A tale of hydrogen abstraction, initially detected via X-ray diffraction

Reaction of a bis-tetra­zinyl pyridine pincer ligand, btzp, with a vanadium(III) reagent gives not a simple adduct but di­chlorido­{3-methyl-6-[6-(6-methyl-1,2,4,5-tetrazin-3-yl-κN²)pyridin-2-yl-κN]-1,4-di­hydro-1,2,4,5-tetra­zin-1-yl-κN¹}oxidovanadium(IV) acetonitrile 2.5-solvate, [V(C₁₁H₁₀N₉)Cl₂O]·2.5CH₃CN, a species which X-ray diffraction reveals to have one H atom added to one of the two tetra­zinyl rings. This H atom was first revealed by a short inter­molecular N...Cl contact in the unit cell and subsequently established, from difference maps, to be associated with a hydrogen bond. One chloride ligand has also been replaced by an oxide ligand in this synthetic reaction. This formula for the complex, [V(Hbtzp)Cl₂O], leaves open the question of both ligand oxidation state and spin state. A computational study of all isomeric locations of the H atom shows the similarity of their energies, which is subject to perturbation by inter­molecular hydrogen bonding found in X-ray work on the solid state. These density functional calculations reveal that the isomer with the H atom located as found in the solid state contains a neutral radical Hbtzp ligand and tetra­valent d¹ V center, but that these two unpaired electrons are more stable as an open-shell singlet and hence anti­ferromagnetically coupled.

S1. Introduction

It is a reliable principle that, as CH groups in benzene are formally replaced by N, the ring as a whole becomes more easily reduced. Ease of reduction increases along the series C₆H₆ < C₅H₅N < C₄H₄N₂ < C₃H₃N₃ < C₂H₂N₄, without regard to which isomeric polyazine is indicated. This can be traced to more N atoms lowering the energy of the ring π* orbital, even if the molecular symmetry decreases, and thus the atomic orbital mixing varies along this series. With this in mind, we set out to explore a new ligand type intended to be redox active, hence to exist in several different oxidation states (Kaim, 2002, 2011).

Tetra­zines (tz; Clavier & Audebert, 2010) are molecules (Scheme 1) that are readily reduced, and the radical monoanion tz^-1 is persistent. This monoanion has been characterized by electron paramagnetic resonance (EPR) and shows hyperfine coupling to all four N atoms (Stone & Maki, 1963). By constructing a ligand carrying two tetra­zines, we double the number of redox states accessible. We chose, for synthetic ease, to link the two tetra­zines via ortho disubstitution on a pyridine, which yields a bis-tetra­zinyl pyridine, btzp (Benson et al., 2013). Btzp has all the advantages of a pincer topology: mandatory mer stereochemistry in a (btzp)ML_n complex and the strong metal-binding ability derived from poly-denticity. The ultimate use of a (btzp)ML_n construct occurs when the metal is also redox active, thus enhancing the electron-storage potential of this species, not only at the pincer ligand but also at the metal. With this in mind, the question becomes when to introduce the redox equivalents – after the synthesis of (btzp)M^q+ or by using a pre-reduced M^(q-n)+, which carries n reducing equivalents prior to pincer-ligand installation. We report here the surprising results of the second approach, which relied heavily on crystallography to reveal the unanti­cipated course of the reaction; this result derives from ligand-centered redox activity. In this regard, it is valuable to recall that tetra­zines, being highly oxidized (i.e. de­hydrogenated compared with hydrazines), are subject to reduction by two equivalents (H₂Tz, Scheme 1); our btzp ligand synthesis proceeds through that di­hydro­tetra­zine stage. However, these reductions are not simply the electron transfer initially envisioned for our metal–ligand assembly, but involve N—H bond formation from a hydrogen source acting as a reducing agent.

In this study, we chose vanadium as our metal since it exists in a broad range of oxidation states from -1 to +5, with all steps in between (hence single-electron redox change is accessible), and because it offers several useful spectroscopic `reporters'. Other groups with similar motivation have reported results on vanadium complexes that contain redox-active ligands that have been shown to be useful as polymerization and di­nitro­gen reduction catalysts (Knijnenburg et al., 2006; Kundu et al., 2013; Milione et al., 2002; Milsmann et al., 2012; Reardon et al., 1999; Vidyaratne et al., 2005). We began with trivalent vanadium, since this can be a one- or even a two-electron redu­ctant. Beginning with VCl<sub>3</sub>(THF)<sub>3</sub>, we envisioned a product [V(btzp)Cl<sub>3</sub>]. Within this molecule we ask, `Where are the electrons?' (Caulton, 2012; Chaudhuri et al., 2001; Lyaskovskyy & de Bruin, 2012; Scarborough et al., 2012; van der Vlugt, 2012), by which we mean: does this molecule persist as V^III and neutral btzp, or is this vanadium reagent a strong enough reducing agent to produce V^IV and the btzp^-1 radical? By a variety of investigative approaches, we found the latter to be the case, and the compound obtained is shown to be dichlorido[3-methyl-6-[6-(6-methyl-2,3-di­hydro-1,2,4,5-tetra­zin-3-yl-κN⁴)pyridin-2-yl-κN]-1,4-di­hydro-1,2,4,5-tetra­zin-1-ido-κN¹]oxidovanadium(V), [V(C₁₁H₁₀N₉)Cl₂O].2.5CH₃CN or [V(Hbtzp)Cl₂O].2.5CH₃CN, (I).

S2. Experimental

S2.1. Synthesis and crystallization

Reaction of equimolar btzp with VCl₃(THF)₃ in aceto­nitrile occurs within minutes, with a color change from pink to dark purple. Full synthetic details are given in the supporting information. Dark-purple crystals of (I) were formed by layering an aceto­nitrile solution of the product with Et₂O and were shown to crystallize with the entire chemical formula as the asymmetric unit.

S2.2. Refinement

Crystal data, data collection and structure refinement details are summarized in Table 1. The H atom participating in classical hydrogen bonding (H₃N) was located in a difference map and refined, including its positional parameters and U_iso. The refined N3—H3N distance is 0.91 (2) Å. All other H atoms were placed in ideal positions, with C—H = 0.95 Å for aromatic (CH) groups or 0.98 Å for methyl (CH₃) groups, and refined using a riding model, with U_iso(H) = 1.2 or 1.5U_eq(C), respectively. The formula of (I) is thus established, purely by crystallography, as [V(Hbtzp)Cl₂O].2.5CH₃CN, where the unligated solvent occupies inter­stices.

S3. Results

S3.1. Crystallographic characterization

Our first X-ray data set showed an o­cta­hedral geometry around V, with a meridional pincer. The first surprise was that the monodentate ligands were not three Cl^- but instead two trans Cl^- and one O. The V/O distance, \sim 1.6 Å, is too short to be a single bond, so we discounted this as being an OH ligand and concluded that a V═O bond had been formed. When we found that refinement with an initial in-house data set gave markedly [~0.17 (1) Å] different V—N distances to the two tetra­zine N atoms, we were both surprised and skeptical. We rationalized this result as an error derived from an incomplete data set due to above-average icing of the crystal during low-temperature data collection, so we remained skeptical of the asymmetric character of the two tetra­zines bonding to V.

When a second, smaller, crystal was grown (using the same crystal-growth method), data were collected with synchrotron radiation and proceeded unexceptionally with regard to data quality. This crystal had the same space group as the first and refined well. It was the same compound as studied previously, again with a VCl₂O subunit. It is of note that it showed (Fig. 1, Table 2) the same unequal V—N(tetra­zine) distances as seen in the data set that we had considered `flawed', so this feature was robust even to imperfect data, and the earlier data were reliable in indicating that there was something exceptional about the product molecule.

With the synchrotron data, following refinement with the pyridyl and methyl H atoms in idealized positions and using a riding model for these H atoms, we saw not only a difference in the two V—N(tetra­zine) distances of 0.23 Å, but also inequalities in the distances within the tetra­zine rings: the two rings do not have the same pattern of C—C and C—N bond lengths. Nevertheless, each nitro­gen-rich ring is essentially planar. Remarkably, analogous C—C and C—N differences were also seen in the earlier `imperfect' data set; again, even these subtle structural features were present in what we had originally considered to be only a `rough' data set.

A key observation with both data sets is a sub-van der Waals contact of 3.1306 (9) Å between atom N3 and a Cl^- ligand on V in a neighboring molecule [N3—H3N···Cl1ⁱ; symmetry code: (i) x + 1/2, -y + 1/2, z - 1/2]. In spite of our expe­cta­tion of the product identity as simply [V(btzp)Cl₃], we considered that this short contact indicated that the ring carried a single H on the N atom involved in this contact. Additionally, the nature of the structure and its symmetry might lend itself to disorder via a 180° rotation around the V—N(pyridyl) vector. The presence of this weak N—HCl hydrogen bond locks the structure in a preferred orientation, thus preventing disorder. Indeed, the H atom appeared clearly and unambiguously in a difference map [Fig. 2 and also in the supporting information (SI); here and in the SI, only the refinement based on the synchrotron data is reported].

With regard to `where the electrons are' (i.e. oxidation state assignment), this opens questions about what charge state the `Hbtzp' ligand possesses. If neutral, this Hbtzp is a radical. Within the singly hydrogenated ring there are two short C—N distances (cf. H₂Tz in Scheme 1) consistent with double bonds, and the remaining distances are all long compared with distances in the second tetra­zine ring in the Hbtzp pincer. In addition, this `modified' ring shows the shorter of the two distances to V consistent with amide character, and the amide lone pair at that N atom presumably π donates to the otherwise unsaturated V, which creates multiple V—N bond character. In sum, we were led by the data to believe that the hydrogenated ring should be viewed as a two-electron reduced species, so the ligand charge state is Hbtzp^-1. With the assumption that the oxo ligand is conventional, hence O^2-, this yields a V oxidation state of +5. Confident in these conclusions, we sought nevertheless to test some of them with density functional theory (DFT) calculations to further extend our understanding. Such calculations not only offer the possibility of establishing the ligand charge state through analysis of bond orders and orbital occupations, but also evaluate alternative charge states such as a neutral Hbtzp radical and V^IV. If Hbtzp^-1/V^V is favored, we expect a singlet with all orbitals doubly occupied. If an Hbtzp radical and V^IV is favored, anti­ferro- or ferromagnetic coupling to afford either a singlet or a triplet state, respectively, is possible.

