The size distributions of nanoparticles of the oxides of Mg, Ba and Al in flames: Their measurement and dependence on the concentrations of free radicals in the flame

• Published in: Proceedings of the Combustion Institute Vol. 31; no. 2; pp. 1939 - 1945
• Main Authors: Fennell, P.S., Dennis, J.S., Hayhurst, A.N.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 2007
• Subjects: Aluminum; Aluminum oxide; Combustion; Combustion synthesis; Flame-sampling; Free radicals; Inorganic soots; Magnesium; Nanoparticles; Oxides; Particle size distributions; Pools; Nanoparticles; Inorganic soots; Combustion synthesis; Flame-sampling; Particle size distributions
• ISSN: 1540-7489; 1873-2704
• DOI: 10.1016/j.proci.2006.07.137

Flat, pre mixed, laminar, and very O 2-rich flames of C 2H 2 + O 2 + N 2 with [O 2]/[O 2] stoich ∼ 2.8 and a temperature ∼2000 K have been burned at atmospheric pressure. Trace amounts (∼13 ppm) of the metals Mg, Ba or A1 were added to the unburnt gases by nebulising an aqueous solution of a halide of the metal, so that e.g., Mg formed molecules of Mg(OH) 2, MgOH and MgO, as well as free atoms of Mg. The relative abundances of these species were governed by well-characterised equilibria and consequently depended on the temperature and also the concentrations of the flame’s free radicals H, OH and O. Transmission electron microscopy showed that nanoparticles of the oxides of these metals formed from their pool of molecular species in these flames. Particle size distributions were also measured (much less tediously) with a mobility analyser (DMS 500, Cambustion) operating at 0.25 bara. The optimal way of continuously sampling the gases at a point along the flame’s axis was investigated and shown to require expanding the sample (to a pressure of ∼1/3 bara) supersonically through an orifice with a diameter greater than 0.4 mm. In addition, the sample had to be diluted with N 2, with a volumetric flow rate of ∼10–20 times that of the sample, all at 1/3 bara. The sizes of oxide nanoparticles, as measured by transmission electron microscopy, agreed with the values of 6–10 nm from the mobility analyser. With Mg all the metal appeared very rapidly as nearly spherical nanoparticles of MgO early in a flame’s reaction zone. This was also true for Ba, which, according to thermodynamic considerations at the final temperature of the flame, should not form any particles of BaO. That particles do actually form is due to the reaction zone having a relatively low temperature and super-equilibrium concentrations of the free radicals H, OH and O. Aluminium was expected to form particles of A1 2O 3. However, only a small fraction of the Al formed particles; this is attributed to the production of gas-phase molecules of Al 2O 3 (i.e., the nuclei) from AlO and AlO 2 being by a relatively slow three-body reaction, as well as Al 2O 3 being a very minor member of the gas-phase pool of molecular species containing Al.