Comprehensive coverage

On science, transparency and advertising lies. Did Nature fall into the trap on the issue of superconductivity at room temperature?

After already retracting a previous article by the same researcher from 2020 that claimed superconductivity at high temperatures, he again publishes a follow-up article even though it also revealed considerable problems

By Dr. Nimrod Bacher, Department of Physics, Ariel University. The writer was a member of one of the groups that checked the claims in the article and found errors in it

On the same subject: Thunderous silence following the claims of finding a conductor at room temperature

If you thought that crime scenes and telenovelas were limited to the television screen, then you will know that even in the world of research science sometimes interesting and juicy stories like the one I will tell you now happen. Let's start from the end.

Last week in Las Vegas the conference of the American Physical Society (American Physical Society) or as it is known, the March conference since it takes place in... March. At the conference there was a lot of buzz because a researcher named Ranga Dias was going to present his new work on a superconductor at room temperature. The new article was published the day before in the respected journal Nature and Dias could break the bond of silence and give a lecture on the subject. There was of course interest and the room was full to capacity, Dias presented his work but the audience did not seem particularly impressed [1].

You must be asking why? If you were to ask today's physicists, they would tell you that the "holy grail" of physics is to find a superconductor at room temperature. If we do this, we will be able to build an MRI device at low costs, travel in trains that will float on magnetic tracks, transfer electricity from one place to another without losses, build frictionless motors and more, and this without investing in cooling power as is currently the case when we use superconductors. The discovery of a superconductor at room temperatures will surely win the researcher a Nobel Prize, fame and publicity until the end of time and inexhaustible financial investments.

So why was the audience not impressed? Because the same Dias had already "worked" on them in the past. In fact, Dias published an article in 2020 that showed superconductivity at room temperatures in another substance called CSH, a mixture of carbon, sulfur and hydrogen, with the last of the components being the most important in our stories [2]. The article was continued (a gentle form that says deleted from the pages of the newspaper) at the behest of the newspaper Nature, the same newspaper that gave Dias another platform to publish his new work on a substance called NLH, a mixture of nitrogen, lutetium and of course hydrogen [3]. And why did they pull the previous article? Because they discovered that Dias or one of the other authors in the article was not honest in the publication and did not explain or actually lied about the results that seemed to other researchers in the field to be fabricated results.

What is a superconductor?

To understand the story in depth, you first need to understand what a superconductor is and how to prove that a material is a superconductor. Well, a superconductor is a material with two main properties: the first is that it has a resistance equal to zero. Not close to zero, not almost zero but absolute zero. If an electric current (or electricity) flows in your copper wires in the wall, then the power you sent to the wire undergoes losses due to the resistance, and these are mainly expressed as heat. You know the resistance effect from heaters and incandescent lamps of old. There it was a tungsten wire in which a current flowed, and the resistance caused losses to heat energy that we enjoyed in the bath or in the form of emitted light. To prove that a material is a superconductor it is necessary to show that its resistance is zeroed out, that is, there is a transition (phase transition in physical terms) from a conductive material with a finite resistance to a state of zero resistance. In fact, if we take a superconducting wire, flow a current through it and close the wire into a loop, we can leave a letter to our grandchildren who will measure the current in 100 years and the current will remain as it is and there are no losses.

The second property is trickier to explain in popular science and is related to the quantum properties of superconductivity. A superconductor repels the magnetic field trying to penetrate it to such an extent that the magnetic field inside a superconductor is zero. It becomes a perfect diamagnet, a material that produces an internal magnetic field opposite to the one applied to it so that the total internal field is zero. Diamagnetism is not a phenomenon that exists in a perfect conductor, whose resistance can also be zero (if such a material existed), but it does exist in a superconducting material. The repulsion of a magnetic field causes a superconductor to hover over a magnet or vice versa. Therefore, rejection of the magnetic field, or what we call in professional language the Meissner effect, named after the physicist who discovered it, is the second and most important property to prove that the material you have in your hands is really a superconductor.

In order to understand the sequence of events that I will describe to you shortly, you need to understand that Dias did not wake up one morning and suddenly think of the substance CSH, create it and measure it. The family of materials called hydrates, meaning hydrogen-based materials, had already been known for several years. We found in them for the first time [4] an effect of superconductivity, although only in the form of a resistance equal to zero, at temperatures that are not room temperature but still higher than many of the other materials in the group we call superconducting materials. Until that surprising finding in hydric materials by a German researcher of Russian origin named Mikael Armets, the peak temperature of the transition to superconductivity was in the region of 160 Kelvin or minus 113 degrees Celsius. Quite cold, but not for physicists who work with cheap liquid nitrogen and not with expensive liquid helium. Armets broke that record and set off the modern day superconducting hydraulic gold race.

The subject of hydrates has been under discussion for many years in solid state physics. Ashcroft, a world-renowned scientist whose number is studied by all physicists today in their first degree, claimed that the bonds between the hydrogen atoms are strong enough to produce the conditions sufficient for superconductivity at high temperatures [5]. Furthermore, Ashcroft went further and claimed that the hydrogen-hydrogen bond could provide superconductivity at high temperatures and without stimulating new theories on the subject. Everything comes from the same theory that explains superconductivity in lead and aluminum and was limited only to very low temperatures. All that is needed is to turn the hydrogen into a solid substance that will probably also be metallic and then hope that it will also be a superconductor.

Huge pressure

Where is the catch you ask? Well, all these hydric materials that have broken records do so at temperatures close to room temperatures but only when pressed between two diamond heads to insane pressures in terms of scientific research in a laboratory. what is crazy In order for the material to become a superconductor, the pressure chamber is brought to pressures equivalent to those prevailing in the Earth's mantle and the core of the sphere, on the order of 100 gigapascals. We live and breathe at atmospheric pressure, which is about 100 kilopascals, which means we need another 6 zeros in the order of magnitude of the pressure to make the material a superconductor.

If you want to stop reading and say, ok, then what's the point? What is the idea of ​​proving that there is superconductivity at such high pressures if it is impossible to do anything with it at atmospheric pressure where we live? How do we use such material? Well the answer to that is that basic research is designed to show that this is possible. In a practical way, it can be said that if it is possible to bring the material to a certain pressure by external means, surely you will find a way to do it by internal means, for example, replace some atoms in the structure of the material and these will attract each other more strongly so that in the end we will effectively get a pressure equivalent to the pressure the external This action is called chemical pressure and physicists use it a lot to learn about the properties of different substances in nature. Well, if this is true and there is superconductivity at high pressures, nothing would stop physicists from producing a superconductor at room temperature and at a lower pressure than that found in the laboratory pressure chamber. Nor manipulation of data...

We will now return to the hero of our stories, Ranga Dias, a scientist who was born in Sri Lanka and now lives in the USA. He did his PhD research work at the University of Washington in the field of superconductivity at high pressures. He received a senior lecturer position (Assistant Professor) at the University of Rochester in the USA and began working to realize his research ideas, superconductivity at room temperatures in hydrated materials. As I already told you, in 2020 he published the article in Nature about the material CSH in which he managed to be the first researcher who not only shows zero resistance at the temperature of a home air conditioner but also succeeds for the first time in measuring the effect that everyone before him failed to do, the Meissner effect in hydric materials. Rejection of the magnetic field and proof that hydrates are indeed superconductors in all senses. Eureka!

Now we will move on to the second hero of our stories, Messiah on the white donkey, Don Quixote of Spain, the fearless fighting knight Prof. Jörg Hirsch from the University of San Diego. Hirsch, for his own reasons and of course in direct connection to his theoretical explanations, does not believe in superconductivity at high temperatures in these materials. That's why Hirsch made a decision to thoroughly check every article that claims superconductivity in materials that don't really match what he thinks. Hirsch turned to check Dias's article and became suspicious of the graphs drawn there. "The transitions are too sharp", he said to himself... it doesn't fit these materials and previous publications. Something in the graphs seemed suspicious to him and he asked the authors and Medias himself to send him the data of the graphs in the article.

Request to receive the experimental data

Make no mistake, this is a legitimate request and every researcher usually shares the data with his colleagues. Even the newspaper itself publishes a notice that the data of the study is available to the public by personal request to the author of the study. The goal is twofold: the first is to check repeatability, meaning that someone else will be able to do the same experiment because it is physics and it repeats itself. The second is to create transparency in science because in the past science knew cases where data was falsified or invented out of thin air (see the Jan Hendrik Schoen scandal entry). Dias, treated the requests of Hirsch, who was already known as a scientist who examines this issue in depth, as somewhat trollish behavior and did not respond to his request. This of course raised suspicions and in joint work by Hirsch and Prof. Frank Marsilio from the USA, the two wrote an article claiming that the data presented in the article look suspicious and expanded their physical explanation on the basis of previous works [see a later version of the article in reference 6]. Dias finally responded to the requests and at the end of 2020 published the complete data in a file attached to the original article in Nature. The data included what was published and the raw data, i.e. what was actually measured in the laboratory. We'll just add, seemingly...

At that point, Hirsch added Prof. Dirk van der Marl from the University of Geneva to the team and asked him to help him go through the published data. "Something there seems suspicious to me," Hirsch told him. The two went through the data of the magnetic measurements, redrew the graphs and wonder and wonder found discrete jumps in the data. Not only did they find jumps, but these are jumps in complete multiples. The researchers asked themselves, why are there digital jumps in the measurement of a function that should be continuous? Is this a jump that the measuring device might produce?

After a simple analysis and eliminating the jumps in a sophisticated way, wonder and wonder, a graph that looks like a step graph (or rather, a cliff of Mitzpe Ramon) and a transition from one high value to another low value, now looks more like soft hills in the sands of Rishon Lezion. No passing, no falling and no shoes, a continuous graph that looks like it came out of a boring graphing machine.

The two decided to publish an article in which they explained the analysis they had done of the data, showed the digital leaps in information, detailed the data they received from the researchers and asked "what was it like?" Or rather, they concluded by saying "we don't know where these data came from". [7]

Graph from reference [7] showing the discrete jumps in the data (c,d) and how the data looks after removing these jumps (eh).

Full disclosure: in my conversations with Dirk at the time, I tried to be the devil's advocate and try to explain the findings in instrumentation problems. After he proved his claim to me, I said that maybe there is something hidden in the unpublished background data and maybe there are jumps there that "leaked" into the data of the article as published. I just couldn't believe that someone would bother so much to fake it...

The angry Dias decided to answer and put an end to the story. He and his friends wrote a detailed article in which they explained their claims, explained their findings and even hinted that Dirk and Jörg were idiots "who did not understand what we really did in the analysis of the data in the article" [8]. Among other things, in the response, the researchers included the data as published, but a more important part of the story is the background measurement data that was missing.

The background measurement is missing

What is the background measurement? Well, to know that what we are measuring is true, we need to understand that the material being measured is like a small pebble in a huge pool made of concrete and metal and filled with water. If we want to understand the properties of the pebble, we will first have to measure its environment as a calibration measurement and then understand what the pebble does to the entire measurement when it is placed in the pond. In other words - measure the pool with the water without the pebble and then measure the same thing with the pebble. The difference between the measurements, if you did do them correctly, is the contribution of the pebble. True, it sounds difficult, but it is possible and if Dias managed to do it the first time, then he really deserves world fame.

And here is what Dias said in response to the question about the background measurement: "We didn't measure a background measurement at all, as you tried to claim in your claims article, what we actually did was take the original measurement, move points in it and call the new graph we created, a background measurement."

wait, but what????? Yes, that's how Dirk, I and a few other colleagues sounded when we read this comment from Dias...

I want to take you back one step for a moment and mention that Dias stated in the original article that "the calibration (or background) measurement was done on the same material but only at a lower pressure. Such a pressure that the material is still not conductive." This is a completely legitimate claim in the article, and if it was indeed made this way, then the difference between the high pressure measurement and the low pressure measurement is the effect created by the CSH substance. But now in the response article he says something else... he says very clearly, while describing the method rigorously, that he created the background measurement by taking the original measurement in which you see a step, canceled the step, called the new graph "User's background measurement" Background Data) and which he subtracted from the original measurement to "emphasize" the effect.

