This action might not be possible to undo. Are you sure you want to continue?

Statistical Physics

(FYS3130)

Prepared by Yuri Galperin

Spring 2004

2

Contents

1 General Comments 5

2 Introduction to Thermodynamics 7

2.1 Additional Problems: Fluctuations . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.2 Mini-tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

2.2.1 A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

2.2.2 B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

3 The Thermodynamics of Phase Transitions 33

3.1 Mini-tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

3.1.1 A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

3.1.2 B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

4 Elementary Probability Theory . . . 49

5 Stochastic dynamics ... 59

6 The Foundations of Statistical Mechanics 77

7 Equilibrium Statistical Mechanics 79

8 Tests and training 103

A Additional information 107

A.1 Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

A.1.1 Thermodynamic potentials . . . . . . . . . . . . . . . . . . . . . . . . . 107

A.1.2 Variable transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

A.1.3 Derivatives from the equation of state . . . . . . . . . . . . . . . . . . . 107

A.2 Main distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

A.3 The Dirac delta-function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

A.4 Fourier Series and Transforms . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

B Maple Printouts 115

3

4 CONTENTS

Chapter 1

General Comments

This year the course will be delivered along the lines of the book [1]. The problems will also be

selected from this book. It is crucially important for students to solve problems independently.

The problems will be placed on the course homepage. The same page will contain solutions

of the problems. So if a student is not able to solve the problem without assistance, then she/he

should come through the solution. In any case, student have to able to present solutions, obtained

independently, or with help of the course homepage.

FYS 3130 (former FYS 203) is a complicated course, which requires basic knowledge of

classical and quantum mechanics, electrodynamics, as well as basics of mathematics.

Using a simple test below please check if your knowledge is sufﬁcient. Answers can be found

either in the Maple ﬁle.

A simple test in mathematics

Elementary functions

Problem 1.1: Function is given by the deﬁnition

f (x) = x

4

+ax

3

+bx

2

.

(a) Show that by a proper rescaling it can be expressed as

f (x) = a

4

F

η

(ξ) where F

η

(ξ) = ξ

4

+ξ

3

+ηξ

2

,

ξ ≡x/a, η ≡b/a

2

. Function F

η

(ξ) is very important in theory of phase transitions.

(b) Investigate F

η

(ξ).

– How many extrema it has?

– When it has only 1 minimum? When it has 2 minima? At what value of η it has an

inﬂection point?

5

6 CHAPTER 1. GENERAL COMMENTS

– Plot F

η

(ξ) for this value of η.

Problem 1.2: Logarithmic functions are very important in statistical physics. Check you mem-

ory by the following exercises:

• Plot function

f

1

(x) = ln

1−x

1+x

for [x[ ≤1.

Discuss properties of this function at [x[ > 1.

• Plot function

f

2

(x) = ln(tanπx).

• Simplify the expression

e

4lnx

−(x

2

+1)

2

+2x

2

+1.

Problem 1.3: Do you remember trigonometry? Test it!

• Simplify

sin

−2

x −tan

−2

x −1.

• Calculate inﬁnite sums

C(α, β) =

∞

∑

n=0

e

−αn

cosβn, α > 0.

S(α, β) =

∞

∑

n=0

e

−αn

sinβn, α > 0.

Hint: take into account that cosx = Re

_

e

ix

_

, sinx = Im

_

e

ix

_

.

Basic integrals

Problem 1.4: Calculate interglars:

•

I

n

(γ) =

∞

0

x

n

e

−γx

dx, γ > 0.

•

dx

x

2

±a

2

,

dx

x(1−x)

.

•

G(γ) =

∞

0

e

−γx

2

/2

dx, γ > 0.

Chapter 2

Introduction to Thermodynamics

Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Problem 2.1: Test the following differentials for exactness. For those cases in which the dif-

ferential is exact, ﬁnd the function u(x, y).

(a) du

a

=

−ydx

x

2

+y

2

+

xdy

x

2

+y

2

.

(b) du

b

= (y −x

2

)dx +(x +y

2

)dy .

(c) du

c

= (2y

2

−3x)dx −4xydy .

Solution 2.1:

(a) The differential is exact, u

a

(x, u) =−arctan(x/y).

Here one point worth discussion. The function u

a

(x, u) = −arctan(x/y) has a singularity

at x = 0, y = 0. As a result, any close path embedding this point contributes 2π to the vari-

ation of the quantity u(x, y). Consequently, this function cannot serve as a thermodynamic

potential if both positve and negative values of x and y have physical meaning.

(b) The differential is exact, u

b

(x, y) = yx +(y

3

−x

3

)/3.

(c) The differential is not exact.

The function u(x, y) is reconstructed in the following way. For an exact differential, du =u

x

dx+

u

y

dy,

u

x

=

∂u(x, y)

∂x

, u

y

=

∂u(x, y)

∂y

.

If we introduce

u

1

(x, y) =

x

0

u

x

(ξ, y)dξ,

7

8 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

then the difference f ≡u(x, y) −u

1

(x, y) is a function only of y. Consequently,

d f

dy

=

∂u(x, y)

∂y

−

∂u

1

(x, y)

∂y

= u

y

(x, y) −

x

0

∂u

x

(x, y)

∂y

dx.

As a result,

f (y) =

y

0

dηu

y

(x, η) −

y

0

x

0

dξdη

∂u

x

(ξ, η)

∂η

.

Finally,

u(x, y) =

x

0

dξu

x

(ξ, y) +

y

0

dηu

y

(x, η) −

y

0

x

0

dξdη

∂u

x

(ξ, η)

∂η

.

For calculation see the Maple ﬁle it1.mws.

Problem 2.2: Consider the two differentials

1. du

1

= (2xy +x

2

)dx +x

2

dy, and

2. du

2

= y(x −2y)dx −x

2

dy.

For both differentials, ﬁnd the change u(x, y) between two points, (a, b) and (x, y). Compute the

change in two different ways:

(a) Integrate along the path (a, b) →(x, b) →(x, y),

(b) Integrate along the path (a, b) →(a, y) →(x, y).

Discuss the meaning of your results.

Solution 2.2: The calculations are shown in the Maple ﬁle. In the case (b) the results are

different because the differential is not exact.

Problem 2.3: Electromagnetic radiation in an evacuated vessel of volume V at equilibrium

with the walls at temperature T (black body radiation) behaves like a gas of photons having

internal energy U = aVT

4

and pressure P = (1/3)aT

4

, where a is Stefan’s constant.

(a) Plot the closed curve in the P−V plane for a Carnot cycle using black body radiation.

(b) Derive explicitly the efﬁciency of Carnot engine which uses black body radiation as its

working substance.

9

!

"

Vol ume

P

r

e

s

s

u

r

e

6

D

6

?

Figure 2.1: On the Carnot cycle with black-body radiation.

Solution 2.3: We will follow example shown in Exercise 2.2. Let us start with isotherms. At

the isotherms the pressure is V-independent, thus isotherms are horizontal, see Fig. 2.1 Along

the ﬁrst isothermal path,

∆Q

1→2

= ∆U +P∆V = (4/3)aT

4

h

(V

2

−V

1

) =

4

3

aT

4

h

V

1

_

V

2

V

1

−1

_

. (2.1)

In a similar way,

∆Q

3→4

=

4

3

aT

4

c

V

4

_

1−

V

3

V

4

_

. (2.2)

Now let us consider an adiabatic path. Along an adiabatic path,

dQ = 0 = dU +PdV = 4aVT

3

dT +aT

4

dV +(1/3)aT

4

dV = 4aVT

3

dT +(4/3)aT

4

dV .

Consequently,

dT

T

=−

1

3

dV

V

→ VT

3

= const, PV

4/3

= const .

Let us start form the point 2 characterized by the values P

1

,V

2

and adiabatically expand the

gas to the point 3 characterized by the volume V

3

. We have V

2

T

3

h

= V

3

T

3

c

. In a similar way,

V

4

T

3

c

=V

1

T

3

h

. Combining these equalities, we get:

V

2

T

3

h

=V

3

T

3

c

, V

4

T

3

c

=V

1

T

3

h

→

V

2

V

3

=

V

1

V

4

=

_

T

c

T

h

_

3

(2.3)

Combining Eqs. (2.1), (2.2) and (2.3), we ﬁnd

∆W = ∆Q

1→2

+∆Q

3→4

=

4

3

aT

4

h

V

1

_

V

3

V

4

−1

__

1−

T

c

T

h

_

.

10 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

Remember: along a closed path ∆U = 0 and the total heat consumption is equal to mechanical

work.

On the

∆Q

1→2

=

4

3

aT

4

h

V

1

_

V

2

V

1

−1

_

.

As a result,

η =

∆W

∆Q

1→2

=

T

h

−T

c

T

h

as it should be.

Problem 2.4: A Carnot engine uses a paramagnetic substance as its working substance. The

equation of state is

M =

nDH

T

where M is magnetization, H is the magnetic ﬁeld, n is the number of moles, D is a constant

determined by the type of substance, and T is is the temperature.

(a) show that the internal energy U, and therefore the heat capacity C

M

, can only depend on

the temperature and not the magnetization.

(b) Let us assume that C

M

=C = constant. Sketch a typical Carnot cycle in the M−H plane.

(c) Compute the total heat absorbed and the total work done by the Carnot engine.

(d) Compute the efﬁciency of the Carnot engine.

Solution 2.4:

(a) By deﬁnition [see Eq. [1](2.23)],

dU = T dS+HdM.

Thus at M = const the internal energy is independent of the magnetization.

(b) Since C = const, U

0

= Nc

M

T, where c

M

is the speciﬁc heat per one particle while N is the

number of particles. Introducing molar quantities we get U

0

= ncT.

A Carnot cycle is shown in Fig. 2.2. We have:

(c) For an isothermal process at T = T

c

,

Q

1→2

=−

H

2

H

1

H(M)dM =

T

c

2nD

(M

2

1

−M

2

2

).

In a similar way,

Q

3→4

=

T

h

2nD

(M

2

3

−M

2

4

).

11

M

H

1

2

3

4

T

T

h

c

Figure 2.2: Sketch of the Carnot cycle.

The total work is then W = Q

1→2

+Q

3→4

, and the efﬁciency is

η =

W

Q

3→4

= 1+

T

c

T

h

M

2

1

−M

2

2

M

2

3

−M

2

4

.

Now let us discuss adiabatic paths. We have at each path,

0 = dQ = dU −HdM = ncdT −

MT

nD

dM.

Immediately we get

dT

T

=

1

n

2

cD

MdM.

Integrating this equality from point 2 to point 3 we obtain,

2n

2

cD ln

T

h

T

c

= M

3

3

−M

2

2

.

In a similar way, integrating from 4 to 1 we obtain

2n

2

cD ln

T

c

T

h

= M

3

1

−M

2

4

.

As a result,

M

2

3

−M

2

2

= M

2

4

−M

2

1

→ M

2

3

−M

2

4

=−(M

1

−M

2

2

).

(d) Using this expression we obtain the efﬁciency

η =

Q

1→2

+Q

3→4

Q

3→4

=

T

h

−T

c

T

h

.

12 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

Coming back to the item (c) we ﬁnd

W = Q

3→4

η =

M

2

3

−M

2

4

2nD

T

h

−T

c

T

2

h

.

Problem 2.5: Find the efﬁciency of the engine shown in Fig. 2.3 ([1] -Fig.2.18). Assume that

3

1

2

4

P

V

adi abati c

Figure 2.3: Sketch of the cycle.

the operating substance is an ideal monoatomic gas. Express your answer in terms of V

1

and V

2

.

(The processes 1 →2 and 3 →4 are adiabatic. The processes 4 →1 and 2 →3 occur at constant

volume).

v

Solution 2.5: Let us start with the processes at constant volume. The mechanical work during

theses processes does not take place. Consequently,

Q

2→3

= (3/2)nR(T

3

−T

2

),

Q

4→1

= (3/2)nR(T

1

−T

4

),

W = (3/2)nR(T

1

+T

3

−T

2

−T

4

).

The efﬁciency is given by the expression

η =

T

1

+T

3

−T

2

−T

4

T

1

−T

4

= 1−

T

2

−T

3

T

1

−T

4

. (2.4)

Now let us consider the adiabatic processes where TV

2/3

=const (monoatomic ideal gas!). Thus,

T

1

T

2

=

T

4

T

3

=

_

V

2

V

1

_

2/3

≡α.

13

Substituting this expression into Eq. (2.4) we obtain:

η = 1−

1

α

= 1−

_

V

1

V

2

_

2/3

.

Problem 2.6: One kilogram of water is compressed isothermally at 20

◦

C from 1 atm to 20

atm.

(a) How much work is required?

(b) How much heat is rejected?

Assume that the average isothermal compressibility of water during this process is κ

T

= 0.5

10

−4

(atm)

−1

and the average thermal expansivity of water during this process is α

P

= 210

−4

(

◦

C)

−1

.

Solution 2.6: Since for an isothermal process dQ = T dS we have

Q = T

P

2

P

1

_

∂S

∂P

_

T

dP.

Using the Maxwell relation for the Gibbs free [see Eq. [1]-(2.112)] energy we obtain

_

∂S

∂P

_

T

=−

_

∂V

∂T

_

P

=−Vα

P

.

Thus

Q =−T

P

2

P

1

V(P)α

T

(P)dP

Now let us assume that α

T

it P-independent, and

V(P) =V

0

[1−κ

T

(P−P

0

)] = (M/ρ)[1−κ

T

(P−P

0

)] .

As a result,

Q =−(M/ρ)Tα

T

(P

2

−P

1

)[1−κ

T

(P−P

0

)/2] .

We see that since the compressibility of water is very low one can neglect the correction due to

change in the volume and assume V ≈M/ρ. The mechanical work is

W =−

P

2

P

1

PdV =−

P

2

P

1

P

_

∂V

∂P

_

T

dP = (M/2ρ)κ

T

(P

2

2

−P

2

1

).

14 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

2J

0

2L

0

J

0

L

0

a

b

c

Figure 2.4:

Problem 2.7: Compute the efﬁciency of the heat engine shown in Fig. 2.4 (Fig. [1]-2.19). The

engine uses a rubber band whose equation of state is

J = αLT , (2.5)

where α is a constant, J is the tension, L is the length per unit mass, and T is the temperature in

Kelvins. The speciﬁc heat (heat capacity per unit mass) is a constant, c

L

= c.

Solution 2.7: From Fig. 2.4 we see that the path a →b is isothermal. Indeed, since J ∝ L, it

follow from Eq. ( 2.5) that T = const. Then, from the same equation we get,

T

a

= T

b

= J

0

/αL

0

, T

c

= J

0

/2αL

0

= T

a

/2.

As a result,

Q

b→a

= MαT

a

2L

0

L

0

LdL = (3/2)MαL

2

0

T

a

= (3/2)MJ

0

L

0

,

Q

a→c

= Mc(T

c

−T

a

) =−(1/2)McT

a

,

Q

c→b

= Mc(T

b

−T

c

) −MJ

0

L

0

= (1/2)McT

a

−MJ

0

L

0

.

Hence, the total work is

W = Q

b→a

+Q

a→c

+Q

c→b

= (1/2)MJ

0

L

0

and

η =

W

Q

b→a

=

1

3

.

This result is also clear from geometrical point of view.

15

Problem 2.8: Experimentally one ﬁnds that for a rubber band

_

∂J

∂L

_

T

=

aT

L

0

_

1+2

_

L

0

L

_

3

_

,

_

∂J

∂T

_

L

=

aL

L

0

_

1−

_

L

0

L

_

3

_

,

where J is the tension, a = 1.010

3

dyne/K, and L

0

= 0.5 m is the length of the band when no

tension is applied. The mass of the rubber band is held ﬁxed.

(a) Compute (∂L/∂T)

J

and discuss its physical meaning.

(b) Find the equation of state and show that dJ is an exact differential.

(c) Assume that heat capacity at constant length is C

L

= 1.0 J/K. Find the work necessary to

stretch the band reversibly and adiabatically to a length 1 m. Assume that when no tension

is applied, the temperature of the band is T = 290 K. What is the change in temperature?

Solution 2.8:

(a) We have

_

∂L

∂T

_

J

_

∂T

∂J

_

L

_

∂J

∂L

_

T

=−1.

Consequently,

_

∂L

∂T

_

J

=−

__

∂T

∂J

_

L

_

∂J

∂L

_

T

_

−1

=−

_

∂J

∂T

_

L

_

∂J

∂L

_

T

=−

L

T

1−

_

L

0

L

_

3

1+2

_

L

0

L

_

3

.

Physical meaning -

α

J

=

1

L

_

∂L

∂T

_

J

is the thermal expansion coefﬁcient at given tension.

(b) The equation of state has the form

J =

aTL

L

0

_

1−

_

L

0

L

_

3

_

.

The proof of the exactness is straightforward.

16 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

(c) Consider adiabatic expansion of the band,

0 = dQ =C

L

dT +J(L, T)dL.

Consequently,

dT

dL

=−

J(L, T)

C

L

=−

aTL

C

L

L

0

_

1−

_

L

0

L

_

3

_

.

Measuring length in units of L

0

as L = L

0

and introducing β ≡ aL

0

/C

L

we obtain the

following differential equation

dT/T =−β(1−

−3

)d.

Its solution is

ln

T

f

T

0

=−

β

2

_

L

f

L

0

_

2

_

1+2

_

L

0

L

f

_

3

_

.

Here L

f

and T

f

are ﬁnal values of the length and temperature, respectively. This is the

equation for adiabatic process which provides the change in the temperature. The mechan-

ical work is then

W =C

L

(T

f

−T

0

).

Problem 2.9: Blackbody radiation in a box of volume V and at temperature T has internal

energy U = aVT

4

and pressure P = (1/3)aT

4

, where a is the Stefan-Boltzmann constant.

(a) What is the fundamental equation for the blackbody radiation (the entropy)?

(b) Compute the chemical potential.

Solution 2.9:

(a) Let us ﬁrst ﬁnd the Helmholtz free energy. From the relation P =−

_

∂A

∂V

_

T

we get

A =−PV =−

1

3

aVT

4

=−

1

3

U .

Consequently,

S =−

_

∂A

∂T

_

V

=

4

3

aVT

3

=

4

3

U

T

.

As a result,

_

∂S

∂U

_

V

=

∂(S,V)

∂(U,V)

=

∂(S,V)/∂(T,V)

∂(U,V)/∂(T,V)

=

(∂S/∂T)

V

(∂U/∂T)

V

=

1

T

.

(b) Since A is N-independent, µ = 0.

17

Problem 2.10: Two vessels, insulated from the outside world, one of volume V

1

and the other

of volume V

2

, contain equal numbers N of the same ideal gas. The gas in each vessel is originally

at temperature T

i

. The vessels are then connected and allowed to reach equilibrium in such a way

that the combined vessel is also insulated fromthe outside world. The ﬁnal volume is V =V

1

+V

2

.

What is the maximum work, δW

f ree

, that can be obtained by connecting these insulated vessels?

Express your answer in terms of T

i

, V

1

, V

2

, and N.

Solution 2.10: The Gibbs free energy of an ideal gas is given by the equation

G = NkT lnP+Nχ(T)

where χ(T) is some function of the gas excitation spectrum.

1

Consequently,

S =−

_

∂G

∂T

_

P

=−NklnP−Nχ

/

(T).

Before the vessels are connected,

S

i

=−Nkln(P

1

P

2

) −2Nχ

/

(T).

After the vessels are connected the temperature remains the same, as it follows from the conser-

vation law, the entropy being

S

f

=−2NklnP−2Nχ

/

(T).

Consequently, ∆S =−Nkln(P

2

/P

1

P

2

). On the other hand,

1

P

=

V

1

+V

2

2NkT

i

,

1

P

i

=

V

i

NkT

i

→

P

2

P

1

P

2

=

4V

1

V

2

(V

1

+V

2

)

2

.

1

For an ideal gas, the energy of the particle can be written as a sum of the kinetic energy, ε

p

= p

2

/2m, and the

energy of internal excitations, ε

α

(characterized by some quantum numbers α),

ε

pα

= ε

p

+ε

α

.

As we will see later, the free energy of the ideal gas can be constructed as

A = −

kT

N!

ln

_

∑

pα

e

−ε

pα

/kT

_

N

≈−NkT ln

_

eV

N

_

mkT

2π¯ h

2

_

3/2

∑

α

e

−ε

α

/kT

_

= −NT ln(eV/N) +N f (T),

f (T) = −kT ln

_

_

mkT

2π¯ h

2

_

3/2

∑

α

e

−ε

α

/kT

_

.

The Gibbs free energy is then

G = A+PV = NkT lnP+Nχ(T), χ(T) ≡ f (T) −kT lnkT .

18 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

As a result, the maximum work is

∆W

f ree

= T

i

∆S = NkT

i

ln

(V

1

+V

2

)

2

4V

1

V

2

.

Problem 2.11: For a low-density gas the virial expansion can be terminated at ﬁrst order in the

density and the equation of state is

P =

NkT

V

_

1+

N

V

B

2

(T)

_

,

where B

2

(T) is the second virial coefﬁcent. The heat capacity will have corrections to its ideal

gas value. We can write it in the form

C

V,N

=

3

2

Nk −

N

2

k

V

f (T).

(a) Find the form that f (T) must have in order for the two equations to be thermodynamically

consistent.

(b) Find C

P,N

.

(c) Find the entropy and internal energy.

Solution 2.11:

(c) The equation of state under consideration can be obtained from the Helmholtz free energy

2

A = A

ideal

+kTB

2

(T)

N

2

V

.

Then

S =−

_

∂A

∂T

_

V

= S

ideal

+δS, δS ≡−k

N

2

V

_

B

2

(T) +TB

/

2

(T)

¸

.

Since we know both entropy and Helmholtz free energy, we ﬁnd the internal energy as

U = A+TS =U

ideal

+kTB

2

(T)

N

2

V

−kT

N

2

V

_

B

2

(T) +TB

/

2

(T)

¸

= U

ideal

+δU , δU ≡−kT

2

B

/

2

(T)

N

2

V

.

2

Remember that P =−

_

∂A

∂V

_

T

.

19

(a) As a result,

C

V

= T

_

∂S

∂T

_

V

=

_

∂U

∂T

_

V

=C

ideal

V

+δC

V

, δC

V

≡−kT

N

2

V

_

2B

/

2

(T) +TB

//

2

(T)

¸

.

We ﬁnd in this way,

f (T) = 2TB

/

2

(T) +T

2

B

//

2

(T).

(b) Let us express the equation of state as V(P, T). Since the density is assumed to be small in

the correction one can use equation for the ideal gas to ﬁnd the the volume. We have,

V =

NkT

P

+NB

2

(T).

Consequently, the entropy can be expressed as

S = S

ideal

+δS

1

, δS

1

≡−kN

B

2

+TB

/

2

kT/P+B

2

.

Now

C

P

= C

ideal

P

+T

_

∂δS

1

∂T

_

P

= −kN

(B

2

+TB

/

2

)

/

(kT/P+B

2

) −(k/P+B

/

2

)(B

2

+TB

/

2

)

(kT/P+B

2

)

2

= C

ideal

P

−k

N

2

V

_

T(2B

/

2

+TB

//

2

) −(B

2

+TB

/

2

)

¸

= C

ideal

P

+δC

V

+k

N

2

V

(B

2

+TB

/

2

).

Here in all corrections we used equation of state for an ideal gas. As a result,

C

P

−C

V

= (C

P

−C

V

)

ideal

+k

N

2

V

(B

2

+TB

/

2

).

Problem 2.12: Prove that

C

Y,N

=

_

∂H

∂T

_

Y,N

and

_

∂H

∂Y

_

T,N

= T

_

∂X

∂T

_

Y,N

−X .

20 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

Solution 2.12: Let us ﬁrst recall deﬁnitions for X, Eq. ([1]-2.66), U = ST +YX +∑

J

jµ

/

j

dN

j

.

The enthalpy is deﬁned as

H =U −XY = ST +

∑

J

µ

/

j

dN

j

.

Since dU = T dS+Y dX we get

dH = T dS−X dY .

Consequently,

C

Y,N

= T

_

∂S

∂T

_

Y,N

=

_

∂H

∂T

_

Y,N

.

To prove the second relation we do the following

_

∂H

∂Y

_

T,N

=

_

∂H

∂Y

_

S,N

+

_

∂H

∂S

_

Y,N

_

∂S

∂Y

_

T

.

Now,

_

∂H

∂Y

_

S,N

=−X ,

_

∂H

∂S

_

Y,N

= T .

Now we have to use the Maxwell relation, which emerges for the Gibbs free energy G=H−TS.

From

dG =−SdT −X dY

we get

_

∂S

∂Y

_

T

=

_

∂X

∂T

_

Y,N

.

Thus we obtain the desired result.

Problem 2.13: Compute the entropy, enthalpy, Helmholtz free energy, and Gibbs free energy

for a paramagnetic substance and write them explicitly in terms of their natural variables if pos-

sible. Assume that mechanical equation of state is m = (D/T)H and the the molar heat capacity

at constant magnetization is c

m

= c, where m is the molar magnetization, H is the magnetic ﬁeld,

D is a constant, c is a constant, and T is the temperature.

Solution 2.13: Let us start with the internal energy u(T, m) per one mole. We have the mag-

netic contribution u

mag

=

m

0

H(m)dm. Since H(m) = (T/D)m we get u

mag

= (T/2D)m

2

. The

“thermal” contribution is cT. As a result,

u(T, m) = T(c +m

2

/2D), T(u, m) =

u

c +m

2

/2D

.

The molar entropy s is then derived from the deﬁnition

_

∂s

∂u

_

m

=

1

T

=

c +m

2

/2D

u

→ s(u, m) = (c +m

2

/2D)ln(u/u

0

).

21

Here u

0

is a constant. As a result, in “natural” variables

u(s, m) = u

0

exp

_

s

c +m

2

/2D

_

.

To ﬁnd other thermodynamic potentials we need s(T, m). We can rewrite the above expression

for the entropy as

s(T, m) = (c +m

2

/2D)ln

T(c +m

2

/2D)

u

0

.

In particular, the Helmholtz free energy is

a(T, m) = u−Ts = T(c +m

2

/2D)

_

1−ln

T(c +m

2

/2D)

u

0

_

.

To get enthalpy we have to subtract from u the quantity Hm= (D/T)H

2

and to express m through

H as m = (D/T)H. As a result, we obtain:

h(T, H) = u−Hm = T(c +DH

2

/2T

2

) −(D/T)H

2

= T(c −DH

2

/2T

2

).

Finally, Gibbs free energy is g = a−Hm, which has to be expressed through T and H. We get

g = a−Hm = T(c +DH

2

/2T

2

)

_

1−ln

T(c +DH

2

/2T

2

)

u

0

_

−(D/T)H

2

= T(c −DH

2

/2T

2

) −T(c +DH

2

/2T

2

)ln

T(c +DH

2

/2T

2

)

u

0

.

Problem 2.14: Compute the Helmholtz free energy for a van der Waals gas. The equation of

state is

_

P+α

n

2

V

2

_

(V −nb) = nRT ,

where α and b are constants which depend on the type of gas and n is the number of moles.

Assume that heat capacity is C

V,n

= (3/2)nR.

Is this a reasonable choice for the heat capacity? Should it depend on volume?

Solution 2.14: Let us express pressure through the volume,

P =

nRT

V −nb

−

αn

2

V

2

.

Since P =−∂A/∂V we obtain

A =−

P(V)dV =−nRT ln(V −nb) −(αn

2

/V) +A(T).

22 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

Here A(T) is the integration constant, which can be found from the given speciﬁc heat. Indeed,

S =

T

C

V,n

dT

/

T

/

= (3/2)nRln(T/T

0

).

Consequently,

A(T) =−

T

S(T

/

)dT

/

= (3/2)nRT[1−ln(T/T

0

)] .

Here we omit temperature-independent constant.

The suggestion regarding speciﬁc heat is OK since the difference between the entropies of

van der Waals gas and the ideal gas is temperature independent. (Check!)

Problem 2.15: Prove that

(a) κ

T

(C

P

−C

V

) = TVα

2

P

(b) C

P

/C

V

= κ

T

/κ

S

.

Solution 2.15: We use the method of Jacobians:

(a)

C

V

= T (∂S/∂T)

V

= T∂(S,V)/∂(T,V) =

∂(S,V)/∂(T, P)

∂(T,V)/∂(T, P)

= T

(∂S/∂T)

P

(∂V/∂P)

T

−(∂S/∂P)

T

(∂V/∂T)

P

(∂V/∂P)

T

= C

P

−T

(∂S/∂P)

T

(∂V/∂T)

P

(∂V/∂P)

T

.

Now, from the Maxwell relations (∂S/∂P)

T

=−(∂V/∂T)

P

. Thus,

C

P

−C

V

=−T

[(∂V/∂T)

P

]

2

(∂V/∂P)

T

= TV

α

2

P

κ

T

.

The ﬁrst relation follows from this in a straightforward way from deﬁnitions.

(b) Let us ﬁrst calculate the adiabatic compressibility (∂V/∂P)

S

as

_

∂V

∂P

_

S

=

∂(V, S)

∂(P, S)

=

∂(V, S)/∂(V, T)

∂(P, S)/∂(P, T)

∂(V, T)

∂(P, T)

=

(∂S/∂T)

V

(∂S/∂T)

P

_

∂V

∂P

_

T

.

Consequently,

C

P

C

V

=

(∂V/∂P)

T

(∂V/∂P)

S

=

κ

T

κ

S

.

23

Problem 2.16: Show that

Tds = c

x

(∂T/∂Y)

x

dY +c

Y

(∂T/∂x)

Y

dx,

where x = X/n is the amount of extensive variable, X, per mole, c

x

is the heat capacity per mole

at constant x, and c

Y

is the heat capacity per mole at constant Y.

Solution 2.16: Let us substitute the deﬁnitions

c

x

= T (∂S/∂T)

x

, c

Y

= T (∂S/∂T)

Y

.

Now the combination c

x

(∂T/∂Y)

x

dY +c

Y

(∂T/∂x)

Y

dx can be rewritten as

T (∂S/∂T)

x

(∂T/∂Y)

x

dY +T (∂S/∂T)

Y

(∂T/∂x)

Y

dx

= T (∂S/∂Y)

x

dY +T (∂S/∂x)

Y

dx = T ds.

Problem 2.17: Compute the molar heat capacity c

P

, the compressibilities, κ

T

and κ

S

, and

the thermal expansivity α

P

of a monoatomic van der Waals gas. Start from the fact that the

mechanical equation of state is

P =

RT

v −b

−

α

v

2

.

and the molar heat capacity is c

v

= 3R/2, where v =V/n is the molar volume.

Solution 2.17: Let us start with ther speciﬁc heat. Using the method similar to the Problem

2.15 we can derive the relation

c

P

−c

v

=−T

[(∂P/∂T)

v

]

2

(∂P/∂v)

T

=

R

1−2α(v −b)

2

/RTv

3

.

Now let us compute the compressibility

κ

T

=−

1

v

_

∂v

∂P

_

T

=−

1

v

_

∂P

∂v

_

−1

T

=

(v −b)

2

vRT

1

1−2α(v −b)

2

/RTv

3

.

Given c

v

, other quantities can be calculated using results of the Problem 2.15.

Problem 2.18: Compute the heat capacity at constant magnetic ﬁeld C

H,n

, the susceptibilities

χ

T,n

and χ

S,n

, and the thermal expansivity α

H,n

for a magnetic system, given that the mechanical

equation of state is M = nDH/T and the heat capacity C

M,n

= nc, where M is the magnetization,

H is the magnetic ﬁeld, n is the number of moles, c is the molar heat capacity, and T is the

temperature.

24 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

Solution 2.18: Let us start with susceptibilities. By deﬁnition,

χ

T,n

=

_

∂M

∂H

_

T,n

=

nD

T

.

Now,

_

∂M

∂H

_

S

=

∂(M, S)

∂(H, S)

=

∂(M, S)/∂(M, T)

∂(H, S)/∂(H, T)

∂(M, T))

∂(H, T)

= χ

T,n

C

M,n

C

H,n

.

Thus we have found one relation between susceptibilities and heat capacities,

χ

S,n

χ

T,n

=

C

M,n

C

H,n

.

Now let us ﬁnd C

H,n

. At constant H, M becomes dependent only on temperature. Then,

dM =

dM

dT

dT =−

nDH

T

2

dT .

Consequently, the contribution to the internal energy is

dU =−HdM =

nDH

2

T

2

dT .

As a result,

C

H,n

−C

M,n

=

nDH

2

T

2

.

Given C

H,n

= nc we easily compute C

H,n

and χ

S,n

.

According to Eq. (R2.149), α

H

is deﬁned as

α

H

=

_

∂M

∂T

_

H

=−

nDH

T

2

.

Problem 2.19: A material is found to have a thermal expansivity α

P

= v

−1

(R/P+a/RT

2

) and

an isothermal compressibility κ

T

= v

−1

[T f (P) +b/P]. Here v =V/n is the molar volume.

(a) Find f (P).

(b) Find the equation of state.

(c) Under what condition this materials is stable?

25

Solution 2.19:

(a) By deﬁnition, we have

∂v

∂T

=

R

P

+

a

RT

2

.

∂v

∂P

= −T f (P) −

b

P

.

To make dv an exact differential we need:

−

R

P

2

=−f (p).

Thus

f (p) = R/P

2

.

(b) We can reconstruct the equation of state as:

v =

P

∂v

∂P

dP =

P

dP

_

−

RT

P

2

−

b

P

_

=

RT

P

−blnP+g(T).

Here g(T) is some function of the temperature. Now,

∂v

∂T

=

R

P

+g

/

(T) ≡

R

P

+

a

RT

2

.

Thus g(T) =−a/RT +const. As a result, we can express the equation of state as

v −v

0

=

RT

P

+bln

P

0

P

−

a

RT

.

(c) Since the compressibility must be positive, we have the stability condition

T f (P) +

b

P

> 0 →

TR

P

2

+

b

P

> 0.

Consequently, the stability condition is

P/T < R/b.

Problem 2.20: Compute the efﬁciency of the reversible two heat engines in Fig. 2.5 (R2.20).

Which engine is the most effective? (Note that these are not Carnot cycles. The efﬁciency of a

heat engineis η = ∆W

total

/∆Q

absorbed

.

26 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

6

6

6

5

5

5

6

6

6

5

5

5

=

=

>

>

?

?

=

>

Figure 2.5:

Solution 2.20: Since dQ = T dS, we immediately get for any closed path in the T −S plane:

∆W

total

=

T dS.

This is just the area of the triangle,

∆W

total

= (1/2)(T

2

−T

1

)(S

2

−S

1

).

The heat absorbed in the case (a) is

∆Q

absorbed

= T

2

(S

2

−S

1

).

Thus,

η

a

=

T

2

−T

1

2T

2

.

In the case (b), it easy to show that

∆Q

absorbed

= (1/2)(T

2

+T

1

)(S

2

−S

1

).

Thus

η

b

=

T

2

−T

1

T

2

+T

1

> η

a

.

Problem 2.21: It is found for a gas that κ

T

= Tv f (P) and α

P

= Rv/P+Av/T

2

, where T is

the temperature, v is the molar volume, P is the pressure, A is a constant, and f (P) is unknown

function.

(a) What is f (P)?

(b) Find v = v(P, T).

27

Solution 2.21: The solution is similar to the problem 2.19. We have:

∂v

∂T

= v

2

_

R

P

+

A

T

2

_

,

∂v

∂P

= −v

2

T f (P).

Let us introduce γ(P, T) ≡[v(P, T)]

−1

. We get,

∂γ

∂T

= −

R

P

−

A

T

2

,

∂γ

∂P

= T f (P).

Again, from the Maxwell relation we get

R

P

2

= f (p). Then we can express γ as

γ =

P

∂γ

∂P

dP = TR

P

dP

P

2

=−

RT

P

+g(T).

Then, g

/

(T) =−A/T

2

, or g(T) = A/T +const. As a result,

γ = γ

0

+

A

T

−

RT

P

.

Consequently,

v(P, T) =

1

γ

0

+A/T −RT/P

.

Problem 2.22: A monomolecular liquid at volume V

L

and pressure P

L

is separated from a gas

of the same substance by a rigid wall which is permeable to the molecules, but does not allow

liquid to pass.The volume of the gas is held ﬁxed at V

G

, but the volume of the liquid cam be

varied by moving a piston. If the pressure of the liquid is increased by pushing in on the piston,

by how much does the pressure of the gas change? [Assume the liquid in incompressible (its

molar volume is independent of the pressure) and describe the gas by the ideal gas equation of

state. The entire process occurs at ﬁxed temperature T].

Solution 2.22: Let us consider the part of the system, which contains both liquid and gas

particles. In this part the chemical potentials must be equal, µ

L

= µ

G

. On the other hand, the

chemical potentials and pressures of gas in both parts must be equal. Thus we arrive at the

equation,

µ

L

(P

L

, T) = µ

G

(P

G

, T).

If one changes the pressure of liquid by δP

L

, then

_

∂µ

L

∂P

L

_

T

δP

L

=

_

∂µ

G

∂P

G

_

T

δP

G

.

28 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

For an ideal gas, (∂µ/∂P)

T

= kT/P

G

=V

G

/N

G

≡ ˜ v

L

. the quantity ˜ v

L

has a physical meaning of

the volume per particle. For a liquid, the relation is the same,

_

∂µ

∂P

_

T

=

∂

∂P

_

∂G

∂N

_

P,T

=

∂

∂N

_

∂G

∂P

_

N,T

=

_

∂V

∂N

_

T

= ˜ v

L

.

The last relation is a consequence of incompressible character of the liquid. As a result,

δP

G

δP

L

=

˜ v

L

˜ v

G

.

2.1 Additional Problems: Fluctuations

Quick access: 23 24 25 26 27

Problem 2.23: Find the mean square ﬂuctuation of the internal energy (using V and T as inde-

pendent variables). What is the mean square ﬂuctuation of the internal energy for a monoatomic

ideal gas?

Solution 2.23: We have

∆U =

_

∂U

∂V

_

T

∆V +

_

∂U

∂T

_

V

∆T =

_

T

_

∂P

∂T

_

V

−P

_

∆V +C

V

∆T .

Here we use Maxwell relation which can be obtained from Helmholtz free energy. Squaring and

averaging we obtain (note that ¸∆V ∆T) = 0)

¸(∆U)

2

) =

_

T

_

∂P

∂T

_

V

−P

_

2

¸(∆V)

2

) +C

2

V

¸(∆T)

2

).

Now,

¸(∆T)

2

) = kT

2

/C

V

, ¸(∆V)

2

) =−kT

_

∂V

∂P

_

T

.

Thus

¸(∆U)

2

) =−kT

_

T

_

∂P

∂T

_

V

−P

_

2

_

∂V

∂P

_

T

+C

V

kT

2

.

For the ideal gas,

P = NkT/V , C

V

= (3/2)Nk.

Thus,

¸(∆U)

2

) = (3/2)N(kT)

2

.

2.1. ADDITIONAL PROBLEMS: FLUCTUATIONS 29

Problem 2.24: Find ¸∆T ∆P) (with variables V and T) in general case and for a monoatomic

ideal gas.

Solution 2.24: We have

∆P =

_

∂P

∂V

_

T

∆V +

_

∂P

∂T

_

V

∆T .

Squaring and averaging we obtain (note that ¸∆V ∆T) = 0)

¸∆T ∆P) =

_

∂P

∂T

_

V

¸(∆T)

2

) =

_

kT

2

C

V

_ _

∂P

∂T

_

V

.

For an ideal gas,

¸∆T ∆P) =

2

3

kT

2

V

.

Problem 2.25: Find ¸∆V ∆P) with variable (V and T).

Solution 2.25: Using results of the previous problem we get

¸∆V ∆P) =

_

∂P

∂V

_

T

¸(∆V)

2

) =−kT .

Problem 2.26: Using the same method show that

¸∆S∆V) = kT

_

∂V

∂T

_

P

, ¸∆S∆T) = kT .

Solution 2.26: Starightforward.

Problem 2.27: Find a mean square ﬂuctuation deviation of a simple pendulum with the length

suspended vertically.

Solution 2.27: Let m be the pendulum mass, and φ is the angle of deviation from the vertical.

The minimal work is just the mechanical work done against the gravity force. For small φ,

W

min

= mg (1−cosφ) ≈mgφ

2

/2.

Thus,

¸φ

2

) = kT/mg.

30 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

2.2 Mini-tests

2.2.1 A

The Helmholtz free energy of the gas is given by the expression

A = Nε

0

−NkT ln(eV/N) −NcT lnT −NζT .

Here e =2.718. . . is the base of natural logarithms, N is the number of particles, V is the volume,

T is the temperature in the energy units, while ε

0

, c and ζ are constants.

(a) Find the entropy as function of V and T.

(b) Find internal energy U as a function of the temperature T and number of particles N.

(c) Show that c is the heat capacity per particle at given volume, c

v

.

(d) Find equation of state.

(e) Find Gibbs free energy, enthalpy. Find the entropy as a function of P and T.

(f) Using these expression ﬁnd the heat capacity at constant pressure, c

P

.

(g) Show that for an adiabatic process

T

γ

P

1−γ

= constant, where γ = c

P

/c

v

.

Solution

By deﬁnition,

S =−

_

∂A

∂T

_

V,N

= Nln

eV

N

+Nc(1+lnT) +NζT .

Now,

U = A+TS = Nε

0

+NcT ,

P = −

_

∂A

∂V

_

T,N

=

NT

V

,

G = A+PV = A+NT = Nε

0

−NT ln

V

N

−NcTlT +NζT

= Nε

0

+NT lnP−NT(1+c)lnT −NζT ,

W = U +PV =U +NT = Nε

0

+N(c +1)T ,

S = −NlnP+N(1+c)lnT +N(ζ+1+c),

c

P

= T

_

∂S

∂T

_

P,N

= 1+c.

2.2. MINI-TESTS 31

Immediately, to keep entropy constant we get

−NlnP+Nc

P

lnT = const → T

c

P

/P = const .

Since c

P

−c

V

= c

P

−c = 1, we obtain

T

c

P

P

c

P

−c

V

= const , → T

γ

P

1−γ

= const .

2.2.2 B

Problem 1

Discuss entropy variation for

(a) adiabatic process,

(b) isothermic process,

(c) isochoric process,

(d) isobaric process.

Problem 2

(a) Discuss the difference between Gibbs and Helmholtz free energy.

(b) Prove the relation

U =−T

2

_

∂

∂T

A

T

_

.

(c) A body with constant speciﬁc heat C

V

is heated under constant volume from T

1

to T

2

. How

much entropy it gains?

(d) Discuss the heating if the same body is in contact with a thermostat at T

2

. In the last case the

heating is irreversible. Show that the total entropy change is positive.

(e) Two similar bodies with temperatures T

1

and T

2

brought into contact. Find the ﬁnal tempera-

ture and the change in entropy.

Solution

Problem 1

(a) Constant

(b)

S = Q/T .

(c)

S =

T

2

T

1

C

V

(T)

T

dT =C

V

ln

T

2

T

1

.

32 CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

(d)

S =

T

2

T

1

C

P

(T)

T

dT =C

P

ln

T

2

T

1

.

Problem 2

Mechanical work under isothermic process is given by

dW = dU −dQ = dU −T dS = d(U −TS).

The function of state

A =U −TS

is called Helmholtz free energy. We have

dA = dU −T dS−SdT = (T dS−PdV) −T dS−SdT =−SdT −PdV.

Thus,

S =−

_

∂A

∂T

_

V

, P =−

_

∂A

∂V

_

T

.

Thus, A is the thermodynamic potential with respect to V and T.

The thermodynamic potential with respect to P and T is called Gibbs free energy. We have

G =U −TS−PV →dG =−SdT +V dP.

(b) Substituting

S =−

_

∂A

∂T

_

V

into deﬁnition of A we get the result.

(c) We have

S =

T

2

T

1

C

V

T

dT =C

V

ln

_

T

2

T

1

_

.

(d) Use 2nd law of thermodynamics

(e) Energy conservation law yields

C

V

(T

2

−T

B

) =C

V

(T

B

−T

1

) → T

B

= (T

1

+T

2

)/2.

for the entropy change we have

δS =C

V

ln

T

B

T

1

+C

V

ln

T

B

T

2

= 2C

V

ln

_

T

1

+T

2

2

√

T

1

T

2

_

≥0.

Chapter 3

The Thermodynamics of Phase Transitions

Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13

Problem 3.1: A condensible vapor has a molar entropy

s = s

0

+Rln

_

C(v −b)

_

u+

a

v

_

5/2

_

,

where c and s

0

are constants.

(a) Compute equation of state.

(b) Compute the molar heat capacities, c

c

and c

P

.

(c) Compute the latent heat between liquid and vapor phases temperature T in terms of the

temperature T, the gas constant R and gas molar volumes v

l

and v

g

. How can you ﬁnd

explicit values of v

l

and v

g

if you need to?

Solution 3.1:

a) Let us ﬁrst ﬁnd the temperature as function of u and s. We have

1

T

=

_

∂s

∂u

_

v

=

5R

2

_

u+

a

v

_

−1

.

Thus,

s = s

0

+Rln

_

C(5RT/2)

5/2

(v −b)

_

,

u =

5

2

RT −

a

v

,

a = u−Ts =

5

2

RT −

a

v

−T

_

s

0

+Rln

_

C(v −b)(5RT/2)

5/2

__

.

33

34 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

As a result,

P =−

_

∂a

∂v

_

T

=

RT

v −b

−

a

v

2

.

This is the van der Waals equation.

(b) We have,

c

v

= T

_

∂s

∂T

_

v

=

5

2

RT .

The result for c

P

can be obtained in the same way as in the problem 2.17.

(c) The latent heat q is just

q = T(s

g

−s

l

) = RT ln

v

g

−b

v

l

−b

.

Since the pressure should be constant along the equilibrium line, P

l

= P

G

, we have

RT

v

g

−b

−

a

v

2

g

=

RT

v

l

−b

−

a

v

2

l

.

Another equation is the equality of chemical potentials, µ

l

= µ

g

. We know that µ =

(∂a/∂n)

T,V

, so everything could be done. Another way is to plot the isotherm and ﬁnd

the volumes using the Maxwell rule.

Problem 3.2: Find the coefﬁcient of thermal expansion, α

coex

= v

−1

(∂v/∂T)

coex

, for a gas

maintained in equilibrium with its liquid phase. Find an approximate explicit expression for

α

coex

, using the ideal gas equation of state. Discuss its behavior.

Solution 3.2: It is implicitly assumed that the total volume of the system is kept constant. We

have,

∂v

∂T

=

_

∂v

∂T

_

P

+

_

∂v

∂P

_

T

dP

dT

.

For an ideal gas, v = RT/P. Thus

_

∂v

∂T

_

P

= R/P,

_

∂v

∂P

_

T

=−RT/P

2

.

Consequently,

α

coex

=

1

v

_

∂v

∂T

_

coex

=

1

T

_

1−

T

P

_

∂P

∂T

_

coex

_

.

According to the Clapeyron-Clausius formula,

_

∂P

∂T

_

coex

=

q

T(v

g

−v

l

)

≈

q

Tv

g

,

35

where q = (∆h)

lg

, we can rewrite this expression as

α

coex

≈

1

T

_

1−

q

RT

_

.

The coefﬁcient of thermal expansion is less because with the increase of the temperature under

given pressure the heat is extracted from the system.

Problem 3.3: Prove that the slope of the sublimation curve of a pure substance at the triple

point must be greater than that of the vaporization curve at the triple point.

Solution 3.3: The triple point is deﬁned by the following equations for the 2 phases:

P

1

= P

2

= P

3

, T

1

= T

2

= T

3

, µ

1

= µ

2

= µ

3

.

The deﬁnitions of the sublimation and vaporization curves are given in Fig. 3.4 of the text-

book [1]. Consequently, the slopes of the vaporization and sublimation curves are given by the

relations

_

∂P

∂T

_

lg

=

s

g

−s

l

v

g

−v

l

≈

s

g

−s

l

v

g

,

_

∂P

∂T

_

sg

=

s

g

−s

s

v

g

−v

s

≈

s

g

−s

s

v

g

.

Since solid state is more ordered, s

s

< s

l

, and

_

∂P

∂T

_

lg

−

_

∂P

∂T

_

sg

=

s

s

−s

l

v

g

< 0.

Problem 3.4: Consider a monoatomic ﬂuid along its ﬂuid-gas coexistence curve. Compute the

rate of change of chemical potential along the coexistence curve, (∂µ/∂T)

coex

, where µ is the

chemical potential and T is the temperature. Express your answer in terms of s

l

, v

l

and s

g

, v

g

which are the molar entropy and molar volume of the liquid and gas, respectively.

Solution 3.4: We have µ(P, T) = µ

l

(P, T) = µ

g

(P, T). Thus,

dµ

dT

=

_

∂µ

g

∂T

_

P

+

_

∂µ

g

∂P

_

T

dP

dT

.

Consequently,

_

dµ

dT

_

coex

=−s

g

+v

g

s

g

−s

l

v

g

−v

l

=

s

g

v

l

−v

g

s

l

v

g

−v

l

.

36 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

Problem 3.5: A system in its solid phase has a Helmholtz free energy per mole, a

s

= B/Tv

3

,

and in its liquid phase it has a Helmholtz free energy per mole, a

l

= A/Tv

2

, where A and B are

constants, v is the volume per mole, and T is the temperature.

(a) Compute the Gibbs free energy density of the liquid and solid phases.

(b) How are the molar volumes, v, of the liquid and solid related at the liquid-solid phase

transition?

(c) What is the slope of the coexistence curve in the P−T plane?

Solution 3.5:

(a) By deﬁnition, g = a+Pv, and P =−(∂a/∂v)

T

. Thus,

g = a−v(∂a/∂v)

T

, g

s

= 4B/Tv

3

s

, g

l

= 3A/Tv

2

l

.

(b) Since P =−(∂a/∂v)

T

we have,

P

s

= 3B/Tv

4

s

, P

l

= 2A/Tv

3

l

.

Since at the phase transition P

s

= P

l

= P, we obtain

v

3

l

v

4

s

=

2A

3B

.

(c) Now we can express the Gibbs free energies per mole in terms of P,

g

s

=

4

3

(3B)

1/4

P

3/4

T

−1/4

, g

l

=

3

2

(2A)

1/3

P

2/3

T

−1/3

.

Since at the transition point g

l

= g

s

, we have

P

T

=

_

3

7/4

2

8/3

A

1/3

B

1/4

_

12

=

3

21

2

24

A

4

B

3

.

Problem 3.6: Deduce the Maxwell construction using stability properties of the Helmholtz

free energy rather than the Gibbs free energy.

Solution 3.6: Since the system in the equilibrium, the maximum work extracted during the

process of the phase transition is ∆A−P

0

∆V. Let us look at Fig. 3.1.

37

P

V

a

b

c

v

l

v

g

P

0

Figure 3.1:

We see that P

0

∆V = P

0

(V

g

−V

l

) is just the area of the dashed rectangle. To ﬁnd ∆A let is follow

isotherm. Since T = const,

∆A =

c

a

PdV ,

which is just the area below the isotherm a−b−c. Subtracting areas we reconstruct the Maxwell

rule.

Problem 3.7: For a van der Waals gas, plot the isotherms in the in the

¯

P−

¯

V plane (

¯

P and

¯

V

are the reduced pressure and volume) for reduced temperatures

¯

T = 0.5,

¯

T = 1.0, and

¯

T = 1.5.

For

¯

T = 0.5, is

¯

P = 0.1 the equilibrium pressure of the liquid gas coexistence region?

Solution 3.7: We use Maple to plot the curves. The graphs have the form We see that for

¯

T = 0.5 there is no stavle region at all. To illustrate the situation we plot the curves in the

vicinity of

¯

T = 1.

Problem 3.8: Consider a binary mixture composed of two types of particles, A and B. For this

system the fundamental equation fog the Gibbs energy is

G = n

A

µ

A

+n

B

µ

B

, (3.1)

the combined ﬁrst and second laws are

dG =−SdT +V dP+µ

A

dn

A

+µ

B

dn

B

(3.2)

(S is the total entropy and V is the total volume of the system), and the chqemical potentials µ

A

and µ

B

are intensive so that

µ

A

= µ

A

(P, T, x

A

) and µ

B

(P, T, x

B

).

38 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

–4

–2

0

2

4

6

8

10

12

0.6 0.7 0.8 0.9 1 1.1 1.2

v

Figure 3.2: The upper curve corresponds

¯

T = 1.5, the lower one - to

¯

T = 0.5.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

P

1 2 3 4 5

v

Figure 3.3: The upper curve corresponds

¯

T = 1.05, the lower one - to

¯

T = 0.85.

Use these facts to derive the relations

sdT −vdP+

∑

α=A,B

x

α

µ

α

= 0 (3.3)

and

∑

α=A,B

x

α

(dµ

α

+s

α

dT −v

α

dP) = 0, (3.4)

where s =S/n, n =n

A

+n

B

, s

α

= (∂S/∂n

α

)

P,T,n

β,=α

, and v

α

= (∂V/∂n

α

)

P,T,n

β,=α

with α =A, B and

β = A, B.

Solution 3.8: Let us express n

α

as n

α

= nx

α

and divide Eq. (3.1) by n. We get

dg = d

_

∑

α

x

α

µ

α

_

=

∑

α

(x

α

dµ

α

+µ

α

dx

α

) (3.5)

39

The let us divide Eq. (3.2) by n to obtain

dg =−sdT +vdP+

∑

α

µ

α

dx

α

. (3.6)

Equating the right hand sides of these equation we prove Eq. (3.3). Further, the entropy is an

extensive variable. Consequently,is should be a homogeneous function of n

α

.

S =

∑

α

(∂S/∂n

α

)

P,T,n

β,=α

n

α

= n

∑

α

s

α

x

α

.

Substituting these expressions into Eq. (3.3) we prove Eq. (3.4).

Problem 3.9: Consider liquid mixture (l) of particles A and B coexisting in equilibrium with

vapor mixture (g) of particles A and B. Show that the generalization of of the Clausius-Clapeyron

equation for the coexistence curve between the liquid and vapor phases when the mole fraction

of of A in the liquid phase is ﬁxed is given by

_

∂P

∂T

_

x

l

A

=

x

g

A

(s

g

A

−s

l

A

) +x

g

B

(s

g

B

−s

l

B

)

x

g

A

(v

g

A

−v

l

A

) +x

g

B

(v

g

B

−v

l

B

)

, (3.7)

where s

α

= (∂S/∂n

α

)

P,T,n

β,=α

, and v

α

= (∂V/∂n

α

)

P,T,n

β,=α

with α = A, B and β = A, B. [Hint:

Equation (b) of the Problem (3.8) is useful.]

Solution 3.9: Let us apply Eq. (3.4) to the gas phase. We get

∑

α

x

g

α

_

dµ

g

α

+s

g

α

dT −v

g

α

dP

g

_

= 0.

Dividing this equation by dT we obtain

∑

α

x

g

α

_

dµ

g

α

dT

+s

g

α

−v

g

α

dP

g

dT

_

= 0.

Now, let us take into account that at the coexistence curve µ

g

α

= µ

l

α

. Thus

dµ

g

α

dT

=

dµ

l

α

dT

=−s

l

α

+v

l

α

dP

l

dT

.

Since at the equilibrium P

g

= P

l

, we prove Eq. (3.7).

Problem 3.10: A PVT system has a line of continuous phase transitions (a lambda line) sepa-

rating two phases, I and II, of the system. The molar heat capacity c

P

and the thermal expansivity

α

P

are different in the two phases. Compute the slope (dP/dT)

coex

of the λ line in terms of the

temperature T, the molar volume v, ∆c

P

= c

I

P

−c

II

P

, and ∆α

P

= α

I

P

−α

II

P

.

40 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

Solution 3.10: At a continuous phase transition the entropy is continuous, ∆s = 0. At the same

time, s = s(P, T) and along the coexistence curve P is a function of the temperature. For each

phase, we get

d s

dT

=

∂s

∂T

+

∂s

∂P

dP

dT

.

We know that (∂S/∂T)

P

= c

P

/T and from the Maxwell relation

(∂s/∂P)

T

=−(∂v/∂T)

P

.

Since ∆s = 0,

(∆c

P

)/T = (dP/dT)∆(∂v/∂T)

P

= (dP/dT)v(∆α

P

).

The answer is

_

dP

dT

_

coex

=

∆c

P

T ∆α

P

.

Problem 3.11: Water has a latent heat of vaporization, ∆h = 540 cal/gr. One mole of steam is

kept at its condensation point under pressure at T

1

= 373 K. The temperature is then lowered to

T

2

= 336 K. What fraction of the steam condenses into water? (Treat the steam as an ideal gas

and neglect the volume of the water.)

Solution 3.11: Let us use the Clausius-Clapeyron equation for the case of liquid-vapor mixture.

Since v

g

¸v

l

we get

_

dP

dT

_

coex

=

q

Tv

g

.

Here q is the latent heat per mole. Considering vapor as an ideal gas we get

dP

dT

=

qP

RT

2

→ P = P

0

e

−q/RT

.

As a consequence,

P

1

P

2

= exp

_

q(T

1

−T

2

)

RT

1

T

2

_

.

Since the volume is kept constant,

n

g

1

n

g

2

=

P

1

P

2

= exp

_

q(T

1

−T

2

)

RT

1

T

2

_

= exp

_

18∆h(T

1

−T

2

)

RT

1

T

2

_

.

Here we have taken into account that the olecular weight of H

2

O is 18. The relative number of

condensed gas is then

n

1

−n

2

n

1

= 1−exp

_

−

18∆h(T

1

−T

2

)

RT

1

T

2

_

.

41

Problem 3.12: A liquid crystal is composed of molecules which are elongated (and often have

ﬂat segments). I behaves like a liquid because the locations of the center of mass of the molecules

have no long-range order. It behaves like crystal because the orientation of the molecules does

have long range order. The order parameter of the liquid crystal is given by the diatic

S = η[nn−(1/3)I] ,

where n is a unit vector (called the director), which gives the average direction of alignment of

the molecules. The free energy of the liquid crystal can be written as

Φ = Φ

0

+

1

2

AS

i j

S

i j

−

1

3

BS

i j

S

jk

S

ki

+

1

4

CS

i j

S

jk

S

kl

S

kl

, (3.8)

where A = A

0

(T −T

∗

), A

0

, B and C are constants, I is the unit tensor so

ˆ x

i

I ˆ x

j

= δ

i j

, S

i j

= ˆ x

i

S ˆ x

j

,

and the summation is over repeated indices. The quantities ˆ x

i

are the unit vectors

ˆ x

1

= x, ˆ x

2

= y, ˆ x

3

= z.

(a) Perform the summations in the expression for Φ and write Φ in terms of η, A, B, C.

(b) Compute the critical temperature T

c

, at which the transition from isotropic liquid to liquid

crystal takes place.

(c) Compute the difference between the entropies between the isotropic liquid (η = 0) and the

liquid crystal at the critical temperature.

Solution 3.12:

(a) First, for brevity, let us express the matrix S as η( ˆ s −I/3), where

ˆ s = nn =

_

_

n

2

1

n

1

n

2

n

1

n

3

n

2

n

1

n

2

2

n

2

n

3

n

3

n

1

n

3

n

2

n

2

3

_

_

.

The Eq. (3.8) can be expressed as

Φ = Φ

0

+

1

2

Aa

2

η

2

−

1

3

Ba

3

η

3

+

1

4

Ca

4

η

4

,

a

2

= Tr

_

ˆ s −

1

3

I

_

2

, a

3

= Tr

_

ˆ s −

1

3

I

_

3

, a

4

=

_

Tr

_

ˆ s −

1

3

I

_

2

_

2

. (3.9)

It is easy to calculate b

m

= Tr

_

ˆ s −

1

3

I

_

m

. Indeed,

Tr

_

ˆ s −

1

3

I

_

m

=

m

∑

k=0

m!

k!(m−k)!

(−1)

k

3

k

Tr

_

ˆ s

k

I

m−k

_

.

42 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

It is straightforward that for m > 1 and k ,= 0 we get

Tr

_

ˆ s

k

I

m−k

_

= Tr

_

ˆ s

k

I

_

= Tr

_

ˆ s

k

_

= 1, Tr I

m

= Tr I = 3.

Here we have used the properties of trace and of ˆ s operators. Since trace is independent of

the presentation, let us direct the axis 1 along the vector n. Then

ˆ s =

_

_

1 0 0

0 0 0

0 0 0

_

_

→ Tr

_

ˆ s

k

_

= Tr ˆ s = 1.

Consequently,

b

m

=

_

1−

1

3

_

m

+2

_

−1

3

_

m

, a

2

= b

2

=

2

3

, a

3

= b

3

=

2

9

, a

4

= b

2

2

=

4

9

.

Finally, we get

Φ = Φ

0

+

1

3

Aη

2

−

2

27

Bη

3

+

1

9

Cη

4

. (3.10)

(b) At the transition point the dependence Φ(η) has two equal minima. To ﬁnd this point, let

us shift the origin of η by some η

0

in order to kill odd in η−η

0

items. It we put η =η

0

+ξ

we have

Φ−Φ

0

=

1

3

A(ξ+η

0

)

2

−

2

27

B(ξ+η

0

)

3

+

1

9

C(ξ+η

0

)

4

. (3.11)

In order to kill the term proportional to ξ

3

one has to put (see the Maple ﬁle) η

0

= B/6C.

Substituting this value into the expression for the coefﬁcient ξ,

(2/3)Aη

0

−(2/9)Bη

2

0

+(4/9)Cη

3

0

,

and equating the result to 0 we get A = B

2

/27C. Since A = A

0

(T −T

∗

) we have

T

c

= T

∗

+

B

2

27A

0

C

.

(c) Since S =−∂Φ/∂T =−(1/3)A

0

η

2

we get

∆S =−(1/3)A

0

(η

+

−η

−

)(η

+

+η

−

) =−(2/3)A

0

η

0

(ξ

+

−ξ

−

),

where η

±

are the roots of the equation ∂Φ/∂η = 0 while ξ

±

are the roots of the equation

∂Φ/∂ξ = 0. Note that the potential Φ(ξ) is symmetric, thus ξ

+

+ξ

−

= 0. The important

part Φ(ξ) at the critical point can be obtained substituting η

0

and A into Eq. (3.11). It has

the form,

δΦ(ξ) =

C

9

_

ξ

4

−

B

2

18C

2

ξ

2

_

.

43

Thus,

ξ

±

=±

B

6C

, ξ

+

−ξ

−

=

B

3C

.

As a result,

∆S =−

1

27

A

0

B

2

C

2

.

Problem 3.13: The equation of state of a gas is given by the Berthelot equation

_

P+

a

Tv

2

_

(v −b) = RT .

(a) Find values of the critical temperature T

c

, the critical molar volume v

c

, and the critical

pressure, P

c

in terms of a, b, and R.

(b) Does the Berthelot equation satisfy the law of corresponding states.?

• Find the critical exponents β, δ, and γ from the Berthelot equation.

Solution 3.13:

(a) Let us express the pressure as a function of volume,

P =

RT

v −b

−

a

Tv

2

.

For the critical temperature we have

_

∂P

∂V

_

T

= −

RT

c

(v

c

−b)

2

+

2a

T

c

v

3

c

= 0,

_

∂

2

P

∂V

2

_

T

=

2RT

c

(v

c

−b)

3

−

6a

T

c

v

4

c

= 0.

From this we obtain:

v

c

= 3b, T

c

=

_

8a

27Rb

, P

c

=

RT

c

8b

.

(b) We put

P = P

c

¯

P, T = T

c

¯

T , v = v

c

¯ v

to get:

_

¯

P+

3

¯

T ¯ v

2

_

(3¯ v −1) = 8

¯

T .

44 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

(c) First let us simplify the equation of state near the critical point. Putting

¯

P = 1+ p, ¯ v = 1+ν,

¯

T = 1+τ

and expanding in p, ν, and τ up to lowest non-trivial order (see Maple ﬁle), we get

p =−

3

2

ν

3

−12ντ +7τ. (3.12)

The derivative (∂p/∂ν)

τ

is

_

∂p

∂ν

_

τ

=−

9

2

ν

2

−12τ. (3.13)

Since this function is even in ν the Maxwell construction

ν

g

ν

l

νdp =

ν

g

ν

l

ν

_

∂p

∂ν

_

τ

dν = 0

requires −ν

l

= ν

g

. Since along the coexistence curve p(ν

g

, τ) = p(−ν

g

, τ) we get

ν

g

=

√

−8τ → β = 1/2.

Substituting this value of ν

g

into Eq. (3.13) we observe that

_

∂p

∂ν

_

τ

= 24τ → γ = 1.

To obtain the critical exponent δ let us introduce ρ = 1/¯ v, the equation of state being

¯

P(ρ,

¯

T) =

8

¯

Tρ

3−ρ

−

3ρ

2

¯

T

.

Expanding near the critical point

¯

P =

¯

T = ρ = 1 we get

¯

P = 1+(3/2)(ρ−1)

3

→ δ = 3.

3.1 Mini-tests

3.1.1 A

Properties of Van der Waals (VdW) liquid

(i) Find the critical temperature T

c

at which the Van der Waals isotherm has an inﬂection

point. Determine the pressure P

c

and volume, V

c

, for a system of N particles at T = T

c

.

3.1. MINI-TESTS 45

(ii) Express the VdW equation in units of T

c

, P

c

, and V

c

. Show that it has the form

_

P

/

+

3

V

/2

_

(3V

/

−1) = 8T

/

(3.14)

where P

/

≡P/P

c

, V

/

≡V/V

c

, and T

/

≡T/T

c

(iii) Analyze the equation of state (3.14) near the critical point. Assume that

P

/

= 1+ p, T

/

= 1+τ V

/

= 1−n

and show that for small p, τ and n the equation of state has the approximate form

p = 4τ +6τn+(3/2)n

3

. (3.15)

(iv) Plot p(n, τ) versus n for τ =±0.05 and discuss the plots.

(v) Using the above equation ﬁnd the stability region. Show this region in the plot.

(vi) Show that the Maxwell relation can be expressed as

n

r

n

l

n(∂p/∂n)

τ

dn = 0 (3.16)

along the equilibrium liquid-gas line. Using this relation and the equation of state ﬁnd n

l

and n

r

.

(vii) Discuss why the stability condition and Maxwell relation lead to different stability criteria?

Illustrate discussion using the plot.

Solution

(i) Writing the VdW equation as

P =

NkT

V −Nb

−

N

2

a

V

2

and requiring

_

∂P

∂V

_

T

c

= 0,

_

∂

2

P

∂V

2

_

T

c

= 0

we obtain

T

c

=

8

27

a

bk

, V

c

= 3Nb, P

c

=

1

27

a

b

2

.

(ii) The result is straightforward in one substitutes P = P

c

P

/

, V =V

c

V

/

and T = T

c

T

/

.

(iii) Straightforward.

(iv) Straightforward.

46 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

(v) The stability region is determined by the equation (∂p/∂n)

τ

= 0. We have

(∂p/∂n)

τ

= 6τ +(9/2)n

2

.

Consequently

n

st

=±2

_

−τ/3.

(vi) Since in the main approximation (∂p/∂n)

τ

is an odd function of n, n

l

= −n

r

. Using

Eq. (3.15) we get (see the Maple ﬁle)

n

r

=

√

−4τ.

3.1.2 B

Curie-Weiss theory of a magnet

The simplest equation for a non-ideal magnet has the form

m = tanh[β(Jm+h)] (3.17)

where β = 1/kT, J is the effective interaction constant, while h is the magnetic ﬁeld measured in

proper units.

(i) Consider the case h = 0 and analyze graphically possible solutions of this equation.

Hint: rewrite equation in terms of an auxiliary dimensional variable ˜ m ≡ βJm. Show that

a spontaneous magnetization appears at T < T

c

= J/k.

(ii) Simplify Eq. (3.17) at h = 0 near the critical temperature and analyze the spontaneous

magnetization as a function of temperature.

Hint: Put T = T

c

(1+τ) and consider solutions for small ˜ m and τ.

(iii) Analyze the magnetization curve m(h) near T

c

. Hint: express Eq. (3.17) in terms of the

variables ˜ m and

˜

h ≡βh ≈h/J and plot

˜

h as a function of ˜ m.

(iv) Plot and analyze the magnetization curves for T/T

c

= 0.6 and T/T

c

= 1.6.

Solution

(i) At h = 0 the equation (3.17) has the form:

(βJ)

−1

˜ m = tanh ˜ m.

From the graph of the function F( ˜ m) ≡(βJ)

−1

˜ m−tanh ˜ m (see the Maple ﬁle) one observes

that only trivial solution ˜ m=0 exists at βJ <11, or at T ≥T

c

=J/k. At T <T

c

there exists

2 non-trivial solutions corresponding spontaneous magnetization. Thus (βJ)

−1

= T/T

c

3.1. MINI-TESTS 47

(ii) The dimensionless function F( ˜ m) ≡(T/T

c

) ˜ m−tanh ˜ m, near T

c

acquires the form F( ˜ m) =

τ ˜ m+ ˜ m

3

/3. Thus the solutions have the from

˜ m = 0, ±

√

−3τ.

Since near T

c

the ratio T

c

/T can be substituted by 1, the same result is true for m.

(iii) In general, Eq. (3.17) can be rewritten as

(T/T

c

) ˜ m = tanh( ˜ m+

˜

h) → ˜ m+

˜

h = artanh[(1+τ) ˜ m] .

Consequently,

˜

h = artanh[(1+τ) ˜ m] − ˜ m

To plot the magnetization curve one has to come back to the initial variables.

48 CHAPTER 3. THE THERMODYNAMICS OF PHASE TRANSITIONS

Chapter 4

Elementary Probability Theory and Limit

Theorems

Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Problem 4.1: A bus has 9 seats facing forward and 8 seats facing backward. In how many

ways can 7 passengers be seated if 2 refuse to ride facing forward and 3 refuse to ride facing

backward?

Solution 4.1: Three people refuse ride facing backward, and they should be deﬁnitely placed

to 9 seats facing forward. The number of ways to do that is 9 8 7. In a similar way, 2 people can

occupy 8 seats facing backward in 8 7 ways. Now 17−5 = 12 seats left. They can be occupied

by 7−3−2 = 2 nice people in 12 11 ways. As a result, we get

9 8 78 712 11 = 3725568

ways.

Problem 4.2: Find the number of ways in which 8 persons can be assigned to 2 rooms (A and

B) if each room must have at least 3 persons in it.

Solution 4.2: The number of persons in the room is between 3 and 5. Let us start with the

situation where room A has 3 persons. The number of ways to do that is 8 7 6. The total

number of ways is then

8 7 6+8 7 6 5+8 7 6 5 4 = 8736.

49

50 CHAPTER 4. ELEMENTARY PROBABILITY THEORY . . .

Problem 4.3: Find the number of permutations of the letters in the word MONOTONOUS. In

how many ways are 4 O’s together? In how many ways are (only) 3 O’s together?

Solution 4.3: The total number of the letters is 10, we have 4 O’s, 2 N’s, and other letters do not

repeat. Thus the total number of permutations is 10!/(4!2!) = 75600. To get the number of ways

to have 4 O’s together we have to use model word MONTNUS which has 7 letters with 2 repeated

N’s. Thus the number of ways is 7!/2! = 2520. To have three O’s together we have 6 places to

insert a separate O to the word MONTNUS, thus the number of ways is 62520 = 15120.

Problem 4.4: In how many ways can 5 red balls, 4 blue balls, and 4 white balls be placed in a

row so that the balls at the ends of the row are of the same color?

Solution 4.4: Let us start with 5 red balls and choose 2 of them to be at the ends. The rest

5+4+4−2 =11 balls are in the middle. They have to be permuted, but permutation of the balls

of the color must be excluded. Among them the have 3 rest red balls, 4 blue balls and 4 white

balls. Thus we have 11!/(3!4!4!) = 11550 ways. Now let us do the same with white balls. Now

we have 2 rest white balls, but 5 blue balls to permute. We get 11!/(2!5!4!) = 6930 ways. The

same is true for blue balls. The total number of ways is 138600+283160 = 25410 ways.

Problem 4.5: Three coins are tossed.

(a) Find the probability of getting no heads.

(b) Find the probability of getting at least one head.

(c) Show that the event “heads on the ﬁrst coin” and the events “tails on the last coin” are

independent.

(d) Show that the event “only two coins heads” and the event “three coin heads” are dependent

and mutually exclusive.

Solution 4.5:

(a) The answer is (1/2)

3

= 1/8.

(b) The sum of the probabilities is 1. Since the probability to get no heads is 1/8, the answer

is 1−1/8 = 7/8.

(c) We assume that result of coin tossing is fully random.

(d) The event “3 heads” contradicts the requirement “only 2”.

51

Problem 4.6: Various 6 digit numbers can be formed by permuting the digits 666655. All

arrangements are equivalently likely. Given that the number is even, what is the probability that

two ﬁves are together?

Hint: You must ﬁnd a conditional probability.

Solution 4.6: Since the number must be even, the last number is 6, and there are 4 ways to

choose this number. There are 5 digits left, and the total number of their permutations give 5!.

We have to divide it by 3! of permutations of 6 and 2! permutations between 5. We end at the

number of different numbers is 4 5!/3!2! = 40. If we want to have two ﬁves together, it is the

same as permute 4 units, the number of permutations 4 4!/3! = 16. Thus the probability is

P = 16/40 = 2/5.

Problem 4.7: Fifteen boys go hiking. 5 get lost, 8 get sunburned, and 6 return home without

problems.

(a) What is the probability that a sunburned boy got lost?

(b) What is the probability that a lost boy got sunburned?

Solution 4.7: See. pp. 176-177 of the book [1].

Let us deﬁne the event to get sunburned as A, and the event to get lost - B. The probability to

get sunburned is P(A) = 8/15, while the probability to get lost is P(B) = 5/15 = 1/3. Thus

We know that 15 −6 = 9 boys had problems. Thus, the joint probability to be either sun-

burned or lost is, P(A∪B) = 9/15. Thus the probability to be both sunburned and lost is

P(A∩B) = P(A) +P(B) −P(A∪B) = (8+5−9)/15 = 4/15.

Consequently, conditional probabilities are

P(A[B) = P(A∩B)/P(A) = 1/2 (a), P(B[A) = P(A∩B)/P(B) = 4/5 (b) ,

respectively.

Problem 4.8: A stochastic variable X can have values x = 1 and x = 3. A stochastic variable

Y can have values y = 2 and y = 4. Denote the joint probability density

P

X,Y

(x, y) =

∑

i=1,3

∑

j=2,4

p

i, j

δ(x −i)δ(y − j).

Compute the covariance of X and Y for the following 2 cases:

(a) p

1,2

= p

1,4

= p

3,2

= p

3,4

= 1/4;

(b) p

1,2

= p

3,4

= 0 and p

1,4

= p

3,2

= 1/2.

For each case decide if X and Y are independent.

52 CHAPTER 4. ELEMENTARY PROBABILITY THEORY . . .

Solution 4.8:

(a) First let us calculate

¸XY) = (1 2+1 4+3 2+3 4)/4 = 6.

At the same time ¸X) = 2 and ¸Y) = 3. Since ¸XY) =¸X) ¸Y) the variables are indepen-

dent.

(b) We have,

¸XY) = (1 4+3 2)/2 = 5.

It is not equal to ¸X) ¸Y), so the variables are dependent.

Problem 4.9: The stochastic variables X and Y are independent and Gaussian distributed with

ﬁrst moment ¸x) =¸y) = 0 and standard deviation σ

X

=σ

Y

= 1. Find the characteristic function

for the random variable Z = X

2

+Y

2

, and compute the moments ¸z), ¸z

2

) and ¸z

3

). Find the ﬁrst

3 cumulants.

Solution 4.9: First let us ﬁnd the distribution function for Z. Let us do the problem in two

ways.

Simple way:

Let us transform the problem to the cylindrical coordinates, x, y →ρ, φ. We have

P(x)P(y) =

1

2π

e

−(x

2

+y

2

)

=

1

2π

e

−ρ

2

, dxdy = ρdρdφ, .

Thus,

P

Z

(z) =

1

2π

ρdρdφe

−ρ

2

/2

δ(z −ρ

2

) =

1

2

e

−z/2

.

General way:

P

Z

(z) =

dxP

X

(x)

dyP

Y

(y)δ(x

2

+y

2

−z)

=

√

z

−

√

z

dx

2

√

z −x

2

P

X

(x)

∑

±

P

Y

_

±

_

z −x

2

_

=

1

π

√

z

0

dx

√

z −x

2

exp

_

−

x

2

2

−

z −x

2

2

_

=

1

2

e

−z/2

.

Now, the characteristic function is

f

Z

(k) =¸e

ikz

) =

1

2

∞

0

dze

(ik−1/2)z

=

1

1−2ik

.

53

Then,

¸z

n

) = (−i)

n

d

n

f

Z

(k)

dk

n

¸

¸

¸

¸

k→0

.

Performing calculations (see the Maple ﬁle), we obtain:

¸z) = 2, ¸z

2

) = 8, ¸z

3

) = 48.

Thus,

C

1

(Z) = ¸z) = 2, , C

2

(Z) =¸z

2

) −¸z)

2

= 4,

C

3

(Z) = ¸z

3

) −3¸z)¸z

2

) +2¸z)

3

= 48−3 2 8+3 2

3

= 24.

Problem 4.10: A die is loaded so that even numbers occur 3 times as often as odd numbers.

(a) If the die is thrown N = 12 times, what is the probability that odd numbers occur 3 times?

If is is thrown N = 120 times, what is the probability that odd numbers occur 30 times?

Use the binomial distribution.

(b) Compute the same quantities as in part (a), but use the Gaussian distribution

Note: For parts (a) and (b) compute your answers to four places.

(c) Compare answers (a) and (b). Plot the binomial and Gaussian distributions for the case

N = 12.

Solution 4.10:

(a) For the random process, the parial probability for an odd number p = 1/4 and for an even

number is q = 1−p = 3/4. The probability to have k odd numbers, according to binimial

distribution is

P

b

N

(k) =

N!

k!(N−k)!

p

k

q

N−k

.

Consequently (see Maple ﬁle),

P

b

12

(3) =

12!

3!9!

3

9

4

12

= 0.2881.

In a similar way, P

b

120

(30) = 0.08385.

(b) The number of odd numbers is ¸k) = N/4 and the dispersion is σ = Np q = 3N/16. Thus

P

G

N

(k) =

1

_

3Nπ/8

exp

_

−

(k −N/4)

2

3N/8

_

.

As a result,

P

G

12

(3) = 0.2660, P

G

120

(30) = 0.0841.

(c) Plots are shown in the Maple ﬁle.

54 CHAPTER 4. ELEMENTARY PROBABILITY THEORY . . .

Problem 4.11: A book with 700 misprints contains 1400 pages.

(a) What is the probability that one page contains no mistakes?

(b) What is the probability that one pages contains 2 mistakes?

Solution 4.11: n mistakes can be distributed between g pages by

w

g

(n) =

(g+n−1)!

(g−1)!n!

.

If one has a page without mistakes, it means that n mistakes are distributed between g−1 pages.

The probabiliy of that event is

P(0) =

(g+n−2)!

(g−2)!n!

n!(g−1)!

(g+n−1)!

=

g−1

g+n−1

.

In our case g = 1400 and n = 700, thus P(0) = 1399/2099 ≈0.67. In a similar way, if we have

two misprints at a page, then n −2 misprints are distributed between g −1 pages. Thus, the

number of ways to do that is w

g−1

(n−2). Consequently,

P(2) =

w

g−1

(n−2)

w

g

(n)

=

(g+n−4)!

(g−2)!(n−2)!

n!(g−1)!

(g+n−1)!

=

(g−1)n(n−1)

(g+n−1)(g+n−2)(g+n−3)

.

We get P(2) ≈0.07.

Problem 4.12: There old batteries and a resistor, R, are used to construct a circuit. Each battery

has a probability p to generate voltage V = v

0

and has a probability 1 − p to generate voltage

V = 0 Neglect internal resistance in the batteries.

Find the average power, ¸V

2

)/R, dissipated in the resistor if

(a) the batteries are connected in the series and

(b) the batteries are connected in parallel.

In cases (a) and (b), what would be the average power dissipated in all batteries were

certain to generate voltage v

0

.

(c) How you realize the conditions and results in this problem in the laboratory?

55

Solution 4.12: Let us start with series connection and ﬁnd the probability to ﬁnd a given value

of the voltage. We have 4 possible values of the voltage. In units v

0

they are 0, 1, 2, and 3. We

have,

P(0) = (1−p)

3

, P(1) = 3p(1−p)

2

, P(2) = 3p

2

(1−p), P(3) = p

3

.

Consequently,

¸W) =¸V

2

)/R =∗(v

2

0

/R)

_

3p(1−p)

2

+12p

2

(1−p) +9p

3

¸

= (3v

2

0

/R) p(1+2p).

Now let us turn to parallel connection. Let us for a while introduce conductance G

i

= 1/R

i

of

the batteries, as well as conductance G = 1/R of the resistor. The voltage at the resistor is given

by the expression

V =

∑

i=1,2,3

E

i

G

i

G+∑

i=1,2,3

G

i

.

Here E

i

are electro-motive forces of the batteries. Now the crucial point is to assume something

about the resistances of the batteries. Let us for simplicity assume that they are equal, and

G

i

¸ G. Actually, it is a very bad assumption for old batteries. Under this assumption, V =

(E

1

+E

2

+E

3

)/3. The the result is just 9 times less that that for series connection.

To check this kind of phenomena in the lab one has to ﬁnd many old batteries at a junk-yard,

ﬁrst check distribution of their voltages and then make circuits.

The problem is badly formulated.

Problem 4.13: Consider a random walk in one dimension. In a single step the probability of

displacement between x and x +dx is given by

P(x)dx =

1

√

2πσ

2

exp

_

−

(x −a)

2

2σ

2

_

dx. (4.1)

After N steps the displacement of the walker is S = X

1

+X

3

+X

N

where X

i

is the displacement

after ith step. After N steps,

(a) what is the probability density for the displacement, S, of the walker, and

(b) what is his standard deviation?

Solution 4.13:

(a) By deﬁnition,

P

S

(s) =

dx

1

P(x

1

)

dx

2

P(x

2

). . .

dx

N

P(x

N

)δ(s −x

1

−x

2

. . . −x

N

) .

To calculate this integral we expand δ-function into Fourier integral

δ(x) =

∞

−∞

dk

2π

e

ikx

.

56 CHAPTER 4. ELEMENTARY PROBABILITY THEORY . . .

In this way we get

P

S

(s) =

∞

−∞

dk

2π

e

iks

dx

1

e

−ikx

1

P(x

1

)

dx

1

e

−ikx

2

P(x

2

). . .

dx

N

e

−ikx

N

P(x

N

)

=

∞

−∞

dk

2π

e

iks

_

∞

−∞

dxe

−ikx

P(x)

_

N

=

∞

−∞

dk

2π

e

iks

[P(k)]

N

. (4.2)

Here

P(k) =

∞

−∞

dxe

−ikx

P(x)

is the Fourier component of the single-step probability. From Eq. (4.1) we obtain

P(k) = exp

_

−ika−k

2

σ

2

/2

_

→ [P(k)]

N

= exp

_

−iNka−k

2

Nσ

2

/2

_

.

Now,

P

S

(s) =

∞

−∞

dk

2π

exp

_

ik(s −Na) −k

2

Nσ

2

/2

¸

=

1

√

2πNσ

2

exp

_

−

(s −Na)

2

2Nσ

2

_

.

(b) We have (see Maple ﬁle),

¸S) =

dssP

S

(s) = Na, ¸S

2

) −¸S)

2

= Nσ

2

.

Problem 4.14: Consider a random walk in one dimension for which the walker at each step

is equally likely to take a step with displacement anywhere in the interval d −a ≤ x ≤ d +a,

where a < d. Each step is independent of others. After N steps the displacement of the walker is

S = X

1

+X

3

+ +X

N

where X

i

is the displacement after ith step. After N steps,

(a) what is the average displacement, ¸S), of the walker, and

(b) what is his standard deviation?

Solution 4.14: Starting from the deﬁnition, we have

¸S) =

sP

S

(s)ds =

sds

dx

1

P(x

1

)

dx

N

P(x

N

)δ(s −x

1

−. . . −x

N

)

=

dx

1

P(x

1

)

dx

N

P(x

N

)(x

1

+x

2

+. . . x

N

) = N¸X) = Nd .

In a similar way,

¸S

2

) =

dx

1

P(x

1

)

dx

N

P(x

N

)(x

1

+x

2

+. . . x

N

)

2

.

57

We know that

_

∑

i

x

i

_

2

=

∑

i

x

2

i

+

∑

i,=j

x

i

x

j

.

The number of pairs with non-equal i and j is N(N−1). After averaging the obtain

¸S

2

) = N¸X

2

) +N(N−1)¸X)

2

.

To calculate ¸X

2

) let us specify the normalized probability as

P(x) =

1

2a

_

1, d −a ≤x ≤d +a

0, otherwise

.

Then,

¸X

2

) =

1

2a

d+a

d−a

x

2

dx = d

2

+a

2

/3.

Summarizing we get (see Maple ﬁle)

¸S

2

) −¸S)

2

= N(d

2

+a

2

/3) +N(N−1)d

2

−(Nd)

2

= Na

2

/3.

Problem 4.15: Consider a random walk in one dimension. In a single step the probability of

displacement between x and x +dx is given by

P(x)dx =

1

π

a

x

2

+a

2

dx. (4.3)

Find the probability density for the displacement of the walker after N steps. Does it satisfy the

Central Limit Theorem? Should it?

Solution 4.15: Using the method of the Problem 4.13 we get

P(k) = e

−a[k[

→ P

S

(s) = 2

∞

0

dk

2π

e

−Nka

cosks =

1

π

Na

2

Na

2

+s

2

.

It does not satisfy the Central Limit Theorem. It should not because ¸X

2

) is divergent.

58 CHAPTER 4. ELEMENTARY PROBABILITY THEORY . . .

Chapter 5

Stochastic Dynamics and Brownian Motion

Quick access: 1 2 3 4 5 6 7 8 9 10 11 Test 1

Problem 5.1: Urn A has initially 1 red and 1 white marble, and urn B initially has 1 white

and 3 red marbles. The marbles are repeatedly interchanged. In each step of the process one

marble is selected from each urn at random and the two marbles selected are interchanged. Let

the stochastic variable Y denote “conﬁguration of the units”. There conﬁgurations are possible.

They are shown in Fig. 5.1

3

1

2

Figure 5.1: To the problem 5.1

We denote these 3 realizations as y(n) where n = 1, 2 and 3.

(a) Compute the transition matrix Q and the conditional probability matrix P(s

0

[s).

(b) Compute the probability vector, ¸P(s)[, at time s, given the initial condition stated above.

What is the probability that there are 2 red marbles in urn A after 2 steps? After many

steps?

(c) Assume that the realization, y(n), equals to n

2

. Compute the moment, ¸y(s)), and the

autocorrelation function, ¸y(0)y(s)), for the same initial conditions as in part (b).

59

60 CHAPTER 5. STOCHASTIC DYNAMICS ...

Solution 5.1: The conﬁgurations are shown in Fig. 5.1. By inspection of the ﬁgure we get,

Q

11

= 0, Q

12

= 1, Q

13

= 0,

Q

21

=

1

2

1

4

=

1

8

, Q

22

=

1

2

1

4

+

1

2

3

4

=

1

2

, Q

23

=

1

2

3

4

=

3

8

,

Q

31

= 0, Q

32

=

1

2

, Q

33

=

1

2

.

Thus the transition matrix Q is given by the expression

Q =

_

_

0 1 0

1/8 1/2 3/8

0 1/2 1/2

_

_

,

while the initial condition corresponds to the probability vector

¸p(0)[ = (0, 1, 0).

Below we proceed as in exercise 5.1 of the book [1].

The following procedure is clear form the Maple printout. First we ﬁnd eigenvalues of Q

which are equal to 1, ∓1/4. The corresponding “right” eigenvectors are:

[1, 1, ¦[1, 1, 1]¦], [

−1

4

, 1, ¦

_

6,

−3

2

, 1

_

¦], [

1

4

, 1, ¦[4, 1, −2]¦]

Thus we denote λ

1

= 1, λ

2

=−1/4, λ

3

= 1/4, and

[1) = (1, 1, 1), [2) = (6, −3/2, 1), [3) = (1, 2, −3).

The “left” eigenvectors are just the right eigenvalues of the transposed matrix, Q

T

. We have for

them the Maple result

[

1

4

, 1, ¦[1, 2, −3]¦], [

−1

4

, 1, ¦[1, −2, 1]¦], [1, 1, ¦[1, 8, 6]¦]

Thus,

¸1[ = (1, 8, 6), ¸2[ = (1, −2, 1), ¸3[ = (1, −2, 1).

It is easy to prove that

¸i[k) = n

i

δ

ik

, where n

1

= 15, n

2

= 10, n

3

= 12.

Now we construct matrices P

i

=[i)¸i[ to get

P

1

=

_

_

1 8 6

1 8 6

1 8 6

_

_

, P

2

=

_

_

6 −12 6

−3/2 3 −3/2

1 −2 1

_

_

, P

3

=

_

_

4 8 −12

1 2 −3

−2 −4 6

_

_

.

61

Now, P(s

0

[s) = P(s −s

0

) with

P(s) =

∑

i

λ

s

i

n

i

P

i

=

_

¸

¸

¸

¸

¸

¸

_

1

15

+

3

5

(

−1

4

)

s

+

1

3

(

1

4

)

s

8

15

−

6

5

(

−1

4

)

s

+

2

3

(

1

4

)

s

2

5

+

3

5

(

−1

4

)

s

−(

1

4

)

s

1

15

−

3

20

(

−1

4

)

s

+

1

12

(

1

4

)

s

8

15

+

3

10

(

−1

4

)

s

+

1

6

(

1

4

)

s

2

5

−

3

20

(

−1

4

)

s

−

1

4

(

1

4

)

s

1

15

+

1

10

(

−1

4

)

s

−

1

6

(

1

4

)

s

8

15

−

1

5

(

−1

4

)

s

−

1

3

(

1

4

)

s

2

5

+

1

10

(

−1

4

)

s

+

1

2

(

1

4

)

s

_

¸

¸

¸

¸

¸

¸

_

As it should be, P(0) = I. In this way we solve the part (a).

Now we come to the point (b). Straightforward calculation gives

¸P(s)[ = ¸p(0)[P(s)

=

_

1

15

−

3

20

(

−1

4

)

s

+

1

12

(

1

4

)

s

,

8

15

+

3

10

(

−1

4

)

s

+

1

6

(

1

4

)

s

,

2

5

−

3

20

(

−1

4

)

s

−

1

4

(

1

4

)

s

_

.

We are asked to ﬁnd the probability of the conﬁguration 3. After 2 steps it is 3/8, and at s →∞

we get 2/5.

The part (c) is calculated in a straightforward way. We obtain

¸y(s)) =

29

5

−

3

10

(

−1

4

)

s

−

3

2

(

1

4

)

s

,

¸y(0)y(s)) =

79

50

(−1)

(1+s)

4

(−s)

−

33

4

16

(−s)

+

81

100

(−1)

(1+2s)

16

(−s)

+

11

2

(−1)

(1+s)

16

(−s)

−

3

2

4

(−s)

+

841

25

Problem 5.2: Three boys, A,B, and C, stand in a circle and play catch (B stands to the right of

A). Before throwing the ball, each boy ﬂips a coin to decide whether to throw to the boy on his

right or left. If “heads” comes up, the boy throws to his right. If “tails” come up, the boy throws

to his right.

The coin of boy A is “fair” (50% heads and 50% tails), the coin of boy B has head of both

sides, and the coin of boy C is weighted (75% heads and 25% tails).

(a) Compute the transition matrix, its eigenvalues, and its left and right eigenvalues.

(b) If the ball is thrown in regular intervals, approximately what fraction of time each boy have

the ball (assuming they throw the ball many times)?

(c) If boy A has the ball to begin with, what is the chance it has it after 2 throws? What is the

chance he will have in after s throws.

62 CHAPTER 5. STOCHASTIC DYNAMICS ...

Solution 5.2: The problem is similar to the previous one. Let us map the notations

A →1, B →2, C →3,

and note that “right” means counterclockwise. Then

Q

12

= Q

13

= 1/2, Q

23

= 1, Q

21

= 0, Q

31

= 3/4, Q

32

= 1/4.

At the same time O

11

= Q

22

= Q

33

= 0. As a result, the transition matrix has the form

Q =

_

_

0 1/2 1/2

0 0 1

3/4 1/4 0

_

_

.

The following solution is similar to the previous problem (see the Maple printout). Note that the

eigenvalues and eigenvectors are complex, but conjugated. It means that the probability is real,

but oscillates in time).

The answers are: after many throws the chance for the boy A to hold the ball is 6/19; starting

from the given initial conditions his chance to hold the ball after 2 throws is 3/8.

Problem 5.3: A trained mouse lives in the house shown in Fig. 5.2.

C (3)

B (2)

A (1)

Figure 5.2:

A bell rings at regular intervals (short compared with the mouses lifetime). Each time it rings,

the mouse changes rooms. When he changes rooms, he is equally likely to pass through any of

the doors of the room he is in.

Let the stochastic variable Y denotes “mouse in a particular room”. There 3 realizations of

Y:

1. “Mouse in room A”

2. “Mouse in room B”

3. “Mouse in room C”

63

which we denote as y(1), y(2), and y(3), respectively.

(a) compute the transition matrix, Q, and the conditional probability matrix, P(s

0

[s).

(b) . Compute the probability vector, ¸P(s)[, at time s, given that the mouse starts in room C.

Approximately, what fraction of his life does he spend in each room?

(c) Assume that the realization, y(n) equals to n. Compute the moment, ¸y(s)), and the auto-

correlation function, ¸y(0)y(s)), for the same initial condition as in part (b).

Solution 5.3: The problem is very much similar to the previous ones. We have

Q

12

= Q

13

= 1/2, Q

21

= 1/3, Q

23

= Q

32

= 2/3, Q

31

= 1/3, Q

ii

= 0.

Thus

Q =

_

_

0 1/2 1/2

1/3 0 2/3

1/3 2/3 0

_

_

.

Now we do the same as in previous problems (see Maple printout). The answers are obvious

from the printout.

Problem 5.4: The doors in the mouse’s house shown in Fig. 5.2 so thay get periodically larger

and smaller. This causes the mouse’s transition between the rooms to become time periodic. Let

the stochastic variable Y have the same meaning as in Problem 5.3. The transition matrix is now

given by

Q

11

(s) = Q

22

(s) = Q

33

(s) = 0, Q

12

(s) = cos

2

(πs/2), Q

12

(s) = sin

2

(πs/2),

Q

21

(s) = 1/4+(1/2)sin

2

(πs/2), Q

23

(s) = 1/4+(1/2)cos

2

(πs/2)

Q

31

(s) = (1/2)cos

2

(πs/2), Q

32

(s) = (1/2) +(1/2)sin

2

(πs/2),

(a) If initially the mouse is in room A, what is the probability to ﬁnd it in room A after 2s

room changes?

(b) If initially the mouse is in room B, what is the probability to ﬁnd it in room A after 2s room

changes? In room B?

Solution 5.4: The solution is rather tedious, but obvious from the Maple ﬁle.

64 CHAPTER 5. STOCHASTIC DYNAMICS ...

Problem 5.5: Consider a discrete random walk on a one-dimensional periodic lattice with

2N+1 sites (label the sites from −N to N). Assume that the walker is equally likely to move on

the lattice site to the left or right at each step. Treat this problem as a Markov chain.

(a) Compute the transition matrix, Q, and the conditional probability, P(s

0

[s).

(b) Compute the probability P

1

(n, s) at time s, given the walker starts at site n = 0.

(c) If the lattice has 5 sites (N = 2), compute the probability to ﬁnd the walker on each site

after s = 2 steps and after s = ∞ steps.Assume that the walker starts at site n = 0.

Hint: Let us understand the word periodic as peridically extended, i. e. possessing the property

P[n+(2N+1), t] = P(n, t).

Solution 5.5: The transition probabilities are given by the expressions

w

n−1,n

= w

n,n−1

= w

n+1,n

= w

n,n+1

= 1/2.

The Master equation has the form

dP

n

/ds = w

n−1,n

P

n−1

+w

n+1,n

P

n+1

−P

n

(w

n,n−1

+w

n,n+1

)

= (1/2)(P

n−1

+P

n+1

−2P

n

) .

Let us look for the solution in the form

P

n

(s) ∝ e

λs+iκn

.

Then we get

λ =−[1−cos(κ)] =−2sin

2

(κ/2).

Now we have to ﬁnd κ. If we apply cyclic boundary conditions (assuming that the word periodic

in the problem formulation means that) we have to require: P(n+2N+1) = P(n) at any s. Thus

e

iκ(2N+1)

= 1 → κ(2N+1) = 2πk,

where k is the integer number ranging from −N to N. Thus, the general solution is

P(n, s) =

N

∑

k=−N

P

k

exp

_

−in

2πk

2N+1

−2ssin

2

πk

2N+1

_

.

The quantities P

k

should be found from the initial conditions. At s = 0 we can rewrite the l. h. s.

of the requirement P(n, 0) = δ

n,0

in the form

P(n, 0) =

N

∑

k=−N

P

k

e

−in2πk/(2N+1)

.

65

On the other hand,

δ

n,0

=

1

2N+1

N

∑

k=−N

e

−in2πk/(2N+1)

.

From that we immediately get P

k

= const(k). The proper constant 1/N (s) is the normalization

coefﬁcient found from the condition

N

∑

n=−N

P(n, s) = 1.

Finally

P(n, s) =

1

N (s)

N

∑

k=−N

exp

_

−in

2πk

2N+1

−2ssin

2

πk

2N+1

_

.

Here

N (s) =

N

∑

n=−N

exp

_

−in

2πk

2N+1

−2ssin

2

πk

2N+1

_

= 2N+1.

Now let us ﬁnd the conditional probability P(n, s[m, 0). Now we have to apply the initial condi-

tion P(n, 0) = δ

m,0

, or

P(n, 0) =

N

∑

k=−N

P

k

e

−in2πk/N

= δ

n,m

=

N

∑

k=−N

P

k

e

−i(n−m)2πk/(2N+1)

.

Thus,

P

k

=∝ e

im2πk/(2N+1)

.

Consequently,

P(s −s

0

[0)

mn

= P(n, s[m, 0) =

1

2N+1

N

∑

k=−N

exp

_

−i(n−m)

2πk

2N+1

−2ssin

2

πk

2N+1

_

.

As s →∞ the probability tends to 1/(2N +1). The calculation for N = 2 is straightforward, see

Maple ﬁle. Finally, the matrix Q

mn

= P(1[0)

mn

.

Problem 5.6: At time t, a radioactive sample contains n identical undecayed nuclei, each with

the probability for the unit time, λ, of decaying. The probability of decay during the time t →

t +∆t is λ∆t. Assume that at time t = 0 there are n

0

undecayed nuclei present.

(a) Write down and solve the master equation of the process [ﬁnd P

1

(n, t)].

(b) Compute the the mean number of undecayed nuclei and the variance as a function of time.

(c) What is the half-life of this decay process?

66 CHAPTER 5. STOCHASTIC DYNAMICS ...

Solution 5.6: The master equation has the form

dP

dt

=−λnP(n, t) +λ(n+1)P(n+1, t).

This is a linear master equation, and we solve it by the method of generating functional, p. 262

of the book [1]. We introduce

F(z, t) =

∞

∑

n=−∞

z

n

P

1

(n, t).

Then we multiply the master equation by z

n

and sum over n. We get

∂F

∂t

=−λ(z −1)

∂F

∂z

. (5.1)

Indeed,

∑

n

z

n

(n+1)P(n+1, t) =

d

dz

∑

n

z

n+1

P(n+1, t) =

∂F

∂z

.

In a similar way

∑

n

z

n

nP(n, t) = z

∑

n

z

n−1

nP(n, t) = z

∂F

∂z

Introducing τ = λt we rewrite Eq. 5.1 in the form

∂F

∂τ

+(z −1)

∂F

∂z

= 0. (5.2)

The characteristics of this equation meet the equation

dτ =

dz

z −1

→ ln(z −1) −τ = const .

Consequently,

F(z, t) = Φ

_

e

−τ+ln(z−1)

_

= Φ

_

(z −1)e

−τ

¸

.

Since P(n, 0) = δ

n,n

0

we get

Φ(z −1) = z

n

0

→ Φ(u) = (u+1)

n

0

.

As result,

F(z, t) =

_

(z −1)e

−τ

+1

¸

n

0

.

Now,

¸n(t)) =

_

∂F

∂z

_

z=1

= n

0

e

−τ

.

In a similar way,

¸n

2

(t)) −¸n(t))

2

= n

0

e

−τ

_

1−e

−τ

_

.

67

To ﬁnd the probabilities let us express the quantity

_

(z −1)e

−τ

+1

¸

n

0

= e

−n

0

τ

[z +(e

τ

−1)]

n

0

= e

−n

0

τ

n

0

∑

n=0

n

0

!

n!(n

0

−n)!

z

n

(e

τ

−1)

n

0

−n

.

Thus,

P

1

(n, τ) =

n

0

!

n!(n

0

−n)!

e

−n

0

τ

_

1−e

−τ

_

n

0

−n

.

Problem 5.7: Consider a random walk on the lattice shown in Fig. 5.3. The transition rates are

4

1

2 3

Figure 5.3: A sketch to the problem 5.7.

w

12

= w

13

= 1/2, w

21

= w

23

= w

24

= 1/3, w

31

= w

32

= w

34

= 1/3,

w

42

= w

43

= 1/2, w

14

= w

41

= 0.

(a) Write the transition matrix, W, and show that that this system obey detailed balance.

(b) Compute the symmetric matrix, V, and ﬁnd its eigenvalues and eigenvectors.

(c) Write P

1

(n, t) for the case P(n, 0) = δ

n,1

. What is P(4, t)?

Solution 5.7: The solution is similar to the Exercise 5.2 in p. 245 of the book [1]. The answers

are obvious from the Maple ﬁle.

Problem 5.8: Consider a random walk on the lattice shown in Fig. 5.4. The site P absorbs the

walker. The transition rates are

w

12

= w

13

= 1/2 w

21

= w

23

= w

2P

= 1/3,

w

31

= w

32

= w

3P

= 1/3, w

P1

= w

P2

= 1/2.

(a) Write the transition matrix, M, and compute its eigenvalues and eigenvectors.

(b) In the walker starts at site n = 1 at time t = 0, compute the mean ﬁrst passage time.

68 CHAPTER 5. STOCHASTIC DYNAMICS ...

P

1

2 3

Figure 5.4: A sketch to the problem 5.8.

Solution 5.8: The solution is similar to the Exercise 5.4 of the book [1]. The answers are clear

from the Maple ﬁle.

Problem 5.9: Let us consider on RL electric circuit with resistance, R, and inductance, L,

connected in series. Even there is no average electromotive force (EMF) exists across the resistor,

because of the discrete character of electrons in the circuit and their random motion, a ﬂuctuating

EMF, ξ(t), exists whose strength is determined by the temperature, T. This, in turn, induces the

ﬂuctuating current, I(t), in the circuit. The Langevin equation for the current is

L

d I(t)

dt

+RI(t) = ξ(t). (5.3)

If the EMF is delta-correlated,

¸ξ(t

2

)ξ(t

1

)) = gδ(t

2

−t

1

)

and ¸ξ(t))

ξ

= 0, compute g and the current correlation function ¸¸I(t

2

)I(t

1

))

ξ

)

T

. Assume that

the average magnetic energy L¸I

2

0

)

T

/2 = kT/2 and ¸I

0

)

T

= 0 where I(0) = I

0

.

Solution 5.9: Let us divide Eq. (5.3) by R, denote L/R ≡τ and measure time in units of τ and

the current in units I

1

≡ R

−1

√

g. Namely, deﬁne θ = t/τ, and j(t) = I(t)/I

1

. Then the equation

for the current has the form

d j/dθ+ j = η(θ), ¸η(θ

2

)η(θ

1

)) = δ(θ−θ

1

).

Its solution has the form

j(θ) = j

0

e

−θ

+

θ

0

dse

s−θ

η(s).

Multiplying solutions for j(θ

2

) and for j(θ

1

) and averaging over realizations of η we get

¸ j(θ

2

) j(θ

1

))

η

= j

2

0

e

−(θ

2

+θ

1

)

+

θ

2

0

ds

2

θ

1

0

ds

1

δ(s

2

−s

1

)e

s

1

+s

2

−θ

1

−θ

2

=

_

j

2

0

−

1

2

_

e

−θ

1

−θ

2

+

1

2

e

−[θ

2

−θ

1

[

.

