Sapucaia nut (Lecythis pisonis Cambess.) flour as a new industrial ingredient: Physicochemical, thermal, and functional properties

Food Research International 109 (2018) 572–582

Contents lists available at ScienceDirect

Food Research International journal homepage: www.elsevier.com/locate/foodres

Sapucaia nut (Lecythis pisonis Cambess.) flour as a new industrial ingredient: Physicochemical, thermal, and functional properties

T

Gerson Lopes Teixeira, Suelen Ávila, Polyanna Silveira Hornung, Rafaela Cristina Turola Barbi, ⁎ Rosemary Hoffmann Ribani Food Engineering Graduate Program, Federal University of Paraná, Polytechnic Center, Jardim das Américas, Curitiba, Paraná 81531-980, Brazil

A R T I C LE I N FO

A B S T R A C T

Keywords: Defatted nut meal Brazilian edible nut Pressurized fluids Functional properties Emulsions

The aim of this work was to investigate the physicochemical, thermal, and functional properties of partially defatted sapucaia nut (Lecythis pisonis Cambess.) flours (PDSF) degreased by subcritical propane (20–60 °C; 20–100 bar) and supercritical CO2 + ethanol (1:1 w/w) as co-solvent (60 °C; 200 bar) in comparison to the PDSF obtained through Soxhlet extraction with petroleum ether. Under the conditions studied herein, compressed propane has a minor effect on the granules' morphology (average particle size between 22 and 32 μm) or in the physicochemical characteristics of the PDSF. It caused a minimum impact on the nutritional profile of the samples; unlike, the thermogravimetric analysis revealed that there is an influence on the thermal stability of the PDSF. Functional characteristics, such as emulsifying (8–20 m2/g), foam (6–12%), and high water (0.35–1.38 g/ g flour) and oil (0.72–1.57 g/g flour) absorption capacity, were observed in PDSF. Defatted flours were found to be effective in the production of emulsions with structures that showed micrometric-sized droplets (up to 85% droplet size < 15.0 μm) with alleged stability. PDSF is a source of proteins (31–49%) and carbohydrates (17–31%), thus it can be used as an ingredient to produce foodstuff in bakery and confectionery aiming to increase their nutritional value and functional properties.

1. Introduction

phenolic compounds with antioxidant characteristics (Teixeira et al., 2017). Presenting low moisture content (~3 to 10%), sapucaia also stands out for its carbohydrate and protein content, which can range from 5 to 11%, and 18 to 29%, respectively. The mineral residue varies around 3%, and the fiber content is ~7% (de Carvalho et al., 2012; de Carvalho, da Costa, de Souza, & Maia, 2008; Vallilo et al., 1999). Extraction of oil from sapucaia nut has been performed from different methods using organic solvents (Teixeira et al., 2017; Vallilo et al., 1999), by cold pressing (Costa & Jorge, 2012; Demoliner et al., 2018), or using green solvents such as compressed propane and supercritical carbon dioxide (Teixeira et al., 2018), but the available studies didn't focus on the defatted residue. Researches featuring de-oiled nut residues such as physic nut (Jatropha curcas) (Das et al., 2011) hazelnut (Turan, Capanoglu, & Altay, 2015) and Brazil nut (Carvalho et al., 2015) shows the potential of this type of by-products. Targeting agroindustry matrices for the production of foodstuffs is crucial for the valorization of the national productive chain. Thereby, the economy is linked to the production of that raw materials, resulting in higher added value. Besides, the indirect reduction in the actions of deforestation occurs as a consequence of the valorization of the vegetable species (Teixeira et al., 2018; Teixeira, Züge, Silveira, Scheer, &

Brazil has a wide variety of edible nuts and walnuts, utilized in a range of food segments including the use of raw material (in natura), to be used as seasonings, condiments and in the extraction of oils or beverage production. Most nuts are a rich in macro and micronutrients, as well as bioactive compounds. Furthermore, nut intake is directly related to the prevention of cardiovascular diseases and hypertension (Carvalho et al., 2015). Some underutilized Brazilian nuts, as the sapucaia (Lecythis pisonis Cambess.), have been studied along the past twenty years. The reports indicate a composition of nutritional importance, besides presenting great potential for the food industry (Demoliner et al., 2018; Denadai et al., 2007; Naozuka, Vieira, Nascimento, & Oliveira, 2011; Teixeira, Ávila, Silveira, Ribani, & Ribani, 2017; Teixeira, Ghazani, Corazza, Marangoni, & Ribani, 2018; Vallilo, Tavares, Pimentel, Badolato, & Inomata, 1998). The sapucaia nut has a chemical composition which changes according to the place of cultivation and harvesting, and conditions of production, among other factors. The lipid content can reach up to 63% (Vallilo et al., 1998; Vallilo, Tavares, Aued-Pimentel, Campos, & Neto, 1999), and shows a composition rich in unsaturated fatty acids and

⁎

Corresponding author at: Food Engineering Graduate Program, Federal University of Paraná, Polytechnic Center, Jardim das Américas, Curitiba, Paraná 81531-980, Brazil. E-mail address: [email protected] (R.H. Ribani).

https://doi.org/10.1016/j.foodres.2018.04.071 Received 14 March 2018; Received in revised form 17 April 2018; Accepted 30 April 2018 Available online 01 May 2018 0963-9969/ © 2018 Elsevier Ltd. All rights reserved.

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

content was determined by Soxhlet (AOCS, 1997). The total carbohydrate content was estimated by difference. Energetic value was calculated by multiplying the carbohydrate and protein content by 4 kcal/g and that of lipids by 9 kcal/g. The color parameters were measured using the MiniScan XE plus colorimeter (HunterLab, Germany) and expressed in values of the CIELab color system, where L* = luminosity, a* = red/green coordinate, and b* = yellow/blue coordinate. The Chroma (color saturation or intensity), and Hue angle were obtained through the formulas: Chroma = [(a2 + b2)1/2]; Hue = [arc tangent (b/a)].

Ribani, 2016). To exemplify, the defatted meal or flour resulting from the extraction of oil from nuts presents a high industrial potential, as a rich source of protein and carbohydrates. Instead of treating them as a residue, they are used as a new product in some companies in Brazil. Additionally, the remaining lipid content in defatted flour is a positive factor, carrying beneficial characteristics of its fatty acids to the defatted meal. These types of flours have proven to show several characteristics of industrial interest, either for use in the production of humans foodstuffs (Bhise, Kaur, & Aggarwal, 2015; Chan, Khong, Iqbal, Mansor, & Ismail, 2013; Huiping Chen et al., 2015; Ganorkar, Patel, Shah, & Rangrej, 2016; Ling, Zhang, Li, & Wang, 2016; Sawada, Venâncio, Toda, & Rodrigues, 2014; Xu, Lui, Luo, & Diosady, 2003), for animal feed (Suprayudi et al., 2016), mushroom cultivation (Pardo-Giménez et al., 2016), or even in the production of adhesives (Zheng et al., 2017). Other products such as beet-sugar pectin (Chen et al., 2016), hemicellulose and cellulose from by-products as sorghum bran, bagasse and biomass (Qiu, Yadav, & Yin, 2017) present functional properties being successfully applied in production of emulsions. In order to have an adequate use of defatted flour for human consumption, in-depth studies on its characteristics are necessary, mainly to ascertain its functional properties, and discover its benefits and possible damages caused by the degreasing processes (e.g. sub- supercritical fluid extractions), and its applicability limitations for industry. Analytical techniques produce relevant results for industry, including thermal analysis (thermogravimetry), microscopic investigation, and colorimetric techniques, answering the diverse issues related to the studied matrix, guiding a correct application (Guimarães et al., 2012; Teixeira et al., 2016; Turan et al., 2015). In this contex, no information regarding the functional properties of sapucaia nut defatted flour has been found in the literature. Thus, this work aimed to evaluate the physicochemical composition and functional properties of sapucaia nut flours partially defatted by pressurized fluids (propane and CO2), as well as the influence of degreasing processes on the characteristics of the obtained products, comparing them with the flour obtained by classic Soxhlet extraction with petroleum ether.

