Slashdot Mirror
High Temperature Bose-Einstein Condensation Observed
ultracool writes "Two separate research groups claim to have observed Bose-Einstein condensation (BEC) in quasiparticles at much higher temperatures than atomic BEC — one at 19 Kelvin and the other at room temperature. The 19 K BEC was composed of half-matter, half-light quasi-particles called polaritons, and the room temperature condensate was composed of 'magnons' (packets of magnetic energy). There is some skepticism among physicists as to whether these really are BECs. If they are true BECs, these experiments are the first evidence of them in the solid state." Just in case you need a brush up on BEC, like I did, check out the Wikipedia article on Bose-Einstein condensation.
That's strange, because when it comes to only pretending to read something I stare at for a few minutes, I can hold my own against the best of them, and I didn't notice anything of the sort. Then I checked the history page, and unless you're writing this comment from Wednesday (you're not), nobody's changed anything...
Anyway, thanks for having me, dinner was marvelous, we should do this again sometime...
hugs & kisses & hearts & flowers,
~ken
How can a mishmash of atoms collapsed into the same space (b-e condensate) have a 'solid state'? Their radius' overlap. Is this more like a gas freezing without any other transition?
Ryan Fenton
I thought Bose-Einstein condensate was a completely different state of matter. How then, could it appear in a "solid state"?
:)
:)
Good question. And damn hard to explain in terms that don't sound insane to the layman
Thing is, the condensed particles here aren't the particles that make up the solid. They're not quite real particles, even. They're so-called quasiparticles, which are a fancy way condensed-matter physicists have of describing what the rest of us call "interactions". Each interaction has its own kind of quasiparticle, (and some silly name ending with -on) and they're basically described just like real particles are. The trick is you can describe the system in terms of these virtual particles instead of the real ones and simplify the problem.
To give an analogy, you could think about a bubble moving through some liquid. The bubble isn't actually a real particle - it's just the overall effect of a bunch of gas molecules pressing and bouncing against the liquid molecules. But thinking of it as just a "bubble particle" is a lot simpler.
Anyway. So the condensate here isn't made up of the solid's atoms. It's made up of quasiparticles. And this is why there's some debate on whether this should be called a BEC or not. On one hand, they can, and do have coherence here. On the other hand, they're just not really real!
But it's also pointed out they're extremely short-lived. It's indeed questionable if you can call something a BEC if it's short-lived, because a BEC is supposedly a low and stable state. (So the question becomes "How stable should it be to be a BEC?") But regardless of that, it's no less interesting.
My guess is, people will probably continue to call every BEC-like kind of condensate a BEC. When the need arises to distinguish the two, they'll have to invent a new term for that context, like "quasiparticle condensate" or something.
Even better, the idiots who insist on using as a reference a website any 12-year-old can change whenever he feels like it could at least learn to use it right.
% 80%93Einstein_condensate&diff=78635928&oldid=78633 658
Dear Wikipedia fanboys,
Learn to fucking reference it right. When you make a link to it, include the full link to the timestamp of the state it is in when you read it.
Example: http://en.wikipedia.org/w/index.php?title=Bose%E2
would have been the correct way to reference Wikipwdia for the grandparent wiki fanboy.
That way, while the content may or may not be either excellent material written by an expert on the field, or the ramplings of a moronic 12-year-old who felt like he knew how things 'work' better than the Ph.D. in the field whose entry he just erased, at least you know the reader will be looking at the same content you did.
Ok, since your "bogometer" seems to go off at one of the most highly respected scientific publications, on the planet... let me do a little physics-to-layman translation for ya.
quanta: packets of things that are quantized, like, you know, everything that happens at the atomic scale
magnetic excitations: increase in magnetic energy, for example by periodically flipping the moment like an oscillator
magnetically ordered: lined up
ensemble: group of stuff
magnetic moments: little tiny magnets formed by the electron spin.
A photon is a quanta of light energy, an energy level in a hydrogen atom is a quantized electron orbit, and a magnon is a quanta of magnetic energy. It's described using perfectly normal terminology that most undergrad physics students should be comfortable with.
Let me attempt a hopefully-understandable explanation. I'm a graduate student in experimental condensed-matter physics.
You can think about it in a coneptually-easier way by thinking about vibrations, which is more intuitive. The simplest model in which to think about vibrations would be in one dimension. Imagine you have a collection of some equal masses, equally spaced, with equal springs between each of those masses. If you excite the system anwhere (ie, push some of the masses), it will vibrate throughout the whole system because each 'atom' is coupled through the springs. The individual excitations of such a system would be the collective 'modes' of oscillation of the system. A mode is a specific oscillation that once set up will continue uninterrupted (without friction). For a simple one-dimensional system like the modes would be a sinusoidal oscillations of the system, where the wavelength of each mode would be the twice the length of the 'crystal' divided by an integer. See the wiki page on the Normal Mode with a cute animation.