69

The ﬁrst item must vanish since in the equilibrium the process must be stationary, i. e. it must

depend only on the difference θ

2

−θ

1

. Thus,

¸ j

2

0

)

T

=

1

2

→ ¸I

2

0

)

T

=

g

2R

2

.

On the other hand,

L¸I

2

0

) = kT → g = 2R

2

kT/L.

Consequently,

¸¸ j(θ

2

) j(θ

1

))

η

)

T

=

1

2

e

−[θ

2

−θ

1

[

, .

In dimensional variables,

¸¸I(t

2

)I(t

1

))

ξ

)

T

= I

2

1

¸¸ j(θ

2

) j(θ

1

))

η

)

T

=

kT

L

e

−[t

2

−t

1

[/τ

.

Problem 5.10: Due to the random motion and discrete nature of electrons, an LRC series

circuit experience a random electromotive force (EMF), ξ(t). This, in turn, induces a random

varying charge, Q(t), on the capacitor plates and a random current, I)t) = dQ(t)/dt, through the

resistor and inductor. The random charge, Q(t), satisﬁes the Langevin equation

L

d

2

Q

dt

2

+R

dQ

dt

+

Q

C

= ξ(t).

Assume that EMF is delta-correlated,

¸ξ(t

2

)ξ(t

1

)) = gδ(t

2

−t

1

)

and ¸ξ(t))

ξ

= 0. Assume that the circuit is at the temperature T and the average magnetic energy

in the inductor and average electric energy in the capacitor satisfy the equipartition theorem,

1

2

L¸I

2

0

)

T

=

1

2C

¸Q

2

0

)

T

,

where Q(0) = Q

0

, I(0) = I

0

. Assume that ¸I

0

)

T

= 0, ¸Q

0

)

T

= 0, ¸Q

0

I

0

)

T

= 0.

(a) Compute the current correlation function ¸¸I(t

2

)I(t

1

))

ξ

)

T

.

(b) Compute the variance of the charge distribution, ¸¸(Q(t) −Q

0

)

2

)

ξ

)

T

.

70 CHAPTER 5. STOCHASTIC DYNAMICS ...

Solution 5.10: The problem is exactly the same as in the Exercise 5.5 of the book [1]. After

mapping

Q →x, I →v, L →m, LC →ω

2

0

, R →γ

we obtain the same equation. Thus,

¸¸I(t +τ)I(t))

ξ

)

T

=

kT

L

e

−Γ[τ[

C([τ[),

where

C(τ) = coshδτ −(Γ/δ)sinhδτ, Γ = R/L, δ =

_

Γ

2

−ω

2

0

.

The variance of charge distribution can be calculated by integrating the formal solution

I(t) = I

0

e

−Γt

−Q

0

ω

2

0

δ

e

−Γt

sinhδt +

1

L

t

0

dt

/

ξ(t

/

)e

−Γ(t−t

/

)

C(t −t

/

)

over time:

Q(t) = Q

0

+I

0

1−e

−Γt

Γ

−Q

0

ω

2

0

2δ

_

1−e

−(Γ−δ)t

Γ−δ

−

1−e

−(Γ+δ)t

Γ+δ

_

+

1

L

t

0

dt

//

t

0

dt

/

ξ(t

/

)e

−Γ(t

//

−t

/

)

C(t

//

−t

/

)

Now we do averages over ξ and T to obtain (linear terms vanish!)

¸¸Q

2

(t))

ξ

)

T

=¸Q

2

0

)

T

_

1−

ω

2

0

2δ

_

1−e

−(Γ−δ)t

Γ−δ

−

1−e

−(Γ+δ)t

Γ+δ

__

2

+¸I

2

0

)

T

_

1−e

−Γt

Γ

_

2

+

g

L

2

t

0

dt

1

t

0

dt

2

t

0

dt

3

t

0

dt

4

e

−Γ(t

1

−t

2

+t

3

−t

4

)

C(t

1

−t

2

)C(t

3

−t

4

)δ(t

2

−t

4

)

Now we substitute ¸Q

2

0

)

T

=CkT, ¸I

2

0

)

T

= kT/L, and g = 4RkT, and then apply δ function. The

last term becomes

4RkT

L

2

t

0

dt

1

t

0

dt

2

t

0

dt

3

e

−Γ(t

1

−2t

2

+t

3

)

C(t

1

−t

2

)C(t

3

−t

2

).

The rest integrations must be done explicitly which results in lengthy expressions.

Problem 5.11: Consider a Brownian particle of mass m in one dimension in the presence of

a constant force f

0

in a ﬂuid with force constant γ and in th4e presence of a delta-correlated

random force ξ(t) such that ¸ξ(t

1

)ξ(t

2

))

ξ

= gδ(t

1

−t

2

) and ¸ξ(t))

ξ

= 0. Assume that the velocity

and displacement of the Brownian particle at t = 0 are v

0

and x

0

, respectively.

(a) Compute the velocity correlation function ¸v(t

1

)v(t

2

))

ξ

.

(b) Compute the displacement variance ¸(x(t) −x

0

)

2

).

71

Solution 5.11: The differential equation for the particle is

d v(t)

dt

+

γ

m

v(t) −

f

0

m

=

1

m

ξ(t),

dx

dt

= v(t).

Its solution is

v(t) = v

0

e

−(γ/m)t

+

f

0

γ

_

1−e

−(γ/m)t

_

+

1

m

t

0

dt

/

e

(γ/m)(t

/

−t)

ξ(t

/

).

The quantity

v

d

(t) =

f

0

γ

_

1−e

−(γ/m)t

_

is just the drift velocity in the force f

0

. Using results of the Sec. 5.E.1 of the book [1] we obtain

¸¸[v(t +τ) −v

d

(t +τ)] [v(t) −v

d

(t)])

ξ

)

T

=

kT

m

e

−(γ/m)[τ[

.

Since ¸v(t))

T

= v

d

(t) we obtain

¸¸v(t +τ)v(t))

ξ

)

T

= v

d

(t +τ)v

d

(t) +

kT

m

e

−(γ/m)[τ[

.

In a similar way, we cam introduce drift displacement

x

d

(t) =

t

0

dt

/

f

0

γ

_

1−e

−(γ/m)t

/

_

=

f

0

γ

_

t −

m

γ

_

1−e

−(γ/m)t

_

_

.

Finally we obtain

¸¸[x(t) −x

0

−x

d

(t)]

2

)

ξ

)

T

=

2kT

γ

_

t −

m

γ

_

1−e

−(γ/m)t

_

_

.

Tests for training

Problem 5.12: Stochastic processes

Consider an electron (spin S = 1/2) bounded to an impiruty center and assume that the tempera-

ture T is much less than the energy difference between the levels of orbital motion. Consequently,

only spin degrees of freedom are excited.

Let the system be placed in external magnetic ﬁeld B, the additional energy in magnetic ﬁeld

is ±βB where ± corresponds to the values ±1/2 of the spin component along the ﬁeld. Here

β =[e[ ¯ h/2mc is the Bohr magneton, c is the velocity of light since the Gaussian system is used.

The magnetic moment of the states with S

z

=±1/2 is ∓β.

(a) Find the stationary probabilities for the states with S

z

=±1/2, P

±

.

- Find average magnetic moment of the system as function of temperature T and magnetic ﬁeld

B.

72 CHAPTER 5. STOCHASTIC DYNAMICS ...

(b) Find mean square ﬂuctuation of magnetic moment, ¸(∆M)

2

) and the ratio ¸(∆M)

2

)/¸M)

2

.

- Show that

¸(∆M

2

)) =

β

2

cosh

2

(βB/T)

. (5.4)

- Discuss limiting cases of large and low temperatures. Explain these results qualitatively.

(c) Write down the Master Equation for the probability P

i

(t) to ﬁnd the system in the state i =±

at time t.

- Using the result for the stationary case ﬁnd the relation between the transition probabilities

W

+−

and W

−+

for the transition from the state “+” to the state “-” and for the reverse transition,

respectively. Express the probabilities W

+−

and W

−+

through the quantity W =W

+−

+W

+−

.

- Derive the Master Equation for the population difference, P(t) ≡w

+

(t) −w

+

(t).

(d) Write down the Master Equation for the conditional probability, P( f , t[i, t

0

), to ﬁnd the system

in the state f at time t under the conditions that at time t

0

it was in the state i.

- Specify initial conditions to this equation.

- Show that solution of this equation with proper initial condition is

P( f , t[i, t

0

) = w

f

+(δ

i f

−w

f

)e

−W[t−t

0

[

(5.5)

where w

f

is the stationary probability to ﬁnd the system in the state f .

Hint: It is convenient to express the set of Master Equations for the quantities P( f , t[i, t

0

) at

different i and f as an equation for the 22 matrix

ˆ

P with matrix elements P

f i

(t[t

0

) ≡P( f , t[i, t

0

).

The solution of the matrix Master Equation ∂

ˆ

P/∂T =

ˆ

W

ˆ

P at t ≥t

0

can then be searched in in the

matrix as

ˆ

P ∝ ∑

k

ˆ

C

(k)

e

−λ

k

(t−t

0

)

. Here λ

k

are eigenvalues of the matrix W, while

ˆ

C

(k)

are time-

independent matrices. They are determined from the intial conditions and from the fact that at

t −t

0

→∞ the conditional probabilities tend to the stationary ones.

(e) Using Eq. (5.5) show that the correlation function ¸∆M(t)∆M(0)) has the form

¸∆M(t)∆M(0)) =¸(∆M)

2

)e

−W[t[

(5.6)

where ¸(∆M)

2

) is given by Eq. (5.4) while W =W

+−

+W

−+

.

- Find the ﬂuctuation spectrum.

Solution 5.12: (a) Denote w

±

the probabilities of the states S

z

=±1/2. We have

M =−β(w

+

−w

−

) . (5.7)

According to the Gibbs distribution,

w

+

/w

−

= e

−2βB/T

. (5.8)

73

Since only 2 spin levels are involved,

w

+

+w

−

= 1. (5.9)

Combinig Eqs. (5.8) and (5.9) we obtain,

w

+

=

1

e

2βB/T

+1

, w

−

=

e

2βB/T

e

2βB/T

+1

, w

−

−w

+

= tanh

_

βB

T

_

(5.10)

As a result,

M = βtanh

_

βB

T

_

(5.11)

(b) We get:

¸M

2

) = β

2

(w

+

+w

−

) = β

2

¸M)

2

= β

2

tanh

2

(βB/T). (5.12)

As a result,

¸(∆M

2

)) =

β

2

cosh

2

(βB/T)

,

¸(∆M)

2

)

¸M)

2

=

1

sinh

2

(βB/T)

(5.13)

At low temperatures, T ¸βB, ﬂuctuations are exponentially suppressed because spin is aligned

to magnetic ﬁeld.

(c) The equation reads,

∂w

i

(t)

∂t

=−

∑

s=±

[W

i→s

w

i

(t) −W

s→i

w

s

(t)] (5.14)

- In the stationary situation, using the detailed equilibrium principle,

W

i→s

w

i

(t) =W

s→i

w

s

(t)

we obtain

W

+−

W

−+

=

w

−

w

+

= e

2βB/T

, , W

−+

=

W

1+e

2βB/T

, W

+−

=

W

1+e

−2βB/T

(5.15)

- Coming back to Eq. (5.14) we get

∂w

+

(t)

∂t

= −W

+−

w

+

(t) +W

−+

w

−

(t),

∂w

−

(t)

∂t

= −W

−+

w

−

(t) +W

+−

w

+

(t). (5.16)

Subtracting equations we arrive at the result,

∂P(t)

∂t

=−WP(t) (5.17)

74 CHAPTER 5. STOCHASTIC DYNAMICS ...

(d) We consider a stationary random process, and P( f , t[i, t

0

) is a function of t −t

0

. Let us assume

t

0

= 0, t ≥0 and denote Let us for brevity denote P( f , t[i, 0) ≡P

f i

(t). The Master Equation for

P

f i

(t) has the form

∂P

f i

(t)

∂t

=−

∑

s

[W

f →s

P

f i

(t) −W

s→f

P

si

(t)] . (5.18)

The initial condition to this equation is

P

f i

(t) = δ

i f

at t →0. (5.19)

Let us introduce the matrix

ˆ

P(t) with matrix elements P

i f

(t) = P( f , t[i, 0). Then Eq. (5.18) can

be rewritten as

∂

ˆ

P(t)

∂t

=

ˆ

W

ˆ

P(t), with

ˆ

W =

_

−W

f →s

W

s→f

W

f →s

−W

s→f

_

. (5.20)

It is natural to search the solution of Eq, (5.18) in the form

ˆ

P(t)∑

k

ˆ

C

(k)

e

−λ

k

t

. Then λ

k

are the

eigenvalues of the matrix

ˆ

W while

ˆ

C

k

are the numerical matrices to be determined from intial

conditions at t = 0 and from the fact that P

f i

(t) → w

f

at t → ∞. Here w

f

is the stationary

probability of the state f .

Equating the determinant of

ˆ

W −λ

ˆ

I we get

λ

1

= 0, λ

2

=−W

f →s

+W

s→f

≡−W .

Here

ˆ

I is the unit matrix Then, in general,

P

f i

(t) =C

(1)

f i

+C

(2)

f i

e

−Wt

where

ˆ

C

(k)

are time-independent matrices. From the initial conditions (5.19),

C

(1)

f i

+C

(2)

f i

= δ

f i

.

On the other hand, at t →∞ the conditions probability tends to the stationary one, w

f

. Conse-

quently,

C

(1)

f i

= w

f

.

Finally, because of the time reversibility, P

f i

(t) depends only on [t[. As result,

P

f i

(t) = w

f

+(δ

f i

−w

f

)e

−W[t[

This expression coincides with Eq. (5.5).

(e) By deﬁnition,

¸∆M(t)∆M(0)) =¸M(t)M(0)) −¸M)

2

=

∑

i f

M

i

M

f

[P( f , t[i, 0) −w

f

]w

i

.

75

Using Eq. (5.5), one can rewrite the above equation as

¸∆M(t)∆M(0)) = e

−W[t[

∑

i f

M

i

M

f

(δ

i f

−w

f

)w

i

=¸(∆M)

2

)e

−W[t[

This expresssion coincides with Eq. (5.6).

The ﬂuctuation spectrum is

(∆M)

ω

=

∞

∞

dt e

−ωt

¸∆M(t)∆M(0)) =¸(∆M)

2

)

∞

∞

dt e

−ωt−W[t[

.

The result reads as,

(∆M)

ω

= 2¸(∆M)

2

).

W

ω

2

+W

2

76 CHAPTER 5. STOCHASTIC DYNAMICS ...

Chapter 6

The Foundations of Statistical Mechanics

This chapter will be skipped this year

77

78 CHAPTER 6. THE FOUNDATIONS OF STATISTICAL MECHANICS

Chapter 7

Equilibrium Statistical Mechanics

Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Problem 7.1: Compute the structure function for N noninteracting harmonic oscillators, each

with frequency ω and mass m. Assume the system has total energy E. Using this structure

function and the microcanonical ensemble, compute the entropy and the heat capacity of the

system.

Solution 7.1: One can use either classical, or quantum Hamiltonian. Let us do the problem

classically. For brevity let us assume the the oscillators are one-dimensional. Then the Hamilto-

nian is

H =

N

∑

i=1

_

p

2

i

2m

+

mω

2

x

2

i

2

_

= E .

The total number of states is

Ω(E) =

V

dx

1

V

dx

2

. . .

V

dx

N

dp

1

dp

2

. . .

dp

N

Θ

_

E −

N

∑

i=1

_

p

2

i

2m

+

mω

2

x

2

i

2

_

_

.

Here

Θ(x) =

_

1, x > 0

0, x < 0

is the Heaviside unit step function. We have

dΘ(x)/dx = δ(x),

thus the structure function Σ(E) = ∂Ω(E)/∂E is given by the equation

Σ(E) =

V

dx

1

V

dx

2

. . .

V

dx

N

dp

1

dp

2

. . .

dp

N

δ

_

E −

N

∑

i=1

_

p

2

i

2m

+

mω

2

x

2

i

2

_

_

.

79

80 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Let us use the following trick. Denote q

i

= x

i

_

mω

2

/2, i = 1..N, q

i+N

= p

i

_

1/2m and R

2

= E.

Then

Ω =

_

2

ω

_

N

dq

1

. . .

dq

2N

Θ

_

R

2

−

2N

∑

j=1

q

2

j

_

≡

_

2

ω

_

N

˜

Ω(R).

From the dimensionality considerations,

˜

Ω(R) = A

2N

R

2N

.

To ﬁnd the coefﬁcient A

2N

let us consider the integral

∞

−∞

dq

1

. . .

∞

−∞

dq

2N

exp

_

−

2N

∑

j=1

q

2

j

_

=

_

∞

−∞

dqe

−q

2

_

2N

= π

N

.

On the other hand, the same integral is

∞

0

dR

∂

˜

Ω

∂R

e

−R

2

= 2NA

2N

∞

0

dRR

2N−1

e

−R

2

= NA

2N

Γ(N)

Thus

A

2N

=

π

N

NΓ(N)

.

Finally,

Ω(E) =

1

NΓ(N)

_

2π

E

ω

_

N

, Σ(E) =

1

E Γ(N)

_

2π

E

ω

_

N

.

Now, the entropy is

S/k = ln

Ω(E)

N!h

N

= ln

_

E

¯ hω

_

N

1

(N!)

2

= Nln(E/¯ hω) −2NlnN+2N.

Here we used the Stirling’s formula. Now,

1

T

=

∂S

∂E

=

Nk

E

→E = NkT, C = kN.

This is the classical answer. The quantum calculation is demonstrated in the Exercise 7.1 of the

book [1].

Problem 7.2: A system consists of N noninteracting, distinguishable two-level atoms. Each

atom can exist in one of two states, E

0

= 0, and E

1

=ε. The number of atoms in energy level E

1

is n

1

. The internal energy of the system is U = n

0

E

0

+n

1

E

1

.

(a) Compute the entropy of the system as a function of internal energy.

(b) Compute the heat capacity of a ﬁxed number of atoms, N.

81

Solution 7.2: Since E

0

= 0 the energy U cam be realized in U/ε ≡n

1

particles are in the state

1, and (N−U/ε) = N−n

1

particles are in the state 0. The number of ways to do that is

W =

N!

n

1

!(N−n

1

)!

Consequently,

S

k

= ln

N!

(N−n

1

)!

≈NlnN−N−n

1

lnn

1

+n

1

−(N−n

1

)ln(N−n

1

) +(N−n

1

).

Finally

S

k

≈NlnN−

U

ε

ln

U

ε

−

_

N−

U

ε

_

ln

_

N−

U

ε

_

.

To ﬁnd temperature we do usual procedure:

1

T

=

∂S

∂U

≈

k

ε

ln

Nε −U

U

.

As result,

U =

Nε

e

ε/kT

+1

, C = Nk

_

ε

kT

_

2

e

ε/kT

_

e

ε/kT

+1

_

2

.

Problem 7.3: A lattice contains N normal lattice sites and N interstitial lattice sites. N identical

atoms sit on the lattice, M on the interstitial sites and N −M on the normal sites (N ¸M ¸1.

If an atom occupies a normal sites, its energy E = 0. If an atom occupies an interstitial site, its

energy is E = ε. Compute the internal energy and heat capacity as a function of temperature for

this lattice.

Solution 7.3: The internal energy is U = Mε, so we have to express M as a function of temper-

ature. The number of ways in which one can place M atoms at N interstitial sites is

W

i

=

N!

M!(N−M)!

.

The rest N−M atoms can be placed at N normal states in the same number of ways. Thus

W =

_

N!

M!(N−M)!

_

2

→

S

k

= 2ln

N!

M!(N−M)!

.

Using the Stiling’s formula we get

S/k ≈ 2[NlnN−N−MlnM+M−(N−M)ln(N−M) +(N−M)]

≈ 2[NlnN−MlnM−(N−M)ln(N−M)] ,

T

−1

= (2k/ε)ln(N−M)/M, M =U/ε.

Thus,

U =

Nε

e

ε/2kT

+1

≈Nεe

−ε/kT

, C ≈2kN(ε/2kT)

2

e

−ε/2kT

.

82 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Problem 7.4: Consider a lattice with N spin-1 atoms with magnetic moment µ. Each atom can

be in one of 3 spin states:, S

z

= −1, 0, +1. Let n

−1

, n

0

and n

1

denote the respective number of

atoms in each of those spin states. Find the total entropy and the conﬁguration which maximizes

the total entropy. What is the maximum entropy? (Assume that no magnetic ﬁeld is present,

so all atoms have the same energy. Also assume that atoms on different lattice sites cannot be

exchanged, so they are distinguishable).

Solution 7.4: We have

W =

N!

n

−1

!n

1

!(N−n

−1

−n

1

)!

,

lnW ≈ NlnN−n

−1

lnn

−1

−n

1

lnn

1

−(N−n

−1

−n

1

)ln(N−n

−1

−n

1

).

Differentiating this function by n

−1

and n

1

and equating the derivatives to 0 we get:

n

1

= (N−n

1

−n

−1

) = n

−1

→ n

1

= n

0

= n

−1

= N/3

The maximum entropy is

S

max

= kNln3.

Problem 7.5: A system has 3 distinguishable molecules at rest, each with a quantized magnetic

moment which can have z-components +µ/2 or −µ/2. Find an expression for the distribution

function, f

i

(i denotes the conﬁguration), which maximizes entropy subject to the conditions

∑

i

f

i

= 1,

∑

i

M

i,z

f

i

= γµ,

where M

i,z

is the magnetic moment of the system in the i-th conﬁguration. For the case γ = 1/2

compute the entropy and compute f

i

.

Solution 7.5: Let us measure magnetic moment in units µ. We have 4 conﬁgurations with

magnetic moments in parentheses

↑↑↑ (3/2), ↑↑↓ (1/2), ↑↓↓ (−1/2), ↓↓↓ (−3/2).

The conﬁgurations with magnetic moments ±1/2 are triply degenerated. Denoting the conﬁgu-

rations as 1,2, 3 and 4, we construct the functional

˜

S

k

=

∑

i

g

i

f

i

(α+βM

i,z

−ln f

i

).

From this expression we get

α+βM

i,z

−ln f

i

−1 = 0 → f

i

= exp(1−α−βM

i,z

) .

83

Here g

i

is the degeneracy factor: g

1,4

= 1, g

2,3

= 3. We have 2 equations for α and β:

∑

i

g

i

exp(1−α−βM

i,z

) = 1,

∑

i

g

i

M

i,z

exp(1−α−βM

i,z

) = γ.

or

2e

1−α

_

cosh

3β

2

+3cosh

β

2

_

= 1

2e

1−α

_

3

2

sinh

3β

2

+

3

2

sinh

β

2

_

= γ.

For γ = 1/2 that yields:

cosh

3β

2

+3cosh

β

2

= 3sinh

3β

2

+3sinh

β

2

Denoting x = e

β/2

we can rewrite this equation as

x

6

−3x

2

−4 = 0,

the meaningful solution being

x =

√

2, or β = ln2.

Consequently,

α = 1+ln

27

√

2

4

.

As a result, we have

f

i

=N e

M

i,z

ln2

, N =

2

3/2

27

.

The entropy is

S

k

=−

∑

i

g

i

f

i

ln f

i

=−

∑

i

g

i

f

i

(lnN +M

i,z

ln2) =−lnN −

1

2

ln2 = ln

27

4

.

Problem 7.6: A ﬂuid in equilibrium is contained in the insulated box of volume V. The ﬂuid is

divided (conceptually) into m cells. Compute the variance of the enthalpy ﬂuctuations, ¸(∆H

i

)

2

)

in the i-th cell. (For simplicity assume the ﬂuctuations occur at ﬁxed particle number, N

i

).

Hint: Use P and S as independent variables.

84 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Solution 7.6: We have

∆H = T ∆S+V ∆P.

Thus

¸(∆H)

2

) = T

2

¸(∆S)

2

) +V

2

¸(∆P)

2

).

From the general theory the ﬂuctuation probability is proportional to

w ∝ e

W

min

/kT

, W

min

= ∆U −T ∆S+P∆V (7.1)

is the minimum work needed to carry out reversibly the given change in the thermodynamic

quantities of small part considered (relative to which the remainder of the body acts as a medium).

Expanding ∆U in the series, we obtain

W

min

=

1

2

_

∂

2

U

∂S

2

(∆S)

2

+2

∂U

∂S∂V

∆S∆V +

∂

2

U

∂V

2

(∆V)

2

_

=

1

2

_

∆S∆

_

∂U

∂S

_

V

+∆V ∆

_

∂U

∂V

_

S

_

=

1

2

(∆S∆T −∆P∆V) .

Thus

w ∝ exp

_

∆P∆V −∆S∆T

2kT

_

. (7.2)

One can use this general formula and choose different couples as independent variables. In our

case the proper pair is SP. We get

∆V =

_

∂V

∂P

_

S

∆P+

_

∂V

∂S

_

P

∆S,

∆T =

_

∂T

∂P

_

S

∆P+

_

∂T

∂S

_

P

∆S

=

_

∂T

∂P

_

S

∆P+

T

C

P

∆S.

The Maxwell relation following from dH = T dS+V dP is

_

∂V

∂S

_

P

=

_

∂T

∂P

_

S

.

Thus

∆V =

_

∂V

∂P

_

S

∆P+

_

∂T

∂P

_

S

.

Substituting the above relations into Eq. (7.2) we obtain

w ∝ exp

_

1

2kT

_

∂V

∂P

_

S

(∆P)

2

−

1

2C

P

(∆S)

2

_

.

85

As a result,

¸∆S∆P) = 0, ¸(∆S)

2

) = kC

P

, ¸(∆P)

2

) =−kT (∂P/∂V)

S

.

Using these relations we get

¸(∆H)

2

) = kT

2

C

P

−V

2

kT (∂P/∂V)

S

.

Problem 7.7: A ﬂuid in equilibrium is contained in the insulated box of volume V. The ﬂuid

is divided (conceptually) into m cells. Compute the variance of the internal energy ﬂuctuations,

¸(∆U

i

)

2

) in the i-th cell. (For simplicity assume the ﬂuctuations occur at ﬁxed particle number,

N

i

). What happens to the internal energy ﬂuctuations near the critical point?

Solution 7.7: We have,

∆U =

_

∂U

∂V

_

T

∆V +

_

∂U

∂T

_

V

∆T =

_

T

_

∂P

∂T

_

V

−P

_

∆V +C

V

∆T .

Now we can use Eq. (7.2) and choose V, T as independent variables. We have:

∆S =

_

∂S

∂T

_

V

∆T +

_

∂S

∂V

_

T

∆T =

C

V

T

∆T +

_

∂P

∂T

_

V

∆T ,

∆P =

_

∂P

∂T

_

V

∆T +

_

∂P

∂V

_

T

∆V .

The we get

w ∝ exp

_

−

C

V

2kT

(∆T)

2

+

1

2kT

_

∂P

∂V

_

T

(∆V)

2

_

,

¸∆T ∆V = 0) ¸(∆T)

2

) = kT

2

/C

V

, ¸(∆V)

2

) =−kT (∂V/∂P)

T

.

Finally,

¸(∆U)

2

) =−kT

_

T

_

∂P

∂T

_

V

−P

_

2

_

∂V

∂P

_

T

+kT

2

C

V

.

Near the critical point ¸(∆U)

2

) →∞.

Problem 7.8: What is the partition function for a van der Waals gas with N particles? Note

that the result is phenomenological and might involve some guessing. It is useful to compare it

to the partition function for an ideal gas. Remember that the particles are indistinguishable , so

when using a partition function one must insert a counting factor. Use this partition function to

compute the internal energy, U(N, T,V), the pressure, P(N, T,V), and the entropy, S(N, T,V).

86 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Solution 7.8: The partition function is (β ≡(kT)

−1

),

Z

N

=

1

N!h

3N

dr

1

. . . dr

N

dp

1

. . . dp

N

e

−βH(¦r,p¦)

.

First we can integrate out the momenta,

Z

(k)

N

=

1

N!h

3N

dp

1

. . . dp

N

e

−β∑

i

p

2

i

/2m

=

1

N!λ

3N

T

,

where λ

T

= h/

√

2πmkT. Since the partition function for the ideal gas is Z

(0)

N

=V

N

Z

(k)

N

Then we

are left with

Z

N

Z

(0)

N

=

1

V

N

dr

1

. . . dr

N

e

−βV(¦r

i

¦)

= 1+

1

V

N

dr

1

. . . dr

N

_

e

−βV(¦r

i

¦)

−1

_

≡1+z

N

.

Let us try to consider rareﬁed gas where only pair interaction is important. Then we can choose

N(N−1)/2 pair to get

z

n

≈

N

2

2V

2

dr

1

dr

2

_

e

−βV(r

1

−r

2

)

−1

_

.

Introducing the virial coefﬁcient

B(T) =

1

2

dr

_

1−e

−βV(r)

_

,

where r = r

1

−r

2

, we obtain

z

N

=−N

2

B(T)/V .

We get

A = A

0

−kT ln(1+z

n

) ≈A

0

−kTz

N

= A

0

+kTN

2

B(T)/V , A

0

≡−kT lnZ

(0)

N

.

In the spirit of the van der Walls model let us assume that the interaction is large at r ≤2r

0

, and

at r > 2r

0

the interaction is small. Then

B(T) = 2π

2r

0

0

r

2

dr

_

1−e

−βV(r)

_

+2π

∞

2r

0

r

2

dr

_

1−e

−βV(r)

_

≈ 16πr

3

0

/3+2πβ

∞

2r

0

r

2

drV(r) ≡b−a/kT ,

where b = 16πr

3

0

/3 and a = π

∞

2r

0

r

2

drV(r). Finally,

z

n

=−

N

2

V

_

b−

a

kT

_

.

For an ideal gas,

A

0

=−NkT ln

eV

N

_

mkT

2π¯ h

2

_

3/2

= N f (T) −NkT ln(eV/N).

87

Thus,

A = N f (T) −NkT ln(e/N) −NkT(lnV −Nb/V) −N

2

a/V .

Since ln(V −Nb) ≈−Nb/V +lnV we get

A ≈A

0

−NkT ln(1−Nb/V) −N

2

a/V .

From this we get

P = −

∂A

∂V

=

NkT

V −b

−

N

2

a

V

, S =−

∂A

∂T

= S

0

+kNln

_

1−

Nb

V

_

,

U =U

0

−

N

2

a

V

=

3

2

kT −

N

2

a

V

.

Problem 7.9: Consider a solid surface to be a two-dimensional lattice with N

s

sites. Assume

that N

a

atoms (N

a

¸N

s

) are adsorbed on the surface, so that each site has either 0 or 1 adsorbed

atom. At adsorbed atom has energy E =−ε, where ε > 0. Assume that atoms on the surface do

not interact with each other.

(a) If the surface is at temperature T, compute chemical potential of the adsorbed atoms as a

function of T, ε, and N

a

/N

s

(use the canonical ensemble).

(b) If the surface is in equilibriumwith an ideal gas of similar atoms at temperature T, compute

the ratio N

a

/N

s

as a function of pressure, P of the gas. Assume the gas has number density

n. (Hint: equate chemical potentials of the adsorbed atoms and the gas).

Solution 7.9:

(a) The canonical partition function is

Z =

1

N

a

!

N

s

∑

i=1

e

−βE

i

=

N

s

!

N

a

!(N

s

−N

a

)!

e

βN

a

ε

.

Here β = 1/kT. Since N

a

¸N

s

,

Z ≈

N

N

a

s

N

a

!

e

βN

a

ε

.

Equating this expression to e

−N

a

βµ

we get

µ =−ε −kT ln(N

s

/N

a

).

88 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

(b) For an ideal gas,

µ = kT ln

_

Pλ

3

T

kT

_

.

Equating chemical potentials, we obtain

−

ε

kT

−ln

N

s

N

a

= ln

_

Pλ

3

T

kT

_

→

N

a

N

s

=

Pλ

3

T

kT

e

ε/kT

.

This expression is valid at high temperatures when

nλ

3

T

¸e

−ε/kT

.

Problem 7.10: Consider a two-dimensional lattice lattice lattice in the x−y plane with sides of

length L

x

and L

y

which contain N atoms(N is very large) coupled by nearest neighbor harmonic

forces.

(a) Compute the Debye frequency for this lattice.

(b) In the limit T →0, what is the heat capacity?

Solution 7.10: We follow the section 7.E of the book [1]. Since N ¸1 the boundary conditions

are not important, and we can apply zero boundary conditions. As in this exercise, we use

continual approximation and replace

∑

k

. . . →

L

x

L

y

(2π)

2

dk

x

dk

y

. . . .

The the density of states is

g(ω) =

∑

k,α

δ(ω−ω

k,α

) =

L

x

L

y

(2π)

2

∑

α

dk δ(ω−ω

k,α

) .

Here α denotes the vibrational mode. In two-dimensional lattice we have two modes, longitudi-

nal, l, and transverse, t. In the Debye approximation it is assumed that the dispersion law ω

k,α

is

linear and isotropic:

ω

k,α

= c

α

k, k

i

=

π

L

i

s

i

, s

i

= 1, 2, . . . .

where c

α

is the sound velocity for each mode. Since in two-dimensional case

dk . . . =

kdkdφ . . . = 2π

kdk . . .

we obtain

g(ω) =

L

x

L

y

2π¯ h

2

_

1

c

2

l

+

1

c

2

t

_

ω.

89

Since the total number of vibrations should be normalized to 2N, where 2 is the number of modes

and N is number of atoms, we should get

ω

D

0

g(ω)dω = 2N.

Consequently

2N =

L

x

L

y

4π¯ h

2

_

1

c

2

l

+

1

c

2

t

_

ω

2

D

and

1

ω

D

=

¸

L

x

L

y

8Nπ¯ h

2

_

1

c

2

l

+

1

c

2

t

_

.

The density of states can be expressed as

g(ω) =

4N

ω

2

D

ω.

As result,

C =

4kN

(¯ hβω

D

)

2

T

D

/T

0

dx

x

3

e

x

(e

x

−1)

2

=

_

2kN, T →∞

24ζ(3)kN(T/T

D

)

2

, T →0.

Here ζ(x) is the Riemann ζ-function, ζ(3) ≈1.20, T

D

= ¯ hω

D

/k.

Problem 7.11: A cubic box (with inﬁnitely hard walls) of volume V =L

3

contains an ideal gas

of N rigid HCl molecules (assume the effective distance between the H atom and the Cl atom

d = 1.3

˚

A).

(a) If L = 1.0 cm, what is the spacing between the translation energy levels?

(b) Write the partition function for this system (include the translational and rotational contri-

butions). At what temperature do rotational degrees of freedom become important.

(c) Write expressions for the Helmholtz free energy, the entropy, and the heat capacity of

the system for temperatures where the rotational degrees of freedom make a signiﬁcant

contributions.

Solution 7.11: Let us put the origin of the reference frame in the cube’s corner. Since at the

walls the wave function should vanish, for each component α of the momentum p we get:

sin(p

α

L/¯ h) = 0 → k

α

= n

α

π/L, k

α

≡ p

α

/¯ h.

Here n

α

is an integer number. Thus the energy levels of translational motion are given by the

expression

ε

n

= ε

0

(n

2

x

+n

2

y

+n

2

z

), ε

t

=

¯ h

2

π

2

2mL

2

.

90 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Thus the spacing between the levels is

ε

n

i

+1

−ε

n

i

= (2n

i

+1)ε

t

.

The rotational degrees of freedom are determined by the moment if inertia,

I =

m

1

m

2

m

1

+m

2

d

2

,

the energy of rotation being

ε

K

= ε

r

K(K+1), K = 0, 1, . . . , ε

r

=

¯ h

2

2I

.

Thus the spacing between the levels is

ε

K+1

−ε

K

= 2(K+1)ε

r

.

It is important that each rotational level is degenerate by the factor (2K+1).

Thus Z = Z

tr

Z

rot

where

Z

tr

=

_

∞

∑

n

x

=1

e

−βε

t

n

2

x

_

3

=

_

V/λ

3

T

, βε

t

¸1,

exp(−βε

t

), βε

t

¸1,

Z

rot

=

∞

∑

K=0

(2K+1)e

−βε

r

K(K+1)

=

_

kT/ε

r

, βε

r

¸1,

1+3e

−2βε

r

, βε

r

¸1.

The following is straightforward.

A = −kT lnZ

tr

−kT lnZ

rot

;

S = −

_

∂A

∂T

_

V

, C = T

_

∂S

∂T

_

V

=−T

_

∂

2

A

∂T

2

_

V

.

The calculations are obvious.

Problem 7.12: An ideal gas is composed of N “red” atoms of mass m , N “blue” atoms of mass

m, and N “green” atoms of mass m. Atoms of the same color are indistinguishable. Atoms of

different color are distinguishable.

(a) Use the canonical ensemble to compute the entropy of this gas.

(b) Compute the entropy of ideal gas of 3N “red” atoms of mass m. Does if differ from that of

the mixture. If so, by how much.

91

Solution 7.12: Since the atoms of different color distinguishable the partition function is just

the product of partition function. For each component

Z

N

=

_

∑

k

e

−βp

2

/2m

_

N

=

1

N!

_

V

λ

3

T

_

N

.

From this expression,

A

N

=−NkT ln

eV

N

_

mkT

2π¯ h

2

_

3/2

= N f (T) −NkT ln(eV/N).

Here

f (T) =

_

mkT

2π¯ h

2

_

3/2

.

Thus, Z = Z

3

N

, and

A = 3N f (T) −3NkT ln(eV/N), S = 3NKln(eV/N) −3N f

/

(T).

In the case of single-color gas of 3N particles we have

S

1

= 3Nkln(eV/3N) −3N f

/

(T) → S−S

1

= 3Nln3.

Problem 7.13: An ideal gas consists of a mixture of “green” and “red” spin-1/2 particles. All

particles have mass m. A magnetic ﬁeld, B, is applied to the system. The “green” particles have

magnetic moment γ

G

, and the “red” particles have magnetic moment γ

R

, where γ

R

<γ

G

. Assume

that temperature is high enough that Fermi statistics can be neglected. The system will be in

equilibrium if chemical potentials the the “red” and the “green” gases are equal. Compute the

ratio N

R

/N

G

, where N

R

is the number of “red” atoms and N

G

is the number of “red” particles.

Use the canonical ensemble (no other ensembles will be accepted).

Solution 7.13: Let us consider one of the gases since they are noninteracting. We have Z =

Z

tr

Z

s

where Z

tr

is the usual transport partition function while

Z

s

=

_

e

−βγB

+e

βγB

_

N

= 2

N

cosh

N

(βγB).

Since

µ =−

kT

N

ln

1

N!

_

V cosh(βγB)

λ

T

_

N

=−kT ln

_

eV cosh(βγB)

Nλ

T

_

.

Equating chemical potentials for “red” and “green” gases we get

cosh(βγ

R

B)

N

R

=

cosh(βγ

G

B)

N

G

→

N

R

N

G

=

cosh(βγ

R

B)

cosh(βγ

G

B)

.

92 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Problem 7.14: Consider a one-dimensional lattice with N lattice sites and assume that i-th

lattice site has spin s

i

=±1. The Hamiltonian describing this lattice is

H =−ε

N

∑

i=1

s

i

s

i+1

.

Assume periodic boundary condition, so s

N+1

= s

1

. Compute the correlation function, ¸s

1

s

2

).

How does it behave at very high temperature and at very low temperature?

Solution 7.14: We have,

Z

N

(T) =

∑

s

1

=±1

. . .

∑

s

N

=±1

exp

_

βε

∑

i

s

i

s

i+1

_

.

As it is recommended in the book [1], we introduce the matrix

¯

P =

_

e

βε

e

−βε

e

−βε

e

βε

_

= e

βε

I +e

−βε

σσσ

1

.

We have

¸s

i

[

¯

P[s

i+1

) = e

βs

i

s

i+1

.

Then

Z

N

(T) =

∑

s

1

=±1

. . .

∑

s

N

=±1

¸s

1

[

¯

P[s

2

)¸s

2

[

¯

P[s

3

). . . ¸s

N

[

¯

P[s

1

) = Tr

¯

P

N

.

The matrix

¯

P has eigenvalues

λ

1

= 2coshβε, λ

2

= 2sinhβε.

Thus

Z

N

(β) = 2

N

(cosh

N

βε +sinh

N

βε).

We have

¸s

i

s

i+1

) =

1

Nε

∂lnZ

N

∂β

=

cosh

N−1

(βε)sinh(βε) +sinh

N−1

(βε)cosh(βε)

cosh

N

(βε) +sinh

N

(βε)

.

At low temperature, βε ¸1, cosh(βε) ≈ sinh(βε) and ¸s

i

s

i+1

) = 1. Thus we have ferromagnet

ordering.

At high temperature,

cosh(βε) →1, sinh(βε) →0 → ¸s

i

s

i+1

) = βε ¸1.

93

Problem 7.15: In the mean ﬁeld approximation to the Ising lattice, the order parameter, ¸s),

satisﬁed the equation

¸s) = tanh

_

¸s)

T

c

T

_

,

where T

c

=νε/2k where ε is the strength of the coupling between lattice sites and ν is the number

of nearest neighbors.

(a) Show that ¸s) is the following temperature dependence

¸s) ≈

_

1−e

−2T

c

/T

, T →0,

_

3(1−T/T

c

), T →T

c

.

(b) Compute the jump in the heat capacity at T = T

c

.

(c) Compute the magnetic susceptibility, χ(T, N)

B=0

, in the neighborhood of T

c

both for T >

T

c

and T < T

c

. What is the critical exponent for both cases?

Solution 7.15:

(a) Since tanhξ ≈1−2e

−2ξ

at large ξ, we get

¸s) ≈1−2e

−2¸s)T

c

/T

→ ¸s) ≈1−2e

−2T

c

/T

.

At small ξ, tanhξ ≈ξ−ξ

3

/3, we obtain

¸s) ≈¸s)

T

c

T

+

¸s)

3

3

_

T

c

T

_

3

.

Thus,

¸s) ≈

_

3(1−T/T

c

).

(b) We will use notations of Sec. 7.F.2 of the book [1]. The the Hamiltonian is

H =−

N

∑

i=1

E(ε, B)s

i

, E(ε, B) =

νε

2

¸s) +µB. (7.3)

Thus U =−NE¸s) =−NE¸s), and

C =−N

_

∂E(ε, B)¸s)

∂T

_

N,B

=−N

_

E +¸s)

∂E

∂¸s)

__

∂¸s)

∂T

_

N,B

.

At T →T

c

−0 and B = 0 we get

¸s) =

_

3(1−T/T

c

) → C[

T

c

−0

=

3

2

Nνε

T

c

.

At T > T

c

the average spin is zero, thus the jump in speciﬁc heat is given by C[

T

c

−0

.

94 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

(c) To get magnetic susceptibility we have to differentiate the magnetic moment N = Nµ¸s)

with respect to B:

χ = Nµ

_

∂¸s)

∂B

_

N,B→0

.

Deﬁning η = (∂¸s)/∂B)

N,B→0

we get the following equation at B →0:

η =

1

cosh

2

(¸s)T

c

/T)

_

βνε

2

η+βµ

_

.

Its solution is

η =

2βµ

2cosh

2

(¸s)T

c

/T) −T

c

/T

.

Finally,

χ =

Nµ

2

νε

T

c

/T

cosh

2

(¸s)T

c

/T) −T

c

/T

.

Above T

c

we have ¸s)=0, and

χ =

Nµ

2

νε

T

c

T −T

c

.

Below T

c

we get

cosh

2

(¸s)T

c

/T) ≈

_

1+

¸s)

2

2

_

T

c

T

_

2

_

2

≈1+3

_

1−

T

T

c

_

.

As a result, below T

c

χ =

Nµ

2

νε

T

c

2(T

c

−T)

.

Thus there is a divergence at T →T

c

, χ ∝ [T −T

c

[

−1

, with critical exponent equal to -1.

Problem 7.16: The density of states of an ideal Bose-Einstein gas is

g(E) =

_

αE

2

, E > 0,

0, E < 0,

where a is a constant. Compute the temperature for Bose-Einstein condensation.

Solution 7.16: We have the equation for β

c

= 1/kT

c

:

N =

∞

E

min

g(E)dE

e

β

c

E

−1

=

α

β

3

c

∞

0

x

2

dx

e

x

−1

=

2αζ(3)

β

3

c

.

Here ζ(x) is the Riemann zeta-function. Finally,

kT

c

=

_

N

2αζ(3)

_

1/3

.

95

Problem 7.17: An ideal Bose-Einstein gas consists of noninteracting bosons with mass m

which have an internal degree of freedom which can be described by assuming, that the bo-

son are two-level atoms. Bosons in the ground state have the energy E

0

= p

2

/2m, while bosons

in the excited state have the energy E

1

= p

2

/2m+∆, where p is the momentum and ∆ is the

excitation energy. Assume that ∆ ¸kT. Compute the Bose-Einstein condensation temperature,

T

c

, for this gas of two-level bosons. Does the existence of internal degree of the internal degree

of freedom raise or lower the condensation temperature?

Solution 7.17: We start with calculation of density of states.

g

0

(E) =

dpδ

_

E −

p

2

2m

_

=

4π

h

3

p

2

dpδ

_

E −

p

2

2m

_

= a

√

E , a ≡

m

3/2

√

2π

2

¯ h

3

.

In a similar way,

g

1

(E) = a

√

E −∆.

The equation for β

c

= 1/kT

c

is:

Nβ

3/2

c

a

=

∞

0

√

xdx

e

x

−1

+

∞

β

c

∆

_

x −β

c

∆dx

e

x

−1

≈

ζ(3/2)

√

π

2

+

√

π

2

e

−β

c

∆

.

We can solve this equation by iterations since ∆ is large. Denoting

β

0

=

_

aζ(3/2)

√

π

2N

_

2/3

we get

β

c

= β

0

_

1+

√

πa

2Nβ

3/2

0

e

−β

0

∆

_

2/3

≈β

0

+

2

√

πa

6N

_

β

0

e

−β

0

∆

.

Thus the internal degree of freedom decreases the condensation temperature.

Problem 7.18: Compute the Clausius-Clapeyron equation for an ideal Bose-Einstein gas and

sketch the coexistence curve. Show that the line of transition points in the P−v plane obeys the

equation

pv

5/3

=

2π¯ h

2

m

g

5/2

(1)

[g

5/2

(1)]

5/3

.

96 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Solution 7.18: Let us ﬁrst recall the notations. The quantity z = e

βµ

is called the fugacity, the

distribution function is written as

¸n

l

) =

1

e

β(ε

l

−µ)

−1

=

z

e

βε

l

−z

.

The functions g

i

(z) are introduced as

g

5/2

(z) = −

4

√

π

∞

0

x

2

dx ln

_

1−z

−x

2

_

=

∞

∑

α=1

z

α

α

5/2

, (7.4)

g

3/2

(z) = z

d

dz

g

5/2

(z) =

∞

∑

α=1

z

α

α

3/2

. (7.5)

Now, the grand canonical potential is

Ω =

V(kT)

5/2

m

3/2

√

2π

2

¯ h

3

∞

0

dy

√

y ln

_

1−ze

−y

_

=−V

kT

λ

3

T

g

5/2

(z).

Thus the equation of state is

P =

kT

λ

3

T

g

5/2

(z).

In equilibrium with the condensate at T = T

c

,

P

c

=

kT

c

λ

3

T

g

5/2

(1).

The critical particle density in the condensation point is

¸n

c

) =

g

3/2

(1)

λ

3

T

.

Denoting λ

T

= a/T

1/2

where a =

_

2π¯ h

2

/mk we get at the critical point:

N

V

=

g

3/2

(1)T

3/2

c

a

3

→ T

c

=

_

Na

3

Vg

3/2

(1)

_

2/3

.

Thus,

P

c

=

kT

c

g

5/2

(1)

λ

3

T

= ka

2

g

5/2

(1)

[vg

3/2

(1)]

5/3

=

2π¯ h

2

m

g

5/2

(1)

[vg

3/2

(1)]

5/3

.

Problem 7.19: Show that the pressure, P, of an ideal Bose-Einstein gas can be written in the

form P = αu, where “u” is the internal energy per unit volume and a is a constant.

(a) what is u?

(b) What is α?

97

Solution 7.19: Let us do integration by parts in Eq. 7.5 to get

g

5/2

(z) = −

2

√

π

∞

0

dy

√

y ln

_

1−ze

−y

_

=

2

√

π

2

3

∞

0

dyy

3/2

z

e

y

−z

Calculation the average energy

U =

∞

0

g(ε)εdε

z

e

βε

−z

and recalling that g(ε) ∝

√

ε we immediately see that

PV =

2

3

U → P =

2

3

u.

Thus, α = 2/3.

Problem 7.20: Electrons in a piece of Cu metal can be assumed to behave as an ideal Fermi

gas. Cu metal in the solid state has mass density ρ = 9 gr/cm

3

. Assume that each Cu atom

donates an electron to the Fermi gas. Assume the system is at T = 0 K.

(a) Compute the Fermi energy, ε

F

, of the electron gas.

(b) Compute the Fermi temperature, T

F

= ε

F

/k.

Solution 7.20: Let us start with the calculation of the density of states:

g(ε) =

2V 4π

h

3

p

2

dpδ

_

ε −

p

2

2m

_

=

√

2Vm

3/2

π

2

¯ h

3

√

ε ≡Vg

0

√

ε.

Here g

0

=

√

2m

3/2

/π

2

¯ h

3

, we have taken into account that spin degeneracy for electrons 2. Inte-

gration expression for densoty of states from 0 to ε

F

and equationg to the number of electrons

we obtain

(2/3)g

0

ε

3/2

F

= n

e

→ ε

F

= (3n

e

/2g

0

)

2/3

, .

Here n

e

= N

e

/V = ρ/m

Cu

is the electron density.

98 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Problem 7.21: The density of states of an ideal Fermi-Dirac gas is

f (E) =

_

D, if E > 0,

0, if E < 0,

where D is a constant.

(a) Compute the Fermi energy.

(b) Compute the speciﬁc heat at low temperature.

Solution 7.21: We have,

N = D

∞

0

dE

e

β(E−µ)

+1

=

D

β

∞

−βµ

dx

e

x

+1

=

D

β

ln

_

1+e

βµ

_

.

Thus

e

βµ

= e

βN/D

−1 → µ =

1

β

ln

_

e

βN/D

−1

_

≈

N

D

−kTe

−N/DkT

.

The energy is then

U = D

∞

0

E dE

e

β(E−µ)

+1

.

Let us use the trick called the Sommerfeld expansion. Consider an auxiliary integral,

I =

∞

0

f (E)dE

e

(E−µ)/kT

+1

=

∞

−µ/kT

f (µ+kTz)dE

e

z

+1

= kT

µ/kT

0

f (µ−kTz)dE

e

−z

+1

+kT

∞

0

f (µ+kTz)dE

e

z

+1

=

µ

0

f (E)dE +kT

∞

0

dE

f (µ+kTz) − f (µ−kTz)

e

z

+1

Here we have used the fact that (e

−z

+1)

−1

= 1−(e

z

+1)

−1

and that e

µ/kT

¸1. Expanding the

integrand, we obtain

I ≈

µ

0

f (E)dE +2(kT)

2

f

/

(µ)

∞

0

zdz

e

z

+1

+. . . =

µ

0

f (E)dE +

π

2

6

(kT)

2

f

/

(µ) +. . . .

In our case, f (E) = DE, f

/

(µ) = D, and we get

U

D

≈

µ

2

2

+

π

2

6

(kT)

2

.

The speciﬁc heat is then

C

D

= µ

∂µ

∂T

+

π

2

3

k

2

T .

The ﬁrst item is exponentially small at low temperatures, and

C ≈

π

2

3

Dk

2

T ≈k

π

2

3

kT

µ

.

99

Problem 7.22: Compute the magnetization of an ideal gas of spin-1/2 fermions in the presence

of magnetic ﬁeld. Assume that fermions each have magnetic moment µ

e

. Find the expression for

magnetization in the limit of weak magnetic ﬁeld and T →0 K.

Solution 7.22: We will calculate here only spin (Pauli) susceptibility. Since the spin degree of

freedom leads to additional energy ±µ

e

B we get

Ω(µ) =

1

2

Ω

0

(µ−µ

e

B) +

1

2

Ω

0

(µ+µ

e

B)

where Ω

0

is the grand potential in the absence of magnetic ﬁeld,

Ω

0

(µ) =−

2

3

U =−

2A

3

∞

0

E

3/2

dE

e

(E−µ)/kT

+1

, A =

√

2Vm

3/2

π

2

¯ h

3

.

