The photon, time and entropy

[Curious45]

The photon, "time" and entropy

(ignore original title, in fact id appreciate if a mod would change it. It was not my intention to have a conversation about time, lol!)

Can we view a closed system of just photons, as ever, or usually, being subject to entropy (defined by the whole 2nd law, rather than merely just quantum statistical mechanics)

1. Say for example, the light from the big bang at the "edge of the universe", traveling in "vacuum"

2. Or a laser that we point out into space that does not encounter any objects or mass-bearing particles?

I have heard it said that photons do not naturally decay on their own (not sure if that's right)?

And that redshift or blueshift is not a loss in energy, that light/the photon generally does not lose energy (not sure if that's right)?

Light travels in a straight line (well a waveshaped one), so without any interaction with a mass bearing object or particle, a group of photons, it would not lose or gain order?

(Although if the photons were heading in different directions, they would get larger/more distributed, but retain the same shape I am thinking, so not more statistically distributed)

Is it appropriate to think of photons as a system? (I believe they can interact, even if they can't collide?)

Are there any systems that do not have entropy? (I guess the center of a black hole might be one, although its not closed, so that's not the best example - perhaps if a black hole tore away from all other mass during expansion?)

...

Oh, and one last final question, if the photon is massless, how does it push a solar sail, or light windmill?

 

[Dale]

That isn't a good definition of time. You can easily think of clocks with no change in entropy. You might use entropy to define which direction in time is future or past, but most clocks don't use entropy to actually measure time.

You call this a frequent definition, but I have never seen it.

 

[Curious45]

That isn't a good definition of time. You can easily think of clocks with no change in entropy. You might use entropy to define which direction in time is future or past, but most clocks don't use entropy to actually measure time.

You call this a frequent definition, but I have never seen it.

What kind of clock doesn't experience loss of energy or heat, or increasing disorder in its time measuring process? At least all the ones I have seen require batteries, or winding up etc, motions that result in direct loss of energy, increase in disorder, and very much to rely on entropy to measure time.

Frequent or not, at minimum, I have not heard any more accurate definitions. I tend to think it is a very excellent definition of time, and its "arrow".

Hopefully someone will come along later and answer my questions ;)

 

[Dale]

What kind of clock doesn't experience loss of energy or heat, or increasing disorder in its time measuring process?

A gyroscope in space, for example. Many other clocks do not increase their own entropy while functioning as a clock, although they may certainly may require an input of energy and therefore increase the total entropy of the universe. E.g. An atomic clocks entropy remains constant while it is functioning.

In fact, I cannot think of a single clock that actually uses changes in entropy as it's operating principle.

 

[1977ub]

Entropy is statistical, isn't it? Thus the entropy of a small system (not in isolation from larger systems) may in fact decrease. To the extent that entropy does increase, there's no guarantee regarding the rate at which it will do so in the small system, right?

 

[Curious45]

A gyroscope in space, for example. Many other clocks do not increase their own entropy while functioning as a clock, although they may certainly may require an input of energy and therefore increase the total entropy of the universe. E.g. An atomic clocks entropy remains constant while it is functioning.

Well, hopefully this dialogue doesn't detract from the original question.

You might have to explain to me how an atomic clock does not rely on, or involve, entropy.

I also don't actually see how a gyroscope in space measures time, via its motion.

If you are using only statistical mechanics to define entropy, I think you would still see that a time keeping system should approach entropy over time, and that it depends on entropy. Even though a gyroscope doesn't in itself measure time (looking as it, in random intervals), won't in itself tell you how much time has passed...how is it set in motion? Would it not lose energy even as an isolated system, albiet very slowly? An atomic clock is hardly a perpetual motion machine either.

But the 2nd law defines entropy in a broader way, also referring to chemical, thermal, radiative and mechanical equilibrium.

"In a state of thermodynamic equilibrium, there are no net flows of matter or of energy, no phase changes, and no unbalanced potentials (or driving forces), within the system."

If you can find a clock, that operates from a state of total chemical, thermal, radiative, mechanical and statistical and energetic entropy, I would be most impressed :)

 

[Dale]

You might have to explain to me how an atomic clock does not rely on, or involve, entropy - loss of heat, order or energy.

They measure the radiation put out by an electron when it changes energy state. Don't electrons involve entropy when they do this? Or is it zero point or free energy that creates the microwaves? Doesn't that increase disorder anyway, via the microwave? Atomic clocks also rely on heat, don't they?

