When you meet a colleague at #apsmarch, you discuss each other's work, family and war in 🇺🇦.
Then you may say 'oh, have you seen ...' I suggest the subject of that discussion my work with @VincentMourik on investigating unreliable research on the topic of Majorana.
A summary 🧵
1. The retracted Nature paper from last year is widely known. But do you know that there are several others hopefully headed in the same direction? They are...
3. A two-part @NatureComms + @NatureNano saga on 'Ballistic' Majorana. We have already shared a full post-publication analysis on the first part, the second part has...
5. It is not just Delft, @QuTech_news , but also another @MSFTQuantum center in Copenhagen, which has published this paper we take apart in a 38-page post publication review.
When it comes to the Copenhagen team, we have questions about several other papers, but the PI Charlie Marcus refuses to share data. The papers don't add up though, from the physics point of view and from what we know about these nanowires.
6. Last but not least, the infamous 'Chiral Majorana' paper @UCLA. Widely panned and praised at the time. It is totally unreliable and MUST be retracted immediately @ScienceMagazine . This is from a different research group and not investigated by us
We found another Delft paper unreliable - Nature Nanotechnology 2018🧵. The stated innovation of this work is in the word 'ballistic'. (Majorana evidence echoes earlier work and is based on zero-bias peaks.) What is the meaning of 'ballistic' here?
According to authors, there are no quantum dots (claim familiar from Nature 2017) and there is 'quantized conductance' (familiar from Nature 2018). Illustrated by Fig1d (black trace plateau at 1) and quotes from the text.
We cannot confirm this central 'ballistic' claim.
When we uncrop Fig1 using Zenodo data (zenodo.org/record/4721357) the first thing we find is vertical resonances indicative of quantum dots. This contradicts 'ballistic' transport electrons fly smoothly through the wire. Quantum dots are puddles that hold and scatter particles.
I celebrate #PiDay2021 in my own way. To me pi will always be about this.
My first physics result: "Measurement of the Current-Phase Relation of SFS pi-Josephson junctions" with magnificent junctions from Chernogolovka, and experiments done in Urbana: arxiv.org/abs/cond-mat/0…
We also hunted second-order Josephson effect at 0-pi transition, where that randomly wiggling purple trace is. I did not find it. I remember Leggett asking 'why is he graduating?'. The team found it shortly after I left. But it took 10 years to publish! arxiv.org/abs/1805.12546
The most important figure in the retracted 'Quantized Majorana Conductance' paper was Figure 2e. A tiny technical panel based on the same raw data as the infamous 'plateau'. What's the big deal? Read on...
First, let's take it in. You see how perfectly the red datapoints all line up at 1 (the quantized value). Members of the Academic Community already memed about it: even the totally commanding theory could not get so close to 1.
Another thing we noticed that was strange: if these data came from the 'plateau' trace, why so few data points? And also, how come in panel e they are all pinned to 1 but panel b from which they were taken shows about 10% scatter around 1?