Crystal structure and vibrational spectra of bis(2-isobutyrylamidophenyl)amine: a redox noninnocent ligand

The molecular structure of bis(2-isobutyrylamidophenyl)amine (H3LNNN) has been determined from single-crystal X-ray diffraction data. The crystal packing of H3LNNN is governed by the N–H···O and C–H···O hydrogen-bonding and C–H···π stacking interactions between the vicinal molecules. The intermolecular interactions in the crystal structure of H3LNNN have been also examined via Hirshfeld surface analysis and fingerprint plots. The Hirshfeld surface analysis showed that the important role of N–H···O and C–H···π interactions in the solid-state structure of H3LNNN. The molecular structure, vibrational frequencies, and infrared intensities of H3LNNN were computed by ab initio HF and DFT (B3LYP, B3PW91, and BLYP) methods using the 6–31G(d,p) basis set. The computed theoretical geometric parameters were compared with the corresponding single crystal structure of H3LNNN. The harmonic vibrations calculated for the title compound by the B3LYP method are in good agreement with the experimental IR spectral data. The theoretical vibrational spectrum of the H3LNNN compound was interpreted through potential energy distributions using the SQM Version 2.0 program. The performance of the used methods and the scaling factor values were calculated with PAVF Version 1.0 program.


Introduction
In recent years, a large amount of work has been devoted to the study of transition metal redox processes for electron transfer processes due to the importance of the electron transfer process in the development of industrial useful catalysts [1][2][3][4][5][6].The coordination of ligands to metal ions is one way of attenuating metal-based redox processes.During a transition metal-mediated redox process, an electron can be accepted by or released from a metal center.When redox noninnocent ligands are coordinated to transition metal ions, the ligand can also participate in electron transfer processes.So, ligand design is very important.Recently, we have focused on the design and catalytic activity of Co(II) complexes formed with tridentate redox-innocent compounds as a ligand [7,8].In the light of these findings and the continuation of our research studies on the tripodal ligand system, our team produced and characterized a number of substituted tridentate ligands and their metal complexes [8][9][10][11].We demonstrate that our synthesized transition metal complexes by using bis(2-isobutyrylamidophenyl)amine as the tripodal redox noninnocent ligand are capable of catalytic oxidation reactions using dioxygen.This ligand system has two N-amidate donor atoms and one amido donor and supports coordinatively unsaturated metal centers with open coordination sites available for small molecule binding.This ligand stabilizes both mononuclear and dinuclear cobalt(II) complexes able to catalytically oxidize PPh₃ to Ph₃PO with much better catalytic efficiencies than those previously observed for cobalt(II) complexes in the presence of excess dioxygen under ambient conditions.Performing these reactions with the large substrate to catalyst loading ratio (500:1) gives maximum turnover numbers of 185 and 345 mol product/mol catalyst for the cobalt(II) complexes.In addition, the most recent application of this ligand system derivatived with different functional groups is the ability for catalytic C-H amination to form indolines from aryl azides by cobalt(II) complexes of them [12].In that, the study of redox behavior of ligands is important for the development of new catalysts.The most suitable markers for determining the redox behavior of the ligand are the C-X (X = C, N, O, S, Se) stretching vibration modes and bond distances.If the ligand is redox noninnocent, the coordination of the metal decreases C-X bond distances, and C-X stretching vibrations shifts to lower frequencies via radical parts formed on the ligand skeletal.The experimental vibrational spectra are accurately reproduced by the calculations, which show that C-C, C-N, and C-X vibration modes are extensively mixed with other modes, and thus unsuitable to work as vibrational markers [13].Therefore, in this study, we aim that learn more information about the structure of the redox noninnocent ligands due to their role in catalytic processes.To achieve this aim, we selected bis (2-isobutyrylamidophenyl)amine as a sample redox noninnocent ligand.We have calculated the structural parameters and vibration modes of H 3 L NNN in the ground state to distinguish the fundamentals from the structural parameters and experimental vibrational frequencies by using the HF [14], B3LYP [15,16], BLYP [15,16], and B3PW91 [15,17], with the standard 6-31G(d,p) basis set.The calculated structural parameters and vibration modes were analyzed and compared with obtained experimental results.In the current work, we also investigated the relative performance of B3LYP, BLYP, and B3PW91 methods, as well as of HF for comparison, at the 6-31G(d,p) level taking as a test compound bis (2-isobutyrylamidophenyl)amine.On the other hand, the role of intermolecular interactions of bis(2-isobutyrylamidophenyl)amine has been analyzed through single-crystal structure studies, and these intermolecular interactions in the single crystal structure of bis (2-isobutyrylamidophenyl) amine have been visualized via Hirshfeld surface analysis and fingerprint plots.

Instrumentation
1 H and 13 C NMR were obtained on a Bruker Avance III 400 MHz Ultrashield Plus Biospin spectrometer.The deuterated solvent DMSO-d 6 was used as purchased.FT-IR spectra were recorded on a Perkin Elmer Spectrum 100 series FT-IR spectrometer in KBr disc and were reported in cm -1 units (4000-400 cm -1 ; number of scans: 250; resolution: 1 cm -1 ).X-ray diffraction studies were carried out in the X-ray Crystallography Laboratory at Emory University on a Bruker Smart 1000 CCD diffractometer.Mass spectra were recorded on an Agilent 6460 series LC-MS/MS trap with electrospray ionization (ESI) source and triple quadrupole ion trap mass analyzer by direct infusion and ESI operated in the positive and negative mode in Advanced Technology Research and Application Center, Mersin University, Mersin, Turkey.Acetonitrile: water (0.1% formic acid) (95:5, %) was used as mobile phase and 2 μL of the sample injected at 0.3 mL/min flow rate [Column: Zorbax Eclipse XDB-C18 (4.6 mm I.D. × 50 mm L., 1.8 μm)].

Theoretical studies
Theoretical calculations were made with the Gaussian 03W program [20].The molecular structure of H 3 L NNN in the ground state was optimized by using BLYP, B3LYP, B3PW91, and HF methods with 6-31G(d,p) basis set.The vibrational frequencies were also computed with the same methods and basis set.The frequency values computed at these levels contain known systematic errors [21].These differences can be corrected using scaling factor values of 0.8992, 0.9614, 1.0072, and 0.9573 for HF, B3LYP, BLYP, and B3PW91, respectively [22][23][24][25][26][27].The scaled quantum mechanical procedure has been widely used in the identification of the vibrational bands of IR and RAMAN spectrums [28].The vibrational modes were assigned using SQM Version 2.0 program on the principle of potential energy distribution analysis [29].The performance of the methods used was quantitatively characterized using the PAVF Version 1.0 program [30,31].

Hirshfeld surface analysis
Analysis of Hirshfeld surfaces and their associated 2D fingerprint plots of H 3 L NNN were computed by using CrystalExplorer 3.1 [32].The Hirshfeld surfaces are mapped with different properties such as shape index, d norm , etc.The d norm is normalized contact distance, defined in terms of d e , d i , and the vdW radii of the atoms.The combination of d e and d i in the form of a 2D fingerprint plot displays a summary of intermolecular contacts in the crystal.

Results and discussion
The synthesis of the title compounds involves the reaction of an isobutyryl chloride with bis(2-aminophenyl)amine in dichloromethane in the presence of triethylamine.The compound was recrystallized by layering hexane onto a concentrated CH 2 Cl 2 solution of the product and characterized by 1 H NMR, 13 C NMR, LC-MS/MS, FT-IR, and X-ray single-crystal diffraction method.All data obtained are consistent with the expected structure.