2.2.2. Thermogravimetric analysis Thermogravimetric analyzes were performed on a TGA 4000 equipment (PerkinElmer Inc. Waltham, USA). Approximately 5 mg of the sample was placed in the platinum pan and then positioned in the oven where it was heated from 30 to 900 °C (10 °C/min) in a synthetic air atmosphere (50 mL/min flow). Data with sample weight changes were obtained and analyzed using Pyris™ software. The thermogravimetric curve (TG) and the derivative thermogravimetric (DTG) were analyzed with the software Origin 8.6 (OriginLab, Massachusetts, USA). The thermal stability was measured from the extrapolation of the initial temperature of the first thermal decomposition event of the respective TG curves. The initial and final temperatures of the respective DTG peaks were used as temperature limits for the instrument data analysis software. 2.2.3. Scanning electron microscopy Scanning electron microscopy (SEM) analysis was performed using a JEOL JSM 6360-LV microscope (Jeol Company, Tokyo, Japan). Sapucaia nut and PDSF fine powder samples were fixed on copper supports using a double-sided adhesive tape and then covered with gold (Au) coating. The micrographs were obtained under vacuum and 15 kV of voltage acceleration with a magnification of 1700×. The area of the granules was calculated using ImageJ 1.51 s software (ImageJ for Windows). 2.2.4. X-ray diffractograms (XRD) The X-ray diffraction patterns of the sapucaia nut and PDSF fine powder samples were investigated using a D8-Advance X-ray diffractometer (Bruker, USA) at 25 °C, employing Cu Kα radiation (λ = 1.5406 Å) from 5° to 60° (2θ), at the flow rate of 2°/min, and a step size of 0.060°.

2. Material and methods 2.1. Sapucaia nut samples The sapucaia nut flour was partially defatted through different degreasing processes presented in details in our previous work (Teixeira et al., 2018), by using three different methods: a) subcritical extraction using propane (L1-L5, 20–60 °C, 20–100 bar); b) supercritical extraction using carbon dioxide (CO2) + ethanol (1:1, w/w) as co-solvent (LC, 60 °C, 200 bar); c) classical Soxhlet extraction (AOCS, 1997) using petroleum ether (LS, control sample). The conditions for each process are summarized in Table 1. The partially defatted sapucaia nut flour (PDSF) samples were ground with the aid of a knife mill (MA630/1, Marconi Ltda., Brazil) for 30 s, resulting in a fine powder. All samples were vacuum packed in LDPE plastic bags and kept under refrigeration until further analysis.

2.2.5. Infrared spectroscopy Fourier transform infrared spectroscopy analysis in the diffuse reflectance mode (DRIFTS) was performed using a Vertex-70 spectrometer (Bruker, USA) with a diffuse reflectance accessory at 25 °C. Data were recorded in the range of 500 to 4000 cm−1 wavenumbers with a spectral resolution of 4 cm−1, and 1024 scans. The reflectance spectra were analyzed after the transformation of the percentage of reflectance into absorbance [log10 (1/Reflectance)] (Carioca & Ferreira, 2011). 2.3. Functional properties of the PDSF 2.3.1. Ultraviolet spectra For each sample, a solution containing 1 mg/mL of the PDSF was prepared in a 50-mL flask, slightly shaken, and then a 2-mL aliquot was transferred to Eppendorf and centrifuged (Heraeus Fresco 21, Fisher Scientific) at 5000 ×g for 5 min. The supernatant was then used to obtain the UV–Vis spectrum (UV-1800, Shimadzu, Kyoto, Japan) in the range of 190 to 400 nm.

2.2. Characterization of sapucaia nut and its PDSF 2.2.1. Proximate composition and color analysis Moisture and ash were determined by thermogravimetry (method described in Section 2.2.2), where moisture was estimated by mass loss between 30 and 150 °C, and the residue remaining at the end of the analysis at 900 °C was considered ashes (Kaspchak et al., 2017). The total N concentration was determined according to the AOCS Ba 4e-93 official method using a LECO FP-528 dry combustion Carbon/Nitrogen analyzer system (LECO, Michigan, USA). The crude protein content was calculated by multiplying the total N content by the factor 5.46. Lipid

2.3.2. Turbidity of the dispersions Dispersions containing 1% (w/v) PDSF were vortexed for 2 min, and then an aliquot was transferred to a quartz cuvette, and the absorbance measured in a Spectro 3000 W spectrophotometer (Scientific Mars, 573

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

Table 1 Proximate composition and color parameters for the sapucaia nut (NS) and its partially defatted flours. Parameter (%)

Sample NS

L1

(raw sample) g

(20°C; 20 bar) e

L2

(60°C; 20 bar) e

L3

L4

(20°C; 100 bar) d

(60°C; 100 bar) b

L5

(40°C; 60 bar) c

LC

(60°C; 200 bar) f

LS

(boiling point)

Moisture Protein Lipids Ashes Carbohydrates Energetic value1

4.19 ± 0.01 20.45 ± 0.41f 58.71 ± 0.30a 2.89 ± 0.01e 13.76 ± 0.70f 665.22 ± 1.49a

6.12 ± 0.01 32.00 ± 0.69d 36.02 ± 0.02c 2.08 ± 0.01h 23.79 ± 0.74d 547.25 ± 0.09c

6.30 ± 0.01 31.89 ± 0.46d 31.41 ± 0.58d 2.37 ± 0.01g 28.04 ± 0.11c 522.39 ± 2.93d

6.63 ± 0.01 36.16 ± 0.87b 26.91 ± 0.13e 2.91 ± 0.00d 27.40 ± 0.74c 496.41 ± 0.62e

7.93 ± 0.01 36.05 ± 0.86b 18.33 ± 0.46f 3.04 ± 0.01b 34.65 ± 1.32a 447.74 ± 2.30f

7.00 ± 0.24 34.08 ± 0.26c 27.12 ± 0.17e 2.87 ± 0.02f 29.93 ± 0.32c 496.12 ± 1.78e

4.79 ± 0.01 28.95 ± 0.28e 46.39 ± 0.55b 2.93 ± 0.01c 16.95 ± 0.83e 601.07 ± 2.75b

11.50 ± 0.01a 49.28 ± 0.54a 3.46 ± 0.06e 4.71 ± 0.02a 31.05 ± 0.47b 352.44 ± 0.29g

L*,2 a*,2 b*,2 C*,3 h*,3

60.27 ± 0.12g 2.66 ± 0.42a 55.84 ± 1.39a 63.00 ± 3.65a 1.52 ± 0.01b

75.84 ± 0.00e 1.50 ± 0.00bc 16.18 ± 0.71c 18.43 ± 0.71cd 1.48 ± 0.00e

76.40 ± 0.00d 1.30 ± 0.00cd 15.71 ± 0.00c 17.40 ± 0.00cde 1.49 ± 0.00d

78.62 ± 0.28b 1.00 ± 0.01d 14.14 ± 0.09d 15.13 ± 0.08e 1.50 ± 0.00c

77.74 ± 0.21c 1.06 ± 0.01d 14.25 ± 0.06d 15.36 ± 0.04de 1.50 ± 0.00c

75.80 ± 0.09e 1.67 ± 0.01b 16.65 ± 0.03c 19.44 ± 0.02c 1.47 ± 0.00f

72.54 ± 0.30f 2.61 ± 0.01a 21.31 ± 0.05b 28.09 ± 0.09b 1.45 ± 0.00g

84.12 ± 0.12a 0.21 ± 0.01e 9.29 ± 0.00e 9.33 ± 0.00f 1.55 ± 0.00a

NS TG DTG

20

80

80

40

L2

20

TG DTG

40

80

L4 TG DTG

0

Mass (%)

80

40

40

80

LC TG DTG

0 0

TG DTG

0

80

20

L5

20

100

40

TG DTG

60

100

60

L3

0 100

20

Mass (%)

60

100

60

TG DTG

20

Mass (%)

Mass (%)

0

Mass (%)

100

DTG (mg min-1)

100

60

L1

20 0

DTG (mg min-1)

Mass (%)

0

40

60 40

LS

20

TG DTG

0

100 200 300 400 500 600 700 800 900

0

Temperature (°C)

DTG (mg min-1)

40

60

DTG (mg min-1)

60

DTG (mg min-1)

80

DTG (mg min-1)

80

Mass (%)

100

DTG (mg min-1)

100

DTG (mg min-1)

Mass (%)

Defatted flours: L1 to L5 = samples degreased by compressed propane; LC = sample degreased by supercritical carbon dioxide (CO2) + ethanol (1,1, w/w) as cosolvent; LS = sample degreased by Soxhlet using petroleum ether. Lowercase letters indicate statistical difference according to Duncan's test (p < 0.05). 1 kcal/100 g. 2 CIE L*a*b* color system. 3 CIE L*C*h* color system.

