Effect of rotating liquid samples on dynamic propagation and aqueous activation of a helium plasma jet

Han XU (徐晗),Shaoshuai GUO (郭韶帅),Hao ZHANG (张浩) and Kai XIE (谢楷),∗

1 School of Aerospace Science and Technology,Xidian University,Xi’an 710071,People’s Republic of China

2 State Key Laboratory of Electrical Insulation and Power Equipment,Centre for Plasma Biomedicine,Xi’an Jiaotong University,Xi’an 710049,People’s Republic of China

Abstract This paper aims to explore the effects of a rotating plasma-activated liquid on the dynamic propagation and biomedical application of a helium plasma jet.The spatial distribution of reactive species and the associated physico-chemical reactions are altered by the rotating liquid,which shows a significant weakening in the axial propagation of the plasma bullet and a strengthening in its radial expansion at the liquid surface.The phenomenon is prompted by the nonzero rotational velocity of the liquid and is regulated by airflow,target distance and liquid permittivity.The concentrations of aqueous reactive species,especially OH and O- 2 ,and the inactivation effectiveness on cancer cells are weakened,indicating that a rotating liquid is not conducive to water treatment of the plasma jet although the treatment area of the plasma jet increases dynamically.This finding is of significance for the plasma–liquid interaction and the biomedical-related applications of plasma jets.

Keywords: atmospheric-pressure plasma jet,plasma–liquid interaction,reactive species,biological inactivation function

1.Introduction

When using cold atmospheric-pressure plasmas in biological applications,the tissue being treated is often covered by a thin liquid layer,making plasma–liquid interactions inevitable and crucial [1–3].The ability to deliver to the liquid surface a high-density plasma,a high electric field and surface bombardment by ions well beyond the confines of the generating electrodes ensure that cold plasma jets are competitive technologically [4,5].However,their relatively small size(∼mm2)does limit the activation efficiency of plasma jets on liquids,prohibiting applications that necessitate large-area processing [6,7].Recent reports have highlighted that a jet array or electromagnetic control helps to improve the plasma range and chemical activity of a plasma jet [8,9].However,costs and process requirements of large-scale plasma jets increase considerably because of an increase in consumption of the working gas and discharge inhomogeneity [10].The treatment area of plasma jets also increases dynamically when the the jet source or the target being treated are continuously moving.This method appears to be more flexible in overcoming the disadvantages of the small size of the plasma jet without additional process costs.Nevertheless,given the plasma–liquid interaction,a number of questions arise: how does dynamic treatment affect the discharge characteristics and discharge products of plasma jets and is this method suitable in raising the activation efficiency of plasma jets on liquids? These issues are important for the study of plasma jet–liquid interaction and biomedical-related applications.However,as far as we know,there are no experimental or simulation studies concerning plasma jet treatments of dynamic aqueous solutions.The corresponding changes in plasma–liquid interactions and the underpinning physics remain unclear.This paper reports on our experimental study of the influence of rotation of the liquid on the dynamic characteristics of the plasma jet,the chemical reactivity of radicals in liquids and the effect in biomedical applications.

2.Experimental setup and methods

In the experiments,a plasma jet (shown in figure 1(a)) was formed emanating from a hollow coaxial stainless-steel needle [high voltage,inner diameter (ID) 2 mm,outer diameter(OD)2.5 mm]inside a quartz tube(ID 3 mm,OD 4 mm)and a ring copper foil (grounded,5 mm width) placed around the tube wall.The electrode was 20 mm away from the exit of the glass capillary.A steady flow of helium(99.999%)at a rate of 2 or 3 l min−1passed through the inner tube of the highvoltage electrode.Five samples (6 ml),namely deionized water,cell culture medium,physiological saline solution,absolute ethanol and acetonitrile,were placed in separate quartz Petri dishes (ID 35 mm,depth 10 mm).Three gap distances,specifically,4,6 and 10 mm,were used between the liquid surface and the jet nozzle.The Petri dish was centred on an insulated rotating platform,the speed of which was adjustable between 3 and 180 rpm.The jet nozzle was offset from the centre of the Petri dish to ensure that the contact point between the plasma plume and liquid was constantly moving as the liquid rotated (figure 1(a)).

