Solutions to Peskin and Schroder – Andrzej Pokraka

Problem 7.1
In Section 7.3 we used an indirect method to analyze the one-loop s-channel diagram for boson-boson
scattering in φ4 theory. To verify our indirect analysis, evaluate all three one-loop diagrams, using
the standard method of Feynman parameters. Check the validity of the optical theorem.

The total scattering amplitude is given by

iM = iM0 + iM1 (s) + iM1 (u) + iM1 (t) + iM2 + . . . (1)

where p1 p2

k k
iM1 (s) = 2 +q 2 −q k = k1 + k2 (2)

k1 k2
and the t- and u-channel diagrams can be obtained by suitable replacements of s = (k1 + k2 )2 .

dd q
Z
4−d i i
iM1 (s) = (−iλµ 2 )2
(2π) d (k/2 − q)2 − m2φ + i (k/2 + q)2 − m2φ + i
Z 1
dd q
Z
1
= λ2 µ4−d d
dx h i2
(2π) 0
k 2 /4 + (2x − 1)k · q + q 2 − m2φ + i
Z 1
dd `
Z
2 4−d 1
=λ µ dx 2 (3)
(2π)d 0 2
[` − ∆ + i]

where ` = q + (2x − 1)k/2 and ∆ = −x(1 − x)k 2 + m2φ . Preforming the momentum integration we
obtain
Z 1 
4πµ2

i
iM1 (s) = λ2 dx Γ()
0 16π 2 ∆
Z 1
λ2 i
  2 
1 µ̃
= dx + log . (4)
16π 2 0  ∆

Recalling that s = k 2 = (k1 + k2 )2 we have

!!
1
λ2 i µ̃2
Z
1
iM1 (s) = dx + log 2 . (5)
16π 2 0  mφ − x(1 − x)s

Using Mathematica to preform the x integral and dividing by i we obtain

 √ 
 s !
λ2  1 4m2φ s µ̃2 
M1 (s) = + 2 − 2 − 1 ArcSin + log . (6)
16π 2  s 2mφ m2φ

1
 Solutions to Peskin and Schroder – Andrzej Pokraka

For s < 4m2φ the square roots are real and M1 (s) is real. However, for s > 4m2φ the square roots
develop branch cuts =⇒ that M1 (s) has a branch cut from threshold to infinity, s ∈ {4m2φ , ∞}.
To find the imaginary part it is easiest to use the identity Im log(−x ± i) = ±π (for x > 0) in
equation (5) and preform then x integration. M1 acquires an imaginary part when the argument
of the logarithm becomes negative q (branch cut from 0 to infinity). For any s this happens for
1
2 − 12 β < x < 1
2 + 12 β where β = 1 − 4m2φ /s. With these remarks we have

1
−λ2
Z
dx log m2φ − x(1 − x)s ± i


ImM1 (s ± i) =
16π 2 0
1 1
−λ 2 2+2β
Z
= dx (±π)
16π 2 1
2−2β
1

1
∓λ2 2β
Z
= dy
16π − 12 β
s
∓λ2 4m2φ
= 1− . (7)
16π s

Working in the centre of mass frame s > 0, t = 0 and u < 0. This implies that only the s-channel
diagrams contribute to the imaginary part of M1 (i.e., ImM1 = ImM1 (s)).
!!
λ2 1 µ̃2
M1 (t) = + 2 + log ,
16π 2  m2φ
 s p ! !
2 4m 2 2
λ 1 φ |u| µ̃
M1 (u) = +2+2 + 1 ArcSinh + log . (8)
16π 2  |u| 2mφ m2φ

To validate the optical theorem we need to relate the imaginary part to the amplitude of the
tree-level scattering squared. The tree-level scattering amplitude squared, |M0 |2 , is just λ2 .

2
 Solutions to Peskin and Schroder – Andrzej Pokraka

Problem 7.2: Alternative regulators in QED

In Section 7.5, we saw that the Ward identity can be violated by an improperly chosen regulator.
Let us check the validity of the identity Z1 = Z2 , to order α, for several choices of the regulator.
We have already verified that the relation holds for Pauli-Villars regularization.
(a) Recompute δZ1 and δZ2 , defining the integrals (6.49) and (6.50) by simply placing an upper
limit Λ on the integration over `E . Show that, with this definition, δZ1 6= δZ2 .

(b) Recompute δZ1 and δZ2 , defining the integrals (6.49) and (6.50) by dimensional regularization.
You may take the Dirac matrices to be 4 x 4 as usual, but note that, in d-dimensions,

g µν γµ γν = d.

Show that, with this definition, δZ1 = δZ2 .

Preamble to parts (a) and (b)

We will take the dimension to be d so that we may use our formulas for both parts (a) and (b).
We will need to know the integral,
∞ d−1+2n
dd ` `2n (−1)m+n i (−1)m+n i
Z Z
`E n
= Ωd d`E = Ωd Im , (9)
(2π)4 (`2 − ∆ + i)m (2π)4 0 (`2E + ∆)m (2π)4

in order to evaluate δZ1 and δZ2 . In the text the integral, Im , is given by equations (6.49) and
(6.50), and, regulated by Pauli-V regularization. We are asked to evaluate these integrals by (a)
placing an upper bound Λ on the momentum `E and (b) using dimensional regularization.
The renormalization factor Z1 is defined by the relation Γµ (p + k, p)|k→0 = Z1−1 γ µ where Γµ is
the electron vertex. The one-loop correction to the electron vertex is given by
p0
k0 = k + q
p−k µ = ū(p0 )δΓµ (p0 , p)u(p). (10)
q
k
p

From Peskin and Schroder we know that this correction is given by equation (6.38)

dd k −igνρ i(k
0 + m)
Z
i(k
+ m)
ū(p0 )δΓµ (p0 , p)u(p) = 4 2
ū(p0
)(−ieγ ν
) 2 γµ 2 (−ieγ ρ )u(p)
(2π) (k − p) + i k − m + i k − m2 + i
0 2

dd k ū(p0 )N u(p)
Z
= 2ie2 , (11)
(2π)4 ((k − p)2 − µ2 + i)(k 0 2 − m2 + i)(k 2 − m2 + i)

where N = − 21 γ ν [k
γ µ k
+ m2 γ µ + m(k
0 γ µ + γ µ k
)]γν . Evaluating the above correction at p0 = p =⇒

k 0 = k will yield the correction to the renormalization constant Z1

dd k
Z
ū(p)N u(p)
ū(p)δΓµ (p, p)u(p) = 2ie2 , (12)
(2π)4 ((k − p)2 − µ2 + i)(k 2 − m2 + i)2

3
 Solutions to Peskin and Schroder – Andrzej Pokraka

where N = (1 − 4−d µ d−2 2 µ µ

γ k
+ 2 m γ − dmk . Combining the propagators via use of Feynman
2 )k
parameters and simplifying the numerator using the Dirac equation, we obtain

dd k ū(p)[(1 − 4−d
2 )k
µ d−2 2 µ µ

γ k
+ 2 m γ − dmk ]u(p)
Z Z Z
µ 2
ū(p)δΓ (p, p)u(p) = 2ie dx dy yδ(x + y − 1)
(2π)4 (k 2 − 2xk · p + xp2 − xµ2 − ym2 + i)3
Z Z
= 2ie2 dx dy yδ(x + y − 1)

dd ` ū(p)[(1 − 4−d
2 )(

` + xp)γ µ

(
` + xp)
d−2 2 µ µ

+ 2 m γ − dm(` + xp) ]u(p)
Z
×
(2π)4 (`2 − ∆ + i)3
Z Z
→ 2ie2 dx dy yδ(x + y − 1)

dd ` (d − 2)2 ū(p)[(1 − 4−d d−2 2 µ

2 )(− d ` γ + x pγ
2 µ

p)
d−2 2 µ µ

+ 2 m γ − dmxp ]u(p)
Z
× 4 2 3
(2π) 2d (` − ∆ + i)
Z Z
= 2ie2 dx dy yδ(x + y − 1)

dd ` ū(p) − d−2 4−d 2 4−d d−2

 2 2

µ 
d (1 − 2 )` + ((1 − 2 )x + 2 )m γ − dmxpµ u(p)
Z
×
(2π)4 (`2 − ∆ + i)3
Z Z
= 2ie2 dx dy yδ(x + y − 1)
h  i
(d−2)2 2 d−2 2 2
Z d
d ` ū(p) − 2d ` + ( 2 (1 + x ) − dx)m γ µ u(p)
× . (13)
(2π)4 (`2 − ∆ + i)3

where ∆ = −x(1 − x)p2 + xµ2 + ym2 = −x(1 − x)p2 + xµ2 + (1 − x)m2 . In the last line of equation
(5) we have used the Gordon identity. Now,

γµ
Γµ (p, p) = γ µ + δΓµ (p, p) ≡ =⇒ δZ1 = −δΓµ (p, p) (14)
1 + δZ1
Thus, the one-loop correction to the renormalization constant Z1 is
(d−2)2 2
dd ` − 2d ` + (( d−2 2
2 (1 + x ) − dx)m
2
Z Z
2
δZ1 = −2ie dx (1 − x) (15)
(2π)4 (`2 − ∆ + i)3

Note that up to this part we have not imposed any regulators on the integral and worked in d-
dimensions. This is the starting point for the evaluation of δZ1 for both parts (a) and (b).
Now, let us set up the calculation of the correction to Z2 . This renormalization constant is
defined as
1 dΣ
≡1− |p→0 . (16)
Z2 dp



The first order contribution to the electron self-energy is given by the diagram
p−k

=
− m0 ) [−iΣ2 (p)] i(p

i(p
− m0 ) . (17)
p k p p2 − m20 p2 − m20

4
 Solutions to Peskin and Schroder – Andrzej Pokraka

where
dd k µ i(k
+ m0 ) −i
Z
−iΣ2 (p) = (−ie)2 d
γ 2 2 γµ
(2π) k − m0 + i (p − k) − µ2 + i
2
Z 1 Z
dd ` −(d − 2)xp

+ dm0
= −e2 dx d 2 − ∆ + i)2
(18)
0 (2π) (`

where ` = k − xp and ∆ = −x(1 − x)p2 + xµ2 + (1 − x)m20 . To find δZ2 we must evaluate Σ2 by
regulating the integral, differentiate wrt p
→m

and take the limit p
dΣ2
δZ2 = (Z2 − 1) = |p→m . (19)
dp

Part (a)
First we set d = 4 then regulate the integral of equation (1) by placing an upper bound Λ on the
momentum, `E :
Λ Λ2
`3+2n 1 2 (`2E )n+1
Z Z
n
Im → d`E 2 E = d(` ) . (20)
0 (`E + ∆)m 0 2 E (`2E + ∆)m

In particular, we will need:

Λ
(`2 )
Z
1 2
I30 = d(`E ) 2 E 3
0 2 (`E + ∆)
1 Λ4
=
4 ∆(∆ + Λ2 )2
  
1 1 ∆
= +O , (21)
2∆ 2 Λ2

and
Λ2
d(`2E ) (`2E )2
Z
I31 =
0 2 (`2E + ∆)3
∆ + Λ2 3∆2 + 4∆Λ2
   
1 3
= log + −
2 ∆ 2(D + Λ2 )2 2
  2  
1 Λ 3 ∆
= log − +O (22)
2 ∆ 2 Λ2

for the evaluation of δZ1 , and,

Λ2
`2
Z
1 2
I20 = d(`E ) 2 E 2
0 2 (`E + ∆)
∆ + Λ2 Λ2
 
1
= log −
2 ∆ 2∆(∆ + Λ2 )2
  2  
1 Λ ∆
= log −1+O , (23)
2 ∆ Λ2

for the evaluation of δZ2 .

5
 Solutions to Peskin and Schroder – Andrzej Pokraka

Substituting equation (1) (with equations (11) and (12)) into (6) with d = 4 we obtain δZ1

d4 ` − 21 `2 + (1 − 4x + x2 )m2
Z Z
2
δZ1 = −2ie dx (1 − x)
(2π)4 (`2 − ∆ + i)3
3 Z Z  
2 (−1) i 1 1 2 2 0
= −2ie Ω4 dx dy yδ(x + y − 1) I + (1 − 4x + x )m I3
(2π)4 2 3
Λ2
Z       
α 1 3 2 2 1 ∆
=− dx (1 − x) log − + (1 − 4x + x )m +O
π 4 ∆ 2 4∆ Λ2
Z   2 2

α Λ 3 m
≈− dx (1 − x) log − + (1 − 4x + x2 )
4π ∆ 2 ∆
Z   2 2

α Λ 3 m
=− dx (1 − x) log − + (1 − 4x + x2 ) . (24)
4π ∆ 2 ∆

Setting p2 = m2 we have

Λ2 (1 − 4x + x2 )m2
Z    
α 3
δZ1 = − dx (1 − x) log − +
4π (1 − x)2 m2 + xµ2 2 (1 − x)2 m2 + xµ2
   2   2 
α 1 Λ 3 5 m
=− 1 + log − + − 2 log
4π 2 m2 4 2 µ2
  2  2 
α 5 Λ m
=− + log − 2 log (25)
4π 4 m2 µ2

where
1
Λ2
Z    2  
1 Λ
dx (1 − x) log = 1 + log (26)
0 (1 − x)2 m2 + xµ2
2 m 2
Z 1 2 2
 2
(1 − x)(1 − 4x + x )m 5 m
dx 2 2 2
= − 2 log . (27)
0 (1 − x) m + xµ 2 µ2

1
d4 ` −2xp
Z Z
Σ2 (p) = −ie2 dx
+ 4m0
0 (2π) (` − ∆ + i)2
4 2
1
(−1)2 i
Z
= −ie2 dx (−2xp

+ 4m0 )
4
Ω4 I20
0 (2π)
Z 1
−e2 i ∆ + Λ2 Λ2
   
1
= Ω4 dx (−2xp
+ 4m0 ) 2 log −
(2π)4 0 ∆ 2∆(∆ + Λ2 )2
2 Z 1   2
  
e Λ ∆
= 4
Ω4 dx (−2xp + 4m0 ) log −1+O
2(2π) 0

∆ Λ2
Z 1   2 
α Λ
≈ dx (−2xp + 4m0 ) log −1
4π 0
∆
Z 1
Λ2
   
α
= dx (−2xp + 4m 0 ) log − 1 . (28)
4π 0
−x(1 − x)p2 + xµ2 + (1 − x)m20

6
 Solutions to Peskin and Schroder – Andrzej Pokraka

Now,
dΣ2
δZ2 = |p→m
dp

Z

1
Λ2
   
α
= dx(−2x) log − 1
4π 0 (1 − x)2 m20
Z 1  
α −1
+ dx (2 − x) m0 (−2x(1 − x)m0 )
4π 0 (1 − x)2 m20 + xµ2
Z 1 
Λ2 x(2 − x)(1 − x)m20
  
α
= dx x − x log +
2π 0 (1 − x)2 m20 + xµ2 (1 − x)2 m20 + xµ2
   2    2 
α 1 1 Λ 1 m
= − 3 + log + −1 + 2 log
2π 2 2 2 2 µ2
  2  2 
−α Λ m
= 3 + log − 2 log (29)
4π 2 µ2

where
1
Λ2 Λ2
Z     
1
dx x log |µ→0 = 3 + log . (30)
0 (1 − x) m20 + xµ2
2 2 2

Clearly, δZ1 6= δZ2 as δZ1 − δZ2 = 7α/16. Therefore, the Ward identity is violated.

Part (b)
We are asked to repeat part (a) but regulate the integrals using dimensional regularization:
Z ∞ d−1+2n
dd ` `2n (−1)m+n i
Z
`E
= Ω d d` E
(2π)4 (`2 − ∆ + i)m (2π)d 0 (`2E + ∆)m
m−n−d/2
(−1)n+m i 1
 
Γ(m − d/2 − n)Γ(d/2 + n)
= (31)
(4π)d/2 ∆ Γ(d/2)Γ(m)

where
2π d/2
Ωd = , (32)
Γ(d/2)
Z ∞
n (`E )d−1+2n
Im = d(`E ) 2
0 (`E + ∆)m
 m−n−d/2
1 1 Γ(m − d/2 − n)Γ(d/2 + n)
= . (33)
2 ∆ Γ(m)

7
 Solutions to Peskin and Schroder – Andrzej Pokraka

With equations (20-22), δZ1 becomes

(d−2)2 2
dd ` − + ( d−2 2
2 (1 + x ) − dx)m
2d `
2
Z Z
δZ1 = −2ie2 dx (1 − x)
(2π)4 2
(` − ∆ + i) 3
 2−d/2
−2e2 (d − 2)2 1 Γ(2 − d/2)Γ(d/2 + 1)
Z
= d/2
dx (1 − x)
(4π) Γ(d/2) 2d ∆ Γ(3)
   3−d/2 !
d−2 1 Γ(3 − d/2)Γ(d/2)
+ (1 + x2 ) − dx m2
2 ∆ Γ(3)

−e2 (1 − )2 4π
  
Γ()Γ(3 − )
Z
= 2
dx (1 − x)
16π 2− ∆ Γ(2 − )
  

m2
 
2 4π
+ (1 − ) (1 + x ) − 2 (2 − ) x Γ(1 + )
∆ ∆

(1 − )2 4π
 
−α Γ()Γ(3 − )
Z
= dx (1 − x)
4π 2− ∆ Γ(2 − )
  

m2

−α
Z
2 4π
+ dx (1 − x) (1 − ) (1 + x ) − 2 (2 − ) x Γ(1 + )
4π ∆ ∆


m2
   
−α −α
Z Z
2 4π 2
= dx (1 − x)(1 − ) Γ() + dx (1 − x) 1 − 4x + x
4π ∆ 4π ∆

m2
     
−α −α
Z Z
1 4π 2
= dx (1 − x) − γE + log −2 + dx (1 − x) 1 − 4x + x
4π  ∆ 4π ∆
2
)m2
 
−α −α − −
Z Z
1 (1 x)(1 4x + x
− log ∆ + log 4πe−γE − 2 +


= dx (1 − x) dx .
4π  4π (1 − x)2 m2 + xµ2
(34)

Adding in the mass scale Λ2 (and absorbing the log (4πe−γE )) we get

Λ2 (1 − x)(1 − 4x + x2 )m2
   
−α −α
Z Z
1
δZ1 = dx (1 − x) + log 2 2 2
− 2 + dx
4π  (1 − x) m + xµ 4π (1 − x)2 m2 + xµ2
  2  2 
−α 1 1 1 Λ 5 m
= + + log − 1 + − 2 log
4π 2 2 2 m2 2 µ2
  2  2 
−α 1 1 Λ m
= + 2 + log − 2 log
4π 2 2 m2 µ2
  2  2 
−α 1 1 Λ m
= + 4 + log − 4 log , (35)
4π 2  m2 µ2

where
1
(1 − x){1, x, x2 }m2 m2 m2 m2
Z       
3
dx |µ→0 = log , −1 + log , − + log (36)
0 (1 − x)2 m2 + xµ2 µ2 µ2 2 µ2
1
(1 − x)(1 − 4x + x2 )m2 m2
Z  
5
dx |µ→0 = − 2 log . (37)
0 (1 − x)2 m2 + xµ2 2 µ2

8
 Solutions to Peskin and Schroder – Andrzej Pokraka

We now calculate the self-energy to 1-loop

Z 1 Z
2 dd ` −(d − 2)xp
+ dm0
Σ2 (p) = −ie dx d 2 − ∆ + i)2
0 (2π) (`
Z 1  2−d/2
i 1 Γ(2 − d/2)
= −ie2 dx ((2 − d)xp + dm) d/2
0

(4π) ∆ Γ(2)
Z 1  
α 4π
= dx (−2(1 − )xp
+ 2(2 − )m) ∆ Γ()
4π 0
Z 1    
α 1 −γE
= dx (2m − xp)
 − log ∆ + log(4πe ) − (m − xp)
+ O()
2π 0
Z 1    
α 1 −γE
= dx (2m − xp) − log ∆ + log(4πe ) − (m − x p) + O() (38)
2π 0


Adding mass scale in logarithm to get a dimensionless quantity yields

Z 1 
Λ2
    
α 1 −γE
Σ2 (p) = dx (2m − xp)
 + log (1 − x)2 m2 + xµ2 + log(4πe ) − (m − xp)

+ O()
2π 0
Z 1 
Λ2
   
α 1
→ dx (2m − xp) + log − (m − x p) + O() . (39)
2π 0
 (1 − x)2 m2 + xµ2

Differentiating yields
Z 1 
Λ2 2m2 x(1 − x)(2 − x)
   
α 1
δZ2 = dx −x + log + + x
2π 0  (1 − x)2 m2 + xµ2 (1 − x)2 m2 + xµ2
   2   2 
α 1 1 Λ m 1
= − + 3 + log − 1 + 2 log +
2π 2  m2 µ2 2
  2  2 
−α 1 Λ m
= + 4 + log − 4 log (40)
4π  m2 µ2

where
1
(1 − x){1, x, x2 }m2 m2 m2 m2
Z       
3
dx |µ→0 = log , −1 + log , − + log (41)
0 (1 − x)2 m2 + xµ2 µ2 µ2 2 µ2
1
2x(1 − x)(2 − x)m2 m2
Z  
dx |µ→0 = −1 + 2 log . (42)
0 (1 − x)2 m2 + xµ2 µ2

There is a mistake δZ1 is 1/2 of its true value. δZ1 = δZ2 when regulated via dimensional regular-
ization.

9
 Solutions to Peskin and Schroder – Andrzej Pokraka

Problem 7.3
Consider a theory of elementary fermions that couple both to QED and to a Yukawa field φ:
Z  
λ
Hint = d3 x √ φψ̄ψ + eψ̄ Aψ .
2
(a) Verify that the contribution to Z1 from the vertex diagram with a virtual φ equals the contri-
bution to Z2 from the diagram with a virtual φ. Use dimensional regularization. Is the Ward
identity generally true in this theory?
(b) Now consider the renormalization of the φψ̄ψ vertex. Show that the rescaling of this vertex at
q 2 = 0 is not canceled by the correction to Z2 . (It suffices to compute the ultraviolet-divergent
parts of the diagrams.) In this theory, the vertex and field-strength rescaling give additional
shifts of the observable coupling constant relative to its bare value.

Part (a)
The scalar particle couples to the charged fermion field via the three point vertex

p0
iλ
= −√ . (43)
q 2
p

The renormalization factor Z1 is defined by the relation Γµ (p + q, p)|q→0 = Z1−1 γ µ where Γµ is

the electron vertex. The one-loop scalar correction to the electron vertex is given by
p0
k0 = k + q
p−k µ = ū(p0 )δΓµ (p0 , p)u(p). (44)
q
k
p
Here,

iλ2 µ(4−d) dd k ū(p0 )N µ u(p)

Z
0 µ 0
ū(p )δΓ (p , p)u(p) = 2 (45)
2 (2π)d (k 0 − m2e + i)(k 2 − m2e + i)((p − k)2 − m2φ + i)

where N µ = (k
0 + me )γ µ (k
+ me ). Also note that the mass dimension of λ in d-dimensions is
(4 − d)/2. Thus, to keep λ dimensionless in d-dimensions we rescale λ → µ(4−d)/2 λ (where µ is of
mass dimension 1) for each coupling constant in the loop. We evaluate equation (37) at q → 0 to
obtain the vertex correction. In this limit we have
iλ2 µ(4−d) dd k ū(p)N µ u(p)
Z
ū(p)δΓµ (p, p)u(p) =
2 (2π) (k − me + i)2 ((p − k)2 − m2φ + i)
d 2 2
Z 1 Z 1
iλ2 µ(4−d) dd ` ū(p)N µ u(p)
Z
= dx dy 2yδ(x + y − 1) (46)
2 (2π)d 0 0 (`2 − ∆ + i)3
where ` = k − xp and ∆ = −x(1 − x)p2 + xm2φ + (1 − x)m2e . Use of the Gordon identity and Dirac
equation simplifies the numerator, N µ =
`γ µ
` + xme (
`γ µ + γ µ
`) + (1 + x)2 m2e γ µ + 2me `µ . Averaging

10
 Solutions to Peskin and Schroder – Andrzej Pokraka

over the loop momentum we discard terms linear in `, and write `µ `ν = `2 g µν /d. With this, the
numerator becomes N µ → 2−d 2 µ 2 µ
d ` γ + (1 + me ) γ .
Thus, the vertex correction is
1 1 2−d 2
iλ2 µ4−d dd ` d ` + (1 + x)2 m2e
Z Z Z
δZ1 = − dx dy 2yδ(x + y − 1)
2 (2π)d 0 0 (`2 − ∆ + i)3
1 2−d 2
iλ2 µ4−d + (1 + x)2 m2e
dd ` d `
Z Z
=− dx 2(1 − x)
2 0 (2π)d
− ∆ + i)3 (`2
1   
iλ2 i 4πµ2 Γ() (1 + x)2 m2e 4π
Z   
Γ(1 + )
=− dx 2(1 − x) −(1 − ) +
2 16π 2 0 ∆ 2 ∆ ∆ 2
2 Z 1 2 −γE 2 2
    
λ 1 4πµ e (1 + x) me
=− dx 2(1 − x) + log −1 +
64π 2 0  ∆ ∆
" ! #
1
λ2 µ̃2 (1 + x)2 m2e
Z
1
= dx (1 − x) 1 − − log − (47)
32π 2 0  (1 − x)2 m2e + xm2φ (1 − x)2 m2e + xm2φ

where µ̃2 = 4πµ2 e−γE .

Next, we evaluate the contribution of the scalar to the electron self-energy
k−p

p p =
− me ) [−iΣ(p)] i(p
i(p
− me ) , (48)
k p2 − m2e p2 − m2e

where
iλ2 µ4−d dd k i(k
+ me )
Z
i
Σ(p) = −
2 (2π) k − me + i (k − p) − m2φ + i
d 2 2 2

iλ2 µ4−d 1 dd ` (
` + y p

Z Z
= dx
+ me )
2 (2π) d (`2 − ∆ + i)2
0
iλ2 µ4−d 1 dd `
Z Z
(xp

+ me )
= dx (49)
2 0 (2π) (` − ∆ + i)2
d 2

where ` = k − yp and ∆ = −(1 − x)xp2 + (1 − x)m2e + xm2φ . Evaluating the ` integral yields the
self-energy,

iλ2 µ4−d 1 dd `
Z Z
(xp
+ me )
Σ(p) = dx
2 0 (2π) (` − ∆ + i)2
d 2
 2−d/2
iλ2 µ4−d 1 Γ(2 − d/2)
Z
i 1
= dx (xp
+ me ) (4π)d/2 ∆
2 0 Γ(2)
2 Z 1 2 
 
λ 4πµ
=− 2
dx (xp + me ) Γ()
32π 0
∆
Z 1
λ2
  2 
1 µ̃
=− dx (x p + m e ) + log + O() . (50)
32π 2 0
 ∆

11
 Solutions to Peskin and Schroder – Andrzej Pokraka

→ me ,
To get the electron vertex correction we differentiate with respect to p and take the limit p
dΣ
δZ2 = |p→me
dp


Z
1
λ2
   2  
1 µ̃ 1 d∆
=− dx x + log − (xp
+ me ) ∆ dp
32π 2 0  ∆
" !!
p→me #
1
λ2 µ̃2 2(1 − x2 )xm2e
Z
1
=− dx x + log + . (51)
32π 2 0  (1 − x)2 m2e + xm2φ (1 − x)2 m2e + xm2φ

To test if the Ward identity holds we take the difference between δZ1 and δZ2
Z 1
λ2
 2
2x(1 − x)(1 + x)m2e

x µ̃
δZ2 − δZ1 = dx − − log −
32π 2 0  ∆ ∆
 2
(1 − x)(1 + x)2 m2e

1−x µ̃
−(1 − x) + + (1 − x) log +
 ∆ ∆
2 Z 1
 2
(1 − x)(1 + x)2 m2e
 
λ 1 − 2x µ̃
= dx + (1 − 2x) log − (1 − x) +
32π 2 0  ∆ ∆
2 Z 1
 2
(1 − x)(1 + x)2 m2e
 
λ 1 − 2x µ̃
= dx + (1 − 2x) log − (1 − x) +
32π 2 0  ∆ ∆
2 Z 1 2 2 2
   
λ µ̃ (1 − x)(1 + x) me
= dx (1 − 2x) log − (1 − x) +
32π 2 0 ∆ ∆
= 0, (52)

where
1 1
µ̃2
   
−1 d∆
Z Z
(1 − 2x) log =− x(1 − x)
0 ∆ 0 ∆ dx
(1 − x)(1 + x)2 m2e
= (1 − x) − . (53)
∆

Part (b)
At the one-loop level the φψ̄ψ vertex renormalization constant has contributions from the diagrams
p0 p0
k0 = k + q k0 = k + q
λ
iM = p−k + p−k = −i √ ū(p0 )δVφψ̄ψ (p0 , p)u(p), (54)
q q 2
k k
p p
where

ū(p0 )δVφψ̄ψ (p0 , p)u(p) =

dd k i(k
0 + me )

−i
Z
0 (4−d)/2 µ i(k
+ me )
d
ū(p ) (−ieµ γ ) 2 2 2
(−ieµ(4−d)/2 γµ )
(2π) k − me + i k − me + i
0 2 (p − k) − m2γ + i
2
)
λµ(4−d)/2 i(k
0 + me ) λµ(4−d)/2
   
i(k
+ me ) i
+ −i √ −i √ u(p). (55)
2 k 0 2 − m2e + i k 2 − m2e + i 2 (p − k)2 − m2φ + i

12
 Solutions to Peskin and Schroder – Andrzej Pokraka

Setting q = 0 we obtain

ū(p)δVφψ̄ψ (p, p)u(p) =

Z 1
dd `
Z   2 4−d  
2 4−d

N1 λ µ N2
dx 2(1 − x) ū(p) −ie µ + i u(p),
0 (2π)d (` − ∆1 + i)3 2 (` − ∆2 + i)3
(56)
where
` = k − yp,
N1 = d`2 + (1 + x2 )d − 2x(d − 2) m2e ,

N2 = `2 + (1 + x)2 m2e ,
∆1 = −x(1 − x)p2 + (1 − x)m2e + xm2γ ,
∆2 = −x(1 − x)p2 + (1 − x)m2e + xm2φ . (57)
Preforming the d-dimensional momentum integral we obtain
Z 1 2−d/2
iµ4−d
d2
  
1 Γ(2 − d/2) 2 2

2 (2 − d/2)
δZ1 = − dx 2(1 − x) −ie − (1 + x )d − 2x(d − 2) me
0 (4π)d/2 ∆1 Γ(3) 2 ∆1
Z 1 2−d/2
iµ4−d Γ(2 − d/2) λ2
     
1 d (2 − d/2)
− dx 2(1 − x) d/2
i − (1 + x)2 m2e
0 (4π) ∆ 2 Γ(3) 2 2 ∆1
Z 1 
!
2 2 2(2 − )(1 + x ) − 4(1 − )x m2e
2

2

i 4πµ Γ() 4(2 − )
= − −ie dx 2(1 − x) −
16π 2 0 ∆1 2 2 ∆1
Z 1 2 
 2    2 2

λ i 4πµ Γ() (1 + x) me
− i 2
dx 2(1 − x) (2 − ) − 
2 16π 0 ∆2 2 ∆1
!
2 Z 1 2 


(2 − )(1 + x ) − 2(1 − )x m2e
2
 
−e 4πµ 2
= dx (1 − x) Γ() 2(2 − ) − 2
16π 2 0 ∆1 ∆1
1 
λ2 4πµ2 (1 + x)2 m2e
Z    
+ dx (1 − x) Γ() (2 − ) − 
32π 2 0 ∆2 ∆1
!
−e2 1


µ̃2 (2 − )(1 + x2 ) − 2(1 − )x m2e
Z   
1 2
= dx (1 − x) +  log (2 − ) − 
8π 2 0  ∆1 ∆1
1
λ2 µ̃2 (1 + x)2 m2e
Z     
1
+ dx (1 − x) +  log (2 − ) − 
32π 2 0  ∆1 ∆1
2
−α λ
= + + finite terms. (58)
π 32π 2 
The wave function renormalization factor Z2 comes from the sum
k−p k−p

p p + p p . (59)
k k

We have already calculated these contributions in the text and in question 7.2:
−α −λ2
δZ2 = + + finite terms. (60)
4π 64π 2 

13
 Solutions to Peskin and Schroder – Andrzej Pokraka

Taking the difference we see that

−α −λ2 −α λ2
δZ2 − δZ1 = + 2
− − + finite terms
4π 64π  π 32π 2 
−3α −3λ2
= + + finite terms
4π 64π 2 
6= 0. (61)

14