100 200 300 400 500 600 700 800 900

Temperature (°C)

Fig. 1. Thermograms showing the mass loss versus the derivative thermogravimetric profile of raw sapucaia nut (NS), and the resulting PDSF samples.

574

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

foam stability (FS). Each sample was evaluated in duplicate. FC and FS were calculated according to the equations:

Table 2 Results of the thermogravimetric (TG) and derivative thermogravimetric (DTG) analyzes of the sapucaia nut (NS), and its partially defatted flours analyzed under a synthetic air atmosphere. Sample

NS

L1

L2

L3

L4

L5

LC

LS

TG

FC (%) =

V1 − V0 V0

(2)

FS (%) =

V2 − V1 V0

(3)

DTG

Stage of thermal degradation

Δm (%)

ΔT (°C)

Tp (°C)

1 Stability 2

3.51 – 77.75

30–113 113–179 179–480

3 1 Stability 2

16.23 5.49 – 66.83

480–690 30–117 117–148 148–485

3 1 Stability 2 3 1 Stability 2 3 1 Stability 2 3 1 Stability 2 3 1 Stability 2

22.30 5.73 – 64.22 25.38 5.72 – 64.05 23.27 7.28 – 61.88 26.03 6.17 – 64.99 23.04 4.25 – 71.39

485–714 30–118 118–150 150–482 482–797 30–117 117–155 155–481 481–707 30–117 117–158 158–474 474–798 30–117 117–159 159–483 483–734 30–114 114–160 160–492

3 1 Stability 2 3

19.70 9.98 – 48.67 29.18

492–701 30–100 100–158 158–414 414–687

66.17 endo – 374.68 endo 406.30 exo 559.50 endo 54.61 endo – 361.66 endo 409.22 exo 578.86 endo 55.05 endo – 358.54 endo 577.14 endo 51.55 endo – 354.99 endo 578.82 endo 27.91 endo – 353.55 endo 578.89 endo 60.03 endo – 362.92 endo 575.60 endo 64.47 endo – 365.93 endo 408.94 exo 574.32 endo 63.30 endo – 305.33 endo 528.00 endo

where V0 = volume before stirring (mL); V1 = volume after stirring (mL); V2 = volume after standing (mL). 2.3.5. Emulsifying activity and emulsion stability Oil-in-water emulsions were prepared by dispersing 25% of soybean oil in a solution containing 1% of the PDSF, followed by stirring (PT 3100 D Polytron homogenizer) at 15000 rpm per 5 min. Each emulsion was immediately analyzed after preparation. The emulsifying activity index (EAI) and emulsion stability index (ESI) of the PDSF were determined according to the method of Pearce and Kinsella (1978) with small modifications proposed by Zhang et al. (2014). An aliquot of 50 μL of each fresh emulsion was collected at 0 and 10 min after homogenization, to which 5 mL of 0.1% (w/v) sodium dodecyl sulfate (SDS) solution were added. The absorbance of these solutions was measured in a spectrophotometer at 500 nm immediately (A0) and 10 min (A10) after preparation of the emulsion, whose values were used to calculate the EAI and the ESI according to the following equations:

EAI (m2/g) = ESI (%) =

A10 × Δt ΔA

(4) (5)

where F is the volumetric fraction of oil (0.25); A0 is the absorbance at time 0, and A10 is 10 min after homogenization; Δt = 10 min; ΔA = A0 − A10. The analyzes were performed at 25 °C in triplicate. 2.3.6. Stability test by phase separation, microstructure and droplet size of emulsions Immediately after the preparation, three aliquots of 10 mL of each emulsion were also transferred to 15 mL graduated centrifuge tubes, and the separated phase (mL) at the bottom of the tube was measured at 0, 10, 20, 30, 60 and 120 min. Droplet size distribution (DSD) was investigated according to Teixeira et al. (2016) with minor modifications, using an inverted Zeiss Axio Observer D1 microscope (Zeiss Vision GmbH, Germany) with magnification of 80×. DSD was obtained with ImageJ software with the aid of the particle size measurement tool, in three micrographs for each sample, in order to measure at least 1000 droplets per sample.

Δm = mass loss; ΔT = temperature range; Tp = peak temperature. Defatted flours by compressed propane (L1 to L5), supercritical CO2 + ethanol as cosolvent (LC), or Soxhlet using petroleum ether (LS).

Brazil) at 500 nm. The obtained value was used as turbidity parameter. This method was adapted from Zhang et al. (2014). 2.3.3. Water or oil holding capacity Water holding capacity (WHC) or oil holding capacity (OHC) was determined using the method described by Ling et al. (2016) with minor modifications. Approximately 1.0 g sample (dry basis) of each flour (m1) was weighed into a 50-mL centrifuge tube of known weight (m2), and 20 mL of soybean oil or water were added. The resulting dispersion was then vortexed for 2 min and allowed to stand for 30 min at room temperature, followed by centrifugation (Excelsa II, Fanem, Brazil) at 3493 ×g for 10 min. The supernatant was gently discarded, and the sludge drained for 2 min before weighing the tube again (m3). The OHC and WHC were expressed as the amount of oil/water retained per gram of PDSF, according to the following equation:

m − m2 − m1 OHC/WHC (g/g of flour) = 3 m1

2 × 2.303 × A 0 F×C

2.4. Statistical and data analysis The results were submitted to the Duncan test at 95% confidence level using Statistica 10.0 software (StatSoft™, Inc.). Graphs were obtained using the Origin 8.6 software (Originlab Corporation). 3. Results and discussion 3.1. Proximate composition and color analysis Unavoidably, after removal of the lipid content from NS, the proportions of the other constituents in the PDSF naturally tend to increase. Due from the difference in the lipid content removed in each extraction condition, the samples L1–L5 showed significant variation in their physicochemical composition (Table 1). Although the compressed propane presents a great efficiency in the extraction of lipids from the sapucaia nuts (Teixeira et al., 2018), a significant amount of remaining fat ranging from 18.33 to 36.02% was still found. Moisture reached a maximum of 7.93%, while protein, carbohydrate, and ash contents

(1)

2.3.4. Foam properties A 25-mL of 1% (w/v) PDSF solution was homogenized with a PT 3100 D Polytron homogenizer (Kinematica AG, Switzerland) at 15000 rpm for 3 min. The foam and solution obtained were transferred to a 50-mL graduated test tube. The volume of the foam portion was measured at 0 min for foaming capacity (FC) and after 30 min rest for 575

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

Fig. 2. Sapucaia nut raw sample, and the resulting defatted flours sided by the corresponding scanning electron micrographs at magnification of 1700×. Bar represents 10 μm.

between 447.74 and 547.25 kcal/100 g. High-calorie value is commonly reported for other walnuts such as Brazil nut, almond, macadamia, pecan and hazelnut (Venkatachalam & Sathe, 2006). The proximate composition for the NS agrees with previous reports (de Carvalho et al., 2008; de Carvalho et al., 2012; Denadai et al., 2007; Vallilo et al., 1998), also presenting great similarity with the composition of Brazil nut (Yang, 2009). Table 1 also presents the parameters of the instrumental analysis of color. Except for the h* angle, all the color parameters presented a high correlation with the oil content of the sapucaia (Teixeira et al., 2017, 2018). Visually, it is possible to verify that the raw sample (NS) has much more yellowish color than the flours. The color parameter L*, referring to luminosity varied significantly (p < 0.05) between 60.27 and 84.12, indicating coloration with light shades; consequently, the correlation of L* with fat content was inversely proportional (r = −0.92). As shown in Table 1, the high lipid content of NS resulted in a lower L* value, while PDSF present higher luminosity values, highlighting LS sample, considered the “lightest” between them. Both NS and PDSF presented positive and different values for the parameter a* (0.21–2.66), indicative of reddish tones, and a high correlation (r = 0.94) with the lipid content. The parameter b* confirms the previous observations regarding the yellowish characteristic of the sapucaia nut, which presented b* value up to 6 times higher than the PDSF samples, and 81% correlation with the amount of fat in the samples. Similarly, the C* parameter, which represents the saturation, and the h* (hue angle) also showed significant variation among the samples. With an 84% correlation with the oil content, the values of C* indicate that the NS sample is the furthest from the light axis, thus the

