I have read that if in the quantum harmonic oscillator, $n$ is very large, then the probability density is similar to the classical one. In the case of a simple harmonic oscillator:

$$P_{clas}=\frac{1}{\pi \sqrt{A^2-x^2}}$$ $$P_{qho}\propto\frac{1}{\sqrt{2n-1-x^2}}.$$

But what about the energies? I know that if i take the limit with n tends to infinite: $\frac{\Delta E_n}{E_n}\rightarrow 0$. So the energy spectrum becomes continuous as in the classical case. Is that the only thing you can know about energy?

If $A^2=2n-1$, then: $$E_{clas}=\frac{1}{2}\omega^2mA^2$$ $$E_{qho}=(2n+1)\hbar\omega/2$$ So in order to become equal: $\omega m =\hbar$. Which means that the mass and the frequency must be very small and this would lead us to have to use quantum mechanics instead of classical mechanics. Is this correct?

  • 2
    $\begingroup$ Hello. You should define the quantities $P_{\text{clas}}$ and $P_{\text{qho}}$. Also, what is $n$? Is it the number of photons in a fock state? The classical limit of the quantum harmonic oscillator is usually understood for a quantum state being the coherent state (in the large average photon number). $\endgroup$
    – StarBucK
    Commented Dec 15, 2023 at 12:30
  • $\begingroup$ Linked. $\endgroup$ Commented Dec 15, 2023 at 14:13

2 Answers 2


Yes, it is correct. Instead of thinking of a spectrum, it's easier to bridge the two in terms of density of states. In quantum mechanics, the density of states is just: $$ D(E) = \sum_n \delta(E-E_n) $$ with the sum over an energy eigenbasis.

In the case of the harmonic oscillator, you have for $n\in\mathbb N$: $$ E_n = \hbar\omega(n+1/2) $$ So as $\hbar\to 0$, you as a weak limit: $$ D(E)\to \frac{1}{\hbar\omega} $$

This is consistent with classical mechanics. The density of states in this case is (using the semiclassical limit, or using the heuristic that a patch of phase space has $\frac{dxdp}{h}$ number of states): $$ D(E) = \frac{1}{h}\int\delta(H-E)dxdp $$ with $H$ the Hamiltonian. In the case of the harmonic oscillator, the isolines are ellipses and you do get: $$ D(E) = \frac{2\pi}{h\omega} $$

These results are typically most useful in statistical physics.

Hope this helps.


$$P_{clas}=\frac{1}{\pi \sqrt{A^2-x^2}}$$ $$P_{qho}\propto\frac{1}{\sqrt{2n-1-x^2}}$$

Your formula for the classical probability (with $A$ being the maximum amplitude of $x$) is correct. But in the formula for the quantum probability density someone has deliberately omitted all the constants. The correct formula including the constants would be (because $\sqrt{\frac{\hbar}{m\omega}}$ has the dimension of a length): $$P_{qho}\propto\frac{1}{\sqrt{2n-1-\frac{m\omega}{\hbar}x^2}}$$ or $$P_{qho}\propto\frac{1}{\sqrt{\frac{\hbar}{m\omega}(2n-1)-x^2}}$$

If $A^2=2n-1$, then ...

Therefore you cannot simply equate $A^2$ and $2n-1$. Instead, you can equate $$A^2=\frac{\hbar}{m\omega}(2n-1)$$

Then, using $$E_{clas}=\frac{1}{2}\omega^2mA^2$$ $$E_{qho}=(2n+1)\hbar\omega/2$$ you find $$E_{qho}=E_{clas}+\hbar\omega$$

So, in the limit $\hbar\to 0$, you have confirmed $E_{qho}\approx E_{clas}$, as it should be due to the correspondence principle.


Your Answer

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

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