Sorption isotherms and physical properties of three quinoa varieties grown at two altitudes

Abstract

Quinoa seeds grown in the high Andean regions are susceptible to adulteration with seeds grown in other regions due to the increasing global demand. Therefore, characterising quinoa will help to develop a process for certifying origin. The aim was to determine the physical properties, sorption isotherms, and microstructural characteristics of three quinoa varieties grown at 2600 m and 3818 m above sea level. Samples were taken at both altitudes and their shape, size, colour, density, and adsorption isotherms were determined using the standard static gravimetric method. The results were fitted to six models and their microstructure was analysed using Scanning Electron Microscopy (SEM). The size varies from 1.30 – 1.83 mm, the density is 966 - 1097 kg/m³. The isothermal curves show a type II behaviour according to the Brunauer-Emmett-Teller classification, the water in the monolayer ranges between 0.038 and 0.047 g H₂O/g d.b. The water uptake was higher in seeds grown at 2600 m above sea level than at 3818 m. The starches have spherical and oval shapes with 25 μm and 30 μm diameter, respectively. The study demonstrated the feasibility of establishing physical differentiation criteria for altitude-adapted varieties to determine storage conditions, quality control, traceability, and denomination of origin.

Keywords:
Quinoa; Grain cultivation; Adulteration; Origin marking; Traceability

Highlights

Quinoa seeds from different altitudes show distinct physical and microstructural properties

Water uptake is higher in quinoa grown at 2600 m than at 3818 m above sea level

Mathematical models and AI can identify altitude-based patterns and certify their origin

1 Introduction

Quinoa (Chenopodium quinoa Willd.) is an Andean grain of great importance due to its nutritional properties, food security, agricultural sustainability, cultural value, and adaptability to the extreme conditions of high-altitude regions. This grain originated in the Andes of South America, where it was the “mother grain” for the Incas (Miller et al., 2021).

Physical properties and sorption isotherms are of great importance for traceability. Therefore, the grains have oval shapes, with a 1.0 – 2.6 mm diameter, and the color varies from pale yellow and red to dark or even black (Aluwi et al., 2017; Graf et al., 2016). The density for six quinoa genotypes was 0.63 – 0.73 g/cm³ (Ghumman et al., 2021; Graf et al., 2016).

Quinoa starch granules are relatively small (0.5 - 3 μm in diameter) and present an unimodal size distribution (Timgren et al., 2013). Variations in starch granule size are influenced by geographical origin and growing conditions (Jan et al., 2017). This may influence viscoelastic properties, as varieties grown at higher altitudes have higher elasticity and viscosity, indicating a stronger gel network compared to those grown at lower altitudes. This could be related to structural differences in the starches present in quinoa, as affected by environmental conditions (Polo-Muñoz et al., 2023). The small starch granules in quinoa can stabilise emulsions and these starch granules have applications in food, due to their high stability (Li et al., 2020). In addition, sorption isotherms can be experimentally constituted and fitted to mathematical models to predict storage moisture in seeds (Rosa et al., 2015; Villa Vélez et al., 2015).

Globally, it has become a new alternative crop in many places due to its nutritional properties and environmental adaptability (Castro‐Alba et al., 2019). In addition, it can adapt to different climatic and soil conditions (Gęsiński, 2008), even some quinoa genotypes can grow under high salt concentrations, such as in seawater (Manaa et al., 2019; Miller et al., 2021). It is remarkable that this crop presents frost resistance and can survive extreme temperatures ranging from -8 °C to 40 °C (González et al., 2011; Martínez et al., 2009).

In 2016, Peru was one of the major producers and exporters of quinoa worldwide, exporting to 62 countries, among which the United States of America (USA) accounted for 35% of total exports, even though this meant a decrease of 30% compared to the previous year (Agrodata, 2016). This situation has affected producers in the Puno region (3818 m.a.s.l.) due to the adulteration of organic products with conventional quinoa grown in coastal regions where higher yields and up to two annual harvests are obtained. Quinoa is currently grown in South America, North America, Asia, Europe, and elsewhere. However, Peru has been the largest producer and exporter of quinoa in recent years, reaching a total of 160,000 metric tons (Fathi & Kardoni, 2020; Gamboa et al., 2020).

In Peru, quinoa production is concentrated mainly at altitudes between 2500 and 3900 meters above sea level, where the climate is temperate to cold and there are frequent frosts. These regions are highly dependent on rainfall for the development of the crop. On the other hand, in the coastal regions, which are below 500 meters above sea level, the agroecological conditions favor higher quinoa yields. Therefore, the grains grown in different geographical environments have variations in their physical and chemical properties, as well as in their yield per hectare (Bazile et al., 2013).

The physicochemical characteristics of grains are fundamental parameters for evaluating their quality and complying with market regulations. However, there is little information on these characteristics in seeds grown at different geographical altitudes. These physicochemical properties are essential tools to differentiate the origin of the grains and establish storage conditions, quality control, traceability, and designation of origin. Therefore, this study aimed to analyze physical properties such as geometric, gravimetric, and chromatic properties. In addition, the sorption isotherms and microstructure of quinoa grains grown at different altitudes.

2 Materials and methods

2.1 Quinoa

Quinoa grains of three varieties (Table 1) were purchased from the National Institute for Agrarian Innovation (INIA). They were then cleaned, washed with water to eliminate saponin, dried at 70 °C to 12% moisture content, and stored in a dry environment at 49% relative humidity at room temperature of 15 °C until further processing and analysis.

Table 1
Sample of quinoa grains.

Varieties	Altitude	Experimental Station	Classification	Coordinates
	(m.a.s.l.)
Salcedo INIA (SI-P)	3818	INIA – P	Suni or Jalca	S16° 1 4' 50”.
Negra Collana (NC-P)	3818	INIA – P		O69° 05' 30”.
Pasankalla (PA-P)	3818	INIA – P
Salcedo INIA (SI-A)	2600	INIA - A	Yunga	S16°23'56”
Negra Collana (NC-A)	2600	INIA - A		O71°32'6”
Pasankalla (PA-A)	2600	INIA - A

2.1.1 Physical properties

2.1.1.1 Geometric properties

The main dimensions, such as major axis diameter (l), minor axis diameter (a), and height (h) of the grains (10 g), were measured using a digital caliper (Caldi-6MP, Truper, Mexico). Geometric diameter (Dg) and sphericity (φ) of seeds were calculated using Equation 1 and Equation 2, following the methodology of Mohsenin (1986).

$Dg\mathrm{}=\mathrm{}(l\mathrm{*}a\mathrm{*}h{)}^{1/3}$ (1)

$\phi \mathrm{}=\mathrm{}\frac{Dg}{l}$ (2)

2.1.1.2 Gravimetric properties

They were calculated using Equations 3, 4, and 5, presented in Table 2. To measure the true density, the toluene displacement method was applied in a pycnometer; the bulk density was determined by filling a 250 mL test tube, and the porosity was determined by the volumetric fraction of air inside.

Table 2
Equations for determining gravimetric properties.

Models	Equations		Reference
True Density (ρr)	$\rho r=\mathrm{}\frac{Ms\mathrm{*}\rho t}{Mt}$	(3)	Konak et al. (2002)
Bulk Density (ρa)	$\rho a=\mathrm{}\frac{Ms}{V}$	(4)	Konak et al. (2002)
Porosity (ɸ)	$\mathrm{ɸ}=\mathrm{}(1-\frac{\rho a}{\rho r}\mathrm{})\mathrm{*}100$	(5)	Barbosa-Cánovas et al. (2005)

• Where: true density kg/m³ (ρr), seed mass kg (Ms), toluene mass kg (Mt), toluene density kg/m³ (ρt), bulk density kg/m³ (ρa), and occupied volume m³ (V).

2.1.1.3 Seed weight

To obtain the mass, each seed was weighed on a precision electronic balance (reading at 0-001 g). The weight of 1000 seeds was determined by weighing 100 seeds in triplicate and then extrapolating this weight to 1000 seeds in each quinoa variety (Vilche et al., 2003).

2.1.1.4 Chromatic properties

The color of quinoa samples (n=3) was measured using a Hunterlab Colour Quest II colorimeter (Hunter Associates Laboratory, Reston, VA, USA). Results were expressed in a*, b*, and L* color coordinates of the International Commission of Ecology of the Commission Internationale de l'Eclairage (CEI). Purity (C) Equation 6 and Hue (H) Equation 7 (Kowalski et al., 2016).

$C=({{a^{*}}^2+{b^{*}}^2)}^{\frac{1}{2}}$ (6)

$H={tan}^{-1}\left(\frac{b^{*}}{a^{*}}\right)$ (7)

2.1.2 Sorption isotherms

The static gravimetric method Aviara (2020) was used to acquire sorption data concerning moisture content in reference to saturated water activity solutions. Measurements were carried out at a temperature of 15 ± 0.5 °C. Saturated solutions were prepared using different levels of water activity, as shown in Table 3, and stored in jars with airtight lids. In addition, 5 g (±0.0001) of each variety of quinoa grain were weighed for 30 days.

Table 3
Salt concentration and water of salts from different water activities.

Salts	aw	Salt/gr	Water/mL
Lithium chloride	0.112	15	8.5
Sodium chloride	0.657	20	6
Potassium sulfate	0.979	20	9

2.1.3 Analysis of sorption isotherms with mathematical modeling

2.1.3.1 Mathematical models of isotherms

The models utilized are presented in Table 4, where the corresponding equations, specifically Equations 8 to 13, are detailed for each model.

Table 4
Mathematical models applied to adsorption isotherms in quinoa grains.

Models	Equations
GAB (van den Berg, 1985)	$\mathit{m}=\frac{{\mathit{m}}_0\mathit{C}\mathbf{}\mathit{k}{\mathbf{}\mathit{a}}_{\mathit{w}}}{\left[(1-\mathit{k}\mathbf{}{\mathit{a}}_{\mathit{w}})(1-\mathit{k}{\mathbf{}\mathit{a}}_{\mathit{w}}+{\mathit{C}\mathbf{}\mathit{k}\mathbf{}\mathit{a}}_{\mathit{w}})\right]}$	(8)
Henderson (1952)	$\mathit{m}={\left[-\frac{\mathit{l}\mathit{n}(1-{\mathit{a}}_{\mathit{w}})}{\mathit{C}}\right]}^{1/\mathit{n}}$	(9)
Halsey (1948)	$\mathit{m}={\left[-\frac{\mathit{A}}{\mathit{l}\mathit{n}\mathbf{}{\mathit{a}}_{\mathit{w}}}\right]}^{1/\mathit{B}}$	(10)
Oswin (1946)	$\mathit{m}=\mathit{A}\mathbf{}(\frac{{\mathit{a}}_{\mathit{w}}}{1-{\mathit{a}}_{\mathit{w}}}{)}^{\mathit{B}}$	(11)
Peleg (1993)	$\mathit{m}=\mathit{A}{\mathbf{}\mathit{a}}_{\mathit{w}}^{\mathit{B}}+\mathit{C}\mathbf{}{\mathit{a}}_{\mathit{w}}^{\mathit{D}}$	(12)
Smith (1947)	$\mathit{m}=\mathit{A}+\mathit{B}\mathit{L}\mathit{n}(1-{\mathit{a}}_{\mathit{w}})$	(13)

• Where: m: equilibrium moisture content (%); m_o: monolayer water, a_w: water activity, C and K: GAB model constants; A, B, C, D and n are model constants (Equations 8, 9, 10, 11, 12 and 13).

The criteria for determining the best-fit model were the percentage error (%E) (Equation 14) and the correlation coefficient (R²). A sorption model is considered acceptable when it has a %E of less than 10% and an R² close to 1, which generally indicates an adequate fit for practical purposes (Lomauro et al., 1985). In addition, for validation, Python codes were developed for the simulation and modelling of sorption isotherms in the 0.01 to 0.95 water activity range.

$\mathrm{E}\mathrm{}\left(\mathrm{\%}\right)=\mathrm{}\frac{100}{\mathrm{N}}\sum \left|\frac{\mathrm{M}\mathrm{e}-\mathrm{M}\mathrm{p}}{\mathrm{M}\mathrm{e}}\right|$ (14)

Where “Me” and “Mp” correspond to the experimental and predicted moisture content, respectively, and “N” is the number of experimental values.

