Introduction
Atmospheric pressure plasma jets (APPJs) based on dielectric barrier discharge (DBD) have shown significant potential for implementing them in low-temperature gas processing applications at atmospheric pressure, thereby eliminating the need for a vacuum chamber (Jeong et al., Reference Jeong, Babayan, Tu, Park, Henins, Hicks and Selwyn1998; Stoffels et al., Reference Stoffels, Flikweert, Stoffels and Kroesen2002, Reference Stoffels, Kieft and Sladek2003; Benhansour et al., Reference Benhansour, Nikravech, Morvan, Amouroux and Chapelle2004; Deng et al., Reference Deng, Shzi, Shama and Kong2005; Walsh et al., Reference Walsh, Shi and Kong2006; Divya Deepak et al., Reference Divya Deepak, Joshi and Prakash2018a, Reference Divya Deepak, Joshi, Prakash and Pal2018b). These plasma jets are close to room temperature, and hence, popularly referred to as atmospheric pressure cold plasma jets (APCPJs) (Deng et al., Reference Deng, Shzi, Shama and Kong2005). The APCPJs have acquired considerable attention due to their pivotal role in various plasma processing applications. The other intriguing feature of cold plasma jets is that they are not restricted by electrodes (Eliasson and Kogelchatz, Reference Eliasson and Kogelchatz1991; Kruger et al., Reference Kruger, Owano and Laux1997; Sladek et al., Reference Sladek, Stoffels, Walraven, Tielbeek and Koolhoven2004) APCPJs have further been utilized in various biomedical and technological applications (Schutze et al., Reference Schutze, Jeong, Babayan, Park, Selwyn and Hicks1998; Xu, Reference Xu2001; Duan et al., Reference Duan, Huang and Yu2005; Roth et al., Reference Roth, Nourgostar and Bonds2007; Xu et al., Reference Xu, Ma and Zhang2008; Kim et al., Reference Kim, Kim, Hong and Yang2010). The most imperative property of this plasma type is that it functions close to room temperature and also lets the treatment of uneven surfaces with the minor depth of penetration. Various researchers (Frame et al., Reference Frame, Wheeler, Detemple and Eden1997; Selwyn et al., Reference Selwyn, Herrmann, Park and Henins2001; Chirokov et al., Reference Chirokov, Gutsol and Fridman2005; Becker et al., Reference Becker, Schoenbach and Eden2006). Raizer (Reference Raizer1991) had elaborated upon the fundamental discharge phenomenon of DBD-based plasmas. It is mentioned that in case of air gaps produced by the plane electrodes if p.d > 5000 Torr cm, steamer breakdown is employed and Townsend mechanism of avalanche multiplication is implemented for p.d < 200 Torr cm, where the distance between the electrodes is denoted by d and p, refers to the pressure. Streamer breakdown is characteristic of comparatively large air gaps. As the value of p.d is less than 200 Torr cm for the ring electrode structure mentioned in this current research work; therefore, this mechanism of discharge is mostly directed by a Townsend breakdown mechanism which is related to the expansion of multiple avalanches. The ring electrode structure has been studied here to optimize the operational factors (supply frequency, supply voltage) to produce cold plasma jets that have long jet lengths. This electrode configuration may be used for possible technological applications.
