- Email: [email protected]
Influence of oil extraction method on properties of canola biodiesel, epoxies, and protein-based plastics
Industrial Crops and Products 77 (2015) 133–138
Contents lists available at ScienceDirect
Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop
Influence of oil extraction method on properties of canola biodiesel, epoxies, and protein-based plastics Wajira A. Manamperi a , Judith D. Espinoza-Perez a , Darrin M. Haagenson a , Chad A. Ulven b , Dennis P. Wiesenborn a , Scott W. Pryor a,∗ a b
Department of Agricultural and Biosystems Engineering, North Dakota State University, Fargo, ND 58108, USA Department of Mechanical Engineering, North Dakota State University, Fargo, ND 58108, USA
a r t i c l e
i n f o
Article history: Received 12 March 2015 Received in revised form 16 August 2015 Accepted 22 August 2015 Keywords: Canola Oil extraction Biobased plastic Epoxy Biodiesel
a b s t r a c t Canola oil was extracted using a liquid carbon dioxide extraction process (LCE), laboratory screw-pressing, and industrial solvent extraction. Oilseed meal generated from these methods was used to extract proteins for protein-based plastics, and the oil was used to produce both epoxy resins and biodiesel. Protein isolates obtained from the LCE-generated meal produced plastics with higher toughness and elongation, but lower tensile strength and modulus, than those using meal obtained from screw pressing and solvent extraction. The oils extracted using the LCE process produced cured epoxy resins with higher flexural modulus than those produced from solvent-extracted oil. However, dynamic mechanical properties such as storage and loss modulus showed no significant differences with respect to oil extraction method. Oils obtained from extraction methods had similar fatty acid profiles and produced biodiesel with no significant differences in properties. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Efficient use of renewable resources to produce higher value products in advanced biorefineries will help reduce global dependence on petroleum. Oilseeds such as canola (Brassica napus L.) can be used to produce a diverse range of bioproducts. Biopolymers and biofuels can be produced from its oil and oilseed meals are rich in protein which can be used in the production of bioplastics. Bioproduct markets are expected to grow at an annual rate of 20–37% and the biobased plastics industry is predicted to reach $7.02 billion by 2018 (Lucintel, 2014). Biobased materials such as epoxidized vegetable oils are already used as plasticizers to provide flexibility and durability in PVC and other materials. A major proportion (more than 90%) of the epoxy-based plasticizers and neat plastics are produced using soy oil. However, competing products such as canola oil-based plastics not only have room for growth, but also will make the industry more vibrant with new products having a variety of properties. Finding efficient ways to extract canola oil with minimum use of petrochemical solvents while preserving quality of the by-products will benefit the canola industry as well as other oilseed processors.
∗ Corresponding author. E-mail address: [email protected] (S.W. Pryor). http://dx.doi.org/10.1016/j.indcrop.2015.08.050 0926-6690/© 2015 Elsevier B.V. All rights reserved.
Current biodiesel biorefineries are focused on fuel yield and quality with coproduct quality considered of secondary importance. Optimized production of both biofuels and coproducts may strengthen biofuel industries and reduce fossil fuel use. Oilseed meals produced in biodiesel production are rich in protein, and can be used for a variety of applications ranging from low-value animal feed to higher-value food additives and industrial applications such as plastics, biocomposites, and adhesives. Given the fact that oilseed meals are produced in large volumes, finding higher value uses would have a particularly large impact on biofuel producers during periods of high feedstock prices. The standard technology for extracting oil from oilseeds such as canola includes screw-pressing and solvent extraction with hexane. Oilseed meal quality can be negatively affected from high temperatures generated during screw-pressing and from exposure to solvents (Camire, 1991; Sun et al., 1999; L’Hocine et al., 2006). Proteins processed in this manner become partially denatured altering their functionality and reducing flexibility and value for use in food or industrial processes. Protein quality is typically given little consideration in oil extraction processes but developing higher value uses for proteins could change that. Also, adopting less severe processing conditions including lower temperatures will provide the required flexibility to use the by-products in a wider variety of high-value applications.
134
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
Another approach for oilseed extraction uses a subcritical liquid carbon dioxide extraction method (LCE) (Ferreira et al., 1993; Stahl et al., 1980). Compared to supercritical carbon dioxide extraction it is carried out at lower pressures (<72.9 atm) and temperatures (<31.3 ◦ C) in the liquid phase (Mangold, 1983). The process can use chemical adjuncts such as ethanol to replace hexane, which is a known carcinogen and can be very difficult and dangerous to work with. Also, industries utilizing oilseed coproducts are beginning to request or specify non-hexane processing for marketing, safety, and regulatory reasons. The goal of this research was to explore how oil extraction technology influences the ability to develop more diverse oilseed biorefineries with increased coproduct options. The specific objective was to compare the effects of canola oil extraction using LCE, screw-pressing, and hexane extraction on the properties of plastics, epoxies, and biodiesel produced from the resulting meal and oil. 2. Materials and methods 2.1. Materials Raw canola seed and processed canola meal were obtained from Archer Daniels Midland (ADM) Company’s crushing plant at Velva, ND, USA. Seeds were cleaned according to USDA–GIPSA recommendations for canola (USDA, 2004). Hexane, acetone, ethanol, and isopropanol were purchased from EMD Chemicals (Gibbstown, NJ, USA). A biodegradable aliphatic co-polyester (PBI 001® ) (Natureplast Inc., Caen, France) was blended with proteins in the extrusion process; it had a density of 1.26 g cm−3 , a melt-mass flow rate of 0.45 g min−1 , and a melt temperature of 115 ◦ C. Glycerol, polyvinylpyrrolidone (PVP), and zinc sulfate (ZnSO4 ) were used without further treatment as plasticizer, compatibilizer, and crosslinker, respectively, in plastic preparation. A petroleum-based epoxy resin (Resinfusion 8603) was purchased from Huntsman Corp. (The Woodlands, TX, USA). The catalyst, Amberlite IR120H resin, was obtained from Acros Organics (Morris Plains, NJ, USA), and anhydrous magnesium sulfate was purchased from Mallinckrodt Baker, Inc. (Phillipsburg, NJ, USA). The amine curing agent used for epoxy testing was PACM-20 [Bis(paminocyclohexyl methane)] (TCI America, Portland, OR, USA). The mold release agent Frekote 770-NC was obtained from Henkel Corp. (Rocky Hill, CT, USA). An aqueous solution of hydrogen peroxide (50%), anhydrous sodium carbonate, and other chemicals (NaOH, KOH, methanol, acetic acid, and HCl) were purchased from EMD Chemicals (Gibbstown, NJ, USA). 2.2. Preparation of canola oil and meal flour Canola oil and meal were prepared using three oil extraction methods: (1) liquid carbon dioxide extraction (LCE) process at Cool Clean Technologies, Inc. (Eagan, MN, USA), (2) screw pressing followed by hexane extraction (SPE) process at North Dakota State University (Fargo, ND, USA), and, (3) industrial solvent extraction process (ISE) at ADM (Velva, ND, USA). The ISE method was a conventional hexane extraction process using the extraction parameters that are quite similar throughout the industry. Also, ground canola meal and oil obtained from ADM were used as industry standards to prepare protein isolates/protein-based plastics and biodiesel/epoxy resins, respectively. LCE oil and meal were prepared by grinding canola seed for at least 1 min using a coffee grinder and then drying for 24 h at 50 ◦ C until moisture content was less than 1% (measured using a Mettler Toledo LJ16 moisture analyzer, Mettler Toledo Inc., Columbus, OH, USA). Ground seed was stored in an air-tight container
until further processing. Before oil extraction, the ground seed was placed in a convection oven at 50 ◦ C for at least 1 h. A 50-g batch of ground seed was measured into a 500-mL beaker and mixed immediately with 250 mL of co-solvent (acetone, ethanol, or isopropanol). The mixture was then blended for 10 min using an Oster model BPST02-B (Jarden Corp., Boca Raton, FL, USA) professional series blender. The mixture was then transferred again to a 500-mL beaker and seed oil bodies were further ruptured and homogenized for 20 min using a Hielscher ultrasonic homogenizer. The mixture was immediately transferred to a 1-L LCE unit, filled with liquid CO2 (approximately 700 mL) and agitated for 5 min. Reactor pressure was approximately 6.2 MPa with a corresponding saturation temperature of 23 ◦ C. After soaking, liquid CO2 was drained from the material through a 1-m filter. The pressure in the separation vessel was slowly lowered causing the CO2 to evaporate leaving the extractables for collection. Canola oil and meal extracted using ethanol were used further to prepare biodiesel, epoxies, and protein-based plastics based on higher oil extraction efficiency, and higher protein content in the resultant protein isolates. SPE process of meal preparation used screw-pressing of seeds followed by solvent extraction. The moisture content of canola seeds was determined gravimetrically using a Mettler Toledo LJ16 moisture analyzer. Seed moisture content was adjusted to 7% by addition of distilled water followed by overnight equilibration. Seeds were then fed at 80 g min−1 to a model S87G Komet screw press (Monchengladbach, Germany) with a press head temperature of 70 ◦ C. A compression screw and a restriction die with a 6-mm (diameter) die opening were used. Screw rotation speed was maintained at 24 rpm. The resulting canola meal was immediately pressed again using the same conditions (except for moisture adjustment) to further remove oil. The meal was then ground using a Retsch ZM1 mill (Brinkmann Instruments Inc., Westbury, NY) and passed through a 25-mesh screen. The residual oil in the resulting canola flour was removed by solvent extraction with hexane for 24 h using a Soxhlet extraction unit. The defatted flour was desolventized in a fume hood at room temperature for 2 d. 2.3. Preparation of protein isolates Protein isolates were prepared as described previously (Manamperi et al., 2011). Canola meal flour (100 g) was dispersed in 400 mL of distilled water. The pH of the suspension was adjusted to 11 using NaOH (6 mol L−1 ) and stirred for 1 h to solubilize the proteins. The suspension was centrifuged at 5000 × g for 30 min to remove fiber and other suspended solids. The protein-rich supernatant was further filtered through cheese cloth and through a Whatman 41 filter paper. In the case of LCE isolate, supernatant was centrifuged again at 5000 × g for 30 min and carefully filtered using cheese cloth and Whatman 41 filter paper to remove the top fat layer (due to high fat content in meal) from the protein rich bottom layer. Protein in the supernatant was then precipitated by drop-wise addition of HCl (6 mol L−1 ) to lower the pH to 5.5. The precipitated proteins were recovered by centrifugation at 5000 × g for 30 min and freeze-dried at a freezing temperature of −25 ◦ C and a drying shelf temperature of 25 ◦ C. Lyophilized protein isolates were used for preparation of plastic specimens. The protein content in both isolates and meals were determined by Kjeldahl analysis (AOAC, 1995). 2.4. Preparation of protein-based plastics Canola protein-based plastic specimens were prepared according to the method described by Mungara et al. (2002) with modifications. Canola protein isolate (35% by weight), plasticizer (15%), synthetic co-polyester (40%), compatibilizer (2%), water (7%), and cross-linker (1%) were mixed mechanically until a homoge-
