Application of plasma-activated fog (PAF) in postharvest treatments to reduce spoilage by fungal pathogens and pesticide residues in fruits

Abstract

Background

See More

During storage, fruits and vegetables are susceptible to the pathogens responsible for postharvest decay. Various tools are available to manage these issues, but not all are environmentally sustainable. Low-temperature plasma (LTP) has garnered significant attention among the most promising and eco-friendly solutions. LTP can be applied directly or indirectly, offering versatile applications. One notable indirect application is the utilization of plasma-activated water (PAW). In this study, we investigated the efficacy of an aerosol made by droplets of water nebulized by the effluent gases of a plasma discharge as a delivery method of PAW to substrates. We named this novel application, reported for the first time, plasma-activated fog (PAF). In this work, it was tested as a new alternative technology for fruit decontamination against postharvest fungal pathogens and pesticide residues.

Results

PAF was generated via volume dielectric barrier discharge (VDBD) in a jet-like configuration and was applied to evaluate the in vitro effects on the conidial germination of major fungal postharvest pathogens, such as Alternaria alternata, Aspergillus carbonarius, Botrytis cinerea, Cladosporium sp., Monilinia fructicola, Penicillium italicum, Penicillium expansum and Rhizopus sp. Differences in fungal sensitivity to PAF were recorded, with A. alternata showing the lowest sensitivity to treatments. For most species, complete spore inhibition was obtained after 3–5 min of exposure. The efficacy of PAF against fungal rot was assessed on table grapes and strawberries, revealing a reduction in the percentage of rotted fruits exposed to 10 min of treatment, ranging from 45 to 80% on table grapes and from 52 to 74% on strawberries. PAF treatments also reduced pesticide residues on grape bunches and strawberry fruits, with various results depending on the active ingredient, with reductions of up to 96% for abamectin among insecticides and acaricides, and up to 38% for the fungicide fenhexamid.

Conclusions

The results obtained in the present work have the potential to refine and optimize PAF treatment conditions for the antimicrobial decontamination of plant products.

Graphical abstract

Background

Postharvest pathogens are responsible for significant economic losses every year, leading to both qualitative and quantitative losses in a wide range of fruits and vegetables, and can be responsible for human health risks due to the production of mycotoxins [1, 2].

Strawberry (Fragaria × ananassa Duch.) and table grape (Vitis vinifera L.) are highly susceptible to postharvest decay caused by a range of pathogenic fungi. The long-term preservation of fresh grapes requires appropriate methods of cold storage and chemical treatment [1, 3, 4]. Storage of strawberries is a challenge, as harvested fruits are often affected by mechanical damage and fungal rot, which can rapidly reduce fruit quality [5].

Strawberry production is approximately 10 million tons worth more than 22 billion USD from approximately 400,000 ha worldwide [6]. In Italy, strawberries cover 4,200 ha in 2025, according to CSO Italy [7], with farming that is mainly concentrated in southern Italy (i.e. Campania and Basilicata regions). Grape cultivation is widespread throughout the world [8], and the Apulia region is one of the most important areas in Italy in terms of grape production. In 2024, the region produced 592.600 tons over an area of 25,455 ha, compared with 1,041.8 tons and 47,514 ha for the whole country [9].

The adoption of innovative technologies that prioritize sustainability and human safety becomes imperative to mitigate yield losses resulting from plant diseases, both in the field and postharvest. Various tools are available to manage postharvest losses, and among them, the excessive use of chemicals poses significant risks for the environment and human health because of the presence of chemical residues in food [10, 11]. The European Commission has set a concrete strategic plan aimed at reducing the use of synthetic pesticides and associated risks, with the implementation of the Farm to Fork strategy aimed at reducing pesticide usage by 50% by 2030. Additionally, the use of plant protection products must not only comply with current European regulations but also be influenced by the marketing strategies of mass market retailers, which often impose limits below the legal thresholds [12, 13].

In recent years, interest in low-temperature plasma (LTP) applications has increased in the agriculture sector and food industry [14] because of their innovative approach, which is linked to eco-friendly sustainability [15] and the absence of chemical residues released by treatments in the environment and products [16].

Many recent papers have proven the effectiveness of direct [17,18,19] and indirect methods [20,21,22,23] for applying LTP. Among indirect plasma applications, plasma-activated water (PAW) has attracted increasing interest from the scientific community for a variety of applications, such as plant irrigation, which promotes growth and defense responses [20] and decontamination of agricultural and food products from microbial pathogens [24, 25]. PAW production involves the exposure of a water reservoir to a plasma discharge. Reactive oxygen and nitrogen species (RONS) that are produced by plasma are dissolved in water, where their chemistry can advance to stable, long-lived species (i.e. H₂O₂, NO₂⁻ and NO₃⁻) that can be conveyed to a substrate. In the case of decontamination of food products, the antimicrobial effect of PAW has been widely demonstrated, highlighting the combined role of low pH and high contents of reactive oxygen and nitrogen species (mainly NO₂, H₂O₂, and reaction intermediates) [25]; however, no clear consensus on the exact mode of action has been reached. The shortest possible latency between PAW production and its application allows the efficacy of treatments to be maximized. LTP treatment was also shown to induce the degradation of pesticide residues [26,27,28,29].

In this work, we propose a novel method for PAW production and application to products that involve the nebulization of PAW at the same time as its production. Other studies reported the advantages of using a finely dispersed aerosol as a delivery medium of PAW or other plasma-activated media to the target substrate. Different methods for producing plasma-activated aerosols have been tested, but these methods involve nebulizing the previously produced PAW [30,31,32,33] or generating an aerosol starting from a plasma-activated hydrogen peroxide [34,35,36,37]. For example, de Oliveira Mallia et al. reported the antimicrobial efficacy of aerosols produced by surface acoustic waves and subsequently activated by using them as feed gas in a plasma reactor [23]. In the present work, effluent gas from a cylindrical coaxial volume dielectric barrier discharge (VDBD) reactor was used to nebulize double distilled water, obtaining a finely dispersed fog (droplet diameter < 5 μm). We named this novel technology plasma-activated fog (PAF) and investigated whether it allows the inhibition of fungal pathogens and the degradation of pesticide residues on the fruit surface. The key advantage of this approach with respect to prior works [30,31,32,33,34,35,36,37] is that a larger water surface interacts with plasma species, implying higher production of RONS delivered by aerosols. Moreover, the absence of humidity in the discharge chamber enables improved chemical stability of the plasma effluents, whereas a greater number of micrometric droplets enhances delivery and efficacy.

The aim of this study was to assess PAF as a decontamination tool for microbial and chemical contaminants on fruits. To this end, we investigated its antimicrobial potential against major postharvest fungal pathogens both in vitro and in vivo, its effect on fruit spoilage in strawberries and table grapes, and its ability to reduce various pesticide residues in fruits.

