ANSWERS TO THE PROBLEMS IN

A First Course in String Theory

Solved by Zan Pan

A First Course in String Theory
by Barton Zwiebach, 2004
Cambridge University Press
ANSWERS TO THE PROBLEMS IN
A First Course in String Theory
1st edition, August 2010
Zan Pan
Department of Modern Physics
University of Science and Technology of China
E-mail address: panzan@mail.ustc.edu.cn

This document is typeset by LATEX. Thanks to Donald Knuth, Leslie Lamport and many others
who contributed to creating this wonderful system. The color used here is blue!40!green.
Contents

I BASICS 1
1 A Brief Introduction 3

2 Special Relativity and Extra Dimensions 5

3 Electromagnetism and Gravitation in Various Dimensions 9

4 Nonrelativistic Strings 15

5 The Relativistic Point Particle 17

6 Relativistic Strings 19

7 String Parameterization and Classical Motion 23

8 World-sheet Currents 27

9 Light-cone Relativistic Strings 31

10 Light-cone Fields and Particles 35

11 The Relativistic Quantum Point Particle 39

12 Relativistic Quantum Open Strings 43

13 Relativistic Quantum Closed Strings 49

II DEVELOPMENTS 53
14 D-branes and Gauge Fields 55

Appendix A Elements of E8 59
A.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Appendix B Modular Forms 61

B.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Bibliography 63

i
ii CONTENTS
 Part I

BASICS

1
Chapter 1

A Brief Introduction

 Summary and Supplement

1. Two significant unifications
• Maxwell’s equations : unification of electricity and magnetism
• Glashow-Weinberg-Salam model : unification of electromagnetism and the weak force
2. Two successful quantum theories
• QED: the quantization of electromagnetism
• QCD: the quantization of strong color force
3. The Standard Model
• Quarks: (u, d), (c, s) and (t, b)
• Leptons: (νe , e− ), (νµ , µ− ) and (ντ , τ − )
• Force carriers: W ± , Z 0 , eight gluons and photon
4. Quantum gravity
• String theory
• Loop quantum gravity
• Causal dynamical triangulation
• Canonical general relativity
• Noncommutative geometry
• Twistor theory
5. String Theory
• Each particle is identified as a particular vibrational mode of an elementary string.
• String theory does not have adjustable dimensionless parameters.
• The dimensionality of spacetime is fixed.
• There are two kinds of strings: open strings and closed strings.
• The graviton appears as a vibrational mode of closed strings.
• Bosonic strings live in 26 dimensions while superstrings in 10.
• M-theory is eleven-dimensional.
• Our world is part of the D-branes.
• The AdS/CFT correspondence is a remarkable physical equivalence between a certain four-
dimensional gauge theory and a closed superstring theory.
• String theory has made good strides towards a statistical mechanics interpretation of black
hole entropy.

3
4 CHAPTER 1. A BRIEF INTRODUCTION
Chapter 2

Special Relativity and Extra

Dimensions

 Summary and Supplement

1. Intervals and Lorentz transformations

xµ = (x0 , x1 , x2 , x3 ) = (ct, x, y, z) (2.1)

a · b = aµ bµ = ηµν aµ bν = −a0 b0 + a1 b1 + a2 b2 + a3 b3 (2.2)

 
γ −γβ 0 0
−γβ γ 0 0
x′µ = Lµ xν , Lµ =   0
 (2.3)
0 1 0
0 0 0 1

2. Light-cone coordinates
1 0 1
x+ = (x + x1 ), x− = (x0 − x1 ) (2.4)
2 2
 
0 −1 0 0
−1 0 0 0
− ds2 = η̂µν dxµ dxν , η̂µν = 
0
 (2.5)
0 1 0
0 0 0 1
a · b = a+ b + + a− b − + a2 b 2 + a3 b 3 , a+ = −a− , a− = −a+ (2.6)
1 0 1 0
p+ = (p + p1 ) = −p− , p− = (p − p1 ) = −p+ (2.7)
2 2
 Quick Calculations
2.1 With the relation γ = (1 − β 2 )−1/2 , it is easy to verify that

(x′0 )2 − (x′1 )2 − (x′2 )2 − (x′3 )2 = γ 2 (x0 − βx1 )2 − (−βx0 + x1 )2 − (x2 )2 − (x3 )2

 
(2.8)
= (x0 )2 − (x1 )2 − (x2 )2 − (x3 )2

2.2 Using Eq. (2.2), we can obtain

a′µ b′µ = −γ 2 (a0 − βa1 )(b0 − βb1 ) + γ 2 (−βa0 + a1 )(−βb0 + b1 ) + a2 b2 + a3 b3

= −a0 b0 + a1 b1 + a2 b2 + a3 b3 (2.9)
= aµ b µ

2.3 Suppose that x+ = √12 (x0 + x1 ) = γ(x0 − βx1 ), then we have γ = √1

2
and β = −1. Therefore,
such a transformation does not exist.
E2
2.4 c2 − p~ · ~p = γ 2 m2 c2 − γ 2 m2 v 2 = γ 2 (1 − β 2 )m2 c2 = m2 c2 .
2.5 Consider the plain (x, y) with the identification (x, y) ∼ (x + 2πR, y + 2πR). The resulting space
is a two-dimensional torus.

5
6 CHAPTER 2. SPECIAL RELATIVITY AND EXTRA DIMENSIONS

 Solutions to the Problems

1
2.1 (a) 1 C = (109 × 8.99 × 109 ) 2 esu = 3 × 109 esu.
(b) If we set the Bolzmann’s contant kB = 1, then [K] = [N · m] = M L2 T −2 .
(c) [e] = [esu] = M 1/2 L3/2 T −1 , [~] = M L2 T −1 , [c] = LT −1 , so we have

e2 (1.602 × 10−19 × 3 × 109 )2 × 10−9 1

= −34 8
≃
~c 1.054 × 10 × 3 × 10 137
In the Heaviside-Lorentz system of units, such a dimensionless number becomes e2 /4π~c.
2.2 (a) According to Eq. (2.4), we have
s
1 1 1−β +
x′+
= √ (x′0 + x′1 ) = √ γ(1 − β)(x0 + x1 ) = x (2.10a)
2 2 1+β
s
1 1 1+β −
x′− = √ (x′0 − x′1 ) = √ γ(1 + β)(x0 − x1 ) = x (2.10b)
2 2 1−β
x′2 = x2 (2.10c)
′3 3
x =x (2.10d)

(b) From Eq. (2.4), we can obtain x0 = √1 (x+ + x− ), x1 = √1 (x+ − x− ). Then,

2 2

1 1 + cos θ + 1 − cos θ − sin θ 2

x′+ = √ (x0 + cos θ x1 − sin θ x2 ) = x + x − √ x (2.11a)
2 2 2 2
1 1 − cos θ + 1 + cos θ − sin θ 2
x′− = √ (x0 − cos θ x1 + sin θ x2 ) = x + x + √ x (2.11b)
2 2 2 2
sin θ
x′2 = sin θ x1 + cos θ x2 = √ (x+ + x− ) + cos θ x2 (2.11c)
2
x′3 = x3 (2.11d)

(c) This case differs strikingly from that of (a).

1   γ +1 + γ−1 − γβ
x′+ = √ γ(x0 − βx3 ) + x1 = x + x − √ x3 (2.12a)
2 2 2 2
1   γ −1 + γ+1 − γβ 3
x′− = √ γ(x0 − βx3 ) − x1 = x + x − √ x (2.12b)
2 2 2 2
x′2 = x2 (2.12c)
γβ
x′3 = γ(−βx0 + x3 ) = − √ (x+ + x− ) + γx3 (2.12d)
2

2.3 (a) With the relations a0 = −a0 , a1 = a1 , a2 = a2 and a3 = a3 , we can obtain

a′0 = γ(a0 + βa1 ), a′1 = γ(βa0 + a1 ), a′2 = a2 , a′3 = a3 (2.13)

(b) We refer to the inverse Lorentz transformation.

∂ ∂ ∂x0 ∂ ∂x1 ∂ ∂
= + = γ( 0 + β 1 ) (2.14a)
∂x′0 ∂x0 ∂x′0 ∂x1 ∂x′0 ∂x ∂x
∂ ∂ ∂x0 ∂ ∂x1 ∂ ∂
= 0
+ 1
= γ(β 0 + 1 ) (2.14b)
∂x′1 ∂x ∂x ′1 ∂x ∂x ′1 ∂x ∂x
∂ ∂
= (2.14c)
∂x′2 ∂x2
∂ ∂
= (2.14d)
∂x′3 ∂x3
(c) Using the first quantization method, we have
E 1 ∂ ∂ ∂ ∂ ~ ∂
pµ = (− , px , py , pz ) = (− i~ , −i~ , −i~ , −i~ ) = (2.15)
c c ∂t ∂x ∂y ∂z i ∂xµ
 7

2.4 (a) The identification yields a semi-circle. There are two fixed points: x = 0, 1. A foundamental
domain can be chosen as [0, 1].
(b) As we know, x = ±1 are identified and y = ±1 are identified. Then it is obvious that there
are four fixed points: (0, 0), (0, 1), (1, 0) and (1, 1).

2.5 Let tan θ = β, then we can obtain the following from the inverse Lorentz transformation:

x0 = tan θ x1 + γ −1 x′0 , x0 = cot θ(x1 − γ −1 x′1 ). (2.16)

It is obvious that the x′0 and x′1 axes appear in the original spacetime diagram as oblique axes.
The angle between the x′0 axis and the x0 aixs is equal to that between the x′1 axis and the x1
axis, i.e. θ = arctan β. Diagrams for the axes are drawn in Fig. 2.1.

x0 x′0 x′0 x0

θ x′1 x′1 θ

θ θ
O x1 O x1

Fig. 2.1 The left illustrates how the axes appear when β > 0, while the right is for β < 0.

√
2.6 (a) In light-cone coordinates, it can be written as (x+ , x− ) ∼ (x+ , x− − 2 2πR).
(b) With the relations ct = γ(ct′ + βx′ ) and x = γ(βct′ + x′ ), we can obtain
 ′ s
βct + x′
  ′
βct + x′
    ′  ′  
2π R x x 1+β R
∼ + ⇒ ∼ + 2π (2.17)
ct′ + βx′ ct′ + βx′ γ −R ct′ ct′ 1 − β −R
p
(c) From the following, we can see that the velocity parameter β = −R/ R2 + Rs2 and that the
compactification radius Rc = Rs .
 ′ p
βct + x′
  ′
βct + x′
 p   ′  ′ 
2π R2 + Rs2 x x R2 +pRs2 + βR
∼ + ⇒ ∼ + 2πγ
ct′ + βx′ ct′ + βx′ γ −R ct′ ct′ −R − β R2 + Rs2
p
(d) For example, (0, 0) and (−2πR, 2π R2 + Rs2 ) are related by the identification.
(e) Lightlike compactification with p
Radius R arises by boosting a standard compactification with
radius Rs with Lorentz factor γ ∼ R2 + Rs2 /Rs , in the limits as Rs → 0.

2.7 (a) Using the result in 2.2 (a), we can rewrite the identification (x0 , x1 ) ∼ (x′0 , x′1 ) as
s
1+β
(x+ , x− ) ∼ (e−λ x+ , eλ x− ), where eλ = (2.18)
1−β

The range of λ is (−∞, ∞) and the orbifold fixed point is (0, 0).
(b) The spacetime diagram refers to Fig. 2.2. From Eq. (2.18), we see that e−λ x+ ·eλ x− = x+ x− .
So the idenfitication above relates points on the curves of x+ x− = a2 .
2
(c) −ds2 = −2dx+ dx− = −2dx+ d( xa+ ) = 2( xa+ )2 (dx+ )2 > 0. Therefore, the interval is spacelike.

(d) We choose (eλ , e2λ ) as the integrating interval, then we have

e2λ
√
2a − √
Z
dx = 2aλ (2.19)
eλ x−
8 CHAPTER 2. SPECIAL RELATIVITY AND EXTRA DIMENSIONS

x0
x− x+

x1

Fig. 2.2 The spacetime diagram for the x± axes and the family of curves x+ x− = a2 .

~2
( kπ 2 l 2
 
2.8 (a) The energy eigenvalues are Ekl = 2m a ) + ( R ) , so we have

∞Z ∞  E  mkT aR
Z
kl
Z(a, R) = exp − dk dl = (2.20)
0 0 kT 2~2

The results for a particle in a two-dimensional box with sides a and 2πR are the same.
(b) Since R ≪ a, the lowest new energy level can be seen as E01 . Then, we have
 nπ 2 2mkT 1
En0 < kT < E01 ⇒ < < 2 (2.21)
a ~2 R
And Z(a, R) in this regime with the leading correction due to the small extra dimension is
Z ∞ √
 ~2 h ~2  kπ 2 i a 2mkT  ~2 
Z(a, R) = exp − exp − dk = exp −
2mkT R2 0 2mkT a ~ 2mkT R2
Chapter 3

Electromagnetism and Gravitation

in Various Dimensions

 Summary and Supplement

1. Classical electrodynamics

• Maxwell’s equations in the Heaviside-Lorentz system of units

1 ∂B
∇×E =− (3.1a)
c ∂t
∇·B = 0 (3.1b)
∇·E = ρ (3.1c)
1 1 ∂E
∇×B = j+ (3.1d)
c c ∂t

• The Lorentz force

dp 1

=q E+ v×B (3.2)
dt c
• Vector and scalar potentials

B =∇×A (3.3a)
1 ∂A
E=− − ∇Φ (3.3b)
c ∂t

2. Relativistic electrodynamics

Aµ = (Φ, A1 , A2 , A3 ), Aµ = (−Φ, A1 , A2 , A3 ) (3.4)

 
0 −Ex −Ey −Ez
Ex 0 Bz −By 
Fµν = ∂µ Aν − ∂ν Aµ =    (3.5)
Ey −Bz 0 Bx 
Ez By −Bx 0
∂λ Fµν + ∂µ Fνλ + ∂ν Fλµ = 0 (3.6)
1 µ
j µ = (cρ, j 1 , j 2 , j 3 ), ∂ν F µν = j (3.7)
c
3. Gravitation and Planck’s length

− ds2 = gµν (x)dxµ dxν , g µα (x)gαν (x) = δ µν (3.8)

gµν (x) = ηµν + hµν (x), ∂ 2 hµν − ∂α (∂ µ hνα + ∂ ν hµα ) + ∂ µ ∂ ν h = 0 (3.9)

′
xµ = xµ + ǫµ (x), δhµν = ∂ µ ǫν + ∂ ν ǫµ (3.10)
r r r
~G ~G ~c
ℓP = 3
= 1.61 × 10−33 cm, tP = 5
= 5.4 × 10−44 s, mP = = 2.17 × 10−5 g (3.11)
c c G

9
10 CHAPTER 3. ELECTROMAGNETISM AND GRAVITATION IN VARIOUS DIMENSIONS

 Quick Calculations

3.1 Under the gauge transformations, E is invariant:

1 ∂ 1 ∂ ǫ
1 ∂A
E′ = −


A + ∇ǫ − ∇ Φ − =− − ∇Φ = E (3.12)
c∂t c ∂t c ∂t

3.2 As we know, Fµν is antisymmetrical: Fµν = −Fνµ . Therefore, we have

Tλµν + Tµλν = ∂λ (Fµν + Fνµ ) + ∂µ (Fνλ + Fλµ ) + ∂ν (Fλµ + Fµλ ) = 0 (3.13a)

Tλµν + Tλνµ = ∂λ (Fµν + Fνµ ) + ∂µ (Fνλ + Fλµ ) + ∂ν (Fλµ + Fµλ ) = 0 (3.13b)

3.3 With the definition F µν = η µα η νβ Fαβ , we can obtain

F µν + F νµ = η µα η νβ Fαβ + η να η µβ Fαβ = η µα η νβ (Fαβ + Fβα ) = 0 (3.14a)

0i 0j ik ik
F + F0i = η η Fjk + F0i = −δ F0k + F0i = 0 (3.14b)
ij ik jl ik jl
F − Fij = η η Fkl − Fij = δ δ Fkl − Fij = 0 (3.14c)

The last one holds because the indices i, j = 1, 2, 3.

