Formulation and Stability Assessment of Bakery Snacks Enriched with Encapsulated Phenolic Compounds from Lemnian Tomatoes and Pumpkin (Cucurbita moschata)

Abstract

In recent years, the health-promoting properties of plant-derived compounds have garnered increasing scientific interest. Notably, tomatoes and pumpkins (Cucurbita moschata), renowned for their abundant phytochemicals and associated biological activities, have become focal points of research. This study investigated the extraction of phenolic compounds from tomatoes and pumpkins cultivated on Lemnos, an Aegean Island, aiming to enhance the nutritional profiles of food products. We established an extraction protocol for total phenolics and evaluated the antioxidant capacity using response surface methodology. Utilizing a central composite design, we optimized the extraction parameters, with time and ethanol concentration identified as critical factors (p < 0.05). The optimized extracts demonstrated substantial phenolic content (9.47 ± 0.08 and 4.52 ± 0.05 mg GAE/g for tomatoes and pumpkins, respectively) and antioxidant capabilities as determined by DPPH (7.65 ± 0.08 and 5.78 ± 0.05 μmol TE/g, respectively), ABTS (9.27 ± 0.02 and 3.95 ± 0.04 μmol TE/g, respectively), FRAP (5.25 ± 0.09 and 2.99 ± 0.03 μmol TE/g, respectively), and CUPRAC assays (2.3 ± 0.04 and 1.25 ± 0.03 μmol TE/g, respectively). Following extraction, the phenolic compounds were encapsulated using maltodextrin and subsequently freeze-dried, yielding high encapsulation efficiency. In alignment with a comprehensive strategy aimed at fostering functional snacks that enhance local economic and public health outcomes, vegetables sourced from local farms were employed to develop a savory cereal bar enriched with tomato extract and a sweet cookie infused with pumpkin extract.

1. Introduction

In recent years, the integration of natural compounds with antioxidant properties to augment the nutritional value of foods has gained considerable academic and industrial interest. Phenolic compounds, prevalent in a diverse array of fruits and vegetables, have attracted substantial attention for their potential health advantages and their contribution to oxidative stability. Specifically, Lemnian tomatoes and pumpkins (Cucurbita moschata) have emerged as notable sources of these compounds. These sources not only provide significant antioxidative benefits but also enhance the sensory qualities of food products with their distinct flavor profiles [1,2].

Given the public’s escalating concerns regarding food quality, there has been a significant surge in research exploring the properties and functionalities of diverse foods. Over recent decades, extensive studies have sought to substantiate the medicinal benefits associated with pumpkins, examining their potential to mitigate conditions such as diabetes, hypertension, tumors, and elevated cholesterol levels, as well as their role in preventing chronic diseases [3,4,5]. Additionally, studies have investigated the antibacterial, anthelmintic, and anti-inflammatory properties of pumpkins, attributes linked to its rich composition of vitamins, fibers, bioactive phytochemicals (including carotenoids and phenolic compounds), and minerals. Correspondingly, the health benefits of tomatoes have been thoroughly examined. These benefits include the anticancer potential of lycopene, its beneficial effects on diabetes, atherosclerosis, and cardiovascular diseases, and its role in modulating cellular pathways related to cancer progression, which is facilitated by the presence of fiber, vitamin C, and phenolic compounds [6,7].

The extraction of total phenolic compounds from plant sources has garnered significant attention due to their potential applications within the food and pharmaceutical sectors. Nevertheless, the stability of these compounds during storage, especially when integrated into food matrices, continues to pose a significant challenge. In response, encapsulation has emerged as a viable strategy to enhance the stability and bioavailability of phenolic compounds. This technique not only prolongs their shelf life but also amplifies their functional benefits in food products [8,9].

The integration of biologically active phytochemicals, particularly polyphenols, into industries such as food and pharmaceuticals underscore the importance of adopting environmentally sustainable extraction methods from plant materials that also preserve the integrity of these compounds [10]. In the realm of environmentally sustainable extraction techniques, several advanced methodologies have emerged, such as ultrasound-assisted extraction (UAE), supercritical fluid extraction (SFE), and microwave-assisted extraction (MAE). Among these, UAE has attracted substantial interest for its efficacious extraction of bioactive compounds. This technique offers numerous benefits, such as improved yield of extracts, decreased extraction time, and reduced operational costs in comparison to traditional extraction methods. To optimize the UAE process, researchers have extensively employed the response surface methodology (RSM). This statistical technique is utilized to explore the interrelationships and interactions among various experimental parameters, thereby enhancing the efficiency and effectiveness of the extraction process [11,12].

Microencapsulation has emerged as a pivotal technique in food science, aimed at enhancing the distribution of bioactive constituents within food matrices. This approach involves the use of carriers such as maltodextrin (MD), which protect bioactive compounds from degradative digestive processes while improving their bioactivity and bioavailability for controlled release. Moreover, these carriers enable targeted delivery to consumers. Microcapsules, a common choice for this purpose, provide robust protection against various environmental stressors such as oxygen, light, humidity, and other relevant factors. As a protective barrier, microencapsulation not only delays the immediate release of bioactive substances but also increases their solubility and bioavailability, thereby facilitating easier handling and transportation of these compounds [12,13].

A range of methodologies has been employed to effectively microencapsulate the bioactive compounds derived from vegetables and herbs. These techniques include dry mixing, complex coacervation, inclusion complexation, spray-drying, and freeze-drying. Among these, freeze-drying is particularly preferred for encapsulating fragile bioactive substances. This technique is advantageous because it avoids exposure to high temperatures, thereby preserving the structural integrity of phenolic compounds, which is essential for maintaining their biochemical activity [14]. The selection of encapsulation agents, coating materials, or carrier substances is critically important in microencapsulation processes, as these components substantially influence the effectiveness of encapsulation and the resulting physicochemical characteristics that govern product stability after freeze-drying. Consequently, the advantageous properties of maltodextrins, including their gel-forming capabilities, low viscosity, and high solubility, make them particularly suitable for the microencapsulation of biologically active molecules, such as polyphenolic compounds, through freeze-drying techniques [15].

In accordance with a holistic approach designed to promote functional snack foods that improve local economic and public health outcomes, this research aims to examine the extraction of total phenolic compounds and assess the antioxidant activities of Lemnian tomatoes and pumpkins (Cucurbita moschata) cultivated on Lemnos Island. Additionally, it investigates the feasibility of encapsulating these phenolic compounds to enhance their stability, as well as explores their potential for enrichment in bakery products by developing a savory cereal bar enriched with tomato extract and a sweet cookie infused with pumpkin extract. A critical aspect of this study is the understanding of storage stability of the encapsulated phenolic compounds to ensure that their antioxidant properties are preserved when incorporated into bakery products, thereby maintaining their functional efficacy.

2. Materials and Methods

2.1. Chemicals and Reagents

Tomato and pumpkin (Cucurbita moschata) specimens, cultivated on Lemnos—an Aegean Island in Greece—were procured from a local market on the island in August and September of 2023, respectively.

The chemical reagents employed in this study were purchased from various international suppliers. Anhydrous sodium carbonate, Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid), and Folin–Ciocalteau reagent were obtained from SDS (Peypin, France). ABTS (2,2′-Azino-bis[3-ethylbenzothiazoline-6-sulphonic acid]) was sourced from Applichem (Darmstadt, Germany). Neocuproine was supplied by Acros Organics (Fair Lawn, NJ, USA). Copper chloride dihydrate, ammonium acetate, sodium dihydrogen phosphate dihydrate, and sodium chloride were procured from Penta (CZ Ltd., Chrudim, Czech Republic). Additionally, potassium persulfate was acquired from Chem-Lab (Zedelgem, Belgium).

2.2. Sample Preparation

To prepare the samples of tomato and pumpkin, the pre-dried plant materials, previously possessing moisture contents of 93.1% and 86.0%, respectively, were subjected to grinding. This was carried out using a laboratory grinder, model IKA A 10 basic (manufactured by IKA Works, Wilmington, NC, USA), for a duration of one minute. This procedure yielded a homogeneous and thickened ground sample.

2.3. Extraction of Samples Assisted by Ultrasounds

The phenolic compounds were extracted from the tomato and pumpkin samples using a VCX-750 ultrasonic processor, equipped with a closed converter (Sonics & Materials, Inc., Newtown, Australia). The extraction process utilized a cup-horn attachment, and the ultrasonic pulsation settings were adjusted to alternate between one second on and one second off at 60% amplitude and 20 kHz frequency. The temperature during preliminary extractions ranged from 25 °C to 80 °C, and the temperature of 55 °C, where total phenolics were in higher yield, was chosen for optimization [16,17]. For the sample preparation, 0.5 g of dried samples were accurately weighed and transferred into 20 mL test tubes. Each tube was then supplemented with 10.0 mL of a solvent, which was either distilled water, ethanol, or a varying mixture of both. Following the extraction, the samples were subjected to centrifugation at 3000× g for 10 min at 25 °C to facilitate phase separation.

2.4. Evaluation of Antioxidant Activity in Extracted Samples

In this study, the antioxidant capacity of the optimized extracts was rigorously assessed using a comprehensive suite of assays, namely ABTS, DPPH, CUPRAC, and FRAP. Each sample underwent triplicate analysis to ensure the reliability of the results. For quantification purposes, various concentrations of Trolox were utilized as a standard, and the results were expressed as Trolox equivalents (TE) per gram of tomato and pumpkin across all testing methods. This standardized approach facilitated a precise comparison of antioxidant activity among the samples.

To evaluate the scavenging potential against the DPPH free radicals, we utilized a method adapted from Miller et al. [18]. Extract volumes ranging from 2.5 to 15.0 μL, or an appropriate Trolox standard, were diluted with methanol to achieve a final volume of 0.9 mL. Subsequently, 0.1 mL of a 0.6 mM solution of DPPH reagent in methanol was introduced to the mixture, which was then vigorously agitated. Following a 15 min incubation period in the absence of light, the absorbance was recorded at a wavelength of 515 nm, using a reference sample for comparison.

To evaluate the scavenging potential against the ABTS radical, we employed a method adapted from Brand-Williams et al. [19]. The ABTS^•+ cation radical was generated by the oxidation of ABTS using potassium persulfate. Sample extracts ranging in volume from 2.5 to 15.0 µL, as well as Trolox standards, were diluted to a final volume of 1.0 mL with the ABTS^•+ solution. The samples were then incubated for 15 min in the dark at room temperature. Following incubation, the absorbance of each sample was quantified at a wavelength of 734 nm.

The Ferric Reducing Antioxidant Power (FRAP) assay, as described by Benzie and Strain [20], was utilized to assess the reducing potential of the samples. This assay involves the reduction of ferric tripyridyltriazine (Fe³⁺-TPTZ) in an acidic environment. Sample volumes not exceeding 15.0 µL were dispensed into test tubes. These were then diluted with distilled water and the FRAP reagent in a volumetric ratio of 3:1 to achieve a total volume of 1.2 mL per tube. After incubating for 10 min at 37 °C, the absorbance was measured at a wavelength of 593 nm.

Furthermore, the reducing capacity of the samples was evaluated using the Cupric Ion Reducing Antioxidant Capacity (CUPRAC) assay, adhering to the methodology described by Özyürek et al. [21]. Incremental volumes of extracts up to 15.0 µL were combined with a reaction buffer comprising CuCl₂•2H₂O, neocuproine, and CH₃COONH₄ in distilled water. The resultant mixtures were then incubated at ambient temperature for 30 min, followed by the measurement of absorbance at 450 nm.

2.5. Determination of Total Phenolics

The phenolic compounds in the tomato and pumpkin samples were quantified after extraction using a modified method outlined by Singleton and Rossi, tailored to the optimized extraction conditions. This quantification was performed in triplicate. Adjustments specific to the samples were incorporated into the Folin–Ciocalteu method to ensure an accurate measurement of phenolic content [22]. The experimental protocol entailed the addition of various volumes of tomato and pumpkin extracts, not exceeding 15.0 µL, to a mixture composed of distilled water and Folin–Ciocalteu reagent in a volumetric ratio of 18:1. The resultant mixture was promptly homogenized and subsequently subjected to a controlled environment with diminished light exposure for a duration of 2.0 min. Following this initial phase, a 20% (w/v) sodium carbonate (Na₂CO₃) aqueous solution was introduced to adjust the total volume to 2.2 mL. The samples were then vigorously shaken and incubated at 40 °C for 30 min within a water bath. Absorbance was measured at 765 nm using a Lambda 25 Spectrophotometer (Perkin Elmer, Norwalk, CT, USA). A standard curve was generated using gallic acid concentrations ranging from 1.14 to 18.18 μg/mL, and the results were quantified and expressed in terms of gallic acid equivalents (GAE), in milligrams per gram of tomato/pumpkin tissue.

