Experimental comparison of soot formation in turbulent flames of Diesel and surrogate Diesel fuels

Soot formation is compared in turbulent diffusion flames burning a commercial Diesel and two Diesel surrogates containing n-decane and α-methylnaphthalene. A burner equipped with a high-efficiency atomisation system has been specially designed and allows the stabilisation of liquid fuels flames with similar hydrodynamics conditions. The initial surrogate composition (70% n-decane, 30% α-methylnaphthalene) was previously used in the literature to simulate combustion in Diesel engines. In this work, a direct comparison of Diesel and surrogates soot tendencies is undertaken and relies on soot and fluorescent species mappings obtained respectively by Laser-Induced Incandescence (LII) at 1064 nm and Laser-Induced Fluorescence at 532 nm. LIF was assigned to soot precursors and mainly to high-number ring Polycyclic Aromatic Hydrocarbons (PAH). The initial surrogate was found to form 40% more soot than the tested Diesel. Consequently, a second surrogate containing a lower α-methylnaphthalene concentration (20%) has been formulated. That composition which presents a Threshold Soot Index (TSI) very close to Diesel one is also consistent with our Diesel composition that indicates a relatively low PAH content. The spatially resolved measurements of soot and fluorescent soot precursors are quite identical (in shape and intensity) in the Diesel and in the second surrogate flames. Furthermore the concordance of the LII temporal decays suggests that a similar growth of the primary soot particles has occurred for Diesel and surrogates. In addition, the comparison of the LII fluence curves indicates that physical/optical properties of soot contained in the different flames might be similar. The chemical composition present at the surface of soot particles collected in Diesel and surrogate flames has been obtained by laser-desorption ionisation time-of-flight mass spectrometry. An important difference is found between Diesel and surrogate samples indicating the influence of the fuel composition on soot content.     

A comparative study of nanoparticles in premixed flames by scanning mobility particle sizer, small angle neutron scattering, and transmission electron microscopy

Scanning mobility particle sizer (SMPS) and transmission electron microscopy (TEM) studies were conducted for TiO₂ and soot particles. The TiO₂ particles were produced from a premixed stagnation ethylene-oxygen-argon flame (? = 0.36) doped with titanium tetraisopropoxide. Soot was generated from a burner-stabilized premixed ethylene-oxygen-argon flame (? = 2.5). The close agreement among SMPS, TEM, and X-ray diffraction results for TiO₂ nanoparticles demonstrates that the probe sampling/mobility measurement technique is accurate for on-line analysis of the size distribution of particles as small as 3 nm in diameter. In the case of soot, notable disagreement between the SMPS and TEM sizes was found and attributable to the fact that the soot taken from the flame studied herein is liquid-like and that upon deposition on the TEM grid, the primary particles do not retain their sphericity. This interpretation is supported by measurements with photo ionization aerosol mass spectrometry, small angle neutron scattering, and thermocouple particle densitometry.     

Sooting tendencies of nonvolatile aromatic hydrocarbons

We have measured sooting tendencies of 72 nonvolatile aromatic hydrocarbons, only five of which have been previously reported in the literature. The tested compounds include long-chain alkylbenzenes up to tridecylbenzene, methyl-substituted benzenes, naphthalenes, biaryls, and polycyclic aromatic hydrocarbons (PAH) with up to four rings. Sooting tendency was defined as the maximum soot concentration f_v,max in a methane/air coflow nonpremixed flame with 5-80 ppm of the aromatic added to the fuel. The f_v,max were converted into Yield Sooting Indices (YSI’s) by the equation YSI = C^∗f_v,max + D, where C and D are constants chosen so that YSI-2-heptanone = 17 and YSI-phenanthrene = 191. The aromatics were dissolved in 2-heptanone and added to the fuel mixture with a syringe pump. Soot concentrations were measured with laser-induced incandescence (LII). The burner and fuel lines were heated; time-resolved soot measurements verified that all of the test compounds were quantitatively transmitted to the flame without losses to the walls. The uncertainties in the results range from ±3 to ±10%.     

