Influence of methane addition on soot formation in pyrolysis of acetylene

Influence of methane addition on soot formation in pyrolysis of acetylene

Combustion and Flame 193 (2018) 83–91 Contents lists available at ScienceDirect Combustion and Flame journal homepage: www.elsevier.com/locate/combu...

2MB Sizes 0 Downloads 53 Views

Combustion and Flame 193 (2018) 83–91

Contents lists available at ScienceDirect

Combustion and Flame journal homepage: www.elsevier.com/locate/combustflame

Influence of methane addition on soot formation in pyrolysis of acetylene Alexander Eremin a, Ekaterina Mikheyeva a,b,∗, Ivan Selyakov a a b

Joint Institute for High Temperatures RAS, Izhorskaya 13 Bld. 2, Moscow 125412, Russia Bauman Moscow State Technical University, 2nd Baumanskaya Street 5, Moscow 105005, Russia

a r t i c l e

i n f o

Article history: Received 8 December 2017 Revised 17 January 2018 Accepted 7 March 2018

Keywords: Soot formation Soot diagnostic Acetylene pyrolysis Methane pyrolysis

a b s t r a c t Time-resolved laser-induced incandescence for particle sizing and laser light extinction for soot volume fraction was applied simultaneously to study the influence of methane addition on soot formation in acetylene pyrolysis. Three series of the experiments with initial mixtures of 2% C2 H2 + Ar, 1% CH4 + Ar and 2% C2 H2 + 0.5/1/2% CH4 + Ar in the temperature range of 160 0–230 0 K and the pressure range of 4–5 bar behind reflected shock waves were carried out. The kinetic characteristic of the soot formation process— the induction time of soot particle inception as well as the temperature dependences of final values of soot volume fraction and particle sizes have been determined and analyzed. An essential increase of soot volume fraction, particle sizes and a decrease of induction time of soot inception at methane addition to acetylene were observed. The gas phase kinetic modeling of the investigated processes up to the soot nuclei precursors formation has been performed. Analysis of gas kinetic stages of acetylene decomposition with methane addition has demonstrated the significant increase of the rates of pyrene formation followed by PAH growth due to effective propargyl C3 H3 formation. © 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

1. Introduction The problem of soot formation during the combustion and pyrolysis of hydrocarbons has been in the focus of attention for many decades. Despite a large amount of experimental and theoretical studies (see for example reviews [1–4] and references within), a number of important issues remain insufficiently studied. One of such problems is the mechanism of soot formation in the pyrolysis of acetylene. A unique feature of acetylene is the presence of a triple bond between carbon atoms at a ratio C/H = 1, which provides an extremely effective formation of polyyne chains during its pyrolysis. However, polyyne molecules are very thermodynamically stable at high temperatures that cause the overprediction of the soot yield if the higher polyynes are directly involved in the formation of soot particle nuclei [5]. In recent years, a number of works have attempted to adapt the HACA mechanism and supplement it with elements from the polyyne model [5–6]. However, there is no universal model that satisfactorily qualitatively and quantitatively describes the formation of soot in acetylene-containing mixtures.

∗ Corresponding author at: Joint Institute for High Temperatures RAS, Izhorskaya 13 Bld. 2, Moscow 125412, Russia. E-mail address: [email protected] (E. Mikheyeva).

An even greater level of complexity arises describing the formation of soot in binary mixtures. Experimentally, the process of soot formation in binary hydrocarbon mixtures was studied in shock tube pyrolysis [7–8] and flames [9–12] and flow reactor [12]. Observed results indicated that the data on soot formation in various binary mixtures are rather conflicting [7]. Strong decrease of soot was found with addition of hydrogen to shock tube pyrolysis of acetylene [7,8,13] and for the ethylene/air diffusion flame [14], but no affect was found for premixed ethylene flame [9]. Methane addition to ethylene resulted in the increase of PAH in flow reactor and diffusion flame and the increase of soot in diffusion flame whereas experiments in premixed flame did not show any synergetic effect [12]. Oxygen-containing additives caused the oxidation of carbon and mostly the decrease of soot yield in shock tube pyrolysis [15–16] and premixed flames [17]. On the other hand, a small amount of oxygenated additives can promote soot formation due to the peculiarity of blend pyrolysis kinetics [18]. The increase in soot yield was found in non-premixed ethylene flame with partial ethanol replacement as a fuel [19]. For aromatic-acetylenic mixtures both the synergistic and the negative effects of soot yield were observed for pyrolysis and flame conditions [7,8,10]. In this view, it becomes particularly important to obtain new experimental data in the mixtures of acetylene with alkanes, primarily with methane. It is known that the formation of soot in the pyrolysis of methane due to C/H = 0.25 is extremely difficult. On

https://doi.org/10.1016/j.combustflame.2018.03.007 0010-2180/© 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

84

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91

Fig. 1. Experimental setup.

the other hand, when adding methane, additional carbon atoms appear in the mixture, which can lead to an increase in the soot yield, but at the same time a fourfold more hydrogen atoms appear, which, as it is known, should reduce the soot yield. Moreover, the CH4 molecules in the reacting mixture give different hydrocarbon radicals, which also could have a controversial effect on soot growth. Therefore, the influence of methane additives on the soot yield in acetylene is a priori not obvious. The preliminary experiments [20] showed that the soot yield in such mixtures substantially increases. Therefore, the purpose of this work was a more detailed experimental and numerical study of this interesting and unexpected effect. 2. Experiment 2.1. Apparatus and mixtures The experiments were carried out behind reflected shock waves in a conventional stainless steel diaphragm type shock tube with an inner diameter of 50 mm. The driver and the driven section had 1.5 and 3.5 m in length respectively. The values of temperature (TRSW ) and pressure (PRSW ) behind the reflected shock wave (RSW) were determined based on the measured incident shock wave velocity by applying one-dimensional gas-dynamic theory and assuming frozen reaction conditions. The shock tube was equipped with two PCB113B piezoelectric pressure gauges to measure the incident shock wave velocity with the accuracy of 0.5% that resulted in the TRSW uncertainty about 25 K. The TRSW was varied in the range of 1650–2250 K and PRSW was in the range of 4–5 bar. The shock tube was evacuated by a fore-vacuum pump to a pressure of 4 × 10−2 mbar. After every experiment the inner walls of tube were cleaned several times with an ethanol. The test gas mixtures were prepared manometrically in a stainless steel mixing vessel. The composition of investigated mixtures was 2% C2 H2 + 98% Ar, 1% CH4 + 99% Ar, 2% C2 H2 + 0.5% CH4 + 97.5% Ar, 2% C2 H2 + 1% CH4 + 97% Ar and 2% C2 H2 + 2% CH4 + 96% Ar. The purity for C2 H2 was certified to be 99.9%, for CH4 —99.99%, for Ar—99.999%. Strong

dilution with inert gas in the shock tube reactor is due to two reasons: first, better gas dynamics in a shock tube, which allows obtaining more accurate kinetic data, and, second, the limitations imposed by optical diagnostic methods: the LII and laser extinction. Our experience [13] has shown that in order to avoid saturation of laser extinction, mixtures with no more than 3% acetylene at pressures up to 10 bar should be used. The optical access in the measurement section was given by four calcium fluoride windows of 6 mm in a diameter mounted perpendicular to each other. The frontal diagnostic quartz glass window of 85 mm in diameter (more than inner diameter of shock tube) was mounted in the end plate. The measurement section of the shock tube and the diagnostic are presented schematically in Fig. 1. The pressure has been controlled by PCB113B piezoelectric pressure gauge in the measurement section. During the experiment, the pressure did not change noticeably. The temperature evolution behind the reflected shock wave in the given experiments was not controlled. However, based on our previous measurements in 3% acetylene [13], we assumed that the temperature did not change significantly. 2.2. Extinction measurements The extinction diagnostic is one of the most informative methods of investigation of the condensed phase formation process [4]. The beam of a conventional 20 mV HeNe laser was passed through two calcium fluoride windows of the shock tube (Fig. 1) and focused on the active photodiode PDA10A-ES (THORLABS) with a rise time of 10 ns. The detector was optically blocked by an interference filter of λ = 632.8 nm (FWHM 1 nm) to suppress the thermal radiation of the reacting gas-particle mixture. The experimental signals were registered by Tektronix TDS 2014B digital scope with 100 MHz band width. Particle extinction, determined by attenuation of passing radiation, is linked with the soot volume fraction by Lambert–Beer’s law [4]. Due to uncertainties of optical properties of growing soot [21–23] in present study the “relative soot volume fraction” that is the product of volume fraction of condensed phase fV and the function of refractive index of soot was chosen as

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91

85

quantitative value of extinction measurements:

fV E (m )/[C ] =

ln

 I(t )  I0

λ

−6π l[C ]

(1)

where t is the reaction time, I0 and I(t) are the incoming and transmitted laser light intensities respectively, λ is a diagnostic wavelength, l is the optical path length, [C] is total carbon atom concentration. Since the total amount of carbon in the system is not the constant in the various mixtures under study, the normalization of soot volume fraction on the total concentration of carbon atoms has been made. The extinction measurements also allow analyzing the induction time of particle inception. The absolute value of the induction time τ ind was defined as the intersection of the inflectional tangent of the soot volume fraction profile with the time axis [4,8]. The induction period for the mixture of 1% CH4 was not determined due to too small magnitude of detected signals. For the mixtures of 2% C2 H2 and 2% C2 H2 + 1% CH4 the error caused by the value of the signal-to-noise ratio is in the range from 5% for the maximum absolute magnitude of signals (signal-to-noise ratio of about 20) to 25% for the minimum signals for which the values were determined (signal-to-noise ratio is not less than 2).

