Hyperpolarised Naphthalene and Levitated Optomechanics for Fundamental Tests of Physics

Abstract

Levitated nanoparticles, free from the clamping losses and decoherence channels imposed by mechanical support, offer a growing versatile platform for precision sensing and tests of fundamental physics at increasingly larger masses. This thesis follows complementary directions based on this platform: the coupling of optically controllable nuclear spin ensembles to the mechanical and rotational degrees of freedom of levitated solids, and the identification of decoherence mechanisms that limit levitated systems in tests of macroscopic quantum superpositions. In the first part, we study levitated solids doped with non-permanent, optically addressable electron spins, with pentacene-doped naphthalene as a prime exam ple, whose nuclear spins can be hyperpolarised using dynamic nuclear polarisation techniques to over 80% at cryogenic temperatures, with lifetimes extending over weeks. Exploiting the homogeneous spin distribution and absence of a preferred nuclear quantisation axis in such materials, we design a multi-spin Stern-Gerlach interferometry protocol that circumvents key limitations of colour centres in nan odiamonds like NV centres, and use it to place improved bounds on objective collapse models. We further show that the unconstrained rotation available to levitated particles enables magic angle spinning at frequencies beyond current NMRstandards, and propose a protocol for reading out spin ensemble polarisation directly from the particle’s mechanical motion. Building on the same hyperpo larisation mechanism, we propose experiments to observe the nuclear Einstein-de Haas effect, using the hyperpolarised bath of nuclei as a reservoir for angular mo mentum. Furthermore, we propose a protocol to investigate the nuclear Barnett effect in the solid state, using the MHz-scale rotation rates accessible to levitated particles. We find that the proposed Stern-Gerlach interferometer could improve upon existing bounds on collapse-model parameters and that the nuclear Einstein de Haas and Barnett effects could be observable using near-term experimental technology, within parameter ranges already accessible to the community. In the second part, we turn to a central obstacle facing proposals for gravitationally induced entanglement (GIE) between levitated masses, namely the conducting and magnetic shields required to suppress direct electromagnetic coupling. We show that residual Casimir and magnetic-dipole interactions between the particles and the shield, averaged over run-to-run positional and orientational fluctuations of the experimental set-up, generate decoherence, and that the shield can itself mediate spurious entanglement between the particles that might mimic a genuine gravitational-entanglement signature. From this, we derive quantitative stability thresholds considerably stricter than previously assumed.