Table 3 XRD and SEM results for sapucaia nut in natura (NS), and its defatted flours. Sample

NS L1 L2 L3 L4 L5 LC LS

XRD

SEM

Degree of relative crystallinity (%)

% Amorphous

Average size (μm)

62.10 54.00 52.70 49.10 50.20 52.25 58.70 32.80

37.90 46.00 47.30 50.90 49.80 47.75 41.30 67.20

33.56 28.66 32.63 25.44 25.07 23.17 20.11 24.54

± ± ± ± ± ± ± ±

2.81a 1.83ab 3.82a 4.15bc 0.54bc 1.88bc 1.50c 2.59bc

Lowercase letters indicate statistical difference according to Duncan's test (p < 0.05). Defatted flours by compressed propane (L1 to L5), supercritical CO2 + ethanol as co-solvent (LC), or Soxhlet using petroleum ether (LS).

were at least 31.89, 23.79, and 2.08%, respectively. Due to the high efficiency in the extraction of the lipid content using petroleum ether, the LS sample had the lowest residual fat (~3%), while moisture, protein, and ash contents were the highest between all PDSF. Compared to hexane extraction, a 49.50 ± 0.51% oil removal is reported (Teixeira et al., 2018), which implies in ~9% remaining fat, 3× higher than that obtained with petroleum ether. After the degreasing processes, the samples naturally reduced their caloric value. NS sample presented a high caloric value, considering its high-fat content, while the LS sample had the lowest caloric value among the PDSF samples. For L1–L5 samples the caloric values were 576

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

40

(a)

35 30

Diffraction intensity (cps)

much more differentiated from the other flours. Table 2 highlights these differences in more details. DTG curves revealed the existence of 4 peaks for mass loss (Δm) events, well-defined for samples NS, L1, L2, and LC, while the others presented only 3 peaks, but all of them comprised a total of 3 thermal events. The first event related to the loss of moisture and volatile compounds comprised a temperature range (ΔT) of 30–113 °C to the raw sample (NS), with < 4% of Δm, while the PDSF showed a lower ΔT, varying between 100 and 118 °C, but a higher mass loss (Δm between 4.25 and 9.98%). This event indicated a higher Δm for LS, which also revealed a moisture gain at the beginning of the analysis. A stability range was observed between 100 and 179 °C in the samples. The second event related to the thermal degradation of oil, carbohydrates and proteins in the samples occurring between 179 and 480 °C, with a peak at 374 and 406 °C, and a Δm of 77.75% for NS. In the PDSF samples L1–L5, the third event occurred between 474 and 798 °C, with a Δm ranging from 22.30–26.03%, whereas the LC and LS samples had ΔT between 414 and 701 °C, and Δm of 19.70, and 29.18%, respectively. This event is also attributed to the cleavage of the covalent bond between the peptide bonds of amino acid residues (Zhang et al., 2017). The third peak (3rd event) may be related to the typical decomposition of proteins and the oxidation of partially decomposed proteins (Kaspchak et al., 2017), or may even be due to the cleavage of the SeS, OeN and OeO bonds of the protein molecules in the samples (Zhang et al., 2017). As shown in Fig. 1, there is a similarity in the mass loss profile of the samples L1-L5, evidencing the different stability patterns other than the PDSF; the LS has a distinctly Δm smaller than the others, basically due to its lower lipid residual, making it less thermally stable. Thus, it is possible to infer that the remaining fat in the PDSF contributes to a higher thermal stability. PDSF showed resistance to thermal decomposition higher than that of soybean meal (Zhang et al., 2017), and similar to that of baru flour (Guimarães et al., 2012), and isolated quinoa protein (Kaspchak et al., 2017).

NS L1 L2 L3 L4 L5 LC LS

25 20 15 10 5 0 10

20

30

40

50

60

Diffraction angle (2 )

(b)

5 6

4

2 3

7

8

9

Absorbance (a.u.)

1

10 NS L1 L2 L3 L4 L5 LC LS

4000

(c)

3500

3000

2500 2000 1500 Wavenumber (cm-1)

1000

3.0 L1 L2 L3 L4 L5 LC LS

2.5 264-269

Absorbance (a.u)

500

2.0

3.3. Morphology of the sapucaia nut and its defatted flours Fig. 2 illustrates the appearance of NS, and the PDSF used in this research right after the oil removal, together with the scanning electron micrographs (SEM), which depicts the morphology of the raw sample, as well as the effects of the degreasing treatments on the microstructure of the PDSF. The average particle diameter of each sample is presented in Table 3. The SEM shows that NS is composed mainly of globular structures, possibly lipid and protein bodies incorporated in squamous tissues that form cell walls, very similar to the parenchymatic tissues exhibited by Brazil nut, as elucidated by Scussel, Manfio, Savi, and Moecke (2014). It also shows that thin-walled, irregular-shaped tissues make up most of the edible tissue of the nuts. Structures with irregular morphology may also be fibers, since sapucaia is a source of this carbohydrate (de Carvalho et al., 2008). The LC sample presented a less damaged parenchyma structure, because of the lower effectiveness in removing fat by CO2 + ethanol (Teixeira et al., 2018), while the other PDSF presented similar shapes. As seen in the micrographs (Fig. 2), there are spherical and other concave-convex structures, due to the high carbohydrate content present in the PDSF (Table 1). It is suggested that some of them may be composed of starch, whose granules have the abovementioned characteristic formats, like those reported by Ling et al. (2016) for defatted pistachio flour. Table 3 shows that the raw sample NS presented the higher particle diameters, which could be due to the oil soaked in the nut granules. There was a significant variation in the mean diameter of the PDSF particles, ranging from 20.11 μm to 33.56 μm. Sample LC showed the smallest particle size. Martínez, Ganesan, Pilosof, and Harte (2011) explain that pressure and temperature are the major factors, which

1.5 1.0 0.5 0.0 240

260

280

300

320

340

360

380

400

Wavelength (nm)

Fig. 3. (a) Diffractogram showing the X-ray profile of NS and the PDSF. (b) DRIFTS spectrum of NS and the PDSF in the region of 4000–500 cm−1. (c) UV–Vis spectrum of 1 mg/mL solutions of PDSF in the wavelength range between 220 and 400 nm.

“darkest”; while the LS stands out as the lightest, in agreement with the described characteristics for the L* parameter. The h* angle (1.45–1.55°) indicated that all the samples had a hue close to red.

3.2. Thermal stability Fig. 1 shows the results of thermogravimetric (TG) and derivative thermogravimetric (DTG) analyses, representing the loss of mass versus temperature for the raw sapucaia nut (NS) and each PDSF sample. The subcritical propane under different conditions has a negligible impact on the thermal stability of the PDSF, whose TG/DTG curves are very similar to each other, whereas the LS sample presents a TG/DTG profile 577

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

2.8

d c

2.7

100

14

FC FS

12

d

Foam capacity (%)

b

b

a 2.6

2.5

a a

10

b

b 60

b

8 6

b

c

c

c

c

c

c

40

4 20

d d

2 0.5

0 0

0.0

L1

L2

L3

L4

L5

LC

LS

L1

L2

L3

L4

Sample

a

OHC (g oil/g flour, d.b.)

1.4

c

e

1.0

b

0.8

c 0.6

1.4 1.2

b d

f

g

b

0.8

c

0.6

d e

0.4

0.4

0.2

0.2

0.0

0.0

L1

L2

L3

L4

L5

LC

(d) 20

1.6

Emulsifying activity index (m2/g)

a

OHC WHC

1.2

L5

LC

LS

Sample

WHC (g water/g flour, d.b.)

