Soy protein isolate-based films cross-linked by epoxidized soybean oil

Abstract

Epoxidized soybean oil (ESO) is an environmentally friendly cross-linking agent derived from soybean, having multiple epoxy groups in its molecules. It can effectively improve tensile strength and water resistance of soy protein isolate (SPI)-based films. The properties of the SPI-based films were characterized by X-ray diffraction and attenuated total reflectance Fourier transform infrared spectroscopy. The best performance of the SPI-based films was achieved when the ESO addition was 2.5%, for which tensile modulus, tensile strength and 10% offset yield strength were increased to 265.0 MPa, 9.8 MPa and 6.8 MPa, respectively. Compared to untreated SPI-based films, these were increases of 695.6%, 139.8%, and 246.6%, respectively. However, the elongation at break was decreased by 67.6% due to the cross-linking between SPI and ESO. The SPI-based film modified by 5% ESO had the best water-resistance property and reduced the 24 hour water absorption from 209.1% to 45.9%, which was a significant decrease of 78.1%.

---

Introduction

Driven by the environmental concerns caused by using petrochemical-based synthetic polymers,^1,2 interest has arisen in using renewable, degradable, and compostable films and coatings from polysaccharide, protein and lipid biopolymers for packaging, mulching and other industrial applications.^3–5 As a by-product in the edible oil industry, soy protein isolate (SPI) has advantages of being inexpensive, biocompatible, renewable, etc., providing an environmentally conscious alternative to fossil fuel sources.^6–9 However, the drawbacks, such as low strength and poor water resistance,^10,11 limit the applications of SPI-based products.

Many efforts have been made to enhance the performances of SPI-based films, such as improving the processing methods,^12–14 blending or compositing with natural materials,¹⁵ treating with enzymes,¹⁶ cross-linking using chemicals,^6,17–21 etc. Among them, chemical cross-linking of proteins has proved to be one of the most promising ways for enhancing the properties of the films. The most commonly used cross-linking agents are aldehydes, such as formaldehyde, glutaraldehyde and glyoxal,^19,22–24 phenolic compounds,^18,25 and epoxy compounds.^6,26 However, the use of these compounds to fabricate food packaging is not particularly safe considering their cytotoxicity. To address this concern, some biomass-based cross-linking agents, such as genipin²⁰ and dialdehyde starch,²¹ were developed. However, the properties of the cross-linked films, especially the tensile strength, are still expected to be improved.

To improve the tensile strength and water-resistance properties of SPI-based films without introducing very toxic chemicals, epoxidized soybean oil (ESO) was employed as a cross-linking agent in this research. As one of the important vegetable oil-based polymers, ESO is synthesized by the epoxidation of the double bonds of soybean oil.²⁷ Compared with petroleum-based polymers, ESO is a preferred bio-material owing to the advantages of it being bio-renewable, biocompatible, and biodegradable.²⁸ The multiple epoxy groups of the ESO molecules are able to cross-link with SPI molecules owing to their abundant amino groups, which enables the formation of a network structure between SPI and ESO to enhance the tensile strength and water-resistance properties of SPI-based films.

The objective of this study was to fabricate cross-linked SPI-based films using the centrifugal casting method. The effects of ESO on the mechanical properties and water resistance behavior of the SPI-based films were investigated.

Experimental

Materials

SPI with a protein content of over 90% was from Sausage Maker, Inc., USA. Vikoflex® 7170 ESO with an oxirane value of more than 6.80% was supplied by Arkema Chemicals Company, USA. 3-Glycidoxypropyltrimethoxysilane (GPTMS) with a purity of 97% and biotechnology grade glycerol were supplied by Acros Organics and bioWORLD, USA, respectively. Sodium hydroxide (NaOH) solution (10%, w/v) was prepared using NaOH beads (≥97%, Acros Organics) and deionized (DI) water that was from a Millipore Milli-Q Integral Water Purification System.

Preparation of the SPI-based films

Using the method of centrifugal casting, a spin caster (AMT-SC5052, Affine Materials Technology, USA) with a motor (BQB 56C34D2096F P, Marathon, USA) of 3450 RPM was employed to fabricate the films. The centrifugal casting method is able to accelerate the evaporation of water from the mixture. The experiments were performed in a fume hood with a relative humidity of about 50% and a temperature of 20 °C. A three-step process was carried out:

(1) The SPI films were made by mixing 4 g SPI with 40 mL DI water, 2 g glycerol and 0.5 g 10% NaOH solution. Different contents of ESO and/or GPTMS were added to produce different types of films as shown in Table 1. All the mixtures were magnetically stirred for 1 h before the next process.

Table 1 Feedstocks for preparing the SPI-based films

Film	SPI^a (g)	DI water^b (mL)	NaOH^c (g)	ESO^e		GPTMS^f
				(g)	(%)^d	(g)	(%)^d
a SPI = soy protein isolate.b DI water = deionized water.c 10% NaOH aqueous solution.d The percentages were calculated based on the amount of SPI (4 g).e ESO = epoxidized soybean oil.f GPTMS = 3-glycidoxypropyltrimethoxysilane.
SPI	4	40	0.5	—	—	—	—
SPI/GPTMS	4	40	0.5	—	—	0.4	10
SPI/GPTMS/ESO-1%	4	40	0.5	0.04	1	0.4	10
SPI/GPTMS/ESO-2.5%	4	40	0.5	0.1	2.5	0.4	10
SPI/GPTMS/ESO-5%	4	40	0.5	0.2	5	0.4	10
SPI/GPTMS/ESO-10%	4	40	0.5	0.4	10	0.4	10
SPI/GPTMS/ESO-15%	4	40	0.5	0.6	15	0.4	10
SPI/ESO-2.5%	4	40	0.5	0.1	2.5	—	—

---

(2) The mixtures were pulled into the spin caster and then centrifuged for 4 h. The films were formed on the surface of a cylindrical container of 5 cm in height and 12.7 cm in diameter.

(3) The films were thermally treated at 105 ± 3 °C for 4 h and stored in a conditioning room with a relative humidity of 50 ± 2% and a temperature of 20 ± 3 °C.

ATR FT-IR analysis

Attenuated total reflectance (ATR) Fourier transform infrared (FT-IR) spectroscopy was used to examine the chemical structures of the films. A Nicolet 6700 FT-IR spectrometer (Thermo Scientific, USA) with an ATR accessory was employed. The wavelength range was between 650 and 4000 cm⁻¹.

XRD analysis

X-ray diffraction (XRD) analysis was carried out with an XRD-6000 diffractometer (Shimadzu, Japan, Cu Kα radiation, λ = 1.54060 Å), operating at 40 kV and 40 mA. The samples were scanned in the continuous scanning mode with 2θ from 10° to 60° (2θ) at a rate of 2° min⁻¹, where θ is the incident angle of the X-ray beam on the sample. The relative crystallinity index (RCI) was calculated directly using the XRD instrument with the equation:

RCI (%) = A_c × (A_c + A_a)⁻¹ × 100 (1)

where A_c is the area of the crystalline region and A_a is the area of the amorphous region. Three replicates were carried out for each composition.

Mechanical property tests

The tensile tests of the specimens were carried out at room temperature in accordance with the procedures described in ASTM D882 standard using a Shimadzu AGS-X tester with a 76.2 mm (3 inch) gauge length. Twelve replicates were used for each film. The specimen dimensions were 18 × 120 mm and the cross section of the narrow section was 76.2 × 12.7 mm (length × width). A crosshead speed of 50 mm min⁻¹ was used.

SEM observation

A Quanta 200 environmental scanning electron microscope (SEM) with an accelerating voltage of 15 kV and a magnification of 2000× was used to observe the tensile fracture failure mode of the specimens. Prior to the observations, the specimens were sputter-coated with gold for 5 minutes to avoid charging under the electron beam.

Water absorption tests

Six specimens of each film with dimensions of 20 × 20 mm were used for the moisture content measurements and water absorption tests. Twenty-four-hour water submersion tests were carried out in accordance with the ASTM D1037 standard for the determination of water absorption. In addition, the moisture contents after equilibration in a conditioning chamber at a relative humidity of 50 ± 2% and a temperature of 20 ± 3 °C for 10 days and water submersion for 24 h were determined in accordance with the ASTM D4442 standard.

Results and discussion

Characterization of the SPI-based films

SPI contains more than 90% proteins, including 7.7% arginine (Arg), 6.8% lysine (Lys), and 2.5% cysteine (Cys) that have abundant –NH₂ groups.²⁹ The total percentage of those types of amino acids can be 17.0% in the protein. ESO contains multiple epoxy groups per molecule.^12,27,30 The –NH₂ groups and the epoxy groups can react with each other through ring-opening polymerization.^6,31 The cross-linking reactions are shown in Fig. 1. The multiple –NH₂ groups of the SPI and multiple epoxy groups of the ESO were able to cross-link the two feedstocks, resulting in a cross-linked film.

Fig. 1 Cross-linking reaction mechanism of SPI and ESO.

Images of the films are displayed in Fig. 2. SPI, SPI/GPTMS, SPI/GPTMS/ESO-1%, and SPI/GPTMS/ESO-2.5% films showed very good transparency. With an increase in ESO amount, the films (SPI/GPTMS/ESO-5%, SPI/GPTMS/ESO-10%, and SPI/GPTMS/ESO-15% films) became less transparent. Compared to the good transparency and uniform distribution of the SPI/GPTMS/ESO-2.5% film, the SPI/ESO-2.5% film without GPTMS was close to opaque and very heterogeneous, suggesting a poor distribution of ESO in the SPI matrix. To improve the compatibility between SPI and ESO, a silane coupling agent, GPTMS, was employed to disperse the hydrophobic ESO into hydrophilic SPI.^32–34 Comparing the images of SPI/GPTMS/ESO-2.5% and SPI/ESO-2.5% films (Fig. 2), it is easy to see the effect of GPTMS on the compatibility of the films, reflecting in the relatively clear and transparent SPI/GPTMS/ESO-2.5% film.

Fig. 2 Images of the SPI-based films.

The ATR FT-IR spectra of SPI powder and SPI-based films are presented in Fig. 3. The broad absorption band observed around 3275 cm⁻¹ is attributable to free and bound O–H and N–H groups. An increase in intensity of the band at 3275 cm⁻¹ was observed after the SPI was fabricated into the films, due to the abundant –OH groups being introduced into the system. The characteristic C–H stretching bands of CH₂ and CH₃ groups of saturated structures were observed at 2925 and 2855 cm⁻¹, and C–H bending band was seen at 1455 cm⁻¹. The peak at 1743 cm⁻¹ corresponded to the stretching of C=O groups in COOH or COOR. The spectra of the SPI films showed relevant peaks at 1627, 1537 and 1233 cm⁻¹ (Fig. 3), which were characteristic of amide I (C=O stretching), amide II (N–H bending), and amide III (C–N and N–H stretching), respectively. This result was consistent with the findings in the literature.³⁵ The peak at 1039 cm⁻¹ belongs to C–O stretching. Compared to the SPI powder, the FT-IR peaks of the SPI film at 3275 and 1039 cm⁻¹ were increased, owing to the glycerol in the SPI film. The intensity of the peak at 1743 cm⁻¹ (C=O groups in COOH or COOR) became stronger after adding the GPTMS ((a) to (b) in Fig. 3), due to the amphiphilic property of GPTMS. The same phenomenon was found when sodium dodecylsulfate was employed.³⁶ The intensities at 1743 cm⁻¹ were decreased after adding 1% and 2.5% ESO ((b) to (c) and (d) in Fig. 3), while they increased for the films containing 5%, 10%, and 15% ESO. There could be two reasons: (1) the reactions between ESO and SPI reduced the density of C=O groups (in COOH or COOR) on the surface of the film; (2) the increased amounts of ESO introduced abundant C=O groups (in COOR) existing on the surface. Comparing the FT-IR spectra of the SPI/GPTMS/ESO-2.5% film ((d) in Fig. 3) with that of the SPI/ESO-2.5% film ((h) in Fig. 3), and that without GPTMS (the SPI/ESO-2.5% film), ESO was located on the surface of the film with poor distribution and presented a relatively strong intensity at 1743 cm⁻¹, indicating that GPTMS can help disperse the hydrophobic ESO into hydrophilic SPI in accordance with the peak intensity at 1743 cm⁻¹ of the SPI/GPTMS/ESO-2.5% film.

Fig. 3 ATR FT-IR spectra of SPI powder, and films of SPI (a), SPI/GPTMS (b), SPI/GPTMS/ESO-1% (c), SPI/GPTMS/ESO-2.5% (d), SPI/GPTMS/ESO-5% (e), SPI/GPTMS/ESO-10% (f), SPI/GPTMS/ESO-15% (g) and SPI/ESO-2.5% (h).

