# I Is this another Maxwell's demon?

#### pkv

Summary
What happens when applying a molecular sieve between two chambers?
A though experiment:

We have two gas chambers at the same temperature and pressure, separated by a wall punctuated by very small pores. Let's say small enough to allow the pass of only a few gas molecules at a time (so, no complex interactions between gas molecules inside the pore).

These pores are designed to allow molecules passing from one chamber to the other easy, and hard or very hard in reverse.

This design (for the pore) comes to mind:

Non-imaging optics CEC

Another one could be a version of a Tesla valve, although I am not so sure of it's validity for single molecules.

The question is: would pressure/temperature in both chambers remain stable, or would the gas slowly migrate to the "easy pass" chamber, raising it's pressure?
If the second option happens, the systems seems to me akin to the Maxwell demon, as it violates the second law.
Such a system would allow to extract mechanical energy from the gases without a cold reservoir, until it's kinetic energy is exhausted, so I guess it's impossible.

Where is the catch?

Related Classical Physics News on Phys.org

Staff Emeritus
All that matters is the size of the hole between chambers. Any baffling you put in the chambers is irrelevant.

#### pkv

Yep, that was correct.
I simulated a gas and different connections between chambers and eventually the gas density always equilibrates...
I used this program as base, and added code for vector walls.

Simple passage

Using a CEC

Using a Tesla valve

...didn't see any noticeable difference in any scenario.

An interesting thing I noticed (for any configuration, like the simple passage), for low quantities of particles (<100) you cant get temporal density irregularities, sometimes quite high (like 20 particles in one side, 30 in the other).

That makes me think you can define a different type of Maxwell demon, one which only needs to count how many particles pass in each direction in the passage, and close it just once whenever the densities are different enough.

That would cause different densities in both chambers, but not necessarily different temperatures I guess. In that case, it wouldn't violate 2n law, right?

#### sophiecentaur

Gold Member
Using a Tesla valve
The name "Tesla" seems to grab my attention whenever I see it and it so often is associated with apparent 'Magic'. That valve is only an extension (surely) of the same effect that allows an aerodynamically shaped car body to move forward at a given speed than it can move backwards (with the same mpg). Fluid flow is not perfectly reversible so that 'party trick' valve is not particularly remarkable.
Am I over-sensitive to that guy or did he really promise much more than has been actually achieved in his name?

#### pkv

I was trying to answer the question about if it's possible to build a Maxwell demon only using non-moving parts. A kind of diode device that allows particles to pass on on direction but not in the other, without having any information about such particles (unlike Maxwell's demon).

The only check valve I know using only non-moving parts is the Tesla valve (which is being used in microfluidics). I don't think it's magic, but I was obviously misunderstanding its working principles.

I guess a Tesla valve acts as a diode only with a coherent and dense stream of particles.

#### anorlunda

Mentor
Gold Member
A bit off topic but ...

Am I over-sensitive to that guy or did he really promise much more than has been actually achieved in his name?
Yes and no. Tesla was a true genius, and a prolific inventor. The 3 phase induction motor, the Tesla turbine, and the Tesla Valve are examples of his contributions. He had more than 300 patents

But later in life he became erratic, kooky or worse. His infamously bad idea was wireless power transmission. That made him a cult hero and a scientist's devil forevermore.

Both descriptions of the man are true making him a true enigma.

#### sophiecentaur

Gold Member
He had more than 300 patents
There's nothing about a patent to say a device is of any use or that it will actually work.
a prolific inventor.
. . . and successful self-publicist, I think.
erratic, kooky or worse
He obviously filled a need at that time. I guess that one can't judge him in isolation. But there is a serious lack of numerical evidence to support his ideas. His fans tend to ignore this and they 'all' quote his Wireless Power Transmission idea as if it actually worked. I wonder how he would have fared if he were alive these days.
a true enigma.
Hmm. That term can be interpreted in an over appreciated way. But I guess he was relatively harmless and not many were killed as a result of his efforts.

#### Lord Jestocost

Gold Member
2018 Award
That makes me think you can define a different type of Maxwell demon, one which only needs to count how many particles pass in each direction in the passage, and close it just once whenever the densities are different enough.
In order to count molecules, the demon must be able to see them. He is in a container in equilibrium at constant temperature, where the radiation must be that of a blackbody. As it is impossible to see anything in the interior of a black body (the demon has the same temperature), he needs photons with energy above the thermal background to locate the molecules. Gaining information by this way is thermodynamically costly, so that the second law of thermodynamics cannot be violated on the whole as has been proved by L. Brillouin.

L. Brillouin in “Maxwell's Demon Cannot Operate: Information and Entropy. I” (Journal of Applied Physics 22, 334 (1951)):

In an enclosure at constant temperature, the radiation is that of a blackbody,'' and the demon cannot see the molecules. Hence, he cannot operate the trap door and is unable to violate the second principle. If we introduce a source of light, the demon can see the molecules, but the over‐all balance of entropy is positive.

#### vanhees71

Gold Member
A quantum version of Maxwell's demon has recently been realized, and the famous information theoretical work by Landauer and Szilard has been confirmed (the entropy of a qubit is $k_{\text{B}} \ln 2$:

https://www.pnas.org/content/114/29/7561

"Is this another Maxwell's demon?"

### Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving