Packard highlights

Some things I learned at my final Packard meeting:

• The density of stars in a globular cluster is just absurd - something like 10⁴-10⁶ stars in a volume 10 ly on a side. Wow.
• The joint between the stem and base of a wine glass is a perfect lens for demonstrating the types of Einstein rings that one sees in gravitational lensing.
• In the protoplanetary disk phase of solar system formation, elements get mixed on very rapid timescales (like around 1000 years).
• Bacteria are much better at using ⁴⁰Ca in their metabolism than ⁴⁴Ca, and it's not at all clear how this works kinetically.
• 3-5 million years ago, in the early Pliocene, the global climate is a good test case for comparison with global warming models. Bad news for me: if the trends can really be mapped onto today, the hurricane rate is likely to increase by a factor of two.
• Using isotopic analysis (!), it is possible to put an error bound on how many people the lions in the Fields Museum actually ate: 41 +- 11. Anecdotal evidence had put the number between 15 and 135.
• We're all going to be able to get our genomes sequenced very soon, since the rate at which DNA can be sequenced (base pairs per day, for example) has gone up by five orders of magnitude in the last five years.

More later....

Packard meeting 2008

I'm in Park City for my last annual meeting as a Packard Fellow. As I've said before, I can't speak highly enough of the Packard Foundation. Their support has jumpstarted or otherwise contributed to pretty much my entire research program, and through their annual meetings I've gotten to know meet some fascinating people and hear excellent talks in the sciences, mathematics, and engineering. This is a particularly special meeting because it's the 20th anniversary of the Packard fellowship program, and they've invited back all the previous fellows. I'll try to post some highlights over the next few days.

Follow-ups

First, following up on my earlier post about the field effect.... I'd seen some of this before but had some really good conversations at the workshop in Japan last week about electrochemical gating. The field effect, as I'd said, is a great way of tuning the density of charge carriers at a surface without the disorder associated with chemical doping. By cranking up some gate voltage you can in some sense just rely on the attraction of opposite charges to accumulate carriers, for example. One major limitation to this technique, though, is the amount of charge that you can really get in there using reasonable dielectrics between the gate and the surface of interest. A given insulator can only take a certain amount of electric field across it before leakage current (and eventual breakdown due to damage from "hot" electrons) starts. Calling that limiting field E_max, you can find the maximum gated charge density to be \epsilon_0 \kappa E_max, where \epsilon_0 is the permittivity of free space in SI units (8.85 x 10^-12 Farad/m) and \kappa is the (unitless) relative dielectric constant. If you've got a really good oxide you can get this product up near 10¹³ carriers per square cm. Usually there is a tradeoff - materials with a big \kappa have a smaller breakdown field. One way around this is to use electrochemical gating. Instead of a dielectric, use either a polymer electrolyte or an ionic liquid. If you don't care about speed of response, this is a great idea because you can get a layer of counterions right next to the surface of interest. As a result, you can accumulate carrier densities exceeding 10¹⁴/cm². That's a huge density that can let you do some fun things even in strongly correlated materials, where you're now talking about adding or removing more than one carrier per unit cell.

A very brief follow-up to my post about the weather: Aww, not this crap again. Hurricane Gustav is really starting to look like a potential annoyance. Here's hoping that it hits neither New Orleans nor Houston.

Sad. Just sad.

According to a news article in Nature, there are now only four PhD scientists left at Bell Labs doing basic research. Four. Leaving aside that I know those guys personally, we should all be saddened by this.

I know that in industry there's always a tension between the immediate company bottom line and longer term investment. I can't help but wonder, though, in these days of institutional investors, mutual funds, and huge executive compensation, if we've really screwed up. Let me put it this way.... Once upon a time there were long-term investors who bought, e.g., AT&T, and really cared about whether AT&T was going to be competitive in ten or fifteen or twenty years. Now most stock is held by institutional investors and mutual funds who really don't care whether AT&T exists in ten years - they just care that there's something with a risk/reward profile like that in ten years. Furthermore, the executive compensation system is designed to massively reward short-term results (how much is this quarter's rate of growth greater than last year's? Note that making a big profit isn't enough - you have to be increasing your profit rate, not just the absolute amount of money the company makes.). As far as I can tell, we have effectively removed much of the economic incentive for long-term investment. No wonder any research and development with a 10-year horizon is almost gone from the American technological landscape. The only exceptions, as is often the case, are companies with so much money that a small research investment is negligible and can give decent PR, like Intel, Google, and Microsoft. Yeah, I know that HP Labs still exists, and I know that IBM still has people playing with STM, and I know that Exxon and Dupont and 3M have lots of talented chemists, but it looks like the days are largely gone of having a staff with a critical mass of tens of physics and chemistry PhDs doing cutting edge long-term research in an industrial setting.

