Quality Evaluation and Shelf-Life Prediction of a Mixed Mango and Passion Fruit Smoothie Under Dimethyl Dicarbonate Treatment and Packaging Interventions

Abstract

This study investigated shelf-life prediction of a cold-stored mixed mango–passion fruit smoothie (60:40) using kinetic modeling to compare the effects of dimethyl dicarbonate (DMDC, 250 ppm), pasteurization (90 °C for 100 s), and packaging type (glass vs. polyethylene terephthalate (PET)) during six weeks at 4 °C. Physicochemical parameters, functional properties (total phenolic content, total flavonoid content, and antioxidant activity by 2,2-diphenyl-1-picrylhydrazyl (DPPH) and Ferric Reducing Antioxidant Power assay (FRAP), and microbial stability were monitored weekly. Zero- and first-order kinetic models were applied to describe quality changes, with the first-order model showing superior fit (average R² = 0.936). pH remained relatively stable (p > 0.05), while total soluble solids (TSS) gradually declined in all treatments from approximately 16–17 °Brix to 13–14 °Brix by week 6. PET packaging resulted in a significantly higher total color difference (ΔE) than glass by the end of storage (p ≤ 0.05), particularly in DMDC-treated samples. Pasteurization reduced initial polyphenol oxidase (PPO) activity by 44–56% compared with untreated and DMDC-treated samples (p ≤ 0.05), whereas PET generally exhibited higher residual PPO activity than glass. DMDC treatment better preserved antioxidant capacity, phenolics, and flavonoids, with significantly higher DPPH and FRAP values than controls at week 6 (p ≤ 0.05). Microbiologically, DMDC effectively suppressed total viable counts (<5 log CFU/mL) and yeast and mold (<3 log CFU/mL), outperforming pasteurization. Shelf-life was estimated at 27–29 days for pasteurization and 41–42 days for DMDC (250 ppm), particularly when combined with glass packaging. Overall, the DMDC–glass combination demonstrated strong potential as a non-thermal preservation approach for fruit beverages.

1. Introduction

The World Health Organization recommends a daily intake of at least 400 g of fruits and vegetables to reduce the risk of chronic diseases such as diabetes, cardiovascular disorders, and cancer [1]. In response to growing consumer demand for convenient, nutrient-rich foods, blended fruit products like smoothies have gained popularity due to their fresh taste, high bioactive compound content, and health benefits [2]. Homemade or minimally processed tropical fruit smoothies typically exhibit a shelf-life of only 1–2 weeks under refrigeration (4 °C) due to rapid microbial growth and enzymatic browning in the absence of hurdles beyond hygiene and cold chain. Commercial ready-to-drink smoothies often achieve longer stability (30–45 days) through additional preservation steps [3].

Current preservation methods, such as thermal pasteurization, can extend shelf-life but often compromise nutritional and sensory qualities [4]. Non-thermal alternatives include dimethyl dicarbonate (DMDC), an FDA-approved antimicrobial processing aid (21 CFR 172.133), which is permitted at up to 250 ppm in certain beverages, including non-carbonated 100% juice products and dilute juice-containing drinks (juice content ≤ 50%). Its use in pulpy smoothies is consistent with these limits, as it hydrolyzes rapidly to methanol and CO₂ [5,6]. Previous studies have demonstrated that DMDC at 250 ppm can ensure microbial quality in passion fruit juice without substantially affecting quality attributes during 24 days of refrigerated storage [7].

While prior work has examined DMDC in mango–passion fruit smoothies [2], limited data exist on kinetic-based shelf-life prediction across physicochemical, functional, and microbial attributes.

Packaging further modulates product shelf-life by affecting oxygen permeability, light exposure, and chemical migration [8,9]. While glass is impermeable to gases, polyethylene terephthalate (PET) bottles are lightweight and cost-effective but may accelerate oxidation [10]. Prior research on packaging effects has centered on single-strength juices [3] or thermally processed beverages, with limited data on smoothies preserved with non-thermal agents like DMDC.

Changes in beverage characteristics during storage can lead to the depletion of essential nutrients and a decline in consumer preference. Therefore, ensuring food safety throughout the product’s shelf-life is crucial to maintaining its quality until consumption. A widely used approach to assess beverage quality during storage is determining its shelf-life, which represents the maximum period a food product can be stored under specific environmental conditions without significant deterioration in quality or consumer acceptability, or before it becomes unsuitable for consumption [11,12].

Shelf-life determination is a complex process influenced by multiple factors, including the type of food product, storage conditions, packaging, and manufacturing methods. Predicting food quality degradation requires identifying key quality indicators through empirical research [2]. One effective approach to assessing quality changes over time is the use of mathematical models [13,14]. Zero-order and first-order models are commonly applied to describe the formation or degradation of quality attributes during storage. Previously developed predictive models have been used to estimate the shelf-life of various products based on quality indicators and consumer preferences [14].

This study addresses the gaps in the literature regarding non-thermal preservation methods for multi-fruit smoothies by evaluating (a) the effects of DMDC treatment and packaging type on quality attributes of mango–passion fruit smoothie at different storage intervals, (b) the efficacy of DMDC in suppressing microbial growth and increasing the shelf-life compared to traditional pasteurization and (c) the role of packaging in mitigating quality degradation (e.g., color, PPO activity) during cold storage. The novelty lies in the direct comparison of DMDC (non-thermal) and pasteurization treatments evaluated under two packaging systems (glass and PET) at different storage intervals, together with kinetic modeling and shelf-life prediction in a multi-fruit smoothie matrix. By elucidating these, this work provides actionable insights for optimizing smoothie preservation, balancing safety, quality, and sustainability in the beverage industry.

2. Materials and Methods

2.1. Preparation of Mixed Mango and Passion Fruit Smoothie

Ripe mangoes (cv. Nam Dok Mai, Thailand origin, maturity index ~14–16 °Brix, peel 70–80% yellow–red) and passion fruits (Passiflora edulis f. flavicarpa, purple type, maturity stage, full color with wrinkled skin) exhibiting uniform size and free from visible defects, were sourced from a retail store in Bangkok and transported to the Department of Food Technology, Faculty of Science, Chulalongkorn University. Fruits were immersed in 200 ppm sodium hypochlorite solution (pH 6.5, 25 °C) for 5 min. Free chlorine was verified using test strips (LaMotte, Chestertown, MD, USA). Mangoes were peeled and passion fruit pulp was sieved to remove seeds. The mango pulp and passion fruit were homogenized using a food processor (KitchenAid, 5KPM5, Springfield, OH, USA) at speed setting ‘2’ (approx. 170 rpm blade speed) for 3 min to achieve a uniform mixture. The blended puree was then combined at a 60:40 (v/v) ratio of mango to passion based on preliminary sensory trials to achieve an optimal balance of sweetness and acidity using graduated cylinders and stored at 4 °C. Treatment (DMDC addition or pasteurization) was applied within 30 min of puree preparation to minimize pre-treatment enzymatic degradation.

2.2. DMDC Treatment and Packaging Types

The optimal DMDC concentration (250 ppm (mg/L beverage)) was selected based on regulatory limits as well as findings from the previous research [2]. DMDC (≥99.5% purity, Sigma-Aldrich, St. Louis, MO, USA) was added directly to the smoothie at a concentration of 250 ppm using a micropipette under constant stirring. Mixed mango and passion fruit smoothie samples without DMDC addition served as the control and were compared with pasteurized samples (90 °C holding temperature for 100 s in a water bath (Memmert, Schwabach, Germany), with a come-up time of ~2–3 min and immediate ice-water cooling to <10 °C). The control group (no DMDC, no pasteurization) underwent identical handling procedures, including the 30 min hold time and stirring, to isolate the effect of the preservation treatments. A 100 mL portion from each sample group (control, DMDC-treated at 250 ppm, and pasteurized) was transferred into two types of sterilized packaging: glass bottles (clear, 120 mL, wall thickness 2.2 mm) and PET bottles (clear, 120 mL, wall thickness 0.5 mm, oxygen transmission rate: 0.12 cc/pkg/day). Headspace volume was standardized to 10 mL. Bottles were sealed with polypropylene screw caps with ethylene vinyl alcohol (EVOH) liners, which were sterilized via UV-C irradiation (254 nm, 15 min per side) in a laminar flow cabinet. Filling was performed cold (4 °C) under ambient air; no headspace flushing or vacuum sealing was applied to simulate standard industry practice for refrigerated products. Bottles were stored in a dark, refrigerated incubator (Binder, Tuttlingen, Germany) at 4.0 ± 0.5 °C under static conditions. Physicochemical properties and microbial populations were analyzed as a destructive sampling design. For each treatment × packaging × time point combination, three independent bottles (n = 3) were removed from storage and analyzed.

2.3. Physicochemical Properties of Treated Mixed Mango and Passion Fruit Smoothie During Storage at 4 °C

The effects of changing DMDC and packaging on the physicochemical properties of mango and passion fruit smoothie during storage at 4 °C were checked weekly for a period of 6 weeks as follows; The value of pH was recorded by digital pH meter (Mettler Toledo, S220, Greifensee, Switzerland). pH was measured by direct probe insertion into the undiluted smoothie after gentle inversion to re-suspend solids. Measurements were taken at 20 ± 1 °C. Total acid (%) was conducted by titration with sodium hydroxide solution according to [15]. Titratable acidity was determined by diluting 10 g of smoothie with 90 mL distilled water and titrating with 0.1 N NaOH to an endpoint of pH 8.2 using an automated titrator. Results were expressed as % citric acid. Total soluble solids were obtained by digital refractometer (Hanna Instrument, 96801, Woonsocket, RI, USA) according to the AOAC method [15]. In brief, the smoothie was centrifuged at 4000× g for 10 min; the supernatant was filtered through Whatman No. 1 paper. The clarified serum was used for °Brix measurement with temperature compensation to 20 °C. Turbidity was measured as absorbance at 660 nm using a UV-Vis spectrophotometer (Thermo Scientific GENESYS 20, Waltham, MA, USA) against a distilled water blank as described by Jafari et al. [2]. Color was measured using a Chroma Meter (Minolta CR-400, Ramsey, NJ, USA) in reflectance mode. Samples (20 mL) were placed in a standard optical glass cuvette (2 cm path length) against a white background. The instrument was calibrated using a white tile. Values are reported as mean ± SD of five readings per sample. ∆E value (color difference) was obtained from the formula:

∆E* = [(L*₁ − L*₂)² + (a*₁ − a*₂)² + (b*₁ − b*₂)²]^1/2 (1)

by making subscript “1” as the default color value of the sample and subscript “2” as the color value measured at each time. The activity of polyphenol oxidase (PPO) was conducted from the modified method of Matsui et al. [16]. For PPO extraction, 5 g of smoothie was homogenized with 10 mL of cold 0.1 M sodium phosphate buffer (pH 6.5) containing 1% polyvinylpolypyrrolidone (PVPP; Boai NKY Pharmaceuticals Ltd., Jiaozuo, China). The homogenate was centrifuged at 10,000× g for 20 min at 4 °C; the supernatant was used as the crude enzyme extract. The buffer was substituted for the sample to produce the blank solution. At 420 nm, the absorbance was measured every 10 s for 5 min throughout the PPO experiment. PPO activity (%) was expressed relative to the activity of the untreated control sample at week 0 (defined as 100%).

2.4. Functional Properties of Treated Mixed Mango and Passion Fruit Smoothie During Storage at 4 °C

Bioactive compounds were extracted by mixing 5 g of smoothie with 20 mL of 80% (v/v) aqueous methanol. The mixture was vortexed for 1 min, sonicated at 25 °C for 15 min, and centrifuged at 4000× g for 10 min. The supernatant was collected for analysis.

Total phenolic content (TPC) was quantified using the Folin–Ciocalteu method. A 0.5 mL aliquot of the extract was mixed with 10 mL of distilled water and vortexed for 3 min. Then, 0.5 mL of 10% Folin–Ciocalteu phenol reagent (Sigma-Aldrich, St. Louis, MO, USA) was added. After a 5 min incubation, 2 mL of 10% (v/v) sodium carbonate solution was introduced, and the samples were incubated in the dark for 120 min. Absorbance was recorded using a UV-visible spectrophotometer (Thermo Scientific™ GENESYS 20, Waltham, MA, USA) at 765 nm, and TPC was expressed as mg GAE/100 g.

Total flavonoid content (TFC) was measured using the aluminum chloride method without NaNO₂ as described by Jafari et al. [1]. Briefly, 1 mL of extract was mixed with 1 mL of 2% AlCl₃·6H₂O in methanol. Absorbance was read at 430 nm after 10 min against a blank of 1 mL extract mixed with 1 mL methanol using a UV-visible spectrophotometer (Thermo Scientific™ GENESYS 20, USA), and TFC was expressed as mg QCE/100 g.

The antioxidant capacity was assessed using the DPPH assay. A 250 μL sample was mixed with 4.75 mL of DPPH solution and incubated in the dark for 15 min. Absorbance was recorded at 515 nm, and the antioxidant activity was calculated using the equation:

DPPH inhibition (%) = [(A_cont − A_sample)/(A_cont)] × 100 (2)

where

A_control is the absorbance of the DPPH solution without sample;

A_sample is the absorbance of the DPPH solution with sample extract.

For quantification, a Trolox standard curve was prepared using concentrations ranging from 0 to 500 μM Trolox under identical assay conditions. The calibration curve was constructed by plotting % inhibition against Trolox concentration, and results were expressed as mM Trolox equivalents per 100 g dry weight (mM TE/100 g).

Ferric Reducing Antioxidant Power (FRAP) assay was determined according to the method described by Shiekh et al. [7]. A 50 μL sample was mixed with 950 μL of FRAP reagent and incubated in the dark for 4 min. Absorbance was recorded at 593 nm. FRAP reagent was prepared fresh daily by mixing 300 mM acetate buffer (pH 3.6), 10 mM TPTZ in 40 mM HCl, and 20 mM FeCl₃·6H₂O in a 10:1:1 (v/v/v) ratio. Results were quantified against a FeSO₄·7H₂O standard curve (100–1000 µM) and expressed as mM Trolox/100 g.

2.5. Microbiological Analysis

The microbiological quality of the smoothie samples—including control (untreated), pasteurized, and DMDC-treated (250 ppm)—was assessed over a 6-week refrigerated storage period. At each sampling interval, 1 mL of each sample was aseptically withdrawn and serially diluted in 0.85% sterile sodium chloride (NaCl) solution. For microbial enumeration, appropriate dilutions were plated in duplicate using the pour plate method. Plating volume was 0.1 mL per plate, spread-plated in duplicate. The dilution series ranged from 10⁰ to 10⁻⁵. The limit of detection was 1.0 log CFU/mL. For statistical analysis, values below the limit of detection were recorded as 0.99 log CFU/mL. Total viable counts were determined using plate count agar (PCA) (AppliChem, Darmstadt, Germany), with plates incubated at 37 °C for 48 h. While 30 °C is common for ambient spoilage, 37 °C was selected to enumerate mesophilic organisms capable of growth at human body temperature, a conservative approach for refrigerated products intended for consumer handling. Yeast and mold counts were assessed using potato dextrose agar (PDA) (Coolaber, Beijing, China) acidified with tartaric acid (pH ~ 3.5), and incubated at 25 °C for 5 days. After incubation, colonies were counted manually, and results were expressed as log colony-forming units per milliliter (log CFU/mL). All microbiological analyses were conducted following procedures outlined in AOAC Official Method 966.23 [15].

2.6. Kinetics Modeling

Quality changes during refrigerated storage—including degradation of bioactive compounds, color loss, reduction in antioxidant activity, and microbial growth—were described using zero-order and first-order kinetic models. To avoid ambiguity in sign convention, degradation and growth processes were expressed using separate rate equations, while maintaining positive rate constants [13,14].

$$-{\displaystyle \frac{dC}{dt}}=k{C}^n$$ (3)

From Equation (3), the following equations can be generated.

ProcessType	Kinetic Order	Integrated Form
Degradation (quality loss)	Zero-order	$C=C_0-k_{0}t$
	First-order	$lnC={lnC}_0-k_{0}t$
Growth (increase/accumulation)	Zero-order	$C=C_0+k_{0}t$
	First-order	$lnC={lnC}_0+k_{0}t$

In the given equations, C and C₀ denote the property value at a specific time and its initial value, respectively, while t represents the storage duration. The parameter k, in zero-order model, signifies the reaction rate constant (day⁻¹ or property·day⁻¹ for zero or first-order model, respectively) and n indicates the reaction order of the changes.

2.7. Predicting the Shelf-Life by Integration of the Quality Properties

Shelf-life was defined as the time required for a given quality attribute to reach a predetermined critical limit. The acceptance criteria used in this study were established based on practical quality deterioration thresholds and literature-supported standards for fruit-based beverages. To determine the kinetic orders, a least-square fitting method was applied by aligning the experimental data with the general expression. The experimental data were fitted to both the zero-order integrated model and the first-order integrated model. Model selection was based on the best statistical fit, considering the highest coefficient of determination (R²) and lowest root mean square error (RMSE). The kinetic order providing the superior fit was subsequently used for rate constant estimation and shelf-life prediction, following the approach established in previous research by Fikry et al. [13].

From the zero-order model, the shelf-life can be determined using this Equation (4).

$$t_{sL}={\displaystyle \frac{C_0-C_{crit}}{k_0}}$$ (4)

while this following equation (Equation (5)), generated from the first-order model, can be used for estimating the shelf-life.

$$t_{sL}={\displaystyle \frac{\mathrm{l}\mathrm{n}(C_0/C_{crit})}{k_1}}$$ (5)

where $t_{sL}$ is shelf-life (days) and $C_{crit}$ is predefined critical limit.

Shelf-life ($t_s$) for each quality attribute was estimated based on its specific critical limit ($C_{crit}$). For color, the threshold was defined as ΔE ≥ 3, representing the minimum perceptible color change. Antioxidant activity (DPPH and FRAP) and bioactive compounds (TPC and TFC) were considered acceptable until they retained at least 50% of their initial values. Microbial shelf-life was determined using a maximum allowable total microbial count of 5 log CFU mL⁻¹. These thresholds provided the reference points for calculating the shelf-life of each attribute using the corresponding zero- or first-order kinetic model.

2.8. Statistical Analysis

The experiment was conducted using three independent smoothie production batches prepared on separate days under identical formulation and processing conditions. These batches were considered biological replicates (n = 3). For each batch, smoothie samples were filled into glass and PET bottles and stored at 4.0 ± 0.5 °C. Sampling was destructive; a separate unopened bottle was used for each treatment × packaging × storage-time combination. For each biological replicate at each sampling time, all physicochemical, antioxidant, color, enzymatic, and microbiological analyses were performed in duplicate (technical replicates). The mean of the duplicate analytical measurements was calculated and used as the value representing that biological replicate. Statistical comparisons were performed separately at each storage week. For each time point, data were analyzed using one-way analysis of variance (ANOVA) to compare treatment × packaging combinations. When significant differences were detected (p < 0.05), mean separation was conducted using Tukey’s honestly significant difference (HSD) test at a 95% confidence level. Statistical analyses were performed using SPSS version 22 (IBM SPSS Statistics, Armonk, NY, USA). Regression analysis was conducted independently for each treatment × packaging condition to describe quality changes over storage time. Zero-order and first-order kinetic models were fitted to the experimental data. Model selection was based on goodness-of-fit criteria, including the coefficient of determination (R²) and root mean square error (RMSE). Models with higher R² and lower RMSE were considered superior. Residual plots were visually inspected to confirm the absence of systematic deviation. This approach ensured reliable model fitting for shelf-life estimation.

3. Results and Discussion

3.1. Effect of DMDC and Packaging on Physicochemical Properties of Mixed Mango and Passion Fruit Smoothie During Storage at 4 °C

Packaging materials exhibit varying barrier properties—such as resistance to oxygen and carbon dioxide permeation—which significantly affect product’s shelf-life, physicochemical stability, and sensory attributes [10]. Hence, the choice of packaging plays a pivotal role in maintaining juice quality during storage. Although PET is generally reported to have higher oxygen permeability than glass, oxygen transmission and light exposure were not directly measured or controlled in this study. Furthermore, treatment × packaging interactions were not statistically modeled. Therefore, interpretations regarding packaging-related differences are limited to observed within-week statistical comparisons and should not be construed as mechanistic evidence of oxygen- or light-driven effects. Table 1 shows physicochemical properties of mixed mango and passion fruit smoothie during storage at 4 °C.

Table 1.

Effect of DMDC and packaging on physicochemical properties of mixed mango and passion fruit smoothie during storage at 4 °C.

		Glass			PET
	Storage Period (Week)	Control	Pasteurization	DMDC 250 ppm	Control	Pasteurization	DMDC 250 ppm
pH ns	0	3.36 ± 0.01	3.37 ± 0.02	3.36 ± 0.02	3.39 ± 0.01	3.38 ± 0.01	3.36 ± 0.03
	1	3.30 ± 0.02	3.34 ± 0.01	3.34 ± 0.02	3.30 ± 0.01	3.35 ± 0.10	3.34 ± 0.01
	2	3.33 ± 0.01	3.34 ± 0.01	3.33 ± 0.01	3.30 ± 0.03	3.32 ± 0.01	3.29 ± 0.02
	3	3.30 ± 0.02	3.36 ± 0.04	3.36 ± 0.02	3.30 ± 0.03	3.46 ± 0.02	3.31 ± 0.01
	4	3.30 ± 0.02	3.36 ± 0.01	3.36 ± 0.03	3.30 ± 0.05	3.33 ± 0.06	3.31 ± 0.01
	5	3.37 ± 0.02	3.31 ± 0.05	3.30 ± 0.01	3.37 ± 0.02	3.34 ± 0.04	3.36 ± 0.01
	6	3.41 ± 0.02	3.44 ± 0.01	3.43 ± 0.02	3.43 ± 0.01	3.36 ± 0.03	3.43 ± 0.02
Total acid content ns (% citric acid )	0	1.45 ± 0.08	1.52 ± 0.07	1.45 ± 0.03	1.45 ± 0.02	1.45 ± 0.01	1.45 ± 0.04
	1	1.45 ± 0.03	1.56 ± 0.03	1.45 ± 0.04	1.55 ± 0.14	1.53 ± 0.05	1.45 ± 0.03
	2	1.55 ± 0.14	1.75 ± 0.03	1.55 ± 0.14	1.55 ± 0.14	1.75 ± 0.13	1.65 ± 0.14
	3	1.84 ± 0.14	1.95 ± 0.08	1.65 ± 0.14	1.84 ± 0.14	1.93 ± 0.12	1.65 ± 0.14
	4	1.94 ± 0.14	1.95 ± 0.02	1.75 ± 0.03	2.04 ± 0.13	1.93 ± 0.15	1.75 ± 0.11
	5	1.84 ± 0.14	1.75 ± 0.03	1.75 ± 0.01	2.04 ± 0.11	1.81 ± 0.09	1.75 ± 0.06
	6	1.65 ± 0.14	1.60 ± 0.01	1.75 ± 0.01	1.75 ± 0.08	1.65 ± 0.07	1.75 ± 0.08
Soluble solid content ns ( °Brix )	0	17.17 ± 0.24	16.17 ± 0.24	16.83 ± 0.85	16.83 ± 0.24	17.50 ± 0.41	16.17 ± 0.24
	1	17.00 ± 0.06	15.83 ± 0.24	16.33 ± 0.24	16.67 ± 0.47	17.33 ± 0.47	16.00 ± 0.82
	2	16.67 ± 0.47	15.83 ± 0.24	16.17 ± 0.24	16.67 ± 0.47	17.33 ± 0.47	16.00 ± 0.82
	3	16.00 ± 0.05	15.50 ± 0.01	16.00 ± 0.00	15.83 ± 0.24	15.50 ± 0.20	16.33 ± 0.24
	4	16.17 ± 0.24	15.00 ± 0.02	15.17 ± 0.24	15.67 ± 0.47	15.17 ± 0.24	15.83 ± 0.24
	5	15.17 ± 0.24	14.17 ± 0.24	15.17 ± 0.24	15.00 ± 0.20	15.17 ± 0.24	15.83 ± 0.24
	6	13.67 ± 0.24	13.17 ± 0.24	13.50 ± 0.41	13.17 ± 0.24	13.67 ± 0.47	14.00 ± 0.41
Turbidity (A660) ns	0	1.35 ± 0.08	1.29 ± 0.06	1.27 ± 0.08	1.30 ± 0.03	1.24 ± 0.03	1.24 ± 0.03
	1	1.18 ± 0.06	1.17 ± 0.02	1.16 ± 0.04	1.19 ± 0.02	1.17 ± 0.03	1.18 ± 0.02
	2	1.04 ± 0.01	1.07 ± 0.02	1.03 ± 0.02	1.04 ± 0.02	1.04 ± 0.01	1.08 ± 0.02
	3	0.86 ± 0.00	0.89 ± 0.03	0.94 ± 0.03	0.84 ± 0.01	0.86 ± 0.00	0.96 ± 0.04
	4	0.77 ± 0.03	0.79 ± 0.05	0.89 ± 0.01	0.76 ± 0.01	0.87 ± 0.02	0.87 ± 0.04
	5	0.67 ± 0.02	0.79 ± 0.01	0.79 ± 0.01	0.66 ± 0.01	0.72 ± 0.01	0.77 ± 0.01
	6	0.62 ± 0.05	0.65 ± 0.00	0.68 ± 0.00	0.58 ± 0.02	0.65 ± 0.03	0.68 ± 0.01
L*	0	50.09 ± 0.25 a	48.05 ± 0.15 b	50.20 ± 0.28 a	50.05 ± 0.25 a	48.35 ± 0.28 b	50.20 ± 0.20 a
	1	48.03 ± 0.51 a	47.64 ± 0.25 ab	48.51 ± 1.79 a	48.78 ± 0.26 a	47.92 ± 0.76 b	49.27 ± 0.20 a
	2	47.94 ± 0.53 a	47.61 ± 0.43 ab	46.02 ± 0.07 b	47.79 ± 0.27 a	47.43 ± 0.23 a	46.55 ± 0.20 b
	3	46.35 ± 0.16 b	47.40 ± 0.27 a	45.51 ± 0.03 c	46.81 ± 0.28 a	46.83 ± 0.10 a	44.38 ± 0.21 b
	4	45.85 ± 0.05 b	47.26 ± 0.08 a	44.90 ± 0.04 c	45.68 ± 0.29 b	46.06 ± 0.04 a	42.24 ± 0.03 c
	5	44.26 ± 0.12 b	46.78 ± 0.48 a	43.53 ± 0.05 c	43.04 ± 0.30 b	45.37 ± 0.37 a	41.40 ± 0.25 c
	6	41.85 ± 0.05 b	46.75 ± 0.39 a	41.58 ± 0.33 b	41.06 ± 0.31 b	44.65 ± 0.02 a	41.37 ± 0.30 b
a*	0	1.12 ± 0.12 ab	1.39 ± 0.05 a	1.24 ± 0.03 ab	1.41 ± 0.25 a	0.97 ± 0.05 b	1.37 ± 0.04 a
	1	1.08 ± 0.10 b	0.70 ± 0.08 c	1.61 ± 0.13 a	1.21 ± 0.15 b	1.71 ± 0.10 a	1.45 ± 0.14 ab
	2	1.04 ± 0.02 b	0.82 ± 0.05 c	1.30 ± 0.06 a	1.10 ± 0.04 b	1.66 ± 0.01 a	1.45 ± 0.14 ab
	3	1.08 ± 0.02 b	0.78 ± 0.03 c	1.88 ± 0.07 a	1.03 ± 0.02 b	1.66 ± 0.03 a	1.86 ± 0.01 a
	4	1.80 ± 0.21 a	1.26 ± 0.07 b	1.62 ± 0.02 ab	1.18 ± 0.04 b	1.56 ± 0.02 a	1.58 ± 0.05 a
	5	1.31 ± 0.43 ab	1.13 ± 0.06 b	1.35 ± 0.07 ab	1.15 ± 0.04 b	1.19 ± 0.03 b	1.19 ± 0.03 b
	6	1.73 ± 0.01 a	1.19 ± 0.09 b	1.20 ± 0.04 b	1.38 ± 0.03 a	1.18 ± 0.01 b	1.19 ± 0.06 b
b*	0	29.76 ± 0.45 a	29.37 ± 0.22 ab	29.40 ± 0.58 ab	29.39 ± 0.60 a	28.93 ± 0.44 b	29.85 ± 0.04 a
	1	29.25 ± 0.20 a	29.02 ± 0.24 a	29.94 ± 1.79 a	29.35 ± 0.23 a	28.84 ± 0.37 a	28.58 ± 0.69 a
	2	28.54 ± 0.37 ab	29.62 ± 0.13 a	28.40 ± 0.14 b	28.50 ± 0.29 ab	28.61 ± 0.05 a	28.46 ± 0.37 b
	3	29.25 ± 0.20 a	28.27 ± 0.75 ab	27.01 ± 0.13 b	28.37 ± 0.66 a	27.60 ± 0.17 ab	27.30 ± 0.21 b
	4	27.94 ± 0.13 a	27.91 ± 0.17 a	27.89 ± 0.40 a	27.05 ± 0.66 ab	26.56 ± 0.22 b	26.50 ± 0.12 b
	5	26.43 ± 0.48 b	27.84 ± 0.78 a	26.26 ± 0.39 b	26.78 ± 0.01 a	25.20 ± 0.01 b	26.41 ± 0.47 ab
	6	26.94 ± 0.13 a	27.99 ± 0.02 a	25.74 ± 0.30 b	26.78 ± 0.01 a	25.20 ± 0.01 b	25.14 ± 0.01 a
∆ E	0	0 ± 0.00	0 ± 0.00	0 ± 0.00	0 ± 0.00	0 ± 0.00	0 ± 0.00
	1	2.20 ± 0.70 b	0.98 ± 0.11 c	3.17 ± 1.35 a	1.46 ± 0.10 b	1.25 ± 0.48 c	1.69 ± 0.55 a
	2	2.50 ± 0.80 b	0.81 ± 0.19 c	4.29 ± 0.36 a	2.56 ± 0.46 b	1.33 ± 0.35 c	3.93 ± 0.05 a
	3	3.82 ± 0.21 b	1.52 ± 0.32 c	5.33 ± 0.34 a	3.58 ± 0.66 b	2.15 ± 0.38 c	6.37 ± 0.12 a
	4	4.69 ± 0.38 b	1.68 ± 0.11 c	5.53 ± 0.30 a	4.99 ± 0.55 b	3.36 ± 0.52 c	8.64 ± 0.11 a
	5	6.73 ± 0.22 b	2.25 ± 0.36 c	7.41 ± 0.18 a	7.5 ± 0.49 b	4.82 ± 0.49 c	9.45 ± 0.43 a
	6	8.74 ± 0.37 b	1.94 ± 0.44 c	9.4 ± 0.37 a	9.38 ± 0.41 b	5.27 ± 0.42 c	10.01 ± 0.23 a
PPO (%)	0	100.22 ± 25 a	44.05 ± 20 c	96.06 ± 50 b	100.37 ± 10 a	43.25 ± 30 c	93.44 ± 30 b
	1	86.01 ± 40 a	47.22 ± 30 c	65.33 ± 30 b	99.45 ± 70 a	42.44 ± 10 c	84.88 ± 20 b
	2	79.04 ± 20 a	40.14 ± 40 c	64.94 ± 60 b	72.56 ± 20 a	46.17 ± 20 c	70.15 ± 10 b
	3	70.21 ± 40 a	35.18 ± 20 c	66.11 ± 30 b	67.76 ± 30 a	48.56 ± 20 c	66.34 ± 20 b
	4	63.30 ± 60 a	47.27 ± 20 c	50.25 ± 20 b	54.33 ± 10 a	48.19 ± 30 c	60.22 ± 60 b
	5	51.21 ± 30 a	49.33 ± 70 c	50.12 ± 40 b	53.21 ± 10 a	51.47 ± 60 c	60.15 ± 80 b
	6	66.11 ± 40 a	53.02 ± 10 c	59.66 ± 40 b	61.23 ± 40 a	56.12 ± 40 c	66.39 ± 70 b

