Yukawa potential

November 11, 2014

1 The Yukawa potential

We consider properties of the Yukawa potential,
k −r
V (r) = e a
r
This potential is the static, spherically symmetric solution to the Klein-Gordon equation,

1 ∂2V 2 m2 c2
− + ∇ V = V
c2 ∂t2 ~2
To see this, let V = V (r) and write the Laplacian in spherical coordinates. Then we have

m2 c2
 
1 d 2 dV
r = V
r2 dr dr ~2

Differentiating on the left side, we have

d2 V 2 dV m2 c2
+ − V = 0
dr2 r dr ~2
Let U = rV . Then
d2 U
 
d dV
= V +r
dr2 dr dr
dV d2 V
= 2 +r 2
dr dr
so that we may write the equation as

1 d2 U m2 c2 U
− = 0
r dr2 ~2 r
2
d U m 2 c2
2
− 2 U = 0
dr ~
and this has exponential solutions,  mc 
U = U0 exp ± r
~
Choosing the decaying exponential for our solution, we have
k r
V = − e− a
r
~
where a = mc , the reduced Compton wavelength of a particle of mass m.

1
2 Bound orbits
We know from our general results that the conserved energy and angular momentum are given by

1 2 L2 k r
E = µṙ + − e− a
2 2µr2 r
2
L = µr ϕ̇

and from the energy expression we see that the radial motion is described by the effective potential

L2 k r
Vef f = 2
− e− a
2µr r
Bound orbits exist if there is a minimum of the effective potential:

L2 k r
V (r) = − e− a
2µr2 r
dV
0 =
dr
L2 k r k r
= − 3 + 2 e− a + e− a
µr r ar
L2
 
k k r
= − 3+ + e− a
µr r2 ar

This has solutions iff

L2  r −r
= kr 1 + e a
µ a
Since the right side is positive definite, there is always a value of L small enough that the equation is satisfied.
Computing the second derivative at this point, we have

L2
 
k k r
V0 = − 3 + + e− a
µr r2 ar
d2 V 3L2
   
2k k −ar 1 k k r
= + − 3 − 2 e − + e− a
dr2 µr4 r ar a r2 ar
3L2
 
k 2 2 1 r
= − + + e− a
µr4 r r2 ar a2
 
3  r −r k 2 2 1 r
= kr 1 + e a − + + e− a
r4 a r r2 ar a2
  
k 3 3 2 2 1 r
= + − + + e− a
r r2 ar r2 ar a2
r2
 
k r r
= 1 + − 2 e− a
r3 a a

so we have a minimum provided

a2 + ar − r2 > 0
This is satisfied as long as
√

a 1+ 5
r <
2

2
 r
and always if r < a. We take r < a in the following, since it allows us to expand in powers of a, and
guarantees that the extremum is a minimum.
r
Notice that to first order in ar , e− a ≈ 1 − ar and the circular orbits lie at approximately

L2  r −r
= kr 1 + e a
µ a
 r r
≈ kr 1 + 1−
 a a
 r 2 
= kr 1 +
a
≈ kr

the same value as for the Newtonian potential.

3 Nearly circular orbits

Now consider the precession of nearly circular orbits in the Yukawa potential,
k r
V (r) = − e− a
r
We know from our general results that the conserved energy and angular momentum are given by

1 2 L2 k r
E = µṙ + − e− a
2 2µr2 r
2
L = µr ϕ̇

For circular orbits, r = r0 , so

L2 k r0
E = 2 − e− a
2µr0 r0
L = µr02 ϕ˙0

Combining with the condition for a minimum,

L2  r −r
= kr 1 + e a
µ a
this gives
 
1  r0  k r0
E = 2 kr 0 1 + − e− a
2r0 a r0
k  r0  − r0
= − 1− e a
2r0 a
L = µr02 ϕ˙0

Therefore, both L and E are determined by r0 ,

k  r0  − r0
E = − 1− e a
2r a
r 0 
r0  − r0
L = µkr0 1 + e a
a

3
or correct to second order,
 
k  r0  r0 1  r0 2
E = − 1− 1− +
2r0 a a 2 a
 
k 2r0 3 r0
  2
= − 1− +
2r0 a 2 a
r s 
r0 1  r0 2
L = µka 1−
a 2 a
 
p 1 r0 2

= µkr0 1 −
4 a
This also gives us the frequency of the circular orbit,

ω0 = ϕ˙0
√ 
2 
µkr0 1 − 14 ra0
=
µr02
s  
k 1  r0 2
= 1−
µr03 4 a

Now suppose we give the system slightly higher energy by instantaneously increasing the angular mo-
mentum by δ. Then

L = µr02 ϕ˙0 + δ
= L0 + δ
2
(L0 + δ) k r0
E = 2 − e− a
2µr0 r0
2
L0 k r0 2L0 δ + δ 2
= 2 − e− a +
2µr0 r0 2µr02
2L0 δ + δ 2
= E0 +
2µr02
Therefore, at general r, ϕ,

2L0 δ + δ 2 1 2 L2 k r
E0 + = µṙ + − e− a
2µr02 2 2µr2 r
The minimum of the potential is still given by

L2  r1  − r1
= kr1 1 + e a
µ a
Expanding about r1 , so that r = r1 + ε with ε  r1 < a

2L0 δ + δ 2 1 2 L2 k r
E0 + = µṙ + − e− a
2µr02 2 2µr 2 r
1 2 L2 k r1 +ε
= µε̇ + 2 − e− a
2 2µ (r1 + ε) r1 + ε
2
1 2 L k
 
r
− a1 1+ rε
= µε̇ + 2 −   e 1
2

2µr12 1 + rε1 r1 1 + rε1

4

r1
1 2 kr1 1 + ra1 e− a k
 
r
− a1 1+ rε
= µε̇ + 2 −  e 1
2

2r12 1 + rε1 r1 1 + rε1
 
r
− a1 r1


1 2 ke 1 + 1 r 1 ε
= µε̇ +
 a
2 −   e − a r1 
2 r1
  
ε
2 1 + r1 ε 1 + r1
r1  2 !
1 2 ke− a 1  r1  2ε ε
= µε̇ + 1+ 1− +3
2 r1 2 a r1 r1
r1  2 ! 2 !
ke− a

ε ε r1 ε 1 r1 ε
− 1− + 1− +
r1 r1 r1 a r1 2 a r1
r1  2 !
1 2 ke− a 1  r1   r1  ε ε
= µε̇ + 1+ −2 1+ +3
2 r1 2 a a r1 r1
r1   2 !
ke− a 1 r12

 r1  ε r1 ε
− 1− 1+ + 1+ +
r1 a r1 a 2 a2 r1
r1   2 !
1 2 ke− a r2

1 r1   r1  r1  ε 1 r1 ε
= µε̇ + − 1− + 1+ − 1+ + 1+ − 12
2 r1 2 a a a r1 2 a a r1
r1 r1  2
r1  ke− a r2 ke− a
 
1 2 1 1 r1 ε
= µε̇ − 1− + 1+ − 12
2 2 a r1 2 a a r1 r1

so that r1 r1  2
2L0 δ + δ 2 r1  ke− a r12 ke− a
 
1 1 1 r1 ε
E0 + + 1 − = µε̇2 + 1+ − 2
2µr02 2 a r1 2 2 a a r1 r1
This has the general form
1 2 1 2 2
Eδ = µε̇ + κ ε
2 2
and this is the energy of a simple harmonic oscillator, oscillating around a point at r1 with angular frequency
s r1
r2 ke− a

r1
ω1 = 1+ − 12
a a µr13

while the orbital frequency is

r  r1  − r1
L = µkr1 1 + e a
a
r
1  r1  − r1
ϕ̇1 = µkr 1 1 + e a
µr2 a
r
1  r1  − r1
≈ µkr 1 1 + e a
µr12 a

The ratio of frequencies is

v
u  − r1
r2
u 1 + ra1 − a12 keµr3 µ2 r14
u a
ω1 1
≈ t
r1
ϕ̇1 µkr1 1 + ra1 e− a

5
 v
u 1 + r1 − r12
u
a a2
= t
1 + ra1
s
r2
 
r1 r1  r1 2
≈ 1+ − 12 1− +
a a a a
r
r2
≈ 1 − 12
a
1 r12
≈ 1−
2 a2
and this varies continuously with r1 and is therefore generically irrational, so the orbit doesn’t close. The
oscillation frequency in r is less than the orbital frequency, so the perigee advances. The advance per orbit
r2
is π a12 .