Supramolecular Packing of a Series of N-Phenylamides and the Role of NH···O=C Interactions.

A series of seven N-phenylamides [R-C(O)NHPh, in which R: CH3, C(CH3)3, Ph, CF3, CCl3, CBr3, and H] were used as models in this study. Molecular packing and intermolecular interactions were evaluated by theoretical calculations, solution NMR, and quantum theory of atoms in molecules analyses. Crystallization mechanisms were proposed based on the energetic and topological parameters using the supramolecular cluster as demarcation. Concentration-dependent 1H NMR experiments corroborated the proposed interactions between molecules. For all compounds (except for R: H, which initially formed tetramers), layers (two-dimensional) or chains (one-dimensional) were formed in the first stage of the proposed crystallization mechanisms. The presence of strong intermolecular NH···O=C interactions promoted the first stages. The study in solution provided different values of association constant (K ass) governed by the hydrogen bond NH···O=C, showing that the stronger interactions are directly influenced by the substituent steric hindrance. A correlation between K ass(NH···O=C) from the solution and the NH···O=C interaction energy in the crystal showed a good trend.

Introduction

The design of new molecular solids with desired properties and functions propels the development of crystal engineering.[1] Several molecular models promote the development of network predictions and understanding of the role of the interactions in the stabilization of molecular solids. This knowledge is widely explored in the construction of molecular machines,[2] molecular catalysts,[3] self-association,[4] cocrystals,[5] solvates, and polymorphs.[6]

Hydrogen bonds and weak interactions, such as n class="Disease">van der Waals interactions, play an important role in the crystalline packing of organic compounds. Furthermore, theoretical studies have contributed to the characterization and description of these interactions. The QTAIM (quantum theory of atoms in molecules) analysis is a powerful tool used to analyze the strength of intra-/intermolecular interactions that stabilize crystal structures.[7,8] Similarly, the molecular reactivity can be obtained by measuring the molecular electrostatic potential (MEP) based on regions of positive and negative potentials.[9]

In addition to theoretical tools, experimental approaches using 1H n class="Chemical">NMR are used to explore interactions, including π···π or hydrogen bonds, reflecting the aggregation of the molecules in the solution. Titration experiments of host–guest[10] and self-assembly[11] reflect the strength of interactions in supramolecular chemistry. Then, the data are compared and fitted to binding models in order to obtain information such as the association constant Kass.[12]

Among the various classes of compounds, the study of amides is of great relevance. The n class="Chemical">hydrogen-bonding abilities of amides are important in biological systems, drugs, and commercial fields, in addition to being much appreciated in crystal engineering. The amide linkage provides structural rigidity and selectivity in crystals. In recent investigation, Chopra et al.[13] investigated the characteristics of intermolecular interactions in a series of N-aryl-2-naphthamides and evidenced the strong intermolecular NH···O=C interactions, CH···π, and π···π stacking which help to stabilize the molecules. This approach describes the crystalline structure and uses motifs to explain packing.

Some approaches of crystalline packing are based on pure geometrical considerations and overlook energy data.[14,15] Moreover, a little is known on crystallization mechanisms that involve the stages of molecular aggregation to form crystals. In this context, our research group has proposed crystallization mechanisms for organic compounds (e.g., isoxazoles and n class="Chemical">triazenes).[16,17] These proposals are based on energetic and topological contributions related to the interactions between the molecules in each stage of crystallization mechanisms. Moreover, the supramolecular cluster was first used as the demarcation of the smallest portion required for the full characterization of all intermolecular interactions in the crystal structure.[18]

In this context, this study aimed to systematically and quantitatively explore the packing of a series of N-phenylamides 1–7 (Figure ) containing substituents with different characteristics and volumes. In addition, crystallization mechanisms are proposed, addressing the main stages of the entire process while corroborated by the solution NMR. This study was carried out in solid and solution NMR, single-crystal X-ray diffraction, and theoretical [density functional theory (DFT) calculations, MEP, and QTAIM] data.

Figure 1

Series of N-phenylamides 1–7 used in this study.

Results and Discussion

Molecular Structure

N-phenylamides 1–7 used in this study have their molecular structures comprised a phenyl ring linked to the substituents by a n class="Chemical">NHC(O) amide unit (Figure ). All compounds were synthesized in our laboratory according to the methodology already described in the literature (see Experimental Section). A set of structures 1–3 and 5–7 were taken from the Cambridge Structural Database (CSD).[19] Crystals of compound 4 were obtained from crystallization by vapor diffusion using hexane and chloroform for a single-crystal X-ray analysis. The ORTEP of compound 4 is shown in Figure , and details of X-ray measurements and crystal data are given in Table S1 (Supporting Information).

Figure 2

ORTEP diagram for compound 4 (named 4A and 4B). Thermal ellipsoids are shown with 50% probability.

The two molecules of the asymmetric unit in Figure have different torsions of the amide group in relation to the aromatic ring. Molecule 4A has an angle between the planes of the phenyl ring and the n class="Chemical">amide unit of 32.2°, and molecule 4B has an angle of 36.3°. Compounds 1–3 and 5–6 presented one molecule in the asymmetric unit (Z′ = 1), and compounds 4 and 7 presented two molecules (Z′ = 2). Compound 7 has two conformational isomers: Z (7A) and E (7B), which were presented as a cocrystal by Levendis et al.[20]

Cable et al.[21] reported that the Z isomer adopts a planar structure and is 2.5 kcal mol–1 more stable than its E isomer. In the solution, 1H n class="Chemical">NMR experiments determined the existence of the mixture of species Z and E with almost the same abundance in a chloroform solution.[22] On the other hand, compound 1 was found almost entirely as the Z isomer, existing as the effect of steric interactions between the methyl and phenyl groups.[21]

Solid-state NMR [n class="Chemical">13C cross-polarization/magic-angle spinning (CP/MAS) NMR] spectroscopy experiments were performed in order to investigate the presence of conformers or polymorphs for the compounds. The comparison between solution and solid-state NMR spectra for compound 7 is shown in Figure a,b, respectively. Similar to the solution spectrum, duplicated signals can be observed, which indicates that the conformers Z and E can be present in the polycrystalline sample. Another possibility is the existence of more than one crystalline phase. Considering that compounds 1–6 did not present duplicated signals, it is plausible that there is only one crystalline phase. All other compounds have their spectra in the Supporting Information (Figures S22–S27). Except for compound 7, all other compounds used in this study showed only Z conformers in both solution and crystal phases.

Figure 3

13C NMR in CDCl3 at 298 K (a) and 13C CPMAS NMR (b) for compound 7.

The crystal structure of the compounds shows that the amide unit tends to increase its deviation in relation to the plane of the phenyl ring influenced by the increase of the substituent volume. The angles (θ1) obtained between the constructed planes of the n class="Chemical">amide unit (N–C=O) and phenyl ring are illustrated and described in Table . The isomers of compound 7 exhibited differences between them. The Z form (7A) had a deviation of about 11°, and the E form (7B) was practically planar. The remaining compounds presented deviations that ranged from 16° to 39°, regarding the increase of the substituent volume. The first value corresponded to the methyl substituent (1) and the last one to the tribromomethyl substituent.

Table 1

Angles Between the Planes of the Ring and the Amide Unit

compound	R	θ1 (deg)a	RMSb	υc
1	CH3	16.03		0.52
2	C(CH3)3	29.45	0.124	1.24
3	Ph	32.34	0.367	1.66
4A	CF3	32.20	0.373	0.91
4B		36.34	0.176
5	CCl3	34.88	0.181	1.38
6	CBr3	39.22	0.191	1.56
7A	H	10.86	0.219	0
7B		0.99	0.574

Angle between planes N–C=O/phenyl.

Obtained from the overlay of C–C(O)N–Cphenyl atoms of acetanilides 1–6. For compound 7, the C(O)N–Cphenyl atoms were used.

Charton’s steric parameters were taken from ref (23).

Obtained from the overlay of C–C(O)N–n class="Chemical">Cphenyl atoms of acetanilides 1–6. For compound 7, the C(O)N–Cphenyl atoms were used.

The overlay diagram for compounds 1–7 can be observed in Table , in which the phenyl ring is used as a rigid part to overlay all structures. According to the overlay results, the structures reveal considerable conformational flexibility between the phenyl ring and amide unit. Another data obtained from the overlay of the molecules are the root mean square (RMS) as a measure of the magnitude of the overlay. Compound 1 was used as reference while considering the methyl substituent as a pattern [because the n class="Chemical">hydrogen-substituted compound (7) has Z′ = 2]. The highest values refer to compounds 3 and 4A because of the twisting of the plane to the opposite direction of the others. The 0.574 value of compound 7B reflects the unique E conformation adopted.

The most widely used quantitative measurement of steric effects is the Es values, proposed by Taft[24] and subsequently modified by Charton (υ).[25] In this manner, a good tendency can be obtained by correlating the steric effect presented by Charton (υ) (Table ) with the angles (θ1) between the N–C=O/phenyl planes (θ1 = 15.362υ + 13.206; r = 0.82), shown in Figure . Therefore, the substituents may have direct influence in the torsion between the measured planes. Compound 7B was not considered in the correlation because it only involves molecules in the Z conformation.

Figure 4

Correlation between the dihedral angle of N–C=O/phenyl (θ1) and Charton’s steric parameter (υ) for compounds 1–7.

Additionally, MEP[26−28] maps were generated for the molecules investigated. The MEPs of 4, 5, and 6 highlighted the presence of the σ-hole in the direction of the C–X bond.[29] The values of Vmax and Vmin and other important regions of the molecules are presented in the Supporting Information (Figure S1 and Table S2).

