Chapter 2

Conformal Field Theory

Following “String Theory” by J. Polchinski, Vol.I.

Notes written by students (work still in progress).
For more information contact George Siopsis
[email protected]

2.1 Massless scalars in two dimensions

We will start by looking at the Polyakov action with one change. The metric has been replaced with a Euclidean
metric δa,b with signature (+,+).
Z
T
S= dτ dσ(Ẋ µ Ẋµ + X 0µ Xµ0 ) (2.1)
2
where T, Ẋ µ , X 0µ represent the string tension, ∂τ X µ , ∂σ X µ respectively.
We can derive the equation of motion by varying the action with respect to the coordinate X. We find:

δX S = 0 ⇒ Xµ00 + Ẍ µ = ∇2 X µ = 0, where ∇2 = ∂σ21 + ∂σ22

We can define z and z̄ as linear combinations of σ1 and σ2 . These will represent the new worldsheet coordinates.

z = σ1 + iσ2 , z̄ = σ1 − iσ2 (2.2)

The bar denotes complex conjugate. We can also invert the coordinate transformation.
z + z̄ z − z̄
σ1 = , σ2 =
2 2
Define differentiation:
1 1
∂z = ∂ =
(∂1 + i∂2 ) and ∂z̄ = ∂¯ = (∂1 − i∂2 ) (2.3)
2 2
¯ and the volume element is given by
∇2 can be written as 4∂ ∂,

1 −1
d2 z = dσdτ = 2dσdτ. (2.4)
1 1

Also define
Z
d2 z δ 2 (z, z̄) = 1 (2.5)

1
2 CHAPTER 2. CONFORMAL FIELD THEORY

so that δ(σ 1 )δ(σ 2 ) = δ 2 (z, z̄). The action in complex coordinates is

Z
¯ µ.
S = T d2 z∂X µ ∂X (2.6)

After varying the action with respect to the coordinate Xµ , we get the equation of motion.

¯ µ=0
∂ ∂X (2.7)
¯
This implies ∂X is a holomorphic function, ie a function of z. Also, ∂X is an antiholomorphic function, ie a
function of z̄.
Another useful result is the divergence theorem for complex coordinates. First, let’s look at the divergence
theorem for three dimensions ie, electrostatics. The divergence theorem states that for any well behaved vector
field E(x) defined within a volume V surrounded by the closed surface S the relation

dS

Z I
d3 x ∇ · E = E · n da
V S

holds between the volume integral of the divergence of E and the surface integral of the outwardly directly
normal component of E.
In 2D:
Z I
¯ z̄ ) = i
d2 z (∂E z + ∂E (E z dz − E z̄ dz̄) where n̂ = (−dz̄, dz). (2.8)
∂R
We can now write the mode expansions in terms of the complex coordinates.
r
µ µ z − z̄ pµ α0 X 1 µ 2nπiz −2nπiz̄
X (z, z̄) = x − i +
+i {αn e l + α̃nµ e l } (2.9)
2 p 2 n
n6=0

As one can see, this mode expansion can be broken into two pieces (left and right handed).

X µ (z, z̄) = XLµ (z) + XR

µ
(z̄)

We see the left handed piece cooresponds to a holomorphic function and the right handed piece to a antiholo-
morphic function. We will never look at the mode expansion itself. Instead we will always look at various
derivatives of the mode expansion.
i pµ π√ 0 X µ inπz
∂XLµ = + − 2α αn exp[ ], (2.10)
2 p+ l l
n6=0
µ √
¯ µ = i p + − π 2α0 inπz̄
X
∂XR +
α̃nµ exp[ ] (2.11)
2p l l
n6=0

We can absorb the pµ by defining α0µ a certain way.

l pµ
Let α0µ = −i √
2π 2α0 p+
2.2. SOLUTION TO THE BOUNDARY-VALUE PROBLEM WITH GREEN FUNCTION 3

Now the derivatives on the fields simplify.

π√ 0 X µ inπz
∂XLµ = − 2α αn exp[ ], (2.12)
l l
all n

√
¯ µ = − π 2α0 inπz̄
X
∂XR α̃nµ exp[ ] (2.13)
l l
all n

2.2 Solution to the Boundary-Value Problem with Green function

The solution to the Poisson or Laplace equation in a finite volume V with either Dirichlet or Neumann boundary
conditions on the bounding surface S can be obtained by means of Greens theorem. In general, we want to solve
the equation,

∇02 G(x, x0 ) = −4πδ(x − x0 ) (2.14)

where G(x,x’) is the potential and the delta function is a point source. The solution for G is given as

1
G(x, x0 ) = + F(x, x0 ) (2.15)
|r − r0 |

with F satisfying the Laplace equation inside the volume V:

∇02 F (x, x0 ) = 0

In two dimensions the Poisson equation is given by

∇2 G(z, 0; z̄, 0) = −2πδ(z)δ(z̄) where ∇2 = ∂ ∂¯

= −2πδ 2 (z, z̄) (2.16)

where

G(z, z̄) = ln |z|2 = ln |z| + ln |z̄| (2.17)

This is just the solution for the potential to a line charge in two dimensions. We can prove that G is a solution
of the Poisson equation in problem 1.

¯ = ∂ ∂¯ ln |z|2 = ∂ 1 ¯1
∂ ∂G + ∂ = 2πδ 2 (z, z̄), z=0 (2.18)
z̄ z
Now that we see G is directly related to the potential, we can take the gradient to get the electric field.