XRD patterns of SPI powder and the SPI-based films are presented in Fig. 4. Peaks at 2θ ≈ 8.8° and 19.8° represent the α-helix and β-sheet structures of SPI secondary structure, respectively.^6,37 The β-sheet XRD peaks of SPI-based films were slightly greater than that of the SPI powder (19.0°), because of the effect of the small molecule (glycerol) in the SPI-based films, which was able to reduce the lattice constant. The RCIs were calculated according to eqn (1), and the results are shown in Table 2. The lowest RCI was found for the SPI/GPTMS/ESO-2.5% film (RCI = 22.9%), which was reduced by 22.4% compared with the SPI film (RCI = 29.5%). The results of ANOVA tests showed that the decrease was significant at the probability level of α = 0.01 (P-value = 0.0022). The crystallinity of the SPI/GPTMS/ESO-2.5% film was decreased significantly, since the molecular chains could not move smoothly after the formation of the cross-linking networks.⁶

Fig. 4 XRD patterns of SPI powder, and films of SPI (a), SPI/GPTMS (b), SPI/GPTMS/ESO-1% (c), SPI/GPTMS/ESO-2.5% (d), SPI/GPTMS/ESO-5% (e), SPI/GPTMS/ESO-10% (f), SPI/GPTMS/ESO-15% (g) and SPI/ESO-2.5% (h).

Table 2 Crystallinity of the SPI powder and SPI-based films

Sample	RCI^a (%)
a RCI = relative crystallinity index.b Mean (standard deviation).
SPI powder	30.1 (0.7)^b
SPI	29.5 (0.9)
SPI/GPTMS	34.0 (0.4)
SPI/GPTMS/ESO-1%	23.5 (0.9)
SPI/GPTMS/ESO-2.5%	22.9 (1.0)
SPI/GPTMS/ESO-5%	28.0 (1.0)
SPI/GPTMS/ESO-10%	36.4 (0.4)
SPI/GPTMS/ESO-15%	34.1 (1.0)
SPI/ESO-2.5%	33.7 (0.4)

---

Mechanical properties

The results of tensile tests (tensile modulus (TE), tensile strength (TS), 10% offset yield strength (YS), and elongation at break) are summarized in Table 3 and the average engineering strain–stress curves of the films are shown in Fig. 5. Generally, it was found that the TE, TS and YS of SPI-based films tended to increase with the added ESO amount to reach a maximum at 2.5% ESO, and then decreased with a further increase in ESO. Obviously, the SPI/GPTMS/ESO-2.5% film had the greatest TE (265.0 MPa), TS (9.8 MPa) and YS (6.8 MPa), which represented increases of 695.6%, 139.8%, and 246.6%, respectively, compared with those mechanical properties of the SPI film (TE = 33.3 MPa, TS = 4.1 MPa, and YS = 2.0 MPa). TE, TS and YS were analyzed by ANOVA tests and the results showed the increases were significant at the probability level of α = 0.001, for which the P-values were 2.4 × 10⁻⁵, 1.6 × 10⁻⁴, and 3.4 × 10⁻⁶, respectively. However, the elongation at break was reduced by 67.6%, owing to the formation of the cross-linking network between SPI and ESO (Fig. 1).⁶ The results indicated that ESO could significantly enhance the TE, TS, and YS of SPI-based film, but reduced the elongation at break. The TE, TS, and YS of the SPI/GPTMS/ESO-2.5% film were increased by 231.9%, 163.6%, and 142.2%, respectively, compared with those of the SPI/ESO-2.5% film. This suggested that the compatibility between the SPI and ESO was improved by the addition of GPTMS, because GPTMS helps to disperse the hydrophobic ESO into the hydrophilic SPI, which was consistent with the images in Fig. 2.

Table 3 Mechanical properties (tensile modulus, tensile strength, 10% offset yield strength, and elongation at break) and thicknesses of the SPI-based films

