Dehydrated Hibiscus sabdariffa Calyces as Anthocyanin-Rich Natural Colorants: Influence of Food-Grade Extraction and Syrup Formulation on Stability and Technological Performance

TL;DR

This study explores using dried hibiscus calyces as natural colorants, showing that they can be effectively extracted and used in food products with good stability.

Contribution

The study introduces a method for extracting stable anthocyanins from dehydrated hibiscus calyces and evaluates their performance in syrup formulations.

Findings

Dehydration reduced moisture and improved storage stability of hibiscus calyces.

The 5% sucrose syrup formulation showed the highest anthocyanin concentration and minimal loss over 30 days.

Color stability in food matrices was pH-dependent, with strong correlations to hue and saturation.

Abstract

The clean-label trend demands stable natural colorants. H. sabdariffa L. calyces is a sustainable anthocyanin source; however, the effects of dehydration and food-grade extraction on pigment recovery and performance remain unclear. Fresh and dehydrated calyces were physicochemically characterized, and dehydrated material with aqueous and hydroalcoholic food-grade solvents was used for extraction. Extraction efficiency was evaluated through total phenolic compound (TPC) content and anthocyanin characterization by HPLC-DAD-ESI-MS/MS. The aqueous extract with the highest anthocyanin concentration was selected for the syrup formulation containing 5–35% sucrose. The physicochemical stability of the formulations was monitored during refrigerated storage (4 °C/30 days) and subsequently tested in food matrices with different pH values. Dehydration calyces reduced moisture content (11.61%) and water activity (aw = 0.56), indicating improved storage stability. The anthocyanin concentration was highest in the 5% sucrose syrup (495.12 ± 40.66 mg mv-3,5-glc·100 g−1), with a 12.71% loss over 30 days. Application in food matrices demonstrated a pH-dependent color response, with correlations between hue angle (h°, r = 0.86) and color saturation (C*, r = −0.77).These results demonstrate that hibiscus calyx syrups are promising natural colorants and bioactive ingredients, particularly for acidic food systems, offering stability.

Figures

Full text

1. Introduction

The growing demand for healthy and natural products has significantly influenced consumer choices. In this context, plant- and fruit-based extracts, particularly those rich in phenolic compounds (PC), have gained considerable attention because of their potential health-promoting properties and dual functionality as natural colorants and bioactive ingredients [1,2]. Beyond their technological role in enhancing color and sensory appeal [3].

Brazil hosts one of the world’s greatest biodiversities, with nearly 45,000 plant species, of which only about 10% are edible, and a substantial portion remains underexplored or underutilized [4]. In this context, the prospecting of parts of neglected and underutilized species (NUS) can significantly contribute to dietary diversification, promote the development of a more sustainable food system, and strengthen food security. The valorization of NUS has emerged as a strategic approach to diversifying food systems, enhancing dietary quality, and promoting sustainability [4,5]. Recent studies have emphasized that NUS integration into food chains can improve micronutrient intake, reduce dependence on major staple crops, and contribute to food and nutritional security, particularly in vulnerable populations [6,7]. Additionally, the resilience of many NUS to adverse environmental conditions and their lower agronomic input requirements reinforce their potential role in sustainable and climate-resilient food production systems [8].

Among NUS with recognized industrial and nutritional potential, Hibiscus sabdariffa L., commonly known as roselle, stands out due to its wide geographical distribution and several edible parts, including leaves, seeds, and red calyces [3,9]. Roselle calyces (RCs) are particularly valued because of their high concentration of anthocyanins, mainly delphinidin- and cyanidin-based derivatives, along with relevant levels of flavonols and phenolic acids [10,11,12]. These compounds confer a distinctive coloration to RC, making their extracts promising for food applications, particularly as natural colorants and bioingredients [13].

Owing to the structural diversity and polarity of PC present in RCs, numerous studies have focused on optimizing extraction processes to maximize yield and preserve compound stability. Conventional extraction techniques, such as infusion, maceration, and decoction, remain widely employed because of their simplicity and compatibility with food applications, typically using temperatures between 25 °C and 100 °C, extraction times ranging from minutes to several hours, and solid-to-solvent ratios varying from 1:5 to 1:40, occasionally including pasteurization (90 °C, 20 min) [14,15,16,17,18,19]. However, recent advances have demonstrated that emerging and intensified extraction technologies, including ultrasound-assisted extraction (UAE), cold plasma–assisted extraction (including low-vacuum configurations), and green solvent-based approaches, can enhance mass transfer, reduce processing time, and improve extraction efficiency while minimizing the thermal degradation of anthocyanins and other PC [20,21,22,23].

These technologies can significantly increase the recovery of bioactive compounds compared with conventional methods, highlighting their relevance for industrial applications [21,23]. In a study targeting food applications, UAE was performed for 45 min at 65 °C using 25% ethanol, resulting in bioactive compound concentrations that were 2–4 times higher than those obtained via conventional extraction, indicating the greater efficiency of the method [15].

Nevertheless, the translation of these techniques to food-grade processes remains challenging, mainly due to solvent safety requirements, cost, scalability, and regulatory constraints, which limit their direct application in food formulations [20,22].

In addition to extraction efficiency, the stabilization and incorporation of phenolic-rich extracts into food matrices remain key challenges. The susceptibility of anthocyanins to environmental factors, such as pH, temperature, light, and oxygen, can significantly affect color stability and functional performance during processing and storage [21,24]. Consequently, recent research has increasingly focused on developing technological strategies to improve the stability and applicability of these compounds, including encapsulation, formulation in semi-solid systems, and incorporation into intermediate food ingredients such as syrups [21,22]. Syrup-based formulations have gained attention owing to their versatility, ease of incorporation into multiple food matrices, and potential to modulate sensory and rheological properties, including sweetness, viscosity, and color intensity, while serving as carriers of bioactive compounds [25,26,27]. However, studies integrating extraction with the technological performance of phenolic-rich ingredients in model food systems, particularly those evaluating the influence of raw material processing, extraction using food-grade solvents, and formulation strategies on phenolic stability and color performance across different food matrices, remain scarce.

Therefore, this study aimed to investigate the technological potential of H. sabdariffa L. calyces as a source of anthocyanin-rich natural ingredients. Fresh and dehydrated calyces from the first harvest (H1) were characterized in terms of physicochemical and phytochemical properties (moisture content, water activity (a_w_), pH, total acidity (TA), soluble solids (SS), instrumental color, total phenolic compounds (TPC), and total anthocyanins (TAC)) to evaluate the effects of dehydration on raw material quality and stability. Dehydrated calyces from a second harvest (H2) were subsequently used to evaluate the efficiency of aqueous and hydroalcoholic food-grade solvents for bioactive compound extraction, employing a hydromethanolic extract as a reference control. The extract presenting the most suitable bioactive profile, based on TPC values and anthocyanin composition determined by HPLC-DAD-ESI-MS/MS, was selected for syrup formulation. Syrups containing different sucrose concentrations were then produced, and their physicochemical stability was monitored during refrigerated storage (4 °C) for 30 days. Finally, the control extract and the syrup with the highest PC retention after storage were applied to food matrices with different pH values to assess their technological applicability and initial color behavior.

2. Materials and Methods

2.1. Chemical Reagents and Their Products

All reagents used in the physicochemical characterization analyses were of analytical grade (>95%): HCl (Dinâmica, Indaiatuba, Brazil), Folin–Ciocalteu reagent (Ls Chemicals, Ribeirão Preto, Brazil), anhydrous sodium carbonate (Ls Chemicals, Ribeirão Preto, Brazil), and NaOH (Dinâmica, Indaiatuba, Brazil). Chromatographic analyses were performed using ultrapure water obtained from a Milli-Q purification system and high-performance liquid chromatography (HPLC)-grade solvents. An analytical-grade (>95%) cyanidin (Cy) 3-O-glucoside (Glc) standard was obtained from Extrasynthese (Genay, France) and used as a reference. Sucrose (Camil Alimentos, Barra Bonita, Brazil), tonic water (Refrigerantes Arco Íris, São José do Rio Preto, Brazil), fermented milk (Frutap Alimentos, Bernardino de Campos, Brazil), Greek yogurt (Lactalis do Brasil, Carambeí, Brazil), powdered milk (Goiasminas Indústria de Laticínios, Tapejara, Brazil), and powdered dessert mixes for pudding, whipped topping, and maria-mole (Dr. Oetker Brasil, São Paulo, Brazil) were purchased from a local store.

2.2. Acquisition of Fresh and DC

Both fresh and dehydrated RC were obtained from H1 at an organic flower farm located in São Carlos, SP, Brazil (21°59′21.84″ S, 47°55′47.78″ W), where cultivation was conducted under greenhouse conditions. The DC used for extract and syrup preparation was obtained from H2 from the same property. Dr. Daniela Sampaio Silveira identified the plant species, and a voucher specimen (no. 36287) was deposited at the Herbarium of São José do Rio Preto (SJRP), IBILCE/Unesp, São Paulo State, Brazil. Both harvests were carried out in March during the summer.

The process for obtaining the DC is described below and illustrated in Figure 1. After harvesting the immature fruits and manually removing the capsules, the RC underwent a mixed drying process consisting of direct sunlight exposure for 3 days followed by shade drying in a ventilated environment for 7 days, aiming for a mass loss greater than 88%, in accordance with the moisture content limits established for commercialization [28].

2.3. Physicochemical Characterization of RC (Fresh and DC)

The physicochemical characterization of fresh and DC (H1) was performed in triplicate, following the methodologies described by the Association of Official Analytical Chemists [29]. The following analyses were conducted: moisture content, determined by the thermogravimetric method in a vacuum oven at 70 °C (TE-395, Tecnal, Piracicaba, Brazil); SS, measured by refractometry (K52-032, Kasvi, São José dos Pinhais, Brazil), with results expressed in °Brix at 25 °C; pH, determined directly using a pH meter (TEC-5, Tecnal, Piracicaba, Brazil); TA, assessed by potentiometric titration and expressed as g of malic acid·100 g^−1^ of sample; and a_w_, measured at 25 °C using an electric hygrometer (Aw Sprint, Novasina, Lachen, Switzerland). Additionally, TPC was determined according to Singleton et al. [30] and TAC according to Ribéreau-Gayon and Stonestreet [31] using a spectrophotometer (Cary 60 UV-Vis, Agilent Technologies, Santa Clara, CA, USA), with results expressed as mg gallic acid equivalents (GAE) g^−1^ and mg malvidin-3,5-glucoside (mv-3,5-glc) g^−1^, respectively.

For spectrophotometric analyses, both fresh and DC from H1 were subjected to preliminary PC extraction, following the method adapted from Santos and Martins [22]. Water was selected as the extraction solvent for preliminary analyses due to its low environmental impact, cost-effectiveness, and suitability for future food applications. Briefly, the RC samples were ground using a hand mixer (PMX2000, PHILCO, Joinville, Brazil), and 1 g of each of the homogenized samples, in triplicate, was mixed with 25 mL of water. The mixtures were subjected to an ultrasound bath (USC-1850, Unique, Indaiatuba, Brazil) at a frequency of 25 kHz and a power of 135 W for 30 min at room temperature (~25 °C) and subsequently centrifuged (CR-G111, Hitachi, Tokyo, Japan) at 14,700× g for 10 min at 4 °C. The resulting supernatants were stored under refrigeration (~7 °C), and the resulting precipitates were subjected to five additional consecutive extractions, following the procedure described in the previous section. Finally, the supernatants from each extraction were combined and concentrated in a rotary evaporator (Hei-Vap Advantage, Heidolph, Schwabach, Germany) at 36 °C, and the final volume was adjusted to 50 mL using a volumetric flask to ensure sufficient concentration and uniformity for subsequent analyses.

