QFT Ii

Class Notes for Quantum Field Theory: Section II
S -Matrix, Wick’s Theorem, Feynman Rules, Specific Sample Calculations
Primary Reference: Mandl & Shaw, “Quantum Field

Theory”
See also: Peskin and Schroeder, “Quantum Field
Theory”
Jack Gunion
U.C. Davis
230A – Part II
J. Gunion
The S-matrix
• Ideally, we would like to solve the coupled differential equations that result
from introducing interactions among the free fields, as for instance E&M
interactions found by the minimal substitution rules.
• An exact solution has not been found, but we have made much progress
using a perturbative approach. So far, we have been fortunate that nature
has always allowed such a perturbative approach to be fruitfully compared
to experiment.
QED, in particular, has passed extremely precise tests because we can
compute very precisely due to the small size of α ∼ 1/137.
• The perturbative approach is most conveniently derived in the interaction

picture in which most of the time evolution of the states (namely, that
associated with the free particle part of the Hamiltonian) is removed
leaving behind only the (small) time evolution from the perturbatively
treated interaction(s).
J. Gunion 230A – Part II 1

So far, we have been using the Heisenberg picture in which state vectors
(in the Foch space sense) are constant in time while the field operators
contain all the time dependence.
• The solution to the full problem can be formulated using the Dyson
expansion which is ideally suited for obtaining perturbation results systematically
The Interaction Picture
• Consider QED as the example theory.
• We divide up the full L into
L = L0 + LI (1)
with
1
L0 =: ψ(x)(i∂ γµ − m)ψ(x) − (∂µAν )(∂ µAν ) :
µ
(2)
2
(using the Lorentz condition formulation for the Aµ fields) and
LI =: eψ(x)Aµγµψ(x) : . (3)

• A full (non-perturbative) solution to the theory would arise if we could
compute
U ≡ U (t, t0) = e−iH (t−t0) (4)
where U defines the time evolution of the Schroedinger picture states,
|A, tiS = U |A, t0iS (5)
from which one has

d
i |A, tiS = H|A, tiS . (6)
dt
U also defines the relation between the Schroedinger and Heisenberg
pictures:
|A, tiH = U †|A, tiS = |A, t0iS (7)
and
O H (t) = U †O S U , (8)
where the Schroedinger picture is defined by O S being independent of time.
Of course, O = H is time-independent and is the same in both pictures.
At t = t0, the states and operators in the two pictures are the same.
However, the Schroedinger state changes with time whereas the Heisenberg
picture state is constant in time: one could write |A, tiH = |AiH , but MS
doesn’t.

• Of course, expectation values are the same in the two pictures since U is
unitary; i.e.
S
S hB, t|O |A, tiS =H hB, t|O H (t)|A, tiH . (9)
• From Eq. (8),

d
i O H (t) = [O H (t), H] . (10)
dt
• Now write H = H0 + HI , where the I subscript is for interaction and not

for interaction picture!
• The I.P. is defined relative to the S.P. by using
U0 ≡ U0(t, t0) = e−iH0(t−t0) (11)
to define the I.P. state
|A, tiI = U0†|A, tiS (12)

and I.P. operators
O I (t) = U0†O S U0 . (13)
Here the subscript I on |A, tiI and the superscript I on O I (t) refer to the
interaction picture (I.P.) and not to the interaction part of the Hamiltonian.
Note that H0I = H0S ≡ H0 .
• Differentiating Eq. (13) gives
d
i O I (t) = [O I (t), H0] (14)
dt
and we also find
d d h i
i |A, tiI = i e+iH0(t−t0)|A, tiS
dt dt
+iH0 (t−t0 ) +iH0 (t−t0 ) d
= −H0e |A, tiS + e i |A, tiS
dt
= −e+iH0(t−t0)H0S |A, tiS + e+iH0(t−t0)H S |A, tiS
= e+iH0(t−t0)HIS |A, tiS

= e+iH0(t−t0)HIS e−iH0(t−t0)|A, tiI
= HII |A, tiI , (15)
where: 1) we used the fact that H S ≡ H in this notation; 2) we used the

fact that H0S = H0 (see below Eq. (13)); 3) we used Eq. (6); and 4) we
were very careful to preserve the placing of the H operator next to |A, tiS .
• Note that we can rewrite the last few lines above as
HII = e+iH0(t−t0)HIS e−iH0(t−t0)

= e+iH0(t−t0)e−iH (t−t0)HIH (t)eiH (t−t0)e−iH0(t−t0)
†
= U 0 HIH U 0 , (16)
where U 0 is a unitary operator that takes us from the H.P. to the I.P. for
any operator (not just the Hamiltonian).

The S-matrix expansion
• This expansion is based upon the I.P. reviewed above, in which:

– the operators satisfy the Heisenberg-like equations of motion but involving
the free Hamiltonian H0 only, not the complete H, as algebraically stated
in Eq. (14):
d I
i O = [O I (t), H0] ; (17)
dt
– if the interaction LI does not involve field derivatives (as for all cases
of interest to us), the fields canonically conjugate to the interacting
fields and those conjugate to the free fields are identical, e.g. in QED
∂L0
πα = ∂∂Lψ̇α
= ∂ ψ̇α
, where we are in the H.P. when writing such equations
and thinking about conjugate fields.
This implies that since the I.P. and the H.P. are related by a unitary
transformation, then in the I.P. the interacting fields satisfy the same
commutation or anticommutation relations as the free fields. Using scalar
field notation, this is shown algebraically as follows:
† †
[φI (~ y , t)] = [U 0 φH (~
x, t), π I (~ x, t)U 0, U 0 π H (~
y , t)U 0]

†
= U 0 [φH (~ y , t)]U 0
x, t), π H (~
†
= U 0 iδ 3(~ ~ )U 0
x−y
= iδ 3(~
x−y
~) . (18)
This in turn implies that we will 2nd quantize the I.P. fields in a manner
that is essentially identical to what we did in the free field case.
– Note: there is a subtlety since the asymptotic states in the I.P. can have
a different mass and different normalization than the asymptotic states
in the free-particle case.
This difference arises because the asymptotic states must contain not
just the free particle behavior, but the full collection of virtual processes
associated with the presence of interactions (think bare line + line with
all possible insertions, loops, ....).
This is a subtlety that we will eventually address, but not in detail until
we come to renormalization theory.
• At this point, MS simplifies the notation a bit and writes in the I.P.
d
i |Φ(t)i = HI (t)|Φ(t)i (19)
dt
where
HI (t) ≡ eiH0(t−t0)HIS e−iH0(t−t0) = HII (t) ; (20)
that is, we will be dropping the I superscript reminding us that everything
is in the I.P.
• It will also be important to note that if HIS is a product of fields in the S.P.
(i.e product of operators in the S.P.), then HI will be the same product of
fields/operators where all the fields/operators are in the I.P. (This is shown
by inserting a whole bunch of 1’s in between the operators in the form
1 = e−iH0(t−t0)eiH0(t−t0).)
• If we start with a state at some initial time, |Φ(ti)i = |ii, the solution of
Eq. (19) gives the state |Φ(t)i at any other time t.
Since HI (t) is hermitian, this time development is given by a unitary
transformation. Accordingly, it preserves the normalization of the state:
hΦ(t)|Φ(t)i = const. (21)
• This formalism we are going to develop will not be appropriate for bound
states. It will apply to the situation where we consider an initial state in

which there are a certain number of widely separated particles which then
converge upon a small interaction region, interact via the interactions, and
then at large times a (usually different) set of widely separated particles
emerge from the interaction region.
Eq. (19) determines the state |Φ(t = ∞)i into which |Φ(t = −∞)i
evolves long after the scattering is over and all particles are far apart again.
• The S-matrix is defined by
|Φ(t = ∞)i = S|Φ(t = −∞)i = S|ii . (22)
Note that Eq. (21), conservation of normalization, implies
hΦ(∞)|Φ(∞)i = hΦ(−∞)|S †S|Φ(−∞)i = hΦ(−∞)|Φ(−∞)i (23)
which requires that S †S = 1, i.e. S is a unitary operator.
• S|ii can consist of a large selection of possible final states |f i.

The probability that after the collision (i.e. at t = ∞) the system is in
state |f i is given by
|hf |Φ(∞)i|2 , (24)

assuming unity normalizations for |Φ(−∞)i = |ii (and hence also for
|Φ(∞)i given conservation of normalization as noted above). The
corresponding probability amplitude is
hf |Φ(∞)i = hf |S|ii ≡ Sf i . (25)
in terms of which X
|Φ(∞)i = |f iSf i . (26)
f
The unitarity of the S-matrix can be written in this basis as
X
|Sf i|2 = 1 , (27)
f
as follows from
1 = hi|ii = hi|S †S|ii = hi|S †|f ihf |S|ii = Sf∗iSf i . (28)
This equation expresses the conservation of probability. It is more general

than the corresponding conservation of particles in NRQM since now
particles can be created and/or destroyed.

Calculating the S-matrix
• We must solve
d
i |Φ(t)i = HI (t)|Φ(t)i (29)
dt
with initial condition |Φ(−∞)i = |ii. Equivalently we must compute
Z t
|Φ(t)i = |ii + (−i) dt1HI (t1)|Φ(t1)i . (30)
−∞
This form obviously satisfies the above differential equation and has the
correct initial boundary condition at t = −∞.
The perturbative series is based on the iterative solution of this equation
where we plug in for |Φ(t1)i the same form as given above for |Φ(t)i and
so forth.
Z t
|Φ(t)i = |ii + (−i) dt1HI (t1)|ii
−∞

Z t Z t1
+(−i)2 dt1 dt2HI (t1)HI (t2)|ii + . . . (31)
−∞ −∞
yielding (as t → ∞)
∞
X Z ∞ Z t1 Z tn−1
S= (−i)n dt1 dt2 . . . dtnHI (t1)HI (t2) . . . HI (tn) .
n=0 −∞ −∞ −∞
(32)
• Now comes a very crucial trick that partly motivates the use of time
ordering. The above expression can be rewritten as
∞ ∞ ∞ ∞
(−i)n
X Z Z Z
S= dt1 dt2 . . . dtnT {HI (t1)HI (t2) . . . HI (tn)} .
n=0
n! −∞ −∞ −∞
(33)
Here, the time ordering operation has been generalized to include an
arbitrary list of operators. It is necessary to insert the T instruction, since
in the original form, Eq. (32), the HI operators (which do not commute
with one another when evaluated at different times) were very definitively

ordered so that HI ’s with earlier times were always written to the right of
HI ’s with later times.
Let me show how one arrives at this using the 2nd order term as an example.
We write (dropping the subscript I for the moment):
Z t Z t1
dt1 dt2H(t1)H(t2)
−∞ −∞
Z t Z t
= dt2 dt1H(t1)H(t2) interchange of integration order
−∞ t2
Z t Z t
= dt1 dt2H(t2)H(t1) relabel t1 ↔ t2
−∞ t1
Z t Z t
= dt1 dt2T {H(t1)H(t2)} no change because of T instruction .
−∞ t1
(34)

Of course, it is also trivially true that
Z t Z t1 Z t Z t1
dt1 dt2H(t1)H(t2) = dt1 dt2T {H(t1)H(t2)} (35)
−∞ −∞ −∞ −∞
since the H’s are already in the correct time ordering.
Thus, we can write

Z t Z t1
dt1 dt2H(t1)H(t2)
−∞ −∞
t t1
1
Z Z
= dt1 dt2T {H(t1)H(t2)}
2 −∞ −∞
t t
1
Z Z
+ dt1 dt2T {H(t1)H(t2)}
2 −∞ t1
t t
1
Z Z
= dt1 dt2T {H(t1)H(t2)} . (36)
2 −∞ −∞

Given that we want to include − signs in the definition of T when fermionic
fields are involved, the absence of any extra signs in the 2nd version of the
above requires that HI contains an even number of fermion factors (as in
QED) and that HI be written in terms of a local interaction (all fields at
same space-time x point). With this latter point in mind we can write
∞ ∞ ∞ ∞
(−i)n
X Z Z Z
S = dt1 dt2 . . . dtnT {HI (t1)HI (t2) . . . HI (tn)}
n=0
n! −∞ −∞ −∞
∞
(−i)n
X Z Z Z
= d4 x 1 d4 x 2 . . . d4xnT {HI (x1)HI (x2) . . . HI (xn)} .
n=0
n!
(37)
R
where the d4x integrals are over all space and all times.
• MS now embarks on a discussion of why it is justified to use non-interacting

states to specify |ii and |f i instead of those including the photon and loop
clouds. This discussion is simply incorrect. In fact, in the interacting theory,
|ii and |f i are not described using a free-particle state Foch state picture
and basis. They will be defined using creation and annihilation operators

that create single particle states in the sense that the states can exist on
their own even in the presence of interactions. Such states will have a mass
that is different from the bare mass appearing in H0. This difference arises
from the fact that as an isolated particle moves along in the interacting
theory, it is continually emitting and reabsorbing virtual quanta in a manner
dictated by Feynman diagrams. This cloud of virtual processes alters the
physical particle mass.
It is only these states which include these virtual particle clouds that can be
used to construct a proper basis for the theory after including interactions.
So let us imagine that we have creation and annihilation operators for these
isolated single particle states including their separated virtual clouds. The
states |ii and |f i will be defined by using these operators. Effectively, we
will mimic this by using I.P. basis operators to form |ii and |f i, but with
the mass appearing in the ωp~ contained in the φI (x) exponential factors
being the full mass obtained after including the virtual particle clouds.
In addition, the creation operator, used to quantize φI (x) in the usual way,
will lead to a strange normalization of the states |ii and |f i when they are
defined using this same creation operator. The state created will have a
single particle component present with some reduced normalization Z < 1.
Very roughly this means that |~pi = a†(~p)|Ωi, where |Ωi is now the vacuum

in the presence of interactions, has normalization h~
p|~
pi = Z < 1. This factor
of Z arises because the creation operator a† appearing in the field φI (x)
(which we are assuming has conventional normalization, as required for the
standard canonical commutation 2nd quantization) also ’knows’ that the
particle it is creating has connections to multi-particle states as well as to a
single particle state, and it is only after summing over all such possibilities
that full unity normalization holds.
√
The Z factor arising for each single particle state used to construct |ii
or |f i must be carefully treated. It used to be that Peskin and Schroeder
tried to give a naive discussion of this in their earlier chapters. I see that
in the latest edition of their book they have given up trying to do this, just
as I will and as the Mandl-Shaw book implicitly does.
√
Thus, we will simply drop the Z factors that should really be present in
the scattering formulas. Perturbatively, this can actually be justified if one
is only interested in computing so-called tree-level amplitudes in which no
closed loops of virtual particles appear. The reason is that in perturbation
theory Z ∼ 1 + O(α) (e.g. in QED) while the tree-level amplitudes define
the lowest order in perturbation theory at which a process can take place.
Thus, including the correction to Z ∼ 1 is a higher order correction.
In contrast, we must use the physical mass (which includes loops ....) of

a free-particle-like state in defining our I.P. states and fields. It is not the
same as the mass of the free-particle theory.
Wicks Theorem
• We must now develop a computational technique for writing down an

expression for S given a certain form for HI (x).
• To get a non-zero contribution to Sf i we must have an appropriate number

of creation and/or annihilation operators to match those appearing in the
definitions of the states |ii and |f i.
Creation operators in the product of HI ’s for particles that do not appear
in |ii or |f i must be compensated by annihilation operators; the result is
emission and later absorption of a virtual particle that only appears in an
intermediate state (i.e. one that mediates between |ii and |f i, but is not
part of either).
• The only way to make this language clear is to give a specific example.
MS chooses e−(p) + γ(k) → e−(p0) + γ(k0) at lowest possible order in
perturbation theory.

