Physical, chemical and nutritional characteristics of puffed quinoa

Abstract

Puffed quinoa can be used as ready-to-eat breakfast food or as an ingredient in snack formulations. In this study, puffed quinoa products with and without starch–chitosan coating were developed by gun, extrusion and microwave puffing at different process conditions (pressure, power, moisture content and energy consumption). Size, bulk density, colour, expansion index, water absorption and solubility, microstructure, mechanical and thermal properties, chemical composition and in vitro digestibility of organic matter and proteins of popped quinoa were assessed. Optimal process conditions for gun puffing were maximum 1.31 MPa after 780 s, 500 r.p.m. and 180 s for extrusion puffing and 1200 W for 60 s applying microwave puffing at 18–20% moisture contents. Gun and extrusion puffing yielded high-quality popped quinoa with a biological availability of organic matter between 84–88% and 79–90% for proteins. Extrusion and gun puffing are the most promising processes to prepare quinoa snacks.

Graphical Abstract

Coating, extrusion puffing, gun puffing, microwave puffing, quinoa, snack

Introduction

Quinoa (Chenopodium quinoa Willd) has been cultivated for centuries in the Andes region of South America. Nowadays, the cultivation of quinoa has been extended to other countries, such as Australia, Canada, China, United Kingdom, amongst others, followed by commercialisation without any value-added processing. This pseudo-cereal is appreciated because of its nutritional value and bioavailability in regard to proteins, carbohydrates, minerals, fatty acids and tocopherols (Pereira et al., 2019).

Gun puffing consists of the sudden release of accumulated superheated vapour by changing elevated internal pressure to atmospheric pressure that results in an expanded grain matrix with porous structure. Hoke et al. (2007) studied the puffing operation of barley reporting quality product formation at operational temperature and pressure of 550 °C and 0.9 – 1.0 MPa, respectively, using 16.5% feed moisture. Pressure was inversely related to product humidity and bulk density (Lee et al., 2019). In addition, puffing exhibits incremental energy consumption being critical post-operational hygiene.

Extrusion is a versatile technology that subjects raw material to mechanical energy, pressure and heat as it is passed through a screw enclosed in a barrel and finally forced through a die to give shape to the end product (Ganjyal & Hanna, 2002). Extrusion of quinoa using 250–500 kJ kg⁻¹ of mechanical energy results in an extrudate with the following characteristics: expansion ratio ranging between 1.17 and 1.55 m m⁻¹, unit density from 0.45 to 1.02 g cm⁻³, water absorption index from 2.33 to 3.05 g g⁻¹ and water solubility index from 14.5 to 15.9% (Kowalski et al., 2016). However, the expansion of quinoa was rather low compared to corn and wheat.

Microwave puffing consists of the sudden application of heat at atmospheric pressure so that water is vaporised inside the grain, increasing internal pressure. Swarnakar et al. (2014) studied rice popping by microwave heating giving a maximum of 63.47% of grain expansion at 14% feed moisture using 2667 kJ kg⁻¹ of energy. Relevant parameters for microwave processing are power input and residence time. However, the lack of a barrier, such as the pericarp, against humidity transport enables moisture loss before and after expansion (van der Sman, 2016). Therefore, the introduction of an additional coating layer to quinoa grains should improve quality of microwave puffed quinoa snack.

Few investigations have been done on the quality of puffed quinoa; moreover, the properties of quinoa snacks have been scarcely studied. Thus, the aim of this study was to evaluate process conditions and energy use for gun, extrusion and microwave puffing and to characterise physical, chemical and nutritional properties of expanded quinoa with and without coating.

Materials and methods

Materials

Quinoa (Chenopodium quinoa Willd, ecotype Roja INIA) was grown in the Ñuble region, Chile (36.834°S, 72.134°W), harvested on 16 March 2018, cleaned, washed with tap water to remove saponins, dried at 70 °C until 15% moisture content (wet basis) and stored in a dry environment at room temperature until further processing and analysis. Corn starch was purchased from Adiplus (Lima, Peru). Chitosan powder (degree of deacetylation 88%) with an average molecular weight of 32 kDa was provided by UDT (Coronel, Chile). D-sorbitol (≥98%) was purchased from Sigma-Aldrich (St. Louis, MO, USA). All other reagents were of analytical grade and supplied by Merck (Darmstadt, Germany); if not, it will be specified otherwise.

