains of complexes was not reported.

Figure 2
View showing a chain running parallel to the b axis generated by the 2 1 screw axis. Colour scheme: C = grey, Cl = green, Cu = purple, N = blue, O = red. 2018). In these structures, the metal center adopts a distorted octahedral geometry, with bonds to each of the three pyridines, two bonds to the amines in the bispidine backbone and a final bond to the carboxylate oxygen. Again, then, the carboxylate does not bridge between two Cu centers in neighbouring complexes and so extended chains are not observed.

Synthesis and crystallization
The synthesis and characterization of the complex [Cu(OH 2 )(TMPA-CO 2 )](ClO 4 ) have been reported previously (Cioncoloni et al., 2018). Crystals of [Cu(TMPA-CO 2 )](ClO 4 ) were grown by vapor diffusion of diethyl ether into a 14 mL vial containing 2 mg of [Cu(OH 2 )(TMPA-CO 2 )](ClO 4 ) dissolved in 2 mL of acetonitrile. The 14 mL vial was sealed with a plastic cap, which was pierced by a needle, thus retarding the rate of mixing of the antisolvent into the acetonitrile solution. Crystals suitable for diffraction were obtained after 2-3 weeks.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 3. The perchlorate anion showed signs of disorder and its oxygen atoms were modelled over two sets of partially occupied sites, the occupancy of which was competitively refined to 0.564 (12)/0.436 (12). Similarity distances restraints were applied to Cl-O bond lengths and OÁ Á ÁO separations, as well as to the oxygen displacement parameters. Hydrogen atoms were placed in geometrically calculated positions (C-H = 0.93-0.97 Å ) and refined as part of a riding model with U iso (H) values set at 1.2U eq of the parent carbon atoms, except the methyl hydrogen atoms of the acetonitrile which were refined as a rigid rotor with U iso (H) set at 1.5U eq of the methyl carbon atom. The crystal studied was refined as a two-component inversion twin.

Special details
Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.