Experimental and Modeling High-Pressure Study of Ammonia–Methane Oxidation in a Flow Reactor

The present work deals with an experimental and modeling analysis of the oxidation of ammonia–methane mixtures at high pressure (up to 40 bar) in the 550–1250 K temperature range using a quartz tubular reactor and argon as a diluent. The impact of temperature, pressure, oxygen stoichiometry, and CH4/NH3 ratio has been analyzed on the concentrations of NH3, NO2, N2O, NO, N2, HCN, CH4, CO, and CO2 obtained as main products of the ammonia–methane mixture oxidation. The main results obtained indicate that increasing either the pressure, CH4/NH3 ratio, or stoichiometry results in a shift of NH3 and CH4 conversion to lower temperatures. The effect of pressure is particularly significant in the low range of pressures studied. The main products of ammonia oxidation are N2, NO, and N2O while NO2 concentrations are below the detection limit for all of the conditions considered. The N2O formation is favored by increasing the CH4/NH3 ratio and stoichiometry. The experimental results are simulated and interpreted in terms of an updated detailed chemical kinetic mechanism, which, in general, is able to describe well the conversion of both NH3 and CH4 under almost all of the studied conditions. Nevertheless, some discrepancies are found between the experimental results and model calculations.


■ INTRODUCTION
Nowadays, the rise in the global concentration of carbon dioxide and harmful pollutants due to the growing energy demand has become a serious problem.Future sustainable scenarios involve the use of carbon-free fuels.In this sense, species such as hydrogen and ammonia play a role as possible substitutes for fossil fuels, 1 to solve the climate change problem.An alternative solution is the use of noncarbon fuels, such as ammonia or hydrogen, as fuels or as chemical storage. 2 Hydrogen has attracted attention as a fuel with zero CO 2 emissions, 3 but it has some application inconveniences associated with its high storage cost.Proof of this is the fact that storing H 2 at high scales in practical applications implies compression to between 350−700 bar or cryogenic cooling to 20 K. 4 For these reasons, H 2 storage is more difficult and expensive than storing NH 3 . 5Moreover, following the interest in ammonia as a hydrogen carrier, there is nowadays an increased interest in ammonia as a fuel.Since ammonia is one of the most produced chemicals in the world, it exhibits a mature production technology, transportation, and storage infrastructure. 6,7Ammonia has been suggested as a substitute for hydrogen, 8 and since the previous decade, initiatives to accelerate global decarbonization have increasingly focused on the use of NH 3 as a feasible alternative fuel. 9dditionally, ammonia can ideally be burned in an environmentally benign way, producing N 2 and H 2 O (R1) 10 : −14 In this way, ammonia can react with NO by the following mechanisms (NH 3 → NO → NNH → N 2 and NH 3 → NO → N 2 ) within the combustion chamber, as pointed out by Alzueta et al. 12 through reactions (R2−R5): However, ammonia presents some challenges for its practical implementation associated with its poor combustion characteristics, such as its high ignition temperature and low flammability, 7,10 potential NO x emissions, 12 and a low laminar burning velocity which is about five times lower compared to CH 4 /air flames. 15−46 From a practical point of view, pressure is an important variable since turbines and certain engines will operate under high pressure conditions.
Increasing pressure has been reported to contribute to reduce the unburned NH 3 and NO emission. 47In this sense, it is essential to examine whether the same effect happens for other reaction products, such as HCN.In this regard, studies in the literature on NH 3 /CH 4 blends at high pressure (from 5 to 100 bar) are rather scarce, and among those we can mention a study on turbulent burning velocity and flame region in a nozzle burner at 5 bar by Ichikawa et al., 48 an autoignition delay time study from 20 to 70 bar and temperatures from 930 to 1140 K of different CH 4 /NH 3 mixtures (0, 5, 10 and 50% of CH 4 in the mixture) in a rapid compression machine by Dai et al., 43 an experimental and numerical study of autoignition at 1. 75  and 10  bar in a shock tube, in which CH 4 /NH 3 ratios of 0, 0.1, and 0.5 were studied, 49 an experimental and numerical study on laminar burning velocity in a pressurized chamber up to 5 bar by Wang et al. 50and chemical kinetic modeling studies. 34,41In these studies, it was found that, at higher pressures, there is less unburned NH 3 and lower NO emissions at the outlet of the reactor compared to what occurs at atmospheric pressure.Additionally, the ignition delay time decreases as CH 4 is added to the mixture, even if added in a small ratio, compared to the conversion of net ammonia.The oxygen excess ratio is also important since the ignition delay time is higher at lower oxygen concentrations.−50 To our knowledge, there is a lack of studies about NH 3 /CH 4 mixture oxidation under relevant conditions carried out in plug flow reactors and high pressures (10 to 40 bar) in a variety of stoichiometries ranging from reducing to oxidizing atmospheres.In this context, the aim of this work is to extend the knowledge of NH 3 /CH 4 mixtures oxidation using the experimental system described above, by analyzing the effect of temperature (from 550 to 1250 K), pressure (from 10 to 40 bar), oxygen excess ratio (from reducing, λ = 0.7, to oxidizing conditions, λ = 3) and CH 4 /NH 3 ratio (0.5, 1 and 2).Experiments are performed using argon as a bath gas, which allows us to accurately determine the molecular nitrogen formed during the reaction and therefore to be able to perform nitrogen balances.Additionally, in order to understand the chemistry and product species formation, the Energy & Fuels results are interpreted in terms of a chemical kinetic mechanism for ammonia/hydrocarbon mixtures, which has been compiled from the literature and updated in the present work, and the main reaction pathways through which reactions proceed have been determined and discussed.

