Structure Collapse, Recrystallization and Melting

Due to their rigid framework, zeolites are potentially stable from a thermal point of view, but this is not true in every case as thermal stability depends upon many factors, such as structure, Si/Al ratio and the cationic form of the microporous material. 

The reason why one zeolite structure is more stable than another is generally difficult to explain, but, at least for some types, it is connected to the presence of water in the framework. As discussed by Barrer ([4] pp. 54-60) on a thermodynamic basis, the stabilization of porous crystals is strictly related to the presence of guest molecules, i.e. water molecules, some oxygens of which share electron pairs with the extra-framework cations. When a zeolite undergoes dehydration, this involves a structure distortion, due to the shift of the cations 

Thermal stability usually increases with increasing Si/Al ratio, but, as seen immediately above, there are notable exceptions, such as zeolite Linde A, characterized by a good thermal stability, in spite of its low Si/Al ratio (=1). On the contrary, this rule is true especially for ZSM-5, so explaining, at least partly, its wide usage in catalysis applications. It is to be observed that thermal stability may sometimes be enhanced through a dealumination process of the phase obtained by synthesis, as is commonly attained for the so-called ultrastable zeolite Linde Y [45]. 

Cation type and concentration in zeolites may play a decisive role on thermal stability, because, as pointed out above, cations are an essential part of the zeolite structure ([1] pp. 493-496). This is why a normally stable zeolite, such as Linde A, becomes unstable after cation-exchange for barium [35] on the contrary, phillipsite, normally considered unstable, may substantially improve its thermal stability further after cation-exchange for potassium [46]. 

Thermal analysis, namely DTA, can help identify structural collapse, only if this occurs sharply, as in the above-mentioned case of Ba-exchanged Linde A (see the sharp endotherm at about 450 K in Fig. 8b, from [35]). Otherwise, if thermal degradation takes place slowly, which is normal in zeolites, no evident thermal effects are recognizable in DTA or similar thermal techniques. 

---