Samarium-neodymium dating

Samarium-neodymium (Sm-Nd) dating is a radiometric method that exploits the radioactive decay of samarium-147 (¹⁴⁷Sm) to neodymium-143 (¹⁴³Nd) by alpha emission. With a half-life of approximately 106 billion years — roughly seven times the current age of the universe — the ¹⁴⁷Sm-¹⁴³Nd system is one of the longest-lived decay schemes used in geochronology and is particularly well suited to dating ancient rocks from the Archean and Hadean eons.^{1, 2} The method occupies a distinctive niche in radiometric dating because samarium and neodymium are both rare earth elements (REE) with very similar ionic radii and geochemical properties, making the Sm-Nd system exceptionally resistant to disturbance during metamorphism and hydrothermal alteration — conditions that commonly reset the rubidium-strontium (Rb-Sr) and potassium-argon (K-Ar) systems.^{1, 9}

Beyond direct age determination, Sm-Nd isotope systematics have become one of the most powerful tools in geochemistry for tracing the long-term evolution of Earth's mantle and crust. The epsilon-Nd notation, depleted mantle model ages, and the extinct ¹⁴⁶Sm-¹⁴²Nd chronometer together provide a framework for understanding when and how the planet's major geochemical reservoirs — the depleted upper mantle, the continental crust, and potentially hidden deep reservoirs — differentiated from an initially homogeneous silicate Earth.^{3, 4, 11}

The ¹⁴⁷Sm-¹⁴³Nd decay system

The physical basis of Sm-Nd dating rests on the alpha decay of ¹⁴⁷Sm, a naturally occurring isotope that constitutes approximately 15 percent of natural samarium. During alpha decay, the ¹⁴⁷Sm nucleus emits an alpha particle (a helium-4 nucleus) and transforms directly into ¹⁴³Nd in a single step, without intermediate daughter products — a considerably simpler decay path than the multi-step uranium decay chains used in U-Pb dating.^{1, 9} The decay energy of ¹⁴⁷Sm is only 2.31 MeV, among the lowest of any naturally occurring alpha emitter, which accounts for its extremely long half-life.

The half-life of ¹⁴⁷Sm has been the subject of careful measurement over several decades. The value adopted by the international geochronological community following the 1977 convention of Steiger and Jäger is 1.06 × 10¹¹ years (106 billion years), corresponding to a decay constant (λ) of 6.54 × 10^-12 per year.^{2, 10} This value was originally determined by Lugmair and Marti (1978), who selected the most precise results from earlier direct counting experiments and validated them against lunar basalt chronology.² The IUPAC-IUGS Subcommission on Geochronology confirmed this recommended value in 2020, noting that although some more recent measurements have yielded slightly different values, the Lugmair and Marti determination remains consistent with the majority of geological cross-calibrations against the U-Pb system.¹⁷

Neodymium has seven naturally occurring isotopes, of which ¹⁴⁴Nd is stable and non-radiogenic. The ratio ¹⁴³Nd/¹⁴⁴Nd is the fundamental measured quantity in Sm-Nd geochronology, because ¹⁴⁴Nd serves as a normalising reference isotope whose abundance does not change over time. The present-day ¹⁴³Nd/¹⁴⁴Nd ratio in any sample reflects the initial ratio at the time the system closed plus the radiogenic ¹⁴³Nd accumulated from ¹⁴⁷Sm decay since that time.^{1, 9}

Resistance to metamorphic resetting

The defining advantage of the Sm-Nd system over other long-lived radiometric methods is its remarkable resistance to isotopic disturbance during geological events that routinely reset competing chronometers. This resistance arises from the fundamental geochemistry of the rare earth elements. Samarium (atomic number 62) and neodymium (atomic number 60) are adjacent members of the lanthanide series, with nearly identical ionic radii (Sm³⁺ = 1.08 angstroms, Nd³⁺ = 1.11 angstroms in eightfold coordination) and the same trivalent oxidation state in virtually all geological environments.^{1, 9} Because of this close chemical similarity, geological processes that mobilise and redistribute elements — such as metamorphic fluid flow, hydrothermal alteration, and low-temperature weathering — affect samarium and neodymium almost identically. When both parent and daughter move together, the Sm/Nd ratio of a rock or mineral changes little, and the isotopic clock is preserved.

