Thursday, December 27, 2018

Ask me something.

As we approach the end of another year, I realize two things:

  • Being chair has a measurable impact on my blogging frequency - it's dropped off appreciably since summer 2016, though fluctuations are not small. 
  • It's been almost 2.5 years since I did an "Ask me something" post, so please have at it.


Anonymous said...

What advice would you give to junior faculty in their first few years of the job, especially junior faculty who were hired in by a very friendly chair and will be going for tenure under a hostile chair?

Douglas Natelson said...

Hi anon. Tricky to answer in detail without more information. Hostile to you personally for some reason? To your research area?

Ideally, the processes in place at your university and in your department should make the personal attitude of the chair largely irrelevant, but I’m enough of a realist to know that this is not true everywhere. It goes almost without saying that it’s best to have good research results, external funding, solid teaching, and enough of a profile in the community that external letter writers will know who you are. Hopefully your department has a faculty mentoring program that involves senior faculty other than the chair, and those are people that you can go to for advice.

At my institution, the chair comes into the tenure process in a couple of ways. First, the chair does annual one on one reviews w the junior faculty members, though the resulting one page memo to the dean is seen by and signed by the junior person, who has an opportunity to respond or offer edits. Second, the chair works with a senior faculty member in the junior person’s research area to come up with some of the external letter writers, with the junior person supplying others; both lists go to the dean and are combined to make a final list. Third, when the time comes for the departmental discussion and vote, the chair runs the tenured faculty meeting, and then prepares a chair memo for the dean and the promotion and tenure committee that summarizes the case and describes the meeting discussion and vote.

Having a senior colleague (perhaps the previous chair) who can keep an eye on things and be an advocate is probably the best.

Anonymous said...

What would you say have been the most practical results of strongly correlated materials in daily life for non-physicists in the past three decades? Off-hand I can think of solid state lithium ion batteries and some other oxides for semiconductor industry uses, but wanted to know what you had in mind.

Douglas Natelson said...

Anon@8:42, good question. I think it's fair to say that broad applications of correlated materials that genuinely depend on their correlated electronic properties are very limited so far. The cobalt oxides used in Li-ion batteries are correlated and widespread, but their use in energy storage is related to their structure related lithiation properties, not strong correlations. High temperature superconducting (HTS) wire has achieved limited use that touches on broader populations - magnet leads for MRI and NMR systems; some cryogen free MRI systems are starting to appear (e.g., here). HTS products have so far been held back from widespread adoption in, e.g., MRIs and turbines, by the usual combination of engineering and economic challenges (scaling up reliable manufacturing to industrial scales at a price point where the additional infrastructure makes economic sense). One broad area that is less obvious: electrochromic and thermochromic coatings, like this. These are metal oxides, usually some combination of tungsten and vanadium, that are switched either thermally or through ion transport and intercalation across a metal-insulator transition; in the more conductive state, the materials better reflect/absorb visible+IR.

The space of correlated materials remains vast, and these systems routinely feature competing ordered states with energy scales that can be room temperature or above. The chromic applications are one example of the hoped-for idea of responsive materials, where we can take advantage of the competition between states with vastly different properties to create devices that can be switched or tipped one way or the other by some stimulus. It is worth noting, though, that these materials tend to be compositionally complicated. We've been spoiled rotten by silicon, III-V semiconductors, and plastics in terms of comparative ease of industrial-scale deployment.

Ali Ibn Rafi said...

Great to see an open Q&A!

I've often heard the joke that condensed matter is like the parable of the blind men and the elephant. I often feel this is partially due to a weakness in the kinds of observables we can measure. In your opinion, what kinds of experimental tools to measure new observables are we missing?

For example, I feel that real-space versions of scattering techniques like Neutron scattering and ARPES would vastly help the field, especially because the momentum space picture is limited as soon as you move away from ideal crystals.

Douglas Natelson said...

Hi Ali, great question, though I have to be careful not to descend into pure wish fulfillment when thinking about possible answers.

In terms of observables/things we'd like to be able to measure, two jump to mind, though I may draw criticism for my choices and descriptions. The first is entanglement. It would be great if there were some straightforward experimental procedure that could be done that would allow some quantitative measure of the entanglement within the electrons in a system. Even the ordinary Fermi gas is highly entangled (since the many-body ground state can be written as a totally antisymmetric linear combination of Slater determinants). As many-body effects become important and the relevant effective degrees of freedom change (say, the emergence of local moments in the Hubbard problem), presumably this could be seen in some measure of the overall entanglement. (In some ways this feels a bit like entropy. There's no nice way to measure entropy directly, though at least with heat capacity measurements over a range of temperatures it is possible to quantify changes in entropy.)