The ring electrode structure mentioned in the current research work fulfills all the essential conditions for its applications in biomedical applications, which are low gas temperature, rich active species, no arcing, and jet ejecting in air. Teschke et al. (Reference Teschke, Kedzierski, Finantu-Dinu, Korzec and Engemann2005) examined an APPJ using an intensified charge-coupled device (ICCD) camera. It is found that the formation of APPJ is chiefly an electrical phenomenon and not related to flow. Maximum power of 4 W was consumed by this device, while the ring electrode structure mentioned in this research paper has a maximum power consumption of only 1.27 W. Further, Lu and Laroussi (Reference Lu and Laroussi2006) examined the atmospheric pressure dynamics produced by sub-microsecond pulses with the help of V–I characteristics and also used CCD camera for examining the spatial and temporal evolutions of the plasma within the discharge chamber. Honga and Uhm (Reference Honga and Uhm2006) produced microplasma jet using nitrogen gas at atmospheric pressure with the help of two aluminum disks in which the gas temperature was found to be less than 3000 K and it was recommended to be implemented for medical applications. Lu et al. (Reference Lu, Jiang, Xiong, Tang, Hu and Pan2008) created an 11 cm cold plasma jet using an ac power supply (kHz) for technological applications; however, optimization of electrical parameters (supply frequency and supply voltage) of the device had not been done. Nie et al. (Reference Nie, Ren, Wang and Zhang2008) produced a plasma jet of argon gas with the help of the tungsten pin electrode pair which was 2.5 mm apart. Among these, one of the tungsten pin electrodes had been kept at a floating potential. The effect of the applied potential and velocity of flow on plasma jet was investigated; however, the power consumed by the plasma jet was not estimated. Pal et al. (Reference Pal, Gulati, Kumar, Prakash and Srivastava2012) examined the multiple-peak discharge occurrence via an equivalent circuit. Radu et al. (Reference Radu, Bartnikas, Czeremuszkin and Wertheimer2003) investigated an APPJ in the DBD reactor under variable experimental conditions (electric field intensity, gap length, and supply frequency) which assisted them in the electrical characterization of the DBD. Kim et al. (Reference Kim, Brockhaus and Engemann2009) produced plasma based on argon gas with the help of a piezo transformer which had a maximum power consumption of 2 W.
Hong Yi et al. (Reference Yi, Lu, Jing, Jie and Yan2013) was successful in generating a DBD-based plasma jet using argon gas by implementing a screw ring-ring electrodes, here the emission spectra of argon had been investigated to find out the physical parameters (supply voltage and frequency) of the discharge, like, the temperature of gas and consumption of power. Joh et al. (Reference Joh, Kang, Chung and Kim2014) created a plasma jet based on helium with the help of pin-shaped electrode where the effects of nozzle shape, electrode materials, and grounded ring electrode on the plasma were investigated; however, the particulars about the consumption of power by the device at different supply voltages and frequencies was not mentioned. Shashurin and Keidar (Reference Shashurin and Keidar2015) examine the research work done in the domain of cold atmospheric pressure plasma jets (C-APPJs) with specific emphasis on experimental methods. Divya Deepak et al. (Reference Divya Deepak, Joshi, Pal and Prakash2016) implemented ring electrode configuration to investigate DBD-based argon cold plasma jets and characterized them electrically. Divya Deepak et al. (Reference Divya Deepak, Joshi, Kumar Pal and Prakash2017) produced a cold plasma jet using a floating helix electrode configuration which had a power consumption of the range of few mW. Kelly and Turner (Reference Kelly and Turner2011) investigated an atmospheric pressure plasma which was based on alternating current and generated a two-dimensional (2D) fluid model of the same. Tendero et al. (Reference Tendero, Tixier, Tristant, Desmaison and Leprince2006) had completed the modeling of a DBD-based reactor with the incorporation of gas flow using COMSOL Multiphysics 5.2a. RF-based plasma needle modeling had been completed using a fluid model having a small fraction of N2 and He by Sakiyama and Graves (Reference Sakiyama and Graves2006).
Preliminary work on DBD-based modeling was only completed for the static discharge of gas. Modeling of streamer like APPJ was completed by Naidis (Reference Naidis2012). The fluid model developed by Gadkari et al. (Reference Gadkari, Tu and Gu2017) using COMSOL Multiphysics for a partially packed DBD-based plasma reactor that was based on pure He. Only a few studies have been done for modeling of DBD incorporating the flow of gas. Both electrical characterization and modeling of a DBD-based APPJ using argon gas in the pin electrode geometry has been completed by Divya Deepak et al. (Reference Divya Deepak, Joshi and Prakash2018a, Reference Divya Deepak, Joshi, Prakash and Pal2018b). This paper focuses on DBD-based APPJs including the flow of argon gas using fluid dynamics and the plasma module of COMSOL Multiphysics (Divya Deepak et al., Reference Divya Deepak, Joshi and Prakash2018a, Reference Divya Deepak, Joshi, Prakash and Pal2018b).COMSOL is multiphysics software is based on a finite element method (FEM). It contains a plasma module in which we can perform 1D, 2D, and 3D modeling, and simulation of APPJs generated using different electrode configurations. Each module of the COMSOL Multiphysics has a defined group of domain equations that are detailed to describe the physics of the process. The COMSOL modules contain all the attributes that are essential for modeling plasma discharge, starting with a Boltzmann Equation, 2-Term Approximation solver that estimates transport properties of source coefficients and electrons from a set of electron impact collision cross-sections (Hagelaar and Pitchford, Reference Hagelaar and Pitchford2005).
In this paper, modeling work has been carried out for DBD-based APPJs for a ring electrode configuration in both static discharge condition and with the flow of Ar gas. The effect of supply voltage and frequency on the plasma parameters like electron density, electron temperature, and the electrical potential of the DBD-based APPJs has been analyzed with the help of simulation results obtained using COMSOL Multiphysics software. The simulation program stops at the instant of discharge initiation due to convergence of the iterative algorithm of electron energy distribution function (EEDF; Hagelaar and Pitchford, Reference Hagelaar and Pitchford2005) and does not provide the results of plasma parameters (electron density, electron temperature, and electrical potential) during the plasma discharge process (Divya Deepak et al., Reference Divya Deepak, Joshi and Prakash2018a, Reference Divya Deepak, Joshi, Prakash and Pal2018b). It is also imperative to mention that in this research work, the plasma diagnostics has not been done during the experiments to determine plasma density and temperature.
Geometry and experimental setup
The experimental setup of ring electrode geometry is presented in Figure 1.
Fig. 1. Experimental setup of ring electrode-based APPJ.
The double ring electrode arrangement is used in this study consists of a long quartz tube with a nozzle, as shown in Figure 1. The nozzle of the quartz tube has two metal sleeves wound on it which acts as the ring electrodes. It is imperative to note that the green metal sleeve colored is connected to the pulsed power supply, while the other red-colored metal sleeve is grounded. The outer diameter and length of the quartz tube are 25 and 155 mm, respectively. The quartz tube thickness is 1.5 mm. Argon gas flows through the inlet of the quartz tube indicated as 1 (see Fig. 2) with an inner diameter is 22 mm. The outer diameter of the nozzle outlet is 6 mm, while its inner diameter is 3 mm which has been indicated as 2 (see Fig. 2). The length of the ring electrodes (indicated as 3) (see Fig. 2) and thickness are 18 and 4 mm, respectively. The ring electrodes indicated as 3 are separated by a distance of 3 mm and are clearly shown in Figure 2. A quartz sleeve of 7 mm outer diameter and 15 mm length was placed on the nozzle of the quartz tube (see Fig. 2) to observe the length of the plasma jet without the effect of the surrounding air.
Fig. 2. Geometry of ring electrode-based APPJ.
A high-voltage probe (Tektronix P 6015 A) (Bandwidth 0–75 MHz) was used to measure the applied voltage. The applied currents were measured using Rogowski-type Pearson current monitor (Model 110) (0.1 V/A, 1 Hz–20 MHz, 20 ns usable rise time) which was connected to a digital storage oscilloscope (Tektronix DPO 4054, bandwidth: 500 MHz). Experiments were implemented using argon gas with a gas flow rate of 1 l/min in a hermetically sealed setup at atmospheric pressure and the further same gas flow rate was used for simulation analysis also. A typical cold plasma jet produced using ring electrode geometry is depicted in Figure 3.
Fig. 3. Cold plasma jet generated using a ring electrode configuration (Divya Deepak et al., Reference Divya Deepak, Joshi, Pal and Prakash2016).
Governing equations for discharge simulations
The mean electron energy and electron density are computed by solving a pair of drift-diffusion equations for the mean electron energy and electron density.
