Undoped Flames Research Articles

Emission Spectra of Hydrocarbon Flames Doped with Phosphorus-Containing Compounds

Phosphorus-containing compounds, including organophosphorus substances, are valuable laboratory flame additives for understanding the combustion behavior of sarin. Chemiluminescence features of excited intermediate species are useful markers of combustion, but an effective chemiluminescence-based combustion diagnostic for phosphorus-based compounds is yet to be developed. In this study, chemiluminescence spectra of atmospheric-pressure flames doped with three phosphorus-containing compounds, namely, dimethyl methylphosphonate, diisopropyl methylphosphonate, and phosphoric acid (), were obtained in two flame configurations, namely, spherically expanding flames and piloted liquid-spray flames. The spectra are compared with those obtained from undoped flames fueled by methane and methanol, respectively. Spectra were recorded for the wavelength range of 200–800 nm and for three equivalence ratios (, 1.0, 1.2) of spherical flames. For both flame configurations, the intensity of the doped flame emission increased relative to the neat flames by over an order of magnitude, but no clear spectral features besides a broad, featureless continuum emission could be identified. No significant difference was observed between the spectra of the spray flames doped with and with diisopropyl methylphosphonate. The observance of a continuum emission for all the doped flames is attributed to blackbody emission, likely resulting from high-temperature soot particles in the flame originating from condensed, phosphorus-containing combustion products. However, unique, narrow chemiluminescence features specific to phosphorus combustion could not be identified at this time.

Soot Formation and Growth in Toluene/Ethylene Combustion Catalyzed by Ruthenium Acetylacetonate

ABSTRACT Ruthenium-based compounds are efficient catalysts to enhance combustion performance and suppress soot emission. However, systematic mechanism and in-situ research on reducing soot are rarely addressed. In this study, ruthenium acetylacetonate (Ru(C5H7O2)3, Ru(acac)3) was utilized as the catalyst precursor in different concentration conditions to investigate its impacts on the size, volume, and morphology of the soot particles in the ethylene flames with central toluene injection. The soot particles were detected by in-situ small-angle X-ray scattering (SAXS) and ex-situ transmission electron microscopy (TEM). This study confirms that Ru(acac)3 can suppress the surface growth for the primary soot particles, but no significant change of the morphology has been discovered for the aggregates. Compared with the undoped flame, the volume of the soot particles in the Ru(acac)3-doped flames is lower and the particle size is smaller, indicating a valid inhibition effect of Ru(acac)3 on the soot emission by affecting its surface growth and oxidation process. Despite the remarkable distinction between the undoped and doped flames, no obvious difference is found between the flames with different catalyst concentrations.

Effect of nickel acetylacetonate addition on soot inception and growth in an ethylene flame studied by using in situ small-angle X-ray scattering

Small-angle X-ray scattering (SAXS) technique has been used to study soot formation and growth in an ethylene flame doped with Ni(acac)2 (nickel acetylacetonate) at atmospheric pressure. An undoped flame under the same conditions was also studied to compare with the doped flame. The results show that the Ni(acac)2 addition affects particle size, shape and morphology significantly at certain flame heights. It has reducing effect on the primary soot particle size at the flame heights between 17 and 34 mm. It delays the growth of primary soot particles at soot growth region. The soot volume fraction in the doped flame is smaller than that in the undoped flame by factors of 1.5–4 at different heights above the burner. In the most heights of the soot surface growth region (HAB=13–30 mm), the Porod exponents for the doped flame are smaller than those in the undoped flame, suggesting that the addition of Ni(acac)2 in the ethylene flame induces a rougher surface and more irregular shape of the soot primary particles.

Synchrotron Radiation Studies of Additives in Combustion, III: Ferrocene

Small-angle X-ray scattering (SAXS) has been used to examine the growth of soot primary particles (∼15–20 nm) and aggregates of primaries (>30 nm) formed in an ethylene flame doped with ferrocene. It is found that, early in the flame, the size of the primary particles formed when ferrocene is added tracks that for an undoped flame. Higher up (at height z above the burner greater than 18 mm), both primary and aggregate sizes are significantly smaller with ferrocene doping compared to the undoped flame. Within experimental error, the number densities of primary (N1) particles in the ferrocene flame are about the same as for the pure ethylene flame at low heights and higher (factor of 10) at larger heights (above 18 mm). On the other hand, the number concentration of aggregate particles (N2) in the doped flame is always greater than in the pure ethylene flame regardless of the height but with drops and increases in the ferrocene-doped flame case that can be correlated with corresponding increases and decreases in aggregate size. SAXS data at sufficiently long residence times in the flame show, through the value of the Porod exponent, that the particle surfaces of primaries and aggregates are much rougher in the doped versus undoped flame, which we interpret as being the effect of iron oxide inclusion that produces enhanced oxidation inside the soot material.

Modeling of 1,3-hexadiene, 2,4-hexadiene and 1,4-hexadiene-doped methane flames: Flame modeling, benzene and styrene formation

In this work, we have developed a detailed chemical kinetic model and reacting flow simulation for the hexadiene-doped 2-d methane diffusion flames studied experimentally by McEnally and Pfefferle. The GRI-Mech 2.11 methane oxidation and Lawrence Livermore butane oxidation mechanisms were used as the base mechanism to which hexadiene chemistry generated by Reaction Mechanism Generator (RMG) was added. Some important chemically activated pathways leading to aromatic species formation, including the reactions on C 5H 7, C 6H 10, C 6H 9, C 6H 7, C 8H 8 and C 8H 9 potential energy surfaces, are examined in great detail using quantum chemistry (CBS-QB3) and master equation analysis as implemented in Variflex. An efficient program to solve the doped methane diffusion flame was developed. The solver uses the method of lines to solve the species mass balance equation arising in the diffusion flame. It assumes that the temperature and velocity profiles of the doped flame are the same as those of the undoped flame. The mole fractions of various species as predicted by our model are compared to the experimentally measured mole fractions. The agreement between theory and experiments is quite good for most molecules. The added hexadiene dopants to the flame decompose to produce significant amount of cyclopentadienyl radical, which combines with methyl radical to produce benzene. We also show that styrene is formed primarily by recombination of cyclopentadienyl and propargyl radicals, a pathway which to our knowledge, has not been included in prior flame simulations.

Allene addition to a fuel-lean ethylene flat flame

Two fuel-lean C 2H 4/O 2/Ar flames, one doped with allene, were analyzed using molecular-beam mass spectrometry (MBMS) and modeled. Flame conditions were ϕ = 0.70 and 56.4% Ar for the undoped flame and ϕ = 0.69, 56.5% Ar and [allene]/[C 2H 4] = 0.0242 for the doped flame, both at 4.000 kPa (30 Torr) with a 30.6 cm/s burner velocity at 300 K. Modeling and measurements agreed well for the major species, but H, O, and OH were overpredicted. Perturbation effects by allene addition included an increase in propargyl (C 3H 3) and C 6H 6 species in both experiments and the model. Modeling suggested that the major C 6H 6 species is benzene. Measurements done on the Chemical Dynamics Beamline of the Advance Light Source, within Lawrence Berkeley National Laboratory, using photoionization molecular-beam mass spectrometry supported the model predictions by determining that the major C 6H 6 species is benzene (45%), followed by 1,5-hexadiyne (35%), and fulvene (20%). Using the rate constants of Miller and Klippenstein [J. Phys. Chem. A 107 (39) (2003) 7783], the model predicted C 3H 3 self-combination reactions to be the major benzene formation route through direct formation as well as through formation of thermal fulvene and linear C 6H 6, which can then isomerize to benzene. Benzene is destroyed mainly through hydrogen abstraction to form phenyl and through O attack to form phenoxy. A major phenyl destruction path is to benzene. The ultimate destruction of benzene goes through phenyl and phenoxy. Both species undergo CO elimination by O addition to phenyl and decomposition of phenoxy to form cyclopentadienyl, which undergoes further oxidation to form CO, CO 2, and H 2O.

Effect of ferrocene addition on sooting limits in laminar premixed ethylene–oxygen–argon flames

The critical sooting C/O ratio was measured for a series of atmospheric-pressure, laminar premixed ethylene–oxygen–argon flames doped with a small amount of ferrocene. Comparisons of the doped and undoped flames show marked increases in sooting tendency for flames doped with ferrocene. Under the same flame conditions, the critical C/O ratios in doped flames were uniformly lower than those of undoped flames. A five-step reaction mechanism of ferrocene decomposition, leading to the formation of the cyclopentadienyl, was proposed. Detailed kinetic modeling of the experimental flames showed little changes in the flame temperature and the aromatic concentration upon ferrocene doping. The experimental and computational results support the early suggestion that in premixed flames the increase in sooting tendency due to ferrocene addition is the result of induced nucleation by iron oxide nanoparticles. These particles provide a surface to initiate soot surface growth.

Measurement of CrO in flames by cavity ringdown spectroscopy

CrO is an important intermediate in the high temperature oxidation chemistry of chromium containing species. This work reports the first detection of CrO in a flame. The B 5Π-X 5Π electronic transition was probed by cavity ringdown spectroscopy (CRDS) in a lean (ϕ=0.38), low-pressure, flat, laminar hydrogen-oxygen-argon flame seeded with Cr(CO)6. The previous B 5Π-X 5Π CrO spectrum of Hocking et al. (605.0 nm-606.5 nm) is extended from the band head located at 605.6 nm to 614.4 nm. The temperature profiles of the doped and undoped flames were obtained from measurements of OH laser- induced fluorescence. Seeding the flame with Cr(CO)6 increased the flame temperature by approximately 150 K. The concentration profile of CrO was measured as a function of height above the burner. CrO absorption signals were converted to concentration using the measured temperature profile and absorption cross-section calculated from lifetimes by Hedgecock et al. A lower limit peak CrO concentration of 1.6 ppb was found in the flame. Some uncertainty in the cross-section remains. Comparisons to calculations using STANJAN indicate that CrO is present in flames at super equilibrium concentrations.

