Physics 523, Quantum Field Theory II

Homework 3
Due Wednesday, 28st January 2004
Jacob Lewis Bourjaily
The Rosenbluth Formula

We are to prove the Rosenbluth Formula by considering the elastic scattering of a relativistic electron
off of a proton while correcting the vertex function of the proton. The amplitude for this process is,

p0 k0
iM = = u(k 0 )(−ieγµ )u(k) −i 0 µ
q 2 u(p )(−ieΓ )u(p).
←
−
q k
p

p+ e−
a) Let us simplify the amplitude using the Gordon identity. Recall that we showed in class that
the generalized vertex function Γµ may be written in terms of functions F1 (q 2 ) and F2 (q 2 ) as
iσ µν qν
Γµ = γ µ F1 (q 2 ) + F2 (q 2 ).
2m
Inserting this into the amplitude and recalling the Gordon identity, we see that
e2
iM = i u(k 0 )γµ u(k)u(p0 )Γµ u(p),
q2
µ ¶
e2 0 0 µ iσ µν qν
= i 2 u(k )γµ u(k)u(p ) γ F1 + F2 u(p),
q 2m
µ ¶
e2 0 0 µ iσ µν qν (p0 + p)µ (p0 + p)µ
= i 2 u(k )γµ u(k)u(p ) γ F1 + F2 + F2 − F2 u(p),
q 2m 2m 2m
µ ¶
e2 (p0 + p)µ
= i 2 u(k 0 )γµ u(k)u(p0 ) γ µ (F1 + F2 ) − F2 u(p),
q 2m
(p0 + p)µ
∴ Γµ = γ µ (F1 + F2 ) − F2 .
2m
b) Let us compute the spin-averaged amplitude squared directly. We see that
µ ¶ µ ¶
e4 X (p0 + p)µ (p0 + p)ν
|M |2 = 4 u(k 0 )γµ u(k)u(p0 ) γ µ (F1 + F2 ) − F2 u(p)u(p) γ ν (F1 + F2 ) − F2 u(p0 )u(k)γν u(k 0 ),
4q 2m 2m
spin
4
e
= Tr [(6 k 0 + me )γµ (6 k + me )γν ] ×
4q 4
½
(p0 + p)ν
(F1 + F2 )2 Tr [(6 p0 + m)γ µ (6 p + m)γ ν ] − F2 (F1 + F2 ) Tr [(6 p0 + m)γ µ (6 p + m)]
2m
¾
(p0 + p)µ (p0 + p)µ (p0 + p)ν
−F2 (F1 + F2 ) Tr [(6 p0 + m)(6 p + m)γ ν ] + F22 Tr [(6 p0
+ m)(6 p + m)] ,
2m 4m2
4e4 ¡ ¢
= 4 kµ0 kν + kν0 kµ − gµν (k 0 · k − m2e ) ×
q
½
¡ ¢
(F1 + F2 )2 p0µ pν + p0ν pµ − g µν (p0 · p − m2 ) − F2 (F1 + F2 )(p0 + p)µ (p0 + p)ν
¾
F2
+ 2 2 (p0 + p)µ (p0 + p)ν (p0 · p + m2 ) ,
4m
4 ¡ ¢
4e
= 4 kµ0 kν + kν0 kµ − gµν (k 0 · k − m2e ) ×
q
½ µ 0 ¶¾
¡ ¢ p · p + m2 2
(F1 + F2 )2 p0µ pν + p0ν pµ − g µν (p0 · p − m2 ) + (p0 + p)µ (p0 + p)ν F2 − F 2 (F 1 + F 2 ) ,
4m2
·
8e4 ¡ ¢
∴ |M | = 4 (F1 + F2 )2 k 0 · p0 k · p + k 0 · pk · p − k 0 · km2 − p0 · pm2e + 2m2 m2e
2
q
µ 0 ¶µ ¶¸
p · p + m2 2 0 0 0 1 0 2 0 2
+ F2 − F2 (F1 + F2 ) k · (p + p)k · (p + p) − (k · k − me )(p + p) .
4m2 2
1
2 JACOB LEWIS BOURJAILY

