On chemical inhibition of shock wave ignition of hydrogen–oxygen mixtures

In this work an influence of the wide range of various inhibitors, namely CCl4, CF3H, C2F4Br2, (CH3O)3P, CF3I and C3F7I on shock-induced ignition of hydrogen was experimentally investigated. Observed temperature dependencies of induction times indicates that CF3H and (CH3O)3P do not show noticeable inhibiting activity at given conditions, while the effectiveness of halogen-containing specie dramatically increases in a row Cl → Br → I. It is shown that the most effective inhibitors of ignition of hydrogen–oxygen mixtures are iodinated hydrocarbons CF3I and C3F7I.


Introduction
The prevention of catastrophic explosions of hydrogen (particularly during the accidents at nuclear power plants) is the actual problem for existing and developing energy technologies. An introduction of chemically active inhibitors which suppress the combustion and detonation development is widely considered. It is known that their effectiveness is caused not only by the removal of oxygen from the fire zone and the high endothermic effect of their evaporation and dissociation, resulting in a temperature reduction in the combustion zone, but also by their chemical reactivity. The generalized mechanism for the chemical inhibition of ignition proposed in the classical paper [1] relates the flame-retardant effect to the reactions O + X + M → OX + M, and inhibit the development of combustion. However, it was noted already in [1], that under certain conditions, chemically active admixtures may lose their flame-retardant properties and even lead to a reduction in the ignition delay. In the recent papers, it has been shown that 2 1234567890 ''"" at elevated temperatures, halogenated and phosphorus-containing fire suppressants can reduce the ignition delay in methane-oxygen mixtures [2][3][4][5], accelerate detonation in acetylene [6], and, moreover, are themselves capable of combustion [7]. Experimental studies on the ignition of haloalkane-containing mixtures in a wide range of parameters can support the development of a comprehensive model of pyrolysis and oxidation of haloalkanes and their influence of combustion development. This is also a scientific task for the development of modern fire extinguishing systems and the analysis of industrial risks. One should note that shock tube experiment differ fundamentally from other types of experiments such as initiation of ignition at room temperature by a local energy source, or the study of thresholds and the rates of flame propagation. In these conditions the processes of heat conductivity and diffusion of the active radicals play the main role, while ignition in a shock tube occurs at the total homogeneity of process in the volume and completely defined by kinetics of chain reactions. Therefore, only experiments behind shock waves allow the most reliable study of the kinetic mechanisms of the effect of various inhibitors on the ignition of hydrogen.
Therefore, the goal of this work was to perform an experimental and numerical study of the effect of various flame-retardant additives, namely CCl 4 , CF 3 H, C 2 F 4 Br 2 , (CH 3 O) 3 P, CF 3 I and C 3 F 7 I, on hydrogen-oxygen mixtures ignition behind shock waves of different intensity.
Carbon tetrachloride, also known as Halon-104, was used in the beginning of the twentieth century, but was later rejected because of its noticeable toxicity.
Lately other, safer halogenoalkanes, particularly fluoroform CF 3 H (Freon-23) [8] and 1.2-dibromoperfluoroethane C 2 F 4 Br 2 (Freon-114B2) [9], were considered as agents to prevent the explosion and detonation of combustible mixtures and were widely used in fire-extinguishing systems. Nowadays, many halogenated compounds are considered toxic, ozone-depleting gases and forbidden by Montreal Protocol. Therefore it is recommended to use more safe iodinated hydrocarbons such as CF 3 I and C 3 F 7 I. One should note, though, that brominated hydrocarbons still can be applied for rare emergency use on unmanned facilities such as nuclear reactors containments if their effectiveness is high enough. Phosphorus-containing compounds such as trimethylphosphite (CH 3 O) 3 P are another promising, but much less studied type of chemical reagents which were shown to be effective flame retardants for hydrocarbons [10][11][12][13], syngas [14], and hydrogen [15].

Experimental and modeling methods
Experiments were carried out in stainless steel shock tube of standard design in stoichiometric hydrogen-oxygen mixtures diluted with argon to 10-20% and doped with 1-3% of studied suppressants. The investigated section had optical windows and was located at a distance of 13 mm from the end plate. The shock tube was equipped with several PCB113B piezoelectric pressure gauges to measure the incident shock wave velocity with an accuracy 0.5%. The actual values of temperature (T RSW ) and pressure (P RSW ) of shock-heated flow behind reflected shock wave were derived using a common iteration method based on one-dimensional shock tube theory [16].
The temperature dependencies of the ignition delay time (the induction time) are the key features of the mixtures investigated. OH radicals are the characteristic species indicating the hydrogen-oxygen and hydrocarbon-oxygen mixtures ignition. During the experiments, an excited OH* chemiluminescence signal was recorded by a Hamamatsu H9307-03 photomultiplier module equipped with an interference filter (310 ± 5 nm) to determine the ignition delay times. In combustible mixtures, the energy release in ignition is quite abrupt and leads to a dramatic increase in the OH concentration in both ground and excited states. Thus, a rapid rise in OH* chemiluminescence was considered as the end of induction time. The exact ignition moment was determined as the intersection of the inflectional tangent line of the OH* radiation intensity plot with the time axis. The increase in pressure was simultaneously recorded by a pressure gauge. The ignition development in hydrogen-oxygen mixtures was modeled using ChemKin  software package [17] by a recently suggested 20-reaction scheme for hydrogen combustion which describes that process well, particularly at low temperatures [18]. Figure 1 presents the measured and calculated ignition delay times in the reference mixtures of 6.7%H 2 + 3.3%O 2 + Ar and 13.3%H 2 + 6.6%O 2 + Ar at pressures of 2.2-2.7 bar. For modeling the pressure value P = 2.5 bar was used. The results show a good agreement.

Results
Unlike to other studied admixtures, trimethylphosphite (CH 3 O) 3 P being added to 13.3%H 2 + 6.6%O 2 + Ar mixture caused notable two-stage ignition development; typical signals of pressure and OH* radiation intensity are presented in figure 2. One can see the first gradual rise of OH* radiation and corresponding slow pressure increase at the time τ 200 µs (which one could call 'pre-ignition'), and the secondary steep OH* radiation and pressure rise at τ 500 µs. While the ignition delay times measured by the steep rise of both signals in presensce of 1%(CH 3 O) 3 P coincide with the measurements in the test mixtures, the pre-ignition occurs much earlier (figure 3). Trimethylphosphite thus certainly lack inhibiting activity in shock tube conditions. This result is not in contradiction with known flame-suppressant properties of phosphorus-containing species, as flame and shock tube kinetics differ significantly as was mentioned above. Figures 4 and 5 summarize the temperature dependencies of the ignition delays measured in 6.7%H 2 +3.3%O 2 +Ar and 13.3%H 2 +6.6%O 2 +Ar mixtures doped by studied flame suppressants.
Recently developed kinetic mechanism of CCl 4 and CF 3 H pyrolysis and oxidation [5] being used for modeling provided a good agreement of experimentally measured and calculated ignition delay times and allowed determining key reactions of influence of studied admixtures on hydrogen combustion. One should note that though radical-terminating reactions (1)   reaction Cl + H 2 → HCl + H also plays quite an important role in combustion development. Thus, admixture pyrolysis which was previously shown to have dramatic influence on methane ignition [5] is significant even at temperatures below 1200 K. According to NIST Chemical Kinetics Database many important reactions of C 2 F 4 Br 2 , CF 3 I and C 3 F 7 I pyrolysis and oxidation mechanisms are uninvestigated. The available data do not allow for reliable estimation of the numerous unknown rate constants, thus it is clear that the development of quantitative kinetic mechanisms describing hydrogen combustion inhibition requires new data on the kinetics of C 2 F 4 Br 2 , CF 3 I and C 3 F 7 I pyrolysis and oxidation. In particular, one could expect a great importance of pyrolysis reactions producing atomic bromine and iodine and consequent active radical formation in reactions Br + H 2 → HBr + H and I + H 2 → HI + H. Therefore an analysis of possible mechanisms of interaction of brominated and iodinated hydrocarbons with hydrogen-oxygen mixtures is the task for the future studies.

Conclusion
Influence of various halogenated and phosphorated inhibitors, namely CCl 4 , CF 3 H, C 2 F 4 Br 2 , (CH 3 O) 3 P, CF 3 I and C 3 F 7 I on shock-induced ignition of hydrogen was experimentally investigated. Observed temperature dependencies of induction times indicate that CF 3 H and (CH 3 O) 3 P certainly lack inhibiting activity (well known for flames) at given shock wave conditions, while the effectiveness of halogen-containing specie dramatically increases in a row Cl → Br → I. C 3 F 7 I provided an unique combination of combustion suppression and safety for human health and ecology.