Cold atmospheric plasma source for surface sterilization: design and simulation

Abstract

Plasmas have several applications in industry, mainly in surface treatment. This work aims to apply the properties of cold atmospheric plasmas for the sterilization of surfaces, objects, and environments. For this an atmospheric plasma source has been developed using commercial-off-the-shelf (COTS), which works with high frequencies, on the order of KHz, utilizing DC-pulsed signals. This approach allowed for a low-cost prototype that can be utilized in classrooms of physics, chemistry, and biology courses. Simulations were made to evaluate the circuit behavior and then compared with the prototyped circuit. This looped interaction allowed for fine-tuning of the simulation parameters, in order to achieve a similar result between simulation and bench measurements. Freecad $0.21$, an open-source CAD software, was utilized to design the structure of the prototype, with several of its parts being 3D printed, in both filament and resin printers. Preliminary results of the biocide action using Peribacillus simplex demonstrate that the plasma source has a biocidal capability after 10 minutes of exposition.

Keywords:
Cold atmospheric plasma; high voltage power supply; plasma sterilization; DC-pulsed circuit.

1.Introduction

Plasmas, often referred to as the fourth state of matter, have a wide range of applications, spanning from metallurgy to semiconductors, and from space to biomedical applications. Their unique properties, such as the ability to generate reactive species and ionize gases, make them an ideal tool for addressing contemporary challenges in science and education.

In the Plasma Physics Laboratory (LFP/UnB), at the University of Brasilia, we study and develop several plasma technologies, such as Hall Effect electric thrusters, Pulsed Plasma Thrusters, Ambipolar thrusters, plasma etching, thin film deposition, complex nanostructures manufacturing, specifically Plasma Enhanced Chemical Vapor Deposition (PECVD) of carbon nanotube. One of these fields involves using cold atmospheric plasma to eliminate biological contaminants.

Since the recent SARS-CoV-2 pandemic, new and affordable ways to sterilize surfaces, objects, and environments have been researched. To this end, there are several mechanisms, such as ovens, antiseptics, disinfectants, autoclaves, and ultraviolet, among others [¹, ²]. Recent studies apply Cold Atmospheric Plasmas (CAP) in surface treatment [³], PFFF3 masks [⁴], as well in wounds, dressings, and medical equipment [⁷, ³, ⁵, ⁶, ⁸, ⁹]. But plasma applications can spread much further, being used in food preservation processes [¹⁰], textile manufacturing, and cosmetic applications [¹¹].

Plasmas can exist in ambient pressure, high or extremely low-pressure ambiances, being, for example, used in satellites electric thrusters. In ambient and high-pressure environments, the plasma reactor can be open to the atmospheric air, or made of a closed chamber. It can utilize different gases or atmospheric air, usually utilizing high-voltage potentials to ionize the precursor gas. Mixtures can be made, with atmospheric air and other gases, to facilitate ionization and plasma generation, such as a mixture of pure hydrogen, helium, or argon gas with atmospheric air [¹², ¹³]. This process can utilize other gaseous combinations [¹⁴].

Rarefied atmosphere, or extremely low-pressure plasmas, requires a closed chamber and a low-pressure environment to ionize a specific gas, like hydrogen peroxide, occupying the entire internal space of the chamber due to diffusion. Although in atmospheric process plasmas also spread due to diffusion, the higher particle density affects the ions and electron’s mean path, collision frequency, and recombination process, altering the volume in which the plasma byproducts can reach, with highly reactive species having a lower reach than the more stable plasma byproducts.

This work focuses on Cold Atmospheric Plasmas (CAP), an atmospheric process. The ionization of atmospheric air produces reactive species of nitrogen and oxygen (RNS and ROS), which are capable of eliminating bacteria and fungi and deactivating viruses [¹, ², ¹⁵]. Its effectiveness depends on the duration of exposure to plasma [⁷].

1.1. General applications of Cold Atmospheric Plasmas

Cold Atmospheric Plasmas can be employed in a number of ways. In the past decade, among several streams of applications, studies were made in biomedical applications, exploring the plasma potential. But CAPs are not limited to health-related uses, with ongoing research utilizing CAPs in the textile industry to give hydrophobic or hydrophilic characteristics to fabrics, in the food industry in order to substitute the pasteurization process in the conservation of puree, or as an anti-mold agent in strawberry crops [¹⁰].

But in the biomedical field, CAPs are used in cosmetic treatment, with plasma jet skin treatment [¹¹], in odontology as an auxiliary tool for root canal treatment [¹⁶], in-loco targeted delivery of cancer medication [¹⁸, ¹⁷], wound healing in chronic or burned patients [⁹, ²⁰, ¹⁹], as an effective tool against resistant pathogens [⁷, ²¹], among others.

Generally CAP source works by means of an electrical field between two electrodes, which may, or not, be surrounded by the working gas. This electrical field can be of continuous nature (DC), or an alternating signal (AC). From this, we can classify the vast majority of CAP sources in plasma jet, dielectric barrier discharge (DBD), and spark (or arc in some cases) [⁸]. Figure 1 brings a simplified diagram of these three CAP sources.

Figure 1
Simplified diagram of common CAP sources. From left to right we can see three ionization methods, being a plasma jet, spark, and dielectric barrier discharge. The electrode polarization depends on the characteristics of the electric field needed for atmospheric air ionization.

1.2 Sterilization processes

The inactivation of biological contaminants using plasmas occurs through several mechanisms, such as free radicals, reactive nitrogen and oxygen species, electric fields, charged particles, and ultraviolet (UV) photons [²², ²]. Each one can have a direct, or indirect, effect on the surface alongside any biological contaminants. CAPs can also be used in the process of making plasma-activated water, whereas the reactive byproducts of atmospheric air ionization are diluted in aqueous solution, retaining some interesting properties.

For instance, free radicals and reactive species can alter the biochemistry of cells, such as membrane lipid peroxidation, plasmid DNA damage, chromosomal DNA/RNA damage, induced oxidative stress, protein denaturation, amino-acid oxidation, or be dissolved in a liquid solution to be later used as a disinfectant [²², ²]. These reactive species created in atmospheric plasmas can be atomic oxygen, hydroxyl radicals, hydrogen peroxide, superoxide, and ozone.

The reactive species in CAPs can be: atomic oxygen, hydroxyl radicals, hydrogen peroxide, superoxide, and ozone. They interact with the liquid medium and can create different species in the plasma, plasma-liquid interface, and liquid medium.

