Ammonia synthesis and by-product formation from H₂O, H₂ and N₂ by dielectric barrier discharge combined with an Ru/Al₂O₃ catalyst

Abstract

NH₃ synthesis from H₂O + N₂, H₂ + N₂ or H₂O + H₂ + N₂ by dielectric barrier discharge combined with an Ru/Al₂O₃ catalyst is studied at atmospheric pressure and room temperature. Additionally, the effects of reaction gas composition, energy density, and discharge frequency on NH₃ yield and by-product formation are investigated. The results show that NH₃ can be formed from reaction gases consisting of H₂O and N₂. The NH₃ yield of the reaction gas containing H₂ is much higher than that of the reaction gas consisting of H₂O and N₂. The presence of H₂O has a promotion effect on NH₃ synthesis from H₂ and N₂, especially for reaction gases with an H₂ content less than 10%. The NH₃ yield first increases and then decreases with an increase in the energy density. The maximum yield of 680 mg kW⁻¹ h⁻¹ occurs at the energy density of 1400 J L⁻¹ and in the reaction gas of 0.14% H₂O, 40% H₂ and N₂. The discharge frequency has a great effect on the NH₃ yield, and maximum NH₃ yield is obtained at the frequency of 13 kHz. Besides NH₃, some by-products, such as N₂O and NO₂, are also formed with reaction gases containing H₂O. However, their formation can be suppressed in the presence of H₂, which is attributed to the reduction effect of H₂ on the by-products.

---

1. Introduction

As the most important post treatment techniques for NO_x emission from combustion processes, selective catalytic reduction (NH₃-SCR) and selective non-catalytic reduction (NH₃-SNCR) with NH₃ as the reductant have gained wide attention both in the research field and in practical applications.^1–4 Although their excellent denitration (DeNO_x) performance has been generally recognized, the supply of NH₃ remains a problem because of its toxicity and explosivity. To ensure safety, harsh measures have to be adopted for the transport, storage and usage of liquid and aqueous NH₃. One effective approach to overcome these shortcomings is to generate NH₃ onsite, where and when the need arises. Essentially, the on-board synthesis of NH₃ for NO_x removal in truck and car exhausts has already been investigated.⁵

The conventional ammonia synthesis technology, particularly the Haber–Bosch process, is only suitable for large scale and continuous production since it requires high operating temperatures (400–500 °C) and pressures (150–300 bar) over magnetite (Fe₃O₄) catalysts.^6,7 Currently, the non-thermal plasma process, which is characterized by operating/stopping instantaneously, has emerged as an alternative method of treating air pollutants,^8,9 as well as a potential way to synthesize gas products at atmospheric pressure and close to ambient gas temperature.^10,11 Essentially, NH₃ synthesis from N₂–H₂ gas mixtures induced by non-thermal plasma at ambient temperature was reported by Brewer as early as 1929.¹² Currently, various forms of plasma, including microwave discharge plasma,¹³ radio frequency discharge plasma,¹⁴ ECR (electron cyclotron resonance) plasma,¹⁵ and expanding thermal plasma,⁷ are employed for the synthesis of NH₃ from gas mixtures of N₂ and H₂. The apparatus for generating these types of plasma, however, operate under reduced pressure (several hundred Pa atmospheric pressure); thus they are unsuitable for practical applications.

Recently, the application of atmospheric pressure DBD plasma in NH₃ synthesis was reported by Mizushima using a catalyst-loaded membrane as a dielectric material,¹⁶ Hong used a DBD packed bed reactor filled with dielectric spheres¹⁷ and Bai applied a microgap DBD reactor.¹⁸ Furthermore, Bai also pointed out that the NH₃ yield increased by approximately 1.54–1.75 times when MgO powder was coated on the surface of the electrode plate.¹⁹ Improvement in NH₃ yield was also achieved by loading the alumina tube with Ru, Pt, Ni, and Fe catalysts placed between the electrodes.²⁰ Previous researchers generally thought that the reaction occurs on the surface of catalysts.^21,22 Excited nitrogen molecules are absorbed on the surface of the catalyst and then dissociated to adsorbed nitrogen atoms, N (a). The nitrogen atoms, N (a), combine with hydrogen from the gas phase or on the surface of the catalyst to successively form NH (a) and NH₂ (a) and finally become ammonia.²²

In this study, we focus on the synthesis of NH₃ from H₂O, H₂ and N₂ by DBD-type plasma combined with an Ru/Al₂O₃ catalyst at atmospheric pressure and room temperature. Ru was selected as the catalyst and dielectric barrier material because it is seen as a second generation non-iron catalyst for NH₃ synthesis from the reaction gases of H₂ and N₂.^6,23 Moreover, H₂O also acts as the source of hydrogen because hydrogen atoms can be produced by dissociating water molecules with the help of non-thermal plasma, in which abundant reactive species such as high-energy electrons, ions, excited atoms and molecules, and active radicals are contained. For this reason, a gas stream containing H₂O, H₂ and N₂ as the reaction gases was used. Furthermore, the effects of reaction gas composition, energy density, and discharge frequency on NH₃ yield are investigated.

2. Experimental

2.1 Experimental setup

A schematic of the experimental setup is shown in Fig. 1. It consists of a gas feeding system, a cylinder DBD reactor with an AC high-voltage power supply (Dalian University of Technology, China), a digital oscilloscope (Tektronix DPO3034, America) with a discharge parameter detection unit and a set of analytical instruments. H₂ + N₂, H₂O + N₂ or H₂O + H₂ + N₂ were used as the reaction gases and the gas flow rates were controlled with mass flow controllers (MFC). H₂O was introduced by bubbling N₂ in deionized water at 25 °C and the relative humidity of the reaction gases was detected by a humidity meter (Cole Parmer, America). Then, the H₂O content was calculated based on the relative humidity. The total gas flow rate was 500 mL min⁻¹ except otherwise stated.

Fig. 1 Schematic of the experimental setup.

2.2 Plasma reactor

Fig. 2 shows a schematic of the concentric DBD reactor used in this study. An aluminium rod, which had the discharge length and outer diameter of 100 mm and 21 mm, respectively, served as the high-voltage electrode and was placed in a quartz tube with the length and inner diameter of 200 mm and 25 mm, respectively. Aluminum tape, wrapped on the outer surface of the quartz tube, was connected to a 10 nF capacitor before grounding. The Ru/Al₂O₃ catalyst was packed in the space between the high-voltage electrode and the quartz tube. By applying a high AC voltage, intense discharge plasma could be generated around the contact point of the catalyst without arcing.

Fig. 2 Schematic of the DBD plasma reactor.

2.3 Preparation of Ru/Al₂O₃ catalyst

Aluminum oxide powder was provided by Aladdin Co. Ltd (Shanghai, China) and RuCl₃ was obtained from Shenyang Jinke Reagent Co. Ltd (China). All chemicals were used as received without any further purification. Ru (1 wt%) was added by incipient wetness impregnation with an aqueous RuCl₃ solution, followed by evaporation in a Rotavapor and drying (105 °C) in an oven overnight. Then, the as-prepared sample was reduced in hydrogen at 400 °C for 4 h to obtain Ru/Al₂O₃.

2.4 Characterization of the catalyst

X-ray powder diffraction (XRD) measurements were performed using a PANalytical X' Pert PRO diffractometer (Cu Pd radiation) at 40 kV and 40 mA. HRTEM micrographs were obtained with a JEM-2100F microscope at 200 kV. The surface chemical states of the Ru/Al₂O₃ catalyst were investigated via XPS (PHI Quantro SXM ULVAC-PHI, Japan) using an Al Kα X-ray source (1486.7 eV) at 15 kV and 25 W with the binding energy calibrated by C 1s at 284.8 eV.