Methods

Coating preparation

Quinoa coating was performed according to Bourtoom & Chinnan (2008) with slight modifications. Corn starch (4%w/v) was dissolved in distilled water by heating (85 °C for 5 min) the mixture on Super Nuova hot plate (Thermo Fisher Scientific, Waltham, MA, USA) and stirring at 333 r.p.m. using heavy duty laboratory GT 21-18 stirrer (Heller, Floral Park, NY, USA) until gelatinisation and cooling to 27 °C. Then, chitosan (2%w/v) was dissolved in acetic acid (1%w/v) by stirring at 500 r.p.m. and heating at 80 °C followed by ultrasonic degassing (E30H; Elma Schmidbauer, Singen, Germany). A starch–chitosan blend was prepared by mixing 100 mL of starch solution and 100 mL of chitosan solution with 4 g of sorbitol as plasticiser, adding finally 250 g of quinoa. After homogenising, grains were stored at 55 °C for 10 h, allowing the coated film to dry until 15%w/w moisture content and to cool at ambient temperature.

Gun puffing

A sample of 250 g of quinoa was loaded into the chamber of an artisanal Peruvian puffing machine. Once sealed, the chamber was manually driven while heating by a propane gas flame for 12, 14, 16 and 20 min until an internal pressure of 0.27, 0.62, 0.96 and 1.31 MPa, respectively, was achieved, followed by sudden release of pressure to 0.10 MPa. After puffing, samples were cooled to ambient temperature and stored in airtight polyethylene bags until further analysis. Energy consumption was calculated from the heat of combustion of propane and the amount of gas used for heating.

Extrusion

Extrusion was carried out using a 34 mm single screw extruder (UDPH-7; Kaifeng Youdo Machinery, Henan, China). Overall length of extruder barrel was 121 mm giving a length-to-diameter ratio of 3.57. A cylindrical die with a diameter of 2.72 mm was used for all extrusion trials. Extrusion was done at a fixed rate of quinoa seeds of 43 kg h⁻¹ and a screw speed of 500 r.p.m. Moisture contents of raw material varied from 14 to 30%. Moisture content of test material was fixed by adding the required amount of water to quinoa with a moisture content of about 14%w/w and conditioning both in a hermetically sealed container at ambient temperature for 2 h and intermittent stirring. Then, extrudates were collected and stored in airtight polyethylene bags at room temperature for further analysis. Specific energy consumption (SEC) was calculated according to Godavarti & Karwe (1997).

Microwave popping

A domestic microwave oven (UT-Classy720, Ursus Trotter, Chile) was used to heat 60 g of quinoa (20%w/w moisture content), applying a power output of 120, 600, 960 and 1200 W for 60 s, respectively. Expanded samples were stored in airtight polyethylene bags at room temperature for further analysis. Specific energy consumption due to microwave oven use was calculated from power output, residence time and the amount of product, considering an oven efficiency of 95% at sample level (Motevali et al., 2014).

Geometrical properties

Main dimensions, such as diameter of major axis (l), diameter of minor axis (w) and height (h) of grains (50 g), were measured by a digital Vernier caliper (Caldi-6MP, Truper Herramientas, Mexico). Equivalent diameter (d_e) was calculated by:

$$d_{\mathrm{e}}={(l·w·h)}^{1/3}$$

Bulk density

Samples (n = 3) were poured into a graduated cylinder, gently tapped, filled to 100 mL and measuring the weight of known sample volume by an analytical balance (M214A, BEL Engineering, Monza, Italy). Bulk density (ρ) was calculated by dividing weight and volume of quinoa samples.

Expansion index (EI)

Equivalent diameter of expanded grain samples (n = 25) was divided by equivalent diameter of samples without expansion (n = 25) or die diameter in case of extrusion (Kowalski et al., 2016).

