Probing Attractive OH−π Interactions and Repulsive n−π Interactions in a Phenol Molecular Balance

TL;DR

A new molecular balance system helps study how phenol interacts with aromatic rings, revealing insights useful for drug and catalyst design.

Contribution

A 2,6-diarylphenol molecular balance is introduced to compare OH−π and repulsive O−π interactions experimentally and computationally.

Findings

Phenolic OH groups prefer electron-rich aromatic rings via OH−π interactions.

Avoiding repulsive O−π interactions significantly influences noncovalent interactions.

The system demonstrates molecular recognition based on aromatic ring electron density.

Abstract

Molecular balances have emerged as invaluable chemical systems for examinations of noncovalent interactions. Here, we report physical‐organic chemistry studies on a 2,6‐diarylphenol molecular balance that enables the examination of competing attractive OH−π interactions and repulsive O−π interactions between two different flanking aromatic rings. Integrated structural, NMR spectroscopic, and quantum chemical analyses revealed that the phenolic hydroxyl group preferentially interacts with the electron‐rich aromatic ring over the electron‐poorer counterpart via through‐space OH−π interactions. Quantum chemical analyses based on canonical energy decomposition analysis furthermore showed that the avoidance of the repulsive O−π interactions provides an important contribution to the noncovalent interactions between the phenol and aromatic rings that constitute the molecular balances. The work is important because it demonstrates that subtle differences in the electron density of aromatic rings can be recognized by the phenolic OH group, a useful piece of knowledge for the design of drugs and catalysts.

This work introduces a 2,6‐diarylphenol molecular balance that enables direct comparison of competing intramolecular attractive OH–π interactions and repulsive O–π interactions. Through combined experimental and computational studies, we show that the phenolic OH group selectively interacts with electron‐rich aromatic rings by avoiding repulsive O–π interactions, overall highlighting a sensitive mode of molecular recognition relevant to catalyst and drug design.

Figures

The 2,6‐diarylphenol scaffold as a molecular balance for examinations of the conformational preference of attractive OH−π interactions and repulsive n−π interactions.

Synthesis of the 2,6‐diarylphenols.

Crystal structures of phenols 1 (top) and 2 (bottom).

VT_NMR analyses of phenol 1. (a) 1H NMR spectrum of compound 1 in Et2O‐d 10 from −75°C to −100°C. (b) 2D COSY spectrum at −100°C.

VT_NMR analyses of phenol 2. (a) 1H NMR spectrum of compound 2 in Et2O‐d 10 at −85°C and −100°C. (b) 2D COSY spectrum at −85°C.

Relative energy ∆E (in kJ mol−1) of the substituted 2,6‐diarylphenols 1 and 2 as a function of the rotation of the dihedral angle φOH (in degrees) in diethylether (blue) and in vacuo (red).

EDA fragments used for the analysis of the OH‐π interaction in substituted 2,6‐diarylphenols, based on the isomer in which the OH points towards the more electron‐rich ring (OMe, top) or the more electron‐poor ring (CF3/H, bottom). EDA data is enclosed in Table 2.

VDD charges (top, in milli‐electrons) of the aryl‐ring carbon atoms considered in the EDA analysis for the isomers of compounds 1 and 2, and molecular electrostatic potential isosurfaces of the substituted aryl rings (bottom, isovalue = 0.03). Computed at ZORA‐BLYP‐D3(BJ)/TZ2P in vacuo.

VDD charges (top, in milli‐electrons) of the aryl‐ring carbon atoms of isomers 1, OMe, and 2, OMe, and their molecular electrostatic potential isosurfaces (bottom, isovalue = 0.03). Computed at ZORA‐BLYP‐D3(BJ)/TZ2P in diethylether.

EDA fragments used for the analysis of the n‐π interaction in substituted 2,6‐diarylphenols, based on the isomer in which the OH points towards the more electron‐rich ring (OMe, top) or the more electron‐poor ring (CF3/H, bottom). EDA data is enclosed in Table 3.

Tables

TABLE 1. Relative energy (in kJ mol−1) and structural data (Å, deg.) of the isomers in which the hydroxyl O–H points to the OMe‐substituted aryl ring of the substituted 2,6‐diarylphenols 1 and 2 relative to the isomer in which it points to the other aryl ring. Geometries of the isomers are depicted on top. a

TABLE 2. Energy decomposition analysis (in kJ mol−1) of the OH‐π interaction in substituted 2,6‐diarylphenols. Fragments used in the EDA are depicted in Figure 6. a

System	Isomer	interaction	∆E int	∆E Pauli	∆V elstat	∆E oi	∆E disp
1	OMe	OH···aryl‐OMe	−5.3	39.5	−22.0	−13.3	−9.5
	CF3	OH···aryl‐CF3	−4.6	36.6	−19.6	−12.2	−9.4
2	OMe	OH···aryl‐OMe	−5.2	39.6	−22.0	−13.3	−9.5
	H	OH···aryl‐H	−5.1	38.9	−21.5	−13.2	−9.4

TABLE 3. Energy decomposition analysis (in kJ mol−1) of the n‐π interaction in substituted 2,6‐diarylphenols. Fragments used in the EDA are depicted in Figure 9. a

System	Isomer	Interaction	∆E int	∆E Pauli	∆V elstat	∆E oi	∆E disp
1	OMe	O···aryl‐CF3	0.8	34.5	−18.9	−6.6	−8.1
	CF3	O···aryl‐OMe	5.0	35.5	−16.0	−6.5	−8.1
2	OMe	O···aryl‐H	3.6	34.4	−16.5	−6.3	−8.0
	H	O···aryl‐OMe	4.8	35.1	−15.8	−6.4	−8.1

Full text

Introduction

1

Noncovalent interactions involving aromatic rings play key roles in chemical and biological systems [1, 2, 3, 4, 5, 6, 7]. Polar−π interactions, a common type of weak interaction in which a polar functionality favorably interacts with the π system of aromatic rings, have been shown to serve as dominant driving forces in numerous supramolecular structures, protein‐ligand associations, and substrate‐catalyst complexations [1, 2, 3, 4, 5, 6, 7]. Diverse types of polar−π interactions have been established, including π−π interactions, cation−π interactions, anion−𝜋 interactions, SH−π interactions, OH−π interactions, halogen−𝜋 interactions, and CH−π interactions [4, 6]. Among these, OH−π interactions represent a special class of unconventional hydrogen bonds in which the hydroxyl group acts as an H‐bond donor and the aromatic ring as an H‐bond acceptor [8, 9, 10, 11]. In addition, the hydroxyl group can also engage with the aromatic ring via lone pair−π interactions (n−π interactions), where a lone pair from oxygen interacts with the π‐electron cloud of the aromatic ring [12, 13]. Weak OH−π interactions are generally more favorable with electron‐rich aromatic rings, while n−π interactions are preferred with electron‐deficient aromatic rings [8, 13]. Using the ortho‐substituted phenols, it was recently demonstrated that phenols can be stabilized by a balanced contribution of the stabilizing OH−R interactions and the avoidance of repulsive HO−R interactions [14]. While experimental and theoretical studies have provided invaluable insight into the nature of OH−π interactions in molecular recognition, the preferential orientation of the OH−π interactions has not been investigated to a similar level of molecular detail [15, 16, 17, 18].

Several simple molecular models have been designed to provide quantitative and qualitative information about the involvement of aromatic rings in stabilization of polar groups, including 2,6‐diarylaromatics, molecular balances, and related model systems [8, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41]. 2,6‐Diarylaromatics, in which the electronic properties of the two flanking aromatic rings can be fine‐tuned by substituents at a distant para‐position, are particularly suitable for investigations of the conformational preference of through‐space polar−π interactions. Here, we developed a molecular balance based on the 2,6‐diarylphenol scaffold in which two different flanking aromatic rings bear different substituents of opposite electronegativities, leading to an altered electron density of the aromatic rings (Figure 1). Molecular balances display an equilibrium between two different conformations, with the preferred conformer typically existing due to more favorable intramolecular noncovalent interactions [4, 5, 29, 31, 32, 34, 35, 36, 40, 41]. We hypothesized that the flanking aromatic rings possessing para‐substituents with large opposing Hammett values (e.g., −0.27 for OMe and +0.54 for CF_3_) might fine‐tune the electron density and electrostatic potential sufficiently so that the phenolic hydroxyl group might display a preference for the more electron‐rich aromatic ring over the electron‐deficient counterpart. In addition, the repulsive n−π interactions (i.e., O−π interactions involving the electron pair of the phenolic O atom) might be energetically less favorable for electron‐rich aromatic rings than with electron‐deficient aromatic rings, overall contributing to the preferred conformation of the molecular balances. We further aimed to probe whether an even smaller difference in the electronic properties between the flanking aromatic rings, caused by two more similar substituents (e.g., OMe and H), could still be recognized by the adjacent hydroxyl group of the central phenol.

The 2,6‐diarylphenol scaffold as a molecular balance for examinations of the conformational preference of attractive OH−π interactions and repulsive n−π interactions.

Results and Discussion

2

Substituted 2,6‐diarylphenols 1 and 2 were synthesized in several steps as shown in Scheme 1. 2,6‐Dibromophenyl acetate 3 was initially coupled with para‐OMe‐substituted boronic acid 4 to afford a monoaryl‐substituted product 5, which then reacted with para‐CF_3_‐substituted boronic acid under Suzuki conditions to produce the phenol acetate 6 that possesses two different aromatic substituents. In the final step, the O‐acetyl group of 6 was removed under basic conditions, affording phenol 1 in 64% overall yield. 2‐Phenylphenol 7 first reacted with N‐bromosuccinimide (NBS) in the presence of diisopropylamine to afford a brominated product 8 in 51% yield. The O‐acetyl protection of 8 yielded ester 9 that further reacted with para‐OMe‐substituted boronic acid through a palladium‐catalyzed Suzuki reaction to yield a doubly substituted phenol 10. The final step required a deprotection of the acetyl group in 10 with sodium hydroxide to afford phenol 2 in 78% yield over the last two steps.

Synthesis of the 2,6‐diarylphenols.