Slow blogging....

I'm off to Japan for this week, primarily to speak here, and secondarily to guarantee that I'm a jetlagged wreck for the first week of classes. So, unless I do some Lost in Translation blogging, it'll be a quiet week here.

Cryptophysicists

I think that we need to coin an official term, "cryptophysicist", to describe people who do physics research outside the mainstream. Ronald Mallett is an example of a credentialed cryptophysicist - he wants to build a time machine using circulating optical beams. His tragic motivations aside, this is a scientifically wacky idea - the energy density that you would need in the beams to produce any significant distortion of spacetime is completely unachievable with foreseeable equipment. On the theory side, Harold Puthoff is another example. Puthoff wants to explain things like inertia in terms of interactions between matter and zero-point fluctuations of the electromagnetic field. Mainstream theorists consider this to be a wacky idea for a long list of reasons.

One major difference between cryptophysicists and cryptozooligists is that the public is generally able to perceive that the latter are outside the mainstream. Everyone knows from daily experience that there probably aren't yeti or sea monsters hanging around. Modern physics is abstracted enough from everyday lives and intuition, though, that many people, including some journalists, honestly can't tell when someone's waaay out there. Also, the concept of the lone genius toiling away in obscurity fighting The Scientific Establishment, which makes for good TV, sounds better when applied to a garage tinkerer than to someone camping out looking for the chupacabra. Still, occasionally the biologists do get to have fun with media coverage of this stuff.

The US election and science

For anyone interested in the US presidential candidates' positions on science, I recommend this NPR piece from today's "All Things Considered". What I find most revealing is near the end. Obama supports the goals of the America COMPETES act, which (among other things) was intended to double support for basic science research over the next several years. McCain does not consider science research to be as high a priority, given fiscal constraints. Bear in mind that as a percentage of GDP the US spending on physical science research has been falling for decades, and we've also witnessed the near-death of industrial long-term research in the sciences. I'm not trying to start a political flamewar here - I just wanted to point people to this information.

Update: From the Department of the Obvious, this article from USA Today about general science literacy and cultural perception in the US. How depressing. 44% of those surveyed couldn't name a single scientist, living or dead, that could be a role model for young people. Wow.

Whoa.

This article (thinks, Incoherent Ponderer, for pointing this out and your insightful post) is really distressing. Prof. Plummer is a class act and chaired a session of the APS Focus Topic that Eric Isaacs and I put together for this past March Meeting. I hope Tennessee is able to find the resources to support him fully.

Random samples

Three little items....

First, the Foundational Questions folks have announced their funding recipients. These grants, intended for fundamental questions in physics and cosmology, are supposed to target topics "unlikely to be otherwise funded by conventional sources". I'm a bit skeptical about that, since it sure looks like a number of these people are getting funded for topics that sound just like their regular research. Still, I do think it's a good thing to put some resources into real foundational questions. For the condensed matter people in the audience, our token representation on the list is Subir Sachdev, who is extending his theory formalism to cover black holes (!), and Keith Schwab, who is going to do true quantum mechanics by building micromechanical systems that exist in a quantum regime.

Second, this paper is one of the more bizarre things I've seen on the arxiv lately. It's from some folks trying to find efficient ways to compute pagerank numbers (useful in search algorithms for the web). They claim to have a Schroedinger-like equation that describes the pagerank. I was intrigued, in part because of all the lame joke opportunities this suggested. Funky.

Lastly, it's disappointing to read articles like this one and this one. I don't know what's more sad - that the the Holmdel Bell Labs site is being sold to developers, or that the Wall Street Journal can run multiple long articles about the discord and tribulations at the top of Alcatel-Lucent and not even mention Bell Labs. I can't help but think that the US will really miss industrial long-term research.

Ahh, the weather.

It's that time of year again, when we get a brief break from high-90s (F) heat and high humidity by having a tropical weather system blow through the area. Hopefully this will bring us some much needed rain, but not too much. It's interesting to watch the evolution of computer models here, and also to keep track of historical trends. I never had to think about this stuff growing up in Pennsylvania....