As a result,

−

Ω(µ)

A

=

1

3

∞

0

E

3/2

dE

e

(E−µ−µ

e

B)/kT

+1

+

1

3

∞

0

E

3/2

dE

e

(E−µ+µ

e

B)/kT

+1

.

In small magnetic ﬁelds or at large temperature,

Ω(µ) ≈Ω

0

(µ) +

(µ

e

B)

2

2

∂

2

Ω

0

(µ)

∂µ

2

.

Consequently,

M =−

_

∂Ω

∂B

_

T,V,µ

= µ

2

e

B

∂

2

Ω

0

(µ)

∂µ

2

.

Since

∂Ω

∂µ

=−N we get

M = µ

2

e

B

_

∂N

∂µ

_

T,V

.

This formula works in the wide region

µ

e

B ¸µ.

In very strong magnetic ﬁelds the analysis is more complicated. First the chemical potential is

given by the expression

N

A

=

1

3

_

(µ+µ

e

B)

3/2

+(µ−µ

e

B)

3/2

_

.

In a similar way,

−

Ω(µ)

A

=

2

15

_

(µ+µ

e

B)

5/2

+(µ−µ

e

B)

5/2

_

.

These equations are written assuming that µ

e

B≤µ, otherwise only one itemis present. Physically

it means that only spins-up-states are occupied. Let us consider the case of very strong magnetic

ﬁeld, µ

e

B ¸µ. Then

Ω(µ) =−

2A

15

(µ+µ

e

B)

5/2

→ M =

Aµ

e

3

(µ+µ

e

B)

3/2

= Nµ

e

.

100 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Problem 7.23: Show that the entropy for an ideal Fermi-Dirac gas (neglecting spin) can be

written in the form

S =−k

∑

l

[¸n

l

)ln¸n

l

) +(1−¸n

l

))ln(1−¸n

l

))] ,

where ¸n

l

) =

_

e

β(ε

l

−µ)

+1

_

−1

.

Solution 7.23: The number of ways to distribute N

l

identical particles among G

l

states with

not more than 1 particle per state is just

W

l

=

G

l

!

N

l

!(G

l

−N

l

)!

.

Using the Stirling’s formula we obtain

S = k

∑

l

ln[G

l

lnG

l

−N

l

lnN

l

−(G

l

−N

l

)ln(G

l

−N

l

)] .

Using the mean occupation number ¸n

l

) = N

l

/G

l

we can rewrite this formula as

S =−k

∑

l

G

l

[¸n

l

)ln¸n

l

) +(1−¸n

l

))ln(1−¸n

l

))] .

If we ignore spin, then G

l

= 1.

Problem 7.24: To the lowest order in the density, ﬁnd the difference in the pressure and the

isothermal compressibility between an ideal boson and ideal fermion gas. Assume that the

fermions and bosons have the same mass and both are spin-less. (Note: You are now consid-

ering fairly high temperature).

Solution 7.24: Assuming that the density of states is given by the expression

g(ε) = g

0

V ε

1/2

we obtain

Ω =−

2

3

g

0

V

∞

0

ε

3/2

dε

e

β(ε−µ)

±1

≈−

2

3

g

0

V

∞

0

ε

3/2

dεe

β(µ−ε)

_

1∓e

β(µ−ε)

_

.

Introducing the fugacity z = e

βµ

we get

P =

2

3

g

0

(kT)

5/2

z

∞

0

ξ

3/2

dξe

−ξ

_

1∓ze

−ξ

_

=

√

π

2

g

0

(kT)

5/2

z

_

1∓z

√

2

8

_

.

101

The normalization condition in this case gives

N = g

0

V(kT)

3/2

z

∞

0

ξ

1/2

dξe

−ξ

_

1∓ze

−ξ

_

=

√

π

2

zg

0

V

_

1∓z

√

2

4

_

.

As result,

z

0

=

2

√

πg

0

(kT)

3/2

N

V

.

We obtain

P ≈

NkT

V

_

1∓z

0

√

2

8

_

_

1∓z

0

√

2

4

_ ≈

NkT

V

_

1±z

0

√

2

8

_

≈

NkT

V

_

1±

√

2

4

√

πg

0

(kT)

3/2

N

V

_

.

The sign “+” hold for the Fermi statistics, the sign “-” holds for the Bose statistics. The com-

pressibility is calculated in a straightforward way since

V ≈

NkT

P

_

1±

√

2

4

√

πg

0

(kT)

5/2

P

_

.

102 CHAPTER 7. EQUILIBRIUM STATISTICAL MECHANICS

Chapter 8

Tests and training

Here we put some exercises for self-testing.

103

104 CHAPTER 8. TESTS AND TRAINING

Problem 8.1:

(a) Discuss the difference between Gibbs and Helmholtz free energy.

(b) Prove the relation

U =−T

2

_

∂

∂T

A

T

_

,

where U is the internal energy.

(c) A body with constant speciﬁc heat C

V

is heated under constant volume from T

1

to T

2

. How

much entropy it gains?

(d) Discuss the heating if the same body is in contact with a thermostat at T

2

. In the last case

the heating is irreversible. Show that the total entropy change is positive.

(e) Two similar bodies with temperatures T

1

and T

2

brought into contact. Find the ﬁnal tem-

perature and the change in entropy.

Solution 8.1:

(a) Mechanical work under isothermic process is given by

dW = dU −dQ = dU −T dS = d(U −TS).

The function of state

F =U −TS

is called Helmholtz free energy. We have

dF = dU −T dS−SdT = (T dS−PdV) −T dS−SdT =−SdT −PdV.

Thus,

S =−

_

∂F

∂T

_

V

, P =−

_

∂F

∂V

_

T

.

Thus, F is the thermodynamic potential with respect to V and T.

The thermodynamic potential with respect to P and T is called Gibbs free energy. We have

G =U −TS−PV →dG =−SdT +V dP.

(b) Substituting

S =−

_

∂F

∂T

_

V

into deﬁnition of F we get the result.

105

(c) We have

S =

T

2

T

1

C

V

T

dT =C

V

ln

_

T

2

T

1

_

.

(d) Use 2nd law of thermodynamics

(e) Energy conservation law yields

C

V

(T

2

−T

B

) =C

V

(T

B

−T

1

)−> T

B

= (T

1

+T

2

)/2.

for the entropy change we have

δS =C

V

ln

T

B

T

1

+C

V

ln

T

B

T

2

= 2C

V

ln

_

T

1

+T

2

2

√

T

1

T

2

_

≥0.

Problem 8.2: Find ﬂuctuation of the energy per particle for the Boltzmann gas of spinless

particles at a given temperature . Assume that ε

p

= p

2

/2m.

Solution 8.2:

¸(δε)

2

) =

¸

(ε

i

−¸ε))

2

_

=¸ε

2

) −¸ε)

2

.

Now

¸ε) =

1

Z

1

∑

i

ε

i

e

−βε

i

=−

1

Z

1

∂Z

1

(β)

∂β

.

¸ε

2

) =

1

Z

1

∑

i

ε

2

i

e

−βε

i

=

1

Z

1

∂

2

Z

1

(β)

∂β

2

.

Thus

¸(δε)

2

) =

1

Z

1

∂

2

Z

1

(β)

∂β

2

−

_

1

Z

1

∂Z

1

(β)

∂β

_

2

=

∂

2

logZ

1

∂β

2

.

Now we have to specify Z

1

for ε

i

= p

2

/2m.

Z

1

=

∑

p

exp(−βε

p

) =

1

(2π¯ h)

3

∞

0

4πp

2

dpexp(−βε

p

).

It is convenient to introduce the density of states as number of states per energy interval.

g(ε) =

∑

p

δ(ε −ε

i

) =

4π

(2π¯ h)

3

∞

0

p

2

dp

∑

i

δ(ε −ε

i

) =

m

3/2

√

2π

2

¯ h

3

√

ε.

106 CHAPTER 8. TESTS AND TRAINING

In this way,

Z

1

=

m

3/2

√

2π

2

¯ h

3

∞

0

dεε

1/2

exp(−βε) =

m

3/2

√

2π

2

¯ h

3

β

3/2

∞

0

dxx

1/2

exp(−x).

As

∞

0

dxx

1/2

exp(−x) = Γ(3/2) = (1/2)

√

π

we get

Z

1

=

a

β

3/2

, a =

_

π

8

m

3/2

π

2

¯ h

3

→ logZ

1

= loga−

3

2

logβ.

Thus

¸(δε)

2

) =

3

2β

2

.

Problem 8.3: Two-level systems with random energy spacings ε are distributed with a constant

probability

p(ε) = P

0

.

Find speciﬁc heat of the system.

Solution 8.3: As we know from the lecture,

n

+

n

−

= e

−βε

, n

+

+n

−

= 1.

Thus

n

±

=

1

exp(±βε) +1

.

The average energy is (we start from the middle of the distance between the levels)

¸ε)

ε

=

ε

2

(n

+

−n

−

) =

ε

2

tanh

βε

2

.

The contribution to speciﬁc heat is

c(ε) =

∂¸ε)

ε

∂T

==−β

2

∂¸ε)

ε

∂β

=

1

4

1

cosh

2

(βε/2)

.

The total speciﬁc heat is

C =

∞

0

p(ε)dεc(ε) = P

0

∞

0

dε

1

4

1

cosh

2

(βε/2)

=

P

0

2β

∞

0

dx

cosh

2

x

=

P

0

2β

.

Note that it is proportional to T.

Appendix A

Additional information

A.1 Thermodynamics

A.1.1 Thermodynamic potentials

Thermodynamic potential Notation Independent variables Differential

Internal energy U S, V, N T dS−PdV +µdN

Heat function (enthalpy) H S, P, N T dS+V dP+µdN

Helmholtz free energy F T, V, N −SdT −PdV +µdN

Gibbs free energy G T, P, N −SdT +V dP+µdN

Landau free energy Ω T, V, µ −SdT −PdV −Ndµ

Table A.1: Thermodynamic potentials (summary)

A.1.2 Variable transformation

Jacobian ∂(u, v)/∂(x, y) is deﬁned as the determinant

∂(u, v)

∂(x, y)

=

¸

¸

¸

¸

∂u/∂x ∂u/∂y

∂v/∂x ∂v/∂y

¸

¸

¸

¸

(A.1)

The following relations can be useful:

∂(v, u)

∂(x, y)

=−

∂(u, v)

∂(x, y)

,

∂(u, y)

∂(x, y)

=

_

∂u

∂x

_

y

,

∂(u, v)

∂(x, y)

=

∂(u, v)

∂(t, s)

∂(t, s)

∂(x, y)

. (A.2)

A.1.3 Derivatives from the equation of state

Important relations arise from the properties of partial differentiations. Consider 3 quantities,

X,Y, Z, related by the equation of state K(X,Y, Z) = const. Now let us consider X,Y as indepen-

107

108 APPENDIX A. ADDITIONAL INFORMATION

dent variables, while Z = Z(X,Y). We have

_

∂Z

∂X

_

Y

dX +

_

∂Z

∂Y

_

X

dY −dZ = 0. (A.3)

If Y, Z are taken as independent variables, then

−dX +

_

∂X

∂Y

_

Z

dY +

_

∂X

∂Z

_

Y

dZ = 0 (A.4)

Now we multiply Eq. A.3 by

_

∂X

∂Y

_

Z

and Eq. A.4 by

_

∂Z

∂Y

_

X

and subtract the results. We obtain

__

∂Z

∂X

_

Y

_

∂X

∂Y

_

Z

+

_

∂Z

∂Y

_

X

_

dX +

_

−

_

∂X

∂Y

_

Z

−

_

∂X

∂Z

_

Y

_

∂Z

∂Y

_

X

_

dZ = 0.

Since dX and dY are independent, this equation is compatible if

_

∂Z

∂X

_

Y

_

∂X

∂Y

_

Z

+

_

∂Z

∂Y

_

X

= 0

_

∂X

∂Y

_

Z

+

_

∂X

∂Z

_

Y

_

∂Z

∂Y

_

X

= 0,

or

_

∂X

∂Y

_

Z

_

∂Y

∂Z

_

X

_

∂Z

∂X

_

Y

=−1, (A.5)

_

∂X

∂Y

_

Z

=

_

∂Y

∂X

_

−1

Z

. (A.6)

General scheme for transformation: Consider any quantity F (X,Y), the differential of which

can be expressed as

dF =

_

∂F

∂X

_

Y

dX +

_

∂F

∂Y

_

X

dX .

Then we divide it by dZ and assume Y = const, (dY = 0). We get

_

∂F

∂Z

_

Y

=

_

∂F

∂X

_

Y

_

∂X

∂Z

_

Y

. (A.7)

Another important relation arises if one divides the expression for dF by xX,

_

∂F

∂X

_

Z

=

_

∂F

∂X

_

Y

+

_

∂F

∂Y

_

X

_

∂Y

∂X

_

Z

. (A.8)

Equations (A.7), (A.8) together with Eqs. (A.5), (A.6) and the Maxwell relations are usually

used for transformations and computation of derivatives from the equation of state.

A.2. MAIN DISTRIBUTIONS 109

A.2 Main distributions

Binomial distribution

W

N

(m) =

N!

m!(N−m)!

p

m

(1−p)

N−m

. (A.9)

Poisson distribution

P(m) =

¯ m

m

m!

e

− ¯ m

. (A.10)

Gaussian distribution

G(x) =

1

√

2πσ

e

−(x−¯ x)/2σ

. (A.11)

Distribution of displacemnet for 1D random walk with the step l

G(x) =

1

2l

_

2πNp(1−p)

e

−[x−(p−q)l]

2

/8Nl

2

p(1−p)

. (A.12)

Gaussian integrals

I

1

(a) =

∞

−∞

dze

−az

2

/2

=

_

2π/a. (A.13)

∞

−∞

z

2

dze

−z

2

/2

= 2

_

dI

1

(a)

da

_

a=1

= I

1

(a)

√

2π. (A.14)

Gaussian distribution for more than one variables Deﬁne entropy as S(x

1

, . . . , x

n

. Then

S−S

0

=−

1

2

∑

i,k

β

ik

x

i

x

k

, β

ik

= β

ki

.

For convenience, let us assume summation over repeated subscripts and rewrite the above equa-

tion

S−S

0

=−

1

2

β

ik

x

i

x

k

. (A.15)

Consequently,

w = Ae

−

1

2

β

ik

x

i

x

k

.

Let us ﬁrst calculate the normalization factor A from

dwdx

1

dx

n

= 1.

To calculate the integral let us introduce the linear transform

x

i

= a

ik

y

k

110 APPENDIX A. ADDITIONAL INFORMATION

to make the quadratic form (refmgs01 diagonal. In order that

β

ik

x

i

x

k

= y

i

y

i

= y

i

y

k

δ

ik

the relation

β

ik

a

il

a

km

= δ

km

(A.16)

should be valid. Denoting determinants of the matrices

ˆ

β and ˆ a as β and a, respectively, we get

the relation βa

2

= 1. The Jacobian of the transformation x

i

→y

i

is just a,

J =

¸

¸

¸

¸

∂(x

1

, . . . , x

n

)

∂(y

1

, . . . , y

n

)

¸

¸

¸

¸

=[ ˆ a[ = a.

Consequently,

Aa

_

dye

−y

2

/2

_

n

=

A(2π)

n/2

_

β

= 1.

Thus

w =

_

β

(2π)

n/2

exp

_

−

1

2

β

ik

x

i

x

k

_

. (A.17)

Let deﬁne generalized forces as

X

i

=−∂S/∂x

i

= β

ik

x

k

, (A.18)

and ﬁrst calculate

¸x

i

X

k

) =

_

β

(2π)

n/2

x

i

β

ik

x

k

e

−β

ik

x

i

x

k

/2

dx

1

dx

n

. (A.19)

The easiest way to calculate this integral is to calculate to assume for a while that ¯ x

i

,= 0. Then

¯ x =

_

β

(2π)

n/2

x

i

e

−β

ik

(x

i

−x

i0

)(x

k

−x

k0

/2

dx

1

dx

n

= x

i0

.

Then we can differentiate both sides with respect to x

k0

and then put x

i0

= x

k0

= 0. The l.h.s. is

just ¸x

i

X

k

) while the r.h.s. is δ

ik

. Thus

¸x

i

X

k

) = δ

ik

, or β

ik

¸x

i

x

k

) = δ

ik

, ¸x

i

x

k

) =

_

ˆ

β

−1

_

ik

. (A.20)

A.3 The Dirac delta-function

The main property is to single-out one particular value x −x

0

of a variable x. It is deﬁned by the

characteristic properties

δ(x −x

0

) =

_

0 for x ,= x

0

∞ for x = x

0

(A.21)

A.4. FOURIER SERIES AND TRANSFORMS 111

but in such way that for any ε > 0

x

0

+ε

x

0

−ε

δ(x −x

0

)dx = 1. (A.22)

Since delta-function has a very (inﬁnitely) sharp peak at x = x

0

and a unit area

B

A

f (x)δ(x −x

0

)dx =

_

f (x

0

) if A < x

0

< B

0 otherwise

. (A.23)

Representations for the Dirac δ-function

Let introduce a positive parameter γ and at ﬁnal stage tend it to zero. Physically it means that γ

is less than all other involved scales. The main representations are the following: rectangular,

δ(x) =

_

γ

−1

for −γ/2 < x < γ/e

0 otherwise

, (A.24)

Lorentzian,

δ(x) =

1

π

γ

x

2

+γ

2

, (A.25)

Gaussian,

δ(x) =

1

γ

√

2π

e

−x

2

/2γ

2

, (A.26)

and integral,

δ(x) =

1

2π

∞

−∞

e

ikx

dk (A.27)

We will also need a representation for the Kronekker’s symbol

δ

n,0

=

1

2π

π

−π

e

−nφ

dφ. (A.28)

A.4 Fourier Series and Transforms

Fourier series

Generally the Fourier series is deﬁned as an expansion of a real function in series as

f (x) =

a

0

2

+

∞

∑

k=1

(a

k

sinkx +b

k

coskx) . (A.29)

The sum of the series is periodic with the period 2π.

Another form is complex Fourier series

f (x) =

∞

∑

k=−∞

c

k

e

ikx

, c

±k

= (a

k

∓b

k

)/2, c

0

= a

0

/2. (A.30)

112 APPENDIX A. ADDITIONAL INFORMATION

Since

2π

0

sinmxsinnxdx =

_

πδ

mn

, m ,= 0,

0, m = 0,

2π

0

cosmxcosnxdx =

_

πδ

mn

, m ,= 0,

2π, m = n = 0,

2π

0

sinmxcosnxdx =

_

0, for integer m, n

we have

_

_

a

k

b

k

c

k

_

_

=

1

π

2π

0

dt f (t)

_

_

coskt

sinkt

1

2

e

−ikt

_

_

, k ,= 0. (A.31)

At k = 0,

a

0

= c

0

=

1

2π

2π

0

dt f (t).

If the function has a discontinuity at x = x

0

then the series gives

f (x

0

) =

1

2

[ f (x

0

+0) + f (x

0

−0)] .

If one is willing to change the interval 2π → 2L than k has to be replaced by πk/L and the

normalization factor is 1/L rather than 1/π.

Advantages of the Fourier series: it can represent discontinuous functions. Let us take an

example of periodic function with the period 2π deﬁned as

F(x) =

_

1, −π/2 < x < π/2

0, −π < x <−π/2, π/2 < x < π,

This function is even, thus

b

k

= 0, a

k

= (2/πk)sinπk/2.

Below we plot the sum of 5 and 50 Fourier harmonics. It is useful to know the Parseval’s identity

1

π

π

−π

[ f (x)]

2

dx =

a

2

0

2

+

∞

∑

k=1

(a

2

k

+b

2

k

).

Fourier transform

We assume L →∞ and replace

∞

∑

k=−∞

(. . .) →L

∞

−∞

dk(. . .).

g(k) =

1

2π

∞

−∞

f (t)e

−ikt

dt . (A.32)

A.4. FOURIER SERIES AND TRANSFORMS 113

0

0.2

0.4

0.6

0.8

1

–1 –0.8 –0.6 –0.4 –0.2 0.2 0.4 0.6 0.8 1

x

Figure A.1:

The completeness condition

f (x) =

1

2π

∞

−∞

dke

ikt

∞

−∞

dx f (x)e

−ikx

. (A.33)

The Dirac delta-function is represented as

δ(x −t) =

1

2π

∞

−∞

dke

ik(t−x)

. (A.34)

We have several useful rules for the Fourier transform. If g(k) is the Fourier transform of f (x)

then

f

/

(x) ≡d f /dx → ikg(k);

f

(n)

(x) ≡d

n

f /dx

n

→ (ik)

n

g(k);

∞

−∞

dy f

1

(y) f

2

(x −y) → g

1

(k) g

2

(k) (convolution theorem) ;

∞

−∞

f

1

(t) f

2

(t)dt →

∞

−∞

g

1

(k) f

2

(−k)dk;

∞

−∞

f

1

(t) f

∗

2

(t)dt →

∞

−∞

g

1

(k) f

∗

2

(k)dk the Parseval’s theorem. (A.35)

These properties are very important for solving differential equations, etc.

114 APPENDIX A. ADDITIONAL INFORMATION

f (x) =

1

√

2π

∞

−∞

F(ω)e

−iωx

dx F(ω) =

1

√

2π

∞

−∞

f (x)e

−iωx

dx

1

a

2

+x

2

, a > 0

_

π

2

e

−a[ω[

a

f (x) =

_

1, [x[ < a

0, [x[ > a

_

2

π

sin(aω)

ω

e

−a

2

x

2

1

a

√

2

e

−ω

2

/4a

Table A.2: Table of Fourier transform pairs

Examples

One-dimensional diffusion: The equation has the form

∂n(x, t)

∂t

=−D

∂

2

n(x, t)

∂x

2

.

Let us assume that we add a particle at time t = 0 at the point x = 0. Then the equation together

with initial condition can be written as

∂n(x, t)

∂t

+D

∂

2

n(x, t)

∂x

2

= δ(x)δ(t).

The Fourier transform of this equation has the form:

(Dk

2

−iω)n(k, ω) = 1 → n(k, ω) =

1

(2π)

2

1

Dk

2

−iω

.

Now we can come back to real space-time representation:

n(x, t) =

∞

−∞

dk

2π

∞

−∞

dω

2π

1

Dk

2

−iω

=

∞

−∞

dk

2π

e

−Dk

2

[t[

=

1

2

_

πD[t[

e

−x

2

/4D[t[

.

Appendix B

Maple Printouts

Quick access: 1, 2.1, 2.2, 2.17, 3.7, 3.12, 3.13, 4.9, 4.10, 4.13, 4.14, 5.1, 5.2, 5.3, 5.4, 5.5, 5.7,

5.8

Test in mathematics

>

F:=(xi,eta)->xi^4+xi^3+eta*xi^2:

>

F1:=(xi,eta)->subs(y=xi,diff(F(y,eta),y)): solve(F1(xi,eta)=0,xi);

0, −

3

8

+

1

8

√

9−32η, −

3

8

−

1

8

√

9−32η

One extremum at eta >9/32, otherwise 3 extrema

>

x1:=-3/8+1/8*sqrt(9-32*eta): x2:=-3/8-1/8*sqrt(9-32*eta):

>

F2:=(xi,eta)->subs(y=xi,diff(F1(y,eta),y)):

>

g0:=eta->simplify(F2(0,eta)): g1:=eta->simplify(F2(x1,eta)):

>

g2:=eta->simplify(F2(x2,eta)):

>

g0(eta); g1(eta); g2(eta);

2η

9

8

−

3

8

√

9−32η−4η

9

8

+

3

8

√

9−32η−4η

>

plot(¦g0(eta),g1(eta),g2(eta)¦,eta=-0.1..9/32);

115

116 APPENDIX B. MAPLE PRINTOUTS

0.5

1

1.5

2

2.5

–0.1 –0.05 0.05 0.1 0.15 0.2 0.25

eta

x0-> maximum at eta <0, otherwise minimum; x1-> minimum at eta>0, maximum at

9/32>eta>0; x2-> minimum at 9/32>eta.

Two speciﬁc values: eta=0 and eta=9/32.

>

plot(F(xi,0), xi=-1.1..0.45); plot(F(xi,9/32), xi=-0.7..0.25);

117

–0.1

–0.05

0.05

0.1

–1 –0.8 –0.6 –0.4 –0.2 0.2 0.4

xi

0.005

0.01

0.015

0.02

0.025

0.03

0.035

–0.6 –0.4 –0.2 0.2

xi

Problem 1.2

>

f1:=x->ln((1-x)/(1+x)): plot(f1(x),x=-1..1);

118 APPENDIX B. MAPLE PRINTOUTS

–6

–4

–2

0

2

4

6

–1 –0.8 –0.6 –0.4 –0.2 0.2 0.4 0.6 0.8 1

x

>

f2:=x->ln(abs(tan(Pi*x))): plot(f2(x),x=0..1,f2=-5..5);

–4

–2

0

2

4

f2

0.2 0.4 0.6 0.8 1

x

119

>

simplify(exp(4*ln(x))-(x^2+1)^2+2*x^2+1);

0

Problem 1.3

>

simplify((sin(x))^(-2)-(tan(x))^(-2)-1,trig);

0

>

assume(alpha>0): assume(beta>0):

>

k:=alpha->sum(exp((I*beta-alpha)*n),n=0..infinity):

>

a1:=alpha->evalc(Re(k(alpha))): a2:=alpha->evalc(Im(k(alpha))):

>

a1(alpha); a2(alpha);

−

e

α˜

(cos(β˜) −e

α˜

)

(cos(β˜) −e

α˜

)

2

+sin(β˜)

2

e

α˜

sin(β˜)

(cos(β˜) −e

α˜

)

2

+sin(β˜)

2

Problem 1.4

>

assume(zeta>0): assume(n,natural):

>

simplify(int(x^n*exp(-zeta*x),x=0..infinity));

ζ˜

(−n˜−1)

Γ(n˜ +1)

>

int((x^2+a^2)^(-1),x); int((x^2-a^2)^(-1),x);

arctan(

x

a

)

a

1

2

ln(x −a)

a

−

1

2

ln(a+x)

a

>

int(exp(-zeta*x^2/2),x=0..infinity);

1

2

√

2

√

π

_

ζ˜

Problem 2.1

(a)

>

uax:=(x,y)->-y/(x^2+y^2): uay:=(x,y)->x/(x^2+y^2):

>

simplify(diff(uax(x,y),y)-diff(uay(x,y),x));

0

The differential is exact

>

ua:=(x,y)->int(uax(xi,y),xi=0..x): ua(x,y);

−arctan(

x

y

)

120 APPENDIX B. MAPLE PRINTOUTS

(b)

>

ubx:=(x,y)->(y-x^2): uby:=(x,y)->(x+y^2):

>

simplify(diff(ubx(x,y),y)-diff(uby(x,y),x));

0

The differential is exact

>

ub:=(x,y)->int(ubx(xi,y),xi=0..x)+int(uby(x,eta),eta=0..y)-int(int(di

>

ff(ubx(xi,eta),eta),eta=0..y),xi=0..x);

ub := (x, y) →

x

0

ubx(ξ, y)dξ+

y

0

uby(x, η)dη−

x

0

y

0

∂

∂η

ubx(ξ, η)dηdξ

>

ub(x,y);

yx −

1

3

x

3

+

1

3

y

3

(c)

>

ucx:=(x,y)->(2*y^2-3*x): ucy:=(x,y)->-4*x*y:

>

simplify(diff(ucx(x,y),y)-diff(ucy(x,y),x));

8y

The differential is not exact

Problem 2.2

General deﬁnitions

>

ua:=(x,y)->int(ux(xi,b),xi=a..x)+int(uy(x,eta),eta=b..y);

ua := (x, y) →

x

a

ux(ξ, b)dξ+

y

b

uy(x, η)dη

>

ub:=(x,y)->int(uy(a,eta),eta=b..y)+int(ux(xi,y),xi=a..x);

ub := (x, y) →

y

b

uy(a, η)dη+

x

a

ux(ξ, y)dξ

>

ux1:=(x,y)->2*x*y+x^2: uy1:=(x,y)->x^2:

>

ua1:=(x,y)->int(ux1(xi,b),xi=a..x)+int(uy1(x,eta),eta=b..y):

>

ub1:=(x,y)->int(uy1(a,eta),eta=b..y)+int(ux1(xi,y),xi=a..x):

>

ua1(x,y); ub1(x,y); simplify(ua1(x,y)-ub1(x,y));

b(x

2

−a

2

) +

1

3

x

3

−

1

3

a

3

+x

2

(y −b)

a

2

(y −b) +y(x

2

−a

2

) +

1

3

x

3

−

1

3

a

3

0

>

ux2:=(x,y)->y*(2-2*y): uy2:=(x,y)->-x^2:

121

>

ua2:=(x,y)->int(ux2(xi,b),xi=a..x)+int(uy2(x,eta),eta=b..y):

>

ub2:=(x,y)->int(uy2(a,eta),eta=b..y)+int(ux2(xi,y),xi=a..x):

>

ua2(x,y); ub2(x,y); simplify(ua2(x,y)-ub2(x,y));

b(2−2b)(x −a) −x

2

(y −b)

−a

2

(y −b) +y(2−2y)(x −a)

2bx −2ba−2b

2

x +2b

2

a−x

2

y +x

2

b+a

2

y −a

2

b−2yx +2ya+2y

2

x −2y

2

a

Why?

>

simplify(diff(ux2(x,y),y)-uy2(x,y),x);

2−4y +x

2

The differential is not exact.

Problem 2.17

>

P:=(T,v)->R*T/(v-b)-alpha/v^2:

>

num:=(diff(P(T,v),T))^2: den:=diff(P(T,v),v);

den :=−

RT

(v −b)

2

+

2α

v

3

>

simplify(-T*num/den);

−

v

3

R

2

T

−RT v

3

+2αv

2

−4αvb+2αb

2

>

factor(2*alpha*v^2-4*alpha*v*b+2*alpha*b^2);

2α(−v +b)

2

Problem 3.7

>

P:=(v,T)->(8*T/(3*v-1))-3/v^2;

P := (v, T) →8

T

3v −1

−

3

v

2

>

plot(¦P(v,1.5),P(v,1),P(v,0.5)

>

¦,v=0.5..1.2,thickness=3,color=black);

122 APPENDIX B. MAPLE PRINTOUTS

–4

–2

0

2

4

6

8

10

12

0.6 0.7 0.8 0.9 1 1.1 1.2

v

For T=0.5 there is no stable region at all. The phse transition can exist only nes T=1.

>

plot(¦P(v,1.05),P(v,1),P(v,0.85)

>

¦,v=0.45..5.3,P=0..2,thickness=3,color=black);

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

P

1 2 3 4 5

v

>

?plot,color

Problem 3.12

>

phi:=xi->(A/3)*(xi+eta)^2-(2*B/27)*(xi+eta)^3+(C/9)*(xi+eta)^4:

>

series(phi(xi),xi=0,5);

(

1

3

Aη

2

−

2

27

Bη

3

+

1

9

Cη

4

) +(

2

3

Aη−

2

9

Bη

2

+

4

9

Cη

3

)ξ+(

1

3

A−

2

9

Bη+

2

3

Cη

2

)ξ

2

+

(−

2

27

B+

4

9

Cη)ξ

3

+

1

9

Cξ

4

>

c3:=eta->-2/27*B+4/9*C*eta: solve(c3(eta)=0,eta);

1

6

B

C

>

c1:=(A,eta)->2/3*A*eta-2/9*B*eta^2+4/9*C*eta^3;

c1 := (A, η) →

2

3

Aη−

2

9

Bη

2

+

4

9

Cη

3

>

solve(c1(A,B/(6*C)),A);

1

27

B

2

C

>

c2:=subs(¦eta=B/(6*C),A=B^2/(27*C)¦,1/3*A-2/9*B*eta+2/3*C*eta^2);

c2 :=−

1

162

B

2

C

>

delta:=xi->(C/9)*xi^4+c2*xi^2:

>

delta1:=xik->subs(y=xi,diff(delta(y),y)): delta1(xi);

4

9

Cξ

3

−

1

81

B

2

ξ

C

>

solve(4/9*C*xi^2-1/81*B^2/C,xi);

−

1

6

B

C

,

1

6

B

C

Problem 3.13

>

P:=(v,T)->8*T/(3*v-1)-3/(T*v^2);

P := (v, T) →8

T

3v −1

−

3

T v

2

>

p1:=(nu,epsilon)=P(1+nu,1+epsilon)-1;

p1 := (ν, ε) = 8

1+ε

2+3ν

−

3

(1+ε)(1+ν)

2

−1

>

aux:=series(8*(1+epsilon)/(2+3*nu)-3/(1+epsilon)/(1+nu)^2-1,nu=0,4):a

>

ux;

(3+4ε −

3

1+ε

) +(−6−6ε +

6

1+ε

)ν+(−9

1

1+ε

+9+9ε)ν

2

+(12

1

1+ε

−

27

2

−

27

2

ε)

ν

3

+O(ν

4

)

>

aux1:=(epsilon)->3+4*epsilon-3/(1+epsilon)+(-6-6*epsilon+6/(1+epsilon

>

))*nu+(-9*1/(1+epsilon)+9+9*epsilon)*nu^2

>

+(12*1/(1+epsilon)-27/2-27/2*epsilon)*nu^3;

aux1 := ε →3+4ε −

3

1+ε

+(−6−6ε +

6

1+ε

)ν+(−9

1

1+ε

+9+9ε)ν

2

+(12

1

1+ε

−

27

2

−

27

2

ε)ν

3

>

aux2:=series(aux1(epsilon),epsilon=0,2);

aux2 :=−

3

2

ν

3

+(7−

51

2

ν

3

+18ν

2

−12ν)ε +O(ε

2

)

Simpliﬁed equation of state

>

p:=(nu,epsilon)->-3/2*nu^3+(7-12*nu)*epsilon;

p := (ν, ε) →−

3

2

ν

3

+(7−12ν)ε

Critical point - epsilon=0;

>

d1:=(nu,epsilon)->subs(y=nu,diff(p(y,epsilon),y)): d1(nu,epsilon);

−

9

2

ν

2

−12ε

The curve is symmetric. That leads to vg=-vl=v

>

solve(p(v,epsilon)= p(-v,epsilon),v);;

0, 2

√

−2ε, −2

√

−2ε

Critical exponent beta=1/2

Along the coexitence curve, vˆ2=-8epsilon. exponent gamma=1

Function of pressure

>

P1:=(rho,T)->8*T*rho/(3-rho)-3*rho^2/T:

Near the critical point

>

P1(1+xi,1);

8

1+ξ

2−ξ

−3(1+ξ)

2

>

series(8*(1+xi)/(2-xi)-3*(1+xi)^2,xi=0);

1+

3

2

ξ

3

+

3

4

ξ

4

+

3

8

ξ

5

+O(ξ

6

)

the exponent delta=3

Problem 4.9

>

assume(z>0): int(1/sqrt(z-x^2),x=0..sqrt(z));

1

2

π

>

f:=k->1/(1-2*I*k): d1:=k->subs(y=k,diff(f(y),y)):

>

d2:=k->subs(y=k,diff(d1(y),y)): d3:=k->subs(y=k,diff(d2(y),y)):

>

d1(0); d2(0); d3(0);

2I

−8

−48I

Problem 4.10

>

P:=(N,k)->(N!/(k!*(N-k)!))*(1/4)^k*(3/4)^(N-k):

>

evalf(P(12,3));

.2581036091

>

evalf(P(120,30));

.08385171464

>

G:=(N,k)->(1/sqrt(3*N*Pi/8))*exp(-(k-N/4)^2*8/(3*N));

G := (N, k) →

e

(−8/3

(k−1/4N)

2

N

)

_

3

8

Nπ

>

evalf(G(12,3));

.2659615201

>

evalf(G(120,30));

.08410441740

>

plot(¦P(12,k),G(12,k)¦,k=0..12);

0

0.05

0.1

0.15

0.2

0.25

2 4 6 8 10 12

k

>

plot(¦P(120,k),G(120,k)¦,k=0..120);

0

0.02

0.04

0.06

0.08

20 40 60 80 100 120

k

Problem 4.13

beta = sigmaˆ2

>

assume(beta>0): assume(a,real): assume(N>0):

>

P:=k->(2*Pi*beta)^(-1/2)*int(exp(-(x-a)^2/(2*beta)-I*k*x),x=-infinity.

>

.infinity): simplify(P(k));

e

(1/2I k(−2a˜+I kβ˜))

>

P1:=s->(2*Pi)^(-1)*int(exp(I*(s-N*a)*k-k^2*N*beta/2),k=-infinity..inf

>

inity): P1(s);

1

2

e

(−1/2

(−s+N˜ a˜)

2

N˜ β˜

)

√

2

√

π

_

N˜ β˜

>

int(s*P1(s),s=-infinity..infinity);

N˜ a˜

>

int(s^2*P1(s),s=-infinity..infinity)-N^2*a^2;

N˜ β˜

Problem 4.14

>

assume(a>0): assume(d>a): assume(N>0):

>

simplify((2*a)^(-1)*int(x^2,x=d-a..d+a));

d˜

2

+

1

3

a˜

2

>

simplify(N*(d^2+1/3*a^2)+N*(N-1)*d^2-N^2*d^2);

1

3

N˜ a˜

2

Problem 5.1

>

with(linalg):

>

Q:=array([[0,1,0],[1/8,1/2,3/8],[0,1/2,1/2]]);

Q :=

_

¸

¸

¸

¸

_

0 1 0

1

8

1

2

3

8

0

1

2

1

2

_

¸

¸

¸

¸

_

>

QT:=transpose(Q):

>

eigenvectors(Q);

[1, 1, ¦[1, 1, 1]¦], [

−1

4

, 1, ¦

_

6,

−3

2

, 1

_

¦], [

1

4

, 1, ¦[4, 1, −2]¦]

>

eigenvectors(QT);

[

1

4

, 1, ¦[1, 2, −3]¦], [

−1

4

, 1, ¦[1, −2, 1]¦], [1, 1, ¦[1, 8, 6]¦]

>

r1:=vector([1, 1, 1]); r2:=vector([6, -3/2, 1]); r3:=vector([4, 1,

>

-2]);

r1 := [1, 1, 1]

r2 :=

_

6,

−3

2

, 1

_

r3 := [4, 1, −2]

>

l1:=vector([1, 8, 6]); l2:=vector([1, -2, 1]); l3:=vector([1, 2,

>

-3]);

l1 := [1, 8, 6]

l2 := [1, −2, 1]

l3 := [1, 2, −3]

>

n1:=15: n2:=10: n3:=12:

>

P1:=array([[l1[1]*r1[1], l1[2]*r1[1],l1[3]*r1[1]], [l1[1]*r1[2],

>

l1[2]*r1[2],l1[3]*r1[2]],[l1[1]*r1[3], l1[2]*r1[3],l1[3]*r1[3]]]);

P1 :=

_

_

1 8 6

1 8 6

1 8 6

_

_

>

P2:=array([[l2[1]*r2[1], l2[2]*r2[1],l2[3]*r2[1]], [l2[1]*r2[2],

>

l2[2]*r2[2],l2[3]*r2[2]],[l2[1]*r2[3], l2[2]*r2[3],l2[3]*r2[3]]]);

P2 :=

_

¸

¸

_

6 −12 6

−3

2

3

−3

2

1 −2 1

_

¸

¸

_

>

P3:=array([[l3[1]*r3[1], l3[2]*r3[1],l3[3]*r3[1]], [l3[1]*r3[2],

>

l3[2]*r3[2],l3[3]*r3[2]],[l3[1]*r3[3], l3[2]*r3[3],l3[3]*r3[3]]]);

P3 :=

_

_

4 8 −12

1 2 −3

−2 −4 6

_

_

>

P:=s->(1/n1)*P1+(1/n2)*(-1/4)^s*P2+(1/n3)*(1/4)^s*P3;

>

Question (a)

P := s →

P1

n1

+

(

−1

4

)

s

P2

n2

+

(

1

4

)

s

P3

n3

>

Pr:=s->evalm(P(s)): Pr(s); Pr(0);

_

¸

¸

¸

¸

¸

¸

_

1

15

+

3

5

(

−1

4

)

s

+

1

3

(

1

4

)

s

8

15

−

6

5

(

−1

4

)

s

+

2

3

(

1

4

)

s

2

5

+

3

5

(

−1

4

)

s

−(

1

4

)

s

1

15

−

3

20

(

−1

4

)

s

+

1

12

(

1

4

)

s

8

15

+

3

10

(

−1

4

)

s

+

1

6

(

1

4

)

s

2

5

−

3

20

(

−1

4

)

s

−

1

4

(

1

4

)

s

1

15

+

1

10

(

−1

4

)

s

−

1

6

(

1

4

)

s

8

15

−

1

5

(

−1

4

)

s

−

1

3

(

1

4

)

s

2

5

+

1

10

(

−1

4

)

s

+

1

2

(

1

4

)

s

_

¸

¸

¸

¸

¸

¸

_

_

_

1 0 0

0 1 0

0 0 1

_

_

>

p0:=vector([0,1,0]);

p0 := [0, 1, 0]

>

PS:=s->evalm(p0&*P(s)): PS(s);

Question (b)

_

1

15

−

3

20

(

−1

4

)

s

+

1

12

(

1

4

)

s

,

8

15

+

3

10

(

−1

4

)

s

+

1

6

(

1

4

)

s

,

2

5

−

3

20

(

−1

4

)

s

−

1

4

(

1

4

)

s

_

>

PS(2); PS(infinity);

>

Answer: after 2 steps 3/8, after many steps 2/5

_

1

16

,

9

16

,

3

8

_

_

1

15

,

8

15

,

2

5

_

>

av:=s->1*PS(s)[1]+4*PS(s)[2]+9*PS(s)[3]: av(s);

>

Moment

29

5

−

3

10

(

−1

4

)

s

−

3

2

(

1

4

)

s

>

av:=sum(m^2*PS(s)[m],m=1..3);

29

5

−

3

10

(

−1

4

)

s

−

3

2

(

1

4

)

s

>

corr:=s->simplify(sum(n^2*sum(m^2*PS(s)[m]*Pr(s)[m,n],

>

m=1..3),n=1..3)): corr(s);

79

50

(−1)

(1+s)

4

(−s)

−

33

4

16

(−s)

+

81

100

(−1)

(1+2s)

16

(−s)

+

11

2

(−1)

(1+s)

16

(−s)

−

3

2

4

(−s)

+

841

25

>

corr(0);

16

>

corr(infinity);

841

25

Problem 5.2

>

with(linalg):

Warning, the protected names norm and trace have been redefined and

unprotected

>

Q:=array([[0,1/2,1/2],[0,0,1],[3/4,1/4,0]]);

Q :=

_

¸

¸

¸

¸

_

0

1

2

1

2

0 0 1

3

4

1

4

0

_

¸

¸

¸

¸

_

>

QT:=transpose(Q):

>

r:=eigenvectors(Q):

>

l:=eigenvectors(QT):

>

r[1][1]; r[2][1]; r[3][1];

1

−

1

2

+

1

4

I

√

2

−

1

2

−

1

4

I

√

2

>

l[1][1]; l[2][1]; l[3][1];

1

−

1

2

+

1

4

I

√

2

−

1

2

−

1

4

I

√

2

>

r[1][3]; r[2][3]; r[3][3];

¦[1, 1, 1]¦

¦

_

−

1

6

−

1

3

I

√

2, 1, −

1

2

+

1

4

I

√

2

_

¦

¦

_

−

1

6

+

1

3

I

√

2, 1, −

1

2

−

1

4

I

√

2

_

¦

>

l[1][3]; l[2][3]; l[3][3];

¦

_

6

5

, 1,

8

5

_

¦

¦

_

1, −

1

3

−

1

3

I

√

2, −

2

3

+

1

3

I

√

2

_

¦

¦

_

1, −

1

3

+

1

3

I

√

2, −

2

3

−

1

3

I

√

2

_

¦

>

r1:=vector([1, 1, 1]): r2:=vector([-1/6-1/3*I*sqrt(2), 1,

>

-1/2+1/4*I*sqrt(2)]): r3:=vector([-1/6+1/3*I*sqrt(2), 1,

>

-1/2-1/4*I*sqrt(2)]): l1:=vector([6/5, 1, 8/5]): l2:=vector([1,

>

-1/3-1/3*I*sqrt(2), -2/3+1/3*I*sqrt(2)]): l3:=vector([1,

>

-1/3+1/3*I*sqrt(2), -2/3-1/3*I*sqrt(2)]):

>

n1:=evalm(l1&*r1): n2:=evalc(evalm(l2&*r2)):

>

n3:=evalc(evalm(l3&*r3)):

Probability matrices.

>

P1:=array([[l1[1]*r1[1], l1[2]*r1[1],l1[3]*r1[1]], [l1[1]*r1[2],

>

l1[2]*r1[2],l1[3]*r1[2]],[l1[1]*r1[3], l1[2]*r1[3],l1[3]*r1[3]]]);

P1 :=

_

¸

¸

¸

¸

¸

¸

_

6

5

1

8

5

6

5

1

8

5

6

5

1

8

5

_

¸

¸

¸

¸

¸

¸

_

>

P2:=array([[l2[1]*r2[1], l2[2]*r2[1],l2[3]*r2[1]], [l2[1]*r2[2],

>

l2[2]*r2[2],l2[3]*r2[2]],[l2[1]*r2[3], l2[2]*r2[3],l2[3]*r2[3]]]);

P2 :=

_

¸

¸

¸

¸

¸

¸

_

−

1

6

−

1

3

I

√

2 (−

1

3

−

1

3

I

√

2)(−

1

6

−

1

3

I

√

2) (−

2

3

+

1

3

I

√

2)(−

1

6

−

1

3

I

√

2)

1 −

1

3

−

1

3

I

√

2 −

2

3

+

1

3

I

√

2

−

1

2

+

1

4

I

√

2 (−

1

3

−

1

3

I

√

2)(−

1

2

+

1

4

I

√

2) (−

2

3

+

1

3

I

√

2)(−

1

2

+

1

4

I

√

2)

_

¸

¸

¸

¸

¸

¸

_

>

P3:=array([[l3[1]*r3[1], l3[2]*r3[1],l3[3]*r3[1]], [l3[1]*r3[2],

>

l3[2]*r3[2],l3[3]*r3[2]],[l3[1]*r3[3], l3[2]*r3[3],l3[3]*r3[3]]]);

P3 :=

_

¸

¸

¸

¸

¸

¸

_

−

1

6

+

1

3

I

√

2 (−

1

3

+

1

3

I

√

2)(−

1

6

+

1

3

I

√

2) (−

2

3

−

1

3

I

√

2)(−

1

6

+

1

3

I

√

2)

1 −

1

3

+

1

3

I

√

2 −

2

3

−

1

3

I

√

2

−

1

2

−

1

4

I

√

2 (−

1

3

+

1

3

I

√

2)(−

1

2

−

1

4

I

√

2) (−

2

3

−

1

3

I

√

2)(−

1

2

−

1

4

I

√

2)

_

¸

¸

¸

¸

¸

¸

_

>

P:=s->(1/n1)*P1+(1/n2)*(r[2][1])^s*P2+(1/n3)*(r[3][1])^s*P3;

>

Question (a)

P := s →

P1

n1

+

r

21

s

P2

n2

+

r

31

s

P3

n3

>

Pr:=s->evalm(P(s)):

>

p0:=vector([1,0,0]);

p0 := [1, 0, 0]

>

PS:=s->evalm(p0&*P(s)): PS(s)[1];

6

19

+

(−

1

2

+

1

4

I

√

2)

s

(−

1

6

−

1

3

I

√

2)

−

1

3

−I

√

2

+

(−

1

2

−

1

4

I

√

2)

s

(−

1

6

+

1

3

I

√

2)

−

1

3

+I

√

2

>

evalc(PS(infinity)[1]); evalc(PS(2)[1]);

6

19

3

8

Problem 5.3

>

with(linalg):

Warning, the protected names norm and trace have been redefined and

unprotected

>

Q:=array([[0,1/2,1/2],[1/3,0,2/3],[1/3,2/3,0]]);

Q :=

_

¸

¸

¸

¸

¸

¸

_

0

1

2

1

2

1

3

0

2

3

1

3

2

3

0

_

¸

¸

¸

¸

¸

¸

_

>

QT:=transpose(Q):

>

r:=eigenvectors(Q):

>

l:=eigenvectors(QT):

>

r[1][1]; r[2][1]; r[3][1];

−1

3

−2

3

1

>

l[1][1]; l[2][1]; l[3][1];

−1

3

1

−2

3

>

r[3][3]; r[1][3]; r[2][3];

¦[1, 1, 1]¦

¦[−3, 1, 1]¦

¦[0, −1, 1]¦

>

l[2][3]; l[1][3]; l[3][3];

¦

_

1,

3

2

,

3

2

_

¦

¦[−2, 1, 1]¦

¦[0, −1, 1]¦

>

r1:=vector([1, 1, 1]): r2:=vector([-3, 1, 1]): r3:=vector([0, -1,

>

1]): l1:=vector([1, 3/2, 3/2]): l2:=vector([-2, 1, 1]): l3:=vector([0,

>

-1, 1]):

>

n1:=evalm(l1&*r1): n2:=evalc(evalm(l2&*r2)):

>

n3:=evalc(evalm(l3&*r3)):

Probability matrices.

>

P1:=array([[l1[1]*r1[1], l1[2]*r1[1],l1[3]*r1[1]], [l1[1]*r1[2],

>

l1[2]*r1[2],l1[3]*r1[2]],[l1[1]*r1[3], l1[2]*r1[3],l1[3]*r1[3]]]);

P1 :=

_

¸

¸

¸

¸

¸

¸

_

1

3

2

3

2

1

3

2

3

2

1

3

2

3

2

_

¸

¸

¸

¸

¸

¸

_

>

P2:=array([[l2[1]*r2[1], l2[2]*r2[1],l2[3]*r2[1]], [l2[1]*r2[2],

>

l2[2]*r2[2],l2[3]*r2[2]],[l2[1]*r2[3], l2[2]*r2[3],l2[3]*r2[3]]]);

P2 :=

_

_

6 −3 −3

−2 1 1

−2 1 1

_

_

>

P3:=array([[l3[1]*r3[1], l3[2]*r3[1],l3[3]*r3[1]], [l3[1]*r3[2],

>

l3[2]*r3[2],l3[3]*r3[2]],[l3[1]*r3[3], l3[2]*r3[3],l3[3]*r3[3]]]);

P3 :=

_

_

0 0 0

0 1 −1

0 −1 1

_

_

>

P:=s->(1/n1)*P1+(1/n2)*(r[3][1])^s*P2+(1/n3)*(r[2][1])^s*P3;

>

Question (a)

P := s →

P1

n1

+

r

31

s

P2

n2

+

r

21

s

P3

n3

>

Pr:=s->evalm(P(s)): Pr(s);

>

Conditional probability

_

¸

¸

¸

¸

_

1 0 0

0

1

2

+

1

2

(

−2

3

)

s

1

2

−

1

2

(

−2

3

)

s

0

1

2

−

1

2

(

−2

3

)

s

1

2

+

1

2

(

−2

3

)

s

_

¸

¸

¸

¸

_

>

p0:=vector([0,0,1]):

>

PS:=s->evalm(p0&*P(s)): PS(s);

Probability vector

_

1

4

−

1

4

(

−2

3

)

s

, −

1

8

+

1

8

(

−2

3

)

s

,

7

8

+

1

8

(

−2

3

)

s

_

>

PS(infinity); PS(0);

_

0,

1

2

,

1

2

_

[0, 0, 1]

>

moment:=1*PS(s)[1]+2*PS(s)[2]+3*PS(s)[3]: moment;

5

2

+

1

2

(

−2

3

)

s

>

corr:=s->simplify(sum(n*sum(m*PS(s)[m]*Pr(s)[m,n], m=1..3),n=1..3)):

>

corr(s);

25

4

+

5

4

3

(−2s)

(−1)

(2s)

2

(2s)

+(−1)

s

3

(1−s)

2

(−1+s)

>

corr(0); corr(infinity);

9

25

4

Problem 5.4

>

with(linalg):

Warning, the protected names norm and trace have been redefined and

unprotected

>

Q0:=s->array([[0,(cos(Pi*s/2))^2,(sin(Pi*s/2))^2],[1/4

>

+1/2*(sin(Pi*s/2))^2,0,1/4+1/2*(cos(Pi*s/2))^2],[1/2*(cos(Pi*s/2))^2,1

>

/2+1/2*(sin(Pi*s/2))^2,0]]): Q0(s);

_

¸

¸

¸

¸

¸

¸

_

0 cos(

1

2

πs)

2

sin(

1

2

πs)

2

1

4

+

1

2

sin(

1

2

πs)

2

0

1

4

+

1

2

cos(

1

2

πs)

2

1

2

cos(

1

2

πs)

2

1

2

+

1

2

sin(

1

2

πs)

2

0

_

¸

¸

¸

¸

¸

¸

_

>

simplify(Q0(2*s));

_

¸

¸

¸

¸

_

0 cos(πs)

2

1−cos(πs)

2

3

4

−

1

2

cos(πs)

2

0

1

4

+

1

2

cos(πs)

2

1

2

cos(πs)

2

1−

1

2

cos(πs)

2

0

_

¸

¸

¸

¸

_

Since [cos(Pi*s)]ˆ2=1, we get

>

Q:=array([[0,1,0],[1/4,0,3/4],[1/2,1/2,0]]);

Q :=

_

¸

¸

¸

¸

_

0 1 0

1

4

0

3

4

1

2

1

2

0

_

¸

¸

¸

¸

_

>

QT:=transpose(Q):

>

r:=eigenvectors(Q):

>

l:=eigenvectors(QT):

>

r[1][1]; r[2][1]; r[3][1];

1

−

1

2

+

1

4

I

√

2

−

1

2

−

1

4

I

√

2

>

l[1][1]; l[2][1]; l[3][1];

−

1

2

+

1

4

I

√

2

−

1

2

−

1

4

I

√

2

1

>

r[3][3]; r[1][3]; r[2][3];

¦

_

1, −

1

2

−

1

4

I

√

2, −

1

6

+

1

3

I

√

2

_

¦

¦[1, 1, 1]¦

¦

_

1, −

1

2

+

1

4

I

√

2, −

1

6

−

1

3

I

√

2

_

¦

>

l[3][3]; l[1][3]; l[2][3];

¦

_

1,

8

5

,

6

5

_

¦

¦

_

−

1

3

−

1

3

I

√

2, −

2

3

+

1

3

I

√

2, 1

_

¦

¦

_

−

1

3

+

1

3

I

√

2, −

2

3

−

1

3

I

√

2, 1

_

¦

>

r1:=vector([1, 1, 1]): r2:=vector([1, -1/2+1/4*I*sqrt(2),

>

-1/6-1/3*I*sqrt(2)]): r3:=vector([1, -1/2-1/4*I*sqrt(2),

>

-1/6+1/3*I*sqrt(2)]): l1:=vector([1, 8/5, 6/5]):

>

l2:=vector([-1/3-1/3*I*sqrt(2), -2/3+1/3*I*sqrt(2), 1]):

>

l3:=vector([-1/3+1/3*I*sqrt(2), -2/3-1/3*I*sqrt(2), 1]):

>

n1:=evalm(l1&*r1): n2:=evalc(evalm(l2&*r2)):

>

n3:=evalc(evalm(l3&*r3)):

Probability matrices.

>

P1:=array([[l1[1]*r1[1], l1[2]*r1[1],l1[3]*r1[1]], [l1[1]*r1[2],

>

l1[2]*r1[2],l1[3]*r1[2]],[l1[1]*r1[3], l1[2]*r1[3],l1[3]*r1[3]]]);

P1 :=

_

¸

¸

¸

¸

¸

¸

_

1

8

5

6

5

1

8

5

6

5

1

8

5

6

5

_

¸

¸

¸

¸

¸

¸

_

>

P2:=array([[l2[1]*r2[1], l2[2]*r2[1],l2[3]*r2[1]], [l2[1]*r2[2],

>

l2[2]*r2[2],l2[3]*r2[2]],[l2[1]*r2[3], l2[2]*r2[3],l2[3]*r2[3]]]);

P2 :=

_

¸

¸

¸

¸

¸

¸

_

−

1

3

−

1

3

I

√

2 −

2

3

+

1

3

I

√

2 1

(−

1

3

−

1

3

I

√

2)(−

1

2

+

1

4

I

√

2) (−

2

3

+

1

3

I

√

2)(−

1

2

+

1

4

I

√

2) −

1

2

+

1

4

I

√

2

(−

1

3

−

1

3

I

√

2)(−

1

6

−

1

3

I

√

2) (−

2

3

+

1

3

I

√

2)(−

1

6

−

1

3

I

√

2) −

1

6

−

1

3

I

√

2

_

¸

¸

¸

¸

¸

¸

_

>

P3:=array([[l3[1]*r3[1], l3[2]*r3[1],l3[3]*r3[1]], [l3[1]*r3[2],

>

l3[2]*r3[2],l3[3]*r3[2]],[l3[1]*r3[3], l3[2]*r3[3],l3[3]*r3[3]]]);

P3 :=

_

¸

¸

¸

¸

¸

¸

_

−

1

3

+

1

3

I

√

2 −

2

3

−

1

3

I

√

2 1

(−

1

3

+

1

3

I

√

2)(−

1

2

−

1

4

I

√

2) (−

2

3

−

1

3

I

√

2)(−

1

2

−

1

4

I

√

2) −

1

2

−

1

4

I

√

2

(−

1

3

+

1

3

I

√

2)(−

1

6

+

1

3

I

√

2) (−

2

3

−

1

3

I

√

2)(−

1

6

+

1

3

I

√

2) −

1

6

+

1

3

I

√

2

_

¸

¸

¸

¸

¸

¸

_

>

P:=s->(1/n1)*P1+(1/n2)*(r[1][1])^(2*s)*P2+(1/n3)*(r[2][1])^(2*s)*P3;

>

Question (a)

P := s →

P1

n1

+

r

11

(2s)

P2

n2

+

r

21

(2s)

P3

n3

>

Pr:=s->evalm(P(s)):

>

Conditional probability

>

p01:=vector([1,0,0]): p02:=vector([0,1,0]):

>

PS1:=s->evalm(p01&*P(s)): PS2:=s->evalm(p02&*P(s)):

>

PS1(s)[1];

5

19

+

−

1

3

−

1

3

I

√

2

−

1

3

−I

√

2

+

(−

1

2

+

1

4

I

√

2)

(2s)

(−

1

3

+

1

3

I

√

2)

−

1

3

+I

√

2

>

PS1(s)[2];

8

19

+

−

2

3

+

1

3

I

√

2

−

1

3

−I

√

2

+

(−

1

2

+

1

4

I

√

2)

(2s)

(−

2

3

−

1

3

I

√

2)

−

1

3

+I

√

2

>

PS2(s)[1];

5

19

+

(−

1

3

−

1

3

I

√

2)(−

1

2

+

1

4

I

√

2)

−

1

3

−I

√

2

+

(−

1

2

+

1

4

I

√

2)

(2s)

(−

1

3

+

1

3

I

√

2)(−

1

2

−

1

4

I

√

2)

−

1

3

+I

√

2

>

PS2(s)[2];

8

19

+

(−

2

3

+

1

3

I

√

2)(−

1

2

+

1

4

I

√

2)

−

1

3

−I

√

2

+

(−

1

2

+

1

4

I

√

2)

(2s)

(−

2

3

−

1

3

I

√

2)(−

1

2

−

1

4

I

√

2)

−

1

3

+I

√

2

The results can be transformed to polar form by the command polar.

Problem 5.5

>

assume(N,natural): assume(k,real): assume(s>0); assume(n,integer):

>

P:=(n,s,N)->(1/(2*N+1))*(1+2*sum(cos(2*Pi*k*n/(2*N+1))*exp(-2*s*(sin(

>

Pi*k/(2*N+1)))^2),k=1..N)):

>

P(n,s,N);

1+2

_

N˜

∑

k˜=1

cos(2

πk˜ n˜

2N˜ +1

)e

(−2s˜ sin(

πk˜

2N˜+1

)

2

)

_

2N˜ +1

>

Pm:=[seq(P(i,2,2),i=-2..2)]: evalf(Pm);

[.1220644066, .2223516301, .3111679263, .2223516301, .1220644066]

Problem 5.8

>

with(linalg):

Warning, the protected names norm and trace have been redefined and

unprotected

>

w:=matrix(3,3,[0,1/2,1/2,1/3,0,1/3,1/3,1/3,0]);

w :=

_

¸

¸

¸

¸

¸

¸

_

0

1

2

1

2

1

3

0

1

3

1

3

1

3

0

_

¸

¸

¸

¸

¸

¸

_

>

M:=matrix(3,3,[-1,1/2,1/2,1/3,-1,1/3,1/3,1/3,-1]);

M :=

_

¸

¸

¸

¸

¸

¸

_

−1

1

2

1

2

1

3

−1

1

3

1

3

1

3

−1

_

¸

¸

¸

¸

¸

¸

_

>

MT:=transpose(M):

>

eigenvectors(M);

[−

5

6

+

1

6

√

13, 1, ¦

_

−

1

2

+

1

2

√

13, 1, 1

_

¦], [−

5

6

−

1

6

√

13, 1, ¦

_

−

1

2

−

1

2

√

13, 1, 1

_

¦],

[

−4

3

, 1, ¦[0, −1, 1]¦]

>

eigenvectors(MT);

[

−4

3

, 1, ¦[0, −1, 1]¦], [−

5

6

+

1

6

√

13, 1, ¦

_

−

1

3

+

1

3

√

13, 1, 1

_

¦],

[−

5

6

−

1

6

√

13, 1, ¦

_

−

1

3

−

1

3

√

13, 1, 1

_

¦]

>

lambda:=vector([-5/6+1/6*sqrt(13),-5/6-1/6*sqrt(13),-4/3]):

>

r1:=vector([-1/2+1/2*sqrt(13), 1, 1]); r2:=vector([-1/2-1/2*sqrt(13),

>

1, 1]); r3:=vector([0, -1, 1]);

r1 :=

_

−

1

2

+

1

2

√

13, 1, 1

_

r2 :=

_

−

1

2

−

1

2

√

13, 1, 1

_

r3 := [0, −1, 1]

>

l1:=vector([-1/3+1/3*sqrt(13), 1, 1]); l2:=vector([-1/3-1/3*sqrt(13),

>

1, 1]); l3:=vector([0, -1, 1]);

l1 :=

_

−

1

3

+

1

3

√

13, 1, 1

_

l2 :=

_

−

1

3

−

1

3

√

13, 1, 1

_

l3 := [0, −1, 1]

1

>

n1:=evalm(l1&*r1): n2:=evalm(l2&*r2): n3:=evalm(l3&*r3):

>

Q1:=-sum((1/n1)*r1[1]*l1[k]/lambda[1],k=1..3):

>

Q2:=-sum((1/n2)*r1[1]*l1[k]/lambda[2],k=1..3):

>

Q3:=-sum((1/n3)*r1[1]*l1[k]/lambda[1],k=1..3):

>

evalf(Q1+Q2+Q3);

>

First passage time

13.64536110

>

:

Bibliography

[1] L. E. Reichl A modern course in statistical physics, 1998.

141

2

Contents

1 2 General Comments Introduction to Thermodynamics 2.1 Additional Problems: Fluctuations 2.2 Mini-tests . . . . . . . . . . . . . 2.2.1 A . . . . . . . . . . . . . 2.2.2 B . . . . . . . . . . . . . 5 7 28 30 30 31

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

3

The Thermodynamics of Phase Transitions 33 3.1 Mini-tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.1.1 A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.1.2 B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Elementary Probability Theory . . . Stochastic dynamics ... The Foundations of Statistical Mechanics Equilibrium Statistical Mechanics Tests and training 49 59 77 79 103 107 107 107 107 107 109 110 111 115

4 5 6 7 8

A Additional information A.1 Thermodynamics . . . . . . . . . . . . . . . A.1.1 Thermodynamic potentials . . . . . . A.1.2 Variable transformation . . . . . . . . A.1.3 Derivatives from the equation of state A.2 Main distributions . . . . . . . . . . . . . . . A.3 The Dirac delta-function . . . . . . . . . . . A.4 Fourier Series and Transforms . . . . . . . . B Maple Printouts

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

3

4 CONTENTS .

1: Function is given by the deﬁnition f (x) = x4 + ax3 + bx2 .Chapter 1 General Comments This year the course will be delivered along the lines of the book [1]. electrodynamics. In any case. which requires basic knowledge of classical and quantum mechanics. The same page will contain solutions of the problems. The problems will also be selected from this book. (b) Investigate Fη (ξ). (a) Show that by a proper rescaling it can be expressed as f (x) = a4 Fη (ξ) where Fη (ξ) = ξ4 + ξ3 + ηξ2 . ξ ≡ x/a. student have to able to present solutions. obtained independently. – How many extrema it has? – When it has only 1 minimum? When it has 2 minima? At what value of η it has an inﬂection point? 5 . Using a simple test below please check if your knowledge is sufﬁcient. Function Fη (ξ) is very important in theory of phase transitions. A simple test in mathematics Elementary functions Problem 1. It is crucially important for students to solve problems independently. FYS 3130 (former FYS 203) is a complicated course. then she/he should come through the solution. as well as basics of mathematics. or with help of the course homepage. Answers can be found either in the Maple ﬁle. The problems will be placed on the course homepage. So if a student is not able to solve the problem without assistance. η ≡ b/a2 .

β) = n=0 ∞ n=0 ∑ e−αn cos βn . ∞ • Calculate inﬁnite sums C(α. . • G(γ) = e−γx . ∑ e−αn sin βn . CHAPTER 1.2: Logarithmic functions are very important in statistical physics.4: • In (γ) = • 0 Calculate interglars: ∞ xn e−γx dx . Check you memory by the following exercises: • Plot function 1−x 1+x Discuss properties of this function at |x| > 1. x(1 − x) dx .6 – Plot Fη (ξ) for this value of η. Basic integrals Problem 1. γ > 0. 2 /2 γ > 0. • Plot function • Simplify the expression f2 (x) = ln(tan πx) . e4 ln x − (x2 + 1)2 + 2x2 + 1 . Hint: take into account that cos x = Re eix . α > 0. f1 (x) = ln for |x| ≤ 1 .3: • Simplify Do you remember trigonometry? Test it! sin−2 x − tan−2 x − 1 . GENERAL COMMENTS Problem 1. β) = S(α. α > 0. sin x = Im eix . Problem 1. dx x2 ± a2 ∞ 0 dx .

y) = yx + (y3 − x3 )/3. ﬁnd the function u(x. ∂u(x. ∂x ∂y If we introduce u1 (x. As a result. u) = − arctan(x/y). this function cannot serve as a thermodynamic potential if both positve and negative values of x and y have physical meaning. ub (x.1: (a) The differential is exact. u) = − arctan(x/y) has a singularity at x = 0. (a) dua = −y dx x2 +y2 x dy + x2 +y2 .1: Test the following differentials for exactness. ua (x. (c) duc = (2y2 − 3x) dx − 4xy dy . 7 . any close path embedding this point contributes 2π to the variation of the quantity u(x. The function u(x. For those cases in which the differential is exact. du = ux dx + uy dy. (c) The differential is not exact. y).Chapter 2 Introduction to Thermodynamics Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Problem 2. y). The function ua (x. Consequently. y) = 0 x ux (ξ. uy = . Solution 2. Here one point worth discussion. y) is reconstructed in the following way. For an exact differential. y) dξ . (b) The differential is exact. y = 0. (b) dub = (y − x2 ) dx + (x + y2 ) dy . y) ∂u(x. y) ux = .

Problem 2. y) ∂u1 (x. ∂y dη uy (x. y) → (x. y) = 0 x dξ ux (ξ. b) → (x. Solution 2.8 CHAPTER 2. y) df = − = uy (x.2: The calculations are shown in the Maple ﬁle. Compute the change in two different ways: (a) Integrate along the path (a. ∂η x Finally. and 2. η) . . where a is Stefan’s constant. INTRODUCTION TO THERMODYNAMICS then the difference f ≡ u(x. y) − u1 (x. f (y) = 0 y y 0 0 x x 0 ∂ux (x. b) → (a. (a. Problem 2. y). y) + y 0 dη uy (x. b) → (x. y). ∂η For calculation see the Maple ﬁle it1. Consequently. For both differentials. ﬁnd the change u(x. u(x. du1 = (2xy + x2 ) dx + x2 dy.2: Consider the two differentials 1. η) − dξ dη ∂ux (ξ. y) between two points. y) − dy ∂y ∂y As a result. In the case (b) the results are different because the differential is not exact.mws. (b) Derive explicitly the efﬁciency of Carnot engine which uses black body radiation as its working substance. η) − y 0 0 dξ dη ∂ux (ξ. b) and (x. η) . (a) Plot the closed curve in the P −V plane for a Carnot cycle using black body radiation. (b) Integrate along the path (a. ∂u(x. y) dx . Discuss the meaning of your results. y) is a function only of y.3: Electromagnetic radiation in an evacuated vessel of volume V at equilibrium with the walls at temperature T (black body radiation) behaves like a gas of photons having internal energy U = aV T 4 and pressure P = (1/3)aT 4 . du2 = y(x − 2y) dx − x2 dy. y).

2) (2.9 P re s s u re 6 D " ! 6 ? V o lu m e Figure 2.V2 and adiabatically expand the gas to the point 3 characterized by the volume V3 .1: On the Carnot cycle with black-body radiation. dT 1 dV =− → V T 3 = const. Consequently. Let us start with isotherms. we ﬁnd 4 V3 ∆W = ∆Q1→2 + ∆Q3→4 = aTh4V1 −1 3 V4 1− Tc Th . 3 V1 In a similar way. (2. . 2. V2 4 ∆Q1→2 = ∆U + P∆V = (4/3)aTh4 (V2 −V1 ) = aTh4V1 −1 . thus isotherms are horizontal. 4 V3 ∆Q3→4 = aTc4V4 1 − . see Fig. T 3 V Let us start form the point 2 characterized by the values P1 .2. we get: V2 Th3 = V3 Tc3 . V4 Tc3 = V1 Th3 . V4 Tc3 = V1 Th3 → V2 V1 = = V3 V4 Tc Th 3 (2. PV 4/3 = const . Combining these equalities.1) (2. We have V2 Th3 = V3 Tc3 . Solution 2. 3 V4 Now let us consider an adiabatic path. In a similar way. Along an adiabatic path.2) and (2. (2.1 Along the ﬁrst isothermal path. At the isotherms the pressure is V -independent.3) Combining Eqs.1).3: We will follow example shown in Exercise 2. dQ = 0 = dU + P dV = 4aV T 3 dT + aT 4 dV + (1/3)aT 4 dV = 4aV T 3 dT + (4/3)aT 4 dV .3).

U0 = NcM T . On the 4 V2 ∆Q1→2 = aTh4V1 −1 . We have: (c) For an isothermal process at T = Tc .2. dU = T dS + H dM . Q1→2 = − In a similar way. (b) Since C = const. Problem 2. where cM is the speciﬁc heat per one particle while N is the number of particles. Q3→4 = Th 2 2 (M3 − M4 ) . (d) Compute the efﬁciency of the Carnot engine. 3 V1 As a result. n is the number of moles. (c) Compute the total heat absorbed and the total work done by the Carnot engine. Sketch a typical Carnot cycle in the M − H plane. 2nD H2 H1 ∆W Th − Tc = ∆Q1→2 Th H(M) dM = Tc 2 (M 2 − M2 ) . [1](2.4: (a) By deﬁnition [see Eq. A Carnot cycle is shown in Fig. D is a constant determined by the type of substance. Introducing molar quantities we get U0 = ncT . INTRODUCTION TO THERMODYNAMICS Remember: along a closed path ∆U = 0 and the total heat consumption is equal to mechanical work. (b) Let us assume that CM = C = constant. H is the magnetic ﬁeld. 2. Thus at M = const the internal energy is independent of the magnetization.10 CHAPTER 2.4: A Carnot engine uses a paramagnetic substance as its working substance. η= as it should be. Solution 2. and therefore the heat capacity CM .23)]. The equation of state is nDH M= T where M is magnetization. (a) show that the internal energy U. 2nD 1 . and T is is the temperature. can only depend on the temperature and not the magnetization.

The total work is then W = Q1→2 + Q3→4 . Q3→4 Th . 0 = dQ = dU − H dM = nc dT − Immediately we get dT 1 = 2 M dM . 2n2 cD ln Th 3 2 = M3 − M2 . 2 2 Q3→4 Th M3 − M4 Now let us discuss adiabatic paths. Th 2 2 2 M3 − M4 = −(M1 − M2 ) .11 M 4 T h 3 2 1 T c H Figure 2. T n cD Integrating this equality from point 2 to point 3 we obtain. We have at each path. and the efﬁciency is 2 2 W Tc M1 − M2 η= = 1+ .2: Sketch of the Carnot cycle. Tc MT dM . nD In a similar way. integrating from 4 to 1 we obtain 2n2 cD ln As a result. 2 2 2 2 M3 − M2 = M4 − M1 Tc 2 3 = M1 − M4 . → (d) Using this expression we obtain the efﬁciency η= Q1→2 + Q3→4 Th − Tc = .

the operating substance is an ideal monoatomic gas.12 CHAPTER 2. 2. Consequently.5: Let us start with the processes at constant volume. Express your answer in terms of V1 and V2 .4) Now let us consider the adiabatic processes where TV 2/3 = const (monoatomic ideal gas!).2. INTRODUCTION TO THERMODYNAMICS Coming back to the item (c) we ﬁnd W = Q3→4 η = 2 2 M3 − M4 Th − Tc . Assume that 2 = @ E= > = JE? " ! 8 Figure 2. (The processes 1 → 2 and 3 → 4 are adiabatic. T1 T4 = = T2 T3 V2 V1 2/3 ≡ α. T1 − T4 T1 − T4 (2. The mechanical work during theses processes does not take place. W = (3/2)nR(T1 + T3 − T2 − T4 ) . . The efﬁciency is given by the expression η= T1 + T3 − T2 − T4 T2 − T3 = 1− .18). Thus.5: Find the efﬁciency of the engine shown in Fig. 2nD Th2 Problem 2. Q4→1 = (3/2)nR(T1 − T4 ) . The processes 4 → 1 and 2 → 3 occur at constant volume). v Solution 2.3 ([1] -Fig.3: Sketch of the cycle. Q2→3 = (3/2)nR(T3 − T2 ) .

T Using the Maxwell relation for the Gibbs free [see Eq. As a result. T . Problem 2.112)] energy we obtain ∂S ∂P Thus Q = −T =− T P2 P1 ∂V ∂T = −V αP . One kilogram of water is compressed isothermally at 20◦ C from 1 atm to 20 (a) How much work is required? (b) How much heat is rejected? Assume that the average isothermal compressibility of water during this process is κT = 0.6: Since for an isothermal process dQ = T dS we have Q=T P2 P1 ∂S ∂P dP .4) we obtain: 1 V1 η = 1− = 1− α V2 2/3 . Solution 2. (2.5 × 10−4 (atm)−1 and the average thermal expansivity of water during this process is αP = 2 × 10−4 (◦ C)−1 . The mechanical work is W =− P2 P1 P dV = − P2 P P1 ∂V ∂P 2 2 dP = (M/2ρ)κT (P2 − P1 ) .6: atm.13 Substituting this expression into Eq. [1]-(2. Q = −(M/ρ)T αT (P2 − P1 ) [1 − κT (P − P0 )/2] . and V (P) = V0 [1 − κT (P − P0 )] = (M/ρ) [1 − κT (P − P0 )] . We see that since the compressibility of water is very low one can neglect the correction due to change in the volume and assume V ≈ M/ρ. P V (P)αT (P) dP Now let us assume that αT it P-independent.

2. (2. Qa→c = Mc(Tc − Ta ) = −(1/2)McTa .4: 2 L 0 Problem 2. The engine uses a rubber band whose equation of state is J = αLT . The speciﬁc heat (heat capacity per unit mass) is a constant. since J ∝ L. and T is the temperature in Kelvins. Indeed. η= . 2. cL = c. Ta = Tb = J0 /αL0 . the total work is W = Qb→a + Qa→c + Qc→b = (1/2)MJ0 L0 and W 1 = . J is the tension. Then. ( 2.7: Compute the efﬁciency of the heat engine shown in Fig. from the same equation we get. Solution 2.14 CHAPTER 2. As a result.5) that T = const. Qc→b = Mc(Tb − Tc ) − MJ0 L0 = (1/2)McTa − MJ0 L0 .5) where α is a constant.4 we see that the path a → b is isothermal. Qb→a 3 This result is also clear from geometrical point of view.7: From Fig. it follow from Eq. [1]-2. INTRODUCTION TO THERMODYNAMICS 2 J 0 a J 0 b c L 0 Figure 2. Tc = J0 /2αL0 = Ta /2 . Hence.19).4 (Fig. Qb→a = MαTa 2L0 L0 2 L dL = (3/2)MαL0 Ta = (3/2)MJ0 L0 . L is the length per unit mass.

T Consequently.0 × 103 dyne/K. T L where J is the tension. What is the change in temperature? Solution 2.5 m is the length of the band when no tension is applied. Physical meaning αJ = 1 L ∂L ∂T J is the thermal expansion coefﬁcient at given tension.8: (a) We have ∂L ∂T ∂T ∂J ∂J ∂L · J · L = −1 . (a) Compute (∂L/∂T )J and discuss its physical meaning. . 3 . (b) Find the equation of state and show that dJ is an exact differential. ∂L ∂T ∂T ∂J ∂J · ∂L L −1 ∂J ∂T =− J =− T L ∂J ∂L T L 1− =− T 1+2 L0 L L0 L 3 3 . (c) Assume that heat capacity at constant length is CL = 1. .0 J/K. The mass of the rubber band is held ﬁxed. Find the work necessary to stretch the band reversibly and adiabatically to a length 1 m. (b) The equation of state has the form aT L L0 J= 1− L0 L The proof of the exactness is straightforward. a = 1. the temperature of the band is T = 290 K.15 Problem 2.8: Experimentally one ﬁnds that for a rubber band ∂J ∂L ∂J ∂T aT L0 = 1+2 L0 L aL L0 = 1− L0 L 3 3 . Assume that when no tension is applied. and L0 = 0.

. The mechanical work is then W = CL (T f − T0 ) .V )/∂(T.V ) ∂(U. ∂(U.16 CHAPTER 2. Here L f and T f are ﬁnal values of the length and temperature. ∂S ∂U = V ∂A ∂V T we get ∂A ∂T V 4U 4 . 3 3 Consequently. respectively. µ = 0. This is the equation for adiabatic process which provides the change in the temperature. dT J(L. (a) What is the fundamental equation for the blackbody radiation (the entropy)? (b) Compute the chemical potential. Problem 2.V ) = = = . Solution 2. L0 Lf 3 1+2 .V )/∂(T.V ) (∂S/∂T )V 1 ∂(S. Measuring length in units of L0 as L = · L0 and introducing β ≡ aL0 /CL we obtain the following differential equation dT /T = −β (1 − Its solution is ln Tf β =− T0 2 Lf L0 2 −3 )d .9: Blackbody radiation in a box of volume V and at temperature T has internal energy U = aV T 4 and pressure P = (1/3)aT 4 . 0 = dQ = CL dT + J(L. where a is the Stefan-Boltzmann constant. INTRODUCTION TO THERMODYNAMICS (c) Consider adiabatic expansion of the band. Consequently. = aV T 3 = 3 3T ∂(S.V ) (∂U/∂T )V T (b) Since A is N-independent. From the relation P = − 1 1 A = −PV = − aV T 4 = − U . T ) L0 aT L =− =− 1− dL CL CL L0 L 3 .9: (a) Let us ﬁrst ﬁnd the Helmholtz free energy. T ) dL . S=− As a result.

P 2NkTi 1 1 Vi = Pi NkTi → P2 4V1V2 = . the free energy of the ideal gas can be constructed as kT A = − ln N! N ∑e pα −εpα /kT ≈ −NkT ln eV N mkT 2π¯ 2 h 3/2 ∑ e−εα /kT α = −NT ln(eV /N) + N f (T ) . f (T ) = −kT ln The Gibbs free energy is then G = A + PV = NkT ln P + Nχ(T ) . δW f ree .10: Two vessels. the energy of the particle can be written as a sum of the kinetic energy.10: The Gibbs free energy of an ideal gas is given by the equation G = NkT ln P + Nχ(T ) where χ(T ) is some function of the gas excitation spectrum. and N. V1 . the entropy being S f = −2Nk ln P − 2Nχ (T ) . εpα = εp + εα . As we will see later.17 Problem 2. χ(T ) ≡ f (T ) − kT ln kT . . The ﬁnal volume is V = V1 +V2 . εp = p2 /2m. mkT 2π¯ 2 h 3/2 ∑ e−εα /kT α . What is the maximum work. ∂T P Before the vessels are connected. V2 . P1 P2 (V1 +V2 )2 For an ideal gas. After the vessels are connected the temperature remains the same. The vessels are then connected and allowed to reach equilibrium in such a way that the combined vessel is also insulated from the outside world. Consequently. insulated from the outside world. The gas in each vessel is originally at temperature Ti . one of volume V1 and the other of volume V2 . On the other hand. Si = −Nk ln(P1 P2 ) − 2Nχ (T ) . εα (characterized by some quantum numbers α). that can be obtained by connecting these insulated vessels? Express your answer in terms of Ti . 1 V1 +V2 = . as it follows from the conservation law. ∆S = −Nk ln(P2 /P1 P2 ). and the energy of internal excitations. contain equal numbers N of the same ideal gas.1 Consequently. Solution 2. ∂G S=− = −Nk ln P − Nχ (T ) .

The heat capacity will have corrections to its ideal gas value. Solution 2.N . V δS ≡ −k N2 B2 (T ) + T B2 (T ) . V V where B2 (T ) is the second virial coefﬁcent. . 2 V (a) Find the form that f (T ) must have in order for the two equations to be thermodynamically consistent. (c) Find the entropy and internal energy. V that P = − ∂A ∂V T .18 CHAPTER 2.11: For a low-density gas the virial expansion can be terminated at ﬁrst order in the density and the equation of state is P= NkT N 1 + B2 (T ) . the maximum work is ∆W f ree = Ti ∆S = NkTi ln (V1 +V2 )2 . 4V1V2 Problem 2. V Since we know both entropy and Helmholtz free energy.11: (c) The equation of state under consideration can be obtained from the Helmholtz free energy2 A = Aideal + kT B2 (T ) Then S=− N2 . INTRODUCTION TO THERMODYNAMICS As a result. we ﬁnd the internal energy as U = A + T S = Uideal + kT B2 (T ) = Uideal + δU . 2 Remember N2 N2 − kT B2 (T ) + T B2 (T ) V V N2 δU ≡ −kT 2 B2 (T ) . V ∂A ∂T = Sideal + δS . (b) Find CP.N = Nk − f (T ) . We can write it in the form 3 N 2k CV.

CV = T ∂S ∂T = V ∂U ∂T ideal = CV + δCV .N ∂H ∂Y =T T. V δCV ≡ −kT N2 2B2 (T ) + T B2 (T ) . V We ﬁnd in this way. Since the density is assumed to be small in the correction one can use equation for the ideal gas to ﬁnd the the volume.N ∂X ∂T −X . (b) Let us express the equation of state as V (P. V Problem 2. Now ideal CP = CP + T δS1 ≡ −kN B2 + T B2 . the entropy can be expressed as S = Sideal + δS1 . kT /P + B2 ∂δS1 ∂T P (B2 + T B2 ) (kT /P + B2 ) − (k/P + B2 )(B2 + T B2 ) = −kN (kT /P + B2 )2 N2 ideal = CP − k T (2B2 + T B2 ) − (B2 + T B2 ) V N2 ideal = CP + δCV + k (B2 + T B2 ) . f (T ) = 2T B2 (T ) + T 2 B2 (T ) . CP −CV = (CP −CV )ideal + k N2 (B2 + T B2 ) .12: Prove that CY. Y. V Here in all corrections we used equation of state for an ideal gas. As a result. V= NkT + NB2 (T ) . We have. T ). P Consequently.N = ∂H ∂T and Y.N .19 (a) As a result.

As a result. where m is the molar magnetization. m) = (c + m2 /2D) ln(u/u0 ) . H is the magnetic ﬁeld. m) = T (c + m2 /2D) . The enthalpy is deﬁned as H = U − XY = ST + ∑ µ j dN j .N ∂H ∂S =T. Y. m) = u .13: Compute the entropy. T ∂H ∂Y = −X .13: Let us start with the internal energy u(T. Helmholtz free energy.N To prove the second relation we do the following ∂H ∂Y Now. c + m2 /2D The molar entropy s is then derived from the deﬁnition ∂s ∂u = m c + m2 /2D 1 = T u → s(u.20 CHAPTER 2. and T is the temperature. Consequently.12: Let us ﬁrst recall deﬁnitions for X. T (u. INTRODUCTION TO THERMODYNAMICS Solution 2. J Since dU = T dS +Y dX we get dH = T dS − X dY .N ∂H ∂T ∂H ∂S . The “thermal” contribution is cT . Assume that mechanical equation of state is m = (D/T )H and the the molar heat capacity at constant magnetization is cm = c. D is a constant.N Now we have to use the Maxwell relation. c is a constant.N = T ∂S ∂T ∂H ∂Y = Y. Y. = T. CY. Solution 2.66).N ∂S ∂Y . u(T. Since H(m) = (T /D)m we get umag = (T /2D)m2 . From dG = −S dT − X dY we get ∂S ∂Y = T ∂X ∂T .N + S. Problem 2. U = ST +Y X + ∑J jµ j dN j . Y. We have the magnetic contribution umag = 0m H(m) dm. S.N Y. enthalpy. m) per one mole. ([1]-2. Eq. and Gibbs free energy for a paramagnetic substance and write them explicitly in terms of their natural variables if possible. which emerges for the Gibbs free energy G = H − T S.N Thus we obtain the desired result. .

m) = (c + m2 /2D) ln . V − nb V . which has to be expressed through T and H. H) = u − Hm = T (c + DH 2 /2T 2 ) − (D/T )H 2 = T (c − DH 2 /2T 2 ) .14: state is Compute the Helmholtz free energy for a van der Waals gas. As a result. Is this a reasonable choice for the heat capacity? Should it depend on volume? Solution 2. We get g = a − Hm = T (c + DH 2 /2T 2 ) 1 − ln T (c + DH 2 /2T 2 ) − (D/T )H 2 u0 T (c + DH 2 /2T 2 ) = T (c − DH 2 /2T 2 ) − T (c + DH 2 /2T 2 ) ln . Finally. u0 In particular. αn2 nRT − 2 . the Helmholtz free energy is T (c + m2 /2D) a(T. m). V2 where α and b are constants which depend on the type of gas and n is the number of moles. Assume that heat capacity is CV.21 Here u0 is a constant. in “natural” variables u(s. The equation of P+α n2 (V − nb) = nRT . u0 2 To get enthalpy we have to subtract from u the quantity Hm = (D/T )H 2 and to express m through H as m = (D/T )H. we obtain: h(T. m) = u − T s = T (c + m /2D) 1 − ln . m) = u0 exp s c + m2 /2D .n = (3/2)nR. To ﬁnd other thermodynamic potentials we need s(T. P= Since P = −∂A/∂V we obtain A=− P(V ) dV = −nRT ln(V − nb) − (αn2 /V ) + A (T ) . Gibbs free energy is g = a − Hm. As a result. u0 Problem 2.14: Let us express pressure through the volume. We can rewrite the above expression for the entropy as T (c + m2 /2D) s(T.

(b) Let us ﬁrst calculate the adiabatic compressibility (∂V /∂P)S as ∂V ∂P Consequently. INTRODUCTION TO THERMODYNAMICS Here A (T ) is the integration constant. (∂V /∂P)T We use the method of Jacobians: Now. T ) ∂(P. T CV. P) (∂S/∂T )P (∂V /∂P)T − (∂S/∂P)T (∂V /∂T )P = T (∂V /∂P)T (∂S/∂P)T (∂V /∂T )P = CP − T . CP −CV = −T [(∂V /∂T )P ]2 α2 = TV P . T ) (∂S/∂T )P ∂P CP (∂V /∂P)T κT = = . S)/∂(V. P) ∂(T. S) ∂(V. Indeed. T ) ∂(V. T S(T ) dT = (3/2)nRT [1 − ln(T /T0 )] . from the Maxwell relations (∂S/∂P)T = − (∂V /∂T )P . T .V )/∂(T.22 CHAPTER 2.n dT = (3/2)nR ln(T /T0 ) . A (T ) = − T Here we omit temperature-independent constant. (∂V /∂P)T κT The ﬁrst relation follows from this in a straightforward way from deﬁnitions.V )/∂(T.15: (a) CV = T (∂S/∂T )V = T ∂(S.V ) = ∂(S. Thus. S) ∂(P.15: Prove that (a) κT (CP −CV ) = TV α2 P (b) CP /CV = κT /κS . which can be found from the given speciﬁc heat.V )/∂(T. S)/∂(P. (Check!) Problem 2. CV (∂V /∂P)S κS . Solution 2. T ) (∂S/∂T )V ∂V = · = · ∂(P. The suggestion regarding speciﬁc heat is OK since the difference between the entropies of van der Waals gas and the ideal gas is temperature independent. = S ∂(V. S= Consequently.

where x = X/n is the amount of extensive variable.15. where M is the magnetization.15 we can derive the relation cP − cv = −T [(∂P/∂T )v ]2 R = .17: Let us start with ther speciﬁc heat. Problem 2.n . per mole.n and χS. and the thermal expansivity αP of a monoatomic van der Waals gas.17: Compute the molar heat capacity cP . Now the combination cx (∂T /∂Y )x dY + cY (∂T /∂x)Y dx can be rewritten as T (∂S/∂T )x (∂T /∂Y )x dY + T (∂S/∂T )Y (∂T /∂x)Y dx = T (∂S/∂Y )x dY + T (∂S/∂x)Y dx = T ds . X. Using the method similar to the Problem 2. the susceptibilities χT. v−b v and the molar heat capacity is cv = 3R/2. other quantities can be calculated using results of the Problem 2. vRT 1 − 2α(v − b)2 /RT v3 Given cv .16: Show that T ds = cx (∂T /∂Y )x dY + cY (∂T /∂x)Y dx . Problem 2. Solution 2. and the thermal expansivity αH.n = nc. (∂P/∂v)T 1 − 2α(v − b)2 /RT v3 Now let us compute the compressibility κT = − 1 v ∂v ∂P =− T 1 v ∂P ∂v −1 = T 1 (v − b)2 . .n .18: Compute the heat capacity at constant magnetic ﬁeld CH. cx is the heat capacity per mole at constant x. where v = V /n is the molar volume. Start from the fact that the mechanical equation of state is RT α P= − 2. and T is the temperature.23 Problem 2. and cY is the heat capacity per mole at constant Y . κT and κS . c is the molar heat capacity.16: Let us substitute the deﬁnitions cx = T (∂S/∂T )x . Solution 2. cY = T (∂S/∂T )Y . n is the number of moles. the compressibilities. given that the mechanical equation of state is M = nDH/T and the heat capacity CM. H is the magnetic ﬁeld.n for a magnetic system.

n .n = ∂M ∂H = T.n −CM. S) ∂(H. = · = χT. χT. T Now.n ∂(H. M becomes dependent only on temperature.n = Given CH. ∂M ∂H = S CM. S) ∂(M. Here v = V /n is the molar volume.n Now let us ﬁnd CH. T ) ∂(H. χS. Then. INTRODUCTION TO THERMODYNAMICS Let us start with susceptibilities.n ∂(M. the contribution to the internal energy is dU = −H dM = As a result. αH is deﬁned as αH = ∂M ∂T =− H nDH 2 dT . dM = nDH dM dT = − 2 dT .n nD .24 Solution 2.n . χT. S)/∂(H. S)/∂(M.18: CHAPTER 2. (R2.n = .n CH. T2 Problem 2. According to Eq. By deﬁnition. T2 nDH 2 . T )) .149). T ) CH. T ) ∂(M. (b) Find the equation of state.19: A material is found to have a thermal expansivity αP = v−1 (R/P + a/RT 2 ) and an isothermal compressibility κT = v−1 [T f (P) + b/P]. (c) Under what condition this materials is stable? . T2 nDH .n and χS. At constant H.n Thus we have found one relation between susceptibilities and heat capacities. (a) Find f (P). CH.n = nc we easily compute CH. dT T Consequently.n CM.

Now. P P RT (c) Since the compressibility must be positive.19: (a) By deﬁnition. 2. P2 ∂v dP = ∂P P dP − RT b − P2 P RT − b ln P + g(T ) .5 (R2. The efﬁciency of a heat engineis η = ∆Wtotal /∆Qabsorbed . we can express the equation of state as v − v0 = RT P0 a + b ln − .20). As a result. we have the stability condition T f (P) + b >0 P → TR b + > 0. we have ∂v R a = + . P2 P Consequently.20: Compute the efﬁciency of the reversible two heat engines in Fig. ∂v R R a = + g (T ) ≡ + . the stability condition is P/T < R/b . P Here g(T ) is some function of the temperature. ∂T P RT 2 ∂v b = −T f (P) − .25 Solution 2. Problem 2. (b) We can reconstruct the equation of state as: v = = P R = − f (p) . Which engine is the most effective? (Note that these are not Carnot cycles. ∂P P To make dv an exact differential we need: − Thus f (p) = R/P2 . . ∂T P P RT 2 Thus g(T ) = −a/RT + const.

v is the molar volume. and f (P) is unknown function.21: It is found for a gas that κT = T v f (P) and αP = Rv/P + Av/T 2 . . Thus ηb = T2 − T1 > ηa . (a) What is f (P)? (b) Find v = v(P. The heat absorbed in the case (a) is ∆Qabsorbed = T2 (S2 − S1 ) . we immediately get for any closed path in the T − S plane: ∆Wtotal = This is just the area of the triangle. Problem 2. where T is the temperature. ∆Wtotal = (1/2)(T2 − T1 )(S2 − S1 ) . A is a constant.20: Since dQ = T dS.5: Solution 2. ηa = In the case (b). P is the pressure. T2 + T1 T2 − T1 . INTRODUCTION TO THERMODYNAMICS = > 6 6 = > ? = 6 ? 5 5 5 5 5 > 5 Figure 2. T ). 2T2 T dS .26 6 6 6 CHAPTER 2. Thus. it easy to show that ∆Qabsorbed = (1/2)(T2 + T1 )(S2 − S1 ) .

but does not allow liquid to pass. γ = γ0 + Consequently. Then we can express γ as P ∂γ dP = T R ∂P dP RT =− + g(T ) . We get. which contains both liquid and gas particles. µL = µG . the chemical potentials and pressures of gas in both parts must be equal. As a result. or g(T ) = A/T + const. Solution 2. T P Problem 2. γ0 + A/T − RT /P A RT − . T . then ∂µL ∂PL δPL = T ∂µG ∂PG δPG . T ) = µG (PG . g (T ) = −A/T 2 . T ) . If the pressure of the liquid is increased by pushing in on the piston. 2 P P Then. In this part the chemical potentials must be equal. T ) = 1 . If one changes the pressure of liquid by δPL . = f (p). v(P. ∂γ R A = − − 2. by how much does the pressure of the gas change? [Assume the liquid in incompressible (its molar volume is independent of the pressure) and describe the gas by the ideal gas equation of state. from the Maxwell relation we get γ= P R P2 . ∂P Let us introduce γ(P. T ) ≡ [v(P.27 Solution 2. µL (PL . On the other hand. The entire process occurs at ﬁxed temperature T ].22: Let us consider the part of the system. ∂T P T ∂γ = T f (P) . but the volume of the liquid cam be varied by moving a piston.21: The solution is similar to the problem 2.The volume of the gas is held ﬁxed at VG .22: A monomolecular liquid at volume VL and pressure PL is separated from a gas of the same substance by a rigid wall which is permeable to the molecules. ∂P Again.19. Thus we arrive at the equation. We have: ∂v R A = v2 + ∂T P T2 ∂v = −v2 T f (P) . T )]−1 .

P = NkT /V .T ∂V ∂N = vL . T −P V ∂V ∂P +CV kT 2 . vL ˜ δPG = . δPL vG ˜ 2. As a result. the quantity vL has a physical meaning of ˜ ˜ the volume per particle. (∆T )2 = kT 2 /CV .1 Additional Problems: Fluctuations Quick access: 23 24 25 26 27 Problem 2. (∆U)2 = (3/2)N(kT )2 .23: Find the mean square ﬂuctuation of the internal energy (using V and T as independent variables). Thus (∆U)2 = −kT T For the ideal gas. What is the mean square ﬂuctuation of the internal energy for a monoatomic ideal gas? Solution 2. ∂P ∂T (∆V )2 = −kT 2 2 = T ∂P ∂T 2 −P V 2 (∆V )2 +CV (∆T )2 . T . ∂µ ∂P = T ∂ ∂P ∂G ∂N = P.28 CHAPTER 2. ∂V ∂P .23: We have ∂U ∂V ∆V + T ∆U = ∂U ∂T ∆T = T V ∂P ∂T − P ∆V +CV ∆T . Squaring and averaging we obtain (note that ∆V ∆T = 0) (∆U) Now. (∂µ/∂P)T = kT /PG = VG /NG ≡ vL . INTRODUCTION TO THERMODYNAMICS For an ideal gas. Thus. For a liquid. V Here we use Maxwell relation which can be obtained from Helmholtz free energy. ˜ T The last relation is a consequence of incompressible character of the liquid. the relation is the same. CV = (3/2)Nk .T ∂ ∂N ∂G ∂P = N.

**2.1. ADDITIONAL PROBLEMS: FLUCTUATIONS
**

Problem 2.24: ideal gas. Solution 2.24:

29

Find ∆T ∆P (with variables V and T ) in general case and for a monoatomic

We have ∆P =

∂P ∂V ∂P ∂T

∆V +

T

∂P ∂T

∆T .

V

**Squaring and averaging we obtain (note that ∆V ∆T = 0) ∆T ∆P = For an ideal gas, ∆T ∆P = 2 kT 2 . 3 V (∆T )2 =
**

V

kT 2 CV

∂P ∂T

.

V

Problem 2.25: Solution 2.25:

**Find ∆V ∆P with variable (V and T ). Using results of the previous problem we get ∆V ∆P = ∂P ∂V (∆V )2 = −kT .
**

T

Problem 2.26:

**Using the same method show that ∆S ∆V = kT ∂V ∂T ,
**

P

∆S ∆T = kT .

Solution 2.26:

Starightforward.

Problem 2.27: Find a mean square ﬂuctuation deviation of a simple pendulum with the length suspended vertically. Solution 2.27: Let m be the pendulum mass, and φ is the angle of deviation from the vertical. The minimal work is just the mechanical work done against the gravity force. For small φ, Wmin = mg · (1 − cos φ) ≈ mg φ2 /2 . Thus, φ2 = kT /mg .

30

CHAPTER 2. INTRODUCTION TO THERMODYNAMICS

2.2 Mini-tests

2.2.1 A

The Helmholtz free energy of the gas is given by the expression A = Nε0 − NkT ln(eV /N) − NcT ln T − NζT . Here e = 2.718 . . . is the base of natural logarithms, N is the number of particles, V is the volume, T is the temperature in the energy units, while ε0 , c and ζ are constants. (a) Find the entropy as function of V and T . (b) Find internal energy U as a function of the temperature T and number of particles N. (c) Show that c is the heat capacity per particle at given volume, cv . (d) Find equation of state. (e) Find Gibbs free energy, enthalpy. Find the entropy as a function of P and T . (f) Using these expression ﬁnd the heat capacity at constant pressure, cP . (g) Show that for an adiabatic process T γ P1−γ = constant, where γ = cP /cv .

Solution

By deﬁnition, S=− Now, U = A + T S = Nε0 + NcT , ∂A NT P = − , = ∂V T,N V G = A + PV = A + NT = Nε0 − NT ln = W = S = cP = V − NcT lT + NζT N Nε0 + NT ln P − NT (1 + c) ln T − NζT , U + PV = U + NT = Nε0 + N(c + 1)T , −N ln P + N(1 + c) ln T + N(ζ + 1 + c) , ∂S T = 1+c. ∂T P,N ∂A ∂T = N ln

V,N

eV + Nc (1 + ln T ) + NζT . N

2.2. MINI-TESTS

Immediately, to keep entropy constant we get −N ln P + NcP ln T = const Since cP − cV = cP − c = 1, we obtain T cP PcP −cV = const , → T γ P1−γ = const . → T cP /P = const .

31

2.2.2 B Problem 1

Discuss entropy variation for (a) adiabatic process, (b) isothermic process, (c) isochoric process, (d) isobaric process.

Problem 2

(a) Discuss the difference between Gibbs and Helmholtz free energy. (b) Prove the relation ∂ A U = −T 2 . ∂T T (c) A body with constant speciﬁc heat CV is heated under constant volume from T1 to T2 . How much entropy it gains? (d) Discuss the heating if the same body is in contact with a thermostat at T2 . In the last case the heating is irreversible. Show that the total entropy change is positive. (e) Two similar bodies with temperatures T1 and T2 brought into contact. Find the ﬁnal temperature and the change in entropy.

Solution

Problem 1 (a) Constant (b) S = Q/T . (c) S=

T2 T1

T2 CV (T ) dT = CV ln . T T1

T T1 Mechanical work under isothermic process is given by dW = dU − dQ = dU − T dS = d(U − T S) . A is the thermodynamic potential with respect to V and T . T2 T1 ∂A ∂T V CV T2 dT = CV ln T T1 . INTRODUCTION TO THERMODYNAMICS S= Problem 2 T2 T1 T2 CP (T ) dT = CP ln . . → TB = (T1 + T2 )/2 . The thermodynamic potential with respect to P and T is called Gibbs free energy. The function of state A = U −TS is called Helmholtz free energy. (b) Substituting S=− into deﬁnition of A we get the result. T Thus. V P=− . (c) We have S= (d) Use 2nd law of thermodynamics (e) Energy conservation law yields CV (T2 − TB ) = CV (TB − T1 ) for the entropy change we have δS = CV ln TB T1 + T2 TB +CV ln = 2CV ln √ T1 T2 2 T1 T2 ≥ 0. We have G = U − T S − PV → dG = −S dT +V dP . Thus. We have dA = dU − T dS − S dT = (T dS − P dV ) − T dS − S dT = −S dT − P dV. S=− ∂A ∂T ∂A ∂V .32 (d) CHAPTER 2.

(a) Compute equation of state. 2 v 5 a a = u − T s = RT − − T s0 + R ln C(v − b)(5RT /2)5/2 2 v 33 ∂s ∂u = v a 5R u+ 2 v −1 . We have 1 = T Thus. . (b) Compute the molar heat capacities. s = s0 + R ln C(5RT /2)5/2 (v − b) .1: a) Let us ﬁrst ﬁnd the temperature as function of u and s. where c and s0 are constants. How can you ﬁnd explicit values of vl and vg if you need to? Solution 3. cc and cP .1: A condensible vapor has a molar entropy s = s0 + R ln C(v − b) u + a v 5/2 . u = 5 a RT − . the gas constant R and gas molar volumes vl and vg . (c) Compute the latent heat between liquid and vapor phases temperature T in terms of the temperature T .Chapter 3 The Thermodynamics of Phase Transitions Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13 Problem 3. .

