Half-Life

Half-life (customarily denoted t_1/2) is the time interval required for a quantity undergoing decay, elimination, or chemical reaction to fall to half (50%) of its initial value. Introduced in 1904 by Ernest Rutherford (originally termed the "half-value period") to characterize the invariant rate at which radioactive elements undergo radioactive decay, the concept of half-life applies across nuclear physics, elementary particle physics, chemical reaction kinetics, environmental toxicology, and clinical pharmacology.

In any first-order exponential decay process—where the instantaneous rate of loss is directly proportional to the amount of substance currently present (dN/dt = −λ*N*)—the half-life t_1/2 = ln(2) / λ ≈ 0.693147 / λ is a constant characteristic of the system, independent of the starting amount N₀. By contrast, in zero-order and second-order chemical or enzymatic reactions, the half-life depends explicitly on the initial reactant concentration [A]₀, providing a diagnostic test for determining reaction order in physical chemistry and pharmacodynamics.

First-Order Exponential Decay Law (t_1/2 = ln(2) / λ = τ ln 2)

Consider a population of N(t) unstable particles, excited fluorophores, or reactant molecules whose rate of disappearance obeys first-order kinetics with rate (or decay) constant λ (units of inverse time, s⁻¹):

Integrating from N(0) = N₀ at time t = 0 to N(t) at time t yields the exponential decay law:

Setting N(t_1/2) = N₀ / 2 and taking the natural logarithm of both sides defines the first-order half-life:

Closely related to the half-life is the mean lifetime (τ), defined as the statistical expectation value of the survival time t of an individual particle integrated over the normalized probability density P(t) = λ exp(−λ*t*):

At t = τ (one mean lifetime), the remaining fraction of the population is N(τ) / N₀ = exp(−1) ≈ 0.367879 (36.79%), whereas at t = t_1/2 = τ ln(2) ≈ 0.6931 τ, the remaining fraction is exactly 50%. Expressing the decay law in base-2 powers of the number of elapsed half-lives n = t / t_1/2 gives the universal scaling identity:

Thus, after n = 1 half-life, 50% remains; after n = 2 half-lives, 25% (1/4) remains; after n = 3 half-lives, 12.5% (1/8) remains; after n = 5 half-lives, 3.125% (1/32) remains; and after n = 10 half-lives, 2⁻¹⁰ = 1/1,024 ≈ 0.0977% (<0.1%) of the original sample remains.

Zero-Order, First-Order, and Second-Order Half-Lives

In chemical thermodynamics and chemical kinetics, a reaction aA → products governed by a power-law rate expression −*d*[A]/dt = k_m [A]^m exhibits qualitatively different half-life behavior depending on the reaction order (m):

1. First-Order Reactions (m = 1, e.g., thermal isomerization of cyclopropane, N₂O₅ decomposition, and nuclear decay): Because t_1/2 = ln(2) / k₁ contains no concentration term, each successive 50% reduction (1.0 M → 0.5 M → 0.25 M → 0.125 M) takes the exact same time interval.
2. Zero-Order Reactions (m = 0, e.g., heterogeneous catalysis on saturated Pt/Au surfaces or hepatic oxidation of ethanol by alcohol dehydrogenase at blood concentrations above K_m): The reaction proceeds at a constant velocity k₀ regardless of reactant concentration until the substrate is exhausted. Consequently, t_1/2 = [A]₀ / (2k₀) shrinks by half with each successive half-life, and the reactant reaches complete zero concentration at finite time t_end = [A]₀ / k₀ = 2 t_1/2.
3. Second-Order Reactions (m = 2, e.g., 2 HI → H₂ + I₂ or homodimerization): Because molecules must collide bimolecularly, collisions become rarer as concentration drops. The half-life t_1/2 = 1 / (k₂ [A]₀) is inversely proportional to initial concentration, meaning each successive half-life is twice as long as the preceding one (e.g., 10 min for 1.0 M → 0.5 M, 20 min for 0.5 M → 0.25 M, 40 min for 0.25 M → 0.125 M).

