Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

When dealing with the Quantum Harmonic Oscillator Operator $H=-\frac{d^{2}}{dx^{2}}+x^{2}$, there is the approach of using the Ladder Operator:

Suppose that are two operators $L^{+}$ and $L^{-}$ and define $f_{n}$ such that



then $$L^{+}L^{-}(f_{n})=L^{+}(\sqrt{n}f_{n-1})=nf_{n}$$ $$L^{-}L^{+}(f_{n})=L^{-}(\sqrt{n+1}f_{n+1})=(n+1)f_{n}$$



Now since $$H=-\frac{d^{2}}{dx^{2}}+x^{2}=\left(x+\frac{d}{dx}\right)\left(x-\frac{d}{dx}\right)$$



If we let $$L^{+}=\frac{1}{\sqrt{2}}\left(x+\frac{d}{dx}\right)\quad\text{and}\quad L^{-}=\frac{1}{\sqrt{2}}\left(x-\frac{d}{dx}\right)$$

then $$H=L^{+}L^{-}+L^{-}L^{+}$$

Since $$L^{+}(f_{n})=\sqrt{n+1}f_{n+1}\implies$$


Thus if we can a $f_{0}$ such that

$$f_{n}=\sqrt{n!}L^{n}(f_{0})\implies L^{-}(f_{n})=\sqrt{n}f_{n-1}\tag{1}$$ would be satisfied, then we have found a set of eigenfunctions for H, which is $f_{n}$ and the corresponding eigenvalue is 2n+1.

Now my instructor says that if we let $f_{0}$ be in the kernel of $L^-$, meaning that:

$$L^{-1}f_{0}=0\tag{2},$$ then (1) would be true.

Specifically, this means that


As it is already verified (and commonly used in quantum mechanics) that this $f_{0}(x)$ does satisfy (1). But I can't really see how does (2) leads to (1). Is this choice $f_0$ really a lucky guess or actually a choice the generally works for all operators that can be expressed as $H=L^{+}L^{-}+L^{-}L^{+}$? A different way of asking this question is: are the ladder operator really just operators that tells us how to get from one state to another, or it is actually something that would be used as a general method of solving some differential equations?

share|cite|improve this question
Are you sure about your definition of $L^-$? Check this! – draks ... Jan 16 '12 at 9:59
The line $$H=-\frac{d^{2}}{dx^{2}}+x^{2}=\left(x+\frac{d}{dx}\right)\left(x-\frac{d}{dx}\‌​right)$$ is wrong (though what comes after it is right). You can't factor $H$ like this because $x$ and $\mathrm d/\mathrm dx$ don't commute. Also I believe you've mixed up $L^+$ and $L^-$; see my answer. – joriki Jan 18 '12 at 17:19
I've never understood this either: having an operator wrote down as $H = L^{+}L^{-}$, only knowing the commutation rules of the "ladder operators", is it possible to have multiple sequences of eigenfunctions starting from different eigenfunctions $f_{0,i}$ being annihilated by $L^{-}$? (so that you can't say you obtain $f_n=\sqrt{n!}L^n(f_0)$ for the most general $f_n$). – user115065 Dec 11 '13 at 7:12
up vote 1 down vote accepted

I would say everything you need for (1) is given in (2): From $L^-$ you get $L^+$ and $f_0$ is given by (2).

If it's possible to express your Hamiltonian as ladder operators, this approach would also help to simplify the way to get the solution of your differential equation. They are for example also used in (finite dimensional) spin physics.

Concerning your question, if $f_0$ is just a lucky guess, you can check here, that the Hermite function $f_n$ fulfill the recursion relations $$ x\;f_{n}(x) = \sqrt{\frac{n+1}{2}}f_{n+1}(x) + \sqrt{\frac{n}{2}}f_{n-1}(x) $$ and $$ f_n'(x) = -\sqrt{\frac{n+1}{2}}f_{n+1}(x) + \sqrt{\frac{n}{2}}f_{n-1}(x) . $$ and

Together with definition of $L^-=\frac{1}{\sqrt{2}}(x+\frac{d}{dx})$, you see that $L^- f_0$ exactly cancels to $0$.

share|cite|improve this answer

I think the confusion begins where you write "and define $f_n$ such that" at the very beginning. The two lines after that are two ways of defining $f_n$, and part of what needs to be shown is that they're equivalent.

Also you seem to have got $L^\pm$ mixed up -- the one that contains the plus sign is the annihilation operator and the one that contains the minus sign is the creation operator, as you can tell from the fact that your $f_0$ is actually in the kernel of what you called $L^+$. I'll use your definitions so as not to confuse things further, but keep in mind that that means that $L^+$ is now the annihilation operator and $L^-$ is now the creation operator.

I'd proceed as follows. The commutator of the ladder operators is


If $f_n$ is an eigenfunction of $H$ with eigenvalue $E_n$, then $L^-f_n$ is an eigenfunction of $H$ with eigenvalue $E_n+2$:

$$ \begin{eqnarray} HL^-f_n&=&(L^+L^-+L^-L^+)L^-f_n \\ &=& (L^+L^-L^-+L^-L^+L^-)f_n \\ &=& (L^-L^+L^-+L^-+L^-L^-L^++L^-)f_n \\ &=& L^-(L^+L^-+1+L^-L^++1)f_n \\ &=& L^-(H+2)f_n \\ &=& L^-(E_n+2)f_n \\ &=& (E_n+2)L^-f_n\;. \end{eqnarray} $$

In the same way we can show that $L^+f_n$ is an eigenfunction of $H$ with eigenvalue $E_n-2$.

Now the eigenvalues of $H$ are non-negative. (You can show this for instance by taking the expectation value of $H$ and integrating by parts.) Thus successively applying $L^+$ to lower the eigenvalue must at some point annihilate the eigenfunction. Thus we've shown a) that any eigenfunction gives rise to an entire succession of eigenfunctions with equally spaced eigenvalues and b) this succession must end with a function that is annihilated by $L^+$, and this yields a differential equation whose solution is your $f_0$.

Then it only remains to normalize the resulting eigenfunctions, and to derive the non-recursive forms with the factorial factors becomes a straightforward exercise in induction.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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