The solid‐state conformation of phenols 1 and 2 was first examined by x‐ray crystallography (Figures 2, S1–S6, and Tables S1–S2). The crystals of phenols 1 and 2 were obtained by slow evaporation from a solvent mixture of petroleum ether and dichloromethane (PE/DCM, v/v 3:1). In both structures, the OH group is positioned towards the more electron‐rich aromatic ring possessing the para‐methoxy substituent (Figures 2 and S3). Structural analyses provide evidence that none of the distant para‐substituents interacts with the phenolic OH group. The crystal system of 1 is triclinic, with one molecule in the asymmetric unit. The flanking rings display a staggered orientation when viewing from the plane of the central benzene ring. The dihedral angles of the flanking rings are 47° (φ_1_) and 42° (φ_2_), whereas the dihedral angle of the central hydroxyl group is 12° (φ_OH_) (Figure 2). The distance (d_r‐H_) between the OH group and the centroid of the aromatic flanking ring is 3.31 Å, which is consistent with through‐space OH−π interactions. The hydrogen of the hydroxyl group in 1 could be located in the difference Fourier map with no sign of a second preferred location (Figure S3). The OH group can also form intermolecular hydrogen bonds with the oxygen of the para‐methoxy group of another molecule (Figure S4). These hydrogen‐bonded dimers form layered structures along the ac‐plane (Figure S5). The crystal system of 2 is monoclinic, with one molecule in the asymmetric unit. The flanking rings are also found in a staggered conformation when viewing from the plane of the central benzene ring. The dihedral angles of the flanking rings are 46° (φ_1_) and 54° (φ_2_), and the dihedral angle of the hydroxyl group is 29° (φ_OH_). The shortest distance (d r‐H) between the OH group and the centroid of the aromatic flanking ring is 3.44 Å, a distance again indicative of through‐space OH−π interactions. The intermolecular hydrogen bonds between the OH and OMe groups can also exist in the crystal (Figure S6). The dimers display π−π stacking between the two pOMe rings with a central distance of 3.63 Å. The hydrogen of the hydroxyl group in 2 could be unambiguously assigned and is located close to the electron‐rich aromatic ring (Figure S3). Overall, the crystallographic results reveal a structural preference that is consistent with the phenolic OH group preferentially interacting with the more electron‐rich aromatic ring (MeO‐Ar) over the related electron‐poorer aromatic rings (H‐Ar or CF_3_‐Ar), indicating that the 2,6‐arylphenols possessing two different flanking rings can serve as molecular balances for probing OH−π interactions.

Crystal structures of phenols 1 (top) and 2 (bottom).

To determine whether these observations are unique to crystal packing or represent a shallow thermodynamic minimum, low‐temperature solution‐state NMR experiments were pursued. Compounds 1 and 2 were dissolved in diethyl ether‐d 10 and inserted into a probe pre‐cooled to −85°C and ^1^H NMR spectra were acquired for each. The OH peak of 1 moves substantially with temperature, which is typical behavior for hydrogen‐bonding hydroxyl protons (Figure 3a). The assignment of the aromatic protons was aided by a 2D COSY experiment (Figures 3b and S7). The most upfield doublet belongs to the H m as its ortho to the electron‐donating OMe group. This proton can be observed to couple to another doublet 7.59 ppm (H o). The protons of the p‐CF_3_ phenyl ring can be observed as an AB quartet between 7.8 and 7.9 ppm and thus engaged in second‐order strong coupling.

VT_NMR analyses of phenol 1. (a) 1H NMR spectrum of compound 1 in Et2O‐d 10 from −75°C to −100°C. (b) 2D COSY spectrum at −100°C.

The chemical shift dispersion was sufficient at −85°C to attempt 1D ROESY build‐up experiments to assess whether the OH ^1^H lies closer to one ring or the other (Figure S8). As the dipolar relaxation rate is inversely proportional to the distance between the two protons, the slope of the ROE vs. mix time (i.e., rate) provides information about the distance between two nuclei. The OH ^1^H was selectively inverted, and then a ROESY spinlock was applied for either 100, 200, or 500 ms (Figure S8). The integral of each ROE was normalized by the number of protons it represents and then divided by the total magnetization (positive ROEs—integral of the OH peak) in order to calculate the build‐up slopes (Table S3). The slope for H o belonging to the p‐OMe ring was found to be 15.6% s^−1^, while that for the protons of the p‐CF_3_ ring was found to be 7.7% s^−1^. The 2‐fold magnitude in rate for the p‐OMe protons over the p‐CF_3_ indicates that the OH lies closer to the p‐OMe than the p‐CF_3_ ring. As only one OH resonance is observed across the temperatures measured, it indicates that the directional preference observed is in fast‐exchange between the two positions, but that the one pointing towards the p‐OMe is more largely populated. A lack of significant broadening of the OH resonance also indicates that the barrier to interconversion is extremely fast, even at −100°C.

Compound 2 unfortunately, did not display the same chemical shift dispersion that allowed for the selective inversion of the OH resonance, most likely due to the more similar ring electronics between the p‐OMe and p‐H rings (Figure 4). Similar to 1, the protons of the p‐OMe could be deduced from the coupling of the most upfield doublet at 7.05 ppm with the doublet at 7.60 ppm. The p‐H ring protons were elucidated by the 2D COSY as well as the 2:2:1 doublet:triplet:triplet pattern expected for a monosubstituted benzene (Figures 4b and S9).

VT_NMR analyses of phenol 2. (a) 1H NMR spectrum of compound 2 in Et2O‐d 10 at −85°C and −100°C. (b) 2D COSY spectrum at −85°C.

