<p>Amorphous boron nitride doped with oxygen, boron oxynitride, BNO, is a porous material stable at high pressures and elevated temperatures with potential uses in adsorption-based separation processes at the industrial scale. We present here a molecular model capable of accurately predicting gas sorption in porous BNO solely from the knowledge of the basic</p>
<p>experimental characteristics, i.e. overall chemical composition and porosity. With this information, the adsorbent is described atomistically by a complex 3-D pore network built by random packing of nanoflakes. The adsorption may then be evaluated by employing Grand Canonical Monte Carlo with classical forcefields. We report sorption isotherms for CO2, N2 and CH4 on BNO at low (< 1 bar) and high (0 – 20 bar) pressures, across a range of temperatures (283 – 313 K), which are well predicted by the molecular model. While the experimental measurement of multi-component isotherms under such conditions is a challenging task, molecular simulations provide predictions without the need of additional information. As an example, CO2/N2 and CO2/CH4 binary mixture isotherms, at conditions relevant to post combustion CO2 capture and natural gas sweetening, are computed. Overall, the model provides fundamental insight, which is useful in the design and optimization of porous BNObased</p>
<p>adsorbents for molecular separations.</p>