Preparation of Nanoscale Semi-IPNs with an Interconnected Microporous Structure via Cationic Polymerization of Bio-Based Tung Oil in a Homogeneous Solution of Poly(ε-caprolactone).

Nanoscale semi-interpenetrating polymer networks of bio-based poly(ε-caprolactone) (PCL) and polymerized tung oil have been prepared via in situ cationic polymerization and compatibilization in a homogeneous solution. This novel blending technique produced a nanoscale morphology of poly(ε-caprolactone) with average particle sizes as small as 100 nm dispersed in a cross-linked tung oil matrix for 20 and 30 wt % PCL blend compositions. In addition, the exothermic cationic polymerization of tung oil in the presence of the PCL homogeneous solution created a microporous morphology with open three-dimensional interconnected cluster structures. The porous morphology was found to be composition-dependent (the pore size and interconnectivity decreased with increasing PCL content in the blend). The values of the cross-link density and storage modulus in the glassy state for fully cured samples increased significantly and reached a maximum for the 20 wt % PCL blend. This simple, versatile, low-cost strategy for preparing nanoscale and interconnected three-dimensional cluster structures with a microporous morphology and desired properties should be widely applicable for new polymer systems.

Introduction

The societal drive to achieve economic and environmental sustainability has led to an increased emphasis on the use of alternative environmentally friendly, bio-based products from biorenewable resources or biomass.[1−7] Fabrication of new classes of engineering materials with outstanding mechanical, thermal, dielectric, and biological properties from bio-based sustainable products has received considerable attention recently.[8−15] Plant oils are the most abundant and cost-effective biorenewable resources worldwide.[16−22] They have been widely used to synthesize bio-based polymers with no toxicity and inherent biodegradability.[8−26] Saturated and unsaturated fatty acids are the main components of triglyceride vegetable oils that are the platform chemicals for polymer synthesis. Seeds of the tung tree have large quantities of tung oil, which contains approximately 84% α-eleostearic acid triglyceride with a large number of conjugated C=C bonds.[27] Tung oil has been used in many industrial applications, such as paints, coatings, varnishes, and related applications. The conjugated C=C bonds of tung oil are reactive and can be easily polymerized thermally, free-radically, and cationically to produce bio-based thermosetting polymers with different properties.[28−30] Heating tung oil at high temperatures up to 300 °C with no catalyst leads to dimerization into a weak rubbery material. The cationic polymerization of tung oil using boron trifluoride diethyl etherate (BFE) as an initiator causes an aggressive exothermic reaction and commonly produces a very brittle, dark, thermosetting material with very poor mechanical properties. Free-radical copolymerization of tung oil with styrene and divinylbenzene at 85–160 °C with different stoichiometry, oxygen uptake, peroxides, and metallic catalysts was found to be an effective method to greatly improve the mechanical properties of tung oil.[16] This copolymerization process was successfully employed to obtain thermosetting copolymers with a wide range of desirable mechanical properties, ranging from rubbery to tough and rigid plastics. Regardless of the significant improvements in the material properties of tung oil copolymers, both styrene and divinylbenzene are nondegradable petroleum-based products.

Poly(ε-caprolactone) (PCL) is a biodegradable and biocompatible thermoplastic semicrystalline polyester with a glass-transition temperature (Tg) of approximately −60 °C and a melting point of about 60 °C.[31,32] Due to the biocompatibility and biodegradability of PCL in physiological media via hydrolysis of its ester linkages, it is widely used in many biomedical applications, such as sutures, adhesion barriers, scaffolds for tissue engineering, and long-term implants.[33,34] Numerous studies have been carried out to modify PCL with large numbers of different thermoplastics materials, such as poly(styrene-co-acrylonitrile) (SAN), tetramethyl polycarbonate (TMPC), poly(lactic acid) (PLA), poly(methyl methacrylate) (PMMA), polycarbonate (PC), and poly(vinyl chloride) (PVC).[35−41]

Compared to the synthesis of new materials for specific industrial applications, polymer blending is an easy way to mix two or more polymers of different properties to produce new materials with diverse, tailored properties compared to those of the pure polymer components. Most polymer blends are immiscible and require a certain type of compatibilizer to improve the interfacial interaction between the two polymer components and reduce the particle size. Phase-separated polymer blends with large particle size and poor interfacial adhesion are excluded from many industrial applications. Compatibilizers, such as block, graft, or star copolymers, are commonly employed to improve the compatibility and increase the thermodynamic stability of the immiscible polymer components. Most of these compatibilizers are expensive and not available commercially. Therefore, the polymerization of monomers in the presence of a homogeneous solution of another thermoplastic polymer could be an excellent approach for fabrication of compatible polymer blends with a nanoscale morphology and improved properties without using any of the expensive compatibilizers.

In this work, cationic polymerization of tung oil in a homogeneous solution of PCL and chloroform was carried out at room temperature to create a nanoscale morphology of semi-interpenetrating polymer networks and a microporous morphology with open, three-dimensional interconnected cluster structures. Using chloroform in this cationic polymerization is very important and a key parameter for the success of this work. Generally, the cationic polymerization of tung oil is very aggressive at room temperature and produces a dark and brittle material with very poor mechanical properties even with a very small concentration of the cationic initiator (less than 0.1 wt % of BFE). This extremely fast cationic polymerization can be inhibited dramatically by diluting the concentration of BFE in chloroform. In addition, chloroform is an excellent solvent for PCL, tung oil, and BFE at room temperature.

The concentration of BFE was kept constant at 2 wt % relative to the concentration of tung oil in chloroform for all blend compositions. The cationic polymerization of tung oil started immediately after mixing tung oil with PCL and BFE in chloroform (25% solid content) and produced transparent, yellow gels. The obtained gels were left to dry for about three days under vacuum at room temperature and for another two days at 60 °C to evaporate all residual solvent. The evaporated chloroform was recycled with a cold trap and reused in the blend preparation process.

Results and Discussion

Figure shows the scheme for cationic polymerization of pure tung oil in chloroform (25 wt % tung oil) at room temperature using 2 wt % BFE initiator. The figure also shows the Fourier transform infrared (FT-IR) spectra for pure, unreacted and fully cured tung oil after cationic polymerization in chloroform. FT-IR spectroscopy is employed to determine the C=C bonds of tung oil before and after the cationic polymerization. The FT-IR spectrum of uncured tung oil has C–H stretching bonds at 3013 cm–1. This peak can be used for monitoring the cationic polymerization of tung oil. For fully cured tung oil, the 3013 cm–1 peak totally disappears, indicating that there is no remaining C=C bonds in the fully cured sample, as clearly seen in the figure. The dynamic mechanical analysis (DMA) measurement for fully cured and dried tung oil thermoset is also demonstrated in Figure . The DMA measurement was investigated in three-point bending mode at 1 Hz and 2 °C/min heating rate. It is apparent that tung oil can be cationically polymerized in chloroform solution to produce a thermoset of bio-based tung oil polymer with storage modulus in the glassy state of about 1000 MPa and Tg of about 33 °C (calculated from the temperature of the peak maximum of tan δ as seen in Figure ).

Figure 1

Tung seeds and tung oil. Scheme for cationic polymerization of tung oil using BFE initiator. FT-IR spectra for uncured (A) and fully cured (B) tung oil thermoset. DMA measurement for fully cured tung oil thermoset obtained from cationic polymerization. Photograph courtesy of The Wood Works Book Tool Co. Free domain: https://www.tungoil.com.au/.

The curing kinetics of the cationic polymerization of tung oil in a homogeneous solution of PCL and chloroform was investigated using dynamic rheology by monitoring the change in viscoelastic material functions (G′, G″, η*, and tan δ) as a function of curing time and composition. For example, the dynamic rheology for the cationic polymerization of the tung oil/PCL 70/30 wt % blend in chloroform (25 wt % solid content) is demonstrated in Figure . This figure shows the curing time dependence of G′, G″, η*, and tan δ at 20 °C and 1 rad/s angular frequency. At the beginning of the cationic polymerization reaction and before the gelation point, G′ is much lower than G″ (i.e., the value of G′ is more than one order of magnitude lower than G″ before the gelation point). In addition, the value of η* is also very low and tan δ has a maximum value before the gelation point. With increasing curing time, G′, G″, and η* increase significantly, while tan δ decreases due to the formation of a cross-linked network structure of tung oil via a cationic polymerization reaction. It is also obvious that G′ increases more rapidly than G″, and the gel point (tgel) can be obtained from the crossover point of G′ and G″ (approximately 50 min, as seen by the arrow in Figure ). At higher curing times, higher than 80 min, the values of G′, G″, and η* leveled off and became almost constant regardless of the increase in curing time due to the formation of an equilibrium three-dimensional network of thermoset tung oil. The value of tan δ also reached a very low and constant value at the end of the curing process. The inset schematic diagram of Figure depicts how the thermoplastic PCL mixed and physically interacted with the tung oil during its cationic polymerization in a homogeneous solution of chloroform. The schematic diagram shows the entangled chains of PCL (blue) with cross-linked tung oil (green) as semi-interpenetrating polymer networks. Similar behavior was observed for other compositions, and tgel was found to increase with decreasing concentration of tung oil in the blends at a constant temperature and angular frequency, as seen in Table .

