Introduction

Ozone is well established as an important oxidizing agent, used in a wide range of industrial applications, such as disinfection of drinking water, treatment of wastewater, extension of food storage lifetimes, destruction of gaseous pollutants, removal of odours, biomedical applications, etc. As a consequence of its attractive properties, many methods for enhancing its production rate by means of electrical discharges have been investigated. These methods involve selection of the proper type of discharge and optimization of its parameters [1–6], application of various catalysts, ferroelectrics, dielectrics in the discharge chamber [7–12], or applications of other fields to the discharge [13–15].

In order to enhance needle-to-mesh DC corona discharge ozone production we proposed in [16] a simple method based on insertion of the hollow needle electrode into the dielectric tube with a supply of air into the discharge through the needle. Supply of air through the hollow needle electrode, however, requires compression of air, which represents a drawback in some applications. In order to eliminate this drawback we tried to use reverse supply of air to the discharge through the mesh electrode. In this case, the pressure of air before supply to the discharge could be decreased.

Supply of the gas flow on the corona discharge is a well-known technique, which can dramatically affect discharge physics [17–22]. Thus in [17] it was shown that gas flow oriented perpendicular to the pin-to plate electrode affects the threshold currents that mark corona-to-glow and glow-to-spark transitions. An important contribution to the study of the effects of external airflows on electric corona discharge in point-mesh configuration is presented in [18]. It was shown that airflow depending on its orientation tends to force ions to travel faster or slower, and it may have considerable effect on the electrical characteristics of the discharge. The effect of gas velocity on corona discharge characteristics was investigated in [19]. It was found that the flowing gas blows out the space charge downstream, changing the space charge distribution between the discharge electrodes. The effects of the magnitude and direction of airflow on the characteristics of the corona discharge generated in a wire-duct reactor were studied in [20]. Two air flow directions were attempted: one parallel to and the other normal to the discharge wire in the reactor. The results showed that DC and AC corona current–voltage characteristics depend significantly on the magnitude of the airflow in the reactor, whatever the direction of the airflow. In [21] it was shown that ozone and NOx generated by the corona discharge are effectively flushed by the airflow from the ionization region. The importance of the gas flow to ozone generation when the high voltage electrode was made of a metal mesh was shown in [22]. Because of the presence of the metal mesh electrode the gas is mixed, which improves both heat and mass exchange in the discharge chamber.

For the discharge with the needle electrode perpendicular to the mesh electrode we compared the effect of air supply to the discharge through the needle and the reverse supply of air through the mesh on the electrical characteristics and production of ozone of the discharge. The experiments were performed with a simple needle and with a needle inserted in the dielectric tube.

As mentioned above, placing a suitable catalyst into the discharge chamber can enhance discharge ozone production. For the needle-to-mesh electrode system when the needle is inserted in the dielectric tube, the effect of the catalyst on ozone production of the discharge has never been studied. We therefore investigated the effect of placing a TiO2 photocatalyst on the mesh electrode for various positions of the needle electrode with respect to the end of the dielectric tube and for supply of air into the discharge through the needle.

The discharge produces the highest ozone concentrations just before its transition from the glow to the streamer regime. When this transition occurs discharge ozone production breaks down dramatically. We therefore recorded the voltages and currents with which this transition occurred for the discharge with supply of air through the needle electrode, for the discharge with reverse supply of air through the mesh, and for the discharge with TiO2 globules on the mesh for the simple needle and various positions of the needle electrode with respect to the end of the dielectric tube.

Experimental Arrangement

Figure 1 shows the experimental arrangement for studying the effect of various methods of air supply to the discharge chamber for discharge with the dielectric tube on the needle electrode and the effect of the TiO2 photocatalyst on the mesh electrode. This arrangement is a modification of that described in [16]. The electrode system was placed in a stainless steel discharge chamber with four openings. From the upper opening was inserted holder with the hollow needle electrode. From opposite opening, that is from the lower part of the discharge chamber was introduced another holder carrying the mesh. Depending on the way, how air was supplied to the discharge, the left and right openings were either closed or open.

