Drying of Yacon (Smallanthus sonchifolius) as a potentialfood product for international commercialization

This paper describes the drying process conditions with the objective to minimise the negative effect over the final quality of the dry Yacon (mallanthus sonchifolius) for international commercialization. Yacon is a tuberous root that grows throughout the Andean mountains in different South American countries. During the last years the interest for this product has been increasing and researchers have collected information that indicates the great potential of yacon. The product cultivation has been promoted over the past few years, due to the health benefits found after consumption. This agricultural product contains FOS (Fructo Oligo Sacarides) of low molecular weight, which is used as alternative sweeteners and, due to the small amount of calories is also attractive and good for overweight and diabetic people. Due to the outstanding importance of its health properties, this studied was carried out to achieve the best parameters of the drying process. For this paper, an experiment was performed considering two factors of the drying process: thickness of the layer (2cm, 4cm and 6cm) and temperature (60°C, 70°C and 80°C), with measurements of the physicochemical properties in fresh, before drying and after the drying process, with three replicates for each combination of levels.


Introduction
Yacon is a tuberous rhizome that growths in the Andean region spreading throughout the humid mountains of Peru, Colombia, Venezuela, Ecuador, Argentina and Bolivia [1]. It is a high-producing crop that yields an average of 30 tons per hectare and has a lifetime of 15-20 days in normal conditions. Yacon is most commonly consumed fresh, in juices, extracts or capsules, and in cooked dishes [2]. The yacon contains FOS ()ructo 2ligo 6acarides) of low molecular weight, normally used as alternative sweeteners with small amounts of calories. It is attractive and is good for overweight and diabetic people since the human body cannot absorb the sugars it contains. The sugars are called inulin and oligofructose, and contribute to a healthy diet in line with the World Health Organization's recommendations regarding nutrition, physical activity and health [3].
Global circumstances and new directions in food consumption are seen as essential components in order to try to find opportunities of commercialization of fruits and vegetables in Latin American countries [4]. The demand of agricultural products by countries around the world has increased over the years and has shown great opportunities to agricultural products like yacon, which is part of all the potential products in developing countries such as Colombia [5]. A need for conservation of food in different countries has also open a new way of consumption and a demand for processed food [6]. The significant interest of people with purchasing power is focused on natural and processed superfoods, which can assure easy consumption for a healthy diet [7].
International markets register an increase on the demand of functional food, forcing countries around the world to increase and diversify the agricultural products [8]. The production and distribution of functional products is risky, expensive and sometimes difficult, due to the individual requirement of the global commerce [6]. A way to minimize the risk is by processing and extending the life time of the products by using different agro industrial transformation processes [9].
Drying is one of the oldest processes used in the conservation of foods with notable effects on the quality of the product [10]. Not all products have the same features in the drying process [11]. That is why it is so important both to carry out research to define the drying parameters of each product in order to minimize the negative effects on quality, and finding the drying conditions that guarantee a quality product competitive in the international market [12]. Others studies in the costumer perceptions have already gave this researchers a starting point in the objective of a product that have good acceptance in a global level of commercialization [13].

Materials
Yacon roots (Smallanthus sonchifolius) were purchased in local markets in the city of Bucaramanga (Colombia), and were selected taking into account similarity in both color and firmness. The yacons were washed and peeled, and immediately rinsed with cold water to remove the remains.
After washing the \acon it was cut into slices of 2 mm, 4 mm and 6 mm, immersed immediately in a solution of citric acid at 1.25% for 2 minutes and then put RYHU SDSHU WRZHOV GXULQJ one minute

Experimental design.
To evaluate the effect on the quality characteristics of the factors, drying temperature (levels 60,70 and 80 °C and thickness (levels 2, 4 and 6 mm), a factorial design of 3 was constructed. The measurement of the effect of the 9 treatments was carried out with three replicatHs and, to homogenize the environmental factors on the drying process, the order of the resulting 27 replications was randomized (9x3). To guarantee the homogeneRXV VDPSOHV of the product used in each replica, the initial characterization RI WKH fresh SURGXFW was measured and compared before statistical DQDO\VLV of the treatments.

Moisture content:
The moisture content was determined according to norm AOAC 2013, using the oven method, that is, to carry a sample of 100 grams of (Smallanthus sonchifolius) at 103°C +-2 to constant weight and then using equation ‫ܪ‬ = ௐ ିௐ ௐ * 100 Zhere Wi is the weight of the solid + water (total kg of water plus dry solid) before the drying process, Wf is the weight of the sample after the drying process. The moisture content was determined before the drying process LQ WULSOLFDWH, and to ascertain PRLVWXUH FRQWHQW at the end of the drying process Hstimations were carried E\ in triplicate and by combinations of the levels of the factors.

