Ethylene-Propane and Ethylene-Dimethyl Ester Synergistic Effects on Soot Formation

 

 

M. Mbarawa^I; W. Lee^II; Y.W. Nam^II; S.H. Chung^III

^IDepartment of Mechanical Engineering. Tshwane University of Technology, Pretoria. South Africa, E-mail: mbarawamm@tut.ac.za
^IIDepartment of Mechanical Engineering. Dankook University, Seoul 140-714, Korea
^IIIDepartment of Mechanical Engineering. Seoul National University, Seoul 140-742, Korea

 

 

---

ABSTRACT

In this study, the synergistic effects of ethylene-propane and ethylene-dimethyl ester (DME) mixtures on soot formation were investigated experimentally using a co-flow diffusion flame burner. The soot volume fraction, soot particle diameter and number density were measured and compared to the homogenous mixture. Addition of DME and propane to the ethylene fuel increased soot volume fraction in the ethylene flames. The ethylene-propane has more pronounced synergistic effect in comparison to the ethylene-DME flames. This is due to the fact that during the decomposition of propane some methyl radicals are generated. The reactions related to these methyl radicals promote the formation of propargyl radicals and consequently the formation of benzene through propargyl self-reaction and finally to the soot formation. Although DME decomposition produces methyl, the C-0 bond in the DME removes some carbon from the reaction path that produces soot. Hence the soot formation in ethylene-DME mixture is much slower than that in ethylene-propane mixture.

---

 

 

Nomenclature

Roman

d Soot particle size [nm]

f Soot volume fraction

i,d Internal diameter [mm]

N Total number density of soot particles per unit volume [#/cm³]

Z Height above the burner exit [cm]

Subscript

p particle

V volume

 

1. Introduction

Soot formation and emission are dominant features of hydro-carbon-air diffusion combustion equipment such as reflects poor combustion and loss of efficiency. Furthermore, soot is a major pollutant and raises health concerns related to inhalation of soot particles. As a result, it is important to gain a fundamental understanding of complex soot formation processes in flames and their control mechanisms.

The effect of fuel structure on the formation of polycyclic aromatic hydrocarbon (PAH) and soot has been emphasized in diffusion flames^1-5, since radicals influencing the formation and growth, such as acetylene (C₂H₂) and propargyl (C₃H₃), are affected by fuel structure and its pyrolysis characteristics. The species generated from fuel pyrolysis lead to incipient ring formation such as benzene (C₆H₆) and naphthalene (C₁₀H₈), which grow to PAHs and finally to soot^6-9.

It has been shown that the mixing of ethylene and propane considerably enhances the formations of PAHs and soot as compared to those with respective pure fuels¹. This enhancement has been explained based on the interaction between acetylene and propargyl species. Acetylene has been known to be an important species for the hydrogen-abstraction-C₂H₂-addition (HACA) mechanism and propargyl for both the benzene ring formation through the propargyl recombination reaction and PAH growth through odd-carbon chemistries^10-12. The mixing of propane in ethylene fuel could enhance the production of soot and PAHs by providing extra propargyl radicals from the dehydrogenation of propane, together with the abundant acetylene.

Recently, it has been observed that the addition of dimethyl ester (DME: CH₃OCH₃) in ethylene fuels in coflow diffusion flames increased the maximum soot volume fraction in the ethylene flames, even though ethylene is a much sootier fuel than DME¹³. The detailed species measurements by McEnally & Pfefferle¹³, suggest that the DME increases concentrations because it decomposes to methyl radical, which promotes the formation of propargyl radical (C₃H₃) through C₁ + C₂ addition reactions and consequently the formation of benzene through propargyl self-reaction. The present study aims to provide better insight into the synergistic effects of propane and DME addition on soot formation of ethylene-air flames.

