Fabrication of Zinc Oxide Nanoparticles Based on a Novel Technique of In-Liquid Plasma

INTRODUCTION
In recent decades, zinc oxide (ZnO) nanoparticles have attracted significant attention and have become an important material in various applications due to their unique optical, electronic, and physicochemical features [1]. This metal oxide nanoparticle can efficiently absorb and emit light, due to the direct band gap, approximately at about 3.37 eV at room temperature. This property makes ZnO an amazing semiconductor for various applications, including advanced photocatalytic systems to remove environmental pollutants and gas sensors to detect the harmful gases in the atmosphere [2]. In addition, the antibacterial features of zinc oxide nanoparticles have also been considered as a promising substance for combating microbial growth and the development of antimicrobial coatings [3]. Hence, the various applications of ZnO nanostructures, including photocatalysts and gas sensors, antibacterial coatings, and optoelectronic materials, highlight their importance in both scientific research and industrial applications [4, 5].
Various methods have been developed to produce ZnO nanostructures, including wet chemical methods, chemical vapor deposition, hydrothermal, and combustion synthesis [6, 7]. However, many of these methods require hazardous chemical precursors, high temperatures, or specific controlled conditions, which limit their industrial scalability [8]. In response to these challenges, researchers are now exploring alternative methods that offer better benefits [5, 9]. In this regard, interest in physical and non-equilibrium methods, such as liquid pulse laser ablation, in liquid plasma discharge, and acoustic cavitation has increased due to their lower cost, greater safety, and more environmental compatibility [10, 11].
Although zinc oxide nanostructures have been synthesized by various methods, the liquid plasma discharge has attracted great attention. Plasma discharge in liquids using the spark or arc method is considered a novel technique for providing activation energy for the reactions at typical ambient conditions [12, 13]. In this method, electrical energy is directly applied to a localized region of the liquid, resulting in rapid evaporation and the formation of vapor bubbles [14]. Since direct phase transition from the liquid phase to the plasma phase is thermodynamically impossible, the plasma is formed in this vapor region. The active species produced by the plasma in this method include OH•, H•, and O• -, which play a key role in breaking bonds and carrying out redox and oxidation reactions [15-17]. This method allows the synthesis of nanomaterials with proper physicochemical properties and high activity, and also presents the advantages, such as control over particle size and shape, the possibility of production under a wide range of reaction conditions, and the possibility of mass production. provides [16, 18]. On the other hand, ultrasonic cavitation is also known as an effective mechanism for the production of nanostructures.  The rapid collapse of created gas bubbles in the liquid phase leads to localized shocks, hot spots, and microcurrents, resulting both aid nucleation and prevent particle aggregation [18]. Therefore, it seems that combining plasma and cavitation phenomena in a reaction medium can provide more efficient and controllable synthetic processes.
In the plasma-in-liquid method, the yield of reactions and final properties of nanoparticles such as size, shape, and specific surface area are highly dependent on the design of the plasma generation device and operating conditions. The most important of these factors include discharge voltage, electrode spacing, pulse frequency and duration, electrode material and type, chemical composition, and pH of the reaction solution [19]. High discharge voltage is the main factor for plasma generation. Increasing voltage increases the energy of active species, reduces particle size, and increases the specific surface area, but at very high voltages there is a possibility of regrowth or local melting. Short, high-energy pulses also create fast vapor bubbles and stronger plasmas [20]. Also, a smaller distance between the electrodes creates a stronger electric field and easier discharge. The presence of ions (such as OH⁻ or NO₃⁻) changes the reaction path. For example, an alkaline medium is usually more suitable for the synthesis of zinc oxide [21]. On the other hand, the type and material of the anode also affect the final properties of the nanoparticles [21, 22]. The anode can play the role of a precursor. In addition, the material of the electrode, the temperature of the solution and the electrical conductivity also affect the properties of the nanoparticles, impurities and even the efficiency of the production of nanoparticles [21].
The power supply is a critical component in the in-liquid plasma method, as the high voltage and power pulses required to generate plasma in a liquid medium can only be provided by power supplies specifically designed for this purpose [23]. The power supply must be capable of generating high-voltage pulses with controlled amplitude and frequency to provide the optimal conditions for nanoparticle synthesis control over particle size and shape, the possibility of production under a wide range of reaction conditions, and the possibility of mass production. Provides [20]. The importance of the power supply is due to the voltage and pulses produced directly affect the physical and chemical properties of the nanoparticles produced, including their size, shape, and surface properties [24]. The use of a Marx generator as a source of high-voltage pulses that allows for the creation of very sharp, short, and frequency-controlled electrical discharges in an aqueous environment is important. This feature can create optimal conditions for the nucleation process and controlled growth of zinc oxide nanoparticles, and compared to conventional power sources, it can improve crystal quality, reduce particle aggregation, and increase specific surface area [20]. 
This study seeks a new method for the synthesis of ZnO nanoparticles from zinc metal electrode in aqueous medium, which is based on pulse liquid plasma discharge medium using a Marx circuit. The main challenge in plasma-in-liquid methods is to achieve scalable and energy-efficient routes for producing the nanoparticles. In this work, the effectiveness of the designed Marx generator was investigated in terms of creating very sharp and short voltage pulses in an aqueous medium. In this circuit, the capacitors are placed in parallel in the charging path and are connected in series at the moment of discharge. The result of this arrangement can be releasing the entire stored power in a short time interval and the creation of a large electric field in the electrode gap. It seems that this high-energy discharge causes rapid local evaporation, vapor bubble formation, and ultimately plasma production. After the end of the pulse, the hot bubble subsides into the cold liquid medium, and the cavitation phenomenon occurs naturally without the need for additional equipment such as ultrasonic generators or plasma jet systems.

