Structural aspects of intermolecular interactions in the solid state of 1,4-di­benzylpiperazines bearing nitrile or amidine groups

The crystal structures of the title 1,4-bis(4-cyanobenzyl)piperazine (1) and 1,4-bis(4-amidinobenzyl)piperazine tetrahydrochloride tetrahydrate (2) are reported. Compound (1) crystallizes in the triclinic space group and compound (2) in the monoclinic space group P2₁/n. In both (1) and (2) the asymmetric unit contains one half of the molecule because the central piperazine rings were located across a symmetry center. The packing of both molecules was dominated by hydrogen bonds. The crystal lattice of (1) was formed by weak C—H⋯N and C—H⋯π interactions. The crystal structure of (2) was completely different, with cations as well as chloride anions and water molecules taking part in intermolecular interactions. Single-crystal X-ray diffraction studies combined with density functional theory (DFT) calculations allowed the characterization of the intermolecular interactions in those two systems having different types of very strong electrophilic groups: non-ionic nitrile and ionic amidine. Chemical shift data from ¹³C CP/MAS (Cross Polarization Magic Angle Spinning) NMR spectra were analyzed using the different procedures for the theoretical computation of shielding constants.

1. Introduction

The increasing demands of the pharmaceutical industry for rapid molecular structure determination of pharmaceutical solids have prompted the development of joint analysis methods spanning X-ray diffraction, ¹³C CP/MAS NMR and molecular modeling. The solid-state form of a drug can have a dramatic effect on its bioavailability and physical properties, and the regulatory approval for many drugs is granted only for the defined polymorph (Barrett et al., 2013; Geppi et al., 2008; Pereira Silva et al., 2011). Moreover, because the existence of solvates (called pseudopolymorphs) is an abiding problem in pharmaceutical chemistry (the solid-state form of a solvate can also be treated as the separate form of the drug), it is important to structurally characterize these. Solid-state structural studies of new substances which are designed as potential chemotherapeutic agents has also aroused great interest (Harris, 2007). Hydrogen bonds are crucial to the interactions between biomolecules, with the macromolecular target and their analysis contributing to expanding the information about biomolecular interactions.

The molecules analyzed in this investigation can be considered as pentamidine analogs and are of interest because of their potential as chemotherapeutic agents against pneumocystis pneumonia (PCP) caused by the fungus Pneumocystis jiroveci in patients with compromised immune systems (Ponce et al., 2010; Furrer et al., 1999; Maini et al., 2013) or as anticancer and antimicrobial agents (Barrett et al., 2013).

The screening of new pentamidine analogs led to the selection of less toxic and highly active molecules (with in vitro IC₅₀ values as low as 0.002 µM compared with 0.5 µM for pentamidine) which contain functional groups such as the piperazine ring that increase the rigidity of the molecule (Vanden Eynde et al., 2004; Huang et al., 2009; Cushion et al., 2004, 2006; Mitsuyama et al., 2008). The intermolecular interactions in the solid state of piperazine-type pentamidine analogs were not examined, and the results obtained here could be useful in the future explanation of their biological features.

In this investigation we analyzed and compared the solid-state structures of new pentamidine analogs containing the piperazine moiety (Fig. 1), i.e. 1,4-bis(4-cyanobenzyl)piperazine (1) and 1,4-bis(4-amidinobenzyl)piperazine tetrahydrochloride tetrahydrate (2) with particular attention paid to hydrogen bonding, using different methods: single-crystal X-ray diffraction analyses combined with molecular modeling and ¹³C CP/MAS NMR spectroscopy. The bis-nitrile compound (1) is an intermediate for the synthesis of bis-amidine (2) via the Pinner reaction (Pinner & Klein, 1877). Diffraction-quality single crystals of bis-amidines are difficult to grow, and the crystal structures of few bis-amidines related to pentamidine have been reported (Maciejewska et al., 2006; Lowe et al., 1989; Srikrishnan et al., 2004; Donkor et al., 1995). Some authors used the information obtained for the crystals of bis-nitriles to explain the properties of bis-amidines (Cui et al., 2003). (The authors designed the inhibitors of key enzymes in the coagulation cascade, and to experimentally check the geometry of the amidine inhibitors they determined their cyanosynthetic intermediates by X-ray diffraction.) In this work we show that such an approach to the problem is inappropriate due to the limited similarity between those two groups.

Figure 1 Molecular conformation and atom-numbering for (1) and (2). Displacement ellipsoids are drawn at 50% probability for non-H atoms. The molecules lie across inversion centers according to the symmetry operations -x, 2-y, 1-z in (1) and -x, -y, -z in (2).

2. Experimental

2.1. Chemistry

All chemicals were purchased from major chemical suppliers as high or the highest purity grade and were used without any further purification. The solvents, K₂CO₃, HCl, NaOH and ammonia, were obtained from POCH (Gliwice, Poland). The substrates 4-cyanobenzyl bromide and piperazine were obtained from Alfa AESAR (Karlsruhe, Germany). The scheme below presents the synthetic route to bis-amidine (2) via bis-nitrile (1). In the first step 1,4-bis(4-cyanobenzyl)piperazine (1) was prepared by a modification of the procedure given by Gruenenthal (2008). The Pinner reaction of (1) to form the bis-amidine (2) was conducted for 2 weeks due to the poor solubility of (1) in ethanolic HCl.

Compound (1) was mentioned in the paper by Spychała (1999), but the detailed synthetic and spectral information was not given and therefore it is described in §2.1.1. Melting points were determined with an Electrothermal 9001 digital melting point apparatus. Elemental analyses were performed on a Vario EL III CHNS element analyzer, and were averaged from two independent determinations. Chemical shifts (p.p.m.) in CDCl₃ (1) or DMSO-d₆ (2) were referenced to tetramethylsilane (TMS) .

2.1.1. 1,4-Bis(4-cyanobenzyl)piperazine (1)

4-Cyanobenzyl bromide (19.6 g 0.1 mol), K₂CO₃ (13.8 g, 0.1 mol), piperazine (4.3 g, 0.05 mol) and DMF (235 ml) were stirred for 1 h at room temperature and then stirred and heated for 5 h at 353–363 K. The reaction mixture was cooled to room temperature, ice-water (700 ml) was added, and stirring was continued for 0.5 h at 272–278 K. The white precipitate was filtered off, washed with cold water (2 × 400 ml) and dried in vacuo. The crude product was crystallized from acetone to give 27.2 g (86%) of fine colorless crystals of (1). M.p. 479.5–480.5 K; C₂₀H₂₀N₄ (M_r = 316): calc. C 75.95, H 6.33, N 17.72%; found C 75.99, H 6.36, N 17.68%. ¹H NMR (299.87 MHz, CDCl₃): δ = 2.48 (br s, 8H, 9-CH₂, 9′-CH₂, 10-CH₂, 10′-CH₂), 3.56 (s, 4H, 8-CH₂, 8′-CH₂), 7.43–7.46 (d, J = 8.1 Hz, 4H, 2-CH, 2′-CH, 6-CH, 6′-CH), 7.59–7.62 (pd, J = 8.1 Hz, 4H, 3-CH, 3′-CH, 5-CH, 5′-CH) p.p.m. ¹³C NMR (50.28 MHz, CDCl₃): δ = 53.24 (C9, C9′, C10, C10′), 62.54 (C8, C8′), 111.15 (C4, C4′), 119.12 (C7, C7′), 129.71 (C2, C2′, C6, C6′), 132.32 (C3, C3′, C5, C5′), 144.22 (C1, C1′) p.p.m.