The second is some measure of topology. Right now, how can you tell, experimentally, if a system you have possesses some interesting electronic topological character? You can measure dispersion relations of electronic excitations with techniques like ARPES and compare to calculations (hunt for Dirac cones, say, or Weyl nodes). While that has produced a large number of Science and Nature papers, that approach works best for weakly interacting systems, ARPES (and quasiparticle interference in STM) is a surface technique, and resolution is always a concern. Alternately, you can try to perform non-local transport measurements of various kinds, but device fabrication on such materials is very nontrivial, and the distance scales over which topological protection survives so far are not macroscopic. Some global test for topological character would be excellent.

(I'll write more later today.)

Douglas Natelson said...

Following on.... You mention real-space versions of ARPES and neutron scattering. I'm not 100% sure what you mean; if the desire is to learn about dispersion relations (energy v (crystal) momentum) of excitations, then purely local measurements are not going to get you there. Now, ARPES and neutrons have other particular limitations. ARPES is often energy resolution-limited at a level that can't see really fine low-energy scales; it's inherently a surface technique, requiring UHV interfacial preparation; and in the end you're always pulling a whole electron out into the vacuum (so subtle many-body dynamics require serious interpretation). Neutrons give you spin information but small cross-sections mean they're not useful for micro/nanoscale samples (e.g. atomically thin materials) and can't readily be used for all elements.

Some ultrafancy version of EELS (e.g., with big advancements in energy resolution) could be very appealing. Atomic-scale spatial resolution via TEM technology, perhaps combined with some flavor of electron holography/tomography/ptychography, could give both real-space and energy/momentum information. Spin-polarized electron sources might give spin information as well, though the limitation of TEM-like sample thickness is severe, and you can't do such measurements at low temperatures and in significant magnetic fields. Fun to speculate, though.

Grumpy said...

1. Is the widespread study of topological effects in solids a fad? If funders realize they have virtually no relevance to quantum computing, will CM physicists move on? If so, to what?

2. what do you consider to be a quantum material? It seems the definition in the past dealt primarily with correlated many-body systems. Now ppl start mixing in various defect/qubit systems under the same umbrella. Both types of systems obey quantum mechanics but they seem very different to me... what definition do you use?

3. What is your take on metasurface optics? Fascinating field of physics inquiry or rehash of basic engineering principles?

Love the blog, sorry for taking advantage and asking 3 questions.

Douglas Natelson said...

Hi Grumpy. Good questions.

1) While CMP certainly has fad tendencies (quick everyone, drop what you're doing and all start working on graphene!), I think topological effects in solids has more staying power than that. It's been less than 15 years since the widespread beginning of the broader appreciation that topology is important in electronic CM systems beyond the quantum Hall effect. People are really just getting started on the experimental exploration of the number and variety of materials systems of interest. There are also still big questions about strongly interacting topological systems, such as topologically nontrivial Mott insulators, for example. I also think it's premature to argue that there is virtually no relevance to quantum computing. While I think the Majorana approaches are long shots (much farther from implementation than other platforms right now), topological encoding of quantum information is an idea that is not going to go away. In terms of the Next Big Thing for condensed matter, whatever it is will probably be a surprise.

2) I've joked in the past that quantum materials are like obscenity - hard to define, but you know it when you see it. One rough, rather restrictive definition would be materials in which the dynamical degrees of freedom are highly quantum mechanical (e.g., entanglement, interactions, + quantum fluctuations are important). The vernacular definition is evolving to be much more broad, for example encompassing materials with quantum degrees of freedom that could be useful for quantum information processing, sensing, or communications (e.g., some people argue that SiC hosting color centers is a quantum material, since those color centers can be used for quantum sensing and communication). Seems to me those are rather different definitions.

3) Metasurface optics is quite neat. Like metamaterials, I'm not sure that there are big open physical questions associated with metasurfaces themselves as the underlying physics is well known, but certainly they might open up new regimes for other experiments (strong coupling quantum optics or optomechanics, say). Potential applications are impressive. Sure, one could claim that they're basically a rehash of phased antenna array physics, and that the basic physics is just Maxwell's equations in the presence of media, but that is an unfair characterization - it'd be like saying all of hard condensed matter is just statistical mechanics + quantum.

Anonymous said...

