Exploring the Effect of Poly(propylene carbonate) Polyol in a Biobased Epoxy Interpenetrating Network.

Poly(propylene carbonate) (PPC) polyol derived from carbon dioxide has been used to make a tough biobased interpenetrating polymer network (IPN). PPC polyol (10, 20, and 30 phr) was added to an epoxy/poly(furfuryl alcohol) IPN, and the effect of PPC polyol on the tensile modulus, tensile strength, tensile toughness, and notched Izod impact strength was determined. Dynamic mechanical analysis (DMA) was used to investigate the effect of PPC polyol on the glass-transition temperature. Loss area (LA) as a characteristic of IPN damping properties was evaluated using DMA. Scanning electron microscopy (SEM) and atomic force microscopy (AFM) were used to obtain more information on the morphology of IPNs on the micro- and nanoscale. It was found that the incorporation of PPC polyol improved the notched Izod impact strength and tensile toughness up to 190 and 220%, respectively. The damping factor peak was broadened with the addition of PPC polyol, and the glass-transition temperature was decreased as the amount of PPC polyol increased. The IPN with 20 phr PPC polyol exhibited better damping properties than neat epoxy and the epoxy/PFA IPN. SEM and AFM images revealed that PPC polyol domains were dispersed in the epoxy phase with an average diameter of around 280 nm.

Introduction

Climate change agreement by the United Nations Framework Convention on Climate Change (UNFCCC) was adopted in December 2015, and parties have committed to limit the increase in global temperature to below 2 °C. This requires stabilization of the concentration of greenhouse gases in the atmosphere (carbon sequestration), particularly carbon dioxide.[1,2] Carbon dioxide is a byproduct of combustion of fossil fuels,[3] and fossil fuels are still an integral part of our energy sources. It is predicted that it will account for about 80% of energy consumption in 2040.[4] On the other hand, it is predicted that raw materials derived from fossil fuels will be depleted in the next few decades.[5] Thus, it is important to implement approaches to utilize renewable resources and mitigate CO2 emission for a sustainable society.[5] In addition to all attempts that are being made to tap renewable energy and develop green products, reduction of carbon dioxide emission and development of carbon neutral strategies should be necessitated to alleviate catastrophic climate change effects. Trapping CO2 as an inexpensive C1 feedstock and converting it into chemicals and fuels can help in reducing emission across the world.[4]

Utilization of poly(propylene carbonate) (PPC) and PPC polyol can be one of the complementary alternatives toward carbon neutralization. PPC was synthesized by alternating copolymerization of carbon dioxide and propylene oxide in the presence of a rudimentary catalyst at the end of 1960s. Due to its biodegradability, biocompatibility, flexibility, and high fixation of carbon dioxide, its commercialization reached to 1000 tons/year.[6−8] PPC can be used in the development of polymer scaffolds, nanofibers, plastics, elastomers, foams, adhesives, and coatings.[9,10] Despite the above-mentioned great potential, practical application of PPC is still a challenge and limited to specialty polymers due to its low glass transition.[7] Therefore, finding new approaches to utilize PPC is an ongoing challenge for both academy and industry.

Epoxy resins are among the most widely used thermosets, and their projected production by the year 2017 is 3 million tons.[11] They have various applications in coating, adhesive, aerospace, and automotive industries. However, their applications have been limited due to their intrinsic brittleness as a result of their high crosslink density.[12] Fracture toughness can be improved by incorporation of a second phase in the epoxy matrix, either hard or soft, such as rubber, thermoplastics, and nanoparticles.[13]

On the other hand, most of the commercial epoxy resins are petroleum based and are not the lowest cost thermoset resins.[11] Although, in the recent years, much effort has been devoted to developing epoxy resins derived from renewable resources, including tannins, cardanol, lignin, and vegetable oils,[11] there are not many biobased epoxy resins available in the market. Partial replacement of petroleum based epoxy resins with commercially available biobased thermosets[14] and making biobased interpenetrating polymer networks (IPNs) can be alternative measures to increase the biobased content.

