Advanced Quantum Field Theory

Roberto Casalbuoni
Dipartimento di Fisica Universit` di Firenze a Lezioni date all’Universita’ di Firenze nell’a.a. 2004/2005.

Contents
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4 4 6 6 9 10 12 14 17 27 27 29 30 33 36 36 38 45 52 54 56 63

1 Notations and Conventions 1.1 Units . . . . . . . . . . . . . . . . . . . . . . . 1.2 Relativity and Tensors . . . . . . . . . . . . . 1.3 The Noether’s theorem for relativistic fields . 1.4 Field Quantization . . . . . . . . . . . . . . . 1.4.1 The Real Scalar Field . . . . . . . . . . 1.4.2 The Charged Scalar Field . . . . . . . 1.4.3 The Dirac Field . . . . . . . . . . . . . 1.4.4 The Electromagnetic Field . . . . . . . 1.5 Perturbation Theory . . . . . . . . . . . . . . 1.5.1 The Scattering Matrix . . . . . . . . . 1.5.2 Wick’s theorem . . . . . . . . . . . . . 1.5.3 Feynman diagrams in momentum space 1.5.4 The cross-section . . . . . . . . . . . . 2 One-loop renormalization 2.1 Divergences of the Feynman integrals . . . 2.2 Higher order corrections . . . . . . . . . . 2.3 The analysis by counterterms . . . . . . . 2.4 Dimensional regularization of the Feynman 2.5 Integration in arbitrary dimensions . . . . 2.6 One loop regularization of QED . . . . . . 2.7 One loop renormalization . . . . . . . . . . 3 Vacuum expectation values and the 3.1 In- and Out-states . . . . . . . . . 3.2 The S matrix . . . . . . . . . . . . 3.3 The reduction formalism . . . . . . S . . .

. . . . . . . . . . . . . . . integrals . . . . . . . . . . . . . . .

matrix 73 . . . . . . . . . . . . . . . . . . 73 . . . . . . . . . . . . . . . . . . 79 . . . . . . . . . . . . . . . . . . 81

4 Path integral formulation of quantum mechanics 85 4.1 Feynman’s formulation of quantum mechanics . . . . . . . . . . . . . 85 4.2 Path integral in configuration space . . . . . . . . . . . . . . . . . . . 88

1

4.3 4.4 4.5 4.6 4.7 4.8 4.9 5 The 5.1 5.2 5.3 5.4 5.5 5.6 5.7 6 The 6.1 6.2 6.3 6.4 6.5 7 The 7.1 7.2 7.3 8 The 8.1 8.2 8.3 8.4 8.5 8.6

The physical interpretation of the path integral . The free particle . . . . . . . . . . . . . . . . . . The case of a quadratic action . . . . . . . . . . . Functional formalism . . . . . . . . . . . . . . . . General properties of the path integral . . . . . . The generating functional of the Green’s functions The Green’s functions for the harmonic oscillator

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

91 97 99 104 107 113 116 123 123 127 129 135 137 142 145

path integral in field theory The path integral for a free scalar field . . . . . . . . . . . . The generating functional of the connected Green’s functions The perturbative expansion for the theory λϕ4 . . . . . . . . The Feynman’s rules in momentum space . . . . . . . . . . . Power counting in λϕ4 . . . . . . . . . . . . . . . . . . . . . Regularization in λϕ4 . . . . . . . . . . . . . . . . . . . . . . Renormalization in the theory λϕ4 . . . . . . . . . . . . . .

renormalization group The renormalization group equations . . . . . . . . . . . . . . . . . Renormalization conditions . . . . . . . . . . . . . . . . . . . . . . . Application to QED . . . . . . . . . . . . . . . . . . . . . . . . . . Properties of the renormalization group equations . . . . . . . . . . The coefficients of the renormalization group equations and the renormalization conditions . . . . . . . . . . . . . . . . . . . . . . . . . . path integral for Fermi fields Fermionic oscillators and Grassmann algebras . . . . . . . . . . . . Integration over Grassmann variables . . . . . . . . . . . . . . . . . The path integral for the fermionic theories . . . . . . . . . . . . . . quantization of the gauge fields QED as a gauge theory . . . . . . . . . . . . . . . . . . . . . Non-abelian gauge theories . . . . . . . . . . . . . . . . . . . Path integral quantization of the gauge theories . . . . . . . Path integral quantization of QED . . . . . . . . . . . . . . Path integral quantization of the non-abelian gauge theories The β-function in non-abelian gauge theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

152 . 152 . 156 . 160 . 161 . 167 169 . 169 . 174 . 178 180 . 180 . 182 . 186 . 196 . 200 . 204

9 Spontaneous symmetry breaking 9.1 The linear σ-model . . . . . . . . . . . 9.2 Spontaneous symmetry breaking . . . . 9.3 The Goldstone theorem . . . . . . . . . 9.4 The Higgs mechanism . . . . . . . . . 9.5 Quantization of a spontaneously broken 2

213 . . . . . . . . . . . . . . . . . 213 . . . . . . . . . . . . . . . . . 219 . . . . . . . . . . . . . . . . . 223 . . . . . . . . . . . . . . . . . 225 gauge theory in the Rξ -gauge 230

9.6 ξ-cancellation in perturbation theory . . . . . . . . . . . . . . . . . . 235 10 The 10.1 10.2 10.3 Standard model of the electroweak interactions The Standard Model of the electroweak interactions . . . . . . . . . The Higgs sector in the Standard Model . . . . . . . . . . . . . . . The electroweak interactions of quarks and the Kobayashi-MaskawaCabibbo matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4 The parameters of the SM . . . . . . . . . . . . . . . . . . . . . . . 10.5 Ward identities and anomalies . . . . . . . . . . . . . . . . . . . . . A.1 Properties of the real antisymmetric matrices 239 . 239 . 244 . 249 . 258 . 259

A

265 . . . . . . . . . . . . . 265

3

Chapter 1 Notations and Conventions
1.1 Units

In quantum relativistic theories the two fundamental constants c e / , the light h velocity and the Planck constant respectively, appear everywhere. Therefore it is convenient to choose a unit system where their numerical value is given by c=/ =1 h (1.1)

For the electromagnetism we will use the Heaviside-Lorentz system, where we take also (1.2) 0 = 1 From the relation
0 µ0

= 1/c2 it follows µ0 = 1 (1.3)

In these units the Coulomb force is given by |F | = 1 e1 e2 4π |x1 − x2 |2 (1.4)

and the Maxwell equations appear without any visible constant. For instance the gauss law is ∇·E =ρ (1.5) dove ρ is the charge density. The dimensionless fine structure constant α= is given by α= 4 e2 4π (1.7) e2 4π 0 / c h (1.6)

using it follows from (1.10) (1.18) 5 .16) (1.17) (1. using the fact that also the elementary particle masses are usually given in MeV . using c = 3 · 1023 fermi · sec−1 (1.5 MeV (1.05 · 10−34 J · sec = 3. For instance.13) Therefore the approximate relation to keep in mind is 1 = 200 MeV · fermi. In fact from our choice the following equivalence relations follow ct E E Et ≈ =⇒ time ≈ lenght ≈ mc2 =⇒ energy ≈ mass ≈ pv =⇒ energy ≈ momentum ≈ / =⇒ energy ≈ (time)−1 ≈ (lenght)−1 h (1.6 · 10−13 1 MeV−1 = 197 fermi (1.602 · 10−19 J 3.Any physical quantity can be expressed equivalently by using as fundamental unit energy. it is enough to notice that the product ch has dimensions [E· ]. Furthermore.12) Using this relation we can easily convert a quantity given in MeV (the typical unit used in elementary particle physics) in fermi. the wave lenght of an electron is given by λe Compton = 200 MeV · fermi 1 1 ≈ ≈ 400 fermi ≈ me 0.58 · 10−22 sec 1 barn = 10−24 cm2 1 GeV−2 = 0.8) / In practice.12) 1 fermi = 3.15) (1. lenght or time in an equivalent fashion. mass.14) we get and Also.15 · 10−26 J · mt Recalling that it follows / ch = From which 1 eV = e · 1 = 1.3 · 10−24 sec 1 MeV−1 = 6.9) (1.389 mbarn (1.15 · 10−26 MeV · mt = 197 MeV · fermi 1.11) (1.5 MeV 0. Therefore / ch = 3 · 108 mt · sec−1 · 1.

3 The Noether’s theorem for relativistic fields We will now review the Noether’s theorem. −1.20) and can be raised by using the inverse metric tensor g µν ≡ gµν . −1). −i∇) ∂xµ (1. given a transformation involving both the fields and the coordinates. x). ∇ ∂x ∂t ∂x The four-momentum operator in position space is pµ → i ∂ = (i∂t .19) where x and p are the three-dimensional position and momentum. The position and momentum four-vectors are given by xµ = (t. 3 (1. then a conservation law follows.22) = ∂t . When both types of transformations are involved the total variation of a local quantity F (x) (that is a function of the space-time point) is given by ˜ ˜ ∆F (x) = F (x ) − F (x) = F (x + δx) − F (x) ∼ F (x) − F (x) + δxµ ∂F (x) = ˜ ∂xµ 6 (1.23) The following relations will be useful p2 = pµ pµ → − ∂ ∂ =− ∂xµ ∂xµ (1.1. (1. 1. More precisely. µ = 0. The scalar product between two four-vectors is given by a · b = aµ bν gµν = aµ bµ = aµ bµ = a0 b0 − a · b where the indices have been lowered by aµ = gµν aν (1. −1.21) (1. When the transformations are limited to the fields one speaks about internal transformations. if it happens that the action is invariant under this transformation. 2.24) (1. This allows to relate symmetries of the action with conserved quantities.26) . The four-gradient is defined as ∂ ∂ ∂ ∂µ = µ = .2 Relativity and Tensors Our conventions are as follows: the metric tensor gµν is diagonal with eigenvalues (+1.25) x · p = Et − p · x 1. p). pµ = (E.

i = 1. then ˜ SV = SV This gives rise to the conservation equation ∂µ Lδxµ + ∂L ∂L ∆φi − δxν i φi = 0 i ∂φ.µ 7 ∂µ J µ = 0 (1. It is then convenient to define a local variation δF .29) and let us consider a generic variation of the fields and of the coordinates.µ ∂φi . one gets different kind of conserved quantities.28) d4 x L(φi .30) ∂x If the action is invariant under the transformation. . .µ . From the general result in eq.µ ∂φ. N (1.35) .34) Pµ is the four-momentum of the system. . x). In the case of internal symmetries we take δxµ = 0. . In the case we take δxµ = aµ with aµ independent on x e ∆φi = 0.32) we get the following local conservation law µ Tν = ∂L i µ φ − Lgν . i ∂φ. ∂φi . The conserved current will be Jµ = ∂L ∂L i ∆φi = δφ . Let us start with an action invariant under space and time translations. From its local conservation we get four constant of motion Pµ = 0 d3 xTµ (1.31) This is the general result expressing the local conservation of the quantity in parenthesis. (1.The total variation ∆ keeps into account both the variation of the reference frame and the form variation of F . and of the corresponding symmetries of the action.ν .ν (1.33) Tµν is called the energy-momentum tensor of the system. x µ = xµ + δxµ ∂φ ˜ ∆φi (x) = φi (x ) − φi (x) ≈ δφi (x) + δxµ µ (1. According to the choice one does for the variations δxµ and ∆φi .µ µ ∂µ Tν = 0 (1.32) (1. depending only on the form variation ˜ δF (x) = F (x) − F (x) Then we get ∂F (x) ∂xµ Let us now start form a generic four-dimensional action ∆F (x) = δF (x) + δxµ S= V (1.27) (1.

j) which can be shown to be the representative of the infinitesimal generators of the Lorentz group in the field representation. a vector field does under rotations. we may have more that one conserved charge Q. As well known. whereas the remaining three correspond to Lorentz boosts. the transformation law of the fields φi under an infinitesimal Lorentz transformation can be written as 1 ∆φi = − Σij 2 µν µν φj (1. if the system has more that one internal symmetry.µ νρ j φ µ µ Tν xρ − Tρ xν + ∂L ij j Σ φ ∂φi νρ . three of them correspond to spatial rotations. In general. The last case we will consider is the invariance with respect to Lorentz transformations. This means that under a Lorentz transformation the components of the field mix together. that is we have a conserved charge for any ∆.39) We see that the number of independent parameters characterizing a Lorentz transformation is six. ν)) define a matrix in the indices (i. for instance. Let us recall that they are defined as the transformations leaving invariant the norm of a four-vector 2 x2 = x (1.38) =− νµ (1.41) and defining (watch at the change of sign) µ µ Mµ = xρ Tν − xν Tρ − ρν ∂L ij j Σ φ ∂φi ρν .42) 8 .36) In general.with an associated constant of motion given by Q= d3 xJ 0 (1. The coefficients Σµν (antisymmetric in the indices (µ.37) For an infinitesimal transformation xµ = Λµν xν ≈ xµ + with µν ν µν x (1.µ (1.µ ∂µ νρ xρ + 1 ∂L ij Σ 2 ∂φi νρ . Using this equation and the expression for δxµ we get the local conservation law 0 = ∂µ = 1 2 νρ ∂L i µ φ − Lgν ∂φi .ν .40) where we have required that the transformation of the fields is of first order in the Lorentz parameters µν . Therefore. the relativistic fields are chosen to belong to a representation of the Lorentz group ( for instance the Klein-Gordon field belongs to the scalar representation).µ (1. as.

9 . we can add the following steps • expansion of φα (x) in terms of a complete set of solutions of the Klein-Gordon equation. or for an interacting field in the interaction picture (and therefore evolving with the free hamiltonian).it follows the existence of six locally conserved currents (one for each Lorentz transformation) ∂µ Mµ = 0 (1. t). the Dirac and the electromagnetic field. t)]± = 0 (1.44) Three of these constants (the ones with ν and ρ assuming spatial values) are nothing but the components of the angular momentum of the field. In the following we will give the main results about the quantization of the scalar. allowing the definition of creation and annihilation operators.47) In the case of a free field.46) (1. 1. t)]± = iδαβ δ 3 (x − y) [φα (x.45) Πα (x) = ∂φα (x) where the index α describes both the spin and internal degrees of freedom • quantization through the requirement of the equal time canonical commutation relations for a spin integer field or through canonical anticommutation relations for half-integer fields [φα (x. Πβ (y. t). t)]± = 0. Πβ (y. • construction of the Fock space through the creation and annihilation operators. t). [Πα (x.43) νρ and consequently six constants of motion (notice that the lower indices are antisymmetric) Mνρ = d3 xM0 νρ (1.4 Field Quantization The quantization procedure for a field goes through the following steps: • construction of the lagrangian density and determination of the canonical momentum density ∂L (1. φβ (y.

57) (−) ˙ f ∗ ∂t g = f ∗ g − f˙∗ g (1. φ(y. ni ∈ Z (1.56) and L being the side of the quantization box (V = L3 ). t)] = iδ 3 (x − y).48) which can be derived from the variation of the action S= where L is the lagrangian density L= 1 ∂µ φ ∂ µ φ − m2 φ2 2 (1. t).51) From this we get the equal time canonical commutation relations ˙ [φ(x. the following quantity is a constant of motion and defines a scalar product (not positive definite) f |g = i where d3 xf ∗ ∂t g (−) |k|2 + m2 2π n L (1. t)] = [φ(x.55) (1.4. with k 0 = ωk = and k= with Z n = n1 i1 + n2 i2 + n3 i3 .54) (1. φ(y.52) By using the box normalization. a complete set of solutions of the Klein-Gordon equation is given by 1 −ikx 1 fk (x) = √ √ e . φ(y. t)] = 0 (1.1. ˙ ˙ [φ(x.53) corresponding respectively to positive and negative energies.1 The Real Scalar Field + m2 )φ(x) = 0 The relativistic real scalar free field φ(x) obeys the Klein-Gordon equation ( (1. t).49) giving rise to the canonical momentum Π= ∂L ˙ =φ ˙ ∂φ (1. V 2ωk 1 ikx 1 ∗ fk (x) = √ √ e V 2ωk (1. t). For any two solutions f and g of the Klein-Gordon equation.58) 10 .50) d4 xL (1.

(−) a† (k) = i d3 xφ(x)∂t fk (x) (−) (1.63) and obtain the commutation relations [a(k).68) d3 x Π2 + |∇φ|2 + m2 φ2 (1.66) for any k. a(k )] = [a† (k).53) are orthogonal with respect to this scalar product. The expano sion of the field in terms of these solutions is then (in the continuum) φ(x) = or.33) for the energymomentum tensor.69) . (1. and of total four-momentum k = k1 + · · · kn . more explicitly φ(x) = d3 k 1 (2π)3 2ωk [a(k)e−ikx + a† (k)eikx ] (1.The solutions of eq.64) (1.61) using the orthogonality properties of the solutions one can invert this expression a(k) = i ∗ d3 xfk (x)∂t φ(x). that for the Klein-Gordon field is Tµν = ∂µ φ∂ν φ − from which P0 = H = d3 xT 00 = 1 ∂ρ φ∂ ρ φ − m2 φ2 gµν 2 1 2 11 (1.59) We can go to the continuum normalization by the substitution 1 √ → V 1 (2π)3 (1.ni (1.65) From these equations one constructs the Fock space starting from the vacuum state which is defined by a(k)|0 = 0 (1. for instance 3 fk |fk = δk.67) 2 describe n Bose particles of equal mass m2 = ki .62) ∗ d3 k[fk (x)a(k) + fk (x)a† (k)] (1. This follows by recalling the expression (1. a† (k )] = 0 (1.k ≡ i=1 δni .60) and sending the Kr¨necker delta function into the Dirac delta function. a† (k )] = δ 3 (k − k ) [a(k). The states a† (k1 ) · · · a† (kn )|0 (1.

4. ∂xi (P = − d3 xΠ∇φ) (1.72) The propagator for the scalar field is given by 0|T (φ(x)φ(y))|0 = −i∆F (x − y) where ∆F (x) = − and ∆F (k) = k2 d4 k −ikx e ∆F (k) (2π)4 1 − m2 + i (1. t). t)] = [φi (x.2 The Charged Scalar Field The charged scalar field can be described in terms of two real scalar fields of equal mass 2 1 L= (∂µ φi )(∂ µ φi ) − m2 φ2 (1.71) : Pµ : a† (k)|0 = kµ a† (k)|0 (1.80) 1 φ† = √ (φ1 − iφ2 ) 2 (1. t)] = iδij δ 3 (x − y) ˙ ˙ [φi (x.78) The theory is invariant under the phase transformation φ → eiα φ. φj (y.76) i 2 i=1 and we can write immediately the canonical commutation relations ˙ [φi (x. t). t)] = 0 The charged field is better understood in a complex basis 1 φ = √ (φ1 + iφ2 ).Pi = d3 xT 0i = − d3 xΠ ∂φ . φj (y.81) . 2 The lagrangian density becomes L = ∂µ φ† ∂ µ φ − m2 φ† φ (1. t).74) (1.35)) jµ = i (∂µ φ)φ† − (∂µ φ† )φ 12 (1.77) (1. φj (y.70) and for the normal ordered operator Pµ we get (the normal ordering is defined by putting all the creation operators to the left of the annihilation operators) : Pµ := implying d3 k kµ a† (k)a(k) (1. (1.79) (1.73) (1.75) 1. and therefore it admits the conserved current (see eq.

with the corresponding constant of motion Q=i d3 x φ† ∂t φ (−) (1. t).88) (1. b† (k )] = 0 13 (1. t)] = iδ 3 (x − y) and [φ(x.87) ∗ d3 k fk (x)ai (k) + fk (x)a† i (k) (1. ˙ ∂φ Πφ† = ∂L ˙ =φ ˙ ∂ φ† (1. t). φ† (y. φ(y.85) Using the expansion for the real fields φi (x) = we get φ(x) = 1 1 ∗ d3 k fk (x) √ (a1 (k) + ia2 (k)) + fk (x) √ (a† 1 (k) + ia† 2 (k)) 2 2 (1. b(k )] = [a(k). t)] = [φ† (x. t). since Πφ = ∂L ˙ = φ† . t)] = 0 (1. 2 it follows φ(x) = φ† (x) = ∗ d3 k fk (x)a(k) + fk (x)b† (k) ∗ d3 k fk (x)b(k) + fk (x)a† (k) 1 b(k) = √ (a1 (k) − ia2 (k)) 2 (1.82) The commutation relations in the new basis are given by ˙ [φ(x. φ† (y. t). t)] = [φ† (x.83) (1. t)] = 0 ˙ ˙ ˙ ˙ [φ(x.91) . t).90) (1.89) from which we can evaluate the commutation relations for the creation and annihilation operators in the complex basis [a(k).84) Let us notice that these commutation relations could have also been obtained directly from the lagrangian (1. φ† (y. φ(y. b† (k )] = δ 3 (k − k ) [a(k). a† (k )] = [b(k).86) Introducing the combinations 1 a(k) = √ (a1 (k) + ia2 (k)).80).

97) The canonical momenta do not depend on the velocities. 1.3 The Dirac Field The Dirac field is described by the action S= V ¯ ˆ d4 x ψ(i∂ − m)ψ (1. the result one gets is the same as proceeding in a naive way. From the general expression for the energy momentum tensor (see eq. this creates a problem in going to the hamiltonian formalism. For this reason we will avoid to describe this extension. For the normal ordered charge operator we get : Q := d3 k a† (k)a(k) − b† (k)b(k) (1. In principle.95) where we have used the following notations for four-vectors contracted with the γ matrices (1.100) .96) v ≡ vµ γ µ ˆ The canonical momenta result to be Πψ = ∂L = iψ † .93) showing explicitly that a† and b† create particles of charge +1 and −1 respectively.ν The momentum of the field is given by Pk = d3 x T 0k =⇒ P = −i 14 d3 x ψ † ∇ψ (1.92) Therefore the operators a† (k) e b† (k) both create particles states with momentum k.33) we get µ µ ¯ ˆ ¯ Tν = iψγ µ ψ. as the original operators a† i . The only non vanishing propagator for the scalar field is given by 0|T (φ(x)φ†(y))|0 = −i∆F (x − y) (1. and we will proceed as in the standard case. In this particular case.ν − gν (ψ(i∂ − m)ψ) (1. ˙ ∂ψ Πψ† = ∂L =0 ˙ ∂ψ† (1.98) (1.94) 1.4.99) and using the Dirac equation µ ¯ Tν = iψγ µ ψ. In fact a rigorous treatment requires an extension of the classical hamiltonian approach which was performed by Dirac himself.We get also 2 : Pµ := d kkµ i=1 3 a† i (k)ai (k) = d3 kkµ a† (k)a(k) + b† (k)b(k) (1.

n)e−ipx + b† (p. n)e−ipx + d† (p. We will collect here the various properties of the spinors.108) m d(p. n)v(p. The decomposition of the Dirac field in plane waves is obtained by using the spinors u(p. the field variation is defined by 1 ∆φi = − Σij 2 µν In the Dirac case one has i ∆ψ(x) = ψ (x ) − ψ(x) = − σµν 4 from which µν µν φj (1. n)eipx γ0 v u Ep ±n (1.104) i 1 ¯ ¯ Mµ = iψγ µ xρ ∂ν − xν ∂ρ − σρν ψ = iψγ µ xρ ∂ν − xν ∂ρ + [γρ . the ¯ current ψγ µ ψ resulting from the phase invariance ψ → eiα ψ. 0. where Ep = |p| The four-vector nµ is the one that identifies the direction of the spin quantization in the rest frame.102) ψ(x) (1. γν ] ψ ρν 2 4 (1.106) where 12 is the identity matrix in 2 dimensions.103) i 1 Σµν = σµν = − [γµ . M 31 . 1). γν ] 2 4 Therefore the angular momentum density (see eq. Then nµ is obtained by boosting from the R rest frame to the frame where the particle has four-momentum pµ .105) By taking the spatial components we obtain J = (M 23 . n)¯(p. and we have defined σ ⊗ 12 = (1. ±n) solutions of the Dirac equation in momentum space for positive and negative energies respectively. ψ † (x) = 15 . n)eipx Ep (1.107) The expression of J shows the decomposition of the total angular momentum in the orbital and in the spin part. For instance. by quantizing the spin along the third axis one takes nµ in the rest frame as nµ = (0. We write ψ(x) = ±n d3 p (2π)3 d3 p (2π)3 m b(p.109) 2 + m2 . 0. M 12 ) = 1 d3 x ψ † −ix ∧ ∇ + σ ⊗ 12 ψ 2 σ 0 0 σ (1. The theory has a further conserved quantity.and the hamiltonian H= d3 x T 00 =⇒ H = i d3 x ψ † ∂t ψ (1.42)) is (1. ±n) e v(p. (1. n)¯(p.101) For a Lorentz transformation. n)u(p.

ψ(y. n)(ˆ − m) = 0 p ¯ p (ˆ + m)v(p. n)u(˜. n ) = v † (p. n ) = v † (p. p). t). n) = v (p. n)v(p. n)v(p. n ) = 0 ¯ p ˜ where.116) If we couple the Dirac field to the electromagnetism through the minimal substitution we find that the free action (1. ψ † (y. −p).112) v(p. n)u(p. n)¯(p.• Dirac equation (ˆ − m)u(p. • Completeness u(p. if pµ = (Ep . n)] (1.114) (1.113) (1. t). t). n) = u(p. n) + d† (p. n)(ˆ + m) = 0 p ¯ p • Orthogonality u(p. n ) = −¯(p. then pµ = (Ep . t). n)u(p. n)u(p. t)]+ = iδ 3 (x − y) or ψ(x. n)d(p. n) = v ±n The Dirac field is quantized by canonical anticommutation relations in order to satisfy the Pauli principle [ψ(x.110) (1. n)b(p. ψ † (y. t) + + =0 (1.118) . n)¯(p. t)]+ = ψ † (x.115) = δ 3 (x − y) The normal ordered four-momentum is given by : Pµ := ±n d3 p pµ [b† (p. t) and [Πψ (x. n ) = δnn ¯ v Ep δnn u† (p. n) = u ±n (1.95) becomes S= V ¯ ˆ ˆ d4 xψ(i∂ − eA − m)ψ (1.111) p+m ˆ 2m p−m ˆ 2m (1. n ) = m v (p.117) Therefore the electromagnetic field is coupled to the conserved current ¯ j µ = eψγ µ ψ 16 (1. ψ(y.

4 The Electromagnetic Field The lagrangian density for the free electromagnetic field. n)b(p.128) (1. F 20 . −F 31 . The propagator for the Dirac field is given by ¯ 0|T (ψα(x)ψβ (y))|0 = iSF (x − y)αβ where ˆ SF (x) = −(i∂ + m)∆F (x) = and SF (k) = d4 k −ikx e SF (k) (2π)4 (1.127) B = (−F 23 .123) L = − Fµν F µν 4 where Fµν = ∂µ Aν − ∂ν Aµ (1.121) ˆ 1 k+m = ˆ k 2 − m2 + i k−m+i (1. In fact. The resulting equations of motion are Aµ − ∂µ (∂ ν Aν ) = 0 The potentials are defined up to a gauge transformation Aµ (x) → Aµ (x) = Aµ (x) + ∂µ Λ(x) (1.122) 1. ∂t B =∇∧A (1.119) The expressions for the momentum angular momentum and charge operators show that the operators b† (p) and d† (p) create out of the vacuum particles of spin 1/2. n)d(p.This forces us to say that the integral of the fourth component of the current should be the charge operator. we find :Q: = e = ±n d3 x ψ † ψ d3 p e b† (p. n) − d† (p. It is possible to use the gauge invariance to require some 17 . −F 12 ) (1.4.120) (1. F 30 ). n) (1.126) ∂A . four-momentum pµ and charge e and −e respectively. expressed in terms of the four-vector potential is 1 (1. Aµ and Aµ satisfy the same equations of motion and give rise to the same electromagnetic field.125) In fact.124) and E = −∇A0 − from which E = (F 10 .

k µ = (E. In fact if we gauge transform Aµ (x) (1.135) (1.130) we see that Aµ has only two degrees of freedom. Since in this gauge all the gauge freedom is completely fixed (in contrast with the Lorentz gauge).134) The term in b(k) cancels. 2 (1.129) This is called the Coulomb gauge. λ = 1. Let us consider the four dimensional Fourier transform of Aµ (x) Aµ (x) = d4 k eikx Aµ (k) (1. and two further four vectors eλ (k).130) (1.136) The arbitrariness of b(k) is a consequence of the gauge invariance. Or we can choose a gauge such that A0 = ∇ · A = 0 (1. therefore it is left undetermined by the equations of motion. µ µ orthogonal to k k µ eλ = 0. we can perform a gauge transformation in such a way that the transformed field satisfies ∂µ Aµ = 0 This is called the Lorentz gauge. 2. For instance. λ = 1. it is easy to count the independent degrees of freedom of the electromagnetic field.132) Let us now decompose Aµ (k) in terms of four independent four vectors.138) .131) Substituting this expression in the equations of motion we get −k 2 Aµ (k) + kµ (k ν Aν (k)) = 0 (1.133) µ The decomposition of Aµ (k) reads ˜ Aµ (k) = aλ (k)eλ + b(k)kµ + c(k)kµ µ From the equations of motion we get ˜ ˜ −k 2 (aλ eλ + bkµ + ckµ ) + kµ (bk 2 + c(k · k)) = 0 µ (1. From the equations (1. For the other quantities we have k 2 aλ (k) = c(k) = 0 (1.particular condition on the field Aµ . −k). which can ˜ be chosen as k µ = (E. k).137) Aµ (x) → Aµ (x) + ∂µ Λ(x) then Aµ (k) → Aµ (k) + ikµ Λ(k) 18 (1. Another way of showing that there are only two independent degrees of freedom is through the equations of motion.

we can always to choose it equal to zero. Πν (y.139) Since the gauge transformation amounts to a translation in b(k) by an arbitrary function of k. Furthermore these amplitudes are different from zero only if the dispersion relation k 2 = 0 is satisfied. 2. t). Πν (y.141) (1.µ 4 2 2 Therefore ∂L = −Aµ.ν + Aµ. Let us consider now the quantization of this theory. t).123) in the following form 1 1 1 L = − [Aµ.µ ] = − Aµ.µ = F µν ∂Aµ.144) (1. the field Aµ (k) becomes Aµ (k) = aλ (k)eλ (k) µ (1. That is ν [Aµ (x. t)] = 0 with Πµ = ∂L ˙ ∂ Aµ To evaluate the conjugated momenta is better to write the lagrangian density (see eq. where however we will encounter other problems.145) implying (1.ν Aµ. t)] = [Πµ (x.µ ][Aµ. λ = 1.ν − Aν. t)] = igµ δ 3 (x − y) (1.ν − Aν. Due to the lack of manifest covariance of the Coulomb gauge we will discuss here the quantization in the Lorentz gauge. If we want to maintain the explicit covariance of the theory we have to require non trivial commutation relations for all the component of the field.ν Aν.143) [Aµ (x. This shows that the corresponding quanta will have zero mass.146) (1.142) (1. Therefore the choice b(k) = 0 is equivalent to the choice of the Lorentz gauge. With the choice b(k) = 0. Therefore we are left with the two degrees of freedom described by the amplitudes aλ (k).ν Πµ = It follows Π0 = ∂L = F µ0 ˙µ ∂A ∂L =0 ˙ ∂ A0 (1.ν + Aν.140) showing that k µ Aµ (k) = 0. (1. Aν (y.147) We see that it is impossible to satisfy the condition [A0 (x.where Λ(x) = d4 k eikx Λ(k) (1. Π0 (y. t). t)] = iδ 3 (x − y) 19 (1. t).148) .

For instance.ν Aν. modifying the lagrangian density in such a way to recover the equations of motion in a particular gauge.150) (1. (1. In fact from ∂L = −Aµ. More generally. To this end we observe that the difference between the two lagrangian densities is nothing but the second term of eq.ν we get 0 = − Aµ + ∂ µ (∂ ν Aν ) − ∂ µ (∂ λ Aλ ) = − Aµ the term (1.151) Then.ν )Aν 2 2 2 1 1 1 = ∂ µ Aµ. up to a four divergence.149) (just think to the Klein-Gordon case). (1.144) 1 1 1 Aµ.ν + Aν. given in eq. But doing so we will not recover the Maxwell equation.µ − g µν (∂ λ Aλ ). in the Lorentz gauge we have Aµ (x) = 0 and this equation can be obtained by the lagrangian density 1 L = − Aµ. we can write the new lagrangian density in the form 1 1 L = − Fµν F µν − (∂ µ Aµ )2 4 2 (1. ∂Aµ.ν Aν − (∂ µ Aµ.ν 2 (1.We can try to find a solution to this problem modifying the lagrangian density in such a way that Π0 = 0. we could add to the original lagrangian density a term of the form λ − (∂ µ Aµ )2 2 The corresponding equations of motion would be Aµ − (1 − λ)∂ µ (∂ λ Aλ ) = 0 20 (1.153) 1 − (∂ µ Aµ )2 (1.152) One can check that this form gives the correct equations of motion.155) 2 which is not gauge invariant. We now express this lagrangian density in terms of the gauge invariant one. is called the gauge fixing term.154) ∂L =0 ∂Aµ (1.µ = ∂ µ Aµ.ν Aµ.123). However we can take advantage of the gauge symmetry.157) (1.ν Aν − ∂ ν (∂ µ Aµ )Aν + (∂ µ Aµ )2 2 2 2 (1. The minus sign is necessary to recover a positive hamiltonian density.156) .

we can proceed with our program of canonical quantization.141) gives rise to ˙ [Aµ (x. (1.159). From eq.158) In the Lorentz gauge we find again Π0 = 0. In a given frame. Notice that now k 2 = 0. 0).in the plane orthogonal to nµ and k µ . n2 = 1. Therefore kµ (λ) (λ) µ = nµ (λ) µ = 0.In the following we will use λ = 1. and we will choose (λ) n0 > 0. nµ = (1. This must be a time-like vector. On the contrary. 2. To avoid the corresponding problem we can ask that ∂ µ Aµ = 0 does not hold as an operator condition. Then we take two four vectors µ . A further bonus is that in this way one has to do with local commutation relations. The physical states span a subspace which is defined by the previous relation.since we are considering solutions of the wave equation. We need four independent four vectors in order to expand the solutions in the momentum space.159) The price to pay to quantize the theory in a covariant way is to work in a Hilbert space much bigger than the physical one. The canonical momentum densities are ∂L Πµ = = F µ0 − g µ0 (∂ λ Aλ ) (1. 0. but rather as a condition upon the physical states phys|∂ µ Aµ |phys = 0 (1. λ = 1. then they will be chosen orthogonal and normalized in the following way (λ) (λ ) µ µ = −δλλ (1. explicitly ˙ Π0 = −∂ λ Aλ = −A0 − ∇ · A ˙ Πi = ∂ i A0 − ∂ 0 Ai = −Ai + ∂ i A0 (1. t)] = −igµν δ 3 (x − y) (1. 0. We will come back later to the condition (1. λ = 1. let us consider the unit four vector which defines the time axis. one needs to introduce non local commutation relations for the canonical variables.163) The four vectors µ . Since we don’t have to worry any more about the operator condition Π0 = 0.164) 21 .162) To get the quanta of the field we look for plane wave solutions of the wave equation. in the Coulomb gauge.161) Since the spatial gradient of the field commutes with the field itself at equal time. 2 (1.160) ˙ ∂ Aµ or. For instance. t). Aν (y. the canonical commutator (1.153) we see that Π0 = ∂L = −∂ µ Aµ ˙ ∂ A0 (1. being orthogonal to nµ are space-like.

166) By construction µ is orthogonal to the previous conditions.63) (λ) µ (k)aλ (k) =i ∗ d3 x fk (x)∂t Aµ (x) (−) (1. we have (1) µ = (0. (1. For any fixed µ.165) with (3) (3) (3) µ µ (λ) µ . 1. and we get (3) µ This four vector is completely fixed by = kµ − (n · k)nµ (n · k) (1. 1. (2) µ = (0. k). (1.169) and linearly independent. 0. nµ (3) µ = 0 (1.53).174) d3 x (λ ) (k)fk (x)∂t Aµ (x) aλ (k) = igλλ and analogously a† (k) = igλλ λ d3 x (λ ) µ (k)Aµ (x)∂t fk (x) (−) (1.172) where we have included the hermiticity condition for Aµ (x). this expansion is the same as the one that we wrote for the Klein-Gordon field. with the (λ) substitution µ aλ (k) → a(k). Then they satisfy the completeness relation (λ) (λ ) µ ν gλλ = gµν (1. 0.170) In the frame where nµ = (1.167) As a last unit four vector we choose nµ (0) µ = nµ (1. Using the orthogonality of the (λ ) µ ’s we find µ (−) ∗ (1. 0).Next. 0. 0. Then. 0. (3) µ = (0. orthogonal to nµ and lying in the plane (k. 0) and k µ = (k. 1) (1. 0. we define a unit space-like four vector. = −1 (1.173) with the functions fk (x) defined as in eq.171) The plane wave expansion of Aµ is Aµ (x) = d3 k 2ωk (2π)3 3 (λ) µ (k) λ=0 aλ (k)e−ikx + a† (k)eikx λ (1.168) These four vectors are orthonormal (λ) (λ ) µ µ = g λλ (1. 0). n) . from eq.175) 22 .

are physical states. it could destroy the linearity of the Hilbert space. Not even the vacuum state satisfies it. a† (k )] = 0 λ λ (1. being bilinear in the states.180) d3 k f (k)a† (k)|0 λ (1. In fact. our expectation is that only the transverse states (λ = 1. λ = = d3 k d3 k f (k)f (k ) 0|aλ (k)a† (k )|0 λ d3 k d3 k f (k)f (k ) 0|[aλ (k). Let us consider a one-particle state |1. a† (k )] = λ = ∗ d3 x d3 y − fk (x)f˙ (y) igµν gλλ k (λ ) µ (λ ) µ (λ ) ν (k) (k )gλ λ δ 3 (x − y) ∗ − f˙k (x)fk (y) −igµν gλλ (k) (λ ) ν (k )gλ λ δ 3 (x − y) (1. λ|1. aλ (k )] = [a† (k). 2).179) = −gλλ Therefore the states with λ = 0 have negative norm.176) = − ∗ d3 x fk (x)i∂t fk (x)gλλ (−) and using the orthogonality relations for the functions f )k [aλ (k). µ (k)aλ (k)e (−) Aµ (x) = (A(+) (x))† (1. if we consider the positive and negative frequency parts of the field A(+) (x) = µ d3 k 2ωk (2π)3 3 (λ) −ikx . So we will try to modify the condition in a linear one ∂ µ Aµ |phys.181) But this would be a too strong requirement.178) The commutation rules we have derived for the operators aλ (k) create some problem.177) (1.From these expressions we can evaluate the commutator [aλ (k). let us notice that formulated in the way we did. Therefore the relevant thing is to show that the states satisfying the Lorentz condition have positive norm. λ = its norm is given by 1. a† (k )|0 λ d3 k |f (k)|2 (1. = 0 (1. a† (k )] = −gλλ δ 3 (k − k ) λ Analogously [aλ (k). For the moment being we have ignored the gauge fixing condition phys|∂ µ Aµ |phys = 0. To discuss the gauge fixing condition. However. but its meaning is that only part of the total Hilbert space is physical. This problem does not come out completely unexpected.182) λ=0 23 .

190) More generally we can make the following observation. and it ensures that N has positive eigenvalues.3 (λ) µ (k)aλ (k) (1.189) These states have vanishing norm ||Φk ||2 = 0 (m) (1.192) 24 .183) Φk (m) = 1 † (a0 (k) − a† (k))m Φk (0. Then the following states satisfy the condition (1.188) Let us denote by Φk (n0 . 0) 3 m! (1.167). and require ∂ µ A(+) (x)|phys. = 0 µ This allows us to satisfy automatically the original requirement (−) phys. = 0 µ (1.it is possible to weaken the condition.187) (1.183) (1. = 0 Notice that [a0 (k) − a3 (k).185) = −(n · k). we get kµ from which [a0 (k) − a3 (k)]|phys. a† (k ) − a† (k )] = −δ 3 (k − k ) + δ 3 (k − k ) = 0 0 3 (3) µ d3 k e−ikx kµ 2ωk (2π)3 λ=0. kµ (0) µ = (n · k) (1. and with n3 longitudinal photons (that is with polarization µ (k)). For instance Na† (k)|0 = − 0 d3 k a† (k )[a0 (k ).|(∂ µ A(+) + ∂ µ Aµ )|phys.191) Notice the minus sign that is a consequence of the commutation relations.186) (1.184) (+) To make this condition more explicit let us evaluate the four divergence of Aµ i∂ µ A(+) (x) = µ Using eq. a† (k)]|0 = a† (k)|0 0 0 0 (1. Let us consider the number operator for scalar and longitudinal photons N= d3 k (a† (k)a3 (k) − a† (k)a0 (k)) 3 0 (1. (1. n3 ) the state with n0 scalar photons (that is with polarization (0) (3) µ (k)).

|H|phys. consider the hamiltonian. = phys.193) since a0 and a3 act in the same way on a physical state (see eq. Then ϕn |N|ϕn = 0 (1.187)). we get 2 phys. are completely arbitrary.| d3 k ωk λ=1 a† (k)aλ (k)|phys.194) Therefore all the physical states with a total definite number of scalar and longitudinal photons have zero norm. (1. we have H = = ˙ d3 x : [Πµ Aµ − L] : 1 1 ˙ ˙ d3 x : F µ0 Aµ − (∂ λ Aλ )A0 + Fµν F µν + (∂ λ Aλ )2 : 4 2 (1. the coefficients ci .197) The proof that a physical state has a positive norm can be extended to the case in which also transverse photons are present. except for the vacuum state (n = 0).195) A generic physical state with zero transverse photons is a linear superposition of the previous states ci |ϕi (1.0 (1.Let us consider a physical state with a total number n of scalar and longitudinal photons.200) 25 . Then ϕn |ϕn = δn. It follows n ϕn |ϕn = 0 (1. λ (1. Of course. For instance.196) |ϕ = c0 |ϕ0 + i=0 This state has a positive definite norm ϕ|ϕ = |c0 |2 ≥ 0 (1.199) Since on the physical states a0 and a3 act in the same way. but this is not going to modify the values of the observables. appearing in the expression of a physical state.198) One can easily show the hamiltonian is given by the sum of all the degrees of freedom appearing in Aµ 1 H = 2 = 3 dx : i=1 3 3 2 ˙ ˙ A2 + (∇Ai )2 − A2 − ∇A0 : i 0 d3 k ωk : λ=1 a† (k)aλ (k) − a† (k)a0 (k) : 0 λ (1.

m = 1 and n = 1.196).) n·k 0 (1. we can always choose |ϕ proportional to |ϕ0 . m = 0 ϕ|Aµ (x)|ϕ = c0 c1 0| d3 k 2ωk (2π)3 e−ikx [ (3) µ (k)a3 (k) + (0) µ (k)a0 (k)]|ϕ1 + c.204) From eqs.c. (1.] (1.168) we have (3) µ + (0) µ = kµ (k · n) (1. This is crucial for the explicit covariance and locality of the theory. Since only |ϕT . The arbitrariness in defining the state |ϕ has. (1. a very simple interpretation. in fact.201) Since Aµ change the occupation number by one unit and all the states |ϕn have zero norm (except for the state with n = 0). contributes to the evaluation of an observable quantity . Consider the following matrix element ϕ|Aµ (x)|ϕ = n. that is it corresponds to perform a gauge transformation. this does not mean that we are always working in the restricted physical space. However. It corresponds to add to Aµ a four gradient.c.c. because in a sum over the intermediate states we need to include all the degrees of freedom.m cn cm ϕn |Aµ (x)|ϕm (1. the only non vanishing contributions come from n = 0. since Λ=0 (1.The generic physical state is of the form |ϕT ⊗ |ϕ . (1.167) and (1.202) In order to satisfy the gauge condition the state |ϕ1 must be of the form |ϕ1 = and therefore ϕ|Aµ (x)|ϕ = d3 k 2ωk (2π)3 [ (3) µ (k) d3 q f (q)[a† (q) − a† (q)]|0 3 0 (1.206) (1. with |ϕ defined as in eq.205) from which ϕ|Aµ (x)|ϕ = ∂µ Λ(x) with Λ(x) = d3 k 2ωk (2π)3 1 (ic c1 e−ikx f (k) + c.207) It is important to notice that this gauge transformation leaves Aµ in the Lorentz gauge.208) 26 .203) + (0) −ikx f (k) µ (k)][c0 c1 e + c.

we are interested to evaluate the probability amplitude of detecting at t = +∞ a given set of free particles specified by a vector state Φf .213) where the state Φi will be specified by assigning the set of incoming free particles in terms of eigenstates of momentum. In practice the preparation and the detection processes are made at some finite times. From the expansion in eq.because the momentum k inside the integral satisfies k 2 = 0. Once we know Φ(+∞).214) We will define the S matrix as the operator that give us |Φ(+∞) once we know |Φ(−∞) |Φ(+∞) = S|Φ(−∞) (1. the spin projection of fermions and the polarization of the photons.1 Perturbation Theory The Scattering Matrix In order to describe a scattering process we will assign to the vector of state a condition at t = −∞ |Φ(−∞) ≡ |Φi (1. The equations of motion will tell us how this state evolves with time and it will be possible to evaluate the state at t = +∞. The amplitude is Sf i = Φf |Φ(+∞) (1.209) (1.5 1.215) The amplitude Sf i is then Sf i = Φf |S|Φi 27 (1. It follows that our ideal description will be correct only if these times are much bigger than the typical interaction time of the scattering process. in QED we will have to specify how many electrons.211) 1. we will detect the final states.212) d4 k −ikx e D(k) (2π)4 1 k2 + i (1. (1.210) d4 k e−ikx (2π)4 k 2 + i (1. For instance.5.172) one gets immediately the expression for the photon propagator 0|T (Aµ (x)Aν (y)|0 = −igµν By defining D(x) = and D(k) = − we get 0|T (Aµ(x)Aν (y)|0 = igµν D(x − y) (1. spin and so on. positrons and photons are in the initial state and we will have to specify their momenta. where.216) . ideally.

28 . The result in terms of ordered T -products is ∞ +∞ (−i)n +∞ S =1+ dt1 · · · dtn T H I (t1 ) · · · H I (tn ) (1. if t1 ≥ t2 ≥ · · · ≥ tn .222) This expression is a symbolic one and it is really defined by its series expansion. then T H I (t1 ) · · · H I (tn ) = H I (t1 ) · · · H I (tn ) (1. For instance.217) 0 where the index S identifies the Schr¨dinger representation and HS is the free hamilo tonian.221) This result can be written in a more compact form by introducing the T -ordered exponential +∞ −i S=T e dt H I (t) −∞ (1. also the S matrix enjoys the same property. The motivation for introducing the T -ordered exponential is that it satisfies the following factorization property t3 t3 t2 O(t)dt T e t1 O(t)dt =T e t2 O(t)dt T e t1 (1.219) One can verify that this is indeed a solution.224) It follows that if the theory is Lorentz invariant.Therefore Sf i is the S matrix element between free states (|Φi and Φf | are called in and out states respectively). and that it satisfies explicitly the boundary condition at t = −∞.223) If there are no derivative interactions we have +∞ −i S=T e dt H I (t) −∞ +i =T e d4 x Lint (1.218) |Φ(t) = HI |Φ(t) ∂t whereas the operators evolve with the free hamiltonian. The perturbative expansion consists in evaluating |Φ(t) by iterating this integral equation. the state vectors satisfy the Schr¨dinger equation with the interaction hamilo tonian (evaluated in the interaction picture) ∂ (1. In the interaction representation defined by 0 |Φ(t) = eiHS t |ΦS (t) .220) n! −∞ −∞ n=1 The T -product of n terms means that the factors have to be written from left to right with decreasing times. To evaluate the S matrix we first transform the Schr¨dinger equation in the interaction representation in an o integral equation i |Φ(t) = |Φ(−∞) − i t −∞ dt1 H I (t1 )|Φ(t1 ) (1. 0 0 O(t) = eiHS t OS (t)e−iHS t (1.

it is enough to consider all the possible contractions of the fields appearing in the T -product. and recalling that the VEV of a normal product is zero. that is in terms of Feynman propagators.2 Wick’s theorem The matrix elements of the S matrix between free particle states can be expressed as vacuum expectation values (VEV’s) of T -products. we get the Wick’s theorem. The Wick’s theorem is summarized in the following equation −i T e −i =: e d4 x j(x)φ(x) d4 x j(x)φ(x) 1 − :e 2 d4 x d4 y j(x)j(y) 0|T (φ(x)φ(y))|0 (1.226) T (φ1 φ2 ) = : φ1 φ2 : + 0|T (φ1φ2 )|0 3 (1. (1.228) 4 T (φ1 φ2 φ3 φ4 ) = + : φ1 φ2 φ3 φ4 : + i=j=k=l=1 : φi φj : 0|T (φk φl )|0 (1. Therefore to evaluate the VEV of a T -product of an arbitrary number of free fields.1. We will use the simplified notation φi ≡ φ(xi ). The Wick’s theorem is then obtained by expanding both sides of this equation in powers of j(x) and taking the VEV of both sides. We get T (φ) = : φ : (1.5. fermionic and photon fields. These V EV s satisfy an important theorem due to Wick that states that the T -products of an arbitrary number of free fields (the ones we have to do in the interaction representation) can be expressed as combinations of T -products among two fields. For instance.229) 0|T (φiφj )|0 0|T (φk φl )|0 and so on. The T -product of two field operators is sometimes called the contraction of the two operators. Let us now expand both sides of eq.230) 29 . By taking the VEV of these expression.225) in a series of j(x) and compare term by term. from the last of the previous relations we get 4 0|T (φ1φ2 φ3 φ4 )|0 = i=j=k=l=1 0|T (φiφj )|0 0|T (φk φl )|0 (1.227) T (φ1 φ2 φ3 ) = : φ1 φ2 φ3 : + i=j=k=1 : φi : 0|T (φj φk )|0 (1.225) where φ(x) is a free real scalar field and j(x) an ordinary real function. The previous formula can be easily extended to charged scalar.

For the fermions one has to remember that the T -product is defined in a slightly different way. 3.An analogous theorem holds for the photon field.231) where σP = ±1 is the sign of the permutation (i. As an illustration the previous formula becomes 4 0|T (ψ1 ψ2 ψ3 ψ4 )|0 = i=j=k=l=1 σP 0|T (ψi ψj )|0 0|T (ψk ψl )|0 (1. More explicitly 0|T (ψ1 ψ2 ψ3 ψ4 )|0 = − + 0|T (ψ1 ψ2 )|0 0|T (ψ3ψ4 )|0 0|T (ψ1 ψ3 )|0 0|T (ψ2ψ4 )|0 0|T (ψ1 ψ4 )|0 0|T (ψ2ψ3 )|0 (1. j. 4) appearing on the right hand side.5.237) .235) (1.232) 1.3 Feynman diagrams in momentum space We will discuss here the Feynman rules for the scalar and spinor QED.234) (1. Let us recall that the electromagnetic interaction is introduced via the minimal substitution ∂µ → ∂µ + ieAµ For instance.238) (1. k.236) (1.233) but another interaction term appears (the seagull term). = −ie φ† ∂µ φ − (∂µ φ† )φ Aµ + e2 A2 φ† φ Therefore the electromagnetic field is coupled to the current jµ = ie φ† ∂µ φ − (∂µ φ† )φ (1. for the charged scalar field we get 1 Lfree = ∂µ φ† ∂ µ φ − m2 φ† φ − Fµν F µν → 4 1 → [(∂µ + ieAµ )φ)]† [(∂ µ + ieAµ ) φ] − m2 φ† φ − Fµν F µν 4 The interaction part is given by Lint. l) with respect to the fundamental one (1. In the spinor case we have a simpler situation 1 1 ¯ ˆ ¯ ˆ ˆ Lfree = ψ(i∂ − m)ψ − Fµν F µν → ψ(i∂ − eA − m)ψ − Fµν F µν 4 4 giving the interaction term ¯ Lint = −eψγµ ψAµ 30 (1. 2. This gives a minus sign any time we have a permutation of the fermion fields which is odd with respect to the original ordering.

Graphical representation for both internal and external particle lines. and we are using the box normalization. The Feynman amplitude M (often one defines a matrix T wich differs from M by a factor i) can be obtained by drawing at a given perturbative order all the connected and topologically inequivalent Feynman’s diagrams. polarization. etc. 1. At the same time we measure momenta and polarizations of the final states. The diagrams are obtained by combining together the graphics elements given in Figures 1.5. after extraction of the factor (2π)4 δ 4 ( p) in (1. 1.2 .239) VE Here f = i.1 . .5.5.239). QED vertex Scalar QED seagull Fig.In a typical experiment in particle physics we prepare beams of particles with definite momentum. For this reason it is convenient to work in a momentum representation. The amplitude M is given by the sum of the amplitudes associated to these diagrams.Graphical representation for the vertices in scalar and spinor QED. . fermions m M (1. Fermion line Photon line Scalar line Fig. The result for a generic matrix element of the S matrix can be expressed in the following form f |S|i = (2π)4 δ 4 ( in pin − out pout ) ext.5. bosons 1 2EV ext.2.1 and 1. 31 .

5. field.5.A fermion loop can appear inside any Feynman diagram (attaching the rest of the diagram by photonic lines).At each diagram we associate an amplitude obtained by multiplying together the following factors • for each ingoing and/or outgoing external photon line a factor λ (or its comµ plex conjugate for outgoing photons if we consider complex polarizations as the circular one) • for each ingoing and/or outgoing external fermionic line a factor u(p) and/or u(p) ¯ • for each ingoing and/or outgoing external antifermionic line a factor v (p) ¯ and/or v(p) • for each internal scalar line a factor i∆F (p) (propagator) • for each internal photon line a factor ig µν D(p) (propagator) • for each internal fermion line a factor iSF (p) (propagator) • for each vertex in scalar QED a factor −ie(p + p )(2π)4 δ 4 ( in pi − out pf ).3) . 1.5.m. 1. 1. where p and p are the momenta of the scalar particle (taking p ingoing and p outgoing) • For each seagull term in scalar QED a factor 2ie2 gµν (2π)4 δ 4 ( • for each vertex in spinor QED a factor −ieγµ (2π)4 δ 4 ( ingoing in pi − out pf ) pf ) pi − outgoing • Integrate over all the internal momenta with measure d4 p/(2π)4 . This gives rise automatically to the conservation of the total four-momentum. In the following we will use need also the rules for the interaction with a static external e. The correspending graphical element is given in Fig. • In a graph containing closed lines (loops).3 .4. Fermion loop Fig. a factor (−1) for each fermionic loop (see Fig. The 32 .

whereas each external fermionic line contributes with (m/V E)1/2 .5. Fig. Therefore. Furthermore the conservation of the total four momentum gives a term (2π)4 δ 4 ( i pi − f pf ). From the rules of the previous Chapter we know that each external photon line contributes with a factor (1/2V E)1/2 .4 . since the static external field breaks translational invariance.rule is to treat it as a normal photon.242) where M is the Feynman amplitude which can be evaluated by using the rules of the previous Section.243) . The upper end of the line can be attached to any fermion line. . pi ) which collide and produce a set of final particles with four momenta pf = (Ef . 1.5.Graphical representation for an external electromagnetic field. except that one needs to substitute the wave function of the photon 1/2 1 (λ) (1.4 The cross-section Let us consider a scattering process with a set of initial particles with four momenta pi = (Ei . pf ). 1. we loose the factor (2π)3 δ( in pin − out pout ). Let us consider the typical case of a two particle collision giving rise to an N particles final state.240) µ (q) 2EV with the Fourier transform of the external field Aext (q ) = µ d3 qe−iq · x Aext (x ) µ (1.241) where q is the momentum in transfer at the vertex where the external field line is attached to. Also. If we exclude the uninteresting case f = i we can write Sf i = (2π) δ 4 4 i pi − f pf fermioni m VE 1/2 bosoni 1 2V E 1/2 M (1. the probability per unit time of the transition is given by w = |Sf i |2 with w = V (2π)4 δ 4 (Pf − Pi ) i 1 2V Ei 33 f 1 2V Ef (2m)|M|2 fermioni (1.

Since in the volume V the momentum is given by p = 2πn/L. Furthermore.244) w gives the probability per unit time of a transition toward a state with well defined quantum numbers. in the final equation.where. since in our normalization we have a particle in the volume V . the number of final states is given by L 2π 3 d3 p (1. and vrel is the relative velocity of the ingoing particles.245) The cross-section is defined as the probability per unit time divided by the flux of the ingoing particles. For the bosons this follows from the normalization condition of the functions fk (x).250) Notice that the dependence on the quantization volume V disappears.249) vrel vrel f (2π)3 We obtain dσ = V vrel 1 V d 3 pf V (2π)4 δ 4 (Pf − Pi ) 2 (2π)3 4V E1 E2 1 4E1 E2 vrel (2m) fermioni f f f 1 2V Ef 3 (2m)|M|2 fermioni = (2π)4 δ 4 (Pf − Pi ) d pf |M|2 (2π)3 2Ef (1. 34 .248) Then the cross-section for getting the final states with momenta between pf e pf +dpf is given by V V V d 3 pf dσ = w dNF = w (1. and has the dimensions of a length to the square [t−1 2 t] = [ 2 ] (1. the total cross-section. we have written 1 m = (2m) VE 2V E (1.246) In fact the flux is defined as vrel ρ = vrel /V . but we are rather interested to the final states having momenta between pf and pf + dpf . as it should be. which is obtained by integrating the previous expression over all the final momenta. is Lorentz invariant. For the fermions recall that in the box normalization the wave function is m u(p)e−ipx EV from which d3 xρ(x) = V V (1. for reasons of convenience.247) d3 x m † u (p)u(p) = 1 VE (1.

252) We see that also this factor is Lorentz invariant. Furthermore p1 p2 vrel = v1 − v2 = − (1. from which E1 E2 |vrel | = E1 m2 = |p1 | = m2 |p1 | = m2 E1 2 E1 − m2 1 2 m2 E1 − m2 m2 = 2 1 2 (p1 · p2 )2 − m2 m2 1 2 (1. as the factors d3 p/2E.In fact. 0) and vrel = v1 . 35 . M is invariant.251) E1 E2 but in the frame where the particle 2 is at rest (laboratory frame) we have p2 = (m2 . as it follows from the Feynman rules.

.1 .The Feynman diagram for the Coulomb scattering at the third order in the electric charge and at the first order in the external field. 2. 2. we can easily see that the relevant Feynman diagrams are the ones of Fig.p ' ' ' p p' p.p p p p'. 36 .Chapter 2 One-loop renormalization 2.1.p a) p k c) p' p-k k+q p.p p p.1 Divergences of the Feynman integrals Let us consider the Coulomb scattering.p ' ' ' p b) p' k p.k p' k p.k k p.1.1 .p ' ' d) Fig. and we assume that the external field is weak enough such to take only the first order. p p' = p.k p. If we expand the S matrix up to the third order in the electric charge.

5) (2. but. from which we get an expansion of the type ephys = e + a2 e3 + · · · = e(1 + a2 e2 + · · ·) (2. The first problem we encounter is that we would like to have the results of our calculation in terms of measured quantities as ephys .8) . Let us write explicitly the amplitudes for the other diagrams ¯ Mb = u(p )ie2 Σ(p ) Mc = u(p )(−ieγ µ ) ¯ where ie2 Πµν (q) = (−1) i d4 k i Tr (−ieγ µ ) (−ieγ ν ) 4 ˆ+q−m+i ˆ−m+i (2π) k ˆ k (2.3) or Σ(p) = i (2. To show this. k. ρ q2 + i (the minus sign originates from the fermion loop) and therefore Πµν (q) = i The last contribution is ¯ Md = u(p )(−ie)e2 Λµ (p . p)u(p)Aext (p − p) µ 37 (2. Analogously one can check that all the other third order contributions diverge.The Coulomb scattering can be used. 2.4) For large momentum.1a). for instance the self-energy contribution to one of the external photons (as the one in Fig.1.6) −igµν 2 νρ ie Π (q)Aext (p − p)u(p). to define the physical electric charge of the electron. This could be done by inverting the previous expansion. consider. in principle.2) (2.9) d4 k 1 1 γµ γν Tr ˆ+q−m k−m+i ˆ (2π)4 k ˆ (2.1) in terms of the parameter e which appears in the original lagrangian. We have Ma = u(p )(−ieγ µ )Aext (p − p) ¯ µ where ie2 Σ(p) = i d4 k −ig µν (−ieγν ) (−ieγµ ) 2 4 (2π) k +i p−k−m+i ˆ ˆ d4 k 1 1 γµ = −e2 γ 4 µ k2 + i (2π) p−k−m+i ˆ ˆ d4 k p−k+m ˆ ˆ 1 γµ 2 γµ 4 2 − m2 + i (2π) (p − k) k +i i p−m+i ˆ ie2 Σ(p) u(p) (2. This is done assuming that the amplitude is linear in ephys .7) i (−ieγ µ )Aext (p − p)u(p) µ p −m+i ˆ q =p −p (2. and here comes the second problem. the integrand behaves as 1/k 3 and the integral diverges linearly. the coefficient of the expansion are divergent quantities.

otherwise we would have to look for some physical meaning of the regularization procedure we choose. the previous considerations. and removing the cut-off (that is by taking the limit for the cut-off going to the infinity). as for instance QED. As far QED is concerned we will study in detail the renormalization at one-loop. This bring us to the other problem.2 Higher order corrections In this section we will illustrate. the inversion of eq.10) The problem of the divergences is a serious one and in order to give some sense to the theory we have to define a way to define our integrals. in a simplified context. (f = i) m Sf i = (2π)δ(Ef − Ei ) M(1) (2.where e2 Λµ (p . 2. p) = or Λµ (p .1). 2. We will give these criteria without a proof but we will try only to justify them On a physical basis. the result should be finite. Of course. we want that the theory does not depend on the way in which we define the integrals. Thinking to the regularization in terms of a cut-off this means that considering the observable quantities in terms of ephys . However. That is we give a prescription in order to make the integrals finite. This can be done in various ways. (2. However there is a restrict class of renormalizable theories. By doing so.1. Consider again the Coulomb scattering. We will not discuss the renormalization at all order and neither we will prove which criteria a theory should satisfy in order to be renormalizable.12) VE where M(1) = u(p )(−ieγ µ )u(p)Aext (p − p ) ¯ (2.13) µ 38 . Since now the coefficients are finite we can indeed perform the inversion and obtaining e as a function of ephys and obtain all the observables in terms of the physical electric charge (that is the one measured in the Coulomb scattering). This is what is called the regularization procedure of the Feynman integrals. or. p) = −i d4 k α 1 1 1 γµ γα 2 γ 4 ˆ (2π) k +i p −k−m+i ˆ p−k−m+i ˆ ˆ (2. as we shall see later by the more convenient means of dimensional regularization. as introducing an ultraviolet cut-off. We will say that the theory is renormalizable when this dependence cancels out.11) d4 k i −igαβ i (−ieγ α ) γµ (−ieγ β ) 2 4 ˆ (2π) k +i p −k−m+i ˆ p−k−m+i ˆ ˆ (2.1). a priori we will introduce in the observables a dependence on the renormalization procedure. At the lowest order the Feynman amplitude in presence of an external static electromagnetic field is simply given by the following expression (see Fig. and in fact in most of the theories this does not happen. this cancellation is not obvious at all.

This quantity could be measured and then we could compare the result with the theoretical result obtained from the previous expression. and again extract the value of e from the data. However the result is tied to a theoretical equation evaluated at the lowest order in e.18) (2. just to exemplify how the renormalization procedure works. This is usually referred to as the vacuum polarization contribution and in this Section we will consider only this correction. In order to get a better determination one has to evaluate the higher order corrections.m.21) 39 . the diagram in Fig.15) µ 4π|x| For the following it is more convenient to re-express the previous amplitude in terms of the external current generating the external e. field. 0) (1) (2. of charge Ze we have Ze Aext (x) = (− . Since we have Aµ = jµ we can invert this equation in momentum space. but with the photon propagator modified in the following way −ig µν −ig µν −ig µρ 2 −ig λν −ig µν −ie2 Πµν → + (ie Πρλ (q 2 )) 2 = + q2 q2 q2 q q2 q4 (2.19) From these expressions one can get easily the differential cross-section dσ/dΩ.16) gµν ν j (q) q2 + i (2. For instance. the next-to-leading contributions ar at the order e4 .1.14) is the Fourier transform of the static e. (2. In the case of a point-like source. In this way we get a measure of the electric charge. The total result is M(1) + Mc = u(p )(−ieγµ )u(p) ¯ −ig µν −ig µρ 2 −ig λν + 2 (ie Πρλ ) 2 q2 + i q +i q +i xt (−ijν (q)) (2.1c gives one of these contribution.and Aext (q) = µ d3 x Aext (x)e−iq · x µ (2. The lowest order result is at the order e2 in the electric charge. This contribution is given in eqs.6) and (2. by getting Aµ (q) = − from which M with ext jµ (x) = (Zeδ 3 (x). field. 0) (2.20) The sum gives a result very similar the to one obtained from M(1) alone.7).m.17) −ig µν ext = u(p )(−ieγµ )u(p) 2 ¯ (−ijν (p − p)) q +i (2. 2.

As we shall see later on. The functions I(q 2 ) can be evaluated by various tricks (we will evaluate it explicitly in the following) and the result is 1 Λ2 1 I(q ) = log 2 − 2 2 12π m 2π 2 1 dz z(1 − z) log 1 − 0 q 2 z(1 − z) m2 (2.The vacuum polarization correction to the propagator.2. at high values of the momentum it behaves as d4 p p2 (2.27) + Fig. but in any case the integral is divergent. 2. In fact. they do not contribute to the final result. .23) p4 The corresponding divergence is only logarithmic. This follows by evaluating the integral by cutting the upper limit by a cut-off Λ.26) Summarizing the propagator undergoes the following modification −ig µν −ig λν → [1 − e2 I(q 2 )] q2 q2 This relation is shown in graphical terms in Fig.24) The terms omitted are proportional to q µ q ν . (2. the real divergence is weaker due to the gauge invariance. We will see that the correct behaviour is given by d4 p (2.As discussed in the previous Section Πµν is a divergent quantity.28) 40 .1. 2. By evaluating it with a cut-off we get Πµν (q 2 ) = −gµν q 2 I(q 2 ) + · · · (2. The total result is then ¯ M(1) + Mc = u(p )(−ieγ0 )u(p) −i q2 1− e2 Λ2 e2 q 2 log 2 − 12π 2 m 60π 2 m2 (−iZe) (2. and since the photon is always attached to a conserved current.1 .2.25) For small values of q 2 (−q 2 << m2 ) one can expand the logarithm obtaining I(q 2 ) ≈ 1 Λ2 1 q2 log 2 + 12π 2 m 60π 2 m2 (2.22) and therefore it diverges in a quadratic way.

30) q 60π m Now we compare this result with the experiment where we measure dσ/dΩ in the limit q → 0. On the contrary. Notice that in this expression there are no more divergent quantities. Furthermore. is compensated by the divergence implicit in e.2. such that Q2 = −q 2 = µ2 .2. 2. the various terms in the expansion turn out to be finite and the theory is said to be renormalizable. We can extract again from the data the value of the electric charge. It must be stressed that the quantity that is fixed by the experiments is not the ”parameter” e appearing in the original lagrangian. This means that the quantity e is not really defined unless we specify how to relate it to measured quantities.2. As already stressed this is not automatically satisfied. If this is possible. In the previous example we were using q → 0. It follows that the parameter e. which is obtained from the experimental data. In this way the renormalized charge depends on the scale µ. must diverge for Λ → ∞. the divergent part coming from the self-energy and vertex corrections cancel out (see later). This relation is usually referred to as the renormalization condition. For the moment being we have included only the vacuum polarization corrections. In other words.2 sum up to give (omitting the contribution of the fermionic 41 . Rather the class of renormalizable theories (that is the theories for which the previous procedure is meaningful) is quite restricted. rather than eR as done previously.in the limit q → 0 the result looks exactly as at the lowest order. but this time we will measure eR . We will define the charge by choosing an arbitrary value of the square of the momentum in transfer qµ . however if we restrict our problem to the charge renormalization. also eR . It will be convenient in the following to call this quantity eB (tha bare charge). the divergence which we have obtained in the amplitude. The graphs in Fig. the renormalization procedure is a way of reorganizing the terms of the perturbative expansion in such a way that the parameter controlling it is the renormalized charge eR . 2. Neglecting other corrections the total amplitude is given by Fig. As we have noticed the final expression we got does not contain any explicit divergent quantity. We have seen that the measured electric charge is different from the quantity appearing in the expression of the vertex. is a finite quantity. the renormalized charge will be called e. but rather eR . Said that. except that we need to introduce a renormalized charge eR given by e2 Λ2 eR = e 1 − log 2 12π 2 m 1/2 (2. since they have been eliminated by the definition of the renormalized charge. Therefore the procedure outlined here has meaning only if we start with an infinite charge. Let us consider now a process like e− µ− → e− µ− .29) In term of eR and taking into account that we are working at the order e4 we can write e2 q 2 −i (1) M + Mc = uf (ieR γ0 )ui 2 1 − R 2 2 (−iZeR ) ¯ (2. In fact renormalizability is not a generic feature of the relativistic quantum field theories.

we get the amplitude in the form M(eB ) = e2 F (Q2 )[1 − e2 I(Q2 ) + e4 I 2 (Q2 ) · · ·] B B B (2.35) . 2.) e2 B −ig λν −ig µρ −ig λσ −ig τ ν −ig µν −ig µρ 2 + 2 (ieB Πρλ ) 2 + 2 (ie2 Πρλ ) 2 (ie2 Πστ ) 2 +· · · B B q2 q q q q q Q2 =µ2 (2.The vacuum polarization contributions to the e− µ− → e− µ− scattering. e− e− eB e− eB eB eB eB eB q eB + eB eB + eB eB eB + µ− µ− µ− Fig. we get e2 B −igµν 1 − e2 I(q 2 ) + e4 I 2 (q 2 ) + · · · B B q2 Q2 =µ2 (2. (2.33) obtaining at the order e4 e2 = e2 [1 + e2 I(µ2 ) + · · ·] B 42 (2. following the discussion of the previous Section we can try to re-express everything in terms of the renormalized charge (2.33) Since the fermionic contribution is the same for all the diagrams.2 . However.. As we know I(Q2 ) is a B B divergent quantity and therefore the amplitude is not defined. lines.2.31) By considering only the gµν contribution to Πµν . by calling it F (Q2 ).34) where e2 F (Q2 ) is the amplitude at the lowest order in e2 .33).32) This expression looks as the lowest order contribution to the propagator be defining the renormalized electric charge as e2 = e2 [1 − e2 I(µ2 ) + e4 I 2 (µ2 ) + · · ·] B B B (2. We can easily invert by series the eq.

33) and therefore it depends also on µ. the renormalized charge e will be measured through the cross-section. we see that this is given by the tree amplitude e2 F (Q2 ) times a geometric series arising from the iteration of B 43 .36) dz z(1 − z) log 1 + 0 Q2 z(1 − z) m2 (2. and tells us how we have to change the coupling. and therefore.38) µ 3π µ dz z(1 − z) log Q2 α log 2 + · · · 3π µ and in this limit MR (e) = e2 F (Q2 ) 1 − (2. when we change the definition scale µ. In fact. What is going on is that the implicit dependecnce of e on µ cancels out the explicit dependence of MR (e). the charge renormalization makes the amplitude finite. In fact.25) for I(Q2 ) we get I(Q2 ) = 1 Λ2 1 log 2 − 2 2 12π m 2π 1 (2.37) Since the divergent part is momentum independent it cancels out in the expression for MR (e). This can be formalized in a way that amounts to introduce the concept of renormalization group that we will discuss in much more detail later on.39) At first sight MR (e) depends on µ. Evaluating I(Q2 ) − I(µ2) per large Q2 (and µ2 ) we get e2 (I(Q2 ) − I(µ2 )) = − 2α π 2α → − π 1 0 1 0 m2 + Q2 z(1 − z) → m2 + µ2 z(1 − z) Q2 α Q2 dz z(1 − z) log 2 = − log 2 (2.By substituting this expression in M(e0 ) we get M(eB ) = e2 [1 + e2 I(µ2 ) + · · ·]F (Q2 )[1 − e2 I(Q2 ) + · · ·] = = e2 F (Q2 )[1 − e2 (I(Q2 ) − I(µ2 )) + · · ·] ≡ MR (e) From the expression (2. This follows from MR (e) = M(eB ) and on the observation that M(eB ) does not depend on µ. This equation is the prototype of the renormalization group equations. (2. Going back to the original expression for M(e0 ). but actually this is not the case. Therefore. µ) By differentiating this expression with respect to µ we get dM(eB ) ∂MR ∂MR ∂e = + =0 dµ ∂µ ∂e ∂µ (2.41) (2. e is defined through the eq.40) This equation say formally that the amplitude expressed as a function of the renormalized charge is indeed independent on µ. by definition. let us be more explicit by writing M(eB ) = MR (e(µ). Furthermore. it will be finite.

42) Also this expression is cooked up in terms of the divergent quantities eB and I(Q2 ).49) (2. (2. Therefore.48) We see that α(Q2 ) increases with Q2 .47) It follows that the true expansion parameter in the perturbative series is the running coupling e(Q2 ).50) 44 . it comes natural to define a coupling running with the momentum of the photon. By using eq. that is when 1− or 3π Q =e µ2 2 Q2 α(µ2 ) log 2 = α(µ2) 3π µ 1 −1 α(µ2 ) (2. but again e2 (Q2 ) is finite.43) We can put the last two equations in the form 1 e2 (Q2 ) and subtracting them 1 e2 (Q2 ) we get e2 (Q2 ) = e2 (µ2 ) 1 + e2 (µ2 )[I(Q2 ) − I(µ2 )] (2. e2 B 1 e2 (µ2 ) = 1 + I(µ2 ) e2 B (2.44) By using this result we can write MR (e) as MR (e) = e2 (Q2 )F (Q2 ) (2.38) we obtain α(Q2 ) = α(µ2 ) Q2 α(µ2 ) log 2 1− 3π µ (2.46) − 1 e2 (µ2 ) = I(Q2 ) − I(µ2 ) (2. in terms of the bare one e2 (µ2 ) = e2 B 1 + e2 I(µ2 ) B (2.45) = 1 + I(Q2 ). therefore the perturbative approach looses validity for those values Q2 of such that α(Q2 ) ≈ 1.the vacuum polarization diagram. This can be shown explicitly by evaluating the renormalized charge. e(µ2 ). e2 (Q2 ) = e2 [1 − e2 I(Q2 ) + e4 I 2 (Q2 ) + · · ·] = B B B e2 B 1 + e2 I(Q2 ) B (2.

see later). in fact. they can be absorbed in the original parameters.3 The analysis by counterterms The previous way of defining a renormalizable theory amounts to say that the original parameters in the lagrangian. This discussion shows clearly that the behaviour of the running coupling constant is determined by I(Q2 ).52) This value of Q2 is referred to as the Landau pole. In fact. We will show now that at one-loop the divergent contributions come only from the three diagrams discussed before. because they have to be fixed by the experiments (renormalization conditions). that is the theory becomes more and more perturbative. The infinite contributions are called counterterms. as e. We see that the theory will be renormalizable if the counterterms we add to make the theory finite have the same structure of the original terms in the lagrangian. We will see that in the so called non-abelian gauge theories the running is opposite.For α(µ2 ) ≈ 1/137 we get Q2 ≈ e1281 ≈ 10556 µ2 (2. We get D = 4 − IF − 2IB (2. by increasing the energy.51) Actually α(Q2 ) is divergent for 3π Q2 2 = e α(µ ) ≈ e1291 ≈ 10560 2 µ (2. Then one can try to separate the infinite from the finite part of the parameters (this separation is ambiguous. let us define D as the superficial degree of divergence of a one-loop diagram the difference between 4 (the number of integration over the mimentum) and the number of powers of momentum coming from the propagators (we assume here that no powers of momenta are coming from the vertices. and by definition they have the same operator structure of the original terms in the lagrangian.53) 45 . that is from the vaccum polarization contribution. However they can also be regularized in the same way as the other integrals. In particular. should be infinite and that their divergences should compensate the divergences of the Feynman diagrams. if this is the case. On the other hand. This means that the coefficients of these counterterms have to be infinite. which however are arbitrary. the procedure of regularization can be performed by adding to the original lagrangian counterterms cooked in such a way that their contribution kills the divergent part of the Feynman integrals. Therefore in the case of QED the perturbative approach maintains its validity up to gigantic values of the momenta. α(Q2 ) increases with Q2 since in the expression e2 (µ2 )(I(Q2 ) − I(µ2 )) the coefficient of the logarithm is negative (equal to −α(µ2 )/(3π)). 2. which is not true in scalar QED).

there are only three divergent one-loop diagrams and all the divergences can be brought back to the three functions Σ(p).54) Also. In fact.57) (2. all the diagrams with an odd number of external photons vanish. The divergence in Πµν . In such a case one has only to show that eliminating these three divergences (primitive divergences) the theory is automatically finite. Furthermore.1 with four external photons (light-light scattering) has D = 0 (logarithmic divergence). 2. p). Also the vacuum polarization diagram has an effective degree of divergence equal to 2.56) From these equations we can get V . gauge invariance often lowers the divergence. since the photon is an eigenstate of C with eigenvalue equal to -1. Πµν (q) e Λµ (p . This is trivial at one-loop. the other will contain the counterterms. one written in terms of the physical parameters. if at each vertex there are Ni lines of the particle of type i. then we have Ni V = 2Ii + Ei from which 2V = 2IF + EF . And finally the divergence of Λµ (p . 2 V = EF + EB (2.3.1. V equal to the number of internal lines V = IF + IB (2.where IF and IB are the number of internal fermionic and bosonic lines. but gauge invariance makes it finite. but it can be made finite if the previous functions are such.3. the photon self-energy. 2 from which 1 IB = EF . For instance the diagram of Fig. We will call also the original parameters and fields of the theory the bare parameters and the bare fields and we will use an index B in order to distinguish them from the physical quantities.55) 3 D = 4 − EF − EB (2. This does not mean that an arbitrary diagram is not divergence. Therefore. IF and IB in terms of EF and EB 1 IF = EF + EB . To realize this program we divide up the lagrangian density in two parts. can be absorbed in the wave function renormalization of the photon (the mass of the photon is not renormalized due to the gauge invariance). 2. V = 2IB + EB (2. since the trace of an odd number of γ matrices is zero. p) goes into the definition of the parameter e. In fact. Therefore the 46 . In general (Furry’s theorem) it follows from the conservation of charge conjugation.58) 2 It follows that the only one-loop superficially divergent diagrams in spinor QED are the ones in Fig. In particular we will show that the divergent part of Σ(p) can be absorbed into the definition of the mass of the electron and a redefinition of the electron field (wave function renormalization). the effective degree of divergence can be less than the superficial degree. Going through the loop one has to have a number of vertices.

for sake of simplicity.. we get 1+C ¯ˆ ¯ ¯ Fµν F µν + gauge − fixing L = (1 + B)iψ ∂ψ − (m + A)ψψ − e(1 + D)ψγµ ψAµ − 4 (2. 47 1/2 Z2 = (1 + B). Defining the renormalization constant of the fields Z1 = (1 + D).59) 4 2 and the counter terms piece Lc.t.t. 2. Adding together Lp and Lc. C E ¯ˆ ¯ ¯ˆ Lc.The superficially divergent one-loop diagrams in spinor QED two pieces of the lagrangian should look like as follows: the piece in terms of the physical parameters Lp 1 1 ¯ ˆ ¯ Lp = ψ(i∂ − m)ψ − eψγµ ψAµ − Fµν F µν − (∂µ Aµ )2 (2.3. = iB ψ ∂ψ − Aψψ − Fµν − (∂µ Aµ )2 − eDψ Aψ 4 2 (2.60) and we have to require that la sum of these two contributions should coincide with the original lagrangian written in terms of the bare quantities.63) .t. Z3 = (1 + C) (2. we write the bare fields as ψB = Z2 ψ.62) Aµ = Z3 Aµ B 1/2 (2. EF 0 EB 2 2 D 0 4 0 2 0 1 2 1 0 Fig.61) where. we have omitted the gauge fixing term.1 .

are determined recursively at each perturbative order in such a way to eliminate the divergent parts and to respect the renormalization conditions. We will see later how this works in practice at one-loop level. Then we regularize the theory giving some prescription to get finite Feynman integrals. because the finite piece is fixed to be an observable quantity. as the muon. This looks very unnatural. phrased in a different way.. which come about when removing the regularization. Therefore if we consider the theory for a different particle. one starts dividing the bare lagrangian in two pieces. and then remove the regularization. eB = (2. the relation between the bare and the physical electric charge is universal (that is it does not depend on the kind of charged particle under consideration). The counter terms A. to fix the finite part. Summarizing.65) 1/2 Z2 Z2 Z3 we get 1 µν ¯ ˆ ¯ ¯ L = iψB ∂ψB − mB ψB ψB − eB ψB γµ ψB Aµ − FB..µν FB + gauge − fixing 1/2 B µ B B Z2 4 Z2 Z3 (2. at fixed regulator. Another observation is that Z1 and Z2 have to do with the self-energy of the electron.we obtain m+A ¯ eZ1 ¯ 1 µν ¯ ˆ L = iψB ∂ψB − ψB ψB − ψ γ ψ Aµ − FB. one in the coupling and the other in some parameter (regulator) which defines the regularization. That is we have a double limit. we use the renormalization conditions. by requiring that the divergences of the Feynman integrals. This requirement gives the renormalization conditions. B. and since Z3 comes from the photon self-energy. one would get a different bare electric charge for the two particles. As a consequence 1/2 eB = e/Z3 . Notice that although the counter terms are divergent quantity when we remove the cut off. Since the separation of an infinite quantity into an infinite plus a finite term is not well defined.64) and putting m+A eZ1 mB = . If one considers only theory involving scalar. The part containing the counter terms is determined. are cancelled out by the counter term contributions. and as such they depend on the electron mass. Before going into the calculations for QED we want to illustrate some general results about the renormalization. Notice that the division of the parameters in physical and counter term part is well defined.66) B 4 So we have succeeded in the wanted separation. order by order. which has the same interactions as the electron and differs only for the value of the mass (mµ ≈ 200me ). one would have to tune the bare electric charge at different values in order to get the same physical charge. 48 . but the gauge invariance of the theory implies that at all the perturbative orders Z1 = Z2 . we will order them according to the power of the coupling in which we are doing the perturbative calculation. After evaluating a physical quantity we remove the regularization. The order of the limit is first to work at some order in the coupling. Or. .µν FB + gauge − fixing (2.

1a and 2. since all the terms in the lagrangian density have dimensions smaller or equal to 4 ¯ dim[ψψ] = 3.72) (2. Therefore. The mass dimensions of the fields can be easily evaluated from the observation that the action is dimensionless in our units / (h = 1). In fact each monomial Oi will appear multiplied by a coupling gi L= i gi Oi (2. It turns out also that these are renormalizable theories ( a part some small technicalities). we see that QED is renormalizable. as 1/p3 . From this counting one can see that only the lagrangian densities containing monomials in the fields with mass dimension smaller or equal to the number of space-time dimensions have a finite number of divergent diagrams. it is not difficult to construct an algorithm which allows to evaluate the ultraviolet (that is for large momenta) divergence of any Feynman diagram. 2 dim[ψ] = n−1 2 (2. In the case of the electron self-energy (see Figs.67) has a mass dimension n. dim[(∂ µ Aµ )2 ] = 4 (2. we see that dim[φ] = dim[Aµ ] = n − 1. 2. defined as dn x L (2.68) In particular.71) 49 . coming from the propagators.fermion and massless spin 1 (as the photon) fields.69) The condition on the dimensions of the operators appearing in the lagrangian can be translated into a condition over the coupling constants. ¯ dim[ψγµ ψAµ ] = 4. Looking at the kinetic terms of the bosonic fields (two derivatives) and of the fermionic fields (one derivative). Then.1b) one has an integration over the four momentum p and a behaviour of the integrand.73) that is the couplings must have positive dimension in mass are to be dimensionless.1. the lagrangian density.70) therefore dim[gi ] = 4 − dim[Oi ] The renormalizability requires dim[Oi ] ≤ 4 from which dim[gi ] ≥ 0 (2. giving a linear divergence (it turns out that the divergence is only logarithmic). In QED the only couplings are the mass of the electron and the electric charge (2. in 4 dimensions the bosonic fields have dimension 1 and the fermionic 3/2.1. in n space-time dimensions.

However one could think that this requirement is too technical and one could imagine other ways of giving a meaning to lagrangians which do not satisfy this 50 .3. We see that the linear σ-models are renormalizable theories.77) −δg2 (ψψ)2 The other two need counter terms of the type ¯ ¯ δg3 (ψψ)3 + δg4 (ψψ)4 (2. if we have written the most general lagrangian. Therefore adding to L the counter term Lc. The renormalization requirement restricts in a fantastic way the possible field theories.which is dimensionless. In the non renormalizable case. the only divergent diagrams which appear are the ones corresponding to the processes described by the operators appearing in the lagrangian. The first diagram modifies the already introduced counter term ¯ (ψψ)4 . order by order. Therefore the theory is a predictive one.74) Here ρ has dimension 1 and λ is dimensionless. At each order we have to introduce new counter terms having an operator structure different from the original one. the number of divergent diagrams increase with the perturbative order. The divergence of the first diagram can be absorbed into a counter term of the original type ¯ (2. As a further example consider a single scalar field. 2. but the second one needs a new counter term ¯ δg5 (ψψ)5 (2. Giving these facts let us try to understand what makes renormalizable and non renormalizable theories different.3.3.76) Lint = −g2 (ψψ)2 At one loop the theory gives rise to the divergent diagrams of Fig. The theory depends on a finite number of arbitrary parameters equal to the number of parameters gi. As an example consider a fermionic theory with an ¯ interaction of the type (ψψ)2 . as for instance the ones in Fig. At the end the theory will depend on an infinite number of arbitrary parameters.t. Since this term has dimension 6. The most general renormalizable lagrangian density is characterized by two parameters 1 1 L = ∂µ φ∂ µ φ − m2 φ2 − ρφ3 − λφ4 2 2 (2. In the renormalizable case.78) These counter terms originate new one-loop divergent diagrams. = i δgiOi (2.2.75) we can choose the δgi in such a way to cancel.79) This process never ends. the relative coupling has dimension -2 ¯ (2. the divergences. 2.

Since dim g2 = −2. if we consider the scattering ψ + ψ → ψ + ψ in the high energy limit (where we can neglect all the masses). on dimensional ground we get that the total cross-section behaves like 2 σ ≈ g2 E 2 (2. We will show that this does not happen.3.3. But the cross-section has to do with the S matrix which is unitary.2 .. 2. a) b) c) ¯ Fig. 51 . simply requiring that it gives rise to a consistent theory at any energy. being there couplings with negative dimensions. a) b) ¯ ¯ Fig.80) Analogously. But suppose we try to give a meaning to a non renormalizable lagrangian.Divergent diagrams coming from the interaction (ψψ)2 .Divergent diagrams coming from the interactions (ψψ)2 and (ψψ)4 . in any non renormalizable theory.3 . the cross-section will increase with the energy. Consider again a theory with a four-fermion interaction. 2. Since a unitary matrix has eigenvalues of . condition.

modulus 1.4 Dimensional regularization of the Feynman integrals As we have discussed in the previous Section we need a procedure to give sense at the otherwise divergent Feynman diagrams. In fact. This way of regularizing is very convenient because it respects all the symmetries.81) where c is some constant. One kind of regularization which nowadays is very much used is the dimensional regularization. in the spirit of renormalization we have first to perform the perturbative expansion and then take the limit over the cut-off. For the previous example we get g2 E 2 ≤ c This implies a violation of the S matrix unitarity at energies such that E≥ c g2 (2. but only for a limited range of values of the energy. The simplest of these procedures is just ∞ Λ to introduce a cut-off Λ and define our integrals as 0 → limΛ→∞ 0 ). The idea is that integrating on a lower number of dimensions the integral improves the convergence properties in the ultraviolet. For 52 . Although this procedure is very simple it results to be inadequate in situations like gauge theories. It is not difficulty to realize that non renormalizability and bad behaviour of the amplitudes at high energies are strictly connected. it follows that its matrix elements must be bounded. k) (2. k) ≈ p−2 or p−4 .84) with F (p. This consists in considering the integration in an arbitrary number of space-time dimensions and taking the limit of four dimensions at the end. In fact one can show that the cut-off regularization breaks the translational invariance and creates problems in gauge theories. This range is fixed by the value of the non renormalizable coupling. Let us what is dimensional regularization about.83) (2. except for very few cases the symmetries do not depend on the number of space-time dimensions.82) It follows that we can give a meaning also to non renormalizable theories. Translating this argument in the cross-section one gets the bound σ≤ c2 E2 (2. 2. Of course. We want to evaluate integrals of the type I4 (k) = d4 p F (p.

This the Weierstrass representation of the Euler Γ(z). but it is defined also for Re z < 0 where it has simple poles located at z = −n. 53 .89) = n=0 (−1)n αn+z + n! n + z dt e−t tz−1 The second integral converges for any z since α > 0. Then. This complex function is defined for Re z > 0 by the integral representation ∞ Γ(z) = 0 dt e−t tz−1 (2. This expression coincides with the representation for the Γ function for Re z > 0.However it is easy to get a representation valid also for Re z ≤ 0. k) ≈ p−4 . Therefore. k) on the entire complex plane. A simple example of this procedure is given by the Euler’s Γ. if we can define a complex function I (ω. Let us divide the integration region in two parts defined by a parameter α α Γ(z) = 0 dt e−t tz−1 + ∞ α dt e−t tz−1 (2. but defined on a larger domain including the point ω = 2. • Construct an analytic function identical to I(ω. the integral diverges as dt t1+|Re z| (2. However the result does not depend on the particular value of this parameter.85) to be regarded as a function of the complex variable ω. k) is convergent. Notice that in order to isolate the divergences we need to introduce an arbitrary parameter α.instance. then by analytic continuation I and I define the same analytic function. k) (2.87) in the limit t → 0. if F (p. with definite singularity. we would like to introduce a quantity I(ω. Therefore it is a meaningful expression on all the complex plane z. • At the end of the calculation take the limit ω → 2. k) in the domain of convergence. Typically this will be for Re ω < 2.86) If Re z ≤ 0.88) Expanding the exponential in the first integral and integrating term by term we get ∞ Γ(z) = n=0 ∞ (−1)n n! α 0 dt tn+z−1 + ∞ α ∞ α dt e−t tz−1 (2. From this example we see that we need the following three steps • Find a domain where I(ω. and as such that it coincides with I on some common domain. k) = d2ω p F (p. the integral is convergent in 2 and in 3 dimensions.

Since the integrand is invariant under rotations of the N-dimensional vector p.90) Let us consider the integral with N an integer number and p a vector in an Euclidean N-dimensional space.94) The same integral can be evaluated in polar coordinates π By putting x = ρ2 we have 1 π N/2 = SN 2 ∞ 0 N/2 ∞ = SN 0 2 ρN −1 e−ρ dρ (2. we can perform the angular integration by means of dN p = dΩN pN −1 dp where dΩN is the solid angle element in N-dimensions.5 Integration in arbitrary dimensions IN = dN pF (p2 ) (2. 54 .92) The value of the sphere surface can be evaluated by the following trick.97) Γ N 2 and π N/2 IN = Γ N 2 ∞ 0 xN/2−1 F (x)dx (2.95) 1 xN/2−1 e−x dx = SN Γ 2 N 2 (2.98) with x = p2 . Then ∞ (2.93) By taking N of these factors we get IN = 2 2 dx1 · · · dxN e−(x1 + · · · + xN ) = π N/2 (2. with IN = SN 0 pN −1 F (p2 )dp (2. Consider +∞ I= −∞ 2 √ e−x dx = π (2.96) where we have used the representation of the Euler Γ function given in the previous Section.2.91) dΩN = SN . Therefore 2π N/2 SN = (2. Therefore SN the surface of the unit sphere in N-dimensions.

At the end we will have to consider the limit ω → 2.106) This shows that I2ω has simple poles located at ω = A. and we use the definition p2 → −p2 = N p2 .108) . The original integral in Minkowski space is then 1 Γ(A − ω) d2ω p = iπ ω (−1)A (p2 − a2 )A Γ(A) (a2 )A−ω 55 (2.99) with p a vector in a Ndimensional Minkowski space.104) B(x. Then we do a change of variables p0 → ip0 . · · ·. A + 1.107) Therefore our integral will be perfectly defined at all ω such that ω = A.103) 0 and recalling the integral representation for the Euler B(x. and therefore we can extend this formula to complex dimensions N = 2ω I2ω = π ω 1 Γ(A − ω) 2 )A−ω Γ(A) (a (2.100) where IN is an integral of the kind discussed at the beginning of this Section with F (x) given by F (x) = (x + a2 )−A (2. y) function (valid for Rex.105) Γ(A) (a2 )A−N/2 We have obtained this representation for N/2 > 0 and Re(A − N/2) > 0.102) By putting x = a2 y we get IN = (a2 )N/2−A y N/2−1 (1 + y)−A dx (2. A + 1. y) = Γ(x + y) 0 it follows 1 Γ(A − N/2) IN = π N/2 (2. We can perform an anticlockwise rotation of 900 (Wick’s rotation) in the complex plane of p0 without hitting any singularity . obtaining i=1 i IN (M ) =i dN p = i(−1)A IN (−p2 − a2 )A (2. y > 0) ∞ Γ(x)Γ(y) = tx−1 (1 + t)−(x+y) dt (2. But we know how to extend the Euler gamma-function to the entire complex plane.101) It follows π N/2 IN = Γ N 2 π N/2 Γ N 2 ∞ 0 xN/2−1 dx (x + a2 )A ∞ (2.The integrals we will be interested in are of the type (M ) IN = dN p (p2 − a2 + i )A (2. · · · (2.

5772. that is Σ(p). Furthermore. Πµν (q) e Λµ (p. p ). It will be also convenient to introduce a parameter µ such that these quantities have the same dimensions in the space with d = 2ω as in d = 4. by using dimensional regularization.115) (2.111) (2.113) (2. the relevant divergent quantities in QED. then 1 Γ(A − ω) d2ω p = iπ ω (−1)A 2 + 2p · k + k 2 − a2 )A 2 )A−ω (p Γ(A) (a from which d2ω p Γ(A − ω) 1 = iπ ω (−1)A (p2 + 2p · k + b2 )A Γ(A) (k 2 − b2 )A−ω Differentiating with respect to kµ we get various useful relations as d2ω p and pµ pν π ω (−1)A =i (p2 + 2p · k + b2 )A Γ(A)(k 2 − b2 )A−ω 1 × Γ(A − ω)kµ kν − gµν (k 2 − b2 )Γ(A − ω − 1) 2 d2ω p −kµ Γ(A − ω) pµ = iπ ω (−1)A (p2 + 2p · k + b2 )A Γ(A) (k 2 − b2 )A−ω (2. is the Euler-Mascheroni constant.116) 2. Let us put in the previous equation p = p + k and b2 = −a2 + k 2 .109) (2.112) Since at the end of our calculation we will have to take the limit ω → 2. it will be useful to recall the expansion of the Gamma function around its poles Γ( ) = where γ = 0. in order to define the counter terms we will determine the expressions which become divergent in the limit of ω → 2.For the following it will be useful to derive another formula..6 One loop regularization of QED In this Section we will regularize.110) (2.. 56 .114) 1 − γ + O( ) (2. and for (n ≥ 1): Γ(−n + ) = where ψ(n + 1) = 1 + (−1)n 1 + ψ(n + 1) + O( ) n! 1 1 +···+ −γ 2 n (2.

The algebra of the Dirac matrices is also easily extended to arbitrary dimensions d. For instance, starting from [γµ , γν ]+ = 2gµν (2.117) we get γ µ γµ = d and γµ γν γ µ = (2 − d)γν (2.119) Other relations can be obtained by starting from the algebraic properties of the γ-matrices. Let us start with the electron self-energy which we will require to have dimension 1 as in d = 4. From eq. (2.4) we have Σ(p) = iµ4−2ω p−k+m ˆ ˆ 1 d2ω k γµ 2 γµ 2ω 2 − m2 + i (2π) (p − k) k +i (2.120) (2.118)

In order to use the equations of the previous Section it is convenient to combine together the denominators of this expression into a single one. This is done by using a formula due to Feynman 1 = ab which is proven using 1 1 1 1 1 = − = ab b−a a b b−a and doing the change of variables x = az + b(1 − z) We get
1 b a 1 0

dz [az + b(1 − z)]2

(2.121)

dx x2

(2.122)

(2.123)

Σ(p) = iµ4−2ω
0

dz

p−k+m ˆ ˆ d2ω k γµ γµ (2π)2ω [(p − k)2 z − m2 z + k 2 (1 − z)]2

(2.124)

The denominator can be written in the following way [...] = p2 z − m2 z + k 2 − 2p · kz (2.125)

and the term p · k can be eliminated through the following change of variables k = k + pz. We find [...] = (p2 − m2 )z + (k + pz)2 − 2p · (k + pz)z = k − m2 z + p2 z(1 − z) (2.126)
2

57

It follows (we put k = k)
1

Σ(p) = iµ

4−2ω 0

dz

ˆ d2ω k p(1 − z) − k + m ˆ γµ 2 γµ 2ω 2 z + p2 z(1 − z)]2 (2π) [k − m p(1 − z) + m ˆ d2ω k γ γµ 2ω µ [k 2 − m2 z + p2 z(1 − z)]2 (2π)

(2.127)

The linear term in k has vanishing integral, therefore
1

Σ(p) = iµ

4−2ω 0

dz

(2.128)

and integrating over k
1

Σ(p) = iµ By defining

4−2ω 0

dz

Γ(2 − ω) p(1 − z) + m ˆ 1 γµ 2 iπ ω γµ 2ω 2 z(1 − z)]2−ω (2π) Γ(2) [m z − p

(2.129)

= 4 − 2ω we get
1

Σ(p) = −µ
0

dz

p(1 − z) + m ˆ 1 π (4− )/2 Γ( /2)γµ 2 γµ 4− (2π) [m z − p2 z(1 − z)] /2
1

(2.130)

Contracting the γ matrices Σ(p) = − we obtain Σ(p) =
1

1 Γ( /2) 16π 2 1 Γ( /2) 16π 2

dz(4πµ2 ) /2
0

( − 2)ˆ(1 − z) + (4 − )m p 2 z − p2 z(1 − z)] /2 [m

(2.131)

×
0

m2 z − p2 z(1 − z) dz[2ˆ(1 − z) − 4m − (ˆ(1 − z) − m)] p p 4πµ2 m2 z − p2 z(1 − z) 4πµ2

− /2

(2.132)

Defining A = 2ˆ(1 − z) − 4m, p and expanding for → 0 Σ(p) = = = − =
1 2A 1 + 2B − γA 1 − log C dz 2 16π 0 2 1 1 2A dz − A log C + 2B − γA 2 16π 0 1 1 [ˆ − 2m + γ(ˆ − 4m)] p p (ˆ − 4m) − p 2 8π 16π 2 1 1 m2 z − p2 z(1 − z) dz[ˆ(1 − z) − 2m] log p 8π 2 0 4πµ2 1 (ˆ − 4m) + finite terms p 8π 2

B = −ˆ(1 − z) + m, p

C=

(2.133)

(2.134)

58

Consider now the vacuum polarization (see eq. (2.8)) Πµν (q) = iµ4−2ω = iµ4−2ω 1 1 d2ω k γµ γν Tr 2ω ˆ+q−m k−m ˆ (2π) k ˆ 2ω ˆ ˆ ˆ d k T r[γµ(k + m)γν (k + q + m)] (2π)2ω (k 2 − m2 )((k + q)2 − m2 )

(2.135)

Using again the Feynman representation for the denominators
1

Π (q) = iµ

µν

4−2ω 0

dz

ˆ ˆ ˆ T r[γµ(k + m)γν (k + q + m)] d2ω k (2.136) (2π)2ω [(k 2 − m2 )(1 − z) + ((k + q)2 − m2 )z]2

We write the denominator as [...] = k 2 + q 2 z − m2 + 2k · qz (2.137)

through the change of variable k = k − qz we cancel the mixed term obtaining [...] = (k − qz)2 + 2(k − qz) · qz + q 2 z − m2 = k + q 2 z(1 − z) − m2 and therefore (we put again k = k) ˆ ˆ ˆ ˆ d2ω k T r[γµ(k − q z + m)γν (k + q (1 − z) + m)] (2π)2ω [k 2 + q 2 z(1 − z) − m2 ]2 0 (2.139) Since the integral of the odd terms in k is zero, it is enough to evaluate the contribution of the even term to the trace
1 2

(2.138)

Πµν (q) = iµ4−2ω

dz

ˆ ˆ ˆ ˆ T r[...]even = T r[γµ(k − q z + m)γν (k + q (1 − z)]pari + m2 T r[γµγν ] ˆ ˆ ˆ ˆ (2.140) = T r[γµkγν k] − T r[γµ q γν q ]z(1 − z) + m2 T r[γµ γν ] If we define the γ as matrices of dimension 2ω × 2ω we can repeat the standard calculation by obtaining a factor 2ω instead of 4. Therefore T r[...]even = 2ω [2kµ kν − gµν k 2 − (2qµ qν − gµν q 2 )z(1 − z) + m2 gµν ] = 2ω [2kµ kν − 2z(1 − z)(qµ qν − gµν q 2 ) − gµν (k 2 − m2 + q 2 z(1 − z)](2.141) and we find Πµν (q) = iµ4−2ω 2ω 2kµ kν d2ω k (2π)2ω [k 2 + q 2 z(1 − z) − m2 ]2 0 2z(1 − z)(qµ qν − gµν q 2 ) gµν − − 2 2 + q 2 z(1 − z) − m2 ]2 2 z(1 − z) − m2 ] [k [k + q dz 2pµ pν Γ(1 − ω) = −igµν π ω 2 1−ω [p2 − a2 ]2 (a ) 59
1

(2.142)

From the relations of the previous Section we get d2ω p (2.143)

1 Γ(1 − ω) = −iπ ω 2 1−ω (2.144) 2] −a (a ) Therefore the first and the third contribution to the vacuum polarization cancel out and we are left with d2ω p [p2 d2ω k 1 2ω [k 2 + q 2 z(1 − z) − m2 ]2 (2π) 0 (2.145) Notice that the original integral was quadratically divergent, but due to the previous cancellation the divergence is only logarithmic. The reason is again gauge invariance. In fact it is possible to show that this implies q µ Πµν (q) = 0. Performing the momentum integration we have Πµν (q) = −iµ4−2ω 2ω (qµ qν −gµν q 2 ) dz 2z(1−z) iπ ω Γ(2 − ω) 2ω [m2 − q 2 z(1 − z)]2−ω (2π) 0 (2.146) By putting again = 4 − 2ω and expanding the previous expression
1 1

whereas

Π (q) = −i2µ
µν

4−2ω ω

2 (qµ qν − gµν q )
2

dz z(1 − z)

1

Π (q) = 2µ 2 =

µν

2− /2

(qµ qν − gµν q )
2 0

Γ( /2) π 2− /2 dz z(1 − z) 4− [m2 − q 2 z(1 − z)] /2 (2π)
− /2

1 m2 − q 2 z(1 − z) 2 2− /2 2 (qµ qν − gµν q 2 ) dz z(1 − z)Γ( /2) 16π 2 4πµ2 0 2 = (4 − 2 log 2)(qµ qν − gµν q 2 ) 16π 2 1 2 (2.147) dz z(1 − z)( − γ) 1 − log C × 2 0

where C is definite in eq. (2.133). Finally Πµν (q) = 8 1 − 4γ − 4 log 2 (q q − gµν q 2 ) dz z(1 − z) 2 µ ν 8π 1 (qµ qν − gµν q 2 ) = 2π 2 γ m2 − q 2 z(1 − z) 1 − − dz z(1 − z) log × 3 6 2πµ2 1− 2 log C

(2.148)

or, in abbreviated way Πµν (q) = 1 1 (q q − gµν q 2 ) + finite terms 2 µ ν 6π (2.149)

We have now to evaluate Λµ (p , p). From eq. (2.11) we have Λ (p , p) = −iµ
µ 4−2ω

ˆ ˆ ˆ ˆ 1 d2ω k α p − k + m µ p − k + m γ γ γα 2 (2π)2ω (p − k)2 − m2 (p − k)2 − m2 k 60

(2.150)

the general formula to reduce n denominators to a single one is
n

i=1

1 = (n − 1)! ai

1 n 0

δ(1 − n βi ) i=1 dβi [ n βi ai ]n i=1 i=1

(2.151)

To show this equation notice that
n n ∞ n 0 i=1

i=1

1 = ai

− dαi e

αi ai
i=1

(2.152)

Introducing the identity 1=
0

n

dλδ(λ −
i=1

αi )

(2.153)

and changing variables αi = λβi
n n

i=1

1 = ai

∞ n 0 i=1

n

dαi dλδ(λ −
i=1 n

− αi )e
n

αi ai
i=1

∞ n

=
0 i=1

dβi λ

n dλ

λ

−λ e

βi ai
i=1

δ(1 −
i=1

βi )

(2.154)

The integration over βi can be restricted to the interval [0, 1] due to the delta function, and furthermore
∞ 0

(n − 1)! dλλn−1 e−ρλ = ρn

(2.155)

In our case we get Λµ (p , p) = −2iµ4−2ω × d2ω k (2π)2ω
1 1−x

dx
0 0

dy (2.156)

ˆ γ α (ˆ − k + m)γ µ (ˆ − k + mγα p p ˆ [k 2 (1 − x − y) + (p − k)2 x − m2 x + (p − k)2 y − m2 y]3

The denominator is [...] = k 2 − m2 (x + y) + p2 x + p y − 2k · (px + p y) Changing variable, k = k + px + py [...] = (k + px + p y)2 − m2 (x + y) + p2 x + p y − 2(k + px + p y) · (px + p y) 2 (2.158) = k 2 − m2 (x + y) + p2 x(1 − x) + p y(1 − y) − 2p · p xy 61
2 2

(2.157)

161) [m2 (x + y) − p2 x(1 − x) − p 2 y(1 − y) + 2p · p xy]2−ω γ α γλ γµ γσ γα Using the relation γ α γλ γ µ γ λ γα = (2 − d)2 γ µ we obtain ( = 4 − 2ω) Λ(1) (p µ 1 . p) = −2iµ4−2ω µ dx 0 0 dy d2ω k (2π)2ω kλ kσ [k 2 − m2 (x + y) + p2 x(1 − x) + p 2 y(1 − y) − 2p · p xy]3 1 1−x iπ ω (−1)3 1 = −2iµ4−2ω dx dy − Γ(2 − ω) Γ(3) 2 0 0 γ α γλ γ µ γ λ γα [m2 (x + y) − p2 x(1 − x) − p 2 y(1 − y) + 2p · p xy]2−ω ω 1 1−x 1 4−2ω 1 = Γ(2 − ω) dx dy µ 2 4π 0 0 γ α γλ γ µ γ λ γα (2. from the convergent one.162) Γ( /2)( − 2) γµ 2 0 dx 0 dy 1 [m2 (x + y) − p2 x(1 − x) − p 2 y(1 − y) + 2p · p xy] /2 1 1−x 1 2 = dx dy − γ [4 − 4 ]γµ 32π 2 0 0 m2 (x + y) − p2 x(1 − x) − p 2 y(1 − y) + 2p · p xy 4πµ2 1 1−x 1 1 1 = 2 γµ − (γ + 2)γµ − 2 γµ dx dy 8π 16π 2 8π 0 0 62 − /2 . Λµ = Λ(1) + Λ(2) µ µ we get for the first term 1 1−x (2. Λµ .159) The odd term in k is zero after integration.Letting again k → k Λµ (p . Separating the divergent piece. (1) (3) Λµ . p) = −2iµ4−2ω d2ω k (2π)2ω 0 0 ˆ ˆ ˆ p ˆ p γ α (ˆ (1 − y) − px − k + m)γ µ (ˆ(1 − x) − p y − k + m)γα × 2 [k − m2 (x + y) + p2 x(1 − x) + p 2 y(1 − y) − 2p · p xy]3 dx dy 1 1−x (2. p) = µ 2 1 4π 2− /2 1 1−x (2.160) Λ(1) (p . whereas the remaining part is convergent. the term in k 2 is logarithmically divergent.

p) = 2 γµ + Λf (p .165) 2. 2.× log and finally m2 (x + y) − p2 x(1 − x) − p 2 y(1 − y) + 2p · p xy 4πµ2 Λ(1) (p . the effect of Σ(p) is to correct the electron propagator. i i i + ie2 Σ(p) +··· p−m p−m ˆ ˆ p−m ˆ (2.168) µ 8π where the functions with the superscript f represent the finite contributions.7 One loop renormalization 1 (ˆ − 4m) + Σf (p) p 2 8π We summarize here the results of the previous Section Σ(p) = Πµν (q) = (qµ qν − gµν q 2 ) (2.170) .164) (2.2). p) = µ (2. In fact we have (see Fig.7.163) 1 γµ + finite terms 8π 2 In the convergent part we can put directly ω = 2 Λ(2) (p µ 1 1−x i iπ 2(−1)3 . (2. at the same order in the perturbative expansion SF (p) = i p−m ˆ 1+ e2 Σ(p) p−m ˆ 63 −1 = i p − m + e2 Σ(p) ˆ (2.1 .167) 6π 2 1 Λµ (p .7. p) (2.169) Fig.1): SF (p) = .166) 1 + Πf (q) ≡ (qµ qν − gµν q 2 )Π(q 2 ) (2. Let us start discussing the electron self-energy. 2. p) = − 4 dx dy 8π 0 Γ(3) 0 α p ˆ p ˆ γ (ˆ (1 − y) − px + m)γ µ (ˆ(1 − x) − p y + m)γα 2 (x + y) − p2 x(1 − x) − p 2 y(1 − y) + 2p · p xy m 1 1−x 1 dx dy = − 16π 2 0 0 γ α (ˆ (1 − y) − px + m)γ µ (ˆ(1 − x) − p y + m)γα p ˆ p ˆ 2 (x + y) − p2 x(1 − x) − p 2 y(1 − y) + 2p · p xy m (2.The loop expansion for the electron propagator. from which. As shown in eq.

taking into account (L1 )ct .2. is i i Bp − A ˆ ≈ 1− (1 + B)ˆ − (m + A) p p−m ˆ p−m ˆ i i i ≈ + (iB p − iA) ˆ p−m p−m ˆ ˆ p−m ˆ (2. At this order we get SF (p) = i i i + ie2 Σ(p) + iB p − iA ˆ p−m p−m ˆ ˆ p−m ˆ (2. 2. -iA iBp / Fig.174) and the correction to the free propagator is given by e2 Σ(p) + B p − A = ˆ e2 +B p− ˆ 8π 2 e2 m + A + finite terms 2π 2 (2. These can be easily evaluated noticing that the expression of the propagator.3. 2. consistently with our expansion we have taken only the first order terms in A and B. 2.2 .7.171) This allows us to define the counter terms to be added to the lagrangian expressed in terms of the physical parameters in such a way to cancel these divergences ¯ˆ ¯ (L1 )ct = iB ψ ∂ψ − Aψψ (2.176) 64 .7. ˆ ˆ . with contributions −iA to the mass term. We can associate to these two terms the diagrams of Fig.175) We can now fix the counter terms by choosing B=− e2 8π 2 1 + F2 m µ (2.7.173) where. and iB p to p.172) (L1 )ct modifies the Feynman rules adding two particles interacting terms.The counter terms for the self-energy (p in the figure should be read as / p). ˆ The propagator at the second order in the coupling constant is then given by adding the diagrams of Fig.Therefore the effect of the divergent terms is to modify the coefficients of p and m: ˆ −1 iSF (p) e2 = p − m + e Σ(p) = p 1 + 2 ˆ ˆ 8π 2 e2 −m 1+ 2 2π + finite terms (2.

177) with F2 andFm finite for → 0. = Fig. 2.7. given −1 ˆ ˆ iSF (p) ≡ Γ(2) (p) = p − m + B p − A + e2 Σ(p) (2.3 .. However they can be determined by the renormalization conditions. that is by fixing the arbitrary constants appearing in the lagrangian.The second order contributions at the electron propagator. The first one is 0 = Γ(2) (ˆ = m) = p from which e2 Σf (ˆ = m) − p The second condition is e2 e2 (ˆ−4m)+e2 Σf (ˆ = m)− 2 p p 8π 2 8π 1 + F2 p + ˆ me2 2π 2 1 + Fm (2.181) ∂Γ(2) (p) ∂p ˆ (2.180) me2 me2 F2 + Fm = 0 8π 2 2π 2 =1 p=m ˆ (2. the propagator coincides with the ˆ free propagator i . In fact. A=− me2 2π 2 1 + Fm m µ (2. Notice also that these two functions are dimensionless and arbitrary up to this moment.178) we can require that at the physical pole. p = m. for p = m ˆ (2.179) SF (p) ≈ p−m ˆ From here we get two conditions.182) 65 .

7. However. from which Dµν (q) = −igµν −igµλ −igρν + +··· (ie2 )(q λ q ρ − g λρq 2 )Π(q 2 ) 2 2 q q q2 −igµν qµ qν 1 − e2 Π(q 2 ) − i 4 e2 Π(q 2 ) + · · · = (2. In fact one finds some illdefined integral in the infrared region. 2. It follows that we should add to Lp 1 1 1 L2 = − Fµν F µν − (∂µ Aµ )2 = Aµ gµν Aν 4 2 2 (2.The loop expansion for the photon propagator. as the electron mass.183) One should be careful because these particular conditions of renormalization gives some problems related to the zero mass of the photon.4 . As far as the vacuum polarization is concerned. because giving a small to the photon and letting it to zero at the end of the calculations gives rise to finite results.which gives e2 ∂Σf (p) ∂p ˆ p=m ˆ − e2 F2 = 0 8π 2 (2. In this case the parameters comparing in Lp are not the directly measured parameters.4) Dµν (q) = .7. Notice that these conditions of renormalization have the advantage of being expressed directly in terms of the measured parameters.185) 2 q q We see that the one loop propagator has a divergent part in gµν . but they can be correlated to the actual parameters by evaluating some observable quantity. Πµν gives rise to the following correction of the photon propagator (illustrated in Fig.184) Fig. M. 2. −igµν −igµλ 2 λρ −igρν + ie Π (q) 2 + · · · q2 q2 q (2. From this point of view one could avoid the problems mentioned above by choosing a different point of renormalization. Therefore the propagator is not any more in the Feynman gauge.186) 66 . one could renormalize at an arbitrary mass scale. However these are harmless divergences. and also divergent and finite pieces in the term proportional to the momenta.

including the corrections at the first order in C and E is Dµν (q) = and the total propagator Dµν (q) = −igµν qµ qν [1 − e2 Π(q 2 ) − C] − i 4 [e2 Π(q 2 ) + C − E] 2 q q (2. since this choice does not change the physics. and evaluate the corresponding Feynman rules. This is because the terms proportional to qµ qν . and then expand in the counter terms.189) The free propagator. + C) β=− i C−E 4 (1 + C)(1 + E) q (2. we can look at these two contributions as perturbations to the free lagrangian.190) (2. as it follows from the gauge invariance.188) C E (L2 )ct = − Fµν F µν − (∂µ Aµ )2 = −C 4 2 1 1 Fµν F µν + (∂µ Aµ )2 4 2 − We solve this equation by putting Dµν (q) = α(q 2 )gµν + β(q 2 )qµ qν Substituting in the previous equation we determine α e β. We have Dµν (q) = qµ qν −igµν −igµν qµ qν ˜ ˜ 1 − F3 − e2 Πf − i 4 F3 + e2 Πf ≡ [1 − Π] − i 4 Π 2 2 q q q q (2. The effect of these two terms to the equation which defines the propagator in momentum space is λ [(1 + C)q 2 gµν − (C − E)qµ qν ]D νλ (q) = −igµ (2. if we have a vertex with a virtual photon (that is the vertex is connected to an internal photon line) and two external electrons.194) u(p )γµ u(p). bay taking into account the Dirac equation.191) We can now choose E = 0 and cancel the divergence by the choice C=− e2 + F3 6π 2 m µ (2.187) As for the electron propagator. ¯ The result is zero. The result is α=− q 2 (1 i .193) In fact we are free to choose the finite term of the gauge fixing.192) −igµν qµ qν (1 − C) − i 4 (C − E) q2 q (2. For instance.195) 67 . the term proportional to qµ qν is saturated with q =p −p (2. and the conservation of the electromagnetic current.the two counterterms E −C (∂µ Aµ )2 2 (2. Let us see what the full propagator says about the mass of the photon. or evaluate the effect on the propagator.

197) ˜ and we see that the propagator has a pole at q 2 = 0. Therefore this is modified by a constant term in momentum space. or by a term proportional to the delta function in the real space ∆12 = e2 dt e2 e4 e2 d4 q −iq(x1 − x2 ) 1 =− − e − δ 3 (r) (2π)4 q 2 60π 2m2 4πr 60π 2 m2 (2. that is ˜ Π(0) = 0 (2. the propagator has the free form.203) The first term. An exception is the case of the anomalous symmetries. which are symmetries at the classical level. Also in this case we will require that at the physical pole.198) from which F3 = −e2 Πf (0) For small momenta we can write dΠ (q) ˜ Π(q) ≈ e2 q 2 dq 2 f q 2 =0 (2.200) since F3 is a constant.204) 68 . Therefore the photon remains massless after renormalization. This is part of a rather general aspect of renormalization which says that if the regularization procedure respects the symmetries of the original lagrangian. gives rise to the Coulomb potential. we get Πf (q) ≈ − and ˜ Π(q) ≈ from which Dµν 1 2π 2 γ 1 + log 6 6 m2 2πµ2 + 1 2π 2 1 dz z 2 (1 − z)2 0 q2 +··· m2 (2. since Π(q 2 ) is finite for q 2 → 0. 1/q 2 . but are broken by the quantum corrections. Using the expression for Πf from the previous Section.202) e2 q 2 gµν + gauge terms = −i 2 1 − q 60π 2 m2 (2. e2 /4πr.where ˜ Π = e2 Πf + F3 (2.196) At the second order in the electric charge we can write the propagator in the form Dµν (q) = −i qµ qν ˜ gµν + 2 Π(q) ˜ q q 2 (1 + Π(q)) (2. All the anomalous symmetries can be easily identified in a given theory.201) e2 q 2 60π 2 m2 (2. q 2 → 0.199) (2. the symmetries are preserved at any perturbative order.

The one-loop vertex is then F2 (Λµ ) = γµ + e2 (Λf − 2 γµ ) (2. which produces a splitting of the levels 2S1/2 and 2P1/2 .205) p is ¯ (L3 )ct = −eDψγµ ψAµ The complete vertex is given by (see Fig.7. The contribution we have just calculated is only −27.206) (counter-term) Fig. 2. 2. and this is proportional to γµ . The counter term to add to interacting part of Lp ¯ Lint = −eψγµ ψAµ (2. This equality follows from the conservation of the current. We have now to discuss the vertex corrections.5 . We fix the counter term by D=− e2 1 + F2 8π 2 (2. and turns out to be about 1057.208) with F2 the same as in eq. (2.7). but it is important since the agreement between experiment and theory is of the order of 0.210) ˆ 69 .1 MHz. and it is one of the contributions to the Lamb shift. and therefore we can use the arbitrariness in the finite part of the vertex counter term to guarantee it.7.207) (2.209) µ 8π One can see that this choice of F2 is such that for spinors on shell the vertex is the free one u(p)(Λµ ) u(p)|p=m = u(p)γ µ u(p) ¯ ¯ (2. The reason is that with this choice we get the equality of the wave function renormalization factors Z1 = Z2 .The one loop vertex corrections. We have seen that the divergent (1) contribution is Λµ (p .5) −ie γµ + e2 Λµ + Dγµ .1 MHz.9 MHz. (2. p).This modification of the Coulomb potential modifies the energy levels of the hydrogen atom. The total Lamb shift is the sum of all the self-energy and vertex corrections.

217) Using pγα = −γα p + 2pα. Here g is the gyromagnetic ratio.219) 70 α .218) (2.213) (2.211) 2m 2m con q = p − p. the denominator of Λµ is given by [.. let us define ¯ p ˆ p ˆ Vµ = u(p )γ α [ˆ (1 − y) − px + m] γ µ [ˆ(1 − x) − p y + m] γα u(p) (2. it is enough to evaluate the matrix element e2 u(p )Λ(2) (p . which the Dirac equation predicts to be equal to be two. and in the previous limit they have to build up the free vertex.212) An analogous operation on the barred spinor leads to the result.215) To evaluate the correction to this term from the one loop diagrams. obtaining ˆ ¯ ˆ ˆ Vµ = u(p ) myγ α + 2(1 − y)p − γ α px γ µ [mxγα + 2(1 − x)pα − p yγα] u(p) (2.214) (2.. and an analogous equation for p . we can bring p to act on ˆ ˆ ˆ ˆ the spinor at the right of the expression. We observe also that the Gordon identity shows immediately that g = 2. To this end we first need to prove the Gordon identity for the current of a Dirac particle pµ + pµ i + σµν q ν u(p) u(p )γµ u(p) = u(p ) ¯ ¯ (2. p)u(p) ¯ µ (1) (2. For momenta (2) on shell. Therefore we will ignore all the terms proportional to γµ and we will take the first order in the momentum q. For the proof we start from pγµ u(p) = (−mγµ + 2pµ )u(p) ˆ and γµ pu(p) = mγµ ˆ Subtracting these two expressions we obtain γµ u(p) = pµ i − σµν pν u(p) m m (2.216) in the limit p → p and for on-shell momenta. as implied by the renormalization condition. because it implies that the coupling with the electromagnetic field is just e σµν F µν (q) 2m (2. In fact Λµ contributes only to the terms in γµ .] = m2 (x + y) − m2 x(1 − x) − m2 y(1 − y) + 2m2 xy = m2 (x + y)2 In order to evaluate the numerator. and p on the spinor at the left.We will evaluate now the radiative corrections to the g − 2 of the electron.

p)u(p) → − dx dy u(p ) pµ ¯ 16π 2 0 m (x + y)2 0 2 µ (x − xy − y ) u(p) (2.Making use of γα γ µ γα = −2γ µ γα γµ γν γ α = 4gµν ˆ ˆ p ˆ γα pγ µ p γ α = −2ˆ γ µ p we get Vµ = u(p ) − 2m2 xyγ µ + 2my(1 − x)(−mγ µ + 2pµ ) ¯ − 4my 2p + 2mx(1 − y)(−mγ µ + 2p ) + 4(1 − x)(1 − y)m2 γ µ − 2y(1 − y)m2 γ µ − 4mx2 pµ − 2m2 x(1 − x)γ µ − 2xym2 γ µ u(p) (2.221) (2. ie2 u(p )σµν q ν u(p) ¯ → 2m 16π (2.222) (2. and eliminating the further contribution in γµ we obtain ¯ e2 u(p )Λ(2) (p .225) + p (x + y)2 By changing variable z = x + y we have 1 1−x dx 0 1 0 dy y − xy − x2 = (x + y)2 1 1 dx 0 x dz 1−x x − 2 z z = 0 [−(1 − x) log x + x − 1] x2 x2 x2 log x − + −x 2 4 2 1 = −x log x + x + from which = 0 1 4 (2.223) and for the piece which does not contain γµ Vµ = 4m¯(p ) pµ (y − xy − x2 ) + p (x − xy − y 2) u(p) u Therefore the relevant part of the vertex contribution is e u(p ¯ 2 µ µ µ (2. magn. p)u(p)|mom.220) (2.229) . which coincides with the free vertex u(p )[γµ + ¯ pµ + pµ i ie2 σµν q ν ]u(p) = u(p )[ + ¯ 16π 2 m 2m 2m 71 1+ e2 8π 2 σµν q ν ]u(p) (2.227) 16π 2 m Using the Gordon in this expression. p)u(p) → − µ ¯ e u(p 2 )Λ(2) (p µ .224) )Λ(2) (p µ 1 1−x e2 4 (y − xy − x2 ) .228) Finally we have to add this correction to the vertex part taken at p = p.226) e2 u(p )[pµ + p µ ]u(p) ¯ (2.

32848 2 2π π aexp = (1159652.2) × 10−9 (2.49( )3 + · · · = (1159652.233) 72 .4) × 10−9 π (2. we get g α =1+ + O(α2 ) 2 2π (2.232) to be compared with the experimental value 2 α 1 1α ath = (g − 2) = − 0.4 ± 0.230) This correction was evaluated by Schwinger in 1948.231) (2.4 ± 0.Therefore the correction is e e α → 1+ 2m 2m 2π Recalling that g is the ratio between S · B and e/2m. Actually we know the first three terms of the expansion α + 1.

1) 2. we are not sure that the assumptions we are going to make are consistent.Chapter 3 Vacuum expectation values and the S matrix 3. Such a formulation is called LSZ (from the authors l (Lehaman.and Out-states For the future we will need a formulation of field theory showing that any S-matrix element can be reduced to the evaluation of the vacuum expectation value of a T product of field operators. It is interesting to see how far one can go in determining the features of a field theory by using only invariance arguments. It should be also kept in mind that since no exact solution of an interacting field theory is known. Said that. This is called the vacuum state Φ0 ≡ |0 By convention.217) by sending the free hamiltonian into the full hamiltonian. (3. Formally this representation can be obtained from eq. we assume that the corresponding energy is zero P0 |0 = 0 From eq. There exists a nondegenerate Lorentz invariant states of minimum energy.1) we get also P|0 = 0 → Pµ |0 = 0 73 (3. that is the operators will be time-dependent and evolving with the full hamiltonian. This formalism is based on a series of very general properties as displacement and Lorentz invariance. Symanzik and Zimmermann). (1. In the following we will work in the Heisenberg picture.1 In.2) .4) (3.3) (3. p0 ≥ 0 (3. our assumptions are: 1. The eigenvalues of the four-momentum lie within the forward light cone p2 = pµ pµ ≥ 0.

Ideally these experiments are realized by preparing at 74 ep ar tic le sta te . assuming that the interactions do not modify the spectrum in a violent way. Remember that the pion is not a stable particle and decays into µ + νµ with ¯ −8 a lifetime of the order of ≈ 2 · 10 sec. but this number is very large with respect to the natural decay frequency 1/mπ ≈ 5·10−24 sec.5) where Φ1 is a momentum eigenstate with eigenvalue pµ such that p2 = m2 .Since pµ = 0 is a Lorentz invariant condition. Notice that this assumption is very much in the spirit of perturbation theory. For instance. 4. xxx Continuum xxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xx sin gl E 2m xx xx m p xx xx Vacuum Fig. by neglecting the weak interactions (responsible for the decay) we can assume that the pion is stable. The same argument can be applied to quarks and leptons that will be the main actors in the following.1. However we mentioned also the pions. 3. the case of the π mesons is given in Fig. The main tools of investigation in the field of elementary particle physics are the scattering experiments. it follows that Φ0 appears as the vacuum states in all the Lorentz frames.1 . Neglecting the complications from the zero mass states we can add a further requirement: the vacuum and the single particle states give a discrete spectrum in pµ . there exists a stable stae of single particle Φ1 ≡ |p (3. That is we introduce a field for each particle. 3. For each stable particle of mass m. Therefore. to say that the arguments presented here can be applied (in the appropriate range of energy) also to composite particles.. 3. . (mπ ≈ 140 MeV ).Energy-momentum spectrum for particles of mass m.

The interaction time will be of the order of the time necessary to cross the atomic dimensions for the case of a single scattering.6) 2. the particle will travel since it will reach the target where it will be subject to the interaction with the atoms of target itself. In particular we will require 1. or the Dirac free equation for a spinor field. In particular. imagine an electron prepared in an eigenstate of momentum p. After that we will let again the electron to move freely since it will reach the detector where we will measure its properties as momentum and polarization. For instance. the scattering process will occur around the time t = 0 (assuming this as the time scale) and then we will measure the outgoing particles at a time T .) corresponding to the physical mass m ( + m2 )φin (x) = 0 (3. In conclusion we can think to describe the particles at t = ±∞ in terms of free fields. φin (x) must satisfy the free Klein-Gordon equation (for a scalar field. with respect to translation it must satisfy [Pµ .8) 75 .and out-fields describe correctly the incoming and outgoing free particles we have to require some properties. etc. φin (x) must transform as the corresponding φ(x) with respect to the symmetries of the theory.7) ∂φin (x) ∂xµ (3. at maximum. After that scattering process we imagine that at t = +∞ the emerging particles will behave again as free particles. The interacting field φ(x)) can be thought of as constructed in terms of these free fields operators and such that at t → ±∞ it reduces to φout (x) or φin (x). cannot be neglected. that is one has always to remember that the experiments is taking place between to finite times −T and T . That is we prepare the beam at some finite time −T .t = −∞ a system of free particles. impinging over an atom. The ideal description considered before can be considered as correct if the interaction time ∆t around t = 0 is much smaller that T . Having prepared the electron in the initial state at the time −T . The same consideration holds for the energy. of the order of the thickness of the target for a multiple scattering. as we know. In real life the times t = ±∞ correspond to laboratory time. These particles will interact with a target at some finite time. The free fields that we are going to use will be denoted by φin (x) for t → −∞ and by φout (x) for t → +∞. Here there is a further simplification that we are doing since in a finite time or a finite volume we cannot prepare a state of definite momentum. That is assuming that they do not interact each other (this requires that the particles are well separated in space). This is one of the reasons why analyzing a scattering process it is always convenient to quantize in a box of the dimension of the experiment itself (that is about the distance between the source of the beam and the detector). or. ] = −i equivalent to require φin(x + a) = eiP · a φin (x)e−iP · a (3. except that they will be subject to the self-interaction which. In order that the in.

We have ∂ −i µ p|φin(x)|0 = p|[Pµ . ain (k)] = = − ∗ d3 x fk (x) ∗ ∗ d3 x fk (x)∂0 φin(x) − ∂0 fk (x)φin (x) (3. (2π)3 2ωk ωk = k 2 + m2 (3. we find ain (k) = i from which [Pi . we see that the operators defined in eq.14) ∂ ∂φin (x) ∂x0 ∂xi − ∂ ∗ f (x) ∂x0 k ∂φin (x) ∂xi φin (x) (3.13) Inverting this relation. This follows immediately. From our starting hypotheses we have that this is the state of single particle with mass m. By using these two properties it follows at once that the in-field creates from the vacuum the physical one particle state from the vacuum.and out-fields are free fields we can apply all the corresponding formalism.9) ∂x By applying −i∂/∂xµ once more. let |p an eigenstate ofPµ with eigenvalue pµ . In particular they can be Fourier expanded φin(x) = with fk (x) = ∗ d3 k ain (k)fk (x) + a† (k)fk (x) in (3. In fact. we find − p|φin (x)|0 = p2 p|φin(x)|0 (3. The same properties are required for the out-fields.12) 1 e−ikx .14) are time independent. by 76 . (3. Since the in.15) d3 x ∗ ∂fk (x) ∂xi ∂ φin (x) − ∂x0 ∗ ∂ ∂fk (x) ∂x0 ∂xi = −ki ain (k) ∗ Since fk (x) and φin (x) both satisfy the free Klein-Gordon equation.10) and using the Klein-Gordon equation (p2 − m2 ) p|φin(x)|0 = 0 (3.11) We see that the only state that φin(x) can create out of the vacuum is the one with pµ such that p2 = m2 .Notice that this requirement implies that we have taken into account the selfinteractions of the field which are responsible for the mass renormalization (we will mention the wave function renormalization later). φin(x)]|0 = pµ p|φin(x)|0 (3.

· · · . otherwise. and therefore the charge Q= d3 xj0 (3. it follows [P0 . · · · . this state would have a negative eigenvalue of the energy. We get also (3. a† (k)] = kµ a† (k) in in (3. ain (k1 )a† (k2 )]|0 = (k2 − k1 )µ ain (k1 )a† (k2 )|0 in in in (3.25) where j(x) (the scalar current) contains the self-interactions of the fields and possible interactions with other fields in the theory. ain (k)] = −kµ ain (k) Also.21) (3.20) It follows that a† (k) creates states of four-momentum kµ whereas ain (k) annihilates in states with momentum kµ . kN . kN ).16) ∂x ∂x is a conserved current. pM ) is identical to the set (k1 . Pµ a† (k1 )a† (k2 )|0 = [Pµ . We need to solve this equation with the 77 .18) (3. Let us now look for a relation between the in.19) dx ∗ ∂fk (x) ∂x0 ∂ φin (x) − ∂x0 ∗ ∂ ∂fk (x) ∂x0 ∂x0 = −k0 ain (k) In conclusion [Pµ . For instance. a† (k1 )a† (k2 )]|0 = (k1 + k2 )µ a† (k1 )a† (k2 )|0 in in in in in in Pµ ain (k1 )a† (k2 )|0 = [Pµ . pM . ain (k)] = = − ∗ d3 x fk (x) 3 ∂ ∂φin (x) ∂x0 ∂x0 − ∂ ∗ f (x) ∂x0 k ∂φin (x) ∂x0 φin (x) (3. The field φ(x) will satisfy an equation of the type ( + m2 )φ(x) = j(x) (3. then the four-vector ∂φ1 ∂φ2 jµ = φ1 µ − µ φ2 (3. · · · .recalling that if φ1 (x) and φ2 (x) are both solutions of the Klein-Gordon equation.and the φ-field. by hermitian conjugation we get [Pµ . in|k1 .17) is a constant of motion. Using this.22) By using the condition on the spectrum of the momentum operator we get also ain (k)|0 = 0 (3.24) p1 . in = 0 unless M = N and the set (p1 . · · · .23) since.

Notice that in writing the previous equation we have included the effects of mass renormalization inside the current j(x). The function ∆ret (x. t) an arbitrary normalizable solution of the Klein-Gordon equation ( + m2 )f (x) = 0 Then the asymptotic condition is formulated as √ lim α|φf (t)|β = Z lim α|φf (t)in |β t→−∞ t→−∞ (3. see eq. the required asymptotic conditions are √ √ lim φ(x) = lim Zφin(x). m2 )j(y) (3.6)).boundary conditions at t = ±∞ relating φ(x) with the in. for x0 < 0 (3. m2 ) = 0. but we know that at the same time we have also a wave function renormalization.31) (3. t)) (3. both 1|φin(x)|0 e 1|φ(x)|0 have the same x-dependence (they must be both proportional to exp(ipx). This is because it is not possible to isolate the operator φ(x) from the scalar current j(x) at t = −∞. In fact.26). m2 ) is the retarded Green’s function.and out-fields but it will differ by a factor Z. (3. m2 ) = δ 4 (x) such that ∆ret (x. Therefore asymptotically the field φ(x) cannot be the √ √ same as the in.32) where |α and |β are any two normalizable states. but φ(x) may create out of the vacuum states other than the single particle state.and out-fields. It can be shown that the asymptotic condition (3. t) − ∂0 f ∗ (x. lim φ(x) = lim Zφout (x) (3. where Z is the wave function renormalization. since this includes also the self-interactions. However the statement can be given in a weak form for the matrix elements.29) The need of Z in this context. cannot hold as an operator statement.30) (3.26) t→−∞ t→−∞ t→+∞ t→+∞ The solution for φ(x) is then φ(x) = √ Zφin (x) + d4 y∆ret (x − y. Therefore. defined by ( + m2 )∆ret (x. is because we want the the in-field properly normalized in order to produce single particle states.28) with f (x. The correct way of doing it is first to smear out the field over a space-like region φf (t) = i d3 x (f ∗ (x. t)φ(x. and therefore the normalization of the two previous matrix elements cannot be the same. t)∂0 φ(x.27) We will see in a moment the meaning of Z in this context. 78 .

in in in (3.35) + m2 )∆adv (x.27) we find √ √ Zφin (x) = Zφout (x) + d4 y∆(x − y. m2 ) − ∆ret (x − y. After the scattering process we are interested to evaluate the probability amplitude for the final state to be a state of non-interacting particles described by the outoperators |p1 . 3.38) where α stands for the all collection of the quantum numbers of the initial state. · · · . out| such that Sβα = β. out = a† (p1 ) · · · a† (pn )|0. in (3. m2 ) (3. pn . In particular we can find a relation between the in.Similar considerations can be made for the out-field. pn . in|S|α. m2)j(y) (3. out ≡ |β.2 The S matrix We have shown in the previous Section that an initial state of n non-interacting particles can be described by |p1 . m2 ) = δ 4 (x) √ Zφout (x) + d4 y∆adv (x − y.34) (3. m2 ) = ∆adv (x − y. ( for x0 > 0 (3. in|S = β. m2 ) = 0. out (3.42) From the definition we can derive some important properties for the operator S (3. in (3.37) The construction of the out-operators of creation and annihilation goes along the same lines discussed for the in-operators.39) out out The probability amplitude is given by the S matrix elements Sβα = β. m2 )j(y) (3. out|α. in ≡ |α.40) We can also define an operator S transforming the in-states into the out-states β.36) with ∆(x − y.33) is the advanced Green’s function. · · · . (3.and out-fields φ(x) = where ∆adv (x. By comparing this expression with eq.41) 79 . in = a† (p1 ) · · · a† (pn )|0.

out|φout = β. in = |p.51) (3.53) 5. The operator S maps the in. out|φout |α. in|Sφout |α. in which shows the relation (3. as it follows from the fact that transform in. (3.45) (3.to the out-fields φin(x) = Sφout (x)S −1 In fact.41) implying S † |α.43) By choosing the phase equal to zero we get |0.44) From eq.46). and therefore 0in|S = 0out| = eiφ 0in| (3. S is Lorentz invariant. in|φin(x)S|α. out from which showing that β. out|φout is an out-state and therefore β. For the same reason the S matrix is invariant under any symmetry of the theory. in = β. out = |0. in|φin S Then β. in = β.50) (3. in = β. 80 . out|φout (x)|α.1. 4.49) (3. in ≡ |0 2.46) (3. This follows from the very definition (3. out|α. in = |α.47) (3. in|SS †|α. out = S|α. it follows |S00 | = 1. in = β. let us consider the following matrix element β. in = S † |α.50) and the unitarity of S we get |α. In an analogous way the stability of the one particle state requires |p. out = δαβ SS † = 1 (3.into outfields having the same Lorentz properties. in But β. Since the vacuum is stable and not degenerate.52) (3.48) (3. out 3. out −1 (3. The S is unitary. in|Sφout (x)|α.

in out out in β − p. it is possible to extract the particle of momentum pµ from the state by introducing the corresponding field operator. in − ∂x0 = ∂ β. out|α. Let us consider the following identities β. if present in the state. p. in ∂ β.αp = β. in in β. in ∂x0 (3. As we know the integral appearing in the previous expression is time-independent since fp (x). Let us start considering the following S matrix element Sβ. out|α. We have denoted |β − p. out|α. in β − p. in (3. out|a† (p)|α. p. p. out the outstate obtained by removing the particle with momentum pµ . by using the asymptotic condition. in + β. p. (3.54) where |α. Therefore we are allowed to make the following substitutions (in the weak sense. out|a† (p) − a† (p)|α. out|φin(x) − φout (x)|α. out|a† (p)|α. Here we will consider the case of scalar particles. in is the state of a set of particles α plus an additional incoming particle of momentum pµ . but the formalism can be easily extended to other cases.14). out|φin(x) − φout (x)|α.3 The reduction formalism In this Section we will describe a systematic way of evaluating the S matrix elements in terms of the vacuum expectation values of T -products of field operators.55) ∂x0 = = = where we have used the adjoint of eq. φin (x) and φout (x) are solutions of the Klein-Gordon equation.57) By using the identity +∞ d4 x −∞ ∂ g1 (x) ∂x0 ∂ g2 (x) − ∂x0 81 ∂ g1 (x) g2 (x) = ∂x0 . in i d3 x fp (x) lim − 0lim + √ x0 →+∞ x →−∞ Z ∂ fp (x) β.56) obtaining β. in (3. otherwise it must be taken as the null vector. out|α. out|α. in β. out|φ(x)|α.3. out|φ(x)|α. in − i d3 x fp (x) ∂x0 ∂ − fp (x) β. We want to show that. since we are taking matrix elements) 1 lim φin(x) = 0lim √ φ(x) x →−∞ Z 1 lim φout (x) = 0lim √ φ(x) x0 →+∞ x →+∞ Z x0 →−∞ (3.

in i ¨ ¨ + √ d4 x β. This exercise is interesting because it shows how the T -product comes about. We want to extract this particle from the state β. in 1 lim − lim γ.and out-particles from the states. p. p. out|φ(x)|α. out|α. out|φ(x)|α. in ∂ γ.61) We can iterate this procedure in order to extract all the in.63) . in = √ Z y0 →+∞ y0 →−∞ 82 (3. with p the momentum of a single particle. in = β − p. Let us see also the case in which we want to extract from the previous matrix element an out-particle. in (3. in γ. out| fp (x)φ(x) − (∇2 − m2 )fp (x) φ(x) |α. in i d3 y fp ∗ (y) ∂y 0 ∂ ∗ f (y) γ. in γ.60) (3. in = β − p. out|(φout (y)φ(x) − φ(x)φin(y))|α. out|φ(x)|α − p .58) d4 x [g1 (x)¨2 (x) − g1 (x)g2 (x)] g ¨ we can write β. out|α. out|aout (p )φ(x)|α.= = x0 →+∞ +∞ −∞ lim − 0lim x →−∞ d3 x g1 (x) ∂ g2 (x) − ∂x0 ∂ g1 (x) g2 (x) ∂x0 (3. in i ¨ d4 x β. in (3. out|α. out|(φout (y)φ(x) − φ(x)φin(y))|α. p . out|T (φ(y)φ(x))|α. Let us suppose that the out state is of the type β = (γ. out| fp (x)φ(x) − fp (x)φ(x) |α. out|α. in = γ. in i + √ d4 xfp (x)( + m2 ) β. p. in Z (3. out|φ(x)|α. in β − p. out|α. in + γ. out|φ(x)ain(p )|α. in = γ.59) Z = By using the Klein-Gordon equation for fp (x) we get β. in + √ Z Integrating by parts β. out|[φout (y)φ(x) − φ(x)φin (y)]|α. p ). out|[aout (p )φ(x) − φ(x)ain (p )]|α.62) ∂y 0 p = = + − Using again the time independence of the integral we have γ. out|α.

out|φ(x)|α − p . in ∂y 0 p γ. pm . we can reduce further the matrix elements of the type β. out|φ(x)|α. out|α. pn ) with the condition pi = qj we get p1 .and eq. · · · . for spinor fields. √ the Dirac case one In √ √ halso to change i/ Z to −i/ Z2 for fermions and to +i/ Z2 for antifermions. Therefore we have shown that it is possible to evaluate all the S matrix elements once one knows the vacuum expectation values of the T -products (called also the n-point Green’s function). out|α.66) 0|T (φ(y1) · · · φ(yn )φ(x1 ) · · · φ(xm ))|0 If some of the qi ’s are equal to some of the pj ’s. · · · . in (3. in . (3. out|α−p. · · · . out|T (φ(y)φ(x))|α.58) β. in = γ. p. in i d4 xfp (x)( + m2 )x γ. out|T (φ(y)φ(x))|α. The result obtained here can be easily extended to generic fields simply by substituting to the wave functions fp (x) the corresponding solutions of the wave equation (for instance. qm ) and in (p1 . out|q1 . in = i √ Z m+n m. in i ∗ √ (3.64) d4 yfp (y)( + m2 )y γ. out|T (φ(y)φ(x))|α. out|φ(x)|α − p . out|φ(x)|α − p . in = β − p. 83 . out|φ(x)|α − p . · · · . qn .61) β.j=1 + m2 )xi ( + m2 )yj · (3. out| fp (y) 2 T (φ(y)φ(x)) − fp (y)T (φ(y)φ(x)) |α. in d3 y fp (y) lim − 0lim 0 →+∞ y →−∞ ∂y 0 Z y ∂ ∗ f (y) γ. in ∂y0 Z γ. in + √ Z 2 i √ + d4 xd4 y Z ∗ fp (x)fp (y)( + m2 )x ( + m2 )y γ. out|T (φ(y)φ(x))|α. where Z2 is the wave function renormalization for the spinor field. in ∂2 i ∗ ¨∗ √ d4 y γ.n ∗ d4 xi d4 yj fqi (xi )fpj (yj )( i. in ∂ i ∗ √ γ.65) More generally if we want to remove all the particles in (q1 . spinors times plane waves) and to the Klein-Gordon operator the corresponding wave / operators (for the spinor case the Dirac operator (i∂ − m)). in Z + − = + = + Substituting inside the (3.

We will show in the following that a simple technique to evaluate perturbatively the vacuum expectation value of n fields. a n-point Green’s function. can be obtained in the context of the path-integral formulation of quantum mechanics. 84 .

1) for ∆t infinitesimal. although using a mathematical formalism which was far from being well defined. and we will see later how this can be extended also to infinite degrees of freedom as needed in quantum field theory. This formulation did not make use of operators and of Schr¨dinger o equation. but rather it was giving an explicit expression for the quantum-mechanical amplitude.1 Feynman’s formulation of quantum mechanics In 1948 Feynman gave a formulation of quantum mechanics quite different from the standard one. t + ∆t|q. which was obtained directly from the physical idea that the probability amplitude to go from one state to another can be obtained by summing together all the possible ways with an appropriate weight. at the beginning of the 70’s it turned out to be a crucial tool for the quantization of gauge theories. the probability amplitude q . First of all we notice that the quantum mechanical problem is to evaluate 85 . Also it is one the few methods which can be used for studying situations outside the perturbative approach. for one degree of freedom. Although the path integral formulation (as the Feynman’s approach is usually called) is just a different way of formulating quantum mechanics. The key observation to get the appropriate weight was taken by the Dirac’s book where it was shown that.Chapter 4 Path integral formulation of quantum mechanics 4. to the probability amplitudes (remember that the probability is the square of the amplitude). is given by the eiS where S is the classical action to go from q at the time t to q at the tome t + ∆t. In fact we will show that the Feynman’s formulation can be derived from the usual formulation of quantum mechanics. This was as apply the ordinary composition laws in probability theory. t (4. In order to derive the Feynman’s formulation we will start studying a quantum system described by one degree of freedom.

0 e−iH+ (qk+1 . p). let us use again the completeness in momentum space with states taken ˜ at a time t in between tk e tk+1 : qk+1 . for instance in configuration space. tk + |qk . t where tk = t + k . tn−1 |qn−2 . but now for an infinitesimal time interval . tk + |p.2) It will be convenient to work in the Heisenberg representation. Therefore q .6) dqk k=1 k=0 (4. n−1 · · · q1 . t = n→∞ lim dq1 · · · dqn−1 q .3) It follows immediately that the relation between these states at different times is given by |q. By doing so we get ˜ qk+1 .the time evolution. t ≥t (4.4) for a time independent hamiltonian. 0 86 (4.10) . 0|e−iH(p. tk + |p. t = eiHt |q. p)(tk + − t) ˜ qk+1. The matrix element (4. 0|p. its matrix elements are given by q . t = qk+1 . t1 |q.9) (remember / = 1). t|qk . t q (4.2). we get q . tk (4. tn−2 · (4. which is given by the propagator. tk + |p. This can be evaluated in any basis.2) can be evaluated by slicing the time interval t − t in infinitesimal pieces and using the completeness relation. t |q.7) We have still to evaluate the matrix element (4.5) t0 ≡ t. q)(tk + − t) |p. The result will be denoted by the symbol H+ (q. t |qn−1 . t . tk = ˜ ˜ dp qk+1.8) The mixed matrix elements can be written as ˜ ˜ qk+1 . tk+1 |qk . that is q(t)|q. The states of the previous equation are defined as eigenstates of the operator q(t) at the time t. the hamiltonian can be written in such a way that all the operators q are at the left of all the operators p. t p. t ˜ 1 − iH+ (qk+1 . 0|p. t = lim n→∞ n−1 tn ≡ t qk+1 . 0 (4. making use of the canonical commutation relations for p and h q. that is using timedependent operators q(t). t |q. 0 (4. t |q. t = |q. p)(tk + − t) qk+1. tk (4. tn−1 qn−1 .

11) (4. p)) e 2π (4. p)) 2 → 0 we have qk+1 .The function H+ (q. p) is obtained by substituting in H+ (q. tk + |qk .18) q . p) are the same operator. tk where ˙ L(qk . p)(tk+1 − t)) 2π 1 ˜ √ ei(−pqk − H− (qk . qk . p) e 2π (4. p) and H+ (q. tk + |p. in general. p) ≡ H(qk . qk . tk 1 ˜ √ ei(pqk+1 − H+ (qk+1 . Notice that although H(q. p) = pqk − H(qk . Coming back to (4. Therefore qk+1 . t |q. In this way we get ˜ qk+1.7) and using (4. t = lim n→∞ dqk k=1 dpk 2π k=0 i L(qk . p) (4. p) the operators q and p with their eigenvalues. t where dp i (p(qk+1 − qk )/ − Hc (qk+1 . p) is nothing but the hamiltonian of the system evaluated in the phase space. H(q. tk → δ(qk+1 − qk ) (4. t and ˜ p. p) are. qk . p)(t − tk )) 2π (4.12) where H− is the eigenvalue of the operator H− (p. p) + H− (qk .16) dp i L(qk .15) it is reasonable to define a variable velocity as qk = ˙ and write Hc (qk+1 . tk + |qk .18) we obtain n−1 n−1 n−1 qk+1 − qk (4.14) (4. different functions. By choosing ˜ t = (tk + tk+1 )/2 and the previous two equations.19) is the lagrangian in phase space.17) In the same limit H(qk . pk ) k=0 e (4. we get qk+1 .20) 87 . tk + |qk . q) defined by putting the operators q to the right of p by using the canonical commutation relations. p) e H+ (q. p) = (H+ (qk+1 . t|qk .13) Since in the limit 1 Hc (qk+1 . p) (4.

On the contrary there are no limitations on p(t ). p) = The momentum integration can be done naively n−1 p2 + V (q) 2m p2 + V (q) 2m (4. t<t <t dµ(p(t)) = t≤t ≤t dp(t ) 2π (4.t q. therefore we get the canonical action t S= t dt L(q. t = · = · n→∞ n−1 lim dqk k=1 dpk i[pk (qk+1 − qk ) − p2 /2m − (V (qk+1 ) + V (qk ))/2] k e = 2π k=0 n−1 n→∞ lim dqk k=1 1 2π 2mπ i n n−1 2 e(−2m/i )(qk+1 − qk ) /4 k=0 e−i (V (qk+1 ) + V (qk ))/2 = 88 . the multiple integral in eq.25) q . as it required by the uncertainty principle.2 Path integral in configuration space The path integral in eq.22) dq(t ). p) = H(q. t = where dµ(q(t)) = i dµ(q(t))dµ(p(t))e dtL(q. t |q.20) is called the functional integral and we will denote it symbolically as t q . p) = H− (q.23) The asymmetry between the integration in q(t) and p(t) implies that the integral is not invariant under a generic canonical transformation. Notice that the integration is made over all the functions q(t ) such that q(t) = q e q(t ) = q . (4. (4. t |q.In the limit.21) In the limit n → ∞.t q .24) (4. p) = In such a case H+ (q. p) t (4.20) can be easily written as a functional integral in configuration space when the hamiltonian of the system is of the type H(q. p) (4. the argument of the exponential gives the integral between t and t of the lagrangian in the phase space. 4.

q) ˙ t (4. N H(qi .30) It is the expression (4. .28) where L(q. However for arbitrary system one has to start with the formulation in the phase space of the previous Section. . . q) is the lagrangian in configuration space.31) 2m i=1 the expression (4. t |q. pi ) = (4. .26) k=0 with the result q . t = i D(q(t))e dtL(q. But before discussing the interpretation let us notice that for N degrees of freedom and with hamiltonians of the type N p2 i + V (qi ) i = 1.29) with the functional measure given by D(q(t)) = lim m 2iπ n n−1 n→∞ dqk k=1 (4.= · n→∞ n−1 lim m 2iπ n n−1 dqk k=1 2 ei [m((qk+1 − qk )/ ) /2 − (V (qk+1 ) + V (qk ))/2] (4. Therefore we will write ˙ t q . q(t ) = q (4. let us recall the boundary conditions q(t) = q.32) 89 . For discussing the interpretation of the expression (4.29). t = lim m 2iπ n n−1 n→∞ n−1 dqk k=1 · We have also n−1 n→∞ i e [m((qk+1 − qk )/ )2 /2 − (V (qk+1 ) + V (qk ))/2] k=0 (4.t q . t |q.t q.27) lim k=0 m 2 qk+1 − qk 2 − V (qk+1 ) + V (qk ) = 2 t dtL(q.27)) which was originally formulated by Feynman.29) (really defined by (4. In fact it has a simple physical.29) can be easily generalized. q) ˙ t (4. which also can be easily generalized to an arbitrary number of degrees of freedom.

1. t' t q q' Fig. (4. This path is approximated in eq..1 . Each of these amplitudes (neglecting the normalization factor m/2iπ ) is given by the exponential of i times the action evaluated along the path. t ) is obtained by multiplying together the amplitudes corresponding to the infinitesimal straight paths.The Feynman definition for the path integral.2. t) to (q .2. 4. where S is the classical action evaluated along the path. Furthermore we need to sum (or integrate functionally) over all the possible paths joining the two given end points.29) is nothing but the action evaluated for a continuous path starting from q at time t and ending at q at time t . 90 . Therefore the interpretation is that in order to evaluate the amplitude to go from (q.27) by many infinitesimal straight paths as shown in Fig. 4. In a more concise way one says that the amplitude to from one point to another is obtained by summing over all the possible paths joining the two points with a weight proportional to the exponential of iS. The expression in the exponent of (4.

4. 4. P .3. From that one can derive directly the final formula in the following way. Under the conditions of Fig.2. where they are detected by D.The double slit experiment. In order to explain this distribution it would seem reasonable to make the following assumptions: .3 The physical interpretation of the path integral Feynman arrived to the path integral formulation in configuration space on a more physical basis. 4. with both slits open.1.The experimental probability distribution. 4. where S is the electron source. By counting the electrons with the detectpr D we can determine.2 . that is with both slits 1 and 2 open. P x Fig.3.3. and therefore the probability distribution of the electrons on the screen C. Let us consider the classical double slit experiment. D 1 x S 2 A B C Fig. The electrons come out of the slit A and arrive to the screen C. one gets the distribution given in Fig.3. . • Each electron must go through the slit 1 or through the slit 2. 4. as illustrated in Fig. with S the classical action evaluated along the infinitesimal path.1.1 . as a function of x. 91 . 4. First of all he knew that the amplitude for infinitesimal time differences is proportional to the exponential of iS. going through one of the two silts A or B.3. how many electrons arrive on C.

but the result that we get follows from our ignorance about the path followed by the particle. Therefore. In fact. this is the result that one obtains when doing the experiment with a further detector telling us which of the holes have been crossed by the electron.3. .• As a consequence the probability for the electron to reach the screen at the point x must be given by the sum of the probability to reach x through 1 with the probability to go through 2.3. The results obtained in the two cases are given in Fig. quantum mechanics tells us that we can say that an electron went through one of the two slits with absolute certainty. the total amplitude for the process is given by the sum of the amplitudes corresponding to the two different possibilities. In fact. He says that it is still possible to speak about trajectories of the electron if one gives up to the usual rules of combining probabilities. These hypotheses can be easily checked by closing one of the two slits and doing again the experiment. only if we detect its passage through the slit. quantum mechanics tells us that the usual rules for combining together probabilities are correct. 4.3 . that is P = |A|2 = |A1 +A2 |2 = |A1 |2 +|A2 |2 +A1 A2 +A1 A2 = |A1 |2 +|A2 |2 = P1 +P2 (4.The experimental probability distributions. We see that P = P1 + P2 . On the contrary the experiment tells us that when we do not detect the hole crossed by the electron. This means that with this experimental setting the interference gets destroyed. under such circumstances.3. P2 P1 x Fig. In particular one defines the following alternatives 92 . therefore one of our hypotheses must be wrong. define a concept as the path followed by a particle. This point of view requires taking into account different type of alternatives according to different experimental settings. That is it does not make sense. The Feynnman’s interpretation is a little bit different although at the end the theories turn out to be equivalent.33) The two expressions differ for a term which is not positive definite and which is the responsible for the interference pattern. P1 and P2 obtained keeping open the sli1 1 or the slit 2 respectively. unless we measure it. 4. Only in this case one would get P = P1 + P2 .

That is we can construct a lattice of points between A and C and construct 93 . For instance one we use a detector to determine the slit crossed by the electrons. In this case the total probability will be given by b P = a dx|A1 (x) + A2 (x)|2 (4. Then.This is the situation where the alternative chosen by the system is not detected by an experimental apparatus. if we want to know the total probability of having an electron in C within the interval (a. suppose that we want to evaluate the amplitude to go from S to the screen C when the screen B is eliminated. P = i Pi . From this point of view it is well conceivable to speak about trajectories of the electron.We will speak about exclusive alternatives when we use an apparatus to determine the alternative chosen by the system. the total amplitude will be given by A(x) = dyA(x.1. This can be done by increasing the number of holes in the screen B. There are situations in which one may have both type of alternatives. In this case the total probability is given by the sum of the probabilities corresponding to the different alternatives. (4.36) Therefore the probability rules are strictly related to the experimental setting.34) 2. In this case we associate to each alternative an amplitude Ai and assume that the total amplitude is given by the sum of all the amplitudes Ai A= i Ai (4. Exclusive alternative . y) (4. y) the amplitude for the electron to reach x going through the hole placed at a distance y from the center of B. For instance. b). Interfering alternatives . if we denote by A(x.35) The total probability for the process is evaluated as P = |A|2 = | i Ai |2 (4.37) Since the alternatives of getting the electron at different points in C are exclusive ones.38) We can continue by using more and more screen of type B with more and more holes. having a detector in C and not looking at the slit crossed by the electrons. but we have to associate to each possible path an amplitude and then sum all over the possible paths. For instance.

39) C . for the single path C the amplitude will be given by exp(iS(C). 4. C) (4. as shown in Fig. C From this point it is easy to get the path integral expression for the total amplitude.39). The canonical momentum is given by p = mqf (q) ˙ and the corresponding hamiltonian is H= 1 p2 + V (q) 2mf (q) (4. By denoting with C the generic path we get that the total amplitude is given by A(x) = A(x.3.4. This is the final result. For instance.all the possible paths from A to the point x on C joining the sites of the lattice. 4. Therefore. since each path is obtained by joining together many infinitesimal paths. This is important because in many cases the simple result outlined above is not the correct one. and for each of these paths the amplitude is proportional to exp(iS) where S is the action for the infinitesimal path.42) (4.4 .40) where f (q) is an arbitrary function of q. D C1 S C2 x A Fig.3. consider the following lagrangian L= m 2 q f (q) − V (q) ˙ 2 (4.41) 94 .The lattice for the evaluation of equation (4. This is easy to get if one starts from the phase space formulation. except for a proportionality factor.

44) We can rewrite this expression in a way similar to the one used in eq.43) e 2π k=0 This expression is formally the same as the one in eq. t |q.46) The formal expression for the path integral is always the same. (4. where qk = (qk + q ¯ qk+1 )/2. In the continuum limit we have n−1 n→∞ lim f (¯k ) = q k=0 n→∞ lim e 1 2 n−1 log f (¯k ) q k=0 n−1 = log f (¯k ) q 1 δ(0) = e2 t = where we have used n−1 n→∞ n→∞ lim e 1 2 dt log f (q) t k=0 (4. t = lim n−1 n→∞ dqk k=1 · dpk i[pk (qk+1 − qk ) − p2 /(2mf (¯k )) − (V (qk+1 ) + V (qk ))/2] q k (4. but the functional measure is now ˙ D(q(t)) = lim mf (¯k ) q 2iπ n n−1 n→∞ dqk k=1 (4.47) t lim = k=0 t dt . (4. n→∞ lim δhk = δ(th − tk ) (4. form eq.48) 95 . t |q. Therefore we get q q . Therefore. but the integration measure must be determined by using the original phase space formulation. (4.29) t q .20) we get n−1 q . t = i D(q(t))e dtL(q.t q. q) given in (4. t = lim mf (¯k ) q 2iπ n n−1 n→∞ n−1 dqk k=1 · i e [mf (¯k )((qk+1 − qk )/ )2 /2 − (V (qk+1 ) + V (qk ))/2] q k=0 (4.26) after sending m → mf (¯k ). t |q.Notice that for qk ≈ qk+1 we have f (¯k ) ≈ (f (qk ) + f (qk+1 ))/2.45) with L(q.40). q) ˙ t (4.t q .

54) (4. we get ∆S = 1. For a macroscopical particle (m ≈ 1 gr and times of h about 1 sec) the typical action is about 1027 / . For a macroscopic particle (say m = 1 gr) and recalling that h ≈ 10−27 erg·sec. To find the total amplitude we must sum over all the possible / paths C. The classical path (the one making S stationary) is t xclass (t) = x0 (4.49) The reason why we say that this is a quantum modification is because the second term must be proportional to h. for an electron m ≈ 10−27 gr.51) Let us now consider another path joining the origin with x0 as. Let us recall that the amplitude for a single path is exp(iS(C)// ) (where we have h put back the factor h). It follows that in the microscopic world.6 · 1026/ and this / h non-classical path can be safely ignored. We get ∆S = m/6. and ∆S = h/6. for instance. take a free particle starting from the origin at t = 0 and arriving at x0 at the time t0 . whereas they will cancel going away from this point due to the fast variation of the phase. However. Since the amplitudes are phase factors the will cancel except for the paths which are very close to a stationary points of S.53) (4.50) t0 The corresponding value for S is 1 S(xclass ) = m 2 t0 0 1 x2 q2 = m 0 ˙ 2 t0 (4. / An interesting question is how one gets the classical limit from this formulation. 96 . as it follows from dimensional arguments. many paths may contribute / to the amplitude.52) Let us take t0 = 1 sec and x0 = 1 cm. For instance. and the amplitude gets contribution h only from the classical path. Around this point the phases will add coherently. t2 x(t) = x0 2 t0 The corresponding action is 2 x2 S(x) = m 0 3 t0 and we get 1 x2 ∆S = S(x) − S(xclass ) = m 0 6 t0 (4. of the order of one. Roughly we can expect that the coherence effect will take place for those paths such that their phase differs from the stationary value. S(Cclass )// .Therefore the result can be also interpreted by saying that the classical lagrangian gets modified by quantum effects as follows i L → L − δ(0) log f (q) 2 (4.

t |q. In this case we can easily make the integrations going back to the original definition (eq. after n − 1 steps we find n−1 n−1 dqk k=1 i e m((qk+1 − qk )/ )2 /2 k=0 1 2 = 2πi m n 2 m 2πi(n ) 97 2 eim(qn − q0 ) /(2(n )) (4. t = lim n→∞ m 2iπ n n−1 dqk k=1 i e m((qk+1 − qk )/ )2 /2 k=0 (4.58) let us start by integrating q1 2 2 dq1 eim[(q2 − q1 ) + (q1 − q0 ) ]/(2 ) = iπ im(q2 − q0 )2 /(2(2 ) e m (4. a Re a ≤ 0 (4.61) . The expression we wish to evaluate is n−1 q .59) 2πi = m m 2πi(2 ) 1 2 2 eim(q2 − q0 ) /(2(2 ) Multiplying by the exponential depending on q2 and integrating over this variable we get 2πi m 2πi = m = 2πi m m 2πi(2 ) m 2πi(2 ) 3 2 1 2 2 2 dq2 eim[(q3 − q2 ) + (q2 − q0 ) /2]/(2 ) 1 2 2πi(2 ) 3m 1 2 1 2 2 eim(q3 − q0 ) /(2(3 )) m 2πi(3 ) 2 eim(q3 − q0 ) /(2(3 )) (4.60) By repeating this procedure.55) since the lagrangian is 1 ˙ (4.27).4.57) which allows us to evaluate the expression 2 2 dxea(x1 − x) + b(x2 − x) = − π ab(x1 − x2 )2 /(a + b) e a+b (4. (4.56) L = mq 2 2 The integrations can be done immediately starting from the gaussian integral ∞ −∞ 2 dxeax + bx = 2 π − e−b /4a .4 The free particle We shall now consider the case of the one-dimensional free motion.

(4.t 0. However we will not do it here.62) This expression can be generalized to a free system with N degrees of freedom N N 2 qi . Of course. t |q.55)) m q . t = 2πi(t − t) 1 2 eiScl (4.69) . Then we define the wave function as ψ(q .63) t →t lim q . we take a fixed initial point of propagation.68) mq 2 i e 2t (4. t |q. t |0. for instance (q. N (4. t) to (q . i = 1. t ) = q . 0 = For the free particle we get m ψ(q. t) = 2πit 98 1 2 q . one could formulate quantum mechanics directly in terms of the path integral.65) Scl = t m 2 m (q − q)2 qcl dτ = ˙ 2 2 (t − t)2 t dτ = t m (q − q)2 2 (t − t) (4. t = δ(q − q) (4. that is that the amplitude is proportional to the exponential of iS with S the action evaluated along the classical path. but we will limit ourselves to notice that the formulations agree in the free case. t ) for the free particle is given by qcl (τ ) = q + The corresponding classical action is t τ −t (q − q) t −t (4. t |q. is generally true.0 D(q(t))eiS (4. 0).66) Therefore our result can be written as m q . t |qi . t = 2πi(t − t) 1 2 2 eim(q − q) /(2(t − t)) (4. t = Notice also that m 2πi(t − t) im e (qi − qi )2 /(2(t − t)) i=1 .67) We shall see that for all the lagrangians quadratic in the coordinates an in the momenta this result.The first factor in the right hand side of this equation cancels out an analogous factor in the integration measure (see eq. t) = (0.64) The classical path connecting (q. It is not difficult to show that it is possible to recover all the results of the ordinary formulation and prove that one has complete equivalence. · · · .

78) . We then perform the change of variables η(τ ) = q(τ ) − qcl (τ ) which in terms of our original definition (4.75) qcl (t ) = q (4.29) means ηk = qk − qcl k .72) Since this is a translation the functional measure remains invariant. 4.This is indeed the wave function for a free particle in configuration space.29) let us start by the Euler-Lagrange equations associated to eq. The new variable η(τ ) satisfies the boundary condition η(t) = η(t ) = 0 The lagrangian in terms of the new variables is given by 1 1 ˙ ˙ L(q) = L(qcl + η) = L(qcl ) + mη 2 + bηη − cη 2 2 2 + dη − eη + mqcl η + bqcl η + bηqcl − cqcl η ˙ ˙ ˙ ˙ ˙ 99 (4. as it can be checked by going in the momentum space m ψ(p. t) = 2πit 1 2 mq 2 i dq e−ipq e 2t = m 2πit 1 2 2πit m 1 2 p2 t −i e 2m = e−iEt (4.73) Let us denote by qcl (τ ) the solution of the classical equations of motion satisfying the boundary conditions qcl (t) = q.5 The case of a quadratic action In this Section we will consider the case of a quadratic lagrangian for a single degree of freedom (it can be easily extended to many degrees of freedom) 1 1 ˙ L = mq 2 + b(τ )qq − c(τ )q 2 + d(τ )q − e(τ )q − f (τ ) ˙ ˙ 2 2 (4. n − 1 (4.71) d [mq + b(τ )q + d(τ )] − [b(τ )q − c(τ )q − e(τ )] = 0 ˙ ˙ dτ that is ˙ m¨ + bq + d + cq + e = 0 q ˙ (4.77) (4.70) where E = p2 /2m. · · · .76) (4.74) (4. k = 1.71) To evaluate the integral (4. (4.

t |q.80) 2 2 Therefore we can write the path integral as follows t q . In fact the path integral which remains to be evaluated depends only on t and t due to the boundary conditions satisfied by η. let us notice that the term bη η is a trivial one. First. However it is important for systems with a finite number of degrees of freedom and therefore we will give here an idea how one can evaluate it in general.t i D(η(τ ))e t 1 1 ˙ ( mη 2 + bη η − cη 2 )dτ ˙ 2 2 (4. this term turns out to be irrelevant.82) bη − bη 2 dτ 2 2 and defining ˙ c=c+b ˜ (4. q ) 0.t 0.84) n→∞ lim n→∞ lim iη ση k=1 T where we have used the boundary conditions on η (η0 = ηn = 0) and we have introduced the vector ⎤ ⎡ η1 ⎢ η2 ⎥ ⎥ ⎢ (4.79) ˙ ˙ ˙ q L(q) = L(qcl ) + mη 2 + bηη − cη 2 + dη − eη − m¨cl + bqcl + cqcl η 2 2 Recalling the equations of motion for qcl and performing a further integration by parts we find 1 1 ˙ L(q) = L(qcl ) + mη 2 + bηη − cη 2 + dη − eη − −d − e η ˙ ˙ ˙ 2 2 1 1 = L(qcl ) + mη 2 + bηη − cη 2 ˙ ˙ (4.83) In this way we get t n−1 n→∞ i t 1 2 1 2 mη − cη dτ = ˙ ˜ 2 2 = = lim i k=0 n−1 m 1 (ηk+1 − ηk )2 − ck ηk ˜ 2 2 2 im i 2 2ηk − 2ηk ηk+1 − ck ηk ˜ 2 2 2 (4. through ˙ an integration by part d 1 2 1˙ bη η = ˙ (4.Inside the action we can integrate by parts the terms containing both qcl and η. For the future purposes of extension of this approach to field theory. since we can absorb it into a re-definition of c(τ ).85) η=⎢ · ⎥ ⎥ ⎢ ⎣ · ⎦ ηn−1 100 . By using the boundary conditions for η we get 1 1 ˙ (4.81) Notice that here all the dependence on the variables q and q is in the classical action. t = eiScl (q.

