Characterization of pea protein-based bioplastics processed by injection moulding

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

Contents lists available at ScienceDirect

Food and Bioproducts Processing journal homepage: www.elsevier.com/locate/fbp

Characterization of pea protein-based bioplastics processed by injection moulding Victor Perez, Manuel Felix ∗ , Alberto Romero, Antonio Guerrero Departamento de Ingeniería Química, Universidad de Sevilla, Facultad de Química, 41012 Sevilla, Spain

a r t i c l e

i n f o

a b s t r a c t

Article history:

This study assesses the behaviour of pea protein isolate (PPI) as a potential candidate for

Received 15 June 2015

the development of biobased plastic materials processed by injection moulding. Around

Received in revised form 18

30–40% of glycerol as plasticizer was required to obtain good processability of PPI/GL blends

September 2015

to produce bioplastics. A mixing rheometer that allows recording of torque and temperature

Accepted 2 December 2015

during mixing and a small-scale-plunger-type injection moulding machine were used to

Available online 14 December 2015

obtain PPI/GL blends and PPI-based bioplastics, respectively. Rheological and differential scanning calorimetry measurements were made to guide the selection of suitable conditions

Keywords:

for injection and moulding. For injection, we selected a temperature relatively close to the

Bioplastic

maximum of the loss tangent, but moderate enough to avoid crosslinking effects (50 ◦ C), and

Dynamic mechanical thermal

for moulding, a high temperature (130 ◦ C) to favour crosslinking in the mould. An increase

analysis

in the PPI/GL ratio leads to an enhancement of elastic bending and tensile properties of

Mixing

bioplastic specimens, as well as an increase in their ability to absorb mechanical energy

Pea protein

before rupturing. On the other hand, the PPI/GL specimens become less transparent. In

Tensile strength test

addition, water uptake of these bioplastics has been found to be very high and fast.

Water absorption

© 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1.

Introduction

Over recent years, a huge amount of non-degradable plastic waste has been generated, leaving behind an undesirable human footprint. This environmental problem is worsened by the fact that there is a continuous growth in the demand for plastic products. As a consequence, naturally derived biodegradable polymers (bioplastics) are attracting growing interest as materials that could reduce the environmental impact, this waste, while themselves being made from renewable sources (Shand et al., 2009). Specifically, since 2011, the market for eco-friendly bioplastics has grown exponentially (Byun and Teck Kim, 2014). Proteins (from various sources such as whey, egg, blood meal, soybean, gluten, pea, etc.) and polysaccharides have been proposed as attractive raw materials for the production of bioplastics for a range of applications (di Gioia and Guilbert, 1999; Pommet et al., 2003; Ribotta et al., 2012; Chao et al., 2013). In order to reduce

∗

intermolecular forces between polymer chains, protein-based bioplastic processing generally requires a mixing stage with a plasticizer (Gennadios, 2002; Feeney and Whitaker, 1988; Mohammed et al., 2000), which leads to an increase in the mobility of protein chains and a reduction in the glass transition (Matveev et al., 2000; Pouplin et al., 1999). Regarding type of protein, plant proteins have become fairly attractive for a wide range of potential applications (Plastics Europe, 2008). In particular, legume seeds are cheap sources of protein with a relatively high nutritional value, which make them a very good raw material for the production of protein-based products (De Graaf and Kolster, 1998; Siracusa et al., 2008). However, the worldwide market for bio-based plastic materials is dominated by soy protein (commercially available as soy flour, soy concentrate and soy isolate). This is attributable to its low price, high quality and fairly versatile properties, factors that make it difficult to compete with (Pearson, 1983). However, utilization of pea protein can

Corresponding author. Tel.: +34 954557179; fax: +34 954556447. E-mail address: [email protected] (M. Felix). http://dx.doi.org/10.1016/j.fbp.2015.12.004 0960-3085/© 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

also have economic benefits since its price ($2.5–2.8 Kg−1 ) is lower than that of other protein isolates like whey protein ($13.5–28 Kg−1 ) and even soy protein ($3–3.8 Kg−1 ) (Kowalczyk and Baraniak, 2011). In addition, pea protein offers other advantages such as its lack of genetic modifications in commercial species as well as its relatively low allergenicity and associated rates of feeding intolerance (Directive, 2001/18/EC). In any case, a large part of the literature is focused on soy protein-based films (Kokoszka et al., 2010; Guerrero et al., 2011a; Guerrero et al., 2011b), while data on the use of pea protein is scarce, in spite of its great potential due to its excellent properties (Choi and Han, 2002; Shi and Dumont, 2014). With regard to processing, protein systems have typically been processed by casting to obtain protein-based films, using an appropriate solvent to transform proteins into a liquid phase. However, only a limited number of plant proteins (gliadin and zein) are soluble in common solvents and using solvents or alkaline solutions increases the cost and makes the process environmentally unfriendly (Reddy and Yang, 2013). In response to this, efforts have been made over recent years to apply classical thermoplastic polymer processing techniques, including thermoforming, extrusion, compression and moulding, to obtain protein-based bioplastic materials (Orliac et al., 2003; Pommet et al., 2003; Liu et al., 2005; Tummala et al., 2006; Jerez et al., 2007; Hernandez-Izquierdo and Krochta, 2008; González-Gutiérrez et al., 2011; Zárate-Ramírez et al., 2011, 2014a). Among the range of thermomechanical techniques used to process plastics, injection moulding is one of the most important for use with thermoplastic or thermosetting polymers to obtain a wide variety of products of different shapes, sizes and functionalities. It seems reasonable to assume that injection moulding could also be used for polymer systems such as proteins which may be of a mixed character (between thermoplastic and thermosetting). However, the use of injection moulding for processing protein-based materials has only recently been considered, and for successful processing of protein/plasticizer blends by this technique, a suitable selection of processing parameters has proven to be essential (Fernández-Espada et al., 2013; Felix et al., 2014, 2015; Martín-Alfonso et al., 2014). Among these parameters, the preinjection temperature of the cylinder, the injection pressure and the moulding temperature are the most important in this process (Beltrán-Rico and Marcilla-Gomis, 2012). In the case of pea-based bioplastics, injection moulding is still an unexplored approach. The overall objective of this work is to explore the characteristics of pea protein for the potential development of biobased plastic materials processed by injection moulding. A small-scale plunger-type injection-moulding machine was used in this study to obtain pea protein-based specimens from pea protein/glycerol blends, previously mixed in a mixing rheometer that allows recording of both torque and temperature during the mixing process. Rheological and differential scanning calorimetry (DSC) measurements of these blends were also carried out to obtain information to guide the selection of processing parameters suitable for injection moulding.

