X-ray structure and density functional theory studies of an unexpected product: trans-bis­{2-[(2-cyano­ethyl)imino­methyl]­phenolato}copper(II)

The title compound, [Cu(C₁₀H₉N₂O)₂] or [Cu^II(CYMB)₂], (I), was obtained in an attempt to reduce trans-bis­(2-{[3,5-bis(trifluoro­methyl)phenyl]imino­methyl}phenolato)copper(II), [Cu(TIMB)₂], (II), with bis­(penta­methyl­cyclo­penta­dienyl)­cobalt(II) [deca­methyl­cobaltocene, Cp*₂Co, (III)]. The mol­ecular structure of (I) has the Cu^II centre located on an inversion centre of the C2/c space group. A density functional theory (DFT) analysis at the B3LYP/Lanl2dz(CuF);6-31G**(CHNO) level performed in order to optimize the structures of the free ligands CYMB^- and TIMB^-, and the metal complexes [Cu^I/II(CYMB)₂]^-/0 and [Cu^I/II(TIMB)₂]^-/0, reproduced well the X-ray diffraction structure and allowed us to infer the insertion of the cyano­methide anion on the 3,5-bis­(trifluoro­methyl)phenyl system from an evaluation of the Mulliken atomic charges and the electronic energies.

S1. Comment

The interest in Schiff base salicylaldehyde copper(II) complexes of the type shown in the first scheme stems in a large part from their redox properties. In fact, chemical and/or electrochemical reactions have been shown to produce the corresponding Cu^I and Cu^III species. In recent years, several studies have been reported in the literature on the bioinorganic behaviour of these complexes as models for biological systems (Abdel-Latif et al., 2007, and references cited therein; Toyota et al., 2001; Filomeni et al., 2007) and their photochromic, thermochromic (Chatziefthimiou et al., 2006) and magnetic properties (Bian et al., 2003). The molecular structure of such complexes has been accurately investigated via spectroscopic and X-ray diffraction (XRD) methods. The Cambridge Structural Database (CSD, Version?; Allen, 2002) shows hundreds of hits for this type of complex. The original papers combine the structural data with redox, electron paramagnetic resonance and UV–Vis properties. Recently, some of us reported the properties of a series of derivatives with R₃–R₆ = H and R₁ = R₂ = 3,4,5-trisubstituted-phenyl (where the substituents consist of a variety of atoms or groups). The derivative with R₁ = R₂ = 3,5-bis(trifluoromethyl)phenyl has been investigated for its redox and spectroscopic properties but not through XRD, while the analogous 3,5-dimethylphenyl derivative has been investigated via XRD (see Corsini et al., 2003, and references cited therein). Against this background, the study of the reactivity of trans-bis({[3,5-bis(trifluoromethyl)phenyl]iminomethyl}phenolato)copper(II), [Cu(TIMB)₂], (II), with reducing agents could be worthwhile. We report here the results of a study performed first by reacting (II) with Cp*₂Co, (III), in acetonitrile at reflux under a nitrogen atmosphere. The reaction produced crystalline (I), the title compound, which surprisingly contains the unreduced Cu^II–CYMB chelate. This, to the best of our knowledge, constitutes a new ligand, with the CH₂-CH₂-CN function linked to the imine N atom in the place of the 3,5-bis(trifluoromethyl)phenyl group, and its structure has not yet been deposited in the CSD or elsewhere.

