Hydrophobicity Study of Melamine, 1,3,5-Benzenetricarbonitrile, and 1,3,5-Triaminobenzene

Abstract

The molecular hydrophobicity of melamine (MA), 1,3,5-triaminobenzene (TAB), and 1,3,5-benzenetricarbonitrile (BTCN) was investigated using various methods. Hydrophobicity is influenced by charge distribution, polar functional groups, and solvent interactions. Our findings indicate that at ultralow concentrations, these compounds are fully dispersed in aqueous solutions, with hydrophobicity decreasing in the order MA > TAB > BTCN. Interestingly, when the concentration surpasses a critical threshold, the hydrophobicity order undergoes a complete reversal: BTCN > MA > TAB. This concentration-dependent inversion offers a fundamental theoretical basis for the molecular design of stimuli-responsive hydrophobic materials.

1. Introduction

Hydrophobic materials, owing to their unique interfacial repulsion properties, possess irreplaceable application value in cutting-edge fields such as self-cleaning coatings, , anticorrosion films, − drug delivery, − hydrophobic coatings, , selective adsorption membranes, − catalytic supports, , oil–water separation membranes, , biocompatible interfaces, and microfluidic devices. The essence of molecular hydrophobicity originates from the distribution of weakly polar/nonpolar groups in its chemical structure, the characteristics of electron cloud density, and the intermolecular stacking patterns. Melamine (MA, 1,3,5-triazine-2,4,6-triamine), 1,3,5-benzenetricarbonitrile (BTCN), and 1,3,5-triaminobenzene (TAB) are three typical nitrogen-containing aromatic ring compounds. They have become emerging research objects due to their controllable hydrophobic behavior and multifunctionality. However, the structural basis of their hydrophobic mechanism has not been systematically elucidated.

The rigid symmetrical triazine ring (C₃N₃) of MA has electronic-rich properties. Its planar structure can form a hydrophobic interface through π–π stacking. However, the three amino groups (−NH₂) around the triazine ring have strong hydrogen bond donor capacity and are prone to form a hydrophilic network with water molecules, resulting in the hydrophilicity of the MA monomer. This contradictory characteristic makes its hydrophobicity highly dependent on the molecular self-assembly form. In supramolecular frameworks (such as hydrogen-bonded organic frameworks, HOFs), amino groups are “shielded” through intramolecular/intermolecular hydrogen bonds, exposing the hydrophobic surface of the triazine ring, thereby significantly enhancing macroscopic hydrophobicity.

The three cyanogroups (−CN) of the benzene ring in the BTCN structure have an electron-withdrawing effect and low polarity, significantly reducing the electron cloud density of the benzene ring. The linear geometric structure of the cyano group weakens the intermolecular interactions, , while its high bond energy (∼891 kJ/mol) endows it with chemical inertness. This characteristic enables BTCN to exhibit excellent hydrophobic stability, with a contact angle of over 120°. Its hydrophobicity stems from the fact that the electron loss in the benzene ring conjugated system weakens the electrostatic interaction with the water dipole.

Although TAB contains three hydrophilic amino groups, the central symmetry of its benzene ring prompts the molecule to form two-dimensional or three-dimensional supramolecular networks through N–H···N hydrogen bonds. This self-assembly can encapsulate the amino group within the structure, exposing the nonpolar benzene ring plane and forming a graphene-like hydrophobic surface. , It is worth noting that the hydrophobicity of TAB is dynamically dependent on the stability of the hydrogen bond network.

Current research mostly focuses on the hydrophobic properties of single molecules, but there is a lack of comparative analysis of the structural origins of the differences in hydrophobicity among the three. This study will combine quantum chemical calculations with molecular dynamics simulations to systematically reveal the hydrophobic structure–activity relationship among MA, BTCN and TAB, providing a theoretical model for the directional design of functionalized hydrophobic materials.

2. Results and Discussion

2.1. XLog P Values

The XLog P values of MA, BTCN and TAB calculated by XLOGP3 software are shown in Table .

