Addition of norbixin microcapsules obtained by spray drying in an isotonic tangerine soft drink as a natural dye

Abstract

Annatto seeds (Bixa orellana L.) are a natural source of norbixin, a carotenoid with antioxidant activity and an intense yellow–orange color which is a commonly used food and beverage colorant. However, it is susceptible to environmental factors such as light, oxygen, and temperature. Microencapsulation presents an alternative for improving the bioactive compound’s stability. In this study, norbixin microcapsules (MCN) were added to isotonic tangerine soft drinks in a quantity not exceeding food additive regulations. The final concentration was 2.86 ± 0.02 µg norbixin/mL, and according to the CIELab system, the beverage acquired the expected orange tonality. The addition of MCN favorably affects beverage stability during storage under accelerated conditions (heat and light), and the half-life time was more significant (29.71 days) than when non-encapsulated norbixin was used (393.39 min). In conclusion, MCN should be considered as an additive with potential use in processed beverage industries instead of synthetic dyes.

Introduction

One of the aspects which has the most influence on processed food’s selection and acceptance is its color. It is furthermore considered an indicator of quality, and a visual aspect frequently associated with flavor identification. In recent years, the increase of consumer concern and food legislation regarding the negative effects of synthetic dyes on human health have helped to raise interest in the development of natural dyes from natural sources (Mesnier et al. 2014; Özkan and Bilek 2014).

Norbixin is an apocarotenoid (9′-cis-6,6′-diapocarotene-6,6′-dioic acid), cis-norbixin water-soluble, and its solubility and stability properties are a result of its chemical structure (Scotter 2009). Annatto seed extract is less expensive compared with other natural pigments, and it is used as a natural dye in different categories of food (dairy products, baked goods, sausages, extruded products, salad dressings, cereals, sodas and liquors) (Cardarelli et al. 2008).

Depending on its final concentration, norbixin can confer an intense coloration in the orange–yellow–red range (Scotter 2009). Since it has no adverse toxicological effects on humans when ingested daily in minimal quantities, it is used as a food additive (Hagiwara et al. 2003). According to JECFA (2006), the estimated acceptable daily intake (ADI) is 0.6 mg/kg of body weight. In order to achieve a typical fruit coloration in non-alcoholic beverages, the following synthetic dyes are often used: Sunset Yellow (E110), Tartrazine (E102) and Amaranth (E123). These contain azo (N=N) functional groups and aromatic ring structures that are harmful to human health. The high consumption of products with synthetic dyes is related to allergy-linked diseases in those susceptible (De Andrade et al. 2014). The Brazilian National Agency for Public Health Oversight (ANVISA, in its Portuguese acronym) thus established legal regulations regarding the use of synthetic dyes and allowable levels in all food products (ANVISA 2007). Norbixin could be used an alternative to replace the use of these synthetic dyes in isotonic soft drinks.

Due to its multiple conjugated double bonds, norbixin is an amphipathic molecule with antioxidant activity (Guan and Zhong 2014). However, the chemical structure is prone to degradation when exposed to adverse factors (high temperatures, oxygen, light, low pH), and this reduces its stability and limits its application, shelf life and bioavailability in food products, due to a loss of quality during storage (Kiokias and Gordon 2003; Rodriguez-Amaya et al. 2006).

Encapsulation techniques are considered one of the most important alternatives to increase carotenoid stability. Microencapsulation is defined as a drying process in which solid microparticles, gas compounds or liquid droplets are surrounded by a coating material that forms a polymeric film around the core, acting as a protective barrier (Gharsallaoui et al. 2007). Many encapsulation processes have been developed. Nevertheless, due to its efficiency and low cost, spray-drying is the most commonly used method in food industry (Kandansamy and Somasundaram 2012).

Due to its emulsifying ability, film-forming capacity and low viscosity at high temperatures, Gum Arabic (GA) is the wall material most frequently used in the encapsulation process. This carbohydrate shows a higher encapsulation efficiency (EE) than maltodextrin (MD), which was used as coating material to encapsulate norbixin (Tupuna et al. 2018) and other bioactive compounds (Barbosa et al. 2005; Rodrigues et al. 2012; Stoll et al. 2016). GA’s primary properties are related to its chemical structure: it is a ramified carbohydrate chain with a small amount of glycoproteins that interact with hydrophilic and hydrophobic sections of the molecules, making GA into an excellent emulsifying agent compatible with other polymers. It has an excellent film-forming effect, which allows it to reach a better EE (Krishnan et al. 2005). The large amount of norbixin retained inside the microcapsules results in good antioxidant activity, that is, activity which is directly correlated with EE. Several studies showed good antioxidant activity when GA was used at a higher ratio than MD (Stoll et al. 2016; Tonon et al. 2010; Tupuna et al. 2018).

The aim of this study was to assess the stability of an isotonic tangerine soft drink with norbixin microcapsules added as natural dye during storage to analyze degradation kinetics when beverages were exposed to accelerated light and heat conditions for an extended period of time. This was done to define estimated shelf-life for future applications of norbixin as a natural additive in beverages.

Materials and methods

Materials

The annatto seeds were purchased at a local market in Porto Alegre, RS, Brazil. As an encapsulation carrier agent, an edible polysaccharide, Gum Arabic SYNTH (Porto Alegre, RS, Brazil), was used. All reactants were analytical grade.

Preparation of the norbixin crystals

To obtain a purified norbixin extract, the procedure described by Rios and Mercadante (2004) was followed. The annatto seeds were first washed with hexane and methanol in order to remove impurities (lipids and polar compounds, respectively). The seeds were then separated by filtration and an extraction of bixin with ethyl acetate was performed; the extract was dried in a M802 rotary evaporator (Fisatom, São Paulo, Brazil). Dichloromethane was then used to dissolve the extract, and absolute ethanol was immediately added to form bixin crystals. Finally, saponification of bixin crystals was performed using a methanolic KOH solution (10%). The norbixin salt obtained was separated through a separation funnel, and the norbixin crystals (NC) were dried in a rotary evaporator and stored at − 18 °C.

Norbixin quantification

A concentration of norbixin was quantified by spectrophotometry using a UV SPECTROPHOTOMETER UV-1800 (Shimadzu, Tokyo, Japan) at 453 nm and absorptivity value E^1%_1cm = 3473 (Tupuna et al. 2018). First, norbixin crystals were dissolved in ethanol (99.7%), and an aliquot was diluted in a KOH (0.5%) aqueous solution, used as a standard. It was not necessary to perform the first dissolution in ethanol to measure norbixin in isotonic beverages. An aliquot of each beverage sample was taken and diluted in KOH (0.5%). A standard curve was developed for norbixin quantification with a correlation factor (R²) of 0.99. Each calculation was done using the following equation:

$$x_{\left(\mathrm{\mu }\text{g}\right)}\frac{Abs\times \gamma \times {10}^6}{E\times 100}$$

where Abs absorbance; γ dilution (mL KOH 0.5%); E absorptivity coefficient

Microencapsulation by spray-drying

A formulation was prepared based on the conditions which achieved the best norbixin encapsulation efficiency in the study carried out by Tupuna et al. (2018). The solution created by dissolving GA in 80 mL of distilled water until a level of 40% total soluble solids (TSS) was reached, and 30 mg NC diluted in 20 mL ethanol (99.7%) was added. It was then magnetically stirred at 40 °C for 30 min until a homogeneous emulsion was achieved, which was stored at − 4 °C.

The drying process was performed in a B-290 Mini Spray Dryer (Büchi, Flawil, Switzerland). A 6.6 mL/min feed flow and an inlet temperature of 150 °C were used as operation parameters for the equipment, and the experiments were carried out in duplicate. The MCN were immediately vacuum packed in plastic bags and stored at − 4 °C.

The MCN showed a low moisture value (3.00 ± 0.50%), monomodal particle size distribution (span = 1.96 ± 0.25) and volume-weighted mean diameter (D_(4.3) = 8.25 ± 0.83 µm). An ABTS test was also performed, and MCN exhibited antioxidant activity (87.65 ± 0.46 µmol TE/g dry sample). MCN had a high water solubility index (92.04 ± 0.75%) and high microencapsulation efficiency (70.97 ± 1.48%–386.25 ± 0.18 µg/g microcapsules) as a result of the successful encapsulation process.

Preparation of the simulated isotonic tangerine soft drink (ITSD)

The colorimetry of a commercial isotonic tangerine soft drink (ITSD) was analyzed using a Chromameter CR-400 colorimeter (Konica Minolta, Tokyo, Japan) in order to define the quantity of norbixin to be added in ITSD to obtain a beverage with similar CIELab color parameter values (L* = 48.27 ± 0.01, a* = 6.75 ± 0.04, b* = 22.76 ± 0.01), without exceeding the limit established for norbixin as an additive in non-alcoholic beverages by the Brazilian food regulation for norbixin E160b (0.005 g/100 mL) (ANVISA 2007). Following several preliminary experiments, the norbixin concentration which was found to achieve the desired color tonality (orange) according to the CIELab color chart was in the 2.50 to 2.90 µg/mL range. We have thus aimed for a beverage with norbixin concentration in this range. Two ITSD working stock solutions were then prepared, each with 1250 mL of distilled water containing 50 g of crystalline sugar, 875 mg of KCl, 1125 mg of NaCl, 375 mg of citric acid, 375 mg of ascorbic acid and 375 µL of artificial tangerine flavor. In one of these, 10 g of MCN were dissolved until a beverage with a final norbixin concentration of 2.86 ± 0.02 µg/mL was achieved. The goal was to have an ITSD colored with MCN (Isotonic-MCN). For control, 650 µg of NC was added to the other stock solution until a beverage with a final norbixin concentration of 2.56 ± 0.05 µg/mL was achieved. This was an ITSD colored with NC (Isotonic-NC). Finally, all solutions were stored in the dark at − 4 °C.

Storage stability studies

Acidity, pH, total soluble solids (°Brix)

Immediately after preparation, the Isotonic-MCN and Isotonic-NC beverages were chemically characterized to obtain the initial value of acidity, pH, °Brix. During storage, a sample of each beverage was periodically taken and analyzed. Acidity was determined using standard procedures and was expressed as mg citric acid/L of sample. The TSS was measured using a PAL-3 refractometer (ATAGO, Tokyo, Japan) and was expressed in °Brix. A pH meter (QUIMIS, São Paulo, Brazil) was used to evaluate pH.

Colorimetric analysis

The color change of the beverages was monitored using a Chromameter CR-400 colorimeter (Konica Minolta, Tokyo, Japan). Measurements were recorded according to the CIELAB system with the parameters L*, a*, b*; where L* is the variation from lightness (0 = white) to darkness (100 = black); a* the variation between green (< a*) and red (> a*); and b* the variation between blue (< b*) and yellow (> b*). A white ceramic disc was used for calibration prior to carrying out measurements. These values were then used to calculate the Chroma value (C*), which indicates color saturation vividness, Hue angle system (h°_ab), which implies the color tone of the sample (0 or 360 = red, 90 = yellow, 180 = green and 270 = blue), and the total color difference (ΔE) throughout storage, as follows (Kim et al. 2002):

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

$$\mathrm{\Delta }E={\left[{\left(L{o}^{\ast }-L^{\ast }\right)}^2+{\left(a{o}^{\ast }-a^{\ast }\right)}^2+{\left(b{o}^{\ast }-b^{\ast }\right)}^2\right]}^{1/2}$$

$$h_{\mathit{ab}}^{\circ }={tan}^{-1}\frac{b^{\ast }}{a^{\ast }}$$

Beverage stability in accelerated storage conditions

Trials were conducted to assess the loss of color in beverage samples submitted to accelerated storage conditions (temperatures from 30 to 60 °C and exposure to light) during 8 h at ambient moisture, in order to establish temperature and period of experimental assays. The results showed that norbixin was more stable in MCN than when tested in free form as an NC. A higher temperature and a higher storage time were thus established for experiments with Isotonic-MCN samples. In the case of non-encapsulated systems, Isotonic-NC samples were used. The beverages were distributed in transparent plastic bottles with screw caps, 30 mL in each, and were immediately flushed with nitrogen, closed, and sealed with plastic paraffin film. Two sets of experiments were carried out at a specific temperature in a laboratory oven in order to assess the wall material’s influence on light stability of norbixin. For Isotonic-NC, the samples were stored under fluorescent light (ca. 750 lx) or in the dark at 35 ± 2 °C, and the norbixin concentration, color, °Brix, acidity and pH in samples were measured at 0, 30, 60, 180, 240, 300, 420 and 500 min during storage. For Isotonic-MCN, the samples were stored under fluorescent light (ca. 750 lx) or in the dark at 40 ± 2 °C, and the same parameters in samples were measured at 0, 1, 3, 5, 7, 10, 12, and 15 days during storage. All experiments were performed in triplicate.

Norbixin’s degradation kinetics were analyzed using the first-order reaction as has previously been reported by Scotter (2009) and Tupuna et al. (2018). For Isotonic-NC and Isotonic-MCN, an aliquot of each beverage sample was taken at 0, 30, 60, 180, 240, 300, 420 and 500 min, and 0, 1, 3, 5, 7, 10, 12, and 15 days during storage, respectively, and the norbixin retention (%) was quantified following the procedure described in Sect. 2.3. To calculate kinetic parameters, norbixin retention (%) was plotted against time. The degradation rate constant (k_t) was determined from the first derivative of the curves plotted with the Origin 8.0 (Origin Lab Co., MA, USA) software (De Rosso and Mercadante 2007). Half-life time (t_1/2) was calculated using the equation (Bustos-Garza et al. 2013):

$$\left[\mathit{norbixin}\right]={\left[\mathit{norbixin}\right]}_0\times ex{p}^{\left(k_t\times t\right)}$$

$$t_{1/2}=\frac{ln2}{k_t}$$

where t_1/2 half-life time value, k_t kinetic constant, t time

Statistical analysis

Statistical parameters were calculated using STATGRAPHICS Centurion XVI software (Statgraphics, Virginia, USA). One-way variance analyses (ANOVA) and a multiple-range test were performed to correlate the data and establish statistically significant differences (p < 0.05) with a 95% confidence interval. All results were recorded and expressed as mean value ± standard deviation.

