Chapitre

5
Superstring Theory
***

1 Bosonic string theory problems . . . . . . . . . . . . . . . . . . . . . . . . . 53

2 Supersymmetric string or (superstrings) . . . . . . . . . . . . . . . . . . . . 53
3 Spectrum of open superstrings from quantization . . . . . . . . . . . . . . . 56
3.1 Open superstrings in NS-sector . . . . . . . . . . . . . . . . . . . . . 56
3.2 Open superstrings in R-sector . . . . . . . . . . . . . . . . . . . . . . 57
4 Spectrum of superstring models . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.1 Type IIA and Type IIB superstrings . . . . . . . . . . . . . . . . . . 58
4.2 Heterotic superstrings E8 × E8 and SO(32) . . . . . . . . . . . . . . . 59
4.3 Type I superstring . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5 Classification of superstrings . . . . . . . . . . . . . . . . . . . . . . . . . . 61

1 Bosonic string theory problems

Bosonic string theory suffers from many problems.
— It contains tachyons.
— It does not have fermionic states describing matter fields.
— It can not be used to unify particle physics and gravity.
— It does not involve non abelain symmetries.
To overcome such problems, one should add fermions ψ µ (τ, σ) on the world-sheet.

2 Supersymmetric string or (superstrings)

The action of this theory can be obtained by adding a fermionic term to the bosonic string
action. Indeed, one has

1 Z µ
S= 0 dσdτ (∂α X µ ∂ α Xµ + iψ ρα ∂α ψµ ) (5.1)
2πα Σ=W orldSheet

53
  
µ
ψ+
where ψ µ = ψAµ =  µ  are M ajorana spinors (real two-compent) with respect to
ψ−
world-sheet and vectors with respect to space-time. It is recalled that ρα = ραAB are
two-dimensional spin matrices satisfying the following anti-commutator relations

{ρα , ρβ } = 2η αβ .

Remarks

1. The formal supersymmetric transformations are illustrated by

X µ ⇐⇒ ψ µ .

2. The superstring action is invariant under

δX µ = χψ µ

δψ µ = −iρα ∂α X µ whereχ is a Majorana spinor.

Using similar methods used in the previous chapters, one can finds the equations of motion
given by
∂α ∂ α X µ = 0, ρα ∂α ψ µ = 0.

The boundary conditions of open superstring can be fixed by

Z
µ µ µ µ σ=π
δSF = dτ [ψ+ δψ+ − ψ− δψ− ]σ=0 = 0. (5.2)

The associated solutions read as

µ µ
ψ+ = ±ψ−
(5.3)
µ µ
δψ+ = ±δψ−
providing two kinds of fermions for open superstrings. The first one is given by

ψ+ (τ, 0) = ψ− (τ, 0)
(5.4)
ψ+ (τ, π) = ψ− (τ, π).

This sector is called Ramond, R . The second one defined

ψ+ (τ, 0) = −ψ− (τ, 0)

(5.5)
ψ+ (τ, π) = −ψ− (τ, π)

called Neveu-Schwarz (NS).

54
Remarks

1. The R-sector which is associated with the periodic fermions generates the following
models
1 +∞
Ψµ+ (τ, σ) dµn e−in(τ +σ)
X
= √
2 n=−∞
(5.6)
1 +∞
Ψµ− (τ, σ) dµn e−in(τ −σ) , n ∈ Z.
X
= √
2 n=−∞

2. The NS-sector which corresponds to the anti-periodic fermions provides the

following modes
1 +∞
Ψµ+ (τ, σ) bµr e−ir(τ +σ)
X
=√
2 r=−∞
(5.7)
1 +∞
Ψµ− (τ, σ) bµr e−ir(τ −σ) , r ∈ Z+ 1 .
X
=√
2 r=−∞ 2

µ µ
3. For closed superstrings, one should add d˜n and b˜r modes, needed to build the
associated spectrum.

It turns out that for closed strings, one can have four sectors
   
R R
 ⊗  (5.8)
NS NS

controlled by the following modes

R-R dµn ,d˜µn

R-NS dµ , b˜µr
n

NS-R µ
br , d˜µn
NS-NS bµr , b˜µr

It should be noted that calculations show that superstrings live in 10-dimensional

space-time
D = 10.

It is interesting to indicate other superstring theory models can be constructed by

combining a bosonic string in d = 26 and a fermionic string in d = 10. This is represented
by

Lef t Right
z }| { z }| {
   
Bosonic F ermionic
   
string string
   
   
 ⊗ 
in in
   
   
   
D = 26 D = 10

55
where one should use the following classical field content

XLµ , µ = 0, . . . , 25 ψRµ , XRµ , µ = 0, . . . , 9. (5.9)

To get a possible theory, one can should compactify the bosonic extra directions

XLµ µ = 0, . . . , 9 XRµ , µ = 0, . . . , 9 (5.10)

XLI I = 10, . . . , 25 ψRµ , µ = 0, . . . , 9. (5.11)

The combined theory is called hetorotic superstring. After compactifications, the action of
such superstrings takes the following form
Z
µ
S= dτ dσ(∂α X∂ α Xµ − 2iψ− ∂+ ψ− − 2iλA A
+ ∂λ+ ). (5.12)

In fact, one has two hetorotic superstrings :

— hetorotic superstring, with the SO(32) Lie symmetry
— hetorotic string, with the E8 × E8 Lie symmetry.

3 Spectrum of open superstrings from quantization

In this section, we would like to discuss the spectrum of superstrings. For open superstring,
one has two types of quantized relations.

3.1 Open superstrings in NS-sector

For the NS sector, one has

{bµr , bµs } = η µν δr+s,0 .

The associated right moving relations are

µ µ
{b˜r , b˜s } = η µν δr+s,0 .

Remarks

— bµ−r and bµ+r are the creation and annihilation modes, respectively.
— b̃µ−r and b̃µ+r are the creation and annihilation operators, respectively.
It has been shown that the spectrum of such sector is given by

αnµ |0 >= 0, bµr |0 >= 0 n, r > 0.

56
The general state reads as

(α−ni )kj (b−rj )kj |0 > .

Y Y
|ψ >= (5.13)
i j

It has been revealed that the mass term is given by

∞ +∞
1
M2 =
X X
α−n αn + rb−r br − . (5.14)
n=1 r= 1
2
2

The ground state of NS sector is a tachyon. However, the first excited level generated a
massless vector
Aµ ≡ bµ− 1 |0 > (5.15)
2

This sector behaves a bosonic one.

3.2 Open superstrings in R-sector

Concerning the Ramond sector, the commutation relations are given by

{dµm , dνn } = η µν δm+n,0

(5.16)
{d˜ν , d˜µn } = η µν δm+n,0 .
m

The associated mass term reads

∞ +∞
M2 =
X X
α−n αn + nd−n dn . (5.17)
n=1 n=1

For n, m = 0, we have
{dµ0 , dν0 } = η µν .

In this way, the ground state of R-sector is massless fermion

|0 > .

Non-massive fermions can be obtained by

(dµ0 )k |0 > .

In D = 10, the matrices dµ0 are 32 × 32 matrices.

Remarks
— The ground state of NS sector is a tachyon. One should remove it using the G.S.O
projection.
G.S.O|ψ >= |ψ > .

57
 — For NS sector, one has
P
1 b−r br
(−1)FN S = −(−1) r= 2
.

— For R-sector, one has

P
(−1)FR = Γ11 (−1) n
d−n d−n
.

4 Spectrum of superstring models

A close inspection shows that there is a nice classification of superstrings.

1. Models with N = 2 supersymmetry

(a) Type II A

(b) Type II B

2. Models with N = 1 supersymmetry

(a) Heterotic, SO(32)

(b) Heterotic, E8 × E8

(c) Type I, SO(32).

4.1 Type IIA and Type IIB superstrings

To get the spectrum of such models, it is useful to recall a formal definition of superalgebra

BB = B, BF = F, F B = F, F F = B. (5.18)

This gives the following statements

— NS-NS, being is a bosonic sector
— R-R, being is a bosonic sector
— NS-R, being a fermionic sector
— R-NS, being a fermionic sector.
The states sectors can be built from the modes

α, α̃, b, b̃ d, ˜
d. (5.19)

Remarks

1. For supersymmetric theories, one can deal with only the bosonic sector.

2. For the bosonic sector NS-NS, one can have the spectrum

8v ⊗ 8v = 1 ⊕ 35 ⊕ 28, (5.20)

58
 for both theories IIA and IIB. One can interpret the above decomposition as

1→φ (5.21)
35 → gµν (5.22)
28 → Bµν . (5.23)

For R-R-sector, one has two possibilities

8s ⊗ 8c → 8 ⊕ 56 → Aµ , C µνσ , (odd orders)

(5.24)
8s ⊗ 8s → 1 ⊕ 28 ⊕ 35, → χ, B̃, Dµνρσ , (even orders).

4.2 Heterotic superstrings E8 × E8 and SO(32)

It is recalled that heterotic superstring can be constructed from a hybrid of a bosonic
string and a fermionic one. In this way, the symmetry group of the bosonic sector can be
composed
SO(1, 25) −→ SO(1, 9) × SO(16)
26 −→ (10, 1) ⊕ (1, 16) (5.25)
PL26 −→ PL10 ⊕ PL16

where PL10 is identified with PR . However, 16 of PL16 can be interpreted as the generator of
Cartan super Lie algebra of Lie algebra. Calculations show that these Lie algebras can be
identified with so(32) and E8 × E8 . It is recalled that

rank(so(32)) = rank(E8 × E8 ) = 16.

The mass of states is given by

1 2
M2 = P + NL − 1 = NR − aR
2 A

where aR = 0, 1/2 for R and NS sectors, respectively. For right moving the massless
states of such superstrings contain
V ⊕ S

where V is a vector of SO(1, 9) and S is a spinor of SO(10) with a fixed chirality.

For left moving massless states, we have two solutions
1. NL = 1 and PA2 = −m2 = 0
The massless states are given by

A
α−1 |0 > A = 1, ..., 16

identified with U (1)16

2. NL = 0 and P 2 = 2

59
 We have the following following sum

V ⊕ adjG

where G is E8 × E8 or SO(32). The closed superstring spectrum can be obtained by

(V ⊕ adjG) ⊗ (V ⊕ S)

The bosonic states are given by

(V ⊕ adjG) ⊗ V

providing Bµν , gµν , Φ and Aµ in the adjoint representation of G

Aµ = Aaµ Ta , a = 1, . . . , dim G.

The fermionic sector is given by

(V ⊕ adjG) ⊗ S (5.26)

4.3 Type I superstring

This model (theory) can be obtained from Type IIB by using a symmetry Ω acting on the
word sheet as follows
Ω : σ −→ 2π − σ,

Ω2 = 1(5.27)This symmetry acts also

XLµ → XRµ (5.28)

Ωαnµ Ω−1 → α̃nµ . (5.29)

For the NS-NS sector, this symmetry acts as follows.

µ µ
Ωα−n ν
α̃−n Ω−1 = α̃−n ν
α−n (5.30)

leaving only symmetric fields

gµν , Φ (5.31)

For the R-R sector, it leaves only anti-symmetric fields

B̃µν .

Certain calculations show that the gauge symmetry of this theory coincides with SO(32).

60
5 Classification of superstrings
The spectrum of superstings contains dilaton φ, where gs = eφ is the coupling constant, the
graviton gµν (spin two), antisymmetric tensor Bµν e and gauge tensors Aµ1 ...µp+1 . One can
write

(gµν , φ, Bµν ) ⊕ Aµ1 ...µp+1 . (5.32)

The fields Aµ1 ...µp+1 generalizing the vector potential Aµ to tensors with p + 1 ((p + 1)-forms,
p = 1, 2, ...) can couple to p-branes. One can give the following classifications (5.1).

Superstring Type IIB Type IIA Heterotic Heterotic Type I

E8 × E8 SO(32)
String Type closed closed closed closed open
(and closed)
Supersymmmetry N =2 N =2 N =1 N =1 N =1
of space-time chiral non-chiral
Lie Symmetry - - E8 × E8 SO(32) SO(32)
D-branes −1, 1, 3, 5, 7 0, 2, 4, 6 - - 1, 5, 9

Table 5.1 – Classification of strings and branes.

61
62
 Chapitre
6
String Theory Revolutions
***

1 From zero to one dimensional objects . . . . . . . . . . . . . . . . . . . . . . 63

1.1 Particle physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
1.2 Lie algebras in particle physics . . . . . . . . . . . . . . . . . . . . . . 64
2 String theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1 Classical and quantum version of string theory . . . . . . . . . . . . . 64
2.2 Quantum string theory . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3 String theory revolutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.1 First revolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.2 Second revolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.2.1 Partial solutions . . . . . . . . . . . . . . . . . . . . . . . . 66
4 Compactification of higher dimensional theories . . . . . . . . . . . . . . . . 66
4.1 Toroidal compactification of superstring theory . . . . . . . . . . . . 66
4.2 Connection with particle physics and Calabi-Yau manifolds . . . . . 66
4.3 Compactification of M-theory . . . . . . . . . . . . . . . . . . . . . . 67
5 Building models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

1 From zero to one dimensional objects

1.1 Particle physics

Particle physics concerns the study of particles and fundamental interactions. It has been
shown that there is a nice interplay between such a physics and Lie algebra structures. It is
interesting to note this bridge
• Strong interaction : su(3) Lie algebra
• Weak interaction : su(2) Lie algebra
• Electromagnetic interaction : u(1) Lie algebra
• Gravitation theory : Geometric deformation of the univers
Other Lie symmetries have been used in gauge theories

63
 • Grand Unified theory (GUT) : E6 , so(10), su(5) Lie algebras
• An , Dn , E6,7,8 gauge theories.

1.2 Lie algebras in particle physics

To apply Lie alegebras in physics, one should understand
• Lie structure
• Representation theory
• Coexter and Dynkin diagrams
• Cartan matrices.

Remarks
— Particle physics data are hidden in Lie algebra structures.
— The main problem is in the gravity theory.
— More details on such problems can be found in
— M. Green, J. Schwarz and E. Witten, Superstring Theory, 2 Vols., Cambridge
U.P.1987
— A.B, Introduction to String Theory, arXiv :0808.2957.

2 String theory
The motion of particles (zero dimensional objects) could be extended to the motion of
higher dimensional objects : strings and branes
Two theories have been dealt with :
• Open string theory
• Closed string theory

2.1 Classical and quantum version of string theory

The action of the classical theory reads as

S = Sb + Sf . (6.1)

This generates the following bosonic equation of motion

(∂τ2 − ∂σ2 )X µ (τ, σ) = 0. (6.2)

2.2 Quantum string theory

The quantization gives massless states

(gµν , Bµν , Aµ , φ, ......)

64
Graviton, photons, and others particles could take places in such a theory. It could be
considered as a possible unified theory of
• Strong interaction,
• Weak interaction,
• Electromagnetic interaction,
• Gravitation interaction.
Two configurations have been investigated to provide non-trivial results

• Open string theory : (D-branes) → Gauge theories

• Closed string theory : → Gravity theory.

3 String theory revolutions

3.1 First revolution

This concerns the classification of superstring models. Here, we give the main results.
• Models beyond particle physics
• 5 models

• Type IIA

• Type IIB

• Heterotic superstring theory with SO(32) = D16 gauge symmetry

• Heterotic superstring theory with E8 × E8 gauge group

• Type I with SO(32) = D16 . gauge symmetry.

We give some problems. Among others, one has

• Dimension of the space-time : D = 10
• Gauge symmetries (Lie algebras) :

• E8 × E8

• SO(32) = D16 .

3.2 Second revolution

This concerns the superstring problems and solutions. Many problems appear in
superstring models

1. Space-time dimension : D = 10 > 4

2. Five models

3. Large gauge symmetries : SO(32) = D16 , E8 × E8

4. ........

65
3.2.1 Partial solutions

Partial solutions have been reported in many places

• Compactification
D = 10 → D = 4

• String dualities

Connection between string models in lower dimensions.

• New theories
1. M-theory in D = 11, Witten, 1995
2. F-theory in D = 12, Vafa, 1996.

4 Compactification of higher dimensional theories

Four dimensional models can be obtained using the following scenario

R1,D−1 → R1,3 × X D−4

where X D−4 is a (D − 4) dimensional compact manifold (finite size). One has

1. String theory lives in D = 10
10 = 4 + 6

2. M-theory lives in D = 11
11 = 4 + 7

3. F-theory lives in D = 12
12 = 4 + 8

4.1 Toroidal compactification of superstring theory

Here, one considers
R1,9 → R1,3 × X 6

where X 6 = 6 is a 6-dimensional compact manifold given by

X 6 = T 2 × T 2 × T 2.

4.2 Connection with particle physics and Calabi-Yau manifolds

Calabi-Yau compactifications could be needed for establishing contact with particle
physics. In this way, one has
R1,9 → R1,3 × X 6 (6.3)

66
where X 6 is a 3-dimensional Calabi-Yau manifold. The Calabi-Yau geometries are defined
by

1. Kahler (complex) manifolds

2. SU(3) holonomy group

3. Ω = dz1 dz2 dz3 .

There are many constructions of Calabi-Yau manifolds. One has

1. Orbifolds : T 6 /G, ⊂ SU (3)

2. Hypersurfaces in CP 4 (z1 , z2 , z3 , z4 , z5 ) projective pace

P5 (z1 , z2 , z3 , z4, z5 ) = 0 (6.4)

3. Hypersurfaces in toric varieties, using toric geometry method.

4.3 Compactification of M-theory

The bosonic sector of the corresponding supergravity effective theory at low energies is
given by Z √ 
1

1Z
16πG11 S = d11 x −h R − |F4 |2 − A3 ∧ F4 ∧ F4 , (6.5)
2 6
where h is the space-time metric determinant and R represents the scalar curvature. The
field strength F4 = dA3 originates from the gauge potential 3-form A3 and G11 is the eleven
dimensional gravitational constant. A close inspection reveals that M-theory involves two
brane solutions, the M2 and the M5-branes. The associated compactifications read as

R1,10 → R1,3 × X 7 (6.6)

X 7 where are 7-dimensional compact manifolds. Possible models are

1. T 7 , ...

2. G2 manifolds (Standard model)

It is recalled that G2 manifolds are

1. 7-dimensional real manifolds

2. G2 holonomy group

3. Special 3-form
w = fijk dwi dxj dxk (6.7)

This form is linked to the octonian algebra given by

ti tj = −δij + fijk tk (6.8)

67
5 Building models
Semi-realistic models in four dimensions have been built using
• 1996 : Geometric engineering method : Vafa group
1. Singularities of Calabi-Yau Manifolds
2. Lie algebras
3. D-brane physics
• 2001 : Intersecting branes : Madrid group
1. Intersection theory
2. D-brane physics
• 2008 : New method based on F-theory : Vafa group
1. F-theory
2. Varieties with elliptic fibrations
More details on compact manifolds can be found in the following reference : D. Joyce,
Compact Manifolds with Special Holonomy, Oxford U.P. 2000.

6 Conclusion
• There are many things...............
• String theory needs : Third revolution.

68