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
neous blend was obtained. Glycerol was used as plasticizer and PVP was used as the compatibilizer to mediate the interaction between protein isolates and synthetic co-polyester. Zinc sulfate was used as a cross-linker. The blend was allowed to equilibrate for at least 24 h at room temperature before further processing. The formulation was melt-compounded using a Leistritz Micro18/GL-40D (American Leistritz Extruder Corp., Somerville, NJ, USA) co-rotating twin-screw extruder. The screw rotation speed was set at 300 rpm, and the temperature profile for the extruder was maintained at 95, 116, 126, 136, 136, and 141 ◦ C from feeder to die. The extruded polymer strands were quenched immediately in room-temperature water and pelletized continuously using a model BT-25 pelletizer. The pellet moisture content was analyzed as described for raw canola seeds and adjusted to 7% (wet basis) by mixing distilled water and allowing to equilibrate overnight. Pellets were then fed to a Technoplas SIM 5080 (Technoplas, Inc., St. Marys, Australia) single-screw injection molder to form ASTM D638 TypeI tensile test specimens. The injection molder temperature profile was maintained at 116, 124, 127, 132, and 124 ◦ C from feeder to nozzle. 2.5. Tensile properties testing Tensile strength, modulus, toughness, and elongation of the injection molded specimens were analyzed using a model 5567 Instron Universal Testing Machine (Instron Corporation, Canton, MA, USA) according to the ASTM D638 testing method (ASTM, 2008). Five replicates were analyzed for each treatment. 2.6. Conversion of canola oil to epoxy resins Canola oil obtained from LCE and ISE processes were used to make canola oil-based epoxy resins as described previously (Espinoza-Perez et al., 2009). Canola oil (300 g), acetic acid (37.8 g), and catalyst Amberlite IR120H resin (ion exchange capacity of 1.8 eq L−1 ; mean particle size: 0.62 to 0.83 mm) (54.6 g) were added to the flask and mixed for 5 min. A 50% aqueous solution of hydrogen peroxide (180 g) was added dropwise in to the flask at a rate of 2 mL min−1 .The reaction was continued for 5 h at a temperature of 60 ◦ C while stirring at 500 rpm. Once the reaction was completed, the catalyst was separated using cheesecloth and the filtrate was held in a separatory funnel for 2 h. The aqueous layer at the bottom was discarded and the organic layer was washed three times with an equal volume of saturated sodium carbonate solution (50 ◦ C) followed by three washings with an equal volume of water (50 ◦ C). Anhydrous magnesium sulfate was added to the organic layer at a 1:5 weight ratio and the organic layer was dried overnight. The organic layer (filtrate) was filtered through Whatman No. 4 filter paper under vacuum and the residual magnesium sulfate was discarded. 2.7. Preparation of cured epoxy resins Epoxy formulations comprised of canola oil-based epoxy resin (30%), petroleum-based epoxy resin (Resinfusion 8603) (20%), and the amine curing agent (50%). The ingredients were weighed into polypropylene beakers and mixed manually for 5 min. The blends were then degassed under vacuum at 450 mm Hg for 5 min. A 65-g batch was used for each treatment. Aluminium molds (10 cm × 10 cm × 1 cm) were used to prepare samples. To enhance the easy release of samples from the mold, 3 layers of a mold release agent (Frekote 770-NC) was applied. Then the epoxy blends were poured into the molds and cured at 80 ◦ C for 1 h. After curing, the samples were removed from the mold and specimens were cut and sanded using a belt grinder and
135
a dual grinder/polisher according to the required dimensions for thermomechanical testing. 2.8. Dynamic mechanical properties testing A dynamic mechanical analyzer DMA Q800 (TA Instruments, Wilmington, DE, USA) operated with a dual cantilever clamp was used to measure the loss and storage modulus at 35 ◦ C, and the glass transition temperature (Tg ) of the neat resin specimens (3 mm × 12 mm × 60 mm). The data were collected in the temperature range of 30–160 ◦ C. The heating rate was maintained at 5 ◦ C min−1 at 1 Hz and amplitude of 15 m. Three replicates were used for each treatment. The data were analyzed by the Universal Analysis 2000 software (TA Instruments, Wilmington, DE) and the Tg was measured as the peak temperature of tan ı curve. 2.9. Flexural properties testing Three-point flexural tests were carried out on 5 specimens of each treatment using an Instron 5567 load frame. The tests were carried out according to the ASTM D790-03 (procedure A) using specimens with dimensions 5 mm × 10 mm × 100 mm (ASTM, 2003). The loading rate and displacement rate (which is a function of specimen thickness) was calculated according to the procedure prescribed by the standard. Data was analyzed using Bluehill 2 (Instron Corporation, Canton, MA, USA) software. 2.10. Transesterification of canola oil Transesterification was completed as described previously by Haagenson et al. (2010). The reaction mixture was prepared using a 30:5:1 molar ratio of methanol:canola oil:KOH. The reaction was carried out at 60 ◦ C while mechanically stirring for 1 h. After completion of the reaction, the mixture was transferred to a 2L separatory funnel and allowed to settle for at least 1 h. The glycerol phase was collected from the bottom of the funnel and discarded. The unrefined biodiesel (fatty acid methyl esters) was heated at 65 ◦ C under vacuum (17 kPa) for 30 min to remove residual methanol. Crude biodiesel was then transferred to a separatory funnel and washed 3 times with water. The washed biodiesel was transferred again to a vacuum flask and dried under vacuum (17 kPa) at 95 ◦ C for 30 min. The refined biodiesel samples were stored in dark bottles for further analysis. 2.11. Biodiesel quality analysis The fatty acid profiles of the samples were obtained using a gas chromatographic method described by Vick et al. (2004) with minor modifications. The kinematic viscosity was determined according to ASTM D445using a capillary viscometer (ASTM, 2010). The acid number was determined according to ASTM D664 (2009), by titrating with KOH using phenolphthalein as an indicator. Cloud point (CP) and pour point (PP) were determined according to ASTM standards D2500 and D97, respectively (ASTM, 2011, 2012). The oxidative stability was determined according to ASTM D6751 (2009) specifications (EN 14112) using an Omnion OSI apparatus. Total glycerol was measured by SafTest (MP Biomedicals, LLC; Solon, OH) according to manufacturer recommendations. 3. Results and discussion 3.1. Meal and protein properties The proximate analysis of canola meal samples and resultant protein isolates are shown in Table 1. Crude protein content of the meal preparations were comparable with each other ranging from
136
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
Table 1 Proximate analysis of canola meal preparations from different oil extraction protocols and corresponding protein isolates from those meal samples. Standard deviations are listed in parentheses. Material
Ash (%)
Crude protein (%)
Fat (%)
Canola meal
LCE–ethanol LCE–acetone LCE–isopropanol ISE SPE
6.6 (0.2) 6.6 (0.2) 6.6 (<0.1) 9.8 (<0.1) 7.6 (<0.1)
41.5 (0.6) 39.8 (0.4) 42.4 (0.1) 40.8 (0.3) 42.9 (0.2)
11.5 (0.3) 8.5 (0.1) 7.8 (0.1) 3.8 (0.2) 0.04 (<0.01)
Protein isolate
LCE–ethanol ISE SPE
2.5 (0.2) 6.7 (0.4) 4.7 (0.2)
86.5 (0.8) 81.4 (0.1) 86.5 (0.1)
0.51 (0.05) 6.9 (0.3) 0.08 (0.04)
10
18
800
16
700
14
600
8
500
6
400 300
4
200 2 0 LCE
ISE
Denaturation effects of protein isolates were reflected prominently in protein-based plastics. Denaturation helps unfold protein molecules enabling a higher degree of interaction between protein molecules with synthetic polyester molecules (Graiver et al., 2004; Mo and Sun, 2001). This is directly reflected in tensile strength and tensile modulus of the plastics (Fig. 1). SPE meal is expected to have the highest degree of denaturation due to use of both screw-pressing and a lengthy solvent extraction process (Soxhlet) to remove oil. The highest tensile strength and modulus values were seen in the plastics prepared using SPE-extracted protein isolates. The low temperature process adopted by LCE oil extraction is expected to impart minimum denaturation on the protein, and both tensile strength and modulus of those plastics were significantly lower than that of ISE and SPE protein-based plastics. The lower tensile strength values indicate a lower degree of protein denatu-
1.5
6
1
4
0
3.2. Protein-based plastic properties
2
8
0
approximately 40–43%, but the residual fat content of samples varied from near 0% in SPE meal to 11.5% in LCE ethanol-extracted meal. High residual oil content is a concern in terms of economics of oil extraction as well as in the protein isolation process due to the requirement of additional removal steps. However, the protein isolates prepared from the LCE ethanol-extracted meal showed very low fat content and the highest crude protein content. Therefore, this process was used to further extract oil and meal to be used in epoxies, biodiesel, and protein-based plastic analysis. Material from ISE and SPE processes were also used for comparison purposes with the LCE ethanol-extracted meal. Protein isolates from SPEextracted meal also showed very low fat contents and high crude protein content while isolate from ISE-extracted meal showed comparable but slightly higher fat content and relatively low crude protein content.
