Plasma-enhanced N₂ fixation in a dielectric barrier discharge reactor: effect of packing materials

Figure 1 shows the schematic of the experimental setup. The NO_x synthesis was conducted in a coaxial DBD reactor. A quartz tube with an external diameter of 13 mm and an inner diameter of 10 mm served as the dielectric layer. A 50 mm long stainless-steel mesh was wrapped around the quartz tube and used as a ground electrode, and a stainless-steel rod with a diameter of 4 mm served as a concentric inner high voltage electrode. The discharge gap was 3 mm, and the discharge volume was 3.3 ml. Different packing materials (Al₂O₃ and BaTiO₃) of 1 mm in diameter were fully packed into the discharge area. The dielectric constants of Al₂O₃ and BaTiO₃ are ∼9 and >1000, respectively. The DBD reactor was connected to a high voltage AC power supply with a frequency of 9.3 kHz and a maximum peak voltage of 30 kV. The applied voltage was measured by a high voltage probe (Testec, TT-HVP 15 HF), while the current was recorded by a current monitor (Magnelab CT-E0.5). The voltage on an external capacitor C_ext (0.47 μF) was measured to determine the total charge accumulated in the plasma. These electrical signals were sampled by a four-channel digital oscilloscope (Tektronix, MDO3024). The discharge power was calculated using the area of the Q–U Lissajous figure. The gas temperature in the DBD reactor was measured by a fibre optic thermometer (Omega, FOB102) with the fibre inserted 10 mm into the plasma region through the outlet of the reactor.

Zoom In Zoom Out Reset image size

The gas products were analysed using an online Fourier-transform infrared (FTIR) spectrometer (FTIR-4200, Jasco) with a resolution of 0.5 cm⁻¹. Air (Zero grade, BOC) was used as the reactant and introduced into the DBD reactor by a mass flow controller (Omega, FMA-2404). The total air flow rate was fixed at 100 ml min⁻¹ for all measurements. The effluent gases passed a standard gas cell with a 10 cm path length placed in the FTIR. The products from the air DBD were determined as NO, NO₂, and N₂O in all working conditions and the sum of their concentrations was defined as NO_x concentration. Calibration gases with known concentrations of NO, NO₂, and N₂O diluted in N₂ were used to quantify the concentration of these products. Each experiment was repeated 3 times, and the margin of error in this work was within 3%.

The selectivity of the NO_x product i was calculated using the following equation:

$$\begin{equation}S\mathrm{i}\enspace (\mathrm{\%})=\frac{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}\enspace \mathrm{o}\mathrm{f}\enspace \mathrm{N}\enspace \mathrm{i}\mathrm{n}\enspace \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\enspace i}{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\enspace \mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}\enspace \mathrm{o}\mathrm{f}\enspace \mathrm{N}\enspace \mathrm{i}\mathrm{n}\enspace \mathrm{N}{\mathrm{O}}_{\mathrm{x}}\enspace \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\mathrm{s}}{\times}100.\end{equation} \tag{ 1 }$$

The energy consumption (EC) of the NO_x synthesis was determined as:

$$\begin{align}& EC\enspace (\mathrm{M}\mathrm{J}\mathrm{m}\mathrm{o}{\mathrm{l}}^{-1})\\ & \qquad =\frac{\mathrm{d}\mathrm{i}\mathrm{s}\mathrm{c}\mathrm{h}\mathrm{a}\mathrm{r}\mathrm{g}\mathrm{e}\enspace \mathrm{p}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}(\mathrm{W})}{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}\enspace \mathrm{o}\mathrm{f}\enspace \mathrm{N}\enspace \mathrm{i}\mathrm{n}\enspace \mathrm{N}{\mathrm{O}}_{x}\enspace \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\mathrm{s}\enspace \mathrm{p}\mathrm{e}\mathrm{r}\enspace \mathrm{s}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{d}\enspace (\mathrm{m}\mathrm{o}\mathrm{l}\enspace {\mathrm{s}}^{-1})}\\ & \qquad \quad {\times}\frac{1}{{10}^{6}}.\end{align} \tag{ 2 }$$

Figure 2(a) illustrates the classic equivalent circuit of a DBD reactor, in which C_diel and C_gap correspond to the capacitance of both the dielectric layer and the capacitance of the discharge gap, respectively. In the presence of packing materials, C_gap includes the contribution of the gas–solid integration in between the electrodes. When the breakdown conditions are reached, and conductive filaments take place, the switch 'K' in the parallel resistive channel will be on.

