Low-temperature chemistry in plasma-driven ammonia oxidative pyrolysis

Abstract

Ammonia is gaining increasing attention as a green alternative fuel for achieving large-scale carbon emission reduction. Despite its potential technical prospects, the harsh ignition conditions and slow flame propagation speed of ammonia pose significant challenges to its application in engines. Non-equilibrium plasma has been identified as a promising method, but current research on plasma-enhanced ammonia combustion is limited and primarily focuses on ignition characteristics revealed by kinetic models. In this study, low-temperature and low-pressure chemistry in plasma-assisted ammonia oxidative pyrolysis is investigated by integrated studies of steady-state GC measurements and mathematical simulation. The detailed kinetic mechanism of NH3 decomposition in plasma-driven Ar/NH3 and Ar/NH3/O2 mixtures has been developed. The numerical model has good agreements with the experimental measurements in NH3/O2 consumption and N2/H2 generation, which demonstrates the rationality of modelling. Based on the modelling results, species density profiles, path flux and sensitivity analysis for the key plasma produced species such as NH2, NH, H2, OH, H, O, O(1 D), O2(a1 Dg), O2(b1 Sg þ), Ar*, H, Arþ, NH3 þ, O2in the discharge and afterglow are analyzed in detail to illustrate the effectiveness of the active species on NH3 excitation and decomposition at low temperature and relatively higher E/N values. The results revealed that NH2, NH, H as well as H2 are primarily generated through the electron collision reactions e þ NH3 / e þ NH2 þ H, e þ NH3 / e þ NH þ H2 and the excited-argon collision reaction Ar* þ NH3 þ H / Ar þ NH2 þ 2H, which will then react with highly reactive oxidative species such as O2*, O*, O, OH, and O2 to produce stable products of NOx and H2O. NH3 / NH is found a specific pathway for NH3 consumption with plasma assistance, which further highlights the enhanced kinetic effects.

Headnote

Abstract

Ammonia is gaining increasing attention as a green alternative fuel for achieving large-scale carbon emission reduction. Despite its potential technical prospects, the harsh ignition conditions and slow flame propagation speed of ammonia pose significant challenges to its application in engines. Non-equilibrium plasma has been identified as a promising method, but current research on plasma-enhanced ammonia combustion is limited and primarily focuses on ignition characteristics revealed by kinetic models. In this study, low-temperature and low-pressure chemistry in plasma-assisted ammonia oxidative pyrolysis is investigated by integrated studies of steady-state GC measurements and mathematical simulation. The detailed kinetic mechanism of NH3 decomposition in plasma-driven Ar/NH3 and Ar/NH3/O2 mixtures has been developed. The numerical model has good agreements with the experimental measurements in NH3/O2 consumption and N2/H2 generation, which demonstrates the rationality of modelling. Based on the modelling results, species density profiles, path flux and sensitivity analysis for the key plasma produced species such as NH2, NH, H2, OH, H, O, O(1 D), O2(a1 Dg), O2(b1 Sg þ), Ar*, H, Arþ, NH3 þ, O2in the discharge and afterglow are analyzed in detail to illustrate the effectiveness of the active species on NH3 excitation and decomposition at low temperature and relatively higher E/N values. The results revealed that NH2, NH, H as well as H2 are primarily generated through the electron collision reactions e þ NH3 / e þ NH2 þ H, e þ NH3 / e þ NH þ H2 and the excited-argon collision reaction Ar* þ NH3 þ H / Ar þ NH2 þ 2H, which will then react with highly reactive oxidative species such as O2*, O*, O, OH, and O2 to produce stable products of NOx and H2O. NH3 / NH is found a specific pathway for NH3 consumption with plasma assistance, which further highlights the enhanced kinetic effects.

Keywords: Non-equilibrium plasma; Ammonia fuel; Oxidative pyrolysis; Pathway flux analysis; Sensitivity analysis

1. Introduction

In response to climate change and global warming, countries around the world have put forward carbon-neutral targets and designed the path to realize carbon neutrality [1]. According to the International Energy Agency [2], achieving net zero emissions requires huge advances in clean production technologies, major changes in clean energy residential behavior as well as energy structure dominated by renewable energy. Therefore, the transition from traditional fossil fuels to low-carbon, clean energy sources has become an important task on a global scale. Ammonia (NH3) is the second largest chemical in the world, with an annual production of over 200 million tons [3–5]. In addition to its traditional use as a fertilizer or an important starting material for the production of commercially important nitrogen compounds and hydrogen, it will serve as a new generation of green fuel for power generation [6–8]. NH3 has a much lower liquefaction pressure than H2 at room temperature and can be an excellent carrier of hydrogen energy [9–11], in addition to the fact that the products of complete combustion of NH3 are N2 and H2O. NH3 is as such recognized as an excellent alternative fuel for engines [12]. Despite aforementioned advantages, the relatively harsh ignition conditions and slow flame propagation of ammonia compared to conventional engine fuels pose a great challenge to the use of ammonia in engines [13,14].

Non-equilibrium plasma has been demonstrated as an effective method for molecular activation. Non-equilibrium plasma, such as corona discharge, nanosecond pulsed discharge, sliding arc discharge, dielectric barrier discharge, glow discharge, microwave discharge and spark discharge [15–18], can produce a large number of active species and heat and then modify the transport processes. New reaction pathways such as O, O(1 D), and O2(a1 Dg) generated by collisions of high-energy electrons/ions with oxygen molecules are introduced into the combustion system to change the low temperature reaction pathway and promote the occurrence of chain reactions, which makes it easy to realize the low temperature activation of fuel molecules [19] and thus accelerate the pyrolysis and oxidation of fuel. Non-equilibrium plasma is so considered to be a promising method for NH3 oxidative pyrolysis, NH3 ignition and combustion enhancement. Other than producing active species and new reaction pathways, plasma activated ammonia decomposition can easily yield a great amount of H2, which will help in speeding up the NH3 flame front propagation and stabilizing the NH3 combustion.

In spite of the huge appeals, few studies have attempted to apply non-equilibrium plasma for NH3 decomposition as well as NH3 ignition and combustion up to now. Alexander Fateev et al. [20] investigated the NH3/Ar dielectric barrier discharge (DBD) plasma chemistry at atmospheric pressure. Emission and absorption spectroscopy measurements, as well as kinetic analysis results, showed that the major products of plasmaassisted NH3 pyrolysis are H2, N2, and N2H4. However, it may provide quantitative rather than qualitative measurements. Congwei Yao et al. [21] have studied the discharge mode and photo-electricity characteristics of Ar/NH3 DBD and discussed the discharge stability under the influence of Penning ionization and attachment reactions. The results need a detailed kinetic map for plasma-assisted NH3 decomposition. The effect of pulse number and frequency on the ignition delay time of NH3/O2/He mixtures at atmospheric pressure and wide temperature ranges (600 K–1500 K) was simulated by using the ZDPLasKin-CHEMKIN zero-dimensional solvers by Galia Faingold et al. [22] Plasma kinetic model for NH3 ignition was initially constructed in this work and some important reaction pathways were revealed. Regretfully, the modelling results have not been verified experimentally. Jinhoon Choe et al. [23] conducted the first experimental study on plasma-assisted ammonia combustion. This work reports that plasma can simultaneously reduce NOx emission and extend the lean blow-off limits of ammonia flames. However, the decomposition process before NH3 ignition and combustion was not reported. Taneja T.S. et al. [24] computationally investigate plasma-assisted combustion of ammonia-air mixtures in constant volume and constant pressure reactors, determining the impact of the reduced electric field (E/N ), equivalence ratios, pressure, pulse frequency, and energy density on ignition delays and NOx emissions. Although abundant modelling results on ignition characteristics were reported in detail, model accuracy was not assessed experimentally. In summary, these studies mentioned above have made pioneering contributions in plasma-assisted NH3 decomposition, ignition and combustion, however, they were mostly limited to qualitative analysis and simulation calculations. As far as we know, there is not enough experimental data so far to verify the theory and the computational model. Plasma-assisted NH3 ignition and combustion is a complex physicochemical process of plasma-coupled ammonia combustion kinetics, dynamics, and chemistry. Ju et al. [25] proposed multiple enhancement paths for non-equilibrium plasma combustion, including thermodynamic enhancement, dynamic enhancement, and transport enhancement. Since the temperature rise of non-equilibrium plasma is low, the enhancement effect of thermal effect on combustion is relatively small, so the dynamic enhancement effect generated by active groups becomes the most important path of combustion enhancement. The strengthening mechanism of non-equilibrium plasma in oxidative pyrolysis and fuel conversion at low temperatures also follows the above path. Being a new research area, kinetic insights into plasma-initiated NH3 oxidative pyrolysis, H2 generation, ignition and combustion have not been fully discovered. Specifically, the kinetic nature in NH3 decomposition as well as NH3 ignition and combustion caused by plasma-introduced new reaction pathways, the complicated synergy of plasma and chemistry and so on have drawn little attention and need to be further revealed.

Accordingly, detailed kinetics in a repetitively-pulsed nanosecond discharge assisted ammonia pyrolysis and oxidation by integration of experimental measurements and mathematical simulation is carried out in this paper. Firstly, experiments are carried out at a pressure of 8000 Pa and a temperature of 350 K using a mixture highly diluted in an Argon gas. The consumption of NH3 and O2 and the production of N2 and H2 are analyzed from the results in a flow plasma reactor with the gas compositions measured by an offline Gas Chromatography (GC-TCD). Secondly, a detailed kinetic model for plasma-assisted NH3 oxidative pyrolysis is developed and calculated based on the zero-dimensional plasma chemistry simulation platform [26] and validated experimentally. Thirdly, path flux and sensitivity analysis are carried out to demonstrate the kinetic effects of key species in the formation of important intermediates, and to demonstrate the strong enhancement of plasma on NH3 oxidative pyrolysis.

2. Experimental setup and simulation method

2.1. Experimental setup As described in Fig. 1, a DBD laminar flow reactor with 15 mm I.D. and 19 mm O.D. is adopted to isolate the kinetic mechanism of plasma-assisted NH3 oxidative pyrolysis from the transport effects. The hollow stainless-steel tube with 2 mm O.D. is a high-voltage electrode and supported by seals at both ends of the quartz tube and placed in the center of the reactor. The outer wall is wrapped with a 30 mm wide copper sheet as a low voltage electrode and grounded.