3 2.5
10
2
Fig. 1. Tensile strength and modulus of plastics prepared from protein isolates resulting from LCE, ISE, and SPE extraction protocols. Error bars represent standard deviation from replicate sample testing. All treatments were significantly different at the 95% confidence level.
Toughness
12
100 SPE
3.5 Elongation
Toughness (J)
Tensile modulus
900
Elongation (%)
Tensile strength (MPa)
12
Tensile strength
Tensile modulus (MPa)
14
0.5 0 LCE
ISE
SPE
Fig. 2. Elongation and toughness of plastics prepared from protein isolates resulting from LCE, ISE, and SPE extraction protocols. Error bars represent standard deviation from replicate sample testing. All treatments were significantly different at the 95% confidence level.
ration. Such proteins have a higher degree of processing flexibility as they can be readily denatured further to the desired level, while any denaturation already done cannot be reversed (Manamperi and Pryor, 2012). Opposite trends were observed in ductile properties of proteinbased plastics compared to strength and modulus (Fig. 2). Elongation and toughness are main indicators of the ductility of protein-based plastics. Flexibility of plastics decreases with the increasing denaturation of protein molecules, since unfolded protein molecules interact more with polyester and other protein molecules to form more entanglements and cross-linking throughout the polymer matrix. This increases rigidity of the plastics thus reducing elongation and toughness (Mo and Sun, 2001; Manamperi and Pryor, 2012). This is reflected in the low ductility (elongation and toughness) of plastics made from the more denatured SPE protein-isolates and high ductility in plastics made from the leastdenatured LCE protein isolates. Ductility is an important property which is often lacking in protein-based plastics. Therefore, results show that the LCE process produces protein isolates that are more versatile for diverse applications. If the final application requires a ductile material, the protein isolates can be used without modification. If higher strength and modulus are primary requirements, the isolates can be modified using other methods to achieve the required denaturation level, (Manamperi and Pryor, 2012; Graaf, 2000). The oils extracted using the LCE process were also compared with oils produced using the ISE process. These oils were used to prepare biodiesel as well as epoxidized neat resin plastics in order to compare their behavior in industrial applications. 3.3. Oil properties The fatty acid profiles of the oils extracted using ISE and LCE (ethanol, isopropyl alcohol, and acetone) processes are compared
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
137
Table 2 Fatty acid profile (%) of canola oil samples prepared with different oil extraction methods.
ISE LCE–ethanol LCE–isopropanol LCE–acetone
Palmitic16:0
Stearic18:0
Oleic18:1
Linoleic18:2
Linolenic18:3
Arachidic20:0
Eicosanoic20:1
4.4 4.1 3.9 4.0
2.1 2.0 2.0 2.0
64.7 65.4 65.6 65.4
18.3 17.7 17.4 17.5
8.2 7.8 8.3 8.6
0.7 0.6 0.7 0.6
1.1 1.1 1.2 1.2
Table 3 Thermo-mechanical properties of epoxy neat resin plastic specimens based on oil extraction process. Numbers in parentheses represent sample standard deviation. (Superscript letters in each column represent statistical difference at a p-value of 0.05).
Table 4 Mechanical properties of epoxy neat resin plastic specimens from oil generated from LCE (ethanol) and ISE extraction protocols. Sample standard deviations from replicate sample testing are shown in parentheses. (Superscript letters in each column represent statistical difference at a p-value of 0.05).
Property
LCE
ISE
Property
LCE
ISE
Tg (◦ C) Storage modulus (GPa) Loss modulus (MPa)
64.3 (0.3)B 1.3 (0.1)A 50.5 (1.6) A
65.7 (0.5)A 1.5 (<0.01)A 52.6 (0.6) A
Flexural strength (MPa) Flexural modulus (MPa) Toughness (J)
27.4 (1.1)A 775 (44)A 0.43 (0.01)A
30.6 (1.6)A 883 (15)B 0.34 (0.08)A
with that of commercial refined canola oil in Table 2. These results show that the degree of unsaturation in the oils used to prepare epoxies and biodiesel were similar and that the LCE extractions did not result in a different fatty acid profile compared to standard industrial practices. The Iodine Value, which varies with fatty acid profile, (Espinoza-Perez et al., 2009) for the LCE-ethanol and ISE oils were 113.3 and 114.7, respectively. Similar fatty acid composition and degree of unsaturation implies that the properties of epoxy resins and biodiesel prepared from these samples may not exhibit noticeable differences since both epoxy and biodiesel properties rely partially on fatty acid profile. 3.4. Epoxy properties Dynamic mechanical analysis (DMA) was carried out in order to analyze the viscoelastic and damping behavior of the epoxybased plastics. The properties such as glass transition temperature (Tg ), storage modulus (E ), and loss modulus (E ) were obtained using the stress response of the material as a function of time, temperature, and frequency (Table 3). The storage modulus measures the dynamic response (elastic component) of the material whereas the loss modulus measures the viscous or damping component. Both elastic and viscous components undergo changes when the phase transitions take place from crystalline to amorphous structures in the polymer. Tg was measured using the peak temperature of the tan ı curve in DMA analysis. Tg is a qualitative measure of the molecular weight distribution and the homogeneity of the polymer matrix (Miyagawa et al., 2004). The oil extraction method did not have a strong effect on the epoxy thermodynamic properties, which is logical based on the similar fatty acid profiles. Glass transition temperature of the neat resin plastics prepared from ISE oil was statistically higher than that of the LCE oil. However, this difference may not be due to the method of oil extraction, but rather to the differences in starting material. The minor differences in Tg values of neat resin plastics bear little practical significance. The storage modulus values of epoxidized neat resins are an indication of the energy stored in the material at a given temperature (35 ◦ C). At 35 ◦ C, the plastics are hard since the temperature is well below the amorphous-to-crystalline transition temperature (Tg ). Therefore, higher storage modulus at this temperature shows higher plastic rigidity. Storage modulus of epoxy-based plastics did not show any significant differences with respect to the type of oil used in the formulation (Table 3). The loss modulus is an indication of the ability of the polymer to dissipate energy. The loss modulus measured by DMA using a wide range of time and temperature may act as a fingerprint of
Table 5 Comparison of biodiesel properties from oil generated from LCE (ethanol) and ISE extraction protocols (n = 2). Sample standard deviations from replicate sample testing are shown in parentheses. Property
LCE
ISE
OSI (h) Cloud point (◦ C) Pour point (◦ C) Acid value (mg KOH −1 ) Total glycerin (wt%) Kinematic viscosity (mm2 s-1 )
12.0 (1.0) −2 (0.4) −18 (<0.1) 0.58 (0.06) 0.03 (0.01) 5.2 (0.2)
18.3 (2.7) −2 (0.7) −15 (2) 0.57 (0.26) 0.03 (0.01) 5.2 (0.1)
the material providing useful information of the molecular relaxations and hence the ultimate properties of the final product in real applications. Similar to storage modulus, the loss modulus showed no significant differences with respect to different processes from which the canola oil was obtained (Table 3). Results of mechanical property testing of LCE and ISE oil-based plastics are shown in Table 4. Differences between cured epoxy resins using oil from the different extraction processes were minimal. No significant difference was seen for either flexural strength or toughness. Flexural strength represents the maximum stress the material can withstand before rupture when subjected to a 3-point bending force. Flexural strength did not show significant differences with respect to type of material used in plastics. Toughness is the total energy that a material can absorb before break and is a measure of the ductility of the material. Flexural modulus is calculated by taking the ratio between stress and strain within the elastic limit. ISE oil-based cured epoxy resins had a significantly higher flexural modulus than LCE resins, indicating higher material stiffness. 3.5. Biodiesel properties The properties of biodiesel samples prepared from LCE and ISE canola oils are listed in Table 5. The oxidative stability is an important biodiesel quality parameter that impacts the storage capability (shelf life) of biodiesel. Oxidative stability depends on the properties such as fatty acid profile and the method of oil extraction, since the natural antioxidants can be extracted to different degrees (Catha et al., 2012). Usually high saturation in oil gives better oxidative stability while high unsaturation is better for cold flow properties such as a lower cloud point and pour point. However, the predominant fatty acid (oleic) in canola (Table 2) provides a balance between these properties hence meeting standards in both properties. Biodiesel from ISE oil had a significantly higher oxidative stability but values for both easily met the requirements of ASTM standards (>3 h). A similar study by Moser (2008) reported