91) G(t .93) ··· 0 ··· 0 ··· 0 ··· · · · · cj−1 ˜ ··· 0 101 . t).t 0 c2 ˜ · · 0 0 ··· 0 ··· 0 ··· · ··· · · · · cn−2 ˜ ··· 0 0 0 · · 0 ⎤ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ cn−1 ˜ (4. We get n−1 dηk k=1 T eiη ση = n−1 dζk k=1 T eiζ σD ζ = n−1 k=1 π = (−iσk ) π n−1 2 det(−iσ) (4. it can be diagonalized by an orthogonal transformation σ = U T σD U (4. · · · .86) G(t .92) ⎤ 0 0⎥ ⎥ 0⎥ ⎥ · ⎥ ⎥ 0⎦ cj ˜ (4. t) = lim then n→∞ n→∞ m 2πi n π n−1 det(−iσ) 1 2 = lim n→∞ m 1 1 2i n−1 2πi ( m ) det(−iσ) det(−iσ) (4. since det|U| = 1.88) The eigenvectors of σ. t) define (for j = 1.87) Since σ is a symmetric matrix. t) = lim Let us define f (t . t) ≡ 0.89) 1 2 We get G(t .90) 2i m n−1 (4.and the matrix ⎤ ⎡ ⎡ 2 −1 0 · · · 0 0 0 c1 ˜ ⎥ ⎢ −1 2 −1 · · · 0 ⎢0 0 0 ⎥ ⎢ ⎢ ⎢ m ⎢ 0 −1 2 · · · 0 0 0 ⎥ ⎥− ⎢ · ⎢ σ= · · ··· · · · ⎥ 2⎢ · 2 ⎢ · ⎥ ⎢ ⎢ ⎣ 0 ⎣0 0 0 · · · −1 2 −1 ⎦ 0 0 0 · · · 0 −1 2 0 Therefore t 0. ζ = Uη. let us ⎡ 2 −1 0 · · · 0 ⎢ −1 2 −1 · · · 0 ⎢ ⎢ 0 −1 2 · · · 0 Pj = ⎢ ⎢ · · · ··· · ⎢ ⎣ 0 0 0 · · · −1 0 0 0 ··· 0 m 2πif (t . t) = In order to evaluate f (t .t i D(η(τ ))e m 2πi n 2 t 1 1 ˙ ( mη 2 + bη η − cη 2 )dτ ˙ 2 2 n−1 T eiη ση = n→∞ lim dηk k=1 (4. are real and the Jacobian of the transformation is one. n − 1) ⎤ ⎡ c1 0 0 ˜ 0 0 ⎢ 0 c2 0 0 0 ⎥ ˜ ⎥ ⎢ 2 ⎢ ⎥ 0 0 ⎥ ˜ ⎢ 0 0 c3 ⎥− m⎢ · · · ⎥ · · ⎢ ⎣0 0 0 2 −1 ⎦ −1 2 0 0 0 (4.

97) where we define p0 = 1.100) in the limit in which we keep τ fixed. ˜ j = 1.96) m Therefore we find the recurrency relation pj+1 = 2− 2 m cj+1 pj − pj−1 .and pj = det|Pj | Clearly 2i m n−1 (4. pj is given by p1 = 2 − p2 = p3 = 2 m c1 ˜ 2 2− 2− m 2 c2 p1 − 1 ˜ c3 p2 − p 1 ˜ (4.99) a time τ = t + j .98) It is also convenient to introduce the finite difference operator ∆pj = pj+1 − pj (4.95) For the first values of j.101) The equation (4.94) det(−iσ) = pn−1 (4.102) therefore we get the following differential equation for φ(τ ) c(τ ) ˜ d2 φ(τ ) φ(τ ) =− 2 dτ m 102 (4. It follows lim ∆ pj = →0 dφ(τ ) dτ (4.98) can be written as ∆2 pj−1 = − cj+1 pj ˜ m (4. and a function φ(τ ) such that lim pj = φ(τ ) →0 (4. n − 1 (4. It is convenient to put this expression in the form pj+1 − 2pj + pj−1 2 =− cj+1 pj ˜ m (4.103) . · · · .

106) 1 sin ω(τ − t) ω 1 sin ωT ω (4. t). t) = φ(t) = 0. t = with f (t . Therefore d2 φ(τ ) = −ω 2 φ(τ ).110) (4. t |q. dφ(τ ) |τ =t = 1 dτ (4. t) (4. (4. φ(τ ) satisfies the following boundary conditions φ(t) = lim p0 = 0 →0 2 p1 − p0 dφ(τ ) |τ =t = lim = lim 2 − c1 − 1 = 1 ˜ →0 →0 dτ m (4. The function φ(τ ) satisfies d2 φ(τ ) = 0.112) 103 .111) (4. for quadratic lagrangian we get the general result q .91) we have f (t . In this case b = 0 and c = mω 2 .108) φ(t) = 0.109) (4.105) Therefore.Furthermore. The harmonic oscillator. We need only to evaluate f (t . dτ 2 from which φ(τ ) = by putting T = t − t: f (t . dτ 2 with solution φ(τ ) = τ − t from which f (t .107) m eiScl 2πif (t . t) = t − t in agreement with the direct calculation. t) = lim pn−1 = φ(t ) →0 (4. Examples: The free particle. t) = φ(t ).104) Recalling eq. dφ(τ ) |τ =t = 1 dτ (4.

the space of the square integrable real functions of a real variable.116) (4.To evaluate the classical action we need to solve the classical equations of motion m¨ + mω 2 q = 0 q with boundary conditions q(t) = q.120) 1 mω (q 2 + q 2 ) cos ωT − 2qq 2 sin ωT (4. The general solution is given by q(τ ) = A sin ω(τ − t) + B cos ω(τ − t) From the boundary conditions A= q − q cos ωT .119) In the limit ω → 0 we get back correctly the free case. There are many spaces of functions. 4. t = i mω (q 2 + q 2 ) cos ωT − 2qq mω 2 sin ωT e 2πi sin ωT (4. t |q.117) Recalling the following integrals t t sin 2ωT dτ cos 2ω(τ − t) = .115) q(t ) = q (4. The concept of functional is nothing but an extension of the concept of function.113) and substituting inside the lagrangian 1 1 1 ˙ L = mq 2 − mω 2 q 2 = mω 2 (A2 − B 2 ) cos 2ω(τ − t) − 2AB sin 2ω(τ − t) 2 2 2 (4. 2ω t t dτ sin 2ω(τ − t) = 1 − cos 2ωT 2ω (4.118) we get Scl = and q . A real functional is defined as an application from a space of functions (and therefore an ∞-dimensional space) to R1 . For instance.114) (4. Therefore a real 104 . sin ωT B=q (4. that a law which associates a number to each point of the manifold. Recall that a function is simply a mapping from a manifold M to R1 .6 Functional formalism The most natural mathematical setting to deal with the path integral is the theory of functionals.

121). +∞]= the space of the integrable functions in R1 with respect to the measure w(x)dx: +∞ F1 [η] = dxw(x)η(x) −∞ (4.121) to denote the functional of defined in eq. A simple way of understand this point is to consider some path on the definition manifold of the functions.122) L → R1 (4. Examples of functionals are • L=L1 [−∞. The next step is to define the derivative of a functional. as we did for defining the path integral. ηn ) obtained by evaluating the functions at the discrete points in which the interval has been divided. Then we can approximate a function η(t) with a convenient limit of its discrete approximation (η1 . F [·]. · · · .126) Then we can consider the variation of F (ηi ) when we vary the quantities ηi n F (ηi + δηi ) − F (ηi ) = j=1 ∂F (ηi ) δηj ∂ηj (4. i = 1. For instance consider functions from R1 to R1 . (4. · · · . F. η2 . ηn ) ≡ F (ηi ). η∈L (4.123) • L=L2 [−∞. n (4. +∞]= space of the square integrable functions in R1 : 1 − F2 [η] = e 2 • L=C= space of the continuous functions: F3 [η] = η(x0 ) (4. t ) of R1 made of many infinitesimal pieces.127) Let us put Ki = 1 ∂F ∂ηi (4. In this approximation we can think of the functional F as an application from Rn → R1 .functional associate a real number to a real function. If we denote the space of functions by L.124) F3 is the functional that associates to each continuous function its value at x0 .128) 105 .125) dxη 2 (x) (4. and take an interval (t. That is F [η] → F (η1 . · · · . a real functional is the mapping F : We will make use of the notation F [η]. We need to understand how a functional varies when we vary the function.

Looking at the eq.128) we get δF [η] 1 ∂F (ηi ) = lim →0 δη(x) ∂ηi xn − x1 . n (4.130) Since δη(x) is the variation of the function evaluated at x. · · · .138) . we will define the functional derivative as δF δF [η] = δη(x)dx (4. xi = x1 + (i − 1) .134) w(x)δη(x)dx (4.135) (4. Then. for → 0 n F (ηi + δηi ) − F (ηi ) = j=1 Kj δηj → Kδηdx (4. to evaluate the functional derivative of a functional we first evaluate its infinitesimal variation and then we can get the functional derivative by comparison with the previous formula. = n−1 ηi = η(xi ). i = 1.132) let us now evaluate the functional derivatives of the previous functionals: δF1 [η] = giving δF1 [η] = w(x) δη(x) Then 1 − δF2 [η] = e 2 dxη 2 (x) (− η(x)δη(x)) (4. (4.133) and therefore δF2 [η] = −η(x)F2 [η] δη(x) (4.131) δη(x) In other words. and δF3 [η] = δ(x − x0 ) δη(x) 106 (4.where is the infinitesimal interval where we evaluate the discrete apoproximation to the function.137) In this way F3 looks formally as a functional of type F1 .129) that is δF [η] = K(x)δη(x)dx (4.136) In order to evaluate the derivative of F3 let us write F3 [η] = η(x)δ(x − x0 )dx (4.

This generalization is called the Volterra expansion of a functional ∞ F [η0 + η1 ] = k=0 1 k! dx1 · · · dxk δ k F [η] δη(x1 ) · · · δη(xk ) η1 (x1 ) · · · η1 (xk ) η=η0 (4. y)η(y)F4[η] (4.Let us consider also 1 − F4 [η] = e 2 then 1 − δF4 [η] = e 2 dxdyK(x. We get ∂L d ∂L δS = − δq(t) ∂q(t) dt ∂ q(t) ˙ (4.142) By varying it t δS[q] = t ∂L ∂L δq + δ q dt = ˙ ∂q ∂q ˙ t t ∂L d ∂L − ∂q dt ∂ q ˙ δqdt (4.144) Starting from the functional derivative one can generalize the Taylor expansion to the functional case.139) where we have used the symmetry of the integral with respect to x ↔ y.141) A further example is the action functional. y)η(x)η(y) (− dxdyK(x. We obtain δF4 [η] =− δη(x) dyK(x. In fact the action is a functional of the path t S[q] = t L(q.140) dxdyK(x. or from a manifold M to a manifold N. 4. y)η(x)η(y) (4. q)dt ˙ (4.143) Since we consider variations with respect to functions with fixed boundary conditions. y)δη(x)η(y)) (4.145) All these definitions are easily generalized to the case of η a function from Rm → Rn . we have δq(t) = δq(t ) = 0. namely 107 .7 General properties of the path integral In this Section we would like to stress some important properties of the path integral.

t i dµ(q(τ ))dµ(p(τ ))e q”.1 . 4.146) • Factorization of the path integral. sum up over all the paths fromq a q” and from q” to q . t (4.The factorization of the path integral.t q”. t”|q. 4.1)).t q . π(t ) (4. t = q.7.• Invariance of the functional measure under translations . t' t" t q' q q" Fig.It follows immediately from the definition that q(τ ) → q(τ ) + η(τ ).t” i dµ(q(τ ))dµ(p(τ ))e t Ldτ t × = i dµ(q(τ ))dµ(p(τ ))e Ldτ t” dq” q .7. t |q.t” Ldτ + i t t” t” Ldτ = dq” q. t |q”. . t” q”. and eventually integrating over all the possible intermediate points q” (see Fig.147) as it follows from the factorization properties of the measure dµ(q(τ )) = t<τ <t dq(τ ) = dq” t<τ <t” dq(τ ) t”<τ <t dq(τ ) 108 . From the very definition of the path it follows that in order to evaluate the amplitude for going from q at time t to q at time t we can first go to q” to a time t ≤ t” ≤ t . p(τ ) → p(τ ) + π(τ ). η(t) = η(t ) = 0 for any π(t). In equations we get t” t qt q .

t q.t q.t q”.149) where F is an arbitrary functional of q. o The next important property of the path integral is that it gives an easy way of evaluating expectation values. t (4.t dµ(q(τ ))dµ(p(τ ))eiS q(t”) = = dq” q .152) dq” q . By proceeding as before we will get the expectation value of q(t1 )q(t2 ) if t1 ≥ t2 .dµ(p(τ )) = t≤τ ≤t” dp(τ ) = t≤τ ≤t” dp(τ ) t”≤τ ≤t dp(τ ) (4.t q .t q. or q(t2 )q(t1 ) if 109 .t q.t dµ(q(τ ))dµ(p(τ ))F [q(τ ). Now let us study the average value of q(t1 )q(t2 ) con t ≤ t1 . t” q” q”. p(τ )]eiS (4. t” q”. t |q(t”)|q”. q(t ) = q . t |q”.153) Therefore the average value of q(t”) evaluated with exp(iS) coincides with the expectation value of the operator q(t”).151) since q(t) satisfies the boundary conditions q(t) = q. To this end we will consider expressions of the type (which will be called the average value of F ) q .148) It is clear that the factorization property is equivalent to the completeness of the states. t (4. Using these properties it is possible to show that the wave function.68) satisfies the Schr¨dinger equation. t ≤ t” ≤ t (4. t |q. t2 ≤ t .t q. as defined in (4. t is an eigenstate of q(t) it follows q . t q .t dµ(q(τ ))dµ(p(τ ))eiS q(t ) = q q . By using the factorization we get q .t dµ(q(τ ))dµ(p(τ ))eiS = q q . t Since |q.t We begin by the observation that q .t dµ(q(τ ))dµ(p(τ ))eiS q(t”) = · = q”. t”|q.t dµ(q(τ ))dµ(p(τ ))eiS q(t”). t |q(t”)|q.t” dµ(q(τ ))dµ(p(τ ))eiS q(t”) dµ(q(τ ))dµ(p(τ ))eiS (4. t”|q.t” dq” q. p and of their time derivatives. Let us start with q .150) q.

t dµ(q(τ ))dµ(p(τ ))eiS q(t1 ) · · · q(tn ) = q . t |T q(t1 )q(t2 ) |q. t (4.155) Let us now consider a functional of q(t)such to admit an expansion in a Volterra’s series q .154) Analogously q . ˙ For instance. t (4.t dµ(q(τ ))dµ(p(τ ))eiS q(t1 )q(t2 ) ˙ q(t2 + ) − q(t2 ) dµ(q(τ ))dµ(p(τ ))eiS q(t1 ) = lim →0 q.157) If the functional depends also on the time derivatives a certain caution is necessary. t |T q(t1 )q(t2 + ) − T q(t1 )q(t2 ) |q.156) T (F [q]) ≡ k=0 1 k! dt1 · · · dtk δ k F [q] δq(t1 ) · · · δq(tk ) T (q(t1 ) · · · q(tk )) q=0 (4. t2 ≤ t ): q .t q. let us study the average value of q(t1 )q(t2 ) (t ≤ t1 .159) 110 . t |T (q(t1 ) · · · q(tk ))|q.t q. t dt2 Let us define an ordered T -product (or covariant T-product) as T q(t1 )q(t2 ) = ˙ d T q(t1 )q(t2 ) dt2 (4.t q. t q=0 1 k! ≡ where q . t ∞ (4.t ∞ dµ(q(τ ))dµ(p(τ ))F [q]eiS dt1 · · · dtk δ k F [q] δq(t1 ) · · · δq(tk ) = · = k=0 k=0 q .t ∞ 1 k! q=0 dµ(q(τ ))dµ(p(τ ))eiS q(t1 ) · · · q(tk ) dt1 · · · dtk δ k F [q] δq(t1 ) · · · δq(tk ) q .t = lim = 1 →0 q .t q.t dµ(q(τ ))dµ(p(τ ))eiS q(t1 )q(t2 ) = q . t |T (q(t1 ) · · · q(tn ))|q. In terms of the T -product we get q . t |T (F [q])|q.t q .t q. t |T (q(t1 )q(t2 )|q. t (4.158) d q .t2 ≥ t1 .

To this end one has to define the integral in an appropriate way as it will be shown later on).t q. t ˙ ˙ (4. inte˙ ˙ grating by parts and using (4. q] in a Volterra’s series of q and q.p η(t) i q . By performing explicitly the functional derivative in eq. p(τ )]eiS = 0 δq(τ ) (4.t q. (4.162) ⎥ ∂qk ⎢ ⎣ ⎦ (of course we are assuming that the integrand goes to zero for large values of the qi ’s.164) Proving the validity of eq. that is dµq. t |T (F [q. The T product is obtained as in (4. t (4.160) The proof can be done by expanding F [q. qn−1 )e k=0 ⎥ = 0 (4.t dµ(q(τ ))dµ(p(τ ))F [q.p A[q] = dµq. q]eiS = q . (4.t dµ(q(τ ))dµ(p(τ )) q.p A[q + η] = dµq. t |T F [q] δS δq(τ ) |q.t The importance of this relation follows from the fact that we can use this relation to get all the properties of quantum mechanics by selecting the functional F in a 111 . t |T δF [q] δq(τ ) |q.161)we get a relation which is the basis of the Schwinger’s formulation of quantum mechanics. the commutation relations and so on so forth. t = − q .161) This is a simple consequence of the definition of the path integral and of the functional derivative ⎡ ⎤ n−1 n−1 n→∞ n−1 lim dqk k=1 k=0 dpk 2π ⎢ i L(qk .158). At the end one has to integrate again by parts.t q .t q.Then it is not difficult to show that for an arbitrary functional of q(t) e q(t) one has ˙ q . q])|q.165) q .163) It follows for an infinitesimal function η such that η(t) = η(t ) = 0 δA[q] dt δq(t) q.p = dµ(q(τ ))dµ(p(τ )) (4. The fundamental relation follows from the following property of the path integral q . let us introduce the following notation dµq.t δ F [q(τ ). · · · . The power of this formulation of quantum mechanics comes about when we want to get the fundamental properties of quantum mechanics as the equations of motion for the operators.t q.156).t q . In fact.161).p A[q]+ dµq.t (4. pk ) ⎥ ⎥ 1 ∂ ⎢ ⎢F (q1 .t q . On the other hand the previous equation follows from the invariance under translation of the functional measure.

175) (4. by choosing F [q] = 1 we get q .174) ∂V ∂q(t2 ) = −iδ(t1 − t2 ) (4. t |q.168) For simplicity.171 ˙ q −mδ(t1 − t2 )[q(t1 ). let us consider the following action t S= t 1 2 mq − V (q) dt ˙ 2 (4. q(t2 )] + T q(t1 ) m¨ (t2 ) + and using the equations of motion. p(t)] = i 112 (4.171) ˙ ˙ q = δ(t1 − t2 )q(t1 )q(t2 ) + δ(t1 − t2 )q(t2 )q(t1 ) + T (q(t1 )¨ (t2 )) ˙ = δ(t1 − t2 )[q(t1 ).169) then ∂V (q(t2 )) δS = − m¨(t2 ) + q δq(t2 ) ∂q(t2 ) T q q(t1 ) m¨ (t2 ) + (4. t = 0 δq(t1 ) (4. t δ(t1 − t2 ) δq(t2 ) δS δq(t2 ) (4. With the choice F [q] = q we get q .173) .166) that is the quantum equations of motion.convenient way. mq(t)] = i Therefore we get the canonical commutator [q(t).167) and since this relation holds for arbitrary states T q(t1 ) = iδ(t1 − t2 ) (4. t |T q(t1 ) δS |q. we get ˙ [q(t). t = i q .170) ∂V (q) = −iδ(t1 − t2 ) ∂q(t2 ) By using the definition of the T -product we get T (q(t1 )¨ (t2 ) = q d2 T (q(t1 )q(t2 )) dt2 2 d2 = [θ(t1 − t2 )q(t1 )q(t2 ) + θ(t2 − t1 )q(t2 )q(t1 )] dt2 2 d − δ(t1 − t2 )q(t1 )q(t2 ) + δ(t1 − t2 )q(t2 )q(t1 ) = dt2 ˙ ˙ + theta(t1 − t2 )q(t1 )q(t2 ) + θ(t2 − t1 )q(t2 )q(t1 ) from which (4.172) Inserting this inside the eq. (4. For instance. t |T δS |q. q(t1 )] + T (q(t1 )¨ (t2 )) q (4.

t |T (q(t1 ) · · · q(tn ))|q.182) (4.t dµ(q(τ ))dµ(p(τ ))eiSJ (4. t|0 (4. for the moment being. By differentiating this expression with respect to the external source and evaluating the derivatives at J = 0 we get q . t) = e−iE0 t q|0 = q|e−iHt |0 = q.181) dq dqΦ0 (q . Notice that more generally one can consider the matrix element q .177) These matrix elements can be derived in a more compact form. Let us introduce the wave function of the ground state Φ0 (q. by introducing the generating functional q . t δJ(t1 ) · · · δJ(tn ) J (4. t q . t |q.t dµ(q(τ ))dµ(p(τ ))eiS q(t1 ) · · · q(tn ) (4.176) Here |0 is the ground state of system. t = 1 i n δn q .t J = q.8 The generating functional of the Green’s functions In field theory the relevant amplitudes are the ones evaluated between t = −∞ and t = +∞. to a single degree of freedom) 0|T (q(t1 ) · · · q(tn ))|0 (4. t ) q .176) to the knowledge of the generating functional.179) and J(τ ) is an arbitrary external source. t q. t |T (q(t1 ) · · · q(tn ))|q. t) 113 . Therefore we will consider matrix elements of the type (limiting ourselves.178) t dτ [pq − H + Jq] ˙ (4.180) J=0 Now we want to see how it is possible to bring back the evaluation of the matrix element in (4. t |q. t Φ0 (q. t |T (q(t1 ) · · · q(tn ))|q. t where SJ = t q . t = q . We have seen that using the reduction formulas we are lead to evaluate the vacuum expectation value of T -products of local operators.t q. t|0 then 0|T (q(t1 ) · · · q(tn ))|0 = = dq dq 0|q .4. t |T (q(t1 ) · · · q(tn ))|q.

t .183) (4. (4. t |q. that is Z[J] = It follows 0|T (q(t1 ) · · · q(tn ))|0 = 1 i n dq dqΦ0 (q .185) wit q and q arbitrary and with the ground state wave functions taken at t = 0. t )Φ0 (q) And finally we get the result imt→+i∞.190) dq dq Φ0 (q)Φ0 (q )Φ0 (q . t ) with t > t > t > t. t ) q . t |q. t |q. t with q .184) J=0 We want to show that it is possible evaluate Z[J] as a path integral with arbitrary boundary conditions on q and q .189) Φn (q )Φn (q)e−iEn (t − t) and t→+i∞ lim e−iE0 t q . t |q . t = t→+i∞ lim Φn (q )Φn (q)e−iEn t ei(En − E0 )t = n = Φ0 (q )Φ0 (q)e−iE0 t = Φ0 (q . t )Φ0 (q ) Analogously q .t ) = (4. That is we want to prove the following relation Z[J] = lim eiE0 (t − t) q . t |q . t |q . t = n (4.187) Since E0 is the lowest eigenvalue we get t →−i∞ lim eiE0 t q . t = t →−i∞ lim Φn (q )Φn (q )eiEn t e−i(En − E0 )t = n = Φ0 (q )Φ0 (q )eiE0 t = Φ0 (q . t |q. we can write q .t →−∞ eiE0 (t − t) q . t |q . the generating functional of the vacuum amplitudes.191) = Φ0 (q)Φ0 (q )Z[J] 114 . t J Φ0 (q. To this end let us consider an external source J(t) vanishing outside the interval (t . Then. t q . t J q . t |q. t t →−i∞ Φ0 (q)Φ0 (q ) J t→+i∞.Next we define the functional Z[J]. t = q |e−iH(t − t ) |q = n J = dq dq q . t = J J Φ0 (q (4. t |q. t ) q . t (4. t) = 0|0 δn Z[J] δJ(t1 ) · · · δJ(tn ) J (4.186) Φn (q )Φn (q )e−iEn (t − t ) (4. t |q . t |q.188) (4.

199) LE = −L(q.τ ZE [J] = τ →+∞.τ →−∞ lim N q.193) where N is a J-independent normalization factor.198) Notice that the euclidean lagrangian coincides formally with the hamiltonian.τ →−∞ lim N q. Then we can write t q .195) − V (q) (4. i ) dτ Therefore q . i dq 1 )=− m dτ 2 dq dτ 2 (4. i τ dq ) + Jq)dτ dτ (4.197) LE dτ (4.200) 115 .t Z[J] = t→+i∞. the relevant physical quantities are the ratios 0|T (q(t1) · · · q(tn ))|0 = 0|0 1 i n δn 1 Z[J] Z[0] δJ(t1 ) · · · δJ(tn ) (4. and furthermore that the coefficient in front of the matrix element is J-independent. This is obtained by introducing an euclidean time τ = it (Wick’s rotation) τ q . Therefore it is physically irrelevant.192) J=0 where the J-independent factor cancels out. More generally dq (4.t i D(q(τ ))e (L + Jq)dτ t (4.196) It is then convenient to define the euclidean lagrangian and euclidean action 1 LE = m 2 SE = τ dq dτ τ 2 + V (q) (4. In fact.Let us notice again that the values q and q are completely arbitrary.τ D(q(τ ))e (L(q.t →−i∞ lim N q. This expression suggests the definition of an euclidean generating functional ZE [J].194) Consider for instance the case 1 dq L = m( )2 − V (q) 2 dt from which L(q.τ −SE + D(q(τ ))e τ Jqdτ τ (4.τ ZE [J] = τ →+∞.

203) (4. 4.9 The Green’s functions for the harmonic oscillator It is interesting to evaluate the generating functional for the harmonic oscillator starting from both the equations (4.183). t )eiSJ Φ0 (q. the path integral turns out to be well defined and convergent if SE is positive definite. t ) = π and SJ = t 1 4 D(q(τ ))Φ0 (q . for the moment being.9 and.204) 1 2 i − mωq ωt e 2 e2 1 2 1 mq − mω 2 q 2 + Jq dτ ˙ (4. t)dqdq 1 i − mωq 2 − ωt e 2 e 2 1 4 (4. We have 1 1 arg = − mω[x(t)2 + x2 (t )] − mω[x0 (t)2 + x2 (t )] − mω[x(t)x0 (t) + x(t )x0 (t )] 0 2 2 116 .202) (4. t) = π mω Φ0 (q . (4.Since e−SE is a positive definite functional.185) and (4. let us consider the argument of the exponential in eq.206) where we will try to arrange x0 (τ ) in such a way to extract all the dependence from the source J out of the integral.τi =iti Since the values of q and q are arbitrary a standard choice is to set them at zero. Let us start from eq. Then. where we do not need to specify the boundary conditions Z[J] = with mω Φ0 (q.207) t t 1 2 1 mq − mω 2 q 2 + Jq dτ ˙ 2 2 (4. (4.201) J=0.205) 2 2 t In this calculation we are not interested in the normalization factors but only on the J dependence.the vacuum expectation values of the operators q(t) can be evaluated by using ZE and continuing the result for real times δ n Z[J] 1 Z[0] δJ(t1 ) · · · δJ(tn ) = (i)n J=0 δ n ZE [J] 1 ZE [0] δJ(τ1 ) · · · δJ(τn ) (4.200). let us neglect the time dependent terms from the oscillator wave functions 1 2 arg = − mω(q 2 + q ) + i 2 let us perform the change of variable q(τ ) = x(τ ) + x0 (τ ) (4.

209) and choosing x0 as a solution of the wave equation in presence of the source J x0 + ω 2 x0 = ¨ we get im[x(t )x0 (t ) − x(t)x0 (t)] ˙ ˙ Let us take the following boundary conditions on x0 : ix0 (t ) = ωx0 (t ).210) In this way.212). the term (4.211) 1 J m (4.214) Keeping in mind the boundary conditions (4.208) + i t Integrating by parts the last integral can be written as im[xx0 ]t − i ˙ t t t [m¨0 + mω 2 x0 − J]xdτ x (4. (4.213). and therefore 1 arg = − mω[x2 (t) + x2 (t )] + i 2 t t 1 1 i ˙ [ mx2 − mω 2 x2 ]dτ + 2 2 2 117 t Jx0 dτ t (4. ˙ ix0 (t) = −ωx0 (t) ˙ (4.213) (4.212) (4. the term coming from the integration by parts cancels the second term in eq.215) .211) cancels the mixed terms 1 1 arg = − mω[x2 (t) + x2 (t )] − mω[x2 (t) + x2 (t )] 0 0 2 2 t 1 1 ˙ + i [ mx2 − mω 2 x2 ]dτ 2 2 t t 1 1 ˙0 + i [ mx2 − mω 2 x2 + Jx0 ]dτ 0 2 2 t Integrating by parts the last term we get 1 1 ˙0 [ mx2 − mω 2 x2 + Jx0 ]dτ = 0 2 2 t t 1 1 1 i [− mx0 (¨0 + ω 2 x0 − J) + Jx0 ]dτ + m[x0 x0 ]t = x ˙ t = 2 m 2 2 t 1 i t mω[x2 (t ) + x2 (t)] + = Jx0 dτ 0 0 2 2 t i t (4.t + i t t 1 1 mx2 − mω 2 x2 dτ + i ˙ 2 2 mxx0 − mω 2 xx0 + Jx dτ ˙˙ t t 1 2 1 mx − mω 2 x2 + Jx0 dτ ˙ 0 2 0 2 (4.

We have ˙ ∆(τ ) = (A − B)δ(τ ) − iAωθ(τ )e−iωτ + iBωθ(−τ )e+iωτ and ˙ ¨ ∆(τ ) = (A − B)δ(τ ) − iω(A + B)δ(τ ) − ω 2∆(τ ) from which A = B. The first two term give rise to the functional at J = 0.216) with x0 (τ ) solution of the equations of motion (4.219) These equations are easily solved fort t > s and t < s ∆(τ ) = Ae−iωτ . We obtain i Z[J] = e 2 t J(τ )x0 (τ )dτ t Z[0] (4. and we can extract it from the integral.223) (4.221) τ >0 τ <0 (4.225) dsds J(s)∆(s − s )J(s ) Z[0] (4.212). ∆(τ ) = Be+iωτ .The J dependence is now in the last term.224) A= 1 2ωm (4.222) (4.220) The constants A e B are fixed by the requirement that ∆(τ ) satisfies eq.218) and (4. Therefore we have ∆(τ ) = Aθ(τ )e−iωτ + Bθ(−τ )e+iωτ (4.217) with d2 i + ω 2 ∆(τ − s) = − δ(τ − s) 2 dτ m ˙ i∆(t − s) = ω∆(t − s). For t < τ < t this solution can be written as t x0 (τ ) = i t ∆(τ − s)J(s)ds (4. ˙ i∆(t − s) = −ω∆(t − s) (4.226) 118 . and finally 1 θ(τ )e−iωτ + θ(−τ )e+iωτ 2ωm The functional we were looking for is ∆(τ ) = 1 − Z[J] = e 2 t t (4.218).210) with boundary conditions (4. (4.

We have already noticed that here the boundary conditions are arbitrary and therefore we choose q(τ ) → 0 for τ → ±∞.We see that ∆(s − s ) is the vacuum expectation value of the T -product of the position operators at the times s and s : ∆(s − s ) = − δ 2 Z[J] 1 Z[0] δJ(s)δJ(s ) = J=0 0|T (q(s)q(s ))|0 0|0 (4.234) (4.230) (4.233) We can evaluate DE (τ ) starting from the Fourier transform DE (τ ) = Substituting inside (4.230) is +∞ q(τ ) = −∞ DE (τ − s)J(s)ds (4.227) The analogue calculation in the euclidean version is simpler since the path integral is of the type already considered (quadratic action) ZE [J] ≈ e−SE (cl) (cl) (4.232) then.235) dνe−iντ DE (ν) (4. we have to solve the equations of motion in presence of the external source δ SE − δq(τ ) that is +∞ −∞ Jqdτ = −m¨(τ ) + mω 2 q(τ ) − J = 0 q (4. the solution of the eq.231) lim DE (τ ) = 0 (4.229) 1 J m Let us now define the euclidean Green’s function q − ω2q = − ¨ d2 1 − ω 2 DE (τ ) = − δ(τ ) 2 dτ m with the boundary condition τ →±∞ (4.228) Here SE is th euclidean classical action.236) .231) we get − from which DE (ν) = 1 dνe−iντ [ν 2 + ω 2 ]DE (ν) = − δ(τ ) m 1 1 2 + ω2 2πm ν 119 (4. (4.

237) To evaluate this integral we close the integral with the circle at the ∞ in the lower ν-plane for τ > 0 or in the upper ν-plane for τ < 0.240) and therefore 1 ZE [J] = e 2 It follows +∞ Jqdτ −∞ 1 Z[0] = e 2 +∞ −∞ J(s)DE (s − s )J(s )dsds Z[0] (4.245) δ 2 ZE [J] 1 =− ZE [0] δJ(τ1 )δJ(τ2 ) (4.244) To compare these two expressions let us introduce the Fourier transform of ∆(t).242) and using the eq.201) 1 δ 2 Z[J] Z[0] δJ(t1 )δJ(t2 ) we get ∆(t) = DE (it) (4.247) .243) J=0.241) δ 2 ZE [J] 1 ZE [0] δJ(s)δJ(s ) J=0 = DE (s − s ) (4.218) we obtain ∆(t) = with (−ν 2 + ω 2 )∆(ν) = − that is ∆(ν) = dνe−iνt ∆(ν) i 2πm (4. (4. (4.238) By doing the same calculation for τ < 0 we finally get DE (τ ) = The classical action is given by +∞ +∞ 1 −ω|τ | e 2mω (4.and DE (τ ) = 1 2πm +∞ dν −∞ ν2 1 e−iντ + ω2 (4.239) SE − Jqdτ = −∞ −∞ 1 1 1 1 ¨ − m q − ω 2q + J q − Jq dτ = − 2 m 2 2 +∞ Jqdτ −∞ (4. Since it satisfies eq.246) 1 i 2 − ω2 2πm ν 120 (4. The result is DE (τ ) = 1 e−ωτ 1 −ωτ (−2πi) = e 2πm (−2iω) 2mω (4.τi =iti J=0 (4.

Im ν (4.9.250) We have omitted here the limit since the expression is now regular.237) does not present any problem (ν 2 + ω 2 > 0). in eq.246) we need to define the singularities present in the denominator. For instance. 4. since we do not encounter any singularity (see Fig. (4. it is necessary to specify the integration path as in Fig.245).9.251) Therefore we have shown how it works the analytic continuation implicit in eq.244).249) With this prescription we can safely rotate the integration path anti-clockwise by 900 .The integration path for equation (4. 4. We get ∆(t) = lim + →0 i 2πm +∞ −∞ e−iνt dν ν 2 − ω2 + i (4. (4. We would get the same result by integrating along the real axis but translating the denominator of +i ( > 0) in such a way to move the pole at ω in the lower plane and the one at −ω in the upper plane.1 . By doing so we have also seen that the Feynman prescription (the +i term) is equivalent to the euclidean formulation. In fact. (4. and therefore we need to specify how the integral is defined. One could arrive to this connection also 121 .1. this expression has two poles at ν = ±ω. In order to reproduce the euclidean result. Performing the change of variable ν = iν ∆(t) = − 1 2πm +∞ −∞ e+ν t dν = DE (it) −ν 2 − ω 2 (4. 4.2) ∆(t) = i 2πm +i∞ −i∞ e−iνt dν ν 2 − ω2 (4.Whereas eq.9. for t > 0 by closing the integral in the lower plane we get i 1 1 −iωt (−2πi) e−iωt = e 2πm 2ω 2mω .248) −ω +ω Re ν Fig.

253) Therefore the Feynman prescription is there to make the integral convergent. the path integral can be formulated as an euclidean integral..9.2 . Therefore we have shown that. in the case of quadratic actions. by noticing that the Feynman prescription is equivalent to substitute in the path integral the term i − mω 2 q 2 (t)dt (4.254) The same convergence is ensured by the euclidean formulation.The Wick rotation in the ν-plane. Im ν − ω + iε ω − iε Re ν Fig. making it a well defined mathematical expression. since we will be expanding around the quadratic case. This fact is enough to guarantee that all the manipulations that we make with the path integral in the perturbative expansion are well defined. 122 . 1 − m e 2 q 2 (t)dt (4.252) e 2 with i − m(ω 2 − i ) e 2 q 2 (t)dt (4. 4.

and its dynamics is described by the lagrangian S= V d4 x 1 ∂µ ϕ∂ µ ϕ − m2 ϕ2 ≡ 2 t dt t d3 x 1 ∂µ ϕ∂ µ ϕ − m2 ϕ2 2 (5. t)|2 − ωk |q(k. t) is a real field (5.Chapter 5 The path integral in field theory 5. t) (5.5) .2) By substituting inside eq. Here we will denote the classical field by ϕ(x). ˙ 32 (2π) 123 (5. t) (5.1) A free field theory can be seen as a continuous collection of non-interacting harmonic oscillators. t) = 1 (2π)3 d3 keik · x q(k. We have denoted the quantum operator by φ(x). t) ˙ ˙ (2π)3 (2π)3 + (k · k − m2 )q(k. Let us start with a free scalar field.1) we get S = 1 2 d4 x V d3 k d3 k [q(k.1 The path integral for a free scalar field We will extend now the previous approach to the case of a scalar field theory. t)q(k . t)q(−k. As it is well known this can be seen by looking for the normal modes. t)|2 .3) Since ϕ(x. t)]eik · x + ik · x 1 = 2 t dt t d3 k [q(k. t)q(k . To this end. let us consider the Fourier transform of the field ϕ(x. t) − (|k|2 + m2 )q(k. t)] ˙ ˙ (2π)3 q ∗ (k. We will show later how to extend the formulation to the case of Fermi fields. t) = q(−k. (5. t)q(−k.4) 2 ωk = |k|2 + m2 and S= t t dt d3 k 1 2 |q(k. For a neutral particle this is a real function.

t)|2 + J(−k.7) From the Fourier decomposition we get t S= t dt d3 k (2π)3 1 2 |q(k. t) ˙ 2 (5. t) = l(k. Let us separate q(k.11) where we have made use of eq. t)|2 − ωk |q(k.Therefore the system is equivalent to a continuous set of complex oscillators q(k. s ) = − 1 2 t dsds t d3 k J(−k.9) It follows from the reality conditions that the real parts are even functions of k whereas the imaginary parts are odd functions x(k. t) + iy(k. t) + ωk q(k. The equations of motion are 2 q (k. t). t) m(k. (5. t) = l(−k. s ) (2π)3 + m(k. it will be enough to take the mass equal to one and sum over all the degrees of freedom.9). t) + im(k. (4. s ) (2π)3 (5. ωk )m(k. t) = 0 ¨ (5.226) 1 − 2 t t dsds J(s)∆(s − s )J(s ) → 1 2 t →− dsds t d3 k l(k. t) (or a pair of real oscillators)).8) where we have also expanded J(x) in terms of its Fourier components. t) (5. t) into their real and imaginary parts q(k. t). t) = x(k. t) and J(k. ωk )l(k. In particular we get for the exponent in eq. t) = −y(−k. we get − 1 2 ≡− d4 xd4 yJ(x) i 2 d3 k ∆(s − s . s)∆(s − s . t) (5. l(k.10) Now we can use the result found in the previous Section for a single oscillator. ωk )J(k. t) = x(−k. s)∆(s − s . t)q(k.12) d4 xd4 yJ(x)∆F (x − y. ωk )eik · (x − y) J(y) (2π)3 (5. Going back to J(x. m2)J(y) 124 . s)∆(s − s . t) = −m(−k.6) We are now in the position of evaluating the generating functional Z[J] by adding an external source term to the action S= V d4 x 1 ∂µ ϕ∂ µ ϕ − m2 ϕ2 + Jϕ 2 (5. t) J(k. y(k. t).

where we have defined xµ = (x. (5. xN ) = 0|0 In the free case we can get evaluate easily the generic G(N ) In fact G0 (the index recalls that we are in the free case) is obtained by developing the generating functional Z[J] = Z[0] ∞ ∞ (N ) d4 x1 · · · d4 xn J(x1 ) · · · J(xn ) n=0 (i)n 0|T (φ(x1 ) · · · φ(xn ))|0 ≡ n! 0|0 (5. Therefore. y µ = (y.14) = i∆F (x1 − x2 . m ) = 4 k 2 + m2 (2π) We go from Minkowskian to the Euclidean description through the substitution 2 −i∆F → G0 (5. xn ) n! 0 By comparison with the expansion of eq. m2 ) = d3 k ∆(s. m2 )J(y) Z[0] (5.17) G0 (x. · · · .18) For the future it will be convenient to use the following notation for the N-points Green’s functions 0|T (φ(x1 ) · · · φ(xN ))|0 (5. the generating functional is i − Z[J] = e 2 In particular we get 1 δ 2 Z[J] 1 0|T (φ(x1)φ(x2 ))|0 = − 0|0 Z[0] δJ(x1 )δJ(x2 ) 1 ZE [J] = e 2 where J=0 d4 xd4 yJ(x)∆F (x − y. s ) and i∆F (x.15) We could repeat the exercise in the euclidean formulation. ωk )eik · x = lim →0+ (2π)3 d4 k i e−ikx 4 k 2 − m2 + i (2π) (5. · · · . with the result d4 xd4 yJ(x)G0 (x − y. m2 ) (5.21) . k).14) i Z[J] = 1− d4 x1 d4 x2 J(x1 )J(x2 )∆F (x1 − x2 ) Z[0] 2 1 − d4 x1 d4 x2 J(x1 )J(x2 )∆F (x1 − x2 ) 8 · d4 x3 d4 x4 J(x3 )J(x4 )∆F (x3 − x4 ) + · · · 125 (5.16) d4 k 1 e−ikx (5.13) Again we have put k µ = (ν. m2)J(y) ZE [0] (5.19) G(N ) (x1 .20) ≡ n=0 d4 x1 · · · d4 xn J(x1 ) · · · J(xn ) (i)n (n) G (x1 . s).

(5. To evaluate G0 we have first to notice that (N ) all the G0 ’s are symmetric in their arguments.1.24) If we represent G0 (x. In the free case we can put . G0 will be given by Fig. y). · · · .The expansion of the four point Green’s function G0 (x1 . · · · .5. Z[J] = eiW [J] 126 (5. 5. x4 ).1. x1 x2 + x1 x2 + x1 x2 x3 x4 x3 x4 x3 x4 (4) Fig.23) + ∆F (x1 − x3 )∆F (x2 − x4 ) + ∆F (x1 − x4 )∆F (x2 − x3 ) It follows G0 (x1 . We will call disconnected a diagram in which we can isolate two subdiagrams with no connecting lines.13).2 since it is obtained in terms of products of G0 .The two point Green’s function G0 (x. x2k+1 ) = 0 (4) (5. (2) (2n) 5.22) For G0 we get back eq. 5.1. y) by a line as in Fig. x4 ) = − ∆F (x1 − x2 )∆F (x3 − x4 ) + ∆F (x1 − x3 )∆F (x2 − x4 ) + ∆F (x1 − x4 )∆F (x2 − x3 ) (2) (4) (4) (5. We see that in the free case all the non vanishing Green’s functions are disconnected except for the two point function. .1. therefore in order to extract the coefficients we need to symmetrize. x2 . 5.1 . x y (2) Fig. x3 .2 . Therefore it is natural to introduce a functional generating only the connected Green’s functions.25) .Since Z[J] is even in J we have G0 (2) (2k+1) (x1 . Therefore G0 is given by a combination of products of n two point functions. Therefore we write d4 x1 · · · d4 x4 J(x1 ) · · · J(x4 )∆F (x1 − x2 )∆F (x3 − x4 ) = 1 3 d4 x1 · · · d4 x4 J(x1 ) · · · J(x4 ) ∆F (x1 − x2 )∆F (x3 − x4 ) (5.

and its Volterra expansion gives rise to the connected Green’s functions ∞ iW [J] = n=0 (i)n n! d4 x1 · · · d4 xn J(x1 ) · · · J(xn )G(n) (x1 . xpn1 )G(n2 ) (xq1 .26) 1 δ 2 Z[J] Z[0] δJ(x1 )δJ(x2 ) =i J=0 δ 2 W [J] δJ(x1 )δJ(x2 ) J=0 = −i∆F (x1 − x2 ) (5. That is W [J] is defined through eq. xn ) = c 1 0|T (φ(x1) · · · φ(xn ))|0 0|0 conn (5. xrnk ) c c c (5.28) where G(n) (x1 . (5. Notice that once we find the connected Green’s functions the theory is completely solved. · · · .27) Therefore the derivatives of W [J] give the connected diagrams with the proper normalization(that is divided by 1/Z[0]). A graphical 127 G(n) (x1 . x3 ) + G(1) (x3 )G(2) (x1 .29) The index ”conn” denotes the connected Green’s functions. c c c c c (3) (3) (1) (2) G (x1 . x2 ) + G(1) (x1 )G(1) (x2 ) ≡ G(2) + G(1) . · · · . x2 ) = G(2) (x1 . · · · . x3 ) + Gc (x1 )Gc (x2 . xn ) c (5. and then we write G(n1 ) (xp1 . · · · . x2 .30) On the right hand side we have used a symbolic notation by adding together the (n) topological equivalent configurations. · · · .with W [J] = − and we have 1 2 d4 x1 d4 x2 J(x1 )J(x2 )∆F (x1 − x2 ) + W [0] (5. · · · . xn ) = . (5. pn1 . x3 ) = Gc (x1 . since the generating functional is recovered by exponentiation of W [J]. 5.2 The generating functional of the connected Green’s functions G(1) (x1 ) = G(1) (x1 ) ≡ G(1) . In the next Section we will see that this property generalizes to the interacting case. To define Gc (x1 . xqn2 ) · · · G(nk ) (xr1 . · · · .25). etc. · · · . xn ) in all the possible subsets. · · · . c c G(2) (x1 . x2 . x2 ) + G(1) (x1 )G(1) (x2 )G(1) (x3 ) ≡ c c c c c c c ≡ G(3) + 3G(1) G(2) + G(1) c c c c 3 2 Let us start by defining recursively the connected Green’s functions. x3 ) +G(1) (x2 )G(2) (x1 .31) where we sum over all the indices ni such that n1 + · · · + nk = n and over all the permutations from 1 to n of the indices in the set p1 . xn ) we decompose the set (x1 .

1 . we get Z[J] = Z[0] ∞ n=0 (i)n n (n) J G = ei(W [J] − W [0]) n! = exp iG(1) J + c ≈ + + + i2 (2) 2 i3 (3) 3 i4 (4) 4 G J + Gc J + Gc J + · · · 2! c 3! 4! 2 3 4 i i i 1 + iG(1) J + G(2) J 2 + G(3) J 3 + G(4) J 4 c c c 2! 3! 4! c 1 2 (1) 2 2 i4 (2) 2 4 i4 i Gc J + Gc J + i3 G(1) G(2) J 3 + G(1) G(3) J 4 c c 2! 4! 3 c c 1 3 (1) 3 3 3 4 (1) 2 (2) 4 i Gc J + i Gc Gc J 3! 2 1 4 (1) 4 4 i Gc J + · · · 4! 128 . xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx = c xxxxx xxxxx xxxxx xxxxx xxxxx = c + c c xxxxx xxxxx xxxxx xxxxx xxxxx c c = c + 3 c + c c xxxxxx xxxxxx xxxxxx xxxxxx xxxxxx c c = c + 3 c + 4 c + c c c c + 6 c c + c Fig. 5.2. By defining G(0) = 1 and using the same symbolic notation as before. We will limit ourselves to an explicit check up to the 4th order. 5.1. description is given in Fig. It (n) is enough to compare the expansion of Z[J] in series of Gc with the expansion from the G(n) .. This can be done by induction.2.The connected Green’s functions. By using these definitions it is not difficult to check that W [J] is the generating functional of the connected Green’s functions.

33) the quadratic part of the action. (5.= 1 + iJG(1) + c + i2 2 G(2) + G(1) c c 2! i3 3 (3) 3 J Gc + 3G(1) G(2) + G(1) c c c 3! i4 4 (4) 2 2 4 + J Gc + 3G(2) + 4G(1) G(3) + 6G(1) G(2) + G(1) + · · · (5. A typical example is the theory λϕ4 with the potential given by λ (5.33) where V (ϕ) is the interaction potential.14). (5. from the interacting part and then we use the previous identity Z[J] = N iS + i D(ϕ(x))e −i D(ϕ(x))e d4 xV d4 xJ(x)ϕ(x) d4 xJ(x)ϕ(x) = N −i = e d4 xV (ϕ) iS0 + i e 1 δ i δJ Z [J] 0 (5.32) c c c c c c 4! 5.35) iS + i D(ϕ(x))e We separate in eq. Z0 [J] is the generating functional for the free case. evaluated in eq.3 The perturbative expansion for the theory λϕ4 S= V Let us consider an interacting scalar field. S0 .34) V (ϕ) = ϕ4 4! The derivation of the perturbative expansion is very simple in this formalism. We take advantage of the following identity valid for any functional F iS + i D(ϕ(x))F [ϕ]e 1 δ = F i δJ d4 xJ(x)ϕ(x) d4 xJ(x)ϕ(x) (5. The action will be of the type d4 x 1 ∂µ ϕ∂ µ ϕ − m2 ϕ2 − V (ϕ) 2 (5. Therefore we get −i Z[J] = eiW [J] = Z[0]e d4 xV 1 δ i − i δJ e 2 d4 xd4 yJ(x)∆F (x − y)J(y) (5.37) 129 .36) Notice that we have suppressed the temporal limits in the expression for the generating functional.

43) (5. 2)J(2) e 2 Z[0] eiW0 [J] (5. · · · . · · · . by defining ⎛ ⎞ 1 δ −i V ⎜ ⎟ i δJ − 1⎠ eiW0 [J] δ[J] = e−iW0 [J] ⎝e (5.42) Let us now consider the case of λϕ4 .37) can be rewritten in the following way −i V Z[J] = eiW [J] = e −i V = e 1 δ i δJ 1 δ i δJ i − J(1)∆F (1.In this expression we have suppressed the mass in the argument of ∆F .38) This gives rise to the following expression for the generating functional of the connected Green’s functions ⎡ ⎛ ⎤ ⎞ 1 δ −i V ⎢ ⎜ ⎥ ⎟ i δJ W [J] = −i log ⎣eiW0 [J] + ⎝e − 1⎠ eiW0 [J] ⎦ −i V ⎢ −iW0 [J] ⎜e = W0 [J] − i log ⎣1 + e ⎝ ⎡ ⎛ 1 δ i δJ ⎞ ⎤ ⎥ ⎟ − 1⎠ eiW0 [J] ⎦ (5.41) we see that we can expand W [J] in a series of δ.44) .39) (5. In fact. Notice also that the expansion is in term of the interaction lagrangian. For the following calculation it will be useful to introduce the following shorthand notation d4 x1 · · · d4 xn F (x1 . the functional δ[J] can be in turn expanded in a series of the dimensionless coupling λ δ = λδ1 + λ2 δ2 + · · · It follows 1 2 W [J] = W0 [J] − iλδ1 − iλ2 δ2 − δ1 2 130 +··· (5. xn ) Then. At the second order in δ we get 1 W [J] = W0 [J] − i δ − δ 2 + · · · 2 (5. eq.40) This expression can be easily expanded in V . (5. xn ) ≡ F (x1 .

3)∆(y. 2)∆(y. x)∆(x. 4)J(1)J(2)J(3)J(4) 4! + 6i ∆(y. 1)∆(x. 3)∆(x.47) The quantity δ2 can be made in terms of δ1 by writing δ2 = − 1 e−iW0 [J] 2(4!)2 1 δ i δJ 1 δ i δJ 4 4 eiW0 [J] e−iW0 [J] eiW0 [J] δ1 1 δ i δJ 4 eiW0 [J] (5. 1)∆(x.45) By using the following list of functional derivatives δ eiW0 [J] = −i∆(x. 1)∆(y. x)2 + 6i∆(x. y) (5. 4 is understood. 2)J(1)J(2) − 3 ∆2 (y. 3. 2)∆(x. 1)∆(y. (5.48) i −iW0 [J] e = − 2(4!) Then we have 1 δ i δJ 4 eiW0 [J] δ1 δ 3 eiW0 [J] δδ1 δ 4 eiW0 [J] δ1 + 4 δJ(x)4 δJ(x)3 δJ(x) δ 2 eiW0 [J] δ 2 δ1 δeiW0 [J] δ 3 δ1 + 6 +4 δJ(x)2 δJ(x)2 δJ(x) δJ(x)3 4 iW0 [J] δ δ1 + e (5. 2. 1)J(1) δJ(x)3 + i∆(x. x) − ∆(x. we obtain δ1 = − i ∆(y.43) we get i δ1 = − e−iW0 [J] 4! 1 δ i δJ 4 eiW0 [J] 4 1 e−iW0 [J] δ2 = − 2 2(4!) 1 δ i δJ 1 δ i δJ 4 eiW0 [J] (5. 2)J(1)J(2)) eiW0 [J] 2 δJ(x) δ3 eiW0 [J] = − 3∆(x.Then.46) +∆(x. 1)∆(x. x)∆(x. 1)∆(x. from eq.49) δJ(x)4 = 131 . 4)J(1)J(2)J(3)J(4))eiW0 [J] where the integration on the variables 1. y)∆(y. 2)J(1)J(2) δJ(x)4 (5. 2)∆(x. 3)J(1)J(2)J(3) eiW0 [J] δ4 eiW0 [J] = (−3∆(x. 1)J(1)eiW0 [J] δJ(x) δ2 eiW0 [J] = (−i∆(x.

x) + 4 − i∆(x. 2)J(1)J(2) + 12i∆(y.51) By omitting the terms independent on the external source we have 1 2 δ2 − δ1 = 2 1 1 i J(1)∆(1. 6)J(4)J(5)J(6) (5. x)∆(x. 2)J(2) + 24∆4 (y. x)∆(x. 3)∆(y. 1)∆(x. x)∆(2. y)∆(y. 6)J(4)J(5)J(6) + 12i∆(y. x)∆(x. y)∆(y. y)∆(y. (5. x)∆(x. 5)∆(y. y)∆(y. x)∆(2. 1)J(1) + i∆(x. 2)J(2) 8 1 J(1)∆(1. 3)J(1)J(2)J(3) · 4∆(y. x)∆(3. y)∆(x. 4)J(2)J(3)J(4) 12 1 J(1)J(2)∆(1. x) (5. y)∆2(y. x) ∆3 (x. y) ∆(y. 4)J(4) 12∆2 (y. 2)J(2) 2 6 4 i 2 J(1)∆(1. 3)∆(y. y)∆(y. 1)J(1) 24∆3 (y. 4)J(3)J(4) 16 i J(1)J(2)J(3)∆(1. 3)∆(y. x)∆(y.52) = + + + − · 132 .and therefore δ2 = 1 2 i −iW0 [J] δ 3 eiW0 [J] δδ1 4 δ1 − e 2 2 · 4! δJ(x)3 δJ(x) δeiW0 [J] δ 3 δ1 δ 2 eiW0 [J] δ 2 δ1 +4 + 6 δJ(x)2 δJ(x)2 δJ(x) δJ(x)3 δ 4 δ1 + eiW0 [J] δJ(x)4 (5. By using the expression for δ1 and eqs. x)∆(x.46) we get 1 2 δ2 − δ1 = 2 i i = − − 2 · 4! 4! 4 − 3∆(x. x)∆(y. 5)∆(y. y)∆(y. 4) 72 ∆(y.50) 2 The δ1 gives a disconnected contribution since there are no propagators connecting the two terms. x) − ∆(x. x)∆2 (x. 2)∆(y. x)∆(y. (5. x)∆ (x. 1)∆(x. In fact the previous expression shows that δ2 contains a term which 2 cancels precisely the term δ1 in eq. 2)∆(x. 4)∆(y. y) + ∆(x. x)∆(y.44). 4)J(3)J(4) + 6 − i∆(x.

the internal one. and a factor (−iλ/4!) for each interaction vertex. The analytic expression is obtained associating a factor i∆F (x − y) to each line connecting the point x with the point y. and the vertex at x.55) 4! 2 133 . For instance.44) and putting J = 0 we can evaluate the connected Green’s functions.3. The numerical factors appearing in eq. x2 ) = i∆F (x1 − x2 ) − c − iλ2 + d4 xd4 y∆F (x1 − x) λ 2 d4 x∆F (x1 − x)∆F (x − x)∆F (x − x2 ) δ 2 W [J] = δJ(x1 )δJ(x2 ) (5. y). 5. as we see the result has an easy interpretation. (5. 5.54) can be obtained by counting the possible ways to draw a given diagram. after that we have only a possibility to attach the vertex at the internal propagator. The diagrams of Fig. Therefore 1 1 ·4·3 = (5.1 . 5.1) are given in terms of two elements. the propagator ∆F (x − y).1) we give a graphical form to this expression. In Fig. by differentiating functionally eq.The graphical expansion for the two point connected Green’s function (2) Gc (x.54) x1 xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xx xx x x x2 = x1 x x2 x + x x1 x xx xx x xx xx x2 + xx xx y xx xx + x1 x x xxx xx y xx xx xx xx + x2 xx xx x1 x x x x2 + xx xx x1 x x x xx xx y x2 xx xx Fig.3. (5.3. corresponding to the lines and the vertex (n) λϕ4 corresponding to a point joining four lines. xx xx (5. This diagram contains the two external propagators. In order to get Gc we have to draw all the possible connected diagrams containing a number of interaction vertices corresponding to the fixed perturbative order. 5.53) J=0 1 3 ∆ (x − y) 6 F 1 ∆F (x − y)∆F (x − x)∆F (y − y) ∆F (y − x2 ) 4 λ2 − i d4 xd4 y∆F (x1 − x)∆2 (x − y)∆F (y − y)∆F (x − x2 ) F 4 . recalling that iG(2) (x1 . For instance let us consider the diagram of Fig. x2 ) c we get G(2) (x1 . To get the numerical factor we count the number of ways to attach the external propagators to the vertex. There are four ways to attach the the vertex at the first propagator and three for the second.Now.3.2.