Fig. 2. Time profiles of relative soot volume fraction measured in different mixtures at similar temperatures and pressures.

2.3. Laser induced incandescence for particle sizing The laser induced incandescence (LII) diagnostic is now widely used for particle sizing [24]. The LII sizing technique is based on the particle heating with a laser pulse and the analysis of the decay time of thermal radiation of particles. The Nd:YAG laser LQ215 (SOLAR Laser Systems) at a wavelength of 1064 nm with pulse duration of 6 ns FWHM was applied for the particle heating. The Nd:YAG laser was triggered by a pressure gauge through a time delay generator that provided different delays after reflected shock wave arrival. The laser fluence was in the range of 0.25–0.3 J/cm2 . Four channels for LII measurements were applied: two channels through the frontal diagnostic window and two via side windows (see Fig. 1). The LII signals were detected using narrow band pass filters with a Hamamatsu H6780-20 and H6780-04 photomultiplier modules (rise time of 0.78 ns) coupled with a LeCroy WaveRunner 6060A oscilloscope (500 MHz bandwidth). The wavelengths of 450 and 770 nm were chosen both for frontal and side window channels. The particle size was determined by approximation of the experimental signal with the calculated curve derived from LII model described in [21]. The typical TEM images of soot particles obtained during pyrolysis of 3% acetylene in a shock tube are given in [21]. Analysis of a large number of TEM images had showed that the particles obey a lognormal distribution in size with a sigma value equal to 1.1. This value is used in present paper to calculate the particle size from the LII measurements. The estimation of the total error, caused by the uncertainty of soot particles properties, showed the values between −34% and +27% [21]. The quantitative size values were determined by averaging of results from all four channels. Typical time profiles of LII signals were shown in our previous papers [13,21]. 3. Experimental results 3.1. Results of extinction measurements The typical time profile of soot volume fraction measured in acetylene, methane and mixture of C2 H2 + CH4 at the temperature range 1900–1953 K is shown in Fig. 2. It is clearly seen that while in 2% C2 H2 the values fV E(m)/[C] is very weak and in 1% CH4 it is negligible, the addition of 1% CH4 to 2% C2 H2 results in the essential increase of soot formation.

Fig. 3. Temperature dependence of relative volume fraction experimentally observed in mixtures of 2% C2 H2 + 98% Ar, 1% CH4 + 99% Ar, 2% C2 H2 + 0.5% CH4 + 97% Ar, 2% C2 H2 + 1% CH4 + 97% Ar and 2% C2 H2 + 2% CH4 + 97% Ar at the reaction times of 1 ms after RSW arrival.

Next Fig. 3 demonstrates the temperature dependence of soot volume fraction observed in different mixtures, representing the well-known bell-shape dependence reported in a number of works on hydrocarbons pyrolysis (see e.g., [4]). The process of soot formation under pyrolysis conditions in the shock tube strongly depends on temperature. The low temperature signals show a definite induction time followed by a rapid signal increase up to a steady state value. The high temperature signals look different: a small sudden signal increase behind the reflected shock is followed by a slow continuous signal increase, which does not end in a steady state value. Unfortunately, the shock tube reactor significantly limits the time of observation. And in classical studies on the formation of soot in pyrolysis of hydrocarbons [25– 26], it is generally accepted to present the results at a specific measurement time. One can clearly see that in the methane pyrolysis there was practically no formation of soot, in acetylene pyrolysis—distinctive bell of the soot volume fraction, and the addition of methane to acetylene led to a significant increase in the soot formation. Note that in the pyrolysis of acetylene the maximum of fV E(m)/[C] lies around 20 0 0 K (empty circles in Fig. 3) and the addition of 0.5–2% methane to 2% acetylene caused the shift of maximum of fV E(m)/[C] to the lower temperatures of about 1850–1900 K (squares in Fig. 3). A monotonic increase in the yield of soot can be observed with increasing amount of methane addition.

86

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91

Fig. 4. Temperature dependence of induction time of particle inception experimentally observed in mixtures of 2% C2 H2 + 98% Ar and 2% C2 H2 + 1% CH4 + 97% Ar.

Fig. 6. Time profiles of particle count medium diameter (CMD) at the temperature range of maximum soot yield in mixtures of 2% C2 H2 + 98% Ar and 2% C2 H2 + 1% CH4 + 97% Ar.

CH4 + 97% Ar nanoparticles grow slightly faster than in the mixture of 2% C2 H2 + 98% Ar. 4. Modeling 4.1. Approach and kinetic scheme

Fig. 5. Temperature dependence of the particle count medium diameter (CMD) measured by LII in mixtures of 2% C2 H2 + 98% Ar and 2% C2 H2 + 1% CH4 + 97% Ar at the reaction times of 1 ms after RSW arrival.

