Unstable and Metastable Mesophases Can Assist in the Nucleation of Porous Crystals

Abstract

Porous crystalsincluding zeolites, metal–organic frameworks, and clathratesare widely used as catalysts and molecular sieves and in energy applications. These materials are synthesized from solution through an intermediate amorphous phase. However, the structural gap between the amorphous and porous crystal phases can lead to large nucleation barriers and difficulties in the control of crystal polymorphs. Previous reports indicate that porous mesophases can facilitate the amorphous to porous crystal transformation, directing the crystallization toward specific crystal structures. To date, it is not known how mesophase stability and synthesis temperature impact the facilitation. Here, we use molecular simulations and nucleation theory to investigate the crystallization of a zeolite in a family of models with tunable mesophase stability. We find that the nucleation mechanism evolves from non-classical to classical as the mesophase stability approaches that of the crystal. The simulations reveal that even an unstable mesophase can facilitate porous crystal nucleation through the formation of a transient fluctuation within which the crystal is originated. We conclude that tapping into the medium-range order of mesophases that have ordered pores without crystalline tiling is promising to increase the crystallization rates of porous crystals while directing the synthesis toward specific polymorphs.

Coarse-Grained Model for the Hydrothermal Synthesis of Zeolites

“…While a simpler, cosine quadratic angular dependence can also nucleate and grow the zeolite, 60 the use of a polynomial three-body interaction has the flexibility to represent the multiple preferred Si−Si−Si angles found in zeolites (Figure S1a), increasing the stability of the crystals. 61 The TS−Z model parameterizes the angular dependence to reproduce the population of angles in zeolite Z1 (Section A of Supporting Information), although the preferred T−T−T angles are common to most zeolites. Details of the potential TS−Z with λ = 4 model and its parameterization are given in section A of the Supporting Information.…”

“…Mixtures of T and S with a molar fraction of x T = 0.74 modeled with TS−Z with λ = 4 spontaneously nucleate and grow zeolite Z1 60 at 716 K when cooled at 1 Kns −1 at 1 bar. 61 The zeolite melts to an amorphous phase at T m = 832 K at 1 bar. 61 We hydrate the TS−Z with λ = 4 model with the CG monatomic water model mW, which mimics hydrogenbonding interactions through two-and three-body SW interactions 65 and reproduces the thermodynamic, dynamical, structural, and crystallization properties of real water.…”

“…61 The zeolite melts to an amorphous phase at T m = 832 K at 1 bar. 61 We hydrate the TS−Z with λ = 4 model with the CG monatomic water model mW, which mimics hydrogenbonding interactions through two-and three-body SW interactions 65 and reproduces the thermodynamic, dynamical, structural, and crystallization properties of real water. 65−77 Moreover, mW has a range of interactions and degree of coarsening compatible with the silica-like T and SDA S of the TS−Z with λ = 4 model, which have been shown to matter in coarse graining.…”

“…Despite the ability of these simple models to nucleate and grow a variety of zeolites, they are not known to model correctly the polymerization of silica, and they do not quantitatively reproduce the experimental speciation of silicate species in the amorphous phase from which the zeolites form. 61 Our aim in this work is to extend the CG model of ref 61 to capture the polymerization of silica, the nucleation and growth of zeolites, and the desorption of water on heating. We start with the two-component TS−Z with λ = 4 model of Bertolazzo et al 61 and expand it to incorporate water as a third component.…”

“…61 Our aim in this work is to extend the CG model of ref 61 to capture the polymerization of silica, the nucleation and growth of zeolites, and the desorption of water on heating. We start with the two-component TS−Z with λ = 4 model of Bertolazzo et al 61 and expand it to incorporate water as a third component. In the next sections, we present the parameterization of the model and show that it reproduces the experimental speciation of the silicates a function of degree of condensation in the polymerization of silica 32 and is able to spontaneously nucleate and grow a zeolite from the hydrated amorphous phase, as observed in experiments.…”