Explosive versus Thermite Behavior in Iron(0) Aerogels Infiltrated with Perchlorates
journal contributionposted on 08.12.2015, 00:00 authored by Nicholas Leventis, Suraj Donthula, Chandana Mandal, Michael S. Ding, Chariklia Sotiriou-Leventis
Monolithic nanoporous iron was prepared via carbothermal reduction of interpenetrating networks of polybenzoxazine and iron oxide nanoparticles. Excess carbon was burned off at 600 °C in air, and oxides produced from partial oxidation of the Fe(0) network were reduced back to Fe(0) with H2 at different temperatures (temp) ranging from 300 to 1300 °C. Samples were carbon-free, for temp > 400 °C also oxide-free, and are referred to according to the final H2-reduction temperature as Fe-temp. Fe-temp monoliths were infiltrated with perchlorates, dried exhaustively and were ignited with a flame in open air. Most experimentation was conducted with LiClO4. Depending on temp, monoliths fizzled out (≤400 °C), exploded violently (500–900 °C) or behaved as thermites (≥950 °C). Samples sealed in evacuated tubes did not explode, while if sealed under N2 the explosive effect was intensified. Thus, explosive behavior was attributed to rapid heating and expansion of gas filling nanoporous space. However, although that condition was necessary for explosive behavior, it was not sufficient. Based on SEM, particle sizes via N2 sorption, electrical conductivity measurements and mechanical strength data under quasi-static compression, it was concluded that the boundaries between the three types of behavior after ignition were associated with (a) mild sintering (fizzling/explosive boundary at around 500 °C); and, (b) melting-like fusion of skeletal nanoparticles (explosive/thermite boundary at around 950 °C). Overall, mechanically weaker networks fizzled out; too strong behaved as thermites; networks of intermediate strength exploded. For thermite behavior in particular, other factors may be also at play, such as a combination of reduced porosity, a substoichiometric amount of LiClO4 and a slower heat release rate. The latter was supported by TGA data in O2 and was attributed to a slower rate of oxidation of progressively thicker nanostructures as the H2-reduction temperature increased.
Read the peer-reviewed publication
interpenetrating networksPerchloratesMonolithic nanoporous ironconductivity measurementsN 2monolithoxidationThermite Behaviornanoporous spaceH 2iron oxide nanoparticlesFeSEMparticle sizesLiClO 4O 2carbothermal reductiontempthermite behaviorExcess carbonTGA dataheat release rateLiClO 4.N 2 sorptionstrength datasubstoichiometric amountSample