2.6. Statistical Analysis

The data are presented as mean ± standard deviation (SD). The response surface design outcomes were analyzed using the trial version of the Minitab^® software (https://www.minitab.com/en-us/products/minitab/free-trial/ (accessed on 26 June 2024), Minitab Ltd., Coventry, UK). Model verification was conducted employing SPSS Version 28.0.10 (IBM Corp., Armonk, NY, USA), through the application of a one-sample t-test. Independent t-tests were applied to compare the antioxidant activities and total phenolics between the optimized tomato and pumpkin extracts. For all the other comparisons, one-way ANOVA was used to compare the means of more than two independent groups. Post hoc comparisons were conducted using Tukey’s test. A p-value of less than 0.05 was set as the threshold for statistical significance.

2.7. The Design of Experimentation

A central composite design (CCD), a prominent design within the framework of the response surface methodology (RSM), was utilized to facilitate the estimation of both linear and quadratic effects, as well as the interactions of influential factors, enhanced by the inclusion of axial points. This design was specifically selected for this study to explore the optimal extraction conditions for quantifying the total phenolic content. The investigation focused on two critical variables, namely the concentration of ethanol (X₁) and the extraction time (X₂).

The coding for each variable was assigned at three levels as follows: −1, 0, and +1. The experiment was performed once. The choice of variables and their corresponding levels was determined through preliminary experiments, a review of the literature, and the constraints of the analytical instrumentation available. Both the coded and actual values of these variables are presented in

Table 1. BBD along with independent factors and their respective levels.

		Factor Levels and Range
Factors	Codes	−1	0	1
Time (min)	X2	20	40	60
Ethanol (%, v/v)	X1	30	50	70

.

The central composite design was employed to perform a regression analysis and second-order polynomial multiple regression as in Equation (1).

Y = β₀ + β₁X₁ + β₂X₂ + β₁₁X₁₂ + β₂₂X₂₂ + β₁₂X₁X₂ + ϵ

(1)

Equation (1) delineates the correlation between the response variable, symbolized as Y, and the quantifications of total phenolic content (TPC), measured in mg GAE/g. These values stem from evaluations of the total phenolic levels. The independent variables influencing the dependent variable, Y, are X₁ and X₂, as illustrated in Table 1. The model incorporates regression coefficients β₀, β₁, β₂, β₁₁, β₂₂, and β₁₂, which correspond to the intercept, linear effects, quadratic effects, and interaction terms, respectively. The model also includes ε, which represents the random error component.

The analysis of variance (quadratic regression analysis ANOVA) was employed to assess the adequacy of the quadratic approximation within the central composite design for the response surface. F-statistics and the corresponding p-values were used to determine the statistical significance of the different term groups, including linear terms, two-factor interactions, and quadratic terms, along with the regression coefficients obtained from the model. The terms were considered statistically significant if they displayed p-values below 0.05, indicating a 95% confidence level.

2.8. Verification of the Statistical Model

Predictive equations formulated using response surface methodology (Central Composite Design, CCD) were utilized to ascertain the optimal extraction conditions for quantifying total phenolic content in the tomato and pumpkin samples. This involved optimizing the extraction time and solvent composition. The phenolic compound concentrations were measured following extraction under these ideal conditions. The efficacy of the predictive model was evaluated by comparing its forecasts to the experimental outcomes thus providing a comprehensive assessment.

2.9. Encapsulation of the Optimized Extracts

The extraction of tomato and pumpkin samples was conducted on a scaled-up basis using optimized conditions (50 g of sample in 1000 mL of diluent), ensuring adequate extract volume for subsequent encapsulation. The organic fraction was concentrated using a rotary evaporator (LabTech, Inc., Hapkinton, MA, USA). For encapsulation, maltodextrin with a dextrose equivalent of 19 was selected as the encapsulating agent. It was dispersed in distilled water to attain a 9.0% solid concentration (w/v) by means of magnetic stirring.

The coating material solutions were formulated and subsequently merged with a concentrated extract of phenolic compounds, serving as the core, at a concentration of 9.0% w/v. This combination utilized specified core-to-coating volume ratios of 1:10, 1:15, 1:20, 1:25, and 1:30. Each mixture was subjected to homogenization using a Unidrive X1000D Homogenizer (CAT Scientific, Paso Robles, CA, USA), which was equipped with a 17 mm diameter shaft (Ingenieurbuero CAT, M. Zipper GmbH, Oberösterreich, Austria), operating at 14,000× g for a duration of 5 min. The operational parameters were established based on the initial experimental data and corroborated by the existing literature.

The encapsulated extracts were stored in an ultra-low temperature freezer (model DW-HL388, Zhongke Meiling Cryogenics Corp., Hefei, China) at −86 °C for 24 h. Subsequently, they were lyophilized for 48 h using a freeze dryer (model BK-FD10PT, Biobase Biodustry Co., Ltd., Jinan, China).

Encapsulation Yield and Efficiency

The quantification of total and surface phenolic compounds was conducted using the Folin–Ciocalteau method, incorporating specific modifications for this study. An amount of 150 mg of encapsulated freeze-dried sample of phenolic extract was dissolved in 3.0 mL of a solvent mixture composed of ethanol, acetic acid, and water in the volumetric proportions of 50:8:42, respectively. The solution was homogenized by vortexing for one minute and subsequently filtered through 0.45 μm regenerated cellulose (RC) syringe filters. The total phenolic content (TPC) was then assessed as described in Section 2.5 of the methodology.