You thus start with an initial state
p)a†s(~
|ii = c†r (~ k)|0i (38)
and end with a final state
|f i = c†r0 (~
p 0)a†s0 (~
k0)|0i . (39)
• So, now the trick is to go to the form of S given in Eq. (33) and to use
an n such that there are enough HI ’s in between hf | and |ii as to get a
non-zero result. If you insert the minimal number, i.e. use the smallest n,
then this defines the “tree-level” amplitude.
Since hf | has a different photon in it than does |ii, you will need two
A fields, one to annihilate the initial photon and one to create the final
photon. This means you will need two HI ’s.
• Since the minimal substitution interaction is HI = −LI = −e : ψA/ ψ :,

this also implies that that you will have an extra ψ and ψ in addition to
the ψ and ψ needed to annihilate the electron present in the initial state
and create the one present in the final state.

These extra ψ and ψ will have inside them creation and annihilation
operators that will basically annihilate one another.
• For now, let us not worry about these two extra guys and focus only on the
fields necessary to get from the initial state to the final state.
The only non-zero contribution to

hf | −e : ψA/ ψ :x1 −e : ψA/ ψ :x2 |ii (40)
will come from the structure (notice I am not yet specifying whether the
fields below are at x1 or x2 – we will come to the appropriate mixtures):
− −
hf |ψ A−ψ +A+|ii = h0|as0 (~
k0)cr0 (~ p)a†s(~
p 0)ψ A−ψ +A+c†r (~ k)|0i (41)
where the A+ has an a in it to “kill” the a†s(~ k) used to define |ii and ψ +
has the c in it to “kill” the c†r (~
p) that is also part of defining |ii. Similarly,
−
the c† in ψ will “kill” the cr0 (~ p 0) and the a† in A− will “kill” the as0 (~
k0).
By “kill” I mean the operation (to give one of the 4 kills)

ai(~l)a†s(~
k)|0i = [ai(~l), a†s(~
k)] + a†s(~
k)ai(~l) |0i = δisδ~l,~k|0i + 0 . (42)

Here i, ~l are the dummy summation variables used in defining
X 1
A µ+
(x2) = q µ
i (~l)ai(~l)e−il·x2 . (43)
i,~
l
~
2V |l|
• Note: The ordering which makes the killing simple, as given above, is the
normal ordering of the required operators.
Thus, in general, we will want to expand the S-matrix as a sum of normal
products. The method for doing so is due to Dyson and Wick.
• First, we repeat the general definition of a normal product. Let Q, R, . . . , W

be operators of the type ψ ±, A±, etc. Each contains a creation or
annihilation operator. Then,
: QR . . . W := (−1)P × (reordering) (44)
+
where the reordering is such that all absorption operators (i.e.
components) are to the right of all creation operators (i.e. − components).

The exponent P is the number of interchanges of fermion operators required
for this reordering.
We also require that normal ordering be a distributive operation: : RS . . . +
V W . . . :=: RS . . . : + : V W . . . :.
• Now, we note that in all cases HI (x) =: A(x)B(x) . . . : is already normal

ordered. (Remember if A = A+ + A− then use distributive rule to do the
normal ordering.)
What appears in S is T {HI (x1) . . . HI (xn)}. We must learn how to deal
with this structure.
• First, note that (with A = A(x1) and B = B(x2))
[A+, B −]+ ,

for two fermion fields
AB− : AB := (45)
[A+, B −] , otherwise
Proof, e.g. for two fermions fields:

+ + + − − + − − + + − + − + − −
h i h i
AB− : AB : = A B +A B +A B +A B − A B −B A +A B +A B
+ − − + + −
h i
= A B + B A = A ,B . (46)
+

• Now, the above commutators or anticommutators are c-numbers and so
their vacuum expectation values are the same as the numbers themselves.
So consider, for example, (remembering A+|0i = 0 since A+ contains the
annihilation operator)
h0|[A+, B −]+|0i = h0|[A+B − + B −A+]|0i

= h0|A+B −|0i
= h0|(A+ + A−)(B + + B −)|0i
= h0|AB|0i (47)
where the next to last line follows since h0|A− = 0 and B +|0i = 0. The
commutator case follows a similar sequence of steps:
h0|[A+, B −]|0i = h0|[A+B − − B −A+]|0i

= h0|A+B −|0i
= h0|(A+ + A−)(B + + B −)|0i
= h0|AB|0i (48)

As a result, in both cases we have
AB =: AB : +h0|AB|0i. (49)
• Further, we note that : AB := ± : BA :, with the − sign applying in the

case of two fermion fields, and the plus sign otherwise.
Just to make sure you understand this statement, lets do a fermion case.
We must remember that two creation operators anticommute as do two
annihilation operators.
: AB : = : (A+ + A−)(B + + B −) :
= A+ B + + A− B + + A− B − − B − A+
= −B +A+ + A−B + − B −A− − B −A+
= − : (B + + B −)(A+ + A−) :
= − : BA : , (50)
where we used − : B +A− := −(−)A−B + = A−B +.

• Using the above, it follows that if x01 > x02 we have
T {A(x1)B(x2)} = A(x1)B(x2)
= : A(x1)B(x2) : +h0|A(x1)B(x2)|0i
= : A(x1)B(x2) : +h0|T {A(x1)B(x2)}|0i (51)
whereas if x01 < x02, then
T {A(x1)B(x2)} = ±B(x2)A(x1)
= ± [: B(x2)A(x1) : +h0|B(x2)A(x1)|0i]
= ± [± : A(x1)B(x2) : ±h0|T {A(x1)B(x2)}|0i]
= : A(x1)B(x2) : +h0|T {A(x1)B(x2)}|0i , (52)
i.e. exactly the same result since the ± sign changes cancel one another.
In short, we have
T {A(x1)B(x2)} =: A(x1)B(x2) : +A(x1)B(x2) (53)

where we have introduced the shorthand notation:
A(x1)B(x2) ≡ h0|T {A(x1)B(x2)}|0i (54)
which is simply the Feynman propagator that we have discussed.

This Feynman propagator is, of course, only non-zero when A and B have
compensating creation and annihilation operators. Sometimes the above
structure is also called the “contraction” of the two fields.
• The Feynman propagators we have so far considered are:
φ(x1)φ(x2) = i∆F (x1 − x2)
φ(x1)φ†(x2) = φ†(x2)φ(x1) = i∆F (x1 − x2)
ψα(x1)ψ β (x2) = −ψ β (x2)ψα(x1) = iSF αβ (x1 − x2)
Aµ(x1)Aν (x2) = iDFµν (x1 − x2) . (55)

• We now need to generalize this beyond just 2 operators in the following
way. We define a generalized normal product containing many operators,
some of which have been contracted, as
: ABCDEF . . . J KLM . . . :
= (−1)P AKBCEL . . . : DF . . . J M . . . : . (56)
Here, P is the number of interchanges of neighboring fermion operators

required to change the order ABC . . . to AKB . . .; for example,
: ψα(x1)ψβ (x2)Aµ(x3)ψ γ (x4)ψ δ (x5) :
= (−1)ψβ (x2)ψ δ (x5) : ψα(x1)Aµ(x3)ψ γ (x4) : (57)
• With this definition, Wick has proven the following generalization of

T {AB} =: AB : +AB (assuming that none of the times are the same in
the operators):
T {ABCD . . . W XY Z}

= : ABCD . . . W XY Z :
+ : ABC . . . Y Z : +all other single contraction cases
+ : ABCDEF . . . Y Z : +all other double contraction cases
+...
+all maximal contraction cases . (58)
In the maximal contraction cases, if there are matching fields then no field
will be left over.
The proof of this theorem is by induction and can be found in the old book
by Bjorken and Drell on Relativistic Field Theory (2nd volume). Another
form of the inductive proof appears at the end of Peskin and Schroeder,
section 4.3. We will not go through the proof here, but will give some
illustration below of how it works in a specific case of interest.
• Finally, we must figure out what the correct procedure is for the case of
interest where we are dealing with
T {HI (x1) . . . HI (xn)} = T {: AB . . . :x1 . . . : AB . . . :xn } . (59)

Let us consider a simplified version of this sort of structure. For x01 > x02
we have, assuming bosonic fields for simplicity,
T {A(x1) : B(x2)C(x2) :}
= (A+ + A−)(B +C + + B −C − + B −C + + C −B +)
= A−(B +C + + B −C − + B −C + + C −B +)
+(B +C + + B −C − + B −C + + C −B +)A+
+[A+, B −]C − + B −[A+, C −] + [A+, B −]C + + [A+, C −]B +
= : ABC : +[A+, B −]C + [A+, C −]B
= : ABC : +h0|T {AB}|0iC + h0|T {AC}|0iB
= : ABC : +ABC + ACB (60)
where we of course wrote : BC : in normal ordered form to begin with, and

then moved the A+ part of the A field from left to right until we also got
complete normal ordering. In so doing, we had to cross the B − and C −
fields with which A+ has non-zero contractions (assuming the same kind
of field). But, it was never necessary to commute a B component past a
C component, since they were already in normal ordering. In the above,

we also used the fact that [A+, B −] = h0|T {AB}|0i by virtue of the fact
that (by our initial assumption) A had later time than B. A very similar
proof would have given the same general result had the time ordering been
reversed.
The net result in all cases is that Eq. (58) applies with the proviso that no
equal time contractions are to be included, which in the present context
means that fields inside the same HI are not to be contracted.
• You should note that this result is critically dependent upon using : HI :
(i.e. normal ordering the interaction Hamiltonian from the beginning) rather
than HI . In the latter case, you would include equal time contractions in
applying Wicks theorem. Using normal ordering for HI , in a certain sense,
obscures what is really happening.
Peskin and Schroeder discuss the situation without normal ordering HI . In
this PS approach (also quite common in other texts), all the contractions of
fields in the same HI with one another are kept. Such contractions, as we
already know, are associated with infinities. We have discussed how such
infinities can be thrown away in the linear free particle context (i.e. when
considering just one H0 in the free particle case). It is quite another matter
to show that they don’t matter when considering HI multiply repeated

inside the S matrix.
In fact, such a proof can be carried out and is contained in Sec. 4.4 of
Peskin-Schroeder. In order to understand this proof, you must follow the
development of the concept of the true vacuum state. The true vacuum
state contains all sorts of infinities associated with not normal ordering HI .
But, when computing an S matrix element, the infinities arising in the S
matrix calculation associated with not normal ordering HI are canceled by
the infinities associated with simply defining the true interacting vacuum
|Ωi relative to the non-interacting vacuum |0i.
Pictorially, the vacuum should be visualized as a bare vacuum plus all kinds
of Feynman graphs containing various sorts of closed bubbles, the latter
arising only when we include contractions of fields within the same HI .
Meanwhile any given S matrix calculation will have a “connected” part
associated with the process at hand, multiplied by this same collection of
closed bubble graphs. But the closed bubble factor is simply absorbed into
the definition of the “true” vacuum and thus does not affect the final result
for the physical calculation.
Well, I am sure that all this discussion is somewhat mysterious; it is only in
230C that it will really be clarified. For now, we must be satisfied with the
fact that at tree-level all the naive manipulations are ok.

Feynman Diagrams and QED rules
Calculation of the Matrix element
• Let us return to the the process
e−(p) + γ(k) → e−(p0) + γ(k0) (61)
process. We wish to evaluate
hf |S|ii = h0|as0 (~
k0)cr0 (~ p)a†s(~
p 0)Sc†r (~ k)|0i (62)
with the part of S that we keep being the minimal version capable of
yielding the process. As described earlier this is what MS calls S (2), i.e. the
2nd order component:
e2
Z Z
S (2) = − d4 x 1 d4x2T {: ψ 1A/ 1ψ1 :: ψ 2A/ 2ψ2 :} (63)
2
and, in fact, only certain components of this S (2) that emerge in the Wick
expansion survive: namely, those with two contracted fermion fields with
the others left over to “kill” the creation and annihilation operators defining
the initial state and the hermitian conjugate of the final state:
2Z
e
Z
S (2) 3 − d4x1 d4x2 : ψ 1A/ 1ψ1ψ 2A/ 2ψ2 :
2
e2
Z Z
− d4x1 d4x2 : ψ 2A/ 2ψ2ψ 1A/ 1ψ1 : (64)
2
From the symmetry under 1 ↔ 2 above, it should be obvious that these
two terms are equal and that we need only evaluate one of these two terms
with a factor of −e2 — let us take the 1st term to evaluate.
−
• In this 1st term we must employ the particular fields components ψ 1 3 c†
and ψ2+ 3 c. We also need one A+ 3 a and one A− 3 a†, leaving:
Z Z
−
S (2)
3 −e 2
d x1 d x2 : ψ 1 A/ −
4 4
1 ψ1 ψ 2 A/+ +
2 ψ2 :
Z Z
− − +
−e2 d4x1 d4x2 : ψ 1 A/ + ψ
1 1 2ψ A/ 2 ψ2 : (65)

• Now we need a few intermediate results related to the “killing” process.
Our first example is:
X 1
ψ +
(x)c†r (~
p)|0i = p ct(~ q )e−iq·xc†r (~
q )ut(~ p)|0i
q
~,t
2V Eq~
1
= p p)e−ip·x|0i
ur (~ (66)
2V Ep~
q ), c†r (~
after using [ct(~ p)]+ = δrtδq~,~p.
Similarly, we find
− 1
c†t (~
X
0 0
p )ψ (x) = h0|cr0 (~
h0|cr0 (~ p ) p q )e+iq·x
q )ut(~
q
~,t
2V Eq~
1 0 +ip0 ·x
= p ur0 (~
p )e h0| . (67)
2V Ep~ 0

Finally, we have
X 1
A µ+
(x)a†s(~
k)|0i = p µ
t (~
q )at (~
q )e−iq·x † ~
as(k)|0i
t,~
q
2V |~
q|
1 ~ −ik·x|0i ,
= q µ
s (k)e (68)
2V |~
k|
and
X 1 †
~0
h0|as0 (k )A µ−
(x) = h0|as0 (k ) ~0 p µ
t (~
q )a q )e+iq·x
t (~
t,~
q
2V |~
q|
1 0 ik0 ·x
= q µ
s0 (~k )e h0| . (69)
2V |~
k0|
So let us use these results to first evaluate the contribution of the first term
in the part of S (2) that appears in Eq. (65). Using h0|0i = 1, we obtain,
dropping temporarily the √21V ... factors:

Z Z
−
h i
~0
h0|as0 (k )cr0 (p
~ ) −e
2 04
d x1
4
d x2 : ψ 1 A
−
/ 1 ψ1 ψ 2 A
+ + †
/ 2 ψ2 : cr (p
†
~)as (~
k)|0i
Z Z
−
h i
~ 0 0 2 4 4 − + + † †
= h0|as0 (k )cr0 (p
~ ) −e d x1 d x2 ψ 1 A/ 1 ψ1 ψ 2 A ~)as (~
/ 2 ψ2 cr (p k)|0i
Z Z
−
h i
~ 0 0 2 4 4 µ− ν+ + † †
= h0|as0 (k )cr0 (p
~ ) −e d x1 d x2 ψ 1 γµ A1 ψ1 ψ 2 γν A2 ψ2 cr (p ~)as (~k)|0i
ix1 ·(k0 +p0 )−ix2 ·(k+p)

h Z Z i
0 µ 0 2 4 4 ν
= ~ )γµ 0 (~
ur0 (p k ) −e d x1 d x2 ψ1 ψ 2 e γν s (~
k )ur ( p
~)
s
ix1 ·(k0 +p0 )−ix2 ·(k+p)

h Z Z i
0 ~ 0 2 4 4
= ur0 (p
~ )/ s0 (k ) −e d x1 d x2 ψ1 ψ 2 e s (~
/ k )ur ( p
~)

i 4 q/ + m −iq·(x1 −x2 )
h Z Z
0 0 2 4 4
= ur0 (p
~ )/ s0 (~
k ) −e d x1 d x2 d q 2 e
(2π )4 q − m2
ix ·(k0 +p0 )−ix2 ·(k+p)
i
e 1 s (~
/ k)ur (p
~)
i 4 q/ +m
h Z i
0 ~0 2 4 4 0 0 4 4
= ur0 (p s0 (k ) −e
~ )/ d q 2 (2π ) δ (k + p − q )(2π ) δ (q − p − k)
(2π )4 q − m2
s (~
/ k)ur (p
~)

2 4 4 0 0 q
/ +m 0 ~0
= −e (2π ) δ (p + k − p − k)ur0 (p
~ )/
s 0 (k ) i 2 2
s (~
/ k)ur (p
~)
q − m q =p +k
1
×q q q q (70)
2V Ep
~ 2V Ep
~0 2V |~
k| 2V |~
k0 |

where we have extended the d4x1 and d4x2 integrals over all space and
time, but have temporarily kept the volume normalization for the external
factors (writing them explicitly in the last line of the equation after having
dropped them earlier). This latter will be convenient when it comes time
to compute a cross section.
Of course, we could have kept finite volume in all the non-external stuff
as well and arrived at an exactly analogous answer. The integral over
continuous q ~ values traces all the way back to QFT-I where we wrote for
the scalar field case
+ − 1 X 1 ~ † ~0 −ik·x+ik0 ·y
[φ (x), φ (y)] = √ [a(k), a (k )]e
2V ~
ω~kω~k0
k~k0
1 d3~
k
Z
V →∞
→ e−ik·(x−y)
2(2π)3 ω~k
≡ i∆+(x − y) . (71)
1 1 −ik·(x−y )
We could have simply written i∆+(x − y) =
P
2V k ω~ e
~ . When
k
0
we reexpressed as a contour integral, the dk would still have been a
3~
continuous integral, but the (2dπk)3 would have been replaced by V1 ~k . And
P

R 3
R
in the algebra on the preceding page, the d x1 d3x2 would yield
Z Z
3 3 ix1 ·(k0 +p0 )−ix2 ·(k+p)−iq·(x1 −x2 ) 2
d x1 d x2e = V δ~k 0+p~ 0,~q δq~,~p+~k , (72)
(the dx01dx02 would still be over all time and you would have Dirac delta
functions from these integrals still). The final stages of the calculation
would have then been:
ix1 ·(k0 +p0 )−ix2 ·(k+p)
h Z Z i
0 ~0
2 4 4
ur0 (p
~ )/ s0 (k ) −e d x1 d x2 ψ1 ψ 2 e s (~
/ k)ur (p
~)
 
i q
/ + m −iq·(x1 −x2 ) 
h Z  Z
0 0 2 4 4 0
X
= ur0 (p
~ )/ ~
s0 (k ) −e d x1 d x2 dq e
 (2π )V q 2 − m2 
q
~
ix1 ·(k0 +p0 )−ix2 ·(k+p)

i
e s (~
/ k)ur (p
~)
i X q
/ +m
h Z i
0 2~0 0 0 0 0 0
= ur0 (p s0 (k ) −e
~ )/ dq (2π )δ ([k + p − q ] )(2π )δ ([q − p − k] )
(2π )V q 2 − m2
q
~
n o
2
× V δ~k 0 +p
~ 0 ,~
δ~,~
q q p+~
k
s (~
/ k)ur (p
~)

2 0 0 0 q
/ +m 0 ~0
= −e (2π )V δ ([p + k − p − k] )δp
~ 0 +~
k 0 ,~
p+~
k
ur 0 ( p
~ )/
s 0 (k ) i 2 2
s (~
/ k)ur (p
~)
q − m q =p +k
1
×q q q q (73)
2V Ep
~ 2V Ep
~0 2V |~
k| 2V |~
k0 |

This shows that the infinite volume limit corresponds to
a − ~b) .
V δ~a,~b → (2π)3δ(~ (74)
In any case, putting aside the normalization factors and exact version
of three momentum conservation delta function, the structure we have
obtained makes a lot of sense. We observe an over all coupling strength
of e2, overall 4-momentum conservation δ functions, and a very specific
algebraic structure that can be associated with a momentum-space Feynman
diagram as shown in the first diagram of the figure.

/0 1 3456
1 /(0
- .

!

8
5 19

. -
/20 7 /(0
"#$&%('*))+, /#5$&%'*))+:,
Figure 1: The two Feynman diagrams contributing to e−γ → e−γ.

The 2nd term of Eq. (65) gives a similar result. We just have to be a little
more careful with the Dirac structure. We have
Z Z
−
h i
~0
h0|as0 (k )cr0 (p
0
~ ) −e
2 4
d x1
4
d x2 : ψ 1 A
+
/ 1 ψ1 ψ 2 A
− + † †
~)as (~
/ 2 ψ2 : cr (p k)|0i
Z Z
−
h i
0 0 2 4 4 ν + µ − + † †
= h0|as0 (~ ~ ) −e
k )cr 0 ( p d x1 d x2 : ψ 1 γν A1 ψ1 ψ 2 γµ A2 ψ2 : cr (p ~)as (~
k)|0i
Z Z
µ− −
h i
~ 0 0 2 4 4 + ν+ † †
= h0|as0 (k )cr0 (p ~ ) −e d x1 d x2 A2 ψ 1 γν ψ1 ψ 2 γµ ψ2 A1 ~)as (~
cr (p k)|0i
ix2 ·k0 +ix1 ·p0 −ix1 ·k−ix2 ·p

h Z Z i
µ 0 0 2 4 4 ν
= s0 (~ ~ )γν −e
k )ur0 (p d x1 d x2 ψ1 ψ 2 e ~)s (~
γµ ur (p k)

i q
/ + m
h Z Z
0 2 4 4 4 −iq· ( x −x )
= ur0 (p s (~
~ )/ k) −e d x1 d x2 4
d q 2 2
e 1 2
(2π ) q −m
ix2 ·k0 +ix1 ·p0 −ix1 ·k−ix2 ·p
i
0
e s0 (~
/ k )ur (p~)
i 4 q/ +m
h Z i
0 ~ 2 4 4 0 4 4 0
= ur0 (p s (k) −e
~ )/ d q 2 (2π ) δ (p − k − q )(2π ) δ (q + k − p)
(2π )4 q − m2
0
s0 (~
/ k )ur ( p
~)

2 4 4 0 0 q
/ + m 0 0
= −e (2π ) δ (p + k − p − k)ur0 (p s (~
~ )/ k) i 2 /
s 0 (~k )ur (p
~)
q − m2 q =p−k0
1
×q q q q (75)
2V Ep
~ 2V Ep
~0 2V |~
k| 2V |~
k0 |

• So, the rules that you would use to get the two contributions to hf |S|ii
are the following.
1. Draw the diagrams involving two external photons, a final e− and an
initial e− (the latter connected by a continuously uni-directional e−-
fermion flow line) and only e− → e−γ “vertices” as specified by the form
of HI and having no loops.
You should quickly understand that you must have exactly two vertices
and that there are only the two topologically distinct (including the arrow
on the electron line) diagrams that we have drawn.
2. Now write down an algebraic expression for each of the two diagrams
using the following rules. For diagram 1 we do the following.
(a) For the outgoing e−(p0, r 0) write on the far left of the Dirac structure
a ur0 (~ p 0).
(b) With the outgoing external photon associate a polarization vector
µ
s 0 (~k 0
) (in a complex polarization basis this would actually be ∗
).
(c) Attach the outgoing photon µ s 0 (~k 0
) to a ieγµ written just to the right
of the ur0 (~ p 0).
(d) Then write down the momentum space virtual e− propagator
q/ + m i
i = (76)
q2 − m2 + i q/ − m + i

with the momentum q carried by the virtual e− given by momentum
conservation, i.e. q = p + k for the first diagram.
(e) Then comes the ieγν to which we attach the incoming photon νs (~ k).
(f) Finally write down to the far right of the Dirac structure the spinor
p) associated with the incoming e−.
ur (~
3. For the 2nd diagram, we follow a very similar procedure except for the
following.
(a) We first attach the incoming νs (~ k) to a ieγν just to the right of the
p 0).
ur0 (~
(b) We employ the e− propagator using the appropriate momentum
conservation value of q = p − k0.
(c) We then attach µ s 0 (~k 0
) for the outgoing photon to a ieγµ written to
the right of the Feynman propagator.
(d) We finish the Dirac structure with the ur (~ p).
4. These rules define the amplitudes (denoted Ma and Mb in MS).

q/ + m
p 0)/ s0 (~
Ma = −e2ur0 (~ k0) i / s(~
k)ur (~
p) (77)
q2 − m2 q = p+ k

q/ + m
p 0)/ s(~
Mb = −e2ur0 (~ k) i 2 2

/ s 0 (~
k 0
)ur (~
p) (78)
q − m q=p−k0

The full amplitude is the sum M = Ma + Mb, and is to be multiplied
by (2π)4δ 4(p0 + k0 − p − k), the overall
√ momentum conservation delta
function, and the product of the 1/ 2V E normalization factors.
• One of the key ingredients above is the ieγµ “vertex”. Algebraically, you
can think of this factor as coming from:
1. −i for each power of HI appearing in the required level of S-matrix
expansion;
2. −eγµ as appearing in each HI in association with constructing the A/
appearing therein.
This vertex is associated with each interaction of a photon and a fermion
whether these are real or virtual.
In the above case, the initial or final e− is real, the photons are real, but
the intermediate e− is virtual.
Regardless, one writes down (assuming incoming or outgoing electrons, and
not positrons)
ur (p)ieγµus(q)µ
t (k) (79)
for each vertex, even when one of the particles is virtual.

In addition, each internal virtual propagator is then associated with the
i/(q 2 − m2) factor. We will shortly understand why I have left out the
q/ + m part.
In the above process, this procedure works as follows, using the example of
diagram 1.
1. We have a µ 0
s0 (k )ur (~
0
0 p )ieγµ ut (q) for the upper vertex. (I am free to
write the , which is just a number, anywhere I want.)