This contrasts sharply with the Rb-Sr system, where the parent (rubidium, an alkali metal) and daughter (strontium, an alkaline earth metal) have very different geochemical behaviours. Rubidium is highly mobile in aqueous fluids and preferentially concentrated in micas and potassium feldspars, while strontium substitutes readily into plagioclase and carbonates. Metamorphic fluids can selectively remove rubidium or strontium from a rock, resetting the Rb-Sr clock while leaving the Sm-Nd system undisturbed.^{1, 6} For this reason, Sm-Nd whole-rock isochrons frequently preserve the original igneous crystallisation age of rocks that have been through one or more episodes of regional metamorphism — events that partially or completely reset their Rb-Sr and K-Ar systematics.^{7, 9}

The resistance of Sm-Nd to resetting is not absolute, however. At very high metamorphic grades, or during prolonged high-temperature events, the rare earth elements can be redistributed at the mineral scale, resetting mineral-mineral isochrons while leaving whole-rock compositions relatively undisturbed. In such cases, Sm-Nd mineral isochrons may record the age of the metamorphic event rather than the original crystallisation, a distinction that must be carefully evaluated through petrographic and geochemical context.¹

The Sm-Nd isochron method

Like other long-lived radiometric systems, Sm-Nd dating is most rigorously applied through the isochron method, which simultaneously determines both the age and the initial isotopic composition of a suite of cogenetic samples without requiring prior assumption about either quantity. The isochron equation for the Sm-Nd system is:

(¹⁴³Nd/¹⁴⁴Nd)_measured = (¹⁴³Nd/¹⁴⁴Nd)_initial + (¹⁴⁷Sm/¹⁴⁴Nd)_measured × (e^λt − 1)

This equation has the form of a straight line (y = b + mx), where the measured ¹⁴³Nd/¹⁴⁴Nd ratio is plotted on the vertical axis against the measured ¹⁴⁷Sm/¹⁴⁴Nd ratio on the horizontal axis. For a suite of samples that formed at the same time from a source with a homogeneous ¹⁴³Nd/¹⁴⁴Nd ratio but variable Sm/Nd ratios, the data define a straight line whose slope is proportional to the age and whose y-intercept gives the initial ¹⁴³Nd/¹⁴⁴Nd ratio at the time of formation.^{1, 9}

In practice, Sm-Nd isochrons are constructed from suites of whole-rock samples of varying composition (for example, different members of a layered igneous intrusion) or from separated mineral phases within a single rock. Minerals that preferentially incorporate the heavy rare earth elements, such as garnet and clinopyroxene, have higher Sm/Nd ratios than minerals enriched in the light rare earth elements, such as plagioclase and apatite. This spread in ¹⁴⁷Sm/¹⁴⁴Nd provides the leverage needed to define a precise isochron slope.^{1, 19} A tight linear alignment of data points on the isochron diagram confirms that the analysed suite has behaved as a closed isotopic system since formation, while scatter from the line signals open-system behaviour and alerts the analyst that a simple age interpretation is not warranted.

A significant limitation of Sm-Nd isochrons compared with U-Pb concordia diagrams is that the fractionation of samarium from neodymium during geological processes is relatively small. Typical crustal rocks have ¹⁴⁷Sm/¹⁴⁴Nd ratios in the range of 0.10 to 0.13, with mantle-derived rocks extending to about 0.20 — a total variation of only a factor of two.^{1, 9} This limited spread means that even small analytical uncertainties on the measured isotopic ratios translate into relatively large age uncertainties, and the precision of Sm-Nd ages is generally lower than that of U-Pb zircon ages by roughly an order of magnitude. For ancient rocks, however, the long time available for radiogenic ingrowth partially compensates for the small parent-daughter fractionation, and Sm-Nd isochrons on Archean samples can achieve age precisions of 30 to 100 million years.^{1, 15}