Zoom In Zoom Out Reset image size

The shape of a Q–U Lissajous figure typically approximates a parallelogram where the internal electrical discharge characteristics can be derived from its dimensions, including the breakdown voltage, reactor charge, and reactor capacitances [23, 24]. Figure 2(b) shows a typical Q–U Lissajous plot where the peak-to-peak charge (Q_p–p) accumulated in the DBD can be determined. The lines AB and CD in a Lissajous figure represent the discharge-off phase, during which only displacement current occurs and therefore no charge transfers between the electrodes. The slope dQ/dU of lines AB and CD is equal to C_cell in this period, which is composed of the gap capacitance C_gap and the dielectric capacitance C_diel, as demonstrated in the following equation:

$$\begin{equation}\frac{1}{{C}_{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}}}=\frac{1}{{C}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{l}}}+\frac{1}{{C}_{\mathrm{g}\mathrm{a}\mathrm{p}}}.\end{equation} \tag{ 3 }$$

The lines BC and DA correspond to the discharge phase, where the plasma is ignited, and the air gap is shortcut. The slope dQ/dU of BC and DA represents the effective capacitance C_eff which equals the dielectric capacitance C_diel in a fully bridged gap [25]. The gap capacitance C_gap is then calculated as:

$$\begin{equation}{C}_{\mathrm{g}\mathrm{a}\mathrm{p}}=\frac{{C}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{l}}{\times}{C}_{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}}}{{C}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{l}}-{C}_{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}}}.\end{equation} \tag{ 4 }$$

The total charge in the plasma Q can be acquired from the voltage U_c on an external capacitor C_ext of 0.47 μF. The voltage on the dielectric layers U_d can be calculated using charge Q and C_diel according to equation (6) [26].

$$\begin{equation}Q={C}_{\mathrm{e}\mathrm{x}\mathrm{t}}{\times}{U}_{\mathrm{c}}\end{equation} \tag{ 5 }$$

$$\begin{equation}{U}_{\mathrm{d}}=\frac{Q}{{C}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{l}}}.\end{equation} \tag{ 6 }$$

Then the gas voltage (also called gap voltage) U_g of the discharge region will be the difference between applied voltage U_a and dielectric voltage U_d:

$$\begin{equation}{U}_{\mathrm{g}}={U}_{\mathrm{a}}-{U}_{\mathrm{d}}.\end{equation} \tag{ 7 }$$

The breakdown voltage U_b can be determined using equation (8) where U_min stands for the minimum external voltage derived from the Lissajous figure (Figure 2(b)) [24, 27].

$$\begin{equation}{U}_{\mathrm{b}}=\frac{\mathrm{1}}{\mathrm{1}+{C}_{\mathrm{g}\mathrm{a}\mathrm{p}}/{C}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{l}}}{\times}{U}_{\mathrm{min}}.\end{equation} \tag{ 8 }$$

The average electric field (E) can be calculated by the ratio of breakdown voltage to the discharge gap distance d_gap:

$$\begin{equation}E\enspace ({\mathrm{k}\mathrm{V}\mathrm{c}\mathrm{m}}^{-1})=\frac{{U}_{\mathrm{b}}\enspace (\mathrm{k}\mathrm{V})}{{d}_{\mathrm{g}\mathrm{a}\mathrm{p}}\enspace (\mathrm{c}\mathrm{m})}.\end{equation} \tag{ 9 }$$

In the presence of packing materials, the discharge gap is significantly changed. To date, there is no standard approach for the approximation of the discharge gap in a packed-bed DBD reactor. The current established simplified geometrical models based on spherical beads enable the estimation of the average electric field in a packed DBD reactor [22, 28] but do not consider the pores, small voids, and surface curvature of the pellets, and thus readily underestimate the volume of the discharge gap. Here, a method based on water absorption [29] is used to estimate the discharge volume of the packed-bed DBD reactor. In this case, the average electric field is determined as follows:

$$\begin{equation}{E}_{\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}-\mathrm{g}\mathrm{a}\mathrm{p}}\enspace (\mathrm{k}\mathrm{V}\mathrm{c}{\mathrm{m}}^{-1})=\frac{{U}_{\mathrm{b}}\enspace (\mathrm{k}\mathrm{V})}{{d}_{\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}-\mathrm{g}\mathrm{a}\mathrm{p}}\enspace (\mathrm{c}\mathrm{m})},\end{equation} \tag{ 10 }$$