According to Ref. [²], among the chemical reactions in CAPs for the production of reactive species, we can highlight:

• Atomic Oxygen: $M^{\ast }+O_2\to M+2O,(M:He,Ne,Ar)$ $M_2^{\ast }+O_2\to 2M+2O$ $e^{-}+O_2\to O+O+e^{-}$

• Hydroxyl Radical:$M^{\ast }+H_{2}O\to M+H^{\bullet }+O^{\bullet }H,(M:He,Ne,Ar)$ $M_2^{\ast }+H_{2}O\to 2M+H^{\bullet }+O^{\bullet }H$ $e^{-}+H_{2}O\to O^{\bullet }H+H+e^{-}$ $UV+H_{2}O\to H_2{O}^{\ast }$ $UV+H_2{O}^{\ast }\to H^{\bullet }+O^{\bullet }H$ $H_2{O}^{\ast }\to O{H}^{-}+H^{+}$ $O{H}^{-}\to O^{\bullet }+e^{-}$

• Hydrogen Peroxide:$e^{-}+O_2\to O_2^{-}$ $O^{\bullet }H+O_3\to H{O}_2+O{H}^{-}$ $O_2^{-}+H_{2}O\leftrightarrow H{O}_2+O{H}^{-}$ $OH+OH\to H_2{O}_2$ $H_2{O}_2+OH\to H{O}_2+H_{2}O$

• Ozone:$O+O_2+M^{\prime }\to M^{\prime }+O_3$

At the Laboratório de Física dos Plasmas (LFP/UnB) a dielectric barrier discharge was obtained from atmospheric air utilizing a Neon Transformer, at 16.3 kV and 60 Hz. In that experiment, the biocidal action was characterized using spores of Geobacillus stearothermophilus, with no remaining viable spores detected after 40 minutes of plasma exposition, in a plasma process of a rarefied atmosphere (between ${10}^{-5}$ and ${10}^{-4}$ Torr) and with nitrogen ($N_2$) as working gas [¹]. Nowadays, the atmospheric plasma source utilizes a switched DC circuit to excite a fly-back transformer and obtain air ionization, operating at 5 kHz and achieving around 5 kV.

In this paper we present details of this CAP source, utilizing DC switching topologies and low-cost easily available components. The focus of this work is the analysis and comparison between numerical simulations and electronic circuit measurements of this CAP source. The free software LTSPICE XVII was used in the circuit simulations, and a Hantek digital oscilloscope, armed with a Tektronix P6015A high-voltage probe, was utilized to acquire a signal from the fly-back transformer output. A key design parameter of this system is its low-cost, simple-to-use, and easily available materials.

1.3 Educational potential of CAP technology

This kind of plasma technology has great potential to give students a hands-on approach to electronic circuit topology and functioning, plasma physics, biology, and chemistry, among other interdisciplinary topics.

For instance, gas diffusion can be treated as the movement of molecules in order to fill every corner of a closed space. However, in open spaces, this process does not reach the outer boundaries (depending on the size of the room and the power applied in the ionization process), and as the ionized gas spreads, it gets thinner, rarefied, having a maximum distance of occurrence, named diffusion halo.

The plasma-liquid interface is the region between the two media, analogous to the interface between water and oil in a glass cup, for example, or rather, the dissolving front of a sugar cube in coffee. In this last case, there will be a thin interaction area delimited by the molecule’s agitation state, which is useful for studying the imagetic properties of matter. Back to the plasma context, in this region compressed between the ionized gas and the water surface, some molecules can interact with each other, giving room for chemical reactions in this small space. These reactions can generate other molecules or dissolve some compounds within the aqueous solution.

Reactive species are by-products of atmospheric air ionization, that has a higher tendency to participate in chemical reactions. Reactive species of Oxygen (ROS) and of Nitrogen gases (RNS) can interact with the biological structure in several diverse ways. For instance, some reactive species are known as free radicals, responsible for cell oxidation. Others can rupture the cell membrane. These molecules can act directly on the biological contaminant or be dissolved in aqueous solution for later use as a disinfectant.

Biological contaminants are the main focus of this work, and can be bacteria, virus or fungus. In other applications of CAPs, these can also include living tissue, such as the skin in cosmetic applications, or oral tissues in dentistry, even cancer cells and chronic wounds, such as skin cancer and diabetic feet wounds.

The main force driving the prototype, described further ahead, involves electric arcs and the ionization through dielectric rupture. This arc regime behaves similarly as thunderstorms, were electric charges accumulates until its electric potential can rupture the atmospheric air, ionizing it and forming a stream of plasma. The stored charges can then move through this low-resistance pathway. This prototype can be used as a means of showing miniaturized “thunders” inside a class room, demonstrating experimentally some concepts as dielectric and conducting materials, air resistance, the influence of air humidity or the distance between electrodes in these electric arcs.

It can also be used to demonstrate concepts involving electromagnetism, such as wireless energy transfer, by utilizing fluorescent bulbs above the discharge, right hand rules when teaching induced current and magnetic fields, by utilizing a ring shaped magnet around the discharge electrode.

As a by product of the ionization of atmospheric air, ozone is produced, giving a distinctive smell to the discharge. In chemistry and geography classes, this device can be used to demonstrate the formation of ozone molecules high above the atmosphere due to the interaction between solar radiation and atmospheric oxygen.

2. Experimental Setup

Our experiments had two different steps. First, electric measurements and waveform acquisition were done in LFP’s electronic workbench. Once the circuit was operating as expected, the prototype was brought to the UnB Microbiology Laboratory, where its biocidal capabilities were evaluated against Peribacillus Simplex SDF0016 biological contaminants.

During the characterization of the circuit, the following materials and equipment were utilized:

• Prototyped circuit on a universal perforated printed circuit board;

• DC bench power-supply, model HIKARI HF 3203S, capable of outputting $32volts$ at $3amperes$;

• Digital storage oscilloscope, model HANTEK DSO2D15, with $1GSa/s$ and $150MHz$ of max bandwidth;

• General purpose oscilloscope probe, model HANTEK P6100 x1/x10 $100MHz$ CAT II;

• High-voltage oscilloscope probe, model TEKTRONIX P6015A x1000 $20kV$ DC or $40kV$ Pulsed, and $75MHz$ of max bandwidth;

• Infrared thermometer, model KLX GM320, $-{50}^{\circ }C$ up to ${400}^{\circ }C$;

• Infrared digital camera, model UNI-T UTi120S;

• Bench digital multimeter, model HP 34401A;

• Soldering iron, model Hikari HK-930;

• Air quality monitor, model 2CO9, with ambient temperature and humidity measurements;

• Set of pliers and screwdrivers.

In this characterization, several components were monitored with both infrared thermometer and thermal camera, with different power levels and operation times, in order to ensure safe operational levels for the power components. During this period, after one hour of continuous operation, with a maximum supply power of around 38 watts (being $25volts$ and $1,5amps$), the power MOSFET was stable between ${70}^{\circ }C$ and ${80}^{\circ }C$, while the rest of the circuit hovered around ${40}^{\circ }C$. At this time there was no forced air flow onto the aluminium heatsinks, and ambient conditions at the time were ${28}^{\circ }C$ at 53% humidity. Adjustments in the simulation models were made after comparing the waveforms obtained during these tests, with the simulation results.

Then, at a later date, the system was brought to the Microbiology Laboratory, at the Institute of Biological Sciences, where the biological properties of the corona discharge were tested. There, alongside thermal sensors and the DC bench power supply, the following materials and equipment were utilized:

• Prototyped circuit and Plasma Wand;

• 9 cm in diameter plastic Petri dishes;

• $40mL$ of LB medium with $1.2\%$ agar, per Petri dish;

• $10\mu L$ of ${10}^{\text{6}}cells/mL$ of Peribacillus simplex SDF0016 [²⁴], in log phase, per Petri dish;

• Incubator, model THERMO SCIENTIFIC MaxQ 6000;

• Biological Safety Cabinet;

• Set of laboratory equipment.

After inoculation, the dishes were exposed for different times, of 1, 3, 6, and 10 minutes. A distance of $50mm$ was set for all exposures, and each time was done in duplicate. Afterward, all dishes, including both control ones, were incubated at ${30}^{\circ }C$ for 24 hours.

During testing with the biological contaminants, the following steps were done, with a few alterations in time of exposure, or incubation time, but with the same steps:

1. The room and prototype were sterilized with 30 minutes of exposure to UV radiation. All work surfaces and counters utilized were cleaned with $70\%$ alcohol;

2. Inside the biological safety cabinet, we produced the Petri dishes with a new, and sterile, disposable petri dish, filled with $40mL$ of LB medium with $1.2\%$ agar, per Petri dish;

3. Inside the biological safety cabinet, we have inoculated all the dishes with the biological contaminant, being in the first test, $10\mu L$ of ${10}^{\text{6}}cells/mL$ of Peribacillus simplex SDF0016 [²⁴], in log phase, per Petri dish. The following tests after were done with $200\mu L$ of ${10}^{\text{8}}cells/mL$ of Peribacillus simplex SDF0016 [²⁴];

4. The samples, and their duplicates, were organized on a clean counter, with the control samples and test dishes separated in collums according to the exposure time of the test;

5. After sample preparation, the room, and biological safety cabinet, were again sterilized with 30 minutes of exposure to UV radiation;

6. The first samples were put in the specific location on the prototype, that was already inside the biological safety cabinet, and then the discharge was activated for the duration of each test. Was common to start first with the longer duration exposures, moving in a decreasing manner through the different exposure times.

7. At the end of each exposure time, the dish’s lid was put back on, and each set of duplicates was stored on a clean counter;

8. At the end of the last exposure test, all dishes were brought to an incubator, and kept there for 24 hours, at a temperature of ${30}^{\circ }C$;

9. At last, after 24 hours of incubation, all samples, including their duplicates and the control dishes, were photographed and analyzed;

10. The disposal of these dishes was done in proper trash containers, being incinerated afterward.

2.1 Safety precautions

When working with biological contaminants, and with high-voltage circuitry, there are some precautions that has to be followed, in order to ensure the safety of everyone involved.

First, we’re gonna to look at safety precautions concerning high-voltage circuits. With every experiment utilizing electronic circuitry, either low or high voltage, a set of preparations has to be made. The prototype has to be on a dry surface, preferably made of an insulating material, away from liquids and moisture. The operator has to keep its distance from exposed circuitry and remove rings, bracelets, necklaces, long earrings, and any other conductive adornment that can accidentally fall, or make contact, with the electric circuit. Prefer to utilize cotton fabrics and shoes with rubber soles.

With the high voltage end of the prototype, the precaution has to be doubled. Although this kind of output power shouldn’t be fatal, or cause some sort of persistent damage, it is always best to be on the safe side. For this, the one-hand rule is paramount. When interacting with parts of the prototype, when it’s turned on, always keep one hand close to your body, for instance in you pocket, and utilize only the other hand to interact with the live circuitry. Also, elevate one foot (the same side as the hand in the pocket) with a dielectric material, like a cardboard box or a block of wood. This is done to discourage the electric current from traversing through your heart, by giving a direct path to the ground (for instance, keeping your left hand in your pocket, and elevating the left foot, the electric arc will prefer to go through your right hand and right foot, straight to the ground, without crossing your chest).

Respect the operating boundaries of the circuitry and heat dissipation. Some elements of this circuit dissipate a lot of heat, requiring the use of heat sinks. Touching these components during operation, or immediately after, can cause burns. Also, inappropriate heat dissipation can cause the component to fail. This prototype was developed to work with a certain input voltage and current levels, but it can handle higher values, for a short amount of time. Higher voltages and currents will impact the output voltage, giving space to arcs inside the flyback transformer, and higher heat generation, due to higher currents moving through the drive circuit. Either way, these operating conditions can impact drastically the longevity of the components. When operating within the parameters described in this work, the prototype can function for hours at a time, without any incident.

Biological safety precautions will encompass two main focuses, first the safety of the operator, second the safety of the samples, to avoid external contamination during testing.

It is paramount that the operator utilizes individual protective equipment, such as safety goggles, surgical gloves, and a lab coat. Depending on the type of biological contaminants, further protection is needed. All waste generated in these procedures has to be properly treated, mostly as bio-hazard waste. All work surfaces have to be cleaned with disinfectants, both before and after the operation. Also, proper instructions have to be given to the operator, to avoid own contamination, for instance scratching your eye during sample inoculation. The contaminants utilized in this study were not pathogenic but still could cause eye inflammation, diarrhea, gastric discomfort, and nausea, among other symptoms. After incubation and result analyses, all dishes were properly disposed of, being collected in specific containers, and incinerated.

With proper care and handling, these experiments are safe to conduct.

3. Decontamination Prototype

Several prototypes were built during the development process of this plasma source. Figure 2 brings the CAD design of the CAP source prototype, Figure 3 shows a detailed view, and Figure 4 shows the version constructed and analyzed in this work. It’s comprised of two plasma wands constructed with $5mL$ disposable syringes, copper electrodes, and dielectric materials such as polyacetal, aluminium oxide, and PLA 3D printed parts. Threaded rods were utilized as structural support for the system. Numbers were engraved at the circumference of the Petri dish holder, with ${90}^{\circ }$ separation, for repeatability, indicating which quadrant of the substrate was treated, or its spatial orientation regarding the high-voltage electrode.

Figure 2
CAD model of this prototype iteration. This design was done in the open software FREECAD 0.21. The base, in gray, was printed with PLA filament. The white dielectrics were machined in polyacetal.

Figure 3
A closer view of the copper electrode inside the disposable syringe. Here we can observe its cylindrical geometry and crown shape structure at one side.

