"Smoking gun" signatures of topological milestones in trivial materials by measurement fine-tuning and data postselection

TL;DR

This paper shows that trivial effects in nanostructures can mimic topological signatures, emphasizing the need for careful data analysis to distinguish true topological phenomena from artifacts.

Contribution

It demonstrates how fine-tuning and data postselection can produce topological signatures in non-topological systems, challenging the reliability of 'smoking gun' indicators.

Findings

Trivial effects can reproduce topological signatures in nanostructures.

Data fine-tuning can create false positives for topological phenomena.

Guidelines are proposed for distinguishing genuine topological effects from artifacts.

Abstract

Exploring the topology of electronic bands is a way to realize new states of matter with possible implications for information technology. Because bands cannot always be observed directly, a central question is how to tell that a topological regime has been achieved. Experiments are often guided by a prediction of a unique signal or a pattern, called "the smoking gun". Examples include peaks in conductivity, microwave resonances, and shifts in interference fringes. However, many condensed matter experiments are performed on relatively small, micron or nanometer-scale, specimens. These structures are in the so-called mesoscopic regime, between atomic and macroscopic physics, where phenomenology is particularly rich. In this paper, we demonstrate that the trivial effects of quantum confinement, quantum interference and charge dynamics in nanostructures can reproduce accepted smoking gun signatures of triplet supercurrents, Majorana modes, topological Josephson junctions and fractionalized particles. The examples we use correspond to milestones of topological quantum computing: qubit spectroscopy, fusion and braiding. None of the samples we use are in the topological regime. The smoking gun patterns are achieved by fine-tuning during data acquisition and by subsequent data selection to pick non-representative examples out of a fluid multitude of similar patterns that do not generally fit the "smoking gun" designation. Building on this insight, we discuss ways that experimentalists can rigorously delineate between topological and non-topological effects, and the effects of fine-tuning by deeper analysis of larger volumes of data.