Microsoft quantum lab retracts published paper: Readings that cast doubt on crucial discovery went AWOL
Quasiparticle eggheads were 'caught up in the enthusiasm of the moment'
Updated A paper published in Nature two years ago and spearheaded by a Microsoft scientist has been retracted after it emerged that the data presented simply didn't add up.
The work was produced at a quantum computer lab set up by Microsoft and QuTech, a research center co-founded by the Delft University of Technology (TU Delft) in the Netherlands. The study, led by Microsoftie and TU Delft Professor Leo Kouwenhoven, reported the discovery of a theoretical quasiparticle the academics believed would prove useful for future quantum computers.
"A 2018 academic paper published in Nature and led by one of our scientific directors, primarily in his capacity as a Professor at TU Delft, was retracted,” Zulfi Alam, a Microsoft Quantum unit veep, told The Register on Monday.
“As part of proposing the retraction, the authors of the paper took feedback from the scientific community, re-analyzed the data, wrote a new paper based on the analysis, and embraced the paper’s examination by independent experts in the field. This is an excellent example of the scientific process at work.
"We remain confident in our topological approach to scaled quantum computing, as we nurture a culture of continuous improvement to innovate across every component of the Quantum Machine.”
What went wrong?
That 2018 paper described a lab experiment that attempted to detect Majorana zero mode particles by measuring the conductance of superconductors at cryogenic temperatures.
That's important because scientists want to use Majorana zero mode particles as qubits in quantum computers as they believe they will be more robust at storing information than today's qubits. Majorana zero mode particles bring stability, in other words.
It would be a significant milestone in quantum computing to find and use these qubit candidates, and the Microsoft-Delft team claimed they had discovered these ghostly quasiparticles in their superconductor.
“Majorana zero-modes — a type of localized quasiparticle — hold great promise for topological quantum computing,” they wrote in their 2018 paper.
“Tunnelling spectroscopy in electrical transport is the primary tool for identifying the presence of Majorana zero-modes, for instance as a zero-bias peak in differential conductance.”
Microsoft: We've made a coding language for a quantum computer that may or may not existREAD MORE
Multiple graphs appeared to show the presence of these wonder particles. But it was too good to be true. The retraction note, published in Nature this week, said that after the team was told of shortcomings in their paper, they were unable to replicate their study’s results, and so decided they hadn't found the Majorana particles after all.
The note, headed by Microsoft's Kouwenhoven, admitted the team had left out measurements that would have cast doubt on their findings. It also acknowledged two researchers – Sergey Frolov and Vincent Mourik – who had scrutinized the paper and highlighted errors.
“Several inconsistencies were pointed out by Sergey Frolov and Vincent Mourik between the raw measurement data that was made available to them and the figures that were published in the paper,” it said.
“We therefore re-analysed all the existing raw data for our original measurements and rebuilt the original experimental setup for a recalibration of the conductance values ... When the data are replotted over the full parameter range, including ranges that were not made available earlier, points are outside the 2-sigma error bars. We can therefore no longer claim the observation of a quantized Majorana conductance, and wish to retract this letter.”
Frolov, an associate professor at the department of physics and astronomy at the University of Pittsburgh in the US, and Mourik, a research associate at the University of New South Wales in Australia, however, were a bit more blunt about the whole debacle.
We have found that the original source experimental data may have been manipulated, namely cut, as well as cut out and pasted together. Furthermore, entire datasets that contradict the central claim of the Nature paper were suppressed
“We have found that the original source experimental data may have been manipulated, namely cut, as well as cut out and pasted together. Furthermore, entire datasets that contradict the central claim of the Nature paper were suppressed,” they said in a series of slides showing how the data had been apparently massaged.
The pair, which had requested and studied the raw data from the team's experiments, noticed data points had been shifted on graphs' axes as well as an omission of evidence that contradicted the suggestion of a Majorana quasiparticle discovery.
"Taking the Nature paper and the additional data together, no evidence of zero-bias conductance quantization exists in this work," Frolov and Mourik concluded.
TU Delft conducted its own probe into the study, and while its Scientific Integrity Committee, made up of independent experts, agreed with the results of Frolov and Mourik’s academic sleuthing, we're told the panel found “no evidence that [the data manipulation] was intentional.”
The authors were caught up in the enthusiasm of the moment and thus did not pay enough attention to data that did not suit their purpose
“The experts consider the most plausible explanation to be that the authors were caught up in the enthusiasm of the moment and thus did not pay enough attention to data that did not suit their purpose,” the university said in a statement.
There were rumblings that something was amiss with the 2018 paper when a study shared on arXiv last month, led by Kouwenhoven and 21 co-authors, admitted the quantum computing lab had failed to find Majorana particles, Wired first reported. The team stated their original study was due to be formally retracted.
“The retraction of the article is obviously a setback in Majorana research towards the development of a quantum computer,” said Lieven Vandersypen, a scientific director at QuTech and TU Delft.
“Reflection on the methods used, must now run its course within the scientific community. Within QuTech we have already started that discussion. At the same time, we are continuing to work hard on our various lines of research towards the realization of a quantum computer and a quantum Internet.” ®
Updated to add
"I immediately understood that the paper was wrong," Frolov told The Register when we asked him what his first impressions of the Microsoft-Delft study were.
"There were so many problems that I could see them even without seeing extra data that were hidden."
"The paper, even if valid, should have minimal impact on quantum computing," he continued.
"The reason people thought it did was a publicity campaign. If Majorana paticles were firmly established – and this is something I am working on – it would be a tremendous basic physics breakthrough. From that to quantum computing is still a very long road, not five years and not even ten.
"And by then, we may have such advanced quantum computers that Majorana qubits will have even longer to catch up."