Film	Thickness (mm)	Tensile modulus (MPa)	Tensile strength (MPa)	10% offset yield strength (MPa)	Elongation at break (%)
a Mean (standard deviation).
SPI	0.261 (0.012)^a	33.3 (2.6)	4.1 (0.2)	2.0 (0.1)	213.5 (12.4)
SPI/GPTMS	0.241 (0.009)	78.3 (6.9)	5.7 (0.6)	3.5 (0.1)	65.7 (13.9)
SPI/GPTMS/ESO-1%	0.239 (0.019)	96.6 (6.0)	7.0 (0.3)	4.0 (0.2)	75.1 (3.7)
SPI/GPTMS/ESO-2.5%	0.261 (0.013)	265.0 (14.5)	9.8 (0.6)	6.8 (0.2)	69.1 (10.9)
SPI/GPTMS/ESO-5%	0.259 (0.022)	132.1 (13.7)	8.9 (0.2)	4.8 (0.1)	83.1 (3.6)
SPI/GPTMS/ESO-10%	0.264 (0.004)	177.8 (20.0)	7.6 (0.3)	5.2 (0.1)	64.8 (7.2)
SPI/GPTMS/ESO-15%	0.328 (0.008)	62.0 (4.8)	4.8 (0.3)	3.1 (0.1)	61.0 (7.0)
SPI/ESO-2.5%	0.342 (0.015)	79.8 (6.5)	3.7 (0.2)	2.8 (0.1)	97.7 (19.4)

---

Fig. 5 Stress–strain curves of SPI-based films.

The tensile strengths of the SPI-based films with different cross-linking agents were compared and are presented in Table 4. The SPI-based films cross-linked by ESO had the highest TS of 9.8 MPa with an increase of 139.8% among the biomass-based agents including genipin (4.6 MPa with a 42.9% increase)²⁰ and dialdehyde starch (7.84 MPa with a 23.7% increase).²¹ Compared to the SPI-based films cross-linked by other chemicals, including 1,2,3-propanetrioldiglycidyl ether (TS = 6.21 MPa with a 197.1% increase),⁶ ferulic acid (TS = 2.602 MPa with a 62.8% increase),¹⁷ and formaldehyde (TS = 1.69 MPa with a 164.1% increase),¹⁹ the TS of the SPI-based film fabricated in this work was still larger (Table 4), though the increase might not be the largest. Based on a detailed literature review, the best cross-linking agent was resorcinol,¹⁸ and the TS (24.7 MPa with a 404.1% increase) of a resorcinol-treated SPI-based film was higher than that in this work. However, resorcinol is a cytotoxic chemical which may cause some environmental concerns. In general, by taking the film performance and environmental issues into account, the SPI-based film cross-linked by ESO is a unique technology because it improves the tensile strength without compromising the environmental sustainability.

Table 4 Comparison of tensile strengths of SPI-based films cross-linked by various agents

Cross-linking agent	TS of films (MPa)^c		Increase^d (%)	Reference
	Control	Treated
a ESO = epoxidized soybean oil.b PTGE = 1,2,3-propanetrioldiglycidyl ether.c TS = tensile strength.d Increase was calculated from the TS of control and treated samples from relevant literature.
ESO^a	4.1	9.8	139.8	This work
PTGE^b	2.09	6.21	197.1	Xu^6
Ferulic acid	1.598	2.602	62.8	Ou^17
Resorcinol	4.9	24.7	404.1	Reddy^18
Formaldehyde	0.64	1.69	164.1	Chen^19
Genipin	3.22	4.6	42.9	Gonzalez^20
Dialdehyde starch	6.34	7.84	23.7	Rhim^21

---

Micromorphology

The cross-sectional morphologies of tensile fracture surfaces are presented in Fig. 6. Defects and inhomogeneous characteristics were obvious in the SPI film. After GPTMS was introduced, the surface was not improved. Relatively homogeneous surfaces were observed for the SPI/GPTMS/ESO films with 1%, 2.5%, and 5% ESO. When the addition of ESO was further increased, the surfaces of SPI/GPTMS/ESO films with 10% and 15% ESO became inhomogeneous gradually. Furthermore, compared to the SPI/GPTMS/ESO-2.5% film, the film without GPTMS (the SPI/ESO-2.5% film) contained abundant defects, owing to the poor ESO distribution in the SPI matrix, which was consistent with the results of mechanical properties.

Fig. 6 SEM images of cross-sections of the SPI-based films.

Water resistance

The results of moisture content determination, 24 h water absorption, and total soluble matter of the SPI-based films are summarized in Table 5 for comparison. Since the main soluble matter in the films was glycerol and the content remained unchanged relative to SPI (50%), the total soluble matter of the films ranged from 24.3 to 30.0%, without significant difference among the films. The moisture content of the SPI film (35.8%) was decreased by 36.2% after adding GPTMS (22.8%, SPI/GPTMS film). The moisture content was further reduced by 55.5% after adding 2.5% ESO, compared with the SPI film. The ANOVA test showed a significant reduction at the level of α = 0.001 (P-value = 5.1 × 10⁻⁶). The results of 24 h water absorption tests (Table 5) showed that the SPI/GPTMS/ESO-5% film had the best water-resistance property (45.9%), the water absorption being significantly reduced by 78.1% (ANOVA test, α = 0.001, P-value = 5.6 × 10⁻⁷) compared with the SPI film (209.1%). The 24 h water absorption of the SPI/GPTMS/ESO-2.5% film having the best TE, TS, and YS properties was 66.8%, which was significantly decreased by 68.1% (ANOVA test, α = 0.001, P-value = 4.0 × 10⁻⁷) compared with the SPI film. The 24 h water absorption of the SPI/ESO-2.5% film, the film without GPTMS, was 207.2%, which was similar to that of the SPI film (209.1%). The reason could be the poor distribution of ESO in the SPI matrix without the GPTMS effect, resulting in abundant defects (pores) in the film (Fig. 6). The results showed that the water resistance of the SPI film was dramatically improved through the formation of a cross-linking network in the modified films.

Table 5 Moisture content after conditioning, water absorption and total soluble matter after 24 h submersion of the SPI-based films

Film	Moisture content (%)	Water absorption (%)	Total soluble matter (%)
a Mean (standard deviation).
SPI	35.8 (1.2)^a	209.1 (3.1)	30.0 (0.4)
SPI/GPTMS	22.8 (0.6)	85.2 (10.0)	27.9 (0.7)
SPI/GPTMS/ESO-1%	19.4 (1.1)	78.1 (0.5)	25.2 (0.5)
SPI/GPTMS/ESO-2.5%	15.9 (1.0)	66.8 (1.7)	26.8 (0.6)
SPI/GPTMS/ESO-5%	19.1 (2.4)	45.9 (2.1)	24.7 (1.2)
SPI/GPTMS/ESO-10%	20.4 (0.5)	93.0 (2.2)	26.5 (0.3)
SPI/GPTMS/ESO-15%	19.5 (2.1)	80.5 (0.4)	24.3 (0.6)
SPI/ESO-2.5%	18.7 (0.7)	207.2 (5.0)	28.4 (0.6)

---

Conclusions

The tensile modulus, tensile strength, 10% offset yield strength and water resistance of the ESO-modified SPI-based films were significantly improved with the presence of GPTMS. When 2.5% ESO was added, an optimal overall performance of films was achieved. Compared to the unmodified SPI film, TE, TS and YS of the SPI/GPTMS/ESO-2.5% film were increased by 695.6%, 139.8%, and 246.6%, respectively, and the elongation at break was decreased by 67.6%, resulting from the cross-linking reactions between SPI and ESO. Furthermore, the water resistance of the modified films was improved. The results of 24 h water absorption tests showed that the SPI/GPTMS/ESO-5% film had the best water-resistance property, with a 78.1% reduction in water absorption compared with the unmodified SPI film.

Acknowledgements

This research was supported by the Special Fund for Forestry Research in the Public Interest (Project 201504502), the Beijing Natural Science Foundation (Project 2151003), the National Natural Science Foundation of China (Project 31000268/C160302), and the University of North Texas Research Initiation Fund.

