Friday, July 30, 2021

Workshop highlights: Spins, 1D topo materials from carbon, and more

 While virtual meetings can be draining (no breaks to go hiking; no grabbing a beer and catching up, especially when attendees are spread out across a 7 timezones), this workshop was a great way for me to catch up on some science that I'd been missing.  I can't write up everything (mea culpa), but here are a few experimental highlights:

  • Richard Berndt's group has again shown that shot noise integrated with STM is powerful, and they have used tunneling noise measurements to probe where and how spin-polarized transport happens through single radical-containing molecules on gold surfaces.
  • Katharina Franke's group has looked at what happens when you have a localized spin on the surface of a superconductor.  Exchange coupling can rip apart Cooper pairs and bind a quasiparticle in what are called Yu-Shiba-Rusinov states.  With STM, it is possible to map these and related phenomena spatially, and the states can also be tuned via tip height, leading to very pretty data.
  • Amazing polymers from here.
    Pavel Jelinek gave a talk with some really eye-popping images as well as cool science.  I had not realized before that in 1D conjugated systems (think polyacetylene) it is possible to see a topological transition as a function of length, between a conjugated state (with valence-band-like orbitals filled, and conduction-band-like orbitals empty) and another conjugated state that has an unpaired electron localized at each end (equivalent to surface states) with effectively band inversion (empty valence-band-like states above filled conduction-band-like states) in the middle.  You can actually make polymers (shown here) that show these properties and image the end states via STM.  
  • Latha Venkataraman spoke about intellectually related work.  Ordinarily, even with a conjugated oligomer, conductance falls exponentially with increasing molecular length.   However, under the right circumstances, you can get the equivalent topological transition, creating resonant states localized at the molecular ends, and over some range of lengths, you can get electronic conduction increasing with increasing molecular length.  As the molecule gets longer the resonances become better defined and stronger, though at even larger lengths the two end states decouple from each other and conductance falls again.
  • Jascha Repp did a really nice job laying out their technique that is AFM with single-charge-tunneling to give STM-like information for molecules on insulating substrates.  Voltage pulses are applied in sync with the oscillating tip moving into close proximity with the molecule, such that single charges can be added or removed each cycle.  This is detected through shifts in the mechanical resonance of the AFM cantilever due to the electrostatic interactions between the tip and the molecule.  This enables time-resolved measurements as well, to look at things like excited state lifetimes in individual molecules.
The meeting is wrapping up today, and the discussions have been a lot of fun.  Hopefully we will get together in person soon!

Monday, July 26, 2021

2021 Telluride workshop on Quantum Transport in Nanoscale Systems

This week I'm (virtually) attending this workshop, which unfortunately is zoom-based because of the ongoing pandemic and travel restrictions.  As I've mentioned in previous years, it's rather like a Gordon Conference, in that it's supposed to be a smaller meeting with a decent amount of pre-publication work.   I'll write up some highlights later, but for now I wanted to feature this image.  At left is the molecular structure of [5]triangulene, which can be assembled by synthetic chemistry methods and surface catalysis (in this case on Au(111) surface).  At right is an AFM image taken using a tip functionalized by a carbon monoxide molecule.  In case you ever doubted, those cartoons from chemistry class are sometimes accurate!

Tuesday, July 20, 2021

Quantum computing + hype

 Last Friday, Victor Galitski published a thought-provoking editorial on linkedin, entitled "Quantum Computing Hype is Bad for Science".  I encourage people to read it.

As a person who has spent years working in the nano world (including on topics like "molecular electronics"), I'm intimately familiar with the problem of hype.  Not every advance is a "breakthrough" or "revolutionary" or "transformative" or "disruptive", and that is fine - scientists and engineers do themselves a disservice when overpromising or unjustifiably inflating claims of significance.  Incentives often point in an unfortunate direction in the world of glossy scientific publications, and the situation is even murkier when money is involved (whether to some higher order, as in trying to excite funding agencies, or to zeroth order, as in raising money for startup companies).   Nano-related research advances overwhelmingly do not lead toward single-crystal diamond nanofab or nanobots swimming through our capillaries.  Not every genomics advance will lead to a global cure for cancer or Alzheimers.  And not every quantum widget will usher in some quantum information age that will transform the world.  It's not healthy for anyone in the long term for unsupported, inflated claims to be the norm in any of these disciplines.

I am more of an optimist than Galitski.  

I agree that we are a good number of years away from practical general-purpose quantum computers that can handle problems large enough to be really interesting (e.g. breaking 4096-bit RSA encryption).  However, I think there is a ton of fascinating and productive research to be done along the way, including in areas farther removed from quantum computing, like quantum-enhanced sensing.  Major federal investments in the relevant science and engineering research will lead to real benefits in the long run, in terms of whatever technically demanding physics/electronics/optics/materials work force needs we will have.  There is very cool science to be done.  If handled correctly, increased investment will not come at the expense of non-quantum-computing science.  It is also likely not a zero-sum game in terms of human capital - there really might be more people, total, drawn into these fields if prospects for employment look more exciting and broader than they have in the past.  

