Eugenol-Based Epoxy Vitrimers: Caffeine and Zinc Acetate as Potential Alternative Catalysts in Curing Kinetics and Dynamic Network Properties

Abstract

The development of sustainable vitrimers from bio-based sources addresses the need for high-performance recyclable materials. This research describes eugenol-derived epoxy vitrimers cross-linked with adipic acid as a curing agent, focusing on comparative effects of caffeine and zinc acetate as transesterification catalysts at 5 and 10% concentrations versus a non-catalyzed control. Both catalysts acted as curing accelerators, confirmed by FTIR and DSC analyses, revealing polyhydroxyester network formation through associative ester exchange enabling topological reorganization. Zinc acetate at 10% proved most efficient, achieving the lowest apparent activation energy (116.0 kJ/mol), highest crosslinking density (ν_e = 3.42 × 10⁻³ mol/cm³), improved thermal stability with unimodal degradation profile, and substantially reduced topology freezing transition temperature (T_v = 132 °C), confirming enhanced dynamic properties. Caffeine demonstrated catalytic activity, reducing apparent activation energy to 124.4 kJ/mol at 10% and promoting rapid epoxide conversion during initial curing at moderate temperatures. Although its catalytic efficiency is moderate compared to zinc acetate, its bio-based origin and non-toxic nature make it a promising green alternative for sustainable vitrimer applications. Results demonstrate that catalyst selection is crucial for tailoring curing kinetics, network structure, and final vitrimeric properties, providing key guidelines for designing advanced circular materials from bio-based precursors.

1. Introduction

The development of sustainable, high-performance materials has become increasingly critical in the face of global environmental challenges and the demand for innovative solutions across various industries [1,2,3]. Among these materials, epoxy resins have garnered considerable interest due to their beneficial characteristics, including superior structural integrity, resistance to chemical attack, and stable thermal behavior [4,5]. However, traditional epoxy formulations are constrained by their inability to be recycled and their vulnerability to heat-induced breakdown, potentially limiting their suitability for extended-use applications [6,7,8,9]. To address these challenges, vitrimer technology has emerged as a groundbreaking material class that merges the durability of cross-linked polymers with the reformability of meltable plastics. This innovation, pioneered by Leibler et al. [10] in 2011, relies on integrating reversible covalent linkages, particularly those formed through transesterification processes in epoxy-carboxylic acid networks [11,12]. These bonds enable the topological reorganization of the network in response to stimuli such as thermal changes, facilitating the exchange of polymer chains and allowing the material to flow, which in turn enables remodeling, self-repair, and recyclability without loss of effective cross-linking, aligning these materials with the principles of the circular economy [13,14,15].

The synthesis of epoxy vitrimers from renewable resources represents a booming line of research aimed at developing sustainable materials with reprocessing capabilities [16,17,18,19]. Among the most promising precursors is eugenol, a naturally occurring phenolic compound extracted from clove oil [20,21,22,23], whose suitability as a monomer is based on its high availability, low toxicity, and versatile chemical structure, which facilitates its functionalization into glycidyl ethers [24,25]. This structural versatility has enabled the development of eugenol vitrimers reported in the literature, where, after epoxidation and subsequent curing with zinc-catalyzed succinic anhydride, dynamic networks with self-repairing, shape memory, and physical and chemical recycling capabilities are obtained [26]. Likewise, variants with renewable carbon contents exceeding 75% have been achieved, balancing robust thermomechanical performance with efficient reprocessability [27]. Taken together, these findings establish eugenol as an essential building block for designing advanced and sustainable epoxy materials.

The curing reaction between epoxy groups and carboxylic acids generates a polymer network through the formation of β-hydroxy esters, which contrasts with the more traditional curing with amines or anhydrides [28,29,30,31,32,33]. The presence of a catalyst is critical not only for achieving efficient curing that imparts high T_g and thermal stability but also for activating the network dynamics characteristic of vitrimer materials [34,35]. In this context, the selection and dosing of the catalyst emerge as a fundamental challenge for optimizing the final properties. This study contrasts two catalysts of distinct nature: zinc acetate, a Lewis acid that operates through metal coordination mechanisms [36,37,38], and caffeine, a mild organic base capable of deprotonating the carboxylic acid to facilitate the opening of the epoxy ring, representing an innovative and sustainable alternative in the synthesis of dynamic materials due to its advantages as a non-toxic, naturally derived, low-cost compound with documented environmental biodegradability [39,40,41].

Although each has shown catalytic effectiveness, there are still knowledge deficiencies regarding how their molecular characteristics and dosage levels similarly influence the polymerization rate, cross-linking architecture, and final vitrimer characteristics, particularly with respect to caffeine. Therefore, the present work aimed to synthesize and characterize epoxy vitrimer materials derived from eugenol, cured with adipic acid, to evaluate the effect of zinc acetate and caffeine used as catalysts at two concentrations (5 and 10 mol% of COOH groups) in promoting dynamic transesterification reactions (DTERs), compared to a control system without catalysts. Thus, the study integrates the principles of vitrimer chemistry and the use of renewable raw materials in a unified polymeric system, contributing to laying the foundations for the development of next-generation functional and sustainable vitrimeric materials.

2. Results

2.1. Synthesis and Characterization of Eugenol Epoxy Monomer (Eu-Epx)

The first step involved a Williamson ether reaction (Scheme 1a) between dibromoethane and eugenol, yielding a diene intermediate (Eu-Di) with two allyl groups. In the FTIR spectrum of Eu-Di (Figure 1a), the absorption band corresponding to the O-H stretching vibrations of the hydroxyl groups of eugenol, located at approximately 3522 cm⁻¹ [42,43], disappeared completely. In contrast, the C=C stretching vibration at 1638 cm⁻¹ [44] remained unchanged, confirming the successful synthesis of the eugenol-derived diene, specifically through the modification of the phenolic hydroxyl group. In addition, the spectrum showed bands at 1472 cm⁻¹ attributed to CH₂ bending vibrations [45,46], bands in the regions of 1220 cm⁻¹ and 1235 cm⁻¹ associated with C-O stretching [47], and a band at 965 cm⁻¹ corresponding to the stretching of the C-O-C bond of the ester group [48,49]. ¹H and ¹³C NMR spectroscopy further confirmed the structure of Eu-Di. In the ¹H NMR spectrum (Figure 1b), the characteristic peak of the phenolic hydroxyl (Ar–OH) of eugenol, located between 5.65 and 5.63 ppm [50], disappeared and was replaced by a new peak between 4.39 and 4.37 ppm, which corresponded to the protons of OCH₂CH₂O [51]. Likewise, the characteristic signals of the allyl groups were detected between 5.03–5.12 ppm (CH₂=CH–CH₂) [52], 5.90–6.06 ppm (CH₂=CH–CH₂) [53], and 3.36–3.31 ppm (CH₂=CH–CH₂–) [53]. The structure was also verified by ¹³C NMR spectroscopy (Figure S2). These results demonstrate the successful synthesis of Eu-Di.

