Potassium-argon dating

Potassium-argon (K-Ar) dating is one of the most widely applied radiometric dating methods in the earth sciences, exploiting the radioactive decay of potassium-40 (⁴⁰K) to argon-40 (⁴⁰Ar) to determine the ages of potassium-bearing rocks and minerals. Because potassium is the seventh most abundant element in Earth's crust and is a major constituent of common rock-forming minerals such as feldspars, micas, and hornblende, the K-Ar system is applicable to an extraordinarily wide range of geological materials and settings.^{1, 2} The method can date materials from approximately 100,000 years old to the age of the Earth, spanning nearly the entire range of geological time. Its refined successor, the argon-argon (⁴⁰Ar/³⁹Ar) technique, introduced by Merrihue and Turner in 1966, measures both parent and daughter isotopes simultaneously on the same sample, dramatically improving precision, enabling step-heating experiments that reveal thermal histories, and extending the method's reach into the historical realm.^{3, 14}

Together, K-Ar and ⁴⁰Ar/³⁹Ar dating have played a transformative role in establishing the absolute timescale of Earth history. They provided the chronological framework for the geomagnetic polarity timescale, demonstrated the progressive age increase of volcanic islands along the Hawaiian chain that confirmed the hotspot hypothesis, and dated the volcanic tuffs that bracket key hominin fossil sites in East Africa, fundamentally shaping our understanding of human evolution.^{1, 8, 13}

The decay of potassium-40

The physical basis of K-Ar dating rests on the radioactive decay of ⁴⁰K, a naturally occurring but rare isotope that constitutes approximately 0.0117 percent of all potassium. Potassium-40 is unusual among geochronologically important radionuclides in that it decays by two competing pathways, a property known as branching decay. Approximately 89.5 percent of ⁴⁰K atoms undergo beta-minus (β⁻) decay to calcium-40 (⁴⁰Ca), while approximately 10.5 percent decay by electron capture to argon-40 (⁴⁰Ar).^{4, 15} A small fraction of the electron-capture events proceed to an excited state of ⁴⁰Ar, which then de-excites by emitting a 1.46 MeV gamma ray — the characteristic gamma-ray signature used to detect ⁴⁰K in geochemical and environmental studies.¹⁷

The total half-life of ⁴⁰K, combining both decay branches, is 1.2504 ± 0.0030 billion years, as determined by the IUGS Subcommission on Geochronology and recommended by Steiger and Jäger in 1977.⁴ The partial decay constant for the electron-capture branch to ⁴⁰Ar (λ_e) is 5.81 × 10⁻¹¹ yr⁻¹, while the partial decay constant for the beta decay to ⁴⁰Ca (λ_β) is 4.962 × 10⁻¹⁰ yr⁻¹. Only the electron-capture branch is useful for geochronology, because the ⁴⁰Ca produced by beta decay is indistinguishable from the vastly more abundant non-radiogenic ⁴⁰Ca already present in virtually all minerals.^{1, 2} The branching ratio — the fraction of ⁴⁰K decays that produce ⁴⁰Ar rather than ⁴⁰Ca — must therefore be known accurately, as it enters directly into the age equation. Renne and colleagues (2010) jointly redetermined the ⁴⁰K decay constants and the ⁴⁰Ar*/⁴⁰K ratio for the Fish Canyon sanidine standard, producing revised values that reduce systematic discrepancies between ⁴⁰Ar/³⁹Ar and U-Pb ages from approximately 1 percent to less than 0.2 percent.⁵

The age equation for conventional K-Ar dating takes the form t = (1/λ) × ln[1 + (λ/λ_e) × (⁴⁰Ar*/⁴⁰K)], where λ is the total decay constant, λ_e is the partial decay constant for the electron-capture branch, ⁴⁰Ar* is the radiogenic argon accumulated since the mineral last cooled through its closure temperature, and ⁴⁰K is the amount of potassium-40 remaining in the sample.^{1, 18} The equation assumes that no radiogenic argon was present at the time of mineral formation (or that any initial argon can be identified and corrected for), and that the system has remained closed to both potassium and argon since that time.