Characterization of adsorbed species on soot formed behind reflected shock waves

The present work addresses the soot formation parameters behind reflected shock waves and the identification of adsorbed species on their surface. Soot induction delay times and yields have been experimentally determined in the case of toluene pyrolysis highly diluted in argon for the following conditions: the initial carbon atoms concentration was kept constant around 1 × 10¹⁸ C atoms cm⁻³, reflected shock pressure and temperature ranges of 1135-1600 kPa and 1470-2230 K, respectively. The decrease of the induction time, as the temperature is raised, was described using an Arrhenius type expression while, for the bell-shaped evolution of the soot yield versus the temperature, a modified Gaussian expression was derived. Using TEM analysis, the mean particle diameter was found to decrease from 35 to 20 nm as the temperature is raised from 1475 to 2135 K. The micro-texture of the soot sample was found to vary as the temperature is raised, leading to a more organised structure. The adsorbed species on these soot were characterized using laser desorption/ionization time of flight mass spectrometer. Results indicate that for temperatures below 1600 K, PAHs in the 178-572 atomic mass units (amu) range were identified. PAHs range was limited to 178-374 amu above 1900 K and they were of benzenoid type above 1600 K. The amount of species adsorbed on the soot surface was found to be inversely proportional to the soot yield with a maximum for the lower temperature domain.     

Transported JPDF modelling and measurements of soot at elevated pressures

Accurate measurements and modelling of soot formation in turbulent flames at elevated pressures form a crucial step towards design methods that can support the development of practical combustion devices. A mass and number density preserving sectional model is here combined with a transported joint-scalar probability density function (JDPF) method that enables a fully coupled scalar space of soot, gas-phase species and enthalpy. The approach is extended to the KAUST turbulent non-premixed ethylene-nitrogen flames at pressures from 1 to 5 bar via an updated global bimolecular (second order) nucleation step from acetylene to pyrene. The latter accounts for pressure-induced density effects with the rate fitted using comparisons with full detailed chemistry up to 20 bar pressure and with experimental data from a WSR/PFR configuration and laminar premixed flames. Soot surface growth is treated via a PAH analogy and soot oxidation is considered via O, OH and O₂ using a Hertz-Knudsen approach. The impact of differential diffusion between soot and gas-phase particles is included by a gradual decline of diffusivity among soot sections. Comparisons with normalised experimental OH-PLIF and PAH-PLIF signals suggest good predictions of the evolution of the flame structure. Good agreement was also found for predicted soot volume statistics at all pressures. The importance of differential diffusion between soot and gas-phase species intensifies with pressure with the impact on PSDs more evident for larger particles which tend to be transported towards the fuel rich centreline leading to reduced soot oxidation.     

Experimental and kinetic modeling study of sooting atmospheric-pressure cyclohexane flame

Experimental data and modelling results of the main products and intermediates from a fuel-rich sooting premixed cyclohexane flame were presented in this work. Model predictions well agree with experimental data both in sooting and non-sooting flames. Major and minor species are properly predicted, together with the soot yield. The initial benzene peak was demonstrated to be due to the fast dehydrogenation reactions of the cycloalkane, which gives rise to cyclohexene and cyclohexadiene both via molecular and radical pathways. Once formed cyclohexadiene quickly forms benzene whereas in the postflame zone, benzene comes from the recombination and addition reactions of small radicals, with C₃H₃ + C₃H₃ playing the most important role in these conditions. An earlier soot inception was detected in the cyclohexane flame with respect to a n-hexane flame and this feature is not reproduced by the model that foresees soot formation significant only in the second part of the flame. The model insensitivity of soot to the reactant hydrocarbon was also observed comparing the predictions of three flames of cyclohexane, 1-hexene and n-hexane with the same temperature profile. A sensitivity analysis revealed that soot primarily comes from the HACA mechanism for the three flames, acetylene being the key species in the nucleation. Experimental data on soot inception seem to indicate the importance of the early formation of benzene, that depends on the fuel structure. It is thus important to further investigate the role of benzene and aromatics in order to explain this discrepancy.     

