Synthesis and special characterization through X-ray analysis of 1,8-dioxooctahydroxanthenes

2.1 Generalities

All reagents were purchased from commercial sources (Sigma Aldrich - St. Louis, MO, US and Vetec - Rio de Janeiro, Brazil) and were employed as received. Solvents were procured from Vetec (Rio de Janeiro, Brazil) and were used as received. The ¹H and ¹³C NMR spectra were recorded on a Varian Mercury 300 instrument (300 MHz and 75 MHz respectively), using deuterated chloroform as solvent. Hydrogen nuclear magnetic resonance (NMR) data are presented as follows: chemical shift (δ) in ppm, number of hydrogen atoms, multiplicity, J values in Hertz (Hz) and assignment. Multiplicities are shown as the following abbreviations: s (singlet), d (doublet), t (triplet), m (multiplet). Infrared spectra (IR) were obtained employing the Varian 660-IR equipment with accessory GladiATR. Mass spectra were recorded on a SHIMADZU GCMS-QP5050A instrument under electron impact (70 eV) conditions. The mass was acquired in positive mode scanning from 50 to 1000 Da. Melting points were determined using MQAPF-301 melting point apparatus (Microquimica, Rio de Janeiro, Brazil). Analytical thin layer chromatography analysis was carried out on TLC plates recovered with 60GF254 silica gel.

2.2 Synthesis

2.2.1

2.2.1 Synthesis of 9-(4-bromophenyl)-3,3,6,6-tetramethyl-3,4,5,6,7,9-hexahydro-1H-xanthene-1,8(2H)-dione (3)

In a typical procedure, a round-bottomed flask (25 mL) was charged with 5,5-dimethylcyclohexan-1,3-dione (294.9 mg, 2 mmol), 4-bromobenzaldehyde (186.9 mg, 2 mmol) and ZrOCl₂·8H₂O (12 mg, 2 mol%). The mixture was stirred at 85 °C and the progress of the reaction was monitored by TLC analysis. After completion of the reaction, the mixture was cooled down to room temperature. Thereafter, it was added to the flask 50 ml of dichloromethane and the mixture was kept under stirring for about 30 min. Then, the catalyst, which is insoluble in dichloromethane, was separated by filtration. After that, 50 mL of ethanol was added and the system was kept undisturbed for recrystallization process. This procedure afforded compound 3 as a pale yellow solid in 80% yield (344.0 mg, 0.80 mmol). Crystals suitable for X-ray diffraction studies were obtained from the above mentioned recrystallization method. The obtained crystals were thoroughly washed with cold ethanol and dried. Structure of 3 is supported by the following data. See Fig. 1 for numbering.

Close

Fig. 1 Synthesis of xanthenodiones 3–11. ⁱ Compound 12 was not synthetized in this study even though its geometry is key to understand the xanthenodione structures.

Export to PPT

White solid (344.3 mg, 0.80 mmol, 80% yield), Mp 239–240 °C. IR (ATR): 2952 (ν_as CH₃), 2876 (ν_s CH₃), 1659 (ν C⚌O), 1624 (ν C⚌C), 1360 (δ CH₃), 1197 (ν_as ⚌C—O—C), 1165 (skeletal vibration of CH₃), 1139 (skeletal vibration of CH₃) cm⁻¹. ¹H NMR (300 MHz, CDCl₃) δ: 7.32 (2H, d, J = 8.0 Hz, Ar—H), 7.16 (2H, d, J = 8.0 Hz, Ar—H), 4.69 (1H, s, 3-H), 2.46 (4H, s, 8-CH₂ and 11-CH₂), 2.23 (d, 2H, J = 16.5 Hz, 6-H_a and 13-H_a), 2.15 (d, 2H, J = 16.5 Hz, 6-H_b and 13-H_b), 1.09 (6H, s, 14-CH₃ and 17-CH₃), 0.98 (6H, s, 14-CH₃ and 16-CH₃). ¹³C NMR (75 MHz, CDCl₃) δ: 196.3 (C1 and C5), 162.5 (C9 and C10), 156.6 (C—Ar), 143.2 (C—Ar), 131.1 (C—Ar), 130.2 (C—Ar), 120.2 (C—Ar), 115.1 (C2 and C4), 50.7 (C6 and C13), 40.8 (C8 and C11), 32.2 (C3), 31.5 (C7 and C12), 29.3 (C15 and C16), 27.3 (C14 and C17). MS, m/z (%): 430 (M+2, 30), 428 (C₂₃H₂₅BrO₃, M⁺, 30), 349 (21), 273 (100), 217 (26), 161 (17), 55 (22), 41 (31).

2.3 X-ray diffraction analysis

Well-shaped single crystals of xanthenodione derivatives (Fig. 1) were chosen and mounted on a κ-goniostat and exposed to X-ray beam (Mo Kα, λ = 0.71073 Å) using a Bruker-AXS Kappa Duo diffractometer with an APEX II CCD detector. X-ray diffraction experiments were carried out at room temperature (see Table 1). Data collection strategy was calculated by setting φ scans and ω scans with κ offsets using the software APEX2 (Bruker, 2012). Other crystallographic softwares were used as follows: SAINT (indexing, integration and scaling of raw data) (Bruker, 2012), SHELXL-97 (structure solving) (Sheldrick, 2008), SHELXL-97 (structure refinement) (Sheldrick, 2008) and MERCURY (Macrae et al., 2008) (structure analysis and graphical representations). Direct methods of phase retrieval were used to solve the crystal structures. All non-hydrogen atoms of asymmetric unit were promptly located from the electronic density Fourier map. The early solved model was refined by full-matrix least squares method based on F². In the refinements, free anisotropic and fixed isotropic thermal displacement parameters were adopted for non-hydrogen and hydrogen atoms, respectively. The isotropic thermal displacement parameters of hydrogens were 20% greater than the equivalent isotropic parameter of the bonded carbon (except those hydrogens in methyl groups which were 50%). Concerning the position of all hydrogens, bond distances and angles were stereochemically constrained according to riding model.

2.4 Computational details

The geometries of synthesized compounds (Fig. 1) were optimized at ωb97xD/6-311++∗∗ level of theory by Gaussian 09-D01 software (Frisch et al., 2009). In order to obtain the temperature effect on the electronic energy, the calculations were carried out in gas phase at 25 °C and in water at 85 °C using the PCM solvation model. The conformational preference of cyclohex-2-enone moieties in the synthesized compounds was studied enforcing their respective opposite syn or anti conformer and calculating their local energy minima and thermodynamics. The conformational preference of aromatic groups was evaluated through potential energy curves scanning the O-C-C-X dihedral angle, where X = O or S, with torsions of 45°.

3.1 Preparation of xanthenediones 3–11

The compounds 3–11 were synthesized via Knoevenagel condensation between different aromatic aldehydes (1) and β-diketones (2) (Mosaddegh et al., 2012). The reactions were performed free of solvents and catalyzed by ZrOCl₂·8H₂O, a versatile catalyst that has been involved in several synthetically useful transformations (Baltork et al., 2007; Sangshetti et al., 2008; Bardajee, 2013). From Fig. 1, the compounds were obtained in good yields within 20 min to 2 h.

The reactions were monitored by TLC analysis. The completion of the reactions was verified by the total consumption of the aldehydes. All the 1,8-dioxooctahydroxanthenes were purified via recrystallization using dichlomethane and ethanol. The structures of the compounds were confirmed upon NMR (¹H and ¹³C) and IR spectroscopy, as well as mass spectrometry. In the ¹H NMR spectra, the hydrogen atoms of the pyran ring was observed within the 4.70–5.27 ppm range. In the ¹³C NMR spectra, the signals for the carbonyl groups were observed near 196 ppm. In the IR spectra, carbonyl stretching frequencies were noticed in the range 1653–1670 cm⁻¹ which is compatible with a conjugated carbonyl group. The molecular formulas of the xanthenodiones were confirmed by mass spectrometry analysis. The structures of compounds 3–11 were also investigated by single-crystal X-ray diffraction.

3.2 Crystal structures

Crystal structures of compounds 3 and 4 have been previously elucidated (Bigdeli et al., 2007; Tu et al., 2002). Their determined unit cell metrics matched those reported in the literature confirming, therefore, all conformational and intermolecular features already described for these xanthenodiones. Crystal structures of all other compounds have not been described and thus are herein reported for the first time. Crystallographic data of these compounds are presented in Table 1.

In addition to C1, compound 5 has two stereocenters and it crystallizes in the centrosymmetric space group (C2/c), as expected from its achiral synthesis. In the chosen asymmetric unit, chiral carbons labeled as C8 and C12 have S and R configurations; however, the enantiomeric counterpart is also present in the unit cell. Compounds 6–11, each one presenting one stereocenter, have been crystallized in the centrosymmetric space groups. The crystallographic asymmetric unit of all xanthenodiones is composed by just one molecule, as viewed in Fig. 2. From Fig. 2, non-hydrogen atoms are shown with their 50% probability ellipsoids and arbitrary labeling scheme. Labeling of the rings is also depicted in Fig. 2.

Close

Fig. 2 50% ellipsoid plot of the non-hydrogen atoms in the asymmetric units of compounds 5–11 elucidated by the single-crystal X-ray diffraction technique. The labeling scheme of the rings as well as non-hydrogen atom numbering follow those shown for compound 5.

Export to PPT

For all compounds reported herein, the three fused rings formed an almost completely planar xanthenodione core. However, the side rings A and C adopt a half-chair conformation with C8 and C12 carbons away from the least-squares (l.s.) planes calculated through the other five coplanar atoms. The distance of these carbons from the l.s. planes and the root-mean square deviation of the l.s. fitted atoms are shown in Table 2.

Taking a xanthenodione mean plane as reference, it is interesting to observe that the flaps C8 and C12 of both half-chairs are oriented toward the aromatic substituent in compounds 5– 8, and 11, which corresponds to a syn conformation of cyclohex-2-enone moieties. On the contrary, in xanthenodiones 3, 4, and 10, C8 points toward the aromatic substituent while C12 points to the opposite direction, which relates to an anti- conformation of the half chairs. This trend observed for 3, 4, and 10 was also noticed for xanthenodione having a para-carboxymethoxy phenyl group (CSD ref. code GIWZAP, reference (Kumar et al., 2014)) as well as in that bearing only one phenyl moiety (herein numbered as compound 12; CSD ref. code VOQSEA, reference (Reddy et al., 2009)). Compound 9 also presents alternated flaps; however, C8 points opposite to the para-substituted phenyl ring while C12 points toward it.

Another difference among compounds 5–11, which is also responsible for the conformational distinction in terms of rings A and C, corresponds to the rotation around the axis of the bond connecting the xanthenodione core to its substituent at C1. Such an intramolecular feature can be viewed in Fig. 3, a molecular overlay of all asymmetric units (only non-hydrogen atoms) through the xanthenodione core.

Close

Fig. 3 Molecular overlay of all asymmetric units (only non-hydrogen atoms) through the xanthenodione core and labeling scheme of its rings.

Export to PPT

Even so, the plane of the substituent at C1 is pointed toward the ring A in all compounds, but with different bent levels. Such conformational predilection is more prominent in compounds with either furan (6 and 7) or thiophene (only 8) rings attached at C1. The dihedral angle among the atoms O1, C1, CX (X = 14 in 6–9; X = 16 in 5; X = 18 in 10 and 11) and Y (Y = S1 in 8 and 11; Y = O4 in 6 and 7; Y = C15, C17, or C19 in 9, 5, and 10) describes this conformational feature (Table 2).

The deviation from 0° indicates that the mean plane of the substituent at C1 coinciding with the transverse plane of the xanthenodione core is more pronounced in compounds 6–8 than in the others. In the case of the compounds possessing furan and tiophene rings, the heteroatom of the substituent at C1 is positioned on the same side of O1, in a so-called syn conformation, relative to the xanthenodione core. Also in these compounds, the para-halogenated phenyl ring is coplanar to furan in 6 and 7 and their l.s. mean planes form angles of 2.53(11)° and 4.11(7)°, respectively. On the other hand, the non-substituted phenyl ring and thiophene are twisted in 8 and 11, with angles of 33.08(8)° and 16.67(8)°, respectively, between their l.s. mean planes. In compound 9, there is an electronic conjugation between the 4-N,N-dimethylamine group and the phenyl ring bonded to C1, which can be observed by looking at the coplanarity between these moieties. There is an angle of 9.3(2)° between the planes calculated through the atoms C20-N1-C21 and C14-C15-C15-C17-C18-C19 in this compound. Furthermore, ring B is slightly distorted in all structures, with C1 carbon atom as the most deviated from the l.s. mean plane through the six ring atoms (Table 2).

In all structures, only the non-classical hydrogen bonds (CH…halogen and CH…O) and other weak contacts involving π-systems (π…π, CH…π) are responsible for keeping the molecules together in the crystals. The isostructural compounds 6 and 7 are featured by alternation of the centrosymmetrically-related molecules into one-dimensional (1D) chains running parallel to the [100] direction through C7-H7A (or B in 7)…O4, π_{(ring D)}…π_{(ring D)}, C12-H12A…Br (or Cl in 7), and C8-H8B…O2 (Fig. 4).