Chromatic properties

Colour of quinoa samples (n = 3) was measured using the Colour Quest II Hunterlab colorimeter (Hunter Associates Laboratory, Reston, VA, USA). Results were expressed in a*, b* and L* colour coordinates of the Commission Internationale de l'Eclairage (CEI). Hue (h°), chrome (C*) and total colour change (ΔE) were calculated by the following equations (Kowalski et al., 2016):

$$h^0={tan}^{-1}\left(b^{\ast }/a^{\ast }\right)$$

$$C^{\ast }=\sqrt{\left({a^{\ast }}^2+{b^{\ast }}^2\right)}$$

$$\mathrm{\Delta }E=\sqrt{{\left(\mathrm{\Delta }{a}^{\ast }\right)}^2+{\left(\mathrm{\Delta }{b}^{\ast }\right)}^2+{\left(\mathrm{\Delta }{L}^{\ast }\right)}^2}$$

Water absorption index and water solubility index

Samples (2.5 g) were ground (Oster Blender, 4172-51, Mexico), sized (60 mesh) and mixed intermittently in 30 mL of water (30 °C) for 30 min. After centrifuging at 3000 × g for 10 min (Damon IEC HN-SII, Needham, MA, USA), water absorption index (WAI) was calculated as the ratio of the mass of precipitate to the mass of original dry sample. After removing supernatant and drying overnight at 70 °C, water solubility index (WSI) was expressed as percentage of remaining solids in the original sample weight. All experiments were performed in triplicate.

Mechanical properties

Texture characteristics of quinoa samples were measured by the Instron Universal Testing Machine (ID 4467 H 1998; Instron, Norwood, MA, USA). At least fifteen replicates per treatment were carried out. Compression tests were performed using a cylindrical stainless steel plunger (3.25 mm) and a crosshead speed of 5 mm min⁻¹, where the applied force was plotted against deformation. Length of test extrudates was 30 mm. Texture was evaluated in terms of maximum force to rupture (F_crit), apparent modulus of elasticity (E_ap) and fractal toughness (FT) or work before cracking (ASAE, 1998).

Scanning electron microscopy (SEM)

Raw and puffed grains (n = 9) were observed as whole kernels or as sections made along the transverse axis. Samples were mounted on aluminium stubs and sputter-coated with gold. The observation of the microstructure was carried out using the Jeol JSM-6380LV scanning electron microscope (Joel, Akishima, Japan) at an accelerating voltage of 20 kV. Digital SEM images were processed by open-source Java-based ImageJ software to determine size distribution of 50 randomly selected air cavities and starch granules twice.

Thermal analysis

Thermal stability of quinoa grains and insufflates was determined by thermogravimetric (TG-DTG) analysis (209 F3 TG; Netzsch-Gerätebau GmbH, Selb, Germany). All samples (3–5 mg) were heated from 20 to 600 °C under a nitrogen atmosphere at heating rate of 10 °C min⁻¹. Gelatinisation enthalpy was assessed by differential scanning calorimetry (DSC) (204 F1 Phoenix; Netzsch-Gerätebau GmbH) at the same operating conditions but heating samples from 20 to 350 °C.

Proximate analysis

Water, proteins, fat and ash contents were determined in duplicate according to AOAC (1990) (925.10, 920.87, 923.05 and 293.03, respectively). Carbohydrates contents were calculated by difference as: 100 – (proteins + fat + water + ash). Values were expressed on dry weight basis.

In vitro digestibility

In vitro digestibility of organic matter (OMD) was determined in sixfold according to the method described by Boisen & Fernández (1997), which is based on a three-stage enzymatic incubation using porcine pepsine, porcine pancreatin and a mixed multi-enzyme complex of arabinase, cellulase, β-glucanase, hemicellulase, xylanase and pectinase, followed by a filtration step to minimise intestinal protein absorption. In vitro digestibility of proteins (PD) was determined by the Dumas combustion procedure using Leco's TruSpec NC analyser (Leco, St. Joseph, MI, USA), analysing afterwards total nitrogen in digested quinoa samples (Etheridge et al., 1998).