Volume of Distribution (V_d), Clearance (CL), and Dosing Steady State

In clinical pharmacology and drug discovery, the plasma elimination half-life (t_1/2) of a drug eliminated by first-order renal excretion and hepatic cytochrome P450 metabolism is governed by two primary physiological parameters—the apparent volume of distribution (V_d, in liters) and total systemic clearance (CL, in L h⁻¹):

This relationship guides clinical dosing regimens and clinical trial phases: when a drug is administered at a constant infusion rate or repeated regular dosing intervals, plasma drug concentrations approach pharmacokinetic steady state (C_ss) asymptotically as 1 − 2^−*n*, reaching 87.5% of steady state after 3 half-lives and 93.75% to 96.88% of steady state after 4 to 5 half-lives (and likewise requiring ~5 half-lives for >96.8% washout after discontinuation).

Parallel Physical and Biological Clearance (t_eff) in Nuclear Medicine

When a radiopharmaceutical (such as ¹³¹I-sodium iodide in thyroid therapy or ^99mTc-MDP in bone scintigraphy) is administered to a patient, the total radioactivity inside the body decreases simultaneously through two independent parallel first-order pathways: physical radioactive decay (with decay constant λ_phys = ln(2)/t_phys) and biological excretion via urine, feces, and sweat (with biological elimination constant λ_biol = ln(2)/t_biol). Because parallel first-order rate constants add linearly (λ_eff = λ_phys + λ_biol), the reciprocal of the effective half-life (t_eff) is the sum of the reciprocals:

Consequently, the effective half-life t_eff is always shorter than both the physical half-life and the biological half-life. For example, ¹³¹I has a physical half-life of t_phys = 8.02 days and a thyroidal biological half-life of t_biol ≈ 24 days, yielding an effective patient clearance half-life of t_eff = (8.02 × 24) / (8.02 + 24) ≈ 6.01 days.

Why doesn't a radioactive sample completely disappear after two half-lives (50% + 50% = 100%)?

In a first-order decay process, each half-life reduces the currently remaining population by 50%, not 50% of the original starting total. During the first half-life, 50% of the original nuclei decay, leaving 50% intact. During the second half-life, half of that remaining 50% decays (which is 25% of the original sample), leaving 25% intact; during the third half-life, half of that 25% decays, leaving 12.5%, and so on as N(t) = N₀ (1/2)ⁿ until only a few discrete atoms remain and decay stochastically to zero.

What is the numerical difference between half-life (t_1/2) and mean lifetime (τ)?

For any exponential decay N(t) = N₀ exp(−λ*t*), the mean lifetime (τ = 1/λ) is the arithmetic average time that an individual particle survives before decaying (the time at which 1/e ≈ 36.79% of the sample remains), whereas the half-life (t_1/2 = τ ln 2) is the median survival time (the time at which 50% of the sample remains). Because some particles survive for many half-lives in the long exponential decay tail of the survival distribution, the mean lifetime τ is always longer than the median half-life by a factor of 1 / ln(2) ≈ 1.4427 (t_1/2 ≈ 0.6931 τ).

Why does alcohol (ethanol) elimination in humans not have a constant half-life?

Most therapeutic drugs are present at micromolar concentrations well below the Michaelis constant (K_m) of liver metabolizing enzymes, so their metabolism is first-order with a constant half-life. By contrast, social consumption of ethanol produces blood concentrations (10–20 mM) that completely saturate hepatic alcohol dehydrogenase (ADH1B) (K_m ≈ 0.05–2 mM). Because the enzyme operates at its maximum velocity (V_max), ethanol is cleared by zero-order kinetics at a flat rate of roughly 15 to 20 mg dL⁻¹ per hour (~7–10 g h⁻¹), meaning its effective half-life increases proportionally with blood alcohol concentration.

1. Rutherford, E. (1904). Radio-activity. Cambridge University Press. https://doi.org/10.5962/bhl.title.19393
2. Atkins, P., & de Paula, J. (2014). Atkins' Physical Chemistry (10th ed.). Oxford University Press. https://global.oup.com/academic/product/atkins-physical-chemistry-9780199697403
3. Rowland, M., & Tozer, T. N. (2011). Clinical Pharmacokinetics and Pharmacodynamics: Concepts and Applications (4th ed.). Lippincott Williams & Wilkins. https://shop.lww.com/Clinical-Pharmacokinetics-and-Pharmacodynamics/p/9780781750097
4. Audi, G., Kondev, F. G., Wang, M., Huang, W. J., & Naimi, S. (2017). "The NUBASE2016 evaluation of nuclear properties." Chinese Physics C, 41(3), 030001. https://doi.org/10.1088/1674-1137/41/3/030001