In Freeman Dyson's classic 1979 Time Without End paper he points out that, if proton decay does not occur, then normal matter will spontaneously fuse to iron on a timescale of $10^{1500}$ years, and heavier elements will undergo fission or alpha emission, again producing iron (although he does not calculate the timescale). My question is how correct this picture is.

It is a well-known quibble that strictly speaking, $^{56}$Fe is not the most tightly bound nucleus: it has 8790.4 keV binding energy per particle but $^{58}$Fe has 8792.3 and $^{62}$Ni has 8794.5. More subtly, this depends on whether we consider adding the right number of nucleons or the right number of protons and neutrons. So it might appear that the long-term composition of matter would hence be nickel-dominated.

However, just like the stellar formation processes favour $^{56}$Fe over $^{62}$Ni despite the energy difference because of the relative scarcity of neutrons to bridge the gap between them, the actual decay processes in the far future also seem to lead to trapping in other isotopes. $\alpha$-decays require a positive Q-value for $^{A}_{Z}X\rightarrow ^{A-4}_{Z-2}Y+\alpha+Q$: this becomes common above $A>102$, but below the energy difference is too small to allow $\alpha$-decay. $\beta$-decay to the rescue! They (and electron capture) happen across the entire $A$ range, but only change $Z$ by $\pm1$. So this appears to lead to a lock-in where heavy element decay get stuck in the region $62>A>102$. In this argument I have assumed that any energetically allowed transition will eventually happen; at least some $\beta$-decays are presumably blocked by spin considerations.

Plot of potential decays in (A,Z) plane: blue beta decay/electron capture, red alpha decays. Green isotopes do not have any energetically allowed decays. White circles mark iron-56 and nickel-62.

Plot of hypothetical decays in (A,Z) plane: blue beta decay/electron capture, red alpha decays. Green isotopes do not have any energetically allowed decays. White circles mark iron-56 and nickel-62.

I seriously doubt Dyson would have made a mistake (even though this is an extremely minor part of the paper), so what is going on? Spontaneous fission reducing everything to the iron peak? Or is the long-term chemical composition of the universe a mixture of stable heavyish elements?

[ Some other minor issues: in $\Lambda$CDM cosmology rapid expansion also makes many atoms and molecules isolated so their fusion processes end long before iron. There are also issues with spontaneous ionization and possibly nuclear decay due to the Herzfeld "paradox" of divergent partition functions for completely isolated atoms. And, as Dyson pointed out, iron is metastable to tunnelling into neutron star or black hole states. ]

  • $\begingroup$ My question is how correct this picture is. Doesn't this require us to wait about $10^{1500}$ years as otherwise we're just opinion based ? Note if the standard model requires changes for e.g. dark matter then we can't know what impact that has on the ultimate fate of matter. Likewise if dark energy proves to require changes to our understanding of fundamental processes. particularly on the cosmic scale you're describing, then this complete unknown renders any answer purely speculative. If there's some more islands of stability for nuclei - similar issue possible. $\endgroup$ Mar 29, 2020 at 23:28
  • 1
    $\begingroup$ This is a question about what our current understanding of physics predicts. That is not more speculative than asking about the dynamics of the Milky Way-Andromeda merger in a few gigayears or what how the scale factor in $\Lambda$CDM changes. $\endgroup$ Mar 29, 2020 at 23:38
  • 2
    $\begingroup$ More recent "duplicate": physics.stackexchange.com/q/803817/226902 $\endgroup$
    – Quillo
    Feb 26 at 13:20

2 Answers 2


I went through your link, and this is an answer to your title,

What is the extreme long-term elemental composition of the universe?

The summary answer is the the table given in the link:

dyson universe

To clarify, all matter turns to iron is talking about matter in stars,

After the time ($=10^{1500}$ yrs) elapsed, most of the matter in the universe is in the form of ordinary low-mass stars that have settled down into white dwarf configurations and become cold spheres of pure iron.

not all matter in the universe, which still will have in the huge empty spaces hydrogen mainly but also a lot of atomic matter debris which at the temperature of zero will not meet another debris to fuse into anything. (Of course radiation will be there also). He proposes iron t so he can do some calculations.

Whether stars become iron or another element , or many elements is immaterial to his discussion which leads to the alternatives of neutron stars and black holes, which is the point of his perambulation. All this assuming that the protons do not decay, and he gives references to later theoretical papers, where proton decay is calculated.

