Possible Room Temperature Superconductor Achieved

TechkNighT_1337 sends news that surfaced on the Next Big Future blog, concerning research out of the University of Bengal, in India. The report is of a possible superconducting effect at ambient room temperatures. Here is the paper on the ArXiv. (Note that this research has not been peer-reviewed or published yet.) "We report the observation of an exceptionally large room-temperature electrical conductivity in silver and aluminum layers deposited on a lead zirconate titanate (PZT) substrate. The surface resistance of the silver-coated samples also shows a sharp change near 313 K. The results are strongly suggestive of a superconductive interfacial layer, and have been interpreted in the framework of Bose-Einstein condensation of bipolarons as the suggested mechanism for high-temperature superconductivity in cuprates. ... The fact that the results described above have been obtained from very simply-fabricated systems, without the use of any sophisticated set-up and any special attention being given to crystal purity, atomic perfection, lattice matching, etc. suggests that the physical process is a universal one, involving only an interface between a metal and an insulator with a large low-frequency dielectric constant. We note in passing that PZT and the cuprates have similar (perovskite or perovskite-based) crystal structures. This resemblance may provide an added insight into the basic mechanism of high-temperature superconductivity."

Comments

Of course!

[Anonymous Coward]

After reading the summary, everything is plainly obvious...

(walks away slowly before anyone can notice I didn't understand anything)

Re:Of course!

[thrawn_aj]

*sigh* It's even worse than that. IAAP and I was very excited to see this ... at first. The article by the way is very well written (serious science - not a crank). The problem is that the data (figure 2 in the arxiv paper - everyone should check this out btw) on which the author hangs all his hopes is seriously noisy (compared to the size of the "kink" that he superposes on the graph). In other words, if you imagine erasing the drawn-in kink, such artifacts occur several places in the data and are generally not above the noise level.

So, I would say that the conclusion is highly unwarranted given the state of the existing signal to noise. However, if the author truly feels there's something promising, he simply has to go about improving his signal. To be fair, the /. title is far more ambitious compared to the original article (indications of ...). He's merely putting this out in the wild to get feedback from other researchers in the field (which is solely what Arxiv is used for by serious researchers, not as a publication destination).

As it stands, the "kink" seems to be nothing more than (one of several) noise bumps. I'll be keeping an eye on this guy of course. Maybe something might come out of this, who know?

Re:Of course!

[getuid()]

*sigh* It's even worse than that. IAAP and I was very excited to see this ... at first. The article by the way is very well written (serious science - not a crank). The problem is that the data (figure 2 in the arxiv paper - everyone should check this out btw) on which the author hangs all his hopes is seriously noisy (compared to the size of the "kink" that he superposes on the graph). In other words, if you imagine erasing the drawn-in kink, such artifacts occur several places in the data and are generally not above the noise level.

Not necessarily. When analysing experimental data, keep in mind that it's not only the ~5 points of the kink that carry relevant information, it's *all* the points! Thus, the proper way to look at the graph would be to focus first the lower half (up to the kink), and then on the upper half, and see what's changed. If, for example, linear fits to the separate data regions give separate straight lines, this could mean that there is something in the data.

That having been said: although IAAS (I am a scientist), I'm not a transport measurements guy and I'm not familiar with the state-of-the-art methods in this particular experimental technique... The guys improving their experimental technique would certainly not hurt at all, but for now, I'd leave it to the peer reviewers to estimate the relevance of *this* particular graph ;-)

Re:Of course!

[RelativeKny]

IAAP too, although in experimental LOW-Tc superconductivity. I agree, although with even more reservations. Not only is the data noisy, but the author claims, that microvolts is "extremely low voltage". This is absolutely rubbish from an experimentalist point of view. It has (obviously from the plots) not been filtered, and the authors claim of a "homebuilt amplifier built on an AD620" is not confidence inspiring. Although the AD620 is not horrible, noise-wise it is used at DC (20 Hz is close enough), which means that 1/f noise will kill his signal no matter what he does. This is especially silly considering, that nothing in the setup should require this low frequency sweeps. Hook up a lock-in amplifier and run ia at a few kHz at least to get decent noise characteristics. Also, all the experimental details of the setup are missing - this alone will get the paper rejected from any peer-reviewed journal. The author might be on to something interesting - superconductivity or not - but the experiment is done like a theorist would, not like an experimental physicist would. Back to the lab, and get some better results - I would love to see this with better measurements. PS: I'm not trying to re-ignite ye olde theorist vs. experimentalist battle, btw - I believe it is possible to be both. Just saying, that the author of this article is not. Seems he has a decent grasp of the theory though, hence my comment to theorists ;-)

Re:Of course!

[arachnoprobe]

I think he knows that his experimental data is crap. The note on the dirtiness of the procedures in the abstracts hints to the fact, that he put out one sample and accidentally found what could be something hyper-interesting. Out of fear of being out-published by someone else, he put out this paper, that - if this is an RT superconductor - he can (rightly) claim having discovered it (leading to wealth and nobel price). Now he can go back an do some proper experiments.

Wait until it has been repeated.

[BLToday]

until the experiment has been repeated by someone else, I'm not holding any hope.

Re:Wait until it has been repeated.

[vbraga]

Maybe we could just redefine what room temperature is!

...really?

[linuxgeek64]

Not peer-reviewed and not published = why the fuck is this on Slashdot?!