(b) 1.6

1.0

80

100

80

b

15

c

d

e

10

a

f

g

b

d

60

40

c

c

5

20

e

f 0

0

L1

LS

EAI ESI

a

Emulsion stability index (%)

Turbidity (A500nm)

(c)

e

Foam stability (%)

(a)

L2

L3

Sample

L4

L5

LC

LS

Sample

Fig. 4. (a) Turbidity of 1% PDSF solutions measured at 500 nm; (b) Oil (OHC) or Water Holding Capacity (WHC), (c) Foam capacity (FC), and foam stability (FS), (d) emulsifying activity index (EAI), and emulsion stability index (ESI) of PDSF. The error bar indicates the standard deviation (n = 3). Lowercase letters indicate statistical difference (p < 0.05).

hemicellulose may increase the percentage of the crystalline region, since these constituents make up a large part of the amorphous region (Qi et al., 2015).

influence the particle size in the degreasing processes since they can cause both size reduction due to the release of the fat globules and the swelling of the granules under extreme conditions (e.g. high pressure and temperature). Consequently, the particle diameter of these flours can also influence its functional properties such as foam and emulsification capacities (Zhang et al., 2014). The SEM study proves that the kinetic energy of the degreasing processes with sub- and supercritical fluids can break the cellular structure of the sapucaia, facilitating the release of the oil (Pereira, Hamerski, Andrade, Scheer, & Corazza, 2017; Teixeira et al., 2018), thereby reducing the residual lipid content. Fig. 3a shows the X-ray diffractograms of the PDSF compared to the NS, while the degree of relative crystallinity is shown in Table 3. The difference in the intensity of the peaks is noticeable. Both NS and PDSF samples showed two strong characteristic peaks close to 9° and 19°, such as the profile of isolated soybean proteins (Zhang et al., 2014) and linseed (Kaushik et al., 2016). The lack of crystallinity or ordered arrangement in the protein structure of the extracted material may be revealed by the low peak intensity at 19° (Kaushik et al., 2016), which is confirmed in Table 3, where LS is the sample with the lowest degree of crystallinity, proving that the treatment with petroleum ether has a great effect on the amorphous nature of the PDSF. The L4 sample showed the greatest reduction in the peak intensity at 19°, suggesting that the crystalline structure of PDSF also collapsed after treatment with compressed propane. The extent of the structural damage became stronger when higher pressure and temperature were applied in the degreasing process (L3 and L4 samples). The degree of crystallinity of the samples was found to be highly correlated (r = 0.93) with the fat content. Crystallinity values of PDSF were smaller than the NS sample, proving that the percentage of oil has a direct impact on this characteristic (Table 3). The lower the level of residual fat, the lower the degree of relative crystallinity of PDSF. For samples with a low degree of crystallinity, removal of starch and

3.4. DRIFTS spectra Fig. 3b depicts the DRIFTS absorbance spectrum of the NS and PDSF samples evaluated in the region of 4000 to 500 cm−1. It was verified that the NS and LC samples, which have the highest lipid content, also present similar IR patterns, showing low intensity in the absorption of the IR spectrum, with little definition of the peaks referring to the functional groups. On the other hand, all the other samples, which achieved a higher percentage of fat removal, exhibited well-defined peaks, with similar IR spectrum pattern among them. The main absorption peaks between 3700 and 3000 cm−1 (band 1) are attributed to the stretching vibration of NeH and free hydroxyl (Chen et al., 2013; Zhong et al., 2017) and may be associated with residual moisture in the samples (Belchior, Franca, & Oliveira, 2016; Zhang et al., 2014). The band at 2933 cm−1 (band 2) corresponds to the vibration of asymmetrical stretching of CeH2 (Chen et al., 2013). Band 3 at 2854 cm−1 is attributed to the elongation of the CeH bonds of the methyl group (–CH3) of fatty acids (Belchior et al., 2016; Naumann, 2000). In bands 4 (1749 cm−1) and 5 (1672 cm−1) there are the functional groups C]O of esters, and β-pleated sheets of proteins, respectively. Bands were also recorded in 1670 cm−1 (amide I, CeO stretch). Band 6 (1548 cm−1, NeH bond) is the amide II band, while band 7 (1458 cm−1) is attributed to CeH vibrations of glycerol (Naumann, 2000). Absorption bands near 1402 cm−1 and 1240 cm−1 are attributed to the NeH bond (amide III, CeN stretch) (Zheng et al., 2017) and subsequent bands (below 1100 cm−1) are the fingerprints of the methyl esters of long chain fatty acids present in the lipid fraction of the 578

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

(a)

(b) Frequency (%)

100

L1

80 60 40

a

f

a

20

ab

b

Frequency (%)

100

L2

80 60

e

b

40

b

20

bc

de

Frequency (%)

100 80

L3

60 40

e

20

100

Frequency (%)

b

e

d

L4

80 60 40

f

20

e

d

100

Frequency (%)

f

a

ef

L5

80 60

d

40

c

20

c

cd

cd

Frequency (%)

100

LC

80 60 40

a

f e

20

a

d

Frequency (%)

100 80

LS

c

60 40

d

20

c

c

c

0

0.1-15.0

15.1-30.0 30.1-45.0 45.1-60.0

>60.0

Diameter ( m)

(c) Separated volume (mL)

6 5 4 3

L1 L2 L3 L4 L5 LC LS

2 1 0

0

20

40

60

80

100

120

Time (min) Fig. 5. (a) Micrographs of emulsions prepared with soybean oil and 1% PDSF, under magnification of 80× (bar represents 200 μm); (b) droplet size distribution (DSD) of the emulsions; (c) phase separation test of PDSF emulsions (dotting is only to guide the eyes). Error bar is the SD (n = 3). Lowercase letters indicate statistical difference (p < 0.05).

579

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

Table 4 Phase separation (mL) of emulsions prepared with soybean oil and 1% (w/v) PDSF solution (1,3, v/v), homogenized at 15,000 rpm for 5 min. Sample

Time (min.) 0

L1 L2 L3 L4 L5 LC LS

ns⁎ ns ns ns ns ns ns

10 1.83 4.03 3.50 3.10 1.57 4.17 1.57

20 ± ± ± ± ± ± ±

c

0.15 0.06a 0.00b 0.36b 0.12c 0.58a 0.06c

2.67 4.73 4.43 3.50 2.33 5.00 2.17

30 ± ± ± ± ± ± ±

c

0.29 0.32a 0.12a 0.50b 0.12c 0.50a 0.29c

2.73 4.80 4.53 3.80 2.47 5.17 2.40

60 ± ± ± ± ± ± ±

d

0.23 0.26ab 0.06b 0.44b 0.06d 0.31a 0.52d

3.37 5.23 5.07 4.00 3.00 5.43 3.03

120 ± ± ± ± ± ± ±

c

0.21 0.12a 0.06a 0.44b 0.00c 0.51a 0.42c

3.70 5.57 5.27 4.33 3.23 5.47 3.17

± ± ± ± ± ± ±

0.26c 0.40a 0.12a 0.32b 0.06c 0.57a 0.29c

⁎ ns = no separation; lowercase letters indicate statistical difference according to Duncan's test (p < 0.05). Defatted flours by compressed propane (L1 to L5), supercritical CO2 + ethanol as co-solvent (LC), or Soxhlet using petroleum ether (LS).

3.5.2. Turbidity of PDSF solutions Fig. 4a shows the turbidity of the PDSF suspensions and reveals that there is a difference (p < 0.05) between the samples, highlighting LS, L3 and L5, which exhibited higher turbidity than the others. Hua, Cui, Wang, Mine, and Poysa (2005) explain that the presence of residual lipids in the defatted meal causes problems in the functional properties of proteins (e.g. emulsification power), and turbidity is a way of verifying this defect. It may be one of the factors, due to which there were significant differences in the turbidity of solutions containing PDSF. Pigments and other compounds responsible for the yellowish color of the sapucaia nut can also be another impacting factor in the variation of this parameter.

samples (Zhang et al., 2013). The region 8 (bands between 1310 and 1240 cm−1) is related to the components of the amide III band of proteins, while the region 9 (1200–900 cm−1) has polysaccharide ring vibrations with C–O–C, CeO, C–O–P, P–O–P. Region 10, in turn, is associated with the sample's fingerprint region (Naumann, 2000). With the removal of oil from the samples, the intensity of the bands changed, indicating that the number of functional groups is affected by the applied degreasing processes. These differences between FT-IR spectra of NS and PDSF confirm that the extraction method caused changes in their chemical profiles. The results also suggest that the composition of the defatted flour can affect the exact position of the bands, besides influencing the changes in the IR spectrum. The FT-IR analysis confirms that the LS sample suffered the greatest structural damage, once its spectrum showed the greatest variation in the IR absorption bands. FTIR profiles similar to NS and PDSF were reported for defatted soybean meal (Zheng et al., 2017), defatted rice bran (Qi et al., 2015), rapeseed meal (Shi et al., 2016), and biofilms of peanut protein isolate (Zhong et al., 2017).