$$\displaystyle{\partial \over {\partial t}}\lpar n_{\rm e}\rpar + \nabla \cdot \lsqb {-}n_{\rm e}\lpar {\rm \mu }_{\rm e}\cdot E\rpar -D_{\rm e}\cdot \nabla n_{\rm e}\rsqb = R_{\rm e}$$
$$\displaystyle{\partial \over {\partial t}}\lpar n_{\rm \varepsilon }\rpar + \nabla \cdot \lsqb {-}n_{\rm \varepsilon }\lpar {\rm \mu }_{\rm \varepsilon }\cdot E\rpar -D_{\rm \varepsilon }\cdot \nabla n_{\rm \varepsilon }\rsqb + E\cdot \Gamma _{\rm e} = R_{\rm \varepsilon }$$
$$\Gamma _{\rm e} = {-}\lpar {\rm \mu }_{\rm e}\cdot E\rpar n_{\rm e}-D_{\rm e}\cdot \nabla n_{\rm e}$$Here, n e is the electron density, n ɛ is the electron energy density, μe is the electron mobility, με is the electron energy mobility, E is the electric field, R e is the electron rate expression, R ε is the energy loss/gain due to inelastic collisions, D e is the diffusion coefficient of electrons, D ε is the electron energy diffusivity, and Γe is the electron flux.
Plasma chemistry using rate coefficients is used to determine the source coefficient in the equations mentioned above.
In the case of rate coefficients, the electron source term is specified by
$$R_{\rm e} = \sum\limits_{\,j = 1}^M {x_jk_j} N_nn_{\rm e}$$were M denotes the reactions that contribute to the growth or decay of electron density.
The mole fraction of the target species for reaction j is given by xj, the rate coefficient for reaction j is given as kj, and the total neutral number density is Nn. The electron energy loss is obtained by summing entire collisional energy loss over all reactions and written as follows:
$$R_{\rm \varepsilon } = \sum\limits_{\,j = 1}^P {x_jk_j} N_nn_{\rm e}\Delta {\rm \varepsilon }_j$$where Δεj is the energy loss from reaction j and P denotes the inelastic electron–neutral collisions. In general, P ≫ M. The rate coefficients can be computed from cross-section data by the following integral
$$k_j = {\rm \gamma }\int_0^\infty {{\rm \varepsilon }{\rm \sigma }_j\lpar {\rm \varepsilon }\rpar f\lpar {\rm \varepsilon }\rpar d{\rm \varepsilon }} $$where γ = (2q/m e)1/2, m e is the mass of an electron, ε is energy (V), f is the electron energy distribution function, and σj is the cross section of collision.
In this case, the EEDF is assumed Maxwellian because in DBD-based plasmas, the degree of ionization is high and as a consequence, the electron–electron collisions drive the distribution toward the Maxwellian shape. The EEDF can be computed by solving the Boltzmann equation. The evolution of the distribution function, f, in a six-dimensional phase space is expressed as follows:
$$\displaystyle{{\partial f} \over {\partial t}} + {\rm \nu }\Delta f-\left({\displaystyle{{eE} \over m}} \right)\nabla _{\rm \nu } = C\lsqb f\rsqb $$where f is the electron distribution function, v is the velocity coordinate, e is the elementary charge, m is the electron mass, E is the electric field, ∇v is the velocity–gradient operator, and C represents the rate of change in f due to collisions. In the code, EEDF is assumed to be almost spherically symmetric, so the series can be truncated after the second term (a so-called two-term approximation) (Divya Deepak et al., Reference Divya Deepak, Joshi and Prakash2018a, Reference Divya Deepak, Joshi, Prakash and Pal2018b). An anisotropic perturbation is also applied due to inelastic collisions under consideration. A detailed explanation of two-term approximation is given elsewhere (Hagelaar and Pitchford, Reference Hagelaar and Pitchford2005).
The electron diffusivity, energy mobility, and energy diffusivity are calculated from the electron mobility using the following equations:
$$D_{\rm e} = {\rm \mu }_{\rm e}T_{\rm e}\comma \quad{\rm \mu }_{\rm \varepsilon } = \left({\displaystyle{5 \over 3}} \right){\rm \mu }_{\rm e}\comma \quad D_{\rm \varepsilon } = {\rm \mu }_{\rm \varepsilon }T_{\rm e}$$where T e is the temperature of the electron.
$${\rm Mean}\,{\rm electron}\,{\rm energy}\,{\rm \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\leftharpoonup}$}}\to {\varepsilon } } = \displaystyle{{n_{\rm \varepsilon }} \over {n_{\rm e}}}$$
$${\rm Electron}\,{\rm temperature}\;T_{\rm e} = \left( {\displaystyle{2 \over 3}} \right){\rm \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\leftharpoonup}$}}\to {\varepsilon } }$$and electrostatic potential is given by the expression,
$$-\nabla {\rm \varepsilon }_0.{\rm \varepsilon }_r\nabla V = {\rm \rho }$$where ρ is the space charge density and V is the applied sinusoidal electrical potential (Hagelaar and Pitchford, Reference Hagelaar and Pitchford2005).