The flow rates of NH3, O2, and Ar are controlled by a mass flow meter (Seven Star Huachuang Co.) and mixed uniformly into the DBD reactor. Ar is used as a dilution gas to prevent the generation of filamentary discharge and ensure the uniformity of the discharge. During the experiment, the air pressure in the reactor is 8 kPa, and a vacuum pump (rotary vane vacuum pump 2XZ-4) is used to control the air pressure in the system. A K-type thermocouple is placed at the end of the reaction zone to measure the gas temperature in the reaction zone. Under natural convection conditions, the temperature in the reaction zone is measured at 350 K. The gas mixtures consist of 20% NH3/80% Ar and 20% NH3/15% O2/ 65% Ar with a total flow rate of 100 mL min 1 . The nanosecond pulse power supply (Xi'an Smart Maple Electronic Technology Co., Ltd, HVP-15) has a maximum discharge frequency of 100 kHz, a pulse width of 1 ms and a peak voltage of 20 kV. The experiments are performed with a voltage probe (RIGOLRP1018H, 1000:1, whose error range was 0–3.0%) and a current probe (CP8030B, 30A/50 MHz, whose error range is within 0–1.0%) to monitor the voltage and current signals during the discharge process in real time and display the output through a digital oscilloscope (RIGOLDS1104Z). The concentration of the main products H2, N2, and the reactants NH3, and O2 are measured quantitatively at the reactor outlet using a gas chromatography (GC9860). The spectral data are collected by using a spectrometer (SP2500i) through a window on the external electrode to detect the plasma-generated species during the discharge.

Fig. 2(a) shows the variation of voltage and current values with time triggered by a single-pulse of a nanosecond pulse power supply. The peak voltage and the pulse width of the single-pulse discharge are set as 5 kV and 100 nm, respectively. Fig. 2(b) suggests the corresponding reduced electric field intensity and the nanosecond pulsed power supply input energy, which will be used as modelling inputs. The voltage U of the approximate square waveform is used to calculate the reduced electric field E/N. The electric field strength E is calculated by E ¼ U/(L þ l/e), where L is the gap distance between the inner electrode and the dielectric layer (6.5 mm in this work), l is the thickness of the dielectric layer (2 mm in this work), and e is the relative permittivity of quartz. Note that the reduced electric field E/N used as the modeling input is determined by equating the area under the actual E/N waveform curve with the area under the approximate square waveform, as shown in Fig. 2(b). Further observation in Fig. 2(b) suggests that the deposition energy per pulse in this study is on the order of 10 2 mJ cm 3 , which will be treated as a controlled parameter to adjust the predicts to the measurements.

In order to illustrate the ammonia molecule activation and formation of intermediate species in a 20% NH3/15% O2/65% Ar plasma, optical emission spectroscopy (OES) is used. It can be seen in Fig. 2(c) that high-intensity Ar spectral lines appear in the wavelength ranges of 650–850 nm due to a large amount of argon atoms' transition in the discharge system [27]. Some N2 spectral lines of the first negative band system such as N2 þ(B2 Su þ)/N2 þ(X2 Sg þ) and the second positive system such as N2*(C3 Pu)/N2*(B3 Pg) appear in the wavelength range of 315.9–434.3 nm [27]. The emission spectral lines of NH*(A3 P/X3 S ), NH2*(A~3 A1/X3 B1), Ha(3d2 D/2p2 P0 ) radicals are observed at 336.1 nm, 661.5 nm and 656.3 nm, respectively [28–30]. The above observations suggest that plasma-generated electrons have sufficient energy to break the N–H bond of NH3 molecules to generate highly reactive NH, NH2, and H radicals, thus significantly promoting ammonia decomposition. Furthermore, OH spectral line is observed at 308.0 nm [27], H2O spectral line at 960 nm, and relatively weak NO spectral lines between 220 and 300 nm [31]. All of these indicate that e-NH3 collision-generated NH, NH2, and H radicals can react with highly reactive oxidative radicals such as O2*, O*, O, OH, and O2 to produce stable products of NO and H2O.

2.2. Kinetic model and numerical method

Detailed plasma chemistry mechanisms governing the oxidative pyrolysis processes in Ar/NH3 and Ar/NH3/O2 mixtures are constructed by including a set of electron impact reactions, dissociative recombination reactions, reactions involving vibrationally- and electronically-excited species and ions, and some important three-body recombination reactions, which is available for below 1 atm and 1500 K. A set of elementary reactions of ground states in Ar/NH3 and Ar/NH3/ O2 mixtures are adopted from Glarborg et al. [32]. The transport coefficients for electrons and the rate constants for electron collision reactions are based on collision cross sections by solving the Boltzmann equation via BOLSIGþ [33]. Collision cross sections for NH3, O, and H2O are taken from the Morgan database in the Lxcat database [34], collision cross sections for O2 from the Morgan [34], Phelps [35], and TRINITI [36] databases in the Lxcat database. The collision cross sections for Ar are taken from the Phelps [35] database. For the vibrationally excited molecules, the processes of vibrational-translational (VT) relaxation, vibrational–vibrational (VV) energy transfer, and chemical reactions stimulated by vibrational states are considered. The kinetic model includes 72 species and 998 elementary reactions, which are provided in the supplementary material.

