Bio-based imine- and disulfide-containing epoxy vitrimers with practical self-healing properties at room temperature

Abstract

Bio-based vitrimers have attracted considerable attention because of their carbon neutrality, healability, and recyclability, which contribute to resource saving and energy saving. The reactions of vanillin with cystamine and 1,6-diaminohexane yielded a phenolic hardener containing both imine and disulfide groups (DVNCTA) and a phenolic hardener containing only imine groups (DVNDAH), respectively. These hardeners were utilized to cure epoxy resin mixtures comprising bio-based polyglycerol polyglycidyl ether (PGPE) and petroleum-based flexible poly (ethylene glycol) diglycidyl ether (PEGDGE) in varying molar ratios. The cross-linking density, glass transition temperature, and mechanical strength of the DVNCTA-cured epoxy vitrimers diminished as the ratio of PGPE to PEGDGE decreased. Remarkably, all cured products underwent at least three successful self-healing cycles by standing at room temperature for 24 h. The tensile strength-based healing efficiency (η_σ) of the DVNCTA-cured epoxy vitrimers improved with decreasing PGPE to PEGDGE ratios, reaching a maximum η_σ (100%) at PGPE/PEGDGE ratio of 1/2. Notably, when comparing DVNCTA- and DVNDAH-cured epoxy vitrimers with identical PGPE/PEGDGE ratios, the former exhibited significantly superior healing performance.

Keywords

Bio-based epoxy vitrimer vanillin cystamine dynamic imine and disulfide bonds self-healing properties thermal and mechanical properties

Introduction

The thermosetting epoxy curing system, comprising epoxy resins, hardeners (e.g., polyamines, polyphenols, and acid anhydrides), and catalysts, is widely applied in coatings, adhesives, electronic materials, industrial tooling, and aerospace industry owing to its excellent adhesion, chemical resistance, and electric insulation properties.¹ However, most conventional epoxy resins and their hardeners are derived from petroleum resources. Furthermore, the cured products are inherently infusible and insoluble, making them challenging to recycle and reprocess. To address these issues, bio-based epoxy vitrimers (renewable epoxy networks derived from biomass and containing exchangeable dynamic covalent bonds) have drawn significant attention. These materials can repair cracks that have physical damage, thereby extending service life and conserving petroleum resources.^2–6 Several types of exchangeable dynamic covalent bonds have been investigated for epoxy vitrimers, transesterification,^7,8 disulfide metathesis,^9–11 imine exchange reactions,^12–15 and boronic acid ester exchange.^16,17 Cystamine (CTA), a decarboxylated derivative of the amino acid cystine, is a promising bio-based disulfide-containing diamine that can function as an amine-based epoxy hardener.¹⁸ Although studies on CTA-cured bio-based epoxy vitrimers containing exchangeable disulfide bonds are limited, several noteworthy contributions exist. For example, Roig et al. reported bio-based epoxy vitrimers obtained by curing eugenol-derived diglycidyl ether with CTA.¹⁹ Similarly, Guggari et al. developed bio-based epoxy vitrimers by curing diglycidyl ether of vanillyl alcohol with CTA,²⁰ while our team reported curing sorbitol polyglycidyl ether with CTA to produce bio-based epoxy vitrimers.²¹ Besides CTA, vanillin (VN) is a promising candidate for preparing bio-based vitrimers owing to its status as one of the few industrially available bio-based aromatic compounds.²² Its aldehyde group can readily form exchangeable imine bonds through a reaction with primary amines.^23,24 Wang et al. developed an imine- and disulfide-containing bio-based epoxy vitrimer by curing diglycidyl ether of an imine-containing bisphenol prepared from VN and p-phenylenediamine with CTA. This vitrimer was reprocessable via compression molding at 120–130°C.²⁵ Guggari et al. reported an imine- and disulfide-containing bio-based epoxy vitrimer obtained by curing a mixture of diglycidyl ether of bisphenol A and diglycidyl ether (DDBB) derived from vanillin glycidyl ether and CTA at the weight ratio of 70/30 with petroleum-based 4-aminophenyl disulfide. The epoxy vitrimer was reprocessable at 170°C.²⁶ Verdugo et al. reported curing DDBB with isophorone diamine to create a vitrimer reprocessable at 140°C.²⁷ Despite these advancements, the healing properties of these imine- and disulfide-containing bio-based epoxy vitrimers prepared using CTA and VN have not been explored. Recently, we investigated the healing properties of a bio-based epoxy vitrimer by curing bio-based polyglycerol polyglycidyl ether (PGPE) with an imine- and disulfide-containing diamine hardener derived from an ethylene-bridged bisvanillin and CTA.²⁸ While this vitrimer demonstrated healability for at least three cycles, healing required compression molding at 120°C under 1 MPa. Self-healing materials that can heal under ambient temperature and pressure conditions are highly desirable.