2. We have a ut(q)ieγν ur (~ p)νs (~
k) for the lower vertex.
3. We have the i/(q 2 − m2) factor with q = p + k by momentum
conservation.
4. We sumP over possible intermediate spin states t for the virtual electron.
5. We use t ut(q)ut(q) = q/ + m which is true even if we allow the q to
have an off-shell energy.
6. The result,
X i
µ
s 0 (~k 0
)ur 0 (~
p 0
)ieγµut(q) ut (q)ieγ ν ur (~
p) ν ~
s (k)
t
q2 − m 2

q/ + m
= µ
s 0 (~k 0
)ur 0 (~
p 0
)ieγµi p)νs (~
ieγν ur (~ k)
q 2 − m2 q =p+k=p0 +k0

q/ + m
p 0)ie/ s0 (~
= ur0 (~ k0)i ie/ s(~
k)ur (~
p) , (80)
q2 − m2 q =p+k=p0 +k0
is precisely what we had before.
• There are a few more rules that we can only get by considering a few other
processes. For the rules for external positrons, we need the appropriate
“killing” operations. For an incoming positron contained in |ii we have:
+ X 1
ψ (x)d†r (~
p)|0i = p dt(~ q )e−iq·xd†r (~
q )v t(~ p)|0i
q
~,t
2V Eq~
1
= p p)e−ip·x|0i
v r (~ (81)
2V Ep~
q ), d†r (~
after using [dt(~ p)]+ = δrtδq~,~p.
Similarly, for an outgoing positron in |f i, we find
1
d†t (~
X
0 − 0
h0|dr0 (~
p )ψ (x) = h0|dr0 (~
p ) p q )e+iq·x
q )vt(~
q
~,t
2V Eq~

1 0 +ip0 ·x
= p vr0 (~
p )e h0| . (82)
2V Ep~ 0
Note how these differ from the corresponding electron results. Where we
have a u(~ p) for an incoming electron we find v(~ p) for an incoming positron,
and where we had u(~ p) for an outgoing electron, we find v(~ p) for an
outgoing positron. As a result, the v and u spinors both appear at the end
of the line directed in the direction of electron (i.e. fermion vs. antifermion)
flow. For instance, an incoming positron has the arrow of fermion flow
directed in the outgoing direction, and one writes v.
• As an example of how to use the above results, let us consider e+(p)γ(k) →

e+(p0)γ(k0) scattering. There would be the same two types of contributions,
but we would have to focus on different contributions to the S matrix. One
contribution would be:
Z Z
+
h h i i
~0 0
h0|as0 (k )dr0 (p
~ ) −e
2 4
d x1
4
d x2 : ψ 1 α A
−
/ 1 ψ1 ψ 2 A
/2
+ − †
ψ2 β : dr (p
†
~)as (~
k)|0i
αβ
Z Z
+
h h i i
~ 0 0 2 4 4 − − + † †
= h0|as0 (k )dr0 (p
~ ) +e d x1 d x2 ψ2 β A / 1 ψ1 ψ 2 A
/2 ~)as (~
ψ 1 α dr ( p k)|0i
αβ
Z Z
+
h h i i
~ 0 0 2 4 4 − µ− ν+ † †
= h0|as0 (k )dr0 (p
~ ) +e d x1 d x2 ψ2 β γµ A1 ψ1 ψ 2 γν A2 ψ1 α ~)as (~
dr ( p k)|0i
αβ

ix2 ·p0 +ix1 ·k0 −ix1 ·p−ix2 ·k
h h Z Z i i
0 µ 0 2 4 4 ν ~
= ~ )β γµ 0 (~
vr0 (p k ) +e d x1 d x2 ψ1 ψ 2 e γν s (k) v r (p
~)α
s αβ
ix2 ·p0 +ix1 ·k0 −ix1 ·p−ix2 ·k

h Z Z i
= v r (p
~)/ ~0
s0 (k ) +e
2 4
d x1
4
d x2 ψ1 ψ 2 e s (~
/ k)vr0 (p
0
~ )

i q
/ + m
h Z Z
0 2 4 4 4 −iq· ( x −x )
= v r (p
~)/ s0 (~
k ) +e d x1 d x2 d q 2 e 1 2
(2π )4 q − m2
ix2 ·p0 +ix1 ·k0 −ix1 ·p−ix2 ·k
i
0
e s (~
/ k)vr0 (p
~ )
i 4 q/ +m
h Z i
= v r (p
~)/ ~0
s0 (k ) +e
2
d q 2
4 4 0 4 4 0
(2π ) δ (k − p − q )(2π ) δ (q + p − k)
(2π )4 q − m2
0
s (~
/ k)vr0 (p
~ )

2 4 4 0 0 0 q
/ + m 0
= +e (2π ) δ (p + k − p − k)v r (p s0 (~
~)/ k) i 2 2
s (~
/ k)vr0 (p
~ )
q − m q =k0 −p
1
×q q q q (83)
2V Ep
~ 2V Ep
~0 2V |~
k| 2V |~
k0 |
Note how, in going from the 1st line to the 2nd line above, I moved the
+
ψ 1 α all the way to the right, past the [. . .]αβ matrix and past ψ2−β . This
is allowed since I retain proper Dirac matrix contractions so long as I keep
track of the α index sums later, as I do. A similar statement applies to
moving the ψ2−β over to the left.
The final expression above can be mapped to the second of the Feynman

diagrams shown in the figure below.

021 3 4
091 3
. /

"!
5 "!

/ .
01 5 6 0:1 3;
#$%'&)(+**,- 7$%'&)(8**,-
Figure 2: The two Feynman diagrams contributing to e+γ → e+γ.
This figure will be confusing at first. The arrows show the direction of
fermion flow. But we must remember that the incoming (antifermion)
e+ has momentum p entering the diagram from the bottom right, and
the outgoing e+ has momentum p0 exiting the diagram. I.e. anti-
fermion momenta and fermion flow are oppositely directed. Momentum
conservation for the virtual propagator is correctly given (for the direction
of fermion flow) by k0 − p for the 2nd diagram. That is, we always write
the Feynman propagator in the canonical form where q is given by the
momentum carried in the fermionic direction, rather than the antifermionic
direction.
I will not work out the details of the 2nd contribution, which gives you the
first (left-hand) diagram of the above figure. Using our Feynman rules we

will get

2 4 4 0 0 q
/ + m 0 0
+e (2π ) δ (p + k − p − k)v r (p s (~
~)/ k) i 2 2
s0 (~
/ k )vr0 (p
~ )
q − m q =−k−p
1
×q q q q . (84)
~ ~ 0
2V Ep ~ 0 2V |k| 2V |k |
~ 2V Ep
A homework problem will be to verify this result in the same kind of detail
as given for the right-hand diagram.
• One important note on the above. We were careful to keep the sign
resulting from strict application of Wick’s theorem, which ended up giving
us a +e2 overall sign for this positron case as opposed to the −e2 overall
sign for the electron case.
Such an overall sign is not actually observable, since we will eventually
square these amplitudes.
• However, one should be very careful with signs in cases where different
diagrams correspond to the exchange of identical fermions in the final state
An example of this, which is also useful for getting the correct result for a
virtual photon propagator, is provided by e−e− → e−e− scattering.

For this process, we need two ψ fields and two ψ fields in order “kill”
the creation operators defining the incoming and outgoing states. More
+ −
precisely, we will need 2 ψ ’s and 2 ψ ’s. This means we are again dealing
with S (2) the 2nd order term in the S-matrix expansion, which I repeat
below. Note that since each of the two HI ’s required contains an A field,
we will have to contract these two A fields against one another in order to
generate a non-zero contribution to the scattering process of interest. The
relevant stuff is thus:
2Z
e
Z
S (2) = − d4x1 d4x2T {: ψ 1A/ 1ψ1 : : ψ 2A/ 2ψ2 :}
2
e2
Z Z
− ν −
3 − d4x1 d4x2 : ψ 1 α(γµ)αβ ψ1+β Aµ 1 A ψ
2 2 δ (γ ν ) +
δρ 2 ρ :
ψ
2
2Z
e
Z
− −
= − d4x1 d4x2ψ 2 δ ψ 1 α(γµ)αβ Aµ 1 A ν
2 (γ ν ) + +
δρ 1 β 2 ρ , (85)
ψ ψ
2
where two changes of sign occurred when we passed two fermionic operators
past one another in order to get to the explicitly normal ordered form given.
We must now adopt a convention for our |ii and |f i states. We use p, r
and k, s for the incoming guys and p0, r 0 and k0, s0 for the outgoing guys

so that we define
p)c†s(~
|ii = c†r (~ k)|0i , |f i = c†r0 (~
p 0)c†s0 (~
k0)|0i ,
⇒ hf | = h0|cs0 (~ p 0) ,
k0)cr0 (~ (86)
where the precise order of the c† operators must be carefully adhered to

because of the minus signs from various anticommutators.
Now, let us consider the structure
p)c†s(~
ψ2+ρc†r (~ k)|0i
X 1
= p ct(~l)ut(~l)ρe−il·x2 c†r (~
p)c†s(~
k)|0i (87)
~
2V E~l
t,l
Contained within this sequence is the structure

h i
p)c†s(~
ct(~l)c†r (~ k)|0i = δtr δ~l~pc†s(~ p)ct(~l)c†s(~
k) − c†r (~ k) |0i
h i
= δtr δ~l~pc†s(~ k) − c†r (~
p)δtsδ~l~k |0i . (88)

If we now bring into play the
X 1
ψ1+β = p cu(~ q )β e−iq·x1
q )uu(~ (89)
u,~
q
2V Eq~
we will need the structure

h i
q )ct(~l)c†r (~
cu(~ p)c†s(~ q ) δtr δ~l~pc†s(~
k)|0i = cu(~ k) − c†r (~
p)δtsδ~l~k |0i
h i
= δtr δ~l~pδusδq~~k − δur δq~p~δtsδ~l~k |0i . (90)
Obviously, the two terms differ by the interchange of t, ~l ↔ u, q ~ or,

~ ↔ s, ~
equivalently, r, p k and the minus sign which is keeping track of the
fermionic statistics.
P P
Using these δ functions to eliminate the u,~q and t,~l yields
1 h
p p e−ik·x1−ip·x2 us(~
k)β ur (~
p)ρ
2V Ep~ 2V E~k
i
−e −ip·x1 −ik·x2
p)β us(~
ur (~ k)ρ (91)

~ ↔ s, ~
These clearly differ by the interchange r, p k and the fermionic
statistics sign.
We would now have to carry out an exactly equivalent game for the left
hand part of the expression in Eq. (85). Without giving the details, I hope
it is relatively straightforward to understand the result. One simply has to
switch everything to primes and do the barring (which includes a complex
conjugation of the exponential factors and u → u) and change the Dirac
indices appropriately (β → α and ρ → δ):
− −
h0|cs0 (~
k0)cr0 (~
p 0)ψ 2 δ ψ 1 α
1 h
+ik0 ·x1 +ip0 ·x2
= p p e us0 (~
k0)αur0 (~
p 0)δ
2V Ep~ 0 2V E~k0
0 0
i
p 0)αus0 (~
− e+ip ·x1+ik ·x2 ur0 (~ k 0 )δ . (92)
Note the antisymmetry under r 0, p ~ 0 ↔ s0 , ~k 0 corresponding to Fermi

statistics. This and the initial-state antisymmetry under r, p ~ ↔ s, ~ k
guarantee that we get zero if the initial or final electrons are in exactly the
same physical state.

We now combine the results of Eqs. (91) and (92) with the remainder of
the stuff in Eq. (85) to obtain the result
−e2 1
Z Z
(2) 4 4
hf |S |ii = q q q q d x1 d x2 ×
2 2V Ep~ 0 2V E~ k0
2V Ep~ 2V E~ k
" #
0 0
+ik ·x1 +ip ·x2 0 0 0 0
+ip ·x1 +ik ·x2 0 0
e us0 (~ ~ )δ − e
k )α ur0 (p ~ )α us0 (~
ur 0 ( p k )δ
µν
−iq·(x1 −x2 ) −g

i
Z
4
(γµ )αβ d qe (γν )δρ
(2π )4 q 2 + i
" #
−ik·x1 −ip·x2 −ip·x1 −ik·x2
e us (~ ~)ρ − e
k)β ur (p ~)β us (~
ur (p k )ρ (93)
where we inserted the Fourier representation of Aµ

1 A ν
2 = iD µν
F (x1 − x2 ).
Of the four terms in the above expression, two are equal to corresponding
ones of the remaining two. This will cancel the 21 appearing in e2/2. This
always happens in QED. We saw another example of this in our earlier
calculations. There is equivalence under interchange of the x1 and x2
vertices (i.e under H1 ↔ H2). Let us focus on two inequivalent terms. We
begin with (dropping some external factors for the moment) taking the 1st

term in each of the large brackets of the above expression:
" #
+ik0 ·x1 +ip0 ·x2
Z Z
4 4 0 0
d x1 d x2 e us0 (~ ~ )δ (γµ )αβ
k )α ur 0 ( p
" #
µν
−iq·(x1 −x2 ) −g

i
Z
4 −ik·x1 −ip·x2
d qe (γν )δρ e us (~
k )β ur ( p
~)ρ
(2π )4 q 2 + i
µν
−ig

4 4 0 0 ~0 0
= (2π ) δ (p + k − p − k )us0 (k )γµ us (~
k) ur 0 ( p
~ )γν ur (p
~) (94)
q 2 + i q =k0 −k=p−p0
where we did the d4x1 and d4x2 integrals to get (2π)4δ 4(k0 − q − k)
and (2π)4δ 4(p0 + q − p) respectively and used one of these to do the d4q
integral. Now, we take the 2nd term of the 1st large bracket and the 1st
term of the 2nd large bracket. We have
" #
+ip0 ·x1 +ik0 ·x2
Z Z
4 4 0 0
d x1 d x2 −e ~ )α us0 (~
ur0 (p k )δ (γµ )αβ
" #
µν
−iq·(x1 −x2 ) −g

i
Z
4 −ik·x1 −ip·x2
d qe (γν )δρ e us (~
k)β ur (p
~)ρ
(2π )4 q 2 + i
µν
−ig

4 4 0 0 0 0
= ~ )γµ us (~
−(2π ) δ (p + k − p − k )ur0 (p k) us 0 (~k )γν ur (p
~)(95)
q 2 + i q =p0 −k=p−k0

where we did the d4x1 and d4x2 integrals to get (2π)4δ 4(p0 − q − k)
and (2π)4δ 4(k0 + q − p) respectively and used one of these to do the
d4q integral. A quick inspection of Eq. (94) compared to Eq. (95) shows
that the relation corresponds to interchanging the r 0, p
~ 0 final state e− with
the s0, ~
k0 final state e− and introducing a relative minus sign between the
two contributions. The remaining two contributions obtained by using the
2nd term of the 2nd large bracket in Eq. (93) simply duplicate the two
contributions that we have already obtained. Thus, after extracting the
standard (2π)4δ 4(p0 + k0 − p − k) overall
√ momentum conservation factor
with its (2π)4 factor and the four 1/ 2V E factors, we end up with the
two invariant amplitudes:
µν
−ig

Ma = −e2us0 (~ k0)γµus(~k) 2 ur0 (~p 0)γν ur (~
p)
q + i q=k0−k=p−p0
µν
−ig

Mb = +e2ur0 (~ p 0)γµus(~
k) 2 us0 (~
k0)γν ur (~p)(96)
q + i q=p0−k=p−k0
the factor of 21 having been canceled by the above-noted factor of 2

duplication. The relative − sign becomes part of our Feynman rules:
– Whenever there are two contributing diagrams that are related to one

another by interchanging fermions of the same type, one writes the same
type of expression for each but introduces a relative minus sign.
We note that while the relative sign between Ma and Mb is determined, the
signs of both could be switched without any physical impact in computing
probabilities (related to |Ma + Mb|2). The overall sign is pure convention,
related to how we defined |ii and |f i. Had we reordered the c† operators
in the definition of one of these two states, both Ma and Mb would have
changed sign.
The Feynman diagrams associated with Ma and Mb are shown in the
figure below.

