Use of the Grand Canonical Transition-Matrix Monte Carlo Method to Model Gas Adsorption in Porous Materials Daniel W. Siderius* and Vincent K. Shen Chemical Sciences Division, National Institute of Standards and Technology, Gaithersburg, Maryland 20899, United States ABSTRACT: We present grand canonical transition-matrix Monte Carlo (GC-TMMC) as an ecient method for simulating gas adsorption processes, with particular emphasis on subcritical gas adsorption in which capillary phase transitions are present. As in other applications of TMMC, the goal of the simulation is to compute a particle number probability distribution (PNPD), from which thermophysical properties of the system can be computed. The key advantage of GC-TMMC is that, by appropriate use of histogram reweighting, one can generate an entire adsorption isotherm, including those with hysteresis loops, from the PNPD generated by a single GC-TMMC simulation. We discuss how to determine various thermophysical properties of an adsorptive system from the PNPD, including the identication of capillary phases and capillary phase transitions, the equilibrium phase transition, other free energies, and the heat of adsorption. To demonstrate the utility of GC-TMMC for studies of adsorption, we apply the method to various systems including cylindrical pores and a crystalline adsorbent to compute various properties and compare results to previously published data. Our results demonstrate that the GC-TMMC method eciently yields adsorption isotherms and high-quality properties of adsorptive systems and can be straightforwardly applied to more complex uids and adsorbent materials. I. INTRODUCTION As a general rule, the properties of a uid can be altered signicantly from their bulk values when conned in tight spaces, with the magnitude of such alterations dependent on the characteristics of the conning walls, specically the anity of the uid to the surface. 1 Of particular scientic and technical interest is the eect of connement on a uids phase behavior (or phase boundaries). 1-3 For example, uid adsorption in porous materials, the focus of this work, serves as the basis for potential viable carbon capture technologies. 4-7 While theory and simulation have played a historically signicant role in the characterization of porous materials, 1,8,9 they have more recently been identied as key tools in screening and developing potential carbon capture materials. 10-12 Thus, the advancement of carbon capture technologies and other applications of gas adsorption will depend on the availability and further development of computationally ecient and precise methods to predict the thermodynamic and dynamic properties of uids in porous materials. At present, the two primary molecular modeling tools used in studies of adsorption are density functional theory 13 (with which we include the closely related lattice mean eld theories 14-17 ) and various forms of Monte Carlo (MC) molecular simulation. 18,19 These methods have proven essential to the advancement of the fundamental understanding of adsorption processes 8 by, for example, suggesting the existence of cavitation-induced capillary evaporation in ink-bottle pores prior to its observance experimentally 20 and conrming the relationship between subcritical adsorption hysteresis and uid metastability. 21,22 One of the main goals in the simulation of adsorption phenomena is the calculation of the adsorption isotherm, which in turn requires identifying and determining metastable and stable uid phases. For example, in the quintessential adsorption problem at a subcritical temperature, the adsorption isotherm may include two main density branches, one that is vapor-like and the other that is liquid-like. Both branches of the isotherm will exist at some range of pressures below the bulk saturation pressure, forming a so-called hysteresis loop. Figure 1 shows an example of one such isotherm, known as a type IV isotherm, 23 in which the conned uid exhibits two capillary Received: January 15, 2013 Revised: February 20, 2013 Published: February 21, 2013 Article pubs.acs.org/JPCC This article not subject to U.S. Copyright. Published 2013 by the American Chemical Society 5861 dx.doi.org/10.1021/jp400480q | J. Phys. Chem. C 2013, 117, 5861-5872