Punicalagin‐enriched edible coatings from revalorized pomegranate peels for quality enhancement of fresh‐cut celery

Abstract

BACKGROUND

Pomegranate peel is a rich source of phenolics, particularly punicalagin, but its valorization for fresh produce preservation requires stabilization. This study aimed to develop and evaluate freeze‐dried encapsulates of pomegranate peel extracts as ingredients in edible coatings for fresh‐cut celery. Pomegranate peels were extracted with ethanol and encapsulated using maltodextrin, inulin, or pectin at wall‐to‐core ratios of 1:1 and 0.5:1. Extracts and encapsulates were characterized for color, total phenolic content (TPC), antioxidant capacity, and punicalagin (Pn). Edible coatings containing the encapsulates were applied to minimally processed celery sticks by dipping, and quality was monitored over 13 days at 4 °C for physicochemical and nutritional quality.

RESULTS

Ethanolic extraction concentrated phenolics (TPC 246.8 mg kg⁻¹; Pn 85.3 g kg⁻¹), and encapsulation with inulin at a 0.5:1 ratio yielded the highest phytochemical retention (TPC 120.8 mg kg⁻¹; Pn 43.4 g kg⁻¹). Coated celery showed reduced discoloration (color change (ΔE) approximately 8 vs approximately 13 in an uncoated control (CTRL)) and weight loss (~0.5% vs ~2% in CTRL). Enriched coatings preserved TPC (~0.12 mg GAE kg⁻¹ vs ~0.08 mg GAE kg⁻¹ in CTRL) and increased celery Pn content, which peaked at ~1.8 g kg⁻¹ on day 9.

CONCLUSION

Encapsulated pomegranate peel extracts provided bioactive fortification and helped to maintain the physicochemical quality of fresh‐cut celery during refrigerated storage, offering a sustainable valorization strategy for agro‐industrial by‐products. © 2026 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

INTRODUCTION

Celery (Apium graveolens L.) is commonly consumed in fresh‐cut form but it is also very perishable and sensitive to changes after processing.^{1, 2} Cutting and handling can lead to dehydration, browning, and microbial growth, which reduce shelf life and quality. Refrigeration and modified‐atmosphere packaging can delay these problems but they are not always sufficient to maintain freshness.^{1, 2} Recent studies have evaluated natural compounds such as propolis, tea tree oil, and gallic acid for the preservation of fresh‐cut celery; however, these approaches often rely on purified ingredients rather than ingredients obtained from food by‐products,^{3, 4, 5, 6} an approach that remains largely unexplored.

In this scenario, it has been estimated that between 30% and 50% of fresh produce from the fruit and vegetable industry becomes by‐products or is discarded. ⁷ These residues, often treated as waste, contain a wide range of valuable compounds, including polyphenols, carotenoids, and other plant‐based antioxidants, which have potential applications in food, cosmetics, and health‐related products. ⁸

Among fruit by‐products, pomegranate (Punica granatum L.) peel has attracted attention because of its high phenolic compound content, especially punicalagin, which has shown strong antioxidant properties. ⁹ These compounds are sensitive to heat, oxygen, and light, which can reduce their effectiveness. Encapsulation techniques are therefore being developed to protect them and improve their stability and bioactivity in food applications. ¹⁰

Edible coatings, a novel technology, are thin layers made from natural materials such as polysaccharides, proteins, or lipids that are applied directly to the surface of food to provide protection.^{11, 12} In fresh‐cut products they help to reduce moisture loss, control microbial growth, and slow oxidation, while maintaining appearance and texture. The incorporation of plant extracts can further improve their protective effects. Several studies have reported promising results using coatings containing essential oils or flavonoids to preserve fruits and vegetables, ¹³ including apples, ¹⁴ strawberries, ¹⁵ tomatoes, ¹⁶ and carrots. ¹⁷ However, the effectiveness of these coatings depends on the stability of the incorporated compounds, which can be difficult to maintain during storage. Nanoencapsulation has emerged as a useful approach to address this limitation by enabling the incorporation of bioactive compounds into edible coatings in a more stable and controlled manner. ¹⁸ Although pomegranate peel extracts have been studied in functional foods and as natural preservatives, their use in edible coatings, particularly for fresh‐cut vegetables, remains limited.

In this context, the hypothesis underlying this research was that edible coatings enriched with nanoencapsulated punicalagin from pomegranate peel can provide a promising and sustainable approach for maintaining the physicochemical quality of fresh‐cut celery during refrigerated storage. This strategy combines the reuse of food waste with protective coating technologies, while addressing current demands for clean‐label products and more environmentally friendly food preservation solutions. Accordingly, the aim of this work was to develop and characterize a new coating formulation and to evaluate its effect on weight loss, color, firmness, soluble solids, pH, phenolic content, antioxidant capacity, and polyphenol oxidase (PPO) activity in fresh‐cut celery.

MATERIALS AND METHODS

Procedure to obtain pomegranate by‐products, extraction, and encapsulation

‘Wonderful’ and ‘Mollar de Elche’ pomegranates were supplied by the Cambayas Cooperative Society (Elche, Spain). The inedible tissue was separated from the fresh fruit, freeze‐dried using a Telstar LyoBeta lyophilizer (Syntegon Telstar Technologies S.L.; Terrassa, Spain), and ground with an IKA A11 basic mill (IKA‐Werke GmbH & Co. KG; Berlin, Germany) to obtain a homogeneous material (Ø < 56 μm). The resulting by‐product was subjected to solid–liquid (1:19) extraction using 96% ethanol at room temperature, assisted by ultrasound for 1 h in a Sonorex Digiplus DL 514 BH bath (BANDELIN electronic GmbH & Co. KG; Berlin, Germany) at 720 W and 35 kHz as previously described. ¹⁹ All extracts obtained in the previous study ¹⁹ were mixed and diluted with water to 25%, and the ethanol was removed using a rotary evaporator at 60 °C for 20 min until constant mass was achieved, concentrating the bioactive compounds dissolved in water.

