Influence of Different Catalysts on Ammonia Synthesis Performance in Coaxial DBD Plasma

Abstract

In the renewable energy-driven “green electricity–green hydrogen–green ammonia” pathway, the development of low-temperature and low-energy-consumption ammonia synthesis technologies is of great significance. In this work, a plasma-catalytic ammonia synthesis system was established using a coaxial dielectric barrier discharge (DBD) reactor. The effects of different catalysts, including Ag, Cu, γ-Al₂O₃, BaTiO₃ and Co/BaTiO₃, Ni/BaTiO₃ on ammonia synthesis performance were systematically investigated. The reaction process was analyzed using voltage–current waveforms, Lissajous figures, and optical emission spectroscopy (OES). The results show that different catalytic systems have a significant influence on ammonia synthesis performance, with the promotional effect ranked as follows: Ni/BaTiO₃ > Co/BaTiO₃ > BaTiO₃ > Ag > γ-Al₂O₃ > Cu. Among them, Ni/BaTiO₃ exhibited the best performance. Under the conditions of N₂:H₂ = 1:1 and a gas flow rate of 2.5 L/min, the NH₃ synthesis rate reached 259.48 μmol/min, and the maximum energy efficiency reached 1.40 g-NH₃/kWh. Catalyst characterization results indicate that the BaTiO₃ support maintained a stable crystal structure, while the loaded metal species were highly dispersed and uniformly distributed on the support surface, which is beneficial for the adsorption and conversion of reactive species on the catalyst surface. Discharge characteristic analysis shows that the introduction of BaTiO₃ enhanced the local electric field and improved the uniformity of micro-discharges, while the further incorporation of metal active components strengthened the micro-discharge behavior. OES results reveal that the intensities of characteristic emission lines, such as NH, N₂⁺, and Hα, were significantly enhanced in the Ni/BaTiO₃ system, facilitating the formation and conversion of NH_x intermediates. The superior performance of Ni/BaTiO₃ is attributed to the coupling between BaTiO₃-induced dielectric enhancement and Ni-promoted surface hydrogenation and NH₃ desorption. This work provides mechanistic insight into catalyst-dependent DBD plasma-catalytic ammonia synthesis and offers an experimental basis for the further optimization of plasma-based ammonia production.

1. Introduction

The flexible conversion between green electricity and green hydrogen is widely regarded as an effective pathway for mitigating the challenges associated with renewable energy integration. Hydrogen energy, characterized by cleanliness, high efficiency, and low carbon emissions, is emerging as a crucial energy carrier for replacing fossil fuels and facilitating the transition toward a zero-carbon energy system [1]. However, hydrogen still faces technical bottlenecks in storage and transportation, such as low volumetric energy density and stringent safety requirements, which limit its large-scale application. In contrast, ammonia (NH₃), as an important basic chemical feedstock, is widely used in fertilizers, environmental protection, refrigeration, and defense industries. It also possesses advantages such as high hydrogen content, easy liquefaction, and convenient storage and transportation, and is therefore regarded as a highly promising hydrogen storage medium and energy carrier [2,3]. Therefore, the development of ammonia-based hydrogen energy utilization pathways is of great significance.

Currently, industrial ammonia production mainly relies on the Haber–Bosch process, which operates under high-temperature (400–500 °C) and high-pressure (10–30 MPa) conditions [4]. This process also depends on fossil-fuel-derived hydrogen, resulting in high energy consumption and significant CO₂ emissions [5,6]. In addition, it is dominated by large-scale centralized production, making it difficult to adapt to the intermittency and distributed nature of renewable energy. Therefore, the development of novel ammonia synthesis technologies that can operate under mild conditions and be coupled with renewable energy has become a key research focus.

In recent years, various emerging ammonia synthesis routes have been proposed, including electrochemical nitrogen reduction, photocatalytic ammonia synthesis, and low-temperature thermos-catalytic methods [7]. However, these technologies generally suffer from low ammonia production rates, insufficient energy efficiency, or poor reaction stability, which limits their practical applications [8,9]. In contrast, non-thermal plasma (NTP) technology can activate and dissociate N₂ molecules through high-energy electrons under ambient temperature and pressure, thereby bypassing the high energy barriers associated with conventional thermal catalysis and providing a new reaction pathway for ammonia synthesis. In non-equilibrium plasma systems, the electron temperature is significantly higher than the gas temperature, enabling the generation of vibrationally excited N₂, N radicals, and H radicals, which promote NH₃ formation [10,11]. Among various plasma types, dielectric barrier discharge (DBD) is widely used in plasma-assisted ammonia synthesis due to its simple structure, stable operation at atmospheric pressure, and ease of coupling with catalysts [12]. The DBD plasma-catalytic ammonia synthesis process involves multiple processes, including gas-phase discharge activation, catalyst surface reactions, and plasma–catalyst interactions. Therefore, the research focus of this system is not only to improve the NH₃ synthesis rate and energy efficiency, but also to further understand the effects of different catalysts on discharge characteristics, reactive species generation, and surface reaction pathways. A systematic investigation of the influence of different types of catalysts on DBD behavior, reactive species formation, and NH₃ synthesis pathways is of great significance for deepening the understanding of the mechanism of plasma-catalytic ammonia synthesis.

In terms of reactor configuration, coaxial DBD reactors coupled with catalysts for ammonia synthesis have been well established and extensively reported in the literature [13,14,15,16,17]; this reactor configuration is suitable for plasma-catalytic ammonia synthesis under low-temperature conditions and represents one of the promising approaches for distributed small-scale ammonia synthesis technologies. Compared with simple optimization of reactor structure, the introduction of catalysts is considered a key factor for further improving the NH₃ synthesis rate and energy efficiency. In plasma-catalysis synergistic systems, heterogeneous catalysts not only provide surface reaction sites but also modulate discharge characteristics, thereby influencing the generation and conversion of reactive species [18]. Previous studies have demonstrated that different types of catalysts have a significant impact on ammonia synthesis performance. Hong et al. [19] employed a nanodiamond-coated catalyst deposited on α-Al₂O₃ spheres, achieving a maximum energy efficiency of 0.29 g/kWh. Wang et al. [13] investigated the role of a Ni/Al₂O₃ catalyst in plasma-enhanced catalytic ammonia synthesis and found that the synergistic effect between Ni active sites and weak acidic sites on the Al₂O₃ surface significantly promoted the formation of NH₂ intermediates, leading to an ammonia synthesis rate of 471 μmol/g/h. In addition, Rouwenhorst et al. [20] studied the catalytic effect of Ru-K/MgO catalysts on DBD plasma ammonia synthesis and found that the synergistic effect between Ru active sites and potassium promoter modification resulted in a maximum ammonia synthesis rate of 1750 μmol/g/h and a maximum energy efficiency of 1.23 g/kWh. Andersen et al. [21] combined a zero-dimensional kinetic model to investigate the influence of M/MgAl₂O₄ (Ru, Co) catalysts on DBD ammonia synthesis, obtaining a maximum NH₃ synthesis rate of 1476 μmol/h and a maximum energy efficiency of 1.19 g/kWh, and demonstrated that N₂ dissociation is the rate-determining step in ammonia synthesis. Liu et al. [22,23] compared the effects of dielectric materials with different dielectric constants (BaTiO₃, TiO₂, SiO₂) on DBD ammonia synthesis and found that BaTiO₃ achieved the highest NH₃ concentration of 1344 ppm, which is 106% higher than that of SiO₂, with an energy efficiency of 0.68 g/kWh. They further investigated catalysts consisting of Fe, Co, and Ni loaded BaTiO₃ supports, among which Ni/BaTiO₃ exhibited the best performance, reaching a maximum NH₃ concentration of 2912 ppm and a maximum energy efficiency of 0.78 g/kWh. It was found that high-dielectric-constant materials can enhance the reduced electric field in the plasma and increase the proportion of high-energy electrons, thereby strengthening electron-impact reactions, promoting the generation of reactive species such as N and H, and accelerating NH₃ formation. The relevant references mentioned in this study, along with recent research on DBD plasma–catalyst coupling for ammonia synthesis, are summarized in