Ez = ∂G, ¯
Ez̄ = ∂G

Z I
3 1
d x∇·E = E · dn̂ = 2πR( ) = 2π (2.19)
V S R
4 CHAPTER 2. CONFORMAL FIELD THEORY

2.3 Amplitudes

h0|X µ1 (z1 )X µ2 (z2 ) · · · X µn (zn )|0i

Simplest amplitude:
q P2 point function.
0
X µ = XLµ = i α2 1 µ
n6=0 n αn exp[
−2πinz
l ].

Aµν = h0|X µ (z)X ν (z 0 )|0i (2.20)

α0 X 1 −2πi(n − m)z
= exp[ ]h0|αnµ αm
ν
|0i
2 nm l
n,m6=0

We can write the expectation value as a commutator plus a term which kills the vacuum.

[αnµ , αm
ν
] = nη µν δm+m,0

α0 X 1 2πin(z − z 0 )
Aµν = exp[− ] (2.21)
2 n l
n6=0
P ωn
This is a sum of the form n , but only converges for |ω| < 1.

2πi(z − z 0 ) 2π(z − z 0 )
|exp[− ]| = |exp[ ]| < 1 ⇒ z2 − z20 < 0, z2 < z20
l l
So we see the Xs must be time ordered to give a correct result i.e.,

h0|XX|0i ⇒ h0|T [XX]|0i, T − time ordering.

We define time ordering as

X µ (z)X ν (z 0 ) for z2 < z20


T [X µ (z)X ν (z 0 )] = , (2.22)
X ν (z)X µ (z 0 ) for z2 > z20
or in terms of the step function.

T [X µ (z)X ν (z 0 )] = θ(z20 − z2 )X µ (z)X ν (z 0 ) + θ(z2 − z20 )X ν (z)X µ (z 0 )

Evaluating the sum:

α0 X en α0 ¯ µν ∼ δ(z − z 0 ) → Aµν is a Green fn.

Aµν = = ln|1 − e−β | : ∂ ∂A (2.23)
2 n n 2
If we apply the D’Alembertian to A, we see A is a two-point Green function.

¯ [X µ (z)X ν (z 0 )]
∂ ∂T = (∂ 2 + ∂¯2 )T [X µ (z)X ν (z 0 )]
= T [∂12 X µ X ν ] + T [∂1 (−δ(z2 − z20 )X µ X ν + δ(z2 − z20 )X µ X ν )]
= T [∂12 X µ X ν ] + T [∂2 δ(z2 − z20 )[X µ , X ν ]] + T [∂2 X µ X ν )]
= T [∂12 X µ X ν ] + δ(z20 − z2 )[∂2 X µ , X ν ] + T [∂22 X µ , X ν ]
= T [(∂12 X µ + ∂22 X µ )X ν ] + δ(z20 − z2 )[∂2 X µ , X ν ]
= δ(z20 − z2 )[∂2 X µ , X ν ]
= πα0 δ(z − z 0 , z̄ − z̄ 0 )η µν (2.24)
2.3. AMPLITUDES 5

A must be of the form of a Green function, Aµν = η µν G(z, z 0 ).

For future reference X will imply only the holomorphic piece, unless specified otherwise.
Write the time-ordered product in terms of the normal ordered product minus the singularity.

α0
T [X µ X ν ] =: X µ X ν : − ln|z − z 0 |η µν (2.25)
2
As z → z 0 :

Im z

. z’

Re z

∞
α0 µν X (z − z 0 )k
X µ (z)X ν (z 0 ) = − η ln |z − z 0 | + : X ν ∂k X µ (z 0 ) : (2.26)
2 k!
k>0
0
α0
X 1 Z Z
1α δ δ k δ δ
(− dzdz 0 η µν ln |z − z 0 | µ ν
) = exp[− dzdz 0 η µν ln |z − z 0 | ] (2.27)
k! 2 2 δX δX 4 δX δX ν
µ
k

Define an operator Ô such that Ô = X µ X ν .

α0
Z
δ δ
: Ô := exp[− dzdz 0 η µν ln |z − z 0 | ]Ô (2.28)
4 δX δX ν
µ

This needs to be inverted.

α0
Z
δ δ
Ô = exp[ dzdz 0 η µν ln |z − z 0 | ] : Ô : (2.29)
4 δX µ δX ν
Now Ô should have no singularities. We have to define the product between two Ôs. This product will have
singularities, unless the product of the two is normal ordered.
:Ô1 ::Ô2 : ← has singularities. :Ô1 Ô2 : ← has no singularities.

α0
Z
δ δ
: Ô1 [X]Ô2 [Y ] := exp[− dzdz 0 η µν ln |z − z 0 | ] : Ô1 :: Ô2 : (2.30)
2 δX µ δX ν
invert

α0
Z
δ δ
: Ô1 [X] :: Ô2 [Y ] := exp[ dzdz 0 η µν ln |z − z 0 | ] : Ô1 Ô2 : (2.31)
2 δX µ δY ν
Example:

Let Ô1 = Ô2 = ∂X µ ∂Xµ = T

6 CHAPTER 2. CONFORMAL FIELD THEORY

: T (z) :: T (z 0 ) :=: ∂X µ ∂Xµ (z) :: ∂ 0 X ν ∂ 0 Xν :

There are four possible single contractions, and two possible double contractions.