101

determined in triplicate as % N × 6.25 using a LECO CHNS-932 nitrogen micro analyser (Leco Corporation, St. Joseph, MI, USA) (Etheridge et al., 1998), was 89.5 wt%. According to Pearson classification (1983), it may be considered as a protein isolate rather than a protein concentrate because its protein content is ca. 90%. Ash content was determined by heating a small amount of pea protein isolate (PPI) at 150 ◦ C by putting a muffle in an oven and weighing the content after 3 h. The other components of the pea protein isolate include 3.5 wt% ash, 1.4 wt% lipids and 5.1 wt% moisture. Glycerol (GL) was purchased from Panreac Química, S.A. (Spain).

2.2.

Sample preparation

Bioplastics were manufactured by a thermo-mechanical procedure including two stages. Firstly, blends were mixed at four different PPI/GL ratios (80/20; 70/30; 60/40 and 50/50) using a two-blade counter-rotating batch mixer, Haake Polylab QC (ThermoHaake, Karlsruhe, Germany), at 25 ◦ C and 50 rpm for 60 min, monitoring the torque and temperature during mixing to obtain a dough-like blend. Secondly, two blends, selected after analysis of the mixing results, were subsequently processed by injection moulding, using a MiniJet Piston Injection Moulding System (ThermoHaake) to obtain bioplastic specimens. The processing conditions are given below, being selected after characterization of the pea protein and its blends. Two moulds were used to prepare two different specimens: (1) a 60 × 10 × 1 mm3 rectangular-shaped specimen for dynamic mechanical analysis (DMA) experiments, water absorption capacity and transparency measurements and (2) a dumb-bell-type specimen by ISO 527-1:2012 for tensile properties of plastics.

2.3.

Characterization of pea protein isolate

2.3.1. Protein solubility Protein solubility at different pH values was determined. Aqueous dispersions (ca. 1.00 g protein/40 mL) were prepared and pH of different aliquots was adjusted to alkaline pH values with 6 N NaOH, and to acid pH with 2 N HCl. Samples were homogenized and subsequently centrifuged for 20 min at 10,000×g and 10 ◦ C. The supernatant were collected for protein content determination by means of the Markwell method (Markwell et al., 1978). Solubility was expressed as a percentage (g soluble protein/100 g isolate in sample).

2.3.2.

Z potential measurements

2.

Material and methods

Isoelectric point was measured using a “Zetasizer 2000” (Malvern Instruments, U.K.). Therefore, different flour samples were prepared at 1 wt% prepared at pH value with buffers. Prior to analysis, the samples were stirred at 20 ◦ C and, then, samples were centrifuged at 10,000×g for 10 min in a RC5C Sorvall centrifuge (Sorvall Instruments, Wilmington, DE, USA). After that, the samples were measured in triplicate at 20 ◦ C. The zeta potential was calculated from electrophoretic mobility using the Henry equation and the Smoluchowski approximation. The isoelectric point corresponds to the point where the potential value is zero, at which all charges of particles are neutralized (Tan et al., 2008).

2.1.

Materials

2.3.3.

Pea protein concentrate was delivered by Roquette (Lestrem, France). The protein content of pea protein concentrate,

SDS-page electrophoresis

Electrophoresis tests were performed using polyacrylamide gels (10%) in presence of sodium dodecyl sulphate (SDS-PAGE) according to Laemmli method (1970). Molecular weights of

102

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

extracted protein fractions were determined by using SDSPAGE gels, considering the relationship between the logarithm of protein molecular weight and electrophoretic mobility, using as analytical standard “Protein Plus Protein Standards” (Bio-Rad, Richmond, CA, USA). Dilutions at three different buffers have been prepared in order to maintain a constant pH. Phosphoric acid has been used to maintain the solution at pH 3. In the same way, monosodium phosphate and ethanolamine were used in order to stabilize pH 7 and 10, respectively. All buffer solutions were carried out at a concentration of 10 mM, and the ionic strength of each solution was adjusted at 154 mM by adding NaCl. Finally, 2-Mercaptoethanol was added to minimize chemical interactions among different protein fractions.

2.3.4.