2.1.4 Environmental Scanning electron microscopy (ESEM)

The grains at different water activities were observed in cross-sectional sections. They were mounted in aluminum sockets. Microstructure observation was carried out using a Jeol JSM -6380LV scanning electron microscope (SEM) (Joel, Akishima, Japan) at an accelerating voltage of 20 kV. The digital SEM images were processed using Java-based, open-source ImageJ software to determine starch size and shape distribution.

2.1.5 Statistical analysis

The sorption isotherm analysis was performed analytically with the six mathematical models: G.A.B., Oswin, Henderson, Peleg, Halsey, and Smith. For physical properties, the statistical method of analysis of variance (ANOVA) was applied with a significance level of 95%. In addition, the statistical program Statgraphics Centurion XVI, version 16.1.11, was used with a Tukey’s test to identify possible differences between treatments.

3 Results and discussion

3.1 Geometric properties

Table 5 shows the geometric properties of the quinoa grains. It can be observed that the geometric diameters of the PA-A variety are superior to PA-P. On the other hand, SI-P and NC-P varieties are superior to SI-A and NC-A varieties.

Table 5
Geometric properties of three quinoa varieties grown at two geographic altitudes.

Quinoa varieties	Geometric properties				Sphericity
	Major	Minor	Thickness	Geometric
	Axis (l)	Axis (a)	Axis (h)	Diameter (Dg)
SI-A	1.8 ± 0.2bc	1.8 ± 0.3bc	1.0 ± 0.2ab	1.5 ± 0.2bc	0.8 ± 0.01a
NC-A	1.6 ± 0.1ab	1.5 ± 0.1ab	0.9 ± 0.1ab	1.3 ± 0.1ab	0.8 ± 0.01a
PA-A	2.1 ± 0.1c	2.0 ± 0.1c	1.2 ± 0.1c	1.7 ± 0.1d	0.8 ± 0.01a
SI-P	2.0 ± 0.2c	1.9 ± 0.3c	1.0 ± 0.2bc	1.6 ± 0.2cd	0.8 ± 0.01a
NC-P	1.6 ± 0.1a	1.6 ± 0.1a	1.0 ± 0.1a	1.4 ± 0.1a	0.9 ± 0.01b
PA-P	1.9 ± 0.2c	1.9 ± 0.1c	1.1 ± 0.1bc	1.6 ± 01cd	0.8 ± 0.01a

• Values are means of three replicates ± standard deviation. Values in the column with the same letter in superscript are not significantly different from each other (p ≤ 0.05).

The PA-A variety is classified as large, while the SI-A, SI-P, NC-P and PA-A varieties are classified as medium, and the NC-A variety as small, according to the Peruvian Technical Standard (Instituto Nacional de Calidad, 2009). This standard establishes that large grains have a diameter greater than 1.70 mm, medium grains between 1.7 mm and 1.4 mm, and small grains between 1.4 mm and 1.2 mm, respectively. This variation is due to factors such as soil type, water, rainfall, geographic altitude, and environmental conditions. The largest and smallest diameters were 1.6 - 2.1 mm and 1.5 - 2.0 mm, respectively. These diameters are in the range reported by (Abderrahim et al., 2015; Vilche et al., 2003).

The quinoa plant produces flat, oval-shaped grains (like a flattened sphere) Figure 1. In addition, they have main parts such as the embryo, perisperm, and episperm (Tapia et al., 1979). Significant variations may be due to the genotypes of quinoa (Ghumman et al., 2021).

Figure 1
Photograph of whole-grain quinoa. Negra Collana (A), Pasankalla (B), Salcedo INIA (C).

In addition, the seeds had different shapes, such as circles, prolate spheroidal, lenticular, and ellipsoidal, with a slight protrusion corresponding to the radicle. In Figure 1, traces of saponin can be observed in the pericarp and embryo. The adherence of saponin could vary among varieties of each genotype and geographical location where it was grown (Azcón-Bieto & Talón, 2003). On the other hand, high-altitude crops had a higher amount of saponin.

3.2 Gravimetric properties

3.2.1 Density

The bulk density, true density, and porosity of quinoa grains ranged from 715 to 805 kg/m³, 966.7 to 1079 kg/m³, and 18.98% to 36.69%, respectively (Table 6). The varieties PA-P and NC-P showed a higher porosity varying between 33.16% to 33.69%, and SI-P showed a low value of 18.98%, while SI-A, NC-A, and PA-A ranged between 22.42% and 24.66%. The results are similar to those reported by (Vilche et al., 2003).

Table 6
Density of the three quinoa varieties grown at two different altitudes.

Quinoa Varieties	Bulk Density kg/m3	True Density kg/m3	Porosity (%)
SI-A	750 ± 0.1c	966.7 ± 0.1a	22.42 ± 0.1b
NC-A	785 ± 0.1d	1050.5 ± 0.1c	24.66 ± 0.1c
PA-A	805 ± 0.1e	1068.5 ± 0.1d	24.66 ± 0.1c
SI-P	805 ± 0.1e	993.6 ± 0.1b	18.98 ± 0.1a
NC-P	695 ± 0.1a	1097.8 ± 0.1d	36.69 ± 0.1e
PA-P	715 ± 0.1b	1069.7 ± 0.1e	33.16 ± 0.1d

• Values are means of three replicates ± standard deviation. Values in the column with the same letter in superscript are not significantly different from each other (p ≤ 0.05).

Quinoa seed densities were lower than cumin and caper seeds (Dursun & Dursun, 2005; Singh & Goswami, 1996), while bulk density was higher. Ghumman et al. (2021) evaluated six quinoa genotypes; the range of grain density was 630 - 730 kg/m³. Bulk density values provide useful information for analyzing heat transfer through grains, quality control, design of transport systems, cleaning and classification, and true density for designing storage, packaging, dehydration, and transport systems.

Grain weights are shown in Figure 2. SI-P and NC-P samples have higher weights than SI-A and NC-P; however, PA has no difference. Li & Zhu (2018) showed that the weights of 1000 grains of four quinoa cultivars grown in southwest Germany ranged from 1.2 to 3.3 g; however, the samples differ, presenting more weight up to 3.6 g. On the other hand, Contreras-Jiménez et al. (2019) determined grain weights varying between 2.5 and 4.1g, and the variations are attributed to variety, size, and moisture content.

Figure 2
Weight of one thousand grains (g) of quinoa varieties grown at different altitudes.

3.3 Chromatic properties

The colors correspond to the outer layer (pericarp) of the seeds under normal physiological conditions (Table 7). The maximum value of lightness (L) was recorded for the SI-P variety, being 85.704 (white quinoa), while the minimum value of L* recorded was 34.14 for the NC-P variety (black quinoa). As expected, these values indicate a reduced lightness of the black grains and the absence of a predominant color. The chroma values varied in the range of 1.6274 to 25.6130. The SI-P variety presented low values, and the PA-A presented the highest value.

Table 7
Color of three quinoa varieties grown at two altitudes.

Varieties	L	a*	b*	C	H
SI-A	80.8304	-0.3332	2.1168	2.1429	98.9454
NC-A	43.1217	3.0441	-1.2876	3.3052	337.0725
PA-A	65.841	8.6704	24.1008	25.6130	109.9771
SI-P	85.704	-0.556	1.5295	1.6274	109.9771
NC-P	34.1472	2.7447	-2.3774	3.6312	319.1016
PA-P	54.5424	8.5654	19.80799	21.5806	66.6153

• Seed color varies from white, gray, and pale yellow to red and dark. Seeds of some genotypes have hints of pink and magenta shades. Seeds are held on the plant in five-lobed perianths.

The Salcedo INIA variety grown at 3818 m.a.s.l. had a high lightness of L 85.704, while the one grown at 2600 m.a.s.l. had L 80.8304. This variety is characterized by low colors in a*, indicating the absence of red to green intensity, and b* values that refer to a pale yellow to cream background. In contrast, the Negra Collana and Pasankalla varieties grown at 2600 - 3818 m.a.s.l. showed higher L 43.1217 and 65.84 than those grown at 3818 m.a.s.l., with L values of 34.14 and 54.54, respectively. The decrease in L of the different quinoa varieties indicates a higher contribution of betacyanins (Escribano et al., 2017). The parameters of b* color yellow demonstrate a contribution of betaxanthins for those quinoa red-violet colors containing betaxanthins and betacyanins. Escribano et al. (2017) showed that as the portion of betaxanthin is reduced, the b* parameter value decreases, which measures the yellow color. At the same time, the value of the a* parameter increases due to the higher content of betacyanins. These values can be compared in Table 7. A change in hue angle (H) is also linked to the lower proportion of betaxanthins decreasing from yellow varieties to those containing more betacyanins and thus exhibiting red to violet color.

The results of L*, a*, and b* (Figure 3) are similar to those reported by Arapa Carcasi & Padrón Pereira (2015) who determined the color of quinoa grains, obtaining the following results: L* 76.16; a* 3.79, and b* 25.99; and Escribano et al. (2017) who determined the color of the white varieties (Blanca Junín, Rosada Huancayo), with values of L* 72.04-75.6, Black Collana variety with values of L* 41.57- 45.26, a* 0.89-2.26, b* 2.80- 8.51 and yellow grains, with values of L* 60.76-67.01, a* 6.41-12.94, and b* 26.81-28.78.

Figure 3
Distribution of a* and b* of the three quinoa varieties: Salcedo INIA (SI), Negra Collana (NC) and Pasankalla (PA).

The diversity of colors present in quinoa grains offers multiple possibilities for characterization through image analysis, which is an eco-friendly alternative of low cost and practical use, based on these data, to give added value by taking advantage of its color characteristics, e.g., in the extraction of vegetable color powders or flours suitable for different applications in the food industry. According to the results of the comparison between the quinoa varieties, there is a statistically significant difference between the varieties at a 95% confidence level.

3.4 Adsorption isotherms

The initial moisture content of quinoa grains was 0.04 – 0.05 g water/g d.m. Figure 4 shows the experimental equilibrium moisture content, based on the water activity of three varieties of quinoa grown at two geographical altitudes. It can be seen that type-II isotherms of sigmoid or S-type were obtained in this image, which are the most frequent in the food.

Figure 4
The equilibrium moisture content of quinoa grains grown at two altitudes.

After Aw greater than 0.75, a cross-linking of the curves is observed, independence between the working temperatures, since it seems that above 0.75, there could be an increase in enzymatic activity and movement between water, carbohydrate, and protein molecules.

According to the results obtained in this research, a relative humidity of 75% and a temperature between 7 and 22 °C to prevent enzymatic reactions of deterioration of carbohydrates, proteins, vitamins, etc., could be considered optimal conditions for flour storage.

3.5 Isotherm models

Table 8 shows the results of the calculations of the mathematical models of G.A.B., Oswin, Henderson, Peleg, Smith, and Halsey, evaluated at 15 °C, the values of R² and % E., showed particular characteristics of each model at certain a_w ranges.

Table 8
Results of adsorption isotherms.

Model	Constant	SI-P	SI-A	NC-P	NC-A	PA-P	PA-A
G.A.B.	Xm	0.0448	0.0380	0.0470	0.0450	0.0430	0.0390
	C	43.5086	22.3100	35.7420	42.5120	55.9420	20.1795
	K	1.4634	1.5470	1.3170	1.3920	1.5650	1.8150
	R2	0.7530	0.7530	0.7750	0.7710	0.7100	0.7313
	%E	1.7364	1.1817	2.3373	1.3988	2.9344	1.0101
OSWIN	A	10.6200	12.0149	13.8495	13.0878	12.0470	10.6927
	B	0.3875	0.3841	0.3035	0.2993	0.3523	0.3615
	R2	0.9966	1.0000	0.9985	0.9980	0.9992	0.9990
	%E	4.9228	2.5945	11.4676	13.5838	10.7780	14.0104
HENDERSON	K	-1.9930	-1.9120	-2.1780	-2.1220	-1.8970	-1.7800
	N	1.3510	1.2700	1.4840	1.4790	1.2850	1.2510
	R2	0.8830	0.8490	0.8870	0.8940	0.8770	0.8870
	%E	1.0000	0.8486	0.8630	0.8488	0.9459	0.8014
PELEG	k1	36.1920	43.1193	32.0116	27.4206	36.9248	33.9075
	n1	5.2053	6.0448	4.9655	6.0346	5.2304	5.3976
	k2	14.8558	14.7785	15.7956	17.4181	13.7472	12.7974
	n2	0.5188	0.4921	0.4823	0.5671	0.5215	0.5842
	R2	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000
	%E	0.0064	0.0001	0.0165	0.1942	0.0010	0.0002
SMITH	A	7.2227	3.5254	4.2618	3.8508	3.0494	2.3144
	B	10.3220	12.6902	10.4006	9.7033	11.2837	10.4997
	R2	0.9998	0.9986	0.9983	0.9975	1.0000	1.0000
	%E	30.3709	3.6369	2.9245	3.4933	0.0505	0.0633
HALSEY	A	2.0017	2.1950	5.5922	4.8403	2.4647	2.5210
	B	1.1803	1.1709	1.4691	1.4589	1.2374	1.2711
	R2	0.9588	0.9759	0.9589	0.9550	0.9655	0.9672
	%E	52.3772	51.1560	43.6762	43.9921	49.1140	46.1168