./'01# 40 .
67

( ) !#"%)
$'*#&+-, ( ) !9#"%$8*#&+-,
.2'03 5 .:403
Figure 3: The two Feynman diagrams contributing to e−e− → e−e−.
• So, by combining the results for the various processes we have computed

we obtain an almost complete set of Feynman rules for QED.
1. Draw all possible tree-level (lowest possible order) diagrams that can give
rise to the process of interest making use of just the fermion → fermion
+ photon “vertex”.
In so doing, keep track of the fermion line direction. An incoming
fermion line must be traced continuously to an outgoing fermion line (or,
equivalently, to an incoming antifermion line).
2. For each vertex, write ieγµ, where a different Lorentz index µ must be
used for each vertex and this µ must be connected either to the µ on a
virtual photon propagator or to the µ of an external photon polarization
vector.
3. For an outgoing photon γ, p, s, write ∗s (~ p), where I here allow for a
complex polarization basis.
4. For an incoming photon γ, p, s write s(~ p).
5. For an incoming e−, p, r write ur (~ p).
6. For an outgoing e−, p, r write ur (~p).
7. For an incoming e+, p, r write v r (~
p).
8. For an outgoing e+, p, r write vr (~p).
9. For the positron cases, note that the momentum direction will be opposite
the fermion arrow direction.

10. For an internal (virtual) electron propagator write
q/ + m
i (97)
q2 − m2 + i
where q is the momentum in the direction of the fermion arrow and the
value of q is obtained by momentum conservation at the vertices.
11. For an internal photon propagator, write
−g µν
i (98)
q2 + i
where q can be chosen in any direction and is given by momentum

conservation at the vertices.
12. A fully contracted Dirac structure should be constructed for each fermion
line, beginning with a u or v to the far left of the expression and working
back against the fermion arrow flow direction until terminating on a u or
v spinor.
Along the way, one will encounter first a vertex and then, possibly, an
internal fermion line, and then, possibly, another vertex, and so forth
until ending with a vertex and then the final u or v.

13. If there are diagrams that differ by simply interchanging:
– two initial e−’s;
– or two initial e+’s;
– or two final state e−’s;
– or two final state e+’s;
– or an initial e+ with a final e−;
– or a final e+ with an initial e−;
then a relative minus sign should be introduced between the corresponding
Feynman amplitudes.
14. For each closed loop, there will be an integration over an unconstrained
1
R 4
momentum, call it qi. One should perform the integral (2π)4 d qi for
each such unconstrained loop momentum.
If the closed loop is a continuous fermion line, an explicit minus sign
should be introduced.
• A note on quickly checking the sign of the vertex. We began with L =

ψiD/ ψ with Dµ = ∂µ + iqAµ to obtain LI = −qψA/ ψ. The convention
for QED is that q = −e so we end up with LI = eψA/ ψ = −HI . The
evolution operator appearing in the S matrix of the interaction picture is
basically e−iHI t which implies that S contains (−iHI )n powers. The lowest

order is just −iHI = +ieψA/ ψ. If you remove the external ψ, ψ, Aµ fields
the corresponding vertex in Dirac space is just +ieγµ.
• It is the last of the Feynman rules above that we have yet to derive using
an explicit example.
We choose to use the virtual e− loop correction to the photon propagator
as our basic example. The basic Feynman diagram that corresponds to
our calculation is below. There, I use the label p for the free momentum
associated with the loop.

"! !

Figure 4: The e− loop correction to the photon propagator.

The computation begins with |ii = a†r (k)|0i and ends with |f i = a†r (k)|0i
and contains two vertices, so that means we are looking at S (2). We then

have
" #
2
e
Z Z n o
4 4 µ ν †
h0|ar (~
k) − d x1 d x2 T : ψ 1 γµ A1 ψ1 :: ψ 2 γν A2 ψ2 : ar (~
k)|0i
2
(99)
To reduce this expression, we must employ Wick’s theorem. In this case,

ψ1 must contract with ψ 2 and vice versa. (Recall: no contractions between
fields in same HI .) In getting to the contraction configuration given below,
I must do an odd number of fermion field interchanges — hence the very
crucial extra − sign! Keeping in mind that we must have one A+ and one
A− in order to kill the a and a† operators, the above reduces to (using the
1st line to simply display all the Dirac indices of the previous form above,
and the 2nd line to display the particular contraction term in Wicks theorem
that will survive as well as the two different choices as to which A is A+
and which is A− – we need one of each and there are two possibilities)
" #
2 Z
e
Z n o
4 4 µ ν †
h0|ar (~
k) − d x1 d x2 T : ψ 1 α (γµ )αβ A1 ψ1 β :: ψ 2 ρ (γν )ρσ A2 ψ2 σ : ar (~k)|0i
2
" #
2 Z
e
Z
4 4 µ+ ν− †
= h0|ar (~
k) + d x1 d x2 : ψ2 σ ψ 1 α (γµ )αβ A1 ψ1 β ψ 2 ρ (γν )ρσ A2 : ar (~ k)|0i
2
" #
2 Z
e
Z
4 4 µ− ν+ †
+h0|ar (~k) + d x1 d x2 : ψ2 σ ψ 1 α (γµ )αβ A1 ψ1 β ψ 2 ρ (γν )ρσ A2 : ar (~ k)|0i
2

" #
2
e
Z Z
4 4 ν− µ+ †
= d x1 d x2 h0|ar (~k)A2 + ψ2 σ ψ 1 α (γµ )αβ ψ1 β ψ 2 ρ (γν )ρσ A1 ar (~ k)|0i
2
" #
2
e
Z Z
4 4 µ− ν+ †
+ d x1 d x2 h0|ar (~
k)A1 + ψ2 σ ψ 1 α (γµ )αβ ψ1 β ψ 2 ρ (γν )ρσ A2 ar (~ k)|0i
2
" #
2
e
Z Z
4 4 ν ik·x2 µ −ik·x1
= d x1 d x2 r (~
k )e + ψ2 σ ψ 1 α (γµ )αβ ψ1 β ψ 2 ρ (γν )ρσ r (~ k )e
2
" #
2
e
Z Z
4 4 µ ik·x1 ν −ik·x2
+ d x1 d x2 r (~
k)e + ψ2 σ ψ 1 α (γµ )αβ ψ1 β ψ 2 ρ (γν )ρσ r (~ k )e
2
(100)
where I used the standard formulae for A+a†|0i and h0|aA−:
µ+ † 1 µ −ik·x
A (x)as (~
k)|0i = q s (~
k )e |0i ,
2V |~
k|
µ− 1 µ ik·x
h0|as (~
k )A (x ) = q s (~
k )e h0| . (101)
2V |~
k|
q
derived earlier, dropping the two 1/ 2V |~k| factors. If I interchange the x1
and x2 variables and relabel dummy indices (µ ↔ ν, α ↔ ρ and β ↔ σ),
then it is easy to see that the 2nd term is identical to the first term.

We thus need only evaluate
Z Z
4 4 ν ik·x2 2 µ −ik·x1
d x1 d x2 r (~
k )e +e ψ2 σ ψ 1 α (γµ )αβ ψ1 β ψ 2 ρ (γν )ρσ r (~
k)e


i q
/ +m
Z Z Z
4 4 ν ik·x2  2 4 −iq·(x2 −x1 )
= d x1 d x2 r (~
k )e +e d qe (γµ )αβ
(2π )4 q 2 − m2 σα

" #
i p
/ +m
Z
4 −ip·(x1 −x2 )  µ ~ −ik·x1
d pe ( γ ν ) ρσ  r (k )e
(2π )4 p2 − m2 βρ


i q
/ +m
Z
ν 2 4
= r (~
k ) + e d q (γµ )αβ

(2π )4 q 2 − m2 σα

" #
i p
/ +m
Z
4 4 4 4 4  µ ~
d p (2π ) δ ( k − q + p )(2 π ) δ ( q − p − k )( γ ν ) ρσ  r (k )
(2π )4 p2 − m2 βρ
 
d4 p 

q
/ +m p
/ +m
Z
4 4 2
= (2π ) δ (k − k)(+e ) i r (~
/ k )i r (~
/ k )

(2π )4 2 2
q − m q = p+ k p2 − m 2
σσ
(102)
q
After including the 1/ 2V |~
k| factors we can summarize our final result

as
~ (2) ~ 4 4 1
hγ(r, k)|S |γ(r, k)i = (2π) δ (k − k) q q M ,(103)
2V |~
k| 2V |~
k|
with
" #
4
d p q/ + m p/ + m
Z
M = +e2 Tr i
/ r (~k)i
/ r (~k)
(2π) 4 2 2
q − m q = p+ k 2
p −m 2
(104)
Note that this result has the opposite sign compared to what you would
write down just naively using ieγµ and ieγν for the two vertices and i qq/2−m
+m
2
and i pp/2−m
+m
2 for the internal fermion propagators.
Note also how the Dirac structure is written by starting to the left with
the p + k propagator and then working “backward” to the p propagator as
shown in the earlier figure.
In this final form, we have also used the Trace notation. The Dirac index
structure was such that one ended up with a continuous connection of

Dirac indices to one another as we moved along the fermion line, with the
result corresponding to the trace of the complicated Dirac matrix product
indicated.

Tree-level QED Processes
Computing a cross section
• We will consider colliding two particles in the initial state |ii with momenta
pi = (Ei, p
~i), i = 1, 2. (Sorry about the confusion of notation; I am simply
following MS here.) The final state |f i is assumed to contain N particles
with p0f = (Ef0 , p
~f0 ), f = 1, . . . N .
• The S-matrix, which always has an overall δ 4 momentum conservation

factor, can be written as
1/2 Y !1/2
X X Y 1 1
Sf i = δf i + (2π)4δ 4( p0f − pi) M
f i i
2V Ei f
2V Ef0
(105)
The δ 4 function in the above is obtained in the limit of an infinite time
interval T → ∞ and an infinite volume, V → ∞. For finite T and V , we

would have obtained the same expression with
X X lim X X
4 4
(2π) δ ( p0f − pi ) = T →∞ δT V ( p0f − pi)
f i V →∞ f i
 
lim Z +T /2 Z X X
≡ T →∞ dt 3
x exp ix · (
d ~ p0f − pi) (106)
V →∞ −T /2 V f i
P 0 P
replaced by δT V ( f pf − i pi). In deriving the cross section, it is helpful
to take T and V finite temporarily.
• For finite T and V , we can define the transition probability per unit of time
as
w = |Sf i|2/T (107)
h P 0 P i2
which will contain the factor δT V ( f pf − i pi) . For one of these
P 0 P
factors, we revert back to the continuum result δT V ( f pf − i pi) →
4 4
P 0 P P 0
(2π)
P δ ( f pf − i pi). For the other factor, we can take ( f pf −
i pi ) = 0 and use δT V (0) = T V , keeping the finite volume and time
interval.

• We then get
!
X X Y 1
Y
1
4 4
w = V (2π) δ ( p0f − pi) |M|2 . (108)
f i i
2V Ei f
2V Ef0
This is the result for transition from state i to a single final state f . It is
of course vanishingly small in the continuum limit of V → ∞ as the final
state possibilities become a continuum or possibilities.
In the V → ∞ limit, it only makes sense to compute the transition

probability to a group of states centered on some central state; one takes
pf0 , p
states in the interval (~ ~f0 + d~
pf0 ), f = 1, . . . , N . The number states in
such an interval is given by
~f0
Y V d3p
. (109)
f
(2π)3
The factors of V , above, cancel those in the denominator in the final state
part of the expression for w, leaving one V in the numerator from the δT V
and two V ’s in the denominator from the initial state killing operations.

• The differential cross section, dσ, is defined to be the transition rate into
this group of final states for one scattering center divided by the flux of
incident particles incident on the volume containing this one scattering
center.
With our choice of normalization for the states, the volume V which we
are considering contains just one scattering center. To show this (again),
just compute
h~
p|~ p)a†(~
pi = h0|a(~ p)|0i = δp~p~h0|0i = 1 , (110)
where we have used the finite volume result for the commutator,
p), a†(~
[a(~ p 0)] = δp~p~ 0 . (111)
And, the incident flux is vrel/V where vrel is the relative velocity of the
colliding particles. To see this, picture a box of volume V traveling with
velocity vrel towards a fixed (for convenience) box of volume V . Orient
the traveling box so that one face is parallel to the front face of the fixed
box. Start at t = 0 with the front face of the traveling box touching the
front face of the fixed box. If the facing faces of the two boxes have area

A, then the fraction of the traveling volume V that passes the front face
of the fixed box per unit time is vrelA/V . Since there is only one particle
in the volume V , this is also the number of particles that pass through the
front face of the fixed box (the number is fractional). The flux of particles
is the number of particles per unit time passing through per unit area, i.e.
it is this number divided by A, giving f lux = vrel/V .
Thus, we obtain
1 ~f0
Y V d3p
dσ = w
[ vVrel ] f
(2π)3
X X 1 Y ~f0
d3 p
= (2π)4δ 4( p0f − pi) 0
|M| 2
.(112)
f i
4E1E2vrel f
(2π)32Ef
Note that this differs from the result in MS (with the extra 2m factors
for fermions) by virtue of the different normalization for fermion fields and
operators. Using my normalizations (which are the same as essentially all
modern treatments), there is no distinction in the basic normalizations for
bosons and fermions.
~f0 are
Note that the δ 4 function above implies that not all of the p

independent. In any given situation, we will have to integrate out a
~f0 to eat up the δ 4 function.
certain number of the p
• The above result holds in any Lorentz fame in which the colliding particles
move collinearly.
In such a frame, and assuming the particles are moving in opposite direction
to one another (it is also ok if one is a rest), the relative velocity is given
by
|~ | |~ |

p1 p2
vrel =
+
E1 E2
1
= |E2|~p1| + E1|~
p2|| . (113)
E1E2
Let us compare this expression with (remember opposite directions for the
colliding particles are assumed)
2 2 2 2 2 2
(p 1 · p 2 ) − m 1 m 2 = (E1 E2 + |~
p1 ||~
p2 |) − m1 m2
2 2 2 2 2 2
= E1 E2 + |~
p1 | |~
p2 | + 2E1 E2 |~
p1 ||~
p2 | − m1 m2
2 2 2 2 2 2 2 2
= (|~
p1 | + m1 )(|~
p2 | + m2 ) + |~
p1 | |~
p2 | + 2E1 E2 |~
p1 ||~
p2 | − m1 m2

2 2 2 2 2 2
= 2|~
p1 | |~
p2 | + m1 |~
p2 | + m2 |~
p1 | + 2E1 E2 |~
p1 ||~
p2 |
2 2 2 2
= E1 |~
p2 | + E2 |~
p1 | + 2E1 E2 |~
p1 ||~
p2 |
2
= (E2 |~
p1 | + E1 |~
p2 |) (114)
In short, we have
2 2 2 1/2

E1E2vrel = (p1 · p2) − m1m2 . (115)
This last result, substituted in the general form of dσ, implies that dσ is a
d3 p
Lorentz invariant. (Recall that (2π)3~2E is a Lorentz invariant.)
p
~
• Some useful special cases are the com system, with |~

p1| = |~
p2| resulting in
E1E2vrel = |E2|~
p1| + E1|~
p2|| = |~
pcom|Etot , (116)
(where Etot ≡ E1 + E2) and the laboratory frame in which p

~2 = 0 and
E1E2vrel = |E2|~
p1| + E1|~
p2|| = |~
p1|m2 , (117)
• Since the center of mass system is so common and since we often consider
2 → 2 processes, it is useful to give some results specific to this case.

~20, the result that |~
We use Eq. (116) and, after eliminating p p10| = |~
p20| ≡
0
|~
pcom | to obtain the form
1 4 0 0 3 0 3 0 2
dσ = 0 0 δ ( p 1 + p 2 − p 1 − p 2 )d p
~1 d p
~2 |M|
pcom |Etot (2π )2 E1 E2
16|~
1 0 0 0 2 0 0 2
= 2 0 0 δ (E1 + E2 − Etot )|~ p1 | d|~ p1 |dΩ1 |M|
64π |~ pcom |Etot E1 E2
#−1
∂ (E10 + E20 )
"
1 0 2 0 2
= 0 0 0 |
|~
pcom | d Ω 1 |M|
64π 2 |~
pcom |Etot E1 E2 ∂|~
pcom
 q q − 1
1 ∂ (m01 )2 + |~ 0 |2 +
pcom (m02 )2 + |~ 0 |2
pcom
0 2 0 2
= |~
pcom | dΩ1 |M|
 
pcom |Etot E10 E20
64π 2 |~ 0
pcom |
 
∂|~
#−1
0 0
"
1 |~
pcom | |~
pcom | 0 2 0 2
= + |~
pcom | dΩ1 |M|
pcom |Etot E10 E20
64π 2 |~ E1 0 E2 0
0
1 |~
pcom | 0 2
= 2 0 0 dΩ1 |M|
64π Etot (E1 + E2 ) |~
pcom |
0
1 |~
pcom | 0 2
= 2 |~
d Ω1 |M| , (118)
64π 2 Etot pcom |

where in the last step we used E10 + E20 = Etot. Thus, our final result is
0
|~ |

dσ 1 pcom 2
0
= 2 2
|M| . (119)
dΩ1 com 64π Etot |~pcom|
• The above and other results for differential cross sections apply irregardless
of whether or not we are dealing with identical particles.
However, if there are two or more identical particles in the final state, we
must not duplicate integration regions in obtaining the total cross section.
For example, in the 2 → 2 process just computed, if the final particles
were 2 e−’s, then the case where (θ10 , φ01) = (α, β) is not physically
distinguishable from the case where (θ10 , φ01) = (π − α, π + β). In this
case, we should only integrate over 0 ≤ θ10 ≤ 12 π, or, equivalently, we could
integrate over all θ10 and then divide the final result by 2!:

1 dσ
Z
tot
σcom = dΩ01 0
. (120)
2 4π dΩ1 com
In the more general case of n identical particles in the final state, we can
obtain the correct result for σ tot by integrating over all of phase space and
then dividing by n!.

In a very rough sense, the 1/2! is partly compensating for the fact that for
two identical particles there will be two Feynman diagram amplitudes that
differ only by interchanging the two final particles (as in e−e− → e−e−),
which, again very roughly, means that the amplitude-squared would be 4
times as large as compared to the case where the final particles are not
identical (e.g. e−µ− → e−µ−).
(As we shall discuss, the e− and µ− are not identical particles. They carry
a lepton identifier quantum number that distinguishes them. See MS for
more details.)
• As an example of a particularly useful cross section, let us consider the

process e+(p)e−(k) → µ+(p0)µ−(k0).
We consider how to deal with two different types of leptons. As noted

above muons and electrons are experimentally distinguishable — e.g. they
have different mass and have a different lepton number.
What this means in our theoretical framework is that there are creation and
annihilation operators for electrons that are distinct from those for muons.

Each lepton has its own free-particle Lagrangian so that
X
L0 = ψ l(x)(i∂/ − ml)ψl(x) . (121)
l=e,µ,τ
Making the minimal substitution (with q = −e for each) gives

X
HI (x) = −LI (x) = −e : ψ l(x)A/ (x)ψl(x) : (122)
l
Note that in the above there is no term like
− e : ψ µ(x)A/ (x)ψe(x) : (123)
Such a term would constitute what is called a flavor-changing neutral

current (FCNC) interaction. Electromagnetism does not have such FCNC
interactions when constructed using the minimal substitution rule from
free-particle Lagrangians for the individual leptons. FCNC interactions for
the leptons are predicted in the context of the weak interactions now that
we know that neutrinos have mass and that the mass eigenstates are clearly
not the same as the Lagrangian eigenstates.

• So, now let’s get back to the cross section computation making the absolute
lepton number conservation assumption.
The initial and final states are:
|ii = c†e(~
k)d†e(~
p)|0i , |f i = c†µ(~
k 0)d†µ(~
p 0)|0i . (124)
The relevant Feynman diagram is

Figure 5: The one Feynman diagram contributing to e+e− → µ+µ−. The

arrows show the directions for the momenta. For the µ+ and e+, the
fermion-flow direction is opposite the indicated momentum direction.

Can you figure out from the Feynman ’rules’ (generalized to include µ
vertices as well as e vertices) what to write down without going through
the derivation below?
Now, since we need a e+ and e− in the initial state we will need a ψe+ and
+
a ψ e for the killing operations. For the final state µ+ and µ− we will need
− −
a ψ µ and a ψµ for the killing operations. This means we must go to S (2)
which now includes
"
e2
Z Z
(2) 4 4
S 3 − d x1 d x2 T {: ψ µ (x1 )A / (x2 )ψe (x2 ) :}
/ (x1 )ψµ (x1 ) :: ψ e (x2 )A
2
#
+T {µ ↔ e, x1 ↔ x2 } (125)
To see this, we must remember that S (2) contains
T {HI (x1)HI (x2)} = T {[HeI (x1) + Hµ

I (x 1 )][He
I (x 2 ) + Hµ
I (x2 )]} (126)
and that the terms of interest then arise as the two cross terms containing
the product of HeI with HµI . Of course, the x1 ↔ x2 2nd term above gives
exactly the same contribution as the 1st term and so we will work with the
equivalent form
Z Z
(2) 2 4 4
S 3 −e d x1 d x2 T {: ψ µ (x1 )A / (x2 )ψe (x2 ) :}
/ (x1 )ψµ (x1 ) :: ψ e (x2 )A (127)

For the process being considered we keep the already delineated parts of
S (2), namely:
Z Z
(2) 2 4 4 − − + +
S 3 −e d x1 d x2 T {: ψ µ (x1 )A / (x2 )ψe (x2 ) :}
/ (x1 )ψµ (x1 ) :: ψ e (x2 )A
Z Z
2 4 4 − ρ − + σ +
→ −e d x1 d x2 ψ µ (x1 )γ ψµ (x1 )Aρ (x1 )Aσ (x2 )ψ e (x2 )γ ψe (x2 ) (128)
We now insert this between our initial and final states and use the standard
killing operations. Thus, we compute (you will have noted that I am not
attempting to keep spin indices on the creation and annihilation operators
defining |ii and |f i)
" Z Z
(2) 0 0 2 4 4 − ρ −
hf |S |ii = ~ )cµ (~
h0|dµ (p k ) −e d x1 d x2 ψ µ (x1 )γ ψµ (x1 )Aρ (x1 )Aσ (x2 )
#
+ σ + † †
ψ e (x2 )γ ψe (x2 ) ce (~ ~)|0i
k )d e ( p
2 0 ρ 0 σ
= −e uµ (~
k )γ vµ (p ~)γ ue (~
~ )v e ( p k) ×
ix1 ·(k0 +p0 )−ix2 ·(k+p) 4 −gρσ −iq·(x1 −x2 )

i
Z Z Z
4 4
d x1 d x2 e d q 2 e
(2π )4 q + i
1
Z
2 0 ρ 0 σ 4
= −e uµ (~k )γ vµ (p ~)γ ue (~
~ )v e ( p k) 4
d q×
(2π )
−gρσ

4 4 0 0 4 4
(2π ) δ (k + p − q )(2π ) δ (q − p − k) i 2
q + i q =p+k

4 4 0 0
= (2π ) δ (p + k − p − k)M , with
−g

0 0 ρ 0
0 σ ρσ
M = uµ (λ , ~
k )(ieγ )vµ (λ , p ~)(ieγ )ue (λ, ~
~ )v e (λ, p k) i 2 (129)
q + i q =p+k
which is what you would get by “naive” application of the Feynman

rules, generalized to include muons as well as electrons. In the very last
expression, we have included the spin/helicities of the various particles that
had been dropped until now.
• So, now let’s insert this into our 2 → 2 cross section expression:
0
|~ |

dσ 1 pcom 2
= |M| (130)
dΩ01 com 64π 2Etot
2 |~ pcom|
using the approximation of neglecting me and mµ, so that |~ p 0|com =

p|com = Etot/2, to obtain (dropping the 01 notation on Ω, as is conventional)
|~

dσ 1
0 ~0 0
= ρ
~ 0)
uµ(λ , k )(ieγ )vµ(λ , p

dΩ com 64π 2Etot
2
−i
2
~)(ieγρ)ue(λ, ~
v e(λ, p k) (131)

q2

+ i q = p+ k

The above result is for fixed spins for all the initial and final particles.
More typically, we might wish to average over initial spins (corresponding
to unpolarized e+ and e− incoming beams) and sum over final spins
(corresponding to not measuring the final spins). In this case, we can make
some simplifications. To show how this works, we consider for the moment
the electron subcomponent of this process:
X
~)(ieγρ)ue(λ, ~
[v e(λ, p ~)(ieγρ0 )ue(λ, ~
k)][v e(λ, p k)]∗ (132)
λ,λ
where, since ρ is a dummy index eventually to be summed over, we must

use ρ0 for the other half of the absolute square. Now,
~)(ieγρ0 )ue(λ, ~
[v e(λ, p k)]∗ = ue(λ, ~
k)†(ieγρ0 )†(ve(λ, p
~)†γ 0)†
k)†γ 0γ 0(−ieγρ†0 )(γ 0)†ve(λ, p
= ue(λ, ~ ~)
= ue(λ, ~
k)(−ieγρ0 )ve(λ, p
~) (133)
so that we have (exposing Dirac indices in the 3rd line before proceeding)

X
~)(ieγρ)ue(λ, ~
[v e(λ, p ~)(ieγρ0 )ue(λ, ~
k)][v e(λ, p k)]∗
λ,λ
X
= ~)(ieγρ)ue(λ, ~
[v e(λ, p k)][ue(λ, ~
k)(−ieγρ0 )ve(λ, p
~)]
λ,λ
X
= ~)γ (γρ)γδ ue(λ, ~
e2v e(λ, p k)δ ue(λ, ~
k)(γρ0 )β ve(λ, p
~)β
λ,λ
X
= e2ve(λ, p ~)γ (γρ)γδ ue(λ, ~
~)β v e(λ, p k)δ ue(λ, ~
k)(γρ0 )β
λ,λ
= e2(p/ − me)βγ (γρ)γδ (k/ + me)δ(γρ0 )β

= e2Tr [(p/ − me)γρ(k/ + me)γρ0 ] . (134)
Similarly, we find (note the matching of the ρ and ρ0 Lorentz indices is

maintained within M and M∗)
X 0 ρ0 0
~
0 0 ρ
[uµ(λ , k )(ieγ )vµ(λ , p 0 0 ~0
~ 0)]∗
~ )][uµ(λ , k )(ieγ )vµ(λ , p
0
λ0 ,λ
2 0 ρ 0 ρ0
= e Tr [(k/ + mµ)γ (p/ − mµ)γ ] . (135)

• Thus, we now need to figure out how to evaluate this kind of trace of Dirac
matrices. There are a series of rules that we will slowly add to that one
needs to derive.
1. First, just by the definition of the trace, we have
Tr[Γ1Γ2] = Tr[Γ2Γ1] (136)

for any two of our 16 matrices.
2. Next, we can show that the trace of an odd number of γ matrices is
zero:
Tr[γ αγ β . . . γ ν ] = 0 (137)
if there are an odd number of matrices in [. . .]. This is shown as follows:
Tr[γ αγ β . . . γ ν ]
= Tr[γ52γ αγ β . . . γ ν ] (since γ52 = 1)
= −Tr[γ5γ αγ β . . . γ ν γ5] (since γ5 anticommutes with all the γ α, . . .)
= −Tr[γ5γ5γ αγ β . . . γ ν ] (by the cyclic property of the trace)
= −Tr[γ αγ β . . . γ ν ] (since γ52 = 1) (138)

so that the only possibility is that the original Tr was equal to 0.
3. Next, we consider
α β 1 α β β α 1
Tr[γ γ ] = Tr[γ γ + γ γ ] = Tr[[γ α, γ β ]+]
2 2
1
= Tr[2g αβ 14×4] = 4g αβ (139)
2
4. A particularly important one is:
Tr[γ αγ β γ γ δ ]
= Tr[(2g αβ − γ β γ α)γ γ δ ]
= Tr[2g αβ γ γ δ ] − Tr[γ β (2g α − γ γ α)γ δ ]
= Tr[2g αβ γ γ δ ] − Tr[γ β 2g αγ δ ] + Tr[γ β γ (2g αδ − γ δ γ α)]
= Tr[2g αβ γ γ δ ] − Tr[γ β 2g αγ δ ] + Tr[γ β γ 2g αδ ] − Tr[γ β γ γ δ γ α]
= 8[g αβ g δ − g βδ g α + g βg αδ ] − Tr[γ αγ β γ γ δ ] . (140)
Since the last term above duplicates the starting Tr, we can move it to

the lhs of the equation and divide both sides by a factor of 2 to obtain
Tr[γ αγ β γ γ δ ] = 4[g αβ g δ − g βδ g α + g βg αδ ] (141)
Note that if we contract some momenta with some of the free γ indices
we will get things like
Tr[p/ γ β k/ γ δ ] = pαkTr[γ αγ β γ γ δ ]
= 4pαk[g αβ g δ − g βδ g α + g βg αδ ]
= 4[pβ kδ − g βδ p · k + pδ kβ ] (142)
Well, that’s all we need for the moment. The Appendix of MS has more.
• We now return to the initial spin average, final spin sum version of our
cross section:
2
−i

dσ 1 1
X
0 ~0 ρ 0 0 ~

= uµ (λ , k )(ieγ )vµ (λ , p
~ )v e (λ, p
~)(ieγρ )ue (λ, k)

2 4
64π 2 Etot 2

dΩ com q + i q =p+k
0

λ,λ,λ0 ,λ
(143)

We use our conversion of the spin sums to traces in the form
2 2 0 ρ 0 ρ0
/ − me )γρ (k
e Tr [(p / + me )γρ0 ]e Tr [(k / − m µ )γ
/ + m µ )γ (p ]
0 0
4 2 0ρ 0ρ ρρ0 0 0 2 0ρ 0ρ
= 16e [pρ kρ0 − gρρ0 (p · k + me ) + pρ0 kρ ][k p −g (k · p + mµ ) + k p ]
4 0 0 0 0
= 32e [p · p k · k + p · k p · k]
(144)
where we used the above theorems, including Tr[odd number] = 0,

then neglected m2e and m2µ and performed the remaining Lorentz index
contractions.
We insert this to obtain:
2
dσ 1 4 0 0 0 0
1
= e [p · p k · k + p · k p · k]
dΩ com 8π 2Etot
2 (p + k)2 + i
(145)
• We now do a little bit of kinematics, writing (neglecting masses so that

E = E 0 = Etot/2 for all particles — for inclusion of masses in all this, see

MS)
p = (E, 0, 0, E) , k = (E, 0, 0, −E)

p0 = (E, E sin θ, 0, E cos θ) , k0 = (E, −E sin θ, 0, −E cos θ)
(146)
yielding
p · p0 = k · k0 = E 2(1 − cos θ)
p · k0 = k · p0 = E 2(1 + cos θ)
(p + k)2 = 2p · k = 4E 2 (147)
so that we obtain
2
dσ 1 1
= e4E 4[(1 − cos θ)2 + (1 + cos θ)2]
dΩ com 8π 2Etot
2 4E 2
1 2 2
= (4πα) [1 + cos2 θ]
8π 2Etot
2 16

α2
= 2
[1 + cos2 θ] , and
4Etot
4π α2 πα2
σtot = 2
= .
3 Etot 3E 2
(148)
It is common convention to write
s ≡ (p + k)2 = (p0 + k0)2 , (149)
2
where s = Etot = 4E 2, the latter holding when masses are neglected. Also,
we frequently write
t ≡ (p0 − p)2 = (k0 − k)2 , u ≡ (k0 − p)2 = (k − p0)2 . (150)
In the massless limit, t = −2p0 · p = −2k0 · k and u = −2k0 · p =

−2k · p0. Further, since t = −2E 2(1 − cos θ) we have dt = 2E 2d cos θ =
(E 2/π)dΩ = 14 (s/π)dΩ. With these substitutions, our differential cross

section can be expressed in the form
t2 u2

dσ 4π 1 2 1
= (4πα) +
dt s 8π 2s 4 s2 4
" 2 #
2 2
2πα t u
= + . (151)
s2 s s
This is a Lorentz invariant form that actually holds in any frame, since it is
expressed entirely in terms of Lorentz invariants.
A note on dimensions: dσ/dt has ’energy’ dimensions of 1/E 4 in the
h̄ = c = 1 units. In the massless limit, we see that all these dimensions are
provided by the energy scale of the process, as encoded in s. This is always
the case when the underlying theory has a dimensionless coupling constant
such as e, which in turn is always the case if we generate interactions using
the minimal substitution rule. In minimal substitution, ∂µ → ∂µ + iqAµ.
Since ∂µ and Aµ have the same dimension, R q4 must be dimensionless. To
check the dimensions of Aµ remember that d xL must be a dimensionless
action. With L = − 14 (∂µAν − ∂ν Aµ)(∂ µAν − ∂ ν Aµ), it is clear that Aµ
must have the same dimensions as ∂µ in order to cancel the dimensions of
d4x.

Photon polarization sums for a cross section
• Write
M = α ~ β ~ ~ ~
r1 (k 1 ) r2 (k2 ) . . . Mαβ... (k1 , k2 , . . .) . (152)
The µ(~ k) are gauge dependent objects. In particular, if we write (we keep
to a basis in which our initial choices — you will see why I use the word
“initial” in a moment — for the polarization ’s are real)
XX 1 ~
A µ+
(x) = q µ
r ( k)e −ik·x
ar (~
k)
~
k r 2V ω(~
k)
µ−
XX 1 ~ +ik·xa† (~
A (x) = q µ
r (k)e r k) (153)
~
k r 2V ω(~
k)
µ µ µ
and perform the
gauge transformation A (x) → A (x) + ∂ f (x) with
f (x) = f˜(k) ar (~
k)e−ik·x − a†(~
k)eik·x , then the corresponding change
in µ(~
k) is
µ → µ(~
r k) − ikµf˜(k) .
r (154)

(Note that this choice of gauge transformation keeps the Aµ field hermitian).
In QED, we can apply this for each of the external ’s, each with their own
~
ki. Thus, if M is GI, we must have
0 = k1αβ ~ ~ ~
r2 (k2 ) . . . Mαβ... (k1 , k2 , . . .)
~ β
= α
r1 ( k 1 )k 2 . . . Mαβ...(~
k1 , ~
k2, . . .)
= ... . (155)
MS extrapolates this to say that
k1αMαβ...(~ k2, . . .) = k2β Mαβ...(~

k1 , ~ k1 , ~
k2, . . .) = . . . = 0 . (156)
This is true only in the case of Abelian gauge theory (which QED is). In
a non-abelian gauge theory, this final form is not true and only a single
photon can be “removed” at a time — all the other polarizations must be
kept in place for the identity to hold.
• Another point of view that gives this same result is current conservation.
Let us focus on just one of the photons and write M = µ(~ k)Mµ(k).
We know that the photon is seeing an interaction of the form L =

R
d4xejµ(x)Aµ(x), where jµ = ψγµψ. Therefore,
Z
Mµ(k) = d4xeik·xhf |jµ(x)|ii . (157)
Then
Z
µ µ ik·x
4

k Mµ(k) = d x (−i∂ )e hf |jµ(x)|ii
Z
= − d4xeik·x[−i∂ µhf |jµ(x)|ii]
= 0 (158)
by virtue of current conservation, ∂ µhf |jµ(x)|ii = 0. (The current is not

necessarily conserved until placed in the context of definite on-shell states
for the other particles involved in the process, as encapsulated in the hi| and
|f i notation.) We know that such current conservation applies for the free
fields. In the interacting case, this current conservation is a consequence of
gauge invariance. The identity kµMµ(k) = 0 is called the Ward Identity.
• With this result, we can now simply perform polarization sums for the cross

section.
Writing Mr (k) = µ ~ ~
r (k)Mµ (k), the cross section will be proportional to
X 2 X
X= ~ ~ ∗ ~
Mr (k) = Mµ(k)Mν (k) µ (~k) ν ~
r (k) . (159)

r
r =1,2 r =1,2
For simplicity, we go to a Lorentz frame (M is a Lorentz invariant so

we can choose any frame we like) where kµ = (k, 0, 0, k) has only a
direction 3 vector component. Then, in the Lorentz gauge we know that
1(~
k) = (0, 1, 0, 0) and 2(~
k) = (0, 0, 1, 0), i.e. 11 = 1 and 22 = 1, all
others zero. So,
X = |M1|2 + |M2|2 , (160)
where the subscripts are the Lorentz indices. Now, for the above k, the
Ward identity reduces to
k(M0 + M3) = 0 , ⇒ M0 = −M3 . (161)
Then we can write equally well
X = |M1|2 + |M2|2 + |M3|2 − |M0|2 = −g µν MµM∗ν = −Mν M∗ν .

(162)

So, effectively, we have made the replacement
X
µ (~k) ν ~
( k) → −g µν
. (163)
r r
r =1,2
This is not an actual equality. In fact (for k2 = 0)
X
~ ν ~ 1
µ
r (k) r (k) → −g µν
− 2
[k ν µ
k −(k·n)(k µ ν
n +k ν µ
n )] , (164)
r =1,2
(k · n)
where nµ = (1, 0, 0, 0), but the other terms don’t contribute because of
the Ward identity.
• Let us now use this technology for the case of Compton scattering. This
was discussed beginning with Eq. (61) in some detail with the result that
we developed the two contributing amplitudes

q/ + m
p 0)/ s0 (~
Ma = −e2ur0 (~ k0) i 2 / s(~
k)ur (~
p) (165)
q − m2 q = p+ k

q/ + m
p 0)/ s(~
Mb = −e2ur0 (~ k) i 2 2

/ s 0 (~
k 0
)ur (~
p) (166)
q − m q=p−k0
corresponding to the Feynman diagrams

/0 1 3456
1 /(0
- .

!

8
5 19

. -
/20 7 /(0
"#$&%('*))+, /#5$&%'*))+:,
Figure 6: The two Feynman diagrams contributing to e−γ → e−γ.
We need X X X
1
4
2
|Mrsr0s0 | = 1
4
Mαβ M∗αβ (167)
pol, s,s0 spin, r,r0 spin
where we have written (dropping the r and r 0 indices for the moment)
~ β ~0
Mss0 = α
s ( k) s0 (k )Mαβ (168)

and used the following procedure:
0
X X h ih 0 i
2 ~ β ~0 α ~ β ~0 ∗
|Mss0 | = α
s ( k) s 0 ( k )M αβ s ( k) s 0 (k )M α0 β 0
s,s0 s,s0
! !
α ~ α0 ~ β ~ 0 β0 ~ 0
X X
= s (k)s (k) s0 (k )s0 (k ) Mαβ M∗α0β0
s s0
αα0 ββ 0
= (−g )(−g )Mαβ M∗α0β0
= Mαβ M∗αβ . (169)
In the above, we have
Mαβ = Ma
αβ + M b
αβ (170)
with

p/ + k/ + m
Ma
αβ p 0)γβ
= −ie2ur0 (~ γαur (~
p)
(p + k)2
− m2
p/ − k/ 0 + m

Mbαβ p 0)γα
= −ie2ur0 (~ γβ ur (~
p) (171)
(p − k 0 )2 − m2

2 2
Using p2 = p0 = m2 and k2 = k0 = 0, the denominators above reduce to
2p · k and −2p · k0, respectively, and we find
X X X
1
4
|M| 2
= 1
4
Mαβ M∗αβ
e4

Xaa Xbb Xab + Xba
= + − (172)
16 (pk)2 (pk0)2 (pk)(pk0)
where
0 β α 0 ∗
X
Xaa = [ur0 (p
~ )γ (f
/ 1 + m)γ ur (p
~)][ur0 (p
~ )γβ (f
/ 1 + m)γα ur (p
~)]
spin
0 β α 0
X
= [ur0 (p
~ )γ (f
/ 1 + m)γ ur (p
~)][ur (p
~)γα (f
/ 1 + m)γβ ur0 (p
~ )]
spin
 !  
β α 0 0
X X
= Tr γ (f
/ 1 + m )γ ~)ur (p
ur ( p ~) γα (f
/ 1 + m)γβ  ~ )ur0 (p
ur0 (p ~ )
r r0
h i
β α 0
= Tr γ (f/ 1 + m )γ (p/ + m)γα (f/ 1 + m)γβ (p/ + m) (173)
0 α β 0 ∗
X
Xbb = [ur0 (p
~ )γ (f
/ 2 + m)γ ur (p
~)][ur0 (p
~ )γα (f
/ 2 + m)γβ ur (p
~)]
spin
h i
α β 0
= Tr γ (f
/ 2 + m )γ (p
/ + m )γβ (f
/ 2 + m )γα (p
/ + m) (174)

0 β α 0 ∗
X
Xab = [ur0 (p
~ )γ (f
/ 1 + m)γ ur (p
~)][ur0 (p
~ )γα (f
/ 2 + m)γβ ur (p
~)]
spin
h i
β α 0
= Tr γ (f/ 1 + m )γ (p/ + m )γβ (f
/ 2 + m )γα (p/ + m) (175)
0 α β 0 ∗
X
Xba = [ur0 (p
~ )γ (f
/ 2 + m)γ ur (p~)][ur0 (p
~ )γβ (f
/ 1 + m)γα ur (p
~)]
spin
h i
α β 0
= Tr γ (f
/ 2 + m )γ (p
/ + m )γα (f
/ 1 + m )γβ (p
/ + m) (176)
where f1 = p + k and f2 = p − k0.
Under the substitutions k ↔ −k0, α ↔ β we find f1 ↔ f2 and hence

Xaa ↔ Xbb and Xba ↔ Xab. Thus, we need only compute Xaa and Xab
and then use the substitution rules to get Xbb and Xba.
If we look closely at Xab as compared to Xba, we see that they are

related by trace reversal. In the Appendix of MS, we see that the trace
of a product of γ matrices is equal to the trace of the reverse product of
the γ matrices by virtue of the fact that there is a matrix C such that
Cγ µC −1 = −[γ µ]T . Assuming an even number of γ matrices, the result is
that inserting CC −1 = 1 throughout the trace converts the original trace
to a trace such that the transpose of each γ matrix appears inside the trace
and then we employ Tr[AT B T C T . . . Z T ] = Tr[Z . . . CBA].