Values are expressed as mean ± SD of three independent biological replicates (n = 3). For each biological replicate, analytical measurements were performed in duplicate and averaged prior to statistical analysis. Statistical comparisons were performed separately at each storage week using one-way ANOVA among treatment × packaging combinations. Different superscript letters (a–c) within the same row indicate significant differences among groups at that storage time according to Tukey’s HSD test (p < 0.05). “ns” indicates that no significant differences were detected among treatment × packaging combinations at that storage week (p ≥ 0.05).

pH remained relatively stable throughout the six-week storage period across all treatments, with values starting around 3.35 and showing no statistically significant changes (p > 0.05). The packaging type (glass vs. PET) had no significant impact on pH levels. This stability reflects the acidic nature of the mango–passion fruit blend, which inherently limits microbial activity. These findings are consistent with those of Yu et al. [4], who reported stable pH in orange–carrot–pomegranate blends treated with DMDC and stored at 4 °C. Yu et al. [4] also observed no significant pH change in DMDC- and HPH-treated mulberry juice during 30 days of cold storage.

Total acid showed an increasing trend during the first four weeks, followed by stabilization or slight decline, with no significant differences between packaging types (p > 0.05). The greater acid stability in DMDC-treated samples suggests reduced microbial degradation of organic acids. Roobab et al. [17] reported similar trends in mixed fruit juices treated with chemical preservatives where acid degradation was suppressed due to microbial inhibition. Castillejo et al. [18] also noted early increases in total acidity in refrigerated multi-fruit smoothies, attributed to ongoing enzymatic reactions before stabilization.

Total soluble solids (TSS) values decreased progressively from 16.17 to 17.17 °Brix to 13.17–14.00 °Brix. The largest decrease occurred in untreated controls, while DMDC-treated and pasteurized samples retained higher TSS levels. The decline in TSS may be partially associated with microbial activity, as suggested by its inverse correlation with total viable counts (r = −0.82). However, in a pulp-rich smoothie matrix, changes in °Brix may also arise from sedimentation or phase separation, as well as refractive index alterations due to enzymatic hydrolysis of polysaccharides. Therefore, the observed reduction in TSS cannot be attributed solely to microbial sugar consumption. Similar patterns were documented by Salazar-González et al. [19] in guava nectar, and by Mgaya-Kilima et al. [3] in roselle–mango juice stored under refrigeration. Unluturk and Atilgan [20] also observed a TSS drop in white grape juice stored at 4 °C, attributing it to microbial activity.

Turbidity also declined over time by week 6. DMDC-treated samples retained higher turbidity, possibly due to inhibition of microbial enzymes that break down insoluble particles. These findings are supported by Janzantti et al. [21], who reported better turbidity retention in thermally and chemically stabilized passion fruit beverages. No significant differences in turbidity were observed between glass and PET packaging (p > 0.05), although DMDC-treated samples retained slightly higher turbidity overall.

In terms of color, both L* and b* values decreased during storage, with significant differences (p ≤ 0.05) between packaging types. This reduction is likely due to non-enzymatic browning and pigment degradation over time. DMDC-treated PET samples showed the most pronounced L* loss between weeks 3 and 5. Polyphenol oxidase (PPO) activity, responsible for enzymatic browning and phenolic degradation, decreased in all treatments over the six-week period. Pasteurization resulted in the most significant initial PPO reduction (up to 44%), while DMDC-treated samples showed more gradual inactivation. Although PET is generally characterized by higher oxygen permeability than glass [22], PPO activity was not consistently higher in PET across treatments and storage times, and packaging was not identified as a significant main effect. Therefore, differences in PPO activity cannot be attributed solely to packaging material under the conditions of this study.

3.2. Effect of DMDC and Packaging on Functional Properties of Mixed Mango and Passion Fruit Smoothie During Storage at 4 °C

The results presented in Table 2 provide insights into the functional properties of mixed mango and passion fruit smoothies, specifically total phenolic content (TPC), total flavonoid content (TFC), and antioxidant activity (measured via DPPH and FRAP assays), under different treatments (control, pasteurization, and DMDC at 250 ppm) and packaging types (glass and PET) during a six-week storage period at 4 °C. The study found that TPC decreased over the six-week storage period, with initial values of 126.93 ± 7.72 mg GAE/100 g dropping to 80.89 ± 0.71 mg GAE/100 g by the end of storage. However, DMDC-treated samples (250 ppm) exhibited the highest phenolic retention compared to control and pasteurized samples, regardless of packaging type. Similarly, TFC declined over time, with DMDC-treated samples again showing better retention than the control and pasteurized counterparts. Notably, the type of packaging (glass vs. PET) did not significantly influence either TPC or TFC (p > 0.05). The superior retention of TPC and TFC in DMDC-treated samples can be attributed to DMDC’s antimicrobial action, which likely minimized microbial degradation of phenolic and flavonoid compounds and also the inhibition of residual endogenous PPO activity [23]. Additionally, DMDC’s non-thermal nature avoids the heat-induced degradation of bioactive compounds often observed with pasteurization, which can cause Maillard reactions or the thermal breakdown of phenolics [4]. This is supported by the study’s observation that pasteurized samples exhibited greater phenolic loss, likely due to heat exposure during processing (90 °C for 100 s).

Table 2.

Effect of DMDC and packaging on functional properties of mixed mango and passion fruit smoothie during storage at 4 °C.

			Glass			PET
	Storage Period (Week)	Control	Pasteurization	DMDC 250 ppm	Control	Pasteurization	DMDC 250 ppm
Total phenolic compound ns	0	123.6 ± 4.55	126.99 ± 8.29	126.93 ± 7.72	124.78 ± 11.63	123.48 ± 6.17	123.73 ± 3.10
	1	116.99 ± 2.27	116.52 ± 3.59	124.12 ± 1.72	114.01 ± 10.65	113.23 ± 2.21	122.43 ± 1.68
	2	105.17 ± 2.11	104.56 ± 2.16	110.32 ± 2.74	105.35 ± 1.17	100.72 ± 0.53	104.33 ± 3.30
	3	105.69 ± 2.48	101.48 ± 1.88	107.61 ± 0.71	103.94 ± 0.84	100.73 ± 1.67	107.07 ± 0.33
	4	96.67 ± 1.05	99.51 ± 0.55	105.67 ± 3.05	94.67 ± 1.08	94.63 ± 0.05	101.07 ± 0.87
	5	94.12 ± 2.12	99.02 ± 0.56	96.20 ± 1.56	90.77 ± 0.10	80.47 ± 0.27	96.63 ± 0.26
	6	86.33 ± 0.24	80.89 ± 0.71	92.40 ± 0.73	83.57 ± 0.69	82.20 ± 0.43	91.25 ± 1.55
Flavonoids ns	0	22.95 ± 0.37	18.98 ± 0.54	21.99 ± 0.50	22.71 ± 1.19	19.07 ± 0.92	22.71 ± 1.03
	1	18.95 ± 0.51	16.47 ± 0.03	18.70 ± 0.23	19.25 ± 1.20	16.52 ± 0.33	19.51 ± 0.66
	2	15.93 ± 1.01	13.14 ± 0.56	16.10 ± 0.69	15.25 ± 1.20	12.44 ± 0.31	16.17 ± 0.50
	3	12.34 ± 0.32	11.78 ± 0.38	15.43 ± 0.65	13.69 ± 0.13	11.65 ± 0.97	14.84 ± 0.96
	4	11.48 ± 0.79	11.40 ± 0.21	12.22 ± 0.21	12.34 ± 0.21	11.48 ± 0.19	11.30 ± 0.81
	5	9.15 ± 0.50	9.40 ± 0.21	9.89 ± 0.38	8.92 ± 0.66	8.57 ± 0.15	9.30 ± 0.81
	6	8.85 ± 0.33	8.84 ± 0.12	8.36 ± 0.25	8.89 ± 0.33	8.61 ± 0.13	8.99 ± 0.26
DPPH ns	0	200.24 ± 13.83	187.60 ± 4.14	207.92 ± 0.89	206.79 ± 1.38	186.50 ± 3.32	200.24 ± 3.16
	1	176.05 ± 1.20	184.31 ± 1.88	203.25 ± 6.78	173.90 ± 0.67	174.19 ± 2.30	197.20 ± 2.02
	2	133.06 ± 1.68	179.30 ± 1.92	193.25 ± 6.78	135.64 ± 5.23	173.33 ± 2.95	197.20 ± 2.02
	3	103.41 ± 2.85	165.37 ± 1.03	185.75 ± 3.65	112.70 ± 5.46	167.13 ± 2.51	188.39 ± 1.25
	4	104.09 ± 0.93	161.30 ± 2.32	183.68 ± 4.79	102.62 ± 1.39	155.65 ± 3.39	187.50 ± 0.88
	5	105.84 ± 3.71	152.80 ± 4.46	168.99 ± 1.65	104.08 ± 0.77	145.65 ± 3.39	167.50 ± 0.88
	6	90.51 ± 1.68	145.92 ± 2.38	145.92 ± 2.06	89.49 ± 1.16	145.73 ± 1.94	149.21 ± 2.56
FRAP ns	0	163.66 ± 0.91	150.19 ± 1.70	168.52 ± 6.76	163.10 ± 3.25	145.13 ± 5.99	165.30 ± 1.98
	1	155.19 ± 1.77	147.92 ± 2.46	163.24 ± 1.91	155.97 ± 7.21	146.57 ± 1.51	166.47 ± 1.86
	2	150.14 ± 0.81	146.53 ± 0.39	165.05 ± 2.25	157.87 ± 2.11	147.50 ± 0.80	162.20 ± 0.22
	3	148.87 ± 2.01	148.17 ± 1.32	158.13 ± 1.58	148.83 ± 6.04	148.77 ± 0.25	160.00 ± 1.30
	4	148.56 ± 6.88	145.53 ± 3.07	156.11 ± 0.57	152.47 ± 1.57	151.07 ± 0.29	155.70 ± 0.86
	5	145.21 ± 5.26	148.77 ± 5.83	152.13 ± 3.88	145.83 ± 2.21	149.57 ± 0.62	151.17 ± 0.33
	6	138.52 ± 1.43	132.20 ± 3.45	144.26 ± 2.36	138.87 ± 6.81	138.70 ± 5.40	149.07 ± 0.66

Values are expressed as mean ± SD of three independent biological replicates (n = 3). For each biological replicate, analytical measurements were performed in duplicate and averaged prior to statistical analysis. Statistical comparisons were performed separately at each storage week using one-way ANOVA among treatment × packaging combinations. “ns” indicates that no significant differences were detected among treatment × packaging combinations at that storage week (p ≥ 0.05). Total phenolic compound: mg GAE/100 g = mg gallic acid equivalent/100 g. Flavonoid: mg QCE/100 g = mg quercetin equivalent/100 g. DPPH and FRAP: mM trolox/100 g = mmol trolox/100 g.

The lack of significant packaging effects on TPC and TFC suggests that neither glass nor PET significantly influenced oxidative degradation pathways for these compounds under the storage conditions. Glass is impermeable to oxygen, while PET has higher oxygen permeability, which could theoretically accelerate oxidation [10]. However, the low storage temperature likely slowed oxidative reactions, minimizing differences between the two packaging types. This finding contrasts with the study’s observation of greater color deviation (ΔE) in PET bottles, suggesting that while packaging impacts visual quality, its effect on bioactive compounds is less pronounced under these conditions. Comparing DMDC with other non-thermal preservation methods, high-pressure processing (HPP) and pulsed electric fields (PEF) have also been shown to preserve phenolic content in fruit juices better than thermal treatments. For instance, HPP can retain up to 90% of phenolic content in mango juice after 60 days at 4 °C [24], and high-power ultrasound (HPU) and PEF in the hurdle concept could preserve antioxidant bioactive compounds in strawberry juice [25], which is comparable to the bioactive retention observed with DMDC in this study. However, DMDC offers a cost-effective and simpler alternative to HPP and PEF, which require specialized equipment and higher operational costs. Future studies should explore the synergistic effects of DMDC with HPP or PEF to further enhance phenolic retention while maintaining microbial quality.

