air mixture in microchannels II. Chemical analysis

Fuel 235 (2019) 923–932

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Heterogeneous reaction characteristics and its effects on homogeneous combustion of methane/air mixture in microchannels II. Chemical analysis

T

⁎

Yefeng Wanga, Weijuan Yanga, , Junhu Zhoua, Haolin Yangb, Yanyi Yaoa, Kefa Cena a b

State Key Laboratory of Clean Energy Utilization, Zhejiang University, Hangzhou 310027, Zhejiang Province, PR China Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou 510640, Guangdong Province, PR China

G R A P H I C A L A B S T R A C T

A R T I C LE I N FO

A B S T R A C T

Keywords: Methane Hetero/homogeneous reaction Chemical analysis Catalytic combustion

Catalytic combustion of methane/air mixture in various channels was numerically investigated to illuminate the chemical interaction mechanism between heterogeneous reactions (HTRs) and homogeneous reactions (HRs) with detailed gas-phase and surface reaction mechanisms. In the channels of different widths and with/without catalyst, the gaseous species, such as CH4, O2, CO, CO2, H, H2, and H2O and the reactions involving these species were monitored and the reaction rates along the streamwise direction were analyzed quantificationally. In the catalytic combustion, the HTR can be divided into prepare phase, weak HTR phase, violent HTR phase, second weak HTR phase, and completion phase according to the situation of gaseous species depletion on catalytic surface. The heat released by the homogeneous reaction and the consumption of O(s) can accelerate the absorption of CH4 on Pt surface in the violent HTR phase (phase III). Meanwhile, the stronger competition to CH4 of HR would inhibit the reactions of HTR involving CH4. The similar situation happens to O2, except the larger consumption of O2 by HTR than that of CH4 because of the larger stick coefficient of O2 on Pt surface. The additional reaction pathway of CO through HTR is favorable for the completeness of CO to CO2 during combustion. The consumption of H radical by HTR in the second weak HTR phase (phase IV) leads to the H2 release from Pt surface to the gas-phase and promotes the homogeneous reactions. The decrease of the width of the channel would suppress the HR intensity and enhance the HTR intensity. However, HR is still dominated in all the cases with a large amount of CH4 and O2 consumption and CO2 and H2O production.

⁎

Corresponding author. E-mail address: [email protected] (W. Yang).

https://doi.org/10.1016/j.fuel.2018.08.097 Received 15 June 2018; Received in revised form 19 August 2018; Accepted 23 August 2018 0016-2361/ © 2018 Published by Elsevier Ltd.

Fuel 235 (2019) 923–932

Y. Wang et al.

Fig. 1. . Schematic of micro channel.

1. Introduction

played the main role in catalytic combustion, and the existence of heterogeneous reaction can decrease the yield of byproduct and improve the fuel conversion completeness. By changing the ratio of the catalytically-active area to the geometrical channel surface area, Pizza et al. [22] obtained the flame stability diagrams of oscillating, asymmetric, and symmetric V-shaped flames in a planar channel of 1 mm height with Pt-coated walls under different inlet velocities and catalytic reactivities. Flame stability and heat transfer of methane/air mixtures in catalytic micro combustors was numerically investigated by Chen et al. [31]. Although the catalytic combustion was concerned in micro combustion field, it is still not clear in the understanding of the chemical interaction mechanism while the published work mainly focused on the interaction mechanism of hydrogen combustion, and the combustion overall characteristics, such as the stable limits, the combustion efficiency and so on. Exploring the chemical interaction mechanism in the catalytic combustion of methane can give guide to the practical use of catalytic combustion in microscale and control of the combustion intensity through catalyst. In this work, numerical investigation of catalytic combustion of methane/air mixture in varies channels is carried out to illuminate the characteristics of heterogeneous reaction and its effect on homogeneous reaction. Chemical analysis, involving the consumption and production of main species, such as CH4, O2, CO, CO2, and H2O by homogeneous and heterogeneous reaction is discussed in detail to acquire the chemical interaction mechanism between heterogeneous and homogeneous reaction.

Micro combustion has been investigated widely due to its low cost, high density, and being friendly to the environment [1]. However, problems such as the stability and efficiency maybe huge obstacles in the practical use because of the increasing heat loss and inadequate residence time for fuels and oxidizers in micro combustion. Catalytic combustion is an effective method to overcome these challenges above due to its more stable operation, low ignition temperature, and less heat loss [2–4]. Lots of researches of catalytic combustion in microscale have been carried out in recent decades [5–15]. Still, it is complex and not easy to understand because of the co-exist of homogeneous and heterogeneous reaction in catalytic combustion. The heat released by homogeneous and heterogeneous reactions may promote the process of each other by affecting the ignition, reaction, and quenching process. Meanwhile, the chemical interaction between heterogeneous and homogeneous reaction is also complex because there are reactants and intermediates are consumed or released by the two kinds of various reactions in different combustion stages, leading to the hard judgments of prohibition or enhancement effect for each other. To understand the chemical interaction between homogeneous and heterogeneous reaction, researchers have conducted lots of chemical analysis of catalytic combustion. Ciuparu et al. [16] investigated the fate of hydroxyls at the surface of γ-Al2O3 supported palladium oxide catalysts under lean methane combustion reaction conditions by using in situ diffuse reflectance-fourier transform infrared (DR-FTIR) spectroscopy. They found that hydroxyls was formatted and accumulated at the catalyst surface during the oxidation and the oxygen mobility of the support influenced the dynamic behavior of hydroxyls at the surface. OH* concentration in narrow channels was measured by OH-PLIF in methane/air combustion, and the low concentration of OH confirmed the radical quenching effect of wall [17]. Using situ spatially-resolved Raman measurements and planar laser induced fluorescence (LIF), Sui et al. [18] assessed the catalytic combustion processes of fuel-rich CH4/ O2/N2/CO2 mixtures on Pt and Pd at 5 bars. It was found that the higher H2 production of Rh had a profound impact on gaseous combustion and Pt suppressed homogeneous reaction. While it is not easy to illuminate the chemical interaction mechanism in catalytic combustion due to the limitation of the capability and accuracy of experimental measurements, the CFD method with appropriate models can predict the species concentration and temperature distribution in combustion [19–24]. Chen et al. [25,26] constructed contribution diagrams and made design recommendations by comparing the combustion performance under different operating conditions in a two-dimensional model. The radical flux inhibition on homogeneous ignition caused by heterogeneous reaction was reported in Appel et al.’s work [27]. The combustion products, like H2O has a suppression effect in H2/air catalytic combustion [6]. The suppression mechanism was elucidated by Bui et al. [28]. Additional OH radical can promote fuel conversion rate in catalytic combustion [29]. A numerical study on the catalytic combustion of H2/air in a planar micro combustor conducted by Lu et al. [30] showed that there are three stages in heterogeneous reaction depending on the relative intensity of heterogeneous and homogeneous reaction, while homogeneous reaction

2. Numerical models and simulation approach 2.1. Computing model and boundary conditions The schematic diagram of the micro channel is shown in Fig. 1 with length (l) of 15 mm and wall thickness (δ) of 0.5 mm. Simulations in the channels whose width (d) varying were conducted to analysis the scale effect in catalytic combustion. 316 stainless steel was selected as the wall material, and the inner surfaces of micro channel were covered with Pt foils with a surface site density of 2.72 × 10−8 kmol/m2. Using the commercial CFD code FLUENT [32], the mass, momentum, energy, and heat transfer was simulated in a two-dimensional (2D) simulation model. Knudsen number was less than the critical value of 0.001 in these cases. Thus, the continuous model and no-slip conditions were appropriate, and the Navier–Stokes equations were still applicable. The laminar model was used because of the low Reynolds number (Re = 33.08 in the channel of d = 2.80 mm based on a velocity of 0.3491 m/s). A first-order up-wind scheme is used to discretize the convective terms in steady mass, momentum, energy and species equations while a second-order central differencing schemes is utilized to discretize the diffusion terms. The pressure-velocity coupling is treated using SIMPLE algorithm. The specific heat of species was calculated using a piecewise polynomial fitting method. The specific heat of the mixture was calculated using the mixing law, and the mixture gas density was calculated using the ideal gas law. The thermal conductivity and viscosity were calculated as a mass-fraction-weighted average of all species, and the kinetic theory was selected for the mass 924

Fuel 235 (2019) 923–932

Y. Wang et al.

x < 15 mm, the inlet and rear parts of channel were omitted to make the simulation convenient. Therefore, the physic model was modified as shown in Fig. 3(b). The length of channel (l) and channel height (d) remained 15 mm and 2.62 mm, respectively. The inlet height (din) was set as 1.08 mm. There are two inlet walls with adiabatic boundary conditions aside the inlet. Fig. 4(a) and 3(b) show the contours of molar OH* concentration normalized by its maxima (XOH) in the channel obtained by the simulation and experiment, respectively. In the experiments, the flame was anchored in the channel near the inlet without pulsation or fluctuation. The molar OH* concentration decreased from the core reaction area to the wall and downstream in the simulation and experiment. Near the wall and downstream, because of the thermal and chemical quenching effects [17], XOH was low. Distributions of normalized molar OH* concentration at x = 1.215 mm are plotted in Fig. 4(c). XOH decreased monotonically as y increased from 0 mm to 1.31 mm (near the inner face) in the experiments. The simulation results are in good agreement with the PLIF data, especially near the inner face (y = 1.31 mm), indicating a rational accuracy of the model.

diffusivity calculation of mixture. Convergence of simulation was guaranteed based upon the residuals of all governing equations to be less than 10−6. The boundary condition of inlet was set as velocity-inlet and outflow was used at the of the channel. The methane/oxygen equivalence ratio (Φ) was 1.0, and a uniform velocity profile was used at the inlet with a velocity (vin) of 0.3491 m/s. The temperature of inlet mixture (Tin) was 423 K, whereas the temperature of outer surface (Tw) was fixed at 773 K. The discrete ordinates (DO) model was used to compute the radiation at inner surfaces, and the emissivity of Pt (ε) was 0.996[33]. The wall thermal conductivity was taken to be 15 W/(m·K). Adiabatic boundary conditions were used for the head and end faces of wall. Detailed gas-phase reaction mechanism, containing 16 species and 41 reversible elementary reactions [23,34,35] was used in the computation mode. Detailed surface chemistry mechanism for methane oxidation over Pt was taken from Deutschmann et al. [8]. It contains 24 irreversible elementary reactions with 9 gaseous and 11 surface species. To assure grid independence, a mesh sensitivity analysis was performed with 8000, 18,000, 24,000, and 40,000 grids. The temperature distribution at centerline and x = 1.34 mm in the 2.62-mm-width channel with different mesh densities are shown in Fig. 2(a) and (b). To save computational time and satisfy the precision requirement, the optimized grid with 24,000 cells was used in the computational domain. The locally refined mesh was used to describe the gas-surface reaction near the inner wall and inlet area.

3. Result and discussion 3.1. Key species of heterogeneous reaction Homogeneous and heterogeneous reactions (denotes as HR and HTR below, respectively) occur in catalytic combustion simultaneously with species exchange and heat transfer between the gas-phase and catalyst surface. According to the surface reaction mechanism (Appendix B), CH4, H, and O are absorbed irreversibly on the Pt surface; the desorption of CO2 on Pt surface is reversible; and the absorption and desorption of H2, O2, H2O, OH, and CO exist in the CH4/air catalytic combustion. Therefore, the concentration and reactions involve these species and radicals above is discussed in this paper. The depletion rate of gaseous species on catalytic surface in the channel of d = 2.10 mm is shown in Fig. 5. The depletion rate of gaseous species on catalytic surface is the difference between absorption reaction rate and desorption reaction rate of the same specie or radical in HTRs, and positive denotes the consumption of specie by HTR, while negative denotes production of specie. 5 different phases were used to describe the HTR process according to the situation of gaseous specie depletion on catalytic surface in Fig. 5: I-prepare phase (x = 0 ∼ 0.97 mm): the inlet gas is heated by the inner surface and hot gas downstream. Nearly, there is no reaction occurs in this phase, except the absorption of O2. And because of a higher sticking probability of O2 in comparison to that of CH4, the catalyst surface is covered primarily by oxygen [11,13,38]; II-weak HTR (x = 0.97 ∼ 1.57 mm): weak HTR occurs with the consumption of H2 and O2 and production of H2O on Pt surface;

