Evidence from experimental studies related to sorption on nanoporous materials, coals, and most recently on shales suggests that the sorption of gas in these targets is not instantaneous, i.e. the time it takes for the gas to adsorb to its maximum capacity is not negligible. A recent experimental study showed that the sorption in shale rocks is non-instantaneous, i.e. there is a delayed effect in adsorption due to different mechanisms of sorption in different materials of shale, for e.g. in inorganic and organic matter. The current suite of models used to predict sorption in shale (e.g. Langmuir, BET, etc.) assumes an instantaneous equilibrium, which is not justifiable for sorption in shales. Sorption of hydrocarbons and carbon dioxide in shale stratum is a complex mechanism that is affected by micro-scale affinity between the adsorbent-adsorbate, pore-scale heterogeneity of the surface with respect to the pore distribution, and structural heterogeneity of the system. These mull-phenomena effects are mull-scale in nature, and their impact on sorption cannot be predicted using simple models that assume instantaneous equilibrium. A model for reliable prediction of sorption in shales must take into account above discussed effects. This study proposes a mechanistic model for mull-scale sorption dynamics (referred as "MSSD") in shale that accounts for molecular and pore-scale effects, in addition to the continuum-scale effects, by corresponding parameters that are relevant on those scales. Proposed model is validated against three experimental datasets that depict sorption dynamics in shale. The MSSD model is then used to investigate the efficiency of carbon storage in shale by adsorption when it is injected as a binary mixture with methane. The proposed model makes it convenient to study the effects of molecular-scale to continuum-scale parameters on sorption in shale that are otherwise possible only through expensive means of experiments or molecular dynamic simulations.