Supramolecular Structure

Contact area and stabilization energy data provide important information about the crystalline lattice, which helps to better understand the molecular packing in the crystal (e.g., crystallization mechanism). The first studies in the area were proposed by Kitaigorodsky, who defined the existence of the first sphere of molecular coordination based on molecules that have at least one contact with any given molecule.[30] Subsequently, the Voronoi–Dirichlet polyhedron (VDP), which is based on Dirichlet’s terms, was introduced as an approach where two adjacent molecules sharing the same faces of a polyhedron had a contact area between them.[31]

However, demarcation is necessary to apply these parameters. Our research group has used supramolecular clusters to obtain all data about interactions between neighboring molecules (MN) around a reference molecule M1 (M1···Mn class="Chemical">N).[18] First, the supramolecular clusters for compounds 1–7 were determined. For compounds 4 and 7 with Z′ = 2, one cluster for each molecule was determined. The molecular coordination number (MCN) was obtained, which is the number of molecules that have at least one contact with any given molecule. When considering each molecule of 7 separately, two clusters with MCN = 15 are found. For compounds 1, 3, and 4 (4A and 4B), an MCN = 14 was found, whereas an MCN = 16 was found for compounds 2, 5, and 6 (the clusters are shown in Figures S2–S11, Supporting Information).

In addition to the contact area data obtained from the VDP analysis, the stabilization energies between molecules of each dimer were determined by quantum mechanics calculations at ωB97X-D/cc-pVDZ level theory. Energetic and topological data were normalized to compare different supramolecular clusters in order to compare the contribution of each dimer in the parameters presented [e.g., normalized stabilization energy (n class="Chemical">NG) and normalized contact area (NC) in each cluster].[17,18] Normalization means to reduce all raw data to the same metric (scale) using the MCN as the reference value. As previously presented by our research group, the dimers can be classified into four types.[18] Type I is represented by interactions with high energy (large NGM1···M values) in a small contact area (small NCM1···M values) usually characterized by strong hydrogen bonds. Type II corresponds to high interaction energy in a large contact surface (e.g., π···π interactions). Type III interaction has low NCM1···M and NGM1···M values with a maximum difference of 0.5 between the two parameters, which is the interaction type that most frequently appears in crystalline systems. Type IV interactions have low NGM1···M values and relatively high NCM1···M values. The symmetry codes, raw and normalized values of contact area, and stabilization energy and type of dimer of compound 1 are shown in Table . The data for the other compounds are presented in the Supporting Information (Tables S3–S12). The normalized data of dimers (NG and NC) for compound 1 are presented in Figure , for the other compounds (2–7), see the Supporting Information.

Table 2

Contact Area and Energetic Data of Each Dimer from the Supramolecular Cluster of Compound 1

dimer	symmetry code	CM1···MN (Å2)a	GM1···MNb (kcal mol–1)	NCM1···MNc	NGM1···MNd	typee
M1	x,y,z
M1···M2	1 – x,–y,–z	36.73	–9.97	2.42	2.70	II
M1···M3	1/2 + x,1/2 – y,–z	22.37	–9.34	1.48	2.53	I
M1···M4	–1/2 + x,1/2 – y,–z	22.37	–9.34	1.48	2.53	I
M1···M5	1/2 – x,–1/2 + y,z	22.90	–6.02	1.51	1.63	II
M1···M6	1/2 – x,1/2 + y,z	22.90	–6.02	1.51	1.63	II
M1···M7	1 – x,1 – y,–z	11.72	–2.00	0.77	0.54	III
M1···M8	–1/2 + x,y,–1/2 – z	10.97	–1.79	0.72	0.48	III
M1···M9	1/2 + x,y,–1/2 – z	10.97	–1.79	0.72	0.48	III
M1···M10	1 – x,–1/2 + y,–1/2 – z	11.25	–1.38	0.74	0.37	III
M1···M11	1 – x,1/2 + y,–1/2 – z	11.25	–1.38	0.74	0.37	III
M1···M12	x,1/2 – y,–1/2 + z	11.00	–0.68	0.73	0.18	IV
M1···M13	x,1/2 – y,1/2 + z	11.00	–0.68	0.73	0.18	IV
M1···M14	1/2 – x,–y,1/2 + z	3.38	–0.68	0.22	0.18	III
M1···M15	1/2 – x,–y,–1/2 + z	3.38	–0.68	0.22	0.18	III
total		212.19	–51.74	14.00	14.00

From ToposPro software.[32]

Obtained using the following equation: GM1···M = GM1+M – (GM1 + GM).

NCM1···M = (CM1···M/∑CM1···M) × MCN.

NGM1···M = (GM1···M/∑GM1···M) × MCN.

Classification according to Martins et al.[18]

Figure 5

Normalized contact area (NC) and stabilization energy (NG) of dimers from the supramolecular cluster of compound 1.

Normalized contact area (n class="Chemical">NC) and stabilization energy (NG) of dimers from the supramolecular cluster of compound 1.

Obtained using the following equation: GM1···M = n class="Chemical">GM1+M – (GM1 + GM).