The reaction of (II) with (III) at reflux for 1.5 h under nitrogen and in the absence of light afforded a dark-green solution from which dark-green crystals of (I) were formed (see second scheme). The complex molecule is shown in Fig. 1, and selected bond distances and angles are reported in Table 1. The Cu^II centre lies on a crystallographic inversion centre and is chelated by two ligand molecules through the enolate O atom and the imino N atom, the two ligands being arranged in trans to each other. As a consequence, the coordination environment is square planar. The nitrile function is not involved in any interactions with the metal centre. It has to be noted that the structure of (I) is reminiscent of that of [Cu(DIMB)₂], where DIMB = [(3,5-dimethylphenyl)imino]methyl}phenolato [Please ensure brackets balance], which is an analogue of (II) having methyl groups in place of the trifluoromethyl substituents (Corsini et al., 2003). In the case of [Cu(DIMB)₂], two different coordination molecules are present in the asymmetric unit, the coordination environments of which are significantly tetrahedrally distorted (the distortion being due to the steric hindrance arising from the 3,5-dimethylphenyl substituent on the imino N atom. The Cu—O and Cu—N bond distances are 1.893 (3) and 2.007 (3) Å, respectively, in perfect agreement with the values previously reported for similar compounds: 1.899 (1) and 1.962 (1) Å (Corsini et al., 2003); 1.892 (2) and 1.895 (2), and 1.992 (2) and 1.995 (4) Å (Saha et al., 2003); 1.832 (2) and 1.913 (2) Å (Zhang, 2004); 1.888 (4) and 2.002 (4) Å (You et al., 2004). The angles at the metal atoms are close to the idealized values of 90 and 180°. In fact, O1—Cu1—N1 [91.8 (1)°] and O1—Cu1—N1(-x + 1/2, -y + 1/2, -z + 1/2) [88.2 (1)°] deviate by just 1.8 (1)° (absolute value). The Cu^II centre deviates slightly [0.0692 (1) Å] from the plane defined by the five-atom chelating ring O1/C5–C7/N1 and the deviation is opposite to the nitrile function. The metal centre is also almost coplanar with the C1–C6 aromatic ring [deviation 0.0282 (1) Å] and the coordination plane deviates by ca 3° from both the chelating plane and the aromatic plane.

The bond distances and angles within the ligand have normal values on the basis of comparative analysis with literature data. For example, the C—O and C═ N bond distances are 1.308 (5) and 1.285 (5) Å in the structure of (I), and average 1.301 (4) and 1.291 (4) Å for [Cu(DIMB)₂]. The iminomethylphenolate moiety is planar, the dihedral angle between the benzene ring and the chelating ring being 2.9 (1)°. The propylnitrile branch protrudes out of the plane of the rest of the ligand and the two arms are opposite each other. The Cu1—N1—C8—C9 and N1—C8—C9—C10 torsion angles are -76.1 (4) and -65.7 (4)°, respectively, whereas the angle between the line defined by atoms N2 and C10 and the normal to the coordination plane is 47.6 (4)°. The N2—C10 bond distance is 1.129 (5) Å, in agreement with uncoordinated nitrile functions (Britton et al., 2004), and the C9—C10—N2 bond angle is 178.7 (5)°. It has to be noted that nitrile/isonitrile isomerism is badly distinguished by the C—N bond distance (Britton et al., 2004). As a consequence, the nitrile model for the present structure was checked through a comparative analysis with the corresponding isonitrile moiety. The conventional agreement factors R1 and wR2 at the same final stage of refinement were 0.0474 and 0.1262 for nitrile, and 0.0512 and 0.1411 for isonitrile, respectively. Thus, the nitrile model was accepted as the correct one.

The complex molecules are stacked almost parallel to the b crystallographic cell edge when the line that connects the metal centres is taken into account. However, the molecular planes and the b edge are tilted by 45°. Adjacent columns of molecules are set in a V-type arrangement (see Supplementary Material, Fig. S1). The shortest contact distance between metal centres is 5.400 (2) Å, so that any direct Cu···Cu interaction can be excluded in the solid state. It has to be noted that the stacking mode is stabilized by π–π stacking interactions that mainly involve the benzo ring of one molecule and the imino and enolate system of an adjacent one (Fig. 2). For example, the C2···C7 contact distance between two adjacent molecules is 3.400 (7) Å, whereas the contact distance between atoms O1 and C5 is 3.428 (6) Å. Noticeably, this stacking mode has infinite chains of C4···Cu···C4···Cu atoms, with an interatomic distance of 3.345 (4) Å and an angle at the metal of 180.0 (1)°. On the basis of the values of the van der Waals radii for a Cu atom (1.4 Å) and for a C atom (1.7 Å) (Bondi, 1964), a linking interaction between C4 and Cu seems to be absent in (I). Thus, the linking force between adjacent molecules should stem only from the π–π van der Waals type bond between delocalized electron systems. Interestingly, intermolecular C7—H···N2 hydrogen-bond type interactions involve the imino and nitrile functions, with C···N = 3.549 (7) Å and C7—H···N2 = 177 (1)°. Chains of complex molecules are built up through these interactions (see Supplementary Material, Fig. S1).

A density functional theory study was carried out on the CYMB– and TIMB–Cu^I/II derivatives. The optimized structures for [Cu^I/II(CYMB)₂]^-/0 and [Cu^I/II(TIMB)₂]^-/0 are represented in Fig. 3 (selected computed geometric parameters are reported in the Supplementary Material). The agreement between computed and experimental bond distances for (I) is good, with the computed ones (coordinate bond) being larger by less than 0.048 Å, whereas those for the ligand have an even better accuracy. It must be noted that computed bond distances in the gas phase are usually larger than the corresponding values found in the solid state (Cini et al., 2007; Tamasi et al., 2009). The BS5 basis set is more accurate than BS4, as expected, and has roughly the same accuracy as the more expensive BS6 in terms of geometric parameters. Therefore, the following analysis is mostly based on computations performed at the BS5 level for all the molecules. Even bond angles and torsion angles computed at the BS5 level are very accurate, suggesting that intermolecular forces in the solid state do not have a large effect on ligand conformation.

The computed structure for [Cu^I(CYMB)₂]^- has a pseudo-linear N,N-coordination arrangement. In fact, the computed Cu—N and Cu—O bond distances are 2.029 and 2.303 Å, respectively. This arrangement could not be confirmed from XRD structures for mononuclear copper(I) complexes with salicylideneimine ligands. In fact, no structure for this latter type of complex was deposited in the 2008 release of the CSD and no other structural assignment was found in the literature from spectroscopic data. The electronic energies for complex formation in the gas phase for [Cu^I/II(CYMB)₂]^-/0 from the free metal cation and free anionic ligands are -426.63 (Cu^I) and -856.35 kcal (Cu^II) (1 kcal mol^-1 = 4.184 kJ mol^-1), showing that the Cu^II species is intrinsically much more stable than the Cu^I one. Once computed at BS4, a similar gap between formation energy was found (438.03 instead of 429.72 kcal). The computed structure for [Cu^II(TIMB)₂] (BS5) has Cu—N and Cu—O bond lengths of 2.040 and 1.918 Å, respectively. The corresponding values computed at BS4 are close (2.021 and 1.922 Å) and a good agreement could also be found with the values determined in the solid state for the related [Cu^II(DIMB)₂] complex [1.962 (2) and 1.899 (2) Å, respectively; Corsini et al., 2003]. The computed structure for [Cu^I(TIMB)₂]^- (BS4) has a pseudo-tetrahedral coordination arrangement and the Cu—N and Cu—O bond distances are 2.051 and 2.076 Å, respectively. Interestingly, the formation energy for [Cu^I(TIMB)₂]^- is less favourable by 25 kcal than that for [Cu^I(CYMB)₂]^-, in agreement with the large steric hindrance generated by the (3,5-trifluoromethyl)phenyl substituents at imine N atom in the TIMB^- derivative compared with the cyanoethyl groups for the CYMB^- one. These effects play a part in directing the reaction of [Cu^II(TIMB)₂] with Cp₂*Co^II (in anhydrous acetonitrile) towards a pathway that ends in [Cu^II(CYMB)₂], instead of causing a reduction at the metal centre (and formation of [Cu^I(TIMB)₂]^-).