Experimental evaluation of strategies for quantitative laser-induced-fluorescence imaging of nitric oxide in high-pressure flames (1-60 bar)

Nitric oxide laser-induced-fluorescence (NO-LIF) 2-D imaging measurements using a new multi-spectral detection strategy are reported for high-pressure flames (1-60 bar). This work builds on previous research that identified interference LIF from O₂ and CO₂ in high-pressure flames and optimized the choice of excitation strategies as a function of application conditions. In this study, design rules are presented to optimize the LIF detection wavelengths for quantitative 2-D NO-LIF measurements over a wide range of pressures (1-60 bar) and temperatures. Simultaneous detection of LIF in multiple wavelength regions enables correction of the NO signal for interference from O₂ and CO₂ and allows simultaneous imaging of all three species. New experiments of wavelength-resolved 1-D LIF in slightly lean (? = 0.9) and slightly rich (? = 1.1) methane/air flames are used to evaluate the design rules and estimate the NO detection limits for a wide range of flame conditions. The quantitative 2-D measurements of NO in the burnt gas are compared with model calculations (using GRI-Mech 3.0) versus pressure for slightly lean and slightly rich flames. The discussions and demonstrations reported in this study provide a practical guideline for application of instantaneous 1-D or 2-D NO-LIF imaging strategies in high-pressure combustion systems.     

Soot emission reduction in oxygenated co-flow jet flames

A computational study was performed for ethylene/air non-premixed laminar co-flow jet flames using an axisymmetric CFD code to explore the effect of oxygenation on PAH and soot emissions. Oxygenated flames were established using N₂ diluted fuel stream along with O₂ enriched air stream such that the stoichiometric mixture fraction (Ζ_st) is varied but the adiabatic flame temperature is not materially changed. Simulations were carried out using a spatially and temporally accurate algorithm with detailed chemistry and transport. A detailed kinetic model involving 111 species and 784 reactions and a fairly detailed soot model were incorporated into the code. Two different approaches, one with constant flame height and other with constant inlet velocity are comprehensively examined to bring out the effects of changes in flame structure and residence time on soot emissions with respect to Z_st. With increase in Ζ_st, a drastic reduction in the formation of soot precursors (acetylene and benzene) and thus in soot emissions are observed. In the present study, oxygenated flames with Ζ_st ≥ 0.424 are considered as blue flames or completely soot free. For various oxygenated flames a C/O ratio between 0.45 and 0.6 is found to be most favorable for soot formation.     

Mass spectrometric analysis of large PAH in a fuel-rich ethylene flame

Mass spectrometric analysis by laser desorption-time of flight-mass spectrometry (LDI-TOF-MS) was exploited to extend the detection of flame-formed polycyclic aromatic hydrocarbons (PAH) up to the mass limit of the first soot particles (>2000 Da) in the soot formation region of a premixed fuel-rich (C/O = 1) ethylene flame. The typical decreasing intensity of PAH ion peaks with increasing mass was found in the mass range m/z 500-1700 although a slight enrichment in the heavier part of PAH could be observed to occur along the flame axis. The separation by means of size exclusion chromatography (SEC) into two different classes of PAH followed by UV-visible spectroscopy corroborated the mass spectral identification of large mass PAH. Critical examination of mass spectral features and SEC separation was the starting point for speculation about the changes occurring in PAH growth from planar to concave structures which could be important for soot inception mechanisms.     