The standard undergraduate and graduate curriculum in physics seems to be fairly set in stone for many decades now. Although there are some slight variations from institute to institute, in general, undergrads start with introductory physics course covering mechanics and E&M, then a modern physics course, then some upper division courses in theo mech, E&M, stat mech, and quantum. Those four courses are then taken at an even higher level in the first years of graduate school, and are tested on a graduate qualifying exam.

One could argue that there is a good reason for this, since these foundations work very well and are the basis for everything that professional physicists do. At the same time, however, the current pedagogy leaves many students with the impression that physics is a finished, or nearly finished, subject, and that there is not much more left to learn that will fundamentally alter the standard paradigms of our field. This is in contrast to, say, biology, where the textbooks are being rewritten on a yearly basis as the state-of-the-art in understanding continues to evolve.

My question is, what ideas do you have for how to change the physics curriculum so that students walk out understanding that while physics is built a very solid and robust foundation, it is at the same time a living, breathing science with room for innovation and creativity?

Anonymous said...

What are your personal opinions on the importance of diversity for the physics community? Our field has a reputation, fair or not, for being unfriendly to women and ethnic minorities, among others. In your view, is this reputation justified, and if so, what are some ideas you have for how to make the physics community more inclusive?

Anonymous said...

You seem to be exceptionally efficient at managing your time. Among other things, you serve as department chair, run an active and successful research group, teach classes and regularly post thoughtful blog posts that can't have been easy to write quickly - all while balancing work with your family life! What advice and special tips would you give to the rest of us so that we can try to emulate some of your efficiency? Is this something that came naturally to you, or did you have to learn it as you advanced in your career?

Douglas Natelson said...

Anon@1:37, that’s a very tough one, especially given that you can’t just add extra content to an already packed curriculum. A couple of simple ideas that come up when talking about this issue are (1) including reading assignments and problems based on current literature even in early classes - this is do-able even for mechanics and E&M, though it requires effort. (2) Having a current-topics-in-physics seminar type course very early on. Placing course content in the context of unsolved scientific and societal problems is also an approach, but there are limits to what can be done and still make sure that would-be majors have the in-depth foundation needed.

Anon@1:44, there is no doubt that our community needs to do better on this issue. I’m not sure physics is any worse than parts of engineering, but that’s not a high bar. By not being better at this, as a discipline we are missing out on big swaths of the population that are undoubtedly talented. Dedicated attention to the issue at all levels is at least a decent first step, and that can help with retention in the discipline. Beyond unfortunately slow processes (e.g., increasing the number of well known roll model figures works, but is not fast), I don’t know what the best approach is for quick, lasting progress. My sense is that even by the beginning undergrad stage we have already lost a lot of people from the pipeline.

Douglas Natelson said...

Anon@2:00, you are too kind. I’ve been very very fortunate. Parenting has been a team partnership and my wife is amazing. Lists are a big help - they’re the only way I can keep track of everything. I’ve definitely gotten better at this as time has gone on (and as my kids have gotten older). The fact is, it’s all about trade offs. My group is probably less high profile and my own impact is probably lower than if I traveled as much and focused as much exclusively on research as some of my colleagues.

Gautam Menon said...

A very Happy New Year to you from someone who reads your blog regularly and continues to learn many things from it. Condensed matter does have an image problem and having practitioners of their field actually present what they do as well as related areas in an understandable way really goes a long way towards correcting that. (I only wish there were more (any?) good soft-condensed matter blogs along similar lines.)

I had one specific question and I'd be interested in knowing what you think. What about the general field of "quantum biology", or looking for quantum effects in biological processes? I have been, in general, somewhat sceptical about this (inevitable thermal decoherence, especially at physiological temperatures) but there are some serious people out there who have given some thought to this (e.g. Matthew Fisher) and there might be something genuine there. There's certainly a case for good targeted experiments.

Douglas Natelson said...

Thanks, Gautam - happy new year to you, as well! I need to read more about what quantum biology enthusiasts are claiming might be quantum-mediated (in a nontrivial sense - everything involving any kind of chemistry obviously has some quantum mechanics in there) biological processes. I can buy that processes involving electron transfer and optical excitation of electronic processes could have interesting quantum aspects even at room temperature in a wet, squishy environment. I'm deeply skeptical of anything like claims that quantum computation is relevant to consciousness, or (more specific to Fisher) that nuclear spins in phosphorous are in long-lived entangled states. I admit I haven't read deeply about this, so I might be missing something, but if that were true, wouldn't large magnetic fields have very profound cognitive effects?