Color:
For color measurement, an H[SDQGDEOH SRO\VW\UHQH (36 made chamber was used, with an opening in the upper part to guarantee that the light input and the distance to the sample were always the same. The measurements were made with a Colorimeter Model CR400, Konica Minolta through the CIELAB system. The measurements were taken at the same time of the day to reduce variability. The colorimeter was calibrated with a white pattern. The values of the coordinates were taken; L*(darklight), a* (green -red) b * (blue -yellow) and was calculated ΔE, by means of the following equation:
2.5. Sugars content: sucrose, fructose, glucose and inverted sugar 10 milliliters of the juice of the pulp ZDV extracted E\ FRPSUHVVLRQ RI ;J RI \DFRQ SXOS in a container. The types of VXJDUV ZHUH measured by means of 4 refractometers: %UL[, fructose, glucose and inverted sugar (+DQQ LQVWUXPHQWV 96801 / +DQQ LQVWUXPHQWV 96802 / +DQQ LQVWUXPHQWV 96803 / HDQQ LQVWUXPHQWV 96804 respectively). Tests SHUFHQWDJHV were done E\ triplicate 2.6. Drying process For the drying tests, a Binder )' forced convection oven was used. For the drying loss curves, moisture loss control is carried out, weighing the samples throughout the entire process. The first hour the samples are weighed every 15 minutes, then every half hour for 2 hours and then every hour until constant weight is achieved.
During the process of drying, the color change and size of yacon slices and weight loss were recorded.

Volatile Organic Compounds (VOCs) from Yacon fresh and after a drying process Extraction by Headspace-Solid Phase Micro Extraction (HS-SPME).
The solids obtained from yacRn slices fresh and after drying were homogenized and placed in a 10ml vial. The vial was sealed by an air-tight Teflon septum, an aluminum cap and incubated at 30ºC for 15 min in a water bath. A 1-cm SPME fiber (50/30um DVB/CAR/PDMS; Supelco, Bellefonte, PA, USA) was manually inserted into the headspace of the sample vial and exposed for 15 min. The volatile compounds were thermally desorbed in the GC injection port (splitless mode) for 5 min at 250ºC.

Gas Chromatography-Mass Spectrometry (GCMS) analysis
GCMS analyses of the SPME fiber were conducted using a Shimadzu GC5050 chromatograph coupled to mass spectrometer. After desorption of volatile compounds from SPME fiber, the analyses were separated made on a HP-5 column (30m x 0,25mm i.d., 0,25 um df) (Agilent Technologies, USA) and a DB-WAX column (30m x 0,25mm i.d., 0,32 um df) (J&W Sci) according with the following temperature program: started at 50ºC, then increased to 190ºC at a rate of 6ºC/min, and then to a 240ºC at a rate of 12ºC/min, with a final hold time of 2 min. Helium was used as a carrier gas at a flow rate of 1,5ml/min, injector and detector temperature, 250ºC. MS was operated in electronic impact (EI) mode 700eV, with mass range 40-350 atomic mass units, a source temperature 280ºC, and transfer line temperature 250ºC.

Volatile compound identification
Tentative identification was attempted by comparison of the retention index (RI) and mass spectra against NIST spectrum data base using NIST MS Search 2.0 (NIST, Gaitherburg, MD, USA). Retention indices (RI) were calculated using a homologous series of C-C25 alkanes standard solution. Chemical authentic standards, when available, were analyzed under the same chromatographic conditions and their RI and mass spectra confirmed compounds identities.

Statistical analysis
The data collected on the different characteristics of interest were described by averages and medians (measures of central tendency), standard deviations (SD) and coefficients of variation (homogeneous measures CV% <10%) in the case of quantitative variables. To identify the shape of the distribution of the variables, asymmetry and kurtosis coefficients were used. Between treatments comparisons were made with the test of medians.   Yacon showed very homogeneous initial moisture values, with an average PRLVWXUH FRQWHQW, and a dispersion percentage of only 6.4% with respect to their mean. The minimum value of moisture found was 73.94% and the highest value was 93%. 7DEOH SUHVHQWV WKH SHUFHQWDJH DYHUDJH Yalues of glucose, inverted sugar, fructose and Brix sugar, with coefficients of variation higher than 20%, indicating a high dispersion of these measurements among the different analyzed roots. Despite the dispersion of the sugar measurements, the behavior of the variables is normal, according to the Shapiro-Wilk test (See Table 1). To establish the initial homogeneity of the product conditions, the initial PRLVWXUH FRQWHQW between the different treatments was compared, using the non-parametric median comparison test (see table 2).  The initial PRLVWXUH FRQWHQW did not show significant differences according to the test of the medians comparisons (Χ 2 (8) = 4.33, p = 0.836), while the medians of the PRLVWXUH FRQWHQW in the dry Yacon offered significant differences(߯ ( ଶ ୀ଼) = 12,667; ‫‬ = 0,027). As observed in table 2 the PRLVWXUH FRQWHQW of the dry Yacon at 60°C did not show differences between the diverse thicknesses compared (߯ (ୀଶ) = 0,9; ‫‬ = 0,638), likewise at 70°C; however, at 80°C the slices of 6mm presented higher PRLVWXUH FRQWHQW than the slices of 2mm and 4mm (߯ (ୀଶ) = 6,3; ‫‬ = 0,043). Note. a= median of time.

