Sustainable glucose-based phenolic resin and its curing with a DGEBA epoxy resin

Journal of the Taiwan Institute of Chemical Engineers 71 (2017) 381–387

Contents lists available at ScienceDirect

Journal of the Taiwan Institute of Chemical Engineers journal homepage: www.elsevier.com/locate/jtice

Sustainable glucose-based phenolic resin and its curing with a DGEBA epoxy resin Yongsheng Zhang a,b, Fatemeh Ferdosian b, Zhongshun Yuan b,∗, Chunbao Charles Xu b,∗ a

School of Chemical Engineering and Energy, Zhengzhou University, 100 Science Avenue, Zhengzhou 450001, Henan, China Department of Chemical and Biochemical Engineering, Institute for Chemicals and Fuels from Alternative Resources (ICFAR), Western University, London, ON N6A 5B9, Canada

b

a r t i c l e

i n f o

Article history: Received 23 January 2016 Revised 30 October 2016 Accepted 21 November 2016

Keywords: Phenol–hydroxymethylfurfural resin DGEBA epoxy resin Curing kinetics Thermal stability Renewable polymer Properties

a b s t r a c t A sustainable novolac-type resin – phenol–hydroxymethylfurfural (PHMF) resin was prepared by reacting phenol with HMF, in-situ derived from glucose, at 120 °C by acid catalysis. Bisphenol A type epoxy resin, i.e. bisphenol A diglycidyl ether (DGEBA), was used as a formaldehyde-free curing agent by substituting conventional formaldehyde-based hexamethylene tetraamine (HMTA) to crosslink the PHMF resin. Curing mechanism was probed and the curing proceeded likely with the ring opening reaction between the DGEBA and reactive hydroxyl groups. DGEBA not only made this system truly formaldehyde-free but also helped form a void-free matrix which is an important merit for composites. The kinetic parameters of the curing reaction were evaluated with model-free and model-fitting methods using exothermal peak data from the curing process. The thermo-mechanical characterization of the cured resin and fiber reinforced bio-composites showed good heat resistance and mechanical performance, suggesting its potential for producing void- and formaldehyde-free composite materials. © 2016 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction Due to increasing environmental awareness and the fast depletion of petroleum resources, the demand for the sustainable chemicals and materials (e.g., bio-based plastics and polymers) derived from renewable resources is expected to increase rapidly in the next 20–30 years [1,2]. Phenol–formaldehyde resin (PF), the earliest commercial synthetic resin, has been widely used as an adhesive in engineered wood products and composites owing to its excellent mechanical properties. While most work on bio-based PF resin have been focused on phenol substitution with bio-phenols derived from lignin or lignocellulosic biomass, [3,4] formaldehyde has also provoked increasing environmental concern [5]. In fact, formaldehyde has been classified as a known human carcinogen; its permissible exposure level has been strictly regulated [6]. Due to the toxicity of formaldehyde, it is imperative to develop formaldehyde-free phenolic resins. Procuring formaldehyde-free phenolic resins can be achieved by replacing formaldehyde with green alternatives, in particular chemicals from bio-renewable resources (biomass). Glucose is abundantly available from cellulose and hemicellulose via hydrolysis. The authors’ group has been working on the re-

∗

Corresponding authors. E-mail addresses: [email protected] (Z. Yuan), [email protected] (C.C. Xu).

placement of formaldehyde with HMF, in-situ generated from glucose via a catalytic process, for producing a novolac type phenol– (hydroxymethyl) furfural (PHMF) resins [7]. Conventionally, hexamethylene tetramine (HMTA) has been used as the most common compound for curing novolac type phenolic resins. However, HMTA is regarded as a hazardous air pollutant and restricted by the Federal Environmental Protection Agency of the United States in polymeric composite manufacture as it decomposes into ammonia and formaldehyde upon heating [8-10]. Thus, it is necessary to find an alternative to HMTA for novolac resins. Epoxides have a high reactivity towards a variety of chemicals or functional groups, via either nucleophilic or electrophilic reaction through epoxy ring opening. An example of epoxides applications is a hardener for novolac phenolic resins. Hsieh and Su studied cure kinetics of epoxy–novolac molding while forming secondary alcohols by proton transfer [11]. Researchers further confirmed that the secondary alcohols formed could easily react with the epoxy groups [12,13]. Han and coworkers reported the curing of phenol novolac with biphenyl epoxy resin at an equivalent mass [14].] Polyphenol-generated epoxy resins with a high glass transition temperature are frequently used as a hardener [15]. Gagnebien et al. used bisphenol A and glycidylethers as models of epoxy and phenol, respectively, to study the hydroxyetherification reactions, where the formation of the hydroxy branch was confirmed [16], as similarly reported by Alevey [17] and Burchard et al. [18].