3.4 The following proof will use the property of the gamma function: Γ(x + 1) = xΓ(x).

2π d/2 d−1 d d 2π d/2 Rd π d/2

vol(S d−1 (R)) = d
R = B (R) ⇒ B d (R) = d
= Rd (3.15)
Γ( 2 ) dR Γ( 2 ) d Γ(1 + d2 )

1√ q
3.5 Since Γ( 32 ) = 2 π, it is easy to get E(r) = 4πr 2 .

3.6 [F ] = [q]2 · Ld−1 ⇒ [q] = M 1/2 Ld/2 T −1 .

3.7 Since [G] = M −1 L3 T −2 , [c] = LT −1 and [~] = M L2 T −1 , we have −α + γ = 0, 3α + β + 2γ = 1,

−2α − β − γ = 0. Then, α = γ = 1/2 and β = −3/2.

3.8 EP = mP c2 = 0.511 × 10−3 × 2.17 × 10−5 /0.911 × 10−27 = 1.22 × 1019 GeV.
H RR RR
3.9 g · dl = (∇ × g) · dS = − ∇ × (∇Vg ) · dS = 0.

3.10 Since the units of G(D) ρm are the same in all dimensions, we have

M M [c]3 M G(D) c3 (D)
D−2 ~G(D) 2G

(D)
[G(D) ] = [G] = ⇒ (D) D−1
= ⇒ ℓ P = = (ℓ P )
LD−1 L3 [~] L (D) c3 G


ℓP ~ℓP

 Solutions to the Problems

3.1 We will check the case for µ = 1. As we know, γds = c dt, E10 = Ex , E11 = 0, E12 = Bz , and
F13 = −By , it is easy to obtain

dpx q dx0 dx1 dx2 dx3 

= F10 + F11 + F12 + F13
dt c dt dt dt dt (3.16)
q dy dz   1 
= cEx + Bz − By = q Ex + (vy Bz − vz By )
c dt dt c

For µ = 0, we have F00 = 0, F01 = −Ex , F02 = −Ey , F03 = −Ez and p0 = − Ec . Then,

1 dE q dE
− = − (Ex vx + Ey vy + Ez vz ) ⇒ = qE · v = F · v (3.17)
c dt c dt
Since Fµν is invariant under the gauge transformation and pµ and xν is independent of Aµ , it is
a gauge invariant equation.
 11

3.2 (a) T is nonvanishing only when each of its three indices takes a different value.
1 ∂Bz ∂Ey ∂Ex
∂0 F12 + ∂1 F20 + ∂2 F01 = + − =0 (3.18a)
c ∂t ∂x ∂y
1 ∂By ∂Ez ∂Ex
∂0 F13 + ∂1 F30 + ∂3 F01 =− + − =0 (3.18b)
c ∂t ∂x ∂z
1 ∂Bx ∂Ez ∂Ey
∂0 F23 + ∂2 F30 + ∂3 F02 = + − =0 (3.18c)
c ∂t ∂y ∂z
∂Bx ∂By ∂Bz
∂1 F23 + ∂2 F31 + ∂3 F12 = + + =0 (3.18d)
∂x ∂y ∂z
The first three are the components of Eq. (3.1a) and the last one is just Eq. (3.1b).
(b)
∂F 0ν ∂Ex ∂Ey ∂Ez 1
= + + = j0 = ρ (3.19a)
∂xν ∂x ∂y ∂z c
∂F 1ν 1 ∂Ex ∂Bz ∂By 1
=− + − = j1 (3.19b)
∂xν c ∂t ∂y ∂z c
∂F 2ν 1 ∂Ey ∂Bz ∂Bx 1 2
=− − + = j (3.19c)
∂xν c ∂t ∂x ∂z c
∂F 3ν 1 ∂Ez ∂By ∂Bx 1 3
=− + − = j (3.19d)
∂xν c ∂t ∂x ∂y c
The first one is just Eq. (3.1c) and the last three are the components of Eq. (3.1d).
3.3 (a) Using the ansata Ez = Bx = By = 0, we can easily obtain the following from the Maxwell’s
equations and the force law in four dimensions.
∂Ey ∂Ex 1 ∂Bz ∂Ex ∂Ey
− =− , + =ρ (3.20)
∂x ∂y c ∂t ∂x ∂y
∂Bz 1 1 ∂Ex ∂Bz 1 1 ∂Ey
= j1 + , − = j2 + (3.21)
∂y c c ∂t ∂x c c ∂t
(b) With the Lorentz covariant formulation, we have Aµ = (Φ, A1 , A2 ), j µ = (cρ, j 1 , j 2 ), and
   
0 −Ex −Ey 0 Ex Ey
Fµν = Ex 0 Bz  , F µν = −Ex 0 Bz  (3.22)
Ey −Bz 0 −Ey −Bz 0

Then ∂λ Fµν + ∂µ Fνλ + ∂ν Fλµ = 0 and ∂ν F µν = 1c j µ will give the same equations with thoes
obtained in (a). And the relativistic form of the force law yields
dE dpx 1 dpy 1
= q(Ex vx + Ey vy ), = q(Ex + vy Bz ), = q(Ey − vx Bz ) (3.23)
dt dt c dt c
3.4 (a) Since Aµ is time-independent, we have ∂0 Fij = 0. Then,
T0ij = ∂0 Fij + ∂i Fj0 + ∂j F0i = ∂i Ej − ∂j Ei = 0 (3.24)
This condition is satisfied because ∇ × E = −∇ × (∇Φ) = 0.
(b) With the relations vol(S d1 (r)) = 2π d/2 rd−1 /Γ( d2 ) and E = −dΦ/dr, we have

Γ( d2 ) q Γ( d2 − 1) q
ZZ
E · dS = q ⇒ E(r) = ⇒ Φ(r) = (3.25)
2π d/2 rd−1 4π d/2 rd−2

3.5 ∇ · f = ∇f · r̂ + f ∇ · ( rr ) = f ′ (r) + f (r)[∇( 1r ) · r + r1 ∇ · r] = f ′ (r) + d−1

r f (r).

3.6 By the definition, we have

1 N N Z ∞
n!  X n! 1
Z  X
z−1
dt t e−t − n
+ n z+n
+ dt e−t tz−1
0 n=0
(−1) n=0
(−1) 1
∞ Z 1 (3.26)
X n!  1 n+z−1

= Γ(z) + − t dt = Γ(z)
n=0
(−1)n z + n 0
12 CHAPTER 3. ELECTROMAGNETISM AND GRAVITATION IN VARIOUS DIMENSIONS

The function of the first integral has the order of O(tz+N ). For ℜ(z) > −N − 1, the integral on
[0, 1] will always converge. So the right-hand side above is well defined. And also, we can obtain
the following identity:
Γ(z + N + 1)
Γ(z) = (3.27)
z(z + 1) · · · (z + N )
Obviously, z = 0, −1, −2, . . ., are poles for Γ(z). The value of residue at z = −n is

(−1)n
Res[Γ(z), −n] = lim (z + n)Γ(z) = (3.28)
z→−n n!

3.7 (a) The “gravitational” Bohr radius for a hydrogen atom is ~2 /Gm3 = 2.2 × 1032 m.
(b) Suppose kT = (8πM )−1 Gα cβ ~γ , then we have

[E] = [G]α [c]β [~]γ M −1 ⇒ M L2 T −2 = M −α+γ−1 L3α+β+2γ T −2α−β−γ

(3.29)
⇒ α = −1, β = 3, γ = 1

For M = 106 M⊙ , M⊙ = 2 × 1030 kg, the temperature is T = 6.15 × 10−14 K. And for the black
hole whose temperature is room temperature (300K), its mass will be M = 4.2 × 1020 kg.

3.8 We use the effective potential Veff (r) = Vg (r) + J 2 /2mr2 to discuss the planetary motion.

2Γ( d2 ) G(D) m d
Z Z
g · dS = ∇ · g d(vol) = −4πG(D) m ⇒ g(r) = − d/2−1 d−1
= − Vg (r)
S d−1 Bd π r dr
(3.30)
Γ( 2d − 1) G(D) m
⇒ Vg (r) = − d/2−1
π rd−2
d
 
From the condition dr Veff (r) r=r = 0, we can solve r0 . Then,
0

d2 Veff  (4 − d)J 2
= (3.31)
dr2 r=r0 mr04


For d = 3, it is positive, so the planetary circular orbits in the four-dimensional world are stable
under perturbations; while for d ≥ 4, they are not stable.

3.9 (a) Using the result in Eq. (3.30), we can directly write down the expression:

G(5) M
Vg(5) (r) = − (3.32)
πr2
(b) The circel can be constructed by the identification of R1 : w ∼ w + 2nπa, thus we have
∞
X G(5) M
Vg(5) (x, y, z, 0) =− (3.33)
n=−∞
π[R2 + (2nπa)2 ]

(c) For R ≫ a, the potential becomes

∞
G(5) M dt G(5) M GM
Z
Vg(5) (x, y, z, 0) = − 2 2
= − =− , (3.34)
−∞ π R + (2πat) 2πaR R

in which we have used the relation G(5) = 2πaG.

3.10 (a) Using the following identity

∞
X 1 1 1
2
= coth , (3.35)
n=−∞
1 + (nπx) x x

we can find an exact closed-form expression for the potential:

G(5) M R R GM R

Vg(5) (x, y, z, 0) = − 2
coth =− coth (3.36)
πR 2a 2a R 2a
 13

(b) For R ≫ a, we can expand the result above to the leading correction:

GM 1 + e−λ GM
Vg(5) (x, y, z, 0) = − −λ
≃− 1 + 2e−λ ), (3.37)
R 1−e R
where λ = R/a. When λ = 5.3, the correction is of order 1%.
(c) Using the following identity

1 x x3 2x5
coth x = + − + + ··· , 0 < |x| < π, (3.38)
x 3 45 945
we can expand the gravitational potential when R ≪ a:

GM  2a R G(5) M  R2 
Vg(5) (x, y, z, 0) ≃ − + =− 1 + (3.39)
R R 6a πR2 12a2
The first term has the same form to the gravitational potential discussed in 3.9 (a) .
14 CHAPTER 3. ELECTROMAGNETISM AND GRAVITATION IN VARIOUS DIMENSIONS
Chapter 4

Nonrelativistic Strings

 Summary and Supplement

1. Equations of motion for transverse oscillations
s
∂2y 2
2∂ y T0
= v0 , v0 = (4.1)
∂t2 ∂x2 µ0

y(t, x) = h+ (x − v0 t) + h− (x + v0 t) (4.2)

2. The nonrelativistic string Lagrangian

 ∂y ∂y  1  ∂y 2 1  ∂y 2
L , = µ0 − T0 (4.3)
∂t ∂x 2 ∂t 2 ∂x
∂L ∂L ∂P t ∂P x
Pt = , Px = , + =0 (4.4)
∂ ẏ ∂y ′ ∂t ∂x
 Quick Calculations
4.1 As we vary: y(t, x) → y(t, x) + δy(t, x), the variation is
Z tf Z a h Z tf Z a h
1  ∂y 2 1  ∂y 2 i ∂y ∂(δy) ∂y ∂(δy) i
δS = δ dt dx µ0 − T0 = dt dx µ0 − T0
ti 0 2 ∂t 2 ∂x ti 0 ∂t ∂t ∂x ∂x

4.2 When we vary the motion by δy, the variation of the action is given by
Z tf Z a
dx P t δ ẏ + P x δy ′


δS = dt
ti 0
tf h∂ a
∂P t
∂P x i
∂
Z Z
= dt
dx P t δy − δy + P x δy − δy (4.5)
ti 0 ∂t ∂t ∂x ∂x
Z a Z tf  Z tf Z a  t
t
t=tf
x
x=a ∂P ∂P x 
= P δy  dx + P δy  dt − dt dx + δy
 
0 t=ti ti x=0 ti 0 ∂t ∂x

4.3 With the relations P t = µ0 ẏ and P x = −T0 y ′ , we can rewrite Eq. (4.5) as
Z a Z tf  Z tf Z a 
t=tf x=a ∂P t ∂P x 
δS = P t δy  dx + P x δy  dt − dt dx + δy
 
0 t=ti ti x=0 ti 0 ∂t ∂x
Z a Z tf  Z tf Z a 
∂y t=tf ∂y x=a ∂ 2y ∂2y 
= µ0 δy  dx + − T0 δy  dt − dt dx µ0 2 − T0 2 δy
0 ∂t t=ti ti ∂x x=0 ti 0 ∂t ∂x

 Solutions to the Problems

4.1 For small oscillations, the horizontal force dFh is
h  ∂y 2 i− 12 h  ∂y 2 i− 12 ∂y ∂ 2 y
dFh = T0 1 + − T0 1 + = −T0 dx (4.6)
∂x x+dx ∂x x ∂x ∂x2
Since ∂y/∂x ≪ 1, it is much smaller than the vertical force dFv .

15
16 CHAPTER 4. NONRELATIVISTIC STRINGS

4.2 For the small longitudinal oscillations, the equation can be derived as
∂z  ∂z  ∂2z ∂2z ∂2z τ0 ∂ 2 z
T (x + dx) − T (x) = τ0  − τ0  = τ0 2 dx = µ0 2 ⇒ 2 = (4.7)
∂x x+dx ∂x x ∂x ∂t ∂t µ0 ∂x2
p
Thus, the velocity of the waves is τ0 /µ0 .
4.3 (a) Suppose u = −v0 t and w = u − a, then we can obtain the following:
y(t, 0) = h+ (−v0 t) + h− (v0 t) = 0 ⇒ h+ (u) = −h− (−u) (4.8)
y(t, a) = h+ (a + u) + h− (a − u) = 0 ⇒ h+ (u + a) − h+ (u − a) = 0 ⇒ h+ (w) = h+ (w + 2a) (4.9)
(b) From the initial conditions, we have
∂y 
y|t=0 = h+ (x) + h− (x) = 0, = −v0 h′+ (x) + v0 h′− (x) (4.10)
∂t t=0

The second equation can be rewritten by the integrating in two different intevals, which gives
different functions of h+ (u). For 0 < x < a, we have
Z x 
ξ ξ 1  u3 u2 
− h+ (x) + h− (x) = 1− dξ ⇒ h+ (u) = 2
− −c (4.11)
0 a a 2 3a 2a
For −a < x < 0, it becomes
x
ξ ξ 1  u3 u2
Z 
− h+ (x) + h− (x) = − 1+ dξ ⇒ h+ (u) = + + c (4.12)
0 a a 2 3a2 2a
We can extand h+ (u) for all u with the periodic conditions h+ (u) = h+ (u + 2a).

