Thermal Processing Effects on Bioactive Composition and Physicochemical Parameters of Citrus grandis Juices: A Cultivar-Specific Study

TL;DR

This study examines how pasteurization affects the nutritional and functional qualities of five types of Sicilian grapefruit juice.

Contribution

The study provides cultivar-specific insights into how thermal pasteurization impacts bioactive compounds and functional properties of Citrus grandis juices.

Findings

Pasteurization caused cultivar-dependent reductions in polyphenols, flavonoids, and carotenoids.

Fresh juices showed stronger enzyme inhibitory activities compared to pasteurized juices.

Phenolic compounds were identified as key contributors to antioxidant capacity in the juices.

Abstract

Conventional thermal pasteurization is widely applied to ensure the safety of fruit juices, although its impact on bioactive compounds and functional properties may vary according to cultivar. This study evaluated the effects of conventional pasteurization on physicochemical parameters, bioactive composition, antioxidant capacity, and enzyme inhibitory activities of juices obtained from five Sicilian Citrus grandis cultivars (Todarii, Maxima, Pyriformis, Chadock, and Terracciani). Total polyphenols, flavonoids, and carotenoids were quantified, while flavanone profiles were characterized by means of HPLC analysis. Antioxidant activity was assessed using DPPH, ABTS, FRAP, and β-carotene bleaching assays, and in vitro inhibitory activities against α-amylase, α-glucosidase, and pancreatic lipase were determined. Pasteurization led to cultivar-dependent reductions in total polyphenols (up to ~40%), flavonoids (up to ~45%), and carotenoids (up to ~25%), accompanied by decreased radical scavenging capacity and reducing power. Naringin was identified as the predominant flavanone, with thermal processing inducing both degradation and release phenomena depending on the cultivar. Fresh juices exhibited stronger enzyme inhibitory activities, particularly against α-amylase and α-glucosidase. Multivariate analysis discriminated against fresh and pasteurized juices, identifying phenolics as the main contributors to antioxidant capacity. Despite bioactive reductions, functional quality was partially preserved, supporting targeted cultivar selection for optimized industrial processing.

Figures

Full text

1. Introduction

The Citrus genus (family Rutaceae) encompasses a wide variety of species characterized by distinctive morphological and phytochemical traits. Among the most economically relevant are C. limon (lemon), C. aurantiifolia (lime), C. reticulata (mandarin), C. sinensis (orange), C. medica (citron), and C. paradisi (grapefruit) [1,2]. Within this group, Citrus grandis (L.) Osbeck, also known as C. maxima (Burm.) Merr., or pomelo, is the largest Citrus fruit and a major source of bioactive compounds such as flavonoids, carotenoids, and vitamin C. It is widely cultivated across Asia (China, Thailand, India, Vietnam) and increasingly grown in the Mediterranean region [3].

In Italy, Citrus cultivation is mainly concentrated in southern regions, particularly Calabria and Sicily, which together account for nearly 80% of the national yield [4]. Sicily, recognized as a Mediterranean biodiversity hotspot, hosts a remarkable diversity of Citrus genotypes well adapted to local pedoclimatic conditions [5]. These traditional cultivars serve as a valuable genetic reservoir, supporting both high-quality juice production and innovation in functional foods. In addition to their agronomic value, Sicilian cultivars also represent an untapped source of nutritionally valuable compounds with potential industrial applications.

Citrus juices are appreciated for their flavour and nutritional richness, containing significant levels of phenolic compounds, flavonoids, carotenoids, and vitamin C. These bioactive contribute antioxidant, anti-inflammatory, and enzyme-inhibitory properties that may help prevent chronic diseases linked to oxidative stress, obesity, and type-2 diabetes mellitus [6,7,8].

Thermal pasteurization remains the most widely applied method to ensure microbial stability and enzymatic inactivation in fruit juices. Nevertheless, heat treatment may promote the degradation of thermolabile phytochemicals, particularly carotenoids and phenolic compounds, through oxidation and isomerization reactions [9]. Previous studies on Citrus juices have reported inconsistent outcomes, largely influenced by matrix composition, cultivar, and processing parameters, thereby emphasizing the importance of genotype-specific responses in determining the final quality of the product [10,11,12]. However, comparative investigations focusing on multiple cultivars of Citrus grandis grown under identical pedoclimatic conditions and simultaneously evaluated for both compositional and functional attributes remain scarce.

Sicily represents a unique reservoir of genetic diversity for C. grandis, yet the response of its local cultivars to conventional pasteurization has not been systematically explored. In this context, the present study provides a comprehensive and cultivar-oriented evaluation of the impact of thermal pasteurization on the physicochemical parameters, phytochemical composition, antioxidant capacity, and enzyme-inhibitory activities of juices obtained from five Sicilian C. grandis cultivars (Todarii, Maxima, Pyriformis, Chadock, and Terracciani). Unlike previous reports that predominantly examined single cultivars or focused solely on compositional changes, this work integrates detailed chemical profiling with functional bioactivity assays under standardized processing conditions, thereby enabling a direct comparison of cultivar-dependent responses. By elucidating differential patterns of bioactive compound retention and biological activity after pasteurization, this study provides novel insights into genotype-specific thermal stability and offers practical guidance for the selection of C. grandis cultivars better suited for industrial juice production with improved functional quality.

2. Materials and Methods

2.1. Chemicals and Reagents

All reagents were obtained from Sigma-Aldrich S.p.A. (Milan, Italy), and analytical-grade solvents were purchased from VWR International S.r.l. (Milan, Italy). Acarbose from Actinoplanes sp. was supplied by Serva (Heidelberg, Germany). Flavanone standards (eriocitrin, neoeriocitrin, narirutin, naringin, hesperidin, and neohesperidin) were purchased from Extrasynthese (Genay, France). Hesperidin, naringin, eriocitrin, and neohesperidin showed HPLC purity ≥ 98%, whereas neoeriocitrin and narirutin showed HPLC purity ≥ 99%. Ultrapure water and acetonitrile (HPLC grade) were obtained from Carlo Erba (Italy).

2.2. Plant Materials

The fruits of Citrus grandis L. Osbeck (syn. Citrus maxima (Burm.) Merr.) cultivars Todarii (No. 107982), Maxima (No. 107983), Pyriformis (No. 107984), Chadock (No. 107985), and Terracciani (No. 107986) were harvested in March 2019 in Palermo, Sicily, Italy (38°06′48.39″ N; 13°22′21.68″ E) (Figure 1 and Figure 2).

Fruits were selected based on uniform size, external appearance, and absence of visible defects. For each cultivar, three independent batches (blocks) of ten fruits were collected to account for biological variability.

