Multi-orbital dynamical mean field theory for transition-metal oxides on near-term quantum processors

Quantum embedding continues to be a promising path for modeling correlated materials on near-term quantum processors. Building on our prior single-orbital demonstration of a hybrid quantum-classical DFT+DMFT workflow, we now generalize the quantum impurity solver to multi-orbital systems. The extension treats the full matrix Green's function and self-energy, enabling orbital-resolved correlations in a Wannier basis. Methodologically, we prepare the ground state using a parameterized, symmetry-aware quantum circuit and qubit layout, then compute charged excitations with a quantum equation-of-motion (qEOM) approach using truncated excitation operators. To stabilize hardware runs, we employ advanced error mitigation protocols, including TREX (twirled readout error extinction) and our novel ZNE-C (zero-noise extrapolation calibration), scalable to large systems. We apply our impurity solver to an Anderson impurity model of a representative transition-metal oxide with multiple correlated d-orbitals and discretized baths on the latest IBM Heron devices. We compare the resulting quasiparticle weights with those obtained from established classical impurity solvers. We discuss scaling behavior versus number of orbitals and bath sites, and outline a path toward cluster extensions within the same framework.