 

2. Experimental Method

2.1 General apparatus

The investigation of the synergistic effects on soot formation was carried out in the co-flow ethylene-air diffusion flames. The co-flow burner is of a standard type and consists of an inner brass fuel tube (i.d. 8 mm) for fuel supply surrounded by an outer brass tube (i.d. 80 mm) for the air flow. The fuel passage contains screens and 3 mm glass beads to provide a uniform exit flow profile. The larger air passage also utilizes a series of screens with a section filled of 3.0 mm glass beads. A ceramic honeycomb section is used as the final section of the air-flow passage for flow stabilization and uniformity in the outer tube. The fuel tube extends 4 mm beyond the exit plane of the outer tube. The flow conditions chosen for these studies always result in an over-ventilated diffusion flame. The fuel used was commercially pure grade ethylene (>99.9 %). The oxidizer was 76 % N₂/24 % O₂. Two fuels (propane (C₃H₈) and DME) were selected to compare the synergistic effect on the soot formation of ethylene-air flames. Propane was selected because its synergistic effects on ethylene-air flames are well known¹. The propane results will provide a point of comparison for the synergistic effects of DME. DME was selected because it has been the subject of numerous combustion studies and has been investigated in diesel engines as an additive to reduce soot¹⁴. DME has no carbon-carbon (C-C) bond; the absence of C-C bound is believed to be responsible for the extremely low soot emissions from engines fuelled with DME. Ethylene (C₂H₄) was chosen as the base fuel because first, it has a much higher propensity to soot, which makes it relatively easy to utilize optical diagnostic equipment; second, C₂H₄ has been widely used in studies of soot formation for many years, so a large database exists upon which comparison can be made.

For current experiments, small amounts of propane (14%) and DME (15%) were added to the fuel side of the ethylene diffusion flames, while the total carbon flow rate has been kept constant for each flame. The 15% DME addition was selected because the measurements by two of the authors (Lee and Nam) in diffusion DME-air flames have shown that the 15% DME has strong synergistic effects.

Basic flow rates for the experiments were 180 mL/min and 60 L/min for ethylene and air respectively. These values were chosen because they gave the most stable flame. The flow rates of fuels and air were measured using a flow control system consisting of mass-flow controllers and calibrated mass-flow meters.

2.2 Laser scattering and extinction optics

A standard laser-light extinction technique was employed to measure the soot volume fraction distributions, soot particle diameter and number density of soot particles within the tested flames. Figure 1 shows the optical arrangement for the laser-light scattering and extinction (LS/E) measurements. An Argon-ion laser (Spectra Physics Stabilite 2017-05) emitting vertically polarized light at a wavelength (λ) of 514.5 nm was used as a light source. The laser power was set to either 1 W for scattering tests or 0.5 W for extinction tests. The laser beam was first chopped at 1 kHz with a mechanical chopper. Then, it was focused to a beam waist of 0.3 mm at the burner centre with a 500-mm-focal-length lens. The optics for collecting scattering signals consisted of an optical fibre module, polarizer, circular aperture collection lens, pin-hole aperture and laser line filter of 514.5 nm. Neutral density filters were used to reduce the scattered light intensity when it was required. The scattering signal optics was placed at 90° to the laser beam. The scattering signal from the optical fibre module was measured with a lock-in amplifier (Spectra-Physics Stabilite; SR830). The output signals from the lock-in amplifier were transferred to a computer via a data-acquisition system using the Labview program for storage and subsequent data analysis. For each test location, 5000 data points were collected over 5 s to measure time-averaged scattering and extinction coefficients with less than 10% experimental uncertainties. This gives uncertainties for soot volume fraction and soot particle size at 95% confidence interval of about ± 5 % of the mean values.

 

 

Before the experiments were performed, the scattered light detection systems were calibrated to account for the effects of the incident laser power, sample volume, light collection efficiency, optical fibre module sensitivity and electronic gain of the system. The calibration was accomplished by passing C₂H₄, which has a known Rayleigh scattering cross-section through the fuel passage of the burner and measuring the resulting scattered light intensities. For checking the optical alignment and verifying the operation of the entire optical setup, the scattered light detection systems were calibrated by using nitrogen. Nitrogen was separately supplied to the burner to measure the scattering signals under the exact conditions of the actual flame experiments.

2.3 Theoretical methods

For spherical soot particles which are small compared to the wavelength of light, the volume fraction (fv), was determined from the dimensional extinction coefficient data using the following widely known relationships derived from small particle Mie theory:

where λ(mm) is the laser wavelength, and is a function of the refractive index of the soot, which is calculated as:

The use of this equation assumes that the soot particles are within the Rayleigh size limit, and hence the scattering contribution to the total measured extinction is negligible. For this study, an index of refraction of = 1.57-0.56i was used as suggested by Smyth and Shaddix¹⁵. The true value of soot's refractive index has been debated in the literature, and it has been reported to be slightly affected by such conditions as temperature^15,16, C/H ratio, and extent of agglomeration¹⁵. Thus, for ease of comparing results to previous studies, the common value of = 1.57-0.56/ for soot refractive index was used. This uncertainty in the refractive index propagates into uncertainty of the calculated soot volume fractions.