http://dx.doi.org/10.1016/j.jtice.2016.11.025 1876-1070/© 2016 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

382

Y. Zhang et al. / Journal of the Taiwan Institute of Chemical Engineers 71 (2017) 381–387

Similarly to an O-creasol novolac epoxy resin, the curing of a bisphenol A type epoxy resin and a phenolic novolac resin is believed to proceed with an autocatalytic mechanism [19,20]. Curing novolac resins with epoxy resins exhibits some advantages, such as moderate reaction temperatures, no formation of volatile compounds, and thus no voids created. In this study, DGEBA was used for the first time as a formaldehyde-free cross-linker for PHMF resins. The objectives of this work included: firstly, the investigation of the curing mechanism between the DGEBA and phenolic resins; secondly, the evaluation of the kinetic parameters of the curing reaction with different kinetic methods and simulations using data acquired from differential scanning calorimetry (DSC); and thirdly, the characterization of thermo-mechanical properties of the bio-composites derived from aforementioned matrix. 2. Material and methods 2.1. Materials Phenol (ACS reagent, 99.7 wt%, Mallinckrodt Baker), d-glucose (ACS reagent, Fisher Scientific) and diglycidyl ether of bisphenol A (DGEBA) (ACS reagent, Sigma-Aldrich) were used as received. ACS reagent-grade acetone, tetrahydrofuran (THF) and ethanol were from Caledon Laboratory Chemicals and used as received. Chromium (II) chloride (CrCl2 ), chromium (III) chloride (CrCl3 .6H2 O), tetraethylammonium chloride (TEAC), and hexamethylenetetramine (HMTA) were purchased from Sigma-Aldrich and used as received. BGF fiberglass cloth was purchased from Freeman, Ohio. 2.2. Synthesis and curing of PHMF resin The PHMF resin was synthesized using a proprietary process developed by the authors [7]. Briefly, 35.25 g phenol, 135.00 g glucose, 30.00 g water, and 0.4185 g CrCl2 (0.02 M), 0.4535 g CrCl3 .6H2 O (0.01 M), 1.8695 g TEAC (0.06 M) were charged to a 1 l glass pressurized batch reactor. The reactor was then heated in an oil bath preheated at 120 °C, and stirred with a magnetic stirrer for 8 h. According to our previous work [7], glucose is firstly isomerized to fructose which is subsequently dehydrated to HMF. The reaction is followed by nucleophilic addition of the electron rich carbons of the para and ortho positions of phenol to the electrophilic aldehyde group in HMF. The hydroxymethyl group in HMF can simultaneously react with the para and ortho position of phenol OH through a Friedel–Crafts alkylation. In the resin synthesis, CrCl2 and CrCl3 would play a role as acid catalyst for the glucoseto-HMF conversion reaction and the resinification reaction of HMF with phenol. Phenol and glucose conversion was determined to be 91.5 wt% and 88 wt%, respectively, by HPLC analysis. After the PHMF resin was prepared and purified, the sample was dried in a vacuum oven under 60 °C for more than 12 h. The resin was thermally cured after mixing it homogeneously with a specified amount of DGEBA. The curing temperature program was: 120 °C for 30 min, 150 °C for 30 min, and 180 °C for 1 h. 2.3. Characterization The Fourier transform infrared (FTIR) spectra were collected with the resin or cured resins on a Perkin-Elmer Spectrum Two IR Spectrometers, scanning from 500 to 40 0 0 cm−1 . A Mettler-Toledo Differential Scanning Calorimeter (DSC) was used for curing kinetics study when PHMF resin was uniformly mixed with 20 wt% DGEBA. Dynamic scans were conducted in a temperature range of 25–250 °C, at various heating rates ranging from 5, 10, and 15–20 °C/min. In each test, 5–10 mg of the mixture was used in an aluminum crucible capped with a perforated