MATERIALS AND METHODS
Chemical
Zinc wire (Diameter of 2 mm, 99.99%) and titanium wire (Diameter of 2 mm, 99.9%) were provided by Fara Pooshesh Mahoor Co. and Safir Sanat Co., respectively. Ammonia solution (NH4OH, 25%) was purchased from Merck Co. Ethanol (C2H5OH, 96%) was provided by Hamoon Teb Co. 

Synthesis Procedure 
First, ammonia was added dropwise to 1 L of deionized water and stirred until the pH reached about 10 at room temperature. Then, a Zinc wire was used as the anode electrode, and a titanium wire was used as the cathode electrode, both located in the alkaline water. After that, a high-voltage electric current was applied to the electrodes in the water, which resulted in the production of white gel-like precipitates of zinc hydroxide. The wet precipitates were collected using filtering, washed twice with ethanol, and dried in an oven at 60℃ for 12 h. Dried zinc hydroxide particles were thermally treated in an oven at 400°C for 2 h, obtaining ZnO nanoparticles.  Here, to investigate the effect of voltage on the surface properties, crystal structure, and particle size of nanoparticles, three ZnO nanoparticle samples were prepared using three different voltages of 8000, 10000, and 12000 V, denoted as ZnO-8, ZnO-10, and ZnO-12 samples.
 Here, the Marx generator was designed to produce short-time and high-voltage pulses between 2 and 14 kV in order to generate electrical discharge required in plasma synthesis. Fig. 1 presents a scheme of the designed Marx generator and reaction medium. As is clear, these capacitors were connected in series in such a way that when the circuit is activated, they create a high voltage pulse, which is necessary to create an electric discharge of plasma in the liquid medium. Each part of the Marx circuit was made using capacitors with a capacity of 10 nanofarads per stage and a voltage tolerance of up to 20 kV.  The frequency of the output pulses of the Marx circuit on the pulse output and probe was measured using a digital oscilloscope with a 1:1000 high-voltage probe. The operating frequency was also recorded in the range of 7.5 Hz. The input current was 0.14 A, and the effective voltage was 220 V. The power factor of the system was considered to be nearly unity (close to 1). The nominal capacitance of the Marx circuit capacitors was confirmed to be around 40 nanofarads. The input power and energy of each pulse were calculated to be 30 W and 2 J, respectively. Given a frequency of 7.5 Hz, the pulse output power was 15 W, and consequently, the efficiency was about 50%. The spark gap distance was adjusted at about 0.8, 1, and 1.2 cm to create the high voltage discharge conditions, which resulted in the generation of high voltages at about 8000, 10000, and 12000 V, respectively. 

Characterization
X-ray diffraction (XRD, PAnalytical X’Pert-Pro, Netherlands) analytical technique was used to address the crystal structure of synthesized ZnO nanoparticles, in the range of 10° to 80°. Fourier transform infrared spectroscopy (FT-IR, Magna-IR 550 spectrometer, Nicolet Co.) was applied to identify the chemical properties of samples from 400 to 4,000 cm-1. The physical properties of specific surface area, pore volume and pore size of samples were determined using Brunauer–Emmet–teller (BET) and Barrett-Joyner-Halenda (BJH) techniques, based on recorded N2 adsorption-desorption isotherms (Belsorp 28 system, Japan) using an automatic system. samples were first degassed under vacuum for 4 h at 250 °C. The field emission scanning electron microscopy (FESEM, MIRA3 TESCAN, USA) was also utilized to specify the morphology of synthesized ZnO nanoparticles. Photoluminescence spectroscopy (PL, G9800A, Agilent, USA) at room temperature, in the wavelength range of 300-700 nm, was performed to evaluate the optical properties and electron transfers in ZnO nanoparticles.

RESULTS AND DISCUSSION
The in-liquid plasma discharge method requires a stable power supply capable of generating high-voltage pulses over extended periods. The Marx circuit designed in this project was an ideal choice for creating intense electric fields due to its ability to generate powerful, short-duration pulses at sequential frequencies with high stability. Here, Under the applied discharge conditions (30 W), a representative 30-min experiment produced 2.22 g of ZnO, corresponding to a rate of about 4.4 g h⁻¹. Based on the measured mass loss of the zinc electrode and the final ZnO obtained after calcination, the reaction yield was around 94%. The corresponding energy efficiency was calculated to be approximately 148 g kWh⁻¹.

Reaction Mechanism
The mechanism of nanoparticle formation in the method performed in this study includes two stages of the creation of initial centers (nucleation) and growth, which determine the size, shape, and final properties of the nanoparticles. When plasma is created in a liquid medium under high voltage conditions, the very high temperature and pressure in the discharge zone and the rapid cooling of the surrounding environment provide completely non-equilibrium conditions. In such an environment, the released atoms and ions quickly reach a supersaturated state and tend to form initial nuclei with nanometer dimensions. These nuclei are the main base for further growth and transformation into stable nanoparticles. After the nucleation stage is completed and a significant number of stable primary nuclei are formed in the medium, the medium is no longer capable of creating new nuclei and enters the growth stage. In this stage, the growth of nanoparticles is based on the gradual attachment of ions and active species to the surface of existing nuclei. Here, Zn²⁺ ions released from the anode electrode and active oxygen species from the plasma react with the surface of the nuclei and complete the ZnO crystal lattice. In addition, the high local temperature and non-equilibrium conditions around the plasma bubbles accelerate the diffusion process and stabilize the crystal structure.

Structural Properties
Changing the voltage in the plasma discharge process can have significant effects on the crystal structure and size of the prepared nanoparticles. The impact of voltage changes on the structural and crystalline size basic properties of the synthesized nanoparticles was investigated using X-ray diffraction analysis, as represented in Fig. 2. The diffraction peaks appeared at about 2θ values of 31°, 34°, 36°, 47°, 56.°, 63°, and 67° can be attributed to the formation of the hexagonal structure (wurtzite phase) of zinc oxide, in line with JCPDS No: 36-1451, indexed to (100), (002), (101), (102), (110), (103), and (112) plans, respectively  [10, 25]. In addition, sharp and intense peaks are observed in the XRD pattern of nanoparticles, confirming the completely crystalline nature of ZnO nanoparticles [26]. 
ZnO is inherently a material with a stable crystalline structure. Usually, ZnO nanoparticles crystallize spontaneously even at relatively low temperatures, because the Gibbs free energy for the crystalline state is lower. There are very limited reports of the synthesis of amorphous ZnO, usually under special conditions such as chemical precipitation at very low temperatures or with organic stabilizers that prevent crystallization. However, in most methods (especially plasma in a liquid, which introduces a lot of energy into the system), a completely crystalline product is obtained. ZnO nanoparticles synthesized by plasma-in-liquid discharge exhibit fully crystalline features due to the high-energy plasma environment, which provides sufficient energy for nucleation and crystal growth. Therefore, no amorphous phase is observed in the XRD patterns and the XRD patterns also confirm the absence of an amorphous background.
The average crystallite size of ZnO-8000, ZnO-10000, and ZnO-12000 was estimated based on the Debye-Scherrer equation and Williamson-Hall method, as reported in Table 1.  As is clear, the sample prepared using a voltage of 12000 V (spark gap distance of 1.2 cm) led to a smaller crystallite size. In this regard, as the voltage increases, the plasma energy increases, and the atomization process intensifies. In this case, nucleation occurs more rapidly, and the crystallites have less opportunity to grow. Therefore, at higher voltages, the crystallite size and crystal defects usually reduce. This trend indicates a decrease in crystal size, which is consistent with the LaMer mechanism and an increase in the nucleation rate at intermediate voltage [25].
The results of fitting the data by the Williamson–Hall method showed that lattice strain makes a significant contribution to the peak broadening. For this reason, the crystallite sizes obtained by the Williamson–Hall method were significantly smaller than those calculated by the Debye–Scherrer equation. This is logical, since the Scherrer method assumes that all the peak broadening is due to the crystallite size, while the Williamson–Hall method attributes part of this broadening to lattice strain [27]. In both methods, a general trend of decreasing crystallite size was observed with increasing discharge voltage. The difference in the absolute value of the sizes indicates that lattice strain played a significant role in the structure of ZnO produced by the plasma method in liquid. This strain can be caused by the high growth and nucleation rate in the plasma environment, rapid changes in local temperature and pressure, and the presence of dislocations and lattice defects [28].

Chemical Properties
Fourier transform infrared spectroscopy (FTIR) was used as a powerful tool for identifying the functional groups and chemical bonds in the synthesized zinc oxide nanoparticles. According to the FTIR results in Fig. 3, the sharp absorption band observed around 500 cm-1 refers to the vibrations of zinc–oxygen (Zn–O) bonds, which is the most important indication for the synthesis of zinc oxide [29]. The broad absorption bands around 3300-3500 cm-1 as well as around 1350 cm-1 corresponded to the vibration of hydroxyl (OH) groups on the surface of nanoparticles, which is caused by the absorption of water molecules and surface moisture  [30]. The peaks at the 1386 cm−1 may also be due to oxygen vacancies, which are characterized as hydrogen-related defects on the surface of ZnO in wurtzite hexagonal type ZnO crystals [31]. In addition, the sharp peaks that appeared at around 830 cm-1 can be attributed to the Zn-OH bonds [32]. 
Morphology and Physical Properties
The morphology of ZnO-8000, ZnO-10000, and ZnO-12000 nanoparticles, including shape, size, uniformity, porosity degree, surface roughness, and agglomeration of particles, which is of immense importance in the nanoparticle synthesis field, was evaluated using the FE-SEM micrographs, represented in Fig. 4. As is clear, the formation of spherical particles with almost uniform dimensions in the range of 5 to 50 nm can be observed for all prepared ZnO nanoparticles. The amount of voltage in the liquid plasma discharge method has a direct effect on particle size and particle shape. Here, it is observed that increasing the voltage led to a decrease in the particle size. In the plasma discharge method, local temperatures increase, and rapid cooling occurs. Therefore, a porous structure is obtained for the synthesized nanoparticles [33]. Furthermore, the particle size formed in the sample with a voltage of 12,000 (ZnO-12000) is clearly smaller than ZnO-10000, and ZnO-8000 and has much less agglomeration. Increasing the voltage can causes the agglomerates to break, which reduces the particle size [15]. 
Although increasing the voltage can reduce the initial particle size, at very high voltages, high local temperatures may cause local melting and coalescence of the particles, subsequently leading to agglomeration or increased particle size and non-uniformity. It is also possible to change the morphology from perfectly spherical particles to rod-like or sheet-like shapes, depending on the plasma energy [15].
The specific surface area, pore volume, and pore diameters have a direct relationship with the performance of ZnO nanoparticles, especially in photocatalytic applications [34]. Higher specific surface area means more active sites for adsorbing pollutant molecules. Table 2 indicates the specific surface area, mean pore size, and pore volume of the ZnO-8000, ZnO-10000, and ZnO-12000 samples. As is clear, increasing the voltage from 8000 to 10000 and 12000 V led to an increase in the surface area from 71 m2/g up to 94 and 99 m2/g, respectively. In addition, the mean pore diameters were also decreased with increasing voltage. It seems that with increasing the spark gap distance and consequently the voltage, more plasma energy is generated and atomization occurs more strongly, leading to the formation of finer particles and a higher specific surface area [35, 36]. Furthermore, the high voltage created in-liquid plasma discharge also leads to the breakdown of agglomerates and increases the effective specific surface area of the nanoparticles [37, 38].  
However, it is observed that increasing the voltage from 10000 to 12000 did not lead to a significant improvement in the specific surface area. At very high voltages, particle growth may reach equilibrium, which limits the increase in specific surface area. Also, at high voltages, some of the micropores may disappear or coalesce to form larger mesopores, resulting in a less-than-expected increase in specific surface area [17, 39].
Fig. 5a represents the N2 adsorption-desorption isotherms of the prepared ZnO nanoparticles. The adsorption isotherms of all nanoparticles correspond to type IV according to the IUPAC standard classification, indicating mesopore structures with interconnected pores [40]. The type IV isotherm is usually associated with a rapid increase in N2 adsorption at high relative pressures and shows a hysteresis loop in the adsorption-desorption isotherm, as is clear in Fig. 5a The hysteresis loops in ZnO nanoparticles refer to type H-3, representing the presence of open pores in their structures [41]. 
Fig. 5b indicates the pore size distribution of the prepared ZnO nanoparticles, based on the BJH technique. The results confirm the formation of mesopores in the range of 2-50 nm. The formation of nanoparticles with a mesopore structure and relatively high specific surface area is ideal for photocatalytic applications [42]. In the plasma discharge method, the nanoparticles are usually formed as mesopores because rapid crystal growth is accompanied by the creation of interparticle spaces.

Optical properties
Photoluminescence spectroscopy is one of the key methods in identifying energy band characteristics, determining the band gap, and also detecting trap states and crystal defects in nanostructures. Since the photocatalytic performance of ZnO nanoparticles is strongly affected by electron-hole recombination and the presence of defect centers, the use of photoluminescence testing allows for indirect evaluation of the photocatalytic efficiency of samples [43]. Fig. 6 indicates the PL spectrum of the ZnO=12000 sample. As can be seen, three distinct bands are observed, each representing electron-hole recombination processes at different energy levels and the presence of various crystal/surface defects [44].  The peak around 396 nm can be attributed to near-band edge emission; this peak indicates radiative recombination of conduction band electrons with valence band holes, and its presence indicates the presence of a significant fraction of relatively good crystallinity in the sample [45]. The sharp peak around 452 nm (blue region) is usually attributed to emission associated with shallow donor surfaces or interband states due to interstitial zinc atoms (Zn_i) or deep surface-half transitions.
The peak around 568 nm (orange/yellow) represents emission due to oxygen vacancies, oxygen interstitials or deep surface trap centers, which usually indicate the presence of bulk defects or oxide surface adsorptions and are known as surface-repulsion emissions [46]. The relatively good intensity of the peak (396 nm) indicates a high ratio of radiative recombination and acceptable crystal quality. The blue peak (452 ​​nm) with significant intensity indicates that the sample has a certain number of surface/interstitial conduction carriers that can affect the photocatalytic behavior, for example, they can facilitate charge separation or change recombination pathways [45, 46].

Comparison with Previous Studies
Recently, many attempts have been made to produce nanoparticles using plasma technology, and many reports have also described the optimal operating conditions required for plasma production and the various mechanisms for producing nanoparticles. Table 3 provides a brief comparison between the findings of this study and previous reports on the preparation of zinc oxide nanoparticles using plasma technology. Abdullah EA et al [47] produced the ZnO nanoparticles in spherical particles with particle size of 10-50 nm using atmospheric plasma jet method, and funded that the morphology is completely dependence on concentration of NaOH–HNO₃ solution. In study conducted using microwave plasma in liquid [48], nanoparticles with metallic zinc were formed in hexagonal and cylindrical shapes with particle sizes of about 10-200 nanometers. Saito et al [49] was also investigated the effect of differnte voltages ranging 42-200 V in soluble plasma method on synthesis of ZnO nanoparticles using zinc wire as cathode, and potassium carbonate solution. The resulted nanoparticles were obtained in nanoflowers and nanorods morphology with particle size of less than 100 nm.
The proposed method in the present study, with a strong discharge by a Marx generator designed, produces extremely high pulse energy for plasma generation. This method is expected to have higher energy efficiency than conventional continuous discharges. The nanoparticles produced in this research have also been produced with spherical and uniform morphology with dimensions less than 50 nm.

CONCLUSION
The successful synthesis of zinc oxide nanoparticles was investigated using a novel plasma discharge method in liquid, and the effect of voltage changes was investigated through spark gap changes. In this study, a Marx generator was designed to generate high voltage and short-duration pulses. The formation of zinc oxide wurtzite structure was confirmed in all samples, so that the average crystallite size of the ZnO sample provided using a voltage of 12 kV was reached at about 17 nm. The size of ZnO crystals synthesized by the liquid plasma discharge method was calculated using the Debye-Scherer equation and the Williamson-Hall method. The results of both methods showed that increasing the discharge voltage leads to a decrease in crystal size.” The FESEM micrographs confirmed the formation of spherical particles with almost uniform dimensions in the range of 5 to 50 nm. The FESEM results also showed decreasing particle size and reduced agglomeration with increasing discharge voltage. Also, more uniform and porous nanostructures were obtained by using a higher voltage. However, the results confirmed that although increasing the plasma energy from 8 to 12 kV leads to an increase in the specific surface area from 71 m2/g up to 99 m2/g, an excessive increase may not have much effect on the specific surface area or even have the opposite result. The outcomes indicate that the plasma discharge method in liquid can enable the synthesis of active nanoparticles with suitable properties, along with controlling the size and porosity of the particles through voltage control, without the use of hazardous precursors or auxiliary equipment such as ultrasonics. This flexibility, along with the elimination of dissolved precursors and extraneous gases, makes the present method a suitable option for producing ZnO nanostructures on both a laboratory and semi-industrial scale.

CONFLICT OF INTEREST
The authors declare that there is no conflict of interests regarding the publication of this manuscript.