I am given two concentric conductive spherical shells, one of radius $a$, and the other of radius $b$, with $b>a$. The space between these shells is filled with a dielectric of relative permittivity $\kappa$. The inner shell has a charge of $+q$ and the outer shell is grounded.

I would like to solve for the potential within the dielectric medium as a function of $r$, and use it to compute the surface charge density on the outer plate.

Considering Laplace Equation in Spherical Coordinates, the solution takes the form: $$V(r,\theta)=\sum_{l=0}^{\infty}{\bigg(A_l+\frac{B_l}{r^{l+1}}}\bigg)P_l(\cos\theta)$$

By spherical symmetry, we know that there exists no dependence on $\theta$, thus we need only to consider the first term in the summation: $$V(r)=A+\frac{B}{r}$$ We will need boundary conditions to determine $A$ and $B$. Consider the surface charge density on the inner plate: $$\sigma=\frac{+q}{4\pi a^2}$$ This surface charge is associated with a discontinuity of the Electric Field at $r=a$. Using Gauss Law considered in the vicinity of a surface charge: $$-\sigma=\epsilon_{above}\frac{\partial V_{above}}{\partial r}-\epsilon_{below}\frac{\partial V_{below}}{\partial r}\Biggr|_{r=a}$$ Since the shell is conductive, it is an equipotential i.e. $V_{below} = constant$. Thus the above formula reduces to: $$-\sigma=\kappa \epsilon_0 \frac{\partial V_{above}}{\partial r}\Biggr|_{r=a}$$ Substituting $\sigma$ and $\partial_r V_{above}$ we obtain: $$\frac{-q}{4\pi a^2}=\kappa \epsilon_0 \frac{-B}{a^2}$$ Simplifying: $$B=\frac{q}{4\pi \kappa \epsilon_0}$$ Applying the second boundary condition of the grounded potential $V(b)=0$: $$0=A+\frac{B}{b}$$ Rearranging and subsituting: $$A=\frac{-B}{b}=\frac{-q}{4\pi \kappa \epsilon_0 b}$$ Thus the potential in the bulk of the dielectric is given by: $$V(r)=\frac{q}{4\pi \kappa \epsilon_0}\biggr(\frac{1}{r}-\frac{1}{b}\biggr),\quad a < r < b$$ To compute the surface charge density at the outer shell, we need to consider the potential outside the outer shell $r>b$. We will repeat the process outside the shell.

Following a similar argument: $$V(r)=C+\frac{D}{r}$$ We have the boundary condition $V(b)=0$ thus: $$C+\frac{D}{b}=0$$ But we are missing a boundary condition...and now I'm stuck. We should probably get $C=D=0$, but I cannot justify it. Could someone justify it for me with another boundary condition, and some physical justification for it.

EDIT: I know that this is an overkill technique for a simple problem, the answer is $-q/{4\pi b^2}$ but I am trying to solve it as a boundary value problem for practice.

  • $\begingroup$ ...why the close vote? This isn't a homework question, I already know the answer. $\endgroup$
    – Joeseph123
    Sep 8, 2020 at 12:30
  • $\begingroup$ You can use Gauss law, picking a sphere that encloses both conductors as integration surface. $\endgroup$
    – secavara
    Sep 8, 2020 at 12:55
  • $\begingroup$ It doesn't have to be literally homework, hence the HW&E tag. $\endgroup$
    – Gert
    Sep 8, 2020 at 13:11

1 Answer 1


You're missing the "trivial" boundary condition $V \rightarrow0$ as $r\rightarrow\infty$ for a charge distribution spread over a finite volume/area/length.

$\therefore \hspace{2pt} V(\infty)=C=0 $ by using your first condition $0+\frac{D}{b}=0$ gives $D=0$.

Surface charge density of grounded shell $$-\sigma=\epsilon_{above}\frac{\partial V_{above}}{\partial r}\Biggr|_{r=b^+}-\epsilon_{below}\frac{\partial V_{below}}{\partial r}\Biggr|_{r=b^-}$$ $$\implies-\sigma=0-\kappa\epsilon_0\frac{q}{4\pi \kappa \epsilon_0}\bigg(-\frac{1}{b^2}\bigg)$$ $$\implies\sigma=-\frac{q}{4\pi b^2}$$

  • $\begingroup$ My problem is, how can I ensure that $V(\infty)=0$, you see we set the reference potential arbitrarily, how can I be sure that our choice of the shell $r=b$ at zero potential is consistent with $V$ vanishing at $r\rightarrow \infty$, and not some other constant? $\endgroup$
    – Joeseph123
    Sep 8, 2020 at 18:16
  • $\begingroup$ You can use the arguments in here to feel confident about this solution. $\endgroup$
    – secavara
    Sep 8, 2020 at 20:36

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.