2.5 Experimental method

NH₃ was absorbed by dilute sulphuric acid and then quantitatively analyzed using a visible spectrophotometer based on Nessler's reagent colorimetric method.²⁴ The other gas phase products were qualitatively confirmed using Fourier transform infrared spectrometry (FTIR, Nicolet 6700, America) and/or mass spectrometry (MS, OmniStar GSD300, Germany). The applied voltage and voltage drop through the capacitor were monitored with a high voltage probe (EP-50K, Japan) and a low voltage probe (Tektronix P5100, America), respectively. The Q–U Lissajous method was used to calculate the discharge power of the DBD reactor (P), which is defined as follows:

P (W) = f × C × A (1)

where f, C and A are the frequency of discharge, capacitance of the external capacitor (10 nF) and the area of the Lissajous diagram, respectively.

Energy density (ED) is defined as the energy deposited per unit volume of reaction gas:

where Q denotes the total gas flow rate in L min⁻¹.

In this study, the yield of NH₃ (η, mg kW⁻¹ h⁻¹) is defined as follows:

where C_NH3 (mg L⁻¹) is the concentration of NH₃ detected at the outlet of the reactor.

3. Results and discussion

3.1 Catalyst characterization

Fig. 3 shows the XRD spectra of the Ru/Al₂O₃ catalyst before and after the discharge reaction (5% H₂ + N₂, ED = 1400 J L⁻¹). Both patterns contain only diffraction peaks from alumina. No peak correlated to Ru is observed, which is probably due to its relatively low content.

Fig. 3 X-ray diffraction patterns of Ru/Al₂O₃ catalyst: (a) before discharge and (b) after discharge (5% H₂ + N₂, ED = 1400 J L⁻¹).

The distribution of Ru particles on the Al₂O₃ support was determined by TEM (Fig. 4), which shows that the Ru nanoparticles (∼3 nm) appear to be uniformly distributed on the Al₂O₃ support and the size and distribution of the Ru particles are almost unchanged after the discharge reaction.

Fig. 4 Transmission electron micrograph of Ru/Al₂O catalyst: (a) before discharge and (b) after discharge (5% H₂ + N₂, ED = 1400 J L⁻¹).

XPS analysis was performed on the Ru/Al₂O₃ catalyst both before and after the discharge reaction (Fig. 5). The peak at 461.9 eV can be assigned as metallic Ru.²⁵ As shown in Fig. 5, the valence state of Ru remains the same after the discharge reaction.

Fig. 5 Ru 3p_3/2 electron spectra of Ru/Al₂O₃ catalyst: (a) before discharge and (b) after discharge (5% H₂ + N₂, ED = 1400 J L⁻¹).

3.2 Discharge characteristics

Fig. 6 shows the relationship between energy density and discharge voltage for different reaction gases at the same discharge frequency of 11.0 kHz. For the representative reaction gases, the energy density remains zero until the voltage reaches the discharge inception value and then the energy density dramatically increases with an increase in discharge voltage. The introduction of H₂O results in an increase in the discharge inception value and a decrease in energy density at a given discharge voltage, which can be attributed to the strong electron affinities of the H₂O molecules. The ability of H₂O molecules to attach dissociative electrons is larger than that of H₂.²⁶ Therefore, a higher discharge inception voltage is needed to achieve the same energy density in the reactor for 0.14% H₂O + N₂, compared with the reaction gases consisting of 5% H₂ + N₂ or 0.14% H₂O + 5% H₂ + N₂.

Fig. 6 Energy density as a function of discharge voltage (f = 11.0 kHz).

