ADAPT-QSCI: Adaptive Construction of an Input State for Quantum-Selected Configuration Interaction

Area of Science:

• Quantum Computing
• Computational Chemistry
• Quantum Many-Body Physics

Background:

• Calculating ground-state energies of quantum many-body Hamiltonians is crucial for chemistry.
• Quantum-selected configuration interaction (QSCI) offers a path but requires a suitable input quantum state, which is challenging to prepare.
• Existing methods struggle with efficient and accurate quantum state preparation for QSCI.

Purpose of the Study:

• To propose an adaptive algorithm for constructing input quantum states for QSCI.
• To enable accurate ground-state energy calculations using current noisy quantum devices.
• To advance the application of quantum computation in quantum chemistry.

Main Methods:

• Introduced ADAPT-QSCI, a quantum-classical hybrid algorithm.
• Employed an adaptive state construction by iteratively running QSCI to grow the input state.
• Performed subspace diagonalization of the Hamiltonian via sampling measurements on a quantum computer.

Main Results:

• ADAPT-QSCI accurately computed ground-state energies for small molecules.
• The method demonstrated robustness in a noisy eight-qubit simulation with 1% error rates.
• Achieved high accuracy despite significant noise in quantum gates and measurements.

Conclusions:

• ADAPT-QSCI is a promising approach for leveraging current noisy quantum devices.
• The adaptive state preparation method overcomes a key challenge in QSCI.
• This work facilitates the practical application of quantum algorithms in quantum chemistry.