2.1.2. 1,4-Bis(4-amidinobenzyl)piperazine (2)

A slurry of 1,4-bis(4-cyanobenzyl)piperazine (1) (1.26 g; 4 mmol) in anhydrous ethanol (40 ml) was saturated with anhydrous HCl at 273–278 K. The contents were stirred in a sealed vessel for 2 weeks at room temperature. The reaction was carried out until the starting material was completely consumed (TLC, IR). The solvent was then removed almost to dryness in vacuo at 312 K. The residue was ground with dry diethyl ether (100 ml) until colorless crystals of an unstable intermediate (ethyl imidate) formed; these were quickly filtered off and dried under reduced pressure over NaOH granules.

Dry ethanol (40 ml) was saturated with anhydrous ammonia gas at 273–278 K, the entire amount of ethyl imidate added and the mixture stirred in a sealed vessel for 24 h at room temperature. Ethanol was removed almost to dryness under reduced pressure and a solution of NaOH (1.0 g) in water (40 ml) was added to the residue and stirred for 15 min. The free base, which formed as a fine white precipitate, was filtered, washed thoroughly with water and dried under reduced pressure over anhydrous NaOH granules. The dry powder was washed with chloroform to remove unreacted bis-nitrile, dried again in vacuo and mixed with anhydrous ethanol (10 ml), acidified with an excess of ethanolic HCl and refluxed for 0.5 h. After cooling, dry diethyl ether (30 ml) was added slowly with stirring. After a few minutes the colorless crystals were filtered off, washed with dry diethyl ether and dried. Recrystallization from aqueous ethanol gave 1.90 g (84%) of pure bis-amidine (2). M.p. 564–566 K; C₂₀H₂₆N₆·4HCl·4H₂O (M_r = 568): calc. C 42.25, H 6.69, N 14.79, Cl 25.00%; found C 42.31, H 6.77, N 14.42, Cl 24.57%. ¹H NMR (299.87 MHz, DMSO-d₆): δ = 3.48 (br s, 10H, 9-CH₂, 9′-CH₂, 10-CH₂, 10′-CH₂, 2NH), 4.49 (s, 4H, 8-CH₂, 8′-CH₂), 7.94 (bs, 8H, 2-CH, 2′-CH, 6-CH, 6′-CH, 3-CH, 3′-CH, 5-CH, 5′-CH), 9.38 (s, 4H, 2NH₂), 9.56 (s, 4H, 2NH₂) p.p.m. ¹³C NMR (50.28 MHz, DMSO-d₆): δ = 48.03 (C9, C9′, C10, C10′), 57.84 (C8, C8′), 128.36 (C3, C3′, C5, C5′), 128.51 (C4, C4′), 131.61 (C2, C2′, C6, C6′), 131.24 (C1, C1′), 164.98 (C7, C7′) p.p.m.

2.2. Crystallography

Crystals of (1) suitable for X-ray analysis were grown by slow evaporation from acetone, and crystals of (2) were grown by slow evaporation from 96% ethanol and a few drops of water (which were added to the ethanol mixture until its homogeneity was achieved). Diffraction data were collected on an Oxford Diffraction KM4 CCD diffractometer using Mo Kα radiation at room temperature for (1) and on an Oxford Diffraction SuperNova CCD diffractometer using Cu Kα radiation at 130 K for (2). Data reduction was carried out using CrysAlis Pro (Agilent, 2011, 2012) for (1) and (2), respectively.

The unit-cell parameters were determined by least-squares treatment of the angles of the highest-intensity reflections chosen from all the experiments. The structures were solved by direct methods using SHELXS97 and refined on F² by full-matrix least-squares with SHELXL97 (Sheldrick, 2008). The function was minimized with for (1) and for (2), where P = (F_o² + 2F_c²)/3. For (1) an empirical extinction correction was also applied according to the formula , and the extinction coefficient χ was equal to 0.08 (1).

All non-H atoms were refined with anisotropic displacement parameters. The coordinates of the H atoms of (1) and the majority of hydrogen positions of (2) were generated geometrically. In (2) the hydrogen at N1 and the H atoms of water molecules were found on the difference map. Then the H atom involved in hydrogen bonds in (1) and all the H atoms in (2) were refined isotropically. The other atoms were refined as a riding model with U_iso(H) = 1.2U_eq(carrier atom). All the details of the measurement, crystal data and structure refinement are given in Table 1. ¹

Table 1
Crystal data, data collection and structure refinement for (1) and (2)

	(1)	(2)
Crystal data
Chemical formula	C20H20N4	C20H30N6·4Cl·4H2O
Mr	316.40	568.36
Crystal system, space group	Triclinic, P	Monoclinic, P21/n
T (K)	293	130
a, b, c (Å)	6.6267 (4), 8.3540 (5), 8.6889 (5)	6.1121 (2), 12.8231 (3), 18.3831 (4)
α, β, γ (°)	83.876 (5), 72.881 (5), 70.676 (5)	90, 99.274 (2), 90
V (Å^3)	433.78 (4)	1421.96 (7)
Z	1	2
Radiation type	Mo Kα	Cu Kα
μ (mm^−1)	0.07	4.08
F(000)	168	600
Crystal size (mm)	0.5 × 0.4 × 0.35	0.5 × 0.1 × 0.08
Data collection
Diffractometer	Xcalibur, Sapphire2, large Be window diffractometer	SuperNova, Single source at offset), Atlas diffractometer
Absorption correction	Multi-scan CrysAlis PRO	Multi-scan CrysAlis PRO
Tmin, Tmax	0.965, 1.000	0.632, 1.000
No. of measured, independent and observed [I > 2σ(I)] reflections	6352, 1625, 1409	6224, 2866, 2771
Rint	0.010	0.015
(sin θ/λ)max (Å^−1)	0.610	0.628
Refinement
R[F^2 > 2σ(F^2)], wR(F^2), S	0.039, 0.118, 1.05	0.027, 0.073, 1.04
No. of reflections	1625	2866
No. of parameters	114	230
H-atom treatment	H atoms treated by a mixture of independent and constrained refinement	All H-atom parameters refined
Δρmax, Δρmin (e Å^−3)	0.13, −0.12	0.25, −0.21
Computer programs: CrysAlis PRO (Agilent, 2011), CrysAlis PRO (Agilent, 2012), SHELXS97, SHELXL97 (Sheldrick, 2008), Mercury (Macrae et al., 2006).

2.3. Molecular modeling

Structural optimizations were performed at the density functional theory (DFT) level with B3LYP/6-311(d,p) hybrid functional, and the locations of the true minima were confirmed by vibrational analysis using GAUSSIAN09 (Frisch et al., 2009). The crystal atomic coordinates were used as the starting point for DFT computations. To investigate the intermolecular interactions in the solid state (Gholivand et al., 2013), the optimization of H atoms positions was performed for the cluster C1 built up from three molecules of (1). The target molecule of (1) was surrounded by two neighboring molecules from one layer which were connected to each other via two C3—H3A⋯N2 hydrogen bonds with a donor–acceptor distance of 3.629 (2) Å (Fig. 2). For compound (2) cluster C2 was built up from one cation of molecule (2) surrounded by 14 chloride anions and ten water molecules (Fig. 2), because compound (2) did not form hydrogen bonds directly with another molecule of (2). All hydrogen-bond parameters are given in Table 2. The positions of the H atoms were optimized, while the other atoms were kept fixed during the optimizations. This approach allowed us to perform the analyses of the central molecule when the neighboring molecules were present. The hydrogen-bonding energies were calculated for compound (1) using the equation: E_Hbond = ½(E_cluster − E_two − E_one), taking into account the two hydrogen bonds C3—H3A⋯N2 formed in the layer. E_two is composed of two neighboring molecules at the left-hand side of Fig. 2, and E_one is the remaining part of C1. For compound (2) the average energy of hydrogen bonds was calculated based on the energy difference between the hydrogen-bonded cluster and its fragments as represented by the equation: E_Hbond = 1/28(E_cluster − E_anions+water − E_one).