Epsilon notation and the chondritic reference

One of the most influential conceptual contributions of Sm-Nd isotope geochemistry is the epsilon notation (ε_Nd), introduced by DePaolo and Wasserburg in 1976 to express the ¹⁴³Nd/¹⁴⁴Nd ratio of a sample as a deviation from a chondritic reference value.³ The reference is the Chondritic Uniform Reservoir (CHUR), which represents the ¹⁴³Nd/¹⁴⁴Nd ratio that the bulk silicate Earth would have today if it had evolved with the same Sm/Nd ratio as chondritic meteorites — the undifferentiated building blocks of the Solar System. The present-day CHUR values, as determined by Jacobsen and Wasserburg (1980) from measurements on five chondrites and one achondrite, are ¹⁴³Nd/¹⁴⁴Nd = 0.512638 and ¹⁴⁷Sm/¹⁴⁴Nd = 0.1967.⁴

The epsilon value is defined as the deviation of a sample's ¹⁴³Nd/¹⁴⁴Nd ratio from the CHUR value at the same time, expressed in parts per ten thousand:

ε_Nd(t) = [(¹⁴³Nd/¹⁴⁴Nd)_sample(t) / (¹⁴³Nd/¹⁴⁴Nd)_CHUR(t) − 1] × 10⁴

A sample with ε_Nd = 0 has the same ¹⁴³Nd/¹⁴⁴Nd as CHUR at the time in question, implying that it was derived from a source with a chondritic (undepleted) Sm/Nd ratio. A positive ε_Nd indicates that the source had a higher-than-chondritic Sm/Nd ratio and therefore evolved to a higher ¹⁴³Nd/¹⁴⁴Nd over time — characteristic of the depleted mantle, from which incompatible elements (including the light rare earth elements, which are enriched in neodymium relative to samarium) have been preferentially extracted by partial melting. A negative ε_Nd indicates a lower-than-chondritic Sm/Nd ratio, characteristic of enriched sources such as old continental crust, which is the complementary reservoir to the depleted mantle.^{3, 18}

DePaolo and Wasserburg (1976) demonstrated that the majority of young mid-ocean ridge basalts (MORB) have ε_Nd values of approximately +7 to +12, confirming that the upper mantle source of MORB has been progressively depleted of its incompatible elements over Earth's history by the extraction of continental crust.³ Ocean island basalts (OIB) typically show lower, but still often positive, ε_Nd values, reflecting derivation from mantle sources that are less depleted or that contain recycled crustal components.¹⁸ Ancient continental crust, by contrast, characteristically shows strongly negative ε_Nd values, reflecting its long isolation from the convecting mantle and the progressive ingrowth of radiogenic ¹⁴³Nd from its low Sm/Nd ratio.

Typical ε_Nd values for major geological reservoirs^{3, 18}

Depleted mantle model ages

A particularly valuable application of Sm-Nd isotope systematics is the calculation of model ages — estimates of the time at which a rock's neodymium isotopic composition diverged from a modelled mantle reservoir. The simplest form is the T_CHUR model age, which calculates the time in the past when the sample's ¹⁴³Nd/¹⁴⁴Nd ratio was identical to the CHUR evolution line. This age represents the time at which the sample (or its source) was separated from an undepleted, chondritic reservoir.^{1, 9}

DePaolo (1981) recognised that T_CHUR model ages systematically underestimate the true crustal formation age, because the mantle source from which continental crust is extracted is not undepleted but has already been partially depleted by prior episodes of crust extraction. He proposed the depleted mantle model age (T_DM), which instead calculates the intersection of the sample's neodymium isotopic evolution with a depleted mantle evolution curve that tracks the progressive depletion of the upper mantle over Earth's history.⁵ The depleted mantle curve was empirically calibrated using the ε_Nd values of juvenile crustal rocks of known age — rocks demonstrably derived from the mantle at the time indicated by their U-Pb or Sm-Nd crystallisation ages.⁵