The simulation uses a zero-dimensional program that couples the plasma kinetics solver ZDPLasKin [37] with the chemical reaction kinetics solver CHEMKIN [38] to solve the excitation, dissociation, and ionization reactions, excited state quenching reactions, charge exchange reactions, electron-ion recombination reactions, ion adsorption reactions, and oxidative pyrolysis reactions of NH3 under nanosecond pulsed discharges. It is noted that the charge separation and sheath formation near the electrodes is not considered. The Poisson equation for the electric field is therefore not solved during the kinetic modeling. The deposition energy is considered an adjustable parameter to fit the NH3 and O2 consumption in the experiments due to the strong sensitivity of the e-NH3 collision reaction rates to the electron density and electron energy distribution. More calculation details can be found in references [16,26,39,40].

3. Results and discussion

3.1. Model validation

Fig. 3 shows the comparison of the experimental measurements and the simulation predicts of the products and reactants concentrations in plasma-assisted ammonia pyrolysis and oxidation systems as a function of applied voltage. As shown in the figures, the concentrations of H2 and N2 in the pyrolysis and oxidation products gradually increase and at the same time the concentrations of the reactants NH3 and O2 decrease with the discharge voltage increasing, which is due to the increased dissociation and ionization in the Ar/NH3 and Ar/NH3/O2 mixtures with the increase of the reduced electric fields. From Fig. 3(a) and (c), the production of H2 drops one order of magnitude and the production of N2 drops a little while O2 is added. This is because the addition of O2 makes the electron energy deposited in the NH3 dissociation and ionization channels decreased. This is also because H2, N2, and the plasma-generated H and N radicals react with the O-containing species to generate H2O and NO. Note that compared to H2, N2 is relatively more difficult to react with O-containing species due to the stable N^N, which explains the decreased H2 yield under the condition of O2 adding.

Model validation is conducted by comparing the measured as well as the model predicted results. In the applied voltage range of 4–8 kV, the experimental measurements and simulation results of the NH3, O2 consumption and N2, H2 generation have a good agreement, with an error of less than 10%. The above agreements indicate that the kinetic model is well modeled in this work.

3.2. Electron energy loss distribution

The reduced electric field (E/N) as well as the mixture composition are important influencing factors for the distribution of electron energy in different excitation channels of electron attachment, vibrational excitation, electron excitation, dissociation, and ionization. Fig. 4(a) illustrates the fractions of the electron energy distributed at different molecular degrees of freedom in a 0.20 NH3/0.80 Ar mixture at different reduced electric fields. When the reduced electric field is relatively low (1–8 Td), the electron energy is mainly transferred to the vibrational excitation of NH3. When the reduced electric field is between 8 Td and 50 Td, the electron energy is mainly transferred to the dissociation of NH3. In this interval, the fraction of energy deposited in the NH3 vibrational excitation channel decreases substantially with increasing the reduced electric field. When the reduced electric field is greater than 50 Td, the deposition of electron energy in the NH3 vibrational excitation and dissociation channels decreases with increasing the reduced electric field, while the energy deposited in the ionization channel and the electron excitation channel of Ar increases with increasing the reduced electric field. Note that within the reduced electric field range of this work (297–594 Td), most of the electron energy is localized in the dissociation and ionization of NH3, electronic excitation of Ar.

Fig. 4(b) shows the fraction of electron energy deposited at different molecular degrees of freedom in a 0.20 NH3/0.15 O2/ 0.65 Ar mixture at different reduced electric fields. When the reduced electric field ranges from 1 Td to 8 Td, the electron energy is primary for the vibrational excitation of NH3, vibrational excitation, and electronic excitation of O2. When the reduced electric field is between 8 Td and 70 Td, the electron energy is mainly transferred to the vibrational excitation and dissociation of NH3 and the dissociation and electronic excitation of O2. In this interval, the fraction of electron energy localized in the vibrational excitation of NH3 and the electronic excitation of O2 decreases significantly with the increase of the reduced electric field. When the reduced electric field is greater than 70 Td, the electron energy is transferred to the vibrational excitation of NH3, electronic excitation of O2, and dissociation of NH3 and O2. Note that within the reduced electric field range of this work (297–594 Td), most of the electron energy is localized in the dissociation and ionization of NH3 and O2, and electronic excitation of Ar, indicating ionic chemistry will play dominant roles in oxidative pyrolysis excited by plasma.

