Molecular Structure Studies on Allyl Sulfonamides: Synthesis, Theoretical Treatment and Evaluation of Biological Activity

Departamento de Química, Universidade Federal de Viçosa, 36570-977 Viçosa-MG, Brazil Instituto de Química, Universidade Federal de Uberlândia, 38408-100 Uberlândia-MG, Brazil Instituto de Ciências Exatas, Universidade Federal de Minas Gerais, 31270-910 Belo Horizonte-MG, Brazil Instituto de Física e Química, Universidade Federal de Itajubá, 37500-906 Itajubá-MG, Brazil Departamento de Fitopatologia, Universidade Federal de Viçosa, 36570-977 Viçosa-MG, Brazil Instituto de Física de São Carlos, Universidade de São Paulo, 13566-590 São Carlos-SP, Brazil


Introduction
Sulfonamides are known mainly for their action against bacteria. 1 Nevertheless, structural modifications and additions of other synthetic blocks to sulfonamides can lead to compounds with different biological properties such as antiglaucoma, antifungal and aldose reductase inhibition. [2][3][4] The Morita-Baylis-Hillman (MBH) reaction is an important strategy for the formation of carboncarbon bonds in organic synthesis. 5 The reaction provides polyfunctionalized molecules which can be used as chemical intermediates in the construction of biologically active substances. 3,[6][7][8][9] This work presents synthetic and theorical approaches on allyl sulfonamides derived from MBH adducts (Scheme 1). The new allyl sulfonamides, bearing different aromatic substituents linked to the sulfonyl group, were characterized by high-resolution electrospray ionization mass spectrometry (HRMS-ESI), infrared (IR) and nuclear magnetic resonance (NMR) spectroscopies, and by density functional theory (DFT) calculations. In addition, the structures of compounds 1e and 2f (Scheme 1) were determined by X-ray diffraction, and Hirshfeld surface analyses were performed for intermolecular interaction studies.
The effects of the allyl sulfonamides 1a-1f and 2a-2f (Scheme 1) on the mycelial growth of Botrytis cinerea were evaluated in vitro. The screening for new antifungals is important to increase the structural variety of chemicals available for field applications to control fungal diseases, specially due to the constant emergence of resistant phenotypes. 10 B. cinerea, also known as gray mold, affects hundreds of plants species, being the main cause of losses in strawberry and grape crops. 11,12 It can also affect humans, occasionally causing occupational pneumonitis in wine makers and farmers. 13
After stirring under reflux for 2-6 h (monitored by TLC), the reactions were completed. Then, water (10 mL) was added and the product was extracted with 1,2-dichloroethane (3 × 10 mL). The organic phase was dried over sodium sulfate, filtered, and concentrated under reduced pressure. The crude material was purified by column chromatography on silica gel (hexane:ethyl acetate:dichloromethane, 3:1:3), yielding the products 1a-1f (43-72%) and 2a-2f (56-89%). The melting points and spectroscopic data for compounds 1b-1f and 2e were in accordance with the literature. 3  X-ray crystallography X-ray diffraction measurement of the compound 1e was performed on an Enraf-Nonius Kappa-CCD diffractometer (Delft, Holland) using graphite monochromated Mo Kα radiation (λ = 0.71073 Å) at room temperature (293 K). The collection software used was Bruker AXS Collect and the data processing was made with HKL Denzo-Scalepack program. 15 The structure was solved by direct methods using SIR-92 program and the model was refined by full-matrix least-squares on F 2 with SHELXL-2018. 16,17 Absorption correction (multi-scan) was applied for the compound using the program SORTAV. 18 The diffraction pattern of 2f was collected at room temperature (293 K) on a XtaLAB Mini diffractometer (Houston, USA) using Mo Kα radiation monochromated by graphite. Using Olex2, 19 the structure was solved with the SHELXT structure solution program using Intrinsic Phasing and refined with the SHELXL-2018 refinement package using least squares minimization. 17 For both compounds, all non-hydrogen atoms were refined using anisotropic displacement parameters. The hydrogen atoms in their calculated positions were refined using a riding model. Structural representations were drawn using ORTEP-3 and MERCURY. 20,21 Details of the unit cell, data collection and refinement are summarized in Table 1.
The complete data on the X-ray crystallographic analysis containing the atomic coordinates, bond lengths and bond angles have been deposited with the Cambridge Crystallographic Data Centre (CCDC No. 1936711 and 1997666, compounds 1e and 2f, respectively).

Hirshfeld surface
The Hirshfeld 3D surface was generated using a high (standard) surface resolution with the program Crystal Explorer 17.5. 22 The strength of the interactions was calculated by Hirshfeld d norm surface (normalized contact distance). The 2D fingerprint plots were prepared for the analysis of the relative contribution of different intermolecular interactions of compounds 1e and 2f. The 3D d norm surface was mapped using a color scale of −0.1000 to 1.0000. The shape index was mapped in the color range of −1.000 to 1.000 a.u. The 2D fingerprint plots were displayed using with the d e and d i distance scales displayed on the graph axes (0.6-2.8 Å). All hydrogen bond lengths were automatically modified to typical standard neutron values (C-H of 1.083 Å).

Energy frameworks and lattice energy calculations
Energy framework analysis was used to explore the intermolecular interaction energies between the molecules of the cluster within 3.8 Å. These calculations were  performed using the Crystal Explorer 17.5 at Accurate mode (B3LYP (Becke, 3-parameter, Lee-Yang-Parr) using 6-31G(d,p) basis) (high performance), 22 with total interaction energy between any nearest neighbor molecular pairs given in terms of four components: electrostatic, polarization, dispersion, and repulsion, with scale factors of 1.057, 0.74, 0.871 and 0.618, respectively. The lattice energy calculation was done with cluster within 20 Å, using the same scale factors.

Computational calculations
DFT calculations were carried out using the B3LYP functional, 23,24 valence triple-zeta polarization (Def2-TZVP) basis set, 25 and Grimme's dispersion correction in Orca 4.0. 26,27 Structures were optimized using a 10 -6 a.u. optimization criteria in forces and 10 -8 a.u. in energy. Harmonic frequency analyses were carried out in the optimized structures to confirm that they are local minima in the potential energy surface. All calculations considered the solvent effect included through conductor-like polarizable continuum model (CPCM), 28 using a dielectric constant (ε) of 10.36, which corresponds to 1,2-dichloroethane, the solvent used in the syntheses. The molecular orbitals were plotted using an isovalue of 0.03 eau -3 .

Biological assay
The antifungal activity of the allyl sulfonamides was evaluated against B. cinerea by the poisoned food technique. 3 The potato dextrose agar (PDA) culture medium (Kasvi, São José dos Pinhais, Brazil) was previously sterilized in autoclave (Stermax, Pinhais, Brazil) for 20 min at 121 ºC. Glassware and spatulas were sterilized at 140 ºC for 3.5 h. The fungus was isolated from infected strawberry tissues with gray mold symptoms. Discs of B. cinerea mycelia (diameter of 7.4 mm) were placed on the center of Petri dishes containing 15 mL of PDA homogeneously mixed with the allyl sulfonamides 1a-1f or 2a-2f at three different concentrations (0.5, 1.5 and 3.0 mmol L -1 ), and dimethyl sulfoxide (DMSO) and Tween 80 (LabSynth, Diadema, Brazil), 1% v/v each. Each sample was prepared in four repetitions and the dishes were kept in the incubator chamber (EthickTechnology, Vargem Grande Paulista, Brazil) at 22 ºC for three days. The control (negative check treatment) was prepared with PDA, DMSO and Tween 80 only. The diameter of the colony was measured with the aid of a digital caliper (Lee tools, Santo André, Brazil) on the third day of incubation. The percentages of inhibition were calculated in comparison with the control.

Synthesis and characterization
The twelve allyl sulfonamides 1a-1f and 2a-2f were prepared using the methodology shown in Scheme 1. The molecular formulae of the six unpublished allyl sulfonamides (1a, 2a-2d and 2f) were confirmed by HRMS-ESI, which presented the expected molecular ion peaks in the negative mode. The spectroscopic data obtained for 1b-1f and 2e are in accordance with the literature. 3,29,30 Characteristic bands in the IR spectra of the allyl sulfonamides 1a-1f and 2a-2f indicated the presence of the most relevant groups within each structure. All spectra showed bands at 3301-3234, 1338-1307 and 1172-1139 cm -1 due to the N-H and the SO 2 groups (νNH, νSO 2as and νSO 2sym , respectively). The νC≡N band (2222-2207 cm -1 ), observed in the spectra of 1a-1f, and the νC=O band (1728-1700 cm -1 ), present in the spectra of 2a-2f, were the most important bands for the differentiation of the allyl sulfonamides series 1 and 2.
In the 1 H NMR spectra of the MBH adducts (1 and 2, Scheme 1), two signals due to the sp 2 -CH 2 methylenic hydrogens (H1) are observed at d 6.02 and 6.10 when Y = CN, and at d 5.81 and 6.32 when Y = CO 2 Me. 14 In the spectra of 1a-1f and 2a-2f, these signals are substituted by one doublet at ca. d 4 (H1) confirming the sp 3 -CH 2 -N bond formed upon the substitution reaction. Also, the signal for H3 at d 5.28 or 5.54 in the spectra of 1 and 2 (Y = CN or CO 2 Me, respectively) is shifted to d 7.3-7.7 in the spectra of the allyl sulfonamides, showing that the double bond moved to C2=C3, with the elimination of the water molecule. Thus, the signals for C3 and C1 are observed at d 138-146 and d 40-47, respectively, in the 13 C NMR spectra of the allyl sulfonamides.
The C=O signal is observed at ca. d 168 in the spectra of the compounds 2a-2f and the CN signal appears at ca.