Like 1, the OH resonance moves further downfield with decreasing temperature (Figure 4a). To still try to observe the ROEs between the OH and the ring protons, a high‐resolution 2D ROESY experiment was performed (Figure S10). 2D integral volumes representing magnetization transfer from the ortho protons of each ring to the OH peak were measured, and similar to 1, the magnetization from the p‐OMe protons to the OH was greater than that for the competing phenyl ring, albeit a somewhat smaller difference (2.02 vs. 1.85).

Further insight into the understanding of how the hydroxyl group interacts with the flanking substituted aryl rings can be obtained through quantum chemical computations. For such purpose, the above synthesized substituted 2,6‐diarylphenols 1 and 2 have been analyzed at the ZORA‐BLYP‐D3(BJ)/TZ2P level of theory, both in diethylether and in vacuo (Tables S4–S8) [42]. The equilibrium geometry of both possible isomers presents a staggered conformation, in agreement with experiment. The isomer in which the OH points towards the more electron‐rich ring (OMe, Table 1) is the most stable in all cases, with isomerization energies of 4.0 and 0.5 kJ mol^−1^ for 1 and 2, respectively. Computed equilibrium geometries agree well with the x‐ray crystallographic data above (Figure 2). In addition, computed isomerization energies in vacuo give the same trend, with 5.3 and 0.6 kJ mol^−1^ for 1 and 2, respectively (Table S4), proving a minor role of the solute‐solvent interactions in the conformational stabilization of these compounds. Furthermore, the computed rotational barrier when the φ_OH_ is varied in the range 0–360° amounts to about 24 and 23 kJ mol^−1^ in diethylether for 1 and 2, respectively, connecting both their isomers, i.e., with the OH pointing towards the electron‐richer and towards the electron‐poorer substituted aryl rings (Figure 5). The same barrier in vacuo increases above 26 kJ mol^−1^.

TABLE 1: Relative energy (in kJ mol−1) and structural data (Å, deg.) of the isomers in which the hydroxyl O–H points to the OMe‐substituted aryl ring of the substituted 2,6‐diarylphenols 1 and 2 relative to the isomer in which it points to the other aryl ring. Geometries of the isomers are depicted on top. a

Relative energy ∆E (in kJ mol−1) of the substituted 2,6‐diarylphenols 1 and 2 as a function of the rotation of the dihedral angle φOH (in degrees) in diethylether (blue) and in vacuo (red).

We also carried out additional computations using the crystallographically obtained dimeric 1‐1 and 2‐2 structures in diethylether and compared them with monomeric 1‐Et_2_O and 2‐Et_2_O structures in diethylether to further probe the solid state and solution results. Dimer 1‐1 presents a total bonding energy of −56.7 kJ mol^−1^ (∆E, Table S8), involving two O─H···O‐Me hydrogen bond interactions (Figure S11). For comparison, monomeric 1 interacts with one molecule of diethylether with ∆E = −39.9 kJ mol^−1^, involving one O─H···O hydrogen bond interaction. These results suggest that phenol–diethylether hydrogen bond interactions are stronger than phenol–phenol hydrogen bond interactions when considering the number of phenol units. The O···H length in the 1–1 dimer is 1.872 Å, whereas for 1 with diethylether it is 1.716 Å; this latter is shorter, thus also supporting the stronger hydrogen bond interaction. Through‐space OH−π interactions are observed in both 1–1 dimer and 1–Et_2_O complexes, as exemplified by similar distances between the OH group and the centroid of the aromatic flanking ring (Figure S11). Values of ∆E for phenol 2 are very similar (Table S8). Taken together, our results indicate that phenols 1 and 2 exist in monomeric forms in diethylether.

To further understand the origin of the preference for the isomer with the OH pointing towards the electron‐richer substituted aryl ring, we have carried out a Kohn–Sham molecular orbital analysis together with an energy decomposition analysis. To this end, we constructed a truncated model system that retains only the OH−π interaction. Specifically, we preserved the hydroxyl group and a single substituted aryl ring, maintaining the geometry found in either 1 or 2, and capped both fragments with a hydrogen atom. To prevent steric repulsion between the newly added hydrogen atoms, the water molecule was rotated while preserving the original H–O position of the hydroxyl group (Figure 6).

EDA fragments used for the analysis of the OH‐π interaction in substituted 2,6‐diarylphenols, based on the isomer in which the OH points towards the more electron‐rich ring (OMe, top) or the more electron‐poor ring (CF3/H, bottom). EDA data is enclosed in Table 2.

The hydroxyl group better interacts with the methoxy‐substituted aryl ring by 0.7 and 0.1 kJ mol^−1^ for 1 and 2, respectively (Table 2). For both systems 1 and 2, their preference for the isomer with OH pointing towards the electron‐richer aryl ring is due to their more favorable electrostatic interaction. Notably, the energy decomposition analysis reveals that dispersion interactions (∆E disp) contribute significantly to the overall stabilization of the OH–π complex (Table 2). The slightly larger dispersion contribution for the interaction with the methoxy‐substituted aryl ring in both systems 1 and 2 is consistent with the slightly shorter distance to the more electron‐rich methoxyphenyl ring (see Tables 1 and 2). In particular, ∆V elstat is more favorable for this isomer by 2.4 kJ mol^−1^ for compound 1, compensating, together with the also more favorable orbital interaction, its larger Pauli repulsion caused by the closer proximity of the hydroxyl H atom to the center of the aryl ring (Table 1). Compound 2 shows the same trends, although the differences are smaller, with ∆V elstat being more attractive by 0.5 kJ mol^−1^.