Results and discussion

Chemical characterization of beverages during storage

The initial chemical parameter measurements for working solutions of beverages were taken as the initial values for storage studies. A comparison was made with final values to determine significant variations over time. Beverages were analyzed for acidity, pH and TSS (Table 1).

Table 1.

Measurements of acidity, pH and TSS for isotonic tangerine soft drinks added of non-encapsulated norbixin and norbixin microcapsules during storage under accelerated conditions of light and temperature

Time (min)	pH		Total soluble solids (°Brix)		Acidity (mg citric acid/L)
	Dark	Light	Dark	Light	Dark	Light
Isotonic-NC
0	3.21 ± 0.01aA	3.21 ± 0.01aA	3.60 ± 0.01aA	3.60 ± 0.01aA	0.12 ± 0.01aA	0.12 ± 0.01aA
30	3.20 ± 0.01aA	3.19 ± 0.01bB	3.63 ± 0.06aA	3.83 ± 0.06cdB	0.14 ± 0.01abA	0.16 ± 0.02bA
60	3.18 ± 0.01bA	3.17 ± 0.01cA	3.77 ± 0.06bcA	3.70 ± 0.03abB	0.15 ± 0.01bA	0.17 ± 0.02bcA
180	3.10 ± 0.02efA	3.11 ± 0.01dA	3.73 ± 0.06bA	3.73 ± 0.15bcA	0.15 ± 0.01bA	0.17 ± 0.02bcA
240	3.13 ± 0.01cA	3.11 ± 0.01dB	3.83 ± 0.06cdA	3.97 ± 0.06eB	0.16 ± 0.02bA	0.17 ± 0.03bcA
300	3.12 ± 0.01cdA	3.11 ± 0.01dA	3.90 ± 0.02dA	3.90 ± 0.02deA	0.16 ± 0.01bA	0.18 ± 0.01bcB
420	3.11 ± 0.02deA	3.10 ± 0.01dA	4.00 ± 0.02dA	3.90 ± 0.01deA	0.17 ± 0.03bcA	0.18 ± 0.02bcA
500	3.09 ± 0.02fA	3.09 ± 0.01eA	3.90 ± 0.01dA	3.90 ± 0.03deA	0.19 ± 0.02cA	0.19 ± 0.03cA
Isotonic-MCN
0	3.89 ± 0.01aA	3.89 ± 0.01aA	5.17 ± 0.06aA	5.17 ± 0.06aA	0.20 ± 0.03aA	0.20 ± 0.03abA
1	3.89 ± 0.01aA	3.85 ± 0.03bB	5.73 ± 0.12eA	5.77 ± 0.12dA	0.21 ± 0.02abA	0.20 ± 0.02aA
3	3.82 ± 0.01bA	3.77 ± 0.01cB	5.23 ± 0.06abA	5.37 ± 0.06bA	0.22 ± 0.01abcA	0.22 ± 0.03abcA
5	3.83 ± 0.01bA	3.83 ± 0.03bA	5.57 ± 0.06deA	5.50 ± 0.04cA	0.24 ± 0.03bcA	0.24 ± 0.02bcA
7	3.74 ± 0.02dA	3.75 ± 0.02deA	5.37 ± 0.06bcA	5.40 ± 0.13bB	0.24 ± 0.04bcA	0.24 ± 0.01cB
9	3.78 ± 0.01cA	3.76 ± 0.01cdA	5.50 ± 0.26cdA	5.10 ± 0.08aB	0.24 ± 0.01dA	0.24 ± 0.02cA
12	3.75 ± 0.02dA	3.74 ± 0.01deA	5.33 ± 0.06abcA	5.43 ± 0.06bA	0.24 ± 0.03dA	0.24 ± 0.05bcB
15	3.74 ± 0.02dA	3.73 ± 0.02eA	5.63 ± 0.06bdeA	5.70 ± 0.01dA	0.24 ± 0.02dA	0.24 ± 0.03cA

Mean ± standard deviation (n = 3)

Isotonic-NC isotonic tangerine soft drink with norbixin crystals, Isotonic-MCN isotonic tangerine soft drink with norbixin microcapsules

Different lowercase letters in the same column indicate statistically significant differences between samples at p < 0.05 and different capital letters in the same file indicate statistically significant differences between dark or light measurements for each parameter at p < 0.05

During storage, these parameters showed different behavior in the presence or absence of light in each beverage (Table 1). TSS and acidity were highest for Isotonic-MCN due to the presence of GA. In beverage solutions with both NC and MCN as a natural dye, a statistically significant variation (p < 0.05) between initial and final values can be observed in all parameters during storage time under accelerated temperature and light exposure conditions. However, when the final value was compared with the final value in the dark, the influence of light did not show significant differences (p > 0.05) (Table 1). There was a minimal decrease in pH in all experiments, indicating that norbixin can remain stable in acidic beverages during storage. This agrees with Prabhakara Rao et al. (2002) and Guan and Zhong (2014) who reported the same tendency when norbixin was added to orange juice and to acidified aqueous norbixin solutions, respectively. When the pH stability of astaxanthin encapsulated with different wall materials was evaluated, Bustos-Garza et al. (2013) reported high t_1/2 at pH 3.0 and 4.0 when only GA was used as a coating material.

Beverage stability in accelerated storage conditions

Color changes during storage

The data for the colorimetry analyses of Isotonic-NC and Isotonic-MCN samples at time zero were located in the first quadrant of the CIELab color chart. The beverages showed values from 5.73 to 6.15 and 20.48 to 24.58 for a* and b*, respectively (Fig. 1). The h°_ab values ranged from 74.36 to 75.96 (Table 2) and demonstrated that it was possible to achieve beverages with the typical orange coloration of an ITSD using norbixin as a natural dye. Prabhakara Rao et al. (2002) reported the same tendency when norbixin was used in ready-to-serve orange juice, and the desired orange–yellow color was achieved at very low concentrations. In contrast, when the stability of astaxanthin incorporated as a natural food dye in a model instant drink was evaluated, Villalobos-Castillejos et al. (2013) reported that the carotenoid changed from intense orange to pink when dispersed in water. This behavior confirms that norbixin did not show coloration changes when dissolved in aqueous solution, as was the case with other carotenoids.

Fig. 1.

Changes in the color parameters for L* (Lightness), a* (redness) and b* (yellowness): a ITSD samples prepared with norbixin crystals (Isotonic-NC) during storage at 35 °C, b ITSD samples prepared with norbixin microcapsules (Isotonic-MCN) during storage at 40 °C. (*) Statistically significant differences (p < 0.05) between results for light and darkness

Table 2.

Color parameters Chroma, Hue, ΔE for isotonic tangerine soft drinks added of non-encapsulated norbixin and norbixin microcapsules during storage under accelerated conditions of light and temperature

Time (min)	Chroma		Hue		ΔE
	Dark	Light	Dark	Light	Dark	Light
Isotonic-NC
0	21.27 ± 0.01a	21.27 ± 0.01a	74.36 ± 0.01b	74.36 ± 0.01abc	0.00	0.00
30	20.68 ± 0.38cd	20.63 ± 0.19a	73.93 ± 0.46a	73.72 ± 0.27a	1.29 ± 0.15	1.60 ± 0.51
60	21.11 ± 0.25ab	20.73 ± 0.57a	76.22 ± 0.32c	75.32 ± 0.58bcd	15.19 ± 1.00	14.73 ± 2.18
180	21.06 ± 0.05ab	20.77 ± 0.57a	76.69 ± 0.08d	76.05 ± 0.51d	15.70 ± 0.39	16.58 ± 0.30
240	20.77 ± 0.08cd	18.62 ± 0.54b	76.69 ± 0.19d	75.47 ± 0.19cd	15.90 ± 0.41	17.52 ± 1.00
300	21.17 ± 0.04a	17.60 ± 0.99bc	76.56 ± 0.20cd	74.42 ± 1.41abcd	15.99 ± 0.11	19.39 ± 0.79
420	20.82 ± 0.08bc	16.69 ± 1.06c	76.79 ± 0.09d	75.07 ± 0.93bc	16.59 ± 0.17	21.10 ± 0.78
500	20.46 ± 0.29d	15.29 ± 0.54d	76.74 ± 0.25d	74.16 ± 0.65ab	19.47 ± 0.93	24.01 ± 1.81
Isotonic-MCN
0	25.34 ± 0.26a	25.34 ± 0.26a	75.96 ± 0.16c	75.96 ± 0.16b	0.00	0.00
1	24.88 ± 0.06ab	24.32 ± 0.14b	75.78 ± 0.13de	76.79 ± 0.08de	0.62 ± 0.18	1.19 ± 0.25
3	23.81 ± 0.09cd	23.60 ± 0.12c	75.78 ± 0.07de	77.05 ± 0.19e	1.57 ± 0.25	1.87 ± 0.12
5	23.50 ± 0.65cd	23.78 ± 0.11c	75.36 ± 0.23de	76.55 ± 0.07cd	1.88 ± 0.97	1.68 ± 0.28
7	24.01 ± 0.36bc	23.09 ± 0.13c	76.37 ± 0.21c	76.96 ± 0.27de	1.40 ± 0.52	2.51 ± 0.09
9	23.09 ± 1.16d	22.02 ± 0.27d	76.50 ± 0.59b	77.14 ± 0.54e	2.31 ± 1.39	3.40 ± 0.35
12	24.04 ± 0.41bc	20.14 ± 0.39e	76.55 ± 0.06b	76.12 ± 0.13bc	1.38 ± 0.25	5.22 ± 0.46
15	23.17 ± 0.51cd	17.73 ± 0.11f	77.03 ± 0.27a	75.40 ± 0.39a	2.25 ± 0.51	7.63 ± 0.29

Mean ± standard deviation (n = 3)

Isotonic-NC isotonic tangerine soft drink with norbixin crystals, Isotonic-MCN isotonic tangerine soft drink with norbixin microcapsules)

Different letters in the same column indicate statistically significant differences between samples at p < 0.05

Color stability was determined for norbixin in Isotonic-NC, stored at 35 °C in the light or the dark. For both conditions, significant differences (p < 0.05) were observed between the initial and final values of the CIELab parameters. An increase of L* occurs, and a*, b* showed extensive color fading (Fig. 1a). Nevertheless, in the absence of light, the norbixin color was more stable throughout storage.

An increase in the L* parameter leads to color loss which causes a lightness in beverage tonality, while a decrease in a* and b* parameters indicates a color change from orange to yellow due to redness loss. For experiments with Isotonic-NC samples, the L* and a* parameters showed the greatest change after 60 min of storage, and remain stable for longer (Fig. 1a). In the presence of light, color changes were more pronounced, above all for parameter b*; this means that light was more responsible than temperature for color variation in Isotonic-NC. Other authors also report that the higher decrease of parameter b* which occurs during the degradation of carotenoids is of the greatest to stability studies (Mesnier et al. 2014), because it means a loss of yellowness that causes a change in desired beverage shade.

The results obtained for Isotonic-MCN samples stored at 40 °C showed that encapsulated norbixin may be more stable during storage (Fig. 1b) in the absence of light; although there were significant differences (p < 0.05) for parameters a* and b*, these presented little variation (Fig. 1b). Parameter L* did not show significant changes, which may be due to the presence of GA in MCN. This polysaccharide is commonly used as an emulsifier and thickening agent in soft drinks with the goal of producing cloudy products and improving shelf-life. L* values showed the same behavior when the addition of GA to acidified aqueous solutions of norbixin to improve thermal and acid stabilities was evaluated (Guan and Zhong 2014).

A high decrease for the a* and b* parameters was observed during a storage time of 9 days in the presence of light. The loss of color was therefore greater (Fig. 1b) and a variation from orange (less redness) to yellow (more yellowness) occurred. The MCN were then able to withstand high temperatures but were unable to endure exposure to accelerated light conditions for an extended period of time. Light exposure resulted in the same variation in CIELab parameters during the storage of bixin (De Marco et al. 2013) and norbixin (Tupuna et al. 2018) microcapsules.

The ΔE value denotes a difference in color between the two samples: the 0 to 0.5 range signifies an imperceptible difference in color; 0.5 to 1.5 a slight difference; 1.5 to 3.0 a just-noticeable difference; 3.0 to 6.0 a notable difference; 6.0 to 12.0 an extremely notable difference; and above 12.0 an entirely different shade (Kim et al. 2002). High ΔE thus implies a significant color variation. For Isotonic-NC, the difference in ΔE values between samples at zero time and at 30 min of storage were less than 3.0, but after 60 min there was a high variation of 15.19 and 14.73 in dark or light, respectively (Table 2). The samples showed a different color shade from this point on. As was expected for light exposure, the final ΔE value was higher; this indicates a significant loss of color quality. These results demonstrate that decoloring as a result of light exposure progressed rapidly from 30 min of storage onwards. Otherwise, the difference in ΔE values for Isotonic-MCN samples was less than 3.0 during the entire storage time in the absence of light, while for samples stored under light exhibited a notable difference from 9 days on, and an extremely notable difference after 15 days (Table 2). Mesnier et al. (2014) reported that ΔE > 2 is used as a threshold value to indicate that the shade has begun to change, indicating a visual loss of color. A difference of less than this threshold value is considered unnoticeable to the naked eye. These findings thus show that the conditions of storage light play an important role in beverage color stability. The coating material also offers better protection from decoloring at a higher storage temperature, an evaluation based on the comparatively smaller differences in color occurring during the storage period.

Several characteristics that allow color stability to be assessed, such as stronger saturation, more vivid color, and less yellowness, are considered important to maintain the quality of tangerine beverages; and thus the interest in acheving acceptable C* and h°_ab values according to the CIELAB color chart for desired orange tonality, along with low ΔE*. There were significant differences (p < 0.05) of C* and h°_ab values during storage of Isotonic-NC and Isotonic-MCN (Table 2), but there was a slight decrease in C* when samples of both beverages were stored in the absence of light, and only a slight h°_ab increase, meaning a change from orange to yellow. In contrast, when both were exposed to light there was a high saturation loss. Final samples thus showed a different color shade. C* and h°_ab depend on a* and b*. It thus could be determined that light influences the degradation of these parameters.

