Skip to main content
Molecules logoLink to Molecules
. 2026 Oct 28;31(21):3893. doi: 10.3390/molecules31213893

Molecular Dynamics Study on Combustion Characteristics and Soot Formation Reactions of C4 Isomers Blended with NH3

Yang Wang 1,2, Jinhao Chen 2, Junjie Lu 2, Zhenhui Zhang 2, Yiju Zhao 2, Lumin Chen 3,*, Mingyan Gu 2,*
Editor: David Lee Phillips
PMCID: PMC13765477  PMID: 43029209

Abstract

To achieve clean combustion of C4 hydrocarbon fuels, five typical C4 structural isomers were selected and reactive molecular-dynamics simulations with the ReaxFF force field were performed to compare combustion characteristics under neat-fuel and NH3-blended conditions. The results reveal strong structural dependence in C4 isomer combustion: terminal double bonds and tertiary C-H bonds govern ignition reactivity in the early stage. Carbon-skeleton cleavage controls overall reaction efficiency in the intermediate stage. Olefinic intermediates tend to persist in the late combustion phase, with distinct soot-precursor distributions across fuels. Ammonia addition markedly shortens oxygen-consumption duration for all systems, suppresses CO2 formation, and increases H2 and H2O yields. For soot formation, neat olefins produce far more six-membered rings than alkanes. NH3 blending raises the system-level C/H atomic ratio, and nitrogen atoms are incorporated into soot skeletons via nitrogen-containing functional groups.

Keywords: C4 isomers, NH3-blended combustion, ReaxFF MD, soot formation

1. Introduction

Nowadays, environmental challenges including global warming, frequent extreme meteorological events and aggravated atmospheric pollution have become critical bottlenecks restricting the advancement of human society [1,2,3]. Combustion, the core technology for energy conversion, dominates energy supply and represents the primary contributor to greenhouse gas and pollutant emissions [4,5]. Amid the transition of global energy mix, ammonia (NH3) is recognized as a highly promising alternative fuel due to its carbon-free property and mature storage and transportation technologies [6,7]. Additionally, NH3 possesses a volumetric energy density comparable to conventional fossil fuels along with a superior octane number, which enables its practical engineering deployment [8]. However, pure NH3 combustion is plagued by inherent limitations such as high ignition energy, low burning velocity and poor flame stability, which restricts its standalone application as fuel [9]. In addition, the combustion of pure NH3 tends to produce nitrogen-containing pollutants such as NO, NO2 and N2O. Under fuel-rich conditions, highly toxic HCN can also be formed, causing irreversible damage to the human body [10,11].

C4 hydrocarbons serve as major byproducts generated during catalytic cracking and steam cracking processes in the petroleum refining industry [12,13], and they also constitute the core components of liquefied petroleum gas. At present, C4 resources are primarily utilized as low-grade fuels with a low chemical conversion rate. This approach brings about a series of drawbacks including low added value of resources, inefficient combustion utilization, and great difficulties in the control of pollutant emissions [14,15]. By contrast, the co-combustion of C4 isomers with NH3 delivers dual benefits: high-value utilization of C4 resources and large-scale practical application of NH3 fuels [16]. Such blended fuels can be widely deployed in various engineering facilities such as industrial furnaces, internal combustion engines, and gas turbines, thereby possessing prominent prospects for engineering [5,17].

In recent years, investigations into the combustion of C4 isomers have covered multi-scale analyses ranging from macroscopic flame behaviors to microscopic radical reaction kinetics. Chen et al. [18] adopted an opposed-flow diffusion flame configuration and observed the transition of flames fueled by 1-butene, isobutene, n-butane and isobutane from a soot-free state to a heavily sooting state by gradually increasing the fuel mole fraction in the fuel stream. Experimental results indicated that the critical fuel concentrations required to reach the macroscopic soot transition threshold were 9.0% for 1-butene, 9.5% for isobutene, 19% for isobutane and 25% for n-butane. This concentration sequence demonstrates that C4 olefins exhibit a far greater tendency for soot formation than their corresponding alkanes. Moreover, the branching structure exerts a non-monotonic effect on the soot propensity of alkanes and olefins. For alkanes, branching facilitates soot generation, whereas linear olefin structures possess higher soot-forming tendencies than branched counterparts. Zhang et al. [19] explored the pyrolysis processes of three butene isomers at low pressure within a temperature range of 900~1900 K by employing synchrotron vacuum ultraviolet photoionization mass spectrometry. Reaction flux analysis clarified the dominant decomposition pathways of the three isomers. It was also revealed that the linear double-bond position and branched structure of C4 olefin isomers substantially alter their initial decomposition routes and product distribution profiles.

Co-combustion of hydrocarbon fuels blended with NH3 can remarkably improve the combustion performance and pollutant formation characteristics of pure NH3. Guo et al. [20] measured and compared the laminar burning velocities of NH3 mixtures doped with n-heptane, gasoline and RP-3 kerosene. The results demonstrated that the n-heptane/NH3 blended fuel possessed the highest laminar burning velocity. This phenomenon was attributed to the straight short-chain structure of n-heptane, which facilitates dehydrogenation reactions to generate small hydrocarbon radicals that accelerate combustion reactions. Gain et al. [21] conducted experimental investigations on fuel-rich jet flames fueled by C2H6/NH3 mixtures. Their observations revealed that increasing the NH3 blending ratio systematically reduced the spectral emission intensities of CH, OH and C2, which are characteristic hydrocarbon intermediates. In contrast, the emission intensity of CN, nitrogen-containing intermediates, increased significantly. Further mechanistic analysis indicated that NH3 addition suppresses the propagation of ethane oxidation chains while promoting the formation of HCN and CN species, thereby lowering the yield of soot precursors. Chen et al. [22] explored the gas–solid coupling mechanism during the co-pyrolysis of C4 isomers and NH3 in a tubular furnace at 1473 K, and uncovered the dual chemical effects of NH3 on soot inhibition. From the perspective of gas-phase chemistry, both experimental measurements and numerical simulations confirmed that NH3 addition reduces the concentrations of critical soot precursors including C2H2, C2H4, and C4H2 in pyrolysis exhaust gas. Meanwhile, the concentration of benzene rises because the reaction pathways for benzene to convert into large polycyclic aromatic hydrocarbons (PAHs) are blocked. This can be explained by the fact that nitrogen-containing species decomposed from NH3 readily react with C2-C4 light hydrocarbons to form nitrogen–hydrogen compounds, which consume carbon atoms available for soot particle growth. In summary, the combustion behavior of neat C4 isomers and C4 isomer/NH3 blends is governed not only by C4 molecular structures and NH3 blending ratios, but also by microscale chemical interactions between hydrocarbons and NH3. Consequently, the soot formation mechanism under such blended combustion conditions remains highly complex and insufficiently understood, which necessitates further in-depth research.

As the most widely investigated binary fuel system in the field of NH3 co-combustion, NH3-CH4 serves as a research benchmark for understanding the fuel interaction of NH3-CH4 mixtures. Rocha et al. [23] studied premixed laminar flames of NH3-CH4–air mixtures under various pressures, φ and NH3 blending ratios. They identified a premixed-diffusion dual-flame structure within the flame. Both elevated pressure and increased NH3 fraction in the fuel suppress NO formation. Meanwhile, they proposed a new calculation method for laminar flame speed accounting for flow inhomogeneity and flame stretch. Jin et al. [24] adopted the Chemkin/Premix code combined with the Okafor reaction mechanism to investigate the combustion characteristics and reaction kinetics of high-pressure premixed NH3-CH4–air laminar flames. Their results revealed that increasing the initial pressure reduces the laminar burning velocity while raising the adiabatic flame temperature and net heat release rate. The rise in pressure lowers the peak concentrations of H, NH2, NO, NO2 and HCN radicals and inhibits NO production. Pashchenko et al. [25] carried out CFD modelling of the combustion of hydrogen-rich synthetic fuels in swirling flames and validated the model against experimental data. They systematically analyzed the effects of fuel composition on combustion and emission characteristics of swirling flames, providing references for the application of hydrogen-rich fuels in combustion devices.

It is evident that existing studies on NH3-CH4 co-combustion lay a solid foundation for NH3-C4 combustion research. Nevertheless, CH4 features a simple molecular structure, and the effect of NH3 exhibits relatively limited diversity. In contrast, C4 fuels have a variety of isomeric structures, and NH3 exerts dual effects on the combustion of C4 fuels: it suppresses partial fuel pyrolysis via radical competition, while hydrogen production from NH3 decomposition accelerates the reactions of certain fuel components. Meanwhile, CH4 combustion hardly generates soot, so most relevant investigations focus on NOx emissions. In comparison, C4 fuels produce substantial soot. In view of the above, this work selects five typical C4 isomers, namely n-butane, isobutane, 1-butene, 2-butene and isobutene, as research objects. Reactive molecular dynamics simulations are adopted to thoroughly investigate the combustion characteristics and soot formation mechanisms of neat combustion and NH3-blended combustion at the molecular level, as well as to analyze the evolution of species and radicals during combustion. This research provides theoretical support for the precise optimization of combustion technologies, facilitating the development of combustion systems with low emissions and high efficiency. In view of the above, five typical C4 isomers, namely n-butane, isobutane, 1-butene, 2-butene and isobutene, are selected as the research objects in this work. Reactive molecular dynamics (ReaxFF MD) simulations are adopted to deeply investigate the combustion characteristics and soot formation mechanisms of neat C4 fuels and C4/NH3 blends at the molecular scale. Moreover, the evolution rules of species and radicals throughout combustion are elaborated. This study can provide theoretical support for the precise optimization of combustion technologies, facilitating the development of combustion systems with low pollutant emissions and high energy efficiency.

2. Results and Discussion

2.1. Macroscopic Combustion Characteristics of C4 Isomers

The consumption rates of fuels and oxidants can reflect their combustion reactivity and macroscopic combustion behaviors. As illustrated in Figure 1a, all C4 isomers exhibit a consistent decreasing trend over time, while distinct discrepancies in consumption rates are observed at different reaction stages. The inset of Figure 1a displays the initial consumption rates of the tested isomers. It is evident that butene fuels possess remarkably higher initial consumption rates compared with butane counterparts. During the early reaction stage, 1-butene and isobutene are rapidly consumed via collisions with radicals including OH, H and O, benefiting from the low activation energy of double-bond addition and minor steric hindrance. In contrast, 2-butene suffers from severe steric hindrance induced by methyl groups on both sides of the carbon–carbon double bond. Despite the low energy barrier for addition reactions, its effective collision frequency is greatly suppressed. Among butane fuels, isobutane generates highly stable tert-butyl radicals through C-H bond cleavage with low activation energy for hydrogen abstraction, which yields a markedly faster initial reaction rate than n-butane.

Figure 1.

Figure 1

Evolution of residual molecular masses of C4 isomer fuels and O2 versus reaction time during combustion: (a) fuels, (b) O2.

After 50 ps, the dominant factor governing fuel consumption gradually shifts to the carbon skeleton cleavage and oxidation capacity of C4 isomers under high radical concentrations. At this stage, 1-butene still maintains the highest consumption rate. n-butane undergoes successive β-scission reactions via its linear carbon skeleton, generating abundant alkyl fragments that participate in chain propagation and subsequent oxidation processes. Its average consumption rate ranks second only to that of 1-butene. In comparison, isobutane reduces fuel concentration merely through rapid tert-butyl radical cleavage. Coupled with limited available decomposition pathways at high temperatures, its overall average consumption rate is lower than that of n-butane. As the reaction proceeds to approximately 100 ps, the discrepancy between butanes and butenes is further enlarged. At this moment, the consumption rates of butane fuels are remarkably lower than those of butene fuels. This is attributed to the absence of C=C double bonds in butane skeletons, which prevents the formation of resonance-stabilized structures and renders the carbon backbone more susceptible to oxidative breakdown. On the contrary, butenes generate intermediates with double bonds or allylic characteristics after undergoing addition, hydrogen abstraction and isomerization reactions. Such intermediates possess resonance stabilization effects and are resistant to cleavage under high-temperature environments with abundant radicals. Consequently, the residual molecular quantity of butene fuels remains higher than that of butanes.

Figure 1b depicts the evolution of residual O2 concentration over reaction time during combustion. The initial oxygen consumption rates, sorted in descending order, follow the sequence: 1-butene > isobutane > isobutene ≈ n-butane > 2-butene. Such disparities in reaction rates remain correlated with the molecular structures of fuels as elaborated above. 1-butene contains a terminal carbon–carbon double bond and highly reactive allylic H, which facilitates rapid generation of allylic radicals and their corresponding peroxy intermediates, thereby accelerating O2 consumption. Isobutane efficiently depletes O2 via small alkyl fragments yielded from tert-butyl radical cleavage. Although isobutene simultaneously possesses double bonds and branched carbon skeletons, substantial steric hindrance surrounding the C=C bond reduces the collision frequency between fuel molecules and O2. N-butane primarily consumes oxygen through hydrogen abstraction, β-scission, and peroxidation pathways. 2-butene exhibits the lowest double bond reactivity and steric accessibility among all tested C4 fuels. Moreover, the allylic radicals produced from 2-butene possess relatively high thermodynamic stability, which slows down its reaction kinetics with O2 and results in the minimum initial O2 consumption rate. In summary, the combustion performance of C4 isomers exhibits pronounced structure-dependent characteristics. In the early reaction stage, highly reactive molecular sites dominate ignition behavior and O2 consumption rates. At the intermediate stage, carbon skeleton configurations govern the reaction rates of β-scission pathways and determine the sustained oxidation capacity of the fuel mixture. In the later combustion stage, resonance-stabilized allylic intermediates regulate the oxidation reactivity of residual fuel species.

C4 isomers exhibit dramatic discrepancies in molecular structure, branching degree and double-bond positions. Analysis of the dynamic evolution of combustion intermediates and final products can reveal fuel pyrolysis pathways, oxidation reaction rates and other core combustion characteristics [26,27]. Figure 2 displays the temporal evolution curves of CO, CO2, H2O, and H2 concentrations during the combustion of different C4 isomers. As illustrated in Figure 2a, butene fuels feature faster CO formation rates than butane fuels in the early reaction stage. After 750 ps, the CO generation rates of butane fuels rise sharply and gradually surpass those of butenes. Specifically, n-butane achieves the fastest growth in CO concentration: its CO yield first exceeds that of 2-butene, followed by isobutene and 1-butene in sequence. Isobutane presents a relatively gentle upward trend of CO concentration, yet its CO output outperforms all butene fuels in the late combustion stage. It is evident that butanes generate CO more readily as the reaction proceeds, despite their lower initial CO formation rates compared with butenes. Compared with CO, the generation profile of CO2 serves as a more reliable indicator of fuel oxidation capacity [27,28]. Figure 2b shows that isobutane delivers a rapid CO2 formation rate at the initial stage, which originates from the existence of tertiary C-H bonds in isobutane. The resulting tert-butyl radicals readily release methyl groups during subsequent oxidation and decomposition processes, which are further oxidized to form vital oxidation precursors. By contrast, 2-butene exhibits the slowest early-stage CO2 production. This indicates that its initial carbon conversion pathways fail to achieve complete oxidation, with most carbon skeletons retained in the form of incompletely oxidized intermediates, such as intermediate radicals and unsaturated pyrolysis fragments. As the reaction progresses, the CO2 generation rate of 2-butene rises continuously and eventually exceeds that of all other fuels. At the end of the simulation, the total CO2 yield follows the descending order: 2-butene > isobutane > n-butane ≈ 1-butene > isobutene. This ranking differs distinctly from the CO distribution. 2-butene yields the minimum amount of CO but the maximum amount of CO2, demonstrating its superior conversion efficiency toward fully oxidized products. In contrast, isobutene tends to form stable ketones and branched oxygenated intermediates. These species undergo complete oxidation at low kinetic rates, which accounts for its lowest final CO2 concentration among all tested C4 fuels.

Figure 2.

Figure 2

Temporal evolution curves of CO, CO2, H2O and H2 concentrations during neat combustion of C4 isomers: (a) CO, (b) CO2, (c) H2O, (d) H2.

The evolutionary characteristics of H2O presented in Figure 2c differ distinctly from those of CO and CO2. The molecular population of H2O exhibits an initial rising trend followed by a gradual decline, all simulated systems reach their H2O concentration peak at approximately 1250 ps, after which H2O concentration continuously decreases. This phenomenon indicates that H2O acts not only as a final combustion product but also as an intermediate species consumed in subsequent chemical reactions. From the perspective of molecular structure, each butane molecule contains two additional H compared with butene counterparts. Therefore, butanes convert H into H2O more readily during the early and intermediate reaction stages, yielding higher peak H2O concentrations. Subsequent consumption reactions then lead to a decline in H2O levels. Significant discrepancies in the net consumption of H2O are observed among different butene isomers, which are closely associated with reactive unsaturated oxygenated intermediates generated throughout the reaction. Such species continuously consume H2O formed in the combustion process.

The evolutionary trend of H2 differs markedly from that of H2O. As illustrated in Figure 2d, the concentration of H2 continuously rises over time during the combustion of all five C4 isomers without obvious decline, which indicates that H2 behaves much like a stable accumulated product. At the termination of the simulation, the total yield of H2 follows the descending order: n-butane > isobutane > 1-butene > 2-butene > isobutene. The higher H2 production from butane combustion originates from their greater intrinsic hydrogen storage capacity. Continuous dehydrogenation during butane pyrolysis persistently releases H, which facilitates the accumulation of H2. Among butene isomers, 1-butene generates more H2 than 2-butene and isobutene, demonstrating that the decomposition pathways of 1-butene are more favorable for H release and subsequent H2 formation. In contrast, 2-butene preferentially converts H into oxidized products such as H2O, resulting in a lower degree of H2 accumulation. Isobutene readily forms stable branched intermediates, a large quantity of H is trapped within organic intermediates with prolonged residence time, ultimately yielding the minimum H2 concentration among all tested fuels.

By comprehensively analyzing the evolutionary characteristics of four oxygen-containing and hydrogen-containing final products, namely CO, CO2, H2O and H2, it can be concluded that the combustion progress of C4 isomers is not merely governed by the ignition rate of fuels, but is synergistically regulated by multiple elemental reaction pathways. As a whole, butenes exhibit rapid initial reactions accompanied by prominent partial oxidation in the later stage, whereas butanes display a slightly slower initial reaction rate with more sustained oxidation throughout the subsequent combustion stage.

