What is Radiocarbon Dating: Nuclear Physics, Calibration Science & Archaeological Application?
Carbon-14 (¹⁴C) radiocarbon dating is the single most widely used absolute dating method in archaeology, geology, and environmental science. Developed by Willard Libby in 1949 (Nobel Prize in Chemistry, 1960), the method exploits the radioactive decay of ¹⁴C — a cosmogenic isotope produced in the upper atmosphere when cosmic-ray neutrons react with ¹⁴N — to determine when an organism stopped exchanging carbon with the atmosphere (i.e., when it died). The method is reliable for organic materials up to ~50,000 years old, covering the entire span of human civilization and the Late Pleistocene.
Mathematical Foundation
Radiocarbon Age (Conventional)
= Radiocarbon age in years before present (BP), where 'present' is defined by convention as AD 1950. This is the uncalibrated (conventional) radiocarbon age — it must be calibrated against dendrochronologically-dated tree ring sequences or marine coral records to obtain a true calendar age, because atmospheric ¹⁴C concentration has not been constant over time.
= Decay constant = ln(2) / t₁/₂ = 0.693147 / 5,730 = 1.2097 × 10⁻⁴ per year. This is the probability per unit time that a single ¹⁴C atom will decay. For N atoms: the number decaying per unit time (activity) is A = λN. The decay constant is the reciprocal of the mean lifetime (τ = 1/λ = 8,267 years).
= Half-life of carbon-14 = 5,730 ± 40 years (Cambridge half-life, Godwin 1962). By convention, radiocarbon laboratories still report ages using the Libby half-life of 5,568 years for historical consistency — the difference is corrected during calibration. After one half-life, 50% of original ¹⁴C remains; after two, 25%; after ten (~57,300 years), only ~0.1% remains — effectively the detection limit of conventional decay counting.
= Ratio of measured sample activity (A_s) to the activity of the modern reference standard (A_0). A_0 is defined as 95% of the activity of NBS oxalic acid I in AD 1950 = 13.56 disintegrations per minute per gram of carbon (dpm/gC). For AMS (Accelerator Mass Spectrometry): this ratio is measured directly as ¹⁴C/¹²C isotope ratio rather than by counting decay events. AMS requires only 1–10 mg of carbon (vs 1–5 grams for conventional counting), enabling dating of individual seeds, pollen grains, and tiny textile fragments.
Percent Modern Carbon (pMC)
= Percent modern carbon — the sample's ¹⁴C activity expressed as a percentage of the modern standard (AD 1950). A pMC of 100% = modern (zero age). A pMC of 50% = one half-life old (5,730 years). A pMC of 0% = infinite age (beyond detection). AMS labs commonly report results as pMC or F¹⁴C (fraction modern) rather than conventional radiocarbon age BP, especially for samples younger than AD 1950 where nuclear weapons testing created excess atmospheric ¹⁴C ('bomb carbon'). pMC > 100% ('greater than modern') means the sample contains bomb carbon and post-dates 1955.
Laws & Principles
- The Fundamental Assumption — Atmospheric Equilibrium: Libby's original model assumed that the atmospheric ¹⁴C/¹²C ratio has been constant through time. This is wrong — atmospheric ¹⁴C fluctuates due to: (1) Solar activity cycles (Schwabe 11-year, Gleissberg 80-year, de Vries/Suess 210-year): reduced solar magnetic shielding → increased cosmic ray flux → higher ¹⁴C production. (2) Geomagnetic field variations: a weaker geomagnetic dipole allows more cosmic rays to reach the atmosphere. (3) Carbon cycle changes: ocean upwelling, volcanic CO2 outgassing, and biospheric carbon reservoir changes redistribute ¹⁴C between atmosphere, ocean, and biosphere. (4) Anthropogenic effects: fossil fuel burning since ~1850 has diluted atmospheric ¹⁴C (Suess effect); nuclear weapons testing 1955–1963 nearly doubled atmospheric ¹⁴C (bomb spike). These variations are why raw radiocarbon ages must be CALIBRATED against independently dated records (tree rings, varves, marine corals) to obtain true calendar ages.
- Calibration — IntCal20, Marine20, SHCal20: The current international calibration curves (Reimer et al., 2020) are: IntCal20 — for Northern Hemisphere terrestrial samples, extending to 55,000 cal BP using tree rings (to ~13,910 cal BP), then floating tree-ring chronologies, Cariaco Basin varves, Hulu Cave speleothems, and marine foraminifera beyond tree ring range. Marine20 — for marine samples (shells, corals, marine sediments). Marine samples appear ~400 years older than contemporaneous atmospheric samples due to the Marine Reservoir Effect (MRE). Regional ΔR corrections must be applied because upwelling intensity varies by ocean basin. SHCal20 — for Southern Hemisphere samples, offset ~40 years older than IntCal20 due to greater ocean surface area in the southern hemisphere causing more ¹⁴C exchange between atmosphere and ocean. Calibration converts a single radiocarbon age (e.g., 3,450 ± 30 BP) into a probability distribution of calendar ages. The resulting calibrated age range is typically wider than the measurement uncertainty because the calibration curve is not monotonic (it has wiggles and plateaus).
- AMS vs Conventional β-Counting: Two measurement technologies exist. Conventional β-counting (Libby's original method) detects ¹⁴C decay events using a gas proportional counter or liquid scintillation counter. It requires 1–5 grams of pure carbon and counting times of 24–72 hours to accumulate sufficient statistics. Precision: ±40–80 years. Accelerator Mass Spectrometry (AMS) directly counts ¹⁴C atoms by accelerating a carbon ion beam through a tandem electrostatic accelerator and separating isotopes by mass-to-charge ratio. AMS requires only 0.5–10 mg of carbon and achieves precision of ±15–30 years. AMS is now the dominant method (>95% of all radiocarbon dates). AMS can date individual seeds, pollen concentrates, specific organic compounds (compound-specific radiocarbon analysis — CSRA), and micro-samples from irreplaceable artifacts (Dead Sea Scrolls, Shroud of Turin, Ötzi the Iceman).
- Contamination and Pretreatment (AAA, ABOx-SC): The most critical source of error in radiocarbon dating is contamination — introduction of carbon that was not part of the original sample. Young contamination (rootlets, humic acids, modern handling) makes samples appear younger; old contamination (calcium carbonate, coal, bitumen) makes samples appear older. Standard pretreatment protocols: AAA (Acid-Alkali-Acid) — removes carbonates (HCl), humic/fulvic acids (NaOH), and atmospheric CO2 absorbed during alkali step (second HCl). Used for charcoal, wood, textiles. ABOx-SC (Acid-Base Oxidation stepped combustion) — aggressive pretreatment for old charcoal (>30,000 BP); removes all non-aromatic carbon to isolate the most resistant graphitic fraction. Extends reliable dating to ~55,000 BP. Collagen extraction (for bone) — demineralization in HCl, gelatinization, ultrafiltration (>30 kDa cutoff). Ultrafiltered collagen is mandatory for bone dating beyond 20,000 BP. Always request information about pretreatment method from the dating laboratory — it directly affects result reliability.
Step-by-Step Example Walkthrough
" An archaeologist submits a charcoal sample from a hearth at a Clovis-era site in North America. The AMS lab reports: radiocarbon age = 10,950 ± 30 BP (measured using ABOx-SC pretreatment). What is the calibrated calendar age? "
- 1. Conventional radiocarbon age: 10,950 ± 30 BP (using the Libby half-life of 5,568 years by convention).
- 2. This corresponds to A_s/A_0 = e^(-10950 × ln2/5568) = e^(-1.364) = 0.2558, or ~25.6% modern carbon (pMC).
- 3. Apply IntCal20 calibration curve (Northern Hemisphere terrestrial sample).
- 4. The IntCal20 curve at 10,950 BP maps to a calendar age of approximately 12,750–12,890 cal BP (68.2% probability) or 12,710–12,940 cal BP (95.4%).
- 5. This calibrated range falls within the established Younger Dryas chronozone and is consistent with the terminal Clovis cultural complex.
- 6. Note: the calibrated age is ~1,800–1,940 years OLDER than the raw radiocarbon age. This offset is because atmospheric ¹⁴C was higher than modern levels during the Late Glacial period (so the sample appears younger in radiocarbon years than it truly is in calendar years).
Final Result: Radiocarbon age: 10,950 ± 30 BP. Calibrated age: 12,750–12,890 cal BP (1σ). The ~1,900-year offset between radiocarbon and calendar age demonstrates why calibration is mandatory for archaeological interpretation. Software: use OxCal (Oxford), CALIB (University of Washington), or Bchron (R package) for formal calibration.