Investigation of structure-directing interactions within copper IJ I ) thiocyanate complexes through X-ray analyses and non-covalent interaction ( NCI ) theoretical approach †

Herein, we reported the synthesis of copperIJI) thiocyanate complexes with ortho-pyridinyl carbohydrazones containing a thiophene (L1) or a furyl ring (L2) as a mixture of two different crystals for each compound, linkage isomers of C1N, [CuIJNCS)IJL1)PPh3] and C1S, [Cu(SCN)(L1)PPh3], for L1, whereas monomeric and polymeric structures C2N, [Cu(NCS)(L2)PPh3], and C2P, [–(NCS)Cu(L2)–]n, for L2. Crystallographic information and theoretical calculations, mainly noncovalent interaction reduced density gradient (NCI-RDG) analyses, were pursued to generate a profound understanding of the structure-directing interactions in these complexes. The supramolecular assemblies are first driven by cooperative π⋯π interactions and hydrogen bonds followed by CH⋯π, S⋯S and S⋯π linkages. In the case of the linkage isomers, intermolecular interactions may have a significant role in the formation of the less stable S-bound isomer C1S.


Introduction
][10] They have applications in different areas such as organic light-emitting diodes (OLEDs), [11][12][13][14][15][16][17][18][19][20][21][22] supramolecular assemblies, oxygen sensors and biological probes.The rich structural features and utilitarian considerations have motivated researchers to focus on the synthesis and characterization of CuĲI) complexes with various donor ligands.The formation of structural variations is greatly influenced by several parameters such as the synthetic conditions or steric/electronic effects exerted by the ligand.
5][26] When SCN − acts as a terminal ligand, it affords potential interaction sites to generate non-covalent intermolecular interactions and, accordingly, can direct the crystal packing.Controlling the self-assemblies in the solid state on the basis of molecular structures and through the use of weak interactions is a long-standing goal of supramolecular chemistry. 27ery recently in our previous work, cuprous halide complexes of ortho-, metaand para-pyridinyl carbohydrazones were introduced. 28The influence of ligand structure and halide variations on the molecular structures and supramolecular arrays of the complexes were studied both experimentally and theoretically.In the following, we employed cuprous pseudohalide, CuSCN, for the synthesis of complexes with two ortho-pyridinyl carbohydrazones.][31] In this contribution, we report the structural characteristics of complexes from the reaction of CuSCN with PPh 3 and ortho-pyridinyl carbohydrazones containing a thiophene (L 1 ) or a furyl ring (L 2 ); see Scheme 1.The former resulted in two linkage isomers: C 1N [CuĲNCS)ĲL 1 )PPh 3 ] and C 1S [Cu(SCN)(L 1 )-PPh 3 ], while the later afforded monomeric and polymeric complexes of C 2N [Cu(NCS)(L 2 )PPh 3 ] and C 2P [-(NCS)Cu(L 2 )-] n .We have also used a recently introduced alternative interpretive technique, the non-covalent interaction (NCI) approach, to manifest the diverse NCIs at the crystal packing structures.This method is based on the analysis of the electron density and enables us to identify and visualize the interactions. 32arious non-covalent interactions, including hydrogen bonding, 33
A solution of the ligand in CHCl 3 was added dropwise to a mixture of copperĲI) thiocyanate and PPh 3 while stirring in CH 3 CN and then the mixture was filtered off.After slow diffusion of diethyl ether in the filtered solutions, two different crystals were obtained for each compound including the light orange needle crystals (C 1N ) and clear light red irregular crystals (C 1S ) for L 1 and orange needle crystals (C 2N ) and dark orange hexagonal crystals (C 2P ) for L 2 .A mixture of the isomeric crystals C 1N and C 1S is shown in Fig. 1.
Various ratios of acetonitrile and chloroform solvents were assessed in the crystallization of C 1 .Upon using more chloroform, the percentage of C 1N was dominant, whereas a higher amount of acetonitrile in the reaction pot increased the percentage of C 1S .For C 2 , the formation of crystals depended on the concentration of the reaction mixture.At a high concen-tration, the polymeric compound precipitated fast and we only obtained crystals of C 2N , but slower diffusion of diethyl ether in the more dilute solution afforded crystals of both C 2P (as the dominant product) and C 2N suitable for X-ray diffraction.
ORTEP diagrams of the molecular structures are shown in Fig. 2. The crystallographic data of the complexes are listed in Table 1.Selected bond distances and angles are summarized in Table 2.