The Peleg model best describes the experimental data in all varieties grown at two altitudes (with a value of %E 0.0001 – 0.1942% and R² 1), followed by the Henderson model (%E value 0.8014 - 1 and R² 0.88) and the GAB model (with %E 1.0101 – 2.9344 and R² 0.77). The values were evaluated according to Lomauro et al. (1985). The curve in the GAB model showed a sigmoid shape, typical of Type II. According to the BET classification, the values of the adsorption isotherm are within the values obtained by (Pumacahua Ramos et al., 2016) starch powders (Al-Muhtaseb et al., 2004) and other food products (Bejar et al., 2012). Monolayer moisture content (X_m) (Table 8) presents lower values than those reported by Pumacahua Ramos et al. (2016), namely monolayer values for quinoa seeds are between 0.087-0.059 g water/g db; and also, by Tolaba et al. (2004), 0.62 g water/g bs in quinoa seeds, but it was higher than amaranth grains (0.0102 g water/g bd (Calzetta Resio et al., 1999). X_m are parameters for the control of food storage and spoilage (Lomauro et al., 1985).

The monolayer moisture corresponds to the product moisture when the primary adsorption points are saturated by water molecules; C and k are energy constants. C represents the chemical potential difference of solute molecules between sorption layers above the monolayer, and K is the ratio between the chemical potential of solute molecules in the pure liquid state and the sorption layers above (Timmermann et al., 2001). The Peleg mathematical model demonstrates the ability to predict sorption isotherm data (Ouafi et al., 2015; Pumacahua Ramos et al., 2017; Tolaba et al., 2004) also reached the same conclusion.

Python codes have also been developed to simulate and model sorption isotherms in the range of 0.01 to 0.95 water activity (Figure 5), with the models shown to have a regression coefficient of 0.9 to 1 in the Oswin, Smith and Peleg models. The simulation was carried out at 0.001 water activity (aw) intervals and the results showed that there were significant differences (p < 0.05) for each quinoa variety, which indicates that the water retention capacity is related to the structural composition of the starches, which in turn are influenced by the variation of environmental temperatures in the production zones (altitudes) during the ripening period of quinoa grains. This result, together with the application of artificial intelligence (AI) predictive computational models, can help to discretise the altitudinal origin (production zones) of quinoa grains.

Figure 5
Simulation and modelling of sorption isotherms in the range of 0.01 to 0.95 water activity.

3.5.1 Microstructure of quinoa grains

Starch granule size analysis was carried out using scanning electron microscopy (SEM) at magnifications between 500x and 600x (Figure 6). The Salcedo INIA variety grown at 3818 and 2600 m.a.s.l. had spherical shapes with 9-8.5 µm diameters and oblong shapes with 5-12 µm diameters, respectively. These results agree with those reported by Capriles & Areas (2013).

Figure 6
Quinoa microstructure. Varieties grown at 2600 masl. Negra Collana (A), Pasankalla (B), Salcedo INIA (C); Varieties grown at 3818 masl. Negra Collana (D), Pasankalla (E), Salcedo INIA (F).

On the other hand, the Negra Collana variety exhibits oblong shapes with an average diameter of 20 µm, similar to wheat granules (Dhital et al., 2011). These characteristics remained consistent at both altitudes. In the case of the Pasankalla variety, spherical shapes sized between 12 and 14 µm were observed at 3818 m.a.s.l., while at 2600 m.a.s.l., they showed oblong shapes with 13 to 30 µm diameters. These results agree with those reported by Wolter et al. (2013) and Atwell et al. (1983).

In addition, higher crystallinity was found in the variety grown in the Altiplano region, indicating a high amylopectin content (Pumacahua Ramos et al., 2016). Variation in the shape and size of starch granules is influenced by the botanical source, genetic variety, and growing conditions (Tester et al., 2004; Atwell et al., 1983; Galwey, 1992).

It is relevant to mention that the differences between our results and those reported in other studies, such as those by Atwell et al. (1983) and Galwey (1992), may be due to the starch characterization. In our case, the surface was analyzed naturally, without performing any purification process of the quinoa starch. It has been observed that 18 to 20 µm packages can contain up to 14,000 small starch granules (Varriano-Marston & Defrancisco, 1984).

4 Conclusion

The study revealed variations in the geometric properties of quinoa seeds, with sizes ranging from 1.397 and 1.683 mm and lightness (L) between 80 and 85. The Negra Collana and Pasankalla varieties grown at 3018 m a.s.l. exhibited higher yields compared to those grown at 2600 m a.s.l., whereas the Salcedo INIA variety showed the opposite trend. The varieties grown at low altitude showed a high water adsorption capacity. The Peleg mathematical model was optimal to represent the experimental data of the sorption isotherms. Starch granules have a spherical and oblong shape of 10 to 25 µm. Additionally, the application of Python simulations proved to be a promising tool for non-destructive analysis of the origin of the quinoa grains, contributing to traceability and quality certification.

• Cite as:
  Zapana-Yucra, F., Guerra Lima, R. S., Medina Espinoza, W., & Prieto, J. M. (2025). Sorption isotherms and physical properties of three quinoa varieties grown at two altitudes. Brazilian Journal of Food Technology, 28, e2024053. https://doi.org/10.1590/1981-6723.05324
• Funding: None.