3.3 Effect of H₂O content on NH₃ yield

3.3.1 Reaction gases consisting of H₂O and N₂. Fig. 7 illustrates the variation of NH₃ yield with energy density for the H₂O content of 0.07%, 0.14%, and 0.21%. For all the H₂O contents, the NH₃ yield first increases and then decreases with an increase in energy density, with maximum ammonia yields of 8.5 mg kW⁻¹ h⁻¹, 17.6 mg kW⁻¹ h⁻¹ and 16.1 mg kW⁻¹ h⁻¹ at energy densities of 500 J L⁻¹, 670 J L⁻¹ and 550 J L⁻¹ for H₂O content of 0.07%, 0.14%, and 0.21%, respectively. The NH₃ yield enhances by about 2 times when the H₂O content increases from 0.07% to 0.14%, whereas it almost remains the same for the H₂O content of 0.14% and 0.21%.

Fig. 7 Effect of H₂O content on NH₃ yield (reaction gases: H₂O + N₂, f = 11.0 kHz).

During discharge, dissociation processes can simultaneously occur due to inelastic collisions of H₂O and N₂ molecules with energetic electrons, which produce active species such as , H₂O*, N˙, H˙, ˙OH, O˙, N⁺, N₂⁺ and H⁺ via the reactions (R1)–(R5), which can further react to form several types of products including NH₃ (Fig. 7), H₂ (Fig. 8) and NO_x (Fig. 9).

Fig. 8 Hydrogen ion current for different energy densities (reaction gases: 0.14% H₂O + N₂, f = 11.0 kHz).

Fig. 9 FTIR spectra of the effluents after the DBD reaction (reaction gases: 0.14% H₂O + N₂, f = 11.0 kHz).

As can be seen from Fig. 8, the H₂ yield, which is represented by ion current detected by MS, increases with an increase in energy density, thus indicating that more H atoms can be generated at a higher energy density, which explains the phenomenon that NH₃ yield increases with an increase in energy density (Fig. 7). On the other hand, the FTIR results shown in Fig. 9 prove that NO_x (mainly NO₂ and N₂O) is also produced during the DBD process and the concentration of NO_x is greatly enhanced when the energy density is higher than a certain value. A portion of N species and discharge energy used for NH₃ synthesis can be consumed in NO_x formation (reactions (R6)–(R8)), which thus results in a decrease in NH₃ yield with an increase in energy density after the peak point (Fig. 7). Furthermore, Mizushima pointed out that the occurrence of a backward reaction is a possible reason for the decrease in NH₃ yield with an increase in energy density.¹⁶ In other words, at a higher energy density, more NH₃ molecules are formed, and their decomposition to H₂ and N₂ is also promoted. For this reason, there exists an optimum energy density for obtaining a maximum NH₃ yield in the plasma process.

N₂ + e → N₂⁺(X²∑g⁺, A²Πu, B²∑u⁺, C³∑u⁺) + 2e (R1)

N₂ + e → N⁺(³P) + N⁺(⁴S, ²D) + 2e (R2)

N₂⁺ + H₂O → N₂ + H₂O⁺ (R3)

H₂O⁺ + H₂O → H₃O⁺ + OH (R4)

e + H₂O⁺ → H + OH (R5)

N + ˙OH → NO + H˙, k₁ = 5.05 × 10⁻¹¹ cm³ per molecule per s (R6)

NO + O˙ → NO₂, k₂ = 3.01 × 10⁻¹¹ cm³ per molecule per s (R7)

NO₂ + N → N₂O + O, k₃ = 1.2 × 10⁻¹¹ cm³ per molecule per s (R8)

3.3.2 Reaction gases consisting of H₂O, H₂ and N₂. The effect of H₂O content on NH₃ yield is plotted in Fig. 10 when the basic composition of the reaction gases is 5% H₂ and N₂. With an increase in energy density from 0 to 2000 J L⁻¹, the NH₃ yield increases to a peak value and then decreases to a certain value regardless of the H₂O content in the reaction gases. For the reaction gases consisting of 5% H₂ and N₂ (dry case), the maximum NH₃ yield is 270 mg kW⁻¹ h⁻¹ at 1386 J L⁻¹, which is much higher than the case for H₂O and N₂ (Fig. 7). The introduction of H₂O to the reaction gases of 5% H₂ and N₂ results in an increase in the NH₃ yield. For the same energy density of 1400 J L⁻¹, the NH₃ yield increases from 270 mg kW⁻¹ h⁻¹ (dry case) to 300 mg kW⁻¹ h⁻¹, 380 mg kW⁻¹ h⁻¹ and 310 mg kW⁻¹ h⁻¹ in the presence of 0.07% H₂O, 0.14% H₂O, and 0.21% H₂O, respectively, which can be attributed to the formation of more H-containing species in favor of NH₃ synthesis in the presence of H₂O. Kikuchi studied the effect of H₂O on atomic hydrogen generation in hydrogen plasma and proposed that the relative concentration of hydrogen atoms increases when water vapor is added.²⁷

Fig. 10 Effect of the addition of H₂O on NH₃ yield (basic reaction gases 5% H₂ + N₂, f = 11.0 kHz).

Fig. 11 shows the FTIR spectra of the effluents after the DBD reaction for different H₂O contents under the same energy density of 1400 J L⁻¹. It can be seen that the presence of H₂O promotes NH₃ formation. Moreover, only nitrous oxide (N₂O) is detected as the byproduct and its concentration increases with an increase in H₂O content. NO₂, which is observed when H₂O and N₂ were used as the reaction gases (Fig. 9), is not detected in the presence of H₂ (Fig. 11). This may be due to the fact that more NH and NH₂ radicals exist in the discharge region as precursors of NH₃. These free radicals are strongly reductive and can reduce NO and NO₂ to N₂O rapidly, as shown in eqn (R9)–(R11).

NO + NH → N₂O + H˙, k₄ = 8.39 × 10⁻¹ cm³ per molecule per s (R9)

NO₂ + NH → N₂O + ˙OH, k₅ = 4.1 × 10⁻¹ cm³ per molecule per s (R10)

NO₂ + NH₂ → N₂O + H₂O, k₆ = 1.4 × 10⁻¹ cm³ per molecule per s (R11)

Fig. 11 FTIR spectra of the effluents after the DBD reaction (reaction gases: 5% H₂ + H₂O + N₂, f = 11.0 kHz).

Evidently, the rate constants of these reduction reactions are several orders of magnitude higher than that of nitrogen oxide (NO and NO₂) formation reactions (eqn (R6) and (R7)). As a result, the NO and NO₂ formed during the plasma process are immediately consumed by the NH and NH₂ radicals. On the other hand, an increase in the H₂O content results in an increase in N₂O concentration by providing more oxidizing species (e.g. O and OH) for NO_x formation (eqn (R6) and (R7)), which may not be favorable for NH₃ synthesis, and the NH₃ yield decreases when the H₂O content is higher than the optimum value of 0.14%.

3.4 Effect of H₂ content on NH₃ yield

3.4.1 Reaction gases consisting of H₂ and N₂. The effect of H₂ content on NH₃ yield without H₂O is shown in Fig. 12. The NH₃ yield increases to a peak value at an energy density of about 1400 J L⁻¹ regardless of the H₂ content. At this point, the NH₃ yield reaches 270 mg kW⁻¹ h⁻¹, 440 mg kW⁻¹ h⁻¹, 620 mg kW⁻¹ h⁻¹, 640 mg kW⁻¹ h⁻¹, and 580 mg kW⁻¹ h⁻¹ for H₂ contents of 5%, 10%, 20%, 40%, and 60%, respectively. In a traditional NH₃ synthesis process, the H₂ content of the reaction gases is usually around 75% (N₂ : H₂ = 1 : 3). In this study, the observed optimum H₂ content (40%) is lower than the calculated value from the stoichiometric ratio of H₂ and N₂ in NH₃. Previous studies on NH₃ synthesis in the presence of non-thermal plasma also indicate that the NH₃ yield depends on the N₂/H₂ ratio in the reaction gases and a similar optimal ratio has been reported.^19,28

Fig. 12 Effect of H₂ content on NH₃ yield (reaction gases: H₂ + N₂, f = 13 kHz).