2.2. Effects of NH3 Addition on the Reactivity of C4 Isomers

Figure 3 illustrates the temporal evolution of fuel and O2 concentrations during neat combustion and NH3-blended combustion of C4 isomers. In this study, the blending ratio of NH3 and C4 molecules is fixed at 1:2, and detailed operating conditions are referred to Part 3. As presented in Figure 3a, neat 1-butene is consumed more rapidly within the initial 20 ps of the reaction. After 20 ps, the residual fuel amounts of the neat and NH3-blended cases gradually converge, and the fuels in both cases are completely depleted at approximately 130 ps. The discrepancy observed in the early reaction stage can be attributed to the competitive consumption of active radicals such as H, O and OH between NH3 and 1-butene. This competition slows chain reactions including double-bond addition, hydrogen abstraction and subsequent β-scission of the terminal alkene [29]. Nevertheless, this competitive effect only lasts for a short duration, once the radical pool becomes sufficiently abundant, the inhibitory impact of NH3 on 1-butene is substantially weakened. The molecular population of O2 for NH3-blended 1-butene combustion is higher than that of the neat case at the early reaction stage, indicating that NH3 also competes for oxidative radicals including O and OH. As the NH3 chain reactions proceed, O2 consumption is significantly accelerated. Ultimately, O2 is nearly exhausted at around 700 ps under the NH3-blended condition, much earlier than 1200 ps for neat combustion. In summary, although NH3 addition suppresses the initial pyrolysis of 1-butene, it introduces extra oxygen consumption pathways and elevates the overall O2 consumption rate.

Figure 3.

Figure 3

Figure 3

Temporal evolution of fuel and O2 concentrations during neat and NH3-blended combustion of C4 isomers: (a) 1-butene, (b) 2-butene, (c) isobutene, (d) n-butane, (e) isobutane.

Unlike 1-butene, NH3 blending remarkably accelerates the consumption of 2-butene. As shown in Figure 3b, 2-butene is completely depleted at approximately 180 ps under the NH3-blended condition, while full fuel consumption is not achieved until 250 ps for neat combustion, demonstrating that NH3 exerts a promoting effect on the pyrolysis of 2-butene. On one hand, NH3 competes for partial radicals as elaborated in the preceding section. On the other hand, 2-butene is categorized as an internal alkene with its C=C double bond located at the middle of the carbon chain, which leads to severe steric hindrance and relatively sluggish initial reactivity [30,31]. Atomic hydrogen generated from the decomposition of blended NH3 participates in reactions at an early stage and compensates for the insufficient initial reactivity of internal alkenes. Since the reactivity-enhancing effect derived from the structural characteristics of 2-butene outweighs the radical competitive effect, the overall consumption rate of 2-butene is elevated after NH3 addition. The initial molecular population of O2 in the NH3-blended 2-butene system is higher than that of the neat case, which indicates that NH3 competes for oxidative radicals in the early reaction stage. As the reaction proceeds, however, the O2 consumption rate of the blended system rises drastically. The O2 consumption rate surges after roughly 180 ps, and O2 is ultimately exhausted at nearly the same time as the neat combustion system. Accordingly, the influence of NH3 on O2 consumption during 2-butene combustion presents obvious stage-dependent characteristics.

The effects of NH3 blending on isobutene also exhibit obvious stage-dependent characteristics. As illustrated in Figure 3c, the fuel concentration of the neat case decreases more rapidly within the initial 30 ps of the reaction, which proves that NH3 still competes with isobutene for active radicals such as H, O and OH in the early stage. After 30 ps, the fuel consumption rate of the NH3-blended system gradually surpasses that of the neat system, revealing that radicals supplemented by decomposed NH3 can promote chain reactions. Eventually, the fuel in both cases is completely depleted at approximately 300 ps, indicating that NH3 blending barely alters the total fuel consumption duration of isobutene. The O2 consumption behavior of isobutene is more sensitive to NH3 addition. The O2 consumption rate is significantly accelerated after NH3 blending, with O2 nearly exhausted at around 1000 ps, whereas O2 is not fully consumed until roughly 1600 ps under neat combustion conditions. This phenomenon demonstrates that the oxidation chain reactions of NH3 are initiated rapidly in the isobutene system. O2 acts as the oxidant for isobutene and simultaneously participates in the oxidation of NH3. Accordingly, two parallel O2 consumption pathways—hydrocarbon oxidation and NH3 oxidation—function synergistically, accelerating the overall O2 consumption rate of the blended system.

As shown in Figure 3d, the fuel concentration curves of neat n-butane combustion and NH3-blended combustion are highly similar. The fuel is fully depleted at approximately 100 ps for the NH3-blended case, while complete consumption occurs at 130 ps under neat conditions, which indicates that NH3 only slightly promotes the pyrolysis of n-butane. NH3 blending imposes two counteracting effects on n-butane pyrolysis. On one hand, NH3 competes for active radicals including H, O, and OH, thereby suppressing the decomposition of n-butane. On the other hand, the decomposition of NH3 supplies atomic hydrogen, which accelerates hydrogen abstraction from C-H bonds and facilitates the progression of chain reactions. These two effects maintain a dynamic equilibrium, resulting in a marginal advance in the total burnout time. After NH3 addition, the O2 consumption rate of the n-butane system rises dramatically. O2 is nearly exhausted at roughly 1000 ps, 1500 ps earlier than the neat combustion condition. Consequently, NH3 blending exerts a negligible influence on n-butane consumption yet imposes a remarkable accelerating effect on O2 depletion.

NH3 blending suppresses the early-stage consumption of isobutane. As illustrated in Figure 3e, the molecular population of isobutane under the NH3-blended condition is initially higher than that of neat combustion before both curves converge, which is correlated with the molecular structure of isobutane. Isobutane contains tertiary C-H bonds with low bond dissociation energies, enabling rapid C-H bond cleavage and abundant radical accumulation at the initial stage of neat combustion. After NH3 addition, NH3 competes for active radicals in the early reaction stage, hindering isobutane pyrolysis and slowing down fuel consumption. As the reaction proceeds, continuous dehydrogenation of NH3 replenishes atomic hydrogen and intensifies chain reactions. Ultimately, full fuel consumption is achieved at approximately 100 ps for both working conditions. Thus, NH3 blending merely inhibits pyrolysis in the early stage and does not prolong the total fuel burnout duration. The variation in O2 consumption curves demonstrates that NH3 blending accelerates the O2 consumption rate in isobutane combustion. On one hand, the intrinsic oxidation reactions of NH3 create an additional O2 consumption pathway. On the other hand, nitrogen-containing radicals generated during the reaction participate in various elementary reactions and further expedite O2 depletion. In summary, NH3 blending restrains the early pyrolytic consumption of isobutane fuel but exerts a promoting effect on O2 consumption.

Overall, the effects of NH3 blending on C4 isomers exhibit pronounced molecular structure sensitivity [32]. For 1-butene and isobutane, the radical competitive effect of NH3 dominates the early reaction stage and suppresses the initial fuel consumption. In the case of 2-butene, atomic hydrogen generated by NH3 decomposition effectively accelerates hydrogen abstraction and pyrolysis of internal alkenes, thereby markedly speeding up fuel depletion. For isobutene and n-butane, NH3 blending primarily alters the stagewise consumption rate while exerting limited influence on the total time required for complete fuel consumption. O2 consumption shows higher sensitivity to NH3 addition. NH3 blending substantially shortens the total oxygen burnout time, which can be attributed to two factors. First, the oxidation reactions of NH3 create an extra O2 consumption pathway. Second, nitrogen-containing radicals generated during combustion participate in diverse elementary reactions and further expedite O2 depletion.

2.3. Effects of NH3 Addition on Combustion Products and Intermediates of C4 Isomers

The formation rules of CO and CO2 reflect the carbon oxidation degree of C4 isomers. As illustrated in Figure 4a, within the initial 250 ps of NH3-doped 1-butene combustion, the CO formation rate is lower than that under pure combustion conditions. However, the number of CO molecules exceeds that of the pure combustion case after 250 ps. Meanwhile, the total yield of CO2 remains consistently lower throughout the entire reaction process when NH3 is blended. These results demonstrate that NH3 inhibits the initial oxidation of 1-butene, whereas it suppresses the conversion of CO to CO2 in the middle and later reaction stages, thereby lowering the overall carbon oxidation degree of the fuel.

Figure 4.

Figure 4

Temporal evolution of CO and CO2 during NH3-blended combustion of C4 isomers: (a) 1-butene, (b) 2-butene, (c) isobutene, (d) n-butane, (e) isobutane.

As illustrated in Figure 4b, 2-butene exhibits analogous oxidation behaviors to 1-butene. For the NH3-doped combustion system, the mole number of CO remains consistently lower than that of the neat combustion case before 1200 ps, yet CO concentrations rise rapidly and surpass the neat counterpart after 1200 ps. Throughout the entire simulation duration, CO2 yields in the NH3-blended system are inferior to those of neat 2-butene combustion, with the concentration gap continuously widening after 1000 ps. Compared with 1-butene, the turning point where CO overtakes the neat baseline shifts distinctly to a later simulation time for 2-butene. This phenomenon demonstrates that carbon oxidation proceeds more stably in neat internal olefin combustion systems, the radical-inhibiting effect of NH3 can only manifest after prolonged simulation via O2 depletion and insufficient OH supply.

In Figure 4c, the CO evolution curves of neat isobutene combustion and NH3-doped isobutene combustion nearly overlap, and the discrepancy in CO mole numbers at 4000 ps is marginal. Meanwhile, CO2 concentrations under NH3 doping are only slightly reduced relative to the neat condition. These observations indicate that neat isobutene combustion inherently sustains a high-intensity radical pool, where the radical competitive suppression effect and radical replenishment effect induced by NH3 nearly counteract each other.

