Scribd
Upload
3 views

finalSolutions

The document is a final exam for the Quantum Field Theory course at Georgia Tech, covering various topics including Yukawa potential, QED Feynman rules, and vertex corrections. It includes problems related to calculating potential energy, deriving propagators, and drawing Feynman diagrams. Solutions are provided for each problem, illustrating the application of theoretical concepts in quantum field theory.

Uploaded by

Abel Shiferaw
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

finalSolutions

The document is a final exam for the Quantum Field Theory course at Georgia Tech, covering various topics including Yukawa potential, QED Feynman rules, and vertex corrections. It includes problems related to calculating potential energy, deriving propagators, and drawing Feynman diagrams. Solutions are provided for each problem, illustrating the application of theoretical concepts in quantum field theory.

Uploaded by

Abel Shiferaw
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Georgia Tech PHYS 7147

Quantum Field Theory


Instructor: Predrag Cvitanović
Fall 2018

Final Exam 2:40pm - 5:30pm, Dec 10, 2018


(with solutions, Dec 11, 2018)

Advice: read through the questions, and then pick and chose which points
to do first.

== you have 2 hours and 50 min to complete the exam in the classroom
== closed book, closed lecture notes, no electronics
== remember to put your name on the top, and number the pages

Yukawa potential

Consider two static point sources J ( x ) = δ3 (x − x1 ) + δ3 (x − x2 ) interacting


through a meson field
1
Z Z
W[ J] = − d4 xd4 yJ ( x ) D ( x − y) J (y) (1)
2
d4 k eik( x−y)
Z
D ( x − y) = . (2)
(2π )4 k2 − m2 + ie
In the path integral formalism

h0|e−iHT |0i = e−iET = e−iW [ J ] . (3)


Compute the potential energy E between the sources as a function of their sep-
aration r = |~x1 − ~x2 | . 6 points
Solution: Keep the interaction terms J ( x ) J (y) and dropping self energy
terms J ( x )2 , J (y)2 . The factor of two arises from the two identical interaction
terms.
dy0 ik0 ( x−y)0 d3 x ik(~x−~xi ) d3 y ik(~y−~yi )
ZZ Z Z Z
− dk0 dx0 e e e
2π (2π )3 (2π )3
~
d3 k eik·(~x−~y)
Z
×
(2π )3 k2 − m2 + ie
The rest is worked out in A. Zee Chapter I.4 From Field to Particle to Force. A
student points out that it is exp(+iW [ J ]), not exp(−iW [ J ]) as stated above. Zee
seems to have both signs on his p. 26. Scary.

1
gµν gymnastics

The generators of SO(d) rotations of spin 1 vectors, spin 1/2 spinors are
respectively
ρ ρ i µ ν
( J ρσ )αν = δα δνσ − δν δασ , Sµν = [γ , γ ] . (4)
4
Verify that γµ is an invariant tensor under SO(d) rotations,
[γµ , Sρσ ] = i ( J ρσ ) µ ν γν (5)
5 points
Solution:
i µ ρ σ
[γµ , Sρσ ] = [γ , γ γ − γσ γρ ]
4
i
= ({γµ , γρ }γσ − γρ {γµ , γσ } − {γµ , γσ }γρ + γσ {γµ , γρ })
4
i
= − (γσ δµρ − γρ δµσ − γρ δµσ + γσ δµρ )
2
ρ
= −i (γσ δµρ − γρ δµσ ) = −i (δνσ δµρ − δν δµσ )γν = i ( J ρσ ) µ ν γν .

QED Feynman rules

Setting h̄ = c = 1, the QED Lagrangian density is


1
L = ψ(iγµ Dµ − m)ψ − Fµν F µν , Dµ = ∂µ − ieAµ . (6)
4
(a) Derive the momentum space electron propagator, using

d4 x
Z
g( x ) e−ix· p = g( p) . (7)
(2π )4
4 points
Solution: Electron propagator satisfies

(i∂µ − m) D ( x ) = δ(4) ( x ) ,

so
1 p+m
( p − m) D ( p) = 1 ⇒ D ( p) = = 2 2
p
 − m  −m
p

To get the causality right, one also needs a small imaginary term in the
denominator:

iD ( p) = i p + m
p 2 − m2 + ie


2
(b) Derive the momentum space photon propagator. 6 points
Solution: The Maxwell Lagrangian in momentum space is given by

1  
L(k) = − Aµ (k) k2 gµν − kµ kν Aν (−k)
2

The matrix k2 gµν − kµ kν has eigenvector kν with a zero eigenvalue, so it


is not invertible. Now argue that Aµ component along k µ is not physical,
and by gauge invariance anything of the form
 
i kµ kν
iD µν (k) = − gµν − (1 − ξ ) . (8)
k2 − ie k2

will serve as a photon propagator (in a covariant gauge).