Kinetic behavior during storage

Accelerated storage conditions (heat and light) result in a compound loss over time, and this effect was observed for both Isotonic-NC and Isotonic-MCN (Fig. 2). As was expected, during all kinetic reactions the behavior of norbixin thermal degradation was first-order. Kinetic degradation constants (k_t) were calculated using an exponential regression fit for Isotonic-NC and Isotonic-MCN in the dark and light exposure with correlation factors (0.94 > R²> 0.98). The fast decay of norbixin retention (%) in Isotonic-MCN (Fig. 2b) at the beginning of the heating test may be due to degradation of the non-encapsulated compound located on the surface of microcapsules, which is consequently exposed to direct heat and light (Barbosa et al. 2005; de Lobato et al. 2015; Tupuna et al. 2018).

Fig. 2.

Norbixin retention (%) during storage in accelerated conditions a ITSD samples prepared with norbixin crystals (Isotonic-NC) at 35 °C, b ITSD samples prepared with norbixin microcapsules (Isotonic-MCN) at 40 °C

For each storage condition, the t_1/2 parameter was calculated with the k_t value in order to predict the time required for norbixin to decrease by 50% from its initial measurement. As can be observed, t_1/2 depends on light exposure, and for all experiments, t_1/2 was higher in the dark (Table 3). In the presence of light, t_1/2 was lower due to the use of accelerated treatments that cause faster degradation of compounds.

Table 3.

Rate kinetic parameters: degradation constant (k_t) and half-life time (t_1/2) for non-encapsulated system (Isotonic-NC) and isotonic tangerine soft drink with norbixin microcapsules (Isotonic-MCN)

Sample	Isotonic-NC			Isotonic-MCN
Conditions	Dark	Light		Dark	Light
Rate parameters
kt × 10−3 (min−1)	1.76	2.90	kt × 10−3 (day−1)	23.33	74.64
t1/2 (min)	393.39	238.93	t1/2 (days)	29.71	9.27

Isotonic-NC isotonic tangerine soft drink with norbixin crystals, Isotonic-MCN isotonic tangerine soft drink with norbixin microcapsules, k_t kinetic constant, t_1/2 half-life time

When Isotonic-NC samples norbixin stability was evaluated in the absence of light, the result of t_1/2 was greater than the t_1/2 value in light. This means that the norbixin retention decrease was influenced by light exposure, and that ITSD colorants with free norbixin are unable to maintain their quality for an extended time, even in the dark. Approximately 3 h were sufficient for a significant norbixin loss (Fig. 2a).

For Isotonic-MCN samples, the result of t_1/2 was far higher in comparison with Isotonic-NC (Table 3), thus confirming that microencapsulation using GA as wall material is an alternative to enhance bioactive compounds’ shelf-life. Non-encapsulated norbixin is more sensitive to degradation. Norbixin loss from Isotonic-NC therefore occurs more rapidly than from Isotonic-MCN, showing that the encapsulated compound is more resistant to high temperatures. When the use of GA as a coating material using spray-drying was compared with other polymers (maltodextrin, whey protein isolate, modified starch or their blends) t_1/2 increases and the shelf-life of bioactive compounds is significantly increased (Al-Ismail et al. 2016; Kanakdande et al. 2007). Good emulsifying properties, and these carbohydrates’ ability to form a film around the compound, are responsible for good retention and protection in accelerated storage conditions; thus providing a plasticity that keeps the protection matrix from cracking (Krishnan et al. 2005). GA’s chemical structure has arabinogalactan protein (AGP) molecules, which give the polymer its water-like emulsifying characteristics and its film-forming capability (Castellani et al. 2010). Barbosa et al. (2005) reported enhanced bixin stability when encapsulated with GA compared to other polymers, and greater stability was noted between microcapsules stored with light exposure than those stored in the dark. Stability was higher for encapsulated bixin compared to free bixin. Cano-Higuita et al. (2015) studied curcumin stability during storage at 25 °C under light and reported that encapsulating matrix containing GA led to a high curcumin retention when spray-drying was used as the drying method.

The GA encapsulation process thus resulted in enhanced norbixin stability at the studied heat and light conditions in a food matrix during storage. This agrees with reports that as a coating material, GA can effectively improve the stability of bioactive compounds such as carotenoids (Barbosa et al. 2005; Bustos-Garza et al. 2013; Cano-Higuita et al. 2015; De Marco et al. 2013; Tupuna et al. 2018), anthocyanins (Burin et al. 2011; Chung et al. 2016) and essential oils (Kanakdande et al. 2007).

