Session F70: Sensing, Interactions, and Fundamental Limits

Enhanced deep water acoustic range estimation based on ocean General Circulation Models

Improved acoustic range estimation supports ocean climate monitoring, ocean model improvement, and underwater navigation, and is limited by the availability of accurate models of ocean sound speed variability. I will describe efforts to test the limits of currently available GCM data to accurately estimate absolute range based on data collected during the PhilSea10 experiment using 510 km source-receiver separation. The methods compare observed acoustic records with synthetic records computed through the GCM ocean sound speed model. The most reliable ranging methods, emphasizing the most reproducible parts of the data records, are able to achieve 20 m or better accuracy. The principles underlying these methods are transportable and are expected to provide reliable range estimates in a broad range of deep-water settings. Generalization to ocean regions with stronger, less well characterized dynamics, is a major challenge, requiring new sources of data to constrain the models.   

Physics-based Inverse Design of Elastic Rods with Deep Neural Network

Rod-like structures, such as DNA, climbing plants, and cables, pervade the nature and our daily life, which usually assume a deformed shape based on the competition between elastic (stretching, bending, twisting) and external (e.g. gravity) forces. We take a combination of physics-based and machine learning-driven approach to tackle the problem of obtaining the undeformed shape, given the deformed configuration under gravity. Conventional methods typically couple a numerical simulation of the elastic rods with optimization subroutines. We focus on rapidly solving the inverse problem in real-time using the Discrete Elastic Rods (DER) algorithm to simulate the forward problem and deep neural networks (DNNs) to store a large number of solutions. However, problems involving very long rods cannot be stored in DNNs due to the requirement of vast training dataset and insufficient precision in complex regression problems. We overcome this issue by decomposing a large rod into a series of smaller parts; the solution involving the smaller parts are stored in DNNs. Using the balance of forces and moments at the joints between two smaller parts, these smaller solutions are combined together to construct the larger solution.   

Synthesizing arbitrary lattice models using a single degenerate cavity

We propose a general method to simulate arbitrary lattice models by manipulating optical synthetic dimension in a single degenerate main cavity. Such a cavity supports a large number of degenerate optical modes with different angular momenta. Couplings between different optical modes can be readily controlled. These features allow us to simulate lattice models that are not convenient to realize using other systems, particularly models in high dimensions and with complicated hopping amplitudes. As a concrete example, we demonstrate how to construct two topological lattice models: the two-dimensional Haldane model and a four-dimensional time-reversal invariant model. For the latter case, we show how topological properties can be detected from the outputs of the cavity, where the second Chern number can be extracted. In the presence of open boundaries, the chirality of the Weyl edge modes can also be detected using the input-output formalism of the cavity modes.   

The effect of mass fluctuations on Dirac materials sensors for dark matter

Dirac materials with a non-zero mass term can form a small energy gap that is typically orders of magnitude smaller than accessible with semiconductors. This energy range provides the sensitivity to search for light (i.e. sub-MeV) dark matter while still blocking out the background noise. We study the effect of disorder through a discrete and continuous random mass-term, which both lead to tails (e.g. Lifshitz tails) in the density of states. The presence of disorder puts stringent constraints on the sensor material properties.   

Magnetic proximity and nonreciprocal current switching in a monolayer WTe2 helical edge

The introduction of magnetic order into topological band structure gives to new phenomena such as proximity effect and nonreciprocal magnetoelectric effects. We couple the helical edges states in a 2D topological insulator, monolayer WTe₂, to a 2D layered antiferromagnet, CrI₃. We find that the edge conductance is sensitive to the magnetization state of the CrI₃, and the coupling can be understood in terms of an exchange field from the nearest and next-nearest CrI₃ layers that produces a gap in the helical edge state. We also find that the nonlinear edge conductance depends on the magnetization of the nearest CrI₃ layer relative to the current direction. At low temperatures this produces an extraordinarily large nonreciprocal current that is switched by changing the antiferromagnetic state of the CrI₃.
  

Far-field thermal radiation at the optical topological transition

We develop theoretical description of far-field thermal emission near optical topological transitions (OTTs). OTTs originate from a transformation in the topology of the phase space in the medium which results drastic changes in the photonic density of states. In this work, we show that the unique evolution of this transition leads to a characteristic signature in the far-zone thermal emission. In particular, we find (APL 115, 151906 (2019)) a strong asymmetric pattern of reduced emission near the transition from type-I hyperbolic to dielectric responses present in both naturally available and composite media. The existence of the asymmetric feature should be experimentally observable and can be used as a tool to detect the OTT between type-I hyperbolic and dielectric electromagnetic responses.