Experimental design and statistical analysis

A two-way factorial design with two levels of factor coating (with and without coating), four levels of factor puffing (without puffing, gun, extrusion and microwave puffing) and n = 3 independent replications was used at each of the 2 × 4 treatment combinations. The n = 24 measurements were taken in a completely randomised order. Gun puffing was assessed by varying pressure, extrusion by moisture content and microwave puffing by power output. Data were analysed by analysis of variance and differences between samples and treatments by Tukey's test (P < 0.05) using R software, version R-3.0.1 (Lucent Technologies, Murray Hill, NJ, USA).

Results and discussion

Expansion characteristics

Expansion index of quinoa popped by gun puffing was between 1.60 and 2.13 m m⁻¹ (Fig. 1a). Pressure increase (0.27–1.31 MPa) favoured expansion; however, grain burning limits maximum pressure. Recovering quinoa grains by a coating layer enhanced expansion for operational pressure higher than 0.75 MPa. Maximum expansion index was found for 1.31 – 1.33 MPa. These values are higher than those reported by Lee et al. (2019) for rice and oat (0.5–0.9 MPa), and may be attributed to differences in grain structure and composition.

Figure 1

Expansion index with and without coating. (a) Gun puffing; (b) extrusion puffing; (c) microwave puffing

Expansion index for quinoa extrudates exhibited a negative parabolic trend with moisture content, where maximum values fluctuated between 1.78 and 1.97 m m⁻¹ at 19–20% humidity with higher expansion after coating (Fig. 1b). Feed moisture below 16% decreased significantly expansion index and blocked mass transport within the extruder, while moisture contents higher than 24% affect negatively extrudate expansion. High moisture contents promote screw mixing in contrast to shear generation, while low moisture contents may reduce water absorption promoting dextrinisation instead of gelatinisation of starch. Reported values for expansion index varied between 1.67 m m⁻¹ for Cherry Vanilla quinoa (Kowalski et al., 2016) and 3.8 m m⁻¹ for an unspecified quinoa variety (Dogan & Karwe, 2003). These differences may be due to extruder's screw characteristics, grain size and chemical composition. In particular, insoluble fibres are able to retain water during heating, lowering water vapour pressure during extrusion puffing. Moreover, insoluble fibres are more rigid than starch-based polymers, inhibiting grain expansion (Ganjyal et al., 2004).

Expansion index of quinoa popped in microwave oven increased with microwave power reaching a maximum value of 1.52 m m⁻¹ at 1200 W for uncoated quinoa (Fig. 1c). Preliminary results indicate that moisture contents of quinoa should range between 18 and 20% to promote a correct absorption of microwave energy and grain expansion. However, final product treated by microwave puffing yielded lower expansion index compared to conventional heating methods because of small residence time and non-uniform heating pattern. Additionally, the lack of a barrier against moisture loss in quinoa results in an inverse temperature gradient in the product and a significant loss of moisture before and after popping.

Specific energy consumption from extruder's drive motor with heat generation by friction that results in the extrusion of raw material varied between 425 and 489 kJ kg⁻¹ (Table 1), which is in the same range as reported by Kowalski et al. (2016). Specific energy consumption due to propane combustion during gun puffing was between 6.96 and 9.27 kJ kg⁻¹ (Table 1). Higher specific energy consumption (1200 kJ kg⁻¹) was found for the microwave operation (Table 1). Microwave energy is generated by stepping up alternating current of 50 Hz from domestic power line up to 2450 MHz. Microwave energy is mostly absorbed by polar molecules like water and then transformed into heat. Heat is not transferred to, but generated in puffed material by direct coupling of energy and mass. In this study, microwave energy consumption was unable to achieve sufficient expansion of quinoa.

Table 1

Physical properties of puffed and unpuffed quinoa grains without and with coating layer (CL)

Treatment	SEC (kJ kg−1)	EI (m m−1)	de (mm)	ρ (kg m−3)	WAI (g g−1)	WSI (%)	C*	h° (°)	ΔE
Raw material	-	-	1.6 ± 0.14a	804 ± 7.46 g	1.91 ± 0.45a	10.7 ± 2.5a	21.5 ± 1.2c	42.4 ± 1.2a	-
Gun puffing	9.27	2.05 ± 0.25c	3.2 ± 0.29c	95.8 ± 3.16a	3.45 ± 0.02b	40.4 ± 0.5d	17.5 ± 1.1b	69.1 ± 0.5b	19.8 ± 1.9b
Extrusion puffing	489	1.78 ± 0.07b	4.8 ± 0.18d	237 ± 4.78c	5.27 ± 0.16d	22.8 ± 0.2b	13.7 ± 0.9a	67.0 ± 0.6b	21.6 ± 2.2b
Microwave puffing	1200	1,52 ± 0,34a	2.4 ± 0.51b	475 ± 3.51d	4.23 ± 0.02bc	7.1 ± 0.1a	15.7 ± 0.7ab	49.0 ± 8.8a	11.8 ± 6.4a
Raw material – CL	-	-	1.7 ± 0.14 a	691 ± 0.82f	4.60 ± 0.04c	8.9 ± 0.1a	17.6 ± 0.8ab	43.5 ± 1.6a	-
Gun puffing – CL	6.96	2.13 ± 0.23c	3.3 ± 0.21c	107 ± 0.31a	4.84 ± 0.14 cd	29.0 ± 0.4c	19.8 ± 0.1b	67.7 ± 0.2b	17.9 ± 1.1b
Extrusion puffing – CL	425	1.97 ± 0.20b	5.2 ± 0.57d	132 ± 1.53b	4.36 ± 0.39bcd	29.0 ± 0.0c	15.1 ± 0.4a	66.7 ± 0.3b	20.4 ± 0.9b
Microwave puffing – CL	1200	1.41 ± 0.38a	2.2 ± 0.58b	600 ± 16.2e	4.11 ± 0.01bc	8.3 ± 0.1a	16.8 ± 2.0a	47.2 ± 9.4a	11.5 ± 0.1a

SEC, specific energy consumption; EI, expansion index; d_e, equivalent diameter; ρ, bulk density; WAI, water absorption index; WSI, water solubility index; C*, chrome; h°, hue; ΔE, total colour change.

Data are expressed as mean ± standard deviation. Different letters within the same column mean significant difference (P < 0.05).

Physical properties

Raw material can be classified as medium-sized grains with an equivalent diameter of 1.6 mm (Table 1). Highest values for equivalent diameter were found for extruded and gun puffed quinoa (Table 1), which agrees with improved puffing performance of these operations.

Bulk density of popped quinoa with and without coating layer (CL) varied between 107–600 kg m⁻³ and 96–475 kg m⁻³, respectively, depending on applied puffing operation (Table 1). Quinoa recovery by chitosan–starch layer favoured grain expansion during extrusion, promoting a more uniform distribution of fibres within the chitosan–starch matrix that may avoid premature cell collapse and enhance porosity of extrudates. However, a relatively high percentage of fibres and proteins in uncoated quinoa grains may affect negatively starch gelatinisation, which results in less expansion and increased bulk density (Ganjyal et al., 2004; Brennan et al., 2008). On the other hand, the absence of coating favours quinoa expansion during microwave puffing. Some damage of the quinoa pericarp during desaponification may promote partial starch gelatinisation in the presence of fibres being helpful in microwave puffing.

Water absorption and water solubility indexes of quinoa samples varied between 1.91 and 5.27 kg kg⁻¹ and 7.1 and 40.4%, respectively (Table 1). A significant increase of WAI was found after introducing chitosan–starch coating and after processing. WSI increased significantly after gun puffing and extrusion. Kowalski et al. (2016) reported WAI and WSI values of 2.82 kg kg⁻¹ and 14%, respectively, for extruded quinoa. These values were influenced by process conditions, starch gelatinisation and carbohydrate content. Increased water absorption depends on temperature and moisture content to produce more homogeneous heating and finally increased starch gelatinisation (Gulati et al., 2016), while high temperature and low humidity will cause dextrinisation instead of gelatinisation and therefore lower water absorption (Ding et al., 2006).

Chrome of quinoa samples ranged between 13.7 and 19.8, while hue varied between 42.4 and 69.1 and total colour change was between 11.5 and 21.6 (Table 1). Chrome or colour intensity decreased significantly due to coating and processing compared to raw material, while hue and total colour change showed major increase for quinoa grains processed by extrusion and gun puffing. These colour changes were attributed to pericarp damage, irregular grain expansion during puffing and starch gelatinisation. Temperatures above 76.7 °C and moisture below 16.9%w/w during puffing promote caramelisation and Maillard reactions of reducing sugars and the degradation of quinoa pigments (Coutinho et al., 2013).

