What is the correct separable Schrödinger equation in spherical coordinates?

Some articles use this formula: $$ \Psi(r,\theta,\phi) = R(r)\cdot\Theta(\theta)\cdot \Phi(\phi), $$ and some of them use: $$ \Psi(r,\theta,\phi) =\frac{R(r)\cdot \Theta(\theta)\cdot \Phi(\phi) }{r}. $$

So which is true?


2 Answers 2


Both formulas are correct. And it is essentially a matter of taste which one you prefer:

  • If you use the separation approach $$\Psi(r,\theta,\phi) = R(r) \cdot \Theta(\theta) \cdot \Phi(\phi)$$ then you get the radial differential equation $${\frac {1}{r^{2}}}{\frac {d}{dr}}\left(r^{2}{\frac {dR}{dr}}\right)-{\frac {\ell (\ell +1)}{r^{2}}R+{\frac {2m}{\hbar ^{2}}}\left[E-V(r)\right]R=0}$$ (like in Wikipedia - Particle in a spherically symmetric potential - Derivation of the radial equation) with the normalization condition $$\int_0^\infty dr\ r^2 |R(r)|^2=1.$$
  • If you use the separation approach $$\Psi(r,\theta,\phi) = \frac{u(r)}{r} \cdot \Theta(\theta) \cdot \Phi(\phi)$$ then you get the radial differential equation $$\frac {d^2u}{dr^2}-{\frac {\ell (\ell +1)}{r^{2}}u+{\frac {2m}{\hbar ^{2}}}\left[E-V(r)\right]u=0}$$ with the normalization condition $$\int_0^\infty dr\ |u(r)|^2=1.$$

At the end you get the same solutions $\Psi(r,\theta,\phi)$ in both cases because $R(r)=\frac{u(r)}{r}$. However, the approach with $u(r)$ leads to simpler math and has more similarity with a 1-dimensional Schrödinger equation.


If I am not wrong, then the second one is used for (I guess) when we solve the radial part. We have to put $R(r)=u(r)/r$, so maybe in second one they have written $R(r)/r$ instead of $R(r)$ is because they put directly $R(r)/r$. I am not sure though.

  • $\begingroup$ Firstly I need to find radial, angular and azimuthal parts, then solve them. $\endgroup$ Jun 21, 2022 at 16:09
  • $\begingroup$ Generally the angular and azimuthal parts are solved first and then radial but u can do either. $\endgroup$
    – Souvik
    Jun 21, 2022 at 16:23
  • $\begingroup$ You will get a different equation because you have to derive $R/r$ twice. you can take a look at a laplacian equation. $\endgroup$ Jun 21, 2022 at 16:30
  • $\begingroup$ U can solve radial eqn using the following method, Near r=0 the differential equation for radial part is d^2u/dx^2=l(l+1)/r^2, now let u=r^s,then putting the value of u in the diiferential eqn, we gets(s-1)=l(l+1), therefore s=l+1, and the solution is u(r)=r^(l+1) or R(r)=r^l $\endgroup$
    – Souvik
    Jun 21, 2022 at 17:30

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.