In the second stage, Eu-Epx was prepared by epoxidation of Eu-Di (Scheme 1b). The reaction was carried out at room temperature in a two-phase ethyl acetate/water system, using Oxone^® (potassium peroxymonosulfate) as the epoxidizing agent. This method was chosen due to its economic and ecological advantages over conventional epoxidation with m-chloroperbenzoic acid, as Oxone^® simplifies the separation and purification processes and utilizes non-toxic organic compounds [42,43]. The successful formation of epoxy groups was verified through multiple analytical techniques. FTIR spectroscopy (Figure 1a) provided initial confirmation by showing the elimination of the C=C stretching vibration at 1642 cm⁻¹ [44] and the out-of-plane =C-H bending vibration at 960 cm⁻¹ [54], while simultaneously revealing a new characteristic symmetric C-O-C stretching vibration of the epoxide ring at 842 cm⁻¹ [55]. Analysis by ¹H NMR (Figure 1c) showed the disappearance of the peaks corresponding to the diene in the ranges of 5.03–5.12 ppm, 5.90–6.02 ppm, and 3.31–3.36 ppm [52,53].

Instead, new peaks attributable to the epoxy group appeared between 2.74–2.83 ppm (–CH–CH₂–O), 2.50–2.56 ppm (Ar–CH₂– and –CH–CH₂–O) and 3.08–3.17 ppm (–CH–CH₂–O) [56]. The structure of Eu-Epx was also confirmed by ¹³C NMR spectroscopy (Figure S3). Finally, the epoxide equivalent weight (EEW) of Eu-Epx was determined to be 192.10 g/equiv, which is close to the theoretical value (193.21 g/equiv). The slight deviation observed could be related to a secondary epoxide ring-opening reaction, which could have led to the formation of hydroxyl groups [44,46]. Furthermore, DSC analysis revealed a sharp melting point at 130 °C for the Eu-Epx monomer (Figure S5).

2.2. Study of the Curing Process

Figure 2 shows the FTIR-ATR spectra of the eugenol-based epoxy system (Eu-Epx/AA) with 5% wt of caffeine, 5% wt of Zn(OAc)₂ or no catalyst, at three stages of the curing process: (i) uncured (initial mixture before heat treatment), (ii) cured (after heat treatment at 150 °C for 60 min), and (iii) post-cured (after additional treatment at 190 °C for 120 min). All spectra were normalized with respect to the aromatic ring vibration band at 1580 cm⁻¹, allowing a comparative evaluation of the evolution of the functional groups.

The broad band at ~3460 cm⁻¹, corresponding to the O–H stretching vibration [42], progressively decreases during heat treatment in all formulations. This decrease reflects the consumption of hydroxyl groups from carboxylic acids during the epoxide opening reaction. The reduction is more pronounced in the presence of catalysts, particularly with Zn(OAc)₂, suggesting greater conversion efficiency. In contrast, the formulation without catalyst shows a less pronounced reduction, especially between curing and post-curing, indicating incomplete conversion.

The C=O stretching band at 1722 cm⁻¹ [57] increases in intensity due to the formation of ester bonds. This increase is more pronounced and progressive in the formulation with Zn(OAc)₂, while with caffeine it stabilizes after curing. In the sample without catalyst, the intensity remains similar between the uncured and cured mixtures, increasing significantly only during post-curing, indicating a slower and less efficient curing process.

The bands associated with C–O vibrations in the regions of 1256, 1138, and 1028 cm⁻¹ [58] show different behaviors depending on the formulation. The band at 1138 cm⁻¹ (steric C–O–C) intensifies more clearly in formulations with catalysts, reflecting greater ester bond formation. The band at 1028 cm⁻¹ increases substantially during post-curing in all formulations, indicating densification of the network in the final stage of the thermal process.

The characteristic band of the epoxy ring at 946 cm⁻¹ (C–O–C stretching) [59] shows different behaviors depending on the catalyst. In the formulation with caffeine, this band decreases markedly from the uncured mixture to the cured mixture, remaining constant during post-curing, indicating rapid and complete conversion of the epoxy group in the curing stage. In the Zn(OAc)₂ formulation, the band maintains similar intensity at all stages, suggesting that this catalyst does not significantly affect the conversion of the epoxy group under FTIR analysis conditions. In the sample without catalyst, the decrease is more gradual and continuous throughout the three stages, with residual signals persisting even after darkening, evidencing incomplete conversion.

Comparing the curing (150 °C/60 min) and post-curing (190 °C/120 min) stages, it can be observed that in the formulation with caffeine, most of the structural transformations occur during curing, while post-curing generates only minor adjustments, indicating early and efficient conversion. The Zn(OAc)₂ formulation shows continuous spectral changes between both stages, particularly in the ester bands (C=O, 1722 cm⁻¹ and C–O, 1138 cm⁻¹) [57,58], suggesting a more progressive curing process [60]. In contrast, in the sample without catalyst, the most significant changes, particularly in the ester formation bands (C=O, 1722 cm⁻¹), occur mainly during post-curing, evidencing slower kinetics and limited reactivity at 150 °C [61].

Comparative analysis reveals that both catalysts improve the efficiency of the reaction compared to the uncatalyzed system, but act differently. Caffeine promotes early conversion of epoxide groups and rapid formation of ester bonds during curing, with stabilization in post-curing. Zinc acetate induces a more progressive formation of ester bonds, with continuous spectral changes even during post-curing, suggesting a more sustained catalytic effect that favors a denser polymer network.

Differential scanning calorimetry (DSC) was employed to investigate the thermal behavior and curing characteristics of the formulations. Dynamic DSC experiments were conducted at heating rates of 5, 10, 15, and 20 °C/min over the temperature range of 25–230 °C. Figure 3 shows representative thermograms at 10 °C/min, while Figure S7 (Supplementary Materials) presents the complete set of curves for all heating rates.

The thermograms exhibit two sequential thermal events: an endothermic peak corresponding to the melting of unreacted components (eugenol-based epoxy and adipic acid), followed by a broad exothermic peak corresponding to the curing reactions (epoxy-acid ring-opening and esterification). These events are partially overlapping, with the extent of overlap varying systematically with heating rate (Figure S7), which has important implications for quantitative analysis. At 10 °C/min (Figure 3), the melting peak appears at 117–129 °C depending on the formulation, while the curing exotherm spans approximately 80–90 °C from onset to completion, indicating a complex multi-step process consistent with epoxy-acid chemistry [62,63].