3.3. Time evolution of key species in plasma-assisted NH3 pyrolysis and oxidation

Fig. 5(a) and (b) show the time evolution of the key intermediate species such as N, H, NH, NH2 as well as vibrational states of NH3 in a plasma-assisted NH3 pyrolysis system at a temperature of 350 K and pressure of 8000 Pa. The reduced electric field is 336 Td. As can be seen from Fig. 5(a), the mole concentrations of H, N, NH, and NH2 increase gradually with a pulsed electric discharge occurrence, reaching stability at about 10 3 s. Observed from a single pulse, the time evolution of H, N, NH, and NH2 is characterized by a rapid rise in discharge stage, which suggests that the electron collision will dissociate NH3 into H, NH, and NH2 radicals and at the same time convert NH2 and NH into N and H. H is as such the most produced intermediate species, followed by NH2, NH, and N. As observed in the insert of Fig. 5(a), N2H4, N2H2, N2H3, etc. are the main intermediate species produced during NH3/Ar discharge process. From Fig. 5(b), NH3(v2) is the most produced vibrational state, followed by NH3(v4) and NH3(v13), according to the quenching law of the excited states. The changes of electron temperature and electron number density over time in the pyrolysis process are shown in the supplementary materials (see Fig. S1).

Fig. 5(c)–(e) suggest the time evolution of excited states, main products, and intermediate species in the plasma-assisted NH3 oxidation reaction system at a temperature of 350 K and pressure of 8000 Pa. The reduced electric field is of 336 Td. As can be seen in Fig. 5(c), with or without the addition of O2, generation tendencies of electron collision vibrations of NH3 stay consistent. Further-more for the vibrations of O2, O2(v1) is the most produced vibration, followed by O2(v2), O2(v3) and O2(v4). As for the electronically excited O2, when the reaction time reaches 10 3 s, the concentration of O2(a1 Dg) shows 1–2 orders of magnitude higher than O2(b1 Sg þ). It can be seen from Fig. S2 that O2(a1 Dg) is produced by electron collision with O2, while O2(b1 Sg þ) is produced by electron collision with O2 and by reaction with O(1 D), respectively. Combining Fig. 5(d) and (e) can find that the key intermediate species give a rapid rise in the discharge stage and then slowly decay in the afterglow stage. This is because the intermediates such as H, N, NH, NH2 and so on react with O2 and oxygenates to produce a large amount of steady-state species H2O, H2O2, N2O, NO2, etc. Note that the addition of O2 makes 60.7% of H convert to H2O, OH, HO2, OH and NH3, which has a strong contribution to the consumption of H, as observed in Fig. 5(e) and Fig. S3(b).

The observed trends in Fig. 5 point to the abundant production and quick quenching of vibrational species, which means the electron collision excitation as well as the chemical reactions involving excited states are necessarily important to be considered in the modeling.

3.4. Reaction pathway flux analysis

In order to illustrate the formation and consumption of key species in plasma-assisted NH3 pyrolysis and oxidation, reaction pathway flux analysis is carried out. Fig. 6(a)-(c) shows the predicted pathway fluxes for the detailed reaction kinetics of plasma-assisted NH3 pyrolysis at a temperature of 350 K and pressure of 8000 Pa. The reduced electric field is of 336 Td. It can be seen from Fig. 6(a) that NH2 is mainly generated through the electron collisional dissociation reaction e þ NH3 / e þ NH2 þ H and the quenching reaction Ar* þ NH3 þ H / Ar þ NH2 þ 2H. The two reactions contribute 70.7% and 26.8% to the generation of NH2 respectively. On the other hand, NH2 is mainly consumed through the elementary reactions NH2 þ H(þM) ¼ NH3(þM), NH2 þ N2H3]NH3þ N2H2/H2NN, NH2 þ NH]N2H2 þ H, NH2 þ N] N2 þ H þ H as well as the electron attachment reaction NH2 þ H / NH3 þ e, accounting for 16.1%, 17.6%, 11.5%, 5.4% and 12.6% of the total consumption of NH2 respectively. It can be seen from Fig. 6(b) that NH is mainly produced through the electron collision reaction e þ NH3 / e þ NH þ H2 of NH3, which contributes 96.1% to the production of NH. The ionic reaction NH3 þ NHþ/NH2 þ / NH þ NH3 þ/NH4 þ contributes 3.8% to the formation of NH. On the other hand, the elementary reactions of NH þ NH2] N2H2 þ H, NH þ H]N þ H2 and NH þ N2H2/N2H3]NNH/ N2H2 þ NH2 are responsible for 59.0%, 30.42% and 6.9% consumption of NH. As can be seen in Fig. 6(c), H2 is mainly from the electron collision dissociation reaction of e þ NH3 / e þ NH þ H2 of NH3, which contributes 92.4% to the generation of H2. The extraction reaction between H and other species in addition contributes 9.6% to the generation of H2. 95.5% of H2 is consumed through the ionic reaction H2 þ Arþ / H þ ArHþ.

Adding O2 into a plasma-excited Ar/NH3 system can change the generation and consumption of major species significantly. In order to illustrate the altered reaction pathways for NH, NH2, and H2 with O2 addition, the detailed reaction kinetics in a 0.20 NH3/0.15 O2/0.65 Ar mixture excited by an NSD plasma is performed. Similar to the pyrolysis pathway flux in Fig. 6(a), NH2 generation is mainly through the electron collisional dissociation reaction e þ NH3 / e þ NH2 þ H and the quenching reaction Ar* þ NH3 þ H / Ar þ NH2 þ 2H of Ar*, which suggests electron collision kinetics plays dominant roles in NH2 generation. However different from the plasma-excited NH3 pyrolysis pathway, with the addition of O2, the electron collision dissociation reactions of e þ NH2 / e þ NH þ H and e þ NH2 / e þ N þ H2 make increased contributions to the consumption of NH2. This is due to most of the electron energy transfers to the dissociation and ionization of NH3 and O2 from the electronic excitation of Ar while O2 is added (see Fig. 4). It can be seen from Fig. 6(e) that for the NH radical, the main generation and consumption pathways in NH3 oxidation system are similar as those in NH3 i.e., the NH radical primarily comes from the electron collision reaction, again suggesting the dominant roles of plasma kinetics in NH3 oxidative pyrolysis process.