Fig. 1
figure 1

Experimental arrangement

Full size image

As a cathode we used a stainless chromium-nickel steel hollow needle of outer diameter 0.9 mm and length 70 mm. The experiments were performed either with a simple needle, that is without the dielectric tube on the needle electrode, or with a corundum dielectric tube on the needle electrode. The outer diameter of the dielectric tube was 3.9 mm, the inner diameter was 1.0 mm, and its length was 31.2 mm.

The dielectric tube could be moved with respect to the tip of the needle electrode. As shown in Fig. 2, the position of the end of the dielectric tube relative to the tip of the needle electrode is expressed by the variable h. At h = 0, the tip of the needle matches the end of the dielectric tube, at negative values of h the needle electrode protrudes from the dielectric tube, and at positive values of h the needle is inserted inside the tube. The variable h varied from h = −3 mm to h = +2 mm during our experiments.

Fig. 2
figure 2

Coordinate system showing the position h of the end of the movable dielectric tube with respect to the needle tip

Full size image

A stainless steel mesh, which was used as an anode, was situated perpendicular to the needle electrode. The mesh had rhomboid cells with diagonals 0.60 and 0.50 mm long and 0.15 mm thick. The distance d between the tip of the needle electrode and the mesh was fixed at 8.3 mm.

In case of experiments with photocatalyst we placed on the mesh electrode cylindrical globules of TiO2 photocatalyst Aerolyst 7706. The diameter of globules was ~3 mm and they were ~4 mm in height. The anatase content of this photocatalyst was at least 70 %, the density was ~3.8 g/cm3, and the surface area was 40–50 g/m2.

As source of air we used an oil-free compressor, followed by a moisture trap and a mass flow controller (MFC). Pressure of air before supply to the discharge chamber was measured by a manometer P. Air into the discharge chamber was supplied in two ways. In the first case, which is shown in Fig. 1, air was supplied through the hollow needle electrode. Output of air was through the lower part of the discharge chamber, that is through the mesh. The left and right openings were closed. In the second case, reverse air supply, the needle was closed and air was supplied to the discharge chamber from the bottom, through the mesh electrode. As long as air could not leave the chamber through the needle, the exhaust from the discharge chamber was through the left and right openings, the exhaust pipes were connected and output ozone concentration was measured.

The Technix-regulated DC high-voltage power supply provided a voltage up to 30 kV. The needle, which was negatively biased, was ballasted by a resistor R = 6.9 MΩ. The discharge power was calculated as the product of the discharge voltage and current. The onset of streamers was detected on an oscilloscope by voltage pulsations as well as visually.

The ozone concentration at the output of the discharge chamber was measured by absorption of the 254 nm UV spectral line using an API 450 ozone monitor.

Experimental Results and Discussion

All experiments were performed with the same airflow through the discharge chamber of 1.8 slm. In order to approach the conditions at which commercial ozonizers function, the relative humidity of air was 25 %.

In the case of air supply to the discharge chamber through the needle electrode the pressure of air at the input of the needle was 1.39 × 105 Pa. In the case of air supply to the discharge chamber through the mesh electrode the pressure of air at the input was 1.05 × 105 Pa. In both cases pressure of air at the output from the discharge chamber was about 400 Pa higher than pressure in the laboratory.

Discharge Ozone Production

In this section, we present the results of ozone concentration and ozone production yield as a function of electrical power for various ways of air supply to the discharge and for the discharge with presence of TiO2 photocatalyst. These experiments were performed with the simple needle; with the needle, which protrudes from the dielectric tube; with the needle that matches with the end of dielectric tube, and with the needle, which is inserted in the dielectric tube.

All experiments have been conducted within the glow corona regime and, therefore, the last point in each figure corresponds to the maximum attainable power before transition to the streamer regime. To easily identify the experimental conditions, upper-case Latin (lower-case Greek) letters have been used to designate transition points in case of air supply through the needle (mesh) electrode. When the air was supplied through the needle with TiO2 globules on the mesh the transition points have been further indicated using lower case Latin characters. The alphabetic order of letters also helps to identify the configuration, or position of the needle tip with respect to the end of dielectric tube. Thus a, α and A indicates a simple needle; b, β and B a needle which protrudes 3 mm from the dielectric tube (h = −3 mm); c, γ and C a needle that matches with the end of the dielectric tube (h = 0 mm); d, δ and D a needle which is inserted 1 mm in the dielectric tube (h = +1 mm) and, finally, e and E a needle which is inserted 2 mm in the dielectric tube (h = +2 mm).