T_DM model ages are widely interpreted as crustal residence ages — the average time since the neodymium in a sample was last in isotopic equilibrium with the convecting mantle. For juvenile igneous rocks derived directly from mantle melting, the T_DM age approximates the true time of crust formation. For sedimentary rocks and their metamorphic derivatives, the T_DM age provides an average age of the crustal sources that contributed to the sediment, weighted by their neodymium concentrations.^{1, 8} McCulloch and Wasserburg (1978) pioneered this approach by using Sm-Nd and Rb-Sr systematics to determine the times at which major segments of the Canadian Shield were extracted from the mantle, demonstrating that the Superior, Slave, and Churchill provinces were all formed within the period 2.5 to 2.7 billion years ago.⁶

It is important to note that T_DM model ages are model-dependent and must be interpreted with care. They assume a single-stage evolution from mantle to crust, whereas real crustal rocks may have complex multi-stage histories involving mixing of sources with different ages. Sedimentary rocks, in particular, represent physical mixtures of detritus from multiple crustal sources, and their T_DM ages are weighted averages that do not correspond to any single geological event.^{1, 8} Despite these caveats, T_DM model ages remain one of the most widely used tools for mapping the age of the continental crust on a regional and global scale.

Applications to mafic rocks and mantle geochemistry

The Sm-Nd system finds its most distinctive applications in rock types where other radiometric methods are difficult or impossible to apply. Mafic and ultramafic rocks — basalts, gabbros, peridotites, and komatiites — are rich in magnesium and iron but typically contain very little rubidium, potassium, or uranium, making the Rb-Sr, K-Ar, and U-Pb systems problematic or inapplicable. These same rocks, however, contain appreciable concentrations of the rare earth elements, including samarium and neodymium, in minerals such as clinopyroxene, garnet, and orthopyroxene, making Sm-Nd the method of choice for dating mantle-derived lithologies.^{1, 9}

DePaolo and Wasserburg (1979) demonstrated the power of this approach by determining a Sm-Nd mineral isochron age of 2,701 ± 8 million years for the Stillwater Complex, a large layered mafic-ultramafic intrusion in Montana. This age, obtained from garnet, clinopyroxene, and plagioclase separates, was in excellent agreement with independent U-Pb zircon ages and established the Sm-Nd method as a viable high-precision chronometer for mafic rocks.¹⁹ The initial ε_Nd value of the Stillwater Complex, calculated from the isochron intercept, showed that the magma was derived from a depleted mantle source, providing one of the earliest demonstrations that the upper mantle had already undergone significant depletion by the late Archean.¹⁹

In mantle geochemistry, the correlation between ε_Nd and ⁸⁷Sr/⁸⁶Sr (the "mantle array") has been one of the most important discoveries of isotope geochemistry. DePaolo and Wasserburg (1979) demonstrated that oceanic basalts define a negative correlation between these two isotopic tracers: rocks with high ε_Nd (derived from depleted sources) tend to have low ⁸⁷Sr/⁸⁶Sr, and vice versa.¹⁸ This complementary behaviour arises because the processes of partial melting and melt extraction that deplete the mantle in light rare earth elements (lowering Sm/Nd and thus ¹⁴³Nd/¹⁴⁴Nd in the extracted crust) simultaneously deplete it in rubidium relative to strontium (lowering Rb/Sr and thus ⁸⁷Sr/⁸⁶Sr in the residual mantle). The Nd-Sr mantle array provides a powerful framework for identifying and characterising distinct mantle source reservoirs, including the depleted MORB mantle, enriched mantle domains (EM-I and EM-II), and the high-μ (HIMU) component associated with recycled oceanic crust.^{9, 18}

Goldstein, O'Nions, and Hamilton (1984) extended the application of Sm-Nd isotopes to sedimentary provenance studies by analysing atmospheric dusts and river-borne particulates from major drainage systems worldwide. They demonstrated that the ε_Nd values and T_DM model ages of fine-grained sediment accurately reflect the average crustal age of the source terranes being eroded, establishing Sm-Nd isotopes as a quantitative tracer for sediment provenance and continental weathering patterns.⁸

Complementarity with the Rb-Sr system

The Sm-Nd and Rb-Sr isotopic systems are most powerful when used in combination, because they respond differently to the same geological processes and together provide a more complete picture than either system alone. The contrasting geochemical behaviour of their parent-daughter pairs — the chemical similarity of Sm and Nd versus the dissimilarity of Rb and Sr — means that metamorphism and fluid interaction will often disturb the Rb-Sr system while leaving the Sm-Nd system intact, or vice versa.^{1, 6}

