2.1 Reagents

The subject of the study was polycaprolactone: M_n 70,000–90,000, < 1.0% water, 1.145 g/mL at 25 °C (Aldrich). Epoxidised natural rubber (epoxyprene 50; 50 mol % epoxidation) was obtained from Kumpulan Guthrie Berhad, Malaysia. A mixture of amino acids (98% creatine, 98% cysteine, 98% phenyloalanine), 97% dodecanoic acid, and 97% quercetin from flavonoids, obtained from Sigma–Aldrich, was used as a cross-linking agent; 1,2-dimethylimidazole (98%) was used as a cross-linking catalyst. Aerosil 380 silica was obtained from Degussa (Table 1).

2.2 Measurement methods

Rubber blends were prepared using a laboratory mixer of Brabender with a front-roll rotation speed of V_p = 60 rpm and friction of 1.1. The average temperature in the Brabender mixer was approximately 180 °C, and the mixing time was 60 min.

The vulcanisation of rubber blends was carried out with the use of steel vulcanisation moulds placed between the shelves of an electrically heated hydraulic press. Teflon films were used as spacers to prevent the adherence of the blends to the press plates. Samples were vulcanised at a temperature of 160 °C under a pressure of 15 MPa for 30 min.

The density of cross-links in the vulcanisates network was determined by the method of equilibrium swelling according to standard PN-74/C-04236. The vulcanisates were subjected to equilibrium swelling in hexane for 48 h at room temperature. The swollen samples were then weighed on a torsion balance and dried in a dryer at a temperature of 60 °C to a constant weight, and after 48 h they were reweighed. The cross-linking density was determined based on Flory–Rehner's equation:

The Huggins parameter of the polymer-solvent interaction (χ) was calculated from the equation:

The tensile strength of vulcanisates was tested according to standard PN-ISO 37:1998 by means of a ZWICK tester, model 1435, for dumbbell w-3.

Ageing characteristics were determined according to standard PN-82/C-04216. Samples were exposed to air at elevated temperature (383 K) for 10 days in a dryer with thermo-circulation. UV ageing was performed using a UV 2000 apparatus from Atlas. The measurement lasted 288 h and consisted of two alternately repeating segments with the following parameters: daily segment (radiation intensity 0.7 W m⁻², temperature 60 °C, duration 8 h), night segment (no UV radiation, temperature 50 °C, duration 4 h). Climatic ageing was carried out using a Weather-Ometer (Atlas; Ci 4000). The test was based on two variable segments simulating day and night conditions, and the samples were subjected to two different cycles: daily cycle (radiation intensity 0.4 W m⁻², temperature 60 °C, duration 240 min, humidity 80%, rain water on), night cycle (no radiation, temperature 50 °C, humidity 60%, duration 120 min).

The ageing coefficient was calculated according to the relationship:

The dispersion degree of the fillers in the elastomeric matrix was assessed from images obtained using a LEO 1530 scanning electron microscope (SEM). The samples tested consisted of vulcanisate fractures in liquid nitrogen dusted with carbon.

The oxygen induction time (OIT) test was performed on a Mettler Toledo DSC instrument. Samples with a mass of 4 mg were heated from room temperature to the test temperature, 220 °C, at a rate of 20 °C min⁻¹ under a nitrogen atmosphere. After 5 min at 220 °C, the gas was switched from nitrogen to air at a flow rate of 60 ml·min⁻¹. When all the antioxidants are consumed, the sample starts to oxidise, producing a deviation in the baseline. The OIT was measured as the time between the switch to airflow and the intersection with a tangent from the maximum derivative after oxidation has started. Two analyses of each sample were performed to ensure the accuracy of the results. The OIT was measured according to national and international standards such as ASTM D 3895.

The crystallized specimens were characterized by using differential scanning calorimeter (DSC) and wide-angle X-ray diffraction (WAXD).

WAXD analyses were performed for WX1 and PLAWX1 on an Mxlabo diffractometer (MAC Science Co.; 3 kW, graphite monochromator, Cu Kα radiation (λ_x = 0.154 nm), operated at 40 kV and 20 mA). Samples were scanned in fixed time mode with counting time of 2 s under a diffraction angle 2θ in the range of 5–60°.

The crystallized specimens were characterized by using temperature-modulated DSC (TA 2920; TA Instruments) at the heating rate of 5 °C min⁻¹ with a heating/cooling cycle of the modulation period of 60 s and an amplitude of ± 0.769 °C, to determine the glass transition (Tg) crystallization temperature (T_c) and the melting temperature (T_m). The DSC was calibrated with indium before use. For the measurement of T_c, we used amorphous samples after quenching from malt state (∼200 °C).

3.1 The properties of ENR/PCL composites

The presence of an oxyrane ring in the structure of natural rubber creates opportunities for various reactions to proceed for modification purposes. In polymers containing epoxide groups, cross-linking is carried out using compounds containing amine groups. The cross-linking of composites of epoxidised natural rubber and polycaprolactone was performed using amino acid derivatives (Scheme 1). The choice of the cross-linkers was also dictated by the pro-ecological character of amino acids that are widely used in medicine. The amino acids used had carboxyl and amine groups. We also used fatty acids (dodecane acid). The primary aim was to obtain the best possible elastic properties in ENR/PCL composites. The best cross-linking effects were obtained using quercetin, a natural derivative of flavonoids. The composite tensile strength was 22.50 MPa, the elongation was 436%, and the cross-link density was 21.15·10⁻⁵ mol cm⁻³. Amino acids such as phenylalanine and asparagine or cysteine, used as natural cross-linking agents, yielded decreased cross-linking effectiveness. The best cross-linker was phenylalanine. The composite cross-linked with phenylalanine had a tensile strength of 13.9 MPa, an elongation of 443%, and a cross-link density of 8.89·10⁻⁵ mol cm⁻³. However, the use of asparagine, cysteine and dodecane acid in the ENR/PCL composite allowed us to obtain a lower elongation at break (≤ 350%) and a tensile strength ranging from 13.8 to 14.92 MPa. Undoubtedly, the use of a flavonoid as a cross-linker constitutes an undeniable element of scientific novelty. Quercetin and phenyloalanine were the best cross-linkers for the biocomposites we developed. The pro-ecological matrix of these composites makes them environmentally friendly and human-friendly and, therefore, widely applicable in many fields. Based on the positive results concerning the action of quercetin as a cross-linker, the mechanical properties of composites cross-linked with various quantities of quercetin were determined (Table 2). From the results listed in Table 2, it follows that the optimal content of quercetin in the composite ranges from 0.75 to 1 phr per 50 phr of ENR/PCL. When the content of quercetin used for cross-linking amounted to 0.75–1.5 phr, the composite elongation at break ranged from 18.4 to 19.20 MPa, and with increasing quercetin content, the elongation at break decreased to the minimum value of 250% at a quercetin content of 1.5 phr. The cross-link density of the composite slightly changed with increasing quercetin content, ranging from 19.3 to 24.70·10⁻⁵ mol cm⁻³ (Table 3).

3.2 Effect of cross-linking agents on the stabilisation of biocomposites

Natural cross-linkers such as amino acids, fatty acids and quercetin, from the group of flavonoids, can also be used as anti-ageing compounds. Therefore, the aim of the tests concerning the influence of simulated ageing by various climatic conditions on the properties of biocomposites was to assess their resistance to those conditions.

Biocomposites cross-linked with quercetin and asparagine exhibited greater resistance to UV ageing than the other biocomposites tested, as confirmed by the high ageing coefficients measured: 0.72 for the composite containing quercetin and 0.82 for the composite containing asparagine. The other ENR/PCL composites exhibited UV ageing resistance similar to that of ENR vulcanisate. As in the case of UV ageing (K = 0.52), the sample of ENR/PCL containing phenylalanine was degraded to the greatest extent by thermo-oxidising (K = 0.31) and climatic ageing (K = 0.52). The ENR/PCL biocomposites exhibited lower resistance to thermal ageing than ENR vulcanisate (K = 0.87), with ageing coefficients that ranged from 0.31 to 0.87. The samples tested showed good resistance to climatic ageing, as confirmed by their ageing coefficients, which approached 1. Among all of the biocomposites, the samples cross-linked with asparagine and quercetin were the most resistant to climatic ageing; their ageing coefficients were K = 0.82 and K = 0.72, respectively (Table 4).

Additionally, DSC examinations were conducted to determine the oxygen induction time (OIT), expressed by the OIT index. It appears that the OIT of the ENR/PCL biocomposite cross-linked with dodecane acid began at a temperature of 410 °C and amounted to 0.32 min. The energy of this process amounted to 935 J·g⁻¹, which is similar to the energy value obtained by the oxygen induction process of the ENR/PCL composite cross-linked with quercetin. The OIT index of this composite was 3.35 min. However, the vulcanisate of ENR rubber had an OIT index equal to 1.36 min, and the energy of his process amounted to 2721 J·g⁻¹. It should be emphasised that the oxidation of this sample began at 210 °C. ENR/PCL biocomposites are characterised by low oxidation energy, although their oxidation energy was comparable in the case of cross-linking with quercetin and lower in the case of the remaining samples. However, increasing the quercetin content did not significantly influence the oxygen induction time: at a content of 0.75 phr, the oxidation was more energetic than that of samples with higher quercetin contents.

3.3 The crystallized specimens of biocomposites ENR/PCL

To assess the crystallinity of the biocomposites designed, we used wide-angle X-ray diffraction (WAX) and DSC methods. The results of the WAX measurements indicate a lower crystalline content in the ENR/PCL composite than in polycaprolactone (Fig. 1). The diffractometric diagram of polycaprolactone shows two peaks located at angles 2θ ≈ 21.45 and 2θ ≈ 23.7, indicating a high degree of crystallinity in this polymer. However, the diffractometric diagram of ENR/PCL composite only shows one broad reflection at 2θ ≈ 10 to 2θ ≈ 22, with a much lower intensity than that of PCL. The addition of amorphous epoxidised natural rubber to the crystalline polycaprolactone can cause a considerable decrease in the crystalline phase of such a composite. The crystalline phase decay in the ENR/PCL composite was confirmed by the DSC diagram, showing an intensive peak of PLC at approximately 71.11 °C, which was considerably decreased in the ENR/PCL composite at 63.13 °C (Fig. 2). The crystallization enthalpy is for PCL ΔH_c = 71.11 J·g⁻¹ and for ENR/PCL composite ΔH_c = 24.30 J·g⁻¹ (Table 5). This crystalline phase decay indicates good miscibility and compatibility of these polymers, thus yielding a composite with very good properties.

The compatibility and good miscibility of epoxidised natural rubber with polycaprolactone were confirmed by the images obtained using the SEM technique. These images show that the composite has a compact uniform structure containing evenly dispersed fillers (Figs. 3 and 4). Because of the different entropy and enthalpy among polymers, phase separation generally occurs in most polymer blends and leads to a multiphase morphology. The properties are dependent not only on the physical properties of each component but also on the dispersed phase microstructure and the interfacial chemistry. Figs. 3 and 4 show SEM micrographs of fracture surfaces of ENR/PCL blends. In our work, the ENR/PCL blends revealed a spherical-particle-dispersed morphology, because the non-polar ENR was immiscible with the polar PCL and led to phase separation. That ENR with higher degree of epoxidation can induce more molecular interactions and entanglements with PCL because polarity of the ENR increases with epoxidation content. During processing, the components are melt-blended, and the minor phase is broken up to form the dispersed phase. Fig. 4 shows the SEM micrographs of freeze-fractured surfaces of the ENR/PCL blend. This SEM image presents scaffolds with interconnected porous structures of ENR/PCL blend. ENR/PCL blend (Fig. 3) presents quite coarse phase morphology with small amount of elongated rubber domains. Fig. 3 shows a compatibility increase in ENR/PLA blends, which is indicated by an obscure interphase. No visible agglomerates or irregularities can be observed, and there is a clear isolation of one polymer phase.