Supra­molecular study, Hirshfeld analysis and theoretical study of 6-meth­oxy­quinoline N-oxide dihydrate

In the crystal structure of 6-meth­oxy­quinoline N-oxide di­hydrate, C₁₀H₉NO₂·2H₂O, (I), the presence of two-dimensional water networks is analysed. The water mol­ecules form unusual water channels, as well as two inter­secting mutually perpendicular columns. In one of these channels, the O atom of the N-oxide group acts as a bridge between the water mol­ecules. The other channel is formed exclusively by water mol­ecules. Confirmation of the mol­ecular packing was performed through the analysis of Hirshfeld surfaces, and (I) is compared with other similar iso­quinoline systems. Calculations of bond lengths and angles by the Hartree-Fock method or by density functional theory B3LYP, both with 6-311++G(d,p) basis sets, are reported, together with the results of additional IR, UV-Vis and theoretical studies.

S1. Comment

Aromatic nitrogen heterocycles are widely used as, amongst other things, versatile chelating agents [see, for example, Katakura & Koide (2006)], precursors to pesticides (Kaiser et al., 1996) and bio-active materials (Polshettiwar & Varma, 2008). It is thought that the presence of a quinoline ring is the key factor responsible for imparting a wide range of medicinal properties (Somvanshi et al., 2008), reflected in the behaviour of quinoline derivatives as antibacterial agents (Towers et al., 1981), as antifungal agents (Biavatti et al., 2002), as cytotoxic compounds (Sui et al., 1998) and as drugs (Campbell et al., 1988). The formation of such compounds with N-oxide groups has also been driven by their important pharmacological applications (Ballabio et al., 1992) and has additionally attracted the interest of those developing the field of crystal engineering and the science of organic materials (Desiraju, 1989).

In hydrated crystal systems it is interesting to analyse the presence and role of water molecules in specific structures. The formation of one- or two-dimensional water chains in a crystalline structure can define the behaviour and properties of that system (Cukierman, 2000). Indeed, a large number of biological processes appear to depend on the behaviour of these water chains (Jude et al., 2002). For example, proton translocation processes through membranes are assisted by chains of water molecules functioning as `proton wires' (Tieleman et al., 2001).

These two themes of N-oxide-bearing quinolines and hydrated solid-state structures have been combined by examining the structure of 6-methoxyquinoline N-oxide, MQNO, as its dihydrate, (I). This is part of our ongoing investigation of the structural properties of the isoquinoline matrix and extends earlier work from our research group that reported the structure of the biomolecule 2-amino-3-(N-oxypyridin-4-ylsulfanyl)propionic acid in its dihydrate phase, and which showed that the molecules of the compound are stabilized by the formation of one-dimensional water chains (Moreno-Fuquen et al., 2010). Herein, we describe the crystal structure of (I) and provide spectroscopic (IR, UV–Vis) analysis. Additional theoretical studies and an analysis of Hirshfeld surfaces were also carried out. It is important to consider the presence of water molecules and their interactions within supramolecular synthons through hydrogen bonds and other intermolecular interaction patterns (Desiraju, 1996). Thus, this last analysis was performed to confirm the molecular packing of the system, by examining the behaviour of intermolecular interactions on these surfaces (McKinnon et al., 2004).

Some of the derivatives of the isoquinoline matrix, the 6-methoxyquinoline N-oxide-hydroquinone 2:1 co-crystal (MQNOHQ; Moreno-Fuquen et al., 2007), 2,4-dichloro-6-methoxyquinoline (Subashini et al., 2009) and other 6-methoxyquinoline derivative structures (Chambers et al., 2004), are available as reference systems with which to compare the structural characteristics of (I). The molecular structure of (I) is shown in Fig. 1. Coplanarity between the quinoline ring and the methoxy C5—O2—C10 group is observed in (I). This same coplanarity is observed for 2,4-dichloro-6-methoxyquinoline, and only a small deviation from coplanarity is seen in MQNOHQ [dihedral angle = 3.1 (1)°]. Other bond lengths and angles of the isoquinoline ring of (I) agree with the literature values (Allen et al., 1987).

The inclusion of water molecules in the quinolinic structure can potentially result in the formation of intermolecular hydrogen bonds, which allows a more stable crystalline system. In the anhydrous 6-methoxyquinoline N-oxide system, wherein intermolecular interactions should be relatively weak, this behaviour should not be observed. The presence of water molecules in the quinolinic structure of (I) allows the formation of relatively strong O—H···O hydrogen bonds, thereby stabilizing the crystal system (Table 1; Nardelli, 1995).

The title achiral molecule crystallizes in the monoclinic Sohncke space group P2₁, possibly motivated by a chiral environment imposed by the lattice (Sakamoto, 2004) or by the formation of chains of hydrogen-bonding character (Leiserowitz & Weinstein, 1975). Compound (I) should display relatively strong hydrogen bonds between the methoxyquinoline ring and the water molecules. Indeed, (I) exhibits O2W—H4W···O1, O1W—H2W···O1, O1W—H1W···O2W and O2W—H3W···O1 hydrogen bonds, with O···O distances of 2.721 (2), 2.774 (2), 2.817 (2) and 2.828 (2) Å, respectively. Other weak C3—H3···O2W and C2—H2···O2 interactions complement the strong hydrogen bonds.

A supramolecular analysis of (I) reveals that, in a first substructure, the presence of channels formed by water molecules and atom O1 of the N—O group is observed. Indeed, atom O1 acts simultaneously as a hydrogen-bond acceptor from atom O1W in the molecule at (-x + 2, y - 1/2, -z + 1) and from atom O2W in the molecule at (x, y, z + 1). Atom O2W is linked to another O1W atom and this latter atom is linked to the next O1 atom of the N-oxide group, and so on. Thus, these channels are characterized by the presence of three molecules of water followed by an O atom of the N-oxide group, and this latter atom acts as a bridge between the water molecules running along [001] (Fig. 2). Also in Fig. 2, one can see that (I) interacts with a second water channel through the weak C3—H3···O2W interactions. Indeed, atom C3 acts as a hydrogen-bond donor to atom O2W in the molecule at (-x + 1, y + 1/2, -z + 1). As a result of interactions in this substructure, edge-fused R⁸₈(24) (Etter, 1990) rings running parallel to [100] are detected (Fig. 2). Additionally, organic molecules are intertwined through the C2—H2···O2 interaction. Atom C2 acts as a hydrogen-bond donor to atom O2 in the molecule at (x - 1, y, z + 1) (see Table 1).

In a second substructure, a channel formed exclusively by water molecules along [100] is observed. Indeed, infinite chains of water molecules, where atoms O2W and O1W interact through relatively strong hydrogen bonds, are detected. Thus, the structural organization of (I) shows the formation of two-dimensional water columns. The water molecules form a two-dimensional lattice which is positioned above and below the 6-methoxyquinoline N-oxide molecules and which helps to stabilize and support them in the crystalline structure (Fig. 3).

Theoretical calculations of bond lengths and angles were performed by the Hartree–Fock (HF) method with a 6-311++G(d,p) basis set and by density functional theory (DFT) B3LYP, also with a 6-311++G(d,p) basis set, and these values compared with the experimental values for (I) (Table 2). From these results we can conclude that the DFT basis set 6-311++G(d,p) is better suited in its approach to the experimental data. To enable a better understanding of the properties of (I), we further studied the stability of this compound in the gaseous state, calculating the harmonic frequencies and comparing the results with those observed in the fundamental vibrational frequencies.

The optimized structures, minimum electronic energies and vibrational frequencies were determined using semiempirical ab initio methods and DFT calculations. Geometry optimizations, single-point energies and vibrational frequencies were calculated using the GAUSSIAN09 program (Frisch et al., 2009). Frequencies were calculated for the optimized structure and the absence of imaginary values in the frequencies confirms the minimum energy value in the optimized structure.

The stability of the hydrated complex in the gaseous state was explored by calculating the harmonic frequencies and comparing them with the respective experimental vibrational frequencies. The frequency calculations were performed from the optimized structure, using HF/6-311++g(d) and DFT/6-311++g(d) at the B3LYP level of theory, using SCI-PCM [Reference or definition needed?] as a model of solvation for both bases. The stability of the compound was checked by the absence of imaginary frequencies in the calculation obtained by both methods. The experimental and simulated IR spectra are shown in Fig. 4. Vibrational analysis of (I) identifies the characteristic bands which correspond to the functional groups that are present in the compound. In the experimental spectrum, the most intense and sharp band is at 1213 cm^-1 and it can be seen in the simulated spectrum at 1235 cm^-1. This signal corresponds to the asymmetric O—N—C stretch. Analogously, one can observe the frequency of axial deformation of the N-oxide group at 1279 cm^-1 and in the simulated spectrum at 1273 cm^-1. The axial deformation band of C—O in the methoxy group, which is located at 1016 cm^-1 in the experimental spectrum and at 1038 cm^-1 in the simulated one, can also be assigned. These and other experimental and calculated bands are given in Table 3. Comparing the calculated and experimental values allows a good correlation between the bands to be found.

The optimization of the molecular geometry of (I) in the gaseous state was performed in order to analyse the stability of the molecule. The analysis of the total energy and the energy gap between the frontier molecular orbitals HOMO (highest occupied molecular orbital) and LUMO (lowest unoccupied molecular orbital) of the molecule characterize its molecular chemical stability.

Compound (I) shows an absorption band in the UV at λ = 234 nm in acetonitrile. Electronic transitions in this molecular system were analysed using B3LYP/6-311G**(d,p) with SCI-PCM (water) as a model of solvation, yielding the most intense band at λ = 267.93 nm (oscillator strength = 0.2016). This band corresponds to an electronic transition from the HOMO to the third lowest unoccupied molecular orbital, LUMO+2. According to Fig. 5, the HOMO presents a charge density localized over the quinoline ring, and a positive phase is localized over the O atom of the N-oxide group. The LUMO is also characterized by a charge distribution over the whole molecule, but that over the O atom of the N-oxide group is now negative.

The analysis of the coefficients of the molecular orbitals of the optimized geometry reflects that the HOMO and LUMO+2 orbitals are delocalized over the C—C bonds of the aromatic rings, and therefore the transition involves an electron-density transfer of type π → π* on the molecular plane of the aromatic backbone. The bathochromic shift of the calculated bands with respect to the experimental bands can be explained by the interaction of the N-oxide functional group with the polar solvent (Hsieh et al., 2010).

The Hirshfeld surface provides information not only on the areas of close contact, and therefore of strong interactions, but also on distant contact areas of weak interactions. The structure of (I) has a different environment on each side of the molecule and for this reason the Hirshfeld surface (Fig. 6b) is not symmetrical. Each contact on the surface can be identified individually from the colour pattern on the shape index surface (colour image available in the online version of the journal). The red spots with concave curvature that appear on the surface of the quinoline plane are similar to those occurring in the naphthalene plane (McKinnon et al., 1998). C—H donor regions with convex curvature can be observed, especially around the methyl group, and are identified by a deep blue colour in Fig. 6(b). The shorter of the O···H—O contacts between the O atom of the N-oxide group and the two water molecules is labelled 1 in Fig. 6(b). This red concave surface shows clearly the strong interactions on each side of the Hirshfeld surface. Another significant interaction can be seen in Fig. 6(b) and is labelled as 2. This corresponds to the interaction which links quinoline molecules through the C2—H2···O2 contact.

The fingerprint plot analysis of the title structure was performed by comparison with the other related structures, namely 6-methoxy-8-nitroquinoline (Chambers et al., 2004), MQNOHQ (Moreno-Fuquen et al., 2007) and 6-methoxy-2,4-dichloroquinoline (Subashini et al., 2009). The 6-methoxy-8-nitroquinoline and 6-methoxy-2,4-dichloroquinoline systems present fingerprint plots that are more or less symmetrical, as a result of the absence of strong or intense interactions. This behaviour is similar to that presented by the naphthalene ring (McKinnon et al., 2004). The two-dimensional fingerprint analysis of MQNOHQ, the molecule closest to (I), clearly shows the interaction of the O atom of the N-oxide group with the O atom of the hydroxyl group of the hydroquinone molecule [O···O = 2.6118 (18) Å and O···H = 1.67 (2) Å], through the emergence of an elongated peak projecting towards the bottom of the fingerprint plot. The fingerprint plot analysis of (I) (Fig. 6c) reveals the emergence of two sharp peaks, characteristic of the most important hydrogen-bond interactions (O—H···O or C—H···O) that occur between the molecules. The lower peak (where d_e < d_i) is more pronounced and corresponds to the interactions O1···H4W—O2, O1···H2W—O1W and O2···C2—H2, in which (I) plays the role of hydrogen-bond acceptor. Similarly, the upper peak (d_i < d_e) corresponds to the C3—H3···O2W interaction, where (I) acts as a hydrogen-bond donor. This behaviour is reflected in the Hirshfeld surface through the red concave surfaces, numbered 1 and 2 in Fig. 6(b).

Detailed analysis of the fingerprint plots can evaluate the overlapping contributions from the most important interactions, including O···H, H···H and C···H, facilitating comparison of the surface properties between (I) and MQNOHQ. This analysis shows that the H···H interactions are the largest contributor to the Hirshsfeld surfaces for both systems, 39.2% in MQNOHQ and 47.6% in (I). The contribution of O—H interactions on the surface is similar for both systems [11.8% in (I) and 12.8% in MQNOHQ]. The H···O distances for the O2W—H4W···O1 and O1···H2W—O1W interactions in (I) [H···O = 1.874 and 1.915 Å, respectively] are more elongated than that observed in MQNOHQ (H···O = 1.670 Å), probably because they also participate in the formation of the water channels along [001].

In conclusion, the O—H···O and C—H···O hydrogen-bond interactions in (I) are involved in the construction of the supramolecular architecture, and the water molecules play a major role in the formation of channels along [100] and [001]. The formation of the water channels, between which are positioned the molecules under study, gives greater stability to the crystal system. The IR spectrum of (I) computed at the DFT level with basis set 6-311++G(d,p) reproduces the vibrational wavenumbers and intensities with an accuracy which allows reliable vibrational assignments. Finally, analysis of the Hirshfeld surface and fingerprint plot for (I) allow the visualization of O···H—O hydrogen bonds as close intermolecular contacts within the supramolecular crystal lattice.

Related literature

For related literature, see: Allen et al. (1987); Ballabio et al. (1992); Biavatti et al. (2002); Campbell et al. (1988); Chambers et al. (2004); Cukierman (2000); Desiraju (1989, 1996); Etter (1990); Frisch (2009); Hsieh et al. (2010); Jude et al. (2002); Kaiser et al. (1996); Katakura & Koide (2006); Leiserowitz & Weinstein (1975); McKinnon et al. (1998, 2004); Moreno-Fuquen, Shankland & Fabbiani (2007); Moreno-Fuquen, Valencia, Kennedy, Gilmour & Ribeiro (2010); Nardelli (1995); Polshettiwar & Varma (2008); Sakamoto (2004); Somvanshi et al. (2008); Subashini et al. (2009); Sui et al. (1998); Tieleman et al. (2001); Towers et al. (1981).

S2. Experimental

The reagent was purchased from Aldrich Chemical Co. It was re-crystallized from acetonitrile to give (I), which melted at 376 (1) K.

S3. Refinement

C-bound H atoms were positioned geometrically, with C—H = 0.95 Å for aromatic H or 0.98 Å for methyl H, and refined using a riding model, with U_iso(H) = 1.2U_eq(C) for aromatic H or 1.5U_eq(C) for methyl groups. The water atoms H1W, H2W, H3W and H4W were located in a difference Fourier map and refined freely.

Computing details

Data collection: CrysAlis CCD (Oxford Diffraction, 2009); cell refinement: CrysAlis CCD (Oxford Diffraction, 2009); data reduction: CrysAlis CCD (Oxford Diffraction, 2009); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: ORTEP-3 for Windows (Farrugia, 1997) and Mercury (Macrae et al., 2008); software used to prepare material for publication: WinGX (Farrugia, 1999).

Figures

	Fig. 1. A plot of (I), with the atom-numbering scheme. Displacement ellipsoids are drawn at the 50% probability level.
	Fig. 2. Part of the crystal structure of (I), showing the formation of channels of water molecules along [001]. [Symmetry codes: (i) -x + 1, y + 1/2, -z + 1; (ii) -x + 2, y - 1/2, -z + 1; (iii) x, y, z + 1.]
	Fig. 3. A two-dimensional view of the water channels bridged by atoms O1 to form hydrophilic layers.
	Fig. 4. Comparison of the calculated IR spectrum of (I) (top) with the observed spectrum (bottom).
	Fig. 5. The atomic orbital compositions of the frontier molecular orbital for (I).
	Fig. 6. (a) Stick drawing of (I) in the orientation used to produce (b) the Hirshfeld surface and (c) the two-dimensional fingerprint plot for (I).

6-Ethoxyquinoline N-oxide dihydrate

Crystal data

C10H9NO2·2H2O	F(000) = 224
Mr = 211.21	Dx = 1.322 Mg m^−3
Monoclinic, P21	Melting point: 386(1) K
Hall symbol: P 2yb	Mo Kα radiation, λ = 0.71073 Å
a = 5.0000 (2) Å	Cell parameters from 2837 reflections
b = 17.1211 (6) Å	θ = 3.5–26.0°
c = 6.4421 (3) Å	µ = 0.10 mm^−1
β = 105.864 (4)°	T = 123 K
V = 530.48 (4) Å^3	Prism, pale yellow
Z = 2	0.21 × 0.16 × 0.12 mm

Data collection

Oxford Gemini S diffractometer	1928 independent reflections
Radiation source: fine-focus sealed tube	1760 reflections with I > 2σ(I)
Graphite monochromator	Rint = 0.017
ω scans	θmax = 26.0°, θmin = 3.5°
Absorption correction: multi-scan (CrysAlis PRO, Oxford Diffraction, 2009)	h = −6→6
Tmin = 0.965, Tmax = 1.000	k = −20→21
2836 measured reflections	l = −7→7

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)] = 0.035	Hydrogen site location: inferred from neighbouring sites
wR(F^2) = 0.079	H atoms treated by a mixture of independent and constrained refinement
S = 1.05	w = 1/[σ^2(Fo^2) + (0.0363P)^2 + 0.0351P] where P = (Fo^2 + 2Fc^2)/3
1928 reflections	(Δ/σ)max < 0.001
153 parameters	Δρmax = 0.15 e Å^−3
1 restraint	Δρmin = −0.15 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
O1	0.5089 (3)	0.79875 (8)	0.7368 (2)	0.0280 (4)
O2	1.0341 (3)	1.06063 (8)	0.2866 (2)	0.0269 (4)
O1W	0.7678 (3)	1.25800 (10)	0.6839 (3)	0.0335 (4)
O2W	0.6638 (3)	0.72504 (9)	0.1254 (3)	0.0268 (3)
N1	0.5114 (3)	0.87590 (9)	0.7698 (3)	0.0208 (4)
C1	0.3832 (4)	0.90549 (12)	0.9102 (3)	0.0239 (5)
H1	0.2954	0.8715	0.9880	0.029*
C2	0.3784 (4)	0.98627 (12)	0.9423 (3)	0.0248 (5)
H2	0.2899	1.0068	1.0436	0.030*
C3	0.5005 (4)	1.03582 (12)	0.8289 (3)	0.0224 (4)
H3	0.4942	1.0906	0.8499	0.027*
C4	0.6376 (4)	1.00521 (11)	0.6787 (3)	0.0192 (4)
C5	0.6453 (4)	0.92332 (11)	0.6517 (3)	0.0202 (4)
C6	0.7859 (4)	0.89034 (12)	0.5088 (3)	0.0209 (4)
H6	0.7933	0.8353	0.4934	0.025*
C7	0.9110 (4)	0.93845 (12)	0.3931 (3)	0.0219 (5)
H7	1.0063	0.9163	0.2984	0.026*
C8	0.8998 (4)	1.02078 (11)	0.4132 (3)	0.0204 (5)
C9	0.7674 (4)	1.05362 (12)	0.5544 (3)	0.0215 (4)
H9	0.7629	1.1088	0.5687	0.026*
C10	1.0346 (5)	1.14424 (12)	0.3004 (4)	0.0356 (6)
H10A	1.1272	1.1603	0.4487	0.053*
H10B	0.8426	1.1634	0.2598	0.053*
H10C	1.1343	1.1661	0.2023	0.053*
H1W	0.634 (5)	1.2537 (14)	0.734 (4)	0.037 (7)*
H2W	0.679 (5)	1.2708 (15)	0.555 (4)	0.041 (8)*
H3W	0.838 (6)	0.7354 (14)	0.180 (4)	0.037 (7)*
H4W	0.616 (6)	0.7473 (18)	0.003 (5)	0.060 (9)*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
O1	0.0431 (9)	0.0177 (7)	0.0231 (8)	−0.0011 (6)	0.0089 (7)	−0.0008 (6)
O2	0.0337 (8)	0.0249 (8)	0.0258 (8)	−0.0023 (6)	0.0144 (7)	0.0018 (7)
O1W	0.0269 (9)	0.0466 (10)	0.0286 (9)	0.0045 (7)	0.0101 (8)	0.0107 (8)
O2W	0.0293 (9)	0.0277 (8)	0.0245 (8)	−0.0015 (6)	0.0095 (7)	0.0032 (7)
N1	0.0230 (9)	0.0196 (9)	0.0183 (9)	0.0009 (6)	0.0033 (7)	−0.0003 (7)
C1	0.0250 (11)	0.0297 (12)	0.0173 (11)	−0.0008 (8)	0.0062 (9)	0.0004 (9)
C2	0.0226 (11)	0.0341 (13)	0.0188 (11)	0.0067 (8)	0.0072 (8)	−0.0051 (9)
C3	0.0235 (11)	0.0216 (10)	0.0214 (11)	0.0034 (8)	0.0049 (9)	−0.0038 (8)
C4	0.0165 (10)	0.0208 (10)	0.0190 (10)	0.0028 (8)	0.0028 (8)	−0.0002 (9)
C5	0.0185 (10)	0.0223 (10)	0.0180 (11)	0.0004 (8)	0.0019 (9)	0.0001 (8)
C6	0.0216 (11)	0.0207 (10)	0.0187 (10)	0.0024 (8)	0.0024 (8)	−0.0042 (9)
C7	0.0206 (11)	0.0280 (12)	0.0177 (11)	0.0017 (8)	0.0064 (9)	−0.0041 (9)
C8	0.0158 (10)	0.0252 (12)	0.0190 (11)	−0.0004 (7)	0.0027 (8)	0.0008 (8)
C9	0.0225 (10)	0.0190 (10)	0.0220 (11)	0.0022 (8)	0.0042 (8)	−0.0012 (9)
C10	0.0478 (15)	0.0255 (12)	0.0396 (14)	−0.0049 (10)	0.0225 (12)	0.0045 (11)

Geometric parameters (Å, º)

