TheEEStory.com

The second law of thermodynamics is a statistical 'law'. In my opinion, it is not a fundamental law. If all of the atoms/molecules are moving in the same direction then one could easily construct a device to extract their kinetic energy and turn it into electrical energy. BUT due to the small size and lack of control of the some 10^23 objects their behavior averages out to give you only a mean behavior, a 'normal' distribution (in statistics parlance) or a 'Gaussian' distribution (in physics parlance).

Changing kinetic energy to some other form (such as electric current) doesn't violate energy conservation.

In like manner, I don't view entropy as fundamental. It is only because of the statistical nature of the objects to which it is applied that give it credence. If the marbles of a Chinese checker game are put into a corner of the board are they in a lower entropy state than if they are randomly spread over the board? Every state is just a state. No one of them is more important or special than another one. It is only when one invokes the time dependent process that governs the motion of the particles that some states stand out more than other ones. Change the process and you change the 'entropy' (probability distribution).

---

From the archive paper we find

---

DGDanforth wrote:

The second law of thermodynamics is a statistical 'law'. In my opinion, it is not a fundamental law.

Just because statistics is used does mean it's less fundamental.
Second Law is much much deeper than you think.
It's linked to the foundations of quantum mechanics, which is statistical too, yet you would not call it unfundamental.

If 2nd law is broken quantum mechanics goes down the drain too, time travel and all kind of other esoteric shit will become possible, but you would not be able to appreciate it because one of these "shits" is cessation of intelligent life (any life for that matter) as we know it.

Last edited Sat, 10 Mar 2012, 10:01pm by Y_Po

---

Y_Po wrote:

DGDanforth wrote:

The second law of thermodynamics is a statistical 'law'. In my opinion, it is not a fundamental law.

Just because statistics is used does mean it's less fundamental.
Second Law is much much deeper than you think.
It's linked to the foundations of quantum mechanics, which is statistical too, yet you would not call it unfundamental.

If 2nd law is broken quantum mechanics goes down the drain too, time travel and all kind of other esoteric shit will become possible, but you would not be able to appreciate it because one of these "shits" is cessation of intelligent life (any life for that matter) as we know it.

You take too much for granted.
Is quantum mechanics at its very core random? That is, do things at the, say, Planck length just happen without cause or are there rules that we have not yet discovered that govern the seemingly random quantum behavior?

Is the universe truly 'stochastic' or is it 'deterministic' (which does not say we humans can determine its evolution but rather that there are underlying principles that govern it).

From a practical point of view, just shut up and calculate, right!? Not appealing to me.

A recent paper pointed out the fact that given the inputs and the outputs of a finite state machine one can not determine the internal states for any number of finite measurements AND YET the machine is deterministic.

One could throw up their hands and simply say we will never know the underlying structure of the universe. But as the old joke goes, we may be able to get just close enough!

---

PN wrote:

"Second law of thermodynamics: The entropy of any closed system not in thermal equilibrium almost always increases. Closed systems spontaneously evolve towards thermal equilibrium -- the state of maximum entropy of the system -- in a process known as "thermalization". Equivalently, perpetual motion machines of the second kind are impossible."

This version of the second law is broken by the graphene device. The definition of the closed system allows an initial state of high temperature. The Entropy of the closed system declines and the device does work and its temperature is lowered. However, it is not clear if this statement of the second law is correct. It ignores how the initial conditions got that way. So where you define the system closure occurring is very important. Portions of the second law may yet stand with the closed system definition requiring the inclusion of the initial heat source.

The only sure result is that this law will be argued over for years with so much mis-information and insults that it will make the Bohr - Einstein debate the model of civility.

The wiki statement of the second law fails, but it is not the definitive statement of the second law. A modification in the definition of a closed system may restore it.

Before anyone throws away the 2nd law of thermodynamics it would be a good idea to understand what it really is. The version presented in high-school chemistry lessons is now regarded as a little crude.

It turns out that if you are really determined there are at least two subtle ways of getting heat to flow uphill (from cooler to hotter parts of the system), as well as the obvious ones involving mechanical energy.

The entropy calculations (which can allow heat to flow uphill) have to take into account factors such as correlation and, at a quantum level, entanglement. Of the two effects entanglement seems to have the greater capacity to effect mischief relative to a simple statement of the 2nd law.

For further details see Entanglement and the Thermodynamic Arrow of Time. If anyone understands this paper fully on the first reading, then maybe they could explain it fully to me too. Otherwise I'll have to ask Terry at some point.

In one sense PN is right because the scope of the closed system has to include any part of the universe with which the bit being measure is correlated or entangled. And there's a lot of correlation and entanglement around.

A better way of looking at systems is that entropy is related to information, under the scope of which come correlation or entanglement. Deletion of information (such as during an irreversible rather then reversible process) is what causes the increase in entropy. Quantum systems tend to delete information all the time (for instance as a superposition decoheres into an eigenstate of the system).

A good example of the information concept of entropy is a system in a cylinder divided into two halves by a partition which allow heat to flow and which will move in response to a pressure difference. If I tell you that the two halves of the cylinder are occupied by gas with a specific chemical formula and give you the temperature, pressure and volume for each half then you can do a classical calculation of the entropy for the gas as a whole. You would surmise that if you remove the partition then the entropy remains the same.

If you are now told that the two halves of the same system contain completely different pure atomic isotopes of the gas with the given chemical formula then all of a sudden your view of the entropy is vastly different. But if you are unable to distinguish isotopes with your current technology then that more accurate calculation of entropy is not available to you, and you have no means of telling that the entropy would increase irreversibly if you removed the partition. It is the necessary deletion of the macroscopic information as to where the different isotopes are that cause the entropy to increase.

The net of all this is that that the graphene experiment does not break any law of thermodynamics. It is surely our incomplete understanding of what is going on that leads to a perception that it might.

Regards,
Peter

Last edited Sun, 11 Mar 2012, 7:26am by Technopete

---

DGDanforth wrote:

Y_Po wrote:

DGDanforth wrote:

The second law of thermodynamics is a statistical 'law'. In my opinion, it is not a fundamental law.

Just because statistics is used does mean it's less fundamental.
Second Law is much much deeper than you think.
It's linked to the foundations of quantum mechanics, which is statistical too, yet you would not call it unfundamental.

If 2nd law is broken quantum mechanics goes down the drain too, time travel and all kind of other esoteric shit will become possible, but you would not be able to appreciate it because one of these "shits" is cessation of intelligent life (any life for that matter) as we know it.

You take too much for granted.
Is quantum mechanics at its very core random? That is, do things at the, say, Planck length just happen without cause or are there rules that we have not yet discovered that govern the seemingly random quantum behavior?

Is the universe truly 'stochastic' or is it 'deterministic' (which does not say we humans can determine its evolution but rather that there are underlying principles that govern it).

From a practical point of view, just shut up and calculate, right!? Not appealing to me.

A recent paper pointed out the fact that given the inputs and the outputs of a finite state machine one can not determine the internal states for any number of finite measurements AND YET the machine is deterministic.

One could throw up their hands and simply say we will never know the underlying structure of the universe. But as the old joke goes, we may be able to get just close enough!

I fail to see a point in this random collection of words and sentences.

---

DGD - no-one can eliminate non-local hidden variables theories of QM. So it is possible that some more detailed theory will turn out to be deterministic.

But it is anthropomorphic to expect this to be the case. Determinism, looked at from a many worlds viewpoint, is simply the requirement that the QM worlds that we see should represent the whole universe. No reason to expect this a priori. A many-worlds universe is mathematically deterministic - it just seems non-deterministic because our consciouness records history consistently down a given possible path, so any introspection can only result in consistent measurements with one path, which then appears random.

Y_Po's point is real, and not understood by those who have a naive view of thermodynamics. Increase in entropy is not some macroscopic approximation. At QM level state density determines how likely things are to happen: and this is deeply wired into the universe.

It cannot be got round with Maxwell's demon, as recent work makes clearer the demon must remember the internal state he interferes with, which requires arbitrarily large memory, or erase it to work with finite memory. The erasure takes energy.

