B5 · Publication Volume 10

Geochemistry, Isotopes and Geochronology

Connects element behaviour, isotope systems and geochronology to geological interpretation.

Purpose and boundary of this book

Geochemical measurements do not interpret themselves. A reported concentration is a property of a sampled material, an analytical procedure and a declared basis. An isotope ratio compares particular nuclides after correction, normalisation and reference-material calibration. A numerical age is the result of a physical model applied to a mineral or domain that may have formed, exchanged components, recrystallised or been reset at different times. This book teaches how to connect those results to geological questions without collapsing measurement, model and event into one number.

The sequence begins with periodic behaviour, ionic size, charge and crystal sites. It then develops partitioning, fractionation and mass balance; aqueous speciation and reaction; fluid-rock alteration; stable-isotope notation and fractionation; radioactive decay; commonly used dating systems; closure and resetting; age uncertainty; and the integration of geochemical evidence. The emphasis throughout is conditional reasoning: what material was measured, what reservoir or process the model represents, which assumptions are required, and which observations could contradict the preferred explanation.

This is not a laboratory methods manual, a catalogue of deposit signatures or a collection of automatic discrimination diagrams. It provides a scientific framework for designing observations, checking calculations and writing interpretations. Laboratory-specific preparation, instrument tuning, interference correction, quality control and accreditation requirements must come from the current validated method used for the work.

General and institution-neutral scope

This is a general, institution-neutral tutorial. It has no affiliation with, sponsorship by, endorsement from or curriculum relationship to any company or individual. It is not prepared for a named laboratory, mine, consultancy, university, instrument, software package or private dataset. All numerical examples are synthetic teaching data and all geological scenarios are invented unless a published study is explicitly identified in a source note.

People, standards bodies, public agencies, journals and repositories named in source notes identify technical sources only. Citation does not imply authorship, approval, partnership or participation in this tutorial. The website that hosts this material is only a delivery surface. It is not presented as the textbook's author, publisher, sponsor, scientific authority or curriculum subject.

Commercially named instruments and proprietary workflows are unnecessary for the reasoning developed here. A learner should be able to transfer the evidence chain to any competent laboratory output that reports the material, method, calibration, corrections, uncertainty and quality-control results needed for the intended interpretation.

The geochemical evidence chain

A defensible interpretation preserves at least six linked layers:

  1. Geological object: outcrop, interval, clast, grain, growth zone, vein generation, fluid inclusion, water body or experimental product.
  2. Sampling support: location, volume or mass, orientation, grain-size fraction, mineral domain, compositing rule and selection criteria.
  3. Preparation and measurement: separation, digestion or ablation, measured signals, blanks, interferences, fractionation correction, calibration and reference materials.
  4. Reported quantity: concentration, amount fraction, activity, isotope ratio, delta value, parent-daughter ratio or calculated date, with units and basis.
  5. Model: mass balance, partition law, equilibrium or kinetic reaction, decay equation, isochron, concordia relation, diffusion model or mixing model.
  6. Geological interpretation: source, pathway, reaction, crystallisation, alteration, cooling, deposition, metamorphism, disturbance or another event, with alternatives and uncertainty.

Skipping a layer creates familiar category errors. A spot date becomes “the age of the rock”; a whole-rock concentration becomes “the fluid composition”; a stable-isotope value becomes a unique temperature; an altered sample is plotted as if it were a melt; or a closed-system equation is applied to an open system. The remedy is not merely more decimal places. It is to restore the missing object, process and assumption.

Learning outcomes

After completing the book, you should be able to:

  • explain how ionic radius, charge, coordination and crystal structure influence substitution and element compatibility;
  • calculate simple partitioning, batch-melting, fractional-crystallisation, mixing and alteration mass balances;
  • distinguish concentration from activity and use pH, redox, complexation, saturation and reaction progress in an aqueous interpretation;
  • map fluid-rock alteration as mineral reactions and mass transfer rather than colour zones alone;
  • report stable-isotope data with explicit ratio, reference scale, delta notation and fractionation convention;
  • derive radioactive-decay and parent-daughter age equations and state the closed-system and initial-condition assumptions;
  • select among common geochronological systems according to mineral, half-life, geological question and disturbance behaviour;
  • explain closure temperature as a kinetic model rather than a fixed mineral label;
  • separate analytical dates, interpreted geological events and the different sources of age uncertainty; and
  • integrate field relations, petrography, mineral chemistry, whole-rock data, isotopes and ages into a testable event model.

