Abstract
To strengthen the dimensional stability of enzymatically treated corn stalk (ECS) biocomposites, hybrid modified lignosulfonate (HML) was used as a binder to fabricate reinforced ECS/HML composites with evaluation by response surface methodology. The effects of the preparation treatment on the enzymatic conditions, as well as the modified lignosulfonate dosage on the physicomechanical properties of the ECS/HML composites, were all evaluated. The optimum preparation parameters were determined via the Box-Behnken experimental design. High mass concentrations of laccase-vanillin and an appropriate modified lignosulfonate dosage for a relatively short enzymatic pretreatment time led to reduced residual stresses and improved dimensional properties. The optimum conditions that minimized thickness swelling (TS) and water adsorption (WA) without significantly compromising the biocomposite mechanical properties were determined to be 25 g/L laccase-vanillin, 118.8 min enzymatic pre-treatment time, and 15 wt% modified lignosulfonate. The ECS/HML composites that were treated under the optimal conditions resulted in an approximate 42% reduction in the dimensional properties without any significant decline in mechanical properties when compared to ECS panels. Unlike the loose structure of ECS biocomposites, the ECS/HML composites had a laminar shape with firm morphology.
Download PDF
Full Article
Dimensional Stability Improvement of Corn Stalk Biocomposites Using Two-part Lignin-derived Binder Optimized with Response Surface Methodology
Yuan Yuan,a Sidan Li,b Feng Jiao,c Guinan Shen,a Lei Yan,a and Weidong Wang a,*
To strengthen the dimensional stability of enzymatically treated corn stalk (ECS) biocomposites, hybrid modified lignosulfonate (HML) was used as a binder to fabricate reinforced ECS/HML composites with evaluation by response surface methodology. The effects of the preparation treatment on the enzymatic conditions, as well as the modified lignosulfonate dosage on the physicomechanical properties of the ECS/HML composites, were all evaluated. The optimum preparation parameters were determined via the Box-Behnken experimental design. High mass concentrations of laccase-vanillin and an appropriate modified lignosulfonate dosage for a relatively short enzymatic pretreatment time led to reduced residual stresses and improved dimensional properties. The optimum conditions that minimized thickness swelling (TS) and water adsorption (WA) without significantly compromising the biocomposite mechanical properties were determined to be 25 g/L laccase-vanillin, 118.8 min enzymatic pre-treatment time, and 15 wt% modified lignosulfonate. The ECS/HML composites that were treated under the optimal conditions resulted in an approximate 42% reduction in the dimensional properties without any significant decline in mechanical properties when compared to ECS panels. Unlike the loose structure of ECS biocomposites, the ECS/HML composites had a laminar shape with firm morphology.
Keywords: Corn stalk; Enzymatic pretreatment; Hybrid modified lignosulfonate; Box-Behnken design; Physicomechanical properties
Contact information: a: Heilongjiang Provincial Key Laboratory of Environmental Microbiology and Recycling of Argo-Waste in Cold Region, College of Life Science and Technology, Heilongjiang Bayi Agricultural University, Daqing 163319, China; b: Institute of New Rural Development, Heilongjiang Bayi Agricultural University, Daqing 163319, China; c: College of Agronomy, Heilongjiang Bayi Agricultural University, Daqing 163319, China; *Corresponding author: wwdcyy@126.com
INTRODUCTION
To decrease the emission of toxic volatiles and to protect both the environment and human health, considerable efforts have been devoted to the development of particleboards using natural-derived binders (Widyorini et al. 2005a). Such particleboards are biocomposites that are formed without the presence of synthetic resin. It has been shown that their self-bonding strength could be improved by activating the chemical components of board constituents during steam or heat treatment (Widyorini et al. 2005b). The main methods employed for this purpose are based on heating or biological treatments (González-García et al. 2011; Wang et al. 2011, 2013; Wu et al. 2011), steam explosion (Xu et al. 2004), and combined processes. These approaches are often less hazardous than panel manufacturing with formaldehyde-based adhesives. In addition, bioproducts prepared without synthetic binders are biodegradable, recyclable, and can be disposed of in environmentally friendly ways (Nadhari et al. 2013).