Unlike the traditional NH₃ synthesis process, in which molecules are dissociated on a catalytic surface at high temperatures (400–550 °C) and high pressures (15–35 MPa), under the non-thermal plasma condition, the precursor molecules, H₂O, H₂ and/or N₂, are dissociated to N and H atoms. Thus, the surfaces of the catalyst are exposed to a high concentration of atomic N and H radicals. NH₃ formation turns out to be strongly dependent on the content of atomic N and H radicals.¹⁵ When the H₂ content is lower than the optimum H₂/N₂ ratio, the reaction possibility between H₂ and N₂ on the surface of the catalyst increases with an increase in H₂ content, which is due to more reaction gases dissociated in favor of NH₃ synthesis. However, if the H₂ content further increases above the optimum value, NH₃ synthesis is restrained mainly due to the following two reasons: firstly, more discharge power would be consumed for activating H₂ molecules other than N₂ molecules, since the dissociation energy of H₂ (4.5 eV) is much lower than that of N₂ (9.8 eV), which results in a shortage of nitrogen atoms for NH₃ synthesis; secondly, with an increase in H₂ content, more hydrogen atoms occupy the catalyst surface, thus reducing the number of adsorbed nitrogen atoms and also increasing the possibility of hydrogen atom recombination to hydrogen molecules.

3.4.2 Reaction gases consisting of 0.14% H₂O, H₂ and N₂. Fig. 13 shows the NH₃ yield versus H₂ content of the reaction gases in the presence of 0.14% H₂O and a given energy density of 1400 J L⁻¹. It can be seen that the NH₃ yield also first increases and then decreases with an increase in H₂ the content, with the highest NH₃ yield of 680 mg kW⁻¹ h⁻¹ at the H₂ content of 40%. It is interesting to find that the presence of H₂O has a promotion effect on NH₃ synthesis from H₂ and N₂, and this effect for reaction gases with an H₂ content less than 10% is much stronger than that for reaction gases with an H₂ content higher than 10%, which may be ascribed to the fact that H₂O molecules make more contribution to generate hydrogen atoms for NH₃ synthesis when the H₂ content is lower than 10%.

Fig. 13 Effect of H₂ content on NH₃ yield (reaction gases: H₂ + N₂ or H₂O + H₂ + N₂, f = 13 kHz).

As shown in Fig. 14, N₂O is the main by-product when the H₂ content is less than 40% and it disappears for a higher H₂ content, which is likely due to the reduction of N₂O into N₂ by hydrogen atoms through the reaction (R12) when the H₂ content is high enough.

N₂O + H → OH + N₂, k₇ = 3.0 × 10¹⁴ exp(−16 000/RT) cm³ per molecule per s (R12)

Fig. 14 FTIR spectra of the effluents after the DBD reaction (reaction gases: 0.14% H₂O + H₂ + N₂, ED = 1400 J L⁻¹ and f = 11 kHz).

3.5 Effect of discharge frequency on NH₃ yield

As shown in Fig. 15, the discharge frequency has a great effect on the NH₃ yield at a given energy density, with the maximum NH₃ yield of 680 mg kW⁻¹ h⁻¹ at 13 kHz for reaction gases consisting of 0.14% H₂O + 40% H₂ + N₂. For a discharge frequency lower than 13 kHz, an increase in discharge frequency results in an increase in microdischarge, which favors the formation of N- and H-containing species and the synthesis of NH₃. However, further increase in the discharge frequency over 13 kHz results in a decrease in NH₃ yield, possibly due to the dissociation of NH₃ by the excessive energetic electrons. Thus, the NH₃ yield decreases to 600 mg kW⁻¹ h⁻¹ at 14.5 kHz. These findings are in a fairly good agreement with Bai's experimental results.¹⁸

Fig. 15 Effect of discharge frequency on NH₃ yield (reaction gases: 0.14% H₂O + 40% H₂ + N₂, ED = 1400 J L⁻¹).

