Since Willis just published an essay on refrigeration systems and how the Earth has its own version, I thought this story might be fun and educational. Many people don’t know that Albert Einstein invented a refrigerator system in 1926 after he became world famous for his Theory of Relativity that was proven by solar eclipse measurements in 1922. I mean, after that what do you do for an encore? Build a fridge I guess.
About the same time, Einstein also became the most prominent critic of Quantum Theory which he had helped to create decades earlier. Given that, I think he’d find the idea of a Quantum refrigerator both hilarious and intriguing at the same time. – Anthony
Researchers at the National Institute of Standards and Technology (NIST) have demonstrated a solid-state refrigerator that uses quantum physics in micro- and nanostructures to cool a much larger object to extremely low temperatures.
What’s more, the prototype NIST refrigerator, which measures a few inches in outer dimensions, enables researchers to place any suitable object in the cooling zone and later remove and replace it, similar to an all-purpose kitchen refrigerator. The cooling power is the equivalent of a window-mounted air conditioner cooling a building the size of the Lincoln Memorial in Washington, D.C.
“It’s one of the most flabbergasting results I’ve seen,” project leader Joel Ullom says. “We used quantum mechanics in a nanostructure to cool a block of copper. The copper is about a million times heavier than the refrigerating elements. This is a rare example of a nano- or microelectromechanical machine that can manipulate the macroscopic world.”
The technology may offer a compact, convenient means of chilling advanced sensors below standard cryogenic temperatures—300 milliKelvin (mK), typically achieved by use of liquid helium—to enhance their performance in quantum information systems, telescope cameras, and searches for mysterious dark matter and dark energy.
As described in Applied Physics Letters,* the NIST refrigerator’s cooling elements, consisting of 48 tiny sandwiches of specific materials, chilled a plate of copper, 2.5 centimeters on a side and 3 millimeters thick, from 290 mK to 256 mK. The cooling process took about 18 hours. NIST researchers expect that minor improvements will enable faster and further cooling to about 100 mK.
The cooling elements are sandwiches of a normal metal, a 1-nanometer-thick insulating layer, and a superconducting metal. When a voltage is applied, the hottest electrons “tunnel” from the normal metal through the insulator to the superconductor. The temperature in the normal metal drops dramatically and drains electronic and vibrational energy from the object being cooled.
NIST researchers previously demonstrated this basic cooling method** but are now able to cool larger objects that can be easily attached and removed. Researchers developed a micromachining process to attach the cooling elements to the copper plate, which is designed to be a stage on which other objects can be attached and cooled. Additional advances include better thermal isolation of the stage, which is suspended by strong, cold-tolerant cords.
Cooling to temperatures below 300 mK currently requires complex, large and costly apparatus. NIST researchers want to build simple, compact alternatives to make it easier to cool NIST’s advanced sensors. Researchers plan to boost the cooling power of the prototype refrigerator by adding more and higher-efficiency superconducting junctions and building a more rigid support structure.
This work is supported by the National Aeronautics and Space Administration.
###
Cool
This reminds me of Maxwell’s Demon. The demon was dreamed up by Maxwell as a way to violate the 2nd law of thermodynamics. It sits at a small opening between 2 identical chambers of gas, and allows only the fastest-moving molecules to pass one way through the opening. This has the effect of heating up and increasing the pressure in one chamber, while the other cools down and loses pressure, thereby violating the 2nd law. Apparently it can be shown that in order to perform this simple task, the Demon himself gains entropy, and the 2nd law is safe!
World’s geekiest and coolest BEER fridge!!!
I don’t recall that Eddington’s eclipse measurements were precise enough to really confirm general relativity. Galactic gravitational lensing has done that big time, though.
Richard Barraclough,
“Apparently it can be shown that in order to perform this simple task, the Demon himself gains entropy, and the 2nd law is safe!”
Interesting. I wonder if God gained entropy too, when He created the universe.
Hmmm, a window air conditioner in the Lincoln Memorial wouldn’t cool very much at all.
ActonGuy says:
March 12, 2013 at 4:42 am
How much power is required? i.e., if there was a window-sized unit, how much power would it take to cool a building the size of the Lincoln Memorial, and how much power would conventional AC units use?
=========================================================
The important questions are:
What’s its carbon footprint
and how many solar panels / windtowers will it take to power it
I have nothing relevant to add but still wish to be in the running for the mug.
I understand being careful in language, but the theory of general relativity has been “proven” to be quite useful in improving GPS accuracy and many other things we take for granted daily. In the future we may find, that like Newton before him, he only described a special case of a yet larger whole, but it will probably be a while before science reaches the level of precision required to notice any aspects of that larger whole that show the theory to be incomplete.
I understand that an experiment can only do one of three things: repudiate a theory by showing a counter example in a case in which it should apply; be inconclusive due to error bars larger than the effect being measured or other experimental problem; or support the theory by measuring something uniquely predicted by it.