Finally, the instrumental color of the RC was determined using a Color Flex 45/0 spectrophotometer (HunterLab, Reston, VA, USA), configured with illuminant D65 and a 10° observer angle, calibrated according to the Commission Internationale de l’Éclairage (CIE) Lab* color scale. Universal software version 4.10 was used to determine the absolute values of L* (lightness, ranging from 0 to 100, from black to white), a* (green to red variation), and b* (blue to yellow variation). Using the absolute values of a* and b*, chroma (C*) was calculated in cylindrical coordinates using Equation (1), and the hue angle (h°) was calculated using Equation (2) [32,33].

$$[eqn]$$ $$[eqn]$$

2.4. Preparation and Characterization of DC Phenolic Extracts Using Different Extraction Solvents

Preliminary assays were conducted using DC from H2 to evaluate the efficiency of different extraction solvents, employing aqueous, hydroalcoholic (water: ethanol, 40:60, v/v), and hydromethanolic (methanol:water:formic acid, 50:48.5:1.5, v/v/v) solutions, with the latter considered as the control. The same extraction procedure described in the previous section was applied to all solvents, enabling a comparative assessment of TPC and the anthocyanin profile prior to syrup formulation. The same methodology described in Section 2.3 was used to determine TPC.

An Agilent 1100 series high-performance liquid chromatograph (Agilent Technologies, Waldbronn, Germany) equipped with a diode array detector (DAD, G1315B) and coupled to an LC/MSD Trap VL (G2445C VL) via an electrospray ionization system (ESI-MS/MS), integrated with an Agilent ChemStation data processing unit (version B.01.03), was used to separate, identify, and quantify anthocyanins in the DC extracts. The mass spectral data were processed using Agilent LC/MS Trap software (version 5.3).

Aliquots (5 mL) of the extracts were subjected to solid-phase extraction (SPE) using C18 cartridges (Waters, Milford, CT, USA) to remove sugars and other polar compounds. The resulting extracts were then filtered through a 0.20 μm polyester membrane (Chromafil PET 20/25, Macherey-Nagel, Düren, Germany) and injected (10 μL) into a reversed-phase C18 chromatographic column (Ascentis Express, 2.1 × 150 mm, 2.7 μm particle size; Agilent Technologies, Santa Clara, CA, USA) at 40 °C. The analysis was performed at a flow rate of 0.19 mL·min^−1^, using the following solvents (v/v/v) as the mobile phase: (A) acetonitrile/water/formic acid, 3:88.5:8.5; (B) acetonitrile/water/formic acid, 50:41.5:8.5; and (C) methanol/water/formic acid, 90:1.5:8.5 [34]. The elution followed a linear gradient program as follows: 0 min, 94% A and 6% B; 10 min, 70% A and 30% B; 30 min, 50% A and 50% B; 34–36 min, 100% B; 42 min, 96% A and 4% B, maintained until 50 min. All analyses were performed in duplicate.

Anthocyanin identification was primarily based on spectroscopic data (UV-Vis and MS/MS) reported in previous studies [10,11,12,35]. The mass spectrometry data were acquired in the positive ionization mode. DAD chromatograms were extracted at 520 nm for quantification. The calibration curve was constructed using cyanidin-3-glucoside (cy-3-glc) as an external standard, and the results were expressed in mg of the corresponding standard per kg^−1^ of DC.

2.5. Preparation of Syrups from Selected DC Extracts and Physicochemical Stability Monitoring During Refrigerated Storage

After selecting the most suitable DC extract based on TPC and anthocyanin profile, syrups were prepared by adding sucrose at different concentrations (w/v): E_5_ (5%), E_15_ (15%), E_25_ (25%), E_35_ (35%). A control extract (E_0_) without sucrose addition was also prepared. Each sample (~40 mL) was transferred to amber glass bottles, hermetically sealed, protected from light, and prepared in triplicate. The samples were stored in a refrigerated BOD incubator (CE-300/350-FAU, Cienlab, Campinas, Brazil) at 4 °C for 30 days. During storage, physicochemical characteristics were monitored, including SS, pH, TA, a_w_, TPC, and TAC, as well as instrumental color stability, following the methodologies previously described.

The physicochemical characterization of the syrups and the E_0_ extract was performed at the beginning of the experiment (day 0) and after the storage period (30 days). On the basis of the L*, a*, and b* values obtained from the instrumental color analysis, the total color difference (ΔE) was calculated using Equation (3) [36]:

$$[eqn]$$

where L*, a*, and b* represent the color parameters of the samples after 30 days of storage, and $[eqn]$ , $[eqn]$ , and $[eqn]$ correspond to the initial values of these parameters at day 0.

2.6. Chromatic Characterization of the Selected DC Extracts and Syrups

2.6.1. Evaluation of the Selected DC Extract in Model Buffer Systems and Food Matrices

The coloration of RC extracts is directly associated with anthocyanins, whose molecular structure is highly sensitive to pH, leading to significant chromatic changes. Therefore, the chromatic behavior of the E_0_ extract was evaluated using instrumental color analysis under different pH conditions. The procedure was adapted from Chen et al. [36], in which the extract was added to buffer solutions at selected pH values to simulate different environments. Buffer solutions were prepared as follows: pH 1.5 (glycine, 0.1 mol·L^−1^), pH 3.5 (sodium citrate, 0.1 mol·L^−1^), pH 5.5 (sodium acetate, 0.1 mol·L^−1^), and pH 7.5 and 9.5 (glycine–sodium hydroxide, 0.1 mol·L^−1^). The extract was added in triplicate to each buffer, homogenized, and analyzed for instrumental color parameters (L*, a*, b*, C*, and h°). Additionally, aliquots diluted in distilled water were used as a control for color stability comparison. These data were also used for ΔE calculations, as described previously.

Following buffer system analysis, the E_0_ extract was incorporated into selected commercial food matrices with pH values ranging from 3.0 to 7.0. These included tonic water (pH 3.0), fermented milk and maria-mole (pH 4.0), Greek yogurt (pH 4.5), reconstituted powdered milk and pudding (pH 6.5), and whipped topping (pH 7.0). The powdered products (maria-mole, pudding, and whipped topping) were prepared according to the manufacturer’s instructions, with the addition of water, milk, and sugar as needed. In all cases, 25% (w/w) of the extract was incorporated to ensure consistent testing conditions. The anthocyanin concentration (mv-3,5-glc, g·mL^−1^) was also used as a reference for cross-system comparisons. Instrumental color analysis (L*, C*, and h°) was performed to assess whether the chromatic patterns observed in the buffer systems were replicated in the food matrices with different pH levels.

2.6.2. Comparison of Selected DC Extract and Syrup in Food Matrices

The syrup formulation that demonstrated the highest stability in chromatic and compositional parameters was selected based on the results obtained after 30 days of storage. This syrup was then incorporated into the same food matrices previously tested with the E_0_ extract. Chromatic analysis was conducted using the same instrumental methodology (L*, C*, and h°), directly comparing the extract and syrup and evaluating the potential impact of sucrose addition on color behavior in real food systems.

2.7. Statistical Analysis

All results are expressed as mean ± standard deviation. The physicochemical parameters of fresh and DC (H1) were compared using Student’s t-test (p ≤ 0.05) using Microsoft Excel (Version 2504, Microsoft Corporation, Redmond, WA, USA).

TPC and anthocyanin profiles of DC (H2) extracts from different solvents were analyzed by one-way analysis of variance followed by Tukey’s test (p ≤ 0.05). The physicochemical parameters of the E_0_ and the syrups during storage were analyzed using two-way ANOVA (factors: storage time and sucrose concentration), followed by Tukey’s post hoc test (α = 0.05) and Principal component analysis (PCA). For the data referring to the incorporation of E_0_ and one selected syrup formulation into commercial food matrices, a two-way ANOVA was also conducted in two stages: (i) considering the factors of food matrix and the presence or absence of the extract; and (ii) food matrix and the presence or absence of sucrose in the extract. These analyses were conducted using Statistica version 10.0 (StatSoft Inc., Tulsa, OK, USA).

Additionally, Pearson correlation analysis was performed between pH and the samples with and without sucrose, considering statistical significance at p ≤ 0.05, p ≤ 0.01, and p ≤ 0.001, and correlation coefficients equal to or greater than |0.7|. These analyses were conducted using the R software (version 4.1.3) [37] and the corrplot package [38].

3. Results and Discussion

This section presents the physicochemical characterization of fresh and dehydrated RC from harvest H1, followed by the evaluation of food-grade extraction solvents applied to DC from harvest H2. Then, the most suitable extracts were used to develop syrups with different sucrose concentrations, whose stability during refrigerated storage was assessed. Finally, the E_0_ was used as a reference, and the syrup showing the most favorable results was applied in food matrices with different pH values to evaluate its feasibility of use and initial color behavior.

3.1. Physicochemical Characteristics of RC (Fresh and DC)

Table 1 presents the physicochemical characteristics of the RC (H1). The fresh calyces showed an average moisture content of 81.35% and a_w_ of 0.99, values consistent with those reported in the literature, where moisture content ranges between 78% and 85% [39,40,41] and a_w_ reaches up to 1.00 [39]. The moisture content of the calyces was reduced to 11.61% after dehydration, meeting the Brazilian legal requirement for dehydrated plant products (≤12%) (Brazil, 2005). At the same time, a_w_ dropped to 0.56, a value considered microbiologically safe for storage (<0.6) [39]. These values are consistent with those observed for sun-dried RC, with moisture content ranging from 11.08% to 12.03% and a_w_ values between 0.51 and 0.54 [40,42,43]. Reducing moisture content and a_w_ to appropriate levels is essential to inhibit microbial growth and ensure product stability during storage [44,45].

The TA (g of malic acid 100 g^−1^) was 2.95 in the fresh RC, a value similar to that reported by [46] (2.42), but lower than that described by [40] (5.15). In the DC, TA was 18.66, falling within the range observed for DC in Malaysia using different drying methods (16.02–19.12) [40]. The pH also varied between samples: 2.91 for fresh RC and 2.42 for DC, aligning with the values reported by the same authors [40], who reported pH values of 2.7 and 2.35 for fresh and DC samples, respectively. These results are also consistent with those of other studies [46,47], which reported pH values between 2.33 and 2.49 for fresh samples.

Regarding SS, the fresh RC presented a Brix value of 0.70°, which is higher than the Brix value of 0.40° reported by [39] for samples from Malaysia. For the dehydrated samples, the value was 1.93 °Brix, which is within the range observed for different drying methods, ranging from 1.70 to 3.8 °Brix [39,40]. The color parameters showed that fresh RC had higher values of L* (21.13) and C* (19.10) than the dehydrated samples (L* = 18.93; C* = 15.84). The reduction in these values is commonly observed in dried fruits and flowers, mainly due to factors such as elevated temperature, oxidative reactions, and enzymatic activity during the drying process [48,49,50]. Despite the decrease in L* and C* values, the characteristic red was preserved, with h° values of 22.91° and 24.16° for the fresh RC and DC, respectively.

Regarding compounds with bioactive properties, the content of TPC (as mg GAE g^−1^) was 1.33 in fresh RC, a value lower than that reported by [51] (55.38) for aqueous extracts of fresh RC. After dehydration, the TPC increased to 20.60, which is in agreement with previous literature [52,53]. Similarly, the TAC content (as mg mv-3,5-glc g^−1^) increased from 0.21 to 6.10, reflecting the effect of water removal during drying. The low levels of TPC and TAC in fresh RC may be associated with mucilage, which forms a viscous gel upon contact with water, hindering the extraction of anthocyanins [54]. In addition, the intact cellular structure of the fresh matrix may have limited solvent access to intracellular compartments containing PC. Conversely, dehydration promotes cell wall disruption and bioactive constituent concentration, as previously observed in different plant matrices, where dried and ground samples exhibit higher extraction efficiency than fresh ones [55,56].

The DC samples showed low moisture content and a_w_, ensuring improved stability and microbiological safety while preserving the characteristic red hue. Given these properties, only the DC samples were used as a potential bioingredient for the subsequent syrup development.

3.2. Characterization of Phenolic DC Extracts Using Different Extraction Solvents

Aqueous and hydroalcoholic solvents were evaluated using a hydromethanolic extract as a control to compare extraction efficiency. Extracts were analyzed for TPC (mg GAE g^−1^) and anthocyanin profiles. TPC ranged from 14.51 ± 0.54 to 17.48 ± 0.65, with the aqueous extract (16.15 ± 0.81) showing lower values than the control (17.48 ± 0.65) but higher than the hydroalcoholic extract (14.51 ± 0.54) (p ≤ 0.05), indicating that water is a competitive and sustainable alternative for obtaining PC-rich extracts.

Two major anthocyanins were identified and quantified based on their chromatographic characteristics (molecular and product ions under ESI-MS/MS and retention times): delphinidin (dp)-3-sambubioside and cyanidin (cy)-3-sambubioside (Table 2). The ESI-MS/MS spectrum of peak 1 in positive ionization mode showed the molecular ion ([M]+) for the dp derivative at m/z 597 and a single product ion at m/z 303, generated by the loss of a sambubiose moiety ([M–294]+), consisting of glucose ([M–162]+) linked to xylose ([M–132]+). Peak 2 exhibited a molecular ion ([M]+) for the cy derivative at m/z 581 and a single product ion at m/z 287, also resulting from the loss of the sambubiose unit ([M–294]+). These results are consistent with those of previous studies on roselle varieties from different regions [10,11,12,35].

The total concentration of these anthocyanins is influenced by several factors, including cultivar, harvest year, edaphoclimatic conditions, extraction methods, and solvents [14,57]. In this study, the comparison of the quantitative results obtained using the three extraction solvents (methanol, water, and ethanol) revealed variations in the anthocyanin content. The concentrations (mg cy-3-glc kg^−1^) in the three extracts ranged from 1579.16 to 2755.81 for dp-3-sambubioside and from 1146.93 to 1842.89 for cy-3-sambubioside, being significantly higher (p ≤ 0.05) in the aqueous extract. Moreover, the aqueous extracts presented the highest total anthocyanin content (p ≤ 0.05), extracting 30.7% more anthocyanins than the hydromethanolic extracts and 68.7% more than the hydroalcoholic extracts. Although these values are 56.5% lower than those reported for aqueous infusions from another region [11], they reinforce the efficiency of the aqueous extract, highlighting its potential as a sustainable and effective alternative for food applications, including syrup production.

3.3. Physicochemical Characteristics, PC, and Color of the Control Extract and Syrups During Refrigerated Storage

The physicochemical properties of the syrups prepared from the aqueous phenolic extract with varying sucrose concentrations (E_0_ (0%, control), E_5_ (5%), E_15_ (15%), E_25_ (25%), and E_35_ (35%)) were evaluated. The samples were stored under refrigeration (4 °C) for 30 days, and the analyses were conducted on days 0 and 30.

The physicochemical results (a_w_, SS, TA, and pH) are shown in Table S1. The two-way ANOVA (storage time × sucrose concentration) indicated that there was no significant interaction between the factors (p > 0.05), demonstrating that the observed changes were mainly due to the isolated effects of either time or sucrose concentration (Table S1). Specifically, the sucrose concentration significantly affected TA, SS, and a_w_ (p ≤ 0.05). An increase in sucrose concentration led to a reduction in TA and a_w_ and the expected increase in SS. However, storage time significantly affected TA and a_w_ values (p ≤ 0.05), both of which tended to decrease after 30 days of refrigeration (Figure 2). These variations may be associated with the structural properties of sucrose, which impact water mobility and activity [58,59], as well as the matrix viscosity [60,61], thereby altering the system’s physicochemical balance. Additionally, storage time promoted a significant decrease in TA (p ≤ 0.05), possibly due to organic acid degradation and conversion and their interaction with PC. A similar behavior was observed during the storage of fruit wines, such as mortiño wine, in which the reduction in acidity was attributed to these same mechanisms [62].

In addition to variations in pH, TA, SS, and a_w_, the ANOVA revealed a significant interaction effect (p ≤ 0.05) between storage time and sucrose concentration on TPC (expressed as mg GAE 100 g^−1^). Table S2 summarizes these results, while Figure 3a illustrates the trends observed throughout the storage period.

The E_0_ showed a significant decrease (p ≤ 0.05) in TPC from 1495.86 to 1399.82 mg GAE 100 g^−1^ after 30 days of refrigerated storage, corresponding to a 6.42% loss (Figure 3a). However, this trend was not consistent across the sucrose-containing formulations. In the sample with 15% sucrose (E_15_), degradation was the most pronounced (5.14%), whereas the samples with 5% (E_5_), 25% (E_25_), and 35% (E_35_) sucrose maintained the highest TPC levels at the end of storage, ranging from 1533.86 (E_5_) to 1678.71 mg GAE 100 g^−1^ (E_25_) on day 0 and from 1548.99 (E_5_) to 1852.89 mg GAE 100 g^−1^ (E_35_) on day 30 (Figure 3a). This behavior suggests that sucrose, at both low (5%) and high (25% and 35%) concentrations, may contribute to PC protection, albeit with varying effectiveness depending on the concentration.

The higher TPC retention in these formulations may be associated with multiple factors. Sucrose may influence PC stability by modifying molecular interactions, reducing precipitation, and promoting the formation of more stable complexes [63]. The potential interactions between PC and the aqueous matrix may have contributed to the observed variations in TPC levels throughout storage [64]. However, when analyzing the factors individually, only sucrose concentration showed a significant effect (p ≤ 0.05), as illustrated in Figure 3b, while storage time had no significant impact (p > 0.05). This effect was more evident in the samples with higher sucrose concentrations (E_25_ and E_35_), which had the highest TPC values. This finding suggests that the observed increases may be attributed to PC preservation and potential analytical interferences in the Folin–Ciocalteu method. This method is known to be susceptible to interference from organic compounds such as ascorbic acid, aromatic amines, organic acids, metal ions (e.g., Fe^2+^), and reducing sugars, which are possible sucrose degradation byproducts during storage [65]. Therefore, the elevated TPC values may result from the protective mechanisms conferred by sucrose and potential analytical artifacts related to the method employed.

Similar to what was observed for the TPC values, ANOVA indicated a significant effect (p ≤ 0.05) of the interaction between the factors (storage time × sucrose concentration) on the TAC content (Table S2 and Figure 4a). This result indicates that the variation in TAC levels over time depended on the sucrose concentration, suggesting a modulatory effect, although with non-uniform behavior among the formulations. At the initial time point (day 0), the E_5_ (531.60 mg mv-3,5-glc 100 g^−1^) and E_15_ (532.11 mg mv-3,5-glc 100 g^−1^) formulations showed the highest TAC contents (p ≤ 0.05), with no significant differences compared to E_0_ (495.26 mg mv-3,5-glc 100 g^−1^). After 30 days of refrigerated storage, a reduction in TAC content was observed in all formulations, with varying magnitudes: the greatest percentage loss occurred in E_15_ (21.8%), followed by E_5_ (13.72%) and E_25_ (9.41%). The E_0_ formulation showed a reduction of 9.07%, while the smallest loss was observed in E_35_ (8.88%). Separate evaluation of the factors revealed that both storage time (Figure 4c) and sucrose concentration (Figure 4b) had significant effects (p ≤ 0.05) on TAC levels. Sucrose concentration exerted a modulatory effect, with distinct differences among the formulations, and the E_5_ formulation (495.126 mg mv-3,5-glc 100 g^−1^) exhibited the highest average content, significantly differing from the others (p ≤ 0.05). In contrast, storage time led to a progressive decrease in TAC levels, even under refrigeration, resulting in a final average of 446.02 mg mv-3,5-glc 100 g^−1^ and a total loss of 12.71%.

This behavior suggests that low (5%) and high sucrose concentrations (25% and 35%) were more effective in retaining anthocyanins over time, while intermediate concentrations, such as 15%, may not be ideal for the stability of these pigments. This pattern may be related to the complexity of the aqueous extract matrix used, which has a more complex composition than purified anthocyanin solutions, although it is not representative of complete food systems. This complexity is also reflected in the literature, which demonstrates the varying effects of sucrose on anthocyanin stability depending on multiple experimental factors [58]. The impact of sucrose on anthocyanin stability remains controversial in the literature. Sucrose has been associated with reduced water mobility, limited oxygen diffusion, and deleterious chemical reaction inhibition, such as condensation or enzymatic reactions [58,59,66]. However, the reported effects vary widely depending on the concentration used, the nature of the matrix, the type of sugar employed, and the storage conditions [58,59,67,68,69].

Han et al. [68] observed that purified anthocyanins from four cultivars of colored potatoes (Solanum tuberosum L.) (W2, W8, R69, and R16–5) showed good retention at moderate sucrose concentrations (0–0.1 g/mL). However, a gradual decline in retention occurred over 9 days at 4 °C. On the other hand, Aaby and Amundsen [67] reported that the addition of 10% sucrose to lingonberry juice (Vaccinium vitis-idaea L.) stored at 6 °C for 16 weeks resulted in variations in anthocyanin retention (37–40%), a value similar to that observed after two weeks at 22 °C. Furthermore, while some authors report protective effects resulting from a reduction in a_w_ through the addition of sucrose, which binds water molecules and restricts their mobility, thereby reducing the availability of free water for degradation reactions. These effects were particularly observed at sucrose concentrations between 40 and 60%, which lowered aw to values ranging from 0.91 to 0.84 in model systems containing hibiscus anthocyanins subjected to moderate heating (30–40 °C) [59]. However, others indicate that under certain conditions, such as high temperatures (100–140 °C) combined with low a_w_ (0.34), this same reduction may exacerbate anthocyanin degradation, as observed in blackberry juice [69]. In this context, the decreased water availability favors the formation of reactive compounds, such as 5-hydroxymethylfurfural (HMF), which can interact with anthocyanins, promoting polymerization and browning reactions, thereby accelerating their degradation and highlighting the complexity and multifactorial nature of these interactions.

Among the color parameters (L*, C*, and h°), ANOVA indicated a significant interaction effect (p ≤ 0.05) between the factors (sucrose concentration × storage time) only for L* and h° (Table S3). L* values (Figure 5) increased after storage in E_0_, indicating greater lightness of the samples and possible pigment degradation. In the syrup, formulations with higher sucrose content (E_15_, E_25_, and E_35_) showed a decrease in L* values over storage. In contrast, the E_5_ formulation maintained stable lightness (p > 0.05), indicating that the matrix has greater optical consistency at this concentration. A similar trend was observed for h°, with less variation over time in E_5_, indicating greater color hue stability at this sucrose level.

Considering only the sucrose concentration factor, a significant effect (p ≤ 0.05) was observed on the L* and C* chromatic parameters. The L* values were significantly lower in the E_5_ formulation and higher in E_25_ and E_35_ formulations (Figure 6a). Similarly, the C* parameter showed a linear trend, with the highest values observed in the E_25_ and E_35_ formulations (Figure 6b).

