UCSD's online QM notes, as usual, starts stating that QM operators are Hermitian and says that operator $O$ elements can be computed by

$$O_{ij} = \langle u_j|O|u_i\rangle$$

The $u_i$ are eigenvectors, which are orthogonal to each other, thanks to being Hermitian. So, I suppose that $O$ is a Hermitian operator. But where are the eigenvalues?

I expect that when you apply the operator to its eigenvector, a $\lambda$ must appear. Yet, I see it nowhere. Neither, text says that $O$ is a diagonal, as I may expect if I apply $\langle u_i|$ to $O|u_i\rangle = \lambda_i |u_i\rangle$, I should get $\lambda_i$ on the main diagonal and 0 everyelse. Are $u_i$ eigenvectors of the $O$ or some another operator?

  • $\begingroup$ Put it in this way. Let $\hat{O}$ be an operator. Insert completeness relation $$\sum_i |u_i \rangle \langle u_i | = \hat{1}$$, we have $$\hat{O}=\sum_{ij} |u_i \rangle \underline{\langle u_i |\hat{O} |u_j \rangle} \langle u_j | $$. The underline part is the representation of operator $\hat{O}$ in $\{u_i\}$ basis. If $u_i$ is the eigenfunction of operator $O$, namely $\hat{O} |u_i \rangle = \lambda_i |u_i \rangle$ we have $\hat{O}=\sum_i \lambda_i | u_i \rangle \langle u_i| $, which sometimes called spactral expansion of operator. The matrix element is $$O_{ij}=\lambda_i \delta_{ij} $$ $\endgroup$ – user26143 Dec 10 '13 at 16:41
  • $\begingroup$ The eigenvalue $\lambda_j$ depends on the operator $\hat{O}$ and the eigenfunction $|u_j\rangle$. For example, the quantum harmonic oscillator $\hat{O}\equiv\hat{H}= \hat{p}^2/2m+m\omega^2\hat{x}^2/2$ with $\lambda_j\equiv E_n=(n+\frac12)\hbar\omega$. $\endgroup$ – Kyle Kanos Dec 10 '13 at 16:47
  • $\begingroup$ @user26143 I do not understand a couple of things. What is the completeness relation? $\endgroup$ – Val Dec 10 '13 at 20:15
  • $\begingroup$ It is a standard result in many textbooks of quantum mechanics. E.g. Sakurai's modern quantum mechanics, section 1.3, p19. $\endgroup$ – user26143 Dec 10 '13 at 21:19
  • $\begingroup$ @user26143 Do you mean that this is not a simple consequence of basis orthonormality? $\endgroup$ – Val Dec 10 '13 at 22:01

In those notes you linked to, the $|u_i\rangle$ form an orthonormal basis. They are not necessarily the eigenkets of the operator $O$. By choosing different bases $\{u_i\}$, the "representation" of the operator $O$ (e.g.., how you write it out in a matrix) will change.

You are correct that if one chooses the set of kets that are eigenkets of the operator, you'll get a diagonal matrix. But in those notes they are trying to keep it general.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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