Molecular geometry
The molecular structure of bis(2-isobutyrylamidophenyl)amine was confirmed by the single crystal X-ray structure studies (Figure 1a).For H 3 L NNN , data collection and refinement are summarized in Table 1.Bond lengths, angles, and hydrogen bond details of the title compound are also presented in Tables 2-4, respectively (Tables 1S and 2S).
The bond distance of the carbonyl groups in the title compound is typical for the double-bond character, C7-O1 = 1.228(3)Å, C17-O2 = 1.231(3)Å.However, the CN bond distances for the investigated compound are all shorter than the average single CN bond distance of 1.In the crystal structure of the title compound, the molecules are connected by intermolecular hydrogen bonds: N2-H2A x, y, z; (iii) -1+x, +y, +z] (Figures 2a-2c and 3).
The unit cell of H 3 L NNN contains two independent molecules in the asymmetric unit, represented as A and B in Figure 1a and these molecules are virtually identical conformation as you can see in Figure 1b.Molecules A and B interact via strong N-H⋯O (Table 4) hydrogen bonds between amide hydrogen atom as strong hydrogen bond donor and carbonyl oxygen atom as strong hydrogen bond acceptor in the asymmetric unit.Moreover, the N-H⋯O hydrogen bonds continue infinitely and lead to the formation of infinite dimeric R 2 2 (20) synthons (Figure 2a).These dimeric synthons in the asymmetric unit expand along the crystallographic [010] direction.The formation of dimeric synthons in H 3 L NNN is also supported by additional bifurcated C-H⋯π interactions between phenyl rings and aliphatic hydrogen atoms (Figures 2b  and 2c).
The infinite chain occurring via N-H  (10) dimeric motif is due to the interaction between aliphatic hydrogen atoms and carbonyl group oxygen atom of two neighboring molecules.On the other hand, the R 2 2 (12) and R 2 2 (14) dimeric motifs occur between aryl ring hydrogen atoms and carbonyl group oxygen atom of two neighboring molecules (Figure 3).
The point group symmetry of the molecular structure of the H 3 L NNN compound is C S .We have performed a full structural optimization of the H 3 L NNN compound and the optimized geometrical parameters calculated by HF and DFT methods (Table 5, Figure 4).In addition, we have compared the experimental geometric parameters with the calculated one and we found that the calculated bond distances and angles show good agreement with experiment one.The best agreement with the experimental values was obtained for the HF and B3LYP methods for bond lengths and bond angles, respectively.The largest difference between calculated and experimental bond distances and angles are 0.042 Å and 5.95°, respectively, for DFT/B3LYP-6-31G(d,p) method.From the calculated values, it has been found that most of the optimized bond distances are slightly larger than the experimental bond distances since the calculations are for isolated molecules in the gas phase and the experimental results are for the solid-state molecules [34][35][36][37][38][39].Although there are minor differences between experimental and theoretical values, the calculated geometric parameters represent a good approximation and are the basis for calculating other parameters such as vibrational frequencies and thermodynamic properties.
The computed thermodynamic parameters (such as thermal energy, specific heat capacity, dipole moment, rotational constants, entropy, and zero-point vibrational energy) of H 3 L NNN by all used methods are listed in Table 6 * The atom-numbering scheme of the molecular structure is given in Figure 1a.

Vibrational assignments
FT-IR spectrum of the title compound is given in Figure 6S.Table 7 lists the vibration frequencies obtained using B3LYP calculations along with an approximate description of each of the experimental frequencies and normal modes.The other calculations (HF, B3PW91, and BLYP) were given as supplementary materials (Tables 3S and 4S).
The title compound has 50 atoms; thus, it gives 144 (3n − 6) normal modes of vibration.All vibration modes are active in both infrared and Raman spectrums.Generally, the theoretical vibrational frequencies are higher than the experimental ones, because of anharmonicity of the incomplete treatment of electron correlation and of the use of finite one-particle * The atom-numbering scheme of the molecular structure is given in Figure 1a.basis set [37,40,41].Therefore, these wavenumbers must be scaled by a proper scale factor and, in this research study, we have used the scaling factor values for HF, B3LYP, BLYP, and B3PW91 as 0.8992, 0.9614, 1.0072, and 0.9573, respectively.The identification of the vibration bands was made using the SQM 2.0 program [29] and the animation option of the GaussView 5.0 program [27].All experimental vibrational frequencies are in good agreement with the theoretical ones.