Figure 2

Curing time dependence of G′, G″, η*, and tan δ for the tung oil/PCL 70/30 wt % blend (25 wt % solid content in chloroform) at 20 °C and 1 rad/s angular frequency. The arrow shows the value of tgel. The schematic diagram shows the entangled PCL chains and tung oil before the cationic reaction and how they are mixed to create semi-interpenetrating polymer networks (blue entangled PCL chains and green cross-linked tung oil).

Table 1

Characterization of Tung Oil/PCL Blends

PCL (wt %)	E′ (MPa)	νe (mol/m3)	Tgel (min)	morphology
0	1212	44.2	1.1	one phase
10	1423	71.8	11.9	one phase
20	1832	373	30.7	nanoparticles/microporous
30	1501	25	49.5	nanoparticles/microporous
50	827	20	60.3	co-continuous
100	688			one phase

The temperature dependence of E′ for fully cured and dried tung oil/PCL blends of different compositions is shown in Figure (DMA measurements). Clearly, the value of E′ is both temperature- and composition-dependent. In a low-temperature range (−100 to 0 °C, glassy state), E′ is very high and reaches almost a maximum value for 20 wt % PCL. The maximum increase in E′ might be related to a special morphology or high cross-linking density for this composition. In an intermediate-temperature range (0–50 °C), E′ decreases dramatically due to the glass relaxation process of the different blends. In a high-temperature range (50–100 °C), E′ becomes temperature-independent and reaches a plateau for different blend compositions. For the 30 and 50 wt % PCL blends, E′ decreases significantly again at about 70 °C due to the melting of the PCL-rich phase in the blends. The melting of PCL in the blends with lower PCL compositions (10 and 20 wt % PCL) has no significant effect on E′ at 70 °C. The composition dependence of E′ at −100 °C (in the glassy state) is demonstrated in the inset plot of Figure . Clearly, E′ reaches a maximum value for the 20 wt % PCL blend, as mentioned above. This might be related to the high cross-linking density of this blend. At a high temperature (e.g., 50 °C above Tg), the plateau of E′ (see Figure ) can be used to calculate the cross-linking density (νe) of the fully cured samples based on the kinetic theory of rubber elasticity, E′ = 3νeRT.[42,43] For the fully cured tung oil, the cross-linking density is approximately νe = 44.2 mol/m3. The value of νe is found to be composition-dependent, i.e., it attains a maximum value (νe = 318 mol/m3) for the 20 wt % PCL blend, in good agreement with the composition dependence of E′, as clearly seen in Table and the inset plot of Figure .

Figure 3

Temperature dependence of the storage modulus of fully cured tung oil/PCL blends. The inset plot shows the composition dependence of both storage modulus at −100 °C and cross-linking density calculated from the kinetic theory of rubber elasticity.

The tan δ obtained from DMA measurements can be employed to investigate the miscibility of the tung oil/PCL blends. Figure shows the temperature dependence of tan δ for different blend compositions (20, 30, and 50 wt % PCL blends). Clearly, two glass relaxation processes are observed for the tung-oil-rich and PCL-rich phases in the high- and low-temperature ranges, respectively. A very sharp tan δ peak is detected for the tung-oil-rich phase, while a very broad peak with reduced height is observed for the PCL-rich phase. It appears that the tung oil/PCL blends (20–50 wt % PCL) are immiscible and phase-separated into micro/nanomorphologies based on the different blend compositions. The blends with 20 and 30 wt % PCL have nanoscale morphologies with an average particle size of 100 nm, while the 50 wt % blend has a co-continuous interconnected phase-separated morphology, as clearly seen in the inset plots of Figure a–c.

Figure 4

Temperature dependence of tan δ for fully cured tung oil/PCL blends with different compositions: (a) 20 wt % PCL blend, (b) 30 wt % PCL blend, and (c) 50 wt % PCL blend. The inset plots show the morphologies of the blends for different compositions. Blends with 20 and 30 wt % PCL have a bright PCL dispersed phase with approximately 50 nm in the dark tung oil matrix, while the 50 wt % PCL blend has a co-continuous interconnected phase-separated morphology.