Why the field effect is important.

I'm probably overdue for making a post introducing some physics topic for a nonspecialist audience, so here you go....

What is the field effect? Well, it's the physical process that is used in field effect transistors (FETs), the hundreds of millions of little three-terminal switches that are the basis of the computer you're using right now. In a FET the electrical conduction between two electrodes (a source and a drain) is modulated by applying a voltage to a third electrode (the gate) that is capacitively coupled to (though separated by a thin insulating dielectric layer from) the semiconductor material between the source and drain (the channel). In an ideal (flat band, for the experts) FET, if the gate voltage is adjusted to be positive relative to the source and drain electrodes, a layer of electrons will be capacitively attracted from the source and drain into the channel region. Once there are carriers accumulating in the channel, conduction can take place from the source to the drain. Alternately, one could imagine cranking up the gate voltage to be negative relative to the source and drain, in which case one would expect to accumulate a layer of holes rather than electrons. A device that can really be modulated to have either electrons or holes as the dominant carriers is said to be ambipolar. Of course, real devices are more complicated than this. For example, there can be surface states that live at the semiconductor/dielectric interface close to the gate; the source and drain electrodes may have preferential alignment to either the conduction or valence bands depending on the materials involved; etc. Still, the FET is the basis for modern electronics, and techniques exist for fabricating these little gadgets by the billions in Si with incredible reliability.

That's why the FET is important for industry. Why is the FET important for physics? The FET idea can be rephrased this way: the FET is a way of changing the chemical potential of charge carriers (or, in a related sense, the density of charge carriers) without altering the composition of the material. Think about that for a sec. Chemists understand doping very well. If you're doing chemistry and want to remove some electrons, you add a halogen atom somewhere since those are hugely electronegative and grab an electron. Alternately, you want an extra electron in there? Throw in an alkali metal atom - they love to give up electrons. This kind of doping is commonly done in, e.g., the high temperature superconductors, where the parent copper oxide compound is actually an (antiferromagnetic) insulator that only becomes conducting (and superconducting) if doped chemically. The problem with chemical doping is that adding new atoms to a solid changes its structure and usually increases disorder, since we generally can't control precisely how the extra atoms are distributed. In FET structures, conversely, one can try to tune the carrier density just by turning up a voltage, no chemistry required. In graphene, for example, people have gated readily from having lots of electrons to having lots of holes and back, all in one device. Of course, FET structures do have some serious limitations: they only really put charge onto a surface or interface (rather than the bulk), and it is extremely challenging to get field effect charge densities to be comparable to those achieved easily with chemical doping. Still, the field effect is a tremendous tool for tuning electronic properties without mucking up the disorder or structure. If you're interested in this in more detail, please check out this Rev Mod Phys paper.

Physics virtual swap meet

Now this is a modest proposal whose time has come. It's common that experimental physicists gradually (or not so gradually) accumulate some pieces of equipment over the years that are laboratory "white elephants". These items are typically acquired for some specific research project or direction, and then over the years as research goals and priorities change, they can end up sitting around gathering dust. Surely someone else somewhere could make productive use of these items. Maybe someone should set up a trading post, where we could list these things and arrange reasonable trades or purchases.

Yes, used equipment vendors exist to address these needs, but they're not not always easy to do deal with, and frequently they offer pennies on the dollar. (For example, at Rice we have an 11-year old electron microscope. It's got a problem that is likely to cost about $10K to fix, though annoyingly the microscope vendor refuses to have a reasonable return policy - if we bought the relevant part and that didn't fix the problem, they'd refuse to take it back even for a restocking fee. Two different used equipment vendors have offered around $8K for the whole SEM (!), while they both offer the same item for sale on their sites for more than $100K.) I also know that people buy and sell scientific equipment on E-bay, but that suffers from some of the same problems. The point here isn't for scientists or universities to make money on this - it's to match up scientists/engineers with equipment that they could use and that currently has the wrong home.

Two papers in Nano Letters

Two new papers in Nano Letters caught my eye.