Re:...really?

[nashv]

Because in physics, people have the good sense to let the larger community take a look before these bureaucratic procedures are finished. That is why ArXiv exists,and if Slashdot does its bit, why the hell not?

Cold Fusion

[Fartypants]

This smells of Cold fusion. I was 12 when that scandal erupted and I'm *still* recovering from the disappointment that we hadn't just entered the age of flying cars. This time I think we're better off saving our excitement until the experiment has been repeated.

Re:Cold Fusion

[Idarubicin]

Ah. Well-renowned scientific co-worker endorses cold fusion. Inside scoop from globally-acclaimed Slashdot science critic hAckzOr.

"It's real", hAckzor concludes.

Film at eleven. Take that, scientific establishment.

Re:Cold Fusion

[Anonymous Coward]

Cold Fusion has more than just political problems: it's a matter of energy scales. To overcome the Coulomb barrier between deuterium and tritium, after which the strong interaction takes over, requires an energy of about 4.5 x 10^7 Kelvin. This is the lowest energy fusion reaction. Now imagine how much energy you can get from the strongest chemical reaction. How about thermite: 2500 K, you're still off by four orders of magnitude. That's the main reason why physicists avoid Cold Fusion.

The current superconductivity article is better, in that the underlying physics is at least plausible, but as a previous poster pointed out, the signal to noise ratio is low, even after smoothing has been applied. Also 4000 Angstroms of deposited Al seems to be somewhat on the thick side for the dielectric to have any effect. But it's certainly worth trying to reproduce the results. (IAAP specializing in superconductivity).

The catches

[BlueParrot]

There has been a number of fraud reports of high temperature superconductivity, and while there are some confirmed examples of superconductivity at very high temperatures ( like -70C ) they usually involve some microscopic crystal or other structure which is not very useful for most practical applications.

In addition, that something super conducts does not imply it can handle a very large current at high temperatures. The current creates a magnetic field, and superconductors can only work when the magnetic field is less than some fixed value that depends on the material. If I'm not mistaken this value is at its highest when the temperature is very low, and thus it's quite plausible you could get a room temperature superconductor which can't carry any significant current unless cooled to more traditional temperatures.

Re:The catches

[bertok]

Yes, but the small crystals are usually a side-effect of the technique used to find novel superconducting compounds. What some groups do is create polycrystalline lumps where each crystal has a slightly different formula. Then they test resistivity with changing temperature across the whole lot. If just one crystal superconducts, there will be a 'kink' in the graph. This is like a simple brute-force method for testing many samples in parallel, but doesn't necessarily provide a formula that an be produced in bulk.

It's like a mathematical proof that states that something "must exist" without providing an actual value.

Also, superconductors are inherently useful irrespective of the current carrying capacity. For example, Josephson Junctions and RSFQ digital electronics are both very useful and require very low power.

Even a "thin-film" superconductor like the one described in the article would be very useful, as that can be practical for integrated circuitry, even if it's not possible to make a flexible wire out of it.

Re:Room Temperature in UK, maybe not in India?

[JamesP]

Reminds me of that joke about scientists in Anchorage discovering a room-temperature superconductor :P

Re:Someone didn't get the memo

[thestudio_bob]

...not something made of silver.

Well, apparently you don't have to deal with electricity stealing Werewolves. I for one, am glad someone is finally addressing this problem.

Re:This will later be known as...

[epine]

The word (and concept) of "unobtainium" goes back to the 50s at least, actually.

If the term "unobtainium" wasn't invented by the early heyday of jet fighter engineering (circa the Korean war), I'll eat my carbon-graphite bike frame.

My understanding is that superconductors have current limits independent of resistive effects (possibly due to magnetic field intensity). How much material you need depends on those exact limits. Even silver could be cheap as dirt if the current density is high enough.

The other thing I've heard is that superconductors are generally discovered by observing related effects, not by measuring conductivity itself.

There also seems to be many people here who have never heard of the black swan effect. You can't prove a black swan doesn't exist by observing a sequence of white swans. There's always a first time. This also applies to the possibility that something important is someday discovered or first published independent of peer review.

That said, there's no point in wearing out your salivary glands unnecessarily, although I've heard it's a common ailment to overdose on visual innuendo of the possibility of doing something you're not actually doing (with dim prospects).

For me qualified engineering porn is when the material is officially characterized in important criteria such as current density limits.

I feel the same way about quantum computing. Still haven't seen a formula which describes the ultimate constraint (or cost) on how many qubits can be stacked together (usually the universe puts limits on salivary endeavours). It would be kind of weird if qubits prove to be as stackable as frictionless pulleys.

Extraordinary claims require evidence.

[dr. loser]

I'm a condensed matter physicist. This claim is weak beyond belief, and it pains me to no end to see it get picked up by slashdot and other sites (nextbigfuture.com). To demonstrate superconductivity, you need to show (a) zero resistance over some range of current; (b) the Meissner effect (expulsion of magnetic flux, seen via magnetometry); (c) a characteristic feature of a phase transition in the heat capacity. This paper shows exactly none of these things. The noise level in the resistance measurements is so poor, you could not tell the difference between zero and 0.01 Ohms (which would be totally believable considering there is already a metal film in the system). This paper in its present form is not fit for publication. Seriously, you don't have to be an expert at this stuff to see that this is weak - just look at the noise level in the current-voltage curves and use some common sense!