To determine the surface phenolic content (SPC) of the microcapsules, a 300 mg sample of freeze-dried material was subjected to a washing procedure using a 1:1 (v/v) ethanol–methanol mixture for 5 min. Following this, the microcapsules were filtered through 0.45 μm regenerated cellulose (RC) syringe filters. The assessment of SPC utilized the methodology previously described for total phenolic content (TPC) quantification. Equations (2) and (3) were used to calculate % encapsulation efficiency and % yield.

$$\mathrm{\%}\mathrm{E}\mathrm{n}\mathrm{c}\mathrm{a}\mathrm{p}\mathrm{s}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}=\frac{\mathrm{T}\mathrm{P}\mathrm{C}-\mathrm{S}\mathrm{P}\mathrm{C}}{\mathrm{T}\mathrm{P}\mathrm{C}}\times 100$$

(2)

$$\mathrm{\%}\mathrm{Y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}=\frac{\mathrm{T}\mathrm{P}\mathrm{C}\mathrm{e}\mathrm{x}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{a}\mathrm{l}}{\mathrm{T}\mathrm{P}\mathrm{C}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{o}\mathrm{r}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{l}}\times 100$$

(3)

2.10. Preparation of the Enriched Bakery Product

The biscuits and cereal bars were enriched with pumpkin and tomato extracts, respectively, using both crude and encapsulated forms of the extracts. A fortification level of 5%, based on the dry matter used for extraction, was employed. Consequently, for every 100 g of the unbaked product, extracts coming from 5 g of dry tomato or pumpkin were incorporated to the corresponding products. The products were prepared by adding extracts either as non-encapsulated or as encapsulated freeze-dried powders. Additionally, control products without the extracts were made to measure their baseline phenolic content.

The preparation of the products was based on the ingredients within

Table 2. Preparation of biscuits with pumpkin extract.

Ingredients	Quantity in Grams
Bread flour	112.0
Sugar	2.00
Olive oil	5.40
Fast action yeast	0.60
Sea salt	1.70
Water	50.0

and

Table 3. Preparation of Cereal bars with tomato extract.

Ingredients	Quantity in Grams
Oat flakes	82.0
Sugar	0.50
Olive oil	10.8
Fast action yeast	0.60
Sea salt	1.70
Water	36.6

. The dry ingredients—flour, sugar, and salt—were combined in a mixing bowl. Yeast, warm water, and olive oil were then added, and the mixture was kneaded until the dough became smooth and elastic. Two separate batches of dough were prepared for the enriched products. The encapsulated products contained 9.5 g of encapsulated extract per 100 g of dough, while the non-encapsulated products included 0.25 g of freeze-dried extract per 100 g of dough.

All products were baked for 15 min at 180 °C in a in a heating chamber with natural convection (Binder, Tuttlingen, Germany).

2.11. Stability Evaluation

Crude and encapsulated extracts along with the final products were evaluated for total phenolic content under various storage conditions. Samples were stored (as is) in amber vessels at an ambient temperature of 25 °C. The samples were also subjected to accelerated temperature conditions in a heating chamber with natural convection (Binder, Tuttlingen, Germany) equipped with controlled temperature settings and maintained in darkness at 65 °C [23]. After the completion of stability duration, the experimental procedure described in Section 2.4, Section 2.5 and Section Encapsulation Yield and Efficiency was followed.

3. Results

3.1. Model Fitting

The ultrasound-assisted extraction (UAE) of total phenolics from pumpkin and tomato extracts was optimized using the response surface methodology. This study utilized a central composite design (CCD) to explore the interaction effects between the ethanol concentration, set at 30, 50, or 70% (v/v) (X₁), and the time duration, set at 20, 40, or 60 min (X₂).

The experimental matrix was constructed using the central composite design (CCD) and consisted of 13 experimental combinations, incorporating two central points. This configuration facilitated a randomized sequence of experiments to reduce the influence of extraneous variables on the response measurements. The concentrations of total phenolic compounds measured in the dry extracts were as follows: for the tomato extract, the values ranged from 2.95 to 9.91 mg GAE/g; for the pumpkin extract, the range was from 0.98 to 4.35 mg GAE/g. The results are detailed in

Table 4. CCD design with coded and actual values along with the total phenolics of the tomato and pumpkin extracts.

Run	Independent Factors		Response TPC (mg/g) 1
	X1 Ethanol (% v/v)	X2 Time (min)	Tomato Extract	Pumpkin Extract
1	50 (0)	40 (0)	9.91	2.79
2	30 (−1)	20 (−1)	3.12	3.98
3	50 (0)	20 (−1)	5.56	2.45
4	50 (0)	40 (0)	8.91	3.15
5	30 (−1)	40 (0)	4.21	4.35
6	50 (0)	40 (0)	9.25	3.04
7	70 (+1)	20 (−1)	3.24	0.98
8	70 (+1)	40 (0)	5.12	2.69
9	50 (0)	60 (+1)	6.76	1.68
10	70 (+1)	60 (+1)	6.25	1.89
11	50 (0)	40 (0)	8.56	2.56
12	30 (−1)	60 (+1)	2.95	2.55
13	50 (0)	40 (0)	8.94	2.88

¹ Results are calculated as equivalents of gallic acid (GAE) in mg for each g of dry tomato or pumpkin by TPC test.

.

The model’s overall fit to the data was rigorously evaluated to ascertain the extent to which a combination of the linear, quadratic, and interaction terms could significantly account for the variability observed in the response variables. For tomato extract, the F-value was reported as 92.71 with an associated p-value nearing zero, while for pumpkin extract, an F-value of 32.61 with a p-value of 0.000 was noted. These statistical metrics indicate that both models are highly significant. This significance suggests that the factors, namely ethanol concentration and processing time, and their interactions have a substantial collective impact on explaining the variations observed in the response variables (

Table 5. Outcomes of the ANOVA for the fitting of transformed data from total phenolics (TPC).

	Tomato			Pumpkin
Source	DF	F-Value	p-Value	F-Value	p-Value
1 Model	5	92.71	0.000	32.61	0.000
Linear	2	151.38	0.000	45.62	0.000
Ethanol	1	251.73	0.000	34.14	0.001
time	1	25.61	0.001	67.90	0.000
Square	2	185.50	0.000	24.44	0.001
Ethanol × Ethanol	1	187.21	0.000	10.78	0.013
time × time	1	53.62	0.000	48.41	0.000
2-Way Interaction	1	30.98	0.001	24.56	0.002
Ethanol × time	1	30.98	0.001	24.56	0.002
Error	7
Lack-of-Fit	3	5.48	0.067	1.16	0.429
Pure Error	4
Total	12

¹ Box–Cox data transformation was conducted using optimal λ = −0.495565 for tomato, and λ = 1.30359 for pumpkin.