= T RT a − 2. Another way is to plot the isotherm and ﬁnd the volumes using the Maxwell rule. v−b v cv = T ∂s ∂T 5 = RT . for a gas maintained in equilibrium with its liquid phase. αcoex = 1 v ∂P ∂T ∂v ∂T + P ∂v ∂P T dP . coex According to the Clapeyron-Clausius formula. Solution 3. Find an approximate explicit expression for αcoex . 2 v The result for cP can be obtained in the same way as in the problem 2. Discuss its behavior. we have q = T (sg − sl ) = RT ln RT a RT a − 2= − 2. Thus = R/P . using the ideal gas equation of state. T (vg − vl ) T vg . Pl = PG .34 As a result. = coex q q ≈ . µl = µg . CHAPTER 3. We know that µ = (∂a/∂n)T.2: Find the coefﬁcient of thermal expansion.17.2: have.V . αcoex = v−1 (∂v/∂T )coex . vl − b Since the pressure should be constant along the equilibrium line. Problem 3. v = RT /P. THE THERMODYNAMICS OF PHASE TRANSITIONS ∂a ∂v P=− This is the van der Waals equation. so everything could be done. T ∂v ∂T = coex 1 T 1− T P ∂P ∂T . (b) We have. We ∂v = ∂T ∂v ∂T Consequently. P ∂v ∂P = −RT /P2 . vg − b vg vl − b vl Another equation is the equality of chemical potentials. It is implicitly assumed that the total volume of the system is kept constant. dT For an ideal gas. (c) The latent heat q is just vg − b .

the slopes of the vaporization and sublimation curves are given by the relations sg − sl sg − sl sg − ss sg − ss ∂P ∂P = ≈ . dT sg − sl sg vl − vg sl = . Express your answer in terms of sl . The deﬁnitions of the sublimation and vaporization curves are given in Fig. The coefﬁcient of thermal expansion is less because with the increase of the temperature under given pressure the heat is extracted from the system.4: We have µ(P. Solution 3. = ≈ .35 where q = (∆h)lg . Solution 3. vg which are the molar entropy and molar volume of the liquid and gas. we can rewrite this expression as αcoex ≈ q 1 1− T RT . Problem 3.3: Prove that the slope of the sublimation curve of a pure substance at the triple point must be greater than that of the vaporization curve at the triple point. 3. T ). where µ is the chemical potential and T is the temperature.3: The triple point is deﬁned by the following equations for the 2 phases: P1 = P2 = P3 . T1 = T2 = T3 . vl and sg . dµ dT = −sg + vg coex ∂µg ∂T + P ∂µg ∂P T dP . dµ = dT Consequently. respectively. Consequently. vg − vl vg − vl . Compute the rate of change of chemical potential along the coexistence curve. ss < sl . ∂T lg vg − vl vg ∂T sg vg − vs vg Since solid state is more ordered. T ) = µg (P. µ1 = µ2 = µ3 . T ) = µl (P.4 of the textbook [1]. vg Problem 3. (∂µ/∂T )coex . Thus. and ∂P ∂T − lg ∂P ∂T = sg ss − sl < 0.4: Consider a monoatomic ﬂuid along its ﬂuid-gas coexistence curve.

Let us look at Fig. (b) Since P = − (∂a/∂v)T we have. s Pl = 2A/T v3 . (b) How are the molar volumes.5: (a) By deﬁnition.6: Since the system in the equilibrium. v. v is the volume per mole. 3 Since at the transition point gl = gs . we obtain v3 2A l . we have P = T 37/4 A1/3 28/3 B1/4 12 3 gl = (2A)1/3 P2/3 T −1/3 .1. where A and B are constants. g = a − v (∂a/∂v)T . s gl = 3A/T v2 . the maximum work extracted during the process of the phase transition is ∆A − P0 ∆V . (a) Compute the Gibbs free energy density of the liquid and solid phases. 3. l gs = 4B/T v3 . of the liquid and solid related at the liquid-solid phase transition? (c) What is the slope of the coexistence curve in the P − T plane? Solution 3. THE THERMODYNAMICS OF PHASE TRANSITIONS Problem 3. and in its liquid phase it has a Helmholtz free energy per mole. g = a + Pv. Ps = 3B/T v4 . l Since at the phase transition Ps = Pl = P. Thus. = v4 3B s (c) Now we can express the Gibbs free energies per mole in terms of P. 224 B3 Problem 3.36 CHAPTER 3. and P = − (∂a/∂v)T . . 4 gs = (3B)1/4 P3/4 T −1/4 .5: A system in its solid phase has a Helmholtz free energy per mole. Solution 3.6: Deduce the Maxwell construction using stability properties of the Helmholtz free energy rather than the Gibbs free energy. and T is the temperature. as = B/T v3 . 2 = 321 A4 . al = A/T v2 .

and the chqemical potentials µA and µB are intensive so that µA = µA (P. A and B. T.5.5 there is no stavle region at all.5. The graphs have the form We see that for ¯ T = 0. For this system the fundamental equation fog the Gibbs energy is G = nA µA + nB µB . To illustrate the situation we plot the curves in the ¯ vicinity of T = 1. Subtracting areas we reconstruct the Maxwell rule. . xB ) . is P = 0. ¯ ¯ ¯ ¯ Problem 3. ∆A = c P dV . plot the isotherms in the in the P − V plane (P and V ¯ ¯ ¯ are the reduced pressure and volume) for reduced temperatures T = 0.37 P b a c P0 v l v g V Figure 3.0.7: We use Maple to plot the curves.5. Problem 3. To ﬁnd ∆A let is follow isotherm.2) (3. T. ¯ = 0. Since T = const.1) (S is the total entropy and V is the total volume of the system).8: Consider a binary mixture composed of two types of particles. and T = 1.1 the equilibrium pressure of the liquid gas coexistence region? ¯ For T Solution 3. xA ) and µB (P.7: For a van der Waals gas. the combined ﬁrst and second laws are dG = −S dT +V dP + µA dnA + µB dnB (3. a which is just the area below the isotherm a − b − c.1: We see that P0 ∆V = P0 (Vg −Vl ) is just the area of the dashed rectangle. T = 1.

the lower one .4 1.4 0. 2 1.nβ=α with α = A. Solution 3. and vα = (∂V /∂nα )P.2 v ¯ ¯ Figure 3.8: Let us express nα as nα = nxα and divide Eq. Use these facts to derive the relations s dT − v dP + and α=A.4) where s = S/n. (3.T.T.2 0 1 2 v 3 4 5 ¯ ¯ Figure 3.to T = 0.to T = 0.5) .38 CHAPTER 3.3: The upper curve corresponds T = 1.2: The upper curve corresponds T = 1.8 0.05.8 0.5.2 P 1 0.5. sα = (∂S/∂nα )P.1 1. B. the lower one .1) by n.3) α=A.9 1 1. B and β = A.nβ=α . We get dg = d ∑ xα µα α = ∑ (xα dµα + µα dxα ) α (3.85.B ∑ xα (dµα + sα dT − vα dP) = 0 .6 1. n = nA + nB .7 0.6 0.6 0. THE THERMODYNAMICS OF PHASE TRANSITIONS 12 10 8 6 4 2 0 –2 –4 0.B ∑ xα µα = 0 (3.8 1. (3.

39 The let us divide Eq. and vα = (∂V /∂nα )P.9: Consider liquid mixture (l) of particles A and B coexisting in equilibrium with vapor mixture (g) of particles A and B. we prove Eq. the molar volume v.T. (3.4) to the gas phase. (3. Further. Compute the slope (dP/dT )coex of the λ line in terms of the temperature T . P P P P g . (3. Problem 3.T.nβ=α nα = n ∑ sα xα . the entropy is an extensive variable. I and II. Thus α dPl dµα dµl = α = −sl + vl . α α dT dT dT Since at the equilibrium Pg = Pl . We get g ∑ xα α dµα + sα dT − vα dPg = 0 . α (3. = g g l ) + xg (vg − vl ) xA (vA − vA B B B g g g g (3. of the system.8) is useful. Consequently. (3.nβ=α . (3. and ∆αP = αI − αII .7) l xA where sα = (∂S/∂nα )P. g g g Dividing this equation by dT we obtain ∑ α g xα dµα g g dPg + sα − v α dT dT g = 0. S = ∑ (∂S/∂nα )P.2) by n to obtain dg = −s dT + v dP + ∑ µα dxα .3).7). α α Substituting these expressions into Eq. let us take into account that at the coexistence curve µα = µl . g Now.6) Equating the right hand sides of these equation we prove Eq.T.10: A PV T system has a line of continuous phase transitions (a lambda line) separating two phases.9: Let us apply Eq. [Hint: Equation (b) of the Problem (3.3) we prove Eq.is should be a homogeneous function of nα .nβ=α with α = A. The molar heat capacity cP and the thermal expansivity αP are different in the two phases. ∆cP = cI − cII . Show that the generalization of of the Clausius-Clapeyron equation for the coexistence curve between the liquid and vapor phases when the mole fraction of of A in the liquid phase is ﬁxed is given by ∂P ∂T xA (sA − sl ) + xB (sB − sl ) B A .] Solution 3. (3.4). B. B and β = A. Problem 3.

What fraction of the steam condenses into water? (Treat the steam as an ideal gas and neglect the volume of the water. dT coex T v Here q is the latent heat per mole. we get ∂s ∂s dP ds = + . Since vg vl we get q dP = g. T ) and along the coexistence curve P is a function of the temperature. dP dT = coex ∆cP . dT ∂T ∂P dT We know that (∂S/∂T )P = cP /T and from the Maxwell relation (∂s/∂P)T = − (∂v/∂T )P . THE THERMODYNAMICS OF PHASE TRANSITIONS Solution 3.11: Let us use the Clausius-Clapeyron equation for the case of liquid-vapor mixture. For each phase. The temperature is then lowered to T2 = 336 K. q(T1 − T2 ) P1 = exp .11: Water has a latent heat of vaporization. . The relative number of condensed gas is then 18 ∆h (T1 − T2 ) n1 − n2 = 1 − exp − . n1 P1 q(T1 − T2 ) 18 ∆h (T1 − T2 ) = exp = exp . The answer is (∆cP )/T = (dP/dT ) ∆ (∂v/∂T )P = (dP/dT ) v (∆αP ) . Since ∆s = 0. At the same time. T ∆αP Problem 3. ∆h = 540 cal/gr.10: At a continuous phase transition the entropy is continuous. s = s(P. P2 RT1 T2 Since the volume is kept constant. g = RT1 T2 RT1 T2 n2 P2 Here we have taken into account that the olecular weight of H2 O is 18.) Solution 3. One mole of steam is kept at its condensation point under pressure at T1 = 373 K. ∆s = 0.40 CHAPTER 3. Considering vapor as an ideal gas we get dP qP = dT RT 2 As a consequence. n1 RT1 T2 g → P = P0 e−q/RT .

Indeed. (a) Perform the summations in the expression for Φ and write Φ in terms of η. (b) Compute the critical temperature Tc . C.8) can be expressed as 1 1 1 Φ = Φ0 + Aa2 η2 − Ba3 η3 + Ca4 η4 .12: A liquid crystal is composed of molecules which are elongated (and often have ﬂat segments).12: (a) First. ˆ x3 = z . (3. 1 a3 = Tr s − I ˆ 3 m 3 . B. A. 2 3 4 where A = A0 (T − T ∗ ).41 Problem 3. A0 . let us express the matrix S as η (s − I/3). (3.9) It is easy to calculate bm = Tr s − 1 I ˆ 3 1 Tr s − I ˆ 3 m . The free energy of the liquid crystal can be written as 1 1 1 Φ = Φ0 + A Si j Si j − B Si j S jk Ski + C Si j S jk Skl Skl . ˆ ˆ Si j = xi · S · x j . at which the transition from isotropic liquid to liquid crystal takes place. Solution 3. B and C are constants. which gives the average direction of alignment of the molecules. where ˆ 2 n1 n1 n2 n1 n3 s = nn = n2 n1 n2 n2 n3 . for brevity. = m! (−1)k ∑ k!(m − k)! 3k Tr sˆk Im−k . The order parameter of the liquid crystal is given by the diatic S = η [nn − (1/3)I] .8) ˆ and the summation is over repeated indices. It behaves like crystal because the orientation of the molecules does have long range order. 2 3 4 a2 1 = Tr s − I ˆ 3 2 . ˆ x2 = y . where n is a unit vector (called the director). 1 a4 = Tr s − I ˆ 3 2 2 . k=0 m . (c) Compute the difference between the entropies between the isotropic liquid (η = 0) and the liquid crystal at the critical temperature. I is the unit tensor so ˆ ˆ xi · I · x j = δi j . The quantities xi are the unit vectors ˆ x1 = x . ˆ 2 n3 n1 n3 n2 n2 3 The Eq. I behaves like a liquid because the locations of the center of mass of the molecules have no long-range order. (3.

3 27 9 (3.11) 3 27 9 In order to kill the term proportional to ξ3 one has to put (see the Maple ﬁle) η0 = B/6C. let us direct the axis 1 along the vector n. (3. Substituting this value into the expression for the coefﬁcient ξ. let us shift the origin of η by some η0 in order to kill odd in η − η0 items.10) 1 3 m +2 −1 3 m . a2 = b2 = 2 . where η± are the roots of the equation ∂Φ/∂η = 0 while ξ± are the roots of the equation ∂Φ/∂ξ = 0. we get 1 2 1 Φ = Φ0 + Aη2 − Bη3 + Cη4 . To ﬁnd this point.11). C 4 B2 2 δΦ(ξ) = ξ . (2/3)Aη0 − (2/9)Bη2 + (4/9)Cη3 .42 CHAPTER 3. Then 1 0 0 ˆ ˆ s = 0 0 0 → Tr sk = Tr s = 1 . thus ξ+ + ξ− = 0. 3 a3 = b3 = 2 . 9 a4 = b2 = 2 4 . It has the form. The important part Φ(ξ) at the critical point can be obtained substituting η0 and A into Eq. Here we have used the properties of trace and of s operators. (3. Since A = A0 (T − T ∗ ) we have Tc = T ∗ + (c) Since S = −∂Φ/∂T = −(1/3)A0 η2 we get ∆S = −(1/3)A0 (η+ − η− )(η+ + η− ) = −(2/3)A0 η0 (ξ+ − ξ− ) . ˆ 0 0 0 Consequently. 9 (b) At the transition point the dependence Φ(η) has two equal minima. THE THERMODYNAMICS OF PHASE TRANSITIONS It is straightforward that for m > 1 and k = 0 we get Tr sk Im−k = Tr sk I = Tr sk = 1 . It we put η = η0 + ξ we have 1 2 1 Φ − Φ0 = A(ξ + η0 )2 − B(ξ + η0 )3 + C(ξ + η0 )4 . Since trace is independent of ˆ the presentation. ˆ ˆ ˆ Tr Im = Tr I = 3 . ξ − 9 18C2 B2 . bm = 1 − Finally. 27A0C . Note that the potential Φ(ξ) is symmetric. 0 0 and equating the result to 0 we get A = B2 /27C.

13: (a) Let us express the pressure as a function of volume. ∆S = − 1 A0 B2 . and γ from the Berthelot equation. 8b RTc 2a + = 0.43 Thus.? • Find the critical exponents β. v = vc · v ¯ Tc = 8a . v−b Tv T . and R. 6C ξ+ − ξ− = B . b. 2 (vc − b) Tc v3 c 2RTc 6a = − = 0. ξ± = ± As a result. to get: 3 ¯ ¯ P + ¯ 2 (3v − 1) = 8T . (b) Does the Berthelot equation satisfy the law of corresponding states. Solution 3. Pc in terms of a. T v2 (a) Find values of the critical temperature Tc . δ. 27Rb Pc = RTc . (vc − b)3 Tc v4 c = − T a RT − 2. and the critical pressure. the critical molar volume vc . (b) We put ¯ P = Pc · P . 3C Problem 3. ¯ Tv ¯ ¯ T = Tc · T . 27 C2 B .13: The equation of state of a gas is given by the Berthelot equation P+ a (v − b) = RT . P= For the critical temperature we have ∂P ∂V ∂2 P ∂V 2 From this we obtain: vc = 3b .

¯ ¯ T = 1+τ and expanding in p. 3.1 A Properties of Van der Waals (VdW) liquid (i) Find the critical temperature Tc at which the Van der Waals isotherm has an inﬂection point.13) we observe that ∂p ∂ν = 24τ τ → γ = 1.12) 9 = − ν2 − 12τ . we get 3 p = − ν3 − 12ντ + 7τ . Vc . . the equation of state being ¯ ¯ 8T ρ 3ρ2 ¯ ¯ P(ρ. ν. Since along the coexistence curve p(νg . 2 τ νg νl (3.1 Mini-tests 3. (3. 2 The derivative (∂p/∂ν)τ is ∂p ∂ν νg νl (3. Substituting this value of νg into Eq. T ) = − ¯ . 3−ρ T ¯ ¯ Expanding near the critical point P = T = ρ = 1 we get ¯ P = 1 + (3/2)(ρ − 1)3 → δ = 3. for a system of N particles at T = Tc . Determine the pressure Pc and volume.44 CHAPTER 3. τ) = p(−νg . v = 1+ν. THE THERMODYNAMICS OF PHASE TRANSITIONS (c) First let us simplify the equation of state near the critical point.13) Since this function is even in ν the Maxwell construction νdp = ν ∂p ∂ν dν = 0 τ requires −νl = νg . To obtain the critical exponent δ let us introduce ρ = 1/v. τ) we get νg = √ −8τ → β = 1/2 . and τ up to lowest non-trivial order (see Maple ﬁle).1. Putting ¯ P = 1+ p.

(vii) Discuss why the stability condition and Maxwell relation lead to different stability criteria? Illustrate discussion using the plot. T = 1+τ V = 1−n and show that for small p. and Vc . τ) versus n for τ = ±0. Pc . (iii) Straightforward. Using this relation and the equation of state ﬁnd nl and nr . Show this region in the plot. (vi) Show that the Maxwell relation can be expressed as nr nl (3. V = VcV and T = Tc T . . τ and n the equation of state has the approximate form p = 4τ + 6τn + (3/2)n3 . Show that it has the form P+ 3 (3V − 1) = 8T V2 45 (3. 27 b2 (ii) The result is straightforward in one substitutes P = Pc P . 27 bk NkT N 2a − 2 V − Nb V = 0.14) where P ≡ P/Pc .15) n (∂p/∂n)τ dn = 0 (3. Solution (i) Writing the VdW equation as P= and requiring ∂P ∂V Tc = 8 a . Pc = 1 a .1.3.05 and discuss the plots. and T ≡ T /Tc (iii) Analyze the equation of state (3. (iv) Plot p(n.16) along the equilibrium liquid-gas line. (v) Using the above equation ﬁnd the stability region. Tc ∂2 P ∂V 2 =0 Tc we obtain Vc = 3Nb. V ≡ V /Vc .14) near the critical point. (iv) Straightforward. MINI-TESTS (ii) Express the VdW equation in units of Tc . Assume that P = 1 + p.

We have (∂p/∂n)τ = 6τ + (9/2)n2 . Consequently nst = ±2 −τ/3 . 3.15) we get (see the Maple ﬁle) √ nr = −4τ . ˜ ˜ From the graph of the function F(m) ≡ (βJ)−1 m − tanh m (see the Maple ﬁle) one observes ˜ ˜ ˜ that only trivial solution m = 0 exists at βJ < 11. Hint: express Eq. Hint: rewrite equation in terms of an auxiliary dimensional variable m ≡ βJm. (ii) Simplify Eq. nl = −nr . (i) Consider the case h = 0 and analyze graphically possible solutions of this equation. Solution (i) At h = 0 the equation (3. Show that ˜ a spontaneous magnetization appears at T < Tc = J/k.6. (3.17) at h = 0 near the critical temperature and analyze the spontaneous magnetization as a function of temperature.6 and T /Tc = 1.17) where β = 1/kT . ˜ ˜ (iv) Plot and analyze the magnetization curves for T /Tc = 0. THE THERMODYNAMICS OF PHASE TRANSITIONS (v) The stability region is determined by the equation (∂p/∂n)τ = 0. Using Eq.1. At T < Tc there exists ˜ 2 non-trivial solutions corresponding spontaneous magnetization.17) in terms of the ˜ ˜ variables m and h ≡ βh ≈ h/J and plot h as a function of m.2 B Curie-Weiss theory of a magnet The simplest equation for a non-ideal magnet has the form m = tanh[β(Jm + h)] (3. Hint: Put T = Tc (1 + τ) and consider solutions for small m and τ. (3. or at T ≥ Tc = J/k. (3. while h is the magnetic ﬁeld measured in proper units. ˜ (iii) Analyze the magnetization curve m(h) near Tc . (vi) Since in the main approximation (∂p/∂n)τ is an odd function of n. Thus (βJ)−1 = T /Tc . J is the effective interaction constant.17) has the form: (βJ)−1 m = tanh m .46 CHAPTER 3.

→ m + h = artanh [(1 + τ)m] .1.17) can be rewritten as (T /Tc )m = tanh(m + h) ˜ ˜ ˜ Consequently. near Tc acquires the form F(m) = ˜ ˜ ˜ ˜ 3 /3. the same result is true for m. (iii) In general.3. ± −3τ . ˜ ˜ ˜ ˜ h = artanh [(1 + τ)m] − m ˜ ˜ To plot the magnetization curve one has to come back to the initial variables. Eq. (3. . ˜ Since near Tc the ratio Tc /T can be substituted by 1. Thus the solutions have the from τm + m ˜ ˜ √ m = 0. MINI-TESTS 47 (ii) The dimensionless function F(m) ≡ (T /Tc )m − tanh m.

THE THERMODYNAMICS OF PHASE TRANSITIONS .48 CHAPTER 3.

They can be occupied by 7 − 3 − 2 = 2 nice people in 12 · 11 ways.1: A bus has 9 seats facing forward and 8 seats facing backward. we get 9 · 8 · 7 × 8 · 7 × 12 · 11 = 3725568 ways. In how many ways can 7 passengers be seated if 2 refuse to ride facing forward and 3 refuse to ride facing backward? Solution 4. Problem 4. Let us start with the situation where room A has 3 persons.1: Three people refuse ride facing backward. The number of ways to do that is 9 · 8 · 7. 49 .2: The number of persons in the room is between 3 and 5. Now 17 − 5 = 12 seats left. The number of ways to do that is 8 · 7 · 6. In a similar way. and they should be deﬁnitely placed to 9 seats facing forward.Chapter 4 Elementary Probability Theory and Limit Theorems Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Problem 4. As a result. The total number of ways is then 8 · 7 · 6 + 8 · 7 · 6 · 5 + 8 · 7 · 6 · 5 · 4 = 8736 .2: Find the number of ways in which 8 persons can be assigned to 2 rooms (A and B) if each room must have at least 3 persons in it. Solution 4. 2 people can occupy 8 seats facing backward in 8 · 7 ways.

Thus we have 11!/(3!4!4!) = 11550 ways. Thus the number of ways is 7!/2! = 2520. 4 blue balls. ELEMENTARY PROBABILITY THEORY . (d) Show that the event “only two coins heads” and the event “three coin heads” are dependent and mutually exclusive. Thus the total number of permutations is 10!/(4!2!) = 75600.50 CHAPTER 4. . Since the probability to get no heads is 1/8.4: In how many ways can 5 red balls. but 5 blue balls to permute. . . (c) We assume that result of coin tossing is fully random. Problem 4. the answer is 1 − 1/8 = 7/8. The rest 5 + 4 + 4 − 2 = 11 balls are in the middle. The same is true for blue balls.3: The total number of the letters is 10. (d) The event “3 heads” contradicts the requirement “only 2”. In how many ways are 4 O’s together? In how many ways are (only) 3 O’s together? Solution 4. Now let us do the same with white balls.5: Three coins are tossed. but permutation of the balls of the color must be excluded. and 4 white balls be placed in a row so that the balls at the ends of the row are of the same color? Solution 4.5: (a) The answer is (1/2)3 = 1/8. thus the number of ways is 6 × 2520 = 15120. (b) The sum of the probabilities is 1. (b) Find the probability of getting at least one head.4: Let us start with 5 red balls and choose 2 of them to be at the ends. Solution 4. Among them the have 3 rest red balls. We get 11!/(2!5!4!) = 6930 ways. The total number of ways is 138600 + 2 × 83160 = 25410 ways. Now we have 2 rest white balls. To have three O’s together we have 6 places to insert a separate O to the word MONTNUS. we have 4 O’s. Problem 4. To get the number of ways to have 4 O’s together we have to use model word MONTNUS which has 7 letters with 2 repeated N’s. They have to be permuted. (c) Show that the event “heads on the ﬁrst coin” and the events “tails on the last coin” are independent. 2 N’s. Problem 4. and other letters do not repeat. (a) Find the probability of getting no heads. 4 blue balls and 4 white balls.3: Find the number of permutations of the letters in the word MONOTONOUS.

Solution 4.3 j=2.Y (x. Thus.51 Problem 4.8: A stochastic variable X can have values x = 1 and x = 3. Thus We know that 15 − 6 = 9 boys had problems. For each case decide if X and Y are independent. the number of permutations 4 · 4!/3! = 16.6: Since the number must be even.4 = 0 and p1. Given that the number is even. (b) p1. what is the probability that two ﬁves are together? Hint: You must ﬁnd a conditional probability. Consequently. All arrangements are equivalently likely. 5 get lost. 8 get sunburned. Problem 4. it is the same as permute 4 units. Denote the joint probability density PX. conditional probabilities are P(A|B) = P(A ∩ B)/P(A) = 1/2 (a).B. A stochastic variable Y can have values y = 2 and y = 4. 176-177 of the book [1].2 = p3. y) = i=1. We have to divide it by 3! of permutations of 6 and 2! permutations between 5. and the total number of their permutations give 5!. ∑ ∑ pi.4 = 1/4. Fifteen boys go hiking. pp.2 = p1. Compute the covariance of X and Y for the following 2 cases: (a) p1. . Thus the probability to be both sunburned and lost is P(A ∩ B) = P(A) + P(B) − P(A ∪ B) = (8 + 5 − 9)/15 = 4/15 . P(A ∪ B) = 9/15. Thus the probability is P = 16/40 = 2/5 .6: Various 6 digit numbers can be formed by permuting the digits 666655.4 = p3. The probability to get sunburned is P(A) = 8/15. the last number is 6. respectively.7: problems. Problem 4. There are 5 digits left. We end at the number of different numbers is 4 · 5!/3!2! = 40.4 P(B|A) = P(A ∩ B)/P(B) = 4/5 (b) .2 = 1/2. If we want to have two ﬁves together. j δ(x − i) δ(y − j) .7: See. while the probability to get lost is P(B) = 5/15 = 1/3. and the event to get lost .2 = p3.4 = p3. Let us deﬁne the event to get sunburned as A. and there are 4 ways to choose this number. and 6 return home without (a) What is the probability that a sunburned boy got lost? (b) What is the probability that a lost boy got sunburned? Solution 4. the joint probability to be either sunburned or lost is.

the characteristic function is fZ (k) = eikz = 1 2 ∞ 0 dze(ik−1/2)z = 1 . Let us do the problem in two ways.9: The stochastic variables X and Y are independent and Gaussian distributed with ﬁrst moment x = y = 0 and standard deviation σX = σY = 1. 2π 2π 1 2π ρ dρ dφ e−ρ 2 /2 dx dy = ρ dρ dφ. We have P(x)P(y) = Thus.52 Solution 4. z2 and z3 . It is not equal to X · Y . XY = (1 · 4 + 3 · 2) /2 = 5 . 2 Now. y → ρ. Find the ﬁrst 3 cumulants. . (b) We have. PZ (z) = General way: PZ (z) = = = dx PX (x) √ z 1 −(x2 +y2 ) 1 −ρ2 e = e . φ. XY = (1 · 2 + 1 · 4 + 3 · 2 + 3 · 4) /4 = 6 . δ(z − ρ2 ) = 1 −z/2 e . so the variables are dependent. 2 dy PY (y) δ(x2 + y2 − z) PX (x) ∑ PY ± z − x2 ± √ √ − z 2 z − x2 √ z 1 dx dx π 0 x2 z − x2 √ exp − − 2 2 z − x2 = 1 −z/2 e . . .8: (a) First let us calculate CHAPTER 4. and compute the moments z . Since XY = X · Y the variables are independent. Problem 4.9: First let us ﬁnd the distribution function for Z. Find the characteristic function for the random variable Z = X 2 +Y 2 . ELEMENTARY PROBABILITY THEORY . Solution 4. Simple way: Let us transform the problem to the cylindrical coordinates. x. At the same time X = 2 and Y = 3. 1 − 2ik .

53 Then,

d n fZ (k) dkn Performing calculations (see the Maple ﬁle), we obtain: zn = (−i)n z = 2, Thus, C1 (Z) = C3 (Z) = z2 = 8 ,

.

k→0

z3 = 48 .

z = 2, , C2 (Z) = z2 − z 2 = 4 , z3 − 3 z z2 + 2 z 3 = 48 − 3 · 2 · 8 + 3 · 23 = 24 .

Problem 4.10:

A die is loaded so that even numbers occur 3 times as often as odd numbers.

(a) If the die is thrown N = 12 times, what is the probability that odd numbers occur 3 times? If is is thrown N = 120 times, what is the probability that odd numbers occur 30 times? Use the binomial distribution. (b) Compute the same quantities as in part (a), but use the Gaussian distribution Note: For parts (a) and (b) compute your answers to four places. (c) Compare answers (a) and (b). Plot the binomial and Gaussian distributions for the case N = 12. Solution 4.10: (a) For the random process, the parial probability for an odd number p = 1/4 and for an even number is q = 1 − p = 3/4. The probability to have k odd numbers, according to binimial distribution is N! b PN (k) = pk qN−k . k!(N − k)! Consequently (see Maple ﬁle),

b P12 (3) = b In a similar way, P120 (30) = 0.08385.

12! 39 = 0.2881 . 3!9! 412

**(b) The number of odd numbers is k = N/4 and the dispersion is σ = N p · q = 3N/16. Thus
**

G PN (k) =

1 (k − N/4)2 exp − 3N/8 3Nπ/8

G P120 (30) = 0.0841 .

.

As a result,

G P12 (3) = 0.2660 ,

(c) Plots are shown in the Maple ﬁle.

54 Problem 4.11:

**CHAPTER 4. ELEMENTARY PROBABILITY THEORY . . .
**

A book with 700 misprints contains 1400 pages.

(a) What is the probability that one page contains no mistakes? (b) What is the probability that one pages contains 2 mistakes?

Solution 4.11:

n mistakes can be distributed between g pages by wg (n) = (g + n − 1)! . (g − 1)!n!

If one has a page without mistakes, it means that n mistakes are distributed between g − 1 pages. The probabiliy of that event is P(0) = (g + n − 2)! n!(g − 1)! g−1 = . (g − 2)!n! (g + n − 1)! g + n − 1

In our case g = 1400 and n = 700, thus P(0) = 1399/2099 ≈ 0.67. In a similar way, if we have two misprints at a page, then n − 2 misprints are distributed between g − 1 pages. Thus, the number of ways to do that is wg−1 (n − 2). Consequently, P(2) = wg−1 (n − 2) (g + n − 4)! n!(g − 1)! (g − 1)n(n − 1) = = . wg (n) (g − 2)!(n − 2)! (g + n − 1)! (g + n − 1)(g + n − 2)(g + n − 3)

We get P(2) ≈ 0.07.

Problem 4.12: There old batteries and a resistor, R, are used to construct a circuit. Each battery has a probability p to generate voltage V = v0 and has a probability 1 − p to generate voltage V = 0 Neglect internal resistance in the batteries. Find the average power, V 2 /R, dissipated in the resistor if (a) the batteries are connected in the series and (b) the batteries are connected in parallel. In cases (a) and (b), what would be the average power dissipated in all batteries were certain to generate voltage v0 . (c) How you realize the conditions and results in this problem in the laboratory?

55 Solution 4.12: Let us start with series connection and ﬁnd the probability to ﬁnd a given value of the voltage. We have 4 possible values of the voltage. In units v0 they are 0, 1, 2, and 3. We have, P(0) = (1 − p)3 , Consequently, W = V 2 /R = ∗(v2 /R) 3p(1 − p)2 + 12p2 (1 − p) + 9p3 = (3v2 /R) p (1 + 2p) . 0 0 Now let us turn to parallel connection. Let us for a while introduce conductance Gi = 1/Ri of the batteries, as well as conductance G = 1/R of the resistor. The voltage at the resistor is given by the expression ∑i=1,2,3 Ei Gi . V= G + ∑i=1,2,3 Gi Here Ei are electro-motive forces of the batteries. Now the crucial point is to assume something about the resistances of the batteries. Let us for simplicity assume that they are equal, and Gi G. Actually, it is a very bad assumption for old batteries. Under this assumption, V = (E1 + E2 + E3 )/3. The the result is just 9 times less that that for series connection. To check this kind of phenomena in the lab one has to ﬁnd many old batteries at a junk-yard, ﬁrst check distribution of their voltages and then make circuits. The problem is badly formulated. Problem 4.13: Consider a random walk in one dimension. In a single step the probability of displacement between x and x + dx is given by 1 (x − a)2 P(x) dx = √ exp − 2σ2 2πσ2 dx . (4.1) P(1) = 3p(1 − p)2 , P(2) = 3p2 (1 − p) , P(3) = p3 .

After N steps the displacement of the walker is S = X1 + X3 + XN where Xi is the displacement after ith step. After N steps, (a) what is the probability density for the displacement, S, of the walker, and (b) what is his standard deviation? Solution 4.13: (a) By deﬁnition, PS (s) = dx1 P(x1 ) dx2 P(x2 ) . . . dxN P(xN ) δ (s − x1 − x2 . . . − xN ) .

**To calculate this integral we expand δ-function into Fourier integral δ(x) = dk ikx e . −∞ 2π
**

∞

. After N steps the displacement of the walker is S = X1 + X3 + · · · + XN where Xi is the displacement after ith step. S = ds sPS (s) = Na . S2 = dx1 P(x1 ) · · · dxN P(xN ) (x1 + x2 + .14: Consider a random walk in one dimension for which the walker at each step is equally likely to take a step with displacement anywhere in the interval d − a ≤ x ≤ d + a. Each step is independent of others. . S2 − S 2 = Nσ2 . Problem 4. PS (s) = → [P (k)]N = exp −iNka − k2 Nσ2 /2 . xN )2 .1) we obtain P (k) = exp −ika − k2 σ2 /2 Now.14: S = = In a similar way. . (a) what is the average displacement. From Eq. dk iks e −∞ 2π ∞ dk eiks −∞ 2π ∞ dx1 e−ikx1 P(x1 ) ∞ −∞ dx1 e−ikx2 P(x2 ) . . After N steps. (b) We have (see Maple ﬁle). . . . − xN ) dxN P(xN ) (x1 + x2 + . −∞ 2π ∞ P (k) = dx e−ikx P(x) is the Fourier component of the single-step probability. (4. Starting from the deﬁnition. we have s PS (s) ds = dx1 P(x1 ) · · · s ds dx1 P(x1 ) · · · dxN P(xN ) δ (s − x1 − . where a < d.56 In this way we get PS (s) = = Here CHAPTER 4. . . . of the walker. N dxN e−ikxN P(xN ) (4. S .2) dx e −ikx P(x) ∞ −∞ = dk iks e [P (k)]N . dk (s − Na)2 1 exp − exp ik(s − Na) − k2 Nσ2 /2 = √ 2Nσ2 −∞ 2π 2πNσ2 ∞ . ELEMENTARY PROBABILITY THEORY . and (b) what is his standard deviation? Solution 4. xN ) = N X = Nd . .

Does it satisfy the Central Limit Theorem? Should it? Solution 4. 1 2a 1 2a 1. It should not because X 2 is divergent. otherwise d+a d−a x2 dx = d 2 + a2 /3 .57 We know that 2 ∑ xi i 2 = ∑ xi + ∑ xi x j . d − a ≤ x ≤ d + a .15: Consider a random walk in one dimension. After averaging the obtain S2 = N X 2 + N(N − 1) X To calculate X 2 let us specify the normalized probability as P(x) = Then. 0. . i i= j The number of pairs with non-equal i and j is N(N − 1).15: Using the method of the Problem 4. X2 = Summarizing we get (see Maple ﬁle) S2 − S 2 2 .3) Find the probability density for the displacement of the walker after N steps. 2π π Na2 + s2 It does not satisfy the Central Limit Theorem. Problem 4. 2 + a2 πx (4. = N(d 2 + a2 /3) + N(N − 1)d 2 − (Nd)2 = Na2 /3 .13 we get P (k) = e−a|k| → PS (s) = 2 ∞ 0 dk −Nka 1 Na2 e cos ks = . In a single step the probability of displacement between x and x + dx is given by P(x) dx = a 1 dx .

ELEMENTARY PROBABILITY THEORY .58 CHAPTER 4. . . .

for the same initial conditions as in part (b). and the autocorrelation function. What is the probability that there are 2 red marbles in urn A after 2 steps? After many steps? (c) Assume that the realization.1: To the problem 5.1 We denote these 3 realizations as y(n) where n = 1. 59 . and urn B initially has 1 white and 3 red marbles.Chapter 5 Stochastic Dynamics and Brownian Motion Quick access: 1 2 3 4 5 6 7 8 9 10 11 Test 1 Problem 5. P(s)|. 5. The marbles are repeatedly interchanged. In each step of the process one marble is selected from each urn at random and the two marbles selected are interchanged. given the initial condition stated above.1 1 2 3 Figure 5. at time s. 2 and 3.1: Urn A has initially 1 red and 1 white marble. (b) Compute the probability vector. equals to n2 . y(s) . y(0)y(s) . Let the stochastic variable Y denote “conﬁguration of the units”. There conﬁgurations are possible. They are shown in Fig. y(n). Compute the moment. (a) Compute the transition matrix Q and the conditional probability matrix P(s0 |s).

1) . Now we construct matrices Pi = |i i| to get 1 8 6 6 −12 6 3 −3/2 . 1. By inspection of the ﬁgure we get. 2 4 8 2 4 2 4 2 2 4 8 1 1 Q31 = 0 . −2. The corresponding “right” eigenvectors are: [1. 1 1 1 1 1 1 3 1 1 3 3 Q21 = · = . 1.1. {[1. [1. We have for them the Maple result 1 −1 [ . 0 1/2 1/2 while the initial condition corresponds to the probability vector p(0)| = (0. 8. λ2 = −1/4. P1 = 1 8 6 .1 of the book [1]. P3 = 1 −2 −4 6 2| = (1. 1. 0) . 1) . 1]}]. { 6. . where n1 = 15. −3]}]. 1) . It is easy to prove that i|k = ni δik . STOCHASTIC DYNAMICS . −2. λ3 = 1/4. The “left” eigenvectors are just the right eigenvalues of the transposed matrix. 1]}]. First we ﬁnd eigenvalues of Q which are equal to 1. 1. [ . −2. Q13 = 0 . 1. The following procedure is clear form the Maple printout. Q33 = . 6) . 1| = (1. Q23 = · = . QT .. 1.. |2 = (6. 6]}] 4 4 Thus. {[1.60 Solution 5. 1 }]. 8. Q22 = · + · = . n2 = 10. 3| = (1. 1. 4 8 −12 2 −3 . 2. Below we proceed as in exercise 5. {[1. 1. 5. {[4. 1) . Q32 = . 2. The conﬁgurations are shown in Fig. Q12 = 1 . n3 = 12 . −3/2.1: CHAPTER 5. P2 = −3/2 1 8 6 1 −2 1 . −3) . −2]}] 4 2 4 Thus we denote λ1 = 1. {[1. and |1 = (1. |3 = (1. Q11 = 0 . 1/4. 2 2 Thus the transition matrix Q is given by the expression 0 1 0 Q = 1/8 1/2 3/8 . [ . [ −3 1 −1 . 1. 1.

what is the chance it has it after 2 throws? What is the chance he will have in after s throws. (a) Compute the transition matrix. We obtain y(s) y(0)y(s) 29 3 −1 3 1 − ( )s − ( )s . 15 20 4 12 4 15 10 4 6 4 5 20 4 4 4 We are asked to ﬁnd the probability of the conﬁguration 3. If “heads” comes up. and C. The coin of boy A is “fair” (50% heads and 50% tails). approximately what fraction of time each boy have the ball (assuming they throw the ball many times)? (c) If boy A has the ball to begin with. A. Straightforward calculation gives P(s)| = = p(0)|P(s) 1 3 −1 1 1 8 3 −1 1 1 2 3 −1 1 1 − ( )s + ( )s . − ( )s − ( )s . + ( )s + ( )s . and its left and right eigenvalues. After 2 steps it is 3/8. each boy ﬂips a coin to decide whether to throw to the boy on his right or left.B.61 Now. and at s → ∞ we get 2/5. its eigenvalues. the boy throws to his right. If “tails” come up. The part (c) is calculated in a straightforward way. Before throwing the ball. the boy throws to his right. 5 10 4 2 4 79 33 81 = (−1)(1+s) 4(−s) − 16(−s) + (−1)(1+2 s) 16(−s) 50 4 100 11 3 841 + (−1)(1+s) 16(−s) − 4(−s) + 2 2 25 = Problem 5. (b) If the ball is thrown in regular intervals. .2: Three boys. Now we come to the point (b). In this way we solve the part (a). P(0) = I. the coin of boy B has head of both sides. stand in a circle and play catch (B stands to the right of A). and the coin of boy C is weighted (75% heads and 25% tails). P(s0 |s) = P(s − s0 ) with P(s) = ∑ = λs i Pi ni i 1 3 −1 1 1 8 6 −1 2 1 + ( )s + ( )s − ( )s + ( )s 15 5 4 3 4 15 5 4 3 4 1 3 −1 1 1 8 3 −1 1 1 − ( )s + ( )s + ( )s + ( )s 15 20 4 12 4 15 10 4 6 4 1 1 −1 1 1 8 1 −1 1 1 + ( )s − ( )s − ( )s − ( )s 15 10 4 6 4 15 5 4 3 4 2 3 −1 s 1 + ( ) − ( )s 5 5 4 4 2 3 −1 1 1 − ( )s − ( )s 5 20 4 4 4 2 1 −1 1 1 + ( )s + ( )s 5 10 4 2 4 As it should be.

Q31 = 3/4 . but oscillates in time). but conjugated. Each time it rings. B (2) A (1) C (3) Figure 5. As a result. Q= 0 3/4 1/4 0 The following solution is similar to the previous problem (see the Maple printout). STOCHASTIC DYNAMICS . starting from the given initial conditions his chance to hold the ball after 2 throws is 3/8. Then Q12 = Q13 = 1/2 . At the same time O11 = Q22 = Q33 = 0. When he changes rooms. C → 3. “Mouse in room A” 2. Note that the eigenvalues and eigenvectors are complex. “Mouse in room B” 3. Q21 = 0 . It means that the probability is real. Let us map the notations A → 1. “Mouse in room C” .62 Solution 5. 5. he is equally likely to pass through any of the doors of the room he is in.2: A bell rings at regular intervals (short compared with the mouses lifetime).2. Problem 5. Let the stochastic variable Y denotes “mouse in a particular room”. Q32 = 1/4 . The problem is similar to the previous one. the transition matrix has the form 0 1/2 1/2 0 1 . the mouse changes rooms.3: A trained mouse lives in the house shown in Fig. B → 2. and note that “right” means counterclockwise. The answers are: after many throws the chance for the boy A to hold the ball is 6/19.2: CHAPTER 5. There 3 realizations of Y: 1.. Q23 = 1 ..

The answers are obvious from the printout. P(s)|. Compute the probability vector. (a) If initially the mouse is in room A. and y(3). Q21 (s) = 1/4 + (1/2) sin2 (πs/2) . We have Q21 = 1/3 . y(n) equals to n. for the same initial condition as in part (b). y(0)y(s) . Q23 = Q32 = 2/3 .4: The doors in the mouse’s house shown in Fig. This causes the mouse’s transition between the rooms to become time periodic. Let the stochastic variable Y have the same meaning as in Problem 5. what is the probability to ﬁnd it in room A after 2s room changes? In room B? Solution 5. respectively. Qii = 0 . Q12 = Q13 = 1/2 . and the autocorrelation function. Problem 5. Q12 (s) = cos2 (πs/2) . y(2). . (a) compute the transition matrix. (b) . The transition matrix is now given by Q11 (s) = Q22 (s) = Q33 (s) = 0 . Compute the moment.3: The problem is very much similar to the previous ones. 1/3 2/3 0 Now we do the same as in previous problems (see Maple printout).3. at time s.2 so thay get periodically larger and smaller. and the conditional probability matrix.4: The solution is rather tedious. Solution 5. y(s) . Approximately. Q. P(s0 |s). what is the probability to ﬁnd it in room A after 2s room changes? (b) If initially the mouse is in room B. Q23 (s) = 1/4 + (1/2) cos2 (πs/2) Q31 (s) = (1/2) cos2 (πs/2) . given that the mouse starts in room C. Thus 0 1/2 1/2 Q = 1/3 0 2/3 .63 which we denote as y(1). 5. what fraction of his life does he spend in each room? (c) Assume that the realization. Q31 = 1/3 . Q32 (s) = (1/2) + (1/2) sin2 (πs/2) . Q12 (s) = sin2 (πs/2) . but obvious from the Maple ﬁle.

Assume that the walker is equally likely to move on the lattice site to the left or right at each step. If we apply cyclic boundary conditions (assuming that the word periodic in the problem formulation means that) we have to require: P(n + 2N + 1) = P(n) at any s. s) = k=−N ∑ N Pk exp −in 2πk πk − 2s sin2 . of the requirement P(n.n−1 = wn+1. The Master equation has the form dPn /ds = wn−1. Thus.n+1 ) = (1/2) (Pn−1 + Pn+1 − 2Pn ) . 0) = δn.t] = P(n. (c) If the lattice has 5 sites (N = 2). given the walker starts at site n = 0. h. Treat this problem as a Markov chain. the general solution is P(n. Solution 5. At s = 0 we can rewrite the l.n = wn. Now we have to ﬁnd κ. where k is the integer number ranging from −N to N.. Thus eiκ(2N+1) = 1 → κ(2N + 1) = 2πk . compute the probability to ﬁnd the walker on each site after s = 2 steps and after s = ∞ steps. Q. (b) Compute the probability P1 (n. Problem 5. Let us look for the solution in the form Pn (s) ∝ eλs+iκn .n−1 + wn. e. .n Pn−1 + wn+1.t) . P(s0 |s).Assume that the walker starts at site n = 0. and the conditional probability.64 CHAPTER 5. s) at time s.n = wn. possessing the property P[n + (2N + 1). (a) Compute the transition matrix. Hint: Let us understand the word periodic as peridically extended.0 in the form P(n. s. i. 0) = k=−N ∑ N Pk e−in2πk/(2N+1) .5: Consider a discrete random walk on a one-dimensional periodic lattice with 2N + 1 sites (label the sites from −N to N).n Pn+1 − Pn (wn. Then we get λ = −[1 − cos(κ)] = −2 sin2 (κ/2) .. 2N + 1 2N + 1 The quantities Pk should be found from the initial conditions.n+1 = 1/2 . STOCHASTIC DYNAMICS .5: The transition probabilities are given by the expressions wn−1.

of decaying. or P(n.0 = N 1 ∑ e−in2πk/(2N+1) . As s → ∞ the probability tends to 1/(2N + 1). the matrix Qmn = P(1|0)mn . δn. λ. Consequently. Finally. s|m. The proper constant 1/N (s) is the normalization coefﬁcient found from the condition n=−N ∑ N P(n. The calculation for N = 2 is straightforward. Assume that at time t = 0 there are n0 undecayed nuclei present.6: At time t.65 On the other hand. Now we have to apply the initial condition P(n. 2N + 1 k=−N From that we immediately get Pk = const(k). 0) = Thus. s|m. s) = 1 . 0) = N 1 2πk πk ∑ exp −i(n − m) 2N + 1 − 2s sin2 2N + 1 . each with the probability for the unit time. (c) What is the half-life of this decay process? . 2N + 1 k=−N k=−N ∑ N Pk e−in2πk/N = δn. 0) = δm. (b) Compute the the mean number of undecayed nuclei and the variance as a function of time. see Maple ﬁle. (a) Write down and solve the master equation of the process [ﬁnd P1 (n.m = k=−N ∑ N Pk e−i(n−m)2πk/(2N+1) . a radioactive sample contains n identical undecayed nuclei.t)].0 . s) = Here N 1 πk 2πk ∑ exp −in 2N + 1 − 2s sin2 2N + 1 . N (s) k=−N N N (s) = n=−N ∑ exp −in 2πk πk − 2s sin2 = 2N + 1 . Finally P(n. 2N + 1 2N + 1 Now let us ﬁnd the conditional probability P(n. 0). P(s − s0 |0)mn = P(n. The probability of decay during the time t → t + ∆t is λ ∆t. Pk =∝ eim2πk/(2N+1) . Problem 5.

dz z−1 → ∂F (5. z=1 = n0 e−τ 1 − e−τ . 0) = δn. 262 of the book [1]. We introduce F(z. ∂t ∂z Indeed. .t) = Φ e−τ+ln(z−1) = Φ (z − 1)e−τ ..1 in the form ∂F ∂F + (z − 1) = 0. n2 (t) − n(t) 2 → Φ(u) = (u + 1)n0 .t) = z ∑ zn−1nP(n.n0 we get Φ(z − 1) = zn0 As result. and we solve it by the method of generating functional.. Now. F(z. and sum over n. n0 F(z. We get (5. The master equation has the form dP = −λn P(n.t) . n(t) = In a similar way. 5.66 Solution 5. dt This is a linear master equation.2) ln(z − 1) − τ = const . ∂τ ∂z The characteristics of this equation meet the equation dτ = Consequently. Since P(n.1) ∂F . STOCHASTIC DYNAMICS . = n0 e−τ .t) = Then we multiply the master equation by zn n=−∞ ∑ ∞ zn P1 (n.t) = z ∂z Introducing τ = λt we rewrite Eq.t) = n n d In a similar way n n ∑ znnP(n.t) = (z − 1)e−τ + 1 ∂F ∂z .t) + λ(n + 1) P(n + 1.t) .6: CHAPTER 5. ∂z ∂F ∂F = −λ(z − 1) . p. ∑ zn(n + 1)P(n + 1.t) = dz ∑ zn+1P(n + 1.

M. w21 = w23 = w24 = 1/3.4. (b) In the walker starts at site n = 1 at time t = 0. V. The transition rates are w12 = w13 = 1/2 w21 = w23 = w2P = 1/3. (a) Write the transition matrix. w12 = w13 = 1/2 . and ﬁnd its eigenvalues and eigenvectors. w31 = w32 = w3P = 1/3. W. and compute its eigenvalues and eigenvectors. 5.1 .t) for the case P(n.7: Consider a random walk on the lattice shown in Fig. and show that that this system obey detailed balance.67 To ﬁnd the probabilities let us express the quantity (z − 1)e−τ + 1 Thus. (c) Write P1 (n. 0) = δn. The transition rates are 1 2 3 4 Figure 5. The site P absorbs the walker. w42 = w43 = 1/2.7. Problem 5. The answers are obvious from the Maple ﬁle. w14 = w41 = 0 . τ) = n0 = e−n0 τ [z + (eτ − 1)]n0 = e−n0 τ n=0 ∑ n!(n0 − n)! zn (eτ − 1)n0−n . compute the mean ﬁrst passage time. P1 (n. Problem 5.8: Consider a random walk on the lattice shown in Fig. w31 = w32 = w34 = 1/3.3: A sketch to the problem 5.7: The solution is similar to the Exercise 5.3. (a) Write the transition matrix. wP1 = wP2 = 1/2 .2 in p.t)? Solution 5. . (b) Compute the symmetric matrix. 5. What is P(4. n0 −n n0 n0 ! n0 ! e−n0 τ 1 − e−τ n!(n0 − n)! . 245 of the book [1].

exists whose strength is determined by the temperature. ξ(t2 )ξ(t1 ) = g δ(t2 − t1 ) and ξ(t) ξ = 0. Solution 5. θ ds es−θ η(s) .4 of the book [1]. Multiplying solutions for j(θ2 ) and for j(θ1 ) and averaging over realizations of η we get j(θ2 ) j(θ1 ) η θ2 θ1 0 + 0 ds2 1 2 ds1 δ(s2 − s1 )es1 +s2 −θ1 −θ2 = 2 j0 − 1 e−θ1 −θ2 + e−|θ2 −θ1 | . and inductance. compute g and the current correlation function I(t2 )I(t1 ) ξ 2 the average magnetic energy L I0 T /2 = kT /2 and I0 T = 0 where I(0) = I0 . induces the ﬂuctuating current. 2 . and j(t) = I(t)/I1 .9: Let us consider on RL electric circuit with resistance. Namely. Problem 5. R. Even there is no average electromotive force (EMF) exists across the resistor. L. The answers are clear from the Maple ﬁle. This.3) . The Langevin equation for the current is L If the EMF is delta-correlated. deﬁne θ = t/τ.3) by R. Its solution has the form j(θ) = j0 e−θ + 2 = j0 e−(θ2 +θ1 ) 0 η(θ2 )η(θ1 ) = δ(θ − θ1 ) . ξ(t). T . in the circuit. Then the equation for the current has the form d j/dθ + j = η(θ) .. in turn. connected in series. (5. STOCHASTIC DYNAMICS . denote L/R ≡ τ and measure time in units of τ and √ the current in units I1 ≡ R−1 g. I(t). a ﬂuctuating EMF. dt (5. Assume that Solution 5.9: Let us divide Eq.68 1 CHAPTER 5.4: A sketch to the problem 5. 2 3 P Figure 5. T d I(t) + R I(t) = ξ(t) . because of the discrete character of electrons in the circuit and their random motion..8: The solution is similar to the Exercise 5.8.

T where Q(0) = Q0 . Q0 T . it must depend only on the difference θ2 − θ1 . i. 2 L I0 = kT → g = 2R2 kT /L . ξ(t). Q0 I0 ξ T. j(θ2 ) j(θ1 ) In dimensional variables. Thus. 2R2 On the other hand. This. ξ(t2 )ξ(t1 ) = g δ(t2 − t1 ) and ξ(t) ξ = 0. on the capacitor plates and a random current. induces a random varying charge. through the resistor and inductor. 2 j0 T = 1 2 → 2 I0 T = g . (Q(t) − Q0 )2 . The random charge. ξ T. Assume that I0 T = 0. Q(t). e. satisﬁes the Langevin equation L d2Q dQ Q +R + = ξ(t) .10: Due to the random motion and discrete nature of electrons. dt 2 dt C Assume that EMF is delta-correlated. I)t) = dQ(t)/dt. 1 L I2 2 0 T = T 1 Q2 2C 0 = 0. I(t2 )I(t1 ) ξ T 2 = I1 j(θ2 ) j(θ1 ) η T η T = = kT −|t2 −t1 |/τ e . in turn. 1 −|θ2 −θ1 | e . I(0) = I0 . (a) Compute the current correlation function I(t2 )I(t1 ) (b) Compute the variance of the charge distribution. = 0. 2 Consequently.. Q(t). L Problem 5.69 The ﬁrst item must vanish since in the equilibrium the process must be stationary. an LRC series circuit experience a random electromotive force (EMF). Assume that the circuit is at the temperature T and the average magnetic energy in the inductor and average electric energy in the capacitor satisfy the equipartition theorem.

δ= Γ 2 − ω2 .11: Consider a Brownian particle of mass m in one dimension in the presence of a constant force f0 in a ﬂuid with force constant γ and in th4e presence of a delta-correlated random force ξ(t) such that ξ(t1 )ξ(t2 ) ξ = g δ(t1 −t2 ) and ξ(t) ξ = 0.5 of the book [1].70 Solution 5. (b) Compute the displacement variance (x(t) − x0 )2 . Problem 5. I(t + τ)I(t) where C(τ) = cosh δτ − (Γ/δ) sinh δτ . L → m.10: mapping CHAPTER 5. Thus. 0 R→γ we obtain the same equation. . I0 t 0 t = kT /L.. and then apply δ function. 0 The variance of charge distribution can be calculated by integrating the formal solution I(t) = I0 e−Γt − Q0 over time: Q(t) = Q0 + I0 + 1 L t ξ T = kT −Γ|τ| e C(|τ|) . L ω2 −Γt 1 0 e sinh δt + δ L t 0 dt ξ(t )e−Γ(t−t )C(t − t ) ω2 1 − e−Γt − Q0 0 Γ 2δ t 1 − e−(Γ−δ)t 1 − e−(Γ+δ)t − Γ−δ Γ+δ −t ) dt 0 0 dt ξ(t )e−Γ(t C(t − t ) Now we do averages over ξ and T to obtain (linear terms vanish!) Q (t) + g L2 2 ξ T t 0 = Q2 T 0 t ω2 1− 0 2δ t t 0 1 − e−(Γ−δ)t 1 − e−(Γ+δ)t − Γ−δ Γ+δ 2 + 2 I0 T 1 − e−Γt Γ 2 dt1 0 dt2 T dt3 0 dt4 e−Γ(t1 −t2 +t3 −t4 )C(t1 − t2 )C(t3 − t4 ) δ(t2 − t4 ) T Now we substitute Q2 0 last term becomes 4RkT L2 2 = CkT . The problem is exactly the same as in the Exercise 5. The rest integrations must be done explicitly which results in lengthy expressions. After Q → x. LC → ω2 . and g = 4RkT . I → v.. STOCHASTIC DYNAMICS . The t dt1 0 dt2 0 dt3 e−Γ(t1 −2t2 +t3 )C(t1 − t2 )C(t3 − t2 ) . respectively. Assume that the velocity and displacement of the Brownian particle at t = 0 are v0 and x0 . Γ = R/L . (a) Compute the velocity correlation function v(t1 )v(t2 ) ξ .

= kT −(γ/m)|τ| e . 5. dt m m m Its solution is v(t) = v0 e−(γ/m)t + The quantity f0 1 − e−(γ/m)t γ is just the drift velocity in the force f0 . m = vd (t) we obtain v(t + τ)v(t) ξ T = vd (t + τ)vd (t) + kT −(γ/m)|τ| e .12: Stochastic processes Consider an electron (spin S = 1/2) bounded to an impiruty center and assume that the temperature T is much less than the energy difference between the levels of orbital motion.71 Solution 5. m In a similar way. c is the velocity of light since the Gaussian system is used. . . Using results of the Sec.E. P± . h The magnetic moment of the states with Sz = ±1/2 is β. (a) Find the stationary probabilities for the states with Sz = ±1/2.11: The differential equation for the particle is d v(t) γ f0 1 + v(t) − = ξ(t) . Consequently. we cam introduce drift displacement xd (t) = Finally we obtain [x(t) − x0 − xd (t)]2 ξ T t dt 0 f0 1 − e−(γ/m)t γ = f0 m t− 1 − e−(γ/m)t γ γ . only spin degrees of freedom are excited. = 2kT m t− 1 − e−(γ/m)t γ γ . the additional energy in magnetic ﬁeld is ±βB where ± corresponds to the values ±1/2 of the spin component along the ﬁeld. Let the system be placed in external magnetic ﬁeld B. Tests for training Problem 5.Find average magnetic moment of the system as function of temperature T and magnetic ﬁeld B.1 of the book [1] we obtain vd (t) = [v(t + τ) − vd (t + τ)] [v(t) − vd (t)] Since v(t) T ξ T dx = v(t) . dt t 0 f0 1 1 − e−(γ/m)t + γ m dt e(γ/m)(t −t) ξ(t ) . Here β = |e|¯ /2mc is the Bohr magneton.

. . Hint: It is convenient to express the set of Master Equations for the quantities P( f .Using the result for the stationary case ﬁnd the relation between the transition probabilities W+− and W−+ for the transition from the state “+” to the state “-” and for the reverse transition. to ﬁnd the system in the state f at time t under the conditions that at time t0 it was in the state i.7) . They are determined from the intial conditions and from the fact that at t − t0 → ∞ the conditional probabilities tend to the stationary ones. Explain these results qualitatively. . while C(k) are timematrix as P independent matrices. (e) Using Eq.t0 ). (5. respectively. (5.Specify initial conditions to this equation. w+ /w− = e−2βB/T .72 CHAPTER 5. (5. (∆M)2 and the ratio (∆M)2 / M 2 . According to the Gibbs distribution.t|i. STOCHASTIC DYNAMICS .12: (a) Denote w± the probabilities of the states Sz = ±1/2.6) (5.t|i.8) (5.t0 ). Express the probabilities W+− and W−+ through the quantity W = W+− +W+− .Discuss limiting cases of large and low temperatures.Find the ﬂuctuation spectrum.5) show that the correlation function ∆M(t)∆M(0) has the form ∆M(t)∆M(0) = (∆M)2 e−W |t| where (∆M)2 is given by Eq. (5. ˆ ˆ ˆ The solution of the matrix Master Equation ∂P/∂T = W P at t ≥ t0 can then be searched in in the ˆ ˆ ˆ ∝ ∑k C(k) e−λk (t−t0 ) . .t|i.Show that solution of this equation with proper initial condition is P( f . .4) cosh2 (βB/T ) .. Here λk are eigenvalues of the matrix W .t|i. We have M = −β (w+ − w− ) . P (t) ≡ w+ (t) − w+ (t).5) Solution 5. (5. . (d) Write down the Master Equation for the conditional probability.t0 ) = w f + (δi f − w f )e−W |t−t0 | where w f is the stationary probability to ﬁnd the system in the state f .. (b) Find mean square ﬂuctuation of magnetic moment.4) while W = W+− +W−+ .Derive the Master Equation for the population difference.Show that β2 (∆M 2 ) = . P( f .t0 ) at ˆ different i and f as an equation for the 2×2 matrix P with matrix elements Pf i (t|t0 ) ≡ P( f . (c) Write down the Master Equation for the probability Pi (t) to ﬁnd the system in the state i = ± at time t.

13) βB.12) (5. e2βB/T w− = 2βB/T . ∂P (t) = −W P (t) ∂t (5.9) (5. As a result. Combinig Eqs. (c) The equation reads.In the stationary situation. ∂t Subtracting equations we arrive at the result. . using the detailed equilibrium principle.8) and (5.14) β2 . ∂t ∂w− (t) = −W−+ w− (t) +W+− w+ (t) .15) (5.73 Since only 2 spin levels are involved. Wi→s wi (t) = Ws→i ws (t) we obtain W+− w− = = e2βB/T .9) we obtain.16) .11) (5. (5.10) (5. cosh2 (βB/T ) (∆M)2 1 = 2 2 M sinh (βB/T ) 1 e2βB/T +1 . W−+ w+ . (5.17) W−+ = W . e +1 βB T w− − w+ = tanh βB T (5.14) we get ∂w+ (t) = −W+− w+ (t) +W−+ w− (t) . (∆M 2 ) = At low temperatures. M = β tanh (b) We get: M 2 = β2 (w+ + w− ) = β2 M 2 = β2 tanh2 (βB/T ) . T to magnetic ﬁeld. ∂wi (t) = − ∑ [Wi→s wi (t) −Ws→i ws (t)] ∂t s=± . 1 + e2βB/T W+− = W 1 + e−2βB/T (5. w+ + w− = 1 . w+ = As a result.Coming back to Eq. ﬂuctuations are exponentially suppressed because spin is aligned (5.

if . Let us assume t0 = 0. (5. (5. Pf i (t) = w f + (δ f i − w f )e−W |t| This expression coincides with Eq. C f i +C f i = δ f i . (1) Cfi = wf . Consequently. λ2 = −W f →s +Ws→ f ≡ −W . ∂t ˆ with W = Ws→ f −W f →s W f →s −Ws→ f . 0) − w f ]wi . 0).t|i. in general. As result. STOCHASTIC DYNAMICS .t0 ) is a function of t −t0 . (5..19) ˆ Let us introduce the matrix P(t) with matrix elements Pi f (t) = P( f . at t → ∞ the conditions probability tends to the stationary one.18) ∂t s The initial condition to this equation is Pf i (t) = δi f at t → 0 . ∆M(t)∆M(0) = M(t)M(0) − M 2 (1) (2) (1) (2) = ∑ Mi M f [P( f . because of the time reversibility. Then Eq. ˆ Here I is the unit matrix Then.19). Pf i (t) = C f i +C f i e−Wt ˆ where C(k) are time-independent matrices.t|i. 0) ≡ Pf i (t).20) ˆ ˆ It is natural to search the solution of Eq. t ≥ 0 and denote Let us for brevity denote P( f .t|i.. (e) By deﬁnition.74 CHAPTER 5. (d) We consider a stationary random process. ˆ ˆ Equating the determinant of W − λI we get λ1 = 0 . and P( f .18) in the form P(t) ∑k C(k) e−λk t . Pf i (t) depends only on |t|. The Master Equation for Pf i (t) has the form ∂Pf i (t) = − ∑[W f →s Pf i (t) −Ws→ f Psi (t)] . (5. Here w f is the stationary probability of the state f . Finally. w f . (5. (5. From the initial conditions (5. On the other hand.18) can be rewritten as ˆ ∂P(t) ˆ ˆ = W P(t) . Then λk are the ˆ ˆ eigenvalues of the matrix W while Ck are the numerical matrices to be determined from intial conditions at t = 0 and from the fact that Pf i (t) → w f at t → ∞.t|i.5).

5). (5. (5. one can rewrite the above equation as ∆M(t)∆M(0) = e−W |t| ∑ Mi M f (δi f − w f )wi = (∆M)2 e−W |t| if This expresssion coincides with Eq. ∞ ∞ dt e−ωt ∆M(t)∆M(0) = (∆M)2 W ω2 +W 2 ∞ ∞ dt e−ωt−W |t| .6). The ﬂuctuation spectrum is (∆M)ω = The result reads as.75 Using Eq. (∆M)ω = 2 (∆M)2 . .

..76 CHAPTER 5.. STOCHASTIC DYNAMICS .

Chapter 6 The Foundations of Statistical Mechanics This chapter will be skipped this year 77 .

78 CHAPTER 6. THE FOUNDATIONS OF STATISTICAL MECHANICS .

We have dΘ(x)/dx = δ(x) . 79 d pN δ E − ∑ N 2 p2 mω2 xi i + 2m 2 V dx1 V dx2 . dxN d p1 d p2 . or quantum Hamiltonian.1: Compute the structure function for N noninteracting harmonic oscillators. . . . . For brevity let us assume the the oscillators are one-dimensional. Using this structure function and the microcanonical ensemble. Then the Hamiltonian is N 2 p2 mω2 xi i H =∑ + =E.Chapter 7 Equilibrium Statistical Mechanics Quick access: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Problem 7. Assume the system has total energy E. . Let us do the problem classically. x > 0 0. d pN Θ E − ∑ N i=1 2 p2 mω2 xi i + 2m 2 . thus the structure function Σ(E) = ∂Ω(E)/∂E is given by the equation Σ(E) = dx1 dx2 . 1. . V dxN d p1 d p2 . Solution 7. 2m 2 i=1 The total number of states is Ω(E) = Here Θ(x) = is the Heaviside unit step function. x < 0 . compute the entropy and the heat capacity of the system. each with frequency ω and mass m. . .1: One can use either classical. V V V i=1 .

1 E Ω(E) = 2π NΓ(N) ω Now. E0 = 0. . ∞ −∞ dq2N exp − ∑ 2N q2 j = ∞ −∞ dq e −q2 2N = πN . Each atom can exist in one of two states. Now. ˜ Ω(R) = A2N R2N . qi+N = pi 1/2m and R2 = E. . (a) Compute the entropy of the system as a function of internal energy. EQUILIBRIUM STATISTICAL MECHANICS Let us use the following trick. dq2N Θ R − ∑ q j ≡ ω ω j=1 From the dimensionality considerations. To ﬁnd the coefﬁcient A2N let us consider the integral ∞ −∞ dq1 . The number of atoms in energy level E1 is n1 . (b) Compute the heat capacity of a ﬁxed number of atoms. The internal energy of the system is U = n0 E0 + n1 E1 . This is the classical answer. the same integral is ∞ dR 0 ˜ ∂Ω −R2 e = 2NA2N ∂R ∞ 0 dR R2N−1 e−R = NA2N Γ(N) 2 Thus A2N = Finally. Then 2N 2 N 2 N ˜ 2 2 Ω(R) . the entropy is S/k = ln E Ω(E) = ln N N!h hω ¯ N N πN . C = kN .2: A system consists of N noninteracting. j=1 On the other hand.1 of the book [1].N. The quantum calculation is demonstrated in the Exercise 7. Denote qi = xi mω2 /2. . N. Ω= dq1 . . and E1 = ε. 1 ∂S Nk = = T ∂E E → E = NkT.. . NΓ(N) 1 E Σ(E) = 2π E Γ(N) ω N . i = 1. distinguishable two-level atoms. . 1 = N ln(E/¯ ω) − 2N ln N + 2N . h (N!)2 Here we used the Stirling’s formula.80 CHAPTER 7. Problem 7.

2: Since E0 = 0 the energy U cam be realized in U/ε ≡ n1 particles are in the state 1. The number of ways to do that is N! W= n1 !(N − n1 )! Consequently. k M!(N − M)! S/k ≈ 2 [N ln N − N − M ln M + M − (N − M) ln(N − M) + (N − M)] ≈ 2 [N ln N − M ln M − (N − M) ln(N − M)] . its energy is E = ε. and (N −U/ε) = N − n1 particles are in the state 0. If an atom occupies a normal sites. 2 kT e +1 eε/kT + 1 Problem 7. −1 T = (2k/ε) ln(N − M)/M . Thus N! M!(N − M)! Using the Stiling’s formula we get W= 2 → S N! = 2 ln . N identical atoms sit on the lattice. U= Nε ≈ Nε e−ε/kT . S N! = ln ≈ N ln N − N − n1 ln n1 + n1 − (N − n1 ) ln(N − n1 ) + (N − n1 ) . eε/kT Nε ε 2 U = ε/kT . M = U/ε . Solution 7. Compute the internal energy and heat capacity as a function of temperature for this lattice. M on the interstitial sites and N − M on the normal sites (N M 1. ≈ N ln N − ln − N − k ε ε ε ε To ﬁnd temperature we do usual procedure: 1 ∂S k Nε −U = ≈ ln . k (N − n1 )! Finally U S U U U ln N − . so we have to express M as a function of temperature.81 Solution 7. +1 C ≈ 2kN(ε/2kT )2 e−ε/2kT . M!(N − M)! The rest N − M atoms can be placed at N normal states in the same number of ways.3: The internal energy is U = Mε. Thus. eε/2kT . T ∂U ε U As result. If an atom occupies an interstitial site. its energy E = 0.3: A lattice contains N normal lattice sites and N interstitial lattice sites. The number of ways in which one can place M atoms at N interstitial sites is N! Wi = . C = Nk .

W = Differentiating this function by n−1 and n1 and equating the derivatives to 0 we get: n1 = (N − n1 − n−1 ) = n−1 The maximum entropy is Smax = kN ln 3 .82 CHAPTER 7. Also assume that atoms on different lattice sites cannot be exchanged. +1. n0 and n1 denote the respective number of atoms in each of those spin states. so they are distinguishable). so all atoms have the same energy. 3 and 4. Let n−1 . i i where Mi. k ∑ i From this expression we get α + βMi. Denoting the conﬁgurations as 1. Sz = −1. EQUILIBRIUM STATISTICAL MECHANICS Problem 7. What is the maximum entropy? (Assume that no magnetic ﬁeld is present.z is the magnetic moment of the system in the i-th conﬁguration. We have 4 conﬁgurations with magnetic moments in parentheses ↑↑↑ (3/2).z fi = γµ . fi (i denotes the conﬁguration). ↑↓↓ (−1/2). Each atom can be in one of 3 spin states:. The conﬁgurations with magnetic moments ±1/2 are triply degenerated. For the case γ = 1/2 compute the entropy and compute fi . which maximizes entropy subject to the conditions ∑ fi = 1 .4: Consider a lattice with N spin-1 atoms with magnetic moment µ. 0. .5: A system has 3 distinguishable molecules at rest. ∑ Mi. Solution 7. Solution 7. → n1 = n0 = n−1 = N/3 Problem 7. Find an expression for the distribution function.5: Let us measure magnetic moment in units µ. ↑↑↓ (1/2).z − ln fi − 1 = 0 → fi = exp (1 − α − βMi. we construct the functional ˜ S = gi fi (α + βMi. each with a quantized magnetic moment which can have z-components +µ/2 or −µ/2.z ) .2. Find the total entropy and the conﬁguration which maximizes the total entropy. n−1 !n1 !(N − n−1 − n1 )! lnW ≈ N ln N − n−1 ln n−1 − n1 ln n1 − (N − n−1 − n1 ) ln(N − n−1 − n1 ) .z − ln fi ) .4: We have N! . ↓↓↓ (−3/2) .

z ln 2) = − ln N − ln 2 = ln . . We have 2 equations for α and β: ∑ gi exp (1 − α − βMi. √ 27 2 α = 1 + ln .83 Here gi is the degeneracy factor: g1.4 = 1.6: A ﬂuid in equilibrium is contained in the insulated box of volume V . (∆Hi )2 in the i-th cell. 27 The entropy is 1 S 27 = − ∑ gi fi ln fi = − ∑ gi fi (ln N + Mi. or β = ln 2 . 4 fi = N eMi. (For simplicity assume the ﬂuctuations occur at ﬁxed particle number. we have N = 23/2 .z) i i = 1. g2. or 2 e1−α cosh 2 e1−α For γ = 1/2 that yields: cosh 3β β 3β β + 3 cosh = 3 sinh + 3 sinh 2 2 2 2 3β β + 3 cosh 2 2 3 3β 3 β sinh + sinh 2 2 2 2 = 1 = γ. The ﬂuid is divided (conceptually) into m cells.z ln 2 . √ 2.z exp (1 − α − βMi. As a result. = γ. Compute the variance of the enthalpy ﬂuctuations. Hint: Use P and S as independent variables. k 2 4 i i Problem 7.3 = 3. Ni ).z) ∑ giMi. Denoting x = eβ/2 we can rewrite this equation as x6 − 3x2 − 4 = 0 . the meaningful solution being x= Consequently.

(∆H)2 = T 2 (∆S)2 +V 2 (∆P)2 . In our case the proper pair is S P. Expanding ∆U in the series.84 Solution 7. ∂S P S ∂T ∆P + ∆S ∂S P S T ∆P + ∆S .1) is the minimum work needed to carry out reversibly the given change in the thermodynamic quantities of small part considered (relative to which the remainder of the body acts as a medium). CP S ∆P + ∂T ∂P The Maxwell relation following from dH = T dS +V dP is ∂V ∂S Thus ∆V = ∂V ∂P ∂V ∂P = P . (7. From the general theory the ﬂuctuation probability is proportional to w ∝ eWmin /kT . we obtain Wmin 1 ∂2U ∂U ∂2U 2 = (∆S) + 2 ∆S ∆V + 2 (∆V )2 2 2 ∂S ∂S∂V ∂V 1 ∂U ∂U = ∆S ∆ + ∆V ∆ 2 ∂S V ∂V S 1 = (∆S ∆T − ∆P ∆V ) .2) we obtain w ∝ exp 1 2kT (∆P)2 − S 1 (∆S)2 .2) Thus One can use this general formula and choose different couples as independent variables. 2 w ∝ exp ∆P ∆V − ∆S ∆T 2kT . 2CP . Wmin = ∆U − T ∆S + P ∆V (7. S Substituting the above relations into Eq. We get ∆V = ∆T = = ∂V ∂P ∂T ∂P ∂T ∂P ∂V ∆S . S ∆P + S ∂T ∂P .6: Thus We have CHAPTER 7. (7. EQUILIBRIUM STATISTICAL MECHANICS ∆H = T ∆S +V ∆P .

T ∆T ∆V = 0 Finally. Using these relations we get (∆H)2 = kT 2CP −V 2 kT (∂P/∂V )S . T (∆U)2 Problem 7. (For simplicity assume the ﬂuctuations occur at ﬁxed particle number. T. ∂V ∂P = −kT T → ∞. P(N. T. −P V + kT 2CV . T. (7. (∆P)2 = −kT (∂P/∂V )S . T ∂S ∂T ∂P ∂T V V ∂S ∂V ∂P ∆T + ∂V ∆T + ∆T = T ∂P CV ∆T + T ∂T ∆T . ∆S ∆P = 0 .7: We have. Remember that the particles are indistinguishable .7: A ﬂuid in equilibrium is contained in the insulated box of volume V . What happens to the internal energy ﬂuctuations near the critical point? Solution 7. We have: ∆S = ∆P = The we get w ∝ exp − CV 1 (∆T )2 + 2kT 2kT ∂P ∂V (∆V )2 . (∆Ui )2 in the i-th cell.V ). Ni ). S(N. Use this partition function to compute the internal energy. so when using a partition function one must insert a counting factor. V ∆V .2) and choose V. U(N.8: What is the partition function for a van der Waals gas with N particles? Note that the result is phenomenological and might involve some guessing. Problem 7. Compute the variance of the internal energy ﬂuctuations. The ﬂuid is divided (conceptually) into m cells. (∆U) Near the critical point 2 (∆T )2 = kT 2 /CV . . It is useful to compare it to the partition function for an ideal gas. ∆U = ∂U ∂V ∆V + T ∂U ∂T ∆T = T V ∂P ∂T − P ∆V +CV ∆T . (∆S)2 = kCP .85 As a result.V ). ∂P ∂T 2 (∆V )2 = −kT (∂V /∂P)T . V Now we can use Eq. and the entropy. the pressure.V ). T as independent variables.

Since the partition function for the ideal gas is ZN = V N ZN Then we are left with ZN (0) ZN = 1 VN dr1 . Let us try to consider rareﬁed gas where only pair interaction is important. Then we can choose N(N − 1)/2 pair to get N2 dr1 dr2 e−βV (r1 −r2 ) − 1 . ZN = (k) dp1 . eV A0 = −NkT ln N N2 a b− V kT mkT 2π¯ 2 h 3/2 = N f (T ) − NkT ln(eV /N) . (0) 1 2 dr 1 − e−βV (r) . . and at r > 2r0 the interaction is small. drN dp1 . . N!λ3N T √ (0) (k) where λT = h/ 2πmkT . ZN = 1 N!h3N 1 N!h3N dr1 .86 Solution 7. A0 ≡ −kT ln ZN . ∞ 2 2r0 r drV (r). . . drN e−βV ({ri }) − 1 ≡ 1 + zN . zN = −N 2 B(T )/V . First we can integrate out the momenta. drN e−βV ({ri }) = 1 + 1 VN dr1 . . . . zn ≈ 2V 2 Introducing the virial coefﬁcient B(T ) = where r = r1 − r2 . . dpN e−β ∑i pi /2m = 2 1 . .8: CHAPTER 7.p}) . In the spirit of the van der Walls model let us assume that the interaction is large at r ≤ 2r0 . Finally. . zn = − For an ideal gas. we obtain We get A = A0 − kT ln(1 + zn ) ≈ A0 − kT zN = A0 + kT N 2 B(T )/V . . EQUILIBRIUM STATISTICAL MECHANICS The partition function is (β ≡ (kT )−1 ). dpN e−βH ({r. Then B(T ) = 2π 0 3 ≈ 16πr0 /3 + 2πβ 3 where b = 16πr0 /3 and a = π 2r0 r2 dr 1 − e−βV (r) + 2π ∞ 2r0 ∞ 2r0 r2 dr 1 − e−βV (r) r2 drV (r) ≡ b − a/kT . .

ε. compute the ratio Na /Ns as a function of pressure. A = N f (T ) − NkT ln(e/N) − NkT (lnV − Nb/V ) − N 2 a/V .87 Thus. S=− = S0 + kN ln 1 − ∂V V − b V ∂T V 2a 2a N 3 N U = U0 − = kT − . N Ns a βNa ε e . (Hint: equate chemical potentials of the adsorbed atoms and the gas). At adsorbed atom has energy E = −ε. Since Na Ns . where ε > 0. Assume that atoms on the surface do not interact with each other. (b) If the surface is in equilibrium with an ideal gas of similar atoms at temperature T . Problem 7. From this we get P = − ∂A NkT N 2a ∂A Nb = − . Since ln(V − Nb) ≈ −Nb/V + lnV we get A ≈ A0 − NkT ln(1 − Nb/V ) − N 2 a/V . Assume the gas has number density n. and Na /Ns (use the canonical ensemble). Na ! i=1 Here β = 1/kT .9: (a) The canonical partition function is 1 Ns −βEi Ns ! Z= ∑ e = Na!(Ns − Na)! eβNaε . V 2 V . P of the gas. Z≈ Equating this expression to e−Na βµ we get µ = −ε − kT ln(Ns /Na ) . Na ! . compute chemical potential of the adsorbed atoms as a function of T .9: Consider a solid surface to be a two-dimensional lattice with Ns sites. so that each site has either 0 or 1 adsorbed atom. (a) If the surface is at temperature T . Assume that Na atoms (Na Ns ) are adsorbed on the surface. Solution 7.

t. . Here α denotes the vibrational mode. (a) Compute the Debye frequency for this lattice. .. Since in two-dimensional case dk . . what is the heat capacity? Solution 7. k → Lx Ly (2π)2 dkx dky . In the Debye approximation it is assumed that the dispersion law ωk. Problem 7. Li where cα is the sound velocity for each mode. we use continual approximation and replace ∑. As in this exercise.α = cα k . we obtain Pλ3 ε Ns T − − ln = ln kT Na kT Pλ3 T kT . Ns kT This expression is valid at high temperatures when nλ3 T e−ε/kT . In two-dimensional lattice we have two modes. . . 2. . and we can apply zero boundary conditions. . . .α Lx Ly (2π)2 ∑ α dk δ (ω − ωk. = we obtain g(ω) = kdk dφ . l.α ) . CHAPTER 7. . → Na Pλ3 ε/kT = Te . . (b) In the limit T → 0. The the density of states is g(ω) = ∑ δ (ω − ωk. and transverse. = 2π Lx Ly 2π¯ 2 h 1 1 + 2 c2 ct l kdk ..10: Consider a two-dimensional lattice lattice lattice in the x − y plane with sides of length Lx and Ly which contain N atoms(N is very large) coupled by nearest neighbor harmonic forces.E of the book [1]. EQUILIBRIUM STATISTICAL MECHANICS µ = kT ln Equating chemical potentials. .10: We follow the section 7. Since N 1 the boundary conditions are not important. .α ) = k. si = 1.α is linear and isotropic: π ωk. longitudinal. .88 (b) For an ideal gas. ω. ki = si .

Since at the walls the wave function should vanish. for each component α of the momentum p we get: sin(pα L/¯ ) = 0 h → kα = nα π/L . ¯ Problem 7. C= 4kN (¯ βωD )2 h TD /T Lx Ly 4π¯ 2 h 1 1 + 2 c2 ct l ω2 D Lx Ly 8Nπ¯ 2 h 1 1 + 2 . T →∞ 2. h Here nα is an integer number. εt = 2mL2 . T → 0 . TD = hωD /k. Solution 7. ω2 D 2kN. Consequently 2N = and 1 = ωD The density of states can be expressed as g(ω) = As result.11: Let us put the origin of the reference frame in the cube’s corner. the entropy. 24ζ(3)kN(T /TD ) dx 0 x 3 ex = (ex − 1)2 Here ζ(x) is the Riemann ζ-function. ζ(3) ≈ 1. what is the spacing between the translation energy levels? (b) Write the partition function for this system (include the translational and rotational contributions).3 A). we should get ωD 0 g(ω) dω = 2N . 2 cl ct 4N ω. At what temperature do rotational degrees of freedom become important.0 cm. εn = ε0 (nx + ny + nz ) .20. Thus the energy levels of translational motion are given by the expression h2 π2 ¯ 2 2 2 . (c) Write expressions for the Helmholtz free energy.89 Since the total number of vibrations should be normalized to 2N. and the heat capacity of the system for temperatures where the rotational degrees of freedom make a signiﬁcant contributions. where 2 is the number of modes and N is number of atoms. (a) If L = 1.11: A cubic box (with inﬁnitely hard walls) of volume V = L3 contains an ideal gas of N rigid HCl molecules (assume the effective distance between the H atom and the Cl atom ˚ d = 1. kα ≡ pα /¯ .

βεr −2βεr . 1. by how much. Atoms of different color are distinguishable. It is important that each rotational level is degenerate by the factor (2K + 1). C=T ∂T V ∂T The calculations are obvious. 1. Atoms of the same color are indistinguishable. 2I m1 m2 2 d . Thus Z = Ztr · Zrot where Ztr = Zrot = h2 ¯ K = 0. Does if differ from that of the mixture.90 CHAPTER 7. εr = . Thus the spacing between the levels is εK+1 − εK = 2(K + 1)εr . If so. βεt T exp(−βεt ). . ∂2 A ∂T 2 = −T V . βεt 1.12: An ideal gas is composed of N “red” atoms of mass m . EQUILIBRIUM STATISTICAL MECHANICS Thus the spacing between the levels is εni +1 − εni = (2ni + 1)εt . (a) Use the canonical ensemble to compute the entropy of this gas. and N “green” atoms of mass m. N “blue” atoms of mass m. I= the energy of rotation being εK = εr K(K + 1) . . V Problem 7. 1. m1 + m2 K=0 ∑ (2K + 1) e−βεr K(K+1) = nx =1 ∞ ∑e ∞ 3 −βεt n2 x = V /λ3 . ∂A ∂S S = − . The rotational degrees of freedom are determined by the moment if inertia. . 1. . A = −kT ln Ztr − kT ln Zrot . . (b) Compute the entropy of ideal gas of 3N “red” atoms of mass m. βε 1 + 3e r The following is straightforward. kT /εr .

3/2 mkT 2π¯ 2 h . The “green” particles have magnetic moment γG . ∑e k −βp2 /2m 1 = N! V λ3 T N . In the case of single-color gas of 3N particles we have S1 = 3Nk ln(eV /3N) − 3N f (T ) → S − S1 = 3N ln 3 . 1 V cosh(βγB) N eV cosh(βγB) kT = −kT ln µ = − ln N N! λT NλT Equating chemical potentials for “red” and “green” gases we get cosh(βγR B) cosh(βγG B) = NR NG → NR cosh(βγR B) . . eV AN = −NkT ln N Here mkT 2π¯ 2 h f (T ) = 3/2 = N f (T ) − NkT ln(eV /N) . Solution 7.13: Let us consider one of the gases since they are noninteracting. For each component N ZN = From this expression. where NR is the number of “red” atoms and NG is the number of “red” particles.12: Since the atoms of different color distinguishable the partition function is just the product of partition function.91 Solution 7.13: An ideal gas consists of a mixture of “green” and “red” spin-1/2 particles. Problem 7. Z = ZN . A magnetic ﬁeld. Compute the ratio NR /NG . S = 3NK ln(eV /N) − 3N f (T ) . and A = 3N f (T ) − 3NkT ln(eV /N) . All particles have mass m. Assume that temperature is high enough that Fermi statistics can be neglected. where γR < γG . B. 3 Thus. and the “red” particles have magnetic moment γR . The system will be in equilibrium if chemical potentials the the “red” and the “green” gases are equal. Use the canonical ensemble (no other ensembles will be accepted). We have Z = Ztr · Zs where Ztr is the usual transport partition function while Zs = e−βγB + eβγB Since N = 2N coshN (βγB) . = NG cosh(βγG B) . is applied to the system.

. EQUILIBRIUM STATISTICAL MECHANICS Problem 7. ¯ si |P|si+1 = eβsi si+1 .. Compute the correlation function. s1 s2 . Nε ∂β coshN (βε) + sinhN (βε) 1.. Thus ZN (β) = 2N (coshN βε + sinhN βε) . we introduce the matrix ¯ P= We have Then ZN (T ) = eβε e−βε e−βε eβε = eβε I + e−βε σ 1 . i=1 N Assume periodic boundary condition. How does it behave at very high temperature and at very low temperature? Solution 7. ZN (T ) = s1 =±1 ∑ . . cosh(βε) → 1 . . sN |P|s1 = Tr PN . sN =±1 ∑ ¯ ¯ ¯ ¯ s1 |P|s2 s2 |P|s3 . We have si si+1 = coshN−1 (βε) sinh(βε) + sinhN−1 (βε) cosh(βε) 1 ∂ ln ZN = . Thus we have ferromagnet λ2 = 2 sinh βε . As it is recommended in the book [1]. ¯ The matrix P has eigenvalues λ1 = 2 cosh βε .. At high temperature.92 CHAPTER 7. s1 =±1 ∑ . cosh(βε) ≈ sinh(βε) and si si+1 = 1. At low temperature.14: Consider a one-dimensional lattice with N lattice sites and assume that i-th lattice site has spin si = ±1. βε ordering.14: We have. sN =±1 ∑ exp βε ∑ si si+1 i .. The Hamiltonian describing this lattice is H = −ε ∑ si si+1 . so sN+1 = s1 . sinh(βε) → 0 → si si+1 = βε 1.

. E(ε. we obtain Tc s3 s ≈ s + T 3 Thus. (c) Compute the magnetic susceptibility.F. and C = −N ∂E(ε.15: (a) Since tanh ξ ≈ 1 − 2 e−2ξ at large ξ. What is the critical exponent for both cases? Solution 7. the order parameter. χ(T. T → 0. s ≈ 3(1 − T /Tc ) . H = − ∑ E(ε. s . B) = i=1 N νε s + µB . N. N)B=0 . (a) Show that s is the following temperature dependence s ≈ 1 − e−2Tc /T . The the Hamiltonian is Tc T 3 → s ≈ 1 − 2 e−2Tc /T . satisﬁed the equation Tc s = tanh s .3) Thus U = −NE s = −NE s . T → Tc .93 Problem 7. . thus the jump in speciﬁc heat is given by C|Tc −0 .B At T → Tc − 0 and B = 0 we get s = 3(1 − T /Tc ) → C|Tc −0 = 3 Nνε . B) si . in the neighborhood of Tc both for T > Tc and T < Tc . tanh ξ ≈ ξ − ξ3 /3. 2 Tc At T > Tc the average spin is zero. T where Tc = ν ε/2k where ε is the strength of the coupling between lattice sites and ν is the number of nearest neighbors. (b) Compute the jump in the heat capacity at T = Tc .15: In the mean ﬁeld approximation to the Ising lattice. 7. (b) We will use notations of Sec.B ∂E ∂s ∂s ∂T .2 of the book [1]. 3(1 − T /Tc ). B) s ∂T = −N E + s N. 2 (7. we get s ≈ 1 − 2 e−2 s Tc /T At small ξ.

∂B N. νε 2(Tc − T ) Thus there is a divergence at T → Tc . Compute the temperature for Bose-Einstein condensation. Finally. below Tc χ= Tc Nµ2 . E < 0. . Solution 7. and χ= Below Tc we get s2 cosh ( s Tc /T ) ≈ 1 + 2 2 1 βνε η + βµ . 2 νε cosh ( s Tc /T ) − Tc /T Nµ2 Tc . 0.94 CHAPTER 7. with critical exponent equal to -1. χ ∝ |T − Tc |−1 .B→0 Deﬁning η = (∂ s /∂B)N.16: The density of states of an ideal Bose-Einstein gas is g(E) = αE 2 . where a is a constant. EQUILIBRIUM STATISTICAL MECHANICS (c) To get magnetic susceptibility we have to differentiate the magnetic moment N = Nµ s with respect to B: ∂s χ = Nµ . χ= Above Tc we have s =0.B→0 we get the following equation at B → 0: η= Its solution is η= Finally. cosh ( s Tc /T ) 2 2 2βµ . νε T − Tc Tc T 2 2 ≈ 1+3 1− T Tc . kTc = N 2αζ(3) . x −1 e β3 c 1/3 Here ζ(x) is the Riemann zeta-function. As a result. 2 cosh ( s Tc /T ) − Tc /T 2 Nµ2 Tc /T . E > 0.16: We have the equation for βc = 1/kTc : N= α g(E) dE = 3 βc E − 1 βc Emin e ∞ ∞ 0 x2 dx 2αζ(3) = . Problem 7.

m3/2 a≡ √ . while bosons in the excited state have the energy E1 = p2 /2m + ∆. dp δ E − p2 2m = 4π h3 p2 d p δ E − p2 2m √ =a E. Denoting β0 = we get βc = β0 1 + √ πa 3/2 2Nβ0 2/3 √ aζ(3/2) π 2N 2/3 e −β0 ∆ ≈ β0 + √ 2 πa 6N β0 e−β0 ∆ .18: Compute the Clausius-Clapeyron equation for an ideal Bose-Einstein gas and sketch the coexistence curve. x −1 e 2 2 We can solve this equation by iterations since ∆ is large. pv = m [g5/2 (1)]5/3 . that the boson are two-level atoms. Does the existence of internal degree of the internal degree of freedom raise or lower the condensation temperature? Solution 7. Compute the Bose-Einstein condensation temperature. Assume that ∆ kT . Thus the internal degree of freedom decreases the condensation temperature. Tc .17: g0 (E) = In a similar way.17: An ideal Bose-Einstein gas consists of noninteracting bosons with mass m which have an internal degree of freedom which can be described by assuming. Problem 7. for this gas of two-level bosons. The equation for βc = 1/kTc is: Nβc a 3/2 = 0 ∞ √ x dx + x −1 e ∞ βc ∆ √ √ x − βc ∆ dx ζ(3/2) π π −βc ∆ ≈ + e . where p is the momentum and ∆ is the excitation energy. Show that the line of transition points in the P − v plane obeys the equation 2π¯ 2 g5/2 (1) h 5/3 . Bosons in the ground state have the energy E0 = p2 /2m. 2π2 h3 ¯ √ g1 (E) = a E − ∆ .95 Problem 7. We start with calculation of density of states.

α=1 α ∞ (7. P= Pc = kTc g5/2 (1) . e l −1 e l −z The functions gi (z) are introduced as 4 g5/2 (z) = − √ π g3/2 (z) = z ∞ 0 x2 dx ln 1 − z−x 2 = ∞ zα d g5/2 (z) = ∑ 3/2 .18: Let us ﬁrst recall the notations. the distribution function is written as z 1 nl = β(ε −µ) = βε .4) (7. (a) what is u? (b) What is α? . EQUILIBRIUM STATISTICAL MECHANICS Solution 7. Pc = kTc g5/2 (1) λ3 T → Tc = Na3 V g3/2 (1) 2/3 . = ka2 g5/2 (1) [vg3/2 (1)]5/3 = 2π¯ 2 g5/2 (1) h .96 CHAPTER 7. λT Thus the equation of state is kT g5/2 (z) .19: Show that the pressure. λ3 T In equilibrium with the condensate at T = Tc . P. λ3 T g3/2 (1) λ3 T The critical particle density in the condensation point is nc = Denoting λT = a/T 1/2 where a = . of an ideal Bose-Einstein gas can be written in the form P = αu. where “u” is the internal energy per unit volume and a is a constant. the grand canonical potential is Ω= V (kT )5/2 m3/2 √ 2π2 h3 ¯ ∞ 0 kT √ dy y ln 1 − z e−y = −V 3 g5/2 (z) . The quantity z = eβµ is called the fugacity.5) Now. 2π¯ 2 /mk we get at the critical point: h 3/2 N g3/2 (1) Tc = V a3 Thus. dz α=1 α zα ∑ 5/2 . m [vg3/2 (1)]5/3 Problem 7.

Problem 7. Assume the system is at T = 0 K. Cu metal in the solid state has mass density ρ = 9 gr/cm3 .97 Solution 7.19: Let us do integration by parts in Eq. . 7. Solution 7.5 to get ∞ 2 √ g5/2 (z) = − √ dy y ln 1 − z e−y π 0 2 2 ∞ z = √ · dy y3/2 y e −z π 3 0 Calculation the average energy U= 0 ∞ g(ε) ε dε z eβε − z and recalling that g(ε) ∝ √ ε we immediately see that 2 PV = U 3 → P= 2 u. we have taken into account that spin degeneracy for electrons 2. (a) Compute the Fermi energy. .20: Let us start with the calculation of the density of states: g(ε) = 2V · 4π p2 p2 d p δ ε − h3 2m √ √ 2V m3/2 √ = ε ≡ V g0 ε . α = 2/3. TF = εF /k. π2 h3 ¯ √ Here g0 = 2m3/2 /π2 h3 . Inte¯ gration expression for densoty of states from 0 to εF and equationg to the number of electrons we obtain 3/2 (2/3)g0 εF = ne → εF = (3ne /2g0 )2/3 . 3 Thus. εF .20: Electrons in a piece of Cu metal can be assumed to behave as an ideal Fermi gas. of the electron gas. Assume that each Cu atom donates an electron to the Fermi gas. Here ne = Ne /V = ρ/mCu is the electron density. (b) Compute the Fermi temperature.

β Thus eβµ = eβN/D − 1 The energy is then U =D µ= 1 N ln eβN/D − 1 ≈ − kTe−N/DkT . EQUILIBRIUM STATISTICAL MECHANICS The density of states of an ideal Fermi-Dirac gas is f (E) = D. f (E) = DE.. β D ∞ 0 E dE eβ(E−µ) + 1 . . Solution 7. where D is a constant. = z +1 0 0 e In our case. D 2 6 The speciﬁc heat is then ∂µ π2 2 C =µ + k T. (b) Compute the speciﬁc heat at low temperature. Consider an auxiliary integral. (a) Compute the Fermi energy. we obtain I≈ µ = 1 − (ez + 1)−1 and that eµ/kT µ 0 1. .21: We have. 6 U µ2 π2 ≈ + (kT )2 . D ∂T 3 The ﬁrst item is exponentially small at low temperatures. f (µ) = D. Let us use the trick called the Sommerfeld expansion. and C≈ π2 2 π2 kT Dk T ≈ k . . N=D 0 ∞ dE eβ(E−µ) + 1 → = D β ∞ −βµ dx ex + 1 = D ln 1 + eβµ . 0. Expanding the z dz +. if E > 0 . and we get f (E) dE + 2(kT )2 f (µ) ∞ f (E) dE + π2 (kT )2 f (µ) + . 3 3 µ . I = ∞ f (E) dE f (µ + kT z) dE = (E−µ)/kT + 1 ez + 1 0 e −µ/kT µ/kT f (µ − kT z) dE ∞ f (µ + kT z) dE = kT + kT e−z + 1 ez + 1 0 0 µ ∞ f (µ + kT z) − f (µ − kT z) = f (E) dE + kT dE ez + 1 0 0 −1 ∞ Here we have used the fact that (e−z + 1) integrand.98 Problem 7..21: CHAPTER 7. if E < 0 .

V This formula works in the wide region µe B In very strong magnetic ﬁelds the analysis is more complicated. Assume that fermions each have magnetic moment µe . . Find the expression for magnetization in the limit of weak magnetic ﬁeld and T → 0 K. Then Ω(µ) = − 2A (µ + µe B)5/2 15 → M= Aµe (µ + µe B)3/2 = Nµe .V. Let us consider the case of very strong magnetic ﬁeld.99 Problem 7.µ = −N we get M = µ2 B e ∂N ∂µ µ. − A 15 These equations are written assuming that µe B ≤ µ. 2 ∂µ2 = µ2 B e ∂2 Ω0 (µ) . otherwise only one item is present. Physically it means that only spins-up-states are occupied. 2 Ω(µ) = (µ + µe B)5/2 + (µ − µe B)5/2 . A 3 In a similar way. . √ 2 2A ∞ E 3/2 dE 2V m3/2 Ω0 (µ) = − U = − . First the chemical potential is given by the expression N 1 = (µ + µe B)3/2 + (µ − µe B)3/2 . ∂µ2 ∂Ω ∂B T. Solution 7. Ω(µ) ≈ Ω0 (µ) + Consequently. − = A 3 0 e(E−µ−µe B)/kT + 1 3 0 e(E−µ+µe B)/kT + 1 In small magnetic ﬁelds or at large temperature.22: Compute the magnetization of an ideal gas of spin-1/2 fermions in the presence of magnetic ﬁeld. M=− Since ∂Ω ∂µ (µe B)2 ∂2 Ω0 (µ) . Since the spin degree of freedom leads to additional energy ±µe B we get 1 1 Ω(µ) = Ω0 (µ − µe B) + Ω0 (µ + µe B) 2 2 where Ω0 is the grand potential in the absence of magnetic ﬁeld. A= 3 3 0 e(E−µ)/kT + 1 π2 h3 ¯ As a result. µe B µ.22: We will calculate here only spin (Pauli) susceptibility. 1 ∞ Ω(µ) 1 ∞ E 3/2 dE E 3/2 dE + . T. 3 .

(Note: You are now considering fairly high temperature).100 CHAPTER 7. Nl !(Gl − Nl )! Using the mean occupation number nl = Nl /Gl we can rewrite this formula as S = −k ∑ Gl [ nl ln nl + (1 − nl ) ln(1 − nl )] . Problem 7. l Gl ! . l where nl = eβ(εl −µ) + 1 −1 . ﬁnd the difference in the pressure and the isothermal compressibility between an ideal boson and ideal fermion gas. then Gl = 1.24: To the lowest order in the density. . Solution 7.23: The number of ways to distribute Nl identical particles among Gl states with not more than 1 particle per state is just Wl = Using the Stirling’s formula we obtain S = k ∑ ln [Gl ln Gl − Nl ln Nl − (Gl − Nl ) ln(Gl − Nl )] . Assume that the fermions and bosons have the same mass and both are spin-less. Solution 7. EQUILIBRIUM STATISTICAL MECHANICS Problem 7.24: Assuming that the density of states is given by the expression g(ε) = g0V ε1/2 we obtain 2 Ω = − g0V 3 ∞ 0 ε3/2 dε 2 ≈ − g0V β(ε−µ) ± 1 3 e ∞ 0 ε3/2 dε eβ(µ−ε) 1 eβ(µ−ε) . Introducing the fugacity z = eβµ we get 2 P = g0 (kT )5/2 z 3 ∞ 0 ξ3/2 dξ e−ξ 1 z e−ξ √ π = g0 (kT )5/2 z 1 2 √ z 2 8 .23: Show that the entropy for an ideal Fermi-Dirac gas (neglecting spin) can be written in the form S = −k ∑ [ nl ln nl + (1 − nl ) ln(1 − nl )] . l If we ignore spin.

3/2 ∞ z 0 ξ 1/2 dξ e −ξ 1 ze −ξ π = zg0V 2 √ √ 1 z 2 4 . The compressibility is calculated in a straightforward way since NkT V≈ P √ 2 1± √ 4 πg0 (kT )5/2 P . the sign “-” holds for the Bose statistics. πg0 (kT )3/2 V √ We obtain NkT P≈ V 1 1 2 8 √ z0 42 z0 NkT ≈ V √ 2 1 ± z0 8 NkT ≈ V 2 N 1± √ 3/2 V 4 πg0 (kT ) √ . . 2 N z0 = √ .101 The normalization condition in this case gives N = g0V (kT ) As result. The sign “+” hold for the Fermi statistics.

EQUILIBRIUM STATISTICAL MECHANICS .102 CHAPTER 7.

Chapter 8 Tests and training Here we put some exercises for self-testing. 103 .

∂F ∂T V . The thermodynamic potential with respect to P and T is called Gibbs free energy. (b) Prove the relation U = −T 2 where U is the internal energy. Thus. TESTS AND TRAINING Problem 8. How much entropy it gains? (d) Discuss the heating if the same body is in contact with a thermostat at T2 . V P=− . T Thus. (b) Substituting S=− into deﬁnition of F we get the result. F is the thermodynamic potential with respect to V and T . . Find the ﬁnal temperature and the change in entropy. We have dF = dU − T dS − S dT = (T dS − P dV ) − T dS − S dT = −S dT − P dV. The function of state F = U −TS is called Helmholtz free energy. In the last case the heating is irreversible. Show that the total entropy change is positive. (c) A body with constant speciﬁc heat CV is heated under constant volume from T1 to T2 .1: (a) Mechanical work under isothermic process is given by dW = dU − dQ = dU − T dS = d(U − T S) . (e) Two similar bodies with temperatures T1 and T2 brought into contact. S=− ∂F ∂T ∂F ∂V ∂ A ∂T T .1: (a) Discuss the difference between Gibbs and Helmholtz free energy. Solution 8.104 CHAPTER 8. We have G = U − T S − PV → dG = −S dT +V dP .

105 (c) We have S= (d) Use 2nd law of thermodynamics (e) Energy conservation law yields CV (T2 − TB ) = CV (TB − T1 )− > TB = (T1 + T2 )/2 . It is convenient to introduce the density of states as number of states per energy interval. 2π2 h3 ¯ i .2: Find ﬂuctuation of the energy per particle for the Boltzmann gas of spinless particles at a given temperature . 1 ∂2 Z1 (β) 1 ∂Z1 (β) − Z1 ∂β2 Z1 ∂β = ∂2 log Z1 . for the entropy change we have δS = CV ln TB TB T1 + T2 +CV ln = 2CV ln √ T1 T2 2 T1 T2 ≥ 0. T2 T1 CV T2 dT = CV ln T T1 . Assume that ε p = p2 /2m. g(ε) = ∑ δ(ε − εi) = p 4π (2π¯ )3 h ∞ 0 m3/2 √ p2 d p ∑ δ(ε − εi ) = √ ε. Z1 ∑ i Z1 ∂β2 i 2 = (εi − ε )2 = ε2 − ε 2 . ∂β2 Now we have to specify Z1 for εi = p2 /2m. Z1 i 1 1 ∂2 Z1 (β) ε2 e−βεi = . Solution 8. Problem 8.2: (δε)2 Now ε ε2 Thus (δε)2 = = = 1 1 ∂Z1 (β) ∑ εie−βεi = − Z1 ∂β . Z1 = ∑ exp(−βε p) = p 1 (2π¯ )3 h ∞ 0 4πp2 d p exp(−βε p ) .

β3/2 a a= → (δε)2 = 3 . 2β2 Problem 8. 2 ∂T ∂β 4 cosh (βε/2) p(ε) dε c(ε) = P0 ∞ dε 0 1 1 P0 = 2 4 cosh (βε/2) 2β ∞ 0 dx P0 .3: As we know from the lecture.106 In this way. . m3/2 Z1 = √ 2π2 h3 ¯ As 0 ∞ 0 ∞ CHAPTER 8. n− Thus n± = n+ + n− = 1 .3: probability Two-level systems with random energy spacings ε are distributed with a constant p(ε) = P0 . n+ = e−βε . exp(±βε) + 1 The average energy is (we start from the middle of the distance between the levels) ε ε = ε ε βε (n+ − n− ) = tanh . 2 we get Z1 = Thus . TESTS AND TRAINING dε ε1/2 exp(−βε) = √ m3/2 2π2 h3 β3/2 ¯ ∞ 0 dx x1/2 exp(−x) . Solution 8. = 2 cosh x 2β Note that it is proportional to T . √ dx x1/2 exp(−x) = Γ(3/2) = (1/2) π π m3/2 8 π2 h3 ¯ 3 log Z1 = log a − log β . Find speciﬁc heat of the system. 1 . 2 2 2 The contribution to speciﬁc heat is c(ε) = The total speciﬁc heat is C = 0 ∞ ∂εε ∂εε 1 1 == −β2 = .

2 Variable transformation Jacobian ∂(u.1 Thermodynamics A.1: Thermodynamic potentials (summary) A. N T. Z) = const. u) =− .Y. v)/∂(x. y) ∂(u. V. v) ∂u/∂x ∂u/∂y = ∂v/∂x ∂v/∂y ∂(x. N T. v) ∂(v.1) ∂(u. ∂(x.1. y) = ∂(x. P.3 Derivatives from the equation of state Important relations arise from the properties of partial differentiations. v) ∂(u. P. µ Differential T dS − P dV + µ dN T dS +V dP + µ dN −S dT − P dV + µ dN −S dT +V dP + µ dN −S dT − P dV − N dµ Table A.Y as indepen107 . y) ∂(x. s) = · .Appendix A Additional information A.1. v) ∂(t.1 Thermodynamic potentials Thermodynamic potential Internal energy Heat function (enthalpy) Helmholtz free energy Gibbs free energy Landau free energy Notation U H F G Ω Independent variables S. y) ∂(t. y (A. V. Now let us consider X.2) A. y) ∂u ∂x . X.Y. ∂(x. y) is deﬁned as the determinant ∂(u.1. Z. N S. V. related by the equation of state K (X. y) The following relations can be useful: ∂(u. Consider 3 quantities. s) ∂(x. y) (A. N T.

6) Consider any quantity F (X. (A. We obtain ∂X ∂Z ∂Z ∂Y dZ = 0 . (A.6) and the Maxwell relations are usually used for transformations and computation of derivatives from the equation of state.Y ). X Y ∂Z ∂Y ∂X ∂Y − Z Y Since dX and dY are independent. (dY = 0). X (A. A. Y (A. . this equation is compatible if ∂Z ∂X ∂X ∂Y or ∂X ∂Y ∂X + Y ∂Y Z ∂X + ∂Z Y Z ∂Y ∂Z ∂Z ∂Y ∂Z ∂Y ∂Z ∂X ∂Y ∂X = 0 X = 0.3) If Y. X Then we divide it by dZ and assume Y = const. Y −1 (A.4) Z and Eq. We have ∂Z ∂X dX + Y APPENDIX A. while Z = Z(X. the differential of which ∂F ∂X dX + Y ∂F ∂Y ∂X ∂Z dX .5) Z X ∂X ∂Y General scheme for transformation: can be expressed as dF = = Z .108 dent variables.5). ADDITIONAL INFORMATION ∂Z ∂Y dY − dZ = 0 .4 by dX + − X ∂Z ∂Y X and subtract the results. Z (A.Y ). Z are taken as independent variables.8) together with Eqs.7) Y Another important relation arises if one divides the expression for d F by xX. ∂F ∂X = Z ∂F ∂X + Y ∂F ∂Y X ∂Y ∂X .8) Equations (A.3 by ∂Z ∂X ∂X ∂Y + Z ∂X ∂Y ∂X ∂Y dY + Z ∂X ∂Z dZ = 0 Y (A. X = −1 . (A. Z (A. A. then −dX + Now we multiply Eq.7). We get ∂F ∂Z = Y ∂F ∂X .

. (A. MAIN DISTRIBUTIONS 109 A. √ = I1 (a) 2π .12) Gaussian integrals I1 (a) = ∞ −∞ dz e−az 2 /2 = 2π/a . 2∑ i. .14) z2 dz e−z 2 /2 =2 dI1 (a) da a=1 Gaussian distribution for more than one variables S − S0 = − 1 βik xi xk . (A.13) (A. 2πσ Distribution of displacemnet for 1D random walk with the step l G(x) = 1 2l 2πN p(1 − p) ∞ −∞ (A.10) 1 ¯ G(x) = √ e−(x−x)/2σ .15) 2 Consequently. let us assume summation over repeated subscripts and rewrite the above equation 1 S − S0 = − βik xi xk . For convenience.11) e−[x−(p−q)l] 2 /8Nl 2 p(1−p) .k Deﬁne entropy as S(x1 .2 Main distributions Binomial distribution WN (m) = Poisson distribution P(m) = Gaussian distribution N! pm (1 − p)N−m . xn . Let us ﬁrst calculate the normalization factor A from dw dx1 · · · dxn = 1 . Then βik = βki . . To calculate the integral let us introduce the linear transform xi = aik yk . . m!(N − m)! mm −m ¯ e ¯. 1 w = A e− 2 βik xi xk .9) (A.A. (A. m! (A.2.

. respectively. the relation βa i i J= Consequently. ˆ xi xk = β−1 ik . Then we can differentiate both sides with respect to xk0 and then put xi0 = xk0 = 0. yn ) A(2π)n/2 β = = 1.19) The easiest way to calculate this integral is to calculate to assume for a while that xi = 0. β 1 exp − βik xi xk 2 (2π)n/2 . The Jacobian of the transformation x → y is just a. . Aa Thus w= Let deﬁne generalized forces as Xi = −∂S/∂xi = βik xk . . . ADDITIONAL INFORMATION to make the quadratic form (refmgs01 diagonal.s. In order that βik xi xk = yi yi = yi yk δik the relation βik ail akm = δkm (A.16) ˆ should be valid.s. is just xi Xk while the r. we get ˆ 2 = 1. ˆ ∂(y1 .17) xi βik xk e−βik xi xk /2 dx1 · · · dxn . or βik xi xk = δik . (A.18) dy e −y2 /2 n ∂(x1 .h. is δik . xn ) = |a| = a . . . (A. Then ¯ x= ¯ β (2π)n/2 xi e−βik (xi −xi0 )(xk −xk0 /2 dx1 · · · dxn = xi0 . The l.110 APPENDIX A. and ﬁrst calculate xi Xk = β (2π)n/2 (A.21) δ(x − x0 ) = ∞ for x = x0 . Thus xi Xk = δik . (A.3 The Dirac delta-function The main property is to single-out one particular value x − x0 of a variable x.h. It is deﬁned by the characteristic properties 0 for x = x0 (A. .20) A. . Denoting determinants of the matrices β and a as β and a.

0 otherwise (A. FOURIER SERIES AND TRANSFORMS but in such way that for any ε > 0 x0 +ε x0 −ε 111 δ(x − x0 ) dx = 1 .27) e−nφ dφ .22) Since delta-function has a very (inﬁnitely) sharp peak at x = x0 and a unit area B A f (x) δ(x − x0 ) dx = f (x0 ) if A < x0 < B . 0 otherwise 1 γ . c0 = a0 /2 . δ(x) = Gaussian.25) 2 2 1 δ(x) = √ e−x /2γ . 2 k=1 (A. γ 2π (A. 1 ∞ ikx e dk 2π −∞ We will also need a representation for the Kronekker’s symbol δ(x) = δn. Physically it means that γ is less than all other involved scales. (A. c±k = (ak bk )/2. δ(x) = Lorentzian.A.26) and integral.24) (A. 2 + γ2 πx (A. (A.4 Fourier Series and Transforms Fourier series Generally the Fourier series is deﬁned as an expansion of a real function in series as f (x) = ∞ a0 + ∑ (ak sin kx + bk cos kx) .4.0 = 1 2π π −π (A.23) Representations for the Dirac δ-function Let introduce a positive parameter γ and at ﬁnal stage tend it to zero.30) . γ−1 for − γ/2 < x < γ/e . Another form is complex Fourier series f (x) = k=−∞ ∑ ∞ ck eikx .28) A. (A.29) The sum of the series is periodic with the period 2π. The main representations are the following: rectangular.

k 2 k=1 k Below we plot the sum of 5 and 50 Fourier harmonics. 2π 0 If the function has a discontinuity at x = x0 then the series gives a0 = c0 = f (x0 ) = 1 [ f (x0 + 0) + f (x0 − 0)] . . 0. m = 0. . π/2 < x < π.31) At k = 0. 2π. 2 If one is willing to change the interval 2π → 2L than k has to be replaced by πk/L and the normalization factor is 1/L rather than 1/π. −π/2 < x < π/2 0. m = n = 0. (A. thus bk = 0 . ADDITIONAL INFORMATION sin mx sin nx dx = cos mx cos nx dx = sin mx cos nx dx = π δmn . n we have ak bk = 1 π ck 2π 0 cos kt dt f (t) sin kt . m = 0. It is useful to know the Parseval’s identity 1 π [ f (x)]2 dx = Fourier transform We assume L → ∞ and replace ∞ ∞ −∞ k=−∞ ∑ (.) → L 1 2π ∞ −∞ dk (.) . −π < x < −π/2. m = 0. Advantages of the Fourier series: it can represent discontinuous functions. Let us take an example of periodic function with the period 2π deﬁned as F(x) = This function is even. . for integer m. 1 2π dt f (t) . ∞ a2 0 + ∑ (a2 + b2 ) . 1 −ikt 2e k = 0. π δmn . (A.112 Since 2π 0 2π 0 2π 0 APPENDIX A.32) . g(k) = f (t) e−ikt dt . 0. ak = (2/πk) sin πk/2 . . π −π 1.

. ∞ −∞ ∞ −∞ ∞ −∞ ∞ −∞ −∞ f1 (t) f2 (t) dt → ∗ f1 (t) f2 (t) dt → g1 (k) f2 (−k) dk . dy f1 (y) f2 (x − y) → g1 (k) · g2 (k) (convolution theorem) .6 –0.35) These properties are very important for solving differential equations.34) We have several useful rules for the Fourier transform. etc.4 0. (A.A.33) The Dirac delta-function is represented as δ(x − t) = 1 2π ∞ −∞ dk eik(t−x) . ∞ f (n) (x) ≡ d n f /dxn → (ik)n g(k) . FOURIER SERIES AND TRANSFORMS 113 1 0. ∗ g1 (k) f2 (k) dk the Parseval’s theorem . (A.8 1 Figure A.4 –0.1: The completeness condition f (x) = 1 2π ∞ −∞ dk eikt ∞ −∞ dx f (x) e−ikx .2 0.6 0. If g(k) is the Fourier transform of f (x) then f (x) ≡ d f /dx → ik g(k) .2 0 0.2 –1 –0. (A.4.8 –0.4 x 0.8 0.6 0.

∂t ∂x2 The Fourier transform of this equation has the form: (Dk2 − iω) n(k. ADDITIONAL INFORMATION ∞ √1 F(ω) e−iωx dx 2π −∞ F(ω) = . 2 Dk2 − iω (2π) Now we can come back to real space-time representation: n(x. |x| > a −a2 x2 e π e−a|ω| 2 a ∞ √1 2π −∞ f (x) e−iωx dx 2 sin(aω) π ω 2 1 √ e−ω /4a a 2 Table A.t) ∂2 n(x. |x| < a f (x) = 0.a > 0 1.t) ∂2 n(x.t) +D = δ(x) δ(t) . e−Dk |t| = −∞ 2π 2 πD|t| ∞ .114 f (x) = 1 a2 +x2 APPENDIX A. Then the equation together with initial condition can be written as ∂n(x.t) = = dk ∞ dω 1 2 − iω −∞ 2π −∞ 2π Dk ∞ dk 2 2 1 e−x /4D|t| .t) = −D . ω) = 1 1 . ∂t ∂x2 Let us assume that we add a particle at time t = 0 at the point x = 0. ω) = 1 → n(k.2: Table of Fourier transform pairs Examples One-dimensional diffusion: The equation has the form ∂n(x.

− + 9 − 32 η. 2. 2.y)): g0:=eta->simplify(F2(0. 4. 5. 5.10. 2η 9 3√ − 9 − 32 η − 4 η 8 8 9 3√ + 9 − 32 η − 4 η 8 8 > plot({g0(eta). 2.eta)=0.2.xi).13. 3. 3 1√ 3 1√ 0.eta=-0.eta).9/32).2.13. g1(eta).y)): solve(F1(xi. 3. 4.g2(eta)}. g2(eta). 5.4. 4.eta)): g1:=eta->simplify(F2(x1.8 Test in mathematics > > F:=(xi. 5.5. otherwise 3 extrema > > > > > x1:=-3/8+1/8*sqrt(9-32*eta): x2:=-3/8-1/8*sqrt(9-32*eta): F2:=(xi.diff(F(y.17. − − 9 − 32 η 8 8 8 8 One extremum at eta >9/32.12.7.3.eta)->xi^4+xi^3+eta*xi^2: F1:=(xi.1. 4.diff(F1(y.g1(eta). 115 .Appendix B Maple Printouts Quick access: 1.14.1. 5.eta)->subs(y=xi.7.9.1. 3.eta). 5..eta)->subs(y=xi.eta)): g0(eta). 5.eta)): g2:=eta->simplify(F2(x2.

0).1 0.1.7.9/32). maximum at 9/32>eta>0..0.45).5 1 0.1 –0. .05 0.. plot(F(xi.25).5 –0. Two speciﬁc values: eta=0 and eta=9/32. MAPLE PRINTOUTS 2.2 0.05 0.5 2 1. otherwise minimum. x2-> minimum at 9/32>eta.25 x0-> maximum at eta <0.116 APPENDIX B. xi=-0. x1-> minimum at eta>0.0. xi=-1.15 eta 0. > plot(F(xi.

05 –0.02 0.6 –0.117 0.05 –1 –0.2 0.4 xi –0.01 0.035 0.2 –0..x=-1.8 –0.4 –0.6 –0.2 Problem 1.2 > f1:=x->ln((1-x)/(1+x)): plot(f1(x).025 0.015 0.1 0.03 0.1 0.005 –0. .2 xi 0.4 0.1).

118 APPENDIX B.4 0. 4 f2 2 0 –2 –4 0.2 0 –2 –4 –6 0.6 0.1. MAPLE PRINTOUTS 6 4 2 –1 –0.x=0.6 0.2 0.6 –0.4 –0.f2=-5.5).4 x 0.2 0.8 –0..8 x 1 > f2:=x->ln(abs(tan(Pi*x))): plot(f2(x).8 1 ..

> > ζ˜(−n˜−1) Γ(n˜ + 1) int((x^2+a^2)^(-1).1 (a) > > uax:=(x.x).x).infinity).xi=0.y).y).natural): > simplify(int(x^n*exp(-zeta*x).n=0.trig).119 > simplify(exp(4*ln(x))-(x^2+1)^2+2*x^2+1).y)->int(uax(xi...y)-diff(uay(x. x −arctan( ) y . x arctan( ) a a 1 ln(x − a) 1 ln(a + x) − 2 a 2 a int(exp(-zeta*x^2/2).3 > simplify((sin(x))^(-2)-(tan(x))^(-2)-1.infinity)). √ √ 1 2 π 2 ζ˜ Problem 2.x=0. a2:=alpha->evalc(Im(k(alpha))): eα˜ (cos(β˜) − eα˜ ) − (cos(β˜) − eα˜ )2 + sin(β˜)2 eα˜ sin(β˜) (cos(β˜) − eα˜ )2 + sin(β˜)2 Problem 1.x): ua(x.. 0 The differential is exact > ua:=(x. a2(alpha).x)). int((x^2-a^2)^(-1)..x=0.infinity): > > a1:=alpha->evalc(Re(k(alpha))): a1(alpha).y). 0 Problem 1.y).y)->x/(x^2+y^2): simplify(diff(uax(x.4 > assume(zeta>0): assume(n. 0 > assume(alpha>0): assume(beta>0): > k:=alpha->sum(exp((I*beta-alpha)*n).y)->-y/(x^2+y^2): uay:=(x.

y)+int(ux1(xi.2 General deﬁnitions > ua:=(x.. y) dξ a > ux1:=(x.eta).xi=a.x)+int(uy1(x.b)..y). MAPLE PRINTOUTS ubx:=(x.y)-diff(ucy(x...eta=b.y).x)). η) dη − x 0 0 y ∂ ∂η ubx(ξ.y)->(x+y^2): simplify(diff(ubx(x.eta).y)-diff(uby(x.y).x)+int(uby(x.y).x)+int(uy(x.eta=b. b) dξ + b y uy(x. ub := (x..y).y)->int(ubx(xi.y): ub1:=(x.y)->int(ux(xi.y)->(2*y^2-3*x): ucy:=(x.. y) dξ + 0 y uby(x.xi=0.y).xi=0.eta=0.y)..y)->-x^2: .x).x): ua1(x. ua := (x.y). η) dη + x ux(ξ. y) → a > y b > > > > x ux(ξ. ub1(x.y)->(y-x^2): uby:=(x..y).xi=a.b). ub := (x. 1 1 b (x2 − a2 ) + x3 − a3 + x2 (y − b) 3 3 1 1 a2 (y − b) + y (x2 − a2 ) + x3 − a3 3 3 0 ux2:=(x. 8y The differential is not exact Problem 2. 0 The differential is exact > ub:=(x. y) → 0 > x ubx(ξ.y)->int(uy1(a.eta).eta=b.xi=a.eta).y)->2*x*y+x^2: uy1:=(x..eta). 1 1 y x − x 3 + y3 3 3 (c) > > ucx:=(x.y).120 (b) > > APPENDIX B.y)-int(int(di > ff(ubx(xi.eta). η) dη ub:=(x..y)->int(uy(a.y)->-4*x*y: simplify(diff(ucx(x.y)).y)->y*(2-2*y): uy2:=(x.y).eta).y). η) dη dξ ub(x.xi=a. y) → uy(a.eta=0...y)-ub1(x.y)->x^2: ua1:=(x.y)+int(ux(xi.eta=b.y)->int(ux1(xi. simplify(ua1(x.x).x)).

v3 R2 T − −R T v3 + 2 α v2 − 4 α v b + 2 α b2 > factor(2*alpha*v^2-4*alpha*v*b+2*alpha*b^2).v).0. T ) → 8 T 3 − 2 3v−1 v > > plot({P(v.y)-uy2(x.5.v). RT 2α den := − + 3 (v − b)2 v > simplify(-T*num/den).y).5)..1.y)). 2 − 4 y + x2 The differential is not exact.P(v.y)->int(ux2(xi. b (2 − 2 b) (x − a) − x2 (y − b) −a2 (y − b) + y (2 − 2 y) (x − a) 2 b x − 2 b a − 2 b2 x + 2 b2 a − x2 y + x2 b + a2 y − a2 b − 2 y x + 2 y a + 2 y2 x − 2 y2 a Why? > simplify(diff(ux2(x.T))^2: den:=diff(P(T. 2 α (−v + b)2 Problem 3.y): ub2:=(x..7 > P:=(v.y).P(v.2.y).b)..121 > > > ua2:=(x.5) }.x): ua2(x.1).eta).color=black).y)->int(uy2(a.1.y).x).y)-ub2(x. ub2(x.. simplify(ua2(x.y)+int(ux2(xi.v=0.v)->R*T/(v-b)-alpha/v^2: num:=(diff(P(T.T)->(8*T/(3*v-1))-3/v^2.eta).eta=b. Problem 2.thickness=3.x)+int(uy2(x..eta=b.xi=a. .17 > > P:=(T. P := (v.xi=a.y).v).

122

APPENDIX B. MAPLE PRINTOUTS

12 10 8 6 4 2 0 –2 –4 0.6 0.7 0.8 0.9 1 1.1 1.2

v

For T=0.5 there is no stable region at all. The phse transition can exist only nes T=1. > plot({P(v,1.05),P(v,1),P(v,0.85) > },v=0.45..5.3,P=0..2,thickness=3,color=black);

2 1.8 1.6 1.4 1.2 P 1 0.8 0.6 0.4 0.2 0 1 2 v 3 4 5

>

?plot,color

Problem 3.12

> >

phi:=xi->(A/3)*(xi+eta)^2-(2*B/27)*(xi+eta)^3+(C/9)*(xi+eta)^4: series(phi(xi),xi=0,5);

1 2 1 2 2 4 1 2 2 ( A η2 − B η3 + C η4 ) + ( A η − B η2 + C η3 ) ξ + ( A − B η + C η2 ) ξ2 + 3 27 9 3 9 9 3 9 3 2 4 1 4 (− B + C η) ξ3 + C ξ 27 9 9 > c3:=eta->-2/27*B+4/9*C*eta: solve(c3(eta)=0,eta); 1B 6C > c1:=(A,eta)->2/3*A*eta-2/9*B*eta^2+4/9*C*eta^3; 2 2 4 c1 := (A, η) → A η − B η2 + C η3 3 9 9 > solve(c1(A,B/(6*C)),A); 1 B2 27 C c2:=subs({eta=B/(6*C),A=B^2/(27*C)},1/3*A-2/9*B*eta+2/3*C*eta^2); c2 := −

> >

>

delta:=xi->(C/9)*xi^4+c2*xi^2: delta1:=xik->subs(y=xi,diff(delta(y),y)): 4 3 1 B2 ξ Cξ − 9 81 C solve(4/9*C*xi^2-1/81*B^2/C,xi); 1B 1B − , 6C 6C

1 B2 162 C

delta1(xi);

>

Problem 3.13

>

P:=(v,T)->8*T/(3*v-1)-3/(T*v^2); 3 T − 2 3v−1 T v p1:=(nu,epsilon)=P(1+nu,1+epsilon)-1; 1+ε 3 p1 := (ν, ε) = 8 − −1 2 + 3 ν (1 + ε) (1 + ν)2 aux:=series(8*(1+epsilon)/(2+3*nu)-3/(1+epsilon)/(1+nu)^2-1,nu=0,4):a ux; P := (v, T ) → 8

>

> >

6 1 1 27 27 3 ) + (−6 − 6 ε + ) ν + (−9 + 9 + 9 ε) ν2 + (12 − − ε) 1+ε 1+ε 1+ε 1+ε 2 2 3 + O(ν4 ) ν > aux1:=(epsilon)->3+4*epsilon-3/(1+epsilon)+(-6-6*epsilon+6/(1+epsilon > ))*nu+(-9*1/(1+epsilon)+9+9*epsilon)*nu^2 > +(12*1/(1+epsilon)-27/2-27/2*epsilon)*nu^3; (3 + 4 ε − 3 6 1 + (−6 − 6 ε + ) ν + (−9 + 9 + 9 ε) ν2 1+ε 1+ε 1+ε 1 27 27 + (12 − − ε) ν3 1+ε 2 2 aux2:=series(aux1(epsilon),epsilon=0,2); 3 51 aux2 := − ν3 + (7 − ν3 + 18 ν2 − 12 ν) ε + O(ε2 ) 2 2 aux1 := ε → 3 + 4 ε −

>

Simpliﬁed equation of state > p:=(nu,epsilon)->-3/2*nu^3+(7-12*nu)*epsilon; 3 p := (ν, ε) → − ν3 + (7 − 12 ν) ε 2 Critical point - epsilon=0; > d1:=(nu,epsilon)->subs(y=nu,diff(p(y,epsilon),y)): 9 − ν2 − 12 ε 2 The curve is symmetric. That leads to vg=-vl=v > solve(p(v,epsilon)= p(-v,epsilon),v);; √ √ 0, 2 −2 ε, −2 −2 ε Critical exponent beta=1/2 Along the coexitence curve, vˆ2=-8epsilon. exponent gamma=1 Function of pressure > P1:=(rho,T)->8*T*rho/(3-rho)-3*rho^2/T: Near the critical point P1(1+xi,1); 8 1+ξ − 3 (1 + ξ)2 2−ξ d1(nu,epsilon);

>

12). .30)).> series(8*(1+xi)/(2-xi)-3*(1+xi)^2.3)).k)}.3)).G(12. 3 3 3 1 + ξ3 + ξ4 + ξ5 + O(ξ6 ) 2 4 8 the exponent delta=3 Problem 4.y)): d3:=k->subs(y=k..k=0.k)->(1/sqrt(3*N*Pi/8))*exp(-(k-N/4)^2*8/(3*N)).2659615201 evalf(G(120.x=0.diff(f(y). 1 π 2 f:=k->1/(1-2*I*k): d1:=k->subs(y=k..08385171464 G:=(N. d3(0).diff(d2(y).y)): d1(0).k). . .sqrt(z)).30)). G := (N.xi=0).2581036091 evalf(P(120. > > . .9 > assume(z>0): > > > int(1/sqrt(z-x^2). d2(0). 2I −8 −48 I Problem 4.y)): d2:=k->subs(y=k.10 > > P:=(N.08410441740 plot({P(12. k) → e(−8/3 (k−1/4 N)2 ) N > > 3 Nπ 8 > evalf(G(12.k)->(N!/(k!*(N-k)!))*(1/4)^k*(3/4)^(N-k): evalf(P(12.diff(d1(y).

k=0.06 0.120).k)}.2 0.0.25 0.08 0.04 0..02 0 20 40 60 k 80 100 120 .G(120.05 0 2 4 6 k 8 10 12 > plot({P(120. 0.k).15 0.1 0.

s=-infinity. −2]}] 4 2 4 > eigenvectors(QT). > > > > e(1/2 I k (−2 a˜+I k β˜)) P1:=s->(2*Pi)^(-1)*int(exp(I*(s-N*a)*k-k^2*N*beta/2).0].inf inity): P1(s). [1.. [ .infinity): simplify(P(k)).[1/8. 1.1/2. 1.1. { 6. . 1 }]. > .Problem 4. {[1. 1 N˜ a˜2 3 Problem 5.s=-infinity.1/2.1/2]]).[0. 1.d+a)). N˜ β˜ Problem 4. (−s+N˜ a˜)2 √ ) (−1/2 N˜ β˜ 1e 2 √ 2 π N˜ β˜ int(s*P1(s).x=d-a.real): assume(N>0): > P:=k->(2*Pi*beta)^(-1/2)*int(exp(-(x-a)^2/(2*beta)-I*k*x).k=-infinity.14 > > > assume(a>0): assume(d>a): assume(N>0): simplify((2*a)^(-1)*int(x^2...1 > > > > with(linalg): Q:=array([[0.13 beta = sigmaˆ2 > assume(beta>0): assume(a.. [ −1 −3 1 .infinity)-N^2*a^2. N˜ a˜ int(s^2*P1(s). 0 1 0 1 1 3 Q := 8 2 8 1 1 0 2 2 QT:=transpose(Q): eigenvectors(Q). 1]}]. . 1.x=-infinity. 1 d˜2 + a˜2 3 simplify(N*(d^2+1/3*a^2)+N*(N-1)*d^2-N^2*d^2).3/8]. 1. {[4.infinity).

8. 6 −12 6 −3 −3 3 P2 := 2 2 1 −2 1 P3:=array([[l3[1]*r3[1]. l1[2]*r1[3]. 1. l2[2]*r2[2]. -2]).1 2 r3 := [4. 6] l2 := [1. [1.[l3[1]*r3[3]. 1] l3 := [1.l1[3]*r1[1]]. 6]}] 4 4 > > r1:=vector([1.l2[3]*r2[2]]. {[1. −2. 1]). l2[2]*r2[1]. 1. 1]}].l3[3]*r3[2]]. 1.l2[3]*r2[3]]]). > > > > > > > . 2. r3:=vector([4. l3[2]*r3[1]. {[1. 1]). −2] > > l1:=vector([1. 8. l1[2]*r1[2]. 1. l2:=vector([1. l3[2]*r3[2]. 1] r2 := 6. [l1[1]*r1[2]. 4 8 −12 2 −3 P3 := 1 −2 −4 6 P:=s->(1/n1)*P1+(1/n2)*(-1/4)^s*P2+(1/n3)*(1/4)^s*P3. [l3[1]*r3[2]. Question (a) −1 s 1 s P1 ( 4 ) P2 ( 4 ) P3 P := s → + + n1 n2 n3 Pr:=s->evalm(P(s)): Pr(s). −2. [ . [l2[1]*r2[2]. r2:=vector([6. Pr(0). -3]).[l1[1]*r1[3]. 8. −3]}]. 1. {[1.l1[3]*r1[2]]. l1 := [1.l3[3]*r3[1]]. 1]). 1. -2. 2. −3] n1:=15: n2:=10: n3:=12: > > > P1:=array([[l1[1]*r1[1].[l2[1]*r2[3].l2[3]*r2[1]]. 2. r1 := [1. -3/2. −3 . l2[2]*r2[3]. l3:=vector([1. 1. 1 8 6 P1 := 1 8 6 1 8 6 P2:=array([[l2[1]*r2[1]. 6]).1 −1 [ . l3[2]*r3[3].l1[3]*r1[3]]]).l3[3]*r3[3]]]). l1[2]*r1[1].

0] PS:=s->evalm(p0&*P(s)): PS(s). after many steps 2/5 1 9 3 .3). > > > > > > > 79 33 81 11 3 (−1)(1+s) 4(−s) − 16(−s) + (−1)(1+2 s) 16(−s) + (−1)(1+s) 16(−s) − 4(−s) 50 4 100 2 2 841 + 25 > corr(0).1.0]). 16 . + ( )s + ( )s .. PS(infinity). 29 3 −1 3 1 − ( )s − ( )s 5 10 4 2 4 corr:=s->simplify(sum(n^2*sum(m^2*PS(s)[m]*Pr(s)[m. Answer: after 2 steps 3/8. Moment 29 3 −1 3 1 − ( )s − ( )s 5 10 4 2 4 av:=sum(m^2*PS(s)[m]. 2 3 −1 s 1 + ( ) − ( )s 5 5 4 4 2 3 −1 s 1 1 s − ( ) − ( ) 5 20 4 4 4 2 1 −1 1 1 + ( )s + ( )s 5 10 4 2 4 Question (b) 1 3 −1 1 1 8 3 −1 1 1 2 3 −1 1 1 − ( )s + ( )s ..3)): corr(s). 1. p0 := [0.n]. m=1. .3). 16 16 8 1 8 2 .n=1. 15 15 5 av:=s->1*PS(s)[1]+4*PS(s)[2]+9*PS(s)[3]: av(s).m=1.. . − ( )s − ( )s 15 20 4 12 4 15 10 4 6 4 5 20 4 4 4 PS(2). > > 1 3 −1 s 1 1 s 8 6 −1 2 1 − ( )s + ( )s 15 + 5 ( 4 ) + 3 ( 4 ) 15 5 4 3 4 1 3 −1 s 1 1 s 8 3 −1 s 1 1 s 15 − 20 ( 4 ) + 12 ( 4 ) 15 + 10 ( 4 ) + 6 ( 4 ) 1 1 −1 1 1 8 1 −1 1 1 + ( )s − ( )s − ( )s − ( )s 15 10 4 6 4 15 5 4 3 4 1 0 0 0 1 0 0 0 1 p0:=vector([0.

− − I 2 } 6 3 2 4 l[1][3]. 1 1 1 √ − + I 2 2 4 1 1 √ − − I 2 2 4 > l[1][1]. r[3][1]. 1 1 0 2 2 Q := 0 0 1 3 1 0 4 4 QT:=transpose(Q): r:=eigenvectors(Q): l:=eigenvectors(QT): r[1][1]. 1.[0. r[2][3].1/2. 1. l[2][1].1/4.1/2]. {[1. r[3][3].2 > with(linalg): Warning. 1 1 1 √ − + I 2 2 4 1 1 √ − − I 2 2 4 > r[1][3]. l[3][1].1]. the protected names norm and trace have been redefined and unprotected > > > > > Q:=array([[0. 841 25 Problem 5.0. r[2][1]. 6 8 { . 1. l[2][3]. 1. } 5 5 > .> corr(infinity).[3/4.0]]). 1]} 1 1 √ 1 1 √ { − − I 2. l[3][3]. − + I 2 } 6 3 2 4 1 1 √ 1 1 √ { − + I 2.

l3[3]*r3[1]].l1[3]*r1[3]]]). -2/3-1/3*I*sqrt(2)]): n1:=evalm(l1&*r1): n2:=evalc(evalm(l2&*r2)): n3:=evalc(evalm(l3&*r3)): Probability matrices. -1/3-1/3*I*sqrt(2).> > > > > > > 1 1 √ 2 1 √ { 1. 1. [l2[1]*r2[2]. l1[2]*r1[1]. − + I 2 } 3 3 3 3 1 1 √ 2 1 √ { 1. > l1[2]*r1[2]. [l1[1]*r1[2].[l1[1]*r1[3].l2[3]*r2[1]]. -1/2-1/4*I*sqrt(2)]): l1:=vector([6/5. Question (a) P1 r21 s P2 r3 1 s P3 P := s → + + n1 n2 n3 Pr:=s->evalm(P(s)): . − − I 2 } 3 3 3 3 r1:=vector([1. 1.l1[3]*r1[2]]. 1. 1 1 √ 1 1 √ 1 1 √ 2 1 √ 1 1 √ − + I 2 (− + I 2) (− + I 2) (− − I 2) (− + I 2) 6 3 3 3 6 3 3 3 6 3 2 1 √ 1 1 √ P3 := − − I 2 1 − + I 2 3 3 3 3 1 1 √ 1 1 √ 1 1 √ 2 1 √ 1 1 √ − − I 2 (− + I 2) (− − I 2) (− − I 2) (− − I 2) 2 4 3 3 2 4 3 3 2 4 > > > P:=s->(1/n1)*P1+(1/n2)*(r[2][1])^s*P2+(1/n3)*(r[3][1])^s*P3. 1. [l3[1]*r3[2]. 1]): r2:=vector([-1/6-1/3*I*sqrt(2).l2[3]*r2[3]]]). 8 6 5 1 5 6 8 P1 := 5 1 5 6 8 1 5 5 > P2:=array([[l2[1]*r2[1]. − + I 2.[l3[1]*r3[3]. l3[2]*r3[3]. 1 1 √ 1 1 √ 1 1 √ 2 1 √ 1 1 √ − 6 − 3 I 2 (− 3 − 3 I 2) (− 6 − 3 I 2) (− 3 + 3 I 2) (− 6 − 3 I 2) 2 1 √ 1 1 √ P2 := − + I 2 1 − − I 2 3 3 3 3 1 1 √ 1 1 √ 1 1 √ 2 1 √ 1 1 √ − + I 2 (− − I 2) (− + I 2) (− + I 2) (− + I 2) 2 4 3 3 2 4 3 3 2 4 > P3:=array([[l3[1]*r3[1]. -1/3+1/3*I*sqrt(2). -1/2+1/4*I*sqrt(2)]): r3:=vector([-1/6+1/3*I*sqrt(2).l2[3]*r2[2]]. l2[2]*r2[3]. -2/3+1/3*I*sqrt(2)]): l3:=vector([1.l1[3]*r1[1]]. l3[2]*r3[1].l3[3]*r3[2]]. 8/5]): l2:=vector([1. − − I 2. > l2[2]*r2[2]. l2[2]*r2[1].l3[3]*r3[3]]]). > l3[2]*r3[2].[l2[1]*r2[3]. > P1:=array([[l1[1]*r1[1]. l1[2]*r1[3].

3 > with(linalg): Warning.2/3. 0. evalc(PS(2)[1]).0.1/2. l[3][1]. −1 3 −2 3 1 l[1][1]. 1 1 0 2 2 1 2 Q := 3 0 3 1 2 0 3 3 QT:=transpose(Q): r:=eigenvectors(Q): l:=eigenvectors(QT): r[1][1].0]). 1 1 √ 1 1 √ 1 1 √ 1 1 √ (− + I 2)s (− − I 2) (− − I 2)s (− + I 2) 6 2 4 6 3 2 4 6 3 + + √ √ 1 1 19 − −I 2 − +I 2 3 3 evalc(PS(infinity)[1]).0]]). 0] PS:=s->evalm(p0&*P(s)): PS(s)[1].2/3].> p0:=vector([1. 6 19 3 8 > > Problem 5.[1/3. l[2][1].1/2]. r[2][1]. r[3][1].0. p0 := [1. −1 3 1 . the protected names norm and trace have been redefined and unprotected > > > > > > Q:=array([[0.[1/3.

[l3[1]*r3[2]. 3/2]): l2:=vector([-2. −1.−2 3 > r[3][3]. -1.[l1[1]*r1[3]. 1.[l3[1]*r3[3]. 1]): l3:=vector([0. [l2[1]*r2[2]. } 2 2 {[−2. −1. 1. 1]} > l[2][3].l2[3]*r2[1]]. [l1[1]*r1[2]. Question (a) P1 r3 1 s P2 r21 s P3 P := s → + + n1 n2 n3 . l3[2]*r3[1]. 0 0 0 1 −1 P3 := 0 0 −1 1 > > P:=s->(1/n1)*P1+(1/n2)*(r[3][1])^s*P2+(1/n3)*(r[2][1])^s*P3. 1]): l1:=vector([1.l1[3]*r1[2]].l3[3]*r3[3]]]). 1]): r3:=vector([0. 6 −3 −3 1 1 P2 := −2 −2 1 1 > P3:=array([[l3[1]*r3[1]. 1]} r1:=vector([1.l2[3]*r2[2]].l2[3]*r2[3]]]). 1. > l2[2]*r2[2]. {[1. > l1[2]*r1[2]. 1. . 1]} {[0. l[3][3]. 1]} {[0. 3 3 1 2 2 1 3 3 P1 := 2 2 3 3 1 2 2 > P2:=array([[l2[1]*r2[1]. 3/2. l1[2]*r1[1]. l2[2]*r2[1].[l2[1]*r2[3]. l1[2]*r1[3].l3[3]*r3[2]]. > l3[2]*r3[2]. r[2][3]. r[1][3]. 1. l2[2]*r2[3]. 3 3 { 1. 1.l1[3]*r1[1]]. > P1:=array([[l1[1]*r1[1]. l3[2]*r3[3].l3[3]*r3[1]]. l[1][3].l1[3]*r1[3]]]). 1]): n1:=evalm(l1&*r1): n2:=evalc(evalm(l2&*r2)): n3:=evalc(evalm(l3&*r3)): > > > > > Probability matrices. -1. 1]} {[−3. 1]): r2:=vector([-3.

1 /2+1/2*(sin(Pi*s/2))^2.4 > with(linalg): Warning.0.− + ( ) .[1/4 +1/2*(sin(Pi*s/2))^2.[1/2*(cos(Pi*s/2))^2. PS(0). 25 5 (−2 s) + 3 (−1)(2 s) 2(2 s) + (−1)s 3(1−s) 2(−1+s) 4 4 corr(0).3).. 9 25 4 Problem 5. 2 2 [0.> > Pr:=s->evalm(P(s)): Conditional probability 1 0 0 p0:=vector([0.1/4+1/2*(cos(Pi*s/2))^2]. 0 1 1 −2 s − ( ) 2 2 3 1 1 −2 s + ( ) 2 2 3 > > PS:=s->evalm(p0&*P(s)): Probability vector 1 1 −2 s 1 1 −2 s 7 1 −2 s − ( ) . 5 1 −2 s + ( ) 2 2 3 corr:=s->simplify(sum(n*sum(m*PS(s)[m]*Pr(s)[m.(cos(Pi*s/2))^2.n=1.3)): corr(s).0.. 1 1 0. 0 1 1 −2 s + ( ) 2 2 3 1 1 −2 s − ( ) 2 2 3 PS(s). 1] > > > > moment:=1*PS(s)[1]+2*PS(s)[2]+3*PS(s)[3]: moment. . + ( ) 4 4 3 8 8 3 8 8 3 > PS(infinity).(sin(Pi*s/2))^2]. the protected names norm and trace have been redefined and unprotected > > > Q0:=s->array([[0. m=1.0]]): Q0(s). 0.n].1]): Pr(s). corr(infinity). .

l[2][1].[1/4.1/2. Q := > > > > 0 1 4 1 2 1 0 1 2 0 3 4 0 QT:=transpose(Q): r:=eigenvectors(Q): l:=eigenvectors(QT): r[1][1]. r[3][1]. 1 1 √ 1 1 √ { 1. − + I 2 } 2 4 6 3 . we get > Q:=array([[0.[1/2. r[1][3].1.0]. r[2][1]. > r[3][3]. r[2][3].3/4].0]]).0. l[3][1]. > 0 1 1 + sin( 1 π s)2 0 4 2 2 1 1 1 1 1 cos( π s)2 + sin( π s)2 0 2 2 2 2 2 simplify(Q0(2*s)). − − I 2. 1 1 1 √ − + I 2 2 4 1 1 √ − − I 2 2 4 1 1 √ − + I 2 2 4 1 1 √ − − I 2 2 4 1 > l[1][1]. 0 cos(π s)2 1 − cos(π s)2 3 1 1 1 − cos(π s)2 0 + cos(π s)2 4 2 4 2 1 1 cos(π s)2 1 − cos(π s)2 0 2 2 1 cos( π s)2 2 1 sin( π s)2 2 1 1 1 + cos( π s)2 4 2 2 Since [cos(Pi*s)]ˆ2=1.

. 1]): n1:=evalm(l1&*r1): n2:=evalc(evalm(l2&*r2)): n3:=evalc(evalm(l3&*r3)): Probability matrices. -1/6-1/3*I*sqrt(2)]): r3:=vector([1.l1[3]*r1[2]]. l2[2]*r2[3]. -1/2+1/4*I*sqrt(2).[l3[1]*r3[3]. [l2[1]*r2[2]. 1. l3[2]*r3[2]. 1]): l3:=vector([-1/3+1/3*I*sqrt(2). 1]} 1 1 √ 1 1 √ { 1. } 5 5 2 1 √ 1 1 √ { − − I 2.l1[3]*r1[3]]]).[l2[1]*r2[3]. 1 } 3 3 3 3 2 1 √ 1 1 √ { − + I 2. 8 6 1 5 5 1 8 6 P1 := 5 5 8 6 1 5 5 P2:=array([[l2[1]*r2[1]. [l1[1]*r1[2]. l2[2]*r2[1]. -2/3-1/3*I*sqrt(2). 1 } 3 3 3 3 > > > > > > > r1:=vector([1. > > P1:=array([[l1[1]*r1[1]. 1]): r2:=vector([1. 8 6 { 1.l2[3]*r2[3]]]). -1/6+1/3*I*sqrt(2)]): l1:=vector([1.l3[3]*r3[2]].l1[3]*r1[1]]. 6/5]): l2:=vector([-1/3-1/3*I*sqrt(2). − + I 2. l3[2]*r3[1]. l2[2]*r2[2]. l3[2]*r3[3].[l1[1]*r1[3]. l[2][3].{[1. − − I 2.l2[3]*r2[1]]. -2/3+1/3*I*sqrt(2). 8/5. [l3[1]*r3[2]. − + I 2. l[1][3]. 1 1 √ 2 1 √ − − I 2 − + I 2 1 3 3 3 3 √ √ √ √ √ (− 1 − 1 I 2) (− 1 + 1 I 2) (− 2 + 1 I 2) (− 1 + 1 I 2) − 1 + 1 I 2 P2 := 3 3 2 4 3 3 2 4 2 4 1 1 √ 1 1 √ 2 1 √ 1 1 √ 1 1 √ (− − I 2) (− − I 2) (− + I 2) (− − I 2) − − I 2 3 3 6 3 3 3 6 3 6 3 > > > > P3:=array([[l3[1]*r3[1]. . l1[2]*r1[2].l3[3]*r3[1]].l3[3]*r3[3]]]).l2[3]*r2[2]]. l1[2]*r1[1]. − − I 2 } 2 4 6 3 > l[3][3]. 1. l1[2]*r1[3]. -1/2-1/4*I*sqrt(2).

1. P3 := 1 1 √ 2 1 √ − − I 2 1 − + I 2 3 3 3 3 1 1 √ 1 1 √ 2 1 √ 1 1 √ 1 1 √ (− + I 2) (− − I 2) (− − I 2) (− − I 2) − − I 2 3 3 2 4 3 3 2 4 2 4 1 1 √ 1 1 √ 2 1 √ 1 1 √ 1 1 √ (− + I 2) (− + I 2) (− − I 2) (− + I 2) − + I 2 3 3 6 3 3 3 6 3 6 3 > > P:=s->(1/n1)*P1+(1/n2)*(r[1][1])^(2*s)*P2+(1/n3)*(r[2][1])^(2*s)*P3.real): assume(s>0). 1 1 √ 1 1 √ 1 1 √ 1 1 √ 1 1 √ (− − I 2) (− + I 2) (− + I 2)(2 s) (− + I 2) (− − I 2) 5 3 3 2 4 2 4 3 3 2 4 + + √ √ 1 1 19 − −I 2 − +I 2 3 3 > PS2(s)[2].5 > assume(N. Question (a) P := s → P1 r11 (2 s) P2 r21 (2 s) P3 + + n1 n2 n3 Pr:=s->evalm(P(s)): Conditional probability > p01:=vector([1.s. assume(n.natural): assume(k.0.k=1. Problem 5.0]): > PS1:=s->evalm(p01&*P(s)): PS2:=s->evalm(p02&*P(s)): > PS1(s)[1].0]): p02:=vector([0.N)): . 2 1 √ 1 1 √ 1 1 √ 2 1 √ 1 1 √ (− + I 2) (− + I 2) (− + I 2)(2 s) (− − I 2) (− − I 2) 8 3 3 2 4 2 4 3 3 2 4 + + √ √ 1 1 19 − −I 2 − +I 2 3 3 > > The results can be transformed to polar form by the command polar. 1 1 √ 1 1 √ 1 1 √ − − I 2 (− + I 2)(2 s) (− + I 2) 5 2 4 3 3 + 3 3√ + √ 1 1 19 − −I 2 − +I 2 3 3 > PS1(s)[2].integer): > P:=(n.. 1 1 √ 2 1 √ 2 1 √ − + I 2 (− + I 2)(2 s) (− − I 2) 8 2 4 3 3 + 3 3√ + √ 1 1 19 − −I 2 − +I 2 3 3 > PS2(s)[1].N)->(1/(2*N+1))*(1+2*sum(cos(2*Pi*k*n/(2*N+1))*exp(-2*s*(sin( > Pi*k/(2*N+1)))^2).

1/3. {[0. {[0.1/2.2223516301.. 1 }].3. 1 1 0 2 2 1 1 w := 0 3 3 1 1 0 3 3 M:=matrix(3. 1 1 −1 2 2 1 1 M := 3 −1 3 1 1 −1 3 3 MT:=transpose(M): eigenvectors(M).1/2.-1]).2223516301. { − − 13. [− − 13. { − + 13. { − − 13. 1. 1+2 k˜=1 > ∑ cos(2 2 N˜ + 1 ) e(−2 s˜ sin( 2 N˜+1 ) ) 2 N˜ π k˜ n˜ π k˜ 2 N˜ + 1 Pm:=[seq(P(i.8 > with(linalg): Warning. [− + 13.1/3. −1.1/3.i=-2. [.-1.1/2.0. .1/3. 1 }].3. 1.-5/6-1/6*sqrt(13).2). 1. −4 5 1√ 1 1√ . 1 }] [− − 6 6 3 3 lambda:=vector([-5/6+1/6*sqrt(13). . 1. 1. 1.[-1.0]). . −1. 6 6 2 2 6 6 2 2 −4 [ . 1.> P(n. 1.1/3.3111679263.-4/3]): [ > > . 1 }].2)]: evalf(Pm).2.s. 1.1/3.N). { − + 13. 5 1√ 1 1√ 5 1√ 1 1√ [− + 13.1220644066. the protected names norm and trace have been redefined and unprotected > > > > w:=matrix(3. 1]}]. 3 6 6 3 3 1 1√ 5 1√ 13.1/3.1/3.1/2.1220644066] Problem 5. . 1]}] 3 eigenvectors(MT).[0. 1.

-1.. 1. 1. l2:=vector([-1/3-1/3*sqrt(13).> > r1:=vector([-1/2+1/2*sqrt(13). 1 l1 := − + 3 3 1 1√ l2 := − − 13.k=1. 1.k=1. 1]).3): evalf(Q1+Q2+Q3)..3): Q3:=-sum((1/n3)*r1[1]*l1[k]/lambda[1]. r3:=vector([0. 1. 1 1√ r1 := − + 13. 1. 1. 1 2 2 1 1√ r2 := − − 13. First passage time 13. 1] l1:=vector([-1/3+1/3*sqrt(13)..3): Q2:=-sum((1/n2)*r1[1]*l1[k]/lambda[2]. -1. 1 2 2 r3 := [0. 1]).64536110 : > . 1]). −1.k=1. 1 1√ 13. 1]). r2:=vector([-1/2-1/2*sqrt(13). 1. 1]). 1. l3:=vector([0. −1. 1]). 1 3 3 l3 := [0. 1] 1 n2:=evalm(l2&*r2): > > > n1:=evalm(l1&*r1): n3:=evalm(l3&*r3): > > > > > Q1:=-sum((1/n1)*r1[1]*l1[k]/lambda[1].

.

141 . 1998. E. Reichl A modern course in statistical physics.Bibliography [1] L.

- Transmisor de presion
- sedimentacion
- Informe de practica.docx
- Diseño Espesador en 3D
- Ensayos No Destructivos Utilizando Técnicas de Ultrasonido
- introducción a la mecanica
- Ejercicio Grupal Campo de Concentracion
- Monografia
- Trabajo Experimental 1
- Precios Doc 3
- BigEasyDriver_UserManal
- 7 - PLC_parte 1c - 2ºSem2013 161113
- Valores y Virturdes
- Luis.miel
- Luis.miel
- guia-8--FP
- So Lidos
- Espectro Del Hidrogeno

- Solutions Manual for Introduction to Modern Statistical Mechanicsby Song Bin Cui
- Sakurai J.J. Modern Quantum Mechanics (AW, 1994)(802dpi)(T)(513s)by jamesbondrus
- Fundamental University Physics Vol.3 - Quantum and Statistical Physics (Alonso & Finn) Hazemsakeekby António Carneiro
- Sakurai quantum mechanics solutions 1by glog

- Reichl L.E. a Modern Course in Statistical Physics (2ed, Wiley, 1998)
- Sakurai_Ch4_1_2_3_4_8_10_12
- Introduction to quantum field theory, based on the book by Peskin and Schroeder
- Introduction to Statistical Physics
- Solved Problems of Jackson's Electrodynamics 01
- Classical mechanics problems - MIT course
- Stat Mec Pathria Solutions
- Statistical Physics
- Sakurai Solutions 5-1 5-2
- Kerson Huang Solution_Intro_Statistical Phy
- Classical Electrodynamics 3rd Ed J.D. Jackson - Solutions - 214 Pg
- Problemas Cap 3 Todos
- Kubo - Statistical Mechanics
- Quantum
- 40181627-Jackson-Solutions
- Shankar - Principles of quantum mechanics
- Solutions Manual for Introduction to Modern Statistical Mechanics
- Sakurai J.J. Modern Quantum Mechanics (AW, 1994)(802dpi)(T)(513s)
- Fundamental University Physics Vol.3 - Quantum and Statistical Physics (Alonso & Finn) Hazemsakeek
- Sakurai quantum mechanics solutions 1
- Sakurai quantum mechanics solutions 4
- David Chandler, David Wu Solutions Manual for Introduction to Modern Statistical Mechanics 19871
- Sakurai quantum mechanics solutions 3
- Statistical Mechanics - Pathria Homework 5
- Statistical Mechanics - Pathria Homework 2
- Statistical Mechanics - Pathria Homework 1
- Thermody-2
- 02 Thermodynamic Process Practice Problem1
- 02-Thermodynamic Process (Practice Problem)
- Thermal Physics - 6
- Problemas Resueltos Del Reichl.tmp

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd