# Statistical Mechanics

Dr Alfred Huan

Contents
1 What is Statistical Mechanics?
1.1 Key Ideas in the module . . . . . . . . . 1.2 Why do we study Statistical Mechanics? 1.3 De nitions . . . . . . . . . . . . . . . . . 1.3.1 Microstate . . . . . . . . . . . . . 1.3.2 Macrostate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3
3 4 4 4 5

2 Distribution Law

2.1 Distribution Law & Ensemble Approaches . . . . . . . . . . 2.2 Computation of Most Probable Macrostate Distribution, nj . 2.3 Problems in Derivation of Maxwell-Boltzmann Distribution . 2.3.1 nj not being always large . . . . . . . . . . . . . . . . 2.3.2 What about the other macrostates? . . . . . . . . . . 2.3.3 What happens when we have many energy levels? . . 2.4 Time Evolution of system of microstates . . . . . . . . . . . 2.5 Alternative Derivation for Maxwell-Boltzmann Distribution . 2.6 Examples of Counting in Statistical Mechanics . . . . . . . . 3.1 3.2 3.3 3.4 4.1 4.2 4.3 4.4

6 7 11 11 11 14 15 16 17

6

3 Indistinguishable Particles

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bose-Einstein Distribution for Bosons . . . . . . . . . . . . . . . . . Fermi-Dirac Distribution for Fermions . . . . . . . . . . . . . . . . Alternative Derivation for Bose-Einstein and Fermi-Dirac Statistics First Law . . . . . . . . . . . . . . . . . . . . . Second Law . . . . . . . . . . . . . . . . . . . . Third Law . . . . . . . . . . . . . . . . . . . . . Calculation of Other Thermodynamic Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

20

20 21 23 24 26 27 29 30

4 Statistical Mechanics and Thermodynamic Laws

26

5 Applications of Maxwell-Boltzmann Statistics 6 Paramagnetic Systems

5.1 Classical Perfect Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 5.2 Harmonic Oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 5.3 Equipartition Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 6.2 Spin 1 Quantum System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 2 6.3 Spin J Quantum System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 1

33 42

CONTENTS

2

6.4 Adiabatic Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

7 Applications of Fermi-Dirac Statistics

7.1 Fermi-Dirac Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 7.2 Free Electron Model of Metals . . . . . . . . . . . . . . . . . . . . . . . . . 53 8.1 8.2 8.3 8.4 Introduction to Black Body Radiation Bose-Einstein Distribution . . . . . . . Planck's Radiation Formula . . . . . . Radiation emitted by blackbody . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

51 59

8 Applications of Bose-Einstein Statistics

59 60 61 63

9 The Classical Limit

9.1 The Classical Regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 9.2 Monatomic Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 9.3 Gibbs Paradox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 10.1 10.2 10.3 10.4 10.5 Introduction . . . . . . . . . . . . . . . . Number of impacts on Area A of a wall . E usion of molecules . . . . . . . . . . . Mean Free Path,  . . . . . . . . . . . . Collision cross-section, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

66 71

10 Kinetic Theory of Gases

71 73 74 75 76

1 Key Ideas in the module We begin with the introduction of some important ideas which are fundamental in our understanding of statistical mechanics. 4. We will deal with single particle states and no ensembles. 2. 8. and static macroscopic variables. weakly interacting particles and the behaviour for distinguishable degeneracies and indistinguishable particles. Bose-Einstein In the 3rd year course.Chapter 1 What is Statistical Mechanics? 1. 3 . The list of important terms are as follows : Microstate Macrostate Partition Function Derivation of Thermodynamic laws by Statistical Mechanics The idea of interaction a particle system . 3. 5. we shall restrict to equilibrium distributions. Entropy : Disorder Number 7.external environment b particle within a system of many particles 6. Temperature 9. Distributions : Maxwell-Boltzmann. Equipartition Theorem : the energies are distributed equally in di erent degrees of freedom. 1. Fermi-Dirac.

the possible energies for each particle is : ~2  2 E = 2m L2 n2 + n2 + n2  x y z For example in an ideal gas. for e. For many particles.z and momentum Px. they do not a ect each other's energy levels.2 Why do we study Statistical Mechanics? We apply statistical mechanics to solve for real systems a system for many particles.3. WHAT IS STATISTICAL MECHANICS? 4 1. . Py .. whereas statistical mechanics gives a microscopic perspective of heat in terms of the structure of matter and provides a way of evaluating the thermal properties of matter. Pz  will give 6N degrees of freedom and this is put in a phase space representation. i. Classical Mechanics The position x.CHAPTER 1. Thermodynamics is concerned about heat and the direction of heat ow. how the particles are distributed over the energy levels. Each particle contains a certain energy. From quantum mechanics. the solution will take the form : total = linear combination of a 1b 2c3::: where a means a particle in state a with an energy Ea . We can easily solve the Schrodinger's equation for 1 particle atom molecule. At T 0. the system possesses a total energy. . 1.. how is E distributed among the particles? Another question would be. heat capacity. We apply statistical mechanics to provide the answer and thermodynamics demands that the entropy be maximum at equilibrium. Quantum Mechanics The energy levels and the state of particles in terms of quantum numbers are used to specify the parameters of a microstate. We have 2 perspectives to approach in looking at a microstate : 1. and there are many parameters for each state. V.3 De nitions 1. For example.E.1 Microstate A Microstate is de ned as a state of the system where all the parameters of the constituents particles are speci ed Many microstates exist for each state of the system speci ed in macroscopic variables E. we consider particles con ned in a cubic box of length L.e. We assume that the molecules are non-interacting. 2.. N.g.y. So.

We adopt the a priori assumption which states the macrostate that is the most stable contains the overwhelming majority of microstates. The macrostate includes what are the di erent energy levels and the number of particles having particular energies. for e. . It contains many microstates. WHAT IS STATISTICAL MECHANICS? 5 1..V. for an ideal gas. the magnetization.CHAPTER 1.g.3.2 Macrostate A Macrostate is de ned as a state of the system where the distribution of particles over the energy levels is speci ed. the equilibrium tends towards a macrostate which is the most stable. The equation of state for the system relates the 3 variables to a fourth. In statistical mechanics. However. N : Number of particles and E : Energy.V. The stability of the macrostate depends on the perspective of microstates.N or E. In the equilibrium thermodynamic state. V : Volume. where P : pressure.T or P. T : Temperature.V. the equilibrium macrostate is unknown from thermodynamics. PV = NkT We also have examples of other systems like magnetic systems where we include M.N. This also gives the probabilistic approach in physics. we only need to specify 3 macroscopic variables P.

If they are indistinguishable. i. the volume V.Chapter 2 Distribution Law 2. There are three approaches in de ning the system and we adopt the three ensemble approaches. degeneracy gj .e. every microstate is a macrostate. the sites being distinguishable e. To enumerate . non-localised. 1. Distinguishability : small and identical particles cannot be distinguished during col- 6 . Microcanonical Ensemble : Isolated system with xed energy. E and number of particles. Let us now consider independent particles. energy E and total number of particles. with energy levels xed and the particles don't in uence each other's occupation. 2.at constant temperature with xed N. However. N. we need to decide whether the particles are distinguishable or indistinguishable. which is maximum at equilibrium. The disorder number. For an isolated system. lisions. localised particles are distinguishable because they are xed in space.g. The energy levels are xed by each particle's interaction with its surroundings but not particle-particle interactions. N are xed. is de ned as the number of microstates available to a macrostate. 3. Grand Canonical Ensemble : Nothing is xed. and occupation nj . The disorder number is related to entropy. in a paramagnetic solid. Mutual interaction exist only for energy exchange during collisions. and is very large for the equilibrium macrostate. Each level j has energy "j .1 Distribution Law & Ensemble Approaches The distribution law is dependent on how we de ne the system. Canonical Ensemble : The system is in a heat bath.

Enumerate . n ! 2. Considering the case for distinguishable particles : 1. we must nd the most probable macrostate distribution. nn1.e. DISTRIBUTION LAW X N = nj Xj Total E = nj "j j 7 The A Priori assumption states that all microstates of a given E are equally probable thus the equilibrium macrostate must have overwhelming .1 1 1 For the second level. nj To nd the nj at equilibrium. N! The number of ways to assign n1 particles = n !N .2 Computation of Most Probable Macrostate Distribution. i. N.2 2 1 2 . i. We distribute N particles over the energy levels: For example. how probable for a particular macrostate to be the one corresponding to equilibrium. Ω Equilibrium state Macrostates The disorder number is proportional to thermodynamic probability. 2. ! The number of ways to assign n2 particles = n !N . n ! 2.e.CHAPTER 2. For the rst level.

3 Level j has degeneracy gj so any of the nj particles can enter into any of the gj levels. DISTRIBUTION LAW Number of ways = 8 N! N !  n !N . N."j are unchanged.CHAPTER 2.. it means that the energy levels.4 2.nn1. Maximise by keeping xed E. Number of ways = gjnj For all levels.7 2.8 d =0  dln  = 1 d =0 Hence we can choose to maximise ln instead of ln = ln N ! + X j nj ln gj . N 2. Number of ways = 2. X nj "j = E 2. n1 ! 2 1 2 N! = n1 !n2!::::nj !::::: = QN ! n! j j 2. Maximise subject to X nj = N and At maximum. n !  :::::: n1 !N . X j ln nj ! 2.6 Y j gjnj Y gjnj Total = N ! n ! j j 2.10 .9 By Stirling's approximation for large N : ln N ! = N ln N . For a xed V.5 2. These constraints are introduced through Lagrange multipliers.

Therefore X j g ln nj dnj = 0 j 2. nj lnnj  g nj ln nj  j by assuming large nj also.11 = n ln n .14 X dnj = 0 for xed N X j "j dnj = 0 for xed E Note "j does not vary.13 nj ln gj .N + ln N + 1 + N ln N . n N 1 = N ln N . nj ln nj + nj  2. N + = N ln N + = N ln N + X j X X j j nj ln gj .CHAPTER 2. N  1 = . DISTRIBUTION LAW Proof : ln N ! = = N X 9 Z1 N 1 ln n ln ndn 2. which is not always true. N + 1 Or N X 1 ln n 4 n = Z 1 = N x ln x .15 .12 n where 4n = d N  = dx. Now we come back to ln ln  N ln N . X g dln  = ln nj dnj . ln N = N ln N . x 1N + N ln N 1 . N + 1 N ln n + N ln N d n  1 N N N 1 2. dnj  = 0 j j It is subjected to the constraints : j 2.

"j Maxwell-Boltzmann Distribution Z where z is the partition function g ln nj  + . because k 0.16 2. then T " "  e. "j j gj e = Ngj e. kT "2 "1 . the upper levels get piled by particles provided 4"  kT 2. nj = 0.19 Otherwise. For T 0.1 mol. "j = 0 j 2. kT "1 n1 g1e.1  and T is the temperature. so we get We choose and such that X j j dnj = 0 "j dnj = 0 2. kT . i. Since "2 "1. "j = P Ngj. we get nj ni for gj = gi and T 0 For T = 0.CHAPTER 2. "1 e kT " 2.20 for g1 = g2 .e. "j e. all the particles are at ground state. kT "2 e. population inversion leads to negative temperature. In lasers. the population of upper levels remain small. 2 n2 = g2e. kT 1 = . except for n1.21 0 . "j dnj = 0 j The equilibrium macrostate becomes : nj = gj e e. "2 .23JK . DISTRIBUTION LAW Now we introduce the constraints : 10 X .18 1 We will prove later that is equal to kT where k is the Boltzmann constant k = 1:38  10.17 X j g ln nj  + . For "j "i. kT e. kT . "1 T T 2 1 2.

= N ln N .3 Problems in Derivation of Maxwell-Boltzmann Distribution There are 2 inherent problems present in the derivation of Maxwell-Boltzmann distribution : 1. Many of the higher levels will give ln nj !  0.3. the lower levels will have large nj but higher levels will have small nj . However. N .23 2.25 . If many levels exist.3. Stirling's approximation can be used. nj = nj + nj  Since ln nj  = N ln N . so The most probable nj Maxwell-Boltzmann Distribution  P j n n  nj = P j in j ~ j i j Ngj e kTj nj = P . nj  because many of the lower levels have large nj . 2.22 2. nj is not always large.1 nj not being always large P If the total number of levels j 1 is small and T is large enough. DISTRIBUTION LAW 11 2." 2. The other macrostates of the same xed energy E.24 nj ln nj . Note that 0! = 1! = 1. P 2.CHAPTER 2. the higher levels contribute very negligibly to ln nj ! so that we can apP proximate again by nj ln nj . 2.2 What about the other macrostates? How does the change when we consider other macrostates? Consider the changes from the most probable macrostate nj to another distribution  nj . What are their s? Are they small enough to be neglected? The equilibrium distribution nj should be weighted against of all possible macrostates ~ i. nj  X j X j nj ln nj 2. then every level j has many particles so that for large nj ."j  gj e kT .

if nj  pnj . X j ln nj  = ln nj  . ln N  nj + kT "j nj . nj ln nj + nj + nj ln nj + 1   nj   + :::    2 n j j X X 1X n 2 = ln nj  .. nj + nj ln nj + n j . in deriving properties of the system. nj . by comparison. Now as before. 2  nj  + :::: j j j j where 2. the uctuations allowed are very small.  nj .  P  nj 2 nj  2. 2  nj  + ::: Z j j j X X n 2 1 X = ln nj  .27 It follows that nj  will be small. if  nj  1020. 1  nj  + ::::::  2 j j nj  = nj e 2  1 X 2.28 Since nj are large. 1  nj  + ::: Z j 2 j j j X n 2 = ln nj  . = N ln N .g."j X j nj = 0 nj "j = 0 and Thus. Thus nj nj .  ln nj  = N ln N . nj = N e kT  Z . . minor uctuations will not matter very much. Ultimately. e. 12 X X j nj + nj  lnnj + nj    nj + nj ln nj + ln1 + nj    nj j 2 X n = N ln N . kT  . 1010 for nj  to be large.  "j 1 X n 2 nj ln N .26 neglecting gj for the purpose of simplifying the discusion. so the uctuation is minor.CHAPTER 2. DISTRIBUTION LAW Expand about nj . nj ln nj . 1   nj   + ::::    2 n j j j 2 X = N ln N .

and we assume is the same for all j. In either case.32 . DISTRIBUTION LAW Ω 13 (i) (ii) (ii) Macrostates where i represents nj nj and ii nj large. Then. n2 2 n2 d n2 ::::::: nj  1  2 2. we see that nj  will be small when the uctuation nj is large Now we come back to the total number of macrostates : The total number of macrostates consistent with constraints : n j nj j .CHAPTER 2. n2 1 n1 Y j d n1 e Z . Alternatively.31 N so nj needs to be small in order for nj  to be signi cantly large. pj X  nj  e  = nj   Z . 2 N  which is small unless 2 1 .30 1 2.29 2. 2. we can also argue the problem out by the choice of a fractional deviation 2 = j njj  j .e. n X  nj 2 X 2  nj = j nj  j = 2N Hence nj  = nj e. N . i.

CHAPTER 2. gi and mi can be made large in this case.33  nj  N  G  G ln Now so ln X  ln nj  + G ln N . 2. G.34 2. This means that all  P The total number of microstates are not mainly due to the maximum term nj ."j e kT .." Ngi e kTi = P . They are grouped into bunches of gi with mi occupation.  Ngie kTi mi = P . 6= nj . DISTRIBUTION LAW Stopping the product when nj = 1  14 If total number of energy levels.37 .e.35 2." 2.2..3 What happens when we have many energy levels? If nj  1 for most levels. X 2 2.. then nj  N and j = 1. i. G lnG  2 2 ln N ln G 2.. N . G ."j  i gi e kT . then we have many energy levels G macrostates look roughly the same. G is very much less than N. i=1 first bundle 1 of g1 bundles The energy levels are so closely spaced such that they are grouped into bundles of g i levels i=2 2nd bundle of g2 bundles Using the previous derivation by counting mi energy states. ~ We resort to the following trick : Energy levels are closely spaced in groups because G is large.36 ln nj   N ln N  X  ln nj   so the most probable microstate has overwhelming . Hence Stirling's approximation is invalid in this case..3.

the  variation at each j level appears large. So for 1023 electrons : The Transition Speed = 1012  1023 = 1035s. for every experiment.39  have roughly the same properties and distribution close to the MaxwellBoltzmann distribution ignoring minor changes. We 2 also x E and T for a xed macrostate i. The treatment tries to obtain the average ni over  many di erent but similar macrostates. and each electron have a spin of  1 .e.38 Ne kTj = P . 2. After a series of attack on our fundamental assumptions. k . DISTRIBUTION LAW 15 We take the time average value of nj over a large number of di erent macrostates. X k :Bk E f T  k . we come to establish the basis of statistical mechanics : All macrostates of signi cant probability i.  mi = mi  g i . we consider paramagnetic system where the atomic spins interact with the external magnetic eld .e."j j e kT mi  = N !  ln nj  = ln  2. So we need a statistical assumption that what we observe corresponds to the real average over all microstates.4 Time Evolution of system of microstates How many microstates does the system explore within the time for the experiment? For example. Hence.CHAPTER 2. one can only explore a small proportion of microstates. the energy levels are discretized.Zeeman E ect. the number of microstates :  eN ln N e1035 o 1035 So it takes forever to go through all the microstates. nj .each atom The change in microstates will mean that the spin energy of the electron is exchanged between neighbouring atoms so the atoms interchange energy levels. The typical time for spin ip is 10.12s." 2. Note in Quantum Mechanics. large Y gimi i mi ! X 2.1 Now. Otherwise with nj  1.40 By grouping closely spaced levels. The particles inside this system = 1023 particles are assumed to be non-interacting. the time-averaged variations are small and the Maxwell-Boltzmann distribution describes nj very well.

Tkl!ij = Tij!kl 2.43 Ri out = Ri in Therefore it follows that : 2.k. which is true for almost all transitions.k.44 2. Therefore. the rate Rij!kl = Ni Nj Tij!kl where Tij!kl is the transition probability.45 Ni X j. j E l k E i Both the transitions from j to l or i to k requires the same amount of energy. X j.42 Ri in = At equilibrium.41 Ri out = Ni The total rate in to level i : X j. V and N xed. The total rate out from level i : 2.l Nk Nl Tkl!ij 2. DISTRIBUTION LAW 16 2.l Nj Tij!kl = X j. Suppose we change from 1 macrostate to another macrostate.46 . keeping E.l Nj Tij!kl 2. We need a transition from i to k to be accompanied by another transition from j to l.k.k.l Nk Nl Tkl!ij Now we invoke the time-reversal invariance.5 Alternative Derivation for Maxwell-Boltzmann Distribution We present an alternative method to derive Maxwell-Boltzmann distribution from the perspective of equilibrium.CHAPTER 2.

50 2.48 2.CHAPTER 2. 2.3. which is the Maxwell-Boltzmann Distribution. "i + 17 2. We want to calculate the number of macrostates and microstates in this system.0.0 macrostate.5 and 0.4.6 Examples of Counting in Statistical Mechanics Problem 1 : Given 6 distinguishable particles. k and l. j. 1. 2 energy levels one with a degeneracy of 2 and the other degeneracy of 5. we require NiNj = Nk Nl Principle of Detailed Balancing or ln Ni + ln Nj = ln Nk + ln Nl and "i + "j = "k + "l is required for energy conservation. There are 7 macrostates namely : 6.51 Ni = e e. "i 2. 3. 5.6. 4. = 26 ways .1. both equations will lead to ln Ni = . DISTRIBUTION LAW Thus for every combination i. Since Ni N "i.47 2. we present a diagramatic scheme of the problem : g2 = 5 g1 = 2 6 particles We also note importantly that there are no equilibrium macrostate in this problem because all the macrostates have di erent total energies. First of all.49 2.2. Now we come to calculate the microstates : For 6.