Governing equations flow simulations
In this simulation study, the effect of the argon gas flow rate has been analyzed for the electron density and electron temperature again using COMSOL 5.2a. The fluid dynamics module contains equations, volume forces, and boundary conditions that are essential for modeling freely moving fluids employing the Navier–Stokes equations and solve for the pressure and the velocity field. The laminar flow interface is used primarily to model flows of comparably low Reynolds number. The interface solves the Navier–Stokes equations, for incompressible and weakly compressible flows (Divya Deepak et al., Reference Divya Deepak, Joshi and Prakash2018a, Reference Divya Deepak, Joshi, Prakash and Pal2018b). The single-phase fluid-flow interfaces are computed using Navier–Stokes equations, as mentioned below:
$$\displaystyle{{\partial {\rm \rho }} \over {\partial t}} + \nabla \cdot \lpar {\rm \rho }u\rpar = 0$$
$$\displaystyle{{{\rm \rho }\partial u} \over {\partial t}} + {\rm \rho }\lpar u \cdot \nabla \rpar u = \nabla\cdot \lsqb {-}pI + {\rm \tau} \rsqb + F$$
$${\rm \rho }C_P\left({\displaystyle{{\partial T} \over {\partial t}} + \lpar u\cdot \nabla \rpar T} \right) = {-}\lpar \nabla\cdot q\rpar + {\rm \tau }\colon S-\left({\displaystyle{T \over {\rm \rho }}} \right)\cdot \left({\displaystyle{{\partial \rho } \over {\partial T}}} \right)\left({\displaystyle{{\partial p} \over {\partial t}}} \right) + \lpar u\cdot \nabla \rpar p + Q$$
$$S = 1/2\lpar \nabla u + \lpar \nabla u\rpar ^T\rpar $$
$${\rm \tau } = 2{\rm \mu }S-2/3{\rm \mu }\lpar \nabla\cdot u\rpar I$$Here, density is given as ρ, τ is the viscous stress tensor, u is the velocity vector, p is the pressure, I is the identity matrix, F is the volume force vector, T is the absolute temperature, C p is the specific heat capacity at constant pressure, q is the heat flux vector, S is the strain-rate tensor, and Q contains the heat sources. A contraction between tensors (S and τ) is denoted by operator “:” and is denoted as
$$S\colon \tau = \sum_n {\sum_m {S_{nm} \tau _{nm}} } $$Boundary conditions
Electrons are lost to the wall because of the random movement in the interval of a few mean free paths of the wall and some electrons are acquired due to secondary emission, thus ensuing in the boundary condition for the electron flux:
$$-n\Gamma _{\rm e} = \lpar \lpar 1/2\rpar v_{\rm e}\rpar n_{\rm e}-\sum_{\rm p} {{\rm \gamma }_{\rm p}\lpar \Gamma _{\rm p}n\rpar } $$and the electron energy flux is computed as
$$-n\Gamma _{\rm \varepsilon } = \lpar \lpar 5/6\rpar v_{\rm e}\rpar n_{\rm \varepsilon }-\sum\nolimits_{\rm p} {{\rm \varepsilon }_{\rm p}{\rm \gamma }_{\rm p}\lpar \Gamma _{\rm p}n\rpar } $$The electrical potential at the external boundary of the DBD is 0 V. At the walls, argon metastable quench and transform to neutral argon atoms. Argon ions also transform back to neutral argon atoms while freeing secondary electrons. Where subscript p refers to different species such as ions or neutral species, electrons, n denotes the density of species, Γ is the flux vector, v e is the collision frequency, γp is the secondary emission coefficient, n e denotes the electron density, n ε is the electron energy density, and ɛp being the mean energy of the secondary electrons. The second term in Eq. (18) is the secondary emission energy flux.