Last edited Sun, 11 Mar 2012, 10:20am by ee-tom

---

Y_Po wrote:

Why do they use different electrodes? Silver and Gold?

Asymmetrical electrodes are required because in a thermionic process one of the electrodes has to be the ‘heat source,’ with the other being the ‘heat sink.’

It might be useful to understand the nature of the material that Xu et al. are working with first before continuing on with any discussion about the Second Law of Thermodynamics. Carbon is different than most other materials. See Super-entropic Amorphous Diamond as Thermionic Energy Converters. There is a land in between crystalline diamond consisting only of sp³-bonded carbons (with no distortions) and graphene consisting only of sp²-bonded carbons (with no distortions).

Amorphous diamond is essentially a chaotic carbon mixture with distorted sp2 and sp³ bonds. As such it possesses both metallic character of conductive graphite and semiconductor character of insulating diamond. Moreover, as each carbon atom is unique in its electronic state that is determined by the degree of distortion of its bonds, amorphous diamond contains numerous discrete potential energies for electrons. In fact, amorphous diamond may have the highest density of atoms (1.8 × 10²³ per cubic centimeter) that is several times higher than ordinary materials (e.g. about four times of iron atoms or silicon atoms). Thus, amorphous diamond has the highest configuration entropy for both atoms and valence electrons.

Due to the distribution of discrete electronic energies with high density, amorphous diamond is uniquely capable to generate electricity and emit radiation. It has been demonstrated that amorphous diamond can be made as silicon free solar cells, front panel display field emission source, sensitive thermal sensing by IR detection, and perfect black body for energy conversion. Various amorphous diamond devices are being fabricated to exploit the superb properties of amorphous diamond.

Technopete wrote:

The net of all this is that that the graphene experiment does not break any law of thermodynamics. It is surely our incomplete understanding of what is going on that leads to a perception that it might.

Indeed, from my very brief reading of the topic-featured article, these workers appear to have found that that the interaction (and the presumed absorption of ions from) some ionic solutions alter the work function of their carbon material more than others. Thermionic carbon has been known about for some time. Just Google ‘thermionic carbon’ if you don’t believe me.

---

I thought that energy could only be extracted from temperature differences, and that it extracting energy from cooling the environment was a no-go.

24 Comments are up at the original article http://physicsworld.com/cws/article/news/48889#...

5 of the comments are by the lead author of the paper.

A new/expanded version of the Arxiv paper should be available on Monday according to a comment by the lead author.

Zihan Xu said

The power density I tested one sample with Graphite-Siliver electrodes is about 70kW/kg. The system was just kept still in the lab at room temperature. It is not easy to test if it has cooling effect. I have designed experiments, and working on it.

Theoretically, if you can have enough device, you can get enough electricty to power a Commercial device.

The blind experimental and control experimental results are in the supporet material, which will be ready this monday when Arxiv back to work."

---

70kW/KG! Wow. Too bad graphene currently costs billions per KG.

omni wrote:

70kW/KG! Wow. Too bad graphene currently costs billions per KG.

At that price, give me a stock of 4B pencils and Sellotape and I'll make you some (maybe not a Kg). Sellotape can be used to make X-rays too, so you can have some of them thrown in during the process.

Regards,
Peter

---

DGDanforth wrote:

The second law of thermodynamics is a statistical 'law'. In my opinion, it is not a fundamental law. <snip>

Changing kinetic energy to some other form (such as electric current) doesn't violate energy conservation.

In like manner, I don't view entropy as fundamental. It is only because of the statistical nature of the objects to which it is applied that give it credence.

DGD, thank you for raising the level of this discussion. IMHO you have a robust argument.

This seems to be the same argument, or at least one similar to, the argument that a device which would function as a Maxwell's Demon may actually be possible; that it may be possible to construct a gate which will pass only molecules whose Brownian motion (and therefore temperature) is either greater than or less than a certain limit.

But I've recently read a persuasive argument that the Maxwell's Demon gate would fail, because the control mechanism (a ratchet, for example) of the gate would itself be subject to Brownian motion, defeating the ability of the gate to separate faster-moving molecules from slower-moving ones.

