You are on page 1of 43

Statistical Physics

Dr. A. J. Macfarlane1 Lent 1998

1 A LT

EXed by Paul Metcalfe – comments to soc-archim-notes@lists.cam.ac.uk.

Revision: 1.10 Date: 1998-06-03 20:13:39+01

The following people have maintained these notes. – date Paul Metcalfe

Contents
Introduction 1 Quantum Statistical Mechanics 1.1 Introduction to Quantum Statistical Mechanics 1.2 Canonical ensemble . . . . . . . . . . . . . . . 1.3 Temperature . . . . . . . . . . . . . . . . . . . 1.4 Towards thermodynamic variables . . . . . . . 1.5 Towards applications . . . . . . . . . . . . . . 1.5.1 N particle partition function . . . . . . 1.5.2 Extensive and intensive variables . . . 1.5.3 Density of states . . . . . . . . . . . . 1.5.4 Gas of spinless particles . . . . . . . . 1.5.5 Entropy and the Gibbs paradox . . . . . 1.6 Harmonic oscillator model . . . . . . . . . . . 2 Thermodynamics 2.1 Introduction . . . . . . . . . . . . . . . 2.2 Applications of dE = T dS − P dV . . 2.2.1 Integrability conditions . . . . . 2.2.2 Specific heats . . . . . . . . . . 2.2.3 Adiabatic changes . . . . . . . 2.2.4 Entropy of n moles of ideal gas 2.2.5 van der Waal’s equation . . . . 2.2.6 The Joule effect . . . . . . . . . 2.3 Some thermodynamics . . . . . . . . . 2.3.1 The second law . . . . . . . . . 2.4 Heat flow . . . . . . . . . . . . . . . . v 1 1 2 4 5 7 7 8 8 9 9 10 11 11 12 12 13 14 14 15 15 16 16 17 19 19 21 21 22 24 25 26 27 28

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

3 Grand ensemble methods 3.1 The formalism . . . . . . . . . . . . . . . . . 3.2 Systems of non-interacting identical particles 3.2.1 A little quantum mechanics . . . . . 3.2.2 The partition functions . . . . . . . . 3.2.3 Classical limit . . . . . . . . . . . . 3.3 Black body radiation . . . . . . . . . . . . . 3.4 Degenerate Fermi gas . . . . . . . . . . . . . 3.4.1 Heat capacity at low temperature . . . 3.5 Bose-Einstein condensation . . . . . . . . . . iii

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

iv 3.6

CONTENTS
White dwarf stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 31 31 32 32 33 34 35 37

4 Classical statistical mechanics 4.1 Introduction . . . . . . . . 4.2 Diatomic gases . . . . . . 4.3 Paramagnetism . . . . . . 4.4 Specific heats . . . . . . . 4.5 Weak interparticle forces . 4.6 The Maxwell distribution . Approximations

ac.E. Phys Waves (etc.com/˜pdm/ or http://www. The recommended books for this course are discussed in the bibliography. Discrete Mathematics Further Analysis Quantum Mechanics Quadratic Mathematics Dynamics of D. A.) General Relativity They may be downloaded from http://home. J.ac. Macfarlane in Cambridge in the Lent Term 1998.uk/CambUniv/Societies/archim/notes.’s Electrodynamics Fluid Dynamics 2 Statistical Physics Dynamical Systems v .cam. Other sets of notes are available for different courses.uk to get a copy of the sets you require. These typeset notes are totally unconnected with Dr.arachsys.Introduction These notes are based on the course “Statistical Physics” given by Dr.cam. Macfarlane. At the time of typing these courses were: Probability Analysis Methods Fluid Dynamics 1 Geometry Foundations of QM Methods of Math.htm or you can email soc-archim-notes@lists.

Redistributions of the electronic files must retain the above copyright notice. EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. All rights reserved. WHETHER IN CONTRACT. EXEMPLARY. this list of conditions and the following disclaimer.Copyright (c) The Archimedeans. 4. Neither the name of The Archimedeans nor the names of their contributors may be used to endorse or promote products derived from these notes. INDIRECT. STRICT LIABILITY. are permitted provided that the following conditions are met: 1. although a fee may be required for the physical act of copying. Cambridge University and their contributors. with or without modification. Neither these notes nor any derived products may be sold on a for-profit basis. 6. . IN NO EVENT SHALL THE ARCHIMEDEANS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT. BUT NOT LIMITED TO. Redistributions in printed form must reproduce the above copyright notice. 3. OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THESE NOTES. INCLUDING. OR CONSEQUENTIAL DAMAGES HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY. INCIDENTAL. this list of conditions and the following disclaimer. 2. You must cause any edited versions to carry prominent notices stating that you edited them and the date of any change. All materials derived from these notes must display the following acknowledgement: This product includes notes developed by The Archimedeans. THESE NOTES ARE PROVIDED BY THE ARCHIMEDEANS AND CONTRIBUTORS “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES. THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. Redistribution and use of these notes in electronic or printed form. SPECIAL. Cambridge University. 5.

N be the total energy and total number of particles of S respectively. One meets three types of ensemble.1 Introduction to Quantum Statistical Mechanics Statistical mechanics deals with macroscopic systems of many particles. but do involve the vast numbers of microstates. Consider • microstates of S . The most probable state of E reveals which microstates of S contribute to its macroscopic states of thermodynamic equilibrium and yield excellent approximations to the values of its thermodynamic variables. • microcanonical : each member of E has the same values of E and N . and could solve the resulting system there is no hope that we could organise usefully the vast amount of data. each identical to S in its macroscopic properties.Chapter 1 Quantum Statistical Mechanics 1. No matter how it is prepared. 1 . temperature T . which correspond to states of thermodynamic equilibrium. the stationary quantum states |i . We use quantum mechanical ideas to approach the subject. • macrostates. we aim to treat the possible states of motion of S by some averaging or statistical procedure that allows us to predict values of the variables in Σ that are constant in states of thermodynamic equilibrium of S . Σ includes the pressure P . This allows us to replace time averages for a single system S with averages at a fixed time over a suitable ensemble E of systems. Consider an isolated system S of gas in a vessel whose walls neither let heat in or out and is subject to no mechanical action. We cannot possibly analyse this in detail — even if we knew the forces and initial conditions. At best. Any state of S (before or after thermodynamic equilibrium is reached) contains particles with positions and momenta changing in time. volume V . it is an experimental fact that S reaches a steady state (a state of thermodynamic equilibrium) in which a set Σ of (rather few) thermodynamic variables are constant. The latter are not states in the quantum mechanical sense. Let E . The ergodic hypothesis (which is provable for some systems) is that S passes in time through all its possible states compatible with its state of thermodynamic equilibrium. total energy E and entropy S .

Then ai = A i and i ai Ei = AE. QUANTUM STATISTICAL MECHANICS • canonical : each member of E has N particles but E is not fixed. The microstates of S are the (complete orthogonal) set |i with energy eigenvalues E i . 1 log n! ∼ n log n − n + 1 2 log 2πn . . p m = 2 −N N +m 2 − 2m2 e N . The fact that averaging procedures give reliable estimates of values of physical variables and apparently different procedures yield equivalent results is due to the effect of the very large number of particles involved. . exponential term e 0.1) The average energy of the members of E is thus the fixed value E . This course studies mainly canonical ensembles. We assume the spectrum is discrete but with possible degeneracy. SA . fluctuations about the average values are found to be very small and all three types of ensemble give the same thermal physics. πN N 2 ∼ m using Stirling’s formula 1 for N −1011 is effectively zero. We will do systems with identical bosons/fermions via the grand ensemble for technical simplicity. i A Proof. For instance. . We can distribute a 1 systems in the microstate |1 over E in a ways. ! Given a set {ai } (a configuration of E ) there are W (a) = ÉAa ways of realising i! it.2 CHAPTER 1. Thus W (a) = A a1 A − a1 ··· = a2 A! . . (1. S 1 . For the latter two. • grand : neither E nor N is fixed. although the average total energy is time independent. We associate with S a canonical ensemble E . although the averages are. Then 1 A−a1 we can distribute the a 2 states in |2 in a2 ways (and so on). If N = 10 23 and m = 1017 then the 1. i ai ! We assign equal a priori probability to each way of realising each possible configuration of E and thus there is a probability proportional to W (a) of realising the configuration {a i }. consider the probability pm of obtaining N/2 + m heads from N coin tosses.2 Canonical ensemble We study a system S of N particles in a volume V with N and V fixed. Suppose that in a possible state of E there are ai members of E in the microstate |i . This is a very large number A of distinguishable replicas of S .

V .1). (1. The fraction of members of E in the microstate |i in the macrostate of thermodynamic equilibrium is ρi = 1 ai = e−βEi . ∂β (1.1). For large n. 0 (negligible probability attaches To compute this suppose A 0 such that a i to configurations where this fails). and may be thought of as the probability of finding |i in the state of thermodynamic equilibrium. We eliminate α via A = tion Z= i i (1. ρi X i . We define the average X of a physical variable X taking the value X i in the state |i by X = i 1 For instance E = A i ai Ei = E (reassuringly).1. For instance.4) This is the Boltzmann distribution.2) ai = e−1−α Z . We seek to maximise this subject to the constraints (1. CANONICAL ENSEMBLE 3 We will associate the state of thermodynamic equilibrium of S (for the fixed values N . Stirling’s formula allows log n! ∼ n (log n − 1) and so log W ∼ A log A − A − i ai (log ai − 1) = A log A − i ai log ai . (1.2. defining the canonical partition func- e−βEi ≡ Ej Ω(Ej )e−βEj .6) . A Z (1. E ) with the configuration of E that is most probable in the presence of the constraints (1. Z is very important. It leads directly from quantum mechanical data to calculation of the thermodynamic variables for S in thermodynamic equilibrium.3) where Ω(Ej ) is the degeneracy of the energy level E j . Thus aj = e−1−α−βEj .5) E=− ∂ log Z . We use two Lagrange multipliers α and β and solve ∂ ∂aj 0= log W − α i ai − β i ai Ei and so log aj + 1 + α + βEj = 0.

Thus √ 1 |∆E | N ∝ = N−2 . Suppose that Sa has microstates |i of energy E ai . σ occurs c iσ times in Eab .2 Suppose the |i occurs a i times. |σ occurs bi times and |i. Then we have ai = A i σ bσ = A ciσ i iσ (1. Note that E = − Z ∂β and so 1 ∂2Z ∂E 1 =− + 2 ∂β Z ∂β 2 Z ∂E ∂β ∂Z ∂β 2 = − E2 + E 2 = − (∆E ) .7) ciσ = A (1.4 CHAPTER 1. Nb respectively we place them in contact to form a composite system S ab such that energy can pass from one to the other and a state of thermodynamic equilibrium is reached. except insofar as necessary to allow a a b state of thermal equilibrium to be reached. 1 ∂Z We can calculate the variance in the energy in a similar way.10) AE = iσ ciσ (Eai + Ebσ ) . It gives in effect the average over all possible states of the ensemble. We make a canonical ensemble E ab for Sab by distributing A replicas of S a and A replicas of Sb independently across the A members of E ab so that each member of Eab has a component of type S ab . and as E depends smoothly on β we expect ∝ N as well. For A large (as in all cases of interest) the most probable state is overwhelmingly so.8) ai = σ ciσ bσ = We demand that the average energy or E ab be fixed at AE = i ai Eai + σ bσ Ebσ (1. 1. Now the configuration {c iσ } arises in 2 If the interaction of particles of S with those of S is negligible. Sb has microstates |σ of energy Ebσ and Sab has microstates |i. The idea of associating average values with actual physical predictions depends on the possible variances being negligible. QUANTUM STATISTICAL MECHANICS Here.9) (1. . holding V fixed corresponds to keeping all the E i fixed. σ of energy E ai + Ebσ .3 Temperature Given two systems Sa and Sb with fixed volumes V a . Vb and numbers of particles N a . 2 For typical large systems E ∝ N . E N which is very small for very large N .

1. 3 k is the same because equal volumes of all gases at fixed temperature and pressure are observed to have the same number of molecules. Consider Sgas in thermodynamic equilibrium in a container of volume V and change V by a small amount by slowly moving one of its walls.11) ways and the configuration {a i }{bσ } arises in W (a)W (b) ∝ A! i ai ! A! σ bσ ! (1.4.7) and (1.9) to get ai = e−1−αa −βEai bσ = e−1−αb −βEbσ Defining Zab (β ) = iσ ai A bσ ρσ = . A We can also maximise (1. . Note that E ∝ V − 3 and we generalise this to Ei = Ei (V ). 1 We define T by β = kT where k is a constant. A common value for β characterises the state of thermal equilibrium of S ab and it is natural to assume that β is some function of temperature T and that the average energy (the energy) of a system is a function of N .1.12) ways. We can calculate the most probable distribution of {c iσ } by maximising (1. TOWARDS THERMODYNAMIC VARIABLES 5 W (c) ∝ A! iσ ciσ ! (1. We will define T for a standard system (a thermometer) and for other systems by putting them into a state of thermal equilibrium with our standard system.12) subject to (1. T is in degrees Kelvin.10). V and T . We get ciσ = e−1−αab −β (Eai +Ebσ ) ρiσ = ciσ .8) and (1. spinless particles in a cube of side L have energies 2 2 En = 2mL2 |n|2 . where n ∈ N3 . As this argument applies to any two systems. k should be a universal constant (Boltzmann’s constant) the same for all systems when a universal definition of T is used.4 Towards thermodynamic variables Recall that in quantum mechanics. A ρi = e−β (Eai +Ebσ ) Za (β ) = i e−βEai Zb (β ) = σ e−βEbσ we see that Zab (β ) = Za (β )Zb (β ) and ρiσ = ρi ρσ . 3 This applies to all gases at sufficiently low density. P V = N kT for a sample of N molecules. Recall the ideal gas law.11) subject to (1.

13) Consider next changes in which the ρ i are allowed to vary. Then δE = −P δV + i Ei δρi .15) We might regard this as arising from the comparison of ensembles with infinitesimally different states of thermal equilibrium. (1. Thus the work done to the system is −P Aδl = −P δV . where δQ is the heat supplied to S and δW is the work done on S . (1.. written in the form dE = δQ + δW.6 CHAPTER 1.14) ai log ρi − log A i ai ρi log ρi i at thermal equilibrium.15) states the First Law of Thermodynamics4 . The change in energy is supplied by the work done on S by a piston applied to the right hand wall. QUANTUM STATISTICAL MECHANICS i The Ei will change according to E i → Ei + ∂E ∂V δV and changes in which the ρ i do not change will be examined first. For slow motion the system passes through successive positions of thermal equilibrium and then the applied force just balances the pressure force P A. W at thermal equilibrium A k A log A − = ai log ai A i A log A − i k = A = −k (1. We see that 1 log ρi ρi δS = −k = −k = and thus 1 T δρi log ρi + ρi δρi (−βEi − log Z + 1) Ei δρi i i Ei δρi = T δS . 4 You (1. We have defined S such that δE = T δS − P δV or in terms of exact differentials. and note that i δρi = 0. δE = δW = −P δV . ∂V β ∂V (1.16) can’t win.. . Define the entropy S= k log Wmax that is. Now E = i ρi Ei and so P =− i ρi 1 ∂ ∂Ei = log Z. dE = T dS − P dV.