In contrast to the three butene isomers, CO formation is persistently suppressed when butane is blended with NH3. For n-butane doped with NH3, CO concentrations are maintained at lower levels than the neat n-butane case over the whole simulation period. In terms of CO2, its concentration under NH3 addition is lower than that of neat combustion within 0~2500 ps, the concentration difference narrows and fluctuates alternately from 2500 ps to 3200 ps. Once the simulation time exceeds 3200 ps, CO2 concentrations in the NH3-doped system decline continuously, whereas CO2 yields in neat n-butane combustion keep increasing steadily. For isobutene blended with NH3, both CO and CO2 concentrations are persistently lower than those of neat isobutane combustion, and CO2 formation undergoes stable inhibition throughout the simulation. This reveals that NH3 preferentially consumes oxidative radicals such as OH and O during isobutane combustion, which impedes the full oxidation of carbon-containing intermediates into CO and further restrains the subsequent conversion of CO to CO2.

NH3 can serve as an additional hydrogen source to facilitate the generation of H2O and H2. Figure 5 illustrates the temporal evolution profiles of H2O and H2 during the combustion of C4 isomers blended with NH3. Under all NH3-blended conditions, the molecular count of H2O exceeds that under pure-fuel combustion at corresponding time intervals. In all cases, the H2O concentration rises rapidly at first, reaches a peak, and then decays slowly. In Figure 5a, for the 1-butene/NH3 blended case, the peak molecular number of H2O is approximately 120 at a peak time of around 500 ps. For pure 1-butene combustion, the peak value is roughly 85 molecules, with the peak delayed to 1500 ps. As shown in Figure 5b,c, the H2O peaks of 2-butene and isobutene with NH3 addition occur within 500~1000 ps, with peak molecular counts of about 120 and 118, respectively. Both values are higher than the corresponding pure-fuel peaks of 90 and 88 molecules. Figure 5d,e correspond to n-butane and isobutane mixed with NH3, which exhibit even higher H2O peak quantities of approximately 150 molecules, markedly surpassing the pure-fuel peak values of 105 and 95 molecules, respectively.

Figure 5.

Figure 5

Temporal evolution of H2O and H2 during the combustion of C4 isomers blended with NH3: (a) 1-butene, (b) 2-butene, (c) isobutene, (d) n-butane, (e) isobutane.

The elevated production of H2O originates from the products formed via the reaction between NH3 and OH, as well as oxidation chain reactions involving NH3. Once O2 is fully consumed, the formation rate of OH decreases, and the generated H2O is subsequently consumed by carbon-rich species to produce CO and H2. Accordingly, NH3 addition yields a higher peak concentration of H2O accompanied by a more rapid post-peak decay, which correlates well with the mid-to-late-stage accumulation of CO and the continuous rise in H2 concentration. In contrast, the evolution trend of H2 follows a consistent pattern across all test cases. The molecular population of H2 under NH3-blended combustion of C4 isomers is universally higher than that under pure fuel combustion. Additionally, H2 exhibits no distinct peak and undergoes continuous growth throughout the entire reaction duration. This phenomenon fundamentally arises from the successive dehydrogenation of NH3, which supplies supplementary H for the overall reaction system, and the subsequent recombination of H to form H2. Furthermore, NH3 and its decomposition intermediates preferentially react with OH, reducing the likelihood of H2 being scavenged by OH and thereby further facilitating the continuous accumulation of H2.

Therefore, in terms of reaction products, NH3 blending significantly alters the oxidation level of the reaction system. On the one hand, the formation of CO2 is suppressed. On the other hand, the yields of H2 and H2O rise remarkably, enhancing the reductive nature of the combustion system. Such discrepancies reveal that the effects of NH3 addition on combustion product distribution stem not only from the hydrogen atoms inherent in NH3 itself but also from the molecular structure of the parent C4 fuel.

NH3 blending alters the peak concentrations of small-molecule pyrolysis products and triggers divergent C2/C3 reaction pathways [28,33]. C2H4, an important product from the β-scission of C4 fuels, acts as an essential precursor for acetylene C2H4. Figure 6a reveals that NH3 addition imposes varying degrees of suppression on the peak C2H4 concentration of different isomers. The inhibition effect is the most pronounced for 1-butene, followed by isobutane, isobutene and 2-butene, whereas n-butane is barely affected. This demonstrates that NH3 accelerates the dehydrogenation and oxidative consumption of C2H4 by increasing the concentrations of H, OH and other radicals. Figure 6b illustrates that the evolution of C2H2 displays selectivity. NH3 blending reduces the peak C2H4 values for 1-butene and isobutene, while elevating the C2H4 peak concentration of 2-butene. This is attributed to the enhanced dehydrogenation of small molecules driven by the higher H concentration induced by NH3, whereby the production rate of C2H2 outweighs its consumption rate. In comparison, the peak C2H2 concentrations of butanes show little sensitivity to NH3 addition. Accordingly, NH3 does not exert a universal inhibitory effect across all C4 isomers. Instead, differentiated impacts on individual isomers are achieved by modulating the competitive interplay among C2H4 dehydrogenation, oxidative consumption of C2H2, and fuel pyrolysis pathways.

Figure 6.

Figure 6

Ratios of molecular population and peak time for C4 isomers under pure-fuel combustion and NH3-blended combustion: (a) C2H4, (b) C2H2.

As illustrated in Figure 7, the peak concentration of C3H6 universally declines after NH3 addition. The strongest inhibitory effect is observed for n-butane, followed sequentially by 2-butene, 1-butene and isobutene, while n-butene exhibits nearly negligible variation. From the perspective of reaction pathways, C3H6 serves as the precursor of C3H6. Elevated concentrations of H and OH accelerate the dehydrogenation and oxidative consumption of C3H6, thereby lowering the C3H6 peak value for most C4 isomers. With the exception of isobutene, NH3 blending increases the peak C3H4 concentration of 1-butene, 2-butene, n-butane and isobutane, among which 1-butene presents the most remarkable enhancement. This indicates that for the C3 reaction pathway, the increased H concentration induced by NH3 facilitates successive dehydrogenation of C3H6 to produce C3H4. Nevertheless, the peak C3H4 concentration may also decrease if the parent fuel generates a high yield of C3 precursors, accompanied by intensified oxidative consumption via OH and O.

Figure 7.

Figure 7

Ratios of molecular populations and peak occurrence times for C3 species under pure fuel combustion and NH3-blended combustion: (a) C3H6, (b) C3H4.

Variations in C1 products can reflect the dominant pyrolysis pathways of the reaction system. As depicted in Figure 8, NH3 exerts disparate effects on the peak concentrations of CH3 and CH4. The peak CH3 concentration rises for isobutane, 1-butene and n-butane, whereas it is suppressed for 2-butene and isobutene. In contrast, the peak CH4 concentration is enhanced for 1-butene and 2-butene but inhibited for isobutane, isobutene and n-butane. CH4 is primarily formed via the recombination of CH3 and H. For 1-butene blended with CH3, both CH3 and H concentrations increase simultaneously, which accounts for the most prominent rise in the CH4 peak value. Isobutane and isobutene inherently produce abundant CH3. After NH3 addition, the elevated concentration of OH accelerates the dehydrogenation and consumption of CH4, resulting in a reduction in its peak concentration.

Figure 8.

Figure 8

Ratios of molecular populations and peak occurrence times for C1 species under pure-fuel combustion and NH3-blended combustion: (a) CH3, (b) CH4.

Overall, NH3 exerts dual regulatory effects on the combustion of C4 isomers. On the one hand, NH3 supplies H to promote fuel pyrolysis and dehydrogenation reactions. On the other hand, NH3 modulates the oxidative consumption of carbon intermediates through OH and O. Consequently, the effects of NH3 blending on the pyrolysis pathways of individual C4 isomers are summarized as follows. For 1-butene, the dominant pyrolysis pathway shifts from C2-centered pathways to C1- and C3-centered pathways. In the case of 2-butene, the yields of highly unsaturated C2 products increase, while the formation of C3H6 is suppressed. For isobutene, the generation of most soot precursors is inhibited. As for n-butane and isobutane, NH3 mainly alters the dynamic equilibrium between saturated and unsaturated C3 intermediates.

These trends can be quantitatively compared with published experimental and modeling studies to confirm their physical robustness. Regarding the suppression of CO2, NH3 is a carbon-free fuel, so blending NH3 directly dilutes the carbon source of the system, and the reduction in CO2 emissions with increasing NH3 blending ratio is a general rule in the ammonia co-firing community. For instance, Li et al. [34] carried out three-dimensional pore-scale numerical simulations of NH3-CH4 premixed combustion in a double-layer spherical stacked porous medium burner and found that CO2 emissions reach a peak at an NH3 blending ratio of approximately 10%, decline significantly once the NH3 concentration exceeds 20%, and reach the lowest level at 30%, with CO2 formation being further suppressed under fuel-rich conditions (φ > 1). This is consistent with the present results: under the strongly fuel-rich condition of φ = 4, the CO2 yield of every C4 isomer system blended with NH3 remains lower than that of the corresponding neat combustion throughout the whole reaction process, in which the CO2 formation of the isobutane system is stably suppressed over the entire simulation time, and the CO2 yield of the n-butane system keeps declining after 3200 ps while that of the neat case still rises. Regarding the promotion of H2 and H2O, NH3 is a hydrogen-rich, carbon-free fuel that can decompose into H2 and N2 at elevated temperatures, and its oxidation reactions mainly produce H2O and N2. Under fuel-rich (oxygen-lean) conditions, the pyrolysis and partial oxidation of NH3 supply abundant H to the system and promote H2 accumulation, which agrees with the conclusion that “NH3 promotes H generation” in the porous-medium combustion study [34] and with the summary of hydrogen-rich fuel combustion characteristics in the NH3 combustion review. Quantitatively, the peak H2O concentrations of all systems increase markedly after NH3 blending (e.g., from 85 to 120 molecules for 1-butene with the peak time advanced from 1500 ps to 500 ps, and from 105 to 150 molecules for n-butane), and the final H2 yields also increase significantly, consistent with the mechanisms reported in the above literature. Furthermore, our group has also observed the inhibitory effects of NH3 on CO2 formation and soot precursors as well as the promotion of H2/H2O in C4H10 combustion with H2-NH3 addition [32], NH3-doped ethylene flames [16], and ethylene/hydrogen/ammonia laminar diffusion flames [7] using the same ReaxFF force field, indicating that the present trends are consistent across systems and settings rather than artifacts of the force field or simulation setup.

2.4. Soot Formation Mechanism of NH3-Co-Fired C4 Isomers

Comparisons of soot microstructures help further clarify the effects of isomer configurations and NH3 blending on soot formation. Table 1 illustrates the microstructure, molecular formula, C/H atomic ratio and internal benzene ring number of the maximum soot particles generated within the simulation cycle under five neat combustion conditions, including pure 1-butene, 2-butene, isobutene, n-butane and isobutane flames.

Table 1.

The microstructure, molecular formula, C/H atomic ratio, and internal benzene ring quantity of the largest soot particles generated during the combustion of C4 isomers.

(a) 1-C4H8 (b) 2-C4H8 (c) i-C4H8 (d) n-C4H10 (e) i-C4H10
Micro-
structure
graphic file with name molecules-31-03893-i001.jpg graphic file with name molecules-31-03893-i002.jpg graphic file with name molecules-31-03893-i003.jpg graphic file with name molecules-31-03893-i004.jpg graphic file with name molecules-31-03893-i005.jpg
Molecular formula C108H39 C109H47 C127H42 C58H42 C54H31
C/H atomic ratio 2.77 2.32 3.02 1.38 1.74
Number of benzene rings 18 14 17 2 2

Overall, the number of benzene rings in soot produced by butene fuels is higher than that in soot generated from butane fuels. Among all tested conditions, the flame fueled with 1-butene yields soot containing the maximum benzene ring count of 18, followed by isobutene with 17 benzene rings and 2-butene with 14 benzene rings. In contrast, soot formed under n-butane and isobutane flames only contains 2 and 3 benzene rings respectively, which are considerably lower than those observed for butene flames. This result demonstrates that butene fuels are more prone to form aromatic ring structures and accelerate soot growth under fuel-rich and high-temperature environments. Such discrepancy originates from the distinct molecular configurations of the fuels. Butenes contain C=C double bonds, during high-temperature pyrolysis and dehydrogenation, they can directly produce unsaturated active intermediates including allyl species, alkynes, and propargyl radicals. These intermediates rapidly form benzene rings and PAH structures via addition, cyclization and condensation reactions, thereby supplying abundant precursors for soot inception. By contrast, butanes have no carbon–carbon double bonds. They must undergo sequential dehydrogenation, bond scission and radical recombination reactions to generate unsaturated small molecules and aromatic precursors.

The C/H atomic ratio reflects the maturity of soot particles. In general, a higher C/H ratio corresponds to soot structures with higher carbon abundance and lower hydrogen content. By contrast, a lower C/H ratio indicates limited aromatic condensation and dehydrogenation of soot [33,35]. The flame fueled with isobutene exhibits the maximum C/H ratio, revealing a high carbon enrichment level of its soot molecules, followed by 1-butene and then 2-butene. The C/H ratios under butane combustion conditions are remarkably lower, which implies that the soot molecules generated within the present simulation cycle retain abundant H, accompanied by low aromatization degree and structural maturity. Combined analysis of the C/H atomic ratio and the number of six-membered rings demonstrates that soot from butene flames possesses not only more benzene rings but also generally higher C/H ratios compared with butane-derived soot. This evidence proves that butene fuels readily undergo dehydrogenation, cyclization and aromatic condensation reactions under high-temperature and fuel-rich conditions, thereby forming carbon-enriched microstructures that are closer to mature soot. The above conclusions are consistent with the microstructural configurations of maximum soot particles generated by the combustion of various C4 isomers displayed in Table 1.

Table 2 presents the microstructure, molecular formula, C/H atomic ratio and internal benzene ring count of the maximum soot particles produced within the simulation cycle for NH3-blended flames of 1-butene, 2-butene, isobutene, n-butane and isobutane. Compared with Table 1, the C/H atomic ratios of all test cases rise after NH3 blending. This phenomenon reveals that NH3 addition alters the reaction pathways during soot growth. The elevated C/H ratio indicates an overall carbon-enriched tendency of soot molecules. Nevertheless, nitrogen originating from NH3 can embed into the molecular skeleton of soot, disrupt the cyclization reaction for aromatic carbon structures, and further modify the evolutionary characteristics of soot aromatic skeletons.

Table 2.

Morphology, molecular formula, C/H atomic ratio, N/C atomic ratio and internal benzene ring number of maximum soot particles generated during combustion of C4 isomers under NH3.

(a) 1-C4H8 (b) 2-C4H8 (c) i-C4H8 (d) n-C4H10 (e) i-C4H10
Micro-
structure
graphic file with name molecules-31-03893-i006.jpg graphic file with name molecules-31-03893-i007.jpg graphic file with name molecules-31-03893-i008.jpg graphic file with name molecules-31-03893-i009.jpg graphic file with name molecules-31-03893-i010.jpg
Molecular formula C133H42N8 C117H35N5 C119H34N6 C122H43N9 C124H39N10
C/H atomic ratio 3.17 3.34 3.50 2.84 3.18
N/C atomic ratio 0.06 0.04 0.05 0.07 0.08
Number of benzene rings 21 21 21 17 26

The N/C atomic ratio equals zero for all neat combustion cases, while this ratio increases remarkably under NH3-blended conditions, demonstrating that N can be incorporated into the molecular structure of soot after NH3 participates in combustion reactions [36]. The N doping degree varies with different fuels. Among all test flames, the NH3-blended isobutane flame achieves the highest N/C ratio, followed by n-butane, which indicates stronger N doping effects for NH3-blended butane fuels. The N/C ratios of butene flames are relatively low, yet these values still verify the involvement of nitrogen species in soot formation. Mechanistically, NH3 decomposes into nitrogen-containing active radicals such as NH2, NH and N under high-temperature environments. These nitrogenous species can react with hydrocarbon radicals, aromatic radicals or active sites on soot surfaces to form C-N bonds, nitrogen heterocycles and other nitrogen-containing structures. On the one hand, NH3 blending modifies the elemental composition of soot molecules. On the other hand, it hinders the continuous formation of aromatic carbon networks and further regulates the graphitization degree of soot particles.

Analysis based on the number of six-membered rings reveals that the quantity of six-membered rings under NH3-blended conditions is larger than that of the corresponding neat fuel flames, with a particularly prominent increment observed for butane systems. Soot generated from neat butene flames already contains abundant six-membered rings. After NH3 blending, the number of six-membered rings rises slightly, accompanied by simultaneous increases in C/H and N/C atomic ratios. This result indicates that N is incorporated into soot molecules during soot growth. It can thus be inferred that NH3 blending fails to suppress the formation of six-membered rings, whereas N doping may reduce the ordering degree of aromatic structures. For alkane flames, the increased yield of six-membered rings cannot be fully attributed to the direct promotion effect of NH3. Within the identical simulation duration, soot produced by neat alkane combustion remains at the early growth stage. In contrast, NH3 blending modifies reaction pathways, radical pools and nitrogen-containing intermediates, thereby generating more nitrogen-doped soot structures.

3. Methods

3.1. Operating Conditions

Combustion is a coupled process involving chemical bond cleavage, recombination and free-radical chain reactions, whose transient atomic-scale evolution is difficult to track via experimental approaches. After years of development, ReaxFF MD enables accurate atomic-level characterization of bond breaking and formation throughout reactions, and has emerged as a core simulation technique for elucidating microscale reaction mechanisms [37,38,39]. The theoretical foundation of ReaxFF MD rests on bond-order theory, consisting of a series of bonded and non-bonded energy terms. In practical calculations, these energy terms are expressed as potential function formulas, alongside empirical parameters required for computation—including bond orders, bonded terms and non-bonded terms between all known element pairs derived from quantum mechanical calculations—stored within the force field parameter set. The functional relationship between the total system potential energy (Esystem) and the potential functions of individual energy terms is given in Equation (1).

Esystem= Ebond+Eover+Eunder+Eval+Epen+Ecoa+Etors+Econj+EvdWasls+ECoulomb (1)

In the above equation, Esystem denotes the total potential energy of the simulation system. Specifically, Ebond represents bond energy, Eover and Eunder correspond to the coordination correction energies for over-coordination and under-coordination, respectively. Eval is the valence angle energy, Epen the penalty energy, Ecoa the three-body conjugation energy, Etors the torsion angle energy, and Econj the four-body conjugation energy. The aforementioned eight terms collectively constitute the bonded interaction energy component. The non-bonded interaction energy consists of two contributions: EvdWaals for van der Waals interactions and ECoulomb for Coulombic interactions.

Table 3 summarizes all simulation cases adopted in this work, which were designed following the single-variable control principle. Five typical C4 isomers, namely 1-butene, 2-butene, isobutene, n-butane and isobutane, were selected as research subjects. Two series of simulation cases were established, including baseline groups of pure fuels and comparative groups with NH3 blending, resulting in a total of 10 simulation cases. The initial T was fixed at 2800 K and the ρ was set to 0.3 g/cm3, with a fuel-rich φ of 4 [40,41]. The number of O2 was calculated according to the stoichiometric ratio.