Structural analysis
CuĲNCS)ĲL 1 )PPh 3 ] (C 1N ).The title compound crystallizes in the triclinic space group P1 ¯.Ligand L 1 binds to the copper atom in a bidentate chelating manner via N2 (pyridine) and N3 (imine).An isothiocyanate anion (N-donor) and one PPh 3 occupy the other coordination sites (Fig. 1).Houser and coworkers suggested an angular index (τ 4 ) which determines the geometry of the four-coordinate metal centres as follows: τ 4 = [360 − (α + β)]/141] (α and β are the two largest angles around a four-coordinate metal centre).The values of τ 4 will range from 1.00 for a perfect tetrahedral geometry to zero for a perfect square planar environment.Intermediate structures including trigonal pyramid and seesaws fall within the range of 0 to 1.00. 37According to the τ 4 value for C 1N (0.79), the coordination polyhedra of the copper centre can be described as trigonal pyramid.
In the structure of C 1N , each molecular unit of the complex is joined to the neighbouring unit by means of three 2-fold interactions including classical and non-classical hydrogen bonds N4-H4⋯S1 and C7-H7⋯S1, respectively (Table 3), and π py -π thiophene interactions (Table 4).
The dimers are further connected to each other through π py -π py and C5-H5⋯π PPh3 interactions along the a-direction (Fig. 3a) to afford chains which are laterally linked together via various intermolecular interactions to generate a 3D network.The interactions which link the chains along the b-axis include (i) S1⋯S1, (ii) C3H3⋯S1, (iii) C11-H11⋯π py , (iv) C2-H2⋯π thiophene and (v) C14-H14⋯π thiophene linkages (Fig. 3b).In addition, C21H21⋯S2 H-bonds plus weak C27-H27⋯π PPh3 interactions (C⋯Cg: 4.090 Å) connect them along the c-direction (Fig. 3c).The distance of the S⋯S interaction was found to be about 3.456 Å which is 4% shorter than the sum of the van der Waals radii of two sulfur atoms.A summary of the parameters for the other interactions mentioned above are presented in Tables 3 and 4.
[CuĲSCN)ĲL 1 )PPh 3 ] (C 1S ).The red irregular crystals of C 1S are the second form resulting from the reaction of a 1 : 1 molar ratio of L 1 and the mixture of cuprous thiocyanate and PPh 3 .X-ray diffraction analysis reveals that it crystallizes in the triclinic space group P1 ¯.The central copper atoms are again in trigonal pyramidal environments (τ 4 = 0.82, exactly equal to that for C 1N ) formed by N py and N im (from the chelating ligand), PPh 3 moiety and thiocyanate anion (this time as an S-donor) (Fig. 1).Changing the coordination site of the ambidentate ligand, NCS − , alters the supramolecular architecture of C 1S compared to that of the N-bound isomer C 1N .The crystal structure of C 1S contains hydrogen bonded dimers generated by three pairwise interactions, N4-H4⋯N1, C7-H7⋯N1 and C10-H10⋯N1 (Table 3 and Fig. 4a).It is worth noting that the coordination of sulfur to the CuĲI) atom  and consequently the orientation of the N1 atom direct the formation of hydrogen bonds and lead to supernumerary slippage of molecules on each other.The offset of pyridine and thiophene rings prevents the formation of a π py -π thiophene interaction, unlike in the structure C 1N .Thus, instead of a π py -π thiophene interaction, a CO⋯π interaction is established between the discrete molecules in the dimers.In addition, S⋯S interactions connect the dimers to form [001] chains.
[CuĲNCS)ĲL 2 )PPh 3 ] (C 2N ).This compound crystallizes in the triclinic space group P1 ¯.A trigonal pyramidal configuration of the Cu(I) centres (τ 4 = 0.80) has been formed by the bidentate chelating ligand, PPh 3 and isothiocyanate (N-donor).C 2N has a similar coordination environment to that of C 1N leading to its being isostructural with this complex.The supramolecular organization of C 2N is also disciplined by the formation of dimers through the same kind of interactions in C 1N : N3-H3⋯S1, C6-H6⋯S1 and π py -π furyl stacking interactions.Pyridine rings participate in the other π-π interactions (π py -π py ) which build up a chain of the dimers directed along the c-axis.The chains are further reinforced by C3-H3A⋯π PPh3 and C4-H4⋯π PPh3 contacts (see Table 4).On the other hand, S⋯S synthons accompanied by C2-H2⋯S1, C15-H15⋯S1 and C10-H10⋯N4 hydrogen bonds as well as C1-H1⋯π furyl interactions link the chains to create (110) sheets.The third dimension of the supramolecular assembly results from the connection of the layers via C21-H21⋯π furyl linkages.Crystal packing diagrams of C 2N are presented in Fig. 5.
[-(NCS)CuĲL 2 )-] n (C 2P ).The reaction of a 1 : 1 molar ratio of L 2 with the mixture of CuSCN and PPh 3 results in two kinds of crystals: orange needle crystals, C 2N , and dark orange hexagonal ones, C 2P .X-ray diffraction analysis confirms that C 2P crystallizes in the monoclinic space group P21/c.Unlike the former complexes, PPh 3 does not coordinate to copper(I).In return, the trigonal pyramid geometry of the Cu(I) centre (τ 4 = 0.80) consists of the chelating ligand (L 2 ), thiocyanate (−SCN) and isothiocyanate (−NCS).Indeed, each thiocyanate anion acts as a bridge between the Cu(I) ions through simultaneous binding from S and N atoms.This coordination pattern leads to the formation of infinite polymeric chains extended along the c-axis (Fig. 6a).The Cu⋯Cu distance within the metal chain is 5.323 Å.The chains are further stabilized via intrachain hydrogen bonds (N3-H3⋯O1, C9-H9⋯O2 and C4-H4⋯S1).
Although the coordination structure of C 2P is different from the others, π-π interactions still have an important     stacking and H-bonding interactions, leading to the connection of the (011) layers along the a-direction to complete the overall supramolecular association (Fig. 6b).In other words, each coordination chain is associated with four other chains in a 3D arrangement, from one side through π furyl -π furyl , C11-H11⋯N4 and C11-H11⋯O1 linkages and from the other side by π py -π py stacking and C2-H2⋯N4 interactions (Fig. 6c).In addition, these chains are laterally linked through interesting S⋯π py interactions in the b-direction (Fig. 6d).

