An eco-friendly wood adhesive from soy protein and lignin: performance properties

Abstract

The aim of this study was to improve the water resistance of a soybean meal-based adhesive with a low-cost lignin-based resin (LR). The LR could react with active groups on the protein molecules and itself to form a cross-linking and interpenetrating network, which improved the water resistance of the resultant adhesive. In addition, adding LR increased the thermal stability of the cured adhesive, forming a cross-linked protein molecule-based structure, and created a smooth surface with fewer holes and cracks to prevent moisture intrusion, which further improved the water resistance of the resultant adhesive. Using LR also resulted in an appropriate viscosity which benefited adhesive distribution during the plywood hot press process and formed a stronger interlock with wood, thus creating good wet shear strength of the resultant plywood. Incorporating 10 wt% LR effectively improved the wet shear strength of the resultant plywood by 200% to 1.05 MPa, which met interior-use plywood bond strength requirement. The resultant adhesive had a solid content of 32.65% and viscosity of 499 400 mPa s, which were acceptable for industrial application in plywood fabrication.

---

1 Introduction

Formaldehyde-based resins, such as urea-, phenol-, and melamine-formaldehyde resins, are present in common wood adhesives. They perform very well, but one of their major disadvantages, from the environmental point of view, is that they are derived from non-renewable and limited fossil resources. Furthermore, urea- and melamine-formaldehyde resins also have the issue of formaldehyde emission, which is hazardous to human health.¹ Therefore, there is an urgent need for developing adhesives based on environmentally friendly and renewable resources.²

Soy protein-based adhesives are one of researchers' focuses since soybean is abundant, inexpensive, and environmentally friendly.³ However, the low bond strength and water resistance of the soy protein-based adhesives limit their application.⁴ Efforts to improve adhesion properties of the soy protein-based adhesives include using denaturants and cross-linking agents.⁵ Denaturants, such as urea, sodium dodecyl sulfate, and alkali, can unfold the protein molecules and these molecules rearrange during the plywood hot press process, which improves the water resistance of the adhesive.⁶ But the resultant plywood bonded by adhesives with these denaturants cannot meet the bond strength requirement of interior use. Using crosslinking agents, such as glycidyl methacrylate, polyamidoamine-epichlorohydrin resin, or polyethylene glycol diacrylate, latex-based adhesive could effectively improve the water resistance of soy protein-based adhesive and the resultant plywood met the interior-use bond strength requirement.^7–10 However, those additives are expensive and greatly increase the cost of the resultant adhesive, which limits the industrial application of soy protein-based adhesives.

As the third most abundant natural biological macromolecule, lignin widely exists in plant resources, such as wood, corncob, and sorghum. It is a low-value by-product of the paper-making and lignocellulosic ethanol industries. Lignin has an aromatic and cross-linked structure comprising three main lignin building units, p-hydroxyphenyl, guaiacyl, and syringyl units, linked by carbon–carbon and ether bonds. As an aromatic compound available from renewable resources, the depolymerization of lignin has the potential to produce bio-based polymers. Since some parts of the lignin structure are similar to that of phenol, it is a potential substitute for phenol in phenol-formaldehyde resin synthesis.¹¹ However, the fewer reactive sites and the stronger steric hindrance effect of lignin mean it has low reactivity and the resultant resin with high lignin content has a poor performance. Considering the possibility of generating coupling reactions of side chains and link-bonds of functional groups, a raw sorghum lignin was used to mix with defatted soy flour to improve the water resistance of a soy protein-based adhesive.^12,13 But the improvement was not significant and it was difficult for the resultant adhesive to meet the bond strength requirement of industrial use.

In this study, corncob lignin was used to synthesize a lignin-based resin (LR). And then, the LR was mixed with a soybean meal-based adhesive to develop a plywood adhesive. The performance of the resultant adhesive including the viscosity of the adhesive, and the functional groups, thermostability, and fracture surface of the cured adhesive was characterized. Three-ply plywood specimens were fabricated with the resultant adhesive and their wet shear strength was tested.