The effect of supply of air through the needle electrode or reverse supply of air through the mesh electrode on discharge ozone production is shown in Fig. 3. The data for h = +2 mm for supply of air through the needle electrode are also shown. The corresponding curve for h = +2 mm for the reverse air supply through the mesh electrode is not given, because it was not possible to ignite the discharge in the glow regime.

Fig. 3
figure 3

Concentration of ozone produced by the discharge for supply of air into the discharge through the needle electrode (empty symbols, dashed lines) and for reverse air supply through the mesh electrode (full symbols, full lines). Simple needle—stars; needle protrudes from the dielectric tube (h = −3 mm)—circles; needle that matches with the end of the dielectric tube (h = 0 mm)—up triangles; needle inserted in the dielectric tube (h = + 1; + 2 mm)—down triangles and squares respectively. Transition points from glow to the streamer discharge regime for the discharge with supply of air through the needle electrode are denoted by upper-case Latin letters and those for the discharge with reverse supply of air through the mesh electrode by lower-case Greek letters

Full size image

The first conclusion that can be drawn from Fig. 3 is that ozone concentration produced by the discharge at transition points is higher for supply of air through the needle than for reverse air supply; e.g. for the simple needle points A and α, for h = −3 mm points B and β; for h = 0 mm C and γ; for h = + 1 mm D and δ. Second, when the needle electrode is inserted in the dielectric tube (h = +1), ozone concentration is substantially higher for the supply of air into the discharge through the needle electrode than ozone concentration produced by the discharge with reverse supply of air through the mesh electrode. In the case of the simple needle, the needle that protrudes from the dielectric tube and the needle that matches with the end of the dielectric tube, the ozone concentration for small discharge power for the discharge with supply of air through the needle electrode is almost the same as for the reverse flow through the mesh electrode. For larger power ozone concentration is only slightly higher for the discharge with supply of air through the needle than for the reverse flow through the mesh electrode.

A comparison of the concentration of ozone produced by the discharge and that produced by the discharge with TiO2 globules on the mesh electrode for supply of air into the discharge volume through the needle electrode is shown in Fig. 4. This figure also shows corresponding transition points when the symbols used are the same as the symbols mentioned in the beginning of this section.

Fig. 4
figure 4

Concentration of ozone produced by the discharge (empty symbols, dashed lines) and concentration of ozone produced by the discharge with TiO2 globules on the mesh (semi-full symbols, full lines) for supply of air into the discharge through the needle electrode. Simple needle—stars; needle protrudes from the dielectric tube (h = −3 mm)—circles; needle that matches with the dielectric tube (h = 0 mm)—up triangles; needle inserted in the dielectric tube (h = + 1; + 2 mm)—down triangles and squares respectively. Transition points from the glow to the streamer discharge regime for the discharge without TiO2 are denoted by upper-case Latin letters and for the discharge with TiO2 by lower-case Latin letters

Full size image

The conclusion that can be drawn from Fig. 4 is that for supply of air through the needle electrode the placement of TiO2 globules on the mesh electrode increases ozone concentration produced by the discharge for all needle configurations tested.

For deeper insight into the effect of various ways of air supply to the discharge and TiO2 photocatalyst on the mesh on concentration of ozone produced by the discharge and ozone production yield we summarized these results in Figs. 5 and 6 for the two extreme situations. This first situation relates to the simple needle electrode, which is not inserted in the dielectric tube (Figs. 5, 6). The second situation relates to the needle electrode inserted in the dielectric tube with h = +1 mm (Figs. 7, 8).

Fig. 5
figure 5

Concentration of ozone produced by the discharge with the simple needle electrode together with transition points from the glow to the streamer discharge regime. Air supply through the needle—empty stars, A. Reverse air supply through the mesh—full stars, α. Air supply through the needle with TiO2 on the mesh—semi-full stars, a

Full size image
Fig. 6
figure 6

Ozone production yield for the discharge with the simple needle electrode together with transition points from the glow to the streamer discharge regime. Air supply through the needle—empty stars, A. Reverse air supply through the mesh—full stars, α. Air supply through the needle with TiO2 on the mesh—semi-full stars, a

Full size image
Fig. 7
figure 7

Concentration of ozone produced by the discharge for the needle electrode inserted in the dielectric tube with h = +1 mm together with transition points from the glow to the streamer discharge regime. Air supply through the needle—empty down triangles, D. Reverse air supply through the mesh—full down triangles, δ. Air supply through the needle with TiO2 on the mesh—semi-full down triangles, d

Full size image
Fig. 8
figure 8

Ozone production yield for the discharge with the needle electrode inserted in the dielectric tube with h = + 1 mm together with transition points from the glow to the streamer discharge regime. Air supply through the needle—empty down triangles, D. Reverse air supply through the mesh—full down triangles, δ Air supply through the needle with TiO2 on the mesh—semi-full down triangles, d

Full size image

For the discharge with the needle electrode inserted in the dielectric tube with h = +1 mm concentration of ozone produced by the discharge with supply of air through the needle, with reverse air supply through the mesh and with air supply through the needle with TiO2 on the mesh are shown in Fig. 7. The corresponding ozone production yield is shown in Fig. 8.

From Figs. 5, 6, 7 and 8 it can be concluded that the best performance of the discharge from the standpoint of ozone generation is the discharge with the supply of air through the needle electrode with TiO2 globules on the mesh when the needle electrode is inserted in the dielectric tube.

The explanation of the results concerning discharge ozone production could not be separated from the corresponding changes in electrical parameters of the discharge for various experimental conditions, which are presented in “Electrical Characteristics of the Discharge” section.

Electrical Characteristics of the Discharge

The decisive role in discharge ozone production is played by electrical parameters of the discharge. In Fig. 9 the V–A characteristics of the discharge are shown together with transition points from the glow to the streamer discharge regime for supply of air into the discharge volume through the needle electrode or for reverse air supply through the mesh electrode. In this figure there are no data for the air supply through the needle inserted in the dielectric tube with h = +2 mm, because for this h it was not possible to ignite the discharge in the glow regime. In this figure are also shown transition points for which the discharge changes from the glow to the streamer regime. The symbols used in this figure are the same as the symbols used in Fig. 3. Thus for the discharge with supply of air through the needle electrode A, B, C, D and E are used; for the discharge with reverse air supply through the mesh α, β, γ and δ are used.

Fig. 9
figure 9

V–A characteristics of the discharge together with transition points from the glow to the streamer discharge regime for supply of air into the discharge volume through the needle electrode (empty symbols, dashed lines, A, B, C and D ) and for reverse air supply through the mesh electrode (full symbols, full lines, α, β, γ and δ). Simple needle—stars; needle protrudes from the dielectric tube (h = − 3 mm)—circles; needle that matches with the dielectric tube (h = 0 mm)—up triangles; needle inserted in the dielectric tube (h = + 1 mm)—down triangles

Full size image

From comparison of the electrical characteristics of the discharge with supply of air through the needle electrode and for reverse air supply through the mesh electrode, which is presented in Fig. 9, the following conclusions can be drawn.

  • For the simple needle electrode and for the needle electrode that protrudes from the dielectric tube with h = −3 mm, the V–A characteristics of the discharge are practically the same for supply of air through the needle electrode and for the reverse air supply through the mesh.

  • For the needle electrode that matches with the dielectric tube (h = 0 mm) and for the needle inserted in the dielectric tube (h = +1; +2 mm) discharge with reverse air supply requires higher voltage for particular currents.

In Fig. 9 it can be seen that voltages at which transitions from the glow to the streamer discharge regime for a particular method of air supply into the discharge occur fit into a certain region. Thus the voltage region which corresponds to the transitions for the supply of air through the needle electrode embraces the letters A, B, C and D and the voltage region which corresponds to the transitions for the discharge with reverse air supply through the mesh embraces the letters α, β, γ and δ.

The results for transitions points can be summarized in the following way.