Figure 4
Experimental setup of the CAP source. In this iteration, two Petri dishes could be exposed simultaneously. Dielectric materials (Aluminum Oxide, Teflon, Polyacetal, rubber) were used to isolate the high-voltage line from exposed structural parts made out of metallic materials.

After a few iterations, a crown shape electrode was chosen. The goal was to disturbed as little as possible the flow of atmospheric air inside the disposable syringe, whilst allowing a surface area large enough for efficient ionization. Utilizing a copper foil, with a thickness of $0.1mm$, the shape was cut with scissors, and this template was rolled, and soldered, into a cylindrical geometry with 16mm of external diameter.

As a research prototype, it was necessary to make adjustments in certain parameters, in search of optimal operational conditions. This led to some decisions, such as utilizing threaded rods, which, with a pair of nuts and pressure washes, allowed us to make fine adjustments on the distance between the agar surface and the active electrode.

Another important feature to be pointed out is the need to electrically connect the agar’s surface to the fly-back transformer’s reference electrode. Since the nature of the discharge in this prototype was DC, without grounding the samples, electrostatic charges would rapidly accumulate, repelling the flow of plasma and its by-products electrically. This grounding was achieved by connecting alligator clips to the edge of the samples.

Petri dishes (9 cm) containing 40 mL of Luria-Bertani agar (1.2%) medium were inoculated with 100 $\mu$L of ${10}^{\text{8}}$ log-phase cells of Peribacillus simplex SDF0024 [²⁴]. The samples were exposed to the plasma flow of the corona discharge for 1, 3, 6, or 10 minutes. A fixed distance of 50 mm was maintained between the active electrode and the agar plate surface. All treatments were conducted in duplicate. Figure 5 shows the corona discharge of atmospheric air. The image shows the plasma column reaching down to the agar below.

Figure 5
Atmospheric plasma produced by the CAP prototype. Both plasma wands were active, hitting the agar surface. These exposures were made inside a biological chapel, to avoid exposure to other contaminants.

3.1 Assembly instructions

In order to assemble this prototype, as seen in Figure 2 and in Figure 4, a few items will be needed. Some can be 3D printed, others can be substituted for more convenient materials. All the 3D printable files can be downloaded at github.com/HelbertJunior/CAP-Prototyope-Beta. The complete list can be seen in Table 1.

Table 1
Components list for hardware assembly. The cost may vary from your location and material choices, so it was left out of this list.

Part	Qtd.
M5 Threaded Rod with 25 cm in length	2
M5 Threaded Rod with 7,5 cm in length	2
M5 Hex Nut	12
M5 Washer	10
Plastic Feet (drawer knob or pull handle)	4
Petri Dishes (glass or disposable ones)	2
5 mL disposable syringes (or 3D printed)	2
Aligator clips	2
3 cm $\times$ 5cm Copper foil (or aluminium foil)	2
Dielectric connector (machined or 3D printed)	2
Dielectric tube (alumina or 3D printed)	2
Electric tape	–
Electrical Wires (with 1000V isolation)	3m
Electrical Wires (can be Ethernet cables)	3m

A few items can be exchanged with available ones, such as the plastic feet, high-voltage wires, ground wires, and alligator clips. For instance, the high-voltage wires can be exchanged with CAT III multimeter wires or 2.5mm house electrical installation wires. The ground wires can be obtained from Ethernet cables. Bottle Caps can be utilized to make the insulating feet. All threaded rods can be exchanged with barbecue sticks.

3D printed files for the insulating parts were provided at the link above. When printing the insulating parts, as seen in detail in Figure 11, make sure to configure your slicer software for 100% infill. Since these parts will be interfacing with high-voltage potentials, it is paramount that their interior has the few vacancies as possible. Dielectric air inside these structures will reduce their insulating capacities.

Figure 6
Step 1, assembly of the longer threaded rods.

Figure 7
Step 2, 3D printed base with both threaded rods attached. Hex nuts and washers hold both rods from the upper and lower sides of the base.

Figure 8
Step 3, a set of hex nuts and washers inserted onto both rods, in preparation to receive the dielectric connector.

Figure 9
Step 04, detail of the crown shape high voltage electrode. This can be substituted with a thin wire or a box cutter blade.

Figure 10
Step 05, detail of the modified syringe with the electrode inside.

Figure 11
Step 6, assembled prototype with the dielectric connector. If 3D printed, utilize a 100% infill.

For assembling the prototype, first cut the threaded rods to size (if not purchased already in size). This can be done with a hacksaw or other cutting tool. Put a set of hex nuts and washers at the end of both 25 cm rods, as seen in Figure 6.

Insert both 25 cm rods into the outermost holes, securing them with a set of hex nuts and washers on the other side. Figure 7 illustrates this step. It is important to make a tight connection, but not to tighten both nuts (up and bottom) to the point that breaks the 3D printed base. Loctite or other thread lock can be used here to ensure a sturdy connection.

Next, put a set of hex nuts and washers on both threaded rods, from the other extremity, first the nut followed by a washer. The distance of this set from the top of the 3D printed base has to be around 10 cm (it can also be adjusted later), as seen in Figure 8.

Continuing the assembly process, cut the copper foil to make the electrode in a crown-like shape, as seen in Figure 9. There isn’t a strict rule as how this electrode has to be, only a few characteristics have to be obeyed. It is important not to block completely the flow of atmospheric air inside the syringe, and to have a thin edge, or several points, oriented towards the exit point of this ionization chamber. As a substitute, a thin piece of wire wrapped in a circle, a cutout from an aluminum soda can, or a piece of the blade from a box cutter can be used.

Make a 7mm diameter hole on the top part of the syringe. This will be necessary to insert it, at a later stage, into the supporting threaded rod. If 3D printed, the file provided already has this feature. Also, the disposable syringe can be used as is, with a smaller diameter output, or it can be removed with a box cutter, in order to enlarge the output diameter. It is a modification that improves ionization, the area of sterilization, and visualization of the discharge. These modifications can be seen in Figure 10.

A piece of wire, of around 1.5 m in length, and with the insulating capability of at least 1000V, can be soldered at the end of the electrode (at the opposite side from the points or sharp edge). This will be the High-Voltage electrode. Inserting these parts (electrode $+$ wire) inside the disposable syringe, you can utilize a piece of electrical tape to attach the wire, and electrode, in the desired position. For this, it’s important to obtain a distance of between 20mm and 30mm from the tip of the high-voltage electrode to the exit point of the syringe, as illustrated by Figure 10.

Put the dielectric connector in the threaded rod, which was previously attached to the base, securing it tightly with two sets of hex nuts and washers. Figure 11 shows how it will be after this step. If 3D printed, this connector has to be configured to 100% infill. Any cavities inside this piece will result in a lower insulating capability when dealing with high-voltage electric arcs inside the structure.