In this study, a VN- and CTA-derived bisphenol (DVNCTA) and VN- and 1,6-diaminehexane-derived bisphenol (DVNDAH) were synthesized via the condensation reactions of VN with CTA and DAH at a molar ratio of 2:1 (Scheme 1). The DVNCTA, containing dynamic imine and disulfide bonds, and DVNDAH, containing only dynamic imine bonds, were employed as epoxy hardeners of mixtures of bio-based PGPE (with a functionality of approximately 4) and petroleum-based poly (ethylene glycol) diglycidyl ether (PEGDGE) (Scheme 2). The effects of the PGPE/PEGDGE ratio and hardeners (DVNCTA or DVNDAH) on the thermal and mechanical properties, and self-healing properties driven by the exchange reactions of imine and disulfide bonds for the cured products were systematically studied (Scheme 3). Consequently, we successfully developed bio-based epoxy vitrimers with excellent self-healing capability at room temperature.

Scheme 1.

Preparation of DVNCTA and DVNDAH.

Scheme 2.

Preparation of cured epoxy products by the reaction of mixtures of PGPE and PEGDGE with DVNCTA or DVNDAH.

Scheme 3.

Exchange reactions of (a) imine and (b) disulfide bonds.

Experimental details

Materials

PGPE (trade name: DENACOL^® EX-512, with an epoxy equivalent weight of 167.5 g/eq., an average functional group number of 4, a chlorine content of 6.5%, and a viscosity of 1300 mPa s at 25°C) was supplied by Nagase ChemteX. Corp. (Tokyo, Japan). PEGDGE (average M_n: 500) was purchased from Sigma–Aldrich Japan Co., Ltd (Tokyo, Japan). VN, cystamine dihydrochloride, 1,6-diaminohexane (DAH), and 2-ethyl-4-methylimidazole (2E4MZ) were sourced from the Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Methyl ethyl ketone (MEK, another name: 2-butanone), tetrahydrofuran (THF), and diethyl ether were purchased from Kanto Chemical Co., Inc. (Tokyo, Japan). CTA (colorless viscous liquid) was prepared by the reaction of CTA dihydrochloride and sodium hydroxide, as previously described.²⁹ All commercially available reagents were used as received without further purification.

Synthesis of DVNCTA

VN (14.9 g, 97.8 mmol) was added to a solution of CTA (7.44 g, 48.9 mmol) in diethyl ether. The resulting solid was crushed and thoroughly mixed with the solution at room temperature for 15 min. After allowing the mixture to stand at room temperature, the precipitate was separated by decantation, and dried in a vacuum oven at 60°C for 24 h, yielding DVNCTA as a yellow powder (yield: 19.2 g, 93%).

Synthesis of DVNDAH

VN (8.22 g, 54.0 mmol) was added to a solution of DAH (3.14 g, 27.0 mmol) in ethanol, and the resulting solution was stirred at room temperature for 24 h. The precipitate was filtered under suction, washed with diethyl ether, and dried in a vacuum oven at 80°C for 24 h, yielding DVNDAH as a yellow powder (yield: 9.47 g, 91%).

Preparation of PGPE/PEGDGE/DVNCTA and PGPE/PEGDGE/DVNDAH cured products

A typical preparation method of a PGPE/PEGDGE/DVNCTA cured product at the epoxy/OH ratio of 1:1 and PGPE/PEGDGE epoxy ratio of 1:1 (PG/PEG-1/1-CTA) is as follows: A solution of PGPE (1.60 g, 9.55 mmol-epoxy), PEGDGE (2.39 g, 9.55 mmol-epoxy), DVNCTA (4.01 g, 19.1 mmol-OH), and 2E4MZ (160 mg, 2 wt% of total weight) in MEK (30 mL) was stirred at room temperature for 30 min. The resulting solution was poured into a culture dish (internal diameter: 100 mm) lined with perfluoroalkoxy alkanes, dried at 80°C for 2 h, 100°C for 2 h in an electric oven, and then under vacuum at 140°C for 2 h to yield a PG/PEG-1/1-CTA film (thickness 0.8 mm). PG/PEG-2/3-CTA and PG/PEG-1/2-CTA films were prepared using the same procedure for PG/PEG-1/1-CTA. A PGPE/PEGDGE/DVNDAH cured product at the epoxy/OH ratio of 1:1 and a PGPE/PEGDGE epoxy ratio of 1:2 (PG/PEG-1/2-DAH) was prepared similarly, except that a mixture of MEK (30 mL) and THF (60 mL) was used as the reaction solvent, and the resulting mixture of PGPE/ PEGDGE/DVNDAH was sonicated at room temperature for 30 min until it is dissolved. The feed reactant amounts and imine and disulfide contents of all the cured products are listed in Table 1.