■ METHODOLOGY
Conversion of reactants and formed products during the combustion of a NH 3 /CH 4 mixture are studied at high pressures (10, 20, 30, and 40 bar) under well-controlled laboratory-scale conditions.The experimental setup, which has been used in success in previous studies 3,51−53 is schematized in Figure 1.In the present work, a temperature range of 550−1250 K and an oxygen excess ratio (λ) ranging from 0.7−3 have been considered.
The reactant gases are fed from gas cylinders (providers: Air Liquide, Praxair, or Messer) and premixed before entering the quartz flow reactor (153.8 cm long, inner diameter of 0.6 cm), which is placed inside a three-zone electrically heated oven, allowing an isothermal zone inside the tube of approximately 35 cm.This isothermal zone was determined experimentally through the temperature profiles performed at different temperatures and pressures.Figure 2 shows, as an example, the temperature profiles measured for the pressure of 40 bar, using a gas flow rate of 1000 mL (STP)/min, as has been used in all the experiments.Profiles for the different pressures and temperatures were also determined.The temperature profiles along the reactor were determined using a thermocouple positioned in the space between the quartz tube and the steel shell used to keep the pressure constant.The temperature measurement was taken each 5 cm in the central zone and each 10 cm at both sides of the reactor.
The oxygen excess ratio (λ) is defined on the basis of the NH 3 oxidation reaction to N 2 (R1) (4NH 3 + 3O 2 ⇌ 2N 2 + 6H 2 O) according to eq 1: The flow rate is 1000 mL (STP)/min and implies a temperature-and pressure-dependent gas residence time in the isothermal reaction zone, as described in eq 2.
In the experiments, concentrations of NH The estimated uncertainty of the measurements is within ±5% but not less than 5 ppm for the continuous analyzers and 10 ppm for the gas microchromatograph and FTIR determinations.As mentioned, mixtures are diluted in argon.
Table 1 summarizes the experimental initial conditions.Sets 17, 22 and 17R, 22R correspond to repetition experiments.The experimental results of repeated experiments will be later compared to evaluate the repetitiveness of the experimental results.The influence of pressure at different temperatures has been evaluated in the 10 to 40 bar range for stoichiometries of 1 and 3 (sets 2−6, 8−15, and 17−22 in Table 1).For the highest pressure studied, 40 bar, we have also considered a fuel-rich stoichiometry of 0.7 (sets 1, 7, and 16 in Table 1), as the  a All experiments are performed in the 550−1250 K temperature interval with a total flow rate of 1000 mL (STP)/min and using Ar as a bath gas.The residence time is defined by eq 2.
Energy & Fuels experimental higher pressure conditions allowed us to see the NH 3 reaction at comparatively lower temperatures.

■ KINETIC MODELING
The experimental results of the present work are simulated using a gas-phase chemical kinetic mechanism based on earlier work on nitrogen chemistry by Glarborg et al., 54 drawing on more recent work on amine chemistry by Stagni et al., 55 updated adding an acetonitrile reaction subset by Alzueta et al. 56 as well as a reaction subset of methylamine by Glarborg et al., 57 updated by Marrodań et al., 58 and including modifications and/or recommendations of the recent studies of Burke, 59 Klippenstein et al., 60 Marshall et al., 61 Alzueta et al., 13,44 and Glarborg et al. 62−64 Reactions recently revised such as NH 2 + HO 2 by Klippenstein et al. 65 and NH 2 + NO 2 by Glarborg, 64 as well as steps involved in amine pyrolysis 61,63 and H 2 NO reactions proposed by Stagni et al. 66 were included as well.Other steps involved in amine chemistry are updated from the work of Cobos and Gao. 67,68he mechanism has been extended to include DME conversion including several reaction subsets taken from previous work of Marrodań et al., 69 which were tested and Energy & Fuels validated under atmospheric-pressure conditions 70 and modified to consider high-pressure conditions by Marrodań et at. 69,71,72These modifications include C 1 −C 2 and NO interactions, proposed by Glarborg et al. 73 and have been revised and updated according to more recent studies involving NO x that work developed and validated under high-pressure conditions, 74−78 as well as reaction subsets for compounds such as ethanol, C 2 H 2 proposed by Alzueta et al. 79 for atmospheric conditions and modified by Gimeńez et al. 78 to consider highpressure conditions.The DME subset taken from the work of Alzueta et al. 80 at atmospheric pressure has been updated according to more recent mechanisms 81,82 to take into account the high-pressure effects.Additionally, the mechanism includes a subset for oxidation of formic acid based on the work of Marshal and Glarborg, 83 reaction subsets for methyl formate, 71 dimethoxymethane, 51 and ethanol 72 taken from different sources, as reported.The thermodynamic data come from the same sources as for the kinetic mechanisms used.The full mechanism is available as the Supporting Information.
The full compiled mechanism with the modifications mentioned has been tested, compared, and validated against several experimental data sets using different experimental setups in a wide range of conditions by Marrodań et al. 69 for DME mixtures.
Calculations have been carried out in the present work using the plug-flow reactor (PFR) model of the Chemkin Pro suite (2016), 84 the initial conditions for each experiment as listed in Table 1, and a "fix gas temperature" type problem, using the nominal reaction temperature at the flat temperature zone since similar results were obtained with and without the measured temperature profiles.The validity of the model and the mechanism has been assessed using different literature studies in which flow reactor data at atmospheric pressure were presented. 12,18Comparison of literature experimental results, shown in Table S1 of the Supporting Information, and calculations with the mechanism compiled and updated in this work are included in Figures S1 and S2 of the Supporting Information.The performance of the model is very good as well for the atmospheric flow reactor experimental data of Figures S1  and S2, which indicates that the model does a good job simulating the conversion of ammonia/methane mixtures, as will also be seen as follows.
■ RESULTS AND DISCUSSION Effect of Pressure.Figure 3 shows the conversion of CH 4 as a function of temperature for different pressures (from 10 to 40 bar), different CH 4 /NH 3 ratios (0.5, 1, and 2), and oxygen excess ratios (λ = 1 and 3).Symbols denote experimental results and lines model calculations, from now on.As can be seen, the reaction onset temperature of CH 4 oxidation shows a pressure dependence as also pointed out by other authors. 43Generally, this temperature decreases markedly with increasing temperature at all pressures studied, as also observed with pure ammonia oxidation at high pressure. 53,85or a pressure increase of 20 bar (from 20 to 40 bar), keeping the rest of the conditions similar, the CH 4 oxidation reaction onset temperature decreases under stoichiometric conditions (λ = 1) is 150 K (1025 K → 875 K), 100 K (925 K → 825 K) and 50 K (875 K → 825 K) and under oxidizing conditions (λ = 3) is 75 K (900 K → 825 K), 25 K (850 K → 825 K) and 25 K (825 K → 800 K), for the CH 4 /NH 3 ratio of 0.5, 1, and 2, respectively.The different reaction onset temperatures for both NH 3 and CH 4 are summarized in Table S2 of the Supporting Information.Under the current studied conditions, it is noteworthy that, for the same stoichiometry, the higher the CH 4 /NH 3 ratio, the lower the effect of pressure on the onset temperature of the CH 4 oxidation reaction.This effect is also observed when we switch from stoichiometric to oxidizing conditions, keeping a similar CH 4 /NH 3 ratio.This is in line with the findings of other authors. 12,43he experimental results are compared to modeling predictions using the kinetic mechanism previously described.The model generally reproduces very well the trends of CH 4 consumption, for the different pressures under the stoichiometric and oxidizing conditions of Figure 3. Modeling calculations indicate that the full conversion of CH 4 is obtained approximately at the same temperature as those in the experimental results.
CH 4 consumption takes place mainly through H atom abstraction by OH (R8) and by reaction with the amine radical (R9).At the highest temperatures considered, (R8) becomes the main CH 4 consumption reaction under the studied conditions.Alzueta et al. 12 also found (R8) and (R9) as the main CH 4 conversion reaction in a flow reactor study of NH 3 / CH 4 oxidation at atmospheric pressure.Under certain conditions, CH 4 /NH 3 = 0.5, reaction (R9) is the main CH 4 consumption reaction.This is in line with the findings of Dai et al. 43 who suggest that (R9) becomes more important at CH 4 / NH 3 = 0.05 than for CH 4 /NH 3 = 0.5.
Figure 3 also shows the repeatability of sets 17, 22 and 17R, 22R, respectively.As seen, the reproducibility of the experiments is very good in all the temperature ranges considered, which is an indication of the good performance of the experimental system and experimental procedure.