TABLE 2: Energy decomposition analysis (in kJ mol−1) of the OH‐π interaction in substituted 2,6‐diarylphenols. Fragments used in the EDA are depicted in Figure 6. a

The stronger electrostatic interaction between the hydroxyl and the electron‐richer aryl ring is caused by the more negatively charged atoms of this latter ring when compared to either substituted by electron‐withdrawing CF_3_ (isomer 1, CF_3_) or H (isomer 2, H). This is largely supported by the VDD charges (Figure 7). For instance, the total VDD charge of the aryl‐ring atoms (H also included) amounts to ‐0.005 e. in 1, OMe, compared to +0.046 e. in 1, CF_3_. The molecular electrostatic potential isosurfaces of the rings of the model systems considered in the EDA further support this observation (Figure 7). For completeness and with the aim to support the validity of the proposed model system above, the VDD charges for the untruncated systems 1 and 2 give the same trend (Figure 8).

VDD charges (top, in milli‐electrons) of the aryl‐ring carbon atoms considered in the EDA analysis for the isomers of compounds 1 and 2, and molecular electrostatic potential isosurfaces of the substituted aryl rings (bottom, isovalue = 0.03). Computed at ZORA‐BLYP‐D3(BJ)/TZ2P in vacuo.

VDD charges (top, in milli‐electrons) of the aryl‐ring carbon atoms of isomers 1, OMe, and 2, OMe, and their molecular electrostatic potential isosurfaces (bottom, isovalue = 0.03). Computed at ZORA‐BLYP‐D3(BJ)/TZ2P in diethylether.

As noted in the introduction, although OH–π interactions are favored with electron‐rich aromatic rings, O lone pair–π (n–π) interactions can also contribute. To quantify these interactions in our set of systems, we performed an equivalent Kohn–Sham molecular orbital analysis together with an EDA analysis, following the same protocol described above, but focusing on the aryl ring toward which the OH group is not oriented. To this end, a truncated model system retaining only the n–π interaction was constructed. Specifically, the model includes the hydroxyl group and a single substituted aryl ring, preserving the geometry found in either 1 or 2, with both fragments capped by hydrogen atoms. To avoid steric repulsion between the added hydrogen atoms, the water molecule was slightly rotated while maintaining the original distance between the center of the aryl ring and the hydroxyl oxygen (r–O) (Table 1 and Figure 9).

EDA fragments used for the analysis of the n‐π interaction in substituted 2,6‐diarylphenols, based on the isomer in which the OH points towards the more electron‐rich ring (OMe, top) or the more electron‐poor ring (CF3/H, bottom). EDA data is enclosed in Table 3.

In contrast to the OH–π interaction, the n–π interaction is repulsive. This repulsion increases markedly when the lone pair is oriented toward an electron‐richer aryl ring, such as one bearing the OMe substituent (Table 3). For system 1, the interaction becomes significantly more repulsive, increasing from +0.8 to +5.0 kJ mol^−^ ^1^ when the lone pair of O points toward CF_3_ and OMe substituents, respectively. Similarly, for system 2, ΔE int becomes more destabilizing from +3.6 to +4.8 kJ mol^−^ ^1^ when the lone pair is oriented toward H and OMe substituents, respectively. This enhanced repulsion with electron‐richer aryl rings arises primarily from a reduction in electrostatic stabilization, accompanied by a comparatively smaller increase in Pauli repulsion (Table 3). These results are consistent with a recent work by Smolyar and Cockroft [14], who have demonstrated that avoidance of n–π interactions plays a substantial role in driving attractive OH–π interactions.

TABLE 3: Energy decomposition analysis (in kJ mol−1) of the n‐π interaction in substituted 2,6‐diarylphenols. Fragments used in the EDA are depicted in Figure 9. a

Conclusion

3

Noncovalent interactions involving aromatic rings, like OH−π interactions, play crucial roles in chemical and biological systems, influencing protein–ligand binding and substrate–catalyst complexation. Using a specially designed 2,6‐diarylphenol molecular balance, we demonstrate the preferential interaction of the phenolic hydroxyl group with electron‐richer aromatic rings via through‐space OH−π interactions. Our integrated structural, NMR spectroscopic, and quantum chemical analyses underscore the sensitivity of the OH group to subtle variations in the electron density of the interacting aromatic rings. The x‐ray crystal structures provided direct evidence that the phenolic OH preferentially points towards the more electron‐rich aromatic ring over the competing electron‐deficient aromatic ring. The VT_NMR analyses showed that phenol 1 exhibits a clear preferential interaction of the OH group with the p‐OMe ring over the competing p‐CF_3_ ring. Moreover, phenol 2, in which relatively similar electronic properties exist in the competing p‐OMe and p‐H rings, demonstrated a weaker and less distinct preference for the p‐OMe ring. Additionally, the quantum chemical computations based on quantitative Kohn–Sham molecular orbital theory in combination with an energy decomposition analysis indicated that the more favorable OH−π interaction between the hydroxyl group and the electron‐richer aryl ring is driven by favorable electrostatic and orbital interactions. However, avoidance of repulsive n–π interactions appears to play a determinant role in driving the formation of the OH−π interaction between the two competing aromatic rings. Our research could guide the development of novel drugs with enhanced specificity and affinity, as well as more efficient catalysts with improved selectivity.

