Crystal structures of two mixed-valence copper cyanide complexes with N-methylethylenediamine

Two mixed-valence compounds have been isolated from copper-cyanide-meen systems, Cu4(CN)5meen2 and Cu2(CN)3meen2·H2O, where meen is N-methylethylenediamine. The former crystallizes as a polymer, in which CuIImeen2 moieties are covalently linked via cyanide bridges to a three-dimensional CuI cyanide-bridged array, while the latter is a binuclear monomer.


Chemical context
There is continuing interest in the synthesis and structures of coordination polymers involving CuCN networks (Etaiw et al., 2015(Etaiw et al., , 2016Cai et al., 2011). The structure determinations described here arise from our ongoing syntheses of mixedvalence copper cyanide complexes incorporating various amines, with the aim of the directed synthesis of new polymeric structures. A variety of crystal structures form from Cu I,II -cyanide-multidentate amine systems, ranging from the classic three-dimensional mixed-valence structure Cu 3 (CN) 4 en 2 ÁH 2 O where en is ethylenediamine (Williams et al., 1972) to molecular compounds such as Cu 2 (CN) 3 eten 2 (Corfield & Michalski, 2014), where eten is N-ethylethylenediamine. Syntheses involving N-methylethylenediamine, meen, led to the formation of blue crystals of (II), Cu 2 (CN) 3 meen 2 ÁH 2 O, which formed as elongated plates. Their structure described here is that of a molecular compound very similar to the eten derivative referred to above. Syntheses with meen have also been carried out in the presence of tetrahedral monovalent anions such as BF 4 À and ClO 4 À , in the hope that incorporation of negative ions might induce crystallization of a polymeric structure. The major or sometimes sole product in these preparations were well-formed polyhedral black crystals The asymmetric unit for compound (II), Cu 2 (CN) 3 meen 2 ÁH 2 O. Ellipsoids are drawn at the probability 50% level. The refined N-and O-bound hydrogen atoms are emphasized.
Trigonal coordination at Cu3 in (I) is distorted, with angles ranging from 112.56 (12) to 129.79 (13) ; the coordination is rigidly planar, however. In (II), coordination at the trigonal planar Cu I atom is much more regular, with angles ranging from 117.49 (7) to 122.15 (7) .
Both (I) and (II) contain Cu II atoms coordinated by two bidentate meen ligands and the N atom of a bridging cyanide group, in square-pyramidal coordination. Cu atoms are 0.122 (1) and 0.220 (1) Å from the best plane through the amine N atoms in (I) and (II), respectively. In (I), the Cu-NH(CH 3 ) bonds are 0.05-0.07 Å longer than the Cu-NH 2 bonds (Table 1), whereas the corresponding bond lengths are more similar in (II)( Table 2), as also seen in the N-ethyl complex corresponding to (II), Cu 2 (CN) 3 eten 2 (Corfield & Michalski, 2014).
Coordination of the methylated N atom in meen to Cu4 produces a chiral center. N atoms in the (x,y,z) atoms of (I) have the R configuration, and the chelate rings have the conformation, with N-C-C-N torsion angles of 54.6 (4) and 56.0 (4) . Glide-plane-related rings will have the S combination. Methylated N atoms in the Cumeen 2 units in (II) both have the SS configuration, with N-C-C-N torsion angles of À53.0 (2) and À53.1 (2) . The center of inversion in (II) causes an equal number of molecules with the RR combinations. The CH 3 -N-C-C torsion angles in the chelate rings depend on the R/S and / combination. For an R combination, this angle will be approximately À170 , and for R the angle will be about À90 . These angles are reversed in sign for the S and S combinations. CH 3 -N-C-C angles are À172.8 (3) and À167.4 (3) in (I), and 175.0 (2) and 174.5 (2) in (II). Averages for these angles in 24 Cumeen chelate rings are reviewed in the Database Survey section.

