The Gold Standard of Geochronology | Rb–Sr Isochron Slope, Initial Ratio and the Mixing-Line Trap

Geochronology does not merely attach a number to a rock. It asks which physical event a mineral or rock system can preserve. In an Rb–Sr study the relevant event can be formation of a comagmatic assemblage, mineral crystallisation or later isotopic equilibration. A mathematical age becomes geological history only when the samples and system behaviour support that assignment.

An isochron uses several samples or mineral separates that ideally shared one initial strontium isotope ratio at the event of interest but acquired different parent-to-stable-reference ratios. Their later radiogenic growth then creates a slope in isotope-ratio space. That slope can carry age information if the system remained appropriately closed.

The important distinction is between an analytical correlation and a dated event. A plot can contain four precise numbers that lie on a line but still represent a mixture rather than a closed-system suite. Before looking at regression, identify the sample group, rock textures and geological reason for expecting common initial Sr isotope composition.

Here we use an intentionally constructed table with four exact model points so the logic is visible. They are not new laboratory measurements, and the age is an illustration calculated from a stated isotope-decay constant rather than an age claimed for any named formation.

Contents · Next · Subject root

Radioactive ⁸⁷Rb decays to ⁸⁷Sr. The resulting radiogenic ⁸⁷Sr contribution grows with time, while ⁸⁶Sr provides a stable reference isotope for normalising the measured ratios. This gives an x coordinate of ⁸⁷Rb/⁸⁶Sr and a y coordinate of ⁸⁷Sr/⁸⁶Sr.

An increase in ⁸⁷Sr/⁸⁶Sr can reflect radiogenic growth, but not every Sr-ratio difference is automatically a consequence of time. Different starting reservoirs can already have different isotopic compositions. Mass balance and common initial composition must therefore be considered before converting an x–y pattern into an age.

In the ideal model, the present-day ⁸⁷Sr/⁸⁶Sr ratio is the initial ratio plus the radiogenic addition scaled by remaining parent Rb. That is why we plot the current ratio of radioactive parent to stable normaliser, rather than simply ‘total rubidium’ against ‘total strontium’ with no isotope accounting.

Real analyses include instrumental fractionation corrections, standards, blanks and measurement uncertainties. The geometric tutorial deliberately treats the supplied ratios as already corrected; a real report must explain how those ratios were obtained and propagated into its fit.

Contents · Previous · Next · Subject root

Start with the basic decay-growth relation for one closed system: the number of radiogenic daughter atoms generated since time t is the remaining parent amount multiplied by e^(λt)−1, where λ is the radioactive decay constant. Divide by a stable reference isotope of the daughter element to express the equation with measurable ratios.

The conventional Rb–Sr isochron form is (⁸⁷Sr/⁸⁶Sr)present = (⁸⁷Sr/⁸⁶Sr)initial + (⁸⁷Rb/⁸⁶Sr)present × [e^(λt)−1]. Write y=b+mx, where b is the initial Sr ratio and m=e^(λt)−1.

The line slope is therefore not t itself. To obtain an age, invert the exponential: t=ln(1+m)/λ. The logarithm is dimensionless, and λ must use inverse time units consistent with the desired output. A slope of 0.007 is not ‘seven thousand years’ or ‘0.7% of an age’.

The intercept b does not mean the present-day ⁸⁷Sr/⁸⁶Sr of every sample at x=0 was actually observed. It is the regression’s extrapolated initial composition under the common-initial-ratio model. Treating it as a physically meaningful shared reservoir is an assumption requiring geological support.

Contents · Previous · Next · Subject root

Use an illustrative λ=1.42×10⁻¹¹ year⁻¹ for ⁸⁷Rb. High-precision geochronology requires current, appropriately calibrated decay constants and their uncertainty. We choose this rounded value to make the numerical demonstration exactly reproducible rather than to assert an uncertainty-free universal constant.

Now suppose the underlying event occurred t=500,000,000 years ago. The model slope is m=e^(λt)−1=e^(0.00710)−1≈0.007125265. This value is small because ⁸⁷Rb decays slowly on the time scale considered.

Use an initial strontium ratio b=0.704000. Then the synthetic line is y=0.704000+0.007125265x. Any chosen x value generates one point whose coordinates meet the ideal isochron formula. For example x=1.000 gives y≈0.711125265.

These are mathematical construction inputs, not measured age constraints. Because the points are generated from one exact line, a regression will show zero residual in exact arithmetic; a real dataset with realistic instrumental noise would not normally achieve that artificial perfection.

Contents · Previous · Next · Subject root

Choose four different present-day ⁸⁷Rb/⁸⁶Sr ratios: x=0.20, 0.80, 1.40 and 2.00. The wide spread makes the conceptual slope easy to see. Substitute each x into y=0.704000+0.007125265x to calculate the corresponding ⁸⁷Sr/⁸⁶Sr coordinate.

The values are approximately 0.705425, 0.709700, 0.713975 and 0.718251, respectively. Each point represents a hypothetical sample with a different Rb/Sr isotope ratio. No individual point by itself reveals both the initial Sr ratio and the age unless an additional constraint is supplied.

Plot y vertically and x horizontally. A line through the points has the defined intercept 0.704000 and slope approximately 0.007125. If you accidentally swap the axes, the new regression slope is not the m required by the Rb–Sr isochron equation. The coordinates and their meaning are part of the method.

Students can reproduce the table in a spreadsheet using one cell for b, another for m and a formula y=b+m×x. Then vary b without changing m and observe a vertical shift. Vary m without changing b and observe a change in slope. This separates the two physical inferences.

Contents · Previous · Next · Subject root

Take m≈0.007125265. Apply t=ln(1+m)/λ with λ=1.42×10⁻¹¹ year⁻¹. The numerator ln(1.007125265) is approximately 0.00710, and the resulting t is approximately 500,000,000 years = 500 Ma.

A useful reverse check is to start from 500 Ma and recover the slope using e^(λt)−1. Agreement between the two directions confirms the arithmetic and units for the synthetic exercise. It does not independently confirm a geological event, because the points were generated from that same model.

The intercept 0.704 is a dimensionless isotope ratio, not a time. The slope is also dimensionless; only after division by λ does an elapsed time emerge. Labelling an axis intercept as ‘the rock’s initial age’ confuses isotopic composition with event chronology.

To report a real isochron age responsibly, include the statistical fit, analytical uncertainty, covariance treatment, scatter, decay-constant choice, sample selection and geological interpretation. Printing ‘500.000 Ma’ from a rounded illustrative λ would create false precision.

Contents · Previous · Next · Subject root

Suppose only one sample with x=1.000 and y≈0.711125 is analysed. Without further information, both b and t are unknown in y=b+x[e^(λt)−1]. Infinitely many b–t combinations can satisfy that single equation within its physical domain.

If a researcher arbitrarily assumes b=0.706 instead of the constructed b=0.704, the inferred slope at x=1 becomes m≈0.005125. Its apparent age is ln(1.005125)/λ≈360 Ma, a dramatic departure from the model’s 500 Ma. The isotope measurement did not change; the initial-composition assumption changed.

The issue is especially dangerous for low-Rb/Sr samples, where a modest uncertain initial Sr contribution can represent a large part of the measured y value. An isochron approach tries to estimate b and m together from multiple samples with varied x values, reducing reliance on a single guessed b.

This is not a claim that a multi-sample line always solves every problem. If the samples begin with heterogeneous Sr ratios, experienced open-system changes or form a mixing trend, an impressive regression may still have a physically misleading intercept and age.

Contents · Previous · Next · Subject root

Consider two hypothetical endmember components with isotope-ratio coordinates A=(0.20, 0.705425) and B=(2.00, 0.718251). Intermediate mixtures can be formed between them. If a mixing weight refers to the fractions of total ⁸⁶Sr contributed by each endmember, both the x and y isotope ratios are weighted by that same fraction.

With equal contributions of ⁸⁶Sr from A and B, x_mix=(0.20+2.00)/2=1.10 and y_mix≈(0.705425+0.718251)/2=0.711838. The mixture lies on the straight line connecting A and B. Other weights likewise trace a line in this shared-denominator ratio plot.

Therefore a sequence of mixture proportions can yield a visually excellent straight array even if none of its points shares the age-and-initial-composition history required for an isochron. Two distinct processes—radiogenic growth and mixing—can sometimes produce similar ratio geometry.