Table 1.

The amounts of feed reactants and imine and disulfide contents of the cured products.

Sample	PGPE (mol-epoxy): PEGDGE (mol-epoxy)	PGPE g (mmol-epoxy)	PEGDGE g (mmol-epoxy)	DVNCTA or DVNDAH g (mmol-OH)	Imine content (wt%)	Disulfide content (wt%)
PG/PEG-1/1-CTA^*1	1:1	1.60 (9.55)	2.39 (9.55)	4.01 (19.1)	6.45	7.66
PG/PEG-2/3-CTA^*1	2:3	1.25 (7.49)	2.81 (11.2)	3.94 (18.7)	6.33	7.51
PG/PEG-1/2-CTA^*1	1:2	1.03 (6.16)	3.08 (12.3)	3.89 (18.5)	6.25	7.41
PG/PEG-1/2-DAH^*2	1:2	1.08 (6.43)	3.21 (12.9)	3.71 (19.3)	6.52	—

^*1DVNCTA was used as an epoxy hardener.

^*2DVNDAH was used as an epoxy hardener.

Self-healing of cured products

Rectangular samples (40 × 7 × (0.5–1.2) mm³) of the cured products were cut into halves. The two pieces were aligned with their cut surface in contact, sandwiched between two polymethylpentene plates, secured with clips, and left at room temperature for 24 h to produce self-healed PG/PEG-CTA and PG/PEG-DAH specimens (sh1-PG/PEG-CTA and sh1-PG/PEG-DAH). This self-healing process was repeated three times (sh1, sh2, and sh3). The tensile strength healing efficiency (η_σ) was calculated as the tensile strength recovery rate using the following equation:

η_{σ} (%) = 100 σ_{1} / σ_{0}

(1)

where σ₀ and σ₁ are the average tensile strengths of the original and healed samples, respectively.

Measurements

Fourier-transform infrared (FT-IR) spectra were recorded on a Shimadzu IRAffinity-1S (Kyoto, Japan) in the range of 4000–500 cm⁻¹ using the attenuated total reflectance (ATR) method. Each IR spectra was acquired using 50 scans at a resolution of 4 cm⁻¹.

Proton nuclear magnetic resonance (¹H-NMR) spectra were recorded with a Bruker Ascend 400 MHz spectrometer (Madison, WI, USA) using DMSO-d₆ as the solvent.

Differential thermal analysis (DTA) was performed using a Shimadzu DTA-50 instrument at a heating rate of 10°C min⁻¹ in a nitrogen atmosphere.

To quantify the degree of swelling (Ds), a film (10 × 10 × (0.7 − 1.0) mm³) was soaked in MEK at room temperature for 24 h. The degree of swelling was calculated using the following equation:

D s (%) = 100 (w_{1} - w_{0}) / w_{0}

(2)

where w₀ is the initial weight of the film, and w₁ is the weight of the swollen film after soaking. After soaking, the film was dried at 100°C in an electric oven for 24 h. The gel fraction (Gf) obtained by MEK extraction was determined using the equation:

G f (%) = 100 w_{2} / w_{0}

(3)

where w₀ and w₂ represent the weights of the original and dried films, respectively. Three samples of each film were tested, and the mean values and standard deviations were calculated for both the degree of swelling and gel fraction measurements.

Differential scanning calorimetry (DSC) was performed using a Shimadzu DSC-60Plus instrument under a nitrogen atmosphere. A prepared sample (4–7 mg) was cooled to −50°C, and the heating scan was monitored at a rate of 20°C min⁻¹. The glass transition temperature (T_g) was determined as the mid-point of the change in heat flow.

Dynamic mechanical analysis (DMA) was conducted using a Mettler–Toledo DMA1 instrument (Japan) on a rectangular plate sample (20 × 5 × (0.7 − 1.0) mm³) with a chuck distance of 10 mm, a frequency of 1 Hz, and a heating rate of 10°C min⁻¹. The amplitude for the DMA measurements was set at 10 μm. The loss tangent (tan δ) peak temperature (T_α) associated with the glass transition was obtained from the temperature dependence of tan δ. The cross-linking density (v_e) of the cured products was calculated from the following equation:

ν_{e} = E^{'} / (3 R T)

(4)

where R is the gas constant, T is the absolute temperature corresponding to the storage modulus value, and E′ is the storage modulus in the rubbery plateau region.³⁰ To ensure that the cured resins were in the rubbery plateau region, the temperature at which E′ was measured was set at T_α + 50°C for each sample.³¹ R is the gas constant, T is the absolute temperature, and E′ is the storage modulus in the rubbery state.

