Effects of Gold Nanoparticles on the Antioxidant Power of Gallic Acid: A Computational Investigation Using a Cluster Model

Abstract

Computational approaches within the framework of density functional theory (DFT) are used to probe the effects of gold nanoparticles (AuNPs) on the antioxidant potency of gallic acid (H₄GA), which is a prototypical polyphenolic acid. Four small gold clusters, Au _n with n = 2, 3, 6, and 11, are employed as simple models to simulate the surface of AuNPs. The antioxidant capacity is evaluated through the ability to donate a hydrogen atom and to transfer an electron, which are characterized by the bond dissociation enthalpy (BDE) and ionization energy (IE) of the antioxidant, respectively. The reactions of both H₄GA and its anionic conjugate H₃GA^– toward the hydroperoxyl radical (HOO^•) are examined following the formal hydrogen atom transfer (HAT) mechanism. The assembly of either H₄GA or H₃GA^– on nanostructured Au surfaces can be used as an effective strategy to greatly improve the antioxidant activity, instead of modifying the chemical structure of the antioxidant. Molecular docking computations reveal some key interactions between GA-Au₁₁ clusters and lipoxygenases, highlighting their antioxidant potential. These findings emphasize the importance of specific molecular features in enhancing antioxidant activity and offering insights for the development of more effective antioxidants.

1. Introduction

As a part of normal metabolism, the production of reactive oxygen species (ROS) is a continuous and inevitable process within the human body. An imbalance caused by the overproduction and/or accumulation of such ROS in a body cell in the form of free radicals may result in oxidative stress (OS) and induce negative effects on several molecules of biological importance, such as DNA, proteins, and lipids. The damage caused to these macromolecules gradually increases with age and is likely implicated in several chronic diseases, including cardiovascular disease, atherosclerosis, neurodegeneration, and even cancer. Numerous studies have been conducted with the aim to find novel antioxidants for a more effective treatment of the diseases associated with such pathogens. It has been strongly suggested that the daily intake of different food types that are good radical scavengers is extremely helpful in preventing the cumulative and harmful effects of OS in the human body.

Concerning the radical scavengers present in food, the polyphenols, a group of micronutrients common in a variety of fruits and vegetables, belong to the most important and potent natural antioxidants. , In particular, the 3,4,5-trihydroxybenzoic acid, shown in Figure and more commonly known as gallic acid (H₄GA), is a low-molecular-weight triphenolic compound present in many plants of nutritional, pharmacological, and cosmetic importance. It attracts a special attention owing to its ability to act as a strong scavenger of various free radicals, including the hydroxyl, peroxyl, alkyl peroxyl, singlet oxygen, and other long-lived mutagenic radicals. , Its action as a free radical scavenger has also been reported to be stronger than that of Trolox, a standard compound used for antioxidant assays. In this context, gallic acid has been the subject of intensive studies aimed at elucidating its radical scavenging activities, metal ion chelation, and, more importantly, its ability to maintain endogenous defense systems, inhibit lipid peroxidation, and promote apoptosis in cancer cells. Recently, persistent efforts using theoretical approaches have been devoted to both the kinetic and thermodynamic aspects of its chemical activities toward radicals. ,−

1.

Chemical structure (left) and equilibrium geometry (middle) of H₄GA, along with its aroxyl radical H₃GA^• (right). The C, O, H, and Au atoms are depicted in gray, red, and white, respectively. This atom numbering is applied throughout the following text.

There is no doubt about the strong antioxidant activity of naturally occurring polyphenols such as H₄GA. However, earnest attempts have continuously been made to discover novel derivatives with outstanding radical scavenging activity. To improve the polyphenols’ activity, a common strategy is to modify the structures of these natural products. For example, an alkyl or aryl substitution, a hydroxylation of the phenol ring to form polyols, or a contraction of the hexacyclic residue of the chromanol to a pentacyclic ring have been found to significantly enhance the activities of many phenolic compounds. , Another approach consists of the insertion of a foreign element into phenolic derivatives, giving rise to the flavonoid derivatives. For instance, the ability to trap free radicals of the derivatives such as chrysin and quercetin has been proven to increase substantially following the replacement of the O atom in their chemical structures with the heavier S or Se elements. − Such structural changes could, however, result in some unexpected and negative effects, such as pro-oxidizability and certain toxicity. Indeed, while hydroxylation of the chromanol core brings some positive impacts on the antioxidant activity of Vitamin E, it also incurs a risk of cytotoxic effects via its oxidized product, namely ortho-benzoquinone. Due to the crucial biological safety, it is necessary to identify other approaches that can amplify the antioxidant activity of the parent compound instead of modifying its chemical structures.

Recently, nanostructured materials have been present in most major scientific and industrial areas and thus play a crucial role in many scientific domains, ranging from materials engineering to life sciences. Among the common inorganic nanostructures suitable for medical sciences and biomedical applications, the derivatives containing the gold atoms have attracted much interest owing to their valuable impacts and distinct advantages. Although the gold bulk is inert with respect to chemical reactions, its nanostructural form is well recognized as a biocompatible metal and exhibits much lower toxicity to the human body than many other metals. In ancient times, the upper classes of the society often consumed colloidal gold as a drinkable fluid to treat certain diseases. Nowadays, the presence of gold nanoparticles (AuNPs) in drugs is not only safe as drug carriers but also tends to produce beneficial outcomes, such as an improvement in the therapeutic action of the drug, an increase in the therapeutic retention time in circulation, a more effective drug delivery, and an enhancement of the target specificity of treatments. , Nevertheless, relatively little is known about the action mechanisms of these nanoparticles.

Motivated by both the advances and challenges in the use of antioxidants, we set out to find a novel and more effective strategy for the amplification of the antioxidant activity of phenolic compounds without amending their functional groups. In particular, we conducted a systematic study using quantum chemical calculations to decipher the mechanism involving an antiradical amplification of H₄GA by gold nanoparticles. To chemically represent the gold nanosurfaces, some small gold clusters Au _n are employed as reactant models. Of the theoretical methods applied to probe the adsorption behavior of molecules on nanostructured surfaces, the atomic nanocluster model is of significant interest because it allows the interactions and related properties to be directly simulated through reliable quantum chemical calculations. To date, full geometry optimizations of nanoclusters containing up to several hundred atoms can be carried out, representing the nanoparticle without defining a crystal structure as in bulk approaches. Recent studies − have proven a successful use of small-sized Au _n clusters to model the AuNP surfaces, and such approaches are highly recommended for theoretical studies involving the binding of organic molecules to AuNPs.

The antioxidant capacity of a certain molecule is evaluated via its ability to donate a hydrogen atom or transfer an electron to the free radical. In the present study, in addition to relevant thermochemical data, portions of the potential energy surfaces (PESs) related to the mechanistic pathways of the hydroperoxyl radical (HOO^•) scavenging activity are also examined. It is worth mentioning that this radical is also present in several processes of great interest, including photocatalysis, atmospheric pollutant degradation, and molecular oxygen activation in enzymes. Furthermore, despite the important role of the HOO^• radical in biological operations, information on its reactivity toward phenolic antioxidants remains limited.