For encapsulation, maltodextrin (M) from cornstarch (provided by Power Natural Life SL, Pontevedra, Spain), inulin (I) (from Ingredissimo, Barcelona, Spain), and pectin (P) (from citrus peel by Merck KGaA, Darmstadt, Germany) were selected as carrier agents and they were mixed with the pomegranate extracts at dissolved solids‐to‐carrier ratios of 0.5:1 and 1:1 (carrier:extract; w:w). These food‐grade, clean label compatible carriers were selected based on their specific characteristics: M provides a soluble, low‐viscosity, neutral‐tasting matrix that improves powder yield and handling; P has emulsifying and stabilizing properties that can enhance encapsulation efficiency and oxidative stability; and I additionally provides dietary fiber with prebiotic properties.

The prepared mixtures were homogenized using an T25 digital Ultra‐Turrax (IKA‐Werke GmbH & Co. KG; Barcelona, Spain) for 10 s at 676 × g, then frozen and freeze‐dried to ensure effective encapsulation. The capsules were characterized by measuring the total phenolic content (TPC), the total antioxidant capacity (TAC) using ferric reducing antioxidant power (FRAP), 2,2‐diphenyl‐1‐picrylhydrazyl (DPPH), and ABTS 2,2′‐azino‐bis(3‐ethylbenzothiazoline‐6‐sulfonic acid) (ABTS) assays, as well as the punicalagin (Pn) content. These methods are described in detail below. Encapsulation yield was expressed as a percentage of the initially added Pn retained within the encapsulated system:

$$\left(\mathrm{Pn}\mathrm{\text{content in capsules}}/\mathrm{\text{initial}}\mathrm{Pn}\mathrm{\text{content}}\right)\times \mathrm{100}.$$

Images of the encapsulated pomegranate by‐products were captured using a field emission scanning electron microscope (FE‐SEM), model Crossbeam 350 (Carl Zeiss AG; Oberkochen, Germany), equipped with a Gemini electron column (Carl Zeiss AG; Oberkochen, Germany). Imaging was performed under high‐vacuum conditions with an accelerating voltage of 1 kV, a working distance of approximately 5 mm, and secondary electron (SE) detection. The image acquisition process was carried out with SmartSEM software version 6.07, Service Pack 10 (16 June 2020).

Minimal processing of celery and coatings formulations

Approximately 10 kg of celery (Apium graveolens ‘Apio MP’) were provided by G's España (Love Fresh Company S.L.; Torre Pacheco, Murcia, España), cultivated at the Serreta farm in Torre Pacheco, Murcia, Spain. The celery was cut into 2 cm sticks, washed in chlorinated cold water (100 ppm, pH 6.5, 5 °C, 2 min), rinsed with water at the same temperature for 1 min, and drained for 5 min on filter‐paper‐lined trays.

Coating formulations were prepared by dispersing 2 g of the encapsulated pomegranate peel extract in 50 mL of distilled water with 250 μL of Tween 80, forming a nanoemulsion. This ratio was selected based on preliminary assays evaluating formulation stability, dispersibility, and coating homogeneity. Celery sticks were immersed for 3 min in a 10% dilution of this emulsion, followed by 5 min draining and gentle drying with absorbent paper.

Five treatments were applied: an uncoated control (CTRL), industrial control (I‐CTRL: 1% citric acid solution), and coatings with punicalagin encapsulated in 1:1 maltodextrin (Pn‐M), inulin (Pn‐I), or pectin (Pn‐P). Fresh‐cut celery samples (approximately 50 g) were packaged in 250 mL polypropylene containers under passive atmospheric conditions (65% humidity), and sealed with manually perforated film (with 15 uniformly distributed holes made using a 0.8 mm diameter needle) to allow gas exchange while maintaining high relative humidity. Samples were stored at 4 °C for up to 13 days. Analyses were performed on days 0, 3, 6, 9, and 13 with five replicates per treatment and sampling day.

The coating formulations were characterized in triplicate (n = 3) by measuring temperature using a digital thermometer (Acuinuga; Bertamiráns, A Coruña, Spain), pH using a digital pH‐meter (GLP‐21, Crison; Alella, Catalonia, Spain), electrical conductivity (μS cm⁻¹) using a benchtop conductivity meter (Basic 30, Crison; Lardero, La Rioja, Spain), total soluble solids content (TSS; g kg⁻¹) using a portable digital refractometer (Atago N1; Tokyo, Japan), color using the CIELab system with a Konica Minolta CR‐400 colorimeter (Tokyo, Japan), and density (g mL⁻¹) using a glass pycnometer based on weight difference.