Insert the smaller threaded rod into the center hole of the dielectric connector, as seen in Figure 12. In this step, you can make a threaded tap on the connector, allowing the rod to be fastened like a screw. Or, by utilizing a two-part epoxy adhesive, this threaded rod can be secured tightly. A hex nut can be used as a finishing piece, as well as to ensure no slacking due to transportation.

Figure 12
Step 7, detail of the smaller threaded rod attached to the dielectric connector. This can be done by tapping threads into the connector hole, or by gluing the threaded rod with a two-part epoxy adhesive. A hex-nut can be fitted to secure this connection.

Put a few layers of electrical tape on the smaller threaded rod, and insert the dielectric cylinder. It can be made from several dielectric materials, such as PVC, alumina, PET-G, and others, or it can be 3D printed (if chosen the latter, print it with 100% infill). Figure 13 brings this step.

Figure 13
Step 8, insulating material covering the threaded rod. Alongside electrical tape, a dielectric cylinder was used to guarantee isolation. Teflon tape can also be added, increasing electrical safety in case of undesired arcs.

Insert the plasma reactor assembly, comprised of high-voltage wires and an electrode, and the disposable syringe, into the dielectric cylinder, until it reaches the halfway point of the insulating material. Figure 14 illustrates this step.

Figure 14
Step 9, isolating assembly of the dielectric cylinder, followed by attachment of the disposable syringe with high voltage electrode.

Solder a 1.5m, in length, thin wire (can be a strand of an Ethernet cable) onto a 5cm to 7cm diameter disk (preferably copper). Then, pass this wire through the small hole at the base, leaving the metallic disk flush with the indentation bottom. A double-sided tape can be used here to hold it in place. The other side of the wire will be connected to the high-voltage ground of the flyback transformer. Add the feet and a pair of petri dishes, and the reactor is ready, as illustrated by Figure 15.

Figure 15
Step 10, prototype structure assembled, with copper ground planes and petri dishes.

4. DC-Pulsed Circuit Topology

The object of study in this work is a circuit capable of generating pulses, with its width modulated, or PWM (pulse width modulation) signal. Since our transformer of choice has diodes in its construction, it seemed logical to utilize this DC approach. By controlling the pulse frequency and duty cycle (width) we can alter the current profile on the transformer’s primary coil, thus changing its output high voltage.

The first step was to design an oscillating circuit capable of creating a PWM signal, and that allowed us, to some extent, to change its frequency and duty cycle. Also, in line with the low-cost and easy availability design parameters, this circuit needed to be simple and cheap. The pluripotent 555 timer integrated circuit (IC) filled all these criteria. Operating in the Astable mode, with a few components, we had our square wave source.

Next, we had to adapt its output signal to be capable of switching on and off a transistor. Due to its switching capabilities and high current applications, the decision was made to utilize MOSFET (Metal Oxide Semiconductor Field Effect Transistor) transistors. Therefore, it was needed to amplify the output signal from our 555 oscillator to meet the chosen MOSFET requirements. This was done with a simple push-pull amplifier comprised of three general-purpose transistors. A simple transient protection circuit is placed between the MOSFET’s drain and VCC. Then, it was a matter of fine-tuning the frequencý and duty cycle that produced the desired corona discharge within voltage and current limitations.

Figure 16 brings the circuit schematic, as well as some parameters for the numerical simulation in the LTSPICE XVII software.

Figure 16
Circuit schematic of the DC pulsed CAP source. This was the setup utilized in LTSPICE XVII for the numerical simulations.

A prototype was built utilizing a universal perforated board, with the facility to change each one of the transistors, and 555’s capacitors. This was done to quickly alter some parameters, such as frequency, power, duty-cycle, and modulation, in the search for optimal results. During this test regime, the dependence of these parameters and output voltage on the flyback’s secondary was very clear.

In the lower right part of Figure 16 there can be found the plasma discharge circuit model, which is based on a neon bulb. The DC-pulsed circuit is designed to generate high-voltage pulses with a specified frequency and duty cycle¹. This component on LTSPICE XVII behaves like a voltage-controlled switch, where Vstrike indicates at which voltage it starts to conduct, and Vhold indicates where this switch opens again and what is its residual voltage. L2 represents the secondary of the flyback transformer, with its series resistance representing losses. D3 rectifies our high-voltage signal, being present in its commercial counterpart. C3 has a dual purpose, acting as the electrodes’ capacitance and as an energy bank for the corona discharge. R5 limits the discharge current, and its relation with C3 stipulates the period of the discharge curve. These parameters were adjusted in order to recreate those encountered in the circuit prototype.

Figure 17 brings the circuit version utilized in this study. It was assembled on a Universal Perforated Board, with its components soldered to the board. A few key components were connected via sockets and headers to allow changes in different parameters. This allowed us to test our circuit limits, quickly replacing any damaged component, and to experiment with a vast range of frequencies and input power. One of these changes was the substitution of BC327 and BC337 general-purpose transistors for TIP 29 and TIP 30 power transistors in the push-pull amplifier, resisting harsher operation conditions and allowing for longer plasma applications without faults.

Figure 17
Circuit prototype studied in this work. It counts with simple MOSFET protection circuits, an off-the-shelf DC-DC step-down regulator, sockets, and headers for easy exchange of parts. A repurposed aluminium heat sink is responsible for keeping the MOSFET temperatures at bay, and a bourne connector supplies a computer fan with variable voltages, from 6 to 12 V.

4.1 Building materials

For a list of building materials for this prototype, please refer to Table 2. There it can be found all the electronic components, and hardware, necessary to achieve the functioning prototype showcased in this study, as well as its prices, as seen locally at the beginning of the year 2025.

Table 2
Components list for this study’s prototype.

Part	Qtd.	Unit (BRL)
10x15 Perforated board	1	24,45
CRT Flyback Transf.	1	29,90
DC PSU 25V 2A	1	85,72
10K lin. potentiometer	2	2,90
8 pin dip socket	2	0,25
NE555 timer	1	1,25
TIP29 NPN transistor	2	1,99
TIP30 PNP transistor	1	0,89
IRF540N mosfet	1	6,25
Heatsink	1	10,00
LM2596 stepdown	1	12,56
12V fan (optional)	1	14,80
2.54mm female header	1 strip	1,00
Through male header	1 strip	1,10
KRE connector 2 ports	3	0,89
KRE connector 3 ports	1	1,11
1N4007 diode	3	0,11
470$\Omega$ 1/4W res.	1	0,05
1k$\Omega$ 1/4W res.	1	0,05
2k2$\Omega$ 1/2W res.	3	0,10
100k$\Omega$ 1/4W res.	1	0,05
1nF 50V capacitor	1	0,07
22nF 50V capacitor	1	0,09
100nF 50V capacitor	1	0,11
1$\mu$ F 50V capacitor	1	0,12
22$\mu$ F 50V capacitor	1	0,19
Disposable syringe	2	0,38
M5 Threaded rod	1	4,70
M5 nut and washer	10	0,20
M5 pressure wahser	4	0,19
Rubber feet	4	0,8
3D printed parts	–	50,00
Total Cost	–	214,74

Other everyday materials are needed, such as cables, solder wick, electrical tape, and tools for assembling the prototype. For instance, older ethernet cables can be salvaged for it’s strands of copper wire.

5. Results

In this section, we’ll be able to see the real behavior of this prototype, and our simulations results. Comparing both waves it is possible to observe a similar behavior between the two waveforms, within an error margin for the parasitic losses.

5.1 Simulation results

Circuit simulations were done with the free software LTSPICE XVII. At first, the general circuit was simulated, in order to analyze current drawn from major components. Afterward, parasitic losses and a MOSFET protection circuit were added.

A simplified discharge arrangement, utilizing a neon bulb model, acted as the high voltage discharge, with both low current arc and corona discharge effects. This was refined with the addition of resistance and capacitance emulating the electrode’s characteristics.

The neon bulb in LTSPICE has a similar behavior as the discharge proposed by the Mayr Model [²⁵, ²⁶, ²⁷]. Testing in both the simulation and in the circuit, utilizing high voltage probes, showed a behavior similar between the simulation and the actual arc discharge. Since the Mayr Model has two parameters, Mayr time constant and relaxation power, which have no analytical solution since they depend on several electrode and arc conditions, the simulated model utilized the default parameters, with discrete components being adjusted to approximate the simulated with actual circuit operation. These values were later fine-tuned with the prototyped circuit measurements and oscilloscope waveform analysis.

The output of this simulation can be seen in Figure 18. When there is a pulse signal on the flyback’s primary coil, a high-voltage pulse is observed on its secondary coil. This, in turn, triggers the plasma discharge, allowing current to flow from one electrode to another. Since there are energy storage components in this circuit, we have that characteristic capacitor discharge curve, from $5.5kV$ down to $2kV$, when the cycle repeats itself Giving enough time, this discharge would reach zero volts. This cycle repeats approximately every $150\mu s$, for a frequency of around $5000Hz$.

Figure 18
Simulation results on LTSPICE XVII. Here we have the three circuit parameters analyzed and compared with the oscilloscope readings. The topmost graph brings the control pulses, generated by the 555 circuit, and amplified by the push-pull amplifier. These pulses had a frequency of 5 kHz and a duty-cycle of around 10%. The middle graph brings the ringing effect seen in the flyback transformer primary coil. And the bottom plot shows de capacitor discharge like characteristic of the plasma discharge.

In order to understand the plasma discharge characteristic curve, similar to a capacitor discharge, we have to analyze each step of the corona discharge.

First, a pulsed current energizes the primary coil of a transformer. Then, through oscillating magnetic field lines inside the transformer’s core, a current is induced in its secondary coil. Due to the opposing behavior to current change, the primary coil sees an opposing current emerging in order to counteract any changes in current. During this moment, the magnetic field inside an inductor collapses, creating large spikes in current (and in voltage due to Ohm’s Law). Since both primary and secondary coils are magnetically coupled, this behavior within the first coil is also transferred to the second one, leading to high-voltage spikes, proportionally to the number of coil turns in both primary and secondary.

When the voltage differential between electrodes is high enough, an atmospheric air dielectric rupture can be achieved, ionizing this gas and forming a lower impedance plasma channel. This channel works as a resistor connecting both ground and high-voltage electrode. Since an inductor has the ability to store energy, similar to a capacitor’s behavior, when this lower impedance plasma channel is formed, it can discharge its stored energy. The analog circuit is a Resistor-Inductor discharge (RL), similar to a Resistor-Capacitor circuit (RC), thus having that discharge curve seen in both simulations and oscilloscope readings.

Since the plasma generated is a dynamic entity, it is natural that its time-varying impedance wouldn’t sustain low resistance pathways indefinitely, leaving some energy behind on the parasitic components such as electrode capacitance and transformer inductance, especially due to the ringing effect.

Due to the primary ringing effect, there is still energy being transferred between primary and secondary coils, as well as between inductors and parasitic capacitance. Also, when a plasma channel is created, it has its momentum and characteristic behavior, with ions remaining still along the plasma path, even after the channel’s closure. That buildup of charges leads to a reduced dielectric rigidity, allowing for lower voltage arcs. This, in tandem with the ringing effect and continuous transferring of energy between capacitors and inductors, can lead to secondary and tertiary plasma discharges. That dynamic behavior has its caveats when simulating, therefore omitting the behavior of secondary discharges as seen on the oscilloscope. In fact, most mathematical models focus on specific parts of the discharge dynamics. This can be seen in the Cassie-Mayr models. In the LTSPICE software, the neon light bulb has a simpler model compared to the real plasma discharge as seen on the oscilloscope’s screen.

The voltage and current characteristics were analyzed during plasma operation. The CAP device demonstrated stable discharge behavior, with a peak voltage of $5kV$ and a steady-state current of $0.003A$. These results highlight the device’s ability to maintain consistent plasma generation, a critical factor for both research applications and classroom demonstrations of discharge physics.

5.2 Oscilloscope readings

The oscilloscope utilized for this data acquisition was a Hantek DSO (Digital Storage Oscilloscope) model DSO2D15, with two channels, 1 GS/s, and 150 MHz of maximum bandwidth. For Low-Voltage measurements, its original P610 probe was used, with 10x attenuation, and to acquire the Kilo-Voltage signal, a Tektronix P6015A was utilized, with 1000x attenuation and capable of measuring 20 KV signals (40 KV in short peaks).

The signals measured were the pulses from the 555 timer, after its amplification with the push-pull transistorized amplifier, the voltage waveform on the flyback transformer primary coil, and finally, with Tektronix P6015A, the high voltage output. These waveforms were compared with the simulation results, and then, through parameter estimation, component values on the LTSPICE circuit were adjusted until both behaviors, simulated and measured, achieved the same results. Once our simulations were fine-tuned, some changes in the driving circuit could be first simulated, with their results matching those of the prototyped circuit.

As seen on Figure 19, the MOSFET’s gate is excited by a square wave, with a frequency of around $5KHz$ and a duty-cycle of approximately $10\%$. Also, after amplification from the push-pull amplifier, that signal has a high-level of around $18V$.

Figure 19
555 signal generator circuit output, showing the square wave pulses that excite the IRF 530N MOSFET. Here we can observe a $20\mu s$ pulse, with a duty cycle of around $10\%$.