2
α0 µν α0 µν
= : T (z)T (z 0 ) : − η ln |z − z 0 | ∗ 4 : ∂Xµ ∂ 0 Xν : + η ηµν (∂∂ 0 ln |z − z 0 |)2 ∗ 2
2 2
2
α0 d 2α0 α0
: T (z) :: T (z 0 ) : = 0
− 0
:T :− : ∂ 0 T (z 0 ) : + : T (z)T (z 0 ) : (2.32)
2 (z − z )4 (z − z ) 2 z − z0

Example:

0
δ
Let Ô1 =: eik1 X(z) : Ô2 =: eik2 X(z ) : δX µ Ô1 = ik1µ Ô1

α0
: Ô1 Ô2 : = exp[ ln |z − z 0 |η µν (ik1µ )(ik2ν )] : Ô1 :: Ô2 :
2
α0
= exp[− ln |z − z 0 |k1 · k2 ] : Ô1 :: Ô2 :
2
α0
= (z − z 0 )− 2 k1 ·k2
: Ô1 :: Ô2 : (2.33)

α0
⇒: Ô1 :: Ô2 : = (z − z 0 ) 2 k1 ·k2
: Ô1 Ô2 :
0
− α2
= (z − z 0 ) k1 ·k2
: ei(k1 +k2 )·X (1 + Ô(z − z 0 )) : (2.34)

2.4 Noether’s Theorem

For every symmetry in a theory, there must be some conserved current, ie

∂µ j µ = 0 ⇒ Symmetry(S). (2.35)

We can integrate over the zeroth component of the conserved current to get the charge.
Z Z Z Z
dQ
Q= 3
d x∂0 j → 0
= 3
d xj = 0
d x∇ · ~j =
3
d~s · ~j = 0 (2.36)
dt R

Q generates transformations.

δA = i[Q, A] = i(QA − AQ)  << 1

Example: Let Q=H

δA = i[H, A] = Ȧ

or Q=~
p
~ = i[~
∇A p, A]

Look at A(t0 ):
2.4. NOETHER’S THEOREM 7

t
S+
t+

t0

S−
t−

Region R is bounded by [S+ , S− ].

δA = i(Q(t+ )A(t0 ) − A(t0 )Q(t− ))

Z Z
3 0
= i{ d x j A(t0 ) − d3 x j 0 A(t0 )}
S+ S−
Z
= i dSµ j µ A(t0 )
∂R
Z
= i dSµ T [j µ A(t0 )]
Z∂R
δA = i d4 x ∂µ T [j µ A(t0 )] ”Ward Identity” (2.37)
R

Show the Ward Identity explicitly in two dimensions:

Z
i
δA = d2 z ∂a T [j a A(z0 )], j = (jz , jz̄ )
2π R
I
i
= (dzjz − dz̄jz̄ )A(z0 ) (2.38)
2π ∂R
∂a j a = ∂z̄ jz + ∂z jz̄ = 0 for special case jz is holomorphic, and jz̄ is antiholomorphic.
I
dz λ(z0 )
jz A(z0 ) = λ(z0 ) : jz A(z0 ) = . . . + + ... (2.39)
2π z − z0

δA = λ + λ̄ = −(λ − λ̄) (2.40)

Example:

Define the transformation on X and the current density.

i i
X µ → xµ + aµ : jaµ = ∂a Aµ jzµ = 0 ∂X µ
α0 α

¯ µ=0
∂j remember X is holomorphic
z

Let A(z0 ) =: eik·X :

8 CHAPTER 2. CONFORMAL FIELD THEORY

δA = : eik·(X+a) :
= (1 + ik · a)A(z0 ) (2.41)

i α0 µν
jzµ A(z0 ) = η ∂ ln |z − z0 |(ikν ) : A(z0 ) : +regular terms
α0 2
1 kµ
= − A(z0 )
2 z − z0
1
⇒ λ = k µ A(z0 ) = −λ̄ : Residue (2.42)
2

(λ + λ̄) = (k µ A(z0 ))

aµ (λ − λ̄) = aµ kµ A(z0 )

Example:

δz = −, δX µ = X µ (z − ) − X µ (z)
= −∂Xµ (2.43)

Noether Current
Z
1 ¯ µ
S=− ∂X µ ∂X
2πα0

Z
1 ¯ µ + ∂X µ ∂(∂X
¯
δz = −(z, z̄) δS = ∂(∂X µ )∂X µ)
2πα0
Z
1 ¯ µ
= ∂(∂X ∂Xµ ) (2.44)
πα0

2.5 Conformal invariance

1 ¯ =0
T = : ∂X µ ∂Xµ : ∂T conserved, T is holomorphic (2.45)
πα0
1 ¯ µ ∂X
¯ µ:
T̄ = : ∂X ∂ T̄ = 0 T̄ is antiholomorphic (2.46)
πα0

1
Tτ τ = − : Ẋ 2 + x02 := Tσσ
2α0

1
Tτ σ = Tστ = − Ẋ · X 0 : Traceless Taa = 0.
α0
For an arbitrary function v(z):
2.5. CONFORMAL INVARIANCE 9

j(z) = iv(z)T (z) ¯ = 0 ⇒ infinite number of conservation laws.

∂j

From j = iv(z)T (z), ∂ a ja = 0

¯ = 0 or ∂ j̄ = 0
∂j
¯
∂(v(z)T (z)) = 0
v(z)T(z) is a conserved current, where
v∂xµ
vT xµ ∼ .
z − z0
Let z transform as δz = z + εV , then
xµ −→ xµ − εv∂xµ
and T (z)A(z 0 ) can be expanded as Laurent series
a−1 a−2
T (z)A(z 0 ) = + + ... + regular terms, (2.47)
z − z0 (z − z 0 )2
then λ is

I
dz
λ = iv(z)T (z)A(z 0 )
2π
I
dz a−1 v(z) a−2 v(z)
= i[ 0
+ + ..]
2π z − z (z − z 0 )2
i
= ia−1 v(z 0 ) + ia−2 ∂v(z 0 ) + a−3 ∂ 2 v(z 0 ) + · · · . ie. (2.48)
2!
ε
δA = −εa−1 v − εa−2 ∂v − a−3 ∂ 2 v − · · · ,
2!
∞
X 1
= −ε a−n−1 ∂ n v. (2.49)
n=0
n!

An example for v(z)=z, z → z + δz = (1 + ε)z, then δA = −hεA, ie.

A0 = (1 + ε)−h A
= (ζ)−h A. (2.50)

or A(z, z̃) transforms as

A(z, z̄) −→ ζ −h ζ̄ −h̃ A.
For ζ = reiθ ,

A −→ r−(h+h̃) e−(h−h̃)θ A

h + h̃ represents the magnitude and h − h̃ represents the spin of A. If A is order h, then ∂A is order h+1, ie.

∂A ∂z 0 ∂A
=
∂z ∂z ∂z 0
→ (1 − εz)(1 − hεz)∂A
→ (1 − (h + 1)εz)∂A.
10 CHAPTER 2. CONFORMAL FIELD THEORY

or
¯
∂A, h̃ → h̃ + 1
Compare the coefficient of ∂v with the equations for δA. This implies a−2 will be hA. For simplicity, let
v(z) = 1. Do a translation transformation on z, z → z + ε. Then δA = A(z − ε) − A(z) = ε∂A −→ a−1 = ∂A.
For an arbitary v(z), z → z + εv(z), δA is given by
δA = −hε∂vA
∂ζ −h ∂ ζ̄ −h̃
A0 = ( ∂z ) ( ∂ z̄ ) A
A0 = (1 + ε∂v)−h A.
If δA has only two singularity terms in the form below, call A a primary field, ie.
∂A h∂A
T (z)A(z 0 ) = 0
+ . (2.51)
z−z (z − z 0 )2
Some examples for T (z)A(z 0 )

1 1
T (z)X µ (z 0 ) = − : ∂X ν ∂Xν : X µ (z 0 ) = η µν ∂ ln(z − z 0 )∂Xν
α0 2
1 µ
= ∂X , ie. h = h̃ = 0 (2.52)
z − z0
1 µν 2
T (z)∂ 2 X µ = η ∂∂ ln(z − z 0 )∂Xν
2
2
= ∂X µ (z)
(z − z 0 )3
2 1
= 0 3
[∂X µ (z 0 ) + (z − z 0 )∂ 2 X(z 0 ) + (z − z 0 )2 ∂ 3 X µ (z 0 ) + ...
(z − z ) 2!
2 2 1
= ∂X µ + ∂ 2 X µ (z 0 ) + ∂∂ 2 X µ + ... (2.53)
(z − z 0 )3 (z − z 0 )2 z − z0
1 α0 µν
T (z) : eikX : = ( η kν ∂ ln(z − z 0 ))2 : eik·X : +η µν kν ∂ ln(z − z 0 ) : ∂Xµ eik·X :
α0 2
α0 k2 1
= 0
: eik·X : + : ∂eik·X (2.54)
4 (z − z ) 2 z − z0
0 2 0 2
For A = eik·X , T (z)A(z 0 ) implies h = α 4k and T̄ (z̄)A(z̄ 0 ) implies h̃ = α 4k . Therefore A has weight (h, h̃) =
0 2 0 2
( α 4k , α 4k ). If a translation is applied to z (z → z + εv), then eik·X(z) → eik·X(z−εv) and δ(eik·X(z) ) =
−εv∂eik·X(z) , which means that h=0. We have just shown that h is not equal to zero, therefore this result must
come from a quantum effect, where
0 2
∂Xeik·X −→ h = 1 + α 4k
0 2
∂ 2 Xeik·X −→ h = 2 + α 4k
0 2
∂ m Xeik·X −→ h = m + α 4k
α0 k 2
∂ mn X µn ...∂ m2 X µ2 ∂ m1 X µ1 eik·X −→ h = mn + ... + m2 + m1 + 4

Xµ h=0 h̃ = 0 (0, 0)
∂X µ h=1 h̃ = 0 (1, 0)
¯ µ
∂X h=0 h̃ = 1 (0, 1)
∂2X µ h=2 h̃ = 0 (2, 0)
0 2 0 2
α0 k 2 α0 k 2
eik·X h = α 4k h̃ = α 4k ( 4 , 4 )
0 2 0 2 0 2 0 2
∂Xeik·X h = 1 + α 4k h̃ = 1 + α 4k (1 + α 4k , 1 + α 4k )
.. .. .. ..
. . . .
2.6. FREE CFTS 11

2.6 Free CFTs

2.6.1 Linear dilaton
The TT OPE is given by:
D 2 2
T (z)T (z 0 ) = + T (z 0 ) + 0 : ∂ 2 Xµ ∂X µ :
2(z − z 0 )4 (z − z 0 )2 α (z − z 0 )

where ∂ 2 Xµ ∂X µ can be written as 12 (∂(∂Xµ ∂X µ ))

If T (z) = − α10 : ∂X µ ∂Xµ : +Vµ ∂ 2 X µ , the TT OPE is given by

D α0
T (z)T (z 0 ) ∼ 0 4
+ Vµ V µ ∂ 2 ∂ 02 ln(z − z 0 ) + ...
2(z − z ) 2
D 6Vµ V µ α0
= + + ...
2(z − z 0 )4 2(z − z 0 )4
c
= + ... , (2.55)
2(z − z 0 )4

where c = D + 6α0 Vµ V µ .

