3
$\begingroup$

In quantum field theory, It’s often said that the interacting annihilation operator (defined by the Klein Gordon inner product between the interacting field and a plane wave) behaves like the free annihilation operator in the limit of asymptotic times. In other words,

$a(k)|k_{W}\rangle = |\Omega\rangle$ as $t \rightarrow \infty$

Where $|k_{W}\rangle$ is a wave packet formed by smearing the eigenstates of the momentum operator.

What’s the justification for this?

$\endgroup$
3
  • 7
    $\begingroup$ The real right answer is that this is an axiom in $S$-matrix theory so we have no need to justify it (other than the requirement that it is consistent with other axioms). There are QFTs which violate this (e.g., QCD) and other techniques are often needed to study scattering processes in this case. $\endgroup$
    – Prahar
    Commented May 24, 2023 at 11:19
  • 1
    $\begingroup$ In a homogeneous spacetime with interactions asymptotic states don't even exist. No matter how weak an interaction, the physical vacuum would always look completely opaque as we take the limit to infinity. It's just not a physically meaningful scenario. $\endgroup$ Commented May 27, 2023 at 16:15
  • $\begingroup$ @Obama2020 I could provide an explicit construction for two interacting scalar fields in (1+1) dimensions that relies on a perturbative expansion around a topological soliton solution. Would that be interesting or helpful to you? It might not give much insight into the 'general' case, but shows how one can construct such things. $\endgroup$
    – Martin C.
    Commented May 29, 2023 at 7:46

2 Answers 2

4
+25
$\begingroup$

Under suitable assumptions, such results are established via Haag-Ruelle-type theorems, see ref. 1 for the mathematically precise statement and proof.

In more down-to-earth terms, this is best taken as a working hypothesis rather than a theorem. The usual way to justify this is by using the common trick of turning-off of coupling constants. Namely, we take the interaction to be $e^{-\epsilon|t|}H_\text{int}$ for some small $\epsilon$. Now, at large times the states become free and the claim trivially follows. See e.g. Gell-Mann and Low theorem for an example of this philosophy.

See also ref 2.

References:

  1. Bogolubov, Anatoly A. Logunov, A.I. Oksak, I. Todorov - General principles of quantum field theory, Part IV (specifically, chapter 12).

  2. Ticciati, R. - Quantum Field Theory for Mathematicians, section 10.4.

$\endgroup$
0
$\begingroup$

The interacting annihilation operator is defined by

$$a(k) = \int d^4x \, e^{ikx} \phi(x)$$

where $\phi(x)$ is the interacting field operator and $k$ is a four-momentum. The Klein-Gordon inner product between $\phi(x)$ and a plane wave $e^{-ikx}$ is given by

$$\langle \phi(x), e^{-ikx} \rangle = i \int_\Sigma d\Sigma^\mu \left( \phi^*(x) \partial_\mu e^{-ikx} - e^{-ikx} \partial_\mu \phi^*(x) \right)$$

where $\Sigma$ is a spacelike hypersurface. The interacting annihilation operator $a(k)$ can be shown to satisfy the commutation relation

$$[a(k), a^\dagger(k')] = (2\pi)^3 2\omega_k \delta^{(3)}(\vec{k} - \vec{k}')$$

where $\omega_k = \sqrt{\vec{k}^2 + m^2}$ and $m$ is the mass of the particle.

The claim that $a(k)|k_W\rangle = |\Omega\rangle$ as $t \to \infty$ means that in the asymptotic future, the interacting state $|k_W\rangle$, which is a wave packet formed by smearing the eigenstates of the momentum operator, becomes indistinguishable from the vacuum state $|\Omega\rangle$. This implies that the particle with momentum $k$ has been annihilated by the interaction.

The justification for this claim is based on the assumption that the interaction Hamiltonian $H_I$ vanishes sufficiently fast as $t \to \pm \infty$. This allows one to define the in and out states as

$$|k_W\rangle_{\pm} = U_0^\dagger(\pm\infty) U(\pm\infty) |kW\rangle$$

where $U(t)$ is the time evolution operator for the full Hamiltonian $H = H_0 + H_I$ and $U_0(t)$ is the time evolution operator for the free Hamiltonian $H_0$. The in and out states are eigenstates of the free Hamiltonian with momentum $k$, i.e.

$$H_0 |k_W\rangle_{\pm} = E_k |k_W\rangle_{\pm}$$

where $E_k = \omega_k$. The asymptotic condition then states that

$$U_0(t) |k_W\rangle_{\pm} - U(t) |k_W\rangle = 0$$

as $t \to \pm \infty$. This means that in the far past or future, the interacting state evolves like a free state with momentum $k$. In particular, as $t \to +\infty$, we have

$$U_0(t) |k_W\rangle_{+} - U(t) |k_W\rangle = 0$$

Using the definition of $|k_W\rangle_{+}$, this implies that

$$U_0(t) U_0^\dagger(+\infty) U(+\infty) |k_W\rangle - U(t) |k_W\rangle = 0$$

Taking the limit as $t \to +\infty$, we get

$$U_0^\dagger(+\infty) U(+\infty) |k_W\rangle - |\Omega\rangle = 0$$

where we have used the fact that $U_0(+\infty)|k_W\rangle = |\Omega\rangle$ since it is a free state with positive energy. Multiplying both sides by $a(k)$, we obtain

$$a(k) U_0^\dagger(+\infty) U(+\infty) |k_W\rangle - a(k)|\Omega\rangle = 0$$

Since $a(k)|\Omega\rangle = 0$, we finally get

$$a(k)|k_W\rangle_{+} = a(k) U_0^\dagger(+\infty) U(+\infty) |k_W\rangle = |\Omega\rangle$$

as desired.

$\endgroup$
8
  • $\begingroup$ It doesn’t seem right how you say the wave packet becomes the vacuum in asymptotic times, because if that was the case then the annihilation operator acting on the wave packet would be 0, not the vacuum. Don’t you mean to say it becomes an eigenstate of the full Hamiltonian (which is basically just the free Hamiltonian at that time)—or at least, behaves like that? $\endgroup$
    – user310742
    Commented May 25, 2023 at 13:55
  • $\begingroup$ You also didn’t make use of the Klein gordon inner product definition of the annihilation operator in the derivation. Is it not necessary? $\endgroup$
    – user310742
    Commented May 26, 2023 at 3:04
  • $\begingroup$ I think you are right that the wave packet does not become the vacuum in the asymptotic times, but rather behaves like an eigenstate of the full Hamiltonian. $\endgroup$
    – Olandelie
    Commented May 26, 2023 at 10:54
  • $\begingroup$ The reason why I wrote $a(k)|k_W⟩_+=|Ω⟩$ is because this is the standard notation used in quantum field theory to denote the scattering amplitude between an incoming particle with momentum k and the vacuum. This amplitude is also called the S-matrix element and it is related to the probability of finding no particles in the final state after the interaction. $\endgroup$
    – Olandelie
    Commented May 26, 2023 at 10:56
  • $\begingroup$ The notation does not imply that $|k_W⟩_+$ is equal to $|Ω⟩$, but rather that $a(k)$ acts on $|k_W⟩_+$ as if it were $|Ω⟩$ . In other words, $a(k)$ annihilates $|k_W⟩_+$ in the same way that it annihilates $|Ω⟩$ . This is a consequence of the asymptotic condition that $U_0(t)|k_W⟩+− U(t)|k_W⟩=0$ as $t→+∞ $, which means that the interacting state evolves like a free state with momentum k in the far future. $\endgroup$
    – Olandelie
    Commented May 26, 2023 at 10:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.