N-Cycloamino substituent effects on the packing architecture of ortho-sulfanilamide molecular crystals and their in silico carbonic anhydrase II and IX inhibitory activities

Three o-nitrosulfonamides and three N-cycloamino-o-sulfanilamides have been successfully synthesized and characterized, and the intermolecular interactions analysed, as well as being tested in silico for carbonic anhydrase II (4iwz) and IX (5fl4) inhibitory activities. The results obtained from crystal packing and DFT analysis suggest that the molecules are held together by forces such as hydrogen bonding and π–π interactions.


Introduction
Sulfanilamide (4-aminobenzenesulfonamide) is aptly described as the antecedent of the group of therapeutics known as 'sulfa drugs', which ushered in the modern era of antibacterial chemotherapy (Ajani et al., 2012). Although it had been a component of a staple azo dye in the colour industry since the beginning of the 20th century, it did not gain prominence in medicine until the 1930s when Gerhard Domagk and coworkers patented Prontosil, A (Fig. 1), a sulfanilamide prodrug, which not only revolutionized the treatment of bacterial infections, but chemotherapy as a whole, and led to the development of other drugs for non-infectious diseases.
It has been established that the bacteriostatic properties of sulfanilamides ( Fig. 1) are predicated based on two major motifs: the aryl amine (-NH 2 ) and sulfonamide (-SO 2 NHR) groups (Lesch, 2007). A free or hydrolysable substituted amino (-NHR 0 ) moiety that is para to the sulfonamido group has been reported to be crucial for antibacterial activity, whereas modification of the position to the ortho and/or meta position results in non-antibacterial activities (Ajani et al., 2012). The derivatization of the sulfonamido group with heterocycles has also produced more potent antibiotics (Ajani et al., 2012;Lesch, 2007). In addition, it has been long reported that no correlations exist between the toxicities and therapeutic efficiencies, as well as toxicities and solubilities, of the three isomers of sulfanilamide, as evidenced by the finding that even though meta-sulfanilamide C was the least toxic of the three, only para-sulfanilamide B possessed bacteriostatic activity (Laug & Morris, 1939). Notably, the inhibitions of the Helicobacter pylori -class carbonic anhydrase (hpCA) (Nishimori et al., 2006) and tumour-associated transmembrane carbonic anhydrase IX (CA IX) (Vullo et al., 2003) isozymes have been observed with ortho-sulfanilamide D (orthanilamide). Sulfonamides E are derivatives of sulfanilamide and remain an important class of drugs, with antibacterial and nonantibacterial potencies, such as diuretic, antimicrobial, antiepileptic, antileprotic, antimalarial, hypoglycemic, antiretroviral, antithyroid and anti-inflammatory activities (Gul et al., 2016;Henry, 1943;Casini et al., 2002;Mohan et al., 2006;Alex & Storer, 2010).
They also inhibit carbonic anhydrase (Gul et al., 2016;Ghorab et al., 2014;Nocentini et al., 2016) and have been reported to show in vivo and/or in vitro antitumour activities (Boyland, 1946). Many of these sulfonamide-based (sulfa) drugs, reported to be in clinical trials, are devoid of the side effects plaguing most of the current pharmacological agents (Casini et al., 2002;Owa et al., 2002;Lavanya, 2017;Andreucci et al., 2019).

Synthesis and crystallization
2.2.1. Synthesis of N-cycloamino-o-nitrobenzenesulfonamides 1-3. o-Nitrobenzenesulfonyl chloride (1.00 mmol) was added slowly to a stirring dried toluene solution (30 ml) of the cycloamine (2.20 mmol) at ambient temperature and stirred for 12 h, monitored by TLC. The reaction mixture was then diluted with dichloromethane (30 ml) and washed with distilled water (3 Â 10 ml). The organic layer was separated, dried over anhydrous sodium sulfate, filtered and concentrated to an oil, which was purified by column chromatography on silica gel (dichloromethane/n-hexane, 2:1 v/v). Crystals were obtained by the slow solvent evaporation of the requisite eluates at ambient temperature, except for 5, which was recrystallized from dichloromethane, slowly evaporated and filtered to give single crystals.
2.2.2. N-Cycloamino-o-sulfanilamides 4-6. An evacuated nitrogen-gas-filled round-bottomed flask was charged with N-cycloamino-o-nitrobenzenesulfonamides 1-3 (15.63 mmol) dissolved in ethanol (30 ml), at ambient temperature, and 10% palladium-on-charcoal catalyst (3.35 mol%) was added, with stirring. Hydrogen gas was then introduced via a balloon and stirring continued at ambient temperature for 12 h. The reaction mixture was filtered and the solvent was evaporated in vacuo. The resulting residue was extracted into dichloromethane (50 ml), dried over anhydrous sodium sulfate, filtered and concentrated under reduced pressure to afford an oil, which was purified on a silica-gel column using dichloromethane and n-hexane (2:1 v/v). Crystals were obtained via slow solvent evaporation of the eluates at ambient temperature.
2.3.2. Protein preparation. The protein structures of 4iwz and 5fl4, with resolutions of 1.60 and 1.82 Å , respectively, were downloaded from the Research Collaboratory for Structural Bioinformatics (RCSB) Protein Data Bank (PDB). Retrieved crystal coordinates were prepared in the 'Protein Preparation Wizard' of the Schrö dinger Suite (Schrö dinger, 2022), with default parameters of assigning bond orders, optimizing and minimization using OPLS4. A receptor grid generation module was applied to the prepared proteins by selecting the corresponding cocrystallized ligand to define the binding site. A default parameter for the radii of van der Waals having a scaling factor of 1 Å with a partial charge cut-off of 0.25 Å was used (Panwar & Singh, 2021;Yang et al., 2022).
2.3.3. Molecular docking. Docking calculations were executed in the extra precision (XP) mode of the Glide module in the molecular modelling platform of the Schrödinger Suite (Schrö dinger, 2022). The complexes with the highest negative docking scores have better binding towards the respective proteins 4iwz and 5fl4. Docking calculations of the synthesized N-cycloamino derivatives against the hCA II (PDB entry 5fl4) and XII (4iw7) isoforms will provide a selectivity profile that may be interesting for the development of novel anticancer agents with limited side effects. The hCA II (PDB entry 5fl4) and XII (4iw7) carbonic anhydrase isoforms have recently emerged as excellent targets for the design of novel therapeutic strategies for cancer, due to their involvement in the survival of tumour cells, as well as in the insurgence of resistance to classical anticancer protocols (Milite et al., 2019).

DFT calculations
Theoretical studies were performed for compounds 1-6 whereupon the SC-XRD structures of the compounds were used for optimization and global reactivity descriptor (GRD) calculations. Computational studies and molecular electrostatic potential (MEP) for 1-6 were carried out using the GAUSSIAN16 software package (Frisch et al., 2016), whereas the calculations were performed using the standard hybrid density functional method (B3LYP) with a basis set of the 6-311++G**(p,d) level (Becke, 1993). Optimized molecules were obtained with the Chemcraft visualization program (https://www.chemcraftprog.com/).

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 1. Carbon-bound H atoms were added in idealized geometrical positions in a riding model. Nitrogen-bound H atoms were located in a difference map and refined freely.

Chemistry
The N-cycloamino-o-sulfanilamides 4-6 were prepared via a two-step reaction, starting from the condensation reaction of o-nitrobenzenesulfonyl chloride with alicyclic amines in toluene, at ambient temperature, to afford N-cycloamino-onitrobenzenesulfonamide adducts 1-3 (Scheme S1 in the supporting information). The use of toluene as a nonpolar reaction medium was, amongst other reasons, to drive the forward reaction. In the second step, adducts 1-3 were hydrogenated with hydrogen gas, in ethanol at ambient temperature, in the presence of 10% palladium-on-activated charcoal catalyst to give the target N-cycloamino-o-sulfanilamides 4-6 in 72-86% yield. The reactions were monitored by TLC.

research papers
All the compounds synthesized were characterized by their melting points and IR, 1 H/ 13 C NMR and MS spectra. In the IR spectra of o-nitrosulfonamide adducts 1-3, the strong absorption bands observed at 1355-1342 and 1171-1161 cm À1 were ascribed to the asymmetric and symmetric stretching frequencies, respectively, of the SO 2 -N moiety, thereby  alluding to the formation of the sulfonamide bond. The disappearances of the SO 2 -Cl (1420 and 1220 cm À1 ) and N-H (3286-3265 cm À1 ) stretching bands in the IR spectra of o-nitrobenzenesulfonyl chloride and cycloamines, respectively, were good indicators of a successful condensation reaction. This was corroborated by the shift of the sulfonyl (-SO 2 -) absorption bands from 1420 and 1220 (in o-nitrobenzenesulfonyl chloride) to 1355-1342 and 1171-1161 cm À1 (in 1-3). It is noteworthy that the lower wavenumbers observed in the IR spectra of o-nitrosulfonamides 1-3 for -SO 2 -were not unusual as the Cl atom bonded to it had been replaced by a less electronegative N atom. In the IR spectra of o-sulfanilamides 4-6, the appearance of two N-H stretching bands in the higher frequency region around 3467 AE 20 and 3383 AE 10 cm À1 , and the disappearance of the nitro (NO 2 ) absorption bands (observed at 1550-1536 and 1369-1342 cm À1 ) in the spectra of 1-3 were attributed to the successful catalytic reduction of the nitro group to the amino group.
The 1 H NMR spectra of o-nitrosulfonamides 1-3 were additive of the individual spectra of the starting materials (i.e. o-nitrobenzenesulfonyl chloride and cycloamines), with the disappearance of the nitrogen proton peaks of cycloamines. The aromatic protons of o-sulfanilamides 4-6 resonated upfield in comparison to the same aromatic protons in precursors 1-3. This general shift towards tetramethylsilane (TMS) was credited to the newly formed amino groups whose lone-pair electrons are suspected of having caused the increased mesomeric shielding of the aromatic protons. D 2 Oexchangeable singlets were also observed in the 1 H NMR spectra of 4-6 between 5.13 and 4.99 ppm for the newlyformed amino protons. The success of the catalytic hydrogenation of nitro adducts 1-3 was corroborated by the 13 C NMR spectra of 4-6, where the requisite C atoms (C-NO 2 ! C-NH 2 ) resonated upfield in the range 133.9-130.1 ppm. The spectroscopic data analyses of the synthesized compounds were consistent with the assigned structures of the compounds. Table 2 Hydrogen-bond, C-HÁ Á Á( ring) andstacking interaction geometry (Å , ) for the crystal structures of p-sulfanilamide (B), o-sulfanilamide (D), o-nitrosulfonamides 1-3 and N-cycloamino-o-sulfanilamides 4-6.

Crystal structure
The molecules of 1-3 and 4-6 crystallized in the monoclinic space group P2 1 /n or P2 1 /c (No. 14), except for 5, which crystallized in the orthorhombic space group Pbca (No. 61). In addition, they all had one molecule in the asymmetric unit, with the exception of 4, with two independent molecules per asymmetric unit cell. The two molecules per unit cell of compound 4 were identical but for the conformation of the pyrrolidine group (cf. Fig. S1 in the supporting information). It is noteworthy that the pyrrolidine ring in 1 is disordered. The molecular structures of 1-3 and 4-6 are shown in Fig. 2, while the crystal data collection parameters of o-nitrosulfonamides 1-3 and N-cycloamino-o-sulfanilamides 4-6 are presented in Table 1. They are compared with the crystal structure data of para-sulfanilamide and ortho-sulfanilamide, which crystallize in the orthorhombic Pbca (No. 61) and monoclinic P2 1 /c (No. 14) space groups, respectively (Gelbrich et al., 2008;Shad et al., 2008). Several sulfonamide derivatives have also been reported (El-Gaby et al., 2020). Selected bond lengths and angles, as determined from the SC-XRD experiments, are collected in Table S1 (see supporting information).
It is instructive to note that the amino (NH 2 ) group in N-cycloamino-o-sulfanilamides 4-6 contributed significantly to their hydrogen-bond interactions (cf. Table 2). In all three structures, there were intramolecular N-HÁ Á ÁO S interactions resulting in ring closures that can be described with S(6) graph-set descriptors (Bernstein et al., 1995;Etter et al., 1990). Furthermore, compounds 5 and 6 exhibited infinitechain intermolecular N-HÁ Á ÁO S interactions with C(6) descriptors. Interestingly, no infinite chain interaction was observed in 4; instead, four molecules were linked into a ring structure with an R 4 4 (24) descriptor. The p-sulfanilamide (Gelbrich et al., 2008) and o-sulfanilamide (Shad et al., 2008) structures also each have a number of infinite-chain interactions and ring structures. Fig. 3 shows selected hydrogen-bond, C-HÁ Á Á( ring) andstacking interactions for sulfonamides 1-3 and sulfanilamides 4-6. All the hydrogen bonds were of moderate (mostly electrostatic) strength (Jeffrey, 1997), with 4 giving the strongest hydrogen bonds (Table 2). Additionally, the compounds also exhibited both intra-and intermolecular C-HÁ Á ÁO S interactions, with the length of the shortest interactions varying in the range 2.30-2.48 Å .
The onlystacking interaction of note occurred in 3, where two centroid-to-centroid interactions with distances of 3.6967 (11) Å were observed between the centrosymmetric indoline moieties. An N OÁ Á Á ring interaction of 3.657 (2) Å was also evident in 3, whereas intermolecular C-HÁ Á Á( ring) interactions of 2.97 Å and S OÁ Á Á( ring) interactions of 3.5773 (15) Å were present in the structure of its hydrogenated analogue 6. The packing diagrams of the crystal structures of compounds 1-6 are shown in Fig. S2 in the supporting information.

