-1

I'm reading the $3^{\mathrm{rd}}$ edition of Sakurai and Napolitano's Modern Quantum Mechanics, and I have a brief question about the notation used to describe the eigenstates of the harmonic oscillator.

The authors define the usual creation ($a^{\dagger}$) and annihilation ($a$) operators, and define the operator $N = a^{\dagger}a$.

They then label the eigenvectors of $N$ by their corresponding eigenvalue $n$, i.e., $N|n\rangle = n|n\rangle$.

They go on to note that

\begin{equation*} Na|n\rangle = (n-1)a|n\rangle, \end{equation*}

(this is clear) and that this implies that $a|n\rangle$ and $|n-1\rangle$ are the same up to a multiplicative constant.

What is the best way to see this? If the eigenvectors of $N$ are labeled by their corresponding eigenvalue, is it true that

\begin{equation*} (n-1)a|n\rangle = a(n-1)|n\rangle = a|n-1\rangle = Na|n\rangle? \end{equation*}

Even in that case, how does it follow that $a|n\rangle$ and $|n-1\rangle$ are the same up to a multiplicative constant? I suspect this question might be deeper than just notation, but I'm not sure enough to be certain.

Edit: in response to the first (and second) downvote, how can I improve the question? From a mathematical perspective, it's not clear to me how the result I'm asking about holds. I come from a background of math and not physics, so I apologize for any foregone conclusions I might be missing here.

kandb
  • 363
  • 1
    WP. Normalize states and examine norms. – Cosmas Zachos Feb 09 '23 at 12:21
  • do you understand why $Na|n\rangle = (n-1) a|n\rangle$ and only don't understand how it implies $a|n\rangle \propto |n-1\rangle$? or do you ask why $Na|n\rangle = (n-1) a|n\rangle$? –  Feb 09 '23 at 12:22
  • I understand that $Na|n\rangle = (n-1)a|n\rangle$. I'm just not sure how $a|n\rangle \propto |n-1\rangle$. – kandb Feb 09 '23 at 12:24
  • 1
    @CosmasZachos what? – kandb Feb 09 '23 at 12:26
  • 1
    Why do you say $a(n-1)|n \rangle =a |n-1 \rangle$? This doesn't make sense. – John Feb 09 '23 at 12:53
  • That's one of the things I'm not sure of, and that's why I ask about the correctness of that step. It doesn't make sense to me either, but maybe if $|n\rangle$ doesn't specify any particular eigenvector, then it works... – kandb Feb 09 '23 at 13:02
  • 2
    The equation $\hat{N}\hat{a}|n\rangle = (n-1)\hat{a}|n\rangle$ tells you that $\hat{a}|n\rangle$ is an eigenstate of $\hat{N}$ with eigenvalue $n-1$. $|n-1\rangle$ is defined as being a (normalized) eigenstate of $\hat{N}$ with eigenvalue $n-1$. Therefore, $\hat{a}|n\rangle$ and $|n-1\rangle$ are the same up to a normalization factor. – Gaussian97 Feb 09 '23 at 13:11

3 Answers3

2

You may check that, since the 1D oscillator spectrum is non-degenerate, the eigenstates of N are unique, and normalized as $\langle n| n\rangle$.

As invited in the comment, define $$ |n\rangle = \frac{(a^\dagger)^n}{\sqrt{n!}} |0\rangle ~~~~ \leadsto ~~~~N|n\rangle = n|n\rangle \\ \langle n | aa^\dagger | n \rangle = \langle n|\left([a, a^\dagger] + a^\dagger a\right)| n \rangle = \langle n|(N + 1)|n\rangle = n + 1 \\ \Rightarrow a^\dagger | n\rangle = \sqrt{n + 1} | n + 1\rangle \\ \Rightarrow|n\rangle = \frac{a^\dagger}{\sqrt{n}} | n - 1 \rangle = \frac{(a^\dagger)^2}{\sqrt{n(n - 1)}} | n - 2 \rangle = \cdots = \frac{(a^\dagger)^n}{\sqrt{n!}}|0\rangle. $$

Cosmas Zachos
  • 62,595
1

$[N,a] = -a$

then, $Na|n\rangle = aN|n\rangle - a|n\rangle = an|n\rangle - a|n\rangle = (n-1)a|n\rangle$

We can think of $N$ as operator, $a|n\rangle$ as eigenstate and $(n-1)$ as eigenvalue.

In other words, it means the 'lowering operator' lowers the number $n$ of eigenstate (oscillator).

In conclusion, $a|n\rangle \propto |n-1\rangle$

  • How does $|n-1\rangle$ figure in your calculations? – kandb Feb 09 '23 at 12:41
  • @kandb From $N | n-1\rangle = (n-1)|N-1\rangle $, you could find the analogy – Sasha Shin Feb 09 '23 at 12:42
  • I agree, but how does that justify that $a|n\rangle \propto |n-1\rangle$? Tenuously interpreted analogies shouldn't be part of science... – kandb Feb 09 '23 at 12:44
  • @kandb How about this? In that system, there is only one energy correspondent to one state. For state$|n\rangle$, acting $N$ operator, you can get n for eigenvalue. Acting $(Na)$ operator, you can get $(n-1)$ for eigenvalue. In other words, the state$|n\rangle$ is changed by 'a' operator. The changed state coincides with state $|n-1\rangle$ up to constant. – Sasha Shin Feb 09 '23 at 13:05
1

We label the eigenstates of the Harmonic Oscillator by the respective eigenvalues of the number operator $N=a^{\dagger}a$. So the state $|n\rangle$ means that $N|n\rangle = n |n\rangle$ etc. Similarly $N|n-1\rangle = (n-1) |n-1\rangle$. Since the eigenvalues of the harmonic oscillator (and therefore the number operators) are non-degenerate it means that any state $|\psi\rangle$ that maintains $N|\psi\rangle = (n-1)|\psi\rangle$ must be the state $|n-1\rangle$, up to some constant multiplicative factor. So if we have $N|\psi\rangle = (n-1)|\psi\rangle$ we can conclude that $|\psi\rangle = C |n-1\rangle$ with $C$ some complex number.

As was shown in other answers, the commutation relations imply that $Na|n\rangle = (n-1)a|n\rangle$ meaning that $a|n\rangle$ is the eigenstate of $N$ with eigenvalue $(n-1)$ up to some constant multiplicative factor. The only missing piece is that this state is unique, in other words that the spectrum is non-degenerate, and a proof for that you can see in the linked answer.