Table 1. Current research on DBD plasma-catalytic ammonia synthesis using coupled catalysts.

Year	Catalyst	N2:H2	Flow Rate (mL/min)	Power (W)	Concentration (ppm)	Synthesis Rate (μmol/g/h)	Energy Efficiency (g/kWh)	Ref.
2016	O-ND/Al2O3	1:3	60	98	2300		0.16	[19]
2019	Ni/Al2O3	1:2	56	25.1		471	0.29	[13]
2021	Ru-K/MgO	1:1	20			1750	1.23	[20]
2022	Co-Ni/Al2O3	1:1	200	30.81		3000	0.83	[24]
2022	Ni/BaTiO3	1:1	100	25	2912		0.78	[23]
2023	Co/MgAl2O4	1:1	80	10	3000		1.19	[21]
2024	Co-Ni/MOF-74	1:1	200	30.84		2609	0.72	[17]
2025	CaH2	1:2	200	27		6440	1.2	[25]
2026	Ni/BaTiO3	1:1	2500	196	5015.14	15,569	1.40	This work

.

Although the above studies have made important progress in improving the performance of DBD plasma-catalytic ammonia synthesis, several issues still require further investigation. First, it remains challenging to achieve both a high NH₃ synthesis rate and high energy efficiency under relatively high gas throughput. Second, the roles of different catalyst properties in regulating DBD characteristics, reactive species generation, and surface conversion pathways are still not fully understood. In particular, the respective contributions of metal conductivity, oxide surface properties, dielectric polarization, and transition-metal surface reactivity need to be further distinguished under comparable plasma reaction conditions. To address these issues, the catalysts in this work were selected to represent different material functions in the plasma-catalytic ammonia synthesis system. Ag and Cu were used as representative conductive metal catalysts to examine whether metallic conductivity and metal particle-induced electric-field redistribution alone can promote NH₃ formation. γ-Al₂O₃ was selected as a conventional oxide support with relatively low dielectric constant, while BaTiO₃ was selected as a high-dielectric-constant material to evaluate the role of dielectric polarization and local electric-field enhancement. Co/BaTiO₃ and Ni/BaTiO₃ were further introduced to investigate the combined effect of a high-dielectric support and transition-metal active species. Therefore, the catalyst selection in this work was intended to construct a representative comparison framework involving metal conductivity, oxide surface properties, dielectric enhancement, and supported metal active sites, rather than simply increasing the number of catalyst samples.

Based on this, a coaxial DBD plasma-catalytic ammonia synthesis experimental platform was constructed in this work. Co/BaTiO₃ and Ni/BaTiO₃ supported catalysts were prepared, and typical catalytic materials such as Ag, Cu, and γ-Al₂O₃ were introduced as comparative systems to systematically investigate the effects of different catalysts on NH₃ synthesis performance and plasma behavior. Combined with voltage–current waveforms, Lissajous figures, OES diagnostics, and characterizations including XRD, SEM, EDS, BET, and XPS, further deepens the understanding of the dielectric enhancement effect of BaTiO₃, the surface reactions of Ni/Co metal active species, and the synergistic mechanism of plasma–catalyst coupling in ammonia synthesis.

2. Experimental Section

2.1. Experimental Setup

This work constructed a DBD plasma-catalytic ammonia synthesis experimental platform, as shown in Figure 1. The coaxial DBD reactor chamber used a quartz tube with an outer diameter of 28 mm and a wall thickness of 1.5 mm as the dielectric layer. A copper mesh wrapped around the outer surface of the quartz tube served as the outer electrode, with a total length of 20 cm. A smooth copper rod with an outer diameter of 22 mm was used as the high-voltage inner electrode, ensuring a discharge gap of 1.5 mm. The corresponding discharge setup is shown in Figure 2. A gas mixture of N₂ and H₂ with a purity of 99.999% was introduced into the reactor inlet. The flow rates of N₂ and H₂ were controlled by mass flow controllers (Sevenstar CS200, 0–5 SLM, Sevenstar Flowmeter Co., Ltd., Beijing, China) and fed into the DBD reactor. The outlet gas was analyzed using a Fourier-transform infrared (FTIR) spectrometer to measure gas concentrations. The DBD reactor was powered by an AC high-voltage power supply (CTP-2000KP, Suman Electronic Co., Ltd., Nanjing, China). The discharge voltage was measured using a high-voltage probe (P6015A, Tektronix Inc., Beaverton, OR, USA), while the discharge current was monitored using a current probe (Pearson 2877, Pearson Electronics, Palo Alto, CA, USA). A sampling capacitor (C_M = 0.47 μF) was connected in series at the grounded terminal of the reactor to measure the total transferred charge in the plasma discharge. All output signals were recorded by an oscilloscope (TBS2204B, Tektronix Inc., Beaverton, OR, USA) for voltage-current waveform analysis and discharge power calculation. In this study, the discharge voltage was adjusted within the range of 6–10 kV, the discharge frequency was fixed at 6 kHz, and the gas flow rate was controlled between 0.5 and 3 L/min. In addition, the volumetric ratio of N₂/H₂ was varied from 1:4 to 4:1 to investigate its effect on ammonia synthesis performance. It is noteworthy that the dimensions of the DBD reactor used in this study were determined based on preliminary tests of discharge stability and ammonia synthesis performance. By adjusting the outer diameter of the inner electrode, the effects of different discharge gaps were compared, and the influence of discharge length on the ammonia generation rate was also investigated. Based on these results, the reactor configuration was finally selected by comprehensively considering stable discharge, relatively low energy consumption, appropriate gas residence time, and structural operability. It should be noted that both the electrode material and electrode structure may affect the DBD characteristics and reaction performance. However, the focus of this study is to reveal the influence of different catalysts on plasma-assisted ammonia synthesis performance. Therefore, to ensure the comparability of the catalyst comparison experiments, all experiments were conducted using the same reactor structure and electrode configuration. The optimization of electrode structure and the effects of different electrode materials on ammonia synthesis performance will be systematically investigated in future studies.