Table 2
Hydrogen-bonding geometry (Å, °) for (1) and (2)

D—H⋯A	d(D—H)	d(H⋯A)	d(D⋯A)	∠(DHA)
(1)
C3—H3A⋯N2^i	0.99 (2)	2.65 (2)	3.627 (2)	166 (1)
(2)
N1—H1⋯O1^ii	0.90 (2)	1.79 (2)	2.687 (2)	173 (2)
N2—H2B⋯O2^vi	0.85 (2)	2.05 (2)	2.887 (1)	170 (2)
N2—H2C⋯Cl2^iv	0.86 (2)	2.35 (2)	3.189 (1)	166 (2)
N3—H3C⋯O2^ii	0.88 (2)	2.07 (2)	2.941 (1)	175 (2)
N3—H3B⋯Cl1^vi	0.86 (2)	2.43 (2)	3.285 (1)	172 (1)
O1—H1AW⋯Cl1^iii	0.81 (2)	2.27 (2)	3.067 (1)	167 (2)
O1—H1BW⋯Cl2^i	0.84 (3)	2.27 (3)	3.098 (1)	171 (2)
O2—H2BW⋯Cl1	0.86 (2)	2.25 (2)	3.088 (1)	163 (2)
O2—H2AW⋯Cl2^v	0.84 (2)	2.26 (3)	3.094 (1)	173 (2)
C8—H8A⋯Cl1	0.98 (2)	2.77 (2)	3.678 (1)	155 (1)
C8—H8B⋯Cl2^i	0.95 (2)	2.64 (2)	3.546 (1)	161 (1)
C6—H6A⋯O2	0.94 (2)	2.64 (2)	3.422 (2)	141 (1)
C9—H9A⋯Cl1^ii	0.94 (2)	2.70 (2)	3.471 (1)	140 (1)
C10—H10B⋯Cl2^i	0.96 (2)	2.85 (2)	3.723 (1)	151 (1)
Symmetry codes: for (1): (i) -x+2, -y+1, -z+2; for (2): (i) x+1, y, z; (ii) x-1, y, z; (iii) -x+1, -y, -z; (iv) -x-1, -y+1, -z; (v) ; (vi) .

Figure 2 Views of the clusters C1 and C2 with hydrogen-bonding scheme analyzed for molecules (1) and (2).

The isotropic ¹³C shielding constants σ (p.p.m.) for (1) and (2) in the solid state were computed with the GIAO (gauge including the atomic orbital) method using GAUSSIAN09 (Frisch et al., 2009) at the DFT level with B3LYP/6-311(d, p) hybrid functional. For the assignment of ¹³C CP/MAS NMR resonances, the structures obtained by X-ray diffraction were optimized prior to chemical shielding calculations using four different procedures:

• (i) the positions of all atoms in (1) and (2) were fully optimized;

• (ii) the positions of the non-H atoms were fixed, but the H atoms were allowed to move in (1) and (2);

• (iii) the positions of all atoms in the clusters C1 and C2 were optimized, and the shielding constants were analyzed for the target molecules (1) and (2);

• (iv) the positions of the non-H atoms were fixed in clusters C1 and C2, and the shielding constants were analyzed for the target molecules (1) and (2).

2.4. NMR spectra measurements

Solution ¹H NMR and ¹³C NMR spectra were recorded at 298 K on a Varian NMRS-300 spectrometer and standard Varian software was employed. Solid-state ¹³C CP/MAS NMR spectra were recorded on a Bruker Avance DMX 400 spectrometer at 100.62 MHz using a 4 mm diameter zirconia rotor. Conventional single-contact ¹H → ¹³C cross-polarization (CP) with reversal of spin temperature in the rotating frame, and high proton decoupling during signal acquisition were performed. The acquisition conditions for ¹³C CP/MAS NMR were: pulse duration 2.5 µs; contact time 4 ms; repetition time 48 s for (1) and 50 s for (2); spectral width 24 kHz; spinning speed 8 kHz. Chemical shifts δ (p.p.m.) were referenced to TMS.

3. Results and discussion

3.1. X-ray structure analysis

The crystal and molecular structures of 1,4-bis(4-cyanobenzyl)piperazine (1) and 1,4-bis(4-amidinobenzyl)piperazine tetrahydrochloride (2) were determined by single-crystal X-ray diffraction. A perspective view (Macrae et al., 2006) of the molecular conformations of (1) and (2), together with the atom-numbering scheme, is illustrated in Fig. 1. Hydrogen-bonding parameters are listed in Table 2, and selected bond lengths, bond angles and torsion angles are listed in Table 3.

Table 3
Selected bond lengths (Å), valence angles (°) and torsion angles (°) for (1) and (2)

	(1)	(2)
C4—C7	1.440 (2)	1.483 (2)
C1—C8	1.509 (2)	1.506 (2)
N1—C8	1.458 (2)	1.515 (2)
C2—C1—C8	119.3 (1)	119.9 (1)
C1—C8—N1	114.3 (1)	110.9 (1)
C8—N1—C10	111.1 (1)	110.5 (1)
C6—C1—C8—N1	26.0 (2)	100.0 (1)
C1—C8—N1—C10	71.1 (1)	171.2 (1)
C5—C4—C7—N3	–	−32.4 (2)
C3—C4—C7—N3	–	147.2 (1)

The results show that the compounds 1,4-bis(4-cyanobenzyl)piperazine (1) and 1,4-bis(4-amidinobenzyl)piperazine tetrahydrochloride tetrahydrate (2), the piperazine-derived analogs of pentamidine, crystallize in the triclinic space group and monoclinic space group P2₁/n, respectively. In both (1) and (2), the asymmetric units contain one half of the molecule because the central piperazine rings are located across symmetry centers according to the symmetry operators (−x, 2 − y, 1 − z) in (1) and (−x, −y, −z) in (2). Additionally, in the asymmetric unit of (2) there are two chloride ions and two molecules of water. As a consequence, we found the tetrachloride salt of this derivative in the lattice. The structure of (1) consists of two cyanobenzyl groups which are joined by the piperazine ring which adopts the expected chair conformation. The cyanobenzyl moiety is almost planar with a maximum deviation of 0.045 (1) Å for C8. The orientation of the piperazine ring with respect to the cyanobenzyl fragment is characterized by C1—C8—N1—C9 and C6—C1—C8—N1 torsion angles of −168.6 (1) and 26.0 (2)°, respectively.