The temperature dependences of the induction time of particle inception for acetylene and acetylene/methane mixtures are presented in Fig. 4. One can see the noticeable decrease of the induction time with methane addition in all experimentally observed temperature range. However, the slope of these dependences is approximately the same for both mixtures. 3.2. Results of LII measurements The temperature dependences of particle size measured at the last stage of shock tube experiment (∼1 ms after RSW arrival to the measurement section) for mixtures of acetylene with and without methane addition are presented in Fig. 5. The observed particle sizes do not exceed 7 nm in diameter. The maximum size of the particles formed in pyrolysis of 2% acetylene is about 2–3 nm and it is at the same temperature as maximum of volume fraction—at approximately 20 0 0 K (see Fig. 3), although the shape of temperature dependence is much smoother. For the mixture with methane a noticeable increase in particle size and the same tendency of shift of maximum size to the lower temperatures was observed. The large scattering of the experimental data is associated with a sufficiently large error of the LII diagnostic. The time profiles of nanoparticle sizes at the temperature range of maximum soot yield for acetylene with and without methane addition are presented in Fig. 6. In pyrolysis of 2% C2 H2 + 1%

A complete simulation of the processes of hydrocarbons pyrolysis and subsequent formation of condensed carbon nanoparticles is an independent, very complex task [5,27–30] which was beyond the scope of this paper. Therefore, the aim of performed numerical analysis was to identify the most probable kinetic routes of methane additives influence on the formation of soot precursors in pyrolysis of acetylene. For this aim the recent soot kinetic mechanism [30] was applied. The gas-phase kinetic model [30] describes the high-temperature pyrolysis and oxidation for a wide range of hydrocarbon fuels. This model was recently tested on acetylene pyrolysis at high temperatures [6] and has shown a good agreement with different experiments in shock tubes [31–34] and flow reactors [35–36]. The kinetic mechanism consists of gas-phase reactions including the kinetic route to benzene and phenyl formation, PAHs growth from benzene to aromatic compounds larger than pyrene (C16 H10 ) (e.g., corannulene (C20 H10 )) and the subsequent soot kinetic model based on discrete sectional method. In the present work only gas phase part of mechanism [30] was used and the commonly accepted soot nuclei precursor [27]—pyrene (C16 H10 ) was chosen for presentation of calculation. In addition to pyrene, calculations were made for C20 H16 (BIN1A) and C20 H10 (BIN1B). Results have shown that the behavior of the curves is similar, i.e. it appeared that it is not so important which kind of the large gas phase components is taken as the soot nuclei precursor. Therefore, in the following simulations only the calculations with pyrene have been performed. The kinetic modeling was carried out using Chemkin code. The calculations for experimentally investigated conditions were carried out in the adiabatic approximation at constant pressure. 4.2. Results of calculation Figure 7 shows the time profiles of the mole fractions of pyrene calculated using mechanism [30] at a temperature of 1900 K. One can see that the addition of methane to acetylene results in a significant acceleration of formation and an increase in the number of soot precursors. Figure 8 shows the temperature dependences of

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91

Fig. 7. Time profiles of pyrene mole fractions calculated using mechanism [30] at a temperature of 1900 K in mixtures 2% C2 H2 + 98% Ar and 2% C2 H2 + 1% CH4 + 97% Ar, p = 4.5 bar.

Fig. 8. Temperature dependence of pyrene maximum mole fraction calculated by mechanism [30] for a calculation time of 1 ms for mixtures2% C2 H2 + 98% Ar and 2% C2 H2 + 1% CH4 + 97% Ar, p = 4.5 bar.

The main kinetic characteristic of the particle formation process is the induction time of particle inception. Since the induction times measured in the different mixtures are well approximated by the Arrhenius-like law [4], the effective activation energy of this process could be extracted from the slope of experimental dependencies presented in Fig. 4. One can see that the methane addition does not change the slope of the temperature dependence of measured induction time. By the least square fitting of the experimental points it was found that the effective activation energy value is about 46 kcal/mole for both mixtures. This value, evidently reflecting the “bottle-neck” reaction, is close to the activation energy of the reactions of acetylene molecules recombination leading to formation of C4 H2 and C4 H3 . These reactions are the well-known first reactions of acetylene pyrolysis in the intermediate temperatures corresponding to our experimental conditions [37]. The first main reaction of methane decomposition forming methyl and hydrogen atom is characterized by much higher activation energy and obviously is not the limiting stage of the process in acetylene–methane mixtures. Therefore, in order to analyze the reasons for promoting pyrene formation when methane is added to acetylene, the entire cycle of reactions leading to the formation of aromatic molecules should be considered. Since methane is a saturated aliphatic molecule, the formation of first aromatic molecules like phenyl and benzene during its pyrolysis is quite difficult and includes a number of stages. In Fig. 9 the graphic scheme of main routes to formation of phenyl and benzene during pyrolysis of 1% methane is presented. It is well known that one of the key radicals in the formation of phenyl and benzene is propargyl C3 H3 . First, the important role of propargyl was proposed in [38] in the discussion of pyrolytic formation of arenes from hydrocarbons. Later the reactions with C3 H3 helped to explain the experimental data of shock tube pyrolysis of allene and 1,3 butadiene [39,40], acetylene flames [41] and flow reactor pyrolysis of 1,5-hexadiyne [42]. During pyrolysis of methane several different routes lead to formation of propargyl and all of them include reactions with acetylene molecules (see Fig. 9). The formation of acetylene in methane pyrolysis proceeds through the formation of C2 H5 by two methyl radicals CH3 recombination:

