PHYSICS 513: QUANTUM FIELD THEORY HOMEWORK 6 1

Physics 513, Quantum Field Theory

Homework 6
Due Tuesday, 21st October 2003
Jacob Lewis Bourjaily
1. For the following derivations it will be helpful to recall the following:
Pψ(t, ~x)P † = ηa γ 0 ψ(t, −~x);
P ψ̄(t, ~x)P † = ηa∗ ψ̄(t, −~x)γ 0 ;
CψC † = −i(ψ̄γ 0 γ 2 )| ;
C ψ̄C † = −i(γ 0 γ 2 ψ)| ;
T ψ(t, ~x)T † = γ 1 γ 3 ψ(−t, ~x);
T ψ̄(t, ~x)T † = −ψ̄(−t, ~x)γ 1 γ 3 .
a) We are to verify the transformation properties of Aµ ≡ ψ̄γ µ γ 5 ψ and T µν ≡ ψ̄σ µν ψ under
P.
Let us first consider the axial vector Aµ .
PAµ P † = P ψ̄γ µ γ 5 ψP † = ηa∗ ψ̄γ 0 γ µ γ 5 ηa γ 0 ψ,
= ψ̄γ 0 γ µ γ 5 γ 0 ψ,
= −ψ̄γ 0 γ µ γ 0 γ 5 ψ,
= −ψ̄γµ γ 5 ψ = −Aµ .
The last step can be seen by noticing that
½ µ ¾
0 µ 0 γ µ=0
γ γ γ = = γµ .
−γ µ µ = 1, 2, 3
Now we will consider the transformation of the tensor T µν .
PT µν P † = P ψ̄σ µν ψP † = ηa∗ ψ̄γ 0 σ µν ηa γ 0 ψ,
= ψ̄γ 0 σ µν γ 0 ψ,
= ψ̄σµν ψ = Tµν .
Similar to the axial vector case, the last step is a result of directly verifying the identity
½ µν ¾
0 µν 0 i 0 µ ν 0 0 ν µ 0 σ µ, ν 6= 0 or µ, ν = 0
γ σ γ = (γ γ γ γ − γ γ γ γ ) = = σµν .
2 −σ µν one of µ or ν = 0
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
b) We are to verify the transformation properties of V µ ≡ ψ̄γ µ ψ and Aµ ≡ ψ̄γ µ γ 5 ψ under C.
Let us first consider the transformation of the vector V µ .
CV µ C † = C ψ̄γ µ ψC † = −i(γ 0 γ 2 )| γ µ (−i)(ψ̄γ 0 γ 2 )| ,
= −ψ̄γ 0 γ 2 γ µ| γ 0 γ 2 ψ,
= ψ̄γ 0 γ 2 γ µ| γ 2 γ 0 ψ,
= −ψ̄γ µ ψ = −V µ .
Let us now consider the axial vector Aµ .
CAµ C † = C ψ̄γ µ γ 5 ψC † = −i(γ 0 γ 2 ψ)| γ µ γ 5 (−i)(ψ̄γ 0 γ 2 )| ,
= −ψ̄γ 0 γ 2 γ 5 γ µ| γ 0 γ 2 ψ,
= ψ̄γ 0 γ 2 γ µ| γ 5 γ 0 γ 2 ψ,
= −ψ̄γ 0 γ 2 γ µ| γ 5 γ 2 γ 0 ψ,
= −ψ̄γ 0 γ 2 γ µ| γ 2 γ 0 γ 5 ψ,
= ψ̄γ µ γ 5 ψ = Aµ .
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
2 JACOB LEWIS BOURJAILY

c) We are to confirm the transformation properties of P ≡ iψ̄γ 5 ψ and V µ ≡ ψ̄γ µ ψ under T .

First let us consider the transformation of the pseudo-scalar P .
T P T † = T iψ̄γ 5 ψT † = −i(−ψ̄γ 1 γ 3 )γ 5 (γ 1 γ 3 ψ),
= iψ̄γ 1 γ 3 γ 5 γ 1 γ 3 ψ,
= −iψ̄γ 5 ψ = −P.
Let us now consider the transformation of the vector V µ .
T V µ T † = T ψ̄γ µ ψT † = ψ̄γ 3 γ 1 γ µ∗ γ 1 γ 3 ψ,
= ψ̄γµ ψ = Vµ .
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
2. a) We are to demonstrate the transformation properties of V µ and Aµ , as previously defined,
under CP.
We have almost computed every detail necessary for our solution in question (1) above.
The only transformation that we have not yet confirmed is the transformation of the vector
V µ under P. Let us compute that now.
PV µ P † = P ψ̄γ µ ψP † = ηa∗ ψ̄γ 0 γ µ ηa γ 0 ψ,
= ψ̄γ 0 γ µ γ 0 ψ,
= ψ̄γµ ψ = Vµ .
By simply applying our transformation properties derived above in succession, we observe
that,
P C
V µ = ψ̄γ µ ψ −−−−→ ψ̄γµ ψ −−−−→ −ψ̄γµ ψ = −Vµ
P C
Aµ = ψ̄γ µ γ 0 ψ −−−−→ −ψ̄γµ γ 5 ψ −−−−→ −ψ̄γµ γ 5 ψ = −Aµ
b) Expecting an analogy with the electromagnetic current vector, we will check the transfor-
mation properties of each.
agree? agree? agree?
P C CP
J µ −−−−→ Jµ J µ −−−−→ −J µ J µ −−−−→ −Jµ
P C CP
V µ −−−−→ Vµ yes V µ −−−−→ −V µ yes V µ −−−−→ −Vµ yes
P C CP
Aµ −−−−→ −Aµ no Aµ −−−−→ Aµ no Aµ −−−−→ −Aµ yes
c) We will demonstrate that the weak Lagrangian,
GF
Lweak ≈ √ (Vµ − Aµ )(V µ − Aµ ),
2
is not invariant under C or P, yet is invariant under CP.
Like before, I will directly compute all of the transformations using the table made above
in part (b) above. First note that
Lweak ∝ V 2 − 2Vµ Aµ + A2 .
When we take each of the of transformations from above, we see that
P
V 2 − 2Vµ Aµ + A2 −−−−→ V 2 + 2Vµ Aµ + A2 6= Lweak ;
C
V 2 − 2Vµ Aµ + A2 −−−−→ V 2 + 2Vµ Aµ + A2 6= Lweak ;
CP
V 2 − 2Vµ Aµ + A2 −−−−→ V 2 − 2Vµ Aµ + A2 = Lweak .
So Lweak is not invariant under C or P by is under CP, as we were required to demonstrate.
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
 PHYSICS 513: QUANTUM FIELD THEORY HOMEWORK 6 3

3. Let us define the product of the 3 discrete symmetry transformations as Θ ≡ CPT . We must
show that under Θ, the Dirac field transforms by the rule
Θψ(x)Θ† = γ 5 ψ ∗ (−x),
where
ψ ∗ (x) ≡ (ψ(x)† )| .

Like so many times before, we will proceed by direct calculation.

Θψ(x)Θ† = CPT ψ(t, ~x)T † P † C † ,
= CPγ 1 γ 3 ψ(−t, ~x)P † C † ,
= ηa Cγ 1 γ 3 γ 0 ψ(−x)C † ,
= −iηa γ 1 γ 3 γ 0 (ψ(−x)† γ 2 )| ,
= −iηa γ 1 γ 3 γ 0 γ 2| ψ ∗ (−x),
= −iηa γ 0 γ 1 γ 3 γ 2 ψ ∗ (−x),
= iηa γ 0 γ 1 γ 2 γ 3 ψ ∗ (−x),
= ηa γ 5 ψ ∗ (−x).
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι

4. For the following derivations it will be useful to recall that

µ ¶ µ ¶
0 0 1 i 0 σi
γW = , γW = ,
1 0 −σ i 0
where µ ¶ µ ¶ µ ¶
0 1 0 −i 1 0
σ1 = , σ2 = , σ3 = .
1 0 i 0 0 −1
a) We must show that any new matrices defined by
µ
γ µ = U γW U †,
where U is an arbitrary 4 × 4 unitary matrix, satisfy the dirac algebra. This is proven by
demonstrating that
{γ µ , γ ν } = 2g µν .
Knowing that the Weyl-representation γ µ ’s satisfy the Dirac algebra, we will directly show
that,
µ
{γ µ , γ ν } = {U γW U † , U γW
ν
U † },
µ µ
= U γW U † U γW
ν
U † + U γW
ν
U † U γW U †,
µ ν ν µ
= U (γW γW + γW γW )U † ,
= 2U g µν U † = 2g µν .
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
b) Consider the unitary matrix which produces the Dirac representation
µ ¶
1 1 1
UD = √ .
2 −1 1
We must show that UD is in fact unitary and we must find the matrices γ µ in the Dirac
representation.
The unitarity of UD is trivial
µ ¶µ ¶
† 1 1 1 1 −1
UD UD = = 14×4 .
2 −1 1 1 1
When the matrices are directly computed, we see that
µ ¶
0 1 0
γ = , γ i = γW
i
.
0 −1
4 JACOB LEWIS BOURJAILY

c) We must now show that in a general frame, the Dirac spinor takes the form,
µ √ ¶
E +√mξ s
us (p) = .
~σ · p~ ξ s / E + m
This is demonstrated by showing that it solves the Dirac equation, or, namely, that
γ µ pµ us (p) = mus (p).
This is simple to evaluate directly. Noting our Dirac representation of the γ µ ’s and that
p0 = E, we see
µ ¶µ √ ¶
p0 −~σ · p~ E +√mξ s
γ µ pµ us (p) = ,
~σ · p~ −p0 ~σ · p~ ξ s / E + m
 h √ 2 2
i 
E E+m− E √ −m
E+m i
ξ s
= h √ ,
~σ · p~ E + m − √E~ σ ·~
E+m
p
ξ s

à √ ³ 2
´ !
−m2
E + m E − EE+m ξs
= ,
√~ σ ·~
p
(E + m − E)ξ s
E+m
µ √ ¶
E + m(E −√E + m)ξ s
= ,
m~σ · p~ ξ s / E + m
µ √ ¶
m E +√m ξ s
= ,
m~σ · p~ ξ s / E + m
= mus (p).
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
d) We must show that the solution found in part (c) is normalized in the standard way.
Given that ξ is normalized such that ξξ † = 1, we see that
µ ¶µ √ ¶
¡ √ √ ¢ 1 0 E+
ūu = u† γ 0 u = E + m ξ † ~σ · p~ ξ † / E + m √m ξ ,
0 −1 ~σ · p~ ξ/ E + m
µ ¶
† (~σ · p~)2
= ξ ξ (E + m) − ,
E+m
E 2 + 2mE + m2 − p~2
= ,
E+m
E 2 + 2mE + m2 − E 2 + m2
= ,
E+m
2mE + 2m2
= ,
E+m
= 2m.
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι