"entropy is disorder" = outmoded theory??

Ray Eston Smith Jr

The wikipedia article I quote below is confusing me. I followed the links to Frank Lambert's website, where he claims that Peter Atkins, in the 8th edition of his Physical Chemistry, has come around to Lambert's idea that entropy is not related to disorder. Could this be true? I would be less surprised to learn that Atkins had become a born-again Christian. However, according to my old-fashioned understanding of entropy, even very improbable things happen occasionally, given a sufficiently huge number of trials. Seriously, I was about to start reading the 7th edition of Atkins Physical Chemistry, but now I'm wondering if I should look for a more reliable author.

http://en.wikipedia.org/wiki/Introduction_to_entropy
Traditionally, 20th century textbooks have introduced entropy as order and disorder so that it provides "a measurement of the disorder or randomness of a system". It has been argued that ambiguities in the terms used (such as "disorder" and "chaos") contribute to widespread confusion and can hinder comprehension of entropy for most students. A more recent formulation associated with Frank L. Lambert describing entropy as energy dispersal describes entropy as measuring "the spontaneous dispersal of energy — at a specific temperature."

http://entropysite.oxy.edu/
March 2006
Atkins' "Physical Chemistry" has been the best selling text worldwide in this subject for many years.* The new 8th edition was published in the US March 16. However, in previous editions Atkins described systems and entropy change in terms of order and disorder or chaos.* Even though he long has used the phrase, "the dispersal of energy", it was confined to an order-disorder view of thermodynamics — for example, to spontaneous changes being "always accompanied by a dispersal of energy into a more disordered state".** (Or "to chaos" in his book "The Second Law".)
In contrast to the Second Law chapter of the 7th edition, which had some 27 instances of using "order to disorder" as a rationale for change, "disorder" and "disorderly" are mentioned only 3 times in the new 8th edition.* Atkins, with co-author dePaula, now state that their view of entropy "summarized by the Boltzmann formula is consistent with our previous statement [earlier in the chapter, re the dispersal of energy in classical thermodynamics] that the entropy is related to the dispersal of energy.

brainstorm

How is dispersal of energy an explanation for why a drop of dye disseminates into a glass of water? Isn't there just a certain amount of molecular motion due to heat-energy, which causes particles to tend to move around randomly such that the dye particles are likely to randomly traverse the particles of clear water and vice versa? Is that energy dispersion or just expression of energy as molecular motion?

Naty1

Originators of new ideas are always claiming support from others...who knows what Atkins believes and whether that REALLY meshes with Lambert's idea.

I would not disregard a proven text because of such a claim. I've never even heard of "dispersal" of energy nor do I know whether Atkins/Lambert agree on what it is.

Entropy is a difficult enough concept that if someone has a better explanation, I'd like to read it.

brainstorm

Energy "dispersal" is a generally logical concept. It is what happens when a cue-ball breaks a triangle of billiard balls. If you have a room of 40 degree air and you make a fire, convection is the result of high KE among air molecules next to the fire dispersing energy to other molecules they come in contact with. Eventually, the system temperature will reach equilibrium, except the heated air will be pushed up by the denser cool air, which will grow as parts of the warm air disperse their energy through the windows and walls and subsequently sink.

Dispersal is just a pattern of transfer between high-energy particles and lower-energy ones. It's empirical common-sense, no?

Gerenuk

Actually entropy has a one-to-one relation to the probability of that particular state being a final state after a "sufficient amount of time" has passed. The relation is a very steep exponential so that slightly higher entropy is extraordinarily favoured by probability.
But still lower entropy states do have a finite probability and will occur.

Andy Resnick

There are several potentially confusing things in those sites. First, entropy relates to a given state, and so cannot be used to describe a process ('dispersal' of energy). Second, there seems to be loose control over discussion of entropy and changes to entropy. Third, 'entropy', like 'energy' is not that well defined outside of quantitative relationships.

Personally, I see the entropy of a state as that amount of energy unavailable to perform useful work. The free energy is the total amount of energy available to perform useful work.

alxm

I wouldn't say it's an "outmoded theory". It's not a theory and never was. It's just an interpretation that helps explain it. As such it may be an outmoded way of explaining entropy.

Equating entropy with disorder only makes sense in a few cases. Gas is more 'disordered' than liquid, both are more disordered than a solid. Mixtures are more disordered than pure substances.

But beyond that, equating entropy with disorder doesn't have much use. Another, more general definition of entropy is "energy not available to perform work". How this relates to entropy being 'disorder' isn't obvious. (In fact, I'd say that once you reconcile these views, that's when you 'get' entropy)

James A. Putnam

There are relationships between disorder, unavailable energy, and thermodynamic entropy. However, relationships do not appear to me to serve as tight enough explanations about: What is thermodynamic entropy as discovered and defined by Clausius.

If thermodynamic entropy is a measure of un-usable energy or a measure of disorder then, why are equilibrium conditions necessary to calculate it? In other words: What is the relationship between absorbing and expelling heat under equilibrium conditions, clearly an ideal condition, but, still necessary in the original macroscopic thermodynamic definition, and the amount of heat that leaks away?

There has already been a very detailed excellent discussion about the development of thermodynamic entropy in another thread. My intention in making this post is only to register an opinion and not to reopen the history of thermodynamic entropy. Even my question is posed without expecting an answer. It is intended to highlight something that I think is of significant importance, and, that I think is passed over in explanations.

James

brainstorm

Heat can only "leak away" as long as there is something in the system that is not in equilibrium with whatever is leaking the heat.

James A. Putnam

Nevertheless, that is a condition of the original definition. Is your point that the original definition is inherently mistaken and can therefore be disregarded in favor of other definitions that do not have to take into account that which Clausius discovered?

Naty1

Whoa!!!!!!!!!!!!!!!!!!
entropy as a measure of disorder has WIDE application....to the evolution of the entire universe, for example, as long as one understands the effects of gravity on entropy disorder.

Entropy is a subset of information theory. and is extremely useful in understanding black hole event horizons. and is linked to decoherence in quantum mechanics which like entropy lets you know which way time is running. The idea that strings in string theory have entropy goes back to the earliest days of string theory.

brainstorm

The genealogy and hermeneutics of "the original definition" may be of consequence if your interest is the history of science, but empirical logic makes it clear that heat can't go anywhere unless there is something in disequilibrium with something else. The fact that you define the system in exclusion of the thing that is in disequilibrium with it is a methodological failure to include everything in the system that is influencing the parts of it you are studying.

James A. Putnam

Clausius defined the system.

brainstorm

I don't know which system you're saying Clausius defined, but I doubt it is any specific system that you have access to study empirically. You have to apply the methodology to test it.

James A. Putnam

The system is an irreversible carnot cycle.

brainstorm

In other words, theoretical rather than empirical.

James A. Putnam

Of course. So is a frictionless surface. Still, the question remains. Is your point that the original definition is inherently mistaken and can therefore be disregarded in favor of other definitions that do not have to take into account that which Clausius discovered? This is not my work. Clausius discovered thermodynamic entropy.