O1—N1	1.338 (2)	C3—H3	0.9500
O2—C8	1.372 (2)	C4—C5	1.415 (3)
O2—C10	1.434 (2)	C4—C9	1.426 (3)
O1W—H1W	0.82 (3)	C5—C6	1.420 (3)
O1W—H2W	0.86 (3)	C6—C7	1.371 (3)
O2W—H3W	0.86 (3)	C6—H6	0.9500
O2W—H4W	0.85 (3)	C7—C8	1.418 (3)
N1—C1	1.343 (2)	C7—H7	0.9500
N1—C5	1.401 (3)	C8—C9	1.382 (3)
C1—C2	1.400 (3)	C9—H9	0.9500
C1—H1	0.9500	C10—H10A	0.9800
C2—C3	1.368 (3)	C10—H10B	0.9800
C2—H2	0.9500	C10—H10C	0.9800
C3—C4	1.429 (3)
C8—O2—C10	116.89 (15)	C4—C5—C6	120.58 (17)
H1W—O1W—H2W	98 (2)	C7—C6—C5	119.61 (19)
H3W—O2W—H4W	106 (2)	C7—C6—H6	120.2
O1—N1—C1	119.62 (15)	C5—C6—H6	120.2
O1—N1—C5	118.10 (15)	C6—C7—C8	120.82 (18)
C1—N1—C5	122.26 (17)	C6—C7—H7	119.6
N1—C1—C2	120.20 (18)	C8—C7—H7	119.6
N1—C1—H1	119.9	O2—C8—C9	126.11 (18)
C2—C1—H1	119.9	O2—C8—C7	113.72 (16)
C3—C2—C1	120.39 (18)	C9—C8—C7	120.15 (18)
C3—C2—H2	119.8	C8—C9—C4	120.43 (18)
C1—C2—H2	119.8	C8—C9—H9	119.8
C2—C3—C4	120.06 (18)	C4—C9—H9	119.8
C2—C3—H3	120.0	O2—C10—H10A	109.5
C4—C3—H3	120.0	O2—C10—H10B	109.5
C5—C4—C9	118.37 (16)	H10A—C10—H10B	109.5
C5—C4—C3	118.73 (17)	O2—C10—H10C	109.5
C9—C4—C3	122.90 (17)	H10A—C10—H10C	109.5
N1—C5—C4	118.34 (16)	H10B—C10—H10C	109.5
N1—C5—C6	121.08 (17)
O1—N1—C1—C2	−178.45 (17)	C3—C4—C5—C6	−178.23 (18)
C5—N1—C1—C2	0.3 (3)	N1—C5—C6—C7	178.96 (18)
N1—C1—C2—C3	1.0 (3)	C4—C5—C6—C7	−1.2 (3)
C1—C2—C3—C4	−0.9 (3)	C5—C6—C7—C8	−0.5 (3)
C2—C3—C4—C5	−0.4 (3)	C10—O2—C8—C9	−0.7 (3)
C2—C3—C4—C9	179.60 (18)	C10—O2—C8—C7	−179.23 (17)
O1—N1—C5—C4	177.15 (16)	C6—C7—C8—O2	−179.82 (17)
C1—N1—C5—C4	−1.7 (3)	C6—C7—C8—C9	1.5 (3)
O1—N1—C5—C6	−3.0 (3)	O2—C8—C9—C4	−179.36 (18)
C1—N1—C5—C6	178.22 (17)	C7—C8—C9—C4	−0.9 (3)
C9—C4—C5—N1	−178.36 (17)	C5—C4—C9—C8	−0.7 (3)
C3—C4—C5—N1	1.6 (3)	C3—C4—C9—C8	179.27 (18)
C9—C4—C5—C6	1.8 (3)

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
O2W—H4W···O1^i	0.85 (3)	1.87 (3)	2.721 (2)	179 (3)
O1W—H2W···O1^ii	0.86 (3)	1.92 (3)	2.774 (2)	178 (2)
O1W—H1W···O2W^ii	0.82 (3)	2.00 (3)	2.817 (2)	170 (2)
C3—H3···O2W^ii	0.95	2.45	3.375 (2)	164
O2W—H3W···O1W^iii	0.86 (3)	1.96 (3)	2.828 (2)	177 (2)
C2—H2···O2^iv	0.95	2.46	3.404 (2)	176

Symmetry codes: (i) x, y, z−1; (ii) −x+1, y+1/2, −z+1; (iii) −x+2, y−1/2, −z+1; (iv) x−1, y, z+1.