Free and total sulphydryls

Free and total sulphydryl groups of protein samples were determined using the method developed by Beveridge et al. (1974) and Thannauser et al. (1984), respectively. Samples (1 mg/mL) were suspended in 0.086 mol/L Tris-HCl–0.09 mol/L glycine–4 mmol/L EDTA–8 mol/L urea–pH 8 buffer. Dispersions were stirred at 25 ◦ C during 10 min at 500 rpm in a thermomixer and then centrifuged at 15,000×g (10 min, 10 ◦ C). Supernatant was incubated with Ellman’s reagent (4 mg DTNB/mL methanol) and 1 mL NTSB was used in the case of the total sulphydryls. Absorbance at 412 nm was measured in a Genesys-20 spectrophotometer (Thermo Scientific, Waltham, MA, USA). The molar extinction coefficient of 3-thio6-nitrobenzoate (TNB; 13,600 L mol−1 cm−1 ) was used. Protein concentration of extracts was determined by the Bradford method (Bradford, 1976).

2.4.

Characterization of blends

2.4.1.

2.4.1. Differential scanning calorimetry (DSC)

DSC experiments were performed with a Q20 (TA Instruments, New Castle, DE, USA), using 3–8 mg samples, in hermetic aluminium pans. A heating rate of 10 ◦ C/min was selected. The sample was purged with a nitrogen flow of 50 mL/min.

2.4.2. Rheological measurements Dough-like blends were characterized by small amplitude oscillatory shear (SAOS) measurements, using a controlledstrain rheometer (ARES, TA Instruments, USA). A plate–plate geometry (dia: 40 mm) with a rough surface has been used, selecting a gap between plates of 2 mm. Low viscosity Dow Corning 200 fluid was used as sealant to avoid sample drying. Strain sweep SAOS tests were also performed in order to establish the linear viscoelasticity range. Temperature ramp tests were carried out at 5 ◦ C/min from 20 to 100 ◦ C. In these measurements, complex viscosity (* ) was monitored at a constant frequency of 6.28 rad/s. All the systems studied had the same thermo-rheological history before performing any rheological test.

samples were coated with Dow Corning high vacuum grease to avoid water loss.

2.5.2.

Tensile strength measurements

Tensile tests were performed by using the Insight 10 kN Electromechanical Testing System (MTS, Eden Prairie, MN, USA), according to by ISO 527-2:1993 for tensile properties of plastics. Young’s Modulus, maximum tensile strength and strain at break were evaluated using type IV probes and an extensional rate of 10 mm min−1 at room temperature.

2.5.3.

Water uptake capacity

Water uptake capacity of bioplastics is measured according to the standard method for determining water absorption in plastics ASTM, 2005. Rectangular specimens of 60 × 10 × 1 mm were used. The specimens are subjected to drying (conditioning) in an oven at 50 ± 2 ◦ C for 5–6 h to determine dry weight, then introduced into distilled water and weighed after 2 and 24 h immersion. Finally, it is subjected to drying (reconditioning) again and weighed to determine the loss of soluble material. All the experiments are performed in triplicate at room temperature. According to the methodology used, water absorption capacity and loss of soluble material are determined by the following equations: %Water absorption =

Wet Weight − Initial Dry Weight × 100 Initial Dry Weight (1)

%Loss of soluble material =

Initial Dry weight − Final Dry weight × 100 Initial Dry weight

2.5.4.

Transparency measurements

Transparency measurements are performed on a Genesys-20 (Thermo Scientific, USA) spectrophotometer. In this device, the transmittance (%) of rectangular specimens, 1 mm thickness, at a selected wavelength of 600 nm is measured. Air is used as blank (100% transmittance). In order to compare the transparency of different bioplastics, a transmittance index (IT) was used: IT =

%Transmittance × 100 %Transmittance of reference bioplastic

2.6.

(3)

Statistical analysis

At least three replicates of each measurement were carried out. Statistical analyses were performed using t-test and oneway analysis of variance (ANOVA, p < 0.05) by means of the statistical package SPSS 18 (IBM, Chicago, IL, USA). Standard deviations from some selected parameters were calculated.

3.

Results and discussion

2.5.

Characterization of bioplastics

3.1.

Properties of PPI

2.5.1.

Dynamic mechanical analysis (DMA)

3.1.1.

Protein solubility and zeta potential

DMA tests were carried out with a RSA3 (TA Instruments), on rectangular probes using dual cantilever bending. All the experiments were carried out at constant frequency (6.28 rad/s) and strain (0.05%, within the linear viscoelastic region). The selected heating rate was 5 ◦ C min−1 . All the

(2)

Values of solubility and zeta potential for PPI as a function of pH are shown in Fig. 1. PPI solubility reached a maximum value of ca. 30% and presented a typical U-shaped solubility behaviour between pH 2 and 10, showing a minimum of solubility between pH 4 and 5. This behaviour is similar to

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

3.2.

Fig. 1 – Solubility (%) and Z-Potential (mV) of Pea Protein Isolate (PPI) as a function of pH values.

that observed in previous studies carried out by other authors (Shand et al., 2007; Osen et al., 2014). Zeta potential measurements of protein dispersions for different pH values provide information about the isoelectric point. The electric potential shifted from negative to positive values from basic to acid pH, passing through zero close to pH 5, which corresponds to the isoelectric point (IEP), where all charges are balanced. The IEP tends to coincide with the minimum solubility, as shown in Fig. 1.

3.1.2.

Molecular weights

Fig. 2 shows the electrophoresis pattern of pea protein where seven fractions can be clearly distinguished. As may be observed, there are bands between 13 and 73 kDa, and these can be associated with legumin and vicilin globular proteins, with another at 93 kDa, corresponding to lipoxygenase (Shand et al., 2007; Taherian et al., 2011). The intensity of these bands depends on pH. Bands with higher molecular weight (around 50, 70 and 93 kDa), which are associated with lipoxygenase, convicilin and vicilin, are particularly intense at pH 7, but may also be observed at pH 10, and are less apparent at pH 3. The legumin fraction is particularly sensitive to pH, such that the ˛-acidic subunit (ca. 38 kDa) appears at pH 7 and 10, whereas the lighter ˇ-basic subunit of legumin (ca. 23 kDa) is observed at pH 3, according to Liang and Tang (2013). Two further vicilin protein fractions (at ca. 13 and 19 kDa) can be observed at pH 10, being less visible at pH 7 and disappearing at pH 3 (Shand et al., 2007; Taherian et al., 2011). It is worth noticing that under acidic conditions only basic legumin can be detected. Under such conditions, other fractions such as vicilin or lipoxygenase seem to be forming larger insoluble aggregates.

3.1.3.

Free and total sulphydryl group content

Our measurements indicate that PPI has a markedly higher content of total sulphydryls (1694 ± 280 ␮mol g−1 PPI) than free sulphydryl (S H) (40 ± 9 ␮mol g−1 PPI), indicating that most sulphydryl groups are forming disulphide bonds, and that larger protein units are formed. This is consistent with the low protein solubility observed in the previous experiments. In any case, total sulphydryl content of PPI is higher than that found for other proteins like albumen protein isolate (Felix et al., 2014).

103

Preparation and characterization of blends

As indicated above, mixing is the first stage in the thermochemical processing of the protein-based bioplastics studied. A suitable selection of mixing conditions is very important; however, it is not always easy to identify the optimal conditions. On the one hand, extensive mixing is required to obtain a homogeneous dough-like blend. On the other hand, long mixing periods must be avoided to limit shear-induced structuring. As a consequence, both torque and temperature values must be monitored to control the overall mixing process. Fig. 3 shows torque (Fig. 3A) and temperature (Fig. 3B) as a function of mixing time (up to 60 min) for PPI/GL systems at four different ratios (80/20, 70/30, 60/40 and 50/50). At first, all the systems show an initial peak followed by a steady increase in torque and then values tend to level off, whereas the temperature increases constantly throughout the mixing time. Considering the different systems, the 80/20 system has much higher torque values than the others, with a dramatic increase in temperature. The high consistency of this blend, due to the low plasticizer content, has been found to be detrimental to its further processability. In fact, this system is an agglomerated powder rather than a homogeneous blend, as can be seen in Fig. 3. On the other hand, the 50/50 ratio dough-like material has rather low constant torque values and, although it may be easily injected, the final bioplastic materials tend to exude the excess of glycerol. It is on these grounds that both systems have been rejected, and the 70/30 and 60/40 PPI/GL systems, showing intermediate behaviours between the aforementioned extremes, have been selected as the most suitable for the injection process. From the torque and temperature profiles, it may be deduced that the point of minimum torque value should be selected as the operating time for mixing in order to achieve a compromise with the mixing time (long enough for a suitable homogenization but short enough to avoid premature cross-linking reactions of protein chains). Therefore, mixing time values as short as 10 and 4 min, which correspond to the minimum torque values for the 60/40 and 70/30 systems, respectively, were selected for the preparation of blends to be further processed. In this way, all the crosslinking potentials for structure formation and reinforcement can be exploited after injection into the mould. An interesting parameter that may be considered in this stage is the specific mechanical energy (SME) input for mixing, which is the energy provided by the mixer per unit mass, defined as follows:

SME =

ω m



tmix

M (t) dt

(4)

0

where, ω represents the rotational speed (in rad s−1 ) of the mixer, m is the total sample mass that is introduced (in kg), M is the torque (in N m) and tmix is the mixing time (in s). The values of SME for the different mixing times were calculated to be 611 and 1144 kJ kg−1 for the 60/40 and 70/30 systems, respectively. Obviously, the higher the PPI/GL ratio, the higher the blend consistency and the higher the specific mechanical energy required to achieve a homogeneous blend.

3.2.1.

DSC measurements

Heat flow patterns obtained from DSC measurements in a temperature range between 30 ◦ C and 170 ◦ C are shown in

104

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

Fig. 2 – SDS-PAGE of PPI dispersed in buffers of different pH values: 3, 7 and 10 (two lines for each pH correspond to replicates with different amount of protein). Enclosed, is a table with the molecular weight and protein fraction name of each signal.

Fig. 3 – Evolution of mixing torque and temperature over the mixing process for PPI/GL blends with different PPI/GL ratios. Pictures of the resulting doughs with different PPI/GL ratio are inserted.

Fig. 4 – DSC profiles for PPI and for different PPI/GL blends (60/40 and 70/30 PPI/GL ratio), subjected to thermoplastic mixing at 50 rpm for 10 min and 4 min, respectively. (Dashed line is used as a guide of the eye). Fig. 4 for PPI powder, and 70/30 and 60/40 PPI/GL dough-like blends. Three thermal events can be observed for the PPI profile: two endothermic DSC peaks and an inflection point. Endothermic peaks correspond to changes in the aggregation state of proteins within the isolate or blends. The first peak, which is located at 63 ◦ C and has not been previously associated with

any pea protein fraction, can be attributed to physical ageing phenomena. Physical ageing is a general process that occurs over time in glassy or partially glassy polymers below their Tg and is a manifestation of the non-equilibrium nature of the glassy state (Chartoff, 1997). This effect has also been detected in other protein systems such as crayfish flour (Farahnaky et al., 2008) and soy protein isolate (Felix et al., 2014). In fact, it should be noticed that this first peak no longer appears after mixing with glycerol (for 70/30 and 60/40 dough-like blends), confirming its reversible physical nature. In contrast, the second peak at 120 ◦ C may be related to protein denaturation at high temperatures and as a consequence is relevant for the selection of injection processing conditions. On the other hand, the glass transition temperature (Tg ) is the temperature at which a reversible thermodynamic pseudo-transition occurs in vitreous materials. In the case of PPI, Tg is observed to be around 100 ◦ C, being displaced to lower values as the PPI/GL ratio increases, being a consequence of the plasticizing effect of glycerol, which induces significant changes in the microstructure. It is worth mentioning that the use of temperatures higher than the Tg leads to systems showing higher mobility and lower viscosity, which may be also relevant for processing.

