Valorization of Hibiscus Flower (Hibiscus sabdariffa L.) Anthocyanins to Produce Sustainable Spray-Dried Ingredients

Abstract

The recent increase in sustainability awareness has triggered the industry to establish novel, eco-friendly sources of plant-based ingredients. In the present study, hibiscus flowers (Hibiscus sabdariffa L.) were investigated as a sustainable source of anthocyanins for use in spray-dried ingredients with antioxidant capacity. To this end, the extraction and spray-drying microencapsulation of hibiscus flower anthocyanins were optimized and the final products were evaluated for their oxidative stability index. Initially, preliminary experiments were carried out to evaluate the effects of selected processing parameters on anthocyanin extraction. Next, the extraction was optimized through a 2² central composite design, considering ethanol concentration (44–56% v/v) and extraction time (1.6–6.4 h) as independent factors. The optimum extraction conditions (8 h, 41.6% v/v ethanol concentration) were used to produce anthocyanin-rich extracts, which were microencapsulated by spray drying using a 2² central composite design with the carrier addition rate (1–3% w/v) and inlet temperature (160–192 °C) as factors. Maximum values of solids recovery (60.8%) and anthocyanin retention (96.0%) were reached when 3.2% w/v of starch–alginate carrier blend and a 170.7 °C inlet temperature were used. Finally, when hibiscus microcapsules were added to soybean oil, higher oxidative stability was achieved compared to the control. Overall, we demonstrate an industrially friendly and scalable approach that takes advantage of abundant hibiscus flowers as a viable source of anthocyanins for multiple applications.

1. Introduction

In recent years, justified concerns about the environmental impact caused by the food industry has motivated an unprecedent shift in dietary patterns and in the food market, with consumers increasingly attentive to the consequences of their food choices [1]. This new world centered on sustainability creates a challenge for the food industry, which must deliver clean label ingredients from natural, renewable sources, but without compromising the quality and performance of these reinvented products.

Anthocyanins are natural pigments that are widespread in nature and concentrated in fruits and flowers. Because of their potent bioactivity, they have attracted increasing interest from both the pharmaceutical and food industries [2,3,4]. The use of anthocyanins as food colorants has been extensively investigated, but their instability to pH, temperature and other environmental factors represents a technical challenge that has yet to be addressed before their well-established use as food additives [5,6]. Indeed, color degradation results in substantial product-quality deterioration because when red colors fade, they give way to brownish tones that negatively interfere with consumer acceptance.

Hibiscus flowers (Hibiscus sabdariffa L.), also known as karkade, roselle, sorrel, guinea sorrel, and viña, are the flowers of an anthocyanin-rich shrub that is native to Africa and cultivated in tropical and subtropical regions [7]. Because of their composition, hibiscus flower extracts can be used to produce natural, phytochemical-rich products at low cost [2]. However, hibiscus flowers have a short lifespan, and their marketable utilization depends on technological solutions to deliver products with easier storage and handling, as well as preserved and stabilized bioactivity [8]. If they are made available in a stable and convenient format, anthocyanin-rich hibiscus extracts can become a dependable source of pigments and health-relevant molecules for multiple applications. To take full advantage of anthocyanins and enable their incorporation into a variety of foods, nutraceuticals and herbal medicines, the first step is to obtain preserved extracts that were produced in an efficient and timely manner [6]. Moreover, scalable, straightforward and sustainable operations that lead to optimized extraction yields are necessary [9].

Due to its cost-effectiveness and short processing time and to equipment availability, spray drying is now the most popular drying technique in the food industry [10]. Our research group has demonstrated that spray-drying microencapsulation is an effective technique by which to protect, stabilize and concentrate natural phenolics recovered from different plant sources [11,12,13]. To this end, wall materials (typically complex carbohydrates and proteins) are used to create structures that wrap target materials (phytochemicals) and protect them from extraneous damage inflicted by environmental factors and/or further processing [14,15,16]. Moreover, extensive literature on the encapsulation of biomolecules with coloring attributes has demonstrated the efficiency of this technique in increasing the stability of food pigments [17].

To achieve the best results and avoid low-quality products and poor process efficiency, a comprehensive understanding and optimization of the production process is necessary [18]. Response surface methodology (RSM) is the most popular multivariate statistic technique, and it is extensively used to design, develop and optimize food-processing operations affected by several parameters. It allows for the establishment of polynomial models that accurately fit experimental data with the purpose of defining operational parameters that maximize selected responses. When accurately designed and conducted, this statistical tool enables the improvement of process performance in a cost-effective and timely manner [19].