Mechanical properties

Texture, in particular crispness, of puffed snacks has been associated with sensorial perception during mastication (Roudaut et al., 2002). Here, puffing operations for quinoa result in significant decrease of maximum force for rupture and apparent modulus of elasticity compared to raw material (Table 2). Softer and less rigid quinoa snacks of less mechanical strength are the consequence of high amylose content in quinoa. Amylose favours starch gelatinisation and the degree of expansion after heating (Joshi et al., 2014). Coated quinoa exhibited similar mechanical behaviour, except for fractal toughness in case of microwave heating. Coated quinoa treated by microwave heating yielded lowest expansion index (1.41 m m⁻¹) and therefore least internal grain porosity.

Table 2

Mechanical properties of puffed and unpuffed quinoa grains without and with coating layer (CL)

Treatment	Eap (MPa)	FT (mJ)	Fcrit (N)
Raw material	629 ± 148b	5.64 ± 3.11b	64.8 ± 6.2c
Gun puffing	17.1 ± 6.9a	0.71 ± 0.16a	3.15 ± 0.29a
Extrusion puffing	23.8 ± 11.0a	0.61 ± 0.20a	4.88 ± 0.90a
Microwave puffing	26.9 ± 4.9a	0.43 ± 0.04a	6.14 ± 1.40a
Raw material – CL	821 ± 13.7b	3.27 ± 1.74ab	53.6 ± 11.1b
Gun puffing – CL	19.8 ± 5.3a	0.88 ± 0.60a	4.91 ± 0.77a
Extrusion puffing – CL	22.1 ± 4.6a	0.38 ± 0.14a	3.93 ± 0.97a
Microwave puffing – CL	20.8 ± 3.8a	2.58 ± 1.39ab	3.84 ± 0.29a

E_ap, apparent modulus of elasticity; FT, fractal toughness; F_crit, maximum force to rupture.

Data are expressed as mean ± standard deviation. Different letters within the same column mean significant difference (P < 0.05).

Structural properties

Micrographs from scanning electronic microscopy provide the size distributions of starch granules in unprocessed grains and of cavities in puffed products (Fig. 2). Cavity size of gun puffed quinoa varied between 51 and 350 μm (Fig. 2b), extruded quinoa was between 79 and 1100 μm (Fig. 2d), microwave puffed quinoa was between 50 and 400 μm (Fig. 2f), and granules of raw material fluctuated between 0.69 and 1.49 μm showing a polygonal shape (Fig. 2g). Cavity size distribution of coated quinoa samples ranged in the same scale: 39–380 μm after gun puffing (Fig. 2a), 88–1250 μm after extrusion puffing (Fig. 2c) and 64–372 μm after microwave puffing (Fig. 2e) with octahedral and polyhedral shapes. Mean wall size between cavities ranged from 0.81 to 1.76 μm for gun puffed quinoa, 50 to 149 μm for extrudates and 4.32 to 5.69 μm for microwave puffed quinoa. Trater et al. (2005) reported values between 40 and 180 μm for mean wall size and cavities of 800 μm. These differences may be explained by different process conditions.

Figure 2

Size distribution curves for starch granules and air cavities. (a) Gun puffing with coating; (b) gun puffing; (c) extrusion puffing with coating; (d) extrusion puffing; (e) microwave puffing with coating; (f) microwave puffing; (g) raw material

Thermal properties

Thermogravimetry is a relevant technique to understand thermal properties of food. Figure 3 shows the thermal properties of coated and uncoated quinoa and their inflated samples, while results from TG-DTG curves and DSC analysis are summarised in Table 3. According to Fig. 3a,b, the TG-DTG curves of quinoa samples consisted of three stages. The first stage ranged between 21 and 260 °C and is due to the loss of moisture and other volatile components of low molecular weight. In the second stage between 260 and 350 °C, two consecutive reactions attributed to the decomposition and oxidation of organic matter accounting a mean weight loss of 62%w/w with an immediate decrease in mass in a short period of temperature. The third stage ranged between 350 and 600 °C with the generation of a final residue that is attributed to ash formation from inorganic compounds.

