Radiocarbon dating

• 1. Introduction

  ›

  Radiocarbon dating is the method that gave humanity its first reliable clock for the deep past. Imagine holding a fragment of bone from an ancient burial, or a scrap of linen wrapped around a two-thousand-year-old scroll, and knowing that buried inside that material is a ticking atomic counter that has been running since the moment the creature died. That counter is carbon-14, and reading it changed archaeology forever.

  The technique rests on a beautiful natural coincidence. Cosmic rays from space slam into nitrogen atoms in the upper atmosphere, creating radioactive carbon atoms that filter down through the air, dissolve into oceans, and get pulled into plants through photosynthesis. Every living thing on Earth, as long as it is alive, carries a known proportion of this radioactive carbon alongside the ordinary kind. The moment it dies, that proportion begins to shrink at a fixed, measurable rate.

  The physicist who recognized this clock was Willard Libby. He developed the method in the late 1940s at the University of Chicago, and by 1960 he had received the Nobel Prize in Chemistry for the achievement. The questions that remain are not about whether the method works, but about how precisely it can be pushed, what complications distort its readings, and what it has revealed about the human past that we could never have known otherwise.

• 2. Libby and the First Clock

  ›

  Martin Kamen and Samuel Ruben of the Radiation Laboratory at Berkeley started the chain of discovery in 1939, when they ran experiments to find out whether common organic elements had isotopes with half-lives long enough to be useful in biomedical research. Using the laboratory's cyclotron accelerator, they synthesized carbon-14 and found that its half-life was far longer than anyone had previously thought.

  A few years later, physicist Serge A. Korff, working at the Franklin Institute in Philadelphia, predicted that cosmic rays interacting with atmospheric nitrogen would generate carbon-14 continuously. That insight flipped the question. The isotope was not just an oddity made in a cyclotron; it was constantly being manufactured in the sky above everyone's heads.

  Willard Libby, then at Berkeley, heard of Korff's research during World War II and grasped what it meant for dating. After moving to the University of Chicago in 1945, Libby published a paper in 1946 proposing that living matter contains carbon-14 alongside ordinary carbon. He and his collaborators then tested the idea using methane collected from sewage works in Baltimore. By enriching their samples isotopically, they confirmed the presence of radiocarbon. Methane made from petroleum showed no radiocarbon at all, because petroleum is far too old for any to survive.

  The results, published in Science in 1947, stated clearly that organic materials could be dated by this method. Libby and James Arnold then tested the principle against samples of known age, including material from the tombs of the Egyptian kings Zoser and Sneferu, independently dated by Egyptologists to 2625 BC plus or minus 75 years. The radiocarbon measurement returned an average of 2800 BC plus or minus 250 years, a result close enough to validate the approach. Those findings appeared in Science in December 1949, and within eleven years of publication, more than twenty radiocarbon dating laboratories had opened worldwide.

• 3. The Physics of Decay

  ›

  Carbon exists in nature as three isotopes. Carbon-12 and carbon-13 are stable and make up the overwhelming bulk of all carbon on Earth. Carbon-14, the radioactive form, is present at a ratio of roughly 1.25 parts per trillion relative to carbon-12.

  Galactic cosmic rays, and to a lesser degree solar cosmic rays, generate neutrons as they pass through the atmosphere. Those neutrons strike nitrogen-14 atoms, converting them into carbon-14 and releasing a proton. The newly created carbon-14 combines with oxygen to form carbon monoxide, then carbon dioxide, which diffuses through the atmosphere and dissolves in the ocean, and is absorbed by plants through photosynthesis. Animals that eat those plants distribute the radiocarbon throughout the biosphere.

  Once an organism dies, it stops taking in new carbon-14, and the atoms already in its tissues begin their steady disappearance. Each carbon-14 nucleus eventually emits a beta particle, an electron, and an electron antineutrino, and the nucleus reverts to stable nitrogen-14. The half-life of carbon-14, the time it takes for half of any given quantity to decay, is currently accepted as 5,700 plus or minus 30 years. That means a quarter remains after about 11,400 years, an eighth after about 17,100 years, and so on. Beyond roughly 50,000 years, the amount left in a sample falls below reliable detection. Special preparation techniques can occasionally extend measurements to around 60,000 or even 75,000 years before the present, but those are the outer limits.

  The mean-life of carbon-14, distinct from its half-life, is 8,267 years. This value appears in the core decay equation. An intermediate product called the radiocarbon age, expressed in radiocarbon years, assumes the atmospheric ratio has never changed. Because that assumption is not perfectly true, calibration is required to convert a radiocarbon age into a real calendar date.

• 4. Complications in the Carbon Reservoir

  ›

  Carbon moves through three interconnected reservoirs: the atmosphere, the biosphere, and the oceans. The atmosphere holds about 1.9 percent of the total carbon in the exchange system, and carbon-14 mixes through it in less than seven years. The surface ocean holds 2.4 percent but is connected to the deep ocean, which carries more than 90 percent of the total carbon and takes about 1,000 years to cycle back to the surface.

  That slow circulation has a direct consequence. Marine creatures live in water that is a mixture of recently surfaced water and older, carbon-14-depleted water from the depths. The average result is that marine organisms appear to be about 400 years older than they actually are, a figure known as the marine reservoir effect. This correction is not uniform; local deviations of several hundred years can occur for areas that are geographically close to each other, depending on upwelling patterns, ocean topography, and climate.

  The southern hemisphere adds another layer. Its greater ocean area means more carbon is exchanged between the sea and the atmosphere, and since that sea-surface carbon is older, the atmosphere in the south has a lower carbon-14 ratio than the north. Radiocarbon results from the southern hemisphere carry an apparent additional age of about 40 years compared to the north.

  Freshwater rivers that pass over limestone can dissolve carbonate ions, which contain carbon from ancient rocks with no detectable carbon-14. Plants and organisms living in such water absorb this dead carbon and can appear thousands of years older than they are. This is the hard water effect. Volcanic eruptions create an analogous problem: plants near the Furnas caldera in the Azores, for example, were found to have apparent ages ranging from 250 to 3,320 years simply because they were photosynthesizing carbon-14-depleted gas released by the volcano.

  Burning fossil fuels introduced a different distortion beginning in the nineteenth century. Coal and oil are so old they contain no measurable carbon-14, and the carbon dioxide released by burning them diluted the atmospheric ratio. Hans Suess first reported this fossil fuel effect in 1955; although the global dilution would have been only about 0.2 percent if the carbon spread evenly through the entire reservoir, the actual reduction in measured radiocarbon activity is closer to 3 percent because deep-ocean mixing is so slow. Above-ground nuclear weapons tests between about 1950 and 1963 pushed in the opposite direction, almost doubling the atmospheric carbon-14. The peak level in the northern hemisphere occurred in 1964 and in the southern hemisphere in 1966, and that spike, called bomb carbon or the bomb pulse, has since been spreading gradually through the rest of the reservoir.

• 5. Calibration and Tree Rings

  ›

  Libby noted as early as 1955 that the assumption of a constant atmospheric carbon-14 ratio might not hold. Discrepancies eventually emerged between radiocarbon dates and the established chronology of the oldest Egyptian dynasties, and neither system could be assumed to be wrong. A third possibility, that the atmospheric ratio had shifted over time, opened a new line of research.

  Trees provided the answer. Each year, a tree lays down one new ring on its outermost layer, and only that outermost ring exchanges carbon with the surrounding air. The inner rings preserve the atmospheric carbon-14 ratio from the year they formed and do not replenish. By overlapping sequences of rings from different ancient trees, researchers constructed an uninterrupted record stretching back thousands of years. That sequence now extends to 13,910 years before present for the northern hemisphere.

  In the 1960s, Hans Suess used the tree-ring sequence to show that radiocarbon dates were consistent with dates established by Egyptologists, validating the method's core accuracy while also revealing its systematic offsets. He published the first calibration curve for radiocarbon dating in 1967. The curve showed two distinct patterns of variation: a long-term fluctuation with a period of about 9,000 years, and shorter-term wiggles with a period of decades. Suess said he drew the wiggles by what he called cosmic schwung, meaning he attributed them to extraterrestrial forces. These short-term fluctuations are now called de Vries effects, after Hessel de Vries.

  The modern calibration standard, called IntCal, has been updated repeatedly. The version current as of 2020, IntCal20, draws on data from tree rings, corals, cave deposits called speleothems, lake sediment layers called varves, plant macrofossils, and single-celled organisms called foraminifera. A separate curve, SHCal20, covers the southern hemisphere, and a third curve, MARINE20, addresses marine samples. When multiple radiocarbon dates are available for a sequence of samples, a technique called wiggle-matching can align that sequence against the calibration curve to produce dates far more precise than any single measurement. A stratified volcanic ash layer in New Zealand, thought to predate human settlement of the islands, was dated to 1314 AD plus or minus 12 years by this method.