In order to confirm better the accuracy of the computations, the calculated vibration frequencies were compared with the experimental ones for [Cu^II(CYMB)₂] and [Cu^II(TIMB)₂] (obtained by FT–IR spectroscopy). The absorption band (weak, w) at 2241 cm^-1 for (I) is attributable to the nitrile C—N stretching vibration (computed at BS4: 2276 cm^-1, 19.946 km mol^-1). Thus, the experimental-to-computed frequency ratio (0.985) is good when compared with literature values (Foresman & Frisch, 1996). Other relevant absorption bands for (I) are located at 1623 (strong, s), 1352 (medium, m) and 760 (m) cm^-1. They are assignable to, respectively, imine C—N stretching vibrations coupled with aromatic ring motions (1657 cm^-1, 922.457 km mol^-1), and H—C—C bending in (1339 cm^-1, 381.224 km mol^-1) and out of the plane (790 cm^-1, 179.070 km mol^-1). As expected, the experimental and computed spectra for [Cu^II(TIMB)₂] do not have any appreciable absorption band in the region 1800–3000 cm ^-1, whereas selected strong effects were found at 1603, 1370 and 768 cm^-1 that correspond to imine C—N stretching vibrations coupled with salicyl ring motions (1622 cm^-1, 664.317 km mol^-1) and H—C—C bending in (1413 cm^-1, 282.760 km mol^-1) and out of the plane (791 cm^-1, 153.433 km mol^-1). Computed absorption effects (very weak) due to Cu—N and Cu—O stretching vibrations for (I) occur at 578 and 476, and 615 and 594 cm^-1, respectively.

Another noteworthy result from the computations is the trend in atomic charges. In fact, the values from the Mulliken population analysis for [Cu^II(TIMB)₂] show that atom C8 (Fig. 3b) has the highest positive charge compared with all the other phenyl ring C atoms. The charges computed at BS4 and BS5 for atom C8 are 0.472 and 0.455 e (single-point calculations for a non-optimized structure built up from the XRD structure of [Cu^II(DIMB)₂]), whereas those for the other five C atoms are equal to or lower than 0.303 e (see Supplementary Material). A similar trend was found for the optimized structure. The trend in atomic charges in the free TIMB^- ligand is significantly different. In fact, the positive charge on C8 is smaller than that of the other phenyl ring C atoms. This means that the metal ligation to TIMB^- favours nucleophilic attack on C8. One can expect that, if the reaction media provide suitable nucleophilic agents, the reaction could occur on that site. This could happen when Cp₂*Co^II is refluxed in anhydrous acetonitrile. In fact, the formation of a cyanomethide anion and a Cp₂*Co^III hydride species is reasonable when it is considered that several such compounds have previously been reported in the CSD and other literature databases (e.g. Heeres et al., 1990). Of course the hypothesis must be studied via deeper theoretical and experimental effort. Nevertheless, the computational part of this work has given sound preliminary explanations for the unexpected result of the attempted chemical reduction of a copper(II) complex.

S2. Experimental

A clear yellow–brown solution was prepared by dissolving (II) (14 mg, 0.02 mmol) in de-aerated CH₃CN (7 ml). A second clear yellow solution was prepared by dissolving (III) (7 mg, 0.037 mmol) in de-aerated CH₃CN (25 ml). The two solutions were mixed and the final clear brown mixture was heated at reflux with stirring, in an anaerobic atmosphere (N₂, 1 atm). The colour turned to dark green after ca 30 min reflux; the heating was continued for 60 min. The resulting solution was left to stand, kept under nitrogen and in the dark, for 24 h. Needle-shaped dark-green crystals of (I) deposited on slow cooling and storing. These were filtered off, washed with small portions of cold acetonitrile and then stored in air (yield 6 mg, 76%). The IR spectra were recorded in the solid state from KBr matrixes at 298 K using a Perkin–Elmer 1600 FT–IR spectrometer.

S3. Refinement

All the H atoms placed in geometric positions and allowed to ride on their parent C atoms, with C—H = 0.93–0.97 Å [Please check added text], and with U_iso(H) = 1.2U_eq(C). All density functional calculations were performed using GAUSSIAN03, Revision D.02 (Frisch et al., 2004), implemented on an IBM SP5 machine at CINECA (Inter-University Computing Centre, Casalecchio di Reno, Bologna, Italy). All structures were fully optimized (unless otherwise specified, see text and Supplementary Material) using B3LYP methods and different levels of theory (Frisch & Frisch, 1998) for comparative purposes. The free ligands, CYMB^- and TIMB^-, were calculated at the Lanl2dz (BS1), 6-31G** (BS2) and 6-311++G** (BS3) levels for all atoms. The copper complexes, [Cu^I/II(CYMB)₂]^-/0 and [Cu^I/II(TIMB)₂]^-/0, were calculated at the Lanl2dz(CuCHNOF) (BS4), Lanl2dz(CuF);6-31G**(CHNO) (BS5) and Lanl2dz(CuF);6-311++G**(CHNO) (BS6) levels (Frisch & Frisch, 1998). The input structures for both copper(II) complexes were from experimental X-ray data {complex (I) from this study and the analogous data for complex (II), [Cu^II(DIMB)₂], having CH₃ groups instead of CF₃; see Corsini et al., 2003}. Single-point calculations were performed at the BS4 and BS5 levels of theory in order to estimate the Mulliken atomic charge (e) and the total electronic energies (Hartrees). The optimized structures of complexes [Cu^II(CYMB)₂] and [Cu^II(TIMB)₂] were used as input for the computation of [Cu^I(CYMB)₂]^- at the BS4 and BS5 levels and [Cu^I(TIMB)₂]^- at the BS4 level of theory. All calculations were performed in the gas phase without any constraint. A hessian analysis was also performed. Molecular drawings were obtained using GAUSSVIEW03 software (Dennington et al., 2003) implemented on a Pentium IV machine.

Computing details

Data collection: XSCANS (Siemens, 1994); cell refinement: XSCANS (Siemens, 1994); data reduction: XSCANS (Siemens, 1994) and XEMP (Siemens, 1994); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008) and WinGX (Farrugia, 1999); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008) and WinGX (Farrugia, 1999); molecular graphics: ORTEPIII (Johnson & Burnett, 1996), Mercury (Macrae et al., 2008) and GAUSSVIEW03 (Dennington et al., 2003); software used to prepare material for publication: CIFTAB (Sheldrick, 2008) and PARST97 (Nardelli, 1997)'.

Figures

	Fig. 1. The X-ray structure of complex (I), with the atom-numbering scheme. Displacements ellipsoids are drawn at the 30% probability level. The Cu^II atom lies on an inversion centre.
	Fig. 2. A diagram of the π–π stacking interactions between the benzo ring of one molecule of (I) and the imino and enolate systems of an adjacent molecule at symmetry position (x, y + 1, z).
	Fig. 3. Drawings of (a) the [Cu(CYMB)2] complex, (I), and (b) the [Cu(TIMB)2] complex, (II), as optimized at the BS5 level of theory.

trans-bis{2-[(2-cyanoethyl)iminomethyl]phenolato}copper(II)

Crystal data

[Cu(C10H9N2O)2]	F(000) = 844
Mr = 409.93	Dx = 1.456 Mg m^−3
Monoclinic, C2/c	Mo Kα radiation, λ = 0.71073 Å
Hall symbol: -C 2yc	Cell parameters from 28 reflections
a = 20.655 (6) Å	θ = 5–19.5°
b = 5.400 (1) Å	µ = 1.19 mm^−1
c = 17.653 (3) Å	T = 293 K
β = 108.29 (1)°	Needle, dark green
V = 1869.5 (7) Å^3	0.40 × 0.15 × 0.15 mm
Z = 4

Data collection

Siemens P4 diffractometer	1269 reflections with I > 2σ(I)
Radiation source: fine-focus sealed tube	Rint = 0.047
Graphite monochromator	θmax = 25.0°, θmin = 2.1°
ω scans	h = −1→24
Absorption correction: ψ scan (XEMP; Siemens, 1994)	k = −1→6
Tmin = 0.810, Tmax = 0.835	l = −20→19
2211 measured reflections	3 standard reflections every 97 reflections
1652 independent reflections	intensity decay: none

Refinement

Refinement on F^2	Primary atom site location: structure-invariant direct methods
Least-squares matrix: full	Secondary atom site location: difference Fourier map
R[F^2 > 2σ(F^2)] = 0.047	Hydrogen site location: inferred from neighbouring sites
wR(F^2) = 0.140	H-atom parameters constrained
S = 1.03	w = 1/[σ^2(Fo^2) + (0.0964P)^2 + 0.2198P] where P = (Fo^2 + 2Fc^2)/3
1652 reflections	(Δ/σ)max < 0.001
124 parameters	Δρmax = 0.68 e Å^−3
0 restraints	Δρmin = −0.76 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 F² against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F², conventional R-factors R are based on F, with F set to zero for negative F². The threshold expression of F² > σ(F²) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F² 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 (Ų)

	x	y	z	Uiso*/Ueq
Cu1	0.2500	0.2500	0.0000	0.0390 (3)
O1	0.25407 (14)	−0.0217 (5)	−0.06609 (17)	0.0519 (7)
N1	0.14856 (16)	0.2159 (5)	−0.02631 (19)	0.0392 (7)
N2	0.0662 (2)	−0.1021 (7)	0.1112 (3)	0.0769 (13)
C1	0.11210 (19)	0.0623 (7)	−0.0778 (2)	0.0423 (9)
H1	0.0651	0.0740	−0.0889	0.051*
C2	0.1360 (2)	−0.1249 (7)	−0.1200 (2)	0.0431 (9)
C3	0.0863 (2)	−0.2851 (8)	−0.1710 (2)	0.0514 (10)
H3	0.0404	−0.2596	−0.1773	0.062*
C4	0.1049 (2)	−0.4749 (9)	−0.2107 (3)	0.0578 (11)
H4	0.0721	−0.5796	−0.2433	0.069*
C5	0.1733 (2)	−0.5096 (8)	−0.2019 (2)	0.0554 (11)
H5	0.1862	−0.6381	−0.2293	0.066*
C6	0.2229 (2)	−0.3574 (8)	−0.1534 (3)	0.0499 (10)
H6	0.2685	−0.3847	−0.1486	0.060*
C7	0.2054 (2)	−0.1618 (7)	−0.1110 (2)	0.0417 (9)
C8	0.1101 (2)	0.3858 (7)	0.0088 (2)	0.0452 (9)
H8A	0.0621	0.3772	−0.0216	0.054*
H8B	0.1254	0.5541	0.0056	0.054*
C9	0.1193 (2)	0.3228 (8)	0.0957 (2)	0.0463 (9)
H9A	0.1675	0.3221	0.1254	0.056*
H9B	0.0977	0.4503	0.1183	0.056*
C10	0.0899 (2)	0.0818 (8)	0.1048 (3)	0.0534 (11)