To get some insight into the reaction pathway through which the oxidation conversion of CH 4 in the presence of NH 3 proceeds at high pressure and different stoichiometries, we have made reaction pathway analyses for the different conditions considered. Figure 4 shows the reaction path diagram for CH 4 consumption at [CH 4 ] = 1000 ppm, and λ = 1 and 3 for the highest (40 bar) and lowest (10 bar) studied pressures when 10% of the NH 3 is consumed.The only difference observed occurs at λ = 1 and 10 bar (green).In any case, at the end, the same species are involved, producing CO and CO 2 as a general reactor output product.
The species and pathways marked in green are important only for 10 bar and stoichiometric conditions.As can be seen, the Figure 5 shows the conversion of NH 3 as a function of temperature for different pressures (from 10 to 40 bar), different Energy & Fuels CH 4 /NH 3 ratios (0.5, 1 and 2) and oxygen excess ratios (λ = 1 and 3), i.e., similar conditions as Figure 3.For any pressure studied, the NH 3 consumption shows the same reaction tendency as CH 4 as the temperature increases.The NH 3 concentration is sharply reduced at a given temperature.Both the NH 3 oxidation onset temperature and the decrease of the onset temperature for NH 3 consumption due to the pressure increase coincide with those given above for CH 4 .This effect also happens at atmospheric pressure where both species began to be consumed at the same temperature. 12As can be seen in Figure 5, the pressure effect on the onset reaction temperature is more pronounced under stoichiometric conditions than at oxidizing ones, contrary to what happens for pure ammonia oxidation, 53 where the pressure increase had an effect independent of the stoichiometry.Thus, bearing in mind the use of ammonia as a fuel, its mixtures with methane provide us benefits when working with excess oxygen and increasing the pressure due to the decrease in the NH 3 oxidation reaction onset temperature from 1100 K 53 to 800 K for CH 4 /NH 3 = 0 and 2, respectively, at 40 bar.
For a given stoichiometry, the pressure effect is more pronounced at lower CH 4 /NH 3 ratios.The conditions under which an increase in pressure is least noticeable are CH 4 /NH 3 = 2 and λ = 3 (Figure 5).In all studied cases of NH 3 oxidation, NH 3 starts to be consumed by reaction (R2) as well known in the literature. 12,13,53igure 5 e,f, shows the repeatability of sets 17, 22 and 17R, 22R conditions, respectively.As for CH 4 , the reproducibility of the experiments is very good in all the temperature range considered.
Figure 6 shows the NH 3 consumption reaction path for CH 4 / NH 3 = 1.On the left, for λ = 0.7 at 40 bar and λ = 1 at 10 (green) and 40 bar.On the right, for λ = 3 at 10 and 40 bar (green).In both cases, black lines represent the common path for the different conditions studied.As can be seen in Figure 6, NH 3 → NH 2 → H 2 NO → HNO → NO and NH 3 → NH 2 → N 2 are the major consumption NH 3 reaction channels: NO reacts with HO 2 to form NO 2 , and with NH 2 to form N 2 .The produced NO 2 is quickly consumed by reaction with CH 3 to form CH 3 ONO and with NH 2 to form H 2 NO; thus, no appreciable quantities of this species are found at the reactor outlet.
We also performed sensitivity analysis.Figure 7 shows an example of the results obtained for CH 4 /NH 3 = 1, at 40 bar and λ = 1 at 878 K. CH 4 consumption is promoted by its reaction with NH 2 , OH, and HO 2 to form CH 3 radicals, by the reaction of CH 2 O with HO 2 , OH, and O 2 to form HCO, by reaction of HO 2 radical with nitrogen species (NH 3 , NH 2 , NO) to form OH radicals 43 and by reaction of CH 3 radicals with O 2 also to form OH radicals.Under the same experimental conditions, the reaction of NH 3 with OH to form NH 2  The main products measured in significant amounts during the conversion of NH 3 /CH 4 mixtures are NO, N 2 O, N 2 , H 2 , CO, and CO 2 .N 2 and N 2 O are the most abundant nitrogen species since NO is consumed by reaction with NH 3 . 44NO 2 is below 10 ppm in all studied experimental conditions, which is consistent with the results of the previous studies for pure ammonia oxidation. 53Figures S3 and S4 of the Supporting Information compare, respectively, experimental and simulated results of N 2 and NO obtained during the oxidation of NH 3 in its mixtures with CH 4 at different oxygen excess ratios for each pressure studied.NO is produced under oxidizing conditions and CH 4 /NH 3 = 1 and 2. Model calculations reproduce well the experimental observations.NO production does not follow a clear trend with varying pressure, with NO concentration increasing with pressure for CH 4 /NH 3 = 1 and decreasing for CH 4 /NH 3 = 2 with a similar pressure variation (Figure S4).Also, the difference in the concentration of NO produced is not remarkable.For the same pressure, NO production is favored by CH 4 addition, with a peak of 16 ppm for CH 4 /NH 3 = 1 and 36 ppm for CH 4 /NH 3 = 2. Probably, the diminution of the NH 3 ratio in the mixture provokes this NO increase in the exhaust gases, 28 which is a drawback of using mixtures with high CH 4 / NH 3 ratios.The main production reaction of NO is (R14) and to a minor extent (R15), (R16) and (R17).It is noted that (R15) and (R16) are more noticeable for λ = 3 than λ = 1, and (R17) is only remarkable at the highest pressures.+ + (R15) NO is consumed mainly through reaction (R18) to form NO 2 and to a minor extent via reactions (R3) and (R4) to form N 2 and NNH.