giving 2 3 P V = 3 E . We want to be able to calculate S from the partition function. we can calculate the thermodynamic variables in a state of thermal equilibrium. 2 S = −k i ρi log ρi ρi (−βEi − log Z ) i = −k = 1 E + k log Z. Now.6)).17) Also.18) We have a very tangible definition of S . T ∂β ∂T ∂T (1. S=− ∂ ∂F 1 ∂ log Z + k log Z = (kT log Z ) = − .15) also shows that T = ∂E ∂S ∂E and P = − ∂V . T Define the free energy F = E − T S .1. We see that the one particle partition function is z = zT zrot zvib . Later we will consider a gas of diatomic molecules. then 1 F = −β log Z. Let Z ≡ ZN describe the sum over states of a gas of N very weakly interacting particles.3). F (by (1.17)) and S (by (1.13)). where z is the “one particle partition function”. we have Boyle’s law P V = N kT .α e−βEaα = a e−βE1a α e−βE2α = Z1 Z2 . α .5. Then as H1 and H2 commute. TOWARDS APPLICATIONS (1.1 N particle partition function Consider a composite system with Hamiltonian H = H 1 + H2 and H1 and H2 independent (they commute). Let H 1 |a = E1a |a and H2 |α = E2α |α . E = i Ei ρi ∝ V 3 V . E = 2 N kT (see later). If these are supposed independent (distinguishable) then Z = z N . . 7 −3 E and so P = 2 For a gas of spinless particles. Given the partition function (1. where Wmax is the number of microstates of S contributing to the state of thermal equilibrium. where Eaα = E1a + E2α . α = Eaα |a. at thermal equilibrium. where HT describes the centre of mass motion. Then the 2 particle partition function is Z= a.18)).5. (1. For the latter it seems best to use grand ensemble methods. but fails for systems of identical (indistinguishable) bosons or fermions. Combined with the result for a perfect gas. H rot describes rotation about the centre of mass and Hvib describes vibration along the axis. This allows easy calculations. as proportional to log W max .5 Towards applications 1. P (by (1. E (by (1. H |a. H = H T + Hrot + Hvib . 1.

1. S = S1 + S2 and so on. QUANTUM STATISTICAL MECHANICS 1. Think initially of a system made up of two independent subsystems.19) = V 4π = (2π )3 = 2πV = 2m h 2 md 2 2 (1. We use (1. As Z = Z1 Z2 we see that E = E1 + E2 .20) d d g ( ). but (1. In general.5. This can fail if volume energies do not swamp surface energies or if intermolecular forces are neither weak nor short range. and so P is intensive.19) is needed in kinetic theory. Then V = V1 + V2 . V . We see that L 2π V (2π )3 3 → i d3 k 4πk 2 dk 2m 2 3 2 1 2 1 2 (1.5. ole of In two dimensions we consider a square of side L and A = L 2 enters in the rˆ m g 2 πA is independent of . Instead of using solutions ψ of the Schr¨ odinger equation such that ψ = 0 on the walls we use periodic boundary condi2 2 k .20) in isotropic contexts. We find that g ( ) = 2 S h2 .8 CHAPTER 1. V . There is one such state per unit volume in nspace and in the continuum limit. We insert a factor gS = (2S + 1) to cope with massive particles of spin S . S and E are extensive and T and P are intensive. Thus we can use ψ k ∝ eık·x with = 2m Lk 3 The periodicity gives that 2π ∈ Z . and g ( ) = gS 2πV 2m h2 3 2 1 2 gives the density of states. N . and so T is intensive. In a state of equilibrium there is mechanical equilibrium and so the systems have the same pressure. there are d 3 n states with n in the range n → n + dn. E and S are proportional to N and so extensive. A logical extension of this argument gives that V . whereas intensive variables are independent of the amount of matter. ψ (x + Li) = ψ (x) (and so on). L 3 3 There are thus d3 n states with k in the range k → k + dk and hence 2 d k states π with k in the range k → k + dk.2 Extensive and intensive variables Extensive variables are proportional to the amount of matter in a system at fixed temperature. tions.3 Density of states Consider a spinless particle in a box of side L. In a state of thermal equilibrium the systems have the same β (and hence same T ).

23). g S = 2 and so V (2π )3 4π c 3 0 ∞ →2 i 2 d = 8πV c3 ∞ 0 ν 2 dν.4 Gas of spinless particles Consider N spinless particles in a cube of side length L = V 3 .5. (1. 1. which is not extensive. For photons.21) We also supply a factor g S = 2S + 1 for a particle of spin S with nonzero rest mass. TOWARDS APPLICATIONS For relativistic particles of rest mass m we have (1. where g ( ) = DV 1 2 and D = 4π 3 2 .23) Now z = z N and so log Z = N log z . (1. (1. (1. as V ∝ N . Setting y = β we get −y z = DV (kT ) ∞ 0 dy y e =V 2mπkT h2 3 2 .23) and (1. N If instead we use Z = z N ! .22) 1.5. This is Gibbs’ paradox.5 Entropy and the Gibbs paradox Using Z = z N .1.18) and Stirling’s formula we have V +3 S = N k log N 2 N k log 2mπkT h2 +3 2Nk and we see that S is extensive. Taking Z = z N ! cures this but has no foundation in classical statistical physics. m = 0. (1. ∂β This is the classical result: there is an energy 1 2 kT associated with each degree of freedom of each particle. Then the single particle partition function z= i 1 e −β i → 0 2m h2 1 2 3 2 ∞ d g ( )e−β . Often an additive constant in S does not matter and (1.23) gives all thermodynamics otherwise correctly. Thus E = −N ∂ 3 −3 2 log β = 2 N kT. Combining with P V = 2 3 E we get P V = N kT . It reflects a normal classical view of how to treat indistinguishable N particles. .5. (1.19) to get d3 k = 4πk 2 dk = 4π ( c)3 ∞ 9 = m 2 c4 + k 2 c2 1 2 1 2 and use d mc2 2 − m 2 c4 .18) we have S = N k log V + 3 2 N k log 2mπkT h2 +3 2 N k.

this case the critical temperature T c = k . with energy 2 kT per degree of freedom per “atom”.6 Harmonic oscillator model Model a crystal by placing one atom at each point of some regular lattice with N sites. −1 E 1 If ω kT this gives the classical result N −1 2 ω = kT . E −β ω −1 . and using En = ω (n + 1 2 ) for the simple harmonic oscillator with frequency ω we get e− 2 z (ω ) = 1 − e −β β ω ω . This is a one dimensional version of the Einstein solid. Quantum statistical mechanics “knows” when to count the full classical 1 2 kT value per degree of freedom: do it only when T is large on some scale set by the problem. Even if ωr = ω for all r. simplify by taking a simple harmonic oscillator of frequency ω r instead of the rth atom.10 CHAPTER 1. Now ZN = r z (ωr ). We find that ∂ log z E =− = N ∂β 1 2 ω+ eβ ω ω . QUANTUM STATISTICAL MECHANICS 1. In 1D. the “atoms” are distinguishable as there exists one at each lattice site. In ω . which tends to zero as If ω kT then β ω 1 and N 2 ω ≈ ωe T → 0.

