Synthesis, Experimental and Theoretical Study of Azidochromones

A series of 2-(haloalkyl)-3-azidomethyl and 6-azido chromones has been synthetized, characterized and studied by theoretical (DFT calculations) and spectroscopic methods (UV-Vis, NMR). The crystal structure of 3-azidomethyl-2-difluoromethyl chromone, determined by X-ray diffraction methods, shows a planar framework due to extended π-bond delocalization. Its molecular packing is stabilized by F···H, N···H and O···H hydrogen bonds, π···π stacking and C–O···π intermolecular interactions. Moreover, AIM, NCI and Hirshfeld analysis evidenced that azido moiety has a significant role in the stabilization of crystal packing through weak intermolecular interactions, where analysis of electronic density suggested closed-shell (CS) interatomic interactions.


Conformational Analysis
Geometries of compounds 1-5 in the gas phase were studied, determined and calculated through a conformational analysis using the B3LYP/6-311++G(d,p) level of theory. Potential energy curves were performed by torsions around the C2-C2 and C3-C3 bonds that connect both exocyclic carbons C2 (bearing the halogen atoms) and C3 with the corresponding C2 and C3 carbon atoms of the heterocyclic ring C2 and C3, respectively. The most stable conformations of 1-5 and the main parameters are presented in Table 1 (See Figures S1-S4, Supplementary Materials).

NMR Spectroscopy
Experimental chemical shifts were compared with those derived from chemical calculations using the B3LYP/6-311+G(2d,p) level of theory (Table 2). For protons, a good agreement is observed with ∆ = TM exp − TM calc deviation ranging from −0.49 to 0.27 ppm, except for the CH 2 N 3 of 1 (∆ = −0.74). For carbon, the deviation range is between −32.2 and 10.9 ppm. The linear relationship between computed and experimental data, for all compounds, gives R-square values above 0.978 and 0.945 for protons (see Figure S5, Supplementary Materials) and carbons, respectively. The highest deviation was found for CF 2 Cl and CF 3 of 1 and 2 with ∆ values of −32.2 and −22.1 ppm, respectively; this suggests that the isotropic shielding of the fluorine atoms is underestimated by theoretical calculations, as previously reported [24][25][26][27]. Analysis of signal multiplicity and coupling constants was used, in conjunction with the calculated data, for structural elucidation.

Electronic Spectra
The calculated and experimental (using methanol as solvent (1: Table 3. Electronic spectra were calculated at the TD-B3LYP/6-311++G(d,p) level of theory, implicitly considering the influence of the solvent (methanol, ε = 32.7), with the conductor-like polar-izable continuum model (CPCM). The main experimental absorption bands were correlated with the calculated vertical electronic transitions, with oscillator strengths (f) > 0.076. The main molecular orbitals, involved in the electronic transitions of 1-5, are depicted in Figures S11-S13 (Supplementary Materials).
The most intense absorption band of 1, localized at 204 nm (calc. 211 nm), is attributed to HOMO − 1 → LUMO + 3 transitions principally due to excitations from π-bonding orbitals of the aromatic ring and non-bonding orbitals of N1 and N3 atoms to π*-orbitals of the benzene ring and p-type orbitals of one fluorine atom.