Well here in fact Dias shot himself in the foot and this is where Nature's investigation against the article actually began to grow stronger and reach its conclusions. Dias actually contradicted what he explained in the article and revealed that he used a method that is not only scientifically incorrect or common in the field of these measurements but also a method that he did not actually explain in the original article. This is not done in scientific research.

Nature finally took a very extreme decision and withdrew the article without the consent of the researchers. To understand why I used the word extreme, you need to understand that when a suspicion arises about an article, then Nature suggests that the researchers withdraw the article according to the investigation carried out. Decent researchers understand the hint and agree to the request since it is their reputation and if it turns out that there is a problem with the publication, then it is better to really withdraw the article. This was not the case with Dias, he insisted that everything was true and real.

The article continues but the story continues. In between, arXiv that entered this swamp pulled a different version from the server each time. Once they claimed that Dirk and Hirsch recorded their response in a disrespectful way. A second time they said it was Dias who did it. Each time a draft was advertised and after a while it was hard to find it. They even went so far as to ban Hirsch from registering any article on ArXiv because of this story. Finally, everything fell into place safely, all versions became available again [9] and Nature announced that they had officially retracted the article.

Using tools from the field of communication engineering

Some time later, Dirk and Hirsch publish an official article [10] of about 23 pages in which they lay out all the explanations for the falsification they found in all the data related to the magnetic measurements, including one very interesting explanation, an explanation of the compatibility index in the family [11]. Sounds strange but be patient and I'll explain... every engineer (like me) learns that between two random signals there is no match or correlation. Noise is a random signal. If we check a match between one noisy independent measurement and another noisy independent measurement, then there will be no match between them. On the other hand, if within the measurement there is a signal that is present in both measurements, then checking the correlation between the measurements will emphasize to us that there is a match. This is called cross-correlation in the professional terms and it is a mathematical tool used to find signals in noisy measurements.

So what about the family? Let's assume for a moment that we have a father who is called "Background" and a mother who is called "Galem". If we "subtract" from the mother "incarnate" the father "background" we get the child "letter". In terms of the article published then, the mother is the measurement that was actually performed, that is, the raw measurement or Raw data. The father is the background Background data (I repeat that the background measurement is the measurement of the entire pool without the pebble or in Dias's case the same background he created from the original measurement), which we are trying to get rid of, the other fathers in the audience will forgive me. Finally, the child is the letter or the information that was finally published as a graph in the article, this is the Published data. The child is supposed to be the mother minus the father when there should not be a cross match between the mother and the father but there should be a match between the child and the mother and between the child and the father. In mathematical terms, we write child = mother - father. Well this is a normal family in terms of signals, noise and correlations.

So what happened when running the cross-correlation tool on the Dias data? Well there is a correlation between the mother and the father, it should be clear in the case of Dias because we created the father out of the mother as argued in the Dias response article. There is also a correlation between the child and the mother, this should also be clear because what was published probably also came from the mother. But God forbid, there is no correlation between the child and the father... If you examine this claim in depth, you realize that the mother is actually the product of the child and the father. In mathematical terms we will write down mother = father + child... Not only is it not normal, this is a story of incest with the seasoning of the story of Jesus and Mary (or Moses the father in this case) the saint in all the terminology of signal and noise processing!

Illustration of matches between data from reference [11]. On the left is a "real" experiment and a lack of correlation between the background and the raw measurement. On the right, the situation with the data as found in Dias' article.
Illustration of matches between data from reference [11]. On the left is a "real" experiment and a lack of correlation between the background and the raw measurement. On the right, the situation with the data as found in Dias' article.

In a framed article, Dirk attributes the suggestion to use this tool to an anonymous Reddit user who contacted him and suggested that he check the cross-match between the data. I assume that the same user is also an engineer since in retrospect I explained to Dirk that this is a well-known signal analysis tool in the world of communication engineering and I also studied it in the past in electrical engineering courses. When I suggested to Dirk that it might be useful to introduce mandatory courses for physicists in the field of random signals and noise and signal processing, he replied, "Are you crazy? You want to teach physicists like Dias how to fake better???”