In (2) the benzyl group is essentially planar with a maximum deviation of 0.011 (1) Å for C2. The C7 atom from the amidine group was found to be coplanar with the above fragment and the N2 and N3 atoms of this group were displaced by 0.556 (2) and −0.653 (2) Å, respectively. The location of the amidine group with respect to the benzene ring can also be characterized by the C3—C4—C7—N2 and C3—C4—C7—N3 torsion angles of −31.6 (2) and 147.1 (1)°, respectively. The central piperazine ring adopts a chair conformation and its orientation with respect to the benzyl fragment moiety is characterized by C1—C8—N1—C9 and C6—C1—C8—N1 torsion angles of 66.5 (1) and 100.0 (1)°, respectively.

The packing of both molecules is dominated by hydrogen bonds (Table 2). In the crystal of (1), the nitrile groups take part in intermolecular C3—H3A⋯N2 (−x + 2, −y + 1, −z + 2) hydrogen bonds which link the molecules into extended chains. The molecules are further organized into layers parallel to the (011) plane via weak C9—H9A⋯π(x − 1, y, z) contacts (Figs. 3 and 4). We also found such a participation of the nitrile substituents in the hydrogen bonds for other analogs of pentamidine (Żabiński et al., 2007, 2010; Maciejewska et al., 2008).

Figure 3 The chains of the molecules of (1) within a layer.

Figure 4 The packing arrangement of (1) showing layers parallel to the (011) plane.

The crystal structure of (2) differs in that the packing involves cations, chloride anions and water molecules. Each cation is surrounded by chloride anions and molecules of water and they are linked by N—H⋯O, N—H⋯Cl, C—H⋯O and C—H⋯Cl hydrogen bonds (see Table 2 and Fig. 5). The O atoms of the water molecules also participate in O—H⋯Cl interactions and as a consequence a three-dimensional lattice is obtained. Each water molecule and each chloride anion takes part in intermolecular hydrogen bonds. It is interesting to note that there are no direct hydrogen bonds between neighboring molecules of (2). In contrast, the pentamidine analog with three O atoms in the linker forms direct hydrogen bonds using these atoms as proton acceptors (Maciejewska et al., 2006; Lowe et al., 1989; Donkor et al., 1995) even in the presence of water molecules. The protonated N atoms of the piperazine ring in (2) cannot be involved in intermolecular interactions as proton acceptors, and water molecules serve as both proton donors and proton acceptors, providing the main intermolecular links. So far, X-ray studies of the structurally related bis-amidines did not present detailed analysis of the intermolecular hydrogen bonding (Lowe et al., 1989) or showed only a few intermolecular hydrogen bonds involving the anions or water molecules (Donkor et al., 1995; Maciejewska et al., 2006; Lowe et al., 1989; Srikrishnan et al., 2004).

Figure 5 Projection of the crystal structure of (2) along the a axis.

3.2. Parameters of hydrogen bonding in the clusters

The theoretical hydrogen-bond parameters for the target molecules (1) and (2) in clusters C1 and C2 are presented in Table 4. The donor–acceptor distances for hydrogen bonds in model clusters are equal to the experimental values since the optimizations were performed only for the H-atom positions. The N—H, C—H and O—H bonds are longer by 0.11–0.16 Å than those obtained from the crystal structure determinations, and as a result the hydrogen–acceptor distances H⋯A are shorter, suggesting stronger intermolecular interactions. Theoretical N—H⋯O, O—H⋯Cl and N—H⋯Cl angles are more linear than the crystallographic values.

Table 4
Theoretical hydrogen-bonding parameters for (1) and (2) calculated in clusters C1 and C2, respectively (see Fig. 2)

D—H⋯A	d(D—H)C1	d(D—H)gas	d(H⋯A)C1	d(D⋯A)C1	∠(DHA)C1
(1)
C3—H3A⋯N2	1.083	1.083	2.56	3.628	171

D—H⋯A	d(D—H)C2	d(D—H)gas	d(H⋯A)C2	d(D⋯A)C2	∠(DHA)C2
(2)
N1—H1⋯O1	1.070	1.026	1.62	2.688	177
N2—H2B⋯O2	1.016	1.015	1.91	2.886	168
N2—H2C⋯Cl2	1.029	1.010	2.17	3.189	172
N3—H3C⋯O2	1.035	1.011	1.92	2.942	174
N3—H3B⋯Cl1	1.011	1.015	2.27	3.285	178
O1—H1AW⋯Cl1	0.969	–	2.10	3.067	177
O1—H1BW⋯Cl2	0.974	–	2.12	3.097	179
O2—H2BW⋯Cl1	0.976	–	2.12	3.088	176
O2—H2AW⋯Cl2	0.967	–	2.13	3.093	175
C8—H8A⋯Cl1	1.093	1.091	2.66	3.680	162
C8—H8B⋯Cl2	1.091	1.090	2.50	3.546	165
C6—H6A⋯O2	1.084	1.089	2.50	3.422	153
C9—H9A⋯Cl1	1.089	1.091	2.59	3.472	150
C10—H10B⋯Cl2	1.088	1.091	2.75	3.723	157

The computed hydrogen-bonding energy in the C1 model cluster between the molecules (1) was equal to −8.8 kJ mol⁻¹, which is characteristic of very weak interactions. In the first approximation (the hydrogen bonds presented in Table 4), the average hydrogen-bonding energy in the C2 model cluster between the molecule of (2) and the surrounding water and chloride anions was calculated as −406.8 kJ mol⁻¹, a very high value. In the second approximation, we considered the 48 interactions with d(H⋯A) distances below 3.6 Å (thereby incorporating solvation of the cations by water molecules), and the average intermolecular interaction energy of compound (2) with all surrounding chloride anions and water molecules was −237.3 kJ mol⁻¹. The intermolecular bonds connecting bis-nitriles are much weaker than the intermolecular bonds formed by bis-amidines. The short contacts in the bis-nitrile crystal (1) are due to weak, nonpolar interactions, whereas in the bis-amidine crystal (2) they are dominated by hydrogen bonds between water molecules and chloride anions.

3.3. Solid-state ¹³C CP/MAS NMR spectra analysis of (1) and (2)

Solid-state ¹³C CP/MAS NMR spectra of 1,4-bis(4-cyanobenzyl)piperazine (1) and 1,4-bis(4-amidinobenzyl)piperazine (2) are presented in Fig. 6. The crystals for the ¹³C CP/MAS NMR experiments were collected in the same manner as for the single-crystal X-ray diffraction. In neither spectrum were multiplets observed, i.e. we observed a single resonance for each pair of chemically equivalent C atoms. Only for the piperazine ring was an additional signal detected, in accordance with the crystallographic results. Preliminary assignments were carried out on the basis of solution chemical shifts and on the basis of the computed shielding constants obtained for the fully optimized structures of (1) and (2), as described in §2.3 point (i). For molecule (1) the match between the experimental chemical shifts δ and the theoretical shielding constants σ was very close [the correlation coefficient for the linear correlations of σ = f(δ) was R² = 0.997 – see the supporting information for more details]. High correlation coefficients were also obtained for the shielding constants calculated by procedures (ii) and (iii). The computations correctly predicted higher shielding constants for C9/C9′ than for C10/C10′ in the piperazine ring. As can be seen from Table 5, significant differences between the isotropic chemical shifts measured in solution and in the solid state were found for C7/C7′ and C3/C3′ which are engaged in hydrogen bonding and for the piperazine C atoms. Surprisingly, the calculations on cluster C1 performed using procedure (iv) produced the worst correlation, although the correlation coefficient was still high (R² = 0.991). Apparently, the strength of specific solid-state effects is weak, and a simple comparison between the solution and solid-state resonances, and the simple calculation for the isolated molecule is sufficient for structural analysis. The resonances of C3/C3′ and C5/C5′ overlap at 131.8 p.p.m., although the calculation showed higher shielding for C5/C5′. This can be caused by the impact of intramolecular motions of the benzene rings in the solid state which were not considered in the calculations. The resonances of the ortho pairs C2/C2′ and C6/C6′ to the piperazine linker were clearly separated. This is a well known phenomenon dependent on intermolecular interactions and the nature of the substituent present at the neighboring C atom. The separation of signals indicated that methylene groups are slightly twisted relative to the benzene ring plane. The resonances of C7/C7′ proximal to N2/N2′ are broadened and split into unequal doublets due to a residual coupling to the quadrupolar ¹⁴N atom (Olivieri et al., 1987). For other C atoms proximal to N atoms only some broadening was observed. The ¹³C CP/MAS spectrum of molecule (2) had the same characteristics as the spectrum of (1) in that no multiplets were observed and the C7/C7′ atoms were broadened by a dipolar coupling to the quadrupolar ¹⁴N nucleus. After preliminary assignment based on the computed shielding constants obtained for the fully optimized structure as described in §2.3 point (i), the shielding constants obtained using three different procedures (ii), (iii) and (iv) were compared to the experimental chemical shifts. The correlation coefficients R² for the linear correlations σ = f(δ) were within the range 0.973–0.996. The highest correlation coefficient was obtained for procedure (ii) – see the supporting information for more details. Procedure (iv), which considered cluster C2, produced the worst results as well as for cluster C1. This observation agreed with our earlier findings for bis-nitriles that the shielding constant computation based on the single molecule structure can be informative for the analysis of solid-state structure based on NMR spectra (Maciejewska et al., 2008). In the spectrum of (2) separated resonances for all aromatic C atoms were observed. Both C atoms ortho and both C atoms meta to the amidine group are shielded in different ways. The lack of coplanarity of amidine groups with the benzene ring clearly affects the shielding of these aromatic C atoms. Next we compared the ¹³C resonances in the solid state, δ_solid, and in solution, δ_solution (Table 5). As can be seen, the highest negative differences between the their chemical shifts were found for C2/C2′, C3/C3′, C7/C7′ and C10/C10′. This was attributed to the engagement of these atoms in the intermolecular interactions with water molecules and chloride anions, which are much stronger than in DMSO solution. Linear pentamidine analogs are molecules of pharmaceutical interest which are rather poorly represented in the crystallographic database: chemical shifts from NMR solid-state spectra of the analyzed compounds have provided valuable information on the solid-state structure to complement the crystallographic data, allowing their hydrogen-bonding patterns to be determined.

Table 5
Differences (p.p.m.) between selected ¹³C chemical shifts in the solid state and in solution for (1) and (2): (Δ = δ_solution − δ_solid)

	(1)	(2)
No.	Δ	Δ
C2 (C2′)	0.5	−0.8
C3 (C3′)	0.5	−4.1
C6 (C6′)	1.5	2.5
C7	−2.0	−1.0
C7′	−2.0	−1.0
C8	−0.1	0.2
C8′	−0.1	0.2
C9	1.0	1.4
C9′	1.0	1.4
C10	−2.5	−0.6
C10′	−2.5	−0.6

Figure 6 ^13C CP/MAS NMR spectra of compounds (1) and (2). Sidebands are marked with an asterisk.

4. Conclusions

The structures of 1,4-bis(4-cyanobenzyl)piperazine (1) and 1,4-bis(4-amidinobenzyl)piperazine tetrahydrochloride (2) at 293 and 130 K were solved using single-crystal X-ray diffraction. Compound (1) crystallizes in the triclinic space group, and compound (2) in the monoclinic space group P2₁/n with four chloride anions and four H₂O molecules. The crystal lattice of (1) is formed by weak C—H⋯N and C—H⋯π interactions. The intermolecular interaction energy (evaluated using the equation: E_Hbond = ½(E_cluster − E_two − E_one) for the former was −8.8 kJ mol⁻¹. The crystal lattice formed by (2) is dominated by water molecules and chloride anions, and the average intermolecular interaction energy of compound (2), taking account of all the surrounding chloride anions and water molecules was −237.3 kJ mol⁻¹. It is interesting to note that no direct hydrogen bonds exist between neighboring molecules of (2). Our result clearly indicated that structural information and intermolecular interactions in bis-nitriles are not transferable to the structural analysis of bis-amidines. The computation of shielding constants for isolated molecules together with the solid-state spectrum are of considerable value in understanding the solid-state structures of pentamidine analogs.

Supporting information

Computing details

Data collection: CrysAlis PRO (Agilent Technologies, 2011) for mr-1a; CrysAlis PRO (Agilent Technologies, 2012) for pnt2. Cell refinement: CrysAlis PRO (Agilent Technologies, 2011) for mr-1a; CrysAlis PRO (Agilent Technologies, 2012) for pnt2. Data reduction: CrysAlis PRO (Agilent Technologies, 2011) for mr-1a; CrysAlis PRO (Agilent Technologies, 2012) for pnt2. For both compounds, program(s) used to solve structure: SHELXS97 (Sheldrick, 1990); program(s) used to refine structure: SHELXL97 (Sheldrick, 1997); molecular graphics: Mercury (Macrae et al., 2006); software used to prepare material for publication: SHELXL97 (Sheldrick,1997).

Figures

(mr-1a) 1,4-bis(4-cyanobenzyl)piperazine

Crystal data

C20H20N4	Z = 1
Mr = 316.40	F(000) = 168
Triclinic, P1	Dx = 1.211 Mg m^−3
a = 6.6267 (4) Å	Mo Kα radiation, λ = 0.71073 Å
b = 8.3540 (5) Å	Cell parameters from 3570 reflections
c = 8.6889 (5) Å	θ = 2.5–26.9°
α = 83.876 (5)°	µ = 0.07 mm^−1
β = 72.881 (5)°	T = 293 K
γ = 70.676 (5)°	Block, colourless
V = 433.78 (4) Å^3	0.5 × 0.4 × 0.35 mm

Data collection

Xcalibur, Sapphire2, large Be window diffractometer	1625 independent reflections
Radiation source: Enhance (Mo) X-ray Source	1409 reflections with I > 2σ(I)
Graphite monochromator	Rint = 0.010
Detector resolution: 8.1929 pixels mm^-1	θmax = 25.7°, θmin = 4.7°
ω scans	h = −8→8
Absorption correction: multi-scan CrysAlis PRO, Agilent Technologies, Version 1.171.35.15 (release 03-08-2011 CrysAlis171 .NET) (compiled Aug 3 2011,13:03:54) Empirical absorption correction using spherical harmonics, implemented in SCALE3 ABSPACK scaling algorithm.	k = −10→10
Tmin = 0.965, Tmax = 1.000	l = −10→10
6352 measured reflections

Refinement

Refinement on F^2	Secondary atom site location: difference Fourier map
Least-squares matrix: full	Hydrogen site location: inferred from neighbouring sites
R[F^2 > 2σ(F^2)] = 0.039	H atoms treated by a mixture of independent and constrained refinement
wR(F^2) = 0.118	w = 1/[σ^2(Fo^2) + (0.0601P)^2 + 0.0584P] where P = (Fo^2 + 2Fc^2)/3
S = 1.05	(Δ/σ)max = 0.016
1625 reflections	Δρmax = 0.13 e Å^−3
114 parameters	Δρmin = −0.12 e Å^−3
0 restraints	Extinction correction: SHELXL, Fc^*=kFc[1+0.001xFc^2λ^3/sin(2θ)]^-1/4
Primary atom site location: structure-invariant direct methods	Extinction coefficient: 0.078 (14)