We tend to accept a theory when hundreds and thousands of experiments find the predicted outcome, but we no longer like to call new ideas “laws” like they did in the 17th-19th centuries. Thus we have Newton’s Law of Gravity which isn’t really a law as we have shown it to be incomplete. We have Ohm’s law, which works where it applies, but has a little problem at the micro scales of quantum tunneling semiconductors. In fact most of the classical “laws” have little hitches when we attempt to apply them to extremely small or extremely large phenomena, but work just great in their traditional domains. I wouldn’t do electronics any more if I had to solve quantum statistics for every calculation to get circuits to work properly (which probably is necessary for extreme precision circuits).
What this does for us is it saves our helium reserves, if it can be scaled enough to be used in all universities. Compressing helium into a liquid state is currently the only viable method for achieving such temperatures.
“I understand that an experiment can only do one of three things: repudiate a theory by showing a counter example in a case in which it should apply; be inconclusive due to error bars larger than the effect being measured or other experimental problem; or support the theory by measuring something uniquely predicted by it.”
and dont forget the 4th. Appear to support the theory, where the happy consequence of bad measurement and bad theory work together. Oh and the 5th, an experiment can suggest that a theory is incomplete and additional hypotheses are needed to explain the result. oh and the 6th, appear to repudiate the theory, but actually point to bad data collection.
And of course, nothing in the experimental result tells you what to do next.
[Reply: 7th… PROFIT!! ~mod]
Not cool–COLD! Very, very cold!
policycritic
In what way are “… I guess the worker bees at Wikipedia are hard at work promoting the myth” that Einstein won a Nobel for the Theory of Relativity?
Maybe there are those that would like a ‘crash course’ to allow them to ‘come up to speed’ on where CF A/K/A LENR stands today … this 101-level ‘survey’ course (“IAP short course”) given at MIT in a series of 12 videos by Dr. Peter Hagelstein in January of this year (2013) may just fill that bill:
** Notice the warning issued in the beginning: “Working in this field can destroy your career .. this is a very dangerous field to be associated with” **
#1 Cold Fusion 101 Dr. Peter Hagelstein at MIT 01/22/2013 (Day 1 Part 1)
(Note the tongue-in-cheek ref to Wikipedia in 1 above)
#2 Cold Fusion 101 Dr. Peter Hagelstein at MIT 01/22/2013 (Day 1 Part 2)
.
And so on …
.
What exactly does the temperature of an electron mean?
In any material, the electrons are typically in thermal equilibrium with the material itself. Electrons are fermions, and practically speaking this means that they have a distribution of energies around the so called “Fermi Energy”, the energy of the Fermi Surface where all of the electrons are packed into their ground states subject to the Pauli Exclusion Principle. Without going into the details of electronic band structure, this surface exists in a reciprocal “momentum” like space consisting of the wave-numbers of the electrons in various directions in the lattice.
At finite temperature some of the electrons near the fermi surface are kicked out of their ground state and into the bands of states just above that surface. They leave behind “holes” in the bands of states just below that surface. This, along with whether or not the fermi surface occurs inside a band or in the gaps in between bands, determines things like the conductor/insulator properties of materials. Superconductivity is also closely related to this sort of fermi surface structure, although it is a purely quantum phenomenon that I won’t go into here.
The electrons in the bands above the fermi surface at finite temperature have an energy ABOVE the ground state energy, and hence one can, to the extent that there are enough of them in a slice of a band for thermal averaging to make sense, associate a temperature with their energy using e.g. k_B T = E (where k_B is Boltzmann’s constant, and where the relationship probably isn’t strictly linear). “Hotter” electrons are in basically found in bands with higher energies farther above the fermi surface for the material.
So what the nanoscale cooling elements do is basically create a way for these “hotter” higher energy electrons to be differentially removed from a material along with an applied current. As they are removed, they are replenished with “cooler” lower energy electrons (the material remains electrically neutral). However, they quickly equilibrate with the temperature of the material lattice — the “phonon” temperature associated with the microscopic oscillations of the actual massive atoms that make up the substance. In the process the electrons absorb energy from the lattice, dropping its enthalpy content and hence its temperature. These hotter electrons are then removed, in a more or less continuous cycle as long as current is applied.
This is a crude picture, but should be enough to answer your question — “hotter” electrons are ones that on average have energies farther above the fermi surface for the material, even though individual electrons have no temperature, only energy, and perhaps the secondary description will give insight into how the cooling takes place.
This article might also help, as Peltier coolers have been around for a long time and are functionally quite similar:
http://en.wikipedia.org/wiki/Thermoelectric_cooling
Their basis isn’t exactly the same, as it is more classical, but the IDEA of differentially moving electrons at a higher temperature to obtain cooling is very much the same.
rgb
This is one of the reasons why I support nuclear power! Collect all those alpha particles after they have spent the majority of their kinetic energy, add electrons and poof instant Helium. (or do they only do this in research reactors?)