Conventional K-Ar dating

In the conventional K-Ar method, as developed and refined through the 1950s and 1960s, potassium and argon are measured on separate aliquots of the same sample. The potassium content is typically determined by flame photometry, atomic absorption spectrophotometry, or isotope dilution, while argon is extracted by fusing the sample in a vacuum or ultra-high-purity gas line, purifying the released gas, and measuring the ⁴⁰Ar/³⁶Ar ratio using a gas-source mass spectrometer.^{2, 16}

Atmospheric argon, which has a ⁴⁰Ar/³⁶Ar ratio of 298.56, is ubiquitous at Earth's surface and must be corrected for. The correction is straightforward in principle: any ³⁶Ar detected in the sample is assumed to be atmospheric (since ³⁶Ar has no significant radioactive parent), and the corresponding atmospheric ⁴⁰Ar is subtracted from the total measured ⁴⁰Ar to yield the radiogenic component (⁴⁰Ar*).^{1, 2} This atmospheric correction is reliable for samples with high radiogenic argon contents but becomes increasingly uncertain for very young or potassium-poor samples in which the radiogenic signal is small relative to the atmospheric background.

A significant limitation of the conventional method is that potassium and argon are measured on different portions of the sample. If the sample is heterogeneous — if, for instance, different mineral grains or domains within a rock have different potassium concentrations — the two aliquots may not be perfectly representative of each other, introducing systematic error. This problem motivated the development of the ⁴⁰Ar/³⁹Ar technique, in which both parent-proxy and daughter isotopes are measured simultaneously on the same material.^{1, 3}

The argon-argon method

The ⁴⁰Ar/³⁹Ar dating technique, first proposed by Merrihue and Turner in 1966, represents a major advance over conventional K-Ar dating.³ In this method, the sample is irradiated with fast neutrons in a nuclear reactor, converting a fraction of the stable isotope ³⁹K to ³⁹Ar via the reaction ³⁹K(n,p)³⁹Ar. Because the ratio of ³⁹K to ⁴⁰K in nature is effectively constant, the amount of reactor-produced ³⁹Ar is directly proportional to the amount of ⁴⁰K in the sample. After irradiation, the ⁴⁰Ar/³⁹Ar ratio is measured by mass spectrometry, and because both isotopes are argon, they are extracted, purified, and measured simultaneously from exactly the same material, eliminating the sample-heterogeneity problem inherent in the conventional K-Ar approach.^{1, 3}

The irradiation parameter J, which characterises the neutron fluence received by the sample, is determined by co-irradiating the unknown sample with a mineral standard of independently known age (a flux monitor). The Fish Canyon sanidine from the Fish Canyon Tuff of Colorado is the most widely used ⁴⁰Ar/³⁹Ar standard, with an astronomically calibrated age of 28.201 ± 0.046 million years.^{11, 21} Other commonly used standards include the Alder Creek sanidine and the GA-1550 biotite, each calibrated against the Fish Canyon sanidine or against independent astronomical or radiometric constraints.²¹

The most powerful feature of ⁴⁰Ar/³⁹Ar dating is the step-heating experiment, also called incremental heating. Rather than fusing the entire sample at once, the analyst heats the sample in a series of progressively higher temperature steps, typically using a resistance furnace or an infrared laser, and measures the ⁴⁰Ar/³⁹Ar ratio of the gas released at each step.¹ In an undisturbed sample, all temperature steps release gas with the same ⁴⁰Ar*/³⁹Ar ratio, and the resulting plot of apparent age versus cumulative ³⁹Ar released produces a flat age plateau, confirming that the mineral has behaved as a closed system since it cooled below its closure temperature. A well-defined plateau, typically comprising at least 50 percent of the total ³⁹Ar released across three or more consecutive steps, provides strong evidence for a reliable age.^{1, 17}

If the sample has experienced partial argon loss due to a later thermal event, the lower-temperature steps release gas from the most disturbed crystal domains and yield anomalously young apparent ages, while higher-temperature steps from more retentive domains approach the true crystallisation age. The resulting age spectrum has a characteristic staircase or saddle shape that immediately alerts the analyst to the thermal disturbance and, in favourable cases, allows both the original age and the timing of the disturbance to be constrained.^{1, 7} This diagnostic capability — the ability to distinguish reliable ages from disturbed ages within a single experiment — is the principal advantage of ⁴⁰Ar/³⁹Ar over conventional K-Ar dating and one of the reasons the method has largely supplanted its predecessor in modern geochronological practice.

Suitable minerals and closure temperatures

K-Ar and ⁴⁰Ar/³⁹Ar dating can be applied to any potassium-bearing mineral that retains argon quantitatively below a certain temperature, but some minerals are far more suitable than others. The concept of closure temperature, formalised by Dodson in 1973, defines the temperature below which a mineral becomes effectively closed to diffusive loss of the daughter isotope.⁶ Different minerals have different closure temperatures for argon, reflecting differences in crystal structure, composition, and the diffusion kinetics of argon atoms through the lattice. This variation is not a weakness but a strength: by dating multiple minerals from the same rock, geologists can reconstruct the thermal history of that rock as it cooled through successively lower temperatures, a practice known as thermochronology.^{1, 6}

Sanidine, a high-temperature polymorph of potassium feldspar found in volcanic rocks, is the most prized mineral for ⁴⁰Ar/³⁹Ar dating of eruption ages. Its extremely low closure temperature (on the order of 200 to 300 degrees Celsius, which in rapidly cooled volcanic rocks is exceeded almost instantaneously during eruption) and its tendency to crystallise with negligible initial argon make it an ideal recorder of eruption time.^{1, 11} Hornblende, a common amphibole in intermediate to mafic igneous and metamorphic rocks, has a much higher closure temperature of approximately 500 to 550 degrees Celsius, making it resistant to thermal resetting and well suited for dating crystallisation and high-grade metamorphic events.^{1, 6} Muscovite closes at approximately 350 to 400 degrees Celsius, biotite at approximately 300 to 350 degrees Celsius, and potassium feldspar (microcline, orthoclase) at approximately 150 to 350 degrees Celsius depending on the structural state and cooling rate.^{1, 6}

Argon closure temperatures for common minerals^{1, 6}

Whole-rock basalt can also be dated by the K-Ar and ⁴⁰Ar/³⁹Ar methods, which is particularly valuable because basaltic lavas are among the most common volcanic products but often lack large phenocrysts of datable minerals. However, whole-rock dating of basalts requires particular care because fine-grained groundmass can be susceptible to alteration by weathering and hydrothermal fluids, which may disturb the potassium and argon systematics. Step-heating experiments on irradiated whole-rock samples can often identify and exclude disturbed fractions.^{1, 2}

Assumptions and limitations

Like all radiometric methods, K-Ar and ⁴⁰Ar/³⁹Ar dating rest on a set of physical assumptions. The most fundamental are that the decay constants of ⁴⁰K are known accurately, that the system has remained closed to both potassium and argon since the event being dated, and that any non-radiogenic argon can be identified and corrected for. The step-heating technique provides a powerful internal test of the closed-system assumption, and the decay constants have been measured with increasing precision over several decades, but several sources of error remain important in practice.^{1, 5}

