Deciphering the controlling factors for phase transitions in zeolitic imidazolate frameworks

Area of Science:

• Materials Science
• Computational Chemistry
• Solid-State Chemistry

Background:

• Zeolitic imidazolate frameworks (ZIFs) exhibit complex phase transitions, including melting and vitrification.
• Investigating structural evolution during amorphous phase transitions in ZIFs is experimentally challenging, particularly at medium-range length scales.

Purpose of the Study:

• To develop a computational method for accurately simulating structural evolution in ZIFs during phase transitions.
• To introduce a novel descriptor for predicting crystallization propensity in ZIF glasses.

Main Methods:

• Training a deep learning-based force field to characterize crystalline and non-crystalline ZIF phases.
• Reproducing structural evolution trends during ZIF melting and glass formation using the trained force field.
• Developing a 'ring orientation index' to quantify crystallization propensity.

Main Results:

• The deep learning force field accurately reproduces ZIF structural evolution at various length scales, comparable to ab initio methods but with lower computational cost.
• The ring orientation index effectively captures the crystallization propensity of ZIF-4 glasses and the formation of ZIF-zni from high-density amorphous phases.
• Crystal formation involves imidazole ring reorientation, leading to a loss of order around zinc-centered tetrahedra.

Conclusions:

• The developed deep learning approach provides an efficient and accurate method for studying ZIF phase transitions.
• The ring orientation index is a valuable tool for understanding and predicting the formation of crystalline phases from amorphous ZIFs.
• This work offers insights applicable to phase transitions in other metal-organic frameworks (MOFs) and the design of MOF glasses.