c) Let us consider the kinematics of this reaction in the initial rest frame of the proton. In this
frame we see that p = (m, ~0), k = (E, E ẑ), k 0 = (E 0 , k~0 ), p0 = (E − E 0 + m, −~k) with |k~0 | = E 0 .
We have defined the momentum transfer q such that p0 − p = q = k − k 0 .
Noting that p · p0 = m2 + Em − E 0 m, let us compute p02 .
p02 = (p+q)2 = p2 +2p·q +q 2 = m2 +2p·(p0 −p)+q 2 = −m2 +2p0 ·p+q 2 = m2 +2Em−2E 0 m+q 2 = m2 ,
=⇒ q 2 = 2E 0 m − 2Em,
q2
∴ E0 = E + .
2m
If we write k = (E , 0, E sin θ, cos θ) so that q = (E − E 0 , 0, −E 0 sin θ, E − E 0 cos θ) we see
0 0 0

θ
q 2 = E 02 − 2EE 0 + E 2 − E 02 − E 02 sin2 θ − E 2 + 2EE 0 cos θ − E 02 cos2 θ = 2EE 0 (cos θ − 1) = −4EE 0 sin2 .
2
q2
Using our identity derived above that E 0 = E + 2m , we may conclude that
2
θ q θ
q 2 = −4E 2 sin2
− 4E sin2 ,
2 2m 2
2 2 θ
4E sin
∴ q2 = − 2
2 θ.
1 + 2E
m sin 2
Let us now compute all of the required inner products to compute the desired amplitude
squared. Noting that p2 = p02 = m2 , k 2 = k 02 = 0, and p · k = Em we may derive all of our
necessary identities and inner products indirectly (it’s more fun that way). We notice that
p02 = m2 = p2 + 2p · q + q 2 = −m2 + 2p0 · p + q 2 ,
q2
∴ p0 · p = m2 − .
2
Similarly,
p02 = m2 = p2 + 2p · q + q 2 = m2 + 2p · k − 2p · k 0 + q 2 ,
but we know that p · k = Em,
q2
∴ p · k 0 = Em + .
2
Likewise,
k 02 = 0 = k 2 − 2k · q + q 2 = 2k · k 0 + q 2 = 0,
q2
∴ k0 · k = − .
2
And
k 02 = 0 = k 2 − 2k · q + q 2 = −2k · p0 + 2k · p + q 2 ,
where we know that k · p = Em and
q2
∴ k · p0 = Em + .
2
Similarly,
k 2 = 0 = k 02 + 2q · k 0 + q 2 = 2p0 · k 0 + q 2 ,
∴ p0 · k 0 = Em.
Tabulating our results, we have shown that
q2 q2 q2
k0 · k = − p0 · p = m2 − k 0 · p = Em +
2 2 2
q2
p0 · k = Em + k 0 · p0 = Em p · k = Em.
2
These imply that
q2 q2
k · (p0 + p) = 2Em + , k 0 · (p0 + p) = 2Em + , and (p + p0 )2 = 4m2 − q 2 .
2 2
 PHYSICS 523: QUANTUM FIELD THEORY II HOMEWORK 3 3

d) We are to use the kinematic information derived in part (c) above to rewrite the spin-averaged
amplitude squared into a more convenient form. Recall that
i
·
8e4 ¡z }| ¢{
|M |2 = 4 (F1 + F2 ) k · p k · p + k · pk · p − k 0 · km2 − p0 · pm2e + 2m2 m2e
2 0 0 0
q
µ 0 ¶µ ¶ ¸
p · p + m2 2 0 0 0 1 0 2 0 2
+ F 2 − F 2 (F1 + F 2 ) k · (p + p)k · (p + p) − (k · k − me )(p + p) .
4m2 2
| {z }| {z }
ii iii
2
We note that in the approximation where k ∼ 0, we should set me → 0. Let us compute each
part separately first before combining the results.
q4 q2
i. (k 0 · p0 )(k · p) + (k 0 · p)(k · p0 ) − (k 0 · k)m2 = (Em)2 + (Em)2 + Emq 2 + + m2 .
4 2