where the d_packing-gap denotes the equivalent discharge gap in the packed-bed plasma zone and is calculated by equation (11):

$$\begin{equation}{d}_{\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}-\mathrm{g}\mathrm{a}\mathrm{p}}\enspace (\mathrm{c}\mathrm{m})={d}_{\mathrm{g}\mathrm{a}\mathrm{p}}\enspace (\mathrm{c}\mathrm{m}){\times}\beta ,\end{equation} \tag{ 11 }$$

where the coefficient β stands for the volume fraction of the gas in the fully packed plasma region and can be calculated with the relation:

$$\begin{equation}\beta =\frac{{V}_{\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{g}-\mathrm{g}\mathrm{a}\mathrm{p}}\enspace (\mathrm{m}\mathrm{L})}{{V}_{\mathrm{d}\mathrm{i}\mathrm{s}\mathrm{c}\mathrm{h}\mathrm{a}\mathrm{r}\mathrm{g}\mathrm{e}}\enspace (\mathrm{m}\mathrm{L})}=\frac{{V}_{{\mathrm{H}}_{2}\mathrm{O}}(\mathrm{m}\mathrm{L})}{{V}_{\mathrm{d}\mathrm{i}\mathrm{s}\mathrm{c}\mathrm{h}\mathrm{a}\mathrm{r}\mathrm{g}\mathrm{e}}\enspace (\mathrm{m}\mathrm{L})},\end{equation} \tag{ 12 }$$

where V_packing-gap represents the gas volume in the discharge area of the packed reactor and V_discharge denotes the total discharge volume (3.3 ml in this study). For the measurement of V_packing-gap, the same amount of packing material as used in the experiments was placed inside a measuring cylinder. Deionised water was gradually added into the cylinder using a calibrated volume pipette until it perfectly submerged the pellets and the water level reached 3.3 ml. The volume of deionised water ${V}_{{\mathrm{H}}_{2}\mathrm{O}}$ was recorded from readings of the pipette. In the present study, the coefficient β for Al₂O₃ and BaTiO₃ was determined as 0.78 and 0.43, respectively.

The electrical signals of the DBD plasma operating in dry air with and without packing materials are illustrated in Figure 3. In the absence of pellets, a number of strong current pulses can be observed, with the amplitudes of current peaks reaching more than 400 mA. This suggests the presence of a typical filamentary discharge mode where plasma streamers propagate across the full gas gap in the discharge area. The duration of these filaments is in tens of nanoseconds [30]. Packing pellets into the plasma region significantly influences the discharge behaviour of the DBD reactor. More specifically, a transition from full gap discharges to more localised discharges usually takes place when introducing the packing materials in the discharge region [31]. Two typical discharge modes can be observed in a packed-bed DBD: (a) local filamentary discharges that form in the small voids in between the pellets; and (b) surface discharges that spread over the surface of pellets covering the full gas gap due to the packed-bed effect [32, 33].

Zoom In Zoom Out Reset image size

Notably, changing packing materials forms disparate discharge patterns, as shown in Figure 3. Compared with plasma only, the discharge with Al₂O₃ packing showed similar voltage signals but lower amplitudes of current peaks, indicating the formation of weaker microdischarges when using the Al₂O₃ packing. Due to the typical packed-bed effect, the discharge behaviour transforms from the filamentary discharge of plasma only to a coupled surface discharge alongside localised microdischarges when using the Al₂O₃ packing. In contrast to plasma only and Al₂O₃ packing, the current profiles of the DBD packed with BaTiO₃ demonstrated more numerous pulses with lower amplitudes of less than 40 mA, corresponding to the formation of localised filamentary discharges between bead–bead and bead–quartz wall. Evidenced by intensified charge-coupled device imaging of the BaTiO₃-packed discharge, local microdischarges are observed in between the pellets, while surface discharges are hardly visible [34]. This phenomenon is associated with the lower β coefficient of BaTiO₃ (β = 0.43) when compared with Al₂O₃ (β = 0.78) and plasma only (β = 1); this low β indicates a limited plasma volume in the reactor. The DBD with BaTiO₃ packing showed a lower average gas voltage and thus should be expected to deposit fewer charges per filament, this can be evidenced by the lower amplitudes of the current peaks in Figure 3(c). Simultaneously, packing with BaTiO₃ produced more filaments and total charge deposition compared to Al₂O₃ and plasma only (see Figure 6). The BaTiO₃ with a high dielectric constant significantly decreased the applied voltage within the gas gaps and resulted in a lower gas voltage, this may also benefit the charge trapping [34]. Similar observations of disparate current profiles when using different packing materials were also reported in previous studies [12, 35, 36].

Figure 3 shows the time-resolved gas voltage (red colour) of the air discharge. A flat curve of gas voltage takes place simultaneously with the ignition of plasma, this value was above 2 kV for both plasma only and Al₂O₃ packing and was ∼1.5 kV for BaTiO₃ packing. This phenomenon indicates that the gas voltage remains relatively constant over time after the breakdown of the gas, implying that the formation of individual filaments (with a lifetime of nanoseconds) is not affected by the applied voltage after reaching gas breakdown. This constant gas voltage in multi-filament DBD can be ascribed to the development of deposited charge on the dielectric surface [37]. To better validate this effect, the breakdown voltage (U_b) of the DBD was derived from the equivalent circuit and Lissajous figures (Figure 4). The breakdown voltage remained constant when changing the discharge power and was ∼2.6 kV, ∼2.4 kV, and ∼1.5 kV for plasma only, Al₂O₃ packing, and BaTiO₃, respectively, consistent with the observation from the gas voltage signals. Previous studies [31, 35] also reported that packing BaTiO₃ into a DBD reactor lowered the breakdown voltage compared to the DBD without packing or with Al₂O₃ packing.

Zoom In Zoom Out Reset image size

Figure 5 shows typical Lissajous figures of the air DBD with and without packing at a constant discharge power of 15 W. The Lissajous figure of the discharge without packing and with Al₂O₃ packing exhibits a similar parallelogram shape, while the Lissajous figure of the DBD packed with BaTiO₃ shows an oval shape, which is likely the consequence of enhanced charge deposition when packing BaTiO₃. The changes in the Lissajous shape also demonstrate the transition of the discharge mode. Our previous works proved that packing pellets into a DBD changed the discharge mode due to the reduced discharge volume induced by the packing [22, 38]. At a constant discharge power, the discharge packed with BaTiO₃ showed a much higher peak applied voltage compared to the DBD without packing (Figure 5), which can be associated with the significantly reduced amplitude of current pulses due to the formation of confined discharge when using the BaTiO₃ packing. As derived from the Lissajous plots, the equivalent capacitance of the discharge gap C_gap increased from 4.6 pF without packing to 9.5 pF and 26.0 pF when packed with Al₂O₃ and BaTiO₃, respectively. This significantly higher gap capacitance in the presence of BaTiO₃ packing can be attributed to the great capacity of BaTiO₃ to trap charges due to the low β coefficient.

Zoom In Zoom Out Reset image size

As shown in Figure 6, the peak-to-peak charges accumulated in the DBD increases with the discharge power. Clearly, introducing packing materials into the discharge area greatly affects the charge characteristics of the DBD. The generated charges were significantly enhanced when fully packed with pellets, especially with BaTiO₃ beads. In general, the peak-to-peak charges were almost doubled with BaTiO₃ packing in comparison to no packing. Moreover, the charge depositions of BaTiO₃ packing were also found superior to Al₂O₃ packing. As shown in Figures 3 and 4, packing with BaTiO₃ results in a lower breakdown voltage and generates more filaments than Al₂O₃ and plasma only due to the lower β coefficient and higher dielectric constant, which contributes to a greater total charge deposition.

Zoom In Zoom Out Reset image size

Table 1 illustrates the influence of packing materials on the time-averaged electric field in the DBD at different discharge powers. As discussed above, increasing the applied voltage (and thus the discharge power) at a fixed frequency has no influence on the breakdown voltage, and therefore results in a constant average electric field. Compared with plasma only, introducing packing materials in the plasma zone effectively enhances the electric field. For instance, in this work, the highest average electric field of around 11.5 kV cm⁻¹ was observed in the BaTiO₃-packed reactor, while the average electric field of the DBD without packing was only 8.7 kV cm⁻¹. Gallon et al found that materials with a high dielectric constant can greatly enhance the local electric field near the contact points between packing beads [39]. A modelling study also reported that an enhanced average electric field can be achieved when using packing materials with a higher dielectric constant [35]. The influence of packing materials on the calculated reduced electric field is consistent with the time-averaged electric field, as shown in Figure 7. The enhanced reduced electric field at a higher discharge power could be attributed to the decreased neutral particle concentration N due to the increased plasma temperature. The gas temperatures of the DBD with and without packing at different discharge powers are listed in Table 2.