This is a deliberately constrained mass-balance example. The relevant weight is the ⁸⁶Sr contribution, not automatically a 50:50 mass fraction of two bulk rocks if their Sr concentrations differ. That distinction is essential for an honest mixing calculation and is often where intuitive diagrams become physically inaccurate.

Contents · Previous · Next · Subject root

Any two distinct x–y observations define a straight line exactly. That geometry alone cannot test whether the isotope system is closed, whether all samples started at one Sr isotope ratio, or whether mixing explains the array. Adding samples with a useful spread of Rb/Sr can reveal lack of fit, but even a many-point line needs geological corroboration.

The regression should recognise uncertainty in both isotope ratios and possible correlation between them, rather than casually treating x as error-free in every case. A fit that is too precise because it ignored the x uncertainty can give unjustifiably narrow age intervals.

Samples should represent a meaningful geological assemblage. Mixing unrelated rock units merely to widen the x-axis does not make an isochron more robust. Wide ratio spread is valuable only if the common-initial-ratio and closed-system premises remain credible for that sample group.

Mineral separates may share an initial Sr ratio at formation and later undergo partial resetting. In some metamorphic contexts, homogenisation can create a new interpretable isotope line. The age assignment then depends on understanding which event established the shared isotopic condition.

Contents · Previous · Next · Subject root

A candidate Rb–Sr isochron should be evaluated alongside rock type, petrographic textures, elemental concentrations, mineral chemistry and the geological relationships among samples. For mixing, plot other isotope or elemental ratios and look for systematic covariation consistent with endmember contributions.

Check whether the putative initial Sr ratio is geologically plausible for the source reservoir and whether distinct samples really belonged to the same system at the inferred event. A regression intercept that looks numerically precise is not enough if it represents a physically impossible initial composition.

Compare other chronometers where available, including U–Pb zircon ages interpreted for the correct mineral domains. Agreement can strengthen a history if the systems date the same event, whereas disagreement may reveal inheritance, thermal resetting or unrecognised mixing. Do not force different chronometers to measure identical processes.

The purpose of this review is not to reject Rb–Sr dating. It is to use the isotope system’s strengths responsibly: several ratio measurements can constrain both a model age and an initial composition, provided the geological basis for that model is inspectable.

Contents · Previous · Next · Subject root

Practice A: A fitted synthetic Rb–Sr line has slope m=0.004 with λ=1.42×10⁻¹¹ year⁻¹. Its model age is ln(1.004)/λ≈281 Ma. It is not 0.004 Ma or four million years.

Practice B: The slope remains 0.007125 while the intercept rises from 0.704 to 0.706. What changes? The whole line shifts upward, but the age inferred from its unchanged slope does not change in the ideal isochron calculation. Whether the new intercept is a credible initial reservoir is a separate geological matter.

Practice C: Endmembers have (x,y)=(0.20,0.705425) and (2.00,0.718251). An equal-⁸⁶Sr mixture has x=1.10 and y≈0.711838. Does a perfect line through these three points prove 500 Ma crystallisation? No. It is compatible with the mathematical mixing model as constructed.

Practice D: A publication provides a 500 Ma Rb–Sr ‘age’ without uncertainties or any mention of initial Sr composition. Which questions should be asked? Request the measured x–y data, fitted intercept, decay constant, regression method, sample relationships and tests for mixing and openness.

Contents · Previous · Next · Subject root

First define the geological event to be dated and why the selected rocks or minerals may share one initial Sr isotope ratio. Prepare and analyse samples with documented corrections, standards and uncertainties. Plot ⁸⁷Sr/⁸⁶Sr against ⁸⁷Rb/⁸⁶Sr with units and isotope labels intact.

Second, fit the data using a statistical method appropriate to errors in the measured ratios. Compute the slope-to-age transformation t=ln(1+m)/λ and carry the uncertainty and decay-constant assumptions into the reported result. Interpret the intercept as a candidate initial composition, not just a decorative y-axis number.

Third, test alternate explanations: isotopic mixing, inherited sources, heterogeneous starting ratios, mobility of Rb or Sr, partial closure, and mineral resetting. A straight line earns geological meaning from consistent quantitative and field evidence, not from appearance alone.

Continue with The Gold Standard of Geochronology, then compare U–Pb concordia and discordia and youngest detrital zircon constraints. Their distinct assumptions show why an isotope ratio, model age and geological event should never be collapsed into one unqualified number.

Contents · Previous · Subject root