References

• Abderrahim, F., Huanatico, E., Segura, R., Arribas, S., Gonzalez, M. C., & Condezo-Hoyos, L. (2015). Physical features, phenolic compounds, betalains and total antioxidant capacity of coloured quinoa seeds (Chenopodium quinoa Willd.) from Peruvian Altiplano. Food Chemistry, 183, 83-90. PMid:25863614. http://doi.org/10.1016/j.foodchem.2015.03.029
  » http://doi.org/10.1016/j.foodchem.2015.03.029
• Agrodata. (2016). Exportación de quinua peruana en 2016. Agrodata Perú. Retrieved in 2024, June 19, from https://www.agrodataperu.com/2016/12/quinua-peru-exportacion-2016-noviembre.html
  » https://www.agrodataperu.com/2016/12/quinua-peru-exportacion-2016-noviembre.html
• Al-Muhtaseb, A. H., McMinn, W. A. M., & Magee, T. R. A. (2004). Water sorption isotherms of starch powders. Part 2: Thermodynamic characteristics. Journal of Food Engineering, 62(2), 135-142. http://doi.org/10.1016/S0260-8774(03)00202-4
  » http://doi.org/10.1016/S0260-8774(03)00202-4
• Aluwi, N. A., Murphy, K. M., & Ganjyal, G. M. (2017). Physicochemical characterization of different varieties of quinoa. Cereal Chemistry, 94(5), 847-856. http://doi.org/10.1094/CCHEM-10-16-0251-R
  » http://doi.org/10.1094/CCHEM-10-16-0251-R
• Arapa Carcasi, P., & Padrón Pereira, C. (2015). Determinación de características físicas en semillas de quinua (Chenopodium quinoa Willd. ) mediante procesamiento digital de imágenes. Revista Venezolana de Ciencia y Tecnología de Alimentos, 5(2), 148-165.
• Atwell, W. A., Patrick, B. M., Johnson, L. A., & Glass, R. W. (1983). Characterization of quinoa starch. Cereal Chemistry, 60(1), 9-11.
• Aviara, N. A. (2020). Moisture sorption isotherms and isotherm model performance evaluation for food and agricultural products. In G. Kyzas & N. Lazaridis (Eds.), Sorption in 2020s. London: IntechOpen. http://doi.org/10.5772/intechopen.87996
  » http://doi.org/10.5772/intechopen.87996
• Azcón-Bieto, J., & Talón, M. (2003). Fundamentos de fisiología vegetal. Madrid: McGrawHill.
• Barbosa-Cánovas, G. V., Ortega-Rivas, E., Juliano, P., & Yan, H. (2005). Food powders: Physical properties, processing, and functionality (Vol. 86). New York: Kluwer Academic/Plenum Publishers.
• Bazile, D., Fuentes, F., & Mujica, A. (2013). Historical perspectives and domestication. In A. Bhargava & S. Srivastava (Eds.), Quinoa: Botany, production and uses (pp. 16-35). Wallingford: CABI. http://doi.org/10.1079/9781780642260.0016
  » http://doi.org/10.1079/9781780642260.0016
• Bejar, A. K., Mihoubi , N. B., & Kechaou, N. (2012). Moisture sorption isotherms: Experimental and mathematical investigations of orange (Citrus sinensis) peel and leaves. Food Chemistry, 132(4), 1728-1735. http://doi.org/10.1016/j.foodchem.2011.06.059
  » http://doi.org/10.1016/j.foodchem.2011.06.059
• Calzetta Resio, A., Aguerre, R. J., & Suárez, C. (1999). Analysis of the sorptional characteristics of amaranth starch. Journal of Food Engineering, 42(1), 51-57. http://doi.org/10.1016/S0260-8774(99)00103-X
  » http://doi.org/10.1016/S0260-8774(99)00103-X
• Capriles, V. D., & Areas, J. A. G. (2013). Effects of prebiotic inulin-type fructans on structure, quality, sensory acceptance and glycemic response of gluten-free breads. Food & Function, 4(1), 104-110. PMid:23032642. http://doi.org/10.1039/C2FO10283H
  » http://doi.org/10.1039/C2FO10283H
• Castro‐Alba, V., Lazarte, C. E., Perez-Rea, D., Carlsson, N.-G., Bergenstahl, B., & Granfeldt, Y. (2019). Fermentation of pseudocereals quinoa, canihua, and amaranth to improvemineral accessibility through degradation of phytate. Journal of the Science of Food and Agriculture, 99(11), 5239-5248. PMid:31062366. http://doi.org/10.1002/jsfa.9793
  » http://doi.org/10.1002/jsfa.9793
• Contreras-Jiménez, B., Torres-Vargas, O. L., & Rodríguez-García, M. E. (2019). Physicochemical characterization of quinoa (Chenopodium quinoa) flour and isolated starch. Food Chemistry, 298(May), 124982. PMid:31261014. http://doi.org/10.1016/j.foodchem.2019.124982
  » http://doi.org/10.1016/j.foodchem.2019.124982
• Dhital, S., Shrestha, A. K., Hasjim, J., & Gidley, M. J. (2011). Physicochemical and structural properties of maize and potato starches as a function of granule size. Journal of Agricultural and Food Chemistry, 59(18), 10151-10161. PMid:21838326. http://doi.org/10.1021/jf202293s
  » http://doi.org/10.1021/jf202293s
• Dursun, E., & Dursun, I. (2005). Some physical properties of caper seed. Biosystems Engineering, 92(2), 237-245. http://doi.org/10.1016/j.biosystemseng.2005.06.003
  » http://doi.org/10.1016/j.biosystemseng.2005.06.003
• Escribano, J., Cabanes, J., Jiménez-Atiénzar, M., Ibañez-Tremolada, M., Gómez-Pando, L. R., García-Carmona, F., & Gandía-Herrero, F. (2017). Characterization of betalains, saponins and antioxidant power in differently colored quinoa (Chenopodium quinoa) varieties. Food Chemistry, 234, 285-294. PMid:28551238. http://doi.org/10.1016/j.foodchem.2017.04.187
  » http://doi.org/10.1016/j.foodchem.2017.04.187
• Fathi, A., & Kardoni, F. (2020). The importance of quinoa (Quinoa chenopodium willd.) cultivation in developing countries: A review. Cercetari Agronomice În Moldova, LIII(3), 337-356. http://doi.org/10.46909/cerce-2020-030