Physico‐chemical analysis

Physicochemical analyses were conducted on days 0 (initial quality), 3, 6, 9, and 13 at 5 °C to monitor product quality over time. At day 0, initial quality parameters were evaluated with ten replicates, including weight loss, firmness, dimensions, color, TSS, and pH. Weight loss was calculated proportionally for each sampling day according to this formula:

$$\left(\left(\mathrm{\text{initial weight}}–\mathrm{\text{final weight}}\right)/\mathrm{\text{initial weight}}\right)\times \mathrm{100}.$$

Firmness was measured at room temperature using a texture analyzer (Brookfield CT3‐4500; Toronto, ON, Canada) equipped with a 1 mm cylindrical probe. Celery pieces were placed horizontally on the platform and subjected to a compression test at a test speed of 0.8 mm s⁻¹ and a compression distance of 1 mm. Measurements were performed on ten pieces per treatment and sampling time and expressed in newtons (N) as both maximum and final force values. Longitudinal and equatorial diameters were determined with a digital caliper (Mitutoyo, Neuss, Germany). Color was assessed using a Konica Minolta colorimeter, as described above, and overall color change (ΔE) was calculated using the equation ΔE = $\sqrt{{\mathrm{\Delta L}}^{*2}+{\mathrm{\Delta a}}^{*2}+{\mathrm{\Delta b}}^{*2}}$ as described by Torres‐Sánchez et al. ²⁰ Juice for TSS and pH measurement was obtained using a juice extractor and these parameters were measured using the equipment described. On subsequent sampling days, weight loss and color were measured (n = 5), and photographs were taken to document changes over the storage period.

Bioactive compounds and enzymes

Total phenolic content, Pn, and enzymatic PPO activity were analyzed at days 0, 1, 3, 6, 9, and 13 to evaluate the evolution of bioactive compounds during storage. All spectrophotometric measurements were conducted using an ultraviolet‐visible (UV‐visible) microplate reader (Tecan Infinite M200; Männedorf, Switzerland).

For this purpose, celery samples were frozen in liquid nitrogen and ground in dark conditions at approximately 10 °C. One gram of plant tissue was extracted with 3 mL of 80% methanol. The mixture was homogenized using an Ultra‐Turrax for 10 s and stirred in dark conditions at 4 °C for 1 h using an orbital shaker (Stuart; Stone, UK). After centrifugation (3220 × g, 10 min, 4 °C), supernatants were filtered through 0.2 μm polytetrafluoroethylene (PTFE) filters and stored at –80 °C until analysis.

The TPC was determined using the Folin–Ciocâlteu method. ²¹ A 96‐well polystyrene flat‐bottom microplate (Greiner Bio‐One; Frickenhausen, Germany) was used. In each well, 19 μL of extract was mixed with 29 μL of 1 mol L⁻¹ Folin–Ciocâlteu reagent. After incubation for 3 min in dark conditions, 192 μL of a sodium carbonate (0.4%) and sodium hydroxide (2%) solution were added. The mixture was incubated for 1 h and absorbance was measured at 750 nm. Gallic acid was used for calibration, and results were expressed as mg gallic acid equivalents (GAE) per kg.

Total antioxidant capacity was assessed using three colorimetric assays: DPPH, ²² ABTS, ²³ and FRAP. ²⁴ For the DPPH assay, 194 μL of a 0.7 mmol L⁻¹ DPPH solution was added to 21 μL of extract in a 96‐well plate. After 30 min incubation, absorbance was read at 515 nm. For the ABTS assay, 200 μL of a pre‐activated 32 μmol L⁻¹ ABTS solution was mixed with 11 μL of extract, incubated for 20 min, and absorbance was recorded at 414 nm. For the FRAP assay, the FRAP reagent was prepared daily by mixing sodium acetate buffer (pH 3.6), 10 mmol L⁻¹ 2,4,6‐tripyridyl‐s‐triazine (TPTZ) in 40 mmol L⁻¹ HCl, and 20 mmol L⁻¹ FeCl₃ in a 10:1:1 ratio (v:v:v). Then, 198 μL of FRAP solution was added to 6 μL of extract, incubated for 30 min, and absorbance was read at 593 nm. In all cases, results were expressed as mg Trolox equivalents (TE) per kg.

For separation, identification, and quantification of total Pn (α‐ and β‐isomers), the method described by Cano‐Lamadrid et al. ¹⁹ was followed using an ultra‐high‐performance liquid chromatography (U‐HPLC) system (Shimadzu, Kyoto, Japan) equipped with a DGU‐20A degasser, LC‐30 CE quaternary pump, SIL‐30 AC autosampler, CTO‐10AS column oven, and SPDM‐20A photodiode array detector. The stationary phase was a C18 reversed‐phase column (Gemini C18 110 Å, 4.6 × 250 mm; 5 μm). The mobile phases consisted of water acidified to pH 2.0 with H₃PO₄ (A) and acetonitrile (B). The gradient started at 5% B for 5 min, increased to 30% B at 30 min, and returned to 5% B until 42 min. The flow rate was 0.8 mL min⁻¹, injection volume 10 μL, column temperature 35 °C, and detection wavelength 260 nm. Identification was based on retention time and quantification was performed using calibration curves of Pn standard. Results were expressed as the sum of α‐ and β‐punicalagin (∑Pn), in g kg⁻¹.

For PPO activity analysis, 1 g of celery sample was added to 30 mL of extraction buffer (0.1 mol L⁻¹ sodium phosphate buffer, 4% polyvinylpolypyrrolidone (PVPP), adjusted to pH 6.5 with 37% HCl). Samples were homogenized using an Ultra‐Turrax for 15 s, stirred in dark conditions for 4 h at 4 °C, and then centrifuged at 3220 × g for 10 min at 4 °C. To determine PPO activity, 50 μL of each supernatant was added to 250 μL of a reaction buffer containing 0.07 mol L⁻¹ catechol and 0.5 mol L⁻¹ sodium phosphate buffer in a 96‐well plate. The reaction kinetics was monitored at 420 nm every 15 s for 10 min in the microplate reader. Polyphenol oxidase activity was expressed as the residual rate of PPO activity with respect to the reference sample over time in minutes (%) using the next equation with the final and initial absorbances (Abs) at 420 nm:

$$\left(\left(\mathrm{\text{Final}}\mathrm{Abs}–\mathrm{\text{Initial}}\mathrm{Abs}\right)/\min \right)\times \mathrm{100}.$$

Statistical analysis

A two‐factor experimental design was employed to evaluate the effects of storage time (t) and treatment type (T) on the investigated variables in celery sticks and Pn encapsulates. Data were analyzed using two‐way analysis of variance (ANOVA) with a significance level of P < 0.05, performed using GraphPad Prism software (v. 8.0.2; GraphPad Software, Boston, MA, USA). The normality of residuals and homogeneity of variances were assessed using the Shapiro–Wilk and Levene's tests, respectively. Weight loss, encapsulation yield, and relative PPO activity were expressed as percentages and analyzed without transformation, as these assumptions were met. Tukey's multiple range test was applied post hoc to determine statistically significant differences among group means.

RESULTS

Characterization of pomegranate capsules and encapsulation yield

Table 1 shows the characterization of the pomegranate extracts used to optimize the freeze‐dried encapsulation. Table 2 displays the main components after adding the evaluated carriers to develop the different encapsulated extracts.

Table 1.

Initial characterization of raw material and obtained ethanolic extracts

	Color			TPC (mg kg−1)	TAC (mg kg−1)			Pn (g kg−1)	Visual appearance
	L*	a*	b*		FRAP	DPPH	ABTS
Pomegranate peels	46.5 ± 4.1	33.8 ± 4.9	24.2 ± 3.1	76.8 ± 5.8	125.4 ± 6.8	2823.9 ± 25.7	4013.2 ± 84.1	17.6 ± 1.4
Ethanol extraction a	50.4 ± 4.5	13.7 ± 1.0	15.2 ± 1.2	246.8 ± 12.6	476.6 ± 55.1	11 295.5 ± 554.8	16 855.3 ± 927.1	85.3 ± 1.0

TPC: Total phenolic content; TAC: Total antioxiant capacity measured by FRAP, DPPH, and ABTS; Pn: Punicalagin content.

^a Data are the mean values obtained in previous works. ²²

Table 2.

Color (CIELab), total phenolic content, total antioxidant capacity, and punicalagin content of freeze‐dried encapsulates of pomegranate peels

		Freeze‐dried control	M 1:1	M 0.5:1	I 1:1	I 0.5:1	P 1:1	P 0.5:1
Color	L*	28.5 ± 1.9 g	65.1 ± 5.0e	64.1 ± 4.1f	78.1 ± 4.9b	78.3 ± 5.8a	67.9 ± 3.2d	68.3 ± 2.2c
	a*	6.5 ± 0.7d	13.0 ± 1.1a	12.8 ± 0.2b	6.3 ± 0.3e	7.3 ± 0.1c	1.3 ± 0.1 g	3.0 ± 0.1f
	b*	5.6 ± 0.3 g	19.0 ± 0.8a	18.2 ± 0.9b	14.1 ± 0.6e	15.7 ± 0.7c	13.3 ± 0.8f	14.4 ± 0.5d
TPC (mg GAE kg−1)		26.2 ± 4.3d	52.1 ± 4.8bc	109.9 ± 25.4a	61.1 ± 13.5b	120.8 ± 8.0a	36.3 ± 7.5 cd	64.6 ± 2.9b
TAC (mg TE kg−1)	FRAP	48.0 ± 7.2d	70.4 ± 2.3bcd	98.0 ± 10.2ab	81.5 ± 2.3bc	119.8 ± 2.6a	42.0 ± 1.9d	66.2 ± 6.6 cd
	DPPH	1174.0 ± 40.8e	1705.3 ± 62.3c	1823.1 ± 80.3bc	1835.0 ± 74.6b	3729.0 ± 35.0a	1464.1 ± 112.5d	1834.7 ± 37.4b
	ABTS	1571.9 ± 121.3d	2046.3 ± 100.8bc	2041.4 ± 105.0c	2191.8 ± 44.4b	4174.3 ± 97.5a	2005.5 ± 36.2c	2025.7 ± 52.6c
Pn (g kg−1)		9.2 ± 0.2f	17.7 ± 1.4d	30.0 ± 2.3b	24.3 ± 1.7c	43.4 ± 0.9a	12.1 ± 0.2e	23.3 ± 0.9c

Different letters denote statistical differences (P < 0.05) among pomegranate encapsulates.

As shown, the raw peels contained moderate amounts of bioactive compounds, with TPC reaching 76.8 mg kg⁻¹ and Pn 17.6 g kg⁻¹. Antioxidant activity values determined by FRAP, DPPH, and ABTS assays were consistent with this composition, but remained relatively low when compared with the ethanolic extracts. These lower values observed in pomegranate peels compared to their ethanolic extracts were also observed for TPC and Pn content. After ethanolic extraction, TPC increased twofold, TAC values increased threefold, and Pn content reached 85.3 g kg⁻¹ (3.8‐fold higher than unextracted peels). The observed colorimetric changes (increases in L* and decreases in a* and b* coordinates) also suggest partial removal of dark pigments, leading to a clearer extract, which is advantageous for subsequent formulation as an ingredient or coating.

The encapsulation process affected both the physical and functional attributes of the powders strongly (Table 2). These patterns confirm that the wall material governed optical behavior, with I promoting brightness (L*) and M retaining red–yellow tones (a*–b*). This trend is also evident in Fig. 1, which shows representative images of each encapsulated extract.

Figure 1.

Visual appearance and images of pomegranate encapsulates obtained by field emission scanning electron microscope (FE‐SEM). Freeze‐dried control is shown at 20×, 50×, and 100×.