Excess argon is the most widely discussed complication in K-Ar geochronology. It refers to ⁴⁰Ar that is not atmospheric and was not produced by in situ decay of ⁴⁰K but was instead incorporated into the mineral from external sources — for example, from argon dissolved in magma, trapped in fluid inclusions, or diffused into the crystal from surrounding rocks at depth.⁷ Excess argon causes the measured ⁴⁰Ar*/⁴⁰K ratio to be higher than it should be, yielding an anomalously old apparent age. In conventional K-Ar dating, excess argon is difficult to detect because the method produces only a single age from each analysis. In ⁴⁰Ar/³⁹Ar step-heating experiments, however, excess argon often reveals itself through anomalously old ages in the low-temperature steps (where weakly bound or inclusion-hosted argon is preferentially released) or through saddle-shaped age spectra, providing a diagnostic tool for identifying and in some cases correcting for the problem.^{1, 7} An isochron plot of ⁴⁰Ar/³⁶Ar versus ³⁹Ar/³⁶Ar for the individual heating steps can also reveal excess argon, because the intercept on the ⁴⁰Ar/³⁶Ar axis will be higher than the atmospheric ratio of 298.56 if non-atmospheric, non-radiogenic ⁴⁰Ar is present.^{7, 17}

Argon loss is the complementary problem: the partial escape of radiogenic ⁴⁰Ar from the mineral lattice due to reheating, deformation, or chemical alteration, which causes the apparent age to be younger than the true age. Argon loss is most common in minerals with low closure temperatures, such as biotite and K-feldspar, and in rocks that have experienced post-crystallisation thermal events such as metamorphism, burial heating, or nearby intrusions. In step-heating experiments, argon loss typically produces a rising age spectrum in which the low-temperature steps yield young apparent ages that increase progressively toward a plateau or maximum at higher extraction temperatures.^{1, 2}

Weathering and alteration can also compromise K-Ar ages by adding or removing potassium from the mineral (for example, through the alteration of feldspar to clay minerals) or by opening pathways for argon diffusion. For this reason, careful petrographic examination and sample selection are essential steps in any K-Ar or ⁴⁰Ar/³⁹Ar study, and visibly altered or weathered samples are routinely rejected.^{2, 16}

A further source of uncertainty arises from the decay constants themselves. The Steiger and Jäger (1977) recommended values remained the international standard for decades, but comparison of ⁴⁰Ar/³⁹Ar ages with high-precision U-Pb ages on the same rocks consistently revealed a systematic discrepancy of approximately 1 percent, with ⁴⁰Ar/³⁹Ar ages appearing younger.^{4, 19} Renne and colleagues (2010) resolved much of this discrepancy by jointly optimising the ⁴⁰K decay constants and the age of the Fish Canyon sanidine standard against astronomical age constraints, demonstrating that the Steiger and Jäger total decay constant was slightly too high.⁵ Adoption of the revised decay constants has brought ⁴⁰Ar/³⁹Ar ages into much closer agreement with U-Pb ages, although the geochronological community continues to refine these values.^{5, 11}

Dating human evolution sites

Perhaps no application of K-Ar and ⁴⁰Ar/³⁹Ar dating has had a greater impact outside the earth sciences than the dating of hominin fossil sites in the East African Rift System, where volcanic activity has fortuitously interbedded datable tuffs (volcanic ash layers) with fossil-bearing sedimentary deposits. The potassium-argon method provided the first absolute ages for these sites and, in doing so, revolutionised the understanding of the antiquity and pace of human evolution.^{8, 10}

The landmark application came in 1961, when Leakey, Evernden, and Curtis published K-Ar dates of approximately 1.75 million years for Bed I at Olduvai Gorge, Tanzania, where Mary Leakey had discovered the skull of Paranthropus boisei (then called Zinjanthropus) in 1959.⁸ This age was far older than anyone had anticipated — contemporary estimates of human evolutionary timescales, based largely on extrapolations from mammalian faunal turnover, had placed the earliest hominins at no more than a few hundred thousand years old. The K-Ar dates from Olduvai immediately doubled or tripled the accepted timespan of human evolution and established radiometric dating as an indispensable tool in paleoanthropology.^{8, 16}