138
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
that canola biodiesel prepared using a conventional method had an OSI of 6.4 h prior to degradation. Cold weather performance is a challenge in the biodiesel industry and cloud point and pour point are important measures of the cold flow properties (Dunn, 2009). Cloud point is the temperature at which the first wax crystallization (haziness) is observed and the pour point is the lowest temperature where free surface movement of biodiesel takes place. Both ISE and LCE oils gave desirable low cloud point and pour point values, but did not show any significant differences between two types of biodiesel. Generally canola biodiesel has superior cold properties compared to soy biodiesel due to its lower saturated fatty acid content (Canakci and Sanli, 2008). The acid value and total glycerin did not show significant differences between ISE and LCE biodiesel samples. The total glycerin is well below the maximum value allowed by ASTM standard (0.25%) while the acid value is very close but slightly above the limit of 0.5 mg KOH g−1 . This could be due to the potassium soaps produced during the catalytic reaction, which are then converted back to free fatty acids during the water washing step hence increasing the acid value of the biodiesel samples. Similar to other properties, the kinematic viscosity did not show any significant difference between ISE and LCE samples. Kinematic viscosity of both samples were within the required limits of 1.9–6.0 mm2 s−1 . 4. Conclusions The low-temperature LCE process for oil extraction had significant effects on the quality and subsequent industrial applications of canola meal compared to the ISE and SPE processes. The processing versatility of the protein isolates produced from LCE meal was higher due to its lower denaturation level. Low-denatured protein isolates are readily soluble in water and other solvents, which is beneficial in the food additives industry. LCE isolates gave better ductile properties such as higher elongation and toughness (more than 280% and 230% higher, respectively, than SPE-isolate-based plastics), which are often lacking in protein-based plastics. These properties could be beneficial in plastics requiring higher ductility in the final product such as overmolding applications. The strength and modulus of these plastics were lower than those produced using ISE and SPE isolates, but these properties could be enhanced using protein modification methods. The LCE oil extraction process did not produce canola oil enriched in unsaturated fatty acids. The fatty acid profiles were similar in both LCE and ISE oils. This was reflected in the similarity of both epoxy and biodiesel properties since oil fatty acid profile governs most of subsequent thermomechanical properties. Acknowledgements The authors would like to thank Cool Clean Technologies (Eagan, MN) and ADM (Velva, ND) for oil extraction and supply of canola seeds and meal. Funding from this work came from the ND Agricultural Products Utilization Commission (APUC). References AOAC, 1995. Official Methods of Analysis, fifteenth ed. Association of Official Analytical Chemists, Arlington. ASTM Standard D790-03, 2003. Standard Test Methods for Flexural Properties of Unreinforced and Reinforced Plastics and Electrical Insulating Materials. American Society for Testing and Materials, Philadelphia.
ASTM Standard D638-08, 2008. Standard Test for Tensile Properties. American Society for Testing and Materials, Philadelphia. ASTM D664, 2009. Standard Test Method for Acid Number of Petroleum Products by Potentiometric Titration. American Society for Testing and Materials, Phildaelphia. ASTM Standard D6751-09, 2009. Standard Specification for Biodiesel Fuel Blend Stock (B100) for Distillate Fuels. American Society for Testing and Materials, West Conshohocken. ASTM 445, 2010. Standard Test Method for Kinematic Viscosity of Transparent and Opaque Liquids (and the Calculation of Dynamic Viscosity). American Society for Testing and Materials, Phildaelphia. ASTM Standard D2500-11, 2011. Standard Test Method for Cloud Point of Petroleum Products. American Society for Testing and Materials, West Conshohocken. ASTM Standard D97-12, 2012. Standard Test Method for Pour Point of Petroleum Products. American Society for Testing and Materials, West Conshohocken. Camire, M., 1991. Protein functionality modification by extrusion cooking. J. Am. Oil Chem. Soc. 68, 200–205. Canakci, M., Sanli, H., 2008. Biodiesel production from various feedstocks and their effects on the fuel properties. J. Ind. Microbiol. Technol. 35 (5), 431–441. Dunn, R.O., 2009. Effects of minor constituents on cold flow properties and performance of biodiesel. Prog. Energy Combust. Sci. 35 (6), 481–489. Espinoza-Perez, J.D., Haagenson, D.M., Pryor, S.W., Ulven, C.A., Wiesenborn, D.P., 2009. Production and characterization of epoxidized canola oil. Trans. ASABE 52, 1213–1221. Ferreira, S.R.S., Meireles, M.A.A., Cabral, F.A., 1993. Extraction of essential oil of black pepper with liquid carbon dioxide. J. Food Eng. 20, 121–133. Graaf, L.A., 2000. Denaturation of proteins from a non-food perspective. J. Biotechnol. 79, 299–306. Graiver, D., Waikul, L.H., Berger, C., Narayan, R., 2004. Biodegradable soy protein–polyester blends by reactive extrusion process. J. Appl. Polym. Sci. 92, 3231–3239. Haagenson, D.M., Brudvik, R.L., Lin, H., Wiesenborn, D.P., 2010. Implementing an in situ alkaline transesterification method for canola biodiesel quality screening. J. Am. Oil Chem. Soc. 87, 1351–1358. L’Hocine, L., Boye, J.I., Arcand, Y., 2006. Composition and functional properties of soy protein isolates prepared using alternative defatting and extraction procedures. J. Food Sci. 71, C137–C145. Lucintel Global Bioplastic Industry, 2013–2018. Trend, Forecast, and Opportunity Analysis. http://www.lucintel.com/reports. (accessed Jnauary 2015.). Manamperi, W.A.R., Wiesenborn, D.P., Chang, S.K.C., Pryor, S.W., 2011. Effects of protein separation conditions on the functional and thermal properties of canola protein isolates. J. Food Sc. 76 (3), 266–273. Manamperi, W.A., Pryor, S.W., 2012. Properties of canola protein-based plastics and protein isolates modified using SDS and SDBS. J. Am. Oil Chem. Soc. 89, 541–549. Mangold, H.K., 1983. Liquefied gases and supercritical fluids in oilseed extraction. J. Am Oil. Chem. Soc. 60, 226–228. Miyagawa, H., Mohanty, A.K., Misra, M., Drzal, L.T., 2004. Thermo-physical and impact properties of epoxy containing epoxidized linseed oil, 1 anhydride-cured epoxy. Macromol. Mater. Eng. 289, 629–635. Mo, X., Sun, S., 2001. Thermal and mechanical properties of plastics molded from urea-modified soy protein isolates. J. Am. Oil Chem. Soc. 78, 867–872. Moser, B.R., 2008. Influence of blending canola, palm, soybean, and sunflower oil methyl esters on fuel properties of biodiesel. Energy Fuels 22 (6), 4301–4306. Mungara, P., Chang, T., Zhu, J., Jane, J., 2002. Processing and physical properties of plastics made from soy protein polyester blends. J. Polym. Environ. 10, 31–37. Catha, S.A.S., Arshad, M.U., Arshad, W., Hussain, A.I., Saif, M.J., Majeed, M., 2012. Effect of extraction regimes on antioxidant potential of rice (Oryza sativa) bran extracts. J. Chem. Soc. Pak. 34, 1290–1294. Stahl, E., Schuetz, E., Mangold, H.K., 1980. Extraction of seed oils with liquid and supercritical carbon dioxide. J. Agric. Food Chem. 28 (6), 1153–1157. Sun, X., Kim, H.-R., Mo, X., 1999. Plastic performance of soybean protein components. J. Am. Oil Chem. Soc. 76, 117–123. USDA, GIPSA, 2004. Grain Inspection Handbook—Book II. USDA, Grain Inspection Packers and Stockyards Administration, Washington, D.C, pp. 13 (chapter 3). Vick, B.A., Jan, C.C., Miller, J.F., 2004. Two-year study on the inheritance of reduced saturated fatty acid content in sunflower seed. Helia 27 (41), 25–40.