IPNs are composed of at least two polymers that are not chemically bonded, but the components cannot be separated because of entanglement of chains.[15] Production of IPNs is an effective way to increase the energy absorption ability, improve damping properties, and increase toughness.[16]

Poly(furfuryl alcohol) (PFA) can be a good biobased thermoset candidate to incorporate into an epoxy matrix. It is synthesized by the condensation polymerization of furfuryl alcohol (FA) in the presence of an acidic catalyst. FA is the product of hydrolysis of furfural derived from lignocellulosic feedstock.[17−19] PFA has been used in foundry molds, as adhesives for wood, and as precursors of graphitic materials.[18,20−22] In recent years, scientists have developed PFA composites,[20,23,24] although there are not enough of studies on epoxy and PFA IPNs. Woodson et al.[25] invented a new acid curable composition containing epoxy and furan resin for concrete structures. It was found that the mentioned composition had better compression strength and was more resistant to stress cracking than conventional polymer concrete materials. Pin et al.[26] copolymerized epoxidized linseed oil (ELO) and FA. It was found that the weight ratio of 50% ELO and 50% FA showed a ductile behavior with 40% elongation at break, although a drastic decrease in tensile strength (around 81%) in comparison to that of neat PFA was observed.[26] Mashouf et al.[27] investigated the hybridization effect of PFA and epoxidized soybean oil in a biobased epoxy matrix. It was found that incorporation of 25% PFA and 5% ESO can increase the notched Izod impact strength by 76%.

In this study, we aimed to prepare an epoxy/PFA IPN to increase the biobased content of the final product. Because both PFA and epoxy resin are brittle in nature, addition of a toughening agent is required. Reflecting on the green approach and the value-added product that we are looking for, PPC polyol has been selected to overcome the brittleness of the selected matrix. Because of low Tg and compatibility of its backbone with epoxy and PFA, it can be considered as a good candidate to toughen the epoxy/PFA IPN. To the best of our knowledge, there is only one study on the toughening of an epoxy matrix with PPC,[28] and further in-depth studies are required to evaluate its effect in thermosets. Huang et al.[28] added polypropylene carbonate into an epoxy resin. Methyl tetrahydrophthalic anhydride (MTHPA) was used as a curing agent, and it was found that PPC can improve the toughness of the epoxy resin. In two recent studies,[7,29] a triblock copolymer of PPC and poly(ε-caprolactone) was synthesized and added into the epoxy matrix. It was found that the tensile toughness of the epoxy resin improved upon addition of the triblock copolymer, although the tensile strength reduced.

In the present study, the effect of addition of PPC polyol in the epoxy/PFA matrix was investigated. PPC polyol (10, 20, and 30 phr) was added into the epoxy/PFA crosslinked network. The effect of PPC polyol addition on the tensile strength, tensile modulus, and notched Izod impact strength was studied. Dynamic mechanical analysis (DMA) was used to find the damping properties of the thermoset. The morphology of the thermosets on the micro- and nanoscale was investigated using scanning electron microscopy (SEM) and atomic force microscopy (AFM).

Results and Discussion

Mechanical Properties

Tensile test and notched Izod notched impact tests were performed to understand the mechanical properties of the thermoset samples. Figure shows the stress–strain curve of all cured formulations. A significant transition from brittle behavior in E, EP, and EP-10 to ductile response in EP-20 and EP-30 was observed. The values of tensile strength and modulus can be found in Figure . Also, the area under the stress versus strain curve was calculated and presented as tensile toughness and can be found in Figure . The area under the stress versus strain curve can be considered a measure of the toughness of the material and representative of energy for crack initiation and propagation. For a valid comparison, analysis of variance was performed. Mean comparisons of all mechanical responses were analyzed using Fisher’s least significance difference (LSD) method with a significance level of 0.05 (α = 0.05). Figures S1–S4 show Fisher’s LSD curves of all responses. In the graphs, if an interval does not contain zero, it means that the corresponding response means are significantly different at the determined α level.

Figure 1

Stress vs strain curve of E, EP, EP-10, EP-20, and EP-30.

Figure 2

Tensile strength and tensile modulus of E, EP, EP-10, EP-20, and EP-30.

Figure 3

Tensile toughness of E, EP, EP-10, EP-20, and EP-30.

The tensile strength and modulus of neat epoxy (E) are 54.04 MPa and 2.7 GPa, respectively. Incorporation of 15% PFA into the epoxy matrix increased the tensile strength to 62.48 MPa. In addition, the presence of the PFA network with a higher modulus[20] in the epoxy matrix increased the tensile modulus to 3.95 GPa, as expected. Addition of 10 phr PPC did not change the tensile strength significantly, whereas further addition of PPC polyol reduced the tensile strength of the matrix. A similar trend was observed by addition of the PPC/PCL copolymer into the epoxy matrix.[7] It can be observed that incorporation of soft PPC chains reduced the tensile modulus. It can be found that the addition of PPC polyol had a positive effect on the improvement of the toughness of the epoxy/PFA matrix. Addition of 10, 20, and 30 phr PPC polyol increased the tensile toughness of the epoxy/PFA matrix by around 91, 220, and 84%, respectively. In addition, there is no significant difference between the toughness of E and EP. It is worth mentioning that EP-20 also exhibited better tensile toughness than neat epoxy (i.e., E). Because PPC polyol is a soft phase in the IPN matrix, it is assumed that cavitation of PPC particles and localized shear yielding of the matrix are the main toughening mechanisms.[30] The observed results are in good agreement with those of Chen et al.[7] and Liu et al.[29] They added a block copolymer of PPC and poly(ε-caprolactone) into the epoxy resin and found a decrease in tensile strength and an increase in toughness.