NCI approach
Principle.The non-covalent interaction (NCI) reduced density gradient (RDG) method has been recently developed as a theoretical strategy to visualize weak interactions.Investigation of the interactions using NCI-RDG analysis is quite concordant with the traditional method that recognizes them according to distances and angles.However, the NCI-RDG technique has more accuracy and precision, as it is based on fundamental computation.It provides a rich illustration of strong attractive, van der Waals interactions and also steric repulsions.The theory rests on the analysis and the graphical interpretation of two scalar properties, charge density ρ and its derivatives, namely the λ eigenvalue of its Hessian and its reduced gradient sĲρ), 22 defined as: where ∇ρ is the gradient of ρ.The non-covalent interactions are located in the regions with low RDG and density.Analysis of signĲλ 2 ) of the electron density Hessian can be used to discern different types of interactions.For the strong ones such as H-bonds, sign(λ 2 )ρ < 0; for the weak van der Waals types, sign(λ 2 )ρ ≈ 0; and for the non-bonded interactions like steric repulsion, sign(λ 2 )ρ > 0.
9][40][41][42][43][44] Close contacts between atoms change the behaviour of the reduced gradient signal more compared to the contacts among the atoms present in the tails, leading to troughs in the 2D NCI plots.These troughs, specially the ρ value at the troughs, are the basis of the NCI approach.The 2D NCI plots are then applied as inputs to construct 3D NCI plots, including isosurfaces of the reduced gradient of the density enabling the spatial visualization of the close contacts.
We applied this method to unravel the nature of supramolecular interactions in the title complexes.NCI analysis has been performed on the structure of complexes including the diverse noncovalent interactions.The considered structures were cut out directly from the CIF data.Since dimerization is the prominent feature of the crystal packing in the monomeric complexes (C 1N , C 1S and C 2N ), the main NCIs are related to the interactions involved in the formation of dimers.The 2D and 3D NCI plots of dimers are shown in Fig. 7. Accordingly, we have done calculation once on the dimers including only the carbohydrazone ligands (Cu + and SCN − ions and PPh 3 moieties have been eliminated) and again for the whole dimeric units of complexes.Some of the other interesting intermolecular interactions in the crystal structures have also been investigated by the NCI method.The presence of noncovalent interactions is characterized by spikes at negative to near-zero sign of λ 2 , whereas the peaks at positive sign indicate the repulsive steric contacts due to the ring formation. 45The spikes at the zero area (signĲλ 2 )ρ between ±0.015 a.u.) show vdW interactions.Notable points of the NCI calculations have been illustrated in the following: (i) As shown in Fig. 7, for the ligand dimers, the spikes that appeared at 0.024 a.u.belong to the pyridine ring closure.These spikes shift to lower values (less repulsion) in the whole dimeric units of complexes.It can be explained by the effect of metal ion in the charge redistribution as well as the electrostatic interaction between atoms within the rings.
(ii) In the case of C 1N , the thiophene ring closure spike (signĲλ 2 )ρ) is located between 0.042 and 0.044 a.u.while in the 2D plot and the 3D isosurface of C 2N , the furyl ring has a much lesser repulsion of ring closure than the thiophene al-ternative.It may be caused by the greater charge perturbation due to the presence of an oxygen atom which leads to more electrostatic interactions.Consistent with this, natural bond orbital (NBO) analysis also reveals the stabilizing energy of 54 kcal mol −1 for the electronic delocalization "lone pair (O) → π* (C-C) orbital" which is more than that for the corresponding charge transfer energy in the thiophene ring (LP(S) → π*(C-C): 48 kcal mol −1 ).
(iii) C 1N and C 1S compounds are linkage isomers, in a way that SCN − is coordinated from N or S atoms, respectively.Optimization of the isomers, in the gas phase and also acetonitrile and chloroform solutions, indicates that the stability of C 1S is approximately 3-4 kcal mol −1 less than that of C 1S ; however, it has been also formed in the solid state.The formation of C 1S can be attributed to stronger intermolecular attractions particularly those involved in dimerization which compensate for the lesser stability of the discrete units of C 1S .RDG isosurfaces show stronger interactions in the C 1S dimer rather than in C 1N .Counterpoise calculations at the M06-2X/6-311G* level indicate that the binding energy of two complexes in a dimer, ΔE dimer , for C 1S is 2.8 kcal mol −1 more than for C 1N as well.
(iv) It was thought to be of interest to further investigate the sulfur interactions to figure out their nature in the solid state structures.The sulfur atom, due to its large van der Waals radius and high polarizability, is able to establish several interactions with its local environment. 46Morgan and co-workers first proposed the hypothesis that a strong interaction exists between aromatic rings and divalent sulfur atoms. 47The importance of the S⋯π aromatic interaction is revealed in the high degree of its conservation across members in protein folding and stabilization. 34,48ig. 8 represents the 3D plots of S⋯S and S⋯π interactions in the solid state structures.The compact and small, flat, pill-shaped isosurfaces, concentrated on the NCI critical points indicate that these interactions are significantly attractive and contribute to the crystal packing stability. 45