NCM1···M = (CM1···M/∑CM1···M) × MCn class="Chemical">N.

Nn class="Chemical">GM1···M = (GM1···M/∑GM1···M) × MCN.

Through normalized data, it is possible to propose crystallization mechanisms based on dimer contribution. The crystallization process can be viewed as a stepwise process in which molecule association increases system complexity in the formation of the crystalline solid. Proposals for these mechanisms can be compared to retrosynthesis applied in organic chemistry,[33] in which the structure of a target molecule is decomposed into a sequence of progressively simpler structures along a path ultimately leading to simpler ones. From the three-dimensional (3D) structure, it is possible to propose the most probable stages that direct crystal growth based on energetic criteria while following the concept of retrocrystallization.[16,17] In a recent proposal, some parameters based on normalized data evidenced the predominance and contribution of energy and contact area parameters in each stage of the crystallization mechanism.[17] The NCG% shows the sum of normalized contact area and stabilization energy contribution of each stage. Moreover, the n class="Chemical">NG/NC stage parameter reveals the dominant factor in each stage (e.g., values >1.0 indicate dominance of the stabilization energy and values <1.0 show the contact area as the dominant parameter). In light of this, crystallization mechanisms were proposed for compounds 1–7 in order to evaluate the changes caused in packing by different substituents.

The proposed crystallization mechanism for compound 1 is represented by the formation of two main stages (Figure ). In stage I, a two-dimensional (2D) portion is formed, which involves 6 dimers from the cluster. The most energetic dimers in this portion are very similar. This portion is formed by interconnected chains with dimers M1···M3/M4, which are mainly directed by the NH···O=C interactions (type I interaction), with energy n class="Chemical">NG = 2.53. The interactions between the chains are mainly maintained by a strong dimer M1···M2 with a NG = 2.70. Additionally, other interactions involving the phenyl rings are present in dimers M1···MN, in which N = 5, 6, and 7.

Figure 6

Proposed crystallization mechanism of compound 1. The shaded area represents the portion in the previous stage. The arrows in each stage indicate the direction of growth. NCG% = 100 × (ΣNCstage + ΣNGstage)/(2 × MCN).

Proposed crystallization mechanism of compound 1. The shaded area represents the portion in the previous stage. The arrows in each stage indicate the direction of growth. NCG% = 100 × (Σn class="Chemical">NCstage + ΣNGstage)/(2 × MCN).

Stage I has NCG% = 74, i.e., 74% of all energetic and topologic contributions of the final crystal in this stage. This stage can be defined as a “point of no return” because this stage has enough stability to form the final 3D structure, rather than return to less complex nuclei. Stage II represents the approximation of the formed portions (2D) interacting to form the 3D structure. This stage presents dimers with n class="Chemical">NG below 0.48. By observing the NG/NC parameter of each stage, it is possible to note that the stabilization energy is dominant in the first stage (NG/NC = 1.26). The contact area, i.e., the complementarity between the surfaces of the 2D portions, is the governing parameter in the second stage (NG/NC = 0.51).

In order to obtain additional data to contribute with the proposed crystallization mechanisms, concentration-dependent 1H n class="Chemical">NMR experiments were carried out for all studied compounds. The experiments for compound 1 are observed in Figure . The prominent changes in chemical shifts for both NH···O=C interactions and those involving aromatic hydrogens derive from a concentration of 0.031 M. This may indicate the formation of the 2D block proposed in the crystallization mechanism from the retrocrystallization approach. The subtle change in chemical shifts under 0.031 M was not considered relevant enough to assess the possible formation of other stable portions during crystallization, such as one-dimensional (1D) chains.

Figure 7

Concentration-dependent 1H NMR spectra of compound 1 performed in CDCl3, at 298 K.

Concentration-dependent 1H n class="Chemical">NMR spectra of compound 1 performed in CDCl3, at 298 K.

Compound 3 and the trihalomethylated compounds 4–6 also presented two stages in their proposed crystallization mechanisms, which is also the case for compound 1. For these compounds, the first stage is governed by the formation of the 2D nucleus and stage II is the interaction between these stable layers through less energetic interactions. The difference in systems 5 and 6 will be in how the layers approach each other at the final moment of the crystallization process. This change in interaction in the final step and consequently in crystal packing is seen in the Supporting Information (Figures S14 and n class="Gene">S15).

The dominant parameters for compounds 3–6 are similar to that of compound 1, where the first stage is subtly governed by the energetic parameter with NG/n class="Chemical">NC = 1.12, 1.08, 1.14, and 1.12. The second stage has the contact area as the dominant parameter, with NG/NC = 0.62, 0.76, 0.74, and 0.75. The crystallization mechanisms were also corroborated with the chemical shifts observed by NMR in the solution (see the Supporting Information, Figures S19–S21).