Figure 3

TG and DTG curves from uncoated (a) and coated (b) quinoa grains inflated by different methods and DSC curves from uncoated (c) and coated (d) inflated quinoa grains

Table 3

Thermal properties of puffed and unpuffed quinoa grains without and with coating

Sample	Temperature (°C)			Weight loss (%)	Gelatinisation
	TO	TP	TC		TO-TC (°C)	TP (°C)	ΔH (J g−1)
Without coating
Quinoa	21	217	256	16.0	194–225	200	53
	256	302	366	60.8
	366	392	596	23.2
Explosive gun	21	65	150	8.8	74–155	119	134
	250	310	355	72.1
	355	400	295	19.2
Extruder	21	70	260	15.8	206–238	215	57
	260	305	350	62.2
	350	390	595	22.0
Microwave	21	135	260	12.9	206–235	212	67
	260	305	350	60.9
	350	405	595	26.2
With coating
Quinoa	21	85	230	10.7	180–224	192	59
	230	300	370	42.5
	370	405	595	10.3
Explosive gun	21	85	255	14.1	90–222	163	116
	255	300	350	66.4
	350	405	595	19.5
Extruder	21	225	255	14.8	111–161	139	58
	255	305	360	64.5
	360	400	595	16.8
Microwave	21	130	255	12.1	201–233	205	64
	255	305	355	62.5
	355	400	595	25.4

T_O, onset temperature; T_P, peak temperature; T_C, conclusion temperature.

Gelatinisation properties of quinoa polymers in coated and uncoated samples were studied by DSC (Fig. 3c,d). Samples show both endothermic and exothermic effects. Endothermic phase transition appears to depend on crystallinity loss of starch granules and had been previously related to amylose–lipid complex formation (Czuchajowska et al., 1998). Gelatinisation enthalpies ranged between 57 and 134 J g⁻¹ for uncoated inflated samples and between 58 and 116 J g⁻¹ for coated ones (Table 3). In both cases, most of these values fall within the range published for quinoa starch gelatinisation (54–65 J g⁻¹) analysing samples at heating rate of 10 °C min⁻¹ (Li & Zhu, 2018) and 65–71 J g⁻¹ for Indian quinoa starch (Jan et al., 2017). Such differences had been endorsed to the different chemical composition and structure of quinoa starch (Varnadevan & Bertoft, 2015). In addition, the gelatinisation enthalpy of uncoated quinoa grains was 53 J g⁻¹, while coated raw material had a value of 59 J g⁻¹ (Table 3). This indicates that coating has just a slight effect on gelatinisation enthalpy; consequently, coated samples require a little bit more thermal energy to gelatinise polymers. On the other hand, gelatinisation temperature for uncoated puffed samples ranged from 119 to 215 °C and 139 to 205 °C for coated ones. A gelatinisation temperature range for quinoa starch of 56–65 °C and 96–102 °C for quinoa proteins has been reported (Ruales and Nair, 1994a; Li et al., 2016; Ruiz et al., 2016). This difference in gelatinisation temperature may be due to the complex matrix and the association between different components such as starch, protein and minerals in raw and puffed quinoa, as well as thermal process conditions.

Proximate analysis and in vitro digestibility

Experimental results of proximate analysis of quinoa samples are shown in Table 4. Puffing operations promote moisture loss in most of the quinoa snacks due to heating. Moreover, a decrease in mineral content, in particular, in gun puffed quinoa was detected because of the loss of pericarp as a consequence of the process conditions. Protein contents of quinoa samples varied between 11.06 and 13.28 g 100 g⁻¹ DW (Table 4). A significant loss of proteins was found for quinoa treated by gun puffing compared to microwave and extrusion puffing due to the loss of cotyledon and grain embryo that possess the highest protein contents. Extrusion provokes degradation of lipids because of high temperature (110–120 °C) and screw speed (500 r.p.m.). Additionally, extrusion may induce micro- and macromolecular changes in the starch structure, modifying amylose and amylopectin chain behaviour, which affects starch gelatinisation and crystallisation.