You seem to be confusing energy with entropy. Energy is continuously added to an atomic clock largely to keep the entropy of the atoms constant.

Actually, now that I think of it, the same is true of most clock mechanisms which require input energy. The purpose of the input energy is to keep the state of the clock constant against entropy. E.g. In a pendulum clock the input energy works to overcome friction, the source of entropy in the system, if you reduce the frictional entropy then you reduce the required energy input.

I also don't actually see how a gyroscope in space measures time, via its motion.

We could look at its position, momentum, at random intervals separated across time, and we would have no idea how much time has passed, or what relationship the gyroscope has to time (Unless we looked at its atomic structure and decay etc). You might have to explain that one to me as well.

Are you kidding? This is the original clock, it measures time in units of days.

Maybe I am missing what you are saying, but I am having trouble following the connections you are implying exist between zero entropy exchanges and time measurement.

Then let's try it the other way. Please provide a mainstream scientific reference which defines time the way you suggest.

 

[Curious45]

"You seem to be confusing energy with entropy. Energy is continuously added to an atomic clock largely to keep the entropy of the atoms constant."

Which according to the 2nd law of thermodynamics increases energy equilibrium over time, as defined within thermodynamic equilibrium, no?

http://en.wikipedia.org/wiki/Thermodynamic_equilibrium

It seems a bit more like it is you, who is relying only on perhaps quantum statistical mechanics (which should actually still provide an increase in entropy via these exchanges anyway, and this is actually explanatory for the law, it is not the experimentally proven law itself), and leaving the broader definition within the 2nd law, of chemical, mechanical, energetic etc out, from my POV.

You can't leave an energy source conveniently out of a closed system, otherwise it wouldn't be closed (or mechanical motion for that matter). You certainly can't leave it out of the 2nd law, that's all explicitly included.

"Then let's try it the other way. Please provide a mainstream scientific reference which defines time the way you suggest"

This is one that is proposed in a number of papers I have read, and is mentioned in the book "arrow of time", by Roger Highfield and Peter Coveney, not just referring to "the arrow" but to a definition of time itself.

If you don't like the idea of this defination, don't use it.

Any widely accepted law or property that applies to time, such as the thermodynamic arrow of time, or lorentz tranformation, defines it. Look on wikipedia under time, and youll find the arrow of time.

Einstein says time is "A progressive change in a physical system". I don't think anyone would deny this. It's the whole premise of how clocks work, its utterly fundamental.

Add to that the thermodynamic arrow of time, the 2nd law and we get -

"A progressive change in a physical system"
+
"In an isolated system, entropy tends to increase with time (defined as by statistical mechanics, and particularly thermodynamic equilibrium)" (Ie time assymetry)
=
"Time is a progressive change in a physical system, and is assymetric - in an isolated system, entropy tends to increase with time, as defined by the 2nd law of thermodynamics and thermodynamic equilibrium"

Perhaps I put it very ineloquently, with my short phrase but...its not exactly the point in this thread anyway. I was just explaining my intellectual interest in the topic. Perhaps I should have left that out, and just asked the question...seems to have causes major distraction from the actual topic.

I am only interested in roughly "closed photon systems" or its closest equivilant and whether we can apply "entropy" as a concept to them, either generally, or in specific instances. Or, more practically, a closed system of photons in a "vacuum". If any of those ideas are even meaningful in terms of what we know.

It would be great if we could get back to the actual question, regarding photons and entropy! :)

 

[Dale]

Well the planets and the sun are hardly exceptions to entropy either,

I never said they were. I am not claiming any exceptions to the second law of thermo, I am just claiming that most clocks do not use the second law as their operating principle to measure time. If time were an increase in entropy then devices which purport to measure time would actually measure the increase in entropy of some system. That simply isn't the case.

This is one that is proposed in a number of papers I have read, and is mentioned in the book "arrow of time", by Roger Highfield and Peter Coveney, not just referring to "the arrow" but to a defination of time itself.

That book is not a mainstream scientific reference, it is a pop sci book. I would be interested in any of the number of papers you refer to.

 

[Curious45]

"If time were an increase in entropy then devices which purport to measure time would actually measure the increase in entropy of some system. "

Its clearly easier to use a well controlled/regular, very local, entropic process such as is found in clocks than run around measuring total entropy.