When both systems yield the same age for a suite of samples, the concordance provides strong confirmation that the age is geologically meaningful and that both systems have remained closed since the dated event. When the two systems disagree, the pattern of discordance provides diagnostic information about the thermal and fluid history of the rocks. A Sm-Nd whole-rock isochron age older than the corresponding Rb-Sr whole-rock age, for example, typically indicates that the Rb-Sr system was partially reset by a metamorphic event that did not disturb the Sm-Nd systematics, and the Sm-Nd age preserves the original igneous crystallisation age while the Rb-Sr age records a younger metamorphic overprint.^{1, 9}

The correlation between ε_Nd and initial ⁸⁷Sr/⁸⁶Sr has been particularly valuable for understanding the evolution of Earth's oldest crustal rocks. Hamilton and colleagues (1978) used combined Sm-Nd and Rb-Sr data from the Isua supracrustal belt of West Greenland — one of the oldest known rock sequences on Earth, at approximately 3.7 to 3.8 billion years — to demonstrate that these rocks were derived from a mantle source that had already been depleted by prior crust extraction, providing early evidence that mantle differentiation was well underway by the early Archean.⁷ McCulloch and Wasserburg (1978) subsequently used coupled Sm-Nd and Rb-Sr analyses of composite samples from major structural provinces of the Canadian Shield to map the spatial pattern of crustal formation ages across North America, revealing distinct episodes of continental growth in the Archean and Proterozoic.⁶

The extinct ¹⁴⁶Sm-¹⁴²Nd chronometer

In addition to the long-lived ¹⁴⁷Sm-¹⁴³Nd system, a short-lived samarium isotope — ¹⁴⁶Sm, with a half-life of approximately 103 million years — decays by alpha emission to ¹⁴²Nd.^{16, 17} Because this half-life is short relative to the age of the Solar System, all primordial ¹⁴⁶Sm has long since decayed, making it an extinct radioactivity — a chronometer that was active only during the first few hundred million years of Solar System history. Variations in the ¹⁴²Nd/¹⁴⁴Nd ratio among terrestrial and meteoritic samples are therefore "fossil" signatures of Sm/Nd fractionation events that occurred in the earliest stages of planetary differentiation, before ¹⁴⁶Sm decayed away.^{11, 16}

Harper and Jacobsen (1992) provided early evidence from coupled ¹⁴⁷Sm-¹⁴³Nd and ¹⁴⁶Sm-¹⁴²Nd systematics that the Earth's mantle had undergone significant chemical differentiation within the first 100 to 200 million years of Solar System formation, with Sm/Nd fractionation occurring early enough for the short-lived ¹⁴⁶Sm to imprint detectable ¹⁴²Nd anomalies on the resulting reservoirs.¹⁶ Bennett, Brandon, and Nutman (2007) confirmed and extended this finding by measuring coupled ¹⁴²Nd and ¹⁴³Nd excesses in 3.85-billion-year-old rocks from the Isua supracrustal belt, demonstrating that these rocks sampled mantle material that had been depleted within the first 30 to 75 million years of Earth's formation.¹³

In 2005, Boyet and Carlson reported a landmark discovery: high-precision measurements showed that the ¹⁴²Nd/¹⁴⁴Nd ratio of virtually all accessible terrestrial rocks is approximately 20 parts per million higher than that of chondritic meteorites.¹¹ This offset implies that the accessible silicate Earth does not have a truly chondritic Sm/Nd ratio, but instead has a slightly superchondritic composition — as if an early enriched reservoir (with low Sm/Nd and correspondingly low ¹⁴²Nd/¹⁴⁴Nd) had been segregated and hidden from subsequent mantle convection within the first 30 million years of Earth's history. Boyet and Carlson proposed that this hidden enriched reservoir resides at the base of the mantle, isolated from the convecting upper mantle for over 4.5 billion years.¹¹ This hypothesis has profound implications for estimates of the bulk silicate Earth composition and for models of early planetary differentiation, and has stimulated extensive debate in the geochemical community.¹⁴