Good to see them getting some publicity over at NIST. Those guys have been doing some good work for a while.
This is not something you scale into a common refrigerator. This is for taking things that are really cold and making them really really cold. In general, you just can’t have quantum effects at high temperatures because all the particles involved are switching states way too frequently to allow any effect. That’s why room temperature superconductivity is likely impossible.
Why this is a big deal, and what applications they are used for: the biggest application they have in mind are very low temperature detectors of light from gamma rays all the way through millimeter waves, things like bolometers. The fundamental limit for sensitivity is how cold you can make your detectors. If you can make them really cold for really cheap, then you solve a lot of problems. This has applications for putting arrays of detectors into cameras, acting sort of like ultra-sensitive CCDs that work at all kinds of weird wavelengths. I think the biggest application is in astronomy, but there’s a lot more out there, including dark matter searches, radioactive isotope testing, and even passive security applications (i.e. tell if someone has a knife without bombarding them with x-rays). Before the only way to get colder than 250 or so mK was to use a dilution refrigerator (where you take advantage of the fact that 4He and 3He like to mix together at a certain ratio to cool things to ~10 mK) or an adiabatic demagnetization refrigerator, which is tricky, tempermental and has very low cooling power. If this can be done cheaply and easily, a lot of these applications get a lot easier.
As for the shortage of helium: most instruments no longer just boil off liquid helium (most, not all). Most use pulse tube coolers to get to LHe temperatures, then have closed cycle refrigerators to go lower, so you’re not actually losing helium. The trouble is, compressing 4He can only get you down to 0.8K or so, which is still way too hot for most applications. You need 3He, which is not found naturally, to get down to the ~300mK temperatures even to start. 3He is a byproduct of making nuclear weapons, as I understand it, so it will only get more scarce for us.
Has anyone checked their thermometer siting?
Joking aside, I liked their chip scale atomic clock too:
http://www.nist.gov/public_affairs/releases/miniclock.cfm#
It was rapidly commercialized and you can buy them:
http://www.symmetricom.com/products/frequency-references/chip-scale-atomic-clock-csac/SA.45s-CSAC/
Good science being done at NIST; elsewhere? not so much …
“… chilled a plate of copper, 2.5 centimeters on a side and 3 millimeters thick, from 290 mK to 256 mK. The cooling process took about 18 hours.”
Wait….something bothers me.
Does 290mK to 256mK mean From 290 milli-Kelvin to 256 milli-Kelvin, in 18 hours?
In that case, do you realise how cold 290 milli-Kelvin already is, before starting? It is 0.29 Kelvin above the absolute zero point!!!
And from 290 milli-Kelvin to 256 milli-Kelvin is only 0.034 Kelvin difference…….and it tool 18 hours…
So, what bothers me is; How long would it take to cool something down from 272.29 Kelvin to 272.256 Kelvin? 18 hours too?
I think their could be a market for “quantum can coolies” or on a larger scale “Earth cooling superconducting quantum global warming balancer” . If we could just get the gang to tell us what the perfect temp should be where we won’t have hurricanes, droughts, tornados and such. We could just set the thermostat to the right temp. /sarcasm
Billy Liar says:
” …It was rapidly commercialized and you can buy them:
http://www.symmetricom.com/products/frequency-references/chip-scale-atomic-clock-csac/SA.45s-CSAC/ ”
Oh, God. Where’s the back of the queue of amateurs trying for a complementary sample?
Any comments on the efficiency of the cooling unit? Low cost, high efficiency cooling to cryonic temperatures would have some interesting applications. (for example using LN2 as a way to store “cold” for load leveling summer cooling demands, you can recover meaningful energy from the expansion as well)
Killer series there Rocky … although some (make that: “many” over 30; the artist/producer of that work states his “… purpose is to reach those under 30”) aren’t going to be able to ‘stomach’ the sound/music track … to that end I recommend maybe simply jumping-ahead to Pt. 5 in the series (a little more theory w/experimental results, demos tossed in):
Part 5
BTW, the series doesn’t claim to get mathy until #12 (or so) in the series.
.
Hobbyists have never had it so good … maybe too late for ‘samples’ (when everyone is asking for one!) but easily available via CC debit:
https://www.sparkfun.com/pages/GPS_Guide.
Note the optional update-rates on some of those exceed 10 Hz (10 position updates per second!)
.
Cooling the mass using just the “thermal mass” of the electrons is where the room air conditioner analogy arises. Having played with the Josephson Effect, I can appreciate the excitement that this creates. There’s nothing like having a macro response to a atomic event to get the blood flowing!