lid. Since steric hindrance effects prevent each hydroxyl group from being equally reactive with epoxy ring, the amount of DGEBA was optimized as of the maximum value of the glass transition temperature of cured resin, i.e. 10% (73 °C), 20% (90 °C), 30% (84 °C) and 40% (81 °C) of PHMF resin by weight. The dynamic mechanical analysis (DMA) curves were obtained on a Netzsch 242C analyzer. The measurement was made under flexural modes in a three-point bending clamp with samples of dimensions of 50 mm × 10 mm × 1 mm. The experimental conditions were: 5 μm oscillation amplitude, 1 Hz frequency, and temperature ramped from 30 to 250 °C at 5 °C/min. Composites were prepared by applying homogeneous solution of PHMF and curing agent onto glass fiber at 1/1 ratio (w/w) and keeping at room temperature until solvent evaporated. The sample was cured by hot pressing in a Carver hydraulic hot press using the same procedure as resin curing under the load of 35 MPa (50 0 0 psi). The pressed composites were also shaped into dumbbell-shaped samples and their tensile strengths were measured as of ASTM 638 on an ADMET Expert 7600 universal test machine. 2.4. Theory of curing kinetics Nonisothermal differential scanning calorimetry techniques are commonly applied to study the curing behavior and kinetics for resins and polymers. The results of DSC measurements are then used for evaluation of kinetics parameters with different models, including model-free methods (Kissinger and Flynn–Wall–Ozawa (FWO) methods) and the model-fitting method as described below. The Kissinger method is widely used for the calculation of activation energy [21] and is given by Eq. (1):

Ea β = Ae(−Ea /RTp ) RTp2

(1)

where β is the heating rate, expressed by β = dT/dt, Ea is apparent activation energies (kJ/mol), A is the pre-exponential factor (min−1 ), R is gas constant (8.314 J/mol/K) and Tp is the exothermic peak temperature (K). By taking the logarithm of Eq. (1), one obtains Eq. (2). A and Ea can be obtained by plotting −1n(β /Tp2 ) against 1/Tp . The advantage of the Kissinger method is that Ea can be obtained without the knowledge of mechanism prior to the reaction.



− ln

β





= − ln

Tp2

AR Ea



+



1 Ea Tp R



(2)

The Flynn–Wall–Ozawa method can be employed to quantify Ea [22,23] and is given by Eq. (3).

log β = −

0.4567Ea +C RTp

(3)

where C is a constant, if one plots log β versus 1/Tp , the slope is Ea − 0.4567 . The calculation of Ea is also independent of the thermal R curing mechanism. The isoconversional method is another useful approach to probe the progress of the reaction at varying conversion rates and provide additional insight into the reaction. Similar to Eq. (2), the activation energy at different extents of reaction can be calculated using the temperature at a given conversion degree α .



− ln

β





= − ln

Tα2

AR Eα



+

1 Tα

E  α

R

(4)

One of the most common differential isoconversional method is Friedman model, which is based on the following equation



ln

dα dt



α ,i

= ln [ f (α )Aα ] −

Eα RTα ,i

(5)

Y. Zhang et al. / Journal of the Taiwan Institute of Chemical Engineers 71 (2017) 381–387

383

Scheme 1. Proposed cross-linkage mechanism between PHMF resin and DGEBA.

Fig. 1. PHMF resin (a), mixture of PHMF resin and DGEBA prior to curing (b), and the hardened resin after curing (c).

For linear nonisothermal programs, Eq. (5) is frequently used in the following form:



dα ln βi dT



α ,i

= ln [ f (α )Aα ] −

Eα RTα ,i

(6)

Applying the Kisinger–Akahira–Sunose (KAS) method at Tα – temperature at conversion degree α , [24,25]



ln

βi Tα2,i



= Const −

Eα RTα

(7)