2. Computational Methods

All electronic structure calculations in this work are managed using the Gaussian 16 program package. The popular hybrid B3LYP functional is used in conjunction with a mixed basis set, namely the correlation-consistent basis set plus an effective core potential (ECP), leading to the aug-cc-pVTZ-PP basis set for the gold atom and the aug-cc-pVTZ for nonmetal atoms. Although the B3LYP was recently recommended to be replaced by other functionals such as M06-2X, LC-ωPBE, BHandLYP, or ωB97XD··· in predicting thermochemical and kinetic parameters, it should not be banished and may even be restored to its rightful place in DFT calculations. Indeed, according to a recent analysis, the B3LYP functional remains a reliable tool for gaining insights into the geometric structures and free radical scavenging mechanisms attributed to polyphenolic compounds. Moreover, this functional is also well established for its balance between the expected accuracy and the computational demand, making it suitable for probing large chemical systems or conducting a large number of calculations.

Diffuse orbital-augmented basis sets, i.e., aug-cc-pVTZ-PP and aug-cc-pVTZ, are necessary to determine more accurately the energies of the anions considered. True local minima and transition structures are confirmed by the number of imaginary frequencies, i.e., 0 and 1, respectively. Because the unrestricted formalism (UB3LYP) is applied for open-shell species, the inherent spin contamination for systems with an unpaired electron (doublet multiplicity) is checked. We find that the expectation eigenvalue <S²> of total spin-squared operator (Ŝ²) obtained for these species almost equals 0.75; the spin contamination is accordingly negligible.

To include the effect of biological media, simulations are also performed in both water and pentyl ethanoate (PE) environments using the integral equation formalism-polarizable continuum model (IEF-PCM) at the same levels as gas phase calculations. The potential energy profiles illustrating the reactions between H₄GA and the HOO^• radical are constructed from the energy differences between the reactants, reactant complexes, transition structures, product complexes, and products. Geometries of transition states are located using the synchronous transit-guided quasi-Newton (STQN) method and the eigenvector following (EF) method implemented in the Gaussian package. Once this step is completed successfully, the optimized structure is verified to have only one imaginary frequency corresponding to the motion of atoms along the expected reaction route. For a more reliable confirmation of the identity of each transition structure located, intrinsic reaction coordinate (IRC) calculations following the vibrational motions of its imaginary frequency are carried out to determine the two connecting energy minima.

It has been established in the abundant literature that phenolic compounds basically act as antioxidants either by donating a hydrogen atom or transferring an electron to the radicals. , These processes, which can also be classified as hydrogen atom transfer (HAT) and single electron transfer (SET) reactions, respectively, can be described as follows:

$$\mathrm{HAT}:\mathrm{ArOH}{+\mathrm{R}}^{•}\to {\mathrm{ArO}}^{•}+\mathrm{R}-\mathrm{H}$$

$$\mathrm{SET}:\mathrm{ArOH}{+\mathrm{R}}^{•}\to {\mathrm{ArOH}}^{•+}{+\mathrm{R}}^{-}\to {\mathrm{ArO}}^{•}+\mathrm{R}-\mathrm{H}$$

Wright et al. proposed that both mechanisms can typically take place in parallel at different rates and tend to give a prevalence of the former. The second mechanism is also known as a single electron transfer followed by a proton transfer mechanism (SET-PT). In a SET-PT reaction, the first step (electron transfer), which is characterized by ionization energy values, occurs more slowly and is the rate-determining step. Higher antioxidant potency is therefore characterized by a smaller bond dissociation enthalpy (BDE) or a lower ionization energy (IE). These thermochemical parameters are computed using the following equations:

$$\mathrm{BDE}=\mathrm{H}({\mathrm{ArO}}^{•})+\mathrm{H}({\mathrm{H}}^{•})-\mathrm{H}(\mathrm{ArO}\mathrm{H})$$

$$\mathrm{IE}=\mathrm{H}({\mathrm{ArOH}}^{•+})+\mathrm{H}({\mathrm{e}}^{-})-\mathrm{H}(\mathrm{ArO}\mathrm{H})$$

These definitions of ionization energy and binding energy are widely used in studies of antioxidant properties using quantum chemical computations because the H­(H^•) and H­(e) enthalpy values can be taken from experimental measurements. , In this context, it makes more sense to compare the computed results to measured data. For the consistency in reported data, the binding energy is also defined as the enthalpy change of the adsorption process.

The enthalpy values of the hydrogen atom and electron in a vacuum and in an aqueous solution are extracted from previous reports. In pentyl ethanoate, these values are determined at the same level of theory (Table S1). The H₄GA/H₃GA^• affinity for Au _n clusters is evaluated via the binding energy E _b, which is defined as the enthalpy difference between those of Au _n ·H₄GA/Au _n ·H₃GA^• complexes and isolated H₄GA/H₃GA^• species as follows:

$${\mathit{E}}_{\mathrm{b}}={\mathrm{H}}_{{\mathrm{Au}}_n·{\mathrm{H}}_4\mathrm{GA}/{\mathrm{Au}}_n·{\mathrm{H}}_3{\mathrm{GA}}^{•}}-({\mathrm{H}}_{{\mathrm{Au}}_n}+{\mathrm{H}}_{{\mathrm{H}}_4\mathrm{GA}/{\mathrm{H}}_3{\mathrm{GA}}^{•}})$$

where ${\mathrm{H}}_{{\mathrm{Au}}_n·{\mathrm{H}}_4\mathrm{GA}/{\mathrm{Au}}_n·{\mathrm{H}}_3{\mathrm{GA}}^{•}}$ is the enthalpy of ${\mathrm{H}}_{{\mathrm{Au}}_n·{\mathrm{H}}_4\mathrm{GA}/{\mathrm{H}}_3{\mathrm{GA}}^{•}}$ complexes; ${\mathrm{H}}_{{\mathrm{Au}}_n}$ and ${\mathrm{H}}_{{\mathrm{H}}_4\mathrm{GA}/{\mathrm{H}}_3{\mathrm{GA}}^{•}}$ are the enthalpies of Au _n clusters and species, respectively. Thus, a negative E _b value corresponds to an exothermic adsorption and vice versa. A more negative E _b indicates a stronger affinity for the gold cluster. The basis set superposition error (BSSE) correction is not included in the present calculations because it is not significant. For the interaction of Au₂ with H₄GA and H₃GA^• in the gas phase, its contribution to the final binding energy is computed to be ∼2.0 kcal/mol, a value within the expected error margin of current DFT calculations.

The Gibbs (free) energy of the interaction is computed using the following equation:

$$\mathrm{\Delta }{G}^0(298)=\sum {({\epsilon }_0+G_{\mathrm{corr}})}_{\mathrm{products}}-\sum {({\epsilon }_0+G_{\mathrm{corr}})}_{\mathrm{reactants}}$$

where (ε₀ + G _corr) is the sum of electronic and thermal free energies, and the subscript “corr” stands for the thermal correction to the Gibbs free energy.

The GOLD program version 5.3 is also used to conduct molecular docking studies, and the crystal structure with PDB ID: 1N8Q is collected from the Protein Data Bank (https://www.rcsb.org). The genetic algorithm is used for the protein–ligand docking protocol. The grid box with dimensions of 10 Å is selected in the active site of the receptor with the coordinates (X:Y:Z = 20.87:1.63:20.00). While water molecules are removed from the proteins, H atoms are added, and ligands are optimized and docked. The most stable complex with the largest Au₁₁ cluster considered is selected for interaction analysis. The Discovery Studio Visualizer is utilized to visualize and analyze the interactions within the gold–molecule complexes.

3. Results and Discussion

3.1. Structures and Energetics

Gallic acid has four potential acidic protons from the carboxylic and phenolic groups. In this report, its neutral, mono-, di-, tri-, and tetra-anions are denoted as H₄GA, H₃GA^–, H₂GA^2–, HGA^3– and GA^4–, respectively. Although H₄GA has several exchangeable hydroxyl group H– atoms, in the present study, we consider only the first OH abstraction reaction. The radical formation from H₄GA yields three distinct products, i.e., 3-OH, 4-OH, and 7-OH radicals, and their relative stabilities are presented in Figure S1. In agreement with previous studies, the O4–H bond of H₄GA emerges as the most easily breakable one. Indeed, the 4-OH radical is computed to be 6.5–31.5 kcal/mol lower in energy than the 3-OH and 7-OH radicals (Figure S1). The 4-OH radical is more stabilized as it possesses two hydrogen bonds formed by interactions between the O4 atom and hydrogen atoms of the OH groups in the meta positions. At the B3LYP/aug-cc-pVTZ + ZPE level, the BDE of O4–H site (cf. Figure ) is computed to be ∼73.6 kcal/mol (in vacuum), which is much smaller than those of O3–H (80.2 kcal/mol) and O7–H (105.5 kcal/mol) sites.

The B3LYP functional has been calibrated to be reliable for the prediction of the O–H bond dissociation enthalpies (BDEs) of polyphenols such as gallic acid. In addition, it has been suggested that a triple-zeta basis set with augmented functions, such as the aug-cc-pVTZ, needs to be employed for calculations on the BDE values. The present results computed at the B3LYP/aug-cc-pVTZ level can be regarded as a further evaluation on the performance of different DFT approaches. Applying the IEF-PCM solvation model, the BDE of H₄GA in a neutral aqueous solution is found to marginally increase to 74.4 kcal/mol, which remains smaller than the previous B3LYP/6-311++G­(d,p) value of 79 kcal/mol. , Concerning gas phase values given above, a difference of ∼5 kcal/mol due to the basis set effect and ZPE correction is rather large.

We now examine the influence of Au metals on the ability to donate a hydrogen atom of H₄GA. Four small Au _n clusters with n = 2, 3, 6, 11 are used as reactant models to represent the surface of AuNPs. As previously reported, while Au₃ adopt a bent structure, Au₆ prefers an equilateral triangle. At the B3LYP/aug-cc-pVTZ-PP level of calculations employed in this study, the V-shaped structure is identified as the most stable form of Au₃, while the linear conformation is predicted to be a transition structure with a relative energy of 1.5 kcal/mol higher and a negative frequency around 27i cm^–1. For Au₆, we are even unable to locate the tricapped triangle conformation as previously reported because it invariably converges to the triangular form after structural relaxation. On the contrary, the global minimum of Au₁₁ was reported to be a three-dimensional form containing a trigonal prism (Figure ). The main difference in these four clusters is that while Au₂ and Au₆ are characterized by a singlet ground state, both Au₃ and Au₁₁ exhibit an open-shell doublet ground state.

2.

Equilibrium structures of Au _n (n = 3, 6, 11) clusters, along with their group symmetry and electronic states.

The equilibrium structures obtained for the complexes of H₄GA and its radical H₃GA^• with Au _n (n = 2, 3, 6, and 11) clusters are presented in Figure . For interaction with H₄GA molecule, the Au _n clusters tend to anchor on the oxygen atom of the carboxyl group via the positively charged Au atoms. , On the contrary, the radical is inclined to adsorb on gold surfaces through an Au–O4 bond. In addition, the interaction is further stabilized by weakly coupled O–H···O and O–H···Au couplings. The gallic acid H₄GA tends to exhibit a planar structure, and the dihedral angles are thus equal to zero. Following complexation with Au atoms in both monodentate and bidentate modes, the dihedral angles in these species remain almost unchanged.

3.

Equilibrium structures obtained for the complexes of H₄GA and its radical H₃GA^• with the gold clusters Au _n with n = 2, 3, 6, and 11.

The distribution of unpaired spin density in [Au ₂ ·H ₃ GA^• ] and [Au ₃ ·H ₄ GA] complexes is also examined to evaluate the repopulation of spin density upon adsorption. As shown in Figure S2 the spin density of the Au atom binding to radical H₃GA^• in [Au ₂ ·H ₃ GA^• ] is significantly increased. On the contrary, the spin density of the Au atom at the binding site between Au₃ and H₄GA is greatly reduced. Indeed, the spin density populated on the Au atom at the coordination position in the [Au ₃ ·H ₄ GA] complex is computed to be ∼0.01 which is negligible, whereas the corresponding values of 0.31 and 0.67 are obtained for other Au atoms (Figure S2).

The Au–O bond length in the [Au ₂ ·H ₄ GA] complex is computed to be ∼2.21 Å, slightly longer than the sum of covalent radii (2.17 Å) of oxygen (0.73 Å) and gold (1.44 Å) atoms. For H₄GA binding to larger Au _n clusters, the Au–O distances now amount to 2.19, 2.32, and 2.32 Å for n = 3, 6, and 11, respectively. In [Au _n ·H ₃ GA^•], the O atom at the C4 position is also the site most preferred for the binding of H₃GA^• to gold clusters. The Au–O distances in [Au ₃ ·H ₃ GA^•] and [Au ₁₁ ·H ₃ GA^•] are significantly reduced to 2.05 and 2.09 Å, respectively, compared to the values of 2.18 and 2.33 Å in [Au ₂ ·H ₃ GA^•] and [Au ₆ ·H ₃ GA^•]. Thus, interactions between the radical and odd-numbered Au _n systems are likely to be more effective because of radical recombination.