• 6. Measuring Carbon-14

  ›

  Libby's first detector was a Geiger counter of his own design. He coated the inner surface of a cylinder with lamp black made from the carbon in his sample and inserted a counting wire so that no material stood between the sample and the detector. That was necessary because the beta particles emitted by decaying carbon-14 are so weak that half of them are stopped by an aluminum layer just 0.01 millimeters thick.

  Gas proportional counters soon replaced Libby's device. These record ionization bursts caused by beta particles, and their shielding of lead or steel reduces background radiation. Anticoincidence detectors further filter out stray signals by ignoring any event recorded simultaneously inside and outside the counter. Liquid scintillation counting, invented in 1950 and made competitive with gas counting in the early 1960s through efficient benzene synthesis methods, works by detecting flashes of light as beta particles interact with a fluorescing agent in benzene. After 1970, liquid counters became the preferred choice for newly built laboratories.

  In the late 1970s, accelerator mass spectrometry, known as AMS, offered a fundamentally different approach. Rather than waiting to detect the decay of individual atoms, AMS counts the actual atoms present. The sample, typically shaped into solid graphite, is made to emit negatively charged carbon ions, which are injected into a particle accelerator. The ions gain speed, pass through a stripper that removes electrons and gives them a positive charge, and then travel through a magnet that curves their paths. Heavier ions curve less than lighter ones, so the different carbon isotopes separate into distinct streams and can be counted individually.

  AMS requires samples as small as 0.5 milligrams of carbon, compared to the roughly 10 grams needed for beta counters. An accuracy of one percent can be achieved in minutes with AMS, far faster than older technology could manage. For beta counters, extending a counting run from 250 to 500 minutes halves the detectable carbon-14 needed to achieve the same error margin of plus or minus 80 years at 68 percent confidence. A standard reference material used across many laboratories is the HOxII oxalic acid standard, prepared by the National Institute of Standards and Technology in 1977 from French beet harvests.

• 7. What Radiocarbon Revealed

  ›

  Anthropologist R. E. Taylor described the effect plainly: radiocarbon data made a world prehistory possible by contributing a time scale that transcends local, regional, and continental boundaries. Before the method existed, archaeologists depended on stratigraphy and the typology of objects like stone tools or pottery, both of which are far less precise and cannot link events across distant sites.

  The shift was substantial enough that historians of archaeology refer to it as the radiocarbon revolution. A second revolution followed when radiocarbon dates began contradicting older models of how innovations spread across prehistoric Europe. Researchers had assumed that new ideas moved by diffusion or by the migration of peoples. Radiocarbon dates showed that many changes must have arisen independently in different places. Taylor also noted a third revolution: the advent of AMS and its ability to date samples as small as individual plant seeds opened entirely new research questions that had been practically unanswerable before.

  Two applications illustrate the range. At Two Creeks in Wisconsin, a buried fossil forest had been dated by correlating sediment layers with Scandinavian sequences, yielding estimates of between 19,000 and 24,000 years old. Libby published radiocarbon dates for the site in 1952, returning an average of 11,404 years before present with a standard error of 350 years. Further testing in the 1990s with AMS produced dates ranging from 11,640 to 11,800 years before present, and a large interlaboratory comparison involving more than 70 laboratories yielded a median of 11,788 plus or minus 8 years before present. That result, when calibrated, gives a date range of 13,730 to 13,550 calendar years before present and is now treated as a key marker for the end of North American glaciation.

  In 1947, scrolls discovered near the Dead Sea were found to contain Hebrew and Aramaic texts, most attributed to the Essene sect. A sample of linen wrapping from the Great Isaiah Scroll was included in a 1955 analysis by Libby, who estimated its age at 1,917 plus or minus 200 years. Later AMS testing of 21 scrolls in the 1990s placed their dates between the early 4th century BC and the mid 4th century AD. In all but two cases, the radiocarbon result fell within 100 years of the age estimated from handwriting style alone. The Isaiah Scroll itself showed two possible date ranges at high confidence: a 15 percent probability it dates from 355 to 295 BC, and an 84 percent probability it dates from 210 to 45 BC.

  In 1988, three separate laboratories dated linen samples from the Shroud of Turin and placed its origins in the 14th century, raising serious questions about its claimed status as a 1st-century relic. The Shroud result is a reminder that radiocarbon dating enters public attention most sharply when it challenges cherished assumptions, which is precisely what it was designed to do.