Chapter 2 Thermodynamics 2. Suppose we change from a state with E 1 (S1 . equivalent to the conservation of energy. In the case of the energy. All thermodynamic variables for the sample of gas we are talking about can be regarded as a function of two suitably chosen independent variables. 11 . • equations of state: for instance P V = N kT . V2 ). V ). E arises as E (S. For finite changes from the initial to the final state conservation of energy gives ∆E = ∆Q + ∆W .1 Introduction Consider a volume V of gas with a fixed number N of particles. We also found that dE = δQ + δW . The state of thermal equilibrium of this gas is characterised by T and V . which arose as the average energy and has negligible fluctuations. This gives us ∆Q ≤ S2 S1 T dS. where δQ is the heat supplied to the system and δW is the work done on the system. In general. • dE = T dS − P dV . Our work with quantum statistical mechanics has produced some concepts naturally: • E : the total energy of the system. V2 ) to E2 (S2 . we expect ∆W ≥ − V2 V1 P dV because of wasted effort (against friction or in producing convection or turbulence). • S : the entropy. This is the first law of thermodynamics. which is true for any thermodynamic change between two infinitesimally close states of thermodynamic equilibrium. This has a clear significance as the number of microstates contributing to the macrostate of thermal equilibrium.

but they are interdependent. These give four Maxwell relations. x . F . In induced/allowed changes in real life with δQ = 0. Consider z = z (x. 2. Some (obvious) irreversible processes are fast piston movement. Then δz = 1= 0= ∂z ∂x ∂x ∂z ∂x ∂y δx + y ∂z ∂y δy x and hence y ∂z ∂x ∂z ∂x and y + y z ∂z ∂y . δS > 0 and so the entropy increases. H and G are all extensive. V = T H = E + P V defines the enthalpy H (S. free expansion of a gas into a vacuum and the mixing of samples of different gases.2. This gives the Maxwell relation ∂P ∂T . so that all thermodynamic variables are well-defined throughout the change. We need some rules for shunting partial derivatives around. If (exceptionally) G = 0 (true for a gas of photons) then G = N µ and so the chemical potential µ can vanish for arbitrary finite N. V ) within dE = T dS − P dV = Thus T = ∂2E ∂V ∂S ∂E ∂S V ∂E ∂S dS + V ∂E ∂V S ∂2 E ∂S∂V and P = − ∂E ∂V S . P ) and G = E − T S + P V defines the Gibbs function G(T.2 Applications of dE = T dS − P dV 2. THERMODYNAMICS These inequalities hold as equalities for reversible changes. as dF = dE − T dS − S dT = −S dT − P dV. Quasi-static changes are changes which are done so slowly as to pass through states arbitrarily close to a state of thermodynamic equilibrium. E .1 Integrability conditions The internal energy E is seen as E (S. so ∂T ∂V = and − ∂P ∂S = V S This is a Maxwell relation. P ). y ). dE is an exact differential. Thus S = − ∂F ∂T V and P = − ∂S ∂V ∂F ∂V T .12 CHAPTER 2. The free energy F = E − T S is a natural function of T and V . that is changes which are quasi-static and non-dissipative.

the specific heat at constant volume and C P = T ∂T . Recall T dS = dE + P dV . the ideal gas law for n moles of gas is P V = nNA kT = nRT. Now T dS = dE + P dV = dH − V dP and so . 2.2. In fact.1) = −1. a mole of any ideal gas occupies the same volume. The ideal gas law tells us that at fixed temperature and pressure. Written in terms of moles. ∂S ∂S Define CV = T ∂T . for a perfect gas we need the equation of state and E = CT . For instance. APPLICATIONS OF DE = T DS − P DV This can be rewritten in a slightly different form: ∂x ∂z ∂x ∂y ∂z ∂x 13 =1 y y z ∂y ∂x z ∂x ∂z (2. we would like E as a function of T and V . T ∂2S ∂V ∂T Also.2) R is called the gas constant and is the same for all ideal gases. so that ∂E ∂T ∂E ∂T ∂E ∂V = ∂E ∂V T dS = T T ∂S ∂T ∂S ∂V = V dT + V dV + P dV T and so and V = T + P.2. Thus E is a function of T only. F .2 Specific heats An amount of gas with N A (Avogadro’s number) of molecules is called a mole. we know that ∂2 S ∂T ∂V and so ∂P ∂T − P. V P the specific heat at constant pressure.2. For a sample of 1 mole of ideal gas we define the specific heat (or heat capacity per mole) using δQ = T dS under suitable conditions. P V = N kT and so ∂V T Given the laws of thermodynamics we need two inputs from experiment to specify all thermodynamic variables completely. V ∂E ∂V =T T ∂E = 0. (2. all gases are essentially ideal at sufficiently low density. y When we wrote down E . In practice. For a perfect gas. H and G we wrote them in terms of the mathematically natural independent variables in each case.

5 . Thus dS = nCV T V and so S = nCV log T + nR log V + c1 .3) CP = For one mole of ideal gas we know that E = CT and so C = C V . γ Note that adiabatics P V constant are steeper than isothermals P V constant. .2. and does not care that S is not explicitly extensive. = ∂S ∂T V and so ∂V ∂T ∂P ∂T . 2. V dV dT + nR . S = nC V log P V γ + c and so (as expected). To find CP we need to use P and T as the independent variables. THERMODYNAMICS CV = T ∂S ∂T = ∂H ∂T V ∂E ∂T . . V CP = CV + T ∂S ∂V T ∂V ∂T = CV + T P (2.4) P This allows the calculation of C P − CV from the equation of state: for an ideal gas CP +R CP − CV = R. S is constant on adiabatics (isentropics). (for monatomic. Now 0 = RdE + RP dV = RnCV dT + RP dV = CV (P dV + V dP ) + RP dV = CP P dV + CV V dP. . . Note that γ = CV >1 CV V for an ideal gas. We refer to n moles of ideal gas using E = nC V T and P V = nRT .3 Adiabatic changes These are (defined as) changes which are reversible and satisfy δQ = 0. V dP γ Thus 0 = γ d V + P and so P V is constant on adiabatics. 2 R.4 Entropy of n moles of ideal gas We start (as usual) from T dS = dE + P dV dV . . Thermodynamics cannot determine the constant c1 . P V (2. . Now dS = ∂S ∂T dT + V = Thus ∂S ∂T P ∂S ∂V ∂S ∂T + dV T dT + V ∂S ∂V T ∂V ∂T ∂S ∂V P T ∂V ∂T dT + P ∂V ∂P dP T . . 5 Statistical mechanics gives CV = 3 2 R.2. = nCV dT + nRT .14 CHAPTER 2. diatomic (etc) 5 7 gases). Thus γ = 3 . Define the ratio of specific heats γ = C . 2.

For a van der Waal’s gas we still have dE = 0 (which is true in general) and so E stays constant. T decreases because some of the molecular kinetic energy is lost in the expansion against attractive intermolecular forces. and then to reach a state of thermal equilibrium specified by P 2 .2. As δQ = 0 (adiathermal) and δW = 0 (no work is done). and as P → ∞ at constant T . which is the residual volume of all the molecules. V → nB .5) where A. (2.6 The Joule effect Consider the apparatus shown. dE = 0 and so for a perfect gas T2 = T1 2 and S2 − S1 = nR log V V1 > 0 for V2 > V1 . 2 2 2. APPLICATIONS OF DE = T DS − P DV 15 2. If these are short range then the smaller V is the more important these become. B and R are strictly positive constants. V2 and T2 . for van der Waal’s gas We can show that CV is independent of V for a van der Waal’s gas and we suppose that CV is also approximately independent of T . due to intermolecular attractive forces. pressure P1 and temperature T 1 . • Molecules are not treated as point particles.5 van der Waal’s equation We get a better agreement with experiment by replacing the perfect gas law with P+ n2 A V2 (V − nB ) = nRT. Now ∂T ∂V =− E ∂E ∂V T ∂E ∂T V − T ∂P ∂T V 2 = P nCV n A 1 =− 2 V nCV < 0. with adiathermal walls and containing n moles of gas at volume V1 .2. We can now integrate this equation to get T2 − T1 = nA CV 1 1 − V2 V1 < 0.2. A n A • P+n V 2 reduces the real gas pressure by an amount V 2 . . Pull back the partition and allow the gas to expand (irreversibly) into the total volume V 2 .2.