Isoconversional method is accurate and it determines the activation energy as a function of conversion. Thus, the two aforementioned methods (Friedman, KAS) are used and made comparison, respectively. Unlike the above described model-free methods or the isoconversional method, the model-fitting method involves a reaction model f(α ) related to the reaction mechanism. The reaction model f(α ) had to be selected in advance in order to estimate the kinetic parameters such as overall order of reaction (n), activation energy (Ea ) and the pre-exponential constant (A) by fitting the experimental data with an appropriate reaction rate equation. The reaction between epoxy and novolac resins is an autocatalytic reaction [26] and the equation for this model is given below:

dα = (k1 + k2 α m )(1 − α )n dt

(8)

where k1 = A1 exp (−E1 /RT), and k2 = A2 exp (−E2 /RT), m and n are the kinetic exponents of the reaction and (m + n) gives the overall reaction order [14]. 3. Results and discussion 3.1. Proposed curing mechanism The FTIR spectra of the purified PHMF resin, mixture of PHMF resin and DGEBA prior to curing, and the hardened resin are plotted in Fig. 1. The absorbance peaks of the spectra are assigned in Table 1. Among those assigned absorbance peaks, the CHO (carbonyl) group stretching at 1702 cm−1 confirms that aldehyde groups formed through glucose isomerization and dehydration during PHMF resin synthesis and remained as an end group. The primary difference before (Fig. 1(b)) and after curing (Fig. 1(c)) is the

Fig. 2. DSC curves of the PHMF resin cured with 20 wt% DGEBA at various heating rates.

band at 911 cm−1 from oxiran group was consumed during curing, and thus disappeared in the cured sample. In the mean time, the numbers of hydroxyl groups were significantly reduced, as indicated by the height of peaks in 30 0 0–350 0 cm−1 region. Epoxy resins can convert primary and secondary hydroxyl groups into secondary alcohol and ether, respectively [27]. A crosslinkage can be achieved through an anionic attack of the epoxy ring by hydroxyl groups, whereby the hydroxyl group is converted into an ether group and the epoxy ring forms another hydroxyl group. On the main chains, a large number of OH groups would appear as a consequence of the epoxy ring opening. However, the system may undergo side reaction such as Friedel–Crafts alkylation between hydroxymethyl group and the para- or ortho-position of phenol. The primary curing mechanism is thus proposed in Scheme 1. 3.2. Model-free curing kinetics Understanding of curing kinetics can provide important information for polymer product manufacture and applications. Heat release profiles in curing of PHMF resin with 20 wt% DGEBA were obtained at four heating rates and are shown in Fig. 2. The cure features, such as initial onset curing temperature (Ti ), the peak temperature (Tp ), end set temperature (Te ), and enthalpy (࢞H) were obtained and displayed in Table 3. The cross-linking reactions present a single exothermic deflection from 101 °C to 170 °C and a peak temperature Tp at around 155 °C. Similar behavior was observed in PF novolac and the epoxy system by Lee et al. [28] and Choi et al. [29] at 100–160 °C but with Tp at around 120 °C [28,29]. Curing of the PHMF resin requires relatively higher temperatures likely due to its limited reactivity to epoxy and higher steric hin-

384

Y. Zhang et al. / Journal of the Taiwan Institute of Chemical Engineers 71 (2017) 381–387 Table 1 FT-IR bands assignments in Fig. 1. Band (cm−1 )

Vibration

3100–3550 2830–2980 1700–1715 1660 1450; 1505; 1602 1226 1017 911

Stretching Stretching Stretching Stretching Stretching Stretching Stretching

Assignment O–H C–H C =O C =O C–C C–O(H) + C–O(Ar) C–O(H) + C–O(C)

Phenolic OH + aliphatic OH CH3 + CH2 and OCH3 Unconjugated C=O Conjugated C=O Aromatic skeleton Phenolic OH + ether 1st order aliphatic OH + ether Oxiran (epoxy ring)

Table 2 Curing characteristics of PHMF resins with 20 wt% DGEBA at different heating rates.

β (°C/min)

Tp (°C)

Ti (°C)

Te (°C)

H∞ (J/g)

5 10 15 20

144 153 158 162

101 103 105 108

151 161 166 170

87 98 91 88

Fig. 4. Fractional conversion as a function of temperatures while curing PHMF and 20 wt% DGEBA at various hearing rates of 5 (a), 10 (b), 15 (c), and 20 (d) °C/min.