You can extend this to three-dimensions by considering a three-dimensional grid of massive atoms, connected by springs. Real crystals don't have to be cubic, they can have a number of various arrangements (hexagonal, trigonal, diamond structure), and the effective spring constants can be different in different directions. But N masses, in 3 dimensions, will have 3N distinct modes. What's important to see is that each mode would have its own frequency, and wavelength, and typically the speed of propagation of each mode doesn't have to be the same. Also of note is that each mode has its own energy.
If you now consider a real crystal, and apply these same concepts but within the realm of quantum mechanics, you get a similar result, but each 'mode' now becomes a 'quanta' of lattice vibration. These vibration quanta are called phonons, which are bosons (they have spin 0, and bosons have integer spin). Even a small chunk of crystal will have on the order 10E23 atoms, so this is a huge number of allowed quantizations, and they can be thought of as a continuum. Each allowed 'mode' will again have its own frequency, wavelength, and energy. If you have a chunk of crystal at any non-zero temperature, any of the modes above the ground state (the ground state is the mode with the lowest allowed energy) can be 'occupied' with a finite probability. As you approach zero temperature, the probability of any mode above the ground state being occupied approaches zero.
A Bose-Einstein Condensate refers to an effective phase transition that happens as you cool the system and it becomes harder to excite the higher energy states as system becomes highly occupied in the ground state. There is a phase transition, the presence of which can be manifested by different qualities in things like specific heat, magnetization, magnetic susceptibility, etc. The crystal is still a solid crystal per-se (meaning it has a well-defined atomic ordering) but the occupations of the various modes of the system will drastically change, building into a near divergence at the ground state.
In the 'magnon' case as mentioned in TFA, you can think of it like phonons described above, but instead of two atoms exchanging vibrational energy, they are exchanging magnetic energy. Each electron is a spin-1/2 dipole (a fermion, not a boson), and there are interactions between two neighboring spins. Spin interactions are highly model dependent, meaning the types of atoms and shape of the crystal has huge impact on the interactions, which is why some materials are magnetic and some are non-magnetic. If you quantize the magnetic interactions you get spin-waves or magnons, similar to the sine-wave vibrational modes of the lattice above except the direction of the spin-moment changes instead of the atom displacement in the lattice.
make world, not war
The site make it very easy to pick a version of a page to link. The left side of the page say cite this article. Click it and use the link it provides.
You mad
For people who still don't understand anything of it, there is a very good article here about Bose-Einstein. Even some nice applets to play with sliders to see how it all works.
To repeat what others have said, requires education, to challenge it , requires brains.
Sorry that I'm unable to boil all of quantum mechanics and solid-state physics into a single easily-comprehensible slashdot posting, while spending a maximum of 15 minutes writing it. I included a few mentions to wikipedia (eg on modes) to aid you, and also quoted certain terms for you to look up on your own. Any quoted terms below, please look up yourself if you don't understand. This post can hopefully get you started. But I can't believe I'm being criticized for spending my own time trying to help someone entirely unfamiliar with the field understand something.
A mode is the collective motion of the atoms in the crystal, not a single frequency. A mode will oscillate at a specific frequency, however. If you write the 'equations of motion' for all atoms in the crystal in 'matrix' form, the modes would correspond to the 'eigenvalues' of that matrix. I'm sure these sentence will confuse you, but again, I can't boil linear algebra anad its application to mechanics down into a few understandable sentences to be comprehended in only a few minutes. f I tried to go too basic into all the details that post would evolveee into a textbook sized tome.
So a crystal will have several different modes. This is very much like quantum mechanics, where energy states are quantized, and each so-called 'eigenstate' has a specific 'wavefunction' associated with it. These oscillatory modes are called 'phonons', which are 'bosons'. The 'magnons' referred to in the articles are different modes. In those cases it's not vibrations they're 'quantizing' but magnetic interactions. The electrons on the atoms in the lattice are tiny 'magnetic dipoles', which can rotate, interact with magnetic fields, interact with other nearby electrons, etc. Again, if this paragraph confuses you then look up the terms in quotes.
make world, not war
You say you're an EE, but it seems apparent have you taken any solid-state physics classes yet. That's where you'll see the real utility in talking about holes. When you look at the band structure in the vicinity of an energy gap, from the quantum-mechanical point of view, excitations above the ground zero-temperature state are most easily expressable in terms of electron-occupations and hole occupations.
For example, in a direct-gap semiconductor, at zero temperature the valence band is fully occupied, and the conduction band is fully unoccupied. If you consider this system at finite temperatures, states in the conduction band can be occupied with finite probability, provided that a corresponding momentum-conserving state in the valence band becomes unoccupied. So sure, you can always write the ground state as the sum of all occupied states up to the fermi energy (the Fermi sea), but this gets mathematically very cumbersome. Especially for complicated materials with anisotropic band structures, etc.