On the other hand, the crystallization mechanism proposed for compound 2 is also divided into two distinct stages (Figure ) but with a slight difference, which is the formation of a 1D portion in the first stage. The interactions between dimers with NG = 3.61 take place in the first stage. This first stage presents n class="Chemical">NCG% = 36, indicating the point of no return of the crystal in forming these chains. Stage II is the interaction between the chains to form the 3D network. All remaining dimers (NG < 1.17) are involved in this stage. The dominance of the parameters follows the same tendency of the other compounds, where the first stage is governed by energy and subsequent increase of topology dominance in the second stage. However, in this compound, stronger dominance of the energetic parameter with a NG/NC = 1.65 can be noted in the first stage.

Figure 8

Proposed crystallization mechanism of compound 2. The shaded area represents the portion in the previous stage. The arrows in each stage indicate the direction of growth.

Compound 7 was first discussed as two clusters of molecules Z and E. The most notable contribution was observed from type I dimers involving NH···O=C interactions for both clusters. n class="Chemical">Nonetheless, the proposed crystallization mechanism for compound 7 has a notable difference when compared to compounds 1–6. Instead of forming the first 2D (or 1D) layers with both NH···O=C linking the chains and interactions involving the aromatic systems, a closed configuration is observed through the formation of a tetramer. This tetramer is composed of two monomers in the Z form and two in the E form that interacts through the NH···O=C interactions. This tetramer has already been reported as a N-phenylformamide cocrystal directed by hydrogen-bonding NH···O=C, although without a supramolecular study.[20] Therefore, it is remarkable that this same tetramer is still formed in solution,[22] reinforcing the idea that this can be treated as the minimum unit, i.e., M1 for this compound. The stabilization energy of −33.76 kcal mol–1 was obtained for the tetramer (Gtetramer) by theoretical calculations.

Because the tetramer is the most stable configuration adopted by molecules of compound 7, it must be considered M1, which is different from the other compounds in which there were no closed configurations. Therefore, a different supramolecular cluster was constructed, i.e., cluster of tetramers (Figure ). The contact area and stabilization energy data and their respective normalized values were obtained for the cluster and are shown in the Supporting Information (Table S5).

Figure 9

Tetramer formed by Z and E conformers of compound 7 considered M1 and their respective supramolecular cluster. The shaded area represents tetramer stacking.

The crystallization mechanism proposed for compound 7 has stage I characterized by tetramer formation (Figure ). The tetramer has a high stabilization energy of −33.76 kcal mol–1 because of the strong NH···O=C interactions.

Figure 10

Proposed crystallization mechanism of compound 7. The shaded area represents the portion in the previous stage. The arrows in each stage indicate the direction of growth.

Stage II occurs with parallel displaced stacking between tetramers, forming a 1D displaced column. This growth is governed by two stacking dimers, which are both with NG = 4.33. This stage presents n class="Chemical">NCG% = 40, revealing the high stability of this 1D portion. The next stage (III) presents interactions between these columns in perpendicular arrangement involving the aromatic rings (2D layer) of a set of dimers, with NG = 1.95, 0.61, and 0.60. In this third stage, the NCG% presented an additional value of 30% of the total energetic and topologic contribution. Stage IV represents the interaction between the layers through the hydrogens of the rings, forming the 3D crystal lattice. All dimers involved in the final stage present values of NG < 0.77. The NG/NC stage parameter showed the same behavior as compounds 1–6, which means there is a slight dominance of the stabilization energy in the first step and the subsequent rise of topological governance in the other two. This is in agreement with a study previously reported by our research group, where a series of triazene N-oxides presented the same tendency.[17] The concentration-dependent 1H NMR experiments for compound 7 are observed in Figure , which shows distinct behavior when compared with compounds 1–6.

Figure 11

Concentration-dependent 1H NMR spectra of compound 7 performed in CDCl3 at 298 K.

Concentration-dependent 1H n class="Chemical">NMR spectra of compound 7 performed in CDCl3 at 298 K.

Prominent changes in the chemical shifts from the lowest concentrations (0.008 M), regarding the NH···O=C interactions, are shown in Figure . This behavior corroborates with the initial formation of tetramers in solution, such as the initial nuclei in the crystallization. The chemical shifts induced by the aromatic systems start from concentrations of 0.062 M. This indicates further steps regarding the formation of the 1D displaced columns proposed in the crystallization mechanism (stage II) and other interactions between aromatic systems.

The changes observed in the packing of the compounds were also investigated, regarding the steric effect of the substituents. In this manner, we propose the use of values obtained by Charton’s method in the supramolecular scope because it was done at the molecular level. The steric effects presented by Charton (υ) (Table ) were correlated with the effects observed in the approximation of the molecules in the earlier stages of crystallization. To measure this effect, axes were constructed in two neighboring molecules (e.g., M1 and MN) linking n class="Chemical">carbonyl α-carbon and ipso-carbon of the phenyl ring. Next, the improper angle (θ2) between these two axes was measured. Then, we conventionalized that molecules that remained with the substituent on the same side on the chains have θ2 = 0° and opposite sides θ2 = 180°. The axes for compounds 1, 4, and 5 and the angles θ2 for all compounds are shown in Table .

