Does classical statistical physics predict newer things vs. thermodynamics?

[Aidyan]

I'm wondering if the passage from a classical thermodynamic theory, i.e. which does not resort to an atomistic theory and methods of probability and statistics, to classical (i.e. non-quantum) statistical mechanics, led to new discoveries and especially if it was able to explain properties of matter that in the conventional thermodynamic theory were previously not explainable? If so which?

 

[vanhees71]

I think classical statistical mechanics as a model to explain the phenomenological macroscopic thermodynamical laws in fact revealed more about micro than macro physics, but it also helped to explain the thermodynamical laws (including the 0th-3rd fundamental laws) from more fundamental laws of classical dynamics by statistical means.

For me the very immediate discovery is the necessity to assume the indistinguishability of particles, because otherwise entropy turns out to be not an extensive quantity as it should be (Gibbs's paradox).

Also statistical physics and thermodynamics helped to discover that the classical models are all inadequate to understand matter on a fundamental level and that quantum theory is needed. Of course, that's true from the very birthday of quantum theory, which is December 14, 1900 when Planck introduced the quantization of the exchange energy of electromagnetic waves with matter to explain the black-body spectrum, which he discovered before by analyzing the high-precision measurements by Rubens and Kurlbaum at the Phyikalisch-Technische Reichsanstalt. This was famously further worked out by Einstein, leading to the idea of "wave-particle duality" for em. waves and later also of "particles" like the electron, which finally lead to the development of modern quantum theory, and only with quantum theory all the fundamental laws of thermodynamics, particularly the third Law (Nernst's Law) can be explained.

 

[marcusl]

There are also specific problems that can only be solved statistically. For instance, if the motion of gas molecules is random, why doesn't the air in the room you are sitting in occasionally and randomly move to the opposite side, leaving you gasping? Stat mech gives the answer in terms of numbers of microstates, which lead to probabilities. The probability of that happening is something like 10^-80, which is so stupendously small that we needen't worry--we can "breathe easy." (If we checked the air every microsecond during the 4.5 billion year age of earth, that's still only 10^23 opportunities for a suffocation event. Checking a million rooms on a million earth-like planets gets us up only to 10^34. We'd need to wait another 10^46 microseconds or another 10^32 years to expect to see one event. We can say that the probability of occurrence is indistinguishable from impossible.)
Similar considerations describe the reasoning behind the 2nd law of thermodynamics and the arrow of time. The 2nd law isn't actually a law that entropy can never increase, but a statement of improbability that involves numbers so fantastically small that it might as well be a certainty. As for the time arrow, the reason that a gas let out of a container into a room doesn't randomly return to the container is due to the same improbabilities. Time reversal is not impossible in a system at equilibrium, but it might as well be.

 

[andresB]

...and especially if it was able to explain properties of matter that in the conventional thermodynamic theory were previously not explainable? If so which?

Thermal Fluctuations.

 

[atyy]

and especially if it was able to explain properties of matter that in the conventional thermodynamic theory were previously not explainable? If so which?

Classical statistical mechanics is able to explain universality in critical phenomena, and calculate the critical exponents.