Moving to Figure 20, we can see two waveforms plotted, a yellow one representing the primary coil signal, and the green curve showing the high-voltage behavior. Here, in the yellow signal, it’s possible to observe the ringing effect in the flyback transformer primary, with an exponential decay and up to $120Vpp$. A similar analysis can be made for this exponential decay, since the energy transfer occurs through a medium that has parasitic losses. This lost energy is mostly converted into heat, a part into sound waves, and other means of parasitic loss.

Figure 20
Voltage waveform of both the flyback primary coil, in yellow, and secondary high-voltage output, in green. Here we can see the ringing effect, as shown also in the simulation, of the flyback primary coil. Also, it’s evident that the capacitor discharge-like behavior of the flyback secondary output occurs once the atmospheric air is ionized. In this setup, voltages up to 6 kV were achieved, while momentary primary ringing burst stayed around 40 V, with 120 V bursts.

This energy-heat conversion has to be considered when designing an enclosure for the system, since not only do the transistors need heat dissipation, but the flyback transformer also needs some sort of ventilation, if operating for large periods of time. There is no need for active cooling or a metallic heat sink attached to the transformer, a simple pathway for natural convection is fine.

Looking into the green curve of Figure 20, we have the corona discharge behavior. Following each pulse on the flyback’s primary coil, we have a high voltage pulse on its secondary. This high-voltage pulse is sufficient to ionize a path between the high-voltage and ground electrodes, through the atmospheric air. When the plasma channel is formed, the energy stored on the flyback’s secondary coil can be discharged to ground. Although a plasma impedance varies with time and several other conditions, a rough approximation to a resistor can be made. Therefore, the current stored is discharged through a plasma pathway, and what we measure with the oscilloscope, as seen in the simulation, is the voltage drop through the plasma complex impedance.

5.3 Decontamination evidence

Figure 21 and Figure 22 demonstrate the sterilizing effect of the developed CAP source. In the first one, the Petri dish on the left side is the control group, showing bacterial proliferation in the inoculated areas. Both the center and right side Petri dishes did not present bacterial growth in the inoculated areas after being exposed to CAP for 10 minutes. The second figure has the plasma biocidal action within 4 distinct exposure times, being 1, 3, 6, and 10 minutes. This test was also done in duplicates. The dishes are presented being 1, 3, 6, 10 minutes, from left to right, and the rightmost dishes are the control ones. All dishes, from both tests, were incubated at 30º C for 24 hours, between exposure and analysis.

Figure 21
Preliminary results showing bacterial growth on the control Petri dish (on the left) and the duplicates that were exposed to 10 minutes of plasma treatment (center and right Petri dishes). All dishes were incubated for 24 hours at ${30}^{\circ }C$.

Figure 22
Results of the first test, with the inoculation covering the entire substrate, not just spots. On the left we have the control plate, with the entire surface covered by colonies, and on the right the result of exposure to the plasma beam. Here it is important to observe the diffusion halo of reactive particles, where in the center of the exposure we have no evident growth, and as we move away from the center of the exposure, isolated colonies appear gradually, until total coverage of the substrate in the untreated region.

The first test was done in duplicates, as seen in the center and right Petri dishes of Figure 21. All inoculation points were bathed in $10\mu$L of ${10}^{\text{6}}$ cells/mL. The specimen was Peribacillus simplex SDF0016 [²⁴], in the log phase, and the medium was LB with 1.2% agar, $40mL$ per Petri dish. The second test was also done in duplicates, shown in Figure 23 in the top and bottom lines. Each column corresponds with a different time, respectively 1, 3, 6, 10 minutes, and control dishes. For this test, the medium was LB with 1.2% agar, $40mL$ per Petri dish, and the specimen was Peribacillus simplex SDF0016 [²⁴], in the log phase, with $200\mu L$ of ${10}^{\text{8}}$ cells/mL.

Figure 23
Result of the second test regime. The upper row represents the dishes, from left to right, with exposure times of 1, 3, 6, 10 minutes, with the rightmost dishes being the control group. The bottom row brings the dishes from the duplicate exposure, maintaining the same parameters as its counterparts.

During the test, a distance of 50 mm was set between the positive electrode, inside the disposable syringe, and the ground plane beneath the Petri dish, for the first test, and a distance of 40 mm for the second one. An electrical connection was made between the agar medium inside the Petri dish and the ground plane. This was necessary due to the DC nature of the discharge, which emits mostly one polarity species, causing an electrostatic charge to arise in the Petri dish, which repels the plasma stream. By grounding the dielectric material of the agar substrate and the ground plane, we can counteract this effect. The exit point of the syringe was between 15 mm and 20 mm apart from the surface of the agar substrate for both tests.

From the Figs. 21, 22, 23, was possible to see the biocidal action of this CAP source. After 10 minutes of exposure, there are no visible colony formation at the center of the plasma discharge. Also, in Figure 23, we can see the relationship between duration of exposure and biocidal action, were smaller times have more colony formation units. This inverse proportion between time and effectiveness is in tandem with the plasma dynamics in an atmospheric discharge, and is seem in other plasma sources [²², ²³].

The prototype circuit performed well during this test regime. However, arc discharges were observed during the tests. Arcs can occur due to the reduced dielectric impedance of the mixture of ionized and neutral species, the variance in distance between electrode and substrate, and environmental conditions, such as temperature, pressure, and humidity. When arc discharges indeed occur, either the NE555, LM741 ICs, or both can be catastrophically damaged. Also the IRF840, IRF530 or the IRF540 MOSFETS requires heat dissipation, being damaged when overheated. With an aluminium heatsink, this critical component had its temperature stabilized between ${70}^{\circ }C$ and ${80}^{\circ }C$. In the right conditions, especially with no arc discharges, it can operate for hours without flaws.

6. Final Considerations

This study successfully demonstrated the design and implementation of a novel cold atmospheric plasma (CAP) device, capable of effective surface decontamination and adaptable for diverse applications. Beyond its practical utility, the device provides an excellent platform for teaching fundamental physics concepts, including plasma generation, ionization process, and electromagnetic phenomena. Future work will focus on exploring additional gas types, refining the device for broader educational use, and developing complementary teaching resources to integrate CAP experiments into physics curricula. These advancements will contribute not only to scientific research but also to enhancing physics education by providing students with hands-on, interdisciplinary learning experiences.

Although the relationship between duration of exposure and biocidal action occurred as expected, other parameters can affect both positively, and negatively, such as discharge current, voltage potential, distance from electrodes, frequency, and gas mixture, among others. Therefore, more tests with this setup can be done in order to ascertain operational thresholds for effective decontamination.

Future work will aim to improve the CAP source. Some aspects to be improved can be separated into tests and circuit improvements. For the test regime, different exposure times are needed, to verify the minimum application time necessary, as well as the relation between the input current and treatment time. In addition, other biological contaminants, such as fungi and viruses, can behave differently when exposed to plasma discharge, and more testing is needed.

Focusing on the electronic circuit, the transformer driver is sensitive to electric arcs at the plasma wand output, burning several components throughout the tests. Sturdier components may have a longer lifespan, as well utilizing microcontrollers and control systems to achieve discharge automation. Some of the electrostatic charging effects can be solved by means of AC circuits, like the ZVS circuit, or by connecting the Petri dishes to circuit reference.

Currently, a new and improved circuit has been developed and is being tested to evaluate its reliability. In addition, a new type of electrode is being developed alongside a new prototype with a broader application area.

Acknowledgments

The authors would like to thank the anonymous reviewers who contributed positively to the best version of this work. HOCJ acknowledges COPEI/Finatec (UnB) under grant 7178 for the funding and financial support of the project. RAM acknowledges support from CNPq, Brazil (grants 407341/2022-6, 407493/2022-0) and COPEI/Finatec (UnB) (grant 7178). JLF acknowledges support from CNPq (grant 405907/2022-2).

Data Availability

The cad files for 3D printing are available at github.com/HelbertJunior/CAP-Prototyope-Beta. The circuit schematics and parts list are sufficient to replicate the device shown in this work. Futhermore all data sets that corroborate the results are contained in this work.

References

• ^[1] J.H.C. Souza, Estudos de esterilização por plasma a pressão atmosférica Tese de Doutorado, Universidade de Brasília, Brasília (2012).
• ^[2] N. Kaushik, S. Mitra, E.J. Baek, L.N. Nguyen, P. Bhartiya, J.H. Kim, E.H. Choi and N.K. Kaushik, Journal of Advanced Research 43, 59 (2023).
• ^[3] S. Moritz, A. Schmidt, J. Sann and M.H. Thoma, Journal of Physics D: Applied Physics 53, 325203 (2020).
• ^[4] A. Schmidt, C.Y. Tschang, J. Sann and M. Thoma, arXiv:2109.03054 (2021).
• ^[5] Z. Chen, G. Chen, R. Obenchain, R. Zhang, F. Bai, T. Fang, H. Wang, Y. Lu, R. Wirz and Z. Gu, Materials Today 54, 153 (2022).
• ^[6] Z. Chen, J. Wu, H. Zhang, M. Zhang, S. Zhang and D. Feng, Plasma Science and Technology 22, 085403 (2020).
• ^[7] M. Izadjoo, S. Zack, H. Kim and J. Skiba, Journal of Wound Care 27, S4 (2018).
• ^[8] Z. Chen, G. Garcia Jr., V. Arumugaswami and R.E. Wirz, Physics of Fluids 32, 11 (2020).
• ^[9] P. Thana, A. Wijaikhum, P. Poramapijitwat, C. Kuensaen, J. Meerak, A. Ngamjarurojana, S. Sarapirom and D. Boonyawan, Heliyon 5, 9 (2019).
• ^[10] L. Yinxin, C. Zhao, M. Huang, C. Sun, J. Cao, J. Wang and L. Huang, Food Control 137, 108906 (2022).
• ^[11] V. Vijayarangan, S. Dozias, C. Heusèle, O. Jeanneton, C. Nizard, C. Pichon, J.M. Pouvesle, A. Stancampiano and E. Robert, Frontiers in Physics 11, 1173349 (2023).
• ^[12] A.M. Howatson, An introduction to gas discharges (Elsevier, Amsterdam, 2013).
• ^[13] E. Husain and R.S. Nema, IEEE Transactions on Electrical Insulation 17, 350 (1982).
• ^[14] X. Pei, J. Kredl, X. Lu and J. Kolb, Journal of Physics D: Applied Physics 51, 384001 (2018).
• ^[15] F. Nascimento, A.G. Sampaio, N.V.M. Milhan, A. Gontijo, P. Mattern, T. Gerling, E. Robert, C.K. Ito and K. Kostov, arXiv:2310.20001 (2023).
• ^[16] C.K. Ito, K.G. Kostov, F. Miranda, N.V.M. Milhan, N. Neto, F. Nascimento and R. Pessoa, Plasma Chemistry and Plasma Processing 44, 1393 (2023).
• ^[17] M. Leduc, D. Guay, R.L. Leask and S. Coulombe, New Journal of Physics 11, 115021 (2009).
• ^[18] V. Vijayarangan, A. Delalande, S. Dozias, J.M. Pouvesle, E. Robert and C. Pichon, International Journal of Pharmaceutics 589, 119874 (2020).
• ^[19] G. Isbary, G. Morfill, H.U. Schmidt, M. Georgi, K. Ramrath, J. Heinlin, S. Karrer, M. Landthaler, T. Shimizu, B. Steffes et al., British Journal of Dermatology 163, 78 (2010).
• ^[20] T. Maho, R. Binois, F. Brulé-Morabito, M. Demasure, C. Douat, S. Dozias, P.E. Bocanegra, I. Goard, L. Hocqueloux, C. Le Helloco et al., Applied Sciences 11, 9598 (2021).
• ^[21] H. Qin, H. Qiu, K. Liu, B. Hong, Y. Liu, C. Li, M. Li, X. An, L. Song, E. Robert et al., Exploration 4, 20230012 (2024).
• ^[22] C. Zhintong, C. Guojun, R. Obenchain, R. Zhang, F. Bai, T. Fang, H. Wang, Y. Lu, R. Wirz and Z. Gu, Materials Today 54, 153 (2022).
• ^[23] H. Zhu, L. Guo, M. Li, Y. Jiang and H. Li, Vacuum 184, p. 109982 (2021).
• ^[24] J. Orem, T. Raiol, M. Magalhães, P. Martins, D. Cavalcante, R. Kruger, M. Brigido and M. Souza, International Microbiology 22, 511 (2019).
• ^[25] U. Habedank, IEEE Transactions on Power Delivery 8, 1921 (1993).
• ^[26] A. Parizad, H. Baghaee, A. Tavakoli and S. Jamali, in: International Conference on Electrical Engineering (Busan, 2009).
• ^[27] F. Yang, Z. Tang, Y. Shen, L. Su and Z. Yang, Frontiers in Energy Research 10, 808289 (2022).

• 1
  These pulses are applied to the CAP device to sustain plasma discharge by periodically ionizing the working gas. The circuit comprises key components, including a high-voltage power supply, a pulse generator, and a capacitor-resistor network, wich regulate the pulse characteristics. This setup offers an excelent opportunity for students to study the principles of pulsed circuits and their applications in plasma generation.

Edited by

• Editor-in-Chief:
  Marcello Ferreira https://orcid.org/0000-0003-4945-3169

Publication Dates

• Publication in this collection
  13 Feb 2026
• Date of issue
  2026

History

• Received
  06 Dec 2024
• Reviewed
  23 Sept 2025
• Accepted
  26 Oct 2025