X-ray crystallography
To gain a deeper insight into the structures of the allyl sulfonamides one compound of each series was investigated by single crystal X-ray diffraction. Thus, prism shaped white crystals of 1e and 2f were grown in chloroform/petroleum ether and acetone/water, respectively. The compound 1e crystallizes in the centrosymmetric space group Pcab of the orthorhombic system with eight molecules in the unit cell, while 2f crystallizes in the centrosymmetric space group P2 1 /c of the monoclinic system with four molecules in the unit cell. The crystallographic asymmetric units of the compounds 1e and 2f are shown in Figure 1.
The X-ray data confirmed the proposed Z-configuration for compound 1e, and the E-configuration for 2f ( Figure 1). The structural conformation of 1e shows that the C2-C10≡N2 group is in the same direction of the SO 2 group, with a dihedral angle of just 5.90(1)° between the least-square plane through C1'SN1 groups and the line through C2C10. On the other hand, the ester group in 2f is in the opposite direction with respect to the SO 2 group, the same dihedral angle being 139.68(1)°. Table 2 shows selected bond lengths and angles for 1e and 2f. In order to check the molecular conformation parameters, both crystal structures were analyzed by Mogul software, 31 within The Cambridge Structural Database (CSD), 32 which allows us to compare all the bond lengths and angles of both structures with similar parameters found in the CSD for chemically equivalent compounds ( Figures S1-S7, in the Supplementary Information (SI) section). This study shows that in both compounds the sulfur present a distorted tetrahedral geometry with S-C bond length very close to the expected mean value, 1.760 Å ( Figure S1). The same does not occur with the other values. In this way, even when the S=O bond lengths in 1e are close to the mean value (1.431 Å), the distance S-O1 in 2f is considerable smaller (1.417(4) Å) showing the influences of the intermolecular interaction ( Figure S2). The opposite behavior is observed in the case of the S-N bond. Both compounds present S-N values larger the mean value, 1.611 Å. However, the difference is higher for 1e. This compound also presents a N1-C1 bond length significantly larger than the mean value (1.460 Å) in agreement with a clear intramolecular charge transfer to the SO 2 group induced by the intermolecular interaction ( Figure S5).
The C1-C2 distance of ca. 1.505 Å is typical of a single bond, and the values observed for C2=C3 (ca. 1.310(3) Å) and C3-C4 (ca. 1.470 Å) are consistent with the conjugated π system. 33 The C-C bond lengths in the phenyl ring 1 (C1' to C6') and the phenyl ring 2 (C4 to C9) ranging from 1.366(9) to 1.402(3) Å are in the range of related aromatic compounds. 3 In the compound 1e (Table 2), the C10≡N2 bond length of 1.147(3) Å confirms the value for a triple bond, while the C1-N1 bond distance of 1.481(3) Å is in the range of a single bond. 34 The C2-C10 distance (1.439(3) Å) is consistent with reported values for analogous cyano allyl sulfonamides. 3,34 The C4'-C7' distance is close to values observed in similar compounds. 34,35 In the compound 2f (Table 2), the C10-O3 is a formal double bond, the C10-O4 bond presents intermediate character, while the C11-O4 bond length is in the normal range for a C-O single bond, as observed in related compounds. 9,36 The C2-C10 bond in 2f (1.483 (8)) is longer than in the compound 1e due to the stronger electron-withdrawing effect of the nitrile compared to the carboxyl group.
In both compounds, the phenyl rings are essentially planar with root-mean-square deviations (r.m.s.d.) of 0.0077 Å for 1e and 0.0055 Å for 2f (ring 1), and of 0.0014 Å for 1e and 0.0055 Å for 2f (ring 2) from the leastsquares plane defined by the atoms. The dihedral angle between these planes are 77.0(1)° for 1e and 86.7(1)° for 2f. In the compound 1e, the deviation of the C7' atom to the ring 2 is of 0.015(4) Å. The C1-N1-S angle is greater than the value expected for the sp 3 N atom. In both compounds, the torsion angles C1'-S-N1-C1 and N1-C1-C2-C10 are significantly different (Table 2). These angles around the S-N1 and C1-C2 bonds describe the conformation of fragments in relation to phenyl rings.
The crystal packings of both compounds are stabilized by C-H⋯O intramolecular interactions as well as C-H⋯O and C-H⋯π intermolecular interactions. The crystal packing of 1e also presents C-H⋯N hydrogen bonds (Table 3). In the compound 1e, the C6-H6⋯N2 and C9-H9⋯O2 intermolecular interactions form dimmers connected by translation in the a axis direction, generating (11) ring motifs. 37 These dimmers form chains in the a axis direction (Figure 2). Furthermore, two C-H⋯π interactions link these chains in the c axis direction into a threedimensional supramolecular network. In the compound 2f, the C5'-H5'⋯O3 intermolecular interactions form dimmers between molecules related by an inversion center, generating (20) ring motifs. These dimmers are interlinked by C6'-H6'⋯O2 interactions which also form dimmers between molecules related by an inversion center, generating (10) ring motifs, forming chains in the c axis direction (Figure 3). Furthermore, C-H⋯π interactions link these chains in the a axis direction into a bidimensional layers.
In order to visualize the participation of intermolecular interactions in the crystal structures of compounds 1e and 2f, the atom⋯atom interactions were further investigated by shape index and 2D fingerprint plots. The shape index depicted in Figure 6, allows to identify the complementarity between molecules in the crystal structure. Thus, the red dots indicate closer atoms, indicating interactions between neighboring molecules. These regions suggest significant C-H⋯π and C-H⋯O interactions, which are consistent with the X-ray analysis. The 2D fingerprint plots (Figure 7) are useful for analyzing the relative contribution of different intermolecular interactions in the crystal structures of 1e and 2f. These data show that the crystal structure of 1e is determined by H⋯H, C⋯H, H⋯C, O⋯H, H⋯O, N⋯H and H⋯N contacts, whereas the structure of 2f is determined by H⋯H, C⋯H, H⋯C, O⋯H and H⋯O contacts. The Figure 8 summarizes the selected percentages of contacts in the crystal structures of 1e and 2f. The non-classical H-H interactions constitute the most significant contribution to the total HS (39.1 and 46.6%, respectively for 1e and 2f). It is also important to highlight the N⋯H/H⋯N (13.0% in 1e and ca. 0% in 2f) and O⋯H/H⋯O (17.5% in 1e and 24.8% in 2f) contributions to the crystal packaging. The inspection of contacts between the other types of atoms pointed out that there are no significant π⋯π interactions within the crystals (C⋯C contacts make 3.1 and 1.7% of the surface area, respectively for 1e and 2f).