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
Cu1	0.0385 (4)	0.0392 (4)	0.0431 (4)	−0.0013 (3)	0.0186 (3)	−0.0009 (3)
O1	0.0450 (15)	0.0472 (15)	0.0676 (18)	−0.0034 (13)	0.0235 (14)	−0.0127 (15)
N1	0.0379 (16)	0.0435 (17)	0.0401 (16)	0.0031 (14)	0.0177 (14)	0.0026 (14)
N2	0.074 (3)	0.051 (2)	0.121 (4)	−0.002 (2)	0.052 (3)	0.016 (2)
C1	0.0372 (19)	0.052 (2)	0.0388 (19)	−0.0030 (18)	0.0136 (16)	0.0063 (18)
C2	0.049 (2)	0.046 (2)	0.0355 (19)	−0.0027 (18)	0.0141 (17)	0.0030 (17)
C1	0.0372 (19)	0.052 (2)	0.0388 (19)	−0.0030 (18)	0.0136 (16)	0.0063 (18)
C3	0.049 (2)	0.067 (3)	0.036 (2)	−0.005 (2)	0.0092 (17)	−0.002 (2)
C4	0.066 (3)	0.060 (3)	0.041 (2)	−0.007 (2)	0.008 (2)	−0.005 (2)
C5	0.078 (3)	0.047 (2)	0.044 (2)	0.004 (2)	0.023 (2)	−0.0014 (19)
C6	0.053 (2)	0.048 (2)	0.053 (2)	0.008 (2)	0.022 (2)	0.000 (2)
C7	0.049 (2)	0.0403 (18)	0.0385 (19)	0.0011 (18)	0.0173 (17)	0.0059 (17)
C2	0.049 (2)	0.046 (2)	0.0355 (19)	−0.0027 (18)	0.0141 (17)	0.0030 (17)
C1	0.0372 (19)	0.052 (2)	0.0388 (19)	−0.0030 (18)	0.0136 (16)	0.0063 (18)
C8	0.044 (2)	0.045 (2)	0.051 (2)	0.0049 (17)	0.0210 (18)	0.0004 (19)
C9	0.056 (2)	0.0426 (19)	0.048 (2)	−0.0005 (18)	0.0275 (19)	−0.0054 (17)
C10	0.053 (2)	0.055 (3)	0.065 (3)	0.009 (2)	0.035 (2)	0.003 (2)

Geometric parameters (Å, º)

Cu1—O1	1.893 (3)	C5—C6	1.381 (6)
Cu1—O1^i	1.893 (3)	C5—H5	0.9300
Cu1—N1	2.007 (3)	C6—C7	1.406 (6)
Cu1—N1^i	2.007 (3)	C6—H6	0.9300
O1—C7	1.308 (5)	C7—C2	1.406 (6)
N1—C1	1.285 (5)	C2—C1	1.431 (5)
N1—C8	1.472 (5)	C1—H1	0.9300
N2—C10	1.129 (5)	C8—C9	1.523 (6)
C3—C4	1.363 (6)	C8—H8A	0.9700
C3—C2	1.425 (6)	C8—H8B	0.9700
C3—H3	0.9300	C9—C10	1.466 (6)
C4—C5	1.386 (6)	C9—H9A	0.9700
C4—H4	0.9300	C9—H9B	0.9700
O1—Cu1—O1^i	180.00 (18)	O1—C7—C6	118.6 (4)
O1—Cu1—N1	91.80 (12)	O1—C7—C2	123.3 (4)
O1^i—Cu1—N1	88.20 (12)	C6—C7—C2	118.0 (4)
O1—Cu1—N1^i	88.20 (12)	C7—C2—C3	119.4 (4)
O1^i—Cu1—N1^i	91.80 (12)	C7—C2—C1	123.1 (4)
N1—Cu1—N1^i	180.00 (4)	C3—C2—C1	117.4 (4)
C7—O1—Cu1	130.3 (3)	N1—C1—C2	127.1 (4)
C1—N1—C8	115.3 (3)	N1—C1—H1	116.5
C1—N1—Cu1	124.1 (3)	C2—C1—H1	116.5
C8—N1—Cu1	120.5 (2)	N1—C8—C9	111.9 (3)
C4—C3—C2	121.3 (4)	N1—C8—H8A	109.2
C4—C3—H3	119.4	C9—C8—H8A	109.2
C2—C3—H3	119.4	N1—C8—H8B	109.2
C3—C4—C5	119.1 (4)	C9—C8—H8B	109.2
C3—C4—H4	120.5	H8A—C8—H8B	107.9
C5—C4—H4	120.5	C10—C9—C8	112.6 (4)
C6—C5—C4	121.4 (4)	C10—C9—H9A	109.1
C6—C5—H5	119.3	C8—C9—H9A	109.1
C4—C5—H5	119.3	C10—C9—H9B	109.1
C5—C6—C7	120.8 (4)	C8—C9—H9B	109.1
C5—C6—H6	119.6	H9A—C9—H9B	107.8
C7—C6—H6	119.6	N2—C10—C9	178.7 (5)

Symmetry code: (i) −x+1/2, −y+1/2, −z.