Tens or Operators We
Tens or Operators We
D† Vi D = Rij Vj .
RV = V + δθn̂ × V.
[Vx , Jz ] = −ih̄Vy
1
or more generally
[Vi , Jj ] = ih̄ijk Vk
We can take that last as the definition of a vector operator. A rotation of a vector
operator is accomplished by computing
D† VD
It is convenient if we can write the operator in a basis of angular momentum eigen-
states since we know how to write D in that basis. So we might write the position
operator as
s
r 8π 1
x = (Y1 + Y1−1 ) s
2 3 3 z
s Y10 =
r 8π 1 4π r
y = (Y − Y1−1 ) s
2i 3 1 3 x ± iy
s Y1±1 = ∓ √
4π 0 4π 2r
z = r Y
3 1
√
where r = x2 + y 2 + z 2 . Some examples of vector operators are momentum, posi-
tion, and angular momentum. If we write our vector using spherical harmonics as a
basis then our definition of a vector operator reads
D† Vm D = Dmm
∗1
0 Vm0
where m = ±1, 0.
q1 ,q2
m1 ,m2
And how does a spherical harmonic transform under rotations? First let’s start with
an example. Let’s suppose that we have a cartesian vector operator V = (Vx , Vy , Vz ).
We can also write it as V±1 = 21 (Vx ± iVy ), V0 = Vz . Let’s consider an infinitesimal
rotation of the vector about the z-axis. We will rotate the cartesian version using an
orthogonal matrix (SO(3)) and the spherical version using a representation of SU(2).
The SO(3) rotation of the cartesian vector is here
1 − 0 Vx Vx − Vy
0
V = Rz ()V ∼ 1 0 Vy = Vx + Vy (1)
0 0 1 Vz Vz
3
0 0
Converting Vspherical to Vcartesian we get
0
V10 + V−1
V1 + V−1 − i(V1 − V−1 ) Vx + Vy
0 0 0
Vcart = −i(V1 − V−1
)
=
−i(V1 − V −1 − i(V1 + V ))
−1
= y − Vx
V
(2)
V0 V0 Vz
Comparing Equation 1 with 2 we see that the SU(2) representation of the rotation
by is D(−) when the SO(3) representation is Rz (). We see this again as follows
We can turn Ylm (n̂0 ) into a more general vector operator V by setting V± = |V |Y1±1
and V0 = |V |Y10 .
0
→ D† (R)Ylm (V)D(R) = Ylm (V)Dm
l −1
X
0 m (R )
m0
0
Ylm (V)D∗ lmm0 (R)
X
=
m0
m0 m0
On the left we rotate the state by R and then measure by taking the expectation
value. On the right we first measure the expectation value of each component of the
operator and then rotate those expectation values by R−1 . Then an infinitesimal
rotation gives us
i i X i
k 0
(1 + J · n̂θ)Tq (1 − J · nθ) = km | (1 + J · n̂) | kq Tmk 0
h̄ h̄ m0
h̄
iX
= Tqk + hkm0 | J · n̂θ | kqi Tmk 0
h̄ m0
→ [J · n̂θ, Tqk ] = hkm0 | J · n̂θ | kqi Tmk 0
X
m0
4
If n̂ = ẑ then
[Jz , Tqk ] = qh̄Tqk (3)
or if n̂ = √1 (n̂x ± in̂y ),
2
Xq q
[J± , Tqk ] = h̄ k(k + 1) − q(q ± 1)hkm0 | kq ± 1iTmk 0 = h̄ k(k + 1) − q(q ± 1)Tq±1
k
m0
(4)
Selection Rule
We show that
D E
α0 , j 0 , m0 | Tqk | α, j, m = 0, unless m0 = q + m
5
We can rotate the pieces by applying the rotation operator and we have
X
D(R)| j, mi = D1 (R)D2 (R)| j1 , m1 , j2 , m2 ihj1 , m1 , j2 , m2 | j, mi
or we could write
X
D(R)| j, mi = D1 (R)D2 (R)| j1 , m1 i| j2 , m2 ihj1 , m1 , j2 , m2 | j, mi
and we get
j j1 j2 0 0
X X
Dm 0 ,m = Dm 0 ,m Dm0 ,m hj1 , j2 , m1 , m2 | j, mihj1 , m1 , j2 , m2 | j, mi
1 2
(5)
1 2
m01 ,m02 m1 ,m2
Clebsch-Gordan Series
If we invert the above we have
X
| j1 , m1 , j2 , m2 i = | j, mihj, m | j1 , m1 , j2 , m2 i
j,m
Then rotation and multiplication from the left by the dual gives
j1 j2
hj 0 , m0 | D | j, mi hj1 , m01 , j2 , m02 | j 0 , m0 ihj, m | j1 , m1 , j2 , m2 i
X
Dm 0 ,m Dm0 ,m =
1 2
1 2
j,m,j 0 ,m0
This is called the Clebsch Gordan series or the Kronecker or direct product of rep-
resentations and we can write
or
Dj1 +j2 0 ··· 0
0 Dj1 +j2 −1 ··· 0
Dj1 ⊗ Dj1 = .. .. .. = | 00ih00 | αi
.
. .
0 ··· Dj1 −j2
6
Integration over rotations
We want to consider
Z XZ
dRD(R)| αi = dR D(R)| j, mihj, m | αi
j,m
where Z 4π
Z
dα Z 4π dγ Z π dβ sin β
dR = =1
0 4π 0 4π 0 2
Then Z
j
dR Dmm 0 = δm,0 δm0 ,0 δj,0
This follows because integration over all angles averages over all directions and the
average is zero except for the that state that has no direction, namely j = m = 0.
Now we can use this result along with Equation 6 and write
Z Z
j1 j2 j 0 0 0
X
Dm 0 m Dm0 m dR = dR Dm 0 ,m hj1 , m1 , j2 , m2 | j, m ihj, m | j1 , m1 , j2 , m2 i
1 1 2 2
j,m,m0
= hj1 , m01 , j2 , m02 | 0, 0ih0, 0 | j1 , m1 , j2 , m2 i
It is clear that j1 and j2 must be equal and m1 = −m2 for the coefficients to be non
zero. If for example j1 = j2 = 2 then
No matter how we rotate the state there will be equal parts of all five components so
it must be that all the coefficients are equal magnitude. Then for the normalization
to be right
1
hj, m, j, −m | 0, 0, j, ji = √
2j + 1
and we can write
Z
j1 j2
Dm 0 m Dm0 m dR
1 2
= hj1 , m01 , j2 , m02 | 0, 0ih0, 0 | j1 , m1 , j2 , m2 i
1 2
7
Then using D(R−1 ) = D† (R) we can see that
j 0
Dm,m 0 = (−1)
m−m
D∗ j−m,−m0
J ∗
and multiply both sides by DM M 0 and integrate over R and use Equation 7
Z Z
j1 j2 J ∗ X j 0 0 0 J ∗
dR Dm 0 ,m Dm0 ,m DM M 0
1 2
= dR Dm 0 ,m hj1 , m1 , j2 , m2 | j, m ihj, m | j1 , m1 , j2 , m2 iDM M 0
1 2
j,m,m0
δm0 ,M δm,M 0 δj,J
hj1 , m01 , j2 , m02 | j, m0 ihj, m | j1 , m1 , j2 , m2 i
X
=
j,m,m0
2j + 1
1
= hj1 , m01 , j2 , m02 | J, M ihJ, M 0 | j1 , m1 , j2 , m2 i (8)
2J + 1
where we have used
j ∗ m−m j 0
Dm,m 0 (R) = (−1) D−m,−m 0 (R).
8
Wigner Eckart Theorem
Spherical tensor operators transform according to
q0
†
Tqk D| 0 0
D∗ kqq0 hjm |Tqk0 | j 0 m0 i
X
hjm |D jmi =
q0
hjm |D† | jm1 ihjm1 |Tqk | j 0 m2 ihj 0 m2 |D| j 0 m0 i = D∗ kqq0 hjm |Tqk0 | j 0 m0 i
X X
m1 m2 q0
j j0
D† mm1 hjm1 |Tqk | j 0 m2 iDm D∗ kqq0 hjm |Tqk0 | j 0 m0 i
X X
2m
0 =
m1 m2 q0
−1 j ∗k j0
hjm1 |Tqk | j 0 m2 i = D† |Tqk0 | j 0 m0 iD−1 m0 m2
X
m1 m D qq 0 hjm
q 0 ,m,m0
0
j
D∗ kqq0 hjm |Tqk0 | j 0 m0 iD∗ jm2 m0
X
= Dm 1m
q 0 ,m,m0
9
D
0 0
E hjm1 | k, q, j 0 , m2 i X
α, j, m1 | Tqk | α ,j , m02 = hαjm |Tqk0 | α0 j 0 m0 ihkq 0 j 0 m0 | jmi
2j + 1 q 0 ,m,m0
With the sum over q 0 , m and m0 we can rewrite that last equation as
D E hjm1 | k, q, j 0 , m2 i
α, j, m1 | Tqk | α0 , j 0 , m02 = √ hαj ||T k || α0 j 0 i
2j + 1
cj (α) is a number that depends only on j, j 0 and α. α represents aspects of the state
that do not depend on orientation, like the radial dependence of a wave function. So
for an initial and final
D
state with j, j 0 , α, α0 E
and a spherical tensor operator Tqk , once
we have computed α, j, m1 | Tqk | α0 , j 0 , m02 for a particular m1 , m2 and q, we can
with the help of the Clebsch Gordon coefficient on the right, determine the reduced
matrix element cj,j 0 α,α0 . Then we compute any of the others by multiplying by the
appropriate Clebsch Gordon coefficient.