2.2. Model validation The OH fluorescence signal in a micro channel is measured by a OHPLIF system; it was used to validate the accuracy of the simulation model. The ultraviolet laser radiation at 283.2 nm was produced by a frequency-doubled Nd: YAG-pumped dye laser system and to excite the OH molecules. The details of OH-PLIF system are reported in [10]. After a calibration the OH fluorescence signal was proportional to the molar concentration of species, and the error was within 10% [36,37]. The micro-channel burner is shown in Fig. 3(a). Two 316 stainless steel plates were placed parallel to each other with a 2.62-mm-wide gap. Pt foils of 0.1 mm thickness (Shanghai Yueci Electronic Technology) were stuck on the inner faces of steel as the catalyst. The premixed methane/ air mixture issued from a 1.08-mm-width channel was introduced between two plates and ignited. The velocity and equivalence ratio of inlet mixture were 0.75 m/s and 1.0, respectively. Si3N4 heaters were used to heat the plates from backside. Two K-type thermocouples were inserted into the small holes of stainless steel plates to monitor the wall temperature (Tw). A K-type thermocouple was placed near the inlet to monitor Tin (the temperature of inlet gas, not shown here). Tw and Tin were set as 773 K and 423 K, respectively. Because the flame was always located and combustion almost completed in the front region of

T (K)

1400 1200 1000 800 600 400

T (K)

1600

1900 1800 1700 1600 1500 1400

8,000 nodes 18,000 nodes 24,000 nodes 40,000 nodes

1950 1800 1650

(b)

1500 0.8 1.6 2.4 3.2 4.0

x (mm)

T (K)

1800

1350 1200 1050

(a)

900

8,000 nodes 18,000 nodes 24,000 nodes 40,000 nodes

750 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 y (mm)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

x (mm)

Fig. 2. . Distribution of temperature at centerline (a) and x = 1.34 mm (b) in 2.62-mm-width channel with different mesh densities. 925

Fuel 235 (2019) 923–932

Y. Wang et al.

-4

1.2 1.0 0.8

num. exp.

0.6 0.4 0.2

(c)

0.0 -0.2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3

y (mm) *

Fig. 4. . Contours of normalized OH molar concentration: (a) simulation result, (b) experimental result and distributions of normalized molar OH* concentration at x = 1.215 mm.

5.6x10 -4 4.8x10 -4 4.0x10 -4 3.2x10 -4 2.4x10 -4 1.6x10 -5 8.0x10 0.0

I II III IV

V

100 80

O2 O 100 O(s)

60 40 20 0

-4

4.50x10 -4 3.00x10 -4 1.50x10 0.00 -4 -1.50x10 -4 -3.00x10 -4 -4.50x10

surface coverage (%)

(b) 5.2 4.8 4.4 4.0 wall wall 3.6 3.2 2.8 2.4 2.0 1.6 1.2 0.8 -1.8-1.2-0.6 0.0 0.6 1.2 1.8 y (mm)

The depletioon rate of gaseous species on catalytic surface (kmol/(m2 s))

x (mm)

(a) 5.2 4.8 4.4 4.0 wall wall 3.6 3.2 2.8 2.4 2.0 1.6 1.2 0.8 -1.8-1.2-0.6 0.0 0.6 1.2 1.8 y (mm)

Normalized molar OH* FonFenWraWion, ȤOH

x (mm)

Fig. 3. . Schematic of the experimental setup for OH-PLIF examination of a methane/air mixture flame in a micro channel.

CO CO2 CH4

-4

4.50x10 -4 3.00x10 -4 1.50x10 0.00 -4 -1.50x10 -4 -3.00x10 -4 -4.50x10 0.0

H2 H H2O OH 10

3.0

6.0

9.0

12.0 15.0

x (mm)

III-violent HTR (1.57 ∼ 2.78 mm): HTR turns violent, consumes a large amounts of CH4, O2, H2 and CO with the reaction rate ∼ 10-4 kmol/(m2s-1), and produces H2O and CO2. Meanwhile, the O(s) on Pt surface decrease rapidly. IV-second weak HTR(2.78 ∼ 7.5 mm)-HTR turns weak with the consumption of H, OH, O, and O2 and production of H2 and H2O in low reaction rates. O and OH is consumed in a relatively low rate(∼10-6 kmol/(m2s-1)). V-completion phase(7.5 ∼ 15 mm): the depletion rate of gaseous species on catalytic surface is close to 0 and the combustion is almost complete. As we can see, the peaks of consumption rate of CH4, O2, CO and the consumption rate of H2 share the same location with the peaks of production rate of CO2 and production rate of H2O, respectively, which shows that the HTRs after the species and radicals absorption are really quick comparing to the incoming gas velocity. Fig. 6 shows the molar concentration distribution near the inner surface (data acquired from the imaginary plane between the solid and fluid zone in FLUNT) in the channel of d = 2.10 mm. The molar

Fig. 5. . The depletion rate of gaseous species on catalytic surface in the channel of d = 2.10 mm. Positive: consumption of gaseous species; negative: production of gaseous species.

concentration of CH4 and O2 drops sharply and the molar concentration of CO2 and H2O rises obviously in phase III because of the violent HTRs. The peaks of CO and H2 concentration near the inner surface show up at the boundary of phases III and IV, located more downstream than the positions where their depletion rates reach peaks at x = 2.52 mm in Fig. 5. Thus, it can be speculated that, before CO and H2 are consumed by the Pt catalyst, there are species diffusions not only from gas-phase to inner surface, but also diffusions towards upstream. In addition, O, H, and OH absorb on the catalyst surface easily due to their stick coefficients of almost 1.0, leading to low concentrations near the inner surface in Fig. 6. In HTR, O(s) is consumed in the adsorption of CO, conversion of CO to CO2, adsorption of H2, H and conversion of H2, H to H2O. Fig. 5 926

Fuel 235 (2019) 923–932

Y. Wang et al.

molar concentration (kmol/(m3 s))

-3

4.0x10-3 3.5x10-3 3.0x10-3 2.5x10-3 2.0x10-3 1.5x10-3 1.0x10-4 5.0x10 0.0-4 3.5x10-4 3.0x10-4 2.5x10-4 2.0x10-4 1.5x10-4 1.0x10-5 5.0x10 0.0

heterogeneous reaction is enhanced. Thus, the intensity of heterogeneous and homogeneous reactions could be controlled with the change of the combustor scale in methane/air combustion. Comparing to the consumption of CH4 and O2, HTR takes a higher proportion of the production of CO2 and H2O, which is attributed to the consumption of other free radicals on Pt surface, and this will be discussed in detail in Section 3.3. Overall, relative reaction intensities of the HTRs are low, with a maximum of 23.47% in the channel of d = 2.10 mm. Thus, HR is dominated in the catalytic combustion of methane/air combustion in these 5 cases.