2.2. Parameter Definition and Calculation

The discharge power (P_dis) was determined using the Lissajous figure method:

$$P_{\mathrm{dis}}=f·C_{\mathrm{M}}·S$$

(1)

where C_M is the external capacitance, f is the frequency of the AC power supply, and S represents the area enclosed by the Lissajous figure.

FTIR was used to quantitatively measure the NH₃ concentration in the outlet gas. During the entire measurement process, the mass flow controllers maintained a constant total gas flow rate. The produced NH₃ was introduced into the FTIR gas cell (optical path length: 20 cm, CaF₂ windows). Once the FTIR signal became stable, the absorption spectra were recorded. In this study, the NH₃ concentration was quantified based on a calibration curve established using NH₃ standard gases with nitrogen as the balance gas. Specifically, NH₃ standard gases with concentrations of 600, 1200, 1500, 1800, 2100, 2400, and 3000 ppm were prepared by adjusting the flow rates of a 3000 ppm NH₃ standard gas and high-purity N₂ gas using mass flow controllers. The FTIR absorption spectra of the NH₃ standard gases with different concentrations were then measured and integrated. Based on the Beer–Lambert law, a linear relationship between the NH₃ concentration and the integrated absorbance area was established, thereby obtaining the NH₃ example spectra and calibration curve, as shown in Figure 3. The results show a good linear correlation between the NH₃ concentration and the integrated absorbance area, with a coefficient of determination R² close to 1.

NH₃ synthesis rate is calculated by:

$$Y_{\mathrm{N}{\mathrm{H}}_3}={\displaystyle \frac{Q_{\mathrm{total}}}{24\hspace{1em}\mathrm{L}\cdot {\mathrm{mol}}^{-1}}}\times c_{\mathrm{N}{\mathrm{H}}_3}$$

(2)

where Q_total represents the total flow rate of N₂ and H₂, 24 L·mol⁻¹ is the molar volume of gas at room temperature, and $c_{{\mathrm{NH}}_3}$ denotes the measured NH₃ concentration the NH₃ energy consumption is calculated by:

$$E_{{\mathrm{NH}}_3}={\displaystyle \frac{P_{\mathrm{dis}}\times 60}{Y_{{\mathrm{NH}}_3}}}$$

(3)

2.3. OES

OES was used to evaluate the relative variations in the emission intensities of key reactive species over different catalysts. During the experiments, plasma emission was collected from the side of the discharge region of the DBD reactor via an optical fiber and transmitted to a spectrometer (SR-750-B1, Andor Technology Ltd., Belfast, UK, equipped with an 1800 grooves/mm grating and a 50 μm entrance slit, providing a spectral resolution of 0.05 nm) for analysis. To improve the detection sensitivity of weak emission signals, a high-sensitivity electron-multiplying charge-coupled device (DU970P_BVE, EMCCD, Andor Technology Ltd., Belfast, UK) detector was employed, enabling effective acquisition of weak molecular and atomic emission lines generated in the discharge. All OES measurements were performed after the reactor reached a steady-state operation under each set of experimental conditions. The optical collection position was kept identical for all tests. The EMCCD gain was set to 50, and the exposure time was fixed at 0.6 s to ensure the comparability of spectral data among different catalysts.

2.4. Preparation of Catalysts

In this study, the Ag, Cu, γ-Al₂O₃, and BaTiO₃ catalysts used for comparison were commercially obtained and used without further treatment. Ag was obtained from Macklin Biochemical Co., Ltd. (Shanghai, China, CAS: 7440-22-4), and Cu was obtained from Macklin Biochemical Co., Ltd. (Shanghai, China, CAS: 7440-50-8), γ-Al₂O₃ was obtained from Aladdin Chemical Co., Ltd. (Shanghai, China, CAS: 1344-28-1), and BaTiO₃ was obtained from Aladdin Chemical Co., Ltd. (Shanghai, China, CAS: 12047-27-7). Co/BaTiO₃ and Ni/BaTiO₃ were prepared using commercial BaTiO₃ as the support.

Supported catalysts were prepared via a wetness impregnation method. BaTiO₃ with a high dielectric constant was used as the support, and the metal loading was 3 wt.%, with a total catalyst mass of 1.0 g. The water uptake capacity of the support was determined by dropwise addition of deionized water until complete saturation, yielding a value of 1.2 ± 0.2 mL/g. Metal precursor solutions were then prepared according to the target loading. For the Co/BaTiO₃ catalyst, the precursor solution was obtained by reacting Co₃O₄ (Macklin Biochemical, Shanghai, China, CAS: 1307-96-6) powder with dilute hydrochloric acid. For the Ni/BaTiO₃ catalyst, Ni(CH₃COO)₂·4H₂O (Macklin Biochemical, Shanghai, China, CAS: 373-02-4) was dissolved in deionized water to form the precursor solution. The prepared precursor solution was gradually added to the BaTiO₃ support, followed by thorough stirring to ensure uniform dispersion of the metal species on the support surface. The resulting samples were dried at 120 °C for 5 h and subsequently reduced at 500 °C for 5 h under a 10% H₂ atmosphere with a flow rate of 200 mL/min. Finally, 3 wt.% Co/BaTiO₃ and Ni/BaTiO₃ catalysts were obtained. In all plasma-catalytic packed-bed experiments, the catalysts were supported on quartz wool and packed into the discharge region of the DBD reactor. Specifically, 2.0 g of quartz wool was first uniformly packed into the 20 cm long discharge region. Subsequently, 1.0 g of catalyst powder, including Ag, Cu, γ-Al₂O₃, BaTiO₃, Co/BaTiO₃, or Ni/BaTiO₃, was weighed and uniformly dispersed on the surface of the quartz wool to ensure a consistent catalyst loading per unit length in the discharge region across different experiments. For the blank control, only quartz wool was used. For the quartz bead experiment, quartz beads with an average diameter of approximately 1 mm were directly filled into the discharge region to form a stable packed bed without obvious movement during gas flow.

2.5. Material Characterization

Figure 4 shows the N₂ adsorption–desorption isotherms of the catalysts, and the corresponding textural properties, including the Brunauer–Emmett–Teller (BET) surface area, pore volume, and average pore diameter, are summarized in

Table 2. Physical characteristics of Co/BaTiO₃ and Ni/BaTiO₃ catalysts before and after the reaction.

Number	Catalysts	B-R/A-R	Surface Area (m2/g)	Pore Volume (cm3/g)	Average Pore Diameter (nm)
(a)	Co/BaTiO3	B-R	7.59	0.028	11.04
(b)	Co/BaTiO3	A-R	6.97	0.025	14.09
(c)	Ni/BaTiO3	B-R	3.53	0.010	8.61
(d)	Ni/BaTiO3	A-R	2.73	0.009	11.28

. The measurements were performed using a 3Flex/ASAP 2460 surface area and porosity analyzer (Micromeritics, Norcross, GA, USA).

The Co/BaTiO₃ and Ni/BaTiO₃ catalysts exhibited surface areas of 7.59 and 3.53 m²/g before reaction, respectively, which slightly decreased to 6.97 and 2.73 m²/g after the plasma catalytic reaction. Meanwhile, the pore volume of Co/BaTiO₃ changed from 0.028 to 0.025 cm³/g, while that of Ni/BaTiO₃ changed from 0.010 to 0.009 cm³/g. In addition, the average pore diameters of both catalysts were within the mesoporous range, and the nitrogen adsorption–desorption isotherms exhibited typical type-IV adsorption behavior, indicating mesoporous structural characteristics. The relatively small changes in the BET parameters before and after reaction suggest that the catalysts maintained relatively stable pore structures and accessible reaction surfaces under plasma reaction conditions, which may help maintain gas diffusion pathways and catalytic interfaces during ammonia synthesis.

X-ray diffraction (XRD, XPert Pro, PANalytical B.V., Almelo, Netherlands) was employed to characterize the crystal structures of Co/BaTiO₃ and Ni/BaTiO₃ catalysts, as shown in Figure 5. The results indicate that both catalysts retain the characteristic diffraction peaks of the BaTiO₃ support, which are located at 2θ = 22.240°, 31.657°, 38.895°, 45.379°, 56.282° and 66.119°, corresponding to the (100), (110), (111), (200), (211) and (220) crystal planes of BaTiO₃ (BaTiO₃, PDF#01-075-0460), respectively. This confirms that metal loading and subsequent thermal treatment do not alter the main crystalline phase of BaTiO₃, and the support maintains its tetragonal perovskite structure. In addition, for Ni/BaTiO₃, a weak diffraction peak associated with metallic Ni appeared at 2θ = 44.507°, which can be assigned to the Ni (111) plane (Ni, PDF#04-004-3109). For Co/BaTiO₃, weak Co-related diffraction peaks were observed at 2θ = 51.149° and 75.250°, which can be assigned to the Co (200) and (220) planes, respectively (Co, PDF#01-071-4238). Due to the low loading amount of Co and Ni active species, only 3 wt.%, and the partial overlap between metal-related peaks and BaTiO₃ diffraction peaks, these metal diffraction peaks exhibited relatively low intensities.

The crystallite size estimation was performed based on XRD peak fitting and full width at half maximum (FWHM) analysis, and the crystallite sizes were calculated using the Scherrer equation. The results show that the average crystallite size of the BaTiO₃ support in the Co/BaTiO₃ catalyst increased from 37.3 ± 1.8 nm before reaction to 47.0 ± 2.1 nm after reaction, while the average crystallite size of the BaTiO₃ support in the Ni/BaTiO₃ catalyst changed from 40.3 ± 1.6 nm before reaction to 37.9 ± 1.1 nm after reaction, indicating that both catalysts maintained relatively good structural stability under the plasma reaction conditions. In addition, due to the relatively low loading amount of Co and Ni active species (3 wt.%), the corresponding XRD diffraction peaks were weak, making it difficult to reliably calculate the crystallite sizes of the metallic active species using the Scherrer equation.

Field-emission scanning electron microscopy (SEM, Zeiss SIGMA, Carl Zeiss, Oberkochen, Germany) was employed to characterize the morphologies of the Co/BaTiO₃ and Ni/BaTiO₃ catalysts before and after reaction, as shown in Figure 6.

In addition, energy-dispersive X-ray spectroscopy (EDS, Oxford Instruments plc., Oxford UltimMax 40, Abingdon, UK) was employed to analyze the elemental distribution of Co/BaTiO₃ and Ni/BaTiO₃ catalysts before and after reaction, as shown in Figure 7, and the corresponding elemental composition results are summarized in

Table 3. EDS elemental composition of Co/BaTiO₃ and Ni/BaTiO₃ catalysts before and after reaction.

Sample		B-R		A-R
Co/BaTiO3	element	wt.%	wt.% Sigma	wt.%	wt.% Sigma
	O	20.97	0.94	13.95	1.11
	Ti	25.05	0.70	30.61	1.03
	Co	1.29	0.50	1.49	0.64
	Ba	52.69	1.01	53.94	1.33
Ni/BaTiO3	O	16.39	0.74	18.04	0.75
	Ti	22.70	0.62	23.46	0.64
	Ni	3.12	0.52	2.23	0.51
	Ba	57.79	0.88	56.27	0.90

. The EDS elemental mapping images show that Ba, Ti, and O were clearly distributed in the catalyst regions, indicating that the main elements of the BaTiO₃ support were retained after the plasma-catalytic reaction. Meanwhile, Co and Ni signals were detected in the corresponding catalysts. This indicates that Co and Ni active species were introduced onto the BaTiO₃ support.

Further qualitative analysis of the dispersion state of the active metal species was performed based on SEM morphology, EDS elemental mapping, and elemental composition analysis. The SEM results showed that no obvious large-scale particle agglomeration was observed either before or after the reaction, indicating that the overall catalyst morphology remained relatively stable. The EDS elemental mapping results revealed relatively uniform distributions of Co and Ni on the BaTiO₃ support surface. Meanwhile, the elemental composition analysis showed only slight changes in the Co and Ni contents before and after the reaction, suggesting that the active metal species could be stably retained on the catalyst surface during the plasma reaction process. Therefore, based on the XRD, SEM, and EDS characterization results, the Co/Ni active species can be considered to be relatively uniformly dispersed on the BaTiO₃ support surface.

X-ray Photoelectron Spectroscopy (XPS, ESCALAB Xi+, Thermo Fisher Scientific, Waltham, MA, USA) analysis was further performed to investigate the surface chemical states of the Co/BaTiO₃ and Ni/BaTiO₃ catalysts before and after the plasma-catalytic reaction, as shown in Figure 8. For Co/BaTiO₃, the Co 2p signal was relatively weak, and the Co 2p region may partially overlap with the Ba 3d signal from the BaTiO₃ support, making it difficult to clearly distinguish and deconvolute the main Co 2p peaks. Co-related satellite features were observed before and after reaction, indicating the presence of surface Co species on the BaTiO₃ support. The retention of these Co-related satellite features after reaction suggests that the surface Co species were relatively stable under the present plasma-catalytic reaction conditions. For Ni/BaTiO₃, the Ni 2p spectra exhibited Ni-related peaks and satellite peaks before reaction, confirming the presence of surface Ni species on the BaTiO₃ support. After reaction, the Ni-related signals were still observed without obvious disappearance, suggesting that the surface Ni species were largely retained under the present plasma reaction conditions.

Therefore, the comparison of XPS spectra before and after the reaction indicates that the supported Co, Ni surface species showed reasonable stability during the plasma-catalytic process. Combined with the BET results, these characterizations confirm the real specific surface area, accessible reaction surface, and surface-exposed metal species of the catalysts. These surface-exposed Co/Ni species may serve as potential active sites for plasma-catalytic ammonia synthesis. Combined with the EDS elemental mapping results, Ni and Co species are relatively well dispersed on the BaTiO₃ surface, with no obvious large-scale aggregation observed. The SEM results also show that no severe collapse or sintering of the catalyst morphology occurs before and after the reaction. Therefore, Co and Ni species can serve as potential surface reaction sites for the adsorption of plasma-activated N₂/H₂ species, surface hydrogenation, and NH₃ desorption.

It should be noted that XPS mainly reflects the surface elemental composition and chemical states of the catalyst, while EDS mainly reflects the spatial distribution of elements. These two techniques can provide evidence for the existence and distribution of surface-active metal species, but they cannot be directly regarded as an accurate quantification of the true active site density.

3. Results and Discussion

3.1. Discharge Characteristics with Different Catalysts

To investigate the influence of different packed catalysts on the discharge characteristics of DBD, the real-time discharge voltage and current waveforms during ammonia synthesis were measured. Figure 9 presents the voltage and current waveforms under identical operating conditions (peak voltage of 10 kV and frequency of 6 kHz) for the plasma-only case, quartz wool packing, and quartz wool-supported Ag, Cu, γ-Al₂O₃, BaTiO₃, as well as supported catalysts Co/BaTiO₃ and Ni/BaTiO₃. Under all experimental conditions, the discharge waveforms exhibit typical filamentary micro-discharge characteristics of DBD [26].

When quartz wool was introduced into the discharge gap, both the number and amplitude of current pulses increased compared with the plasma-only condition. This indicates that the porous structure of quartz wool can create numerous micro-discharge channels within the discharge gap, thereby enhancing local electric-field inhomogeneity, increasing the number of micro-discharges, and strengthening the overall discharge intensity. In contrast, when 1 mm quartz beads were packed in the discharge gap, the current pulse amplitude decreased significantly. This is mainly because quartz beads, as insulating dielectrics, accumulate surface charges during discharge and generate a counter electric field, which suppresses the development of electron avalanches within individual micro-discharge channels and thereby inhibits the instantaneous discharge current. Meanwhile, the packing of quartz beads shortens the effective discharge length of each micro-discharge channel, further leading to a reduction in current pulse amplitude [27]. For the Ag catalyst supported on quartz wool, the current pulse amplitude was slightly lower than that observed for quartz wool alone, indicating that the presence of Ag metal particles partially modified the local electric-field distribution and stabilized certain discharge channels. In the Cu catalyst system, both the number and amplitude of current pulses increased significantly, suggesting that Cu generated additional electric-field-enhanced regions within the discharge gap, thereby promoting the formation of more micro-discharge channels. For the γ-Al₂O₃ and BaTiO₃ catalyst systems, the current waveforms exhibit more densely distributed and more uniformly arranged micro-discharge pulses. Owing to their relatively high dielectric constants, γ-Al₂O₃ and BaTiO₃ possess strong dielectric polarization capability, which significantly enhances the local electric field distortion within the discharge gap. This facilitates electron energy gain and promotes gas breakdown, leading to a more uniform and stable micro-discharge distribution in the plasma region [22]. Upon further introduction of the supported catalysts Co/BaTiO₃ and Ni/BaTiO₃, the number of current pulses increases further, accompanied by a more densely distributed pulse pattern. Compared with the BaTiO₃-only system, metal loading introduces additional active sites with enhanced electron emission capability on the catalyst surface. These metallic active centers can promote secondary electron emission, thereby reducing the electric-field strength required for gas breakdown and further facilitating the formation of micro-discharge channels [28].

To further investigate the influence of different catalysts on the DBD characteristics, Lissajous figures were constructed under identical operating conditions based on the measured discharge voltage and the voltage across the sampling capacitor, as shown in Figure 10. It can be observed that the closed-loop shapes under different conditions change from a parallelogram-like form to an elliptical shape, indicating variations in discharge behavior. The Lissajous patterns for Ag-, Cu-, and γ-Al₂O₃-filled systems exhibit similar morphologies, suggesting that these catalysts have no significant effect on the overall discharge characteristics. In contrast, when BaTiO₃, Co/BaTiO₃, and Ni/BaTiO₃ are introduced, the Lissajous figures show a more pronounced elliptical tendency, indicating a transition in the discharge mode from typical filamentary micro-discharges to a combination of surface discharge and weakened micro-discharge behavior [29,30]. The introduction of high-dielectric-constant BaTiO₃ not only facilitates the formation of a more uniform and stable plasma discharge region, but also increases the effective collision probability between electrons and reactant molecules, thereby enhancing N₂ activation efficiency [31]. After the introduction of supported catalysts Co/BaTiO₃ and Ni/BaTiO₃, the effective charge transfer per discharge cycle decreases. This is mainly attributed to the fact that the incorporation of metallic active components alters the charge accumulation and release behavior on the catalyst surface. The metal particles promote surface charge redistribution, thereby reducing the degree of charge accumulation on the dielectric surface.

3.2. FTIR Analysis of NH₃ Products in DBD Plasma Synthesis

To confirm the formation of ammonia products in the plasma synthesis process, the reaction effluent was analyzed by FTIR spectroscopy. Figure 11 illustrates the concentration variation in FTIR absorption spectra of Ni/BaTiO₃ under different applied voltages. Distinct characteristic absorption peaks can be observed at approximately 3333 cm⁻¹, 1625 cm⁻¹, and 967 cm⁻¹, which are assigned to the N-H stretching vibration, bending vibration, and deformation vibration of NH₃ molecules, respectively [32]. In addition, a characteristic peak of NH₄⁺ is detected at around 3017 cm⁻¹ [33], indicating the presence of NH₃ and partially formed NH_X intermediates in the gas-phase reaction products. Furthermore, the NH₃ concentration under different operating conditions was quantified based on the integrated intensity of the characteristic NH₃ peak at 967 cm⁻¹, combined with the calibration relationship established above. This enables quantitative comparison of NH₃ production under different catalysts and reaction parameters.

3.3. Effect of N₂/H₂ Ratio on NH₃ Synthesis Rate and Energy Efficiency

Under the conditions of a discharge voltage of 10 kV, a discharge frequency of 6 kHz, and a total gas flow rate of 1 L/min, Figure 12a–c show the variations in NH₃ synthesis rate, energy consumption, and NH₃ concentration, respectively, under different N₂/H₂ volumetric ratios with various catalyst packings. The results indicate that, for all catalyst systems, the NH₃ synthesis rate initially increases and then decreases with increasing H₂ volume fraction, exhibiting a clear non-monotonic trend. Compared with the plasma-only blank system, all catalyst-assisted systems significantly enhance the NH₃ synthesis rate. Among them, the Ni/BaTiO₃ catalyst exhibits the best performance. At a N₂:H₂ ratio of 1:1, the NH₃ synthesis rate reaches 208.96 μmol/min, which is 237.25% higher than that of the plasma-only condition. According to previous studies, the dissociation energy of the N≡N triple bond is 9.8 eV, which is more than twice that of H₂ (4.52 eV). Therefore, the activation and dissociation of N₂ is considered the rate-determining step in ammonia formation [13]. A moderate increase in H₂ content promotes the generation of reactive H species in the plasma, thereby facilitating the recombination of N and H species in both the gas phase and on the catalyst surface [34]. However, when the H₂ fraction becomes too high, excessive energy is consumed in H₂ dissociation, which suppresses N₂ dissociation and consequently leads to a decrease in NH₃ synthesis rate. Overall, the promotion effect of different catalysts follows the order: Ni/BaTiO₃ > Co/BaTiO₃ > BaTiO₃ > Ag > γ-Al₂O₃ > Cu > quartz wool > plasma-only. This indicates that the supported catalysts Ni/BaTiO₃ and Co/BaTiO₃ exhibit superior performance in plasma-assisted ammonia synthesis. This enhancement can be attributed to the high dielectric constant of BaTiO₃, which strengthens the local electric field in the discharge region and improves micro-discharge behavior, thereby increasing electron energy and electron density. In addition, the loaded metal active species (Ni and Co) provide additional surface-active sites, promoting the adsorption and reaction of reactive nitrogen and hydrogen species on the catalyst surface [35], ultimately enhancing NH₃ formation efficiency. A similar dependence of NH₃ synthesis on the N₂/H₂ ratio was reported by Wang et al. for Ni/Al₂O₃, where the feed composition affected the balance between N₂ activation and hydrogenation of nitrogen-containing intermediates; however, the reported energy efficiency was 0.29 g-NH₃/kWh, lower than that obtained in this work [13]. Liu et al. also reported that Ni/BaTiO₃ showed the best performance among M/BaTiO₃ catalysts under optimized N₂/H₂ conditions, but its maximum energy efficiency was 0.78 g-NH₃/kWh, indicating that the present Ni/BaTiO₃ system provides more efficient energy utilization [23].

3.4. Effect of Total Gas Flow Rate on NH₃ Synthesis Rate and Energy Efficiency

In different catalyst systems, the gas flow rate is a key factor influencing the NH₃ synthesis rate and energy efficiency, as it is directly related to the processing capacity of the DBD reactor [36]. Figure 13a–c show the variations in NH₃ synthesis rate and energy consumption, and NH₃ concentration, respectively, as a function of total gas flow rate under the optimal N₂/H₂ volumetric ratio (N₂:H₂ = 1:1) in DBD plasma-catalytic systems with different catalysts. The results indicate that, for all catalyst systems, the NH₃ synthesis rate first increases and then decreases with increasing gas flow rate, while the energy consumption exhibits an initial decreasing and subsequent increasing trend. When the total gas flow rate increases from 0.5 L/min to 2.5 L/min, the Ni/BaTiO₃ system shows an increase in NH₃ synthesis rate from 138.77 μmol/min to 259.48 μmol/min, accompanied by a decrease in NH₃ energy consumption from 86.47 MJ/mol to 45.78 MJ/mol. However, further increasing the gas flow rate led to a decline in NH₃ production performance. This behavior can be explained as follows: at low gas flow rates, the residence time of the gas in the discharge region is excessively long, allowing the NH₃ formed within the plasma zone to undergo secondary dissociation at the downstream end of the discharge region. As the gas flow rate increases, the reactant flux increases, which enhances the effective collision frequency between high-energy electrons and gas molecules. Meanwhile, the reduced residence time suppresses the secondary decomposition of NH₃, thereby promoting NH₃ formation [24]. However, when the flow rate becomes too high, the residence time in the discharge region becomes insufficient, leading to a reduction in the effective electron–molecule collision time, as well as shortened adsorption and reaction times of active species on the catalyst surface, ultimately inhibiting the overall reaction process. Liu et al. reported that the gas flow rate significantly affected Co-Ni/Al₂O₃-assisted DBD ammonia synthesis by changing the residence time and plasma–catalyst contact process, with a maximum energy efficiency of 0.83 g-NH₃/kWh [24]. Andersen et al. also showed that the feed flow rate influenced NH₃ concentration and energy efficiency in Co/MgAl₂O₄-packed DBD systems, but the reported maximum energy efficiency was 1.19 g-NH₃/kWh, which is lower than the 1.40 g-NH₃/kWh achieved in this work [21].