Supramolecular features
In (I), each dimeric Cu 2 (CN) 6 unit is linked by the C4-N4 cyanide group to a screw-related Cu 2 (CN) 6 unit to form chains of these units parallel to the c axis, Fig. 3. Trigonally coordinated Cu3 also links the Cu 2 (CN) 6 units together via CN bridges into single-stranded chains along the 8.231 (1) Å b axis, Fig. 4, similar to the double-stranded chains along the 8.356 (1) Å a axis seen in the polymeric compound (et 2 oenH)[Cu 2 (CN) 3 ], (Corfield et al., 2016), where et 2 oen is N,N-diethylethanolamine. The columns are further linked together by Cu3 to form a structure with channels, into which projects the coordinated Cumeen 2 unit, Fig. 3. The topology around Cu3 involves three 20-membered rings. There are four close interactions between amine N-H bonds and bridging CN groups, with NÁ Á ÁN distances ranging from 3.257 (3) to 3.479 (3) Å , which may account for the tendency for ordered CN groups in this structure. The shortest HÁ Á ÁH intermolecular contact in (I) is 2.47 Å for H15BÁ Á ÁH18B(À 1 2 À x, À 1 2 + y, À 1 2 + z). Centrosymmetric pairs of discrete molecules of (II) are held together by hydrogen bonding (Table 3) to the water molecules, Fig. 5, with each water molecule forming one donor and one acceptor hydrogen bond. These pairs are linked into chains via hydrogen bonds along [011], N4-H4AÁ Á ÁÁ Á ÁN3(1 À x, 1 À y, 2 À z)-C2(1 À x, 1 À y, 2 À z), where these four atoms are almost co-linear. Two other potentially attractive relationships between N-H bonds and cyanide groups are also shown in  Table 2 Selected bond lengths (Å ) for (II).     The Cumeen set showed the same lengthening of the Cu-NH(CH 3 ) bonds with respect to the Cu-NH 2 bond lengths as found here in (I), with averages of 2.010 (4) and 2.041 (4) Å , respectively. A similar difference was found for the Comeen set, where the corresponding means were 1.962 (8) and 1.998 (8) Å . Cu-N bond lengths showed no correlation with coordination numbers around Cu, which ranged from four through six. N-Cu-N angles in the Cumeen set are in a limited range of 84.0 to 86.4 , and the four such angles in the present study all lie near the middle of this range.
The average of the absolute values of the N-C-C-N torsion angles in the chelate rings for the Cumeen set is 51.6 , with a sample s.u. of 6.5 , excluding one outlier from a flat chelate ring. Corresponding angles in the present work are all within one s.u. of the mean. The mean absolute CH 3 -N-C-C angles for R/S and R/S combinations, respectively, in the Cumeen set are 171 (6) and 89 (5) , where sample s.u.'s are given. Equivalent torsion angles in both structures presented here fall within one s.u. of these means.

Synthesis and crystallization
The compounds were synthesized by air oxidation of CuCN/ NaCN/meen systems. A typical preparation of (II) had CuCN (5.7 mmol) and NaCN (8.3 mmol) stirred in 6 mL of water until all solids dissolved, when 8.6 mmol of N-methylethylenediamine (meen) in approximately 5 mL of water were added. Blue crystals in the form of extended thick plates were recovered after two days at room temperature. Crystals of (I) were obtained in a similar preparation with 11.5 mmmol CuCN, 16.5 mmol NaCN, and 16.2 mmol meen, to which were added an aqueous solution containing 9.9 mmol NaClO 4 . Blue crystals of (II) were obtained after two weeks, but after another six weeks the filtrate yielded large black polyhedral crystals of (I).
Infra-red spectra obtained with a Nicolet iS50 FT-IR machine on the polymer (I) showed three bands in the CN stretching region, with peaks at 2079, 2109, and 2119 cm À1 . In addition, there are strong bands at 3250 and 3312 cm À1 , and a weak, sharp band at 3150 cm À1 , presumably all due to N-H stretching vibrations. For (II), CN stretching frequencies at 2089 and 2115 cm À1 were observed.

