- 1.9k Downloads
KeywordsAccessory Mineral Detrital Zircon Geological Time Scale Isotope Dilution Method Detrital Mineral
Uranium–Lead dating is the geological age-determination method that uses the radioactive decay of uranium (U) isotopes (238U, 235U, and also in this entry 232Th) into stable isotopes of lead (Pb) (206Pb, 207Pb, and 208Pb, respectively). U–Pb geochronology is the science of both the methodology but also the application of these methods to geological problems.
U–Pb Decay System and Age Calculations
There are a number of intermediate isotopes in the decay chain that must first decay before stable Pb isotopes are produced. The most important ones for U–Pb dating are 234U and 230Th in the 238U decay system and 231Pa in the 235U system, all of which have half-lives between 30,000 and 250,000 years. These intermediate isotopes produce complications to the interpretation of U–Pb dates that can be very important for young (i.e., Neogene) ages and particularly minerals with high Th/U such as monazite or allanite. Also there are a few cases where the effects are very dramatic and which may involve 231Pa excesses in the 235U–207Pb decay chain (Oberli et al. 2004; Anczkiewicz et al. 2001).
When the measurements are made, dates are calculated by measuring a daughter/parent ratio (206Pb/238U, 207Pb/235U, 208Pb/232Th) and a radiogenic lead isotope ratio (207Pb/206Pb), resulting in up to four calculated dates, three of which are part of the U–Th–Pb system of decay.
In the U–Pb system, a calculated date using one of the equations above does not necessarily result in the most valid interpreted age of a mineral, partly because some minerals have a tendency to lose Pb as a result of volume diffusion and/or diffusion via radiation damage and partly because some accessory minerals can have multiple age components during their formation (e.g., cores overgrown by younger rims). These combined effects are often referred to as “open-system” behavior, referring to cases where a date does not simply reflect a single growth domain where the parent and daughter isotope ratio has changed only through radioactive decay. Age patterns within single grains of zircon and monazite, in particular, can be quite complicated because of these factors and require careful analysis in the petrographic context, especially in metamorphic rocks (Cottle et al. 2009).
A method of visually displaying measured isotope ratios to better interpret potential complexities resulting from open-system behavior can be attributed to Wetherill (1956), who devised the Concordia diagram. This diagram graphically interrelates all three ratios of the U–Pb system whereby the X and Y axes are 207Pb/235U and 206Pb/238U, and the inverse gradient of any point drawn through the origin is proportional to 207Pb/206Pb. In this diagram, a curve originating at the origin, termed the “concordia curve,” defines the 207Pb/235U and 206Pb/238U values where the calculated dates of these ratios are identical, using the age equations above. Any measured analysis of a mineral that plots within uncertainty of the Concordia curve is considered “concordant,” having its 207Pb/235U and 206Pb/238U calculated dates the same within uncertainty. If falling off this curve, either above or below, an analysis is termed “discordant,” with measured dates not in agreement.
A variant of the Concordia diagram, termed the Tera-Wasserburg diagram (Tera and Wasserburg 1972), plots 207Pb/206Pb as Y and 238U/ 206Pb as X and is particularly useful when measurement of the small 204Pb isotope is imprecise or not possible. Both of these diagrams offer a graphical method of making a correction for disturbances to the U–Pb system arising, for example, from a single moment of Pb loss or a second phase of mineral growth well after a mineral’s formation, in order to calculate the original time of mineral growth. Further variations of the Concordia diagram involve 208Pb/232Th plotted against 207Pb/235U or 206Pb/238U to illustrate behavior involving U, Th, and Pb isotopes. The use of two parent U isotopes and two daughter Pb isotopes is a very useful property; it allows a correction for disturbed systems and recovery of the initial age of a mineral, a feature unique to the U–Pb dating method among all isotope geochronometers.
A correction for initial or “common” lead, that is, the lead incorporated into a material at its time of formation, is implicit in calculation of ages. The theory of this is well-developed in many geochemistry or geochronology textbooks and is not detailed herein.
Historical Development of the Method
The discovery of radioactivity of uranium was a pivotal event, with recognition that transformation of uranium into other elements was inevitable. Early on, lead was suspected of being the ultimate stable end product of this decay (Boltwood 1907). Even before the discovery of isotopes by Frederick Soddy in 1913 (Soddy 1913), U–Pb ages of geological materials were calculated, using initial estimates of the half-life or decay rate of uranium. Although Boltwood (1907) was the first to calculate the age of a mineral based on its U–Pb ratio, Arthur Holmes (Holmes 1911), under the guidance of his supervisor R J Strutt, determined dates of a suite of igneous-related Devonian minerals of variable U–Pb ratio in Norway by measuring the atomic U–Pb ratio in minerals with improved methods. This study used a combination of appropriate geological samples with clever methods of classical wet chemistry and radiochemistry. The dates calculated, approximately 370 Ma, were within 5 % of the correct ages determined later. Holmes also recalculated Boltwood’s dates with a better decay constant, having sought information on the geological context of Boltwood’s samples from the US Geological Survey. These were of real importance because this began the first numerical calibration of the geological timescale as well as proving that the Earth was much older than 1 Ga.