Table 3.

Simulation parameters for ReaxFF MD calculations.

Case 1-C4H8 2-C4H8 i-C4H8 n-C4H10 n-C4H10 NH3 O2 T, K ρ, g/cm3 φ
1 80 0 0 0 0 0 120 2800 0.3 4
2 80 0 0 0 0 40 130
3 0 80 0 0 0 0 120
4 0 80 0 0 0 40 130
5 0 0 80 0 0 0 120
6 0 0 80 0 0 40 130
7 0 0 0 80 0 0 130
8 0 0 0 80 0 40 140
9 0 0 0 0 80 0 130
10 0 0 0 0 80 40 140

It should be noted that ReaxFF MD is a molecular-scale “computational experiment” method aimed at revealing reaction pathways and mechanisms at the atomic level, rather than reproducing the macroscopic operating conditions of real combustion devices. By accelerating reactions at elevated temperatures, ReaxFF enables chemical processes that occur on micro- to millisecond timescales to be fully observed within a nanosecond simulation window. This “high-temperature acceleration” strategy is a common practice in ReaxFF reactive-dynamics studies [37,38]. The rationality of the simulation conditions adopted here is as follows: T = 2800 K lies within the adiabatic flame temperature range of hydrocarbon combustion, providing sufficient activation energy for key processes such as C4 isomer decomposition, oxidation and soot-precursor formation; the high system density of 0.3 g/cm3 increases molecular collision frequency and accelerates reaction progress, thereby yielding sufficient reaction events and statistical samples within limited computational resources; the strongly fuel-rich equivalence ratio φ = 4 is consistent with the physical conditions of soot formation, ensuring that each system generates sufficient and distinguishable soot structures within the simulation window; and the 4000 ps simulation duration guarantees the statistical significance of comparisons among different systems. Meanwhile, ReaxFF itself possesses inherent limitations: (i) it describes bond formation and cleavage based on an empirical bond-order potential, and its reaction enthalpies and energy barriers are less accurate than those of ab initio methods and detailed chemical kinetic mechanisms [38], so this work does not rely on absolute thermochemical quantities and only compares relative trends between systems; (ii) pressure is not an independent control variable in the NVT ensemble, making it difficult to capture third-body collision and pressure-dependent unimolecular dissociation effects, so pressure effects are not addressed here; (iii) at 2800 K the reactions proceed predominantly along high-temperature pathways, which may differ from those at low/intermediate temperatures, and therefore the conclusions of this work are limited to high-temperature fuel-rich conditions; (iv) the simulation system contains only hundreds of C4 molecules (thousands of atoms), and within the 4000 ps timescale soot particles are only in the early nucleation and molecular-growth stage, without covering the complete lifecycle of surface growth, coalescence and oxidation of real soot particles, which limits the statistical representativeness of soot-related conclusions.

3.2. Simulation Details

The construction of the molecular system model was mainly implemented via the ReaxFF module embedded in the Amsterdam Density Functional (ADF) 2021.107 software package, where the CHON-2019 reactive force field was adopted with a time step of 0.25 fs. To guarantee the stability of the reaction ensemble for simulating incomplete combustion of C4 isomers, geometry optimization and energy minimization were sequentially performed to achieve the minimum energy state and reasonable structural configuration of the initial system. Figure 9 presents the optimized ensemble and molecular structures of C4 fuels. It can be observed that fuels, oxidizers and additives are homogeneously mixed after optimization, approximating the state of a premixed flame and facilitating subsequent kinetic analysis.

Figure 9.

Figure 9

Optimized reaction ensemble and molecular configuration of C4 isomers.

All simulation cases in this work adopted the NVT ensemble, where the number of atoms, system volume and temperature were kept constant. The Berendsen thermostat (Intel Xeon E5-2650 V4, Intel Corporation, Santa Clara, CA, USA) with a damping coefficient of 100 fs was utilized for temperature regulation. Trajectory analysis and statistical calculations of products and intermediates were conducted with a bond-order cutoff value of 0.3 Å. Periodic boundary conditions were applied throughout all simulations to eliminate the interference of boundary effects on reaction evolution. All computational cases were performed on a high-performance server equipped with 24 CPU cores (E5-2650 V4). A total simulation duration of 4000 ps via ReaxFF reactive molecular dynamics required a wall-clock time of 12 h. To ensure statistical reliability, three independent simulations with distinct initial configurations are performed for each operating condition, and the final data are averaged from the three independent ensemble-simulation runs. In the study of soot formation, it is generally accepted that the collision-induced nucleation of two tetracyclic aromatic pyrene molecules A4 gives rise to freshly nucleated soot particles containing 32 carbon atoms [42,43]. Therefore, in ReaxFF MD studies, as the reaction proceeds, a so-called young soot particle is considered to have formed once the number of carbon atoms within a given molecule is no less than 32. Furthermore, when the carbon-atom content of a soot particle is no less than 100, it is considered to have developed into a mature soot particle. These classification criteria have been adopted and validated multiple times in our previous studies [40,41]. Consequently, the same classification approach is retained in the present work to discuss the influence of ammonia doping on the soot-formation mechanism of C4 isomers.

Notably, although the present work does not include direct experimental calibration, the reliability of the ReaxFF calculations is supported by three lines of evidence. First, the CHON-2019 reactive force field employed here belongs to the ReaxFF empirical bond-order potential family whose parameters were fitted against quantum-mechanical calculations and has been widely validated in hydrocarbon oxidation, ammonia combustion and soot formation studies [37,38]. Second, the adopted soot-classification criterion (C32 as young soot and C100 as mature soot) follows the generally accepted PAH-collision nucleation framework [42,43] and has been repeatedly reproduced and validated in our previous ReaxFF studies under comparable conditions [40,41]. Third, the key trends observed in the present work—NH3 suppressing CO2 formation and soot-precursor generation while promoting H2/H2O yields—are qualitatively consistent with our group’s and other groups’ ReaxFF results obtained with the same force field across different fuel systems, including methane/ethylene combustion [40,41], C4H10 combustion with H2-NH3 addition [32], and ammonia-doped ethylene flames [7,16,36], as well as with experimental and modeling studies of NH3-CH4 combustion [23,24,34] and reviews of ammonia combustion [5]. This cross-system consistency indicates that the observed trends reflect robust physical chemistry rather than artifacts of the force field or simulation setup. Given the molecular-scale scope of this work, the calculations are used to reveal relative trends and reaction mechanisms rather than to reproduce absolute quantitative emission data, and no direct extrapolation to engineering devices is intended.

The present ReaxFF MD study involves several intentional assumptions and simplifications, which are stated explicitly below together with their justifications. (1) The NVT ensemble is adopted instead of an open-flow or constant-pressure configuration, because the objective is to isolate intrinsic chemical kinetics at a prescribed temperature rather than to reproduce heat-transfer or pressure variations; accordingly, pressure-dependent and third-body effects are not discussed in this work (see Section 3.1). (2) A homogeneous, perfectly premixed initial mixture is assumed. This simplification is appropriate for comparing intrinsic fuel reactivity under identical mixing conditions and deliberately excludes turbulence, fuel-droplet evaporation and non-premixed mixing, which would otherwise obscure the structural-comparative focus of this study. (3) Periodic boundary conditions are applied to mimic an infinitely large, homogeneous system and eliminate surface boundary effects; since only relative trends among the five C4 isomers are compared, finite-size effects influence all systems equally and do not bias the inter-isomer comparison. (4) A bond-order cutoff of 0.3 Å is used for species identification, which is a standard convention in ReaxFF trajectory analysis [38] and ensures that bonded/non-bonded states are defined consistently across all cases. (5) The simulation system contains only hundreds of fuel molecules (thousands of atoms) and evolves over 4000 ps. This finite size and time span are dictated by computational cost and are sufficient to capture fuel decomposition, oxidation and early soot nucleation, but they do not cover later soot growth, coalescence, oxidation or engineering timescales; consequently, soot-related conclusions are restricted to early-stage trends. (6) Only major product families—CO, CO2, H2, H2O, C1-C3 intermediates, PAHs and soot particles—are counted, while trace species and detailed NOx formation are outside the scope of this work. (7) Three independent runs with distinct initial configurations are averaged to control statistical noise. Together, these assumptions represent deliberate trade-offs that make atomic-scale mechanism resolution computationally affordable; because all five isomers and all NH3-blended cases are treated under identical settings, the comparative, mechanism-focused conclusions of this work are not affected by these simplifications.