3.5. Effect of Peak Discharge Voltage on NH₃ Synthesis Rate and Energy Efficiency

Figure 14a–c show the effects of discharge peak voltage on NH₃ synthesis rate, energy consumption, and NH₃ concentration, respectively, under the optimal N₂/H₂ volumetric ratio for different catalyst systems. The results indicate that the NH₃ synthesis rate increases with increasing discharge peak voltage for all catalyst configurations, while clear differences are observed among different catalysts. Overall, the performance follows the order: Ni/BaTiO₃ > Co/BaTiO₃ > BaTiO₃ > Ag > γ-Al₂O₃ > Cu > quartz wool support. At a discharge peak voltage of 10 kV, the Ni/BaTiO₃ system achieves the highest NH₃ synthesis rate of approximately 208.96 μmol/min. Meanwhile, at a lower discharge peak voltage of 6 kV, the NH₃ energy consumption decreases to 43.71 MJ/mol. This behavior can be attributed to the fact that increasing the discharge voltage enhances the electric field strength, thereby promoting the generation of high-energy electrons. As a result, the dissociation, excitation, and ionization of N₂ and H₂ molecules are accelerated, leading to the formation of a greater number of reactive species, including N and H radicals, excited N₂, and N₂⁺ ions [13]. Meanwhile, the increased discharge voltage also enhances the micro-discharge density and expands the plasma discharge region, thereby increasing the effective collision probability among reactive species and further promoting the NH₃ synthesis rate [14]. However, it can also be observed that the NH₃ energy consumption gradually increases with increasing discharge voltage. This is because, in DBD plasma-assisted ammonia synthesis, the input power increases nonlinearly with the applied voltage. Under high-voltage conditions, not only is the target reaction promoted, but undesired processes such as gas heating, excessive ionization, and NH₃ decomposition are also intensified, leading to an increase in energy consumption for ammonia synthesis. Similar voltage-dependent behavior has also been reported in previous DBD plasma ammonia synthesis studies. Chen et al. investigated NH₃ synthesis in a coaxial DBD reactor using porous SiO₂ and nonporous glass beads as catalyst supports, and found that both N₂ conversion and NH₃ energy yield increased with the applied voltage [37]; however, their study mainly focused on inert support materials, whereas the present work further introduces Ni/BaTiO₃ with both dielectric enhancement and metal active sites, leading to higher NH₃ synthesis performance. Xu et al. studied the effect of pulse voltage in a DBD reactor and found that, without noble gas addition, the NH₃ synthesis rate first increased and then decreased with increasing voltage, indicating that excessive voltage could promote NH₃ decomposition [38]; by comparison, the Ni/BaTiO₃ system in this work maintains the highest NH₃ synthesis rate among all tested catalysts and achieves a maximum energy efficiency of 1.40 g-NH₃/kWh under optimized conditions.

3.6. Relationship Between NH₃ Synthesis Rate and Energy Efficiency

Figure 15 illustrates the relationship between NH₃ synthesis rate and energy efficiency for different catalyst systems under a total gas flow rate of 1 L/min, varying N₂/H₂ volumetric ratios and discharge peak voltages. The results show that the scatter distributions of all catalyst systems exhibit a clear nonlinear behavior, indicating that the NH₃ synthesis rate does not vary monotonically with energy efficiency. Instead, relatively high NH₃ synthesis rates are observed within a specific energy efficiency range. Among all catalysts, the data points for Ni/BaTiO₃ are mainly located in the upper-right region of the plot, corresponding to simultaneously high NH₃ synthesis rates and high energy efficiency values. The maximum energy efficiency reaches 1.40 g-NH₃/kWh, significantly outperforming the other catalyst systems. This demonstrates that the Ni/BaTiO₃ catalyst not only significantly enhances the NH₃ synthesis rate but also maintains a relatively high level of energy utilization over a broad operating window. For the single-component catalyst systems, BaTiO₃ exhibits better performance compared with Ag, Cu, and γ-Al₂O₃. Its data points are relatively concentrated in the higher NH₃ synthesis rate region, indicating that high-dielectric-constant supports can effectively enhance the discharge process and improve plasma–catalyst interactions. Similar rate–efficiency trade-offs have been reported in DBD plasma-catalytic ammonia synthesis. Rouwenhorst et al. achieved a maximum energy efficiency of 1.23 g-NH₃/kWh over Ru-K/MgO [20], while Andersen et al. reported 1.19 g-NH₃/kWh over Co/MgAl₂O₄ [21]. Compared with these representative systems, the Ni/BaTiO₃ catalyst in this work reaches a higher maximum energy efficiency of 1.40 g-NH₃/kWh while maintaining a high NH₃ synthesis rate, suggesting a more favorable balance between plasma activation and catalytic surface conversion.

3.7. OES Analysis of DBD Plasma with Different Catalysts

To further elucidate the formation characteristics of reactive species in different catalyst systems, OES was employed for diagnostic analysis. OES measurements were conducted at a discharge voltage of 9 kV, a frequency of 6 kHz, and the optimal N₂/H₂ volumetric ratio corresponding to each catalyst system. All measurements were carried out under identical discharge conditions to ensure comparability among the different catalysts. The OES analysis was primarily intended for qualitative comparison of reactive species generation and to assess the relative influence of each catalyst on key species such as NH, N₂⁺, and Hα. As shown in Figure 16a,b, the emission spectra in the ranges of 300–500 nm and 640–680 nm under different catalyst conditions are presented, respectively. Notably, no significant signals were observed in the 500–640 nm range, and thus this spectral region is not shown in the figure. It can be observed that characteristic emission lines corresponding to NH, N₂, N₂⁺, and H_α are detected in all systems, indicating the presence of multiple reactive species associated with ammonia synthesis during the plasma discharge process. In particular, the emission line at 336.7 nm is attributed to the NH (A³Π→X³Σ⁻) electronic transition [39], This reflects the formation of NH_X intermediate species. The second positive system of N₂ (C³Π_u→B³Πg), distributed in the 300–400 nm region, corresponds to the main emission band of electronically excited nitrogen molecules. In addition, a characteristic peak of the first negative system of N₂⁺ (B²Σ_u⁺→X²Π_g⁺) is observed at around 391.4 nm, indicating that during the discharge process, N₂ molecules undergo not only electronic excitation but also partial ionization [13]. In addition, a distinct H_α emission line is observed at 656.3 nm, corresponding to the radiative transition of hydrogen atoms from the n = 3 to n = 2 energy level, indicating the presence of a considerable amount of reactive hydrogen atoms in the discharge region.

To further compare the effect of different catalysts on the generation of plasma reactive species, the intensities of characteristic emission peaks in each system were normalized, as shown in Figure 16c. Among all systems, the Ni/BaTiO₃ catalyst exhibits the highest intensities of characteristic species, including NH, N₂⁺, and H_α emission lines. Based on the experimental results presented above, the increase in the concentration of reactive species in the gas phase contributes to the enhanced efficiency of NH₃ synthesis in the DBD system.

Beyond the qualitative identification of reactive species, OES was further used to estimate the gas temperature in the discharge region. Since DBD plasma is a non-thermal equilibrium system, where the electron temperature, gas temperature, and catalyst bed temperature are not completely identical, OES was employed to analyze the emission spectra in the discharge region and to characterize the plasma excitation state and gas temperature features. Previous studies have used the N₂ second positive system, N₂ (C³Π_u→B³Πg), for plasma temperature diagnostics by Boltzmann analysis or by fitting the experimental N₂ (C→B) emission bands with simulated spectra [40,41]. In this work, the emission spectra in the range of 340–360 nm were fitted using the Boltzmann method to estimate the gas temperature under different packing systems. Since the rotational temperature of N₂ can be approximately regarded as the gas temperature in atmospheric-pressure DBD plasma, $T_g$ was obtained by achieving the best match between the simulated and experimental spectra, as shown in Figure 17.

The fitting results show that all systems exhibit high fitting accuracy, with $R^2$ values of approximately 99.4–99.6%, indicating the reliability of this method. The gas temperature follows the order of quartz wool < BaTiO₃ < Co/BaTiO₃ < Ni/BaTiO₃, with the Ni/BaTiO₃ system reaching the highest gas temperature of approximately 560 K.

3.8. Mechanistic Analysis of DBD Plasma—Catalyst Coupling

The above OES results provide direct evidence for the plasma activation of N₂/H₂. In the DBD plasma-catalytic ammonia synthesis process in the N₂/H₂ system, energetic electrons first collide with gas-phase N₂ and H₂ molecules, causing N₂ to undergo excitation, ionization, and partial dissociation, while also promoting the dissociation of H₂ to generate reactive hydrogen species. The main processes include N₂ + e⁻⟶ N₂* + e⁻, N₂ + e⁻⟶ N₂⁺ + 2e⁻, N₂ + e⁻⟶ 2N* + e⁻, and H₂ + e⁻⟶ 2H* + e⁻ [12]. Subsequently, these plasma-generated reactive species may participate in NH₃ formation through gas-phase reactions and surface-assisted pathways, including adsorption, stepwise hydrogenation, and desorption on the catalyst surface.

For the plasma-only system, NH₃ formation mainly depends on gas-phase electron-impact activation of N₂/H₂ and subsequent reactions among plasma-generated N- and H-containing species. However, due to the absence of additional catalytic surfaces and dielectric packing, the generation and surface conversion of reactive species are relatively limited. After quartz wool is introduced, its porous structure can increase the number of local micro-discharge channels and provide additional contact interfaces, thereby slightly improving plasma activation. Nevertheless, quartz wool is chemically inert and has limited ability to promote surface hydrogenation reactions, so its enhancement of NH₃ synthesis remains limited. For the conductive metal catalysts, Ag and Cu can modify the local electric-field distribution to some extent owing to their high electrical conductivity. However, the experimental results indicate that single conductive metals cannot effectively promote NH₃ synthesis. This may be attributed to the relatively weak dielectric polarization ability of Ag and Cu, which limits their capacity to enhance the local electric field compared with high-dielectric-constant materials. Meanwhile, Ag and Cu have relatively weak binding ability toward N₂ species, making it difficult to provide efficient surface reaction pathways for the adsorption, stepwise hydrogenation, and NH₃ desorption of nitrogen-containing intermediates [42]. Therefore, metallic conductivity alone is insufficient to achieve efficient plasma-catalytic ammonia synthesis.

For γ-Al₂O₃, as a conventional oxide support widely used in heterogeneous catalysis, its surface can provide certain adsorption sites for plasma-generated reactive species. However, compared with BaTiO₃, γ-Al₂O₃ has a lower dielectric constant and weaker dielectric polarization ability, resulting in a limited effect on local electric-field enhancement and micro-discharge regulation. Therefore, the NH₃ synthesis performance of γ-Al₂O₃ is lower than that of BaTiO₃, indicating that dielectric properties play an important role in the plasma-catalytic ammonia synthesis system.

In contrast, BaTiO₃, as a high-dielectric-constant material, can enhance the local electric field, promote micro-discharge formation, and increase the energy injected into the plasma. This facilitates electron-impact reactions and promotes the generation of chemically reactive species, thereby enhancing N₂/H₂ activation [22]. Therefore, BaTiO₃ exhibits better NH₃ synthesis performance than Ag, Cu, and γ-Al₂O₃, demonstrating the importance of dielectric enhancement in DBD plasma-catalytic ammonia synthesis.

When Co or Ni active species are further loaded onto BaTiO₃, the catalyst can combine the dielectric enhancement effect of BaTiO₃ with the surface reaction ability of transition-metal active sites. The plasma-generated N* and H* species may be adsorbed on the surfaces of Co/BaTiO₃ or Ni/BaTiO₃ and converted to NH₃ through a stepwise hydrogenation pathway, namely N* + H* ⟶ NH*, NH* + H* ⟶ NH₂*, and NH₂* + H* ⟶ NH₃*, followed by the desorption of NH₃* from the catalyst surface to form gaseous NH₃.

In addition, Barboun et al. reported that the NH₃ desorption peak temperature follows the order Ni < Co, indicating that, after reactive nitrogen species have been generated by plasma, Ni is more favorable for the stepwise hydrogenation of surface nitrogen species and NH₃ desorption. In contrast, Co has a stronger affinity for nitrogen species, which may favor nitrogen adsorption or retention; however, excessively strong Co–N interactions may also make subsequent NH_X hydrogenation and NH₃ release relatively difficult [43]. Therefore, Co/BaTiO₃ may be more inclined to capture and stabilize activated nitrogen species, whereas Ni/BaTiO₃ is more favorable for the stepwise hydrogenation of NH_x intermediates and NH₃ release.

Overall, the different catalyst systems investigated in this work provide a comparative basis for understanding how conductive metals, conventional oxide supports, high-dielectric materials, and supported transition-metal catalysts influence plasma activation and surface conversion.

4. Conclusions

This work investigated the role of different catalysts in plasma-assisted ammonia synthesis using a coaxial DBD reactor system. The results demonstrate that the catalyst type has a significant impact on NH₃ synthesis performance, with the activity order of Ni/BaTiO₃ > Co/BaTiO₃ > BaTiO₃ > Ag > γ-Al₂O₃ > Cu. Under the optimized operating conditions, the Ni/BaTiO₃ system exhibited the best overall performance. At a N₂:H₂ ratio of 1:1 and a total gas flow rate of 2.5 L/min, the NH₃ synthesis rate reached 259.48 μmol/min, with an energy efficiency of 1.40 g-NH₃/kWh. Discharge characteristics and OES diagnostics indicate that, compared with the plasma-only system, the incorporation of high-dielectric-constant BaTiO₃ led to a more homogeneous discharge, while metal loading increased the number of micro-discharge events. In particular, the Ni/BaTiO₃ system exhibited higher intensities of NH, N₂⁺, and Hα emission lines, which facilitated the formation and subsequent conversion of NH_X intermediates. Combined with XRD, SEM, and EDS characterizations, it was confirmed that the metallic active species were well dispersed on the BaTiO₃ support, thereby enhancing the interaction between gas-phase reactive species and the catalyst surface. Overall, this study achieved relatively high NH₃ synthesis rates and energy efficiency within the investigated systems, demonstrating the superior performance of the Ni/BaTiO₃ catalyst in plasma-assisted ammonia synthesis and providing experimental insights for the further optimization of such systems. For future large-scale applications, the NH₃ production capacity could be further improved by increasing the effective discharge volume, enhancing the gas processing capacity, and adopting a modular parallel configuration of multiple DBD reactors. This numbering-up strategy may help maintain discharge uniformity and operational stability during scale-up.