The notched Izod impact strength of all thermoset formulations can be found in Figure . The notched Izod impact strength values represent the response of formulations to a high strain rate. Addition of 20 and 30 phr PPC polyol significantly increased the impact energy of the epoxy/PFA matrix by 36 and 191%. Impact strengths of EP-20 and EP-30 are also higher than that of the neat epoxy matrix, exhibiting their better response to the high strain rate than neat epoxy. Huang et al.[28] found that incorporation of PPC into the epoxy resin cured by MTHPA can improve the impact strength of the epoxy matrix.

Figure 4

Notched Izod impact strength of E, EP, EP-10, EP-20, and EP-30.

DMA

To find the glass-transition temperature and time–temperature-dependent motion of IPN segments, DMA studies were performed. Figure shows the storage modulus and tan δ of E, EP, EP-10, EP-20, and EP-30. α-Transitions of E, EP, EP-10, EP-20, and EP-30 are 138, 104, 73, 54, and 42 °C, respectively. A comparison of cured neat epoxy (E) with EP IPN showed a decrease in glass-transition temperature, and it can be related to the lower crosslink density of EP, as shown in Table . The crosslink density was calculated according to rubber elasticity theory and using ν = E′/3RT, where ν, E′, R, and T are the crosslink density (mol m–3), storage modulus at Tg + 30 K, gas constant, and absolute temperature, respectively.[7,31] Moreover, it can be found that the addition of PPC polyol in the epoxy/PFA matrix reduced the glass-transition temperature. Addition of PPC polyol also broadened the damping factor peak. Reduction in glass-transition temperature and broadening of loss factor upon addition of PPC into an epoxy matrix were previously reported by Huang et al.[28] The reduction in glass-transition temperature can be related to the lower crosslink density of samples (see Table ) and the plasticization effect of PPC soft chains. Tensile and impact tests were performed at room temperature (around 25 °C). Hence, EP-20 and EP-30 are in their leathery region. This means that samples are tough and side group motions in the network can happen. In addition, the observed β-transitions in E and EP are mainly related to hydroxyl ether, methyl, and diphenyl ether groups,[7,26] whereas β-transitions cannot be specified in EP-10, EP-20, and EP-30 due to broadening of the damping factor peak. It is worth mentioning that the glass-transition temperature was also obtained using differential scanning calorimetry (DSC), and the results are shown in Table S3. Although the numbers were different due to difference in the measurement method of DSC and DMA, a similar decreasing trend was observed.

Figure 5

Storage modulus and tan δ graphs of E, EP, EP-10, EP-20, and EP-30.

Table 1

Glass-Transition Temperature (Tg), Storage Modulus at Tg + 30 K (E′), and Crosslink Density (ν)

sample	Tg (°C)	E′ at Tg + 30 K (MPa)	ν (×103 mol m–3)
E	138	18.19	1.77
EP	104	14.59	1.55
EP-10	73	9.29	1.08
EP-20	54	6.167	0.75
EP-30	42	3.688	0.47

To quantitatively compare the damping properties of samples, the area under the loss modulus curve was calculated. Loss area (LA) can be considered as a better characteristic to compare damping properties, as the engineering products are used in a range of temperatures, not a specific temperature. The LA was calculated based on the method proposed by Fay et al.[32] and the lower integration limit was chosen to be −25 °C. Figure shows LA of thermoset IPNs. As can be seen, damping properties of the epoxy/PFA IPN improved with addition of soft PPC chains, and EP-20 exhibited the highest LA. Because change in LA is mainly related to the chemical composition and morphology in multicomponent systems,[32] the morphology of samples was studied using SEM and AFM.

Figure 6

Comparison of LA of samples.