0.0. Hence we can talk about a xed macrostate. Problem 2 : Given 5 energy levels. we apply the weighted average for microstates.1 macrostate.0. Let us look at the rst 3 levels : i a particle moves from the 3rd level down to the 2nd level and ii a particle moves from 1st level to the 2nd level.CHAPTER 2. i.0. the change in energy is balanced by exchange of particles for the 2 processes. because the number of microstates in either macrostate is not overwhelming enough.0.52 j i Therefore. P P n = nj i nj  ~ 2. 1 =6 C5 1 C0 1 C1 0 C0 0 C0 = 6 ways =6 C4 2 C2 0 C0 0 C0 0 C0 = 15 ways 2 At equilibrium. We consider the particles to be weakly interacting.1. using the example above. 18 ! n n = n Nn ! g1 1 g2 2 1! 2 6!  25  5 ways = 5!1! Subsequently. Macrostate 4. we can calculate for the remaining macrostates.2. each separated by E . Now we try to calculate Now we consider the schematic picture of this problem : E E E E Macrostate 5. DISTRIBUTION LAW For 5. they exchange energy but do not a ect each other's energy levels. n1 = 5  6 + 4  15 ~ 6 + 15 90 = 21 . In this case.0.e.

.CHAPTER 2. DISTRIBUTION LAW n2 = 0  6 + 2  15 ~ 6 + 15 30 = 21 n3 = 1  6 + 0  15 ~ 6 + 15 6 = 21 n4 = 0 ~ n5 = 0 ~ Note : 19 X which is the same as P j nj = 6 ~ nj for any macrostate.

Symmetric  : s = E.2 .1 s = a 1b2c3 + a 2b1c3 + a 3b 2c1 + a 1b3c2 + a 2b3c1 + a 3b1c2 2. and so are all linear combination of the permutations. Antisymmetric  a = X p . N  3. For example. The exchange of quantum particles cannot be detected. we get six possibilities : a 2b1c3::::: a 2b3c1::::: a 1b3c2::::: etc. we get X p P 1. for three particle permutations.1pP 1. We can write a many particle wavefunction as a linear combination of a 1b2:::::: plus all the possible permutations. N  20 3. 2. 1. for 3 particles. The 's of the individual particles combine to form a total . 2.g.1 Introduction Why are particles indistinguishable? This is a quantum mechanical e ect. All particles are identical so all permutations are solutions to Schrodinger's equations. and they are non-localised. :::::. ::::. There are two physically meaningful solutions.Chapter 3 Indistinguishable Particles 3.

Macrostates Microstates Unless g 1 for some levels. meaning all permutations cancel in pairs. Note when 2 fermions are in the same state. a = b . by the clumping of levels together to make gi large. 1! mi !gi . a 3b2c1 . We can then count microstates  1 because the states in i are distinguishable. for example.a 1.2 Bose-Einstein Distribution for Bosons There is no limit to the occupation number at each level of quantum state. Note that = 1 for all macrostates because of the indistinguishability of particles.1 every time we exchange two particle states. INDISTINGUISHABLE PARTICLES 21 where P is the permutation operator.. To put mi particles in gi states. 2. 3 = . etc. for 3 particles. by arranging mi identical particles and gi. a 1b3c2 + a 2b3c1 + a 3b1c2 If we exchange 1 and 2.1P means . 1! The number of ways = mi !ggi.4 X i X i mi = N mi"i = E .1 identical barriers on a straight line. 1. Bosons : Symmetric wavefunctions with integral spins. . we get s = a 1b 2c3 .CHAPTER 3. 1! mi i Therefore. 3 There are two kinds of quantum particles in nature: 1. imagine arranging mi particles into gi states. a 2b1c3 . 1! To maximise subject to the constraints : 3.3 3. 2. Fermions : Anti-symmetric wavefunctions with half-integral spins. and hence we obtain Pauli's Exclusion Principle. For example. 3. E. then a 2. = mi + gi . total  is zero.g. + . the more the better. In fact.

11  We will nd later that = kT where  is the chemical potential. So X i i dmi = 0 "idmi = 0 3. e mi "i 3. 1 which is the Bose-Einstein Distribution. 1 lngi . mi + gi . gi .6 3.8 X i ln mi + gi  + . dmi = 0 3. 1 . 1 . gi . . mi + gi = e. ln mi! . 1! . Therefore. "idmi = 0 mi Then.10 mi = e. eg"ii . INDISTINGUISHABLE PARTICLES Therefore. ln = 22 X i X i X i lnmi + gi . 1 mi + gi lnmi + gi .7 X mi + gi ln m dmi = 0 i i Add in : X . The systems at chemical particle number equilibrium have the same . mi ln mi . nj = e. gi ln gi 3.5 1   for gi. dmi ln mi . 3. mi dln  = X i dmi lnmi + gi + dmi . e 1"j . 1 . 1 lnmi + gi . 1 where = kT to be proved later. 1! mi + gi .9 3. lngi . mi ln mi + mi . 1 Averaging over the gi levels in each clump. Two systems at thermodynamic equilibrium have the same temperature T.CHAPTER 3.

mi  + .13 = Y i mi!gi . "idmi = 0 m i .dmi lnmi . mi = ln dm =0 i X mi i 3. etc. we allow 0 or 1 particle to each quantum state.14 Therefore. So X i i dmi = 0 "idmi = 0 3.15 Now we add in : X . Either there is a particle or no particle in each quantum state.CHAPTER 3. dmi + dmi lngi .16 X i ln gi . mi lngi . Now we clump the energy levels to produce gi states and mi  gi. ln  where gi X i gi ln gi . INDISTINGUISHABLE PARTICLES 23 3. m ! i i i 3. mi! mi = N gi To maximise subject to the constraints : X i X i mi"i = E 3.. mi  + dmi i X gi .. gi .3 Fermi-Dirac Distribution for Fermions In this case.12 3. Therefore i Number of ways to ll mi boxes out of gi = m !g g. mi mi 1 d ln = . mi ln mi .

23 Applying time-reversal invariance. 24 gi . Riout = Ni Riin = X j. Tij!kl is the transition rate from i ! k accompanied by j ! l to maintain energy conservation. Note : 1 The positive sign in Fermi-Dirac distribution implies that nj  1 which is Pauli  Exclusion principle which states that no 2 fermions can share the same quantum state. Suppose the rate depends on the occupation at k and l.k.  has a value.l Nk Nl Tkl!ij 1  Nj  3.17 3. e mi "i 3. Therefore. Ni X j. 2 If nj = N is not a constraint.k.22 3.  is zero and this is true for open systems.l X = 1  Ni  Nk Nl Tkl!ij 1  Nj  j.18 3.k.4 Alternative Derivation for Bose-Einstein and FermiDirac Statistics For equilibrium consideration.21 At equilibrium.CHAPTER 3.l X 3. For isolated systems with xed N.l Nj Tij!kl1  Nk 1  Nl  3. mi = e.k.19 mi = e. P 3. e 1"j + 1 which is the Fermi-Dirac Distribution. INDISTINGUISHABLE PARTICLES Then.l Nj Tij!kl1  Nk 1  Nl  = 1  Ni  Tij!kl = Tkl!ij X j.20 Nk Nl Tkl!ij 1  Ni 1  Nj  j.k. eg"ii + 1 Averaging over the gi levels in each clump. nj = e. .

32 and with the negative - sign. the more likely the transition. k.33 For fermions. we obtain Ni = e. the more you receive. "i = e e. e 1"i  1 So starting with the positive + sign. 1 Bose-Einstein Distribution 3.30 3. .26 3.28 Ni Nj = Nk Nl 1  Ni 1  Nj 1  Nk 1  Nl N N N N ln 1  iN + ln 1  jN = ln 1  k + ln 1  lN Nk i j l with "i + "j = "k + "l which is the conservation of energy. the rate Riout 1 + Nk 1 . the more particles in k and l. "i 1  Ni 3. Ni = e e.25 3. 3.27 3.24 3. INDISTINGUISHABLE PARTICLES Then for every combination i. Nl .31 Ni = e. Nk 1 . Nl  which demostrates Pauli's Exclusion Principle. "i i then. we need 25 NiNj 1  Nk 1  Nl  = Nk Nl 1  Ni1  Nj  So 3. l. we obtain Ni = e.CHAPTER 3.29 "i Ni  Nie e. i. the rate Riout 1 . and for bosons. e 1"i . The more you have. j. Therefore N ln 1  iN = .e. e 1"i + 1 Fermi-Dirac Distribution 3.

This occurs when there is a change in the boundary conditions V. we have dW = .Chapter 4 Statistical Mechanics and Thermodynamic Laws 4.7 0H = .3 dU = X j "j dnj + X j nj d"j Now. the energy levels change. E = U = Therefore 4. 2 2  "j = ~2m 12 n2 + n2 + n2  V3 x y z 4.5 dW = X j nj d"j . For example : particle in a box Hence So.1 First Law The First law of Thermodynamics : dU = dQ + dW In Statistical Mechanics : Total Energy.g. etc.1 X j nj "j 4.P = or X @"j nj @V j X @"j nj @m j 26 4. e. When the boundary conditions change.2 4.4 4. M.PdV or 0Hdm etc.6 4.

V.Also dnj = 0 so is also constant.13 4. for example.9 Also gE.e. We start from the TdS equation : TdS = dU + PdV .14 S f  so S1 = f  1  and S2 = f  2. During irreversible changes. N  4.2 Second Law We now search for a microscopic explanation for S in terms of . V. the particles may collide with the chamber walls and acquire more energy where the walls are hotter with more heat added. 4. dS = 0 i. N  4. Now f  1 2 = f  1 + f  2 . The total of the Universe also increases because the equilibrium macrostate must be the maximum .S = S1 + S2 Total = Now 1 2 1 and 2 .11 4. volume increase with no heat added dQ = 0. STATISTICAL MECHANICS AND THERMODYNAMIC LAWS 27 dW Original dQ Then dQ= P "j dnj which takes place by rearrangement in the distribution. the total S in the Universe increases. dN 4. In kinetic theory. Now we consider 2 systems with entropy S1 and S2 and disorder number Total entropy. 4.8 so S f E. S is constant.12 4.CHAPTER 4.10 Note : In adiabatic changes.

The lesser the knowledge about a system.16 4.CHAPTER 4. . The value of k comes out from the ideal gas equation. we are unable to deduce the value of k. Therefore 1 +C +C +C 4.19  = @f   @ = @f  2  @ 2 Then 1 @f  1  = 1 @f  2  2 @ 1 1 @ 2 @f  1  = @f  2  = k 2 @ 1 @ 2 where k is a constant by separation of variables.18 4.23 Hence we have a microscopic meaning of entropy. The system is perfectly ordered when = 1 and S = k ln = 0. STATISTICAL MECHANICS AND THERMODYNAMIC LAWS Then 28 @f  @ So Also 1 1 2  = @f  @ 1 1  2 2 2 = @f  1  @ 1 4. 1 So S1 = f  1  = k ln S2 = f  2  = k ln S = k ln 1 + k ln = S1 + S2 which implies that C = 0. From the formula which we obtained for the relation of entropy with . To be proven later 4.17 4. When more microstates are availiable 1 so S 0.15 @f   @ @f  @ 1 2 2 2 = @f  1  @ 1 1 4.22 2 2 S = k ln R where k is the Boltzmann constant NA .21 4.20 4. the more disordered it gets.

CHAPTER 4. However. S = k log E0  = 0 Practically. T=0 is a theoretically correct statement.  that the disorder number tends to 1. The third law of thermodynamics can also be understood in a quantum mechanical . this is not achievable. experimentally. Hence we can only say S ! S0 as T ! 0 where S0 is the entropy associated with the remaining small interactions for which "int kT and independent of all the entropy for all the other interactions degrees of freedom. STATISTICAL MECHANICS AND THERMODYNAMIC LAWS 29 4. Now we come to the formulation of the third law : The Third Law of Thermodynamics states that as the temperature T ! 0.3 Third Law As the temperature T tends to zero. some small interactions e. Note that it is meaningless to talk about the exchange of particles in the same state because they do not exchange their positions. The internal energy UT increases when T increases. When all the particles take the ground state. it implies that there are no further permutations for the Maxwell-Boltzmann distribution. it is approaching ground state as UT also approaches 0.g. the magnetic moments are required to point in the same direction. If kT "int. In order to make T = 0. nuclear paramagnetism may lift the ground state degeneracy. the entropy S ! 0. E0  = 1. The atoms occupying in the ground state are placed in fixed position in a lattice like this. We know from the 2nd law of thermodynamics that heat passes from hot high temperature to cold low temperature if no work is done. quadrapole moments and octapole moments etc. Why is this so? The answer : There exists very small interactions in the nucleus. From the third law. When the macrostate has only one microstate. but it is an impossibility. then S 6= 0 at a given T. made up of magnetic moments. the contribution to the entropy S from the system components also tend to zero.

@Z j = X ." g e. 1 dT kT 2 . the Partition function Z is de ned as : Z= where j represents all levels.CHAPTER 4. d =.N j j j where which is proven in tutorial. By the Schrodinger equation : 30 .26 Since. The depends on the nature of the particles and volume V etc. N @Z jV. free exchange of particles will allow to reach the same value."j 4.25 We also de ne the internal energy. U or E U= nj "j X = N gj e. X j gj e kT . @ V. where will be the same.24 Now we try to nd the relation between P and T. when 2 systems are in thermal equilibrium. 2~m @ + V   = H @x2 2 2 The V  consists of many interaction terms. STATISTICAL MECHANICS AND THERMODYNAMIC LAWS perspective as the freezing out of interactions. = 1 kT "j 4. S as a combination of entropies of many sub-systems : S = Stotal 1 + Stotal 2 + Stotal 3 4.N Z@ j X 4. For the system of same identical particles. Hence we can approach the total entropy.4 Calculation of Other Thermodynamic Variables From chapter 2. "j "j Z j = . Now we proceed to calculate the other thermodynamic variables.

Z "j g Z  nj = N e "j j Then we work out ln .32 .28 Therefore The Helmholtz free energy. N ln N = U + N ln Z g nj ln nj  + nj  j 4.NkT ln Z 4. TdS .29 4. 1 @Z = @ ln Z  Z ln U = .30 4. SdT = TdS .SdT . STATISTICAL MECHANICS AND THERMODYNAMIC LAWS and therefore. SdT = . PdV . ln = N ln N ! + X j nj ln gj . N + j X Xj = N ln N +  nj "j  + N ln Z . ln nj ! = N ln N . F is de ned as : Therefore S = k ln = k U + Nk ln Z = U + Nk ln Z T F = U .N @ @ Z jV.CHAPTER 4.27 We know that nj = N gj e.31 Since dF = dU .N ln = NkT 2 @ @TZ jV. TdS . PdV 4.N 31 4. TS F = .

35 @T . STATISTICAL MECHANICS AND THERMODYNAMIC LAWS From the chain rule Therefore. @F V @T 4.33 Entropy : 4.34 = Nk ln Z + NkT @Z Z @T = Nk ln Z + U T Then we can use the results above to compute the heat capacity at constant volume : Cv =  @U V 4.CHAPTER 4. we can calculate : Pressure : 32 @F dF =  @V T dV +  @F V dT @T @F P = . @V T = NkT  @ ln Z T @V S = .

7.e. The atoms are pointlike .occupying no volume. 2. " kT which is the classical limit. The energy is quantized as in the case of particle in a box in QM. gravity.g.1 Classical Perfect Gas To apply the Maxwell-Boltzmann Statistics to the classical perfect gas.e. 6. but there is no interference on each other's "j . e. There is no interaction with external elds. From quantum mechanics. There is no internal structure for the particles. 4.only for exchanging energy between particles and establishing thermal equilibrium. i. 3. no excited electronic states and the paricles are treated as single particles. 33 . The discrete levels can be replaced by a continuous one if the spacing is small as compared to T. 5. 2 2 "j = 2~  2 n2 + n2 + n2  mL x y z for V = L3 and where nx. ny and nz are integers. magnetic elds.Chapter 5 Applications of Maxwell-Boltzmann Statistics 5. hence we consider only the translational kinetic energy KE. we require a few conditions : 1. i. There is a weak mutual interaction . The gas is monatomic.

because all the states are summed over. 2mkTL2 n2 +n2 +n2  nx ny nz X .~2 22 n2 = e 2mkTL x e 2mkTL y e 2mkTL z ny nz Znx1 Z Z .CHAPTER 5. X "j Z = e. APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS 34 Now we compute the partition function.1 Z 1 .1 2 Z 1 2 =  rdr d.~2 2 z 2mkTL2 x 2mkTL2 y = dnxe dny e dnz e 2mkTL2 n2 = XXX j ~2 2 5.~ 2  2 nx mL kT 5.x2 dx = L h2 2 ~2 2 n2 x 2 2 1 Let x =  2.x2 dx =  = Z 1 1 e.1 0 0 0 To evaluate the integral : Z 0 1 2mL2 kT  1 1 Z 1 e.~2 22 n2 X .2 where .x2 dx e 2mL2 kT dnx =  ~22 2 2 .1 .1 e. Z and in this case gj = 1. kT x y z e.~2 2 n2 1 .~2 22 n2 X .~2 2 n2 1 .x2 +y2 dxdy 2 Z.1 2mkT  1 Z 1 e.

ln Z = lnV + 3 ln T + 3 ln 2mk 2 2 h2 ln P = NkT  @@VZ T = NkT V 5.r2  1 0 0 1 =  1 2 2 2 p = Z.6 .4 5.1 Z 1 e.3 Therefore 2 Z = L3  2mkT  3 h2 3 = V  2mkT  2 2 h 5.y2 dy 21 5.x2 dx 1 Z 1 .5 So.1 e.e.

11 is the density of states. ny and nz making up the grid points with each grid point representing one state. i. Nh3 e kT" dn " n = 3 V 2mkT  2 d" where dn d" 5. The number of particles having energies between " and " + " : .e degeneracy dn ". APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS i.8 For each atom. ny and nz states. we consider the quantum numbers nx."j nj = N e kT Z Nh3 e . ."j kT = 3 V 2mkT  2 where j represent all nx.10 5. the distribution becomes : .9 5. d" To calculate the density of states.CHAPTER 5. PV = NkT since 35 PV = nRT Nk = nR n k = NR R k=N A because N n 5. Also the entropy is given by : S = U + Nk ln Z T 3 Nk + 3 Nk ln 2mk  + Nkln V + 3 ln T  = 2 2 h2 2 Therefore.7 = NA molecules per mole ln U = NkT 2  @ @TZ V 1 = NkT 2 3 T 2 3 NkT =2  U = 3 kT 2 5.e.

dn " = 1 4R2R d" 8 = 1 4R2 dR " 8 d" 2 2 " = 2~  2 R2 mL 5.CHAPTER 5. " N" 2 kT = p  e 3 " 2 4 kT  5.13 Therefore. The number of states lying between energy " and " + " will become : Using we obtain. kT 2v 2m 2 " 2 " 1 n = 3 3 h 2mkT  2 1 . APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS By writing 36 R2 = n2 + n2 + n2 x y z 4 The number of points states with energy up to "R lie within a sphere of volume 3 R3.12 d" = ~22 R dR mL2 dn " =  R mL2 " d" 2 ~232 2 3 1 =  2m3L " 2 " 4 ~3 5.14 For the momentum distribution. Therefore. 2m 1 = 2V h3  " 2 " 3 2 " 3 V Nh3 e. Only positive n's are allowed so we take the positive octant. we use : and 2 " = 2pm d" = p dp m .