An article by researchers alleging forgery in Dias's first article

If you thought this was the end of the story, then you were wrong. After all, it is known that in the cases of fraudulent researchers, the signs of this appear earlier. They only get better over the years as long as they stay under the radar.

Let's start with the first sign that didn't exactly point to the immediate suspect in our story. It turns out that in the continued article there was a researcher who did the magnetic measurements and he signed another article in a newspaper called Physical Review Letters (PRL) in which he showed that a substance called europium also becomes a superconductor at high pressures [12]. Hirsch, our friend from before, has already checked the article and sent PRL his claims about problems with the graphs from this article. It is important to emphasize: Dias is not at all listed as a co-researcher for the above article and is not directly related to it. The common denominator is the researcher named Matthew Debsay who is responsible for the magnetic measurements in the two problematic articles.

PRL did their research and one of the authors, James Hamlin, who was also a PRL signatory, began to check all the data of Debsay's work and found that the data had been manipulated. In addition, since Hamelin works at high pressures, his student discovered that there was indeed a side effect at the temperature at which Debsay reported an effect in europium. When the measurement was repeated with other means that neutralized the side effect, no evidence of europium superconductivity was found and the article was continued.

So if Same Matthew Responsible for the two problematic magnetic measurements in two different papers on different materials, so he is probably the master forger...well, it would be reasonable to assume so but the story keeps twisting and turning.

Playing with the measurement and copying data?

Hamlin's suspicions after the above examination and after seeing the work done by Dirk and Hirsch and conversations with Hirsch on the subject, made him realize that probably the other data in Dias's Nature article are also problematic. He decided to investigate the electrical resistance measurement data in Dias's 2020 article, which also seemed to him to be fake. Since he was not provided with the data, he was able to extract the raw data from the graphs [13] and found that they also have unexplained digital jumps [14].

That is, the conclusion is this, the entire article includes the magnetic measurements that Matthew probably made and also the electrical measurements that Matthew did not make, they are all a story of playing with the measurement data. Apparently all the data was manipulated to produce the graphs in the article and prove the claim of superconductivity. So apparently someone else is pulling the strings here, sorry... in the data.

Hamelin did not stop here, these suspicions increased when he saw another work by Dias on a substance called MnS2 [15]. From his familiarity with the subject of material measurements at high pressures he compared the data of Dias' new article in PRL with Dias's thesis and again found a match between the data. But wait... what's the problem? Well, Dias' thesis from 2013 did not deal with the MnS2 material at all but with the GeSe4 material. But if you draw the data one on top of the other you get a completely identical graph at low temperatures! Simon Kimber, who signed the aforementioned PRL, received an email from Hamlin, and after realizing that there was a forgery problem, he himself decided to write to the PRL and request that they investigate the article he signed.

A comparison made between the data from the PRL article in 2021 and Dias' thesis work from 2013 on completely different materials [18].
A comparison made between the data from the PRL article in 2021 and Dias' thesis work from 2013 on completely different materials [18].

Did the mask of lies stop here??? of course not! When Hamlin examined Dias's thesis, he discovered sentences in it that were very familiar to him. In fact he discovered whole paragraphs that he knew. Hamelin decided to investigate the issue in depth and entered his own thesis and Dias's thesis into the tool for detecting copies (iThenticate). It turns out that entire paragraphs were found in both theses. Dias simply copied entire paragraphs including mistakes he made in referencing some of the quotes in the paragraphs he copied. Furthermore, in the joint response article by Dias and Salmat (the second lead researcher in the CSH article) entire paragraphs were found that were copied from Hamelin's thesis [*]. In fact you are welcome to try it yourself... [16].

Comparison of parts of Hamelin's thesis (left) with the article published by Dias and Salamet (right) [8]. From Dan Gristo's article [*].
Comparison of parts of Hamelin's thesis (left) with the article published by Dias and Salamet (right) [8]. From Dan Gristo's article [*].