Close

Fig. 4 Isostructural supramolecular chains in (a) 6 and (b) 7 growing parallel to the [100] direction. Cg(E) denotes the centroid calculated through ring E atoms. In this picture and hereinafter, only hydrogen atoms involved in the displayed contacts were not omitted. The depicted measurements refer to the hydrogen…acceptor (i. e., oxygen, halogen or Cg) distance and the weak hydrogen bonding angle. When there, Cg…Cg distance is also exhibited.

Export to PPT

Both carbons at the half-chair flaps are involved in the last two contacts, which can indicate a conformational adaptability from these moieties to favor geometrically the intermolecular interactions around the crystal environment. On contrary, these methylene groups do not participate significantly in the crystal packing of compound 8. This structure is marked by formation of a centrosymmetric dimer where the CH moiety at the para-position of the phenyl ring is a non-classical hydrogen bonding donor to O2 (Fig. 5). This contact explains the phenyl ring twist observed in this structure. Such a phenomenon, also found in compound 11, is a consequence of a geometrical balance to form two CH…π contacts involving the π-system of the phenyl ring E (Fig. 5).

Close

Fig. 5 The pattern of weak intermolecular contacts responsible for phenyl ring E twist in (a) 8 and (b) 11.

Export to PPT

Compounds 3, 4, and 10 are isostructurals as expected from their molecular similarities (differing only for the halogen at the para-position of the phenyl ring bonded to C1) (Fig. 6). Even though there are changes in the non-classical hydrogen bonding geometry, they keep the same main supramolecular motifs as, for instance, an 1D chain assembled through the CH…halogen contact having the methyl moiety at C12 as a donor. This chain grows along the [001] direction and is made up of c-glide related molecules (Fig. 6).

Close

Fig. 6 (a) The supramolecular chain of 10 growing along the [001] direction. (b) Isostructural crystal packing of 10 (left) and of its bromine analog (right).

Export to PPT

Crystal packing of 5 is stabilized by means of CH…O and CH…π interactions, which gives rise to chains running parallel to the [101] direction (Fig. 7).

Close

Fig. 7 Supramolecular chain running parallel to the [101] direction in compound 5. Cg(D) denotes the centroid calculated through ring D atoms.

Export to PPT

In the crystal structure of 9, there is formation of a 1D chain along the [010] direction where both carbonyl oxygens are non-classical hydrogen bonding acceptors from the methylene groups close to O1 (Fig. 8a). In this structure, there is also the face-to-face stacking of two centrosymmetry-related molecules through π…π interactions. This last interaction occurs between the π-system of phenyl moiety and the π-cloud around the nitrogen in the 4-N,N-dimethylamine revealing, therefore, a dual intramolecular and intermolecular conjugation phenomenon in this compound (Fig. 8b).

Close

Fig. 8 (a) Supramolecular chain running parallel to the [010] direction in compound 5 and (b) the π_(N1)…π_{(ring D)} interactions stabilizing the centrosymmetric dimer in this structure. Cg(D) denotes the centroid calculated through ring D atoms.

Export to PPT

3.3 Theoretical aspects

The theoretical geometries of compounds 3–11 in the gas phase and in water are depicted in Fig. S1 (see supplementary material). In the crystal phase, molecular backbones have shown a preferential syn conformation for compounds 6, 7, 8, and 11 regarding the O1-C1-CX-Y torsion (Y = O4 in 6 and 7, or S1 in 8 and 11; Fig. 9).

Close

Fig. 9 O1-C1-CX-Y dihedral angle in compounds 6, 7, 8, and 11 (Y = O4 in 6 and 7, or S1 in 8 and 11).

Export to PPT

It is important to mention that anti/syn conformations, regarding the aforementioned dihedral, occur when the aromatic substituent is a furan or thiophene derivative (compounds 6, 7, 8, and 11). When the aromatic portion corresponds to a phenyl group, there is no differentiation considering anti/syn conformations. With no significant geometry difference between the gas phase and water, our calculations converged to similar conformations noticed in Section 3.2. Applying a conformational screening around this dihedral angle, it was possible to investigate why there is predominance for the syn conformations. Thus, from the optimized structure obtained by theoretical calculations, the variation of energy due to torsion of the O1-C1-CX-Y dihedral can be seen in Fig. 10. In this case, the D point is the energy minima and is equivalent to the optimized structure. In addition, compounds 6 and 7 have identical R’ and R’’ groups (Fig. 1), which result in very similar energy profiles. Because of their similarity, it is possible to discuss them together. Figs. S2 and S3 (see supplementary material) illustrate the conformers corresponding to points A, B, C, and D for compounds 6 and 7 along with some information about important dihedrals and atomic distances. The maximum energy observed in the diagrams of compounds 6 and 7 (Fig. 10) is point A. Comparing the structures of the conformers associated with points A and B, it is possible to recognize the fact that red eclipsed oxygens are not the main feature responsible for the higher energy of the point A conformer. Considering compound 6, going from A to B, the dihedral angle decreases from 33.41° to 11.59°. This fact results in increased repulsion between these atoms. However, from Fig. 10, the energy decreases from points A to B. Moreover, the yellow-highlighted dihedrals also decrease from 93.34° to 48.34° for compound 6 which results an eclipsing of the oxygen and hydrogen atoms. Furthermore, the decreasing of the aforementioned dihedrals implies in approximation of oxygen and carbon atoms in the conformer related to point A. In this conformer, the distance between them is 2.70 Å, whereas in the point B conformer, it grows to 3.05 Å. Points C and D are those where two oxygens are in anti and syn conformations, respectively. The energy difference between these two points is more than 2.5 kcal·mol⁻¹. The key to understanding the higher stability of the point D conformer is to look at the oxygen atom of the furan ring. In the point C conformer, this oxygen is eclipsed by hydrogen at a distance of 2.42 Å, whereas in point D the closest atom to the same oxygen is a carbon at a distance of 2.92 Å. By the same reason, it is also relevant to realize that the point C conformer is higher in energy than that for point B, although the distance between the two oxygens is higher in the conformer related to the point C. Therefore, it is expected that compounds 6 and 7 adopt the syn conformation.

Close

Fig. 10 Potential energy curves from the conformational analyses of compounds 6, 7, 8 and 11.

Export to PPT

Figs. S4 and S5 (see supplementary material) illustrate the important aspects for the compounds 8 and 11 (Y = S1) that have differences in terms of R’ and R’’ groups (Fig. 1). From Fig. 10, the energy maximum for compound 11 (point A) is greater than 6 kcal·mol⁻¹, while for compound 8, it is below 4 kcal·mol⁻¹. The superior energy of the point A conformer of compound 11 compared to that of compound 8 was ascribed to the repulsion among the sulfur atom and methyl groups. In the point A conformer of compound 8, there are two carbons which are eclipsed. However, in compound 11, the sulfur atom is eclipsed by the carbon. Therefore, higher repulsion is expected for compound 11 regarding its point A conformer. However, the point C conformer of compound 8 has also sulfur and carbon atoms eclipsed, albeit with almost half of the point A destabilization energy compared to the same point of compound 11. The reason is the existence of methyl groups in compound 11 which increase the steric repulsion, while for compound 8, R’ and R” are hydrogens (Fig. 1). To agree with this argument, an interesting effect happens considering the 11-syn to 11-anti interconversion (Fig. S1). As the O1-C1-CX-S1 dihedrals changes, one methyl group goes down through the plane to relieve the repulsion between the other CH₃ and sulfur. This same effect does not occur in the 8-syn to 8-anti interconversion since R’ and R’’ are hydrogens (Fig. 1). In the anti and syn conformations (B and D points, respectively), it is possible to apply the same arguments used for compounds 6 and 7 to justify the preference for the syn conformation. In the case of compounds 8 and 11, the reason involves the distance of sulfur to other atoms. Compounds 8 and 11 are destabilized by more than 2.0 kcal·mol⁻¹considering point B. For compound 8, the sulfur atom is close to oxygen and a hydrogen by, respectively, 3.28 Å and 2.60 Å in the point B conformer. In the point D conformer, the closest sulfur distance to oxygen and carbon atoms are 3.82 Å and 3.53 Å, respectively Considering compound 11, the respective distances are 3.31 Å and 2.79 Å for the point B conformer and 3.77 Å and 3.55 Å for the point D conformer. In both cases, in the anti-conformer, the repulsions are more severe than in the syn conformation. Thus, the syn conformer predominates for both compounds.