O2 CO2 CH4 H2O O CO H2 H OH

3.3. Effect of catalyst

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 x (mm)

Because the effect of HTR is most obvious in the channel of d = 2.10 mm, the numerical simulations of catalytic combustion and non-catalytic combustion in the channel of d = 2.10 mm were carried out to explore the principles of the depletion and production of gaseous species and radicals on the Pt surface. The detailed situation of the depletion and production of these species in the 5 phases is shown and the cause of the behavior of heter/homogeneous reactions is analyzed based on the gas-phase and surface chemistry mechanism in this part. In the simulation of non-catalytic combustion, the “wall surface reactions” is closed in Fluent software and there are no surface reaction on the inner surface. For convenience, the combustion with Pt catalyst is represented by ‘cata’, while combustion without Pt catalyst is represented by ‘none’ below.

Fig. 6. . The concentration distribution near the inner surface in the channel of d = 2.10 mm.

shows that fast consumption of O(s) and O2 by HTR share the same position. And after O(s) being nearly 0, the consumption of O2 on the Pt surface continues. Considering the depletion of other species and radicals, we can conclude that the intensity of O2 consumption by HTR can be used to characterize the intensity of HTR. 3.2. Scale effect The reactions involving reactants CH4, O2 and products CO2, H2O is analyzed in this section to evaluates the intensity of the homogeneous/ hetero-phasic reactions within the channels of different widths. In the channel of d = 2.00 mm, no violent HR occurs. Once d ≥ 2.10 mm, there is obvious HR in the channel, and co-exist and effect on each other of HTR and HR would show up. The relative reaction intensity of HTR of specie i (wi) is defined as

3.3.1. Effect on CH4 The conversion of CH4 in ‘cata’ is shown in Fig. 7. The conversion of CH4 is calculated by

r = (1−nonsite / ninlet ) × 100% where r is the conversion of CH4, nonsite is the onsite mole of CH4 in the gas-phase, and ninlet is the mole of CH4 at the inlet. Before the intensive consumption of CH4 in the gas-phase (x < 2.0 mm), the conversion of CH4 rises to almost 30%, which is caused by the absorption of CH4 on Pt surface (CH4 + 2Pt = CH3(s) + H(s)) and weak HR in the gas-phase. The same phenomenon also shows up in the ‘cata’ combustion in the channel of d = 2.00 mm, in which no obvious HR occurs. Thus, the conversion curves of CH4 in the channels of d = 2.00 and 2.10 mm is almost overlapping in Fig. 7 before the point (1.36, 26.93%), defined as ‘effect point’. We define the largest depletion rate of gaseous CH4 by HR as the homogeneous ignition point, shown in Fig. 8. The ‘effect point’ is upstream than the homogeneous ignition point. Therefore, upstream the effect point, the depletion of CH4 in catalytic combustion is independent with homogeneous ignition; downstream the effect point, CH4 reacts quickly in the channel of d = 2.10 mm, and the consumption curve of CH4 by HTR in Fig. 8 reaches its peak. In the channel of d = 2.00 mm, the depletion rate of CH4 by HTR is almost constant (around 4 × 107 kmol/(m2 s)). Thus, it can be speculated that violent HR in the channel of d = 2.10 mm promotes the CH4 absorption on Pt surface. Two main reasons cause this: on one hand, the prodigious amounts of heat released by HR would accelerate the absorption

wi = ri, HTR/(ri, HTR + ri, HR) where ri,HTR is the depletion rate of i by HTRs, ri,HR is the depletion rate of i by HRs. As d decreases, the input molar flow rates of reactants decrease with the same inlet velocity (v = 0.75 m/s). To exclude the influence of the decreasing of reactants while d decreasing, we calculated ri,HR/d and rHTR/d, the results are shown in Table 1. The decreasing ri,HR/d with the decreasing d indicates that the decrease of the width of the channel would suppress the HR intensity. Conversely, rCH4,HTR/d and rO2,HTR/d increase with d decreasing, indicating that HTR is enhanced in a narrower channel. It can be conclude that the effect of HTR in narrower channels becomes more obvious when the relative reaction intensities in HTR (wCH4 and wO2) decrease monotonously with d decreasing. This is the reason we analyze the combustion in the channel of d = 2.10 mm in Sec. 3.1. In the channel of d = 2.20 mm, rCH4,HTR and rO2,HTR show the minimum because of the co-effect of decreasing reactants and the enhancement to HTR. Table 2 shows the rHR/d, rHTR/d, and w of CO2 and H2O. Similarly, with d decreasing, homogeneous reaction is suppressed while

Table 1 . The ratio of the consume reaction rate to d and relative reaction intensities of HTR of CH4 and O2. d (mm)

2.10

2.20

2.40

2.62

2.80

CH4

rHR/d (kmol/(s·mm)) rHTR/d (kmol/(s·mm)) w (%)

9.43E−07 1.14E−08 1.19

9.50E−07 6.45E−09 0.67

9.50E−07 6.21E−09 0.65

9.50E−07 5.69E−09 0.59

9.50E−07 5.07E−09 0.53

O2

rHR/d (kmol/(s·mm)) rHTR/d (kmol/(s·mm)) w (%)

1.72E−06 1.97E−07 10.29

1.76E−06 1.55E−07 8.08

1.85E−06 1.46E−07 7.31

1.78E−06 1.35E−07 7.04

1.79E−06 1.25E−07 6.54

927

Fuel 235 (2019) 923–932

Y. Wang et al.

Table 2 . The ratio of the production reaction rate to d and relative reaction intensities of HTR of CO2 and H2O. d (mm)

2.10

2.20

2.40

2.62

2.80

CO2

rHR/d (kmol/(s·mm)) rHTR/d (kmol/(s·mm)) w (%)

8.24E−07 1.26E−07 13.23

8.68E−07 8.36E−08 8.79

8.75E−07 7.75E−08 8.13

8.82E−07 6.64E−08 7.00

8.89E−07 5.71E−08 6.04

H2O

rHR/d (kmol/(s·mm)) rHTR/d (kmol/(s·mm)) w (%)

8.90E−07 2.72E−07 23.47

9.27E−07 2.33E−07 20.09

9.42E−07 2.20E−07 18.96

9.54E−07 2.10E−07 18.07

9.64E−07 1.99E−07 17.15

100

(34) CH4 + O = CH3 + OH

r (%)

80

(36) CH4 + OH = CH3 + H2O

d = 2.00 mm d = 2.10 mm

60

The reaction rates of R(33) and R(34) are lower in ‘cata’ than that in ‘none’, and the reaction rate of R(36) is higher in ‘cata’ because of the higher concentration of OH in ‘cata’, which indicates that the exist of Pt catalyst would change the reaction pathway and lead to a series of changes.