S3.2. Computational inquiry

To better understand the electronic and geometric structure of [V(Hbtzp)Cl₂O] as identified by crystallography, we addressed the following questions with DFT: (i) to which N atom does a single H atom prefer to bind in this vanadium complex; and (ii) what is the electron distribution in this species, thus what is the charge on Hbtzp?

To locate the thermodynamically preferred position for H-atom binding to the Hbtzp ligand in [V(Hbtzp)Cl₂O], we considered each of the three tetra­zine N atoms (N2, N3, N4; Fig. 1) not bound to the V center. For each structure, we considered both singlet and triplet spin states, i.e. six species. To simplify the discussions below, we will refer to these three isomers as 2, 3 and 4 (referring to H bound to N2, N3 or N4, respectively), with a subscript of either S or T to indicate singlet or triplet, respectively. The results are summarized in Table 3.

Isomers 2 and 3 were found to be most stable, with 4 higher in energy by ~7 kcal mol^-1 (1 kcal mol^-1 = 4.184 kJ mol^-1). Thus, we focused our attention on 2 and 3. For both isomers, the singlet is favored over the triplet by 1–2 kcal mol^-1, with 2 favored over 3, but by an energy difference that is within the expected error of our methodology. This energetic ordering may also be influenced by the fact that our model optimizes the structures without solvation effects or inter­molecular inter­actions (which are present in the solid state). We suspected there was artificial stabilization of the N2—H site due to its proximity to the anionic oxo ligand. Indeed, there is a short H···O contact (2.23 Å) as Fig. 3 shows, and hence an intra­molecular hydrogen bond. Further support for this is the fact that the N(pyridyl)—V—O angle is significantly less than 180° [N5—V1—O1 = 176.38 (4)°; Table 2?], which represents distortion of the o­cta­hedral angles in order to facilitate this hydrogen bond.

This inter­action biases the energy landscape for the most favored N atom for the added H atom. The crystallographic work shows inter­molecular N3—H···Cl hydrogen bonding, which is not modeled in our single-molecule calculation. We next compared the intra-ring bond lengths from isomers 2_S and 3_S, and from the experimental X-ray structure. The N—N and N—C bond lengths in the hydrogenated tetra­zine arm of the btzp ligand are most diagnostic, and a comparison of these bond lengths is shown in Fig. 4, and a similar comparison of isomers 2_S, 3_S and 4_T may be seen in Fig. SI-1. The N2—C2 X-ray bond length of 1.3097 (12) Å is closest to that of the shorter 1.304 Å in 3_S, rather than the longer 1.360 Å in 2_S. Similarly, the C2—N3 X-ray bond length of 1.3499 (13) Å is closest to that of the longer 1.365 Å in 3_S, rather than the shorter 1.312 Å in 2_S. Isomer 4_T is the furthest from agreement with experiment. Therefore, the 3_S structure gives the best agreement with experiment.

Given the small energy difference between 3_S and 3_T, the singlet and triplet states, we wanted to use a structural comparison to test our DFT energy-based spin-state assignment further. Fig. SI-3 shows that the structures of both spin states agree well with experiment. The largest difference is in the N1—C3 and C3—N4 bond lengths, where 3_S matches better with experiment by ~0.01 Å in both bonds. A more complete comparison of bond lengths between 3_S and the X-ray structure may be found in Fig. SI-2 and Table SI-1.

Finally, we sought to use the V1—N1 distance to discriminate among NH regioisomers. Table 3 shows the V1—N1 comparison for all six calculated isomers. Although all calculated V1—N1 bond lengths are longer than experiment, the short V1—N1 distance in 3_S agrees well with experiment.

The final question addressed was that of the redox/spin states of the ligand and metal within [V(Hbtzp)Cl₂O], now that we have identified 3_S as the isomer most relevant to the X-ray structure. The singlet was found to have an open-shell wavefunction: the restricted calculation with α and β spin states confined to identical spatial orbitals was found to have a wavefunction instability and to be higher in energy than an `unrestricted' wavefunction where different spatial orbitals are allowed for the α (spin up) and β (spin down) electrons. We evaluated spin densities and the corresponding orbitals (Fig. 5) of the open-shell singlet wavefunction to help with the redox/spin state assignments (Neese, 2004). This revealed the spin density in the singlet to be on both the metal center and the ligand, but with opposite spins: the spin density at V is α (white in Fig. 5) while the ligand is β (blue). It is also noteworthy that there is α spin density at O. However, a Mulliken analysis shows this spin to be only ~0.2 and none of the singly occupied orbitals (see below) is based on the oxo ligand. We thus conclude that this spin density is due to spin polarization rather than oxyl character.