The discharge simulations for the ring electrode geometry have been performed for both, with the flow of argon gas and the static mode. For the ring electrode geometry, the discharge simulations have been performed between time scales of a nanosecond to microsecond in a 99-time step. At the instant of discharge initiation, the simulation program stops due to the iterative algorithm of EEDF that converges. A snapshot of the meshing with a mesh area (1672 mm2) of ring electrode geometry implemented using the FEM is shown in Figure 4. The separation of space charge is resolved using the boundary layer meshing on the plasma volume. The snapshots presented in Figures 5, 6, 9, 10, 13, and 14 only depict the values of electron density, electron temperature, and electrical potential at the instant of discharge initiation only as the iterative algorithm of COMSOL Multiphysics software stops at the instant of discharge and cannot illustrate 2 or more order variance in parameter values in a single snapshot.
Fig. 4. The meshing of ring electrode-based APPJ using FEM.
Fig. 5. Electron density at 4.5 kV/25 kHz (static condition).
Fig. 6. Electron density at 4.5 kV/25 kHz (with flow).
Results and discussion
Results of simulation
In the preliminary set of simulations, supply frequency is increased from 10 to 25 kHz, whereas the supply voltage was fixed at 4.5 kV to understand its influence on plasma parameters. The distribution of density of electrons at the instant of the breakdown of Ar gas in a static mode and with the flow of Ar gas is represented in Figures 5 and 6.
The variation of electron density as the supply frequency is varied from 10 to 25 kHz is shown in Figure 7.
Fig. 7. Electron density versus supply frequency at 4.5 kV.
Figure 7 represents the variation of electron density values obtained for a change of supply frequency from 10 to 25 kHz at 4.5 kV. Figure 7 depicts that electron density rises with an increment of the supply frequency. However, it should be noted from Figures 5 and 6 that the electron density (static condition) is greater than electron density (With flow). Further, the electron density in the static mode is considerably greater in comparison to the electron density in the with flow condition. It is clearly seen from Figure 7 that for all supply frequencies, the electron density values in static condition is higher in comparison to with flow of gas. This primarily as a result of higher collision rate between argon atoms in static condition in comparison to with flow of gas (Raizer, Reference Raizer1991).
It is intriguing to note that the electron density growth is notably greater at higher supply frequency (25 kHz). This is because the electrons have lesser transit time (i.e., ~40 μs) subsequently resulting in ion buildup causing increased power consumption by ring electrode configuration. This rise in power consumption by the plasma at higher frequencies is supported from the experimental results of DBD-based argon plasma jet (see Table 1) that indicates that power consumed by the plasma has risen from 0.46 W at 4.5 kV/10 kHz to 0.88 W at 4.5 kV/25 kHz (Divya Deepak et al., Reference Divya Deepak, Joshi, Pal and Prakash2016).
Table 1. Experimental results of DBD-based argon cold plasma jet (Divya Deepak et al., Reference Divya Deepak, Joshi, Pal and Prakash2016)
In Figure 8, the supply frequency is kept constant at 25 kHz and the supply voltage has been varied from 3.5 to 6 kV, the corresponding values of electron density were noted for both conditions (with the flow of Ar gas and static mode). It is seen that electron density increases with increment in the value of supply voltage this phenomenon occurs as energetic electrons which are already available cause electron avalanche. At lower values of supply voltage, that is, 3.5 kV, the electron density in both cases (with flow and static mode) are having lower values as the seed electrons are not energetic enough to cause further secondary ionization, whereas, at higher supply voltage (6 kV), the electron density in static mode is greater in comparison to with flow of Ar gas due to the numerous collisions between them that happen in static mode (see Fig. 8).
Fig. 8. Electron density versus supply voltage at 25 kHz.
Further, it has been observed from experimental results depicted in Table 1 when the supply voltage is increased more than 5.5 kV, the power consumed by the plasma reduces as energetic electrons cause sufficient secondary ionization of argon atoms to sustain the plasma discharge, this fact is closely in agreement with the simulation results described in Figure 8 that shows highest electron density of 2.25 × 1016 m−3 which signifies that degree of ionization is considerably higher for sustaining the argon cold plasma.