Thermogravimetric analysis (TGA) was conducted on samples weighing approximately 3–5 mg using a Shimadzu TGA-50 thermogravimetric analyzer at a heating rate of 20°C min⁻¹ under a nitrogen atmosphere. The temperatures at which x% mass loss occurred (T_dx%, x = 5, 10, and 50) were determined from the TGA curves.

Tensile tests of rectangular samples (40 × 7 × (0.5 − 1.2) mm³) were carried out at 20–25°C using a Shimadzu Autograph AGS-X instrument. The span length and testing speed were set at 25 mm and 10 mm min⁻¹, respectively. Three or four specimens were tested for each set of samples, and the mean values and standard deviations were calculated.

Results and discussion

Synthesis and characterization of imine- and disulfide-containing phenolic hardeners

DVNCTA and DVNDAH were synthesized through the condensation reactions of VN with CTA and DAH, respectively (Scheme 1). Figure 1 shows the FT-IR spectra of VN, CTA, DAH, DVNCTA, and DVNDAH. The spectrum of VN exhibits absorption bands corresponding to phenolic hydroxy (OH) and aldehyde carbonyl (C=O) stretching vibrations (ν_OH and ν_C=O ) at 3157 and 1660 cm⁻¹, respectively. The spectra of CTA and DAH reveal absorption bands attributed to the primary amine (NH₂) stretching (ν_NH) and scissoring vibrations (δ_NH) at 3354–3178 and 1604–1589 cm⁻¹, respectively. In the FT-IR spectra of DVNCTA and DVNDAH, characteristic bands of C=O and NH₂ groups were absent. Instead, absorption bands corresponding to ν_OH and the imine C=N stretching vibration (ν_C=N) were observed at approximately 3165–3138 and 1643–1639 cm⁻¹, respectively. These results confirm that DVNCTA and DVNDAH contain phenolic hydroxy groups and imine bonds formed by the condensation reaction of the aldehyde group of VN with the primary amino groups of CTA and DAH.

Figure 1.

FT-IR spectra of DVNCTA and DVNDAH, along with those of VN, CTA, and DAH.

Figure 2 shows the ¹H-NMR spectra of DVNCTA and DVNDAH in DMSO-d₆. DVNCTA and DVNDAH displayed the ¹H-NMR chemical shift (δ) of imine protons at 8.21 and 8.18 ppm (s. 2H, H-e), aromatic protons at 7.38 and 7.37 (s, 2H, H-a), 7.15 and 7.13 (dd or dl, 2H, H-b), and 6.81 and 6.80 ppm (d, 2H, H-c), and those of methoxy protons at 3.79 and 3.80 ppm (s, 6H, H-d). The ¹H-signals of the ethylidene group (-CH₂CH₂-) of DVNCTA were observed at 3.81 (t, 4H, H-f) and 3.02 ppm (t, 4H, H-g). The ¹H-signals of hexamethylene group (-(CH₂)₆-) of DVNDAH were observed at 3.50 (t, 4H, H-f), 1.60 (m, 4H, H-g), and 1.38 ppm (t, 4H, H-h). The chemical structures of DVNCTA and DVNDAH were thus confirmed through the FT-IR and ¹H-NMR analyses.

Figure 2.

400 MHz ¹H-NMR spectra of DVNCTA and DVNDAH in DMSO-d₆.

Preparation and characterization of epoxy-cured products

DTA measurements of a PGPE/PEGDGE/DVNCTA mixture at the epoxy/OH ratio of 1:1 and a PGPE/PEGDGE epoxy ratio of 2:3 and a PGPE/PEGDGE/DAH mixture at the epoxy/OH ratio of 1:1 and PGPE/PEGDGE epoxy ratio of 1:2 were carried out to determine the curing temperature. The onset and peak temperatures of the broad exothermic peaks attributed to the reaction of epoxy and phenolic hydroxy groups of the PGPE/PEGDGE/DVNCTA and PGPE/PEGDGE/DVNDAH mixtures were observed at ca. 82 and 134°C, and 84 and 138°C, respectively (Figure S1, Supplementary Material). Based on the DTA results, the curing temperatures of PGPE/PEGDGE/DVNCTA and PGPE/PEGDGE/DVNDAH mixtures were set to 80–140°C.

The preparation of epoxy-cured products involved drying and curing at 80–140°C of MEK solutions of PGPE/PEGDGE/DVNCTA at the epoxy/OH ratio of 1:1 and PGPE/PEGDGE epoxy ratios of 1:1, 2:3, and 1:2 to produce PG/PEG-1/1-CTA, PG/PEG-2/3-CTA, and PG/PEG-1/2-CTA, respectively. For comparison, PG/PEG-1/2-DAH was prepared similarly to PG/PEG-1/2-CTA, except DVNDAH was used instead of DVNCTA (Scheme 2), and a mixture of MEK/THF 1/2 (v/v) was used as the reaction solvent. All epoxy-cured products were obtained as dark brown films (Figure S2, Supplementary Material). The fact that the films were completely cured was confirmed by no observation of exothermic peaks in the DTA curves (Figure S1). Figure 3 shows the FT-IR spectra of all the cured films and their reactants (PGPE, PEGDGE, DVNCTA, and DVNDAH). The spectra of PGPE and PEGDGE exhibit absorption bands characteristic of epoxy groups at 904, 835, and 756 cm⁻¹. These bands were absent in the spectra of the cured films, indicating the consumption of epoxy groups. However, new alcoholic ν_OH bands were observed at approximately 3375 cm⁻¹, confirming that the epoxy-phenolic hydroxy curing reaction occurred as described in Scheme 2.

Figure 3.

FT-IR spectra of PGPE, PEGDGE, DVNCTA, DVNDAH, and the cured products.

Figure 4 presents the Ds and Gf values of all the cured products obtained using MEK as the soaking solvent. The Ds values for PG/PEG-CTA samples increased as the ratio of PGPE to PEGDGE decreased, suggesting a corresponding decrease in cross-linking density. Additionally, a slight decrease of Gf was observed with a decreasing fraction of PGPE relative to PEGDGE, indicating that the reactions of PEGDGE with DVNCTA generated linear polymers that are soluble in MEK. When PG/PEG-1/2-CTA and PG/PEG-1/2-DAH were compared, PG/PEG-1/2-CTA exhibited slightly higher Ds and lower Gf than PG/PEG-1/2-DAH, suggesting that PG/PEG-1/2-CTA has a slightly lower cross-linking density. The formation of cross-linked structures in all the cured products was confirmed by the high Gf values (84–88%).

Figure 4.

Ds and Gf values of (a) PG/PEG-1/1-CTA, (b) PG/PEG-2/3-CTA, (c) PG/PEG-1/2-CTA, and (d) PG/PEG-1/2-DAH.

Thermal properties of the cured products

Figure 5 presents the DSC curves of all cured products. The glass transition temperature (T_g) values derived from the DSC heating curves are listed in Table 2. The T_g values (23–27°C) slightly decreased as the ratio of PGPE to PEGDGE decreased, likely owing to a reduction of the cross-linking density. Notably, the T_g (23°C) of PG/PEG-1/2-CTA was lower than that (27°C) of PG/PEG-1/2-DAH, reflecting that the -SS- bond in PG/PEG-1/2-CTA exhibits greater rotational freedom compared to the -CH₂CH₂- bond in PG/PEG-1/2-DAH.

Figure 5.

DSC curves of all the cured products.

Table 2.

T_g, T_α, ν_e, T_d5%, T_d10%,, T_d50%, and char yield at 480°C for all the cured products.

Sample	T_g (°C)	T_α (°C)	ν_e (µmol cm⁻³)	T_d5% (°C)	T_d10% (°C)	T_d50% (°C)	Char yield at 480°C (%)
PG/PEG-1/1-CTA	27	58	171	271	304	408	34
PG/PEG-2/3-CTA	25	51	33.8	265	296	404	32
PG/PEG-1/2-CTA	23	40	7.09	260	283	403	31
PG/PEG-1/2-DAH	27	47	12.3	302	329	392	22

Figure 6 shows the temperature dependence of the E′ and tan δ obtained by DMA measurements of the cured products. The T_α values obtained from the DMA curves are listed in Table 2. The T_α values of PG/PEG-CTA films decreased with decreasing ratio of PGPE to PEGDGE, and the T_α of PG/PEG-1/2-CTA was lower than that of PG/PEG-1/2-DAH, consistent with the trends observed for T_g by DSC measurements. The E′ rubbery plateau region was observed for all the cured products at temperatures higher than T_α, indicating the formation of a network structure. The ν_e values of PG/PEG-CTA films also decreased with decreasing ratio of PGPE to PEGDGE, reflecting that the epoxy functionality (ca. 4) of PGPE is higher than that (2) of PEGDGE. The ν_e of PG/PEG-1/2-CTA was lower than that of PG/PEG-1/2-DAH, aligning with the result of Ds and Gf.

Figure 6.

DMA curves of all the cured products.

Table 2 lists the T_dx% (x = 5, 10, and 50) values for all cured products. The TGA curves are shown in Figure S3 (see Supplementary Materials). The T_dx% (x = 5, 10, and 50) values of PG/PEG-CTA films decreased with decreasing ratio of PGPE to PEGDGE ratio, reflecting the corresponding decrease in ν_e. The T_dx% (x = 5 and 10) values of PG/PEG-1/2-CTA were significantly lower than those of PG/PEG-1/2-DAH, consistent with the fact that the bond energy of the S–S bond is lower than that of C–C bond. The char yields of all the cured products decreased in the order: PG/PEG-1/1-CTA > PG/PEG-2/3-CTA > PG/PEG-1/2-CTA > PG/PEG-1/2-DAH. This trend likely results from the decreasing weight fraction of the aromatic moiety-containing DVNCTA or DVNDAH in the same order (Table 1).

Mechanical and self-healing properties of cured products

Figure 7 shows the tensile stress-strain curves and tensile properties of all the cured products. The tensile strength and modulus of PG/PEG-CTA films decreased with decreasing ratio of PGPE to PEGDGE, consistent with the trends observed for T_g and T_α. Conversely, the elongation at break increased by a similar order of magnitude. The tensile strength and elongation at break of PG/PEG-1/2-CTA were lower and higher, respectively than those of PG/PEG-1/2-DAH, whereas their tensile moduli were comparable.

Figure 7.

Tensile stress-strain curves, tensile modulus, tensile strength, and elongation at break of (a) PG/PEG-1/1-CTA, (b) PG/PEG-2/3-CTA, (c) PG/PEG-1/2-CTA, and (d) PG/PEG-1/2-DAH.

The self-healing properties of the cured products were quantitatively assessed by comparing the tensile properties of the original and self-healed samples. The as-prepared cured film samples were first cut into two pieces. These pieces were aligned at the cut surfaces, sandwiched between two polymethylpentene plates, secured with clips, and then allowed to stand at room temperature for 24 h to form self-healed (sh1) films (Figure 8).

Figure 8.

Self-healing behavior of all the cured products.

The sh1-samples were self-healed twice using the same procedure to produce sh2- and sh3-samples. Typical tensile stress-strain curves and η_σ values of sh1-, sh2-, and sh3-samples are shown in Figure 9. The tensile strength, tensile modulus, and elongation at break of sh1, sh2, and sh3-samples are listed in Table S1 (Supplementary Material). FT-IR spectra confirmed that the chemical structures of sh1-, sh2-, and sh3-samples remained unchanged after repeated self-healing cycles (Figure S4, Supplementary Material). The maximum tensile stress and strain at the break of all the cured products gradually decreased as the number of self-healing cycles increased to varying degrees. The η_σ values of self-healed PG/PEG-CTA samples decreased with an increasing number of healing cycles (i.e., sh1 > sh2 > sh3) and increased with a decreasing ratio of PGPE to PEGDGE (i.e., PG/PEG-1/1 < -2/3 < -1/2). The η_σ of sh1- PG/PEG-1/2-CTA was 100%, and the decrease of η_σ with additional healing cycles was smaller than that of self-healed PG/PEG-1/1-CTA and PG/PEG-2/3-CTA samples, despite PG/PEG-1/2-CTA having the lowest imine and disulfide content. The healing efficiency of this curing process improves with an increase in flexibility and reduced cross-linking density, as these factors facilitate the imine and disulfide metathesis reaction, regardless of the minor differences in imine and disulfide content. Furthermore, the η_σ values of sh1-, sh2-, and sh3-PG/PEG-1/2-DAH samples were significantly lower than those of the corresponding sh1-, sh2-, and sh3-PG/PEG-1/2-CTA samples, underscoring the importance of incorporating of the dual dynamic bonds (imine and disulfide) to achieve superior self-healing properties.

Figure 9.

Changes of tensile stress-strain curves by repeated self-healing at room temperature for 24 h, and the η_σ values for (a) PG/PEG-1/1-CTA, (b) PG/PEG-2/3-CTA, (c) PG/PEG-1/2-CTA, and (d) PG/PEG-1/2-DAH.

