Sunday, April 18, 2010

Thermodynamic fluctuations and statistical physics

Einstein’s earliest papers were concerned with thermodynamics. He wrote a paper establishing a thermodynamic identity in 1902, and a few other papers which attempted to interpret phenomena from a statistical atomic point of view.

His research in 1903 and 1904 was mainly concerned with the effect of finite atomic size on diffusion phenomena. As in Maxwell’s work, the finite nonzero size of atoms leads to effects which can be observed. This research, and the thermodynamic identity, were well within the mainstream of physics in his time. They would eventually form the content of his PhD thesis.[42]

His first major result in this field was the theory of thermodynamic fluctuations. When in equilibrium, a system has a maximum entropy and, according to the statistical interpretation, it can fluctuate a little bit. Einstein pointed out that the statistical fluctuations of a macroscopic object, like a mirror suspended on spring, would be completely determined by the second derivative of the entropy with respect to the position of the mirror.

Searching for ways to test this relation, his great breakthrough came in 1905. The theory of fluctuations, he realized, would have a visible effect for an object which could move around freely. Such an object would have a velocity which is random, and would move around randomly, just like an individual atom. The average kinetic energy of the object would be kBT, and the time decay of the fluctuations would be entirely determined by the law of friction.

The law of friction for a small ball in a viscous fluid like water was discovered by George Stokes. He showed that for small velocities, the friction force would be proportional to the velocity, and to the radius of the particle (see Stokes’ law). This relation could be used to calculate how far a small ball in water would travel due to its random thermal motion, and Einstein noted that such a ball, of size about a micron, would travel about a few microns per second. This motion could be easily detected with a microscope and indeed, as Brownian motion, had actually been observed by the botanist Robert Brown. Einstein was able to identify this motion with that predicted by his theory. Since the fluctuations which give rise to Brownian motion are just the same as the fluctuations of the velocities of atoms, measuring the precise amount of Brownian motion using Einstein’s theory would show that Boltzmann’s constant is non-zero and would measure Avogadro’s number.

Head and shoulders shot of a young, moustached man with dark, curly hair wearing a plaid suit and vest, striped shirt, and a dark tie.
Albert Einstein, 1905, The Miracle Year. On 30 April 1905, Einstein completed his thesis with Alfred Kleiner, Professor of Experimental Physics, serving as pro-forma advisor. Einstein was awarded a PhD by the University of Zurich. His dissertation was entitled A New Determination of Molecular Dimensions.

These experiments were carried out a few years later, and gave a rough estimate of Avogadro’s number consistent with the more accurate estimates due to Max Planck’s theory of blackbody light, and Robert Millikan’s measurement of the charge of the electron.[43] Unlike the other methods, Einstein’s required very few theoretical assumptions or new physics, since it was directly measuring atomic motion on visible grains.

Einstein’s theory of Brownian motion was the first paper in the field of statistical physics. It established that thermodynamic fluctuations were related to dissipation. This was shown by Einstein to be true for time-independent fluctuations, but in the Brownian motion paper he showed that dynamical relaxation rates calculated from classical mechanics could be used as statistical relaxation rates to derive dynamical diffusion laws. These relations are known as Einstein relations.

The theory of Brownian motion was the least revolutionary of Einstein’s Annus mirabilis papers, but it had an important role in securing the acceptance of the atomic theory by physicists.

0 comments:

Post a Comment