Hirshfeld surface analysis
The Hirshfeld surface analyses (Turner et al., 2017) Table 3 Percentage contributions of selected interatomic contacts to the Hirshfeld surface of compounds 1-6.
large circular depressions (deep red) visible on the d norm surfaces typically indicate that the molecule has a donor site(s) (e.g. amine and/or sulfone) or interactions with proteins. Fingerprint plots of o-nitrosulfonamides 1-3 and N-cycloamino-o-sulfanilamides 4-6 in full and resolved into CÁ Á ÁH, OÁ Á ÁH and NÁ Á ÁH are presented in Fig. S4 (supporting information). The intermolecular OÁ Á ÁH and NÁ Á ÁH interactions appear as two distinct spikes of almost equal length in the 2D (two-dimensional) fingerprint plots in the region 1.2 < (d e + d i ) < 2.9 Å as light-sky-blue patterns in full fingerprint 2D plots and characterized to be 2.56 AE 0.21 Å corresponds to OÁ Á ÁH contacts which contributes the majority of the surface area. 2D fingerprint plots reveal the contributions of these interactions in the crystal structure quantitatively and are presented in Table 4 (with minimum and maximum values of d norm , d i and d e provided). Complementary regions are also visible in the fingerprint plots (Fig. S4), where one molecule acts as a donor (d e > d i ) and the other acts as an acceptor (d e < d i ). This finding was validated by the calculated molecular electrostatic potential of 1-6 ( Fig. S5). The negative potential (acceptor) is indicated as a red surface around the O atoms attached to sulfur (-SO 2 ) and the N atoms attached to oxygen (-NO 2 ). The blue/purple surface area indicates that the positive potential (donor) is mapped in the proximity of the H atoms (Fig. S5).

Global reactivity descriptors (GRDs)
The full geometry of optimized molecules 1-6 presented bond lengths similar to those obtained from the crystal data. A comparison of selected torsion angles of the crystal structures of 1-6 and the DFT-optimized molecules showed that con-  Table 5 Frontier molecular orbital (FMO) energies of synthesized compounds 1-6.  Frontier molecular orbitals for the optimized structures of 1-6.
formation of the molecules did not change significantly in the DFT-optimized state (Fig. S6). Generally, the observed, almost flat, O-S-N-C torsion angle of the DFT-optimized molecules suggest that the lone pairs on sulfur may have contributed to the -electron delocalization that is observed in the DFT molecules. The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) electrons are distributed around various moieties within the various molecules (Fig. 4). Generally, electron distribution is mainly scattered in the HOMO over the phenyl, sulfur and indolinyl/ pyrrolidinyl rings, with the exception of 3 and 5. The LUMO is mainly spread over the phenyl moieties. This indicates that there is a transfer of charge between the indolinyl/pyrrolidinyl rings and the phenyl moieties within the molecule.
The HOMO-LUMO gap, which describes the stability of molecules and predicts reactivity between species by providing the electrical transport properties, as well as electron carrier and mobility in molecules (Rathi et al., 2020), are provided in Table 5. N-Indolinyl-o-nitrobenzenesulfonamide 3 displayed the smallest energy gap (3.24 eV), indicating that it was the softest molecule with good polarizability and reactivity, whereas N-piperidinyl-o-sulfanilamide 5 presented the largest energy gap of 4.924 eV, thereby corroborating its high chemical hardness of 2.462 eV (cf. Table 5). The lowest LUMO energy was obtained from 3 (E LUMO = À3.175 eV), indicating that it is the best electron acceptor of the molecules analyzed, whereas 6 was the best electron donor in the series, with the highest HOMO energy (E HOMO ) of À6.142 eV ( Table 5). The observed large energy gap (4.924 eV) in 5 suggests that charge transfer could promote its bioactivity and ability to form biological interactions at the piperidinyl and phenyl moiety (Al-Wahaibi et al., 2019). Therefore, the predicted order of biological interactions are 5 > 6 > 4 > 2 > 1 > 3.
The ionization potential (I), electron affinity (A), chemical potential (), electronegativity (), global hardness (), global softness (S) and global electrophilicity (!) values were calculated using the HOMO and LUMO energy values and are collated in Table 5. The lowest I value of 6.142 eV originated from sulfanilamide 6, whereas sulfonamide 3 gave the largest A value of 3.175 eV. Amongst the compounds studied, 2 gave the highest value of 5.1795 eV. Interestingly, sulfanilamide 5 displayed the highest value of 2.462 eV and the lowest chemical softness (S) of 0.406 eV, thus alluding to its having the most reactive nature of all the molecules investi-  Table 6 Energies of o-nitrosulfonamides 1-3 and N-cycloamino-o-sulfanilamides 4-6 with the hCA II (PDB entry 4iwz) and hCA IX (5fl4) isoenzymes.  gated. The highest global electrophilicity of 29.597 eV was also recorded for sulfonamide 2, indicating that it is a strong electrophile. In general, the chemical reactivities of compounds 1-6 have been shown to vary with the groups attached to the compounds (Abbaz et al., 2018).
Docking calculations between 4iwz and A (reference drug) displayed a docking score of À2.252 kcal mol À1 , which is higher than for all synthesized compounds 1-6 (cf. Table 6).
To determine the mode of interaction of the synthesized compounds with human carbonic anhydrase IX inhibitor (hCA IX), the synthesized compounds were docked into the active site of 5fl4, and the results obtained were compared with the docked results of the reference drug B. We observed that the reference drug interacts with amino acid residues ASP 13 (1.59-2.73 Å ) and VAL 130 (2.53 Å ) via hydrogen bonding, and with HID 94 (5.49 Å ) via -cation interactions (cf. Table 7). Furthermore, B exhibited a docking score of À1.969 kcal mol À1 , a glide E-model energy of À41.029 kcal mol À1 and a ligand efficiency of À0.082 kcal mol À1 , and is surrounded by several amino acid residues. Some of the residues are TRP 9 , PRO 203 , THR 201 , HID 68 , LEU 199 , HID 94 , GLN 92 , VAL 171 and ZN 264 , with bad contacts or interactions observed on residue ASP 131 (Fig. 6). Benzenesulfonamide 2 presented the highest binding affinity, with a docking score of À1.977 kcal mol À1 , higher than the reference drug. All other synthesized compounds, except for N-cycloamino-o-nitrobenzenesulfonamide 1 (docking score = À0.807), displayed significantly good docking scores; however, they were lower than the reference drug (cf. Table 6). Compound 3 displayed hydrogen-bond interactions with amino acid residue GLN 71 , with a bond length of 2.26 Å , and a -cation interaction with amino acid residue HID 94 , with a bond length of 4.05 Å (Fig.  S15).
We observed that the docking scores of 2 with 4iwz and 5fl4 are close to those obtained for A with 4iwz and B with 5fl4. Docking scores of molecules with ring structures 1 and 3-6 (in the range > À1.67 kcal mol À1 ) also correlated with the electronegativity and electrophilicity values presented in Table 5. This is informed by the HOMO and LUMO properties (Kumar et al., 2018).

Conclusion
o-Nitrosulfonamides 1-3 and N-cycloamino-o-sulfanilamides 4-6 have been successfully synthesized, characterized and the intermolecular interactions analysed, as well as being tested in silico for carbonic anhydrase II (4iwz) and IX (5fl4) inhibitory activities. The results obtained from crystal packing and DFT analysis suggests that the molecules are held together by forces such as hydrogen bonding andinteractions. The results of the DFT study of compounds 1-6 were correlated with the molecular docking data and indicate that electronegativity and electrophilicity of the title compounds play an important role in their interaction with carbonic anhydrase II (4iwz) and IX (5fl4).
O-Nitrosulfonamide 2 displayed a good docking score against 4iwz (lower than the reference drug) and the best against 5fl4 (higher than the reference drug). These results provided a valuable synthesis approach and structural and docking information for compounds 1-6 that may be used for the development of potent antibacterial drugs.

1-[(2-Nitrophenyl)sulfonyl]pyrrolidine (ka097)
Crystal data Refinement. Carbon-bound H atoms were placed in calculated positions and were included in the refinement in the riding model approximation, with U iso (H) set to 1.2 U eq (C). The pyrrolidine group is disordered necessitating the use of the restraints SADI, EADP and SAME. Single-crystal X-ray diffraction (SC-XRD) data were collected at 200 or 296 K on a Bruker APEXII CCD diffractometer with graphite-monochromated Mo Kα radiation using the APEX2 data collection software and SAINT (Bruker 2012) for cell refinement and data reduction. The structures were solved by dual-space methods applying SHELXT2018 (Sheldrick, 2015a) and refined by least-squares procedures using SHELXL2018 (Sheldrick, 2015b). Data were corrected for absorption effects using the numerical method implemented in SADABS (Bruker 2012). All non-H atoms were refined anisotropically. The crystal structure diagrams were drawn with ORTEP-3 for Windows (Farrugia 2012).

1-[(2-Nitrophenyl)sulfonyl]-2,3-dihydro-1H-indole (ja250)
Crystal data An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. Carbon-bound H atoms were placed in calculated positions and were included in the refinement in the riding model approximation, with U iso (H) set to 1.2 U eq (C). Single-crystal X-ray diffraction (SC-XRD) data were collected at 200 or 296 K on a Bruker APEXII CCD diffractometer with graphite-monochromated Mo Kα radiation using the APEX2 data collection software and SAINT (Bruker 2012) for cell refinement and data reduction. The structures were solved by dual-space methods applying SHELXT2018 (Sheldrick, 2015a) and refined by least-squares procedures using SHELXL2018 (Sheldrick, 2015b). Data were corrected for absorption effects using the numerical method implemented in SADABS (Bruker 2012). All non-H atoms were refined anisotropically. The crystal structure diagrams were drawn with ORTEP-3 for Windows (Farrugia 2012).

Special details
Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. Carbon-bound H atoms were placed in calculated positions and were included in the refinement in the riding model approximation, with U iso (H) set to 1.2 U eq (C). The nitrogen-bound H atoms were located on a difference map and refined freely. Single-crystal X-ray diffraction (SC-XRD) data were collected at 200 or 296 K on a Bruker APEXII CCD diffractometer with graphite-monochromated Mo Kα radiation using the APEX2 data collection software and SAINT (Bruker 2012) for cell refinement and data reduction. The structures were solved by dual-space methods applying SHELXT2018 (Sheldrick, 2015a) and refined by least-squares procedures using SHELXL2018 (Sheldrick, 2015b). Data were corrected for absorption effects using the numerical method implemented in SADABS (Bruker 2012). All non-H atoms were refined anisotropically. The crystal structure diagrams were drawn with ORTEP-3 for Windows (Farrugia 2012).

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å 2 )
x y z U iso */U eq S1 0.66247 (2)  where P = (F o 2 + 2F c 2 )/3 (Δ/σ) max < 0.001 Δρ max = 0.27 e Å −3 Δρ min = −0.42 e Å −3 Special details Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. Carbon-bound H atoms were placed in calculated positions and were included in the refinement in the riding model approximation, with U iso (H) set to 1.2 U eq (C). The nitrogen-bound H atoms were located on a difference map and refined freely. Single-crystal X-ray diffraction (SC-XRD) data were collected at 200 or 296 K on a Bruker APEXII CCD diffractometer with graphite-monochromated Mo Kα radiation using the APEX2 data collection software and SAINT (Bruker 2012) for cell refinement and data reduction. The structures were solved by dual-space methods applying SHELXT2018 (Sheldrick, 2015a) and refined by least-squares procedures using SHELXL2018 (Sheldrick, 2015b). Data were corrected for absorption effects using the numerical method implemented in SADABS (Bruker 2012). All non-H atoms were refined anisotropically. The crystal structure diagrams were drawn with ORTEP-3 for Windows (Farrugia 2012).