New Ways to Create and to Detect Nonlocal Cooper Pairs in Solid-State Nanodevices
Alexander Schroer
Abstract
New strategies to generate and to detect spin-entangled pairs of electrons in mesoscopic solid-state heterostructures are proposed. A bilayer-graphene device is designed in which Cooper pairs from an s-wave superconductor are injected into an electrostatically-defined topological channel. Due to a particular band structure, two electrons of a Cooper pair propagate in opposite directions and are hence spatially separated if atomic-scale defects are sparse. Neither energy filtering nor Coulomb repulsion are required. The device can be interpreted as a normal/superconducting/normal junction, in which emission of nonlocal singlets is equivalent to local Andreev reflection, in contrast to the widespread identification of Cooper pair splitting with crossed Andreev reflection. To detect entanglement, the Josephson current through two parallel single-level quantum dots is investigated at all occupations. Combining exact diagonalization and perturbation theory, signatures of nonlocal Cooper pair transport are identified in the critical current, a macroscopic quantity. The model reproduces recent experimental observations and predicts a nonlocal triplet ground state on the quantum dots due to a tunable superconductor-mediated exchange coupling, which, too, is visible in the critical current exactly if nonlocal Cooper pairs are transported. Entanglement detection is pursued, too, by converting spin-entangled Cooper pairs into polarization-entangled photons, which are subsequently probed by a Bell-type measurement. A closed emission cycle is constructed, such that no correlations between the photons and the electrons remain, which are detrimental for entanglement transfer. Inevitable side channels and imperfections are identified and accounted for by a particular measurement protocol. Furthermore, it is discussed how spin-charge separation can be used to extract signatures of nonlocal spin singlets from the average current through an electronic beam splitter. This complements a known mechanism based on exchange statistics by which only the more difficult-to-access noise is sensitive for entanglement. The device can be realized by two crossed nanowires or by quantum-Hall edge states in Corbino geometry. The thesis also contains a review on entanglement and on recent theoretical and experimental progress in solid-state systems towards entanglement detection and generation, focusing on Cooper pair splitting.