Zoom In Zoom Out Reset image size

To get insights into the influence of packing materials on the changes of the electric field and electron energy of the DBD, the rate coefficients of electron impact reactions and electron energy distribution function were calculated using the Boltzmann solver BOLSIG+ (version 12/2019) [40]. The cross-sections of air components were retrieved from the LXCat project database [41]. The electric field has a significant impact on the plasma properties including the mean electron energy and electron density, and thus the prevalence and distribution of reactive species generated in the plasma [42]. Figure 8 shows the calculated mean electron energy as a function of the reduced electric field. Clearly, the mean electron energy increases with the increase of the reduced electric field in the air DBD. The coloured rectangle illustrates the range of the reduced electric field in this study, between 49.1 Td for plasma only and 74.2 Td for BaTiO₃, while the DBD packed with Al₂O₃ has an E/N of around 60 Td.

Zoom In Zoom Out Reset image size

Figure 9 shows the influence of packing materials on the calculated mean electron energy at different discharge powers. Referring to the trend in Figure 8, increasing the discharge power enhanced the reduced electric field and mean electron energy. The mean electron energy of the DBD packed with Al₂O₃ or BaTiO₃ was higher compared to the discharge using plasma only. The DBD with BaTiO₃ packing showed the highest mean electron energy of 1.81 eV at a discharge power of 16 W. More importantly, our calculation implies the presence of packing materials significantly changes the physical properties of the discharge, generating more energetic electrons benefitting the activation of dinitrogen, which is consistent with the improved performance of plasma N₂ fixation (see section 3.3).

Zoom In Zoom Out Reset image size

In this study, NO and NO₂ were identified as the major products in the plasma synthesis of NO_x , while N₂O was also detected. Figure 10 shows the influence of packing materials on the NO_x production at different discharge powers. Clearly, increasing discharge power enhanced the NO_x concentration. It is noteworthy that increasing discharge power from 9 to 16 W almost doubled the NO_x concentration from 5811 to 11 830 ppm using BaTiO₃ packing. Increasing discharge power can generate a greater number of filaments, creating more channels for plasma chemical reactions, thus enhancing the production of NO_x [43]. Patil et al also reported NO_x concentration increased significantly when increasing the discharge power in a DBD reactor [12].

Zoom In Zoom Out Reset image size

Compared with the reaction using plasma only or Al₂O₃, packing BaTiO₃ in the DBD significantly enhanced the formation of NO_x , which can be attributed to the enhanced reduced electric field and mean electron energy induced by the packing, generating more energetic electrons and chemically reactive species (e.g. nitrogen and oxygen excited states) for the activation of N₂ molecules. The positive effect of BaTiO₃ packing on the plasma chemical reactions was also reported in previous studies. For instance, Michielsen et al evaluated the effect of different packing materials (e.g. glass wool, quartz wool, SiO₂, ZrO₂, Al₂O₃ and BaTiO₃) on CO₂ dissociation in a packed-bed DBD reactor and found that the BaTiO₃ packing exhibited the highest CO₂ conversion and energy efficiency [44]. Mei et al also reported that packing BaTiO₃ into a DBD reactor substantially enhanced the conversion of CO₂ and energy efficiency compared to the plasma reaction without packing [45].

Figure 11 illustrates the selectivity of NO, NO₂, and N₂O as a function of discharge power with and without packing materials. Increasing the discharge power enhanced the NO selectivity but decreased the selectivity of NO₂ and N₂O regardless of whether the packing is used or not. NO and NO₂ were the dominant products for the reaction using plasma only or Al₂O₃ packing, while packing BaTiO₃ into the DBD substantially increased the NO selectivity, reaching a maximum of 67.9% at a discharge power of 16 W. This finding indicates that the stronger electric field induced by BaTiO₃ packing is favourable for the formation of NO, which can be explained by the NO formation pathways below.