This is where the mask of lies and forgeries found so far ends. Now you can go to the new article published on NLH and presented at the last March conference. This time Dias published the data together with the article and if you read it then you discover some transparency about the way the data was analyzed. Also in the presentation given by Dias at the March conference he explained his background and deduction from the measurements. Despite this, it turns out that if you take the data and draw it, you find that the background is not that small. Dias did background subtraction operations even from measurements where it is not necessary to do so. Such a measurement is the resistance measurement - if a material becomes a superconductor, it actually shortens the contacts and the measured resistance will be zero. There is also a technique that Dias used to avoid measuring the contacts themselves as many researchers in the field do. Despite this, the raw data shows measurements where the resistance is not zero and in fact Dias subtracts from them a background measurement for which it is not clear what the reason is and then "produces" the beautiful graphs you see in the article where the resistance is zeroed. Many researchers agree that the method used by Dias is simply not particularly legitimate in measuring resistance as done in the article. In light of the history of the aforementioned researcher and the way in which he manipulates his data, there is no doubt that the findings should be taken with a limited guarantee.

On the left - the original measurement according to the data published together with the article. On the right - the same measurement after subtracting "background" from it to produce a phase transition to a superconductor. Is it necessary to do this in a superconductor model in the configuration in which it was measured in Dias' article? No! (From a tweet by Anna Akrap, a colleague of the writer).
On the left - the original measurement according to the data published together with the article. On the right - the same measurement after subtracting "background" from it to produce a phase transition to a superconductor. Is it necessary to do this in a superconductor model in the configuration in which it was measured in Dias' article? No! (From a tweet by Anna Akrap, a colleague of the writer).

¨ × ×

So where did we go? Dias' article describes a material that can be easily obtained in the laboratory unlike the previous materials that only had a few experts in the world who were able to do this. The most important law in physics is repeatability in measurements, so now many groups will work to try and reproduce the material and see if they see superconductivity effects. Dias's group managed to reproduce the CSH substance in the laboratory, but there is no clear evidence of superconductivity [17] or at least we do not see a repetition of the same experiments exactly as they appeared in the ongoing Nature article. We will emphasize that Dias still shows effects after the measured data has been manipulated which is not always explained. Claims about superconducting materials sound like news to readers and not every material that shows any change or jump in the data is a superconducting material. In fact, if we all follow the methods used by Dias, we can find many more superconducting materials in nature that actually only result from a very aggressive analysis of the data.

We have a few questions to ask:

  1. How did Nature decide to let Dias's latest article go out into the world without crucifying the researchers and asking them for the data? In fact, those graphs that show that the data has once again been heavily manipulated, if they were available to the reviewers of the work and if they were serious reviewers in the field, then this article probably would not have passed the screening of the judgment. For some reason Nature not only published another paper with heavy manipulation of the data but allowed Dias to quote his own paper which was already pulled from Nature! This story severely damages the credibility of the once very respectable newspaper.
  2. Will Dias, who received research funds and raised funds (see entry below) based on preliminary data that was later published and continued, be forced to return the money? What is the law of the company that he and Salamat started on the subject of superconductors at room temperatures? In a framed article it is said that Dias stated at another conference that two of the investors in the company are the CEOs of Spotify and OpenAI. It turns out that this is a lie and after confronting Dias with him, he explained that he was just describing potential investors. Amazingly, the YouTube recording of the lecture was subsequently deleted to prevent us from verifying the statement.

The questions we are currently asking ourselves are much more basic and relate to the disease that academia is suffering from today. You can also read about this in detail in Tamar and Oz Almog's book "All the Lies of the Academy" in which there is a chapter that describes the aforementioned topic in detail [19]. The Nature newspaper was known as a respected newspaper in which they published, for example, the discovery of the particle called the neutron and other immeasurably important discoveries. Since then, the newspaper has turned others like it into an advertising monster and an industry turning over millions of dollars. A researcher who wants to publish in these newspapers has to come up with a sensational discovery, one that will make the newspaper sell more according to the opinion of its editors. But the curse is also double for the researcher himself. In order to get a position in the academy or to receive funds from research funds, the researcher must show that he has experience in publishing in such newspapers. Most research funds and universities look at the researcher's CV and list of publications and if the latter has many publications in influential newspapers then this is considered a kind of sign of success. Well, more than once research groups sin by hiding questionable data or analyses, publishing desired results and only those (what we call cherry picking in the professional parlance), giving excessive or incorrect interpretation of the data and so on just to enter the golden gates of the newspaper. Don't get me wrong, there are many good articles in these newspapers and certainly scientifically correct as well. But the urge to move forward sometimes also creates the urge to exaggerate or sometimes to lie. The academy itself discusses these fateful issues every day and it is very difficult to know where the wind will blow in the future when it comes to scientific publications. For sure, this story will affect the journal Nature and the attitude of many researchers regarding the publications coming out of there (see the entry of an ongoing article regarding Majorana Fermions).

