An efficient implementation of the GOSTSHYP pressure model by applying shell-bounding Gaussian 1-electron-3-center...

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Accurate calculation of one-electron-three-center overlap integrals is crucial for quantum chemistry simulations.
• The computational cost of such calculations often limits the size of systems that can be studied, particularly for methods like the Gaussians On Surface Tesserae Simulate HYdrostatic Pressure (GOSTSHYP) model.
• High memory usage in GOSTSHYP restricted previous studies to smaller systems, such as fullerenes up to C40.

Purpose of the Study:

• To implement an efficient integral screening algorithm for one-electron-three-center overlap integrals within the Q-Chem program.
• To reduce the computational scaling of the GOSTSHYP model to a linear relationship with the number of surface tesserae.
• To enable the study of larger and more complex molecular systems, specifically fullerenes, by mitigating high RAM usage.

Main Methods:

• Development and implementation of a screening algorithm for one-electron-three-center overlap integrals using contracted Gaussian-type orbitals.
• Derivation of integral bounds utilizing shell-bounding Gaussians and Obara-Saika recurrence relations.
• Integration of the screening algorithm into the Q-Chem quantum chemistry package and optimization of the GOSTSHYP model.

Main Results:

• The integral screening algorithm successfully reduced the computational scaling of the GOSTSHYP model to be linear with respect to the tesserae used for surface discretization.
• Significant reductions in calculation time and memory usage were achieved.
• The improved GOSTSHYP model allowed for the calculation of fullerene compressibility up to C180, a substantial increase from the previous limit of C40.

Conclusions:

• The implemented integral screening algorithm provides a significant performance enhancement for quantum chemistry calculations.
• This advancement overcomes previous computational limitations, enabling the investigation of larger molecular systems and properties.
• The study demonstrates the practical applicability and effectiveness of the developed algorithm in materials science research, particularly for fullerene systems.