This study thus demonstrates that norbixin is more stable during storage in accelerated conditions when encapsulated than when exposed in free form. Losses were faster in Isotonic-NC (hours) in comparison with Isotonic-MCN, which needed more time (days) and higher temperatures (Table 3).

Accelerated treatments with heat and light exposure are commonly conducted in stability studies for bioactive compounds in order to perform an estimate of shelf-life. The stability assessment of MCN in a model aqueous system at high temperatures (60, 90 and 98 °C) was studied by Tupuna et al. (2018). The authors reported that norbixin is more stable when encapsulated than when exposed to heat in free form. Gallardo-Cabrera and Rojas-Barahona (2015) carried out a stability study of an aqueous solution of norbixin during storage at 30 °C with light exposure and reported the highest shelf-life. Nonetheless, the results obtained in this research could not be directly compared with our study results, because the authors used different concentrations and temperatures, and samples were stored in containers without direct light exposure.

Isotonic-MCN samples stored without light need approximately three times as many storage days to achieve a norbixin loss of 50% when compared to samples stored while exposed to light (Table 3). The samples showed 70% norbixin retention up to 15 storage days (Fig. 3). Samples stored in light exhibited high degradation prior to 10 storage days (Fig. 2b) in conjunction with color loss (Fig. 1b). Light had a more significant effect on norbixin stability was more significant than heating. Parvin et al. (2011) and De Marco et al. (2013), studied stability during storage of norbixin extracts and bixin microcapsules, respectively. The authors reported a higher t_1/2 in the dark than when exposed to light at room temperature in the model aqueous system.

Fig. 3.

Samples of isotonic tangerine soft drink with norbixin microcapsules stored in the dark (a) and when exposed to light (b)

Due to their chemical structure, carotenoids are prone to isomerization and oxidation during food processing and storage. The main cause of carotenoid loss is enzymatic or non-enzymatic oxidation, which depends on oxygen availability, carotenoid structure and food matrix type. This is furthered by light, heat, acids, metals, enzymes, and peroxides. The length of storage time in severe conditions (high temperature, moisture, and light), as well as the use of oxygen and light permeable packing, increases degradation and decreases the shelf-life of processed foods (Rodriguez-Amaya et al. 2006).

The addition of ascorbic acid (AA) may negatively influences norbixin degradation. AA’s impact on carotenoid light and heat stability is beneficial only at high concentrations; at a low concentration, a pro-oxidant effect may occur (Mesnier et al. 2014). Enrichment of beverages with AA could promote the isomerization of carotenoids and contact with organic acids during deterioration (Meléndez-Martínez et al. 2009). To date, only anthocyanins have been applied as a natural dye in isotonic soft drink systems, and color and stability during storage with light exposure or in the dark were evaluated. As expected, light had a harmful effect in all experiments. When the stability of anthocyanins as a natural dye in model beverages was assessed, the authors reported that in the presence of AA, the anthocyanin stability was reduced and extensive color loss occurred during storage due to the hydrogen peroxide formed through ascorbic acid oxidation. It was possible to improve stability with the addition of low levels of GA. This result was attributed to the interaction between anthocyanin and GA through hydrogen bonding. The anthocyanin thus achieves a barrier against AA condensation or oxidation by hydrogen peroxide during storage (Chung et al. 2016). Based on the study results, it could be suggested that a higher GA concentration may increase the beneficial interaction between AA and GA, causing a decrease in AA’s pro-oxidant effect; thus the enhanced light stability of norbixin when applied as a natural dye in beverage systems.

Conclusion

In this study, through the application of MCN obtained by spray-drying with GA, it was possible to achieve the desired orange color for an ISTD, as well as to determine the kinetic rates of degradation during storage. The t_1/2 required for degradation of encapsulated norbixin was extremely high (29.71 days at 40 °C) compared with free norbixin (3.98 h at 35 °C) under heat. The results show that light exposure results in a higher norbixin retention decrease in beverages compared with storage in the dark. The addition of AA promotes norbixin deterioration. The addition of GA could, however, prevent the pro-oxidation effect. It was therefore proven that the encapsulation process is an excellent alternative to enhance norbixin stability in the aqueous model system and thus extend shelf-life. Overall, this study validated the potential use of MCN as a natural dye for application in food matrices, and particularly in beverages, due to their hydro-solubility. The food and beverage industry can use this information to develop products with natural rather than synthetic dyes, thereby contributing to human health.

Author contributions

D.S. Tupuna-Yerovi designed the study, carried out the experiments, interpreted, discussed the results and wrote the manuscript. K. Paese handled the spray drying and particle size analyses equipment. S. Guterres provided equipment and revised the manuscript. S. Flôres supported and interpreted the statistical analyses. A. Rios assisted all experiments of extraction, encapsulation, characterization, evaluation and helped in results interpretation.

Compliance with ethical standards

Conflict of interest

All authors declare that they have no conflict of interest.