The discovery of isotopes by Soddy had little impact initially to U–Pb dating because the development of mass spectrometry capable of measuring U or Pb isotopes was only achieved decades later. This changed dramatically in ∼1935 with advances in technology. Alfred Nier, among the most brilliant of all mass spectrometrists of the twentieth century, designed instruments and ion signal amplification devices that could measure accurately and relatively precisely the isotope composition of U and Pb; indeed, his measurement of U isotopes were pivotal in demonstrating that 235U was the isotope capable of fission by thermal neutron irradiation (Nier 1939). His measurements of Pb isotopes allowed U–Pb isotope ages of U ore minerals to be calculated and a 207Pb/206Pb age of 2.6 Ga to be determined on monazite (Nier et al. 1941). Some years later, Claire Patterson (1956) published accurate and precise measurements of primordial lead in troilite (an iron sulfide with negligible U) from the Canyon Diablo (Meteor Crater, Arizona) Ni–Fe meteorite that allowed an age of ~4.55 Ga for the age of early solar system meteorites to be established.
One of the most important collateral outcomes of atomic energy research was the accurate and precise method of isotope quantification known as isotope dilution. In this method a known quantity of an enriched isotope of an element (i.e., 235U, 208Pb) is added to a sample prior to dissolution, and the concentration of natural U and Pb isotopes in the sample are calculated using the measurement by mass spectrometry of the isotope composition of the resultant mixture. This method is an extension of the original method using radioactive tracers invented by George de Hevesy in 1913, for which he was awarded the Nobel Prize in 1943.
This procedure is simple and superbly reliable. To illustrate how this works, imagine a bucket of unknown quantity of colored balls that are black and white in a ratio of 2:1 black–white (representing the natural isotopes in a sample). If one adds 100 red balls (representing the enriched tracer) and mixes the balls up and then measures a ratio of red–black of 0.5 (by isotope ratio mass spectrometry of a portion of the sample), then one can easily work out that there are 50 black and 25 white balls in the original sample. One does not have to count all the balls to know this; one only needs to know the number of tracer balls added and the ratio (i.e., isotope ratio) of the mixture.
This methodology was developed and applied to a wide range of elements and sample materials. It still represents the most accurate and precise manner of isotope quantification in geological samples. One of the finest early papers to date accessory minerals by these methods is the tour de force of Tilton et al. (1955) in which various minerals were dated in a Precambrian granite; the paper presents tedious but elegant analytical methods, including not just U–Pb but also Th–Pb, and is a required reading for all U–Pb geochronologists who wish to appreciate the fundamental work of the early pioneers of the methods we now use so routinely. This was the standard method of U–Pb geochronology until the mid-1990s and is still widely used today.
Evolution of Methods 1955–1980
In the isotope dilution method of U–Pb dating, decomposition of samples into a homogeneous solution is a requirement so that tracer and sample isotopes can fully mix prior to chemical separation and isotope analysis by mass spectrometry. During the period prior to ~1970, dissolution of various silicate minerals useful for U–Pb geochronology (i.e., zircon) was tractable only with flux melting followed by HCl acid dissolution, and it required a large quantity of sample material. This was a hindrance to widespread application of the U–Pb dating method partly because of the time and expense involved in the procedure but also because zircon commonly loses Pb due to radiation damage of its crystal lattice. When large quantities of zircon were analyzed by this flux method, this inevitably resulted in many discordant dates being produced because of the prevalence of Pb loss in some of the grains; analysis of tiny amounts of high quality zircon was simply not generally possible. This frustrating reality proved to be a strong motivation to find a better way – to be able to analyze much smaller quantities of zircon and ultimately single grains that were of very high quality not subject to significant Pb loss.
In the late 1960s and early 1970s, two young geologists – Tom Krogh and James Mattinson – devised a solution to this conundrum while doing work early in their careers at the Carnegie Institute of Washington. Using the availability of Teflon™, which fortuitously had been discovered by Roy Plunkett of DuPont Corporation in 1938 during refrigerant research, Krogh and Mattinson made several key breakthroughs in geochemical research: (1) the ability to dissolve refractory minerals like zircon in HF acid at high temperatures (~220 °C) in steel-jacketed Teflon™ pressure vessels, (2) the ability to clean Teflon™ vessels so as to introduce very little additional lead contamination, and (3) the development of sub-boiling distillation of acids in Teflon™ to produce ultrapure acids with low lead contamination. Together with the silica gel Pb ionization method on a re-thermal ionization filament, it became possible to analyze single grains of zircon or other minerals; a practice that is routine today. Despite these improvements, the issue of Pb loss in zircon still hindered high-precision age determination. Krogh (1982) addressed the problem of zircon discordance by inventing the air abrasion technique to remove the discordant outer part of zircon crystals, which resulted in production of vastly improved concordance (i.e., closed-system behavior) of zircons. In 1985–1987, R Parrish worked with Tom Krogh and J C Roddick (Parrish and Krogh 1987; Roddick et al. 1987) to synthesize a significant quantity of the artificial isotope 205Pb for distribution to world laboratories and devised an efficient method to dissolve multiple individual zircon samples in a single steel-jacketed pressure vessel. All of these analytical advances conspired to facilitate a very significant growth in the number of U–Pb laboratories and the quantity of high-accuracy, high-precision zircon dates. Using U–Pb zircon dating with ≤ ±4 Ma uncertainty, Precambrian volcanic chronostratigraphy became a commonplace in the 1980s–1990s, especially in Canada. A great advance in Precambrian continental orogenic evolution then proceeded, especially in the Canadian Shield. This made possible seminal papers on continental evolution, such as Paul Hoffman’s United Plates of America synthesis (Hoffman 1988).
At the same time, when these methods were applied to complicated igneous and metamorphic zircon, the complexity in zircon U–Pb systematics became apparent, including the phenomenon of zircon inheritance and multiple metamorphic zircon growth periods during high-grade metamorphism. In such cases, single-grain analysis produced consistently discordant data, and the determination of precise ages of metamorphism or igneous events proved to be a challenge in many samples. Complex zircon growth and alteration were imaged using electron microscopy, suggesting that the complexities in U–Pb data were the result of a combination of lead loss and multiple zircon age domains within single grains. The paper by Corfu et al. (2003) illustrates the variation in textures elegantly. The analysis of zircon on the micron scale became an obvious tantalizing goal, but it presented major technical challenges and was impossible at the time for multi- or single grain zircon analysis by ID-TIMS. One way forward using ID-TIMS came from accessory minerals like monazite with less inheritance and Pb loss to determine igneous and metamorphic ages (Parrish 1990).
With more work applied by ID-TIMS methods to alternative minerals, the effects of chemical fractionation of the intermediate daughter isotopes 230Th and 231Pa relative to 238U and 234U became better documented (Schärer 1984; Parrish 1990; Anczkiewicz et al. 2001) As a result of this effort on alternative minerals, excess 206Pb in monazite and other Th-rich minerals, sometimes of huge relative magnitude, was documented, and smaller magnitude 206Pb deficiencies and much more rare 207Pb excesses (Anczkiewicz et al. 2001) were documented in zircon; all of these features were shown to exist by early work of Mattinson (1973).
Notwithstanding the advances made by laboratories using the ID-TIMS method, efforts to develop in situ methods of U–Pb analysis were accelerating, building on advances in secondary ion mass spectrometry (SIMS).
In Situ Methods of U–Pb Analysis
In this article, in situ U–Pb dating refers to the measurement of age of a mineral in a solid state using microbeam sampling techniques, and it includes first SIMS and later LA-ICP-MS methods, as well as the less-used electron microprobe chemical dating method, all of which are described briefly below.
Among other scientists involved in SIMS (secondary ion mass spectrometry) techniques, William Compston of the Australian National University (ANU) led a team to develop SIMS methods for the analysis of very small (<50 μm) regions of zircon for the purpose of U–Pb dating. The SIMS method consists of firing a primary highly focused ion beam (oxygen or cesium ions) onto the surface of polished zircon in order to produce secondary sample ions which are then accelerated and focused into a beam which could be analyzed by a mass spectrometer. However, in SIMS zircon analysis, a wide range of molecular ions are produced during the primary sputtering process where the beam interacts with the sample. Many of these molecular ions have very similar masses in relation to the Pb isotopes that are of interest for dating, and which occur in very small quantities. In order to cleanly separate and measure the molecular ions from the Pb ions of interest, larger radius (>1 m) double-focusing high mass resolution SIMS instruments had to be designed and built. In the U–Pb trade, these methods have been loosely termed “ion microprobe” methods, and the ANU instrument was termed “SHRIMP” (Sensitive High Resolution Ion Microprobe) and was built in the early 1980s; an analogous instrument was built in France by Cameca Instruments in the early 1990s. In 1984, the first SIMS zircon age was published by Compston et al. (1984), and it determined an age of 4.36 Ga for zircons in lunar breccia sample 73,217 returned from Apollo 17. This development heralded a new period of exploration of zircon as a geochronometer, although the application of the SIMS zircon work to terrestrial rocks did not see wide application until the 1990s.
SIMS U–Pb methods cannot quantify accurately the ratio of Pb to U isotopes within a zircon or other accessory mineral directly. Instead calculation of a date relies on the analysis of a “standard” or “reference” zircon whose U–Pb isotope ratios must have been determined by the ID-TIMS method. Ion signals of unknowns are measured and compared to those of the reference zircons (the Pb–U ratio which is known) to establish a calibration relationship; this in turn is used to calculate the 206Pb/238U ratio of the unknown, then used for age determination. A series of zircon reference samples has evolved over the past ~30 years, with variable ages, U contents, homogeneity, and extent of concordance. The search for, and characterization of, suitable reference materials for isotopic and elemental analysis is an ongoing effort for in situ analysis, and not just for zircon.
Building on earlier work with quadrupole mass spectrometry and ionization using plasma sources, inductively coupled plasma ionization (ICP) was combined in the early 1990s with multiple-collector mass spectrometry (MC-MS) using an additional electrostatic analyzer to make the first ICP multi-collector mass spectrometer (ICP-MC-MS). This instrument combined the highly efficient ICP ion source with multiple collections of ion signals in a mass spectrometer of mass resolution similar to a normal TIMS instrument (mass resolution ~400). Shortly thereafter, laser ablation (LA) micro-sampling was attached to this type of instrument to deliver the first LA-ICP-MC-MS, with the measurement of U–Pb ages using a zircon reference sample. One of the more comprehensive early papers using this method was that of Horstwood et al. (2003) which also included a common Pb correction. Among other applications, this allowed hafnium (Hf) isotopes to be measured more time efficiently than by TIMS methods, and this was exploited by ID-TIMS zircon geochronologists to add Hf model ages to the zircon U–Pb evolution (Davis et al. 2005). These were the precursors to more widespread LA-ICP-MS instruments on the market now, which offer sufficient sensitivity and precision to compete directly with SIMS in situ analysis, but at less cost and with greater sample throughput, though with higher sample consumption. Both SIMS and LA-ICP-MS (including LA-ICP-MC-MS) instruments are dependent on one or more reference samples for date calculation.
SIMS U–Pb methods are very sensitive to surface polish quality and surface-charging characteristics and require a careful matching of the composition (matrix) of sample to the reference material. LA-ICP-MS methods, on the other hand, are more tolerant of surface irregularities, are less sensitive than SIMS methods to sample matrix matching, consume a larger quantity of sample, and have limited ability to measure the common Pb reference isotope 204Pb due to the interference with 204Hg in the carrier gas. Careful work using the most sensitive LA-ICP-MS methods, however, can deliver the same age accuracy and precision as SIMS methods with consumption of sample only a factor of 2–4× greater, with laser ablation pits as small as a few μm deep and ~20 μm in diameter. Depth profiling by both methods is possible but is more controlled via SIMS analysis. As of the time of writing, international efforts are underway to improve interlaboratory inter-comparability of U–Pb data, but a fundamental limitation of both of these in situ methods is that accuracy and precision of the 206Pb/238U ratio in a sample appears limited to ~1–2 % 2σ on weighted mean dates, which places a fundamental limit on the ultimate age precision measured by means of the 206Pb/238U ratio. The exact reason for this limitation is likely some combination of plasma chemical fractionation and laser ablation processes, perhaps compounded by matrix matching issues.
Unfortunately, a misunderstanding of the strengths and weaknesses of both isotope dilution methods and in situ SIMS or LA-ICP-MS methods traditionally led to tension between the two methodological communities. For example, SIMS dates claiming a precision of <1 % of zircon age were published in numerous studies, and some were used to attempt a refinement of the geological time scale; several of these have been refuted and subsequently refined by careful isotope dilution methods. On the other hand, many papers contain published isotope dilution U–Pb zircon data that are discordant and yield non-unique and ambiguous interpretations of age on complex multiple growth-zoned zircons. More often than not, these complex and ambiguous interpretations are from metamorphosed rocks where multiple high-temperature events have produced complicated zircons. The primary lesson from these examples is that both methods have their strengths, and when used appropriately, rocks and events can be dated in the best possible way using one or both methods. All of these methods have their place in U–Pb geochronology.
Two further methodological aspects deserve mention. Electron microprobe U–Th–Pb analysis of monazite has been done since 1994 (Suzuki et al. 1994). This “chemical” (as opposed to isotope) method of dating was refined by Montel et al. (1996) and Williams and Jercinovic (2002) primarily in Precambrian rocks where monazite contains relatively large quantities of radiogenic Pb, sufficient to overwhelm any initial common lead, while preserving the petrographic context. Monazite U–Th–Pb isotope dating in thin sections is the only method able to reveal the ages of very young monazites of Mesozoic-Neogene age, however. Together, EMP, SIMS, and LA-ICP-MS methods offer significant insight into the P-T-t (pressure–temperature–time) conditions and paths of metamorphic rocks of all ages from Archaean to Neogene.
Two More Innovations to Remember
A further refinement was developed over many years by James Mattinson in a procedure loosely termed “chemical abrasion” or “CA-TIMS” (Mattinson 2005). Effectively this is a recipe involving high-temperature (~900 °C) annealing of zircon, followed by a HF–HNO3 acid partial dissolution of annealed grains, which effectively removes the discordant regions (zones, cracks, etc.) of a zircon grain, leaving undissolved zircon that has been a closed system since formation. When the residual zircon is analyzed, concordant analyses often result, allowing very precise concordant ages to be determined. This is such a profoundly important procedure that it has been adopted worldwide as the routine zircon procedure for ID-TIMS methods.
Finally, any U–Pb geochronologist will understand clearly the value in accurate and sophisticated data reduction, linear regression, and graphical plotting software to diversely illustrate and determine uncertainties in ages of U–Pb data, the latter being its most important attribute. The most commonly used software to plot data, termed Isoplot, was conceived, written and refined over several decades primarily by Ken Ludwig of the Berkeley Geochronology Center (Ludwig 1991). It is one of the most important underpinning tools that U–Pb geochronologists use.
Geologists and geochronologists, indeed humans in general, have a tendency to take for granted the technological advances that make our lives easier; it is important to remember the intellectual breakthroughs as well as the sheer effort of scientists who pioneered these advances. These breakthroughs have made the science of U–Pb geochronology as apparently routine as we know it today.
U–Th–Pb Geochronology: Applications to Earth, Planetary, and Environmental Science
The age of meteorites, the Moon, Earth, and the Solar System
The age of igneous events throughout Earth history, plutonic and volcanic, mafic, alkaline, and felsic
Earth differentiation and subsequent crustal evolution
Age of the oldest continental crust and Earth materials
The behavior of intermediate daughter radioactive isotopes and their impact on high-precision dating
Rates of production of continental crust
Continental reconstructions over time involving amalgamation and dispersal of continents
Rates of tectonic events in arcs, collisional orogens, etc.
Age of plumes and large igneous provinces and their global impacts
Quantification of time in pressure–temperature paths of metamorphic rocks (P–T–t paths)
Chronostratigraphy, in particular in Precambrian where biostratigraphy is absent
Calibration of the geological time scale
Rates of cooling of orogens and rocks by using U–Pb dates as thermochronometers in the context of closure temperatures
Identifying the age of sources of igneous rocks by zircon and monazite inheritance
Dating multiple growth zones in complex minerals even at micron-scale dimension
Timing of diagenesis and cementation of sedimentary rocks
Precise dating of extinction events facilitating the search for causality
Dating of calcite and phosphate of Quaternary age for environmental reconstructions and human evolution
Provenance of sediments, sedimentary rocks, dust, river sand, and ice-rafted glacial debris
Some of these topics are elaborated in chapters within this volume, but others are less well described. The following section highlights aspects of applications that are considered particularly interesting additions to complement the other chapters.
Minerals for U–Pb Dating
Titanite ((sphene, CaTiSi05))
Allanite (Ca2(Al3+, Fe3+, Fe2+)3Si3O12[OH]
Of these minerals only uraninite, thorite, and monazite have >1 % U and/or Th. Minerals most suitable for dating have little or no initial “common” Pb; this means that when dated the Pb contained in the mineral is primarily radiogenic. However, when common Pb is significant, either due to young age or low relative ratio of U/Pb (allanite, apatite, calcite, titanite), minerals can still be dated using an isochron approach, making a correction for the common lead isotope composition. Many of these minerals are amenable to either ID-TIMS or in situ SIMS or LA-ICP-MS dating methods, the latter requiring a reference mineral of the same type for the dating procedure.
Dating of Igneous Events and Related Applications
The early quest in U–Pb geochronology was mainly about determining the age of the Earth. By 1956, sufficient U–Pb ages documented Earth’s age to be >2.6 Ga, and work by Patterson (1956) on meteorite primordial Pb combined with information on terrestrial samples demonstrated that the Earth was ~4.5 Ga. Most work in U–Pb geochronology of the 1950s–1960s concerned dating igneous minerals – zircon and titanite – to constrain igneous events and populate the geological time scale and using the punctuated emplacement of felsic volcanic or plutonic rocks to constrain tectonic and orogenic events. These topics are dealt with in other chapters. Zircon became the mineral of choice mainly because it was realized early on that it was very refractory, resisted resetting, and could reveal igneous events in spite of subsequent alteration and metamorphism. Indeed, the work of Gulson and Krogh (1973) and Copeland et al. (1988) documented a phenomenon called “inheritance,” which is the survival (i.e., non-dissolution) of a preexisting phase – zircon and monazite – throughout a melting and magma crystallization event due to saturation of that magma in the mineral in question. These “inherited” grains are used extensively to deduce the age and geochemical characteristics of the source and/or country rock regions of the magmas. Beginning around 1990, the mineral baddeleyite (ZrO2) began to be dated in diabase and gabbroic rocks by U–Pb methods (Heaman and LeCheminant 1993), allowing reliable and precise ages of mafic magmas, including ophiolites to be dated. This development was paired quickly with palaeomagnetism of such rocks (dyke swarms, major mafic igneous provinces) to more robustly reconstruct the whereabouts of continental fragments dispersed by rifting and to determine the positions of continents on the Earth more accurately throughout geological time. Kimberlite dating by U–Pb is dealt with in a related chapter using the minerals perovskite and baddeleyite. U–Pb geochronology of igneous zircon and its application to geological problems are addressed in other chapters of this volume.
Meteorite U–Pb Dating
U–Pb geochronology of meteorites is a discipline in itself, requiring very rigorous sample preparation and chemical/isotope measurement procedures; a chapter is devoted to this topic in this volume. The antiquity of meteorites and the rapid change in the radiogenic 207Pb/206Pb ratio early in Earth history resulting from higher 235U/238U permit very precise dating of crystallization of the earliest materials of the solar system such as chondrules and CAIs (calciumaluminium inclusions). These materials contain relatively high U/Pb ratios, and it has been possible to date these 4,567 Ma old grains to better than ±0.5 Ma uncertainty.
Earth’s Oldest Zircons
In terms of the oldest materials identifiable on the Earth, the detrital zircons of quartzose metasedimentary rocks (Mt Narryer quartzite; Jack Hills conglomerate) of Western Australia have been a gold mine of information. Reasonably abundant (~3 % of detrital zircons) >3.9 Ga zircons occur in these samples, and they have been dated extensively (>100,000 zircon grains have been analyzed). The geochemical characteristics of these old grains (Hf isotopes, rare earth elements (REE) spectra, oxygen isotopes, Ti thermometry, inclusion mineral characteristics) have been researched at length (Amelin and Ireland 2013); these are mentioned in a related chapter. Among the most interesting conclusions of these studies are that the oldest zircons on Earth are ~4.36 Ga, significantly younger than the last stages of crystallization of the lunar magma ocean at ~4.42 Ga (Nemchin et al 2009). Hf isotope studies by many authors on these old zircons reveal that most if not all are derived from mantle reservoirs that are depleted to chondritic in composition, indicating that significant silicate mantle differentiation took place very early in Earth history. The extent to which this resulted in “continental” crust, as opposed to mafic crust, is very difficult to determine and no doubt will continue to be a topic of research well into the future. On the basis of existing work, it appears unlikely that zircons >4.4 Ga will be found on Earth, even though it is certain that scientists will continue to search for them!
U–Pb Dating of Metamorphism and Quantifying Time in P–T Evolution
A related chapter describes the application of U–Th–Pb dating of monazite, xenotime, allanite, rutile, and titanite to the chronology of metamorphism. The chemistry of these accessory minerals makes it a challenge to relate their growth directly to the pressure and temperature conditions of metamorphism, since they do not form directly from reactions of P- and T-sensitive minerals like garnet, aluminosilicates, micas, etc. Nevertheless, studies of Smith and Barreiro (1990) showed how monazite grows in amphibolite facies conditions, appearing usually in the staurolite zone of Barrovian metamorphism, a generalization that still appears to hold. Other minerals such as allanite preserve reaction relationships to apatite and monazite and at times other more abundant rock-forming minerals. Because garnet is a major reservoir for yttrium (Y) and because Y can occur in significant proportions in monazite due to its solid solution with xenotime, Y-zoning is used as a “monitor” of the presence/relative abundance of garnet during monazite growth. This allows monazite chemistry to be used to relate its growth to conditions of metamorphism defined by other pelitic minerals. Clearly, the textural relationships of these accessory minerals in relation to the fabric of other rock-forming minerals are crucial pieces of evidence concerning the relative age of accessory minerals. For example, inclusions of accessory minerals within other P–T-sensitive phases (garnet, coesite, diamond, etc.) can preserve the oldest prograde metamorphic ages. Phases that grow during decompression and retrogression, for example, titanite in rutile-bearing eclogite, can be used to date the exhumation/retrogression part of the P–T history. In these geological situations, the use of in situ methods is usually a requirement, along with data on mineral chemistry, phase equilibria, textural data, geological setting, and field relations. In situ methods need to measure wherever possible all U–Th–Pb isotopes, not just U–Pb, for monazite, due to the problem of excess 206Pb in Th-rich monazite (Parrish 1990), in order to determine accurate ages in young metamorphic rocks. Often, multiple minerals dated in the same sample will reveal a richer and more complete metamorphic history. A particularly common characteristic, which is increasingly documented in detailed studies of metamorphic rocks, is that monazite growth takes place over a long period of time, including prograde, peak, and in part retrograde conditions. Examples published recently in the Himalaya and in the Cordillera have documented tens of millions of years of time in a rock’s P–T evolution. One of the challenges in doing this in situ work is that reference minerals of well-determined age suitable for age calibration may not exist, prompting the search for better mineral standards, for example, for rutile and allanite (Bracciali et al. 2013).