As can be seen in Fig. 3(a) and (c), the amount of H2 generated in plasma-assisted NH3 pyrolysis is one order of magnitude greater than that in plasma-assisted NH3 oxidation. This is because while O2 is added, plasma collision-generated H2 will quickly react with O2 or O2-containing species, resulting in a reduction of the H2 amount. Fig. 6(f) accordingly illustrates the changes in the reaction pathways with or without O2 addition. As shown in Fig. 6(f), H2 is mainly consumed through the ionic reactions H2 þ O2 / OH þ OH and the recombination reaction H2 þ M14 / H þ product, accounting for 60.7% and 34.3% of its consumption respectively.

NOx emission from NH3 fuel combustion is one of the most important characteristics for assessing its application in engines. Reaction pathway flux for NO is therefore predicted to illustrate the generation and consumption of NO species in plasma-assisted NH3 oxidation. It can be seen in Fig. 6(g) that NO mainly comes from the elementary reactions NH2 þ NO2]H2NO þ NO, NH2 þ HNO]NH3 þ NO, and N2H2 þ O]NH2 þ NO, which contributes 30.1%, 21.9% and 15.2% to the production of NO respectively. On the other hand, NO is mainly consumed through the ionic reaction NO þ OH / HNO2 þ e, and the elementary reaction NH2 þ NO]N2 þ H2O, accounting for 44.7% and 40.1% of the consumption. Particularly, as a chain termination reaction, NH2 þ NO]N2 þ H2O contributes significantly to the production of N2 and the consumption of NO, which is generally consistent with the results reported in Refs. [41–43].

Due to the paper length limit, the plot of the reaction network of plasma assisted NH3 pyrolysis and oxidation are presented in the supplementary materials (see Fig. S4 and Fig. S5).

3.5. Sensitivity analysis of major species generation and consumption

Sensitivity analysis reveals the reactions governing the formation and consumption of the key species, which is calculated by doubling the elementary reaction rate constant or the electron collision cross section [26,44,45] Fig. 7(a)-(b) shows the sensitivity analysis for both the electron collision reactions and the elementary reactions for H2, and NH2 species in NSD plasma assisted NH3 pyrolysis at a pressure of 8000 Pa and temperatures of 300 K, 500 K, and 800 K, respectively.

H2 generation in plasma assisted NH3 oxidative pyrolysis is of great significance for the subsequent ignition and combustion of NH3. Sensitivity analysis of H2 is accordingly calculated to determine the key reactions governing the formation and consumption of H2, as shown in Fig. 7(a). It can be seen that the electron collision reactions of e þ NH3 / e þ NH þ H2 and e þ NH3 / e þ NH2 þ H show relatively large positive sensitivities. This again reveals that H2 generation is mainly through the electron collided dissociation of NH3. On the other hand, reactions of N2H2 þ H]NNH þ H2, NH þ H2]NH2 þ H and NH2 þ NH2(þM) ¼ N2H4(þM) present large positive sensitivities, indicating the importance of H, NH, NH2 radicals in H2 generation. Note that the sensitivity coefficients for the electron collision dissociation reactions of e þ NH3 / e þ NH2 þ H and e þ NH3 / e þ NH þ H2, which have the greatest contribution to the generation of H2, gradually decrease with increasing temperature. This demonstrates the effectiveness of plasma-activated NH3 at low temperatures. The sensitivity analysis of the charged particle-related reactions and the neutral specie related reactions for NH2 is shown in Fig. 7(b). Distinguished from H2 and NH species, the electron collision reaction H þ H / H2 þ e and the elementary reaction N2H3/N2H2 þ H]N2H2/NNH þ H2 present high sensitivities to NH2 generation. This indicates that NH2 generation is electron collision dependent. In addition, with the increase in temperature, both the electron collision reactions and the elementary reactions present decreased sensitivity, indicating the effectiveness of plasmaassisted ignition and combustion at low temperatures.

The sensitivity analysis of NH during pyrolysis showed similar results (See the Fig. S6 for detailed analysis).

Fig. 7(c)-7(d) shows the sensitivity analysis for both the electron collision reactions and the elementary reactions for H2, and NO species in NSD plasma-assisted NH3 oxidation at a pressure of 8000 Pa and temperatures of 300 K, 500 K, and 800 K, respectively. Sensitivity analysis of H2 formation and consumption in plasma-assisted NH3 oxidation is calculated. As can be seen in Fig. 7(c), the reactions e þ NH3 / e þ NH2 þ H, e þ NH3 / e þ NH þ H2 present high sensitivity to the generation of H2, indicating that the addition of oxygen has a minor effect on the process of electron collision generation of H2. From the sensitivity analysis of neutral species-related elementary reactions, the reaction of NH3 þ OH]NH2 þ H2O presents the largest sensitivity to H2 generation, indicating that NH2, NH, H, and OH are the key species to enhance the H2 generation while O2 is added.

Fig. 7(d) shows the sensitivity analysis of the electron collision reactions and the elementary reactions to NO generation and consumption. It can be seen in Fig. 7(d) that the reactions with the largest sensitivity to NO generation are NH2 þ NO2]H2NO þ NO and H2NN þ O2]NH2 þ NO. Together with the pathway flux analysis in Fig. 6(g), it is clear that the generation of NH2 plays a dominant role in NO generation. So, the reaction of e þ NH3 / e þ NH2 þ H show positive sensitivity to the production of NO. Detailed sensitivity analyses of NH and NH2 are shown in Fig. S7(a)(b).

In summary, generations of free radicals such as H, NH, NH2, OH, etc., ionic species of H , excited state of O(1 D), and Ar* are responsible for NH3 pyrolysis and oxidation. Furthermore, with the increase in temperature, the electron collision reactions present decreased sensitivity, demonstrating the effectiveness of plasma-assisted NH3 ignition and combustion at low temperatures.

4. Conclusions

DBD flow reactor is used to study the kinetic process of ammonia pyrolysis and oxidation assisted by NSD plasma at a temperature of 350 K and a pressure of 8 kPa. The pyrolysis and oxidation products of ammonia are measured by GC and OES. The steady state and spectral analysis showed the accuracy of the species contained in the mechanism. A detailed mechanism of plasma-assisted pyrolysis and oxidation of ammonia gas is proposed, including a set of electron impact reactions, reactions involving excited species, ionic reactions and chain-branching reactions, available for below 1 atm and 1500 K. By comparing the concentrations of products predicted by the model and measured by experiments, the rationality and validity of the established plasma-assisted ammonia oxidation kinetics model are verified. Through the reaction pathway and the sensitivity analysis, it is found that NH2, NH, and H2 are mainly generated by the electron collision reactions e þ NH3 / e þ NH2 þ H, e þ NH3 / e þ NH þ H2 and the excited argon collision reaction Ar* þ NH3 þ H / Ar þ NH2 þ 2H with or without O2. With the addition of O2, the key plasma-generated species such as H, N, NH, and NH2 will easily react with O2 and Ocontaining species to produce a large number of H2O, H2O2, N2O, NO2, and other steady-state substances, among which, more than half of the H generated oxygenated species. It is also found NO is mainly generated by the elementary reactions, such as NH2 þ NO2]H2NO þ NO and H2NN þ O2]NH2 þ NO. On the other hand, OH and O(1 D) play important roles in the consumption of NO. The calculated electron energy loss fraction suggests that when the reduced electric field is larger than 300 Td, most of the electron energy is localized in the dissociation and ionization channels of NH3, electronic excitation of Ar. In this study, it is found that excited state chemistry and ionic chemistry participated by Ar*, H , Arþ, NH3 þ, O2 , playing a significant role in the oxidative pyrolysis of NH3. It is furthermore found that with the increase in mixture temperature, the electron collision reactions present decreased sensitivity, demonstrating the effectiveness of plasma-assisted NH3 ignition and combustion at low temperatures.

This study provides quantitative diagnostic data of stable products in plasma activated NH3 dissociation and oxidation at low temperature for the first time. More quantitative in situ diagnostics of intermediate species including NH2, NH, H, O, OH and so on are needed to provide validation targets for kinetic models in the future study.

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to thank the grant support from the National Natural Science Foundation of China (No. 21975018, 22278032).

Sidebar

* Corresponding author.

E-mail address: [email protected] (Q. Chen).

Footnote

Received 16 March 2023; revised 28 April 2023; accepted 29 May 2023

Available online 3 June 2023

References

References

[1] A. Huovila, H. Siikavirta, C. Antu~na Rozado, J. R€okman, P. Tuominen, S. Paiho, Å. Hedman, P. Ylen, J. Clean. Prod. 341 (2022) 130912.

[2] https://www.ren21.net/reports/global-status-report/.

[3] H. Liu, J. Li, F. Du, L. Yang, S. Huang, J. Gao, C. Li, C. Guo, Green Energy Environ. 8 (2023) 1619–1629.

[4] Y. Li, H. Wang, B. Chang, Y. Guo, Z. Li, S.H. Talib, Z. Lu, J. Wang, Green Energy Environ. 8 (2023) 1174–1184.

[5] M.I. Ahmed, D.B. Hibbert, C. Zhao, Green Energy Environ. 8 (2023) 1567–1595.

[6] H. Kobayashi, A. Hayakawa, K.D.K.A. Somarathne, E.C. Okafor, Proc. Combust. Inst. 37 (2019) 109–133.

[7] Z. Wan, Y. Tao, J. Shao, Y. Zhang, H. You, Energy Convers. Manag. 228 (2021) 113729.

[8] C.W. Gross, S.-C. Kong, Fuel 103 (2013) 1069–1079.

[9] D. Chen, C. Han, Q. Sun, J. Ding, Q. Huang, T.-T. Li, Y. Hu, J. Qian, S. Huang, Green Energy Environ. 8 (2023) 258–266.

[10] G. Yang, S. Guan, S. Mehdi, Y. Fan, B. Liu, B. Li, Green Energy Environ. 6 (2021) 236–243.

[11] Y. Qin, Y. Chen, X. Zeng, Y. Liu, X. Lin, W. Zhang, X. Qiu, Green Energy Environ. 8 (2023) 1728–1736.

[12] H. Xing, C. Stuart, S. Spence, H. Chen, J. Clean. Prod. 297 (2021) 126651.

[13] O. Kurata, N. Iki, T. Matsunuma, T. Inoue, T. Tsujimura, H. Furutani, H. Kobayashi, A. Hayakawa, Proc. Combust. Inst. 36 (2017) 3351–3359.

[14] E.C. Okafor, Y. Naito, S. Colson, A. Ichikawa, T. Kudo, A. Hayakawa, H. Kobayashi, Combust. Flame 187 (2018) 185–198.

[15] Y.A. Lebedev, Plasma Sources Sci. Technol. 24 (2015) 053001.

[16] X. Tu, J.C. Whitehead, Int. J. Hydrogen Energy 39 (2014) 9658–9669.

[17] G. Sun, W. Liu, C. Li, R. Zhang, J. Phys.: Conf. Ser. 441 (2013) 012023.

[18] V.P. Demkin, S.V. Melnichuk, Russ. Phys. J. 60 (2017) 339–345.

[19] X. Mao, Q. Chen, A.C. Rousso, T.Y. Chen, Y. Ju, Combust. Flame 206 (2019) 522–535.

[20] A. Fateev, F. Leipold, Y. Kusano, B. Stenum, E. Tsakadze, H. Bindslev, Plasma Process. Polym. 2 (2005) 193–200.

[21] C. Yao, S. Chen, S. Wang, Z. Chang, A. Sun, H. Mu, G.-J. Zhang, J. Phys. D Appl. Phys. 51 (2018) 225201.

[22] G. Faingold, J.K. Lefkowitz, Proc. Combust. Inst. 38 (2021) 6661–6669.

[23] J. Choe, W. Sun, T. Ombrello, C. Carter, Combust. Flame 228 (2021) 430–432.

[24] T.S. Taneja, P.N. Johnson, S. Yang, Combust. Flame 245 (2022) 112327.

[25] Y. Ju, W. Sun, Prog. Energy Combust. Sci. 48 (2015) 21–83.

[26] X.Mao,A.Rousso,Q.Chen,Y. Ju,Proc.Combust. Inst. 37 (2019) 5545–5552.

[27] A.A. Panarin, V.S. Skakun, E.A. Sosnin, V.F. Tarasenko, Opt. Spectrosc. 122 (2017) 168–174.

[28] D. Zhang, Q. Gao, B. Li, J. Liu, Z. Li, Appl. Opt. 58 (2019) 1210–1214.

[29] A. Hayakawa, T. Goto, R. Mimoto, T. Kudo, H. Kobayashi, Mech. Eng. J. 2 (2015) 14-00402.

[30] X.Z. Xinluo Zhao, T.O. Tsugio Okazaki, A.K. Atsuo Kasuya, H.S. Hiroshi Shimoyama, Y.A. Yoshinori Ando, Jpn. J. Appl. Phys. 38 (1999) 6014.

[31] J.E. Nicholas, C.A. Amodio, Plasma Chem. Plasma Process. 2 (1982) 331–339.

[32] P. Glarborg, J.A. Miller, B. Ruscic, S.J. Klippenstein, Prog. Energy Combust. Sci. 67 (2018) 31–68.

[33] G.J.M. Hagelaar, L.C. Pitchford, Plasma Sources Sci. Technol. 14 (2005) 722–733.

[34] Morgan database.www.lxcat.net/Morgan.

[35] Phelps database.www.lxcat.net/Phelps.

[36] TRINITI database.www.lxcat.net/TRINITI.

[37] http://www.zdplaskin.laplace.univ-tlse.fr/.

[38] A.E. Lutz, R.J. Kee, J.A. Miller, SENKIN: A Fortran program for predicting homogeneous gas phase chemical kinetics with sensitivity analysis, https://www.osti.gov/biblio/5371815, 1987.

[39] X. Mao, Q. Chen, AIAA J. 56 (2018) 4312–4320.

[40] X. Mao, G. Li, Q. Chen, Y. Zhao, Chin. J. Chem. Eng. 24 (2016) 17191727.

[41] T. Mendiara, P. Glarborg, Combust. Flame 156 (2009) 1937–1949.

[42] Z. Tian, Y. Li, L. Zhang, P. Glarborg, F. Qi, Combust. Flame 156 (2009) 1413–1426.

[43] H. Nakamura, S. Hasegawa, T. Tezuka, Combust. Flame 185 (2017) 1627.

[44] J. Sun, Q. Chen, B. Zhao, C. Guo, J. Liu, M. Zhang, D. Li, J. Phys. D Appl. Phys. 55 (2022) 135203.

[45] J. Sun, Q. Chen, X. Yang, B.E. Koel, J. Phys. D Appl. Phys. 53 (2020) 064001.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.gee.2023.05.010.