To directly verify the accuracy of the molecular simulation method applied in this work, the key qualitative trends obtained from the ReaxFF calculations are compared against independent experimental observations on the same fuel families. First, the structural dependence of soot formation is reproduced: the present simulations show that neat olefins generate far more six-membered rings and soot precursors than alkanes, which is qualitatively consistent with counter-flow diffusion flame experiments [18], where the critical fuel concentrations for the macroscopic soot transition threshold follow the order 1-butene (9.0%) < isobutene (9.5%) < isobutane (19%) < n-butane (25%)—i.e., C4 olefins exhibit a far greater sooting tendency than their corresponding alkanes. Second, the inhibitory effect of NH3 on soot precursors observed in this work agrees with tube-furnace co-pyrolysis experiments of C4 isomers with NH3 at 1473 K [22], in which NH3 addition was found to reduce the concentrations of key soot precursors including C2H2, C2H4 and C4H2, and with experimental soot-morphology observations in ammonia-doped ethylene flames [8,16,36]. Third, the product trends induced by NH3 blending—suppressed CO2 formation and enhanced H2/H2O yields—are consistent with experimental and modeling studies of NH3-CH4 combustion [23,24,34] and with ammonia-combustion reviews [5]. Regarding the system size, each case contains 80 fuel molecules (hundreds of molecules and thousands of atoms in total), which lies within the typical range of ReaxFF combustion simulations reported in the literature [37,40,41] and is sufficient to resolve atomic-scale reaction pathways. Because all ten cases are compared under identical molecule counts, density and temperature, finite-size effects act equally on every system and do not bias the relative rankings among C4 isomers; statistical noise is further controlled by averaging three independent runs. In addition, all quantitative criteria used in this work—including the C32/C100 soot classification [42,43], the 0.3 Å bond-order cutoff [38], and the choice of T, ρ and φ—are grounded in published methods rather than arbitrary assumptions. Together, the qualitative agreement with independent experiments on the same fuel systems and the standard methodological criteria confirm that the molecular simulation method applied in this work is fit for its mechanistic, trend-comparison purpose; the conclusions are accordingly limited to relative trends and reaction pathways rather than absolute quantitative predictions.

4. Conclusions

In this work, ReaxFF MD simulations are performed at T = 2800 K and φ = 4 to investigate the pure combustion and NH3-blended combustion processes of five C4 isomers, namely n-butane, isobutane, 1-butene, 2-butene and isobutene. The main conclusions are summarized as follows:

  • (1)

    The combustion and soot formation of C4 isomers exhibit remarkable structural dependence. In the early reaction stage, the ignition rate is governed by the reactivity of terminal double bonds and tertiary C-H bonds. 1-butene and isobutane possess the fastest initial reaction rates, while 2-butene shows the slowest ignition due to steric hindrance. In the mid-stage of reactions, the cracking capacity of carbon skeletons dominates the reaction efficiency, and linear carbon skeletons deliver superior sustained reactivity. In the late combustion stage, resonance-stabilized allylic intermediates tend to persist in the system, hindering the oxidation of olefins. In terms of combustion products, 2-butene demonstrates the strongest oxidation capacity with the highest CO2 yield. Alkanes contain higher H, leading to notably greater H2 production compared with butenes.

  • (2)

    The regulation effect of NH3 doping on C4 isomers is structurally sensitive. For 1-butene and isobutane, NH3 only imposes transient inhibitory effects via radical competition, exerting limited influence on the total burnout time. By contrast, NH3 remarkably accelerates the cracking of 2-butene through supplementary H supply. NH3 addition drastically shortens the depletion time of O2, suppresses CO2 generation, and facilitates the accumulation of H2O and H2. Overall, NH3 reduces the oxidation level of the combustion system and enhances its reducibility. Furthermore, NH3 enriches the radical pool to alter hydrocarbon cracking pathways, resulting in differentiated evolution of small-molecule intermediates for various C4 isomers.

  • (3)

    With respect to soot formation, the quantity of six-membered rings for all C4 isomers rises under NH3-blended combustion relative to pure fuel combustion, among which n-butane achieves the most significant increment. Nevertheless, the elevated six-membered ring content of butanes cannot be fully attributed to the promotion effect of NH3. Under identical simulation durations, soot generated from pure alkanes remains in the early growth stage. NH3 doping accelerates the formation of nitrogen-containing soot structures by rearranging reaction pathways, modifying radical pools and regulating the evolution of nitrogen-bearing intermediates. N is incorporated into soot skeletons in the form of nitrogen functional groups and nitrogen heterocycles, simultaneously elevating the C/H and N/C atomic ratios. In addition, N doping fails to reduce the total yield of six-membered rings, yet it disturbs the construction of aromatic carbon frameworks and diminishes the ordering degree of aromatic structures within soot particles. It should be noted that the above conclusions represent mechanistic insights at the molecular level, supported by cross-system consistency with existing experimental and modeling studies, rather than direct quantitative calibration against engineering operating conditions.

Author Contributions

Conceptualization, Y.W. and L.C.; methodology, J.C. and Y.Z.; software, J.L. and Y.Z.; validation, Y.W. and J.L.; formal analysis, J.C. and Z.Z.; investigation, Y.W. and L.C.; resources, L.C.; data cu-ration, M.G.; writing—original draft preparation, Y.W.; writing—review and editing, Y.W. and J.L.; visualization, M.G. and Z.Z.; supervision, L.C.; project administration, L.C.; funding acquisition, M.G. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The datasets presented in this article are not readily available because the data belong to confidential content as part of the research supported by the National Natural Science Foundation of China, and relevant data access is restricted in accordance with the project confidentiality management requirements. The data presented in this study are available on request from the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

This work was supported by the National Natural Science Foundation of China (52376088), the Natural Science Foundation of Anhui Province (2308085QE168), the Funding for Postdoctoral Researchers᾽ Scientific Research Activities in Anhui Province (2023B718), and the University Synergy Innovation Program of Anhui Province (GXXT-2022-025).

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

References

  • 1.Yin C., Abatzoglou J.T., Jones M.W., Cullen A.C., Sadegh M., Wang J., Liu Y. Increasing synchronicity of global extreme fire weather. Sci. Adv. 2026;12:eadx8813. doi: 10.1126/sciadv.adx8813. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Shindell D., Faluvegi G., Nagamoto E., Parsons L., Zhang Y. Reductions in premature deaths from heat and particulate matter air pollution in South Asia, China, and the United States under decarbonization. Proc. Natl. Acad. Sci. USA. 2024;121:e2312832120. doi: 10.1073/pnas.2312832120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Gao Y., Guo X., Kou W., Guo X., Zhang S., Gao H., Chen D. Advancing high-resolution modeling to unravel the interplay between extreme weather events and air pollution under global warming. Front. Environ. Sci. Eng. 2025;19:100. doi: 10.1007/s11783-025-2020-9. [DOI] [Google Scholar]
  • 4.Li W., Li Y., Katharina K.-H. Chemistry links combustion, atmosphere, and climate. CCS Chem. 2026;8:1715–1736. doi: 10.31635/ccschem.026.202506986. [DOI] [Google Scholar]
  • 5.Zhang M., Wei X., An Z., Okafor E.C., Guiberti T.F., Wang J., Huang Z. Flame stabilization and emission characteristics of ammonia combustion in lab-scale gas turbine combustors: Recent progress and prospects. Prog. Energy Combust. Sci. 2025;106:101193. doi: 10.1016/j.pecs.2024.101193. [DOI] [Google Scholar]
  • 6.Huang F., Guo S., Wang L., Yang Z., Kong W. Experimental and numerical study on the performances of a free-piston engine generator using ammonia and methane fuel mixtures. Fuel. 2023;341:127654. doi: 10.1016/j.fuel.2023.127654. [DOI] [Google Scholar]
  • 7.Wang Y., Liu K., Luo K., Chang K., Gu M. Study on the mechanism of polycyclic aromatic hydrocarbons and soot formation in ethylene/hydrogen/ammonia laminar diffusion flames. Combust. Flame. 2025;277:114168. doi: 10.1016/j.combustflame.2025.114168. [DOI] [Google Scholar]
  • 8.Hua Y., Xiang X., Qiu L., Gao D. Effect of ammonia addition on the morphology and nanostructure evolution of soot particles in ethylene co-flow diffusion flames. J. Energy Inst. 2023;110:101356. doi: 10.1016/j.joei.2023.101356. [DOI] [Google Scholar]
  • 9.Lee J.H., Lee S.I., Kwon O.C. Effects of ammonia substitution on hydrogen/air flame propagation and emissions. Int. J. Hydrogen Energy. 2010;35:11332–11341. doi: 10.1016/j.ijhydene.2010.07.104. [DOI] [Google Scholar]
  • 10.Mashruk S., Shi H., Mazzotta L., Ustun C.E., Aravind B., Meloni R., Alnasif A., Boulet E., Jankowski R., Yu C., et al. Perspectives on NOx emissions and impacts from ammonia combustion processes. Energy Fuels. 2024;38:19253–19292. doi: 10.1021/acs.energyfuels.4c03381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Mathieu O., Petersen E.L. Experimental and modeling study on the high-temperature oxidation of Ammonia and related NOx chemistry. Combust. Flame. 2015;162:554–570. doi: 10.1016/j.combustflame.2014.08.022. [DOI] [Google Scholar]
  • 12.Atlasov V.R., Palankoev T.A., Dement’ev K.I. Pathways of conversion of middle distillate hydrocarbons in deep catalytic cracking to produce light olefins. ACS Omega. 2025;10:48761–48772. doi: 10.1021/acsomega.5c06758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Akram A.A.-A., Abdullah M.A., Sulaiman S.A.-K. Thermal and catalytic cracking of whole crude oils at high severity. J. Anal. Appl. 2020;145:104705. doi: 10.1016/j.jaap.2019.104705. [DOI] [Google Scholar]
  • 14.Agbaje T.A., Vega L.F., Khaleel M., Wang K., Karanikolos G.N. Membranes and adsorbents in separation of C4 hydrocarbons: A review and the definition of the current upper bounds. Sep. Purif. Technol. 2021;278:119530. doi: 10.1016/j.seppur.2021.119530. [DOI] [Google Scholar]
  • 15.Ma W., Zhu G., Yuan Q., Yang J. Catalytic Dehydrogenative Cracking of C4 Hydrocarbons on a Bifunctional Metal–Acid Catalyst. Catalysts. 2025;15:1011. doi: 10.3390/catal15111011. [DOI] [Google Scholar]
  • 16.Liu Q., Gu M., Ouyang Z., Sun J., Li S., Chen P., Wang Y. Study on the oxidation reactivity and NO reduction characteristics of soot in ammonia-doped ethylene flame. Energy. 2025;319:135082. doi: 10.1016/j.energy.2025.135082. [DOI] [Google Scholar]
  • 17.Eyisse E.F., Nadimi E., Wu D. Ammonia Combustion: Internal Combustion Engines and Gas Turbines. Energies. 2024;18:29. doi: 10.3390/en18010029. [DOI] [Google Scholar]
  • 18.Chen C., Zhao X., Qi D., Yang K., Xu L., Li T., Ying Y., Liu D. Sooting transition diagnostics in counter-flow flames of C4 isomer fuels. Energy. 2023;262:125612. doi: 10.1016/j.energy.2022.125612. [DOI] [Google Scholar]
  • 19.Zhang Y., Cai J., Zhao L., Yang J., Jin H., Cheng Z., Li Y., Zhang L., Qi F. An experimental and kinetic modeling study of three butene isomers pyrolysis at low pressure. Combust. Flame. 2012;159:905–917. doi: 10.1016/j.combustflame.2011.09.005. [DOI] [Google Scholar]
  • 20.Guo L., Zhu Y., Li D., Sun W., Shi Y., Liu C., Zhang H., Liu X., Han J., Guo Y., et al. Study on the combustion characteristics and reaction mechanism of high-reactive-fuels-blended ammonia. J. Energy Inst. 2025;123:102270. doi: 10.1016/j.joei.2025.102270. [DOI] [Google Scholar]
  • 21.Gain B., Bilal M., Wang Y.-F., Mansy A.E., Yousuf M., Jin K.-R., Wang D., Su B., Liaqat A., Khalil M., et al. Elucidating the role of ammonia doping in modulating combustion dynamics in ethane fuel-rich jet flames. Fuel. 2026;416:138515. doi: 10.1016/j.fuel.2026.138515. [DOI] [Google Scholar]
  • 22.Chen C., Ying Y., Yang K., Qi D., Yu R., Chen M., Liu D. Revealing soot formation via gas-solid coupling in C4 isomer fuels pyrolysis with carbon-free ammonia addition. Combust. Flame. 2025;273:113930. doi: 10.1016/j.combustflame.2024.113930. [DOI] [Google Scholar]
  • 23.Rocha R.C., Zhong S., Xu L., Bai X.-S., Costa M., Cai X., Kim H., Brackmann C., Li Z., Aldén M. Structure and Laminar Flame Speed of an Ammonia/Methane/Air Premixed Flame under Varying Pressure and Equivalence Ratio. Energy Fuels. 2021;35:7179–7192. doi: 10.1021/acs.energyfuels.0c03520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Jin T., Dong W., Qiu B., Xu C., Liu Y., Chu H. Effect of Ammonia on Laminar Combustion Characteristics of Methane–Air Flames at Elevated Pressures. ACS Omega. 2022;7:15326–15337. doi: 10.1021/acsomega.1c05938. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Pashchenko D. Hydrogen-rich fuel combustion in a swirling flame: CFD-modeling with experimental verification. Int. J. Hydrogen Energy. 2020;45:19996–20003. doi: 10.1016/j.ijhydene.2020.05.113. [DOI] [Google Scholar]
  • 26.Wang Q., Zhang P., Liu C., Yang B. Co-oxidation of ammonia with ethylene and acetylene: An experimental and kinetic modeling study. Combust. Flame. 2026;284:114630. doi: 10.1016/j.combustflame.2025.114630. [DOI] [Google Scholar]
  • 27.Fenard Y., Dagaut P., Dayma G., Halter F., Foucher F. Experimental and kinetic modeling study of trans-2-butene oxidation in a jet-stirred reactor and a combustion bomb. Proc. Combust. Inst. 2015;35:317–324. doi: 10.1016/j.proci.2014.05.007. [DOI] [Google Scholar]
  • 28.Pan L., Hu E., Zhang J., Tian Z., Li X., Huang Z. A high pressure shock tube study of 1-butene oxidation and its comparison with n -butane and alkenes. Fuel. 2015;157:21–27. doi: 10.1016/j.fuel.2015.04.062. [DOI] [Google Scholar]
  • 29.Li Y., Zhou C.-W., Curran H.J. An extensive experimental and modeling study of 1-butene oxidation. Combust. Flame. 2017;181:198–213. doi: 10.1016/j.combustflame.2017.03.023. [DOI] [Google Scholar]
  • 30.Zhao P., Yuan W., Sun H., Li Y., Kelley A.P., Zheng X., Law C.K. Laminar flame speeds, counterflow ignition, and kinetic modeling of the butene isomers. Proc. Combust. Inst. 2015;35:309–316. doi: 10.1016/j.proci.2014.06.021. [DOI] [Google Scholar]
  • 31.Li Y., Zhou C.-W., Somers K.P., Zhang K., Curran H.J. The oxidation of 2-butene: A high pressure ignition delay, kinetic modeling study and reactivity comparison with isobutene and 1-butene. Proc. Combust. Inst. 2017;36:403–411. doi: 10.1016/j.proci.2016.05.052. [DOI] [Google Scholar]
  • 32.Wang Y., Liu Q., Ouyang Z., Chang K., Sun J., Liu D., Huang X., Gu M. Characterization and mechanism of soot formation in C4H10 combustion with H2-NH3 addition. Fuel. 2024;371:131992. doi: 10.1016/j.fuel.2024.131992. [DOI] [Google Scholar]
  • 33.Wang K., Villano S.M., Dean A.M. Experimental and kinetic modeling study of butene isomer pyrolysis: Part II. Isobutene. Combust. Flame. 2017;176:23–37. doi: 10.1016/j.combustflame.2016.07.036. [DOI] [Google Scholar]
  • 34.Li D., Tan F., Wang X., Xu J., Wang S., Li F., Zhu T., Zhu L., Wu Z., Liao M. Three-Dimensional Pore-Scale Numerical Study of the Impact of Operation Parameters on Ammonia–Methane Premixed Combustion in a Double-Layer Spherical Stacked Porous Medium Burner. ACS Omega. 2026;11:27381–27390. doi: 10.1021/acsomega.6c01628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Johansson K.O., El Gabaly F., Schrader P.E., Campbell M.F., Michelsen H.A. Evolution of maturity levels of the particle surface and bulk during soot growth and oxidation in a flame. Aerosol Sci. Technol. 2017;51:1333–1344. doi: 10.1080/02786826.2017.1355047. [DOI] [Google Scholar]
  • 36.Chang K., Gu M., Wang Y., Zou S., Ouyang Z., Han Y., Sun J. Investigation of soot and NO formation pathway and the interaction between C–N species in ammonia-doped ethylene co-flow diffusion flames. Combust. Flame. 2026;285:114727. doi: 10.1016/j.combustflame.2025.114727. [DOI] [Google Scholar]
  • 37.Shi H., Zhao W., Wang Y., Liu Z., Wang T., Niu Z., Zhou L., Wei H. A ReaxFF molecular dynamics study of pyrolysis and coking inhibition by low/zero-carbon fuel additives in aviation fuels. Combust. Flame. 2026;285:114738. doi: 10.1016/j.combustflame.2025.114738. [DOI] [Google Scholar]
  • 38.Kimberly C., Adri V.D., William G. ReaxFF reactive force field for molecular dynamics simulations of hydrocarbon oxidation. J. Phys. Chem. A. 2008;112:1040–1053. doi: 10.1021/jp709896w. [DOI] [PubMed] [Google Scholar]
  • 39.Chen M., He J., Zhao X., Yu R., Yang K., Liu D. Combustion condition predictions for C2-C4 alkane and alkene fuels via machine learning methods. Fuel. 2024;373:132375. doi: 10.1016/j.fuel.2024.132375. [DOI] [Google Scholar]
  • 40.Wang Y., Gu M., Wu J., Cao L., Lin Y., Huang X. Formation of soot particles in methane and ethylene combustion: A reactive molecular dynamics study. Int. J. Hydrogen Energy. 2021;46:36557–36568. doi: 10.1016/j.ijhydene.2021.08.125. [DOI] [Google Scholar]
  • 41.Wang Y., Gu M., Cao L. Reactive molecular dynamics simulations of soot formation in acetylene combustion with hydrogen addition. Energy Rep. 2021;7:1198–1213. doi: 10.1016/j.egyr.2021.09.144. [DOI] [Google Scholar]
  • 42.Eaves N.A., Zhang Q., Liu F., Guo H., Dworkin S.B., Thomson M.J. CoFlame: A refined and validated numerical algorithm for modeling sooting laminar coflow diffusion flames. Comput. Phys. Commun. 2016;207:464–477. doi: 10.1016/j.cpc.2016.06.016. [DOI] [Google Scholar]
  • 43.Wang Y., Raj A., Chung S.H. A PAH growth mechanism and synergistic effect on PAH formation in counterflow diffusion flames. Combust. Flame. 2013;160:1667–1676. doi: 10.1016/j.combustflame.2013.03.013. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets presented in this article are not readily available because the data belong to confidential content as part of the research supported by the National Natural Science Foundation of China, and relevant data access is restricted in accordance with the project confidentiality management requirements. The data presented in this study are available on request from the corresponding author.


Articles from Molecules are provided here courtesy of Multidisciplinary Digital Publishing Institute (MDPI)

RESOURCES