Figures 9 and 10 depict the electron temperature distribution at the moment of the breakdown of Ar gas in static condition and with the flow at 4.5 kV/25 kHz. It is evident that at all supply frequencies, the electron temperature under static condition is greater than electron temperature with flow condition. This happens as more electrons are available in static discharge mode causing more collisions between them resulting in greater energy exchange between them resulting in higher electron temperature (Raizer, Reference Raizer1991). But, it is further observed that as the supply frequency is increased to 20–25 kHz, then the electron temperature rises significantly as compared to lower supply frequency (10–15 kHz), this phenomenon is due to ion buildup that causes more input power to be absorbed by the electrons.
Fig. 9. Electron temperature at 4.5 kV/25 kHz (static discharge condition).
Fig. 10. Electron temperature at 4.5 kV/25 kHz (with flow condition).
This increase in power consumed by the argon cold plasma can be seen from Table 1 where power consumed has risen from 0.82 W at 4.5 kV/20 kHz to 0.88 W at 4.5 kV/25 kHz. The seed electrons absorb energy that is accountable for further ionization of argon atoms and henceforth ensuing in the avalanche process. Subsequently, there is a growth in electron density and possibly the growth of luminous spots at cathode at supply frequencies of 15–25 kHz as described elsewhere (Raizer, Reference Raizer1991). It is additionally seen from Figure 7 that the electron density rises with increment in the frequency of supply, this phenomenon occurs as the already available energetic electrons cause further ionization of argon atoms. Since the density of electrons is comparatively less in with flow condition in comparison to static mode henceforth the energy exchanged also less between them, thereby resulting in low values of electron temperature. The relationship between supply frequency and electron temperature for the developed ring electrode geometry is shown in Figure 11 which depicts that the electron temperature rises with increment in supply frequency in both static mode and with the flow of Ar gas. The electron temperature is a function of mean electron energy, it relies on electron energy density and density of electrons.
Fig. 11. Electron temperature versus supply frequency at 4.5 kV.
Figure 12 represents the variation of electron temperature with supply voltage at f = 25 kHz. It has been observed from experimental results depicted in Table 1 when the supply voltage is increased more than 5.5 kV, the power consumed by the plasma reduces as energetic electrons cause sufficient secondary ionization of argon atoms to sustain the plasma discharge, this fact is closely in agreement with the simulation results described in Figure 8 that shows highest electron density of 2.6 × 1016 m−3 which signifies that degree of ionization is considerably higher for sustaining the argon cold plasma.
Fig. 12. Electron temperature versus supply voltage at 25 kHz.
Figures 13 and 14 represent the electrical potential that is around 1.1 kV at the instant of discharge initiation which seems to be relatively lower than the experimental conditions. The simulated plasma potential in COMSOL Multiphysics stops at the moment of the breakdown of Argon gas (i.e., discharge initiation) due to its iterative algorithm that converges, henceforth, it is not able to attain the values of the electrical potential of the experimental condition.
Fig. 13. Electrical potential at 4.5 kV/25 kHz (with flow).
Fig. 14. Electrical potential at 4.5 kV/25 kHz (static condition).
However, from the experimental results of Divya Deepak et al. (Reference Divya Deepak, Joshi, Pal and Prakash2016), it can be noted that initiation of discharge happens at nearly 1.02 kV (see Fig. 15) where the discharge current begins flowing. This result is closely identical with the simulation results of electrical potential with the flow of Argon gas in Figure 13. Moreover, the corresponding electron density with the flow of Argon gas is 9 × 1015 /m−3 (see Fig. 6) which shows discharge initiation. Henceforth, our result of the simulation is in agreement with the experimental results (Divya Deepak et al., Reference Divya Deepak, Joshi, Pal and Prakash2016). From Figures 13 and 14, it can be observed that the electrical potential is ensuing from ring electrode-1 to ring electrode-2 at 4.5 kV/25 kHz at the condition with the flow of Ar gas and static condition. The relationship between supply frequency and electrical potential at fixed applied voltage 4.5 kV is shown in Figure 15.
Fig. 15. V–I characteristics at 4.5 kV/25 kHz (argon cold plasma jet) (Divya Deepak et al., Reference Divya Deepak, Joshi, Pal and Prakash2016).