Materials and methods

Plasma-activated fog production and physicochemical characterization

Plasma reactor design

A volume dielectric barrier discharge (VDBD) reactor was used to generate the plasma used to produce the PAF. The reactor was built inside a quartz cylinder with two electrodes in a coaxial configuration.

A schematic representation of the reactor is shown in Fig. 1.

Fig. 1

Schematic representation of the plasma reactor used to produce plasma-activated fog (PAF). The insert presents the front sight of the reactor

The ground electrode was made of copper and envelopes the outer wall of the quartz cylinder, which acts as a dielectric barrier. The high-voltage (HV) electrode consists of a steel rod encapsulated in a quartz sheath positioned at the centre of the reactor. Thus, both electrodes are isolated by a dielectric barrier (quartz), avoiding the exposure of metal to the discharge zone, which could lead to sputtering and, subsequently, delivery of metal nanoparticles to targets. Ambient air was used as feed gas for the discharge and entered the reactor before the discharge volume with a flow of 10 sLm (standard litres per minute) supplied to the reactor through an oil-free compressor CleanAIR CLR 15/25 T (Ceccato Aria Compressa, Montecchio Maggiore, VI, Italy). This high flow guarantees a stable discharge with negligible energy fluctuations from run to run, allowing continuous work for several tens of minutes.

The afterglow of the resulting plasma was then injected into a glass aerosol nebulizer for medical use (GammaDis Farmaceutici, Civitanova Marche, MC, Italy) loaded with 6 mL of double distilled water (Carlo Erba water RPE for analysis), generating a finely dispersed fog with a droplet diameter of less than 5 μm, exploiting the Venturi effect, i.e. the pressure drop occurring when air passes through a narrow section, which draws liquid into the airflow and nebulizes it. Importantly, the produced fog did not wet the surfaces it encountered. The exit of the nebulizer was then positioned at the entrance of the different containers used for in vitro and in vivo treatments alike. The containers were used to create a homogeneous environment saturated with PAF.

Plasma generation parameters

The plasma reactor was powered by an AC power supply based on an Information Unlimited PVM500—Plasma Power Generator. The plasma was ignited by applying an AC voltage (with a peak voltage of 19.5 kV) composed of repetitive bursts obtained through an external trigger realized via a TG 5011 Function Generator (TTi, Fort Worth, Texas). Each burst consisted of 15 AC cycles (f_AC = 25 kHz) and had a repetition rate of 500 Hz (resulting duty cycle of 30%), which ensured homogeneous discharge in the interelectrode gap. We used a Tektronix P6015A high-voltage probe (1000:1@1 MΩ, bandwidth 75 MHz) to monitor voltage waveforms and a voltage probe (100:1@10 MW, bandwidth 1 GHz, Tektronix, Beaverton, Oregon, USA) to measure the potential drop on a transferred charge measuring capacitor (C = 2.5 nF) inserted between the induction electrode and ground. A Tektronix TDS 2014 4-channel digital oscilloscope (bandwidth 100 MHz, up to 1GS/s) was used to monitor the voltage‒charge characteristics of the discharge, whereas electrical characterization was performed via a Keysight InfiniiVision MSOX6004A with a 1-GHz bandwidth operating at 20GSa/s and 12-bit vertical resolution (Keysight Technologies, Santa Rosa, California, USA).

The charge‒voltage characteristics were used to evaluate the plasma energy, and consequently, the gas plasma dose was defined as:

$$D_{{{\text{gas}}}} = \frac{{E_{{{\text{burst}}}} \times \tau \times f }}{{V_{{{\text{discharge}}}} }},$$

(1)

where E_burst is the energy measured over a single voltage burst, f is the frequency (f = 500 Hz for all treatments), and V_discharge is the volume of the treated gas effectively exposed to plasma during its residence time in the reactor. The estimated residence time (τ) was obtained by dividing the discharge physical dimension by the gas flow rate expressed in cm³ s⁻¹, resulting in 6.6 × 10^–3 s.

Optical emission spectroscopy

We took spectra of the discharge ignited in the gap between the inner and outer UV grade quartz walls. The plasma-induced emission (PIE) was collected by a fibre adapter equipped with a UV-fused silica single-lens focusing on an optical fibre bundle model LG-455-020-3 (3 m, 190–1100 nm, with 19 fibres (ϕ = 200 µm), with a 10 mm ferrule at the slit end, Teledyne Princeton Instruments, Trenton, New Jersey, USA). The light was spectrally resolved by a 30 cm spectrometer (Acton Spectra Pro 2300) equipped with a multiple-grating turret with 300/600/1200 grooves mm⁻¹ and blazed at 300 nm. The spectra were acquired with a Princeton Instruments PI-MAX4 1024i ICCD camera equipped with a 1024 × 1024 pixel sensor (size 12.8 µm, active area 13.1 × 13.1 mm²; Teledyne Princeton Instruments, Trenton, New Jersey, USA). The intensities of the emission spectra acquired by the ICCD detector were spectrally calibrated, and the intensities were calibrated via Halogen (Oriel Lightning, Darra, Australia) and calibration lamps (Oriel Lightning, Darra, Australia). Spectra were used to evaluate rotational temperatures and nitrogen vibrational distributions by comparison with spectra simulated via the freely available spectroscopic tool Massive OES [38, 39].

Physical characterization of PAF

The dimensions of the droplets constituting the aerosol were estimated via microscopic imaging. The Nd:YAG laser second harmonic (532 nm) (OPOTEK Opolette™ 355 LD) output was passed through a telescope formed by a plano-convex lens and a cylindrical lens to form a laser sheet at the exit of the nebulizer to image the aerosol. The resulting scattered light images were collected through an Olympus BH2-UMA optical microscope (Olympus Corporation, Hamburg, Germany) to achieve up to 480×magnification and then processed through a custom routine written in MATLAB [40] to estimate the droplet size. The image pixel dimensions were calibrated via a microscope counting slide (Bürker), resulting in a resolution of 0.48 μm/pixel.

The temperature of the plasma reactor was measured via a FLIR ThermaCAM E320 infrared (IR) camera (FLIR Systems, Wilsonville, Oregon, USA). This device operates within a spectral range of 7.5–13 µm and provides a thermal sensitivity of ≤ 0.1 °C at 30 °C. The camera was positioned at a fixed distance of one metre from the reactor to ensure consistent imaging conditions. The camera calibration was previously performed using a blackbody reference source before the measurements were taken to account for emissivity variations. The temperature distributions were recorded under different operational conditions: with or without water in the nebulizer and with the plasma discharge ignited or not ignited, while an airflow of 10 sLm was kept constant. The data were analysed to identify the thermal gradients and evaluate heat dissipation within the reactor.

Chemical characterization of PAF

PAF was collected by placing an ice-refrigerated glass recipient at the exit of the nebulizer and using all the 6 mL contained in the water reservoir of the nebulizer. Thus, ~ 3 mL of liquid was condensed and subjected to chemical analyses to tentatively characterize the PAF contents in terms of RONS. Different treatment times (1, 5 and 10 min) and input voltages were explored. The concentrations of such long-lived dissolved RONS, particularly hydrogen peroxide (H₂O₂), nitrite (NO₂⁻), and nitrate (NO₃⁻) ions, were determined via a multiparametric photometer, Aqualytic AL400 (Aqualytic GmbH & Co., Dortmund, Germany), along with Cell Test Lovibond® Water Testing kits (Tintometer Inc., Dortmund-Aplerbeck, Germany), following the manufacturer’s protocols for quantifying nitrate (Vario Nitra protocol X 535580), nitrite (protocol 512310BT), and hydrogen peroxide (protocol 512381BT). Additionally, the pH values were measured via a HANNA Instruments pH meter (model HI98191) equipped with a HI72911 pH probe (HANNA Instruments, Woonsocket, Rhode Island, USA). The measurements were repeated three times, and the results were averaged.

In vitro inhibitory activity of PAF against postharvest fungal pathogens

The following strains of the main postharvest fungal pathogens were used: Botrytis cinerea SAS56, Monilinia fructicola Mfrc123, Aspergillus carbonarius AC49, Rhizopus sp. FA1, Penicillium italicum PIO1, Penicillium expansum 141, Cladosporium sp. 173, and Alternaria alternata ALT1. The strains were stored at -80 °C as suspensions of conidia and mycelia in aqueous 10% glycerol and were revitalized on potato dextrose agar (PDA: infusion of 200 g of peeled and sliced potatoes kept at 60 °C for 1 h, 20 g of D-(+)-glucose adjusted to pH 6.5, and 20 g of European bacteriological agar (LLG) Labware per liter) to obtain fresh cultures.

The inhibitory effect of PAF treatment on the selected fungi was evaluated through conidial germination assays [17]. Specifically, to promote sporulation, colonies were grown at 25 ± 1 °C in the dark for 2 days and then exposed to a combination of two daylight (Osram, L36W/640) and two near-UV (Osram, L36/73) lamps with a 12 h light/dark photoperiod for 7 days. The conidia were suspended in sterile distilled water by scraping the colony surface with a sterilized loop, filtered through glass wool to remove mycelial fragments, and adjusted to a concentration of 1 × 10⁵ conidia mL⁻¹ by using a hemocytometer. Aliquots (10 μL) of the obtained conidial suspension were then spotted on two replicated PDA disks (6 mm diameter) placed on sterile microscope slides and exposed to the plasma treatment in a 5-L plastic box (19 cm × 29 cm × 14 cm) modified for nebulizer entry on one side of the box.

For all the fungi tested, the efficacies of seven different durations of exposure to PAF, ranging from 15 s to 10 min, and four different plasma application systems were measured and compared. Specifically, insufflation of humid (RH ~ 75%) and ambient (RH ~ 35%) air through the discharge and then into the nebulizer containing water (reported as humid air with water, PAF-HAW, and dry air with water, PAF-DAW, respectively) were the two modes of application of PAF, whereas insufflation of humid and ambient air through discharge and then into a nebulizer without water (reported as humid air with no water, P-HANW, and dry air with no water, P-DANW, respectively) were the two modes used as controls to assess the effects of the plasma effluents alone. The tested experimental conditions are reported schematically in Table 1.

Table 1 Treatment conditions used in the in vitro tests

Untreated controls consisting of untreated PDA disks inoculated with untreated conidia were included in all the experiments. After treatment, the disks were incubated in a moist chamber at 25 ± 1 °C in the dark to allow conidial germination. After 18 h, microscopic observations were conducted at × 200 magnification on random samples of 100 conidia on each of three replicated spots per condition via a Leica DM300 microscope (Leica Microsystems, Wetzlar, Germany). The inhibition rate (%) caused by the plasma treatment was calculated by considering the percentage of germinated conidia on the untreated control disk. After the efficacy evaluation, PAF-DAW was chosen as the default working condition and subsequently, for the sake of clarity, referred to as PAF.

Efficacy of PAF treatment against the postharvest decay of table grape and strawberry fruits

A first evaluation of the effectiveness of PAF treatments in reducing postharvest fruit rot was performed on strawberry fruits (cv ‘Inspire’) collected from a commercial orchard located at Scanzano Jonico (Province of Matera, Southern Italy), which was grown according to organic standards. Healthy and free-of-defects fruits were harvested at the commercial maturity stage, selected for uniformity of size and degree of ripening, and used for the experiment on the day that they were harvested. Fruits were arranged in plastic trays (14.2 cm × 9.5 cm × 5 cm) and exposed to PAF treatment (5 min and 10 min). The exit of the nebulizer (plasma dispenser) was placed directly into the tray through a hole, and during the treatment, the PAF saturated the entire environment where the strawberries were placed. Fruits exposed to the fog produced with the discharge off for 10 min were used as controls. After treatment, the strawberries were stored in the dark at 4 ± 1 °C for 3 days, followed by 3 days of storage at 25 ± 1 °C. During storage, the trays were enclosed in plastic bags to maintain high relative humidity (95–98%). At the end of storage, the number of fruits showing rots caused by B. cinerea, Rhizopus sp. and other secondary fungi (e.g., Alternaria spp., Aspergillus spp., Penicillium spp.) was recorded, and the prevalence (P) was calculated as the percentage of rotten fruits. Each treatment included five replicates, each consisting of 5 fruits per tray, and the experiment was replicated twice. Treatment efficacy (E) was calculated via Abbott’s formula [41]: [(P_untreated − P_treated)/P_untreated] × 100.

A second trial was carried out on fruits of the Candonga strawberry cultivar Sabrosa collected from an organic orchard in Policoro (Province of Matera, Southern Italy). In this case, fruits were arranged in a single layer in open plastic trays (19 cm × 11.5 cm × 6 cm), with approximately 500 g of strawberry per tray, placed into the 5-L plastic box used as the treatment chamber and exposed to 5 min and 10 min of PAF treatment. After the treatment, the storage at 4 ± 1 °C was extended to 5 days, followed by 2 days of storage at 25 ± 1 °C before the first symptom assessment.

To evaluate the efficacy of PAF treatments on table grapes, a first preliminary trial was carried out on bunches of two seedless cultivars, the white cultivar ‘Superior’ and the black cultivar ‘Attica’, which were harvested at their optimum ripeness in a ‘tendone’ vineyard located at Rutigliano (Province of Bari, Southern Italy). After being harvested, the grapes were transported to the laboratory and stored overnight at 4 ± 1 °C. The next day, portions of bunches, each consisting of at least five berries, for a total of 500 g in weight, were placed on a plastic mesh raised 2 cm in the treatment chamber and exposed to PAF for 10 min. Untreated samples were used as controls. Each treatment included five replicates. After treatment, the bunches were packaged in plastic trays (18 cm × 11 cm × 10 cm), enclosed in plastic bags to ensure high relative humidity, stored at 4 ± 1 °C for 3 days and then maintained at 25 ± 1 °C for up to 9 days. The number of berries showing rot symptoms was recorded after 3, 5 and 9 days of storage to determine the disease prevalence and treatment efficacy, which was measured as the E.

Further experiments were carried out on white table grape cultivars ‘Autumn Crisp’ and ‘Autumn King’ to compare the efficacy of different exposure times (5, 7.5, and 10 min) and different modes of grape exposure (unpacked, open tray, and closed perforated tray) with a fixed exposure time (10 min). After treatment, the bunches were maintained for 7 days at 4 ± 1 °C and then stored for 8 and 13 days at 25 ± 1 °C. Daily assessments were conducted to evaluate the progression of berry rot.

Efficacy of PAF treatments in reducing pesticide residues

The efficacy of PAF in reducing pesticide residues was first tested in table grape bunches artificially contaminated with different doses of pesticides. Commercial formulations of abamectin (Vertimec Pro, 18 g L⁻¹ active substance (AS); Syngenta Global AG), acetamiprid (Epik SL, 50 g L⁻¹ AS; Sipcam Italia SpA), fenhexamid (Teldor Plus, 500 g L⁻¹ AS; Bayer CropScience), fludioxonil (Geoxe, 500 g L⁻¹ AS; Syngenta Global AG), and isofetamid (Kenja, 400 g L⁻¹ AS; Certis Belchim BV) were suspended in sterile distilled water. For each product, three different doses were used: the recommended dose (full dose), 50% of that dose (1/2 dose), and 25% of that dose (1/4 dose). Specifically, at harvest, grape bunches cv ‘Italia’ were collected from a commercial vineyard located in Canosa di Puglia (Apulia Region, Southern Italy). From the harvested grapes, a sample of approximately 50 kg was randomly collected for contamination with pesticides. Three replicates of 500 g each were then immersed for 1 min in a 10 L volume of the respective suspensions containing the pesticide, dried at room temperature for 12 h, and packaged in open plastic trays for treatment with PAF. Treatments of 10 min of exposure to PAF were performed as described above. After PAF exposure, the pesticide residue contents of treated and untreated table grapes were analysed by an external service (BonassisaLab SpA, Foggia, Italy) using a multiresidue analytical method via the QuEChERS method [42] following the European Committee for Standardization (CEN) method CEN-EN 15662:2018 [43], on the Agilent 7010C GC/TQ and 6470B LC/TQ coupled with 1260 Infinity II LC systems (Agilent Technologies, Inc., Santa Clara, CA, USA) and limit of quantification (LOQ) of 0.01 mg kg⁻¹.

Further studies were conducted on grape and strawberry samples collected from commercial fields and subjected to conventional crop protection strategies. After harvesting, samples of strawberry (cv ‘Sabrosa’) and table grape (cv ‘Autumn Crisp’) were subjected to 10 min of PAF treatment and then analysed for pesticide residues. Multiresidue analysis was performed by Agro.Biolab Laboratory (Rutigliano, Bari, Italy) using the QuEChERS method following CEN-EN 15662:2018. Two LC-MS/MS systems were used, both achieving a limit of quantification (LOQ) of 0.005 mg kg⁻¹. The first system consisted of an AB Sciex Triple Quad™ 5500 mass spectrometer (SCIEX, Concord, Ontario, Canada) coupled with a Shimadzu UHPLC system comprising a Nexera X2 LC-30AD binary pump, a Nexera X2 SIL-30AC autosampler (Shimadzu Scientific Instruments, Columbia, MD, USA), and a Shim-pack Velox Biphenyl column (2.7 µm, 2.1 × 100 mm). The second system employed a Shimadzu LCMS-8060NX triple quadrupole mass spectrometer, connected to a Shimadzu Nexera X3 UHPLC system equipped with an LC-40B X3 binary pump and an SIL-40C X3 autosampler, using the same Shim-pack Velox Biphenyl column (2.7 µm, 2.1 × 100 mm). Data from treated and untreated control samples were compared, and the efficacy of PAF treatment in reducing each residue was determined.

Statistical analysis

The data were analysed via OriginPro 2024b [44], tested for normality via the Kolmogorov‒Smirnov test and homogeneity of variance via the Levene’s test (square deviations), and subjected to analysis of variance (ANOVA) for a completely randomized design, using software-estimated parameters and setting significance level at 0.05. The mean values of the treatments were compared via Tukey’s honestly significant difference (HSD) test at P ≤ 0.05 and P ≤ 0.01 probability levels.

Results

Plasma electrical characteristics

The discharge charge‒voltage characteristics are shown in panels (a) and (b) of Fig. 2.

Fig. 2

Typical electric characteristics of the plasma discharge during a voltage burst in terms of a the applied voltage; b the developed charge; and c the charge‒voltage Lissajous figure

In panel (c) of Fig. 2, a typical charge‒voltage Lissajous figure, obtained by plotting the charge vs. voltage waveforms averaged over 8 bursts, is represented. The resulting parallelogram-like shape is very close to the ideal shape, and the waveforms at the boundaries fall within a tight margin, indicating that the discharge was homogeneous and that the majority of the energy supplied to the reactor was deposited in plasma and, thus, in feed gas. The area of this figure represents the energy dissipated over each burst estimated to be ̴ 38 mJ, resulting from Eq. 1, at a gas plasma dose (D_gas) of 116.1 J/L_air.

Plasma optical characteristics

Figure 3 shows the experimental mean N₂ second positive system (SPS) (0,0) band profile, with the corresponding standard error, together with synthetic profiles simulated for rotational temperatures of 300 K, 380 K and 450 K for the PIE over the entire voltage burst. The estimated error from multiple simulations of bands, not shown for the sake of clarity, at different temperatures was on average on the order of 10% of the reported measurements, with values of ± 30, ± 38, and ± 45. By comparing the acquired and simulated spectra, we can surmise that the temperature developed in a plasma filament indeed falls within the 300–450 K range, and more specifically, the spectrum simulated at a temperature of 380 K offers a good fit for the experimental points.