How to get the numerical factor for the first order contributions to (2) Gc (x1 . Therefore we get 1 4! ·4·4·3·2= 134 1 6 (5. 5.3.3. x2 ).3.3.The graphical expansion up to the second order of Gc (x1 . . x2 . xx xx xx xx x1 x xx xx xx xx x x x2 = x1 xx xx x x xx xx x xx xx x2 Fig. Three ways for attaching the second leg of the first vertex to the second vertex and two ways of attaching the the third leg of the first vertex to the second one.How to get the numerical factor for one of the second order contribu(2) tions to Gc (x1 .56) . xx xx xx xx x x1 xx xx xx xx y x x2= x1 x x x x x y xx xx y xx xx x2 Fig. 5.4 . 5.3 .3. . x3 . xx x xxxxxxx xx 4 xxxxxxx xxxxxxx xxxxxxx x xxxxxxx x xxxxxxx xxxxxxx xx x 2xx xx xx x x1 xx x1 xx xx xxx xx 4 xx xx = 3 x 2xx xx x1 x x x xx x 3 x1 xxxx xx xx y xx xx + x x 2xx x1 + x x xx xx x4 + xxx xx 3 x1 xx xx + x 2xx xx x1 xx xx + x 2xx xx xx xxxx y xxx xx 4 + xx x xx 3 x y xx x x 2x xxx xx 4 xx xx xx x xx x1xx x xx x xx 4 xx x 3 x 2xx xx xx xx xx xx xxy xx xx xxx 3 xx x xx 4 x 4 + y x xx xx xx y x + x3 xx xx xx x 2xx + x xxx xx 3 x1 xx + x 2xx x xx xx xx xx xxx 4 y x x x3 (4) Fig. We proceed in the same way for the diagram of Fig. 5.. There are four ways to attach the first vertex to the first propagator and four for attaching the second vertex to the second propagator. where the factor 1/4! comes from the vertex. x2 ). x4 ).2 .

The four point function can be evaluated in analogous way and the result is given in Fig. 5.3.4. Of course we can get it by functional differentiation of W [J]. The first order contribution is G(4) (x1 , x2 , x3 , x4 ) = −iλ c d4 x∆F (x1 − x)∆F (x2 − x)∆F (x3 − x)∆F (x4 − x) (5.57)

where the numerical coefficient is one, since there are 4 · 3 · 2 of attaching the vertex to the four external lines.

5.4

The Feynman’s rules in momentum space

The rules for the perturbative expansion of the Green’s functions in space-time that we got in the previous Section can be easily extended to the momentum space. We will consider again the two point function given in eq. (5.54) where we will take (2)(i) separately the various contributions in λ that we will denote by Gc , where i (2) stands for the power of λ. let us also define the Fourier transform of Gc (x1 , x2 ) as G(2) (p1 , p2 ) = c Then we get G(2)(0) (p1 , p2 ) = (2π)4 δ 4 (p1 + p2 ) c and G(2)(1) (p1 , p2 ) = c λ i d4 p −ip(x1 − x) = −i d4 x1 d4 x2 d4 xeip1 x1 + ip2 x2 e 2 (2π)4 p2 − m2 + i d4 k −ik(x − x2 ) d4 q i i · e 4 q 2 − m2 + i 4 2 − m2 + i (2π) (2π) k 4 4 4 d p d q dk λ (2π)4 δ 4 (p − k)(2π)4 δ 4 (p1 − p)(2π)4 δ 4 (p2 + k) = −i 2 (2π)4 (2π)4 (2π)4 i i i · (5.60) 2 − m2 + i q 2 − m2 + i k 2 − m2 + i p The product of the δ 4 functions may be rewritten in such a way to pull out the overall four-momentum conservation (2π)4 δ 4 (p1 + p2 )(2π)4 δ 4 (p1 − p)(2π)4 δ 4 (p − k) By integrating over p and k we get G(2)(1) (p1 , p2 ) c λ = −i (2π)4 δ 4 (p1 +p2 ) 2 i 2 p1 − m2 + i
2

d4 x1 d4 x2 G(2) (x1 , x2 )eip1 x1 + ip2 x2 c i − m2 + i

(5.58)

p2 1

(5.59)

(5.61)

i d4 q (5.62) 4 q 2 − m2 + i (2π)

Also in momentum space we can make use of a diagrammatic expansion with the following rules: 135

• For each propagator draw a line with associated momentum p (see Fig. 5.4.1). .

:

i ___________ 2 _ p m2 + iε

Fig. 5.4.1 - The propagator in momentum space. • For each factor (−iλ/4!) draw a vertex with the convention that the momentum flux is zero (see Fig. 5.4.2). .
p1 p
4

λ : _ i __ , p1 + p 2 + p 3 + p4 = 0 4!

p

2

p

3

Fig. 5.4.2 - The vertex in momentum space. • To get Gc draw all the topological inequivalent diagrams after having fixed the external legs. The number of ways of drawing a given diagram is its topological weight. The contribution of such a diagram is multiplied by its topological weight. • After the requirement of the conservation of the four-momentum at each vertex we need to integrate over all the independent internal four-momenta with weight d4 q (5.63) (2π)4 A more systematic way is to associate a factor iλ − (2π)4 δ 4 ( pi ) 4! i=1
4 (n)

(5.64)

to each vertex and integrate over all the momenta of the internal lines. This gives automatically a factor
n

(2π) δ (
i=1

4 4

pi ),

(n = numero linee esterne)

(5.65)

corresponding to the conservation of the total four-momentum of the diagram. By using these rules we can easily evaluate the 2nd order contribution of Fig. 5.4.3. 136

.

p1

q1 q2 q3

p

2

Fig. 5.4.3 - One of the second order contributions to Gc . We get 4×4×3×2× − iλ 4!
2

(2)

i 2 p2 − m2 + i

3

i=1

d4 qi i 2 (2π)4 qi − m2 + i p2 1 i − m2 + i

·(2π)4 δ 4 (p1 − q1 − q2 − q3 )(2π)4 δ 4 (q1 + q2 + q3 − p2 ) = ·
i=1

1 λ2 (2π)4 δ 4 (p1 + p2 ) 2 6 (p1 − m2 + i )2
3

d4 qi i 2 (2π)4 qi − m2 + i

(2π)4 δ 4 (p1 − q1 − q2 − q3 )

(5.66)

5.5

Power counting in λϕ4

We start by going to the euclidean formulation. Let us denote the time by tR . The relation with the euclidean time is tE = itR . Recalling from the Section 4.9 that for the harmonic oscillator of unit mass one has (we have defined xR = (tR , x), xE = (tE , x)) i∆F (xR ) = = d3 p 2 ∆(tR , ωk )eip · x = (2π)3 d4 p e−ipE xE
E (2π)4

d3 p 2 DE (tE , ωk )eip · x (2π)3 (5.67)

p2 + m2 E

where pE = (EE , p), and we have used eq. (4.237) for unit mass and with ν = EE . From this we get the following modifications to Feynman’s rules in the euclidean version • Associate to each line a propagator p2 1 + m2 (5.68)

137

• To each vertex associate a factor − λ 4! (5.69)

This rule follows because the weight in the euclidean path-integral is given by e−SE instead of eiS . The other rules of the previous Section are unchanged. The reason to use the euclidean version is because in this way the divergences of the integrals become manifest by simply going in polar coordinates. We will now in the position to make a general analysis of the divergences. For simplicity we will consider only scalar particles. First of all consider the number of independent momenta in a given diagram, or the number of ”loops”, L. This is given, by definition, by the number of momenta upon which we are performing the integration, since they are not fixed by the four-momentum conservation at the vertices. Since one of this conservation gives rise to the overall four-momentum conservation we see that the number of loops is given by L = I − (V − 1) = I − V + 1 (5.70)

where I is the number of internal lines and V the number of vertices. As in Section 2.3 we will be interested in evaluating the superficial degree of divergence of a diagram. To this end notice that • The L independent integrations give a factor of space-time dimensions.
L k=1

dd pk where d is the number

• Each internal line gives rise to a propagator containing two inverse powers in momenta I 1 (5.71) 2 p + m2 i=1 i Given that, the superficial degree of divergence is defined as D = dL − 2I (5.72)

Notice that D has to do with a common scaling of all the integration variables, therefore it may well arise the situation where D < 0 (corresponding to a convergent situation), but there are divergent sub-integrations. Therefore we will need a more accurate analysis that we will do at the end of this Section. For the moment being let us stay with this definition. We will try now to get a relation between D and the number of external legs in a diagram. By putting VN equal to the total number of vertices with N lines we have (see Section 2.3) NVN = E + 2I 138 (5.73)

with E the number of external legs and I the number of internal ones (recall we are dealing with scalar particles only). Eliminating I between this relation and the previous one, we get 1 (5.74) I = (NVN − E) 2 from which D = d(I −VN +1)−2I = (d−2)I −d(VN −1) = This relation can be rewritten as 1 D = d − (d − 2)E + 2 In particular, for d = 4 we have D = 4 − E + (N − 4)VN (5.77) N −2 d−N 2 VN (5.76) d−2 (NVN −E)−d(VN −1) (5.75) 2

For the λϕ4 theory this relation becomes particularly simple since then D depends only on the number of external legs D =4−E (5.78)

Therefore the only diagrams superficially divergent, D ≥ 0, are the ones with E = 2 and E = 4. As noticed before, this does not prove that a diagram with E > 4 or D < 0 is convergent. Let us discuss more in detail this point. Let us consider the diagram of Fig. 5.5.1, and let us suppose that the diagrams 1 and 2 have degrees of superficial divergence D1 and D2 respectively. Since they are connected by n lines, we have n − 1 independent momenta, and therefore .

1 D1 n

2 D2

Fig. 5.5.1 - Two diagrams connected by n-lines. D = D1 + D2 + 4(n − 1) − 2n = D1 + D2 + 2n − 4 (5.79)

Therefore we may have D < 0 with D1 and D2 positive, that is both divergent. However for n > 1 we have D ≥ D 1 + D2 (5.80) 139

we need D1 or D2 to be negative enough to overcome the other. The idea is simply to look for the reducibility of the diagram. .4.5.5. Then. We begin with the 3-particle reducibility. . and therefore it cannot give rise to divergences. 1 xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx N Fig. 3-particle reducible diagrams.5.5. 5. We will now stay with the λϕ4 theory. we must have N > 4. for the nature of the vertex it is impossible to have a connected 4-particle irreducible diagram. 5.3. In fact.3 . In Fig. the one-particle reducible diagrams can be easily reconstructed from the 1PI’s.2 we give an example of a one-particle reducible diagram. therefore D = D 1 + D2 + 2 = 3 − N (5.81) In order the superficial degree of divergence is negative.2 . After that we look for the diagrams having D < 0 but yet divergent. In analogous way we look for two-particle reducible diagrams with D < 0 and we extract contributions of the forms given in Fig 5.5.One-particle reducible Feynman’s diagram. xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx Fig. To exclude the case n = 1 means to consider only the one-particle irreducible (1PI) Feynman’s diagrams. In this case we have D1 = 0 and D2 = 1 − N. we can consider the 2-.Three-particle reducible Feynman’s diagram. in order to have a complete classification of the divergences.and therefore in order to have D < 0 with 1 or 2 being divergent diagrams. 140 . In this diagram we have a primitively divergent four point function and a diagram with D < 0 which can be again analyzed in the same way. 5. 5. In this case we look for diagrams of the type of Fig. Furthermore. The reason of excluding these diagrams from our considerations is that the momentum of the internal line connecting the two diagrams is fixed.

In the first case we have D1 = 2 and D2 = 2 − N.4 . therefore D = D 1 + D2 = 4 − N (5. If it happens that all the divergent diagrams (in the primitive sense specified above) correspond to terms which are present in the original lagrangian.82) Again this requires N > 4. we say that the theory is renormalizable. In the second case we have D1 = 0 and D2 = 2 − N. This turns out to be a general theorem due to Weinberg. 5.or 2-particle reducible parts containing two-and four-point divergent functions. 1 xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx N 1 xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx xxxxxxxxx N Fig.. By proceeding in this way we see that truly convergent diagrams do not contain hidden two and four point divergent functions. therefore D = D 1 + D2 = 2 − N (5. What we have shown is that for the λϕ4 theory in 4 dimensions a 1PI Feynman diagram is convergent if D < 0 and it cannot be split up into 3.83) requiring N > 2. In general we can look at this question by studying the equation for the superficial divergence that we got before 1 D = d − (d − 2)E + 2 141 N −2 d−N 2 VN (5.5.Two-particle reducible Feynman’s diagram. In the case of the λφ4 theory the Weinberg’s theorem tells us that all the divergences rely in the two and four point functions. valid in any field theory and it says that a Feynman diagram is convergent if its superficial degree of divergence and that of all its subgraphs are negative.84) . The second diagram has again D2 < 0 and we can iterate the procedure.

6 Regularization in λϕ4 In this Section we will evaluate the one-loop contributions to the two.85) We see that D increases with VN for N > 4. and we recall that for the scalar field we have dim[ϕ] = From this it follows at once that dim[m] = 1. dim[λ] = 4 − 2ω (5. (5.92) (5. As discussed in Section 2. must be less or equal to 4. An interesting case is the one of the theory λϕ3 in 6 dimensions. ϕN .and three-point ones. meaning that the dimension N of the vertex operator.two.88) In our unit system we have (/ = c = 1). Therefore we have infinite divergent diagrams and the theory is not renormalizable.90) d −1=ω−1 2 (5. and therefore the action S2ω should be h dimensionless. We can have a renormalizable scalar theory only for N ≤ 4.91) The Feynman’s rules are slightly modified 142 . let us consider the action in dimensions d = 2ω S2ω = d2ω x 1 1 λ ∂µ ϕ∂ µ ϕ + m2 ϕ2 + ϕ4 2 2 4! (5. here too it is convenient to introduce a dimensionful parameter µ in order to be able to define a dimensionless coupling λnew = λold (µ2 )ω−2 Then the action will be (λ = λnew ) S2ω = d2ω x 1 1 λ ∂µ ϕ∂ µ ϕ + m2 ϕ2 + (µ2 )2−ω ϕ4 2 2 4! (5.and fourpoint Green’s functions by using the dimensional regularization.3 this allows to evaluate the dimensions of the fields. To this end.87) and the only divergent functions are the one. In d = 6 we have dim[ϕ] = 2 and again in order to have renormalizability we must have that the dimensions of the vertex operator must be less or equal to 6.86) 5. For d = 6 we get D = 6 − 2E + (2N − 6)VN and for N = 3 D = 6 − 2E (5.For d = 4 we recall D = 4 − E + (N − 4)VN (5.89) As we did in QED for the electric charge.

1 .94) • The coupling λ goes into λ(µ2 )2−ω . we will do all the calculation in the euclidean version.1.• The scalar product among two four-vectors becomes a sum over 2ω components.93) pi ) = 1 (5.97) 32π 2 2−ω m2 143 .95) p p 1 p 2 Fig. p2 1 1 1 T2 2 + m2 p1 + m2 (5. This is given by (2π)2ω δ (2ω) (p1 + p2 ) .6. We start our calculation from the first order contribution to the two-point function corresponding to the diagram (tadpole) of Fig. 5. Also. with T2 = 1 (−λ)(µ2 )2−ω 2 λ m2 2 (4π)2 4πµ2 m2 1 πω λ 1 d2ω p = − (µ2 )2−ω Γ(1 − ω) 2 1−ω 2ω p2 + m2 2ω (2π) 2 (2π) (m ) 2−ω = − Γ(1 − ω) (5.6.96) For ω → 2 we get T2 ≈ − λ m2 1 4πµ2 + ψ(2) + · · · + · · · × (−1) 1 + (2 − ω) log 2 2 2 (4π) m 2−ω 4πµ2 1 λ + ψ(2) + log = m2 +··· (5. Remember that in Section 2. • In the loop integrals we will have a factor d2ω p (2π)2ω and the δ function in 2ω-dimensions will be defined by d2ω pδ (2ω) ( i (5. 5.The one-loop contribution to the two-point function in the λϕ4 theory.5 we gave the relevant integrals both in the euclidean and in the minkowskian version.

.6. 5.100) [(k 2 + m2 ) + (1 − x)(p2 − 2k · p)]2 0 dx [x(k 2 m2 ) 1 the denominator can be written as (k 2 + m2 ) + (1 − x)(p2 − 2k · p) = [k − (1 − x)p]2 + m2 + p2 (1 − x) − p2 (1 − x)2 = [k − (1 − x)p]2 + m2 + p2 x(1 − x) (5.2 .98) d2ω k 1 1 .6.101) Since our integral is a convergent one.99) We reduce the two denominator to a single one.102) . This expression has a simple pole at ω = 2 and a finite part depending explicitly on µ2 . through the equation (see Section 2.The one-loop contribution to the four-point function in the λϕ4 theory. p 1 p-k p4 p 2 k p 3 Fig.2. with an amplitude 4 (2π) δ with 1 T4 = (−λ)2 (µ2 )4−2ω 2 2ω (2ω) (p1 + p2 + p3 + p4 ) p2 k=1 k 1 T4 + m2 (5.6) 1 1 = 2 + m2 (k − p)2 + m2 k = = 1 + + (1 − x) ((k − p)2 + m2 )]2 0 1 1 dx 2 + m2 ) + (1 − x) (k 2 + m2 − 2k · p + p2 )]2 [x(k 0 1 1 dx (5. it is possible to translate k into k − p(1 − x) obtaining T4 = λ2 2 4−2ω (µ ) 2 1 dx 0 1 d2ω k 2ω (2π) [k 2 + m2 + p2 x(1 − x)]2 144 (5. The corresponding diagram is given in Fig. 5. Let us consider now the one-loop contribution to the four-point function. (2π)2ω k 2 + m2 (k − p)2 + m2 p = p1 + p2 (5.

107) depending on the renormalized parameters m and λ and field ϕ.103) (5.By performing the momentum integration we get λ2 T4 = (µ2 )4−2ω 2 In the limit ω → 2 T4 ≈ 1 1 dx λ2 2 2−ω 1 + (2 − ω) log µ2 (µ ) + ψ(1) 2 2−ω 2 0 (4π) m2 + p2 x(1 − x) × 1 − (2 − ω) log 4π 1 2 1 m2 + p2 x(1 − x) 2 2−ω λ + ψ(1) − dx log ≈ (µ ) 32π 2 2 − ω 4πµ2 0 1 0 dx 1 Γ(2 − ω) ω 2 + p2 x(1 − x)]2−ω (4π) [m (5. and in the counterterms contribution 1 1 δλ Lct = δZ∂µ ϕ∂ µ ϕ + δm2 ϕ2 + µ4−2ω ϕ4 2 2 4! The sum of these two expressions 1 (λ + δλ) 4−2ω 4 1 µ ϕ L = (1 + δZ)∂µ ϕ∂ µ ϕ + (m2 + δm2 )ϕ2 + 2 2 4! 145 (5.7 Renormalization in the theory λϕ4 In order to renormalize the theory λϕ4 we proceed as in QED.106) 5. a 1+a−1 0 The final result is T4 ≈ (µ2 )2−ω − a>0 (5. expressed in terms of bare couplings and bare fields in a part 1 1 λ Lp = ∂µ ϕ∂ µ ϕ + m2 ϕ2 + µ4−2ω ϕ4 2 2 4! (5.108) .109) (5. That is we split the original lagrangian.105) 1 4πµ2 λ2 + ψ(1) + 2 + log 32π 2 2 − ω m2 1+ 1+ 4m2 p2 4m2 p2 4m2 1 + 2 log p +1 −1 +··· (5.104) Also the x integration can be done by using the formula √ 1 √ 1+a+1 4 dx log 1 + x(1 − x) = −2 + 1 + a log √ .

The contributions to the teo-point connected Green’s function.2. 5. the effect of the first order contribution is to produce a mass correction determined by T2 .7. xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xx = xx xx + xx xx xx + + xx xx xx xx xx xx xx xx + xxxxxx xxxxxx Fig. all the 1PI contributions we get the following expression for the two-point connected Green’s function (see Fig. if we redefine couplings and fields as follows 1/2 ϕB = (1 + δZ)1/2 ϕ ≡ Zϕ ϕ (5.113) L = ∂µ ϕB ∂ µ ϕB + m2 ϕ2 + ϕ4 B B 2 2 4! B The counter-terms are then evaluated by the requirement of removing the divergences arising in the limit ω → 2. p2 ) = (2π)2ω δ 2ω (p1 + p2 ) c 1 1 1 + 2 T +··· 2 2 2 p2 + m2 +m p1 + m 1 1 ≈ (2π)2ω δ 2ω (p1 + p2 ) 2 (5.116) 146 . except for a finite part which is fixed by the renormalization conditions.112) m2 + δm2 −1 = (m2 + δm2 )Zϕ = 1 + δZ λ + δλ −2 λB = µ4−2ω = µ4−2ω (λ + δλ)Zϕ 2 (1 + δZ) m2 B In fact 1 λ0 1 (5.1) G(2) (p1 .7. p2 ) = (2π)2ω δ 2ω (p1 + p2 ) c p2 1 1 1 1 + 2 (−δm2 ) 2 2 −T 2 +m p1 + m p1 + m2 2 (5.gives back the original lagrangian. 5. 2 m2 2 64π 2 µ ϕ2 (5. We choose the counterterm as 2 1 2 2 λ m2 δm ϕ = + F1 ω. 5.110) (5.114) p1 + m2 − T2 p2 1 Therefore. From the results of the previous Section. by summing up .7.115) with = 4 − 2ω and F1 an arbitrary dimensionless function Adding the contribution of the counterterm to the previous result we get G(2) (p1 .1 .111) (5. The divergent part can be taken in care by the counterterm δm2 which produces an additional contribution to the one-loop diagram. represented in Fig.

3 (again we have included only 1PI diagrams). at the order λ we have G(2) (p1 . However we will discuss more in detail the renormalization conditions in the next Chapter when we will speak about the renormalization group.119) This expression has a pole in Minkowski space. u = (p1 + p4 )2 (5.118) The finite result at the order λ is then G(2) (p1 . 5. as we shall see later.7. Its diagrammatic expansion at one loop is given in Fig.117) Of course the same result could have been obtained by noticing that the counterterm modifies the propagator by shifting m2 into m2 +δm2 .121) we get 4 G(4) (p1 .122) 3λ 32π 2 + ψ(1) + 2 − log 147 m2 1 − A(s.. and we can define the renormalization condition by requiring that the inverse propagator at the physical mass is p2 + m2 phys (5.The Feynman’rule for the counterterm δm2 . Notice that at one loop there is no wave function renormalization. From the results of the previous Section and introducing the Mandelstam invariants s = (p1 + p2 )2 .2 . p2 . Now the combination δm2 −T2 is finite and given by δm2 − T2 = λ m2 m2 F1 − ψ(2) + log 32π 2 4πµ2 (5. xx x = _ δm2 Fig. p4 ) c × (−µ4−2ω )λ 1 − = (2π) δ 2 2ω (2ω) (p1 + p2 + p3 + p4 ) p2 k=1 k 1 + m2 (5. t = (p1 + p3 )2 . Consider now the four-point connected Green’s function.120) This choice determines uniquely the F1 term. t. 5. p2 ) = (2π)2ω δ 2ω (p1 + p2 ) c 1 p2 + m2 1 + 1 λ 32π 2 m F1 − ψ(2) + log 4πµ2 2 (5. but this comes about at two loops. u) 2 4πµ 3 . p2 ) = (2π)2ω δ 2ω (p1 + p2 ) c p2 1 + m2 1 + δm2 − T2 (5. p3 .7.

.7. p4 ) c = (2π) δ 2ω (2ω) (p1 + p2 + p3 + p4 ) p2 k=1 k 1 × (−µ2−4ω )λ (5. p3 . p2 . 5. t. p4 ) = (2π)2ω δ (2ω) (p1 + p2 + p3 + p4 ) c × 1− 3λ 32π 2 −G1 + ψ(1) + 2 − log p2 k=1 k 1 (−µ2−4ω )λ + m2 (5. 5. t.3 . p3 .123) The divergent part can be disposed off by the counterterm δλ 4−2ω 4 3λ2 2 1 m2 µ + G1 ω. with A(s.4. = _ (µ2 )2 _ω δλ Fig. 5.4 . u) = z=s.The Feynman’s rule for the couterterm δλ. 2 ϕ = µ4−2ω 4! 4! 32π 2 µ ϕ4 (5. By adding the counterterm we get 4 G(4) (p1 ..t. xx xxxxxxx xx xx xxxxxxx xxxxxxx xxxxxxx xx xxxxxxx xx xxxxxxx xx xxxxxxx xx xxxxxxx xx xx xx xx xx x x xx x x xx xx xx xx x x x = + xx xx xx + xx xx xx xx x x + xx xx xx xx xx + xx xx xx xx xx xx xx xx xx xx Fig. s = 4m2 .The one-loop expansion for the four-point Green’s function.u 4m2 log 1+ z 1+ 1+ 4m2 z 4m2 z +1 −1 (5.124) with G1 an arbitrary dimensionless function. This counterrtem produces an additional contribution with a Feynman rule represented in Fig.125) m2 1 − A(s. t = u = 0) to be equal to the lowest phys order value 4 G(4) (p1 .126) 2 +m 148 .7.7. p2 . u) 2 4πµ 3 A possible renormalization condition is to fix the scattering amplitude at some value of the invariants (for instance.

5 4 2 m2 3λ2 m2 2 + ψ(2) + G1 − log 4(16π 2 )2 4πµ2 • Contribution from the diagram d) of Fig 5.127) • Contribution from the diagram b) of Fig 5.5 2 m2 4 λ2 m2 2 + ψ(1) + F1 − log 4(16π 2)2 4πµ2 (5. We will examine only the two-point function.5 − 1 2 6m2 6m2 λ2 p + 2 + 6(16π 2 )2 2 m2 3 + ψ(1) − log 2 4πµ2 (5.7.128) • Contribution from the diagram c) of Fig 5.7.7. 5. conservation of the four-momentum.7.130) (5. xx xx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx xx xx = xx xx xx xx + xx xx xx xx xx xx + xx xx + xx xx + xx xx xx xx + xx xx xx xx x x + a) b) + xx xx + xx x x c) d) Fig.5 − λ2 4(16π 2 )2 4 2 + 2 ψ(2) + ψ(1) − 2 log m2 4πµ2 (5.129) 149 .7. including also the counterterm contributions is given in Fig. By omitting the external propagators and the delta-function of . 5.7.The two-loop expansion for the two-point Green’s function.5 .5. Its complete expansion up to two loops. the various two-loop contributions are (notice that a counterterm inside a loop gives a second order contribution as a pure twoloops diagram): • Contribution from the diagram a) of Fig 5.In any case G1 is completely fixed by this condition. Without entering into the calculation we want now show how the renormalization program works at two loops.

F2 . 5. The non trivial part of 150 . 2 ) 24(16π 2 )2 µ (5. ∂µ ϕ∂ µ ϕ and ϕ4 . Fig. and therefore they can be reproduced by operators as ϕ2 . 2 µ (5. What makes the theory renormalizable is the fact that the new divergences that one encounters can all be disposed off by modifying the counterterms δm2 and λ. Notice also that we are not introducing more and more freedom. (5.133) and F2 a new dimensionless function. that is 1 δZ∂µ ϕ∂ µ ϕ 2 with the choice m2 2 λ2 δZ = − + H2 (ω. The previous procedure can be easily extended to higher and higher orders in the expansion in the coupling constant. Also. This depends from the fact that the divergent terms are constant or quadratic in the external momentum. which are really two conditions: a) the position of the pole. All the functions F1 . b) the residuum at the pole. This expression is still divergent but at the order λ2 .Adding together these terms we find − p2 2λ2 1 λ2 1 λ2 + m2 [F1 + 3G1 − 1] + m2 12(16π 2 )2 (16π 2 )2 2 2(16π 2 )2 (5.115) and 1 δm2 = m2 2 λ2 2(16π 2 )2 4 2 1 m2 + (F1 + 3G1 − 1) + F2 ω.7. Therefore we can make finite the diagram by modifying at the second order the δm2 counterterm 1 δm2 + δm2 ϕ2 (5.134) (5. G1 H2 are finite in the limit ω → 2. The further divergence in p2 can be eliminated through the wave function renormalization counterterm. the function H2 is fixed by the renormalization condition on the inverse propagator.132) 1 2 2 with δm2 given in eq. because only particular combinations of these functions appear in the counterterms.131) where we have used ψ(2) − ψ(1) = 1. .6 .135) with H2 dimensionless.Insertion of counterterms in multi-loop diagrams. and are precisely these combinations which are fixed by the renormalization conditions.

the renormalization program is just the last one. In fact they cancel in adding up 2 the four two-loop contributions. 5.6. 151 . These diagrams contain 1/ terms coming from the counterterms times log p2 terms from the loop integration. If present the log p2 terms would be a real disaster since they are non-local in configuration space and as such they cannot be absorbed by local operators. However if one makes all the required subtractions. In fact consider diagrams of the type represented in Fig. where we have taken out all the possible external lines. For instance. that is one subtracts correctly also the diagrams of the type in Fig. it is possible to show that the log p2 terms cancel out. 5.7.7.6. In the previous two-loop calculation we had no such terms but we got their strict relatives log(m2 /4πµ2 ). in the expression for δm2 there are not such terms.

These considerations can be condensed into a differential 152 . In the previous relation we must regard the renormalized parameters as function of the bare ones. and. λB and mB ). On the other hand we can express the bare parameters in terms of the renormalized ones. In the λϕ4 case. ω) (n) −n/2 n/2 (6. at the same time. p2 . · · · . λ. renormalizing the operators through the wave function renormalization (ϕB = 1/2 Zϕ ϕ). λB . pn ) defined by removing the propagators on the external legs and omitting the delta-function expressing the conservation of the four-momentum n n (2π) δ where 2ω (2ω) ( i=1 pi )Γ (n) (p1 . ω) = Zϕ Γ(n) (p1 . · · · . this means λ and m depending on λB and mB . pn . p2 . p2 . pn . mB . Of course. However the right hand side of the previous equation does not depend on the choice of the parameter µ.Chapter 6 The renormalization group 6.1 The renormalization group equations For the following considerations it will be convenient to introduce the one-particle irreducible truncated n-point functions Γ(n) (p1 . · · · . obtaining the relation −n/2 ΓB (p1 . m. pn ) i=1 DF (pi )−1 (6.1) DF (p) = d2ω xeipx 0|T (ϕ(x)ϕ(0))|0 (6. This implies that also the functions Γ(n) do not depend on µ except for the field −n/2 rescaling factor Zϕ . pn ) = (n) G1P I (p1 . · · · . in a renormalizable theory the Γ(n) ’s are finite by construction.3) The factor Zϕ originates from a factor Zϕ coming from the renormalization (n) −1 factors in Gc and from n factors Zϕ from the inverse propagators. µ.2) is the exact euclidean propagator. p2 . p2 . We have two ways of evaluating these functions. · · · . we can start from the original lagrangian and evaluate them in terms of the bare couplings constants (in the case of λϕ4 .

pn ) c = i=1 d2ω xi ei(p1 x1 + · · · + pn xn ) 0|T (ϕ(x1 ) · · · ϕ(xn )|0 conn. · · · . m. Introducing a common scale s for the momenta we get (from the Euler theorem) µ ∂ ∂ ∂ +s +m − (4 − n + (n − 2)) Γ(n) (spi . ω) = 0 ∂µ ∂µ ∂λ ∂µ ∂m 2 ∂µ (6.equation obtained by requiring that the total derivative with respect to µ of the right hand side of eq. λ. therefore n G(n) (p1 .9) (6. It is convenient to define the following dimensionless coefficients µ µ ∂λ m = β λ. m. m. λ.4) The content of this relation is that a change of the scale µ can be compensated by a convenient change of the couplings and of the field normalization. p2 . it follows that also the derivative of log Zϕ is finite. .7) has dimensions n(ω − 1) − 2nω = −n(ω + 1). (6. In dimensions d = 2ω the delta function has dimensions −2ω and dim[ϕ] = ω − 1.3) vanishes ∂ ∂λ ∂ ∂m ∂ n ∂ log Zϕ (n) +µ +µ − µ Γ (p1 . . ω m ∂µ µ (6.8) Therefore Γ(n) must be a homogeneous function of degree 4 − n + (n − 2)/2 in the dimensionful parameters pi .10) . ω 2 ∂µ µ m µ ∂m = γm λ. (6.6) We can try to eliminate from this equation the term ∂/∂µ. ) = 0 ∂µ ∂s ∂m 2 153 (6. . To this end let us consider the dimensions of Γ(n) . Notice that since Γ(n) is finite in the limit ω → 2. pn . µ. and we get dim Γ(n) − 2ω = −n(ω + 1) + 2n → dim Γ(n) = n + ω(2 − n) and putting = 4 − 2ω dim Γ(n) = 4 − n + (n − 2) 2 (6. λ. The propagator DF (p) has dimensions 2(ω − 1) − 2ω = −2.5) obtaining µ ∂ ∂ ∂ +β + γm m − nγd Γ(n) (p1 . m e µ. · · · . ω) = 0 ∂µ ∂λ ∂m (6. p2 . pn . µ. µ. · · · . ω ∂µ µ m 1 ∂ log Zϕ µ = γd λ.

(6.16) (6.therefore it will be possible to express them as a Laurent series in the renormalized parameters ∞ λB = µ a0 + k=1 ∞ ak k (6. γm and γd . b0 and c0 may depend on but only in an analytical way. µ. In order to use the renormalization group equations we need to know more about the structure of the functions β. ) = 0 ∂s ∂λ ∂m (6. and allows us to evaluate them at momenta spi once we know them at momenta pi . Recalling from the previous Section the expressions for the counterterms at the second order in λ δλ = δm2 = 3λ2 2 + G1 32π 2 (6. m. To this end one needs to know the coefficients β.18) 154 .17) we get for the bare parameters λB = µ λ+ 3λ2 32π 2 2 + G1 + O(λ3 ) (6.14) with the coefficients functions of the dimensionless parameters λ and m/µ. γd and γm .15) λ 1 2 λ2 4 + F1 + m2 m2 2 + (F1 + 3G1 − 1) + F2 2 2 )2 32π 2(16π δZ = − 2 λ2 + H2 2 )2 24(16π (6. For this purpose one needs to specify the renormalization conditions since these coefficients depend on the arbitrary functions F1 . λ. Notice that a0 .Since everything is finite for → 0 we can take the limit and eliminate ∂/∂µ between this equation and eq.11) This equation tells us how the Γ(n) scale with the momenta.13) Z ϕ = c0 + k=1 ck k (6. We have seen that the counterterms are singular expressions in .12) m2 B =m 2 b0 + k=1 ∞ bk k (6. · · ·. G1 .6) −s ∂ ∂ ∂ +β + (γm − 1)m − nγd + 4 − n Γ(n) (spi .

since it is very difficult to solve the group renormalization equations when the coefficients β. They have considered the dependence of Γ(n) on the physical mass.21) (6.24) (6. it is possible to define a scheme where they do not depend on m/µ.22) (6. Furthermore in their equation the terms γm and µ/∂µ do not appear. γm e γd depend on both variables λ and m/µ.m2 = m2 1 + B + λ2 2(16π 2)2 λ2 + 24(16π 2)2 Zϕ = 1 − λ 32π 2 4 2 2 + F1 1 + (F1 + 3G1 − 1) + F2 + H2 2 + O(λ3 ) + H2 + O(λ3 ) (6. 155 . This is an essential feature.23) (6.27) 3λ2 + O(λ3 ) 16π 2 λ λ2 1 F1 + (F2 + H2 ) + O(λ3 ) b0 = 1 + 2 2 )2 32π 2(16π 12 b1 = λ λ2 λ2 + (F1 + 3G1 − 1) + + O(λ3 ) 2 2 )2 2 )2 16π 2(16π 12(16π b2 = c0 = 1 − 2λ2 + O(λ3 ) 2 )2 (16π λ2 H2 + O(λ3 ) 24(16π 2)2 λ2 + O(λ3 ) 12(16π 2 )2 c1 = − Notice that the dependence on m/µ comes only from the arbitrary functions defining the finite parts of the counterterms.26) (6. Callan and Szymanzik have considered a different version of the renormalization group equations.20) Therefore the coefficients are given by a0 = λ + a1 = 3λ2 G + O(λ3 ) 2 1 32π (6. In that case one can show that β and γd depend only on λ.19) 2 λ2 24(16π 2 )2 (6. In their place there is a inhomogeneous term in the large momentum limit.25) (6. On the other hand since these coefficients depend on the renormalization conditions. Another option would be to try to solve the equations in the large momentum limit where the mass dependence can be neglected. This is easily understood since the divergent terms originate from the large momentum dependence where the masses can be neglected.

F2 · · ·.33) . the finite part of Lct can be absorbed into a redefinition of the parameters appearing in L since both have the same operatorial structure.32) Γ(4) (p1 . u) = 6 for pi = 0.30) In the two point function we have omitted the finite terms coming from the second order contributions for which we have given only the divergent part. The renormalization group equations then tell us how to change the couplings when we change µ.31) (6.29) 1 m2 − A(s. p2 . p4 ) = −µ λ 1 − • A first possibility is to renormalize at zero momenta.29) and (6.and four-point truncated amplitudes Γ(2) (p) = λ2 H 2 p2 24(16π 2)2 m2 λ + m2 1 + F1 − ψ(2) + log 32π 2 4πµ2 λ2 + F2 + · · · 2(16π 2 )2 1− 3λ 32π 2 ψ(1) + 2 − log (6. t. Furthermore we have to fix a scale µ at which we define the parameters λ and m. Recall that the lagrangian has been decomposed as L = Lp + Lct (6.6. p4 ) = −µ λA where m2 and λA are quantities to be fixed by comparing some theoretical A cross-section with the experimental data. in order to leave invariant the physical quantities. p2 . let us now discuss various possibilities of renormalization conditions (sometimes one speaks of subtraction conditions) Γ(4) (p1 . In turn these are fixed by relating them to some observable quantity as a scattering cross-section. p3 . (6. A F1 = ψ(2) − log m2 A . This is another way of saying that the finite terms in Lct are fixed once we assign λ and m. Then. u) − G1 4πµ2 3 (6. we get A H2 = 0.28) therefore. As we have said the procedure for fixing m and λ is largely arbitrary and we will exhibit various possibilities. t. To do that it is convenient to write explicitly the expression for the two. 4πµ2 156 GA = ψ(1) − log 1 m2 A 4πµ2 (6. p3 . that is requiring Γ(2) (p) p2 ∼0 = p2 + m2 A pi =0 (6.30) and using A(s.2 Renormalization conditions Let us consider again the renormalization conditions. comparing with eqs. The renormalization procedure introduces in the theory a certain amount of arbitrariness in terms of the scale µ and of the arbitrary functions F1 .

for small (6.Another possibility would be to fix the momenta at some physical value in the Minkowski space. As a consequence a popular choice is to subtract at an arbitrary value of the momenta.39) 157 .34) (6. For instance one requires Γ(2) (p) = p2 + m2 . bi and ci appearing in the Laurent series defining the bare couplings turn out to be mass independent.37) λB = µ λ + k k=1 Differentiating with respect to µ and recalling that λB does not depend on this variable. • To choose the subtraction point at zero momenta is generally dangerous for massless particles (in this way one would introduce spurious infrared divergences). In this way the coefficients ai . GB = ψ(1)+2−log B2 − A(M 2 .35) Notice that the point chosen for the four-point function is a non-physical one. let us start with the bare coupling ∞ ak (λ) (6. In this case we get m2 m2 1 B . F1 = ψ(2)−log • The last way of renormalizing we will consider is due to Weinberg and ’t Hooft and the idea is simply to put equal to zero all the arbitrary functions F1 . B Γ(4) (p1 . p2 = M 2 s = t = u = M2 (6. M 2 . · · · at each order in the expansion.40) k k k=1 k=1 dak (λ) dλ → 0. we get ∞ 0= where λ+ k=1 ak (λ) k +µ ∂λ 1+ ∂µ ∞ ak (λ) k (6.38) k=1 ak (λ) = Since β = µ∂λ/∂µ is analytic for we can write β = A + B . B B H2 = 0. For instance. p4 ) = −µ λB . The Weinberg and ’t Hooft prescription allows to evaluate the renormalization group equations coefficients in a very simple way. F2 . p2 .36) This way of renormalize has however the problem of leaving an explicit dependence on the mass in the coefficients of the renormalization group equations. obtaining ∞ ∞ ak (λ) ak (λ) 0= λ+ + (A + B ) 1 + (6. M 2 ) 1 4πµ2 4πµ 3 (6. p3 . For instance we could require Γ(−p2 = m2 ) = 0 where m is ¯ ¯ (4) the physical mass and then fix Γ with all the momenta on shell.

50) and let us consider the following partial differential equation (we shall see that the renormalization group equations can be brought back to this form) ∂ ∂ F (xi .46) d a1 dλ (6. let us define t = log s (6. t) ∂t ∂xi (6.43) (6. The function λ(µ) is known as the running coupling constant.51) The general solution to this equation is obtained by using the method of the ”characteristics”. That is one considers the integral curves (the characteristic curves) defined by the ordinary differential equations dxi (t) = βi (xi (t)) dt 158 (6.49) This equation can be integrated in order to know how we have to change λ when we change the value of the scale µ.48) From the expression of the previous Section for a1 at the second order in λ.47) (6. t) = βi (xi ) F (xi . In fact. Its knowledge. we get µ 3λ2 ∂λ = β(λ) = ∂µ 16π 2 (6. together with the knowledge of m(µ) allows us to integrate easily the renormalization group equations.44) (6.45) (6.52) .41) (6.42) (6.and identifying the coefficients of the various powers in B+λ=0 a1 + A + Ba1 = 0 ak+1 + Aak + Bak+1 = 0 from which B = −λ A=− 1−λ We then get β(λ) = − 1 − λ and for → 0 β(λ) = − 1 − λ and 1−λ d dak ak+1 = dλ dλ d a1 − λ dλ d a1 dλ 1−λ d a1 dλ (6.

58) with xi (t) satisfying eq. (6. t) = e dt γ(xi (t )) 0 F (xi .53) (6. For the Γ(n) we get (t = log s) −n Γ(n) (spi . the general solution is ¯ F (xi . m(s). (6. the amplitudes Γ(n) do not scale only with the trivial factor s4−n due to their dimensions. but that they undergo also a non trivial scaling. λ. By putting t (6.60) ∂m(s) = m(s)(γm (λ(s)) − 1) (6.56) G(xi . µ)s4−ne with s s ∂λ(s) = β(λ(s)) ∂s s γd (λ(s )) 1 ds s (6. due to γd = 0. µ) = Γ(n) (pi . t) = βi (xi ) G(xi .51) and therefore the general solution will be t G(xi . t) ∂t ∂xi with γ(xi ) a given function. t) = e dt γ(xi (t )) 0 ψ(xi (t)) (6. In fact we can check that ψ satisfies the differential equation dxi (t) ∂ψ ∂ψ ∂ ψ(xi (t)) = = βi ∂t dt ∂xi ∂xi In our case the equation contains a non-derivative term ∂ ∂ G(xi .61) ∂s and λ(s = 1) = λ.59) (6.55) (6. t) satisfies eq. Then. which is necessary to compensate the variations of λ and m with the scale. 159 .54) is an arbitrary function to be determined by the condition at t = 0 required for F . m. t) (6. This result tells us that when we perform a re-scaling on the momenta. 0) x x (6.57) we see that F (xi . t) = ψ(xi (t)) where ψ(¯i ) = F (¯i . λ(s). m(s = 1) = m. t) + γ(xi )G(xi .52).with given boundary conditions xi (0) = xi .

7 we recall that δZ3 = C = − and therefore (6.65) 1 ¯ ˆ ¯ˆ d2ω x ψ(i∂ − m)ψ − eψ Aψ − Fµν F µν 4 1 2 = 3ω − 2 (6.71) ak k (6. we want to check that the running coupling obtained via the renormalization group equations is the same that we have defined in Section 2.67) 4 giving ¯ˆ ¯ ¯ˆ L = i(1 + δZ2 )ψ ∂ψ − (m + δm)ψψ − e(1 + δZ1 )µ /2ψ Aψ − 1 + δZ3 Fµν F µν (6. in particular.2.66) Lp = ψ(i∂ − m)ψ − eµ /2 ψ Aψ − Fµν F µν 4 and δZ3 ¯ˆ ¯ ¯ˆ Lct = iδZ2 ψ ∂ψ − δmψψ − eδZ1 µ /2 ψ Aψ − Fµν F µν (6.72) 160 . First of all let us start with the QED lagrangian in d = 2ω dimensions (neglecting the gauge fixing terms) S2ω = Since ¯ˆ dim[ψ Aψ] = ω − 1 + 2 ω − it follows dim[e] = 2 − ω Therefore we define enew = eold (µ)ω−2 = eold µ− /2 (6. the identification eB = µ /2 e (1 + δZ1 ) = µ /2 e(1 − δZ3 /2) (1 + δZ2 )(1 + δZ3 /2) e2 6π 2 (6.70) e2 12π 2 We proceed as in the previous Section by defining eB = µ /2 e 1 + eB = µ /2 e+ k (6.64) (6.62) (6. From Section 2. In particular.3 Application to QED In this Section we will apply the previous results to QED.63) As before we write the lagrangian as Lp + Lct.69) where we have used δZ1 = δZ2 .68) 4 After re-scaling of the fields one gets back the original lagrangian via.6. with (we ignore here the gauge fixing term) 1 ¯ ˆ ¯ˆ (6.

77) where K is an integration constant which can be eliminated by evaluating e at another momentum scale Q2 . 6.74) e3 12π 2 The differential equation defining the running coupling constant is β(e) = µ e3 ∂e(µ) = β(e) = ∂µ 12π 2 (6.79) which agrees with the result obtained in Section 2.75) (6.73) → 0 we get a1 (6. We get 1 1 1 Q2 − = − log 2 α(Q2 ) α(µ2 ) 3π µ or α(Q2 ) = 1− α(µ2 ) α(µ2 ) 3π (6. and in the limit a1 = β(e) = µ from which 1 ∂e =− ∂µ 2 1−e d de with (6.49) we get − 3 µ 1 1 = log − λ(µ) 16π 2 µ s λs 1 1− 161 3λs 16π 2 µ log µs (6.2.4 Properties of the renormalization group equations µ 3λ2 dλ = + O(λ3 ) dµ 16π 2 (6.e3 12π 2 Requiring that eB is µ-independent.82) .80) Integrating eq.81) where µs is a reference scale and λs = λ(µs ). Therefore λ(µ) = λs (6.76) Integrating this equation we obtain 1 e2 (µ2 ) =− 1 log µ2 + K 2 12π (6.78) log Q2 µ 2 (6. (6.

1)).states.82) should be valid for large values of λ. which can be rather described by bound states of quarks as mesons and baryons. assuming the validity of the expression for the β function up to large momenta. let us expand the β(λ) around its zero β(λ) ≈ (λ − λF )β (λF ) It follows µ dλ = (λ − λF )β (λF ) + · · · dµ (6.From this expression we see that starting from µ = µs . Of course there are no motivations why eq.82) we see that. The pole in eq. and it will become infinite at some value of µ. (6. λ increases with µ. than the running coupling is always increasing.85) 162 . If this happens for a finite µ we say that we have a Landau pole at that scale.and out.84) (6. The point λF is called a fixed point since if we start with the initial condition λ = λF than λ remains fixed at that value. However in this case the coupling is increasing at large distances and this would be a problem for our approach based on th in. however if it would be negative the theory would become perturbative at large mass scales or at small distances. To study the behaviour of λ around the fixed point.82) is called the Landau pole (see Section 2. The increasing of λ is related to the sign of the β(λ) function. or equivalently at large distances. (6.2 in the case of QED).83) which is a very big scale if λs is small. • If β(λ) > 0 for any λ. Let us now discuss several possible scenarios starting from a theory where β = 0 for λ = 0. here the fields that are useful to describe the dynamics at short distances (the quark fields) cannot describe the asymptotic states. In this case one could solve the renormalization group equations at large momenta by using the perturbative expansion at a lower order to evaluate the β and the other coefficients. 6. For the λϕ4 theory and for QED the β function is positive. the coupling constant becomes infinite at the value of µ given by 16π 2 µ = µs e 3λs (6. • Suppose that β(λ) is positive at small λ and that it becomes negative going through a zero at λ = λF (see Fig. Suppose that we start with a small λ in such a way that the perturbative expansion is sensible. In this theory we can give a definite sense to the asymptotic states. As we shall see this is the situation in QCD. (6. However increasing µ we will need to add more and more terms to the expansion due to the increasing of λ. Therefore the λϕ4 theory allows a perturbative expansion only at small mass scales.4. Going back to the eq.

4. The large scale (or small distance) behaviour of a field theory of . starting from λs > λF the theory would become perturbative at large scales (see Fig.2)).4.86) (6.The behaviour of β(λ) giving rise to an ultraviolet fixed point. In the present case the sign is negative since β(λ) > 0 for λ < λF and β(λ) < 0 for λ > λF . this type would be dictated by the value of λF .1).4.4. Otherwise.87) Notice that the sign of β (λF ) plays here a very important role. 6. However this is an infrared (IR) fixed point. For large values of µ we have that λ → λF independently on the initial value λs .1 . β(λ) λS λS λF λS λ Fig. since the coupling goes to zero for 163 .The behaviour of λ close to a UV fixed point. 6. In particular if λF 1 and λs < λF we would be always in the perturbative regime. 6.2 . 6. from which dλ dµ = β (λF ) λ − λF µ Integrating this equation we get λ − λF = λs − λF µ µs β (λF ) (6. For this reason λF is referred to as a ultraviolet (UV) fixed point (see Fig. λ λF µS µ Fig. notice that λ = 0 is a fixed point since β(0) = 0.. About this point.

In higher dimensions other theories enjoy this property.4. whereas it goes away from λs for increasing µ. becomes zero at λF and then positive. • Let us consider now the case of β(λ) < 0 for small values of λ and decreasing monotonically. Therefore λF is an IR fixed point (see Fig. We differentiate the eq.90) and we see that λ → λF µ → 0 in a way independent on λs . suppose β(λ) = −aλ2 with a > 0. Integrating the equation for the running coupling we get λ(µ) = λs 1 µ 1 + aλs log µs (6. Notice that also this time the running coupling has a pole. (6.88) In this case λ is a decreasing function of µ and goes to zero for µ → ∞ (that is λ = 0 is a UV fixed point). whereas a UV fixed point is said to be attractive.89) (6. but at a smaller scale than µs 1 − µ = e aλs µs Therefore the coupling increases at large distances. We say also that the fixed point is repulsive since λ goes away from the fixed point for increasing µ. • Finally let us consider the case of β(λ) < 0 for small values of λ.91) k=1 bk k ∞ + β(λ) k=1 bk k (6. For instance.3)). For instance in d = 6 the theory λϕ3 is asymptotically free. Let us now derive the perturbative expressions for γm and γd using the same method followed for β(λ).µ → 0 independently on the initial value λs . ` possibile ricavare. 6. This property is known as ”asymptoyic freedom” and it holds for non-abelian gauge theories that we will describe later on in the course. these are the only 4-dimensional theories to share this property.13) with e respect to µ (notice that in this renormalization scheme b0 = 1) bk ∂λ ∂m 1+ 0 = 2m + m2 k ∂µ ∂µ k=1 from which 0 = 2γm 1 + k=1 ∞ ∞ ∞ bk k (6. Then β (λF ) > 0.92) 164 . By expanding β(λ) around λF we get (by the same analysis we did before) λ − λF = λs − λF µ µs β (λF ) (6.

93) or (6. (6.. (6.46) we find 2γm − λ db1 =0 dλ dbk d 2γm bk − 1−λ dλ dλ 1 db1 γm = λ 2 dλ λ db1 dbk bk − dλ dλ 1−λ d dλ a1 − λ dbk+1 =0 dλ a1 − λ dbk+1 dλ (6.96) Let us now consider Zϕ (recall that although this quantity is infinite in the limit → 0.24) for b1 we get for γm λ 5 γm (λ) = − 2 32π 12 λ 16π 2 2 + O(λ3 ) (6. where we have made use of the definition γm = (µ/m)(∂m/∂µ).95) Using the eq.46) 1 dc1 γd = − λ 2 dλ 165 (6. By differentiating eq. Using eq. 6. (6.97) ck k ck 1 γd = (A + B ) k 2 k=1 (6.3 .99) .94) and (6. γd is finite).The behaviour of λ close to an IR fixed point.14) we find (c0 = 1) 1 dλ 1 ∂Zϕ = µ Zϕ γ d = µ 2 ∂µ 2 dµ or 1+ k=1 ∞ ∞ ck k k=1 ck 1 = β(λ) k 2 k=1 ∞ ∞ (6.4. λ λF µS µ Fig. (6.98) Using again the expression for β(λ) (see eq.

59) Γ(n) (spi . msγm (λF )−1 . the integral is dominated by the points where there is a zero of the β function.104) Then. integrating the following equation (see eq. 166 . To justify why we have assumed that in the evaluation of γm and γd the only important contribution comes from λ = λF we notice that the integral s 1 ds γm (λ(s )) s (6.102) s ds we have (6.100) At the lowest order. let us study the behaviour of Γ(n) when the theory has a UV fixed point at λ = λF . (6.61)) dm = m(γm (λF ) − 1) (6. λF . λ. for large values of s (see eq. (6.101) To conclude this Section. barring the case in which γm and β have simultaneous zeroes. µ) (6. µ) → s(4−n−nγd (λF )) Γ(n) (pi .106) can be rewritten (using ds/s = dλ/β) λ(s) dλ λ γm (λ ) β(λ ) (6.103) m(s) = m(1)sγm (λF )−1 Analogously 1 s ds γd (λF ) = γd (λF ) log s s (6.λ dc1 dck dck+1 = λck − dλ dλ dλ 1−λ d a1 dλ (6.105) Therefore. (6. Then. m. From this equation we understand also the reason why γd is called the anomalous dimension.27) we get γd = 1 12 λ 16π 2 2 + O(λ3 ) (6. if 1 − γm (λF ) > 0 we can safely neglect the mass on the right hand side of this equation. In non abelian gauge theories there is a UV fixed point at λ = 0 where γm = 0 and therefore one can neglect the masses at large momenta. using eq.107) Therefore. Since for large scales λ → λF we may suppose that γm and γd go to γm (λF ) e γd (λF ) respectively.