Experimental Section

4

General Experimental Procedures

4.1

A Buchi 535 melting point apparatus was used for measurements of melting points. A Bruker Avance III 400 MHz NMR spectrometer was used for acquiring the ^1^H, ^13^C, and ^19^F NMR spectra. A Bruker Daltonics‐micrOTOF‐Q II‐ESI‐Qq‐TOF mass spectrometer was used for acquiring the high‐resolution mass spectra (HRMS). Chemical shifts (δ) are reported in parts per million (ppm) referenced to the resonance of the residual CDCl_3_ solvent (^1^H: 7.26 ppm; ^13^C: 77.16 ppm). Splitting patterns are reported as follows: singlet (s); doublet (d); doublet of doublets (dd); doublet of doublet of doublet (ddd); triplet (t); triplet of doublet (td); quartet (q); multiplet (m). Coupling constants (J) are reported in Hertz (Hz). All reagents and solvents were purchased from commercial vendors.

Synthesis of 3‐bromo‐4'‐methoxy‐[1,1'‐biphenyl]‐2‐yl acetate ( ** 5 **). Aq. Na_2_CO_3_ (2 M, 1.70 mL, 2 equiv) was added to a mixture of 2,6‐dibromophenol acetate 3 (500 mg, 1.71 mmol), 4‐methoxyphenylboronic acid 4 (286 mg, 1.88 mmol, 1.1 equiv) and Pd(PPh_3_)4 (99 mg, 0.08 mmol, 0.05 equiv) in the solvent mixture of toluene/THF (6 mL, v/v 1:1). After heating at 110°C using an oil bath for 24 h under argon atmosphere, the reaction mixture was poured into 20 mL of water and extracted with ethyl acetate (3 × 30 mL). The combined organic layers were washed with brine, dried over MgSO_4_, filtered, and concentrated. The crude residue was purified by flash column chromatography on silica gel (PE/EA, 15:1) to give the 3‐bromo‐4'‐methoxy‐[1,1'‐biphenyl]‐2‐yl acetate 5 (445 mg, 81%) as a colorless oil. ^1^H NMR (400 MHz, CDCl_3_) δ 7.57 (dd, J = 8.0, 1.6 Hz, 1H), 7.37–7.29 (m, 3H), 7.17 (t, J = 7.8 Hz, 1H), 6.99–6.91 (m, 2H), 3.85 (s, 3H), 2.15 (s, 3H); ^13^C NMR (101 MHz, CDCl_3_) δ 168.2, 159.5, 145.8, 137.1, 132.0, 130.0, 129.5, 127.4, 117.5, 113.9, 55.4, 20.7; HRMS (ESI): m/z calcd for C_15_H_13_BrNaO_3_ [M + Na]^+^: 342.9937, found 342.9940.

Synthesis of 4‐methoxy‐4''‐(trifluoromethyl)‐[1,1':3',1''‐terphenyl]‐2'‐yl acetate ( ** 6 **). An aqueous solution of Na_2_CO_3_ (2 M, 0.90 mL, 2 equiv) was added to a mixture of 3‐bromo‐4'‐methoxy‐[1,1'‐biphenyl]‐2‐yl acetate 5 (290 mg, 0.90 mmol), 4‐trifluoromethylphenylboronic acid (257 mg, 1.35 mmol, 1.5 equiv) and Pd(PPh_3_)4 (52 mg, 0.05 mmol, 0.05 equiv) in toluene (5 mL). After it was heated at 110°C using an oil bath for 24 h under argon atmosphere, the reaction mixture was poured into 15 mL of water and extracted with ethyl acetate (3 × 20 mL). The combined organic layers were washed with brine, dried over MgSO_4_, filtered, and concentrated. The crude residue was purified by flash column chromatography on silica gel (PE/EA, 20:1) to give the 4‐methoxy‐4''‐(trifluoromethyl)‐[1,1':3',1''‐terphenyl]‐2'‐yl acetate 6 (280 mg, 80%) as a colorless oil. ^1^H NMR (400 MHz, CDCl_3_) δ 7.61–7.54 (m, 2H), 7.51–7.45 (m, 2H), 7.36–7.20 (m, 5H), 6.92–6.82 (m, 2H), 3.72 (s, 3H), 1.72 (s, 3H); ^13^C NMR (101 MHz, CDCl_3_) δ 169.0, 159.3, 145.2, 141.8, 135.9, 134.6, 130.9, 130.2, 129.96, 129.9, 129.5, 129.5, 126.7, 125.3 (q, J = 3.7 Hz), 121.9 (q, J = 273 Hz), 115.5, 113.9, 55.3, 20.6; ^19^F NMR (376 MHz, CDCl_3_) δ ‐62.4. HRMS (ESI): m/z calcd for C_22_H_17_F_3_NaO_3_ [M + Na]^+^: 409.1022, found 409.1004.