Therefore the central charge for the Linear Dilaton theory can be any number. When the number of dimensions
is one or two the theory is exactly solvable.

α0 2
T (z)X µ (z 0 ) ∼ ∂ ln(z − z 0 )∂X µ + V µ ∂ ln(z − z 0 ) + ...
2
1 µ V µ α0
= ∂X + , h = 0 and Xµ is not a primary field, (2.56)
z − z0 2(z − z 0 )2
V µ α0 v
v(z)T (z)X µ (z 0 ) ∼ [v(z 0 ) + (z − z 0 )∂v + ...] + ∂X µ + .... (2.57)
2(z − z 0 )2 z − z0

For δX µ = −ε, λ = 12 V µ α0 ∂v + v∂X µ .

2.6.2 bc theory
Let b and c be anticommuting fields, ie spinors. The action can be writen as
Z
1 ¯
S= d2 zb∂c (2.58)
2π
The equations of motion are given by: ∂c¯ = 0, ∂b
¯ = 0, where b and c are holomorphic. If we let l = 2π, then
we can write b and c as
X X
b(z) = i bn einz , c(z) = i cn einz , (2.59)
where

{b(z), c(z 0 )} = δ(σ − σ 0 )equaltime , z2 = z20 and 0 < σ < 2π or (2.60)

{bm , cn } = δm+n,0 (2.61)
1
< 0|b(z)c(z 0 )|0 > =
1 − ei(z−z0 )
12 CHAPTER 2. CONFORMAL FIELD THEORY

1
∼ + regular terms, then (2.62)
z − z”
1
: b(z)c(z 0 ) : = b(z)c(z 0 ) − (2.63)
z − z0
If b has weight hb = λ, then hc = 1 − λ. This is known since the action has weight 0 and the volume element
has weight (-1,-1). From the transformation δz = ε(z), b will change to

∂z 0 λ
b0 = ( ) b(z − ε)
∂z
= (1 − λ∂ε)(b − ε∂b), then (2.64)
δb = −λ∂b − ε∂b, and (2.65)
δc = −(1 − λ)∂c − ε∂c (2.66)
Z
δS = ¯
∂ε((∂b)c − λ∂(bc)), (2.67)

where (∂b)c − λ∂(bc) is the Noether current in this case, then T is

T = : (∂b)c : −λ∂(: bc :), and (2.68)

1
c(z)b(z 0 ) = + ... (2.69)
z − z0
1
b(z)c(z 0 ) = + ... (2.70)
z − z0
1 1 1 1
T (z)T (z 0 ) = −( )2 − 2λ∂( ∂ ) + λ2 ∂∂ 0
z − z0 z − z0 z − z0 (z − z 0 )2
−1 + 6λ − 6λ2
= + ...
2(z − z 0 )4
c
= (2.71)
2(z − z 0 )4
c = −2 + 12λ − 12λ2 = 1 − 3(2λ − 1)2 , central charge. (2.72)
From the Linear Dilaton theory c = D + 6αV 2 . Let the charges from the two theories be equal, ie. D + 6αV 2 =
1 − 3(2λ − 1)2 and solve for V. For the case when D=1, one will obtain
1
V = √ (2λ − 1). (2.73)
2π
In the special case (λ = 12 ), V=0, c=1, b and c can be writen as a linear combination of scalar fields Ψ1 and Ψ2
1
b(z) = √ (Ψ1 + iΨ2 )
2

1
c(z) = √ (Ψ1 − iΨ2 )
2
Z
1 ¯ 1 + Ψ2 ∂Ψ
¯ 2)
S= d2 z(Ψ1 ∂Ψ
4π

1
T = − Ψ1 ∂Ψ1
2
Note that if λ = 2 and V=0, then c is -26 from c = 1 − 3(2λ − 1)2 . This is the result obtained in chapter 1.
2.7. THE VIRASORO ALGEBRA 13

2.6.3 βγ theory
The next example, the bosonic case, let β and γ be commuting scalar fields. The action is given by:

Z
1 ¯
S=√ d2 zβ ∂γ, (2.74)
2π

where

¯ = ∂γ
∂β ¯ =0 (2.75)

The procedure is the same as for the spinor case.

1
β(z)γ(z 0 ) = + ... (2.76)
z − z0
1
γ(z)β(z 0 ) = − + ... (2.77)
z − z0
c = −1 + 3(2λ − 1)2 (2.78)

Note that if λ = 23 , then c is 11.

Xµ

D (b, c) −26 d
d D+ − 26 + 11 = 0 =⇒ D = 10.
Ψµ 2 (β, γ) 11 2

From
r
α0 µ α0 1 µ inz
XLµ µ
=x − p z=i Σ αn e
2 2 n

0
< XL (z), XL (z 0 ) > = ln(1 − ei(z−z ) )
∼ ln(z − z 0 ) + ... (2.79)

2.7 The Virasoro algebra

The worldsheet of a free closed string moving through space-time looks like a cylinder whose radius may
fluctuate depending on the excitation mode. We can map this cylinder to the complex plane. This map would
be equivalent to sqeezing one end of the string so it looks like a cone. Then smash the cone into a plane. Now
the worldsheet coordinates can be expressed as complex coordinates, where r will represent tau and the phase
term will represent the position on the string, sigma.

z → z̃ closed strings l = 2π, 0 < σ < 2π

z̃ = e−iz
14 CHAPTER 2. CONFORMAL FIELD THEORY

Im z Im z

t Re z

t=−infinity
Re z
2π

Instead of expanding in Fourier modes, do a Laurent Expansion.