Conclusions

Mixtures of PGPE/PEGDGE with epoxy molar ratios of 1/1, 2/3, and 1/2 were cured with imine- and disulfide-containing DVNCTA to produce PG/PEG-CTA-1/1, -2/3, and -1/2 films. For comparison, a mixture of PGPE/PEGDGE with an epoxy molar ratio of 1/2 was cured with only imine-containing DVNDAH to produce PG/PEG-DAH-1/2. This study investigated the effects of the PGPE/PEGDGE ratio and the difference between DVNCTA and DVNDAH as the epoxy hardener on the thermal, mechanical, and repetitive self-healing properties of the bio-based epoxy networks. A decrease in the PGPE to PEGDGE ratio led to reductions in T_g, T_α, ν_e, and mechanical strength of the PG/PEG-CTA films. The T_g, T_α, ν_e, and mechanical strength of PG/PEG-1/2-CTA were lower compared to those of PG/PEG-1/2-DAH. All the cured films demonstrated self-healing capabilities at least three times by simply standing at room temperature. A lower PGPE to PEGDGE ratio resulted in increased η_σ values of the repetitive self-healed PG/PEG-CTA films. Furthermore, the η_σ values of repetitive self-healed PG/PEG-1/2-CTA samples were significantly higher than those of repetitive self-healed PG/PEG-1/2-DAH samples. These results highlight the importance of molecular chain mobility in facilitating the imine and disulfide metathesis reactions, as well as the critical role of dual dynamic bonds in achieving exceptional repetitive self-healing properties. As such, these epoxy networks hold significant potential for applications in sustainable coating materials and small components with excellent self-healing properties.

Supplemental Material

Supplemental Material - Bio-based imine- and disulfide-containing epoxy vitrimers with practical self-healing properties at room temperature

Supplemental Material for Bio-based imine- and disulfide-containing epoxy vitrimers with practical self-healing properties at room temperature by Takashi Yoshimura, Kaito Sugane and Mitsuhiro Shibata in Journal of Polymers from Renewable Resources.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Mitsuhiro Shibata

Supplemental Material

Supplemental material for this article is available online.

References

Jin

Park

. Synthesis and application of epoxy resins: a review. J Ind Eng Chem 2015; 29: 1–11. DOI: 10.1016/j.jiec.2015.03.026.

Chong

Lai

Rahman

, et al. A review on recent approaches to sustainable bio-based epoxy vitrimer from epoxidized vegetable oils. Ind Crops Prod 2022; 189: 115857. DOI: 10.1016/j.indcrop.2022.115857.

Zhang

Shi

, et al. Recyclable, repairable, and malleable bio-based epoxy vitrimers: overview and future prospects. Curr Opin Green Sustainable Chem 2023; 39: 100726. DOI: 10.1016/j.cogsc.2022.100726.

Yang

Liu

Lin

, et al. From biomass to vitrimers: latest developments in the research of lignocellulose, vegetable oil, and naturally-occurring carboxylic acids. Ind Crops Prod 2023; 206: 117690. DOI: 10.1016/j.indcrop.2023.117690.

Sangermano

Bergoglio

Schögl

. Biobased vitrimetric epoxy networks. Macromol Mater Eng 2024; 309: 2300371. DOI: 10.1002/mame.202300371.

Kumar

Krishnan

Prabakaran

. Renewable resource-based epoxy vitrimer composites for future application: a comprehensive review. ACS Sustain Resour Manag 2024; 1: 2086–2107. DOI: 10.1021/acssusresmgt.4c00200.

Kumar

Connal

. Biobased transesterification vitrimers. Macromol Rapid Commun 2023; 44: 2200892. DOI: 10.1002/marc.202200892.

Xia

Gao

, et al. Bio-based epoxy vitrimers with excellent properties of self-healing, recyclability, and welding. Polymers 2024; 16: 2113. DOI: 10.3390/polym16152113.

Xiao

Wang

, et al. Self-healing silicon-containing eugenol-based epoxy resin based on disulfide bond exchange: synthesis and structure-property relationships. Polymer 2021; 229: 123967. DOI: 10.1016/j.polymer.2021.123967.

10.

Shan

Mai

Lin

, et al. Self-healing, reprocessable, and degradable bio-based epoxy elastomer bearing aromatic disulfide bonds and its application in strain sensors. ACS Appl Polym Mater 2021; 3: 5115–5124. DOI: 10.1021/acsapm.1c00865.

11.

Trtnik

Tanguy

Yan

. Recyclable, self-strengthening starch-based epoxy vitrimer facilitated by exchangeable disulfide bonds. Chem Eng J 2023; 451: 138610. DOI: 10.1016/j.cej.2022.138610.

12.

Liu

Liang

, et al. Water-resistant bio-based vitrimers based on dynamic imine bonds: self-healability, remodelability and ecofriendly recyclability. Polymer 2020; 210: 123030. DOI: 10.1016/j.polymer.2020.123030.

13.

Liu

, et al. Biobased high-performance epoxy vitrimer with UV shielding for recyclable carbon fiber reinforced composites. ACS Sustainable Chem Eng 2021; 9: 4638–4647. DOI: 10.1021/acssuschemeng.1c00231.