Table 4

Nutritional properties (g 100 g⁻¹ DW) and in vitro digestibility of organic matter (OMD) and proteins (PD) for puffed and unpuffed quinoa grains without and with coating layer (CL)

Treatment	Humidity	Ash	Fat	Proteins	Carbohydrates	OMD (%)	PD (%)
Raw material	7.12 ± 0.01c	2.40 ± 0.01e	6.41 ± 0.14d	12.89 ± 0.10 cd	78.30 ± 0.04a	81.3 ± 1.8ab	23.67 ± 0.03a
Gun puffing	6.99 ± 0.21c	2.21 ± 0.02b	6.07 ± 0.13d	12.07 ± 0.08b	79.66 ± 0.19b	83.3 ± 3.6ab	93.37 ± 0.03 g
Extrusion puffing	10.67 ± 0.14e	2.32 ± 0.01d	3.35 ± 0.03a	12.89 ± 0.07 cd	81.44 ± 0.10d	84.0 ± 0.6abc	87.16 ± 0.03e
Microwave puffing	5.17 ± 0.06a	2.37 ± 0.01e	6.37 ± 0.11d	12.91 ± 0.06 cd	78.35 ± 0.17a	83.0 ± 1.1ab	50.17 ± 0.03c
Raw material – CL	8.34 ± 0.03d	2.27 ± 0.01c	4.11 ± 0.01b	13.22 ± 0.03d	80.40 ± 0.01c	-	-
Gun puffing – CL	7.08 ± 0.05c	2.00 ± 0.02a	5.13 ± 0.10c	11.06 ± 0.11a	81.81 ± 0.20d	88.0 ± 0.4c	90.65 ± 0.03f
Extrusion puffing – CL	7.99 ± 0.05d	2.26 ± 0.01c	4.03 ± 0.01b	13.28 ± 0.05d	80.43 ± 0.07c	84.2 ± 0.3bc	79.20 ± 0.03d
Microwave puffing – CL	6.51 ± 0.04b	2.33 ± 0.01d	5.46 ± 0.34c	12.57 ± 0.22c	79.65 ± 0.12b	79.8 ± 3.9a	38.47 ± 0.03b

OMD, in vitro digestibility of organic matter; PD, in vitro digestibility of proteins.

Data are expressed as mean ± standard deviation. Different letters within the same column mean significant difference (P < 0.05).

Nutritional value of food depends on the bioavailability of macro- and micronutrients to the digestive system. In contrast to microwave puffing, gun and extrusion puffing of coated quinoa grains are helpful to improve the in vitro digestibility of organic matter from quinoa. Severe thermal and shear conditions at gun puffing and extrusion, respectively, provoke degradation of molecular components and starch gelatinisation (Rathod & Annapure, 2017). Less intense microwave heating results in lower OMD from coated quinoa that was due to partial starch gelatinisation and presence of fibres. In addition, the digestibility of proteins also increased with the intensity of thermal process conditions. PD values were found in the range of 79–93% for quinoa samples processed by extruder and explosive gun (Table 4) and agree with Ruales & Nair (1994b). The presence of resistant starch, amylose–lipid or starch–protein complexes and insufficient energy consumption are related to low PD after microwave puffing (Eerlingen & Delcour, 1995).

Conclusions

Recovery of quinoa by a coating layer that consists in starch, chitosan and sorbitol improved physicochemical characteristics after gun and extrusion puffing due to higher capacity to retain water vapour within the grains. Although microwave processing yielded high energy consumption, coating of quinoa did not result in sufficient expansion. Further investigations should be done to develop new coating material to retain water within the quinoa grains before their expansion. Gun and extrusion puffing result in satisfactory quality quinoa snacks, avoiding the loss of nutritional properties with suitable bioavailability of organic matter and proteins compared to microwave puffed quinoa and raw material. Our results could be relevant for food industry, but further research is needed in other quinoa varieties for microwave processing.

Acknowledgments

The authors declare no conflict of interest. The authors gratefully acknowledge the research funding from FIA (Project No. PYT-2017-0495).