Development of a dehydrated product with edible film characteristics from mammee apple ( Mammea americana L.) using Refractance Window drying

lhmeller@ufpa.br Abstract The Mammee apple ( Mammea americana L.) is a widely grown fruit, popular in tropical areas, which is mainly consumed fresh. Refractance Window (RW) drying is used to produce food films, flakes, and powders, and may be an alternative to commercially used dry food production methods. The aim of this study was to obtain a dehydrated product from the pulp of mammee apple using RW. The dehydrated product presented a homogeneous microstructure and good mechanical (TS = 1.15 Mpa, %E = 13.93) and technological properties (hygroscopicity = 21.46 g.100 g -1 , water solubility index = 32.73 ± 0.67 g.100 g -1 ). RW drying was found to be a viable and low-cost method for edible film production, generating products with adequate technological characteristics and high nutritional value. Ots dehydrated products can be consumed directly or used as ingredients in the development of food


Introduction
Brazil is characterized by having different climates and ecosystems, ranging from humid tropical, through semi-arid to temperate climates. This allows for the production of a wide variety of fruits rich in bioactive compounds, such as carotenoids and phenolic compounds (Canuto et al., 2010). Northern Brazil presents a great variety of fruit species, both native and exotic, with excellent nutritional and commercial characteristics. Some of these species have gained a prominent position in tropical fruit production, both for fresh consumption and elaboration of processed food products, both in the foreign and national market (Karam et al., 2016;Biazotto et al., 2019).
The choice of an adequate preservation method may be the key to its success, both regarding the preservation of compounds present in the food matrices and for obtaining higher food durability through water reduction. Refractance Window (RW) drying has been gaining prominence and can be used to dry fruit suspensions, pastes, and pulps, generating films, flakes, or powders with good sensory and nutritional properties as final products (Dchoa-Martínez et al., 2012). RW drying was shown to be a good alternative to traditional methods, since it generates dehydrated products of superior quality with lower processing costs, and therefore with higher added value (Putranto et al., 2010;Zotarelli et al., 2015;Minjares-Fuentes et al., 2017;Quinteros et al., 2018;Dliveira et al., 2019).
The aim of the present study was to apply RW to mammee apple pulp (Mammea americana L.) to obtain a dehydrated product of high technological and nutritional quality for use in the food industry.

Obtaining the raw material
Mammee apple (Mammea americana L.) fruits were purchased at the Ver-o-Peso Market (Belém/PA, Brazil) and transported to the Laboratory of Physical Measurements (LAMEFO) at the Federal University of Pará (UFPA) in Belém, Brazil. The fruits were washed with running water to remove dirt, sterilized with 100 ppm sodium hypochlorite for 10 min, and peeled. The pulp was divided into pieces, and placed into polyethylene bags, which were vacuum sealed, and stored in a freezer at -18 ± 2 °C until the analysis. All analyses were carried out in triplicate.
Total carotenoids, antioxidant activity, and color were determined for both the fresh pulp and the dehydrated product. Technological, mechanical, and morphological properties were determined only for the dehydrated product.

Total carotenoid content and antioxidant activity
Total carotenoid content was determined according to the method described by Rodrigues-Amaya (2001). Measurements were conducted using a spectrophotometer (NDVA instruments, NO 2000UV, Brazil), at wavelenghts between 20-500 nm, using petroleum ether as the blank. Antioxidant activity was quantified based on the ABTS radical method as described by Rufino et al. (2007).

Color analysis
Color analysis was performed using a colorimeter CR-400 Chroma Meter (KDNOCA MONDLTA, Japan) with 2° observer angle and D65 illuminant operating in the COE (L*, a*, b*) system, in which L* represents luminosity (L* = 0 black and L* = 100 white), a* defines the transition from green (-a*) to red (+a*), and b* represents the transition from blue (-b*) to yellow (+b*). The chroma index (C*) and hue angle (h*) were determined by the equipment during the measurements (Hunterlab, 2008).

Paste preparation
A paste was prepared by adding water to the mammee apple pulp, to achieve a more homogeneous distribution of the product on the drying equipment. To guarantee reproducibility of the solid concentration in all assays, a 2I:1 pulpI:water ratio (w/w) and a fixed °Brix value of 8.5 were adopted. Pastes were prepared by weighing the set amounts of pulp and water using an analytical scale (Shimadzu, AY220, Brazil), followed by homogenization using an LQ Diamante Black blender (Britania, Curitiba, PR, Brazil) for 1 min. A total 500 g of paste was used in each drying assay.

Refractance Window apparatus and drying process
Drying was performed according to Costa et al. (2019), using similar equipment. The pastes were spread on the equipment over an area of 50 × 75 cm, forming a 3-mm-thick layer, and dehydrated at 70 ± 2 °C.
After drying, the dehydrated products were stored in polyethylene packages, vacuum-sealed, and stored at ambient temperature (25 ± 2 °C) until analyzed.

Technological properties
Water Solubility Ondex (WSO) and Water Absorption Ondex (WAO)I: WSO and WAO were determined according to the methodology described by Anderson et al. (1969), with modifications. 1 g sample of the dehydrated product was weighed in previously tared centrifuge tubes. Next, 10 mL distilled water at 25 °C was added to the tubes, which were stirred in a QL-901 Vortex (BODMOXER, São Carlos, SP, Brazil) shaker to completely homogenize the samples. The tubes were then centrifuged (Thermo Electron Led GMBH, Multifuge X1R, Belo Horizonte, MG, Brazil) for 15 min at 1,046.45 g. The supernatant liquids were placed in previously tared porcelain crucibles and dried in an air circulating oven for 4 h at 105 °C. The crucibles were weighed and the WSO was found from the ratio between the weight of the evaporated residue (g) and the sample's dry weight (g). After the supernatant liquid was removed, the centrifuge tubes were weighed and the WAO was calculated from the ratio of the centrifuged residue (g) (gel.g -1 dry matter).
HygroscopicityI: was determined according to the methodology proposed by Cai & Corke (2000). Approximately 2 g of the sample was weighed in plastic capsules, which were placed in a glass desiccator containing a saturated sodium chloride solution (75% UR) at 25 °C. During the adsorption process, the samples were periodically weighed until they reached hygroscopic equilibrium. Hygroscopicity was expressed as grams of moisture adsorbed per 100 g of dry sample (g.100 g -1 ).

Mechanical properties
Tensile strength (TS) and percent elongation at break (%E) were determined using a QTS texture analyzer (Brookfield, Middleboro, MA, USA) according to the standard ASTMD 882 method (Davanço et al., 2007). The dehydrated product samples were cut into 100 mm long and 25 mm wide pieces. The initial gap between the grips and the test speed were set to 50 mm and 2 mm.s -1 , respectively.
Tensile strength was calculated by dividing the maximum force at product rupture (F M ) by the cross-sectional area (A) (Equation 1). %E was determined by dividing the final distance (d f ) of the probes by the initial distance (d i ) (50 mm) multiplied by 100 (Equation 2) (Zavareze et al., 2012).
The mean thickness (mm) required to calculate the cross-sectional area was determined using five measurements taken along the product with an Onsize 3103-25 digital micrometer (Buenos Aires, Argentina) with ± 0.002 mm precision.   (2) where TS = tensile strength (MPa); F M = maximum force at the moment of product rupture (N); A = cross-sectional area of the film (m 2 ); %E = elongation (%); d f = distance at the time of rupture (mm); d i = initial gap (mm).

Particle morphology
Particle morphology was assessed using a LED-1430 (LEOCA, Cambridge, UK) scanning electron microscope (SEM) using 80 mA and 5 kV. The samples were fixed to aluminum cylinders with double-sided tape and coated with a gold layer (Bal-Tec SCD 050).

Results and discussion
The results obtained in the analyses of mammee apple (Mammea americana L.) pulp are presented in Table 1.
The physicochemical composition of the mammee apple pulp presented similar values to those observed by Braga et al. (2010) and Péroumal et al. (2017). Based on its pH, mammee apple is classified as an acidic fruit (Chisté & Cohen, 2011).
The total RW drying time for mammee apple was 50 min, generating a dehydrated product with final moisture content of 8.32% (dry basis). This is a short drying time compared to those found for traditional drying processes such as freeze-drying, drum-drying, and pulsed vacuum drying (Baeghbali et al., 2016;Link et al., 2017).
Usually, RW rapidly decreases the final moisture content with minimum nutrient loss, since it uses temperatures lower than 100 °C (Costa et al., 2019).
The results for the dry product are presented in Table 2. The drying process retained 64% of the total carotenoid concentration of mammee apple, qualifying the dehydrated product as a source of provitamin A.   The product's hygroscopicity was lower than that observed for mango pulp dried by RW (Caparino et al., 2012). This may be related to differences in composition between the two raw materials, since hygroscopicity is related to the presence of amorphous compounds, such as sugars, in matrices. Low hygroscopicity guarantees chemical and microbiological stability during storage (Tontul et al., 2018).
TS was lower in our product than in films obtained from fruit residues (Farias et al., 2012) and higher than reported for edible starch films with added safflower (Campo et al., 2016). This low TS was explained by conformational changes in the molecular structure of the dehydrated product, formed by the plastifying effect of water used in the formulation (Cuq et al., 2011).

Particle morphology
Micrographs of the dehydrated product in film format, magnified by 100 (A) and 500 (B) times, are presented in Figure 1. Ot presented a regular shape with a regular shape with a smooth surface, and thin elongated structures related to food fibers present in the raw material.

Conclusion
The present results indicate that RW drying is a viable and promising alternative to current drying methods, resulting in the retainment of bioactive compounds, and generating a dry product that can be considered a source of provitamin A.
The dehydrated product presented good tensile strength and percent elongation at break, and a high water solubility index, favoring its application as an edible film. On addition, it presented film characteristics without the need for additives or auxiliary addition.