The variation of electrical potential with supply frequency at 4.5 kV is shown in Figure 16. It is clear now that at lower supply frequencies, the electrical potential reached for discharge initiation is identical for both static mode and with the flow of gas. Consequently, at a supply frequency of 10 kHz, the electrical potential desirable for discharge initiation is nearly at the same voltage level for both cases of with flow of Ar gas and in static modes. At these low supply frequencies, the electrons tend to have a higher transit time of ~100 μs. Nevertheless, at a supply frequency of 20–25 kHz, the electrons possess lesser transit time ~40 μs owing to ion buildup, so the electrical potential reached for discharge initiation is becoming noticeable which may also result in additional power consumption.
Fig. 16. Electrical potential versus supply frequency at 4.5 kV.
Summary
From our simulation results at 4.5 kV and different supply frequencies, it is observed that electron density grows with supply frequency for both (with the flow of argon gas and static condition). However, it is observed that electron density is comparatively higher in the case of the static condition of argon gas than with flow at all supply frequencies since more argon atoms are available for ionization (see Figs 5 and 6). With the increment of supply frequency in both conditions of with flow of argon gas and static mode, the electron temperature rises (see Fig. 7). From the simulation results at f = 25 kHz and different supply voltages, it is seen from Figure 8 that electron density rises with increment in supply voltages. This is because at higher supply voltages, the initially available seed electrons absorb additional energy resulting in electrons that are energetic which causes further ionization of argon atoms.
But it is noted that the electron temperature is more for static condition compared to with flow of argon gas (see Figs 9 and 10) as the particle density is lesser in case of with flow of argon. Thus, lesser collisions and lesser energy exchange (between electrons and argon atoms) in case of with flow of gas which results in lower electron density and lower electron temperature compared to static discharge condition. Further, the electron temperature rises with increment in supply frequency in both static mode and with the flow of Ar gas (see Fig. 11). Also, electron temperature also increases with an increase in supply voltage. It is observed that when the supply voltage is incremented more input power is dumped into the plasma resulting in more energetic electrons being available to cause secondary ionization (see Fig. 12).
Electrical potential rises with increment in supply frequency as more input power is dumped into the plasma as supply frequency increases. It is observed that electrical potential reached is less in the case of with flow of argon gas in comparison to the static mode of discharge (see Fig. 16).
It is also observed that the simulation data has been closely in agreement with the experimental results (Divya Deepak et al., Reference Divya Deepak, Joshi, Pal and Prakash2016) depicted in Table 1 where the rise in power consumed by the argon cold plasma from 0.46 W at 4.5 kV/10 kHz to 0.88 W at 4.5 kV/25 kHz has been justified with Figure 7 which clearly depicts the growth in electron density due to ion buildup that resulting in more consumption of input power. Also, from Figure 11, the increase in electron temperature at higher supply frequencies (20–25 kHz) due to more power absorbed the seed electrons is justified with the experimental results in Table 1 that clearly indicate that power consumed by the argon plasma increased from 0.82 W at 4.5 kV/20 kHz to 0.88 W at 4.5 kV/25 kHz.
Further, it has been observed from experimental results depicted in Table 1 that when the supply voltage is increased more than 5.5 kV, the power consumed by the plasma reduces as energetic electrons cause sufficient secondary ionization of argon atoms to sustain the plasma discharge, this fact is closely in agreement with the simulation results described in Figure 8 that shows highest electron density of 2.6 × 1016 m−3 which signifies that degree of ionization is considerably higher for sustaining the argon cold plasma.
It is imperative to further note that the discharge initiation in the argon plasma occurs at 1.02 kV at 4.5 kV/25 kHz which closely matches the simulation results depicted in Figure 13 where the electrical potential is 1.02 kV.
These simulation and experimental results of DBD-based argon cold plasma jet would be useful for designing and operating ring electrode configuration without any arcing phenomenon. Further, the DBD-based argon cold plasma jet could be used for various biomedical applications. However, this will require in vitro biomedical laboratory testing and study of biomedical changes and other sterilizing effects to establish the relevance.
Acknowledgment
The Director, CSIR-Central Electronics Engineering Research Institute, Pilani is gratefully acknowledged for providing necessary lab facilities in performing these experiments