CH3 + CH3 = C2 H5 + H pyrene maximum mole fractions in the investigated mixtures for a calculation time of 1 ms. This plot also demonstrates an essential increase in pyrene concentration when methane is added to acetylene. So, one can conclude that, in general, the results of simulations are consistent with the tendencies observed in the experiments. The addition of methane to acetylene leads to an extremely large increase in concentration of soot precursors in all observed temperature range.

87

(R1)

And the following reaction chain C2 H5 →C2 H4 →C2 H3 →C2 H2 :

C2 H5 (+M) = C2 H4 + H(+M)

(R2)

C2 H4 + H = C2 H3 + H2

(R3)

C2 H3 (+M) = C2 H2 + H(+M)

(R4)

C2 H4 (+M) = C2 H2 + H2 (+M)

(R5)

5. Discussion The simulation indicates a huge increase in the concentrations of pyrene (assumed as a soot precursor) when methane was added to acetylene (see Figs. 7 and 8), corresponding with the observed increase in the soot volume fraction in the experiments. Obviously, the further growth of condensed particles is determined by a complicated combination of surface growth and coagulation processes, so a direct comparison of these two quantities would be unjustified. However, a noticeable increase in size and essential increase in soot volume fraction observed in experiments are qualitatively consistent with the results of the calculations. Within this consideration, we will focus on the analysis of possible kinetic reasons of the promotion of initial stages of soot growth in acetylene by the methane additives.

Recombination of acetylene with methyl radical leads to formation of C3 H4 (propyne and allene):

C2 H2 + CH3 = C3 H5

(R6)

C2 H2 + CH3 = C3 H4 + H

(R7)

C3 H5 = C3 H4 + H

(R8)

and following C3 H3 (propargyl) formation:

C3 H4 (+M) = C3 H3 + H(+M) C3 H4 + H = C3 H3 + H2

(R9) (R10)

88

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91

Fig. 9. Main routes of first aromatic ring production at pyrolysis of 1% methane at T = 1900 K and p = 4.5 bar. The thickness of the arrows is proportional to the rates of the corresponding reactions.

C2 H2 + C2 H = C4 H3

(R15)

C2 H2 + C2 H2 = C4 H3 + H

(R16)

and following recombination of C4 H3 + C2 H2 :

C4 H3 + C2 H2 → C6 H5

(R17)

results in direct phenyl formation. Another route to phenyl is recombination of C4 H2 and C2 H3 :

C4 H2 + C2 H3 → C6 H5

(R18)

through reactions:

Fig. 10. Main routes of first aromatic ring production at pyrolysis of 2% acetylene at T = 1900 K and p = 4.5 bar. The thickness of the arrows is proportional to the rates of the corresponding reactions.

Besides that, the recombination of acetylene with methylene CH2 also results in propargyl formation:

C2 H2 + C2 H2 = C4 H2 + H2

(R19)

C4 H3 = C4 H2 + H

(R20)

and reaction R4. And the third main channel to first aromatic ring is recombination of C4 H4 with acetylene and ethynyl radical:

C4 H4 + C2 H = C6 H5

(R21) (R22)

C2 H2 + CH2 (S) = C3 H3 + H

(R11)

C4 H4 + C2 H2 → C6 H6

C2 H2 + CH2 = C3 H3 + H

(R12)

C4 H4 forms from the recombination of acetylene molecules or acetylene with vinyl radical:

And recombination of two propargyl radicals results in formation of phenyl and benzene:

C2 H2 + C2 H2 = C4 H4

(R23) (R24)

C3 H3 + C3 H3 = C6 H5 + H

(R13)

C2 H2 + C2 H3 = C4 H4 + H

C3 H3 + C3 H3 (+M) = C6 H6 (+M)

(R14)

In pyrolysis of mixture of 2% acetylene and 1% methane the main routes of first aromatic ring formation is presented in Fig. 11. When methane is added to acetylene, in addition to the routes described above, the reactions of methyl and methylene with acetylene forming propargyl (R6–R12) begin to work, which are much

In 2% acetylene pyrolysis the main routes of first aromatic ring formation is presented in Fig. 10. In this route the formation of C4 H3 via reactions:

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91

89

Fig. 11. Main routes of first aromatic ring production at pyrolysis of mixture of 2% acetylene and 1% methane at T = 1900 K and p = 4.5 bar. The thickness of the arrows is proportional to the rates of the corresponding reactions.