can’t break even either.3 Some thermodynamics Suppose a sample S of n moles of perfect gas is put in thermal contact with a heat bath B and compressed reversibly from volume V 1 to volume V2 .1 The second law (Kelvin) No process can continuously (by going round a cycle) extract heat from a heat bath and perform an equal amount of work. Thus VA VB = VD VC . Now ∆S = nR log V V1 and so the entropy of S decreases. E = E (T ) is unchanged over one complete cycle so that the work done in a cycle by S is −W = ∆Q1 + ∆Q2 by the first law (as δQ = 0 on adiabatics). This can be generalised to other cycles. ∆Q2 ∆Q2 On adiabatics T V γ −1 is constant and so γ −1 γ −1 = T1 VD T2 VA γ −1 γ −1 and T2 VB = T1 VC . On CD the heat bath at T 1 supplies heat ∆Q1 < 0 to S at temperature T1 . because the isolated universe of S and B has ∆S = 0 the entropy of B must increase. (Clausius) No process can continuously transfer heat from a colder to a hotter body. On AB the heat bath at T 2 supplies heat ∆Q2 > 0 to S at temperature T 2 . We know that VD VB B ∆Q2 = nRT2 log V VA > 0 and ∆Q1 = nRT1 log VC = −nRT1 log VA . 2. and clearly 0 < η < 1. Now V2 V1 ∆Q + ∆W = 0 ⇒ ∆Q = −∆W = 2 P dV = nRT log V V1 . The efficiency η is defined η= −W ∆Q1 =1+ < 1.. If the gas is perfect then E = E (T ) = nCV T and so E stays the same. 1 Take a sample S of perfect gas around a closed Carnot cycle by means of adiabatics and isothermals (T 2 > T1 ). (−∆Q 1 ) > 0 of heat is wasted at the low temperature heat bath. Thus ∆W > 0 for compression (V 2 < V1 ) and ∆Q < 0.. However. Heat is given out to B 2 by S . Thus η = 1 You T2 −T1 T2 . (see footnote on page 6) .3. THERMODYNAMICS 2.16 CHAPTER 2. This agrees with the Kelvin statement of the second law.

HEAT FLOW 17 2. x1 (T1 − T ) + x2 (T2 − T ) = 0 and ∆S1 +∆S2 ≥ 0. Put S1 and S2 in thermal contact allowing no change in V 1 + V2 .4 Heat flow Let S1 be a sample of n 1 moles of ideal gas at temperature T 1 in a fixed volume V 1 . As for the entropy. and we can solve for T .2 log TT 1. Similarly for S2 . Then E1 (T1 ) = n1 CV1 T1 = x1 T1 . Suppose the state of thermal equilibrium is reached at a temperature T .4.2. ∆S 1.2 = x1.2 . Then (by conservation of energy).

THERMODYNAMICS .18 CHAPTER 2.

Chapter 3 Grand ensemble methods 3. Suppose that ai members of G are in the microstate |i so that ai = A i i ai Ei = AE i ai Ni = AN. We find this by maximising the same log W as before to get 0= ∂ ∂ai log W − α i ai − β i ai Ei − γ i ai N i . We have thus fixed the average energy and number of particles of members of the ensemble. Each configuration {a i } of G defines a state of the ensemble and we assign to each configuration equal a priori probability. A Z 19 . We now define the grand partition function Z= i e−β (Ei −µNi ) .1) The fraction of members of G in the microstate |i is ρi = e−β (Ei −µNi ) ai = . (3. This gives us that ai = e−(1+α) e−β (Ei −µNi ) . Given a system S of fixed volume we construct a grand (canonical) ensemble G with a large number A of distinguishable replicas of S in microstates |i of S with Ni particles and energies E i . where we have defined the chemical potential µ by βµ = −γ . We associate the state of thermodynamic equilibrium of S with the most probable configuration of G in the presence of our constraints.1 The formalism Our approach is a direct extension of the approach we adopted earlier for the canonical ensemble.

(3. on the whole. β. S. As before we can and E use Z to calculate thermodynamic variables: 1 β ∂ log Z ∂µ N= .2) . (3.µ (3.V (3.6) . Defining the entropy S=− gives that 1 δρi ρi k log Wmax = −k A ρi log ρi i δS = −k i δρi log ρi − i ρi = −k i δρi (−β (Ei − µNi ) − log Z ) . This gives the fundamental thermodynamic relation T δS = δE + P δV − µδN. from consideration of changes at constant ρ i we find that δE = −P δV and 1 ∂Ei = ∂V β ∂ log Z ∂V P =− i ρi .V (3.20 The grand ensemble average is CHAPTER 3. The chemical potential is therefore given by ∂E ∂N µ= . a good thing). β. GRAND ENSEMBLE METHODS ¯= O i ρi Oi ¯ = E and N ¯ = N (which is.4) More general changes in which δρ i = 0 obey δρi = 0 i i Ni δρi = δN δE = −P δV + i Ei δρi .V As before.5) yielding the first law dE = δQ + δWmech + δWchem . where δWchem = µδN is the work done in adding δN particles to the system S .3) E − µN = − ∂ log Z ∂β µ.

The state of thermodynamic equilibrium corresponds to the most probable state in G . dE = T dS − P dV + µdN .7) (3. Now Ω also equals −P V and so P V = kT log Z (3. µ.2 Systems of non-interacting identical particles 3. We can view E as a function of S . N ) and so E=S ∂E ∂S +V V.18). They are also very sharp (for the same reason). namely S= ∂kT log Z ∂T . We return to the first law. V and N . E . λV.1 A little quantum mechanics We will treat such systems with only one type of particle. Some thermodynamic variables arise from averages over microstates weighted by the probability of finding them in the state of thermodynamic equilibrium. 3. SYSTEMS OF NON-INTERACTING IDENTICAL PARTICLES Returning to the entropy S we see that S = kβ (E − µN ) + k log Z which can be put in the same form as (1. S and V are extensive we get E (λS. . λN ) = λE (S.9) allows the calculation of the equation of state from the grand partition function. .N ∂E ∂V +N S. NB energy flows from the system with the higher temperature to that with the lower temperature until thermodynamic equilibrium is reached.3 gives that two systems placed in thermal and “diffusive” contact will reach a thermodynamic equilibrium char1 acterised by common values of β = kT and µ.2.2.N ∂E ∂N S. We keep the idea that at fixed N A .V 21 (3. ) is . We define the grand potential Ω for a state of thermodynamic equilibrium by Ω = E − T S − µN = E − µN − (E − µN + kT log Z ) = −kT log Z . V. The averages are effectively averages over all possible states of the ensemble because the macrostate of thermodynamic equilibrium dominates overwhelmingly.3.V = T S − P V + µN. . and when (as for many large systems).8) A similar argument to that used to define temperature in 1. We add that if the contact allows the diffusion of particles then in the state of thermodynamic equilibrium with constant µ there is (on average) no diffusion. When interactions are neglected then the wavefunction of each stationary state of the system Ψ(x 1 . Thus Z = e−β Ω . x2 .