2 Experimental

2.1 Materials

Soybean meal (SM) was obtained from Xiangchi Grain and Oil Company in Shandong Province of China, and then milled to 200 mesh flour. Components of the SM flour were tested as follows: 46.88% soy protein, 40.24% saccharides, 5.86% moisture, 6.46% ash, and 0.56% fat. Phenol, formaldehyde solution (37–40 wt%) and sodium hydroxide were AR grade reagents and purchased from Tianjin Chemical Reagent Co. Corncob lignin was purchased from Shandong Long Li Biological Technology Co. Ltd, China. The lignin was produced from the corncob residue after hydrolysis of hemicelluloses, which contained 88.5% acid-insoluble lignin, 4.7% acid-soluble lignin, 0.10% arabinose, 0.12% glucose, and 0.08% xylose. The ash and moisture contents of the lignin were 1.3% and 5.2%, respectively. Poplar veneer (40 × 40 × 1.5 cm, 8% moisture content) was provided from Hebei Province of China.

2.2 Preparation of the lignin-based resin (LR)

The LR was synthesized by batch copolymerization according to lignin, phenol, and formaldehyde in a weight ratio of 150 : 140 : 132 (100% solid content). In the first step, lignin (150 g) and phenol (140 g) with formaldehyde (146 g, 37 wt%) and NaOH (45 g, 50 wt%) were mixed and stirred in a flask. The mixture was heated to 80 °C and kept for 1 h. In the second step, a second portion of formaldehyde (97 g, 37 wt%) and NaOH (20 g, 50 wt%) was added into the flask and stirred for 1 h at 80 °C. In the last step, the rest of the formaldehyde (113 g, 37 wt%) and NaOH (20 g, 50 wt%) were added to the flask and further stirred for 1 h at 80 °C. Then the mixture was cooled down to 40 °C and, through a vacuum distillation process, free formaldehyde was removed to get the LR. The solid content of the resultant resin was 60% and the viscosity was 380 mPa s at 20 °C. The reaction equation is presented in Fig. 1.

Fig. 1 The reaction equation for the formation of LR.

2.3 Preparation of soybean meal-based adhesives

For the SM adhesive, SM flour (28 g) was added into deionized water (72 g) and stirred for 10 min at 20 °C. 5, 10, 15, 20, and 25 wt% of LR were mixed with the SM adhesive, and further stirred for 10 min at 20 °C to develop the SM/LR5% adhesive, the SM/LR10% adhesive, the SM/LR15% adhesive, the SM/LR20% adhesive, and the SM/LR25% adhesive, respectively.

2.4 Preparation of the plywood samples

Three-ply plywood samples were prepared under the following conditions: 180 g m⁻² glue spreading for a single surface, 70 s mm⁻¹ hot press time, 135 °C hot press temperature, and 1.2 MPa hot press pressure.¹⁴ After hot pressing, the plywood samples were stored under ambient conditions for at least 12 h before testing.

2.5 Solid content measurement

The adhesive solid content was determined using an oven-drying method. Approximately 3 g (weight α) of the adhesive was placed into an oven and dried at 105 °C for several hours until a constant weight (weight β) was obtained. The value of the solid content was calculated using the following equation. The average value of the solid content was calculated from three parallel samples.

2.6 Dynamic viscoelastic measurement

The apparent viscosity of the different adhesives was determined using a rheometer with a parallel plate fixture (20 mm diameter). The distance was set to 1 mm for all measurements. Experiments were conducted under a steady shear flow at 25 °C. Shear rates ranged from 0.1 to 240 s⁻¹ in 10 s⁻¹ increments. All of the measurements were conducted in triplicate, and the average value was determined.

