Consider the following scenario: there is a magic photon source that emits photons in both directions along a single axis. In one direction is an infinite amount of lead, in the other direction is an infinite amount of osmium. The emitted photons pass directly into these metals and do not interact with anything else, nor are any other forms of radiation present. The energies of these photons range from high ultraviolet to the most energetic gamma ray known to be possible, with every amount of energy being equally represented. The photons are emitted at sufficient intervals that any increase in temperature caused by a photon will dissipate before the next photon is emitted (though the base temperature is unspecified).

In this situation, will the photons, and any secondary energized particles, travel farther through the lead or the osmium?

The description of high-energy photon flux attenuation that I am familiar with says that a high-energy photon passing through conventional matter will lose energy by interaction with atomic nuclei that it passes, with more massive nuclei reducing the photon's energy by a greater amount. Thus lead, being the stable element with the highest atomic number, seems like an obvious choice for shielding against this type of radiation. The question arises because osmium has a close atomic number to lead but is much denser, meaning that a photon will interact with more atomic nuclei in the same distance. In a purely hypothetical sense, this might make osmium a superior shield.

Additionally, if any element or compound will stop these photons more effectively than either lead or osmium, I would like to know about that.


I tried answering this by going to the XCOM database where you can get a calculation of the stopping power of elements and compounds.

First - pick a few likely candidates. I found a table of elements with density which is a good place to start. The highest density elements are also among the highest Z ones:

density  name        Z   m    density
15.4     Thorium    90  232    5.97
16.65    Tantalum   73  181    6.72
18.95    Neptunium  93  237    7.44
19.32    Gold       79  197    7.75
19.35    Tungsten   74  183    7.82
19.84    Americium  95  243    7.76
20.2     Uranium    92  235    7.91
21.04    Rhenium    75  185    8.53
21.45    Platinum   78  195    8.58
22.4     Iridium    77  191    9.03
22.6     Osmium     76  187    9.19

I got the first three columns from the above link, then added the Z of the most stable (or one of the stable) isotopes, and used these to compute proton density (from which electron density follows). Units of density and proton density are g/cm3.

I decided to plot the XCOM output for four elements: Pb, Os, Ir and U. You could repeat for others - but this gives a general direction. The result (log-log scale):

enter image description here

As you can see, lead is much worse than the other three - and the other three are almost indistinguishable at high energies, but have a little bit of difference at lower energies (where the binding energy of the electrons is comparable with the energy of the incoming radiation, and photoelectric effects dominate).

Whether Uranium or Osmium/Iridium "win" depends on the energy distribution of your "all energies" input beam - but this approach should give you an idea on how to tackle the question. Clearly, lead is worse than either of these three for most energies (it's better than Uranium just below Uranium's K edge...).


The likelihood of a photon scattering off a particle in its path is proportional mostly to the amount of mass per unit area in the path of the photon beam. This means that, though higher atomic mass is important, it is really density that makes a good gamma shield.

Wikipedia cites lead as being only 20-30% better at shielding than a similar shield made of concrete or soil - assuming the shields have the same mass - despite lead having a much larger nucleus than the typical components of these materials. The real benefit of lead is that it is dense, so we can make cram all of that mass into a small volume; the lead shield will be much smaller than the soil one.

This means that though lead is likely a(n ever so) slightly better shield per unit mass than osmium, the density of osmium implies better shielding in a given volume than lead.

  • $\begingroup$ "The likelihood of a photon scattering off a particle in its path is proportional mostly to the amount of mass per unit area" It is a little more precise to use the electron number density rather than the mass density. The result is that for two materials of the same mass density the lower atomic number tends to come out ahead because $A$ rises faster than $Z$. (This statement assumes that the photons are high enough in energy that the ionization potential of the materials in question can be ignored.) $\endgroup$ Jul 23 '15 at 20:58
  • $\begingroup$ It depends on the energy of the photons. Just above the K edge, the high Z material works very well because of the high photo fraction (probability of photoelectric absorption). But when the energy range is specified "all the way out to infinity", the binding energy of the electrons becomes negligible and electron density is all that matters. As @dmckee says, that requires you to look at more than just the mass density. Either way Osmium wins. $\endgroup$
    – Floris
    Jul 24 '15 at 2:29
  • $\begingroup$ Helpful paper: researchgate.net/profile/Hans_Bichsel/publication/… $\endgroup$
    – Floris
    Jul 24 '15 at 2:35

Well If you are measuring a shield for example lead 8ft by 8ft and 2 ft thick exactly and same measurements for osmium, lead would be slightly better but if you measuse by atoms eg 10 trillion trillion trillion billion million atoms osmium would be several times better than lead the reason we use lead is it provide more protection per inch but for nuclear bunkers they would be better off with osmium because at that point they don't care about size. But the answer::osmium

  • 2
    $\begingroup$ The reason we use lead instead of osmium is simply costs. I just ordered another 15 tons of lead shielding for less than 100 000 Euros. By way of comparison, you would get only about 60 g of osmium for the same price. $\endgroup$
    – user316055
    Oct 9 at 12:28

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.