15 ZZZ all p space dpxdpy dpz = = Z 2 Z0 1 0 d Z p 2  0 sin . 2mkT p2p = 2 2mkT  3 Note that 37 5. the number of particles have momentum between p and p + p is : p2 d" n = p  N 3 pp e. 2mkT dp p 2 2m 4 kT  p2 N 4e. APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS So.CHAPTER 5.

pd.

p

Z

1

0

p2dp

4p dp

5.16

Therefore we can also express in px, py and pz as p2 N e, 2mkT p p p n = x y z 3 2mkT  2 x p2 + z y N e, p2 2+mkT p2 p p p = x y z 3 2mkT  2 p2 p2 p2 y N e, 2mkT p e, 2mkT p e, 2mkT p z x = x y z 3 2mkT  2 For velocity distribution, we use p = mv, n = f vx; vy ; vz dvxdvy dvz 2 2 2 3 mvx +vy +vz  = Nm 3 e, 2kT vxvy vz 2mkT  2 = gvdv 3 mv2  = Nm 4 3 e, 2kT v2dv 2mkT  2 which are the Maxwell distribution laws for velocities and speeds. We invoke a few integral identities :

5.17

5.18

In =

Z

1

I0 = 2 I1 = 1 2 I2 = 1 I0 2 I3 = I1 I4 = 3 I2 2

0 p

e,x2 xndx

CHAPTER 5. APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS
So,

38

Z
0

1

Nm3 4  2kT  3 Z 1 e,x2 x2dx 2 gvdv = 2 2mkT  3 m 0 =N
1 Z 1 vgvdv v=N  0 m34  2kT 2 Z 1 e,x2 x3 dx = 3 2mkT  2 m 0 r = 8kT m 1 v2 = N

5.19

The mean speed, v 

5.20

and mean square speed is

v2gvdv 0 m34 Z 1 e,x2 x4 dx = 5 2mkT  2 0 = 3kT m

Z

1

5.21

Average energy, 1 mv2 = 3 kT 2 2 Root mean square RMS speed, v2 = The most probable speed :
3 dgv = 0  2v e ,2mv2 , mvp e ,2mv2 = 0 kT kT p dv r kT kT  vp = 2m

5.22

p

r

3kT m

v 

5.23

5.24

CHAPTER 5. APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS
Maxwell Distribution of Molecular Speeds V

39

Vp

V

V2

T

vp v 

p

v2

5.2 Harmonic Oscillator
We now consider the simple 1-D harmonic oscillator in equilibrium at T, with the 2 Hamiltonian, H = 2pm + 1 k0x2 2 where j = 0,1,2,.....

"j = j + 1 ~! 2

! = k0 m Assume the oscillators are independent and distinguishable, ,"j nj = N e kT Z Z=
1 X
j =0

r

e,

j + 1 ~! 2

kT

! = e, 2~kT ! = e, 2~kT

1 X
j =0

e, kT

j ~!

1 ~! 1 , e, kT ~! 1 = e, 2 1 , e, ~!

5.25

1 1 U = N ~! 2 + ~! 1 + kT + ::: ..27 5. ln1 . 1 At low T... Quantization is not important for kT ~!  "= The Equipartition theorem states that all terms in the Hamiltonian with squared coordinates x or px etc will contribute  1 kT to the average energy..28 which is the classical result. ~! ~!e.. ar2.26 ln Z = . ~!  2 U = ...N @lnZ @ 1 = 2 N ~! + 1 . 2 At high T. This is re ected in the classical regime at high T where kT "int.CHAPTER 5. ~! e 1+ 1 = N ~! 2 e ~! . N. 1 ~! . provided quantization 2 is not important. e. APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS Note : for a series of a. it means that the oscillators are in ground state. 1 = N ~! 1 + kT 2 ~! = NkT 5. U  1 N~!. Example 1 : Ideal Gas. The high energy states are occupied.3 Equipartition Theorem . r 5. 1 X n=0 40 a arn = 1 . 1 = N ~! 1 + ~! 1 2 e kT . ar. H = kinetic energy only p2 + p2 + p2 y z x = 2m 2m 2m Therefore U = 3 kT  N atoms 2 = 3 NkT 2 5.

For vibrational motion. p2 + p2 + p2 y z x Htrans = 2m 2m 2m 2 1 L2 Hrot = 1 Lx + 2 I y 2I x y neglecting rotation about z-axis. atoms don't appear to vibrate much at normal temperature.e.CHAPTER 5.29 . Therefore. 2 Example 3 : Diatomic gas. the total internal energy. i. the Hamiltonian is frozen out and the contribution is not 1 kT. 2 The reason is because the vibrational motion is frozen out. So in terms of Hvib .e. the experimental 2 results show that U = 5 NkT. For rotational motion. 41 Therefore H = Kinetic energy + Potential energy p2 x = 2m + 1 k0x2 2 U = kT  N atoms = NkT With quantization setting in at low T. APPLICATIONS OF MAXWELL-BOLTZMANN STATISTICS Example 2 : 1-D Harmonic Oscillator. However. Total H = Hvib + Htrans + Hrot 5. p0z2 + 1 k z0 2 Hvib = 2m 2 0 where z' is the relative coordinate between the 2 atoms. U should be 7 NkT. For translational motion. i. the atoms are in ground state. ~!vib kT.

which are ultimately paired up so that the net J of outer electrons is zero. J = .~  B  ~ ~ ~ = gJ B B  J = gJ B Bz Jz Quantization 6.Chapter 6 Paramagnetic Systems 6.2 The Jz and J are observed from atomic physics. In many solids of transition elements 3d. Note that J = S.gJ B J When the system interacts with an external magnetic eld. Note that the excited electronic states are unoccupied at room temperature because " kT . Hamiltonian. It also acts like the magnetic moment.1 6. so that the net L = 0 for atoms in the ground electronic state.1 Introduction For a paramagnetic spin system. the atoms have unpaired electrons or partially lled ~ orbitals. 4d and 4f orbitals of many elements are not involved in valency. S . the orbital quenching occurs. there is a degeneracy of 2S + 1 states. B = 2e~ Bohr Magneton mc In a solid. Hence in the ground electronic state of 3d elements. H = . particularly 3d. L and spin. the outer electrons are involved in valency for chemical bonds. so magnetism arises from their S. The inner core. 42 . L and J. The net angular momentum J for each atom is equal to both the linear orbital ~ ~ angular momentum.

gsB Sz =  1 gsB 2 The Zeeman interaction with external eld gives energy levels gsB Bz Sz =  1 gsB B 2 where gs  2 and B is the Bohr magneton.4 6.3 The distribution in the spin states is associated with the spin temperature TS which may not be the same as the lattice temperature TL . . and determines the relaxation time .eld to a spin 3 system : 2 For example. 2 The magnetic moment.CHAPTER 6. 6. S = 3/2 system 43 3 2 1 2 -1 2 -3 2 B 0 Hamiltonian.spin-phonon interaction. This is important in spin resonance experiments such as electron spin resonance and nuclear magnetic resonance. A mechanism must exist for TS and TL to establish thermal equilibrium .  = . 6.2 Spin 1 Quantum System 2 We consider the system of N magnetic atoms of spin 1 due to 1 unpaired electron. H = gsB BSz and gs  2 for electron spin.5 The  energies correspond to spin   antiparallel to B or spin "  parallel to B. PARAMAGNETIC SYSTEMS Applying B. 6.

PARAMAGNETIC SYSTEMS "" = . the total energy.12 6.13 where " = B . Z 2kT 6. 2 gsB e.10 where Therefore. 1 gsB B 2 1g  B " = + 2 s B Using the Maxwell-Boltzmann distribution.11 S = U + Nk ln Z T = . 2kT Z B B Z = e+ 2kT + e. BV 6.CHAPTER 6. Z 2kT Ngs B B tanh gsB B  =. NgsB B tanh gsB B  + Nk ln2 cosh gsB B  T 2kT 2kT The magnetization M : 1X M = V nj j j X + j B kT = VN Z j e gs B B gs B B 1 = VN 1 gsB e 2kT . 2kT  Z 2 2 Ngs B B sinh gsB B  =.9 6. 1 gsB Be+ 2kT + 1 gsB Be.6 6. 2kT = 2 cosh gs2B B  kT gs  B gs  B 44 So 6. E : The entropy S : gs B B gs B B E = N . 2kT Z2 N g  sinh gsB B  = VZ s B 2kT 1 N g  tanh gsB B  = 2V s B 2kT E = . gs B B n" = N e+ 2kT Z gs B B n = N e.8 6.7 6.

1. the partition function. J Therefore there are 2J + 1 states. ::::.J + 1 y = 1y 1y e 2 .3 Spin J Quantum System Now we consider a system with total angular momentum J in an external B. +J +(J-1) 0 -(J-1) -J B=0 B mJ = . e.J J X mJ =.Jy = ey . 1. 2 = e. e. "J = gJ B BmJ where gJ is the Lande g-factor an experimental quantity. 0.J.eld. 11  6. PARAMAGNETIC SYSTEMS 45 6.J . 1 2 2 eJ + 1 y .CHAPTER 6. .P sum = ar .yJ 2J X rn .15 = sinhJ + 1 y 2 sinh y 2 . e. ::::::. J . = e. Let y = gJ B B Therefore.J 6. G. 1.14 eymJ eymJ mJ =0 2J +1y . 1 1 eJ +1y .yJ  e ey . J X gJ B B Z= e = mJ =.

lnsinh y  2 2 1 J + 2  coshJ + 1 y 1 cosh y 2 . PARAMAGNETIC SYSTEMS Therefore.17 ln Z = lnsinhJ + 1 y . =Z "J gJ B BmJ 6. U = .NgJ B BJBJ y 6. 2 coth y 2 2 2 Brillouin function 6.20 6. lnsinh y  2 2 Internal energy.CHAPTER 6. 2sinh y 2  M = N gJ B  1 V sinhJ + 2 y 2 1 1 = N gJ B J + 2  cothJ + 1 y . Z N e. M = VN Z gJ B mJ e @Z = VN 1 @B Z N 1 @Z @y = V Z @y @B = VN gJ B @Z Z @y ln = N gJ B @ @yZ V gJ B BmJ 6.18 6.19 where We proceed to evaluate other thermodynamic quantities : Helmholtz free energy.NkT ln Z = .N @ @yZ @ = .21 1 1 BJ y = J J + 1  cothJ + 1 y .NkT lnsinhJ + 1 y . 46 Then nJ = N e. 2 coth y  V 2 2 N g  JB y =V J B J 6.N @ ln Z @ ln @y = . F = .22 .16 X Magnetization.

27 . y =J y+ 3 y 12 = 31 J 2 + J y J y J + 1 = 3 6.1 x2 x3 x2 = 1 + 2x + 42 + :::: + 11 + 2x + 42 + 83! + ::::: .26 6.x 6. PARAMAGNETIC SYSTEMS Entropy.CHAPTER 6. x + x2 . NgJ B BJBJ y + Nk lnsinhJ + 1 y . 1 .1 x2 = 21 2 + 2x + 2x2 + ::::1 + x + 23 + ::: x 2 2 = 1 1 + x + x2 + :::1 . x + 2x + ::: + x + 2x :::::2 :::: x 3 3 1 1 + x . coth x = ex + e. 1  1 + y  2 3 2 y 3 J + 1 y 2 2 1 2 1 1 J + 2  y .x e . x2 .25 so. 1 BJ y = J J + 1  2 Therefore. lnsinh y  T 2 2 At high T. y 1.24 So. 1. 1 1 + J + 2 y  .23 . and for small x. S = U + NK ln Z T = . M = N gJ B J J + 1 y V 3 2 N gJ 2 J J + 1B = V B 3kT @ = 0 @B T 2 2 0J  = N gJ B 3kTJ + 1 V 1 Curie's law T 6.e 2x = e2x + 1 e . x2 + 2:::: = x 3 2 = 1 1 + x + ::::: x 3 1+x = x 3 x 47 6.

e x 6. At low T. 1 .Jy + ln2J + 1  ln2J + 1 48 U = . 1  + y .J + 2 2 2 2 1 y + :::: . y + ::: 1 + J + 2 y + :::: . ln y + :: = . e.33 where all spins are align in the same direction parallel to B.J + 2 2y .J + 2 y e . hence ! 0.29 or 1 1 + y + ::: . M = N gJ B J V . Since.31 .CHAPTER 6. J + 1 + ::: 2 2 2 .x = 1 e . 1 = ln e. y So. ln y 2 2 = ln2J + 1 6. Proof : 2 2 eJ + 1 y . Therefore 1. and for large x. 1 .J + 1 y ln Z = ln 2 1 .30 Then S = Nk ln2J + 1 for 2J + 1 levels. ln ln Z = ln 2 2 = lnJ + 1 y .Jy + ln J + 2 2 = . 1 y. NgJ B BJ J + 1y ! 0 3 ln Z  ln2J + 1 6.28 6.2y e 2 1 2 1 y + ln 1 + J + 1 2y + ::: . PARAMAGNETIC SYSTEMS Since. 1 2 = .32 6. ln e . coth x = ex + e. 1 = 1 2 2 6. ln 1 + y + :::: .x 1 BJ y = J J + 1 .

eld increases. Reduce the entropy by turning on the B.J + 1 y  . e.eld to zero. y . But alignment can also be produced with a B. 1.e. So adiabatic cooling is two-staged.35 6. so cooling reduces the entropy of system.4 Adiabatic Cooling From the third law. U = .NgJTB BJ + NkJy = 0 6. ln 1 = J + 2 2 2 2 = Jy 49 6. ln e y .eld isothermally.eld. Reduce the temperature by turning o the B.34 Internal energy. dipoles align more and energy and entropy is taken out from the system. so we need magnetic dipoles to align in one direction. Ts Spin System TL Lattice Heat Bath at constant T Heat flow Entropy passed to lattice and heat bath = dQ Ts 2. As B. isolated from heat bath. . so we need the alignment to stay as we reduce the B. S ! 0 as T ! 0. i. S = .2y  ln Z = ln 2 2 1 y + ln 1 .eld adiabatically. e . PARAMAGNETIC SYSTEMS 2 2 2 eJ + 1 y .36 6. The paramagnetic system is in contact with the lattice and the heat bath.CHAPTER 6.NgJ B BJ Entropy.

dQ = TdS. The lattice passes heat and entropy back to spin. Isothemal Magnetization Stage (2) : From Q to R. But this entropy is so small because of low TL . As B decreases. Hence both TS and TL will drop drastically. SLat is small. PARAMAGNETIC SYSTEMS Ts Spin System TL Lattice 50 Remove the B-field adiabatically to keep spins align. so the heat passing is very small.CHAPTER 6. S Entropy Curves obeying the third law P B=0 (1) (2) R Q B=B T Note : They obey the 3rd law Stage (1) : From P to Q. Sspin stays constant. so TS decreases. Adiabatic Demagnetization .

n ! n !g g. The two s will reach equilibrium at same temperature. n ! i i i i j j j j 7.1 Fermi-Dirac Distribution From chapter 3. = kT .2 nj = e. Now to identify for Fermi-Dirac distribution. we consider 2 fermion systems in thermal and di usive contact so to allow particle exchange. egj" + 1 2. We also assume that the same type of fermions for both systems. Fermi-Dirac Distribution nj = e.Chapter 7 Applications of Fermi-Dirac Statistics 7. 1 1 For either systems. With the same constraints : X X ni + nj = N X ni "i + nj "j = E = i j Y Y i j = n !g g. Bose-Einstein Distribution 7.3 51 . This can be proven by having the quantum system in 1 contact with a classical perfect gas where = kT . egj" . we obtain two quantum distributions : 1.1 7.

dni + X j 7. gi . nj  . so i = j =  chemical potential Suppose we take one of the systems : Sj = k ln j X = gj ln gj . gj . ln nj dnj . X i Xj "idni + "j dnj  = 0 j i X i ln gi .4 lngj .5 Using the method of Lagrange multipliers. + "idni + n i X j . ni lngi . nj . ni dni + dni . nj  . ln gj n nj . nj ln nj j ni = e. nj  . eg"jj + 1 7. ni . gi . APPLICATIONS OF FERMI-DIRAC STATISTICS ln = + 52 X X j i i gi ln gi . + " = 0 ln n j j So the two systems in thermal contact have the same and the two systems in di usive contact will have the same . ni dni + ln gj n nj dnj ni j j i 7. nj ln nj + nj X X = gi ln gi . ln nidni . nj dnj + dnj .6 Then Choose and . X X  dni + dnj  = 0 .CHAPTER 7. ni . gj . gj . nj  + gj . nj  lngj . nj  lngj . so that Then ln gi . So. gi . dnj =0 X X . = ln gi . eg"ii + 1 nj = e. ni  lngi . + "i = 0 ni gj . nj  lngj . which is related to . + "j dnj = 0 j 7. The di usive contact gives rise to chemical equilibrium. ni + gi . gj . nj ln nj j dln = X i lngi . ni .7 . ni ln ni + ni gj ln gj . ni ln ni + gj ln gj . ni  .

there is free propagation and in nite conductivity of electrons. dW = .e. nj ln nj dnj j X = k .8 TdS = . takes little energy to detach them. The electrons interact with the electrostatic potential of the positive ions in the lattice.2 Free Electron Model of Metals The valence electrons in the outermost shell are loosely bound.k dnj + k dnj "j  . i.g electrical work.CHAPTER 7. k dW j j 7. In metal. e. we apply Fermi-Dirac Statistics . dW because 7.electrons of spin  1 and 2 =  1 . the electrons are con ned in a box body of 2 .11 +1 where  dN is the work done in moving the fermions from one place to another. For perfect stationary lattice. " .k dN + k dU .10 7.dN + dU . However lattice imperfections. APPLICATIONS OF FERMI-DIRAC STATISTICS Then 53 dS = k =k X j lngj . we will obtain the solution to the Schrodinger equation in the form of Bloch waves. nj dnj + dnj . hence they have low ionisation potential. e jkT nj = gj 7. In the free electron model. + "j dnj j X X = . We neglect the lattice potential. nj ln nj .9 dU = TdS + dW + dN Therefore So  = kT same derivation for bosons 7.kT dN + dU . dnj X gj . If we consider the periodicity of the potential. lattice vibrations phonons and the scattering by the other electrons leads to resistance in the lattice. nj d"j = . The energy will form a band. the weakly attached conducting electrons are free" to move in the lattice.

Number of states between k and k+dk = 1  4k2dk  2   L 3 8  V k2dk = 2  7. i.ny . ny .13 nR ny nx 1 4 Number of states from k=0 to k=k = 8  3 n3  2 R 7.16 .nz . APPLICATIONS OF FERMI-DIRAC STATISTICS metal and we neglect the mutual interactionselectron-electron scattering. nz .CHAPTER 7.12 where k is the wavevector 2  and is de ned by the following relation : 2 k2 = L2 n2 + n2 + n2  x y z nz 7. ny and nz are all positive integers.15 V m" Number of states between " and "+d" is g"d" = 2 2~2 dk d" d" V 2m" m d" = 2 ~2 ~2k r V 2m2 " ~2 d" = 2 4  ~ 2m" 3 m2 1 = 2V 2 2~3 " 2 d"  1 = V C" 2 d" 7. . the energy is given by : 54 "nx.14 Note : we multiply the relation by 2 because of the 2 spin states and 1 because we 8 only take the positive octant.e. For single particle state nx. nx. = 2~  2 n2 + n2 + n2  mL x y z 2 = ~ k2 2m 2 2 7.

 7.17 E= = X X j j nj "j e "j . 2 1 2m 3 C = 22 ~3 55 N= = X X j j nj e "j .17 for N. + 1 0 We solve equation 7.19 1 0 EF E For T = 0 = "F and Fermi energy 7. +1 Z 1 g"d" = e "j .18 to obtain ". 1 +1 7. This is di cult to evaluate except at low T.D " = 0 fF .D " = f F-D (E) "j gj e "j .D " = 1 " "F " "F .20 fF .CHAPTER 7. T. We shall try to evaluate this in 2 cases 1. + 1 0 gj 7. V  and use  in 7.18 +1 Z 1 "g"d" = e "j . At T = 0 Probability of occupation of state of energy is fF . APPLICATIONS OF FERMI-DIRAC STATISTICS where Therefore.

23 The meaning of Fermi energy is that all states below "F are occupied and all states above "F are unoccupied at T = 0. + 1 " +1 Z0 1 "e. Z0 1 e" . Total Energy. all electrons occupy the lowest possible state subject to the Exclusion Principle.D"d" 7.21 and g"d" Z "F 1 = VC " 2 d" 0 3 2 = V C 2 "F 3 2 3 2 "F = 3N 2 ~ 3 3 2V 2m 2 2 ~2 3 =  3N  2 2m V 2 ~ 2 = 2m kF 1 kF =  3N  3 V 7.24 N= Therefore.D"d" Z "F 3 =VC " 2 d" 0 2"5 =VC 2 5 3 N" =5 F 0 Z 1 7. g"fF . At T=0. we want to nd Cv and we know that the experimental value "F is proportional to T. Z 1 0 g"fF . kT 1.CHAPTER 7. " g"d" e F = + N"F ". + 1 e 0 Z 1 7. At low T. so the states are lled until "F . APPLICATIONS OF FERMI-DIRAC STATISTICS Hence 56 N= = Z Z 0 0 1 "F g"fF . E = 2.D"d" U= "g"d" ". " +g1"d" " g"d"  F  = + Fj . ".22 2 7.25 .

"F V C" 2 d" Cv = kT 2 7.27 e "."F  + 12 0 Now. " g"e "."F  = . dfF-D (E) dE kT EF E The contribution to the Cv integral comes mainly from values of " near "F ."F  2 Z 01 x2e dx + 1 1 ex 2 where x = " . Z 1 1 1 " . @fF . Z 1 x2 exdx 2 = 3 . . "F  = V C"F k2 T "F ex + 12 The lower limit can be extended to . 7.1 because kT "F . i. "F 2e ".26 1 1 @ F  " .e.CHAPTER 7."F d"d" 2 Cv = kT 2 F ".31 . The value drops to 1 of the maximum value when e " .28 e ". + 12 0 @ Since  does not change much. so we assume @T = 0 and  = "F . kT 2 . "F  kT .1 ex + 12 Therefore 2 Cv =  k2 Tg"F  3 where g"F  is the density of states of Fermi level. e "j . So the function is sharp for small kT  "F . The physical picture : 7. kT @T  d" =."F  + 12 @" The maximum occurs at " = "F . "  "F 1 V C" 1 Z 1 " . "F e ". e ". APPLICATIONS OF FERMI-DIRAC STATISTICS Then 57 Cv =  @U V @T Z 1 " .30 7.D " 7.29 7.

Only those electrons near "F can gain kT to move into unoccupied states. and need to include e ective mass of electron m me because electrons are not completely free but the lattice potential e constrains them. heat capacity = The number of electrons participating  energy T = g"F kT  kT T = f "F k2 T 7.33 Experimental result : Cv T for a fermi gas. most electrons cannot gain kT because the states above them are already occupied. Therefore Energy of excitation = g"F k2T 2 7. For T 0 : We will expect electrons to gain energy kT.CHAPTER 7. The theory is in agreement but does not give the right constant of proportionality. . However. APPLICATIONS OF FERMI-DIRAC STATISTICS g(E)f F-D (E) 58 T=0 T>0 EF E For T = 0 : All states are occupied within " = "F is a truncated parabola. An alternative estimate : Number of electrons involved is g"F kT and they gain kT.32 Contribution to Cv . This results in a rounded-o cli at "F .

In this chapter. The photons do not interact with one another. ! = ck.Chapter 8 Applications of Bose-Einstein Statistics 8. At equilibrium. The hole is very small so that the equilibrium is not disturbed. We note that the photons have the following characteristics : 1. " = ~! = ~ck 59 . we discuss radiation in thermal equilibrium. 3. 2. The photons have zero rest mass and travel at c. The radiation bounces inside the chamber until nally being absorbed. the distribution of photons will follow the temperature T.1 Introduction to Black Body Radiation First. the speed of light. The photons satisfy the dispersion relation. The photons have spin 1 bosons and they obey Bose-Einstein statistics. The radiation is enclosed in vessel. so. 4. in order to study the radiation. we assume a hole is cut in the chamber wall to emit some radiation. The atoms in the wall will absorb and emit electromagnetic radiation. The black body will also absorb all radiation falling on it. This will approximate to a blackbody. we will study black body radiation using statistical mechanics. and all the walls are at temperature T. Second.

1 ln n dnj = 0 j . APPLICATIONS OF BOSE-EINSTEIN STATISTICS 60 8. nj ln nj 8. The energy content of photons are not changed. E= For bosons. 1! nj + gj . 1dnj + dnj . 1 . This implies that and  = 0. 1 lngj . 1 lngj . The blackbody absorbs and emits radiation until thermal equilibrium is established. X j nj "j = constant for the photons = Y nj + gj . 1! j nj !gj . 1 lnnj + gj . The number of photons are not conserved because of absorption and re-emission.CHAPTER 8. nj ln nj . 1! . gj . 1! 8. The features of the system are as follows : 1. 1 . 3. dnj = 0 8. 1 lnnj + gj . ln = = = X X j X j j lnnj + gj . nj ln nj + nj . lngj . The blackbody is so small that heat capacity is negligible. 1 .2 dln  = X j lnnj + gj .2 Bose-Einstein Distribution Perfect Insulating Walls Small Blackbody We consider a system of a small blackbody placed inside a chamber with perfect insulating walls.1 Maximise subject to E = constant only. gj . ln nj ! . nj + gj . 1 + gj . 1 nj + gj .3 X nj + gj . 1 . 2.

2 linear or 2 circular. standing electromagnetic waves result such that n = L 2 Therefore x kx = 2 = nL x 2 = ny  ky =  L y z kz = 2 = nL  z For any photon with ~ = kx. kz .6 1 Prove = kT by having the photon system in thermal contact with another system at temperature T and use walls that can exchange energy. k ! 2 = k2 = k2 + k2 + k2 x y z c2 For any ~ . 1 8.4 8. ky . for large gj and nj .3 Planck's Radiation Formula Consider photons enclosed in a box L3 . 1  . 8.5 X g ln nj +n j . j 61 dnj "j = 0 8. k nz 8. 1 Bose-Einstein function = e "j1.7 nR ny nx . APPLICATIONS OF BOSE-EINSTEIN STATISTICS X Therefore So.CHAPTER 8. there are 2 polarizations . "j dnj = 0 j nj = e "jgj.

8 since V Number of photon states between k and k+dk = 2 k2dk 2 Number of photon states between frequency range ! and !+d! = V2 !2 d!  c c V !2d! = 23 c The number of photons in frequency range ! and !+d! is 8. the maximum of u!. T is given by : e ~! 3!2 .11 8. T) 8. Thus the energy of the photons in the frequency range ! and !+d! is V ~ !3d! ~!N ! d! = 2 3 ~!  c e .1 The energy density per unit volume in the frequency range is 3 ~ Planck's formula u!. Number of photon states for k=0 to k=k = 1  4 n3  2 8 3 R 3 =  L3 k3 3 2 n2 = L2 k2 R  62 8. T d! = 2c3 e !!d! 1 ~ .CHAPTER 8. 12 ~! =0 . u(w.13 T2 T1 T2>T1 w For T2 T1. APPLICATIONS OF BOSE-EINSTEIN STATISTICS Therefore. !2 ~e e ~! .12 8.10 N !d! = Density of states  Bose-Einstein function !2 = 2c3V e ~d! 1  !.9 8.

the colour of the blackbody changes from red to blue : Total Energy Intensity.19 8.18 uT  = aT 4 where Stefan-Boltzmann law 8.16 8. Therefore. or more accurately.CHAPTER 8. ~!max kT = 2:822 Wien's displacement law 8. 1 0 4 4 4 = k2 T 3   c3~ 15 0 Z 1 8.15 8.17 or !max = 2:822k T ~ So. 1 1 x3 dx ~ = c3 41~4 ex .4 Radiation emitted by blackbody d = r sin . APPLICATIONS OF BOSE-EINSTEIN STATISTICS 63  3e ~! . by neglecting -1. uT = 2c = 2:822k T max ~ ~c 1 max = 2:822k T Therefore u!. 1 = xex where x = ~! Solving numerically. T d! Z 1 !3d! ~ = 23  c 0 Ze ~! . 1 = ~e ~!  3ex .822.14 8. x  3. x = 2. as T rises.

d rd.

r2 = sin .

d.

A.d 2 4 a = 15ck~3 3 Consider a tiny hole. with photons coming from direction .

! .

+ d.

. ! + d . where . with a subtended solid angle d .

64 The photons that will pass through in time t are located in a parallelpiped of length ct. APPLICATIONS OF BOSE-EINSTEIN STATISTICS The total solid angle = 4. The volume of the parallelpied is dA  ct cos .CHAPTER 8.

No. θ θ c θ t dA Therefore.. of photons passing per unit time from direction .

! .

+ d.

. ! + d  with a frequency range ! to ! + d! = Volume of Parallelpiped  Fractional solid angle  Number density in frequency interval t = dAc cos .

 d   N !d! 4 V Therefore. the total energy radiated per unit time in frequency interval ! to ! + d! Z 2 Z  2 = d sin .

d.

dAc cos .

 ~! N !d! 4 V 0 0 = 1 c~!dA N !d! 4 V 1 cu!. T d! 4 . T dAd! =4 Therefore Power radiated per unit area in frequency interval from ! to ! + d! = 1 cu!.

8 W m.CHAPTER 8. T d! Total power radiated = 4 0 1 caT 4 =4 = T4 where a = 15c3 ~3 65 2 k4 . Therefore.2 K.4 . 2 4 = 60ck~3 2 = 5:67  10. APPLICATIONS OF BOSE-EINSTEIN STATISTICS 1 c Z 1 u!.

 1 nr F-D. there is a lack of discreteness. So we require N Z 1 9. Fermi-Dirac and Bose-Einstein distributions approximate to Maxwell-Boltzmann distributions in the classical regime. since "0 ! 0 for ground state.Chapter 9 The Classical Limit 9. including r = 0. which are averaged over time. The energy spacing is " kT. and nr unoccupied states. According to the Maxwell-Boltzmann statistics.B-E  e. we consider the continuous variable " rather than the discrete energy "r . nr M-B= N e. 1 which means that there are many In the Fermi-Dirac or Bose Einstein statistics. "r M-B distribution So all the 3 distributionsMaxwell-Boltzmann. = e  e.1 1. 66 . the classical regime is only satis ed if 1 1 e "r . e so that "r .1 The Classical Regime In the classical regime. "r .  1 i.e. "r Z for all r.

2 Monatomic Gas For monatomic gas. . THE CLASSICAL LIMIT 67 9.CHAPTER 9. so.

3 V 2mkT 2 Z = h3 .

1 3 2 Nh3 V 2mkT T 2 .

and high mass. T and low density N . V Assume the energy is equipartitioned. p p = 2m" = h in QM terms dB Then . N 3 h2 V 2mk 1 The conditions are high temperature. m. " = 3 kT for each atom 2 p2 = 2m So.

3 .

1 3 2 2 1 = 4 2mkT 3 m" .

3 3 1 2 = 2 p3 .

3 3 .

 3 2 dB = 2 h and .

3 N 3 2 3 1 V 2 dB i. 1 1 .e.

2 1 .

V 3 .

2 2 2 dB 3 N = 3 l0 where l0 is the average seperation between atoms. Examples : 9.2 9.3 .

7cm At T = 273 K. dB l0.CHAPTER 9. and so dB = 8 10.31 kg and T = 273 K. He gas : so l0 = 2  10.8cm For m = 9:1  10. 2. Hence Maxwell-Boltzmann distribution is applicable. 68 N = 1020 per cm3 V m = 6 41023 g l0 . hence N  1023 per cm3 V l0  2  10.9 cm For liquid He at T = 4 K. g(E)f F-D (E) T=0 Approximately M-B T>0 EF f F-D (E) E 1 T=0 T>0 0 EF E . so Bose Einstein statistics apply. N = 2  1022 per cm3 V so l0 = 4  10. For conduction electrons : Assume 1 electron per atom. THE CLASSICAL LIMIT 1. Since dB l0 . so Fermi-Dirac statistics apply.8cm.

Now we mix the volumes of identical atoms.3 Gibbs Paradox For classical ideal gas. 9. V N T S V N T S S = U + Nk ln Z T  . THE CLASSICAL LIMIT Note that 2 n" " 1 e. 69 " for M-B distributions. we assume the atoms to be distinguishable.CHAPTER 9.

However.5 = 2Sold + 2Nk ln 2 The problem lies with distinguishability and overcounting. Hence we use a corrected . 9.4 3 S = 3 2N k + 2N k ln 2V + 2 ln T + 3 ln 2mk 2 2 h2 9. old Corrected = N ! Y gjnj = n!  . V ! 2V.  3 Nk + Nk ln V + 3 ln T + 3 ln 2mk = 2 2 2 h2 Suppose now N ! 2N. we expect S!2S.

Corrected S = k ln Corrected .  9.6 j j Therefore. kN ln N . N   .

.

7 .  5 Nk + Nk ln V + 3 ln T + 3 ln 2mk = 2 N 2 2 h2 9.

V ! 2V.CHAPTER 9. 70 2V S = 5 2N k + 2Nk ln 2N + 3 ln T + 3 ln 2mk 2 2 2 h2 = 2Sold  . THE CLASSICAL LIMIT Then for N ! 2N.

.

 9.8 .

pz .mvx 2 e 2kT dvx = 1 .z can have any momentum px. The probability that x-velocity lies between vx and vx+dvx is : P vx dvx = A0e 2kTx dvx where . The kinetic energy KE terms are determined by px. The probability that x-momentum which lies between px and px+dpx is : x P px dpx = Ae 2mkT dpx . For the Hamiltonian : 1. velocity. The potential energy PE terms are determined by x. py and pz . i. 1. At thermal equilibrium.1 Introduction Consider the distribution for e.e.Chapter 10 Kinetic Theory of Gases 10. The Boltzmann law applies to every coordinate in the Hamiltonian.1 71 .1 10.2 where A Z 1 . etc of system of gas molecules in thermal equilibrium. py . For example.g.p2 x e 2mkT dpx = 1 . a molecule in any position x. the momentum and spatial coordinates are independent of each other.mv2 10.1 2.y.4 A0 Z 1 . energy. position. momentum. 2. y and z.3 10.p2 10.

CHAPTER 10.5 10. v x 2 2 2 For speed distribution : P c dc where c2 = vx + vy + vz . P[px]. vy . vz dvxdvy dvz = P vx dvxP vy dvy P vz dvz . KINETIC THEORY OF GASES 3. The probability that the molecule is found between and x and x+dx is : 72 P x dx = Be where . For graphs of P px or P vx : the gaussian curves centred at zero. hence we do not require the interaction potential except for establishing thermal equilibrium. dφ dc dθ P vx.1 We concentrate on the momentum and velocity distributions.V x e kT dx = 1 .V x kT dx 10.6 B Z 1 . P[v x ] px.

2mvy e .2mvz2 dv dv dv kT kT kT = 2kT x y z . 3 2 2 m 2 e .2mvx e .

m 3 .mc2 2 e 2kT dccd.

csin.

d  = 2kT Integrating over all angles : 3 .

m 2 . we can calculate : 10.8 .mc2 P c dc = 4 2kT e 2kT c2dc Hence.7 10.

put .CHAPTER 10. KINETIC THEORY OF GASES 1. cmax : For cmax. Most probable speed.

9 c=  3. 1 dP c = 0  c = 2kT 2 max dc m 2. Mean speed c :  73 10. Mean square speed. c2 : Z 0 1 1 .

10 p 4. Mean kinetic energy KE : Note : so kT c2 = 3m . Root mean square speed. 8kT 2 cP c dc = m 10.11 p 5. c2 : c2 = Z 0 1 kT c2P c dc = 3m 10.

12 Mean KE = 1 mc2 = 3 kT 2 2 2 2 2 c2 = vx + vy + vz 2 = 3vx 2 vx = kT m 10.2 Number of impacts on Area A of a wall θ A .14 10. 1 2 10.15 10.13 10.

KINETIC THEORY OF GASES The distribution of speeds : P c dc 74 The fraction of molecules travelling in direction .CHAPTER 10.

! .

+ d.

. ! + d  is cd.

csin.

d = sin.

d.

d 10.16 4c2 4 The fraction that will reach A in time t for above direction = Ac cos .

Number of impacts on area A in time t = NP c dc sin.17 V where V is the volume of the gas.t 10. Therefore.

d.

d Ac cos .

t 4 V Number of impacts on unit area per unit time 2 N Z 1 cP c dc Z  sin .

cos .

d.

19 10. n = N . Z 2 d = 4V 0 0 0 N c 1 2 = 4V 2 1  = 4 nc where the number density.3 E usion of molecules The number of molecules with speeds c!c+dc from direction . V ZZZ 10.18 10.

! .

+ d.

! + d  imping a area in time t is N cP c dcsin..

d.

ce : R 2 P dc ce = R ccP cc dc   . The mean speed of escape.d At 4V The distribution is proportional to c P c dc.

m 1 2 = 9kT 8m = 3kT 1 8kT 2 m 10.20 .

KINETIC THEORY OF GASES Number of molecules escaping per unit time per unit area 1 mc = p 8kT 4  kT m 1 2 m1 So.CHAPTER 10. 75 . the smaller molecules escapese uses faster.

 10. 1. dx Then.22 10.4 Mean Free Path. The distribution of free path : Probability that free path in between x and x+dx We can prove this by a self consistent argument : Probablity that collision occurs between x and x+dx = Probability of no collision from 0 to x  Probability of collisions in dx Comparing with the above equation.24 Probability of collision in dx = dx  According to 10. Probability of collision in dx = 1 .21 10. we write Probability of no collision in dx = e . 1 2 10. e dx + :::: =1. for some are longer and some are shorter.23 10. The more probable the collision.   dx  . the shorter is the free path.dx  x e.22.  the probability of no collision in dx = e. It is not all equal.  dx .  The free path is de ned as the distance between collisions.

dx  10.25 which agrees with 10.23. . .

Mean free path = = 76 10.   5 m.27 10.  1 0 = 10. Also. 11 00 11 00 11 00 11 00 d d The moving gas molecules sweeps out a cylinder of cross section : = d2 where d is the diameter of the each molecule. KINETIC THEORY OF GASES So the argument is self consistent.26 Z 0 1  .  dx  Z 0 1 x e. 10.   7  10.5 Collision cross-section.e.xe.31 . Therefore.30 10. Mean free path.28 So.29 10. we see that Z 1 x dx e.6 mbar. 1  = p n1 2 At stp. Average volume occupied by each molecule = V = 1 N n 10.  = V  1 2 N d 1 = nd2 Taking in account of the motion of gas molecules.  dx x = .8m.CHAPTER 10. x 1 + 0 x xe.   = 1 0 which is as required for probability. At 10.