The melting temperature (T_m) increases with heating rate for all formulations (Figure S7,

Table 1. Curing data were obtained from the dynamic DSC thermograms recorded at 5, 10, 15, 20 °C/min heating rates for all formulations.

Sample	β (°C/min)	Ti (°C)	Tp (°C)	Tf (°C)	ΔHc (mJ/mg)
Caffeine 5%	5	130.3	148.0	198.7	157.1
	10	134.4	153.6	212.2	169.2
	15	138.0	157.2	220.0	191.1
	20	142.1	162.2	219.9	201.9
Caffeine 10%	5	128.7	146.5	204.8	113.1
	10	130.5	153.2	218.1	169.9
	15	137.8	158.9	218.8	208.2
	20	138.8	163.4	219.2	238.1
Zn(OAc)2 5%	5	125.8	137.6	204.2	149.2
	10	128.1	146.2	220.9	157.8
	15	130.7	150.5	224.1	226.0
	20	135.7	154.7	227.0	293.2
Zn(OAc)2 10%	5	130.6	149.6	209.5	117.1
	10	134.9	153.8	213.1	141.2
	15	141.2	159.4	219.7	209.1
	20	144.8	164.0	229.6	216.9
Without catalyst	5	129.2	150.2	200.9	106.2
	10	135.3	158.1	218.1	172.2
	15	140.1	158.3	218.2	220.9
	20	145.2	160.6	218.9	254.1

β = heating rate; T_i = onset temperature; T_p = peak temperature; T_f = end temperature; ΔH_c = reaction enthalpy.

), rising by approximately 5–6 °C from 5 to 20 °C/min. This behavior is consistent with thermal lag effects typical of physical phase transitions and does not depend on catalyst type or concentration [64].

The degree of separation between melting and curing events changes systematically with heating rate (Figure S7). At lower heating rates (5 °C/min), greater temporal overlap is observed, with the curing exotherm beginning during the latter stages of melting. At higher heating rates (20 °C/min), the events become more clearly separated. This improved separation occurs because the curing reaction is displaced to higher temperatures more than the melting transition, resulting in a more distinct baseline between events [65].

The apparent curing enthalpy (ΔH_c) increases systematically with heating rate for all formulations (Table 1), suggesting that the measured values are influenced by kinetic and experimental factors rather than reflecting solely the intrinsic reaction energy. Two main effects explain this behavior.

First, signal overlap: At lower heating rates, the endothermic melting signal partially compensates the exothermic curing heat flow, resulting in underestimation of the net enthalpy. At higher heating rates, improved temporal separation between events reduces this interference, yielding higher apparent ΔH_c values [66]. Second, volatilization losses: Slower heating prolongs exposure to intermediate temperatures, potentially causing volatilization of low-molecular-weight species (e.g., adipic acid), which reduces reactive stoichiometry and measured enthalpy [67]. Faster heating minimizes these losses.

Therefore, the observed increase in ΔH_c with heating rate likely reflects improved thermal resolution and reduced experimental artifacts rather than genuine thermodynamic differences. For this reason, a heating rate of 10 °C/min was selected for comparative analysis to balance event separation, minimize artifacts, and align with standard practice in the literature for epoxy-based systems [68]. All subsequent ΔH_c comparisons are made at this fixed heating rate.

The peak curing temperature (T_p) increases with heating rate for all formulations (Figure 4, Table 1), rising by approximately 13–19 °C from 5 to 20 °C/min. This systematic increase is consistent with Arrhenius-type temperature dependence typical of kinetically controlled reactions, confirming that the curing process is governed by reaction kinetics [69]. The peak curing temperature (T_p) serves as an indicator of catalytic activity [70], representing the temperature at which the reaction rate is maximum (Figure 3, Table 1). All catalyst-containing formulations show T_p values lower than those of the uncatalyzed system (157.1 °C at 10 °C/min), indicating that both zinc acetate and caffeine accelerate the curing reactions under non-isothermal conditions.

Zn(OAc)₂ 5% exhibits the strongest catalytic effect, with T_p approximately 11 °C lower than that of the uncatalyzed system (146.3 °C vs. 157.1 °C at 10 °C/min). This reduction is consistent with the established catalytic activity of metal acetates in epoxy-acid reactions, where zinc species facilitate ring-opening and esterification through coordination and activation of reactive groups [11].

Caffeine-containing formulations show T_p values 3.9–4.5 °C lower than those of the uncatalyzed system (152.6–153.2 °C vs. 157.1 °C at 10 °C/min). While more modest than the Zn(OAc)₂ 5% effect, this reduction represents measurable acceleration of the curing reactions, suggesting that caffeine facilitates the curing process through a mechanism distinct from that of the metal catalyst [71,72]. The similar T_p values for both caffeine concentrations (5% and 10%) suggest that the catalytic effect may approach saturation at the lower loading, or that other factors (e.g., solubility or distribution) limit the impact of increased concentration under dynamic heating conditions [73].

Interestingly, increasing Zn(OAc)₂ concentration from 5% to 10% substantially reduces the catalytic effect (T_p = 153.8 °C, only 3.3 °C lower than uncatalyzed). This non-monotonic behavior suggests possible catalyst deactivation, aggregation, or interference with the reaction mechanism at higher zinc concentrations [74,75].

All formulations complete curing in a similar temperature range, with final temperatures (T_f) at 10 °C/min ranging from 212.2 °C to 220.9 °C (Table 1), independent of catalyst type or concentration. The similarity in T_f despite significant differences in T_p indicates that the catalysts primarily affect reaction kinetics rather than the thermodynamic feasibility or ultimate extent of curing [76].

Comparing apparent curing enthalpies (ΔH_c) at 10 °C/min (Table 1), caffeine-containing systems show moderately lower values than the uncatalyzed system (168–170 mJ/mg vs. 178 mJ/mg, representing 5–6% reduction). In contrast, zinc acetate systems show substantially lower ΔH_c values, with a clear concentration-dependent trend: Zn(OAc)₂ 5% shows 158 mJ/mg (11% reduction) and Zn(OAc)₂ 10% shows 141 mJ/mg (21% reduction) relative to the uncatalyzed system.

These reductions in apparent ΔH_c for Zn(OAc)₂ systems may reflect differences in reaction mechanisms or pathways facilitated by the catalyst [36]. A lower ΔH_c at constant heating rate can indicate faster conversion over a narrower temperature range or alternative, less energy-intensive reaction routes [77]. The caffeine systems, showing ΔH_c values closer to the uncatalyzed reference, appear to have less pronounced effects on overall reaction energetics. However, given the potential for signal overlap and baseline distortions discussed above, these observations must be interpreted as apparent thermodynamic differences rather than definitive mechanistic conclusions. The primary evidence for catalyst effects comes from the more robust T_p parameter, which is less affected by experimental artifacts.