Prerequisites and notation

You should be comfortable with atoms, ions, minerals, crystal sites, logarithms, exponentials, ratios, mass fractions, linear regression, uncertainty and the distinction between observation and inference. You should also understand relative-age relations, igneous and metamorphic processes, sedimentary reworking and fluid-rock reaction at an introductory level.

Symbols are defined locally because conventions vary. Concentration is commonly written C with the material and basis stated. A mineral-melt partition coefficient is written D_i^{min/melt}=C_i^{min}/C_i^{melt}. Activity is a_i=\gamma_i m_i or another declared standard-state expression. Isotope ratios are written explicitly, such as R={}^{87}\!Sr/{}^{86}\!Sr. Delta notation is \delta=(R_{sample}/R_{reference}-1)\times1000\ \text{per mil}. Radioactive decay uses N=N_0e^{-\lambda t} and t_{1/2}=\ln2/\lambda.

An age unit must not be inferred from context: use a, ka, Ma or Ga. Uncertainty statements must identify whether they are standard uncertainties, expanded uncertainties, confidence intervals, analytical-only terms or combinations that include calibration and decay-constant contributions. A correlation coefficient is not a goodness-of-fit verdict, and a small analytical uncertainty is not proof that a date represents the event of interest.

How to use the diagrams and examples

Each lesson has one purpose-built schematic. The diagrams are conceptual models, not measured sections, instrument layouts or universal geological sequences. Rebuild each diagram with a real or instructor-provided dataset only after identifying the measured object, units, normalisation, uncertainty and model boundary. Replace every conceptual arrow with an observation, calculation or explicitly labelled hypothesis.

Every numerical table in the worked examples is synthetic. Values are chosen to make the arithmetic inspectable and are not presented as a real deposit, laboratory run, water sample, standard or geological locality. The examples teach transformations and checks, not expected natural ranges. When applying the workflow to real data, retain raw observations and qualifiers, use the laboratory's validated reporting basis, and preserve the version of every external constant or reference value.

The same number may support more than one explanation. Learners should therefore write a result in three sentences: what was measured or calculated; what interpretation is supported under stated assumptions; and what alternative or unresolved test remains. This discipline is more important than choosing a fashionable plot.

Reproducibility and data records

A reusable dataset requires stable sample and analysis identifiers, immutable raw files, preparation history, method version, calibration and blank records, uncertainty components, reference-material results, qualifier codes and links to geological observations. Derived fields should preserve their equation, constants, units, software environment and input identifiers. Never overwrite a measured ratio with a corrected ratio or a concentration with a normalised value.

For ages, preserve the measured isotope ratios, common-component correction, disequilibrium treatment where relevant, decay constants, tracer or calibration version, covariance information, rejected analyses and rejection rule. For stable isotopes, preserve the measured ratio or signal, normalisation scale, reference materials, drift and memory corrections, and whether exchangeable oxygen, water or another component was removed. For solution chemistry, preserve filtration, preservation, temperature, redox measurement method, alkalinity procedure, major-ion charge balance and detection-limit semantics.

Reproducibility does not mean that every competent analyst must obtain identical last digits. It means that another reviewer can reconstruct the reported result, understand differences and determine whether they matter to the geological conclusion.

Assessment and completion standard

Completion requires more than recalling definitions. The learner should be able to review a supplied age-and-geochemistry dataset and produce an evidence ledger containing:

  • the geological object and sampling support for every analysis;
  • unit, basis, reference scale, quality-control status and uncertainty;
  • at least one mass-balance or reaction check;
  • a distinction between analytical date, interpreted event and regional correlation;
  • an explicit statement of open-system, inheritance, mixing or resetting risks;
  • a figure that displays individual analyses and uncertainty rather than only a preferred mean; and
  • a conclusion whose confidence is tied to evidence and whose decision boundary is stated.

A satisfactory submission remains valid if the preferred event interpretation changes. That is possible only when observations, calculations and assumptions are separated. A polished narrative that cannot be recalculated does not meet the standard.

Core sources