(c) For x and v0 t in the domain D = (x, v0 t)| 0 ≤ x ± v0 t < a , the wave function is
1 h (x − v0 t)3 (x − v0 t)2 i 1 h (x + v0 t)3 (x + v0 t)2 i  x x2 v 2 t2 
y(t, x) = 2
− − 2
− = v0 t − − 0 (4.13)
2 3a 2a 2 3a 2a a a2 3a2
(d) From the function above, we can obtain
∂y  x x2 v 2 t2  h 1  x 1  2 i v 2 t2
= v0 − 2 − 0 2 = −v0 − − − 02 (4.14)
∂t a a a 4 a 2 a
Obviously, at t = 0 the midpoint x = a/2 has the largest velocity. It is easy to concluce that the
velocity of the midpoint reaches zero at t0 = a/2v0 and y(t0 , a/2) = a/12.
4.4
4.5
4.6 (a) The variation δS of the action under a variation δq of the coordinate can be derived as
∂L ∂L 
Z Z 
δS = δ dt L(q(t), q̇(t); t) = dt δq + δ q̇
∂q ∂ q̇
Z h ∂L d  ∂L   d ∂L  i
= dt δq + δq − δq (4.15)
∂q dt ∂ q̇ dt ∂ q̇
Z  ∂L d ∂L 
= dt − δq
∂q dt ∂ q̇
Then, δS = 0 gives the famous Euler-Lagrange equation
∂L d ∂L
− =0 (4.16)
∂q dt ∂ q̇
(b) The derivation is similar to that in part (a):
Z Z h ∂L ∂L i
δS = δ dD x L(φ(x), ∂µ φ(x)) = dD x δφ + δ(∂µ φ)
∂φ ∂∂µ φ
Z h ∂L  ∂L   ∂L  i
= dD x δφ + ∂µ δφ − ∂µ δφ (4.17)
∂φ ∂∂µ φ ∂∂µ φ
Z  ∂L ∂L 
= dD x − ∂µ δφ
∂φ ∂∂µ φ
Then we can obtain the Euler-Lagrange equation for the dynamical field φ(x)
∂L ∂L
− ∂µ =0 (4.18)
∂φ ∂∂µ φ
Chapter 5

The Relativistic Point Particle

 Summary and Supplement

1. Action for a relativistic point particle

r
2 v2 mc2 ∂L mv
L = −mc 1 − 2 , H = p , p= = p (5.1)
c 1 − v 2 /c2 ∂v 1 − v 2 /c2
τf
r
dxµ dxν
Z Z
µ µ
x = x (τ ), S = −mc ds = −mc −ηµν dτ (5.2)
τi dτ dτ

2. Equations of motion
dpµ d2 xµ
δ(dxµ ) = d(δxµ ), = 0, =0 (5.3)
dτ ds2
q
Z Z
S = −mc ds + Aµ (x) dxµ (5.4)
P c P

 Quick Calculations

5.1 Since xi and xf are fixed under the variation, we have

1
Z Z Z
δSnr = δ mv 2 dt = mv · v0 dt = mv0 · dx = 0 (5.5)
2

5.2 For any arbitrary parameter τ ′ (τ ), we have

dpµ dpµ dτ ′ dpµ

= =0⇒ =0 (5.6)
dτ dτ ′ dτ dτ ′

 Solutions to the Problems

5.1 For the new parameter τ = f (s), we have

d2 xµ d  dxµ df  d2 xµ  df 2 dxµ d2 f d2 f
= = + = 0 ⇒ = 0 ⇒ f = as + b, (5.7)
ds2 ds dτ ds dτ 2 ds dτ ds2 ds2
where a and b are constants independent of s.

5.2 If we reexpress the integrand ds using the parameterized world-line, i.e.

r
dxµ dxν
ds = −ηµν dτ (5.8)
dτ dτ
then we can obtain the following

d2 xλ d2 xλ  dτ 2 dxλ d2 τ dxµ dxν d2 xλ 1 dxλ d  dxµ dxν 

= + = 0 ⇒ ηµν − ηµν = 0 (5.9)
ds2 dτ 2 ds dτ ds2 dτ dτ dτ 2 2 dτ dτ dτ dτ

5.3

17
18 CHAPTER 5. THE RELATIVISTIC POINT PARTICLE

5.4 (a) Using the relations Aµ = (−Φ, A) and dxµ /dt = (c, v), we can rewrite the action S as

1 q 1 q
Z Z
2
S= mv dt + (−Φc + A · v) dt ⇒ L = mv 2 − qΦ + A · v (5.10)
2 c 2 c

(b) By the definition, it is easy to get

∂L q
p= = mv + A · v (5.11)
∂v c
(c) The Hamiltonian for the charged particle is given by

q  mv 2 q  mv 2 1  q 2
H = p · v − L = mv 2 + A · v − − qΦ + A · v = + qΦ = p − A + qΦ (5.12)
c 2 c 2 2m c

5.5 The variation of the action for a charged point particle can be derived as
Z τf
dpµ q τf h dxµ  dxµ i
Z
δS = − dτ δxµ + dτ δAµ + Aµ δ
τ dτ c τi dτ dτ
Z iτf Z τf h µ
dpµ q ∂Aµ ν dx d dAµ µ i
=− dτ δxµ + dτ ν
δx + (Aµ δxµ ) − δx
τi dτ c τi ∂x dτ dτ dτ
Z τf
dpµ q τf  ∂Aµ ν dxµ ∂Aµ dAν µ 
Z
=− dτ δxµ + dτ ν
δx − δx (5.13)
τi dτ c τi ∂x dτ ∂xν dτ
Z τf
µ dpµ q τf  ∂Aν ∂Aµ  dxν µ
Z
=− dτ δx + dτ − δx
τ dτ c τi ∂xµ ∂xν dτ
Z τf i 
dpµ q dxν  µ
= dτ − + Fµν δx
τi dτ c dτ

Then, δS = 0 gives the equation of motion:

dpµ q dxν
= Fµν (5.14)
dτ c dτ

5.6
5.7
Chapter 6

Relativistic Strings

 Summary and Supplement

1. Area functional for spatial surfaces
s
Z  ∂x ∂x  ∂x ∂x   ∂x ∂x 2
A = dξ 1 dξ 2 · · − · (6.1)
∂ξ 1 ∂ξ 1 ∂ξ 2 ∂ξ 2 ∂ξ 1 ∂ξ 2

∂x ∂x √
Z
gij = i · i , ds2 = gij dξ i dξ j , A = dξ 1 dξ 2 g, g = det(gij ) (6.2)
∂ξ ∂ξ
2. The Nambu-Goto action
r
∂X ∂X 2  ∂X 2  ∂X 2
Z
A= dτ dσ · − (6.3)
∂τ ∂σ ∂τ ∂σ
τf Z σ1 q
T0
Z
S=− dτ dσ (Ẋ · X ′ )2 − (Ẋ)2 (X ′ )2 (6.4)
c τi 0
∂Xµ ∂Xν T0 √
Z
γαβ = ηµν , S=− dτ dσ −γ, γ = det(γαβ ) (6.5)
∂ξ α ∂ξ β c
3. Equation of motion, boundary conditions, and D-branes
T0
q
L(Ẋ µ , X µ′ ) = − (Ẋ · X ′ )2 − (Ẋ)2 (X ′ )2 (6.6)
c
∂L T0 (Ẋ · X ′ )Xµ′ − (X ′ )2 Ẋµ
Pµτ = =− q (6.7)
∂ Ẋ µ c
(Ẋ · X ′ )2 − (Ẋ)2 (X ′ )2

∂L T0 (Ẋ · X ′ )Ẋµ − (X ′ )2 Xµ′

Pµσ = = − (6.8)
∂X µ′
q
c
(Ẋ · X ′ )2 − (Ẋ)2 (X ′ )2

∂Pµτ ∂Pµσ
Equation for relativistic string: + =0 (6.9)
∂τ ∂σ
∂X µ
Dirichlet boundary condition: (τ, σ∗ ) = 0, µ 6= 0, σ∗ = 0 or σ1 (6.10)
∂τ
Free endpoint condition: Pµσ (τ, σ∗ ) = 0, σ∗ = 0 or σ1 (6.11)

4. Action in terms of transverse velocity

∂X µ ∂X ∂X µ ∂X
X µ (τ, σ) = (ct, X(t, σ)), = (c, ), = (0, ) (6.12)
∂τ ∂t ∂σ ∂σ
∂X  ∂X ∂X  ∂X  ∂X 2  ∂X ∂X 2
2
v⊥ = − · , v⊥ = − · (6.13)
∂t ∂t ∂s ∂s ∂t ∂t ∂s
Z σ1 r r
Z  ds  v 2 Z 2
v⊥
S = −T0 dt dσ 1− ⊥ , L = −T 0 ds 1 − (6.14)
0 dσ c2 c2

19
20 CHAPTER 6. RELATIVISTIC STRINGS

5. Motion of open string endpoints

v 2 − 2 n ∂X ∂X  µ h 2  ∂X 2 i ∂X µ o
1
T0 
P σµ = − 2 1 − ⊥ · Ẋ + c − (6.15)
c c2 ∂s ∂t ∂t ∂s
2 − 1 
T0  v 2 ∂X ∂X  ∂X
P σ0 = − 1− ⊥ · =0⇒v· =0 (6.16)
c c2 ∂s ∂t ∂s
 v 2  12 ∂X µ
P σµ = −T0 1 − 2 = 0 ⇒ v 2 = c2 (6.17)
c ∂s
 Quick Calculations
6.1 Using the chain rule of derivatives, we can obtain
s
Z  ∂x ∂x  ∂ ξ̃ 1 2  ∂x ∂x  ∂ ξ̃ 2 2  ∂x ∂x 2  ∂ ξ̃ 1 ∂ ξ̃ 2 2
A = dξ 1 dξ 2 · · − ·
∂ ξ̃ 1 ∂ ξ̃ 1 ∂ξ 1 ∂ ξ̃ 2 ∂ ξ̃ 2 ∂ξ 2 ∂ ξ̃ 1 ∂ ξ̃ 2 ∂ξ 1 ∂ξ 2
s (6.18)
Z  ∂x ∂x  ∂x ∂x   ∂x ∂x 2
= dξ˜1 dξ˜2 · · − ·
∂ ξ̃ 1 ∂ ξ̃ 1 ∂ ξ̃ 2 ∂ ξ̃ 2 ∂ ξ̃ 1 ∂ ξ̃ 2

6.2 Using the chain rule, we can prove that

∂ξ i ∂ ξ̃ j ∂ξ i ∂ ξ̃ i ∂ξ j ∂ ξ̃ i
Mij M̃jk = k
= k = δ ik , M̃ij Mjk = j
= = δ ik (6.19)
∂ ξ̃ j ∂ξ ∂ξ ∂ξ ∂ ξ̃ k ∂ ξ̃ k

6.3 For a point on the world-sheet where all tangent vectors are spcelike with the exception of one
that is null, we have
∂X ∂X ∂X
= ⇒ v 2 (λ) = (λ + 1)2 ≥0 (6.20)
∂τ ∂σ ∂σ
When λ = −1, the tangent vector v = 0.
6.4 With the relations Ẋ · X ′ = Ẋ µ Xµ′ , (Ẋ)2 = Ẋ µ Ẋµ , and (X ′ )2 = X µ′ Xµ′ , it is easy to verify
Eq. (6.7) and Eq. (6.8).
 Solutions to the Problems
6.1 Since the oscillations are small, we have
ds2 = dX · dX = (dx, dy) · (dx, dy) = dx2 + dy · dy ≃ dx2 (6.21)
Then the following approximation holds:
∂X  ∂X ∂X  ∂X  ∂x ∂y   ∂x ∂y   ∂y 
v⊥ = − · ≃ , − , · (1, 0)(1, 0) = 0, (6.22)
∂t ∂t ∂s ∂s ∂t ∂t ∂t ∂t ∂t
Furthermore, the action reduces to be
r Z tf Z a h
1  ∂y 2 1  ∂y 2 i
Z Z
S ≃ −T0 dt dx 1 − 2 ≃ −T0 dt dx 1 − 2 (6.23)
c ∂t ti 0 2c ∂t
Up to an additive constant −aT0 (tf −ti ), it is just the action for a nonrelativistic string performing
small transverse oscillations, since µ0 = T0 /c2 .
6.2 We start our derivation with the Nambu-Goto action and work in the static gauge:
Z σ1 r
T0 τf  ∂X 2
Z
S≃− dτ dσ 0 − (−c2 )
c τi 0 ∂σ
Z tf Z 1  ∂X 
= −T0 dt dσ 
 
∂σ

ti 0
Z tf Z a r  ∂y 2
= −T0 dt dx 1 + (6.24)
ti 0 ∂x
Z tf Z a h
1 ∂y 2 i

≃ −T0 dt dx 1 +
ti 0 2 ∂x
Z tf Z a
1  ∂y 2
= −aT0 (tf − ti ) − T0 dt dx
ti 0 2 ∂x
 21

6.3
6.4 In this problem, we are discussiong the time evolution of a closed circular string. It is clear to
us that Ṙ(t) = v⊥ , so we have
Z q q
L = −T0 ds 1 − Ṙ2 (t)/c2 = −2πR(t)T0 1 − Ṙ2 (t)/c2 (6.25)

Then, we can obtain the Hamilonian

∂L 2πT0 RṘ  2Ṙ  2πR(t)T0

q
H = Ṙ −L=− q − 2 + 2πRT0 1 − Ṙ2 /c2 = q (6.26)
∂ Ṙ c
2 1 − Ṙ2 /c2 1 − Ṙ2 (t)/c2

Acording to the energy onservation law, Ḣ = 0. It gives

d R(t)  ct 
q = 0 ⇒ Ṙ2 (t) − R(t)R̈(t) = c2 ⇒ R(t) = R cos , (6.27)
dt R
1 − Ṙ2 (t)/c2

which has already satisfied the intial conditions: R(0) = R and Ṙ(0) = 0.
6.5 By the definition, it is easy to obtain

T02 [c2 − (∂t X)2 ]  2

Pµσ P σµ = c − (∂t X)2 (∂s X)2 − (∂s X · ∂t X)2 = 0 ⇒ v = |∂t X| = c (6.28)

2
c2 − v⊥

6.6 From Eq. (6.14), we know that r

v⊥2
ds
L = −T0
2
1− (6.29)
c dσ
Then, the canonical momentum density and the Hamiltonian density are given by

∂L 1 v 2 − 12 h 1 ∂v⊥ 2 i ds T0  2 − 1
v⊥ 2 ds
P(t, σ) = = −T0 1 − ⊥ 2
− 2
= 2
1 − 2
v⊥ (6.30)
∂(∂t X) 2 c c ∂(∂t X) dσ c c dσ
r
2 − 1 v 2 ds
∂X T0  v⊥ 2 ds
2 T0  v 2 − 12 ds
H = P(t, σ) · −L= 2 1− 2 v⊥ + T0 1 − ⊥ 2
= 2 1− ⊥ (6.31)
∂t c c dσ c dσ c c2 dσ
The total Hamiltonian can be written as
T0 v 2 − 12 v⊥2 − 1
Z Z  Z  2
H = dσ H = 2 ds 1 − ⊥ = µ 0 ds 1 − , (6.32)
c c2 c2
where µ0 is the rest mass of a string rasing solely from the tension.

6.7 (a) The conditions satisfied by P0σ , Piσ , and Paσ at the endpoint are stated as follows:

∂Paσ
P0σ (t, 0) = 0, Piσ (t, 0) = 0, (t, 0) = 0 (6.33)
∂σ
(b) If the string ends on a D0-brane, then the σ = 0 endpoint is fixed in the spacetime. Therefore,
all boundary conditions are automatially satisfied.
(c) For a string ending on a D1-brane, we have

T0  v 2 − 21  ∂X ∂X  ∂X
P0σ = 1− ⊥ · =0⇒v· =0 (6.34)
c c2 ∂s ∂t ∂s
(d)
22 CHAPTER 6. RELATIVISTIC STRINGS
Chapter 7

String Parameterization and

Classical Motion

 Summary and Supplement

1. Choosing a σ parameterization
∂X ∂X ∂X
· = 0, v⊥ = (7.1)
∂σ ∂t ∂t
 ds 2
Ẋ · X ′ = 0, X 2 = −c2 + v⊥
2
, X ′2 = (7.2)
dσ
2 − 1 µ
T0  v⊥ 2 ds ∂X
 v 2  12 ∂X µ
Pτµ = 2
1 − 2
, P σµ = −T0 1 − ⊥ (7.3)
c c dσ ∂t c2 ∂s
∂ T0 ds  T0 ds
Z
p = 0, H= p (7.4)
∂t 1 − v⊥ 2 /c2 dσ 1 − v⊥ 2 /c2

r r
2
v⊥ 2
v⊥
T0 1 ∂v⊥ ∂  ∂X 
= T0 1 − , T eff = T 0 1 − (7.5)
c2 1 − v⊥ c2 ∂s c2
p
2 /c2 ∂t ∂s

2. Wave equation and constraints

ds E
dσ ′ = p , σ 1 ∈ [0, σ1 ], σ1 = (7.6)
1 − v⊥2 /c2 T0

∂2X 1 ∂2X
Wave equation: − =0 (7.7)
∂σ 2 c2 ∂t2
∂X ∂X
Parameterization condition: · =0 (7.8)
∂t ∂σ
 ∂X 2 1  ∂X 2
Parameterization condition: + 2 =1 (7.9)
∂σ c ∂t
∂X  ∂X 
Free boundary condition: = =0 (7.10)
∂σ σ=0 ∂σ σ=σ1
 

T0 ∂X µ ∂X µ
Pτµ = , P σµ = −T0 (7.11)
c2 ∂t ∂σ
3. General motion of an open string
1 
X(t, σ) = F (ct + σ) + F (ct − σ) , σ ∈ [0, σ1 ] (7.12)
2
 dF (u) 2 2σ1
 = 1 and F (u + 2σ1 ) = F (u) + v0 (7.13)
 
du c

σ1  πu πu  σ1 πσ  πct πct 
F (u) = cos , sin , X(t, σ) = cos cos , sin (7.14)
π σ1 σ1 π σ1 σ1 σ1

23
24 CHAPTER 7. STRING PARAMETERIZATION AND CLASSICAL MOTION

 Quick Calculations

7.1 With the periodic conditon of F (u), it is easy to show that

1 
X(t = t0 + 2σ2 /c, σ) = F (ct0 + 2σ1 + σ) + F (ct0 + 2σ1 − σ)
2
1  2σ1
= F (ct0 + σ) + F (ct0 − σ) + v0 (7.15)
2 c
2σ1
= X(t = t0 , σ) + v0
c
Therefore, v0 is the average velocity of any point σ on the string calculated over any time interval
of duration 2σ1 /c.

 Solutions to the Problems

7.1 (d) The length ℓ of an open string parameterized with engery is given by
r r
σ1
v2 2
v⊥
Z
ds = 1− ⊥ dσ ⇒ ℓ = 1− dσ (7.16)
c2 0 c2

7.2 (a) For the rotating string, we have v⊥ = ωs = 2cs/ℓ. Then, the following holds

dE T0 T0
E(s) = =p = p (7.17)
ds 2
1 − v⊥ /c 2 1 − 4s2 /ℓ2

It has singularities at the endpoints s = ±ℓ/2. And the total energy is given by
ℓ/2
T0 π
Z Z
E= ds E(s) = ds p = ℓT0 (7.18)
−ℓ/2
2
1 − 4s /ℓ 2 2

(b) The average energy density is πT0 /2, so we have

√
T0 π π2 − 4
p = T0 ⇒ s = ± ℓ (7.19)
1 − 4s2 /ℓ2 2 2π

(c) The energy carried by the string on the interval [−s, s] is given by
s
T0 2s
Z
E= dx p = ℓT0 arcsin (7.20)
−s
2
1 − 4x /ℓ 2 ℓ

7.3 (a) The general solution for X(t, σ) in terms of a vector function F (u) is given by

1 
X(t, σ) = F (ct + σ) + F (ct − σ) (7.21)
2
The following parameterization conditions are required
 ∂X 1 ∂X 2  dF (u) 2
± =1⇒  =1 (7.22)
 
∂σ c ∂t du

∂X c dF (−u) dF (u)
(0, σ) = F ′ (σ) + F ′ (−σ) = 0 ⇒

=− ⇒ F (u) = F (−u) (7.23)
∂t 2 du du
(b) We should impose this condition: X(t, σ + σ1 ) = X(t, σ), i.e.

F (ct + σ + σ1 ) + F (ct − σ − σ1 ) = F (ct + σ) + F (ct − σ)

⇒ F (σ + σ1 ) + F (−σ − σ1 ) = F (σ) + F (−σ) (7.24)
⇒ F (σ + σ1 ) = F (σ)

(c) tP = σ1 /c, X(tP , σ) = X(0, σ).

 25

7.4 (a) Using the boundary condition at σ = σ1 , we find

1 
X(t, σ1 ) = x1 + F (ct + σ1 ) − F (ct − σ1 ) = x2 ⇒ F (u + 2σ1 ) − F (u) = 2(x2 − x1 ) (7.25)
2
(b) From the parameterization conditions, we can obtain
 ∂X 1 ∂X 2  dF (u) 2
± =1⇒  =1 (7.26)
 
∂σ c ∂t du
(c) According to the information above, we have
∂X
(t, 0) = F ′ (ct) = (sin γ cos ωt, sin γ sin ωt, cos γ)
∂σ  uω uω 
⇒ F ′ (u) = sin γ cos , sin γ sin , cos γ (7.27)
c c c
uω c uω 
⇒ F (u) = sin γ sin , − sin γ cos , u cos γ
ω c ω c
(d) Using the result in part (a), it is easy to obtain
c (u + σ1 )ω σ1 ω c (u + σ1 )ω σ1 ω 
2 sin γ cos sin , sin γ sin sin , σ1 cos γ
ω c c ω c c
(7.28)
cπ L0
= 2(x2 − x1 ) = 2(0, 0, L0) ⇒ ω = , σ1 =
σ1 cos γ
(e) Now, we can directly write down the wave function for this relativistic jumping rope:
σ πct πσ σ1 πct πσ 
1
X(t, σ) = sin γ cos sin , sin γ sin sin , σ cos γ (7.29)
π σ1 σ1 π σ1 σ1
2
(f) First, we calculate the value of v⊥ :
 πct πσ πct πσ   πσ 
v⊥ = − c sin γ sin sin , c sin γ cos sin 2
, 0 ⇒ v⊥ = c2 sin2 γ sin2 (7.30)
σ1 σ1 σ1 σ1 σ1
Then, the energy distributed in the string is given by
v 2 − 2
1  i− 12  i− 12
2 πσ 2 πz
 h h
2 2
E(z) = T0 1 − ⊥ = T 0 1 − sin γ sin = T 0 1 − sin γ sin (7.31)
c2 σ1 L0
7.5 (a) This ansatz is consistent with the constraints on F (u).
(b) It is easy to see that X ′ (t, σ) = 21 F ′ (ct + σ) + F ′ (ct − σ) , so we have
 

dy  πσ 
X ′ (0, 0) = F ′ (σ) ⇒ = sin γ cos (7.32)
dσ σ1
(c) X ′ (t, 0) = F ′ (ct).
(d) We will find an integral relation between a, σ1 and γ:
 Z u+2σ1 h πx i
Z u+2σ1 h πx i 
F (u + 2σ) − F (u) = cos γ cos dx, sin γ cos dx
u σ1 u σ1
 Z 2σ1 h πx i
Z 2σ1 h πx i 
= cos γ cos dx, sin γ cos dx (7.33)
0 σ1 0 σ1
 2σ1
Z h πx i  Z σ1  πx 
= cos γ cos dx, 0 ⇒ a = cos γ cos dx
0 σ1 0 σ1
Assume that γ is small, then the following approxiamtion holds:
Z σ1 
1 πx   1 
a≃ 1 − γ 2 cos2 dx = σ1 1 − γ 2 (7.34)
0 2 σ1 4
(e) Using the following identity
π
1
Z
J0 (z) = cos(z cos θ) dθ, (7.35)
π 0

we can obtain
σ1
πx  σ1 π a
Z  Z
a= cos γ cos dx = cos(γ cos θ) dθ ⇒ = J0 (γ) (7.36)
0 σ1 π 0 σ1
26 CHAPTER 7. STRING PARAMETERIZATION AND CLASSICAL MOTION

7.6 (a) We only need to examine the components.

1n h π(ct + σ) i h π(ct − σ) io
cos γ cos + cos γ cos
2 σ1 σ1
γ h π(ct + σ) π(ct − σ) i γh π(ct + σ) π(ct − σ) i
= cos cos + cos cos cos − cos (7.37)
2 σ1 σ1 2 σ1 σ1
 πct πσ   πct πσ 
= cos γ cos cos cos γ sin sin
σ1 σ1 σ1 σ1
1n h π(ct + σ) i h π(ct − σ) io
sin γ cos + sin γ cos
2 σ1 σ1
γh π(ct + σ) π(ct − σ) i γh π(ct + σ) π(ct − σ) i
= sin cos + cos cos cos − cos (7.38)
2 σ1 σ1 2 σ1 σ1
 πct πσ   πct πσ 
= sin γ cos cos cos γ sin sin
σ1 σ1 σ1 σ1
When ct = σ1 /2, the second component is zero. That is to say, the string is horizontal.
(b) It is easy to see that Ẋ/c = 12 F ′ (ct + σ) − F ′ (ct − σ) . Then we have
 

1n h π(ct + σ) i h π(ct − σ) io
cos γ cos − cos γ cos
2 σ1 σ1
γh π(ct + σ) π(ct − σ) i γh π(ct + σ) π(ct − σ) i
= − sin cos + cos sin cos − cos (7.39)
2 σ1 σ1 2 σ1 σ1
 πct πσ   πct πσ 
= − sin γ cos cos sin γ sin sin
σ1 σ1 σ1 σ1
1n h π(ct + σ) i h π(ct − σ) io
sin γ cos − sin γ cos
2 σ1 σ1
γh π(ct + σ) π(ct − σ) i γh π(ct + σ) π(ct − σ) i
= cos cos + cos sin cos − cos (7.40)
2 σ1 σ1 2 σ1 σ1
 πct πσ   πct πσ 
= cos γ cos cos sin γ sin sin
σ1 σ1 σ1 σ1
Therefore, the instantaneous transverse velocity satisfies
 1 ∂X    πct πσ 
 =  sin γ sin sin (7.41)
  
c ∂t σ1 σ1
 

For γ = π/2, the string midpoint σ = σ1 /2 reaches the speed of lignt when ct = σ1 /2, meaning
that the string is horizontal.
√
(c) For γ = 2π/2 and ct = σ1 /4, we have
  π 1 πσ  σ1
 sin √ √ sin =1⇒σ= (7.42)

2 2 σ1 2

For ct = σ1 /3, we have

  π √3 πσ 
 sin √ sin  = 1 ⇒ σ = 0.174σ1 or 0.826σ1 (7.43)

2 2 σ1
Chapter 8

World-sheet Currents

 Summary and Supplement

1. Conserved currents on the world-sheet
Z
S = dξ 0 dξ 1 L(∂0 X µ , ∂1 X µ ), (ξ 0 , ξ 1 ) = (τ, σ) (8.1)

∂L  ∂L ∂L 
jµα = , (jµ
0 1
, jµ ) = , ′ = (Pµτ , Pµσ ) (8.2)
∂(∂α X µ ) ∂ Ẋ µ ∂X µ
Z σ1
dpµ
Z
pµ (τ ) = Pµτ dσ, = 0, pµ = (Pµτ dσ − Pµσ dτ ) (8.3)
0 dτ γ

2. Lorentz symmetry and associated currents

X µ → X µ + δX µ , δX µ = ǫµν Xν , ǫµν = −ǫνµ (8.4)

∂Mτµν ∂Mσµν
Mα α α
µν = Xµ Pν − Xν Pµ , + =0 (8.5)
∂τ ∂σ
Z
Mµν = (Mτµν dσ − Mσµν dτ ), Mµν = −Mνµ (8.6)
γ

3. The slope parameter α′

J E2 1
= α′ E 2 , J= , α′ = (8.7)
~ 2πT0 c 2πT0 ~c

 Quick Calculations
8.1 We will use the divergence theorem to prove the result.
0  1
dQi k ∂ji k ∂ji ∂ji2 ∂jik 
Z Z Z
1 2 1 2
= dξ dξ · · · dξ = − dξ dξ · · · dξ + + · · · + = − j · dA = 0
dξ 0 ∂ξ 0 ∂ξ 1 ∂ξ 2 ∂ξ k ∂V

8.2 For the fixed 2-by-2 matrix Aab that satisfies Aab va vb = 0, we have

A11 v1 v1 + A12 v1 v2 + A21 v2 v1 + A22 v2 v2 = 0 ⇒ A11 = 0, A12 = −A21 , A22 = 0 (8.8)

8.3 For a 4-by-4 matrix ǫµν that satisfies ǫµν vµ vν = 0, the conclusion is the same: ǫµν must be
antisymmetric.
8.4 ǫµν = ηµα ηνβ ǫαβ = −ηµα ηνβ ǫβα = −ηαµ ηβν ǫβα = −ǫνµ .
8.5 For the boost with very small β, we have
′
x 0 − x0 = (γ − 1)x0 − γβx1 ≃ −βx1 ⇒ ǫ01 = −β (8.9)
′
x 1 − x1 = −γβx0 + (γ − 1)x1 ≃ βx0 ⇒ ǫ10 = β (8.10)
And all other values are zero.

27
28 CHAPTER 8. WORLD-SHEET CURRENTS
√
8.5 J = Iω, E = 21 Iω 2 ⇒ J ∼ E.
8.6 [α′ ] = [E]−2 , [~] = [E]T , [c] = LT −1 ⇒ [ℓs ] = L.
 Solutions to the Problems
8.1 (a) The variation can be written as δq(t) = ǫn × q, where ǫ is an infinitesimal constant and n is
the rotation axis. Then we have

q̇ ′ = q̇ + ǫn × q ⇒ q̇ ′2 = q̇ 2 + ǫ2 (n × q̇)2 ≃ q̇ 2 (8.11)

The Lagrangian only depends on q̇ 2 , therefore L is invariant.

(b) The conserved charge associated with this symmetry transformation is given by
∂L
ǫQ = · δq = p · (n × q) = n · (q × p) = n · J, (8.12)
∂ q̇
where J is the vector angular momentum.
8.2 (a) The variation of coordinates and the conserved charges are given by
∂L a
δq i (t) = ǫi (t)h(q(t); t), ǫi Qi = δq (8.13)
∂ q̇ a
Considering the Euler-Lagrange equations, we can obtain
dQi d  ∂L  a ∂L d ∂L ∂L d
ǫi = δq + a (δq a ) = a δq a + a (δq) = 0 (8.14)
dt dt ∂ q̇ a ∂ q̇ dt ∂q ∂ q̇ dt
(b) For a world with no spatial dimentions, α = 0. And the corresponding equations become
∂L d 0
ǫi ji0 = δφa , j = 0, Qi = cqi (8.15)
∂ φ̇a dt i

8.3
8.4 (a) T0 = 8.5 × 1014 GeV · m−1 , ℓs = 1.92 × 10−14 cm.
(b) α′ = 2.58 × 10−33 GeV−2 , T0 = 3 × 1047 GeV · m−1 .
8.5 For the relativistic jumping rope, we have
σ πct πσ σ1 πct πσ 
1
X(t, σ) = sin γ cos sin , sin γ sin sin , σ cos γ (8.16)
π σ1 σ1 π σ1 σ1
T0 ∂X T0 πσ  πct πct 
Pτ = 2 = sin γ sin − sin , cos ,0 (8.17)
c ∂t c σ1 σ1 σ1
Then, the z-component of angular momentum is given by
Z σ1 Z σ1
T0 σ1 2 πσ T0 σ12
Jz = M12 = τ τ
(X1 P2 − X2 P1 ) dσ = sin γ sin2 dσ = sin2 γ (8.18)
0 πc 0 σ1 2πc
Since σ1 = E/T0 , we have found
E2 Jz
Jz = sin2 γ ⇒ = (sin2 γ)α′ E 2 (8.19)
2πT0 c ~

8.6 With the Euler-Lagrange equation, we can show that

dQ d  ∂L  ∂L d dΛ ∂L ∂L
ǫ = δq + (δq) − ǫ = δq + δ q̇ − δL = 0 (8.20)
dt dt ∂ q̇ ∂ q̇ dt dt ∂q ∂ q̇
For the transformation q(t) → q(t) + ǫq̇(t), we have
∂L ∂L h d  ∂L  ∂L i d  ∂L  ∂L
δL = δq + δ q̇ = ǫ q̇ + q̈ = ǫ q̇ ⇒ Λ = q̇ (8.21)
∂q ∂ q̇ dt ∂ q̇ ∂ q̇ dt ∂ q̇ ∂ q̇
∂L ∂L
ǫQ = δq − ǫΛ = q̇ − ǫΛ ⇒ Q = 0 (8.22)
∂ q̇ ∂ q̇
 29