Fig. 12. Time profiles of the mole fractions of propargyl radicals calculated using mechanism [30] at a temperature of 1900 K in mixtures 2% C2 H2 + 98% Ar and 2% C2 H2 + 1% CH4 + 97% Ar, p = 4.5 bar.

hindered in pure methane, but become extremely effective in this mixture. In Figs. 12 and 13 the time and temperature profiles of propargyl mole fraction in acetylene and acetylene/methane mixtures are compared. It is clearly seen that in acetylene–methane mixtures the formation of propargyl occurs much more efficiently and starts at lower temperatures. Thus, the above analysis of the initial stages of pyrolysis demonstrates that addition of methane to acetylene opens up new efficient and kinetically favorable routes for the formation of the first aromatic molecules, in the first place—the reactions of acetylene with methyl and methylene (R7, R10–R12), which directly forms the propargyl radicals. The next point that needs to be explained is a shift in the experimentally observed maximum values of the soot volume fraction and size toward lower temperatures (see Figs. 3 and 5).

Fig. 13. Temperature dependence of propargyl radicals maximum mole fraction calculated by mechanism [30] for a calculation time of 1 ms for mixtures 2% C2 H2 + 98% Ar and 2% C2 H2 + 1% CH4 + 97% Ar, p = 4.5 bar.

In case of pyrolysis of acetylene and methane mixture all the activation energies (Ea ) of the main reactions leading to methyl and methylene formation:

CH4 + H = CH3 + H2

(R25)

CH3 + H = CH2 (S) + H2

(R26)

to propyne and allene formation (R7) and finally to propargyl formation (R10–R12) do not exceed 15 kcal/mole and the rate constants for these reactions are about 4 orders of magnitude more at T < 20 0 0 K than for reactions of acetylene molecules recombination R16, R19, R23 (with Ea ∼ 60–80 kcal/mole) crucial for pure acetylene pyrolysis. This fact could explain the very effective propargyl formation at T < 20 0 0 K in the acetylene–methane mixture (see Fig. 12) resulting in the shift of temperature dependences of soot

90

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91

volume fraction and particle sizes to the lower temperatures observed in experiments. The following formation of PAHs from the benzene/phenyl passes though HACA mechanism [43] and different PAH reactions beyond HACA [28]. The variety of PAH precursor chemistry is dependent on fuel composition and conditions [28]. The addition of methane to acetylene and corresponding effective formation of propargyl radicals leads to emergence of additional route of PAHs growth—through propargyl recombination [44], for example:

C3 H3 + C6 H5 →products

(R27)

C3 H3 + C12 H7 →products

(R28)

C3 H3 + C12 H10 →products

(R29)

C3 H3 + C14 H9 →products

(R30)

In recent work [45], a new mechanism for PAH growth including methylene was found: addition C2 H2 → intramolecular hydrogen migration → addition CH2 → cyclization → H-elimination. It can also start working in case of methane addition. Of course, briefly mentioned pathways of PAH growth additional to HACA do not exhaustively describe the real situation. Therefore the new theoretical studies on this subject are very necessary. Thus, the presented analysis of known kinetic models clearly demonstrates and explains the effects of promoting the soot formation upon addition of methane to acetylene, observed in experiments. The influence of methane addition on the initial stage of soot growth process is manifested both in the formation of first aromatic ring and in the growth of PAH molecules. 6. Conclusions The effect of methane addition on the soot formation in the pyrolysis of acetylene was studied. Essential increase of soot volume fraction in pyrolysis of acetylene with methane addition was experimentally found under shock tube pyrolysis conditions. Qualitative kinetic analysis of the initial stages of soot formation has shown that the reason of this effect is the propargyl recombination leading to the acceleration of first aromatic ring formation and growth of larger PAHs molecules. The main channel of propargyl formation is recombination of methylene CH2 and methyl CH3 radicals (formed at methane decomposition) with acetylene molecules. Acknowledgments This work was supported by the Joint Project of the German Research Foundation (DFG) [grant SCHU 1369/24-1] and the Russian Foundation for Basic Research (RFBR) [grant RFBR-16-58-12014]. References [1] B.S. Haynes, H.G. Wagner, Soot formation, Prog. Energy Combust. Sci. 7 (1981) 229–273. [2] H. Richter, J.B. Howard, Formation of polycyclic aromatic hydrocarbons and their growth to soot – a review of chemical reaction pathways, Prog. Energy Combust. Sci. 26 (20 0 0) 565–608. [3] Z.A. Mansurov, Soot formation in combustion processes (review), Combust. Explos. Shock Waves 41 (2005) 727–744. [4] A.V. Eremin, Formation of carbon nanoparticles from the gas phase in shock wave pyrolysis processes, Prog. Energy Combust. Sci. 38 (2012) 1–40. [5] G.L. Agafonov, I.V. Bilera, P.A. Vlasov, Y.A. Kolbanovskii, V.N. Smirnov, A.M. Tereza, Soot formation during the pyrolysis and oxidation of acetylene and ethylene in shock waves, Kinet. Catal. 56 (2015) 12–30. [6] C. Saggese, N.E. Sanchez, A. Frassoldati, A. Cuoci, T. Faravelli, M.U. Alzueta, E. Ranzi, Kinetic modeling study of polycyclic aromatic hydrocarbons and soot formation in acetylene pyrolysis, Energy Fuels 28 (2014) 1489–1501. [7] M. Frenklach, T. Yuan, M.K. Ramachandra, Soot formation in binary hydrocarbon mixtures, Energy Fuels 2 (1988) 462–480.

