All chemicals and solvents were obtained from commercial sources and used as received.

To a 10 mL aqueous solution of Zn(CH₃COO)₂·2H₂O (0.367 g; 2.0 mmol) was added a 5 mL MeOH solution of 3-Ampy (0.188 g; 2.0 mmol). The mixture was stirred at ambient temperature for 15 minutes. To the mixture was added a 5 mL MeOH solution of KSCN (0.389 g; 4 mmol). The resulting mixture was further stirred for 30 minutes after which a white precipitate was obtained. The precipitate was filtered, washed with MeOH, air-dried, and weighed. The filtrate was allowed in a refrigerator at 4˚C and colourless crystals were obtained. [Zn(3-ampy)₂(SCN)₂]: yield: 72.4%. Element Anal. Calc. for C₁₂H₁₂N₆S₂Zn: C, 38.94; H, 3.25; N, 22.72. Found: C, 39.03; H, 2.93; N, 22.54. IR (KBr dics cm⁻¹): 3449 (m), 3347 (m), 2109 (sh), 2089 (s), 1623 (m), 1581 (m), 1494 (m), 1449 (m), 1354 (m), 1137 (m), 800 (m), 697 (m) 548 (w).

IR spectra: KBr pellets were prepared in a nitrogen-filled glove box and the spectra were recorded on a PerkinElmer System 2000 FTIR spectrometer in the range 400 - 4000 cm^–1. Elemental analyses for C, H, and N were performed on a FlashEA1112 element analyser. Thermogravimetric (TG) and differential thermal analysis (DTA) curves, coupled with mass spectrometry, were obtained using a NETZSCH STA449F1 thermoanalyzer in a dynamic argon atmosphere (heating rate 10˚C·min^–1, flow rate 25 mL·min^–1, aluminium oxide crucible, mass 20 mg, and temperature range from room temperature up to 900˚C).

X-ray data were collected with a GEMINI CCD diffractometer (Rigaku Inc.), λ(Mo-K_α) = 0.71073 Å, T = 130(2) K, ω-scan rotation. Data reduction was performed with CrysAlis Pro including the program SCALE3 ABSPACK for empirical absorption correction [14]. All structures were solved by dual space methods with SHELXT-20xy [25]. Structure refinement was done with SHELXL-2018 [26] [27] by using full-matrix least-square routines against F². Hydrogen atoms and NH were located on difference Fourier maps calculated at the final stage of the structure refinement. The remaining hydrogen atoms of (2) were calculated on idealised positions using the riding model. The pictures were generated with the program Mercury [28]. CCDC 2075665 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via https://www.ccdc.cam.ac.uk/data_request/cif (or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44)1223-336-033; or deposit@ccdc.cam.uk).

Calculations were performed on the complexes to gain further insights into the molecular interactions in the synthesized compound and to predict some of their electronic properties. A single molecule as well as a dimer of the complex was optimized without any symmetric constraints. The calculations on the single molecule were performed using DFT methods and two exchange-correlations functional [29]: the B3LYP that associates the Becke three parameter exchange functional [30] [31] and the Lee-Yan-Parr correlation functional [32] and the wB97XD which integrates a long-range correction term [33] [34]. wB97XD has the advantage being able to model with accuracy short-range and long-range interactions. The molecular orbitals of the studied molecule were described by a polarized and diffused triple split valence orbital Pople basis set 6-311++G(2d,2p) [35]. This basis set has been polarized for all atoms to improve the flexibility of molecular orbitals [36]. The optimization process was based on the X-ray structure data, and making use of ultrafine integral and tight optimization criteron. The harmonic vibrational frequencies of the optimized geometry were calculated to ensure that no imaginary frequency was associated to the optimized molecule. The calculations on the dimer were performed using wB97XD/6-31++G(d,p) level of theory. Gaussian 09 Revision A.02 package [37] was used for all quantum calculations. The non-covalent interactions (NCI) such as hydrogen bond and van der Waals interactions were described by the Contreras-Garcia and collaborators approach referred to as non-covalent interactions index [38] [39]. The multiwfn software [40] was used for this purpose.

Single crystal X-ray structure analysis of the complex (Figure 1) revealed that it is the same as the Zn(3-ampy)₂(SCN)₂ complex previously reported, but with a different orientation of the aminopyridine rings around the metal centre [11]. The complex crystallizes in the orthorhombic space group Pnma with Z = 4 and contains only one crystallographically independent Zn cation that is located, together with the two SCN⁻ ions, on a mirror plane. The Zn cation is tetrahedrally coordinated by two terminal N-bonded thiocyanate anions and two monodentate (through the pyridine N-atom) 3-aminopyridine ligands. The packing diagram is shown in Figure 2. Crystallographic data for the complex complex to that in the literature is presented in Table 1.

An interplay of weak intermolecular interactions and face to face π‒π stacking interactions of coordinated 3‒ampy moieties extend the structure into three dimensions.

The IR spectrum of the complex reveals characteristic bands for 3-ampy and the thiocyanate ligands. These bands are presented in Figure 3. The medium strong vibrations at 3347 and 3449 cm⁻¹ are assigned to ν_s(NH₂) and ν_as(NH₂) vibrations of the NH₂ group, respectively. The medium peak at 1581 cm⁻¹ is assigned to NH₂ deformation vibration [19]. The very strong C ≡ N asymmetric stretching vibrations of the SCN⁻ are observed at 2109 cm⁻¹ in the spectrum of the complex, indicating that it has taken part in bonding [17]. The ν_C=N stretching modes of the pyridine ring shifted from 1595 to 1623 cm⁻¹ in the spectrum of the complex, indicating its participation in bonding. The strong, well resolved and sharp absorption bands found in the region of 1494 - 1000 cm⁻¹ in the spectrum of the complex are assigned to the aryl C-H stretching vibrations of the coordinated pyridine ring. The Zn-Npy stretching mode is present at about 548 cm⁻¹.

The experimental and theoretical values of some frequencies of the Zn-3-ampy complex are presented in Table 2. The values obtained at B3LYP/6-311++G(2d,2p) level of theory have been scaled by 0.944 for NH₂ stretching and bending and Zn-N2 while the others have been scaled by 1.007 [41]. The values obtained at wB97XD/6-311++G(2d,2p) level of theory were scaled by 0.957 (from CCCBDB listing of precalculated vibrational scaling factors (nist.gov)). As shown in Table 2, the scaled and experimental frequencies are different, with the difference less than 44 cm⁻¹. Therefore, there is a good agreement between the experimental and theoretical results.

The scaled frequencies obtained with wB97XD/6-311++G(2d,2p) are over estimated (particularly for NH₂) compared to both experimental and B3LYP/6-311++G(2d,2p) values. However, it is important to note that the defference between the unscaled frequencies obtained at these two levels of theory is not large except for NH₂.

The TG and DTA curves for the complex are shown in Figure 4. Thermal analysis indicates that [Zn(3-ampy)₂(NCS)₂] decomposes in three steps. The first decomposition step, with a mass loss of 26.35%, occurs from 220˚C to 300˚C and is attributed to the release of one 3-ampy molecule (theoretical 25.42%). The second decomposition step from 300˚C to 550˚C (∆m = 25.65%) is attributed to the loss of a second molecule of 3-ampy to afford Zn(SCN)₂. Further decomposition of this intermediate results in a residue of composition ZnS₂ (residual mass: 35.46%, theoretical: 34.62%). This is accompanied by a broad exothermic DTA peak from 550˚C to 800˚C.

The Hirshfeld surface (HS) [39] [42], defined as the electron density boundary surfaces between the molecules in a crystal, are useful to analyse and view the intermolecular contacts. Hirshfeld surface is mapped using the normalized contact distance ( d n o r m ), which is calculated using Equation (1).

d n o r m = d i − r i v d w r i v d w + d e − r e v d w r e v d w . (1)

The 3D d n o r m surface is used for identification of close intermolecular interactions which are mainly responsible for the molecular packing in the crystals. The sign of d n o r m can be used to characterise the intermolecular contact in the crystal. When d n o r m < 0 , the intermolecular contacts are shorter than the van der Waals radii and they are referred to as close molecular interactions. When d n o r m > 0 , intermolecular contacts are longer than the van der Waals radii and when d n o r m = 0 the intermolecular contacts are similar to the van der Waals radii. These three intermolecular interactions are identified on the Hirshfeld surfaces by their colour. Short contacts are represented by the red area, long contacts by the blue and the white area represents contacts with lengths equivalent to the sum of the van der Waal radii of the interacting atoms. The d n o r m 3D Hirshfeld surface of the synthesized molecule was calculated from its X-ray structure using Crystal Explorer 17.5 software and the results are presented in Figure 5. The close contact molecules have been included in the graph to allow the visualization of the atoms involved in the short contacts. The red spots on Figure 5 are observed around sulphur atoms of the thiocyanate and the two hydrogen atoms of the amine groups of the 3-ampy ligand. Therefore, all the close contacts in the crystal occur between sulphur and hydrogen atoms of the amine groups of different molecules. S-H interactions are then, the main interactions responsible for the molecular packing in the crystal. The distance between the sulphur and hydrogen atoms involved in the close contacts are 2.548 Å and 2.816 Å. However, the intensity and the sizes of these red spots are not the same around the two sulphur atoms or the two hydrogen atoms. This indicates that the distance between the atoms involved in the short contact is directly proportional to the intensity and the size of the red spots of HS.

The 2D fingerprint plot [42] [43] can be derived from the 3D Hirshfeld surface. This is a plot of d_i vs d_e at each point of the 3D Hirshfeld surface and it provides a visual summary of the frequency of each combination of d_e and d_i across the surface of a molecule. The 2D fingerprint is used to quantitatively study at the same time, all the molecular contacts in the crystal. It has the advantage that it can be decomposed to highlight particular close contacts. This decomposition enables the separation of the contributions from different interaction types, which commonly overlap in the full fingerprint.

The decomposition of the 2D fingerprint of the studied molecule (presented in Figure 6) reveals that the main interaction that holds molecules together in the crystalline structure of the Zn-3-ampy are X---H (X = H, C, N and S) interactions. These interactions represent 81.2% of all the intermolecular interactions in this crystal. However, their contribution decreases as follows: S-H (30.2%) > H-H (24.3%) > C-H (20.6%) > N-H (6.1). It is important to note that, C-N and C-S interactions have a non-negligible contribution of 8.4% and 4.3%, respectively, to the packing of the crystal.

Molecular packing into crystals involves a broad range of attractive and repulsive interactions. Thus, it is important to know if such interactions are attractive and therefore, stabilizing. NCI analysis in the unit cell was investigated using Multiwfn package and visualized by VMD molecular graphics viewer. Thus, due to the limitation of our computational resources, NCI analysis was performed on two of the four molecules in the unit cell of the complex. The dimer was optimized from the X-ray structure data using the wB97XD functional associated to the 6-31++G(d,p) basis set with ultrafine integral (see data in Tables S1 and S2, electronic supplementary material (ESM)). The wB97XD was chosen because of its capacity to handle weaker NCI [44]. The NCI, the reduced density gradient, was plotted as a function of the density (mapped as isosurfaces) over the molecule of interest. The nature of the specific interactions was identified as follows: strong attractive interaction is indicated in blue whereas red indicates a strong repulsive interaction. Weak interactions such as van der Waals interactions are highlighted by a green isosurface.

The NCI plot of the Zn-3-ampy complex (Figure 7) shows the presence of a very large green surface between the two molecules indicating that the two molecules are held together by van der Waals interactions. In addition, the small red surface at the centre of each pyridine ring indicates the steric interactions in this ring.

Crystal Explorer 17.5 software was also used to calculate and visualize the 3D energy framework along with intermolecular interaction energies. These interaction energies were calculated between a central molecule and the neighbouring molecules situated in a radius of 3.8 Å around the central molecule. These neighbouring molecules were generated by applying crystallographic symmetry operations and 16 neighbour molecules were found. The B3LYP/6-31G(d,p) level of theory, as implemented in Crystal Explorer 17.5, was used to calculate the interaction energy between the central molecule and each of the 16 neighbours. The interaction energy between two molecules (Equation (2)) is a counterpoise corrected energy obtained as the sum of electrostatic energy ( E e l e ) , dispersion energy corrected by Grimme’s D2 dispersion correction ( E d i s ) , polarization energy ( E p o l ) , and exchange-repulsion energy ( E r e p ) [45].

E t o t = E e l e + E d i s + E p o l + E r e p (2)

The value of interaction energy between the different pair of molecules and their four components are presented in Table 3 These energies have been scaled appropriately using for benchmarked energy models [45]. R represents the distance between the centroids (atomic means position) of two molecules and the N the number of molecules situated at the same distance R of the central molecule.

These results indicate that, the interactions between the central molecule and its neighbours can be grouped in six categories. The molecules of each of these categories can be identified by their colours in Figure 8.

The most important interaction energy (−95.4 kJ/mol) is obtained between the molecules that have the strongest S-H close contact with the central molecule. This interaction is dominated by the electrostatic component. The interaction between the yellow molecules and the central molecule is less than the previous one because of the proximity between the Zinc ions of these molecules. This Zn-Zn repulsive interaction can also explain why S-H close contact involving yellow molecules is less important than those involving green molecules. The attraction between the yellow molecules and the central molecule is dominated by dispersion interactions. The interaction energy involving molecules that do not participate in close contacts can be attributed to their distance from the central molecule and are dominated by electrostatic attraction. The interactions between violet molecules and the central molecule are destabilizing interactions probably due to the symmetry between these molecules. For the studied cluster, the total electrostatic energy is −161.9 kJ/mol, dispersion energy is −169.4 kJ/mol, polarization energy is −65.8 kJ/mol, repulsion energy 179.8 kJ/mol and the total interaction energy is −217.3 kJ/mol. Therefore, electrostatic and dispersion interactions contribute equally to the molecular packing of the Zn-3-ampy molecules in the crystal. These observations substantiate the 3D framework energy diagrams (Figure 9) of the concerned molecules. However, this diagram shows that electrostatic and dispersion interactions are predominant in different directions. The size of the tube relating the centroid of two molecules is directly proportional to the intensity of the interaction of the molecules.

The X-ray and theoretical geometry parameters for the complex are presented in Table 4. These results indicate that X-rays and theoretical geometry parameters are different and this difference is most important for bond angles. This observation is attributed to the fact that theoretical results are obtained from a single molecule in the vacuum while X-rays are obtained for a crystal which involves molecular interactions.

A and B refer to the two molecules of the dimer; A represents the molecule under the green surface and B the molecule on top of the green surface (Figure 7).

This result supports the fact that, the environment influences the properties of chemical species including their geometry parameters. However, linear regression analysis shows a linear correlation between the experimental and theoretical bond lengths. The regression coefficient and the standard deviations are 0.9869/0.053 and 0.9876/0.051 for B3LYP/6-311++G(2d,2p) and wB97XD/6-311++G(2d,2p), respectively. Although the B3LYP is less robust than wB97XD, they produced similar results (data for xyz coordinates are in Tables S3 and S4, ESM; optimised geometry parameters are presented in TableS5, ESM). The results also reveal that Zn-N1 and Zn-N2 are shorter than Zn-N3 probably because the thiocyanate is an anion and the charge transfer between this ligand and the metal is more important. In addition, the bond length between the Zn atom and the nitrogen atoms of the 3-ampy ligands (Zn-N3) are identical while the Zn-N1 and Zn-N2 are not. This is due to the van der Waals interaction between one of the hydrogen atoms of each amine group and the carbon atom of the thiocyanate group as revealed by the molecular diagram derived from NCIs analysis (Figure 7) of the electronic density of the concerned molecule. To confirm it, Zn(4-ampy)₂(NCS)₂ have been optimized at wB97XD/6-311++G(2d,2p) level of theory (see data in Tables S6 and S7, ESM) and the analysis of the results showed that Zn-N1(1.927 Å) and Zn-N2 (1.926 Å) are identical in Zn-4-ampy. This result confirms that the interaction between the carbon of the thiocyanate and the hydrogen of the amine groups is responsible for the difference between Zn-N1 and Zn-N2 bond length in Zn-3-ampy compound. These two bonds are shorter in Zn-4-ampy.

The complexes Zn-3-ampy with chloride and bromide co-ligands were optimised at the B3LYP/6-311++G(d,p) level of theory [46]. Comparison of the coordinate bonds between the Zn atom and the pyridine nitrogen atom of 3-ampy in the compounds (Zn(3-ampy)₂(NCS)₂, Zn(3-ampy)₂Cl₂ and Zn(3-ampy)₂Br₂), indicates that the said bond lengths increase as follows: Zn(3-ampy)₂(NCS)₂ (2.017) < Zn(3-ampy)₂Cl₂ (2.144) < Zn(3-ampy)₂Br₂ (2.152).

