B2 · Publication Volume 7

Igneous Geochemical Classification and Series

major elements, alkaline and subalkaline series, and tectonic affinity

A one-way workflow from observation and declared classification to tested interpretation
A one-way workflow from observation and declared classification to tested interpretation

Learning objectives

After this lesson, you should be able to choose between modal and chemical classification routes, normalise modal data, check whole-rock analytical suitability, distinguish classification from petrogenesis and evaluate trace-element tectonic diagrams as conditional evidence.

Start with a field problem

A fine-grained altered rock is labelled “andesite” in a legacy log. A fresh sample from the same unit has 57 percent silica and total alkalis near the boundary between two TAS fields, while a coarse equivalent contains visible quartz and feldspars. Should all samples keep the inherited name, be reclassified chemically, be classified modally or remain at a broader level?

The answer depends on material and data quality. Classification must use the scheme designed for the available attributes. It must also preserve uncertainty near a boundary and avoid pretending that altered chemistry represents original magma.

Core process model

For suitable coarse crystalline igneous rocks, the IUGS QAPF approach uses modal proportions of quartz (Q), alkali feldspar (A), plagioclase (P) and feldspathoid (F), normalised within the appropriate QAP or FAP triangle. Quartz and feldspathoid are generally not plotted together. The modal method requires representative, measurable mineral proportions rather than a visual guess from a weathered face.

For many volcanic rocks whose modes cannot be determined because of glass or fine grain size, the total alkali–silica diagram uses an anhydrous-normalised whole-rock analysis. TAS is a chemical classification, not a direct modal equivalent. It is sensitive to mobility of sodium and potassium, analytical totals, alteration and the presence of accumulated crystals.

Major-element series such as alkaline versus subalkaline, or subdivisions within subalkaline rocks, are defined by stated diagrams and boundaries. A sample close to a line should retain coordinates, analytical uncertainty and boundary version. Do not force a precise name when the data do not support it.

Trace elements and isotopes are mainly interpretive. Normalised multi-element patterns can compare enrichment and depletion; ratios may reduce some dilution effects; isotopes can constrain source or contamination. Yet element mobility, residual mineralogy, crystal accumulation, age, normalisation values and closure all affect patterns. A “tectonic discrimination” field is a model trained on particular datasets, not a plate-setting detector.

Evidence and measurement

Start with field and petrographic description. Determine whether the sample is coherent, fragmental, intrusive, extrusive, cumulate, enclave or altered. For modal classification, use point counting or another reproducible estimate with enough counts and representative area. Report uncertainty and treatment of matrix, glass and accessory phases.

For chemical classification, inspect major-element totals, loss on ignition, duplicates, standards and alteration indicators. State whether iron is reported as FeO, Fe₂O₃ or total iron under another convention. Recalculate to an anhydrous basis only after retaining raw values. A closure sum near 100 percent is a check, not proof of freshness or representativeness.

Keep three output fields: descriptive class, petrogenetic hypothesis and tectonic interpretation. Each needs different evidence and confidence. This simple data design prevents a later user from treating “arc basalt” as though “arc” were an observed mineral.

Worked example

A coarse rock yields 100 quartz, 180 alkali-feldspar and 120 plagioclase points, with no feldspathoid. The Q+A+P total is 400, so the normalised values are


Q=25\%,\qquad A=45\%,\qquad P=30\%.

Quartz lies in the 20–60 percent granitoid interval, and


\frac{A}{A+P}=\frac{45}{75}=0.60.

On the declared IUGS QAP diagram this plots near the boundary between central granite subdivisions. Recounting uncertainty matters. If each feldspar proportion has several percentage points of uncertainty, retain the plotted values and boundary note rather than reporting a falsely exact root name.

Now suppose an altered fine-grained sample has 57.0 percent SiO₂, 4.0 percent Na₂O, 3.5 percent K₂O and 5 percent loss on ignition. Its total alkalis are 7.5 percent, but sodium and potassium may have moved. A TAS name can be recorded as provisional only if alteration screening supports it. Otherwise use a broader descriptive volcanic-rock name and seek a fresher sample.

Misinterpretations and uncertainty

QAPF and TAS are not interchangeable calculators. A normative mineral calculation is not a measured mode. A whole-rock composition from a cumulate, enclave or mixed pyroclastic sample may not represent liquid. Classification boundaries can change among editions, and local legacy terms may not map one-to-one to an international root name.

Trace-element patterns are compositional data with shared denominators and correlated uncertainty. A field on a discrimination plot can be occupied for several reasons. Tectonic setting should also be tested against stratigraphy, age, structure, regional geology, isotopes and competing petrogenetic models.

Practical investigation

Classify five synthetic samples: a coarse quartz–feldspar rock with point counts, a glassy volcanic rock with fresh chemistry, an altered volcanic rock, a crystal-rich cumulate and a mixed tuff. For each, choose a classification route, state why the other route is inappropriate, report the most defensible name and list the data needed to improve confidence.

Then build a one-page interpretation matrix. Place modal class, TAS class, major-element series, trace-element pattern, isotope evidence and regional context in separate rows. Mark which are observations, calculations or interpretations.

Mastery check

  1. When is QAPF preferable to TAS?
  2. Why must Q, A and P be normalised?
  3. How can loss on ignition and alkali mobility affect a TAS name?
  4. Why is a normative mineral proportion not a modal measurement?
  5. What evidence is needed before a tectonic discrimination field becomes persuasive?

Sources and further reading