1. Calculation Results of XLog P Values.

	MA	BTCN	TAB
XLog P	1.83	–0.91	–0.12

As indicated in Table , the XLog P values for MA, BTCN, and TAB are 1.83, −0.91, and −0.12, respectively. By comparing the values of the three, it can be observed that the XLog P of MA is 1.83, indicating hydrophobicity. The XLog Ps of BTCN and TAB are both negative (−0.91 and −0.12 respectively), indicating that they are hydrophilic overall. This difference mainly stems from the different substituents. The amino group (−NH₂) in MA has a strong electron-donating ability on the benzene ring, but the hydrophobic structure of the triazine ring itself dominates, making it hydrophobic as a whole. The cyano group (−CN) in BTCN is a strongly polar group, significantly enhancing the molecular hydrophilicity. Although TAB also contains multiple amino groups, its electronic effect on the benzene ring and molecular conjugated structure make its hydrophilicity weaker than that of BTCN, and thus its XLog P is slightly higher. In summary, the order of hydrophobicity of the three is MA > TAB > BTCN.

2.2. Molecular Structure Optimization

The molecular structures of MA, BTCN, and TAB were subjected to optimization via density functional theory (DFT) calculations. The optimized molecular structures are shown in Figure .

1.

Charge distribution diagrams of (a) MA, (b) BTCN, and (c) TAB. Side views of the molecular structures of (d) MA, (e) BTCN, and (f) TAB. Electrostatic potential distributions of (g) MA, (h) BTCN, and (i) TAB. In the models, blue spheres represent nitrogen atoms, gray spheres represent carbon atoms, and white spheres represent hydrogen atoms.

As illustrated in Figure , the nitrogen (N) atom within the heterocyclic ring of the MA molecule carries a charge of −0.336e, while the N atom in the side chain has a charge of −0.702e. All atoms within the MA molecule reside in the same plane. In the BTCN molecule, the N atom of the side chain exhibits a charge of −0.138e, and all its atoms are coplanar. Regarding the TAB molecule, the N atom of the side chain bears a charge of −0.772e, and the hydrogen atoms on its side chain are not coplanar with the atoms of the benzene ring.

The extent of the charge on the nitrogen (N) atom significantly influences its propensity to form hydrogen bonds with water molecules. Specifically, a higher charge on the N atom facilitates the formation of hydrogen bonds, thereby diminishing its hydrophobicity. Amino groups can not only provide H atoms to form hydrogen bonds (as donors), but also accept hydrogen bonds through the lone pair of electrons of N atoms (as acceptors). The nitrile group has no available H atom and can only serve as a weak acceptor for hydrogen bonds, interacting with H in water molecules through the lone pair of electrons of the N atom. However, this interaction is much weaker than that of the hydrogen bond of the amine group. Consequently, based on the charge distribution of the N atoms within the side chains, the hydrophobicity strength can be ranked as follows: BTCN > MA > TAB.

Hydrophobicity occurs because a molecular surface cannot form strong electrostatic interactions with water molecules. When the molecular surface has either a negligible region of negative electrostatic potential (indicating an inability to attract water, such as with H⁺) or a broad area of weak positive electrostatic potential (which repels water, such as with O^–), it becomes difficult to establish hydrogen bonds or electrostatic attractions with water, resulting in hydrophobic behavior. The TAB molecule contains a wide range of negative electrostatic regions and is prone to attracting water molecules to form hydrogen bonds. BTCN contains a wide range of weak positive electrostatic potential regions, making it difficult to attract water molecules to form hydrogen bonds. The MA molecule contains a small number of negative electrostatic regions and a large number of weak negative electrostatic regions, and its ability to attract water molecules lies between TAB and BTCN. Based on the electrostatic potential distribution diagram, the relative degree of hydrophobicity can be ranked as follows: BTCN > MA > TAB.

Natural bond orbital (NBO) analysis was conducted to clarify the electronic structure and charge distribution within the molecules being studied. The findings revealed that the cyano groups in BTCN exert a significant electron-withdrawing effect, resulting in a decrease in its dipole moment. Specifically, BTCN showed a markedly lower dipole moment (338 au) compared to the reference molecules MA (1040 au) and TAB (1203 au). Although each CN bond has a large dipole moment, in a highly symmetric configuration, the dipole moments of these individual bonds partially or nearly completely cancel each other out on the vector. This leads to a significant reduction in the net dipole moment of the entire molecule. This overall reduction in polarity directly weakens the interaction between BTCN molecules and water molecules. On the one hand, the weakened net dipole is difficult to have effective dipole–dipole interactions with strongly polar water molecules; On the other hand, charge delocalization also reduces the ability of cyano nitrogen atoms to act as hydrogen bond acceptances. In contrast, MA and TAB lack such potent electron-withdrawing groups, resulting in less uniform charge distributions and, consequently, larger dipole moments. These insights provide valuable perspectives on the electronic properties of these molecules and their potential intermolecular interactions. Notably, the dipole moment is closely correlated with molecular polarity and hydrophobicity; a higher dipole moment signifies greater polarity and reduced hydrophobicity. Consequently, based on the dipole moment values, the hydrophobicity order of the molecules can be ranked as BTCN > MA > TAB.

2.3. Molecular Polarity Index

The system’s polarity was assessed using Multiwfn, which analyzed the distribution characteristics of electrostatic potential on the molecular surface. The pertinent parameters for MA, BTCN, and TAB concerning the Molecular Polarity Index (MPI, for short) are presented in Table .

2. Calculation Results of the Molecular Polarity Index.

	total surface area of the system (Å2)	proportion of the nonpolar surface area (%)	proportion of the polar surface area (%)	MPI polarity index (kcal/mol)
MA	156.3696	23.30	76.70	16.7573
BTCN	197.9696	22.72	77.28	18.5450
TAB	169.8235	25.49	74.51	17.0940

According to the data presented in Table , the molecular polarity indices (MPI) of MA, BTCN, and TAB are 16.7573, 18.5450, and 17.0940 kcal/mol, respectively. The corresponding percentages of molecular polar surface area (PSA) relative to the total surface area are 76.70, 77.28, and 74.51%, respectively. A higher MPI value indicates greater overall molecular polarity. For neutral molecules, an increased proportion of polar surface area generally correlates with enhanced molecular polarity. From molecular structural perspective, the observed polarity trend primarily arises from differences in the polarity and electronic effects of the substituents present in each molecule. BTCN exhibits the highest polarity, attributable to the presence of three strong electron-withdrawing cyano groups (−CN) attached to its benzene ring, which significantly enhance intramolecular charge separation. Although TAB contains three electron-donating amino groups (−NH₂), its overall polarity is lower than that of BTCN. This can be explained by the conjugation of the lone pairs on the amino groups with the π-system of the benzene ring, which partially delocalizes the electron density and reduces their localized polar character. MA is based on a symmetric triazine ring structure; despite containing nitrogen atoms, the high symmetry and extensive conjugation within the ring system result in a more uniform distribution of electron density, leading to relatively lower overall polarity.

Thus, the consistent trends observed in both MPI and PSA values confirm the following polarity order: BTCN > TAB > MA. This sequence is inversely correlated with the hydrophobicity strength (MA > TAB > BTCN), which aligns with the fundamental principle that increased molecular polarity leads to decreased hydrophobicity.

2.4. Analysis of Molecular Dynamics Simulation Results

According to the manual, the solubility of MA in water is approximately 0.3 g/100 mL at 20 °C, and it increases significantly with rising temperature. According to the information from the chemical professional database of Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, the calculated values of the inherent solubilities of MA, BTCN and TAB are 6.392, 0.342, and 387.787 g/L, respectively. To evaluate the differences in hydrophobicity among these three substances, appropriate periodic simulation systems were established. Three types of systems were established: the first with 10 research object molecules and 10,000 water molecules, the second with 60 research object molecules and 10,000 water molecules, and the third with 180 research object molecules and 10,000 water molecules.

Using Packmol software, 10, 60, and 180 MA molecules (or BTCN, TAB) and 10,000 water molecules were placed at opposite ends of the rectangular box to establish a periodic model. After energy minimization, followed by NPT and NVT equilibrations, a 10 ns production simulation was conducted to obtain a stable structural model. The models before and after optimization are depicted in Figure . The radial distribution curves are presented in Figure .

2.

Structural models before and after molecular dynamics optimization. Models a–c represent the structural configurations of MA, BTCN, and TAB, respectively, before molecular dynamics optimization. (Consider the systems comprising 60 research object molecules and 10,000 water molecules as examples.) Models d–f represent the optimized structures for MA, BTCN, and TAB, respectively, with 10 research object molecules each. Similarly, models g–i and models j–l correspond to the 60 research object molecule and 180 research object molecule systems, respectively (the optimized models hide water molecules for observation. The subsequent analysis takes the systems comprising 60 research object molecules and 10,000 water molecules as examples).