The BDE values of the predicted O–H bonds in the gas-phase environment are provided in Table . According to the HAT mechanism, a lower BDE value indicates a higher antioxidant potency. In a vacuum, the BDE value of free H₄GA is computed to be 73.6 kcal/mol and is substantially decreased to 56.4–64.7 kcal/mol when the molecule is adsorbed on the open-shell Au₃ and Au₁₁ clusters. However, the difference between the BDE values for free H₄GA and its adsorbed form on the closed-shell Au₂ and Au₆ become less significant. Thus, remarkably, the presence of free electrons on the gold surface leads to a stronger effect on the H-donating ability of H₄GA. By comparing the BDE values obtained for the isolated molecule and its conjugated form with Au _n (Table ), we can expect that the antioxidant activity of H₄GA is enhanced more significantly by the odd-numbered Au _n clusters having a doublet state.

1. O–H Bond Dissociation Enthalpy (BDE) and Ionization Energy (IE) for Free Gallic Acid and Its Adsorption on Au _n Clusters .

	BDE	IE
Species	(kcal/mol)	(kcal/mol)	(eV)
H 4 GA	73.6	188.0	8.2
[Au 2 ·H 4 GA]	75.4	182.4	7.9
[Au 3 ·H 4 GA]	64.7	143.4	6.2
[Au 6 ·H 4 GA]	75.8	167.3	7.3
[Au 11 ·H 4 GA]	56.4	128.6	5.6

^aResults are obtained in gas phase at the B3LYP/aug-cc-pVTZ/aug-cc-pVTZ-PP + ZPE level.

The potency to donate an electron of H₄GA is evaluated via its adiabatic IE value. This thermochemical parameter is computed by taking the enthalpy difference between the cation and the neutral atom at their optimized structures. Their computed results are also included in Table . According to the SET mechanism, a lower IE value reflects a stronger electron-transferring ability, and hence is accompanied by a higher antioxidant power. At the B3LYP/aug-cc-pVTZ + ZPE level, the gas-phase IE for H₄GA is ∼188 kcal/mol (8.2 eV), which is equal to the B3LYP/6-311++G­(df,p) value predicted previously. Especially, the presence of odd-numbered Au _n particles is found to greatly enhance the electron-donating ability of H₄GA. The IE values of [Au ₃ ·H ₄ GA] and [Au ₁₁ ·H ₄ GA] are now reduced to 143.4 kcal/mol (6.2 eV) and 128.6 kcal/mol (5.6 eV), respectively, whereas those of [Au _n ·H ₄ GA] with n = 2, 6 amount to 167.3 and 182.4 kcal/mol (7.3 and 7.9 eV).

The strength of interaction between H₄GA/H₃GA^• and gold metals, which is evaluated via the binding energy and changes in Gibbs energy during adsorption, is also an important factor in determining the antiradical potency of H₄GA in a gold colloidal solution. Table presents the numerical results obtained for the complexes considered. Accordingly, the H₄GA molecule is noted to interact with even-numbered systems, namely Au₂ and Au₆, more strongly than the H₃GA^• radical. The E _b values for H₄GA binding to Au₂ and Au₆ are −15.2 and −8.2 kcal/mol, respectively, that are slightly more negative than the corresponding values of −13.3 and −5.9 kcal/mol for H₃GA^•. In contrast, the radical H₃GA^• binding to Au₃ and Au₁₁ induces a significant change in both E _b and △G ²⁹⁸ values. For example, the E _b and △G ²⁹⁸ values of [Au ₃ ·H ₄ GA] are −15.2 and −5.7 kcal/mol, respectively, which are greatly enlarged to −24.2 and −12.9 kcal/mol for [Au ₃ ·H ₃ GA^•]. The higher affinity toward H₃GA^• of odd-numbered systems is expected to facilitate the H-donating ability of H₄GA.

2. Binding Energy (E _b) and Gibbs Energy (△G ²⁹⁸) in kcal/mol for the Adsorption of H₄GA and Its Radical H₃GA^• on Au _n Surfaces (B3LYP/aug-cc-pVTZ/aug-cc-pVTZ-PP + ZPE).

Complex	E b	△G 298	Complex	E b	△G 298
[Au 2 ·H 4 GA]	–15.2	–5.7	[Au 2 ·H 3 GA•]	–13.3	–4.0
[Au 3 ·H 4 GA]	–15.3	–4.3	[Au 3 ·H 3 GA•]	–24.2	–12.9
[Au 6 ·H 4 GA]	–8.2	1.0	[Au 6 ·H 3 GA•]	–5.9	2.6
[Au 11 ·H 4 GA]	–9.0	3.2	[Au 11 ·H 3 GA•]	–24.9	–10.2

3.2. The Solvent Effects

Table presents the thermochemical parameters for the adsorption of H₄GA and H₃GA^• on Au₂ and Au₃ clusters in different solvents. The BDE values of free H₄GA molecule in the gas phase, water, and pentyl ethanoate are predicted to be 73.6, 74.4, and 73.9 kcal/mol, respectively. The results indicate that the change in the bond dissociation enthalpy of gallic acid in the solvents considered is not significant. A similar trend is also observed for H₄GA adsorbed on Au₂. The O–H BDEs of [Au ₂ ·H ₄ GA] in the gas phase, water, and pentyl ethanoate now amount to 75.4, 75.6, and 75.2 kcal/mol, respectively, which are slightly larger than those obtained for the free H₄GA. In addition, we find that the affinity of Au₂ for H₄GA and H₃GA^• is nearly the same in all investigated solvents. For example, the binding energy E _b and the Gibbs energy △G ²⁹⁸ of forming [Au ₂ ·H ₄ GA] product in water are around −11.6 and −3.2 kcal/mol, respectively, as compared to the values of −10.4 and −1.0 kcal/mol obtained for [Au ₂ ·H ₃ GA^•] (Table ).

3. Thermodynamic Data Characterizing the Antioxidant Potency of Gallic Acid, along with the Binding Energy (E _b) and Gibbs Energy (△G ²⁹⁸) for the Adsorption of H₄GA and H₃GA^• on Au₂ and Au₃ Clusters .

	Gas phase	Water	PE	Gas phase	Water	PE
	BDE			IE
H 4 GA	73.6	74.4	73.9	188.0 (8.2)	118.3 (5.1)	137.2 (6.0)
[Au 2 ·H 4 GA]	75.4	75.6	75.2	182.4 (7.9)	111.7 (4.8)	132.0 (5.7)
[Au 3 ·H 4 GA]	64.7	47.4	52.2	143.4 (6.2)	68.4 (3.0)	89.6 (3.9)
	E b			△G 298
[Au 2 ·H 4 GA]	–15.2	–11.6	–12.6	–5.7	–3.2	–3.4
[Au 2 ·H 3 GA•]	–13.3	–10.4	–11.3	–4.0	–1.0	–2.0
[Au 3 ·H 4 GA]	–15.3	–11.6	–13.0	–4.3	–3.0	–0.1
[Au 3 ·H 3 GA•]	–24.2	–38.5	–34.6	–12.9	–28.9	–20.7

^aResults (kcal/mol; IE values in eV are given in parentheses) are collected at the B3LYP/aug-cc-pVTZ/aug-cc-pVTZ-PP + ZPE level. Solvent effects are examined using the IEF-PCM Solvation Model.

The H₃GA^• radical exhibits a much higher affinity toward Au₃ than the H₄GA molecule, especially in an aqueous medium. Indeed, the E _b value of −38.5 kcal/mol for [Au ₃ ·H ₃ GA^•] in water is much more negative than that of −11.6 kcal/mol predicted for [Au ₃ ·H ₄ GA]. These results clearly stipulate a stronger interaction between Au₃ and the radical H₃GA^• and that the resulting [Au ₃ ·H ₃ GA^•] complex is more stable than the [Au ₃ ·H ₄ GA] counterpart. Due to the entropic effects, the Gibbs energies of forming [Au ₃ ·H ₃ GA^•] and [Au ₃ ·H ₄ GA] become smaller, but such a trend is still similar to that of binding energies. In contrast to H₄GA, the interaction of H₃GA^• with Au₂ becomes more breakable in all of the solvents considered. For example, the E _b and △G values for forming [Au ₂ ·H ₃ GA^•] in water are around −10.4 and −1.0 kcal/mol, as compared to the corresponding values of −11.6 and −3.2 kcal/mol for [Au ₂ ·H ₄ GA]. Overall, the presence of free electrons on a gold nanostructured surface is expected to significantly enhance the antiradical potency of H₄GA molecules, especially in highly polar media like water.

We now consider the solvent effects on the ability of electron donation of H₄GA. Calculated IEs of free H₄GA and its complexes with Au _n clusters (n = 2, 3) in different solvents are included in Table . Overall, IE values obtained for both GAH and [Au _n ·H ₄ GA] complexes decrease in the order of gas phase > pentyl ethanoate > water (Table ). We observe a significant influence of a polar solvent such as water on IE values. The difference between IE values for H₄GA in both the gas phase and aqueous solution amounts to ∼70 kcal/mol (3.0 eV), while the difference between the gas phase and pentyl ethanoate is significantly reduced to 51 kcal/mol (2.2 eV). A more negative solvation energy of the electron is likely the major reason for lower IEs in aqueous solution. Computed results listed in Table also reveal that solvents influence the IEs more drastically compared to their effects on BDEs. This can be understood by the fact that cations are charged species, and they, consequently, emerge as much more sensitive to solvent effects than neutral molecules and radicals.

3.3. Interaction of Au _n Clusters with the Anion H₃GA^–

Gallic acid is a water-soluble acid with a first dissociation constant of pK _a1 = 4.2. Its dominant form in biological media (pH = 7.4) is the monoanion H₃GA^–. Accordingly, the H₄GA molecule tends to exist in an anionic form upon deprotonation of either the carboxylic or phenolic group, as depicted in Figure . The existence of such ionic states can induce a significant effect on the antiradical power of gallic acid in aqueous solution.

4.

Phenolate (left) and carboxylate (right) forms located for the deprotonated gallic acid (H₃GA^–). Values in parentheses are their relative energies (kcal/mol) in aqueous solution (IEF-PCM/B3LYP/aug-cc-pVTZ + ZPE).

In this study, phenolate Iso_1 is computed to be more stable than carboxylate Iso_2 by 5.1 kcal/mol (IEF-PCM/B3LYP/aug-cc-pVTZ + ZPE value). Correspondingly, in aqueous solvent, the former is likely to be the dominant isomer, and it will be considered for all subsequent investigations on the interaction of Au _n clusters with the monoanion H₃GA^–. This remarkable finding, contrary to traditional expectations, may be due to the simultaneous formation of two intramolecular hydrogen bonds (O3–H···O4···H–O5) in Iso_1. The dominance of Iso_1 over Iso_2 is also supported by the acidity of gallic acid, which is rather stronger than that of 4-hydroxybenzoic acid. Experimentally, the pK _a1 values of gallic acid and 4-hydroxybenzoic acid amount to 4.2 and 4.6, respectively. Because the hydroxyl (OH) is a well-known electron-donating group, the presence of three OH groups would be expected to reduce the acidity of gallic acid compared to that of 4-hydroxybenzoic acid if the carboxylate form Iso_2 is the prominent form of H₃GA^–. In practice, gallic acid (pK _a1 = 4.2) exhibits a stronger acidity than 4-hydroxybenzoic acid (pK _a1 = 4.6), and thus the formation of the phenolate form Iso_1 can be expected in an aqueous solution at pH = 7.4.

Optimized geometries of the complexes arising from the interaction between some small Au _n clusters and the deprotonated form Iso_1 are illustrated in Figure . The most stable forms of [Au _n ·H ₂ GA^•–] complexes are generated by anchoring the Au metals on the O4 atom of the phenolate anion. In addition, the binding is partially stabilized by weak O···H–O and Au···H–O interactions. The free radicals [Au _n ·H ₂ GA^•–] are obtained by hydrogen cleavage of either the O3–H or the O5–H group. Notably, the radical H₂GA^•– (Figure S3) tends to act as a bidentate ligand, consisting of two anchoring Au–O bonds, when binding to Au _n clusters.

5.

Equilibrium structures for complexes of the anion H₃GA^– and its radical H₂GA^•– with Au_n (n = 2, 3, 6, 11) clusters.

Computed BDE and IE values for the anion H₃GA^– in a water solution in the absence and presence of Au _n particles are presented in Table . The lowest BDE of H₃GA^– corresponding to the O3/O5–H bond amounts to 73.5 kcal/mol, which is slightly smaller than the corresponding value of 74.4 kcal/mol obtained for neutral H₄GA in water. The IE of H₃GA^–, which corresponds to the difference between the energies of the anion H₃GA^– and the radical H₃GA^•, is greatly reduced to 84.0 kcal/mol (3.6 eV), as compared to 118.3 kcal/mol (5.1 eV) for H₄GA, indicating a much higher electron-transferring potency of the anion. Reversely, this value corresponds to the electron affinity of the radical.

4. Thermodynamic Data (kcal/mol; IE Values in eV are Given in Parentheses) Obtained for the Free Anion H₃GA^– and Its Complexes with Au _n (n = 3, 6) Clusters .

Species	BDE	IE
H 3 GA –	73.5	84.0 (3.6)
[Au 2 ·H 3 GA – ]	75.1	93.2 (4.0)
[Au 3 ·H 3 GA – ]	60.6	63.8 (2.8)
[Au 6 ·H 3 GA – ]	72.3	92.1 (4.0)
[Au 11 ·H 3 GA – ]	55.8	57.8 (2.5)

Species	E b	△G 298
[Au 2 ·H 3 GA – ]	–19.6	–10.9
[Au 2 ·H 2 GA•–]	–18.0	–9.0
[Au 3 ·H 3 GA – ]	–18.3	–11.2
[Au 3 ·H 2 GA•–]	–31.2	–19.7
[Au 6 ·H 3 GA – ]	–11.5	–2.3
[Au 6 ·H 2 GA•–]	–12.7	–2.8
[Au 11 ·H 3 GA – ]	–11.3	0.8
[Au 11 ·H 2 GA•–]	–27.8	–13.3

^aCalculations in the aqueous solution are performed at the IEF-PCM/B3LYP/aug-cc-pVTZ/aug-cc-pVTZ-PP + ZPE level.