3.2.2.

Temperature ramps

Fig. 5 shows the elastic modulus (G ) and the loss tangent (tan ı) obtained from small amplitude oscillatory shear

105

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

Table 1 – Operating conditions selected for injection moulding of specimens with different PPI/GL ratios (60/40 and 70/30).

Pre-injection cylinder Injection Packing stage

T (◦ C)

Pressure (MPa)

Time (s)

50 50 130

0.1 0.1–50 50

100 <1 200

selection of a lower temperature (in the thermoplastic region) seems to be advisable.

3.3.

Fig. 5 – Storage modulus (G ) and loss tangent (tan ı) at 1 Hz and 5 ◦ C min−1 , as a function of temperature for 60/40 and 70/30 PPI/GL ratios.

measurements for 60/40 and 70/30 PPI/GL blends. The objective of these measurements was to select a suitable temperature to achieve moderate rheological properties and, as a consequence, to facilitate injection to the mould from the pre-injection cylinder, making it possible to obtain bioplastics with higher reproducibility and efficiency. The viscoelastic behaviour of the two plasticized blends is quite similar, with a pronounced initial decrease in G with temperature and, after reaching a minimum value (Tmin ), a marked increase, with G tending to recover, at least partially, to its initial level. This behaviour may be explained in terms of two opposite effects: a thermally-induced structural relaxation, which dominates below 60 ◦ C, and an enhancement of the network structure related to occurrence of thermallyinduced protein crosslinking, which becomes dominant at the highest temperature values. In any case, the tan ı remains much lower than 1 for the whole temperature range, reflecting the dominantly elastic character of these blends, with maxima at ca. 70 and 80 ◦ C for 60/40 and 70/30 systems, respectively. The occurrence of a tan ı peak is also related to a glass transition. Although the Tg values from small amplitude oscillatory shear tests are lower than those obtained from DSC measurements (Fig. 4), results from the two techniques may be considered consistent, the differences being attributable to different heating rates. Again, it may be seen that the higher the glycerol content, the higher the plasticization effect and the greater the reduction in Tg . However, the glycerol content seems to produce the opposite effect on parameter Tmin . In other words, Tmin is higher than Tg for the 60/40 blend, as expected, while Tmin is lower than Tg for the 70/30 blend. This result may be explained in terms of protein crosslinking that may occur over 70 ◦ C. Thus, for the former blend the increase in G takes place after reaching the Tg , but starts to take place even before the Tg for the later. According to these results, the optimum injection temperature should be as close as possible to Tmin but not exceeding it. On the other hand, it is important to avoid crosslinking in the cylinder, and for this reason, a moderate temperature should be selected, ensuring a thermoplastic behaviour for the blend over injection. Ideally, the selected temperature should be in the range Tg < T < Tmin if possible. However, in this case, the Tg values are so close to the protein crosslinking region that the

Preparation and characterization of bioplastics

Processing parameters for the pre-injection cylinder are selected according to the results obtained in previous experiments (Sections 3.2.1 and 3.2.2). The conditions selected for the pre-injection cylinder are 50 ◦ C and a residence time of 100 s. As mentioned above, the temperature should not be increased excessively and, in addition, the residence time should not be too long in order to prevent thermally-induced protein cross-linking effects before the injection stage. As for the mould conditions, the moulding temperature and time selected are 130 ◦ C and 200 s, respectively. It is also important to avoid exposure to high temperatures for a long time in order to avoid protein degradation. It should be noted that, according to DSC measurements, protein degradation does not take place at this temperature. Pressure values of 50 and 20 MPa were selected for the injection and packing stages, respectively, to ensure a suitable flow of blend and moulding of specimens. Table 1 summarises the conditions selected for the injection moulding process for each of the blends studied. Similar pressure conditions have been used previously for other proteins (Felix et al., 2014, 2015; Martín-Alfonso et al., 2014). These conditions are used for both systems (60/40 and 70/30 PPI/GL blends) in order to allow for the development of protein crosslinking to achieve the final network structure.

3.3.1.

DMTA

The viscoelastic behaviour of bioplastic specimens can be studied by means of DMTA measurements. Fig. 6A shows how the values of E and tan ı change with temperature for the two bioplastics (60/40 and 70/30 PPI/GL). Both systems exhibit the same behaviour, where, the elastic component (E ) is always above the viscous one (E ), as may be deduced from the values of tan ı, with both moduli decreasing markedly with temperature. The thermal profile for E passes through an inflection point, which may be related to a glass transition, and then tends to reach a rubbery-like plateau region at high temperatures. This plateau region indicates that no thermosetting potential is found under the selected experimental conditions, which evidences that no higher mould temperatures and no longer residence times are required. E values are higher for the specimens with a higher proportion of protein isolate (70%), indicating a more elastic response. Furthermore, the tan ı profile shows one peak, regardless of the proportion of protein used, which reveals a good compatibility between the protein and glycerol.