Fig. 3. DSC curves of the PHMF resin cured with 20 wt% HMTA at different heating rates: 5 (a), 10 (b), 15 (c), and 20 (d) °C/min.

drance than PF novolac. Fig. 2 and Table 2 illustrate how Tp shifts to a higher temperature region along with an increased heating rate, while the enthalpy remains identical during the reactions, suggesting the curing reaction is almost complete. Kissinger and FWO plots were applied to obtain the activation energy (Ea ) and pre-exponential factor (A). The overall activation energies obtained from Eqs. (1) and (2) were higher than other literature results, e.g. between 80 and 100 114 kJ/mol, [20,30] because of the higher molecular weight and steric hindrance of PHMF. For comparison, the PHMF resin was also cured with the conventional agent HMTA (Fig. 3), and the kinetic parameters were close to PHMF–DGEBA curing (Table 3), suggesting the energy barrier upon curing is mainly determined by the resin itself. The end temperature (170 °C) implies post cure at a higher temperature (e.g. 180 °C) can reduce the curing cycle to enhance the productivity.

the degree of reaction α , which increases slowly at the beginning, faster in the middle stage, and then levels off before completion. From the figure, it can be observed that, while increasing the heating rate, the reaction conversion shifts to a higher temperature because a higher heating rate does not allow enough time for the reactive groups to react. Thus a same degree of reaction requires a higher temperature. Furthermore, the model-free isoconversional method was applied to analyze the process. Figs. 5 and 6 present the Kissinger and FWO plots, respectively, for the PHMF resin containing 20 wt% of DGEBA at conversion degree varying between 0.05 and 0.95. As illustrated in Eqs. (4) and (5), the slope of the line fitted on the curves gives the activation energy of the system. Table 4 shows the values of Ea and correlation coefficient R2 along reaction conversion. The correlation coefficients (R2 ) are mostly greater than 0.98. Ea remains constant during the curing process, validating the signal reaction mechanism proposed in Scheme 1. The increased temperature leads to the increased reaction rate and functional groups flexibility, thus lowering the energetic barrier for the ring opening reaction. On the other hand, the higher viscosity engenders a higher energy barrier associated with the molecular diffusion. These two factors compromise the effect from their counterparts, accounting for the stable Ea . The values are a little lower but close to those derived from Kissinger and FWO methods (Table 3). 3.4. Kinetics modeling

3.3. Curing kinetics in isoconversional methods The heat flow under the exothermic plots is integrated against the temperature and then processed to estimate fractional conversion (α ) as well as the rate of reaction (dα /dt) [31,32]. As illustrated in Fig. 4, a series of S-shaped curves were generated, representing

The reaction kinetic parameters were determined by fitting the dynamic DSC conversion data to the autocatalytic equation (i.e. Eq. (6)) by using the least square regression method. As shown in Fig. 7, there is congruence between the experimental data and predicted data obtained from the autocatalytic model for all heating

Y. Zhang et al. / Journal of the Taiwan Institute of Chemical Engineers 71 (2017) 381–387

385

Table 3 Kinetics parameters for curing of PHMF resins with 20 wt% DGEBA or HMTA. Sample

PHMF + DGEBA PHMF + HMTA

Heating rate (°C/min)/peak temperature (°C)

Kissinger

5

10

15

20

Ea (kJ/mol)

A

Flynn–Wall–Ozawa Ea (kJ/mol)

144.2 131.6

153.5 139.1

157.8 143.9

161.7 148.2

113.7 112.7

6.55E + 10 1.49E + 11

114.8 113.7

Table 4 Kinetic parameters obtained by isoconversional methods for curing of PHMF resins with 20 wt% DGEBA.

α 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95

Friedman

KAS 2

Ea (kJ/mol)

R

90.3 99.8 106.0 110.6 113.3 115.1 116.1 118.1 116.8 116.1 117.3 114.4 112.8 109.8 108.4 107.9 99.6 99.6 97.1

0.975 0.992 0.996 0.998 1.0 0 0 1.0 0 0 1.0 0 0 1.0 0 0 0.990 0.999 0.999 0.999 0.998 0.995 0.985 0.981 0.973 0.968 0.957

Ea (kJ/mol)

R2

111.7 103.0 105.7 106.5 107.4 107.7 108.9 111.3 110.0 109.7 112.2 110.8 111.0 111.4 112.6 113.2 110.0 109.0 107.7

0.954 0.966 0.979 0.984 0.985 0.995 0.995 0.996 0.995 0.997 0.998 0.997 0.997 0.997 0.997 0.998 0.995 0.995 0.989

Fig. 5. Friedman plots for the PHMF resin containing 20 wt% of DGEBA at conversion degree varying between 0.10 and 0.95. Table 5 Kinetic parameters derived from the autocatalytic model for curing of PHMF resins with 20 wt% DGEBA.

β

k1

E1 (kJ/mol)

n

k2

E2 (kJ/mol)

m

5 10 15 20 Ave

4.47E + 04 4.50E + 04 4.50E + 04 4.53E + 04 4.50E + 04

62.9 75.0 60.9 64.5 65.8

0.39 0.31 0.51 0.49 0.43

1.92E + 08 1.77E + 08 1.29E + 08 1.24E + 08 1.55E + 08

82.1 81.2 79.3 78.4 80.2

0.69 0.67 0.72 0.84 0.73

3.5. Thermo mechanical property

Fig. 6. KAS plots for the PHMF resin containing 20 wt% of DGEBA at conversion degree varying between 0.10 and 0.95.

rates. From kinetic parameters in Table 5, for the curing reaction at all heating rates, E1 and E2 fall in a relatively narrow range of 63– 75 kJ/mol and 78–82 kJ/mol, respectively. The overall order (n + m) of the reaction is in the range of 1.0–1.3, implying the reaction rate is proportional to the reactant concentration. These values are similar as those reported in literature [33]. Since the higher reaction rate are between 135 and 165 °C, special attention shall be paid to this temperature range during the composite processing.

The PHMF resins can be used as adhesives for the manufacture of fiber-reinforced plastic (FRP) bio-composites. In this study, the FRP bio-composite was produced using glass fiber of an equal mass to that of the admixture of PHMF resin and DGEBA hardener. DMA measurement was performed on the profiles of the prepared FRP bio-composite. From the DMA test, the storage modulus (E’) and tanδ of the composite specimen were calculated and shown in Fig. 8. The authors believe that the relatively low storage modulus values are likely due to the poor crosslink density of the bioresin. In fact, the storage modulus of the bio-composite obtained in this work are comparable to those of biocomposites from tannin– phenolic polymers reinforced with 30% coir fibers [34] and those from lignin cured PHMF composites [35]. The Tg of the cured samples was determined by the peak temperature of tanδ , which is defined by the ratio of storage modulus to loss modulus. The Tg value for the bio-composite determined by DMA was around 174 °C, suggesting great enhancement through the glass fiber addition compared with that (120–130 °C) of the neat PHMF resin determined by DSC, as similarly observed and reported by the authors and other researchers [7,34,35]. In comparison, the Tg of the bio-composite obtained in this study is thus higher than its counterparts, oil palm/glass hybrid

386

Y. Zhang et al. / Journal of the Taiwan Institute of Chemical Engineers 71 (2017) 381–387

Fig. 7. Comparison of curing reaction rate for curing of PHMF resins with 20 wt% DGEBA obtained by experiments (dots) and the model fitting (line) at various heating rates of 5 (a), 10 (b), 15 (c) and 20 (d) °C/min. Table 6 Tensile properties of PHMF FRC cured with DGEBA or HMTA. Sample

Tensile strength (MPa)

Young’s modulus (GPa)

Elongation at break (%)

PHMF + DGEBA PHMF + HMTA

105 ± 8 109 ± 4

14.6 ± 0.8 19.7 ± 0.9

7.2 ± 0.6 5.5 ± 0.5

fiber-reinforced PF composites, which have Tg up to 140 °C. [36] It is comparable to Tg (from 175-215°C) for DGEBA cured PF novolac resins according to a report by Ogata et al. [37]. Table 6 gives the comparison of the tensile properties of woven fiberglass cloth-PHMF resin composites cured with DGEBA with those cured with HMTA. The PHMF–DGEBA composites present a tensile strength at around 105 MPa, which is comparable with that of PHMF–HMTA composite (109 MPa, Table 6) and commercial phenolic sheet molding composites (96 MPa). [38] 4. Conclusions In this work, a formaldehyde-free bio-based phenolic resin, i.e., PHMF resin, was prepared and cured with a bisphenol A type epoxy (DGEBA). The curing mechanism was investigated by