Therefore, this work focuses on optimizing both the extraction and spray drying of anthocyanins recovered from hibiscus flowers (Hibiscus sabdariffa L.) by RSM to produce powdered extracts with preserved antioxidant phenolics. To this end, we (a) determined the optimal conditions that would maximize anthocyanin extraction from hibiscus flowers; (b) established the optimized parameters that would provide the highest efficiency and yield for the spray drying of anthocyanin-rich extracts and finally; (c) evaluated the oxidative stability index of spray-dried anthocyanin powders. This study is aligned with the emerging market for natural ingredients and products and explores anthocyanin-rich edible flowers as a sustainable source of valuable phytochemicals for the booming plant-based market.

2. Materials and Methods

2.1. Materials

Hibiscus flowers were obtained from the local market. Ethanol (EtOH; >98%), cyanidin-3-glucoside (C3G; >98%), butylated hydroxytoluene (BHT; >98%) and all other reagents were of analytical grade and were used without further purification.

2.2. Preparation of Hibiscus-Flower Flour (HFF)

Initially, 225 g of dried hibiscus flower calyxes was weighed and ground to produce a fine powder [20,21]. Specifically, the sample was pulverized using a Thomas hammer mill (Prater Industries, Bolingbrook, IL, USA) with a 2 mm mesh size, then processed in a cyclone mill (model CT 193 Cyclotec FOSS, Hilleroed, Denmark) with a 0.5 mm mesh size to produce HFF, which was stored at −20 °C.

2.3. Preliminary Evaluation of Anthocyanins Extraction

Preliminary experiments were conducted to evaluate the effects of five selected processing parameters (

Table 1. Experimental factors and levels for the preliminary fractional 2^5-1 factorial design for the evaluation of the effects of temperature (°C), pH, EtOH concentration (% v/v), extraction time (h) and solid-to-liquid ratio (% w/v) on the anthocyanin-extraction process.

Experimental Factors	Levels
	−1	0	+1
Temperature (°C)	31	35	39
pH	1.5	2.3	3.2
EtOH concentration (% v/v)	44	50	56
Extraction time (h)	1.6	4.0	6.4
Solid-to-liquid ratio (% w/v)	19	23	27

) on the extraction of anthocyanins from HFF based on previous reports [20,22]. A fractional 2^5-1 factorial design (E: ABCD) with three central points (19 total runs) was carried in order to identify the influence of temperature (31.0–39.0 °C), pH (1.5–3.2), EtOH concentration (44–56% v/v), extraction time (1.6–6.4 h) and solid-to-liquid ratio (S/L ratio; 19.0–27.0% w/v) on the extraction process. For the extraction, 3 g of HFF was added to a centrifuge tube wrapped in aluminum foil and the aqueous ethanol solvent was added at a defined EtOH concentration and S/L ratio, according to the experimental design (Table 1). Then, the pH of the mixture was adjusted with 1 M HCl solution to the defined condition according to the experimental design (Table 1) and the tubes were homogenized for 1 min and incubated at defined temperatures and times (Table 1). Finally, the samples were centrifuged at 4000× g for 30 min and the supernatant was collected for anthocyanin quantification.

2.3.1. Anthocyanin (ANC) Quantification

ANC concentration was determined by the pH differential method [23]. Briefly, 0.5 mL of anthocyanin-rich extracts were separately mixed with 4.5 mL of 0.025 M KCl (pH 1.0) and 0.4 M sodium acetate (pH 4.5) buffers and left to stand for 45 min in the dark at room temperature. Then, the absorbance of each solution was measured in a spectrophotometer (Thermo Fisher, Waltham, MA, USA) with readings at 520 nm and 700 nm for each buffer solution, ensuring an absorbance lower than 0.8. ANC concentration was expressed in cyaniding-3-glucoside (C3G) equivalents, according to Equation (1), as follows:

$$\mathrm{ANC} \mathrm{concentration} \left(\mathrm{mg}\frac{\mathrm{C}3\mathrm{G}}{\mathrm{L}}\right)=\frac{\mathrm{AA}}{\mathrm{MEC}}\times \mathrm{MW}\times 1000\times \mathrm{DF}$$

(1)

$$\mathrm{AA}={({\mathrm{ABS}}_{520}-{\mathrm{ABS}}_{700})}_{\mathrm{pH} 1.0}- {({\mathrm{ABS}}_{520}-{\mathrm{ABS}}_{700})}_{\mathrm{pH} 4.5}$$

(2)

where AA is the adjusted absorbance according to Equation (2), MW is the molecular weight of cyanidin-3-glucoside (457.6 g/mol), MEC is the molar extinction coefficient (29,600, for cyanidin-3-glucoside) and DF is the dilution factor (500).