  » http://doi.org/10.46909/cerce-2020-030
• Galwey, N. W. (1992). The potential of quinoa as a multi-purpose crop for agricultural diversification: A review. Industrial Crops and Products, 1(2–4), 101-106. http://doi.org/10.1016/0926-6690(92)90006-H
  » http://doi.org/10.1016/0926-6690(92)90006-H
• Gamboa, C., Bojacá, C. R., Schrevens, E., & Maertens, M. (2020). Sustainability of smallholder quinoa production in the Peruvian Andes. Journal of Cleaner Production, 264, 121657. http://doi.org/10.1016/j.jclepro.2020.121657
  » http://doi.org/10.1016/j.jclepro.2020.121657
• Gęsiński, K. (2008). Evaluation of the development and yielding potential of Chenopodium quinoa Willd. under the climatic conditions of Europe. Acta Agrobotanica, 61(1), 185-189. http://doi.org/10.5586/aa.2008.026
  » http://doi.org/10.5586/aa.2008.026
• Ghumman, A., Mudgal, S., Singh, N., Ranjan, B., Kaur, A., & Rana, J. C. (2021). Physicochemical, functional and structural characteristics of grains, flour and protein isolates of Indian quinoa lines. Food Research International, 140, 109982. PMid:33648217. http://doi.org/10.1016/j.foodres.2020.109982
  » http://doi.org/10.1016/j.foodres.2020.109982
• González, J. A., Bruno, M., Valoy, M., & Prado, F. E. (2011). Genotypic variation of gas exchange parameters and leaf stable carbon and nitrogen isotopes in ten quinoa cultivars grown under drought. Journal Agronomy & Crop Science, 197(2), 81-93. http://doi.org/10.1111/j.1439-037X.2010.00446.x
  » http://doi.org/10.1111/j.1439-037X.2010.00446.x
• Graf, B. L., Rojo, L. E., Delatorre-Herrera, J., Poulev, A., Calfio, C., & Raskin, I. (2016). Phytoecdysteroids and flavonoid glycosides among Chilean and commercial sources of Chenopodium quinoa: Variation and correlation to physico-chemical characteristics. Journal of the Science of Food and Agriculture, 96(2), 633-643. PMid:25683633. http://doi.org/10.1002/jsfa.7134
  » http://doi.org/10.1002/jsfa.7134
• Halsey, G. (1948). Physical adsorption on non-uniform surfaces. The Journal of Chemical Physics, 16(10), 931-937. http://doi.org/10.1063/1.1746689
  » http://doi.org/10.1063/1.1746689
• Henderson, S. M. (1952). A basic concept of equilibrium moisture. Agricultural Engineering, 33(1), 29-32.
• Instituto Nacional de Calidad – INACAl. (2009). NTP 205.062:2009: Quinua (Chenopodium quinoa Willd.): Requisitos Lima: INACAL.
• Jan, K. N., Panesar, P. S., Rana, J. C., & Singh, S. (2017). Structural, thermal and rheological properties of starches isolated from Indian quinoa varieties. International Journal of Biological Macromolecules, 102, 315-322. PMid:28396270. http://doi.org/10.1016/j.ijbiomac.2017.04.027
  » http://doi.org/10.1016/j.ijbiomac.2017.04.027
• Konak, M., Çarman, K., & Aydin, C. (2002). PH - postharvest technology: Physical properties of chick pea seeds. Biosystems Engineering, 82(1), 73-78. http://doi.org/10.1006/bioe.2002.0053
  » http://doi.org/10.1006/bioe.2002.0053
• Kowalski, R. J., Medina-Meza, I. G., Thapa, B. B., Murphy, K. M., & Ganjyal, G. M. (2016). Extrusion processing characteristics of quinoa (Chenopodium quinoa Willd.) var. Cherry Vanilla. Journal of Cereal Science, 70, 91-98. http://doi.org/10.1016/j.jcs.2016.05.024
  » http://doi.org/10.1016/j.jcs.2016.05.024
• Li, G., & Zhu, F. (2018). Quinoa starch: Structure, properties, and applications. Carbohydrate Polymers, 181, 851-861. PMid:29254045. http://doi.org/10.1016/j.carbpol.2017.11.067
  » http://doi.org/10.1016/j.carbpol.2017.11.067
• Li, S., Zhang, B., Li, C., Fu, X., & Huang, Q. (2020). Pickering emulsion gel stabilized by octenylsuccinate quinoa starch granule as lutein carrier: Role of the gel network. Food Chemistry, 305, 125476. PMid:31525589. http://doi.org/10.1016/j.foodchem.2019.125476
  » http://doi.org/10.1016/j.foodchem.2019.125476
• Lomauro, C. J., Bakshi, A. S., & Labuza, T. P. (1985). Evaluation of food moisture sorption isotherm equations. Part I: Fruit, vegetable and meat products. Lebensmittel-Wissenschaft + Technologie, 18(2), 111-117.
• Manaa, A., Goussi, R., Derbali, W., Cantamessa, S., Abdelly, C., & Barbato, R. (2019). Salinity tolerance of quinoa (Chenopodium quinoa Willd) as assessed by chloroplast ultrastructure and photosynthetic performance. Environmental and Experimental Botany, 162, 103-114. http://doi.org/10.1016/j.envexpbot.2019.02.012
  » http://doi.org/10.1016/j.envexpbot.2019.02.012
• Martínez, E. A., Veas, E., Jorquera, C., San Martín, R., & Jara, P. (2009). Re-introduction of Quínoa into arid Chile: Cultivation of two lowland races under extremely low irrigation. Journal Agronomy & Crop Science, 195(1), 1-10. http://doi.org/10.1111/j.1439-037X.2008.00332.x
  » http://doi.org/10.1111/j.1439-037X.2008.00332.x
• Miller, M. J., Kendall, I., Capriles, J., Bruno, M. C., Evershed, R. P., & Hastorf, C. A. (2021). Quinoa, potatoes, and llamas fueled emergent social complexity in the Lake Titicaca Basin of the Andes. Proceedings of the National Academy of Sciences of the United States of America, 118(49), 1-8. PMid:34845028. http://doi.org/10.1073/pnas.2113395118
  » http://doi.org/10.1073/pnas.2113395118
• Mohsenin, N. N. (1986). Physical properties of plant and animal materials: Structure, physical characteristics and mechanical properties. New York: Gordon and Breach Science Publishers.
• Oswin, C. R. (1946). The kinetics of package life. III. The isotherm. Journal of the Society of Chemical Industry, 65(12), 419-421. http://doi.org/10.1002/jctb.5000651216
  » http://doi.org/10.1002/jctb.5000651216
• Ouafi, N., Moghrani, H., Benaouada, N., Yassaa, N., Maachi, R., & Younsi, R. (2015). Moisture sorption isotherms and heat of sorption of Algerian bay leaves (Laurus nobilis). Maderas. Ciencia y Tecnología, 17(4), 759-772. http://doi.org/10.4067/S0718-221X2015005000066
  » http://doi.org/10.4067/S0718-221X2015005000066
• Peleg, M. (1993). Assessment of asemi-empirical four parameter general model for sigmoid moisture sorption isotherms. Journal of Food Process Engineering, 16(1), 21-37. http://doi.org/10.1111/j.1745-4530.1993.tb00160.x
  » http://doi.org/10.1111/j.1745-4530.1993.tb00160.x
• Polo-Muñoz, M. P., Garcia-Parra, M. Á., & Roa-Acosta, D. F. (2023). Viscoelastic behavior of gels obtained from five cultivars of quinoa at altitude gradient. Frontiers in Sustainable Food Systems, 7, 1-12. http://doi.org/10.3389/fsufs.2023.1222277
  » http://doi.org/10.3389/fsufs.2023.1222277
• Pumacahua Ramos, A., Gomez, J., Telis Romero, J., Villa-Vélez, H., & Lopes, J. (2016). Isotherms and isosteric heat of sorption of two varieties of Peruvian quinoa. Scientia Agropecuaria, 7(4), 409-417. http://doi.org/10.17268/sci.agropecu.2016.04.06
  » http://doi.org/10.17268/sci.agropecu.2016.04.06
• Pumacahua Ramos, A., Limaylla, K. M., Telis Romero, J., & Lopes Filho, J. F. (2017). Isotermas y calor isostérico de adsorción de agua de almidón de quinua. Biotecnologia en el Sector Agropecuario y Agroindustrial, 15(1), 95-104. http://doi.org/10.18684/BSAA(15)95-104
  » http://doi.org/10.18684/BSAA(15)95-104
• Rosa, D. P., Cantú-Lozano, D., Luna-Solano, G., Polachini, T. C., & Telis-Romero, J. (2015). Mathematical modeling of orange seed drying kinetics. Ciência e Agrotecnologia, 39(3), 291-300. http://doi.org/10.1590/S1413-70542015000300011
  » http://doi.org/10.1590/S1413-70542015000300011
• Singh, K. K., & Goswami, T. K. (1996). Physical properties of cumin seed. Journal of Agricultural Engineering Research, 64(2), 93-98. http://doi.org/10.1006/jaer.1996.0049
  » http://doi.org/10.1006/jaer.1996.0049
• Smith, S. E. (1947). The sorption of water vapor by high polymers. Journal of the American Chemical Society, 69(3), 646-651. PMid:20289451. http://doi.org/10.1021/ja01195a053
  » http://doi.org/10.1021/ja01195a053
• Tapia, M. Gandarillas, H. Alandia, S. Cardoso A. Mujica, A. (1979). Quinua y Kañihua cultivos andinos Bogotá: CIID, Oficina Regional para la America Latina.
• Tester, R. F., Karkalas, J., & Qi, X. (2004). Starch structure and digestibility enzyme-substrate relationship. World’s Poultry Science Journal, 60(2), 186-195. http://doi.org/10.1079/WPS20040014
  » http://doi.org/10.1079/WPS20040014
• Timgren, A., Rayner, M., Dejmek, P., Marku, D., Sj, M., & Marilyn Rayner, C. (2013). Emulsion stabilizing capacity of intact starch granules modified by heat treatment or octenyl succinic anhydride. Food Science & Nutrition, 1(2), 157-171. PMid:24804025. http://doi.org/10.1002/fsn3.17
  » http://doi.org/10.1002/fsn3.17
• Timmermann, E. O., Chirife, J., & Iglesias, H. A. (2001). Water sorption isotherms of foods and foodstuffs: BET or GAB parameters? Journal of Food Engineering, 48(1), 19-31. http://doi.org/10.1016/S0260-8774(00)00139-4
  » http://doi.org/10.1016/S0260-8774(00)00139-4
• Tolaba, M. P., Peltzer, M., Enriquez, N., & Pollio, M. L. (2004). Grain sorption equilibria of quinoa grains. Journal of Food Engineering, 61(3), 365-371. http://doi.org/10.1016/S0260-8774(03)00143-2
  » http://doi.org/10.1016/S0260-8774(03)00143-2
• van den Berg, C. (1985). Development of B.E.T.-like models for sorption of water on foods, theory and relevance. In D. Simatos & J. L. Multon (Eds.), Properties of water in foods (pp. 119-131). Dordrecht: Springer.
• Varriano-Marston, E., & Defrancisco, A. (1984). Ultrastructure of quinoa fruit (Chenopodium quinoa Willd). Journal of Food Structure, 3(3), 165-173.
• Vilche, C., Gely, M., & Santalla, E. (2003). Physical properties of quinoa seeds. Biosystems Engineering, 86(1), 59-65. http://doi.org/10.1016/S1537-5110(03)00114-4
  » http://doi.org/10.1016/S1537-5110(03)00114-4
• Villa Vélez, H. A., Ferreira de Souza, S. J., Pumacaua Ramos, A., Polachini, T., & Telis-Romero, J. (2015). Thermodynamic properties of water adsorption from orange peels. Journal Bioenergy and Food Science, 02(2), 72-81. http://doi.org/10.18067/jbfs.v2i2.32
  » http://doi.org/10.18067/jbfs.v2i2.32
• Wolter, A., Hager, A., Zannini, E., & Arendt, E. K. (2013). In vitro starch digestibility and predicted glycaemic indexes of buckwheat, oat, quinoa, sorghum, teff and commercial gluten-free bread. Journal of Cereal Science, 58(3), 1-6. http://doi.org/10.1016/j.jcs.2013.09.003
  » http://doi.org/10.1016/j.jcs.2013.09.003

Edited by

• Associate Editor:
  Rosinelson da Silva Pena.

Publication Dates

• Publication in this collection
  24 Mar 2025
• Date of issue
  2025

History

• Received
  19 June 2024
• Accepted
  22 Jan 2025