Suppose I have a particle of unspecified spin whose states are determined by a single quantum number $k=1,...,N$. In standard quantum mechanics notation, the state such that the particle is in a superposition of all possible states is given by $$|\varphi\rangle=\frac{1}{\sqrt N}\sum_{k=1}^{N}|k\rangle.$$ Would it make sense to use second quantization to describe the same state? So far I've only seen this formalism when dealing with many-body systems. In this case, it would perhaps look something like this: $$a_1^\dagger|0\rangle+...+a_N^\dagger|0\rangle,$$ if $a$ is the annihilation operator for the particle and $|0\rangle$ the vacuum state. There are several points of which I'm unsure:

  • Can I still use the formalism if, a priori, I don't know the spin of the particle?
  • Can I still add different amplitudes to the formal sum $a_1^\dagger|0\rangle+...+a_N^\dagger|0\rangle$, in the form $\alpha_1a_1^\dagger|0\rangle+...+\alpha_Na_N^\dagger|0\rangle$?
  • Does this make sense at all?
  • $\begingroup$ A requirement of QFT (/second quantization) is that it be able to reduce to standard QM. As such, I think the idea behind this question should work in theory. See the definition of the field operator in QFT when expressed in ~momentum basis. $\endgroup$
    – PrawwarP
    Dec 7, 2020 at 14:47
  • $\begingroup$ Does the action of the quantised field operators not do this? $$\hat\phi(x)|0\rangle=\int d^3\tilde p\text{ } (a^{\dagger}(p)e^{-ipx}+a(p)e^{ipx})|0\rangle=\int d^3\tilde p\text{ } a^{\dagger}(p)e^{-ipx}|0\rangle.$$ $\endgroup$
    – Charlie
    Dec 7, 2020 at 14:49
  • $\begingroup$ Charlie's formula can also generalize to apply with amplitudes different from plane waves (though, if I recall correctly, they still have to be solutions of Schrodinger equation [although, I may be too restrictive in my understanding of that memory]) $\endgroup$
    – PrawwarP
    Dec 7, 2020 at 14:53
  • $\begingroup$ This proposal seems pretty different to me from the ordinary 2nd quantization scenario. Here we're taking a Hilbert space spanned by N orthogonal states, and trying to recast it as a Hilbert space spanned by 2^N orthogonal states. Unless I'm missing something, how could vector spaces with a different number of finite dimensions be isomorphic to each other? Ordinarily, we're starting with a direct product of a countably infinite number of uncountably infinite-dimensional Hilbert spaces, and just decomposing that into a direct sum of different sectors each with a well-defined particle number. $\endgroup$ Dec 10, 2021 at 3:32
  • $\begingroup$ The second one doesn't appear to change the dimensionality, while the first one does. $\endgroup$ Dec 10, 2021 at 3:37

1 Answer 1


Yes, it is possible to use second quantization for single-particle problems in the way outlined in the question. There are a few points to pay attention to:

  • In most cases this will be redundant, since the formalism is designed expressly for treating many-particle problems, taking into account the statistics of fermions and bosons. But there are exceptions.
  • Depending on the calculation technique, one may have or may not have to pay attention to constraining the total number of particles. For example, equation-of-motion techniques, applying to a particle conserving Hamiltonian, typically pass without problems. However, all kind of statistical averages do require imposing a constraint - there is in fact a number of techniques used to impose such constrains (although normally sued in more complex settings), such as slave-boson approach, drone-fermions, Schwinger boson, etc.

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.