3.3.2.

Tensile strength measurements

The stress–strain curves obtained from uniaxial strength measurements for both systems (60/40 and 70/30) follow the typical pattern found for other protein-based bioplastic materials (Felix et al., 2014; Martín-Alfonso et al., 2014;

106

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

Fig. 6 – Results from mechanical tests carried out for different PPI/GL biobased specimens (60/40 and 70/30 PPI/GL ratio): (A) Storage modulus (E ) and loss tangent (tan ı) values from Dynamic Mechanical Thermal Analysis (DMTA) temperature ramp measurements performed at constant frequency (1 Hz) and heating rate (3 ◦ C min−1 ) and (B) Young’s modulus, maximum stress and strain at break from Tensile Strength measurements. Zárate-Ramírez et al., 2014a, b). All the curves have an initial linear region with a high constant stress–strain slope (that corresponds to the elastic modulus or Young’s Modulus) due to a good resistance at low deformation values. Then, there is a long plastic region with a continuous decrease in the stress–strain slope towards a constant value, after the elastic limit, reaching the maximum stress and strain at break, the rupture of the sample, at which point, there is a sudden decrease in stress (results not shown). Fig. 6B shows the values of the tensile parameters from tensile tests performed on the (60/40 and 70/30) PPI/GL bioplastics: Young’s modulus, maximum tensile strength and strain at break. Comparing results, a higher PPI/GL ratio results in markedly higher Young’s Modulus and maximum tensile strength values and, conversely, a lower strain at break. In fact, these results are in accordance with the higher values found for the elastic modulus from DMTA measurements. Moreover, as stated by Lagrain et al. (2010), increasing the elongation at break of glassy, amorphous polymers is typically achieved at the expense of the elastic modulus. This being the case, it is worth comparing the toughness, with values of 1.6 and 2.4 MPa m−3 for 60/40 and 70/30 samples, respectively. Toughness is calculated from the area under the stress–strain curve, as follows:



εmax

T=

 (ε) .dε

(5)

0

These results indicate that an increase in protein content leads to a greater ability to absorb energy before rupturing.

3.3.3.

Fig. 7 – Evolution of water absorption capacity (%) after immersion for 2 h and 24 h and loss soluble matter (%), and the transmittance index for 1 mm thick specimens for different PPI/GL ratios (60/40 and 70/30). Notably, the water absorption rate of both bioplastic materials is relatively fast, maximum absorption values being reached within the first 2 h of immersion, with no further increase seen after 24 h. It should also be mentioned that an increase in PPI/GL ratio does not yield any noticeable change in water uptake, which may be the result of two effects: water tends to fill the voids generated by glycerol migration, but it also tends to hydrate the hydrophilic parts of protein surfaces. The former effect is favoured when a higher amount of glycerol is present, but the latter is proportional to protein content. The two effects seem to counterbalance, giving rise to similar water uptake values from the first 2 h.

Water uptake capacity

Fig. 7 shows a comparison of water uptake measurements obtained after immersion of bioplastic samples for 2 and 24 h, as well as the loss of water-soluble matter for both (60/40 and 70/30) bioplastic specimens. For this last parameter, only the values calculated after 2 h are shown, since no significant changes were noticed between 2 and 24 h. As can be seen, the soluble material loss is very similar to the glycerol content in the specimen before immersion, which suggests that virtually all the highly hygroscopic glycerol is lost during the experiment. On the other hand, water uptake capacity is very high, being slightly higher than 100% in some cases, which may be attributed to the hydrophilic character of this protein system.

3.3.4.

Transparency measurements

Fig. 7 also shows the evolution of the transmittance index for the (60/40 and 70/30) PPI/GL specimens. In synthetic polymers, lower transparency is attributed to higher crystallinity of the specimen (Menges, 2002). However, in a biopolymer the concept is more complex, since the transparency is not only attributed to the crystallinity, but also to secondary reactions (mainly heat-induced) such as Maillard reactions that are associated with relevant changes in the colour and hence the transparency of the samples. For this reason, it is very important to use the same processing conditions when comparing different systems. The data obtained with our specimens

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

indicate that decreasing the proportion of PPI increases the transparency, the transmittance index having a lower value for the 70/30 than the 60/40 PPI/GL system. In other words, with a higher PPI content specimens are more crystalline and also more opaque.

4.

Concluding remarks

Characterization of PPI reveals that this is a protein isolate consisting of a wide variety of protein fractions showing different molecular weights, where a significant amount of them are highly denatured, as indicated by its low solubility and low content of free sulphydryls. Around 30–40% of the plasticizer (glycerol) has been found to be required to obtain good processability of PPI/GL blends to produce bioplastics. The excess of plasticizer occurring for a lower protein/plasticizer ratio yields blends with too low consistency for suitable processing and specimens showing glycerol exudation. On the other hand, increasing this ratio would produce some shear-induced crosslinking effects leading to hardly processing blends and excessively brittle specimens. A glass-like transition followed by a minimum in viscoelastic properties, prior to the onset of shear-induced crosslinking effects, has been found in PPI/GL blends. Both thermal events are considered useful to delimit suitable injection processing conditions. Within the aforementioned range, a higher PPI/GL ratio leads to an enhancement of the elastic bending and tensile properties of PPI/GL bioplastic specimens, as well as their ability to absorb mechanical energy before rupturing. It also improves water uptake, though to a lesser extent. On the other hand, the material becomes less transparent. Interestingly, considering their water uptake behaviour, these bioplastics show a fast absorption rate up to a very high capacity, such that they can be considered potential resources for the development of absorbent materials. From the overall results obtained, it is also apparent that PPI, being a by-product of the pea agroindustry, may be useful for producing bioplastics by means of a thermo-mechanical process consisting in two stages (mixing and injection moulding), providing that suitable processing conditions are selected. In addition, according to their mechanical and physicochemical properties, PPI-based bioplastics, coming from renewable resources, can be regarded as promising candidates for the substitution of conventional petroleum plastics in certain applications.

Acknowledgements This work is part of a research project sponsored by MINECO, “Ministerio de Economía y Competitividad”, from the Spanish Government (Ref. MAT2011-29275-C02-02/01) and by the Andalousian Government, (Spain) (project TEP-6134). The authors gratefully acknowledge their financial support. The authors also acknowledge the Microanalysis Service and Functional Characterisation Service (CITIUS-Universidad de Sevilla) for providing full access and assistance to the LECO-CHNS-932 and DSC Q20 Calorimetry (TA instruments), respectively.

107

References ASTM, 2005. Standard test method for plastics. In: Annual Book of ASTM Standards. American Society for Testing and Materials, Philadelphia, PA. Beltrán-Rico, M., Marcilla-Gomis, A., 2012. Tecnología de Polímeros. Publicaciones Universidad de Alicante, Alicante, ISBN: 978-84-9717-232-5. Beveridge, T., Toma, S.J., Nakai, S., 1974. Determination of SH- and SS-groups in some food proteins using Ellman’s reagent. J. Food Sci. 39, 49–51. Bradford, M.M., 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. Byun, Y., Teck Kim, Y., 2012. Innovations in Food Packaging, 2nd ed. Elsevier, London, UK, pp. 369–390, ISBN: 978-0-12-394601-0. Chartoff, R.P., 1997. Thermal Characterization of Polymeric Materials, 2nd ed. Academic Press, New York, pp. 484–744. Chao, D., He, R., Jung, S., Aluko, R.E., 2013. Effect of pressure or temperature pretreatment of isolated pea protein on properties of the enzymatic hydrolysates. Food Res. Int. 54, 1528–1534. Choi, W.S., Han, J.H., 2002. Film-forming mechanism and heat denaturation effects on the physical and chemical properties of pea-protein-isolate edible films. J. Food Sci. 67, 1399–1406. De Graaf, L.A., Kolster, P., 1998. Industrial proteins as a green alternative for ‘petro’ polymers: potentials and limitations. Macromol. Symp. 127, 51–58. di Gioia, L., Guilbert, S., 1999. Corn protein-based thermoplastic resins: effect of some polar and amphiphilic plasticizers. J. Agric. Food Chem. 47 (3), 1254–1261. Directive, 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC. Official Journal L 106, 17/04/2001 P. 0001 - 0039. Available from: http://eur-lex.europa.eu/legalcontent/EN/TXT/HTML/?uri=CELEX:32001L0018&from=ES, Accessed 12.09.10. Etheridge, R.D., Pesti, G.M., Foster, E.H., 1998. A comparison of nitrogen values obtained utilizing the Kjeldahl nitrogen and Dumas combustion methodologies (Leco CNS 2000) on samples typical of an animal nutrition analytical laboratory. Anim. Feed Sci. Technol. 73, 21–28. Feeney, R.E., Whitaker, J.R., 1988. Importance of cross-linking reactions in proteins. Adv. Cereal Sci.Technol. 9, 21–43. Felix, M., Martín-Alfonso, J.E., Romero, A., Guerrero, A., 2014. Development of albumen/soy biobased plastic materials processed by injection molding. J. Food Eng. 125, 7–16. Felix, M., Romero, A., Cordobés, F., Guerrero, A., 2015. Development of crayfish bio-based plastic materials processed by small-scale injection moulding. J. Sci. Food Agric. 95, 679–687. Fernández-Espada, L., Bengoechea, C., Cordobés, F., Guerrero, A., 2013. Linear viscoelasticity characterization of egg albumen/glicerol blends with applications in material moulding processes. Food Bioprod. Process. 91, 319–326. Gennadios, A., 2002. Protein Based Films and Coatings. CRC, New York, pp. 66–115. González-Gutiérrez, J., Partal, P., García-Morales, M., Gallegos, C., 2011. Effect of processing on the viscoelastic, tensile and optical properties of albumen/starch-based bioplastics. Carbohydr. Polym. 84, 308–315. Guerrero, P., Nur Hanani, Z.A., Kerry, J.P., de la Caba, K., 2011a. Characterization of soy protein-based films prepared with acids and oils by compression. J. Food Eng. 107, 41–49. Guerrero, P., Stefani, P.M., Ruseckaite, R.A., de la Caba, K., 2011b. Functional properties of film based on soy protein isolate and gelatin processed by compression molding. J. Food Eng. 105, 65–72. Hernandez-Izquierdo, V., Krochta, J., 2008. Thermoplastic processing of proteins for film formation—a review. J. Food Sci. 73 (2), R30–R39.