Table 4. Hydrogen bonds for the title compound (Å, °).*
According to Table 7, experimental vibrational frequencies are in better agreement with the scaled vibrational frequencies and are found to have a good correlation for B3LYP than BLYP, B3PW91, and HF methods.
In the heterocyclic compounds, ν N-H vibration occurs in the region 3500-3000 cm -1 .The IR band appearing at 3406, 3398, and 3367 cm -1 is assigned to the ν N-H stretching mode of vibrations.These vibration modes are computed at 3451, 3404, and 3404 cm -1 for the B3LYP method.The differences between experimental and computed ν N-H stretching modes  are about 45, 6, and 37 cm -1 (DFT-B3LYP/6-31G(d,p).These striking discrepancies can come from the formation of intermolecular hydrogen bonding with N-H.This interpretation is verified with ν C=O stretching vibration mode.The differences between experimental (1695 and 1679 cm -1 ) and computed (1707 and 1704 cm -1 ) ν C=O are about 12 and 25 cm -1 , respectively.It can be easily observed in the correlation graphics of the computed and experimental frequencies of   2c).
The characteristic CH stretching vibration modes ν CH of the aromatic structure of the H 3 L NNN compound are expected to appear in the frequency range 3100-3000 cm -1 [42][43][44][45].Although eight vibrational modes are calculated in the 3100-3000 cm -1 range, the ν CH stretching vibration modes of H 3 L NNN were assigned to four bands observed in the IR spectrum.This difference between the calculated and observed vibration band numbers is due to the overlapping of the aromatic ν CH  stretching vibrational frequencies.The first two bands (3118 and 3115 cm -1 ) are symmetric ν CH stretching vibration modes and the others (3099 and 3059 cm -1 ) are asymmetric ν CH stretching vibration modes of the aromatic structure [46].For the assignments of methyl group frequencies, 39 fundamental vibration modes can be associated with methyl groups.Twelve stretchings, nine deformations, six rockings, five umbrellas, and seven torsion vibration modes have designated the motion of the methyl group.The methyl symmetric and asymmetric stretching frequencies are observed at 3035, 3001, 2966, 2964, 2938, and 2929 cm -1 in the IR spectrum of the title compound.The minor differences between observed and calculated asymmetric stretching vibrational modes may be due to strong C-H•••π interaction which are observed in the crystal form (Figures 2a-2c, Table 7).The observed bands at 1382 and 1357 cm -1 are attributed to methyl umbrella vibration modes [24].The bands observed at 1109, 1049, and 902 cm -1 are assigned to the rocking vibration modes of the methyl group.The bands due to the δ CH in-plane aromatic ring bending vibration mode interacting with the ν CC stretching vibration mode are observed in the region 1608-950 cm -1 [47].ϒ CH vibration modes are strongly coupled vibrations and occur in the region 964-721 cm -1 .All the δ CH and ϒ CH bending vibration modes of the CH group have been identified and they are given in Table 7.
The identification of CN stretching vibration modes ν CN is difficult because of the mixing of the other vibration modes.However, we solved this problem by the GaussView Version 3.0 and the SQM Version 2.0 programs [29].Therefore, the CN stretching vibration modes are clearly identified and assigned in this research.Some of the vibration bands appearing between 1421 and 1049 cm −1 are assigned as CN stretching vibration modes (Table 7).All the obtained results agree with the literature [48].
Generally, the C=C stretching vibration modes are seen in the region of 1430-1650 cm -1 for aromatic compounds [49][50][51][52].The C-C stretching vibration modes of the title compound are observed at 1608, 1597, 1579, 1568, 1527, 1490, 1436, and 1421 cm -1 .All bands lie in the expected range when compared to the literature values [46].The C-C-C in-plane bending vibration modes are observed between 879 and 520 cm -1 and the ϒ CC vibration modes are calculated between 737 and 466 cm -1 .
A general better performance of B3LYP versus the other methods can be quantitatively characterized by using the root mean square values, the mean absolute percentage error, and the coefficients of correlation (r) between the observed and computed vibration frequencies.All these obtained data were computed in this study by the PAVF Version 1.0 program [30] according to Scott and Radom.The coefficients of correlation values for all DFT methods were greater than 0.9993 and they are very close to those reported in the literature [43][44][45][46][47][48][49][50][51][52][53][54][55].
The root mean square errors of the experimental and calculated vibration bands are found to be 13.49, 14.57, 15.48, and 31.06 for B3LYP, B3PW91, BLYP, and HF methods, respectively.These obtained results indicate that the fundamental frequencies computed by B3LYP, B3PW91, and BLYP methods for the H 3 L NNN compound show good agreement with the experimental values.Especially, B3LYP has the best agreement.A small difference between the calculated and experimental vibrational modes is also observed.These small differences due to the formation of inter-and intramolecular hydrogen bonding.In addition, we note that the theoretical calculations belong to the gaseous phase and the experimental results belong to the solid phase [37].
We also computed the optimal scaling factors, which are crucial for vibrational spectral identification, using the PAVF 1.0 program [30].Only single-uniform scaling factors were calculated without accounting for different vibrations.The single-uniform scaling factor values obtained are 0.9606, 0.9895, 0.9576, and 0.9034 for the B3LYP, BLYP, B3PW91, and HF methods, respectively.These obtained scaling factor values are very close to those recommended by Scott and Radom [22] for the same levels of theory (0.9614, 1.0072, 0.9573, and 0.8992, respectively).Thus, for future vibrational spectral predictions for unknown derivatives of H 3 L NNN , one can recommend scaling factors 0.9606, 0.9895, 0.9576, and 0.9034 for the B3LYP, BLYP, B3PW91, and HF methods, respectively.

Hirshfeld surface analysis
Hirshfeld surface analysis for molecules A and B in the asymmetric unit of H 3 L NNN was calculated by using the program CrystalExplorer 3.1 [32].The Hirshfeld surface was helped to distinguish the similarities and differences between the symmetry-independent molecules A and B present in the asymmetric unit.The Hirshfeld surfaces of H 3 L NNN were investigated to clarify the nature of the intermolecular interactions and are illustrated in Figures 5, 6a and 6b showing the surfaces that have been mapped over a d norm and shape index functions.The surfaces are shown as transparent to allow visualization of the molecular moiety, in a similar orientation for the molecules, around which they were calculated.In the d norm Hirshfeld surface, contacts with distances equal to the sum of the van der Waals radii are represented as white regions and the contacts with distances shorter than and longer than van der Waals radii are shown as red circles and blue areas, respectively [56,57].
In front and back d norm surfaces of molecule A, a total of four dark red spots were observed; these dark red spots are for the short N-H⋯O hydrogen bonds between molecules A and B.Moreover, there is one smaller red spot corresponding to weaker C-H⋯O interactions.On the other hand, in front and back d norm surfaces of molecule B, a total of seven red spots were observed; the four dark red spots in these surfaces are for the short N-H⋯O H-bonds between molecules A and B, and the other three (light red spots) in front d norm surfaces are for C⋯H interactions (also recognizable on Hirshfeld surface mapped with shape index function, Figures 6a and 6b   compound.The decomposed fingerprint plots for the two crystallographically independent molecules A and B are shown in Figure 7S.For both molecules A and B, the H⋯H interactions have the highest contribution of the total Hirshfeld surface with 60.9 and 61.7%, respectively, and the contribution from the H⋯H contact is 0.8% more for molecule B compared to molecule A. Despite the high share of H⋯H interactions, the role of these interactions in the stabilization of crystal structure is quite small in importance because this interaction is between the same species.In both fingerprints plots for molecules A and B, the two sharp spikes responsible for the strong N-H⋯O H-bond formation were observed.These contributions are almost similar with a difference of 0.4% for both molecules.On the other hand, the wings regions were observed which correspond to the C⋯H interactions, attributed to C-H⋯π interactions, in both fingerprints plots of molecules A and B. The contribution from the C⋯H contact is 1.2% more for molecule A in comparison with molecule B.
. The structure optimization and zero-point vibrational energy of H 3 L NNN in HF, BLYP, B3LYP, and B3PW91/6-31G(d,p) are 282.8046,256.6969, 264.7935, and 265.3738 kcal/mol, respectively.The global minimum energy obtained for structure optimization of H 3 L NNN is -1092 a.u.for the B3LYP method.The minimum energy becomes -1085 a.u.for HF.The difference in the amount of energy between the methods is ca.7 a.u.only.