–cv. Todarii: leaves oval–oblong with slightly winged petioles; fruits medium to large, round, and slightly flattened at the base. Peel yellow, moderately thick with a soft albedo; pulp yellowish and mildly acidic.–cv. Maxima: leaves oval–oblong, slightly curled, with narrow winged petioles; fruits very large and cylindrical, constricted near the stem. Peel yellow and thick with a spongy pinkish albedo; pulp yellow–pink, juicy, and aromatic.–cv. Pyriformis: leaves oval–oblong with slightly winged petioles; fruits pear-shaped and medium–small, tapering at the apex. Peel yellow and thin with a soft white albedo; pulp pale yellow, juicy, and distinctly sour.–cv. Chadock: leaves oval–oblong with clearly winged petioles at branch tips; fruits large, slightly flattened, and rounded with a shallow apical depression. Peel yellow and moderately thick with a spongy pinkish albedo; pulp pink, juicy, and slightly bitter.–cv. Terracciani: leaves oval–oblong with slightly winged petioles; fruits small and oval with a visible style and stigma. Peel yellow and thin with a soft white albedo; pulp pale yellow and sour tasting.

A fruit harvested from every variety, identified by Prof. Rosario Schicchi, has been stored in the University of Palermo Herbarium (Voucher No. 107982–107986).

2.3. Sample Preparation

The juice was extracted by manual squeezing using a commercial kitchen juicer (MACOM Just Kitchen 859, Macom, Milan, Italy). The obtained juice was filtered to remove pulp and seeds, then centrifuged at 3000 rpm for 15 min. The resulting supernatant was further clarified through a 0.45 µm membrane filter (Millipore Corporation, Bedford, MA, USA). All operations were performed under subdued light to minimize photodegradation of light-sensitive compounds. For pasteurization, aliquots of 250 mL were placed in a KonfiStar 60 Digital Pasteurizer (Klarstein, Berlin, Germany) and heated to 90 °C for 10 min, according to Etzbach et al. [9]. Both pasteurized and untreated (fresh) juices were transferred into sterilized glass bottles and stored at −18 °C until analysis (Table 1).

2.4. Physicochemical Analysis of C. grandis Fruits and Juice

Fruits were collected in Palermo and inspected to verify their integrity and absence of contamination from insects and dust. Physical characteristics, including fruit weight (g), average width (cm), average height (cm), flavedo thickness (mm), albedo thickness (cm), and juice yield (%), were determined. Samples were stored at −18 °C until analysis (Supplementary Materials, Table S1).

The colour of the fresh juice of C. grandis was evaluated using a colorimeter (model CSM-4, PCE Instruments, Capannori (LU), Italy). Lightness (L*) and chromatic parameters (+a*: red, −a*: green; +b*: yellow, −b*: blue) were measured. Colour characteristics were further described using the Chroma index (C*) and Hue angle, [Equations (1) and (2)], which indicate the degree of saturation (colour intensity) and the relative contributions of redness and yellowness, respectively:

$$[eqn]$$

where C* (chroma) represents the colour intensity derived from the a* (red/green) and b* (yellow/blue) values.

$$[eqn]$$

where Hue° (hue angle) represents the dominant hue of the juice color, obtained from the arctangent of the ratio between the b* (yellowness) and a* (redness) parameters.

The pH was measured at room temperature using a pH meter (Model Basic 20, Crison Instruments, Barcelona, Spain), calibrated with standard buffers at pH 4 and pH 7. Total soluble solids (TSS) were determined using a digital refractometer PR-201α (Atago, Tokyo, Japan), previously calibrated at 20 °C. Results are expressed in degrees Brix (°Brix). Ascorbic acid content was quantified via potentiometric titration of the juice with a 2,6-dichloroindophenol solution. All measurements were performed in triplicate [6].

2.5. Quantification of Phenolic, Flavonoid, and Carotenoid Compounds

The Total Phenolic Content (TPC) was measured using the Folin–Ciocalteu method [13]. In brief, 0.1 mL of C. grandis juice was combined with 0.2 mL of Folin–Ciocalteu reagent, 1 mL of 15% (w/v) sodium carbonate, and 2 mL of distilled water. The mixture was incubated at room temperature for 2 h, and absorbance was recorded at 765 nm using a UV–vis spectrophotometer (Jenway 6003, Nottingham, UK). All measurements were performed in triplicate, and results were expressed as mg gallic acid equivalents (GAE) per 100 mL of juice.

The Total Flavonoid Content (TFC) was determined by mixing 0.1 mL of C. grandis juice with an equal volume of 2% aluminium chloride solution. After incubation at room temperature for 15 min, absorbance was measured at 510 nm [13]. Analyses were carried out in triplicate, and values were expressed as mg quercetin equivalents (QE) per 100 mL of juice.

The Total Carotenoid Content (TCC) was quantified following the procedure described by Gu et al. [14] and [15]. One millilitre of juice was mixed with 0.5 mL of 5% NaCl solution, vortexed for 30 s, and centrifuged at 4500 rpm for 10 min. A 100 µL aliquot of the supernatant was diluted with 0.9 mL of n-hexane, and absorbance was measured at 460 nm. Results were expressed as mg β-carotene equivalents per 100 mL of juice.

2.6. High-Performance Liquid Chromatography (HPLC) Analysis

Citrus grandis juice was filtered using a 0.45 μm membrane filter. The separation of bioactive compounds was carried out using an HPLC system equipped with a Phenomenex C18 column (150 mm × 3 mm, 5 µm), following the official methods of the International Federation of Fruit Juice Producers [6].

Two buffers were employed as the mobile phase: solvent A (acetonitrile/water/phosphoric acid, 70:26:4) and solvent B (20 mM potassium dihydrogen phosphate, pH 3.5). The gradient elution program was as follows: initial condition, 85% A and 15% B; at 5 min, 70% A and 30% B; at 20 min, 50% A and 50% B; at 30 min, 25% A and 75% B; at 35 min, 5% A and 95% B; and at 40 min, returning to 85% A and 15% B. The column was maintained at 25 °C, with a flow rate of 1 mL/min. Chromatograms were monitored at a wavelength of 287 nm.

Compound identification and quantification were performed by comparing retention times, UV spectra, and calibration curves with those of the external standards, using a photodiode array detector. The following compounds were selected as standards: eriocitrin, neoeriocitrin, narirutin, naringin, hesperidin, and neohesperidin.

Calibration was carried out by injecting standard solutions at five different concentrations, in triplicate, for each compound. Concentration ranges were selected based on the levels expected in the samples. Calibration curves passed through the origin and demonstrated good linearity, with coefficients of determination (R^2^) ranging from 0.9975 for naringin to 0.9999 for eriocitrin.

The limits of detection (LOD) (Equation (3)) and quantification (LOQ) (Equation (4)) were calculated as the minimum concentrations at which the compounds could be detected (LOD) and reliably quantified (LOQ), using the following equations:

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

where: SD is the standard deviation of ten replicate injections of the blank and S is the slope of the calibration curve for each analyte.

The method was validated for LOD and LOQ, ensuring reliable sensitivity and reproducibility of the analysis.