(c) Derive the electron/positron-photon 3-vertex. 2 points


Solution: In the Minkowski space, the coupling of spinor-in, photon, and
spinor-out is local, with coupling strength and the invariant coupling of
spinor and Minkowski indices:

ieγµ

The Fourier transform yields


Z
d4 xe−ik1 · x e−ik2 · x eik· x = (2π )4 δ4 (k − k1 − k2 ) (9)

so the 3 vertex in momentum space conserves the momentum:

i eγµ δ4 (k − k1 − k2 ) .

QED vertex corrections

(a) Draw ALL 1-loop corrections to electron-photon vertex 2 points


Solution: There are two electron self-energy graphs and one photon po-
larization graph in the addition to the 1PI vertex graph (Zee figure)

3
(b) Draw all 2-loop 1-particle-irreducible corrections to electron-photon ver-
tex 3 points
Solution: There is one photon polarization graph (electron loop inserted
into the internal photon propagator) + six 2-photon graphs:

One can draw 3-photon electron loops, but they vanish by Furry’s theo-
rem. Nobody got this right.
(c) Draw all tree (no loops) electron-electron scattering diagrams 2 points
Solution: Only two graphs - both with a photon exchange, but the out-
going electrons exchanged in the second graph.
(d) Write down the Feynman integral for the 1PI 1-loop correction to electron-
photon vertex. 4 points
Solution:
d4 k − i
 
i i
Z
ν µ
ie γ ie γ ieγ ν
(2π )4 k2 0
p + k − m p + k − m

(e) Simplify γν p 0 γµ pγν . 2 points


Solution. We worked this one in HW09:

γ α γ λ γ µ γ ν γα = (2gαλ − γλ γα )γµ γν γα
= 2γµ γν γλ − γλ γα γµ γν γα = 2γµ γν γλ − 4γλ gµν
= 2γµ γν γλ − 2γν γµ γλ − 2γµ γν γλ
= −2γν γµ γλ , (10)

where we have used

γ α γ µ γ ν γα = (2gαµ − γµ γα )γν γα = 2γν γµ − γµ γα γν γα


= 2γν γµ + 2γµ γν = 4gµν .
Contracting with p0λ from the left, pν from the right, we have

γα p 0 γµ pγα = −2 pγµ p 0 (11)

4
(f) What is the value of this contribution on the mass-shell, sandwiched be-
tween ū( p0 ) and u( p)? 1 points
Solution: We will need γα γµ γα = −2γµ ,

q2 = ( p0 − p)2 = 2[m2 − ( p0 · p)] , (12)

where q is the external photon 4-momentum, and

( p0 + p)µ = 2mγµ − iσµν qµ , (13)

is the Gordon identity.


In order to use Dirac equation pu( p) = mu( p), we need to move p 0 in
(11) to the left of p, so we redo (10)

γ α γ λ γ µ γ ν γα = (2gαλ − γλ γα )γµ (2gα ν − γα γν )


= γλ γα γµ γα γν + 4γµ gλν − 2γµ γλ γν − 2γλ γν γµ
= −2γλ γµ γν + 4γµ gλν
−2(2gλµ − γλ γµ )γν − 2γλ (2gµν − γµ γν )
= 2γλ γµ γν + 4gλν γµ − 4gλµ γν − 4gµν γλ , (14)

Contracting with p0λ from the left, pν from the right, we have

γα p 0 γµ pγα = 2 p 0 γµ p + 4( p0 · p)γµ − 4p0µ p − 4pµ p 0 (15)

Sandwiched between ū( p0 ) · · · u( p) and using (12) and (13) we get

γα p 0 γµ pγα = [2m2 + 4( p0 · p)]γµ − 4m( p0 + p)µ


= (2m2 + 4m2 − 2q2 )γµ − 4m(2mγµ − iσµν qν )
= −2(m2 + q2 )γµ + 4im σµν qν . (16)

Clearly (e) and (f) deserve more points than the measly 2+1.

You might also like