2.7 Wet shear strength measurement

The wet shear strength of the interior-use plywood (Type II plywood) was determined using a shear strength test in accordance with the description in China National Standards (GB/T 17657-1999). Twelve plywood specimens (2.5 cm × 10 cm) were cut from two plywood panels and submerged into water at 63 ± 2 °C for 3 h, and then dried at a room temperature for 10 min before tension testing. The wet shear strength was calculated by the following equation:

2.8 Residual rate test

The adhesive samples were placed in an oven at 120 ± 2 °C until a constant weight (M) was obtained. The cured adhesives were soaked in tap water for 24 h at ambient temperature, then oven-dried at 105 ± 2 °C for 5 h, until a constant weight was obtained (m). The residual rate is defined as m divided by M, as in eqn (3). The average value of the residual rate was calculated from six parallel samples.

2.9 Fourier transform infrared (FTIR) spectroscopy

The different adhesive samples prepared as described in 2.3 were cured in an oven at 120 ± 2 °C until a constant weight was obtained and then ground into a powder. FTIR spectra of the different cured adhesives were recorded using a Nicolet 7600 spectrometer (Nicolet Instrument Corporation, Madison, WI) from 500 to 4000 cm⁻¹ with a 4 cm⁻¹ resolution using 32 scans.

2.10 Thermogravimetric (TG) measurement

The different adhesives were cured in an oven at 120 ± 2 °C until a constant weight was obtained and ground into a powder. The thermal stabilities of the cured adhesive samples were tested using a TGA instrument (TA Q50, Waters Company, USA). About 5 mg powdered samples were weighed in a platinum cup and scanned from 30 to 600 °C at a heating rate of 10 °C min⁻¹ in a nitrogen environment while recording the weight change.

2.11 Scanning electron microscopy (SEM)

The different samples were poured into a piece of aluminum foil and dried in an oven at 120 ± 2 °C until a constant weight was achieved. A Hitachi S-3400N (Hitachi Science System, Ibaraki, Japan) scanning electron microscope was used to observe fractured surfaces of the different adhesive samples. The surface was sputter coated with gold prior to examining it under the microscope.

3 Results and discussion

3.1 Solid content measurement

Solid content is a basic physical parameter for a wood adhesive that influences the performance of the adhesive during the hot press process. In general, the adhesive properties are improved with solid content. A low solid content of the adhesive indicates a larger water content leading to an excessive diffusion into the veneer resulting in poor bonding at the interface due to insufficient adhesive.^15,16 The average value of solid content of the different SM/LR adhesive formulations was calculated from three parallel samples and is shown in Fig. 2. Soy protein has big molecules with a complex structure and is easy to swell in water, which presents a high viscosity. As expected, for the SM adhesive, the 27.36% solid content caused a significant viscosity and a further increase in the solid content will be determined by the additive in the SM adhesive system. The low solid content of the SM adhesive also led to a lack of adhesive in plywood and a long hot press time, which is not appreciated in the plywood fabrication industry.¹⁷ Because of LR's high solid content, the solid content of the SM/LR adhesive increased from 30.11 to 34.85% with LR addition from 5 to 25 wt%. Moreover, with LR addition up to 10 wt%, the solid content of the adhesive was 32.65%, which was increased by 19.3% compared to the SM adhesive. According to a literature review, the solid content of soy protein-based adhesives ranges from 32 to 36%.¹⁵ Therefore, the adhesive with more than 10 wt% LR addition meets the requirement for plywood adhesive application.

Fig. 2 The solid content of the different adhesive samples.

3.2 Dynamic viscoelastic measurement

The apparent viscosity of the SM/LR adhesives is shown in Fig. 3 and the initial viscosity of the different adhesives is summarized in Table 1. As the LR addition increased from 5 to 15 wt%, the initial viscosity of the resultant adhesive increased from 35 810 to 686 600 mPa s (Table 1). This is because the LR contains free sodium hydroxide which is an effective protein denaturing agent. With the increase of the LR, more protein molecules unfold and increase the friction among these molecules, which greatly increases the adhesive viscosity. In addition, this denaturing process also exposes lots of interior functional groups able to form a large number of intermolecular forces (hydrogen bonds, hydrophobic interactions) between protein molecules, which further increases the viscosity of the adhesives. The viscosity of the adhesive decreased from 686 600 to 73 300 mPa s when the addition of LR increased from 15 to 25 wt%. This is attributed to two factors. The first one is the low viscosity of the LR itself. The other one is that the protein molecule begins to degrade as the sodium hydroxide increases, which further decreases the viscosity of the resultant adhesive.