As in the case of H₄GA, both hydrogen- and electron-donating abilities of H₃GA^– are much improved when combined with the odd-numbered Au _n system. Indeed, the BDE value for [Au ₃ ·H ₃ GA ^– ] amounts to 60.6 kcal/mol, which is much smaller than the 73.5 kcal/mol observed for free H₃GA^–. The conjugation with Au₃ in addition markedly enhances the electron-donating ability of H₃GA^–. The IE values of H₃GA^– binding to Au₃ and Au₁₁ are only 63.8 and 57.8 kcal/mol (2.8 and 2.5 eV), respectively, compared to a value of 84.0 kcal/mol (3.6 eV) obtained for free H₃GA^–. Overall, the anion H₃GA^– is characterized as having a much stronger antiradical potency when bound to the odd-numbered Au₃ cluster. Then, the assembly of H₄GA on nanostructured Au surfaces emerges as an efficient strategy to improve its antioxidant activity.

Also, Table reveals a particularly strong affinity of the radical anion H₂GA^•– toward Au₃ and Au₁₁ as compared to either Au₂ or Au₆. Indeed, the binding energy of −31.2 kcal/mol for the formation of [Au ₃ ·H ₂ GA^•–] is more negative than the corresponding values of −18.0 and −12.7 kcal/mol computed for [Au ₂ ·H ₂ GA^•–] and [Au ₆ ·H ₂ GA^•–], respectively. The difference between E _b values for the radical H₂GA^•– and the anion H₃GA^– when binding to Au₃ is ∼13 kcal/mol. However, such a difference when these species bind to even-numbered Au _n systems (n = 2, 6) turns out to be negligible. For example, the E _b values of −19.6 and −11.5 kcal/mol for [Au ₂ ·H ₃ GA ^– ] and [Au ₆ ·H ₃ GA ^– ], respectively, are comparable to −18.0 and −12.7 kcal/mol for [Au ₂ ·H ₂ GA^•–] and [Au ₆ ·H ₂ GA^•–]. The exceptionally strong affinity for H₂GA^•– of the odd-numbered gold cluster is expected to greatly enhance the H-donating potency of H₃GA^–.

3.4. The HOO^• Scavenging

In agreement with previous investigations on the antioxidant properties of phenolic compounds, the HAT mechanism appears to be a dominant process in both vacuum and solution. While the HAT reactions can be characterized by the O–H bond dissociation enthalpy (BDE), the SET reactions are typically evaluated via the corresponding ionization energy (IE). As shown in Tables , , and , BDE values obtained for either H₄GA/H₃GA^– species or their complexes with Au _n clusters are systematically smaller than the corresponding IE values. Therefore, it can be deduced that the antioxidant properties of gold-supported phenolic compounds are also dominated by the HAT mechanism. In this context, some thermodynamic and kinetic aspects related to the antiradical action of gallic acid are considered via the following reactions:

$${\mathrm{H}}_4\mathrm{GA}{+\mathrm{HOO}}^{•}\to {\mathrm{H}}_3{\mathrm{GA}}^{•}+\mathrm{HOOH}$$

$${\mathrm{H}}_3{\mathrm{GA}}^{-}{+\mathrm{HOO}}^{•}\to {\mathrm{H}}_2{\mathrm{GA}}^{•-}+\mathrm{HOOH}$$

Calculations are performed for both isolated H₄GA/H₃GA^– species and their conjugated forms with Au _n particles as well. The enthalpy (△H _r) and Gibbs energy (△G _r) of these reactions are computed and presented in Table . Overall, these processes are highly exergonic, and thus, both H₄GA and H₃GA^– are willing to scavenge the radical HOO^• via the HAT pathway. In the gas phase, the Gibbs energy of the reaction between HOO^• and H₄GA is computed to be −11.0 kcal/mol. When binding to Au₂ and Au₆, the reaction is also exergonic, but the △G _r values are now slightly decreased to −9.3 kcal/mol (Table ). A similar tendency is also observed for the reaction between H₃GA^– and HOO^• in the presence of Au₂ and Au₆ clusters. It is likely that the HOO^• scavenging ability of the H₄GA/H₃GA^– species is not significantly affected by even-numbered Au _n clusters.

5. Enthalpies (△H _r) and Gibbs Energies (△G _r) in kcal/mol of Reactions between H₄GA/H₃GA^– and the Radical HOO^• in the HAT Mechanism .

Reactions	△H r	△G r	Reactions	△H r	△G r
H4GA + HOO•	–10.9	–11.0	H3GA– + HOO•	–10.9	–11.6
[Au 2 ·H 4 GA] + HOO•	–9.0	–9.3	[Au 2 ·H 3 GA – ] + HOO•	–9.3	–9.7
[Au 3 ·H 4 GA] + HOO•	–19.7	–19.6	[Au 3 ·H 3 GA – ] + HOO•	–23.8	–20.1
[Au 6 ·H 4 GA] + HOO•	–8.6	–9.5	[Au 6 ·H 3 GA – ] + HOO•	–12.1	–11.1
[Au 11 ·H 4 GA] + HOO•	–26.6	–24.2	[Au 11 ·H 3 GA – ] + HOO•	–28.6	–26.7

^aFor the anionic state, calculations are performed in an aqueous solvent (B3LYP with both aug-cc-pVTZ and aug-cc-pVTZ-PP basis sets).

However, a completely different landscape emerges when H₄GA/H₃GA^– binding occurs with odd-numbered Au _n systems. The △H _r and △G _r values of both reactions mentioned above become significantly more negative due to conjugation with Au₃ and Au₁₁. When binding to these odd-numbered Au _n particles, the △G _r values for the reaction between H₃GA^– and HOO^• vary from −20.1 to −26.7 kcal/mol, as compared to −11.6 kcal/mol for the free H₃GA^– anion (Table ). The more negative the Gibbs energy, the more exothermic the reaction. The results clearly indicate that the antioxidant ability of gallic acid molecules in aqueous solution, which reflects the scavenging potency of the anion H₃GA^– toward HOO^•, could greatly be improved by dissolving them in a colloidal gold solution.

As mentioned above, because the radicals are highly reactive, their potential reactions with the Au cluster are further investigated. In aqueous solution, the Gibbs energy change for the formation of the [Au₃·HOO^•] complex is −33.7 kcal/mol, which is more negative than the corresponding value of −28.9 kcal/mol for [Au₃·H₃GA^•]. Similarly, the HOO^• binding to [Au₃·H₃GA^•] forming the complex [Au₃(H₃GA^•)­(HOO^•)] (Figure S4) is predicted to be spontaneous with a Gibbs energy of −4.4 kcal/mol. Overall, the Au cluster is highly reactive toward the HOO^• radical and could also be a good source of HOO^• scavenging. However, the reaction between HOO^• and H₄GA adsorbed on the gold surface remains more energetically preferred, as it is further driven by the proton transfer from the latter to the former with a Gibbs energy change of −19.6 kcal/mol (Table ).

We now consider the potential energy profiles describing the reactions between H₄GA and HOO^•. The effects of Au₂ on the energy profile of the reactions are also examined. Equilibrium structures of the reactant complexes (RCs), transition structures for H abstraction (TSs), and product complexes (PCs) are located and presented in Figure . IRC calculations are carried out at the same level to verify that every transition structure is consistent with its given reactants and products. Generally, the HAT reaction between phenolic compounds and peroxyl radicals like HOO^• undergoes multiple pathways with four main steps. The first step is the formation of a complex from reactants, followed by a transition structure and an intermediate. Then, the product is obtained from the intermediate dissociation without any transition structure. The overall reaction pathway can be briefly depicted as follows:

$$\mathrm{Reactants}\to \mathrm{RC}\to \mathrm{TS}\to \mathrm{PC}\to \mathrm{Products}$$

6.

Gas-phase optimized structures of reactant complexes (RCs), transition structures (TSs), and product complexes (PCs) for the HAT reactions of H₄GA (above) and [Au₂·H₄GA] (below) with HOO^•.

As shown in Figure , both RCs and PCs are stabilized by two hydrogen bonds, either between H₄GA and HOO^• or between H₃GA^• and HOOH. The distances of the O3···HOO and the O4H···OOH are varied in the range of 1.74–1.80 Å for [H₄GA···HOO^•], as compared to 1.77–1.78 Å for [Au₂·H₄GA···HOO^•]. Such values are much shorter than the sum of van der Waals radii (2.72 Å) of oxygen (1.52 Å) and hydrogen (1.20 Å) atoms. In the product complex [H₃GA^•···HOOH], the lengths of H-bonds are around 1.77–1.80 Å, which are slightly shorter than the value of 1.81 Å for [Au₂·H₃GA^•···HOOH].

Only one hydrogen bond in the transition structure (TS) exists. In the TS geometry, the O4H···OOH distances are 1.36 and 1.32 Å for [H₄GA···HOO^•] and [Au₂·H₄GA···HOO^•], respectively. The bond lengths of O4–H are now lengthened to 1.08–1.09 Å, as compared to a value of 0.97 Å in free H₄GA. In particular, the transition structures are characterized by large imaginary vibrational frequencies of 859i cm^–1 for [H₄GA···HOO^•] and 1138i cm^–1 for [Au₂·H₄GA···HOO^•].

The gas-phase Gibbs energy profiles describing the reactions between H₄GA and HOO^• with both in the absence and presence of Au₂ are displayed in Figure , while their enthalpy changes are given in Figure S5. As expected, the enthalpies of RCs are always lower than those of the separated starting points, but in terms of Gibbs energies, the former lies above the latter due to entropy corrections. The relative Gibbs energies of reactant complexes and reactants are about 3.9–4.0 kcal/mol. The relative Gibbs energy of TS, which can also be considered as the Gibbs activation energy, is around 10.4 kcal/mol for [H₄GA···HOO^•] and 10.8 kcal/mol for [Au₂·H₄GA···HOO^•]. The reaction rate, as a consequence, is not significantly affected by the presence of Au₂. However, a comparison of the energies of product complexes reveals that due to the Au₂ attendance, the reaction becomes less thermodynamically favorable. The formation of [Au₂·H₃GA^•···HOOH] is associated with a Gibbs energy change of −5.2 kcal/mol, as compared to a slightly more negative value of −7.4 kcal/mol for [H₃GA^•···HOOH].

7.

Gibbs energy profiles illustrating the gas phase reaction H₄GA + HOO^• in both the absence and presence of Au₂. Complexes RCs and PCs lie above the separated reactants and products, respectively, due to the large entropy effects (energies obtained using the B3LYP functional with both aug-cc-pVTZ and aug-cc-pVTZ-PP basis sets).

The desired intermediates, i.e., RCs and PCs, and transition structures for the reaction of H₃GA^– with HOO^• in aqueous solution are presented in Figure . At the equilibrium point of the product complex [H₃GA^–···HOO^•], the hydrogen atom of HOO^• is partially dissociated and lies closer to the oxygen atom of H₃GA^–. As a result, it appears that one O₂ moiety starts to emerge in [H₃GA^–···HOO^•]. In the optimized structures of TSs, two H-bonds are formed between H₃GA^– and HOO^•. Accordingly, the O3H···OOH and O4···HOO distances are 1.38 and 1.58 Å, respectively, for [H₃GA^–···HOO^•], which shift to 1.28 and 1.65 Å for [Au₂·H₃GA^–···HOO^•]. The length of the O3–H bond increases from 0.97 Å in isolated anion H₃GA^– to 1.07 Å in [H₃GA^–···HOO^•] and 1.27 Å in [Au₂·H₃GA^–···HOO^•]. They are found to have large imaginary frequencies at 1558i cm^–1 for [H₃GA^–···HOO^•] and 2013i cm^–1 for [Au₂·H₃GA^–···HOO^•]. Optimization of these TSs by following the eigenvector of their imaginary frequencies gives rise to product complexes. At the equilibrium point of PCs, two hydrogen bonds with distances varying in the range of 1.68–1.72 Å between the atoms of O3 and O4 of the radical H₂GA^•– and HOOH are located (Figure ).

8.

Optimized structures of reactant complexes (RCs), transition structures (TSs), and product complexes (PCs) for the HAT reactions of H₃GA^– (upper panel) and [Au₂·H₃GA^–] (lower panel) with HOO^•.

In both the RC and TS structures of [Au₂·H₃GA^–···HOO^•], the gold metal tends to anchor on the O4 site of the anion Ha^–GA^–. When the reaction is terminated, one hydrogen atom of H₃GA^– is transferred to HOO^• forming the radical H₂GA^•– and the molecule HOOH. In the [Au₂· H₂GA^•–···HOOH] complex product, the Au₂ moiety binds to H₂GA^•– via the O3 position of the radical. These processes are reversible, as both adsorption and desorption processes may occur in parallel at different rates. As a result, changes take place at binding sites due to geometrical rearrangement and charge density.