Contents lists available at ScienceDirect
Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop
Influence of oil extraction method on properties of canola biodiesel, epoxies, and protein-based plastics Wajira A. Manamperi a , Judith D. Espinoza-Perez a , Darrin M. Haagenson a , Chad A. Ulven b , Dennis P. Wiesenborn a , Scott W. Pryor a,∗ a b
Department of Agricultural and Biosystems Engineering, North Dakota State University, Fargo, ND 58108, USA Department of Mechanical Engineering, North Dakota State University, Fargo, ND 58108, USA
a r t i c l e
i n f o
Article history: Received 12 March 2015 Received in revised form 16 August 2015 Accepted 22 August 2015 Keywords: Canola Oil extraction Biobased plastic Epoxy Biodiesel
a b s t r a c t Canola oil was extracted using a liquid carbon dioxide extraction process (LCE), laboratory screw-pressing, and industrial solvent extraction. Oilseed meal generated from these methods was used to extract proteins for protein-based plastics, and the oil was used to produce both epoxy resins and biodiesel. Protein isolates obtained from the LCE-generated meal produced plastics with higher toughness and elongation, but lower tensile strength and modulus, than those using meal obtained from screw pressing and solvent extraction. The oils extracted using the LCE process produced cured epoxy resins with higher flexural modulus than those produced from solvent-extracted oil. However, dynamic mechanical properties such as storage and loss modulus showed no significant differences with respect to oil extraction method. Oils obtained from extraction methods had similar fatty acid profiles and produced biodiesel with no significant differences in properties. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Efficient use of renewable resources to produce higher value products in advanced biorefineries will help reduce global dependence on petroleum. Oilseeds such as canola (Brassica napus L.) can be used to produce a diverse range of bioproducts. Biopolymers and biofuels can be produced from its oil and oilseed meals are rich in protein which can be used in the production of bioplastics. Bioproduct markets are expected to grow at an annual rate of 20–37% and the biobased plastics industry is predicted to reach $7.02 billion by 2018 (Lucintel, 2014). Biobased materials such as epoxidized vegetable oils are already used as plasticizers to provide flexibility and durability in PVC and other materials. A major proportion (more than 90%) of the epoxy-based plasticizers and neat plastics are produced using soy oil. However, competing products such as canola oil-based plastics not only have room for growth, but also will make the industry more vibrant with new products having a variety of properties. Finding efficient ways to extract canola oil with minimum use of petrochemical solvents while preserving quality of the by-products will benefit the canola industry as well as other oilseed processors.
∗ Corresponding author. E-mail address: [email protected] (S.W. Pryor). http://dx.doi.org/10.1016/j.indcrop.2015.08.050 0926-6690/© 2015 Elsevier B.V. All rights reserved.
Current biodiesel biorefineries are focused on fuel yield and quality with coproduct quality considered of secondary importance. Optimized production of both biofuels and coproducts may strengthen biofuel industries and reduce fossil fuel use. Oilseed meals produced in biodiesel production are rich in protein, and can be used for a variety of applications ranging from low-value animal feed to higher-value food additives and industrial applications such as plastics, biocomposites, and adhesives. Given the fact that oilseed meals are produced in large volumes, finding higher value uses would have a particularly large impact on biofuel producers during periods of high feedstock prices. The standard technology for extracting oil from oilseeds such as canola includes screw-pressing and solvent extraction with hexane. Oilseed meal quality can be negatively affected from high temperatures generated during screw-pressing and from exposure to solvents (Camire, 1991; Sun et al., 1999; L’Hocine et al., 2006). Proteins processed in this manner become partially denatured altering their functionality and reducing flexibility and value for use in food or industrial processes. Protein quality is typically given little consideration in oil extraction processes but developing higher value uses for proteins could change that. Also, adopting less severe processing conditions including lower temperatures will provide the required flexibility to use the by-products in a wider variety of high-value applications.
134
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
Another approach for oilseed extraction uses a subcritical liquid carbon dioxide extraction method (LCE) (Ferreira et al., 1993; Stahl et al., 1980). Compared to supercritical carbon dioxide extraction it is carried out at lower pressures (<72.9 atm) and temperatures (<31.3 ◦ C) in the liquid phase (Mangold, 1983). The process can use chemical adjuncts such as ethanol to replace hexane, which is a known carcinogen and can be very difficult and dangerous to work with. Also, industries utilizing oilseed coproducts are beginning to request or specify non-hexane processing for marketing, safety, and regulatory reasons. The goal of this research was to explore how oil extraction technology influences the ability to develop more diverse oilseed biorefineries with increased coproduct options. The specific objective was to compare the effects of canola oil extraction using LCE, screw-pressing, and hexane extraction on the properties of plastics, epoxies, and biodiesel produced from the resulting meal and oil. 2. Materials and methods 2.1. Materials Raw canola seed and processed canola meal were obtained from Archer Daniels Midland (ADM) Company’s crushing plant at Velva, ND, USA. Seeds were cleaned according to USDA–GIPSA recommendations for canola (USDA, 2004). Hexane, acetone, ethanol, and isopropanol were purchased from EMD Chemicals (Gibbstown, NJ, USA). A biodegradable aliphatic co-polyester (PBI 001® ) (Natureplast Inc., Caen, France) was blended with proteins in the extrusion process; it had a density of 1.26 g cm−3 , a melt-mass flow rate of 0.45 g min−1 , and a melt temperature of 115 ◦ C. Glycerol, polyvinylpyrrolidone (PVP), and zinc sulfate (ZnSO4 ) were used without further treatment as plasticizer, compatibilizer, and crosslinker, respectively, in plastic preparation. A petroleum-based epoxy resin (Resinfusion 8603) was purchased from Huntsman Corp. (The Woodlands, TX, USA). The catalyst, Amberlite IR120H resin, was obtained from Acros Organics (Morris Plains, NJ, USA), and anhydrous magnesium sulfate was purchased from Mallinckrodt Baker, Inc. (Phillipsburg, NJ, USA). The amine curing agent used for epoxy testing was PACM-20 [Bis(paminocyclohexyl methane)] (TCI America, Portland, OR, USA). The mold release agent Frekote 770-NC was obtained from Henkel Corp. (Rocky Hill, CT, USA). An aqueous solution of hydrogen peroxide (50%), anhydrous sodium carbonate, and other chemicals (NaOH, KOH, methanol, acetic acid, and HCl) were purchased from EMD Chemicals (Gibbstown, NJ, USA). 2.2. Preparation of canola oil and meal flour Canola oil and meal were prepared using three oil extraction methods: (1) liquid carbon dioxide extraction (LCE) process at Cool Clean Technologies, Inc. (Eagan, MN, USA), (2) screw pressing followed by hexane extraction (SPE) process at North Dakota State University (Fargo, ND, USA), and, (3) industrial solvent extraction process (ISE) at ADM (Velva, ND, USA). The ISE method was a conventional hexane extraction process using the extraction parameters that are quite similar throughout the industry. Also, ground canola meal and oil obtained from ADM were used as industry standards to prepare protein isolates/protein-based plastics and biodiesel/epoxy resins, respectively. LCE oil and meal were prepared by grinding canola seed for at least 1 min using a coffee grinder and then drying for 24 h at 50 ◦ C until moisture content was less than 1% (measured using a Mettler Toledo LJ16 moisture analyzer, Mettler Toledo Inc., Columbus, OH, USA). Ground seed was stored in an air-tight container
until further processing. Before oil extraction, the ground seed was placed in a convection oven at 50 ◦ C for at least 1 h. A 50-g batch of ground seed was measured into a 500-mL beaker and mixed immediately with 250 mL of co-solvent (acetone, ethanol, or isopropanol). The mixture was then blended for 10 min using an Oster model BPST02-B (Jarden Corp., Boca Raton, FL, USA) professional series blender. The mixture was then transferred again to a 500-mL beaker and seed oil bodies were further ruptured and homogenized for 20 min using a Hielscher ultrasonic homogenizer. The mixture was immediately transferred to a 1-L LCE unit, filled with liquid CO2 (approximately 700 mL) and agitated for 5 min. Reactor pressure was approximately 6.2 MPa with a corresponding saturation temperature of 23 ◦ C. After soaking, liquid CO2 was drained from the material through a 1-m filter. The pressure in the separation vessel was slowly lowered causing the CO2 to evaporate leaving the extractables for collection. Canola oil and meal extracted using ethanol were used further to prepare biodiesel, epoxies, and protein-based plastics based on higher oil extraction efficiency, and higher protein content in the resultant protein isolates. SPE process of meal preparation used screw-pressing of seeds followed by solvent extraction. The moisture content of canola seeds was determined gravimetrically using a Mettler Toledo LJ16 moisture analyzer. Seed moisture content was adjusted to 7% by addition of distilled water followed by overnight equilibration. Seeds were then fed at 80 g min−1 to a model S87G Komet screw press (Monchengladbach, Germany) with a press head temperature of 70 ◦ C. A compression screw and a restriction die with a 6-mm (diameter) die opening were used. Screw rotation speed was maintained at 24 rpm. The resulting canola meal was immediately pressed again using the same conditions (except for moisture adjustment) to further remove oil. The meal was then ground using a Retsch ZM1 mill (Brinkmann Instruments Inc., Westbury, NY) and passed through a 25-mesh screen. The residual oil in the resulting canola flour was removed by solvent extraction with hexane for 24 h using a Soxhlet extraction unit. The defatted flour was desolventized in a fume hood at room temperature for 2 d. 2.3. Preparation of protein isolates Protein isolates were prepared as described previously (Manamperi et al., 2011). Canola meal flour (100 g) was dispersed in 400 mL of distilled water. The pH of the suspension was adjusted to 11 using NaOH (6 mol L−1 ) and stirred for 1 h to solubilize the proteins. The suspension was centrifuged at 5000 × g for 30 min to remove fiber and other suspended solids. The protein-rich supernatant was further filtered through cheese cloth and through a Whatman 41 filter paper. In the case of LCE isolate, supernatant was centrifuged again at 5000 × g for 30 min and carefully filtered using cheese cloth and Whatman 41 filter paper to remove the top fat layer (due to high fat content in meal) from the protein rich bottom layer. Protein in the supernatant was then precipitated by drop-wise addition of HCl (6 mol L−1 ) to lower the pH to 5.5. The precipitated proteins were recovered by centrifugation at 5000 × g for 30 min and freeze-dried at a freezing temperature of −25 ◦ C and a drying shelf temperature of 25 ◦ C. Lyophilized protein isolates were used for preparation of plastic specimens. The protein content in both isolates and meals were determined by Kjeldahl analysis (AOAC, 1995). 2.4. Preparation of protein-based plastics Canola protein-based plastic specimens were prepared according to the method described by Mungara et al. (2002) with modifications. Canola protein isolate (35% by weight), plasticizer (15%), synthetic co-polyester (40%), compatibilizer (2%), water (7%), and cross-linker (1%) were mixed mechanically until a homoge-