  • When there is reverse air supply through the mesh electrode transition from the glow to the streamer regime for simple needle or for any position of the needle in the dielectric tube takes place at lower current in comparison with the discharge with the supply of air through the needle electrode. This is seen from comparison of the transition currents corresponding to the points A and α; B and β; C and γ and D and δ.

  • Insertion of the needle electrode into the dielectric tube decreases the current with which transition from the glow to the streamer regime occurs. The deeper the needle electrode is inserted the lower the current.

  • Voltages at which transition from the glow to the streamer regime occur are not substantially affected by insertion of the needle electrode into the dielectric tube.

  • For the discharge with simple needle electrode and for the discharge with the needle electrode that protrudes from the dielectric tube with h = −3 mm, discharge onset voltages for supply of air through the needle and that through the mesh are practically the same.

  • For the discharge with the needle electrode that matches with the dielectric tube or protrudes from the dielectric tube with h = +1, +2 mm, the discharge onset voltage is higher for the discharge with reverse air supply.

The effect of the TiO2 photocatalyst on the mesh electrode on electrical parameters of the discharge for various positions of the dielectric tube with respect to the needle electrode, with the supply of air through the needle electrode, is shown in Fig. 10. The discharge with the same supply of air into the discharge without TiO2 on the mesh is used as a reference. In this figure transition points as well as voltage regions in which transitions from the glow to the streamer regime for the discharge with TiO2 globules on the mesh occur, are also shown. The symbols used are the same as the symbols used in Fig. 4, that is a, b, c, d and e. From this figure the following conclusions can be drawn.

Fig. 10
figure 10

V-A characteristics of the discharge together with transition points from the glow to the streamer discharge regime (empty symbols, dashed lines, A, B, C, D, E) and the discharge with TiO2 globules on the mesh (semi-full big symbols, full thick lines, a, b, c, d, e) for supply of air into the discharge through the needle electrode. Simple needle—stars; the needle that protrudes from the dielectric tube (h = −3 mm)—circles; the needle that matches with the dielectric tube (h = 0 mm)—up triangles; the needle inserted in the dielectric tube (h = +1; +2 mm)—down triangles and squares respectively

Full size image

  • For the simple needle, as well as for all positions of the needle electrode with respect to the end of the dielectric tube, the discharge with TiO2 globules on the mesh requires, for any current, a lower voltage than the discharge without TiO2 globules on the mesh, both for supply of air through the needle or reverse supply of air through the mesh (this last case is not shown in Fig. 10).

  • For the discharge with TiO 2 globules on the mesh, insertion of the needle electrode into the dielectric tube decreases the current at which transition from the glow to the streamer regime occurs. The deeper the needle electrode is inserted the lower the current at which transition occurs. A similar finding was observed in the case of reverse air supply (not shown in Fig. 10).

  • For supply of air through the needle electrode the discharge with TiO2 globules on the mesh transition from the glow to the streamer regime occurs for simple needle as well as for any position of the needle electrode with respect to the end of the dielectric tube at lower voltage in comparison with the discharge without TiO2.

  • Discharge onset voltage for the discharge with TiO2 globules on the mesh is lower than the discharge onset voltage for supply of air through the needle without TiO2 globules.

  • For a deeper insight into the effect of reverse air supply and TiO2 photocatalyst in comparison with the discharge with supply of air through the needle electrode without TiO2 on electrical characteristics of the discharge we summarized these results for two extreme situations. The first relates to the simple needle electrode which is not inserted in the dielectric tube (Fig. 11) and the second to the needle electrode inserted in the dielectric tube with h = +1 mm (Fig. 12).

    Fig. 11
    figure 11

    V–A characteristics and transition points for the discharge with the simple needle electrode. Air supply through the needle—empty stars, A. Reverse air supply through the mesh—full stars, α Air supply through the needle with TiO2 on the mesh—semi-full stars, a

    Full size image
    Fig. 12
    figure 12

    V–A characteristics and transition points for the discharge with needle electrode inserted in the dielectric tube with h = +1 mm. Air supply through the needle—empty down triangles, D. Reverse air supply through the mesh—full down triangles, δ Air supply through the needle with TiO2 on the mesh—semi-full down triangles, d

    Full size image

From Fig. 11 it can be seen that for the discharge with the simple needle electrode the V-A characteristics are the same for the supply of air through the needle and for reverse supply of air through the mesh electrode. For supply of air through the needle, placement of TiO2 globules on the mesh decreases discharge voltage. For the needle electrode inserted into the dielectric tube with h = +1 mm (Fig. 12) for a particular current the discharge voltage is the lowest for the discharge with TiO2 globules on the mesh, the discharge with the supply of air through the needle electrode needs higher voltage and finally discharge with the reverse air supply requires the highest voltage.

The effect of different air supply to the discharge on its electrical characteristics and ozone generation can be understood at least qualitatively on the basis of the following considerations.

First of all, in the case of the simple hollow needle electrode and the hollow needle electrode, which protrudes from the dielectric tube we have a negative corona discharge, with supply of air either through the cathode (needle) or through the anode (mesh).

Generally speaking, the interelectrode space of the negative corona can be divided into three regions. In the region around the needle electrode, where there is a strong electric field, high-energy electrons inelastically collide with neutral atoms and cause avalanches. The positive ions created are attracted back to the tip of the needle and are the source of secondary electrons caused by the secondary emission. This region is called the ionization region or active region. The conduction electrons then move away from the needle in the direction of the decreasing electric field and attach themselves to oxygen molecules, creating negative ions. Thus the corona plasma region, defined as the region in which corona-enhanced chemical reactions are possible, extends beyond the ionization region [23]. Finally, in the outer, low electric field drift (unipolar) region, negative ions and electrons drift toward the anode and react with neutrals, but with too low energy to ionize and too low density to react with other ionized particles. The space charge field of drifting ions/electrons is the dominating factor determining both the corona current/voltage characteristic and the current density distribution in the discharge gap.

When air is supplied to the discharge, momentum transfer owed to the collisions of neutral molecules and atoms with negative/positive ions perturbs space charge distribution between electrodes, which is reflected in the change of the current–voltage characteristics.

When we supply air to the discharge through the hollow needle electrode (cathode) we supply air to the ionization region. Continuity of electric current from the cathode into the plasma is provided by secondary emission from the cathode mostly induced by ion impact. The secondary emission coefficient, apart from the cathode material, type of gas etc., depends on the energy of ions. Expansion of air from the needle with relatively high velocity acts against the direction of positive ion motion and therefore it could affect secondary emission of electrons. The electrons possess high mobility and flow to the anode with much higher velocities, and therefore they are not easy for the flowing air to deflect.

On the other hand, when we supply air to the discharge through the mesh electrode (anode) the air is supplied to the drift (unipolar) region. For negative corona discharge in air the current is sustained by negative ions, mainly oxygen molecules and atoms. As long as the airflow is oriented opposite to the ion drift direction, it affects the motion of negative ions consequently for particular voltage reverse supply of air through the mesh would decrease discharge current. This effect is nicely seen in Fig. 9 mainly for h = 0 and +1 mm (for h = + 1 mm also in Fig. 12). On the other hand, it could also be said that the discharge with reverse air supply requires higher voltage for particular currents. Thus the charge losses must be compensated by enhanced ionization, which requires higher discharge voltage.

It should also be mentioned that airflow patterns in the discharge chamber would be quite different in the case of air supply through the hollow needle or through the mesh electrode, which could affect turbulent transport of charges to the walls, convective charge transport and also accumulation of molecules active with respect to the detachment. For the case of air supply to the discharge through the needle electrode, an example of the velocity field distribution can be seen in [24]. For the case of air supply through the mesh electrode, an example of the airflow patterns is shown in [18]. In this paper it is also noted that current density on the mesh surface changes when airflow through the mesh changes its direction.

When the needle electrode is inserted into the dielectric tube the discharge cannot be considered as a classical simple corona discharge because, apart from the processes which pertain to negative corona, other effects appear that influence discharge mechanisms. First of all, insertion of the needle electrode into the dielectric tube changes distribution of the electric field near the needle electrode [25]. The insertion of the needle electrode into the dielectric tube also changes airflow patterns at the output of the needle in comparison with the simple needle electrode and may trigger additional disturbances or vortexes in the flow [26].

Enhancement of the discharge ozone production caused by placement of TiO2 globules on the mesh electrode could be explained by taking into account first the effect of this photocatalyst on electrical parameters of the discharge and second the effect of this photocatalyst on plasmachemical processes leading to ozone generation [12].

Placement of TiO2 on the mesh electrode may cause redistribution of the electric field in the vicinity of the mesh electrode. This is a physical effect, which is not directly related to the specific catalytic activity of the catalyst material but to the type of material. The TiO2 is an n-type semiconductor. The electric field redistribution then affects the electron energy distribution function, electron impact dissociation, and ionization rates, and consequently effects a change in the electrical parameters of the discharge. This effect is nicely demonstrated in Fig. 10, where it can be seen that for the simple needle as well as for all positions of the needle electrode with respect to the dielectric tube the discharge with TiO2 globules on the mesh requires lower voltage for a particular current than the discharge with supply of air through the needle without TiO2 on the mesh.

A second effect, which is important for enhancement of discharge ozone production is activation of the TiO2 photocatalyst by radiation emitted from the discharge and the contribution of this photocatalyst to plasmachemical processes leading to ozone generation. In the case of TiO2 the series of its energy levels is associated with covalent bonds among atoms composing the valence band and a second series of spatially diffuse higher energy levels is associated with the conduction band. Both series are separated by a forbidden energy gap of 3.2 eV. The electron of the valence band becomes excited when illuminated by UV radiation. The wavelength of the radiation necessary for this photo-excitation is 388 nm.

The corona discharge in air is an emitter of UV radiation [27, 28]. The strongest ultraviolet emission originates from the second positive system of nitrogen (C3Πu → B3Πg), which emits radiation of wavelength 337.1 nm [28]. The first negative system (B2Σ +u  → X2 Σ +g ) of N2 + emits radiation of wavelength 391.4 nm. It is therefore clear that UV radiation of the corona discharge in air could be used for TiO2 photoactivation in spite of the fact that in the corona glow regime this illumination has relatively low intensity. However, according to [29], even under extremely low-intensity UV illumination from as little as 10 nW/cm2 activation of TiO2 photocatalyst occurs.

When TiO2 absorbs ultraviolet radiation the electron of the valence band becomes excited. The excess energy of this excited electron promotes the electron to the conduction band, thereby creating the negative-electron and positive-hole pair. The electron and positive hole can react with the molecules in the vicinity of the catalyst. In air the electron reacts with an oxygen molecule to form the superoxide anion.

O2 −. These superoxide anions should be added to other plasma components produced by the discharge. All these species at different levels contribute to the processes leading to discharge ozone production, so that the resultant ozone concentration is increased.

Conclusions

For DC negative corona discharge with a hollow needle electrode inserted in the dielectric tube against the mesh electrode we investigated the role of air supply to the discharge through the needle or through the mesh electrode and the effect of the TiO2 photocatalyst on the mesh electrode in discharge ozone production. It was found that reverse supply of air through the mesh electrode decreases the pressure required to transport a given amount of air through the discharge in comparison with supply of air through the needle electrode. However, for the discharge with the same power, ozone concentration, produced by the discharge as well as ozone production yield are lower.

We also showed that for the discharge with the needle electrode that matches with the end of a dielectric tube and for the needle inserted in the dielectric tube reverse air supply requires higher discharge voltage for a particular current. Insertion of the needle electrode into the dielectric tube decreases the current at which transition from the glow to the streamer regime occurs. The deeper the needle electrode is inserted the lower the current. On the other hand, voltages at which transition from the glow to the streamer regime occurs are not substantially affected by insertion of the needle electrode into the dielectric tube. Placement of TiO2 globules on the mesh electrode increases ozone concentration produced by the discharge with supply of air through the needle electrode.

The best performance of the discharge from the standpoint of ozone generation was that of the discharge with the supply of air through the needle electrode with TiO2 globules on the mesh when the needle electrode is inserted in the dielectric tube.

The obtained results could be useful for construction of cheap and simple ozone generators as well as for increasing the efficiency of plasmachemical generators for decomposition of volatile organic compounds.