DSC and FTIR measurements. DGEBA resin proved to be an effective cross-linker for curing of PHMF without emission of any toxic by-product. Curing of the PHMF resin and DGEBA started at 105 °C, with the peak at around 155 °C. The curing activation energy was measured to be ∼114 kJ/mol from Kissinger and FWO methods. The Ea turned to be around 110 kJ/mol, despite of which isoconversional method was used, i.e. Friedman or KAS method. Furthermore, the autocatalytic kinetic model was simulated and matched well with experimental data. The formaldehyde-free PHMF resin hardened with DGEBA was also used to produce novel fiber-reinforced plastic composite materials having high tensile fracture strength of 105 MPa and a glass transition temperature of 174 °C, suggesting the potential of the DGEBA cured PHMF resin for industrial applications as a matrix for bio-composites.

Y. Zhang et al. / Journal of the Taiwan Institute of Chemical Engineers 71 (2017) 381–387

Fig. 8. DMA profiles of the fiber-reinforced plastic bio-composite using PHMF resin cured with DGEBA.

Acknowledgments The authors are grateful for the financial support from NSERC/FPInnovations Industrial Research Chair Program in Forest Biorefinery, Grant IRCSA413630-09, and the Ontario Research FundResearch Excellence (ORF-RE) from Ministry of Economic Development and Innovation, Grant IRCPJ413631-09. Support from the industrial partners including FPInnovations, Arclin Canada and BioIndustrial Innovation Centre is also acknowledged. References [1] Ragauskas AJ, Williams CK, Davison BH, Britovsek G, Cairney J, Eckert CA, et al. The path forward for biofuels and biomaterials. Science 2006;311:484–9. [2] Netravali AN, Chabba S. Composites get greener. Mater Today 2003;6:22–9. [3] Devi A, Srivastava D. Studies on the blends of cardanol-based epoxidized novolac type phenolic resin and carboxyl-terminated polybutadiene (CTPB), I. Mater Sci Eng A Struct 2007;458:336–47. [4] Zhao Y, Yan N, Feng MW. Biobased phenol formaldehyde resins derived from beetle-infested pine barks-structure and composition. ACS Sustain Chem Eng 2012;1:91–101. [5] US Environmental Protection Agency. Extremely hazardous substances: superfund chemical profiles. Norwich, NY: William Andrew Publishing; 1988. [6] Hahnenstein I, Hasse H, Kreiter CG, Maurer G. 1H- and 13C-NMR spectroscopic study of chemical equilibria in solutions of formaldehyde in water, deuterium oxide, and methanol. Ind Eng Chem Res 1994;33:1022–9. [7] Yuan Z, Zhang Y, Xu C. Synthesis and thermomechanical property study of novolac phenol-hydroxymethyl furfural (PHMF) resin. RSC Adv 2014;4:31829–35. [8] Nielsen AT, Moore DW, Ogan MD, Atkins RL. Structure and chemistry of the aldehyde ammonias. 3. Formaldehyde-ammonia reaction. 1, 3, 5-Hexahydrotriazine. J Org Chem 1979;44:1678–84. [9] Richmond HH, Myers GS, Wright GF. The reaction between formaldehyde and ammonia. J Am Chem Soc 1948;70:3659–64. [10] Jacobs DE, Kelly T, Sobolewski J. Linking public health, housing, and indoor environmental policy: successes and challenges at local and federal agencies in the United States. Environ Health Perspect 2007;115:976–82. [11] Hsieh T, Su A. Cure kinetics of an epoxy-novolac molding compound. J Appl Polym Sci 1990;41:1271–80.