).

The individual linear impacts of ethanol and time on the response variables are markedly significant, as evidenced by their substantial F-values (251.73 and 34.14 for ethanol and 25.61 and 67.90 for time, in the cases of tomato and pumpkin, respectively), accompanied by exceedingly low p-values. These data strongly support the existence of a significant linear relationship between each factor and the corresponding response variable. Furthermore, the quadratic effects of ethanol and time also demonstrate considerable significance, with F-values recorded at 187.21 and 53.62 for tomato, and 10.78 and 48.41 for pumpkin, respectively, alongside similarly low p-values. These findings suggest that the associations between the factors and the response variables include a quadratic component, indicating the presence of curvature in the relationships that must be considered (Table 5).

The interaction between ethanol and time is markedly significant, as evidenced by F-values of 30.98 for tomatoes and 24.56 for pumpkins, coupled with low p-values. This finding indicates that the impact of one factor (such as ethanol) on the response variable is significantly moderated by the other factor (such as time), and this influence is reciprocal (Table 5).

Concerning model adequacy, while some deviation from a perfect fit is noted, as evidenced by an F-value of 5.48 and a p-value of 0.067, these results do not reach the threshold of statistical significance at the conventional level of 0.05. Thus, it can be concluded that the model demonstrates an acceptable fit to the data. Additionally, the p-value of 0.429 for the variable “pumpkin” further supports the notion that the model provides a satisfactory fit to the observed data (Table 5).

In conclusion, the analysis of variance (ANOVA) for both extracts demonstrates that the model is highly significant. This significance extends to both the linear and quadratic effects of the factors involved. Furthermore, the interaction between these factors also shows statistical significance. The lack-of-fit test, which is not statistically significant, indicates that the model provides an adequate fit to the data. This is corroborated by the lack-of-fit test results, reinforcing the model’s reliability and the robustness of its conclusions (Table 5). Polynomial predictive equations that are utilized to represent response surfaces of total phenolics (TPC) in tomato and pumpkin are presented in

Table 6. Quadratic models refer to polynomial predictive equations that are utilized to represent response surfaces of total phenolics (TPC) in tomato and pumpkin.

TPC	1 Predictive Equations		R2	R2 Adjusted
Tomato	−Y−0.5 = −1.3419 + 0.02943 ethanol + 0.00969 time − 0.000317 ethanol × ethanol − 0.000170 time × time + 0.000107 ethanol × time	(4)	98.51%	97.45%
Pumpkin	Y = 7.47 − 0.2195 ethanol + 0.1138 time + 0.001166 ethanol × ethanol − 0.002471 time × time + 0.001463 ethanol × time	(5)	95.88%	92.94%

¹ TPC: Total phenolic content presented as mg of Gallic acid equivalents (GAE) per g of dry sample. X₁: Ethanol concentration (%, v/v); X₂: Time in min.

.

Ultrasound-assisted extraction (UAE) is a method that employs acoustic cavitation to improve the efficiency of the extraction process. This technique involves the application of ultrasound waves to a solvent, inducing alternating cycles of high and low pressure within the medium. During the low-pressure cycles, cavitation bubbles form because of the solvent’s rapid expansion. Subsequently, these bubbles collapse abruptly under the high-pressure cycles, producing localized hot spots and high-speed liquid jets. These phenomena significantly enhance the mass transfer and extraction rates.

Cavitation-induced microjets and shock waves are pivotal in ultrasonic-assisted extraction (UAE), exerting significant mechanical forces on plant tissues. This action facilitates the physical breakdown of cell walls, thereby augmenting the liberation of phenolic compounds into the extraction solvent. Additionally, the turbulent flow resulting from cavitation enhances mass transfer, thereby accelerating the kinetics of extraction and improving the overall efficiency of the process.

Ethanol and water are effective solvents for ultrasound-assisted extraction, primarily because of their proficient transmission of ultrasound waves. These solvents possess relatively low acoustic impedances, facilitating the penetration and propagation of ultrasound energy through the solvent with minimal attenuation. Consequently, this allows for effective coupling of ultrasound energy with the solvent, optimizing cavitation effects and thus enhancing the extraction process.

Ethanol and water are highly effective solvents, also due to their capacity to dissolve a diverse array of compounds. Water, a strongly polar solvent, is particularly adept at solubilizing hydrophilic phenolic compounds, including flavonoids and glycosides. Ethanol, although polar, exhibits some non-polar characteristics as well, enabling it to dissolve both hydrophilic and lipophilic phenolic compounds. The combined solubility properties of ethanol and water make their mixtures particularly efficient for the extraction of a wide spectrum of phenolic compounds from botanical sources, such as tomatoes and pumpkins.

The solubility of phenolic compounds in ethanol–water mixtures can be significantly enhanced through the manipulation of the ethanol-to-water ratio. Adjusting this ratio allows for the tuning of the solvent mixture’s polarity, facilitating the selective extraction of distinct phenolic classes. Specifically, increased concentrations of ethanol preferentially extract more lipophilic phenolic compounds, whereas reduced concentrations are optimal for hydrophilic phenolics. This adaptability in solubility parameters renders ethanol–water mixtures highly effective for the extraction of phenolic compounds with diverse chemical characteristics.

3.2. Extraction Optimization and Model Validation

The response surface methodology was employed to evaluate the synergistic effects of two variables to optimize the extraction of phenolic compounds. Figure 1 and Figure 2 present the three-dimensional response surface plots that demonstrate the combined influence of these independent variables on the yield of extracted phenolics. Notably, the graphs focus solely on the variables identified as having a substantial effect.

Phenolic compounds are widely recognized for their broad spectrum of biological activities and their significant implications for human health, which have made them a focal point in scientific investigations. However, the task of unraveling their complex interactions and refining extraction techniques to maximize their benefits remains a formidable challenge. In this context, the response surface methodology (RSM) emerges as a critical statistical tool. RSM provides a structured method to investigate the intricate relationships among multiple variables that affect the extraction of phenolic compounds, facilitating a more effective harnessing of their potential.