Peak temperatures (T_p) for each heating rate were determined by deconvolution of the overlapping melting and curing events using OriginPro2025 software. Gaussian peak fitting was applied to separate the endothermic melting signal from the exothermic curing peak, enabling accurate determination of T_p for the curing reaction.

The activation energy (E_a) of the curing process was determined using the Kissinger method (Equation (1)) [78].

$$\ln {\displaystyle \frac{\mathsf{\beta }}{\left[{\mathrm{T}}_{\mathrm{p}}^2\right]}}={\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}{\mathrm{T}}_{\mathrm{p}}}}+\ln {\displaystyle \frac{\mathrm{AR}}{{\mathrm{E}}_{\mathrm{a}}}}$$

(1)

where β is the heating rate, T_p is the peak curing temperature, R is the universal gas constant (8.314 J mol⁻¹ K⁻¹), A is the pre-exponential factor, and E_a is the activation energy. The value of E_a is derived from the slope of the linear plot of ln(β/T_p²) versus 1/T_p, as shown in Figure 5. The detailed linear fits for all formulations are presented in

Table 2. Kinetic parameters calculated using the Kissinger method for the curing reaction of epoxy vitrimer.

Sample	A (min−1)	Ea (kJ/mol)	R2
Caffeine 5%	9.1585 × 1014	133.1487	0.9944
Caffeine 10%	7.9974 × 1013	124.3774	0.9936
Zn(OAc)2 5%	9.1625 × 1013	122.6481	0.9995
Zn(OAc)2 10%	6.6266 × 1012	116.0136	0.9913
Without catalyst	1.92483 × 1015	135.6679	0.9914

, with correlation coefficients (R²) > 0.99 indicating good linearity.

It should be noted that the Kissinger method assumes a single-step reaction and yields an apparent activation energy for the overall curing process [79]. Given the complexity of the epoxy-acid system (involving ring-opening, esterification, and partial overlap with melting), the E_a values obtained represent the energy barrier for the composite process rather than individual elementary reactions. Nevertheless, these values provide a useful comparative measure of the catalytic effect on the overall curing kinetics.

The activation energy for the non-catalyzed reference system was determined to be 135.7 kJ/mol, which is consistent with values reported in the literature for epoxy-acid curing reactions relying on thermal energy alone [80]. The incorporation of both catalysts reduced the energy barrier for the reaction.

In the caffeine-containing system, increasing the caffeine concentration from 5% to 10% resulted in a decrease in Ea from 133.1 to 124.4 kJ/mol. This concentration-dependent reduction indicates that caffeine acts as a kinetic facilitator for the curing reaction, with its accelerating effect becoming more pronounced at higher concentrations [70,81].

Zinc acetate showed a greater reduction in E_a compared to caffeine at equivalent concentrations. At 5% concentration, Zn(OAc)₂ (E_a = 122.6 kJ/mol) achieved a larger reduction compared to 5% caffeine (133.1 kJ/mol). This effect was more pronounced at 10% concentration, with Zn(OAc)₂ reaching the lowest E_a in the study (116.0 kJ/mol), compared to 124.4 kJ/mol for caffeine 10%.

Based on the activation energies, the following order of catalytic effectiveness can be established (from highest to lowest reduction in E_a): Zn(OAc)₂ 10% > Zn(OAc)₂ 5% > caffeine 10% > caffeine 5% > non-catalyzed system.

It is noteworthy that while Zn(OAc)₂ 10% exhibits the lowest activation energy, this formulation showed a higher peak temperature (T_p) compared to Zn(OAc)₂ 5% in the DSC analysis, indicating non-monotonic behavior. This suggests that the catalytic mechanism at higher zinc concentrations may involve additional factors (e.g., aggregation or coordination effects) that influence the overall reaction kinetics in a complex manner [11,74].

2.3. Influence of the Catalyst on the Structure of Eu-Epx/AA Vitrimers

Figure 6 shows the comparative analysis of the FTIR-ATR spectra of polymeric materials cured without catalyst and in the presence of different catalysts (5% caffeine, 5% Zn(OAc)₂), in order to evaluate the resulting structural transformations. All spectra were normalized at 1580 cm⁻¹ (aromatic ring vibrations) [82], allowing for a quantitative comparison of the functional groups present after the complete curing process.

The broad absorption at ~3430 cm⁻¹ [42] is consistent with the presence of hydroxyl groups, which may originate from both unreacted carboxylic acid and β-hydroxyester formation following epoxy ring opening [76,77].

The broad absorption at ~3430 cm⁻¹ is consistent with the presence of hydroxyl groups, which may originate from both unreacted carboxylic acid and β-hydroxyester formation following epoxy ring opening [76,77]. The higher intensity of this band in the sample without catalyst, compared to the catalyzed samples, suggests incomplete conversion, with higher concentrations of unreacted carboxylic acid groups. Conversely, the reduced intensity in the catalyzed samples indicates more complete curing reactions, with greater conversion of epoxy and carboxylic acid groups. It should be noted that the hydroxyl groups present in the β-hydroxyester bonds formed during curing are essential for the transesterification exchange reactions that enable vitrimer behavior [78].

The C=O stretching vibration at 1722 cm⁻¹ [57], corresponding to both ester and carboxylic acid groups, shows higher relative intensity in the uncatalyzed sample compared to the catalyzed samples. Given the persistence of epoxide bands (926 cm⁻¹, discussed below) in the uncatalyzed sample, this higher C=O intensity likely reflects the coexistence of formed esters and unreacted carboxylic groups, indicating incomplete conversion. A distinctive band at 1540 cm⁻¹, attributed to the asymmetric COO⁻ stretching vibration [83], was detected exclusively in the sample with Zn(OAc)₂, indicating the formation of zinc-carboxylate coordination complexes. This suggests that zinc acetate may catalyze the reaction through coordination with carboxylic acid groups [79,80].

The transformation of the epoxide units was corroborated by the substantial reduction or absence of characteristic vibrations typically associated with epoxide rings at 1470, 1329, and 1296 cm⁻¹ [58] in the cured samples, accompanied by the appearance of a C–O–C stretching vibration at 1160 cm⁻¹ confirming the formation of ester bonds [81].

Finally, the nearly complete disappearance of the epoxide ring C-O-C stretching vibration at 926 cm⁻¹ (as well as the bands at 843 and 741 cm⁻¹) [58] in the catalyzed samples, in contrast to its residual presence in the uncatalyzed material, indicates that the catalysts promote higher degrees of epoxide conversion and more complete curing reactions [82,83].