Where I think Galitski is right on is the concern about what he calls "quantum Ponzi schemes".  Some people poured billions of dollars into anything with the word "blockchain" attached to it, even without knowing what blockchain means, or how it might be implemented by some particular product.  There is a real danger that investors will be unable to tell reality from science fiction and/or outright lying when it comes to quantum technologies.  Good grief, look how much money went into Theranos when lots of knowledgable people knew that single-drop-of-blood assays have all kinds of challenges and that the company's claims seemed unrealistic. 

I also think that it is totally reasonable to be concerned about the sustainability of this - anytime there is super-rapid growth in funding for an area, it's important to think about what comes later.  The space race is a good example.  There were very cool knock-on benefits overall from the post-Sputnik space race, but there was also a decades-long hangover in the actual aerospace industry when the spending fell back to earth.  

Like I said, I'm baseline pretty optimistic about all this, but it's important to listen to cautionary voices - it's the way to stay grounded and think more broadly about context.  

APS Division of Condensed Matter Physics Invited Symposium nominations

Hopefully the 2022 APS March Meeting in Chicago will be something closer to "normal", though (i) with covid variants it's good to be cautious about predictions, and (ii) I wouldn't be surprised if there is some hybrid content.  Anyway, I encourage submissions.  Having been a DCMP member-at-large and seen the process, it's to all of our benefit if there is a large pool of interesting contributions.


The Division of Condensed Matter Physics (DCMP) program committee requests your proposals for Invited Symposium sessions for the APS March Meeting 2022. DCMP hosts approximately 30 Invited symposia during the week of the March Meeting highlighting cutting-edge research in the broad field of condensed matter physics. These symposia consist of 5 invited talks centered on a research topic proposed by the nominator(s). Please submit only Symposium nominations. DCMP does not select individual speakers for invited talks.

Please use the APS nominations website for submission of your symposium nomination.

Submit your nomination

Nominations should be submitted as early as possible, and no later than August 13. Support your nomination with a justification, a list of five confirmed invited speakers with tentative titles, and a proposed session chair. Thank you for spending the time to help organize a strong DCMP participation at next year’s March Meeting.

Jim Sauls, Secretary/Treasurer for DCMP

Friday, July 16, 2021

Slow blogging + a couple of articles

Sorry - blogging has been slow in recent days because, despite it being summer, it's been a very busy time for various reasons.

Here are a couple of articles that I've come across that seem interesting.  On the news/popular writing front:

On the science front, there have been several cool things that I haven't had time to look at in depth.  A couple of quantum info papers:

  • In this Nature paper, the google quantum AI team have used their 53 qubit chip to do proof-of-concept demonstrations of two different quantum error correction approaches.  Perhaps someone more knowledgable that me can chime in below in the comments about how the ratio of physical qubits to logical qubits depends on the fidelity and other properties of the physical qubits.  Basically, I'm wondering if, e.g., ion trap-based schemes would be able to make even better advantage out of the 1D error correction approach here.
  • Meanwhile, in China a large group has demonstrated a 66 qubit system similar in design to the google/Martinis approach.  

Tuesday, July 06, 2021

Infrastructure and competitiveness

With the recent passage in the US Senate of an authorization that would potentially boost certain scientific investments by the US, and the House of Representatives version passing its versions for NSF and DOE, talk of "competitiveness" is in the air.  It took a while, but it seems to have dawned on parts of the US Congress that it would be broadly smart for the country to invest more in science and engineering research and education.  (Note that authorizations are not appropriations - declaring that they want to increase investment doesn't actually commit Congress to actually spending the money that way.  A former representative from my area routinely voted for authorizations to double the NSF budget, and then did not support the appropriations, so that he could claim to be both pro-science and anti-spending.) 

Looking through my old posts on related topics, I came across this one from 2014, about investment in shared research equipment at universities and DOE labs.  Since then, the NSF's former National Nanotechnology Infrastructure Network has been replaced by the National Nanotechnology Coordinated Infrastructure organization, but the overall federal support for this fantastic resource has actually gone down in real dollars, since its annual budget is unchanged since then at $16M/yr.  As I wrote back in 2014, in an era when one high end transmission electron microscope can cost $8M or more, that seems like underinvestment if the goal is to maximize innovation by making top-flight shared research instruments available to the broadest cross-section of universities and businesses.   

I reiterate my suggestion:  Companies (google? Intel? Microsoft? SpaceX? Tesla? 3M? Dupont? IBM?) and wealthy individuals who truly want to have a more competitive science and engineering workforce and innovation base should consider establishing an endowed entity to support research equipment and staffing at universities.   A comparatively modest investment ($300M) could support more than the entire NNCI every year, in perpetuity.