Table 3

Cα–C-ipso (M1) and Cα′–C-ipso′ (MN) Axes for Compounds 1, 4, and 5 and the Improper Torsional Angles Obtained for Compounds 1–6

compound	R	θ2 (deg)a
1	CH3	180
2	C(CH3)3	109
3	Ph	0
4	CF3	0
5	CCl3	97
6	CBr3	102

Improper torsional angles [M1(Cα–Cipso)–MN(Cα′–Cipso′)]. Torsions were measured using the software Mercury.[34] Compound 7 was not considered while obtaining the axes due to tetramer being different from the other packages.

Improper torsional angles [M1(Cα–Cipso)–MN(Cα′–Cipso′)]. Torsions were measured using the software n class="Chemical">Mercury.[34] Compound 7 was not considered while obtaining the axes due to tetramer being different from the other packages.

This correlation exhibited a tendency between the evaluated data, which demonstrates that steric factors may influence the approximation of the molecules in the formation of the first stage of crystallization (see the Supporting Information, Figure S33). Compounds 3 and 4 presented distinct behavior, showing that steric effects are not dominant over these substituents (Ph and CF3) and compound 7 presented a different packing. Subsequently, the improper angle (θ2) was correlated with the volume of the substituents obtained by the Hirshfeld surface. A similar tendency was as observed and is shown in the Supporting Information (Figure S34).

Energy Contribution of Each Type of Interaction

Another important approach that provides a great deal of information about the nature of interactions is the analysis of QTAIM. The theory proposed by Bader[35] is based on the topological analysis of the distribution of electron charge density. Furthermore, correlations using interatomic distance and topological descriptors of the electron density at the bond critical point (BCP) can be used to access the energy properties of n class="Chemical">hydrogen bonds.[36] Our research group has used the fragmentation of GM1···M by the electron density (ρ) in the BCPs of the bond paths between the dimers in order to obtain atom···atom interaction energies.[16,17,37,38] Zou et al.[39] also showed that there is a strong correlation between these parameters for hydrogen and halogen bonds.

Then, using the fragmentation of the intermolecular interactions, it was possible to verify the strong influence of the NH···O=C bond in the stabilization of the tetramer (Figure ). The existence of conformers allowed packing with two main interactions of −7.08 kcal mol–1 and two of −6.55 kcal mol–1 that stabilize this structure. Additionally, five other interaction paths were observed with lower interaction energies, which added −6.51 kcal mol–1 to the tetramer stabilization.

Figure 12

Fragmentation of the stabilization energy (kcal mol–1) using the interaction pathways to the tetramer formed by compound 7. Cage critical points and ring critical points were omitted from the images for better clarity of the data.

This analysis can provide the contribution of each interaction type to assess the understanding, regarding the differences caused by the presence of different substituents. Figure shows the percentage of energy contribution obtained for each cluster separated by the types of interaction observed by QTAIM analysis. The interactions were manually classified based on the atoms and their respective pathways. The hydrogen bonds were divided into n class="Chemical">NH···O=C, CH···π, and CH···O/CH···N. Additionally, π···π, CH···HC, halogen bonds (X interactions), and “others” (N···N, N···π, O···π, and O···O) were considered.

Figure 13

Representation of energy contribution by type of interaction for compounds 1–7.

The NH···O=C interaction contributes with 28% of cluster stabilization for compound 7, with an energy of −7.08 kcal mol–1. Compounds 1, 2, and 4 also have great contribution of this interaction with values of −5.75 kcal mol–1 (22%), −4.07 kcal mol–1 (17%), and −4.53 kcal mol–1 (23%), respectively. The less pronounced contribution of n class="Chemical">NH···O=C interactions is for compounds 3, 5, and 6: −3.05 kcal mol–1 (10%), −2.63 kcal mol–1 (12%), and −3.33 kcal mol–1 (14%), respectively, which have direct relation to the size of the substituent.

The hydrogen bonds classified as CH···O/CH···n class="Chemical">N appear significantly with about 25% of the energy contribution to compounds 1, 2, and 7. In these same compounds, CH···π type interactions increase in relation to the increased availability of hydrogens to interact with π sites (i.e., H, CH3, and C(CH3)3).

Furthermore, structure 3 has about 60% of CH···π interactions, as expected by the presence of two aromatic rings in the structure. These dispersion forces play an important role in the packing of the compounds, as seen in the crystallization mechanism, with the formation of only two stages. The total stabilization energy of organic compounds is increased by the cooperative effect of multiple CH···π interactions.[40] In the structures with halogenated substituents, only compound 4 showed significant CH···π interactions, with the contribution of 23%. The CH···HC interactions contributed with 17 and 18% to stabilize the cluster of compounds 1 and 2, respectively. The methyl and tert-butyl substituents favor the occurrence of these weak interactions, although they have a significant total contribution to cluster stabilization.