[8] V.G. Knorre, D. Tanke, T. Thienel, H.G. Wagner, Soot formation in the pyrolysis of benzene/acetylene and acetylene/hydrogen mixtures at high carbon concentrations, Symp. (Int.) Combust. 26 (1996) 2303–2310. [9] B.S. Haynes, H. Jander, H. Matzing, H.G. Wagner, The influence of gaseous additives on the formation of soot in premixed flames, Symp. (Int.) Combust. 19 (1982) 1379–1385. [10] R.J. Gill, D.B. Olson, Estimation of soot thresholds for fuel mixtures, Combust. Sci. Technol. 40 (1984) 307–315. [11] S.J. Harris, A.M. Weiner, Soot particle growth in premixed toluene/ethylene flames, Combust. Sci. Technol. 38 (1984) 75–87. [12] J.F. Roesler, S. Martinot, C.S. McEnally, L.D. Pfefferle, J.-L. Delfau, C. Vovelle, Investigating the role of methane on the growth of aromatic hydrocarbons and soot in fundamental combustion processes, Combust. Flame 134 (2003) 249–260. [13] A. Eremin, E. Gurentsov, E. Mikheyeva, Experimental study of molecular hydrogen influence on carbon particle growth in acetylene pyrolysis behind shock waves, Combust. Flame 159 (2012) 3607–3615. [14] Ö.L. Gülder, D.R. Snelling, R.A. Sawchuk, Influence of hydrogen addition to fuel on temperature field and soot formation in diffusion flames, Symp. (Int.) Combust. 26 (1996) 2351–2357. [15] A. Eremin, E. Gurentsov, E. Mikheyeva, Experimental study of temperature influence on carbon particle formation in shock wave pyrolysis of benzene and benzene-ethanol mixtures, Combust. Flame 162 (2015) 207–215. [16] M. Frenklach, T. Yuan, Effect of alcohol addition on shock-initiated formation of soot from benzene, Proceedings of XVI International Symposium on Shock Tubes Waves (1987), pp. 487–493. [17] J. Wu, K.H. Song, T. Litzinger, S.-Y. Lee, R. Santoro, M. Linevsky, M. Colket, D. Liscinsky, Reduction of PAH and soot premixed ethylene-air flames by addition of ethanol, Combust. Flame 144 (2006) 675–687. [18] H. Böhm, M. Braun-Unkhoff, Numerical study of the effect of oxygenated blending compounds on soot formation in shock tubes, Combust. Flame 153 (2008) 84–96. [19] C. McEnally, L. Pfefferle, The effects of dimethyl ether and ethanol on benzene and soot formation in ethylene nonpremixed flames, Proc. Combust. Inst. 31 (2007) 603–610. [20] A. Eremin, A. Kiverin, E. Mikheyeva, I. Selyakov, Soot formation in pyrolysis of the mixture of acetylene with methane, 10th Mediterranean Combustion Symposium, 2017 WIP 38. [21] A.V. Eremin, E.V. Gurentsov, E. Popova, K. Priemchenko, Size dependence of complex refractive index function of growing nanoparticles, Appl. Phys. B Laser Opt. 104 (2011) 285–295. [22] H. Bladh, J. Johnsson, N.E. Olofsson, A. Bohlin, P.E. Bengtsson, Optical soot characterization using two-color laser-induced incandescence (2C-LII) in the soot growth region of a premixed flat flame, Proc. Combust. Inst. 33 (2011) 641–648. [23] F. Migliorini, K.A. Thomson, G.J. Smallwood, Investigation of the optical properties of aging soot, Appl. Phys. B Laser Opt. 104 (2011) 273–283. [24] H.A. Michelsen, C. Schulz, G.J. Smallwood, S. Will, Laser-induced incandescence: particulate diagnostics for combustion, atmospheric, and industrial application, Prog. Energy Combust. Sci. 51 (2015) 2–48. [25] M. Frenklach, S. Taki, R.A. Matula, A conceptual model for soot formation in pyrolysis of aromatic hydrocarbons, Combust. Flame 49 (1983) 275–282. [26] R. Starke, B. Kock, P. Roth, A. Eremin, E. Gurentsov, V. Shumova, V. Ziborov, Shock wave induced carbon particle formation from CCl4 and C3 O2 observed by laser extinction and by laser-induced incandescence (LII), Combust. Flame 135 (2003) 77–85. [27] J. Appel, H. Bockhorn, M. Frenklach, Kinetic modeling of soot formation with detailed chemistry and physics: laminar premixed flames of C2 hydrocarbons, Combust. Flame 121 (20 0 0) 122–136. [28] H. Wang, Formation of nascent soot and other condensed-phase materials in flames, Proc. Combust. Inst. 33 (2011) 41–67. [29] D. Chen, Z. Zainuddin, E.K.Y. Yapp, J. Akroyd, S. Mosbach, M. Kraft, A fully coupled simulation of PAH and soot growth with a population balance model, Proc. Combust. Inst. 34 (2013) 1827–1835. [30] C. Saggese, S. Ferrario, J. Camacho, A. Cuoci, A. Frassoldati, E. Ranzi, H. Wang, T. Faravelli, Kinetic modeling of particle size distribution of soot in a premixed burner-stabilized stagnation ethylene flame, Combust. Flame 162 (2015) 3356–3369. [31] J.H. Kiefer, S.S. Sidhu, R.D. Kern, K. Xie, H. Chen, L.B. Harding, The homogeneous pyrolysis of acetylene II: the high temperature radical chain mechanism, Combust. Sci. Technol. 82 (1992) 101–130. [32] Y. Hidaka, K. Hattori, T. Okuno, K. Inami, T. Abe, T. Koike, Shock-tube and modeling study of acetylene pyrolysis and oxidation, Combust. Flame 107 (1996) 401–417. [33] C. Wu, H. Singh, R. Kern, Pyrolysis of acetylene behind reflected shock waves, Int. J. Chem. Kinet. 19 (1987) 975–996. [34] M.B. Colket, In the pyrolysis of acetylene and vinylacetylene in a single-pulse shock tube, Symp. (Int.) Combust. 21 (1988) 851–864. [35] N.E. Sanchez, A. Callejas, A. Millera, R. Bilbao, M.U. Alzueta, Formation of PAH and soot during acetylene pyrolysis at different gas residence times and reaction temperatures, Energy 43 (2012) 30–36. [36] K. Norinaga, O. Deutschmann, K.J. Huttinger, Analysis of gas phase compounds in chemical vapor deposition of carbon from light hydrocarbons, Carbon 44 (20 06) 1790–180 0. [37] J.H. Kiefer, W.A. Von Drasek, The mechanism of the homogeneous pyrolysis of acetylene, Int. J. Chem. Kinet. 22 (1990) 747–786.

A. Eremin et al. / Combustion and Flame 193 (2018) 83–91 [38] C. Hurd, A.R. Macon, J.I. Simon, R.V. Levetan, Pyrolytic formation of arenes. I. Survey of general principles and findings, J. Am. Chem. Soc. 84 (1962) 4509–4515. [39] C.H. Wu, R.D. Kern, Shock-tube study of allene pyrolysis, J. Phys. Chem. 91 (1987) 6291–6296. [40] R.D. Kern, H.J. Singh, C.H. Wu, Thermal decomposition of 1,2 butadiene, Int. J. Chem. Kinet. 20 (1988) 731–747. [41] J.A. Miller, C.F. Melius, Kinetic and thermodynamic issues in the formation of aromatic compounds in flames of aliphatic fuels, Combust. Flame 91 (1992) 21–39.

91

[42] S.E. Stein, J.A. Walker, M.M. Suryan, A. Fahr, A new path to benzene in flames, Symp. (Int.) Combust. 23 (1990) 85–90. [43] M. Frenklach, Reaction mechanism of soot formation in flames, PCCP 4 (2002) 2028–2037. [44] A. Fahr, S.E. Stein, Reactions of vinyl and phenyl radicals with ethylene, ethane and benzene, Symp. (Int.) Combust. 22 (1988) 1023–1029. [45] P. Liu, H. Lin, Y. Yang, C. Shao, B. Guan, Z. Huang, Investigating the role of CH2 radicals in the HACA mechanism, J. Phys. Chem. 119 (2015) 3261–3268.