3.5.3. Oil or water holding capacities of PDSF Some of the main consequences of retention of oil/water on the flour are the reduction in moisture and fat loss, as well as the improvement in taste and mouthfeel of food products, due to the flour's ability to bind to the water/oil (Ling et al., 2016). The oil holding capacity (OHC) and water holding capacity (WHC) of the PDSF ranged from 0.72 to 1.57, and from 0.35 to 1.38, respectively (Fig. 4b). A larger extent of absorption of both water and oil was observed in the LS sample, because this sample had the lowest residual fat content, allowing the exposure of a large number of hydrophilic and hydrophobic parts bound to more water and oil. The lowest values were recorded for the LC sample, whose residual fat was the highest among the studied samples. The same effect was observed for pistachio flours (Ling et al., 2016) and canola (Khattab & Arntfield, 2009), where samples exposed to high temperatures showed higher OHC and WHC, probably due to heat diffusion and denaturation of proteins, which expose additional available binding positions for water and oil.

3.5. Functional properties of PDSF 3.5.1. Ultraviolet spectra Most proteins exhibit characteristic ultraviolet (UV) absorption at 280 nm due to the presence of aromatic amino acids (Moore, DeVries, Lipp, Griffiths, & Abernethy, 2010). Considering the high protein content in PDSF, which participate in the formation of emulsions by their amphiphilic character, a scanning at wavelengths between 190 and 400 nm (Fig. 3c) was carried out to authenticate the presence of these compounds in the aqueous solution used in the production of the emulsions. The maximum absorption peak was observed at 264 nm (L2 and L5), 265 nm (L3, L4, LC, and LS) and 269 nm (L1), and the intensity showed a significant difference between all treatments applied; the sample defatted by Soxhlet being the most affected (lower intensity). The LC and LS samples showed absorption intensity of 1.059 and 0.732, respectively. The absorption intensities of the samples treated by compressed propane were 1.308 (L1), 1.148 (L2), 0.884 (L3), 0.988 (L4), and 0.950 (L5), with standard deviations ≤0.001. Despite the significant difference in the UV absorption intensity of the samples, the ANOVA revealed that the influence of the P and T variables on this parameter is not significant (p > 0.05), but samples treated at lower pressures tend to show higher absorption intensity. Soybean protein treated with subcritical water presents significant changes in its UV profiles, caused mainly by the increase in temperature, which can cause conformational changes in protein molecules; this has a direct impact when using these proteins to produce emulsions. Additionally, the increase in absorption intensity may also be caused by the gradual exposure of the hydrophobic groups due to the high-pressure treatment (Zhang et al., 2014).

3.5.4. Foam properties of PDSF Like the water/oil absorption capacities, the foaming capacity (FC) and foam stability (FS) also showed a significant difference (p < 0.05) among some of the PDSF samples evaluated. In Fig. 4c it can be verified that the solution containing 1% of PDSF produces little foam, reaching a maximum value of 12% (LS), with stability varying between 53 and 83%, the LC sample being the less stable. These low FC and FS values may be due to both the residual fat content and the high protein-protein interaction, leading to the formation of aggregates, harmful to the formation of foam and the decrease of the solubility of the nitrogen due to the thermal denaturation caused by the applied processes (Ling et al., 2016). Protein isolates from cashew nuts had higher FC (45%), but lower FS (55%) than the PDSF (Ogunwolu, Henshaw, Mock, Santros, & Awonorin, 2009); defatted flour of hazelnut, in turn, has an FC of 400%, with FS of 45% (Turan et al., 2015). Factors such as pH and the type of protein present in the flour used in the production of the solution also influence the FC and FS, because if the protein presents low solubility at an isoelectric point, only the soluble protein fractions will be involved in the foaming, and since the concentration of these soluble 580

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

emulsions. It is suggested the isolation of sapucaia proteins aiming to evaluate their technological characteristics, since the isolated proteins, in general, have better functional properties, as well as improved emulsifying power due to their amphiphilic capacity.

fractions is very low, the amount of foaming will be small (Ogunwolu et al., 2009). 3.5.5. Emulsifying properties of PDSF Emulsification ability and stability are important functional properties of food proteins. Factors such as protein type, concentration, pH, ionic strength, and viscosity of the system affect both emulsification ability and stability. When it comes to emulsions of the protein-fatwater type, several chemical and physical factors are involved in the formation, stability and texture properties (Khattab & Arntfield, 2009). The effect of the sub- and supercritical fluid and petroleum ether degreasing processes on the emulsifying properties of PDSF is shown in Fig. 4d, represented by the emulsifying activity index (EAI) and emulsion stability index (ESI). According to the results obtained, the applied treatments had a significant impact on the evaluated characteristics. In addition to the protein content of PDSF, the emulsifying properties may also be affected by the combination between protein and polysaccharide (Chen et al., 2016). The remaining fat in each PDSF sample had also influence in the emulsion properties. Additionally, some lipid components such as monoacylglycerols and diacylglycerols have surfactant properties helping the emulsion formation (McClements & Weiss, 2005). The spectrophotometric method used showed that L1 flour had higher EAI, reaching 20 m2/g, but low stability (7%), while the LS sample would be the most stable, reaching about 56% ESI, indicating that the extraction with petroleum ether provides a flour with emulsifying power more durable than the others. The other PDSF had ESI between 7 and 24%, representing a variation between 2 and 8 times the stability obtained by the LS sample.

4. Conclusions In general terms, partially defatted sapucaia nut flour (PDSF) is a good source of protein and carbohydrates. The results indicate that the processes of degreasing with compressed propane under the studied conditions, little affect the morphology of the PDSF granules, and the physicochemical characteristics of the PDSF, causing a minimum impact on the nutritional profile of the samples. On the other hand, the thermogravimetric analyses revealed that there is an impact on the stability properties of the sapucaia flours to the thermal degradation. Samples L1 (propane extracted) and LS (petroleum ether extracted) stood out for having the greatest emulsifying activity. The PDSF presented other functional properties such as foam capacity, as well as high water and oil absorption capacities. PDSF was effective in the production of emulsions whose structure showed droplets of micrometric size, with alleged stability. Due to the high protein and residual lipid content associated with unsaturated fatty acids and bioactive compounds already reported in the sapucaia nut, PDSF can be used as a new potential ingredient in bakery and confectionery products in order to increase their nutritional value and functional properties. Acknowledgements This work was supported by the Coordination for the Improvement of Higher Education Personnel – CAPES (Brazil), grant n. 1291783 (CAPES-DS) and n. 88881.135997/2016-01 (PDSE).

3.5.6. Microstructure, droplet size distribution and phase separation of PDSF emulsions The micrographs of the oil-in-water emulsions produced with 1% solution of PDSF and soybean oil are shown in Fig. 5a, while the droplet size distribution (DSD) of each sample is shown in the graphs of Fig. 5b. Microscopic analysis confirmed that the PDSF can effectively produce emulsions with micrometric-sized droplets, whose diameter was influenced by the different flours used. This information is reinforced by the DSD frequency plot (Fig. 5b), which also reveals statistical difference (p < 0.05) between the DSD of all samples in the diameter ranges evaluated (0.1 to > 60.0 μm). Here we highlight the emulsions produced with samples L3, L4, L5 and LS, obtained between 60 and 85% of their DSD with droplets smaller than 15.0 μm; emulsions presenting lower droplet size tend to be more stable (Teixeira et al., 2016). On the other hand, one of the least stable, the LC sample presented droplets in larger sizes and almost 40% of its DSD was > 60.0 μm. Excluding LC, all the other emulsions had > 71% of their DSD between 0.1 and 30.0 μm. To compare the colorimetric stability results of the emulsions, a phase separation stability test was performed during 120 min. Table 4 shows the phase separation data of the emulsions prepared with soybean oil and 1% PDSF solution, while Fig. 5c graphically exemplifies the tendency for stability in the quasi-linear separation process after 30 min of rest. As mentioned, lipid components can impact the emulsion formation, being able to alter the droplet size during homogenization, thus affecting the emulsion stability (McClements & Weiss, 2005). Fig. 5c shows that after 120 min rest, the samples LS, L1, and L5 presented the smallest volume separated, with no difference (p < 0.05) between them, highlighting those samples as the most stables. On the other hand, the samples L2, L3, and LC were also less stable, once the result of the phase separation was almost double the others. Although the results found in this test for some samples are different from the spectrophotometric test, this is a simpler way to verify the stability of emulsions, representing an alternative to the former. Despite the rapid phase separation reported herein, it is important to note that the concentration of the PDSF was only 1%, so it is believed that higher proportions of PDSF may produce much more stable