8.7 The proof is very similar to that in Eq. (8.20):

∂jiα ∂ h ∂L i a ∂L ∂δφa ∂Λα

ǫi α
= α a
δφ + a α
− ǫi αi
∂ξ ∂ξ ∂(∂α φ ) ∂(∂α φ ) ∂ξ ∂ξ
(8.23)
∂L a ∂L
= δφ + δ(∂α φa ) − δL = 0
∂φa ∂(∂α φa )

For the transformation φa (ξ β ) → φa (ξ β ) + ǫβ ∂β φa , we have

∂L a ∂L h ∂  ∂L  ∂L i
δL = a
δφ + a
δ(∂α φa ) = ǫβ α a
∂β φa + a
∂α ∂β φa
∂φ ∂∂α φ ∂ξ ∂∂α φ ∂∂α φ
(8.24)
∂  ∂L  ∂L
= ǫβ α ∂β φa ⇒ Λα β = ∂β φa = δβα L
∂ξ ∂∂α φa ∂∂α φa
∂L  ∂L  ∂L
ǫβ jβα = δφa
− ǫ β α
Λ β = ǫ β
∂β φa
− δ α
β L ⇒ jβα = ∂β φa − δβα L (8.25)
∂∂α φa ∂∂α φa ∂∂α φa
It is easy to see that j00 = H, i.e. the Hamiltonian density.
30 CHAPTER 8. WORLD-SHEET CURRENTS
Chapter 9

Light-cone Relativistic Strings

 Summary and Supplement

1. The σ parameterization
X 0 (τ, σ) = cτ, nµ X µ (τ, σ) = λτ (9.1)
√
~ = c = 1, L = T, M = L−1 , ℓs = α′ (9.2)
Z τf Z σ1 q
1
S=− dτ dσ (Ẋ · X ′ )2 − (Ẋ)2 (X ′ )2 (9.3)
2πα′ τi 0
Z σ
open strings: (n · p)σ = π dσ̃ n · P τ (τ, σ̃), n · X(τ, σ) = 2α′ (n · p)τ (9.4)
0
Z σ
closed strings: (n · p)σ = 2π dσ̃ n · P τ (τ, σ̃), n · X(τ, σ) = α′ (n · p)τ (9.5)
0

2. Constraints and wave equations

Ẋ · X ′ = 0, Ẋ 2 + X ′2 = 0, (Ẋ ± X ′ )2 = 0 (9.6)

1 1
Pτµ = Ẋ µ , P σµ = − X µ′ , Ẍ µ − X µ′′ = 0 (9.7)
2πα′ 2πα′
3. Wave equation and mode expansions
∞
X µ (τ, σ) = f0µ + f1µ τ +
X
(Aµn cos nτ + Bnµ sin nτ ) cos nσ (9.8)
n=1

f1µ = 2α′ pµ
p
Aµn cos nτ + Bnµ sin nτ = −i 2α′ /n (aµ∗
n e
inτ
− aµn e−inτ ), (9.9)
∞
√ cos nσ
X µ (τ, σ) = xµ0 + 2α′ pµ τ − i 2α′
X
(aµ∗
n e
inτ
− aµn e−inτ ) √ (9.10)
n=1
n
√ √ √
αµ0 = 2α′ pµ ,αµn = aµn n, αµ−n = aµ∗
n n, n ≥ 1 (9.11)
√ √ 1 µ −inτ
X µ (τ, σ) = xµ0 + 2α′ αµ0 τ + i 2α′
X
α e cos nσ (9.12)
n n
n6=0
√ X
Ẋ µ ± X µ′ = 2α′ αµn e−in(τ ±σ) (9.13)
n∈Z

4. Light-cone solution of equations of motion

 1 1 
nµ = √ , √ , 0, . . . , 0 , n · X = X +, n · p = p+ (9.14)
2 2
σ
2π
Z
+ ′ + +
X (τ, σ) = βα p τ, p σ= dσ̃ pτ + (τ, σ), β = 2, 1 (9.15)
β 0

X I = (X 2 , X 3 , . . . , X d ), Ẋ − ± X −′ = (2βα′ p+ )−1 (Ẋ I ± X I′ )2 (9.16)

31
32 CHAPTER 9. LIGHT-CONE RELATIVISTIC STRINGS

x+
0 = 0, α+ +
n = α−n = 0, n = 1, 2, . . . , ∞ (9.17)
√ 1 ⊥ 1 X 1
2α′ α−
n = + Ln , L⊥
n = αIn−p αIp , 2p+ p− = ′ L⊥ (9.18)
p 2 α 0
p∈Z

1 X ⊥ −in(τ ±σ) 1
Ẋ − ± X −′ = Ln e = (Ẋ − ± X −′ )2 (9.19)
p+ 4α′ p+
n∈Z
∞
1 X I∗ I
M 2 = 2p+ p− − pI pI = na a (9.20)
α′ n=1 n n

 Quick Calculations
9.1 e = ~c/ℓ = 2 × 10−5 × 1018 = 2 × 1013 eV = 20 TeV.

9.2 Since tµ tµ < 0, then we can prove the vector v µ is timelike:

 t · X ′ µ′  t · X′ ′  (t · X ′ )2
v µ vµ = tµ − ′ ′
X tµ − ′ ′
X µ = tµ tµ − ′ <0 (9.21)
X ·X X ·X X · X′

9.3 It is obvious that X µ (τ, σ) is real.

√
9.4 Using the relation 2α′ α− ⊥ +
0 = Ln /p , we can rewrite the expression as

√ √ X1
X − (τ, σ) = x−
0 + 2α′ α−
0 τ + i 2α
′ α− e−inτ cos nσ
n n
n6=0
(9.22)
1 ⊥ i X 1 ⊥ −inτ
= x−
0 + + L0 τ + + L e cos nσ
p p n n
n6=0

−
√
9.5 When all aIn vanish, we have L⊥ −
n = 0 for n ≥ 1. Therefore, X (τ, σ) = x0 + 2α′ α−
0 τ.

 Solutions to the Problems

9.1 (a) (nµ + bµ )(nµ + bµ ) = bµ bµ ≥ 0. Therefore, bµ is spacelike of null.
(b) If bµ is null, we have bµ (nµ −λbµ ) = 0 and λnµ )(nµ −λbµ ) = 0. Then (bµ −λnµ )(nµ −λbµ ) = 0.
Therefore, bµ = λnµ .
(c) The set of vectors bµ that satisfies nµ bµ = 0 is a hyperplane orthogonal to the vector nµ .
Therefore, it is a space of dimension (D1 ). And the subset of null vectors forms a subspace of
dimension one.
(d) For D = 2 and n · X = X + , we can obtain
 1 1   1 1 
nµ = √ ,√ , nµ = −√ ,√ (9.23)
2 2 2 2

9.2 (a) Using Eq. (9.16), we can find

(∂τ X I )2 + (∂σ X I )2 (∂τ X I )(∂σ X I )

∂τ X − = , ∂σ X − = (9.24)
2βα′ p+ βα′ p+

From the consistency condtion ∂σ ∂τ X − = ∂τ ∂σ X − , it is easy to prove that

∂σ (∂τ X I )2 + (∂σ X I )2 = 2∂τ (∂τ X I )(∂σ X I ) ⇒ ∂τ2 X I = ∂σ2 X I

   
(9.25)

(b) If the transverse coordinates X I satisfy the wave function, then it follows

∂τ2 X − − ∂σ2 X − = (βα′ p+ )−1 (∂τ2 X I )(∂τ2 X I − ∂σ2 X I ) = 0 (9.26)

(c) The coordinates satisfying Neumann boundary conditions means that ∂σ X I = 0 at σ = 0, π,

while satisfing Dirichlet boundary conditions indicates that ∂τ X I = 0. Since ∂σ X − is determined
by the product of ∂τ X I and ∂σ X I , it is always zero. Therefore, X − will always satisfy Neumann
boundary conditions.
 33

αµn αµ−n = (α′ )−1 (a2 + a2 ) = 2a2 /α′ .

P∞
9.3 (a) M 2 = (α′ )−1 n=1
√
(b) The length of the string is given by l = 2πM α′ = 2πa 2α′ . And the explicit functions of
X (2) (τ, σ) and X (3) (τ, σ) are given by
(2)
√ √
X (2) (τ, σ) = x0 + i 2α′ a(e−iτ − eiτ ) cos σ = 2 2α′ a sin τ cos σ (9.27)
(3)
√ √
X (3) (τ, σ) = x0 + i 2α′ ia(e−iτ + eiτ ) cos σ = −2 2α′ a cos τ cos σ (9.28)
1
 (2) (2) (3) (3) 
(c) L⊥
0 = 2 α−1 α1 + α−1 α1 = a2 , L ⊥ − 2 +
0 = 0 (n 6= 0). X (τ, σ) = a τ /p .

(d) τ = ωt = πt/l. The value of p+ can be determined by the following

1 √ a2 a
X 1 (τ, σ) = √ (X + − X − ) = ( 2p+ α′ − √ )τ = 0 ⇒ p+ = √ (9.29)
2 2p + 2α′

9.4 (a) √ √
X 2 (τ, σ) = x20 + i 2α′ a(e−iτ − eiτ ) cos σ = 2 2α′ a sin τ cos σ (9.30)
34 CHAPTER 9. LIGHT-CONE RELATIVISTIC STRINGS
Chapter 10

Light-cone Fields and Particles

 Summary and Supplement

1. Classical scalar fields

Z  1 1 
S= dD x − η µν ∂µ φ∂ν φ − m2 φ2 , (∂ 2 − m2 )φ = 0 (10.1)
2 2
p
φ(t, x) = ae−iEp t+ip·x + a∗ eiEp t−ip·x , Ep = p2 + m2 (10.2)
dD p ip·x
Z
φ(x) = e φ(p), (p2 − m2 )φ(p) = 0 (10.3)
(2π)D
 ∂ ∂ ∂ ∂ 
xT = (x2 , x3 , . . . , xd ), −2 + − + I − m 2
φ(x+ , x− , xT ) = 0 (10.4)
∂x ∂x ∂x ∂xI
h ∂ 1
i
pT = (p2 , p3 , . . . , pd ), i + − + pI pI + m2 φ(x+ , p+ , pT ) = 0 (10.5)
∂x 2p
p+ τ h ∂ 1
i
x+ = 2 , i − p I I
p + m 2
φ(τ, p+ , pT ) = 0 (10.6)
m ∂τ 2m2
ap , a†p = 1, a−p , a†−p = 1, a(t), ȧ† (t) = a† (t), ȧ(t) = 2iEp
       
(10.7)

H = Ep a†p ap + a†−p a−p , P = p a†p ap − a†−p a−p

(10.8)

2. Maxwell fields
∂ 2 Aµ − ∂ µ (∂ · A) = 0, p2 Aµ − pµ (p · A) = 0 (10.9)
µ µ µ +
A (p) → A (p) + ip ǫ(p), A (p) = 0 (10.10)
p · A = 0, p+ A− = pI AI , p2 AI = 0 (10.11)

3. Gravitational fields
p2 hµν − pα (pµ hνα + pν hµα + pµ pν h = 0 (10.12)
µν µ ν ν µ
δh (p) = ip ǫ (p) + ip ǫ (p), δh = 2ip · ǫ (10.13)
h++ = h+− = h+I = 0, h = 0, hII = 0, p2 hµν = 0 (10.14)
p+ hI− = pJ hIJ , p+ h−− = pI h−I , p2 hIJ = 0 (10.15)

 Quick Calculations

10.1 By the definition, we have

h1 1 1 i 1 1 1
H = Π∂0 φ − L = Π2 − Π2 − (∇φ)2 − m2 φ2 = Π2 + (∇φ)2 + m2 φ2 (10.16)
2 2 2 2 2 2

10.2 With the mass-shell condition, we have

1
p2 + m2 = −2p+ p− + pI pI + m2 = 0 ⇒ p− = p I p I + m2


(10.17)
2p+

35
36 CHAPTER 10. LIGHT-CONE FIELDS AND PARTICLES

10.3 ∂0 φp and ∇φp are given by

1 i
ȧ(t)eip·x + ȧ∗ (t)e−ip·x , a(t)eip·x − a∗ (t)e−ip·x p (10.18)
   
∂0 φp = p ∇φp = p
2V Ep 2V Ep
Using the quantization condition, we can integrate the action as
Z Z h 1 1 1 i
S = dt dd x ȧ(t)ȧ∗ (t) − a(t)a∗ (t)p2 − a(t)a∗ (t)m2
2V Ep 2V Ep 2V Ep
1 h 1 1
Z Z i
= dt dd x ȧ(t)ȧ∗ (t) − Ep a(t)a∗ (t) (10.19)
V 2Ep 2
Z h 1 1 i
= dt ȧ(t)ȧ∗ (t) − Ep a(t)a∗ (t)
2Ep 2

10.4 With the quantization condition, it is easy to obtain

Z h 1 1 1 i
H = dd x ȧ(t)ȧ∗ (t) + a(t)a∗ (t)p2 + a(t)a∗ (t)m2
2V Ep 2V Ep 2V Ep
(10.20)
1 1
= ȧ(t)ȧ∗ (t) + Ep a(t)a∗ (t)
2Ep 2

10.5 Using the relations a(t), ȧ† (t) = a† (t), ȧ(t) = 2iEp , we can easily check that
   

1  1
a(t) − a† (t), ȧ(t) − ȧ† (t) = −

[q2 (t), p2 (t)] = − (−2iEp − 2iEp ) = i (10.21)
4Ep 4Ep
1  1
a(t) + a† (t), ȧ(t) − ȧ† (t) =

[q1 (t), p2 (t)] = (2iEp − 2iEp ) = 0 (10.22)
4iEp 4iEp

10.6 Using the facts ap , a†k = δp,k and a†p , a†k = 0, we can obtain
   

X †
P a†p1 a†p2 · · · a†pn |Ωi = kak (a†p1 ak + δp1 ,k )a†p2 · · · a†pn |Ωi
k
X
= ka†p1 a†k (a†p2 ak + δp2 ,k ) · · · a†pn |Ωi + p1 a†p1 a†p2 · · · a†pn |Ωi
k
(10.23)
k
X
= pn a†p1 a†p2 · · · a†pn |Ωi
n=1
Pk
Similarly, we can prove that Ha†p1 a†p2 · · · a†pn |Ωi = n=1 En a†p1 a†p2 · · · a†pn |Ωi.