Total phenolic content was lowest in the freeze‐dried control and rose for all carriers, particularly when the wall‐to‐core ratio was reduced to 0.5:1. The highest TPC values were shown by I 0.5:1 and M 0.5:1, followed by I 1:1, P 0.5:1, M 1:1, and P 1:1. This pattern indicates that decreasing the wall fraction from 1:1 to 0.5:1 increased the apparent phenolic load for each carrier through higher loading and reduced dilution. Consistent with TPC, TAC increased relative to the control across the three assays, with I delivering the strongest responses overall. Overall, the results point to a clear carrier effect: I showed the best results followed by M and P for FRAP, DPPH, and ABTS analyses, and the 0.5:1 ratio consistently enhanced antioxidant readouts for each material. Following a similar trend, the control contained the lowest Pn content (9.2 g kg⁻¹). I 0.5:1 achieved the highest value, followed by M 0.5:1, and I 1:1, P 0.5:1, and M 1:1 showed mean values, and P 1:1 was significantly lower.

Consistent with the color results described above, Fig. 1 presents optical (0×) and scanning electron microscope (SEM) images (100×, 500×, 1000×) of the freeze‐dried powders, highlighting clear carrier‐dependent microstructural differences. The freeze‐dried control shows large, smooth agglomerates with limited fragmentation, as expected for an aqueous extract lyophilized without a wall material; for this reason these SEM images are shown in this case at 20×, 50×, and 100×. In contrast, all encapsulated samples (M, I, and P) exhibit markedly smaller, more fragmented particles organized in heterogeneous matrices with fracture planes and evident microporosity. Despite this heterogeneity, clear differences were observed among carriers. Within each carrier, the 1:1 and 0.5:1 ratios display closely related architectures, indicating that polymer type is the principal determinant of morphology, whereas the wall:core ratio mainly modulates the degree of porosity and agglomeration. Qualitatively, M showed granular and plate‐like particles with compact glassy areas, I showed angular flakes with internal porosity, and P formed cohesive structures with folded layers and fibrous agglomerates. Across magnifications up to 1000×, fine sub‐particles appear embedded or coated within the polysaccharide network, supporting the conclusion that the antioxidant‐rich extract – including Pn – is dispersed and physically entrapped by the carbohydrate chains. This particle‐size reduction and matrix encapsulation relative to the control are consistent with the higher retention and protection of bioactive compounds observed for the encapsulated systems, particularly those based on I.

These results indicate that encapsulation efficiency (Fig. 2) also varied considerably depending on the carrier type and wall‐to‐core ratio. The overall encapsulation efficiency was approximately 50%, which is acceptable for freeze‐drying techniques.

Figure 2.

Encapsulation yield (%) of Pn in the different freeze‐dried encapsulates of pomegranate peels. Different letters represent significant differences (P < 0.05) among encapsulates.

Initial characterization of minimally processed celery sticks and edible coatings

Once the encapsulates were developed, they were incorporated as edible coatings in a fresh‐cut celery product. Table 3 shows that the fresh celery sticks had the expected green–yellow coloration with more lightness at the external surface in comparison with the cutting zone, more negative a* externally, and greater b*. Dimensions were within the typical commercial range. Texture analysis yielded a crisp profile. Juice pH was mildly acidic and soluble solids were low, consistent with the low‐sugar nature of celery.

Table 3.

Initial characterization of fresh‐cut celery sticks

External color	L*	53.5 ± 1.5	Diameter (mm)	Equatorial	16.6 ± 3.1
	a*	−13.6 ± 1.3		Longitudinal	25.7 ± 4.5
	b*	27.0 ± 2.2	Firmness (N)	Max	7.89 ± 0.86
Color of cutting zone	L*	46.7 ± 6.1		Final	5.10 ± 0.51
	a*	−9.9 ± 1.1	Juice	pH	5.7 ± 0.0
	b*	18.6 ± 1.6		TSS (g kg−1)	2.9 ± 0.06

TSS, total soluble solids.

Table 4 indicates marked differences among coating solutions. The I‐CTRL coating was essentially colorless as it was water with citric acid, whereas Pn‐M and Pn‐P were more chromatic toward yellow with moderate lightness, and Pn‐I was the darkest with a lower b* value, indicating a slight yellow tone. Acidity and ionic strength also differed: I‐CTRL was strongly acidic and as conductive as tap water due to the high concentration of citric acid. Pn‐M and Pn‐I were closer to celery juice pH with moderate conductivity, and Pn‐P showed intermediate acidity and the highest conductivity. All coatings had very low TSS and similar densities, indicating that they deposited a similar amount of dry solids on the celery surface when used with the same application volume.

Table 4.

Color (CIELab), visual appearance, pH, total soluble solids, conductivity, and density of edible coatings used

		I‐CTRL	Pn‐M	Pn‐I	Pn‐P
Color	L*	55.5 ± 0.27 a	44.0 ± 0.09 c	39.0 ± 0.16 d	51.02 ± 0.12 b
	a*	−0.52 ± 0.01 b	−0.3 ± 0.02 a	−0.7 ± 0.02 c	−3.6 ± 0.01 d
	b*	−0.31 ± 0.06 d	10.2 ± 0.05 a	6.2 ± 0.05 c	10.0 ± 0.03 b
Visual appearance
pH		2.3 ± 0.0 c	5.4 ± 0.0 a	5.3 ± 0.06 a	3.9 ± 0.06 b
TSS (g kg −1 )		0.93 ± 0.06 a	0.6 ± 0.06 b	0.6 ± 0.06 b	0.6 ± 0.0 b
Conductivity (μS cm −1 )		217 ± 0.03 c	564 ± 7.2 b	566 ± 25.7 b	735 ± 13.6 a
Density (g mL −1 )		1.08 ± 0.53 a	1.09 ± 0.2 a	1.08 ± 0.4 a	1.07 ± 0.36 a

Different letters denote significant differences among coatings (P < 0.05).