Conflicts of interest The authors declare no conflict of interest. References AOCS (1997). Official Methods and Recommended Practices of the American Oil Chemists’ Society (5th ed.). Champaign, Illinois: AOCS Press. Belchior, V., Franca, A. S., & Oliveira, L. S. (2016). Potential of Diffuse Reflectance Infrared Fourier Transform Spectroscopy and Chemometrics for Coffee Quality Evaluation. ETP International Journal of Food Engineering, 2(1), 1–8. http://dx.doi. org/10.18178/ijfe.2.1.1-8. Bhise, S., Kaur, A., & Aggarwal, P. (2015). Development of protein enriched noodles using texturized defatted meal from sunflower, flaxseed and soybean. Journal of Food Science and Technology, 52(9), 5882–5889. http://dx.doi.org/10.1007/s13197-0141630-1. Carioca, A. C., & Ferreira, C. M. (2011). Application of diffuse reflectance spectroscopy in the quantification of the constituents of bauxite and iron ore. Mineraçao Mining, 64(2), 199–204. http://dx.doi.org/10.1590/S0370-44672011000200011. Carvalho, R. F., Huguenin, G. V. B., Luiz, R. R., Moreira, A. S. B., Oliveira, G. M. M., & Rosa, G. (2015). Intake of partially defatted Brazil nut flour reduces serum cholesterol in hypercholesterolemic patients—a randomized controlled trial. Nutrition Journal, 14(1), 59. http://dx.doi.org/10.1186/s12937-015-0036-x. Chan, K. W., Khong, N. M. H., Iqbal, S., Mansor, S. M., & Ismail, M. (2013). Defatted kenaf seed meal (DKSM): Prospective edible flour from agricultural waste with high antioxidant activity. LWT - Food Science and Technology, 53(1), 308–313. http://dx.doi. org/10.1016/j.lwt.2013.01.003. Chen, H., Qiu, S., Gan, J., Liu, Y., Zhu, Q., & Yin, L. (2016). New insights into the functionality of protein to the emulsifying properties of sugar beet pectin. Food Hydrocolloids, 57, 262–270. http://dx.doi.org/10.1016/j.foodhyd.2016.02.005. Chen, X., Ru, Y., Chen, F., Wang, X., Zhao, X., & Ao, Q. (2013). FTIR spectroscopic characterization of soy proteins obtained through AOT reverse micelles. Food Hydrocolloids, 31(2), 435–437. http://dx.doi.org/10.1016/j.foodhyd.2012.11.017. Chen, H., Zhao, M., Lin, L., Wang, J., Sun-Waterhouse, D., Dong, Y., ... Su, G. (2015). Identification of antioxidative peptides from defatted walnut meal hydrolysate with potential for improving learning and memory. Food Research International, 78, 216–223. http://dx.doi.org/10.1016/j.foodres.2015.10.008. Costa, T., & Jorge, N. (2012). Characterization and fatty acids profile of the oils from Amazon nuts and walnuts. Nutrition & Food Science, 42(4), 279–287. http://dx.doi. org/10.1108/00346651211248647. Das, M., Uppal, H. S., Singh, R., Beri, S., Mohan, K. S., Gupta, V. C., & Adholeya, A. (2011). Co-composting of physic nut (Jatropha curcas) deoiled cake with rice straw

581

Food Research International 109 (2018) 572–582

G.L. Teixeira et al.

Fluids, 128, 338–348. http://dx.doi.org/10.1016/j.supflu.2017.03.021. Qi, J., Yokoyama, W., Masamba, K. G., Majeed, H., Zhong, F., & Li, Y. (2015). Structural and physico-chemical properties of insoluble rice bran fiber: effect of acid–base induced modifications. RSC Advances, 5(97), 79915–79923. http://dx.doi.org/10. 1039/C5RA15408A. Qiu, S., Yadav, M. P., & Yin, L. (2017). Characterization and functionalities study of hemicellulose and cellulose components isolated from sorghum bran, bagasse and biomass. Food Chemistry, 230, 225–233. http://dx.doi.org/10.1016/j.foodchem. 2017.03.028. Sawada, M. M., Venâncio, L. L., Toda, T. A., & Rodrigues, C. E. C. (2014). Effects of different alcoholic extraction conditions on soybean oil yield, fatty acid composition and protein solubility of defatted meal. Food Research International, 62, 662–670. http://dx.doi.org/10.1016/j.foodres.2014.04.039. Scussel, V. M., Manfio, D., Savi, G. D., & Moecke, E. H. S. (2014). Stereoscopy and Scanning Electron Microscopy of Brazil Nut (Bertholletia excelsa H.B.K.) Shell, Brown Skin, and Edible Part: Part One-Healthy Nut. Journal of Food Science, 79(7), http://dx. doi.org/10.1111/1750-3841.12502. Shi, C., He, J., Yu, J., Yu, B., Mao, X., Zheng, P., & Chen, D. (2016). Physicochemical properties analysis and secretome of Aspergillus Niger in fermented rapeseed meal. PLoS One, 11(4), http://dx.doi.org/10.1371/journal.pone.0153230. Suprayudi, M. A., Alim, S., Fauzi, I. A., Ekasari, J., Setiawati, M., Junior, M. Z., & Tacon, A. G. J. (2016). Evaluation of hydrolysed rubber seed meal as a dietary protein source for Nile tilapia Oreochromis niloticus L. Aquaculture Research, 3801–3808. http://dx. doi.org/10.1111/are.13206. Teixeira, G. L., Ávila, S., Silveira, J. L. M., Ribani, M., & Ribani, R. H. (2017). Chemical, thermal and rheological properties and stability of sapucaia (Lecythis pisonis) nut oils: A potential source of vegetable oil in industry. Journal of Thermal Analysis and Calorimetry, 1–17. http://dx.doi.org/10.1007/s10973-017-6742-1. Teixeira, G. L., Ghazani, S. M., Corazza, M. L., Marangoni, A. G., & Ribani, R. H. (2018). Assessment of subcritical propane, supercritical CO2and Soxhlet extraction of oil from sapucaia (Lecythis pisonis) nuts. Journal of Supercritical Fluids, 133(October 2017), 122–132. http://dx.doi.org/10.1016/j.supflu.2017.10.003. Teixeira, G. L., Züge, L. C. B., Silveira, J. L. M., Scheer, A. D. P., & Ribani, R. H. (2016). The impact of polyoxyethylene sorbitan surfactants in the microstructure and rheological behaviour of emulsions made with melted fat from Cupuassu (Theobroma grandiflorum). Journal of Surfactants and Detergents, 19(4), 725–738. http://dx.doi. org/10.1007/s11743-016-1820-0. Turan, D., Capanoglu, E., & Altay, F. (2015). Investigating the effect of roasting on functional properties of defatted hazelnut flour by response surface methodology (RSM). LWT - Food Science and Technology, 63(1), 758–765. http://dx.doi.org/10. 1016/j.lwt.2015.03.061. Vallilo, M. I., Tavares, M., Aued-Pimentel, S., Campos, N. C., & Neto, J. M. M. (1999). Lecythis pisonis Camb. nuts: oil characterization, fatty acids and minerals. Food Chemistry, 66(2), 197–200. http://dx.doi.org/10.1016/S0308-8146(99)00040-0. Vallilo, M. I., Tavares, M., Pimentel, S. A., Badolato, E. S. G., & Inomata, E. I. (1998). Caracterização química parcial das sementes de Lecythis pisonis. Acta Amazonica, 28(2), 131–140. Venkatachalam, M., & Sathe, S. K. (2006). Chemical composition of selected edible nut seeds. Journal of Agricultural and Food Chemistry, 54(13), 4705–4714. http://dx.doi. org/10.1021/jf0606959. Xu, L., Lui, F., Luo, H., & Diosady, L. L. (2003). Production of protein isolates from yellow mustard meals by membrane processes. Food Research International, 36(8), 849–856. http://dx.doi.org/10.1016/S0963-9969(03)00082-6. Yang, J. (2009). Brazil nuts and associated health benefits: A review. LWT - Food Science and Technology, 42(10), 1573–1580. http://dx.doi.org/10.1016/j.lwt.2009.05.019. Zhang, X., Li, L., Xie, H., Liang, Z., Su, J., Liu, G., & Li, B. (2013). Comparative analysis of thermal behavior, isothermal crystallization kinetics and polymorphism of palm oil fractions. Molecules, 18(1), 1036–1052. http://dx.doi.org/10.3390/ molecules18011036. Zhang, Q., Tu, Z., Wang, H., Huang, X.-Q., Fan, L.-L., Bao, Z.-Y., & Xiao, H. (2014). Functional properties and structure changes of soybean protein isolate after subcritical water treatment. Journal of Food Science and Technology, 52(June), 3412–3421. http://dx.doi.org/10.1007/s13197-014-1392-9. Zhang, J., Ying, D., Wei, Y., Zhang, B., Su, X., & Li, S. (2017). Thermal transition and decomposition properties of pH- and phosphate-induced defatted soybean meals. Journal of Thermal Analysis and Calorimetry, 128(2), 699–706. http://dx.doi.org/10. 1007/s10973-016-5991-8. Zheng, P., Li, Y., Li, F., Ou, Y., Lin, Q., & Chen, N. (2017). Development of defatted soy flour-based adhesives by acid hydrolysis of carbohydrates. Polymer, 9(5), http://dx. doi.org/10.3390/polym9050153. Zhong, T., Liang, Y., Jiang, S., Yang, L., Shi, Y., Guo, S., & Zhang, C. (2017). Physical, antioxidant and antimicrobial properties of modified peanut protein isolate based films incorporating thymol. RSC Advances, 7(66), 41610–41618. http://dx.doi.org/ 10.1039/c7ra07444a.

and different animal dung. Bioresource Technology, 102(11), 6541–6546. http://dx. doi.org/10.1016/j.biortech.2011.03.058. de Carvalho, M. G., da Costa, J. M. C., de Souza, V. A. B., & Maia, G. A. (2008). Avaliação dos parâmetros físicos e nutricionais de amêndoas de sapucaia and castanha-dogurguéia. Revista Ciências Agronômicas, 39(4), 517–523. de Carvalho, I. M. M., Queirós, L. D., Brito, L. F., Santos, F. A., Bandeira, A. V. M., de Souza, A. L., & de Queiroz, J. H. (2012). Caracterização química da castanha de sapucaia (Lecythis pisonis Cambess.) da região da zona da mata Mineira. Bioscience Journal, 28(6), 971–977. Demoliner, F., de Britto Policarpi, P., Ramos, J. C., Bascuñan, V. L. A. F., Ferrari, R. A., Jachmanián, I., ... Block, J. M. (2018). Sapucaia nut (Lecythis pisonis Cambess) and its by-products: A promising and underutilized source of bioactive compounds. Part I: Nutritional composition and lipid profile. Food Research International, 108(October 2017), 27–34. http://dx.doi.org/10.1016/j.foodres.2018.03.028. Denadai, S. M. S., Hiane, P. A., Marangoni, S., Baldasso, P. A., Miguel, A. M. R. D. O., & Macedo, M. L. R. (2007). In vitro digestibility of globulins from sapucaia (Lecythis pisonis Camb.) nuts by mammalian digestive proteinases. Ciência e Tecnologia de Alimentos, 27(3), 535–543. http://dx.doi.org/10.1590/S0101-20612007000300018. Ganorkar, P. M., Patel, J. M., Shah, V., & Rangrej, V. V. (2016). Defatted flaxseed meal incorporated corn-rice flour blend based extruded product by response surface methodology. Journal of Food Science and Technology, 53(4), 1867–1877. http://dx. doi.org/10.1007/s13197-015-2134-3. Guimarães, R. D. C. A., Favaro, S. P., de Souza, A. D. V., Soares, C. M., Nunes, Â. A., de Oliveira, L. C. S., & Honer, M. R. (2012). Thermal properties of defatted meal, concentrate, and protein isolate of baru nuts (Dipteryx alata Vog.). Food Science and Technology (Campinas), 32(1), 52–55. http://dx.doi.org/10.1590/S010120612012005000031. Hua, Y., Cui, S. W., Wang, Q., Mine, Y., & Poysa, V. (2005). Heat induced gelling properties of soy protein isolates prepared from different defatted soybean flours. Food Research International, 38(4), 377–385. http://dx.doi.org/10.1016/j.foodres.2004.10. 006. Kaspchak, E., de Oliveira, M. A. S., Simas, F. F., Franco, C. R. C., Silveira, J. L. M., Mafra, M. R., & Igarashi-Mafra, L. (2017). Determination of heat-set gelation capacity of a quinoa protein isolate (Chenopodium quinoa) by dynamic oscillatory rheological analysis. Food Chemistry, 232, 263–271. http://dx.doi.org/10.1016/j.foodchem. 2017.04.014. Kaushik, P., Dowling, K., McKnight, S., Barrow, C. J., Wang, B., & Adhikari, B. (2016). Preparation, characterization and functional properties of flax seed protein isolate. Food Chemistry, 197, 212–220. http://dx.doi.org/10.1016/j.foodchem.2015.09.106. Khattab, R. Y., & Arntfield, S. D. (2009). Functional properties of raw and processed canola meal. LWT - Food Science and Technology, 42(6), 1119–1124. http://dx.doi. org/10.1016/j.lwt.2009.02.009. Ling, B., Zhang, B., Li, R., & Wang, S. (2016). Nutritional Quality, Functional Properties, Bioactivity, and Microstructure of Defatted Pistachio Kernel Flour. JAOCS, Journal of the American Oil Chemists’ Society, 93(5), 689–6991. http://dx.doi.org/10.1007/ s11746-016-2813-x. Martínez, K. D., Ganesan, V., Pilosof, A. M. R., & Harte, F. M. (2011). Effect of dynamic high-pressure treatment on the interfacial and foaming properties of soy protein isolate-hydroxypropylmethylcelluloses systems. Food Hydrocolloids, 25(6), 1640–1645. http://dx.doi.org/10.1016/j.foodhyd.2011.02.013. McClements, D. J., & Weiss, J. (2005). Lipid emulsions. In F. Shahidi (Ed.). Bailey’s industrial oil and fat products (pp. 457–502). (6th ed.). John Wiley & Sons, Inc.. http:// dx.doi.org/10.1002/047167849X.bio019. Moore, J. C., DeVries, J. W., Lipp, M., Griffiths, J. C., & Abernethy, D. R. (2010). Total protein methods and their potential utility to reduce the risk of food protein adulteration. Comprehensive Reviews in Food Science and Food Safety, 9(4), 330–357. http://dx.doi.org/10.1111/j.1541-4337.2010.00114.x. Naozuka, J., Vieira, E. C., Nascimento, A. N., & Oliveira, P. V. (2011). Elemental analysis of nuts and seeds by axially viewed ICP OES. Food Chemistry, 124(4), 1667–1672. http://dx.doi.org/10.1016/j.foodchem.2010.07.051. Naumann, D. (2000). Infrared spectroscopy in microbiology. Encyclopedia of Analytical Chemistry, 102–131. http://dx.doi.org/10.1002/9780470027318.a0117. Ogunwolu, S. O., Henshaw, F. O., Mock, H. P., Santros, A., & Awonorin, S. O. (2009). Functional properties of protein concentrates and isolates produced from cashew (Anacardium occidentale L.) nut. Food Chemistry, 115(3), 852–858. http://dx.doi. org/10.1016/j.foodchem.2009.01.011. Pardo-Giménez, A., Catalán, L., Carrasco, J., Álvarez-Ortí, M., Zied, D., & Pardo, J. (2016). Effect of supplementing crop substrate with defatted pistachio meal on Agaricus bisporus and Pleurotus ostreatus production. Journal of the Science of Food and Agriculture, 3838–3845. (December 2015) https://doi.org/10.1002/jsfa.7579. Pearce, K. N., & Kinsella, J. E. (1978). Emulsifying properties of proteins—Evaluation of a turbidimetric technique. Journal of Agricultural and Food Chemistry, 26(3), 716. Pereira, M. G., Hamerski, F., Andrade, E. F., Scheer, A.d. P., & Corazza, M. L. (2017). Assessment of subcritical propane, ultrasound-assisted and Soxhlet extraction of oil from sweet passion fruit (Passiflora alata Curtis) seeds. The Journal of Supercritical

582