Conclusions
Diverse coordination structures from the reaction of CuSCN with PPh 3 and ortho-pyridinyl carbohydrazone containing a thiophene (L 1 ) or furyl ring (L 2 ) were presented.A mixture of thiocyanate linkage isomers, C 1N and C 1S , was obtained for L 1 , while the reaction with L 2 rendered two monomeric and polymeric compounds, C 2N and C 2P , respectively.The molecular and supramolecular structures of these systems were elucidated using X-ray diffraction.The structuredirecting interactions were also investigated by NCI-RDG calculations.
Pyridine, thiophene and furyl rings, the polarized aromatic systems, have an important role in governing the supramolecular assembly of the complexes by establishing π-π interactions.However, the coordination of sulfur to the CuĲI) atom in C 1S leads to supernumerary slippage of the neighbouring molecules which prevents the formation of π py -π thiophene connections.CH-π interactions between PPh 3 moieties contribute to further stabilization of the selfassociation in the monomeric complexes (C 1N , C 1S and C 2N ).
A prominent feature of the crystal packing in the monomeric complexes is the formation of the dimeric motifs via hydrogen bonding and π-π stacking interactions.Formation of the less stable S-bound isomer C 1S can be attributed to stronger intermolecular attractions particularly those involved in dimerization which compensate for the lesser stability of the discrete units of C 1S compared to C 1N .
Another interesting feature of the solid state structures is the presence of the lesser known S⋯S and S⋯π interactions.NCI-RDG analysis clearly indicates the significant contribution of these interactions in maintaining favourable packing interactions in the complex.

Materials and methods
All chemicals and solvents used in the syntheses were of reagent grade and were used without further purification. 1H and 13 C NMR spectra were recorded on a Bruker (Avance DRS) 250 MHz spectrometer.IR spectra were recorded on a Nicolet 510P spectrophotometer using KBr disks.Elemental analysis was performed using a Heraeus CHN-O-RAPID apparatus.

Synthesis procedures
The ligands were prepared from the reaction of two equivalent amounts of 2-thiophenecarboxylic acid hydrazide or 2-furoic hydrazide and 2-pyridinecarboxaldehyde in methanol solution.
A solution of ligand (0.20 mmol) in CHCl 3 (4 mL) was added dropwise to a stirred solution of a mixture of copperĲI) thiocyanate (0.20 mmol) and triphenylphosphine (0.20 mmol) in CH 3 CN (2 ml).The colour of the reaction mixture turned from orange to red.The reaction mixture was filtered; slow diffusion of diethyl ether in the filtered solution afforded

Crystal structure determination
Single crystal X-ray diffraction data were collected for all compounds on an Agilent Gemini Ultra diffractometer equipped with an Eos CCD area detector and using either Mo-Kα radiation (λ = 0.71073 Å) or Cu-Kα radiation (λ = 1.5418Å).The data were collected at 173 K using an Oxford Cryosystems Cryostream 600.The data were processed with CrysAlisPro. 49mi-empirical absorption corrections were carried out using the Multi-Scan 50 program.The structures were solved by direct methods using SHELXT 51 and refined with full matrix least squares refinement using SHELXL-2013 (ref.52) within Olex2. 53All non-hydrogen atoms were refined anisotropically.Hydrogen atoms were added at calculated positions and refined using a riding model based on the parent atom.The CIF files have been deposited with the CCDC and have been given the deposition numbers 1401344, 1401345, 1469718 and 1469717 for C 1N , C 1S , C 2N and C 2p , respectively.

Computational details
The NCI technique was carried out through the analysis of the reduced density gradient (RDG) with low densities 32 at the ωB97XD 54 /6-311+G** level using the Gaussian 09 package 55 and Multiwfn program. 56The calculated grid points are plotted for a defined real space function, signĲλ 2 (r))ρ(r) and reduced density gradient (RDG) and a visualization of the gradient isosurface was depicted using the VMD 1.9.2 software. 57The colour of the isosurfaces was decided using the value of sign(λ 2 )ρ.Blue, green and red colour codes are commonly used to describe stabilizing H-bonding, van der Waals and steric interaction, respectively.Pictures are provided for an isosurface value of s = 0.5.
Natural bond orbital (NBO) analysis 58 was performed on the crystal structure of the complexes using the NBO 3.1 module in Gaussian 09 at the B3LYP/6-311+G** level of theory.The binding energy of two complexes in a dimer, ΔE dimer , for C 1N , C 1S and C 2N were calculated at the M062X/6-311G* level based on the energy difference between the dimer and its units.The interaction energies have been corrected for the basis set superposition error (BSSE) using the counterpoise (CP) procedure. 59

Fig. 1
Fig. 1 50× magnification photo of a mixture of orange rectangular cube C 1N and red cube C 1S .

Fig. 3
Fig. 3 (a) Representation of dimeric units in C 1N and their association through π-π and C-H⋯π interactions, creating the [100] chain; (b and c) side views of the crystal packing in the ac and ab planes, respectively, which show how the chains are connected in the 3D network.