Electrical diagnostics were performed with an oscilloscope(Tektronix DPO3000),a high-voltage probe(Tektronix P6015A)and a current monitor(Pearson 2877).Ranging over 200 to 800 nm,optical emission spectra at the gas–liquid interface were obtained using a spectrometer (Ocean Optics MAYA2000).A camera (Nikon D7000) with an exposure time of 2 s was used to take discharge images of the plasma jet.The propagation dynamics of the ionization wavefront in the tube,air gap and its interaction with the liquid surface were captured using an intensified charged-couple device(Princeton Instruments,PI-Max3) with an exposure time of 20 ns.Chemical fluorescent probes were used to measure concentrations of long-lived species in the plasma-activated liquid using a microplate reader(Thermo Scientific Varioskan Flash Reader) with the Amplex Red reagent for H2O2and a Griess reagent kit forNO2-andNO3-[11].With X-band electron spin resonance equipment (EMX,Bruker GmbH),spin-trapping was employed to determine the radical species;here 5,5-dimethyl-1-pyrroline-N-oxide (DMPO;Dojindo,1 mM) and 1-hydroxy-2,2,6,6-tetramethyl-4-oxo-piperidine(TEMPONE-H;Enzo,0.1 mM) were used to react with OH,O2-and ONOO−[12].

In the tumour-cell inactivation experiments,A375 melanoma cells were cultured in RPMI 1640 medium supplemented with 10% foetal bovine serum,50 μg ml−1streptomycin and 100 U ml−1penicillin at 37 °C with humidified air and 5%CO2.Six millilitres of the medium at a concentration of 2 × 105cells added to the Petri dish was exposed to the plasma jet for 6 min.After plasma jet treatment and recultivation,the cell viability of the A375 cells was evaluated using a CellTiter-Glo®luminescent cell viability assay kit (Promega,USA) with a microplate reader.The cell treatment was repeated at least three times using the same procedure.All the data are presented as mean ± the standard deviation (SD).

3.Results and discussions

The electrical properties of a plasma jet interacting with deionized water positioned 6 mm from the jet nozzle and with a gas flow of 2 l min−1are presented in figure 1(b).The green rectangle corresponds to the interval of the spatiotemporal

resolved development of the ionization wavefront during generation,propagation,interaction with the liquid surface and gradual disappearance (corresponding ICCD images in figures 2 and 3).The duration of each positive voltage halfcycle is approximately 13 μs.The development of the ionization wavefront starts from the ring electrode around the moment,marked ‘1’,when the positive current peaks.The subsequent small negative current peak,marked as ‘2’,corresponds to the moment when the ionization wavefront reaches the liquid surface [13].In particular,the position of current peak ‘2’ changes with the velocity of the ionization wavefront for different gap distances,gas flows and other parameters during the plasma–liquid interaction.Current peak‘2’ does not appear while the plasma jet is expanding freely[14].Moreover,during one discharge cycle,each positive current peak ‘1’ is much stronger than the negative current peak (marked as ‘3’),because the dielectric barrier discharge is asymmetric and the enhanced local positive electric field stems from the lower mobility of the positive charge[15].The discharge power was about 300 mW for all settings (excitation voltage fixed at 5 kVpp),although the position of the current peak varies under different discharge conditions.

Figure 2(a) shows the recorded development of the ionization wavefront during the plasma–liquid interaction for different rotational speeds of the liquid.The dotted line in the time-resolved images represents the position of the liquid surface.The results show that the shape of the ionization wavefront is irregular because,when propagating in the quartz tube,its radius gradually decreases as the wave moves away from the ring electrode.The ionization wavefront crosses the entire quartz tube(20 mm)in approximately 2 μs,indicating a velocity of ∼10 km s−1,and its propagation velocity in the air gap increases further [16].When the ionization wavefront (plasma bullet) interacts with the water surface at t=4.5 μs,its volume expands rapidly and its brightness increases significantly;a very weak cathode glow then appears and gradually disappears.Most importantly,there are distinct differences in the interaction behaviour between the plasma bullet and the water surface for different rotational speeds of the liquid.The expanding plasma bullet appears to show significant axial propagation at the water surface if the water being treated is stationary;a radial spreading of the ionization wave over the water surface is not observed,in agreement with previous research(figure 2(a-1))[13].When the treated water rotates with speeds below 20 rpm,the propagation behaviour of the plasma bullet remains basically unchanged,being predominantly axial at the water surface(in figure 2(a-2),the speed is 6 rpm).However,when the rotational speed of the treated aqueous solution is greater than 50 rpm,the axial propagation of the plasma bullet significantly weakens,whereas at the gas–liquid interface its radial propagation strengthens (in figure 2(a-3),the speed is 60 rpm).Of course,the rotational speed of the liquid should not be too high (above 120 rpm),otherwise visible splashing or vortices appear on the liquid surface.The above rotational speed data were obtained from our repeated tests.It was noted that the changes in the propagation behaviour of the plasma bullet happen randomly when the treated water rotates with speeds of 20–50 rpm.In addition,we also observed no changes in the propagation behaviour when the plasma jet impinges on the central position of the water solution and is independent of the direction of rotation.Since propagation of the ionization wavefront represents the development trajectory of the reactive species,the changes in the propagation behaviour of the plasma bullet indicate that the spatial distribution of reactive species and the associated physico-chemical reactions at the water surface,for example electronic excitation and ionization,are affected by the rotating liquid [17,18].