40 20

effect point

0 0.0

3.0

6.0 9.0 x (mm)

12.0

3.3.2. Effect on O2 As Fig. 9 shows, there is a quick consumption of O2 in the phase III of ‘cata’ case. 89.71% O2 is consumed by HR while another 10.29% is consumed by HTR. Similar to the consumption of CH4 by HTR and HR, the peak of the O2 reaction rate curve of HTR in ‘cata’ locates more downstream than that of HR, with the competition to O2 between HTR and HR. The ratio of the whole CH4 reaction rate to O2 reaction rate by HTR is below 0.06 in the channel of d = 2.10 mm in the ‘cata’ case, much lower than the value of 0.5, the ratio of mole CH4 to that of O2 at equivalent ratio. Thus, the competition to CH4 of HTR is more disadvantage that the competition to O2. This is caused by the different stick coefficient of CH4 and O2 on Pt- the stick coefficient of CH4 and O2 on Pt sites is 0.01 and 0.07 [38], leading to an easier absorption of O2. And this is also one of the reasons why CH4 is only 1.19% and O2 is 10.29% in Section 3.2.

15.0

The depletion rate of CH4 (kmol/(m2 s))

Fig. 7. . The conversion of CH4 in ‘cata’ in the channel of d = 2.00 and 2.10 mm along x.

I 6.0x10

-3

5.0x10

-3

4.0x10

-3

3.0x10

-3

2.0x10

-3

1.0x10

-3

IV

III

II

homogeneous ignition

HR in cata HTR in cata

3.3.3. Effect on CO and CO2 As Fig. 10 shows, in ‘cata’ and ‘none’ case, there are high-speed-rate production of CO in the channel after the homogeneous ignition. Then, the CO is consumed by HR and HTR together and CO2 is produced. The HR element reactions R(15) HCO + H = H2 + CO, (18) HCO + O2 = HO2 + CO, (19) HCO + M = H + CO + M are responsible for the production of CO. Although the reaction rate of CO

HR in none

0.0 0.0

1.0

2.0

3.0

4.0

5.0

The depletion rate of O2 (kmol/(m2 s))

x (mm) Fig. 8. . The depletion rate of CH4 by HTR and HR in ‘cata’ and ‘none’ in the channel of d = 2.10 mm along x.

reaction rate in a higher temperature; on the other hand, the intermediate production produced by HR, such as CO, H etc. would react with the O(s) on Pt surface, resulting in the decrease of the surface coverage of O(s) and more empty Pt active sites for CH4 to be adsorbed. At x = 2.5268 mm where the depletion rate of CH4 by HTR reaches its peak, the gaseous CH4 is almost consumed by HR (r = 97.70%). So there is not adequate CH4 for HTR and the peak of the CH4 consumption curve by HTR is small, presenting as only 1.19% CH4 is consumed by HTR. So, HR also shows inhibition effect on the CH4 consumption by HTR because of the more competitive power on CH4 consumption than that of HR. The main reactions of CH4 consumption in HR are:

9.0x10

-3

8.0x10

-3

7.0x10

-3

6.0x10

-3

5.0x10

-3

4.0x10

-3

3.0x10

-3

2.0x10

-3

1.0x10

-3

0.0 0.0

II

I

IV

III

HR in cata HTR in cata HR in none

1.0

2.0

3.0

4.0

5.0

x (mm) Fig. 9. . The depletion rate of O2 by HTR and HR in ‘cata’ and ‘none’ in the channel of d = 2.10 mm along x.

(33) CH4 + H = CH3 + H2 928

Fuel 235 (2019) 923–932

Y. Wang et al.

IV

V

-3

2.5x10 -3 2.0x10 -3 1.5x10 -3 1.0x10 -4 5.0x10 0.0

-3

0.0

The depeltion rate of H, H2, and H2O (kmol/(m2 s))

The depletion rate of CO and CO2 (kmol/(m2 s))

I II III 1.0x10

-3

-1.0x10

HR in cata HTR in cata HR in none

-3

-2.0x10

-3

-3.0x10

-3

-4.0x10

(a) CO

-3

-5.0x10 0.0 -4

-5.0x10

HR in cata HTR in cata HR in none

-3

-1.0x10

-3

-1.5x10

-3

-2.0x10

(b) CO2

-3

-2.5x10

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

x (mm) Fig. 10. . The depletion rate of CO(a) and CO2(b) by HTR and HR in the channel of d = 2.10 mm ‘cata’ and ‘none’ along x.

desorption on Pt surface in HTR are obvious, the reaction rate of CO absorption is higher than that of CO desorption. Thus, the Pt surface only consumes gaseous CO, rather than releases CO to the gas-phase. The main reaction of CO consumption in HR is R(13) CO + OH = CO2 + H. In the ‘cata’ case, 12.18% CO produced by HR is depleted by the HTR, and the absorption production- CO(s) is converted to CO2(s) and then the CO2 desorption occurs. The position of CO starting to be produced in the ‘cata’ case locates more downstream than that of the ‘none’ case, while the position of CO being completely consumed (the reaction rate of CO decreases to 0) is almost the same in the ‘cata’ and ‘none’ cases. The similar phenomena happen to the consumption reactions of CO2. In general, the physical process of the conversion of CO to CO2 is shorten in the ‘cata’ case because of the additional reaction pathway on the Pt catalyst surface. Also, the additional reaction pathway is beneficial to the completeness of the combustion, leading to lower molar CO concentration and higher CO2 concentration at outlet.

I II III

IV

V

HR in cata HTR in cata HR in none

-4

-5.0x10 -3 -1.0x10 -3 -1.5x10-3 1.0x10

(a) H

0.0 -1.0x10

-3

-2.0x10

-3

-3.0x10

-3

HR in cata HTR in cata HR in none (b) H2

-3

-4.0x10 0.0 -1.0x10

-3

-2.0x10

-3

-3.0x10

-3

-4.0x10

-3

-5.0x10

-3

HR in cata HTR in cata HR in none (c) H2O

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

x (mm)

molar concentration of H (kmol/m3)

Fig. 11. . The depletion rate of H(a), H2(b) and H2O(c) by HTR and HR in in the channel of d = 2.10 mm ‘cata’ and ‘none’ along x.

