The journal Science has lifted an expression of concern on a paper claiming evidence of Majorana quasiparticles and added new details

Controversial quantum-computing paper gets hefty correction — but concerns linger

The journal Science has lifted an expression of concern on a paper claiming evidence of Majorana quasiparticles and added new details

An abstract representation of Majoranas, which are their own antiparticles.Credit: Ramon Andrade 3DCiencia/Science Photo Library

A key study¹ claiming to provide evidence of Majorana quasiparticles has received an extensive correction five years after it was published in the journal Science. Two researchers who flagged the paper as problematic say that the correction isn’t sufficient — triggering the latest dispute in a field dogged by controversy.

For decades, physicists have been compelled by the idea that ultracold electrons in microscopic devices could behave collectively to form quasiparticles resistant to noise — both environmental perturbations and the inherent atomic jostling that plagues all quantum systems. The resilience of these Majoranas could make them ideal candidates for forming qubits, the informational units in quantum computers that are analogous to bits in classical machines. Studies to prove their existence have come up short, although recent bold claims by technology giant Microsoft have drawn considerable scrutiny.

In September 2018, a team led by Charlie Marcus, a physicist at the University of Copenhagen in Denmark, who also worked for Microsoft at the time, posted a manuscript to the preprint server arXiv that described a novel approach to generate Majoranas. The researchers made nanowires of indium arsenide surrounded by a shell of aluminium. Applying a small magnetic field, they then measured electrical signals “consistent” with pairs of Majoranas, one at either end of each wire. A year and a half later, they included theoretical simulations to justify their results, and the study was published in Science¹.

Two physicists — Sergey Frolov, at the University of Pittsburgh in Pennsylvania, and Vincent Mourik, now at the Research Centre Jülich in Germany — raised questions about the validity of the data, and in July 2021, Science applied an editorial expression of concern to the paper to warn readers of potential problems. Now, Science is lifting that warning, and the authors are issuing a 20-page correction to the paper’s supplementary material. News of the correction was first reported on 31 July by the technology news site The Register.

The Copenhagen group looked for Majoranas with devices made of indium arsenide wires (green) wrapped with aluminium (blue), as shown in this coloured electron microscope image. Credit: Reprinted with permission from Ref. 1.

The authors say they are relieved by the outcome. “It’s not really correcting any errors,” says co-author Saulius Vaitiekėnas, a physicist at the University of Copenhagen. “We are summarizing and providing additional information.” Frolov, on the other hand, argues that the data in the paper do not give a full picture of electron behaviour in the team’s devices and calls for retraction. “I do not trust this data,” he says.

Jake Yeston, an editor at Science who oversees physical-sciences submissions, says that the journal decided not to retract the paper because there was not a “clear, community-grounded view that it’s obviously wrong”. But, Yeston says, the lack of information in the original paper was a problem, and it has now been fixed. “It shouldn’t be that a reader who wants to know what your protocol was has to go to your lab and talk to you,” he says. “That should be in the paper.”

Questioning the data

Thirteen years ago, Frolov and Mourik were authors on a different study in Science² that reported evidence for Majoranas. But excitement around the result faded after researchers discovered that other mundane phenomena could mimic the quasiparticles.

When the Copenhagen team’s manuscript posted to arXiv in 2018, Frolov and Mourik were dubious so they requested to see all of the data. E-mails reviewed by Nature show that the Copenhagen group released additional data in November 2020. The pair of critics analysed the information provided and concluded that the data were incomplete and contradicted the study’s central claims. An internal inquiry by the university’s physics institute, however, found “no problems with the paper”, and that the Copenhagen team had turned over all of its data. Unsatisfied, editors at Science applied an expression of concern to the paper, and in October 2021, Yeston filed a complaint with the university to request an “independent, transparent investigation by experts”.

In June 2022, the university assembled a panel of independent physicists to undertake the effort: Sophie Guéron, at the University of Paris-Saclay; Allan MacDonald, at the University of Texas at Austin; and Pertti Hakonen, at Aalto University in Finland. They travelled to Copenhagen, conducted interviews and examined data from 60 microscopic devices (the original paper included data from 4). Their year-long investigation³ found no misconduct, but stated that the team’s selection of data led to “conclusions that did not adequately capture the variability of outcomes”. The excluded data, however, did not undermine the paper’s main conclusions, they said.

One sticking point for Frolov and Mourik continues to be the Copenhagen team’s choice of ‘tunnelling regime’ — the range of low electrical conductivities over which the devices were scanned. The Copenhagen researchers said they saw signs of Majoranas persisting “throughout” their chosen tunnelling regime. But Frolov and Mourik said that the additional data they received showed that the tunnelling regime was much wider, and that the telltale Majorana signs were limited to the smaller tunnelling window.