Notes and references

1. L. Averous, J. Macromol. Sci., Part A: Pure Appl.Chem., 2004, C44, 231–274.
2. M. Avella, M. E. Errico, R. Rimedio and P. Sadocco, J. Appl. Polym. Sci., 2002, 83, 1432–1442.
3. J. M. Krochta and C. DeMulderJohnston, Food Technol., 1997, 51, 61–74.
4. M. Cruz-Romero and J. Kerry, CAB Rev., 2008, 3, 25.
5. J. P. Maran, V. Sivakumar, R. Sridhar and V. P. Immanuel, Ind. Crops Prod., 2013, 42, 159–168.
6. F. Xu, Y. Dong, W. Zhang, S. Zhang, L. Li and J. Li, Ind. Crops Prod., 2015, 67, 373–380.
7. P. Lodha and A. N. Netravali, Ind. Crops Prod., 2005, 21, 49–64.
8. J. W. Rhim, A. Gennadios, C. L. Weller and M. A. Hanna, Ind. Crops Prod., 2002, 15, 199–205.
9. X. Ren and M. Soucek, in Soy-Based Chemicals and Materials, ed. R. P. Brentin, American Chemical Society, 2014, pp. 207–254.
10. P. Tummala, W. Liu, L. T. Drzal, A. K. Mohanty and M. Misra, Ind. Eng. Chem. Res., 2006, 45, 7491–7496.
11. W. J. Liu, M. Misra, P. Askeland, L. T. Drzal and A. K. Mohanty, Polymer, 2005, 46, 2710–2721.
12. X. Song, C. Zhou, F. Fu, Z. Chen and Q. Wu, Ind. Crops Prod., 2013, 43, 538–544.
13. J. A. Foulk and J. M. Bunn, Ind. Crops Prod., 2001, 14, 11–22.
14. V. M. Ghorpade and M. A. Hanna, Trans. ASAE, 1996, 39, 611–615.
15. H. Tian, G. Xu, B. Yang and G. Guo, J. Food Eng., 2011, 107, 21–26.
16. G. A. Denavi, M. Perez-Mateos, M. C. Anon, P. Montero, A. N. Mauri and M. Carmen Gomez-Guillen, Food Hydrocolloids, 2009, 23, 2094–2101.
17. S. Y. Ou, Y. Wang, S. Z. Tang, C. H. Huang and M. G. Jackson, J. Food Eng., 2005, 70, 205–210.
18. D. J. P. Reddy, A. V. Rajulu, V. Arumugam, M. D. Naresh and M. Muthukrishnan, J. Plast. Film Sheeting, 2009, 25, 221–233.
19. L. Chen, G. Remondetto, M. Rouabhia and M. Subirade, Biomaterials, 2008, 29, 3750–3756.
20. A. Gonzalez, M. C. Strumia and C. I. A. Igarzabal, J. Food Eng., 2011, 106, 331–338.
21. J. W. Rhim, A. Gennadios, C. L. Weller, C. Cezeirat and M. A. Hanna, Ind. Crops Prod., 1998, 8, 195–203.
22. A. Bigi, G. Cojazzi, S. Panzavolta, K. Rubini and N. Roveri, Biomaterials, 2001, 22, 763–768.
23. C. M. Vaz, P. F. N. M. van Doeveren, G. Yilmaz, L. A. de Graaf, R. L. Reis and A. M. Cunha, J. Appl. Polym. Sci., 2005, 97, 604–610.
24. C. Marquie, J. Agric. Food Chem., 2001, 49, 4676–4681.
25. G. Strauss and S. A. Gibson, Food Hydrocolloids, 2004, 18, 81–89.
26. R. D. Patil, J. E. Mark, A. Apostolov, E. Vassileva and S. Fakirov, Eur. Polym. J., 2000, 36, 1055–1061.
27. S. J. Park, F. L. Jin and J. R. Lee, Mater. Sci. Eng., A, 2004, 374, 109–114.
28. A. M. Díez-Pascual and A. L. Díez-Vicente, ACS Appl. Mater. Interfaces, 2014, 6, 17277–17288.
29. I. Mateos-Aparicio, A. R. Cuenca, M. J. Villanueva-Suarez and M. A. Zapata-Revilla, Nutr. Hosp., 2008, 23, 305–312.
30. Z. Qin, Q. Gao, S. Zhang and J. Li, BioResources, 2013, 8, 5369–5379.
31. H. Lei, G. Du, Z. Wu, X. Xi and Z. Dong, Int. J. Adhes. Adhes., 2014, 50, 199–203.
32. K. Luo, S. Zhou, L. Wu and G. Gu, Langmuir, 2008, 24, 11497–11505.
33. C. Li, H. Li, S. Zhang and J. Li, BioResources, 2014, 9, 5448–5460.
34. Y. Xie, C. A. S. Hill, Z. Xiao, H. Militz and C. Mai, Composites, Part A, 2010, 41, 806–819.
35. E. M. Ciannamea, P. M. Stefani and R. A. Ruseckaite, Food Hydrocolloids, 2014, 38, 193–204.
36. V. Schmidt, C. Giacomelli and V. Soldi, Polym. Degrad. Stab., 2005, 87, 25–31.
37. J. Chen, X. Chen, Q. Zhu, F. Chen, X. Zhao and Q. Ao, J. Sci. Food Agric., 2013, 93, 1687–1691.

---