  

A bizarre state of confined He-3

Atoms under various types of external confinements have been scrutinized by theorists for many years now. Here, we focus on a He-3 atom whose 1s² shell is known to undergo a tiny hyperfine spin-splitting: 1s² →1s↑1s↓. We treat the spin-split He-3 within the known spin-polarized Hartree-Fock approximation which we modify by the incorporation of a spin-dependent adjustable parameter to model the hyperfine splitting. We place He-3 inside a spherical potential box of a height U₀, width △ (U₀ = 5 Ry, △ = 5 a.u., as a case study), and an adjustable inner radius R₀.The resultant atomic potential becomes a double-well potential with the inner and outer wells being separated by about 5 a.u. We model the pressure on He-3 by narrowing the inner radius R₀. We find that, at R₀ ≤ 0.85 a.u., the 1s↓-electron migrates into an outer well. The 1s↑-electron remains in the inner well as long as R₀ > 0.4 a.u. We, thus, unravel a possibility of creating a bizarre state of He-3 with the much different orbital radii and energies of the 1s↑- and 1s↓ electrons. At R₀'s ≤ 0.4 a.u. both electrons reside in the outer well, far away from the nucleus, and He-3 turns into a novel type of a low-n (n = 1) and low-I (I = 0) "Rydberg" atom.   

Benchmark calculations of the doublet S-states of the singly charged carbon ion

We report acurate variational calculations of the five lowest doublet S-states of the C⁺ ion. The wave functions of the six-particle system (five electrons and nucleus) have been expanded in terms of 16000 all-particle explicitly correlated Gaussians whose nonlinear variational parameters were subject to extensive optimization. The motion of the finite-mass nucleus has been explicitly included in the Hamiltonian, while relativistic corrections to the energy levels have been computed in the framework of the perturbation theory. Lowest-order quantum electrodynamics (QED) corrections have also been estimated. The results obtained for the energy levels have enabled the determination of transition frequencies with sub-wavenumber accuracy.   

Compact tunable plasma muon accelerator for ultrashort micron-scale muon pair beams

Experimentally accessible schemes of compact plasma muon acceleration are introduced and modeled using a novel technique of controlled post-processing of cascade showers. These schemes use propagating structures in plasma, driven as wakefields of femtosecond-scale high-intensity laser or dense particle beams, to capture muons of a divergent cascade shower of: (a) hadronic type from proton-nucleon or photo-production reactions or, (b) electromagnetic type. Apart from the direct trapping and acceleration of particles of a raw shower in a plasma wakefield, a conditioning stage is proposed to selectively focus only one of the charge states. Not only is the high gradient of plasma acceleration structures well suited for rapid acceleration to extend the lifetime of short-lived muons but their inherent spatiotemporal scales also make possible production of unprecedented ultrashort, micron-scale muon beams. Compact muon acceleration schemes hold the promise to open up new avenues for applications.   

Reinforcement Learning and the Cost of Observation

Reinforcement learning has recently been used for practical application in both physics and chemistry laboratories. When conducting experiments in these environments, measurements can often be expensive, either due to the resources used or the time that is taken. Due to this, measurements are typically taken intermittently or when exploring new conditions. If we were to use reinforcement learning to optimise a laboratory experiment, it would ideally act in a similar manner. Therefore, it would need to act without perfect information and balance the cost of observation with the need for new data.
In this work, we show how an agent can build an internal hypothesis of its environment, using experience from past measurements, that it can then act on. The measurements it will obtain come from various sensors that each have an associated cost. The informativeness will vary given the current state, so the agent needs to determine the most valuable information to sample. We also demonstrate how the agent’s behaviour changes when certain measurements are limited by cost.   

A modified reference approach implemented in Python (MRAIP) to obtain CO2 emissions in any region

There is a general consensus that good quality estimates of CO₂ emissions from fuel combustion are essential for the construction of inventories since they have direct link between fuel use and emission which is key to strategizing how to reduce these emissions. In the study here, we develop a modified reference approach implemented in Python (MRAIP) to obtain the CO₂ emissions in any location such as a country, region, state or town once its apparent consumptions can be determined. The simplicity of the MRAIP is that it can be used to estimate the amount of CO₂ that is emitted in n-litres of fuel combustion with limited additional effort and data requirements. Therefore, it can be used to obtain both small and large scale inventories even with QPython in Android smartphones. The MRAIP was demonstrated to obtain both states and national CO₂ emission inventories in Nigeria from 2000 to 2019. Further, we discuss simple ways to include other major CO₂ emitters which varies from one location to another such as coal/peat, natural gas, deforestation, cement, and even the minor petroleum products as well as bunker fuel which are not included in the MRAIP.