The funds require publication of data files at the same time as the article

Well, this is a very interesting period and actually takes us back in time to the 2000s when physics learned first hand the price of scientific falsification. Scandal Schoen, named after the German scientist Jan Hendrick Schoen, became a beacon in the world of solid state physics. Sean actually falsified entire sets of data published in papers like Nature and Science. He won many awards for this and his name was even mentioned among the candidates for the Nobel Prize. Very quickly other groups that tried to reproduce his work showed that this was not possible and thus exposed the bluff. In fact, the request to reveal the research data today and publish it with the article was born from that scandal. Many research foundations, including the Swiss National Research Foundation, now force researchers to publish the data in a separate file for each article that will be published with the support of the foundation's funds. European foundations such as ERC are also considering taking similar steps and are currently forcing researchers to publish their findings in "open" journals free of charge. It is interesting to know if the Israeli National Science Foundation will take the same steps following the global scientific introduction on the subject and stories such as the Shawn scandal. All that is left for us now is to wait for the reconstruction work and repeatability from other groups to understand whether we have now experienced the Dias Scandal firsthand or whether it is really the inventor of superconductors at room temperatures. Using a familiar idiom it is said that either Dias will eat the hat or we will take the hat off to his face.


[2] E. Snider et al. Nature 586, 373, (2020)

[3] N. Dasenbrock-Gammon et al., Nature 615, 244 (2023)

[4] AP Drozdov et al., Nature 525, 73 (2015)

[5] NW Ashcroft, Phys. Rev. Lett. 21, 1748 (1968)

[6] JE Hirsch et al., Phys. Rev. B 103, 134505 (2021)

[7] D. van der Marel et al. (January 2022)

[8] R. Dias et al., 2021

[9] D. van der Marel et al. (August 2022)

[10] D. van der Marel et al., International Journal of Modern Physics B 37, 2375001 (2023)

[11] D. van der Marel, personal website,; Van der Marel's talk at the online workshop 'Does Condensed Matter Physics need to worry about a reproducibility crisis?' (March 2023)

[12] M. Debessai et al., Phys. Rev. Lett. 102, 197002, (2009)

[13] J. Hamlin,

[14] J. Hamlin's talk at the online workshop 'Does Condensed Matter Physics need to worry about a reproducibility crisis?' (March 2023)

[15] D. Durkee et al., Phys. Rev. Lett. 127, 016401 (2021)

[16] Daniel Arovas, Facebook account[0]=AZUXc0AunOFqi80LmuIUXiw_R3nrQg9YLPkdM8F2kUVpok9PohEFnWs688vB3IT6KJ4KIoNoOSstUK9Z3VcWQ9Qq_s8QBF3Z-aEC5V1hJ2pV_GCQaGrbr2rA8Tx7dyrAaI7rZqeP-x_AREKde969zvpTAylyrMoa7fJDc0OIRGUa6g&__tn__=%2CO%2CP-R

[17] H. Pasan et al.,



[*] The above was written from close familiarity with the story as a research assistant in Dirk van der Marl's group and personal conversations with Dirk throughout the story and as a follow-up to a recent article by APS reporter Dan Gristo

10 תגובות

  1. Father + mother = child
    What courage do you have to get involved with progress like this...

  2. Well done Nimrod! There are not enough incentives in the scientific publishing system to invest in reproducing articles. They just want to bring out the next new discovery without verifying that what was before was true. Works like yours are important to prevent a scientific crisis in physics like there is in other fields of science.

  3. A long and pointless background article. Where does it lead at all except to stay in place?

    Who brings content without background?
    How does one who uses the Holy Spirit know what to bring?
    And why is that?
    Because the background is full of details that hide the point and this is to ensure that the person remains in a mental prison. A divine genius that leaves everyone else out.
    Christianity is a shell and a shell sucks from a point of truth and you should learn the right way to define what is spirit and what is holiness.
    Now is the time to choose to get out of the straits or stay in the respectable and soporific mental comfort.
    Drawings and graphs just like in the news have long shown that this is not the truth but a tool to make sure you stay where you are wanted.
    Sometimes you have to throw all the history (external science theory) in the trash to be able to smell something new.
    But the truth is that only those who seek the truth will have the right to act in this way, and the truth is the Holy One, blessed be He.

  4. It is impossible to understand the concept of superconductivity, without the concept of "passive time" which is absolute rest and absolute cold.
    Passive tense appears in Asbar's physics.
    To get to Asbar's physics, you must type "Asbar's magic journey on the wings of natural knowledge"

    This innovative physics is adjacent to innovative geometry, which deserves the name "physical geometry"
    In this geometry there is an exact mechanical experiment, which proves that Pi varies in a narrow range, between 3.1416 and 3.164

    Attached is the video about the scope experiment.

    A. Asbar

  5. Hello people,

    There are enough materials in nature that do not need to be under high pressure or harsh conditions to be good conductors. This is not what we are looking for.
    Dodge today has a crisis in the energy world, especially in the matter of electricity supply. We are in a situation where it cannot be solved in other ways (what I like to call the binary situation, either we use a new technology with all its limitations or we don't and then we are left with losses). If we use superconductors we can prevent these losses and therefore we have no problem living with the need to cool them with liquid nitrogen. In fact, there is a project in Germany that tries to pave such supply lines between cities. All MRI devices today work with superconducting magnets despite the high costs of helium gas. If it is impossible to do something in another way, then they will choose the new technology. In the case we are dealing with, it is not a material that is a turning point in its new properties. In any case, it is important to remember that the researcher's report is about resistance and not specific (dimensionless) resistance, so this material is probably very resistant relative to copper.

    Regarding the subject of the analysis and activation of advanced artificial intelligence tools - there is something in what you say. In any case, you have to remember that the models have to practice on something and this is where the "prejudices" of the researcher who builds the model come into play. Physics uses it today to analyze measurement data. It's hard for me to say where it will go, but as they say in the cyber world, the hacker will develop a malicious tool and then the cyber company will develop a defense against him, he will develop a new tool again and God forbid. There will always be a struggle between the researcher who will falsify and the researcher/newspaper who will check the article. That's why in the end what will win (as in this case) is the repeatability test. If it is impossible to reproduce this work anywhere else or alternatively when it is reproduced the phenomena reported by the researcher are not discovered then this indicates the incorrectness of the research.

  6. First of all thanks. Your article, Nimrod Bacher (as my son's name), sheds light on a particularly despicable human phenomenon.
    The article is in-depth and well-reasoned with many examples and all of them are written in clear and good Hebrew for every reader.
    I think that with the advancement of artificial intelligence models it will be more and more difficult to copy and falsify data, without being an expert in this field as well. If in the past it was possible to simply send a photo to a newspaper and that was all the evidence required, then today this is absolutely not the case. Also manipulation of data which can sometimes be very complex and especially if done maliciously, in the future can be detected by the machine.
    I'm not a physicist, but I'm interested if, after all, there is something to be done with such conductors, which may not be called superconductors at room temperature, but which have a very, very low resistance. OK, there will be losses. It is not "superconducting". Aren't these materials that have something to do with them despite these losses? Is there no advantage to a material with a resistance of less than 10 milliohms at room temperature?

    Cool article, thanks

  7. Within about a week since the publication, several works have already been uploaded (albeit only on arXiv) that do not reproduce the work of Dias and his partners but show things in reverse (some show a change in color as seen by Dias, among other things an indication of the same material, but without superconductivity):
    And the hand is still on.

Leave a Reply

Email will not be published. Required fields are marked *

This site uses Akismat to prevent spam messages. Click here to learn how your response data is processed.

Skip to content