Crystal structure analyses have also revealed two distinct conformational preferences for the cyclohex-2-enone moieties of compounds 3–11. Compounds 3, 4, 9, 10, and 12 adopt the anti conformation, while compounds 5, 6, 7, 8, and 11 assume the syn conformation. The geometries of all conformations can be seen in Fig. S1 (see supplementary material). Table 3 shows the energy barrier of a possible interconversion between the syn and anti conformations following the equation:

Applying Eq. (1) for the compounds 3–11, energies were calculated from frozen crystal coordinates, enforcing only their respective opposite cyclohex-2-enone syn or anti conformer. The obtained ΔE values for 3, 9, 10, and 12 were positive due to the higher stability (and therefore lower energy values) of anti conformers. However, our calculations in the gas phase and in water predicts that syn conformers are electronically more stable for all compounds (ΔE < 0). In fact, for an isolated or solvated molecule, these compounds might exist in a syn conformer, which disagrees with the experimental data. This difference might be due to the supramolecular contributions (see Fig. 11).

Close

Fig. 11 Calculated gas phase geometries of the dimmers with either syn or anti cyclohex-2-enone rings conformation.

Export to PPT

Fig. 11 shows the built dimmers of compounds 3, 7, 11, and 12 that were taken as models to understand the solid-state data through the supramolecular model. These dimmers were built following the crystal coordinates. To obtain the hypothetic syn/anti conformers, the dimmers were optimized using molecular mechanics (UFF) with the aromatic groups frozen. The energy was calculated with a single-point at the ωb97xD/6-311++∗∗ level. All aromatic groups have aromaticity and conjugated bonds, and molecules with syn-orientation can potentially perform stacked interactions, such as π-π∗-stacking and T-shaped, which have strong dispersive and charge transfer effects that can only be described by suitable quantum mechanics methods (Josa et al., 2014a; Ramalho et al., 2013; Quin and Tyrell, 2010). The interaction energy of each dimmer was obtained by using (2), where A and B are the monomers. Table 3 also shows the interconversion energy barrier between the syn/anti dimmers in the gas phase (3).

Even though it is not identical to the solid state, the supramolecular approach with two monomers have already demonstrated to agree with the experimental data. The ΔE for compounds 3 and 12 present a positive signal indicating that anti conformation is indeed preferred. The calculation results for compounds 7 and 11 have kept the preference for the syn conformations as the crystal data have shown, which is the most natural occurrence since the monomers adopt this same conformation. One interesting question is why did the conformational change happen in compounds 3, 4, 9, 10, and 12?

Looking at the geometries of the compound 7 dimmers (Fig. 11), it is possible to see that parallel displaced π-π^∗-stacking rules the interaction between the monomers. Even just changing the hydrogens position between its syn and anti conformations, the energy barrier is −1.98 kcal·mol⁻¹, while the ΔE between the monomers was just −0.09 kcal·mol⁻¹ in the gas phase. In this case, the dispersive interactions between Cl and H could be decisive to stabilize the supramolecule along with the stacking interactions.

With a short-conjugated chain and a large substituent as bromine, compound 3 cannot perform π-π^∗-stacking but angular T-shaped σ_C-H → π^∗_Ar and dispersion interactions (Josa et al., 2014b; Santos et al. 2014). In its syn conformation, the methyl group can make these interactions difficult, increasing the system steric repulsion. In addition, the methyl group, oriented on the same side of the aromatic group in 3-anti, contributes to increasing the dispersion interactions with bromine. This fact can be supported by comparing the dimmers of compounds 3 and 12, whose ΔE values are 1.66 kcal·mol⁻¹ and 0.63 kcal·mol⁻¹, respectively. The ΔE in the gas phase between their monomers were almost the same, −0.68 kcal·mol⁻¹ for compound 3 and −0.61 kcal·mol⁻¹ for compound 12, which means that somehow the 3-anti dimmer is more stabilized than the 12-anti dimmer.

The crystal data and the theoretical calculations for compound 11 have also revealed possible T-shaped σ_C-H → π^∗_Ar interactions in the dimmers of compound 11. However, the dispersion can also be important. The ΔE for the monomers is −0.48 kcal·mol⁻¹ and −0.82 kcal·mol⁻¹ for the dimmer. This gain in the stabilization energy can be correlated to the C—H and methyl group dispersive interactions, which are completely erased in the 11-anti dimmer. In this prospect, our findings highlight the importance of the intermolecular interactions in this conformational preference. Indeed, intermolecular interactions are crucial for the solid state matter (Pereira et al., 2011) and it can show more distinct behavior than the solvated molecule if non-classical effects are considered, which can be easily understood with the supramolecular approach. It is worth emphasizing that all these interactions in the dimmer state can be intensified in the solid state by increasing the number of monomers present in the supramolecule.