Notice also that (see eq. To study better this problem. the expression (6.5 The coefficients of the renormalization group equations and the renormalization conditions We have already noticed that the coefficients of the renormalization group equations depend on the subtraction procedure.109) where the dot stands for the derivative with respect to m/µ.112) m (γm − 1)a0 = 0 ˙ µ 3λ G1 32π 2 ≈ −λ 1 − (6. In particular.115) A= 16π 2 µ 32π 2 and for → 0 m 3λ2 ˙ 3λ2 G1 + (6.5) m ∂ m (6. let us consider. when using subtraction procedure 167 . Therefore.12) for the bare coupling ∞ λB = µ a0 + k=1 ak k (6. obtaining a0 + Ba0 = 0 a1 + Aa0 + Ba1 + At the second order in λ B = −λ 1 + and Aa0 = − 3λ G1 32π 2 1+ 3λ G1 16π 2 −1 (6. using the first two renormalization conditions considered in Section 6.108) Differentiating with respect to µ we find ∞ a0 + k=1 ak k ∂λ +µ ∂µ ∞ a0 + k=1 ak k ∂(m/µ) +µ ∂µ ∞ a0 + ˙ k=1 ak ˙ k =0 (6. Similar expression can be obtained for γm e γd .113) 3λ2 3λ +λ 1− G1 2 16π 32π 2 3λ m 3λ2 ˙ G1 − (γm − 1) 8π 2 µ 32π 2 (6.114) Since the right hand side is at least of the second order in λ.110) = (γm − 1) µ ∂µ µ µ With a familiar procedure.111) (6. neglecting higher order terms we get m 3λ2 ˙ 3λ2 G1 + (6.116) β= 16π 2 µ 32π 2 We see directly how the β function depends on the subtraction procedure.6. (6.2 a mass dependence of the coefficients is introduced. we put β = A + B . for generic renormalization conditions.

such that the coefficients of the renormalization group equations depend explicitly on the masses. Of course. For instance we get expressions of the type λ =f Therefore we obtain β =µ ∂λ ∂(m/µ) ˙ m ∂λ =µ f +µ f = βf + (γm − 1)f˙ ∂µ ∂µ ∂µ µ (6.118) λ. m µ = λ + O(λ2 ) (6. Another possibility is to choose the arbitrary scale µ such that m/µ can be neglected. It is possible to show that also the sign of the β function is schemeindependent. Therefore the presence of a fixed point does not depend on the scheme used (at least in this approximation).117) Working in a region where we can neglect the mass we have β (λ ) = β(λ)f (λ) (6.119) and we see that a fixed point at λ = λF implies a fixed point in λ . 168 . changing the renormalization procedure will induce a finite renormalization in the couplings. it is convenient to look for solutions in a region of momenta where the masses can be neglected.

since the path integral is strictly related to the classical description of the theory. (7. θi θj θk . In fact it turns out that it is possible to have a ”classical” (called pseudoclassical) description in a way such that its quantization leads naturally to anticommutation relations. · · · .1). It follows (0) (1) Gn = Gn ⊕ Gn (7. Gn . · · · . i. the elements constructed by all the possible products among the generators form a basis of Gn : 1.2) Notice that there are no other possible elements since 2 θi = 0 (7. θ1 θ2 · · · θn (7. n (7. it could be nice to have such a ”classical” description for fermionic systems.1) Then.Chapter 7 The path integral for Fermi fields 7. θi θj . θi . A Grassmann algebra Gn is a vector space of dimensions 2n and it can be described in terms of n generators θi (i = 1. n) satisfying the relations θi θj + θj θi ≡ [θi .4) 169 . whereas the subset (1) generated by the product of an odd number of generators is the odd part. This type of mathematical objects form the so called Grassmann algebras. · · · .1 Fermionic oscillators and Grassmann algebras The Feynman’s formulation of the path integral can be extended to field theories for fermions.3) from eq. The subset of elements of Gn generated by the product of an even (0) number of generators is called the even part of the algebra. Gn . j = 1. Before proceeding further let us introduce some concepts about these algebras which will turn out useful in the following. In this theory one introduces anticommuting variables already at the classical level. However. First of all we recall that an algebra V is nothing but a vector space where it is defined a bilinear mapping V × V → V (the algebra product). θj ]+ = 0.

If x ∈ Gn . t = ϕ(x)|ϕ(x). · · · . t). and let us recall that this is nothing but a set of harmonic oscillators. However it can be thought as an odd element of a Grassmann algebra. y are homogeneous elements of Gn (that is to say even or odd). and let us try to diagonalize the corresponding annihilation operator a|θ = θ|θ (7. Therefore the eigenvalue θ cannot be an ordinary number.10) . deg(x) = +1.9) Then. This is true also for Fermi fields. θn 170 (7. a† i j = 0. (7.A decomposition of this kind is called a grading of the vector space. + ai . if it is possible to assign a grade to each element of the subspaces. In the case of field theories one takes eigenstates of the field operators in the Heisenberg representation φ(x. we will formulate the path integral starting directly from quantum mechanics. · · · .6) and as a consequence it is possible to diagonalize simultaneously at fixed time the set of operators φ(x.5) (1) (0) Since we have no time to enter into the details of the classical mechanics of the Grassmann variables. since if x. θn = θi |θ1 . t). except for the fact that these oscillators are of Fermi type. If x ∈ Gn . t)|ϕ(x).7) (7. deg(x) = −1.8). as it foillows from eqs. t)]− = 0 (7. 2.8) For greater simplicity take the case of a single Fermi oscillator. then deg(xy) = deg(x)deg(y) (7. we must have θ2 = 0. To gain a better understanding of this point let us stay for the moment with free fields. that is they are described by creation and annihilation operators satisfying the anticommutation relations [ai . Notice that this grading is also consistent with the algebra product. For Fermi fields such a possibility does not arise since they anticommute at equal times. Therefore it is not possible to define simultaneous eigenstates of Fermi fields. If we have n Fermi oscillators and we require ai |θ1 . φ(y. from a2 = 0. aj ]+ = a† . In this case the grading is as follows 1. Let us recall that in the Feynman’s formulation one introduces states which are eigenstates of a complete set of observables describing the configuration space of the system. t this is possible since the fields commute at equal times [φ(x. a† j + = δij (7.

18) In the algebra G2 we can define normalized eigenstates † |θ = eθa |0 N(θ. θ∗ ) = N(θ.16) Here we have used the left-derivative on the Gn defined by ∂ θi θi · · · θik = δi1 . θ∗ ) 171 (7. Then it is convenient to extend the algebra G1 generated by θ to a comples Grassmann algebra.iθi1 θi3 · · · θik ∂θi 1 2 + · · · + (−1)k−1 δik . · · · . θn But ai aj = −aj ai and therefore [θi .14) where we have used since θ2 = a2 = 0.i θi1 θi3 · · · θik−1 (7. θn ) = θj (ai |θ1 . θn ) = θj θi |θ1 . θj ]+ = 0 (7. mathematically such an object is called a modulus over the algebra). We have also (7.iθi2 θi3 · · · θik − δi2 .11) We see that the equations (7.17) Suppose we want to define a scalar product in the vector space of the eigenstates of a (notice that these are vector spaces with coefficients in a Grassmann algebra. generated by θ and θ∗ .10) are meaningful if we identify the eigenvalues θi with the generators of a Grassmann algebra Gn . The operation ” ” is defined as the complex conjugation on the ordinary numbers and it is an involution (automorphism of the algebra with square one) with the property ∗ ∗ ∗ (θi1 θi2 · · · θik )∗ = θik · · · θi2 θi1 (7.we get ai (aj |θ1 .19) where N(θ. · · · .13) where |0 is the ground state of the oscillator. In fact a|θ = a(1 + θa† )|0 = θaa† |0 = θ(1 − a† a)|0 = θ|0 = θ(1 + θa† |0 = θ|θ † eθa = 1 + θa† (7. θ∗ ) (7. θ∗ ) ∈ G2 and the phase is fixed by the convention (0) N ∗ (θ. G2 .12) (7.9) can be solved easily † |θ = eθa |0 (7.20) . In this case the equation (7. · · · .15) a† |θ = a† (1 + θa† )|0 = a† |0 = ∂ ∂ (1 + θa† )|0 = |θ ∂θ ∂θ (7.

In fact (a† a)2 = a† aa† a = a† (1 − a† a)a = a† a from which a† a(a† a − 1) = 0 (7. we can build up only the 2 eigenstate with eigenvalue 1.29) whereas 1] (7.21) the normalization factor N(θ. a† + = 1). θ∗ ) is determined by the condition ∗ † 1 = θ|θ = N 2 (θ. θ∗ )e−θθ (7. 1 0 a† = 0 1 0 0 (7. θ∗ ) 0|eθ a (7. Therefore the space of the eigenstates of a† a is a two-dimensional space.25) Starting from the state with zero eigenvalue for a† a.30) 172 .27) 0 1 (7.23) It follows ∗ N(θ. θ∗ )(1 + θ∗ θ) = N 2 (θ. θ∗ ) = eθθ /2 Let us recall some simple properties of a Fermi oscillator. a† |0 . The state with occupation number equal to 1 is |1 = In this basis the state |θ is given by |θ ∗ † ∗ = eθθ /2 eθa |0 = eθθ /2 (1 + θa† )|0 ∗ ∗ θ = eθθ /2 (|0 + θ|1 ) = eθθ /2 1 ∗ θ| = eθθ /2 [ θ∗ 1 0 (7. θ∗ ) 0|eθ a eθa |0 = N 2 (θ.24) (7. since a† |0 = 0.The corresponding bra will be ∗ θ| = N(θ. θ∗ ) 0|(1 + θ∗ θaa† )|0 ∗ = N 2 (θ. By defining the ground state as |0 = one sees easily that a= 0 0 .22) (7.28) (7. the occupation number operator has eigenvalues 1 and 0. a2 = 0 and a.26) (one writes down a as the most general 2 × 2 matrix and then requires a|0 = 0. θ∗ ) 0|(1 + θ∗ a)(1 + θa† )|0 = N 2 (θ.

35) by requiring dθ · θ = 1. (7. the previous integration rules can be rewritten as integration rules over the algebra G1 requiring [dθ.35) These relations determine the integration in a unique way. 173 dθ · 1 = 0 (7.30) satisfy 1 0 ≡ 12 (7. and therefore we would like its eigenstates to satisfy a completeness relation of the previous type.33) Therefore the integration must satisfy dθ∗ dθ · θθ∗ = 1 dθ∗ dθ · θ = dθ∗ dθ · θ∗ = dθ∗ dθ · 1 = 0 (7. since they fix the value of the integral over all the basis elements of G2 .29) and (7. To this end we recall that the crucial element in order to derive the path integral formalism from quantum mechanics is the completeness relation for the eigenstates of the position operator dq|q q| = 1 (7.37) (7. dθ∗ ]+ = [θ.32) dθ∗ dθ|θ θ| = 0 1 where the identity on the right hand side must be the identity on the two-dimensional vector space. Using the explicit representation for the states dθ∗ dθ|θ θ| = = = ∗ dθ∗ dθeθθ ∗ dθ∗ dθeθθ θ 1 θθ∗ θ∗ ( θ∗ θ 1 1) dθ∗ dθ θθ∗ θ∗ θ 1 + θθ∗ = 1 0 0 1 (7. dθ]+ = [θ∗ . dθ∗ ]+ = [θ. dθ∗ ]+ = 0 Then.34) and (7.34) (7. Therefore we will ”define” the integration over the Grassmann algebra G2 by requiring that the states defined in eqs. we can reproduce the rules (7. Let us now come to the definition of the integral over a Grassmann algebra. In this space the linear combinations of the vectors are taken with coefficients in G2 (or more generally in G2n ).31) What we are trying to do here is to think to the operators a as generalized position operators.We see that in order to introduce eigenstates of the operator a has forced us to generalize the ordinary notion of vector space on the real or on the complex numbers. The integral over a generic element of the algebra is then defined by linearity. dθ]+ = [θ∗ .36) . to a vector space on a Grassmann algebra (a modulus).

θn ) = 0 ∂θj 174 (7.2 Integration over Grassmann variables Let us consider the Grassmann algebra Gn .in (7. · · · .40) 7. · · · . · · · . · · · .···. Noticing also that for the antisymmetry we have fn (i1 . · · · .···. · · · . · · · . in )θi1 · · · θin = n!fn (1.42) (7.45) it follows dθn · · · dθ1 f (θ1 .41) Notice that the measure dθ is translationally invariant. dθj ]+ = 0 (7.43) dθf (θ + α) (7. · · · . in )θi1 · · · θin (7. θn = e i=1 with n θi a† i e i=1 |0 (7. This can be seen either by the observation that a derivative destroys a factor θi in each element of the basis and therefore dθn · · · dθ1 ∂ f (θ1 .46) An interesting property of the integration rules for Grassmann variables is the integration by parts formula. · · · .47) . Its generic element can be expanded over the basis elements as f (θ1 . n) (7. θn | = 1 (7.38) ∗ dθi dθi |θ1 .For n Fermi oscillators we get n ∗ θi θi /2 n |θ1 . In fact if f (θ) = a + bθ and (1) α ∈ G1 dθf (θ) = since’ dθbα = 0 (7. θn ) = f0 + i1 f1 (i1 )θi1 + i1 .12 f2 (11 .in fn (i1 . θj ]+ = [dθi . θn ) = n!fn (1.39) i=1 and integration rules dθi θj = δij Assuming also [dθi .44) Of course the coefficients of the expansion are antisymmetric. If we integrate f over the algebra Gn only the last element of the expansion gives a non vanishing contribution.i2 )θi1 θi2 +· · · i1 . · · · . n)θ1 · · · θn i1 . θn θ1 .

52) from which bij = (a−1 )ij dθn · · · dθ1 = det|a−1 |dθn · · · dθ1 (7. We see also that ∂θ det|a| = det (7.k bki ajh δhk = (ab)ij = δij (7.7).50) dθk bki (7.55) ∂θ and dθn · · · dθ1 = det−1 ∂θ dθn · · · dθ1 ∂θ (7.56) Therefore the rule for a linear change of variables is dθn · · · dθ1 f (θi ) = dθn · · · dθ1 det−1 ∂θ f (θi (θi )) ∂θ (7. Let us now consider the transformation properties of the measure under a change of variable. (7. θi θj = cik cjh θk θh = −cik cjh θh θk == −cjh cik θh θk = −θj θi where cij is the inverse of the matrix aij . For simplicity consider a linear transformation θi = aik θk (7. Therefore we require dθi θj = By putting dθi = k (7. since it can be seen easily that the previous transformation leaves invariant the multiplication rules.58) .either by the property that the integral is invariant under translations (see the previous Section and Section 4.k dθk bki ajh θh = h.54) where we have used the anticommutation properties of dθi among themselves.49) dθi θj = δij (7. In fact.48) k The eq.40) must be required to hold in each basis.57) This rule should be compared with the one for commuting variables dx1 · · · dxn f (xi ) = dx1 · · · dxn det 175 ∂x f (xi (xi )) ∂x (7.53) The transformation properties of the measure are (7.51) we obtain dθi θj = h.

59) The matrix A is supposed to be real antisymmetric AT = −A As shown in the Appendix a real and antisymmetric form ⎡ 0 0 · 0 λ1 ⎢ −λ1 0 0 0 · ⎢ ⎢ 0 0 0 λ2 · ⎢ 0 −λ2 0 · As = ⎢ 0 ⎢ ⎢ · · · · · ⎢ ⎣ · · · · · · · · · · (7.46) we get contribution to the integral only from the term in the n expansion of the exponential (θ T As θ ) 2 for n even. for n even I = = dθn · · · dθ1 eθ As θ /2 = T dθn · · · dθ1 ( θ ( n )! 2 2 1 1 T As θ ) 2 n 1 n dθn · · · dθ1 n ( )!λ1 · · · λ n θ1 · · · θn 2 ( )! 2 √2 = λ1 · · · λ n = detA 2 176 (7. for n even 2 2 1 T θ As θ = λ1 θ1 θ2 + · · · + λ n−1 θn−2 θn−1 . (7.64) The expression in the exponent is given by 1 T θ As θ = λ1 θ1 θ2 + · · · + λ n θn−1 θn .62) θi = j we get θT Aθ = (S T θ )AS T θ = θ SAS T θ = θ As θ and using det|S| = 1 I= dθn · · · dθ1 eθ As θ /2 T T T (7.63) (7. Then.61) ⎥ ⎥ · ·⎥ · ·⎦ · · by means of an orthogonal transformation S.j ≡ T dθn · · · dθ1 eθ Aθ/2 (7. In the case of n odd we get I = 0 since θ T As θ does not contain θn . As in the commuting case. let us do the following change of variables (S T )ij θj (7. the more important integral to be evaluated for the path integral approach is the gaussian one Aij θi θj /2 I= dθn · · · dθ1 e i.We see that the Grassmann measure transforms according to the inverse jacobian rather than with the jacobian itself.60) matrix can be reduced to the ⎤ · · · ·⎥ ⎥ · ·⎥ ⎥ · ·⎥ (7. for n odd 2 2 (7. Therefore.66) .65) From eq.

Since the measure is invariant under translations we find I(A. To evaluate this integral let us put θ = θ + A−1 χ We get 1 T 1 T θ Aθ + χT θ = (θ − χT A−1 )A(θ + A−1 χ) + χT (θ + A−1 χ) 2 2 1 T 1 = θ Aθ + χT A−1 χ 2 2 (7.69) Again. the result for fermionic variables is opposite to the result of the commuting case. χ) = √ T −1 T −1 dθn · · · dθ1 eθ Aθ /2 + χ A χ/2 = eχ A χ/2 detA T (7.68) with A a real and symmetric matrix.72) where we have used the antisymmetry of A and χT θ = −θ T χ. A can be diagonalized by an orthogonal transformation with the result J= T n 1 dx1 · · · dxn e−x Ad x /2 = (2π) 2 √ detA (7. In fact.67) This result is valid also for n odd since in this case det|A| = 0. A similar result holds for commuting variables.70) with χi Grassmann variables.73) again to be compared with the result for commuting variables I(A.71) (7.where we have used det|As | = det|A|.74) The previous equations can be generalized to complex variables † ∗ ∗ dθ1 dθ1 · · · dθn dθn eθ Aθ = det|A| (7. Another useful integral is I(A. J) = T T T −1 (2π) 2 dx1 · · · dxn e−x Ax/2 + J x = eJ A J/2 √ detA n (7. let us define J= T dx1 · · · dxn e−x Ax/2 (7. χ) = T T dθn · · · dθ1 eθ Aθ/2 + χ θ (7. Then √ T dθn · · · dθ1 eθ Aθ/2 = detA (7.75) 177 .

2 we get 1 T (7. We will give here only the result of this analysis. t = where and S= t θ ∗ (t )=θ ∗ . t |θ. θ∗ )eθ θ(t )/2 + θ (t)θ/2 + iS ∗ (7.80) (7.t ∗ D(θ.79) θ Aθ = −iη ∗ A12 η 2 and recalling that (remember that here one has to use the inverse of the jacobian) dθ2 dθ1 = idηdη ∗ we get T dθ2 dθ1 eθ Aθ/2 = i ∗ dηdη ∗ e−iη A12 η i θ2 = − √ (η − η ∗ ) 2 (7. Considering a Fermi oscillator.85) 178 . it would be possible to derive the corresponding path integral formulation.3 The path integral for the fermionic theories In this Section we will discuss the path integral formulation for a system described by anticommuting variables. one could start from the ordinary quantum mechanical description and then by using the completeness in terms of Grassmann variable.76) The second of these equations follows from the first one as in the real case.82) 7.77) and putting 1 θ1 = √ (η + η ∗ ). θ∗ ) = t dθdθ∗ (7.78) (7. Starting from T dθ2 dθ1 eθ Aθ/2 = A12 (7.t θ(t)=θ.83) D(θ. It turns out that the amplitude among eigenstates of the Fermi annihilation operator is given by θ . we will prove the first one for the case n = 1. θ∗ ) dt 2 (7.84) i ∗ ˙ ˙∗ (θ θ − θ θ) − H(θ.81) and ∗ dηdη ∗eη (−iA12 )η = −iA12 (7.and † † † † −1 ∗ ∗ dθ1 dθ1 · · · dθn dθn eθ Aθ + χ θ + θ χ = e−χ A χ det|A| (7.

86) where η and η are the external sources (of Grassmann character).where.89) (7. We will not insist here further on this discussion. the two point function is given by 1 δZ 1 ¯ 0|T (ψ(x)ψ(y))|0 = iSF (x − y) = (7. as it follows from the derivation made in the bosonic case. Correspondingly we can assign only the coordinates at a particular time. The expression is very similar to the one obtained in the commuting case if we realize that the previous formula is the analogue in the phase space using complex coordinates of the type q ± ip. ψ)e ¯ / ¯ [ψ(i∂ − m)ψ + η ψ + ψη] ¯ (7. but all we have done in the bosonic case can be easily generalized for Fermi fields.87) where the determinant of the Dirac operator has been included in the term at zero source. taking into account that the classical field of this case are anticommuting. We know that this is nothing but the Dirac propagator defined by / (i∂ − m)SF (x − y) = δ 4 (x − y) It follows −i Z[η. It should be noticed the particular boundary conditions for the θ variables originating from the fact that the equations of motion are of the first order.73) we get formally −i 1 −(i¯) η (iη) −i¯ η η / / i∂ − m i∂ − m Z[η.90) The Green’s functions are obtained by differentiating the generating functional with respect to the external sources (Grassmann differentiation). and correspondingly the generating functional for the free Dirac theory will be Z[η. η] = ¯ i ¯ D(ψ. In any case. Using the inte¯ gration formula (7.91) Z[0] δ η(x)δη(y) η=¯=0 ¯ η 0|0 SF (x) = lim + The T -product in this expression is the one for the Fermi fields. 179 1 d4 k e−ikx = lim 4/ −m+i →0+ (2π) k ¯ d4 xd4 y η(x)SF (x − y)η(y) (7. η] = Z[0]e ¯ = Z[0]e (7. for the Fermi oscillator H = θ∗ θ. since at the end we are interested in field theory where only the generating functional is relevant and where the boundary conditions are arbitrary (for instance we can take the fields vanishing at t = ±∞). and in any case do not affect the dependence of the generating functional on the external source. In this expression we have to specify how to take the inverse operator. In particular.88) . The extension to the Fermi field is done exactly as we have done in the bosonic case. η] = Z[0]e ¯ with k /+m d4 k e−ikx 4 k 2 − m2 + i →0 (2π) (7.

The idea is 180 . Q is a diagonal matrix having as eigenvalues the ¯ charges of the different fields measured in unit e. This invariance.Chapter 8 The quantization of the gauge fields 8. was called gauge invariance. The physical meaning of this invariance lies in the possibility of assigning the phase to the fields in an arbitrary way.3) It is also referred to as a U(1) symmetry. It looks more physical to require the possibility of assigning the phase in an arbitrary way at each space-time point. As a prototype we can consider the free Dirac lagrangian ¯ / L0 = ψ(x)[i∂ − m]ψ(x) which is invariant under the global phase transformation ψ(x) → ψ (x) = e−iQα ψ(x) (8. The free lagrangian (8.2) (8. This way of thinking is in some sort of contradiction with causality. These are transformations leaving invariant the action of the system and are characterized by a certain number of parameters which are independent on the space-time point. and Q(ψ2 ) = 2.1) If one has more than one field. formulated by Weyl in 1929. since it requires to assign the phase of the fields simultaneously at all space-time points.1) cannot be gauge invariant due to the derivative coming from the kinetic term. is invariant by choosing Q(ψ1 ) = Q(φ) = 1. without changing the observable quantities. This is a so called abelian symmetry since e−iαQ e−iβQ = e−i(α + β)Q = e−iβQ e−iαQ (8.1 QED as a gauge theory Many field theories possess global symmetries. with φ a scalar field. For instance. a term as ψ2 ψ1 φ.

(8.10) . ∂ν + ieQAν ] = ieQFµν with Fµν = ∂µ Aν − ∂ν Aµ transforms in the same way Fµν → e−iQα(x) Fµν eiQα(x) = Fµν 181 (8.8) (8.6) 1 = e−iQα(x) ∂µ + ieQ(Aµ − ∂µ α) ψ e from which The lagrangian density ¯ / ¯ ¯ Lψ = ψ[iD − m]ψ = ψ[iγ µ (∂µ + ieQAµ ) − m)ψ = L0 − eψQγ µ ψAµ 1 Aµ = Aµ + ∂µ α e (8. the covariant derivative undergoes a unitary transformation Dµ → Dµ = e−iQα(x) Dµ eiQα(x) Then.4) (8.9) is then invariant under gauge transformations.simply to generalize the derivative ∂µ to a so called covariant derivative Dµ having the property that Dµ ψ transforms as ψ. To construct the covariant derivative. We see also that by requiring local invariance we reproduce the electromagnetic interaction as obtained through the minimal substitution we discussed before. In order to determine the kinetic term for the vector field Aµ we notice that eq. in the following way Dµ = ∂µ + ieQAµ The transformation law of Aµ is obtained from eq. we need to enlarge the field content of the theory.4) [(∂µ + ieQAµ )ψ] = (∂µ + ieQAµ )ψ (x) = (∂ + ieQA )e−iQα(x) ψ µ µ (8.13) (8.5) will be invariant as the mass term under the local phase transformation. that is Dµ ψ(x) → [Dµ ψ(x)] = e−iQα(x) Dµ ψ(x) In this case the term ¯ ψDµ ψ (8.12) (8.7) (8. Dν ] = [∂µ + ieQAµ .11) (8. the gauge field Aµ . (8. by introducing a vector field.4) implies that under a gauge transformation. also the commutator of two covariant derivatives [Dµ . or under the local group U(1).

in general. The free lagrangian N L0 = a=1 ¯ / ψa (i∂ − m)ψa (8.15) ∂ψ. Lie(G).17) is invariant under the global transformation Ψ(x) → Ψ (x) = AΨ(x) (8. We notice also that gauge invariance prevents any mass term. The generators T A satisfy the algebra AB [T A . In a more general situation the actual symmetry could be a subgroup of U(N). Denoting by U the generic element of G.20) 182 .2 Non-abelian gauge theories The approach of the previous section can be easily extended to local non-abelian symmetries. (that is the vector space spanned by the infinitesimal generators of the group) in the fermion representation. The fields ψa (x) will belong. eliminating the infinitesimal parameter α.14) The gauge principle has automatically generated an interaction between the gauge field and the charged field. For instance.19) where T A denote the generators of the Lie algebra associated to G. So we will consider here the gauging of a subgroup G of U(N). T B ] = ifC T C (8. when the masses are not all equal. The complete lagrangian density is then 1 ¯ L = Lψ + LA = ψ[iγ µ (∂µ + ieQAµ ) − m]ψ − Fµν F µν 4 (8. Therefore. 1 M 2 Aµ Aµ .18) where A is a unitary N × N matrix.16) 8. we will write the corresponding matrix Uab acting upon the fields ψa as A U = e−iαA T .The last equality follows from the commutativity of Fµν with the phase factor. U ∈G (8. Also. and we have denoted by Ψ the column vector with components ψa . the photon field is massless. by using the Noether’s theorem we find the conserved current as ∂L ¯ jµ = δψ = ψγµ (Qα)ψ (8.µ from which. to some reducible representation of G. ¯ Jµ = ψγµ Qψ (8. We will consider the case of N Dirac fields. since the 2 local invariance implies the global ones.

A = 1.21) σA . that now the group does not need to be abelian. In the general case the T A ’s are N × N hermitian matrices that we will choose normalized in such a way that TA = 1 T r(T A T B ) = δ AB 2 (8.19).29) (8.26) implies Dµ Ψ → (∂µ + igBµ )U(x)Ψ = U(x)∂µ Ψ + U(x)[U −1 (x)igBµ U(x)]Ψ + (∂µ U(x))Ψ = U(x)[∂µ + U −1 (x)igBµ U(x) + U −1 (x)∂µ U(x)]Ψ therefore and U −1 (x)igBµ U(x) + U −1 (x)∂µ U(x) = igBµ i Bµ (x) = U(x)Bµ (x)U −1 (x) + (∂µ U(x))U −1 (x) g 183 (8. if G = SU(2). and we take the fermions in the fundamental representation.26) (8. in general A A A A e−iαA T e−iβA T = e−iβA T e−iαA T (8.27) (8. (8.22) 2 where σ A are the Pauli matrices. 2.31) . and therefore.24) Notice.25) Let us now proceed to the case of the local symmetry by defining again the concept of covariant derivative Dµ Ψ(x) → [Dµ Ψ(x)] = U(x)[Dµ ψ(x)] We will put again Dµ = ∂µ + igBµ where Bµ is a N × N matrix acting upon Ψ(x).23) To make local the transformation (8.30) (8. means to promote the parameters αA to space-time functions αA → αA (x) (8. For instance. 3 (8. Ψ= we have ψ1 ψ2 (8.AB where fC are the structure constants of Lie(G).28) The eq. In components µ Dab = δab ∂ µ + ig(B µ )ab (8.

33) (8. transforms as Ψ under gauge transformations.38) (8.41) (8. Bν ] in components µν F µν = FC T C (8. Dν ]Ψ) = igFµν Ψ = igFµν U(x)Ψ = U(x)([Dµ . ∂ν + igBν ] = ig(∂µ Bν − ∂ν Bµ ) − g 2 [Bµ . Dν ]Ψ ≡ igFµν Ψ (8.26). since it does not commute with the gauge transformation as in the abelian case Fµν = U(x)Fµν U −1 (x) The invariant kinetic term will be assumed as 1 LA = − T r[Fµν F µν ] 2 Let us now evaluate Fµν igFµν = [Dµ .34) The difference with respect to the abelian case is that the field undergoes also a homogeneous transformation.For an infinitesimal transformation U(x) ≈ 1 − iαA (x)T A we get 1 δBµ (x) = −iαA (x)[T A .39) (8.40) (8. Dν ] = [∂µ + igBµ . the transformation law is consistent with a Bµ linear in the generators of the Lie algebra.36) in virtue of the eq. that is (B µ )ab ≡ Aµ (T A )ab A The transformation law for Aµ becomes 1 AB δAµ = fC αA Aµ + ∂ µ αC C B g (8. In fact the quantity [Dµ .35) (8. The kinetic term for the gauge fields is constructed as in the abelian case. Bν ] or Fµν = (∂µ Bν − ∂ν Bµ ) + ig[Bµ .32) Since Bµ (x) acquires a term proportional to T A .42) 184 .37) This time the tensor Fµν is not invariant but transforms homogeneously. that is ([Dµ . (8. Dν ]Ψ) = U(x) (igFµν ) Ψ (8. Bµ (x)] + (∂µ αA (x))T A g (8.

with
µν AB FC = ∂ µ Aν − ∂ ν Aµ − gfC Aµ Aν C C A B

(8.43)

The main feature of the non-abelian gauge theories is the bilinear term in the preAB vious expression. Such a term comes because fC = 0, expressing the fact that G is not abelian. The kinetic term for the gauge field, expressed in components, is given by 1 µν FµνA FA (8.44) LA = − 4 A Therefore, whereas in the abelian case LA is a free lagrangian (it contains only quadratic terms), now it contains interaction terms cubic and quartic in the fields. The physical motivation lies in the fact that the gauge fields couple to everything which transforms in a non trivial way under the gauge group. Therefore they couple also to themselves (remember the homogeneous piece of transformation). To derive the equations of motion for the gauge fields, let us consider the total action ¯ / ¯ d4 x Ψ(i∂ − m)Ψ − g Ψγµ B µ Ψ + SA (8.45)
V

where SA = − and the variation of SA

1 2

d4 x T r(Fµν F µν )
V

(8.46)

δSA = −
V

d4 xT r(Fµν δF µν )

(8.47)

Using the definition (8.41) for the field strength we get δFµν = ∂µ δBν + ig(δBµ )Bν + igBµ δ(Bν ) − (µ ↔ ν) from which δSA = −2
V

(8.48)

d4 xT r[F µν (∂µ δBν + ig(δBµ )Bν + igBµ (δBν ))]

(8.49)

where we have taken into account the antisymmetry properties of Fµν . Integrating by parts we obtain δSA = −2
V

d4 xT r[−(∂µ F µν )δBν − igBµ F µν δBν + igF µν Bµ δBν ] d4 xT r [(∂µ F µν + ig[Bµ , F µν ]) δBν ]

= 2
V

=
V

d4 x (∂µ F µν + ig[Bµ , F µν ])A δAνA

(8.50)

185

where we used the cyclic property of the trace. By taking into account also the free term for the Dirac fields and the interaction we find the equations of motion ¯ ∂µ F µνA + ig[Bµ , F µν ]A = g Ψγ ν T A Ψ / (i∂ − m)Ψ = gγµ B µ Ψ

(8.51)

¯ From the first equation we see that the currents Ψγµ T A Ψ are not conserved. In fact the conserved currents turn out to be
A ¯ Jν = Ψγν T A Ψ − i[B µ , Fµν ]A

(8.52)

The reason is that under a global transformation of the symmetry group, the gauge fields are not invariant, said in different words they are charged fields with respect to the gauge fields. In fact we can verify immediately that the previous currents are precisely the Noether’s currents. Under a global variation we have
AB δAµ = fC αA Aµ , C B

δΨ = −iαA T A Ψ

(8.53)

and we get jµ = from which
µν AB ¯ j µ = Ψγ µ αA T A Ψ − FC fC αA AνB

∂L ∂L δΨ + δAνC ∂Ψ,µ ∂Aν,µ C

(8.54)

(8.55)

AB In the case of simple compact Lie groups one can define f ABC = fC with the property f ABC = f BCA . It follows µν FC f ABC AνB T A = i[B ν , Fνµ ] ≡ i[B ν , Fνµ ]A T A

(8.56) (8.57)

Therefore ¯ j µ = Ψγ µ αA T A Ψ − i[Bν , F νµ ]A αA After division by αA we get the Noether’s currents (8.52). The contribution of the gauge fields to the currents is also crucial in order they are conserved quantities. In fact, the divergence of the fermionic contribution is given by ¯ ¯ ¯ ¯ ∂ µ (Ψγµ T A Ψ) = −ig Ψγµ T A B µ Ψ + ig Ψγµ B µ T A Ψ = −ig Ψγµ [T A , B µ ]Ψ (8.58)

which vanishes for abelian gauge fields, whereas it is compensated by the gauge fields contribution in the non abelian case.

8.3

Path integral quantization of the gauge theories

As we have seen at the beginning of this course that the quantization of the electromagnetic field is far from being trivial, the difficulties being related to the gauge 186

invariance of the theory. In this Section we would like to discuss how these difficulties arise also in the path integral quantization. To this end let us recall the general scheme of field quantization in this approach. First of all we will deal only with the perturbative approach, meaning that the only thing that one needs to evaluate is the generating functional for the free case. The free case is defined by the quadratic part of the action. Therefore the evaluation of the path integral reduces to calculate a Gaussian integral of the type given in eqs. (7.73) and (7.73) for the fermionic and bosonic case respectively. For instance, in the bosonic case one evaluates I(A; J) =
T T T −1 (2π) 2 dx1 · · · dxn e−x Ax/2 + J x = eJ A J/2 √ detA
n

(8.59)

In our case the operator A appearing in the gaussian factor is nothing but the wave operator. The result of the integration is meaningful only if the operator A is nonsingular (it needs to have an inverse) otherwise the integral is not well defined. This is precisely the point where we run into problems in the case of a gauge theory. Remember from eq. (1.132) that the wave operator for the electromagnetic field in momentum space is given by (−k 2 g µν + k µ k ν )Aν (k) = 0 We see that the operator has a null eigenvector kν , since (−k 2 g µν + k µ k ν )kν = 0 (8.61) (8.60)

and therefore it is a singular operator. This is strictly related to the gauge invariance, which in momentum space reads Aµ (k) → Aµ (k) + kµ Λ(k) (8.62)

In fact, we see that the equations of motion are invariant under a gauge transformation precisely because the wave operator has kµ as a null eigenvector. Another way of seeing this problem is the following: in order to get physical results, as to evaluate S matrix elements, it is necessary to integrate over gauge-invariant functionals D(Aµ )F [Aµ ]eiS (8.63)

with F [Aµ ] = F [AΩ ] (here AΩ is the gauge transformed of Aµ ). But being the action µ µ and the functional F gauge invariant, it follows that the integrand is invariant along the gauge group orbits. The gauge group orbits are defined as the set of points, in the field space, which can be reached by a given Aµ via a gauge transformation. In other words, given Aµ , its orbit is given by all the fields AΩ obtained by varying the µ parameters of the gauge group (that is varying Ω), see Fig. 8.3.1. For instance, in the abelian case AΩ = Aµ + ∂µ Ω (8.64) µ 187

.

AΩ µ Aµ Ω A'µ A'µ

Fig. 8.3.1 - The orbits of the gauge group in the field space. Since the orbits define equivalence classes (or cosets) we can imagine the field space as the set of all the orbits. Correspondingly we can divide our functional integral in a part parallel and in one perpendicular to the orbits. Then, the reason why our functional integral is not well defined depends from the fact that the integrand along the space parallel to the orbits is infinite (the integrand being gauge invariant does not depend on the corresponding variables). However this observation suggests a simple solution to our problem. We could define the integral by dividing it by the integral along the part parallel to the orbit (notice that is the same as the volume of the gauge group). In order to gain a clear understanding of this problem let us consider the following very simple example in two dimensions
+∞

Z=
−∞

1 − a2 (x − y)2 dxdye 2

(8.65)

This is as reducing the space-time two two points with x and y the possible values of the field in those points. At the same time the exponent can be seen as an euclidean action with the (gauge) symmetry xΩ = x + Ω, yΩ = y + Ω (8.66)

The integral (8.65) is infinite due to this invariance. In fact this implies that the integrand depends only on x − y, and therefore the integration over the remaining variable gives rise to an infinite result. In order to proceed along the lines we have described above, let us partition the space of the field, that is the two-dimensional space (x, y) in the gauge group orbits. Given a point (x0 , y0), the group orbit is given by the set of points (xΩ = x + Ω, y Ω = y + Ω) corresponding to the line x − y = x0 − y0 , as shown in Fig. 8.3.2 We can perform the integral by integrating 188

.
y

(x Ω, y Ω)

(x 0 , y 0) x

f(x,y) = 0 x+y=c

Fig. 8.3.2 - The field space (x, y) partitioned in the orbits of the gauge group. We show also the line x + y = c and the arbitrary line f (x, y) = 0 (dashed). along the perpendicular space, the line x + y = c (c is a constant), and then along the parallel space. It is this last integral which gives a divergent result, since we are adding up the same contribution an infinite number of times. therefore the obvious way to do is to introduce as integration variables x − y and x + y, integrating only on x − y. Or, said in a different way, perform both integrals and dividing up by the integral over x + y, which cancels the divergence at the numerator. This can be done in a more systematic way by using the following identity
+∞

1=
−∞

dΩδ

1 (x + y) + Ω 2

(8.67)

Then we multiply Z per by this factor obtaining ⎤ ⎡ 1 2 2 − a (x − y) 1 ⎦ (x + y) + Ω e 2 Z = dΩ ⎣ dxdyδ 2

(8.68)

The integral inside the parenthesis is what we are looking for since it is evaluated along the line x+y = constant. Of course the result should not depend on the choice of this constant (it corresponds to translate the integration line in a way parallel to the orbits). In fact, by doing the change of variables x → x + Ω, y →y+Ω (8.69)

189

In fact by the transformation x → x + Ω . Then. y Ω ) = (x + y)/2 + Ω = 0. proceeding as in eq. y + Ω ) dΩδ [f (x + Ω + Ω .74) In the previous case we had f (x. y Ω ) = ∆f (x. y + Ω)] e 2 (8. (8. let us choose the line (see Fig. y) 1 = ∆f (x. Multiplying Z by the expression (8.70) As promised the infinite factor is now factorized out and we see also clearly that this nothing but the volume of the gauge group (Ω parameterize the gauge group. 8. y) = 1. y)δ [f (x + Ω.2) f (x. y + Ω ) 190 dΩδ [f (x + Ω. from which ∆f (x. y + Ω)] (8. y) is gauge invariant.73) Then.76) and changing the integration variable Ω → Ω + Ω 1 = ∆f (x + Ω . we define our integral as Z = Z = VG with VG = dΩ (8.77) .3. y + Ω + Ω )] (8.we get Z= ⎡ dΩ ⎣ dxdyδ ⎤ 1 − a2 (x − y)2 1 ⎦ (x + y) e 2 2 (8.74) we get ⎤ ⎡ 1 2 2 − a (x − y) ⎦ Z = dΩ ⎣ dxdy∆f (x. which allows us to define a gauge invariant function ∆f (x.75) Let us show that ∆f (x. y) dΩδ f (xΩ .67) we consider the following identity.71) In this way the integral is defined by integrating over the particular surface x+y = 0. y) = 0 (8. y → y + Ω we obtain 1 = ∆f (x + Ω . y + Ω)] (8. For instance. Of course we would get the same result integrating over any other line (or surface in the general case) having the property of intersecting each orbit only once. y) dΩδ [f (x + Ω. which in the present case is isomorphic to R1 ).72) 1 − a2 (x − y)2 1 (x + y) e 2 2 dxdyδ (8. y) = (x + y)/2 = 0 giving rise to f (xΩ .

79) Z = dΩ ⎣ dxdy∆f (x. As we have already noticed the singularity of A is related to the gauge invariance.79). Therefore we can again factorize the integration over Ω. Therefore we can write δ[f (xΩ . y)δ [f (x.Comparison with eq.74) gives ∆f (x. Clearly ∆f represents the answer of the gauge fixing f = 0 to an infinitesimal gauge transformation.82) ∂Ω Ω=0 where f Ω = f (xΩ . (8.80) The function ∆f (x. y)δ [f (x. y) crosses the gauge orbits only once. As a last remark. the equation f (xΩ . y Ω ) = 0 has the only solution Ω = 0. In fact. y) = Z= with A = a2 1 − xi Aij xj d xe 2 2 (8. y) = (x + y)/2. y Ω ). y) (called the Faddeev-Popov determinant) can be calculated directly from its definition. integrating over x − y only. means to integrate only on the eigenvectors of A corresponding to non zero eigenvalues. and the solution we have given here solves also the problem of the singularity of A. 191 . y)] e 2 and define Z = Z = VG 1 − a2 (x − y)2 dxdy∆f (x. y) = 0. since by hypothesis the surface f (x.81) giving rise to ∂f Ω (8.84) a singular matrix. y Ω )] = 1 ∂f ∂Ω Ω Ω=0 δ[Ω] (8. To this end notice that in (8. As a consequence. But in this case. y + Ω ) (8. (8. y) = ∆f (x + Ω . notice that our original integral can be written as ∆f (x.83) 1 −1 −1 1 (8. ∆f (x. y)] e 2 (8. y) appears multiplied by δ[f (x. being the sum of infinite gauge copies ⎤ ⎡ 1 2 2 − a (x − y) ⎦ (8.75) does not depend on Ω (that is is gauge invariant). y)]. This can be shown exactly as we did before in the case f (x.78) The integral inside the parenthesis in eq. we need to know ∆f only at points very close at the surface f (x.

93) and changing variables Ω Ω = Ω we get ∆−1 (aΩ µ f −1 )= dµ(Ω Ω )δ[fA (AΩ )] = µ 192 dµ(Ω)δ[fA (AΩ )] µ (8. is given by dµ(Ω) = A. Also.85) where n is the number of parameters of LieG. B B = 1. n (8. This can be obtained as follows. therefore it is enough to know the measure around the identity of the gauge group.Since the solution we have found to our problem is of geometrical type. abelian or not.92) having the usual properties extended to the functional case.89) It follows that for infinitesimal transformations dµ(ΩΩ ) = dµ(Ω Ω) = dµ(Ω) next we define the Faddeev-Popov determinant as in (8. defined by the identity 1= dµ(Ω)δ[Ω] (8.90) In this expression we have introduced a functional delta-function. In such a case the transformation Ω can be written as Ω ≈ 1 + iαA (x)T A Since at the first order in the group parameters ΩΩ ≈ 1 + i(αA + αA )T A (8. since ∆−1 [AΩ µ f −1 ]= dµ(Ω)δ[fA (A(Ω µ −1 Ω) ] (8.74) we need also a measure on the gauge group. We start by choosing a surface in the gauge field space given by fB (Aµ ) = 0. We have seen that we are interested in the solutions of fB (AΩ ) = 0 µ (8. To write down the analogue of equation (8.x dαA (x) (8.74) 1 = ∆f [Aµ ] dµ(Ω)δ[fA (AΩ )] µ (8.87) we see that the invariant measure of the group. We can see that ∆f is a gauge invariant functional.91) (8. · · · . it is very easy to generalize it to the case of a gauge field theory. around the identity.94) .88) (8. this surface should intersect the gauge orbits only once.86) around fB (Aµ ) = 0.

the Faddeev-Popov determinant δf [Aµ ] is the answer of the gauge-fixing to an infinitesimal gauge transformation.101) αA =0 = det −1 δfA (AΩ ) µ δαB αA =0 δ[αA ] (8.96) The functional measure is invariant under gauge transformations i Aµ → ΩAµ Ω−1 + (∂µ Ω)Ω−1 g (8. the only solution of fA (AΩ ) = 0 in this neighborhood is µ Ω = 1.102) . and the inhomogeneous part gives rise to a translation.Consider now the naive path integral Z= i D(Aµ )F [Aµ ]e LA d4 x (8. LA and ∆f [Aµ ]. The final expression that we get for Z is Z= δfA (AΩ ) µ D(Aµ )δ[fA (Aµ )]det δαB 193 i e αA =0 LA d4 x (8. We multiply Z by the identity (8. or αA = 0. Then by performing the change of variables Aµ → AΩ .97) since the homogeneous part has determinant one.99) The evaluation of ∆f goes as before. Since in the following F will not play any particular role we will discuss only the case F = 1. and µ using the gauge invariance of D(Aµ ). but all the following considerations are valid for an arbitrary gauge invariant F .100) Once again. therefore δ[fA (AΩ )] µ from which ∆f [Aµ ] = det δfA (AΩ ) µ δαB (8. we get ⎡ ⎤ 4 i LA d x ⎢ ⎥ δ[fA (Aµ )]∆f [Aµ ]⎦ Z = dµ(Ω) ⎣ D(Aµ )e (8.91) ⎡ ⎤ Z= ⎢ dµ(Ω) ⎣ i D(Aµ )e LA d4 x ⎥ δ[fA (AΩ )]∆f [Aµ ]⎦ µ (8. we define the functional integral as Z = Z→ VG Z = dµ(Ω) i D(Aµ )e LA d4 x δ[fA (Aµ )]∆f [Aµ ] (8.95) where F is a gauge invariant functional. Since in the previous equation it appears multiplied by δ[fA (Aµ )].98) Again.

the vacuum expectation value (v.e. does not depend on the gauge fixing fA = 0. of F .This expression gives the right measure to be used in the path integral quantization of the gauge fields.e. Consider the identity 0|T ∗(F [A])|0 f1 = D(Aµ )δ[f1 ]∆f1 eiS F [A] Ω dµ(Ω)δ[f2 ]∆f2 (8. If F is gauge invariant we get at once 0|T ∗(F [A])|0 f1 = 0|T ∗ (F [A])|0 f2 (8.’s of F [A] in due different gauges. Then 0|T ∗(F [A])|0 f1 = Ω dµ(Ω) 0|T ∗(δ[f1 ]∆f1 F [AΩ ])|0 f2 (8. (8. We will prove this result by showing the relation between the v.91)and defined f2 [A] ≡ f2 [AΩ ].105) Ω where we have used eq. but in terms of which we can easily construct the v.v.v.’s of gauge invariant quantities. For the purpose of building up the perturbative theory it is convenient to write the integrand of eq.v. we get µ 0|T ∗(F [A])|0 f1 = = dµ(Ω) dµ(Ω) −1 Ω−1 D(Aµ )δ[f1 ]∆f1 eiS F [AΩ ]δ[f2 ]∆f2 Ω D(Aµ )eiS δ[f2 ]∆f2 δ[f1 ]∆f1 F [AΩ ] (8. where Seff is an effective action 194 .109) in the form exp(iSeff ). that is: if F [A] is a gauge invariant functional F [AΩ ] = F [A] (8. (8.v.107) This is the relation we were looking for. since it is built up in terms of the correct functional measure.) of a gauge field functional will be defined as 0|T ∗ (F [A])|0 f = D(Aµ )δ[f ]∆f eiS F [A] (8.v.103) where the index f specifies the gauge in which we evaluate the v.e. we can prove an important result.108) For the following we will define a generating functional A Zf [ηµ ] = i D(Aµ )δ[f ]∆f eiS e A d4 xηµ Aµ A (8. By change of variable Aµ → AΩ .e.109) which is not gauge invariant.e.106) where in the second line we have made a further change of variables Ω → Ω−1 and used the invariance of dµ(Ω). About this point.104) than its v. In particular.

116) showing again that the v.v. more conveniently by choosing the gauge condition in the form fA (A) − BA (x) = 0.e.113) Multiplying this expression by the functional (8. (8.112) constant = ∆f [A] i − 2β dµ(Ω)D(B)e i − 2β dµ(Ω)e d4 x d4 x A Ω (fA (x))2 δ(fA − BA ) Ω (fA (x))2 A = ∆f [A] (8. We will discuss later the exponentiation of ∆f .taking into account the gauge fixing f = 0 and the Faddeev-Popov determinant.107) is 0|T ∗(F [A])|0 f1 = Ω dµ(Ω) 0|T ∗(∆f1 eiSf1 F [AΩ ])|0 f2 (8. or. allowing us to define a class of generating functionals i D(Aµ )∆f [A]e d4 x(LA − 1 2β (fA )2 ) i e A d4 xηµ Aµ A A Zβ [ηµ ] = A (8.114) Finally. a gauge invariant functional gives rise to a gauge invariant v. with BA (x) a set of arbitrary functions independent on the gauge fields.e. As far as δ[f ] it is concerned. we can exponentiate it by using its Fourier representation.v. defining Sf = − 1 2β d4 x A (fA (x))2 (8. for the corresponding operator.110) (8.115) we find the analogue of eq.111) We can further multiply this identity by the following expression which does not depend on the gauge fields i − 2β D(B)e d4 x A 2 BA (x) constant = We get (8.95)we can show as before that the gauge volume factor factorizes. Then ∆(f −B) [A] = det and therefore 1 = ∆f [A] Ω dµ(Ω)δ[fA − BA ] δ(fA [AΩ ] − BA ) δαB = det αA =0 δfA [AΩ ] δαB αA =0 = ∆f [A] (8. 195 .

8. η. ηµ ] = e ¯ with ¯ Z0 [η.125) .119) (8.124) ¯ [¯ψ + ψη + ηµ Aµ ]d4 x η (8.121) 1 e xδ (x − y) (8. ηµ ] = N Writing SA = − 1 4 Fµν F µν d4 x = 1 2 196 d4 x Aµ (g µν − ∂ µ ∂ ν )Aν (8. ηµ ] = N ¯ ¯ [¯ψ + ψη + ηµ Aµ ]d4 x η (8.122) Then the Faddeev-Popov determinant turns out to be field-independent and it can be absorbed into the normalization of the generating functional. η ] ¯ ¯ µ 0 (8.123) We extract the interaction term in the usual way −ie Z[η. η. η. η.120) 1 α(x) e 4 (8.4 Path integral quantization of QED Let us now discuss the path integral quantization of QED in a covariant gauge. which is given by i ¯ D(ψ.118) (8. We start from the lagrangian 1 ¯ L = ψ [iγ µ (∂µ + ieAµ ) − m] ψ − Fµν F µν 4 with Fµν = ∂µ Aν − ∂ν Aµ invariant under ψ(x) → ψ (x) = e−ieα(x) ψ(x) 1 Aµ (x) → Aµ (x) = Aµ (x) + ∂µ α(x) e The Lorentz gauge is defined by f (Aµ ) = ∂ µ Aµ (x) = 0 and f (AΩ ) = ∂ µ Aµ (x) + µ with δf (AΩ ) µ δα(y) = α=0 (8. ψ)D(Aµ )e d4 xL i δ[∂µ A ]e µ Z[η.117) (8. ψ)D(Aµ )e d4 xL0 i δ[∂µ A ]e µ d4 x − 1 δ i δη γµ 1 δ i δη µ 1 δ i δ η Z [η.126) i ¯ D(ψ.

133) . we can rewrite the integral over C(x) as −i C D(C)e 1 ∂µ η µ + i CC 2 (8.129) i x. ηµ ] = e−i η (x)SF (x − y)η(y) i Aµ (x) Aµ (x) i η (x)Aµ (x) µ 2 (8. ηµ ] = Z[0]e−i η (x)SF (x − y)η(y) e 2 197 gµν − ∂µ ∂ν 2 − ∂µ η i µ (+i) − ∂ν η i ν i µ ∂µ ∂ν ν η η 2 = e2 (8.using δ[∂µ Aµ ] and integrating over the Fermi fields we get ¯ ¯ Z0 [η.127) ·N D(Aµ )δ[∂µ A ]e e µ Then we exponentiate the δ[∂µ Aµ ] through its Fourier transform δ[∂µ Aµ ] = µ D(C)ei C(x)∂µ A (x) = µ D(C)e−i ∂µ C(x)A (x) (8. η. η. ηµ ] = e−i η (x)SF (x − y)η(y) 1 i(ηµ (x) − ∂µ C(x)) 2 ·N D(C)e (8.130) (8.132) ην (8.128) obtaining the following expression for Z0 ¯ Z0 [η.131) and doing again this gaussian integral we find 1 e2 The final result is i − ηµ ¯ ¯ Z0 [η. η .y i(η µ (y) − ∂ µ C(y)) i 1 − ηµ η µ ¯ = e−i η (x)SF (x − y)η(y) e 2 1 i 1 i ∂µ C η µ − ∂µ C ∂ µ C 2 · D(C)e After integrating by parts. η. ηµ ] = e−i η (x)SF (x − y)η(y) ¯ 1 i Aµ (x) Aµ (x) + (ηµ (x) − ∂µ C(x))Aµ (x) ·N D(Aµ )D(C)e 2 Doing the gaussian integral over Aµ we get ¯ ¯ Z0 [η.

114) i ¯ D(ψ. ηµ ] = N ¯ · e ¯ ¯ d4 x η ψ + ψη + ηµ Aµ (8.134) we write i µ ν −i η (x)SF (x − y)η(y)e− 2 η (x)Gµν (x − y)η (y) ¯ ¯ Z0 [η.135) Notice that this propagator is transverse. ηµ ] = Ne−i η (x)SF (x − y)η(y) ¯ i D(Aµ )e d4 x 1 µ A Kµν Aν + ηµ Aµ 2 (8. η. the addition of this term removes the problem of the singular wave operator.136) (8. that is it satisfies the Lorentz condition ∂ µ Gµν (x) = 0 (8.3. Of course.139) In this case the only effect of the gauge fixing is to change the lagrangian L to L = L − (∂µ Aµ )2 /(2β).141) 198 .140) Therefore we have to evaluate the functional integral ¯ Z0 [η. that is using the generating functional (8. ψ)D(Aµ )e d4 x L − 1 (∂µ Aµ )2 2β Zβ [η.Introducing the function Gµν = lim + →0 d4 k kµ kν gµν + 2 − 2 e−ikx 4 (2π) k +i (k + i )2 (8.138) Of course. except for th photon propagator which is −i k2 kµ kν gµν − 2 +i (k + i )2 (8. η. ηµ ] = Z[0]e The photon propagator in this gauge is given by 1 0|T (Aµ(x)Aν (x))|0 = iGµν (x − y) 0|0 (8. We can perform an analogous calculation but using the gauge fixing in the form f − B discussed at the end of the previous Section.137) The Feynman’s rules in this gauge are the same discussed in Section 1. the additional term in the propagator proportional to kµ kν does not affect the physics since the photon is coupled to a conserved current. η.5. In fact we can write d4 xL = 1 2 d4 xAµ gµν − (1 − 1 )∂µ ∂ν Aν β (8.

144) We can see that Kµν is non singular for β = ∞.5. (8.146) This can be solved by writing the most general second rank symmetric tensor function of kµ Gµν (k) = Agµν + Bkµ kν (8. eq. ηµ ] = Z[0]e where ρ Kµν Gνρ (x) = δµ δ 4 (x) (8. Solving these equations Gµν (k) = − gµν kµ kν + (1 − β) 4 2 k k (8.with Kµν = gµν obtaining − β−1 ∂µ ∂ν β (8.148) The singularities are defined by the usual i term.144) gives d4 k (2π)4 or −gµν k 2 + −gµν k 2 + β−1 ρ kµ kν Gνρ (k)e−ikx = δµ β d4 k −ikx e (2π)4 (8.151) The choice corresponding to the quantization made in Section 1.142) i µν ν −i η (x)SF (x − y)η(y) e− 2 ηµ (x)G (x − y)η (y) ¯ ¯ Zβ [η.147) with the coefficients such that −Ak 2 = 1. m2 = 0) (8. Going to momentum space. η. In fact for β → 0 i − lim e 2β d4 x(∂µ Aµ )2 → δ[∂µ Aµ ] β→0 (8.145) β−1 ρ kµ kν Gνρ (k) = δν β (8.3 is β = 1. and we get Gµν (x) = lim + →0 d4 k gµν kµ kν − 2 e−ikx + (1 − β) 2 4 (2π) k +i (k + i )2 (8.150) The discussion made at the beginning of this Section corresponds to β = 0 (the so called Landau gauge).143) (8.152) 199 . In this case Gµν (x) = −gµν ∆F (x.149) β−1 1 A − Bk 2 = 0 β β (8.

Separating the lagrangian in the quadratic part plus the rest. since the fermionic coupling is completely analogue to the abelian case.153) A where LA and LA are the quadratic and the interacting part respectively.156) .155) Consider now ∆f [Aµ ]. 8.This is called the Feynman’s gauge. we get LA = − 1 4 µν FµνC FC C 1 AB DE = − ∂µ AνC − ∂ν AµC − gfC AµA AνB ∂ µ Aν − ∂ ν Aµ − gfC Aµ Aν C C D E 4 1 (2) AB AB DE = LA + gfC AµA AνB ∂ µ Aν − g 2 fC fC Aµ Aν AµD AνE C A B 4 (2) ≡ LA + LI (8. The situation is quite different in the non-abelian case where the fermionic current is not conserved since the gauge fields are not neutral. Therefore we need further contributions in order to cancel the β dependent terms. being proportional to kµ kν does not contribute. In this Section we will consider the pure gauge case. In the Lorentz gauge fA [Aµ ] = ∂µ Aµ A 200 (8. it contains three-linear and fourlinear interaction terms and furthermore the Faddeev-Popov determinant typically depends on the gauge fields. As we see the parameter β selects different covariant gauges.5 Path integral quantization of the non-abelian gauge theories In the case of a non-abelian gauge theory there are several differences with respect to QED. and as a consequence the β dependent part of the propagator. Such contributions arise from the Faddeev-Popov determinant. which is conserved in the abelian case. The interaction terms can be extracted by the functional integration by the usual trick iSA Z[ηµ ] = e with SA = (I) (I) (2) (I) 1 δ A i δηµ Z [η A ] 0 µ (8.154) d4 xLA and i D(Aµ )e d4 x LA − (2) (I) Z0 [ηµ ] = N 1 2β (∂µ Aµ )2 A A i e µ d4 xηA AA µ ∆f [Aµ ] (8. As already stressed the physics does not depend on β since the ¯ photon couples to the fermionic current ψγ µ ψ. The pure gauge lagrangian is not quadratic.

c )e ∗ c∗ (x)MAB (x − y)cB (y)d4xd4 y A (8.158) we have AB fC [AΩ ] = ∂µ Aµ + gfC ∂µ (αA Aµ ) + αC µ C B Then δfC [AΩ (x)] µ δαB (y) αA =0 = δBC µ 2 4 BA 4 x δ (x − y) + gfC ∂µ (δ (x − y)AA ) = MCB (x − y) (8. η ] = e (8.76)) that a determinant can be written as a gaussian integral over Grassmann variables. recalling from eq.161) The ghosts are zero mass particles interacting with the gauge fields through the interaction term (I) AC (8. sending αA → gαA ) AB δAµ = gfC αA Aµ + ∂ µ αC C B (8. δfA [AΩ (x)] µ ∆f [Aµ ] = det δαB (y) =N αA =0 i D(c. η ∗ . It should be noticed that the ghost carry zero spin and therefore in their particle interpretation lead to a violation of the spin-statistics theorem. it is useful to give to it an exponential form. However this fact has no consequence.160) The Grassmann fields cA (x) are called the ghost fields.163) µ 0 µ with Z0 [η.157) (8. In order to get an expression for the generating functional which is convenient fir deriving the Feynman rules. ηµ ] = i ·e · i e D(Aµ )D(c. i δηµ e i δη i δη ∗ Z [η.162) SF P G = −g d4 xc∗ fB ∂µ Aµ cB A C In order to deal with this interaction it turns out convenient to define a new generating functional where we introduce also external sources for the ghost fields 1 δ 1 δ (I) 1 δ (I) iSA iSF P G − . since our generating functional do not generate amplitudes with external ghosts.164) .159) We see that ∆f [A] depends explicitly on the gauge fields. (7. This can be done. The action for the ghost fields can be read from the previous equation and we get SF P G = d4 xc∗ (x) δ AB A AC − gfB ∂µ Aµ cB (x) C (8. c∗) d4 x LA − (2) 1 2β A (∂µ Aµ )2 + ηµ Aµ A A A ∗ d 4 x c∗ cA + c∗ η A + η A cA A A 201 (8. η∗. η ] Z[η.Under an infinitesimal gauge transformation (for later convenience we have changed the definition of the gauge parameters. η ∗ .

. For the propagators we get . kµ kν k2 + i k2 1 +i (8.166) we get i A ν − ηµ (x)Gµν (x − y)ηB (y) −i η A ∗ (x)∆ (x − y)η B (y) AB AB e Z0 [η. A (7.Ghost field propagator: iδAB .1 .2 .167) (8.168) with Gµν = δAB Gµν (x − y) AB It is then easy to read the Feynman’s rules. η . k µ.Gauge field propagator: −iδAB gµν − (1 − β) .165) Defining ∆AB (x) = −δAB lim + →0 d4 k e−ikx (2π)4 k 2 + i (8.5. 202 1 k2 + i (8. For the ghost field we use eq.The integration over Aµ like in the abelian case.76) with the result D(c. ηµ ] = e 2 ∗ (8. 8. The Feynman’s rules for the vertices are obtained by taking the Fourier transform AB of the interaction parts (here we have written fC ≡ fABC ).5.170) B . A Fig.169) ν. 8. c )ei ∗ c∗ A cA + c∗ η A A +η A∗ cA − (iη A ) − =e ∗ i (iη A ) −i η A =e ∗ 1 ηA (8. B k A Fig.

A k1 k3 λ. D k4 k3 ν. 8.5. B k2 Fig.Three-linear gluon vertex: gfABC (2π)4 δ 4 (k1 + k2 + k3 ) [gµν (k1 − k2 )λ + gνλ (k2 − k3 )µ + gλµ (k3 − k1 )ν ] (8.5 . µ. 8. 8.Four-linear gluon vertex: 4 −ig (2π) δ ( 2 4 4 i=1 ki ) fABE fCDE (gµλ gνρ − gνλ gµρ ) (8.4 .173) . A k p B q C Fig. C Fig. A k1 ρ.172) + fACE fBDE (gµν gλρ − gλν gµρ ) + fADE fCBE (gµλ gρν − gρλ gµν ) .5.171) . µ.. B k2 λ.3 . C ν.Ghost-gluon vertex: −gfABC (2π)4 δ 4 (k + p − q)pµ 203 (8.5. µ.

since all the interaction terms have dimensions ≤ 4. In non-abelian gauge theories there are several ways to define the coupling. If the theory is gauge invariant (and the renormalization should preserve this feature).3)) gB = µ /2 g Z1 Z2 Z3 1/2 (8.When we have fermions coupled to the gauge fields we need to add the rule for the vertex (the propagator is the usual one). By a simple comparison with the coupling in QED we see that in this case the rule for the vertex is . in fact we can look at the fermionic coupling.Fermion-gluon vertex: A −igTbc (γµ )βγ (2π)4 δ 4 (p − p − k) (8. 204 . On the other hand we will evaluate the beta function. c Fig. giving rise to asymptotic freedom. 8. all these ways should give the same result. which are the analogue of the equality Z1 = Z2 in QED. Therefore we will study the definition coming from the fermionic coupling. As it is known. QCD the theory of strong interactions is in fact a gauge theory based on the gauge group SU(3). b p' γ. at the trilinear or at the four-linear couplings. µ. and Z@ and Z3 are the wave function renormalization constants for the fermion and the gauge particle respectively. A k p β.5. and it is precisely ths feature of being an asymptotically free theory that has made it very attractive phenomenologically. In this case the relation between the bare coupling and the renormalized one is the same as in QED (see Section 6. This is in fact ensured by a set of identities among the renormalization constants (Ward identities). since in these theories there is the possibility of having β < 0.175) where Z1 is the vertex renormalization constant. However.174) 8.6 The β-function in non-abelian gauge theories We will skip a careful treatment of the renormalization in non-abelian gauge theories.6 . the theory is renormalizable by power counting.

The one-loop contribution is given in Fig. At the same time we will exhibit also the general results for a theory SU(N) with fermions in the fundamental representation (that is the one of dimensions N). 8.One-loop fermion self-energy. (2.179) 2N This result is justified by the observation that the operator A T A T A (called the Casimir operator) commutes with all the generators and therefore it must be proportional to the identity matrix. where all the group factors are easily evaluated.1.6. (8. As a consequence we will need to evaluate all the three renormalization constants.180) Z2 = 1 − 2 C2 (F ) 8π with 205 . we get Σ(p) = A (T A T A )ΣQED (p) = A (T A T A ) 1 (ˆ − 4m) + Σf (p) p QED 2 8π (8. .176) For SU(2) we have (T A T A )ab = A A τA τA 2 2 ab 3 = δab 4 (8.177) In the case of a group SU(N) and fermions in the fundamental representation it is possible to show that (T A T A )ab = C2 (F )δab (8. We will work in the Feynman-’t Hooft gauge.As we shall see.7) g2 (8. recalling the expression for the electron self-energy in QED (see eq. Let us start with the fermion self-energy. 8. From the previous expression for Σ we find immediately (see Section 2.1 . corresponding to the choice β = 1 for the gauge field propagator (see eq. one of the differences with QED is that now Z1 = Z2 .178) A N2 − 1 C2 (F ) = (8.166)). In order to avoid group theory complications. The Feynman rules for this diagram are the same as in QED except for the group factors. we will consider the case of a gauge group SU(2) with fermions in the spinor representation.169)). k p p p _k Fig. Then.6.

6. To this end we need the one-loop expansion of the gauge particle propagator (the vacuum polarization diagram). Let us begin our calculation from the gauge particle contributions to the vacuum polarization. In fact all these diagrams involve the integral d2ω k k2 which. The kinematics is specified in Fig. This is given in Fig. 8.2 are zero (this is not true in other regularizations).181) By taking the limit a → 0.3.. We have ig 2 (a) Πµν.2 . 8. = + a) + + b) + c) + d) + + e) + f) Fig. 8.One-loop expansion of the vacuum polarization diagram in non-abelian gauge theories. e) and f)) of Fig.2. We will evaluate the truncated Green’s function which differs by a factor ig 2 from the vacuum polarization vacuum Πµν as defined in Section 2.182) (8. 8.AB 1 = (−g)2 f ADC f DBC 2 −i −i d2ω k E 2ω µν (p + k)2 k 2 (2π) (8. we see that for ω > 1 the integral vanishes.183) 206 .6.1. Since we are using dimensional regularization we can see immediately that the last three diagrams (diagrams d).6. The next task is the evaluation of Z3 . in dimensional regularization gives (including a mass term) d2ω k = −iπ ω Γ(1 − ω)(a2)ω−1 k 2 − a2 (8.6.

p+k C. It is given by σ ρ Eµν = gµ (p − k)ρ + g σρ (2k + p)µ − gµ (2p + k)σ × [gσν (p − k)ρ − gνρ (2p + k)σ + gρσ (2k + p)ν ] = gµν (p − k)2 + (2p + k)2 + 2ω(2k + p)µ (2k + p)ν − [(2p + k)µ (2k + p)ν − (2k + p)µ (p − k)ν + (p − k)µ (2p + k)ν ](8. ν Fig.The factor 1/2 is a symmetry factor associated to the diagram and the tensor Eµν comes from the trilinear vertices.190) kµ kν → 2ω 207 1 dz 0 1 (k + z(1 − z)p2 )2 2 (8. µ p D. The group factor is easily evaluated for SU(2).184) . taking into account that the integration over the odd terms in k gives zero.6. since in this case f ABC = ABC . and therefore C2 (G) is nothing but the Casimir of the adjoint. σ k p B. 8. By using the standard trick we reduce the two denominators in the integral to a single one 1 1 = 2 (k + p)2 k with k = k − pz (8.3 . Then we get (8.185) ADC DBC = −2δAB The general result for SU(N) is (taking into account that f ABC are completely antisymmetric) f ADC f DBC = −C2 (G)δAB (8. ρ A.One-loop gauge particle contribution to the vacuum polarization.187) This result can be understood by noticing that (T A )BC = if ABC gives a representation of the Lie algebra of the group (the adjoint representation). and that for symmetry we can make the substitution gµν 2 k (8.189) Operating this substitution into the tensor Eµν .186) with C2 (G) = N (8.188) .

as it should be.One-loop ghost contribution to the vacuum polarization. The relevant kinematics is given in Fig. We have ig 2 Πµν. p+k D A.194) By the standard procedure we get (k = k − pz)) ig 2 Πµν.6.AB = (−1)g 2 f ACD f BDC (b) i i d2ω k (p + k)µ kν 2ω (p + k)2 k 2 (2π) (8.4.6.AB = i C2 (G)δAB g p − p p (8.192) By taking only the 1/ term ( = 4 − 2ω) in the expansion and integrating over z we get 19 µν 2 11 µ ν g2 (a) ig 2 Πµν. . we get iπ ω 1 (a) ig 2 Πµν. we will see that adding to this diagram the contribution of the ghost loop we recover the transverse factor.193) 2 16π 6 3 As we see this contribution is not transverse. ν Fig.we get (by calling k 2 with k 2 ) Eµν = gµν k 2 6 1 − 1 + gµν p2 (1 + z)2 + (2 − z)2 2ω + pµ pν 2ω(1 − 2z)2 − 2 (1 + z)2 + (2 − z)(1 − 2z) ≡ aµν k 2 + bµν (8.AB = −g 2 C2 (G)δAB (b) 1 dz 0 d2ω k kµ kν − pµ pν z(1 − z) (2π)2ω (k 2 + z(1 − z)p2 )2 (8.4 .6.195) 208 . 8.2. in fact. 8. µ p C k p B. The ghost contribution is the part b) of the expansion of Fig.5 and putting everything together.191) Performing the integral over k with the formulas of Section 2.AB = − g 2 (−C2 (G))δAB 2 (2π)2ω 1 dz ωz(1 − z)p2 Γ(1 − ω)aµν + Γ(2 − ω)bµν 0 (8. that is it is not proportional to (pµ pν − gµν p2 ). 8.

8.5.AB + Πµν. we get ig 2 Πµν.196) We perform now the integral in z. we perform the integral over k 1 2 p Γ(1 − ω) − pµ pν Γ(2 − ω) 2 0 (8. 2 (a) (Πµν.AB = ig 2 T r(T A T B )ΠQED = i µν (c) ig 2 Πµν.200) 2 where nF is the number of fundamental representations.AB = −ig 2 (b) πω C2 (G)δAB (2π)2ω 1 dz z(1 − z) ig 2 Πµν.AB ) = −i g2 8π 2 5 2 C2 (G) − nF 3 3 209 pµ pν − gµν p2 (8.198) + = −i 2 C2 (G)δAB pµ pν − gµν p2 ig 8π 3 The fermionic loop contribution is given in Fig.199) 12π 2 where we have used the conventional normalization for the group generators in the fermionic representation 1 T r(T A T B ) = nF δAB (8.AB + Πµν. Then. we have three different kind of quarks corresponding to different flavors.6. For instance.AB ) p+k A.AB = i (b) g2 1 1 C2 (G)δAB p2 gµν + pµ pν 2 16π 6 3 (8. 8. in the case of QCD. The leading contribution is g2 nF δAB pµ pν − gµν p2 (8.6. µ p p B.One-loop fermion contribution to the vacuum polarization.6. .AB (b) Πµν.197) Putting together this contribution with the former one we get g2 5 (8. the one-loop leading term contribution to the vacuum polarization is given by ig 2 Πµν. Then taking the limit ω → 2. and keeping only the leading term.AB = ig 2 (Πµν. and except for the group factor is the same evaluated in Section 2.5 . Then.where we have eliminated the terms linear in k .201) (a) (b) (c) . ν k Fig.

Again. .A = (−igγν T B ) g · · ˆ −i d2ω k ik 2ω k 2 (p − k)2 (2π) ABC µν (−g)f [g (q + p − k)ρ − g νρ(p + p − 2k)µ + g ρµ (p − k − q)ν ] −i (−igγρ T C ) (8.A (8.By comparison with Section 2.6.204) 2 2 Therefore we obtain −ig 3 Λµ.7 we get g2 Z3 = 1 + 2 8π 5 2 C2 (G) − nF 3 3 (8.6.Fermion one-loop vertex correction. The fermion contribution is given by the diagram of Fig.7. We get (q = p − p) −ig 3 Λµ. 8.202) We are now left with the vertex corrections. A p' + k B p' k p+k B p Fig.206) (p − k)2 210 . 8.6. the leading contribution is g3 3 µ.6 . 8. this is the same as in QED except for the group factors.205) The final diagram to be evaluated is the vertex correction from the trilinear gauge coupling given in Fig.203) −ig Λf = −i 2 (T B T A T B )γ µ 8π The group factor can be elaborated as follows T B T A T B = T B (T B T A + if ABC T C ) = C2 (F )T A + if ABC T B T C 1 1 = C2 (F )T A − f ABC f BCD T D = C2 (F ) − C2 (G) T A (8.6.A = −i f g3 8π 2 1 C2 (F ) − C2 (G) T A γ µ 2 (8.

210) .A ) = −i g f Again. Therefore the divergent piece is obtained by neglecting all the external momenta. Since we are interested only in the divergent piece. Therefore the expression simplifies to −ig 3 Λµ. A. µ p' . whereas the numerator goes at most as k 2 .A = ig 3 f ABC T B T C g The group factor gives i i f ABC T B T C = f ABC f BCD T D = C2 (G)T A 2 2 whereas the numerator gives simply −3k 2 γ µ .208) ˆ d2ω k γν k(g µν k ρ − 2g νρk µ + g µρ k ν )γρ (2π)2ω (k 2 )3 (8.k B k p' p-k C p Fig.209) (8.7 gives Z1 = 1 − Using eq. 8. the calculation can be simplified a lot by the observation that the denominator goes as (k 2 )3 .One-loop vertex correction from the trilinear gauge coupling.175) we have 1 a1 gB = µ /2 g(1 + ∆Z1 − ∆Z2 − ∆Z3 ) = µ /2 g + 2 and collecting everything together we get a1 = − g3 16π 2 11 2 C2 (G) − nF 3 3 211 (8..A = − g g3 3 C2 (G)T A γ µ (2π)2ω 2 g3 3 d2ω k = −i 2 C2 (G)T A γ µ (k 2 )2 8π 2 (8. Then we have −ig 3 Λµ.A = −ig 3 (Λµ.7 . comparison with Section 2. (8.211) g3 (C2 (F ) + C2 (G)) T A γ µ 8π 2 (8.A + Λµ.207) Adding up the two vertex contributions we obtain −ig 3 Λµ.212) g2 (C2 (F ) + C2 (G)) 8π 2 (8.6.213) (8.

217) We see that QCD is asymptotically free (β(g) < 0) for nF < 33 2 (8.The evaluation of β(g) goes as in the QED case (see Section 6.218) 212 .215) In the case of QCD where the gauge group is SU(3) we have C2 (G) = 3 and therefore βQCD (g) = − g3 16π 2 2 11 − nF 3 (8.216) (8.214) g3 16π 2 (8.3). that is β(g) = µ that is β(g) = − 1 ∂g(µ) =− ∂µ 2 1−g d dg 1 a1 = − (a1 − 3a1 ) = a1 2 11 2 C2 (G) − nF 3 3 (8.

Chapter 9 Spontaneous symmetry breaking 9. with a symmetry O(N). in components. We will start examining the linear σ-model.1 The linear σ-model In this Section and in the following we will study. some field theory with particular symmetry properties.4) (9. · · · .6) (9.7) =0 ji =− 213 . φN ) and leaving invariant its norm N |φ| = 2 i=1 φi φi (9. The lagrangian is given by 1 L= 2 N i=1 1 ∂µ φi ∂ µ φi − µ2 2 N i=1 λ φi φi − 4 N 2 φi φi i=1 (9. φi This is satisfied by ij ij φj (9. This is a model for N scalar fields.3) The condition for letting the norm invariant gives |φ + δ φ|2 = |φ|2 from which φ · δφ = 0 or.1) This lagrangian is invariant under linear transformations acting upon the vector φ = (φ1 .2) Consider an infinitesimal transformation (from now on we will omit the index of sum over the indices which are repeated) δφi = ij φj (9.5) (9. from a classical point of view.

10) (9.15) This is nothing but the Lie algebra of the group O(N).µ µφ ) ∂(∂ i 2 AB AB Tij φj (9. N (9. It is useful for further generalizations to write the infinitesimal transformation in the form δφi = ij φj =− i 2 AB AB Tij φj .13) which is similar to what we did in Section 3. i. O(N). and the only conserved current is given by 12 21 Jµ = −Jµ = φ1. N. · · · . by exponentiating the infinitesimal transformation one gets S=e with T (9.9) (9. For a finite transformation we have |φ |2 = |φ|2 (9.16) since the N(N − 1)/2 parameters AB are linearly independent. we find the N(N − 1)/2 conserved currents AB AB Jµ = −iφi. The matrices S form the rotation group in N dimensions. T CD ] = −iδ AC T BD + iδ AD T BC − iδ BD T AC + iδ BC T AD (9. A.17) In the case of N = 2 the symmetry is the same as for the charged field in a real basis. The Noether’s theorem implies a conserved current for any symmetry of the theory.showing that the rotations in N dimensions depend on N(N − 1)/2 parameters.3 when we discussed the Lorentz transformations. B = 1.µ φ1 (9.11) (9.µ Tij φj (9. and the T AB are the infinitesimal generators of the group.µ φ2 − φ2.8) with φi = Sij φj implying SS T = 1 In fact. By comparison we see that the matrices T AB are given by AB A B A B Tij = i(δi δj − δj δi ) (9. In this case we will get N(N − 1)/2 conserved quantities.14) It is not difficult to show that these matrices satisfy the algebra [T AB .18) 214 . j = 1. · · · . Applying now the Noether’s theorem we find the conserved current jµ = ∂L i δφi = − φi.12) =− implying that S is an orthogonal transformation.

if we have conserved currents given by A A (9. using the canonical commutation relations.28) where L and R are two special (that is with determinant equal to 1) unitary matrices.One can easily check that the charges associated to the conserved currents close the same Lie algebra as the generators T AB . τ1 and τ3 real.25) (9. R ∈ SU(2). More generally. that is L. π2 . π3 . σ) These fields can be arranged into a 2 × 2 matrix M = σ + iτ · π (9. The reason to restrict these matrices to be special is that only in this way the transformed matrix satisfy the condition (9. we get [QA .20) (9. σ) = (π.21) ˙ A d3 x φi Tij φj (9.23) where τ are the Pauli matrices.26) (9.µ Tij φj with [T A . QB ] = if ABC QC with QA = A d3 x j0 (x) = −i (9.27) This lagrangian is invariant under the following transformation of the matrix M M → LMR† (9. and that τ2 anticommutes with τ1 and τ3 .29) . if A is a 2 × 2 matrix with detA = 1. Noticing that τ2 is pure imaginary.22) A particular example is the case N = 4. we get M = τ2 M ∗ τ2 Furthermore we have the relation 1 |φ|2 = σ 2 + |π|2 = T r(M † M) 2 Using this it is easy to write the lagrangian for the σ-model in the form 1 1 1 L = T r(∂µ M † ∂ µ M) − µ2 T r(M † M) − λ T r(M † M) 4 4 16 2 (9. We parameterize our fields in the form φ = (π1 .24) (9. T B ] = if ABC T C then. then τ2 AT τ2 = A−1 215 (9.25). In fact.19) jµ = −iφi.

since −L and −R define the same S as L and R). 2 then we get i i 1 i δL M = (− θL · τ)M = (− θL · τ)(σ + iπ · τ ) = θL · π + (θL ∧ π − θL σ) · τ (9.32) (9.29) τ2 L∗ τ2 = τ2 L† τ2 = L† τ2 RT τ2 = R−1 = R† T −1 ∗ (9.26) we see that M → LMR† leaves the norm of the vector φ = (π. for M = LMR† we get τ2 M τ2 = τ2 L∗ M ∗ RT τ2 = τ2 L∗ τ2 (τ2 M ∗ τ2 )τ2 RT τ2 and from (9. the invariance group in this basis is SU(2)L ⊗ SU(2)R . that is on σ and π. 2 Analogously we obtain 1 δR σ = − θR · π. We can evaluate the effect of an infinitesimal transformation. but from the relation (9. Therefore this transformation must belongs to O(4).31) since the L and R are independent transformations.33) 2 2 2 2 where we have used τi τj = δij + i Since δL M = δL σ + iδL π · τ we get 1 δL σ = θL · π.34) (9.37) The combined transformation is given by 1 δσ = (θL − θR ) · π. 2 1 δL π = (θL ∧ π − θL σ) 2 1 δR π = (θR ∧ π + θR σ) 2 (9. σ) invariant and therefore the same must be true for the linear transformation acting upon the matrix elements of M.38) . In fact this group and O(4) are related by the following observation: the transformation M → LMR† is a linear transformation on the matrix elements of M. To this end we will consider separately left and right transformations. We parameterize the transformations as follows i L ≈ 1 − θL · τ .30) =L (9. This shows that the two groups SU(2) ⊗ SU(2) and O(4) are homomorphic (actually there is a 2 to 1 relationship.36) (9.35) ijk τk i R ≈ 1 − θR · τ 2 (9.Therefore. 2 1 δπ = [(θL + θR ) ∧ π − (θL − θR )σ] 2 216 (9.

[QA .µ θL · π + π. [QR .41) We see that the transformations corresponding to the diagonal SU(2) are the rotations in the 3-dimensional space spanned by π. [QV . From the Noether’s theorem we get the conserved currents 1 1 L jµ = σ. QR ] = 0 i j (9. πj ] = i i ijk πk .40) These transformations span a subgroup SU(2) of SU(2)L ⊗ SU(2)R called the diagonal subgroup.49) These equations show that QV are the infinitesimal generators of a subgroup SU(2) i of SU(2)L ⊗ SU(2)R which is the diagonal subgroup.39) as it must be for a transformation leaving the form σ 2 + |π|2 invariant.50) 217 .µ σ − π.48) and A Jµ = σ. These rotations define a subgroup O(3) of the original symmetry group O(4).44) Using the canonical commutation relations one can verify that the corresponding charges satisfy the Lie algebra of SU(2)L ⊗ SU(2)R [QL . δπ = θ ∧ π (9. σ] = 0 i (9. as it follows from [QV .µ π + π.µ ∧ π (9.µ ∧ π L Jµ = σ.43) (9.42) (9. Of particular interest are the transformations with θL = θR ≡ θ.µ 1 L A R J µ = (J µ − J µ ) 2 (9.µ π − π.and we can check immediately that σδσ + π · δπ = 0 (9. [QL .µ π − π. QA ] = i i j V ijk Qk (9. [QV . QA ] = i i j A ijk Qk . In this case we have δσ = 0.45) By taking the following combinations of the currents 1 L V R Jµ = (Jµ + Jµ ). In this case we have L = R and M → LML† (9.47) (9.µ σ The corresponding algebra of charges is [QV . 2 one has V Jµ = π ∧ π.46) (9. QL ] = i i j L ijk Qk .µ σ − π. QV ] = i i j V ijk Qk . QR ] = i i j R ijk Qk .µ · (θL ∧ π − θL σ) 2 2 and dividing by θL /2 and analogously R Jµ = −σ.

this is well defined only when we are considering the theory close to a minimum the energy of the system. Then the condition for getting a minimum is that the eigenvalues of the matrix of the second derivatives of the potential at the stationary point are definite positive.57) ∂φi In order to have a solution to the stationary condition we must have φi = 0. or |φ|2 = − µ2 λ (9. In the case of the linear σ-model we have 1 λ V = µ2 |φ|2 + (|φ|2 )2 (9.55) This equation shows that the particle masses are given by the eigenvalues of the second derivative of the potential at the minimum.58) 218 .In the following we will be interested in treating the interacting field theories by using the perturbation theory.54) φ=v φi φj + · · · (9. In the case of the linear σ-model the energy is given by N H= d xH = 3 dx i=1 3 ∂L ˙ φ −L = ˙ i ∂ φi 1 dx 2 3 N ˙ (φ2 + |∇φi|2 ) + V (|φ|2 ) i i=1 (9.56) 2 4 Therefore ∂V = µ2 φi + λφi |φ|2 (9. As in the quantum mechanical case.53) The lagrangian density becomes 1 L = ∂µ φi ∂ µ φi − V (|φ + v|2 ) 2 Expanding V in series of φi we get 1 ∂2V V = V (|v| ) + 2 ∂φi ∂φj 2 (9. it follows that the absolute minimum is obtained for constant field configurations.52) ∂φi Let us call by vi the generic solution to this equation (in general it could happen that the absolute minimum is degenerate). In fact if we are going to expand around a maximum the oscillation of the system can become very large leading us outside of the domain of perturbation theory. In this case we define new fields by shifting the original fields by φi → φi = φi − vi (9. such that ∂V (|φ|2 ) =0 (9.51) Since in the last member of this relation the first two terms are positive definite.

barring a phase factor.This equation has real solutions only if µ2 /λ < 0. µ2 cannot be identified with a physical mass. But. is just but an example of a general phenomenon which goes under the name of spontaneous symmetry breaking of symmetry.60) . We will study this case in the following Sections.2 Spontaneous symmetry breaking In this Section we will see that the linear σ-model with µ2 < 0.62) (9. therefore we may have non zero solutions to the minimum condition only if µ2 < 0. The free theory is given by the quadratic terms in the lagrangian density. these are given by the eigenvalues of the matrix of the second derivatives of the potential at the minimum and they are positive definite by definition. are small).61) (9. notice that in this case. Just as an example consider a scalar field described by a lagrangian invariant under parity P : φ → −φ (9. the fluctuations. in order to have a potential bounded from below one has to require λ > 0. both the free and the interacting theories are O(N) symmetric. It follows 0|φ|0 = 0|P −1P φP −1P |0 = 0|P φP −1|0 = − 0|φ|0 219 (9. The idea is very simple and consists in the observation that a theory with hamiltonian invariant under a symmetry group may not show explicitly the symmetry at the level of the solutions.59) The lagrangian density will be of the type 1 L = ∂µ φ∂ µ φ − V (φ2 ) 2 If the vacuum state is non degenerate. 9. • The fundamental state of the theory is degenerate and transforms in a non trivial way under the symmetry group. However. As we shall see this may happen when the following conditions are realized: • The theory is invariant under a symmetry group G. Furthermore. This phenomenon lies at the basis of the modern description of phase transitions and it has acquired a capital relevance in the last years in all field of physics. we must have P |0 = |0 since P commutes with the hamiltonian. and they describe N particles of common mass m. In the case of µ2 > 0 the minimum is given by φi = 0 and one can study the theory by taking the term λ(|φ|2 )2 as a small perturbation (that is requiring that both λ and the values of φi .

In fact.65). φ2 ). In the case µ2 < 0 (degenerate case). if µ2 λ (9. By denoting with R(θ) the operator rotating the fields in the plane (φ1 . in the non-degenerate case we have R(θ)|0 = |0 and 0|φ|0 = 0|R−1RφR−1 R|0 = 0|φθ |0 = 0 R(θ)|0 = |θ (9.67) which now does not imply that the expectation value of the field vanishes.63) Things change if the fundamental state is degenerate.70) (9.69) For µ2 > 0 there is a unique fundamental state (minimum of the potential) φ = 0. v= − µ2 λ (9.72) since φθ = φ. and a lagrangian density with symmetry O(2) 1 λ 1 L = ∂µ φ · ∂ µ φ − µ2 φ · φ − (φ · φ)2 2 2 4 where φ · φ = φ2 + φ2 1 2 (9.68) with v defined as in (9.66) Therefore R|φ|R = R|P −1 P φP −1P |R = − L|φ|L (9.60).71) (9.65) By denoting with |R e |L the two states corresponding to the classical configurations φ = ±v. whereas for µ2 < 0 there are infinite degenerate states given by |φ|2 = φ2 + φ2 = v 2 1 2 (9. we have P |R = |L = |R (9. this potential has two minima located at φ = ±v. In the following we will be rather interested in the case of continuous symmetries.64) V (φ2 ) = φ2 + φ4 2 4 with µ2 < 0.73) where |θ is one of the infinitely many degenerate fundamental states lying on the circle |φ|2 = v 2 . So let us consider two scalar fields. This would be the case in the example (9. Then 0|φi|0 = 0|R−1 (θ)R(θ)φi R−1 (θ)R(θ)|0 = θ|φθ |θ i 220 (9.from which 0|φ|0 = 0 (9.74) . we have (9.

This follows from the existence of the tunnel effect. The equilibrium positions are around the minima positions. For instance. as µ2 for the scalar field. but below the Curie temperature exhibits spontaneous magnetization. and consequently to break the symmetry. One can easily construct classical systems exhibiting spontaneous symmetry breaking. The situation can be described qualitatively saying that the existence of a degenerate fundamental state forces the system to choose one of these equivalent states. since spontaneous symmetry breaking cannot happen in finite systems. and the temperature for the ferromagnet. By denoting with H the matrix of the hamiltonian between these two states. the order parameters were zero in the symmetric phase and different from zero in the broken phase. and the phase transition is characterized by the order parameters which assume different values in different phases. the expectation value of the field (contrarily to the non-degenerate state) does not need to vanish. and recall that we have defined the fundamental states through the correspondence with the classical minima x = x0 → |R x = −x0 → |L (9.77) 221 . due to the transition the hamiltonian acquires a non zero matrix element between the states |R and |L . where x is the particle position. we get H= 0 1 1 0 (9. These situations are typical for the so called second order phase transitions. the lagrangian and the equations of motion preserve the symmetry.75) Again. ±x0 . Let us consider again a particle in a double-well potential. But the breaking is only at the level of the solutions. breaking in this way the symmetry. In fact. This system has parity invariance x → −x. a classical particle in a double-well potential. One can describe them through the Landau free-energy. as the expectation value of the scalar field or as the magnetization. If we put the particle close to x0 . The situation is slightly more involved at the quantum level. A further example is given by a ferromagnet which has an hamiltonian invariant under rotations. which depends on two different kind of parameters: • Control parameters. In the previous examples. • Order parameters.with φθ = R(θ)φi R−1 (θ) = φi i (9. The system goes from one phase to another varying the control parameters. it will perform oscillations around that point and the original symmetry is lost.76) But the tunnel effect gives rise to a transition between these two states and as a consequence it removes the degeneracy.

and the minimum at φ = v. at t = 0. coming back to the scalar field example. This is not an energy eigenstate and its time evolution is given by 1 |R. One can show that 1 < 0 and therefore the fundamental state is the symmetric one. by putting the particle in the right minimum. This situation gives rise to the well known effect of quantum oscillations. We can express the states |R and |L in terms of the energy eigenstates |R = |L = 1 √ (|S 2 1 √ (|S 2 + |A ) − |A ) (9.83) ∆E In nature there are finite systems as sugar molecules.80) with eigenvalue EA = 0 − 1 .85) . In fact. for infinite systems. for t = π/∆E the state |R transforms into the state |L . and we may have spontaneous symmetry breaking.81) Let us now prepare a state. which seem to exhibit spontaneous symmetry breaking. becomes infinite in the limit of infinite volume H(φ = 0) − H(φ = v) = − V d3 x µ2 2 λ 4 µ4 v + v = 2 4 4λ 222 d3 x = V µ4 V 4λ (9. Therefore. The splitting of the fundamental states decreases with the height of the potential between two minima. therefore. its expectation value on the vacuum must be a constant. and 1 |A = √ (|R − |L ) 2 (9. The explanation is simply that the energy difference ∆E is so small that the oscillation period is of the order of 104 − 106 years. 0 − 1) (9. the previous mechanism does not work. The state oscillates with a period given by 2π (9.The eigenvalues of H are ( 0 + 1.78) We have no more degeneracy and the eigenstates are 1 |S = √ (|R + |L ) 2 with eigenvalue ES = 0 (9. In fact one observes right-handed and left-handed sugar molecules. |S .79) + 1. as it follows from the translational invariance of the vacuum 0|φ(x)|0 = 0|eiP xφ(0)e−iP x |0 = 0|φ(0)|0 = v (9.82) with ∆E = EA − ES . t = √ e−iES t |S + e−iEA t |A 2 1 = √ e−iES t |S + e−it∆E |A 2 (9.84) T = and the energy difference between the maximum at φ = 0.

In fact. among which φ = 0 is a maximum.91) 223 . we have only one real solution given by φ = 0. The points of the sphere |φ|2 = v 2 are degenerate minima.9. This theorem says that for any continuous symmetry spontaneously broken. • The theory must be manifestly covariant. (9. • The Hilbert space of the theory must have a definite positive norm. since ∂2V = δlm µ2 > 0 (9. by choosing φl = vδlN as a representative point. v= We have two possibilities • µ2 > 0. The theorem holds rigorously in a local field theory.86) 2 2 4 The conditions that V must satisfy in order to have a minimum are ∂V = µ2 φl + λφl |φ|2 = 0 ∂φl with solutions |φ|2 = v 2 . there exists a massless particle (the Goldstone boson).3 The Goldstone theorem From our point of view. the most interesting consequence of spontaneous symmetry breaking is the Goldstone theorem.88) λ The character of the stationary points can be studied by evaluating the second derivatives ∂2V = δlm (µ2 + λ|φ|2 ) + 2λφl φm (9. there are infinite solutions. under the following hypotheses • The spontaneous broken symmetry must be a continuous one. We will limit ourselves to analyze the theorem in the case of a classical scalar field theory. we get ∂2V = 2λv 2 δlN δmN > 0 ∂φl ∂φm (9. which is a minimum.89) ∂φl ∂φm φl = 0.87) −µ2 (9. Let us start considering the lagrangian for the linear σ-model with invariance O(N) 1 µ2 λ L = ∂µ φ · ∂ µ φ − φ · φ − (φ · φ)2 (9.90) ∂φl ∂φm • µ2 < 0.

94) m2 a = 0. whereas the N − 1 rotations on the planes a − N move away from the surface of the minima. · · · . In fact. are given by m2 = −2µ2 (9. whereas the one we got depends only on the two parameters m and λ.96) In this form the original symmetry O(N) is broken. It must be stressed that this is not the most general potential invariant under O(N − 1).93) ⎣· · · · ⎦ ∂φl ∂φm minimo 0 0 · −2µ2 Therefore the masses of the fields φa . φ χ By defining m2 = −2µ2 we can write the potential as a function of the new fields m4 1 V =− + m2 χ2 + 16λ 2 m2 λ χ 2 N −1 2 (9. and it is a a=1 invariant under rotations around the axis we have chosen as representative for the fundamental state. This can be seen also in terms of generators. · · · . N − 1. Since the field we have chosen as representative of the ground state is φi |min = vδiN . V depends only on the combination N −1 φ2 .98) Therefore we have N − 1 broken symmetries and N − 1 massless scalars. N − 1 . b = 1. Clearly the rotations along the first N − 1 directions leave the potential invariant. the right fields to be used are φl − vδlN . and χ = φN − v.95) φ2 a a=1 +χ 2 λ + 4 N −1 φ2 a a=1 +χ 2 (9. and their mass is just given by the coefficient of the quadratic term ⎡ ⎤ 0 0 · 0 ⎢0 0 · ∂2V 0 ⎥ 2 ⎥ = −2µ2 δlN δmN = ⎢ Mlm = (9. Therefore spontaneous symmetry breaking puts heavy constraints on the dynamics of the system. However a residual symmetry O(N − 1) is left. (0. we have ab a b b a (9.97) Tij φj |min = i(δi δj − δi δj )vδjN = 0 since a. · · · . In fact the most general potential (up to the fourth order in the fields) describing N scalar fields with a symmetry O(N − 1) would depend on 7 coupling constants. We have also seen that we have N − 1 massless scalars.Expanding the potential around this minimum we get V (φ) ≈ V + minimum 1 ∂2V 2 ∂φl ∂φm minimum (φl − vδlN )(φm − vδmN ) (9. a = 1. v). and aN a N N a a Tij φj |min = i(δi δj − δi δj )vδjN = ivδi = 0 (9.92) If we are going to make a perturbative expansion. The generators of O(N) divide up naturally in the generators of the vacuum symmetry 224 .

the Goldstone bosons correspond to the generators of G which are left after subtracting the generators of H. 0). in order to have a Hilbert space spanned by only the physical states. If one goes to a covariant gauge. and where the physical states are extracted through a supplementary condition. But once one couples spontaneous symmetry breaking to a gauge symmetry. the Goldstone bosons. 9. at the same time. Since these states are degenerate the operation does not cost any energy.99) and let us analyze the spontaneous symmetry breaking mechanism. as the Coulomb gauge. if one chooses a physical gauge. v= −µ2 λ (9. the theory is covariant but one has to work with a big Hilbert space. and. This is well known already for the electromagnetic field. If µ2 < 0 the symmetry is broken and we can choose the vacuum as the state φ = (v.100) 225 . gauge invariance forbids the presence in the lagrangian of terms quadratic in the fields. than the theory looses the manifest covariance. which however are fermions.4 The Higgs mechanism We have seen that the spontaneous symmetry breaking mechanism. This is the famous Higgs mechanism. The reason is that it is impossible to quantize a gauge theory in a way which is at the same time manifestly covariant and has a Hilbert space with positive definite metric.(here O(N − 1)). One can say that Goldstone bosons correspond to flat directions in the potential. Also gauge theories lead to massless vector bosons. Intuitively one can understand the origin of the massless particles noticing that the broken generators allow transitions from a possible vacuum to another. and in the so called broken generators. as in the electromagnetic case. as the Lorentz one. each of them corresponding to a massless Goldstone boson. The way in which the Goldstone theorem is evaded is that the Goldstone bosons disappear. if we look back at the hypotheses underlying a gauge theory. What happens is that. Unfortunately in nature the only massless particles we know are the photon and perhaps the neutrinos. with non-definite positive metric. in the case of continuous symmetry leads to massless scalar particles. In fact. where one has to choose the gauge before quantization. if the original symmetry group G of the theory is spontaneously broken down to a subgroup H (which is the symmetry of the vacuum). the gauge bosons corresponding to the broken symmetries acquire mass. In general. Let us start with a scalar theory invariant under O(2) 1 λ µ2 L = ∂µ φ · ∂ µ φ − φ · φ − (φ · φ)2 2 2 4 (9. in fact. From the relativistic dispersion relation this implies that we must have massless particles. things change. it turns out that Goldstone theorem does not hold in this context.

In fact. This is better seen by using polar coordinates for the fields.101) In this case the group O(2) is completely broken (except for the discrete symmetry φ2 → −φ2 ). from which δχ = −αφ2 . 2 θ≈ φ2 φ2 ≈ v+χ v (9. the new fields transform as ρ → ρ. αv. The Goldstone field is φ2 . 1 2 sin θ = φ2 φ2 1 + φ2 2 (9. In fact one can have invariance under translations of the field. δφ2 = αχ + αv (9.106) It should be also noticed that the two coordinate systems coincide when we are close to the vacuum. This is what happens when one moves in a way which is tangent to the surface of the degenerate vacuums (in this case a circle). This has a peculiar way of transforming under O(2).After the translation φ1 = χ+v. the original fields transform as δφ1 = −αφ2 . we get the potential (m2 = −2µ2 ) V =− 1 m4 + m2 χ2 + 16λ 2 m2 λ λ χ(φ2 + χ2 ) + (φ2 + χ2 )2 2 2 4 2 (9.105) Under a finite rotation.104) δφ2 = αφ1 (9. in that case we can perform the following expansion ρ= φ2 + χ2 + 2χv + v 2 ≈ v + χ. if we make the theory invariant under a local transformation.107) Again. In fact. This is the main reason for the Goldstone particle to be massless.108) 226 . we will have invariance under θ(x) → θ(x) + α(x) (9. In other words it must be possible to eliminate the Goldstone field from the theory. that is ρ= φ2 + φ2 . θ →θ+α (9. as when we are doing perturbation theory.103) We see that the Goldstone field undergoes a rotation plus a translation. only if the potential is flat in the corresponding direction. with 0|χ|0 = 0. How do things change if our theory is gauge invariant? In that case we should have invariance under a transformation of the Goldstone field given by δφ2 (x) = α(x)χ(x) + α(x)v (9. it follows that we can choose it in such a way to make φ2 (x) vanish.102) Since α(x) is an arbitrary function of the space-time point.

116) (9.112) By performing the following gauge transformation on the scalars φ → φ = φe−iθ and the corresponding transformation on the gauge fields 1 Aµ → Aµ = Aµ − ∂µ θ g (9.113) (9.110) We know that it is possible to promote a global symmetry to a local one by introducing the covariant derivative ∂µ φ → (∂µ − igAµ )φ from which 1 L = (∂µ + igAµ )φ† (∂ µ − igAµ )φ − µ2 φ† φ − λ(φ† φ)2 − Fµν F µν 4 In terms of the polar coordinates (ρ.99) can be written as L = ∂µ φ† ∂ µ φ − µ2 φ† φ − λ(φ† φ)2 (9. 2 1 φ† = √ ρe−iθ 2 (9.117) In this way the Goldstone boson disappears.114) (9.111) (9. We have now to translate the field ρ ρ = χ + v. 2 2 The O(2) transformations become phase transformations on φ φ → eiα φ and the lagrangian (9. 0|χ|0 = 0 (9.115) the lagrangian will depend only on the fields ρ and Aµ (we will put again Aµ = Aµ ) µ2 λ 1 1 L = (∂µ + igAµ )ρ(∂ µ − igAµ )ρ − ρ2 − ρ4 − Fµν F µν 2 2 4 4 (9.109) φ = √ (φ1 + iφ2 ).By choosing α(x) = −θ(x) we can eliminate this last field from the theory. It is convenient to introduce complex variables 1 1 φ† = √ (φ1 − iφ2 ) (9.118) 227 . θ) we have 1 φ = √ ρeiθ . The only remaining degree of freedom in the scalar sector is ρ(x). Let us study the gauging of this model.

Before we had 4 degrees of freedom. min aN Tij φj min a = ivδi . coming from the covariant derivative.120) It is instructive to count the degrees of freedom before and after the gauge transformation. b = 1. The previous gauge transformation realizes the purpose of making that term vanish. · · · .126) .124) = 0. a.119) 2 Therefore the gauge field acquires a mass m2 = g 2v 2 A (9. because now it is a massive vector field.121) between the Goldstone field and the gauge vector which makes complicate to read the mass of the states.122) g AB (9.100) φi = vδiN Recalling that ab Tij φj (9. The gauge in which such a thing happens is called the unitary gauge.and we see that this generates a bilinear term in Aµ . We will consider now the further example of a symmetry O(N). The lagrangian invariant under local transformations is 1 µ2 λ L = (Dµ )ij φj (D µ )ik φk − φi φi − (φiφi )2 2 2 4 where (9. but three degrees of freedom from the gauge vector. we choose again the vacuum along the direction N. two from the scalar fields and two from the gauge field. with v defined as in (9. given by 1 2 2 g v Aµ Aµ (9. After the gauge transformation we have only one degree of freedom from the scalar sector.125) the mass term for the gauge field is given by 1 AB AB (T CD )ik φk Wµ W µCD − g 2Tij φj 8 min min 1 aN aN = − g 2 Tij φj (T bN )ik φk Wµ W µbN 2 min min 1 1 a b aN aN = g 2v 2 δi δi Wµ W µbN = g 2 v 2 Wµ W µaN 2 2 228 (9.123) (Dµ )ij = δij ∂µ + i (T AB )ij Wµ 2 B A where (T AB )lm = i(δlA δm − δm δlB ). In the case of broken symmetry (µ2 < 0). N − 1 (9. The result looks a little bit strange. but the reason why we may read clearly the number of degrees of freedom only after the gauge transformation is that before the lagrangian contains a mixing term Aµ ∂ µ θ (9.

The situation is represented in Fig. 6.127) where a = 1. G H GW Fig. whereas the non zero action of the broken generators generate a mass term for the other gauge fields. The broken generators in GW correspond to massive vector bosons. From the previous derivation this follows noticing that the generators of H annihilate the minimum of the fields. The other degree of freedom is in the other factor. The correspondence among the fields 229 iN .aN Therefore. We can now show how to eliminate the Goldstone bosons. the symmetry of the vacuum. The broken generators do not belonging to GW correspond to massless Goldstone bosons. and GW . they have associated massless Goldstone bosons. the group of local symmetries. one can divide up the broken generators in two categories. 9. Finally the gauge fields associated to generators of GW lying in H remain massless. GW ∈ G. H ∈ G.4. In fact we can define new fields ξa and χ as aN φi = e−iT ξa (χ + v) (9. whereas Wµ . remain massless. H the subgroup of G leaving invariant the vacuum.1 -This figure shows the various groups. N − 1. G. In the second category fall the other broken generators. In general. leaving the corresponding gauge bosons massless. associated to the unbroken symmetry O(N − 1). · · · . the fields Wµ associated to the broken directions T aN acquire a mass ab g 2 v 2 /2. Th unbroken generators in GW correspond to massless vector bosons. and GW the group of local (gauge) symmetries. they have associated massive vector bosons. the global symmetry of the lagrangian. In the first category fall the broken generators lying in GW . that is the sum is restricted to the broken directions. if G is the global symmetry group of the lagrangian.1.

The free equations of motion are ( + m2 )Wµ − ∂µ (∂ ν Wν ) = 0 Let us study the solutions of this equation in momentum space. on the massive vector fields Wµ and on the ab massless field Wµ .φ and (ξa .129) showing that the ξa ’s are really the Goldstone fields. ni). k) we introduce µ(i) µ(3) = (0. χ) can be seen easily by expanding around the vacuum aN e−iT ξa iN a ≈ δiN − i(T aN )iN ξa = δiN + δi ξa (9.132) 4 2 where Wµ is the massive gauge field (called also a Yang-Mills massive field).134) (9. and 1 + 3(N − 1) + 2(N − 1)(N − 2)/2 = N 2 in the unitary gauge.133) It is convenient to introduce a set of four independent four-vectors.5 Quantization of a spontaneously broken gauge theory in the Rξ -gauge In the previous Section we have seen that in the unitary gauge some of the gauge fields acquire a mass. The unitary gauge is defined through the transformation φi → aN eiT ξa ij aN aN aN aN i (9. χ + v) (9. By putting k µ = (E. We define Wµ (x) = d4 k Wµ (k)e−ikx (9. E 230 i = 1.128) from which φi ≈ (vξa .131) Wµ → eiT ξa Wµ e−iT ξa − ∂µ eiT ξa e−iT ξa g This transformation eliminates the Goldstone degrees of freedom and the resulting aN lagrangian depends on the field χ. 2 k |k| (9. Notice again the counting of the degrees of freedom N +2N(N − 2 1)/2 = N in a generic gauge. φj = δiN (χ + v) (9. The corresponding quadratic part of the lagrangian will be (we consider here the abelian case corresponding to a symmetry U(1) or O(2)) 1 m2 2 L = − Fµν F µν + W (9.130) 9. = 1 m |k|.135) .

138) (9.142) The propagator is defined by the equation ( λ + m2 ) − ∂µ ∂ν Gνλ (x) = igµ δ 4 (x) (9. As a consequence the argument 231 . we can define a fourth unit vector µ(0) = kµ m (9.136) aλ (k) + kµ b(k) (9.144) Therefore. Therefore.λ =0 gλλ = g µν (9. the corresponding Feynman’s rule for the massive gauge field propagator is kµ kν i − 2 (9.137) = 0. When the momentum is on shell.140) In this way we have four independent unit vectors which satisfy the completeness relation 3 µ(λ) ν(λ ) λ.139) implying (k 2 − m2 )aλ (k) = 0. This propagator has a very bad high-energy behaviour. (9. corresponding to transverse and longitudinal polarization. Then we can decompose Wµ (k) as follows Wµ (k) = Since kµ µ(λ) µ(λ) ni · nj = δij (9.143) Solving in momentum space we find Gµν (k) = d4 k −ikx i e 4 2 − m2 + i (2π) k −g µν + kµkν m2 (9.142). In fact it goes to constant for k → ∞.141) Therefore the sum over the physical degrees of freedom gives 3 µ(λ) ν(λ) λ=1 = −g µν + kµ kν m2 (9. for m = 0 the field has three degrees of freedom with k 2 = m2 .145) g µν − k − m2 + i m2 Notice that the expression in parenthesis is nothing but the projector over the physical states given in eq.with k · ni = 0. from the equation of motion we get (−k 2 + m2 )( µ(λ) aλ (k) + kµ b(k)) + kµ k 2 b(k) = 0 m2 b(k) = 0 (9.

φ) = − 1 m4 + m2 χ2 + · · · 16λ 2 (9.4. The first term is the mass term for the W m2 = g 2 v 2 W 232 (9. whereas φ is the Goldstone field.153) The last term is an interaction term and it does not play any role in the following considerations.101)) V (χ. φ2 = φ (9. we introduce the fields φ1 = v + χ. φ) − Fµν F µν 2 2 4 where. (9. we have to take into account that the gauge invariance allows us to make different choices of the gauge. We will discuss this point only for the abelian case. Let us recall that under a transformation of O(2) the two real scalar field of the theory transform as δφ1 = −αφ2 .154) . In fact we will show that there exist an entire class of gauges where the theory is renormalizable by power counting.147) Assuming that the spontaneous breaking occurs along the direction φ1 . Let us consider the O(2) theory as presented in Section 9. but the argument cab be easily extended to the non-abelian case. in order to make it a gauge theory we introduce covariant derivatives Dµ φ1 = ∂µ φ1 + gWµ φ2 Dµ φ2 = ∂µ φ2 − gWµ φ1 with Wµ transforming under a local transformation as 1 δWµ (x) = ∂µ α(x) g (9. In terms of the new variables the lagrangian is v2 = − 1 1 1 L = (∂µ χ + gWµ φ)2 + (∂µ φ − gWµ χ − gvWµ )2 − V (χ. Although in the unitary gauge a spontaneously broken gauge theory does not seem to be renormalizable.152) (9. the potential up to quadratic terms is given by (see eq.148) (9.149) µ2 (9.for the renormalizability of the theory based on power counting does not hold any more. We see that the effect of the shift on φ1 is to produce the terms 1 2 2 2 g v W − gvWµ∂ µ φ + g 2 vW 2χ 2 (9.146) Therefore. δφ2 = αφ1 (9. with gauge group O(2).151) with with m2 = −2µ2 .150) λ Remember that χ is the Higgs field.

Finally the third contribution 1 − ξg 2v 2 φ2 2 is a mass term for the Goldstone field m2 = ξg 2v 2 = ξm2 φ W (9. the ghosts cannot be ignored. that we got rid of by choosing the unitary gauge. In fact remember that it disappears in the unitary gauge. in an example. as it should be. In fact let us consider a gauge fixing of the form f − B = 0. that evaluating a physical amplitude the contribution of the Goldstone disappears.158) it cancels the unwanted mixing term.156) 2β 2β 2β β 2β The first term is the usual covariant gauge fixing.whereas the second term is the mixing between the W and the Goldstone field. For this purpose we need to know the variation of the gauge fixing f under an infinitesimal gauge transformation δf (x) = giving 1 4 δf (x) = δ (x − y) + gvξ(χ(x) + v)δ 4 (x − y) δα(y) g Therefore the ghost lagrangian can be written as c∗ c + ξg 2v 2 c∗ c + ξg 2vχc∗ c (9.160) However this mass is gauge dependent (it depends on the arbitrary parameter ξ) and this is a signal of the fact that in this gauge the Goldstone field is fictitious. We are now in the position of reading the various propagators from the quadratic part of the lagrangian 233 .162) 1 α(x) + gvξα(x)(χ(x) + v) g (9. Finally we have to consider the ghost contribution from the Faddeev-Popov determinant.159) (9. We will see later. By proceeding as in Section 8.3 we know that the effect of such a gauge fixing is to add to the lagrangian a term − 1 1 ξ ξ2 2 2 2 1 2 f = − (∂ µ Wµ + gvξφ)2 = − (∂ µ Wµ )2 − gv∂ µ Wµ φ − g v φ (9. since they are coupled to the Higgs field. The second term.157) + gvWµ∂ µ φ β and we see that with the choice β=ξ (9. f = ∂ µ Wµ + gvξφ (9. after integration by parts reads ξ (9.155) where ξ is an arbitrary parameter and B(x) is an arbitrary function.163) Notice that in contrast with QED. However this choice is not unique.161) (9.

Of course.169) We see that for this class of gauges the asymptotic behaviour of the propagator is 1/k 2 and we can apply the usual power counting.• Higgs propagator (χ): i k 2 − m2 k2 i − ξm2 W • Goldstone propagator (φ): • Ghost propagator (c): − k2 i − ξm2 W The propagator for the vector field is more involved. for ξ → ∞ the ghost and the Goldstone fields decouple since their mass goes to infinity. Also. the singularity of the limit destroys the asymptotic behaviour that one has for finite values of ξ.167) (9.165) (9. let us collect the various quadratic terms in Wµ 1 1 1 − Fµν F µν + m2 W 2 − (∂ µ Wµ )2 W 4 2 2ξ In momentum space wave operator is 1 (g µν k 2 − k µ k ν ) − m2 g µν + k µ k ν W ξ which can be rewritten as g µν − kµ kν k2 (k 2 − m2 ) + W kµ kν 1 2 (k − ξm2 ) W 2 ξ k (9.164) The propagator in momentum space is the inverse of the wave operator (except for a factor −i).168) (9.166) (9. 234 . To invert the operator we notice that it is written as a combination of two orthogonal projection operators P1 and P2 αP1 + βP2 The inverse of such an operator is simply 1 1 P1 + P2 α β Therefore the W propagator is given by i 2 − m2 k W i = − 2 k − m2 W − g µν − g µν kµkν kµ kν iξ − 2 k2 k − ξm2 k 2 W kµkν − 2 (1 − ξ) k − ξm2 W (9. Notice that the unitary gauge is recovered in the limit ξ → ∞. In this limit the W -propagator becomes the one discussed at the beginning of the Section.

