Sunday, May 30, 2021

Ask me something.

 I realized today that I had not had an open "Ask me something" post since December, 2018.  Seems like it's time - please have at it.

16 comments:

Anonymous said...

Thanks Professor for this opportunity.

Today, we have with us more than 1000s of 2dimnensional materials. It is like a vastly unexplored jungle. What are the open questions and new challenges that you foresee for these materials. I know that scalability to industrial scale is a big obstacle but what issues do you think are important as regards the optic, electronics and magnetic properties of these materials.

Anonymous said...

Twistronics, hype or not?

On a more nuanced note, I really think the experimental verification of interesting physics, such as superconductivity, as a function of twist angle in bilayer graphene is both unexpected and just very cool! However, does that justify a creation of a new -tronics field? Will it go the same way as "spintronics"? Can we expect an expansion of the field of twistronics beyond bilayer graphene?

Anonymous said...

What exactly has changed in the past few decades in condensed matter/solid-state physics that has pushed Nature and Science journals as the gold standard instead of journals like Phys. Rev. Letters? If I look at the important historical papers in physics, they seem to all have come from Phys. Rev., not Nature or Science except until very recently. Most other fields of physics do not hold Nature/Science type journals in high regard, unlike in condensed matter

What changed historically to make Nature and Science the "currency" of scientific productivity in condensed matter?

What can be done to reduce the "Buzzfeedification" of condensed matter and Nature/Science type hyped-up publications? This would be a valuable lesson for growing new fields like quantum information.

Anonymous said...

What can be done to stop the publish or perish culture that leads to not only noisy science but also puts pressure on scientists/profs and also on the newcomers as they need to show that they are capable to bring in money via projects.

Also, there are 1000s of articles in fields like plasmonics, photonics where almost every paper claims that the fabricated nanostructure is the so-called "best one" for so and so application. I really do not understand why every group believes that their sample or prototype device is the best one and yet the research stays limited to the academics with very few actually making it to the industry for mass production. What can be done to avoid unnecessary claims?

Douglas Natelson said...

Anon@2:58 - I think the biggest open physics issues in this class of materials involve understanding the variety of competing states (beyond basic semiconductors) even in the monolayer limit; the combo of reduced dimensionality, valley effects, spin-orbit coupling, frustration in some lattices, topology, etc. makes these systems rich. Adding in tunable interlayer couplings and stacking of heterostructures and you have a rich vein to mine. As has been seen in the twisted multilayer experiments, there is clearly a lot of work remaining to be done to understand the role of correlations and the origins of superconductivity in some of these systems. On the technological side, growth of high quality material remains extremely difficult for the vast majority of these systems, and exfoliation isn't going to cut it if one really wants a technology. Understanding point defects and their properties is also potentially interesting, in the same way that color centers in diamond and SiC have taken on new significance recently as atom-like objects for quantum science applications.

Anon@4:09, as I mentioned above, while I'm as cynical of "tronics" marketing as anyone, there is no doubt that there is a huge space of basic science to learn about in moiré systems (twisted or from heterostacking). Yes, I do think that there will be a lot of work beyond just bilayer (or more recently trilayer) graphene. However, the expansion to other materials (e.g. MoSe2 and other TMDs, for example) really is going to be helped by improved material quality. Just from the graphene work it's already clear that lots of delicate states can be obscured due to disorder, and graphene is still clearly the leader in material quality for van der Waals systems. BTW, I wouldn't damn spintronics with too much faint praise yet - the work that's been done and is being done in spin Hall physics is seeing applications in data storage. Sure, the Datta-Das spin transistor hasn't happened, and the c. 2003 hype was overblown, but that doesn't mean there aren't interesting paths forward.

Douglas Natelson said...

Anon@7:56, good question. Somehow Science and Nature have done an incredible job of marketing themselves, while PRL has a reputation (we can debate how deserved or undeserved) for being slow and having a cantankerous and distressingly random refereeing process (not from the editors, but from the referees). In practice, Science and Nature and their progeny are just as random in their refereeing and are not necessarily fast, but somehow they have become ever more highly pursued. I think you've identified a media role - papers in the glossy journals are supposedly constructed to be of broad interest, and those are the ones whose press releases get picked up. One thing that would help would be to encourage the revival of real science journalism - far too much completely credulous repeating of university press release claims out there. There are excellent science journalists, but budget reductions at many outlets (including major newspapers and all the TV news networks) have exacerbated the tendency to sensationalism.