2.7. In Vitro Antioxidant Activity

2.7.1. Radical Scavenging and Reducing Power Assays (DPPH, ABTS, and FRAP Test)

The in vitro antioxidant activities of all C. grandis juices were evaluated using 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), and Ferric Reducing Antioxidant Power (FRAP) assays as reported in Sicari et al. [16].

2.7.2. β-Carotene Bleaching Test

Protection against lipid peroxidation was assessed using the β-carotene bleaching assay as described by Sicari et al. [16] with minor modifications. Briefly, an emulsion was prepared by mixing β-carotene, linoleic acid, and Tween 20. Juice extracts, at concentrations ranging from 2.5 to 100 µg/mL, were added to the emulsion and incubated for 30- and 60-min. Absorbance was recorded at 470 nm using a UV–Vis spectrophotometer (Jenway 6003, Milan, Italy) against a reagent blank. Propyl gallate was used as a positive control. Results were expressed as percentage inhibition (%).

2.8. Carbohydrate Hydrolysing Enzymes and Pancreatic Lipase Inhibitory Effects

The inhibitory activities on key digestive enzymes (α-amylase, α-glucosidase, and pancreatic lipase) were evaluated following previously reported protocols [13]. All assays were performed in triplicate. Appropriate blank corrections were applied to remove background absorbance and ensure accurate quantification of enzyme inhibition, according to the approach described by Lankatillake et al. [17].

For the α-amylase inhibition assay, a 0.5% (w/v) starch solution was prepared by dispersing potato starch in 20 mM sodium phosphate buffer containing 6.7 mM sodium chloride. The enzyme solution was made by dissolving 25.3 mg of porcine pancreatic α-amylase (EC 3.2.1.1, 10 units/mg) in 100 mL of cold distilled water. A mixture of 75 µL enzyme solution and 15 µL fruit juice sample was combined with 150 µL starch solution and incubated at 25 °C for 5 min. The production of maltose was quantified colorimetrically using 3,5-dinitrosalicylic acid (DNS) at 540 nm. Acarbose served as the positive control at an equivalent concentration to the juice sample.

In the α-glucosidase inhibition assay, a 4% (w/v) maltose solution was prepared by dissolving 12 g of maltose in 300 mL of 50 mM sodium acetate buffer. The enzyme solution was prepared by dissolving 1 mg of α-glucosidase from Saccharomyces cerevisiae (EC 3.2.1.20, 10 units/mg) in 10 mL of ice-cold distilled water. C. grandis juices (5 µL) were mixed with 250 µL of maltose solution and incubated at 37 °C for 5 min. Then, 5 µL of α-glucosidase solution was added, and the mixture was incubated at 37 °C for 30 min. The reaction was stopped by adding 50 µL of perchloric acid solution, followed by centrifugation to collect the supernatant. The supernatant was then mixed with 5 µL o-dianisidine (DIAN) solution, a reagent for glucose quantification, and 10 µL of peroxidase/glucose oxidase (PGO) colour reagent solution and incubated at 37 °C for 30 min. Absorbance was measured at 492 nm. Acarbose served as the positive control at an equivalent concentration to the juice sample.

Pancreatic lipase inhibitory activity was determined using a 96-well plate assay [18]. A 5 mM solution of 4-nitrophenyl octanoate (NPC) in dimethyl sulfoxide was prepared, along with an aqueous solution of porcine pancreatic lipase (1 mg/mL) and Tris-HCl buffer (pH 8.5).

C. grandis juices (6 µL) were added to each well, followed by 6 µL of enzyme solution, 6 µL of NPC, and 279 µL of buffer. The mixture was incubated at 37 °C for 30 min, and absorbance was measured at 405 nm. Orlistat was used as a positive control.

Blank correction procedures were identical for all enzyme inhibition assays (α-amylase, α-glucosidase, and pancreatic lipase). Background absorbance due to juice colour and non-enzymatic reactions was corrected according to the method of Lankatillake et al. [17] using the following equations:

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

where $[eqn]$ and $[eqn]$ are the blank-corrected absorbances of the sample and control reactions, respectively. $[eqn]$ and $[eqn]$ are the raw absorbances, while $[eqn]$ (sample blank, juice + substrate without enzyme) and $[eqn]$ (control blank, substrate + buffer without juice or enzyme) are used to correct for the intrinsic colour of the juice and the baseline absorbance of the reaction medium.

The percentage of inhibition was calculated using Equation (7):

$$[eqn]$$

where $[eqn]$ and $[eqn]$ are the blank-corrected absorbances of the sample and control reactions, calculated as described in Equations (6) and (7). These corrections ensured the removal of any background optical interference.

Equations (5) and (7) were applied uniformly for α-amylase, α-glucosidase, and lipase inhibition assays.

2.9. Statistical Analysis

All experiments were performed in triplicate and repeated independently on three different days, resulting in a total of nine measurements per sample (n = 9), corresponding to three independent experiments each including three technical replicates. Data are expressed as mean ± standard deviation (SD) of nine independent values. The dataset containing the parameters analysed in Citrus grandis juice samples was subjected to one-way analysis of variance (ANOVA) using GraphPad Prism version 10.0 for Windows (GraphPad Software, San Diego, CA, USA). Post hoc comparisons were performed using Tukey’s test, and statistically significant differences are indicated by different lowercase letters. A dual statistical approach was adopted to evaluate both the overall effects of pasteurization across all cultivars and the specific differences within each individual cultivar. Unpaired two-tailed t-tests were used to compare treatments within the same cultivar (UT vs. PAST). The same test was also applied by pooling all untreated juices and comparing them with all pasteurized juices combined to assess the overall treatment effect. Statistical significance was set at p < 0.05 (), p < 0.01 (), and p < 0.001 (); non-significant results are indicated as ns. Additional analyses, including assessment of repeatability, linear regression, mean, and relative standard deviation, were performed using GraphPad Prism. The Relative Antioxidant Capacity Index (RACI) was calculated according to the procedure described by Todorovic et al. [19]. This statistical approach integrates the antioxidant capacity values obtained from different in vitro assays to provide a comprehensive comparative evaluation. In the present study, data from the four antioxidant tests were used to calculate the RACI values using the following Equation (8):

$$[eqn]$$

where x represents the raw data, μ is the mean, and σ is the standard deviation.

Pearson’s correlation coefficient (r) was used to assess linear associations among continuous variables, after confirming approximate normal distribution. Principal Component Analysis (PCA) was carried out using SPSS Statistics software, version 22.0 (IBM, Chicago, IL, USA).

3. Results and Discussion

3.1. Juice Quality Parameters

The analysis shows that juice content is not directly proportional to fruit size. The cultivar Maxima, despite having the highest fruit weight and dimensions (width, height, and especially albedo thickness), exhibits a relatively low juice percentage (14.1%). This suggests that the larger size is mainly associated with a greater structural component (peel and albedo) rather than a higher edible liquid fraction.

