Complex chemical reaction mechanisms of high temperature hydrocarbon decomposition are represented as networks and their underlying graph topologies are analyzed as a dynamic system. As model reactants, 1,3-butadiene, acetylene, benzene, ethane, ethylene, methane, methyl isobutyl ketone (MIBK) and toluene are chosen in view of their importance for the global environment, energy technologies as well as their quantum chemical properties. Accurate kinetic mechanisms are computationally simulated and converted to bipartite graphs for the incremental conversion steps of the main reactant. Topological analysis of the resulting temporal networks reveals novel features unknown to classical chemical kinetics theory. The time-dependent percolation behavior of the chemical reaction networks shows infinite order phase transition and a unique correlation between the percolation thresholds and electron distribution of the reactants. These observations are expected to yield important applications in the development of a new theoretical perspective to chemical reactions and technological processes e.g. inhibition of greenhouse gases, efficient utilization of fossil fuels, and large scale carbon nanomaterial production.