387

[12] Mak HD, Rogers MG. Determination of chain branching in epoxy resins by nuclear magnetic resonance spectrometry. Anal Chem 1972;44:837–9. [13] Batzer H, Zahir SA. Studies in the molecular weight distribution of epoxide resins. II. Chain branching in epoxide resins. J Appl Polym Sci 1975;19:601–7. [14] Han S, Yoon HG, Suh KS, Kim WG, Moon TJ. Cure kinetics of biphenyl epoxy-phenol novolac resin system using triphenylphosphine as catalyst. J Polym Sci A Polym Chem 1999;37:713–20. [15] Levchik SV, Weil ED. Thermal decomposition, combustion and flame-retardancy of epoxy resins – a review of the recent literature. Polym Int 2004;53:1901–29. [16] Gagnebien D, Madec P, Marechal E. Synthesis of poly (sulphone-b-siloxane)s – I. Model study of the epoxy-phenol reaction in the melt. Eur Polym J 1985;21:273–87. [17] Alvey FB. Selectivity of the epoxide phenol reaction. J Appl Polym Sci 1969;13:1473–86. [18] Burchard W, Bantle S, Zahir SA. Branching in high molecular weight polyhydroxyethers based on bisphenol A. Macromol Chem Phys 1981;182:145–63. [19] Brunelle DJ, Evans TL. Contemporary topics in polymer science. Plenum Press, New York: Springer US; 1992. p. 108. [20] Park SJ, Seo MK, Lee JR. Isothermal cure kinetics of epoxy/phenol-novolac resin blend system initiated by cationic latent thermal catalyst. J Polym Sci Part A Polym Chem 20 0 0;38:2945–56. [21] Kissinger HE. Reaction kinetics in differential thermal analysis. Anal Chem 1957;29:1702–6. [22] Ozawa T. A new method of analyzing thermogravimetric data. Bull Chem Soc Jpn 1965;38:1881–6. [23] Flynn JH, Wall LA. A quick, direct method for the determination of activation energy from thermogravimetric data. J Polym Sci Part B Polym Phys 1966;4:323–8. [24] Starink M. The determination of activation energy from linear heating rate experiments: a comparison of the accuracy of isoconversion methods. Thermochim Acta 2003;404:163–76. [25] Akahira T, Sunose T. Method of determining activation deterioration constant of electrical insulating materials. Res Rep Chiba Inst Technol 1971;16:22–31. [26] Li T, Qin H, Liu Y, Zhong X, Yu Y, Serra A. Hyperbranched polyester as additives in filled and unfilled epoxy-novolac systems. Polymer 2012;53:5864–72. [27] El-Thaher N, Mekonnen T, Mussone P, Bressler D, Choi P. Nonisothermal DSC study of epoxy resins cured with hydrolyzed specified risk material. Ind Eng Chem Res 2013;52:8189–99. [28] Lee YK, Kim DJ, Kim HJ, Hwang TS, Rafailovich M, Sokolov J. Activation energy and curing behavior of resol-and novolac-type phenolic resins by differential scanning calorimetry and thermogravimetric analysis. J Appl Polym Sci 2003;89:2589–96. [29] Choi WS, Shanmugharaj A, Ryu SH. Study on the effect of phenol anchored multiwall carbon nanotube on the curing kinetics of epoxy/novolac resins. Thermochim Acta 2010;506:77–81. [30] Ren S, Lan Y, Zhen Y, Ling Y, Lu M. Curing reaction characteristics and phase behaviors of biphenol type epoxy resins with phenol novolac resins. Thermochim Acta 2006;440:60–7. [31] Málek J. Kinetic analysis of crystallization processes in amorphous materials. Thermochim Acta 20 0 0;355:239–53. [32] Park BD, Riedl B, Hsu EW, Shields J. Differential scanning calorimetry of phenol–formaldehyde resins cure-accelerated by carbonates. Polymer 1999;40:1689–99. [33] Kim WG, Lee JY. Contributions of the network structure to the cure kinetics of epoxy resin systems according to the change of hardeners. Polymer 2002;43:5713–22. [34] Barbosa V, Ramires EC, Razera IA, Frollini E. Biobased composites from tannin–phenolic polymers reinforced with coir fibers. Ind Crops Prod 2010;32:305–12. [35] Zhang Y, Yuan Z, Xu C. Engineering biomass into formaldehyde-free phenolic resin for composite materials. AIChE J 2015;61:1275–83. [36] Sreekala MS, Thomas S, Groeninckx G. Dynamic mechanical properties of oil palm fiber/phenol formaldehyde and oil palm fiber/glass hybrid phenol formaldehyde composites. Polym Compos 2005;26:388–400. [37] Ogata M, Kinjo N, Kawata T. Effects of crosslinking on physical properties of phenol–formaldehyde novolac cured epoxy resins. J Appl Polym Sci 1993;48:583–601. [38] Pilato L. Phenolic resins: a century of progress. Springer Verlag; 2010.