Antioxidant activity, as measured by DPPH and FRAP assays, exhibited a declining trend across all samples over the six-week storage period (Table 2). However, DMDC-treated samples consistently maintained higher DPPH and FRAP values compared to control and pasteurized samples by week 6. Packaging type did not significantly affect antioxidant activity (p > 0.05), consistent with findings by Kim et al. [26] for green tea and Mgaya-Kilima et al. [3] for roselle-mango juice, where antioxidant degradation occurred regardless of packaging material during cold storage. The better preservation of antioxidant activity in DMDC-treated samples can be linked to the higher retention of phenolic and flavonoid compounds, which are primary contributors to the antioxidant capacity of fruit smoothies [3]. Phenolics and flavonoids scavenge free radicals (DPPH assay) and reduce ferric ions (FRAP assay), and their degradation—whether through microbial activity, enzymatic reactions, or oxidation—directly reduces antioxidant activity. DMDC’s ability to suppress microbial growth (e.g., <5 log CFU/mL for total microorganisms, as shown in Figure 1) likely reduced the activity of oxidative enzymes like PPO, which was also lower in DMDC-treated samples compared to controls (Table 1). Furthermore, the non-thermal nature of DMDC minimized the formation of reactive oxygen species (ROS) that can occur during thermal pasteurization, preserving the smoothie’s antioxidant potential. Pasteurized samples, while initially effective in reducing microbial load, showed greater antioxidant loss over time, possibly due to thermal degradation of heat-sensitive antioxidants like ascorbic acid, which is abundant in passion fruit [2]. Ascorbic acid degradation can also generate ROS, further accelerating antioxidant loss [4]. In contrast, DMDC’s mechanism of action—inactivating microorganisms by alkylating nucleophilic groups in microbial enzymes—does not involve heat, thus avoiding such secondary oxidative effects. The lack of packaging effect on antioxidant activity aligns with the TPC and TFC results, suggesting that oxygen permeability differences between glass and PET were not a dominant factor in antioxidant degradation at 4 °C [10].

Figure 1.

Effect of DMDC and packaging on total microorganism of mixed mango and passion fruit smoothie during storage at 4 °C.

Comparing DMDC with other non-thermal methods, ultrasound treatment has been shown to preserve antioxidant activity in fruit juices by inactivating enzymes like PPO while minimizing thermal damage [27]. However, ultrasound can sometimes degrade ascorbic acid due to cavitation-induced ROS formation, which may offset its benefits. DMDC’s advantage lies in its targeted antimicrobial action without affecting the smoothie’s redox environment, making it a promising alternative for antioxidant preservation.

3.3. Effect of DMDC and Packaging on Total Viable Count and Yeast and Mold Counts of Mixed Mango and Passion Fruit Smoothie During Storage at 4 °C

Figure 1 and Figure 2 illustrate the total viable count (TVC) and yeast and mold count of the smoothies over six weeks of storage at 4 °C. The TVC (Figure 1) increased in all samples over time, with control samples in both glass and PET reaching the highest levels by week 6 (7.94 log CFU/mL in glass and 7.84 log CFU/mL in PET). Pasteurized samples showed a moderate increase, reaching 5.83 log CFU/mL in glass and 5.62 log CFU/mL in PET by week 6, while DMDC-treated samples (250 ppm) maintained the lowest microbial load, with counts below 5 log CFU/mL (e.g., 4.19 log CFU/mL in glass and 3.71 log CFU/mL in PET at week 6). Similarly, yeast and mold counts (Figure 2) followed a comparable trend, with control samples reaching 6.64 log CFU/mL in both glass and PET by week 6, pasteurized samples showing counts of 5.24 log CFU/mL (glass) and 5.02 log CFU/mL (PET), and DMDC-treated samples maintaining counts below 3 log CFU/mL (e.g., 2.17 log CFU/mL in glass and 2.73 log CFU/mL in PET at week 6). Notably, DMDC-treated samples in glass bottles showed no detectable yeast or mold growth until the third week, highlighting the treatment’s strong initial antimicrobial efficacy.

Figure 2.

Effect of DMDC and packaging on yeast and mold count of mixed mango and passion fruit smoothie during storage at 4 °C.

The superior microbial control by DMDC can be attributed to its mechanism of action, which involves the alkylation of nucleophilic groups (e.g., amino and sulfhydryl groups) in microbial enzymes and proteins, leading to rapid inactivation of bacteria, yeasts, and molds [7]. This non-thermal approach contrasts with pasteurization (90 °C for 100 s in this study), which initially reduces microbial load but allows for regrowth over time due to the survival of heat-resistant spores or post-process contamination [4]. The effectiveness of DMDC aligns with findings by Zuehlke et al. [28], who reported significant reductions in yeast populations in grape juice and wine with 200 mg/L DMDC, and Yu et al. [4], who demonstrated that 250 ppm DMDC effectively controlled total microorganisms, yeasts, and molds in mulberry leaf juice for 60 days at 4 °C. These studies support DMDC’s potential as a viable alternative to thermal processing for ensuring microbial quality in fruit-based beverages.

The lack of significant differences (p > 0.05) between glass and PET packaging in terms of microbial growth suggests that packaging material had a limited role in microbial proliferation under the storage conditions at 4 °C. Glass is impermeable to oxygen, while PET has higher oxygen permeability, which could theoretically support aerobic microbial growth [10]. However, the low storage temperature likely slowed microbial metabolism, minimizing the impact of oxygen availability.

Pasteurization, while effective in reducing initial microbial load, showed a notable increase in TVC and yeast/mold counts over time, likely due to the survival of heat-resistant microorganisms or post-process contamination. High-pressure processing (HPP) and pulsed electric fields (PEF) are non-thermal alternatives that have been widely studied for juice preservation. HPP can reduce microbial counts to below 2 log CFU/mL in fruit juices, with effects lasting up to 60 days at 4 °C [24], which is comparable to DMDC’s performance in this study. However, HPP requires significant capital investment and may not be as effective against yeast and mold spores as DMDC. PEF, on the other hand, can achieve a 5-log reduction in microbial counts but often requires combination with mild heat or other hurdles to prevent regrowth, making DMDC a simpler and more cost-effective option [29].

Potassium metabisulfite is another chemical preservative that can control microbial growth in juices. However, it may impart a sulfur-like off-flavor, which could affect sensory quality, and its use is restricted in some markets due to allergen concerns [30]. DMDC, while producing methanol and carbon dioxide upon hydrolysis, generates these byproducts at levels below safety thresholds [5], making it a safer and more consumer-friendly option. Additionally, DMDC’s broad-spectrum antimicrobial activity, as demonstrated by its effectiveness against both TVC and yeast/mold, positions it as a versatile preservative for complex matrices like smoothies, where diverse microbial populations (e.g., bacteria from mango pulp, yeasts from passion fruit) may be present.

This study focused on total viable count and yeast/molds as general hygiene and spoilage indicators. Future work should specifically enumerate acid-tolerant spoilage bacteria (e.g., Lactobacillus, Acetobacter) and relevant pathogens (e.g., Escherichia coli O157:H7, Salmonella) to fully validate process efficacy.

3.4. Kinetic Modeling of Changes in Properties of Smoothie over Storage Period

Kinetic modeling is widely used in food science to describe quality changes during storage, including microbial growth, color variation, and degradation of antioxidant compounds. In the present study, changes in microbiological (total viable count and yeast/mold count), physical (total color difference, ΔE), and antioxidant-related parameters (DPPH, FRAP, total phenolic content (TPC), and total flavonoid content (TFC)) of mango–passion fruit smoothie were evaluated under refrigerated storage. Experimental data were fitted to zero-order and first-order kinetic models using regression analysis, and the estimated kinetic parameters are presented in Table 3.

Table 3.

Statistical parameters of the kinetic models used for describing the changes in the properties of stored smoothie.

Model	Parameters	ΔE	DPPH	FRAP	TPC	TFC	Total Plate Count	Mold and Yeast
Glass-Control
Zero-order Model	Co (CI)	0.0825 (−0.9418–1.1068)	183.8739 (151.5193–216.2285)	161.5993 (157.1618–166.0368)	121.8725 (117.1781–126.5669)	21.3446 (18.8842–23.8051)	3.1436 (2.7341–3.5530)	2.9661 (2.5695–3.3627)
	k (CI)	−0.1912 (−0.2318–0.1506)	2.5437 (1.2618–3.8257)	0.5287 (0.3529–0.7045)	0.8472 (0.6612–1.0332)	0.3385 (0.2410–0.4360)	−0.1005 (−0.1167–0.0843)	−0.0794 (−0.0952–0.0637)
	R2	0.967	0.8388	0.9228	0.9648	0.941	0.9807	0.9712
	RMSE	0.4943	15.6116	2.1412	2.2651	1.1872	0.1976	0.1914
First-order Model	Co (CI)	1.2862 (0.6398–1.9327)	194.3266 (165.9950–222.6582)	161.8731 (157.4291–166.3171)	122.7905 (118.1635–127.4175)	22.6297 (21.2611–23.9982)	3.3872 (3.0821–3.6923)	3.1207 (2.8769–3.3646)
	k (CI)	−0.0463 (−0.0602–0.0323)	0.0212 (0.0132–0.0291)	0.0035 (0.0024–0.0047)	0.0082 (0.0065–0.0099)	0.0250 (0.0215–0.0286)	−0.0192 (−0.0221–0.0163)	−0.0174 (−0.0200–0.0149)
	R2	0.959	0.9069	0.9267	0.9698	0.9868	0.9845	0.9849
	RMSE	0.5509	11.8617	2.0859	2.0995	0.5614	0.1771	0.1388
Glass-Pasteurization
Zero-order Model	Co	0.3225 (−0.2280–0.8730)	190.1636 (185.7509–194.5762)	153.9679 (149.5594–158.3763)	123.2475 (114.1520–132.3430)	17.8193 (15.9932–19.6454)	1.4321 (0.6982–2.1661)	0.9693 (−0.3310–2.2696)
	k	−0.0471 (−0.0689–0.0253)	1.0513 (0.8765–1.2262)	0.3923 (0.2177–0.5670)	0.9099 (0.5496–1.2703)	0.2362 (0.1639–0.3086)	−0.1017 (−0.1308–0.0727)	−0.0734 (−0.1249–0.0218)
	R2	0.8603	0.9795	0.8696	0.8939	0.9337	0.9418	0.7283
	RMSE	0.2656	2.1292	2.1272	4.3887	0.8811	0.3542	0.6274
First-order Model	Co	0.6346 (0.1071–1.1620)	190.8371 (185.6981–195.9762)	154.0221 (149.3889–158.6553)	124.3105 (114.8245–133.7965)	18.5381 (17.1730–19.9032)	1.7688 (1.3977–2.1399)	1.0290 (0.4264–1.6315)
	k	−0.0310 (−0.0557–0.0062)	0.0062 (0.0051–0.0074)	0.0027 (0.0014–0.0039)	0.0088 (0.0054–0.0122)	0.0192 (0.0153–0.0231)	−0.0295 (−0.0358–0.0231)	−0.0359 (−0.0528–0.0189)
	R2	0.7508	0.9747	0.8619	0.8988	0.9715	0.9722	0.8696
	RMSE	0.3548	2.3645	2.189	4.2863	0.578	0.2446	0.4347
Glass-DMDC 250 ppm
Zero-order Model	Co	0.9557 (−0.4132–2.3246)	212.4086 (200.9007–223.9165)	169.3421 (165.1583–173.5260)	126.6157 (121.4944–131.7370)	21.3546 (20.2210–22.4883)	0.3350 (−0.3145–0.9845)	−0.6068 (−1.7111–0.4975)
	k	−0.1935 (−0.2477–0.1392)	1.3476 (0.8916–1.8035)	0.5303 (0.3645–0.6961)	0.8371 (0.6342–1.0401)	0.3183 (0.2734–0.3632)	−0.0734 (−0.0991–0.0476)	−0.0776 (−0.1214–0.0338)
	R2	0.9439	0.9203	0.9312	0.9574	0.9852	0.9148	0.8061
	RMSE	0.6605	5.5527	2.0188	2.4711	0.547	0.3134	0.5328
First-order Model	Co	2.0883 (0.8047–3.3719)	212.9563 (199.3903–226.5222)	169.4715 (164.9631–173.9799)	127.4533 (122.2676–132.6390)	22.1034 (20.6277–23.5790)	0.6769 (0.3860–0.9677)	0.1665 (−0.2174–0.5504)
	k	−0.0363 (−0.0541–0.0185)	0.0072 (0.0044–0.0100)	0.0033 (0.0022–0.0044)	0.0077 (0.0059–0.0095)	0.0216 (0.0180–0.0253)	−0.0410 (−0.0532–0.0289)	−0.0698 (−0.1299–0.0098)
	R2	0.8882	0.9006	0.9241	0.9611	0.9812	0.9552	0.8134
	RMSE	0.9323	6.2	2.1194	2.3609	0.6162	0.2272	0.5228
PET-Control
Zero-order Model	Co	−0.3596 (−1.3512–0.6319)	188.3771 (159.5191–217.2352)	162.3882 (156.9382–167.8382)	121.8118 (117.5953–126.0283)	21.4032 (19.2726–23.5338)	2.9568 (2.5239–3.3896)	2.8414 (2.5326–3.1503)
	k	−0.2176 (−0.2569–0.1783)	2.6763 (1.5329–3.8197)	0.5019 (0.2860–0.7178)	0.9224 (0.7553–1.0895)	0.3318 (0.2474–0.4162)	−0.0993 (−0.1165–0.0822)	−0.0857 (−0.0980–0.0735)
	R2	0.9759	0.8787	0.8771	0.9758	0.9533	0.9779	0.9848
	RMSE	0.4784	13.9245	2.6297	2.0345	1.028	0.2089	0.149
First-order Model	Co	1.1601 (0.5719–1.7483)	199.2252 (176.6652–221.7851)	162.5446 (156.8696–168.2197)	122.9301 (119.0439–126.8164)	22.5067 (21.0360–23.9774)	3.1958 (2.9155–3.4760)	3.0475 (2.7773–3.3177)
	k	−0.0509 (−0.0647–0.0371)	0.0219 (0.0156–0.0281)	0.0033 (0.0018–0.0047)	0.0091 (0.0076–0.0105)	0.0238 (0.0201–0.0275)	−0.0199 (−0.0227–0.0171)	−0.0185 (−0.0214–0.0156)
	R2	0.9689	0.9446	0.8735	0.982	0.9837	0.9864	0.9835
	RMSE	0.5442	9.4096	2.6685	1.7527	0.6072	0.164	0.1556
PET-Pasteurization
Zero-order Model	Co	−0.0793 (−0.7563–0.5977)	180.4975 (173.2112–187.7838)	151.9218 (148.8103–155.0333)	120.2925 (112.9762–127.6088)	17.7886 (15.5106–20.0666)	1.0793 (−0.0772–2.2357)	1.0161 (0.0123–2.0198)
	k	−0.1274 (−0.1543–0.1006)	0.8524 (0.5637–1.1411)	0.2459 (0.1226–0.3691)	0.9972 (0.7073–1.2871)	0.2461 (0.1559–0.3364)	−0.0974 (−0.1433–0.0516)	−0.0758 (−0.1155–0.0360)
	R2	0.9676	0.9201	0.8402	0.9399	0.9076	0.8567	0.8275
	RMSE	0.3267	3.5157	1.5013	3.5302	1.0992	0.558	0.4843
First-order Model	Co	0.8145 (0.2673–1.3618)	180.8315 (172.7879–188.8751)	151.9374 (148.7299–155.1448)	121.7127 (114.5154–128.9101)	18.5950 (16.6432–20.5468)	1.3428 (0.8499–1.8358)	1.1478 (0.7349–1.5607)
	k	−0.0466 (−0.0652–0.0280)	0.0052 (0.0033–0.0071)	0.0017 (0.0008–0.0025)	0.0101 (0.0074–0.0129)	0.0204 (0.0148–0.0261)	−0.0346 (−0.0453–0.0238)	−0.0335 (−0.0441–0.0229)
	R2	0.9331	0.9102	0.8346	0.9499	0.9485	0.944	0.9391
	RMSE	0.4692	3.7291	1.5273	3.2223	0.8204	0.3488	0.2878
PET-DMDC 250 ppm
Zero-order Model	Co	0.3421 (−1.0618–1.7461)	207.6975 (192.7713–222.6237)	167.7504 (164.8919–170.6088)	122.9657 (115.6783–130.2531)	21.8082 (20.0393–23.5772)	0.3618 (−0.2755–0.9991)	−0.3632 (−1.1859–0.4595)
	k	−0.2564 (−0.3121–0.2008)	1.1336 (0.5422–1.7250)	0.4377 (0.3244–0.5510)	0.7773 (0.4886–1.0661)	0.3390 (0.2689–0.4091)	−0.0735 (−0.0988–0.0483)	−0.0714 (−0.1040–0.0388)
	R2	0.9656	0.8292	0.9518	0.9055	0.9687	0.9181	0.8637
	RMSE	0.6774	7.2021	1.3792	3.5163	0.8535	0.3075	0.397
First-order Model	Co	2.2976 (0.3496–4.2456)	207.7619 (190.9923–224.5315)	167.8478 (164.7800–170.9156)	123.7854 (116.3962–131.1745)	22.8381 (21.5600–24.1161)	0.7023 (0.4079–0.9967)	0.2300 (−0.0909–0.5510)
	k	−0.0382 (−0.0625–0.0139)	0.0060 (0.0025–0.0094)	0.0027 (0.0020–0.0035)	0.0074 (0.0047–0.0100)	0.0236 (0.0204–0.0268)	−0.0403 (−0.0522–0.0284)	−0.0616 (−0.0986–0.0247)
	R2	0.8412	0.8033	0.9468	0.913	0.988	0.9552	0.8841
	RMSE	1.4561	7.7299	1.4486	3.3725	0.5282	0.2274	0.3661

Model fitting revealed parameter-dependent behavior across treatments (Control, DMDC, and Pasteurization) and packaging conditions (glass and PET). For microbial counts, first-order growth kinetics in logarithmic form generally provided an adequate description of the data. Control samples exhibited the fastest microbial growth rates, whereas both DMDC and pasteurization treatments significantly reduced growth rates, particularly during early storage.