In contrast, Terracciani, characterized by smaller fruits and reduced flavedo and albedo thickness, shows the highest juice percentage (37%). In this case, the lower incidence of peel tissues appears to favour a greater juice yield.

Chadock and Pyriformis display intermediate conditions: although their fruits are larger than those of Terracciani, they maintain medium–high juice percentages (22.2% and 20.0%, respectively), suggesting a balance between morphological development and juice yield.

Overall, an inverse relationship emerges between peel components (particularly albedo thickness) and juice percentage: cultivars with thicker peel tissues tend to have lower juice yield, whereas smaller fruits with thinner peel tissues are proportionally richer in juice. This relationship has important implications for varietal selection depending on the intended use, especially for processing industries.

Citrus grandis juice colour, pH, total soluble solids (TSS, °Brix), and ascorbic acid content were analysed, and the results are presented in Table 2 and Table S2.

The pH represents a crucial quality attribute influencing both flavour and the stability of bioactive compounds in fruit juices [20]. Across the five cultivars, pH values ranged from 2.3 to 3.1, consistent with the typical acidity of Citrus juices, and were only slightly affected by heat treatment. Pasteurized samples generally exhibited marginally higher or comparable pH values relative to the untreated ones, confirming that the mild thermal conditions applied did not promote acid degradation or decarboxylation reactions.

The TSS values varied between 7.8 and 11.1 °Brix and were comparable before and after pasteurization across all cultivars, with no statistically significant differences observed. These variations remained within analytical variability and aligned with previous findings showing negligible changes in °Brix after short-time pasteurization [21,22].

Ascorbic acid concentrations ranged from 102.8 to 246.7 mg/L, displaying high cultivar-dependent variability but no statistically significant differences between untreated and pasteurized samples (p > 0.05). In our samples, decreases were observed only in the Maxima and Terracciani cultivars, while the other juices exhibited higher or unchanged vitamin C levels after pasteurization. This observed trend, particularly the slight increase in cultivars such as Todarii and Pyriformis, may be explained by the thermal inactivation of ascorbate oxidase, which is the primary enzyme responsible for the enzymatic degradation of ascorbic acid. In unpasteurized samples, this enzyme can remain active during the initial stages of juice preparation, leading to a higher degradation compared to pasteurized juices, where the enzyme is immediately denatured. Furthermore, the heat treatment likely promoted the disintegration of cellular structures, enhancing the analytical extractability of ascorbic acid from the fruit matrix and resulting in a more efficient recovery during the assay [23,24] Additionally, the protective effect of phenolic compounds, which are known to shield vitamin C from oxidative decomposition and can increase in response to processing stress, might have further contributed to the high retention of this nutrient Similar trends have also been described by Lee & Coates [10] and Mandha et al. [25] for other Citrus species, confirming that moderate pasteurization generally preserves ascorbic acid integrity.

Colour is a crucial visual indicator for assessing the quality of fruit juice, and it influences consumer-purchasing decisions. For this reason, this study aimed to measure the effect of pasteurization, on the CIE Lab parameters and Hue angle of our samples (Table 2). Lightness (L*) values, which indicate the brightness of the juice, showed a slight but statistically significant decrease after pasteurization (p < 0.05) when considering all cultivars, suggesting a general trend toward a darker appearance due to heat treatment. Specifically, the Maxima cultivar, L* value decreased from 40.9 to 38.1, and Terracciani from 39.7 to 38.1. However, the differences between untreated and pasteurized samples were not statistically significant within each individual cultivar. These changes were more pronounced following thermal treatment, primarily due to a decrease in lightness at higher temperatures, consistent with previous reports. Rabie et al. [26] observed an L* value of 41.3 for fresh Physalis peruviana juice, which decreased to 38.1 after pasteurization. Similarly, Salar et al. [27] reported a reduction in lightness in Citrus maqui beverages after thermal processing, aligning with the trends observed in our study.

Chroma values, reflecting color saturation, were comparable across treatments, with no statistically significant differences observed among the cultivars; Maxima exhibited a pronounced increase from 8.0 to 13.3, which was statistically significant within that cultivar. This might indicate an enhancement in colour vividness due to changes in pigment concentration or matrix effects [28].

The hue angle (Hue°), representing the perceived color tone of the juice, showed numerically higher values after pasteurization in the Maxima and Terracciani cultivars (about 1.4° vs. 1.3°), although the differences were not statistically significant. This suggests that the juice colour remained stable after heat treatment, although small changes might occur due to pigment transformation or interactions within the juice matrix [29].

3.2. Phytochemical Profile

To investigate the impact of thermal processing on the phytochemical composition of Citrus grandis, juice samples from five cultivars were analysed using HPLC (Table 3). In addition, total polyphenol content (TPC), total flavonoid content (TFC), and total carotenoid content (TCC) were preliminarily assessed as global indicators of bioactive compound levels. The results, summarized in Figure 3, highlight the effects of pasteurization on the retention of these compounds in C. grandis juices.

A comparative analysis between untreated and pasteurized samples was conducted to evaluate the extent of thermal degradation and its implications for the nutritional and functional quality of the final product.

The effect of pasteurization on phenolic and flavonoid contents was not uniform among cultivars, with significance levels ranging from non-significant to highly significant (p < 0.01), depending on the compound and genotype. Regarding total polyphenol content (TPC), Todarii and Terracciani showed the largest decreases, with reductions of approximately 36% (from 25.3 to 16.2 mg gallic acid equivalents (GAE)/100 mL of juice) and 39% (from 19.3 to 11.7 mg GAE/100 mL), respectively. Maxima also exhibited a significant decrease of 21% (from 15.2 to 12.0 mg GAE/100 mL). In Pyriformis and Chadock, values were numerically lower after pasteurization (from 15.2 to 12.1 mg GAE/100 mL and from 11.0 to 10.5 mg GAE/100 mL, respectively), but these differences were not statistically significant. Similarly, TFC showed cultivar-dependent responses to pasteurization. Terracciani and Pyriformis experienced the most pronounced and statistically significant decreases (~45%, from 14.1 to 7.8 mg quercetin equivalents/100 mL and ~37%, from 9.3 to 5.9 mg QE/100 mL, respectively). In Todarii, Maxima, and Chadock, values were numerically lower after pasteurization (from 11.0 to 10.0 mg QE/100 mL, from 12.2 to 11.1 mg QE/100 mL, and from 8.8 to 7.1 mg QE/100 mL, respectively), but these differences were not statistically significant. These findings suggest that flavonoids in Terracciani and Pyriformis are more sensitive to heat treatment, whereas the other cultivars retain these compounds more effectively, likely due to differences in thermal stability and phenolic composition. Flavonoids are influenced by a multitude of factors, both intrinsic and related to processing and storage.

Flavonoid degradation during pasteurization is known to involve oxidative reactions, cleavage of glycosidic bonds, and structural rearrangements promoted by temperature and acidic pH conditions typical of Citrus juices.