By carefully adjusting critical variables, including solvent type and extraction duration, response surface methodology (RSM) facilitates the efficient design of experiments and maximizes phenolic yield while minimizing resource use. Furthermore, RSM aids in the creation of predictive models that illuminate the ideal conditions for extracting phenolic compounds from diverse matrices. Through its repetitive cycle of experimental testing and model enhancement, RSM not only improves the efficiency of extraction but also advances knowledge of the behavior of phenolic compounds and their prospective uses in functional foods.

The response surface plots presented in Figure 1 and Figure 2 demonstrate a preferential extraction of total phenolic compounds at moderate ethanol concentrations in the solvent (X₁). Conversely, these figures also reveal a detrimental effect on the extraction process associated with prolonged experimental durations (X₂).

The extraction process was conducted using solutions with different concentrations of ethanol and varying durations. A 54% (v/v) ethanol solution with an extraction time of 46 min resulted in the highest yield of phenolic compounds, extracting 9.47 ± 0.08 mg GAE per gram of dry tomato extract. Conversely, a 31% (v/v) ethanol solution with a 30 min extraction period achieved a maximum yield of 4.52 ± 0.05 mg GAE per gram of dry pumpkin extract, in terms of total phenolic content.

The optimal conditions were identified utilizing the response optimizer function within the Minitab^® statistical software. The corresponding results are presented in Table 6.

The predictive model’s validation was achieved through a comparative analysis between the nominal values and their experimental counterparts under the optimal conditions detailed in

Table 7. Predicted and experimental values for TPC upon optimal conditions for tomato and pumpkin extracts.

Independent Factor 1 TPC (mg/g DM)	Predicted Values 1	Experimental Values	Desirability
Tomato	9.46	9.47 ± 0.08 a	0.9300
Pumpkin	4.44	4.52 ± 0.05 a	1.0000

¹ The optimal conditions were 31% and 54% (v/v) ethanol content and the duration of 30 min and 46 min for pumpkin and tomato, respectively, for TPC. ^a in the same group of measurements the denote values indicate that are not statistically different. DM: dry matter.

. The comparison between nominal and predicted values and the actual measurements revealed no statistically significant differences (p > 0.05), suggesting that the response optimization process was highly accurate.

The precision of the optimal sample can be characterized by its desirability outcome. Greater precision in optimization correlates with a desirability outcome approaching the value of 1.0000. Consequently, in an ideal setting, there is no statistical difference between the model verification results and the predicted responses [24].

3.3. Determination of Antioxidant Capacity and Total Phenolics

Table 8. Antioxidant capacities of tomato and pumpkin extract upon optimized conditions.

1 Parameters	Tomato Optimized Extract	Pumpkin Optimized Extract
DPPH (μmol TE/g)	7.65 ± 0.08 a	5.78± 0.05 b
ABTS (μmol TE/g)	9.27 ± 0.02 a	3.95 ± 0.04 b
FRAP (μmol TE/g)	5.25 ± 0.09 a	2.99 ± 0.03 b
CUPRAC (μmol TE/g)	2.3 ± 0.04 a	1.25 ± 0.05 b
TPC (mg GAE/g)	9.47 ± 0.08 a	4.52 ± 0.05 b

Results are presented as mean ± SD of three measurements. ¹ DPPH, ABTS, FRAP, CUPRAC: Results are expressed as trolox equivalents (TE) in μmol for each g of dry extract of tomato or pumpkin; TPC: Total phenolics are expressed as gallic acid equivalents (GAE) in mg for each g of tomato or pumpkin dry extract. Different letters on each row denote significant difference (p < 0.05) between the tomato and pumpkin values for each parameter based on independent t-tests statistical analysis.

presents the results for the antioxidant capacity of optimized tomato and pumpkin extracts, evaluated through the DPPH, ABTS, FRAP, and CUPRAC assays, in conjunction with their total phenolic content (TPC).

The ABTS and DPPH assays both evaluate a compound’s ability to neutralize free radicals. Conversely, the FRAP and CUPRAC assays are utilized to assess the capacity of the sample to reduce ferric (Fe) and cupric (Cu) ions, respectively.

3.4. Encapsulation Efficiency and Yield Evaluation

In initial trials, maltodextrin, a predominant encapsulating agent utilized in microencapsulation processes, was assessed as a prospective core material for microencapsulation through freeze-drying, utilizing various core-to-coating material ratios. The choice of coating material was informed by several advantageous properties, including its high solubility in water, low viscosity, ability to form stable solids, resilience during the digestion process, biodegradability, safety, and cost-effectiveness [25].

The results from preliminary experiments indicated that a core–coating ratio of 1:30 (w/w) of maltodextrin was optimal and was therefore selected for subsequent analyses. The yields of total phenolic content in the freeze-dried tomato encapsulated extract varied between 62.2% and 89.2%, as determined using Equation (3) and as detailed in

Table 9. The %Yield of total phenolics in freeze-dried encapsulated extracts.

Core–Coating Ratio (w/w)	Tomato Encapsulated Extract % Yield of Total Phenolics
1:10	72.5 ± 1.0 a
1:15	63.5 ± 0.3 b
1:20	62.2 ± 0.4 c
1:25	67.9 ± 0.7 d
1:30	89.2 ± 0.7 e

Different letters denote significant difference (p < 0.05) between each ratio values based on one-way ANOVA statistical analysis, followed by Tukey’s post hoc analysis. In the experiments that employed a 1:30 weight-to-weight, core-to-coating ratio, where a higher percentage yield was achieved, the encapsulation efficiency of total phenolics was determined to be 99.2%, as calculated using Equation (2). These selected parameters were subsequently utilized for the encapsulation of the pumpkin extract.

.

3.5. Stability Evaluation for the Extracts

The stability of both crude and encapsulated extracts, as well as products formulated using the encapsulated extract, was assessed at temperatures of 65 °C and 25 °C over a period of 12 days. According to Palamutoğlu et al., one day of storage at 65 °C is equivalent to approximately one month of storage under ambient conditions [23]. Therefore, a storage period of 12 days at 65 °C may be equivalent to 12 months of storage at ambient room temperatures.