These results indicate a clear correlation between catalyst presence and curing efficiency, with catalyzed samples showing higher degrees of epoxide conversion and more complete transformation of reactive groups compared to the uncatalyzed system.

2.4. Thermal Stability and Glass Transition

Thermogravimetric analysis (TGA) evaluated the influence of the catalyst on the thermal stability of the polymer systems. TGA curves (Figure 7) showed a multistage degradation characteristic of cured epoxies. The control sample, without catalyst, showed the lowest thermal stability (T_onset 350.3 °C) and two degradation peaks at 357.6 and 396.3 °C (

Table 3. Thermogravimetric analysis results for Eu-Epx/AA vitrimers.

Sample	Tonset (°C)	Tmax 1 (°C)	Tmax 2 (°C)	Char (%)	Degradation Profile
Caffeine 5%	357.11	373.9	400.67	19.96	Bimodal
Caffeine 10%	364.21	380.3	404.6	21.21	Bimodal
Zn(OAc)2 5%	359.4	378.5	404.25	23.56	Bimodal
Zn(OAc)2 10%	354.37	392.5	-	23.61	Unimodal
Without catalyst	350.28	357.61	396.28	27.14	Bimodal

T_onset: Onset degradation temperature (5% weight loss); T_{max 1}: Temperature of maximum degradation rate (first peak); T_{max 2}: Temperature of maximum degradation rate (second peak, if present); Char: Residual mass at 700 °C.

), indicative of an imperfect network containing segments more susceptible to thermal degradation with stepwise decomposition associated with limited crosslinking efficiency [29,84,85].

The incorporation of catalysts improved thermal stability. Both caffeine and zinc acetate increased T_onset, reaching the maximum value (364.2 °C) with 10% caffeine, consistent with a denser and more stable network resulting from more efficient curing [86,87].

The degradation profiles varied depending on the catalyst: the formulations with caffeine maintained two degradation events, while 10% Zn(OAc)₂ showed a single sharp peak at 392.5 °C. This unimodal degradation profile suggests that zinc acetate at higher concentrations promotes a more homogeneous network, where degradation of the main structure occurs in a narrower temperature range and at a higher temperature [86,88].

These results indicate that catalysis not only increases thermal stability but also influences the homogeneity of the polymer network [84].

The glass transition temperature (T_g) of the Eu-Epx/AA polymer networks was determined using differential scanning calorimetry (DSC). As shown in Figure 8, the control sample without catalyst exhibited a T_g of 41.7 ± 1.5 °C. The addition of both catalysts, caffeine and zinc acetate (Zn(OAc)₂), at concentrations of 5% and 10%, resulted in slightly higher T_g values ranging from 43.7 to 44.9 °C. These T_g values are remarkably similar across all formulations (within ~2 °C), suggesting that the presence and concentration of catalyst have a minimal effect on the glass transition temperature [89]. This behavior is consistent with the fact that all formulations share the same chemical composition and curing conditions, leading to polymer networks with comparable structural characteristics [89].

In crosslinked polymer networks, the glass transition temperature is primarily influenced by the chemical structure of the network segments, the crosslink density, and the free volume [90]. Since all formulations are based on the same epoxy–acid system, the similar T_g values observed are expected. The T_g values obtained in this study (41–45 °C) are consistent with those commonly reported for epoxy–carboxylic acid networks derived from aliphatic dicarboxylic acids [89,91], supporting the reliability of the thermal analysis.

These results indicate the effectiveness of caffeine and zinc acetate as catalysts for the curing reaction between epoxy and carboxyl groups. The catalysts, especially at higher concentrations, promote the formation of more thermally stable networks, as indicated by higher onset degradation temperatures (T_onset). Additionally, zinc acetate at 10% promotes a more homogeneous network structure, as evidenced by the transition from bimodal to unimodal degradation profiles [92,93,94].

2.5. Swelling Behavior and Gel Content

Table 4. Gel Content and Swelling Ratio Analysis of Cured Vitrimer Formulations.

Formulation	Gel Content (%, THF)	Swelling Ratio (Q, THF)	Swelling Ratio (Q, DCM)
No catalyst	78.3 ± 2.1	3.85 ± 0.18	4.25 ± 0.20
Caffeine 5%	91.2 ± 1.5	2.42 ± 0.12	2.75 ± 0.14
Caffeine 10%	93.8 ± 1.8	2.18 ± 0.10	2.49 ± 0.12
Zn(OAc)2 5%	92.5 ± 1.3	2.35 ± 0.11	2.62 ± 0.13
Zn(OAc)2 10%	95.1 ± 1.1	1.95 ± 0.09	2.21 ± 0.11

presents the results of the analysis of the swelling ratio and gel content using solvents of varying polarity (THF and DCM) to evaluate the degree of cross-linking and network formation in epoxy vitrimers synthesized from eugenol [27,95].

The swelling ratio results reveal a consistent trend across all formulations: catalyst incorporation reduces the degree of swelling, indicating increased crosslinking density. The uncatalyzed formulation exhibited the highest swelling ratios (Q = 3.85 ± 0.18 in THF; Q = 4.25 ± 0.20 in DCM), reflecting lower crosslinking density and greater polymer-solvent interaction. This is consistent with a less densely crosslinked and more heterogeneous structure, as supported by its bimodal thermal degradation profile and lower thermal stability [96,97,98,99].

In contrast, all catalyzed samples showed lower swelling ratios, with values decreasing as catalyst concentration increased. The formulation containing 10% Zn(OAc)₂ presented the lowest swelling values (Q = 1.95 ± 0.09 in THF; Q = 2.21 ± 0.11 in DCM), corresponding to reductions of 49.4% and 48.0%, respectively, compared to the uncatalyzed system. This reduction is consistent with the formation of a dense, well-developed polymer network that restricts solvent diffusion and polymer chain expansion [100,101,102,103].

The gel content analysis complements the swelling data by quantifying the fraction of insoluble, crosslinked material. The uncatalyzed control exhibited a gel content of 78.3 ± 2.1%, implying that over 21% of the material remains extractable, consistent with lower crosslinking efficiency [104,105,106]. In contrast, the catalyzed formulations showed significant increases in gel content, ranging from 91.2% to 95.1%, depending on catalyst type and concentration.

The highest gel content was achieved with 10% Zn(OAc)₂ (95.1 ± 1.1%), indicating the formation of a tightly crosslinked and highly integrated network. These results support the effectiveness of both caffeine and zinc acetate as catalysts, with the latter showing greater efficiency across all parameters. Using the Flory-Rehner Equation (2) [107], crosslinking density (ν_e) was estimated from the swelling data in THF:

$$ve={\displaystyle \frac{-\left[\ln \left(1-\phi \right)+\phi +\chi {\phi }^2\right]}{\left[V_1\left({\phi }^{\wedge }\left(1/3\right)-\phi /2\right)\right]}}$$

(2)

where φ is the volume fraction of the polymer in the swollen gel, χ is the Flory-Huggins polymer-solvent interaction parameter, and V₁ is the molar volume of the solvent.

The results, shown in

Table 5. Estimated Crosslinking Density from Swelling Analysis.

Formulation	Polymer Volume Fraction (φ)	Crosslinking Density (νe, mol/cm3)	Relative Increase
No catalyst	0.206	1.89 × 10−3	-
Caffeine 5%	0.293	2.87 × 10−3	+51.9%
Caffeine 10%	0.314	3.14 × 10−3	+66.1%
Zn(OAc)2 5%	0.298	2.95 × 10−3	+56.1%
Zn(OAc)2 10%	0.339	3.42 × 10−3	+80.9%

, demonstrate 52% to 81% increases in network density with catalyst incorporation, reaching a maximum of 3.42 × 10⁻³ mol/cm³ for Zn(OAc)₂ at 10%.

The highest value (ν_e = 3.42 × 10⁻³ mol/cm³) was again obtained for the 10% Zn(OAc)₂ system, in agreement with its minimal swelling and maximal gel content. These results support a clear structure-property relationship: enhanced crosslinking yields more robust and solvent-resistant networks, confirming the catalytic role of caffeine and zinc acetate in promoting efficient network formation [94,108,109].

2.6. Influence of the Catalyst on Thermomechanical Behavior

Figure 9 shows the thermal expansion curves of the Eu-Epx/AA vitrimers, comparing the material without a catalyst to those containing zinc and caffeine catalysts at various concentrations. The Topology Freezing Transition Temperature (T_v) was determined from these curves according to methods reported in the literature [38,110], identified as the temperature at which the height of the sample began to decrease following the initial thermal expansion. This behavior indicates the onset of macroscopic flow due to the dynamic exchange of covalent bonds, allowing for the topological rearrangement of the network [10].

The results of this analysis, summarized in

Table 6. TMA results for epoxy vitrimers with different catalysts and concentrations.

Sample	Tv (°C)	ΔTMA (μm)	CTE1 (°C−1)
Caffeine 5%	210	41	1.985 × 10−4
Caffeine 10%	198	44	2.188 × 10−4
Zn(OAc)2 5%	189	52	2.176 × 10−4
Zn(OAc)2 10%	132	47	2.745 × 10−4
Without catalyst	235	55	1.787 × 10−4

, show a clear dependence of T_v on both the type of catalyst used and its concentration. The sample without catalyst exhibited the highest T_v at 235 °C, consistent with limited bond exchange activity in the absence of a catalyst [13]. In contrast, the incorporation of catalysts significantly reduced T_v, with zinc acetate (Zn(OAc)₂) showing greater effectiveness than caffeine. Specifically, increasing the concentration of Zn(OAc)₂ from 5% to 10% resulted in a reduction in T_v from 189 °C to 132 °C, representing a decrease of 57 °C. This confirms that zinc acetate acts as an effective catalyst for transesterification, accelerating the kinetics of dynamic exchange and reducing the temperature required for network rearrangement. This trend aligns with literature findings, which indicate that higher concentrations of catalyst accelerate the kinetics of dynamic transesterification, thereby lowering the temperature necessary for network rearrangement [37,111,112].

Caffeine also induced a concentration-dependent reduction in T_v, from 210 °C (5%) to 198 °C (10%), compared to 235 °C for the non-catalyzed system. This confirms that caffeine catalyzes the dynamic transesterification responsible for the vitrimeric behavior. However, this reduction is smaller than the 103 °C decrease achieved with 10% zinc acetate, highlighting a clear difference in catalytic efficiency between the two systems under these conditions [13,113,114].

The coefficient of thermal expansion in the glassy state (CTE₁) was calculated from the initial linear region of the TMA curve (35 °C to T_v). The values ranged from 1.79 × 10⁻⁴ to 2.75 × 10⁻⁴ °C⁻¹, typical of crosslinked epoxy systems [113]. The highest CTE₁ was observed in the sample with 10% Zn(OAc)₂ (2.75 × 10⁻⁴ °C⁻¹), while the uncatalyzed sample exhibited the lowest value (1.79 × 10⁻⁴ °C⁻¹). The correlation between higher CTE₁ and lower T_v suggests increased segmental mobility in the glassy state for catalyzed systems [111]. These results highlight the role of catalyst selection and concentration in determining T_v.

3. Materials and Methods

3.1. Materials

Eugenol (Eu, CAS 97-53-0) ≥ 98%, 1,2-dibromoethane (CAS 106-93-4) 98%, Oxone^® (potassium peroxomonosulfate, CAS 70693-62-8), Zinc acetate (Zn(OAc)₂, CAS 5970-45-6) ≥ 98%, Triethylbenzylammonium chloride (TEBAC, CAS 56-37-1) 99%, caffeine (CAS 58-08-2) 99%, Potassium carbonate (CAS 584-08-7) ≥ 99.8%, adipic acid (AA, CAS 124-04-9) ≥ 99.5%, acetone (CAS 67-64-1) ≥ 99.5%, and ethyl acetate (CAS 141-78-6) ≥ 99.8% were purchased from Sigma-Aldrich (St. Louis, MO, USA) and used without further purification. Sodium chloride (NaCl, CAS 7647-14-5) ≥ 98%, sodium sulfite (Na₂SO₃, CAS 7757-83-7) ≥ 98%, sodium bicarbonate (NaHCO₃, CAS 144-55-8) ≥ 98%, and anhydrous magnesium sulfate (MgSO₄, CAS 7487-88-9) ≥ 98% were purchased from LABSYNTH (Diadema, SP, Brazil). Potassium iodide (KI, CAS 7681-11-0) ≥ 98%, Tetrahydrofuran (THF, CAS 109-99-9), Dichloromethane (DCM, CAS 75-09-2) and absolute ethanol (CAS 64-17-5) > 99.9% were purchased from NEON (Suzano, SP, Brazil).