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
N1	0.10408 (16)	0.86797 (12)	0.59789 (11)	0.0469 (3)
N2	0.7613 (2)	0.70453 (18)	1.21965 (17)	0.0819 (5)
C1	0.3807 (2)	0.71994 (15)	0.75000 (14)	0.0498 (3)
C2	0.5930 (2)	0.61331 (17)	0.74366 (16)	0.0574 (4)
H2A	0.6718	0.5422	0.6558	0.069*
C3	0.6906 (2)	0.60984 (17)	0.86392 (17)	0.0590 (4)
H3A	0.842 (3)	0.530 (2)	0.8587 (19)	0.077 (4)*
C4	0.5760 (2)	0.71739 (16)	0.99367 (15)	0.0510 (3)
C5	0.3641 (2)	0.82650 (19)	1.00210 (16)	0.0638 (4)
H5A	0.2869	0.8996	1.0887	0.077*
C6	0.2682 (2)	0.82620 (19)	0.88145 (16)	0.0650 (4)
H6A	0.1251	0.8988	0.8882	0.078*
C7	0.6780 (2)	0.71263 (17)	1.12006 (17)	0.0608 (4)
C8	0.2781 (2)	0.71262 (16)	0.61823 (17)	0.0598 (4)
H8A	0.3946	0.6879	0.5174	0.072*
H8B	0.2160	0.6198	0.6416	0.072*
C9	−0.0144 (2)	0.83736 (15)	0.49172 (15)	0.0523 (3)
H9A	−0.0776	0.7475	0.5368	0.063*
H9B	0.0891	0.8011	0.3872	0.063*
C10	0.19728 (19)	1.00347 (15)	0.52785 (14)	0.0499 (3)
H10A	0.3028	0.9687	0.4237	0.060*
H10B	0.2759	1.0257	0.5971	0.060*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
N1	0.0542 (6)	0.0441 (5)	0.0480 (5)	−0.0137 (4)	−0.0255 (4)	0.0031 (4)
N2	0.0873 (9)	0.0795 (9)	0.0863 (9)	−0.0082 (7)	−0.0524 (8)	−0.0101 (7)
C1	0.0557 (7)	0.0434 (6)	0.0484 (7)	−0.0072 (5)	−0.0218 (5)	0.0026 (5)
C2	0.0557 (7)	0.0515 (7)	0.0584 (7)	−0.0031 (5)	−0.0182 (6)	−0.0116 (6)
C3	0.0474 (7)	0.0564 (7)	0.0683 (8)	−0.0029 (6)	−0.0225 (6)	−0.0053 (6)
C4	0.0524 (7)	0.0503 (7)	0.0510 (7)	−0.0112 (5)	−0.0229 (5)	0.0050 (5)
C5	0.0611 (8)	0.0684 (9)	0.0482 (7)	0.0061 (6)	−0.0217 (6)	−0.0098 (6)
C6	0.0542 (7)	0.0712 (9)	0.0546 (7)	0.0107 (6)	−0.0251 (6)	−0.0085 (6)
C7	0.0611 (8)	0.0564 (8)	0.0659 (8)	−0.0087 (6)	−0.0301 (7)	−0.0006 (6)
C8	0.0726 (8)	0.0479 (7)	0.0593 (8)	−0.0051 (6)	−0.0337 (7)	−0.0037 (6)
C9	0.0652 (8)	0.0479 (7)	0.0558 (7)	−0.0233 (6)	−0.0301 (6)	0.0048 (5)
C10	0.0499 (6)	0.0568 (7)	0.0508 (7)	−0.0201 (5)	−0.0222 (5)	0.0011 (5)

Geometric parameters (Å, º)

N1—C8	1.4582 (15)	C4—C7	1.4403 (16)
N1—C10	1.4587 (16)	C5—C6	1.3769 (17)
N1—C9	1.4636 (14)	C5—H5A	0.9300
N2—C7	1.1411 (17)	C6—H6A	0.9300
C1—C2	1.3826 (17)	C8—H8A	0.9700
C1—C6	1.3875 (18)	C8—H8B	0.9700
C1—C8	1.5090 (16)	C9—C10^i	1.5074 (17)
C2—C3	1.3749 (18)	C9—H9A	0.9700
C2—H2A	0.9300	C9—H9B	0.9700
C3—C4	1.3840 (19)	C10—C9^i	1.5074 (17)
C3—H3A	0.993 (16)	C10—H10A	0.9700
C4—C5	1.3837 (17)	C10—H10B	0.9700
C8—N1—C10	111.12 (10)	C1—C6—H6A	119.3
C8—N1—C9	109.95 (9)	N2—C7—C4	178.26 (14)
C10—N1—C9	108.61 (9)	N1—C8—C1	114.29 (10)
C2—C1—C6	117.90 (11)	N1—C8—H8A	108.7
C2—C1—C8	119.29 (11)	C1—C8—H8A	108.7
C6—C1—C8	122.78 (11)	N1—C8—H8B	108.7
C3—C2—C1	121.65 (12)	C1—C8—H8B	108.7
C3—C2—H2A	119.2	H8A—C8—H8B	107.6
C1—C2—H2A	119.2	N1—C9—C10^i	110.36 (10)
C2—C3—C4	119.53 (11)	N1—C9—H9A	109.6
C2—C3—H3A	120.4 (9)	C10^i—C9—H9A	109.6
C4—C3—H3A	120.0 (9)	N1—C9—H9B	109.6
C3—C4—C5	119.95 (11)	C10^i—C9—H9B	109.6
C3—C4—C7	119.41 (11)	H9A—C9—H9B	108.1
C5—C4—C7	120.63 (12)	N1—C10—C9^i	110.14 (9)
C6—C5—C4	119.56 (12)	N1—C10—H10A	109.6
C6—C5—H5A	120.2	C9^i—C10—H10A	109.6
C4—C5—H5A	120.2	N1—C10—H10B	109.6
C5—C6—C1	121.40 (12)	C9^i—C10—H10B	109.6
C5—C6—H6A	119.3	H10A—C10—H10B	108.1
C6—C1—C2—C3	0.8 (2)	C8—C1—C6—C5	178.13 (13)
C8—C1—C2—C3	−177.26 (12)	C10—N1—C8—C1	71.10 (14)
C1—C2—C3—C4	−1.1 (2)	C9—N1—C8—C1	−168.63 (10)
C2—C3—C4—C5	0.4 (2)	C2—C1—C8—N1	−156.04 (12)
C2—C3—C4—C7	179.55 (12)	C6—C1—C8—N1	26.00 (19)
C3—C4—C5—C6	0.5 (2)	C8—N1—C9—C10^i	179.05 (10)
C7—C4—C5—C6	−178.64 (13)	C10—N1—C9—C10^i	−59.17 (14)
C4—C5—C6—C1	−0.8 (2)	C8—N1—C10—C9^i	−179.90 (9)
C2—C1—C6—C5	0.1 (2)	C9—N1—C10—C9^i	59.03 (14)

Symmetry code: (i) −x, −y+2, −z+1.

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
C3—H3A···N2^ii	0.993 (16)	2.654 (17)	3.6273 (19)	166.5 (13)

Symmetry code: (ii) −x+2, −y+1, −z+2.

