The temperature and species concentration history experienced by the gas-borne nanoparticles during their evolution in the flame has a major impact on their size, morphology, composition, and crystallinity. In our recent work (Combust. Flame, 244 (2022) 112251), we have reported optical emission measurements of a Fe(CO)5-doped H2/O2/Ar fuel-lean (ɸ = 0.5) flame, revealing that the temperature of the early-formed nanoparticles exceeds the gas temperature by several hundred degrees, while the particle volume fraction increases sharply, followed by rapid disintegration in the reaction zone. This behavior, modeled by single particle Monte-Carlo simulations indicates involvement of heterogeneous reactive processes at the particle surface, such as particle reduction and oxidation, growth and etching. Within the refined approach of the current study, reactive and non-reactive collisions were treated consistently, assuming rapid thermalization between the impinging molecule and the particle, with subsequent random energy sampling to determine reactivity. In the present work, we test the limits and validity of the heterogeneous flame-particle interaction model by manipulating the oxidation–reduction and growth-etching balance by varying the equivalence ratio (0.25<ɸ<1.5). For the entire range of equivalence ratios studied in experiments and simulations, we find a deviation between the particle and gas phase temperatures with significantly higher particle temperature, which is continued until a full degree of iron oxidation within the particle (O/Fe ratio=3/2) is reached. Validating the simulations against the measurements of particle temperature and volume fraction over a wide range of equivalence ratios, emphasized the necessity to account for gas-phase Fe-atom concentration depletion. We incorporated nucleation theory to estimate initial cluster population, linking Fe-concentration variation in the gas phase and the stochastic particle evolution model. The surface reaction parameters in our current work were updated using density functional theory literature data, and validation of the model predictions against experimental data, across the entire range of equivalence ratios.