• To simplify these traces, we need some more trace identities.
1. First, we have
α α β 1 α β 1 αβ
γ γα = gαβ γ γ = gαβ [γ , γ ]+ = gαβ 2g = 4. (177)
2 2
2. Next, we have
α ν αν ν α ν ν
γ γ γα = (2g − γ γ )γα = (2 − 4)γ = −2γ (178)
3. Thirdly, we have
ν µ α ν ν µ α
γα γ γ γ = (2gα − γ γα )γ γ
µ ν ν µ
= 2γ γ − γ (−2γ )
µ ν
= 2[γ , γ ]+
µν
= 4g (179)
4. Finally, we need
ν µ σ α ν ν µ σ α
γα γ γ γ γ = (2gα − γ γα )γ γ γ
µ σ ν ν µ σ α
= 2γ γ γ − γ γα γ γ γ
µ σ ν µσ ν
= 2γ γ γ − 4g γ
µ σ ν µ σ σ µ ν
= 2γ γ γ − 2(γ γ + γ γ )γ
σ µ ν
= −2γ γ γ . (180)

• Using these identities we may now evaluate our X’s.
β α 0
h i
Xaa = Tr γ (f / 1 + m)γ (/ p + m)γα (f / 1 + m)γβ (/ p + m)
α 0 β
h i
= Tr (f/ 1 + m)γ (/ p + m ) γα ( f
/ 1 + m)γβ (/p + m) γ
0
h i
= Tr (f/ 1 + m)(−2/ p + 4m)(f / 1 + m)(−2/ p + 4m)
0
h i
= 4Tr (f/ 1 + m)(/p − 2m)(f / 1 + m)(/ p − 2m)
0 2 0 0 4
n h i o
= 4 Tr[f/ 1/
pf p ] + m −4Tr[f
/ 1/ p ] − 4Tr[f
/ 1/ / 1/p ] + Tr[/p/p ] + 4Tr[f
/ 1f
/ 1 ] + 4m Tr[1]
(
0 0 2 0
= 16 [2(f1 p)(f1 p ) − (pp )(f1 f1 )] + m [−4(f1 p) − 4(f1 p )
)
0 4
+4(f1 f1 ) + (pp )] + 4m
0 2 4
= 32[(pk)(pk ) + m (pk) + m ] , (181)
where in the last step we used
(f1f1) = (p + k)2 = m2 + 2(pk)

(f1p) = (p + k) · p = m2 + (pk)
(f1p0) = (p0 + k0) · p0 = m2 + (k0p0) = m2 + (kp)
(pp0) = p · (p + k − k0) = m2 + (pk) − (pk0) (182)

2
as follow from p2 = p0 = m2, (pk) = (p0k0) and (pk0) = (p0k).
From the substitution rule, we immediately obtain
Xbb = 32[(pk)(pk0) − m2(pk0) + m4] . (183)
For Xab we find
β α 0
h i
Xab = Tr γ (f
/ 1 + m)γ (/
p + m ) γβ ( f
/ 2 + m)γα (/
p + m)
β 2 0
h i
= / 1 + m)(−2f
Tr γ (f p + 4mf2 β + 4mpβ − m 2γβ )(/
/ 2 γβ / p + m)
"(
= Tr −8(f1 f2 )/
p + 4mf
/ 2/
p + 4mf
/ 2 (f
/ 1 + m) + 4mp
/ (f
/ 1 + m)
) #
2 0
−2m (−2f
/ 1 + 4m) (/
p + m)
"
0 2 0 2 0 2 2 0
= 4 −8(f1 f2 )(pp ) + 4m (f2 p ) + 4m (pp ) + 4m (f1 p ) + 4m (f2 p)
#
2 2 4
+4m (f2 f1 ) + 4m (f1 p) − 8m p 0 , 2nd from m
1st from /
"
2 0 2 0 2 0 2 2 2 2 0
= 4 −8m (pp ) + 4m [(pp ) − (pk)] + 4m (pp ) + 4m [m + (pk)] + 4m [m − (pk )]

#
2 2 2 2 4
+4m m + 4m [m + (pk)] − 8m
4 2 2 0
= 16[2m + m (pk) − m (pk )] , (184)
where we had to use

0 0 0 0 2 0 0 2
(f1 f2 ) = (p + k) · (p − k ) = (p + k) · p − (p + k ) · k = m + (pk) − (p k ) = m
0 0 0 0 0 0 0
(f2 p ) = (p − k ) · p = (pp ) − (k p ) = (pp ) − (pk)
0 0 0 0 0 2 0 0 2
(f1 p ) = (p + k) · p = (p + k ) · p = m + (k p ) = m + (kp)
0 2 0
(f2 p) = (p − k ) · p = m − (pk )
2
(f1 p) = (p + k) · p = m + (pk) .
Using the substitution rule (k ↔ −k0) to get Xba, we verify that

Xba = Xab as expected.
• Using these results in

X X X
1 1
4
|M| 2
= 4
Mαβ Mαβ
e4

Xaa Xbb Xab + Xba
= + − (185)
16 (pk)2 (pk0)2 (pk)(pk0)

we obtain
p · k0
( ! 2 )
p·k

1
X X 2 4 2 1 1 4 1 1
4 |M| = 2e + + 2m − +m −
p · k0 p·k p·k p · k0 p·k p · k0
pol, s,s0 spin, r,r0
(186)
This takes a very simple form in the laboratory system defined by p =
(m, 0, 0, 0) for which we can write k = (ω, ~
k), k0 = (ω 0, ~
k 0) and p0 =
(E 0, p
~ 0). For these definitions, we have
0
( ! 2 )
1
X X 2 4 ω ω 1 1 2 1 1
4 |M| = 2e + + 2m − 0 +m − 0 . (187)
ω0 ω ω ω ω ω
This can be further simplified by using the Compton scattering relation
1 1 1
− = (cos θ − 1) , (188)
ω ω0 m
where θ is the angle between ~

k and ~
k0, which gives
0
X X ω ω
1
4
2
|M| = 2e 4
+ − sin2 θ . (189)
ω0 ω

To derive Eq. (188) we note that since p = p0 + k0 − k we also have
p · k = (p0 + k0 − k) · k = p0 · k + k0 · k = p · k0 + k0 · k , (190)
where we used (p − k0)2 = m2 − 2p · k0 = (p0 − k)2 = m2 − 2p0 · k, which

in turn implies that p · k0 = p0 · k, for the 2nd equality.
If we take ~k along the ẑ axis and ~k0 at an angle θ with respect to the ẑ
axis, then ~
k0 · ~
k = ωω 0 cos θ and the above equation reduces to
ωm = ω 0m + ωω 0(1 − cos θ) (191)
which, in turn, gives Eq. (188). For the cross section, we return to the
expression
 
X X 1 Y ~0f
d3 p
4 4
dσ = (2π) δ ( 0
pf − pi )   |M|2 (192)
4E1E2vrel f
(2π)32Ef0
with E1E2vrel = [(p1 · p2)2 − m21m22]1/2. In our case we can identify particle
1 as a photon with m1 = 0 and particle 2 as the target proton. Then we

get
!
3 0 3~ 0
4 4 0 0 1 d p
~ d k 2
dσ = (2π) δ (p + k − p − k) |M|
4ωm (2π)32E 0 (2π)32ω 0
1 0 0 0 0 2 2
= δ(ω + E − m − ω)ω dω dΩ |M|
64π 2ωω 0mE 0
−1
ω0 ∂(ω 0 + E 0)

2
= dΩ |M|
64π 2ωmE 0 ∂ω 0
(193)
where we have used the notation dΩ for the ~

k0 solid angle. To compute
the required derivative, we note that
E 0 = [m2 + (~
k −~
k0)2]1/2 = [m2 + ω 2 + ω 02 − 2ωω 0 cos θ]1/2 (194)
so that
∂E 0 ω 0 − ω cos θ
= (195)
∂ω 0 E0
implying that
∂(E 0 + ω 0) ω 0 − ω cos θ
= +1
∂ω 0 E0
ω 0 + E 0 − ω cos θ
=
E0
ω + m − ω cos θ
=
E0
mω
= , (196)
E 0ω 0
where for the next-to-last step we used energy conservation, ω 0 + E 0 =

ω + E, and for the last step we used Eq. (188). Altogether, we get
ω0 E 0ω 0 ω0

dσ ω
= 2e4 + − sin2 θ
dΩ 64π 2ωmE 0 mω ω0 ω
4 0 2 0
e ω ω ω
= + − sin2 θ
32π 2m2 ω ω0 ω

2 0 2 0
α ω ω ω
= + − sin2 θ . (197)
2m2 ω ω0 ω
In the non-relativistic limit, ω m, which means ω ∼ ω 0 (see Eq. (188)),

and we get
dσ α2 2
→ 2
(1 + cos θ) , (198)
dΩ 2m
which is the Thomson cross section. Obviously, the full relativistic result
of Eq. (197) has been tested against experimental data in a highly detailed
way and excellent agreement has been found (after including radiative, i.e.
higher order, corrections).
• Finally, we wish to return to the gauge invariance issue. You will recall that
we had the two amplitudes:

q/ + m
Ma = −e2ur0 (~p 0)/ s0 (~
k0 ) i 2 2
/ s (~k)ur (~p) (199)
q − m q=p+k

q/ + m
p 0)/ s(~
Mb = −e2ur0 (~ k) i 2 2

/ s 0 (~
k 0
)ur (~
p) , (200)
q − m q=p0−k

where for q in Mb we write q = p0 − k instead of the equivalent (by
momentum conservation) form q = p − k0 employed earlier.
The gauge invariance claim is that if we replace (~

k0) by k0 or (~k) by k,
then M = Ma + Mb → 0. Let’s check this for the case of (~ k) → k.
We get, dropping spin indices for convenience, and writing in the explicit q
values for each amplitude

p/ + k/ + m
p 0)/ (~
Ma → −e2u(~ k0) i 2 2
k/ u(~p) (201)
(p + k) − m
0
−

p
/ k
/ + m ~
p 0)k/ i 0
Mb → −e2u(~ 2 2

/ ( k 0
)u(~
p) (202)
(p − k) − m
In Ma, we write
p/ + k/ + m p/ + k/ + m
k/ u(~
p) = (k/ + p/ − p/ )u(~
p)
(p + k)2 −m 2 (p + k)2 − m2
p/ + k/ + m
= (k/ + p/ − m)u(~
p)
(p + k)2 − m2

(p + k)2 − m2
= u(~
p)
(p + k)2 − m2
= u(~
p) , (203)
so that
p 0)/ (~
Ma → −ie2u(~ k0)u(~
p) . (204)
In Mb, we write
p/ 0 − k/ + m p/ 0 − k/ + m
p 0)k/
u(~ p 0)(k/ − p/ 0 + p/ 0)
= u(~
(p0 − k)2 − m2 (p0 − k)2 − m2
0 0 p/ 0 − k/ + m
p )(k/ − p/ + m)
= u(~
(p0 − k)2 − m2
0 −(p0 − k)2 + m2
= u(~
p )
(p0 − k)2 − m2
p 0) ,
= −u(~ (205)
so that
p 0)/ (~
Mb → +ie2u(~ k0)u(~
p) . (206)

Obviously, these two results (after the substitution) for Ma and Mb cancel
to give zero, the requirement of gauge invariance.

QFT Ii

Uploaded by

Copyright:

Available Formats

QFT Ii

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

QFT Ii

Uploaded by

Copyright:

Available Formats

Class Notes for Quantum Field Theory: Section II

S -Matrix, Wick’s Theorem, Feynman Rules, Specific Sample Calculations

Primary Reference: Mandl & Shaw, “Quantum Field

• The perturbative approach is most conveniently derived in the interaction

J. Gunion 230A – Part II 1

The Interaction Picture

• Consider QED as the example theory.

• We divide up the full L into

J. Gunion 230A – Part II 2

|A, tiS = U |A, t0iS (5)

from which one has

J. Gunion 230A – Part II 3

• From Eq. (8),

• Now write H = H0 + HI , where the I subscript is for interaction and not

• The I.P. is defined relative to the S.P. by using

U0 ≡ U0(t, t0) = e−iH0(t−t0) (11)

to define the I.P. state

|A, tiI = U0†|A, tiS (12)

J. Gunion 230A – Part II 4

• Differentiating Eq. (13) gives

and we also find

J. Gunion 230A – Part II 5

where: 1) we used the fact that H S ≡ H in this notation; 2) we used the

• Note that we can rewrite the last few lines above as

HII = e+iH0(t−t0)HIS e−iH0(t−t0)

J. Gunion 230A – Part II 6

• This expansion is based upon the I.P. reviewed above, in which:

J. Gunion 230A – Part II 7

hΦ(t)|Φ(t)i = const. (21)

J. Gunion 230A – Part II 9

• The S-matrix is defined by

|Φ(t = ∞)i = S|Φ(t = −∞)i = S|ii . (22)

Note that Eq. (21), conservation of normalization, implies

hΦ(∞)|Φ(∞)i = hΦ(−∞)|S †S|Φ(−∞)i = hΦ(−∞)|Φ(−∞)i (23)

which requires that S †S = 1, i.e. S is a unitary operator.

• S|ii can consist of a large selection of possible final states |f i.

J. Gunion 230A – Part II 10

hf |Φ(∞)i = hf |S|ii ≡ Sf i . (25)

1 = hi|ii = hi|S †S|ii = hi|S †|f ihf |S|ii = Sf∗iSf i . (28)

This equation expresses the conservation of probability. It is more general

J. Gunion 230A – Part II 11

J. Gunion 230A – Part II 12

J. Gunion 230A – Part II 13

J. Gunion 230A – Part II 14

since the H’s are already in the correct time ordering.

Thus, we can write

J. Gunion 230A – Part II 15

• MS now embarks on a discussion of why it is justified to use non-interacting

J. Gunion 230A – Part II 16

J. Gunion 230A – Part II 17

J. Gunion 230A – Part II 18

• We must now develop a computational technique for writing down an

• To get a non-zero contribution to Sf i we must have an appropriate number

J. Gunion 230A – Part II 19

and end with a final state

• Since the minimal substitution interaction is HI = −LI = −e : ψA/ ψ :,

J. Gunion 230A – Part II 20

J. Gunion 230A – Part II 21

write the , which is just a number, anywhere I want.)