X1
Xµ ∼ αµ z̃ −n
n n
What happens to T(z):

1
T (z) = : ∂X µ ∂Xµ :
α0
X
= Tm eimz
?
X
= Tm z̃ −m

T doesn’t transform as easily as guessed.

Subtract term with central charge to make it a primary field (tensor). Then transform each coordinate.
c 2 1
T (z)T (z 0 ) = + T (z 0 ) + ∂T (z 0 ) + . . . (2.80)
2(z − z 0 )4 (z − z 0 )2 z − z0
multiply both sides by v(z)

v(z)T (z)T (z 0 ) = v(z)[. . .]

c 1 3
δT = −λ =  ∂ v − 2∂vT − v∂T (2.81)
2 3!

z → z + v(z) z → z̃ = e−iz

∂ z̃ 2 c
T (z) = ( ) T (z̃) + {z̃, z} (2.82)
∂z 12

where {z̃, z} is known as a Schwarzian derivative, defined below:

2z̃ 000 z̃ 0 − 3(z̃ 00 )2

{z̃, z} = (2.83)
2(z̃ 0 )2
2.7. THE VIRASORO ALGEBRA 15

=1/2 for our example.

c
T = −z̃ 2 T̃ +
24
We can invert this and solve for z̃,

c
T̃ = −z̃ −2 T̃ + z̃ −2
24
X c
= Lm z −m−2 , Lm = −Tm + δm,0 (2.84)
24
Invert the equation and solve for L.
I
dz̃ m+1
Lm = z̃ T̃ (2.85)
2πi
R 2π dσ
The Hamiltonian 0 2π T can now be written in terms of L.
c
H = L0 − (2.86)
24
¯ m = 0 since ∂[z̃
We can see that ∂L ¯ m+1 T̃ ] = 0. This implies all Lm s generate symmetries.
The Ward identity for any operator A:

δA = i[Q, A]
dz
H
implies Q can be written as an integral of the current, j. Q = 2πi j(z)

The OPE of T with some primary field of weight h is given by:

h ∂A(z 0 )
T (z)A(z 0 ) = 0
A(z 0 ) + + ... (2.87)
(z − z )2 z − z0
Look at the variation of A:

δA = −h∂vA − v∂A

choose v = z̃ m+1 j = z̃ m+1 T̃ Q =

H
j ∼ Lm

⇒ δA = i[Lm , A]

[Q, A] = h(m + 1)z̃ m + Z̃ m+1 ∂A = [Lm , A]

expand A in a Laurent expansion:

∂ z̃ X
A = ( )−h Am z̃ −m
∂z
X
= Am z̃ −m−h (2.88)

This is the expansion for a primary field in the z̃ coordinates. Look at the commutator of L with A in these
cordinates.
16 CHAPTER 2. CONFORMAL FIELD THEORY

[Lm , An ] = h(m + 1)An+m − (n + m + h)An+m

= [(h − 1)m − n]Am+n (2.89)
We know the algebra, but what are the representations of the algebra.
c
δT = expected + ∂ 3 v
12

c
[Lm , T ] = expected + (m + 1)m(m − 1)z̃ m−2
12

c
[Lm , Ln ] = expected + (m3 − m)δm+n,0
12
c
= (m − n)Lm+n + (m3 − m)δm+n,0 ”VirasoroAlgebra” (2.90)
12
Look at the case m=0:

[L0 , An ] = −nAn , [L0 , Ln ] = −nLn (2.91)

L0 |ψi = E|ψi |ψ 0 i = Ln |ψi

L0 |ψ 0 i = [L0 , Ln ]|ψi + Ln L0 |ψi

= −nLn |ψi + ELn |ψi
= (E − n)Ln |ψi
= (E − n)|ψ 0 i (2.92)
For n < 0 ⇒ E 0 > E : E’=E-n → E is increasing.
For n > 0 ⇒ E 0 < E: Hence E is decreasing forever and there is no ground state energy. If H isn’t bounded
from below, the theory has a major problem!
Look at n=-1,0,1. Is this analogous to the raising and lowering operators for angular momentum.

[L0 , L1 ] = −L1

[L0 , L−1 ] = L−1

[L1 , L−1 ] = 2L0

We notice that this closed algebra is independent of c, therefore for every CFT we should get this subalgebra.
This is a Lie algebra SL(2,R) and is not compact.
Look at Quantum Mechanics:

[L0 , L+ ] = L+

[L0 , L− ] = −L−

[L+ , L− ] = 2L0
This is close to the algebra above, but not the same. The difference is actually very important!
2.8. MORE EXPANSIONS 17

2.8 More expansions

2.8.1 Free scalars
Now that we know how to transform to these z̃ coordinates, we can look at the same calculations and look for
similarities. The mode expansion is given by,

r
α0 µ α0 X 1 µ −m
XLµ µ
= x − i p ln z̃ + i α z̃ . (2.93)
2 2 m>0 m m
The two-point function which has to be radially ordered is given by,
α0 X 1 z̃ 0 µν 0
h|XLµ (z̃)XLν (z̃ 0 )|i = ( )η |z̃ | < |z̃| → time ordering...i.e. radial ordering (2.94)
2 m z̃
α0 z̃ 0
= ln |1 − |η µν . (2.95)
2 z̃
The normal ordered product is given by,

α0
ln |z − z 0 | in z picture.
: XX := XX = (2.96)
2
Now we can compare our definition for :: to switching a and a† around.