10.7 We only need to prove that N a†p1 a†p2 · · · a†pn |Ωi = na†p1 a†p2 · · · a†pn |Ωi.
 Solutions to the Problems
10.1 Using the results in Eq. (10.18) and the quantizition condition
Z L1 Z Ld
1
dx · · · dxd exp(±2ip · x) = 0, (10.24)
0 0

we can evaluate the integral and obtain

ip
Z
dd x ȧ(t)eip·x + ȧ∗ (t)e−ip·x a(t)eip·x − a∗ (t)e−ip·x
  
P =−
2V Ep
ip
Z
dd x ȧ∗ (t)a(t) − ȧ(t)a∗ (t)
 
=− (10.25)
2V Ep
ip ∗
=− (ȧ a − ȧa∗ )
2Ep
Then, with the exoression of a(t), it is easy to get
ip
(iEp ) a∗p eiEp t − a−p e−iEp t ap e−iEp t + a∗−p eiEp t



P =−
2Ep
(10.26)
− − ap e−iEp t + a∗−p eiEp t a∗p eiEp t + a−p e−iEp t




= p(a∗p ap − a∗−p a−p )

 37

10.2 (a) Using the quantizition condition, we can prove that

1 1 X
Z Z
′ ′ ′
dx′ f (x′ )e−ip·x = dx′ f (p′ )ei(p −p)·x = f (p) (10.27)
V V ′
p

Plug this back into the Fourier series, we can obtain a representation for the delta function:
1 X 1 X ip·(x−x′ )
Z Z
′
′ d
′ ′
dx f (x )δ (x − x ) = dx′ f (x′ )eip·(x−x ) ⇒ δ d (x − x′ ) = e (10.28)
V p V p

(b) For the complete scale field expansion, we have

i Xp
Ep a†p eiEp t−ip·x − ap e−iEp t+ip·x


Π(t, x) = ∂0 φ(t, x) = √ (10.29)
2V p

Then, it is very easy to show that

φ(t, x), π(t, x′ ) = iδ d (x − x′ )

 
(10.30)

10.3 (a) For the Lorentz covariant equation Aµ = B µ , µ = 0, 1, . . . , d, we have

1 1
A+ = √ (A0 + A1 ) = √ (B 0 + B 1 ) = B + (10.31)
2 2
1 1
A− = √ (A0 − A1 ) = √ (B 0 − B 1 ) = B − (10.32)
2 2
It is obvious that AI = B I for I = 2, . . . , d.
(b) By the definition, we can obtain
1 1 1
R++ = A+ B + = √ (A0 + A1 ) √ (B 0 + B 1 ) = (R00 + R01 + R10 + R11 ) (10.33)
2 2 2
1 1 1
R+− = A+ B − = √ (A0 + A1 ) √ (B 0 − B 1 ) = (R00 − R01 + R10 − R11 ) (10.34)
2 2 2
1 1 1
R−+ = A− B + = √ (A0 − A1 ) √ (B 0 + B 1 ) = (R00 + R01 − R10 − R11 ) (10.35)
2 2 2
1 1 1
R−− = A− B − = √ (A0 − A1 ) √ (B 0 − B 1 ) = (R00 − R01 − R10 + R11 ) (10.36)
2 2 2
µν µν
Then, it is easy to see that an equaltiy R = S between Lorentz tensors implies the equality
of the light-cone components.
(c) For the Minkowski metric η = diag(−1, 1, 1, 1), the relations above become

η ++ = η −− = 0, η +− = η −+ = −1 (10.37)

(d) For the antisymmetric electromagnetic field strength

 
0 Ex Ey Ez
−E x 0 Bz −By 
F µν = 

, (10.38)
−Ey −Bz 0 Bx 
−Ez By −Bx 0

we can calculate the light-cone components

F ++ = F −− = 0, F +− = −Ex , F −+ = Ex (10.39)
1 1
F +2 = √ (Ey + Bz ), F +3 = √ (Ez − By ) (10.40)
2 2
1 1
F −2 = √ (Ey − Bz ), F −3 = √ (Ez + By ) (10.41)
2 2
F 22 = F 33 = 0, F 23 = Bx , F 32 = −Bx (10.42)
38 CHAPTER 10. LIGHT-CONE FIELDS AND PARTICLES

10.4 For the uniform constant electric field E = E0 ex , we can choose

Aµ = (−E0 x, E0 x, 0, 0), (10.43)

which automatically satisfies the condition A+ = 0. Then, we have

2 √
A− = √ (A0 − A1 ) = − 2E0 x = E0 (x− − x+ ), AI = 0 (10.44)
2

10.5 A pure gauge of a gravitational field is defined as hµν (p) = ipµ ǫν (p) + ipν ǫµ (p). Then, from the
equation of the field

(−i)(pµ pα hνα + pν pα hµα )

p2 hµν = pα (pµ hνα + pν hµα ) ⇒ hµν = i (10.45)
p2

we can see that: if p2 6= 0, setting ǫµ = pα hµα and ǫν = pα hνα will just yield a pure gauge.
10.6 (a) Using the antisymmetric property of Bµν , we can obtain

Hµνρ + Hνµρ = ∂µ (Bνρ + Bρν ) + ∂ν (Bρµ + Bµρ ) + ∂ρ (Bµν + Bνµ ) = 0 (10.46)

Similarly, Hµνρ + Hρνµ = 0, Hµνρ + Hµρν = 0. Therefore, Hµνρ is totally antisymmetric. Under
the gauge transformation δBµν = ∂µ ǫν − ∂ν ǫµ , we have

δHµνρ = ∂µ (∂ν ǫρ − ∂ρ ǫν ) + ∂ν (∂ρ ǫµ − ∂µ ǫρ ) + ∂ρ (∂µ ǫν − ∂ν ǫµ ) = 0 (10.47)

(b) It is easy to see that

∂µ ǫ′ν − ∂ν ǫ′µ = ∂µ (ǫν + ∂ν λ) − ∂ν (ǫµ + ∂µ λ) = ∂µ ǫν − ∂ν ǫµ (10.48)

(c) In the momentum space, ǫ′µ (p) = ǫµ + pµ λ(p) will generate the same gauge transformation
as ǫµ (p). Then, we have

1 ǫ+
ǫ′+ (p) = √ ǫ0 + p0 λ + ǫ1 + p1 λ = 0 ⇒ λ = − +


(10.49)
2 p

(d)
10.7 (a) Recall that Fµν = ∂µ Aν − ∂ν Aµ , then we have

1
δL = − δ(Fµν F µν ) − m2 Aµ ∂µ ǫ − bm∂µ φ∂ µ ǫ − mφ∂ 2 ǫ − bm2 (∂ · A)ǫ
4 (10.50)
= −m2 ∂µ (Aµ ǫ) − m∂µ (φ∂ µ ǫ)

where we have chosen b = 1. So the action will be invariant under the gauge transformation.
(b) The field equations are given by

∂L ∂L
− ∂µ = 0 ⇒ m2 Aν + ∂µ F µν = 0 (10.51)
∂Aν ∂∂µ Aν

∂L ∂L
− ∂µ = 0 ⇒ ∂ 2 φ − m(∂ · A) = 0 (10.52)
∂φ ∂∂µ φ
(c) If we set ǫ = −φ/m, then φ′ = φ + δφ = 0. The second field equation becomes ∂ · A = 0.
(d) The simplified equations in the momentum space can be written as

(p2 + m2 )Aν − (p · A)pν = 0, p · A = 0 ⇒ (p2 + m2 )Aν = 0 (10.53)

If p2 6= −m2 , we can only have trivial solutions Aν = 0.

Chapter 11

The Relativistic Quantum Point

Particle

 Summary and Supplement

1. Light-cone point particle

Z τf p
ηµν ẋµ ẋν = ẋ2 , S= Ld τ, L = −m −ẋ2 (11.1)
τi

p+ 1
x+ = τ, ẋ2 = − , pµ = m2 ẋµ , p 2 + m2 = 0 (11.2)
m2 m2
1 p− pI
p− = (pI pI + m2 ), x− (τ ) = x−
0 + τ, xI (τ ) = xI0 + τ (11.3)
2p+ m2 m2

2. Quantization of the point particle

xI , x− I +
xI (τ ), x− I +



0 ,p ,p , 0 (τ ), p (τ ), p (τ ) (11.4)

xI (τ ), pJ (τ ) = iη IJ , x− + −+
   
0 (τ ), p (τ ) = iη = −i (11.5)
1 + 1  I
p (τ )p− (τ ) = p (τ )pI (τ ) + m2

H(τ ) = 2 2
(11.6)
2m 2m
∂ 1
i ψ(τ, p+ , pT ) = pI pI + m2 ψ(τ, p+ , pT )


2
(11.7)
∂τ 2m
†
 +
ψ(τ, p+ , pT ) ↔ φ(τ, p+ , pT )

 p , pT ↔ a +
p ,pT |Ωi , (11.8)

3. Light-cone Lorentz generators

δxµ (τ ) = ǫµν xν (τ ), M µν = xµ (τ )pν (τ ) − xν (τ )pµ (τ ) (11.9)

(M µν )† = M µν ,
 µν ρ 
M , x (τ ) = iη µρ xν (τ ) − iη νρ xµ (τ ) (11.10)
[M µν , M ρσ ] = iη µρ M νσ − iη νρ M µσ + iη µσ M ρν − iη νσ M ρµ (11.11)
1 1 I −
M +− = − (x− p+ + p+ x−
0 ), M −I = x− I − I
0 p − (x0 p + p x0 ) (11.12)
2 0 2
 Quick Calculations

11.1 For the state |Ψ, ti = e−iHt |Ψi, we can easily check that

d
i |Ψ, ti = i(−iH)e−iHt |Ψi = H |Ψ, ti (11.13)
dt

11.2 For the corresponding Heisenberg operators, we have

α1 (t), α2 (t) = eiHt α1 e−iHt , eiHt α2 e−iHt = eiHt α1 , α2 e−iHt = α3 (t)

     
(11.14)

39
40 CHAPTER 11. THE RELATIVISTIC QUANTUM POINT PARTICLE

11.3 Since [x− I

0 (τ ), p (τ )] = 0, then we can obtain

dx−
0 (τ ) 1  −
x0 (τ ), pI pI + m2 = 0

i = 2
(11.15)
dτ 2m
11.4 For ǫ− 6= 0 and ǫ+ = ǫI = 0, it is easy to check that
δx− (τ ) = iǫρ pρ (τ ), x− (τ ) = iǫ+ p+ (τ ), x− (τ ) = −ǫ+ = ǫ−
   
(11.16)

11.5 Using the fact [x− +

0 , 1/p ] = i/p
+2
, we can show that
  − 1 I I 2
 1 − 1  I I 2 p I p I + m2 p−
x− −

0 , p = x0 , (p p + m ) = x , (p p + m ) = i = i (11.17)
2p+ 2 0 p+ 2p+2 p+

11.6 According to Eq. (11.5), we have

 µν ρ 
M , x (τ ) = xµ (τ ) pν (τ ), xρ (τ ) − xν (τ ) pµ (τ ), xρ (τ ) = iη µρ xν (τ ) − iη νρ xµ (τ )
   
(11.18)

11.7 We observe that: the first two terms are antisymmetric for the indices µ and ν, and the last two
terms are also antisymmetric for the indices µ and ν.
11.8 By the definition, we can directly prove that
1 0 1
M +− = x+ p− − x− p+ = (x + x1 )(p0 − p1 ) − (x0 − x1 )(p0 + p1 ) = x1 p0 − x0 p1 = M 10 (11.19)
2 2
 Solutions to the Problems
11.1 For the Heisenberg operator, we can prove its equation of motion
dξ(t) d
i = i eiHt ξe−iHt = i(iHeiHt )ξe−iHt + ieiHt ξ(−iHe−iHt ) = [ξ(t), H] (11.20)
dt dt
11.2 (a) According to the Schrödinger equation, we have
d |Ψ, ti dU (t) dU (t)
i =i |Ψi = HU (t) |Ψi ⇒ i = HU (t) (11.21)
dt dt dt
(b) This result has been proven in Eq. (11.20).
(c) This result has been proven in Eq. (11.14).
11.3 Using the Hamilton’s canonical equations
dq ∂H dp ∂H
= , =− , (11.22)
dt ∂p dt ∂q
we can prove the time evolution of an operator in the classical phase space:
dv ∂v ∂v dp ∂v dq ∂v ∂v ∂H ∂v ∂H ∂v
= + + = − + = + {v, H} (11.23)
dt ∂t ∂p dt ∂q dt ∂t ∂p ∂q ∂q ∂p ∂t

11.4 The variation δL can be derived as follows

∂L µ ∂L µ d  ∂L  ∂L
δL = µ
δx + µ
δ ẋ = µ
λτ ∂τ X µ + µ ∂τ (λ∂τ X µ ) = ∂τ (λL) (11.24)
∂x ∂ ẋ dτ ∂ ẋ ∂ ẋ
And the associated charge is given by
∂L µ
λ(τ )Q = δx (τ ) − λ(τ )L = 0 (11.25)
∂ ẋµ
11.5 (a) [M µν , pρ ] = [xµ pν − xν pµ , pρ ] = [xµ , pρ ]pν − [xν , pρ ]pµ = iη µρ pν − iη νρ pµ .
(b) [xµ pν , xρ pσ ] = [xµ , xρ pσ ]pν + xµ [pν , xρ pσ ] = iη µσ xρ pν − iη νρ xµ pσ . Then, we have
[M µν , M ρσ ] = [xµ pν , xρ pσ ] − [xµ pν , xσ pρ ] − [xν pµ , xρ pσ ] + [xν pµ , xσ pρ ]
= (iη µσ xρ pν − iη νρ xµ pσ ) − (iη µρ xσ pν − iη νσ xµ pρ )
(11.26)
− (iη νσ xρ pµ − iη µρ xν pσ ) + (iη νρ xσ pµ − iη µσ xν pρ )
= iη µρ M νσ − iη νρ M µσ + iη µσ M ρν − iη νσ M ρµ
(c) In the light cone coordinates, we have η +− = η −+ = −1, η II = 1 and the others are zero.
 41

11.6 (a) First, we calculate the value of [xI0 , p− ].

1 pI I I ipI
[xI0 , p− ] = [xI0 , (p I I
p + m 2
)] = [x0 , p ] = (11.27)
2p+ p+ p+
Then, it is easy to obtain

I 1 I − − I − I I − 1 I − − I I − ipI
M −I = x−
0 p − (x0 p + p x0 ) = x0 p − x0 p + [x0 , p ] = x0 p − x0 p + (11.28)
2 2 2p+

(b) Now, we shall prove the light-cone gauge commutator

I ipJ I J − I − − J I − J − ipI − J
[M −I , M −J ] = [x−
0p , ] − [x−0 p , x0 p ] − [x0 p , x0 p ] + [x0 p , x0 p ] + [ ,x p ]
2p + 2p+ 0
ipI pJ ixJ pI p− ixI pJ p− ixJ pI p− ixI pJ p− ipI pJ
= − +2 − 0 + + 0 + +( 0 + − 0 + ) + +2 = 0
2p p p p p 2p
(11.29)

where we have used the following relations:

ipI 1 i ip−
[xI0 , p− ] = , [x−
0 , ] = +2 , [x− −
0 ,p ] = (11.30)
p+ p+ p p+

11.7 (a) By the definition, we have

1 p− 1 p− p+ p− 1
M +− = − (x− − 2 τ )p+ − p+ (x− − 2 τ ) = 2
τ − (x− p+ + p+ x− ) (11.31)
2 m 2 m m 2
(b)
42 CHAPTER 11. THE RELATIVISTIC QUANTUM POINT PARTICLE
Chapter 12

Relativistic Quantum Open Strings

 Summary and Supplement

1. Light-cone Hamiltonian and commutators

(xI (σ), x− τI +
xI (τ, σ), x− τI +


0 , P (σ), p ), 0 (τ ), P (τ, σ), p (τ ) (12.1)

[X I (τ, σ), P τ J (τ, σ)] = iη IJ δ(σ − σ ′ ), [x− +

0 (τ ), p (τ )] = −i (12.2)
Z π
X + = 2α′ p+ τ, L⊥ ′ + −
0 = 2α p p , H = 2α′ p+ dσ P τ − (12.3)
0
π
X I′ (τ, σ)X I′ (τ, σ) i
Z h
H(τ ) = πα′ dσ P τ I (τ, σ)P τ I (τ, σ) + (12.4)
0 (2πα′ )2
2. Commutation relations for oscillators
 I d
(Ẋ ± X I′ )(τ, σ), (Ẋ J ± X J′ )(τ, σ ′ ) = ±4πα′ iη IJ δ(σ − σ ′ )

(12.5)
dσ
 I I′ J J′
(Ẋ ± X )(τ, σ), (Ẋ ∓ X )(τ, σ ) = 0, σ, σ ′ ∈ [0, π]
′

(12.6)
X ′ ′ ′ d
e−im (τ +σ) e−in (τ +σ ) [αIm′ , αJn′ ] = 2πiη IJ δ(σ − σ ′ ) (12.7)
dσ
m′ ,n′ ∈Z
√
[αIm , αJ−n ] = mδmn η IJ , αI0 = 2α′ pI (12.8)
√
[xI0 , αJ0 ] = 2α′ iη IJ , [xI0 , αJn ] = 0, n 6= 0 (12.9)
√ √
αIn = aIn n, αI−n = aI† n n, (αIn )† = αI−n (12.10)
[aIm , aJ† IJ
n ] = δmn η , [aIm , aIn ] = [aI† J†
m , an ] = 0 (12.11)

3. Strings as harmonic oscillators

π
1
Z Z Z
S = dτ dσ L = dτ dσ (Ẋ I Ẋ I − X I′ X I′ ) (12.12)
4πα′ 0

∞
√ X cos nσ
X I (τ, σ) = q I (τ ) + 2 α′ qnI (τ ) √ (12.13)
n=1
n
∞
√ X
inτ cos nσ
X I (τ, σ) = xI0 + 2α′ pI τ + i 2α′ (aIn e−inτ − aI†
n e ) √ (12.14)
n=1
n

4. Transverse Virasoro operators

∞
√ 1 ⊥ 1X I
2α′ α−
n = + Ln , L⊥
n = αn−p αIp , (L⊥ † ⊥
n ) = L−n (12.15)
p 2
p∈Z

∞ ∞
1 I I X I I 1 X
L⊥
0 = α0 α0 + α−p αp + (D − 2) p (12.16)
2 p=1
2 p=1

43
44 CHAPTER 12. RELATIVISTIC QUANTUM OPEN STRINGS

∞
1 1 X
2α′ p− = (L⊥ + a), M2 = (a + naI† I
n an ) (12.17)
p+ 0 α′ n=1

∞ ∞
1 X X 1 1
a= (D − 2) p, ζ(s) = , ζ(−1) = − (12.18)
2 p=1 n=1
ns 12

I I I
√
[L⊥ [L⊥ ′ I
m , αn ] = −nαm+n , m , x0 ] = −i 2α αm (12.19)
D−2 3
[L⊥ ⊥ ⊥
m , L−n ] = (m + n)Lm−n + (m − m)δmn (12.20)
12
I τ I σ I′
[L⊥
m , X (τ, σ)] = ξm Ẋ + ξm X (12.21)
τ
ξm (τ, σ) = −ieimτ cos mσ, σ
ξm (τ, σ) = eimτ sin mσ (12.22)

X I (τ + ǫξm
τ σ
, σ + ǫξm ) = X I (τ, σ) + ǫ[L⊥ I
m , X (τ, σ)] (12.23)

5. Lorentz generators
π
1
Z
M µν = (X µ Ẋ ν − X ν Ẋ µ ) dσ (12.24)
2πα′ 0
∞
1 µ µ
M µν = xµ0 pν − xν0 pµ − i
X
(α−n αn − αν−n ανn ) (12.25)
n=1
n
∞
I 1  I ⊥ I i X 1 ⊥ I
M −I = x− ⊥
(L−n αn − αI−n L⊥

0p − x0 (L 0 + a) + (L 0 + a)x0 − √ n ) (12.26)
4α′ p+ ′ +
2α p n=1 n
∞
1 X I
h  D − 2 1 D − 2 i
[M −I , M −J ] = ′ +2
α−m αJm − αJ−m αIm × m 1 − + +a (12.27)
α p m=1 24 m 24

1
D = 26, a = −1, [M −I , M −J ] = 0, 2α′ p− = (L⊥ − 1), H = L⊥
0 −1 (12.28)
p+ 0

6. Constructing the state space

aIn p+ , pT = 0,
 
n ≥ 1, I = 2, . . . , 25 (12.29)

25
∞ Y
Y
(aI† λn.I  +
 
|λi = n ) p , pT (12.30)
n=1 I=2

∞
X 1
N⊥ = naI† I
n an , M2 = (−1 + N ⊥ ) (12.31)
n=1
α′

[N ⊥ , aI† I†
n ] = nan , [N ⊥ , aIn ] = −naIn (12.32)
25
∞ X
X
N ⊥ |λi = Nλ⊥ |λi , Nλ⊥ = nλn,I (12.33)
n=1 I=2

tachyons: N ⊥ = 0, α′ M 2 = −1, number of states = 1 (12.34)

photons: N ⊥ = 1, α′ M 2 = 0, number of states = D − 2 (12.35)

massive tensors: N ⊥ = 2, α′ M 2 = 1, number of states = (D − 2)(D + 1)/2 (12.36)

7. Tachyons and D-brane decay

 Quick Calculations

12.1 Since (xI0 )† = xI0 , (αI0 )† = αI0 , and the index n is summed over all integers except zero, we can

†
see that X I (τ, σ) = X I (τ, σ).
 45

12.2 Using the definition in Eq. (12.12), we can obtain

π ∞ ∞
1 cos2 nσ
Z Z h X X i
S= dτ dσ q̇ I (τ )q̇ I (τ ) + 4α′ q̇nI (τ )q̇nI (τ ) − 4α′ qnI (τ )qnI (τ )n sin2 nσ
4πα′ 0 n=1
n n=1
∞ ∞
1 π nπ i
Z h X X
I I I I
= ′
dτ π q̇ (τ )q̇ (τ ) + 4α′
q̇n (τ ) q̇n (τ ) − 4α ′
qnI (τ )qnI (τ )
4πα n=1
2n n=1
2
h 1 ∞ 
1 I n I
Z X i
I I I I
= dτ q̇ (τ ) q̇ (τ ) + q̇n (τ )q̇n (τ ) − qn (τ )qn (τ )
4α′ n=1
2n 2
(12.37)

12.3 Using the relations in Eq. (12.9), we can see that

I 1X J √ √
[L⊥
m , x0 ] = [αm−p αJp , xI0 ] = [αI0 , xJ0 ]αJm = −i 2α′ η IJ αJm = −i 2α′ αIm (12.38)
2
p∈Z

12.4 For p 6= 1, α2−p αp = αp α2−p , therefore we have

1X 1
L2 = α2−p αp = α1 α1 + (α0 α2 + α−1 α3 + α−2 α4 + · · · ) (12.39)
2 2
p∈Z

Similarly, for p 6= −1, α−2−p αp = αp α−2−p , therefore we have

1X 1
L−2 = α−2−p αp = α−1 α−1 + (α−2 α0 + α−3 α1 + α−4 α2 + · · · ) (12.40)
2 2
p∈Z

12.5 For the oscillators including Lorentz indices, we can verify that
1 I I J J 1
[α1 α1 , α−1 α−1 ] = (αI1 [αI1 , αI−1 ]αJ−1 + αJ−1 [αI1 , αI−1 ]αI1 )
4 2 (12.41)
1 D−2
= (η IJ αI1 αJ−1 + η IJ αJ−1 αI1 ) = + αI−1 αI1
2 2

12.6 It is easy to check the expressions. Note that the numbers n, m − n, p and m − p that appear
on the oscillator are all positive.
Pm
12.7 Suppose that n=1 n2 = am3 + bm2 + cm, then you can use the specila cases for m = 1, 2 , 3
to determine the values of a, b, c.

12.8 For the transformation generated by i(L⊥ ⊥

m + L−m ), the parameters are given by

ξ τ = i(ξm
τ τ
+ ξ−m ) = (eimτ + e−imτ ) cos mσ = 2 cos mτ cos mσ (12.42)
σ σ σ imτ −imτ
ξ = i(ξm + ξ−m ) = i(e −e ) sin mσ = −2 sin mτ sin mσ (12.43)

12.9 Since M µν is τ -independent, it suffices to pick up the τ -independent terms that arise in the
products. Then, we have
√
Z π
1 X1
M µν = dσ 2α′ (xµ0 αν0 − xν0 αµ0 ) + i2α′ (αµn αν−n − ανn αµ−n ) cos2 nσ

2πα 0′ n
n6=0
∞ (12.44)
µ ν ν µ
X 1 µ µ ν ν
= x0 p − x0 p − i (α α − α−n αn )
n=1
n −n n

12.10 If the basis states |λi and |λ′ i are different, then we have

(λ′ , λ) = δ IJ δ(p′+ − p+ )δ(p′T − pT ) = 0 (12.45)

12.11 Substitute the expression for |Ψ, ti into the Schrödinger’s equation satisfied by the general states,
then the result will be obvious.
46 CHAPTER 12. RELATIVISTIC QUANTUM OPEN STRINGS

12.12 Using the Euler-Lagrange equation, we can directly obtain that

∂L ∂L
− ∂µ = 0 ⇒ −V ′ (φ) + ∂µ ∂ µ φ = 0 (12.46)
∂φ ∂∂µ φ

 Solutions to the Problems

12.1
12.2 For the mode expansion of X I (τ, σ), its derivatives are given by
√ X √ X
Ẋ I (τ, σ) = 2α′ αIn e−inτ cos nσ, X I′ (τ, σ) = −i 2α′ αIn e−inτ sin nσ (12.47)
n∈Z n∈Z

Using the commutator in Eq. (12.8), we can easily obtain

 I′ X X
X (τ, σ), X J′ (τ, σ ′ ) = −2α′ αIm e−imτ sin mσ, αJn e−inτ sin nσ ′
 

m∈Z n∈Z
X (12.48)
′ IJ ′
= 2α η m sin mσ sin mσ = 0
m∈Z
X X
Ẋ I (τ, σ), Ẋ J (τ, σ ′ ) = 2α′ αIm e−imτ cos mσ, αJn e−inτ cos nσ ′
  

m∈Z n∈Z
X (12.49)
′ IJ
= 2α η m cos mσ cos mσ ′ = 0
m∈Z

12.3 (a) Since αI0 commutes with all other oscillators, it does not contribute to our calculations
presented here. So we have
√ X 1 1 X J −inτ
[X I (τ, σ), P τ J (τ, σ ′ )] = xI0 + i 2α′ αIm e−imτ cos mσ, √ cos nσ ′
 
αn e
m π 2α′ n∈Z
m6=0
1 i X
= √ [xI0 , αJ0 ] + η IJ cos mσ cos mσ ′
π 2α′ π
m6=0
1 X
= iη IJ cos mσ cos mσ ′
π
m∈Z
(12.50)

(b) By the basic result of Fourier series, we can obtain

Z π Z π
1X
f (σ ′ )δ(σ − σ ′ ) dσ ′ = f (σ) = cos nσ f (σ ′ ) cos nσ ′ dσ ′ (12.51)
0 π 0 n∈Z

Then, it gives the following representation of the delta function

1X
δ(σ − σ ′ ) = cos nσ cos nσ ′ (12.52)
π
n∈Z

P∞
12.4 For |x| < 1, the series n=0 xn = 1/(1 − x) holds. Then, we have
∞ ∞ ∞ ∞ Z ∞
ts − 1 1
Z Z X X
dt t = dt e −nt
= s
dx e−x xs−1 = Γ(s)ζ(s) (12.53)
0 e −1 0 n=1 n=1
n 0

For the small t, the following expansion holds

1 1 1 t t2 t2  1 1 t
t
≃ t2
= 1 − − + + O(t3 ) = − + + O(t2 ) (12.54)
e −1 t(1 + 2t + 6)
t 2 6 4 t 2 12

And it is obvious that

Z 1
1 1 t
1 1 1
dt ts−1 − + − + − + = 0, (12.55)
0 t 2 12 s − 1 2s 12(s + 1)
 47

therefore we can rewrite Eq. (12.53) as

Z 1 Z ∞
1 1 1 t
1 1 1 1
Γ(s)ζ(s) = dt ts−1 t − + − + − + + dt t (12.56)
0 e − 1 t 2 12 s − 1 2s 12(s + 1) 1 e −1
Using the small t expansion, we can see that the function of the first integral has the order of
O(ts+1 ). For ℜ(s) > −2, the integral on [0, 1] will always converge. So the right-hand side above
is well defined. Recall the pole struction of Γ(s): Res[Γ(s), −n] = (−1)n /n!, we have
1 1 1 1
Res[Γ(s), 0]ζ(0) = − ⇒ ζ(0) = − Res[Γ(s), −1]ζ(−1) = ⇒ ζ(−1) = − (12.57)
2 2 12 12
We can also obtain the result from the following formula:
πs
ζ(s) = 2s π s−1 sin Γ(1 − s)ζ(1 − s) (12.58)
2

12.5 For m + n 6= 0, [L⊥ ⊥ ⊥

m , Ln ] = (m − n)Lm+n . Then, we can check that

[L⊥ ⊥ ⊥ ⊥ ⊥ ⊥
m , Ln ] = (m − n)Lm+n = −(n − m)Ln+m = −[Ln , Lm ] (12.59)

[L⊥ ⊥ ⊥ ⊥ ⊥ ⊥ ⊥ ⊥ ⊥
m , [Ln , Lk ]] + [Ln , [Lk , Lm ]] + [Lk , [Lm , Ln ]]

= [L⊥ ⊥ ⊥ ⊥ ⊥ ⊥
m , (n − k)Ln+k ] + [Ln , (k − m)Lk+m ] + [Lk , (m − n)Lm+n ]
(12.60)
= (n − k)(m − n − k) + (k − m)(n − k − m) + (m − n)(k − m − n) L⊥
 
m+n+k

=0
For the Virasoro operators built with just one type of oscillator, we have
1
[Lm , Ln ] = − (n − m)Ln+m + (n3 − n)δn+m,0 = −[Ln , Lm ]
 
(12.61)
12
[Lm , [Ln , Lk ]] + [Ln , [Lk , Lm ]] + [Lk , [Lm , Ln ]]
= [Lm , (n − k)Ln+k ] + [Ln , (k − m)Lk+m ] + [Lk , (m − n)Lm+n ]
 
= (n − k)(m − n − k) + (k − m)(n − k − m) + (m − n)(k − m − n) Lm+n+k (12.62)
1
(n − k)(m3 − m) + (k − m)(n3 − n) + (m − n)(k 3 − k) δm+n+k,0

+
12
=0

12.6 (a) A(−m) = −A(m), A(0) = 0.

(b) Now, we consider the Jacobi identity for Lm , Ln and Lk with m + n + k = 0.
[Lm , [Ln , Lk ]] + [Ln , [Lk , Lm ]] + [Lk , [Lm , Ln ]]
= [Lm , (n − k)Ln+k ] + [Ln , (k − m)Lk+m ] + [Lk , (m − n)Lm+n ]
 
= (n − k)(m − n − k) + (k − m)(n − k − m) + (m − n)(k − m − n) L0 (12.63)
 
+ (n − k)A(m) + (k − m)A(n) + (m − n)A(k)
= (n − k)A(m) + (k − m)A(n) + (m − n)A(k)
(c) From the identity above, we can obtain the following difference equation:
(m − n)A(m + n) = (m + 2n)A(m) − (2m + n)A(n) (12.64)

12.7 (a) If L1 |λi = 0 and L2 |λi = 0, then we have [L2 , L1 ] |λi = L3 |λi = 0, [L3 , L1 ] |λi = 2L4 |λi = 0,
and so on. Therefore, the state is annihilated by all Ln with n ≥ 1.
(b) [L1 , L0 ] = L1 , [L0 , L−1 ] = L−1 , [L1 , L−1 ] = 2L0 . They form a subalgebra of the Virasoro
algebra. There are no central terms here.
12.8 (a) The combination L⊥ ⊥
m − L−m reparametrizes the σ coordinate of the string while keeping
τ = 0. They form a subalgebra of the Virasoro algebra. For m 6= ±n, we have
[L⊥ ⊥ ⊥ ⊥ ⊥ ⊥ ⊥ ⊥
m − L−m , Ln − L−n ] = (m − n)(Lm+n − L−m−n ) − (m + n)(Lm−n − Ln−m ) (12.65)
(b)
48 CHAPTER 12. RELATIVISTIC QUANTUM OPEN STRINGS

12.9 (a) It is easy to verify that

τ
ξ̇m = meimτ cos mσ = ξm
σ′
, ξ˙m
σ
= imeimτ sin mσ = ξm
τ′
(12.66)

(b) For the change of coordinates, we have

∂τ X = (1 + ǫ∂τ ξ τ )∂τ ′ X, ∂σ X = (1 + ǫ∂σ ξ σ )∂σ′ X (12.67)

∂τ X · ∂σ X = (1 + ǫ∂σ ξ σ )2 ∂τ ′ X · ∂σ′ X = 0 ⇒ ∂τ ′ X · ∂σ′ X = 0 (12.68)

(∂τ X)2 + (∂σ X)2 = (1 + ǫ∂σ ξ σ )2 (∂τ ′ X)2 + (∂σ′ X)2 = 0 ⇒ (∂τ ′ X)2 + (∂σ′ X)2 = 0
 
(12.69)

12.10 (a) If the orientation of this second string is the direction of decreasing σ, then it equals the first
one.
(b)

12.11 (b) The critical points are given by V ′ (φ) = 0.