The extinction coefficient at each radial position was determined by using Bouguer's Law:

where I is the incident laser intensity measured by the photo-detector with the flame on, I_o is the laser intensity with the flame off, Kext (1/mm) is the local laser extinction coefficient, and l (mm) is the path length of the laser through the sampling volume.

Average soot particle size (d_p) can be determined from:

At a particular flame location, the volumetric soot scattering coefficient was found from the measured scattering intensities for soot, , and calibration gas , and the known scattering coefficient of calibration gas

here, N is the number of concentration, which is 2.5 × 10¹⁹ for an ideal gas and C_alib is the angular differential cross section.

There were two important considerations during the implementation of equation 5. First, neutral density filters must be used during the flame tests because of the orders-of-magnitude difference in scattering intensities between the soot particles and the calibration gases. Second, soot scattering intensity must be corrected for the attenuation of the laser light as it travelled from the edge of the flame to the measurement location and vice versa. This correction for self absorption through the entire flame ensured that the laser power reaching the measurement volume in the flame was the same as that of the non absorbing calibration gas.

With the available calculated soot particle size (d_p) and the soot volume fraction (f_v), and under the assumption of point contact among particles sizes forming each aggregate, the total number of soot particles sizes per unit volume (n_p) can be calculated of soot particles per unit volume as follows:

 

3. Results and Discussions

3.1 Flame appearance

Figure 2 presents the digital pictures of the pure C₂H₄-air flame, the (14 % C₃H₈ + 86 % C₂H₄)-air flame and (15% DME + 85 % C₂H₄)-air flame under the same conditions. The images of all flames were taken with same exposure time. At the chosen flow rate, the C₂H₄-air flame burns slightly below the smoke point, while the (14 % C₃H₈ + 86 % C₂H₄)-air flame and (15 % DME + 85 % C₂H₄)-air flame burn above their smoke point. The flame heights were 77 mm, 74 mm and 73 mm for the (14 % C₃H₈ + 86 % C₂H₄)-air flame, (15 % DME + 85 % C₂H₄)-air flame and the C₂H₄-air flame respectively. The luminosity of the flame reflects the radiation from soot particles, and the more soot in the flame, the brighter the flame. The (14 % C₃H₈ + 86 % C₂H₄)-air flame were generally brighter than those of C₂H₄-air flame and the (15 % DME + 85 % C₂H₄)-air flame. It can also be seen that there is a gap between burner surface and the luminous region of the flame, which indicates that there is no soot at the location near the burner surface, and the soot begins to appear thereafter.

 

 

3.2 Integrated soot volume fractions

The synergistic effect of the ethylene-propane mixture on soot formation was investigated. Figure 3 shows the variation in integrated soot volume fraction obtained by integrating the soot volume fraction with respect to radius, versus the distance (z) above the burner exit, when propane is added to ethylene at various propane ratios (χ). The results obtained by one of the authors of the current study 1 using a 12 mm internal diameter burner. In this set of experiments, the outer air velocity and the total carbon flow rate were maintained at 20.0 e ratio χ is defined as χ = Q_prop/Q_tot, where Q_prop is the volumetric flow rate of propane and Q_tot is the total volumetric flow rate of ethylene and propane. Soot volume fraction increases initially with height, attains a peak value, then decreases with axial distance and eventually vanishes near the tip of the flame. The results show that the maximum soot volume fraction varies nonmonotonically with c such that the synergistic effects are exhibited. For example, the maximum soot volume fraction for χ = 0.14 are larger than the corresponding maximum value for χ = 0.0 or 1.0. The peak value of the maximum integrated soot volume fraction, occurs at χ = 0.14, as shown in figure 4. The nonlinear synergistic effects of ethylene-propane mixtures on soot formation is clearly demonstrated. Based on these results, 14 % propane addition was selected for current study.

