The presence of polymer-particle interactionsin the composite materials leads to the formation of interfacial layers inwhich its physical properties is significantly different from those of polymermatrix and fillers. Formation of this phase can directly influence on membranepermeability and selectivity in addition to unwanted effects on their mechanical/physicalproperties. In this research for achieving a correct insight about the role andnature of interfacial morphologies on the performance of mixed matrix membranes(MMMs), some new techniques mainly in terms of mathematical modeling weredeveloped. Applying ideal permeation predictive models isthe first approach. Although ideal models often failed in predicting MMMsperformance, these models can provide guidelines for discernment of the typesof formed interfacial morphology, like current methods in characterization.Furthermore, a new modeling approach was proposed to satisfactory predicteffective permeability of MMMs by only adjusting of one parameter. In theapproach, existing non-ideal predictive models are rearranged based onintroducing a reduced permeability factor in the range of -0.5 to 1. Accordingto the proposed method, effective interface thickness is evaluated based onfindings of molecular dynamics and Monte Carlo simulations as well as particlesize and loading data. Moreover, based on the predictions, characteristicparameters of interfacial layer comprising interfacial permeability correctionfactor and permeability reduction factor of affected particles by blockage, areobtained. Although predictive models have been used as auseful tool for permeability prediction of MMMs, in this research a new methodwas developed based on gas permeation test for characterization of interfacialmorphology including the type of polymer-particle interaction, thickness andtraort properties of the interface and the size of aggregates in the case ofparticle aggregation. Capability of the proposed method was investigated bycomparing the results of SAXS analysis for PES based composites containingpristine and modified silica particles. Moreover, this method easily can beapplied for investigation of operational temperature and pressure effects oncharacteristic parameters of interfacial layer. Due to the temperature dependency of polymerchain mobility as well as plasticization effects related to the presence ofpenetrants in the polymer matrix investigation of traort properties ofprepared membranes at different operational temperatures and pressures isrequired. This would allow determining suitable membrane materials for a givengas separation application and optimum operational conditions. As measurementof traort properties of membranes over a wide range of conditions is timeconsuming and costly, in order to optimize membrane process design, anestimation of their separation performance based on existing experimental datais valuable. In this regard, so far different modeling attempts have been doneto take into account the effects of temperature and pressure on the traortproperties of polymeric membranes. While, one of the shortcomings in this areais development of a new predictive model with considering operationaltemperature and pressure effects on activation energy of permeability anddiffusivity as well as heat of sorption. Because analysis of theses energeticparameters can provide a proper insight into the effect of interactions betweenthe membrane and penetrants on the temperature and pressure dependence of thetraort of gases through membranes. Toovercome this limitation, in this study van't Hoff-Arrhenius model was modifiedin which, both effects of temperature and pressure are simultaneouslyconsidered on traort properties of polymeric and MMMs containing glassy andrubbery polymers. Furthermore, to minimize the number of experiments, theproposed model is capable to predict properly the traort properties as wellas the energetic parameters of penetrants through membranes over a wide rangeof operational conditions, only by measuring their properties at two pressures within a suitable temperature range including at least 3 points.