Fig. 3

N₂ SPS(0,0) band signature for plasma temperature measurements

Physical properties of PAF

Droplet dimensions

Microscope images of the aerosol droplets were captured at the nebulizer’s exit to determine the droplet diameter. The acquired data were normalized and grouped into bins, with each bin width corresponding to the resolution of the imaging technique (0.48 μm per pixel). The resulting probability density function (PDF) is presented in Fig. 4.

Fig. 4

The aerosol droplet diameter normalized probability density function (PDF) is represented as a histogram (yellow columns) and fitted through a bimodal lognormal function (magenta curve). Both axes are on a logarithmic scale for better visualization of lower-probability effects

The measured droplet diameters range from just below 1 μm to approximately 50 μm. As shown in Fig. 4, the size distribution exhibits a bimodal structure, with two distinct peaks or regions of higher density. The primary mode, centred at approximately 1 μm and extending to approximately 5 μm, accounts for most droplets, which appear much more frequently in the dataset. A secondary mode, representing less frequent larger droplets, is centred at 10 μm. These components may represent different droplet generation mechanisms. Consequently, most droplets produced by the nebulizer used in this study were smaller than 5 μm.

Thermal images of the reactor

Infrared images of the reactor under various operating conditions were taken to evaluate the temperature distribution and heat dissipation of the apparatus. The thermal images shown in Fig. 5 were all taken with 10 sLm of air flowing through the reactor, without (left column) or with (right column) water in the nebulizer and with the plasma off (top row) or on (bottom row).

Fig. 5

Thermal images of the reactor under different operating conditions: a no water in the nebulizer, plasma turned off; b no water in the nebulizer, discharge ignited (P-DANW as per Table 1); c water in the nebulizer, plasma off; d water in the nebulizer, discharge ignited (PAF-DAW). The numbers in the top-right corner of each picture represent the ambient temperature measured at the centre of the crosshairs. On the right of each picture, a grayscale image of the temperature, calibrated by the device specifically for each picture, is shown

Figure 5 shows that two cooling processes occur in the reactor, specifically inside the nebulizer. The first is the air expansion at the nozzle of the Venturi tube (panel a), and the second is the nebulization of water into an aerosol of micron-sized droplets (panels c and d). When the discharge is ignited under the copper tape ground electrode, the temperature immediately increases, as evidenced by the brighter white region in the discharge zone. Under working conditions for PAF production (panel d), the temperature in this zone is 56 °C. However, this temperature decreases as the effluent gas travels through the reactor, eventually cooling to 17 °C within the nebulizer. Therefore, the PAFs delivered to the substrates were at or below room temperature.

Chemical characterization of the produced PAF

The PAF produced by blowing dry air into the plasma reactor and varying the treatment time and input voltage was collected just at the exit of the nebulizer and characterized in terms of its RONS content and pH. The results are shown in Fig. 6.

Fig. 6

Chemical characterization of PAW in terms of RONS concentration and pH. In the top row, the results obtained with different durations of plasma treatment are shown. On the bottom row, the characterized PAF was produced with 5-min treatments of varying input voltages, which resulted in different dissipated energies per burst (38.7 mJ and 21.9 mJ, respectively) and thus different gas plasma doses, as reported on the abscissa axis. In each graph, columns labelled with the same letters are not significantly different from other columns of the same colour at the probability level P = 0.05, according to Tukey’s HSD test

The concentration of RONS in the liquid phase after plasma treatment increased with increasing treatment time for all the species. Notably, a much larger fraction of nitrate ions was produced, as they are the more stable species in the solution. Double distilled water (initial pH = 6.4) was acidified immediately, and no large difference in pH was observed between the treatment times of 5 and 10 min. Changing the input voltage did not seem to influence the pH or the production of nitrite ions and hydrogen peroxide but had a significant effect on the concentration of nitrate ions, which was much lower at lower voltages.

In vitro evaluation of the inhibitory activity of PAF against fungal pathogens

The four treatments used, i.e. PAF-DAW, PAF-HAW, P-DANW and P-HANW (see Table 1), inhibited conidial germination to different extents in the eight fungal species tested, as shown in Fig. 7. As expected, the efficacy of the treatments progressively increased with increasing exposure time, reaching the highest rates of inhibition after 3 and 5 min of treatment for most of the fungal species tested, except for A. carbonarius and A. alternata. No significant differences in the inhibitory effects were observed among samples located at different positions in the treatment chamber (data not shown), confirming that samples’ exposure to PAF was uniform across the container.

Fig. 7

Inhibition of conidial germination of different fungal species at different exposure times to low-temperature plasma via different application systems: a PAF-DAW = PAF in dry air; b PAF-HAW = PAF in humid air; c P-DANW = plasma in dry air and no water in the nebulizer; d P-HANW = plasma in humid air and no water in the nebulizer

PAF-DAW and PAF-HAW generally presented the highest antifungal efficacy, with complete inhibition mostly achieved after 5 min of treatment (Fig. 7a and b). In particular, M. fructicola, P. italicum, and B. cinerea were almost completely inhibited after 30 s of PAF exposure. Longer exposure times were needed for ≥ 90% inhibition of Cladosporium sp. (3 min), and P. expansum, Rhizopus sp., and A. carbonarius (5 min). A. alternata presented the lowest level of sensitivity to all the treatment conditions, with an inhibition rate of 28% after 10 min of exposure to PAF.

Three-factor ANOVA was used to explore possible interactions between the plasma application system, the exposure time, and the fungal species. The analysis revealed that all three factors had a statistically significant effect on fungal inhibition and that the interactions among the three factors were significant (Table 2).

Table 2 Three-way ANOVA results for the inhibition rates (%) of conidial germination by the four plasma application systems, seven plasma exposures, and eight fungal species

In detail, statistically significant differences were observed among the plasma application systems tested, with the highest efficacy achieved by treatments using PAF, i.e. dry air with water in the nebulizer (PAF-DAW), followed by treatments using humid air and water in the nebulizer (PAF-HAW), whereas the lowest efficacy was recorded by plasma treatments with no addition of water in the nebulizer (P-DAW and P-HAW).

With respect to the exposure time, the inhibitory efficacy significantly increased in a direct proportional manner and reached values higher than 90% at 10 min.

Among the fungal species, M. fructicola was the most sensitive to the various treatments, followed by B. cinerea and P. italicum, which presented similar sensitivities. A. alternata was the most resistant, whereas the remaining fungi (Cladosporium sp., P. expansum, Rhizopus sp. and A. carbonarius) presented intermediate behaviour.

