In an apparent effort to beat back questions raised about their controversial claim last year, Anshu Pandey and Dev Kumar Thapa, two scientists from the Indian Institute of Science, Bengaluru, have updated their preprint paper and made observers sit up with renewed interest.
When they first presented their results 10 months ago, several noted scientists were skeptical because the research text didn’t at all contain the requisite level of detail for what was always going to be a difficult claim to digest. For example, Brian Skinner, a physicist at the Massachusetts Institute of Technology, Boston, had found a pattern in a part of their data that shouldn’t have been there, indicating a major flaw.
More broadly, while many theorists felt they had enough to go with, experimentalists everywhere were frustrated. There was little in the duo’s preprint paper that would allow them recreate the experiment in their own labs and arrive at their own conclusions. Reproducibility is one of the pillars of scientific knowledge, and without it, the claim was always going to be stranded in doubtland.
A significant revision
But this time, armed with more measurements in a much larger sample set and with eight more collaborators in their team, the duo has re-advanced their claim of having observed superconductivity at ambient temperature and pressure in a nanostructured gold-silver composite material, and have opened themselves up to more scrutiny.
Superconductivity is the phenomenon in which a material offers zero, or vanishingly low, resistance to the passage of electrical current below a certain temperature, called its critical temperature (Tc). Since the mid-20th century, scientists have been looking for a material whose critical temperature is near room temperature – near 0º C – but so far without success. Finding such a material would round off a historic quest drawing from multiple branches of the physical sciences, as well as making a dizzying array of futuristic technologies possible, including high-efficiency power transmission lines.
To ascertain if a material has become a superconductor during tests, scientists check for some other properties. For example, the material would also become perfectly diamagnetic and display the Meissner effect: when an external magnetic field is applied to it, it expels the field from its interior. Put another way, a magnet placed above a superconductor will hover over it.
The significant part of the new paper, posted on May 21 on the arXiv preprint server – as compared to the earlier one in July 2018, which prompted so much controversy and doubt – is that the scientists report they have observed the material’s resistance drop to micro-ohms, which is 100x lower than before. They have also provided a description of protocols used to prepare the gold-silver nanostructured samples and details of the electrical and magnetic measurements. This should allow other research groups to try and reproduce this extraordinary claim of a room-temperature superconductor.
Also read: Did IISc Team Find a Room-Temperature Superconductor? The Devil’s in the Details
A prominent complaint against the researchers’ previous paper had been that they hadn’t included details of how they’d prepared their samples, so other research groups couldn’t attempt to reproduce their results. The authors had simply said that the samples were 10-20-nm wide silver globules prepared by “standard colloidal techniques” and embedded in a matrix of gold nanospheres. The latter are then cast into 25-nm-thick films on a substrate or into pellets.
As The Wire reported at the time, the structure, morphology and properties of the nanostructured samples prepared by the colloidal technique dependent crucially on the protocols used to make them.
Moar data
In their new submission, the authors report that they have investigated a total of 128 nanostructured samples of silver-gold composites, compared to just 33 earlier and which had been studied up to February 2018 before the first publication. The second round of investigations seems to have begun in September 2018 and continued up to May 2019, adding another 95 samples to the set. In the earlier set, Pandey and Thapa reported seeing vanishing resistance (less than 0.1 milli-ohms, the limit set by the sensitivity of the instruments used earlier) in five samples.
Now, the 10-member research group has reported seeing four more samples – nine in all – displaying near-zero resistance. They have also reported improved instrument sensitivity that tracks resistance down to 2 micro-ohms, which is two orders of magnitude better.
This in turn corresponds to a resistivity of about 10-12 ohm-metre – 10,000-times lower than that of good metallic conductors like copper, gold and silver. More significantly, the authors also report a much higher superconducting transition temperature of 286 K (~13º C) compared to the 236 K earlier. This is a remarkable jump, and 13º C is room temperature in many parts of the world.
The transition temperatures in the nine cases range from 150 K to 286 K.
Temperature-dependent resistance in two representative samples with transitions at 272.85 K (green circles) and 178.8 K (yellow squares). Source and caption: arXiv:1807.08572
The resolution of resistance in the low resistance state (2 micro-ohms), and the extent of resistance drop at the transition (nearly six decades), calculated from the slope of the I-V curves as shown in the inset. Caption and source: arXiv:1807.08572
Their updated preprint paper’s appendix contains a table detailing the nature of the samples (film or pellet; encapsulated or not), the fabrication protocols, months during which the measurements were made, and their results.
The authors have used two different fabrication protocols. However, the second protocol seems to have been used only during the second round of investigations beginning September 2018. The protocols differ essentially in the order in which reagents containing silver and gold are mixed and added to the gold nanosphere solution (containing gold nanoparticles 10-12 nm wide). Next, the nine samples displaying near-zero resistance include two films encapsulated in epoxy while the other seven were not encapsulated.
Finally, there are more samples in which they have observed significant drops in resistance and magnetic susceptibility, though they are not sufficient to characterise them to be superconducting.
For example, in the new samples, the authors undertook a new electrical measurement – of the inductive response – to see if the resistive and magnetic transitions in the silver-gold nanostructured composites happened at around the same critical temperature. As they write in their paper, in a superconducting transition, the onset of diamagnetism is expected to trigger a shielding current that reduces the amount of mutual inductance. Their paper describes the results of these measurements (made in May 2019) on one sample in which they have observed this concurrence with the critical temperature being 172.5 K and the drop in inductive response at around 165 K.