4. Reaction routes

Based on the results described above, we propose the reaction routes for the NH₃ synthesis and by-product formation via a hybrid plasma-catalytic process, as shown in Fig. 16. It is well known that the rate-limiting step of NH₃ synthesis in the conventional Haber–Bosch process is the dissociative chemisorption of N₂ molecules on the surface of the catalyst and the dissociation process only occurs at a temperature between 400–500 °C due to the high binding energy of 9.6 eV of N₂. In a DBD reactor, atomic N (a) species can be produced on the surface of the catalyst through the dissociative adsorption of molecules excited by electron impacts at atmospheric pressure and room temperature. Plasma discharge also generates activated hydrogen species such as H atoms and molecules. Then, NH₃ can be produced by the reactions of N (a) atoms with H (a) atoms with the help of Ru/Al₂O₃. Essentially, this research shows that a small amount of NH₃ (several mg per kW per h) is detected over alumina when the DBD reactor is on. The presence of Ru significantly promotes the production of NH₃ (Fig. 16(a)).

Fig. 16 Reaction routes for NH₃ synthesis and by-product formation by DBD-type plasma combined with the Ru/Al₂O₃ catalyst under different reaction gases: (a) H₂ + N₂; and (b) H₂O + N₂; (c) H₂O + H₂ + N₂.

NH₃, NO₂, and N₂O were formed from H₂O and N₂, because H₂O is dissociated to give H and O atoms, and OH active species with the help of non-thermal plasma (Fig. 16(b)). For the reaction gas of H₂O, H₂ and N₂, only N₂O, as the by-product, was detected when the H₂ content was less than 40% and N₂O also disappeared when the H₂ content was above 40% (Fig. 16(c)).

5. Conclusions

NH₃ can be synthesized from reaction gases consisting of H₂O + N₂, H₂ + N₂ or H₂O + H₂ + N₂ by DBD-type plasma combined with catalyst at atmospheric pressure and room temperature. The largest NH₃ yields from H₂ + N₂ (640 mg kW⁻¹ h⁻¹) and H₂O + H₂ + N₂ (680 mg kW⁻¹ h⁻¹) are much higher than that from H₂O + N₂ (18 mg kW⁻¹ h⁻¹), which is attributed to the fact that H₂ molecules dissociate into hydrogen atoms more easily than H₂O molecules.

The dissociation degrees of H₂O, H₂ and N₂ enhance with an increase in energy density, which results in an increase in NH₃ yield. However, energy densities higher than 500 J L⁻¹ and 1400 J L⁻¹ for the reaction gases of H₂O + H₂ and H₂ + (H₂O) + N₂, respectively, are unfavorable for NH₃ synthesis because the discharge power is more inclined to dissociate H₂O and H₂ instead of N₂, which results in an N/H ratio below the optimal value.

Besides NH₃, some by-products such as N₂O and NO₂ are also formed in the reaction gases containing H₂O. However, their formation can be suppressed in the presence of H₂, which is attributed to the reduction effect of H₂ on the by-products. Furthermore, the by-products disappear for reaction gases containing higher than 40% H₂.

Acknowledgements

The authors wish to thank Prof. A. Mizuno from the Toyohashi University of Technology for valuable discussions. The authors also gratefully acknowledge the financial support by the National Natural Science Foundation of China (No. 21377009 and 21547004).

References

1. G. Busca, L. Lietti, G. Ramis and F. Berti, Appl. Catal., B, 1998, 18, 1–36.
2. W. Fan, T. Zhu, Y. Sun and D. Lv, Chemosphere, 2014, 113, 182–187.
3. M. T. Javed, N. Irfan and B. M. Gibbs, J. Environ. Manage., 2007, 83, 251–289.
4. L. Wang, H. Chen, M. Yuan, S. Rivillon, E. H. Klingenberg, X. Li and R. T. Yang, Appl. Catal., B, 2014, 152, 162–171.
5. Y. Iitsuka, H. Yamauchi, S. Sato, K. Takashima, A. Mizuno and G. Prieto, IEEE Trans. Ind. Appl., 2012, 48, 872–877.
6. N. Saadatjou, A. Jafari and S. Sahebdelfar, Chem. Eng. Commun., 2015, 202, 420–448.
7. R. Michalsky, B. J. Parman, V. Amanor-Boadu and P. H. Pfromm, Energy, 2012, 42, 251–260.
8. S. Samukawa, M. Hori, S. Rauf, K. Tachibana, P. Bruggeman, G. Kroesen, J. C. Whitehead, A. B. Murphy, A. F. Gutsol and S. Starikovskaia, J. Phys. D: Appl. Phys., 2012, 45, 1–5.
9. X. Zhu, X. Gao, R. Qin, Y. Zeng, R. Qu, C. Zheng and X. Tu, Appl. Catal., B, 2015, 170–171, 293–300.
10. X. Tu and J. C. Whitehead, Appl. Catal., B, 2012, 125, 439–448.
11. K. Zhang, B. Eliasson and U. Kogelschatz, Ind. Eng. Chem. Res., 2002, 41, 1462–1468.
12. A. K. Brewer and J. W. Westhaver, J. Phys. Chem., 1930, 34, 153–164.
13. H. Uyama and O. Matsumoto, Plasma Chem. Plasma Process., 1989, 9, 421–432.
14. H. Uyama and O. Matsumoto, Plasma Chem. Plasma Process., 1989, 9, 13–24.
15. H. Uyama and O. Matsumoto, Plasma Chem. Plasma Process., 1996, 16, 547–562.
16. T. Mizushima, K. Matsumoto, J. I. Sugoh, H. Ohkita and N. Kakuta, Appl. Catal., A, 2004, 265, 53–59.
17. J. Hong, S. Prawer and A. B. Murphy, IEEE Trans. Plasma Sci., 2014, 42, 2338–2339.
18. M. Bai, Z. Zhang, X. Bai, M. Bai and W. Ning, IEEE Trans. Plasma Sci., 2003, 31, 1285–1291.
19. M. Bai, X. Bai, Z. Zhang and M. Bai, Plasma Chem. Plasma Process., 2000, 20, 511–520.
20. T. Mizushima, K. Matsumoto, H. Ohkita and N. Kakuta, Plasma Chem. Plasma Process., 2007, 27, 1–11.
21. J. H. van Helden, W. Wagemans, G. Yagci, R. A. B. Zijlmans, D. C. Schram, R. Engeln, G. Lombardi and G. D. Stancu, J. Appl. Phys., 2007, 101, 101–112.
22. K. Sugiyama, K. Akazawa, M. Oshima, H. Miura, T. Matsuda and O. Nomura, Plasma Chem. Plasma Process., 1986, 6, 179–193.
23. W. Rarog, Z. Kowalczyk, J. Sentek, D. Skladanowski and J. Zielinski, Chem. Lett., 2000, 68, 163–168.
24. GB/T 14668-93, Air quality-determination of ammonia-Nessler's reagent colorimetric method.
25. F. Sen and G. Gokagac, Energy Fuels, 2008, 22, 1858–1864.
26. A. Fridman, in Plasma Chemistry, Cambridge University, New York, 2008, ch. 2, p. 33.
27. J. Kikuchi, S. Fujimura, M. Suzuki and H. Yano, Jpn. J. Appl. Phys., 1993, 32, 70–76.
28. J. Nakajima and H. Sekiguchi, Thin Solid Films, 2008, 516, 4446–4451.

---

Footnote

† These authors contributed equally.

---