Improving the mapping of quantum chemistry problems to quantum computers

A key goal in quantum chemistry for quantum computing is to go beyond active-space methods while still keeping the problem small enough for quantum hardware. Standard active-space approaches capture the strongly correlated orbitals explicitly, but they often leave most of the dynamic correlation to classical post-processing. That is a serious limitation, because higher-lying orbitals can still shift reaction energies, excitation gaps, and spin-state orderings enough to change the chemistry you are trying to predict.

  • What has been developed at HQS is a route to retain dynamic correlation natively in the quantum formulation, without bringing the full virtual orbital space onto the device. The central idea is to separate the problem into a compact active region, which contains the essential multireference physics, and an effective environment that captures the remaining screening and fluctuation effects. In practice, this turns a large orbital problem into a much more structured hybrid representation: a small fermionic system coupled to bosonic modes that encode dynamical correlation.

    This is where the Random Phase Approximation becomes especially useful. Rather than treating RPA as just a classical correction, it is used as a compression strategy for the inactive space. The strongly correlated part stays explicit, while the rest of the molecule is mapped into collective screening modes that are much more natural to handle in a hybrid quantum-classical setting. That makes it possible to preserve the chemistry that active-space-only models often miss, while keeping the quantum resource requirements under control.

    The broader result is a more realistic path for quantum computing in chemistry: not simply shrinking the problem until it fits on a device, but reformulating it so that static and dynamic correlation are both represented in a hardware-aware way. In that sense, the main advance is not only a better approximation scheme, but a better mapping of electronic structure itself onto a quantum computer.

  • System-bath model for quantum chemistry, Dmitry S. Golubev, Reza G. Shirazi, Vladimir V. Rybkin, Benedikt M. Schönauer, Peter Schmitteckert, Michael Marthaler,‍ ‍arXiv:2603.09631

  • Efficient Random Phase Approximation for Diradicals, Reza G. Shirazi, Vladimir V. Rybkin, Michael Marthaler, Dmitry S. Golubev, ‍arXiv:2404.18691