Your also presuming that clocks are the most accurate and direct way of measuring time (not saying you can't measure time by clocks at all, but id argue we have no idea if its the most direct or accurate or not..)

Without a scientific definition of time, a proper mechanical understanding with empirical proof, that is not in such a defination ciricularly_dependant on clocks!_, how can that assumption be tested? At the moment we base our objective sense of time on highly regular changes of a subatomic nature, but there is nothing to say that is the most accurate or direct way of measuring time either.

We can measure the mass of the moon, based on its gravitational effects, but that does not mean that mass and gravitation are identical, or that we are directly measuring mass to achieve this figure.

I might get back to your reply here later Dalespam, ATM, I am concerned its detracting from my core question.

 

[chill_factor]

do refrigerators make time go backwards in their interiors?

 

[Curious45]

do refrigerators make time go backwards in their interiors?

Short answer, of course not.

Longer answer:

1. Not a closed system. Google the 2nd law of thermodynamics.

2. Refrigerators are obviously highly entropic, they constantly lose temperature and burn energy (as well as their chemical entropy).

3. Entropy doesn't stop for the food, it merely slows a bit, even if we ignore the whole system (contrary to the 2nd law) and it certainly doesn't reverse, or "go backwards".

 

[chill_factor]

An ideal photon gas in thermal equilibrium in a closed cavity indeed does not lose energy, just as an ideal mass on a spring oscillating in space doesn't lose energy. However that is not a physical situation - what's the cavity itself in thermal equilibrium with?

Redshift indeed does make photons lose energy. E = hf where f is frequency. lower frequency, lower energy.

A point source emitting EM waves will indeed introduce higher entropy; energy is conserved in a shell around the source, but for each unit area, the further from the source, the lower the energy. Thus, energy is being spread out and higher entropy results.

 

[Curious45]

Thats it, enough stupid distractions from my question, I am re-editing my original post. I am tempted just to start a new thread, if people keep obsessing about time - no one has answered even ONE of my questions regarding entropy or photons.

 

[Curious45]

An ideal photon gas in thermal equilibrium in a closed cavity indeed does not lose energy, just as an ideal mass on a spring oscillating in space doesn't lose energy. However that is not a physical situation - what's the cavity itself in thermal equilibrium with?

Redshift indeed does make photons lose energy. E = hf where f is frequency. lower frequency, lower energy.

A point source emitting EM waves will indeed introduce higher entropy; energy is conserved in a shell around the source, but for each unit area, the further from the source, the lower the energy. Thus, energy is being spread out and higher entropy results.

Thanks so much for your actual answer. I appreciate this :)

Would an ideal mass on a spring oscillating in a vacuum not fall subject to its own centre of mass gravity? (ignore this if you like, its not important)

Interesting about redshift.

I am a bit confused as to how redshift happens if I am honest. Its not like the doppler effect is a good analogy for me, as sound has a carrier, and you can compress the wave of something with a strict relative speed - something that always hits you at the same speed, that's much more confusing to me...i guess the frequency of the wave must shift via some other means?

This last part, is this due to merely "spread"? Or individual photon energy too? What if it was laser light?

Would a universe containing only light, reach a total thermodynamic equilibrium? (Just as a though experiment)

 

[chill_factor]

Interesting about redshift. I am a bit confused as to how redshift happens if I am honest. Its not like the doppler effect is a good analogy for me, as sound has a carrier, and you can compress the wave of something with a strict relative speed - something that always hits you at the same speed, that's more confusing.

This last part, is this due to merely "spread"? What if it was laser light?

Lasers also spread, they just spread out less.

Redshift is easy to understand; you can change frequency in ways other than changing speed. For example, you can change wavelength.

A system of photons will usually increase in entropy for all but the most ideal cases.

Basically, nothing escapes entropy *when there is a mechanism for dissipation*. That is important. Masses on an ideal spring in outer space will not increase in entropy. But basically everything in real life does. Even the solar system will tend to increase in entropy, just that the time scale is literally astronomical. For example, just from classical gravity and treating planets and stars as particles, you'll find that galaxies will eventually throw off most of their stars into intergalactic space. However, there is an important catch: you have to have an entire galaxy, since then you have enough particles to do thermodynamics. With only 1 or 2 things, you can't do thermodynamics.

 

[Dale]

Curious45, Since you have been unable to produce a scientific reference defining time as you claim, this thread is locked. Feel free to PM me with a reference and I will reopen it.