Danilov et al., "Nanoelectromechanical switch operating by tunneling of an entire C60 molecule"
This is a single-molecule electronic experiment, and it's a pretty neat example of using careful measurements to deduce a fair bit of information about a nanoscale system without a direct microscopic imaging probe. In their experiment the authors find bistable switching between two different conducting configurations of a junction thought to contain a single C₆₀ molecule (as inferred by a signature in the electronic conduction of a well-known vibrational mode of the fullerene). Now, telegraph-like switching isn't new by any stretch, and many nanoscale systems exhibit discrete changes in their properties due to motion of one or a few atoms or molecules. Here, by carefully analyzing the voltage and temperature dependence of the switching, they are able to deduce that the most likely mechanism for the change in configuration is motion of the fullerene between two different sites. I'm not sure that I'm 100% convinced by their interpretation, but the data are quite pretty and the analysis is clever.

Stampfer et al., "Tunable graphene single-electron transistor"
In this work, Klaus Ensslin's group starts from a graphene flake (the now-usual scotch-tape approach) and uses standard patterning methods to make a little graphene region connected by narrow graphene constrictions to source and drain electrodes. Using gate electrodes near the constrictions, they could electrostatically shift the local chemical potential there, shifting the graphene from having electrons charge carriers to holes as charge carriers. By further tuning with an additional gate, they could make a complete single-electron transistor - essentially a puddle of confined charges weakly connected by tunnel barriers to larger reservoirs, with the confinement sufficiently strong that the electrostatic energy cost of changing the puddle population by one charge exceeded the available thermal energy. (This was all done at cryogenic temperatures, since the charging energy of the dot was around 3 meV, or about 35 K.). This is a particularly nice, clean example of using graphene and its relatively unique properties as a platform for nanoscale device fabrication. Several groups are getting very good at this, and if efforts to grow large area, high quality, single-layer graphene succeed, there could be some genuine technological implications.

LHC publicity machine

I understand that the folks at CERN feel like it's important for people to be aware of the LHC and get excited about it - at this point, it looks like it's going to be the only game in town in a few years for the frontier of high energy physics. Still, the steady stream of publicity (much of it arguing that they're going to unlock the secrets of the universe, prove string theory, find evidence of extra dimensions, etc.) is getting to be a bit much. Today comes this article discussing the cooldown of the magnets for the collider and the detector. Technologically impressive to be sure, but the whole "colder than deep space" angle is pretty lame - people have been able to reach these temperatures for nearly 100 years, and superconducting magnets are used in thousands of MRI machines the world over. We get it - it's a big machine. If this is the level of publicity hounding that's going on before they even have a single piece of data, the coverage of the actual physics runs is going to be really oppressive.

Scientists, the media, and desperation

I could've predicted this. Given current energy concerns, it's not at all surprising that the various media are ready to give airtime and column space to wacky stories like this one. The temptation must be irresistible: the public is desperate; the story itself is great TV - the lone inventor, persevering in the face of opposition from those stodgy old scientists; they can even put in quotes from the would-be inventor and the scientists and claim to be covering "both sides". You know the drill: "This conventional scientist says that if he drops this pencil it will fall to the ground. Others disagree! The controversy, up next after this commercial message." News flash: sometimes it doesn't make any sense to cover "both sides".

I think the part that frustrates me the most is the misperception by part of the public and some of the media that scientists want these alleged breakthroughs to fail. Nothing could be further from the truth! If someone discovered cheap, inexhaustible energy because of a remarkable revolutionary breakthrough, we'd love it - it'd be the most exciting time in science since the quantum revolution. The problem is, though, that keeping an open mind doesn't mean lowering your scientific standards because you'd like to believe the result. I, for one, am not holding my breath about hydrino power.

slow blogging + pnictide fun

I'll be traveling (work + vacation), so blogging will be slow until July 15 or so.

Before I go, I wanted to point out that the plot continues to thicken regarding the pairing symmetry of the new iron pnictide superconductors. For example, this paper reports scanning SQUID microscopy on a sample of one of the compounds, with no apparent evidence for sign flips in the order parameter that you might expect if the material was, e.g., d-wave like the cuprates. In contrast, this paper argues that scanning tunneling spectroscopy data resemble d-wave expectations. This paper reports photoemission studies showing that the compounds have quite a complicated band structure and suggests that different parts of the Fermi surface may have different phenomenology. That sounds reminiscent of this theory paper, but I haven't read them in detail.

What makes an experiment "good"

Recently I've had some conversations with a couple of people, including someone involved in journalism, about what makes a physics experiment good. I've been trying to think of a good way to explain my views on this; I think it's important, particularly since the lay public (and many journalists) don't have the background to judge realistically for themselves the difference between good and bad scientific results.

