Thanks for your suggestion. We have investigated the sensitivity of the oscillations to change of parameters (both for infinitesimal or finite change). Figure \ref{955477} shows the oxygen profile for three cases with three different activation energies. Case 1 has activation energy of 120000 J/gmole. Case 2 has activation energy that is is slightly bigger than case 1 (120001 J/gmole). Case 3 has larger activation energy (120100 J/gmole). Small change in parameters does not disturb the answer. Also finite change, remains in finite distance with  the original answer. In all cases, refinement in temporal and spatial scale does not change the result.