Crystal structures and hydrogen bonding in the proton-transfer salts of nicotine with 3,5-dinitrosalicylic acid and 5-sulfosalicylic acid

The crystal structures of the 1:1 salts of nicotine with 3,5-dinitrosalicylic acid and with 5-sulfosalicylic acid both show polymeric hydrogen-bonded and π–π-bonded structures but these differ in that in the first example, cations and anions form separate cation chains or anion columns which are unassociated through formal hydrogen bonds while in the second, hydrogen-bonded cation–anion chains are found.


Chemical context
Nicotine [3-(2S-1-methylpyrrolidin-2-yl)pyridine] is well known as a toxic liquid alkaloid which is found in the leaves of the tobacco plants Nicotiana tabacum and N. rustica (Rodgman & Parfetti, 2009). Because of these properties, nicotine and its compounds have been of commercial interest and have been used in the past as insecticides and as veterinary ectoparasiticides (usually as the sulfate) (Ujvá ry, 1999), as well as in limited medical applications as the bitartrate (Eudermol) for the treatment of smoking-withdrawal syndrome (Enzell et al., 1977). However, its veterinary use is restricted due to its toxicity with even topical applications, resulting in the total ban on its use in the USA early in 2014.

Structural commentary
In both the nicotinium salts of DNSA (I) and 5-SSA (II), proton-transfer to the pyrrolidine N-atom of nicotine has occurred as expected, generating an N11(R) chiral centre relative to the known C21(S) centre. Also, in both (I) and (II) (Figs. 1 and 2), the asymmetric units comprise two indepen-dent NIC + cations (C and D) and either, for (I), two DNSA phenolate monoanions or two 5-SSA carboxylate monoanions (A and B) (Figs. 1, 2). With (II), the two independent anion and cation pairs are pseudo-centrosymmetrically related but the presence of the inversion centre is obviated by the fact that both of the NIC cations have the same N11(R), C21(S) absolute configuration.
In (I), the nicotinium C and D cations are conformationally similar but in (II), they are different. However, in both, the pyrrolidinium plane is significantly rotated with respect to that of the benzene ring [the torsion angles C2C/D-C3C/D-C21C/D-N11C/D are À71.9 (4) (C) and À68.8 (4) (D) in (I) and À45.7 (4) (C) and 125.7 (3) (D) in (II)]. This conformation with the two rings antiplanar is usual for cationic nicotine structures, e.g. Arnaud et al. (2007). The substituent carboxyl and nitro groups of the DNSA anions in (I) are essentially coplanar with the benzene ring, with the maximum deviation among the three defining torsion angles for each anion (C2A/B-C1A/B-C11A/B-O2A/B, C2A/B -C3A/B-N3A/ B-O32A/B and C4A/B-C5A/B-N5A/B-O52A/B) being for the C3B nitro group [173.7 (3) ]. In the B anion, there is 25% rotational disorder about the C1Á Á ÁC4 ring vector, which generates a second phenolic O-component (O21B). This phenomenon has precedence in DNSA salt structures, e.g. with the nicotinamide salt (Koman et al., 2003;24% disorder). The C3 nitro group is most often associated with deviation from planarity in the DNSA phenolate salts (Smith et al., 2007) and is the more interactive and sterically crowded group. In the case of (I), the uncommon planarity is probably associated with the presence of anion -bonding associations.
With the 5-SSA anions, the carboxylic acid group is essentially coplanar with the benzene ring, which is expected in this salicylic acid species, invariably having the short intramolecular carboxylic acid O-HÁ Á ÁO phenol hydrogen bond (Table 2) (Smith et al., 2006). This interaction is also present in the phenolate anion in (I) in which the carboxylic acid H-atom is anti-related (Table 1). The molecular conformation and atom labelling for the two NIC cations (C and D) and the two DNSA anions (A and B) in the asymmetric unit of (I), with displacement ellipsoids drawn at the 40% probability level.
Inter-species hydrogen bonds are shown as dashed lines (see Table 1).