It makes much more sense to redefine the ground state (the filled fermi sea) as being the vacuum state (ie, no occupations). Mathematically this makes calculations MUCH easier, as then an excitation will consist of exciting BOTH an electron (in the conduction band) and a hole (forcing a vacancy in the fermi sea). This is highly necessary for making calculations (such as conductivity, magnetization, specific heat, etc) actually possible to do. Now when you consider momentum and spin-dependent phenomena (magnetism, superconductivity, spintronics, etc) you have to carefully consider the excitations of the hole (what is it's momentum and spin). So yes, holes do map exactly to quasiparticles.
When you finally take some solid-state courses you'll see that holes DO HAVE an an effective mass (quite often not the same as the mass of the electron). They also have charge (-e), momentum, energy, and spin. Now regarding the polarons, if you're talking about complex quantum interactions, since any excitation into the conduction band requires similar 'excitation' of a hole, there is no reason to assume these two will act independently, they are of course highly coupled (conserving total momentum, spin, etc). In fact, creation of a particle-hole pair are somtimes called excitons. Now in the BEC systems under study, what reasons do you have a priori to assume that such quantized excitations would NOT consist of particle-hole pairs?
The concept of your post implies that you are intuitively understanding holes only as the lack of the electrons in a classical system. But when you consider the microscopic interactions with proper accounting for quantum mechanics and thermodynamics, your classical view falls far short of being feasibly workable. It becomes much MUCH MUCH easier to talk about holes as excitations of the Fermi sea.
And on one final note that's outside my element, by considering holes as excitations of the Fermi sea, Dirac made similar propositions in the burgeoning field of quantum-electrodynamics to propose the existence of a similar anti-electron (to the vacuum ground state being like the Fermi sea) which is the positron.
make world, not war
In a 3-D crystal, your momentum is a 3-D vector, and therefore 'k' is a vector. Electrons have two available spins, up and down (denote spin by quantum number s=+1 and s=-1). So in the ground state, no two electrons in the system can have the same set of quantum numbers. This means each of the 10E23 electrons has a different 'k' and 's'. The ground state can be thought of as adding electrons to the system by applying the quantum 'creation operator', adding an electron of momentum 'k' and spin 's'. So the Fermi Sea is the state producted by applying the creation operator over ALL allowable k (zero to the 'k' associated w/ the Fermi energy), and over all spins. Now if you want to keep doing this integral from the vacuum state just for simple excitations of a few electrons, you are being ridiculous. Especially when you deal with non-trivial lattice potentials as well as strong electron interactions, the integrals become VERY difficult to solve. But you can always think of small excitations from the ground state in the standard electron-hole picture, which gets quite easy, especially since you can model things as Taylor expansions about the ground state where gaps can be modelled as quadratic, etc.
A thermal or other excitation above this ground state will consist of BOTH annihilating an electron with some given 'k' and 's', and then creating it with some other 'k' and 's'. Each of these operations is done with the quantum-mechanical creation and annhilation operator, which don't necessarily commute with each other (just like position and momentum operators don't commute). This leads to nontrivial quantum phenomena.
Due to the periodic lattice structure (and hence periodicity in momentum space, along with the various Brillouin Zones), there are different allowable energies for a given 'k'. So it's MUCH EASIER to model interactions from the Fermi Sea ground state as both exciting a hole and also exciting an electron, each with their associated 'k' and 's'. Such excitations can come from a variety of sources, such as magnetic interactions, lattice interactions, etc, and become very interesting and difficult to capture. Eg, it took about 50 years to get the BCS model of superconductivity at this level after discovery of superconductivity in 1911.
But anyway, this is why it's highly useful, and thus important, to consider holes. When you run through the details in this way you see holes have an effective mass, momentum, spin, etc. And they certainly can and will interact with the excited electrons as well.
Anyway, I hope this helps, and that you don't get so accusatory when people talk about holes, because physically and mathematically it makes much sense to talk about holes as excitations. Now I have to get to work, I'm spending too much time writing these things out.
make world, not war
Not what I know of. Rb87, the most popular isotope used in these experiments is slightly radioactive, but the halflife is too long to make any difference.
;-). (WTF am I doing on slashdot.)
Identical means the same isotope and same charge. Also only neutral atoms since any charge would make it very hard to achieve the density needed to achieve BEC (ions repel each other, you know). I don't think there is any BEC with ions.
Mostly alkali atoms have been used (the ones leftmost in the periodic table). These are the ones most straightforward to cool with lasers. They have also made BEC with spin polarized He. And yes, BECs with moleculs have been made. But I think the detection of these BECs are indirect. Usually, a BEC with atoms are made and then through manipulating the scattering length (the interaction between the atoms) with a magnetic field (Feschbach resonances) a molecular BEC is made.
If this answer is incomprehensible, it might be because it is Saturday night here and we are having a great party. I am a little (just a little) drunk right know