3.2. Synthesis of Eugenol Diene Compound (Eu-Di)

The eugenol diene compound Eu-Di was synthesized following the methodology proposed by Liu et al. [26] by reacting eugenol with 1,2-dibromoethane, as illustrated in Scheme 1. Eugenol (32.8 g, 0.20 mol), anhydrous potassium carbonate (K₂CO₃, 27.6 g, 0.20 mol), potassium iodide (KI, 1.28 g, 0.008 mol), and absolute ethanol (200 mL) were introduced into a 500 mL three-neck flask and stirred magnetically. The mixture was then heated at 70 °C. When the solution turned dark green, 1,2-dibromoethane (15.03 g, 0.08 mol) dissolved in 20 mL of ethanol was slowly added using a constant pressure funnel. The reaction was maintained at 70 °C for 12 h, after which it was allowed to cool to room temperature and left for an additional 16 h to promote crystallization. The crude product was recovered by filtration, washed with hot water, recrystallized using ethanol, and dried in a vacuum oven at 60 °C for 8 h. The white solid obtained was designated as Eu-Di, with a yield of 91%. ¹H NMR (300 MHz, CDCl₃, δ): 6.92–6.87 (m, 2H, Ar–H), 6.74–6.66 (d, 4H, Ar–H), 6.06–5.90 (ddt, 2H, CH=CH–CH₂), 5.12–5.03 (m, 4H, CH₂=CH–), 4.39–4.37 (s, 4H, OCH₂CH₂O), 3.88–3.77 (s, 6H, ArOCH₃), 3.36–3.31 (dt, 4H, CH₂–CH=CH–). 13C NMR (300 MHz, CDCl3, δ):149.83, 146.72, 137.72, 133.70, 120.75, 115.78, 114.63, 112.84, 67.99, 56.13, 39.96.

3.3. Synthesis of Epoxy Derived from Eugenol (Eu-Epx)

Eu-Di (1 g) was dissolved in a mixture of acetone (60 mL) and ethyl acetate (50 mL) in a 500 mL three-neck flask equipped with magnetic stirring, a thermometer, and a pressure funnel. To this solution, 0.6 g (0.8 mmol) of tetrabutylammonium bromide (TEBAC) and 5.80 g of sodium bicarbonate (NaHCO₃) were added. The flask was kept at 5–8 °C, and then 3.5 g of Oxone^® (equivalent to 9.89 mmol of KHSO₅) dissolved in 20 mL of H₂O was added dropwise while maintaining vigorous stirring and low temperature. The mixture was then maintained at room temperature for 72 h. The organic layer was separated and washed with a saturated aqueous solution of Na₂SO₃. The crude product was recrystallized from ethyl acetate to obtain the final product as a light-yellow powder, with a yield of 60%. The purity of the Eu-Epx monomer was evaluated by quantitative ¹H NMR using eugenol as an internal standard. Spectra were recorded in CDCl₃ with TMS, integrating the epoxide signals (δ ≈ 2.90–3.24 ppm) and the aromatic protons of eugenol (δ ≈ 6.50–7.25 ppm). The monomer purity was determined to be approximately 90%, with minor deviations attributed to residual or partially converted species.

The equivalent epoxide weight (EEW) of Eu-Epx was estimated by combining Gel permeation chromatography (GPC) and 1H-NMR analyses, following the method of Garea et al. [115]. GPC was performed using chloroform as the eluent and polystyrene standards to determine the molecular weight distribution. 1H-NMR spectra (in CDCl₃) were used to integrate the epoxide and aromatic proton signals. The EEW was calculated by correlating the average molecular weight obtained by GPC (Figure S4) with the epoxide proton content determined by NMR. ¹H NMR (300 MHz, CDCl₃, δ): 6.97–6.88 (d, 2H, Ar–H), 6.81–6.74 (m, 6H, Ar–H), 4.41–4.32 (s, 2H, Ar–OCH₂–=), 3.89–3.84 (s, 6H, Ar–OCH₃), 3.08–3.17 (s, 2H, –CH–CH₂–O), 2.74–2.83 (m, 2H, –CH–CH₂–O), 2.50–2.56 ppm (s, 4H, Ar–CH₂– and –CH–CH₂–O). 13C NMR (300 MHz, CDCl₃, δ): 150.01, 147.31, 131.02, 121.30, 114.95, 113.50, 68.18, 56.24, 52.60, 46.85, 38.47.

3.4. Preparation of the Epoxy Acid Curing System

Epoxy-acid networks were prepared through a curing process involving the reaction between Eu-Epx and adipic acid (AA) in a 1:1 molar ratio. Zn(OAc)₂ and caffeine were used as catalysts (Scheme 2). The catalyst concentrations of 5 and 10 mol% (relative to COOH groups) were selected based on literature precedents for transesterification-based vitrimers [31,33,35] and preliminary optimization studies. The 5% level represents a commonly reported threshold for achieving dynamic exchange activity, while the 10% concentration allows the assessment of enhanced catalytic performance without causing issues like excessive catalyst aggregation or plasticization of the polymer network, which can occur at very high loadings. This systematic comparison enables the optimization of catalyst dosage with respect to curing kinetics, network conversion, and thermal properties. A control sample was prepared without adding any catalysts to evaluate the effect of the catalytic agents on the curing process. The components were mixed with mortar to ensure homogeneous distribution and prevent premature curing [21,34,36,38].

Thermal curing protocol: The curing conditions were determined based on preliminary DSC analysis (10 °C/min), which indicated that the curing reaction initiates at 130 °C, peaks at 150 °C, and completes by 190 °C. A three-stage thermal protocol was implemented (

Table 7. Thermal curing protocol for all formulations.

Stage	Temperature (°C)	Time (Min)	Purpose
1. Initial curing	130	60	Gelation and network formation
2. Intermediate curing	150	60	Network densification
3. Post-curing	190	120	Network completion and equilibration

):

The curing temperatures were selected based on preliminary DSC analysis (10 °C/min, see Supplementary Materials Figure S6), which showed that the curing reaction initiates at 130 °C, reaches maximum rate at 150 °C (T_p), and completes by 190 °C. The three-stage protocol ensures controlled gelation (Stage 1), network densification at peak reaction rate (Stage 2), and complete conversion with network equilibration (Stage 3). All formulations were subjected to the identical protocol to enable valid comparative analysis. Samples were analyzed after each stage to verify curing progression.

3.5. Characterization Techniques

The chemical structures of Eu-DI and Eu-Epx were characterized using proton (¹H) and carbon-13 (¹³C) nuclear magnetic resonance (NMR) spectroscopy. The NMR spectra were recorded on a Mercury VX 300 (Varian, Palo Alto, CA, USA) spectrometer operating at a frequency of 300 MHz, using deuterated chloroform (CDCl₃) as the solvent and tetramethylsilane (TMS) as the internal standard. Analyses were performed at a constant temperature of 25 °C. Gel permeation chromatography (GPC) analysis was conducted using an LC-20AD system (Shimadzu, Nakagyo-ku, Japan) equipped with a CTO-20A oven, DGU-20A3R degasser, and RID-20A refractive index detector. Separation was performed using two Phenogel columns in series: 5 μm, 50 Å (300 × 7.8 mm) and 5 μm, 100 Å (300 × 7.8 mm). Chloroform was used as the eluent at 1.0 mL/min, with a column temperature of 30 °C and injection volume of 20 μL. Calibration was based on a 0.2% w/v polystyrene standard mixture.

Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR-FTIR) was employed to analyze the monomers, study the curing reaction, and characterize the chemical structure of the final materials. The measurements were conducted using a Frontier FT-IR/FIR (PerkinElmer, Waltham, MA, USA) instrument over a wavelength range of 500 to 4000 cm⁻¹, with a resolution of 4 cm⁻¹, and 60 scans per sample. The FTIR spectra were processed using PerkinElmer Spectrum 10.4 software to apply the ATR correction.

The curing behavior and glass transition temperature (T_g) of the materials were assessed using differential scanning calorimetry (DSC) with a DSC7000 (Hitachi, Chiyoda-ku, Japan) instrument. Samples weighing 8–10 mg were analyzed in a nitrogen atmosphere at a flow rate of 50 mL/min. Before performing the experiments, the DSC equipment was calibrated for both temperature and enthalpy using a high-purity indium standard to ensure accurate measurement. Dynamic curing studies were conducted by heating the samples from 30 to 260 °C at various rates (5, 10, 15, and 20 °C/min). The total heat of reaction was determined by calculating the area under the exothermic peak in the DSC thermograms using Hitachi TA7000 software. To verify the reliability of the peak curing temperature (T_p) values, a deconvolution analysis was performed on representative DSC thermograms using OriginPro2025 (OriginLab, Northampton, MA, USA) software. Gaussian peak shapes were assumed, and an excellent fit was achieved (R² > 0.99). The T_p values obtained from the deconvoluted curves were consistent with those derived from the original composite peaks, within the expected instrumental precision. The T_g was subsequently determined using a constant heating and cooling rate of 10 °C/min for both cycles.

The thermal stability of the cured samples was evaluated using thermogravimetric analysis (TGA) with a Q-500 analyzer (TA Instruments, New Castle, DE, USA). Samples were heated to a temperature range from room temperature to 700 °C under a constant nitrogen flow of 50 mL/min, with a heating rate of 10 °C/min.

To quantitatively assess the crosslinking density and network integrity of the eugenol-based epoxy vitrimers, swelling ratio and gel content measurements were performed. Cured samples from each formulation, with approximate dimensions of 5 mm × 5 mm × 1 mm and an initial mass (W₀) of approximately 50 mg, were used. The samples were immersed in 5 mL of solvent at room temperature (23 ± 1 °C). To ensure a comprehensive characterization, two solvents with different polarities were employed, Tetrahydrofuran (THF) and Dichloromethane (DCM). The immersion time was 72 h to reach swelling equilibrium. After this period, the samples were carefully removed, the surface solvent was blotted with filter paper, and the swollen mass (W_s) was immediately recorded. Subsequently, the samples were dried under vacuum at 60 °C until a constant weight was achieved to determine the final dry mass (W_ḍ). The swelling ratio (Q) and the gel content (%) were calculated using Equation (3) and Equation (4), respectively. All measurements were performed in triplicate, and the results are reported as the mean ± standard deviation.

$$Q={\displaystyle \frac{W_s}{W_d}}$$

(3)

$$\mathrm{G}\mathrm{e}\mathrm{l} \mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}={\displaystyle \frac{W_d}{W_0}}\times 100\%$$

(4)

The topology-freezing transition temperature (T_v) was determined using a TMA-60 (Shimadzu, Japan) thermomechanical analyzer. The measurements were performed in compression mode using a glass probe, applying a constant force of 0.5 N while maintaining a nitrogen flow of 50 mL/min. Before the measurements, a conditioning cycle was performed, which involved heating the samples from 25 to 80 °C and subsequently cooling them to 30–35 °C to eliminate any internal stresses, as reported in the literature [40]. Following this conditioning, a heating ramp was executed from 35 to 260 °C at a rate of 5 °C/min, adhering to an emerging and increasingly supported method for T_v identification [116].

4. Conclusions

This work reports the synthesis and comprehensive characterization of epoxy vitrimers derived from renewable eugenol cured with adipic acid, demonstrating that catalyst type and concentration are crucial for controlling curing kinetics, network structure, and dynamic properties. Both caffeine and zinc acetate acted as effective curing accelerators, enhancing the epoxy-acid reaction compared to the non-catalyzed system. Multi-technique characterization (FTIR, DSC, TGA, swelling analysis, and TMA) revealed that the catalysts promote the formation of a polyhydroxyester network with enhanced crosslinking efficiency and vitrimer behavior. Zinc acetate, particularly at 10% concentration, proved to be the most efficient catalyst, achieving the lowest apparent activation energy (116.0 kJ/mol, 14.5% reduction from uncatalyzed) and producing the densest network (crosslinking density ν_e = 3.42 × 10⁻³ mol/cm³, 81% increase). This formulation exhibited a unique unimodal thermal degradation profile, indicating superior network homogeneity, and showed the lowest topology freezing transition temperature (T_v = 132 °C, 44% reduction), confirming effectiveness in catalyzing transesterification reactions. However, the non-monotonic behavior observed at 10% concentration (higher peak curing temperature despite lower activation energy) suggests complex catalytic mechanisms that warrant further investigation. Caffeine demonstrated catalytic activity, notably in the initial curing stage at moderate temperatures (150 °C). It reduced the activation energy to 124.4 kJ/mol at 10% (8.3% reduction from uncatalyzed) while promoting rapid epoxide conversion. Although its catalytic efficiency is moderate compared to zinc acetate, its bio-based origin and non-toxic nature make it a promising green alternative for designing vitrimers with lower environmental impact. Further mechanistic studies would be valuable to fully elucidate its catalytic pathways and optimize its application in vitrimer systems. The presence of catalysts directly influenced crosslinking density and vitrimer behavior. Swelling and gel content analyses confirmed superior network integrity in catalyzed systems, with gel content increasing from 78.3% (uncatalyzed) to 95.1% (10% Zn(OAc)₂). Thermal expansion experiments revealed distinct topology freezing transitions in all catalyzed materials, confirming their vitrimer nature and potential for reprocessing and recycling applications. This study validates the use of bio-based precursors for vitrimer synthesis and provides key guidelines for catalyst selection: zinc acetate for maximum crosslinking density and lowest T_v, and caffeine as a promising greener alternative offering a favorable balance between catalytic efficiency and sustainability. These findings contribute to the development of advanced circular materials that combine the performance advantages of thermosets with the recyclability of thermoplastics, addressing critical challenges in sustainable polymer science.