 

 

 

 

Figure 5 shows the integrated soot volume fractions versus the distance (z) above the burner exit for the C₂H₄-air flame, the (14 % C₃H₈ + 86 % C₂H₄)-air flame, and the (15 % DME + 85 % C₂H₄)-air flame under similar conditions. The integrated soot volume fraction increases initially with height attains a peak value in the flame, and thereafter decreases along the flame height. This trend is observed for all flames. The figure shows that the addition of either propane or DME to the fuel enhances the formation of soot in the ethylene-air diffusion flame. However, although the fraction of added propane (14 %) is lower than that of added DME (15 %), the enhancement of soot volume fraction due to the addition of propane is more significant. This synergistic effect of propane in soot formations has been attributed to the competition between the incipient ring formation and the subsequent growths of PAH and soot⁵. For example, the formation of a benzene ring from the propargyl recombination can be more effecttive for propane-rich flames, since propargyl can maintain relatively high concentrations with the addition of propane through dehydrogenation reactions^1,2. Consequently, the production of PAHs can be enhanced by the abundance of incipient rings. Meanwhile the concentration of acetylene and the adiabatic flame temperature will be higher in ethylene-rich flames. Thus, the growths of PAHs and soot through the temperature-sensitive HAC Apathways will be more pronounced. It has been reasoned that the synergistic effect occurs for propane mixture fuels by the interaction between the incipient benzene ring formation from propargyl recombination and the PAH and soot growth through the HACA pathways⁵. On the other hand, the dissociation of DME undergoes hydrogen abstraction followed by bond scission, leading to the formation of methyl and formaldehyde (CH₂O). Formaldehyde is entirely converted into HCO by hydrogen abstraction. HCO is finally converted into CO through hydrogen abstraction reactions. Due to the strength of the C=O bond, CO will not contribute to the production of aromatic species so any carbon from the DME that produces CO is considered to be removed from the reaction pathway leading to aromatic species and soot. Therefore, the carbon in CH₂O does not contribute to PAH formation and soot growth, and thus 50 % of the carbon in DME makes no contribution to soot formation. Though the other 50 % of the carbon in DME is converted into CH₃, most CH₃ makes no contribution to PAH formation. This is due to the active nature of CH₃ radicals. The methyl production/recombination routes convert approximately 30 % of the carbon that makes potential contribution to PAH and soot growth¹⁴. Unlike propane addition, where the synergistic effects are produced by the competition between the incipient ring formation and the subsequent growths of PAH and soot, for DME, it comes mainly from the methyl via propargyl radical (C₃H₃) through C₁+ C₂ addition reactions, and consequently the formation of benzene through propargyl self-reaction. As can be seen in the figure, the addition of DME to ethylene diffusion flame has less of a synergistic effect than propane.

 

 

3.3 Soot volume fraction distributions

Figures 6a-c display the radial variations of the soot volume fraction for the C₂H₄-air flame, the (14 % C₃H₈+ 86 % C₂H₄)-air flame, and the (15 % DME + 85 % C₂H₄)-air flame at different heights above the burner exit. In this study, eight axial flame locations were analyzed. The first two locations represent the soot inception regions (z = 10 mm and 15 mm), the next locations (from z = 20 mm to 40 mm) represent soot surface growth regions and final location (z = 45 mm) represents soot oxidation region. The data presented in the figures was only half of the flame width due to the axisymmetric conditions of the flames. At lower locations, the peak radial soot volume fraction occurs primarily in the outer flame regions. This trend differs at the higher locations, whereby the soot volume fractions occurred at about 2-3 mm from the flame centreline, with significant soot volume fractions occurring inwards to the centreline. This trend is observed for all flames. In the case of the C₂H₄-air flame, the overall maximum soot volume fraction through the entire flame was approximately 10.7 ppm measured at z =35 mm (see figure 6a).

 

 

 

 

 

 

Figure 6b presents the radial distributions of the soot volume fractions in the (14 % C₃H₈ + 86 % C₂H₄)-air flame. The maximum soot volume fraction of approximately 16 ppm was measured at z = 40 mm above the burner.

Figure 6c shows the radial distributions of the soot volume fractions in the (15 % DME + 85 % C₂H₄)-air flame at different axial locations. The overall maximum soot volume fraction through the entire flame appears to be approximately 13.5 ppm and was measured at z = 35 mm. Comparing these three cases, the (14 % C₃H₈ + 86 % C₂H₄)-air flame displays the stronger synergistic effect than the (15 % DME + 85 % C₂H₄)-air flame. As mentioned earlier the synergistic effects of the propane are due to the competition between the incipient ring formation and the subsequent growths of PAH and soot^1,2. DME has less of the synergistic effect than propane because only about 30 % of its carbon contributes to PAH and soot growth¹⁴.

3.4 Soot aerosol properties

Table 1 lists the overall maximum soot particle size (d_p,max), along with the specific axial locations (Z) where these values were measured in the tested flames. The table also lists the values of soot volume fraction (f_v) at the (d_p,max) values as well as the number density of soot particles.

 

 

As seen in the table, d_p,max at z = 20 mm above the burner exit for the (14 % C₃H₈ + 86 % C₂H₄)-air flame is approximately 18.2 % and 6.45 % higher than that for the C₂H₄-air flame and the (15 % DME + 85 % C₂H₄)-air flame respectively. While the (14 % C₃H₈ + 86 % C₂H₄)-air flame has a high f_v, it has the highest d_p,max which results in this flame having the lowest N_p value. These results suggest that the higher amount of soot volume fraction for the (14 % C₃H₈ + 86 % C₂H₄)-air flame is caused by the growth of the soot particles due to the condensation of growth species, C₂H₂ and/or PAH on the soot particle surface. In the case of the (15 % DME + 85 % C₂H₄)-air flame and the C₂H₄-air flame, the lower amount of the soot volume fraction is due to nucleation process and slower particle size growth rates after particles are incepted. However, it should be noted that the contribution of nucleation to total soot is not significant in diffusion flame. At z = 30 mm, for all tested flames, the dominance of surface growth rate is apparent from the increasing of d_p,max and decreasing of the N_p value. At this location, d_p,max of the (14 % C₃H₈15 % + 86 % C₂H₄)-air flame becomes higher than the (15 % DME + 86 % C₂H₄)-air flame and C₂H₄-air flame by 2.4 % and 25.5 % respectively. At the upper location (z = 40 mm), d_p,max for the (14 % C₃H₈ + 86 % C₂H₄)-air flame increases, while N_p value continuously decreases. On the other hand, for the (15 % DME + 85 % C₂H₄)-air flame and C₂H₄-air flame, d_p.max values decrease, while N_p values continuously increase. The shrinking of d_p.max for (15 % DME + 85 % C₂H₄)-air flame and C₂H₄-air flame means the beginning of the oxidation process. Thus soot oxidation process for these two cases starts earlier. In these flames, soot formation and oxidation processes proceed at the same time, with the soot formation being dominant up to the maximum soot volume fraction location, and soot oxidation dominating the process thereafter. The relationship between soot concentration and soot particle size, however, is complex due to a varying degree of soot nucleation growth and oxidation rates. The soot growth rates is higher with the (14 % C₃H₈ + 86 % C₂H₄)-air flame than the (15 % DME + 85 % C₂H₄)-air flame. Thus the synergistic effects are more pronounced in ethylene-propane diffusion flames and less pronounced in ethylene-propane diffusion flames.

 

4. Conclusion

The synergistic effects of ethylene-propane and ethylene-DME mixtures on soot formation were studied in axisymmetric co-flowing acetylene-air laminar diffusion flames using the laser light-extinction technique. The (14% C₃H₈ + 86% C₂H₄)-air flame, the (15% DME + 85 %C₂H₄)-air flame and C₂H₄-air flame were investigated and results on soot volume fraction, soot particle size, and soot number density were compared. The main findings from this study are as follows:

□ Soot volume fractions in ethylene-propane and ethylene-DME mixture flames are higher than those of the pure ethylene flames. The ethylene-propane mixture produced a larger increase in soot than the ethylene-DME mixture flame.

□ The synergistic effects are more pronounced in ethylene-propane diffusion flames than in ethylene-DME diffusion flames. The most reasonable explanation of the weak synergistic effects of ethylene-DME diffusion flames is that during DME decomposition methyl (CH₃) and formaldehyde (CH₂O) are produced. Formaldehyde is entirely converted into HCO by hydrogen abstraction and HCO is finally converted into CO through hydrogen abstraction reactions. Due to the strength of the C=O bond, CO produced will not contribute to the production of aromatic species. Hence, CH₂O does not contribute to PAH formation and soot growth. Most of CH₃ makes no contribution to PAH formation. This is due to the active nature of CH₃ radicals. The CH₃ production/ recombination routes convert approximately 30 % of the carbon that makes potential contribution to PAH and soot growth¹⁷ via propargyl radical (C₃H₃) through Cl + C2 addition reactions, and conseuently the formation of benzene through propargyl self-reaction. This is contrary to the decomposition of propane-ethylene flame, where abundant of propargyl radicals are produced via dehydrogenation process of propane. These proparrgyl radicals together with the abundant acetylene are respon sible for the strong synergistic effects ethylene-propane flame.

 

References

1. Hwang JY, Chung SH, Kang HG and Lee W, Synergistic effect of ethylene-propane mixture on soot formation in laminar diffusion flames, Combustion and Flame, 1997, 191, 427-434.         [ Links ]

2. Hwang JY, Chung SH and Lee W, Twenty-Seventh International Symposium on Combustion, Effects of oxygen and propane addition on soot formation in counterflow ethylene flames and the role of C₃ chemistry, The Combustion Institute, Pittsburgh, 1998, 27, 1531-1538.

3. Vandsburger U, Kennedy I, and Glassman I, Sooting counter-flow diffusion flames with varying oxygen index, Combustion Science and Technology, 1984, 39, 263-285.         [ Links ]

4. Gomez A, and Glassman I, Quantitative comparison of fuel soot formation rates in laminar diffusion flames, Twenty-First International Symposium on Combustion, The Combustion Institute, Pittsburgh, 1986, 21(1), 1087-1095.

5. Hura HS and Glassman I, Fuel oxygen effects on soot formation in counter-flow diffusion flames, Combustion Science and Technology, 1987, 53, 1-21.         [ Links ]

6. Frenklach M, Clary D W, William C, Gardiner JR and Stephen ES, Detailed kinetic modelling of soot formation in shock-tube pyrolysis of acetylene, Proceedings of the Combustion Institute 1984, 20, 887-901.         [ Links ]

7. Wang H and Frenklach M, A detailed kinetic modeling study of aromatics formation in laminar premixed acetylene and ethylene flames, Combustion and Flame, 1997, 110, 173-221.         [ Links ]

8. Frenklach M and Wang H, Detailed modeling of soot particle nucleation and growth, Twenty-Third International Symposium on Combustion, The Combustion Institute, Pittsburgh, 1991, 23, 1559-1566.

9. Frenklach M, On Surface Growth Mechanism of Soot Particles, Twenty-sixth International Symposium on Combustion, The Combustion Institute, Pittsburgh, 1996, 26, 2285 - 2293.

10. D'Anna A and Violi A, A kinetic model for the formation of aromatic hydrocarbons in premixed laminar flames Proceedings of the Combustion Institute, 1998, 27, 425-433.         [ Links ]

11. Miller JA and Melius CF, Kinetic and thermodynamic issues in the formation of aromatic compounds in flames of aliphatic fuels, Combustion and Flame, 1992, 91, 21-39.         [ Links ]

12. Marinov NM, Pit WJ, Westbrook CK, Castaldi MJ and Senkan SM, Modeling of aromatic and polycyclic aromatic hydrocarbon formation in premixed methane and ethane flames, Combustion Science and Technology, 1996, 34-42.

13. McEnally CS and Pfefferle L, The effects of dimethyl ether and ethanol on benzene and soot formation in ethylene nonpremixed flames, Thirty-First International Symposium on Combustion, The Combustion Institute, Pittsburgh, 2007, 31, 603-610.

14. Miyamoto N, Ogawa H and Arima T, Improvement of diesel combustion and emissions with addition of various oxygenated agents to diesel fuel, SAE Paper No. 962115, 1996.

15. Santoro RJ and Shaddix CR, in: Kohse-Höinghaus K and Jeffries JB (Eds.), Applied Combustion Diagnostics, Chapter 9: Laser-Induced Incandescence, Taylor and Frances, New York, 252, 2002.

16. Smyth KC and Shaddix CR, The elusive history of m = 1.57 - 0.56i for the refractive index of soot, Combustion and Flame, 1996, 107(3), 314-320.         [ Links ]

17. Wu J, Ki HS, Litzinger T, Lee S-Y, Santoro R and Linevsky M, Reduction of PAH and soot in premixed ethylene-air flames by addition of dimethyl ether, Combustion Science and Technology, 2006, 178(5), 837-863.         [ Links ]

 

 

Received 7 May 2007
Revised form 4 October 2007
Accepted 7 October 2007