U–Pb Dates of Minerals as Thermochronometers
Although many minerals datable by U–Pb methods were initially regarded as having high closure temperatures, many of these can be effectively used as thermochronometers. In contrast to zircon and monazite, it has long been documented that the retention of Pb by titanite, allanite, rutile, and apatite is not complete during amphibolite facies conditions of metamorphism or reheating. Early work in case studies showed clearly that U–Pb dates on titanite, rutile, and apatite were generally younger than the peak of metamorphism, often much younger. When considered along with dates from other minerals (micas, hornblende) of known closure temperature, the Pb-retention temperature of these minerals was regarded as being in the range of 400–600 °C. More recent experimental diffusion studies have generally confirmed these estimates. Based on empirical and experimental data, the order of retention of Pb for these minerals from highest retention to lowest is allanite > titanite > apatite > rutile, with allanite and titanite being ~600–650 °C and apatite and rutile being closer to 500 °C. When combined with other thermochronometers (Rb–Sr mica, Ar–Ar hornblende and mica, fission track, U–Th–He), a full thermal history for metamorphic rocks can be established from ~700 °C to <100 °C, with U–Pb dating constraining the thermal and mineral growth history above 400–500 °C (Parrish 2001). Together with chemical and phase equilibria data, the P–t history (exhumation and /or burial) can also be constrained, but to a lesser degree of detail.
Provenance Applications of Detrital Mineral Dating
The identification of the source of clastic sedimentary rocks, or provenance analysis, enables a reconstruction of part of the geological history. The first lines of evidence are clast composition in conglomerate and the composition of grains (lithic, quartz, feldspar, heavy minerals) in sandstone. However, without further information, few rocks can be proven to be derived from a specific source. U–Pb dating of detrital minerals provides probably the most important information allowing the discrimination of source areas for clastic rocks. Detrital mineral dating began to be tractable once single minerals could be dated. This work began in the mid–late 1980s using single-grain ID-TIMS dating, but the tedious aspect of the dating meant that few grains could be dated in a given study. The appeal of this type of data captured the imagination of geologists, and a new field of research increasingly gained momentum through the 1990s and in recent years. Some of these studies emphasized the desirability of systematically examining the detrital zircon signatures of formations in continents and in orogens, whereas other studies applied this single-grain dating technique to specific problems, for example, the determination of the maximum age of sedimentation of a Precambrian sedimentary rock for which few depositional age constraints existed. This field has exploded into what one might call “mass production” of detrital zircon data, covering most continents’ major sedimentary rock packages and including modern river detritus.
The proliferation since ~2000 of in situ U–Pb dating methods such as LA-Q-ICP-MS has encouraged many geologists to conduct this type of work as “users” who follow a standard, often somewhat unsophisticated, recipe for analysis in order to address provenance questions. This trend involves generating a huge amount of data often illustrated in probability distribution plots that show both uncertainties and abundance of data in only semiquantified form. Sometimes the data are not uniformly rigorously collected, some having no information about whether single zones in zircon are dated, whether data are concordant or nearly so, or whether the 206Pb/238U or the 207Pb/206Pb ages are being used for the interpreted age of a grain. Although the data is usually published in supplementary tables online, the interpretation of detailed data is a challenge to obtain as an interested reader. This can compromise the robustness of such datasets. A great variety of quality of data from such studies exists, and this diversity in quality continues to be published. Readers are cautioned to examine data thoroughly and question whether interpretations made are sufficiently rigorous!
Many geological questions pertaining to provenance of sediments simply cannot be answered by zircon single-grain dating alone. Approaches that use the most pertinent geochemical and dating methods are likely to be able to answer questions better, for example, by the use of additional detrital minerals (U–Pb dating of detrital monazite and/or rutile in sediments; Ar–Ar dating of detrital mica) and the use of isotope tracers of dated minerals (Nd in detrital monazite, Sr–Nd in detrital titanite and apatite, Hf isotopes in detrital zircon) (Najman et al. 2008). Excellent science demands a broad approach that is tailored to the question being addressed and which uses the most appropriate isotope methodology.
Sedimentological, Environmental, and Paleoclimate Applications of U–Pb Dating
A chapter on U–Pb applications to diagenetic events is part of this volume and presents methods for addressing this topic. A key part of this field is the application of U–Pb methods (both ID-TIMS and in situ methods) to carbonate (calcite, aragonite) and phosphate that occur as cement or as primary sedimentary materials deposited usually from marine or freshwater (fossils, cements, speleothems, lake carbonate, carbonate veins, hominin or other animal phosphate). While there are currently few extensive studies published, it is clear that this is a field that will grow rapidly in the near future.
