Fast and Reasonable Geometry Optimization of Lanthanoid Complexes with an Extended Tight Binding Quantum Chemical...

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Lanthanoid Chemistry

Background:

• Accurate structural prediction of lanthanoid complexes is crucial for understanding their chemical properties.
• Existing low-cost quantum chemical methods have limitations in accuracy and efficiency for these systems.

Purpose of the Study:

• To evaluate the performance of the GFN-xTB method for geometry optimization of lanthanoid complexes.
• To compare GFN-xTB against established quantum chemical methods.

Main Methods:

• A benchmark study involving 80 diverse lanthanoid complexes.
• Comparison with X-ray crystallography data from the Cambridge Structural Database.
• Validation against DFT-D3(BJ) optimized structures for specific promethium systems.

Main Results:

• GFN-xTB achieved a lower average heavy-atom root-mean-square deviation (0.65 Å) compared to Sparkle/PM6 (0.86 Å) and HF-3c (0.68 Å).
• The method produced chemically reasonable structures with fewer outliers.
• GFN-xTB demonstrated competitive computational speed.

Conclusions:

• GFN-xTB is a highly effective and efficient low-cost quantum chemical method for optimizing lanthanoid complex structures.
• Its accuracy and speed make it suitable for exploring complex conformations and reaction mechanisms.