171) (9. since the gauge invariant quantities do not depend on the gauge fixing. We will write ¯ ˆ ¯ ˆ ¯ ¯ Lf = ψL iDψL + ψR i∂ψR − λf (ψL ψR Φ + ψR ψL Φ∗ ) with Dµ = ∂µ − igWµ (9.176) Here we have used again the complex notation for the scalar field. In fact ¯ ψL = and therefore 1 − γ5 2 ψ = ψ† 1 − γ5 ¯ 1 + γ5 γ0 = ψ 2 2 ¯ ¯ ψ = ψL ψR + ψR ψL (9. ψR → ψR where (9.175) .6 ξ-cancellation in perturbation theory In this Section we will discuss the cancellation of the ξ dependent terms in the calculation of a simple amplitude at tree level.To end this Section we observe that with this approach one is able to prove at the same time that the theory is renormalizable and that it does not contain spurious states. as in eq. We will also mimic the standard model by making transform under the gauge group only the left-handed part of the fermion field. But. in such a way to be gauge invariant and to gives rise to a mass term for the fermion when the symmetry is spontaneously broken. That is we decompose ψ as ψ = ψL + ψR 1 − γ5 1 + γ5 ψ.172) A fermion mass term would destroy the gauge invariance.109). they do not depend in particular on ξ. ψR = ψ 2 2 and we assume the following transformation under the gauge group ψL = ψL → eiα ψL .177) (9.173) ¯ ¯ 1 + γ5 + 1 − γ5 ψψ = ψ 2 2 (9. To this end we consider again the O(2) model of the previous Section. and we add to the theory a fermion field. In fact the theory is renormalizable for finite values of ξ and it does not contain spurious states for ξ → ∞. (9. 9. and therefore one gets the wanted result.170) (9.174) Therefore we will introduce a Yukawa coupling between the scalar field and the fermions. since it couples left-handed and right-handed fields. and we remind that it transforms as Φ → eiα Φ 235 (9.

. This expression shows that the fermion field acquires a mass through the Higgs mechanism given by λf v mf = √ 2 (9. φ λf Fig.6. Therefore we will consider only these two contributions. We get ¯ ˆ 1 − γ5 ψ + ψi∂ 1 + γ5 ψ + g ψγµ 1 − γ5 ψW µ ¯ ˆ ¯ Lf = ψi∂ 2 2 2 λf ¯ 1 + γ 5 1 − γ5 ¯ ψ(φ1 + iφ2 ) + ψ ψ(φ1 − iφ2 ) − √ ψ 2 2 2 from which λ ¯ ¯ ¯ ˆ ¯ 1 − γ5 ψW µ − √f ψψ(v + χ) + iψγ5 ψφ L = ψi∂ψ + g ψγµ 2 2 (9. p' k W + q p k' φ + χ Fig.180) The previous expression shows also that the Higgs field is a scalar (parity +1). The ξ dependence arises from the gauge field and the Goldstone field exchange.Goldstone-fermion vertex: √ γ5 2 236 .6.6. The amplitude gets contribution from the diagrams of Fig. 9. whereas the Goldstone field is a pseudoscalar (parity= -1). 9.178) where χ + v = φ1 and φ = φ2 .2 . The Feynman’s rule for the vertex is given in Fig.1. Let us start with the Goldstone exchange. The lagrangian can be rewritten in terms of the field ψ by writing explicitly the chiral projectors (1 ± γ5 )/2.179) (9. 9. We want to show that the ξ contribution cancels out.The fermion-fermion scattering in the U(1) Higgs model.1 . .2. 9.Therefore Lf is gauge invariant.6. Consider now the fermion-fermion scattering in this model at tree level.

The second term can be simplified by using the identity ¯ q µ u(p )γµ = u(p ) ¯ 1 − γ5 1 − γ5 u(p) = (p − p )µ u(p )γµ u(p) = ¯ 2 2 1 − γ5 1 + γ5 p−p ˆ ˆ u(p) = mf u(p )γ5 u(p) ¯ 2 2 (9.3 . The Feynman rule for the W-fermion vertex is given in Fig. 9.182) 2 q − ξmW 2 It is convenient to isolate the ξ dependence by rewriting the gauge field propagator as 1 −i qµqν 1−ξ g µν − 2 + q µ q ν − 2 2 2 − m2 q mW mW q − ξm2 W W qµqν qµqν −i −i = = 2 g µν − 2 + 2 q − m2 mW q − ξm2 m2 W W W (9.6.W-fermion vertex: igγµ The corresponding amplitude is ¯ MW = (ig)2 u(p )γµ × q2 −i − m2 W 1 − γ5 2 1 − γ5 u(p) 2 1 − γ5 qµqν (1 − ξ) u(k )γν ¯ g µν − 2 u(k) (9.184) By using this relation we have 1 − γ5 1 − γ5 u(p)¯(k )γµ u u(k) = 2 2 u u = (mf )¯(p )γ5 u(p)(−mf )¯(k )γ5 u(k) ¯ q µ q ν u(p )γµ 2 λf = − √ 2 v 2 u(p )γ5 u(p)¯(k )γ5 u(k) ¯ u 237 (9.6. Fig.185) . 9.181) where q = p − p = k − k.The corresponding amplitude is Mφ = λf √ 2 2 u(p )γ5u(p) ¯ q2 i u(k )γ5 u(k) ¯ − ξm2 W (9. .183) The first term is ξ independent and it is precisely the W propagator in the unitary gauge.3.

Therefore the result is the same that we would have obtained in the unitary gauge where there is no Goldstone contribution.λ =0 = i k 2 − m2 W 3 (λ) (λ) µ ν λ=1 + kµ kν −i k 2 − m2 m2 W W (9. For the choice ξ = 1 the gauge propagator has the very simple form −ig µν k 2 − m2 W (9.189) 2 − m2 k W it is to cancel the contribution of the time-like component of the gauge field. It is also worth to compare what happens for different choices of ξ.187) This corresponds to the Landau gauge. (9.since they are contracted with the fermionic current which is conserved. On the other hand. Substituting this result in MW we get MW = (ig)2 u(p )γµ ¯ + −i 1 − γ5 u(p) 2 2 q − m2 W λ √f 2 2 g µν − qµqν m2 W u(k )γµ ¯ 1 − γ5 u(k) 2 (9. In ¯ fact the fermionic current ψγµ (1 − γ5 )/2ψ it is not conserved because the fermion acquires a mass through the spontaneous breaking of the symmetry. which (0) is described by the polarization vector µ .180). Now let us make some comment about the result. In QED we have used the fact that the q µ q ν terms in the propagator do not play a physical role.188) This is the Feynman-’t Hooft gauge. which however it is cancelled by the Goldstone pole which also is at k 2 = 0. For ξ = 0 the gauge propagator is transverse −i 2 − m2 k W g µν − kµkν k2 (9. In the present case this does not happens. In fact the gauge propagator can be rewritten as −i −ig µν = 2 2 k 2 − mW k − m2 W 3 (λ) (λ ) µ ν gλλ λ. which has a propagator i (9. 238 .where we have made use of eq. Here the role of the Goldstone boson. The propagator has a pole term at k 2 = 0. using m2 = g 2 v 2 we see that the last term cancels exactly the Mφ ampliW tude. but we get a term proportional to the fermion mass. whereas the second piece is cancelled by the Goldstone boson contribution.186) −i g2v2 q 2 − ξm2 m2 W W u(p )γ5 u(p)¯(k )γ5 u(k) ¯ u Finally. the term that one gets in this way it is just the one necessary to give rise to the cancellation with the Goldstone boson contribution.190) The first piece is the propagator in the unitary gauge.

Chapter 10 The Standard model of the electroweak interactions
10.1 The Standard Model of the electroweak interactions

We will describe now the Weinberg Salam model for the electroweak interactions. Let us recall that the Fermi theory of weak interactions is based on a lagrangian involving four Fermi fields. This lagrangian can be written as the product of two currents which are the sum of various pieces. GF (+) LF = √ Jµ J µ(−) 2 (10.1)

Limiting for the moment our considerations to the charged current for the electron and its neutrino (here ν ≡ νe )), the current is given by
(+) ¯ Jµ = ψe γµ (1 − γ5 )ψν , (−) (+) ¯ Jµ = Jµ = ψν γµ (1 − γ5 )ψe †

(10.2)

To understand the symmetry hidden in this couplings it is convenient to introduce the following spinors with 2 × 4 components L= (ψν )L (ψe )L = 1 − γ5 2 ψν ψe (10.3)

where we have introduced left-handed fields. The right-handed spinors, projected out with (1 + γ5 )/2, don’t feel the weak interaction. This is, in fact, the physical meaning of the V − A interaction. By using ¯ ¯ ¯ ¯ 1 + γ5 ¯ L = (ψν )L , (ψe )L = ψν , ψe 2 (10.4)

239

we can write the weak currents in the following way
(+) ¯ Jµ = 2ψe

1 + γ5 1 − γ5 0 ¯ ¯ γµ ψν = 2 (ψν )L , (ψe )L γµ (ψν )L 2 2 0 0 (ψν )L ¯ ¯ ¯ = 2Lγµ τ− L = 2 (ψν )L , (ψe )L γµ 1 0 (ψe )L τ± =

(10.5)

τ1 ± iτ2 2 the τi ’s being the Pauli matrices. Analogously
(−) ¯ Jµ = 2Lγµ τ+ L

where we have defined

(10.6)

(10.7)

In the fifties the hypothesis of the intermediate vector bosons was advanced. According to this idea the Fermi theory was to be regarded as the low energy limit of a theory where the currents were coupled to vector bosons, very much as QED. However, due to the short range of the weak interactions these vector bosons had to have mass. The interaction among the vector bosons and the fermions is given by g ¯ g (+) (−) Lint = √ Jµ W −µ + Jµ W +µ = √ Lγµ (τ− W −µ + τ+ W +µ )L (10.8) 2 2 2 The coupling g can be related to the Fermi constant through the relation g2 GF √ = 8m2 2 W Introducing real fields in place of W± W± = we get W1 ∓ iW2 √ 2 (10.10) (10.9)

g¯ µ µ (10.11) Lint = Lγµ (τ1 W1 + τ2 W2 )L 2 Let us now try to write the electromagnetic interaction within the same formalism ¯ ¯ Lem = eψe Qγµ ψe Aµ = eψe Q 1 + γ5 1 − γ5 1 − γ5 1 + γ5 γµ + γµ 2 2 2 2 µ ¯ ¯ = e[(ψe )L Qγµ (ψe )L + (ψe )R Qγµ (ψe )R ]A ψe Aµ (10.12)

where

1 + γ5 ¯ ¯ 1 − γ5 ψe , (10.13) (ψe )R = ψe 2 2 where Q = −1 is the electric charge of the electron, in unit of the electric charge of the proton, e. It is also convenient to denote the right-handed components of the electron as R = (ψe )R (10.14) (ψe )R = 240

We can write ¯ ¯ (ψe )L γµ (ψe )L = Lγµ and 0 0 0 1

¯ 1 − τ3 L L = Lγµ 2

(10.15)

¯ ¯ (ψe )R γµ (ψe )R = Rγµ R from which, using Q = −1, 1¯ em ¯ ¯ τ3 Lem = e − Lγµ L − Rγµ R + Lγµ L Aµ = ejµ Aµ 2 2 From this equation we can write 1 Y em 3 jµ = jµ + jµ 2 with
3 ¯ jµ = Lγµ

(10.16)

(10.17)

(10.18)

τ3 L 2

(10.19) (10.20)

and
Y ¯ ¯ jµ = −Lγµ L − 2Rγµ R 3 These equations show that the electromagnetic current is a mixture of jµ , which is Y a partner of the charged weak currents, and of jµ , which is another neutral current. 3 The following notations unify the charged currents and jµ i ¯ τi jµ = Lγµ L 2

(10.21)

Using the canonical anticommutators for the Fermi fields, it is not difficult to prove that the charges associated to the currents Qi = span the Lie algebra of SU(2), [Qi , Qj ] = i
k ijk Q i d3 x j0 (x)

(10.22)

(10.23)

Y The charge QY associated to the current jµ commutes with Qi , as it can be easily verified noticing that the right-handed and left-handed fields commute with each other. Therefore the algebra of the charges is

[Qi , Qj ] = i

k ijk Q ,

[Qi , QY ] = 0

(10.24)

This is the Lie algebra of SU(2) ⊗ U(1), since the Qi ’s generate a group SU(2), whereas QY generates a group U(1), with the two groups commuting among themselves. To build up a gauge theory from these elements we have to start with an initial lagrangian possessing an SU(2) ⊗ U(1) global invariance. This will produce 241

4 gauge vector bosons. Two of these vectors are the charged one already introduce, whereas we would like to identify one of the the other two with the photon. By evaluating the commutators of the charges with the fields we can read immediately the quantum numbers of the various particles. We have [Lc , Qi ] = and [R, Qi ] = 0 [Lc , QY ] = −Lc , [R, QY ] = −2R (10.26) (10.27) d3 x [Lc , L† a τi 2 Lb ] = τi 2 Lb (10.25)

ab

cb

These relations show that L transforms as an SU(2) spinor, or, as the representation 2, where we have identified the representation with its dimensionality. R belongs to the trivial representation of dimension 1. Putting everything together we have the following behaviour under the representations of SU(2) ⊗ U(1) L ∈ (2, −1), R ∈ (1, −2) (10.28)

The relation (10.18) gives the following relation among the different charges 1 Qem = Q3 + QY 2 (10.29)

As we have argued, if SU(2) ⊗ U(1) has to be the fundamental symmetry of electroweak interactions, we have to start with a lagrangian exhibiting this global symmetry. This is the free Dirac lagrangian for a massless electron and a left-handed neutrino ¯ ˆ ¯ ˆ L0 = Li∂L + Ri∂R (10.30) Mass terms, mixing left- and right-handed fields, would destroy the global symmetry. For instance, a typical mass term gives ¯ mψψ = mψ † 1 + γ5 1 − γ5 1 − γ5 1 + γ5 ¯ ¯ γ0 + γ0 ψ = m(ψR ψL + ψL ψR ) (10.31) 2 2 2 2

We shall see later that the same mechanism giving mass to the vector bosons can be used to generate fermion masses. A lagrangian invariant under the local symmetry SU(2) ⊗ U(1) is obtained through the use of covariant derivatives τi i g ¯ ¯ L = Liγ µ ∂µ − ig Wµ + i Yµ L + Riγ µ (∂µ + ig Y µ ) R 2 2 The interaction term is ¯ Lint = g Lγµ τ g ¯ ¯ · W µ L − Lγµ LY µ − g Rγµ RY µ 2 2 242 (10.33) (10.32)

In this expression we recognize the charged interaction with W ± . We have also, as expected, two neutral vector bosons W 3 and Y . The photon must couple to the electromagnetic current which is a linear combination of j 3 and j Y . The neutral vector bosons are coupled to these two currents, so we expect the photon field, 3 Aµ , to be a linear combination of Wµ and Yµ . It is convenient to introduce two orthogonal combination of these two fields, Zµ , and Aµ . The mixing angle can be identified through the requirement that the current coupled to Aµ is exactly the electromagnetic current with coupling constant e. Let us consider the part of Lint involving the neutral couplings
µ 3 LN C = gjµ W3 +

g Y µ j Y 2 µ

(10.34)

By putting (θ is called the Weinberg angle) W3 = Z cos θ + A sin θ Y = −Z sin θ + A cos θ we get 1 Y 1 Y 3 3 LN C = g sin θ jµ + g cos θ jµ Aµ + g cos θ jµ − g sin θ jµ Z µ 2 2 The electromagnetic coupling is reproduced by requiring the two conditions g sin θ = g cos θ = e (10.37) (10.36)

(10.35)

These two relations allow us to express both the Weinberg angle and the electric charge e in terms of g and g . In the low-energy experiments performed before the LEP era (< 1990) the fundamental parameters used in the theory were e (or rather the fine structure constant α) and sin2 θ. We shall see in the following how the Weinberg angle can be eliminated in favor of the mass of the Z, which is now very well known. By using eq. em 3 Y (10.37), and jµ = jµ + 1 jµ , we can write 2
em 3 em 3 LN C = ejµ Aµ + [g cos θjµ − g sin θ(jµ − jµ )]Z µ

(10.38)

Expressing g through g = g tan θ, the coefficient of Zµ can be put in the form
3 em [g cos θ + g sin θ]jµ − g sin θ jµ =

g 3 g em jµ − sin2 θ jµ cos θ cos θ

(10.39)

from which
em LN C = ejµ Aµ +

g g Z µ 3 em em [jµ − sin2 θ jµ ]Z µ ≡ ejµ Aµ + j Z cos θ cos θ µ
Z 3 em jµ = jµ − sin2 θ jµ

(10.40)

where we have introduced the neutral current coupled to the Z (10.41)

243

We get QY (φ0 ) = 2 × 0 − − 1 2 =1 (10. in such a way to have the corresponding gauge particle (the photon) massless.43) 2 e g g 10. we use the relation 1 Qem = Q3 + QY 2 for the lower component of Φ. (10. both the components of the doublet must have the same value of QY . and using again eq.44) To determine the weak hypercharge.45).23 (10.45) Since QY and Qi commute. Since SU(2) ⊗ U(1) has four generators. (10. we will be left with one unbroken symmetry. The simplest choice turns out to be a complex representation of SU(2) of dimension 2 (Higgs doublet).The value of sin2 θ was evaluated initially from processes induced by neutral currents (as the scattering ν − e) at low energy. As we already noticed the vacuum should be electrically neutral.2 The Higgs sector in the Standard Model According to the discussion made in the previous Section. Therefore.37) gives the following relation among the electric charge and the couplings g and g 1 1 1 = 2+ 2 (10.47) which justifies the notation φ+ for the upper component of the doublet.42) This shows that both g and g are of the same order of the electric charge. of Φ. we need three broken symmetries in order to give mass to W ± and Z. The approximate value is sin2 θ ≈ 0. the group U(1)em must be a symmetry of the vacuum (that is the vacuum should be electrically neutral). QY . 244 . that we should identify with the group U(1) of the electromagnetism (U(1)em ). so one of the components of the doublet must also be neutral. we get Qem (φ+ ) = 1 1 + =1 2 2 (10. than we will write Φ= φ+ φ0 (10. Assume that this component is the lower one. To realize this aim we have to introduce a set of scalar fields transforming in a convenient way under SU(2) ⊗ U(1).46) (10. Notice that the eq.

55) From this expression we read immediately the particle masses (φ− = (φ+ )† ) m2 = m2 + = m2 − = 0 η φ φ and m2 = 2λv 2 = −2µ2 h (10.51) 0 φ+ √ √ + ≡ Φ0 + Φ v/ 2 (h + iη)/ 2 (10. λ = m2 /2v 2 ).The most general lagrangian for Φ with the global symmetry SU(2) ⊗ U(1) and containing terms of dimension lower or equal to four (in order to have a renormalizable theory) is LHiggs = ∂µ Φ† ∂ µ Φ − µ2 Φ† Φ − λ(Φ† Φ)2 (10.52) (10. This is called the Higgs field.48) where λ > 0 and µ2 < 0.49) (10.54) (10.57) It is also convenient to express the parameters appearing in the original form of the potential in terms of m2 e v (µ2 = −2m2 . and a massive + 2 2 2 v2 µ2 = =− 2λ 2 µ2 λ 0 √ v/ 2 (10.56) Summarizing we have three massless Goldstone bosons φ± scalar h. 245 . In this way we get h h h VHiggs 1 m2 m2 1 m2 = − m2 v 2 + h h2 + h h |φ+ |2 + (h2 + η 2 ) + h h 8 2 v 2 2v 2 1 |φ | + (h2 + η 2 ) 2 (10.58) and η.50) (10.53) (10. The potential has infinitely many minima on the surface |Φ|2 minimum with v2 = − Let us choose the vacuum as the state 0|Φ|0 = By putting Φ= with 0|Φ |0 = 0 we get 1 1 |Φ|2 = v 2 + vh + |φ+ |2 + (h2 + η 2 ) 2 2 from which 1 µ4 1 1 + λv 2 h2 + 2λvh |φ+ |2 + (h2 + η 2 ) + λ |φ+ |2 + (h2 + η 2 ) VHiggs = − 4 λ 2 2 (10.

h) ≈ (ξ2 .64) 0 √ (v + h)/ 2 (10. and τ2 which are broken. φ2 .68) 1 0 246 . except for the substitution of Φ with √ (0.61) (v + h)/ 2 According to our rules the exponential should contain the broken generators. In fact there are τ1 . η.60) To study the mass generation it is convenient to write Φ in a way analogous to the one used in eq. We will also denote old and new gauge fields with the same symbol. at any rate.59) (10. so the previous expression gives a good representation of Φ around the vacuum. h) By performing the gauge transformation Φ → e−iξ · τ /v Φ Wµ · i τ τ → e−iξ · τ /v Wµ · eiξ · τ /v + 2 2 g Yµ → Yµ ∂µ e−iξ · τ /v eiξ · τ /v (10. by now. −ξ1 . In fact. 1) of SU(2) ⊗ U(1).65) (10. we have g g H Dµ = ∂µ − i τ · Wµ − i Yµ 2 2 and H LHiggs = (Dµ Φ)† D Hµ Φ − µ2 Φ† Φ − λ(Φ† Φ)2 (10. Furthermore it should contain the combination of 1 and τ3 which is broken (remember that (1 + τ3 )/2 is the electric charge of the doublet Φ. −ξ3 .67) (10. Φu = (10. Recalling that Φ ∈ (2.63) the Higgs lagrangian remains invariant in form. we promote the global symmetry to a local one by introducing the covariant derivative.66) (10. But.62) (10. and it is conserved).As usual. τ3 is a broken generator. expanding around this state 1 + iξ3 /v i(ξ1 − iξ2 )/v 1 − iξ3 /v i(ξ1 + iξ2 )/v 1 i(ξ1 − iξ2 ) ≈ √ (v + h − iξ3 ) 2 √ and introducing real components for Φ (φ+ = (φ1 − iφ2 )/ 2) eiξ · τ /v 0 √ (v + h)/ 2 ≈ 1 φ1 − iφ2 Φ= √ v + h + iη 2 we get the relation among the two sets of coordinates (φ1 .127) 0 √ Φ = eiξ · τ /v (10. (v + h)/ 2). Defining 0 1 Φd = . (9.

72) (10.73) From eq.69) Φ → √ Φd 2 We get −i g g τ ·W + Y 2 2 v √ Φd = −i 2 g v g g √ (τ− W − + τ+ W + ) + τ3 W 3 + Y √ Φd 2 2 2 2 g v 1 = −i √ √ W + Φu − (gW 3 − g Y )Φd (10. Φ with its expectation value v (10. The Higgs sector brings in two additional parameters. and of the Weinberg angle. In this way we get g 2v 2 + 2 1 g 2 + g 2 2 2 |W | + v Z 4 2 4 Finally the mass of the vector bosons is given by 1 2 MW = g 2v 2 . in the kinetic term. or from the gauge coupling constants 2 g2 MW 2 = 2 2 = cos θ 2 MZ g +g (10. cos θ = g g2 + g 2 (10. (10. 4 1 2 2 MZ = (g 2 + g )v 2 4 (10. orthogonal to the photon. which can also be expressed in terms of the electric charge (or the fine structure constant). the mass term is v2 g2 + 2 1 |W | + (gW 3 − g Y )2 2 2 4 From eq. the mass of the Higgs.74) Notice that the masses of the W ± and of the Z are not independent.70) 2 2 2 Since Φd and Φu are orthogonals.75) (10.71) allowing us to write the mass term in the form v2 g2 + 2 g2 + g 2 |W | + (W 3 cos θ − Y sin θ)2 2 2 4 (10.the mass terms for the scalar fields can be read by substituting. and therefore sin θ = g g2 + g 2 . (10.76) Let us now discuss the parameters that we have so far in the theory.37) we have tan θ = g /g. since their ratio is determined by the Weinberg angle. 247 . The gauge interaction introduces the two gauge couplings g and g .35) we see that the neutral fields combination is just the Z field.

78) Therefore. −1). Another possibility is to use the mass of the Z 2 MZ = e2 1 1 πα √ =√ 2 2θ 4 2GF sin θ cos 2GF sin2 θ cos2 θ sin2 θ = 1 1− 2 1− √ 4πα 2 2GF MZ (10.83) In the unitary gauge. the trilinear term generates the electron mass.82) we see that the following coupling (Yukawian coupling) is invariant φ+ ¯ ¯ ¯ LY = ge LΦR + h. Φ ∈ (2. sin2 θ and GF . φ0 (10. (ψe )L ] (ψe )R + h.c. −2).65) both for Φ and for the lepton field L. However. due to the breaking of the symmetry.81) Therefore.9) GF g2 √ = 2 8MW 2 2 using the expression for MW . since it is impossible to construct bilinear terms in the electron field which are invariant under the gauge group. In fact. R ∈ (1. we get (10. the mass of the Z is very well known MZ = (91.84) 248 . (10. The last problem we have to solve is how to give mass to the electron. and at the moment we have only an experimental lower bound on mh . we get ¯ ¯ LY = ge [ (ψν )L .80) The first alternative has been the one used before LEP. Once the field Φ acquires a non vanishing expectation value.0021) GeV (10. and the expectation value of the field Φ. from LEP.mh .79) to eliminate sin2 θ (10. Recalling the behaviour of the fields under SU(2) ⊗ U(1) L ∈ (2.1867 ± 0. The parameter v can be expressed in terms of the Fermi coupling constant GF . +1) (10. from eq.c. v. which is given by mh > 60 GeV .c. which is obtained by using the transformation (10. (v + h)/ 2 (10. the three parameters g. = ge [ (ψν )L . (ψe )L ] 0 √ (ψe )R + h. The Higgs particle has not been yet discovered. after LEP1. GF and MZ .77) v2 = √ 1 ≈ (246 GeV )2 2GF (10. The solution is that we can build up trilinear invariant terms in the electron field and in Φ. g and v can be traded for e. the parameters that are now used as input in the SM are α.

7)) Jd with ΨL = uL dL = µ(±) ¯ = 2ΨL γ µ τ∓ ΨL 1 − γ5 2 u d (10. 2 2 (10. p ≈ uud. This has been done with a theory that. First of all we recall that in the quark model the nucleon is made up of three quarks. also the Yukawian coupling has only dimension 3.86) we get ge v ¯ ge ¯ (10.from which ge v ¯ ge ¯ LY = √ (ψe )L (ψe )R + √ (ψe )L (ψe )R h + h. the weak transition n → p + e− + νe is ¯ obtained through ¯ (10. Then.87) LY = √ ψe ψe + √ ψe ψe h 2 2 Therefore the symmetry breaking generates an electron mass given by ge v me = − √ 2 (10. The extension to the quarks is a little bit less obvious and we will follow the necessary steps in this Section. with composition.c. Therefore.91) 249 .88) In summary.5) and (10. we have been able to reproduce all the phenomenological features of the V −A theory and its extension to neutral currents. (10. before spontaneous symmetry breaking is renormalizable. n ≈ udd. In fact.3 The electroweak interactions of quarks and the Kobayashi-Maskawa-Cabibbo matrix So far we have formulated the SM for an electron-neutrino pair.90) (10. 10. In fact the proof was given only at the beginning of the seventies by ’t Hooft.85) Since ¯ 1 + γ5 ψe )† = ψ † 1 + γ5 γ0 ψe = ψe 1 − γ5 ψe = (ψe )R (ψe )L ¯ ¯ ¯ [(ψe )L (ψe )R ]† = (ψe e 2 2 2 (10. The proof that the renormalizability of the theory holds also in the case of spontaneous symmetry breaking (µ2 < 0) is absolutely non trivial. it is natural to write quark current in a form analogous to the leptonic current (see eqs. but it can be extended in a trivial way to other lepton pairs.89) d → u + e− + νe and the assumption is made that quarks are point-like particles as leptons.

101) uL dL cos θC + sL sin θC ≡ uL dC L (10.96) As far as the right-handed components are concerned.92) u s with ΨL = uL sL = 1 − γ5 2 (10.100) 3 2 3 and (Qem (s) = Qem (d) = −1/3) 1 1 1 QY (dC ) = 2(− + ) = L 3 2 3 (10. given by µ(±) ¯ = 2ΨL γ µ τ∓ ΨL Js (10.93) Furthermore the total current contributing to the Fermi interaction had to be of the form µ µ µ Jh = Jd cos θC + Js sin θC (10.03 (10. ) R 3 3 250 .98) This allows us to assign QL to the representation (2. and sin θC = 0.95) The presence of the angle in the current was necessary in order to explain why the weak decays strangeness violating (that is the ones corresponding to the transition u ↔ s.21 ± 0.99) 2 we get 2 1 1 QY (uL) = 2( − ) = (10.97) µ(±) ¯ = 2QL γ µ τ∓ QL (10. Collecting together the two currents we get Jh with QL = and dC = dL cos θC + sL sin θC L (10. as K + → µ+ νµ ) where suppressed with respect to the decays strangeness conserving. we assign them to SU(2) singlets. from the hyperon decays it was known that it was necessary a further current involving the quark s. − ) (10.94) where θC is the Cabibbo angle. from 1 Qem = Q3 + QY (10. 1/3) of SU(2) ⊗ U(1). obtaining 4 2 dC ∈ (1.102) uR ∈ (1. In fact.Phenomenologically.

GIM) that another quark with charge +2/3 should exist. Q2 ] = = i d3 xd3 y [Q† τ1 QL . with K + → µ+ νµ induced by the charged current (see Fig. which from dC† dC = d† d cos2 θC + s† s sin2 θC + (d† s + s† d) sin θC cos θC (10.104) ¯ R 3 3 3 R 3 em Since the Z is coupled to the combination jµ − sin2 θjµ . it is easily seen that the coupling contains a bilinear term in the field dC given by − 1 1 2 1 ¯ ¯ − sin θ Z µ dC γµ dC + sin2 θZ µ dC γµ dC L L R R 2 3 3 (10. Q† τ2 QL ] = L L d3 x Q† τ3 QL = i L d3 x Q† [τ1 . d. we find Γ(K 0 → µ+ µ− ) ≤ 6 × 10−5 + → µ+ ν ) Γ(K µ (10.3). the charm.107) is proportional to sin θC cos θC . In a more general way we can get this result from the simple 3 observation that W3 is coupled to the current jµ which has an associated charge Q3 which can be obtained by commuting the charges Q1 and Q2 [Q1 . Then.109) . one can form two left-handed doublets Q1 = L where sC = −dL sin θC + sL cos θC L 251 (10. However the experimental result is that these transitions are strongly suppressed.108) This problem was solved in 1970 by the suggestion (Glashow Illiopoulos and Maiani. We can compare the strength of a FCNC transition. But if we compare K 0 → µ+ µ− induced by the neutral current . s are given by τ3 3 ¯ (10. τ2 ]QL L (10. with the strength of a flavour changing charged current transition as u → s.106) d3 x(u† uL − dC† dC ) L L L The last term is just the charge associated to a Flavour Changing Neutral Current (FCNC). From this comparison we expect the two transitions to be of the same order of magnitude. the hadronic neutral currents contribution from quarks u.103) jµ = QL γµ QL 2 1¯ 4 2¯ Y jµ = QL γµ QL + uR γµ uR − dC γµ dC (10. 2.105) ¯ This gives rise to terms of the type ds and sd which produce strangeness changing ¯ neutral current transitions . Q2 = L cL sC L (10. using BR(K 0 → µ+ µ− ).110) uL dC L . which is proportional to sin θC .Therefore.

The same considerations apply to QY .112) since the transformation matrix = (10. the bottom or beauty. with mb ≈ 5 GeV . is the combination orthogonal to dC . Finally in 1995 the partner of the bottom. known as J/ψ.113) is orthogonal.3. The mass of the top is around 175 GeV . since we need a cancellation at least at the order G2 in order to cope with F the experimental bounds. the top. bb. with ¯ PC −− = 1 . 10. was discovered at Fermi Lab. as it should be. The charm was discovered in 1974. This mechanism of cancellation turns out to be effective also at the second order in the interaction. As a consequence the expression of Q3 gets L modified d3 x (u† uL + c† cL − dC† dC − sC† sC ) (10. In fact. one can see that the two second order graphs contributing to K 0 → µ+ µ− cancel out except for terms coming from the mass difference between the quarks u and c.111) Q3 = L L L L L L But dC† dC + sC† sC = d† dL + s† sL L L L L L L dC L sC L cos θC − sin θC sin θC cos θC dL sL (10. the Υ. J In 1977 there was the observation of a new vector resonance. One gets A(K 0 → µ+ µ− ) ≈ G2 (m2 − m2 ) F c u (10. interpreted as a bound state ¯ where b is a new quark.The neutral current flavour changing process K 0 → µ+ µ− and he charged current flavour changing process K + → µ+ νµ .5 GeV . We see that in association with three 252 .1 .114) allowing to get a rough estimate of the mass of the charm quark. mc ≈ 1 ÷ 3 GeV . when it was observed the bound state cc. t. The mass of charm was evaluated to be around 1.d K 0 Z s – µ + u K + W s – + µ + µ – νµ Fig.

121) + fR .115) L .119) and the right-handed singlets by eAR . A = 1. This is very important because it guarantees the renormalizability of the SM. dAR (10. Bell. Summarizing we have three generations of quarks and leptons. We will denote the last ones by AL = νA eA . Inside each generation the total charge is equal to zero. L . Each generation includes two left-handed doublets νA uA . 3 (10. the conservation of the currents coupled to the YangMills fields is crucial for the renormalization properties. 3 (10.117) e− L dA L A and the corresponding right-handed singlets (except for the neutrinos if we assume that they are massless).118) i 3 3 f ∈gen where we have also taken into account that each quark comes in three colors. Jackiw anomalies. showing that the mixing of different quarks arises naturally from the fact that the quark mass eigenstates are not necessarily the same fields which couple to the gauge fields.leptonic doublets νe e− there are three quark doublets u d .120) We assume that the neutrinos are massless and that they do not have any righthanded partner. We will complete now the formulation of the SM for the quark sector. L qAL = uA dA . In the case of the SM for the electroweak interactions one can show that the condition for the absence of these corrections (Adler. 2. in general there are quantum corrections to the divergences of the currents destroying their conservation. L (10. L ντ τ− t b (10. 2. However. L νµ µ− c s . In fact. In fact 2 1 − Qf = 0 − 1 + 3 × Qtot = =0 (10. L A = 1. uAR .116) L We will show later that experimental evidences for the quark b to belong to a doublet came out from PEP and PETRA much before the discovery of top. see later) is that the total electric charge of the fields is zero. The gauge interaction is then L = fL τ QY (fL ) ¯ Yµ γ µ fAL fAL i∂µ + g · Wµ + g 2 2 QY (fR ) ¯ fAR i∂µ + g 2 253 γ µ fAR (10.

125) ˜ Observing that QY (Φ) = −1. a mass term for the down quark is generated gd v ¯ gd v ¯ ¯ √ dLdR + dR dL = √ dd (10. QY (eAR ) = −2. (10. dAR . AB (10.129) .c. (10. √ giving mass to the up quark. The most general Yukawian coupling among quarks and Higgs field is then given by LY = AB i where gAB .127) Lfermion mass v =√ 2 e u d ¯ ¯ ¯ gAB eAL eBR + gAB uAL uBR + gAB dAL dBR + h. We start with up and down quarks.123) = where√Φ is the Higgs doublet. v/ 2). u. In order to give mass to the up quark we need to introduce the conjugated Higgs doublet defined as ˜ Φ = iτ2 Φ = (φ0 ) −(φ+ ) (10. uAR . Φ (0. and fAR = eAR .126) gu qL ΦuR + h.c. with i = e. d. another invariant Yukawian coupling is given by ¯ ˜ (10. and a Yukawian coupling given by ¯ gd qL ΦdR + h.124) 2 2 √ corresponding to a mass md = −gd v/ 2.c. When we shift Φ by its vacuum expectation value we get a mass term for the fermions given by e u d ¯ ˜ ¯ gAB ¯AL ΦeBR + gAB qAL ΦuBR + gAB qAL ΦdBR + h. u.128) Therefore we obtain three 3 × 3 mass matrices v i i MAB = − √ gAB . are arbitrary 3 × 3 matrices. qAL . We recall also the weak hypercharge assignments QY ( AL ) = −1.c.122) Let us now see how to give mass to quarks. When this takes its expectation value.where fAL = AL . QY (dAR ) = − 2 3 1 3 4 QY (uAR ) = 3 QY (qAL ) = (10. mu = −gu v/ 2. d (10. 2 254 i = e.

In fact.140) Defining we get V = (S u )−1 S d . U = H −1 M (10.136) (10.132) we see that H is hermitian and positive definite. This follows from the polar decomposition of an arbitrary matrix M = HU (10. V † = V −1 (V ) h(+) Jµ = 2¯ALγµ dAL u with dAL = VAB dBL (V ) The matrix V is called the Cabibbo-Kobayashi-Maskawa (CKM) matrix. ψR = T ψR (10. and that U is unitary. and it describes the mixing among the down type of quarks (this is conventional we could 255 .These matrices give mass respectively to the charged leptons and to the up and down of quarks.139) (10. Then we get ¯ ¯ ¯ ψL MψR = ψL S(S † MT )T † ψR ≡ ψL Md ψR where we have defined the mass eigenstates ψL = SψL .138) (10.131) √ where H † = H = MM † is a hermitian definite positive matrix and U † = U −1 is a unitary matrix.133) with T = U −1 S is a unitary matrix.130) i where S and T are unitary matrices (depending on i) and Md are three diagonal matrices. They can be diagonalized by a biunitary transformation i Md = S † M i T (10.135) We can now express the charged current in terms of the mass eigenstates. We have +(h) Jµ = 2¯AL γµ τ− qAL = 2¯ALγµ dAL = 2¯ALγµ ((S u )−1 S d )AB dBL q u u (10.137) (10. by defining H 2 = MM † . Each of the mass terms has the structure ¯ ψL MψR . if the unitary matrix S diagonalizes H we have Hd = S † HS = S † MU −1 S = S † MT (10.134) −1 (10. Furthermore one can take the eigenvalues to be positive. since U † U = M † H −2M = M † M † M −1 M = 1 UU † = H −1 MM † H −1 = H −1 H 2 H −1 = 1 Therefore.

in the case of two generations V depends only on one real parameter (the Cabibbo angle). For three generations V depends on three real parameters and one phase. We see that its physical origin is that. the matrix V . A violation of CP has been observed experimentally by Christensen et al. implies a CP violation. By relaxing this assumption we may generate a mixing in the leptonic sector producing a violation of the different leptonic numbers. However one is free to choose in an arbitrary way the phase for the 2n up and down quark fields. M d . in general. Since the invariance under the discrete symmetry CP implies that all the couplings in the lagrangian must be real.144) 2 real parameters. We can see that. In fact h(3) ¯ ¯ ¯ ¯ ¯ ¯ jµ ≈ uLuL − dL dL = uL(S u )−1 S u uL − dL(S d )−1 S d dL = uL uL − dL dL (10.143) parameters.145) Then. In fact a real unitary matrix is nothing but an orthogonal matrix which depends on n(n − 1) (10.141) For the charged leptonic current. In the case we have n generations of quarks and leptons. since we have assumed massless neutrinos we get ν ν Jµ(+) = 2¯AL γµ (S e )AB eBL ≡ 2¯AL γµ eAL (10.have chosen to mix the up quarks as well). M u . But an overall phase does not change V . it follows that the SM. in 1964. These authors discovered that the eigenstate of CP . It is interesting to discuss in a more detailed way the structure of the CKM matrix. there are no relations between the mass matrix of the up quarks. Therefore.142) where νL = (S e )† νL is again a massless eigenstate. Therefore the CKM matrix depends only on n2 − (2n − 1) = (n − 1)2 (10. depends on n2 parameters. in general. This means that V will depend on a number of phases given by number of phases = (n − 1)2 − (n − 1)(n − 2) n(n − 1) = 2 2 (10. being unitary. among different generations. it is impossible to choose the phases in such a way to make V real. The neutral currents are expressions which are diagonal in the ”primed” states. there is no mixing in the leptonic sector. due to the unitarity of the various S matrices.146) . therefore their expression is the same also in the basis of the mass eigenstates. However this is tied to our assumption of massless neutrinos. with CP = −1 1 0 ¯ |K2 = √ |K 0 − |K 0 2 256 (10. in the case of three generations. and the mass matrix of the down quarks.

01 ± 0.08 ± 0.004 − 0.0018 (10.224 0.016 and |Vcs | = 1.153) More recently from the branching ratio t → W b.152) (10.002 − 0.9750 0.036 − 0.decays into a two pion state with CP = +1 with a BR 0 BR(K2 → π + π − ) ≈ 2 × 10−3 (10.218 − 0.9757 0.014 0.9745 − 0. The result is |Vcd| = 0.2205 ± 0. using the muon decay as normalization.L.15 ± 0.151) The elements Vub and Vcb are determined from the b → u and b → c semi-leptonic decays as evaluated in the spectator model. if the SM is able to explain the observed CP violation in quantitative terms.97 ± 0.034 − 0.041 ± 0. The elements Vud and Vus are determined through nuclear β-.150) (10.07 (10. One gets |Vub | = 0.9989 − 0.155) 0.224 0.046 ⎠ (10.003 (10.219 − 0.147) It is not clear yet.9736 ± 0. and from the B semi-leptonic exclusive ¯ decay B → D ν . One gets |Vud | = 0. K− and hyperon-decays.02 |Vcb | and |Vcb| = 0.9736 − 0. To complete this Section we give the experimental values for some of the matrix elements of V .005 V = ⎝ 0.148) A more recent comprehensive analysis gives the following 90 % C. CDF (1996) has measured |Vtb | |Vtb | = 0.149) The elements Vcd and Vcs are determined from charm production in deep inelastic scattering νµ + N → µ + c + X.9993 257 .18 (10.224 ± 0.0010 and |Vus | = 0. range for the various elements of the CKM matrix ⎛ ⎞ 0.154) (10.046 0.

We start with the gauge sector. six mass parameters for the quarks (assuming the color symmetry) and four mixing angles.0359895(61) (10. The other parameters as (mt . with MZ = (91. In fact. sin2 θ) to (g. The choice was to use the fine structure constant α= and the Fermi constant GF = 1. in this type of experiments can be safely neglected due to the relatively large experimental errors. most of the experimental research of the seventies and eighties was centered about the determination of sin2 θ. in order to relate the set of parameters (α. The situation has changed a lot after the beginning of running of LEP1. Before LEP1.158) Also the great accuracy of the experimental measurements requires to take into account the radiative corrections. in this indirect way.156) as measured from the β-decay of the muon. δMZ /MZ ≈ 2 × 10−5 . mh ) just because they affect the radiative corrections. MZ ). GF . However. that is that h part of the interaction between fermions and Higgs field giving rise to the quark and lepton masses. the Higgs sector is specified by the parameters µ and the self-coupling λ. the radiative corrections are particularly sensitive to the top mass. v) we can use the tree level relations. which.10. As a consequence one assumes that the mass matrix for the fermions is known. This was not the case till two years ago.166389(22) × 10−5 GeV −2 (10. 258 . Therefore. before the top discovery.0021) GeV (10. v). In particular.1867 ± 0. before the top discovery.157) 1 137. GF . and its mass was unknown. the better choice was to use quantities known with great precision related to the parameters (g. However we notice that one gets informations about parameters as (mt . g . We will discuss this point in more detail at the end of these lectures. Then we have the Yukawian sector. In this case the most convenient set is (α. We will rather use v = −µ2 /λ and m2 = −2µ2 . g . the question arises about the better choice of the input parameters.4 The parameters of the SM It is now the moment to count the parameters of the SM. the mass of the Z has been measured with great precision. mh ) affect only radiative corrections. The masses of the vector bosons can be expressed in terms of the previous parameters. If we assume three generations and that the neutrinos are all massless we have three mass parameters for the charged leptons. In order to test the structure of the SM the important parameters are those relative to the gauge and to the Higgs sectors. where we have the two gauge coupling constants g and g . Then. LEP1 was able to determine the top mass. Then. In fact.

However.162) )ij φj − i ∂L ∂φi.10.159) d4 x ∂L ∂L δφi + δφi. This is not always so and the motivations can be easily understood. The problem has to do with the invariance of the functional integration measure with respect to the symmetry transformation. that is if the 259 .µ A Since for A independent on x we have δS = 0 by hypothesis. Therefore we are left with δS = where A jµ = −i d4 x ∂ µ A A jµ (10. These identities are nothing but the expression of the symmetry in the quantum context. Let us consider a field theory with a lagrangian invariant under the following global transformation φi → φi + δφi with δφi (x) = −i A (T A )ij φj (x) (10.µ (10.µ = −i A (T A )ij ∂µ φj − i(∂µ Substituting we find δS = − i d4 x − i ∂L ∂φi A (T A A )(T A (10.µ ∂φi ∂φi. we will see that there are situations in which the measure is not invariant and therefore the symmetry is broken at the quantum level. if we take A (x) to describe a variation around the classical solutions.164) ∂L (T A )ij φj ∂φi.µ )ij φj (10.5 Ward identities and anomalies Up to now we have assumed that the symmetries for the classical lagrangian are still valid at the quantum level.µ A (T A )ij ∂µ φj (10.160) Then.165) are the Noether’s currents after factorizing the infinitesimal parameters A . let us consider the transformation of the action under the same transformation of the fields but with A → A (x) a space-time function. In fact. Before dealing with this problem it will be convenient to rephrase the Noether’s theorem in a way more suitable to derive the Ward identities in a quantum field theory. it follows that the first two terms must vanish.161) (10.163) ∂L (T A )ij φj ∂µ ∂φi. We find δS = where δφi.

174) This derivation is correct only if the integration measure is invariant under the change of variables (10.168) for any A (x). Then. writing δS = − we get µ ∂µ jA = 0 d4 x µ A ∂µ jA (10.170) By expanding this expression at the first order in A (x) and using the fact that the previous expression must agree with eq.172) By differentiating once (remember that one gets the T ∗ product) we have µ A ∂x 0|T (jµ (x)φi (y))|0 = −δ 4 (x − y) 0|(T A)ij φj (y)|0 (10.171) From this equation we can generate the Ward identities by differentiating with respect to ηi and then putting ηi = 0. then we must have δS = 0. We will illustrate a simple situation where the measure 260 .166) (10. For instance.168) If we perform the change of variables φi → φi + δφi = φi − i A (x)(T A )ij φj and assume that the integration measure is invariant. differentiating N times µ A ∂x 0|T (jµ (x)φi1 (x1 ) · · · φiN (xN ))|0 N = p=1 δ 4 (x − xp ) 0|T (φi1 (x1 ) · · · (−(T A )ip j φj (xp )) · · · φiN (xN ))|0 (10. for ηi = 0 we get A ∂ µ 0|jµ (x)|0 = 0 (10.173) and. we get iS[φ] + i D(φ)e d4 x ηi φi i e d4 x( ∂ µ A A jµ (10.169) + Z[η] = A ηi (T A )ij φj ) (10.169). we get iS[φ] + i D(φ)e d4 x ηi φi A −i∂ µ jµ + ηi (T A )ij φj 0= (10.167) Let us now consider the generating functional associated to this theory iS[φ] + i D(φ)e d4 x ηi φi Z[η] = (10.fields inside the currents satisfy the equations of motion. (10.

If the gauge field Aµ is quantized we have to insert a further functional integration in Aµ but this will not play any role in the following. To this end we proceed as before by performing the change of variables inside Z ψ(x) → ψ (x) = (1 + iα(x)γ5 )ψ(x) ¯ ¯ ¯ ψ(x) → ψ (x) = ψ(x)(1 + iα(x)γ5 ) The variation of the action is given by δS = ¯ d4 x α(x)∂µ (ψγ µ γ5 ψ) (10.177) We will show that the quantum corrections destroy the conservation law.176) As we will be interested only in checking the current conservation we can put to zero the external sources for the fermion fields (see the previous derivation).178) (10. so for simplicity we will take Aµ as an external field. In fact for an external gauge field we define the generating functional as ¯ ˆ i d4 x ψiDψ ¯ Z = DψD ψe (10.182) Notice that they become negative definite after being Wick rotated 2 2 λ2 → −k4 − |k|2 = −kE m (10.183) 261 . In the case of a zero gauge field Aµ the eigenvalues λm are given by 2 λ2 = k0 − |k|2 m (10. For the following considerations the gauge field can be taken as an external field. The lagrangian is invariant under the global transformation ψ → eiαγ5 ψ giving rise to the classically conserved current ¯ Jµ = ψγµ γ5 ψ (10.179) In order to evaluate the change of the integration measure it is convenient to expand ˆ the fermion field in a basis of eigenvectors of D ˆ iDφm (x) = λm φm (x) ˜ ˜ −iDµ φm (x)γ µ = λm φm (x) (10. but the same applies also to the quantized case.181) where we have used a discrete notation for the eigenvalues.is not invariant.175) where Dµ = ∂µ + igAµ (10.180) (10. Consider a massless fermion in a gauge field.

from the orthogonality relation d4 x φ† φn = δmn m we get am = with Cmn = d4 x φ† (x)iα(x)γ5 φn (x) m (10.By hypothesis the eigenvectors φm span an orthonormal basis and therefore we can ¯ expand ψ and ψ as ψ(x) = m am φm (x). Since we are interested in the infinitesimal transformation.192) The coefficient of α(x) is the trace of γ5 taken over all the Hilbert space. Then. except for a possible constant we have ¯ DψD ψ = dam dbm (10.184) where am and bm are Grassmann coefficients. we have det|I| = eT r log(1 + C) ≈ eT r C that is log det|I| ≈ T rC = i d4 x α(x) n (10.186) Then.187) d4 x φ† (x)iα(x)γ5 φn (x)an = am + m n Cmn an (10. from the transformation rules for the Grassmann measure we get ¯ Dψ D ψ = 1 ¯ DψD ψ det|I|2 (10. Therefore.191) φ† (x)γ5 φn (x) n (10. Therefore this expression is ill defined and it 262 . ¯ ψ(x) = m ˜ bm φm (x) (10.189) d4 x φ† (x)ψ (x) = am + m n (10. Although the trace of γ5 is zero on the spinor space it gets multiplied by an infinite factor coming from the total Hilbert space.185) m We can evaluate the effect of the change of variables on the Grassmann coefficients by starting from ψ (x) = (1 + iα(x)γ5 )ψ(x) = m am φm (x) (10.188) A similar transformation holds for bm .190) where I = 1 + C is the jacobian of the transformation.

D ν ]− = D 2 + σµν F µν 2 2 (10. A natural choice for the regularization is the following φ† (x)γ5 φn (x) = lim n n M →∞ 2 2 φ† (x)γ5 φn (x)eλn /M n n (10.194) ˆ where tr is the trace over the Dirac matrices.200) (10. and therefore we should regulate the expression in a gauge invariant way.196) The matrix element can be easily evaluated by inserting momentum eigenstates and performing a Wick’s rotation 2 x|e− /M |x = lim x→y d4 k k 2 /M 2 −ik(x − y) e e =i (2π)4 d4 kE −k 2 /M 2 M4 e E =i (2π)4 16π 2 (10.197) (10. The evaluation of D)2 gives 1 1 ˆ (D)2 = γµ γν D µ D ν = [γµ . γν ]+ D µ D ν + [γµ . Therefore we get φ† (x)γ5 φn (x) = lim tr γ5 n n M →∞ g 1 − σµν F µν 2! 2M 2 2 2 x|e− /M |x (10.198) Therefore φ† γ 5 φn = n n ig 2 tr[γ5 σµν σρλ ]F µν F ρλ 8 · 16π 2 In our definitions γ5 = iγ 0 γ 1 γ 2 γ 3 The trace is easily evaluated by calculating first tr[γ5 σ01 σ23 ] = −itr[γ 0 γ 1 γ 2 γ 3 γ0 γ1 γ2 γ3 ] = −4i 263 (10. Furthermore. we would like to maintain the gauge invariance.193) This is in fact a convergence factor in the euclidean space. γν ]− D µ D ν 2 2 i g = D 2 − σµν [D µ .needs to be regularized.195) In order to get a contribution from the trace over the Dirac indices we need at least four γ matrices.199) . The previous expression can be rewritten as φ† (x)γ5 φn (x) = n n M →∞ lim 2 ˆ 2 φ† (x)γ5 e(iD) /M φn (x) n n = M →∞ 2 ˆ 2 lim x|tr γ5 e(iD) /M |x (10. Therefore the leading term is obtained by expanding the exponential up to the order (σµν F µν )2 and neglecting Aµ in all the other places.

One has to check that the factor in eq. It follows that the standard model is consistent only if the anomaly vanishes.206) where the term T A comes in the chiral transformation in conjunction with γ5 .205) This is called the Adler-Bell-Jackiw (ABJ) anomaly.206) vanishes for all the generators of SU(2) ⊗ U(1). in fact Qelectron + Qneutrino + 3(Qup + Qdown ) = −1 + 0 + 3 2 1 − 3 3 =0 (10.and using Then 0123 = 1 we get tr[γ5 σµν σρλ ] = 4i φ† γ 5 φn = − n n µνρλ (10. whereas the other two matrices arise from the regulator.202) Therefore the determinant we are looking for is given by −i det |I| = e g2 d x α(x) 32π 2 4 µνρλ F µν F ρλ (10. T C ]+ ] (10. It is not difficult to show that this is indeed the case only if the sum of the electric charges of the fermions is zero Q=0 fermions (10.208) 264 .207) This condition holds for any generation. The ABJ anomaly is modified by a factor T r[T A [T B . (10. for instance. The previous result can be extended to the case of non-abelian chiral symmetries as.201) µν g2 32π 2 µνρλ F F ρλ (10. in the standard model where the SU(2) symmetry acts only on the left-handed fermions and therefore the corresponding transformations involve γ5 .203) and after the change of variables the generating functional is given by i ¯ DψD ψe ¯ ˆ d4 x ψiDψ i e g2 ¯ d4 x α(x)(∂µ (ψγ µ γ5 ψ) + 16π 2 µνρλ F µν F ρλ ) (10. It is a crucial point in the renormalization of the gauge theories that the gauge symmetry is preserved at the quantum level.204) Z= leading to the anomalous conservation law µ ∂µ j5 = − g2 16π 2 µνρλ F µν F ρλ (10.

if λ is an eigenvalue.1) e’. let us notice that iA is a hermitian matrix and therefore it can be diagonalized through a unitary transformation U: U(iA)U † = Ad where Ad is diagonal and real.3) (A.61) by means of an orthogonal transformation. lo written in the form ⎡ λ1 0 0 0 · ⎢ 0 −λ1 0 0 · ⎢ ⎢ 0 0 · 0 λ2 ⎢ ⎢ 0 0 0 −λ2 · Ad = ⎢ ⎢ · · · · · ⎢ ⎣ · · · · · · · · · · (A.Each submatrix of the type λi 0 0 −λi 265 (A. To this end.5) . It follows that Ad can be · · · · · · · ⎤ · ·⎥ ⎥ ·⎥ ⎥ ·⎥ ⎥ ·⎥ ⎥ ·⎦ · (A. Ad has necessarily a zero eigenvalue.1 Properties of the real antisymmetric matrices Given a real and antisymmetric matrix n × n A. We will assume also to order the matrix in such a way that all the λi are positive. we want to show that it is possible to put it in the form (7.Appendix A A. The eigenvalues of iA satisfy the equation det|iA − λ · 1| = 0 Since AT = −A it follows det|iA − λ · 1| = det|(iA − λ · 1)T | = det| − iA − λ · 1| = 0 Therefore. the same is for −λ.4) Notice that if n is odd.2) (A. that we will write as (Ad )nn .

10) with As defined in (7.9) iλ1 0 0 0 · · · 0 0 0 −iλ1 · · · 0 0 iλ2 0 0 · · · (A.12) But A and As are real matrices. 266 .can be put in the form 0 −iλi through the unitary 2 × 2 matrix 1 i 1 V2 = √ 2 1 i In fact V2 Defining λi 0 0 0 V2† = −iλi −λi ⎡ 0 V2 · · · · · · · · · · · · · iλi 0 (A. we get ⎡ 0 ⎢ −iλ1 ⎢ ⎢ 0 ⎢ V Ad V † = ⎢ 0 ⎢ ⎢ · ⎢ ⎣ · · ⎤ · ·⎥ ⎥ ·⎥ ⎥ ·⎦ · · · · · · · · · · · · · · · ⎤ · ·⎥ ⎥ ·⎥ ⎥ · ⎥ = iAs ⎥ ·⎥ ⎥ ·⎦ · (A.11) (A. Therefore (V U)(iA)(V U)† = iAs or (V U)(A)(V U)† = As (A. However being V U real and unitary it must be orthogonal.8) (A. and therefore also V U must be real.7) iλi 0 (A.61).6) V2 ⎢ 0 ⎢ V =⎢ · ⎢ ⎣ · · with Vnn = 0 for n odd.

Sign up to vote on this title
UsefulNot useful