1234567890''""
Regarding drying time, significant differences were observed in the median time taken by the Yacon slices to find the equilibrium moisture, which is affirmed with a confidence level of 99% (see table 3). At 60°C half of the samples elapsed less than 14.4 hours while the other half had a longer drying time. In the case of the samples dried at 70°C, the drying time of 50% of the samples was less than 12.5 hours and, finally the samples dried at 80°C were 8 hour in the oven.
When comparing the observed drying time as a function of the temperature factor, it was found that the median drying time at 80°C is significantly lesser than the time needed to dry the product at 60°C 70°C (χ (ୀଶ) ଶ = 11,22; ‫‬ = 0,003). No significant differences were observed between the required drying time at 60°C and 70°C.
The loss of PRLVWXUH FRQWHQW of the Yacon presented greater dispersion in the slices dried at 60°C and the most representative variations were observed up to five hours after entering lthe convection oven; for drying at 70°C, the time in which Yacon shows the greatest moisture loss is four hours. Yacon slices dried at 80°C drop moisture with a higher proportion until three hours. Under the different drying temperatures used, the thickness of 2mm was always the one registering the highest moisture loss in the shortest time; the thicknesses of 4mm and 6mm under the temperature of 70°C showed well differentiated moisture loss, while the thicknesses of 4mm and 6mm at 80°C showed a very similar index of moisture loss (See figure 3).
The color of Yacon was recorded at the beginning and at the end, using coordinates L* a* b*. It was observed that the coordinates a* and b* showed greater variability (CV%) at the beginning. The lLJKWQHVV (L*) is presented as the most homogeneous characteristic recorded. The coordinates showed a symmetrical behavior with respect to the average (asymmetry less than 1 in absolute value) and a form of meso-acoustic Gaussian bell (kurtosis less than one in absolute value) as observed in table 4.

Figura.3. a.
Curves of the drying process at 60°C. b Curves of the drying process at 70°C. c Curves of the drying process at 80°C As seen in table 4, the values of the color coordinates L*, a*, b* of the fresh yacon and, after the drying process, it can be affirmed that there was a color change due to the drying process. Initially the fresh yacon is more luminous and greenish than the dry yacon.   -0,167 -0,44 Notes. L* lLJWKQHVV. a* coordinates red/green (+a indicates red, -a indicates green). b* coordinates yellow/blue (+b indicates yellow, -b indicates blue). sd standard dev. CV% percent coefficient variation.
The temperature of the drying air has a significant effect on the color characteristic of the Yacon. The parameter that had the highest sensitivity was L* (OLJKWQHVV). The value varies on average between 61 for fresh and 46.5 for dry, having a decrease of 25% represented in a darkening, as can be seen in Figure 4. With respect to the parameters a* and b* it is observed that the green color was accentuated, the values varied from 3 to 8. For the parameter b, the values varied from 21 and 28, accentuating the yellow color. To establish the change in the coloration of the product with the drying process, the median comparison test was used, finding no significant differences in the color change in the L* and b* axes (see table 5). In the coordinate a*, the color presented variations marginally significant, which implies that the magnitudes of the variations are slightly different between the treatments used (p≈0.05).
When analyzing the color changes that occurred due to the drying process, it was observed that the greatest color change was presented for the combination of 60°C and 4mm with a value ΔE = 39 followed by the combination 60°C and 6mm with a value of ΔE = 34, the lower temperature and greater thickness of the yacon slice, which leads to a longer drying time favoring browning; this is comparable with that reported by Cuervo-Andrade [14]. The smallest color change was presented for the combination of temperature parameters of 70° C and 2mm thickness.
The negative values of ΔH* for all combinations of parameters indicate that the drying process increased the yellow color in the samples analyzed, the values between -1.26 and -1.45. In the case of saturation, it is observed that for all combinations of parameters the effect of the drying process was a decrease in saturation, which is reflected in the opaque color of the yacon slices as can be seen in  Nota. ∆ = difference. ߯ ଶ = statistic Ji-square. df=degrees of freedom. p=p-hypothesis value. L*=lLJWKQHVV. a*= coordinates red/green (+a indicates red, -a indicates green). b*= coordinates yellow/blue (+b indicates yellow, -b indicates blue).
When comparing the median value of ΔE* between the different treatments, it was observed that there is no significant difference in color variation, since within each treatment the median is similar to the overall median of the set of measurements, regardless of the treatments employed, affirmed with a confidence of 95%. A similar behavior was observed in ΔC* and ΔH*, color parameters that did not show important variations between the drying treatments used. Table 6. Comparison of VOCs from Yacon fresh and after drying process at 60ºC at different thickness