2.4. Optimization of Anthocyanins Extraction

The extraction of anthocyanins from HFF was optimized following a 2² full factorial central composite experimental design (

Table 2. Experimental factors and levels for the 2² central composite experimental design for the optimization of ethanol concentration (% v/v) and extraction time (h) for the extraction of anthocyanins from HFF.

Experimental Factors	Levels
	−α	−1	0	+1	+α
EtOH concentration (% v/v)	39.9	44	50	56	60.1
Extraction time (h)	0.1	1.6	4.0	6.4	8

) with five central points (13 total runs). Based on preliminary experiments (Section 2.3), ethanol concentration (44–56% v/v) and extraction time (1.6–6.4 h) were identified as independent factors influencing anthocyanin extraction. For the extraction trials, 3 g of HFF were weighed into centrifuge tubes wrapped in aluminum foil. This step was followed by the addition of the aqueous ethanol extraction solution at a concentration chosen according to Table 2 and at a fixed solid-to-liquid ratio of 23% w/v. The tubes were homogenized for 1 min and incubated at room temperature for defined time periods (Table 2). Then, samples were centrifuged at 4000× g for 30 min and the supernatant was collected for analysis (Section 2.3.1). The optimization goal was to maximize the concentration of anthocyanins extracted from HFF.

2.5. Optimization of Production by Spray Drying

The spray-drying production of anthocyanin-rich ingredients from hibiscus flowers was carried out using a 1:1 w/w blend of starch and sodium alginate as the drying carrier. The influence of carrier addition rate (1–3% w/v) and inlet temperature (160–192 °C) on the spray-drying performance was optimized according to a 2² full factorial central composite experimental design (

Table 3. Experimental factors and levels for the 2² central composite experimental design for the optimization of carrier ratio (% w/v) and inlet temperature (°C) for the spray drying of anthocyanin-rich extracts from HFF.

Experimental Factors	Levels
	−α	−1	0	+1	+α
Carrier ratio (% w/v)	0.6	1.0	2.0	3.0	3.4
Inlet temperature (°C)	153.4	160	176	192	198.6

) with three central points (11 total runs). Anthocyanin-rich extracts were prepared under optimized conditions (item 2.3) and mixed with the drying-carrier blend at defined ratios (Table 3) for 20 min at 40 °C with magnetic stirring for complete dissolution.

Spray drying was performed in a laboratory-scale Labplant SD-Basic Spray Dryer (Lab Plant, Filey, UK) equipped with a two-channel 0.5 mm jet nozzle. The feed solution was constantly homogenized and was fed to the system with a peristaltic pump at a flow rate of 26.7 mL/min. Air was used in co-current flow at an operating pressure of 3 atm with the flow rate fixed at 70 m³/h. Inlet temperature was defined according to Table 3, and outlet temperatures ranged between 95–115 °C.

Solids Recovery (%) and ANC Retention (%)

Spray-drying yield, expressed as solids recovery (%), was calculated as the ratio between the total solids in the final powdered product and the total solids in the feed solution, according to Equation (3).

$$\mathrm{Solids} \mathrm{recovery} \left(\%\right)=\frac{\mathrm{Total} \mathrm{solids} \mathrm{of} \mathrm{spray-dried} \mathrm{product} \left(\mathrm{g}\right)}{\mathrm{Total} \mathrm{solids} \mathrm{in} \mathrm{the} \mathrm{feed} \mathrm{solution} \left(\mathrm{g}\right)}\times 100$$

(3)

ANC retention (%) was calculated as the ratio between the ANC concentration in the spray-dried product and the ANC concentration in the feed solution, according to Equation (4).

$$\mathrm{ANC} \mathrm{retention} \left(\%\right)=\frac{\mathrm{ANC} \mathrm{concentration} \mathrm{in} \mathrm{the} \mathrm{spray-dried} \mathrm{product} (\mathrm{mg} \mathrm{C}3\mathrm{G}/\mathrm{g})}{\mathrm{ANC} \mathrm{concentration} \mathrm{in} \mathrm{the} \mathrm{feed} \mathrm{solution} (\mathrm{mg} \mathrm{C}3\mathrm{G}/\mathrm{g})}\times 100$$

(4)

Optimized conditions were determined with the goal of maximizing both the solids recovery (%) and ANC retention (%) and were used to prepare spray-dried extracts for the evaluation of oxidative stability in soybean oil.