Composite panels can be made of several crops (Thamae et al. 2008; Shah 2013), such as corn stalk, wheat straw (Wang and Sun 2002), kenaf core (Xu et al. 2003), cotton stalk (Zhou et al. 2010), rice husk (Ndazi et al. 2006), and castor stalk (Grigoriou and Ntalos 2001). Corn is the most produced cereal worldwide, surpassing wheat and rice (Jarabo et al. 2013). The production of corn and its wastes increases each year. Corn residues are industrial raw material sources that can potentially be used in several applications, including human consumption, energy (Yan et al. 2012), fuel (Gao et al. 2017), carbon materials, and chemical production (Ioannidou et al. 2009). The industrial usage of agricultural residues from the forest industry may effectively reduce the production cost, increase the sustainability of the panel manufacturing industry (Juárez et al. 2007), and minimize air pollution caused by the combustion of corn residues. However, farmers harvest only the grains, and most of the residues, including the stalks and husks, are burned or disposed of due to several limitations like collection cost and farming conditions (Çöpür et al. 2007). For biocomposites, Wu et al. (2011) manufactured binderless fiberboards from corn stalk that was pretreated with white-rot fungus. The pretreatment process significantly increased the mechanical properties and improved the crystallinity of the fiberboards without using any synthetic adhesives. However, most binderless biocomposites exhibit deficiencies that limit their industrial applications (Yuan et al. 2011). Therefore, the methods for reinforcing the interactions between components are key for improving the physicomechanical properties of biocomposites prepared without synthetic binders.
Lignin significantly affects the manufacturing of binderless biocomposites. Most studies have focused on technical lignin, such as kraft lignin (Velásquez et al. 2003; Mancera et al. 2011a, 2011b) and lignosulfonate (Jin et al. 2011; Ji and Guo 2018; Ji et al. 2018), both of which are generated from the papermaking industry. However, the low reactivity of lignin has limited its commercial use; thus it is often discarded or burned to generate energy or recover chemicals. Less than 5% of the world’s global lignin supply has been used as low-value compounds (Huet al. 2011). Lignin oxidation with H2O2 may effectively improve its safe utilization because water is used to break it down, replacing organic solvents and reducing environmental issues. Alkaline aqueous solutions have been determined to be the best reaction medium compared to acidic or neutral environments (Hu and Guo 2015).
Hybrid modified lignosulfonate (HML) is a type of lignin-based binder with modified ammonium lignosulfonate and polyethylenimine (Yuan and Guo 2014; Yuan et al. 2014). Hybrid modified lignosulfonate has a better environmental impact in wood fiber composites because it is a formaldehyde-free binder (Yuan and Guo 2016, 2017). However, only a handful of studies so far have reported on applications related to modified technical lignin in biocomposites using enzymatic treatments. In this study, the feasibility of using HML as a natural binder and enzymatic pretreated corn stalk (ECS) particles as raw materials to manufacture biocomposites was evaluated at a representative factory that had an approximate production capacity of 145,000 m3/year in Harbin, an urban region in the northeast of China. The Box-Behnken design (BBD) was applied to investigate the effects of variable interactions on the physicomechanical properties of ECS/HML composites. Fourier transform infrared spectrometry (FTIR) and scanning electron microscopy (SEM) were also employed to compare the chemical structures of the binderless biocomposites before and after the introduction of HML.
EXPERIMENTAL
Materials
Corn stalks obtained from Anda (Heilongjiang province, China) were air-dried and stored in jute bags. The corn stalks were chipped through a hacker chipper and then reduced into smaller particles using a knife ring flaker (FW-100 high-speed shredder; HuaYi Instrument Co., Ltd., Changzhou, China). The obtained particles were dried to a moisture content of 5% and passed through 40-mesh to 60-mesh sieve for separation, then stored for the manufacturing of corn stalk biocomposites. The content of ash according to GB/T 742 (2008), ethanol-benzene extractives according to GB/T 10741 (2008), Klason lignin according to GB/T 747 (2003), holocellulose according to GB/T 2677.10 (1995), and α-cellulose according to GB/T 744 (2004) were measured. The hemicelluloses content was calculated by subtracting the cellulose content from the holocellulose content. The average chemical compositions of the initial particles were determined to be 4.6% ash, 14.9% extractives, 16.7% lignin, 45.6% cellulose, and 22.5% hemicelluloses. Due to the overlap in the test parameters, the total chemical composition exceeded 100% (104.4%), which was expected (Angles et al. 1997). Ammonium lignosulfonate from Shenyang Xingzhenghe Chemical (Shenyang, China) was used as it was received, with the composition content determined to be 51.9% total lignin, 27.1% carbohydrates, 10.6% ash, and 4.6% moisture. Laccase was purchased from Wuhan Yuancheng Technology Development (Wuhan, China) and stored at -10 °C. The activity of the commercial laccase powder was approximately 4000 U/g. Vanillin (4-hydroxy-3-methoxybenzaldehyde) was purchased from Shanghai Adams Reagent Co., Ltd. (Shanghai, China). Polyethylenimine (PEI) was obtained from Shanghai UN Chemical (Shanghai, China). The molecular weight of PEI was 75,000 in 50 wt% aqueous solution. All other chemicals used were of analytical grade.
Enzymatic treatment of the corn stalks
The corn stalk particles were first pretreated with laccase-vanillin according to previous reports (Yuan and Guo 2013). Briefly, the dried corn stalk particles were suspended in deionized water to yield a 5.0 wt% suspension at a pH of 5, and they were then mixed using a JJ-1 precision-power motor stirrer (Changzhou Wanfeng Instrument Co., Ltd., Changzhou, China). Next, the suspension was stirred with an oxygen stream at temperatures from 45 °C to 47 °C for 180 min. The mass ratio of laccase to vanillin was set to 2.56:100 for 1 L of the desired oxidation system. After the pretreatment, the particles were transferred to gauze for dewatering and were air-dried at 30 °C. They were then dried for 1 h at 100 °C to eliminate enzyme reaction. The target moisture content of the pretreated corn stalk particles was set between 4% and 6%.
Preparation of the HML
Two-part lignin-derived binder was prepared through an oxidation reaction and combination technology using ammonium lignosulfonate (AL) and PEI according to previous reports (Yuan et al. 2014). Briefly, 50 g of AL powder was alkalinized in 100 g water adjusted to a pH of 10. Next, 30% H2O2 dosage based on dry weight to lignosulfonate was added to the solution. After heating for 30 min at 60 °C, the mixture was concentrated to form a 20 wt% modified ammonium lignosulfonate (MAL) solution. Hybrid modified lignosulfonate was prepared by mixing MAL with PEI at a weight ratio of 7:1 for 30 min.
Manufacturing of the ECS/HML composites
The ECS particles were mixed with HML at different proportions in a SHR-10A high-speed blender (Zhangjiagang Yunfan Machinery Co., Ltd., Zhangjiagang, China). The mixed particles were fixed into the mat of a 250 mm × 250 mm forming box. The target density of all biocomposites was determined to be 0.8 g/cm3 ± 0.03 g/cm3 with a target thickness of 5 mm. For reproducibility, each group of experiments under these sets of conditions was replicated three times. The corn stalk biocomposites issued from ECS were prepared and used as controls.
Methods
Physicomechanical properties of ECS/HML composites
For each test, the ECS/HML composites were cut into three test samples according to GB/T 17657 (2013) after conditioning at 20 °C ± 2 °C and 65% ± 5% relative humidity (RH). The test samples were set to 200 mm × 50 mm for the modulus of rupture (MOR) and the modulus of elasticity (MOE) tests, and 50 mm × 50 mm for the internal bonding strength (IB) test. A loading speed of 5 mm/min was selected for the MOR and the MOE tests and 2 mm/min for the IB test. The thickness swelling (TS) and water absorption (WA) were also set to 50 mm × 50 mm using three replicates. The specimens were first immersed in water at 20 °C ± 1 °C and then their thickness and weight changes were measured after 24 h. The load-bearing particleboard properties of GB/T 4897 (2015) were followed as MOR ≥ 15 MPa, MOE ≥ 2.2 GPa, IB ≥ 0.45 MPa, and 24 h TS ≤ 22%.
FTIR and SEM characterization
The FTIR spectroscopy of prepared corn stalks, lignosulfonates, and ECS/HML specimens were characterized using a Nicolet Magna-IR 560 (ThermoFisher Scientific, Madison, USA). The spectra were recorded at wavelengths ranging from 4000 cm-1 to 500 cm-1. Each sample was scanned 40 times at resolution of 4 cm-1. Following the IB tests, the SEM results were collected to evaluate the morphological changes of the ECS panels and the ECS/HML composites on a Sirion 200 (FEI Company, Hillsboro, USA) microscope.
Experimental design
A Box-Behnken experimental design with three independent parameters and three various levels was adopted using Design-Expert 8.0.6 software (Stat-Ease Inc., Minneapolis, USA). A total of 17 experiments at a central point were employed to determine the variables that influence the ECS/HML composites’ performance. This method allows the establishment of statistical relationships between the experimental variables and response variables to describe the nature of the response surface and elucidate the optimal manufacturing conditions. These features should, in turn, allow for predicting the optimal board properties. Table 1 lists the design matrix and mechanical properties data of the obtained ECS/HML composites.
The three critical parameters that affected the physicomechanical properties of the ECS/HML composites were the mass concentration of laccase-vanillin (XA), the enzymatic pretreated time (XB), and the modified lignosulfonate dosage (XC). These parameters were selected as the independent variables based on preliminary experiments. The dependent variables (response), which include MOR, MOE, IB, 24 h TS, and 24 h WA, were then evaluated. An analysis of variance (ANOVA) was performed for each response at a confidence level of 95%. All data were expressed using the average of three replicates along with their coefficient of variation (CV). The CV observations for the samples from the ECS/HML composites compared well with those of the control.
Table 1. Experimental Design of Coded Factors and Results of BBD for Physicomechanical Properties of ECS/HML Composites
*Coefficient of variation
RESULTS AND DISCUSSION
Data Analysis and Regression Models
The analysis of variance p-values for the mechanical properties of the ECS/HML composites are presented in Table 2. All p-values below 0.05 revealed significant model terms, while values above 0.05 indicated insignificant model terms (Alslaibi et al. 2013). Meanwhile, p-values below 0.0001 would imply that all models of mechanical properties were significant and there is only a 0.01% chance that such values could occur due to noise.
Table 2. Analysis of Variables for p-Value of Parameters and Their Interactions
ns Not signficant
From the ANOVA results (Table 1), the enzymatic pretreated time levels showed no significant effect on the physicomechanical properties, except for IB and WA. The factors positively affected the physicomechanical properties under different conditions, except for IB that was lower at low-level preparation conditions.
Table 3. Regression Models of Mechanical Properties and Dimensional Stability of ECS/HML Composites
Fig. 1. Response surface plots for MOR as a function of: (a) XA vs. XB, (b) XA vs. XC, and (c) XB vs. XC
A series of estimates yielded five quadratic models associating the mechanical properties to the preparation conditions. These included XA, XB, and XC (Table 3). The models fit well, and all R2 values were higher than 0.98. All predicted R2 values agreed with the adjusted R2 values. Values of adequate precision greater than 4 are desirable (Muthukumar et al. 2003). The effects of MOR, MOE, and IB on the mechanical properties and dimensional stability (TS and WA) were investigated more deeply in later sections.
Mechanical Properties
To further analyze the effects of parameters XA, XB, and XC on the mechanical properties of ECS/HML composites, the response surfaces were plotted as three-dimensional (3D) plots. From the generated data obtained from the same test, the MOR and MOE were analyzed together. The fitted models for MOR and MOE yielded R2 values of 0.9918 and 0.9827, respectively. Table 2 shows that the variables XA and XC significantly affected the MOR and MOE while variable XB was insignificant.
Figures 1a to 1c show the effects of 3D response surfaces of XA, XB, and XC on MOR according to the quadratic mathematical model. At a constant enzymatic pretreatment time (Fig. 1a), the MOR values significantly increased as the mass concentration of laccase-vanillin increased. At a constant mass concentration of laccase-vanillin (1) (Fig. 1b), changes in the trends of MOR were divided into two stages. First, the modified lignosulfonate dosage (XC) ranged from -1 to 0. As the XC increased, the MOR value increased, indicating that the crosslinking of the modified lignosulfonate with PEI improved the mechanical strength. Therefore, lignin cross-linked with PEI could be used as a corn panel adhesive. Second, the XC ranged from 0 to 1.
The MOR values decreased as the modified lignosulfonate dosage increased, suggesting that high amounts of HML could deteriorate the mechanical performance of the ECS/HML composites. The negative value of XB meant that the decreased enzymatic pre-treatment time increased the MOR values (Fig. 1c). Therefore, excess lignin-based binder can be cured before the final formation of ECS/HML composites. The high amounts of cured lignin-based binder made the samples fragile, which led to the mechanical damage. The same influencing trend was investigated for the MOE and the same explanations of MOR were valid for MOE (Figs. 2a to 2c). The maximum MOR and MOE values of the ECS/HML composites determined at a constant enzymatic pretreatment time of 120 min were 29.18 MPa and 3.148 GPa, respectively.
The response surface plots for the IB are shown in Figs. 3a to 3c. The IB represented the strength of the bonding between the particles and should be considered to ensure that the panels do not delaminate during post-processing. The fitted model for IB yielded an R2 value of 0.9890. Thus, all single factors and interactions between the mass concentration of laccase-vanillin and other variables (XAXB, XAXC) were significant for the IB (Table 2). The effects of all variables on IB depicted similar trends. The internal bonding strength generally increased with XAwith XB and with the enzymatic pretreatment time (XA). However, Fig. 3 shows that IB was significantly affected by XA and XC but moderately affected by time (XB) in the ECS/HML composites. This suggested that the lignin of corn stalk particles melted well at selected enzymatic pretreated concentrations and selected times, forming strong interparticle bonds at lignin and cellulosic surface areas (Back 1987). Hence, the mechanical properties of ECS/HML composites decreased in the presence of excess amounts of HML. Thus, increased modified lignosulfonate dosage and enzymatic concentration of laccase-vanillin improved the mechanical properties of ECS/HML composites below 15 wt% HML.
Fig. 2. Response surface plots for the MOE as a function of: (a) XA vs. XB, (b) XA vs. XC, and (c) XB vs. XC
Fig. 3. Response surface plots for IB as a function of: (a) XA vs. XB, (b) XA vs. XC, and (c) XB vs. XC