Synthesis of 4‐methoxy‐4''‐(trifluoromethyl)‐[1,1':3',1''‐terphenyl]‐2'‐ol ( ** 1 **). A solution of NaOH (49 mg, 1.09 mmol, 3 equiv) in water (3 mL) was added to a solution of 4‐methoxy‐4''‐(trifluoromethyl)‐[1,1':3',1''‐terphenyl]‐2'‐yl acetate 6 (140 mg, 0.36 mmol) in THF/MeOH (6 mL, v/v 1:1) at an ice bath. After stirring for another 3 h, the reaction was quenched with aq. HCl (2 M, 20 mL). The mixture was extracted with ethyl acetate (3 × 10 mL). The combined organic layers were washed with sat. aq. NaHCO_3_, brine, dried over MgSO_4_, filtered, and concentrated to give the 4‐methoxy‐4''‐(trifluoromethyl)‐[1,1':3',1''‐terphenyl]‐2'‐ol 1 (245 mg, 98%) as a white solid. mp 115–116°C; ^1^H NMR (400 MHz, CDCl_3_) δ 7.78–7.71 (m, 4H), 7.49–7.43 (m, 2H), 7.30 (ddd, J = 8.6, 7.6, 1.8 Hz, 2H), 7.15–7.01 (m, 3H), 5.44 (s, 1H), 3.89 (s, 3H); ^13^C NMR (101 MHz, CDCl_3_) δ 159.7, 149.6, 142.0, 130.6, 130.5, 130.0, 129. 9, 129.0, 128.9, 127.4, 125.4 (q, J = 3.8 Hz), 124.4 (q, J = 273 Hz), 121.0, 114.9, 55.5; ^19^F NMR (376 MHz, CDCl_3_) δ ‐62.4; HRMS (ESI): m/z calcd for C_20_H_15_F_3_NaO_2_ [M + Na]^+^: 367.0916, found 367.0902.

Synthesis of 3‐bromo‐[1,1'‐biphenyl]‐2‐ol ( ** 8 **). Diisopropylamine (83 uL, 0.59 mmol, 0.1 equiv) was added to a solution of [1,1'‐biphenyl]‐2‐ol 7 (1.0 g, 5.88 mmol) in anhydrous DCM (10 mL) under an argon atmosphere. Then NBS (1.05 g, 5.88 mmol, 1 equiv) was added in four portions. After stirring at room temperature overnight, the reaction was treated with aq. HCl (1 M, 25 mL). The mixture was extracted with DCM (3 × 30 mL). The combined organic layers were washed with sat. aq. NaHCO_3_, brine, dried over MgSO_4_, filtered, and concentrated. The crude residue was purified by flash column chromatography on silica gel (PE/EA, 50:1) to give the 3‐bromo‐[1,1'‐biphenyl]‐2‐ol 8 (747 mg, 51%) as a pale yellow liquid. ^1^H NMR (400 MHz, CDCl_3_) δ 7.58–7.53 (m, 2H), 7.49 (ddt, J = 9.1, 6.6, 1.1 Hz, 3H), 7.41 (tt, J = 7.26, 1.4 Hz, 1H), 7.28 (dd, J = 7.6, 1.6 Hz, 1H), 6.91 (t, J = 7.8 Hz, 1H), 5.71 (s, 1H); ^13^C NMR (101 MHz, CDCl_3_) δ 149.4, 137.3, 131.6, 130.3, 129.8, 129.3, 128.7, 127.9, 121.8.

Synthesis of 3‐bromo‐[1,1'‐biphenyl]‐2‐yl acetate ( ** 9 **). Ac_2_O (831 uL, 8.80 mmol, 1.5 equiv) was added to a solution of 3‐bromo‐[1,1'‐biphenyl]‐2‐ol 8 (1.46 g, 5.86 mmol) and DMAP (72 mg, 0.59 mmol) in anhydrous pyridine (25 mL) at 0°C using an ice bath under argon atmosphere. The reaction was allowed to warm to room temperature and further stirred at room temperature overnight. The solvent was then evaporated, and the residue was dissolved in ethyl acetate (50 mL), and washed with aq. HCl (1 M, 50 mL), brine. The separated organic layer was dried over MgSO_4_, filter and concentrated. The crude residue was purified by flash column chromatography on silica gel (PE/EA, 50:1) to give the 3‐bromo‐[1,1'‐biphenyl]‐2‐yl acetate 9 (1.61 g, 95%) as a colorless oil. ^1^H NMR (400 MHz, CDCl_3_) δ 7.61 (dd, J = 8.0, 1.6 Hz, 1H), 7.45–7.32 (m, 6H), 7.20 (t, J = 7.8 Hz, 1H), 2.13 (s, 3H); ^13^C NMR (101 MHz, CDCl_3_) δ 168.3, 145.8, 137.5, 137.2, 132.5, 130.1, 128.9, 128.5, 128.0, 127.5, 117.6, 20.7.

Synthesis of 4‐methoxy‐[1,1':3',1''‐terphenyl]‐2'‐yl acetate ( ** 10 **). Aq. Na_2_CO_3_ (2 M, 0.35 mL, 2 equiv) was added to a mixture of 3‐bromo‐[1,1'‐biphenyl]‐2‐yl acetate 9 (100 mg, 0.35 mmol), 4‐methoxyphenylboronic acid (104 mg, 0.69 mmol, 2 equiv) and Pd(PPh_3_)4 (20 mg, 0.02 mmol, 0.05 equiv) in the solvent mixture of toluene/THF (3 mL, v/v 1:1). After it was heated at 110°C using an oil bath for 24 h under argon atmosphere, the reaction mixture was poured into 10 mL water and extracted with ethyl acetate (3 × 15 mL). The combined organic layers were washed with brine, dried over MgSO_4_, filtered, and concentrated. The crude residue was purified by flash column chromatography on silica gel (PE/EA, 50:1) to give the 4‐methoxy‐[1,1':3',1''‐terphenyl]‐2'‐yl acetate 10 (105 mg, 96%) as a colorless oil. ^1^H NMR (400 MHz, CDCl_3_) δ 7.52–7.31 (m, 10H), 7.00–6.93 (m, 2H), 3.84 (s, 3H), 1.83 (s, 3H); ^13^C NMR (101 MHz, CDCl_3_) δ 169.0, 159.2, 145.3, 138.1, 135.9, 135.6, 130.3, 130.2, 130.1, 129.8, 129.1, 128.3, 127.5, 126.5, 113.8, 55.3, 20.6; HRMS (ESI): m/z calcd for C_21_H_19_O_3_ [M + H]^+^: 319.1329, found 319.1318.