When compared with H₄GA, the anion H₃GA^– follows a different energy landscape when it reacts with the radical HOO^•. The RC and PC are now more stable than the reactants by 6.2 and 11.3 kcal/mol, respectively. The reaction passes through a small energy barrier of only 2.6 kcal/mol, which is much smaller than the corresponding value of 10.4 kcal/mol predicted for H₄GA and HOO^•. Therefore, one can expect that the anion H₃GA^– exhibits more effective antioxidant activity than the neutral H₄GA. When binding to Au₂, while the RC lies 0.8 kcal/mol above the reactants, the PC now becomes more energetically favorable with a relative energy of −6.4 kcal/mol (Figure ). The formation of products [Au₂·H₂GA^•–] and HOOH is less exothermic with a Gibbs energy of −9.7 kcal/mol, as compared to a value of −11.6 kcal/mol for forming H₂GA^•– and HOOH. Numerous attempts have also been devoted to studying the potential energy profiles of the reactions between H₄GA and HOO^• species with the open-shell Au₃ cluster. However, we were not able to locate the transition structures (TS) for these reactions. Previous attempts to determine the TSs for several radical scavenging processes of polyphenolic acids were also unsuccessful; all this suggests barrierless reaction processes. −

9.

Gibbs energy profiles for the reaction H₃GA^– + HOO^• in both the absence and presence of Au₂. Calculated results are obtained in aqueous solution using the IEF-PCM model.

3.5. Molecular Docking Model of Gallic Acid-AuNPs

As stated above, polyunsaturated fatty acid (PUFA) metabolites significantly influence inflammatory diseases and cancer progression related to oxidative stress. Inhibiting the enzymes responsible for PUFA metabolism, such as dioxygenases (cyclooxygenases and lipoxygenases), can offer a strategy for managing these conditions if their molecular mechanisms are fully understood. Previous studies have demonstrated that the use of molecular docking facilitates a visualization of interaction mechanisms between active compounds and their target sites. − Therefore, this technique allows us to elucidate the binding interactions of the compounds with specific molecular targets for the optimization of therapeutic interventions. Chemical systems with strong antioxidant and free radical scavenging abilities including polyphenols may effectively block PUFA activities. , The H₄GA and Au₁₁·H₄GA complexes are expected to exhibit such properties. In particular, structural analysis suggests that the Au₁₁·H₄GA complexes effectively bind to the active site of lipoxygenases (ID PDB entry 1N8Q). Key residues in the lipoxygenase active site (Gln506, His510, Trp511, His515, Val558, Ile564, His701, Asn705, Leu765, Ile849) are essential for creating interactions between ligands and the protein.

Previous studies have demonstrated that these residues are involved in interactions with the Au₁₁·H₄GA complexes, as illustrated in Figure , indicating their potential to block PUFA metabolism at the molecular level.

10.

Chemical structures and models of the molecular docking study of Au₁₁·H₄GA complexes in lipoxygenases (ID PDB: 1N8Q). Ligand and enzyme structures are shown in stick-shape (above) and ribbon-shape (below). Au₁₁·H₄GA complexes are colored green and yellow, respectively. Key amino acid is shown as a cyan model.

Specifically, the carboxylic group of H₄GA shows H-bond interactions with both His510 and Ile849, highlighting its strong binding affinity in the active site of lipoxygenases. Moreover, the C₃–OH group displays an H-bond with Gln506, further stabilizing its interaction with the enzyme. These hydrogen bonds are essential in enhancing Au₁₁·H₄GA complexes’ potential inhibitory effect. Furthermore, Pi–alkyl interactions and Pi-Pi T-shaped stacking between the benzene ring of H₄GA and key residues such as Ala553, Leu557, and His515 were also observed. These interactions are particularly significant as they stabilize the complex and contribute to the overall binding energy, ensuring a more robust engagement of the Au₁₁·H₄GA complexes within the lipoxygenase active site. In summary, this model provides us with a closer look at the molecular interactions between small gold clusters and gallic acid when they enter the body and demonstrates their effectiveness at specific molecular targets. This visualization not only points out insights into the mechanisms of action but also aids the optimization and improvement of treatment methods based on the gold cluster model.

4. Concluding Remarks

In this study, the effects of gold nanoparticles (AuNPs) on the antioxidant power of gallic acid (H₄GA), in both neutral and anionic states, are thoroughly analyzed by means of DFT calculations. Four small Au _n clusters are employed as simple models to simulate the surface of the AuNPs. The investigations are carried out in both the gas phase and solution using the IEF-PCM approach.

The O–H bond dissociation enthalpy in a vacuum of free H₄GA is larger by ca. 9–17 kcal/mol in comparison to its adsorbed forms with the open-shell Au₃ and Au₁₁ clusters. Similarly, the electron-donating ability of H₄GA is also significantly enhanced when the molecule binds to these particles. The gas-phase IE values of H₄GA binding to Au₃ and Au₁₁ clusters are in the range of 143–129 kcal/mol (6.2–5.6 eV), compared to 188 kcal/mol (8.2 eV) in the free state. The radical H₃GA^•, which results from the central O–H bond dissociation, is found to exhibit a particularly high affinity with the odd-number clusters, especially in an aqueous medium. The stronger interaction between this radical and Au metals is an important factor that contributes to an improvement in the H-donating potency of the antioxidant. We further observe that the electron-donating ability is affected more drastically by a polar solvent such as water than the hydrogen-donating potency.

Overall, computed results for BDE and IE values clearly show that the antioxidant activity is enhanced more effectively by odd-numbered Au _n systems that act as radicals. Under biological conditions, the H₄GA molecule tends to emerge as the phenolate form by proton cleavage of the central phenolic group. The difference between the BDE values for H₄GA and its phenolate anion, resulting from further O–H bond dissociation, is found to be negligible. However, the latter exhibits a much higher electron-transferring potency than the former. Similar to H₄GA, the antioxidant power of the conjugated anion H₃GA^– is considerably enhanced when combined with the odd-numbered Au _n radical system.

The HOO^• scavenging of both H₄GA and H₃GA^– via the HAT mechanism is also examined in terms of both thermodynamic and kinetic aspects. Overall, the reactions are highly exergonic, and thus such species are willing to scavenge the radical HOO^• via the HAT pathway. In addition, computed results show that the HOO^• scavenging of H₄GA/H₃GA^– becomes much more efficient with the presence of odd-numbered Au _n particles. However, both thermodynamic and kinetic analyses confirm a lower HOO^• scavenging ability of H₄GA/H₃GA^– with the presence of even-numbered Au _n counterparts, as the reactions release smaller reaction energies and must overcome higher energy barriers.

The molecular docking reveals that the Au₁₁·H₄GA complexes contain multiple key interactions with lipoxygenases, suggesting that this complex has significant antioxidant potential. The Au₁₁·H₄GA complexes can be regarded as an effective candidate for further exploration in the treatment of diseases associated with oxidative stress, such as cancer or inflammatory disorders. The strong antioxidant and free radical scavenging properties of H₄GA, in combination with the catalytic activity of the AuNPs, offer an interesting avenue for the pharmaceutical development of therapeutic strategies aimed at mitigating oxidative damage and controlling disease progression.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c03489.

• Figures show the optimized structures and potential energy profiles and Tables list the shapes and Cartesian coordinates of the lowest-lying structures located for all systems considered (PDF)

The authors declare no competing financial interest.