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
neous blend was obtained. Glycerol was used as plasticizer and PVP was used as the compatibilizer to mediate the interaction between protein isolates and synthetic co-polyester. Zinc sulfate was used as a cross-linker. The blend was allowed to equilibrate for at least 24 h at room temperature before further processing. The formulation was melt-compounded using a Leistritz Micro18/GL-40D (American Leistritz Extruder Corp., Somerville, NJ, USA) co-rotating twin-screw extruder. The screw rotation speed was set at 300 rpm, and the temperature profile for the extruder was maintained at 95, 116, 126, 136, 136, and 141 ◦ C from feeder to die. The extruded polymer strands were quenched immediately in room-temperature water and pelletized continuously using a model BT-25 pelletizer. The pellet moisture content was analyzed as described for raw canola seeds and adjusted to 7% (wet basis) by mixing distilled water and allowing to equilibrate overnight. Pellets were then fed to a Technoplas SIM 5080 (Technoplas, Inc., St. Marys, Australia) single-screw injection molder to form ASTM D638 TypeI tensile test specimens. The injection molder temperature profile was maintained at 116, 124, 127, 132, and 124 ◦ C from feeder to nozzle. 2.5. Tensile properties testing Tensile strength, modulus, toughness, and elongation of the injection molded specimens were analyzed using a model 5567 Instron Universal Testing Machine (Instron Corporation, Canton, MA, USA) according to the ASTM D638 testing method (ASTM, 2008). Five replicates were analyzed for each treatment. 2.6. Conversion of canola oil to epoxy resins Canola oil obtained from LCE and ISE processes were used to make canola oil-based epoxy resins as described previously (Espinoza-Perez et al., 2009). Canola oil (300 g), acetic acid (37.8 g), and catalyst Amberlite IR120H resin (ion exchange capacity of 1.8 eq L−1 ; mean particle size: 0.62 to 0.83 mm) (54.6 g) were added to the flask and mixed for 5 min. A 50% aqueous solution of hydrogen peroxide (180 g) was added dropwise in to the flask at a rate of 2 mL min−1 .The reaction was continued for 5 h at a temperature of 60 ◦ C while stirring at 500 rpm. Once the reaction was completed, the catalyst was separated using cheesecloth and the filtrate was held in a separatory funnel for 2 h. The aqueous layer at the bottom was discarded and the organic layer was washed three times with an equal volume of saturated sodium carbonate solution (50 ◦ C) followed by three washings with an equal volume of water (50 ◦ C). Anhydrous magnesium sulfate was added to the organic layer at a 1:5 weight ratio and the organic layer was dried overnight. The organic layer (filtrate) was filtered through Whatman No. 4 filter paper under vacuum and the residual magnesium sulfate was discarded. 2.7. Preparation of cured epoxy resins Epoxy formulations comprised of canola oil-based epoxy resin (30%), petroleum-based epoxy resin (Resinfusion 8603) (20%), and the amine curing agent (50%). The ingredients were weighed into polypropylene beakers and mixed manually for 5 min. The blends were then degassed under vacuum at 450 mm Hg for 5 min. A 65-g batch was used for each treatment. Aluminium molds (10 cm × 10 cm × 1 cm) were used to prepare samples. To enhance the easy release of samples from the mold, 3 layers of a mold release agent (Frekote 770-NC) was applied. Then the epoxy blends were poured into the molds and cured at 80 ◦ C for 1 h. After curing, the samples were removed from the mold and specimens were cut and sanded using a belt grinder and
135
a dual grinder/polisher according to the required dimensions for thermomechanical testing. 2.8. Dynamic mechanical properties testing A dynamic mechanical analyzer DMA Q800 (TA Instruments, Wilmington, DE, USA) operated with a dual cantilever clamp was used to measure the loss and storage modulus at 35 ◦ C, and the glass transition temperature (Tg ) of the neat resin specimens (3 mm × 12 mm × 60 mm). The data were collected in the temperature range of 30–160 ◦ C. The heating rate was maintained at 5 ◦ C min−1 at 1 Hz and amplitude of 15 m. Three replicates were used for each treatment. The data were analyzed by the Universal Analysis 2000 software (TA Instruments, Wilmington, DE) and the Tg was measured as the peak temperature of tan ı curve. 2.9. Flexural properties testing Three-point flexural tests were carried out on 5 specimens of each treatment using an Instron 5567 load frame. The tests were carried out according to the ASTM D790-03 (procedure A) using specimens with dimensions 5 mm × 10 mm × 100 mm (ASTM, 2003). The loading rate and displacement rate (which is a function of specimen thickness) was calculated according to the procedure prescribed by the standard. Data was analyzed using Bluehill 2 (Instron Corporation, Canton, MA, USA) software. 2.10. Transesterification of canola oil Transesterification was completed as described previously by Haagenson et al. (2010). The reaction mixture was prepared using a 30:5:1 molar ratio of methanol:canola oil:KOH. The reaction was carried out at 60 ◦ C while mechanically stirring for 1 h. After completion of the reaction, the mixture was transferred to a 2L separatory funnel and allowed to settle for at least 1 h. The glycerol phase was collected from the bottom of the funnel and discarded. The unrefined biodiesel (fatty acid methyl esters) was heated at 65 ◦ C under vacuum (17 kPa) for 30 min to remove residual methanol. Crude biodiesel was then transferred to a separatory funnel and washed 3 times with water. The washed biodiesel was transferred again to a vacuum flask and dried under vacuum (17 kPa) at 95 ◦ C for 30 min. The refined biodiesel samples were stored in dark bottles for further analysis. 2.11. Biodiesel quality analysis The fatty acid profiles of the samples were obtained using a gas chromatographic method described by Vick et al. (2004) with minor modifications. The kinematic viscosity was determined according to ASTM D445using a capillary viscometer (ASTM, 2010). The acid number was determined according to ASTM D664 (2009), by titrating with KOH using phenolphthalein as an indicator. Cloud point (CP) and pour point (PP) were determined according to ASTM standards D2500 and D97, respectively (ASTM, 2011, 2012). The oxidative stability was determined according to ASTM D6751 (2009) specifications (EN 14112) using an Omnion OSI apparatus. Total glycerol was measured by SafTest (MP Biomedicals, LLC; Solon, OH) according to manufacturer recommendations. 3. Results and discussion 3.1. Meal and protein properties The proximate analysis of canola meal samples and resultant protein isolates are shown in Table 1. Crude protein content of the meal preparations were comparable with each other ranging from
136
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
Table 1 Proximate analysis of canola meal preparations from different oil extraction protocols and corresponding protein isolates from those meal samples. Standard deviations are listed in parentheses. Material
Ash (%)
Crude protein (%)
Fat (%)
Canola meal
LCE–ethanol LCE–acetone LCE–isopropanol ISE SPE
6.6 (0.2) 6.6 (0.2) 6.6 (<0.1) 9.8 (<0.1) 7.6 (<0.1)
41.5 (0.6) 39.8 (0.4) 42.4 (0.1) 40.8 (0.3) 42.9 (0.2)
11.5 (0.3) 8.5 (0.1) 7.8 (0.1) 3.8 (0.2) 0.04 (<0.01)
Protein isolate
LCE–ethanol ISE SPE
2.5 (0.2) 6.7 (0.4) 4.7 (0.2)
86.5 (0.8) 81.4 (0.1) 86.5 (0.1)
0.51 (0.05) 6.9 (0.3) 0.08 (0.04)
10
18
800
16
700
14
600
8
500
6
400 300
4
200 2 0 LCE
ISE
Denaturation effects of protein isolates were reflected prominently in protein-based plastics. Denaturation helps unfold protein molecules enabling a higher degree of interaction between protein molecules with synthetic polyester molecules (Graiver et al., 2004; Mo and Sun, 2001). This is directly reflected in tensile strength and tensile modulus of the plastics (Fig. 1). SPE meal is expected to have the highest degree of denaturation due to use of both screw-pressing and a lengthy solvent extraction process (Soxhlet) to remove oil. The highest tensile strength and modulus values were seen in the plastics prepared using SPE-extracted protein isolates. The low temperature process adopted by LCE oil extraction is expected to impart minimum denaturation on the protein, and both tensile strength and modulus of those plastics were significantly lower than that of ISE and SPE protein-based plastics. The lower tensile strength values indicate a lower degree of protein denatu-
1.5
6
1
4
0
3.2. Protein-based plastic properties
2
8
0
approximately 40–43%, but the residual fat content of samples varied from near 0% in SPE meal to 11.5% in LCE ethanol-extracted meal. High residual oil content is a concern in terms of economics of oil extraction as well as in the protein isolation process due to the requirement of additional removal steps. However, the protein isolates prepared from the LCE ethanol-extracted meal showed very low fat content and the highest crude protein content. Therefore, this process was used to further extract oil and meal to be used in epoxies, biodiesel, and protein-based plastic analysis. Material from ISE and SPE processes were also used for comparison purposes with the LCE ethanol-extracted meal. Protein isolates from SPEextracted meal also showed very low fat contents and high crude protein content while isolate from ISE-extracted meal showed comparable but slightly higher fat content and relatively low crude protein content.