Efficacy of PAF treatment against the postharvest decay of table grape and strawberry fruits

The efficacy of PAF against postharvest rots on strawberry fruits of two different cultivars (‘Inspire’ and ‘Sabrosa’) and berries of table grapes of four different cultivars (‘Superior’, ‘Attica’, ‘Autumn Crisp’ and ‘Autumn King’) was evaluated.

The results obtained for strawberry are shown in Fig. 8. In particular, for the cultivar ‘Inspire’, after 3 days of cold storage (4 ± 1 °C) and 3 days of shelf storage at room temperature (25 ± 1 °C), the majority of the fruits in the untreated control showed rot, which was predominantly caused by Rhizopus spp.. Compared with the untreated control, both treatments with PAF significantly reduced rot symptoms by 40% and 74% after 5 min and 10 min of treatment, respectively.

Fig. 8

Prevalence (%) of postharvest rots on untreated and PAF-treated fruits of two strawberry cultivars (‘Inspire’ and ‘Sabrosa’) after different exposure times. Fruits rots assessed after 3 days of cold storage at 4 ± 1 °C and 3 days of shelf life at 25 ± 1 °C on cultivar ‘Inspire’ and after 5 days of cold storage at 4 ± 1 °C and 2 days of shelf life at 25 ± 1 °C on cultivar ‘Sabrosa’. For each cultivar, the same letters are not significantly different at the probability level P = 0.05, according to Tukey’s HSD test

In the Candonga strawberry cultivar ‘Sabrosa’, which was cold stored for 5 days (4 ± 1 °C) and then exposed to 2 days of shelf life (25 ± 1 °C), fruit rots in the untreated control group were caused predominantly by B. cinerea. A significant reduction in rot prevalence, as compared with that of the control, was recorded after PAF exposure for 5 min (E = 54%) and 10 min (E = 51%).

For table grapes, the first evaluation of the efficacy of treatments was carried out on two different cultivars (‘Superior’ and ‘Attica’) exposed to PAF for 10 min compared with the untreated control. The first rot symptoms were observed in the untreated control after 3 days of cold storage at 4 ± 1 °C, with 14% of the berries of the black cultivar ‘Attica’ showing symptoms of secondary rot caused mainly by Penicillium sp. and Aspergillus sp. and 24% of the berries of the white cultivar ‘Superior’ showing symptoms of grey mould caused by B. cinerea. Treatments with PAF significantly reduced the percentage of symptomatic fruits (P = 5%), reaching an efficacy of 80% for the cultivar ‘Superior’, whereas no significant difference in rot incidence was recorded for the treated cultivar ‘Attica’ (P = 13%) (data not shown). The efficacy of treatments, however, progressively increased on ‘Attica’ grapes after five days of shelf life (dsl) (E = 45%), whereas grey mould symptoms also appeared in the untreated control (P = 3.6%), while good efficacy levels (75%) were maintained on ‘Superior’, where infected berries in the untreated control, at that time point, reached 42% (Fig. 9). For both cultivars, a slight decrease in efficacy was recorded after 7 dsl (E = 32% and 64% for ‘Attica’ and ‘Superior’, respectively) (data not shown).

Fig. 9

Prevalence (%) of postharvest rots on untreated and PAF-treated table grapes of the two cultivars (‘Attica’ and ‘Superior’), assessed after 3 days of cold storage at 4 ± 1 °C and 5 days of shelf life at 25 ± 1 °C. For each cultivar, the same letters are not significantly different at the probability level P = 0.05, according to Tukey’s HSD test

The efficacy of PAF was then tested on two other cultivars of table grapes (‘Autumn Crisp’ and ‘Autumn King’) with three different exposure times (5, 7.5, and 10 min) and three different modes of grape exposure (unpacked, open tray, and closed perforated tray) with a fixed exposure time of 10 min.

In ‘Autumn Crisp’, the prevalence of rot in the untreated control was 8% at 8 dsl and reached 20% at 13 dsl, whereas in ‘Autumn King’, it was 34% and 67% at the two time points. A statistically significant disease reduction was observed after PAF treatment in both cultivars.

At 8 dsl, the efficacy of the treatments on ‘Autumn Crisp’ reached 87% after 5 min of exposure, and a lower efficacy (~ 50%) was recorded after 7.5 min and 10 min of treatment, whereas on ‘Autumn King’ it was 36–52% with no statistically significant difference among the treatment times (data not shown). At 13 dsl, the efficacy was always significant, with E values ranging from 47 to 63% on ‘Autumn Crisp’ and from 26 to 49% on ‘Autumn King’ (Fig. 10). Longer in vivo treatments of 15 min (split into two 7.5-min sessions) resulted in slightly reduced efficacy (data not shown).

Fig. 10

Prevalence (%) of postharvest rots on untreated and PAF-treated table grapes cv ‘Autumn Crisp’ and ‘Autumn King’, comparing different durations of exposure to PAF (5, 7.5, and 10 min) after 7 days of cold storage at 4 ± 1 °C followed by 13 days of shelf life at 25 ± 1 °C. For each cultivar, the same letters are not significantly different at the probability level P = 0.05, according to Tukey’s HSD test

When different modes of grape exposure to PAF treatments were compared, the prevalence of rots on the untreated control was very different between the two cultivars with P = 3% on ‘Autumn Crisp’ and P = 78% on ‘Autumn King’ at 13 dsl. The bunches packaged in open trays had the greatest effect on ‘Autumn Crisp’, with E = 85%. Treatments on bunches unpacked and in closed trays showed no efficacy, with the latter leading to an increase in rot prevalence (Fig. 11). For ‘Autumn King’, the unpacked bunch and open tray treatments showed significant efficacy that was similar for the two conditions (E = 45–52%), whereas the closed tray treatment provided the worst results (Fig. 11).

Fig. 11

Prevalence (%) of postharvest rots on untreated and PAF-treated table grapes cv ‘Autumn Crisp’ and ‘Autumn King’, comparing different modes of grape exposure (unpacked, open tray, and closed perforated tray) with a fixed exposure time of 10 min after 7 days of cold storage at 4 ± 1 °C followed by 13 days of shelf life at 25 ± 1 °C. For each cultivar, the same letters are not significantly different at the probability level P = 0.05, according to Tukey’s HSD test

Efficacy of PAF treatments in reducing pesticide residues

The efficacy of PAF in reducing pesticide residues on table grapes and strawberries was evaluated, revealing an overall reduction in the pesticides tested after 10 min of treatment. However, the extent of decontamination varies for different pesticides, likely due to their differences in molecules and physicochemical properties.