Anon@8:52, I wish I had a great answer for you. As long as funding is competitive and people need some quantifiable (albeit crudely) way to demonstrate that they are making progress, there is pressure to publish. Most program managers will agree that one solid paper per supported student per year is good output, but they will also be elated at much higher publication rates. Regarding plasmonics and applications, I do think there will be more transitions from the lab to industry, but there are some reasons why progress is slow. For example, a lot of plasmonics work is done with gold and silver, and large-scale semiconductor fabs do not like to work with those materials (and don't like additive lift-off processing either). TiN is an interesting plasmonics material in the near-IR that is compatible with CMOS manufacturing, but it's not trivial for a lot of academics to work with. Remember, it took about 20 years between the batch synthesis of really high quality II-VI semiconductor nanocrystals and the sale of "quantum dot display" TVs. In terms of claims, this gets back to the perceived need to argue that new milestones have been achieved. Greater frankness ("this structure teaches us something important, but will require considerable engineering to be technologically applied") should be rewarded.

Pizza Perusing Physicist said...

I don't have a question per se, but I just wanted to thank you, Doug, for all your advice and support via this blog throughout the years. Since I last commented on one of your posts, I have obtained a faculty position and have had a great start, submitting several grant applications and forming great network connections with colleagues in my department despite COVID-19. I wanted you to know that your wisdom as shared through this blog played no small part in my success. Thank you.

Anonymous said...

Thanks professor,
What do you think about the quantum computing trend, especially with the quantum supremacy claims. Is it just hype of Quantum Computers will actually be made. I have heard a lecture from Prof. Gil Kalai and he seemed convinced it is not possible

Anonymous said...

How would you recommend someone with an experimental background in (hard) condensed matter (Phd level in cuprates) in industry getting their foot in the door at these quantum information companies? I would love to leave my tech job at Facebook (well paying to be fair) to do some real quantum physics again...

Anonymous said...

What's the timeline for high Tc high field magnets looking like? It always feels like they are just 5-10 years away and ready to revolutionize nuclear fusion reactor designs.

Douglas Natelson said...

PPP, thanks very much for the kind words - they mean a lot. Congratulations on the faculty position, and best of luck!

Anon@7:20, I think quantum computers will actually be made. I'm not sure what the best approach is in the short term, but I don't know of any deep fundamental reason why the whole notion is wrong-headed. As for Prof. Kilai, my reading of his argument is that it claims that with noisy intermediate-scale quantum systems, it is harder to implement robust error correction than it is to demonstrate large-scale quantum advantage. By that argument (which I can't evaluate with my knowledge of the field), it would not be possible for, e.g., google's superconducting qubit approach (with surface code error correction) to scale to a really useful level. I guess I wonder: if someone had qubits and gates with really good fidelities, would they evade his concerns by not really being NISQ systems anymore?

Anon@9:24, I have no special knowledge here. My sense is that if you know how to measure Josephson junctions and have low-T technical skills, then google, microsoft, and intel may all potentially be interested. There are likely people that you know directly or indirectly at some of the QIS companies. You may want to try seeing if you have any networking contacts.

Anon@3:59, they're getting there. The NHMFL has one, for example: https://nationalmaglab.org/magnet-development/magnet-science-technology/magnet-projects/32-tesla-scm
and here is a UK spheromak company making high-Tc magnets: https://www.youtube.com/watch?v=VHl02g5DUAU
and MIT is pushing the same thing. Fusion goals aside, it sounds like the technical aspects of the magnets are getting under control.




Anonymous said...

Dear Prof. Natelseon, A quicky :-) Where would you recommend a young person to work nowadays? Is there anywhere left with the free rolling spirit of e.g. Bell labs etc? It seems that the universities and national labs are now dominated by the 'normal' types who thrive on management and power structures, which in my view retard scientific progress.

Anonymous said...

Anon@9:24, for Intel at least, the quantum information work takes place within research labs which are picky wrt to hiring. Depending on how keen you are to leave facebook, you can join Intel as a process engineer and then try to internally transfer. However, process engineering is a demanding career.

Anonymous said...

Anon@2:21, if possible, can you please elaborate on the job profile of a process engineer. I may have a job offer in the future as one.

Anonymous said...

What is the current status of our understanding of high-Tc superconductors?

Douglas Natelson said...

Anon@2:59, good question, and perhaps some readers have strong opinions on this. In terms of large companies, it’s very hard to have really free wheeling research these days because of the trend away from the Bell/IBM old school approach. I do know that in the quantum arena there is a lot of good work done at Northrop Grumman, Hughes Research, Honeywell, and others. National labs do have some excellent groups, though management structures can lead to lack of fire-in-the-belly. NIST and other places like Lincoln Labs seem to be doing particularly well. There are start-ups, of course, but that’s very different….

Anon@7:54, continually evolving. The overly short answer is, there is still not great consensus on the mechanism, and there are many competing ordered states at play in the phase diagram. The normal state itself remains complicated, even in the overdoped limit where for years naively it had been assumed that Fermi liquid response eventually takes over. There is also a lot of other excitement about superconductors these days - twisted vdW materials, high pressure hydrides, triplet sc in UTe2, etc.