Morphology

Morphology is an important characteristic in IPNs with good damping properties. Figure A–C shows SEM micrographs of etched, impact-fractured E, EP, and EP-20. Figure A showed a surface with sharp edges. This micrograph is attributed to a brittle fracture that is well known in an epoxy matrix.[26] Because PFA chains can swell when in contact with acetone, the crosslinked PFA network can be easily distinguished in the epoxy matrix (Figure B). Moreover, PPC polyol can be dissolved in acetone. It is worth mentioning that 4 days of etching was chosen in this study based on our preliminary observations. It was difficult and even impossible to observe the PFA network on fractured surfaces with shorter etching time in acetone (i.e., 24 h). Figure B shows interconnected PFA chains in the epoxy matrix, but the fractured surface has sharp edges, confirming the lower impact strength of EP in comparison to that of EP-20.

Figure 7

SEM micrograph of etched, impact-fractured E (A), EP (B), and EP-20 (C) and cryo-fractured surface of EP-10 (D), EP-20 (E), and EP-30 (F).

A study of the EP-20 fractured surface showed rough surfaces in epoxy regions. This means that PPC polyol chains are mainly distributed in the epoxy matrix and made it tougher; however, the PPC polyol distribution cannot be clearly distinguished from impact specimens. To investigate the distribution of PPC polyol in the matrix, cryo-fractured samples with 30 min etching in acetone are shown in Figure D–F. From cryo-fractured surface morphology, it can be concluded that PPC polyol distributed as spherical domains with an average size of around 280 nm, and the size of PPC polyol domains did not vary with change in PPC polyol content. It is worth noting that with a lower amount of PPC polyol (i.e., EP-10), the epoxy matrix showed a brittle fracture, which is in good agreement with the observed mechanical properties (see Figure S5). Although no sign of distribution of PPC polyol in the PFA matrix was observed, AFM was also used to make sure that the observed gap between the swelled PFA phase and the epoxy phase is not related to PPC polyol etching.

Due to the higher amount of PPC polyol, EP-30 was chosen for AFM scans. PQNM was used to better differentiate between the effect of topography and mechanical heterogeneity. Figure shows the height and adhesion images of EP-30, and the deformation image can be found in Figure S6. The dark circles in Figure A represent the PPC polyol phase, and image processing of the height image showed an average diameter of 210 nm, which is in close agreement with the average diameter found in SEM images of both EP-20 and EP-30. The adhesion force measured by AFM is mostly related to the difference in chemical interactions of the tip and surface[33] and can be used to distinguish epoxy and PFA phases. From Figure B, it is evident that PPC polyol domains are mainly distributed in the epoxy phase and are responsible for toughening the epoxy/PFA IPN. Smaller scan size (Figure S7) revealed inhomogeneity in the epoxy matrix. Spatial heterogeneity in an epoxy matrix was a challenging issue for decades and has been confirmed recently by Morsch et al. using AFM-IR.[34] They related the internal nodular morphology of epoxy to heterogeneous network connectivity in an epoxy matrix. In our case, scans of EP-30 in the epoxy region and also scans of E and EP samples revealed nanosize nodular heterogeneity in the structure.

Figure 8

AFM height (A) and adhesion (B) images of EP-30.

Conclusions

In summary, we developed a biobased tough IPN through addition of PPC polyol in an epoxy/PFA matrix. Mechanical properties, morphology, and dynamic mechanical properties of the samples with a change in the amount of PPC polyol (10, 20, and 30 phr) were studied. Addition of 20 and 30 phr of PPC polyol into the IPN increased its notched Izod impact strength by 36 and 191%, respectively. The improvement was related to the presence of a PPC soft phase in the brittle matrix that could absorb energy. The observed behavior was further described using SEM and AFM. SEM micrographs showed PPC polyol was mainly distributed in the epoxy phase with an average diameter of 280 nm. AFM images confirmed the same finding. Using SEM and AFM techniques, it was found that the PFA phase is randomly distributed in the epoxy matrix. Also, EP-10, EP-20, and EP-30 showed around 91, 220, and 84% increase in tensile toughness, respectively. The LA of EP-20 increased in comparison to that of EP, confirming the improved damping properties of EP-20. DMA results showed a reduction in glass-transition temperature and confirmed that EP-20 and EP-30 were in their leathery region, which provides more evidence for their ability to damp energy.

Experimental Section

Materials

Diglycidyl ether of bisphenol A epoxy resin (Epon 828) and triethylenetetramine (Epikure 3234) were purchased from Hexion. The amount of hardener for all samples was chosen based on stoichiometry. Epoxy cured with the amine hardener is referred to as “E” hereafter. PPC polyol with a molecular weight of 2000 (Converge polyol 212-20) was purchased from Novomer. The polyol is a diol, and around 40% of its mass contains carbon dioxide. Water-insoluble homopolymer of FA resin (QuaCorr 1300) was purchased from Penn A Kem. QuaCorr 2001, a mixture of 1,2-benzenedicarbonyl dichloride and dimethyl phthalate, was purchased from Penn a Kem, and 5% of it (based on the weight of PFA in the formulation) was used as a catalyst. Weight ratio of 85/15 (epoxy/PFA) was chosen for all IPNs. This ratio was chosen based on our preliminary studies and was selected to reach the highest possible biobased content, minimize uncrosslinked furan resin, and also balance the mechanical properties, including the tensile, flexural, and impact strengths. More information on the effect of epoxy/PFA ratio can be found in the Supporting Information. Cured epoxy/PFA IPN was called “EP”. To study the effect of PPC polyol, 10, 20, and 30 phr of PPC polyol were used, and cured samples are referred to as “EP-10”, “EP-20”, and “EP-30”, respectively. Table shows a summary of sample designations.

Table 2

Summary of Weight Ratio and Sample Designations

sample designation	epoxy/PFA (wt %)	PPC polyol (phr)
E	100/0	0
EP	85/15	0
EP-10	85/15	10
EP-20	85/15	20
EP-30	85/15	30

Sample Preparation

In the case of E and EP samples, materials were mixed properly using a mechanical stirrer. EP-10, EP-20, and EP-30 were prepared by mixing epoxy and PPC polyol in an oil bath at 80 °C until complete disappearance of PPC polyol. Then, the mixture was degassed overnight at room temperature. Then, QuaCorr 1300, QuaCorr 2001, and Epikure 3234 were added to the mixture and mixed properly. A Carver hydraulic press (model 4128) was used to cure all samples, and all samples were cured under the same conditions. The mixture was poured into a rectangular mold (23 cm × 23 cm × 3 mm) and held at 90 °C under 6000 psi gauge pressure for 2 h and at 150 °C under 6000 psi for 4 h. It is worth noting that the curing temperature and time were chosen according to preliminary studies. The cured samples were cooled to room temperature by the circulation of water through platens, and the cooling cycle was 5 min. Complete curing of samples was confirmed by the absence of any exothermic peak in DSC graphs, and all tests were performed after at least 48 h of curing.

Methods

Mechanical Test

Tensile and impact samples were cut using a computer numeric controlled mill (VF-4, Haas Automation Inc.) according to ASTM D638 and D256. Tensile tests were carried out at room temperature using an Instron testing machine (model 3382) with a 5 kN load cell. Type IV samples were used for the tensile test, and the crosshead speed was set at 5 mm/min. Young’s modulus and tensile strength values are average of at least five measurements.

Notched Izod impact energies of samples were measured using a Testing Machine Inc. (TMI) instrument according to ASTM D256, and the values are an average of at least six replicates. The impact test was performed after at least 40 h of notching.

DMA measurements were conducted using a Q800 machine from TA instruments in dual cantilever mode. Rectangular specimens with dimensions of 3.2 × 12.5 × 65 mm3 were used. The tests were carried out by heating the samples from −80 to 150 °C at a heating rate of 3 mm/min and oscillating stress of 1 Hz.

SEM

A Phenom ProX scanning electron microscope (Phenom-World BV, Eindhoven) was used to study the fracture surface of samples. The accelerating voltage was 10 kV to collect images. Impact-fractured samples were etched in acetone for 4 days and were used for morphology studies. Also, cryo-fractured samples were prepared by soaking rectangular bars in liquid nitrogen for 1 h. The surfaces of all samples were gold coated by sputtering using a Cressington 108 Sputter Coater (Cressington Scientific Instruments Ltd., Watford, UK).

AFM

Nanoscope Multimode 8 (Bruker) with a Nanoscope V controller and Nanoscope 8.10 software in peak force quantitative nanomechanical (PFQNM) mode was used. An RTESPA-300 silicon probe was used to scan the surface of samples. To make a flat surface, samples were microtomed using Leica Ultracut (Leica, Germany). Deflection sensitivity was calibrated using ramp on sapphire. Relative calibration method on the polystyrene film (Bruker) was used before the scan. The tip radius and spring constant were estimated to be 40 nm and 20 N/m, respectively. The image resolution was 512 per line, and the scan rate was 1 Hz. Adhesion and height channel images were exported using Nanoscope analysis 1.50, and Adobe Photoshop CS6 was used for image processing.