1
V1′ (φ) = φ(φ − φ0 ) = 0 ⇒ φc = 0, φ0 (12.70)
α′ φ0

1  φ2  φ0
V2′ (φ) = − ′
|φ| 1 + ln 2 = 0 ⇒ φc = 0, ± √ (12.71)
2α φ0 e
1
V3′ (φ) = φ(φ2 − φ20 ) = 0 ⇒ φc = 0, ±φ0 (12.72)
2α′ φ20
(c) The mass of the scalar particle for the critical point φ̄ is given by m(φ̄) = V ′′ (φ̄).
1 1 1
V1′′ (φ) = (2φ − φ0 ) ⇒ m(0) = − ′ , m(φ0 ) = ′ (12.73)
α′ φ0 α α

1  φ2   φ 
0 1
V2′′ (φ) = − 2 + ln ⇒ m(0) = ∞, m ±√ =− ′ (12.74)
2α′ φ20 e 2α
1 1 1
V3′′ (φ) = (3φ2 − φ20 ) ⇒ m(0) = − ′ , m(±φ0 ) = ′ (12.75)
2α′ φ20 2α α
Chapter 13

Relativistic Quantum Closed

Strings

 Summary and Supplement

1. Mode expansions and commutation relations

X µ (τ, σ) = XLµ (τ + σ) + XR
µ
(τ − σ), X µ (τ, σ) = X µ (τ, σ + 2π) (13.1)
r r
µ 1 Lµ α′ µ α′ X 1 µ −inu
XL (u) = x0 + ᾱ0 u + i ᾱ e (13.2)
2 2 2 n n
n6=0
r r
µ 1 Rµ α′ µ α′ X 1 µ −inv
XR (v) = x0 + α0 v + i α e (13.3)
2 2 2 n n
n6=0

α′ µ
ᾱµ0 = αµ0 , αµ0 =
p , xLµ Rµ
0 = x0 = x0
µ
(13.4)
2
√
r
µ µ µ α′ X 1 −inτ µ inσ
′
X (τ, σ) = x0 + 2α α0 τ + i e (αn e + ᾱµn e−inσ ) (13.5)
2 n
n6=0

µ′
√ X
Ẋ µ + X µ′ = 2XL (τ + σ) = 2α′ ᾱµn e−in(τ +σ) (13.6)
n∈Z
µ′
√ X
Ẋ µ − X µ′ = 2XR (τ − σ) = 2α′ αµn e−in(τ −σ) (13.7)
n∈Z

[ᾱIm , ᾱJ−n ] = mδmn η IJ , [αIm , αJ−n ] = mδmn η IJ , [αIm , ᾱJn ] = 0 (13.8)

[āIm , āJ† IJ
n ] = δmn η , [aIm , aJn ] = δmn η IJ , [xI0 , pJ ] = iη IJ (13.9)
+ ′ + ′ + ′ + −
X = α p τ, ∂τ = α p ∂X + , H =αp p (13.10)

2. Closed string Virasoro operators

1
Ẋ − ± X −′ = (Ẋ − ± X −′ )2 (13.11)
2α′ p+
X X
(Ẋ − + X −′ )2 = 4α′ L̄⊥
ne
−in(τ +σ)
, (Ẋ − − X −′ )2 = 4α′ L⊥
ne
−in(τ −σ)
(13.12)
n∈Z n∈Z
√ 2 √ 2
2α′ ᾱ− ⊥
n = + (L̄n − 1), 2α′ α− ⊥
n = + (Ln − 1) (13.13)
p p
α′ I I α′ I I
L̄⊥
0 = p p + N̄ ⊥ , L⊥0 = p p + N ⊥ , L̄⊥ 0 = L0
⊥
(13.14)
4 4
∞
X ∞
X
N̄ ⊥ = nāI† I
n ān , N⊥ = naI† I
n an , N̄ ⊥ = N ⊥ (13.15)
n=1 n=1
1 2
(L⊥ + L̄⊥ ′ −
0 − 2) = α p , M 2 = −p2 = (N ⊥ + N̄ ⊥ − 2) (13.16)
p+ 0 α′

49
50 CHAPTER 13. RELATIVISTIC QUANTUM CLOSED STRINGS

3. Closed string states space

S
RIJ = ŜIJ + AIJ + S ′ δIJ , trŜIJ = 0, S′ = (13.17)
D−2
ŜIJ aI† J†  +
X  
Graviton fields: 1 ā1 p , pT (13.18)
I,J

AIJ aI† J†  +
X  
Kalb-Ramond fields: 1 ā1 p , pT (13.19)
I,J

S ′ aI† J†  +
 
Dilaton fields: 1 ā1 p , pT (13.20)

4. A brief look at superstring theories

G(10) ∼ g 2 (α′ )4 , G ∼ g 2 α′ , go2 ∼ g, g ∼ eφ (13.21)
ψ1I (τ, σ) = ΨI1 (τ − σ), ψ2I (τ, σ) = ΨI2 (τ + σ) (13.22)
ψ1I (τ, 0) = ψ2I (τ, 0), ψ1I (τ, π) = ±ψ2I (τ, π) (13.23)
I I
Ramond boundary condition: Ψ (τ, π) = +Ψ (τ, −π) (13.24)
Neveu-Schwarz boundary condition: ΨI (τ, π) = −ΨI (τ, −π) (13.25)
X
ΨI (τ, σ) = bIr e−ir(τ −σ) , bI−r bI−r = −bI−r bI−r = 0 (13.26)
r∈Z+1/2
9 Y
Y ∞ 9
Y Y
(αI−n )λn ,I (bJ−r )ρr,J |NSi ⊗ p+ , PT
 
NS sector: (13.27)
I=2 n=1 J=2 r= 1 , 3 ,···
2 2
X
I
dIn e−ir(τ −σ) , ground states: RA , A = 1, . . . , 16
 
Ψ (τ, σ) = (13.28)
n∈Z
9 Y
Y ∞ 9 Y
Y ∞
(αI−n )λn ,I (dJ−m )ρm,J RA ⊗ p+ , PT
   
R sector: (13.29)
I=2 n=1 J=2 m=1
1
NS Sector: α′ M 2 = − + N ⊥ , R Sector: α′ M 2 = N ⊥ (13.30)
2
bosonic states: b−1/2 |NSi ⊗ p , pT , fermionic states: |Ra1 i ⊗ p+ , pT
I
 +   
 (13.31)
NS-NS massless fields: gµν , Bµν , φ, bI−1/2 b̄J−1/2 |NSiL ⊗ |NSiR ⊗ p+ , pT
 
(13.32)
Type IIA, Type IIB, E8 × E8 heterotic, SO(32) heterotic, Type I (13.33)
 Quick Calculations
13.1 By the definition of L̄⊥ ⊥
m and Lm , we can obtain
r r
 ⊥ I 1 X  J J α′ IJ J α′ I
ᾱp ᾱm−p , xI0 = ᾱJ0 , xI0 ᾱJm = −i
  
L̄m , x0 = η ᾱm = −i ᾱ (13.34)
2 2 2 m
p∈Z
r r
 ⊥ I 1 X  J J α′ IJ J α′ I
αp αm−p , xI0 = αJ0 , xI0 αJm = −i
  
Lm , x0 = η αm = −i α (13.35)
2 2 2 m
p∈Z

13.2 First, we point out that the following relations hold

 ⊥ I
L̄m , ᾱn = −nᾱIm+n ,
 ⊥ I
Lm , αn = −nαIm+n , I
  ⊥ I
L⊥

m , ᾱn = L̄m , αn = 0 (13.36)
Then, using Eq. (13.5), we can prove that
r
I I α′ X 1 −inτ  ⊥ I  −inσ
L̄⊥ L̄⊥
   
0 , X (τ, σ) = 0 , x0 +i e L̄0 , ᾱn e
2 n
n6=0
r r
α′ I α′ X I −in(τ +σ) (13.37)
= −i ᾱ0 − i ᾱn e
2 2
n6=0
i
= − (Ẋ µ + X µ′ )
2
 51
r
I I α′ X 1 −inτ  ⊥ I  inσ
L⊥ L⊥
   
0 , X (τ, σ) = 0 , x0 +i e L0 , αn e
2 n
n6=0
r r
α′ I α′ X I −in(τ −σ) (13.38)
= −i α0 − i αn e
2 2
n6=0
i
= − (Ẋ µ − X µ′ )
2
13.3 The 16 ground states are listed as follows:

|Ra1 i : |0i , ξ2 ξ1 |0i , ξ3 ξ1 |0i , ξ4 ξ1 |0i , ξ3 ξ2 |0i , ξ4 ξ2 |0i , ξ4 ξ3 |0i , ξ4 ξ3 ξ2 ξ1 |0i (13.39)
|Ra2 i : ξ1 |0i , ξ2 |0i , ξ3 |0i , ξ4 |0i , ξ3 ξ2 ξ1 |0i , ξ4 ξ2 ξ1 |0i , ξ4 ξ3 ξ1 |0i , ξ4 ξ3 ξ2 |0i (13.40)

We can see that the eight ground states |Ra1 i have an even number of fermionic operators and
the other eight states |Ra2 i have an odd number of fermionic operators.

13.4 From Eq. (13.26), we can conlcude that all states in the turncated NS sector have half-integer
N ⊥ eigenvalues and integrally valued α′ M 2 .

13.5 The numbers of graviton, Kalb-Ramond, and dilaton states in ten dimensions are 35, 28, and 1.
Add these numbers up and we just get 64.

 Solutions to the Problems

13.1 (a)
(b) Using Eq. (13.7), we have
X X ′
[(Ẋ I − X I′ )(τ, σ), (Ẋ J − X J′ )(τ, σ ′ )] = 2α′ e−im(τ −σ) [αIm , αJn e−in(τ −σ )
m∈Z n∈Z
IJ im(σ−σ′ )
X
′
= 2α mη e (13.41)
m∈Z
d
= −4πα′ iη IJ δ(σ − σ ′ )
dσ
Then, it is obvious that the following holds
1 X im(σ−σ′ )
δ(σ − σ ′ ) = e (13.42)
2π
m∈Z

13.2

13.3 (a) Define f (σ0 ) = e−iP σ0 X I (τ, σ)eiP σ0 , then we have

df  ∂X I  d2 f  ∂2X I 
= e−iP σ0 eiP σ0 , 2 = e−iP σ0 eiP σ0 , ... (13.43)
dσ0 ∂σ0 dσ0 ∂σ02

By the Taylor’s theorem, we can obtain

∞ ∞
X σ0n  dn f  X σ0n  ∂ 2 X I 
f (σ0 ) = f (0) + n = 2 = X I (τ, σ + σ0 ) (13.44)
n=0
n! dσ 0 σ0 =0
n=0
n! ∂σ 0

(b) The relaion holds for the following reasons

∂ h −iP σ0 I i
e X (τ, σ)eiP σ0 = e−iP σ0 Ẋ I (τ, σ)eiP σ0 (13.45)
∂τ
∂ h −iP σ0 I i
e X (τ, σ)eiP σ0 = e−iP σ0 X I′ (τ, σ)eiP σ0 (13.46)
∂σ
(c) Using Eqs. (13.6) and (13.7), we can obtain
√ X √ X
2α′ e−iP σ0 ᾱIn eiP σ0 e−in(τ +σ) = 2α′ ᾱIn (τ, σ)eiP σ0 e−in(τ +σ+σ0 ) (13.47)
n∈Z n∈Z
52 CHAPTER 13. RELATIVISTIC QUANTUM CLOSED STRINGS
√ X √ X
2α′ e−iP σ0 αIn eiP σ0 e−in(τ −σ) = 2α′ αIn (τ, σ)eiP σ0 e−in(τ −σ−σ0 ) (13.48)
n∈Z n∈Z

e−iP σ0 ᾱIn eiP σ0 = ᾱIn e−inσ0 , e−iP σ0 αIn eiP σ0 = αIn einσ0 (13.49)
(d) Using the reults above, we have

e−iP σ0 |U i = (e−iP σ0 αI−m eiP σ0 )(eiP σ0 ᾱI−n eiP σ0 ) |U i = αI−m ᾱI−n e−i(m−n)σ0 |U i (13.50)

The invariance of |U i under σ0 transformation requires σ0 = 2kπ.

13.4 (a) Using Eq. (13.12), we can obtain
Z 2π XZ 2π
I I′ ′ inσ
dσ Ẋ X =α dσ e−inτ (L̄⊥
n −e
−inσ
− L⊥
ne ) = 2πα′ (L̄⊥ ⊥
0 − L0 ) (13.51)
0 n∈Z 0

13.5
13.6
13.7 (a) bi1 bi2 : 8 × 7 + 1 = 57. bi1 bi2 bi3 : 8 × 7 × 6 + 1 = 337. bi1 bi2 bi3 bi4 : 8 × 7 × 6 × 5 + 1 = 1681.
 Part II

DEVELOPMENTS

53
Chapter 14

D-branes and Gauge Fields

 Summary and Supplement

1. Quantizing open strings on Dp-branes

DD: X a (τ, σ)σ=0 = X a (τ, σ)σ=π = x̄a , a = p + 1, . . . , d

 
(14.1)
NN: X m′ (τ, σ)σ=0 = X m′ (τ, σ)σ=π = 0, m = 0, 1, . . . , p
 
(14.2)

(14.3)
(14.4)
(14.5)

2.
(14.6)
(14.7)

3.
(14.8)
(14.9)

 Quick Calculations
14.1
14.2
14.3
 Solutions to the Problems
14.1
14.2
14.3

55
56 CHAPTER 14. D-BRANES AND GAUGE FIELDS
Appendices

57
Appendix A

Elements of E8

A.1 Introduction
Source from: http://en.wikipedia.org/wiki/E8 (mathematics)
In mathematics, E8 is the name given to several closely related exceptional simple Lie groups and Lie
algebras of dimension 248; the same notation is sometimes used for their root lattice, which has rank 8.
The designation E8 comes from Killing and Cartan’s classification of the complex simple Lie algebras,
which fall into four infinite families labeled An , Bn , Cn , Dn , and five exceptional cases labeled E6 , E7 ,
E8 , F4 , and G2 . The E8 algebra is the largest and most complicated of these exceptional cases.
The E8 Lie group has applications in theoretical physics, in particular in string theory and supergravity.
The group E8 × E8 serves as the gauge group of one of the two types of heterotic string and is one of
two anomaly-free gauge groups that can be coupled to the N = 1 supergravity in 10 dimensions. E8
is the U-duality group of supergravity on an eight-torus (in its split form). One way to incorporate
the standard model of particle physics into heterotic string theory is the symmetry breaking of E8 to
its maximal subalgebra SU (3) × E6 . In 1982, Michael Freedman used the E8 lattice to construct an
example of a topological 4-manifold, the E8 manifold, which has no smooth structure. In February
2008, Garret Lisi published a particle physics theory based on the E8 Lie group.

59
60 APPENDIX A. ELEMENTS OF E8
Appendix B

Modular Forms

B.1 Introduction
Source from: http://en.wikipedia.org/wiki/Modular form
In mathematics, a modular form is a (complex) analytic function on the upper half-plane satisfying
a certain kind of functional equation and growth condition. The theory of modular forms therefore
belongs to complex analysis but the main importance of the theory has traditionally been in its
connections with number theory. Modular forms appear in other areas, such as algebraic topology and
string theory. A modular function is a modular form of weight 0: it is invariant under the modular
group, instead of transforming in a prescribed way, and is thus a function on the modular region (rather
than a section of a line bundle). Modular form theory is a special case of the more general theory of
automorphic forms, and therefore can now be seen as just the most concrete part of a rich theory of
discrete groups.

61
62 APPENDIX B. MODULAR FORMS
Bibliography

[1] B. Zwiebach, A First Course in String Theory. Cambridge University Press, 2004.
[2] M.B. Green, J.H. Schwarz, and E. Witten, Superstring Theory, Volume 1: Introduction. Cambridge
University Press, 1987.

63