O'Neil, Carlson, Francis, and Stevenson (2008) applied the ¹⁴⁶Sm-¹⁴²Nd chronometer to mafic amphibolites from the Nuvvuagittuq greenstone belt in northern Quebec, Canada. They found that ¹⁴²Nd/¹⁴⁴Nd ratios in these rocks correlate positively with Sm/Nd ratios, producing a ¹⁴⁶Sm-¹⁴²Nd isochron age of 4,280 million years — potentially making them the oldest preserved crustal rocks on Earth, predating even the Acasta Gneiss by more than 200 million years.¹² Although this interpretation has been debated, the Nuvvuagittuq result illustrates the unique capability of the extinct ¹⁴⁶Sm-¹⁴²Nd system to resolve differentiation events in the Hadean eon that are inaccessible to any extant long-lived chronometer.

Crustal evolution and continental growth

One of the most consequential applications of Sm-Nd isotope systematics has been the reconstruction of the growth history of the continental crust over Earth's 4.5-billion-year history. Because the extraction of continental crust from the mantle by partial melting systematically fractionates Sm from Nd (enriching the crust in Nd relative to Sm), the ε_Nd values and T_DM model ages of crustal rocks record the timing and volume of crust-mantle differentiation through geological time.^{5, 6}

Regional compilations of T_DM model ages have been used to map the age structure of the continental crust on every continent. These maps reveal that the continents are assembled from crustal blocks of distinctly different ages, separated by orogenic belts where younger material was added. In North America, for example, T_DM model ages define a broadly concentric pattern centred on the Archean cratons of the Superior and Slave provinces (T_DM > 2.5 Ga), surrounded by progressively younger Proterozoic terranes (T_DM 1.8 to 2.0 Ga) and Phanerozoic accreted material along the continental margins.^{5, 6}

The question of whether continental crust has grown progressively over time or was largely generated in a few major pulses remains debated, but T_DM model age compilations from sedimentary rocks worldwide suggest that significant volumes of continental crust were in place by the late Archean, with at least 60 to 70 percent of the present crustal volume extracted from the mantle before 2.5 billion years ago.^{1, 8} The complementary depletion of the mantle is recorded in the ε_Nd evolution curve: the average ε_Nd of juvenile crustal rocks increases from near-chondritic values in the early Archean to +8 to +12 for modern MORB, tracking the progressive enrichment of the mantle in ¹⁴³Nd as ¹⁴⁷Sm in the depleted residue decays over geological time.^{3, 5}

The debate over whether the Earth's bulk composition is truly chondritic, sparked by Boyet and Carlson's (2005) discovery of a terrestrial ¹⁴²Nd excess relative to chondrites, has important consequences for crustal growth models. If the accessible Earth has a superchondritic Sm/Nd ratio, then T_DM model ages calculated using a chondritic reference may be systematically too old, and the volume of ancient crust may be overestimated.^{11, 14} Caro and Bourdon (2010) explored the implications of a non-chondritic bulk Earth for the coupled evolution of the mantle-crust system, demonstrating that the choice of reference composition significantly affects estimates of the rate and timing of continental growth.¹⁴

Limitations and sources of uncertainty

Despite its unique strengths, the Sm-Nd method has several inherent limitations that restrict its precision and scope of application. The most fundamental is the small degree of Sm/Nd fractionation during geological processes. Because samarium and neodymium are geochemically so similar, the Sm/Nd ratio varies by only about a factor of two across common rock types, compared with orders-of-magnitude variation in the U/Pb or Rb/Sr ratios. This limited spread in the parent-daughter ratio means that the isochron slope — and hence the age — is determined with lower precision than in systems with greater elemental fractionation.^{1, 9}

The extremely long half-life of ¹⁴⁷Sm, while advantageous for dating ancient rocks, means that the accumulation of radiogenic ¹⁴³Nd is very slow. For rocks younger than about 100 to 200 million years, the amount of radiogenic ¹⁴³Nd produced is too small to measure with sufficient precision to yield a useful age, effectively restricting the method to the Mesozoic and older.^{1, 15} This lower age limit is substantially higher than for the U-Pb, Ar-Ar, or Rb-Sr methods, all of which can date Phanerozoic and even Quaternary materials.