108

food and bioproducts processing 9 7 ( 2 0 1 6 ) 100–108

Jerez, A., Partal, P., Martinez, I., Gallegos, C., Guerrero, A., 2007. Protein-based bioplastics: effect of thermo-mechanical processing. Rheol. Acta 46, 711–720. Kokoszka, S., Debeaufort, F., Hambleton, A., Lenart, A., Voilley, A., 2010. Protein and glycerol contents affect physico-chemical properties of soy protein isolate-based edible films. J. Food Eng. 11, 503–510. Kowalczyk, D., Baraniak, B., 2011. Effect of plasticizers, pH and heating of film-forming solution on the properties of pea protein isolate films. J. Food Eng. 105, 295–305. Lagrain, B., Goderis, B., Brijs, K., Delcour, J.A., 2010. Molecular basis of processing wheat gluten toward biobased materials. Biomacromolecules 11, 533–541. Liang, H-N., Tang, C-H., 2013. pH-dependent emulsifying properties of pea [Pisum sativum (L)] proteins. Food Hydrocoll. 33, 309–319. Liu, W., Misra, M., Askeland, P., Drzal, L.T., Mohanty, A.K., 2005. ‘Green’ composites from soy based plastic and pineapple leaf fiber: fabrication and properties evaluation. Polymer 46, 2710–2721. Markwell, M.A.K., Haas, S.M., Bieber, L.L., Tolbert, N.E., 1978. Modification of Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal. Biochem. 87, 206–210. Martín-Alfonso, J.E., Félix, M., Romero, A., Guerrero, A., 2014. Development of new albumen based biocomposites formulations by injection moulding using chitosan as physicochemical modifier additive. Compos. Part B Eng. 61, 275–281. Matveev, Y., Grinberg, V., Tolstoguzov, V., 2000. The plasticizing effect of water on proteins, polysaccharides and their mixtures. Glassy state of biopolymers, food and seeds. Food Hydrocoll. 14, 425–437. Menges, G., 2002. Werkstoffkunde Kunststoffe, 5 Aufl. Haner Verlag, Müchen, ISBN: 3-446-21257-4. Mohammed, Z.H., Hill, S.E., Mitchell, J.R., 2000. Covalent crosslinking in heated protein systems. J. Food Sci. 65, 221–226. Osen, R., Toelstede, S., Wild, F., Eisner, P., Schweiggert-Weisz, U., 2014. High moisture extrusion cooking of pea protein isolates: raw material—characteristics, extruder responses, and texture properties. J. Food Eng. 127, 67–74. Pearson, A.M., 1983. In: Hudson, B.J.F. (Ed.), Developments in Food Proteins. , 2nd ed. Applied Science Publishers Ltd., Essex, England, pp. 67–108 (Soy proteins). Plastics Europe. 2008. The Compelling Facts about Plastics. An Analysis of Plastics Production, Demand and Recovery for 2006 in Europe: EUPC (the European Plastics Converters), EPRO

(European Association of Plastics Recycling and Recovery Organisations), & EuPr (the European Plastics Recyclres). Pommet, M., Redl, A., Morel, M.H., Guilbert, S., 2003. Study of wheat gluten plasticisation with fatty acids. Polymer 44, 115–122. Pouplin, M., Redl, A., Gontard, N., 1999. Glass transition of wheat gluten plasticized with water, glycerol, or sorbitol. J. Agric. Food Chem. 47, 538–543. Reddy, N., Yang, Y., 2013. Thermoplastic films from plant proteins. J. Appl. Polymer Sci. 130 (2), 729–738. Ribotta, P.D., Colombo, A., Rosell, C.M., 2012. Enzymatic modifications of pea protein and its application in protein cassava and corn starch gels. Food Hydrocoll. 27, 185–190. Shand, P.J., Ya, H., Pietrasik, Z., Wanasundara, P.K.J.P.D., 2007. Physicochemical and textual properties of heat-induced pea protein isolate gels. Food Chem. 102, 1119–1130. Shen, L., Haufe, J., Patel, M., 2009. Product Overview and Market Projection of Emerging Bio-Based Plastics. Utrecht University, Utrecht—The Netherlands, pp. 227 (PRO-BIP 2009 final report). Siracusa, V., Rocculi, P., Romani, S., Dalla Rosa, M., 2008. Biodegradable polymers for food packaging: a review. Trends Food Sci. Technol. 19, 634–643. Taherian, A.R., Mondor, M., Labranche, J., Drolet, H., Ippersiel, D., Lamarche, F., 2011. Comparative study of functional properties of commercial and membrane processed yellow pea protein isolates. Food Res. Int. 44, 2505–2514. Tan, W.F., Koopal, L.K., Weng, L.P., van Riemsdijk, W.H., Norde, W., 2008. Humic acid protein complexation. Geochim. Cosmochim. Acta 72, 2090–2099. Thannauser, T.W., Konishi, Y., Scheraga, H.A., 1984. Sensitive quantitative analysis of disulfide bonds in polypeptides and proteins. Anal. Biochem. 138, 181–188. Tummala, P., Liu, W., Drzal, L.T., Mohanty, A.K., Misra, M., 2006. Influence of plasticizers on thermal and mechanical properties and morphology of soy.based bioplastics. Ind. Eng. Chem. 45, 7491–7496. Zárate-Ramírez, L.S., Martínez, I., Romero, A., Partal, P., Guerrero, A., 2011. Wheat gluten-based materials plasticised with glycerol and water by thermoplastic mixing and thermomoulding. J. Sci. Food Agric. 91, 625–633. Zárate-Ramírez, L.S., Romero, A., Martínez, I., Bengoechea, C., Partal, P., Guerrero, A., 2014a. Effect of aldehydes on thermomechanical properties of gluten-based bioplastics. Food Bioprod. Process. 92, 20–29. Zárate-Ramírez, L.S., Romero, A., Bengoechea, C., Partal, P., Guerrero, A., 2014b. Thermo-mechanical and hydrophilic properties of polysaccharide/gluten-based bioplastics. Carbohydr. Polym. 112, 24–31.