2.6. Oxidative Stability Test

The Rancimat analysis was conducted by adding powdered ANC-rich hibiscus extracts or BHT to soybean oil and comparing their effects on the oxidative stability of the oil. Samples (0.5 g) of spray-dried hibiscus extracts obtained under optimized conditions were mixed with soybean oil and results were compared with those obtained using 0.5 g of either synthetic antioxidant BHT or a control sample (no addition, soybean only). The test was performed (Rancimat, Metrohm, Switzerland) under the following conditions: 3 g of soybean oil, air flow of 10 L/h and temperature of 110 °C. Measurements of the electrical conductivity of volatile degradation products was used to determine the induction time (h) and consequent oxidative stability of soybean oil [24].

2.7. Statistical Analysis

Results are shown as average ± standard deviation. Response surface experimental design, analysis of variance (ANOVA) evaluation and desirability function for response optimization were implemented using DesignExpert software (version 13.0; Stat-Ease, Minneapolis, MN, USA).

3. Results and Discussion

3.1. Preliminary Evaluation of ANC Extraction

Solid–liquid extraction is the most straightforward approach to ANC extraction from plant tissues, and the process efficiency is directly influenced by key processing parameters that need to be assessed for each selected matrix. Temperature, solvent composition (EtOH concentration, pH), extraction time (also referred as contact time) and solid-to-liquid ratio are directly related to the efficiency of mass transfer from the plant tissue to the extraction solvent; they also influence solute solubility and, therefore, the extraction equilibrium [25]. The influence of selected process parameters (temperature, pH, EtOH concentration, extraction time and solid-to-liquid ratio) on the extraction of anthocyanins from HFF was preliminarily evaluated through a fractional 2^5-1 factorial design. The goal of this step was to identify the factors that significantly (p < 0.05) affected the extraction process so that they could be further optimized.

Table 4. Analysis of variance (ANOVA) of regression model from preliminary fractional 2^5-1 factorial design for the evaluation of the effects of temperature (°C), pH, EtOH concentration (% v/v), extraction time (h) and solid-to-liquid ratio (% w/v) on the anthocyanin-extraction process.

Factor	df	SS	MS	F-Value	p-Value
Model	5	1.424 × 107	2.849 × 106	4.39	0.0146 *
A: Temperature (°C)	1	1.568 × 105	1.568 × 105	0.2416	0.6312
B: pH	1	4.442 × 105	4.442 × 105	0.6844	0.4230
C: EtOH concentration (% v/v)	1	3.907 × 106	3.907 × 106	6.02	0.0290 *
D: Extraction time (h)	1	8.034 × 106	8.034 × 106	12.38	0.0038 **
E: S/L ratio (% w/v)	1	1.702 × 106	1.702 × 106	2.62	0.1294
Residual	13	8.438 × 106	6.491 × 105
Lack of Fit	11	8.217 × 106	7.470 × 105	6.75	0.1360
Pure Error	2	2.213 × 105	1.106 × 105
Cor Total	18	2.268 × 107

Statistically significant factors are indicated as follows: * p < 0.05, ** p < 0.01. Legend: HFF—hibiscus-flower flour; EtOH—ethanol; S/L ratio—solid-to-liquid ratio; df—degrees of freedom; SS—sum of squares; MS—mean square.

shows the ANOVA for the linear regression model (R² = 0.6280) obtained from the fractional 2^5-1 factorial design. EtOH concentration (% v/v) and extraction time (h) were found to be significant parameters (p < 0.05) in ANC extraction and were prioritized for further optimization with the aim of maximizing the ANC concentration in the extracted solutions. Indeed, ethanol is a polar solvent for efficient ANC extraction [25]. As in our study, Rocha et al. [26] has previously reported finding that both ethanol concentration and extraction time significantly influence the extraction of ANC from red raspberries. The two selected processing parameters (EtOH concentration and extraction time) were further optimized through a full factorial 2² central composite experimental design.

