Parallel computation of coupled-cluster hyperpolarizabilities.

Area of Science:

• Computational chemistry
• Quantum chemistry
• Molecular modeling

Background:

• Static hyperpolarizabilities are key molecular properties for nonlinear optical (NLO) applications.
• Accurate theoretical prediction of these properties is essential for designing novel NLO materials.
• Evaluating the performance of different computational methods is vital for reliable NLO studies.

Purpose of the Study:

• To assess the accuracy of coupled-cluster response theory and four density functional theory (DFT) methods for calculating static hyperpolarizabilities.
• To determine appropriate computational methods and basis sets for NLO studies across diverse molecular types.
• To establish a framework for estimating quantum chemical approximation errors in QM/MM calculations.

Main Methods:

• Coupled-cluster response theory (CCRT) calculations with large basis sets.
• Comparison with four common density functional theory (DFT) methods.
• High-performance computing for large-scale electronic structure calculations.

Main Results:

• CCRT results served as a benchmark for evaluating DFT method performance.
• Specific DFT methods and basis sets were identified as suitable for different molecular systems.
• The study demonstrated the feasibility of large-scale electronic structure calculations on modern supercomputers.

Conclusions:

• The findings guide the selection of accurate and efficient computational strategies for NLO property prediction.
• This work provides insights into error estimation for advanced computational chemistry applications.
• Modern computational resources enable complex quantum chemical calculations for materials science research.