Fig. 3 The apparent viscosity of the different adhesive samples.

Table 1 The initial viscosity of the different adhesive samples

LR addition (wt%)	0	5	10	15	20	25
Initial viscosity (mPa s)	35 810	296 700	499 400	686 600	181 200	73 300

---

3.3. Wet shear strength measurement

Soy protein can be used as an adhesive because the protein molecules are able to disperse and unfold in solution. The unfolded molecules increase the contact area and adhesion to other surfaces. In addition, these molecules entangle with each other during the curing process to produce bond strength.¹⁸ However, this kind of bond is easily broken by moisture, which is the main disadvantage of the adhesive. The wet shear strength of plywood bonded by the different adhesives and the residual rate of the different adhesives are shown in Fig. 4 and 5, respectively. The wet shear strength of the plywood bonded by the SM adhesive was 0.35 MPa, which failed to meet the interior-use bond strength requirement (≥0.7 MPa). From Fig. 5 the residual rate of the SM adhesive was the lowest one, 76.6%, indicating that the SM adhesive had the lowest cross-linking structure among all the adhesive formulations. The wet shear strength of the plywood bonded with adhesive containing 5 wt% LR was 0.78 MPa, which met the interior-use panel bond strength requirement. However, from the standard deviation at 5 wt% addition, the wet shear strength of the plywood specimens cannot be guaranteed to be over 0.70 MPa. After further increasing the LR addition to 10 wt%, the wet shear strength increased by 200.0% compared with the SM adhesive and reached a maximum value of 1.05 MPa which met the interior-use plywood bond strength requirement. Also, the residual rate of the adhesive was improved by 3.8% with LR addition from 5 to 10 wt%. This was due to three reasons. First of all is the denaturing process caused by the NaOH in the LR. Native soy proteins have a highly ordered global structure with hydrophilic groups exposed outside and hydrophobic groups buried inside. NaOH, as an effective protein denaturing agent, unfolds the soy protein molecules and exposes the inside hydrophobic groups, which could prevent moisture intrusion to improve the water resistance of the adhesive.¹⁹ In addition, according to our previous research on X-ray diffraction patterns of the soy protein adhesive and its denatured adhesive,²⁰ after using a denaturing agent, the unfolded molecules of the protein rearranged during the curing process and formed a more crystalline structure which improved the water resistance of the adhesive. The second reason is the cross-linking effect of LR. The amino acids could dominate soy protein properties because many side chains connect to these molecules and interact with various inorganic, organic materials and cellulosic fibers. In this adhesive formulation, the functional groups in lignin, such as phenolic, alcoholic hydroxyl group, active hydrogen of benzene ring, are transferred to phenolic hydroxyl methyl groups (Fig. 1) which could react with the amino and hydroxyl groups on the soy protein molecules and form a cross-linking network between LR and soy protein molecules, which results in a solid structure to improve the water resistance. The third reason is that the LR molecules also could react with themselves to form an interpenetration structure with soy protein molecules, which further increased the water resistance of the adhesive (Fig. 7). The cross-linked structure caused by LR/protein and LR itself also increased the residual rate of the adhesive.

Fig. 4 The wet shear strength of the different adhesive samples (error bars represent standard deviations from twelve replicates).

Fig. 5 The residual rate of the different adhesive samples (error bars represent standard deviations from six replicates).