Zoom In Zoom Out Reset image size

In the NO_x synthesis from N₂ and O₂, the primary step is the formation of NO, as shown in (R1).

$$\begin{equation}{\mathrm{N}}_{2}+{\mathrm{O}}_{2}\enspace {\rightleftharpoons}\enspace 2\mathrm{N}\mathrm{O}.\end{equation} \tag{ R1 }$$

However, this direct bimolecular reaction is symmetry-forbidden. It is well recognised that the NO molecules are generated through a combination of the elementary reactions (R2) and (R3) following the Zeldovich mechanism. The first reaction (R2) is typically regarded as the rate-determining step [46].

$$\begin{equation}{\mathrm{O}+\mathrm{N}}_{2}\to \mathrm{N}\mathrm{O}+\mathrm{N}\end{equation} \tag{ R2 }$$

$$\begin{equation}{\mathrm{N}+\mathrm{O}}_{2}\to \mathrm{N}\mathrm{O}+\mathrm{O}.\end{equation} \tag{ R3 }$$

In NTPs, collisions between the plasma-induced energetic electrons and neutral particles can generate numerous highly reactive species that can participate in the synthesis of NO_x . As a result, NO_x synthesis reactions that only occur at combustion temperatures can be achieved at low temperatures by overcoming the high energy barrier of NO formation. To be more specific, the rate-limiting reaction (R2) can be significantly accelerated by the plasma-generated nitrogen excited species, as shown in (R4).

$$\begin{equation}{\mathrm{O}+\mathrm{N}}_{2}^{{\ast}}\to \mathrm{N}\mathrm{O}+\mathrm{N}.\end{equation} \tag{ R4 }$$

Besides, the enhancement of the electric field and mean electron energy could effectively facilitate the vibrational and electronical excitation of nitrogen molecules, as proved by the enhanced rate coefficients calculated by BOLSIG+ (Figure 12), which explains the improved NO selectivity with the increase of discharge power in the presence of the BaTiO₃ packing.

Zoom In Zoom Out Reset image size

Previous studies of plasma NO_x synthesis [19, 47, 48] showed that reaction (R5) is the dominant pathway for NO₂ formation in NTPs, while the generated NO₂ molecules may be converted back to NO via reaction (R6). (M: third body)

$$\begin{equation}\mathrm{N}\mathrm{O}+\mathrm{O}(+\mathrm{M})\to {\mathrm{N}\mathrm{O}}_{2}(+\mathrm{M})\end{equation} \tag{ R5 }$$

$$\begin{equation}{\mathrm{N}\mathrm{O}}_{2}+\mathrm{O}\to {\mathrm{N}\mathrm{O}+\mathrm{O}}_{2}.\end{equation} \tag{ R6 }$$

The formation of N₂O can be mainly attributed to the reactions (R7)–(R9) [18].

$$\begin{equation}{\mathrm{N}}_{2}{(\mathrm{A})+\mathrm{O}}_{2}\to {\mathrm{N}}_{2}\mathrm{O}+\mathrm{O}\end{equation} \tag{ R7 }$$

$$\begin{equation}{\mathrm{N}\mathrm{O}}_{2}+\mathrm{N}\to {\mathrm{N}}_{2}\mathrm{O}+\mathrm{O}\end{equation} \tag{ R8 }$$

$$\begin{equation}{\mathrm{N}}_{2}+\mathrm{O}({}^{1}\mathrm{D})\to {\mathrm{N}}_{2}\mathrm{O}.\end{equation} \tag{ R9 }$$

As the initial energy of ground-state molecules can be elevated by plasma-generated chemically reactive species, the energy barriers of the aforementioned reactions can therefore be effectively reduced.

Figure 13 shows the effect of packing materials on the energy consumption of NO_x synthesis at a constant discharge power of 16 W. Note the overall energy consumption was almost constant when increasing the discharge power. The NO_x synthesis using plasma only exhibited the highest energy consumption of 51.3 MJ mol⁻¹, significantly higher than when using Al₂O₃ packing (33.4 MJ mol⁻¹) and BaTiO₃ packing (15.9 MJ mol⁻¹). This result indicates that the energy cost of the plasma NO_x synthesis using DBDs can be significantly reduced by using packing materials. More importantly, introducing proper packing materials into the plasma region can benefit the NO_x generation from both physical and chemical standpoints.

Zoom In Zoom Out Reset image size

Table 3 compares the reaction performances of the NO_x synthesis using DBD plasmas. The concentration of NO_x produced in this study is higher than that reported in previous works using the same type of NTP. Without packing materials, the energy consumption of NO_x synthesis is comparable or higher than that using similar DBD systems, while the energy consumption (15.9 MJ mol⁻¹) for NO_x synthesis using BaTiO₃ is lower compared with the previous studies using packed-bed DBDs.