Samples were collected at three-day intervals.

Table 10. Stability results for tomato extract.

Days	Extract Stability Testing at 25 °C (%Content)/±SD		Stability Testing at 65 °C (%Content)/±SD
	Crude	Encapsulated	Crude	Encapsulated
0	100.0 ± 0.4 a	100.0 ± 1.1 a	100.0 ± 0.4 a	100.0 ± 1.1 a
3	90.2 ± 1.1 b	97.1 ± 0.5 c	77.9 ± 0.5 d	98.1 ± 1.2 c
6	77.2 ± 1.0 e	96.2 ± 0.4 c	65.9 ± 0.9 f	92.8 ± 1.5 g
9	59.2 ± 0.7 h	93.1 ± 1.2 i	52.8 ± 0.8 j	89.1 ± 1.5 k
12	49.2 ± 0.9 l	89.8 ± 0.8 m	39.9 ± 1.1 n	77.4 ± 0.5 o

Different letters on each row and column denote statistical difference (p < 0.05) based on one-way ANOVA analysis, followed by Tukey’s post hoc analysis.

and

Table 11. Stability results for pumpkin extract.

Days	Extract Stability Testing at 25 °C (%Content)/±SD		Stability Testing at 65 °C (%Content)/±SD
	Crude	Encapsulated	Crude	Encapsulated
0	100.0 ± 0.8 a	100.0 ± 0.9 a	100.0 ± 0.8 a	100.0 ± 0.9 a
3	82.2 ± 0.7 b	99.1 ± 1.2 a	79.5 ± 1.1 b	95.4 ± 0.4 a
6	78.7 ± 0.6 c	92.3 ± 1.7 b	61.1 ± 1.7 c	87.8 ± 1.5 b
9	61.1 ± 0.9 d	88.4 ± 1.1 b	45.2 ± 1.3 d	81.7 ± 1.0 c
12	49.4 ± 0.8 e	79.9 ± 0.8 c	38.1 ± 0.5 e	71.8 ± 1.4 d

Different letters on each row and column denote statistical difference (p < 0.05) based on one-way ANOVA analysis, followed by Tukey’s post hoc analysis.

, along with Figure 3 and Figure 4, display the results as percentage correlations between the concentrations measured at each time point and the initial total phenolic content for each condition and sample under evaluation.

Table 10 and Table 11, as well as Figure 3 and Figure 4, display the changes in total phenolic content throughout the stability testing conducted, under both the ambient and accelerated conditions for the crude and encapsulated extracts. The data reveal that the degradation rate of phenolics in the crude extract was significantly greater compared to that of the encapsulated extract. This observation highlights the critical role of encapsulation in preserving the stability of bioactive compounds. These results support previous studies that have identified encapsulation as a valuable strategy for protecting various bioactive substances [26,27,28,29].

The concentration of phenolic molecules demonstrated a marginal decline over successive time intervals under both sets of stability conditions for the encapsulated extracts. The stability results for both crude and encapsulated extracts reinforce the idea that encapsulation plays an essential role in reducing degradation and substantial damage to bioactive compounds. Overall, the data suggest a negative correlation between temperature and phenolic concentration. Nonetheless, the impact of thermal effects on phenolic content remains ambiguous, as evidenced by the disparate findings of multiple studies [30,31,32].

3.6. Enriched Products Evaluation

The bakery products—specifically, biscuits enriched with pumpkin extract and cereal bars enriched with tomato extract (Figure 5)—were produced following the methodologies outlined in Section 2.10. This process employed encapsulated extracts to ensure the phenolics’ relative stability, maintaining concentrations above 80% for the entirety of the stability testing period.

The products were formulated both with and without encapsulated extract to assess the percentage increase in total phenolic compounds and evaluate the antioxidant capacity, as measured by the ABTS, DPPH, CUPRAC, and FRAP assays (

Table 12. Results of enriched and control bakery products.

Test	Tomato Cereal Bars			Pumpkin Biscuits
	Enriched	Control	%Increase	Enriched	Control	%Increase
Total phenolic content	1.51 ± 0.03 a	1.20 ± 0.01 b	25.8	1.72 ± 0.02 a	1.67 ± 0.03 a	3.0
DPPH	3.67 ± 0.04 a	2.65 ± 0.03 b	38.5	3.37 ± 0.02 a	2.28 ± 0.04 b	47.8
ABTS	11.63 ± 0.04 a	9.68 ± 0.02 b	20.2	9.55 ± 0.05 a	8.70 ± 0.03 b	9.7
FRAP	17.59 ± 0.05 a	13.92 ± 0.06 b	26.4	9.81 ± 0.06 a	7.38 ± 0.04 b	32.9
CUPRAC	19.35 ± 0.04 a	13.79 ± 0.02 b	40.3	11.75 ± 0.02 a	11.32 ± 0.03 a	3.8

Significant differences (p < 0.05) based on independent t-test for each test between enriched and control samples of tomato cereal bars or pumpkin biscuits are denoted by different letters within in each row. Results are presented as mean ± SD of three measurements. DPPH, ABTS, FRAP, CUPRAC: Results are expressed as Trolox equivalents (TE) in μmol for each g of prepared bakery product; TPC: Total phenolics are expressed as gallic acid equivalents (GAE) in mg for each g of prepared bakery product. ^a in the same group of measurements the denote values indicate that they are not statistically different.

).

4. Discussion

This research utilized a central composite design to optimize the extraction process and determine the optimal conditions for extracting phenolic compounds from tomatoes and pumpkins grown on Lemnos Island, aiming to maximize total phenolic content. Additionally, this study verified the effectiveness of encapsulation in protecting these bioactive compounds from degradation during storage. The successful encapsulation of phenolic compounds from the extracts of Lemnian tomatoes and pumpkins (Cucurbita moschata) has notably enhanced the stability of phenolic-enriched bakery products. Nonetheless, evaluating their bioavailability is crucial to ascertain the commercial feasibility of these fortified products.