By increasing the LR content to 15 wt%, the residual rate of the adhesive further increased by 2.9% when compared with the adhesive with 10 wt% LR, indicating a more cross-linked structure in the adhesive. In general, the water resistance of the adhesive increased with the cross-linking structure in the adhesive. However, from Fig. 4, the wet shear strength of the plywood bonded by the adhesive with 15 wt% LR addition decreased to 0.91 MPa and the viscosity of the adhesive increased by 37.5% compared with the adhesive with 10 wt% LR (Table 1). This was because a very high viscosity was detrimental to the flowability of the adhesive and caused an adhesive distribution issue which reduced the wet shear strength of the plywood. Also, it was difficult for the adhesive with a very high viscosity to penetrate into wood gaps and form an interlock with the wood, which further reduced the wet shear strength of the resultant plywood. By further increasing the LR content to 25 wt%, the wet shear strength continued decreasing and then levelled off (0.65 MPa). This was attributed to three reasons. One was that the excess NaOH along with LR addition decomposed the soy protein molecules into small pieces and reduced the bond property of the adhesive. The second was because using the LR only as a wood adhesive gave a low water resistance. The last one was the very low viscosity of the adhesive, which caused an over-penetration into wood and low bond strength.

3.4. FTIR spectroscopic analysis

The FTIR spectra of the SM adhesive and its hybrid adhesives are presented in Fig. 6. A peak observed at approximately 3319 cm⁻¹ was related to the bending vibrations of free and bound O–H and N–H groups, which could form hydrogen bonds with the carbonyl group of the peptide linkage in the protein. The peak observed at approximately 2930 cm⁻¹ was attributed to the symmetric and asymmetric stretching vibrations of the –CH₂ group in the different adhesives. The main absorption bands of the peptide were related to peaks at approximately 1656, 1536, and 1241 cm⁻¹, which were characteristic of amide I (C=O stretching), amide II (N–H bending) and amide III (C–N and N–H stretching), respectively.²¹ The bands corresponding to C–O bending were located at 1052 cm⁻¹.

Fig. 6 FTIR spectra of the different adhesive samples.

With LR addition increasing from 0 to 10 wt% in the adhesive, the absorption peaks of amide II and amide III (1536 and 1396 cm⁻¹) gradually decreased, which was attributed to the reaction between functions of the LR and the –NH groups on soy protein molecules. This reaction decreased the amount of hydrophilic groups in the adhesive, which increased the water resistance of the adhesive. In addition, LR cross-linked the soy protein molecules to form a solid network, which increased the cross-linking degree of the adhesive and the water resistance of the adhesive was further improved.²² (Most likely the specific reaction shown in Fig. 7.) Furthermore, as the LR increased, the LR could react with itself to form a LR network and penetrate the soy protein network to form an interpenetrating network, which also improved the water resistance of the adhesive. After further increasing the LR addition to 25 wt%, the absorption peak of amide II became negligible and eventually disappeared in the spectrum of the SM/LR adhesive with 25 wt% LR addition, indicating that the addition of LR was excessive. This excessive LR formed a key structure with low water resistance, so that the water resistance of the adhesive system was reduced. The absorption peak of amide I showed a minor decrease with increasing LR content, which was attributed to the protein content in the adhesive formulation decreasing as the LR content increased.

Fig. 7 The curing process of adhesive enhanced with LR.

3.5. TG analysis

Fig. 8 shows the thermogravimetric (TG) and derivative thermogravimetric (DTG) curves of the SM/LR adhesives. The thermal degradation process of the adhesives could be divided into three stages. The first stage (I) was attributable to the evaporation of residual moisture at a temperature region of 50–220 °C. No degradation of soy protein was found and the weight loss ratio was very small. The second stage (II) was the initial degradation stage from 210 °C to 280 °C, which resulted from the weight loss of the small molecules subject to degradation and the breaking of some unstable chemical bonds. The third stage (III) was the skeleton structure degradation stage, at a temperature region of 280–360 °C, attributed to the cross-linking network structure degradation. Before the first degradation stage, the small weight loss was attributable to the evaporation of residual moisture.²³ After the third degradation stage, a further heating caused breakages of C–C, C–N, and C–O linkages, soy protein backbone peptide bonds were decomposed, and gases such as CO, CO₂, NH₃, and H₂S were produced.²⁴ The decomposition of residual modifier should be taken into consideration and belonged to this degradation stage.