Figure 1.(a) Schematic of the plasma jet–liquid experimental setup(HV,high voltage;ICCD,intensified charged-couple device;Mass Flow Controller (MFC),).(b) I–V characteristics of the plasma jet interacting with deionized water at a target distance of 6 mm and a gas flow of 2 l min−1.

Figure 2.(a)Spatiotemporal development of the ionization wavefront during the plasma–liquid interaction for different rotational speeds of the liquid(liquid sample deionized water;gap distance 6 mm,gas flow 2 l min−1).(b)Optical emission spectrum of the plasma jet at the gas–liquid interface for different liquid rotation speeds.(c) Discharge images during plasma jet treatment of deionized water for stationary and rotating liquid.

Figure 3.Spatiotemporal development of the ionization wavefront during the plasma jet–liquid interaction under different experimental conditions compared with those of figure 2:(a) increasing gas flow,(b) various gap distances,(c) various liquid samples.

The optical emission spectra of the plasma jets show that although the propagation behaviour of the plasma bullet at the liquid surface changes with different rotational speeds,the overall radiation intensity of the plasma jet is basically unaffected (figure 2(b)).The emission spectra are rich in hydroxide species and nitrogen species from the air,as well as helium species.Based on a simplified collision–radiation model,the spatial-average electron density of the plasma jet is estimated to remain roughly constant at approximately 7.82 × 1011cm−3in all settings through the relative intensity ratio of the specific nitrogen emission lines 371.1 nm and 380.5 nm (I371.1nm/I380.5nm) [19].In addition,although the propagation of the plasma bullet varies axially and radially on the gas–liquid interface at higher rotational speeds,the overall changes in light intensity and the size of the plasma jet plume are also invisible to the naked eye(figure 2(c)).These results indicate that gas discharges of the plasma jet over the gas–liquid interface,corresponding to the generation of gaseous reactive species,are unaffected by the rotating liquid.This is manifested in that the propagation characteristics of the plasma jet plume at the macroscopic level and the species radiation intensity masked through spatial averaging remain unchanged.However,the charge distribution and related chemical reactions at the microscopic level are definitely influenced by the rotating liquid,which corresponds to a weakening in the axial propagation of the plasma bullet at the liquid surface.

To regulate the propagation behaviour of the plasma bullet at the liquid surface and understand the underpinning physics,further measurements of the dynamic propagation of ionization wavefront under different gas flows,gap distances and liquid samples were performed (figure 3).It should be noted that the form of the liquid surface was not affected by the rotating liquid,i.e.the liquid surface was horizontal in all settings.Figure 3(a)shows the development of the ionization wavefront when the deionized water is posted at the same distance as in figure 2(a),i.e.6 mm away,but the gas flow is increased from 2 to 3 l min−1.The results show that the size and velocity of the ionization wavefront increase significantly with the increased working gas flow.Specifically,the propagation velocity of the ionization wavefront in the quartz tube increases nearly two-fold to 20 km s−1,although the velocity of the airflow(below 10 m s−1)is much smaller than that of the ionization wavefront [20].Such a change shortens the propagation time of the ionization wavefront and advances the moment when the ionization wavefront strikes the water surface to t=3.5 μs.The rotating liquid with a rotational speed of 60 rpm also makes the plasma bullet appear and expand radially out over the gas–liquid interface,whereas the axial propagation of the plasma bullet becomes slightly stronger with increasing gas momentum (compared with figure 2(a-3)).The propagation dynamics of the ionization wavefront with a gas flow of 2 l min−1for different gaps(4 mm and 10 mm) are presented in figure 3(b).The changes in the air gap have little effect on the dynamic development of the ionization wavefront,compared with figure 2(a),except that the increased air gap (10 mm) prolongs the propagation duration by approximately 0.5 μs.The interaction behaviours between the plasma bullet and the water surface with the rotating liquid are similar to the results in figure 2(a) but,more importantly,the decreased air gap(4 mm)also enhances the axial propagation of the plasma bullet to a certain extent.From schlieren results of plasma jet–liquid interaction in previous work,the axial momentum increased significantly with increasing gas flow.In addition,the reduced air gap distance also suggests that the longitudinal extension of the gas momentum increases,and the lateral extension decreases above the liquid surface during the plasma jet–liquid interaction [13,21].Therefore,on the basis that the propagation behaviour of the plasma bullet at the liquid surface changes with air gap and airflow,the radial expansion of the plasma bullet induced by the rotating liquid may be attributed to the change in flow field at the gas–liquid interface.The contact point of the plasma jet and the treated liquid is moving at a higher speed with the rotating liquid.The change in the flow field at the gas–liquid interface arising from the viscous resistance and rotational force affects the distribution of the ionization waves,specifically,the distribution of reactive species and associated physico-chemical reactions,which contributes to the radial dispersion of these waves over the liquid surface.

Figure 4.Variation of (a) the concentrations of aqueous short-lived species OH and O2- (DI,deionized).(b) The concentrations of aqueous long-lived species H2O2 and ONOO−.(c)The cell viability of A375 melanoma cells after 6 min of plasma jet irradiation under different experimental conditions (n=3,*p < 0.05,**p < 0.01,***p < 0.001).

Furthermore,we compared the changes in the propagation behaviour of the ionization wavefront induced by the rotation of various common liquids:cell culture medium,physiological saline solution,absolute ethanol and acetonitrile.The working gas flow and air gap were set to 2 l min−1and 6 mm,respectively,consistent with figure 2(a).The results show that weakening of the axial propagation of the plasma bullet and strengthening of the radial expansion occur at the gas–liquid interface for all liquids with rotational speed of 60 rpm.However,this phenomenon is more pronounced in two organic solutions,acetonitrile and ethanol (see the 4.5 μs frames in figure 3(c)).By categorizing the liquid properties,the dielectric constants of absolute ethanol (∼24) and acetonitrile(∼38) are much smaller than that of deionized water.Therefore,the prominent radial expansion of the plasma bullet in organic solutions may be attributed to a smaller permittivity at the target surface that contributes to the horizontal propagation of the electric field,which promotes the formation and propagation of surface ionization waves [13,22].

Finally,we studied whether the aqueous activity and biomedical effectiveness of the plasma jet is improved by the rotating liquid.The concentrations of aqueous reactive species classified with short-lived species(OH and O2-)and longlived species (H2O2,ONOO−,NO2-andNO3-) characterize aqueous activity (figures 4(a) and (b)),whereas cell viability of A375 melanoma cells after plasma jet treatment evaluates application effectiveness (figure 4(c)).The plasma–liquid interaction time was 6 min.Note that the concentrations of aqueousNO2-andNO3-induced by the helium plasma jet are too low (around 2 μM) to effectively determine the variation trends,which is in line with previous reports [23,24].As shown in figure 4(a),the concentrations of aqueous shortlived species (OH and O2-) decrease significantly at higher rotational speeds (60 rpm) compared with stationary water,but remain basically unchanged at lower speeds (6 rpm).For instance,the concentrations of aqueous OH and O2-decrease by 10.7 μM(18.6%)and 2.1 μM(26.2%),respectively,when the treated deionized water rotates with speed of 60 rpm compared with stationary water (gas flow 2 l min−1,gap distance 6 mm).Similar phenomena are also found when the helium flow rate,air gap or liquid composition varies,but more importantly the reduction in the concentrations of aqueous OH and O2-becomes weaker with increasing gas flow and decreasing gap distance.Specifically,the concentration of aqueous OH only decreases by 6.5 μM (10.3%) and 3.8 μM(7.5%) at higher rotational speeds (60 rpm) compared with stationary water,respectively,when the helium gas flow is increased to 3 l min−1or the air gap distance decreased to 4 mm.