Figure 1 .
Figure 1.(a) Crystal structure of bis(2-isobutyrylamidophenyl)amine.Thermal ellipsoids are shown at the 50% probability level and hydrogen atoms have been removed for clarity.(b) Overlay diagram of two independent molecules.
) between phenyl carbon atom of molecule B and phenyl/methyl hydrogen atom vicinal molecule and C-O⋯H interactions between carbonyl O atom of molecule A and aliphatic H atom of molecule B. This indicates that these interactions play a very important role in the formation of crystals.The analysis about C•••H interactions of the molecules A and B was done using the Hirshfeld surface shape index (Figures 6a and 6b

Figure 5 .
Figure 5. d norm Hirshfeld surface and d norm Hirshfeld surface surrounded by one neighboring molecule associated with close contacts of molecules A and B.

Figure 6 .
Figure 6.Hirshfeld surface of molecules A (left) and B (right) mapped with shape index function.

Table 1 .
Crystal data and structure refinement for H 3 L NNN .

Table 2 .
Selected bond lengths for H 3 L NNN .*

Table 3 .
Selected bond angles for H 3 L NNN .*

Table 5 .
Selected optimized and experimental geometries of H 3 L NNN in the ground state.*

Table 6 .
The calculated thermodynamic parameters of H 3 L NNN .

information Crystal structure and vibrational spectra of bis(2-isobutyrylamidophenyl)amine: a redox noninnocent ligand
3 L NNN shows N-H•••O, C-H•••O, and C-H•••π inter-molecular interactions.The N-H•••Ointeractions between molecules are among the strongest reported interactions for H 3 L NNN .The Hirshfeld surfaces analysis has been used for more investigation of intermolecular interactions as a driving force for the crystal structure of the H 3 L NNN compound formation has been demonstrated.In addition, the relative contribution of intermolecular interactions in H 3 L NNN is analyzed by fingerprint plots of the Hirshfeld surface.The ground state geometries were optimized using the B3LYP, BLYP, B3PW91, and HF methods.The vibration modes were also computed with these methods.The theoretical vibrational modes are in good agreement with its observed FT-IR spectrum of H 3 L NNN .Optimal uniform scaling factors were also computed for the H 3 L NNN compound.The three hybrid functions can be equally successful for vibrational spectrum predictions for the H 3 L NNN compound type derivatives.Taking small variations of the scaling factors into account for the derivatives of H 3 L NNN , one can recommend scaling factors of 0.9606, 0.9895, 0.9576, and 0.9034 for the B3LYP, BLYP, B3PW91, and HF methods, respectively, for future vibrational spectral assignments for unknown compounds of this class.Emrah ASLANTATAR 1 , Savita K. SHARMA 2 , Omar VILLANUEVA3, Cora E. MACBETH 2 , Ilkay GUMUS1,4and Hakan ARSLAN 1,2,4, * 1 Department of Chemistry, Faculty of Arts and Science, Mersin University, Mersin, Turkey 2 Department of Chemistry, Emory University, 1515 Dickey Drive, Atlanta, USA 3 School of Science and Technology, Georgia Gwinnett College, Lawrenceville, USA 4 Advanced Technology Research and Application Center, Mersin University, Mersin,

Table 4S .
Optimized and experimental geometries of H3L NNN in the ground state.*

Table 1S .
All bond lengths for H3L NNN .*Theatom-numberingscheme of the molecular structure is given in Figure1a.

Table 4S .
Optimized and experimental geometries of H3L NNN in the ground state.