Fig. 8 The thermogravimetric (TG) and derivative thermogravimetric (DTG) curves of the different adhesive samples.

Compared with the SM adhesive, the adhesives with 5 and 10 wt% LR additions showed a strong peak (derivative weight) in stage II, which was attributed to the unstable bonds formed by LR and protein molecules during the curing process of the resultant adhesive. Further increasing the LR addition decreased that peak, which meant the main structure of the cured adhesive changed as the LR addition increased. For the adhesives with 5 and 10 wt% LR addition, LR reacted with protein molecules and formed a cross-linked protein molecule network presenting a major structure in the adhesive system, which improved the water resistance of the resultant adhesive. As the LR addition further increased, the reactive groups on the protein deceased (FTIR analysis) and the protein decomposed (viscosity analysis), so that the LR self-cross-linked and formed a network, being a major structure in the adhesive system. This LR network was easily broken by moisture, which presented a low water resistance. In addition, the protein was more likely acting as a filler in the adhesive system with 25 wt% LR addition, which further decreased the water resistance of the adhesive.

In the third stage, the thermal degradation behavior showed evident distinction with and without LR. The SM adhesive had the highest degradation rate which was decreased significantly with LR addition gradually increasing in the adhesive formulations. The relative weight losses in the third stage of the adhesive with LR addition from 0 to 25 wt% were 37.6, 32.0, 25.6, 22.8, 17.2 and 15.5%, respectively, which suggested a better thermal stability of the SM/LR adhesive.

3.6. SEM analysis

The fracture surface micrographs of the cured SM/LR adhesives are shown in Fig. 9. A large number of holes and cracks were observed on the fracture surface of the SM adhesive. In addition, the whole fracture surface appeared very loose and disordered. These holes and cracks were formed by water evaporation in the adhesive during the hot press process. Moisture could intrude into those cracks and swell to break bonds, which reduced the water resistance of the SM adhesive.²⁵ After 10 wt% of LR was introduced, no holes and cracks were observed, and the fracture surface of the cured adhesive became smoother and more compact. This was caused by the reaction between the LR and soy protein to form a cross-linked network, which could effectively prevent moisture intrusion to improve the water resistance of the adhesive. On further increasing the LR addition to 25 wt%, holes, cracks, and disordered surface were observed again on the facture surface. This is because the soy protein molecules decomposed as the LR addition increased. These decomposed protein molecules with a lot of hydrophilic groups, acting as filler, filled in the LR and reduced the water resistance of the composite adhesive. Also, the holes and cracks helped moisture attachment and swelling, which further decreased the water resistance of the adhesive.

Fig. 9 The fracture surface micrographs of the different cured adhesive samples.

4 Conclusions

Using LR at an appropriate level effectively increased the water resistance of the SM-based adhesive and the wet shear strength of the resultant plywood. This is attributed to the following reasons: (1) a cross-linking network formed by the reaction between the LR and the active groups on soy protein molecules; (2) a smoother fracture surface formed preventing moisture intruding; (3) an interpenetrated network formed by cross-linked protein molecules and self-crosslinking LR molecules; (4) the improved thermal stability of the adhesive; and (5) the resultant appropriate viscosity which benefited adhesive distribution during the hot press process and formed a stronger interlock with the wood surface, thus creating better water resistance. Further increasing the LR addition decreased the water resistance of the adhesive because the resultant very low viscosity by adding LR and protein decomposition caused the adhesive to over-penetrate into the wood surface making effective bonding difficult.