The results of the preliminary evaluations performed under laboratory conditions on table grape are shown in Table 3. Compared with the untreated control, the PAF treatment removed ≥ 95.7% of abamectin (an insecticide and acaricide), with a statistically significant difference at all three doses tested. This allowed us to obtain residue levels on the treated samples that were always equal to or below the maximum residue level (MRL) legally allowed on table grape. The treatment was mildly effective in removing the insecticide acetamiprid, with an effectiveness ranging from 32.0% (1/4 dose) to 39.7% (full dose), and the fungicide fenhexamid, with an effectiveness ranging from 34.7% (1/2 dose) to 37.8% (1/4 dose). In this case, the statistically significant differences between treated and untreated samples were achieved only at the full dose of fungicide. The effectiveness of the fungicides fludioxonil and isofetamid was not significant. It ranged from 4.5% (1/4 dose) to 24.5% (1/2 dose) for fludioxonil and from zero (full dose) to 29.1% (1/4 dose) for isofetamid.

Table 3 Efficacy of pesticide, fungicide, insecticide and acaricide reduction in artificially contaminated table grape samples under laboratory conditions

The results obtained for the decontamination activity of PAF on table grape and strawberry contaminated by standard applications of pesticides under commercial farming conditions are depicted in Fig. 12. On table grape, decontamination activity against acetamiprid (45.6%) was confirmed for PAF treatment, which could also reduce the residues of the fungicides dimethomorph (36.4%) and metalaxyl (7.4%) and trace amounts of penconazole (19.2%) and proquinazid (5.6%) (Fig. 12a). In strawberries, PAF treatments were slightly effective in reducing the concentrations of the fungicides fludioxonil (34.3%), penconazole (38.6%) and penthiopyrad (36.1%) and were very effective in reducing the residues of insecticides and acaricides, such as bifenazate and spinetoram (approximately 68% reduction), hexythiazox (52.5%), and trace amounts of cyflumetofen (41.2%) and pirimicarb (54%) (Fig. 12b).

Fig. 12

Efficacy of PAF treatment (10 min) on table grape cultivar ‘Autumn Crisp’ (a) and strawberry fruit cultivar ‘Sabrosa’ (b) contaminated by standard spray applications in commercial fields. The labels on the “Treated with PAF” columns show the pesticide reduction (%) in treated vs. untreated fruits

Discussion

The PAF proposed in this study showed significant potential for applications on fruits, consisting of direct antifungal activity, a reduction in fruit decay during the postharvest period, and an effective reduction in pesticide residues.

The plasma discharge characteristics revealed homogeneous energy deposition within the reactor, as indicated by the near-ideal Lissajous figure, confirming efficient plasma generation. The chosen configuration, with the discharge zone positioned before the nebulization chamber, reduces the variance between experiments by limiting the detrimental effect of humidity on the discharge electrical characteristics and the consequent plasma parameter variability. Furthermore, the double dielectric barrier design of the reactor isolates both electrodes from the discharge itself, preventing sputtering phenomena that could lead to metal nanoparticle contamination of treated products and metallic electrode degradation. While the temperature of the gas in the plasma phase could increase to 107 °C (380 K, as shown in Fig. 3), it is important to emphasize that the filamentary nature of the discharge means that the temperature measured via optical emission reflects only the gas volume crossed by streamers, not the entire chamber. Thus, the global temperature of the gas inside the reaction chamber should be closer to 60 °C, as measured via the IR camera. Moreover, the gas dynamic expansion that occurs at the nozzle positioned at the end of the Venturi tube, after the discharge zone, significantly reduces the plasma temperature. Moreover, the formation of aerosols from water by the Venturi effect further reduces the temperature to sub-room values (~ 17 °C) so that the treatment targets are never exposed to warm air. The produced aerosol consists of droplets smaller than 5 μm, largely falling in the 1–2 μm range, in which the reactive species from the plasma efficiently mix with the water droplets, advancing their chemistry to useful intermediates and becoming more readily available for delivery to the substrates. The observed droplet size distribution at the exit of nebulizers follows a bimodal trend. The first peak falls within the descending slope of the accumulation mode, with median diameters typically ranging from 0.85 to 2.73 µm, depending on factors such as nebulizer design and operational parameters. The second peak is due to a small fraction of coarse droplets (> 5 µm) that can emerge because of droplet coalescence or inefficiencies in atomization [45]. Notably, the employed device does not wet the treated products, as the microdroplets are readily absorbed or quickly evaporate.

Chemical characterization of the condensed PAFs revealed measurable concentrations of RONS, such as H₂O₂, NO₂⁻, and NO₃⁻, which are key contributors to their antimicrobial efficacy [25, 46,47,48]. These RONS play pivotal roles in disrupting microbial cell walls, suggesting that the plasma system produces an adequate content of bioactive species for pathogen inhibition [49]. Importantly, the production of RONS is overwhelmingly displaced towards nitrates. This result could be due to the relatively high temperature, although it is not comparable to warm plasma sources such as microwaves or gliding arc reactors, which are observed in the plasma, and to an accumulation effect taking place because of the PAF collection methodology that we utilized. In fact, to completely dissipate the water reservoir of 6 mL, PAF is condensed in an ice-cooled glass recipient over an ~10-min period, which could lead to the accumulation of excess nitrates from the gas phase into the condensed liquid phase as well as to an advancement in the RONS chemistry towards nitrates, i.e. the most stable species. This characterization is preliminary and does not accurately represent the PAF chemical composition, which should be assessed in future studies with analytical techniques capable of investigating aerosols (e.g., FTIR and Raman spectroscopy). Decreasing the mean energy dissipated per burst resulted in the same absolute production of nitrites and hydrogen peroxide, but the fraction of nitrates over the total produced RONS was significantly reduced, suggesting that the discharge regime shifted to the “ozone mode”, as lower temperatures are reasonably achieved in the reactor [50]. The high acidity (pH = 1.9–2.7) of the solution may play a role in enhancing the antimicrobial activity, but the exact contribution of the acid pH or its interaction with RONS must be further investigated. Fungi can adapt to environmental pH to modulate their pathogenetic activity [51], demonstrating a lower dependence on pH changes for survival. This suggests a predominant role of RONS in decontaminant efficacy [25], but accurate physicochemical characterization of the produced PAF is necessary.

The study revealed that all the tested fungal species presented a statistically significant response in terms of sensitivity to PAF treatment, with fungal species-specific inhibition rates dependent on exposure time, the plasma feeding gas and the production of aerosols. The highest inhibition rates were observed with the use of ambient air with water (PAF-DAW or PAF), with a mean efficacy of 60.4%, which was significantly different from that of the other treatment conditions (P < 0.05). Additionally, longer plasma exposure times were directly correlated with increased inhibition rates, with 91.3% treatment efficacy after 10 min of PAF exposure. Different levels of PAF efficiency were also revealed among the different fungal species analysed in in vitro experiments, with M. fructicola being the most sensitive fungus under all the investigated treatment conditions, followed by P. italicum and B. cinerea. The variation in treatment efficacy among fungal species likely reflects differences in their cellular structure and composition, with A. carbonarius and A. alternata exhibiting the highest resistance, as was also proven via direct plasma application [17, 18]. These results revealed the antifungal efficacy of PAF, with almost complete inhibition of fungal conidia for all the systems used for plasma application and for most of the fungi after 3 min and 5 min of treatment. Longer PAF exposure times or multiple applications of PAF may be needed for more resistant species (such as A. carbonarius and A. alternata).

The observed antifungal effects are exerted primarily through the action of RONS generated during the plasma discharge [25, 52]. These species disrupt fungal cell membranes, induce oxidative stress, and impair intracellular functions, including mitochondrial activity and redox homeostasis. Studies have shown that PAW can inhibit fungal spore germination, reduce metabolic activity, and suppress the synthesis of mycotoxins [52, 53]. The lower efficacy of PAF treatments against fungi such as A. carbonarius and A. alternata is likely due to the higher content of melanin in spores than in other species, which makes them more resistant to oxidant-mediated damage [18].

In postharvest applications on fruit, PAF treatment demonstrated good efficacy in reducing the incidence of rot in strawberries and table grapes.

For table grape, 10 min of PAF exposure significantly reduced the percentage of rotten berries up to 79.5% in the cultivar ‘Superior’ after refrigerated storage, which was greater than that in the cultivar ‘Attica’. Despite the higher incidence of the disease (24.3%), better treatment efficacy was recorded for the ‘Superior’ cultivar. This difference may be attributed to the difference in the performance of the two cultivars during shelf storage, which is related to their skin thickness; the black grape ‘Attica’ presents a thinner skin. Treatments were less efficient for long shelf-life periods, both in untreated and treated samples. In the case of the ‘Autumn Crisp’ and ‘Autumn King’ cultivars, the treatment efficacy was maintained at 5–10 min of PAF exposure and across various packaging methods of grapes before treatment, with the open tray showing better disease control for both cultivars, with similar efficacy in the case of unpacked grapes for the ‘Autumn King’.

For strawberry, 10 min of PAF treatment showed satisfactory efficacy in reducing fruit rot. This demonstrates the activity of RONS in the plasma phase, which enriches the fog, making it active. Overall, PAF treatments effectively reduced postharvest rot across both strawberry and table grapes, highlighting the potential of PAF as a versatile postharvest treatment adaptable to different fruit types. However, the degree of effectiveness may depend on the cultivar, growing and storage conditions.

This work also demonstrated that PAF can effectively reduce pesticide residues on table grape and strawberry fruits, according to previous reports of plasma treatments [29]. Notable reductions in insecticide and acaricide residues such as abamectin (up to 97.8%) and acetamiprid (39.7%) were detected in artificially contaminated grapes, and spinetoram (68.6%) and bifenazate (68.3%) were detected in field-contaminated strawberries. Fungicide residues were also reduced, although to a lesser extent, by the treatments, with a maximum decrease of 36.4% for dimethomorph on field-contaminated grapes and 38.6% for penconazole on strawberries. This difference in effectiveness could be attributed to the different complexities of pesticide molecules, which may expose more functional groups to react with PAF, likely through oxidative degradation. Additionally, the distribution of pesticide residues on fruit surfaces, which are affected by interactions with the plant and by environmental conditions, can impact their accessibility to PAF, which acts primarily as a surface treatment. These factors are particularly relevant in field-contaminated samples, where the timing of application and field variability further contribute to inconsistent degradation. Further study of the complete fragmentation patterns of the treated molecules could help elucidate the reaction mechanisms. Nonetheless, these reductions indicate that PAF could be a valuable tool for reducing pesticide residues on fruits, contributing to safer food products.

The antifungal efficacy of PAF in this study makes it competitive with other plasma-based technologies. Direct cold plasma treatments have demonstrated strong antimicrobial activity against postharvest pathogens such as B. cinerea and Penicillium spp., but often require direct exposure and longer treatment times, which can limit scalability and risk surface damage [19, 54, 55]. Although PAW has been successfully applied for antimicrobial decontamination and pesticide degradation on fruits [56,57,58], its efficacy against microbes may be lower and influenced by more variables (e.g. different methods of water production, PAW storage and stability) and generally requires longer treatment times [59,60,61,62]. The use of PAW on fruits postharvest might be a limitation in the potential industrial application of the technology, since producing adequate liquid volumes to uniformly wash the products might prove challenging. Moreover, some commodities should not be wet to prevent fungal postharvest contamination and fruit spoilage. In contrast, PAF combines the advantages of indirect plasma application with efficient aerosol delivery, ensuring homogeneous distribution of reactive species without thermal or mechanical stress on the fruit surface.

Compared with other plasma-activated aerosol systems, such as those generated via surface acoustic wave nebulization [30, 31, 33], which primarily involve surface disinfection and bacterial inactivation, our PAF approach demonstrated superior antifungal efficacy on fruits and additional benefits, such as pesticide residue reduction. While aerosolized hydrogen peroxide systems enhanced by cold plasma [34,35,36,37] have shown strong bactericidal effects, they require chemical additives and longer dwell times, whereas PAF achieves comparable or better results without added chemicals, ensuring safer and more sustainable postharvest applications.

These findings position PAF as a competitive alternative for postharvest disease control and pesticide residue mitigation.

Conclusions

PAF has emerged as a highly promising and environmentally sustainable technology for addressing postharvest challenges in fruits. In this study, our approach completely inhibited fungal growth and reduced fruit rot, improving food safety and extending the shelf life of fresh products, while also significantly lowering pesticide residues. These results demonstrate the potential of PAF as a non-invasive alternative to conventional methods, ensuring uniform RONS delivery without thermal damage. Our findings highlight effective treatment times and modes of application of PAF for different commodities and for different aims. Nonetheless, further studies are needed to fully explore its effects on different produce and its potential scalability. Future research efforts should focus on detailed physicochemical characterization of PAF, including real-time aerosol composition and RONS speciation via FTIR, Raman spectroscopy, and mass spectrometry, with the aim of elucidating the mechanisms of antifungal action and pesticide degradation pathways. From a practical perspective, optimization of plasma power and gas flow and composition is needed to maximize the efficacy while minimizing energy consumption. Pre-industrial trials should evaluate scalability through continuous-flow or modular systems, integration into existing packing lines, and cost-effectiveness analyses. Additionally, post-treatment studies on fruit quality, organoleptic properties, and regulatory compliance are essential to support industrial adoption. These steps will ensure that PAF can transition from laboratory validation to a robust, economically viable solution for large-scale postharvest applications.