3.

Pair radial distribution functions (RDFs) of solute molecules from molecular dynamics simulations in aqueous solution. Curves a–c represent the radial distribution functions for MA, BTCN, and TAB in systems comprising 60 research object molecules and 10,000 water molecules, respectively.

As shown in Figure , following the optimization of the MA-water mixture model, a small number of MA molecules aggregate to form clusters. In the case of the BTCN-water mixture model after optimization, the tendency to agglomerate is significantly more pronounced. These compounds show complete dispersion in water at extremely low concentrations, as indicated by the small red circles in Figure g,h. In contrast, upon optimizing the TAB-water mixture model, the TAB molecules display a relatively higher degree of dispersion.

Analysis of the radial distribution functions, g(r), and coordination numbers, N(r), for the three compounds reveals striking differences in their aggregation behaviors in aqueous solution. MA exhibits distinct g(r) peaks between 0.5 and 1.0 nm, indicative of strong intermolecular interactions, and an N(r) of ∼20 at 2.5 nm, consistent with an extended aggregated network. BTCN shows an intensified version of this trend, with more densely spaced g(r) peaks and a markedly larger N(r) of ∼50 at 2.5 nm, signifying the formation of larger and denser aggregates. Conversely, TAB presents a qualitatively different profile: its g(r) features are broader and weaker (maximum g(r) ≈ 8), and its N(r) plateaus at only ∼14. These data collectively establish a clear trend in aggregate size and interaction strength: BTCN > MA > TAB.

In the MA-water mixture model, there are 50 hydrogen bonds between MA molecules. The number of hydrogen bonds between MA and water molecules is 378, of which 225 are formed by the side chains and water molecules. In the BTCN-water mixed model, there are 75 hydrogen bonds between BTCN and water molecules, and no hydrogen bonds between BTCN molecules. In the TAB-water mixed model, there are 359 hydrogen bonds between TAB and water molecules, and 2 hydrogen bonds between TAB molecules. MA forms more hydrogen bonds with water molecules, mainly because hydrogen bonds can not only be formed between the N atoms the side chains and water molecules, but also be formed between the N atoms on the heterocyclic rings and water molecules. In these systems, the connection between solutes and water molecules mainly relies on hydrogen bonds. When the strength of hydrogen bonds is not significantly different, the more hydrogen bonds there are between the solute and water molecules, the stronger the attraction between them and water, and the weaker their hydrophobicity. Based on the results of molecular dynamics simulations, the order of hydrophobicity strength is BTCN > MA > TAB.

In the systems containing 10 solute molecules and 10,000 water molecules, all the solutes are dispersed in the water. Calculate the solvent-accessible area of each solute molecule and obtain the average value after statistics. The average solvent accessible areas per molecule of BTCN, MA and TAB are 3.4547, 2.9138, and 3.0898 nm², respectively. In the systems containing 60 solute molecules and 10,000 water molecules, the solvent-accessible surface areas (SASA) of BTCN, MA, and TAB were 76.419, 109.181, and 166.280 nm², respectively. The total surface area is calculated based on the average solvent-accessible area of each molecule, and the proportions of sasa in the cluster to the total surface area of BTCN, MA and TAB are 0.3687, 0.6245, and 0.8969, respectively. The proportion of SASA provides insights into the extent to which clusters are exposed to solvents. A low proportion of SASA indicates that the internal environment of the cluster is isolated from water, which may have an impact on its solubility, reactivity and biological activity. Conversely, a high SASA proportion reflects a more open and extended molecular conformation, which facilitates extensive hydration. In such cases, the molecules’ surface area is more accessible to water molecules, leading to increased solvent interaction and potentially influencing its physical and chemical properties. Consequently, the observed disparities in solvent-accessible surface area (SASA) among BTCN, MA, and TAB underscore that the hydrophobicity strength follows the order: BTCN > MA > TAB.

2.5. Comparative Analysis of Hydrophobicity Determined via Diverse Analytical Approaches

The sequence of molecular hydrophobicity obtained by various analytical methods is shown in Table .

3. Sequence of Molecular Hydrophobicity Obtained by Different Analytical Methods.

judging basis	the order of hydrophobicity strength
XLog P	MA > TAB > BTCN
the charge distribution of the N atoms within the side chains	BTCN > MA > TAB
natural bond orbital (NBO) analysis	BTCN > MA > TAB
the molecular polarity index (MPI)	MA > TAB > BTCN
the solvent accessible surface area (SASA)	BTCN > MA > TAB
hydrogen-bond counts	BTCN > MA > TAB

XLog P serves as a widely employed metric for assessing the lipid solubility of molecules. The Molecular Polarity Index (MPI) is a parameter utilized to quantify molecular polarity. A higher MPI value corresponds to greater molecular polarity and stronger interactions with water molecules. XLog P and MPI serve as metrics for assessing the hydrophobic characteristics of individual molecules within an aqueous environment. Among the compounds under investigation, the order of hydrophobicity strength is MA > TAB > BTCN.

This order (MA > TAB > BTCN) reflects the intrinsic hydrophobicity of isolated molecules. The charge distribution of nitrogen (N) atoms within the side chains significantly influences the molecule’s hydrophobicity. Notably, BTCN exhibits the least pronounced charge distribution of N atoms, which may result in weaker interactions with water molecules, thereby conferring stronger hydrophobicity. Natural bond orbital (NBO) analysis provides insights into the extent of electron delocalization within molecules. The NBO analysis of BTCN reveals the highest degree of electron delocalization, which may facilitate the reduction of electron cloud overlap with water molecules, consequently enhancing its hydrophobicity. The solvent-accessible surface area (SASA) represents the surface area of a molecule that is accessible to solvent molecules. Although a larger SASA value implies a greater contact area between the molecule and the solvent, it does not invariably correlate with stronger hydrophobicity. In the case of BTCN, however, its largest SASA value may be attributed to its unique molecular structure, which could potentially mitigate direct interactions with water molecules, thus manifesting stronger hydrophobicity. The number of hydrogen bonds serves as a crucial indicator of the interaction strength between molecules and water. An increased number of hydrogen bonds between the examined molecule and water, coupled with a decreased number of intramolecular hydrogen bonds, generally signifies stronger molecular-water interactions and weaker hydrophobicity.

Molecular hydrophobicity is influenced by various factors, including charge distribution, polar groups, and the solvent environment. Based on the aforementioned analyses, when the concentrations of MA, BTCN, and TAB in water are sufficiently low to ensure complete dispersion, the strength of hydrophobicity is determined primarily by the magnitude of the interaction between solute and water molecules. The order of hydrophobicity strength can be inferred as MA > TAB > BTCN at low concentrations. However, at high concentrations, it depends on the relative strength of the interaction between solute molecules and the interaction between solute and water molecules. When the concentration of these substances in water reaches a certain threshold, the order of hydrophobicity strength shifts to BTCN > MA > TAB. This phenomenon profoundly reveals that “hydrophobicity” is not an absolutely unchanging property, but rather a dynamic manifestation of the complex competitive relationships within molecules, between molecules, and between molecules and solvents under different systems and conditions.

These findings offer a valuable foundation for the rational design of hydrophobic materials, which are applicable in areas such as drug delivery systems , and waterproof coatings. , The molecular-level insights into surface energy modulation and contact angle hysteresis allow for precise control over material wettability, a critical factor in various fields. For instance, in targeted drug delivery, adjusting hydrophobicity to match specific physiological environments, such as through pH-responsive coatings, promotes controlled release kinetics and enhances bioavailability. In the field of advanced coatings, , the elucidated structure–property relationships guide the development of next-generation waterproof materials with self-cleaning capabilities. , Future research directions involve exploring bioinspired hierarchical structures and developing computational models to predict material performance under dynamic wetting conditions.

2.6. Limitations and Prospects of Simulation Methods

It is important to note that the B3LYP density functional theory selected for this study inherently involves approximations, which may lead to inaccuracies in the depiction of electron density and charge distribution within the system. These inaccuracies could potentially affect the interpretation of the calculated results. Nonetheless, the validity of our conclusions fundamentally relies on the comparative consistency of results across various systems. As long as consistent methodologies and computational precision are employed, the relative energy rankings and trends observed are typically reliable and appropriate for qualitative, and possibly semiquantitative, analysis.

The outcomes of molecular dynamics (MD) simulations are largely dependent on the parameters of the chosen force field. Conventional force fields, such as AMBER, CHARMM, and OPLS-AA, utilize predefined parameter sets to model atomic interactions, typically derived from fitting to a limited set of experimental or high-level quantum mechanical data. Consequently, they may not fully capture the energetic nuances across all chemical environments. In this study, the OPLS-AA force field was employed, which has been extensively validated for its applicability to organic small molecule systems. Nonetheless, for specific interactionssuch as those involving unconventional protonation statespotential inaccuracies may arise, thereby affecting the quantitative reliability of the results. While force field limitations may impact the accuracy of absolute values, we maintain that comparisons of relative trends among different systems or mutants remain robust and generalizable, as systematic errors tend to partially cancel out. Furthermore, the molecular dynamics simulation, comprising a system of 30,000 atoms and a production time of 10 ns, was deemed adequate to achieve system equilibration and provide sufficient conformational sampling for a quantitative assessment of molecular stacking and intermolecular interactions.

3. Conclusions

Molecular hydrophobicity, dictated by charge distribution, polar functional groups, solvent interactions, and various other factors, displays concentration-dependent behavior in aqueous systems. This study reveals a notable inversion in the hydrophobicity hierarchy for MA, BTCN, and TAB as concentration increases:

• 1.At extremely low concentrations (in a fully dispersed state): MA > TAB > BTCN.
• 2.Beyond a critical concentration threshold: BTCN > MA > TAB.

Hydrophobicity is not an absolutely unchanging property, but rather a dynamic manifestation of the complex competitive relationships within molecules, between molecules, and between molecules and solvents under different systems and conditions. These results offer essential insights for the strategic design of hydrophobic materials that respond to changes in concentration.

4. Methods

4.1. Hydrophobicity Assessment Utilizing the XLOGP3 Method

The hydrophobic parameters [XLOGP3] for MA, BTCN, and TAB were calculated using the XLOGP3 software, developed by the Chinese Academy of Sciences. XLog P is a computational model based on the atomic contribution method, developed to predict the octanol–water partition coefficient (Log P) of chemical compounds. The model employs a refined classification of atomic types, whereby the Log P of a molecule is calculated as the sum of contributions from each atom type, supplemented by structural correction factors that capture specific intramolecular interactionssuch as hydrogen bonding. A higher XLog P value indicates stronger hydrophobicity.

4.2. Molecular Structure Optimization

Geometric optimizations of MA, BTCN, and TAB were initially performed at the B3LYP/6-31G+ level. Subsequently, further optimization was carried out at the B3LYP/DGTZVP level to ascertain their lowest-energy conformations and corresponding dipole moment values. The molecular polarity was evaluated by analyzing the molecular surface electrostatic potential distribution with the aid of the Multiwfn software.

4.3. Molecular Dynamics Simulation

Classical molecular dynamics simulations were conducted to model the behavior of MA, BTCN, and TAB in aqueous solution under normal temperature and pressure conditions. The GROMACS software package was used for the simulation calculations. The molecular dynamics simulations utilized the OPLS-AA , force field with RESP II charges. , Visualization was accomplished through the application of VMD.

Three mixed systems, namely MA-water, BTCN-water, and TAB-water, were constructed by incorporating the molecules of interest into individual periodic simulation boxes filled with water molecules. After energy minimization, the systems were equilibrated under NPT (constant number of particles, pressure, and temperature) and NVT (constant number of particles, volume, and temperature) ensembles. Each system then underwent a 10 ns production molecular dynamics simulation. This protocol resulted in stable configurations at 298 K and 0.1 MPa, ensuring thermodynamic equilibrium under ambient conditions. ,

Hydrophobicity is characterized by several key indicators, including the solvent-accessible surface area (SASA), which quantifies the surface area of a molecule accessible to a solvent; radial distribution functions (RDFs), which describe the spatial distribution of atoms or molecules around a reference point; and hydrogen-bond counts, which reflect the number of hydrogen bonds formed by a molecule with its surroundings.

The authors declare no competing financial interest.