No abstract

The combustion of hydrogen and silane is studied. It is established that the chain initiation reaction on quartz in the zone of hydrogen and silane combustion is manifested as an autocatalytic reaction which is able to initiate a chain explosion and participate in the initiation of a thermal explosion. It is shown that in the case of an oxyhydrogen gas, the assumption of a branching-chain nature of the third limit is inconsistent with Semenov's law, which includes double exponential dependences of the chain reaction rate on time and temperature. A criterion for the participation of branching chains in complex processes is proposed based on the presence or absence of short delays of a thermal explosion (≈1 sec). According to the criterion, the explosion of an oxyhydrogen gas at atmospheric pressure with delays markedly exceeding 1 sec proceeds without the participation of branching chains and is consistently explained by the joint action of autocatalytic processes on the reactor wall and gas-phase processes.

Previously, the theory of branched chain processes was developed [1][2][3][4] and detailed agreement between theory and experiment was demonstrated [3,4] for the combustion of a hydrogen-oxygen mixture in a closed reactor within the ignition peninsula in the case where the initial pressure P 0 was several times as high as the pressure P 1 at the first ignition limit. However, near the first ignition limit, the behavior of branched chain processes differed fundamentally from that predicted theoretically [3,5]. In 1971 was proposed to seek the cause of the discrepancy between theory and experiment in the possibility of heterogeneous chain propagation or in experimental errors [7, p. 631]. The systematic error due to the unforeseen interaction between the vacuum lubricant and the rarefied flame was indeed detected and eliminated [2].Earlier, the possibility of branching of active sites on various surfaces was discussed by Alyea and Haber [8], Semenov [1], Hinshelwood et al. [9], and Nalbandyan [10]. The concept of heterogeneous chain propagation, which is allied to branched chain processes and heterogeneous catalysis, was formulated by Voevodskii in 1953 and 1955 [11] and by Semenov in 1956 in his Nobel Lecture [12]. Alyea and Haber [8] hypothesized the chain propagation on the surface since their experiments demonstrated the ignition of a hydrogen-oxygen mixture on introduction of a quartz rod into the zone of intersection of the hydrogen and oxygen jets at 600°ë . Nalbandyan showed [10] that the quartz surface indeed emitted active reaction sites decreasing the ignition delay in the zone of intersection of the hydrogen and oxygen jets. These active sites were a product of heterogeneous chain initiation rather than heterogeneous branching as assumed by Alyea and Haber [8]. Over more than half a century , a number of attempts were made to experimentally detect the accelerating effect of the wall on the hydrogen combustion in the gas and to relate the obtained results to the concept of heterogeneous chain branching [6][7][8]13]. However, these explanations were not unambiguous and the observed kinetic phenomena later generally turned out to have other causes (the effect of the lubricant, self-heating, and various inhomogeneities within the reactor [2,14]). For example, the combustion of silane, whose product ( SiO 2 ) deposited on the reactor wall and caused an enhanced glow of the surface, was regarded in [13] as a process involving heterogeneous chain branching. However, Azatyan and Aivazyan in 1986 [13] were likely to only reproduce, under new conditions, the phenomenon observed by Alyea, Haber, and Nalbandyan in 1930, and the enhanced glow of pure quartz formed in silane combustion is probably related not to heterogeneous branching but to rapid heterogeneous chain initiation on fresh quartz [8].The elimination of the systematic error due to the vacuum lubricant and the futility of attempts made over many years to detect heterogeneous chain branching [6-8, 14] meant, with a high probability, the nonexistence o...

No abstract