26.12.2012 Views

Introduction to Conformal Field Theory: With Applications to String ...

Introduction to Conformal Field Theory: With Applications to String ...

Introduction to Conformal Field Theory: With Applications to String ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

2.6 Opera<strong>to</strong>r Algebra of Chiral Quasi-Primary <strong>Field</strong>s 29<br />

• Applying the relation (2.45) <strong>to</strong> the Virasoro algebra (2.16) for values<br />

m = −1, 0, +1, we see that the chiral energy–momentum tensor is a quasiprimary<br />

field of conformal dimension (h, h) = (2, 0). In view of Eq. (2.25), this<br />

observation also explains the form of the Laurent expansion (2.42).<br />

• It is worth <strong>to</strong> note that the Schwarzian derivative S(w, z) vanishes for SL(2, C)<br />

transformations w = f (z) in agreement with the fact that T (z) is a quasi-primary<br />

field.<br />

2.6 Opera<strong>to</strong>r Algebra of Chiral Quasi-Primary <strong>Field</strong>s<br />

The objects of interest in quantum field theories are n-point correlation functions<br />

which are usually computed in a perturbative approach via either canonical quantisation<br />

or the path integral method. In this section, we will see that the exact twoand<br />

three-point functions for certain fields in a conformal field theory are already<br />

determined by the symmetries. This will allow us <strong>to</strong> derive a general formula for the<br />

OPE among quasi-primary fields.<br />

2.6.1 <strong>Conformal</strong> Ward Identity<br />

In quantum field theory, so-called Ward identities are the quantum manifestation of<br />

symmetries. We will now derive such an identity for the conformal symmetry of<br />

two-dimensional CFTs on general grounds. For primary fields φi, we calculate<br />

�� dz<br />

2πi ɛ(z) T (z) φ1(w1,<br />

�<br />

w1) ...φN (wN , wN )<br />

(2.46)<br />

=<br />

=<br />

N� �<br />

�<br />

φ1(w1, w1) ...<br />

�<br />

i=1<br />

N� �<br />

�<br />

φ1(w1, w1) ...<br />

i=1<br />

C(wi )<br />

dz<br />

2πi ɛ(z) T (z) φi(wi,<br />

�<br />

�<br />

wi) ...φN (wN , wN )<br />

hi �ɛ(wi) + ɛ(wi) �wi<br />

�<br />

�<br />

φi(wi, wi) ...φN (wN , wN )<br />

where we have applied the deformation of the con<strong>to</strong>ur integrals illustrated in Fig. 2.5<br />

and used Eq. (2.37). Employing then the two relations shown in Eq. (2.38), we can<br />

write<br />

�<br />

dz<br />

0 =<br />

2πi ɛ(z)<br />

�� �<br />

T (z) φ1(w1, w1) ...φN (wN , wN )<br />

(2.47)<br />

−<br />

N�<br />

�<br />

i=1<br />

hi 1<br />

+ �wi<br />

(z − wi) 2 z − wi<br />

�� �<br />

φ1(w1, w1) ...φN (wN , wN )<br />

�<br />

Since this must hold for all ɛ(z) of the form ɛ(z) =−z n+1 with n ∈ Z, the integrand<br />

must already vanish and we arrive at the <strong>Conformal</strong> Ward identity

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!