Aromatic formation pathways in non-premixed methane flames

A kinetic mechanism, previously developed and successfully applied to the prediction of the formation of benzene and aromatics in different flame conditions, was applied to assess the importance of the various benzene and aromatic formation pathways in non-premixed flames. Four sets of data were tested: the methane flame and the same flame doped with toluene, ethylbenzene, and tert-butylbenzene, as studied by Anderson and co-workers. The model predicts, with good accuracy, the growth of hydrocarbons and the formation of benzene and aromatic species. The modeling shows that in the undoped methane flame, benzene formation is controlled by propargyl radical combination. Acetylene addition to C4 radicals contributes a moderate amount, whereas toluene decomposition is insignificant. The predictions are almost unaffected by the fulvene pathway. Benzene is strongly perturbed by dopant addition to methane. Predictions agree quite well with benzene concentrations in the undoped flame and agree with the increase in benzene concentration when alkylbenzenes are added. Key reactions leading to the formation of naphthalene are the propargyl addition to benzyl radicals, and, to a lesser extent, the hydrogen-abstraction acetylene-addition mechanism. Cyclopentadienyl radical combination, which is the dominant route in premixed and partially premixed flames, is insignificant in these flame conditions.

Detailed modeling and laser-induced fluorescence imaging of nitric oxide in a NH 3-seeded non-premixed methane/air flame

In this paper, we study the formation of NO in laminar, nitrogen-diluted methane diffusion flames that are seeded with ammonia in the fuel stream. We have performed numerical simulations with detailed chemistry as well as laser-induced fluorescence (LIF) imaging measurements for a range of ammonia injection rates. For comparison with the experimental data, synthetic LIF images are calculated based on the numerical data accounting for temperature and fluorescence quenching effects. We demonstrate good agreement between measurements and computations. The LIF corrections inferred from the simulation are then used to calculate absolute NO mole fractions from the measured signal. The NO formation in both doped and undoped flames occurs in the flame sheet. In the undoped flame, four different mechanisms contribute to NO formation. The present calculations show that the most important pathway is prompt NO, followed by the NNH mechanism, thermal NO, and the N2O mechanism. As the NH3 seeding level increases, fuel-NO becomes the dominant mechanism and N2 shifts from being a net reactant to being a net product. Nitric oxide in the undoped flame as well as in the core region of the doped flames is underpredicted by the model; we attribute this mainly to inaccuracies in the NO recycling chemistry on the fuel-rich side of the flame sheet.

The effects of ferrocene addition on soot particle inception and growth in premixed ethylene flames

The effects of the addition of dicyclopentadienyl iron, ferrocene, on soot particle inception and growth were studied in atmospheric pressure, premixed ethylene flames with C O = 0.71–0.83 at 0.015–0.46% ferrocene by weight of the fuel. The rate of surface growth of soot in undoped and doped flames was measured using laser light scattering and light extinction techniques to determine the particle size and number density as a function of residence time in the flame. The final soot yields in the ferrocene doped flames were enhanced by factors of 13.5-1.2 over the range of flame C O ratios studied. This enhancement decreased with increasing C O ratio for the same concentration of iron in the flames. The spatial distribution of the elements iron and carbon in the particles was determined with Auger spectroscopy. The iron was found to be concentrated in the cores of the particles. The results of a steady-state one-dimensional nucleation model which indicated that the iron oxide particles would nucleate well in advance of the soot inception time were consistent with this elemental stratification and the observation of the onset of carbon deposition in doped flames at earlier residence times than in undoped flames of the same C O ratio. The iron oxide particle sizes and number densities predicted from the model indicated that these particles were both numerous and small enough to be seed particles on which carbon deposition could take place. The importance of the activity of the surfaces of the solid in determining the rate of surface growth of soot was investigated. Lower rates of surface growth measured in doped flames than in undoped flames of the same C O ratio at early residence times indicated that the nucleated iron oxide particles were a less active surface on which carbon deposition took place than were the young soot particles. The rate of surface growth decreased by a factor of two over the residence time in the flames in all undoped flames and in doped flames of C O ratio above 0.80. For the doped flames of C O ratio below 0.80, the rate of surface growth did not decay in the late residence time regions of the flames. The break between the doped flames which did and did not display the persistent growth rates correlates with the relative magnitudes of the characteristic diffusion times for iron through carbon based on the final particle size formed in the flames and the particle growth time.

Formation and interconversion of fixed-nitrogen species in laminar diffusion flames

Laminar diffusion flames, fed with fuel mixtures containing CH4, were probed and the analyses reduced by Bilger's simplified method to get approximate estimates of the rates of formation of the fixed-N species: NO, HCN, NH3. The results show that the Zeldovich mechanism is the main source of the sum of fixed-N. Much of the local fixed-N is present as HCN (less as NH3) in hydrocarbon-containing parts of the flame; but these originate mainly from NO diffusing into such regions, and they mostly revert to NO again after the hydrocarbons burn out. There is little destruction of the sum of fixed-N species in the rich parts of undoped flames. In the rich parts of doped flames, however, where the concentrations of all forms of fixed-N are increased, a significant destruction of fixed-N is observed.