Energy framework calculation
The energy partitioning showed that the maximum contribution to the lattice stabilization is coming from the dispersion component, with contributions around 55% for 1e and 53% for 2f. The coulomb component contributes with ca. 34% in both compounds, the remaining contribution coming from the polarization energy. The   electrostatic, polarization, dispersion, repulsion and total energy of interaction of these compounds are listed in Table 4. The Figures S8 and S9 in the SI section show these different intermolecular interactions, with the geometrical parameters and interaction energies partitioned into the energy components.
Through the analysis of the energy structure diagrams (Figures 9 and 10), it is possible to observe that the coulomb and dispersion terms of energy frameworks for both compounds indicate that they are similar. However, the dispersion term is more dominant when compared to the coulomb term.
The lattice energy (-336.5 and -333.6 kJ mol -1 , for 1e and 2f, respectively) showed that the expected relationship between the calculated lattice energy and the melting points are practically preserved. Thus, the highest energy in the network corresponds to the highest melting point, in this case, of the compound 1e.

Molecular modeling
DFT calculations were carried out in the presence of a reaction field to simulate the solvent effect and provide an understanding about the geometrical structures in     1,2-dichloroethane. The DFT calculations agree with the X-ray results. The main geometrical parameters for compounds 1a-1f and 2a-2f are highlighted in Table S1, in the SI section. The complete set of optimized structures for 1a-1f and 2a-2f are also provided (Figures S10 and S11) together with the optimized cartesian coordinates (Table S2). The atom numbering is the same used in the ORTEP diagram in Figure 1, for clarity. The frontiers orbitals were also evaluated by the DFT calculations. For the compounds 1a-1f, the gap between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) is around 4.5-4.6 eV, with exception of 1a, which gap is of 4.3 eV (Figure 11). The frontiers orbitals for the compound 1c shown in Figure 12 are similar to those calculated for 1b-1f. The results indicate that both HOMO and LUMO orbitals are typically antibonding orbitals, located in the aromatic ring and the nitrile. For compound 1a, the HOMO orbital has some contribution from the p-orbital of iodine, as shown in Figure 12b, and there is a node plane between the iodine atom and the aromatic ring in this molecular orbital. This could be related to the different reactivity of this molecule towards the biological targets. It is also interesting to note the concentration of the LUMO orbital in the nitrile, indicating that this is the more susceptible site to a nucleophilic attack.
The calculated HOMO-LUMO gap for the esters 2a-2f is in the range of 4.6-4.8 eV ( Figure 11) and the frontier orbitals are similar for all the components of the series, being antibonding orbitals located in the aromatic ring and the ester group, as exemplified in Figure 13 for the compound 2c.
The synthetic route favors the formation of the diastereomers Z for the cyano compounds 1a-1f and the E configuration for the ester compounds 2a-2f. The DFT calculations confirm these isomers as the most stables in each series (Table 5).
It is important to highlight that the difference in energy between the E and Z forms for the series 2a-2f is small, with the E isomer being, in most cases, less than 10 kJ mol -1 more stable than the Z isomer. In the series 1a-1f, compounds 1b and 1c are remarkably more stable in the Z form (Table 5).