3 2.5
10
2
Fig. 1. Tensile strength and modulus of plastics prepared from protein isolates resulting from LCE, ISE, and SPE extraction protocols. Error bars represent standard deviation from replicate sample testing. All treatments were significantly different at the 95% confidence level.
Toughness
12
100 SPE
3.5 Elongation
Toughness (J)
Tensile modulus
900
Elongation (%)
Tensile strength (MPa)
12
Tensile strength
Tensile modulus (MPa)
14
0.5 0 LCE
ISE
SPE
Fig. 2. Elongation and toughness of plastics prepared from protein isolates resulting from LCE, ISE, and SPE extraction protocols. Error bars represent standard deviation from replicate sample testing. All treatments were significantly different at the 95% confidence level.
ration. Such proteins have a higher degree of processing flexibility as they can be readily denatured further to the desired level, while any denaturation already done cannot be reversed (Manamperi and Pryor, 2012). Opposite trends were observed in ductile properties of proteinbased plastics compared to strength and modulus (Fig. 2). Elongation and toughness are main indicators of the ductility of protein-based plastics. Flexibility of plastics decreases with the increasing denaturation of protein molecules, since unfolded protein molecules interact more with polyester and other protein molecules to form more entanglements and cross-linking throughout the polymer matrix. This increases rigidity of the plastics thus reducing elongation and toughness (Mo and Sun, 2001; Manamperi and Pryor, 2012). This is reflected in the low ductility (elongation and toughness) of plastics made from the more denatured SPE protein-isolates and high ductility in plastics made from the leastdenatured LCE protein isolates. Ductility is an important property which is often lacking in protein-based plastics. Therefore, results show that the LCE process produces protein isolates that are more versatile for diverse applications. If the final application requires a ductile material, the protein isolates can be used without modification. If higher strength and modulus are primary requirements, the isolates can be modified using other methods to achieve the required denaturation level, (Manamperi and Pryor, 2012; Graaf, 2000). The oils extracted using the LCE process were also compared with oils produced using the ISE process. These oils were used to prepare biodiesel as well as epoxidized neat resin plastics in order to compare their behavior in industrial applications. 3.3. Oil properties The fatty acid profiles of the oils extracted using ISE and LCE (ethanol, isopropyl alcohol, and acetone) processes are compared
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
137
Table 2 Fatty acid profile (%) of canola oil samples prepared with different oil extraction methods.
ISE LCE–ethanol LCE–isopropanol LCE–acetone
Palmitic16:0
Stearic18:0
Oleic18:1
Linoleic18:2
Linolenic18:3
Arachidic20:0
Eicosanoic20:1
4.4 4.1 3.9 4.0
2.1 2.0 2.0 2.0
64.7 65.4 65.6 65.4
18.3 17.7 17.4 17.5
8.2 7.8 8.3 8.6
0.7 0.6 0.7 0.6
1.1 1.1 1.2 1.2
Table 3 Thermo-mechanical properties of epoxy neat resin plastic specimens based on oil extraction process. Numbers in parentheses represent sample standard deviation. (Superscript letters in each column represent statistical difference at a p-value of 0.05).
Table 4 Mechanical properties of epoxy neat resin plastic specimens from oil generated from LCE (ethanol) and ISE extraction protocols. Sample standard deviations from replicate sample testing are shown in parentheses. (Superscript letters in each column represent statistical difference at a p-value of 0.05).
Property
LCE
ISE
Property
LCE
ISE
Tg (◦ C) Storage modulus (GPa) Loss modulus (MPa)
64.3 (0.3)B 1.3 (0.1)A 50.5 (1.6) A
65.7 (0.5)A 1.5 (<0.01)A 52.6 (0.6) A
Flexural strength (MPa) Flexural modulus (MPa) Toughness (J)
27.4 (1.1)A 775 (44)A 0.43 (0.01)A
30.6 (1.6)A 883 (15)B 0.34 (0.08)A
with that of commercial refined canola oil in Table 2. These results show that the degree of unsaturation in the oils used to prepare epoxies and biodiesel were similar and that the LCE extractions did not result in a different fatty acid profile compared to standard industrial practices. The Iodine Value, which varies with fatty acid profile, (Espinoza-Perez et al., 2009) for the LCE-ethanol and ISE oils were 113.3 and 114.7, respectively. Similar fatty acid composition and degree of unsaturation implies that the properties of epoxy resins and biodiesel prepared from these samples may not exhibit noticeable differences since both epoxy and biodiesel properties rely partially on fatty acid profile. 3.4. Epoxy properties Dynamic mechanical analysis (DMA) was carried out in order to analyze the viscoelastic and damping behavior of the epoxybased plastics. The properties such as glass transition temperature (Tg ), storage modulus (E ), and loss modulus (E ) were obtained using the stress response of the material as a function of time, temperature, and frequency (Table 3). The storage modulus measures the dynamic response (elastic component) of the material whereas the loss modulus measures the viscous or damping component. Both elastic and viscous components undergo changes when the phase transitions take place from crystalline to amorphous structures in the polymer. Tg was measured using the peak temperature of the tan ı curve in DMA analysis. Tg is a qualitative measure of the molecular weight distribution and the homogeneity of the polymer matrix (Miyagawa et al., 2004). The oil extraction method did not have a strong effect on the epoxy thermodynamic properties, which is logical based on the similar fatty acid profiles. Glass transition temperature of the neat resin plastics prepared from ISE oil was statistically higher than that of the LCE oil. However, this difference may not be due to the method of oil extraction, but rather to the differences in starting material. The minor differences in Tg values of neat resin plastics bear little practical significance. The storage modulus values of epoxidized neat resins are an indication of the energy stored in the material at a given temperature (35 ◦ C). At 35 ◦ C, the plastics are hard since the temperature is well below the amorphous-to-crystalline transition temperature (Tg ). Therefore, higher storage modulus at this temperature shows higher plastic rigidity. Storage modulus of epoxy-based plastics did not show any significant differences with respect to the type of oil used in the formulation (Table 3). The loss modulus is an indication of the ability of the polymer to dissipate energy. The loss modulus measured by DMA using a wide range of time and temperature may act as a fingerprint of
Table 5 Comparison of biodiesel properties from oil generated from LCE (ethanol) and ISE extraction protocols (n = 2). Sample standard deviations from replicate sample testing are shown in parentheses. Property
LCE
ISE
OSI (h) Cloud point (◦ C) Pour point (◦ C) Acid value (mg KOH −1 ) Total glycerin (wt%) Kinematic viscosity (mm2 s-1 )
12.0 (1.0) −2 (0.4) −18 (<0.1) 0.58 (0.06) 0.03 (0.01) 5.2 (0.2)
18.3 (2.7) −2 (0.7) −15 (2) 0.57 (0.26) 0.03 (0.01) 5.2 (0.1)
the material providing useful information of the molecular relaxations and hence the ultimate properties of the final product in real applications. Similar to storage modulus, the loss modulus showed no significant differences with respect to different processes from which the canola oil was obtained (Table 3). Results of mechanical property testing of LCE and ISE oil-based plastics are shown in Table 4. Differences between cured epoxy resins using oil from the different extraction processes were minimal. No significant difference was seen for either flexural strength or toughness. Flexural strength represents the maximum stress the material can withstand before rupture when subjected to a 3-point bending force. Flexural strength did not show significant differences with respect to type of material used in plastics. Toughness is the total energy that a material can absorb before break and is a measure of the ductility of the material. Flexural modulus is calculated by taking the ratio between stress and strain within the elastic limit. ISE oil-based cured epoxy resins had a significantly higher flexural modulus than LCE resins, indicating higher material stiffness. 3.5. Biodiesel properties The properties of biodiesel samples prepared from LCE and ISE canola oils are listed in Table 5. The oxidative stability is an important biodiesel quality parameter that impacts the storage capability (shelf life) of biodiesel. Oxidative stability depends on the properties such as fatty acid profile and the method of oil extraction, since the natural antioxidants can be extracted to different degrees (Catha et al., 2012). Usually high saturation in oil gives better oxidative stability while high unsaturation is better for cold flow properties such as a lower cloud point and pour point. However, the predominant fatty acid (oleic) in canola (Table 2) provides a balance between these properties hence meeting standards in both properties. Biodiesel from ISE oil had a significantly higher oxidative stability but values for both easily met the requirements of ASTM standards (>3 h). A similar study by Moser (2008) reported