Modelling of aromatics and soot formation from large fuel molecules

There is a need for prediction models of soot particles and polycyclic aromatic hydrocarbons (PAHs) formation in parametric conditions prevailing in automotive engines: large fuel molecules and high pressure. A detailed kinetic mechanism able to predict the formation of benzene and PAHs up to four rings from C₂ fuels, recently complemented by consumption reactions of decane, was extended in this work to heptane and iso-octane oxidation. Species concentrations measured in rich, premixed flat flames and in a jet stirred reactor (JSR) were used to check the ability of the mechanism to accurately predict the formation of C₂ and C₃ intermediates and benzene at pressures ranging from 0.1 to 2.0 MPa. Pathways analyses show that propargyl recombination is the only significant route to benzene in rich heptane and iso-octane flames. When included as the first step of a soot particle formation model, the gas-phase kinetic mechanism predicts very accurately the final soot volume fraction measured in a rich decane flame at 0.1 MPa and in rich ethylene flames at 1.0 and 2.0 MPa.     

The effects of dimethyl ether and ethanol on benzene and soot formation in ethylene nonpremixed flames

Soot volume fractions, C1-C12 hydrocarbon concentrations, and gas temperature were measured in ethylene/air nonpremixed flames with up to 10% dimethyl ether (CH₃OCH₃) or ethanol (CH₃CH₂OH) added to the fuel. The measurement techniques were laser-induced incandescence, photoionization mass spectroscopy, and thermocouples. Oxygenated hydrocarbons have been proposed as soot-reducing fuel additives, and nonpremixed flames are good laboratory-scale models of the fuel-rich reaction zones where soot forms in many full-scale combustion devices. However, addition of both dimethyl ether and ethanol increased the maximum soot volume fractions in the ethylene flames studied here, even though ethylene is a much sootier fuel than either oxygenate. Furthermore, dimethyl ether produced a larger increase in soot even though neat dimethyl ether flames produce less soot than neat ethanol flames. The detailed species measurements suggest that the oxygenates increase soot concentrations because they decompose 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. Dimethyl ether has a stronger effect than ethanol because it decomposes more completely to methyl radical. Ethylene does not decompose to methyl, so its flames are particularly sensitive to this mechanism; the alkane-based fuels used in most practical fuels do decompose to methyl radical, so the mechanism will be much less important for practical devices.     

Effect of ceria nanoparticles on soot inception and growth in toluene-oxygen-argon mixtures

Soot formation from the combustion of toluene (C₆H₅CH₃) and of two concentrations of nano-sized-ceria-laden toluene was monitored using a shock tube to observe the effect of the organometallic additive on the formation of soot from its point of inception. Two concentrations of ceria, of chemical composition CeO_1.63, were employed to examine the effect on soot production of toluene over the range of temperature 1588-2370 K using two levels of inert gas dilution in which reflected-shock pressure was maintained near 1.5 atm. The ceria nanoparticles were synthesized using a microemulsion technique which employs sodium dioctyl sulfosuccinate (AOT), a surfactant, to retard agglomeration. Introduction of the nanoparticles into the shock tube is achieved using a novel, two-stage injection procedure. Soot yield measurements reveal that the presence of ceria has no direct implications on peak soot concentration near 1950 K. A shift in the parabolic soot profile of toluene in the direction of increased temperature was observed for each concentration of ceria with a larger shift occurring for increased concentration of ceria, although the same effect was exhibited for the toluene-AOT mixtures in absence of ceria, supporting an inefficaciousness of ceria on soot suppression on kinetic timescales. It is evidenced in measured soot delay times that the presence of the surfactant in absence of ceria significantly slows the rate of soot growth for T < 2000 K, while the presence of ceria has a relatively negligible impact. Under conditions of higher fuel concentration, a remarkable decrease in soot accumulation on the shock tube walls was observed in experiments using the ceria-toluene mixtures over that yielded by pure toluene combustion. In the present paper, the authors report the first measurements of nanoparticle-influenced combustion of a hydrocarbon as performed in a shock tube.     

Cyclopentadiene combustion in a plug flow reactor near 1150 K

Cyclopentadienyl (CPDyl) was generated for study by oxidizing and pyrolizing 1,3-cyclopentadiene (CPD) in Princeton’s adiabatic, atmospheric pressure flow reactor. This study used nitrogen carrier gas, initial CPD concentrations from 1000 to 3000 ppm by volume (ppmv), equivalence ratios from fuel lean (? = 0.6) to pyrolytic conditions (? = 100) and initial temperatures from 1100 to 1200 K. The reaction progress was followed from 5 to 150 ms using a water cooled sample probe and GC-FID analysis of C₁-C₁₄ species. The oxidation results show that CPD and CPDyl react via 19 pathways to yield 22 hydrocarbon intermediates. Analysis of the oxidative CPDyl ring opening pathways reveals the importance of the 2,4-cyclopentadienoxy (c-C₅H₅O) β-scission reaction: c-C₅H₅O ↔ CHCH-CHCH-CHO. The fastest theoretical mechanism has a calculated unimolecular high-pressure rate constant of 2.00 × 10¹³e^−7215/T s⁻¹ which is seven orders of magnitude larger at 1150 K than the previous literature estimate. Cyclopentadienone (CPDone) has been assumed to be an important intermediate in C₅ ring oxidation even though it has not been unambiguously identified in the combustion environment. A detection limit of 20 ppmv for CPDone in the present apparatus failed to note any CPDone. A set of mechanistic pathways for the C₅ ring oxidation includes steps to avoid unrealistic CPDone production is presented. The complex mechanism illustrates the need for detailed models to understand the combustion of aromatics and soot precursors. The article stresses the importance of CPDyl in the formation of aromatic rings during combustion, which subsequently leads to polycyclic aromatic hydrocarbons (PAH) and soot precursors.     

Addition of NO2 to a laminar premixed ethylene-air flame: Effect on soot formation

Experiments were conducted on a laminar premixed ethylene-air flame at equivalence ratios of 2.34 and 2.64. Comparisons were made between flames with 5% NO₂ added by volume. Soot volume fraction was measured using light extinction and light scattering and fluorescence measurements were also obtained to provide added insight into the soot formation process. The flame temperature profiles in these flames were measured using a spectral line reversal technique in the non-sooting region, while two-color pyrometry was used in the sooting region. Chemical kinetics modeling using the PREMIX 1-D laminar flame code was used to understand the chemical role of the NO₂ in the soot formation process. The modeling used kinetic mechanisms available in the literature. Experimental results indicated a reduction in the soot volume fraction in the flame with NO₂ added and a delay in the onset of soot as a function of height above the burner. In addition, fluorescence signals—often argued to be an indicator of PAH—were observed to be lower near the burner surface for the flames with NO₂ added as compared to the baseline flames. These trends were captured using a chemical kinetics model that was used to simulate the flame prior to soot inception. The reduction in soot is attributed to a decrease in the H-atom concentration induced by the reaction with NO₂ and a subsequent reduction in acetylene in the pre-soot inception region.     

Computational and experimental investigation of the interaction of soot and NO in coflow diffusion flames