All this is, of course, highly theoretical. The reason I think that we will find it difficult or impossible to take advantage of the statistical nature of thermodynamics-- why we will likely never invent a practical "free energy" device or process, nor a functional analog of a Maxwell's Demon-- is because nature has never invented one. Nature and evolution have had billions of years to invent more efficient biological processes. We are only now beginning to understand how, for example, photosynthesis achieves such an astonishingly high efficiency, on the quantum level, at converting solar energy to chemical energy.

Yet no organism we have discovered runs its metabolism on "free energy"; none has dispensed with the need for a constant input of energy, either chemical energy (food) or sunlight. If Nature had ever invented a quantum method of harvesting useful work from ambient heat, the organism in which it first appeared would have easily out-competed every other type; there would be no life form left in the world except descendents of that one lifeform!

Now, I don't consider this "proof". After all, humans have invented things Nature never did. No organism uses radio for communication; Nature never invented the transistor; no critter's nervous system uses digital information processing or storage.

But for something as basic as finding a quantum method of harvesting energy from ambient heat, or Brownian motion, or any sort of random quantum process-- energy which can be use to perform work, or decrease entropy-- I think if it were possible, Nature would have already found a way.

Again, I don't consider this proof that we'll never be able to take advantage of the statistical nature of thermodynamics to "cheat" entropy-- I just find it highly unlikely that we ever will.

DGDanforth wrote:

If the marbles of a Chinese checker game are put into a corner of the board are they in a lower entropy state than if they are randomly spread over the board? Every state is just a state. No one of them is more important or special than another one. It is only when one invokes the time dependent process that governs the motion of the particles that some states stand out more than other ones. Change the process and you change the 'entropy' (probability distribution).

I don't think marbles on a Chinese Checkers board is a good analogy of physical processes on the molecular level. Moving marbles one by one into one corner of the board does not increase the energy necessary to move the next marble to the same corner. Yet if you were to try the same thing with molecules of gas in an enclosed container, trying to move them all into one corner, you'd find that the more molecules you moved into that corner, the higher the temperature in that corner would become. Each molecule moved there would increase the amount of energy necessary to move the next molecule there.

Last edited Sun, 11 Mar 2012, 9:21pm by Lensman

---

omni wrote:

70kW/KG! Wow. Too bad graphene currently costs billions per KG.

Not for long :-)

---

Looks like about USD $50,000/KG right now for Graphene.

http://www.cheaptubes.com/carbon-nanotubes-pric...

And FarmEER's link shows Graphene nano powder for about $20/gram. So looks like Graphene is plummeting in price. also, it is very light and not much is needed for a cell.

seslaprime wrote:

And FarmEER's link shows Graphene nano powder for about $20/gram. So looks like Graphene is plummeting in price. also, it is very light and not much is needed for a cell.

The stuff they are selling there still has to be reduced. Can't be used as is.

Ready to use graphene sheets are still going for about $100 per square inch of .142 nm thick sheets. 30 million sheets for 1 cm thick stack. Specific gravity of 1.7. Do the math. Billions per kilo.

Graphene cheap enough to use for anything but a science experiment is still 10 to 15 years away.

omni wrote:

seslaprime wrote:

And FarmEER's link shows Graphene nano powder for about $20/gram. So looks like Graphene is plummeting in price. also, it is very light and not much is needed for a cell.

The stuff they are selling there still has to be reduced. Can't be used as is.

Ready to use graphene sheets are still going for about $100 per square inch of .142 nm thick sheets. 30 million sheets for 1 cm thick stack. Specific gravity of 1.7. Do the math. Billions per kilo.

Graphene cheap enough to use for anything but a science experiment is still 10 to 15 years away.

Interesting, do you work in the industry?

---

I am glad to hear omni's estimate of cheap graphene as just 10 to 15 years away. I would guess shorter for high value applications and longer for low value applications.

High temperature superconductors were discovered in the mid 80's. It is only after 25 years of work that High Temperature superconductor wire is reaching a price where it might be usable in a product. Probably another 10-20 years before it is used in all the big motors and generators.

This power generating discovery will have a similar development path but it is drawing on an vastly improved technology base when compared to the mid 80's.

Predicting progress in a technology development with deadlines is very hard (Actually its a job for a fool since it is impossible). Ask any project manager, or use DW as an example.

The best I can predict here is its likely to be a 20 year slog to get any graphene generator to market. Minus 10 to plus 20 years as 1 sigma estimates on the probability of 20 years being reached. (And for the nit pickers, its not a Gaussian distribution)

My other prediction is that the argument over the second law will follow a similar path. In this case, the geezers have to die first. However, I do agree that this device, if proven to work, has grave implications for the foundations of Quantum Mechanics. As far as I am concerned, Mead destroyed the foundation 10 years ago. But we all know that I am considered a Quack for suggesting this. The last laugh will be a robust one.

---

DGDanforth wrote:

The second law of thermodynamics is a statistical 'law'. In my opinion, it is not a fundamental law. If all of the atoms/molecules are moving in the same direction then one could easily construct a device to extract their kinetic energy and turn it into electrical energy. [...]

I don't really like the 2nd law, either. Liouville's theorem of Hamiltonian mechanics is fundamental, though, and it can be used to prove the impossibility of devices like this.

After all, what makes an energy source _useful_ is not just that it releases energy, but that it does so in a way that lets you harness that energy to perform whatever useful work you like, like running a compressor.

Doing such useful work takes the environment from a general state (specifically, any state selected from a large volume of phase space) to a more specific state (selected from a smaller volume of phase space). That is to say that useful work transforms something from whatever it is to specifically what you want.

Now, Liouville's theorem says that systems do not evolve this way. The volume of phase space that bounds the state of a system is constant over time. If you want to set up a system, part of which will go from a more general to a more specific state, then you have to allow another part of the system to go from a more specific to a more general state -- the electrons in a discharging capacitor for example.

Perpetual motion machines that violate the 2nd law would typically allow you to construct an isolated system that evolves from a more general to a more specific state. Such machines are therefore impossible by Liouville's theorem according to the argument above.

This self-charging battery seems to be just such a machine.

Last edited Mon, 12 Mar 2012, 11:52am by matt

So matt, if the self-charging battery works - you have proven that Liouville's theorem is wrong.

Now I have never really studied the Hamiltonian, but I suppose that if Liouville is wrong, the Hamiltonian is wrong too.

So Classical and Quantum mechanics get thrown out the window if the self-charging battery works. Since the second law is Classical - there goes the neighborhood.

Is this getting to be fun!

The reformulation of everything in terms of Wave Function Coupling - will work and produce surprises!

---

omni wrote:

seslaprime wrote:

And FarmEER's link shows Graphene nano powder for about $20/gram. So looks like Graphene is plummeting in price. also, it is very light and not much is needed for a cell.

The stuff they are selling there still has to be reduced. Can't be used as is.

Ready to use graphene sheets are still going for about $100 per square inch of .142 nm thick sheets. 30 million sheets for 1 cm thick stack. Specific gravity of 1.7. Do the math. Billions per kilo.

Graphene cheap enough to use for anything but a science experiment is still 10 to 15 years away.

I would say more like 2-5 years away. The technology to mass produce Graphene already exists. Also, The Carbon nano-tube technology has matured. (nano-tube is graphene sheets rolled into tubes). you can purchase industrial level nano-tubes right now in bulk at Cheaptubes.com.

There is a growing demand for Graphene world wide. there are hundreds of applications and today would be a trillion dollar industry if the bulk product were available.

I remember when the Nano-tubes were invented just a few years ago. it was MIT if I recall. the same was said that it would be 20 years before they would be mass produced.

well look at them now. cheaptubes.com.

Whenever there is a fortune involved, industry seems to move at a more rapid pace.

this article shows the Korean's already scaling up a system created in Texas to mass produce large sheets of graphene for use in touch-screen displays.

http://news.sciencemag.org/sciencenow/2010/06/g...

I think, Like nano-tubes, Graphene will be mass produced at Cheap prices within a few years.

Graphene Nanoplatelets (GNPs) Prices
GNPs Grade 3 is Industrial Grade

GNPs Grade 4 is Research Grade

Quantity ------- GNPs ----- GNPs
----------------- Grade 3 -- Grade 4
01-49 grams -- $15.0/g - $70/g
50-99 grams -- $2.50/g - $60/g
500-999 gms -- $1.25/g - $20/g
1 Kilogram ---- $450/KG - $4,500/KG

Read more: http://www.cheaptubes.com/graphene-nanoplatelet...

seslaprime wrote:

...well look at them now. cheaptubes.com.

Whenever there is a fortune involved, industry seems to move at a more rapid pace.

---

Self charging battery exists already today. A solar array feeding a standard battery or capacitor. There must be some physics law this violates.

dvelasco68 wrote:

Graphene Nanoplatelets (GNPs) Prices
GNPs Grade 3 is Industrial Grade

GNPs Grade 4 is Research Grade

Quantity ------- GNPs ----- GNPs
----------------- Grade 3 -- Grade 4
01-49 grams -- $15.0/g - $70/g
50-99 grams -- $2.50/g - $60/g
500-999 gms -- $1.25/g - $20/g
1 Kilogram ---- $450/KG - $4,500/KG

Read more: http://www.cheaptubes.com/graphene-nanoplatelet...

seslaprime wrote:

...well look at them now. cheaptubes.com.

Whenever there is a fortune involved, industry seems to move at a more rapid pace.

yes, I read the whole site.

farmEEr wrote:

omni wrote:

70kW/KG! Wow. Too bad graphene currently costs billions per KG.

Not for long :-)

There may be commercial suppliers of graphene but the prices are still very high. I really have not seen anyone claiming they can produce graphene films in a low cost manner and in an easily manufactured format. By it's very nature the films are extremely fragile. This is the reason that most films are processed on very costly substrates. If they can't produce this on a low cost substrate then this is purely a science fair experiment and nothing more. A easy way to produce expansive sheets of monolayer films is required and this is difficult to do. Sure I could some up with a cool battery IF...

The devil is in the details and this one manufacturing issues is a killer as far as I'm concerned.

---

Prof Neilson wrote:

My other prediction is that the argument over the second law will follow a similar path. In this case, the geezers have to die first.

It's not us "geezers" who are preventing perpetual motion machines from working. It's the nature of the universe, and that is not likely to magically change just because you want it to. No more than it has for previous generations just because they wanted it to, and they wanted it every bit as badly as you.

---

jnissen wrote:

farmEEr wrote:

omni wrote:

70kW/KG! Wow. Too bad graphene currently costs billions per KG.

Not for long :-)

There may be commercial suppliers of graphene but the prices are still very high. I really have not seen anyone claiming they can produce graphene films in a low cost manner and in an easily manufactured format. By it's very nature the films are extremely fragile. This is the reason that most films are processed on very costly substrates. If they can't produce this on a low cost substrate then this is purely a science fair experiment and nothing more. A easy way to produce expansive sheets of monolayer films is required and this is difficult to do. Sure I could some up with a cool battery IF...

The devil is in the details and this one manufacturing issues is a killer as far as I'm concerned.

most likely because you have not looked. in fact, there are several patents for producing mass quantities of large graphene sheets at low cost. the technology exists today.

I suspect that as we speak, methods are scaling up and within a year or 2, we will see Graphene being produced in quantity and price.

Everyone is wanting Graphene right now. the Industry is tooling up to switch from Silicon to Graphene. iridium is targeted to be replaced with graphene. everywhere we look in electronics industry there is application to enhance or outright replace with Graphene. it is a potential Trillion Dollar industry.

as I said, like the carbon nano-tube industry that was suppose to take 20-30 years to take off only took a few years. When there is vast riches involved, the industry hath no boundaries.

So Lens

If the graphene battery is shown to work - what is the hypothetical?

Hari-Kari?

Moving to another dimension where things work the Lensman way?

---