With regards to the antioxidant activity evaluation, the ABTS assay is versatile as it can be used for testing both hydrophilic and lipophilic antioxidant systems. The minor discrepancies observed in the results from these assays under optimized conditions indicate that the phenolic compounds influencing free radical scavenging capacity tend to be hydrophilic. On the other hand, the DPPH assay is more suited for assessing hydrophobic systems [33]. The results of the present study were in accordance with previous studies that referred comparable results for tomato and pumpkin total phenolic content. Specifically, in the context of tomato extracts, the total phenolic content varied from 312.2 mg GAE per kg to 557.8 mg per kg of fresh tomato, following extraction with methanol. Correspondingly, the antioxidant activity, as measured by DPPH assay, ranged between 0.81 and 1.74 mmol Trolox per kg of fresh tomato, as documented by Violeta Nour et al. [34]. Similarly, in the study conducted by Ramandeep Toor et al. [35], the quantification of hydrophilic and lipophilic phenolic compounds yielded a concentration of 24.7 mg/100 g of fresh tomato. Comparable promising outcomes were observed in pumpkin extracts, where the phenolic compounds were quantified to be between 26.31 and 79.86 mg/100 g of fresh pumpkin in the investigation by Daniela Priori et al. [36], while antioxidant activity ranged from 71.09 to 357.742 μg of Trolox per g of fresh pumpkin. Moreover, the study by Meriem Mokhtar et al. [37] reported total phenolic content of up to 97.4 mg/100 g of fresh pumpkin, alongside antioxidant activity measuring up to 0.065 and 0.074 μmol Trolox per g of fresh pumpkin for DHHP and ABTS assays, respectively.

Concerning total phenolic content (TPC), the data show that TPC in tomato extract exhibits a saddle-shaped response to ethanol concentration. At moderate ethanol levels (50% v/v), TPC tends to be higher. Higher ethanol concentrations (70% v/v) generally result in lower TPC, while moderate ethanol levels (50% v/v) are associated with higher TPC values. TPC generally increases with the extraction time but shows an optimum at 40 min for 50% ethanol concentration. Beyond 40 min, especially at higher ethanol concentrations, the TPC does not significantly increase and may even decrease. This saddle-shaped response indicates that there is an optimum ethanol concentration and extraction time that maximizes TPC in tomato extracts. For pumpkin extract, the data suggest that increasing ethanol concentration negatively impacts TPC. Higher ethanol levels (70% v/v) consistently result in lower TPC values. Lower ethanol concentrations (30% v/v) are associated with higher TPC. Extraction time appears to have a less pronounced effect on TPC in pumpkin extract compared to ethanol concentration. While there is some variation, the impact is not as significant as that of ethanol concentration. The negative effect of higher ethanol concentrations on TPC in pumpkin extract may be due to the nature of the phenolic compounds in pumpkin, which might be more soluble in water.

The predictive model was thoroughly evaluated and found to be adequately robust. The experimental setup led to optimized conditions that refer to the usage of a 54% (v/v) aqueous ethanol solution as solvent and an extraction duration of 46 min for tomato, and 31% (v/v) aqueous ethanol solution and an extraction duration of 30 min for pumpkin. These optimized conditions facilitated significant extractions of total phenolics, quantified as 9.47 ± 0.08 and 4.52 ± 0.05 mg GAE per gram from the dry tomato and pumpkin (Cucurbita moschata) extracts, respectively. The results demonstrate the successful isolation of phenolic compounds from Lemnian tomatoes and pumpkins through ultrasound-assisted extraction, utilizing eco-friendly solvents such as ethanol and water. These extracts, which are abundant in antioxidants, are particularly valuable for culinary uses, given the increasing preference for natural substitutes to synthetic antioxidants.

The rate at which the total phenolics decreased was greater in the crude extract than in the encapsulated extract, highlighting the protective effect of encapsulation. Both extracts demonstrated substantial stability in antioxidant activities despite exposure to storage conditions and elevated temperatures. Nevertheless, during the accelerated stability testing, a minor reduction in total phenolic compounds was observed in the encapsulated extract, when compared to the crude extract, with the conditions maintained at 25 °C.

Encapsulation successfully maintained the phenolic compounds within the extracts, indicating promising applications within the food industry. The growing demand for natural antioxidants justifies additional investigation into extracts that possess advantageous properties. This emphasizes the significance of employing such extracts to enhance the nutritional content of food products.

The comparative analysis between enriched products and traditional controls has demonstrated a markedly beneficial effect from the incorporation of encapsulated extracts into conventional products. Specifically, this enhancement was evident in the significant increase in total phenolic compounds and antioxidant capacity. These results highlight the substantial potential of using encapsulated extracts to augment the nutritional and functional qualities of traditional bakery items, thereby providing superior health advantages to consumers.

Utilizing local vegetables like tomatoes and pumpkins (Cucurbita moschata) in the formulation of functional foods is a promising strategy for regional health and economic growth, but it is not without challenges. One critical issue is the variability in the bioactive compound content of vegetables, which can be influenced by cultivation conditions, harvest time, and post-harvest handling. This variability can affect the consistency and efficacy of the functional foods produced. Moreover, research on bioavailability is essential for understanding the absorption, metabolism, and physiological effects of encapsulated phenolic compounds following ingestion. It is vital to evaluate variables such as their interaction with the bakery product matrix, the influence of digestive processes, and the kinetics of absorption. By investigating the bioavailability of these compounds, we can determine their efficacy and safety for human consumption. Furthermore, there is a need for robust scientific research to substantiate the health claims associated with functional foods. Regulatory frameworks must also be developed to ensure these products are both safe and effective for consumer use. Additionally, consumer education is crucial as it fosters an understanding of the benefits of functional foods, which is essential for market acceptance and growth [38,39].

5. Conclusions

In conclusion, this study has highlighted the efficacy of encapsulation techniques in enhancing the stability of phenolic compounds within bakery products enriched with extracts from Lemnian tomatoes and pumpkins (Cucurbita moschata). The encapsulation process effectively preserves these bioactive elements during storage. Utilizing local vegetables to create functional foods not only promises substantial benefits for regional economic growth and public health improvements but also demands rigorous attention to agricultural methodologies, scientific substantiation, and adherence to regulatory standards to fully capitalize on these advantages. The successful integration of these components is crucial for the widespread acceptance and ongoing consumption of functional foods. Further research into the bioavailability of these encapsulated phenolic compounds could lead to the development of bakery products that provide superior nutritional and health benefits, which is essential for their successful market introduction.