Manthey and Grohmann [30] demonstrated that the stability of Citrus flavonoids is closely related to their glycosylation pattern, reporting that rutinoside derivatives generally exhibit greater resistance to acid- and heat-induced hydrolysis compared to neohesperidoside forms. This structural aspect may explain the higher susceptibility observed in Terracciani and Pyriformis if characterized by a less thermally stable glycosidic profile.

Furthermore, Patras et al. [31] highlighted that thermal degradation of phenolic compounds in fruit matrices depends not only on intrinsic molecular structure but also on matrix-related factors such as oxygen availability, pH, and interactions with polysaccharides, which can either protect phenolics from oxidation or increase their exposure during heating. Similarly, Sánchez-Moreno et al. [32] reported significant reductions in flavonoid content and antioxidant activity in thermally processed Citrus juices compared to non-thermal treatments, attributing these losses to oxidation and structural degradation of phenolic compounds.

Collectively, these findings support the hypothesis that the differential retention observed among Citrus grandis cultivars is primarily genotype-dependent and mediated by both flavonoid structural features and matrix organization, rather than by temperature alone.

Compared to our samples, Brasili et al. [33] reported a slight increase in total phenolic content in pasteurized C. sinensis (L.) Osbeck juice from the Bahia cultivar (639.5 vs. 621.6 mg/L), likely due to enhanced solubilization of phenolic compounds following pasteurization. This comparison highlights how the impact of heat treatment can vary depending on cultivar and processing conditions. These cultivar-specific differences are consistent with trends reported in the literature. In the study by Shahid et al. [34], conventional pasteurization of Citrus juices resulted in a measurable reduction of bioactive compounds, particularly TPC and TFC. The pasteurized samples showed a 20–30% decrease in TPC compared with untreated juices, while TFC values also declined, although to a variable extent depending on the fruit species. These reductions were attributed to thermal degradation and oxidation of phenolic structures, as well as possible polymerization reactions during heat exposure.

In the TCC assay, Todarii was the only cultivar to exhibit a statistically significant decrease (p < 0.05), with carotenoid levels dropping from 13.2 to 10.1 mg β-carotene equivalents/100 mL, corresponding to a reduction of approximately 23%. The other cultivars showed non-significant changes, with Maxima even displaying a slight numerical difference (+8%, from 11.8 to 12.8 mg β-carotene equivalents/100 mL), while Pyriformis, Chadock, and Terracciani experienced decreases ranging from 8 to 25%. This variability highlights a cultivar-dependent response of carotenoids to pasteurization, with Todarii carotenoids being more sensitive to heat. The relative stability observed in most cultivars may be attributed to differences in carotenoid composition, matrix effects, or protective factors limiting degradation during thermal exposure [35]. Thermal pasteurization is widely applied to extend the shelf life of Citrus juices by inactivating microbes and enzymes. However, this process can also accelerate carotenoid oxidation and degradation, although findings in the literature are inconsistent [12]. Several studies report reduced total carotenoid content (TCC) after pasteurization [10,11,35] while others found no significant changes [9]. Reduced carotenoid levels may lead to lower bioactivity and affect the functional quality of the juice.

When considering statistical analyses across all cultivars for total polyphenols, flavonoids, and carotenoids, no significant overall differences were detected between untreated and pasteurized juices (Figure 3a–c). This indicates that, while individual cultivars exhibited specific responses, the overall effect of pasteurization on these compounds was not statistically significant at the population level, suggesting that the thermal treatment affected each cultivar differently but did not substantially alter the general phytochemical profile of C. grandis juices.

The HPLC method demonstrated robust analytical performance, characterized by low limits of detection (LOD) and quantification (LOQ) for all analytes (Table S3), as well as excellent linearity across the tested concentration ranges (R^2^ ≥ 0.9975). The chromatographic run time of 40 min provided an optimal balance between resolution and throughput, while the use of a reversed-phase C18 column ensured efficient retention and separation of the target flavonoids. These features confirmed the method’s suitability for accurate quantification of minor flavonoid constituents in Citrus matrices. Six marker flavonoids, eriocitrin, neoeriocitrin, narirutin, naringin, hesperidin, and neohesperidin, were quantified in Citrus grandis juices from five cultivars (Todarii, Maxima, Pyriformis, Chadock, and Terracciani). The analysis was carried out by high-performance liquid chromatography (HPLC, Table 3 and Table S3, Figure S1), allowing precise identification and quantification based on retention time, spectral characteristics, and external standard calibration. Flavonoid content varied markedly among C. grandis cultivars, reflecting genotype-dependent phytochemical diversity and highlighting the relevance of cultivar selection when assessing functional and nutritional properties. The flavanone distribution in the juices obtained from the five cultivars is reported in Table 3. When considering the total flavanone content, pasteurization resulted in an overall increase in four out of five cultivars, with Todarii, Pyriformis, Chadock, and Terracciani showing increases ranging from approximately 10 to 30%, while only the Maxima cultivar exhibited a net decrease. This indicates that pasteurization did not induce a generalized degradation of flavanones, but rather exerted cultivar- and compound-specific effects, likely reflecting a balance between thermal degradation and enhanced extractability from the juice matrix. The dataset further reveals cultivar-specific variations in flavanone composition and distinct responses to pasteurization, related to the structural diversity and thermosensitivity of flavanone glycosides and aglycones. Although the overall treatment effect was not statistically significant (p > 0.05), clear numerical trends were observed. In juice from the Todarii cultivar, pasteurization produced a consistent increase in all major flavanones, eriocitrin, neoeriocitrin, narirutin, naringin, and hesperidin, with concentrations increasing by approximately 25–50% compared with the untreated juice. This behaviour suggests that moderate heat exposure disrupted the cell wall matrix, enhancing extractability or solubilization of vacuole-bound glycosides and resulting in higher apparent concentrations in the aqueous phase. Similar effects were reported by Londoño-Londoño et al. [36], who observed improved recovery of Citrus flavanones after mild thermal treatment due to matrix softening and increased solvent accessibility.

In the Maxima cultivar, pasteurization led to a marked reduction in several flavanones. Neoeriocitrin and narirutin decreased to non-detectable levels, while naringin and hesperidin showed significant declines compared to the untreated juice. Conversely, neohesperidin concentration increased after pasteurization. The overall decrease in most flavanones suggests thermal degradation and/or acid-promoted hydrolysis of glycosidic bonds. The concomitant rise in neohesperidin may indicate structural rearrangements or isomerization processes under heat and acidic conditions, as previously described by Manthey & Grohmann [30] under acidic and heat-assisted conditions, although further targeted analyses would be required to confirm this transformation.