The π···π interactions appear in compounds 3–5 and 7. In compound 7, the π···π interactions occur because of the parallelly displaced tetramers in the cluster, as seen in its mechanism in Figure . Compound 3, in addition to the CH···π interactions, two π···π interactions with around −1.30 kcal mol–1 also contribute to the stabilization of the cluster. For the halogenated compounds, only the fluorinated and chlorinated compounds have π···π interactions. The first one presents one interaction that represents 7% of the stabilization. The second has a parallel stacking dimer with two interactions of −2.17 kcal mol–1 (13%).

Compounds 4–6 had their contributions involving halogen interactions fragmented (Figure ), which is because they represent the biggest contributions to their respective clusters. For compound 4, which had Z′ = 2, only one molecule is present due to similarity between clusters. The percentage of contribution by X···X interactions increases on the order of F < Cl < n class="Chemical">Br, i.e., increasing the atomic radius increases the number of interactions between halogens. The same order was observed for X···O interactions. The CH···X interactions have equivalent contribution in 4 and 6 clusters (both 18%) and increase the contribution in cluster 5 (28%). Finally, the presence of interactions Xσ-hole···π type follows the σ-hole values presented in the MEP maps of Table S2 wherein compound 4 did not show such interactions, despite the small value of σ-hole. Halogen interactions play an important role in the total cluster stabilization of compounds 4−6. However, these interactions are not essential in the first stage of crystallization.

Figure 14

Representation of energy contribution by type of halogen interactions for compounds 4–6. X = F (4A), Cl (5), or Br (6).

Representation of energy contribution by type of halogen interactions for compounds 4–6. X = F (4A), Cl (5), or n class="Chemical">Br (6).

Exploring NH···O=C Interactions in Solution

NMR procedures are of general interest because of their importance in n class="Chemical">hydrogen bond studies, especially proteins. The correlation of hydrogen bond lengths and 1H NMR chemical shifts have proven that the crystal structure of proteins is preserved in solution.[41] In this manner, previous concentration-dependent 1H NMR experiments were used to obtain the association constant (Kass).[11,12] The Kass was obtained for the strong hydrogen bonds (N–H···O=C) related to the amide group and is listed in Table .

Table 4

Data of Kass Constants for Compounds 1–7

compounds	1	2	3	4	5	6	7
Kass(N–H···O=C) (mol L–1)	4.25	2.23	0.67	0.76	0.51	0.33	4.58

Analysis of the Kass values for all compounds suggests that as the size of substituent increases, the Kass value decreases. The steric hindrance previously correlated with Charton’s steric effects may hinder to approach the molecules (see the Supporting Information, Figure S35). The Kass results are in agreement with results reported by Hunter et al.[11,12] because the first aggregates formed in solution are similar to the arrangements found in the crystal.

The association constant (Kass(N–H···O=C)) was correlated with the n class="Chemical">hydrogen interaction energy (GNH···O=C), obtained by the fragmentation of the dimer energy, as previously demonstrated (Figure ). To the best of our knowledge, no correlation between the strength of the hydrogen interaction (from the crystal) with the Kass in solution has been reported in the literature. The good correlation found for compounds 1–7 (GNH···O=C = −0.7941Kass – 2.8364; r = 0.91) is shown in Figure . This indicates a direct relation strength of interactions in solution and the stabilization energy of the interactions in the crystal lattice. Therefore, with this data, it is possible to obtain, by either methods, information on the strength of the hydrogen bonds in these systems. Additionally, this approach can be used in the field of supramolecular chemistry to assist in the understanding of molecular self-assembly.

Figure 15

Correlation between Kass(NH···O=C) and GNH···O=C for compounds 1–7.

Correlation between Kass(NH···O=C) and Gn class="Chemical">NH···O=C for compounds 1–7.

Conclusions

The effect of the substituent on a series of N-phenylamides 1–7 leads to some changes in both the strength of the n class="Chemical">hydrogen bond energy and crystal packing. The proposed crystallization mechanisms for the compounds were divided into two types, which are two stages (1–6) and stacking of tetramers (7). Compound 2 presented the formation of a 1D portion (chains) in the first stage, where 1 and 3–6 showed the formation of 2D blocks. The different approaches of these forms lead to the formation of 3D structures. Compound 7 presented tetramers in both solution and solid state which were elucidated by the proposed crystallization mechanism. The concentration-dependent 1H NMR experiments corroborated the stages proposed in the crystallization mechanisms. The packing of the compounds was evaluated in relation to the steric effect of the substituents and correlated with the steric parameter of Charton, thus showing some tendency. The correlation between Kass and G for the NH···O=C interactions was proposed, and good correlation was observed (r = 0.91). This data suggest that the magnitude of this interaction can be elucidated through experimental and theoretical data in the supramolecular environment.

Experimental Section

Synthesis of N-Phenylamides (1–7)

The N-phenylamides (1–7) were synthesized to allow the study of the compounds in solution and obtain the crystal of compound 4. The synthesis was performed according to acylation reactions described in the literature.[42−44] In summary, 1 equiv of n class="Chemical">aniline was reacted with 1.2 equiv (2 equiv for compound 7) of the acylating agent, with a base in the presence of a solvent (when necessary). All compounds were purified using recrystallization or chromatographic column techniques. The yields of the isolated products ranged from 21 to 99%. For more details about synthesis, see the Supporting Information.

X-ray Structure Determination

The structures used in this study are available in the CSD through the following numbers: (1) 785065,[45] (2) 239624,[46] (3) 965773,[47] (5) 1314188,[48] (6) 747306,[49] (7) 682820.[20] Crystal structure parameters of compound 4 are listed in Table S1 in the Supporting Information. Experimental essays for crystallization of compound 4 failed with the aim to obtain crystals with good quality for X-ray data collection. Several crystallization tests were performed using pure polar solvents (methanol and n class="Chemical">chloroform) and combinations of these with an apolar solvent (n-hexane) in various proportions. In all cases, colorless crystals in the form of elongate needles were observed, so in ways that a fragment of a thicker needle was chosen for data collection. Starting from the best available sample, data were collected at low temperature (100 K) because the crystal presented low stability at room temperature. Preliminary attempts to index the symmetry of the crystal system reveal to be monoclinic as suggestion with the best reliability. Therefore, the strategy for data collection was based on this monoclinic system involving the half of the Ewald sphere. Attempts to perform the data collection based on an orthorhombic crystal system resulted fruitless. In this context of analysis—based on the original diffraction data—was maintained the monoclinic space group P21 for compound 4, without the possibility of symmetry increase to the orthorhombic system. In addition, on the basis of the reasonable quality of the diffraction data and on the observed uncertainty on the x Flack parameter, a final refinement using the Twin instruction was performed (SHELXL version 2016/6). The initially attributed BASF scaling factor converges to 0.04439. Diffraction data were collected on the Bruker D8 Venture Photon 100 diffractometer equipped with a graphite monochromator Mo Kα radiation (λ = 0.71073 Å) was used for collection, which was controlled at 100 K using an Oxford Cryosystems Cryostream 800 low temperature unit. The frames were integrated with Bruker SAINT software package.[50] Absorption effects were corrected using the multiscan method (SADABS).[51] The structure was solved by direct methods using the software package WinGX.[52]

Theoretical Calculations

All calculations in this study were performed using the Gaussian 09 software package.[53] DFT calculations were performed in a single point mode with the ωB97XD/cc-pVDZ theory level.[54] The counterpoise method of Boys and Bernardi was employed to minimize basis set superposition error.[55]G represents the electronic energy obtained from the calculations, in kcal mol–1. The stabilization energy for the dimers in the cluster was obtained, considering the dimers and monomers using the following equation: n class="Chemical">GM1···M = GM1+M – (GM1 + GM). The energy of tetramer was obtained using the following equation: Gtetramer = GM1+M2+M3+M4 – (GM1 + GM2 + GM3 + GM4). The fragmentation was obtained correlating Gtetramer and ρ (BCPs). MEP maps were built on the electron density 0.001 au isosurface with GaussView.[56] The Hirshfeld surface was obtained using the CrystalExplorer software.[57]

The wave functions used in the QTAIM analysis were generated at the ωB97XD/cc-pVDZ level of theory. The QTAIM analyses were performed with the aid of the AIMAll program package.[58] Further information regarding the QTAIM data is described in the Supporting Information.

Determination of the Association Constant (Kass)

Determination of the association constant was done by 1H NMR dilution experiments by preparing different dilutions of each compound at known concentrations (1, 0.5, 0.25, 0.125, 0.062, 0.031, 0.015, 0.008, and 0.004 mol L–1). All of the 1H NMR spectra were recorded for each concentration using a Bruker Avance III 600 MHz [1H at 600.130 MHz in 5 mm sample tubes at 298 K in CDCl3/tetramethylsilane (TMS). Nonlinear curve fitting software was used to analyze NMR signals.

1H and 13C NMR and Solid-State NMR (SSNMR)

1H and n class="Chemical">13C NMR spectra were recorded on a Bruker Avance III (1H at 600.13 MHz and 13C at 150.903 MHz) spectrometer in CDCl3/TMS solutions at 298 K. All spectra were acquired in a 5 mm tube at natural abundance. The SSNMR data were recorded in a Bruker Avance III (600 MHz) in the Larmor frequency for 13C NMR of 150.903 MHz. The 13C CPMAS spectra were acquired on a 4.0 mm probe at a spinning rate of 12 kHz. The 1H and 13C 90° pulses were set to 3.4 and 4.85 μs corresponding to radio frequency (RF) field strengths of 74 and 50 kHz, respectively. The CPMAS was acquired using a contact time of 3 ms with 1H and 13C RF of 72 (70–100% RAMP-CP shape) and 74 kHz, respectively. During the acquisition, a SPINAL-64 1H decoupling with a pulse length of 7.25 μs at a RF field strength of 74 kHz was applied.