3.2. Anthocyanin Recovery from HFF

Hibiscus calyxes are readily available source of valuable anthocyanins for sustainable production of food ingredients. However, the stabilization of these ephemeral compounds in a convenient format is a crucial step in the formulation of functional ingredients that can be incorporated into complex food matrices. Solvent-assisted extraction has been described as a well-established, convenient approach for the recovery of natural polyphenols, including anthocyanins from fruits and vegetable products [27]. In this regard, ethanol is a suitable solvent owing to its non-toxic, food-grade status and availability [28]. Ethanol concentration (% v/v) and extraction time were found to influence ANC extraction from HFF, so these two parameters were optimized by a 2² central composite experimental design (

Table 5. Full factorial 2² central composite experimental design for the optimization of ethanol concentration (% v/v) and time (h) for the extraction of anthocyanins from HFF.

EtOH Concentration (% v/v)	Time (h)	ANC Concentration (mg C3G/L)
50 †	4 †	1453
50	8	1956
50 †	4 †	1415
44	6.4	1963
50 †	4 †	1654
50 †	4 †	1500
50	0.1	951
60.1	4	1337
44	1.6	1445
56	6.4	1724
56	1.6	1020
39.9	4	2211
50 †	4 †	1616

^† Central points. Legend: HFF—hibiscus-flower flour; EtOH—ethanol; ANC—anthocyanins; C3G—cyanidin-3-glucoside.

), using the ANC concentration in the extract as goal response for the optimization.

A second-order model (Equation (5), non-coded factors) was obtained to estimate the ANC concentration from HFF extraction, and the ANOVA of the response surface experimental design (Table S1) showed ethanol concentration and extraction time as linear factors and ethanol concentration as a quadratic factor significantly affecting (p < 0.05) the concentration of extracted ANC. A high correlation coefficient (R² = 0.9554) was found, indicating that the experimental data fit well to the model. Anthocyanins are water-soluble pigments; however, a high concentration of water in the solvent can cause the degradation of anthocyanins by affecting the deprotonation/reprotonation rates of these molecules [29]. At lower water concentrations, the lack of free water molecules surrounding the ANC molecules can stabilize the protonated form, which increases color stability [30]. Therefore, water-ethanol mixtures must provide an appropriate equilibrium between extraction/stabilization capacities for satisfactory anthocyanin extraction. Response surface analysis (Figure 1) and the desirability function showed that the maximum ANC extraction was obtained at an ethanol concentration of 41.6% v/v, beyond which no significant increase in ANC extraction was observed.

$$\begin{gathered} \mathrm{ANC} \mathrm{concentration} \\ \left(\mathrm{mg}\frac{\mathrm{C}3\mathrm{G}}{\mathrm{L}}\right)=9086.97 - 277.26 \times \mathrm{EtOH} \mathrm{concentration}\left(\%\right)+15.25 \times \mathrm{Time} \left(\mathrm{h}\right)+2.27 \times {\left[\mathrm{EtOH} \mathrm{concentration} \left(\%\right)\right]}^2 \end{gathered}$$

(5)

The extraction time had a linear and positive effect on ANC extraction (p < 0.05, Table S1). Results showed that a longer extraction time favored the ANC mass transfer from HFF to the extraction solution (Figure 1). The desirability function proved that a maximized ANC extraction occurs at 8 h, and under optimized conditions (41.6% v/v ethanol, 8 h), an extract containing 2489 mg C3G/L was obtained. The ANC concentration obtained in this study (1078 mg C3G/100g HFF) was comparable to that obtained with similar water-ethanol extraction of Hibiscus sabdariffa, as reported by Escobar-Ortiz et al. [21] (1000–1500 mg C3G/100g), and higher than that obtained with ultrasound-assisted extraction of blackberry pomace by Santos et al. [31] (45–120 mg C3G/100g).

3.3. Production of Spray-Dried Anthocyanin-Rich Components

The ANC-rich hibiscus extract prepared under optimized conditions (41.6% v/v ethanol, 8 h) was used for the spray-drying production of anthocyanin-rich ingredients using a 1:1 w/w blend of starch and sodium alginate as the drying carrier. For spray drying, a 2² central composite experimental design was used to optimize the carrier addition ratio (1–3% w/v) and inlet temperature (160–192 °C) to maximize both the solids recovery (%) and ANC retention (%).