14.

Liu

Zhang

Feng

, et al. Degradable bio-based epoxy vitrimers based on imine chemistry and their application in recyclable carbon fiber composites. J Mater Sci 2021; 56: 15733–15751. DOI: 10.1007/s10853-021-06291-5.

15.

Roig

Hidalgo

Ramis

, et al. Vitrimeric epoxy–amine polyimine networks based on a renewable vanillin derivative. ACS Appl Polym Mater 2022; 4: 9341–9350. DOI: 10.1021/acsapm.2c01604.

16.

Zych

Tellers

Bertolacci

, et al. Biobased, biodegradable, self-healing boronic ester vitrimers from epoxidized soybean oil acrylate. ACS Appl Polym Mater 2021; 3: 1135–1144. DOI: 10.1021/acsapm.0c01335.

17.

Sangaletti

Ceseracciu

Marini

, et al. Biobased boronic ester vitrimer resin from epoxidized linseed oil for recyclable carbon fiber composites. Resour Conserv Recycl 2023; 198: 107205. DOI: 10.1016/j.resconrec.2023.107205.

18.

Khalafi

Ehsani

Khonakdar

. Investigation of the cure kinetics and thermal stability of an epoxy system containing cystamine as curing agent. Polym Adv Technol 2020; 32: 1251–1261. DOI: 10.1002/pat.5174.

19.

Roig

Agizza

Serra

, et al. Disulfide vitrimeric materials based on cystamine and diepoxy eugenol as bio-based monomers. Eur Polym J 2023; 194: 112185. DOI: 10.1016/j.eurpolymj.2023.112185.

20.

Guggari

Magliozzi

Malburet

, et al. Vanillin- based epoxy vitrimers: looking at the cystamine hardener from a different perspective. ACS Sustainable Chem Eng 2023; 11: 6021–6031. DOI: 10.1021/acssuschemeng.3c00379.

21.

Ogawa

Sugane

Shibata

. Self-healing amine- and carboxy-cured bio-based epoxy vitrimers driven by the disulfide metathesis reaction. Polym Bull 2025; 82: 625–647. DOI: 10.1007/s00289-024-05538-w.

22.

Fache

Boutevin

Caillol

. Vanillin, a key-intermediate of biobased polymers. Eur Polym J 2015; 68: 488–502. DOI: 10.1016/j.eurpolymj.2015.03.050.

23.

Qiang

Wang

Liu

, et al. From vanillin to biobased aromatic polymers. Polym Chem 2023; 14: 4255–4274. DOI: 10.1039/d3py00767g.

24.

Liguori

Hakkarainen

. Designed from biobased materials for recycling: imine-based covalent adaptable networks. Macromol Rapid Commun 2022; 43: 2100816. DOI: 10.1002/marc.202100816.

25.

Wang

Xie

, et al. Extrudable, robust and recyclable bio-based epoxy vitrimer via tailoring the topology of a dual dynamic-covalent-bond network. Polymer 2023; 289: 126487. DOI: 10.1016/j.polymer.2023.126487.

26.

Guggari

Magliozzi

Malburet

, et al. Vanillin-based dual dynamic epoxy building block: a promising accelerator for disulfide vitrimers. Polym Chem 2024; 15: 1347–1357. DOI: 10.1039/d4py00038b.

27.

Verdugo

Santiago

De la Flor

, et al. A biobased epoxy vitrimer with dual relaxation mechanism: a promising material for renewable, reusable, and recyclable adhesives and composites. ACS Sustainable Chem Eng 2024; 12: 5965–5978. DOI: 10.1021/acssuschemeng.4c00205.

28.

Abe

Shibata

. Bio-based self-healing epoxy vitrimers with dynamic imine and disulfide bonds derived from vanillin, cystamine, and dimer diamine. Molecules 2024; 29: 4839. DOI: 10.3390/molecules29204839.

29.

Hamid

ZAA

Blencowe

Ozcelik

, et al. Epoxy-amine synthesised hydrogel scaffolds for soft-tissue engineering. Biomaterials 2010; 31: 6454–6467. DOI: 10.1016/j.biomaterials.2010.05.008.

30.

Hill

. Calculation of crosslink density in short chain networks. Prog Org Coating 1997; 31: 235–243. DOI: 10.1016/S0300-9440(97)00081-7.

31.

Echeverria-Altuna

Ollo

Larraza

, et al. Effect of the biobased polyols chemical structure on high performance thermoset polyurethane properties. Polymer 2022; 263: 125515. DOI: 10.1016/j.polymer.2022.125515.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.82 MB