The Age of a Diamond and the Time of Its Journey
The question “How old is a diamond?” sounds simple, but in geology it may refer to several different times.
At a minimum, we must distinguish:
- the time when the diamond, or one of its zones, grew;
- the time it spent stored in the mantle;
- the time of the kimberlite or lamproite eruption;
- the time of later weathering, sediment transport, and deposition.
These times may be separated by billions of years.
Why diamond is not dated with carbon-14
Radiocarbon dating uses the isotope carbon-14, which has a half-life of approximately 5,730 years. This is excellent for archaeological and geologically young organic materials, but entirely unsuitable for diamonds that are hundreds of millions or billions of years old.
Stable carbon isotopes in diamond are important for studying the source and cycle of carbon, but by themselves they do not provide a calendar age for the crystal.
Diamond itself generally lacks a good radiometric clock
Diamond is almost entirely carbon and usually contains too little of the elements needed for standard radiometric dating.
Geochronologists therefore most often analyze mineral inclusions that contain suitable parent and daughter isotopes.
The most important systems in diamond geochronology are:
- Sm–Nd in garnet and clinopyroxene;
- Rb–Sr in certain silicate inclusions;
- Re–Os in sulfides;
- in special cases, other methods such as U–Pb, Ar–Ar, or systems applicable to specific inclusion or fluid materials.
Re–Os dating of sulfide inclusions has become especially important because extremely small sulfide phases can contain enough rhenium and osmium for dating.
Sm–Nd and Rb–Sr were historically crucial to the first reliable ages obtained from silicate inclusions. Their strength increases when multiple garnets and clinopyroxenes of the same paragenesis can be analyzed with a sufficiently large range of isotopic ratios.
Re–Os is particularly useful for sulfides because rhenium and osmium partition strongly into sulfide phases. The development of more sensitive techniques has enabled the analysis of individual microscopic sulfides, and in rare cases multiple sulfides from the same diamond can define an internal isochron and link the resulting date much more directly to the specific crystal.
Isochrons and population ages
A radiometric clock is not simply a matter of “measure one number and read the date.”
An isochron analyzes samples that should belong to the same geological event but have different parent-to-stable-isotope ratios. If they were in isotopic equilibrium and then remained sufficiently closed, their relationship may define a line whose slope yields an age.
In diamond geochronology, this often means combining inclusions from multiple diamonds of the same locality and paragenesis.
The resulting age therefore often describes:
- one growth episode;
- one diamond population;
- a geological event associated with a particular fluid.
It need not establish the age of every diamond from the mine.
If there are too few samples for an isochron, researchers may calculate model ages from an assumed evolutionary model for the source. A model age can be geologically useful, but it is not evidentially equivalent to a well-defined isochron. The final text must therefore always identify the type of age, not merely the number.
Red flag “Diamonds from Mine X are 3.3 billion years old” is often too broad a claim. It is more precise to say that a population with an approximate age of 3.3 Ga has been identified and dated at that deposit.
Protogenetic and syngenetic inclusions
One of the greatest interpretive challenges is the relationship between an inclusion and its host diamond.
- A syngenetic inclusion crystallized during the same growth event.
- A protogenetic inclusion existed before the diamond enveloped it.
- An epigenetic inclusion formed after the principal growth phase.
At first glance, a protogenetic inclusion appears problematic: its crystallization age may be older than the diamond. During a diamond-forming event, however, the mineral may isotopically re-equilibrate with the fluid. In such a case, the isotopic system may record the time of diamond growth even though the mineral grain itself is older.
“Protogenetic” therefore does not automatically mean “unsuitable for dating.” It is necessary to determine what a particular isotopic system actually recorded.
Uncertainty is part of the result
A diamond age should not be written as a perfectly precise date. In many classic studies, uncertainties are on the order of hundreds of millions of years, especially when very small inclusions and ancient systems are analyzed.
Such uncertainty can still be highly useful geologically. The difference between events at 3.5 Ga, 2.9 Ga, and 1.1 Ga is sufficient to relate diamond-forming events to different phases of continental construction, subduction, or orogeny.
The historic breakthrough: Kimberley
Classic analyses from the 1980s showed that mineral inclusions in diamonds from the Kimberley area of South Africa yielded ages on the order of 3.2–3.3 billion years, while the kimberlites that transported them were far younger, on the order of tens of millions of years.
This made it clear that kimberlite did not create most of the diamonds it carries. It entrained them from a much older mantle reservoir.
[VISUAL 9.1: Three clocks—diamond growth several billion years ago → long mantle residence → much younger kimberlite eruption]
The oldest well-dated populations
The oldest well-studied and directly dated populations include diamonds from Canada’s Slave craton, especially Diavik and Ekati, with growth events approximately 3.5–3.3 billion years ago.
As an especially strong example, a Re–Os regression of 11 peridotitic sulfide inclusions from five diamonds from the Panda kimberlite in the Ekati area yielded an age of 3.52 ± 0.17 Ga. Such a result dates a defined population and event; it does not mean that every diamond from Ekati or the entire Slave craton is the same age.
This does not mean that every diamond from those mines is that old. A single deposit may contain multiple populations and multiple growth episodes.
Diamonds do not necessarily grow for billions of years
If a diamond is three billion years old, that does not mean it grew continuously for three billion years.
Crystallization may be a geologically relatively brief episode, after which the diamond remained stored in a cold cratonic root until a much younger eruption.
The interval from growth to eruption is therefore principally mantle residence time, not the duration of crystallization.
Multiple zones, multiple times
A single diamond may have a core and coats that formed in separate episodes. Nitrogen zoning, cathodoluminescence, inclusion chemistry, and spectroscopy may reveal boundaries between those events.
If only one inclusion is analyzed, the result does not automatically date the entire crystal. At best, it dates the relationship among that inclusion, a particular zone, and the geological event recorded by the isotopic system.
Nitrogen is a thermal record, not a simple calendar
Over time, nitrogen in diamond may change from isolated atoms into pairs and more complex aggregates. The rate of aggregation depends strongly on temperature, so nitrogen state can provide information about a time–temperature history.
The same degree of aggregation, however, may result from different combinations of time and temperature. Nitrogen must therefore not be used as a stand-alone precise clock without independent geological constraints.
Dating is often destructive
Classic inclusion geochronology often requires the mineral to be:
- physically extracted from the diamond;
- cleaned and chemically processed;
- dissolved;
- analyzed by mass spectrometry.
Large, valuable gemological diamonds are therefore almost never sacrificed solely to obtain an age. Scientific ages generally come from research samples and populations, not from market stones accompanied by a grading report.
Age is not market quality
A laboratory report generally does not state a diamond’s geological age. Even when the age of a population from a particular deposit is known, it is not automatically a characteristic that can be assigned to an individual stone without evidence of origin and appropriate analysis.
Age is also not part of the 4Cs system and does not in itself determine market value.
Chapter summary
- The age of diamond growth, mantle residence time, and the age of the kimberlite are three different geological clocks.
- Carbon-14 is not applicable to diamonds that are millions or billions of years old.
- Diamond is most often dated indirectly through mineral inclusions.
- Sm–Nd, Rb–Sr, and Re–Os are the most important classic isotopic systems in diamond geochronology.
- An isochron most often provides the age of a population or event, not automatically every stone from a mine.
- A protogenetic inclusion may be older than the diamond, but its isotopic system may be reset during the diamond-forming event.
- Uncertainties of hundreds of millions of years can be normal and geologically useful.
- The oldest well-dated Diavik/Ekati populations reach approximately 3.5–3.3 Ga.
- A three-billion-year-old diamond did not necessarily grow for three billion years; it may have spent most of that time stored in the mantle.
- A single crystal may contain multiple growth phases and zones.
- Nitrogen provides a time–temperature record but is not a stand-alone precise clock.
- Dating inclusions is often destructive and is therefore not a standard gemological service.
[VISUAL 9.2: Evidence hierarchy for age—a specific dated inclusion → diamond/zone → population → deposit; what may and may not be generalized]