Comparison of Sm-Nd with other radiometric dating systems^{1, 9, 15}

Additional uncertainties arise from the decay constant itself. Although the Lugmair and Marti (1978) value of 6.54 × 10^-12 per year has been the standard for nearly half a century, some more recent direct measurements have yielded half-lives ranging from approximately 106 to 117 billion years, a spread of more than 10 percent that is considerably larger than the uncertainties quoted on individual measurements.¹⁷ The IUPAC-IUGS Subcommission has recommended continued use of the 106-billion-year value pending resolution of this discrepancy, but the uncertainty in the decay constant remains a systematic limitation on the accuracy (as distinct from the precision) of Sm-Nd ages.¹⁷

The interpretation of T_DM model ages is subject to additional caveats. These ages are calculated under the assumption of a single-stage evolution from mantle to crust, and they are sensitive to the choice of depleted mantle evolution curve and the reference composition (chondritic or superchondritic) used for the calculation. Rocks that have experienced mixing of sources with different crustal residence times, or that have had their Sm/Nd ratios modified by secondary processes such as weathering of accessory minerals, will yield T_DM ages that do not correspond to any real geological event.^{1, 14} For these reasons, T_DM model ages are most reliably interpreted in combination with independent age constraints from U-Pb or other methods.

Key discoveries and continuing significance

The Sm-Nd isotope system has contributed to several landmark discoveries in Earth science. The pioneering work of DePaolo and Wasserburg in 1976 establishing the epsilon notation and the chondritic reference framework fundamentally changed how geochemists think about mantle-crust evolution, providing a quantitative language for describing the degree of depletion or enrichment of any geological sample relative to the primordial bulk Earth.³ Their subsequent demonstration of the Nd-Sr mantle array (1979) revealed the existence of distinct geochemical reservoirs in the mantle and laid the foundation for modern mantle geochemistry.¹⁸

The dating of the Isua supracrustal belt by Hamilton and colleagues (1978) provided one of the earliest Sm-Nd constraints on the antiquity of Earth's crust, demonstrating that rocks approaching 3.8 billion years in age preserved primary mantle signatures indicating early depletion.⁷ McCulloch and Wasserburg's (1978) systematic application of coupled Sm-Nd and Rb-Sr analyses to major crustal provinces demonstrated the power of isotope geochemistry for mapping the age structure of continents, an approach that has since been applied worldwide.⁶

The discovery by Boyet and Carlson (2005) that the accessible terrestrial silicate Earth has a ¹⁴²Nd/¹⁴⁴Nd ratio approximately 20 parts per million higher than chondritic meteorites has stimulated one of the most active debates in modern geochemistry. If confirmed as a result of early planetary differentiation rather than nucleosynthetic heterogeneity in the solar nebula, this finding implies that a substantial enriched reservoir — possibly constituting 5 to 30 percent of the silicate Earth — has remained hidden and isolated at the base of the mantle since the Hadean eon, with far-reaching implications for models of mantle convection, the bulk composition of the Earth, and the thermal evolution of the planet.^{11, 14}

The Sm-Nd method continues to evolve. Improvements in mass spectrometric precision, particularly through the development of multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) and thermal ionisation mass spectrometry (TIMS) with higher sensitivity, have pushed the resolution of ¹⁴²Nd measurements to the parts-per-million level, enabling detection of differentiation events within the first tens of millions of years of Solar System history.^{12, 13} The O'Neil and colleagues (2008) determination of a ¹⁴⁶Sm-¹⁴²Nd age of 4.28 billion years for the Nuvvuagittuq rocks represents one of the most dramatic applications of this capability, potentially identifying the oldest surviving fragment of Earth's primordial crust.¹² As analytical techniques continue to improve and new samples from ancient terranes are investigated, the samarium-neodymium system — in both its extant and extinct forms — will remain an essential tool for understanding the earliest and most fundamental chapters of Earth's geological history.