If I got this paper as a referee, I would reject it for lack of proof. arxiv.org/abs/2201.03453 (For anyone who is interested, I can be more technical in DM or email, details are not twitter-friendly…) Here are a few general thoughts:
First, there is the context - the subject is under scrutiny. E.g. one paper from the same group is under an Editorial Expression of Concern. science.org/doi/10.1126/sc… The concern is about non-representative data selection, but scientific root is - alternative explanation!
There exists a well-established NON-MAJORANA explanation for sticky zero bias peaks and related phenomena (subgap features). Here is a thread from a year ago where I explain Andreev Bound States in quantum dots, which are often unintended: x.com/spinespresso/s…
There exists a well-established NON-MAJORANA explanation for sticky zero bias peaks and related phenomena (subgap features). Here is a thread from a year ago where I explain Andreev Bound States in quantum dots, which are often unintended: x.com/spinespresso/s…
Is this known to the authors? Yes! But, like the Science paper above, the new manuscript does not discuss the potent alternative explanation. It is easy to see: papers about Andreev States mimicking Majorana are absent from the list of references. x.com/spinespresso/s…
Is this known to the authors? Yes! But, like the Science paper above, the new manuscript does not discuss the potent alternative explanation. It is easy to see: papers about Andreev States mimicking Majorana are absent from the list of references. x.com/spinespresso/s…
To me, data like these are a tell-tale sign of Andreev states from multiple quantum dots: I see several states criss-crossing zero source-drain voltage, with differing slopes Most likely, the sample has significant disorder that creates unintentional dots.
But wait, you say, they looked at not one, but two ends of a nanowire. That can only be Majorana? Of course not. In the 11-dimensional space of parameters, you can find two points with similar looking features. We do not call that correlation. An illustration from our paper: