What accounts for a Lyman-break for all wavelengths shorter than 91.2nm if the Lyman limit is the highest energy photon that neutral hydrogen absorbs?

Question

From this description of Lyman-break galaxies, I don't understand how:

...radiation at higher energies than the Lyman limit at 912 Å is almost completely absorbed by neutral gas around star-forming regions of galaxies. In the rest frame of the emitting galaxy, the emitted spectrum is bright at wavelengths longer than 912 Å, but very dim or imperceptible at shorter wavelengths—this is known as a "dropout", or "break".

But the wikipedia page for the Lyman series states that the highest limit of radiation absorbed or emitted by neutral hydrogen is 91.2 nm:

the Lyman series is a ... series of transitions and resulting ... emission lines of the hydrogen atom as an electron goes from n ≥ 2 to n = 1 ... The greater the difference in the principal quantum numbers, the higher the energy of the electromagnetic emission.

It then states that there is an asymptotic limit to this energy as the difference between transition levels approaches infinity:

There are infinitely many spectral lines, but they become very dense as they approach n = ∞ (the Lyman limit)... "91.1753 nm"

I don't understand how neutral hydrogen can absorb light from a photon emitted with wavelength shorter than 912 Angstroms if this wavelength is precisely the highest energy photon hydrogen can absorb.

So my question is: How can there exist a wide drop-out in the spectrum of galaxies at wavelengths shorter than about 91.2 nm, if the highest-energy electromagnetic radiation a hydrogen atom can emit or absorb (at n = ∞) is 91.1753 nm?

How is the hydrogen interacting with photons of higher-energies than this?

It must be a conceptual issue I'm not understanding. Should I not view absorption energies as conceptually the same as emission energies?

Answer 1

The Lyman-limit cross section

The Lyman limit is not a narrow line, like it is for electronic transitions. There is a minimum energy needed to ionize a hydrogen atom — 13.6 eV, corresponding to a wavelength of 912 Å — but higher energies are not a big problem, because — as Jon Custer comments — the excess energy instead goes into kinetic energy of the particles.

If the energy of the photon becomes too large, however, it will eventually be unable to ionize the atom. In fact, the probability of ionization — the cross section — decreases with the cube of the wavelength, i.e. $\phi(\lambda) \propto \lambda^3$. In the figure below I've sketched the absorption cross section where you can see that significant absorption actually extends all the way to (a few) 100 Å.

The intrinsic galaxy spectrum

But normal galaxies don't emit much light at such high energies. So the reason that the spectra of Lyman-break galaxies "stay" black blueward of the 912 Å Lyman limit is that, at wavelengths shorter than what may be effectively absorbed, there simply isn't much light emitted in the first place.

However, if you consider instead a galaxy with an active galactic nucleus, the emission spectrum extends all the way to soft X-rays (and beyond). For these galaxies you do indeed see the transmitted intensity begin to rise again at short wavelengths.

High-redshift galaxies

In my original answer I considered only galaxies at high redhifts (because those are the ones I'm used to thinking of). Here, another effect comes into play which tends to erase even very short wavelengths:

As light travels away from the galaxy, it is redshifted by the expansion of the Universe. Hence, after a little while light that had a too short wavelength will be redshifted closer to the Lyman limit, and if there happens to be neutral hydrogen around at this point (which invariably there is in the early Universe), then it will be absorbed after all.

The Lyman alpha forest

This effect it more pronounced for Lyman alpha photons, i.e. the bound-bound transition from the ground state to the first excited state. For this line, only photons with energies very close to the "correct" energy (10.2 eV, or 1216 Å) are able to excite the hydrogen. But expansion eventually redshifts the more energetic part of the spectrum to 1216 Å, so every time there's a hydrogen cloud, this cloud causes a narrow dip in the spectrum.

For intermediate redshifts ($z\sim 1\text{–}6$), you will see a bunch of these lines, comprising the so-called "Lyman alpha forest". At high redshifts ($z\gtrsim6$), the Universe is so neutral that all these lines overlap, and no light is transmitted, resulting in the "Gunn-Peterson trough". At low redshifts ($z\lesssim1$), most of the spectrum is transmitted.

This evolution is seen in the figure below:

^{Credit: Ross McLure.}