What percent of the Earth's core is uranium? And how much of the heat at the core is from radioactive decay rather than other forces?


3 Answers 3


Good question! Geochemists and geophysicists agree to disagree, sometimes quite strongly. There are also disagreements within each group as well as between the two groups.

It's not just uranium. There are four isotopes whose half-lives are long enough that they can be primordial and whose half-lives are not so long that they don't produce much heat. These four isotopes are

  • Uranium 235, with a half-life of 0.703 billion years,
  • Potassium 40, with a half-life of 1.277 billion years,
  • Uranium 238, with a half-life of 4.468 billion years, and
  • Thorium 232, with a half-life of 14.056 billion years.

The consensus view amongst geochemists is that there is very little, if any, of any of these isotopes in the Earth's core. Potassium, thorium, and uranium are chemically active. They readily oxidize. In fact, they readily combine chemically with lots other elements -- but not iron. They are strongly lithophilic elements. Moreover, all three are "incompatible" elements. In a partial melt, they have a strong affinity to stay in the molten state. This means that relative to solar system abundances, all three of these elements should be strongly enhanced in the Earth's crust, slightly depleted in the Earth's mantle, and strongly depleted in the Earth's core.

Geophysicists look at the amount of heat needed to drive the Earth's magnetic field, and at the recent results from neutrino observations. From their perspective, the amount of residual heat from the Earth's formation is not near enough to drive the geomagneto. The growth of the Earth's inner core creates some heat, but not near enough to sustain the geodynamo. Geophysicists want a good amount of heat flux across the core mantle boundary to sustain the geodynamo, and to them the only viable source is radioactivity. Recent geoneutrino experiments appear to rule out uranium or thorium in the Earth's core, but not potassium 40. The neutrinos generated from the decay of potassium 40 are not detectable using current technology.

  • 1
    $\begingroup$ Your response is confusing in that you write about the lack of uranium and the reasons it's lacking and then bring up that a lot more heat is needed to support the geomagneto effect. It sounds like the geophysicists are on the right track; maybe the uranium is deep in the mantle if not the core. $\endgroup$ May 4, 2015 at 3:33
  • 6
    $\begingroup$ Why is it confusing? We don't really know, unless someone comes up with new evidence or a better interpretation of the existing one. The two competing theories presented are based on different interpretations of different data. $\endgroup$ May 4, 2015 at 9:08
  • 1
    $\begingroup$ I think the last paragraph is what he means. On the one hand, there is comparatively very little of the listed radioactive elements in the earths core, more in the mantle, the highest percentage near the surface, on the other hand, there's enough of of these radioactive elements to heat up the earth's core. It's a curious combination of points, not necessarily contradicting, but I see what Steve is asking. I also liked David's answer - wanted to point that out and not sound critical. $\endgroup$
    – userLTK
    May 4, 2015 at 9:17
  • 3
    $\begingroup$ Sometimes, scientists agree to disagree. That's a good thing: If we knew everything, science would be dead. A couple of related areas where scientists agree to disagree are the heat flow across the core-mantle boundary (estimates range from less than 4 TW to over 17 TW) and the age of the Earth's solid inner core (estimates range from not possibly less than 1.2 billion years old and most likely over 2.2 billion years old, to not possibly more than 1 billion years old and most likely less than 700 million years old). Competing theories, or better, competing hypotheses. $\endgroup$ May 4, 2015 at 9:27
  • 1
    $\begingroup$ @Owen - Those are the only four primordial nuclides that count with regard to being a significant heat source for the Earth. Shorter lived unstable nuclides are not primordial, while longer-lived ones have such long half lives that they do not represent a significant heat source. Platinum 190 and Bismuth 209 have half lives of $6.5\times10^{11}$ and $19\times10^{18}$ years, respectively, far too long to make them anything close to significant heat sources. $\endgroup$ Sep 24, 2017 at 14:35

One possibility is that enough uranium is present to provide a substantial heat source -- in the deep mantle rather than the core.

Gautron et al.[1] study the inclusion of uranium in an aluminum-doped calcium silicate perovskite, which is believed to exist in the lower mantle (the authors cite Ref. 2). With the aluminum doping, the silicate perovskite becomes compatible with uranium(IV), and thereby "all the uranium present in this region could be easily stored via its insertion in the Al-CaSiO3 perovskite." The authors suggest that thorium, which also favors the +4 oxidation state and forms a similarly large cation, may similarly be incorporated, although they directly studied only uranium.


  1. Laurent Gautron, Steeve Greaux, Denis Andrault, Nathalie Bolfan-Casanova, Nicolas Guignot, M. Ali Bouhifd (2006). "Uranium in the Earth's lower mantle". Geophysical Research Letters 33(23), L23301. https://doi.org/10.1029/2006GL027508.

  2. Hirose, K. (2002). "Phase transitions in the pyrolitic mantle around 670–km depth: Implications of upwellings of plumes from the lower mantle". J. Geophys. Res., 107(B4), 2078. https://doi.org/10.1029/2001JB000597.


He's correct, there is precious little uranium or thorium in the Earth's core. He's also right that an extra source of heat is required to drive the core's magnetism. Note however that, as has been known for a long time, the Earth's core is less dense than would be the case if it was pure Ni-Fe alloy. The answer is that there is a lot of sulphur down there, in fact about 10% of the moon's weight in sulphur, which most likely exists as high pressure Fe-sulphide phases. Potassium is a lithophile element and would not normally exist in the core, but potassium is soluble in Fe-sulphide. The radionuclide 40-potassium, in the sulphide, in the core, is the source of the missing heat which drives the dynamo which has created the Earth's magnetism for more vthan 3.5 billion years.

  • 4
    $\begingroup$ You state this as if it's fact. Citation needed. As far as I know, scientists do not know whether there's a significant amount of potassium-40 in the Earth's core. The inverse beta reaction used in current neutrino detectors cannot detect the neutrinos emitted by potassium-40 decay. The peak energy in that decay is less than the minimum energy needed to trigger the inverse beta reaction. $\endgroup$ Sep 18, 2015 at 15:06
  • 1
    $\begingroup$ There is a scientific consensus that the outer core must contain some light element (or light elements) in addition to iron, nickel, and trace amounts of heavier elements. From what I've read, geologists and geophysicists agree to disagree on which light element is most responsible for the reduced density of the outer core compared to that of an iron/nickel. Some say it's oxygen, others silicon, yet others, sulfur. Or perhaps carbon. As far as I can tell, there is no scientific consensus on this, either. $\endgroup$ Sep 18, 2015 at 15:10
  • 1
    $\begingroup$ True, it's a deduction rather than a fact, but the longevity of radioactive heat in the core has to be explained somehow, and the potassium solution is much more plausible than any of the three alternatives (thorium and two isotopes of uranium). $\endgroup$ Sep 25, 2015 at 23:32
  • 1
    $\begingroup$ @DavidHammen as far as we know, potassium, uranium and thorium are some of the most lithophile elements there are. If you want to suggest they reside in the core, it's up to you to propose a mechanism of why they would be there in the first place. $\endgroup$
    – Gimelist
    Sep 29, 2015 at 0:30
  • 1
    $\begingroup$ @Michael - I am not saying that K, U, and Th are present in substantial amounts in the core. I said exactly the opposite! However, explaining the amount of heat needed to explain the geodynamo without radioactivity has been problematic. I don't know whether the recently published (March 2016) article entitled The deep Earth may not be cooling down by D. Andrault et al. is heretical, but it certainly does address several of the core issues. $\endgroup$ Jun 21, 2016 at 9:53

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