2 The partition functions Let the one particle wavefunctions of the particles of S be ψ r (x) with energy pose that in the microstate |i . µ2 ) 3! ψ3 (x1 . x2 . 3. µ2 ) ψ1 (x3 . ψ3 (x3 . µ3 ) . For spin 1 2 fermions we have φ(x) → ψn (x)α or ψn (x)β. µ3 ) ψ2 (x3 . We can use determinants to build the antisymmetric wavefunctions. . 2 For identical bosons we cannot use wavefunctions φ 1 (x1 )φ1 (x2 )φ2 (x3 ) which correspond to particles 1 and 2 in state 1 ≡ n 1 and particle 3 in state 2 ≡ n2 . µ1 ) ψ2 (x2 . 2 ≡ n2 and µ = α or β ) we get the antisymmetrized version Ψ(x1 . of one particle wavefunctions. n r of the particles have wavefunction φ r (x). Consider spin 0 bosons in a cube of side L. µ ) ψ1 (x2 . . µ1 ) ψ3 (x2 . Ψ= ψ (x . µ2 ) . GRAND ENSEMBLE METHODS obtained by either symmetrization (for bosons) or antisymmetrization (fermions) of the product φ1 (x1 )φ2 (x2 ) . µ) and ψ2 (x.22 CHAPTER 3. 3! All we can say about this is that there are two particles in state |1 and one particle in state |2 . µ2 ) Ψ reflects the Pauli exclusion principle which forbids any two of the ψ i or any two sets (x.2. µ2 ) 1 1 1 1 ψ2 (x1 . For instance. Sup- nr and Ei = r r nr . µ3 ) 1 ψ1 (x1 . given ψ 1 (x. . but we must use instead the symmetrized wavefunction Ψ= 1 (φ1 (x1 )φ1 (x2 )φ2 (x3 ) + φ1 (x1 )φ2 (x2 )φ1 (x3 ) + φ2 (x1 )φ1 (x2 )φ1 (x3 )) . µ) (where 1 ≡ n1 . 2! ψ2 (x1 . µ1 ) ψ1 (x2 . where α is “spin up” ↑ and β is spin down ↓ with H and E n independent of α and β . Then the one particle wavefunctions are ψn (x) = with corresponding energies En = 2 2 L 3 2 πx πx πz sin n1L sin n2L sin n3 L 2mL2 |n| . µ2 ) = Similarly for three particles. µ) from being the same. Each Ψ is determined fully by the number of times each one particle wavefunction is used in the product term that we (anti)symmetrize. µ2 . µ1 ) ψ2 (x2 . Ψ is fully determined by the fact that φ 1 is used twice and φ2 used once. Then Ni = r r.

SYSTEMS OF NON-INTERACTING IDENTICAL PARTICLES 23 We see that the microstate |i of S is fully determined by i ≡ {n r }. r −µ) + 1 s=r β.10) We use this to find N= Now n r e −β ( Z . n ¯ r is the average occupation number of the r th one We see that N = rn particle state at thermal equilibrium. 0 ≤ nr < ∞ and Z= r 1 1− e −β ( r −µ) . ) = = r nr e−βnr ( r −µ) . We obtain the full set of microstates of S by letting the n r vary without restriction over their allowed range of values. r −µ) ∓1 . r −µ) 1 β ∂ log Z ∂µ = β. so that Z= r 1 + e −β ( r −µ) . For fermions we have that n r can only be zero or one.11) This is the Bose-Einstein distribution.. This is the Fermi-Dirac distribution.n2 =− = 1 β eβ ( ∂ log Z ∂ r 1 ... This gives that n ¯r = 1 eβ ( r −µ) −1 r . If there are g r one particle states of energy particles with energy r as n ¯( r ) = gr eβ ( we can write the average number of . s ¯r . n ¯r = n1 .n2 .. 1 +n2 2 +···−µn1 −µn2 −.µ.3.... (3. (Thus N i and Ei cannot themselves be restricted — this is why we use the grand ensemble method..2. For bosons.) We can now write down the grand partition function Z= i e−β (Ei −µNi ) e−β (n1 n1 . (3.V r 1 eβ ( r −µ) +1 .

4 2 . 5 d ∓1 eβ ( −µ) and E = DV 0 3 ∞ eβ ( −µ) Putting z = β we find N = DV (kT ) 2 I 1 (−βµ) and E = DV (kT ) 2 I 3 (−βµ). DV 2π 2m h2 3 2 as in section 1..2. ∞ 0 N= g( ) d β ( e −µ) ∓ 1 1 2 E= 0 ∞ eβ ( −µ) g( ) d .5. The average number of particles with energy in the range → + d is n( )d = g ( )d . eβ ( −µ) ∓ 1 Thus in the continuum limit.3. If ey We can approximate the integrals (expand the integrand) to get E∼ 3 2 N kT eβµ 1 ∓ √ + . ∓1 In three dimensions.9) we find that P V = 2 3 E .3 Classical limit For a volume V of N particles with energy E .. GRAND ENSEMBLE METHODS for N large we pass to the continuum limit. The in g ( ). we use g ( ) = DV N = DV 0 ∞ 1 2 1 2 3 2 and so d ∓1 . . 2 2 with In (y ) = 0 ∞ z n dz . e z +y ∓ 1 1 we can neglect the 1 in the denominator and find that n( ) ∝ e −β g ( ). → r d g ( ). the grand partition function Z is given by log Z = ∓ 0 ∞ g ( ) log(1 ∓ e−β ( −µ) )d . 2 3 comes from the 1 2 factor 3. where D = gS and gS = 2S + 1 is the spin degeneracy.24 CHAPTER 3. g ( ) = DV and so (integrating by parts) we see that ∞ 0 log Z = DV 2 = β . We can now use the results of section 3. 3 2 3 eβ ( 3 2 −µ) ∓1 Combining this with (3. 1 2 where g ( ) is the density of states factor.1 to find things like N and E .

Thus n ¯r = eβ 1 −1 and E = r r 1 r eβ r −1 . Using the full formula for the energy we find that . This condition is satisfied at low density N and at high temperature. It fails to holds for electrons in solids due to the fact that their effective mass inside the solid is much less than the free electron mass. Most real gases liquefy before quantum mechanical effects set in. Using this result. where e kT − 1 ≈ kT and dE = 8πV 2 ν dνkT = kT g (ν ) dV. (kT ) D 8π 3 2 NP 3. The number of photons depends on the temperature.3. BLACK BODY RADIATION 2 In the lowest approximation E = 3 2 N kT and N = DV (kT ) The condition e −βµ 1 is thus 3 25 √ π 2 . These quanta of the electromagnetic field are relativistic bosons of rest mass 0. which is called the ultraviolet catastrophe. d dν ∝ ν gives a divergent energy. Thus the total number of photons is not conserved and so no constraint can be applied to them. so P V = N kT ∓ 1 √ .22) to get N= 8πV c3 ∞ 0 ν 2 dν eβhν − 1 and E = 8πV c3 ∞ 0 hν 3 dν . The atoms of the perfectly black body absorb all photons incident on them and independently emit photons such that even in thermodynamic equilibrium the number of photons varies significantly.3 Black body radiation Consider a cubic cavity of side L = V 3 inside a perfectly black body. This is a classical limit — it holds when h is very small on a scale defined by m V and the mean energy per particle 3 2 kT . We pass to the continuum limit using (1. c3 which agrees with the classical result of energy kT per normal mode of radiation.3. Therefore µ = 0. This is Planck’s law. 2mπkT h2 3 2 V N 1 if gS = 1. E 2 This is the classical equipartition of energy. Our previous work applies. We must consider a gas of photons in a volume V at a temperature T . eβhν − 1 where = hν . It is true for all real gases except helium at very low temperature and very high density. the walls of which are maintained at a fixed temperature T . We can use our formulae for E and N to get the lowest order quantum correction to the equation of state — recall that P V = 2 3 E . We expect classical behaviour at high T or hν hν low ν . setting µ = 0.

E T S V The entropy density = 4 3 3 σT → 0 as 3S 4 4 3 1 (V σ ) 1 3 ∝ V − 3 at constant entropy. e βµ We assume that µ is finite and positive and µ = E F = kTF is constant at T = 0. The energy E = V σT = 4 4E 3T . n = V ∝ T (see example sheet 3). • electrons in white dwarf stars.26 CHAPTER 3. . For N electrons in a volume V at a temperature T we have N= 0 ∞ g ( )F ( ) d 2m h3 and E = 0 3 2 ∞ g( ) F ( ) d . We see that the energy density is a function of T only. We further assume that µ = E F + O(kT ) for small T . where g ( ) = DV √ . and σ = 15 h3 c3 is Stefan’s constant.7) with µ = 0 we see that 5 4 S = k log Z + and we can evaluate k log Z = T → 0. 3. Using (3. It can be shown that µ = EF + O(kT )2 . GRAND ENSEMBLE METHODS e= 8π E = (kT )4 V (hc)3 = σT 4 . It occurs for • electrons in solids. F( ) = eβ ( −µ) +1 The range 0 ≤ e −βµ < ∞ is allowed without blowup of F . ∞ 0 z 3 dz ez − 1 8π k This is Stefan’s law. = 1 4 3 σT . pronounced at low temperature. The region e −βµ 1 is the classical realm and we expect that quantum mechanical effects will be most 1.4 Degenerate Fermi gas This is the extreme quantum limit. • neutrons in neutron stars and • nucleons in nuclear matter. 1E 3V 1 ∂E The pressure P can now be calculated from P = − ∂V = S N 3 Finally the density of photons. D = 4π and 1 .

4. 3N 2DV 2 3 ∝ V −3 . The Pauli principle requires particles to have non-zero momentum and so there is pressure even at T = 0. until the Fermi energy is reached. In fact 2 2 P = 33 3 N V 5 3 h2 ∝ m−1 2m and so at T = 0 the pressure is much bigger for lighter particles.13) E = DV We can solve (3. ∂ ∂T ∞ We want to find C = ∂E ∂T of both of the above to find V (which is proportional to the usual C V ). Classical kinetic theory The equation of state P V = 2 3 E becomes P V gives P → 0 as T → 0. 3.12) (3. DEGENERATE FERMI GAS Now lim eβ ( −µ) 27 T →0 = 0 ∞ < EF > EF . 5 2 3 = const. so that 1 0 < EF > EF .12) for E F to get EF = 2 2 3 EF 5 = 3 5 N EF .4. ∂T . We take ∞ 0 C= g( ) ∂F ( − EF ) d . We can now perform the integrals for N and E to get 2 N = DV 2 3 EF 3 (3. one spin up and one spin down) with increasing energy according to the Pauli principle. At T = 0 it is energetically most favourable for electrons to fill up the one particle energy eigenstates (two electrons at a time.1 Heat capacity at low temperature We have ∞ 0 N= E= 0 g ( )F ( ) d and g ( )F ( ) d .3. T →0 lim F ( ) = Thus at T = 0 F ( ) = θ( − EF ). E F is the highest energy occupied at T = 0.

This will be at T = T B . This leaves a convergent integral whose value is C∼ T π2 Nk . so that −βE F → −∞ as T → 0. given by √ N ˆ = D πf (0) . We approximate g ( ) ≈ g (EF ) and µ ≈ EF . ez−βµ − 1 2 where ∞ f (−βµ) = n=1 enβµ n2 3 . C ∼ g (EF )k 2 T ∞ −∞ z 2 ez dz (ez + 1) 2. π2 3 . 2 TF so that where TF (the Fermi temperature) is defined by E F = kTF . f (−βµ) is convergent iff βµ ≤ 0. the Bose-Einstein denominator does not vanish for any point in the range of integration. Our equation for N is OK as T decreases since f (−βµ) can increase by βµ getting less negative. The 2 factor in g ( ) assigns 0 weight to the particles in the = 0 state. GRAND ENSEMBLE METHODS ≈ EF At low temperature ∂F ∂T is very like a delta function and so the region dominates the integral. This approximation is expected to be good for T TF .612) at βµ = 0 and decreases monotonically as βµ decreases through negative values from βµ = 0.5 Bose-Einstein condensation Recall the result N = DV 0 ∞ 1 2 d −1 eβ ( −µ) = DV (kT ) 3 2 ∞ 0 √ 1 z 2 dz 3 π 2 = DV (kT ) f (−βµ). We therefore approximate the lower limit of the integral by −∞. +1 2( − EF )2 = g (EF )k 2 T z e dz (ez + 1) We know that EF is finite and positive. This gives us g (EF ) kT 2 ∞ 0 ∞ −βEF 2 z C∼ d ( − EF ) eβ ( −E F ) e β ( −E F ) 2. When βµ reaches 0 problems occur as f cannot increase any more. It is easy to see that f has a maximum value (of 2.28 CHAPTER 3. Why? 1 More care is called for in the passage to the continuum limit. 3. This would be no problem for . Assuming this. Now suppose that N V is fixed. ˆ (kTB ) 3 2 D =D V 2 Our equation for N appears to fail beyond this.

6. so that the integral for can be done to get N 8π x3 = 3 3 (mc2 )3 F . producing an electron gas. WHITE DWARF STARS 29 fermions (which have the Pauli principle). and ρ ≈ 107 ρsun . Now kTF = EF = h2 2m 3N 8πV 2 3 ≈ 2 × 106 eV. N V ˆ (kTB ) 3 2 .6 White dwarf stars White dwarfs are abnormally faint: they are stars in which the hydrogen supply has run out so that they are composed mainly of helium. For T < T =D B B then NC 3 ˆ (kT ) 2 =D . We write N = N 0 + NC . but no law stops bosons from condensing into the = 0 state if need be. where N0 is the number of particles in the = 0 state and N C is the number of particles in > 0 states. T ≈ Tsun = 107 K M ≈ Msun . For T ≈ 10 7 K. Treating the electrons relativistically we have 8π N = 3 3 V c Let √ 2 EF d ( mc2 2 − m 2 c4 ) 2 . V c 3 . N V 1 − m2 c4 = mc2 x. 3. We regard them as being a mass of helium at T ≈ 10 7 K and under extreme compression. To a good approximation this may be viewed as a Fermi gas in the degenerate (negligible T ) limit. V giving NC = N T TB 3 2 . kT ≈ 103 eV is much greater than the binding energy of electrons to helium. This may be seen by computing T F . 11 Since N T. so that thermal collisions completely ionize all the atoms. V For T = TB . the natural temperature scale of the problem (the Fermi temperature). Typically. defining T .3. V is so large TF ≈ 10 K The helium nuclei neutralise the charge of the star and produce the gravitational attraction which counteracts the extreme zero-point pressure of the electron gas (which dominates the zero-point pressure from the nuclei). We can calculate the total energy E = E elec (R) + Egrav (R) and find the radius R 0 of the star by minimising this. Then for T > TB √ N = D 2π f (−βµ).

GRAND ENSEMBLE METHODS The integral for the energy density is E 8π = 3 3 (mc2 )4 V c and so E N xF 0 x2 dx 1 + x2 = 3 2 4 mc xF + R−1 . Hence 1 xF 2 3 + O(x− F ) . R R where a. We know that V ∝ R and so xF ∝ E= a γ + bR − . . which leads to the Chandrasekhar upper limit on the mass of a white dwarf. This only has a minimum if γ ≥ a. b and γ are positive constants.30 CHAPTER 3.

This result shows that E = 3 2 N kT .q) .1 Introduction Recall the result Z = z N for a system of distinguishable non-interacting particles. q ) is the classical Hamiltonian as a function of classical variables p and q . V ∂ log Z ∂β = −N V Z= i=1 d3 p d3 q h3 e −β È N i=1 H (pi .qi ) .q) .1) ˆ (ˆ where H p. The average for the system of a physical variable f (p.Chapter 4 Classical statistical mechanics 4. q)e−βH (p. 31 . q ˆ) is the quantum mechanical Hamiltonian and H (p. d3 p d3 q e−βH (p.q) ∂ log z ∂β V agrees with the result ∂ log z ∂β . (4. qi ) is the Hamiltonian of the N -particle system. which is the common classical result of energy per degree of freedom per particle. It can be shown that in the classical limit z= r e −β r = r r|e−β H |r ˆ gives z = h −3 d3 p d3 q e−βH (p. q) is f (p. where N i=1 H (pi . q) = We can see that H (p. p2 For a monatomic gas with H = 2 m we have V z= 3 h 1 2 kT dp e − βp 2m 2 3 V = 3 h 2πm β 3 2 . q) = − E=− Now N d3 p d3 q f (p.