A combined computational and experimental investigation that examines the relationship of soot formation and NO in coflow ethylene air diffusion flames is presented. While both NO and soot formation are often studied independently, there is a need to understand their coupled relationship as a function of system parameters such as fuel type, temperature and pressure. The temperature decrease due to radiative losses in systems in which significant soot is produced can affect flame length and other temperature-dependent processes such as the formation of NO. The results of a computational model that includes a sectional representation for soot formation with a radiation model are compared against laser-induced fluorescence measurements of NO. The sooting characteristics of these flames have been studied previously. Experimentally, a laser near 225.8 nm is used to excite the γ(0, 0) band in NO. Spectrally resolved fluorescence emission is imaged radially, for the (0, 0), (0, 1), (0, 2), (0, 3), and (0, 4) vibrational bands, at varying axial heights to create a two-dimensional image of NO fluorescence. A reverse quenching correction is applied to the computational results to determine an expected fluorescence signal for comparison with experimental results. Modeling results confirm that Fenimore NO is the dominant mechanism for NO production and suggest that for lightly sooting flames (peak soot volume fraction < 0.5 ppm), soot reduces only the Zeldovich NO formation (by a factor of two). For flames with increased soot levels (peak soot volume fraction ∼ 4 ppm), the model indicates not only that Zeldovich NO decreases by a factor of 2.5 through radiation loss, but that non-Zeldovich NO is reduced in the top center of the flame by about 30% through the oxidation of soot.     

The effects of burner stabilization on Fenimore NO formation in low-pressure, fuel-rich premixed CH4/O2/N2 flames

We investigate the effects of varying the degree of burner stabilization on Fenimore NO formation in fuel-rich low-pressure flat CH₄/O₂/N₂ flames. Towards this end, axial profiles of flame temperature and OH, NO and CH mole fractions are measured using laser-induced fluorescence (LIF). The experiments are performed at equivalence ratios between 1.3 and 1.5. The flame temperature is seen to decrease by 200-300 K, with a concomitant decrease in OH mole fraction, upon reducing the total flow rate from 5 to 3 L/min, thus increasing stabilization. At equivalence ratios between 1.3 and 1.5, this decrease in flow rate lowers the maximum CH mole fraction by a factor of 2, and the NO mole fraction by ∼40% in all flames studied. Integrating the reaction rate for CH + N₂ to estimate Fenimore NO formation, using the rate coefficient in GRI-Mech 3.0, and the measured temperatures and CH profiles show very good agreement with the measured NO mole fraction for ? = 1.3 and 1.4, supporting the current choice for this rate. This agreement also shows that the increase in residence time caused by increased stabilization is an important factor in the ultimate impact of the changes in CH mole fraction on NO formation. The results at ? = 1.5 suggest that substantial quantities of fixed nitrogen species, e.g., HCN, are only slowly oxidized in the post-flame zone under these conditions, leading to a significant discrepancy between the measured NO mole fraction and that obtained by integrating over the CH profile. Detailed calculations using GRI-Mech 3.0 predict the experimental results at ? = 1.3 nearly quantitatively, but show increasing differences with the measurements for both CH and NO profiles with increasing equivalence ratio.     

A computational study of soot formation and flame structure of coflow laminar methane/air diffusion flames under microgravity and normal gravity

A numerical study is conducted of methane–air coflow diffusion flames at microgravity (μg) and normal gravity (1g), and comparisons are made with experimental data in the literature. The model employed uses a detailed gas phase chemical kinetic mechanism that includes PAH formation and growth, and is coupled to a sectional soot particle dynamics model. The model is able to accurately predict the trends observed experimentally with reduction of gravity without any tuning of the model for different flames. The microgravity sooting flames were found to have lower temperatures and higher volume fraction than their normal gravity counterparts. In the absence of gravity, the flame radii increase due to elimination of buoyance forces and reduction of flow velocity, which is consistent with experimental observations. Soot formation along the wings is seen to be surface growth dominated, while PAH condensation plays a more major role on centreline soot formation. Surface growth and PAH growth increase in microgravity primarily due to increases in the residence time inside the flame. The rate of increase of surface growth is more significant compared to PAH growth, which causes soot distribution to shift from the centreline of the flame to the wings in microgravity.     