(pnt2) 1,4-bis(4-amidinobenzyl)piperazine tetrahydrochloride

Crystal data

C20H30N6·4(Cl)·4(H2O)	F(000) = 600
Mr = 568.36	Dx = 1.327 Mg m^−3
Monoclinic, P21/n	Cu Kα radiation, λ = 1.54184 Å
a = 6.1121 (2) Å	Cell parameters from 1496 reflections
b = 12.8231 (3) Å	θ = 2.4–61.6°
c = 18.3831 (4) Å	µ = 4.08 mm^−1
β = 99.274 (2)°	T = 130 K
V = 1421.96 (7) Å^3	Needle, colourless
Z = 2	0.5 × 0.1 × 0.08 mm

Data collection

SuperNova, Single source at offset), Atlas diffractometer	2866 independent reflections
Radiation source: SuperNova (Cu) X-ray Source	2771 reflections with I > 2σ(I)
Mirror monochromator	Rint = 0.015
ω scans	θmax = 75.4°, θmin = 4.2°
Absorption correction: multi-scan CrysAlis PRO, Agilent Technologies, Version 1.171.36.20 (release 27-06-2012 CrysAlis171 .NET) (compiled Jul 11 2012,15:38:31) Empirical absorption correction using spherical harmonics, implemented in SCALE3 ABSPACK scaling algorithm.	h = −4→7
Tmin = 0.632, Tmax = 1.000	k = −14→15
6224 measured reflections	l = −22→19

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)] = 0.027	Hydrogen site location: mixed
wR(F^2) = 0.073	All H-atom parameters refined
S = 1.04	w = 1/[σ^2(Fo^2) + (0.0403P)^2 + 0.5773P] where P = (Fo^2 + 2Fc^2)/3
2866 reflections	(Δ/σ)max = 0.001
230 parameters	Δρmax = 0.25 e Å^−3
0 restraints	Δρmin = −0.21 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
C1	0.4602 (2)	0.25750 (9)	0.10510 (7)	0.0191 (2)
C2	0.3358 (2)	0.33382 (10)	0.06331 (7)	0.0243 (3)
H2A	0.363 (3)	0.3484 (13)	0.0133 (9)	0.027 (4)*
C3	0.1766 (2)	0.38987 (10)	0.09256 (7)	0.0242 (3)
H3A	0.090 (3)	0.4387 (15)	0.0638 (10)	0.035 (5)*
C4	0.1431 (2)	0.37163 (9)	0.16469 (7)	0.0188 (2)
C5	0.2698 (2)	0.29650 (10)	0.20730 (7)	0.0194 (2)
H5A	0.253 (3)	0.2858 (12)	0.2573 (9)	0.021 (4)*
C6	0.4263 (2)	0.23956 (10)	0.17714 (7)	0.0197 (2)
H6A	0.515 (3)	0.1891 (14)	0.2048 (9)	0.027 (4)*
C7	−0.0282 (2)	0.43176 (10)	0.19541 (7)	0.0204 (3)
C8	0.6300 (2)	0.19590 (10)	0.07227 (7)	0.0207 (3)
H8A	0.739 (3)	0.1636 (13)	0.1107 (9)	0.024 (4)*
H8B	0.701 (3)	0.2381 (13)	0.0408 (9)	0.023 (4)*
C9	0.4224 (2)	0.02784 (10)	0.06722 (7)	0.0197 (2)
H9A	0.316 (3)	0.0590 (13)	0.0918 (9)	0.025 (4)*
H9B	0.542 (3)	0.0025 (12)	0.1037 (9)	0.022 (4)*
C10	0.6855 (2)	0.05874 (10)	−0.01800 (7)	0.0195 (2)
H10A	0.808 (3)	0.0340 (12)	0.0180 (8)	0.018 (4)*
H10B	0.732 (3)	0.1131 (14)	−0.0483 (9)	0.026 (4)*
N1	0.52109 (17)	0.10900 (8)	0.02358 (6)	0.0178 (2)
N2	−0.0708 (2)	0.52765 (9)	0.17266 (7)	0.0247 (2)
H2B	−0.174 (3)	0.5625 (16)	0.1866 (10)	0.038 (5)*
H2C	0.015 (3)	0.5567 (16)	0.1460 (11)	0.042 (5)*
N3	−0.13564 (19)	0.38704 (9)	0.24356 (6)	0.0238 (2)
H3B	−0.238 (3)	0.4209 (14)	0.2605 (9)	0.029 (4)*
H3C	−0.117 (3)	0.3209 (15)	0.2551 (9)	0.026 (4)*
Cl1	1.00818 (5)	0.00448 (2)	0.175051 (17)	0.02406 (10)
Cl2	−0.20790 (5)	0.32769 (3)	−0.080094 (18)	0.02796 (10)
O1	1.19156 (19)	0.17886 (9)	−0.07956 (7)	0.0365 (3)
O2	0.90505 (17)	0.16698 (8)	0.28880 (6)	0.0255 (2)
H1	0.408 (3)	0.1363 (16)	−0.0085 (10)	0.038 (5)*
H2AW	1.001 (4)	0.1687 (18)	0.3271 (13)	0.055 (6)*
H2BW	0.961 (4)	0.1271 (18)	0.2587 (12)	0.052 (6)*
H1AW	1.127 (4)	0.1378 (18)	−0.1093 (12)	0.046 (6)*
H1BW	1.093 (4)	0.223 (2)	−0.0764 (13)	0.065 (7)*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
C1	0.0206 (6)	0.0146 (5)	0.0229 (6)	−0.0007 (5)	0.0057 (5)	−0.0026 (4)
C2	0.0351 (7)	0.0188 (6)	0.0204 (6)	0.0050 (5)	0.0091 (5)	0.0004 (5)
C3	0.0317 (7)	0.0187 (6)	0.0221 (6)	0.0076 (5)	0.0045 (5)	0.0009 (5)
C4	0.0200 (6)	0.0146 (6)	0.0221 (6)	−0.0016 (4)	0.0048 (5)	−0.0047 (4)
C5	0.0214 (6)	0.0183 (6)	0.0190 (6)	−0.0020 (5)	0.0051 (5)	−0.0010 (5)
C6	0.0193 (6)	0.0172 (6)	0.0224 (6)	0.0008 (5)	0.0031 (5)	0.0011 (5)
C7	0.0199 (6)	0.0186 (6)	0.0226 (6)	−0.0003 (5)	0.0029 (5)	−0.0045 (5)
C8	0.0210 (6)	0.0165 (6)	0.0260 (6)	−0.0012 (5)	0.0082 (5)	−0.0027 (5)
C9	0.0234 (6)	0.0156 (6)	0.0228 (6)	0.0002 (5)	0.0124 (5)	0.0001 (5)
C10	0.0205 (6)	0.0154 (6)	0.0257 (6)	0.0003 (5)	0.0127 (5)	−0.0006 (5)
N1	0.0189 (5)	0.0139 (5)	0.0225 (5)	0.0009 (4)	0.0092 (4)	−0.0005 (4)
N2	0.0270 (6)	0.0184 (5)	0.0314 (6)	0.0044 (5)	0.0124 (5)	−0.0008 (5)
N3	0.0237 (6)	0.0192 (6)	0.0309 (6)	0.0012 (4)	0.0118 (5)	−0.0021 (5)
Cl1	0.02477 (17)	0.02354 (17)	0.02534 (17)	−0.00187 (11)	0.00850 (12)	0.00067 (11)
Cl2	0.02792 (18)	0.02546 (18)	0.03057 (18)	0.00152 (12)	0.00495 (13)	0.00601 (12)
O1	0.0331 (6)	0.0279 (6)	0.0443 (7)	0.0077 (5)	−0.0063 (5)	−0.0080 (5)
O2	0.0254 (5)	0.0246 (5)	0.0271 (5)	−0.0031 (4)	0.0063 (4)	−0.0005 (4)

Geometric parameters (Å, º)

C1—C2	1.3920 (18)	C9—N1	1.4987 (15)
C1—C6	1.3922 (17)	C9—C10^i	1.5145 (18)
C1—C8	1.5056 (16)	C9—H9A	0.939 (17)
C2—C3	1.3862 (19)	C9—H9B	0.966 (17)
C2—H2A	0.977 (17)	C10—N1	1.5022 (14)
C3—C4	1.3938 (18)	C10—C9^i	1.5145 (18)
C3—H3A	0.928 (19)	C10—H10A	0.970 (15)
C4—C5	1.3949 (18)	C10—H10B	0.964 (17)
C4—C7	1.4835 (17)	N1—H1	0.90 (2)
C5—C6	1.3877 (17)	N2—H2B	0.85 (2)
C5—H5A	0.952 (16)	N2—H2C	0.86 (2)
C6—H6A	0.939 (18)	N3—H3B	0.859 (19)
C7—N2	1.3112 (18)	N3—H3C	0.877 (19)
C7—N3	1.3155 (17)	O1—H1AW	0.81 (2)
C8—N1	1.5146 (16)	O1—H1BW	0.84 (3)
C8—H8A	0.979 (17)	O2—H2AW	0.84 (2)
C8—H8B	0.946 (17)	O2—H2BW	0.86 (2)
C2—C1—C6	119.30 (11)	N1—C9—C10^i	111.30 (10)
C2—C1—C8	119.87 (11)	N1—C9—H9A	109.4 (10)
C6—C1—C8	120.83 (11)	C10^i—C9—H9A	109.3 (10)
C3—C2—C1	120.37 (12)	N1—C9—H9B	106.3 (10)
C3—C2—H2A	120.0 (10)	C10^i—C9—H9B	112.1 (9)
C1—C2—H2A	119.6 (10)	H9A—C9—H9B	108.4 (14)
C2—C3—C4	120.13 (12)	N1—C10—C9^i	110.86 (10)
C2—C3—H3A	120.1 (11)	N1—C10—H10A	107.5 (9)
C4—C3—H3A	119.8 (11)	C9^i—C10—H10A	112.4 (9)
C3—C4—C5	119.78 (11)	N1—C10—H10B	105.6 (10)
C3—C4—C7	119.73 (11)	C9^i—C10—H10B	109.0 (10)
C5—C4—C7	120.49 (11)	H10A—C10—H10B	111.3 (13)
C6—C5—C4	119.67 (11)	C9—N1—C10	109.52 (9)
C6—C5—H5A	120.0 (10)	C9—N1—C8	111.77 (10)
C4—C5—H5A	120.3 (10)	C10—N1—C8	110.50 (9)
C5—C6—C1	120.73 (12)	C9—N1—H1	106.7 (12)
C5—C6—H6A	121.4 (10)	C10—N1—H1	109.8 (12)
C1—C6—H6A	117.8 (10)	C8—N1—H1	108.5 (12)
N2—C7—N3	121.90 (12)	C7—N2—H2B	121.1 (13)
N2—C7—C4	118.81 (12)	C7—N2—H2C	118.8 (14)
N3—C7—C4	119.28 (12)	H2B—N2—H2C	120.0 (19)
C1—C8—N1	110.93 (10)	C7—N3—H3B	119.6 (12)
C1—C8—H8A	111.3 (10)	C7—N3—H3C	121.7 (11)
N1—C8—H8A	107.3 (10)	H3B—N3—H3C	118.3 (16)
C1—C8—H8B	110.9 (10)	H1AW—O1—H1BW	102 (2)
N1—C8—H8B	105.1 (10)	H2AW—O2—H2BW	105 (2)
H8A—C8—H8B	111.1 (14)
C6—C1—C2—C3	−1.3 (2)	C5—C4—C7—N2	148.80 (13)
C8—C1—C2—C3	178.98 (12)	C3—C4—C7—N3	147.18 (13)
C1—C2—C3—C4	1.3 (2)	C5—C4—C7—N3	−32.38 (18)
C2—C3—C4—C5	−0.2 (2)	C2—C1—C8—N1	−80.34 (14)
C2—C3—C4—C7	−179.79 (12)	C6—C1—C8—N1	99.97 (13)
C3—C4—C5—C6	−0.85 (18)	C10^i—C9—N1—C10	−57.07 (14)
C7—C4—C5—C6	178.72 (11)	C10^i—C9—N1—C8	−179.86 (10)
C4—C5—C6—C1	0.83 (19)	C9^i—C10—N1—C9	56.80 (14)
C2—C1—C6—C5	0.26 (19)	C9^i—C10—N1—C8	−179.65 (10)
C8—C1—C6—C5	179.94 (11)	C1—C8—N1—C9	−66.53 (13)
C3—C4—C7—N2	−31.64 (18)	C1—C8—N1—C10	171.23 (10)

Symmetry code: (i) −x+1, −y, −z.

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
C8—H8A···Cl1	0.979 (17)	2.766 (17)	3.6783 (14)	155.2 (13)
C6—H6A···O2	0.939 (18)	2.639 (17)	3.4216 (16)	141.2 (13)
O2—H2BW···Cl1	0.86 (2)	2.25 (2)	3.0882 (11)	163 (2)
C8—H8B···Cl2^ii	0.946 (17)	2.641 (17)	3.5465 (13)	160.6 (13)
C10—H10B···Cl2^ii	0.964 (17)	2.849 (18)	3.7231 (13)	151.3 (13)
O1—H1BW···Cl2^ii	0.84 (3)	2.27 (3)	3.0977 (12)	171 (2)
N3—H3C···O2^iii	0.877 (19)	2.066 (19)	2.9411 (15)	174.7 (16)
C9—H9A···Cl1^iii	0.939 (17)	2.701 (17)	3.4710 (12)	139.7 (13)
N1—H1···O1^iii	0.90 (2)	1.79 (2)	2.6875 (16)	173.2 (18)
O1—H1AW···Cl1^iv	0.81 (2)	2.27 (2)	3.0673 (12)	167 (2)
N2—H2C···Cl2^v	0.86 (2)	2.35 (2)	3.1894 (13)	166.4 (19)
O2—H2AW···Cl2^vi	0.84 (2)	2.26 (2)	3.0939 (11)	173 (2)
N2—H2B···O2^vii	0.85 (2)	2.05 (2)	2.8870 (15)	170.4 (19)
N3—H3B···Cl1^vii	0.859 (19)	2.432 (19)	3.2847 (12)	171.8 (15)

Symmetry codes: (ii) x+1, y, z; (iii) x−1, y, z; (iv) −x+2, −y, −z; (v) −x, −y+1, −z; (vi) x+3/2, −y+1/2, z+1/2; (vii) −x+1/2, y+1/2, −z+1/2.

 

Footnotes

¹Supporting information for this paper is available from the IUCr electronic archives (Reference: BM5065 )