Now H= α ˙ and pφ = I sin2 θφ. L= 1 2 mx 2 I θ + sin θ φ (V ≡ 0 in this rigid case. Thus z = zt zr = V 2mπkT h2 3 2 8π 2 IkT h2 1 . Each molecule has a dipole moment m and gives 1 Which ought to be expected! . This gives E = 5 2 N kT . ˙ pθ = I θ q ˙α pα − L = p2 p2 p2 φ + θ + . d pd qe 3 3 2 −1 2 mβ p =V 2mπ βh2 3 2 = 8π 2 IkT . To apply (4. 2m 2I 2I sin2 θ and z= where 1 zt = 3 h and 1 zr = 2 h dpφ dφ dpθ dθ e β −2 I φ p2 θ + sin2 θ p2 1 h5 d2 p d3 q dpφ dφ dpθ dθ e−βH = zt zr . CLASSICAL STATISTICAL MECHANICS 4. ∂L We define the generalised momenta p α = ∂ q ˙α : p i = mx ˙ i. pφ and φ integrals and then doing the θ integral. h2 We evaluate the above integral by first doing the p θ . 4.1) we need the Hamiltonian H .2 Diatomic gases We will study the rigid dumbbell model as shown.3 Paramagnetism Each molecule of an N molecule solid acts as a little magnet fixed at its own lattice site and free only to rotate about it. To find the Hamiltonian we first write down the Lagrangian 2 ˙2 ˙2 ˙2 + 1 .32 CHAPTER 4.) Note that the Lagrangian splits neatly into translational motion of G and rotational motion about the centre of mass. We want to write down the partition function Z = z N .

0. We can calculate the magnetisation of the solid: M = (0. Thus ∞ zr = l=0 2 (2l + 1)e− l(l+1)Tr T (4. . . . .4 Specific heats The result E = 5 2 N kT for a diatomic gas is found to be accurate at sufficiently high temperatures. The contribution to the energy from rotational motion is Tr . . 0. . B ). therefore N kT if T T r or 0 if T For high temperatures. E is found to be 3 2 N kT at lower temperatures. We will look at zr . This is explained by quantum mechanics.4.2) are exponentially small and we take only the l = 0 term to get z r ≈ 1 — there is no rotational contribution to the energy (or heat capacity). For Tr T (most gases at normal temperatures) we can turn this sum into an integral to get zr = T = Tr 8π 2 h2 kT as before. −l + 1. For small y (high T ) we thus have Curie’s law. l − 1. Tr is typically about 50K and is experimentally accessible. It is as if the rotational degrees of freedom are frozen out. using Hr = 21 I L (the quantum mechanical angular momentum operator). . M ) = N M . 2 We have Z = z N and z = zr zt . Now Hr |l m = l(l + 1) 2I 2 |l m for m = −l. 1. Only the third component m cos θ is non-zero: m cos θ = β −1 ∂ log z ∂B = m coth y − y −1 . . l for each of l = 0. There is an extra term in the Hamiltonian Hv = 1 2 mω 2 Q2 P + m 4 and an extra factor z v in the one particle partition function. 2. .4. E rises to 7 2 N kT due to vibrational motion along the axis of the dumbbell. M = 1 3 N my = 4. SPECIFIC HEATS 33 a contribution −m · B to the energy when in an applied magnetic field B = (0. β N m2 B 3kT . For T r T all of the terms with l = 0 in (4.2) with Tr = 2Ik . Now H= and so z= 2πI 2 sinh y 2π h2 β y p2 p2 p2 φ + θ + − mB cos θ 2m 2I 2I sin2 θ where y = mβB .

(as N − 1 ∼ N ) and N 2 f (T ) 2V N 2 f (T ) + stuff which does not depend on V . q a +q b 2 =1+ N (N − 1) 2V 2 d3 qa d3 qb λab This integral can be evaluated using the change of variables x = qa − qb . We suppose that λrs = e−βφ(Rrs ) − 1 is small (weak interaction) and we treat our results to lowest non-trivial order. V β we get with f (T ) = 0 ∞ 1− N f (T ) 2V 4πy 2 dy e−βφ(y) − 1 . Then K = 1 + log Z = N log V + Using βP = P V = N kT ∂ log Z ∂V and y = d3 y e−βφ(y) − 1 ≡ V f (T ). (4. CLASSICAL STATISTICAL MECHANICS 4. 2m r<s where Rrs = |qr − qs |. Then K = V −N r d3 qr 1+ r<s φ(Rrs ) (no summation). We use d3 pr d3 qr −βH e h3 r =1 2πmkT h2 3N 2 N Z= = V NK = Z0 K.5 Weak interparticle forces Consider a classical gas of N molecules with Hamiltonian N H= r =1 p2 r + φ(Rrs ).3) If we take a specific φ we can recover the van der Waals’ equation of state (see page 15). N where Z0 = z0 and z0 is the partition function for one molecule of a spinless gas of N non-interacting molecules and K = V −N r d3 qr e−β È r<s φ(Rrs ) . . so that d3 qa d3 qb λab = V defining f (T ).34 CHAPTER 4.

5) is completed by setting A = N 2 a and B = N b. . 35 Then the contribution to f (T ) from region 1 where φ is infinite is d 0 4πy 2 dy = −2b.4. Note that B is the volume of all the molecules. − βH 3 e d p 4πv 2 dve− 2 βmv = N 1 2 f (v ) dv. Now f (T ) = −2(b − βa) and (4. The constant ∞ is (of course) chosen to make 0 f (v ) dv = 1.6 The Maxwell distribution p Consider a gas with (one particle) Hamiltonian 2 m. b For N V small we can (approximately) take 1 + equation to get Nb V to the left hand side of the P+ N 2a V2 (V − N b) = N kT and the correspondence with (2. We can define averages in the obvious way. that is g (v ) = 0 ∞ 1 2 f (v )g (v ) dv. The number of molecules in the region p → p + dp is 2 n(p) d3 p = Now N= n(p) d3 p = m3 C β p2 N e−βH = Ce− 2m . 4. a and b are positive constants. f (v ) is the probability of finding a particle with speed in v → v + dv . which defines the Maxwell distribution of speeds.6. f (v ) = const × v 2 e− 2 βmv . THE MAXWELL DISTRIBUTION Consider the φ shown. For large T and weak attractive forces the contribution from region 2 is ∞ d 4πy 2 dy (−βφ(y )) = 2βa.3) becomes P+ N 2a V2 V = N kT 1+ Nb V .

CLASSICAL STATISTICAL MECHANICS .36 CHAPTER 4.

so √ 2πne−n nn . so that n! = nn+1 0 n(log v −v ) 1) n −1− (v− 2 2 ∞ en(log v−v) dv. e ∼e . ∞ −∞ e− nu2 2 du = 37 . Now for large n. tn e−t dt = 0 ∞ en log t−t dt.Approximations Stirling’s formula We derive the approximation for Stirling’s formula from the Γ function using Laplace’s method. Then ∞ 0 tz−1 e−t dt n! = Let v = t n. n! ∼ nn+1 e−n This is Stirling’s formula. Recall that ∞ 0 Γ(z ) = and Γ(n + 1) = n!.