Nano/microscale morphologies could be obtained by controlling the polymerization conditions and composition of the system, i.e., the content of PCL, and an appropriate choice of the initiator and use of organic solvent, specific for this system and not for any other polymer blends. We used the term compatibilization because the cationic polymerization of tung oil and the mixing process with PCL solution were carried out simultaneously. Hence, the cationic polymerization process of tung oil starts, while tung oil, PCL, and the cationic initiator are totally miscible in chloroform. Once the cationic polymerization process starts, the whole system is converted into a solid gel instantaneously. In the solid gel, the phase-separation process due to an increase in molecular weight (i.e., polymerization-induced phase separation (PIPS)) will be minimized. Therefore, this system is different from other epoxy/PCL systems due to the fact that epoxy/PCL blends are commonly controlled by PIPS.[44−46] In the case of PIPS, the phase separation is carried out in the melt at high temperatures for an extended curing time. Hence, the phase separation is related to the increase in the molecular weight, and consequently, the Gibbs free energy of mixing will be positive and the blend will be phase separated. In the current system, the organic solvent will be trapped inside the blend structure. Hence, the phase separation in the solid blend will be very minimized and unique nano/microscale phase-separated morphologies are created. The blends are compatibilized without using any compatibilizers, such as block or graft copolymers. The phase-separation process and the obtained nano/micromorphologies of the current system are unique to this relatively new technique of mixing. More details about the blend morphology will be provided in the following section.

The morphologies of the fully cured and dried blends were investigated using scanning electron microscopy (SEM). The fully cured blends were fractured in liquid nitrogen and then sputtered with gold. The SEM morphology of all blends was investigated using a field emission scanning electron microscope (FE-SEM, FEI Quanta 250) operating at 10 kV under high vacuum. Figure shows the SEM morphologies for tung oil/PCL blends with 20 and 30 wt % PCL. Clearly, the two blend compositions have a fine porous structure. With high magnification, nanoscale morphologies of PCL bright nanoparticles in the dark tung oil matrix were detected. The average nanoscale particle size of the dispersed phase (PCL particles) was about 100 nm. With a higher concentration of PCL, no nanoscale morphology was observed. In addition, for blends with 10 wt % PCL or lower PCL content, no morphology was observed and completely miscible blends were obtained.

Figure 5

SEM morphologies for fully cured tung oil/PCL blends with 20 and 30 wt % PCL at different magnifications ((a) and (b) for 20 wt % PCL and (c) and (d) for 30 wt % PCL).

The porous structure of these blends might be related to the evaporation of the trapped solvent during the drying process. More details about the porous structure can be seen in Figure . One can see that the three-dimensional pore size decreased with increasing PCL content in the blends. In addition, the tung oil formed interconnected, clustered particles. The size of the interconnected, clustered particles for the blend with 30 wt % PCL is much smaller than that of the blend with 20 wt % PCL.

Figure 6

SEM morphologies for fully cured porous tung oil/PCL blends with 20 and 30 wt % PCL at different magnifications (a–c).

For the blend of 50 wt % PCL, no nanomorphology or porous structures were observed. Figure demonstrates a comparison between the porous structures of different blend compositions. It appears that the tung oil and the PCL phase separated into a co-continuous structure for the 50/50 wt % blend with no porous structure, as seen in Figure c (the bright PCL is co-continuous with the dark tung oil).

Figure 7

SEM morphologies of porous tung oil/PCL blends with different compositions: (a) tung oil/PCL 80/20 wt % blend, (b) tung oil/PCL 70/30 wt % blend, and (c) tung oil/PCL 50/50 wt % blend.

The porous structure of bio-based materials is necessary for many applications. A blowing agent and an additional processing step are required to create a foam or porous structure. The organic solvent in the current system has many important functions, such as reducing the reactivity of the cationic initiator, enhancing the miscibility of tung oil with PCL, and producing a porous structure. The trapped organic solvent inside the blend structure after the gelation process will be evaporated to form a unique three-dimensional interconnected porous structure that cannot be produced using the traditional foaming techniques. These materials could find potential applications in the biomedical field if they are biocompatible and if the solvent is totally evaporated without any residue. Additional potential applications, e.g., as an adsorbent or a support for a specific catalyst, might also be considered for the current materials.