Hirshfeld Surface Analysis
To explore the features associated with the role of intermolecular contacts, a Hirshfeld surface analysis of 4 was performed (Figures 3 and 4) [32]. Figure 3a evidences contacts shorter than van der Waals, as highlighted by the red dots on the d norm surface, where the hydrogen and fluorine atoms of the -CH 2 N 3 , -CF 2 H and -C Ar -H moieties promote intermolecular contacts. In this sense, Tables S9 and S10 and Figure S14 (Supplementary Ma-terials) show the crystal lattice energy calculated with the CE-B3LYP/6− 31 G(d,p) energy model; for the C-H···F contacts, a character significative of contribution-dispersive (E dis. ) and -repulsive interaction (E rep. ) energies was observed. Particularly, the C3 -H3 A···F2 intermolecular contact displayed an important contribution with a relative high impact of dispersive, repulsive and total energy of −23.4 kcal/mol, 12.9 kcal/mol and −21.2 kcal/mol, respectively. The interactions' energy results are in agreement with those observed in 3-dibromomethyl-2-difluoromethylchromone [5]. Moreover, shape index and curvedness properties evidence π···π stacking and C-O···π interactions [33], which arise from planar stacking arrangements between the chromone rings. Features such as 'bow-tie' patterns of large red and blue triangles (see red circle in Figure 3b) and large green flat regions delineated by a blue outline (Figure 3c) reveal these close contacts associated with weak interactions [30].   Figure 4 shows the 2D-fingerprint plot of the main intermolecular contacts of 4. The pair of narrow spikes, labeled 1, corresponds to the shortest F···H distance associated to C-H···F interactions, while 2 evidences the N···H contacts that arise from C5-H5···N1 interactions. The F···H and N···H contacts have major contributions (23% and 22%) due to the relatively high proportion of fluorine and nitrogen atoms interacting in the crystal structure. Moreover, related chromones with -CF 3 , -CF 2 H moieties revealed a high proportion of weak F···H contacts that provide stability to the crystal structures [3,25]. On the other hand, labels 3 and 5 show H···H and C···C contacts with lower percentages of relative contri-bution (14% and 9%, respectively), related to π···π and C-H···π arrangements. Moreover, a moderate percentage of O···H (10%) hydrogen bonds, due to the oxygen atom of the carbonyl group and hydrogen of benzene moiety [30], is observed. A similar percentage of relative contribution was evidenced for N···F (8%) and C···O (6%) contacts with labels 6 and 7, whose characters will be discussed later through AIM and NCI analysis.
2.6. AIM and NCI Analysis of Intermolecular Contacts of 4 Figure 5 shows the self-assembled tetramer of compound 4, where the theory of atoms in molecules (AIM) has been included by visualization of the noncovalent interactions (NCI) by means of the critical points (CPs) and bond paths [34]. Seven CPs were taken into account in order to understand and reveal the weak interactions, where the reduced density gradient (RDG) [35] is visualized with a color. Isosurfaces of the H-bond and halogen bond (blue color), van der Waals interactions (green color) and strong repulsion areas (red color) are shown. In this sense, the seven CPs (1-7) illustrate vdW interactions, where the azide group and their three nitrogen atoms play a significant role, with large RDG isosurfaces (green color), and participate in several N···H and N···F contacts [36]. On the other hand, the combined AIM/NCI plot and topological parameters (Table 4) were calculated using the Multiwfn program [37], considering the main (3, −1) CPs according to Bader s theory of AIM [34]. Therefore, the topological criteria of electron density [ρ(r)] and Laplacian of the electron density ∇ 2 ρ showed values that suggest a weak interaction  Table 4 −V(r) G(r) < 1 and Hr > 0 .

General
All solvents and reagents were from Aldrich (St. Louis, MO, USA) and used without further purification. The melting points (uncorrected) were determined on a Büchi Melting Point M-560 (Büchi Labortechnik AG, Flawil, Switzerland). Infrared absorption spectra (KBr disk) were recorded on Varian 660-IR FT-IR (Agilent Technologies, Santa Clara, CA, USA) spectrometer with 2 cm −1 of resolution in the range of 4000-400 cm −1 (see Figure  S15a, Supplementary Materials). The Raman spectra of the solid were performed in the range of 3500-100 cm −1 at room temperature on a Thermoscientific DXR Raman microscope (Thermo Scientific, Madison, WI, USA) using a diode-pumped solid-state laser of 780 nm, with spectral resolution of 5 cm −1 (see Figure S15b, Supplementary Materials). The 1 H-, 19 F-and 13 C-NMR spectra were recorded at 25 • C on a Bruker Avance II 500 spectrometer (Bruker, Billerica, MA, USA) using CDCl 3 and CD 3 OD as solvent. Chemical shifts (δ) are expressed in ppm relative to TMS for 1 H-and 13
The bromine-substituted 3-polyhaloalkylchromone (0.33 mmol), sodium azide (1.26 mmol) and acetone (10 mL) were added to a 50 mL bottom flask with ground glass joint (Scheme 2). The reaction conditions for each compound are described in the Table S1 (Supplementary Materials). The end of reaction was monitored by TLC (hexane-EtOAc, 9:1). The mixture was filtered and washed with cold acetone, then crude product was dried under vacuum using a rotary evaporator and recrystallized in hexane to yield the pure compound.

X ray Diffraction Date and Structural Refinement of 4
The measurements were performed on an Oxford Xcalibur, Gemini, Eos CCD diffractometer (Agilent Technologies XRD Products, Yarnton, UK) with graphite-monochromated MoKα (λ = 0.71073 Å) radiation. X-ray diffraction intensities were collected (ω scans with ϑ and κ-offsets), integrated and scaled with CrysAlisPro [41] suite of programs. The unit-cell parameters were obtained by least-squares refinement (based on the angular setting for all collected reflections with intensities larger than seven times the standard deviation of measurement errors) using CrysAlisPro. Data were corrected empirically for absorption employing the multi-scan method implemented in CrysAlisPro. The non-H structures were solved by the intrinsic phasing procedure implemented in SHELXT [42] and the molecular model refined by full-matrix least-squares with SHELXL of the SHELX suite of programs [43]. The hydrogen atoms were determined in a Fourier difference map phased on the heavier atoms and refined at their found positions with isotropic displacement parameters. Crystal data and structure refinement results are summarized in Table S2 (Supplementary Material). Crystallographic structural data have been deposited at the Cambridge Crystallographic Data Centre (CCDC, Cambridge, UK). Any request to the Cambridge Crystallographic Data Centre for this material should quote the full literature citation and the reference number CCDC 2119625.

Computational Details
Quantum chemical calculations were performed for the ground state (gas phase) of 1-5 with the Gaussian 09 [44]. Scans of the potential energy surface were carried out with the B3LYP/6-311++G(d,p) level of theory. Potential energy curves were performed around the dihedral angles involving the nitrogen, chlorine and fluorine atoms (C3C2-C2 F, C4C3-C3 N, C2C3-C3 N and C2C3-C3 Cl) (see Figures S1-S4, Supplementary Materials).
The geometry optimizations calculations were carried out with the Density Functional Theory (B3LYP) method, employing the 6-311++G(d,p) basis set. In all cases, the calculated vibrational properties correspond to potential energy minima with no imaginary values for the frequencies. The 1 H-and 13 C-chemical shifts were calculated for the optimized geometries (B3LYP/6-311+G(2d,p)) using the GIAO method (Gauge Including Atomic Orbital), with the corresponding TMS shielding calculated at the same level of theory. The electronic transitions were calculated with the Time-Dependent Density Functional Theory (TD-DFT), implicitly considering the solvent effect (methanol).

Hirshfeld Surface Calculations of 4
The 2D-fingerprint plots of the Hirshfeld surface, energy frameworks and lattice interaction energies of 4 were generated using CrystalExplorer v17.5 software [45]. The incidence of each intermolecular interaction in the crystal was visualized and decoded from the 2D-fingerprint plot. The normalized contact distance surface (d norm ) based on d i and d e , (contact distances normalized by the van der Waals (vdW) radii) can be visualized by red spots, the 3D d norm surfaces are mapped over a fixed color scale of −0.243 au (red)-0.824 Å au (blue), shape index in the color range of −1.0 au (concave)-1.0 au (convex) Å and curvedness index in the range of 14−4.0 au (flat)-0.4 au (singular) Å. The surface properties mentioned above (d norm , shape and curvedness index) were used to identify planar stacking. Moreover, the intermolecular interaction energies of (4) were calculated using TONTO program, integrated in the CrystalExplorer v17.5 software. The interaction energies (Table S8 and Figure S26, Supplementary Materials) between the molecules are obtained using CE-B3LYP model (B3LYP/6-31G(d,p)).

Conclusions
Five azidochromones were fully characterized by NMR ( 1 H-, 13 C-, 19 F-) and electronic (UV-Vis) spectroscopies and GC-MS and UHPLC-MS/MS spectrometric methods. From the most stable conformations (theoretical calculations) for each of the compounds, the NMR and electronic spectra were simulated to aid the interpretation and compare these data with experimental results. A good accordance was found between experimental and calculated spectra. The crystal packing of 4 showed weak intermolecular interactions that promote the stabilization of the supramolecular assembly. Therefore, theoretical approaches such as Hirshfeld surface, AIM and NCI analysis were useful to understand the relative contributions of contacts and the character of weak intermolecular interactions in the context of electronic density. The analysis revealed that the F···H (23%), N···H (22%) and O···H (10%) contacts associated to -CF 2 H, -N 3 and -C=O groups and H···H (14%), C···C (9%) contacts associated with π···π stacking and C-O···π interactions are the main driving forces in crystal-packing formation. Moreover, the topological parameters studied by NCI/AIM methods evidenced a significant role of the azide (-N=N=N) group contributing through three nitrogen atoms to several bond-critical points. In addition, the reduced density gradient (RDG) characterized the weak intermolecular interactions as van der Waals interactions, where the electronic density can be considered as weak closed-shell (CS) interatomic interactions.