Figure 2
The molecular conformation and atom labelling for the two NIC cations (C and D) and the two 5-SSA anions (A and B) in the asymmetric unit of (II), with displacement ellipsoids drawn at the 40% probability level.
Inter-species hydrogen bonds are shown as dashed lines (see Table 2).

Supramolecular features
In the supramolecular structure of (I), the two independent NIC cations C and D interact through N1C -HÁ Á ÁN11D i and N1D -HÁ Á ÁN11C hydrogen bonds (Table 1), giving zigzag chains extending along a (Fig. 3). With the DNSA anions, there are no formal hydrogen-bonding interactions either between A and B anions or with the NIC chain structures. Instead, these anions form --bonded stacks which are parallel to the NIC + chains down a [ring-centroid separation = 3.857 (2) Å ]. The presence ofstacking is unusual in DNSA cation structures. In the crystal, there are a number of intermolecular CC/D-HÁ Á ÁOA/B hydrogen-bonding interactions, which give an overall three-dimensional structure.

Figure 3
The alternating hydrogen-bonded C-D cationic columns and -bonded A-B anion stacks in the structure of (I), viewed along the stacks in the unit cell.

Figure 4
The hydrogen-bonded A-C and B-D chain structures in (II), extending along b. Non-associative H atoms have been omitted. For symmetry codes, see Table 2.

Synthesis and crystallization
The title salts (I) and (II) were prepared by refluxing equimolar quantities of nicotine (160 mg) and the respective acids, 3,5-dinitrosalicylic acid (230 mg) for (I) or 3-carboxy-4-hydroxybenzenesulfonic acid (220 mg) for (II) in 30 ml of ethanol for 10 min, after which room temperature evaporation of the solutions gave, for (I), thin yellow needles and for (II) colourless prisms, from which specimens were cleaved for the X-ray analyses.

Refinement details
Crystal data, data collection and structure refinement details are summarized in Table 3. H atoms on all potentially interactive O-H and N-H groups in all molecular species, were located by difference-Fourier methods but these and the carbon-bound H-atoms were subsequently set as riding on the parent atoms in the refinement in calculated positions [O-H = 0.88, N-H = 0.94, C-H = 0.95-1.00 Å ] and with U iso (H) = 1.5U eq (O or methyl-C) or 1.2U eq (C, N). The site occupancy factors for the rotationally disordered phenolate components (O2B) and its other component (O21B) in anion B of (I) were determined as 0.752 (4): 0.248 (4) and were subsequently set at 0.75:0.25 in the refinement.
In both structures, the known C21(S) absolute configuration was invoked. The Flack parameter for (I) [0.2 (16)] has no physical meaning. The absolute structure of compound (II) was confirmed by resonant scattering [Flack parameter = À0.02 (9)]. Crystal structures and hydrogen bonding in the proton-transfer salts of nicotine with 3,5-dinitrosalicylic acid and 5-sulfosalicylic acid

Special details
Geometry. Bond distances, angles etc. have been calculated using the rounded fractional coordinates. All su's are estimated from the variances of the (full) variance-covariance matrix. The cell e.s.d.'s are taken into account in the estimation of distances, angles and torsion angles Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.

(II) (1R,2S)-1-Methyl-2-(pyridin-3-yl)pyrrolidin-1-ium 3-carboxy-4-hydroxybenzenesulfonate
Crystal data where P = (F o 2 + 2F c 2 )/3 (Δ/σ) max < 0.001 Δρ max = 0.49 e Å −3 Δρ min = −0.36 e Å −3 Absolute structure: Flack (1983), 4361 Friedel pairs Absolute structure parameter: −0.02 (9) Special details Geometry. Bond distances, angles etc. have been calculated using the rounded fractional coordinates. All su's are estimated from the variances of the (full) variance-covariance matrix. The cell e.s.d.'s are taken into account in the estimation of distances, angles and torsion angles Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.