Table 6. Full factorial 2² central composite experimental design for the optimization of carrier ratio (% w/v) and inlet temperature (°C) for the spray drying of anthocyanin-rich extracts from HFF.

Carrier Ratio (% w/v)	Inlet Temperature (°C)	Solids Recovery (%)	ANC Retention (%)
1	160	50.0	88.7
2	198.6	72.5	92.7
0.6	176	59.3	85.7
3	192	59.3	94.8
2 †	176 †	65.3	93.5
2 †	176 †	63.0	93.7
3.4	176	57.4	96.0
2 †	176 †	60.6	94.3
1	192	58.0	88.4
3	160	56.7	95.1
2	153.4	46.5	94.0

^† Central points. Legend: HFF—Hibiscus flower flour; ANC—anthocyanins.

shows the results from the experimental design for spray-drying optimization.

Spray-drying yield, assessed as the total solids recovered from the feed solution, is a useful index by which to evaluate the performance of the spray-drying operation, as it dictates the potential feasibility and scalability of the process [32]. Solids recovery ranged from 46.5% to 72.5% (Table 6), with most treatment combinations showing results above 50% (except when 2% w/v and 153.4 °C were used), which is the lower limit considered to show acceptable performance for lab-scale spray-drying protocols [33]. A first-order linear model was found to best describe the behavior of the experimental data (Equation (6), non-coded factors). ANOVA (Table S2) showed the linear effect of the inlet temperature to significantly (p < 0.05) affect solids recovery, and there was a non-significant (p > 0.05) lack of fit. The linear and positive influence of the inlet temperature on the solids recovery can be seen in the shape of the response surface of the obtained model (Figure 2A), which shows a clear upward inclination towards higher temperature ranges and no curvatures.

$$\mathrm{Solids} \mathrm{recovery} \left(\%\right)=- 7.53+0.37 \times \mathrm{Inlet} \mathrm{temperature} (°\mathrm{C})$$

(6)

The carrier blend starch–sodium alginate was chosen due to its solubility, low toxicity, biocompatibility, biodegradability and low cost, which are reflected in its well-established utilization by the food and pharmaceutical industries [34]. Both starch and sodium alginate have gel forming capacities that make them suitable options as microencapsulation by spray drying materials of sensitive target molecules, such as hibiscus-derived anthocyanins [35,36]. Similar solids-recovery results were reported by Baltrusch et al. [37] (50–58%) with regard to the microencapsulation by spray drying of tea extracts with similar carbohydrate polymers (starch, alginate and carrageenan) and by Vergara et al. [38] (20–81%) with regard to the production of spray-dried anthocyanin-rich products from purple-fleshed potatoes using maltodextrin as the drying carrier. The solids recovery obtained in this work under optimized conditions (3.2% w/v carrier ratio, 170.7 °C inlet temperature, 60.8% solids recovery) is much higher than that reported by Martins et al. [39] (15.8–42.7%), which underlines the satisfactory performance of the developed spray-drying protocol—higher solids recovery indicates a more efficient drying process.

The ANC retention (%) reflects the ability of the spray-drying process to preserve the ANC content originally found in the hibiscus extract, carrying the valuable compounds from the feed solution to the final dried ingredient [40]. Remarkable ANC retention, above 85%, was found for all treatments (Table 6), indicating that the chosen carrier addition (1–3% w/v) and inlet temperatures (160–192 °C) ranges were able to efficiently deliver ANC-rich spray-dried products. A second-order model (Equation (7), non-coded factors) was adjusted to the ANC-retention experimental data, and a high correlation coefficient (R² = 0.9891) and non-significant lack of fit (p > 0.05, Table S3) indicate the predictive capacity of the model. ANOVA analysis revealed that the carrier addition rate significantly (p < 0.05, Table S3) affected the ANC retention in both linear and quadratic terms, while no significant effect of temperature (p > 0.05) was found. Response surface analysis of ANC retention results (Figure 2B) revealed a slight curvature associated with the carrier addition rate (Equation (7)), as well as the maximization of ANC retention at higher carrier addition rates.