We see that this is the same recursion relationship as for the Clebsch Gordon coef-
ficients. Therefore it is reasonable to conclude that the matrix elements are all pro-
portional to Clebsch Gordon coefficients which is precisely what the Wigner Eckart
10
Theorem says, namely
D E
D
0 0 0 k
E
0 0
α0 j 0 | |T k | | α, j
α , j , m | Tq | α, j, m = hjmkq | j , m , j, ki √ 0
2j + 1
A = A0 ek·r
E2ij = pi rj
Implications
Matrix elements of a scalar operator T00 are zero unless m = m0 and j = j 0
11
Example of Wigner Eckart symmetry
Consider the dipole operator ~r. The matrix element
Z
hf | ~r | ii = Rf∗ (r)Ylf∗ ,mf (rY1,m )Ri (r)Yli ,mi d3 r
Z Z
= Rf∗ (r)rRi (r)r2 dr Ylf∗ ,mf (Y1,m )Yli ,mi dΩ
v
u (2li
+ 1)(2l + 1)
Z u
= Rf∗ (r)rRi (r)r2 drhli , mi , 1, m | lf , mf ihli , 0, 1, 0 | lf , 0it
4π(2lf + 1)
v
u (2li + 1)(2l + 1) Z ∗
u
= hli , mi , 1, m | lf , mf ihli , 0, 1, 0 | lf , 0it Rf (r)rRi (r)r2 dr
4π(2lf + 1)
1 hf, lf | |T 1 | | i, li i
rm = hli , mi , 1, m | lf , mf i q
2lf + 1)
where
r± = rY1,±1 (θ, φ), r− = rY1,0 (θ, φ)
Projection theorem
The theorem shows that an expectation value taken between states with the same
j, that a rank one tensor can be written as λJ where λ is some constant that is
independent of the z-component of angular momentum of the initial and final states.
D
0
E hα0 , jm | J · V | α, jmi
α , jm | Vq1 | α, jm = 2 hjm0 | Jq | jmi .
h̄ j(j + 1)
First note that J · V = J0 V0 − J+ V− − J− V+ Then
hα0 , j, m | J · V | α, j, mi = mh̄ hα0 , jm | V0 | α, j, mi
h̄ q
− j(j + 1) − m(m − 1) hα0 , jm − 1 | V− | α, jmi
2
h̄ q
− j(j + 1) − m(m + 1) hα0 , jm + 1 | V+ | α, jmi
2
= cjm hα0 , j ||V|| αji
where the Wigner-Eckart theorem is used in the last step. But J · V is a scalar
operator so its expectation value can have no m dependence. So cjm → cj . We
might have done the same exercise with V → J and then we would get
D E
α, jm | J2 | α, jm = cj hα0 , j ||J|| α, ji
12
Meanwhile we could have written from the WE theorem
hα0 jm0 | Vq | α, jmi hα0 , j ||V|| α, ji
=
hα, jm0 | Jq | α, jmi hα, j ||J|| αji
or
hα0 jm0 | Vq | α, jmi hα0 , jm |J · V| α, jmi
=
hα, jm0 | Jq | α, jmi hα, jm |J · J| αjmi
which implies
0 0 hα0 , jm |J · V| α, jmi
hα jm | Vq | α, jmi = 2 hα, jm0 | Jq | α, jmi
h̄ j(j + 1)
13