There are different kinds of experiments, of course, each with its own special requirements. I'll limit myself to condensed matter/AMO sorts of work, rather than high energy or nuclear. Astro is a whole separate issue, where one is often an observer rather than an experimenter, per se. In the world of precision measurement, it's absolutely critical to understand all sources of error, since the whole point of such experiments is to establish new limits of precision (like the g factor of the electron, which serves as an exquisite test of quantum electrodynamics) or bounds on quantities (like the electric dipole moment of the electron, which is darned close to zero as far as anyone can tell, and if it was nonzero there would be some major implications). Less stringent but still important is the broad class of experiments where some property is measured and compared quantitatively with theoretical expectations, either to demonstrate a realization of a prediction or, conversely, to show that a theoretical explanation now exists that is consistent with some phenomenon. A third kind of experiment is more phenomenological - demonstrating some new effect and placing bounds on it, showing the trends (how the phenomenon depends on controllable parameters), and advancing a hypothesis of explanation. This last type of situation is moderately common in nanoscale science.

One major hallmark of a good experiment is reproducibility. In the nano world this can be challenging, since there are times when measured properties can depend critically on parameters over which we have no direct control (e.g., the precise configuration of atoms at some surface). Still, in macroscopic systems at the least, one should reasonably expect that the same experiment with identical sample preparation run multiple times should give the same quantitative results. If it doesn't, that means (a) you don't actually have control of all the parameters that are important, and (b) it will be very difficult to figure out what's going on. If someone is reporting a surprising finding, how often is it seen? How readily is it reproduced, especially by independent researchers? This is an essential component of good work.

Likewise, clarity of design is nice. How are different parameters in the experiment inferred? Is the procedure to find those values robust? Are there built-in crosschecks that one can do to ensure that the measurements and related calculations make sense? Can the important independent variables be tweaked without affecting each other? Are the measurements really providing information that is useful?

Good analysis is also critical. Are there hidden assumptions? Are quantities normalized in sensible ways? Do trends make sense? Are the data plotted in ways that are fair? In essence, are apples being compared to apples? Are the conclusions consistent with the data, or truly implied by the data?

I know that some of this sounds vague. Anyone more eloquent than me want to try to articulate this more clearly?

New comic

Thanks to Tom for having a link on his page to this. Too true! This is also brilliant. Good to have a laugh, despite the passing of George Carlin. One of my favorite Carlin quotes: [W]e have flamethrowers. And what this indicates to me, it means that at some point, some person said to himself, "Gee, I sure would like to set those people on fire over there. But I'm way too far away to get the job done. If only I had something that would throw flame on them."

No "Singularity" for you.

I wasn't going to even mention the idea of a Singularity, but then the IEEE made a point of dedicating an issue of their magazine to the concept. For those who don't know, the term "Singularity" originates with sci-fi author Vernor Vinge, who has written some compelling novels. Proponents of the concept believe that we live in an era of exponentially accelerating technological change, and that at some point (the Singularity) there will be a complete break in the nature of our species and societies, ushering in what some call a transhumanist future. The technologies typically associated with this idea are (1) Drexlerian molecular nanotechnology, so that we can eliminate scarcity by building anything we want anytime we want via (self-reproducing) nanomachines; (2) immortality via nanotechnological or biochemical control over biological processes that lead to senescence; and (3) strong AI, often including the concept of people uploading their minds to constructed hardware. The thing that continues to surprise me about this idea is that so many people seem to take it so seriously.

Hey, I'm all for optimism, and I'm generally bullish on the future of the species despite current scariness and some scientific arguments, but asserting that we will have a transhumanist utopia in twenty or thirty years is a wee bit of a reach, to put it mildly.

New physics building - suggestions? ideas? horror stories?

My university is in the design phase on a new physics building. This is exciting - first, it's a rare opportunity to design new lab space literally from the ground up. Second, new space will make possible some targeted expansion in the experimental directions in our department as well as in the experimental physicsy part of our electrical and computer engineering department.

Anyone out there have suggestions on building design, particularly with regards to laboratory facilities, utilities, HVAC, electrical service, vibrations, etc.? We're already looking at several recently constructed buildings elsewhere to learn lessons about best practices. If you have thoughts on physics buildings that you think are particularly well done (e.g., the electrical wiring system for the labs at the new nano building at Purdue looks extremely clever and well done), or, conversely, specific examples of design ideas that are lousy in practice or implementation, please post in the comments or email me.