Refinement details
Crystal data, data collection and structure refinement details are summarized in Table 4. In (I), it was apparent that several low-order reflections were partially or completely obscured by the backstop and/or subject to overload. We recollected a fast dataset to = 15 with the backstop pushed back, obtained the scale factor between the two datasets using reflections with above 5 , and replaced 27 low-angle reflections in (I) with data from the fast dataset. Three low-angle reflections were not obtained in the fast dataset, and these have been omitted in the final refinement.
In (I), 3 -CN cyanide groups C1 N1 and C2 N2 were found to be ordered, with the C atom bridging two Cu atoms, as in Corfield et al. (2016). In addition, C5 N5 was found to have a clearly preferred orientation and was refined as an ordered group. C,N occupancy factors were refined for the two other cyanide groups, with preferential occupancies of 79 (3)% and 78 (3)% found for C3 N3 and C4 N4, respectively. Only the major C or N atoms are listed in the cif tables of bond lengths and bond angles. In (II), all the CN groups are ordered; CN orientations were checked by refinements with interchange of each CN group in turn, in each case resulting in significantly higher R factors.  Table 3 Hydrogen-bond geometry (Å , ) for (II). (2) 3.302 (3) 170 (2) Symmetry codes: (i) Àx þ 1; Ày; Àz þ 1; (ii) Àx þ 1; Ày þ 1; Àz þ 2; (iii) x þ 1; y þ 1; z.

Figure 5
Packing diagram for compound (II), Cu 2 (CN) 3  In both compounds, C-bound H atoms were constrained to idealized positions with C-H distances of 0.97 Å for CH 2 groups and 0.96 Å for CH 3 groups, and U eq values fixed at 1.2 times the U iso of their bonded C atoms. The methyl torsion angles were refined. In (II), the N-and O-bound hydrogen atoms were clearly visible in the difference-Fourier map and were refined independently. The N-bound hydrogen atoms in (I) were clearly seen in a near-final difference map, and could be independently refined, but we chose to constrain them to idealized positions, with N-H distances of 0.90 Å for NH 2 groups, 0.91 Å for NH groups, and U eq values treated the same as for the C-H atoms.
For both (I) and (II), data had been previously collected with a CAD-4 system (Enraf- Nonius, 1994), on three crystals in the case of (I), and two crystals for (II). For (I), final R 1 factors for the CAD-4 data were 0.045 for 2228 data with F 2 > 2, while for (II), R 1 was 0.036 for 2245 data with F 2 > 2. It was felt instructive to compare refined parameters obtained by the two methods. We defined Á/ for a given parameter as the absolute value of the difference between the parameters determined by the two instruments divided by the square root of the sum of the squares of the standard deviations for the two parameters. For (I), the structural parameters agreed very well, for the mean and maximum Á/ for all parameters were 0.74 and 2.60. The maximum deviation for bond lengths was 2.1. For (II), there were differences of 4-5 between positional parameters for the water oxygen atom, O1, which was much better defined in the data set from the KappaCCD instrument. Apart from parameters for O1, the agreement was excellent, with average Á/ for positional parameters 0.79, and no Á/ greater than 3. There were differences in the U ij for the two Cu atoms because an extinction parameter was refined for the KappaCCD data set. For all other atoms, the mean Á/ for the thermal parameters was 0.83 with only one Á/ greater than 3. -Absolute structure parameter 0.010 (9) -Computer programs: KappaCCD Server Software (Nonius, 1997), DENZO and SCALEPACK (Otwinowski & Minor, 1997), SHELXS97 (Sheldrick, 2008), SHELXL2014 (Sheldrick, 2015), ORTEPIII (Burnett & Johnson, 1996) and publCIF (Westrip, 2010 For both compounds, data collection: KappaCCD Server Software (Nonius, 1997); cell refinement: SCALEPACK (Otwinowski & Minor, 1997); data reduction: DENZO and SCALEPACK (Otwinowski & Minor, 1997); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL2014 (Sheldrick, 2015); molecular graphics: ORTEPIII (Burnett & Johnson, 1996); software used to prepare material for publication: publCIF (Westrip, 2010).

(I) Poly[bis(µ 3 -cyanido-κ 3 C:C:N)tris(µ 2 -cyanido-κ 2 C:N)bis(N-methylethane-1,2-diamine\κ 2 N,N′)tricopper(I)copper(II)]
Crystal data Special details Experimental. Scalepack values for Tmin and Tmax are normalized to unity. Values given here were obtained by multiplying them by exp(-µd) where d= 0.45mm. 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. 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 > 2sigma(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. (

Special details
Experimental. Scalepack values for Tmin and Tmax are normalized to unity. Values given here were obtained by multiplying them by exp(-µd) where d= crystal_size_mid. 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.