This result supports the hypothesis that different sucrose concentrations promote changes in color intensity, possibly due to system optical density modifications and interactions between dissolved solids and the aqueous matrix [70,71]. Such changes may cause a hyperchromic effect, increasing light absorption by the pigments and enhancing color perception, as previously reported in studies with roselle anthocyanins [59]. These interactions may also directly affect the refractive index and light scattering, thereby affecting the visual perception of color [71].

Furthermore, a significant effect was observed on h° values (Figure 6c), with a trend of increasing average values with the addition of sugar to the extract, although no clear linear pattern was identified. The E_5_ (35.11°) and E_25_ (35.29°) samples showed the lowest average h° values among the syrup formulations, with no statistically significant differences compared to the other formulations (p > 0.05). Additionally, storage time also significantly influenced (p ≤ 0.05) h° values, with an average increase after 30 days (Figure 6d), although the magnitude of this variation was minor.

Finally, we calculated the total color difference (∆E), a parameter widely used to quantify perceptible changes in the color of food matrices during storage [33]. After 30 days, ∆E values ranged from 0.44 (E_5_) to 2.96 (E_0_), indicating varying degrees of color change among the formulations. According to the criteria proposed by Pathare et al. (2013) [33], ∆E variations below 1.5 are classified as imperceptible or barely perceptible, whereas values between 1.5 and 3 indicate moderate chromatic differences. E_5_ (0.44) and E_15_ (1.33) showed high color stability, whereas E_35_ (2.05), E_25_ (2.80), and notably E_0_ (2.96) exhibited visually perceptible changes, although still below the threshold considered highly distinguishable (∆E > 3) (Figure 5).

PCA was applied to facilitate multivariate visualization of the data for the E_0_ and the syrup formulations containing different sucrose concentrations, evaluated at 0 and 30 days of storage. As shown in Figure 7, the first three principal components together explained 87.91% of the total data variance, with 57.79%, 16.08%, and 14.14% attributed to the first (PC 1), second (PC 2), and third (PC 3) components, respectively. Supplementary Table S4 provides a complete description of variable contributions to PC 1, PC 2, and PC 3. The selection of these components is supported by the scree plot of eigenvalues (Figure S1).

PC 1 was primarily associated with C* and h°, as well as TPC and SS concentrations, with factor loadings ≥ 0.7. These variables characterized the syrup formulations with higher sucrose concentrations (E_25_ and E_35_), located in the positive quadrant of the PC1 axis. Conversely, in the negative quadrant, PC 1 was correlated with TA and a_w_ values (factor loadings ≤ −0.7), describing the E_0_ and those with lower sucrose concentrations (E_5_ and E_15_). Thus, variables related to increased sucrose content exhibited behavior opposite to those associated with lower sucrose concentrations, highlighting a negative correlation between these groups. These results indicate that the addition of sucrose distinctly influenced the physicochemical and color parameters, emerging as the main factor responsible for the sample separation in the PCA.

Although accounting for a smaller proportion of the total variance, PC 2 was negatively correlated with pH values (factor loadings ≤ −0.7). Samples with higher pH values were located in the negative scores of this component, whereas those with lower pH values were in the positive scores. Thus, although with less impact than the variables related to PC 1, pH also contributed to the separation of the samples observed in the analysis, independent of the sucrose concentration. The separation between storage times was mainly observed along PC 3, which was negatively correlated with TAC content, showing factor loadings ≤ −0.7. Samples evaluated at time zero had negative scores on this component, reflecting higher TAC concentrations. Moreover, those stored for 30 days were located in the positive scores, indicating degradation of these compounds over time. Therefore, although storage time had a less expressive effect on the total explained variance, it was still a relevant factor for separating the samples along PC 3.

Based on the obtained results, an integrated analysis of physicochemical parameters, bioactive compound stability, and color attributes during 30 days of refrigerated storage supported the selection of the syrup containing 5% sucrose (E_5_) for food applications. Syrup E_5_ exhibited superior technical performance among the evaluated formulations, achieving the highest TAC concentration considering the influence of sucrose content, with a relative loss of 13.72% after storage, while maintaining high TPC levels and minimal color changes. From a chromatic standpoint, E_5_ showed good color stability, with the lowest total color difference (ΔE = 0.44) among all formulations, indicating that changes during storage are imperceptible to the human eye. L* and h° values also remained stable, suggesting consistent visual appearance and preservation of optical characteristics of the pigment.

From a technological standpoint, the formulation containing 5% sucrose offers a balanced compromise. While lower sucrose levels may reduce caloric contribution, they may also affect microbial stability; thus, their application should be evaluated according to the intended product formulation and storage conditions.

Therefore, considering the balance between functionality, physicochemical, and visual stability, the syrup containing 5% sucrose is a viable and technically sound option for food applications, particularly in functional products or those targeting consumers seeking lower-sugar alternatives.

3.4. Chromatic Characterization of Extracts in Model Buffer Systems and Food Matrices

In the buffered systems, the extracts exhibited reddish hues under acidic conditions (pH 1.5–3.5), whereas higher pH values (≥5.5) led to marked visual changes, with the appearance of yellowish to greenish tones (Figure 8 and Table S5). Based on these results, extract E_0_ was applied to various commercial food matrices with different pH ranges (2.96–68.5) to assess its color behavior in real systems (Table 3).

The addition of the extract resulted in significant chromatic changes (p ≤ 0.05) across all matrices tested. Color intensification was observed in acidic systems, such as tonic water (pH = 3.0), as evidenced by the high C* value (31.77), along with a reddish hue indicated by the low h° values (25.88). In contrast, in products with pH closer to neutrality, such as milk and whipped topping (pH between 5.7 and 6.4), a significant decrease in C* (4.40 to 5.27; p ≤ 0.05) was recorded, accompanied by a chromatic shift toward yellowish hues, as indicated by higher h° values (68.18 to 80.20).

Furthermore, the addition of the extract significantly reduced the L* values (p ≤ 0.05) in all matrices, with values ranging from 25.70 (tonic water) to 86.23 (whipped topping). However, among samples containing the extract, significantly higher L* values (p ≤ 0.05) were found in matrices with pH above 4 (≥64.94), indicating that higher acidity is associated with darker and more saturated systems. These findings are consistent with the results obtained in buffered systems, where lower pH values (≤3.5) were associated with lower L* values (43.43–45.15), more reddish hues (h° = 28.14–58.52), and higher color saturation (C* = 6.82–4.40) (Table S4), reinforcing the role of pH as a determining factor in the visual expression of the pigments present.

3.5. Addition of the Control Extract and Syrup to the Food Matrices

Based on the chromatic characterization results of E_0_ in buffered systems and food matrices, the influence of sucrose (syrup containing 5% sucrose—E_5_) on the color parameters of various commercial food products was evaluated during storage (Table 4). The extract was added to matrices with acidic pH (3.0–4.5), previously identified as more favorable for the expression of extract coloration. Overall, the presence of sucrose promoted significant chromatic changes (p ≤ 0.05), particularly increasing L* and C* values, along with shifts in h° values, indicating hue modifications. For example, in tonic water, sucrose increased the L* (from 25.70 to 27.63) and C* (from 31.77 to 35.04) values and shifted the h° (from 25.88 to 29.02). Similar results were observed in fermented milk, where L* increased from 64.94 to 65.59 and h° from 45.30 to 49.68. Other matrices also showed notable changes: in maria-mole, h° decreased from 34.75 to 33.75; in yogurt, L* remained high (79.47 and 79.24 without and with sucrose, respectively), with a slight increase in C* (from 8.32 to 8.45).

This behavior is consistent with that previously observed in syrup formulations, reinforcing the fact that sucrose can influence chromatic parameters, especially L* and C* values. However, in addition to the changes in optical density and interactions between dissolved solids and the aqueous matrix introduced by sucrose [70,71], the intrinsic composition and characteristics of the matrices may also affect chromatic parameters. Fermented milk and yogurt have distinct inherent coloration, which may enhance or mask the added pigment. Moreover, in the case of anthocyanin extracts applied to dairy matrices, such as creamy desserts and milkshakes, [72] reported anthocyanin losses due to interactions between these compounds and other food components, potentially leading to the formation of insoluble aggregates or masking of the pigments’ chromophoric properties.

This data also supported a Pearson correlation analysis aimed at exploring potential linear relationships between chromatic variables and the pH values of the matrices to understand their influence on pigment color expression (Figure 9). A significant correlation was observed between chromatic parameters and the pH of each matrix: a positive correlation with h° (r = 0.89, p ≤ 0.001), a negative correlation with C* (r = 0.77, p ≤ 0.01), and a positive, though weaker, correlation with L* (r = 0.66, p ≤ 0.05). A negative correlation was also observed between L* and C* (r = 0.95; p ≤ 0.001).

In less complex anthocyanin extracts, such as those from RC, which contain diglycosylated anthocyanins, namely dp-3-sambubioside and cy-3-sambubioside [35], an increase in pH favors the hydration of the flavylium cation form, leading to the formation of colorless hemiketal species. This structural transformation is directly associated with increased L* values, decreased C*, and perceptible h° changes, including hue shifts and transitions toward less saturated regions of the visible spectrum [73,74]. Understanding the interactions between pH, sucrose, and the food matrix provides valuable insights for the technological application of RC extracts and their derived bioingredients, such as syrups, supporting the formulation of foods with stable and visually appealing coloration. Furthermore, various parts of Hibiscus extracts have demonstrated functional potential in certain products, particularly dairy-based ones, by providing antioxidant, antihypertensive, and antimicrobial properties [75,76], further reinforcing their use as promising bioingredients.

4. Conclusions

The physicochemical characterization of the DC revealed that the drying process promoted the concentration of bioactive compounds, such as anthocyanins and TPC, while significantly reducing moisture content and a_w_, achieving safe levels that favor microbiological stability. The evaluation of different extraction solvents for DC, supported by HPLC-DAD-ESI-MS/MS analysis, confirmed dp-3-sambubioside and cy-3-sambubioside as the predominant anthocyanins and supported the selection of the aqueous extract as a suitable food-grade option for syrup formulation, balancing phenolic content and food applicability.

The formulation of syrups using DC extracts with varying sucrose concentrations demonstrated that this sugar plays a modulatory role in the physicochemical properties of the system, contributing to reductions in a_w_ and TA and exerting a non-uniform influence on anthocyanin retention. PCA revealed a clear separation of samples according to sucrose concentration, which was associated with changes in colorimetric and physicochemical parameters. The effect of storage time on anthocyanin degradation was also underscored, confirming its importance for system stability. Although the formulation with 35% sucrose showed the lowest absolute anthocyanin loss after 30 days of refrigerated storage, the 5% sucrose formulation stood out for its superior color stability (∆E = 0.44) and lower relative anthocyanin losses when considering the isolated effect of time, indicating greater stability throughout storage.

Applying the E_0_ in various food matrices revealed pH-dependent color sensitivity, with red hues preserved in acidic systems and color shifts toward yellowish or greenish tones at higher pH values. Incorporating the syrup (E_5_) confirmed these findings, maintaining reddish hues in acidic food systems while affecting L* and C* parameters. Pearson’s correlation analysis supported these results, revealing a positive correlation between pH and h°, a negative correlation with C*, and a negative correlation between L* and C*, indicating reduced saturation and increased lightness in less acidic environments.

These findings underscore the potential of H. sabdariffa L. syrup as a natural and versatile bioingredient. It has promising applications in the development of acidic foods and beverages that align with the growing demand for healthier and more sustainable products. Moreover, the demonstrated physicochemical and chromatic stability during storage confirms its feasibility, ensuring the preservation of bioactive and visual characteristics.