Conclusions

Semi-interpenetrating polymer networks of bio-based tung oil/PCL blends were prepared via cationic polymerization of tung oil in a homogeneous solution of PCL. Co-occurrence of a nanoscale morphology and interconnected, clustered microporous structures was observed for 20 and 30 wt % PCL blends. For blends with higher PCL content (e.g., 50 wt % PCL), neither nanoscale nor interconnected, clustered microporous structures were detected and only a highly interconnected co-continuous morphology was observed. Both the storage modulus in the glassy state and the cross-link density of fully cured blends were found to be blend-composition-dependent, i.e., they increase to reach maximum values at 20 wt % PCL. The kinetics of cationic polymerization of tung oil in PCL homogeneous solution was inhibited greatly by increasing the concentration of PCL and the organic solvent in the blend. This simple, versatile, and low-cost cationic polymerization of tung oil in a homogeneous solution of PCL to prepare nanoscale and interconnected, co-continuous morphologies with desired properties should be widely applicable.

Experimental Section

Materials

The PCL used in this work was provided by Union Carbide Corporation (PCL-767). The weight-average molecular weight and polydispersity of PCL are 40 400 g/mol and 2.61, respectively.

Tung oil, boron trifluoride diethyl etherate (BFE) (cationic initiator), and chloroform were obtained from Sigma-Aldrich and were used as received. Tung oil is a low-viscosity yellow liquid with a specific gravity of 0.937 at 25 °C. The chemical component of tung oil is α-eleostearic fatty acid (cis-9-, trans-11-, trans-13-octadecatrienoic acid).

Sample Preparation

A homogeneous solution of PCL, tung oil, and BFE was prepared in chloroform at room temperature with 25 wt % solid content. The content of BFE was kept constant at 2 wt % for different blend concentrations with respect to the concentration of tung oil. The cationic polymerization of tung oil in the homogeneous solution of PCL generated transparent, yellow gels for all blends after different curing times based on the concentration of tung oil in the blend. The higher the tung oil concentration, the shorter the curing time to obtain the yellow gel. The homogeneous solution changed into the yellow gel without any significant reduction in volume. The obtained gels were left to dry at room temperature in a vacuum oven for three days. Complete drying of the gels was accomplished at 60 °C in another two days in the vacuum oven. The evaporated chloroform was recycled using a cold trap and reused in the blend preparation process. The exothermic and extremely aggressive cationic polymerization reaction of tung oil was inhibited dramatically in the homogeneous solution of PCL and chloroform. In addition, chloroform was necessary to dissolve PCL and BFE.

Characterization of the Blends

FT-IR Spectral Analysis

The cationic polymerization of tung oil was confirmed by the evaluation of C=C bonds before and after polymerization using FT-IR analysis. Homogeneous solutions of tung oil and PCL in chloroform with and without the cationic initiator were placed on a KBr salt plate. The FT-IR spectra of the blend were recorded on a Bruker IFS-66V spectrometer (Billerica, MA) after evaporation of chloroform. The completely reacted tung oil/PCL 70/30 wt % blend was crushed into powder in liquid nitrogen, and the powder/KBr mixture was compressed into a plate, which was characterized with the same spectrometer.

Rheological Measurements

Curing kinetics of cationic polymerization of tung oil in a homogeneous solution of PCL and chloroform was investigated using an AR200ex rheometer (TA Instruments) with 25 mm diameter parallel plates. The air/sample interface was covered with a thin layer of low-viscosity silicone oil to prevent the evaporation of chloroform during the curing process. In this study, time sweeps at different constant angular shear frequencies (ω = 1–20 rad/s) at 20 °C in the linear viscoelastic regime (at 2% strain) were carried out to evaluate the gelation process of the cationic polymerization of tung oil and its blends with PCL. Strain sweep at a constant temperature and frequency was also carried out to obtain the linear viscoelastic range.

DMA Measurements

A dynamic mechanical analyzer (DMA, Q800) from TA Instruments was used to investigate the thermomechanical properties of the fully cured and dried tung oil/PCL blends in three-point bending mode. Rectangular samples with approximately 0.9 mm thickness, 7 mm width, and 12 mm length were heated from −100 to 150 °C at a heating rate of 2 °C/min. A frequency of 1 Hz and displacement amplitude of 5 μm were used in all DMA measurements. The storage modulus in the glass state at −100 °C was evaluated as a function of blend composition from the DMA measurements. The cross-link density of the blends was also determined from the plateau modulus at 50 °C above the Tg of each composition based on the rubber elasticity theory.

Morphology of the Blends

The morphology of the tung oil/PCL blends was investigated using scanning electron microscopy (SEM). The materials were fractured in liquid nitrogen, fixed on special SEM holders, and then sputtered with gold. The prepared samples were investigated using a field emission scanning electron microscope (FE-SEM, FEI Quanta 250) operating at 10 kV under high vacuum.