Synthesis of 4‐methoxy‐[1,1':3',1''‐terphenyl]‐2'‐ol ( ** 2 **). A solution of NaOH (40 mg, 0.99 mmol, 3 equiv) in water (3 mL) was added to a solution of 4‐methoxy‐[1,1':3',1''‐terphenyl]‐2'‐yl acetate 10 (105 mg, 0.33 mmol) in THF/MeOH (6 mL, v/v 1:1) at an ice bath. After stirring for another 3 h, the reaction was quenched with aq. HCl (2 M, 20 mL). The mixture was extracted with ethyl acetate (3 × 10 mL). The combined organic layers were washed with sat. aq. NaHCO_3_, brine, dried over MgSO_4_, filtered, and concentrated to give the 4‐methoxy‐[1,1':3',1''‐terphenyl]‐2'‐ol 2 (87 mg, 96%) as a white solid. mp 114−116°C; ^1^H NMR (500 MHz, CDCl_3_) δ 7.62–7.57 (m, 2H), 7.54–7.48 (m, 4H), 7.44–7.39 (m, 1H), 7.31–7.27 (m, 2H), 7.10–7.02 (m, 3H), 5.44 (s, 1H), 3.88 (s, 3H); ^13^C NMR (126 MHz, CDCl_3_) δ 159.3, 149.5, 137.8, 130.6, 130.0, 129.8, 129.7, 129.5, 128.9, 128.8, 128.5, 127.7, 120.8, 114.4, 55.4; HRMS (ESI): m/z calcd for C_19_H_17_O_2_ [M + H]^+^: 277.1223, found 277.1214.

Single Crystal X‐Ray Crystallography

4.2

The data for compounds 1 and 2 were collected on a Bruker D8 Quest diffractometer with a sealed tube and Triumph monochromator (λ = 0.71073 Å). Saint (v8.40a) was used for the intensity integration [43]. SADABS was used for absorption correction [44]. Both structures were solved with direct methods using SHELXT‐2014/5 [45] and the least‐squares refinement was carried out by SHELXL‐2018/3 [46] against |Fho|2 of all the reflections. Non‐hydrogen atoms were refined freely with anisotropic displacement parameters. Hydrogen atoms were placed on calculated positions or located in difference Fourier maps. A riding model was used for refining all calculated hydrogen atoms.

Deposition Numbers 2429368 (for 1) and 2429369 (for 2) contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service.

NMR Analyses

4.3

Samples were dissolved in approximately 0.6 mL Et_2_O‐d 10 and transferred to a Bruker 300 MHz equipped with a BBFO probe that had been pre‐cooled to −85°C using a heat exchanger submerged in LN_2_ dewar. 1D ^1^H spectra were acquired with the zg30 pulse sequence, NS = 8 and a D1+AQ = 1 + 8.4 = 9.4 s. Selective ROESY experiments were performed by integrating the OH resonance and setting up the selective experiment through the sel1d command in TopSpin. Spectra were acquired with various mix times, 32 scans, and a D1+AQ = 2 + 4.2 = 6.2 s. 2D ROESY spectra were acquired with a spectral width of 1.5 ppm focused around the aromatic protons centered at 7.15 ppm. A mix time of 300 ms, 8 scans per increment, and 192 increments in the indirect dimension were chosen, yielding a resolution of approximately 2.34 Hz in the indirect dimension. 2D COSY spectra were acquired with a spectral width of 1.44 ppm focused around the aromatic protons centered at 7.29 ppm, 1 scan per increment with 64 increments in the F1 dimension.

Quantum Chemical Analyses

4.4

All quantum chemical computations were performed using the Amsterdam Density Functional (ADF) program within dispersion‐corrected density functional theory at the ZORA‐BLYP‐D3(BJ)/TZ2P level of theory [42, 47, 48]. Solvent effects in diethylether were modeled using the Conductor‐Like Screening Model (COSMO) as incorporated in ADF. Both the OH−π‐ring and O−π‐ring interactions were examined using quantitative Kohn–Sham molecular orbital theory combined with a quantitative energy decomposition analysis (EDA) in vacuo [49, 50, 51, 52]. The interaction energy (ΔE int) between the referred fragments was decomposed into electrostatic interactions (ΔV elstat), Pauli repulsion between occupied orbitals (ΔE Pauli), stabilizing orbital interactions (ΔE oi), and dispersion contributions (ΔE disp). Finally, the atomic charges were determined using the Voronoi deformation density (VDD) method [53, 54].

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: asia70652‐sup‐0001‐SuppMat.pdf.

Supporting File 2: asia70652‐sup‐0002‐Data.zip.