Subsequent ⁴⁰Ar/³⁹Ar work has refined and extended the chronology of East African hominin sites with dramatically improved precision. Walter (1994) used single-crystal laser-fusion ⁴⁰Ar/³⁹Ar dating of volcanic feldspars to date the deposits at Hadar, Ethiopia, that contain the famous Australopithecus afarensis skeleton known as "Lucy" to 3.18 million years ago, and the "First Family" assemblage to 3.20 million years ago.⁹ Deino (2011) applied ⁴⁰Ar/³⁹Ar dating to tuffs at Laetoli, Tanzania, establishing that the Upper Laetolil Beds containing A. afarensis fossils and the famous hominin footprint trail span 3.85 to 3.63 million years, with the Footprint Tuff (Tuff 7) dated to 3.66 million years.²⁰ These dates provide the temporal framework within which the emergence and evolution of australopithecines, the earliest members of the genus Homo, and the development of stone tool technology are understood.¹⁰

The ⁴⁰Ar/³⁹Ar method has also been applied to sites well beyond East Africa. Volcanic deposits at hominin sites in Indonesia, China, the Caucasus, and Europe have been dated, extending the chronological reach of the method across the global range of Homo erectus and other archaic human species.¹⁰ The ability to date volcanic minerals as young as a few thousand years was dramatically demonstrated by Renne and colleagues (1997), who obtained a ⁴⁰Ar/³⁹Ar isochron age of 1,925 ± 94 years for sanidine from the AD 79 eruption of Vesuvius, in close agreement with the historically documented date provided by the Roman author Pliny the Younger.¹⁴

Calibrating the geologic time scale

K-Ar and ⁴⁰Ar/³⁹Ar dating have been central to establishing the absolute calibration of the geologic time scale. Because potassium-bearing minerals are common in volcanic ash beds that can be correlated with the biostratigraphic and magnetostratigraphic record, the method provides critical tie-points that convert relative geological sequences into absolute chronologies.^{1, 11}

One of the most important contributions of K-Ar dating to global stratigraphy was the establishment of the geomagnetic polarity timescale. In the 1960s, Allan Cox, Richard Doell, and Brent Dalrymple at the U.S. Geological Survey combined K-Ar ages of young volcanic rocks with measurements of their magnetic polarity to construct the first quantitative record of geomagnetic field reversals. By dating many lavas from around the world and determining whether each preserved normal or reversed magnetic polarity, they demonstrated that the Earth's magnetic field had reversed at specific, globally synchronous times, and they established the sequence of polarity epochs and events (now called chrons) that constitute the geomagnetic polarity timescale.^{2, 16} This timescale, in turn, was critical to the recognition of magnetic anomaly patterns on the ocean floor and the confirmation of seafloor spreading, becoming one of the key lines of evidence for plate tectonics.¹⁶

K-Ar dating of lavas in the Hawaiian Islands by McDougall (1964) revealed a systematic age progression along the island chain, from the oldest volcanoes in the northwest (Kauai, approximately 5.1 million years) to the youngest in the southeast (the Big Island of Hawaii, less than 0.7 million years). This age gradient provided the first quantitative support for the hotspot hypothesis proposed by J. Tuzo Wilson, in which a stationary mantle plume generates volcanoes that are carried progressively away from the heat source by the moving Pacific Plate.¹³

The precision of ⁴⁰Ar/³⁹Ar dating has continued to improve, allowing it to contribute to the calibration of geological boundaries at the level of tens of thousands of years. Kuiper and colleagues (2008) used astronomical tuning of marine sedimentary cycles in Morocco to recalibrate the age of the Fish Canyon sanidine standard, reducing the absolute uncertainty of ⁴⁰Ar/³⁹Ar ages from approximately 2.5 percent to approximately 0.25 percent and establishing a mutually consistent age for the Cretaceous-Paleogene boundary of approximately 65.95 million years.¹¹ Renne and colleagues (2013) subsequently used this improved calibration to demonstrate synchrony between the Cretaceous-Paleogene boundary mass extinction, the Chicxulub bolide impact, and the eruption of the Deccan Traps to within 32,000 years, providing the tightest temporal constraints yet achieved on the relationship between these events.¹²

Cross-validation with other methods

The reliability of the K-Ar and ⁴⁰Ar/³⁹Ar methods is demonstrated by their systematic agreement with independent dating techniques applied to the same geological materials. Cross-validation between ⁴⁰Ar/³⁹Ar and U-Pb ages on co-existing minerals from the same volcanic units has been a major focus of the geochronological community, both as a test of the methods themselves and as a means of improving the accuracy of the decay constants.^{5, 19}

Min and colleagues (2000) compared ⁴⁰Ar/³⁹Ar biotite and U-Pb zircon ages from a 1.1-billion-year-old rhyolite, finding agreement to within 1 percent and demonstrating that any systematic error in the ⁴⁰Ar/³⁹Ar method over this vast timespan is small.¹⁹ The residual discrepancy was consistent with a slight inaccuracy in the then-accepted ⁴⁰K decay constants, a finding that contributed to the subsequent revision by Renne and colleagues.^{5, 19} At the young end of the age range, the agreement between the ⁴⁰Ar/³⁹Ar age of the AD 79 Vesuvius eruption and the historically documented date provides a direct calibration check on the entire method, from neutron irradiation and mass spectrometry to the decay constants and standard ages.¹⁴

Cross-validation of ⁴⁰Ar/³⁹Ar ages against independent methods^{5, 12, 14, 19}

The agreement between ⁴⁰Ar/³⁹Ar ages and those obtained by U-Pb, astronomical tuning, and historical documentation across timescales spanning from two thousand to over one billion years constitutes one of the most compelling demonstrations of the reliability of radiometric dating. The systematic convergence of independent methods, each based on different physical principles and measured on different materials, leaves no reasonable doubt that these techniques are recording real elapsed time.^{5, 16}

Historical significance

The development of K-Ar dating in the 1950s and 1960s filled a critical gap in the geochronologist's toolkit. While the U-Pb method, applied primarily to the mineral zircon, provided unparalleled precision for dating ancient igneous and metamorphic rocks, zircon is relatively uncommon in mafic volcanic rocks — precisely the rock type most abundant at Earth's surface and most directly associated with volcanic eruptions, magnetic reversal records, and fossil-bearing sedimentary sequences. Potassium, by contrast, is present in virtually every common rock-forming mineral, and the gaseous daughter product argon is efficiently released upon heating, making measurement straightforward with mid-twentieth-century mass spectrometry technology.^{2, 16}

The pioneering work of Garniss Curtis and Jack Evernden at the University of California, Berkeley, in the late 1950s and early 1960s extended K-Ar dating from geological applications to the dating of human evolutionary sites in East Africa, a move that transformed paleoanthropology from a discipline that relied on relative faunal dating to one anchored in absolute chronology.⁸ G. Brent Dalrymple and Marvin Lanphere at the U.S. Geological Survey produced the definitive monograph on the method in 1969, establishing the protocols and error analysis that guided a generation of K-Ar laboratories.² The subsequent development of the ⁴⁰Ar/³⁹Ar technique, advanced by Turner, McDougall, Harrison, Renne, Deino, and their colleagues, elevated the method to a level of precision and diagnostic power that rivals U-Pb geochronology for many applications.^{1, 3}

Today, ⁴⁰Ar/³⁹Ar dating is applied across an extraordinary range of problems in Earth and planetary science: from the dating of lunar basalts and Martian meteorites to the calibration of the Phanerozoic time scale, from the determination of cooling rates in metamorphic terranes to the forensic dating of young volcanic hazards. Its versatility, combined with the diagnostic power of the step-heating experiment and the continuously improving precision of the technique, ensures that potassium-argon geochronology will remain a cornerstone of geological dating for decades to come.^{1, 10, 12}