Michel Carlier

A rapid compression machine investigation of oxidation and auto-ignition of n-heptane: measurements and modeling

n-Heptane oxidation and auto-ignition in a rapid compression machine is studied in the low and intermediate temperature regimes at high pressures. Experimental ignition delay times and some phenomenological aspects related to knock in engines are presented, providing additional information at lower temperatures on previously published delays from shock tube experiments. The products of oxidation are identified and time profiles are measured during a two-stage ignition process. Eight C7 heterocycles, heptenes, lower 1-alkenes, aldehydes, and carbon monoxide are the main species. Their origin is discussed in relation to the isomerization and decomposition of heptylperoxy radicals. The high selectivity observed in the formation of lower 1-alkenes is explained by the scission of the β CC bond of the β-hydroperoxyheptyl radicals weakened by the presence of oxygen atoms. Numerical simulation of the experiments with Warnatzs comprehensive chemical mechanism gives satisfactory results for cool flame and total ignition delays, but fails to reproduce the detailed chemistry before auto-ignition.

Experimental and Modeling Study of Oxidation and Autoignition of Butane at High Pressure

Oxidation and autoignition of stoichiometric, lean (ϕ = 0.8), and rich (ϕ = 1.2) butane-“air” mixtures are studied in a rapid compression machine between 700–900 K and 9–11 bar. Information is obtained concerning cool flames and ignition delays. Product profiles for selected major and minor species are measured during a two-stage ignition process. The presence of C4 heterocycles may be connected to isomerization and decomposition of butylperoxy radicals. The experimental results are compared with numerical predictions of an homogeneous adiabatic model based on the Pitz-Westbrook comprehensive chemical mechanism of 1990. The experimental and predicted delays are in the same order of magnitude. A relatively good agreement is found for the major species profiles. Improvement of the mechanism is needed to account for the minor products. The different paths of OH formation are discussed.

The production of benzene in the low-temperature oxidation of cyclohexane, cyclohexene, and cyclohexa-1,3-diene

Abstract The oxidation and auto-ignition of cyclohexane, cyclohexene, and cyclohexa-1,3-diene have been studied by rapid compression between 600 K to 900 K and 0.7 MPa to 1.4 MPa to identify the low-temperature pathways leading to benzene from cyclohexane. Auto-ignition delay times were measured and concentration-time profiles of the C 6 intermediate products of oxidation were measured during the auto-ignition delays. Cyclohexane showed two-stage ignition at low temperatures, but single-stage ignition at higher temperatures, and a well-marked negative-temperature coefficient. By contrast there was neither a cool flame, nor a negative-temperature coefficient for cyclohexa-1,3-diene. Cyclohexene behaved in an intermediate way without a cool flame, but with a narrow, not very marked negative-temperature coefficient. The identified C 6 products belong to three families: the bicyclic epoxides and cyclic ketones, the unsaturated aliphatic aldehydes, and the conjugated alkenes, which are always the major products. The formation of C 6 products from cyclohexane is explained by the classical scheme for low-temperature oxidation, taking into account addition of O 2 to cyclohexyl radicals and the various isomerizations of the resulting peroxy radicals. Most of the C 6 products from cyclohexene are predicted by the same scheme, beginning with the formation of the allylic cyclohexenyl radical. However, addition of HO 2 to the double bond has to be included to predict the formation of 1,2-epoxycyclohexane. For cyclohexa-1,3-diene, the classical scheme is not valid: the C 6 oxygenated products are only formed by addition of HO 2 to the double bond. For all three hydrocarbons, the pathways to benzene are those leading to conjugated alkenes, and they are always more efficient than those producing oxygenated products, either by adding HO 2 to double bonds, or by addition of O 2 to the initial cyclic radical.

Comparison of oxidation and autoignition of the two primary reference fuels by rapid compression

New experimental data on autoignition delays and product distributions during two-stage autoignitions for the two primary reference fuels n -heptane and iso -octane (2,2,4-trimethylpentane) have been obtained by rapid compression in the low and intermediate range of temperature for enginelike conditions of stoichiometry and dilution. The lower reactivity of iso -octane has been compensated by a four times increase in pressure. A good correlation between our data and that published is obtained when the compressed charge density of the core gas is considered. Both fuels show many common features in this temperature range: a marked negative temperature coefficient region that shifts to higher temperatures as the pressure is increased and a similarity in the nature of the intermediate species. However, the importance of the cool flame zone is greater for n -heptane, and the negative temperature coefficient region extends toward higher temperatures. The evolution of the main intermediate products formed during the two-stage autoignition is presented and discussed according to a common generic mechanism that takes into account the various isomerizations of alkylperoxy radicals and scissions of the hydroperoxyalkyl radicals. Cyclic ethers are important intermediates. For both hydrocarbons, tetrahydrofurans are the major O heterocycles formed in cool flames, especially in the case of iso -octane. The observed high selectivity in the lower alkenes demonstrates the importance of β carbon-carbon scission of the hydroperoxyalkyl radicals that leads to terminal alkenes in the case of n -heptane and to methylpropene and substituted pentenes in the case of iso -octane. These channels have to be taken into account in the improvement of detailed mechanisms for good predictions of pollutants.

