10.07.2015 Views

Information Theory, Inference, and Learning ... - Inference Group

Information Theory, Inference, and Learning ... - Inference Group

Information Theory, Inference, and Learning ... - Inference Group

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Copyright Cambridge University Press 2003. On-screen viewing permitted. Printing not permitted. http://www.cambridge.org/0521642981You can buy this book for 30 pounds or $50. See http://www.inference.phy.cam.ac.uk/mackay/itila/ for links.404 31 — Ising ModelsN Mean energy <strong>and</strong> fluctuations Mean square magnetization16Energy0.50-0.5-1-1.5-21 10Temperature0.5Mean Square Magnetization10.80.60.40.201 10Temperature1Figure 31.3. Monte Carlosimulations of rectangular Isingmodels with J = 1. Mean energy<strong>and</strong> fluctuations in energy as afunction of temperature (left).Mean square magnetization as afunction of temperature (right).In the top row, N = 16, <strong>and</strong> thebottom, N = 100. For even largerN, see later figures.100Energy0-0.5-1-1.5Mean Square Magnetization0.80.60.40.2-21 10Temperature01 10TemperatureContrast with Schottky anomalyA peak in the heat capacity, as a function of temperature, occurs in any systemthat has a finite number of energy levels; a peak is not in itself evidence of aphase transition. Such peaks were viewed as anomalies in classical thermodynamics,since ‘normal’ systems with infinite numbers of energy levels (such asa particle in a box) have heat capacities that are either constant or increasingfunctions of temperature. In contrast, systems with a finite number of levelsproduced small blips in the heat capacity graph (figure 31.4).Let us refresh our memory of the simplest such system, a two-level systemwith states x = 0 (energy 0) <strong>and</strong> x = 1 (energy ɛ). The mean energy isexp(−βɛ)E(β) = ɛ1 + exp(−βɛ) = ɛ 11 + exp(βɛ)<strong>and</strong> the derivative with respect to β isSo the heat capacity is(31.22)dE/dβ = −ɛ 2 exp(βɛ)[1 + exp(βɛ)] 2 . (31.23)C = dE/dT = − dEdβ1k B T 2 =ɛ2k B T 2exp(βɛ)[1 + exp(βɛ)] 2 (31.24)<strong>and</strong> the fluctuations in energy are given by var(E) = Ck B T 2 = −dE/dβ,which was evaluated in (31.23). The heat capacity <strong>and</strong> fluctuations are plottedin figure 31.6. The take-home message at this point is that whilst Schottkyanomalies do have a peak in the heat capacity, there is no peak in theirfluctuations; the variance of the energy simply increases monotonically withtemperature to a value proportional to the number of independent spins. Thusit is a peak in the fluctuations that is interesting, rather than a peak in theheat capacity. The Ising model has such a peak in its fluctuations, as can beseen in the second row of figure 31.5.Figure 31.4. Schematic diagram toexplain the meaning of a Schottkyanomaly. The curve shows theheat capacity of two gases as afunction of temperature. Thelower curve shows a normal gaswhose heat capacity is anincreasing function oftemperature. The upper curve hasa small peak in the heat capacity,which is known as a Schottkyanomaly (at least in Cambridge).The peak is produced by the gashaving magnetic degrees offreedom with a finite number ofaccessible states.TRectangular Ising model with J = −1What do we expect to happen in the case J = −1? The ground states of aninfinite system are the two checkerboard patterns (figure 31.7), <strong>and</strong> they have

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!