Simultaneous measurement of resistance (top) and inductive response (bottom) in the sample as a function of temperature. The inset in the upper panel shows the schematic lead configuration used for electrical measurements. Caption and source: arXiv:1807.08572
The doubts close in
There is no other substantive new information or detail in the revised paper. In fact, many of the diagrams and graphs are the same as in the earlier version. This implies that the authors stand by their original result and are now reinforcing their claim – of having discovered a room-temperature superconductor – with more data and improved resistance measurements.
“The authors have taken great care in adding several details; obviously they would have realised the importance of those parameters,” T. Pradeep, a nanomaterials expert at IIT Madras, told The Wire. “I would think that it is now possible to reproduce the preparation in many labs. However, nanomaterials are sensitive to the purity of the reagents and the conditions of preparation. They decide the details of the particles formed, such as their morphology, surface species, etc. To what extent these parameters determine the measured properties remains to be seen.”
He also mentioned that his group may attempt to reproduce the IISc team’s results.
According to Pushan Ayyub, a superconductivity expert at the Tata Institute of Fundamental Research (TIFR), Mumbai, “Just the large increase in the author list indicates that more of the measurements were entrusted to specialists.” “I would completely trust all the electrical transport measurements when Arindam is one of the authors.”
“Earlier it was easy to dismiss the results due to poor measurements and characterisation,” said S. Ramakrishnan, Ayyub’s colleague at TIFR. “Now we see a drop in resistivity and weak diamagnetism. One needs more investigations and measurements, like the Josephson effect, etc. to confirm this result. Perhaps the same team will do these.”
Also read: Under a Microscope, a Bizarre Electronic Dance Comes Into Better Focus
It seems the new results are more credible than the older ones, but complaints and some doubts remain. Perhaps most important among them is the ‘repeat background noise’ in independent experiments that Skinner had pointed out. It had triggered widespread doubt on the data presented in the older paper. Since then, the wider research community’s pall of suspicion over the Pandey-Thapa claim hasn’t gone away.
In fact, the team seems to have reproduced the same controversial magnetic measurement dataset (taken with a pellet sample prepared in November 2017) in the present paper.
The authors acknowledge its presence and note in their paper: “[W]e also observe a strong repeatable noise in [magnetic] susceptibility … The origin of this unique ‘noise’ remains uncertain…”. They discuss the issue in the ‘supplementary information’ part of the paper but haven’t been able to explain the repetition.
Next, a closer look at the susceptibility plot suggests they have simply carried over the data from the earlier paper but plotted it with different symbols, in different colours and with small vertical and horizontal offsets. The noise patterns also look the same, as Thouis (Ray) Jones – a computational biologist at the Broad Institute, Massachusetts – has pointed out on Twitter.
Still with weird noise offsets, sometimes vertically (two green plots), sometimes shifted (yellow and black). pic.twitter.com/9EAwvtzmuZ
— Thouis (Ray) Jones (@thouis) May 22, 2019
In fact, there is no indication in the new paper that the researchers have undertaken any new magnetic measurements on the gold-silver sample that had a critical temperature of of 236 K and displayed diamagnetic transition over a range of magnetic fields at temperatures around that number. In the same vein, the value of the diamagnetic volume fraction (of -0.056) is also identical to that in the previous paper.
The strange susceptibility data
Further, the researchers have studied the material’s magnetic susceptibility only under zero field cooled (ZFC) conditions, which yield an upper limit on the fraction of the sample that is superconducting. They would have had to examine it under field-cooled conditions as well to determine the lower bound. And the susceptibility measured in ZFC conditions implies that a maximum of just 6% of the sample volume was superconducting.
“I am very gratified to see this response [by the authors],” Skinner told The Wire. “It looks very thorough, and they have provided many of the additional details that people had been asking for. I’m sure that a number of people … will wade through all the data and protocols very carefully, and look for places where false signals might appear.”
However, he found it “strange” that the team had presented the same magnetic susceptibility data, at least at first sight. “I wonder why they didn’t just take new data, since this particular data set aroused so much suspicion. Were they unable to reproduce the measurement?”
The variation of volume magnetic susceptibility of pellet P11117PE0_01 with temperature at different external magnetic fields (B). Caption and source: arXiv:1807.08572v2
The contested noise data. Source: arXiv:1807.08572
Skinner continued:
One interesting thing to notice is that, now that they have plotted the same data again (using different symbols), you can see that all five curves are really the same and not just the bottom three. Some are shifted vertically and some have a horizontal shift also. … If I were reviewing the paper, I would want to have a high-level of additional evidence that everything is in good shape. But mostly I am happy that the authors have made a response, and we can now move forward with evaluating and discussing the results in a scientific way.
Pratap Raychaudhuri, another physicist at TIFR, had doubts about the inductive response data. He said in a comment on Facebook:
In my view, this is the most important data in the revised manuscript. The diamagnetic shielding response is a very precise tool to establish macroscopic phase coherence. But looking at the data, I am unsure. When the resistance of the film is zero, the shielding … currents should set up simultaneously, and the real part of the mutual inductance should show a rapid large drop. Here, [it] does not drop well below this temperature [of 165 K] and even when it does, the drop is tiny, smaller than I have seen in any superconductor.
Extraordinary claims do require extraordinary evidence, but if that isn’t in the offing, then the doubts raised are bound to be extraordinary as well. This time, however, with sufficient details available to go with the revised and renewed claim, scientists around the world can inspect the IISc data more seriously. And as laboratories around the world try and reproduce the results, real academic discussion can begin around this most celebrated quest of materials science.
R. Ramachandran is a science writer.