: XLµ (z̃)XLν (z̃ 0 ) : = : (XL+ + XL− )µ (XL+ + XL− )ν :

= XLµ (z̃)XLν (z̃ 0 ) + [X (−) (z 0 ), X (+) (z)]
n
α0 X 1 µν z̃ 0
= XLµ (z̃)XLν (z̃ 0 ) − η (2.97)
2 n n z̃

α0 µ α0
[xµ , −i p ln z̃] = XX + η µν ln |z̃ − z̃ 0 |. (2.98)
2 2
From normal ordering
α0 µν
X µ (z)X ν (z 0 ) = η ln(z − z 0 )+ : X µ (z)X ν (z 0 ) :,
2
the operator product can be writen as

product = singularity + normal ordering product

From I
dz m+1
Lm = z T (z), charge
2πi
where z m+1 T (z) is a conserved current and

r ∞
µ α0 X µ −m−1
∂X = −i α z
2 m=−∞ m
r
µ 2 µ
P = α
α0 0
X 1
Xµ = xµ + pµ ln z + i αµ z −m
m m
m6=0
18 CHAPTER 2. CONFORMAL FIELD THEORY

1
T (z) = : ∂X µ ∂Xµ :
α0
1 X −n1 −1 −n2 −1
= z z : αnµ1 αn2 µ :
2 n ,n
1 2

1X µ
Lm = : αm−n αnµ : (2.99)
2 n

The Hamiltonian (L0 ) is given by:

∞
α0 2 X µ
L0 = p + α−n αnµ (2.100)
4 n=1

The commutation relations are:

[X µ (σ), Πν (σ 0 )] = 2πiδ(σ − σ 0 ) (2.101)

µ ν 0
[X (σ), X (σ )] = 0, where X = XL (z) + XR (z̄) (2.102)
[XLµ (z), XLν (z 0 )] = XLµ (z)XLν (z 0 ) − XLν (z 0 )XLµ (z)
α0 µν α0
= η ln(z − z 0 ) − η µν ln(z − z 0 )
2 2
πi 0 µν d
= αη (step function)
2 dz
πi 0 µν
= α η δ(z − z 0 ), (2.103)
2
where differentiating a step function, a delta function is obtained.

Increasing time

c
c
Increasing time

An example for : eikX : is

eA eB = e[A,B] eB eA , then
α0
: eik1 X1 :: eik2 X2 : = e±iπ 2 k1 ·k2 : eik2 X2 :: eik1 X1 :,
0
q
For the special case D=1, X µ = α2 Ψ, k1µ = ± α20 = k2µ and let O1 = e±iψ and O2 = e±iψ , then

O1 O2 = e±iπ O2 O1 = −O2 O1

or
2.9. VERTEX OPERATORS 19

{O1 O2 , O2 O1 } = 0 (2.104)
0 0
ψ(z)ψ(z ) ∼ ln(z − z ), then (2.105)
iψ −iψ 0 − ln(z−z 0 ) iψ −iψ 0
: e (z) :: e (z ) : ∼ e : e (z)e (z ) :
1 iψ −iψ 0
∼ : e (z)e (z ) : (2.106)
z − z0
0
: eiψ (z) :: eiψ (z 0 ) : ∼ eln(z−z ) : eiψ (z)eiψ (z 0 ) :
∼ (z − z 0 ) : eiψ (z)eiψ (z 0 ) : . (2.107)

2.8.2 bc CFT
We can write b and c in terms of ψ: b =: eiψ : and c =: eiψ :,. The stress tensor is given as:

Tψ =: ∂ψ∂ψ : +V ∂ 2 ψ (2.108)
−m−λ −m−1+λ 0 1
P P
and from b(z) = bm z and c(z) = cm z and if the weight λ is an integer, b(z)c(z ) ∼ z−z 0 , then
the anticommutator between b and c is

{bm , cn } = δm+n,0 . (2.109)

bn and cn are annihilation operators for n > 0. For zero mode, n = 0
{b0 , c0 } = 1
If let |Ψi be a null state of b, ie. b0 |ψi = 0 and c0 |ψi = |χi,then

c0 |χi = c0 c0 |ψi = 0, {bm , bn } = {cm , cn } = 0 (2.110)

b0 |χi = b0 c0 |ψi = {b0 , c0 }|ψi = |ψi, then (2.111)
hψ||ψi = hχ|b0 b0 |χi = 0 (2.112)
hχ||χi = hψ|b0 b0 |ψi = 0, and (2.113)
hψ||χi 6= 0. (2.114)
ie. defining the inner product hχ|b0 |χi =
6 0 if |χi 6= 0

2.9 Vertex operators

Vertex operators are one-to-one corresponding to their states, I ' |0 > for an operator ∂X µ and momentum
operator pµ
r
µ α0 X µ −m−1
∂X (z) = −i α z
2 m m
pµ |0i = 0,
µ
p |0; ki = k µ |0; ki, then
r ∞
α0 X µ −m−1
∂X µ (z)|0i = −i α z |0i
2 m=1 −m
r
α0 µ µ µ
= −i (α + α−2 z + α−3 z 2 + ...)|0i letting z −→ 0
2 −1
r
α0 µ
∂X µ (0) = −i α |0i (2.115)
2 −1
20 CHAPTER 2. CONFORMAL FIELD THEORY

From |Ai ' A(z) or |Ai = A(0)|0i, ∂X µ operating at a bra is

r ∞
α0 X µ
h0|∂X (z) µ
= −i h0| α−m z −m−1 , letting z −→ ∞
2 m=1
r
µ α0 µ
h0|∂X (∞) = −i h0|α−1 , (2.116)
2
then the product can be consider as
h0| time order stuffed |0i
|{z} |{z}
time at +∞ time at −∞

From the equation of ∂X ,µ

α1µ is
I
µ dz
α−1 = z −1 ∂X µ (z) , (2.117)
2πi | {z }
conservedcurrent

z → e−iz

z’1
z z’1
The inner circle approaches z
the outside.
z
z’2

z’2

µ
α−2 |0i also can be obtained from
r
2 µ α0 µ µ µ
∂ X (z)|0i = −i (α + 2α−3 z + ...)|0i, for any α−m (2.118)
2 −2
r
α0 µ
∂ m X µ (0) = −i ((m − 1)!α−m + O(z))|0i, letting z −→ 0 (2.119)
2
r
µ 2 1
α−m (0) ' i ∂ m X µ (z). (2.120)
α0 (m − 1)!
µ ν
α−m α−n |0i can be obtained from

α0 µ
: ∂ m X µ ∂ ν X ν : |0i = − (m − 1)!(n − 1)!α−m ν
α−n + ... |0i (2.121)
2 |{z}
go to 0 in the infinite pass
Some examples of vertex operators are

eik·X |0i = |0; ki

lim : eik·X(z) : |0i = eik·X |0i, then
z−→0
: ∂X m1 ∂X m2 ...eik·X : ' α−m1 α−m2 ....|0; ki
2.10. PRIMARY FIELDS 21

2.10 Primary fields

A primary field A, A ⇒ |Ai, A(0)|0i = |Ai, with a state |m1 + m2 + ..., k > can be writen as

µ2
|Ai = α−m1 µ1 α−m 2
...|0; kiwhere (2.122)
1 1
A(z) = ... : ∂ n1 X µ1 ∂ n2 X µ2 ...eik·X : . (2.123)
(n1 − 1)! (n2 − 1)!

If we let λ = 2 for the bc theory:

1
b−m |ψi −→ ∂ m−2 b, b0 |ψi = 0 (2.124)
(m − 2)!
1
c−m |ψi −→ ∂ m+1 c (2.125)
(m + 1)!
Tbc (z) = : (∂b)c : −λ∂ : (bc) :, (2.126)

where

X
b(z) = bm z −m−λ
X
c(z) = cm z −m+λ

The Virasoro operator in this case is

I
dz m+1
Lbc
m = z Tbc
2πi
X
= −(n + λ)bn cm−n + λ(m + 1)bn cn−m + aδn,0
n
X
= (mλ − n) : bn cn−m : +aδm,0 (2.127)
n
c 2 1
Tbc (z)Tbc (z 0 ) ∼ 0
+ 0
Tbc + ∂Tbc (2.128)
2(z − z )4 (z − z )2 z − z0

where c = 1 − 3(2λ − 1)2 , and the commutation relations for L are given by:

c
[Lbc bc bc
m , Ln ] = (n − m)Ln+m + (m3 − n)δn+m,0 (2.129)
12
For m=1 and n=-1, the commutator is

[Lbc bc
1 , L−1 ]|Ψi = 2Lbc
0 |Ψi
L1 L−1 |Ψi − L−1 L1 |Ψi = λb0 c1 (1 − λ)b−1 c0 |Ψi
= λ(1 − λ)c1 b−1 |Ψi
= λ(1 − λ){c1 , b−1 }|Ψi
= λ(1 − λ)|Ψi (2.130)
22 CHAPTER 2. CONFORMAL FIELD THEORY

2.11 Operator product expansion

Consider the commutator between Lm and A

[Lm , A] = z m+1 ∂A + h(m + 1)z m A, where (2.131)

X
A(z) = An z −n−h (2.132)

For m = 0,

[L0 , A] = z∂A + hA (2.133)

[L0 , An ] = −nAn (2.134)
as z → 0, [L0 , A(0)] = hA(0) (2.135)
L0 |Ai = [L0 , A(0)]|0i + A(0)L0 |0i, L0 |0i = 0,
= hA(0)|0i = h|Ai (2.136)
as z → 0, [Lm , A] = 0, for m > 0, then (2.137)
Lm |Ai = 0, m > 0, (2.138)

where L0 is bounded from below.

2.12 Unitary CFTs

To correct the inner product with a negative norm, h...||...i ≥ 0 , consider a Virasoro operator L+
m = L−m or

hψ|Lm |χi = hL+

m ψ||χi (2.139)
and also X µ with h0; k||0; k 0 i = 2πδ(k − k 0 ). From [αmµ
, αnν ] = η µν δm+n,0 , for µ = ν = 0, we know that
µ 2 0
k α−1 |0; ki k < 0. This can be corrected by letting Φ → X , the action will be
Z
1 ¯
S= d2 z∂φ∂φ (2.140)
2πα0
for h ≥ 0, then

hA|[L1 , L−1 ]|Ai = hA|2L0 |Ai (2.141)

k L−1 |Ai k2 = 2h ≥ 0, L0 |vaci = 0, c > 0, then (2.142)
c
h0|[Lm , L−m] = 2mL0 + (m3 − m)|0i (2.143)
12
c
k L−m |0i k2 = (m3 − m) > 0 (2.144)
12
And we have gotten what we wanted; a positive norm for the highest weight state.