Juice from the Pyriformis cultivar showed a relatively simple flavanone profile in the untreated state, dominated by narirutin, naringin, and hesperidin, while pasteurization led to the appearance of eriocitrin and neoeriocitrin and to moderate increases in all detectable compounds. This behaviour supports the hypothesis that thermal treatment can promote the release of conjugated or insoluble flavanones from the cellular matrix, converting bound or polymerized phenolics into soluble, measurable forms. The Chadock cultivar yielded juice with the highest eriocitrin concentration among all samples, exceeding 34 mg/L in the untreated state and increasing slightly after pasteurization. The persistence and modest enhancement of naringin content in this cultivar may indicate a greater resistance of this flavonoid to heat-induced degradation under the applied processing conditions. Furthermore, the increase in neohesperidin indicates a potential rearrangement or interconversion of glycosidic forms, consistent with previously observed thermally induced isomerization of flavanones [21].

Finally, in the Terracciani cultivar, narirutin and hesperidin were the predominant flavanones in both treatments, and pasteurization caused a marked rise in hesperidin concentration (from 13.5 to 32.2 mg/L). This substantial increase may result from the release of cell wall-bound hesperidin, which becomes more soluble following pectin degradation and disruption of polysaccharide phenolic interactions during heating. A minor decrease in eriocitrin and the appearance of neohesperidin in the pasteurized juice further support the coexistence of degradation and conversion mechanisms. Although ANOVA did not indicate significant treatment effects, the observed compositional patterns highlight a dynamic balance between degradation, transformation, and release processes that govern flavanone stability under thermal conditions. The variability among cultivars can be attributed to differences in juice matrix composition, pH, and the initial ratio of glycosylated to aglycone forms, which collectively influence the behaviour of flavanones during heat processing [22]. Overall, these findings confirm that pasteurization exerts compound and cultivar-specific effects: depending on the structural characteristics of the flavanones and the surrounding matrix environment, heat may act either as a degrading or as an extracting factor.

3.3. Antioxidant Activity

The antioxidant activity of Citrus grandis juices was assessed through multiple complementary in vitro assays, including DPPH (2,2-diphenyl-1-picrylhydrazyl), ABTS (2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid)), FRAP (ferric reducing antioxidant power), and β-carotene bleaching methods. Antioxidant capacity values of pasteurized juice samples, compared with untreated controls, are presented in Figure 4a–c and Table 4 expressed as the average across all evaluated C. grandis cultivars.

For both DPPH and ABTS assays, pasteurization significantly reduced the radical scavenging activity in all cultivars (p < 0.001 for individual comparisons). In the DPPH assay, untreated juices exhibited strong radical-scavenging activity, ranging from 78 to 100% in Pyriformis and Chadock, respectively. On average, pasteurization decreased activity by approximately 22 percentage points across cultivars (Figure 4a). Comparable reductions have been reported in other fruit matrices. For instance, fresh Physalis peruviana L. juice exhibited a 39% decline in DPPH activity after pasteurization and storage [26]. Similarly, in grapefruit juices, pasteurization led to a 15% loss of total phenols and a 40% reduction in DPPH activity [22]. In contrast, Brasili et al. [33] observed an increase in DPPH activity in the soluble fraction of pasteurized C. sinensis juices (Cara Cara and Bahia cultivars), while the corresponding pellet fractions exhibited a marked decrease, suggesting differential heat responses between soluble and bound antioxidants.

Consistent with DPPH results, the ABTS assay also revealed a significant decline in antioxidant capacity after pasteurization. Untreated juices ranged from 78 to 96% in Pyriformis and Todarii, respectively, and decreased significantly to between 58 and 80% after heat exposure (Figure 4b).

The FRAP assay confirmed this trend, showing a significant reduction in reducing power (p < 0.001) across all cultivars. Untreated juices ranged from 23.8 µM Fe (II)/g (Pyriformis) to 93.9 µM Fe (II)/g (Terracciani), while pasteurized samples decreased by 25% to 45% (Figure 4c). Comparable FRAP values have been reported in other Citrus species, such as C. aurantium (no activity), C. clementina (2.8–5.7 µM Fe (II)/g), and higher levels in juice concentrates (122.8–364.2 µM Fe (II)/g) [15,37,38].

The β-carotene bleaching results in Table 4 showed a clear reduction in antioxidant activity following pasteurization. Untreated juices displayed inhibition values ranging from 18 to 60% for Pyriformis and Maxima, respectively, after 30 min incubation and from 22 to 78% after 60 min incubation. Heat treatment reduced inhibition in all cultivars, especially in Todarii (from 25 to 11%) and Pyriformis (from 18 to 5%). Maxima and Chadock showed moderate reductions of 25–30%, and Terracciani exhibited an intermediate decline of about 50% at both time points. These results indicate that pasteurization diminishes the ability of C. grandis juices to inhibit β-carotene oxidation, likely due to degradation of thermolabile antioxidants such as flavanone glycosides and ascorbic acid. The differing magnitudes of reduction among cultivars reflect matrix-dependent variability in phenolic and carotenoid stability, consistent with the cultivar-specific responses observed in other antioxidant assays.

Xu et al. [21] also reported strong variability in Citrus juices, with DPPH inhibition up to 61.6%, FRAP values of 307–899 mg/L AEAC, and total carotenoids (as β-carotene equivalents) between 0.06 and 10.02 mg/L, generally higher in mandarin juices.

Overall, these consistent and significant decreases across all antioxidant tests demonstrate that pasteurization markedly affects the antioxidant capacity of the juices. Statistical analyses confirmed significant overall differences for DPPH (p < 0.01; **), ABTS (p < 0.05; *), and no significant change for FRAP (ns) (Figure 4a–c), indicating that the extent of antioxidant loss depends on both the assay system and the mechanisms involved.

The RACI (Relative Antioxidant Capacity Index) values obtained for Citrus grandis juice demonstrated substantial variability among the tested Citrus grandis juices (samples 1–5) and between untreated (UT) and pasteurized (PAST) conditions (Figure S2). The results clearly indicate that the antioxidant capacity is predominantly associated with cultivars Todarii and Maxima. Untreated C. grandis cv. Maxima recorded the highest RACI value (≈+35), followed by cv. Todari (≈+17), confirming that these two untreated juices possessed the strongest intrinsic antioxidant potential within the dataset. Pasteurization led to a marked decrease in antioxidant capacity for both cultivars, reducing RACI values to approximately +10 (1-PAST) and +23 (2-PAST), respectively. Despite this decline, both samples retained positive RACI values after thermal processing, indicating that a portion of their antioxidant compounds remained active. The observed reduction can be attributed to the thermal degradation of phenolic compounds, vitamin C, and other heat-labile constituents, which are known to contribute significantly to the antioxidant properties of Citrus juices [39]. In contrast, cultivars Pyriformis, Chadock, and Terracciani displayed very low RACI values, with some even approaching or falling below zero. This suggests limited antioxidant potential in these matrices, possibly due to lower concentrations of bioactive compounds or differences in cultivar composition, processing stage, or juice extraction yield. The minimal differences between UT and PAST conditions for these samples further indicate that pasteurization had little measurable impact on their already low antioxidant capacity. The decrease in antioxidant capacity observed following pasteurization, particularly in C. grandis juice from cv. Todarii and Terraciani, can be attributed to the thermal degradation of key bioactive compounds such as phenolics and vitamin C. These constituents are known to play a dominant role in the antioxidant potential of Citrus juices and are highly sensitive to heat-induced oxidation and structural alteration [40]. Nevertheless, the persistence of positive RACI values in pasteurized samples suggests that a fraction of thermally stable antioxidants remains active after processing.

