Every periodic table on every classroom wall gives carbon an atomic weight of 12.011. It is printed as a fact, in the same typeface as the atomic number, and the atomic number really is a fact — carbon has six protons everywhere in the universe.
The weight is not that kind of number. For carbon and thirteen other elements, the body that sets these values does not publish a single figure at all. It publishes a range.
What IUPAC publishes instead
The Commission on Isotopic Abundances and Atomic Weights gives fourteen elements their standard atomic weight as an interval:
hydrogen [1.007 84, 1.008 11] · lithium [6.938, 6.997] · boron [10.806, 10.821] · carbon [12.0096, 12.0116] · nitrogen [14.006 43, 14.007 28] · oxygen [15.999 03, 15.999 77] · magnesium [24.304, 24.307] · silicon [28.084, 28.086] · sulfur [32.059, 32.076] · chlorine [35.446, 35.457] · argon [39.792, 39.963] · bromine [79.901, 79.907] · thallium [204.382, 204.385] · lead [206.14, 207.94]
The reason is that these elements are not isotopically uniform on Earth:
"Many elements can be found on Earth in a variety of materials with substantially different genesis. As a consequence, the values of the atomic weight for some elements vary significantly depending upon the origin and age of these materials."
Where that variation is larger than the measurement uncertainty and well documented, a single number would be asserting a precision the Earth does not supply. So the Commission stopped asserting it, starting in 2009.
The list is still moving
Most accounts of this date it all to 2009 and 2011 and put the count at twelve. Both are out of date. Argon became an interval in 2017 and lead in 2020, and lead's is the widest of the fourteen by a long way — [206.14, 207.94], a span of nearly two whole mass units, because lead is the end point of three separate radioactive decay chains and a rock's lead depends on what decayed in it.
⚠️ The values move too. In 2024 the Commission revised zirconium from 91.224 to 91.222, along with gadolinium and lutetium. A printed table is a snapshot with a date, whether or not it prints one.
Lithium, which is the famous case for the wrong reason
Lithium has the widest relative spread of the fourteen, and it is usually explained as geology. The Commission's own account says the opposite:
"Although Li occurs in diverse geological associations and although the relative mass difference of the isotopes is large, the variability of the atomic-weight values of lithium in most terrestrial sources appears to be smaller than 0.002."
Natural lithium barely varies. The interval spans 0.059 — thirty times the natural range — because of what industry does to it:
"Lithium depleted in 6Li may be distributed in commerce, with abundances of 6Li as low as 2 %, and atomic weights in excess of 6.99."
Lithium-6 is a nuclear source material, and the depleted residue from extracting it is sold as ordinary lithium. So the honest answer for a bottle of lithium reagent is not "it depends where on Earth it came from" but "it depends what was done to it before it reached you" — which is why the Commission flags it with an annotation meaning the material may have been subjected to "undisclosed or inadvertent isotopic fractionation".
Where the single number went
There is still a single value, and this part is commonly described wrongly.
IUPAC used to publish a separate table of "conventional atomic weights" for education, trade and commerce. That table was discontinued in the 2021 report; the single value you want now is the abridged standard atomic weight, quoted to five significant figures with a conservative uncertainty — carbon 12.011 ± 0.002, lithium 6.94 ± 0.06, lead 207.2 ± 1.1.
⚠️ It is not the midpoint. Argon's abridged value is 39.95 against an interval midpoint of about 39.88; lead's is 207.2 against 207.04. Many correspond instead to a reference material. And the interval is not a more accurate measurement waiting to be narrowed — the Commission is explicit that "no particular value in the interval can be regarded as more representative than any other", and that the notation "does not alter the meaning of the standard atomic weight".
What this does not mean
It does not mean 12.011 is wrong, or that anyone using it is making an error. For nearly every purpose — a stoichiometry problem, a molar mass, a school exam — the abridged value is the value, and the interval would change no answer you could measure.
Nor does it mean the wall chart is hiding something. IUPAC's own periodic table of the elements and isotopes prints both: the single value above, the interval beneath it. The chart tells the truth; the copies of it usually drop the second line.
What to do with it
If you are quoting an atomic weight in anything durable, quote the uncertainty with it and say which table it came from — standard, abridged, or a textbook of unknown vintage. The three differ in the digits that a precise calculation will use.
If the work is isotope-sensitive — dating, forensics, tracing a material's origin, anything involving commercial lithium or boron — the interval is the warning that a tabulated value cannot substitute for measuring your own sample. That is what the Commission says to do when the number matters: determine it on the material concerned.
For everything else, take the single number, and know that it is a considered convenience rather than a constant of nature.
Where this comes from
The intervals, the abridged values and the 2024 revisions are from the Commission on Isotopic Abundances and Atomic Weights, from its standard and abridged atomic weight tables headed "Standard Atomic Weights 2024", read on 16 September 2026. The explanatory passages and the discontinuation of the conventional table are from "Standard atomic weights of the elements 2021", the IUPAC Technical Report published in Pure and Applied Chemistry volume 94; the lithium passages are from the Commission's own element page for lithium. The dual-value chart is IUPAC's Periodic Table of the Elements and Isotopes, dated June 2019 — which shows thirteen intervals, because it predates lead.
⚠️ The count of fourteen was taken by parsing the published table and counting bracketed rows, not from a summary: 14 intervals, 70 single values and 34 elements with no standard atomic weight, which sums to 118 and so accounts for every element. That check was worth doing. An automated reading of the same page returned fifteen, having folded neon — a single value carrying an uncertainty — into the interval list. The word "conventional" appears zero times across the Commission's three current tables, against "atomic" 34, 25 and 12 times in the same files.
View author profile → · Editorial policy