Physicochemical quality of fresh‐cut celery sticks

Figure 3 shows the instrumental color change (ΔE), visual appearance, and weight loss data used to track quality over 13 days at 4 °C following the application of edible coatings to the fresh‐cut celery sticks. Over time, ΔE increased in all treatments, but coatings consistently moderated the change relative to CTRL. At day 3, the CTRL already showed a larger variation (approximately 8–9) compared with the coated samples (approximately 5–7). By day 13 the deviation was most evident, with the CTRL showing the highest discoloration (approximately 12–14), whereas coated treatments remained lower (approximately 6–10). As shown in the photographs, the CTRL exhibited pronounced browning at the cut surfaces together with an overall loss of brightness after 13 days, while I‐CTRL offers only partial mitigation, Pn‐M and Pn‐P maintained an intermediate appearance with some yellowing and localized darkening, and Pn‐I preserved the green color most effectively, showing brighter, more uniform pieces and cleaner cut faces.

Figure 3.

Color evolution (ΔE), visual appearance, and weight loss (%) of the celery sticks under different edible coatings. Different capital letters denote significant differences (P < 0.05) among treatments while different lower case letters denote significant differences (P < 0.05) among sampling times during the shelf‐life of the study.

The separation is clearer when focusing on weight loss, as coated samples displayed significantly lower losses than the CTRL throughout storage. At day 3, CTRL had the largest loss (approximately 0.9% to 1.1%), whereas the coated samples clustered near 0.2% to 0.4%. By day 9, the CTRL approached 1.5% to 1.8% whereas the coated samples remained below approximately 0.6% to 0.8% and at day 13 the CTRL was highest (approximately 1.9% to 2.2%) with all coatings significantly lower. In particular, phenolic‐rich samples tended to have the smallest losses, with Pn‐I among the lowest (approximately 0.3% to 0.4%).

Bioactive compounds presence against browning

The results for bioactive compounds in coated and uncoated celery sticks (Fig. 4) highlight three important aspects of the system: the inherently low phenolic content of celery, the effect of edible coatings on phenolic maintenance, and the role of Pn in enriching and stabilizing the product during storage.

Figure 4.

Total phenolic content (mg GAE kg ⁻¹ ), polyphenol oxidase (PPO) activity (%), and punicalagin (Pn) content (g kg ⁻¹ ) of studied celery sticks. Different capital letters denote significant differences (P < 0.05) among treatments while different lower case letters denote significant differences (P < 0.05) among sampling times during shelf‐life study.

The TPC results confirm that celery has intrinsically low phenolic levels: at day 0, all lots started near 0.09–0.11 mg GAE kg⁻¹. Over storage, ΔTPC was limited but coatings helped to maintain or slightly increase values in comparison with CTRL. By day 9, CTRL trended lower, whereas Pn‐coated treatments remained higher. This pattern persisted at day 13, with coated samples generally near 0.10–0.15 mg GAE kg⁻¹ and CTRL closer to 0.10 mg GAE kg⁻¹. The better retention in the enriched coated samples, most notably Pn‐I samples, was consistent with the TPC provided by the pomegranate peel encapsulates used to prepare the coatings (high TPC and Pn, especially for the I‐based powders as shown in Table 2), and with the barrier effect of the film that limited oxidative and enzymatic depletion.

The PPO activity was also higher in the Pn‐coated samples than in CTRL and I‐CTRL celery sticks, particularly around days 3 and 13, when Pn‐I and Pn‐P reached approximately 1.0% to 1.6% whereas CTRL and I‐CTRL were approximately 0.4% to 0.8%.

After an insignificant initial level, Pn was measurable from the first day (approximately 0.05–0.7 g kg⁻¹) in Pn‐coated samples. The Pn content increased through day 6 (approximately 0.8–1.2 g kg⁻¹) and peaked around day 9, when Pn‐I showed the highest values (approximately 1.6 g kg⁻¹), followed by Pn‐M and Pn‐P (approximately 1.8 g kg⁻¹). At day 13, Pn remained substantial in all Pn‐coated samples (approximately 1.2–1.7 g kg⁻¹), confirming retention over refrigerated storage. This hierarchy mirrored the encapsulate composition (Table 2), where I‐based powders carried the greatest Pn concentration and strongest antioxidant values.

DISCUSSION

Encapsulation process

The solvent extraction step (Table 1) produced the expected enrichment of hydrolysable tannins and associated antioxidant capacity: relative to raw peels, the ethanolic extract showed increases in TPC, TAC, and Pn. This pattern is consistent with prior studies showing that hydroalcoholic systems efficiently recover pomegranate peel polyphenols, particularly Pn and other ellagitannins, and deliver higher in vitro antioxidant readouts than water alone, depending on cultivar and solvent affinity.^{19, 25} The concurrent shift in color coordinates (higher L*, lower a*, and b*) is consistent with partial removal of dark/brown pigments during extraction and with the predominance of colorless tannins in the ethanolic fraction. ⁹ Together, these observations are in line with reviews ²⁶ and experimental studies^{19, 27} that have reported that ethanolic solvents are suitable media for concentrating peel phenolics into functional extracts while preserving redox activity.