3.3.4. Effect on H, H2 and H2O The depletion rate of H, H2 and H2O by HTR and HR in the channel of d = 2.10 mm ‘cata’ and ‘none’ along x is shown in Fig. 11. The H radical involves in the mixture flow, diffusion, HR in gas-phase and absorption/desorption on Pt surface. Fig. 12 shows the molar concentration of H radicals at centerline and inner surface, and the arrows point to the position where H is consumed most quickly in ‘none’ and ‘cata’ case by HR. It is obvious that the arrows positions of H radicals locate more upstream than the positions of the highest molar concentration of H radicals at centerline. Considering the inlet gas only consists of CH4, O2, and N2, the H radical consumed at the arrows position in Fig. 12 is diffused from the downstream area. For example, in ‘none’ case in the channel of d = 2.10 mm, the H radical is produced between x = 0.58–1.29 mm in the gas-phase; almost 56.64% of the H radical diffuses upstream and participates in HRs, and the main reactions are R(1)H + O2 = OH + O and R(33)CH4 + H = CH3 + H2; the another 43.36% H radical is mainly consumed by R(7) H + O2 + M = HO2 + M and R(8) HO2 + H = OH + OH between x = 1.36–5.20 mm. In ‘cata’ case, the amount of H radicals involving in diffusion toward upstream accounts for 92.56% of the production of H radical in gas-phase. Different to the situation in ‘none’ case, there is no another H radical consumption in HR downstream the production area. Instead, the another 7.44% H radical is consumed by HTR on Pt surface.

5.0x10

-5

4.0x10

-5

3.0x10

-5

2.0x10

-5

1.0x10

-5

0.0 0.0

centerline in cata inner surface in cata centerline in none inner surface in none

1.0

2.0

3.0

4.0 5.0 x (mm)

6.0

7.0

8.0

Fig. 12. . Molar concentration of H radical on the centerline and near the inner surface.

In ‘none’ case, the quick consumption of H radical in HR around the centerline results in the lower molar concentration of H radicals at centerline than that near the inner surface. In contrast, the molar concentration of H radical near the inner surface in the ‘cata’ case is lower than that at inner surface due to the quick absorption of H radicals on Pt surface. As for H2, there are two peaks in the curve of H2 depletion rate by HR in ‘cata’ case, locating in phase III and phase IV, respectively. The first peak overlaps with the peak in the curve of H2 depletion rate by HTR. As for the ‘none’ case, there is only one peak in the curve of H2 depletion rate by HR. Thus, the second peak of H2 depletion rate by HR in ‘cata’ case is caused by HTR: HTR produces H2 between x = 2.7958–4.6317 mm and then the H2 is converted to H2O through HR. The H2 release by HTR is attributed to the high concentration of H radical in the gas-phase. The gaseous H radical is absorbed by Pt surface 929

Fuel 235 (2019) 923–932

Y. Wang et al.

overwhelming proportion of CH4 and O2 consumption and CO2 and H2O production. It will be interesting to study the optimal size of the combustor and control the contribution of HR and HTR to acquire the needed combustion characteristic. (3) HR is beneficial to the CH4 depletion by HTR which may be caused by the heat released by the homogeneous reaction and the consumption of O(s) is beneficial to accelerate the absorption of CH4 on Pt surface. Meanwhile, the stronger competition to CH4 of HR would inhibit the reactions of HTR involving CH4 in phase III. The similar situation happens to O2, except the larger consumption of O2 by HTR than that of CH4 because of the larger stick coefficient of O2 on Pt surface. (4) In phase III, HTR consumes 12.18% gaseous CO. Compared to the ‘none’ case, the additional reaction pathway through which CO is converted to CO2 on Pt surface, is favorable for the completion of CO to CO2 in catalytic combustion, leading to lower molar CO concentration and higher CO2 concentration at outlet. In ‘none’ case, the gaseous H radical would diffuse upstream and downstream and then participates in HRs, while in ‘cata’ case, the H consumption downstream is replaced by the HTR in phase IV, leading to the H2 release from Pt surface to the gas-phase and promotes the homogeneous reactions in return. The production of H2O by HR and HTR happens during phases II, III, and IV, and the combustion intensity can be judged according to the generating rate of H2O.

and converted to H2 through the element reaction 2H(s)= > H2 + Pt (s). H2 is able to assist the combustion of methane/air with the thermal and chemical effects [3,39–41]. Therefore, HTR promotes the combustion from the perspective of H2. In the ‘cata’ case, the production of H2O in HR and HTR happens during phases II, III, and IV. That is to say, H2O is always generated once the methane/air combustion proceeding. So, the combustion intensity can be judged according to the generating rate of H2O. 4. Conclusion A 2D simulation of methane/air mixture combustion was used in micro channels to illuminate the chemical interaction mechanism between HTR and HR. The conclusions are as follows: (1) The HTR can be divided into 5 phases according to the situation of gaseous species depletion by HTR: I-prepare phase: the inlet gas is heated by the inner surface and hot gas downstream with the almost fully coverage of O2 on Pt sites; II-weak HTR: weak HTR occurs with the consumption of H2 and O2 and production of H2O; IIIviolent HTR: HTR turns violent, consumes a large amount of CH4, O2, H2 and CO, and produces H2O and CO2. Meanwhile, the O(s) on Pt surface decrease rapidly; IV-second weak HTR: HTR turns weak with the consumption of H, OH, O, and O2 and production of H2 and H2O in low reaction rates; V-completion phase: the combustion is almost completed. (2) As d decrease, ri,HR/d decreases while ri,HTR/d increases and increases. Thus, the decrease of the width of the channel would suppress HR and enhance HTR. However, HR is dominated in the catalytic combustion of methane/air combustion with an

Acknowledgments The work is supported by the National Natural Science Foundation of China [51336010].

Appendix A Skeletal mechanism for methane oxidation by ANSYS Fluent (16 species and 41 reactions). Reaction mechanism rate coefficients in the form kf = ATβexp(−E0/RT). Units are moles, seconds, Kelvins and calories/mole. species H2O CO2 O2 CH4 CO H H2 OH O CH3 HCO HO2 H2O2 CH2O CH3O N2 Reactions Ak

βk

Ek

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

−0.927 2.7 1.51 1.4 −1 −2 −0.8 0 0 0 0 0 1.3 0 0 0 0 0 −1 1.62 0 2.46 0 7.4 0 0

16,874 6262 3430 −397 0 0 0 1004 693 0 0 45,411 −758 3011 0 0 0 0 16,993 2175 3513 −970 39,914 −956 76,482 0

