The researchers, Parthiban Santhanam and coauthors from MIT, have published their study in a recent issue of Physical Review Letters.
As the researchers explain in their study, the key to achieving a power conversion efficiency above 100%, i.e., unity efficiency, is to greatly decrease the applied voltage. According to their calculations, as the voltage is halved, the input power is decreased by a factor of 4, while the emitted light power scales linearly with voltage so that its also only halved. In other words, an LEDs efficiency increases as its output power decreases. (The inverse of this relationship - that LED efficiency decreases as its output power increases - is one of the biggest hurdles in designing bright, efficient LED lights.)
In their experiments, the researchers reduced the LEDs input power to just 30 picowatts and measured an output of 69 picowatts of light - an efficiency of 230%. The physical mechanisms worked the same as with any LED: when excited by the applied voltage, electrons and holes have a certain probability of generating photons. The researchers didnt try to increase this probability, as some previous research has focused on, but instead took advantage of small amounts of excess heat to emit more power than consumed. This heat arises from vibrations in the devices atomic lattice, which occur due to entropy.
This light-emitting process cools the LED slightly, making it operate similar to a thermoelectric cooler. Although the cooling is insufficient to provide practical cooling at room temperature, it could potentially be used for designing lights that dont generate heat. When used as a heat pump, the device might be useful for solid-state cooling applications or even power generation.
Theoretically, this low-voltage strategy allows for an arbitrarily efficient generation of photons at low voltages. For this reason, the researchers hope that the technique could offer a new way to test the limits of energy-efficiency electromagnetic communication.
Explore further:
Darpa seeks new power dynamic for continuation of Moore's Law
More information:
Parthiban Santhanam, et al. Thermoelectrically Pumped Light-Emitting Diodes Operating above Unity Efficiency. Phys. Rev. Lett. 108, 097403 (2012). DOI: 10.1103/PhysRevLett.108.097403
Physics Synopsis
Deathclock
Lurker2358
So LEDs are over-unity machines.
Wait. I thought the U.S,. patent office stance was that over unity machines don't work, which is why they turned down Rossi.
230% efficiency? Really?
So you invented a reverse-entropy machine eh?
Let's toss out all them text books, along with everything this site represents I suppose...
Deathclock
Callippo
If the cold fusion is overunity machine, then the hydrogen bomb is perpetuum mobile. Cold fusion is not an overunity machine - it's a process, which is thermodynamically perfectly feasible, it's just believed to proceed too slowly under normal conditions. Rossi didn't get a patent, because he's the US government does want to give a monopoly to the future energy production to man, who isn't USA citizen. It's a political thing, as many similar devices were patented already in USA w\out problem.
El_Nexus
Not quite. We tend to assume that because luminescence is useful it must be low in entropy, which it's not. It's less disordered than heat, but far more disordered than, say, mechanical work. By emitting luminescence, the LED reduces its own internal entropy but still increases the entropy of the universe as a whole.
Callippo
Lurker2358
I know Rossi's device is not an over unity machine.
But the U.S. patent office classified it as one nevertheless.
Callippo
rwinners
Deathclock
No... This does not violate the first law of thermodynamics.
El_Nexus
Energy is being conserved. Heat is a form of energy, and it is being converted into light by this process.
rbrtwjohnson
kochevnik
Cave_Man
I have a sneaking suspicion that when they say 30 picowatts to 69 picowatts they are simply using a standard of electrical energy which needs to be incorporated into quantum mechanical units.
kind of like how a charged hydrogen atom may yield a certain electron volt quanta but the total energy in the atom system is much larger than the available charge. One can assume since an electron is more massive than a photon that the higher the conversion efficiency the more the complex heavy electron is converted to simple lower energy photons.
They aren't making any free energy only freeing up what is there with greater efficacy.
nuge
antialias_physorg
I think it's much simpler than that. During operation the LED gets hot. When you reduce the voltage some of the heat gets converted into infrared radiation (additionally to the light generated by the LED). So for a short while we get more than is being put in from the wall plug - but not more than was put in over the entire time of operating the LED.
GrandM4x
Energy is conserved as it will always be.
Lurker2358
That's too easy to control for, and I hope these researchers aren't that stupid.
All you'd need do to control for that is use very short on/off pulses and measure the energy in vs energy out.
This way you avoid any hidden buildup of heat or other energy coming from the original power supply.
Graeme
Richardmcsquared
Skepticus
gwrede
They take liberties, such as "forgetting" the thermal engergy use. That's like saying that "my bike that has an electric assistant motor, is 200% efficient" when actually I do half the work myself. I think this is downright childish.
Oh, and BTW, ever thought what kind of light you get with 10-8 watts anyway. So the whole thing is actually a lot of talk about nothing at all.
_ucci_oo
antonima
Right, only the second law is getting screwed.. its not like its the first time!
Unless it is meant to be a shock article, it is completely commonplace for a diode to emit light, even when no voltage is applied at all! Quoting a photonics textbook:
"At room temperature, the intrinsic concentration of electrons and holes in GaAs is n1 == 1.8 * 10^6 cm^-3. Since the radiative electron-hole recombination coefficient rr == 10^-10 cm^3/s, the electroluminescence rate Rrnp=RrN^2i == 324 photons/ cm^3-s... this corresponds to an optical power density == 7.4 * 10^-17 W/cm^3."
Photodiodes will emit light on their own, and a semiconductor with a lower bandgap will emit even more equilibrium light than GaAs. Since clearly this is a non-equilibrium system, there is probably more to it than just that.
Kedas
The question is who is heating it.
ShotmanMaslo
SincerelyTwo
So the fallacy here is that most of you decided to assume that the title of the article was truthful.
In point of fact the title is an out-right lie, and the technology is an open system and that, you fucking twats, does not violate the laws of physics. The implications of the title does, but again, the title is a lie.
Every comment here except one or two are a waste of time and life to read. If you are remotely intelligent and manage to get to my comment I urge to to move on to something else and forget this place, it's obvious that YouTube users have made this place their new home.
Kinedryl
baudrunner
Kinedryl
JRi
tpb
Think of the LED as a peltier junction which when used to heat instead of cool, produces more heat on it's hot side than electrical energy put into it. The excess comes form the cold side and ambient.
The LED has a junction like the peltier device and can convert ambient thermal energy into photons instead of heat.
Callippo
antonima
Uh....................WHAT??
That is more than 100% awesome :-P
Grallen
69 picowatts - 30 = 39 picowatts of heat are becoming less entropic by becoming light.
How much more entropic is heat than light? How much more entropic is light than electricity?
They say this continues to scale...
I pose this question: Is there a point where we can choose to make light, but do so without increasing entropy any more than it would have increased if left undisturbed?
This is little more than choosing the type of entropy created, but still intriguing.
Also another open question...
Electricity Heat converted to light then converted to electricity.
Potentially more efficient than: Heat converted to kinetic energy converted to electricity?
wdhowellsr
I'm am certainly not against amazing discoveries that change everything but it always seems that when we try to scale up to the normal world, everything changes.
There is not and will never be a free lunch in Physics.
Callippo
Callippo
5thabove
Time is happening now, Space is a 4th phase in time and the force carrier of gravity and is an element of time and part of energy that makes an atom. Atoms like ANYTHING that oscillate have 4 phases in time. The hole is not only the missing graviton in the S model, its a 4th part of the atom, a 4th part of the observer. U R all real after all!
JFH^^
wiyosaya
"LED's Coefficient of Performance (COP)Exceeds 1".
As I see it, either the author of the article did not understand the process for one reason or another, or the author wanted to be sensational. If the latter, that would be unfortunate, IMHO.
"Over unity" would require that energy is created from nowhere and an accounting of that energy output would reveal that there is energy that that seemingly came for nowhere. The article does not state that.
The article does indicate the the "extra luminosity" is derived from heat. Therefore, energy is conserved, and a detailed, accurate analysis of the system would reveal that all energy in the output can be accounted for from some source in the system.
There is no over unity here - unfortunately.
Rohitasch
James_Mooney
Seriously though - we are converting ambient heat into light - which does seem to go against entropy.
Moebius
If the PV panel is 100% efficient not 50% it sounds like you do have over unity gain.
If you have an individual 100% efficient receptor for each LED you should have a net gain of power and probably a thermocouple.
Szkeptik
So they used the entropy of one part of the system to lower the entropy of the next part of the system? How does that even work?
DirtySquirties
Callippo
paul856
So here's what is going on. The authors ignored the device's blackbody radiation, which at the temperature quoted is indeed more than the light power observed. The device needs only to exhibit a modulation of "blackness" with the applied current to account for the observed behavior.
Case closed.
Callippo
The thermoluminescent pigments glow in visible light, while cooling its environment. The closed system composed of LED does the very same: it glows in visible light, while cooling its environment.
Strictly speaking, only if we would embeds the LED with graphene battery into black box, the energy of light absorbed in its walls would compensate the cooling of battery, so that this device wouldn't violate the thermodynamics in strict sense.
antialias_physorg
The article says that the the LED cools.
Notice the following passages:
and
There is no miracle here. Excess heat is turned into light. In addition to the light produced via the electric current this gives you more light of that frequency than the current would alone.
paul856
I postulate that the authors are measuring blackbody radiation from their device which is modulated by the DC applied.
If their entire measurement system was at the LED temperature, then the LED would absorb as much light from the system as it thermally radiated. Its net light out would not be over unity of electric power applied.
Paul
antialias_physorg
Again:
How is 'excess heat' not a gradient?
No thermodynamic laws are violated anywhere in this.
paul856
I calculated also that the blackbody radiation is indeed greater than the light output observed (making some assumptions of the size of the device; it wasn't exactly specified).
Energy will flow from the hot LED to the colder lab. This is called blackbody radiation, not a hyper efficient LED.
Callippo
DustinS
Has anyone here actually read the article to see if this is correct? (a copy would be nice but not for $30).
If this is correct, then a lot depends on the intensity & wavelength of the emitted photons relative to the LED's blackbody radiation (both of which affect the maximum theoretical solar panel efficiency). Unfortunately, neither the wavelength distribution nor the LED size are mentioned in the summary.
It should be noted that the 2nd law of thermodynamics is an unproven mathematical theorem even for classical mechanics (this was mentioned in my statistical mechanics textbook and the professor put it in a list of things that you would get a Nobel prize for either proving or disproving).
paul856
By the way, I don't know of any proven 2nd law violations, And I have disproven some other claimed violations before.
Paul
Fumfum
2) This is not a "unity" device.
3) However, this is showing efficiency in energy conservation by breaking yet another testing mechanism.