This is consistent with the results obtained in our previous work on high-pressure ammonia consumption in a flow reactor 53 in which NO was not produced under stoichiometric conditions.The typical thermal DeNO x reaction NH 2 + NO ⇌ NNH + OH has not been found to be very important under the studied conditions, as other authors have. 20n contrast, we found a clear effect of pressure on the N 2 O concentration at the reactor output, which consists of obtaining a higher N 2 O concentration at higher working pressures.A direct relationship has also been found between the CH 4 /NH 3 ratio and the N 2 O production, which will be discussed below.Figure 8 (rescaled as Figure S5 in the Supporting Information) shows the results of N 2 O concentration as a function of temperature.In that figure, a maximum of 140, 215, and 290 ppm of N 2 O at 40 bar at λ = 3 for CH 4 /NH 3 ratios of 0.5, 1, and 2, respectively is observed.Model calculations reproduce the main trends observed experimentally, even though they underpredict the specific values.The model simulations indicated that the N 2 O concentration is very low under all studied conditions, but, as can be seen in Figure 8, this does not occur experimentally, where concentrations are higher than calculations.According to the model, the production of N 2 O takes place through reactions (R19) and (R20), in comparison to atmospheric pressure where this occurs exclusively via (R20). 18 In the presence of CH 4 with high concentration levels of CO, N 2 O consumption mainly occurs through reaction (R21) as found by Sun et al. 20 and to a minor extent through reaction.
Concerning N 2 , this species is produced by reaction (R3) and to a minor extent by reaction (R5).It is noticeable that at intermediates and high temperatures, in which the NH 3 concentration is below 50%, (R21) starts to be important in N 2 production.
Regarding CO, in Figure 9, CO is produced through reactions (R23) and (R24) under all studied conditions.Reaction (R24) is favored by the increase in pressure, which is even more noticeable under oxidizing conditions.Also, as the CH 4 /NH 3 ratio increases, this effect is accentuated, with the (R23) reaction being negligible at CH 4 /NH 3 = 2.
As far as consumption reactions are concerned, while (R25) is always present, (R26) is important only for the lowest pressure studied and (R27) and (R28) for the higher pressure conditions considered.
We have evidenced some instability in calculations with the model under the given conditions.This issue is more relevant as the pressure increases, in particular, for stoichiometric

Energy & Fuels
conditions.In order to analyze this in certain detail, we performed some tests.It is found that the reaction NH 2 + HO 2 → products is a key step as pointed out 43,85 and the possible product channels of this reaction are (R13), (R29), and (R30).
Therefore, optimizing the mechanism of NH 3 conversion for high pressure needs to take into account the rate constant of H 2 NO-related reactions, which have been reported to be relevant. 85At present, we do not know the reason for the Energy & Fuels discrepancy between the experimental and calculation results, but this is an aspect that deserves further specific studies.
Figure 9 shows the experimental and simulation results of consumption and production of the species suffering from these instabilities (i.e., NH 3 , N 2 , and CO) with three different modifications of the current mechanism used.First, the mechanism of the present work using the reaction rates constants proposed by Klippenstein et al. 65 for reactions (R13), (R29), and (R30) without modifications (version 1 of the mechanism).Then, we have used two other versions of the mechanism: one using the rate constants of Sumathi et al. 86 for reactions (R13), (R29), and (R30) (version 2 of the mechanism), and the same mechanism of version 1, in which reaction (R30) has been multiplied by 10 (version 3 of the mechanism).With both versions 2 and 3 of the mechanism, instabilities disappear (Figure 9), and the changes made in version 3 resulted in a large overprediction of N 2 and CO 2 .Thus, we can state that the most critical reaction when it comes to produce instabilities is reaction (R30), where the rate value appears to be critical for the model performance.