Encapsulation by freeze‐drying (Table 2) further shaped the chemical and functional profile of the powders in a carrier‐dependent manner. First, depending on the carrier used, I and M formed amorphous, glassy matrices that could limit oxygen and moisture diffusion and reduce phenolic oxidation.^{28, 29} In several systems, such as lycopene encapsulates from tomato by‐products, ³⁰ oleuropein encapsulates with higher in vitro bioaccessibility, ³¹ or powders obtained from Saskatoon berries, ²⁹ I outperformed M for bioactive compound retention during drying and storage. On the other hand, lowering the wall:core ratio increased the apparent dilution and can improve the proportion of TPC and Pn per gram of powder, even though very high wall fractions sometimes raise encapsulation efficiency. The data reported – higher TPC/Pn and stronger FRAP/DPPH/ABTS capacities at 0.5:1 – agreed with a direct relationship between higher core yields and higher compositional signals, as has been widely described.^{32, 33} The direct relationship between high Pn and stronger TAC, especially in I 0.5:1, suggests that Pn contributes substantially to the redox capacity of these powders.

As Fig. 1 illustrates, the freeze‐dried powders presented irregular particle sizes and morphologies. This is typical of lyophilization processes. In this sense, and in relation to carrier encapsulation yield, I was the most effective protector of Pn, followed by M and P, confirming its superior ability to form a protective matrix against oxidation and degradation. A consistent trend was observed whereby formulations with a lower proportion of carrier yielded higher relative antioxidant content, as the extract accounted for a larger fraction of the encapsulate. However, this effect must be balanced against the lower stability generally associated with reduced wall material.

Regarding Pn encapsulation yield (Fig. 2), the values observed fell within the broad range reported for polyphenol encapsulation by freeze‐drying, which varies with matrix composition and solids loading. ³³ Examples in the literature show that, depending on the carrier and formulation, freeze‐dried systems can achieve moderate to high encapsulation efficiency (approximately 40% to 50% for certain anthocyanins) when the matrix and process are optimized.^{33, 34} In this sense, pomegranate peel extracts encapsulated with maltodextrin or combined carriers likewise showed variable but substantial retention of phenolics and their functional activity. ³⁵ Hence, the superiority of inulin‐based formulations in this study, particularly at 0.5:1, was consistent with reports that inulin can provide efficient protection for sensitive polyphenols during dehydration and storage, translating into higher measured payloads and stronger antioxidant signals.^{29, 31}

Application in fresh‐cut celery

The initial characterization of fresh‐cut celery sticks (Table 3) confirmed that the raw material displayed the physicochemical and sensory traits expected for this vegetable. Similar differences in color between epidermal and inner tissues have been reported in fresh‐cut celery and other minimally processed vegetables, and are typically attributed to chlorophyll gradients, tissue hydration, and structural differences between supporting collenchyma and parenchyma.^{36, 37} In fact, differences between the external color and color cutting zone indicated that the freshly cut parenchymatous tissue was less bright, less green, and more yellow than the external fibrous surface. Firmness values indicated a crisp texture comparable with previous reports,^{38, 39} confirming that the selected raw material met commercial expectations. Likewise, juice pH and TSS results were consistent with the naturally low sugar content of celery, ¹ which explains its mild sensory profile and low caloric value.

The characterization of the coating solutions in which the celery sticks were dipped (Table 4) revealed significant differences in physicochemical properties depending on the pomegranate encapsulates, as described above. The color values obtained for Pn‐coatings reflected the natural pigmentation of pomegranate phenolics, especially anthocyanins and ellagitannins, which confer a yellow‐orange or reddish tone. ⁴⁰ Pn‐I was the formulation with the lowest luminosity value, which may relate to the strong affinity of inulin matrices for polyphenols and the formation of more pigmented complexes, as has been described for inulin‐based encapsulation systems. ⁴¹

Although I‐CTRL was strongly acidic, Pn‐M and Pn‐I showed values close to those of celery juice, which may facilitate sensory compatibility and reduce the risk of acid‐induced tissue damage during storage. ⁴² By contrast, Pn‐P exhibited an intermediate pH, reflecting the acidic nature of pectin solutions.^{43, 44} Conductivity was highest for Pn‐P, likely due to the ionizable groups in pectin molecules, whereas Pn‐M and Pn‐I were moderately conductive. These values fell within the expected range for biopolymer‐based coatings containing phenolic‐rich extracts. ⁴⁵

Given the similar TSS and density values obtained, it can be hypothesized that the coatings had a similar thickness, explaining why none of them produced noticeable changes in the initial physicochemical or visual attributes of the celery sticks, despite their marked differences in pH, electrical conductivity, and chromaticity. This observation is consistent with previous studies showing that edible coatings with a low solids content (<1%) are typically transparent, imperceptible, and do not affect consumer acceptance when applied to fresh‐cut vegetables.^{11, 12}

The physicochemical parameters of fresh‐cut celery sticks coated with Pn‐rich edible coatings (Fig. 3) demonstrated that these films effectively slowed both discoloration and water loss during refrigerated storage. Color variation (ΔE) values confirmed that visual quality degradation progressed in all treatments, but the extent was significantly reduced in the coated samples compared with the untreated celery sticks (CTRL). By the end of the 13 day storage, CTRL sticks exhibited severe browning at cut surfaces and an overall loss of brightness, which agrees with previous studies showing that celery sticks are particularly prone to enzymatic browning and tissue dehydration during storage.^{1, 2} The I‐CTRL coating provided partial reduction in browning due to its citric acid content, but it lacked phenolic enrichment. By contrast, Pn samples preserved the sticks’ appearance moderately, with some yellowing and localized darkening. This effect can be attributed to the Pn content and other ellagitannins present in the encapsulates.^{40, 46} I has also been shown to have a protective role by forming more stable matrices that limit oxygen penetration and stabilize phenolics during storage compared with other carriers, as described above.^{29, 41} These results are consistent with reports that edible coatings enriched with antioxidants or antimicrobials can delay the development of enzymatic browning and surface discoloration in fresh‐cut vegetables by forming a semipermeable barrier and by delivering active phenolic compounds that interfere with oxidative pathways responsible for this visible deterioration in quality.^{12, 36}

Weight‐loss data provided even clearer evidence of coating efficacy. Such reductions in water loss in comparison with uncoated treatment (CTRL) are in line with the established role of hydrocolloid‐based edible films as semipermeable barriers that reduce moisture transfer.^{11, 12} Importantly, the phenolic‐enriched systems tended to perform slightly better than I‐CTRL, which can be explained not only by the physical barrier effect but also by possible synergistic interactions between phenolic compounds and the biopolymer networks that enhance film cohesiveness and reduce permeability. ¹² Again, Pn‐I showed the lowest final weight loss, which reinforces the superior protective effect of I compared with other carriers.^{29, 31}

Regarding the bioactive compound content, the data confirm that celery has intrinsically modest levels of phenolics, consistent with previous reports that classify celery among the vegetables with relatively low polyphenolic content compared with leafy greens or cruciferous vegetables. ⁴⁷ Nevertheless, Pn‐coated samples were able to maintain and improve their TPC. This can be explained by the additional phenolic content from the encapsulates, as previously characterized in Table 2 (notably the I‐based powders), and by the barrier function of the coatings, which limits oxygen diffusion and thus slows enzymatic oxidation. Similar stabilization of phenolic levels through phenolic‐enriched edible films has been described in lettuce and spinach,^{12, 36} reinforcing the effectiveness of this strategy.

The results related to PPO activity were expected because the coated samples contain more phenolic substrates (Fig. 4 and Table 2), which can both sustain PPO‐mediated reactions and upregulate activity in wounded tissues. ⁴⁸ However, these results should be interpreted carefully, as the high concentration of phenolic compounds incorporated through the coating may have acted as additional substrates or may have interfered with the in vitro spectrophotometric assay, potentially leading to an overestimation of the measured PPO activity. Importantly, the higher PPO in Pn‐coated celery did not translate into worse visual quality or increased browning (Fig. 3). This may be explained by the antioxidant and metal‐chelating capacity of Pn and other ellagitannins, which could limit the progression of oxidative browning reactions, together with the barrier effect of the coating in reducing oxygen and moisture transfer. Comparable findings have been reported in apple, melon, ¹⁴ and pear slices ⁴⁹ coated with polyphenol‐rich films, where PPO activity remained measurable but visual browning was attenuated.

As also shown in Fig. 4, Pn was absent in the CTRL and I‐CTRL coatings but was detectable from the first day in Pn‐coated celery, peaking around day 9, and remained substantial at day 13. This reflects both efficient deposition and a degree of stability during refrigerated storage. Again, this behavior can be related to the encapsulate composition (Table 2), with I‐based encapsulates carrying the greatest Pn load and translating this into the highest measured levels in celery. As Pn is one of the most potent antioxidants in pomegranate peel, ⁵⁰ its incorporation into celery sticks not only improved physicochemical quality appearance but also enhanced nutritional and functional value. These findings are consistent with previous research showing that by‐products of pomegranate, particularly the peel, are rich in hydrolysable tannins with strong antioxidant capacity, ⁵¹ and that their inclusion in edible matrices can deliver measurable enrichment in fresh‐cut produce.

CONCLUSIONS

This study demonstrated that pomegranate peel by‐products, when extracted, encapsulated, and incorporated into edible coatings, can be successfully applied to fresh‐cut celery to enhance its functional and storage quality. The encapsulation process, particularly when using I at a 0.5:1 wall‐to‐core ratio, resulted in powders with the highest Pn content and strongest antioxidant activity, which translated into improved results once they were applied as coatings for fresh‐cut celery. During refrigerated storage, Pn‐coatings effectively reduced weight loss and attenuated discoloration compared with uncoated celery, with the I‐based formulation providing the most consistent preservation of visual and antioxidant quality. Moreover, Pn‐enriched coatings fortified celery sticks with Pn and contributed to a better maintenance of TPC, retarding browning processes despite increased PPO activity.

These findings highlight the potential of revalorizing pomegranate by‐products as natural functional ingredients within edible coatings, offering a dual benefit of reducing food waste and improving the nutritional and physicochemical properties of minimally processed products. Improvements in application methods are still needed to maximize bioactive retention but the major limitation of this study is that microbiological quality and sensory acceptability were not evaluated. Further studies should assess aerobic mesophilic bacteria, yeasts, and molds, to assess the shelf‐life of coated fresh‐cut products, and consumer acceptance using a sensory analysis to propose the coating for a practical application on fresh‐cut celery. Overall, these results provide preliminary evidence of the potential of polyphenol‐rich encapsulates in sustainable fresh‐cut vegetable preservation strategies.

FUNDING INFORMATION

This work was supported by grant PID2021‐123857OB‐I00 funded by the Spanish Ministry of Science, Innovation and Universities (Ministerio de Ciencia, Innovación y Universidades, MICIU), the Spanish State Research Agency (Agencia Estatal de Investigación, AEI) 10.13039/501100011033 (AGROAL‐NEXT project), and the European Regional Development Fund (FEDER/ERDF/EU).

CONFLICT OF INTEREST

The authors declare no conflict of interest.

AUTHOR CONTRIBUTIONS

Lorena Martínez‐Zamora: conceptualization, methodology, software, validation, formal analysis, investigation, data curation, writing – original draft preparation, and visualization. Marina Cano‐Lamadrid: methodology, formal analysis, and investigation. Rosa Zapata: conceptualization, methodology, software, formal analysis, investigation, and data curation. Francisco Artés‐Hernández: conceptualization, resources, writing – review and editing, supervision, project administration, and funding acquisition. All authors have read and agreed to the published version of the manuscript.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.