Autoignition Delays of a Series of Linear and Branched Chain Alkanes in the Intermediate Range of Temperature

A set of ignition data of linear and branched chain alkanes (n-butane, n-pentane, neopentane, n-heptane, and isooctane) measured in an original rapid compression machine is provided. It allows a comparison of the ignition conditions of pressure, temperature and equivalence ratio for these hydrocarbons. Detailed mechanisms from different research groups based on a similar generic scheme of hydrocarbon oxidation are tested against the experimental ignition delays. The differences between experimental and modeling results are discussed.

Rate constant for the reaction OH + H2S in the range 243–463 K by discharge-flow laser-induced fluorescence

The absolute rate constant k1 of the reaction of OH with H2S has been measured in helium in the range 243–363 K by discharge-flow laser-induced fluorescence. A few measurements have also been performed with the same flow tube by resonance fluorescence in the range 294–463 K. The parabolic behaviour of k1vs. temperature, with a minimum value of (33 ± 5)× 10–13 cm3 molecule–1 s–1 at 294 K, is confirmed. Our data above 294 K can be fitted with an Arrhenius form: k1= 1.32 × 10–11 exp [–(394 + 190)/T] cm3 molecule–1 s–1.

Influence of equivalence ratio on the structure of low-pressure premixed methanolair flames

Abstract Five low-pressure premixed methanolair flames were analyzed at different equivalence ratios between 0.82 and 1.50. Mole fraction profiles were measured for nine species using a sampling probe/gas ghromatography/electron spin resonance technique. Experimental data were compared with predictions from a one-dimensional flame model using a reduced chemical kinetic mechanism. This mechanism, including only 18 species and 34 pairs of elementary reactions, and the rate constants were in accord with recently recommended values. For the species examined in the five flames, a satisfactory agreement was obtained in shape, magnitude and also in location (except for the lean flame). The removal of CH 3 OH was dominated by CH 2 OH and HCO reactions, especially their decomposition and consumption by O 2 , with fast subsequent production of CH 2 O and CO. A reaction path analysis, including the calculation of net reaction rates and the individual contributions of the elementary reactions, is used to interpret the model.

Detection by E.S.R. of peroxy radicals: The importance of radical-radical reactions in the slow oxidation of butane

Abstract Butylperoxy radicals have been detected by E. S. R. in the slow oxidation of butane. In the first stages of the reaction, the concentration of the radicals and the consumption of oxygen grow roughly exponentially with time. The importance of the biradical reaction 2 R O 2 • → 2 R O • + O 2 is discussed by taking into account the experimentally determined radical concentration (around 10 −7 - 10 −8 mole 1 −1 ).

The Influence of Hydrogen Sulfide on the Combustion of Methanol in a Stoichiometric Premixed Flame. Experimental and Numerical Studies

Abstract The kinetic mechanism for converting H2S to SO and S02 has been studied in a low-pressure (80 Torr) stoichiometric CH3OH-Air flame doped with H2S in the range up to 2.4%. The experimental measurements are compared with the predictions of a reduced and updated comprehensive chemical kinetic model including 23 species and 47 pairs of elementary reactions. The agreement between model and experiment is generally good except for H atom. Analysis of the sulfur reaction mechanism clearly shows the importance of reactions H2S + H = SH + H2, H2S + OH = SH + H20, SO + OH = SO2 + H, SO + O2 = SO2 + O and S + O2 = SO + O to describe the evolution of H2S, and SO and SO2 species.

Quantitative measurements of hydroperoxy and alkylperoxy radicals in gas-phase oxidation reactions from overlapping electron paramagnetic resonance spectra

Two numerical methods are presented for the analysis of overlapping electron spin resonance spectra of peroxy radicals obtained by freezing. One is devoted to the determination of a simple lineshape parameter and the other uses a data collecting and processing system and a least-squares fitting treatment. Examples are provided for quantitative measurements of peroxy species obtained during gas-phase oxidation of hydrocarbons and heterogeneous decomposition of hydrogen peroxide.