According to recent reports,90% of the aqueous OH is due to the deposition of positive ions at the gas–liquid interface and 10% to the photolysis of water by (vacuum)ultraviolet photons.Aqueous O2-is derived in the capture of solvated electrons by dissolved O2[24,25].Thus,the concentrations of these two aqueous reactive species are closely related to the density distribution and chemical reaction of charged particles at the water surface.The weakening in the axial propagation of the plasma bullet at the liquid surface is caused by the rotating liquid,which likely suppresses the electron solvation and charge exchange at the gas–liquid interface,and even lowers the chemical reaction rate with regard to the charged particles [17].This is why the concentrations of aqueous OH and O2-decrease significantly at a higher rotational speed.Meanwhile,the axial propagation of the plasma bullet at the liquid surface is enhanced to a certain extent if the helium gas flow is increased or the air gap distance decreased,corresponding to a less significant decrease in the concentrations of aqueous OH and O-2.The finding is of significance for the regulation of reactive species distribution and associated physico-chemical reactions at the gas–liquid interface during the plasma jet–liquid interaction.

Interestingly,a slight downwards trend is also found in the variation in concentration of aqueous H2O2,but the concentration of aqueous ONOO−remains basically unaffected by the rotation of the aqueous solution (figure 4(b)).Specifically,when the treated deionized water rotates with a speed of 60 rpm,the concentration of aqueous H2O2decreases by 5.2 μM,3.6 μM and 1.9 μM,respectively,corresponding to three experimental settings: (1) gas flow 2 l min−1,gap distance 6 mm;(2) gas flow 3 l min−1,gap distance 6 mm;(3)gas flow 2 l min−1,gap distance 4 mm.By comparing the variations of aqueous H2O2and OH,the concentration of aqueous OH decreases approximately twice as much as that of aqueous H2O2.Since the primary loss pathway for aqueous OH is the reaction forming aqueous H2O2(OH+OH→H2O2),we consider that the decrease in the concentration of aqueous H2O2is caused by the change in the concentration of aqueous OH [26].The majority of aqueous H2O2and ONOO−derives from the solvation of gaseous neutral particles (H2O2and NxOy),and gaseous H2O2and NxOyare not formed on the surface of the spreading plasma so these molecules must convect and diffuse from more remote locations thereby taking a few microseconds[27].The density of gaseous particles,which is closely related to the propagation of the plasma jet in the atmosphere,remains essentially unchanged,corresponding to the experimental results on the discharge characteristics (figures 2(b) and (c)).In addition,this process is less sensitive to the charge distribution and chemical reaction around the gas–liquid interface,which leads to the concentrations of aqueous H2O2and ONOO−being less affected by the liquid rotation.

Moreover,the change in the viability of melanoma cells shows that solution activation of the plasma jet through the rotating liquid is weaker than that for stationary liquids(figure 4(c)),which correlates with the reduction in concentrations of aqueous OH and O2-which can induce strong oxidative stress in biological tissue [28].Specifically,the inactivation of cancer cells in the plasma jet is reduced by approximately 7%–15% at a rotational speed of 60 rpm.This result solves the problem we posed earlier,namely a rotating liquid is not conducive to water treatment of the plasma jet although the treatment area of the plasma jet increases dynamically.

4.Conclusion

In summary,the purpose of our study was to explore the effects of a rotating plasma-activated liquid on the dynamic characteristics and biomedical application of a helium plasma jet.The results indicate that the propagation characteristics of the plasma plume at the macroscopic level and the species radiation intensity masked through spatial averaging remain unchanged by the rotating liquid,but the axial propagation of the plasma bullet at the liquid surface is significantly weakened.The behavioural changes of the plasma bullet can be ascribed to changes in reactive species distribution and associated physico-chemical reactions at the gas–liquid interface,which are regulated by the rotational velocity of the liquid and airflow,as well as target distance and liquid permittivity.The high-speed rotating liquid hinders the generation of aqueous reactive species,where the drop in the concentrations of aqueous OH and O2-is more significant than that of aqueous H2O2and ONOO−.Cancer cell inactivation by a plasma jet is diminished due to the rotating liquid despite the increase in treatment area,which indicates that a rotating liquid is not conducive to the biological inactivation functions of plasma jets.Although this finding does not positively promote water treatment by plasma jet,it has important significance for the study of the interaction between plasma jets and dynamic aqueous solutions.

Acknowledgments

This work was supported by National Natural Science Foundation of China (No.52107162),the Science and Technology Projects of Shaanxi Province (No.2022CGBX-12) and the Science and Technology Projects of Xi’an City(No.2021SFCX0005).

ORCID iDs