Using 10 wt% of LR effectively improved the water resistance of the adhesive by 3.7% and the wet shear strength of the resultant plywood bonded by this adhesive by 200% to 1.05 MPa, which met interior-use plywood bond strength requirements. The solid content and viscosity of the resultant adhesive were 32.65 wt% and 499 400 mPa s, respectively, which were acceptable for the industrial use of plywood adhesive. This is an effective solution for promoting soybean meal-based adhesive application.

Acknowledgements

The authors are grateful for financial support from the Beijing Natural Science Foundation (2151003) and the Special Fund for Forestry Research in the Public Interest (Project 201404501).

Notes and references

1. T. J. Sellers, For. Prod. J., 2001, 51, 12–22.
2. A. Pizzi and K. L. Mittal, Wood adhesives, CRC Press, 2011.
3. C. R. Frihart, Handb. Wood Chem. Wood Compos., 2005, 215.
4. R. Montgomery, Bioresour. Technol., 2004, 91, 1–29.
5. C. R. Frihart and M. J. Birkeland, Soy Properties and Soy Wood Adhesives, Soy-Based Chemicals and Materials, 2014, ch. 8.
6. A. Bacigalupe, A. K. Poliszuk, P. Eisenberg and M. M. Escobar, Int. J. Adhes. Adhes., 2015, 62, 1–6.
7. G. Y. Qi, N. B. Li, D. H. Wang and X. S. Sun, Ind. Crops Prod., 2013, 46, 165–172.
8. C. H. Park, S. W. Lee, J. W. Park and H. J. Kim, React. Funct. Polym., 2013, 73, 641–646.
9. C. S. Gui, G. Y. Wang, D. Wu, J. Zhu and X. Q. Liu, Int. J. Adhes. Adhes., 2013, 44, 237–242.
10. S. H. Imam, S. H. Gordon, L. Mao and L. Chen, Polym. Degrad. Stab., 2001, 73, 529–533.
11. S. Yang, T.-Q. Yuan, M.-F. Li and R.-C. Sun, Int. J. Biol. Macromol., 2015, 72, 54–62.
12. W. Zhang, Y. Ma, C. Wang, S. Li, M. Zhang and F. Chu, Ind. Crops Prod., 2013, 43, 326–333.
13. Z. Xiao, Y. Li, X. Wu, G. Qi, N. Li, K. Zhang, D. Wang and X. S. Sun, Ind. Crops Prod., 2013, 50, 501–509.
14. J. Luo, J. Luo, Q. Gao and J. Li, Ind. Crops Prod., 2015, 63, 281–286.
15. Q. Gao, S. Q. Shi, J. Z. Li, K. W. Liang and X. M. Zhang, BioResources, 2012, 7, 946–956.
16. W. Qiao, S. Li, G. Guo, S. Han, S. Ren and Y. Ma, J. Ind. Eng. Chem., 2015, 21, 1417–1422.
17. L. Glavas, Starch and Protein based Wood Adhesives, 2011.
18. K. Yang, X. Wang and Y. Wang, J. Ind. Eng. Chem., 2007, 13, 485.
19. N. Hettiarachchy, U. Kalapathy and D. Myers, J. Am. Oil Chem. Soc., 1995, 72, 1461–1464.
20. J. Luo, C. Li, X. Li, J. Luo, Q. Gao and J. Li, RSC Adv., 2015, 5, 62957–62965.
21. L. Chen and M. Subirade, Biomacromolecules, 2009, 10, 3327–3334.
22. W. Weihong and X. Guoliang, J. For. Res., 2007, 2, 021.
23. G. Y. Qi and X. S. Sun, J. Am. Oil Chem. Soc., 2011, 88, 271–281.
24. R. Kumar, V. Choudhary, S. Mishra and I. Varma, J. Therm. Anal. Calorim., 2004, 75, 727–738.
25. J. A. Gerrard, Trends Food Sci. Technol., 2002, 13, 391–399.

---