138
W.A. Manamperi et al. / Industrial Crops and Products 77 (2015) 133–138
that canola biodiesel prepared using a conventional method had an OSI of 6.4 h prior to degradation. Cold weather performance is a challenge in the biodiesel industry and cloud point and pour point are important measures of the cold flow properties (Dunn, 2009). Cloud point is the temperature at which the first wax crystallization (haziness) is observed and the pour point is the lowest temperature where free surface movement of biodiesel takes place. Both ISE and LCE oils gave desirable low cloud point and pour point values, but did not show any significant differences between two types of biodiesel. Generally canola biodiesel has superior cold properties compared to soy biodiesel due to its lower saturated fatty acid content (Canakci and Sanli, 2008). The acid value and total glycerin did not show significant differences between ISE and LCE biodiesel samples. The total glycerin is well below the maximum value allowed by ASTM standard (0.25%) while the acid value is very close but slightly above the limit of 0.5 mg KOH g−1 . This could be due to the potassium soaps produced during the catalytic reaction, which are then converted back to free fatty acids during the water washing step hence increasing the acid value of the biodiesel samples. Similar to other properties, the kinematic viscosity did not show any significant difference between ISE and LCE samples. Kinematic viscosity of both samples were within the required limits of 1.9–6.0 mm2 s−1 . 4. Conclusions The low-temperature LCE process for oil extraction had significant effects on the quality and subsequent industrial applications of canola meal compared to the ISE and SPE processes. The processing versatility of the protein isolates produced from LCE meal was higher due to its lower denaturation level. Low-denatured protein isolates are readily soluble in water and other solvents, which is beneficial in the food additives industry. LCE isolates gave better ductile properties such as higher elongation and toughness (more than 280% and 230% higher, respectively, than SPE-isolate-based plastics), which are often lacking in protein-based plastics. These properties could be beneficial in plastics requiring higher ductility in the final product such as overmolding applications. The strength and modulus of these plastics were lower than those produced using ISE and SPE isolates, but these properties could be enhanced using protein modification methods. The LCE oil extraction process did not produce canola oil enriched in unsaturated fatty acids. The fatty acid profiles were similar in both LCE and ISE oils. This was reflected in the similarity of both epoxy and biodiesel properties since oil fatty acid profile governs most of subsequent thermomechanical properties. Acknowledgements The authors would like to thank Cool Clean Technologies (Eagan, MN) and ADM (Velva, ND) for oil extraction and supply of canola seeds and meal. Funding from this work came from the ND Agricultural Products Utilization Commission (APUC). References AOAC, 1995. Official Methods of Analysis, fifteenth ed. Association of Official Analytical Chemists, Arlington. ASTM Standard D790-03, 2003. Standard Test Methods for Flexural Properties of Unreinforced and Reinforced Plastics and Electrical Insulating Materials. American Society for Testing and Materials, Philadelphia.
ASTM Standard D638-08, 2008. Standard Test for Tensile Properties. American Society for Testing and Materials, Philadelphia. ASTM D664, 2009. Standard Test Method for Acid Number of Petroleum Products by Potentiometric Titration. American Society for Testing and Materials, Phildaelphia. ASTM Standard D6751-09, 2009. Standard Specification for Biodiesel Fuel Blend Stock (B100) for Distillate Fuels. American Society for Testing and Materials, West Conshohocken. ASTM 445, 2010. Standard Test Method for Kinematic Viscosity of Transparent and Opaque Liquids (and the Calculation of Dynamic Viscosity). American Society for Testing and Materials, Phildaelphia. ASTM Standard D2500-11, 2011. Standard Test Method for Cloud Point of Petroleum Products. American Society for Testing and Materials, West Conshohocken. ASTM Standard D97-12, 2012. Standard Test Method for Pour Point of Petroleum Products. American Society for Testing and Materials, West Conshohocken. Camire, M., 1991. Protein functionality modification by extrusion cooking. J. Am. Oil Chem. Soc. 68, 200–205. Canakci, M., Sanli, H., 2008. Biodiesel production from various feedstocks and their effects on the fuel properties. J. Ind. Microbiol. Technol. 35 (5), 431–441. Dunn, R.O., 2009. Effects of minor constituents on cold flow properties and performance of biodiesel. Prog. Energy Combust. Sci. 35 (6), 481–489. Espinoza-Perez, J.D., Haagenson, D.M., Pryor, S.W., Ulven, C.A., Wiesenborn, D.P., 2009. Production and characterization of epoxidized canola oil. Trans. ASABE 52, 1213–1221. Ferreira, S.R.S., Meireles, M.A.A., Cabral, F.A., 1993. Extraction of essential oil of black pepper with liquid carbon dioxide. J. Food Eng. 20, 121–133. Graaf, L.A., 2000. Denaturation of proteins from a non-food perspective. J. Biotechnol. 79, 299–306. Graiver, D., Waikul, L.H., Berger, C., Narayan, R., 2004. Biodegradable soy protein–polyester blends by reactive extrusion process. J. Appl. Polym. Sci. 92, 3231–3239. Haagenson, D.M., Brudvik, R.L., Lin, H., Wiesenborn, D.P., 2010. Implementing an in situ alkaline transesterification method for canola biodiesel quality screening. J. Am. Oil Chem. Soc. 87, 1351–1358. L’Hocine, L., Boye, J.I., Arcand, Y., 2006. Composition and functional properties of soy protein isolates prepared using alternative defatting and extraction procedures. J. Food Sci. 71, C137–C145. Lucintel Global Bioplastic Industry, 2013–2018. Trend, Forecast, and Opportunity Analysis. http://www.lucintel.com/reports. (accessed Jnauary 2015.). Manamperi, W.A.R., Wiesenborn, D.P., Chang, S.K.C., Pryor, S.W., 2011. Effects of protein separation conditions on the functional and thermal properties of canola protein isolates. J. Food Sc. 76 (3), 266–273. Manamperi, W.A., Pryor, S.W., 2012. Properties of canola protein-based plastics and protein isolates modified using SDS and SDBS. J. Am. Oil Chem. Soc. 89, 541–549. Mangold, H.K., 1983. Liquefied gases and supercritical fluids in oilseed extraction. J. Am Oil. Chem. Soc. 60, 226–228. Miyagawa, H., Mohanty, A.K., Misra, M., Drzal, L.T., 2004. Thermo-physical and impact properties of epoxy containing epoxidized linseed oil, 1 anhydride-cured epoxy. Macromol. Mater. Eng. 289, 629–635. Mo, X., Sun, S., 2001. Thermal and mechanical properties of plastics molded from urea-modified soy protein isolates. J. Am. Oil Chem. Soc. 78, 867–872. Moser, B.R., 2008. Influence of blending canola, palm, soybean, and sunflower oil methyl esters on fuel properties of biodiesel. Energy Fuels 22 (6), 4301–4306. Mungara, P., Chang, T., Zhu, J., Jane, J., 2002. Processing and physical properties of plastics made from soy protein polyester blends. J. Polym. Environ. 10, 31–37. Catha, S.A.S., Arshad, M.U., Arshad, W., Hussain, A.I., Saif, M.J., Majeed, M., 2012. Effect of extraction regimes on antioxidant potential of rice (Oryza sativa) bran extracts. J. Chem. Soc. Pak. 34, 1290–1294. Stahl, E., Schuetz, E., Mangold, H.K., 1980. Extraction of seed oils with liquid and supercritical carbon dioxide. J. Agric. Food Chem. 28 (6), 1153–1157. Sun, X., Kim, H.-R., Mo, X., 1999. Plastic performance of soybean protein components. J. Am. Oil Chem. Soc. 76, 117–123. USDA, GIPSA, 2004. Grain Inspection Handbook—Book II. USDA, Grain Inspection Packers and Stockyards Administration, Washington, D.C, pp. 13 (chapter 3). Vick, B.A., Jan, C.C., Miller, J.F., 2004. Two-year study on the inheritance of reduced saturated fatty acid content in sunflower seed. Helia 27 (41), 25–40.