Intra-cavity second harmonic generation with Nd:YVO4/BIBO laser at 542 nm

We report, for the first time, an efficient intra-cavity second-harmonic generation (SHG) at 1084 nm in a nonlinear optical crystal, BiB₃O₆(BIBO) at the direction of (θ, ?) = (170.1°, 90°), performed with a LD end-pumped cw Nd:YVO₄ laser. With 590 mW diode pump power, a continuous-wave (cw) SHG output power of 19 mW at 542 nm yellow-green color has been obtained using a 1.5 mm-thick BIBO crystal. The optical conversion efficiency was 3.22%. It was found that the output wavelength could be 532 nm, 537 nm or 542 nm according to regulating the angle of BIBO.     

Criteria for selection of components for surrogates of natural gas and transportation fuels

The present paper addressed the production of soot precursors, acetylene, benzene and higher aromatics, by the paraffinic (n-, iso-, and cyclo-) and aromatic components in fuels. To this end, a normal heptane mechanism compiled from sub-models in the literature was extended to large normal-, iso-, and cyclo-paraffins by assigning generic rates to reactions involving paraffins, olefins, and alkyl radicals in the same reaction class. Lumping was used to develop other semi-detailed sub-models. The resulting mechanism for components of complex fuels (named the Utah Surrogate Mechanism) includes detailed sub-models of n-butane, n-hexane, n-heptane, n-decane, n-dodecane, n-tetradecane and n-hexadecane, and semi-detailed sub-models of i-butane, i-pentane, n-pentane, 2,4-dimethyl pentane, i-octane, 2,2,3,3-tetramethyl butane, cyclohexane, methyl cyclohexane, tetralin, 2-methyl 1-butene, 3-methyl 2-pentene and aromatics. Generic rates of reaction classes were found adequate to generate reaction mechanisms of large paraffinic components. The predicted maximum concentrations of the fuel, oxidizer, and inert species, major products and important combustion intermediates, which include critical radicals and soot precursors, were in good agreement with the experimental data of three premixed flames of composite fuels under various conditions. The relative importance in benzene formation of each component in the kerosene surrogate was found to follow the trend aromatics > cyclo-paraffins > iso-paraffins > normal-paraffins. In contrast, acetylene formation is not that sensitive to the fuel chemical structure. Therefore, in formulation of surrogate fuels, attention should be focused on selecting components that will yield benzene concentrations comparable to those produced by the fuel, with the assurance that the acetylene concentration will also be well approximated.     

Pressure and temperature dependence of soot in highly controlled counterflow ethylene diffusion flames

Soot volume fraction and dispersion index were measured by pyrometry in a series of highly controlled counterflow diffusion flames, with peak temperatures, T_max, spanning a few hundred degrees and pressure covering the 0.1–0.8 MPa range. An unprecedented level of control was implemented by selecting flames with a self-similar structure to ensure that the normalized temperature-time history experienced by the reactants was the same, regardless of pressure. The self-similarity was verified by suitably rescaling the transverse coordinate with respect to a characteristic diffusion length. At constant T_max, the soot volume fraction increases approximately by two orders of magnitude as the pressure is raised from 1 atm to 4 atm, and by one to two additional orders of magnitude with an additional doubling of the pressure to 8 atm. At constant pressure, the soot load spans two to three orders of magnitude and soot formation exhibits increased sensitivity to temperature as the pressure is raised. Soot inception occurs near the flame, with an increase in soot concentration that becomes steeper at higher T_max. The increase is accompanied by a decrease in the dispersion exponent that is suggestive of dehydrogenation and aging of the particles and is sharper at higher T_max. Soot experiences continuous growth in a monotonically decreasing temperature field until it is convected away radially at the stagnation plane, with essentially no opportunity for oxidation. Evidence of two distinct mechanisms for soot formation was found: the classic high temperature, high activation energy process affecting soot formed in the vicinity of the flame and followed by dehydrogenation; and a relatively low-temperature, zero activation energy process, associated with the increase in volume fraction at low-temperatures in proximity of the stagnation plane. The latter is tentatively attributed to dimerization of aromatics, as revealed by the concurrent increase in the dispersion index corresponding to an increase in the particle hydrogen content.