The instability issue may also be attributed to thermochemical data of participating species, mainly HNO, or to the evolution of the H 2 NO species which has been identified as important in other ammonia studies. 54s seen in Figure 9, the best match of the experimental results and calculations made with the 3 versions of the mechanism is obtained with version 3. As has been mentioned, multiplying the rate constant of reaction (R30) by 10 acts to avoid the instabilities happening with the mechanism proposed in the present work, i.e., version 1, which includes without changes the rate constant proposed by Klippenstein et al. 65 Version 2 of the mechanism includes the rate constants of Sumathi et al. 86 for reactions (R13), (R29), and (R30), and also avoids instabilities, even though it largely overpredicts the conversion of NH 3 and production of N 2 and CO 2 .For the N 2 O production cases, none of the 3 versions solve the problem of instabilities.Additionally, it has to be noted that the recent work by Klippenstein and  Glarborg on the reaction rate of NH 2 + HO 2 is probably more accurate than the factor of 10 used in version 3 of the mechanism, as necessary to minimize instabilities, since that determination would exhibit an uncertainty lower than an order of magnitude.However, the use of the 3 modified versions of the mechanism does intend to show that the model still needs improvement, and more work on this is desirable.A possible way for improvement may rely on the formation of C/N species, such as methylamine or a higher amount of nitromethane, whose description may also need refinement.A hypothetical formation of PAH may also happen, in particular under fuel-rich conditions.
As discussed in previous work by our group, 3,51,53,58,69,71,72,87 it is important to bear in mind that the changing in pressure in our experimental system and procedure, implies the change of the residence time according to eq 2, as has been discussed and analyzed in earlier studies. 53,69,87In these studies, it was concluded that both pressure and residence time simultaneously affected the formation and concentration of products.Compared to the work of Marrodań et al., 69,87 in the present results, pressure has a major influence on the methane and ammonia conversion, as can be seen in Figures S6 and S7 of the Supporting Information.
Effect of the Oxygen Excess Ratio. Figure 10 includes the results of varying the oxygen excess ratio (λ = 0.7, 1 and 3) at 40

Energy & Fuels
bar for CH 4 /NH 3 = 1, to show the effect of the availability of O 2 availability.Experimental results show a clear sensitivity to oxygen availability.Reactant conversion is shifted to lower temperatures when the O 2 concentration is increased.Full conversion, where possible under the conditions studied, is also achieved at lower temperatures and higher oxygen concentrations.As the working pressure decreases, the decrease in reaction onset temperature is more pronounced, and the same effect occurs at lower CH 4 /NH 3 ratios.Switching from stoichiometric to oxidizing conditions, at 20 bar, the reaction onset temperature reduction is 125, 75, and 50 K, for CH 4 /NH 3 = 0.5, 1, and 2. Changing from λ = 0.7 to 1 produces only a shift (50 K) for CH 4 /NH 3 = 0.5 and 1.In the case of 40 bar and at CH 4 /NH 3 = 1 (Figure 10), a shift to a lower temperature of the onset of the CH 4 and NH 3 conversion reaction has been found (50 K, 875 → 825) when moving from reducing to oxidizing conditions.In previous work of our group, it was found that the conversion of pure ammonia happens at approximately 20−25 K 53 less under oxidizing conditions compared to stoichiometric conditions.However, when CH 4 is added, this difference increases, as seen in the present results.Methane is mainly consumed by H-abstraction-producing CH 3 radicals, which leads to the increase of the O/H radical pool.Also, the addition of methane leads to the production of CH 3 , CH 2 O, and HCO, species that promote ammonia consumption, as can be seen in the results of the sensitivity analysis of Figure 7.
The maximum peak of CO emissions is reached at lower temperatures and in lower concentrations with increasing excess oxygen ratio, which logically promotes full CO conversion; therefore, CO 2 production is shifted to lower temperatures.From an environmental point of view, it is remarkable that the increase in pressure leads to an increase in the maximum N 2 O concentration observed, as pointed out before, increasing it by a factor of more than 2 when switching from stoichiometric to oxidizing conditions.
One key consideration to bear in mind when discussing a deficiency or excess of oxygen is the production of HCN.In our experimental system, under the studied conditions, it has been found that HCN is only produced under reducing conditions.Under these conditions, the HCN concentration at the outlet is higher as the proportion of CH 4 is higher in the mixture.Figure 11 includes the results of HCN production varying the oxygen excess ratio (λ = 0.7 and 3) at 40 bar for CH 4 /NH 3 = 2 (the most favorable CH 4 /NH 3 ratio for HCN formation) and varying the CH 4 /NH 3 ratio under reducing conditions, to show the effect of both O 2 availability and CH 4 /NH 3 ratio.Results show that, for λ = 0.7, an appreciable formation of HCN happens during the interaction between NH 3 and CH 4 , in particular, above 1000 K when significant conversion of CH 4 occurs.The model underpredicts the concentration of HCN by a factor of 2, even though the main trends are reasonably well captured.However, no significant formation of HCN is found for fuel-lean conditions, λ= 3.
Effects of the CH 4 /NH 3 Ratio.From a practical point of view, it is interesting to evaluate different CH 4 /NH 3 ratios in order to assess the use of possible different mixtures containing both ammonia and methane.
The consumption and production of species during the oxidation of CH 4 /NH 3 mixtures are also measured as a function of the CH 4 /NH 3 ratio, at different oxygen excess ratios (λ = 1 and 3), at 40 bar as an example.To give an idea of the CH 4 addition effect, Figure 12 shows the consumption of CH 4 , NH 3 , and the production of N 2 O.As seen in Figure 12, for a given pressure and oxygen excess ratio, the addition of CH 4 always results in a decrease of the reaction onset temperature of both NH 3 and CH 4 oxidation.This effect is more noticeable for stoichiometric conditions than for oxidizing ones.This fact is of interest, from a practical point of view, because it appears that the addition of carbon combustibles helps to diminish the ignition temperature of ammonia.For the highest pressure studied, 40 bar, pure ammonia starts to convert at approximately 1165 K under the same experimental conditions and [NH 3 ] = 1000 ppm.In Figure 12, it can be seen that the onset temperature for NH 3 conversion decreases (875 K, 825 K, 825 K, at 40 bar and CH 4 /NH 3 = 0.5, 1, and 2, respectively) when NH 3 is oxidized in the presence of CH 4 , for all studied CH 4 / NH 3 ratios and stoichiometries.This is due to the increased production of OH radicals and CH 3 radicals, which will subsequently produce more OH radicals, as discussed above.On the other hand, the major disadvantage is the production of NO and N 2 O occurs at high CH 4 /NH 3 ratios.As seen above, the combustion of NH 3 and CH 4 under the studied conditions produces CH 3 ONO which decomposes into NO, which reacts to form NO 2 at high pressure, and this later produces N 2 O.According to the mechanism, under the present conditions, the increase in N 2 O was not found to come from the production of HCN oxidation, as found by other authors, 88,89 even though the specific operating conditions considered are different.Anyway, HCN production is not important under the conditions in which N 2 O is formed.In this sense, working with CH 4 /NH 3 ratios higher than 1 is not desirable because it does not offer great benefits compared to CH 4 /NH 3 ratios of 1 or 0.5, while it leads to a considerable increase of N 2 O emitted, which is a greenhouse gas with 273 times more global warming potential than that of CO 2 . 90Furthermore, a high CH 4 /NH 3 concentration leads to increased emissions of CO 2 .
Mass Balances.In order to evaluate the quality of our experiments and to determine if the measured species are dominant under the studied conditions, we decided to do nitrogen balances for the experiments performed.We can do that because we have used argon as a bath gas, allowing in this way the precise determination of N 2 as a product gas.Figure 13 shows, as an example, a nitrogen atom balance for different experimental conditions.The N balance is calculated by considering the nitrogen atoms of the following species: NH 3 , N 2 , HCN, NO, N 2 O, and NO 2 .Even though NO 2 has been accounted for in the nitrogen balance determined, their experiment profiles have not been shown in the figures because these species are around the uncertainty of the equipment measurements lower than 5 ppm in all cases.The N balance (in percentage) calculated with the model considering the same species mentioned above is also shown in Figure 13 as a continuous line.As seen, the calculated N balance is between 90 and 100%, while the experimental one closes between 90 and 105% along the whole temperature range.This indicates a reasonable agreement between species determined and calculated, even though a small mass of other species not analyzed experimentally may also be present.
Similarly, Figure 14 shows an example of a carbon atom balance for the same experiments shown in Figure 13.The C balance is calculated considering the carbon atoms of the following species: CH 4 , CO, HCN, and CO 2 .The C balance (in percentage) calculated with the model of the species mentioned above is also shown in Figure 14 as a line, and it is between 85 and 100%, while the experimental one closes between 90 and 105% for all studied temperatures.Results indicate a reasonable closing of the C balance as well.

■ CONCLUSIONS
An experimental and simulation study of the main features of the oxidation of mixtures of ammonia with methane at high pressure (from 10 to 40 bar), under reducing, stoichiometric, and oxidizing conditions in the 550−1250 K temperature interval, in a quartz tubular flow reactor with, roughly, 1000 ppm of inlet NH 3 and 500, 1000, and 2000 ppm of CH 4 , in NH 3 /CH 4 mixture, and using argon as a diluent, has been performed.
The main product of ammonia conversion is N 2 , followed by N 2 O, and under certain conditions NO (CH 4 /NH 3 = 1 and 3 at λ = 3) while the NO 2 concentration is negligible and below the uncertainty of the measurements under all conditions studied.The use of high pressure acts to favor the formation of N 2 and CO 2 from ammonia oxidation compared to what happens at atmospheric pressure.This is a positive outcome for the use of ammonia as a fuel in pressure applications such as turbines.However, the N 2 O concentration in the exhaust gases is significantly higher than that in pure NH 3 oxidation, which may be a drawback.
The onset of both NH 3 and CH 4 oxidation occurs at higher temperatures for reducing and stoichiometric conditions than for oxidizing conditions for all the pressures considered, indicating the importance of oxygen stoichiometry for ammonia conversion.In addition, working at higher oxygen excess ratios minimizes HCN production.
Compared to the oxidation of pure ammonia, the presence of CH 4 acts to shift ammonia conversion to lower temperatures, up to 300 K under certain conditions.
Pressure is seen to have an important influence on both the NH 3 and CH 4 oxidation regimes in the mixture, shifting them to lower temperatures as the pressure increases.However, the influence of pressure is seen to be significantly more important at low pressures compared to high pressures, and, at the same time, this influence is lower at higher CH 4 /NH 3 ratios.
The production of OH, CH 3 , and HCO species promotes the NH 3 and CH 4 consumption and, on the counterpart, the inhibition of NH 3 and CH 4 combustion is favored by reactions that produce HO 2 , H 2 O, and O 2 species, under the studied conditions.
The mechanism compiled from the literature and updated in the present work, used to carry out the simulations, is able to describe well the conversion of both NH 3 and CH 4 under almost all of the studied conditions.Nevertheless, discrepancies, mainly in the prediction of NH 3 , NO, N 2 , N 2 O, and CO 2 under certain   3, 6, 8, 11, 22, and 22R in Table 1.experimental conditions, have been observed, with higher discrepancies between the experimental and calculated results seen for N 2 O.The best agreement between experimental results and calculations is found for oxidizing conditions.

FFigure 6 .
Figure 6.Reaction path diagram for the consumption of NH 3 at CH 4 / NH 3 = 1 for reducing conditions (λ = 0.7) at 40 bar and stoichiometric conditions (λ = 1) at 10 and 40 bar (left) and, oxidizing conditions (λ = 3) at 10 and 40 bar (right).Black lines represent the common path for the different conditions and green lines show the additional path happening at 10 bar (left) and 40 bar (right).

Figure 13 .
Figure 13.Experimental and calculated N balance during the oxidation of NH 3 , as a function of the reactor temperature, for different oxygen excess ratios, compositions, and pressures.Species included in the balance are NH 3 , NO, NO 2 , N 2 O, HCN, and N 2 .Sets 1, 3, 6, 8, 11, 22, and 22R are shown in Table1.

Figure 14 .
Figure 14.Experimental and calculated C balance during the oxidation of CH 4 , as a function of the reactor temperature, for different oxygen excess ratios, compositions, and pressures.Species included in the balance are CH 4 , CO, HCN, and CO 2 .Sets 1,3, 6, 8, 11, 22, and 22R in Table1.

Table 1 .
Matrix of Experimental Conditions a and H 2 O, and the resulting chain reaction of NH 2 with NO to form N 2 and H 2 O and with HO 2 to form NH 3 and O 2 , together with the HO 2 recombination to form H 2 O 2 and O 2 , 43 represent the major inhibition reactions for the consumption of CH 4 .In a minor extent, CH 4 + O 2 ⇌ CH 3 + HO 2 (R10) and 2CH 3 (+M) ⇌ C 2 H 6 (+M) (R11) appear as inhibition reactions as well.In the case of NH 3 , under the same experimental conditions, its conversion is promoted by the reaction of CH 4 with radicals (NH 2 , OH, and HO 2 ) to form CH 3 radicals, by reactions of HO 2 with other species (NH 2 , NO and NH 3 ) to form OH radicals, by reaction of CH 3 radicals with O 2 to form also OH radicals, by reactions of CH 2 O with oxygenated species (OH, HO 2 and O 2 ) to produce HCO and by reaction of NH 2 radicals with NO 2 to form NO. As can be seen in the CH 4 case, the consumption of NH 3 is promoted by the production of OH, CH 3, and HCO species.In this case, for the inhibition of NH 3 consumption, we find reaction (R10) CH 4 + O 2 ⇌ CH 3 + HO 2 and the recombination reactions (R11) 2CH 3 (+M) ⇌ C 2 H 6 (+M) and (R12) 2HO 2 ⇌ H 2 O 2 and the NH 3 chain reaction of (R2) NH 3 + OH ⇌ NH 2 + H 2 O, followed by (R3) NH 2 + NO ⇌ N 2 + H 2 O or (R13) NH 2 + HO 2 ⇌ NH 3 + O 2 .As with CH 4 , the inhibition of ammonia combustion is favored by reactions that produce HO 2 , H 2 O, and O 2 species.It can be considered that as the main trends of species conversion are well captured by the model, the main reaction pathways are feasible.
Validation of kinetic mechanism; table of experimental onset reaction NH 3 and CH 4 temperatures; N 2 and NO species concentration figures; comparison of pressure and residence time effect; and N 2 O species concentration rescaled figure (PDF) María U. Alzueta − Department of Chemical and Environmental Engineering, Aragón Institute of Engineering Research (I3A), University of Zaragoza, 50018 Zaragoza, Spain; orcid.org/0000-0003-4679-5761;Email: uxue@ unizar.es