  • $\begingroup$ Yes, but my question is for a more detailled answer about the elemental composition. The iron claim, which I think is partially fallacious, is widely reported but we should be able to do better 41 years after the initial paper. $\endgroup$ Mar 30, 2020 at 6:31
  • $\begingroup$ It is not necessary for Dysons argument. it would just make it more difficult to do the calculations. $\endgroup$
    – anna v
    Mar 30, 2020 at 8:43
  • $\begingroup$ Yes, but I am interested in the stability of particular kinds of matter. That is why I want to be able to do the calculations. $\endgroup$ Mar 30, 2020 at 12:25
  • $\begingroup$ It seems to me not worth it. For example, a lot of first order effects can be computed without knowing the exact potentials, by using the harmonic oscillator potential. I think the assumption "it is all iron" is a first order approximation that allows estmates , and assuming complicated possibilities will not change much. the outcomes . $\endgroup$
    – anna v
    Mar 30, 2020 at 13:38
  • $\begingroup$ @Anders Sandberg Given the long association of iron with artifice in human society, I was surprised by Dyson's conclusion that it might be the prevalent matter in the late times of an open universe. Might the fallacy that you're imagining bear on some possible re-evolution of a universe with a more naturalistic resemblance to our own (especially given Anna's description of an iron universe as only a 1st approximation)? $\endgroup$
    – Edouard
    Mar 31, 2020 at 1:40

Well, according to John Baez, you'd eventually wind up with a whole lot of really cold, really empty, space. He seems think that by 10^23 years the universe will consist mainly of, "black dwarfs, neutron stars, and black holes together with atoms and molecules of gas, dust particles, and of course planets and other crud, all very close to absolute zero." He also agrees with the spontaneous ionization you cited in your original post. Black holes should evaporate by 10^99 years. The smallest black dwarves should be converted entirely into iron by 10^3200 years.

Then, in 10^10^26 years, "if smaller black dwarfs and planets and the like don't evaporate and their protons don't decay, they may quantum-tunnel into becoming solid iron. And then, if this iron doesn't evaporate and nothing else happens, these balls of iron will eventually quantum-tunnel into becoming black holes, which then Hawking-radiate away."

So, I guess, eventually and for a very long time, chemical composition of the universe will be pretty much just iron, then black holes, then photons, and then... nothing. Which, I guess, could bring us back to Penrose's CCC and the birth of a new universe. And if we take things just a little farther, then I guess we're probably back to hydrogen, with a sprinkling of helium and a dash of lithium ;-)

  • $\begingroup$ Baez doesn't do the isotope analysis I discuss in my question, and just assumes Dyson was right. $\endgroup$ Oct 5, 2021 at 23:52
  • $\begingroup$ That's true, but he also makes another assertion that changes the game (in my opinion) a little. The assertion that "in a universe whose expansion keeps accelerating, each pair of freely falling observers will eventually no longer be able to see each other, because they get redshifted out of sight." If we accept this, then we need to redefine "universe", right? If we're talking about the observable universe, then, when we get that far into the future, the answer depends on which universe we're looking at. The one full of ionized hydrogen gas or the one with a big ball of iron... $\endgroup$
    – Thor
    Oct 6, 2021 at 0:06
  • $\begingroup$ Universe definitions aside, I think I see what you're asking and I'm afraid that I'm not capable of calculating what the isotopic make up of the "multiverse" would be that far out... I mean, I guess you'd have to calculate how many generations of stars could exist before everything finally falls apart and use those numbers to calculate what elements you'd have and in what quantities. $\endgroup$
    – Thor
    Oct 6, 2021 at 0:14
  • $\begingroup$ Nope, I'm an idiot and my mind goes fuzzy whenever I see math symbols... You're not really not asking about the elemental composition of the universe at the end of time, you're asking about the iron limit and if the "iron peak" should be called the "nickel peak" instead and why there aren't a bunch of heavier elements mixed in there. Ummm... no clue. $\endgroup$
    – Thor
    Oct 6, 2021 at 1:01
  • $\begingroup$ At a guess, the answer would have to do with the value of the fine structure constant, whether it's actually constant, and what its exact value (or final value) at "[zero energy][1]" will be, whatever the heck that means... Or perhaps it's just the spontaneous fission you mention in your original post. I mean, if matter can spontaneously quantum tunnel to become a black hole, then getting from heavier element to the iron/nickel peak should be a piece of cake (no math to back this up)! [1]: en.wikipedia.org/wiki/… $\endgroup$
    – Thor
    Oct 6, 2021 at 1:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.