NBO [47] [48] analysis is a self-consistent theoretical framework for analysing the bonding nature of molecules and materials. It also analyses interaction between occupied orbitals (bond and lone pair) and unoccupied orbitals (anti-bonding and Rydberg orbitals) and between unoccupied orbitals. These interactions are interpreted as charge transfer between NBO donor and NBO acceptor or as the extension of the conjugation by electronic density delocalisation between the NBO donor and NBO acceptor. This type of interaction has the advantage to stabilize the molecule and the stabilisation energy associated (E⁽²⁾) with it is estimated by the second-order perturbation theory (Equation (3)). More intense is the interaction between the NBO donor and the NBO acceptor, greater is the extend of conjugation and higher is the E⁽²⁾.

E ( 2 ) = q i ( F i , j ) 2 ε j − ε i (3)

The values of the stabilisation energies of some NBO orbital interactions of Zn-3-ampy have been determined in order to characterise its coordination bond and to identify the interactions that greatly contribute to the stability of the synthetized complex. The values E⁽²⁾ higher than 5 kJ/mol for some NBO donor and NBO acceptor are presented in Table 5.

The analysis of NBO reveals that the most important interaction between ligand and Zn²⁺ occurs mainly between the lone pairs of bonding atoms (s and p orbitals) and 4s and 4p orbitals of Zn²⁺. The stabilisation energy (E⁽²⁾) of this charge transfer is the same for the two pyridine nitrogen atoms of the ring while it is different for the two nitrogen atoms of thiocyanate. The values due to metal-ligand charge transfer increase as follows N(3) < N(2) < N(3). The most important interaction occurs with N(1) while the less important is with N(3). This result is consistent with observations that the N(1)-M bond is the shortest coordination bond in this molecule and the N(3)-M bond is the longest one.

In this studied complex, there is also important charge transfer from LP* and RY* anti-bonding orbitals of metal ions with a stabilisation energy of 92 kcal/mol of each ligand. This points out the importance of empty s, p, d and f Zn orbitals adding to polarization and diffusion of the basis set, in the bonding process. NBO analysis also shows that, the delocalisation of the electronic density of the lone pair on sulphur atoms also play an important role in the stabilization process of this complex.

The analysis of NBO results also reveals that there is no orbital overlap between the metal ion and the chelating atoms of the ligand. Therefore, the bonding between the chelating atoms and the Zn(II) ion is mostly electrostatic interaction.

The HOMO and LUMO of the complex studied in this work are shown on Figure 10.

The HOMO of the complex is concentrated on the C-S and N-C bonds on one of the thiocyanate groups while the LUMO is distributed on the two pyridine cycles. The distribution of HOMO and LUMO indicates the nucleophilic and electrophilic part of the molecule, respectively. Energy values of these frontier molecular orbitals, the energy gap between these orbitals, the chemical potential, the global hardness and the dipole moment are used in quantum mechanics to describe the global reactivity of molecules. The data for the complex studied are presented in Table 6.

A high energy value for the HOMO indicates the tendency of the molecule to donate electrons and a low energy value for the LUMO shows a greater tendency to accept electrons. The energy values for the HOMO and LUMO of the complex studied indicates that it has the tendency to accept electrons than to donate. Compared to the B3LYP functional, the wB97XD functional overestimates the tendency of this molecule to accept electrons while it under estimates its tendency to donate electrons. The results (Table 6) also indicate that the wB97XD functional overestimates the chemical stability (high values of gap and η) of Zn-3-ampy and underestimates its polarizability (low dipole moment).

The Molecular electrostatic potential (MEP) [49] of the studied molecule (Figure 11) shows a blue area around the hydrogen atoms of the amine group, a yellow area around the sulphur and nitrogen atoms of the thiocyanate groups, and the rest of the surface is almost green. Then, the most electrophilic parts of the complex are the hydrogen atoms of the anime groups while the most nucleophilic are the sulphur and nitrogen atoms of the thiocyanate.

This result suggests that, the concerned molecule can act as hydrogen bond donor through hydrogen atoms and as hydrogen bond acceptor through sulphur and nitrogen atoms. However, comparing the intensity of the colours we can infer that, it acts as a good hydrogen donor but a poor hydrogen bond acceptor. The MEP of the complex also reveals that S1 and N1 are more nucleophilic than S2 and N2. This observation corroborates the distribution of the HOMO on the surface of the molecule and the atomic charges (Table 7). This result can be attributed to the fact that, S2 and N2 are involved in van der Waals contacts with the hydrogen atoms of the amine groups.