The finding of spin density at only one tetra­zine arm suggests that the added single H atom in this product injects an electron into the π* orbital of tetra­zine upon binding to the ligand. For further clarification, we calculated the corresponding orbitals to determine which orbitals are singly occupied in this species. These results (Fig. 5, bottom) show a β spin orbital on the metal center and an α spin orbital on the ligand with a small overlap of 0.36. A small overlap means that these two orbitals are spatially distinct, as is visually evident from Fig. 5. It is also clear from the SOMO (singly occupied molecular orbital) of the ligand that this is a π*_NN orbital. Because there is one singly occupied V d orbital, this complex is best described as a V^IV center anti­ferromagnetically coupled to a tetra­zinyl radical. Contrary to our working hypothesis at the end of the crystallographic work, the Hbtzp ligand is an uncharged radical in this species. Note that the metal SOMO is also of π symmetry with respect to the pincer plane, yet this does not mix significantly with the ligand SOMO (Sαβ = 0.36); spatial separation of opposite spins is found to give a more stable electronic structure. It is worth noting that the analogous triplet species, 3_T, has essentially the same metal and ligand redox states but with the two electrons ferromagnetically coupled (Fig. SI-4).

S4. Discussion and conclusion

The synthesis of (I) is entirely reproducible. The synthetic origins of the H and O atoms remain to be determined. Together, adventitious H₂O might be the answer, but one must not neglect the possibility that the H might come from the CH₃CN solvent. The identified product is oxidized at V and reduced at the btzp ligand, so the reducing power initially present in VCl₃(THF)₃ has certainly been used but, given the conclusion of a V^IV product, it has functioned as only a one-electron redu­ctant.

The computational results have discerned the relative stabilities of regioisomeric hydrogenated tetra­zine rings in isolated complexes, but cannot capture the solid-state reality of inter­molecular hydrogen bonding to Cl^- in neighboring VCl units. Nevertheless, the energetic similarity of the various isomers established here already serves as a reminder that these tautomers are all viable participants in any reaction chemistry, and that such reactivity may exhibit effects of environmental influence (e.g. solvent, concentration, inter­molecular hydrogen bonding). Moreover, solution-phase results may include a mixture of these tautomers, and further studies are actively being pursued experimentally and computationally to test this hypothesis.

This case study proves that structural `anomalies', even when first seen in flawed X-ray diffraction data sets obtained under non-ideal (but real world) conditions, carry important messages to anyone willing to listen. Outliers are sometimes truly `noise', but in select cases can be the invitation to discovery.

Related literature

For related literature, see: Benson et al. (2013); Caulton (2012); Chaudhuri et al. (2001); Clavier & Audebert (2010); Kaim (2002, 2011); Knijnenburg et al. (2006); Kundu et al. (2013); Lyaskovskyy & de Bruin (2012); Milione et al. (2002); Milsmann et al. (2012); Neese (2004); Reardon et al. (1999); Scarborough et al. (2012); Stone & Maki (1963); Vidyaratne et al. (2005); van der Vlugt (2012).

Computing details

Data collection: APEX2 (Bruker, 2005); cell refinement: APEX2 (Bruker, 2005); data reduction: APEX2 (Bruker, 2005); program(s) used to solve structure: SIR2004 (Burla et al., 2005); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: SHELXTL (Sheldrick, 2008); software used to prepare material for publication: SHELXTL (Sheldrick, 2008).

Figures

	Fig. 1. The molecular structure of [V(Hbtzp)OCl2]·2.5CH3CN, (I), showing selected atom labeling. Displacement ellipsoids are drawn at the 50% probability level. Only the H atom on N3 is shown.
	Fig. 2. Residual electron density map for the complex [V(Hbtzp)OCl2], showing electron density (green contours) due to H near N3 and trans to V. The electron density was assigned as H3N and refined as an H atom. The map also shows sigma bonding density on most of the bonds in this ring, as well as elsewhere in this plane.
	Fig. 3. Top-down views of the optimized structures, showing the hydrogen bonding (dotted line) in isomer 2S (left) versus 3S (right). The distance d is 2.232 Å. Visualized using CYLview (Legault, 2009).
	Fig. 4. Comparison of [V(Hbtzp)Cl2O] bond lengths (Å) within the hydrogenated tetrazine arm for the two lowest energy singlets, 2S and 3S, with data from the X-ray experiment.
	Fig. 5. Isosurface plots for the spin density (top, 0.002 au) and the corresponding singly occupied orbitals (bottom, 0.05 au) of isomer 3S. The overlap between the α (left) and β (right) orbitals is 0.36.

Dichlorido[3-methyl-6-[6-(6-methyl-2,3-dihydro-1,2,4,5-tetrazin-3-yl-κN⁴)pyridin-2-yl-κN]-1,4-dihydro-1,2,4,5-tetrazin-1-yl-κN¹]oxidovanadium(IV)

Crystal data

[V(C11H10N9)Cl2O]·2.5C2H3N	F(000) = 1036
Mr = 508.76	Dx = 1.482 Mg m^−3
Monoclinic, P21/n	Synchrotron radiation, λ = 0.41328 Å
Hall symbol: -P 2yn	Cell parameters from 9905 reflections
a = 8.1651 (5) Å	θ = 2.3–24.4°
b = 23.7109 (14) Å	µ = 0.37 mm^−1
c = 11.8159 (7) Å	T = 100 K
β = 94.801 (1)°	Acicular, violet
V = 2279.6 (2) Å^3	0.01 × 0.01 × 0.01 mm
Z = 4

Data collection

Bruker Platform goniometer diffractometer	14694 independent reflections
Radiation source: synchrotron	12834 reflections with I > 2σ(I)
Silicon 1 1 1 and 3 1 1 monochromator	Rint = 0.043
Detector resolution: 83.33 pixels mm^-1	θmax = 22.5°, θmin = 1.0°
ω and φ scans	h = −10→15
Absorption correction: multi-scan (SADABS; Sheldrick, 2003)	k = −43→36
Tmin = 0.996, Tmax = 0.997	l = −21→21
59575 measured reflections

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)] = 0.041	Hydrogen site location: inferred from neighbouring sites
wR(F^2) = 0.114	H atoms treated by a mixture of independent and constrained refinement
S = 1.06	w = 1/[σ^2(Fo^2) + (0.0323P)^2 + 1.6092P] where P = (Fo^2 + 2Fc^2)/3
14694 reflections	(Δ/σ)max = 0.002
318 parameters	Δρmax = 0.61 e Å^−3
25 restraints	Δρmin = −0.54 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	Occ. (<1)
V1	0.131905 (17)	0.356982 (6)	0.489027 (13)	0.01227 (3)
Cl1	0.23950 (3)	0.292340 (10)	0.63063 (2)	0.01909 (5)
Cl2	0.06759 (3)	0.440656 (10)	0.38539 (2)	0.01689 (4)
O1	−0.04438 (9)	0.32779 (3)	0.46638 (7)	0.01852 (12)
N1	0.25390 (9)	0.32034 (3)	0.37180 (7)	0.01453 (11)
N2	0.18319 (10)	0.27892 (4)	0.30187 (8)	0.01807 (13)
N3	0.44034 (11)	0.26508 (4)	0.23370 (8)	0.01971 (14)
H3N	0.514 (3)	0.2443 (10)	0.198 (2)	0.036 (6)*
N4	0.51447 (10)	0.30645 (4)	0.29828 (8)	0.01854 (13)
N5	0.37828 (9)	0.39357 (3)	0.51142 (7)	0.01392 (11)
N6	0.14268 (10)	0.41415 (3)	0.63545 (7)	0.01432 (11)
N7	0.01847 (10)	0.41997 (4)	0.69888 (7)	0.01620 (12)
N8	0.17504 (12)	0.48807 (4)	0.80923 (8)	0.02243 (16)
N9	0.29894 (12)	0.48153 (4)	0.74574 (8)	0.02146 (15)
C1	0.21224 (15)	0.20607 (5)	0.16185 (11)	0.0270 (2)
H1B	0.0920	0.2063	0.1596	0.040*
H1C	0.2542	0.1698	0.1918	0.040*
H1D	0.2460	0.2115	0.0850	0.040*
C2	0.27955 (12)	0.25237 (4)	0.23641 (9)	0.01813 (15)
C3	0.41806 (10)	0.33233 (4)	0.36352 (8)	0.01480 (13)
C4	0.48566 (10)	0.37705 (4)	0.43989 (8)	0.01440 (12)
C5	0.64294 (11)	0.40034 (4)	0.44307 (9)	0.01724 (14)
H5	0.7187	0.3889	0.3907	0.021*
C6	0.68481 (11)	0.44077 (4)	0.52552 (9)	0.01881 (15)
H6	0.7910	0.4574	0.5299	0.023*
C7	0.57270 (11)	0.45711 (4)	0.60183 (9)	0.01779 (14)
H7	0.6011	0.4842	0.6593	0.021*
C8	0.41800 (11)	0.43241 (4)	0.59100 (8)	0.01468 (12)
C9	0.28135 (11)	0.44380 (4)	0.66128 (8)	0.01527 (13)
C10	0.03889 (13)	0.45659 (4)	0.78558 (8)	0.01833 (15)
C11	−0.09928 (16)	0.46450 (6)	0.85812 (10)	0.0258 (2)
H11A	−0.1896	0.4391	0.8327	0.039*
H11B	−0.1377	0.5036	0.8523	0.039*
H11C	−0.0618	0.4560	0.9373	0.039*
N11	0.5298 (3)	0.38639 (9)	0.84407 (16)	0.0527 (4)
C14	0.4669 (2)	0.34802 (7)	0.88029 (14)	0.0365 (3)
C15	0.3865 (3)	0.29962 (7)	0.92437 (17)	0.0504 (5)
H15A	0.2737	0.3096	0.9388	0.076*
H15B	0.4466	0.2874	0.9954	0.076*
H15C	0.3845	0.2689	0.8689	0.076*
N12	0.9574 (4)	0.31986 (15)	−0.0534 (3)	0.0702 (8)	0.795 (3)
C16	0.9213 (3)	0.33514 (9)	0.0323 (2)	0.0392 (4)	0.795 (3)
C17	0.8700 (6)	0.3543 (2)	0.1415 (3)	0.0454 (6)	0.795 (3)
H17A	0.7777	0.3806	0.1286	0.068*	0.795 (3)
H17B	0.9620	0.3734	0.1842	0.068*	0.795 (3)
H17C	0.8358	0.3218	0.1849	0.068*	0.795 (3)
N12D	1.1311 (15)	0.3776 (6)	0.0697 (10)	0.0702 (8)	0.205 (3)
C16D	1.0138 (10)	0.3681 (4)	0.1122 (8)	0.0392 (4)	0.205 (3)
C17D	0.862 (2)	0.3578 (10)	0.1640 (16)	0.0454 (6)	0.205 (3)
H17D	0.8226	0.3931	0.1954	0.068*	0.205 (3)
H17E	0.8808	0.3300	0.2250	0.068*	0.205 (3)
H17F	0.7787	0.3433	0.1065	0.068*	0.205 (3)
N13	0.3344 (7)	0.3972 (2)	0.1260 (4)	0.0685 (13)	0.50
C18	0.4016 (5)	0.43273 (17)	0.0847 (3)	0.0450 (8)	0.50
C19	0.4858 (5)	0.47791 (17)	0.0296 (3)	0.0449 (8)	0.50
H19A	0.4401	0.4815	−0.0495	0.067*	0.50
H19B	0.6033	0.4692	0.0314	0.067*	0.50
H19C	0.4707	0.5135	0.0697	0.067*	0.50

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
V1	0.00857 (5)	0.01259 (6)	0.01593 (6)	−0.00130 (4)	0.00270 (4)	−0.00209 (4)
Cl1	0.01342 (8)	0.01821 (9)	0.02603 (10)	0.00005 (6)	0.00402 (7)	0.00681 (7)
Cl2	0.01609 (8)	0.01563 (8)	0.01860 (9)	0.00017 (6)	−0.00066 (6)	0.00035 (6)
O1	0.0108 (2)	0.0187 (3)	0.0263 (3)	−0.0020 (2)	0.0031 (2)	−0.0032 (2)
N1	0.0103 (2)	0.0151 (3)	0.0186 (3)	−0.0012 (2)	0.0037 (2)	−0.0029 (2)
N2	0.0135 (3)	0.0190 (3)	0.0225 (3)	−0.0023 (2)	0.0061 (2)	−0.0076 (3)
N3	0.0145 (3)	0.0216 (3)	0.0241 (4)	0.0005 (3)	0.0080 (3)	−0.0054 (3)
N4	0.0125 (3)	0.0201 (3)	0.0238 (3)	0.0000 (2)	0.0064 (2)	−0.0025 (3)
N5	0.0105 (2)	0.0140 (3)	0.0172 (3)	−0.0013 (2)	0.0010 (2)	0.0004 (2)
N6	0.0131 (3)	0.0148 (3)	0.0151 (3)	−0.0006 (2)	0.0009 (2)	−0.0007 (2)
N7	0.0162 (3)	0.0177 (3)	0.0149 (3)	−0.0002 (2)	0.0028 (2)	−0.0017 (2)
N8	0.0220 (4)	0.0251 (4)	0.0200 (3)	−0.0011 (3)	0.0003 (3)	−0.0073 (3)
N9	0.0196 (3)	0.0230 (4)	0.0213 (3)	−0.0022 (3)	−0.0009 (3)	−0.0072 (3)
C1	0.0223 (4)	0.0283 (5)	0.0318 (5)	−0.0053 (4)	0.0110 (4)	−0.0152 (4)
C2	0.0153 (3)	0.0188 (3)	0.0212 (4)	−0.0009 (3)	0.0068 (3)	−0.0045 (3)
C3	0.0109 (3)	0.0155 (3)	0.0183 (3)	0.0001 (2)	0.0034 (2)	0.0002 (2)
C4	0.0094 (3)	0.0149 (3)	0.0189 (3)	−0.0007 (2)	0.0019 (2)	0.0018 (2)
C5	0.0092 (3)	0.0179 (3)	0.0247 (4)	−0.0007 (2)	0.0017 (3)	0.0033 (3)
C6	0.0102 (3)	0.0186 (3)	0.0271 (4)	−0.0027 (3)	−0.0016 (3)	0.0040 (3)
C7	0.0132 (3)	0.0173 (3)	0.0220 (4)	−0.0031 (3)	−0.0034 (3)	0.0011 (3)
C8	0.0120 (3)	0.0148 (3)	0.0169 (3)	−0.0019 (2)	−0.0010 (2)	0.0007 (2)
C9	0.0142 (3)	0.0153 (3)	0.0160 (3)	−0.0016 (2)	−0.0007 (2)	−0.0009 (2)
C10	0.0208 (4)	0.0196 (4)	0.0146 (3)	0.0005 (3)	0.0016 (3)	−0.0021 (3)
C11	0.0285 (5)	0.0296 (5)	0.0204 (4)	−0.0016 (4)	0.0088 (4)	−0.0063 (4)
N11	0.0573 (11)	0.0572 (10)	0.0438 (8)	−0.0118 (8)	0.0051 (7)	0.0126 (7)
C14	0.0426 (8)	0.0352 (7)	0.0323 (6)	0.0085 (6)	0.0076 (5)	0.0085 (5)
C15	0.0851 (16)	0.0294 (7)	0.0401 (8)	0.0058 (8)	0.0249 (9)	0.0098 (6)
N12	0.0735 (19)	0.0784 (19)	0.0622 (16)	0.0048 (15)	0.0269 (14)	−0.0096 (14)
C16	0.0388 (9)	0.0351 (8)	0.0435 (10)	0.0000 (7)	0.0020 (7)	−0.0020 (7)
C17	0.0607 (13)	0.0318 (10)	0.0425 (19)	0.0035 (9)	−0.0029 (12)	−0.0108 (12)
N12D	0.0735 (19)	0.0784 (19)	0.0622 (16)	0.0048 (15)	0.0269 (14)	−0.0096 (14)
C16D	0.0388 (9)	0.0351 (8)	0.0435 (10)	0.0000 (7)	0.0020 (7)	−0.0020 (7)
C17D	0.0607 (13)	0.0318 (10)	0.0425 (19)	0.0035 (9)	−0.0029 (12)	−0.0108 (12)
N13	0.083 (3)	0.065 (3)	0.062 (2)	0.024 (2)	0.033 (2)	0.007 (2)
C18	0.0490 (19)	0.0524 (19)	0.0337 (14)	0.0207 (16)	0.0049 (13)	−0.0082 (13)
C19	0.0460 (17)	0.0504 (18)	0.0353 (14)	0.0033 (14)	−0.0137 (13)	−0.0165 (13)

