B5 · Publication Volume 10

Stable Isotopes

fractionation, temperature and tracing fluid or source

Learning goals

After this lesson, you should be able to write an isotope ratio explicitly, calculate and interpret delta notation, distinguish a reference material from a reference scale, calculate a fractionation factor, and separate equilibrium, kinetic, mixing and source effects. You should also be able to design sampling that preserves mineral domain, fluid fraction and exchange history.

Stable-isotope measurements are comparative. A delta value is not a concentration, age or direct temperature. It expresses the relative difference between a sample ratio and a defined reference ratio after calibration. Geological interpretation requires an isotope-bearing element budget, a fractionation model and evidence that the measured material preserves the process of interest.

Ratios, delta notation and reference scales

Stable-isotope measurement and interpretation chain: ratio, reference scale, fractionation, mixing and process
Stable-isotope measurement and interpretation chain: ratio, reference scale, fractionation, mixing and process

For a heavy-to-light isotope ratio R, such as {}^{18}O/{}^{16}O, delta notation is

$\delta=\left(\frac{R_{sample}}{R_{reference}}-1\right)1000\ \text{per mil}.$

A value of +10\ \text{per mil} means the sample ratio is 1.010 times the reference ratio; it does not mean the sample contains 1% heavy isotope. The absolute abundance depends on the ratio and total amount of the element. Always identify both isotopes, the material analysed and the reporting scale.

Reference materials realise and connect a scale. Measurements may be normalised using two or more materials to correct scale compression and laboratory bias. A material used for daily quality control is not automatically the scale definition. Record accepted values, batch, preparation, normalisation equation and uncertainty.

Different elements and isotope systems have different reference conventions. Never transfer a delta symbol without its isotope pair and scale. A table headed only “delta” is not reusable data.

Fractionation and exchange

For materials A and B, a fractionation factor can be defined

$\alpha_{A-B}=\frac{R_A}{R_B} =\frac{\delta_A/1000+1}{\delta_B/1000+1}.$

The quantity 1000\ln\alpha is often convenient because sequential fractionations add approximately in logarithmic form and because it is close to the delta difference for small fractionations. The direction convention must be stated; \alpha_{A-B} is the reciprocal of \alpha_{B-A}.

Equilibrium fractionation reflects different isotope distributions among coexisting phases at equilibrium and commonly varies with temperature. Kinetic fractionation arises when forward and reverse rates do not balance, such as rapid diffusion, evaporation or incomplete reaction. Rayleigh fractionation models progressive removal of a product from a reservoir under specified fractionation and mixing assumptions. These are model classes, not labels inferred from the sign of delta alone.

Exchange can reset a mineral or fluid without visible bulk replacement. The extent depends on diffusion, dissolution-reprecipitation, fluid access, grain size and time. A temperature calculated from two phases is meaningful only if the phases equilibrated with each other at the event of interest, retained their compositions, and satisfy the calibration domain.

Sources, mixing and geological interpretation

Isotope compositions can distinguish reservoirs when source ranges are sufficiently different and subsequent fractionation is constrained. They rarely identify a source uniquely on their own. Two sources can overlap, and one source can evolve through reaction, evaporation or exchange. Pair isotope data with concentrations, mineral hosts, field relations and independent tracers.

Mixing must conserve isotope-bearing amounts. For reservoirs j with element amount n_j and isotope ratio R_j, the exact mixed ratio depends on heavy and light isotope inventories. For small natural variations and similar total isotope abundance, an amount-weighted delta approximation may be adequate, but the weights are element amounts, not automatically water volume or rock mass.

In water systems, evaporation, condensation, recharge, rock exchange and mixing can overlap. In carbonate or silicate minerals, precipitation temperature, fluid composition, kinetic growth and later alteration all affect values. In sulfur-bearing minerals, oxidation state and microbial or abiotic pathways may matter. A bivariate plot is a starting geometry, not a unique process solution.

Temperature and mass balance

Many equilibrium calibrations have a form related to 1000\ln\alpha=A/T^2+B/T+C, with coefficients specific to phases and calibration. Extrapolation beyond the experimental or empirical domain can be severe because temperature is recovered from an inverse nonlinear relation. Propagate uncertainty from both measured values and calibration parameters, and test sensitivity to disequilibrium.

Sampling and measurement design

Define the isotope-bearing domain: bulk rock, mineral separate, growth zone, structural carbonate, water fraction, dissolved species, gas or organic compound. Petrographic screening is essential where alteration, inclusions or multiple generations are possible. Microsampling should follow textures rather than a regular grid that mixes events.

Record preparation that can exchange or remove the element of interest. Drying, acid reaction, heating, washing and storage may alter water, carbonate or organic components. For waters, record filtration, preservation, headspace, evaporation risk and field conditions. For minerals, retain images and maps linking each analysis to a domain.

Quality control includes blanks, memory, linearity, drift, reference-material normalisation, replicate preparation and matrix effects. Report the full isotope ratio or delta notation, scale, uncertainty and whether replicates describe measurement repeatability, preparation reproducibility or natural heterogeneity. These are different variance components.

Worked synthetic examples

Delta and fractionation. A synthetic phase A has \delta_A=+10\ \text{per mil} and phase B has \delta_B=-20\ \text{per mil} on the same scale. Then

$\alpha_{A-B}=\frac{1.010}{0.980}=1.030612,$

and

$1000\ln\alpha_{A-B}=30.15\ \text{per mil}.$

The simple delta difference is 30\ \text{per mil}, close but not identical. The result says A has the higher heavy-to-light ratio. It does not identify equilibrium, temperature or process without an applicable calibration and evidence of coexisting exchange.

Mixing. Two synthetic waters have \delta=-80\ \text{per mil} and +10\ \text{per mil}. If they contribute 30% and 70% of the isotope-bearing oxygen inventory, and the small-difference approximation is acceptable,

$\delta_{mix}\approx0.30(-80)+0.70(10)=-17\ \text{per mil}.$

If the waters contribute equal volumes but different dissolved or bound oxygen inventories for the analysed fraction, those are not the correct weights. A mixing calculation must match the analysed element pool.

Interpretation workflow and uncertainty

  1. Name isotope pair, analysed material and spatial domain.
  2. Preserve raw measurement, reference materials and scale normalisation.
  3. Separate measurement repeatability, preparation variability and geological heterogeneity.
  4. Write fractionation direction and equation explicitly.
  5. Test equilibrium, kinetic, mixing and exchange models against textures and concentrations.
  6. Use amount balance for mixing and reaction.
  7. Apply calibrations only within their material and temperature domain.
  8. Propagate analytical and calibration uncertainty and explore open-system alternatives.
  9. Report what the isotope data exclude as well as what they support.

Failure modes

Common failures are treating delta as a concentration, omitting the scale, averaging unlike materials, subtracting delta values as if exact in all cases, using rock mass instead of element inventory in mixing, and calculating a temperature from phases that did not equilibrate. Another is interpreting a published source field as timeless and universal while ignoring local variation and later exchange.

Uncertainty should include sampling domain, reference scale, normalisation, fractionation calibration, exchange and source variability. Geological uncertainty often dominates instrument precision. Display individual domains and textures rather than only a pooled mean.

Practice and review

  1. Convert \delta=-25\ \text{per mil} to R_{sample}/R_{reference}.
  2. Calculate \alpha_{A-B} and 1000\ln\alpha for two supplied delta values, then reverse the convention.
  3. Construct a three-component isotope-and-concentration mixing problem and identify whether the solution is unique.
  4. Design a microsampling path across a zoned vein that avoids mixing growth generations.
  5. Write competing explanations for an isotope shift accompanied by constant concentration, and for a shift accompanied by large concentration change.

For the completion exercise, create a synthetic paired-mineral dataset containing an equilibrium pair, a reset rim and a mixed analysis. Calculate fractionations, flag domains, and write an interpretation that does not assign temperature to the invalid pairs.

Sources and further reading