U is partitioned into aragonite in marine fossils, and calcite includes sufficient U to also be dated, when the initial incorporation of lead into the mineral is low, as is often the case with precipitation of calcite from groundwater. Thus when a material has a high U/Pb ratio, these carbonate or phosphate minerals can be dated, by either ID-TIMS methods (Smith and Farquhar 1989; Rasbury et al. 1997; Richards et al. 1998) or by a combination of ID-TIMS and in situ methods (Woodhead et al. 2012). Indeed it is now tractable to apply LA-ICP-MS using high-sensitivity instrumentation to date calcite in situ even when the U content is not particularly high, via the use of a calcite standard. This field of research will likely explode in popularity and demand over the next 10 years.
Collaboration and Improvement of Methods
The field of U–Pb dating has come a long way since its inception by the work of Boltwood and Holmes about 100 years ago in the period immediately following the discovery of radioactivity. With the development of sufficiently sensitive mass spectrometry by Nier in the 1930–1940s and the availability of isotope tracers of U and Pb in the 1950s and 1960s, the field of ID-TIMS exploded with refinement of methods, use of Teflon acid distillation and mineral dissolution vessels and multi-collector high-sensitivity mass spectrometers. Recognition of complex zircon growth histories demanded the development of better ways to analyze concordant zircons and develop in situ methods, which in the past 20 years have exploded in availability and popularity. This has brought U–Pb geochronology within reach to geologists, some of whom never worked in a laboratory previously. Nevertheless, the advancement in methods is very impressive, both intellectually and technologically. Innovative applications of U–Pb dating to geological problems have generally kept pace with technology, because geological questions usually precede the availability of efficient and accurate instrumentation and methods. For example, widespread detrital single-grain dating had to wait until efficient in situ methods of analysis of large numbers of grains became available. The use of single grain rutile U–Pb dating in provenance studies will become more common following the recent development of rutile reference materials and sensitive LA-ICP-MS methods (Bracciali et al. 2013). Similarly, the availability of calcite U–Pb reference materials and sensitive mass spectrometry capable of measuring young ages precisely in minerals with low U contents is just now tractable; this will cause a rapid increase in demand and application of U–Pb dating to calcite and Quaternary materials, hardly addressed at the present time.
Collaboration among U–Pb laboratories has been essential because of the inherent ability of ID-TIMS methods to measure isotope ratios apparently more precisely than one can confidently calibrate isotope tracer solutions. The EARTHTIME project (www.earth-time.org) is one of these international efforts that has recognized the need for more collaboration and use of common solutions to monitor and measure interlaboratory data, to ensure that it is comparable, and to encourage laboratories to quote uncertainties that are accurate and realistic. For example, age solutions and mixed 233U–235U–205Pb isotope tracers are available to world laboratories for the purpose of improving the precision and accuracy of data produced. This effort has been extended, for example, to U-series dating and to in situ methods of U–Pb dating, so that numerous world laboratories are able to compare data, assess their own performance against international standards, and share the best practice. This type of collaborative philosophy, combined with continued technological and method innovation, is the way forward for this important area of geochronology.
- Boltwood, B. B., 1907. On the ultimate disintegration products of the radioactive elements, part II: the disintegration products of uranium. American Journal of Science, 23, 77–88.Google Scholar
- Heaman, L., and Parrish, R. R., 1991. U–Pb geochronology of accessory minerals. In Applications of Radiogenic Isotope Systems to Problems in Geology. Short course handbook, J Ludden and L Heaman, eds., Mineralogical Association of Canada, Vol. 19, pp. 59–102.Google Scholar
- Ludwig, K., 1991. ISOPLOT – A Plotting and Regression Program for Radiogenic Isotope Data. Denver: US Geological Survey. US geological survey open file report, 91–445.Google Scholar
- Mattinson, J. M., 1973. Anomalous isotopic composition of lead in young zircons. Carnegie Institution of Washington Yearbook, 72, 613–616.Google Scholar
- Najman, Y., Bickle, M., Carter, A., Garzanti, F., Wilbrans, J., Willett, S., Oliver, G., Parrish, R. R., Akhter, S. H., Allen, R., Ando, S., Chisty, E., Reisberg, L., and Vessoli, G., 2008. The Palaeogene record of Himalayan erosion. Earth and Planetary Science Letters, 273, 1–14.CrossRefGoogle Scholar
- Parrish, R. R., 2001. The response of mineral chronometers to metamorphism and deformation in orogenic belts. In Miller, J. A., Holdsworth, R. E., Buick, I. S., and Hand, M. (eds.), Continental Recactivation and Reworking. London: Geological Society. Special publications, Vol. 184, pp. 289–301.Google Scholar