$$\mathrm{ANC} \mathrm{retention} \left(\%\right)=+44.85+9.85 \times \mathrm{Carrier} \mathrm{ratio} (\%) - 1.61 \times {[\mathrm{Carrier} \mathrm{ratio} (\%\left)\right]}^2$$

(7)

Similar ANC-retention results were reported by Nguyen et al. [41] with regard to the spray drying of hibiscus anthocyanins with maltodextrin (>90%). The desirability function showed the optimal operation conditions to be at 3.2% w/v carrier addition ratio and 170.7 °C inlet temperature. Interestingly, high solids recovery and high ANC retention were obtained with very low amounts of drying carrier, which contributes to decreasing production costs. Indeed, high ANC retention had been previously reported for the spray drying of hibiscus extracts using carbohydrate-based carriers such as maltodextrin and gum arabic; however, addition rates varied from 10% up to 50% w/v [22,39]. The choice of spray-drying carrier can directly affect not only the process performance and the product quality, but also the environmental performance and overall sustainability of the drying process [42]. Therefore, the ability to produce spray-dried ANC-rich ingredients using biodegradable, low-cost carriers at low addition rates highlights the success of the optimized spray-drying protocol. Spray-dried ANC-rich ingredients prepared under these conditions yielded 60.8% and 96.0% solids recovery and ANC retention, respectively, and were used in the oxidative-stability test.

3.4. Rancimat Test—Assessment of Oxidative Stability Index

The ability of spray-dried ANC-rich hibiscus extract to prevent heat-induced oxidation of soybean oil was assessed through the Rancimat test. In this protocol, samples were subjected to heating in the presence of an air stream that induces autoxidation and the consequent formation of low-molecular-weight organic acids that are measured through changes in conductivity, correlating the induction time (time to the onset of oxidative decomposition), with the antioxidant capacity of the ingredient [43].

Longer induction times indicate higher antioxidant activity, i.e., demonstrate that a particular system is less prone to be oxidatively deteriorated. The induction time of soybean oil without any antioxidant addition was 22.49 ± 2.97 h, while the addition of BHT significantly increased the induction time to 35.50 ± 0.42 h (p < 0.05). The addition of microencapsulated hibiscus extracts to the soybean oil resulted in an induction time of 30.39 ± 0.83, a significantly greater value than was found for the control, but still lower than the value found for BHT (p < 0.05). These results show that anthocyanin-rich hibiscus extracts could be an effective natural antioxidant agent.

Phenolic-based compounds, such as anthocyanins, have been previously reported for preventing heat-induced autoxidation of oil, as assessed by the Rancimat test. Souza et al. [44] reported that freeze-dried ANC-rich extracts from Saskatoon berry (Amelanchier alnifolia Nutt.) were able to increase the induction time of polyunsaturated borage oil up to 65%, while Serrano-Díaz et al. [45] reported that freeze-dried extracts from Crocus sativus flowers delayed the autoxidation of sunflower oil up to 12%, under similar test conditions. Spray-dried hibiscus extracts obtained under optimized conditions in this work delivered a 35% delay in the induction time of pure soybean oil, which shows that a rational, scalable spray-drying protocol can deliver high-quality, shelf-stable products with significant antioxidant capacity. Moreover, the antioxidative effect expected from such natural products is dose-dependent [43,44], so additional studies with varying addition levels of ANC-rich hibiscus extracts would be needed to clarify what concentration would be necessary to outperform BHT.

4. Conclusions

In this work, we demonstrated the optimized extraction and microencapsulation by spray drying of abundant, low-cost hibiscus flowers anthocyanins using well-established and scalable techniques. Our results determined that maximized anthocyanin extraction can be achieved after an 8 h extraction, using a 41.6% v/v aqueous ethanol solution to produce an anthocyanin-rich extract containing 3000 mg C3G/L. When optimized hibiscus extracts were spray-dried with 3.2% w/v of starch–sodium alginate carrier blend and 170.7 °C inlet temperature, remarkable values of 60.8% and 96.0% were achieved for solids recovery and anthocyanin retention, respectively. Moreover, anthocyanin-rich ingredients were proven to delay the onset of induced oxidation in soybean oil, although more detailed studies are required to elucidate dose-dependent effects and antioxidant potential. Overall, our results show that hibiscus flowers are a sustainable and readily available source of anthocyanins, molecules with market and health relevance that can be feasibly used to supply ingredients for value-added applications.