For a decade, scientists have been scratching their heads when trying to put a date on primeval events like the crystallization of the magma ocean on the moon or the early formation of Earth’s continental crust.
Their problem? A revised estimate of the half-life of a radioactive isotope called samarium-146 that is used to gauge the age of ancient rocks.
The updated value, published in 2012 in Science, shortened samarium-146’s half-life by a whopping 35 million years, down to 68 million years from the standard estimate of 103. This reset the clock on the solar system’s early history and suggested the oldest rocks on Earth could have formed tens of millions of years earlier than previously thought.
“It means everything was forming more quickly,” Michael Paul (pictured) of the Hebrew University of Jerusalem, a coauthor of the Science paper, told New Scientist at the time.
But as researchers began using the new half-life in their calculations, their results turned out not to match those obtained with other so-called isotopic clocks. The ensuing uncertainty led many to take a belt-and-suspenders approach, using both the old and the new half-life estimates, according to Richard W. Carlson, staff scientist emeritus at the Carnegie Institution for Science’s Earth & Planets Laboratory.
“As dating methods improved, the discrepancy when using the 68-million-year half-life was becoming clear,” Carlson, who was not involved in the Science study, told Retraction Watch.
The confusion may be down to a lab error, it now appears.
According to Paul, he and his colleagues were recently alerted to an “inconsistency” in their paper by “an independent group” that asked not to be identified. (Paul would not provide any specifics unless we shared our story ahead of publication, which in the interest of editorial integrity we refused to do.) He told us that the finding:
relied on an analysis of the information contained in the original publication and the detailed supplementary material adjoined to it. It is utterly surprising that it had eluded up to this “late point” all co-authors, reviewers and readers of the original article who could have reached the same conclusion based on published material.
Paul’s group determined that the problem had been caused by a mass-spectrometry measurement that had gone wrong for unknown reasons. After deciding to pull their report, they notified members of the research community, including Carnegie’s Carlson, about the impending retraction. (Paul’s more detailed comments to us can be found here.)
Their paper, “A shorter 146Sm half-life measured and implications for 146Sm-142Nd chronology in the Solar System,” was retracted on March 30. The article has been cited 127 times, according to Clarivate’s Web of Science.
“I give the authors credit for being upfront about their mistake and retracting the result when they recognized the error in their measurement,” Carlson told us. “These are very difficult measurements on an extremely rare isotope.”
He added: “I’m not sure that 103 million years is the accurate value of the 146Sm half-life, but it clearly is closer to the true value than is the 68-million-years value.”
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at team@retractionwatch.com.
few millions doesn’t matter. who cares? nobody knows
This is why accurate research and measurements are so important, as they can lead to the inaccuracy and throwing off on other research topics.
I did check informally with Richard Carlson about which value of the half life to use. He pointed out to me that the 68 million year value did not agree well with other dating methods. He has been great help to those of us that need to use accurate half-lifes but do not have the expertise to evaluate the primary publications.
FYI I did cite both values in my publication.
Researchgate alerts when publication that one cited is retracted.