In the Bargmann-Fock case, there is an analog of the distributional states
|q〉, given by taking states that are eigenvectors fora, but unlike the|α〉, are
not normalizable. We define
|δw〉=ewa
†
| 0 〉=
∑∞
n=0
wn
√
n!
|n〉=ewz=e
|w|^2
(^2) |w〉
Instead of equation 23.3, such states satisfy
〈δw 1 |δw 2 〉=ew^1 w^2
The|δw〉behave in a manner analogous to the delta-function, since the Bargmann-
Fock analog of computing〈q|ψ〉using the function space inner product is, writ-
ing
ψ(w) =
∑∞
n=0
cn
wn
√
n!
the computation
〈δz|ψ〉=
1
π
∫
C
ezwe−|w|
2
ψ(w)d^2 w
=
1
π
∫
C
ezwe−|w|
2 ∑∞
n=0
cn
wn
√
n!
d^2 w
=
1
π
∫
C
∑∞
m=0
zm
√
m!
wm
√
m!
e−|w|
2 ∑∞
n=0
cn
wn
√
n!
d^2 w
=
∑∞
n=0
cn
zn
√
n!
=ψ(z)
Here we have used the orthogonality relations
∫
C
wmwne−|w|
2
d^2 w=πn!δn,m (23.4)
It is easily seen that the Bargmann-Fock wavefunction of a coherent state is
given by
〈δz|α〉=e−
|α|^2
(^2) eαz (23.5)
while for number operator eigenvector states
〈δz|n〉=
zn
√
n!
In section 23.5 we will compute the Bargmann-Fock wavefunction〈δz|q〉for
position eigenstates, see equation 23.13.
Like the|α〉(and unlike the|q〉or|p〉), these states|δw〉are not orthogonal
for different eigenvalues ofa, but they span the state space, providing an over-
complete basis, and satisfy the resolution of the identity relation
1 =
1
π
∫
C
|δw〉〈δw|e−|w|
2
d^2 w (23.6)