ii. p0 · p + m2 2 2 1 2 q2 2
F s − F1 F 2 − F2 = F − F − F1 F2 − F22 ,
4m2 2 2· 8m2 2 ¸
1 q2 2
= − (F22 + 2F1 F2 + F12 − F12 + F ,
2 4m2 2
· µ ¶¸
1 2 2 q2 2
= − ((F1 + F2 ) − F1 − F .
2 4m2 2

1 q4 q4
iii. (k 0 · (p0 + p))(k · (p0 + p)) − (k 0 · k)(p0 + p)2 = 4(Em)2 + 2Emq 2 + + q 2 m2 − .
2 4 4

Combining these results, we see that the coefficient for the (F1 + F2 )2 term will be
q4 q2 q2 q4
2(Em)2 + Emq 2 + + m2 − 2(Em)2 − Emq 2 − m2 = ,
4 2 2 4
which can be written,
q4 q2 q2 2E 2 m2 q2
= =− 2 .
4 2 2 1 + 2E
m sin
θ
2
2m2
q2
Similarly, we will combine the results above to compute the coefficient for the (F12 − 2
4m2 F2 )
term.
q2 4E 3 m sin2 θ2 2E 2 m2 sin2 θ2
2(Em)2 + Emq 2 + m2 = 2E 2 m2 − 2 θ − 2 θ ,
2 1 + 2E
m sin 2 1 + 2E
m sin 2
2E 2 m2 + 4E 3 m sin2 θ
2 − 4E 3 m sin2 θ
2 − 2m2 E 2 sin2 θ
2
= ,
1+ 2E
m sin2 θ
2
1 − sin2 θ
2E 2 m2 cos2 θ
= 2E 2 m2 2E
2
2 θ = 2E
2
2 θ .
1+ m sin 2 1+ m sin 2

Therefore, combining all of these results, the total spin-average amplitude squared becomes
·µ ¶ ¸
16e4 E 2 m2 2 q2 2 2 θ q2 2 2 θ
|M | = 4 ¡
2 ¢ F1 − F cos − (F1 + F2 ) sin .
q 1 + 2E 2 θ
m sin 2
4m2 2 2 2m2 2
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
4 JACOB LEWIS BOURJAILY

dσ ¯
¯
e) Let us compute the differential cross section, d cos θ lab . To do this, we will compute the cross
section in most general terms. From elementary considerations, we calculated that
 
1 Y d3 p f 1 ³ X ´
dσ =   |M |2 (2π)4 δ (4) pA + pB − pf ,
2EA 2EB |vA − vB | (2π)3 2Ef
f
3 0 3 0
1 d pd k 1
= |M |2 δ (4) (p + k − p0 − k 0 ).
4mE (2π)2 4E 0 Ep0
We see that this is so because EA = m, EB = E, |vA − vB | = 1 and there are two final states.
Let us now integrate over dσ to find its dependence on cos θ. During the derivation, we will
make use of the fact that E+m = Ep0 +E 0 by energy conservation enforced by the dirac δ function.
2
q 4E 2 sin2 θ
We will also call upon our results above to use the identities E 0 = E + 2m and q 2 = 1+ 2E sin22θ .
m 2
Notice the insertion of the Jacobian for the change of variables to integrate over the energy
portion of the δ function in line 4. We will now proceed directly by first integrating over the p0
part of the integral.
Z Z 3 0 3 0
1 d pd k 1
σ = dσ = |M | 2 δ (4) (p + k − p0 − k 0 ),
4mE (2π) 4E 0 Ep0
2
Z 3 0 ³ p ´
1 d k 1 (1) 0
= |M |2 δ E − E − m + m 2 + E 2 + E 02 − 2EE 0 cos θ ,
4mE (2π)2 4E 0 Ep0
Z 02 ³ p ´
1 E dEdΩ 1 (1) 0
= |M |2 δ E − E − m + m 2 + E 2 + E 02 − 2EE 0 cos θ ,
4mE (2π)2 4E 0 Ep0
Z µ ¶−1
1 dΩ E 0 E 0 − E cos θ
= |M |2 1 + ,
4mE (2π)2 4Ep0 Ep0
Z µ ¶
1 2
dΩ E 0 Ep0
= |M | ,
4mE (2π)2 4Ep0 Ep0 + E 0 − E cos θ
Z µ ¶
1 d cos θ E 0 Ep0
= |M |2 ,
4mE (2π) 4Ep0 Ep0 + E 0 − E cos θ
Z
1 2
E0
= |M | d cos θ ,
32πmE m + E(1 − cos θ)
Z
1 E0
= |M |2 d cos θ ,
32πm E 2
1 + m sin2 θ2
2E

Z q2
1 2
E + 2m
= |M | d cos θ 2 θ,
32πm2 E 1 + 2E m sin 2
Z 2E 2 sin2 θ2
1 E− m(1+ 2E 2 θ
m sin 2 )
= |M |2 d cos θ 2E 2 θ ,
32πm2 E 1 + m sin 2
Z 2
2 θ 2E 2 2 θ
1 E + 2E
m sin 2 − m sin 2
= |M |2 d cos θ 2 θ 2 ,
32πm2 E (1 + 2E m sin 2 )
Z
1 E
= |M |2 d cos θ ,
32πm2 E (1 + 2E
m sin2 θ2 )2
Z
1 2
= 2 θ 2 |M | d cos θ,
32πm2 (1 + 2E
m sin 2 )
1 2
= 2 θ 2 |M | cos θ,
32πm2 (1 + 2E
m sin 2 )
¯
dσ ¯¯ 1
∴ = ¡ ¢ |M |2 .
d cos θ ¯lab 32πm 1 + 2E
2 sin 2 θ 2
m 2
‘ ’
óπ²ρ ²́δ²ι δ²ιξαι
 PHYSICS 523: QUANTUM FIELD THEORY II HOMEWORK 3 5

f ) We will now derive the Rosenbluth formula. From our work above, we see that
h³ ´ i
¯ 16e4 E 2 m2 F12 − 4m q2 2
cos2 θ2 − 2m q2 2 2 θ
dσ ¯¯ 2 F2 2 (F1 + F2 ) sin 2
= ¡ ¢ ¡ ¢ ,
d cos θ ¯lab 32πm2 1 + 2E 2 θ 2 4
q 1 + 2E 2 θ
m sin 2 m sin 2
h³ ´ i
q2 q2 2 θ
e4 E 2 F12 − 4m 2 F2
2
cos2 θ2 − 2m 2
2 (F1 + F2 ) sin 2
= ¡ ¢2 ¡ ¢ ,
16E 4 sin4 θ2
2π θ 2
1 + 2Em sin2 θ2 1 + 2E m sin2 θ2
(1+ m sin 2 )
2E 2
h³ ´ i
q2 q2 2 θ
e4
F12 − 4m 2 F2
2
cos2 θ2 − 2m 2
2 (F1 + F2 ) sin 2
= ¡ ¢ ,
32πE 2 sin4 θ2 1 + 2E m sin 2
2 θ

h³ ´ i
¯ πα2 F12 − q2 2 q
cos2 θ2 − 2m
2
2 2 θ
dσ ¯¯ 4m2 F2 2 (F1 + F2 ) sin 2
∴ = ¡ ¢ .
d cos θ ¯ lab 2E 2 sin4 θ2 1 + 2E
m sin 2
2 θ

‘ ’
óπ²ρ ²́δ²ι δ²ιξαι