H + O2 = OH + O O + H2 = OH + H OH + H2 = H2O + H OH + OH = O + H2O H + H + M = H2 + M H + OH + M = H2O + M H + O2 + M = HO2 + M HO2 + H = OH + OH HO2 + H = H2 + O2 HO2 + O = O2 + OH HO2 + OH = H2O + O2 H2O2 + M = OH + OH + M CO + OH = CO2 + H CO + O + M = CO2 + M HCO + H = H2 + CO HCO + O = OH + CO HCO + OH = H2O + CO HCO + O2 = HO2 + CO HCO + M = H + CO + M CH2O + H = HCO + H2 CH2O + O = HCO + OH CH2O + OH = HCO + H2O CH2O + O2 = HCO + HO2 CH2O + CH3 = HCO + CH4 CH2O + M = HCO + H + M CH3 + O = CH2O + H

1.59E+17 3.87E+04 2.16E+08 2.10E+08 6.40E+17 8.40E+21 7.00E+17 1.50E+14 2.50E+13 2.00E+13 6.02E+13 1.00E+17 1.51E+07 3.01E+14 7.23E+13 3.00E+13 1.00E+14 4.20E+12 1.86E+17 1.26E+08 3.50E+13 7.23E+05 1.00E+14 8.91E−13 5.00E+16 8.43E+13 930

Fuel 235 (2019) 923–932

Y. Wang et al.

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

CH3 + OH = CH2O + H2 CH3 + O2 = CH3O + O CH3 + O2 = CH2O + OH CH3 + HO2 = CH3O + OH CH3 + HCO = CH4 + CO CH4(+M)= CH3 + H(+M) CH4 + H = CH3 + H2 CH4 + O = CH3 + OH CH4 + O2 = CH3 + HO2 CH4 + OH = CH3 + H2O CH4 + HO2 = CH3 + H2O2 CH3O + H = CH2O + H2 CH3O + OH = CH2O + H2O CH3O + O2 = CH2O + HO2 CH3O + M = CH2O + H + M

8.00E+12 4.30E+13 5.20E+13 2.28E+13 3.20E+11 6.30E+14 7.80E+06 1.90E+09 5.60E+12 1.50E+06 4.60E+12 2.00E+13 5.00E+12 4.28E−13 1.00E+14

0 0 0 0 0.5 0 2.11 1.44 0 2.13 0 0 0 7.600–3528.0 0

0 30,808 34,895 0 0 104,000 7744 8676 55,999 2438 17,997 0 0 25,096

Appendix B Skeletal mechanism for CH4-O2 surface mechanism on Pt methane oxidation by ANSYS Fluent. Reaction mechanism rate coefficients in the form kf = ATβexp(−E0/RT). Units are moles, seconds, Kelvins and joules/mole. SITE/PT_SURFACE/ SDEN/2.72E-9/ Pt(S) H(S) H2O(S) OH(S) CO(S) CO2(S) CH3(S) CH2(S) CH(S) C(S) O(S) Reactions

Ak

βk

Ek

1

4.46E+10

0.5

0

3.70E+21

0

67,400

1

0

0

1.80E+21

−0.5

0

0.023

0

0

3.70E+21

0

213,200

1

0

0

0.75

0

0

1.00E+13 1

0 0

40,300 0

1.00E+13 3.70E+21 3.70E+21 3.70E+21 1.62E+20

0 0 0 0 0.5

192,800 11,500 17,400 48,200 0

1.00E+13 1.00E+13 3.70E+21 4.63E+20

0 0 0 0.5

125,500 20,500 105,000 0

3.70E+21 3.70E+21 3.70E+21 3.70E+21 1.00E+18

0 0 0 0 0

20,000 20,000 20,000 62,800 184,000

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

H2 + 2Pt(s) = > 2H(s) FORD/Pt(s) 1/ 2H(s) = > H2 + 2Pt(s) COV/H(s) 0 0–6000.0/ H + Pt(s) = > H(s) STICK O2 + 2Pt(s) = > 2O(s) DUPLICATE O2 + 2Pt(s) = > 2O(s) DUPLICATE STICK 2O(s) = > O2 + 2Pt(s) COV/O(s) 0.0 0.0–60000.0/ O + Pt(s) = > O(s) STICK H2O + Pt(s) = > H2O(s) STICK H2O(s) = > H2O + Pt(s) OH + Pt(s) = > OH(s) STICK OH(s) = > OH + Pt(s) H(s) + O(s) = OH(s) + Pt(s) H(s) + OH(s) = H2O(s) + Pt(s) OH(s) + OH(s) = H2O(s) + O(s) CO + Pt(s) = > CO(s) FORD/Pt(s) 2/ CO(s) = > CO + Pt(s) CO2(s) = > CO2 + Pt(s) CO(s) + O(s) = > CO2(s) + Pt(s) CH4 + 2Pt(s) = > CH3(s) + H(s) FORD/Pt(s) 2.3/ CH3(s) + Pt(s) = > CH2(s) + H(s) CH2(s) + Pt(s) = > CH(s) + H(s) CH(s) + Pt(s) = > C(s) + H(s) C(s) + O(s) = > CO(s) + Pt(s) CO(s) + Pt(s) = > C(s) + O(s)

research. Prog Energy Combust Sci 2011;6:669–715. [2] Wang Y, Zhou Z, Yang W, Zhou J, Liu J, Wang Z, et al. Combustion of hydrogen-air in micro combustors with catalytic Pt layer. Energy Convers Manage 2010;6:1127–33.

References [1] Ju Y, Maruta K. Microscale combustion: technology development and fundamental

931

Fuel 235 (2019) 923–932

Y. Wang et al.

[21] Karagiannidis S. Experimental and numerical investigation of the hetero-/homogeneous combustion of lean propane/air mixtures over platinum. Proc Combust Inst 2009;2:1947–55. [22] Pizza G, Mantzaras J, Frouzakis CE, Tomboulides AG, Boulouchos K. Suppression of combustion instabilities of premixed hydrogen/air flames in microchannels using heterogeneous reactions. Proc Combust Inst 2009;2:3051–8. [23] Turkeli-Ramadan Z, Sharma RN, Raine RR. Two-dimensional simulation of premixed laminar flame at microscale. Chem Eng Sci 2015:414–31. [24] Yan Y, Huang W, Tang W, Zhang L, Li L, Ran J, et al. Numerical study on catalytic combustion and extinction characteristics of pre-mixed methane–air in micro flatbed channel under different parameters of operation and wall. Fuel 2016:659–67. [25] Chen J, Xu D, Song W. Optimal combustor dimensions for the catalytic combustion of methane-air mixtures in micro-channels. Energy Convers Manage 2017:197–207. [26] Chen J, Gao X, Liu B, Xu D. Management of the gas-phase and surface chemistry in methane-fueled catalytic micro-combustors. Int J Hydrogen Energy 2017. [27] Appel C, Mantzaras J, Schaeren R, Bombach R, Inauen A, Kaeppeli B, et al. An experimental and numerical investigation of homogeneous ignition in catalytically stabilized combustion of hydrogen/air mixtures over platinum. Proc Combust Inst 2002;1(2002):1031–8. [28] Bui PA, Vlachos DG, Westmoreland PR. Homogeneous ignition of hydrogen-air mixtures over platinum. Symposium on Combustion 1996;1:1763–70. [29] Lu Q, Pan J, Yang W, Pan Z, Tang A, Zhang Y. Effects of products from heterogeneous reactions on homogeneous combustion for H2/O2 mixture in the micro combustor. Appl Therm Eng 2016:897–903. [30] Lu Q, Pan J, Yang W, Tang A, Bani S, Shao X. Interaction between heterogeneous and homogeneous reaction of premixed hydrogen–air mixture in a planar catalytic micro-combustor. Int J Hydrogen Energy 2017. [31] Chen J, Song W, Xu D. Flame stability and heat transfer analysis of methane-air mixtures in catalytic micro-combustors. Appl Therm Eng 2016:837–48. [32] F. Inc, FLUENT 6.3 User’s Guide. (2006). [33] Emissivity v Temperature Graph For Platinum, 2018. [34] Yang B, Pope SB. An investigation of the accuracy of manifold methods and splitting schemes in the computational implementation of combustion chemistry. Combust Flame 1998;1:16–32. [35] Rieger J, Weiss C, Rummer B. Modelling and control of pollutant formation in blast stoves. J Cleaner Prod 2015;3:254–61. [36] Yamamoto K, Ozeki M, Hayashi N, Yamashita H. Burning velocity and OH concentration in premixed combustion. Proc Combust Inst 2009;1(2009):1227–35. [37] Yamamoto K, Isii S, Ohnishi M. Local flame structure and turbulent burning velocity by joint PLIF imaging. Proc Combust Inst 2011;1(2011):1285–92. [38] Deutschmann O, Schmidt R, Behrendt F, Warnat J. Numerical modeling of catalytic ignition. Symposium on Combustion 1996;1:1747–54. [39] Tang A, Xu Y, Pan J, Yang W, Jiang D, Lu Q. Combustion characteristics and performance evaluation of premixed methane/air with hydrogen addition in a microplanar combustor. Chem Eng Sci 2015:235–42. [40] Chen J, Gao X, Xu D. Kinetic effects of hydrogen addition on the catalytic selfignition of methane over platinum in micro-channels. Chem Eng J 2016:1028–34. [41] M. X, L.X. F, X, L.Y. X, Lmaz, Experimental analysis of the effects of hydrogen addition on methane combustion. Int. J. Energy Res., 5 (2012) 643–647.

[3] Yan Y, Tang W, Zhang L, Pan W, Li L. Thermal and chemical effects of hydrogen addition on catalytic micro-combustion of methane–air. Int J Hydrogen Energy 2014;33:19204–11. [4] Federici JA, Norton DG, Brüggemann T, Voit KW, Wetzel ED, Vlachos DG. Catalytic microcombustors with integrated thermoelectric elements for portable power production. J Power Sources 2006;2:1469–78. [5] Haruta M, Souma Y, Sano H. Catalytic combustion of hydrogen—II. An experimental investigation of fundamental conditions for burner design. Int J Hydrogen Energy 1982;9:729–36. [6] Vlachos DG, Bui PA. Catalytic ignition and extinction of hydrogen: comparison of simulations and experiments. Surf Sci 1996;3:L625–30. [7] Mantzaras J, Appel C. Effects of finite rate heterogeneous kinetics on homogeneous ignition in catalytically stabilized channel flow combustion. Combust Flame 1999;4:336–51. [8] Deutschmann O, Maier LI, Riedel U, Stroemman AH, Dibble RW. Hydrogen assisted catalytic combustion of methane on platinum. Catal Today 2000;1:141–50. [9] Zheng X, Mantzaras J, Bombach R. Hetero-/homogeneous combustion of ethane/air mixtures over platinum at pressures up to 14 bar. Proc Combust Inst 2013;2:2279–87. [10] Yang H, Feng Y, Wang X, Jiang L, Zhao D, Hayashi N, et al. OH-PLIF investigation of wall effects on the flame quenching in a slit burner. Proc Combust Inst 2013;2:3379–86. [11] Deng C, Yang W, Zhou J, Liu Z, Wang Y, Cen K. Catalytic combustion of methane, methanol, and ethanol in microscale combustors with Pt/ZSM-5 packed beds. Fuel 2015:339–46. [12] Yu S, Yong F, Suzuki Y. Radical quenching of metal wall surface in a methane-air premixed flame. Combust Flame 2015;10:4036–45. [13] Yang W, Wang Y, Zhou J, Zhou J, Wang Z, Cen K. Catalytic self-sustaining combustion of the alkanes with Pt/ZSM-5 packed bed in a microscale tube. Chem Eng Sci 2016:30–6. [14] Sarli VD, Barbato PS, Benedetto AD, Landi G. Start-up behavior of a LaMnO 3 partially coated monolithic combustor at high pressure. Catalysis Today PA 2014:200–10. [15] Landi G, Benedetto AD, Barbato PS, Russo G, Sarli VD. Transient behavior of structured LaMnO3 catalyst during methane combustion at high pressure. Chem Eng Sci 2014:350–8. [16] Ciuparu D, Perkins E, Pfefferle L. In situ DR-FTIR investigation of surface hydroxyls on γ-Al2O3 supported PdO catalysts during methane combustion. Appl Catal A 2004;2:145–53. [17] Yu S, Suzuki Y. Effect of wall surface reaction on a methane-air premixed flame in narrow channels with different wall materials. Proc Combust Inst 2013;2:3395–402. [18] Sui R, Mantzaras J, Bombach R. A comparative experimental and numerical investigation of the heterogeneous and homogeneous combustion characteristics of fuel-rich methane mixtures over rhodium and platinum. Proc Combust Inst 2016;3:4313–20. [19] Norton DG, Vlachos DG. Combustion characteristics and flame stability at the microscale: a CFD study of premixed methane/air mixtures. Chem Eng Sci 2003;21:4871–82. [20] Kaisare NS, Stefanidis GD, Vlachos DG. Comparison of ignition strategies for catalytic microburners. Proc Combust Inst 2009;2:3027–34.

932