With the current fabrication process, they're only getting a chunk with LK-99 particles sprinkled in. Since nobody yet knows how to fabricate a pure chunk of LK-99, it'll be hard to measure the true resistance of LK-99.
>Most likely, the material LK99 as reported in [1-3] is a heterophase structure, with co-existent non-superconducting constituents. This may yield superconducting droplets surrounded by nonsuperconducting material...
>In fact, I find that Cu on this Pb(2) is 1.08 eV more energetically favorable than Cu on the Pb(1) site, suggesting possible difficulties in robustly obtaining Cu substituted on the Pb(1) site.
The paper from Dr. Griffin at LBNL suggests copper atoms have to be placed in a specific (but less likely) position in the molecule to result in the desired flat band characteristic. Also, the original authors and the labs who were able to replicate LK-99 are reporting they had to make multiple batches to even find a tiny piece that shows levitation. This suggests that you just have to be very lucky to produce a sample with high enough concentration of LK-99 to observe levitation.
If we can somehow confirm that LK-99 is truly a room temperature superconductor, billions of dollars of R&D fund will pour in to improve the fabrication process. When the first transistor was invented, people probably weren't imagining that we'll be mass producing them in nanometer scale in the future. Or maybe LK-99 will be stuck in a lab like graphene. Who knows?
From a purely scientific angle, so presumably less amenable to mass production, why doesn't anyone use FT-ICR mass spectrometers to assemble say a unit cell?
There is a good classic primer on FT-ICR, mostly focused on analysis (mass spectrometry) but also mentioning activation energies for reactions and measurement of kinetics etc.
If you dunk a bunch of chemicals (for simplicity think wet chemistry) in a vial, all reactions and side reactions are simultaneously occuring, so one has little control over what happens on an atomic scale.
FT-ICR can be used to observe the state AND to manipulate the state. Its like having a compact particle collider, but instead of the high (TeV) energy in CERN etc. its just chemical energy levels.
It happens in high vacuum, so low densities of species, hence not amenable to mass production.
But the instrument is both eyes and hands: one can identify the frequencies corresponding to each ionized molecule, and selectively energize or de-energize specific species to encourage or prevent main and competing reactions, by pumping or damping specific frequencies.
One may build up a molecule in elementary steps and eject finished molecules. Those steps can occur at the same time in the same vessel. Its like having a miniature digitally controlled chemical plant, without having to redo all the pipework if you decide to use a different pathway here or there.
The good news is that we should know within a few weeks if rock surgery is effective. If that works they can increase the purity and sample size without needing to find a new creation process
There are a ton of suggestions floating around already on how the efficiency could be improved and some of those are very clever and involve using the superconducting properties of the individual grains to help sort them out from the bulk material.
How do you explain the almost three orders of magnitude drop in resistance in Fig. 6d in one of the original articles ( https://arxiv.org/pdf/2307.12037.pdf ) with a few LK-99 particles sprinkled here an there? There must be a current path along which more than 99.8% of the material is in the supposed superconducting state. So the particles almost touch but not quite yet?
I think the more likely explanation is that the particles do touch each other but the interface is not superconductive. In other words, it is a polycrystalline material, and most of it is LK-99, but the grain boundaries are not a very good conductor. In conventional superconductors grain boundaries don't disrupt superconductivity because they are 3D superconductors, but in this allegedly 1D superconductor the superconducting channels in most cases don't meet at the grain boundaries, so the current has to overcome the resistance of some material that is almost an insulator.
If that is the case it will be difficult to produce a material that is macroscopically superconducting. But I hope researchers will be able to make single crystals that are large enough for resistance measurements so that finally it can be determined if this material is a superconductor or not. For practical uses the best result that can be achieved with this material may be a metal-LK-99 composite where the LK-99 particles lower the resistivity of the metal by 50-90%.
I believe that with the synthesis method proposed by the Korean team, i.e. by the chemical reaction between a certain kind of lead sulfate with copper phosphide, the chances of progress are slim.
The Korean team appears to have been stuck for several years by the lack of reproducibility of this synthesis method. While it was a great discovery that has shown that this material must have some very interesting properties, perhaps even superconductivity at ambient temperature and pressure, in order to be able to measure its properties and be able to evaluate the possible practical applications, a much more precise method for enforcing the desired crystal structure is required, than mixing powders and baking them into a ceramic.
Perhaps such a method for producing samples with deterministic properties would be to develop first a method to grow monocrystals of the special kind of lead phosphate that forms the base crystal structure, maybe by drawing the crystals from melt.
Once monocrystals of this kind of lead phosphate are available, they could be doped with copper, e.g. by ion implantation. By controlling and varying the parameters of the process, e.g. the angle of incidence and the velocity of the ions and the thermal profile used for annealing, it is likely that reproducible samples can be produced, where the copper ions substitute lead in the useful places and not in the others.
By this method it would be possible to produce only thin layers of LK-99, but that should be enough to enable the characterization of the material.
Moreover, because LK-99 is very fragile, it is unlikely that it could be used to make cables or coils. Practical uses where LK-99 would be deposited as thin films are much more likely.
As an alternative to ion implantation, which might be able to produce thicker layers, perhaps once monocrystals of the base lead phosphate are available it may be possible to develop some method of chemical vapor deposition, to grow epitaxially a layer of LK-99 over the base crystal, but with such a method it is less obvious if there is any way to control which lead atoms are substituted, though this may depend on the orientation of the base crystal.
Drawing from melt to achieve a consistent crystal orientation was my first thought as well.
Looking at the Griffin paper, the Pb(2) site is described as being 1.08eV 'more energetically favorable'. I am having trouble understanding what this means.
Years back I did MOCVD semiconductor fabrication research, I never reached a mastery of it but I am still trying to leverage that understanding here.
During growth, adatoms that incorporate into proper crystal lattice locations enter a lower energy state compared to those in imperfect locations. The energy state is lower in the sense that it requires more energy to remove them from that location. Hence careful control of temperature allows you to selectively favor incorporation into these low energy locations e.g. choose a temperature high enough to remove adatoms from 'imperfect' locations but low enough to not remove them from 'perfect' (low energy) locations.
So when the author says 'energetically favorable' am I to understand this means the Pb(2) location represents a lower energy state (i.e. more difficult to remove Cu from this location) or the opposite? Or something else entirely?
