With $Y_{lm}(\vartheta,\varphi)$ being the Spherical Harmonics and $z_l^{(j)}(r)$ being the Spherical Bessel functions ($j=1$), Neumann functions ($j=2$) or Hankel functions ($j=3,4$) defining $$\psi_{lm}^{(j)}(r,\vartheta,\varphi)=z_l^{(j)}(r)Y_{lm}(\vartheta,\varphi),$$ what are representations of the Poincaré transformations applied to the Vector Spherical Harmonics

$$\vec L_{lm}^{(j)} = \vec\nabla \psi_{lm}^{(j)},\\ \vec M_{lm}^{(j)} = \vec\nabla\times\vec r \psi_{lm}^{(j)},\\ \vec N_{lm}^{(j)} = \vec\nabla\times\vec M_{lm}^{(j)}$$

? Does any publication cover all Poincaré-transformations, i.e. not only translations and rotations but also Lorentz boosts? I'd prefer one publication covering all transformations at once due to the different normalizations sometimes used.


Unfortunately, I'm not familiar with the optics literature on this kind of scattering, so I don't know where this result would be available in the literature. It does seem like the papers of Stein and of Cruzan are not online. I can only make a suggestion for how to rederive the transformation formulas independently.

A promising starting point is the expansion of a plane wave in spherical harmonics which can be found as formula (10.43) in section 10.3 of the 3rd edition of Jackson's textbook on Electrodynamics. Multiplying it by $e^{-i\omega t}$ on both sides gives $$ e^{-i\omega t + i\mathbf{k}\cdot\mathbf{r}} = 4\pi e^{-i\omega t} \sum_{l=0}^\infty i^l j_l(kr) \sum_{m=-l}^l Y^*_{lm}(\theta,\phi) Y_{lm}(\theta',\phi') , $$ where $(r,\theta,\phi)$ and $(k,\theta',\phi')$ are the spherical coordinates of the vectors $\mathbf{r}$ and $\mathbf{k}$ respectively. The left hand side essentially gives the functions $\psi^{(1)}_{lm}(r,\theta,\phi)$ that you are interested in, when expanded in the basis of $Y_{lm}(\theta',\phi')$. I'm not sure which expressions would generate $\psi^{(j)}_{lm}(r,\theta,\phi)$ for your other $j$-cases.

The same left hand side also transforms nicely with respect to Poincaré transformations. I think you should be able to use that property to get the transformation formulas that you need. For example, if you consider the translation by $\mathbf{r}=\mathbf{s}+\mathbf{d}$, with $(s,\chi,\eta)$ the spherical coordinates of $\mathbf{s}$, you get $$ e^{i\mathbf{k}\cdot\mathbf{r}} = e^{i\mathbf{k}\cdot \mathbf{d}} e^{i\mathbf{k}\cdot\mathbf{s}} . $$ Expanding each of the exponentials in the basis of $Y_{lm}(\theta',\phi')$ and using the Clebsch-Gordan formula for expanding products of spherical harmonics will give a formula for $\psi^{(1)}_{lm}(r,\theta,\phi)$ in terms of $\psi^{(1)}_{l'm'}(s,\chi,\eta)$.

Lorentz transformations could be handled similarly.

  • $\begingroup$ Thanks for your feedback (which I cannot even upvote yet...), the problem with this method is that plane waves are expanded fully in Bessel functions, their Neumann-part is zero since they are free of the singularity at the origin. $\endgroup$ Mar 5 '12 at 10:17
  • $\begingroup$ As I mentioned in the answer, I'm not sure which generating function would give you spherical Neumann functions instead of spherical Bessel ones. I'm still not sure, but I think these formulas might help: dlmf.nist.gov/10.56. They are not exactly in the same form as the plane wave I used above, but they should be related. $\endgroup$ Mar 5 '12 at 12:32
  • $\begingroup$ Thanks again (I found these also in Abramowitz/Stegun: people.math.sfu.ca/~cbm/aands/page_439.htm) $\endgroup$ Mar 5 '12 at 14:09

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.