3.4. Enzyme Inhibitory Activity

Several studies have established a strong association between oxidative stress and the onset of chronic conditions such as obesity and type 2 diabetes mellitus. In this context, juices from different Citrus grandis cultivars, considered as an average across all genotypes, were evaluated for their potential to inhibit key enzymes involved in the pathophysiology of metabolic syndrome. Specifically, inhibitory effects on α-amylase and α-glucosidase, the main enzymes responsible for carbohydrate hydrolysis and key therapeutic targets for glycaemic control, were assessed. The percentage inhibition of these enzymes was determined by comparing pasteurized juice samples with their corresponding untreated controls. The juices exhibited inhibitory activity against all tested enzymes, although to varying degrees. Pasteurization significantly affected these activities, leading to a marked decrease in α-amylase and α-glucosidase inhibition across all cultivars (p < 0.001), except for Chadock in the α-amylase assay, which showed a slightly lower significance (p < 0.01). α-Amylase inhibition declined from 44–95% in fresh juices to 36–62% after pasteurization, with the largest decreases recorded for Maxima (95 to 59%) and Pyriformis (77% to 44%).

Pasteurization significantly reduced α-glucosidase inhibition (p < 0.001), from 35–77% in untreated samples to 20–51%, with the greatest reductions in Todarii and Chadock. Sapkota et al. [3] reported comparable inhibition in Citrus maxima juices, with α-amylase inhibition ranging from 75.5% to 79.5% and α-glucosidase from 70.8% to 72.8%.

Pancreatic lipase, the enzyme responsible for the digestion of triglycerides and absorption of free fatty acids, was also evaluated. Lipase inhibition decreased significantly after pasteurization, with values dropping from 13–45% in untreated juices to 3–30% following heat treatment. All cultivars exhibited highly significant reductions (p < 0.001), except Maxima, which showed a significant decrease at p < 0.05. These findings indicate that pasteurization markedly diminishes the inhibitory potential of these juices against digestive enzymes, likely due to the degradation or structural alteration of bioactive compounds responsible for enzyme inhibition [41]. Comparable to previous findings on Citrus maxima albedo, which demonstrated notable anti-diabetic potential through α-glucosidase (45%) and α-amylase (35%) inhibition at 25 mg/mL, with only modest lipase inhibition (20.9%), the extracts in the present study similarly exhibited differential enzyme inhibition profiles [42]. These results are also consistent with previous studies on C. clementina juices, which reported pronounced inhibitory activity against α-glucosidase [6,15]. However, in contrast to C. pomelo cultivars, which showed weak α-glucosidase and α-amylase inhibition but strong lipase inhibition [41,42], our results demonstrate a more balanced inhibitory profile across the tested enzymes, suggesting cultivar-specific differences in bioactive compound composition and enzyme interaction.

Overall, when considering the statistical analysis across all cultivars for α-amylase, α-glucosidase, and lipase inhibitory activities, significant overall differences were observed for α-amylase and α-glucosidase (p < 0.05; *), whereas no significant change was detected for lipase inhibition (ns), as illustrated in Figure 5a–c.

3.5. Principal Component Analysis

As shown in Figure 6, the Principal Component Analysis (PCA) provided an integrative interpretation of the compositional and functional differences among Citrus grandis cultivars and their response to pasteurization. The first two principal components, PC1 and PC2, explained ~44% and ~32% of the total variance, respectively, accounting together for 76% of the dataset variability.

PC1 clearly discriminated untreated (UT) from pasteurized (PAST) samples for all cultivars, confirming that thermal treatment represented the dominant source of variation. Samples positioned on the positive side of PC1 corresponded to untreated juices and were associated with higher total phenolic content (TPC), total flavonoid content (TFC), total carotenoid content (TCC), and α-amylase inhibitory activity. These variables contributed most strongly to the positive loading of PC1 and may therefore be regarded as biochemical markers of the fresh juice condition.

Conversely, pasteurized samples were consistently shifted toward negative PC1 values, reflecting the reduction of thermolabile phenolics and partial loss of enzyme inhibitory capacity induced by heat exposure. Importantly, the magnitude of cultivar displacement along PC1 closely paralleled the quantitative HPLC findings (Table 3), strengthening the mechanistic interpretation of the multivariate pattern. Maxima exhibited one of the largest negative shifts, consistent with its marked decrease in total flavanone content (264.7 to 205.7 mg/L) and the disappearance of specific compounds such as narirutin and neoeriocitrin after pasteurization. This extensive compositional alteration supports the high thermal sensitivity observed in its antioxidant and enzyme inhibitory responses. Todarii also showed a pronounced PCA displacement, despite an apparent increase in total flavanone concentration (307.0 to 397.2 mg/L). This behaviour suggests that heat-induced matrix disruption enhanced flavanone extractability without proportionally preserving functional activity, indicating qualitative rather than purely quantitative modifications. In contrast, Chadock and Pyriformis exhibited smaller displacements along PC1, reflecting greater compositional stability. Chadock maintained high eriocitrin levels and showed moderate increases in total flavanone content (251.7 to 306.7 mg/L) without drastic compound losses, while Pyriformis displayed controlled compositional adjustments (167.4 to 201.5 mg/L) with preservation of dominant flavanones. This relative stability in the chemical profile was mirrored by more contained variations in antioxidant and enzyme inhibitory parameters.

PC2 primarily represented an axis contrasting antioxidant performance and acid-related variables. Positive PC2 loadings were associated with pH, FRAP, β-carotene bleaching (30 and 60 min), DPPH, ABTS, and eriocitrin, whereas ascorbic acid exhibited a strong negative loading. This distribution suggests that antioxidant capacity in these matrices is more strongly influenced by phenolic composition and specific flavanones than by vitamin C content alone. Samples richer in structurally stable flavanones tended to cluster toward the upper region of the plot, whereas those characterized by higher ascorbic acid levels occupied lower PC2 positions.

The differential clustering of cultivars may be explained by structural and matrix-related factors. Variations in flavanone glycosylation patterns likely influence thermal behaviour, since rutinoside derivatives generally exhibit greater resistance to acid-catalysed hydrolysis and heat-induced degradation compared with neohesperidosides [30,32]. Furthermore, differences in pectin architecture and polysaccharide–phenolic interactions may modulate oxygen diffusion and the microenvironment surrounding bioactive compounds during heating [31,43]. In pigmented cultivars such as Chadock, the coexistence of carotenoids and phenolics may also generate synergistic antioxidant networks that attenuate oxidative degradation through radical-scavenging interactions [21,44], thereby buffering multivariate displacement after pasteurization.

Overall, PCA confirms that pasteurization induces coordinated modifications in phenolic composition, antioxidant activity, and enzyme inhibition capacity, while preserving cultivar-specific chemical fingerprints. The close correspondence between multivariate clustering and HPLC-derived compositional trends supports the conclusion that thermal stability in Citrus grandis juices is primarily governed by flavanone profile, glycosylation pattern, and matrix organization rather than by a uniform degradation mechanism.

3.6. Correlation Matrix

The correlation heatmap highlights the relationships among physicochemical and antioxidant parameters in untreated (UN) and pasteurized (PAST) juice samples of Citrus grandis. A clear clustering pattern was observed, showing strong positive correlations among the main antioxidant indices, particularly between total phenolic content (TPC), total flavonoid content (TFC), and the antioxidant assays (DPPH, ABTS, and FRAP) (Figure 7). This indicates that the overall antioxidant capacity of the juices is mainly driven by their phenolic and flavonoid composition. In both untreated and pasteurized samples, these variables tend to move in the same direction, confirming that phenolic compounds largely contribute to the radical scavenging potential of C. grandis juice.

Positive correlations were also observed between TCC (total carotenoid content) and the β-carotene bleaching assays at 30 and 60 min, suggesting that carotenoids contribute significantly to the protection against lipid peroxidation in these systems. However, the magnitude of correlation coefficients between carotenoids and phenolic-based antioxidant assays (DPPH, ABTS, FRAP) was generally lower, implying that these classes of compounds act through distinct antioxidant mechanisms.

The negative correlations between amylase activity and the major antioxidant markers (TPC, TFC, DPPH, ABTS, FRAP) in both treatments may indicate that higher enzymatic activity corresponds to lower antioxidant retention, possibly due to degradation of bioactive compounds during thermal processing or subsequent storage. This trend is more evident in pasteurized samples, reflecting the effect of thermal treatment on enzymatic denaturation and phenolic stability.

When comparing UN and PAST samples, the correlations remain qualitatively similar, although a general reduction in the strength of associations is observed after pasteurization. This attenuation may be attributed to partial degradation of thermolabile compounds such as flavonoids and carotenoids, leading to a more heterogeneous distribution of antioxidant properties. Nonetheless, the retention of significant positive correlations between TPC, TFC, and FRAP in PAST samples suggests that, despite heat exposure, a substantial fraction of phenolic compounds remains active.

Overall, the correlation pattern underscores the central role of phenolic and flavonoid compounds in determining the antioxidant potential of Citrus grandis juices, and it also reflects the modulatory effect of pasteurization on the biochemical interactions among antioxidant constituents. These findings are consistent with previous reports on Citrus juices, which demonstrate that mild thermal processing can alter quantitative levels without completely disrupting the intrinsic relationships among bioactive parameters.

4. Conclusions

This study provides novel evidence that the impact of thermal pasteurization on Citrus grandis juices, in terms of biochemical and functional attributes, is strongly cultivar-dependent and cannot be generalized across pomelo varieties.

Specifically, we show that thermal treatment does not simply cause a uniform loss of bioactive compounds in terms of total polyphenols, flavonoids, and carotenoids, but rather triggers a dynamic balance between matrix disruption and thermolabile compound degradation. In some cultivars (e.g., Todarii), pasteurization promoted the release of matrix-bound flavonoids, partially compensating for thermal losses, whereas in others (notably Maxima and Terracciani) hydrolysis and isomerization phenomena predominated, leading to more pronounced reductions. This differential behavior has not been previously clarified for pomelo juices.

HPLC analysis enabled the precise identification and quantification of key marker flavonoids, eriocitrin, neoeriocitrin, narirutin, naringin, hesperidin, and neohesperidin, revealing distinct qualitative and quantitative differences across cultivars. Chadock displayed the highest eriocitrin levels and remarkable stability after pasteurization, while Maxima suffered significant losses of several flavanones, indicating greater thermal sensitivity. Pasteurization induced contrasting effects, promoting the release of matrix-bound compounds in some cultivars (e.g., Todarii) while triggering hydrolysis or isomerization of flavonoids in others (notably Maxima and Terracciani). These findings suggest that the degradation of bioactive compounds during pasteurization does not follow a simple unidirectional trend but rather reflects a dynamic balance between release, transformation, and loss processes.

Certain phenolic and carotenoid constituents demonstrated relative thermal stability, contributing to the preservation of antioxidant potential under moderate heating. Overall, C. grandis juice retained measurable antioxidant activity after pasteurization, though the extent of thermal degradation was strongly influenced by initial phenolic composition, vitamin C content, and matrix structure. Hence, optimizing processing parameters is essential to balance microbial safety with the retention of nutritional and functional quality.

Enzyme inhibitory activities toward α-amylase, α-glucosidase, and lipase were also affected, with the most pronounced reductions observed in Maxima and Pyriformis, whereas Chadock maintained higher inhibitory capacity. These variations likely stem from structural modifications and partial degradation of phenolic, flavonoid, and carotenoid compounds responsible for these effects.

Multivariate analysis further confirmed that pasteurization-induced changes are primarily driven by coordinated shifts in phenolic and carotenoid composition rather than by simple physicochemical alterations. The clear separation between untreated and pasteurized samples in PCA underscores the sensitivity of thermolabile bioactive compounds, while also revealing differential resilience among cultivars.

Among cultivars, Maxima and Todarii showed the largest compositional shifts, indicating higher thermal sensitivity, whereas Chadock and Pyriformis retained greater stability.

Although sensory evaluation was beyond the scope of this study, the slightly bitter and acidic profile described for some cultivars suggests that technological approaches such as blending with sweeter fruit matrices or incorporation into functional beverage formulations could represent viable strategies for enhancing consumer acceptance. From an application perspective, these juices may be strategically exploited through blending, formulation as functional ingredients, or integration into industrial beverage matrices, where their bioactive richness could be valorized while mitigating potential sensory limitations.

Furthermore, considering that the juices were tested undiluted and under physiologically relevant concentrations, the observed enzyme inhibitory activities provide nutritionally meaningful preliminary evidence of potential functional effects, although further studies incorporating simulated gastrointestinal digestion models and in vivo validation are necessary to confirm their effective bioavailability and physiological impact.

Overall, moderate thermal treatment modulates but does not compromise the nutritional and functional quality of Citrus grandis juices. The integration of advanced analytical tools, together with the selection of heat-tolerant cultivars (particularly Chadock and Pyriformis), offers a practical framework for designing pasteurization protocols that preserve varietal identity and functional quality while ensuring microbial safety. These findings support a shift from a generic processing strategy toward cultivar-tailored thermal optimization in pomelo juice production.