We present a systematic computational study of Mg-MOF-74, CuBTC and zeolite 13X for CO2 separation from multi-component flue gas mixtures. The impurities' impact was evaluated at the molecular level and process conditions. Adsorption isotherms and isosteric heats of adsorption of pure (CO2, N2, O2, H2O, SO2 and NO2) components, binary and ternary mixtures were obtained from Grand Canonical Monte Carlo simulations. Working capacities, purities, recoveries and exergetic performances were evaluated for VSA/PSA/TSA processes. Results show that NO2 has a negligible effect in the studied range. For H2O and SO2 the energy requirements are reduced as the impurity content increases and recovery and purity increase, up to an 'optimal' point where a competition for CO2 preferred adsorption sites produces a sharp drop in purity and the energetic index grows exponentially. The minimum energy requirement were obtained for TSA at a desorbing temperature of 443 K in the three materials, with impurities of 1% H2O for CuBTC, 0.5% H2O for Mg-MOF-74 and 0.02% H2O for 13X, obtaining values of 1.13, 0.55 and 0.58 GJ/tCO2, respectively. Hybrid VTSA processes with impurities content in the feed mixture and CCS specifications achieve energy performances of 0.36 GJ/tCO2 and 0.46 GJ/tCO2 with Mg-MOF-74 and 13X, respectively. Mg-MOF-74 stands up as an attractive material for VTSA processes, presenting higher working capacities, purities and second-law efficiencies, with lower energy consumptions, also showing a better 'buffer' behavior than zeolite 13X when trace impurities are present. This work represents the first quantitative assessment of the process performance of MOFs adsorbents in swing adsorption process for CO2 capture considering impurities effects. Results reinforce the validity of molecular simulations for guiding the optimization of these processes.