For color change (ΔE), first-order kinetics more consistently described the progressive increase observed during storage, indicating that color degradation was proportional to the remaining quality state. Pasteurized samples generally showed slower color change compared to untreated controls, while DMDC-treated samples exhibited intermediate behavior.

Antioxidant-related parameters (DPPH, FRAP, TPC, and TFC) declined over time across all treatments. In most cases, first-order kinetics provided a better fit based on combined evaluation of R², RMSE, and residual distribution, suggesting that degradation of bioactive compounds was concentration-dependent. Pasteurization tended to slow antioxidant degradation relative to control samples, likely due to partial inactivation of oxidative enzymes. DMDC treatment effectively controlled microbial growth but did not consistently provide superior antioxidant retention compared to thermal treatment.

Packaging effects were parameter-specific and, where significant, occurred primarily through interaction with treatment and storage time rather than as a consistent independent main effect. Previous studies have indicated that both zero-order and first-order models effectively describe quality changes in various products over cold storage periods, including date-based energy drink [13], coconut water [31], carrot juice [32], sweet lime juice [29] and pomegranate juice [33]. The present findings support the applicability of kinetic modeling for predicting quality changes in mixed-fruit smoothies during cold storage.

3.5. Prediction of Shelf-Life of Smoothie

The predicted shelf-life of the fruit-based beverages varied considerably depending on the applied treatment and packaging material, with microbial quality serving as the limiting factor across all quality attributes (Table 4). Table 4 provides a detailed overview of the predicted shelf-life of the mixed mango and passion fruit smoothie, based on various quality and microbiological parameters, under different treatments (control, pasteurization, and DMDC at 250 ppm) and packaging types (glass and PET) during storage at 4 °C. The shelf-life was estimated using kinetic modeling, incorporating the reaction rate constants from the selected kinetic models (zero-order; first-order) into the respective shelf-life equations and considering key quality attributes such as color stability (ΔE), antioxidant activity (DPPH and FRAP), total phenolic content (TPC), total flavonoid content (TFC), total plate count (TVC), and mold and yeast counts. The microbial shelf-life threshold of 5 log CFU/mL (equivalent to 10⁵ CFU/mL) was used as a benchmark.

Table 4.

Predicted shelf-life of smoothie by integration of quality properties.

Property	Glass			PET
	Control	Pasteurization	DMDC 250 ppm	Control	Pasteurization	DMDC 250 ppm
ΔE	15.2 (13.7–16.7)	25.0 (22.5–27.5)	20.0 (18.0–22.0)	22.0 (19.8–24.2)	30.0 (27.0–33.0)	39.0 (35.1–42.9)
DPPH	18.2 (16.4–20.0)	28.0 (25.2–30.8)	36.0 (32.4–39.6)	20.0 (18.0–22.0)	32.0 (28.8–35.2)	35.0 (31.5–38.5)
FRAP	16.8 (15.1–18.5)	27.0 (24.3–29.7)	38.0 (34.2–41.8)	19.0 (17.1–20.9)	33.0 (29.7–36.3)	36.0 (32.4–39.6)
TPC	19.0 (17.1–20.9)	30.0 (27.0–33.0)	40.0 (36.0–44.0)	21.0 (18.9–23.1)	35.0 (31.5–38.5)	38.0 (34.2–41.8)
TFC	20.0 (18.0–22.0)	31.0 (27.9–34.1)	39.0 (35.1–42.9)	22.0 (19.8–24.2)	34.0 (30.6–37.4)	37.0 (33.3–40.7)
Microbial-limited	20.2 (18.2–22.2)	33.0 (29.7–36.3)	42.0 (37.8–46.2)	21.8 (19.6–24.0)	36.0 (32.4–39.6)	39.0 (35.1–42.9)

Critical limit for total viable count (TVC) or yeast and mold counts = 5 log CFU/mL (10⁵ CFU/mL). Critical limit for color difference (ΔE) = 3.5. Critical limit for DPPH, FRAP, TPC, TFC = 50% of its initial value. Predicted shelf-life (days) of mango–passion fruit smoothie based on individual quality attributes and integrated shelf-life. The integrated shelf-life was defined as the minimum predicted value among all attributes according to their predefined acceptance criteria.

In control samples, both glass and PET containers reached the critical microbial limit of 5 log CFU/mL at 16 days, which constrained the shelf-life of other attributes such as color (ΔE), antioxidant activity (DPPH and FRAP), and bioactive compounds (TPC and TFC). This indicates that, under untreated conditions, microbial growth primarily dictates the safe storage period, and chemical changes in the beverage remain secondary within this timeframe.

Thermal pasteurization significantly improved the predicted shelf-life by reducing the microbial load and slowing chemical degradation. As shown in Table 1, pasteurized beverages in glass maintained microbial quality for up to 29 days, whereas PET-packaged beverages were safe for 27 days. Within these limits, color changes were slower, with ΔE remaining within acceptable thresholds, and antioxidant activity (DPPH and FRAP) as well as total phenolic and flavonoid contents (TPC and TFC) were better preserved. These results reflect the combined effect of microbial control and reduced oxidative degradation. Pasteurization also slightly favored glass over PET in terms of retention, likely due to glass’s superior barrier properties against oxygen and light.

Chemical preservation with DMDC at 250 ppm provided the most substantial extension of predicted shelf-life. According to Table 1, glass containers with DMDC-treated beverages maintained microbial quality for 42 days, whereas PET containers were safe for 41 days. Under these conditions, all quality attributes—including ΔE, DPPH, FRAP, TPC, and TFC—remained within acceptable limits throughout the microbial shelf-life. The k values for chemical degradation were lowest for DMDC-treated samples, indicating slower loss of functional properties. These findings highlight that chemical preservatives, combined with effective packaging, can markedly enhance both the safety and functional stability of fruit-based beverages.

Overall, microbial quality is the key determinant of predicted shelf-life, which in turn constrains the stability of all other quality parameters (Table 1). Glass packaging consistently offered slightly longer shelf-life than PET across all treatments. Among treatments, DMDC 250 ppm provided the longest predicted shelf-life (Glass: 42 days; PET: 41 days), followed by pasteurization (Glass: 29 days; PET: 27 days), while untreated controls had the shortest safe storage period (16 days). These results emphasize the importance of combining chemical treatment, thermal processing, and high-barrier packaging to extend the shelf-life of perishable fruit-based beverages while maintaining their nutritional and sensory qualities.

The study’s findings also align with previous research on similar beverages. For instance, Visuthiwan and Assatarakul [6] reported a shelf-life of 40 days for lychee juice treated with UV radiation which is comparable to DMDC’s performance. However, previous studies often focused on single-fruit juices, whereas the complex matrix of a mango-passion fruit smoothie (with diverse microbial and enzymatic challenges) makes DMDC’s efficacy particularly notable.

4. Conclusions

This study confirms that dimethyl dicarbonate (DMDC, 250 ppm), particularly in glass packaging, effectively improves the stability and safety of a mixed mango–passion fruit smoothie stored at 4 °C for six weeks. DMDC significantly reduced losses in antioxidant capacity (DPPH and FRAP), preserved total phenolic and flavonoid contents, and maintained microbial loads below 5 log CFU/mL for total viable counts and 3 log CFU/mL for yeast and mold. In comparison, pasteurization (90 °C for 100 s) provided moderate microbial control but was associated with greater degradation of bioactive compounds and noticeable microbial regrowth during storage. Packaging influenced color stability, as PET bottles exhibited higher total color difference (ΔE) values than glass, although both materials showed minimal differences in pH and total soluble solids. Kinetic analysis indicated that the first-order model (R² = 0.874–0.986) best described quality deterioration, supporting reliable shelf-life estimation. Microbial growth was identified as the principal shelf-life limiting factor. Untreated samples remained acceptable for approximately 16 days, whereas pasteurization extended microbial stability to 27–29 days and DMDC treatment to 41–42 days, while preserving key quality attributes. Glass packaging demonstrated slightly superior protective performance, likely due to improved barrier properties. Overall, DMDC showed comparable or, in some aspects, superior performance to thermal pasteurization in maintaining antioxidant activity and color stability. Nevertheless, regulatory compliance, labeling considerations, and consumer perception of chemical preservatives should be addressed before commercial application. The absence of formal sensory evaluation represents a limitation, and future research should assess sensory acceptance and explore potential synergistic combinations of DMDC with other non-thermal preservation technologies to support broader industrial implementation.

Author Contributions

S.J.: Investigation, Formal analysis, Data curation and Writing—original draft. N.R.: Conducting the lab work. M.F.: Data curation and Writing—original draft. M.U.: Data curation and Writing—original draft. K.A.S.: Review and revise. I.K.: Review and revision. S.C.: Writing—review and editing. D.K.M.: Writing—review and editing. K.A.: Conceptualization, Data curation, Funding acquisition, Project administration, Supervision, Writing—original draft and Writing—review and editing. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

This research was funded by The Thailand Research Fund, Thailand under the project ‘Research and Researchers for Industries-RRi (MSD61I0062),’ and the Department of Food Technology, Faculty of Science, Chulalongkorn University.