A potential function is given by $V(r)=\frac{Ae^{-\lambda r}}{r}$ Find charge density and hence charge.

I first took the gradient of potential to get $\vec{E}(r)=\frac{Ae^{-\lambda r}}{r}[\lambda+\frac{1}{r}] \hat{r}$

Now using Gauss's law $\nabla \cdot \vec{E}=\frac{\rho}{\epsilon_0} $
$\implies \rho=-\epsilon_0 \frac{Ae^{-\lambda r}}{r}[\frac{1}{r^2}+\frac{\lambda}{r}+\lambda^2]$

Now to find the charge $Q=\int \rho 4\pi r^2 dr$ .
Did I do it correctly? What limits should I choose for the integration?


At least in the first $r$-derivation, you missed some $\lambda$'s. Generally, always check your units if you derive solutation like that, as for exmple "$1+\frac{1}{r}$" just must be flawed.

And you'll solve your problem more directly using the Poisson equation $\Delta V(r)=-\frac{\varrho(r)}{\varepsilon_0}$, and that with a $\Delta$ in explicitly spherical coordinates.

To take a short track here, your $V(r)$ looks like a Yukawa potential, i.e. $$G(r)=\frac{\text{e}^{-\lambda r}}{4\pi\ r}$$ solves $(\Delta-\lambda^2)G(r)=-\delta(\vec r)$ and hence $$\Delta V(r)=4\pi A\left(-\delta(\vec r)+\frac{\lambda^2}{4\pi}\frac{\text{e}^{-\lambda r}}{r}\right)\equiv -\frac{\varrho(r)}{\varepsilon_0}.$$

Integrate over all of space, if you want to find the total charge. If you have a density like $\propto r^{-n}$, then clearly there are charges everywhere. Remark: For integration, notice that you're dealing with a three dimensional delta-function here (but from the Coulomb potential, you know that it gives you a single charge anyway). For the exp-integration, up to some small numbers, you can figure out the result just by power counting in $\lambda$'s.

  • $\begingroup$ Sorry I missed out $\lambda$ in my equation I've now edited that $\endgroup$ – pranphy Nov 9 '12 at 15:25
  • $\begingroup$ Sorry I am not much aware of delta function. Keeping $\delta(\vec r)$ in the result keeps the work of taking Laplacian still intact. So why should I prefer using $\delta(\vec r)$ in my answer of charge density? $\endgroup$ – pranphy Nov 9 '12 at 15:30
  • $\begingroup$ @PrakashGautam: Not sure what you mean. Work of taking the Laplacian? I skipped taking derivatives, essentially, by "knowing" the Green function of a differential equation. The Delta-function is easy to integrate. In any case, have you used the gradient in Sperical cooridnates? $\endgroup$ – Nikolaj-K Nov 9 '12 at 18:17

Gauss' law relates the integrated flux through a closed surface to the integrated charge inside it.

The symmetry of the situation lets you do the flux integral trivially, so I assume that you mean the volume one. Well, you integrate over the full angular ranges and from the center to the bounding surface.


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.