Improving the Runtime of Quantum Phase Estimation for Chemistry through Basis Set Optimization.

Area of Science:

• Quantum computing
• Computational chemistry
• Quantum algorithms

Background:

• Quantum phase estimation (QPE) offers high accuracy for molecular ground-state energies.
• QPE's computational cost scales with molecular size, limiting its application for dynamic correlation.
• Enlarging the active space for dynamic correlation increases computational demands.

Purpose of the Study:

• To investigate basis set optimization strategies for reducing QPE computational costs.
• To assess methods for enhancing the efficiency of quantum simulations for molecular energies.
• To enable scalable and accurate quantum simulations by mitigating resource requirements.

Main Methods:

• Investigated adjusting Gaussian basis function coefficients to minimize the Hamiltonian 1-norm (λ).
• Employed a large-basis-set frozen natural orbital (FNO) strategy to reduce QPE resource requirements.
• Studied 58 small organic molecules and the N₂ dissociation curve.

Main Results:

• Adjusting basis function coefficients yielded minor, system-dependent reductions in λ (up to 10%).
• The FNO strategy significantly reduced QPE resources, achieving up to an 80% reduction in λ.
• The FNO approach also decreased the number of orbitals by 55% without compromising accuracy.
• Active spaces from larger basis sets more effectively captured correlation effects.

Conclusions:

• Basis set optimization, especially using FNOs, is a viable strategy to reduce QPE costs.
• Improved orbital basis quality, not just size, is key for incorporating dynamical correlation.
• This work advances scalable and accurate quantum simulations for chemistry with practical resource needs.