Antifungal assay
The cyano-sulfonamides 1b, 1c, 1d and 1f are active in vitro against Colletotrichum gloeosporioides, the causal agent of anthracnose in plants. They inhibit from 15% (1b) to 48% (1f) the mycelial growth of C. gloeosporioides, at the concentration of 1.5 mmol L -1 . 3 In this work, the influences of the twelve allyl sulfonamides in the mycelial Figure 11. Diagram of frontiers molecular orbitals calculated at B3LYP/ def2-TZVP level of theory with the HOMO-LUMO gap indicated in eV. The energy of each orbital is indicated in relation to HOMO, which was set to zero.  growth of B. cinerea were investigated through a similar methodology. Among the necrotrophic and polyphage fungi, B. cinerea is one of the most studied, 38 especially due to the occurrence of resistance to available fungicides. 10,39 The bioassay results are summarized in Table 6.
It is clear that the inhibition was dose dependent for all compounds. The activity becomes prominent at 3 mmol L -1 , where the inhibition percentages are close or superior to 50%. The results for each concentration are similar, showing that the variation of the groups X (I, Br, Cl, F, CH 3 or H) is of minor importance for the observed activity. Nevertheless, at 3 mmol L -1 , it was possible to differentiate the behavior of 1d and 1e, which presented the best results within the series 1a-1f. The compounds 2d and 2e, are also among the most actives (2a, 2d-2f) in this series, although the results were more homogeneous within the 2a-2f series.
It seems that the ester group has a greater impact on the antifungal activity towards B. cinerea than the nitrile, as the results of the series 2a-2f were, in general, superior to those of the series 1a-1f, at 3 mmol L -1 . This behavior could be related to the differences in conformation of those groups (trans to the SO 2 group in 2f and cis in 1e in relation to the ester and cyano groups, respectively) as discussed in the X-ray crystallography sessions. In lower concentrations, the differentiation of the various compounds was less evident.

Conclusions
The reactions between MBH adducts and primary sulfonamides in the presence of sulfuric acid furnish allyl sulfonamides in good yields. The compounds 1a-1f and 2a-2f were synthetized and characterized by spectroscopy techniques and investigated by molecular modeling, and compounds 1e and 2f were also analyzed by single crystal X-ray. The DFT and X-ray diffraction studies confirmed that the reactions are stereospecific, providing the Z-configuration for the series of the cyanosulfonamides 1a-1f and E-configuration for the series of the carbomethoxy-sulfonamides 2a-2f. Moreover, theoretical calculations confirmed that both HOMO and LUMO orbitals of these allyl sulfonamides are typically antibonding orbitals, located in the aromatic ring and in the nitrile or the ester group, with a unique difference observed for 1a, which has a larger contribution of the iodine p orbital in the HOMO. The energy framework studies showed greater significance of the dispersion term and the calculated lattice energy is in agreement with the melting points. The crystal packing of 1e is stabilized by C-H···O and C-H···π interactions while 2f is stabilized by C-H···O, C-H···N and C-H···π interactions. The 2D fingerprint plots showed that in 1e, the N⋯H/H⋯N and O⋯H/H⋯O interactions are the most important contacts for the crystal lattice stability, with a contribution of 30.5% for the HS area while in 2f, the H···O/O···H contacts are important for the crystal lattice, with a contribution of 24.8% for the HS area. The energy partitioning showed that the maximum contribution to the lattice stabilization is coming from the dispersion component with contribution greater than 50% while the coulomb component contributes around 34%. The calculated lattice energy is in agreement with the melting points. The compounds inhibited the mycelial growth of the pathogenic fungus B. cinerea, with emphasis on the series 2a-2f (50-60% inhibition) and the compounds 1d-1e (52-60% inhibition) at 3 mmol L -1 in vitro. The differences in the molecular configuration in these groups (E for 2a-2f and Z for 1a-1f) led to different intermolecular interactions, which might be related to the differences observed in the biological activity evaluated. Thus, the class of allyl sulfonamides present a potential application for the control of fungal diseases and is worth of further investigation.

Supplementary Information
Crystallographic data (excluding structure factors) for the structures in this work were deposited in the Cambridge Crystallographic Data Centre as supplementary publication number CCDC 1936711 and CCDC 1997666, compounds 1e and 2f, respectively. Copies of the data can be obtained, free of charge, via https://www.ccdc.cam. ac.uk/structures/.