Geometric parameters (Å, º)

V1—O1	1.5994 (7)	C6—C7	1.3926 (15)
V1—N1	1.9740 (8)	C6—H6	0.9500
V1—N5	2.1864 (8)	C7—C8	1.3885 (13)
V1—N6	2.1939 (8)	C7—H7	0.9500
V1—Cl2	2.3673 (3)	C8—C9	1.4706 (13)
V1—Cl1	2.3821 (3)	C10—C11	1.4849 (15)
N1—N2	1.3786 (11)	C11—H11A	0.9800
N1—C3	1.3816 (11)	C11—H11B	0.9800
N2—C2	1.3097 (12)	C11—H11C	0.9800
N3—C2	1.3499 (13)	N11—C14	1.145 (2)
N3—N4	1.3542 (13)	C14—C15	1.441 (3)
N3—H3N	0.91 (2)	C15—H15A	0.9800
N4—C3	1.3008 (12)	C15—H15B	0.9800
N5—C4	1.3269 (12)	C15—H15C	0.9800
N5—C8	1.3366 (12)	N12—C16	1.137 (4)
N6—N7	1.3178 (12)	C16—C17	1.462 (4)
N6—C9	1.3459 (12)	C17—H17A	0.9800
N7—C10	1.3426 (13)	C17—H17B	0.9800
N8—N9	1.3180 (14)	C17—H17C	0.9800
N8—C10	1.3487 (15)	N12D—C16D	1.141 (11)
N9—C9	1.3391 (13)	C16D—C17D	1.451 (13)
C1—C2	1.4840 (15)	C17D—H17D	0.9800
C1—H1B	0.9800	C17D—H17E	0.9800
C1—H1C	0.9800	C17D—H17F	0.9800
C1—H1D	0.9800	N13—C18	1.138 (6)
C3—C4	1.4696 (13)	C18—C19	1.456 (6)
C4—C5	1.3955 (12)	C19—H19A	0.9800
C5—C6	1.3889 (15)	C19—H19B	0.9800
C5—H5	0.9500	C19—H19C	0.9800
O1—V1—N1	101.07 (4)	N5—C4—C5	121.44 (9)
O1—V1—N5	176.38 (4)	N5—C4—C3	111.71 (7)
N1—V1—N5	75.31 (3)	C5—C4—C3	126.84 (8)
O1—V1—N6	112.05 (4)	C6—C5—C4	117.64 (9)
N1—V1—N6	146.72 (3)	C6—C5—H5	121.2
N5—V1—N6	71.56 (3)	C4—C5—H5	121.2
O1—V1—Cl2	96.81 (3)	C5—C6—C7	120.64 (8)
N1—V1—Cl2	96.26 (3)	C5—C6—H6	119.7
N5—V1—Cl2	83.80 (2)	C7—C6—H6	119.7
N6—V1—Cl2	83.26 (2)	C8—C7—C6	117.72 (9)
O1—V1—Cl1	96.48 (3)	C8—C7—H7	121.1
N1—V1—Cl1	91.81 (3)	C6—C7—H7	121.1
N5—V1—Cl1	83.62 (2)	N5—C8—C7	121.39 (9)
N6—V1—Cl1	81.69 (2)	N5—C8—C9	111.89 (7)
Cl2—V1—Cl1	162.824 (11)	C7—C8—C9	126.72 (9)
N2—N1—C3	117.66 (7)	N9—C9—N6	123.63 (9)
N2—N1—V1	121.57 (6)	N9—C9—C8	120.29 (8)
C3—N1—V1	120.57 (6)	N6—C9—C8	116.08 (8)
C2—N2—N1	117.01 (8)	N7—C10—N8	124.31 (9)
C2—N3—N4	122.55 (8)	N7—C10—C11	118.05 (10)
C2—N3—H3N	125.0 (15)	N8—C10—C11	117.61 (9)
N4—N3—H3N	111.8 (15)	C10—C11—H11A	109.5
C3—N4—N3	114.30 (8)	C10—C11—H11B	109.5
C4—N5—C8	121.14 (8)	H11A—C11—H11B	109.5
C4—N5—V1	117.13 (6)	C10—C11—H11C	109.5
C8—N5—V1	121.68 (6)	H11A—C11—H11C	109.5
N7—N6—C9	119.29 (8)	H11B—C11—H11C	109.5
N7—N6—V1	122.08 (6)	N11—C14—C15	179.2 (2)
C9—N6—V1	118.62 (6)	C14—C15—H15A	109.5
N6—N7—C10	116.69 (8)	C14—C15—H15B	109.5
N9—N8—C10	118.40 (9)	H15A—C15—H15B	109.5
N8—N9—C9	117.59 (9)	C14—C15—H15C	109.5
C2—C1—H1B	109.5	H15A—C15—H15C	109.5
C2—C1—H1C	109.5	H15B—C15—H15C	109.5
H1B—C1—H1C	109.5	N12—C16—C17	178.4 (3)
C2—C1—H1D	109.5	N12D—C16D—C17D	177.8 (15)
H1B—C1—H1D	109.5	C16D—C17D—H17D	109.5
H1C—C1—H1D	109.5	C16D—C17D—H17E	109.5
N2—C2—N3	122.70 (9)	H17D—C17D—H17E	109.5
N2—C2—C1	119.68 (9)	C16D—C17D—H17F	109.5
N3—C2—C1	117.62 (9)	H17D—C17D—H17F	109.5
N4—C3—N1	125.71 (9)	H17E—C17D—H17F	109.5
N4—C3—C4	119.30 (8)	N13—C18—C19	178.9 (5)
N1—C3—C4	114.95 (8)
O1—V1—N1—N2	−3.09 (8)	N3—N4—C3—N1	0.93 (15)
N5—V1—N1—N2	176.79 (8)	N3—N4—C3—C4	178.36 (9)
N6—V1—N1—N2	171.44 (7)	N2—N1—C3—N4	−2.78 (15)
Cl2—V1—N1—N2	−101.31 (7)	V1—N1—C3—N4	172.14 (8)
Cl1—V1—N1—N2	93.87 (7)	N2—N1—C3—C4	179.69 (8)
O1—V1—N1—C3	−177.82 (7)	V1—N1—C3—C4	−5.39 (11)
N5—V1—N1—C3	2.07 (7)	C8—N5—C4—C5	−1.42 (14)
N6—V1—N1—C3	−3.28 (11)	V1—N5—C4—C5	176.28 (7)
Cl2—V1—N1—C3	83.96 (7)	C8—N5—C4—C3	177.37 (8)
Cl1—V1—N1—C3	−80.85 (7)	V1—N5—C4—C3	−4.93 (10)
C3—N1—N2—C2	2.96 (14)	N4—C3—C4—N5	−171.20 (9)
V1—N1—N2—C2	−171.91 (8)	N1—C3—C4—N5	6.50 (11)
C2—N3—N4—C3	0.62 (15)	N4—C3—C4—C5	7.51 (15)
O1—V1—N5—C4	3.6 (6)	N1—C3—C4—C5	−174.79 (9)
N6—V1—N5—C4	178.77 (7)	N5—C4—C5—C6	1.30 (14)
Cl2—V1—N5—C4	−96.29 (7)	C3—C4—C5—C6	−177.30 (9)
Cl1—V1—N5—C4	95.43 (7)	C4—C5—C6—C7	0.01 (14)
N1—V1—N5—C8	179.55 (8)	C5—C6—C7—C8	−1.15 (14)
N6—V1—N5—C8	−3.54 (7)	C4—N5—C8—C7	0.19 (14)
Cl2—V1—N5—C8	81.40 (7)	V1—N5—C8—C7	−177.41 (7)
Cl1—V1—N5—C8	−86.88 (7)	C4—N5—C8—C9	−179.39 (8)
O1—V1—N6—N7	3.62 (8)	V1—N5—C8—C9	3.01 (10)
N1—V1—N6—N7	−170.59 (7)	C6—C7—C8—N5	1.08 (14)
N5—V1—N6—N7	−176.05 (8)	C6—C7—C8—C9	−179.40 (9)
Cl2—V1—N6—N7	98.26 (7)	N8—N9—C9—N6	2.69 (16)
Cl1—V1—N6—N7	−90.05 (7)	N8—N9—C9—C8	−176.90 (9)
O1—V1—N6—C9	−176.85 (7)	N7—N6—C9—N9	−3.29 (14)
N1—V1—N6—C9	8.94 (10)	V1—N6—C9—N9	177.17 (8)
N5—V1—N6—C9	3.48 (7)	N7—N6—C9—C8	176.31 (8)
Cl2—V1—N6—C9	−82.21 (7)	V1—N6—C9—C8	−3.23 (11)
Cl1—V1—N6—C9	89.48 (7)	N5—C8—C9—N9	179.82 (9)
C9—N6—N7—C10	1.06 (13)	C7—C8—C9—N9	0.26 (15)
V1—N6—N7—C10	−179.41 (7)	N5—C8—C9—N6	0.21 (12)
C10—N8—N9—C9	0.02 (16)	C7—C8—C9—N6	−179.35 (9)
N1—N2—C2—N3	−1.64 (16)	N6—N7—C10—N8	1.61 (15)
N1—N2—C2—C1	178.02 (10)	N6—N7—C10—C11	179.49 (10)
N4—N3—C2—N2	−0.22 (17)	N9—N8—C10—N7	−2.19 (17)
N4—N3—C2—C1	−179.89 (11)	N9—N8—C10—C11	179.93 (11)

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
N3—H3N···Cl1^i	0.91 (2)	2.24 (2)	3.1306 (9)	166 (2)

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